Healthcare analytics, AI solutions for biological big data, providing an AI platform for the biotech, life sciences, medical and pharmaceutical industries, as well as for related technological approaches, i.e., curation and text analysis with machine learning and other activities related to AI applications to these industries.
Can Blockchain Technology and Artificial Intelligence Cure What Ails Biomedical Research and Healthcare, Volume 2 (Volume Two: Latest in Genomics Methodologies for Therapeutics: Gene Editing, NGS and BioInformatics, Simulations and the Genome Ontology), Part 1: Next Generation Sequencing (NGS)
Can Blockchain Technology and Artificial Intelligence Cure What Ails Biomedical Research and Healthcare
Curator: Stephen J. Williams, Ph.D.
Updated 12/18/2018
In the efforts to reduce healthcare costs, provide increased accessibility of service for patients, and drive biomedical innovations, many healthcare and biotechnology professionals have looked to advances in digital technology to determine the utility of IT to drive and extract greater value from healthcare industry. Two areas of recent interest have focused how best to use blockchain and artificial intelligence technologies to drive greater efficiencies in our healthcare and biotechnology industries.
More importantly, with the substantial increase in ‘omic data generated both in research as well as in the clinical setting, it has become imperative to develop ways to securely store and disseminate the massive amounts of ‘omic data to various relevant parties (researchers or clinicians), in an efficient manner yet to protect personal privacy and adhere to international regulations. This is where blockchain technologies may play an important role.
A recent Oncotarget paper by Mamoshina et al. (1) discussed the possibility that next-generation artificial intelligence and blockchain technologies could synergize to accelerate biomedical research and enable patients new tools to control and profit from their personal healthcare data, and assist patients with their healthcare monitoring needs. According to the abstract:
The authors introduce new concepts to appraise and evaluate personal records, including the combination-, time- and relationship value of the data. They also present a roadmap for a blockchain-enabled decentralized personal health data ecosystem to enable novel approaches for drug discovery, biomarker development, and preventative healthcare. In this system, blockchain and deep learning technologies would provide the secure and transparent distribution of personal data in a healthcare marketplace, and would also be useful to resolve challenges faced by the regulators and return control over personal data including medical records to the individual.
The review discusses:
Recent achievements in next-generation artificial intelligence
Basic concepts of highly distributed storage systems (HDSS) as a preferred method for medical data storage
Open source blockchain Exonium and its application for healthcare marketplace
A blockchain-based platform allowing patients to have control of their data and manage access
How advances in deep learning can improve data quality, especially in an era of big data
Advances in Artificial Intelligence
Integrative analysis of the vast amount of health-associated data from a multitude of large scale global projects has proven to be highly problematic (REF 27), as high quality biomedical data is highly complex and of a heterogeneous nature, which necessitates special preprocessing and analysis.
Increased computing processing power and algorithm advances have led to significant advances in machine learning, especially machine learning involving Deep Neural Networks (DNNs), which are able to capture high-level dependencies in healthcare data. Some examples of the uses of DNNs are:
Prediction of drug properties(2, 3) and toxicities(4)
Generative Adversarial Networks (https://arxiv.org/abs/1406.2661): requires good datasets for extensive training but has been used to determine tumor growth inhibition capabilities of various molecules (7)
Recurrent neural Networks (RNN): Originally made for sequence analysis, RNN has proved useful in analyzing text and time-series data, and thus would be very useful for electronic record analysis. Has also been useful in predicting blood glucose levels of Type I diabetic patients using data obtained from continuous glucose monitoring devices (8)
Transfer Learning: focused on translating information learned on one domain or larger dataset to another, smaller domain. Meant to reduce the dependence on large training datasets that RNN, GAN, and DNN require. Biomedical imaging datasets are an example of use of transfer learning.
One and Zero-Shot Learning: retains ability to work with restricted datasets like transfer learning. One shot learning aimed to recognize new data points based on a few examples from the training set while zero-shot learning aims to recognize new object without seeing the examples of those instances within the training set.
Highly Distributed Storage Systems (HDSS)
The explosion in data generation has necessitated the development of better systems for data storage and handling. HDSS systems need to be reliable, accessible, scalable, and affordable. This involves storing data in different nodes and the data stored in these nodes are replicated which makes access rapid. However data consistency and affordability are big challenges.
Blockchain is a distributed database used to maintain a growing list of records, in which records are divided into blocks, locked together by a crytosecurity algorithm(s) to maintain consistency of data. Each record in the block contains a timestamp and a link to the previous block in the chain. Blockchain is a distributed ledger of blocks meaning it is owned and shared and accessible to everyone. This allows a verifiable, secure, and consistent history of a record of events.
Data Privacy and Regulatory Issues
The establishment of the Health Insurance Portability and Accountability Act (HIPAA) in 1996 has provided much needed regulatory guidance and framework for clinicians and all concerned parties within the healthcare and health data chain. The HIPAA act has already provided much needed guidance for the latest technologies impacting healthcare, most notably the use of social media and mobile communications (discussed in this article Can Mobile Health Apps Improve Oral-Chemotherapy Adherence? The Benefit of Gamification.). The advent of blockchain technology in healthcare offers its own unique challenges however HIPAA offers a basis for developing a regulatory framework in this regard. The special standards regarding electronic data transfer are explained in HIPAA’s Privacy Rule, which regulates how certain entities (covered entities) use and disclose individual identifiable health information (Protected Health Information PHI), and protects the transfer of such information over any medium or electronic data format. However, some of the benefits of blockchain which may revolutionize the healthcare system may be in direct contradiction with HIPAA rules as outlined below:
Issues of Privacy Specific In Use of Blockchain to Distribute Health Data
Blockchain was designed as a distributed database, maintained by multiple independent parties, and decentralized
Linkage timestamping; although useful in time dependent data, proof that third parties have not been in the process would have to be established including accountability measures
Blockchain uses a consensus algorithm even though end users may have their own privacy key
Applied cryptography measures and routines are used to decentralize authentication (publicly available)
Blockchain users are divided into three main categories: 1) maintainers of blockchain infrastructure, 2) external auditors who store a replica of the blockchain 3) end users or clients and may have access to a relatively small portion of a blockchain but their software may use cryptographic proofs to verify authenticity of data.
YouTube video on How #Blockchain Will Transform Healthcare in 25 Years (please click below)
In Big Data for Better Outcomes, BigData@Heart, DO->IT, EHDN, the EU data Consortia, and yes, even concepts like pay for performance, Richard Bergström has had a hand in their creation. The former Director General of EFPIA, and now the head of health both at SICPA and their joint venture blockchain company Guardtime, Richard is always ahead of the curve. In fact, he’s usually the one who makes the curve in the first place.
Please click on the following link for a podcast on Big Data, Blockchain and Pharma/Healthcare by Richard Bergström:
Mamoshina, P., Ojomoko, L., Yanovich, Y., Ostrovski, A., Botezatu, A., Prikhodko, P., Izumchenko, E., Aliper, A., Romantsov, K., Zhebrak, A., Ogu, I. O., and Zhavoronkov, A. (2018) Converging blockchain and next-generation artificial intelligence technologies to decentralize and accelerate biomedical research and healthcare, Oncotarget9, 5665-5690.
Aliper, A., Plis, S., Artemov, A., Ulloa, A., Mamoshina, P., and Zhavoronkov, A. (2016) Deep Learning Applications for Predicting Pharmacological Properties of Drugs and Drug Repurposing Using Transcriptomic Data, Molecular pharmaceutics13, 2524-2530.
Wen, M., Zhang, Z., Niu, S., Sha, H., Yang, R., Yun, Y., and Lu, H. (2017) Deep-Learning-Based Drug-Target Interaction Prediction, Journal of proteome research16, 1401-1409.
Gao, M., Igata, H., Takeuchi, A., Sato, K., and Ikegaya, Y. (2017) Machine learning-based prediction of adverse drug effects: An example of seizure-inducing compounds, Journal of pharmacological sciences133, 70-78.
Putin, E., Mamoshina, P., Aliper, A., Korzinkin, M., Moskalev, A., Kolosov, A., Ostrovskiy, A., Cantor, C., Vijg, J., and Zhavoronkov, A. (2016) Deep biomarkers of human aging: Application of deep neural networks to biomarker development, Aging8, 1021-1033.
Vandenberghe, M. E., Scott, M. L., Scorer, P. W., Soderberg, M., Balcerzak, D., and Barker, C. (2017) Relevance of deep learning to facilitate the diagnosis of HER2 status in breast cancer, Scientific reports7, 45938.
Kadurin, A., Nikolenko, S., Khrabrov, K., Aliper, A., and Zhavoronkov, A. (2017) druGAN: An Advanced Generative Adversarial Autoencoder Model for de Novo Generation of New Molecules with Desired Molecular Properties in Silico, Molecular pharmaceutics14, 3098-3104.
Ordonez, F. J., and Roggen, D. (2016) Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition, Sensors (Basel)16.
December 5, 2018 | The boom of blockchain and distributed ledger technologies have inspired healthcare organizations to test the capabilities of their data. Quest Diagnostics, in partnership with Humana, MultiPlan, and UnitedHealth Group’s Optum and UnitedHealthcare, have launched a pilot program that applies blockchain technology to improve data quality and reduce administrative costs associated with changes to healthcare provider demographic data.
The collective body, called Synaptic Health Alliance, explores how blockchain can keep only the most current healthcare provider information available in health plan provider directories. The alliance plans to share their progress in the first half of 2019.
Providing consumers looking for care with accurate information when they need it is essential to a high-functioning overall healthcare system, Jason O’Meara, Senior Director of Architecture at Quest Diagnostics, told Clinical Informatics News in an email interview.
“We were intentional about calling ourselves an alliance as it speaks to the shared interest in improving health care through better, collaborative use of an innovative technology,” O’Meara wrote. “Our large collective dataset and national footprints enable us to prove the value of data sharing across company lines, which has been limited in healthcare to date.”
O’Meara said Quest Diagnostics has been investing time and resources the past year or two in understanding blockchain, its ability to drive purpose within the healthcare industry, and how to leverage it for business value.
“Many health care and life science organizations have cast an eye toward blockchain’s potential to inform their digital strategies,” O’Meara said. “We recognize it takes time to learn how to leverage a new technology. We started exploring the technology in early 2017, but we quickly recognized the technology’s value is in its application to business to business use cases: to help transparently share information, automate mutually-beneficial processes and audit interactions.”
Quest began discussing the potential for an alliance with the four other companies a year ago, O’Meara said. Each company shared traits that would allow them to prove the value of data sharing across company lines.
“While we have different perspectives, each member has deep expertise in healthcare technology, a collaborative culture, and desire to continuously improve the patient/customer experience,” said O’Meara. “We also recognize the value of technology in driving efficiencies and quality.”
Following its initial launch in April, Synaptic Health Alliance is deploying a multi-company, multi-site, permissioned blockchain. According to a whitepaper published by Synaptic Health, the choice to use a permissioned blockchain rather than an anonymous one is crucial to the alliance’s success.
“This is a more effective approach, consistent with enterprise blockchains,” an alliance representative wrote. “Each Alliance member has the flexibility to deploy its nodes based on its enterprise requirements. Some members have elected to deploy their nodes within their own data centers, while others are using secured public cloud services such as AWS and Azure. This level of flexibility is key to growing the Alliance blockchain network.”
As the pilot moves forward, O’Meara says the Alliance plans to open ability to other organizations. Earlier this week Aetna and Ascension announced they joined the project.
“I am personally excited by the amount of cross-company collaboration facilitated by this project,” O’Meara says. “We have already learned so much from each other and are using that knowledge to really move the needle on improving healthcare.”
November 29, 2018 | The US Department of Health and Human Services (HHS) is making waves in the blockchain space. The agency’s Division of Acquisition (DA) has developed a new system, called Accelerate, which gives acquisition teams detailed information on pricing, terms, and conditions across HHS in real-time. The department’s Associate Deputy Assistant Secretary for Acquisition, Jose Arrieta, gave a presentation and live demo of the blockchain-enabled system at the Distributed: Health event earlier this month in Nashville, Tennessee.
Accelerate is still in the prototype phase, Arrieta said, with hopes that the new system will be deployed at the end of the fiscal year.
HHS spends around $25 billion a year in contracts, Arrieta said. That’s 100,000 contracts a year with over one million pages of unstructured data managed through 45 different systems. Arrieta and his team wanted to modernize the system.
“But if you’re going to change the way a workforce of 20,000 people do business, you have to think your way through how you’re going to do that,” said Arrieta. “We didn’t disrupt the existing systems: we cannibalized them.”
The cannibalization process resulted in Accelerate. According to Arrieta, the system functions by creating a record of data rather than storing it, leveraging machine learning, artificial intelligence (AI), and robotic process automation (RPA), all through blockchain data.
“We’re using that data record as a mechanism to redesign the way we deliver services through micro-services strategies,” Arrieta said. “Why is that important? Because if you have a single application or data use that interfaces with 55 other applications in your business network, it becomes very expensive to make changes to one of the 55 applications.”
Accelerate distributes the data to the workforce, making it available to them one business process at a time.
“We’re building those business processes without disrupting the existing systems,” said Arrieta, and that’s key. “We’re not shutting off those systems. We’re using human-centered design sessions to rebuild value exchange off of that data.”
The first application for the system, Arrieta said, can be compared to department stores price-matching their online competitors.
It takes the HHS close to a month to collect the amalgamation of data from existing system, whether that be terms and conditions that drive certain price points, or software licenses.
“The micro-service we built actually analyzes that data, and provides that information to you within one second,” said Arrieta. “This is distributed to the workforce, to the 5,000 people that do the contracting, to the 15,000 people that actually run the programs at [HHS].”
This simple micro-service is replicated on every node related to HHS’s internal workforce. If somebody wants to change the algorithm to fit their needs, they can do that in a distributed manner.
Arrieta hopes to use Accelerate to save researchers money at the point of purchase. The program uses blockchain to simplify the process of acquisition.
“How many of you work with the federal government?” Arrieta asked the audience. “Do you get sick of reentering the same information over and over again? Every single business opportunity you apply for, you have to resubmit your financial information. You constantly have to check for validation and verification, constantly have to resubmit capabilities.”
Wouldn’t it be better to have historical notes available for each transaction? said Arrieta. This would allow clinical researchers to be able to focus on “the things they’re really good at,” instead of red tape.
“If we had the top cancer researcher in the world, would you really want her spending her time learning about federal regulations as to how to spend money, or do you want her trying to solve cancer?” Arrieta said. “What we’re doing is providing that data to the individual in a distributed manner so they can read the information of historical purchases that support activity, and they can focus on the objectives and risks they see as it relates to their programming and their objectives.”
Blockchain also creates transparency among researchers, Arrieta said, which says creates an “uncomfortable reality” in the fact that they have to make a decision regarding data, fundamentally changing value exchange.
“The beauty of our business model is internal investment,” Arrieta said. For instance, the HHS could take all the sepsis data that exists in their system, put it into a distributed ledger, and share it with an external source.
“Maybe that could fuel partnership,” Arrieta said. “I can make data available to researchers in the field in real-time so they can actually test their hypothesis, test their intuition, and test their imagination as it relates to solving real-world problems.”
Blockchain-based genomic data hub platform Shivom recently reached its $35 million hard cap within 15 seconds of opening its main token sale. Shivom received funding from a number of crypto VC funds, including Collinstar, Lateral, and Ironside.
The goal is to create the world’s largest store of genomic data while offering an open web marketplace for patients, data donors, and providers — such as pharmaceutical companies, research organizations, governments, patient-support groups, and insurance companies.
“Disrupting the whole of the health care system as we know it has to be the most exciting use of such large DNA datasets,” Shivom CEO Henry Ines told me. “We’ll be able to stratify patients for better clinical trials, which will help to advance research in precision medicine. This means we will have the ability to make a specific drug for a specific patient based on their DNA markers. And what with the cost of DNA sequencing getting cheaper by the minute, we’ll also be able to sequence individuals sooner, so young children or even newborn babies could be sequenced from birth and treated right away.”
While there are many solutions examining DNA data to explain heritage, intellectual capabilities, health, and fitness, the potential of genomic data has largely yet to be unlocked. A few companies hold the monopoly on genomic data and make sizeable profits from selling it to third parties, usually without sharing the earnings with the data donor. Donors are also not informed if and when their information is shared, nor do they have any guarantee that their data is secure from hackers.
Shivom wants to change that by creating a decentralized platform that will break these monopolies, democratizing the processes of sharing and utilizing the data.
“Overall, large DNA datasets will have the potential to aid in the understanding, prevention, diagnosis, and treatment of every disease known to mankind, and could create a future where no diseases exist, or those that do can be cured very easily and quickly,” Ines said. “Imagine that, a world where people do not get sick or are already aware of what future diseases they could fall prey to and so can easily prevent them.”
Shivom’s use of blockchain technology and smart contracts ensures that all genomic data shared on the platform will remain anonymous and secure, while its OmiX token incentivizes users to share their data for monetary gain.
Blockchain will secure the DNA database for 50 million citizens in the eighth-largest state in India. The government of Andhra Pradesh signed a Memorandum of Understanding with a German genomics and precision medicine start-up, Shivom, which announced to start the pilot project soon. The move falls in line with a trend for governments turning to population genomics, and at the same time securing the sensitive data through blockchain.
Andhra Pradesh, DNA, and blockchain
Storing sensitive genetic information safely and securely is a big challenge. Shivom builds a genomic data-hub powered by blockchain technology. It aims to connect researchers with DNA data donors thus facilitating medical research and the healthcare industry.
With regards to Andhra Pradesh, the start-up will first launch a trial to determine the viability of their technology for moving from a proactive to a preventive approach in medicine, and towards precision health. “Our partnership with Shivom explores the possibilities of providing an efficient way of diagnostic services to patients of Andhra Pradesh by maintaining the privacy of the individual data through blockchain technologies,” said J A Chowdary, IT Advisor to Chief Minister, Government of Andhra Pradesh.
Other Articles in this Open Access Journal on Digital Health include:
Live Coverage: MedCity Converge 2018 Philadelphia: AI in Cancer and Keynote Address
Reporter: Stephen J. Williams, PhD
3.3.4 Live Coverage: MedCity Converge 2018 Philadelphia: AI in Cancer and Keynote Address, Volume 2 (Volume Two: Latest in Genomics Methodologies for Therapeutics: Gene Editing, NGS and BioInformatics, Simulations and the Genome Ontology), Part 2: CRISPR for Gene Editing and DNA Repair
8:30 AM -9:15
Practical Applications of AI in Cancer
We are far from machine learning dictating clinical decision making, but AI has important niche applications in oncology. Hear from a panel of innovative startups and established life science players about how machine learning and AI can transform different aspects in healthcare, be it in patient recruitment, data analysis, drug discovery or care delivery.
Ayan: working at IBM and Thompon Rueters with structured datasets and having gone through his own cancer battle, he is now working in healthcare AI which has an unstructured dataset(s)
Carla: collecting medical images over the world, mainly tumor and calculating tumor volumetrics
Tufia: drug resistant breast cancer clinician but interested in AI and healthcareIT at Mayo
John: taking large scale datasets but a machine learning skeptic
moderator: how has imaging evolved?
Carla: ten times images but not ten times radiologists so stressed field needs help with image analysis; they have seen measuring lung tumor volumetrics as a therapeutic diagnostic has worked
moderator: how has AI affected patient recruitment?
Tufia: majority of patients are receiving great care but AI can offer profiles and determine which patients can benefit from tertiary care;
John: 1980 paper on no free lunch theorem; great enthusiasm about optimization algortihisms fell short in application; can extract great information from e.g. images
moderator: how is AI for healthcare delivery working at mayo?
Tufia: for every hour with patient two hours of data mining. for care delivery hope to use the systems to leverage the cognitive systems to do the data mining
John: problem with irreproducible research which makes a poor dataset: also these care packages are based on population data not personalized datasets; challenges to AI is moving correlation to causation
Carla: algorithisms from on healthcare network is not good enough, Google tried and it failed
John: curation very important; good annotation is needed; needed to go in and develop, with curators, a systematic way to curate medial records; need standardization and reproducibility; applications in radiometrics can be different based on different data collection machines; developed a machine learning model site where investigators can compare models on a hub; also need to communicate with patients on healthcare information and quality information
Ayan: Australia and Canada has done the most concerning AI and lifescience, healthcare space; AI in most cases is cognitive learning: really two types of companies 1) the Microsofts, Googles, and 2) the startups that may be more pure AI
Final Notes: We are at a point where collecting massive amounts of healthcare related data is simple, rapid, and shareable. However challenges exist in quality of datasets, proper curation and annotation, need for collaboration across all healthcare stakeholders including patients, and dissemination of useful and accurate information
9:15 AM–9:45 AM
Opening Keynote: Dr. Joshua Brody, Medical Oncologist, Mount Sinai Health System
The Promise and Hype of Immunotherapy
Immunotherapy is revolutionizing oncology care across various types of cancers, but it is also necessary to sort the hype from the reality. In his keynote, Dr. Brody will delve into the history of this new therapy mode and how it has transformed the treatment of lymphoma and other diseases. He will address the hype surrounding it, why so many still don’t respond to the treatment regimen and chart the way forward—one that can lead to more elegant immunotherapy combination paths and better outcomes for patients.
hodgkin’s lymphoma best responder to PD1 therapy (nivolumab) but hepatic adverse effects
CAR-T (chimeric BCR and TCR); a long process which includes apheresis, selection CD3/CD28 cells; viral transfection of the chimeric; purification
complete remissions of B cell lymphomas (NCI trial) and long term remissions past 18 months
side effects like cytokine release (has been controlled); encephalopathy (he uses a hand writing test to see progression of adverse effect)
Vaccines
teaching the immune cells as PD1 inhibition exhausting T cells so a vaccine boost could be an adjuvant to PD1 or checkpoint therapy
using Flt3L primed in-situ vaccine (using a Toll like receptor agonist can recruit the dendritic cells to the tumor and then activation of T cell response); therefore vaccine does not need to be produced ex vivo; months after the vaccine the tumor still in remission
versus rituximab, which can target many healthy B cells this in-situ vaccine strategy is very specific for the tumorigenic B cells
HoWEVER they did see resistant tumor cells which did not overexpress PD-L1 but they did discover a novel checkpoint (cannot be disclosed at this point)
Please follow on Twitter using the following #hashtags and @pharma_BI
#MCConverge
#AI
#cancertreatment
#immunotherapy
#healthIT
#innovation
#precisionmedicine
#healthcaremodels
#personalizedmedicine
#healthcaredata
And at the following handles:
@pharma_BI
@medcitynews
Please see related articles on Live Coverage of Previous Meetings on this Open Access Journal
Vyasa Analytics Demos Deep Learning Software for Life Sciences at Bio-IT World 2018 – Vyasa’s booth (#632), Volume 2 (Volume Two: Latest in Genomics Methodologies for Therapeutics: Gene Editing, NGS and BioInformatics, Simulations and the Genome Ontology), Part 1: Next Generation Sequencing (NGS)
Vyasa Analytics Demos Deep Learning Software for Life Sciences at Bio-IT World 2018 – Vyasa’s booth (#632)
Reporter: Aviva Lev-Ari, PhD, RN
BOSTON – May 10, 2018 – Vyasa Analytics, a provider of deep learning software and analytics for life sciences and healthcare organizations, today announces three pre-built deep learning analytics modules for its Cortex software at Bio-IT World Conference & Expo. Cortex enables the secure, scalable application of deep learning-based artificial intelligence (AI) analytics to enterprise data, identifying patterns, relationships and concepts across disparate data sources.
The new Neural Concept Recognition, Image Analytics and ChemVector analytics modules in Cortex enable life sciences organizations to quickly and easily apply deep learning analytics to large data streams of text, images and chemical structures. Like all deep learning analytical modules in Cortex’s library, these new modules allow users to ask complex questions of their data and use the answers to gain critical insights.
“Life sciences and healthcare organizations are using AI tools to advance research and development and deliver better patient care. Deep learning algorithms provide a set of powerful approaches that help us apply analytics more effectively and comprehensively across large scale data sources,” said Dr. Christopher Bouton, founder and CEO of Vyasa. “The idea of AI has been around for decades, but we are now experiencing a perfect storm of GPU-based computing power, deep learning algorithm advances and highly scalable data sources that enables paradigm-shifting machine learning and analytics capabilities.”
Vyasa will be demoing three deep learning analytics modules for Cortex at Bio-IT World 2018 in Boston from May 15 to 17, including:
Neural Concept Recognition. This module can be trained on text concepts (e.g. drugs, diseases, pathways, conditions, side effects, genes) in structured and unstructured data. Users can ask Cortex complex questions across large scale data sets, and discover unexpected relationships between concept types. Concept recognition analytics is applicable to a wide range of use cases from competitive intelligence, to drug repurposing and EHR analytics.
Life Sciences R&D Specialized Image Analytics. Deep learning enables novel, powerful forms of image analytics, capable of being trained to detect patterns and objects in large scale image data sources.With just a few clicks in Cortex, the user can connect large streams of image data and apply analytics to those sources. Vyasa has finely-tuned this analysis for life sciences images, and it is ideal for cell assay screening, drug manufacturing and post-market screening for counterfeit packaging and tablets.
ChemVector de novo Compound Design. This proprietary Cortex module applies deep learning to chemical structures. Users can drag and drop one or more SDF files containing SMILES strings into Cortex, and Cortex can identify and generate novel compounds that optimize critical variables such as log-p, molecular weight and synthetic viability. ChemVector can be used with a range of other chemistry-specific analytical modules also available in Cortex.
Dr. Bouton, Vyasa’s founder and CEO, received his BA in Neuroscience (Magna Cum Laude) from Amherst College in 1996 and his Ph.D. in Molecular Neurobiology from Johns Hopkins University in 2001. Previously Dr. Bouton was the CEO of Entagen a software company founded in 2008 that provided innovative Big Data products including Extera and TripleMap. Entagen’s technologies won numerous awards including the “Innovative Technology of the Year Award for Big Data” from the Massachusetts Technology Leadership Council in 2012 and Entagen was recognized as a Gartner “Cool Vendor” in the Life Sciences in 2013. Entagen was acquired by Thomson Reuters in 2013. Dr. Bouton is an author on over a dozen scientific papers and book chapters and his work has been covered in a number of industry news articles.
Visit Vyasa and demo Cortex at booth #632, and watch the explainer video at www.vyasa.com.
About Vyasa Analytics
Vyasa Analytics provides deep learning software and analytics for life sciences and healthcare organizations. Cortex is Vyasa’s secure, highly scalable software platform for collaborative knowledge discovery and data analytics. Using Vyasa’s proprietary Neural Concept Recognition technology, Cortex identifies trends and patterns across disparate data sources, empowering project teams to gain insights and drive better decision making. Learn more at www.vyasa.com.
Linguamatics announces the official launch of its AI self-service text-mining solution for researchers.
Reporter: Aviva Lev-Ari, PhD, RN
3.4.9 Linguamatics announces the official launch of its AI self-service text-mining solution for researchers, Volume 2 (Volume Two: Latest in Genomics Methodologies for Therapeutics: Gene Editing, NGS and BioInformatics, Simulations and the Genome Ontology), Part 2: CRISPR for Gene Editing and DNA Repair
iScite provides end-users with direct access to powerful AI-driven insights from text
Boston — May 9, 2018 — Linguamatics, the leading NLP-based text analytics provider for biomedical applications, today announced the launch of Linguamatics iScite, a breakthrough innovation in scientific search that puts the precision and power of Linguamatics artificial intelligence (AI) technology directly into the hands of scientists, researchers and other knowledge workers. iScite offers a modern, easy-to-use scientific search engine that provides intuitive access to AI-powered searches across key biomedical data sources and delivers insightful answers to search questions.
iScite is designed as a next-generation search experience that empowers non-technical users to conduct their own NLP-based scientific searches to extract data insights. Rather than rely on time- and/or resource-crunched technical experts to create and perform searches, iScite enables users to quickly and independently find precise answers to their high-value questions.
“Traditional search methods are often time-consuming, expensive and ineffective, and the results are imprecise and difficult to sift through,” said Jane Reed, head of life science strategy for Linguamatics. “With iScite, users can take advantage of the power of NLP without the traditional complexities. Our patent-pending Answer-Routing Engine interprets users’ search terms and guides them to the best possible answers to their questions. Searches are seamless across multiple content sources, and users are quickly pointed to the exact content relevant to their search without having to laboriously read through every word of the source documents.”
iScite uses Linguamatics’ award-winning technology stack to handle the nuances of language and the variety of ways people express the same information, ensuring searches are comprehensive and accurate. Using advanced NLP relationship and pattern matching, iScite rapidly guides users directly to the relevant insights extracted from cloud-hosted scientific content. Results are presented in a structured, semantically-meaningful way, with options for dynamic filtering and faceting, and multiple collaboration features to allow easier sharing of insights with co-workers and key stakeholders. Behind the scenes Linguamatics uses a powerful blend of NLP and machine learning-based methods to achieve the best precision and recall.
“By empowering end-user scientists and clinicians with an easy-to-use search engine, we are speeding their access to the right knowledge for decision-making to advance the discovery, development and delivery of therapeutics,” said Linguamatics Executive Chairman John Brimacombe. “iScite has the potential to revolutionize the search process for the biomedical industry by providing everyone with rapid access to the knowledge they need, while freeing data scientists and informaticians to focus on the most challenging, in-depth search projects. iScite is a breakthrough in scientific research, filling an industry demand for a self-service alternative that delivers deep insights in a single search.”
Linguamatics will demonstrate iScite at Bio-IT World 2018 in Boston May 15-17. Visit us at booth #549, or go to our website, http://www.linguamatics.com/iscite, for more information.
About Linguamatics Linguamaticstransforms unstructured big data into big insights to advance human health and wellbeing. A world leader in deploying innovative text analytics for high-value knowledge discovery and decision support, Linguamatics’ solutions are used by top commercial, academic and government organizations, including 18 of the top 20 global pharmaceutical companies, the US Food and Drug Administration (FDA) and leading US healthcare organizations.
Linguamatics Media contact:
Michelle Ronan Noteboom, Sr. Account Director
Amendola Communications
+ 1 512.426.2870 mnoteboom@acmarketingpr.com
Synopsis Days 1,2,3: 2018 Annual World Medical Innovation Forum Artificial Intelligence April 23–25, 2018 Boston, Massachusetts | Westin Copley Place
Curator: Aviva Lev-Ari, PhD, RN
3.1.6 Synopsis Days 1,2,3: 2018 Annual World Medical Innovation Forum Artificial Intelligence April 23–25, 2018 Boston, Massachusetts | Westin Copley Place, Volume 2 (Volume Two: Latest in Genomics Methodologies for Therapeutics: Gene Editing, NGS and BioInformatics, Simulations and the Genome Ontology), Part 2: CRISPR for Gene Editing and DNA Repair
Synopsis Day 1: 2018 Annual World Medical Innovation Forum Artificial Intelligence April 23–25, 2018 Boston, Massachusetts | Westin Copley Place
Chest pain patients are often evaluated for acute myocardial infarction through troponin testing, which may prompt downstream services (cascades) of uncertain value.
Objective
Determine the association of high-sensitivity cardiac troponin (hs-cTn) assay implementation with cascade events.
Methods
Using electronic health record and billing data, we examined patient-visits to five emergency departments, April 1, 2017 – April 1, 2019. Difference-in-differences analysis compared patient-visits for chest pain (n=7,564) to patient-visits for other symptoms (n=100,415) (irrespective of troponin testing) before and after hs-cTn assay implementation. Outcomes included presence of any cascade event potentially associated with an initial hs-cTn test (primary), individual cascade events, length of stay, and spending on cardiac services.
Results
Following hs-cTn implementation, patients with chest pain had a 2.8% (95%CI 0.72, 4.9) net increase in experiencing any cascade event. They were more likely to have multiple troponin tests (10.5%, 95%CI 9.0, 12.0) and electrocardiograms (7.1 per 100 patient-visits, 95%CI 1.8, 12.4). However, they received net fewer computed tomography scans (-1.5 per 100 patient-visits, 95%CI -1.8, -1.1), stress tests (-5.9 per 100 patient-visits, 95%CI -6.5, -5.3), and cardiac catheterizations (-0.65 per 100 patient-visits, 95%CI -1.01, -0.30) and were less likely to receive cardiac medications, undergo cardiology evaluation (-3.5%, 95%CI -4.5, 2.6), or be hospitalized (-5.8%, 95%CI -7.7, -3.8). Chest pain patients had lower net mean length of stay (-0.24 days, 95%CI -0.32, -0.16) but no net change in spending.
Conclusions
Hs-cTn assay implementation was associated with more net upfront tests yet fewer net stress tests, catheterizations, cardiology evaluations, and hospital admissions in chest pain patients relative to patients with other symptoms.
Background: We assessed whether plasma troponin I measured by a high-sensitivity assay (hs-TnI) is associated with incident cardiovascular disease (CVD) and mortality in a community-based sample without prior CVD.
Methods: ARIC study (Atherosclerosis Risk in Communities) participants aged 54 to 74 years without baseline CVD were included in this study (n=8121). Cox proportional hazards models were constructed to determine associations between hs-TnI and incident coronary heart disease (CHD; myocardial infarction and fatal CHD), ischemic stroke, atherosclerotic CVD (CHD and stroke), heart failure hospitalization, global CVD (atherosclerotic CVD and heart failure), and all-cause mortality. The comparative association of hs-TnI and high-sensitivity troponin T with incident CVD events was also evaluated. Risk prediction models were constructed to assess prediction improvement when hs-TnI was added to traditional risk factors used in the Pooled Cohort Equation.
Results: The median follow-up period was ≈15 years. Detectable hs-TnI levels were observed in 85% of the study population. In adjusted models, in comparison to low hs-TnI (lowest quintile, hs-TnI ≤1.3 ng/L), elevated hs-TnI (highest quintile, hs-TnI ≥3.8 ng/L) was associated with greater incident CHD (hazard ratio [HR], 2.20; 95% CI, 1.64-2.95), ischemic stroke (HR, 2.99; 95% CI, 2.01-4.46), atherosclerotic CVD (HR, 2.36; 95% CI, 1.86-3.00), heart failure hospitalization (HR, 4.20; 95% CI, 3.28-5.37), global CVD (HR, 3.01; 95% CI, 2.50-3.63), and all-cause mortality (HR, 1.83; 95% CI, 1.56-2.14). hs-TnI was observed to have a stronger association with incident global CVD events in white than in black individuals and a stronger association with incident CHD in women than in men. hs-TnI and high-sensitivity troponin T were only modestly correlated (r=0.47) and were complementary in prediction of incident CVD events, with elevation of both troponins conferring the highest risk in comparison with elevation in either one alone. The addition of hsTnI to the Pooled Cohort Equation model improved risk prediction for atherosclerotic CVD, heart failure, and global CVD.
Conclusions: Elevated hs-TnI is strongly associated with increased global CVD incidence in the general population independent of traditional risk factors. hs-TnI and high-sensitivity troponin T provide complementary rather than redundant information.
Siemens Launches High-sensitivity Troponin Test for Faster Diagnosis of Heart Attacks
The new troponin I assays can detect lower levels of troponin compared to conventional testing
July 25, 2018 — The U.S. Food and Drug Administration (FDA) cleared Siemens Healthineers high-sensitivity troponin I assays (TnIH) for the Atellica IM and ADVIA Centaur XP/XPT in vitro diagnostic analyzers from Siemens Healthineers to aid in the early diagnosis of myocardial infarctions.
The new tests can shorten the time doctors need to diagnose a life-threatening heart attacks. The time to first results is 10 minutes. When a patient experiencing chest pain enters the emergency department, a physician orders a blood test to determine whether troponin is present. As blood flow to the heart is blocked, the heart muscle begins to die in as few as 30 to 60 minutes and releases troponin into the bloodstream.
The company said its high-sensitivity performance of the two new Siemens TnIH assays offers the ability to detect lower levels of troponin at significantly improved precision at the 99th percentile, and detect smaller changes in a patient’s troponin level as repeat testing occurs. This design affords clinicians greater confidence in the results with precision that provides the ability to measure slight, yet critical, changes to begin treatment.[1,2]
Chest pain is the cause of more than 8 million visits annually nationwide to emergency departments, but only 5.5 percent of those visits lead to serious diagnoses such as heart attacks.[3] Armed with data to properly triage patients sooner or to exclude myocardial infarctions, the Siemens Healthineers TnIH assays can help support testing initiatives tied to improving patient experience.
“Our emergency department is overcrowded with patients. If we can do a more efficient job at triaging patients to receive the proper level of care and to discharge the patients who do not need to stay in the emergency department, this will have a tremendous economic advantage for our healthcare system,” said Alan Wu, M.D., chief of clinical chemistry and toxicology at Zuckerberg San Francisco General Hospital and Trauma Center.
Siemens is launching the product at the 70th AACC Annual Scientific Meeting and Clinical Lab Expo taking place July 31 to Aug. 2 in Chicago.
Increases in levels of cardiac troponin T by high-sensitivity assay (hs-cTnT) over time are associated with later risk of death, coronary heart disease (CHD), and especially heart failure in apparently healthy middle-aged people, according to a report published June 8, 2016 in JAMA Cardiology[1].
The novel findings, based on a cohort of >8000 participants from the Atherosclerosis Risk in Communities (ARIC) study followed up to 16 years, are the first to show “an association between temporal hs-cTnT change and incident CHD events” in asymptomatic middle-aged adults,” write the authors, led by Dr John W McEvoy (Johns Hopkins University School of Medicine, Baltimore, MD).
Individuals with the greatest troponin increases over time had the highest risk for poor cardiac outcomes. The strongest association was for risk of heart failure, which reached almost 800% for those with the sharpest hs-cTnT rises.
Intriguingly, those in whom troponin levels fell at least 50% had a reduced mortality risk and may have had a slightly decreased risk of later HF or CHD.
“Serial testing over time with high-sensitivity cardiac troponins provided additional prognostic information over and above the usual clinical risk factors, [natriuretic peptide] levels, and a single troponin measurement. Two measurements appear better than one when it comes to informing risk for future coronary heart disease, heart failure, and death,” McEvoy told heartwirefrom Medscape.
He cautioned, though, that the conclusion is based on observational data and would need to be confirmed in clinical trials. Moreover, high-sensitivity cardiac troponin assays are widely used in Europe but are not approved in the US.
An important next step after this study, according to an accompanying editorial from Dr James Januzzi (Massachusetts General Hospital, Boston, MA), would be to evaluate whether the combination of hs-troponin and natriuretic peptides improves predictive value in this population[2].
“To the extent prevention is ultimately the holy grail for defeating the global pandemic of CHD, stroke, and HF, the main reason to do a biomarker study such as this would be to set the stage for a biomarker-guided strategy to improve the medical care for those patients at highest risk, as has been recently done with [natriuretic peptides],” he wrote.
The ARIC prospective cohort study entered and followed 8838 participants (mean age 56, 59% female, 21.4% black) in North Carolina, Mississippi, Minneapolis, and Maryland from January 1990 to December 2011. At baseline, participants had no clinical signs of CHD or heart failure.
Levels of hs-cTnT, obtained 6 years apart, were categorized as undetectable (<0.005 ng/mL), detectable (≥0.005 ng/mL to <0.014 ng/mL), and elevated (>0.014 ng/mL).
Troponin increases from <0.005 ng/mL to 0.005 ng/mL or higher independently predicted development of CHD (HR 1.41; 95% CI 1.16–1.63), HF (HR 1.96; 95% CI 1.62–2.37), and death (HR 1.50; 95% CI 1.31–1.72), compared with undetectable levels at both measurements.
Hazard ratios were adjusted for age, sex, race, body-mass index, C-reactive protein, smoking status, alcohol-intake history, systolic blood pressure, current antihypertensive therapy, diabetes, serum lipid and cholesterol levels, lipid-modifying therapy, estimated glomerular filtration rate, and left ventricular hypertrophy.
Subjects with >50% increase in hs-cTnT had a significantly increased risk of CHD (HR 1.28; 95% CI 1.09–1.52), HF (HR 1.60; 95% CI 1.35–1.91), and death (HR 1.39; 95% CI 1.22–1.59).
Risks for those end points fell somewhat for those with a >50% decrease in hs-cTnT (CHD: HR 0.47; 95% CI 0.22–1.03; HF: HR 0.49 95% CI 0.23–1.01; death: HR 0.57 95% CI 0.33–0.99).
Among participants with an adjudicated HF hospitalization, the group writes, associations of hs-cTnT changes with outcomes were of similar magnitude for those with HF with preserved ejection fraction (HFpEF) and HF with reduced ejection fraction (HFrEF).
Few biomarkers have been linked to increased risk for HFpEF, and few effective therapies exist for it. That may be due to problems identifying and enrolling patients with HFpEF in clinical trials, Dr McEvoy pointed out.
“We think the increased troponin over time reflects progressive myocardial injury or progressive myocardial damage,” Dr McEvoy said. “This is a window into future risk, particularly with respect to heart failure but other outcomes as well. It may suggest high-sensitivity troponins as a marker of myocardial health and help guide interventions targeting the myocardium.”
Moreover, he said, “We think that high-sensitivity troponin may also be a useful biomarker along with [natriuretic peptides] for emerging trials of HFpEF therapy.”
But whether hs-troponin has the potential for use as a screening tool is a question for future studies, according to McEvoy.
In his editorial, Januzzi pointed out several implications of the study, including the possibility for lowering cardiac risk in those with measurable hs-troponin, and that HF may be the most obvious outcome to target. Also, optimizing treatment and using cardioprotective therapies may reduce risk linked to increases in hs-troponin. Finally, long-term, large clinical trials on this issue will require a multidisciplinary team effort from various sectors.
“What is needed now are efforts toward developing strategies to upwardly bend the survival curves of those with a biomarker signature of risk, leveraging the knowledge gained from studies such as the report by McEvoy et al to improve public health,” he concluded.
mRNA Data Survival Analysis, Volume 2 (Volume Two: Latest in Genomics Methodologies for Therapeutics: Gene Editing, NGS and BioInformatics, Simulations and the Genome Ontology), Part 1: Next Generation Sequencing (NGS)
mRNA Data Survival Analysis
Curators: Larry H. Bernstein, MD, FCAP and Aviva Lev-Ari, PhD, RN
SURVIV for survival analysis of mRNA isoform variation
The rapid accumulation of clinical RNA-seq data sets has provided the opportunity to associate mRNA isoform variations to clinical outcomes. Here we report a statistical method SURVIV (Survival analysis of mRNA Isoform Variation), designed for identifying mRNA isoform variation associated with patient survival time. A unique feature and major strength of SURVIV is that it models the measurement uncertainty of mRNA isoform ratio in RNA-seq data. Simulation studies suggest that SURVIV outperforms the conventional Cox regression survival analysis, especially for data sets with modest sequencing depth. We applied SURVIV to TCGA RNA-seq data of invasive ductal carcinoma as well as five additional cancer types. Alternative splicing-based survival predictors consistently outperform gene expression-based survival predictors, and the integration of clinical, gene expression and alternative splicing profiles leads to the best survival prediction. We anticipate that SURVIV will have broad utilities for analysing diverse types of mRNA isoform variation in large-scale clinical RNA-seq projects.
Eukaryotic cells generate remarkable regulatory and functional complexity from a finite set of genes. Production of mRNA isoforms through alternative processing and modification of RNA is essential for generating this complexity. A prevalent mechanism for producing mRNA isoforms is the alternative splicing of precursor mRNA1. Over 95% of the multi-exon human genes undergo alternative splicing2, 3, resulting in an enormous level of plasticity in the regulation of gene function and protein diversity. In the last decade, extensive genomic and functional studies have firmly established the critical role of alternative splicing in cancer4, 5, 6. Alternative splicing is involved in a full spectrum of oncogenic processes including cell proliferation, apoptosis, hypoxia, angiogenesis, immune escape and metastasis7, 8. These cancer-associated alternative splicing patterns are not merely the consequences of disrupted gene regulation in cancer but in numerous instances actively contribute to cancer development and progression. For example, alternative splicing of genes encoding the Bcl-2 family of apoptosis regulators generates both anti-apoptotic and pro-apoptotic protein isoforms9. Alternative splicing of the pyruvate kinase M (PKM) gene has a significant impact on cancer cell metabolism and tumour growth10. A transcriptome-wide switch of the alternative splicing programme during the epithelial–mesenchymal transition plays an important role in cancer cell invasion and metastasis11, 12.
RNA sequencing (RNA-seq) has become a popular and cost-effective technology to study transcriptome regulation and mRNA isoform variation13, 14. As the cost of RNA-seq continues to decline, it has been widely adopted in large-scale clinical transcriptome projects, especially for profiling transcriptome changes in cancer. For example, as of April 2015 The Cancer Genome Atlas (TCGA) consortium had generated RNA-seq data on over 11,000 cancer patient specimens from 34 different cancer types. Within the TCGA data, breast invasive carcinoma (BRCA) has the largest sample size of RNA-seq data covering over 1,000 patients, and clinical information such as survival times, tumour stages and histological subtypes is available for the majority of the BRCA patients15. Moreover, the median follow-up time of BRCA patients is ~400 days, and 25% of the patients have more than 1,200 days of follow-up. Collectively, the large sample size and long follow-up time of the TCGA BRCA data set allow us to correlate genomic and transcriptomic profiles to clinical outcomes and patient survival times.
To date, systematic analyses have been performed to reveal the association between copy number variation, DNA methylation, gene expression and microRNA expression profiles with cancer patient survival16, 17. By contrast, despite the importance of mRNA isoform variation and alternative splicing, there have been limited efforts in transcriptome-wide survival analysis of alternative splicing in cancer patients. Most RNA-seq studies of alternative splicing in cancer transcriptomes focus on identifying ‘cancer-specific’ alternative splicing events by comparing cancer tissues with normal controls (see refs 18, 19, 20, 21, 22, 23 for examples). A recent analysis of TCGA RNA-seq data identified 163 recurrent differential alternative splicing events between cancer and normal tissues of three cancer types, among which five were found to have suggestive survival signals for breast cancer at a nominal P-value cutoff of 0.05 (ref. 21). Some other studies reported a significant survival difference between cancer patient subgroups after stratifying patients with overall mRNA isoform expression profiles24, 25. However, systematic cancer survival analyses of alternative splicing at the individual exon resolution have been lacking. Two main challenges exist for survival analyses of mRNA isoform variation and alternative splicing using RNA-seq data. The first challenge is to account for the estimation uncertainty of mRNA isoform ratios inferred from RNA-seq read counts. The statistical confidence of mRNA isoform ratio estimation depends on the RNA-seq read coverage for the events of interest, with larger read coverage leading to a more reliable estimation14. Modelling the estimation uncertainty of mRNA isoform ratio is an essential component of RNA-seq analyses of alternative splicing, as shown by various statistical algorithms developed for detecting differential alternative splicing from multi-group RNA-seq data14, 26, 27, 28,29. The second challenge, which is a general issue in survival analysis, is to properly model the association of mRNA isoform ratio with survival time, while accounting for missing data in survival time because of censoring, that is, patients still alive at the end of the survival study, whose precise survival time would be uncertain. To date, no algorithm has been developed for survival analyses of mRNA isoform variation that accounts for these sources of uncertainty simultaneously.
Here we introduce SURVIV (Survival analysis of mRNA Isoform Variation), a statistical model for identifying mRNA isoform ratios associated with patient survival times in large-scale cancer RNA-seq data sets. SURVIV models the estimation uncertainty of mRNA isoform ratios in RNA-seq data and tests the survival effects of isoform variation in both censored and uncensored survival data. In simulation studies, SURVIV consistently outperforms the conventional Cox regression survival analysis that ignores the measurement uncertainty of mRNA isoform ratio. We used SURVIV to identify alternatively spliced exons whose exon-inclusion levels significantly correlated with the survival times of invasive ductal carcinoma (IDC) patients from the TCGA breast cancer cohort. Survival-associated alternative splicing events are identified in gene pathways associated with apoptosis, oxidative stress and DNA damage repair. Importantly, we show that alternative splicing-based survival predictors outperform gene expression-based survival predictors in the TCGA IDC RNA-seq data set, as well as in TCGA data of five additional cancer types. Moreover, the integration of clinical information, gene expression and alternative splicing profiles leads to the best prediction of survival time.
SURVIV statistical model
The statistical model of SURVIV assesses the association between mRNA isoform ratio and patient survival time. While the model is generic for many types of alternative isoform variation, here we use the exon-skipping type of alternative splicing to illustrate the model (Fig. 1a). For each alternative exon involved in exon-skipping, we can use the RNA-seq reads mapping to its exon-inclusion or -skipping isoform to estimate its exon-inclusion level (denoted as ψ, or PSI that is Per cent Spliced In14). A key feature of SURVIV is that it models the RNA-seq estimation uncertainty of exon-inclusion level as influenced by the sequencing coverage for the alternative splicing event of interest. This is a critical issue in accurate quantitative analyses of mRNA isoform ratio in large-scale RNA-seq data sets14, 26, 27, 28, 29. Therefore, SURVIV contains two major components: the first to model the association of mRNA isoform ratio with patient survival time across all patients, and the second to model the estimation uncertainty of mRNA isoform ratio in each individual patient (Fig. 1a).
Figure 1: The statistical framework of the SURVIV model.
(a) For each patient k, the patient’s hazard rate λk(t) is associated with the baseline hazard rate λ0(t) and this patient’s exon-inclusion level ψk. The association of exon-inclusion level with patient survival is estimated by the survival coefficient β. The exon-inclusion level ψk is estimated from the read counts for the exon-inclusion isoform ICk and the exon-skipping isoform SCk. The proportion of the inclusion and skipping reads is adjusted by a normalization function f that considers the lengths of the exon-inclusion and -skipping isoforms (see details in Results and Supplementary Methods). (b) A hypothetical example to illustrate the association of exon-inclusion level with patient survival probability over time Sk(t), with the survival coefficient β=−1 and a constant baseline hazard rate λ0(t)=1. In this example, patients with higher exon-inclusion levels have lower hazard rates and higher survival probabilities. (c) The schematic diagram of an exon-skipping event. The exon-inclusion reads ICk are the reads from the upstream splice junction, the alternative exon itself and the downstream splice junction. The exon-skipping reads SCk are the reads from the skipping splice junction that directly connects the upstream exon to the downstream exon.
Briefly, for any individual exon-skipping event, the first component of SURVIV uses a proportional hazards model to establish the relationship between patient k’s exon-inclusion level ψk and hazard rate λk(t).
For each exon, the association between the exon-inclusion level and patient survival time is reflected by the survival coefficient β. A positive β means increased exon inclusion is associated with higher hazard rate and poorer survival, while a negative β means increased exon inclusion is associated with lower hazard rate and better survival. λ0(t) is the baseline hazard rate estimated from the survival data of all patients (see Supplementary Methods for the detailed estimation procedure). A particular patient’s survival probability over time Sk(t) can be calculated from the patient-specific hazard rate λk(t) as . Figure 1b illustrates a simple example with a negative β=−1 and a constant baseline hazard rate λ0(t)=1, where higher exon-inclusion levels are associated with lower hazard rates and higher survival probabilities.
The second component of SURVIV models the exon-inclusion level and its estimation uncertainty in individual patient samples. As illustrated in Fig. 1c, the exon-inclusion level ψk of a given exon in a particular sample can be estimated by the RNA-seq read count specific to the exon inclusion isoform (ICk) and the exon-skipping isoform (SCk). Other types of alternative splicing and mRNA isoform variation can be similarly modelled by this framework29. Given the effective lengths (that is, the number of unique isoform-specific read positions) of the exon-inclusion isoform (lI) and the exon-skipping isoform (lS), the exon-inclusion level ψk can be estimated as . Assuming that the exon-inclusion read count ICk follows a binomial distribution with the total read count nk=ICk+SCk, we have:
The binomial distribution models the estimation uncertainty of ψk as influenced by the total read count nk, in which the parameter pk represents the proportion of reads from the exon-inclusion isoform, given the exon-inclusion level ψk adjusted by a length normalization function f(ψk) based on the effective lengths of the isoforms. The definitions of effective lengths for all basic types of alternative splicing patterns are described in ref. 29.
Distinct from conventional survival analyses in which predictors do not have estimation uncertainty, the predictors in SURVIV are exon-inclusion levels ψk estimated from RNA-seq count data, and the confidence of ψk estimate for a given exon in a particular sample depends on the RNA-seq read coverage. We use the statistical framework of survival measurement error model30 to incorporate the estimation uncertainty of isoform ratio in the proportional hazards model. Using a likelihood ratio test, we test whether the exon-inclusion levels have a significant association with patient survival over the null hypothesis H0:β=0. The false discovery rate (FDR) is estimated using the Benjamini and Hochberg approach31. Details of the parameter estimation and likelihood ratio test in SURVIV are described in Supplementary Methods.
Figure 2: Simulation studies to assess the performance of SURVIV and the importance of modelling the estimation uncertainty of mRNA isoform ratio.
We compared our SURVIV model with Cox regression using point estimates of exon-inclusion levels, which does not consider the estimation uncertainty of the mRNA isoform ratio. (a) To study the effect of RNA-seq depth, we simulated the mean total splice junction read counts equal to 5, 10, 20, 50, 80 and 100 reads. We generated two sets of simulations with and without data-censoring. For each simulation, the true-positive rate (TPR) at 5% false-positive rate is plotted. The inset figure shows the empirical distribution of the mean total splice junction read counts in the TCGA IDC RNA-seq data (x axis in the log10 scale). (b) To faithfully represent the read count distribution in a real data set, we performed another simulation with read counts directly sampled from the TCGA IDC data. Sampled read counts were then multiplied by different factors ranging from 10 to 300% to simulate data sets with different RNA-seq read depth. Continuous and dashed lines represent the performance of SURVIV and Cox regression, respectively. Red lines represent the area under curve (AUC) of the ROC curve (TPR versus false-positive rate plot). Black lines represent the TPR at 5% false-positive rate.
Using these simulated data, we compared SURVIV with Cox regression in two settings, without or with censoring of the survival time. In the setting without censoring, the death and survival time of each individual is known. In the setting with censoring, certain individuals are still alive at the end of the survival study. Consequently, these patients have unknown death and survival time. Here, in the simulation with censoring, we assumed that 85% of the patients were still alive at the end of the study, similar to the censoring rate of the TCGA IDC data set. In both settings and with different depths of RNA-seq coverage, SURVIV consistently outperformed Cox regression in the true-positive rate at the same false-positive rate of 5% (Fig. 2a). As expected, we observed a more significant improvement in SURVIV over Cox regression when the RNA-seq read coverage was low (Fig. 2a).
To more faithfully recapitulate the read count distribution in a real cancer RNA-seq data set, we performed another simulation study with read counts directly sampled from the TCGA IDC data. To assess the influence of RNA-seq read depth on the performance of SURVIV and Cox regression, sampled read counts were then multiplied by different factors ranging from 10 to 300% to simulate data sets with different RNA-seq read depths (Fig. 2b). The TCGA IDC data set has an average RNA-seq depth of ~60 million paired-end reads per patient. Thus, the read depth of these simulated RNA-seq data sets ranged from ~6 million reads to 180 million reads per patient, representing low-coverage RNA-seq studies designed primarily for gene expression analysis32 up to high-coverage RNA-seq studies designed primarily for alternative isoform analysis29. At all levels of RNA-seq depth, SURVIV consistently outperformed Cox regression, as reflected by the area under curve of the receiver operating characteristic (ROC) curve as well as the true-positive rate at 5% false-positive rate (Fig. 2b). The improvement of SURVIV over Cox regression was particularly prominent when the read depth was low. For example, at 10% read depth, SURVIV had 7% improvement in area under curve (68% versus 61%) and 8% improvement in the true-positive rate at 5% false-positive rate (46% versus 38%). Collectively, these simulation results suggest that SURVIV achieves a higher accuracy by accounting for the estimation uncertainty of mRNA isoform ratio in RNA-seq data.
SURVIV analysis of TCGA IDC breast cancer data
To illustrate the practical utility of SURVIV, we used it to analyse the overall survival time of 682 IDC patients from the TCGA breast cancer (BRCA) RNA-seq data set (see Methods for details of the data source and processing pipeline). We chose to analyse IDC because it is the most frequent type of breast cancer33, comprising ~70% of patients in the TCGA breast cancer data set. To control for the effects of significant clinical parameters such as tumour stage and subtype and identify alternative splicing events associated with patient outcomes across multiple molecular and clinical subtypes, we followed the procedure of Croce and colleagues in analysing mRNA and microRNA prognostic signature of IDC33 and stratified the patients according to their clinical parameters. We then conducted SURVIV analysis in 26 clinical subgroups with at least 50 patients in each subgroup. We identified 229 exon-skipping events associated with patient survival in multiple clinical subgroups that met the criteria of SURVIV P-value≤0.01 in at least two subgroups of the same clinical parameter (cancer subtype, stage, lymph node, metastasis, tumour size, oestrogen receptor status, progesterone receptor status, HER2 status and age as shown in Fig. 3). DAVID (Database for Annotation, Visualization and Integrated Discovery) Gene Ontology analyses34 of the 229 alternative splicing events suggest an enrichment of genes in cancer-related functional categories such as intracellular signalling, apoptosis, oxidative stress and response to DNA damage (Supplementary Fig. 1). Table 1 shows a few selected examples of survival-associated alternative splicing events in cancer-related genes. Using two-means clustering of each individual exon’s inclusion levels, the 682 IDC patients can be segregated into two subgroups with significantly different survival times as illustrated by the Kaplan–Meier survival plot (Fig. 4). We also carried out hierarchical clustering of IDC patients using 176 survival-associated alternative exons (P≤0.01; SURVIV analysis of all IDC patients). Using the exon-inclusion levels of these 176 exons, we clustered IDC patients into three major subgroups, with 95, 194 and 389 patients, respectively. As illustrated by the Kaplan–Meier survival plots, the three subgroups had significantly different survival times (Supplementary Fig. 2).
Figure 3: SURVIV analysis of exon-skipping events in the TCGA IDC RNA-seq data set.
IDC patients are stratified into multiple clinical subgroups based on clinical parameters including cancer subtype, stage, lymph node status, metastasis, tumour size, oestrogen receptor status, progesterone receptor status, HER2 status and age. Only clinical subgroups with at least 50 patients are included in further analyses. Numbers of patients in the subgroups are indicated next to the names of the subgroups. Shown in the heatmap are the log10 SURVIV P-values of the 229 exons associated with patient survival (P≤0.01) in at least two subgroups of the same class of clinical parameters. Turquoise colour indicates positive correlation that higher exon-inclusion levels are associated with higher survival probabilities. Magenta colour indicates negative correlation that lower exon-inclusion levels are associated with higher survival probabilities.
Figure 4: Kaplan–Meier survival plots of IDC patients stratified by two-means clustering of the exon-inclusion levels of four survival-associated alternative splicing events.
Clustering was generated for each of the four exons separately. Black lines represent patients with high exon-inclusion levels. Red lines represent patients with low exon-inclusion levels. The P-values are from SURVIV analysis of the TCGA IDC RNA-seq data. (a) ATRIP. (b) BCL2L11. (c) CD74. (d) PCBP4.
Figure 5: Alternative splicing of STAT5A exon 5 is significantly associated with IDC patient survival.
(a) The gene structure of the STAT5A full-length isoform compared to the ΔEx5 isoform skipping the 5th exon. (b) Kaplan–Meier survival plot of IDC patients stratified by two-means clustering using exon-inclusion levels of STAT5A exon 5. The 420 patients in Group 1 (average exon 5 inclusion level=95%) have significantly higher survival probabilities than the 262 patients in Group 2 (average exon 5 inclusion level=85%) (SURVIV P=6.8e−4). (c) Exon 5 inclusion levels of IDC patients stratified by two-means clustering using exon 5 inclusion levels. Group 1 has 420 patients with average exon-inclusion level at 95%. Group 2 has 262 patients with average exon-inclusion level at 85%. (d) STAT5A exon 5 inclusion levels in normal breast tissues versus breast cancer tumour samples. Exon-inclusion levels are extracted from 86 TCGA breast cancer patients with matched normal and tumour samples. Normal breast tissues have average exon 5 inclusion level at 95%, compared to 91% average exon-inclusion level in tumour samples. Error bars represent 95% confidence interval of the mean.
Figure 6: Splicing factor regulatory network of survival-associated alternative splicing events in IDC.
(a–c) Kaplan–Meier survival plots of IDC patients stratified by the gene expression levels of three splicing factors: TRA2B (a, Cox regression P=1.8e−4), HNRNPH1 (b, P=3.4e−4) and SFRS3 (c, P=2.8e−3). Black lines represent patients with high gene expression levels. Red lines represent patients with low gene expression levels. (d) The exon-inclusion levels of a DHX30 alternative exon are negatively correlated with TRA2B gene expression levels (robust correlation coefficient r=−0.26, correlation P=1.2e−17). (e) The exon-inclusion levels of a MAP3K4 alternative exon are positively correlated withHNRNPH1 gene expression levels (robust correlation coefficient r=0.16, correlation P=2.6e−06). (f) A splicing co-expression network of the three splicing factors and their correlated survival-associated alternative exons. In total, 84 survival-associated alternative exons are significantly correlated with the three splicing factors. The positive/negative correlation between splicing factors and alternative exons is represented by blue/red lines, respectively. Exons whose inclusion levels are positively/negatively correlated with survival times are represented by blue/red dots, respectively. The size of the splicing factor circles is proportional to the number of correlated exons within the network.
Figure 7: Cross-validation of different classes of IDC survival predictors measured by the C-index
A C-index of 1 indicates perfect prediction accuracy and a C-index of 0.5 indicates random guess. The plots indicate the distribution of C-indexes from 100 rounds of cross-validation. The centre value of the box plot is the median C-index from 100 rounds of cross-validation. The notch represents the 95%confidence interval of the median. The box represents the 25 and 75% quantiles. The whiskers extended out from the box represent the 5 and 95% quantiles. Two-sided Wilcoxon test was used to compare different survival predictors. The different classes of predictors are: (a) clinical information (median C-index 0.67). (b) Gene expression (median C-index 0.68). (c) Alternative splicing (median C-index 0.71). (d) Clinical information+gene expression (median C-index 0.69). (e) Clinical information+alternative splicing (median C-index 0.73). (f) Clinical information+gene expression+alternative splicing (median C-index 0.74). Note that ‘Gene’ refers to ‘Gene-level expression’ in these plots.
Next, we carried out the SURVIV analysis in five additional cancer types in TCGA, including GBM (glioblastoma multiforme), KIRC (kidney renal clear cell carcinoma), LGG (lower grade glioma), LUSC (lung squamous cell carcinoma) and OV (ovarian serous cystadenocarcinoma). As expected, the number of significant events at different FDR or P-value significance cutoffs varied across cancer types, with LGG having the strongest survival-associated alternative splicing signals with 660 significant exon-skipping events at FDR≤5% (Supplementary Data 3 and 4). Strikingly, regardless of the number of significant events, alternative splicing-based survival predictors outperformed gene expression-based survival predictors across all cancer types (Supplementary Fig. 3), consistent with our initial observation on the IDC data set.
Alternative processing and modification of mRNA, such as alternative splicing, allow cells to generate a large number of mRNA and protein isoforms with diverse regulatory and functional properties. The plasticity of alternative splicing is often exploited by cancer cells to produce isoform switches that promote cancer cell survival, proliferation and metastasis7, 8. The widespread use of RNA-seq in cancer transcriptome studies15, 47, 48 has provided the opportunity to comprehensively elucidate the landscape of alternative splicing in cancer tissues. While existing studies of alternative splicing in large-scale cancer transcriptome data largely focused on the comparison of splicing patterns between cancer and normal tissues or between different subtypes of cancer18, 21, 49, additional computational tools are needed to characterize the clinical relevance of alternative splicing using massive RNA-seq data sets, including the association of alternative splicing with phenotypes and patient outcomes.
We have developed SURVIV, a novel statistical model for survival analysis of alternative isoform variation using cancer RNA-seq data. SURVIV uses a survival measurement error model to simultaneously model the estimation uncertainty of mRNA isoform ratio in individual patients and the association of mRNA isoform ratio with survival time across patients. Compared with the conventional Cox regression model that uses each patient’s mRNA isoform ratio as a point estimate, SURVIV achieves a higher accuracy as indicated by simulation studies under a variety of settings. Of note, we observed a particularly marked improvement of SURVIV over Cox regression for low- and moderate-depth RNA-seq data (Fig. 2b). This has important practical value because many clinical RNA-seq data sets have large sample size but relatively modest sequencing depth.
Using the TCGA IDC breast cancer RNA-seq data of 682 patients, SURVIV identified 229 alternative splicing events associated with patient survival time, which met the criteria of SURVIVP-values≤0.01 in multiple clinical subgroups. While the statistical threshold seemed loose, several lines of evidence suggest the functional and clinical relevance of these survival-associated alternative splicing events. These alternative splicing events were frequently identified and enriched in the gene functional groups important for cancer development and progression, including apoptosis, DNA damage response and oxidative stress. While some of these events may simply reflect correlation but not causal effect on cancer patient survival, other events may play an active role in regulating cancer cell phenotypes. For example, a survival-associated alternative splicing event involving exon 5 of STAT5A is known to regulate the activity of this transcription factor with important roles in epithelial cell growth and apoptosis37. Using a co-expression network analysis of splicing factor to exon correlation across all patients, we identified three splicing factors (TRA2B, HNRNPH1 and SFRS3) as potential hubs of the survival-associated alternative splicing network of IDC. The expression levels of all three splicing factors were negatively associated with patient survival times (Fig. 6a–c), and both TRA2B and HNRNPH1 were previously reported to have an impact on cancer-related molecular pathways40, 41, 42, 43, 44, 45. Finally, despite the limited power in detecting individual events, we show that the survival-associated alternative splicing events can be used to construct a predictor for patient survival, with an accuracy higher than predictors based on clinical parameters or gene expression profiles (Fig. 7). This further demonstrates the potential biological relevance and clinical utility of the identified alternative splicing events.
We performed cross-validation analyses to evaluate and compare the prognostic value of alternative splicing, gene expression and clinical information for predicting patient survival, either independently or in combination. As expected, the combined use of all three types of information led to the best prediction accuracy. Because we used penalized regression to build the prediction model, combining information from multiple layers of data did not necessarily increase the number of predictors in the model. The perhaps more surprising and intriguing result is that alternative splicing-based predictors appear to outperform gene expression-based predictors when used alone and when either type of data was combined with clinical information (Fig. 7). We observed the same trend in five additional cancer types (Supplementary Fig. 3). We note that this finding was consistent with a previous report that cancer subtype classification based on splicing isoform expression performed better than gene expression-based classification25. While this trend seems counterintuitive because accurate estimation of gene expression requires much lower RNA-seq depth than accurate estimation of alternative splicing29, one possible explanation may be the inherent characteristic of isoform ratio data. By definition, mRNA isoform ratio is estimated as the ratio of multiple mRNA isoforms from a single gene. Therefore, mRNA isoform ratio data have a ‘built-in’ internal control that could be more robust against certain artefacts and confounding issues that influence gene expression estimates across large clinical RNA-seq data sets, such as poor sample quality and RNA degradation12. Regardless of the reasons, our data call for further studies to fully explore the utility of mRNA isoform ratio data for various clinical research applications.
The SURVIV source code is available for download at https://github.com/Xinglab/SURVIV. SURVIV is a general statistical model for survival analysis of mRNA isoform ratio using RNA-seq data. The current statistical framework of SURVIV is applicable to RNA-seq based count data for all basic types of alternative splicing patterns involving two isoform choices from an alternatively spliced region, such as exon-skipping, alternative 5′ splice sites, alternative 3′ splice sites, mutually exclusive exons and retained introns, as well as other forms of alternative isoform variation such as RNA editing. With the rapid accumulation of clinical RNA-seq data sets, SURVIV will be a useful tool for elucidating the clinical relevance and potential functional significance of alternative isoform variation in cancer and other diseases.
Novel Discoveries in Molecular Biology and Biomedical Science, Volume 2 (Volume Two: Latest in Genomics Methodologies for Therapeutics: Gene Editing, NGS and BioInformatics, Simulations and the Genome Ontology), Part 1: Next Generation Sequencing (NGS)
Novel Discoveries in Molecular Biology and Biomedical Science
Curator: Larry H. Bernstein, MD, FCAP
UPDATED on 6/1/2016
The following is a collection of current articles on noncoding DNA, synthetic genome engineering, protein regulation of apoptosis, drug design, and geometrics.
No longer ‘junk DNA’ — shedding light on the ‘dark matter’ of the genome
A new tool called “LIGR-Seq” enables scientists to explore in depth what non-coding RNAs actually do in human cells May 23, 2016
he LIGR-seq method for global-scale mapping of RNA-RNA interactions in vivo to reveal unexpected functions for uncharacterized RNAs that act via base-pairing interactions (credit: University of Toronto)
What used to be dismissed by many as “junk DNA” has now become vitally important, as accelerating genomic data points to the importance of non-coding RNAs (ncRNAs) — a genome’s messages that do not specifically code for proteins — in development and disease.
But our progress in understanding these molecules has been slow because of the lack of technologies that allow for systematic mapping of their functions.
Now, professor Benjamin Blencowe’s team at the University of Toronto’s Donnelly Centre has developed a method called “LIGR-seq” that enables scientists to explore in depth what ncRNAs do in human cells.
The study, described in Molecular Cell, was published on May 19, along with two other papers, in Molecular Cell and Cell, respectively, from Yue Wan’s group at the Genome Institute of Singapore and Howard Chang’s group at Stanford University in California, who developed similar methods to study RNAs in different organisms.
Of the 3 billion letters in the human genome, only two per cent make up the protein-coding genes. The genes are copied, or transcribed, into messenger RNA (mRNA) molecules, which provide templates for building proteins that do most of the work in the cell. Much of the remaining 98 per cent of the genome was initially considered by some as lacking in functional importance. However, large swaths of the non-coding genome — between half and three quarters of it — are also copied into RNA.
So then what might the resulting ncRNAs do? That depends on whom you ask. Some researchers believe that most ncRNAs have no function, that they are just a by-product of the genome’s powerful transcription machinery that makes mRNA. However, it is emerging that many ncRNAs do have important roles in gene regulation — some ncRNAs act as carriages for shuttling the mRNAs around the cell, or provide a scaffold for other proteins and RNAs to attach to and do their jobs.
But the majority of available data has trickled in piecemeal or through serendipitous discovery. And with emerging evidence that ncRNAs could drive disease progression, such as cancer metastasis, there was a great need for a technology that would allow a systematic functional analysis of ncRNAs.
“Up until now, with existing methods, you had to know what you are looking for because they all require you to have some information about the RNA of interest. The power of our method is that you don’t need to preselect your candidates; you can see what’s occurring globally in cells, and use that information to look at interesting things we have not seen before and how they are affecting biology,” says Eesha Sharma, a PhD candidate in Blencowe’s group who, along with postdoctoral fellow Tim Sterne-Weiler, co-developed the method.
The human RNA-RNA interactome, showing interactions detected by LIGR-seq (credit: University of Toronto)
The new ‘‘LIGation of interacting RNA and high-throughput sequencing’’ (LIGR-seq) tool captures interactions between different RNA molecules. When two RNA molecules have matching sequences — strings of letters copied from the DNA blueprint — they will stick together like Velcro. With LIGR-seq, the paired RNA structures are removed from cells and analyzed by state-of-the-art sequencing methods to precisely identify the RNAs that are stuck together.
“Most researchers in the life sciences agree that there’s an urgent need to understand what ncRNAs do. This technology will open the door to developing a new understanding of ncRNA function,” says Blencowe, who is also a professor in the Department of Molecular Genetics.
Not having to rely on pre-existing knowledge will boost the discovery of RNA pairs that have never been seen before. Scientists can also, for the first time, look at RNA interactions as they occur in living cells, in all their complexity, unlike in the juices of mashed up cells that they had to rely on before. This is a bit like moving on to explore marine biology from collecting shells on the beach to scuba-diving among the coral reefs, where the scope for discovery is so much bigger.
Actually, ncRNAs come in multiple flavors: there’s rRNA, tRNA, snRNA, snoRNA, piRNA, miRNA, and lncRNA, to name a few, where prefixes reflect the RNA’s place in the cell or some aspect of its function. But the truth is that no one really knows the extent to which these ncRNAs control what goes on in the cell, or how they do this.
Discoveries
Nonetheless, the new technology developed by Blencowe’s group has been able to pick up new interactions involving all classes of RNAs and has already revealed some unexpected findings.
The team discovered new roles for small nucleolar RNAs (snoRNAs), which normally guide chemical modifications of other ncRNAs. It turns out that some snoRNAs can also regulate stability of a set of protein-coding mRNAs. In this way, snoRNAs can also directly influence which proteins are made, as well as their abundance, adding a new level of control in cell biology.
And this is only the tip of the iceberg; the researchers plan to further develop and apply their technology to investigate the ncRNAs in different settings.
“We would like to understand how ncRNAs function during development. We are particularly interested in their role in the formation of neurons. But we will also use our method to discover and map changes in RNA-RNA interactions in the context of human diseases,” says Blencowe.
Abstract of Global Mapping of Human RNA-RNA Interactions
The majority of the human genome is transcribed into non-coding (nc)RNAs that lack known biological functions or else are only partially characterized. Numerous characterized ncRNAs function via base pairing with target RNA sequences to direct their biological activities, which include critical roles in RNA processing, modification, turnover, and translation. To define roles for ncRNAs, we have developed a method enabling the global-scale mapping of RNA-RNA duplexes crosslinked in vivo, “LIGation of interacting RNA followed by high-throughput sequencing” (LIGR-seq). Applying this method in human cells reveals a remarkable landscape of RNA-RNA interactions involving all major classes of ncRNA and mRNA. LIGR-seq data reveal unexpected interactions between small nucleolar (sno)RNAs and mRNAs, including those involving the orphan C/D box snoRNA, SNORD83B, that control steady-state levels of its target mRNAs. LIGR-seq thus represents a powerful approach for illuminating the functions of the myriad of uncharacterized RNAs that act via base-pairing interactions.
Understanding the unknown functions of these genes may lead to the creation of new diagnostic tests for clinical laboratories and anatomic pathology groups
Once again, J. Craig Venter, PhD, is charting new ground in gene sequencing andgenomic science. This time his research team has built upon the first synthetic cell they created in 2010 to build a more sophisticated synthetic cell. Their findings from this work may give pathologists and medical laboratory scientists new tools to diagnose disease.
Recently the research team at the J. Craig Venter Institute (JCVI) and Synthetic Genomics, Inc. (SGI) published their latest findings. Among the things they learned is that science still does not understand the functions of about a third of the genes required for their synthetic cells to function.
JCVI-syn3.0 Could Radically Alter Understanding of Human Genome
Based in La Jolla, Calif., and Rockville, Md., JCVI is a not-for-profit research institute aiming to advance genomics. Building upon its first synthetic cell—Mycoplasma mycoides (M. mycoides) JCVI-syn1.0, which JCVI constructed in 2010—the same team of scientists created the first minimal synthetic bacterial cell, which they calledJCVI-syn3.0. This new artificial cell contains 531,560 base pairs and just 473 genes, which means it is the smallest genome of any organism that can be grown in laboratory media, according to a JCVI-SGI statement.
For pathologists and medical laboratory leaders, the creation of a synthetic life form is a milestone toward better understanding genome sequencing and how this new knowledge may help advance both diagnostics and therapeutics.
“What we’ve done is important because it is a step toward completely understanding how a living cell works,” Clyde Hutchison III, PhD, told New Scientist. “If we can really understand how the cell works, then we will be able to design cells efficiently for the production of pharmaceutical and other useful products.” Hutchison is Professor Emeritus of Microbiology and Immunology at the University of North Carolina at Chapel Hill, Distinguished Professor at the J. Craig Venter Institute, a member of the National Academy of Sciences, and a fellow of the American Academy of Arts and Sciences.
Clyde Hutchison, III, PhD (above), Professor Emeritus of Microbiology and Immunology at the University of North Carolina at Chapel Hill and Distinguished Professor at the J. Craig Venter Institute, stated that his team’s “goal is to have a cell for which the precise biological function of every gene is known.” (Photo credit: JCVI.)
Understanding a Gene’s True Purpose
According to the JCVI researchers, 149 genes have no known purpose. They are, however, necessary for life and health.
“We know about two-thirds of the essential biology, and we’re missing a third,” stated J. Craig Venter, PhD, Founder and CEO of JCVI, in a story published by MedPage Today.
This knowledge is based upon decades of research. JCVI seeks to create a minimal cell operating system to understand biology, while also providing what the JCVI statement called a “chassis for use in industrial applications.”
What Do these Genes Do Anyway?
The JCVI team found that among most genes’ biological functions:
“JCVI-syn3.0 is a working approximation of a minimal cellular genome—a compromise between a small genome size and a workable growth rate for an experimental organism. It retains almost all the genes that are involved in the synthesis and processing of macromolecules. Unexpectedly, it also contains 149 genes with unknown biological functions, suggesting the presence of undiscovered functions that are essential for life,” the researchers told the journal Science.
More research is needed, the scientists say, into the 149 genes that appear to lack specific biologic functions.
Unlocking Mystery of the 149 Genes Could Lead to Advances in Genomic Science
“Finding so many genes without a known function is unsettling, but it’s exciting because it’s left us with much still to learn. It’s like the ‘dark matter’ of biology,” said Alistair Elfick, PhD, Chair of Synthetic Biological Engineering, University of Edinburgh, UK, in the New Scientist article.
Studies such as JCVI’s research is key to broadening understanding and framing appropriate questions about scientific, ethical, and economic implications of synthetic biology.
The creation of a synthetic cell will have a profound and positive impact on understanding of biology and how life works, JCVI said.
Such research may inspire new whole genome synthesis tools and semi-automated processes that could dramatically affect clinical laboratory procedures. It also could lead to new techniques and tools for advanced vaccine and pharmaceuticals, JCVI pointed out.
No single technique has set the molecular biology field ablaze with excitement and potential like the CRISPR-Cas9 genome editing system has following its introduction only a few short years ago. The following articles represent the flexibility of this technique to potentially treat a host of genetic disorders and possibly even prevent the onset of disease.
Scientists recently convened at the CRISPR Precision Gene Editing Congress, held in Boston, to discuss the new technology. As with any new technique, scientists have discovered that CRISPR comes with its own set of challenges, and the Congress focused its discussion around improving specificity, efficiency, and delivery.
With a staggering number of papers published in the past several years involving the characterization and use of the CRISPR/Cas9 gene editing system, it is surprising that researchers are still finding new features of the versatile molecular scissor enzyme.
If a Cas9 nuclease variant could be engineered that was less grabby, it might loosen its grip on DNA sequences throughout the genome—except those sequences representing on-target sites. That’s the assumption that guided a new investigation by researchers at Massachusetts General Hospital.
The gene-editing technology known as CRISPR-Cas9 is starting to raise expectations in the therapeutic realm. In fact, CRISPR-Cas9 and other CRISPR systems are moving so close to therapeutic uses that the technology’s ethical implications are starting to attract notice.
Published: Tuesday, May 24, 2016 A comparison of synthetic gene-activating Cas9 proteins can help guide research and development of therapeutic approaches.
The CRISPR-Cas9 system has come to be known as the quintessential tool that allows researchers to edit the DNA sequences of many organisms and cell types. However, scientists are also increasingly recognizing that it can be used to activate the expression of genes. To that end, they have built a number of synthetic gene activating Cas9 proteins to study gene functions or to compensate for insufficient gene expression in potential therapeutic approaches.
“The possibility to selectively activate genes using various engineered variants of the CRISPR-Cas9 system left many researchers questioning which of the available synthetic activating Cas9 proteins to use for their purposes. The main challenge was that all had been uniquely designed and tested in different settings; there was no side-by-side comparison of their relative potentials,” said George Church, Ph.D., who is Core Faculty Member at the Wyss Institute for Biologically Inspired Engineering at Harvard University, leader of its Synthetic Biology Platform, and Professor of Genetics at Harvard Medical School. “We wanted to provide that side-by-side comparison to the biomedical research community.”
In a study published on 23 May in Nature Methods, the Wyss Institute team reports how it rigorously compared and ranked the most commonly used artificial Cas9 activators in different cell types from organisms including humans, mice and flies. The findings provide a valuable guide to researchers, allowing them to streamline their endeavors.
The team also included Wyss Core Faculty Member James Collins, Ph.D., who also is the Termeer Professor of Medical Engineering & Science and Professor of Biological Engineering at the Massachusetts Institute of Technology (MIT)’s Department of Biological Engineering and Norbert Perrimon, Ph.D., a Professor of Genetics at Harvard Medical School.
Gene activating Cas9 proteins are fused to variable domains borrowed from proteins with well-known gene activation potentials and engineered so that the DNA editing ability is destroyed. In some cases, the second component of the CRISPR-Cas9 system, the guide RNA that targets the complex to specific DNA sequences, also has been engineered to bind gene-activating factors.
“We first surveyed seven advanced Cas9 activators, comparing them to each other and the original Cas9 activator that served to provide proof-of-concept for the gene activation potential of CRISPR-Cas9. Three of them, provided much higher gene activation than the other candidates while maintaining high specificities toward their target genes,” said Marcelle Tuttle, Research Fellow at the Wyss and a co-lead author of the study.
The team went on to show that the three top candidates were comparable in driving the highest level of gene expression in cells from humans, mice and fruit flies, irrespective of their tissue and developmental origins. The researchers also pinpointed ways to further maximize gene activation employing the three leading candidates.
“In some cases, maximum possible activation of a target gene is necessary to achieve a cellular or therapeutic effect. We managed to cooperatively enhance expression of specific genes when we targeted them with three copies of a top performing activator using three different guide RNAs,” said Alejandro Chavez, Ph.D., a Postdoctoral Fellow and the study’s co-first author.
“The ease of use of CRISPR-Cas9 offers enormous potential for development of genome therapeutics. This study provides valuable new design criteria that will help enable synthetic biologists and bioengineers to develop more effective targeted genome engineering technologies in the future,” said Wyss Institute Founding Director Donald Ingber, M.D., Ph.D., who is the Judah Folkman Professor of Vascular Biology at Harvard Medical School and the Vascular Biology Program at Boston Children’s Hospital, and also Professor of Bioengineering at the Harvard John A. Paulson School of Engineering and Applied Sciences.
Engineering T Cells to Functionally Cure HIV-1 Infection
Despite the ability of antiretroviral therapy to minimize human immunodeficiency virus type 1 (HIV-1) replication and increase the duration and quality of patients’ lives, the health consequences and financial burden associated with the lifelong treatment regimen render a permanent cure highly attractive. Although T cells play an important role in controlling virus replication, they are themselves targets of HIV-mediated destruction. Direct genetic manipulation of T cells for adoptive cellular therapies could facilitate a functional cure by generating HIV-1–resistant cells, redirecting HIV-1–specific immune responses, or a combination of the two strategies. In contrast to a vaccine approach, which relies on the production and priming of HIV-1–specific lymphocytes within a patient’s own body, adoptive T-cell therapy provides an opportunity to customize the therapeutic T cells prior to administration. However, at present, it is unclear how to best engineer T cells so that sustained control over HIV-1 replication can be achieved in the absence of antiretrovirals. This review focuses on T-cell gene-engineering and gene-editing strategies that have been performed in efforts to inhibit HIV-1 replication and highlights the requirements for a successful gene therapy–mediated functional cure.
Automated top-down design technique simplifies creation of DNA origami nanostructures
Nanoparticles for drug delivery and cell targeting, nanoscale robots, custom-tailored optical devices, and DNA as a storage medium are among the possible applications
May 27, 2016
The boldfaced line, known as a spanning tree, follows the desired geometric shape of the target DNA origami design method, touching each vertex just once. A spanning tree algorithm is used to map out the proper routing path for the DNA strand. (credit: Public Domain)
MIT, Baylor College of Medicine, and Arizona State University Biodesign Institute researchers have developed a radical new top-down DNA origami* design method based on a computer algorithm that allows for creating designs for DNA nanostructures by simply inputting a target shape.
DNA origami (using DNA to design and build geometric structures) has already proven wildly successful in creating myriad forms in 2- and 3- dimensions, which conveniently self-assemble when the designed DNA sequences are mixed together. The tricky part is preparing the proper DNA sequence and routing design for scaffolding and staple strands to achieve the desired target structure. Typically, this is painstaking work that must be carried out manually.
The new algorithm, which is reported together with a novel synthesis approach in the journal Science, promises to eliminate all that and expands the range of possible applications of DNA origami in biomolecular science and nanotechnology. Think nanoparticles for drug delivery and cell targeting, nanoscale robots in medicine and industry, custom-tailored optical devices, and most interesting: DNA as a storage medium, offering retention times in the millions of years.**
Shape-shifting, top-down software
Unlike traditional DNA origami, in which the structure is built up manually by hand, the team’s radical top-down autonomous design method begins with an outline of the desired form and works backward in stages to define the required DNA sequence that will properly fold to form the finished product.
“The Science paper turns the problem around from one in which an expert designs the DNA needed to synthesize the object, to one in which the object itself is the starting point, with the DNA sequences that are needed automatically defined by the algorithm,” said Mark Bathe, an associate professor of biological engineering at MIT, who led the research. “Our hope is that this automation significantly broadens participation of others in the use of this powerful molecular design paradigm.”
The algorithm, which is known as DAEDALUS (DNA Origami Sequence Design Algorithm for User-defined Structures) after the Greek craftsman and artist who designed labyrinths that resemble origami’s complex scaffold structures, can build any type of 3-D shape, provided it has a closed surface. This can include shapes with one or more holes, such as a torus.
A simplified version of the top-down procedure used to design scaffolded DNA origami nanostructures. It starts with a polygon corresponding to the target shape. Software translates a wireframe version of this structure into a plan for routing DNA scaffold and staple strands. That enables a 3D DNA-based atomic-level structural model that is then validated using 3D cryo-EM reconstruction. (credit: adapted from Biodesign Institute images)
With the new technique, the target geometric structure is first described in terms of a wire mesh made up of polyhedra, with a network of nodes and edges. A DNA scaffold using strands of custom length and sequence is generated, using a “spanning tree” algorithm — basically a map that will automatically guide the routing of the DNA scaffold strand through the entire origami structure, touching each vertex in the geometric form once. Complementary staple strands are then assigned and the final DNA structural model or nanoparticle self-assembles, and is then validated using 3D cryo-EM reconstruction.
The software allows for fabricating a variety of geometric DNA objects, including 35 polyhedral forms (Platonic, Archimedean, Johnson and Catalan solids), six asymmetric structures, and four polyhedra with nonspherical topology, using inverse design principles — no manual base-pair designs needed.
To test the method, simpler forms known as Platonic solids were first fabricated, followed by increasingly complex structures. These included objects with nonspherical topologies and unusual internal details, which had never been experimentally realized before. Further experiments confirmed that the DNA structures produced were potentially suitable for biological applications since they displayed long-term stability in serum and low-salt conditions.
Biological research uses
The research also paves the way for designing nanoscale systems mimicking the properties of viruses, photosynthetic organisms, and other sophisticated products of natural evolution. One such application is a scaffold for viral peptides and proteins for use as vaccines. The surface of the nanoparticles could be designed with any combination of peptides and proteins, located at any desired location on the structure, in order to mimic the way in which a virus appears to the body’s immune system.
The researchers demonstrated that the DNA nanoparticles are stable for more than six hours in serum, and are now attempting to increase their stability further.
The nanoparticles could also be used to encapsulate the CRISPR-Cas9 gene editing tool. The CRISPR-Cas9 tool has enormous potential in therapeutics, thanks to its ability to edit targeted genes. However, there is a significant need to develop techniques to package the tool and deliver it to specific cells within the body, Bathe says.
This is currently done using viruses, but these are limited in the size of package they can carry, restricting their use. The DNA nanoparticles, in contrast, are capable of carrying much larger gene packages and can easily be equipped with molecules that help target the right cells or tissue.
The most exciting aspect of the work, however, is that it should significantly broaden participation in the application of this technology, Bathe says, much like 3-D printing has done for complex 3-D geometric models at the macroscopic scale.
* DNA origami brings the ancient Japanese method of paper folding down to the molecular scale. The basics are simple: Take a length of single-stranded DNA and guide it into a desired shape, fastening the structure together using shorter “staple strands,” which bind in strategic places along the longer length of DNA. The method relies on the fact that DNA’s four nucleotide letters—A, T, C, & G stick together in a consistent manner — As always pairing with Ts and Cs with Gs.
The DNA molecule in its characteristic double stranded form is fairly stiff, compared with single-stranded DNA, which is flexible. For this reason, single stranded DNA makes for an ideal lace-like scaffold material. Further, its pairing properties are predictable and consistent (unlike RNA).
** A single gram of DNA can store about 700 terabytes of information — an amount equivalent to 14,000 50-gigabyte Blu-ray disks — and could potentially be operated with a fraction of the energy required for other information storage options.
Essential role of miRNAs in orchestrating the biology of the tumor microenvironment
MicroRNAs (miRNAs) are emerging as central players in shaping the biology of the Tumor Microenvironment (TME). They do so both by modulating their expression levels within the different cells of the TME and by being shuttled among different cell populations within exosomes and other extracellular vesicles. This review focuses on the state-of-the-art knowledge of the role of miRNAs in the complexity of the TME and highlights limitations and challenges in the field. A better understanding of the mechanisms of action of these fascinating micro molecules will lead to the development of new therapeutic weapons and most importantly, to an improvement in the clinical outcome of cancer patients. Keywords: Exosomes, microRNAs, Tumor microenvironment, Cancer
While cancer treatment and survival have improved worldwide, the need for further understanding of the underlying tumor biology remains. In recent years, there has been a significant shift in scientific focus towards the role of the tumor microenvironment (TME) on the development, growth, and metastatic spread of malignancies. The TME is defined as the surrounding cellular environment enmeshed around the tumor cells including endothelial cells, lymphocytes, macrophages, NK cells, other cells of the immune system, fibroblasts, mesenchymal stem cells (MSCs), and the extracellular matrix (ECM). Each of these components interacts with and influences the tumor cells, continually shifting the balance between pro- and anti-tumor phenotype. One of the predominant methods of communication between these cells is through extracellular vesicles and their microRNA (miRNA) cargo. Extracellular vesicles (EVs) are between 30 nm to a few microns in diameter, are surrounded by a phospholipid bilayer membrane, and are released from a variety of cell types into the local environment. There are three well characterized groups of EVs: 1) exosomes, typically 30–100 nm, 2) microvesicles (or ectosomes), typically 100–1000 nm, and 3) large oncosomes, typically 1–10 μm. Each of these categories has a distinctly unique biogenesis and purpose in cellcell communication despite the fact that current laboratory methods do not always allow precise differentiation. EVs are found to be enriched with membrane-bound proteins, lipid raft-associated and cytosolic proteins, lipids, DNA, mRNAs, and miRNAs, all of which can be transferred to the recipient cell upon fusion to allow cell-cell communications [1]. Of these, miRNAs have been of particular interest in cancer research, both as modifiers of transcription and translation as well as direct inhibitors or enhancers of key regulatory proteins. These miRNAs are a large family of small non-coding RNAs (19–24 nucleotides) and are known to be aberrantly expressed, both in terms of content as well as number, in both the tumor cells and the cells of the TME. Synthesis of these mature miRNA is a complex process, starting with the transcription of long, capped, and polyadenylated pri-miRNA by RNA polymerase II. These are cropped into a 60–100 nucleotide hairpinstructure pre-miRNA by the microprocessor, a heterodimer of Drosha (a ribonuclease III enzyme) and DGCR8 (DiGeorge syndrome critical region gene 8). The premiRNA is then exported to the cytoplasm by exportin 5, cleaved by Dicer, and separated into single strands by helicases. The now mature miRNA are incorporated into the RNA-induced silencing complex (RISC), a cytoplasmic effector machine of the miRNA pathway. The primary mechanism of action of the mature miRNA-RISC complex is through their binding to the 3’ untranslated region, or less commonly the 5’ untranslated region, of target mRNA, leading to protein downregulation either via translational repression or mRNA degradation. More recently, it has been shown that miRNAs can also upregulate the expression of target genes [2]. MiRNA genes are mostly intergenic and are transcribed by independent promoters [3] but can also be encoded by introns, sharing the same promoter of their host gene [4]. MiRNAs undergo the same regulatory mechanisms of any other protein coding gene (promoter methylation, histone modifications, etc.…) [5, 6]. Interestingly, each miRNA may have contradictory effects both within varying tumor cell lines and within different cells of the TME. In this review, we provide a state-of-the-art description of the key role that miRNAs have in the communication between tumor cells and the TME and their subsequent effects on the malignant phenotype. Finally, this review has made every effort to clarify, whenever possible, whether the reference is to the −3p or the -5p miRNA. Whenever such clarification has not been provided, this indicates that it was not possible to infer such information from the cited bibliography.
Angiogenesis and miRNAs Cellular plasticity, critical in the development of malignancy, includes the many diverse mechanisms elicited by cancer cells to increase their malignant potential and develop increasing treatment resistance. One such mechanism, angiogenesis, is critical to the development of metastatic disease, affecting both the growth of malignant cells locally and their survival at distant sites. In the last ten years, miRNAs, often packaged in tumor cell-derived exosomes, have emerged as important contributors to the complicated regulation and balance of pro- and anti-angiogenic factors.
Most commonly, miRNAs derived from cancer cells have oncogenic activity, promoting angiogenesis and tumor growth and survival. The most-well characterized of the pro-angiogenic miRNAs, the miR-17-92 cluster encoding six miRNAs (miR-17, −18a, −19a, −19b, −20a, and −92a), is found on chromosome 13, and is highly conserved among vertebrates [7]. The complex and multifaceted functions of the miR-17-92 cluster are summarized in Fig. 1. Amplification, both at the genetic and RNA level, of miR-17-92 was initially found in several lymphoma cell lines and has subsequently been observed in multiple mouse tumor models [7].
Central role of the miR-17-92 cluster in the biology of the TME. The miR-17-92 cluster encoding miR-17, −18a, −19b, −20a, and -92a is upregulated in multiple tumor types and interacts with various components of the TME to finely “tune” the TME through a complex combination of pro- and anti-tumoral effects
Most commonly, miRNAs derived from cancer cells have oncogenic activity, promoting angiogenesis and tumor growth and survival. The most-well characterized of the pro-angiogenic miRNAs, the miR-17-92 cluster encoding six miRNAs (miR-17, −18a, −19a, −19b, −20a, and −92a), is found on chromosome 13, and is highly conserved among vertebrates [7]. The complex and multifaceted functions of the miR-17-92 cluster are summarized in Fig. 1. Amplification, both at the genetic and RNA level, of miR-17-92 was initially found in several lymphoma cell lines and has subsequently been observed in multiple mouse tumor models [7]. Up-regulation of this particular locus has further been confirmed in miRnome analysis across multiple different tumor types, including lung, breast, stomach, prostate, colon, and pancreatic cancer [8]. The miR-17-92 cluster is directly activated by Myc and modulates a variety of downstream transcription factors important in cell cycle regulation and apoptosis including activation of E2F family and Cyclin-dependent kinase inhibitor (CDKN1A) and downregulation of BCL2L11/BIM and p21 [7]. In addition to promoting cell cycle progression and inhibiting apoptosis, the miR-17-92 cluster also downregulates thrombospondin-1 (Tsp1) and connective tissue growth factor (CTGF), important antiangiogenic proteins [7]. Similarly, microvesicles from colorectal cancer cells contain miR-1246 and TGF-β which are transferred to endothelial cells to silence promyelocytic leukemia protein (PML) and activate Smad 1/5/8 signaling promoting proliferation and migration [9]. Likewise, lung cancer cell line derived microvesicles contain miR-494, in response to hypoxia, which targets PTEN in the endothelial cells promoting angiogenesis through the Akt/eNOS pathway [10]. Lastly, exosomal miR-135b from multiple myeloma cells suppresses the HIF-1/FIH-1 pathway in endothelial cells, increasing angiogenesis [11]. A summary of the studies showing the functions of exosomal miRNAs in shaping the biology of the TME is provided in Table 1.
Table 1
Actions of exosomal miRNAs exchanged between cells of the TME
The most common target of anti-angiogenic therapy is VEGF, and not unsurprisingly, multiple miRNAs (including miR-9, miR-20b, miR-130, miR-150, and miR-497) promote angiogenesis through the induction of the VEGF pathway. The most studied of these is the up-regulation of miR-9 which has been linked to a poor prognosis in multiple tumor types, including breast cancer, non-small cell lung cancer, and melanoma [12]. The two oncogenes MYC and MYCN activate miR-9 and cause E-cadherin downregulation resulting in the upregulated transcription of VEGF [13]. In addition, miR-9 has been shown to upregulate the JAK-STAT pathway, supporting endothelial cell migration and tumor angiogenesis [13]. Both amplification of miR-20b and miR-130 as well as miR-497 suppression regulate VEGF through hypoxia inducible factor 1α (HIF-1α) supporting increased angiogenesis [14, 15, 16, 17]. …..
The pivotal discovery in 2012 by Mitra et al. laid the ground-work for our current knowledge on the interactions between tumor-derived miRNAs and fibroblasts. In combination, the down-regulation of miR-214 and miR-31 and the up-regulation of miR-155 trigger the reprogramming of quiescent fibroblasts to CAFs [32]. As expected, the reverse regulation of these miRNAs reduced the migration and invasion of co-cultured ovarian cancer cells [32]. While the pathway of miR-155’s involvement in CAF biology is still being elucidated, the pathways of miR-214 and miR-31 have been established. In endometrial cancer, miR-31 was found to target the homeobox gene SATB2, leading to enhanced tumor cell migration and invasion [33]. MiR-214 similarly has an inverse correlation with its chemokine target, C-C motif Ligand 5 (CCL5) [32]. CCL5 secretion has been associated with enhanced motility, invasion, and metastatic potential through NF-κB-mediated MMP9 activation and through generation and differentiation of myeloid-derived suppressor cells (MDSCs) [34, 35, 36]. Furthermore, miR-210 and miR-133b overexpression and miR-149 suppression have been subsequently found to independently trigger the conversion to CAFs, possibly through paracrine stimulation, and to additionally promote EMT in prostate and gastric cancer, respectively [37, 38,39]. MiR-210 additionally enlists monocytes and encourages angiogenesis [37]. …
Another function of CAFs is the destruction of the ECM and its remodeling with a tumor-supportive composition and structure which includes modulation of specific integrins and metalloproteinases as some of the most studied miRNA targets. The 23 matrix metalloproteinases (MMPs) are critical in the ECM degradation, disruption of the growth signal balance, resistance to apoptosis, establishment of a favorable metastatic niche, and promotion of angiogenesis [54]. As expected, miRNAs have been found to regulate the actions of MMPs, together working to promote cancer cell growth, invasiveness, and metastasis. In HCC, MMP2 and 9 expression is up-regulated by miR-21 via PTEN pathway downregulation. Similarly, in cholangiocarcinoma it was observed that reduced levels of miR-138 induced up-regulation of RhoC, leading to increased levels of the same two MMPs [55, 56]. ….
As has been shown throughout this review, miRNAs have an important and varied effect on human carcinogenesis by shaping the biology of the TME towards a more permissive pro-tumoral phenotype. The complex events leading to such an outcome are currently quite universally defined as the “educational” process of cancer cells on the surrounding TME. While the initial focus was on the direction from the cancer cell to the surrounding TME, increasingly interest is centered on the implications of a more dynamic bidirectional exchange of genetic information. MiRNAs represent only part of the cargo of the extracellular vesicles, but an increasing scientific literature points towards their pivotal role in creating the micro-environmental conditions for cancer cell growth and dissemination. The nearby future will have to address several questions still unanswered. First, it is absolutely necessary to clarify which miRNAs and to what extent they are involved in this process. The contradictory results of some studies can be explained by the differences in tumor-types and by different concentrations of miRNAs used for functional studies. Understanding whether different concentrations of the same miRNA elicit different target effects and therefore changes the biology of the TME, will represent a significant consideration in the development of this field. It is certainly very attractive (especially in an attempt to develop new and desperately needed better cancer biomarkers) to think that concentrations of miRNAs within the TME are reflected systemically in the circulating levels of that same miRNA, however this has not yet been irrefutably demonstrated. Moreover, the study of the paracrine interactions among different cell populations of the TME and their reciprocal effects has been limited to two, maximum three cell populations. This is still way too far from describing the complexity of the TME and only the development of new tridimensional models of the TME will be able to cast a more conclusive light on such complexity. Finally, the pharmacokinetics of miRNA-containing vesicles is in its infancy at best, and needs to be further developed if the goal is development of new therapies based on the use of exosomic miRNAs. Therefore, the future of miRNA research, particularly in its role in the TME, holds still a lot of questions that need answering. However, for these exact same reasons, this is an incredibly exciting time for research in this field. We can envision a not too far future in which these concerns will be satisfactorily addressed and our understanding of the role of miRNAs within the TME will allow us to use them as new therapeutic weapons to successfully improve the clinical outcome of cancer patients.
Researchers at the Walter and Eliza Hall Institute in Australia have discovered a new way to trigger cell death that could lead to drugs to treat cancer and autoimmune disease.
Programmed cell death (a.k.a. apoptosis) is a natural process that removes unwanted cells from the body. Failure of apoptosis can allow cancer cells to grow unchecked or immune cells to inappropriately attack the body.
The protein known as Bak is central to apoptosis. In healthy cells, Bak sits in an inert state but when a cell receives a signal to die, Bak transforms into a killer protein that destroys the cell.
Triggering the cancer-apoptosis trigger
Institute researchers Sweta Iyer, PhD, Ruth Kluck, PhD, and colleagues unexpectedly discovered that an antibody they had produced to study Bak actually bound to the Bak protein and triggered its activation. They hope to use this discovery to develop drugs that promote cell death.
The researchers used information about Bak’s three-dimensional structure to find out precisely how the antibody activated Bak. “It is well known that Bak can be activated by a class of proteins called ‘BH3-only proteins’ that bind to a groove on Bak. We were surprised to find that despite our antibody binding to a completely different site on Bak, it could still trigger activation,” Kluck said. “The advantage of our antibody is that it can’t be ‘mopped up’ and neutralized by pro-survival proteins in the cell, potentially reducing the chance of drug resistance occurring.”
Drugs that target this new activation site could be useful in combination with other therapies that promote cell death by mimicking the BH3-only proteins. The researchers are now working with collaborators to develop their antibody into a drug that can access Bak inside cells.
Their findings have just been published in the open-access journal Nature Communications. The research was supported by the National Health and Medical Research Council, the Australian Research Council, the Victorian State Government Operational Infrastructure Support Scheme, and the Victorian Life Science Computation Initiative.
Abstract of Identification of an activation site in Bak and mitochondrial Bax triggered by antibodies
During apoptosis, Bak and Bax are activated by BH3-only proteins binding to the α2–α5 hydrophobic groove; Bax is also activated via a rear pocket. Here we report that antibodies can directly activate Bak and mitochondrial Bax by binding to the α1–α2 loop. A monoclonal antibody (clone 7D10) binds close to α1 in non-activated Bak to induce conformational change, oligomerization, and cytochrome c release. Anti-FLAG antibodies also activate Bak containing a FLAG epitope close to α1. An antibody (clone 3C10) to the Bax α1–α2 loop activates mitochondrial Bax, but blocks translocation of cytosolic Bax. Tethers within Bak show that 7D10 binding directly extricates α1; a structural model of the 7D10 Fab bound to Bak reveals the formation of a cavity under α1. Our identification of the α1–α2 loop as an activation site in Bak paves the way to develop intrabodies or small molecules that directly and selectively regulate these proteins.
“Cure” is a word that’s dominated the rhetoric in the war on cancer for decades. But it’s a word that medical professionals tend to avoid. While the American Cancer Society reports that cancer treatment has improved markedly over the decades and the five-year survival rate is impressively high for many cancers, oncologists still refrain from declaring their cancer-free patients cured. Why?
Patients are declared cancer-free (also called complete remission) when there are no more signs of detectable disease.
However, minuscule clusters of cancer cells below the detection level can remain in a patient’s body after treatment. Moreover, such small clusters of straggler cells may undergo metastasis, where they escape from the initial tumor into the bloodstream and ultimately settle in a distant site, often a vital organ such as the lungs, liver or brain.
When a colony of these metastatic cells reaches a detectable size, the patient is diagnosed with recurrent metastatic cancer. About one in three breast cancer patients diagnosed with early-stage cancer later develop metastatic disease, usually within five years of initial remission.
By the time metastatic cancer becomes evident, it is much more difficult to treat than when it was originally diagnosed.
What if these metastatic cells could be detected earlier, before they established a “foothold” in a vital organ? Better yet, could these metastatic cancer cells be intercepted, preventing them them from lodging in a vital organ in the first place?
The implant is a tiny porous polymer disc (basically a miniature sponge, no larger than a pencil eraser) that can be inserted just under a patient’s skin. Implantation triggers the immune system’s “foreign body response,” and the implant starts to soak up immune cells that travel to it. If the implant can catch mobile immune cells, then why not mobile metastatic cancer cells?
We gave implants to mice specially bred to model metastatic breast cancer. When the mice had palpable tumors but no evidence of metastatic disease, the implant was removed and analyzed.
Cancer cells were indeed present in the implant, while the other organs (potential destinations for metastatic cells) still appeared clean. This means that the implant can be used to spot previously undetectable metastatic cancer before it takes hold in an organ.
For patients with cancer in remission, an implant that can detect tumor cells as they move through the body would be a diagnostic breakthrough. But having to remove it to see if it has captured any cancer cells is not the most convenient or pleasant detection method for human patients.
Detecting cancer cells with noninvasive imaging
There could be a way around this, though: a special imaging method under development at Northwestern University called Inverse Spectroscopic Optical Coherence Tomography (ISOCT). ISOCT detects molecular-level differences in the way cells in the body scatter light. And when we scan our implant with ISOCT, the light scatter pattern looks different when it’s full of normal cells than when cancer cells are present. In fact, the difference is apparent when even as few as 15 out of the hundreds of thousands of cells in the implant are cancer cells.
There’s a catch – ISOCT cannot penetrate deep into tissue. That means it is not a suitable imaging technology for finding metastatic cells buried deep in internal organs. However, when the cancer cell detection implant is located just under the skin, it may be possible to detect cancer cells trapped in it using ISOCT. This could offer an early warning sign that metastatic cells are on the move.
This early warning could prompt doctors to monitor their patients more closely or perform additional tests. Conversely, if no cells are detected in the implant, a patient still in remission could be spared from unneeded tests.
The ISOCT results show that noninvasive imaging of the implant is feasible. But it’s a method still under development, and thus it’s not widely available. To make scanning easier and more accessible, we’re working to adapt more ubiquitous imaging technologies like ultrasound to detect tiny quantities of tumor cells in the implant.
Besides providing a way to detect tiny numbers of cancer cells before they can form new tumors in other parts of the body, our implant offers an even more intriguing possibility: diverting metastatic cells away from vital organs, and sequestering them where they cannot cause any damage.
In our mouse studies, we found that metastatic cells got caught in the implant before they were apparent in vital organs. When metastatic cells eventually made their way into the organs, the mice with implants still had significantly fewer tumor cells in their organs than implant-free controls. Thus, the implant appears to provide a therapeutic benefit, most likely by taking the metastatic cells it catches out of the circulation, preventing them from lodging anywhere vital.
Interestingly, we have not seen cancer cells leave the implant once trapped, or form a secondary tumor in the implant. Ongoing work aims to learn why this is. Whether the cells can stay safely immobilized in the implant or if it would need to be removed periodically will be important questions to answer before the implant could be used in human patients.
What the future may hold
For now, our work aims to make the implant more effective at drawing and detecting cancer cells. Since we tested the implant with metastatic breast cancer cells, we also want to see if it will work on other types of cancer. Additionally, we’re studying the cells the implant traps, and learning how the implant interacts with the body as a whole. This basic research should give us insight into the process of metastasis and how to treat it.
In the future (and it might still be far off), we envision a world where recovering cancer patients can receive a detector implant to stand guard for disease recurrence and prevent it from happening. Perhaps the patient could even scan their implant at home with a smartphone and get treatment early, when the disease burden is low and the available therapies may be more effective. Better yet, perhaps the implant could continually divert all the cancer cells away from vital organs on its own, like Iron Man’s electromagnet that deflects shrapnel from his heart.
This solution is still not a “cure.” But it would transform a formidable disease that one out of three cancer survivors would otherwise ultimately die from into a condition with which they could easily live.
New PSA Test Examines Protein Structures to Detect Prostate Cancers
5/16/2016 by Cleveland Clinic
A promising new test is detecting prostate cancer more precisely than current tests, by identifying molecular changes in the prostate specific antigen (PSA) protein, according to Cleveland Clinic research presented today at the American Urological Association annual meeting.
The study – part of an ongoing multicenter prospective clinical trial – found that the IsoPSATM test can also differentiate between high-risk and low-risk disease, as well as benign conditions.
Although widely used, the current PSA test relies on detection strategies that have poor specificity for cancer – just 25 percent of men who have a prostate biopsy due to an elevated PSA level actually have prostate cancer, according to the National Cancer Institute – and an inability to determine the aggressiveness of the disease.
The IsoPSA test, however, identifies prostate cancer in a new way. Developed by Cleveland Clinic, in collaboration with Cleveland Diagnostics, Inc., IsoPSA identifies the molecular structural changes in protein biomarkers. It is able to detect cancer by identifying these structural changes, as opposed to current tests that simply measure the protein’s concentration in a patient’s blood.
“While the PSA test has undoubtedly been one of the most successful biomarkers in history, its limitations are well known. Even currently available prostate cancer diagnostic tests rely on biomarkers that can be affected by physiological factors unrelated to cancer,” said Eric Klein, M.D., chair of Cleveland Clinic’s Glickman Urological & Kidney Institute. “These study results show that using structural changes in PSA protein to detect cancer is more effective and can help prevent unneeded biopsies in low-risk patients.”
The clinical trial involves six healthcare institutions and 132 patients, to date. It examined the ability of IsoPSA to distinguish patients with and without biopsy-confirmed evidence of cancer. It also evaluated the test’s precision in differentiating patients with high-grade (Gleason = 7) cancer from those with low-grade (Gleason = 6) disease and benign findings after standard ultrasound-guided biopsy of the prostate.
Substituting the IsoPSA structure-based composite index for the standard PSA resulted in improvement in diagnostic accuracy. Compared with serum PSA testing, IsoPSA performed better in both sensitivity and specificity.
“We took an ‘out of the box’ approach that has shown success in detecting prostate cancer but also has the potential to address other clinically important questions such as clinical surveillance of patients after treatment,” said Mark Stovsky, M.D., staff member, Cleveland Clinic Glickman Urological & Kidney Institute’s Department of Urology. Stovsky has a leadership position (Chief Medical Officer) and investment interest in Cleveland Diagnostics, Inc. “In general, the clinical utility of prostate cancer early detection and screening tests is often limited by the fact that biomarker concentrations may be affected by physiological processes unrelated to cancer, such as inflammation, as well as the relative lack of specificity of these biomarkers to the cancer phenotype. In contrast, clinical research data suggests that the IsoPSA assay can interrogate the entire PSA isoform distribution as a single stand-alone diagnostic tool which can reliably identify structural changes in the PSA protein that correlate with the presence or absence and aggressiveness of prostate cancer.”
Point of Care, Highly Accurate Cervical Cancer Screening
Fifty-five million times a year, American women go to their gynecologist for a Pap Smear. After waiting a few weeks for the results, more than 3.5 million of them are called back to the physician for a follow up visualization of the cervix. Beyond the stress related to possibly having cancer, the women are then subjected to a colposcopic exam, and all too often, a painful biopsy. Then more stressful waiting for a final diagnosis from the pathologist.
Cervical cancer develops slowly, allowing for successful treatment, when identified on time. Regions with high screening compliancy have low mortality rates from this cancer. In the US, for instance, where screening rates are close to 90%, only 4,200 women die from cervical cancer, annually, or 2.6 women per 100,000. However, the screening process in the developed world is long, complicated and not optimized.
In developing regions however, cervical cancer is a leading cause of women death. Over 85% of the total deaths from this cancer are in developing countries. Regions suffering from low screening rates include not only Africa, India and China, but many Eastern European countries as well. According to an OECD report from 2014, the cervical cancer screening rates in Romania and Hungary are as low as 14.6% and 36.7% respectively. The mortality rates in these countries are high, 16 in 100,000 women in Romania and 7.7 in 100,000 in Hungary.
The current screening process for cervical cancer detection is long, beginning with a Pap or HPV test. Cytology results take weeks to receive. A positive result requires follow-up testing by colposcopy and often biopsy. In countries where there is little access to medical care, or where screening compliancy is low, the chances of successful detection via this multi-step process are small. Developing regions and non-compliant countries require a point of care diagnostic method, which eliminates the need for return visits.
Additional limitations to cervical cancer screening are the low sensitivity and specificity rates of Pap tests and the high false positive rates of HPV test, leading to unnecessary colposcopies. Both cytology and colposcopy testing are highly dependent on operator proficiency for accurate diagnosis.
Biop has developed a new technology for the optimization of this process, into one, three minute, painless optical scan. The vaginal probe uses advanced optical, imaging and non-imaging technologies to identify and classify epithelium based cancers and pre-cancerous lesions. The probe is inserted into the vaginal canal, and scans the entire cervix. The resulting images and optical signatures created from the light, and captured by the sensors, are analyzed by the proprietary algorithm. The result is two pictures, on the physician’s screen; a high resolution photograph of the patient’s cervix, immediately next to a hot/cold map indicating a precise classification and location of any diseased lesions.
Deep learning applied to drug discovery and repurposing
Deep neural networks for drug discovery (credit: Insilico Medicine, Inc.)
Scientists from Insilico Medicine, Inc. have trained deep neural networks (DNNs) to predict the potential therapeutic uses of 678 drugs, using gene-expression data obtained from high-throughput experiments on human cell lines from Broad Institute’s LINCS databases and NIH MeSH databases.
The supervised deep-learning drug-discovery engine used the properties of small molecules, transcriptional data, and literature to predict efficacy, toxicity, tissue-specificity, and heterogeneity of response.
“We used LINCS data from Broad Institute to determine the effects on cell lines before and after incubation with compounds, co-author and research scientist Polina Mamoshina explained to KurzweilIAI.
“We used gene expression data of total mRNA from cell lines extracted and measured before incubation with compound X and after incubation with compound X to identify the response on a molecular level. The goal is to understand how gene expression (the transcriptome) will change after drug uptake. It is a differential value, so we need a reference (molecular state before incubation) to compare.”
The research is described in a paper in the upcoming issue of the journal Molecular Pharmaceutics.
Helping pharmas accelerate R&D
Alex Zhavoronkov, PhD, Insilico Medicine CEO, who coordinated the study, said the initial goal of their research was to help pharmaceutical companies significantly accelerate their R&D and increase the number of approved drugs. “In the process we came up with more than 800 strong hypotheses in oncology, cardiovascular, metabolic, and CNS spaces and started basic validation,” he said.
The team measured the “differential signaling pathway activation score for a large number of pathways to reduce the dimensionality of the data while retaining biological relevance.” They then used those scores to train the deep neural networks.*
“This study is a proof of concept that DNNs can be used to annotate drugs using transcriptional response signatures, but we took this concept to the next level,” said Alex Aliper, president of research, Insilico Medicine, Inc., lead author of the study.
Via Pharma.AI, a newly formed subsidiary of Insilico Medicine, “we developed a pipeline for in silico drug discovery — which has the potential to substantially accelerate the preclinical stage for almost any therapeutic — and came up with a broad list of predictions, with multiple in silico validation steps that, if validated in vitro and in vivo, can almost double the number of drugs in clinical practice.”
Despite the commercial orientation of the companies, the authors agreed not to file for intellectual property on these methods and to publish the proof of concept.
Deep-learning age biomarkers
According to Mamoshina, earlier this month, Insilico Medicine scientists published the first deep-learned biomarker of human age — aiming to predict the health status of the patient — in a paper titled “Deep biomarkers of human aging: Application of deep neural networks to biomarker development” by Putin et al, in Aging; and an overview of recent advances in deep learning in a paper titled “Applications of Deep Learning in Biomedicine” by Mamoshina et al., also in Molecular Pharmaceutics.
Insilico Medicine is located in the Emerging Technology Centers at Johns Hopkins University in Baltimore, Maryland, in collaboration with Datalytic Solutions and Mind Research Network.
* In this study, scientists used the perturbation samples of 678 drugs across A549, MCF-7 and PC-3 cell lines from the Library of Integrated Network-Based Cellular Signatures (LINCS) project developed by the National Institutes of Health (NIH) and linked those to 12 therapeutic use categories derived from MeSH (Medical Subject Headings) developed and maintained by the National Library of Medicine (NLM) of the NIH.
To train the DNN, scientists utilized both gene level transcriptomic data and transcriptomic data processed using a pathway activation scoring algorithm, for a pooled dataset of samples perturbed with different concentrations of the drug for 6 and 24 hours. Cross-validation experiments showed that DNNs achieve 54.6% accuracy in correctly predicting one out of 12 therapeutic classes for each drug.
One peculiar finding of this experiment was that a large number of drugs misclassified by the DNNs had dual use, suggesting possible application of DNN confusion matrices in drug repurposing. FutureTechnologies Media Group | Video presentation Insilico medicine
Abstract of Deep learning applications for predicting pharmacological properties of drugs and drug repurposing using transcriptomic data
Deep learning is rapidly advancing many areas of science and technology with multiple success stories in image, text, voice and video recognition, robotics and autonomous driving. In this paper we demonstrate how deep neural networks (DNN) trained on large transcriptional response data sets can classify various drugs to therapeutic categories solely based on their transcriptional profiles. We used the perturbation samples of 678 drugs across A549, MCF-7 and PC-3 cell lines from the LINCS project and linked those to 12 therapeutic use categories derived from MeSH. To train the DNN, we utilized both gene level transcriptomic data and transcriptomic data processed using a pathway activation scoring algorithm, for a pooled dataset of samples perturbed with different concentrations of the drug for 6 and 24 hours. When applied to normalized gene expression data for “landmark genes,” DNN showed cross-validation mean F1 scores of 0.397, 0.285 and 0.234 on 3-, 5- and 12-category classification problems, respectively. At the pathway level DNN performed best with cross-validation mean F1 scores of 0.701, 0.596 and 0.546 on the same tasks. In both gene and pathway level classification, DNN convincingly outperformed support vector machine (SVM) model on every multiclass classification problem. For the first time we demonstrate a deep learning neural net trained on transcriptomic data to recognize pharmacological properties of multiple drugs across different biological systems and conditions. We also propose using deep neural net confusion matrices for drug repositioning. This work is a proof of principle for applying deep learning to drug discovery and development.
A novel nanoscale organic transistor-based biosensor that can detect molecules associated with neurodegenerative diseases and some types of cancer has been developed by researchers at the National Nanotechnology Laboratory (LNNano) in Brazil.
The transistor, mounted on a glass slide, contains the reduced form of the peptide glutathione (GSH), which reacts in a specific way when it comes into contact with the enzyme glutathione S-transferase (GST), linked to Parkinson’s, Alzheimer’s and breast cancer, among other diseases.
Sensitive water-gated copper phthalocyanine (CuPc) thin-film transistor (credit: Rafael Furlan de Oliveira et al./Organic Electronics)
“The device can detect such molecules even when they’re present at very low levels in the examined material, thanks to its nanometric sensitivity,” explained Carlos Cesar Bof Bufon, Head of LNNano’s Functional Devices & Systems Lab (DSF).
Bufon said the system can be adapted to detect other substances by replacing the analytes (detection compounds). The team is working on paper-based biosensors to further lower the cost, improve portability, and facilitate fabrication and disposal.
The research is published in the journal Organic Electronics.
Abstract of Water-gated phthalocyanine transistors: Operation and transduction of the peptide–enzyme interaction
The use of aqueous solutions as the gate medium is an attractive strategy to obtain high charge carrier density (1012 cm−2) and low operational voltages (<1 V) in organic transistors. Additionally, it provides a simple and favorable architecture to couple both ionic and electronic domains in a single device, which is crucial for the development of novel technologies in bioelectronics. Here, we demonstrate the operation of transistors containing copper phthalocyanine (CuPc) thin-films gated with water and discuss the charge dynamics at the CuPc/water interface. Without the need for complex multilayer patterning, or the use of surface treatments, water-gated CuPc transistors exhibited low threshold (100 ± 20 mV) and working voltages (<1 V) compared to conventional CuPc transistors, along with similar charge carrier mobilities (1.2 ± 0.2) x 10−3 cm2 V−1 s−1. Several device characteristics such as moderate switching speeds and hysteresis, associated with high capacitances at low frequencies upon bias application (3.4–12 μF cm−2), indicate the occurrence of interfacial ion doping. Finally, water-gated CuPc OTFTs were employed in the transduction of the biospecific interaction between tripeptide reduced glutathione (GSH) and glutathione S-transferase (GST) enzyme, taking advantage of the device sensitivity and multiparametricity.
The study offers understanding of potential therapeutic targets.
Building on data from The Cancer Genome Atlas (TCGA) project, a multi-institutional team of scientists have completed the first large-scale “proteogenomic” study of breast cancer, linking DNA mutations to protein signaling and helping pinpoint the genes that drive cancer. Conducted by members of the National Cancer Institute’s Clinical Proteomic Tumor Analysis Consortium (CPTAC), including Baylor College of Medicine, Broad Institute of MIT and Harvard, Fred Hutchinson Cancer Research Center, New York University Langone Medical Center, and Washington University School of Medicine, the study takes aim at proteins, the workhorses of the cell, and their modifications to better understand cancer.
Appearing in the Advance Online Publication of Nature, the study illustrates the power of integrating genomic and proteomic data to yield a more complete picture of cancer biology than either analysis could do alone. The effort produced a broad overview of the landscape of the proteome (all the proteins found in a cell) and the phosphoproteome (the sites at which proteins are tagged by phosphorylation, a chemical modification that drives communication in the cell) across a set of 77 breast cancer tumors that had been genomically characterized in the TCGA project. Although the TCGA produced an extensive catalog of somatic mutations found in cancer, the effects of many of those mutations on cellular functions or patients’ outcomes are unknown.
In addition, not all mutated genes are true “drivers” of cancer — some are merely “passenger” mutations that have little functional consequence. And some mutations are found within very large DNA regions that are deleted or present in extra copies, so winnowing the list of candidate genes by studying the activity of their protein products can help identify therapeutic targets. “We don’t fully understand how complex cancer genomes translate into the driving biology that causes relapse and mortality,” said Matthew Ellis, director of the Lester and Sue Smith Breast Center at Baylor College of Medicine and a senior author of the paper.
“These findings show that proteogenomic integration could one day prove to be a powerful clinical tool, allowing us to traverse the large knowledge gap between cancer genomics and clinical action.” In this study, the researchers at the Broad Institute analyzed breast tumors using accurate mass, high-resolution mass spectrometry, a technology that extends the coverage of the proteome far beyond the coverage that can be achieved by traditional antibody-based methods. This allowed them to scale their efforts and quantify more than 12,000 proteins and 33,000 phosphosites, an extremely deep level of coverage.
Scripps scientists have designed a drug candidate that decreases growth of breast cancer cells.
In a development that could lead to a new generation of drugs to precisely treat a range of diseases, scientists from the Florida campus of The Scripps Research Institute (TSRI) have for the first time designed a drug candidate that decreases the growth of tumor cells in animal models in one of the hardest to treat cancers—triple negative breast cancer.
“This is the first example of taking a genetic sequence and designing a drug candidate that works effectively in an animal model against triple negative breast cancer,” said TSRI Professor Matthew Disney. “The study represents a clear breakthrough in precision medicine, as this molecule only kills the cancer cells that express the cancer-causing gene—not healthy cells. These studies may transform the way the lead drugs are identified—by using the genetic makeup of a disease.”
The study, published by the journal Proceedings of the National Academy of Sciences, demonstrates that the Disney lab’s compound, known as Targaprimir-96, triggers breast cancer cells to kill themselves via programmed cell death by precisely targeting a specific RNA that ignites the cancer.
Short-Cut to Drug Candidates
While the goal of precision medicine is to identify drugs that selectively affect disease-causing biomolecules, the process has typically involved time-consuming and expensive high-throughput screens to test millions of potential drug candidates to identify those few that affect the target of interest. Disney’s approach eliminates these screens.
The new study uses the lab’s computational approach called Inforna, which focuses on developing designer compounds that bind to RNA folds, particularly microRNAs.
MicroRNAs are short molecules that work within all animal and plant cells, typically functioning as a “dimmer switch” for one or more genes, binding to the transcripts of those genes and preventing protein production. Some microRNAs have been associated with diseases. For example, microRNA-96, which was the target of the new study, promotes cancer by discouraging programmed cell death, which can rid the body of cells that grow out of control.
In the new study, the drug candidate was tested in animal models over a 21-day course of treatment. Results showed decreased production of microRNA-96 and increased programmed cell death, significantly reducing tumor growth. Since targaprimir-96 was highly selective in its targeting, healthy cells were unaffected.
In contrast, Disney noted, a typical cancer therapeutic targets and kills cells indiscriminately, often leading to side effects that can make these drugs difficult for patients to tolerate.
Benjamin Zealley and Aubrey D.N.J. de Grey Commentary on Some Recent Theses Relevant to Combating Aging: June 2015
Cancer Autoantibody Biomarker Discovery and Validation Using Nucleic Acid Programmable Protein Array Jie Wang, PhD, Arizona State University
Currently in the United States, many patients with cancer do not benefit from population-based screening due to challenges associated with the existing cancer screening scheme. Blood-based diagnostic assays have the potential to detect diseases in a non-invasive way. Proteins released from small early tumors may only be present intermittently and are diluted to tiny concentrations in the blood, making them difficult to use as biomarkers. However, they can induce autoantibody (AAb) responses, which can amplify the signal and persist in the blood even if the antigen is gone. Circulating autoantibodies are a promising class of molecules that have the potential to serve as early detection biomarkers for cancers. This PhD thesis aims to screen for autoantibody biomarkers for the early detection of two deadly cancers, basal-like breast cancer and lung adenocarcinoma. First, a method was developed to display proteins in both native and denatured conformations on a protein array. This method adopted a novel protein tag technology, called a HaloTag, to immobilize proteins covalently on the surface of a glass slide. The covalent attachment allowed these proteins to endure harsh treatment without becoming dissociated from the slide surface, which enabled the profiling of antibody responses against both conformational and linear epitopes. Next, a plasma screening protocol was optimized to increase significantly the signal-to-noise ratio of protein array–based AAb detection. Following this, the AAb responses in basal-like breast cancer were explored using nucleic acid programmable protein arrays (NAPPA) containing 10,000 full-length human proteins in 45 cases and 45 controls. After verification in a large sample set (145 basal-like breast cancer cases, 145 controls, 70 non-basal breast cancer) by enzyme-linked immunosorbent assay (ELISA), a 13-AAb classifier was developed to differentiate patients from controls with a sensitivity of 33% at 98% specificity. A similar approach was also applied to the lung cancer study to identify AAbs that distinguished lung cancer patients from computed tomography–positive benign pulmonary nodules (137 lung cancer cases, 127 smoker controls, 170 benign controls). In this study, two panels of AAbs were discovered that showed promising sensitivity and specificity. Six out of eight AAb targets were also found to have elevated mRNA levels in lung adenocarcinoma patients using TCGA data. These projects as a whole provide novel insights into the association between AAbs and cancer, as well as general B cell antigenicity against self-proteins.
Comment: There are two widely supported models for cancer development and progression—the clonal evolution (CE) model and the cancer stem cell (CSC) model. Briefly, the former claims that most or all cells in a tumor contribute to its maintenance; as newer and more aggressive clones develop by random mutation, they become responsible for driving growth. The range of different mutational profiles generated is assumed to be large enough to account for disease recurrence after therapy (due to rare resistant clones) and metastasis (clones arising with the ability to travel to distant sites). The CSC model instead asserts that a small number of mutated stem cells are the origin of the primary cell mass, drive metastasis through the intermittent release of undifferentiated, highly mobile progeny, and account for recurrence due to a generally quiescent metabolic profile conferring potent resistance to chemotherapy. In either case, the immunological visibility of an early tumor may be highly sporadic. Clones arising early in CE differ little in proteomic terms from healthy host cells; those that do trigger a response are unlikely to have acquired robust resistance to immune attack, so are destroyed quickly in favor of their stealthier brethren. Likewise, CSCs share some of the immune privilege of normal stem cells and, due to their inherent ability to produce differentiated progeny with distinct proteomic signatures, are partially protected from attacks on their descendants. Consequently, such well-hidden cells may remain in the body for years to decades. The autoantibody panel developed in this study for basal-like breast cancer exhibits exceptional specificity despite a comparatively small training set. Given its ease of application, this suggests great promise for a more exhaustively trained classifier as a populationlevel screening tool.
Condition-Specific Differential Sub-Network Analysis for Biological Systems Deepali Jhamb, PhD, Indiana University
Biological systems behave differently under different conditions. Advances in sequencing technology over the last decade have led to the generation of enormous amounts of condition-specific data. However, these measurements often fail to identify low-abundance genes and proteins that can be biologically crucial. In this work, a novel textmining system was first developed to extract condition-specific proteins from the biomedical literature. The literaturederived data was then combined with proteomics data to construct condition-specific protein interaction networks. Furthermore, an innovative condition-specific differential analysis approach was designed to identify key differences, in the form of sub-networks, between any two given biological systems. The framework developed here was implemented to understand the differences between limb regenerationcompetent Ambystoma mexicanum and regeneration-deficient Xenopus laevis. This study provides an exhaustive systems-level analysis to compare regeneration competent and deficient sub-networks to show how different molecular entities inter-connect with each other and are rewired during the formation of an accumulation blastema in regenerating axolotl limbs. This study also demonstrates the importance of literature-derived knowledge, specific to limb regeneration, to augment the systems biology analysis. Our findings show that although the proteins might be common between the two given biological conditions, they can have a high dissimilarity based on their biological and topological properties in the sub-network. The knowledge gained from the distinguishing features of limb regeneration in amphibians can be used in future to induce regeneration chemically in mammalian systems. The approach developed in this dissertation is scalable and adaptable to understanding differential sub-networks between any two biological systems. This methodology will not only facilitate the understanding of biological processes and molecular functions that govern a given system, but will also provide novel intuitions about the pathophysiology of diseases/conditions.
Comment: We have long advocated a principle of directly comparing young and old bodies as a means to identify the classes of physical damage that accumulate in the body during aging. This approach circumvents our ignorance of the full etiology of each particular disease manifestation, a phenomenally difficult question given the ethical issues of experimenting on human subjects, the lengthy ‘‘incubation time’’ of aging-related diseases, and the complex interconnections between their risk factors—innate and environmental. Repairing such damage has the potential to prevent pathology before symptoms appear, an approach now becoming increasingly mainstream.11 However, a naı¨ve comparison faces a number of difficulties, even given a sufficiently large sample set to compensate for inter-individual variation. Most importantly, the causal significance of a given species cannot be reliably determined from its simple prevalence.12 The catalytic nature of cell biology means that those entities whose abundance changes the most profoundly in absolute terms are quite unlikely to be the drivers of that change and may even spontaneously revert to baseline levels in the absence of on-going stimulation. Meanwhile, functionality is often heavily influenced independently of abundance by post-translational modifications that may escape direct detection. Sub-network analysis uses computational means to identify groups of genes and/or proteins that vary in a synchronized way with some parameter, indicating functional connectivity. The application of methods such as those developed here to the comparison of a wide range of younger and older conditions will facilitate the identification of processes—not merely individual factors—that are impaired with age, and thus will help greatly in identifying the optimal points for intervention.
Development of a Light Actuated Drug Delivery-on-Demand System Chase Linsley, PhD, University of California, Los Angeles
The need for temporal–spatial control over the release of biologically active molecules has motivated efforts to engineer novel drug delivery-on-demand strategies actuated via light irradiation. Many systems, however, have been limited to in vitro proof-of-concept due to biocompatibility issues with the photo-responsive moieties or the light wavelength, intensity, and duration. To overcome these limitations, the objective of this dissertation was to design a light-actuated drug delivery-on-demand strategy that uses biocompatible chromophores and safe wavelengths of light, thereby advancing the clinical prospects of light-actuated drug delivery-on-demand systems. This was achieved by: (1) Characterizing the photothermal response of biocompatible visible light and near-infrared-responsive chromophores and demonstrating the feasibility and functionality of the light actuated on-demand drug delivery system in vitro; and (2) designing a modular drug delivery-on-demand system that could control the release of biologically active molecules over an extended period of time. Three biocompatible chromophores—Cardiogreen, Methylene Blue, and riboflavin—were identified and demonstrated significant photothermal response upon exposure to near-infrared and visible light, and the amount of temperature change was dependent upon light intensity, wavelength, as well as chromophore concentration. As a proof-of-concept, pulsatile release of a model protein from a thermally responsive delivery vehicle fabricated from poly(N-isopropylacrylamide) was achieved over 4 days by loading the delivery vehicle with Cardiogreen and irradiating with near-infrared light. To extend the useful lifetime of the light-actuated drug delivery-on-demand system, a modular, reservoir-valve system was designed. Using poly(ethylene glycol) as a reservoir for model small molecule drugs combined with a poly(N-isopropylacrylamide) valve spiked with chromophore-loaded liposomes, pulsatile release was achieved over 7 days upon light irradiation. Ultimately, this drug delivery strategy has potential for clinical applications that require explicit control over the presentation of biologically active molecules. Further research into the design and fabrication of novel biocompatible thermally responsive delivery vehicles will aid in the advancement of the light-actuated drug delivery-on-demand strategy described here. Comment: Our combined comments on this thesis and the next one appear after the next abstract.
Light-Inducible Gene Regulation in Mammalian Cells Lauren Toth, PhD, Duke University
The growing complexity of scientific research demands further development of advanced gene regulation systems. For instance, the ultimate goal of tissue engineering is to develop constructs that functionally and morphologically resemble the native tissue they are expected to replace. This requires patterning of gene expression and control of cellular phenotype within the tissue-engineered construct. In the field of synthetic biology, gene circuits are engineered to elucidate mechanisms of gene regulation and predict the behavior of more complex systems. Such systems require robust gene switches that can quickly turn gene expression on or off. Similarly, basic science requires precise genetic control to perturb genetic pathways or understand gene function. Additionally, gene therapy strives to replace or repair genes that are responsible for disease. The safety and efficacy of such therapies require control of when and where the delivered gene is expressed in vivo.
Unfortunately, these fields are limited by the lack of gene regulation systems that enable both robust and flexible cellular control. Most current gene regulation systems do not allow for the manipulation of gene expression that is spatially defined, temporally controlled, reversible, and repeatable. Rather, they provide incomplete control that forces the user to choose to control gene expression in either space or time, and whether the system will be reversible or irreversible. The recent emergence of the field of optogenetics—the ability to control gene expression using light—has made it possible to regulate gene expression with spatial, temporal, and dynamic control. Light-inducible systems provide the tools necessary to overcome the limitations of other gene regulation systems, which can be slow, imprecise, or cumbersome to work with. However, emerging light-inducible systems require further optimization to increase their efficiency, reliability, and ease of use.
Initially, we engineered a light-inducible gene regulation system that combines zinc finger protein technology and the light-inducible interaction between Arabidopsis thaliana plant proteins GIGANTEA (GI) and the light oxygen voltage (LOV) domain of FKF1. Zinc finger proteins (ZFPs) can be engineered to target almost any DNA sequence through tandem assembly of individual zinc finger domains that recognize a specific 3-bp DNA sequence. Fusion of three different ZFPs to GI (GI-ZFP) successfully targeted the fusion protein to the specific DNA target sequence of the ZFP. Due to the interaction between GI and LOV, co-expression of GI-ZFP with a fusion protein consisting of LOV fused to three copies of the VP16 transactivation domain (LOV-VP16) enabled blue-light dependent recruitment of LOV-VP16 to the ZFP target sequence. We showed that placement of three to nine copies of a ZFP target sequence upstream of a luciferase or enhanced green fluorescent protein (eGFP) transgene enabled expression of the transgene in response to blue light. Gene activation was both reversible and tunable on the basis of duration of light exposure, illumination intensity, and the number of ZFP binding sites upstream of the transgene. Gene expression could also be patterned spatially by illuminating the cell culture through photomasks containing various patterns.
Although this system was useful for controlling the expression of a transgene, for many applications it is useful to control the expression of a gene in its natural chromosomal position. Therefore, we capitalized on recent advances in programmed gene activation to engineer an optogenetic tool that could easily be targeted to new, endogenous DNA sequences without re-engineering the light inducible proteins. This approach took advantage of CRISPR/Cas9 technology, which uses a gene-specific guide RNA (gRNA) to facilitate Cas9 targeting and binding to a desired sequence, and the light-inducible heterodimerizers CRY2 and CIB1 from Arabidopsis thaliana to engineer a lightactivated CRISPR/Cas9 effector (LACE) system. We fused the full-length (FL) CRY2 to the transcriptional activator VP64 (CRY2FL-VP64) and the amino-terminal fragment of CIB1 to the amino, carboxyl, or amino and carboxyl terminus of a catalytically inactive Cas9. When CRY2-VP64 and one of the CIBN/dCas9 fusion proteins are expressed with a gRNA, the CIBN/dCas9 fusion protein localizes to the gRNA target. In the presence of blue light, CRY2FL binds to CIBN, which translocates CRY2FL-VP64 to the gene target and activates transcription. Unlike other optogenetic systems, the LACE system can be targeted to new endogenous loci by solely manipulating the specificity of the gRNA without having to re-engineer the light-inducible proteins. We achieved light-dependent activation of the IL1RN, HBG1/2, or ASCL1 genes by delivery of the LACE system and four gene-specific gRNAs per promoter region. For some gene targets, we achieved equivalent activation levels to cells that were transfected with the same gRNAs and the synthetic transcription factor dCas9-VP64. Gene activation was also shown to be reversible and repeatable through modulation of the duration of blue light exposure, and spatial patterning of gene expression was achieved using an eGFP reporter and a photomask.
Finally, we engineered a light-activated genetic ‘‘on’’ switch (LAGOS) that provides permanent gene expression in response to an initial dose of blue light illumination. LAGOS is a lentiviral vector that expresses a transgene only upon Cre recombinase–mediated DNA recombination. We showed that this vector, when used in conjunction with a light-inducible Cre recombinase system, could be used to express MyoD or the synthetic transcription factor VP64- MyoD in response to light in multiple mammalian cell lines, including primary mouse embryonic fibroblasts. We achieved light-mediated up-regulation of downstream myogenic markers myogenin, desmin, troponin T, and myosin heavy chains I and II as well as fusion of C3H10T1/2 cells into myotubes that resembled a skeletal muscle cell phenotype. We also demonstrated LAGOS functionality in vivo by engineering the vector to express human VEGF165 and human ANG1 in response to light. HEK 293T cells stably expressing the LAGOS vector and transiently expressing the light-inducible Cre recombinase proteins were implanted into mouse dorsal window chambers. Mice that were illuminated with blue light had increased micro-vessel density compared to mice that were not illuminated. Analysis of human vascular endothelial growth factor (VEGF) and human ANG1 levels by enzyme-linked immunosorbent assay (ELISA) revealed statistically higher levels of VEGF and ANG1 in illuminated mice compared to non-illuminated mice.
In summary, the objective of this work was to engineer robust light-inducible gene regulation systems that can control genes and cellular fate in a spatial and temporal manner. These studies combine the rapid advances in gene targeting and activation technology with natural light-inducible plant protein interactions. Collectively, this thesis presents several optogenetic systems that are expected to facilitate the development of multicellular cell and tissue constructs for use in tissue engineering, synthetic biology, gene therapy, and basic science both in vitro and in vivo.
Comment: Although it is easy to characterize technological progress as following in the wake of scientific discoveries, the reverse is almost equally true; advances in technique open the door to types of experiment previously intractable or impossible. Such is currently the case for the field of optically controlled biotechnology, which has exploded into prominence, particularly over the last half-decade. Light of an appropriate wavelength can penetrate mammalian tissues to a depth of up to a couple of centimeters, rendering much of the living body accessible to optical study and control—still more if the detector/source is integrated into an endoscopic or fiber optic probe. Techniques borrowed from the semiconductor industry allow patterns of illumination to be controlled down to the nanometer scale, ideal for addressing individual cells. The highly controlled time course of such experiments, as compared to traditional means of gene activation, such as the addition of a chemical agent to the medium, eliminates confounding variables, and simplifies data analysis. Furthermore, this level of immediate control opens the door to closed-loop systems where the activity of entities under optical control can be continuously tuned in relation to some parameter(s). In the first of these two illuminating theses, a vehicle is developed that permits light-driven release of a small molecule. Such a system could be employed to target a systemically administered antibiotic or anti-neoplastic agent to a site of infection or cancer while sparing other bodily tissues from toxicity. Because most modern drugs cannot be produced in the body, even given arbitrarily good control of cellular biochemistry, this technique will have lasting value in numerous clinical contexts. In the second thesis, the level of precision achieved is even more profound; the CRISPR/Cas9 system has received much recent attention13 in its own right for its capacity to target arbitrary genetic sequences without an arduous protein-engineering step. The LACE system described stands to permit genetic manipulation with almost arbitrarily good spatial, temporal, and genomic site-specific control, using only means available to a typical university laboratory.
Targeting T Cells for the Immune-Modulation of Human Diseases Regina Lin, PhD, Duke University
Dysregulated inflammation underlies the pathogenesis of a myriad of human diseases ranging from cancer to autoimmunity. As coordinators, executers, and sentinels of host immunity, T cells represent a compelling target population for immune-modulation. In fact, the antigen-specificity, cytotoxicity, and promise of long-lived of immune-protection make T cells ideal vehicles for cancer immunotherapy. Interventions for autoimmune disorders, on the other hand, aim to dampen T cell–mediated inflammation and promote their regulatory functions. Although significant strides have been made in targeting T cells for immune modulation, current approaches remain less than ideal and leave room for improvement. In this dissertation, I seek to improve on current T cell-targeted immunotherapies, by identifying and pre-clinically characterizing their mechanisms of action and in vivo therapeutic efficacy.
CD8+ cytotoxic T cells have potent anti-tumor activity and therefore are leading candidates for use in cancer immunotherapy. The application of CD8+ T cells for clinical use has been limited by the susceptibility of ex vivo– expanded CD8+ T cells to become dysfunctional in response to immunosuppressive microenvironments. To enhance the efficacy of adoptive cell transfer therapy (ACT), we established a novel microRNA (miRNA)-targeting approach that augments CTL cytotoxicity and preserves immunocompetence. Specifically, we screened for miRNAs that modulate cytotoxicity and identified miR-23a as a strong functional repressor of the transcription factor Blimp-1, which promotes CTL cytotoxicity and effector cell differentiation. In a cohort of advanced lung cancer patients, miR- 23a was up-regulated in tumor-infiltrating CD8+ T cells, and its expression correlated with impaired anti-tumor potential of patient CD8+ T cells. We determined that tumor-derived transforming growth factor-b (TGF-b) directly suppresses CD8+ T cell immune function by elevating miR-23a and down-regulating Blimp-1. Functional blockade of miR-23a in human CD8+ T cells enhanced granzyme B expression; and in mice with established tumors, immunotherapy with just a small number of tumor-specific CD8+ T cells in which miR-23a was inhibited robustly hindered tumor progression. Together, our findings provide a miRNA-based strategy that subverts the immunosuppression of CD8+ T cells that is often observed during adoptive cell transfer tumor immunotherapy and identify a TGF-bmediated tumor immune-evasion pathway
Having established that miR-23a-inhibition can enhance the quality and functional resilience of anti-tumor CD8+ T cells, especially within the immune-suppressive tumor microenvironment, we went on to interrogate the translational applicability of this strategy in the context of chimeric antigen receptor (CAR)-modified CD8+ T cells. Although CAR T cells hold immense promise for ACT, CAR T cells are not completely curative due to their in vivo functional suppression by immune barriers—such as TGF-b—within the tumor microenvironment. Because TGF-b poses a substantial immune barrier in the tumor microenvironment, we sought to investigate whether inhibiting miR-23a in CAR T cells can confer immune competence to afford enhanced tumor clearance. To this end, we retrovirally transduced wild-type and miR-23a–deficient CD8+ T cells with the EGFRvIII-CAR, which targets the PepvIII tumorspecific epitope expressed by glioblastomas (GBM). Our in vitro studies demonstrated that while wild-type EGFRvIIICAR T cells were vulnerable to functional suppression by TGF-b, miR-23a abrogation rendered EGFRvIII-CAR T cells immune-resistant to TGF-b. Rigorous preclinical studies are currently underway to evaluate the efficacy of miR-23adeficient EGFRvIII-CAR T cells for GBM immunotherapy.
Last, we explored novel immune-suppressive therapies by the biological characterization of pharmacological agents that could target T cells. Although immune-suppressive drugs are classical therapies for a wide range of autoimmune diseases, they are accompanied by severe adverse effects. This motivated our search for novel immunesuppressive agents that are efficacious and lack undesirable side effects. To this end, we explored the potential utility of subglutinol A, a natural product isolated from the endophytic fungus Fusarium subglutinans. We showed that subglutinol A exerts multimodal immune-suppressive effects on activated T cells in vitro. Subglutinol A effectively blocked T cell proliferation and survival, while profoundly inhibiting pro-inflammatory interferon-c (IFN-c) and interleukin-17 (IL-17) production by fully differentiated effector Th1 and Th17 cells. Our data further revealed that subglutinol A might exert its anti-inflammatory effects by exacerbating mitochondrial damage in T cells, but not in innate immune cells or fibroblasts. Additionally, we demonstrated that subglutinol A significantly reduced lymphocytic infiltration into the footpad and ameliorated footpad swelling in the mouse model of Th1-driven delayed-type hypersensitivity. These results suggest the potential of subglutinol A as a novel therapeutic for inflammatory diseases.
Comment: Immunotherapy is among the most promising approaches to cancer treatment, having the specificity and scope to selectively target transformed cells wherever they may reside within the body and the potential to install a permanent defense against disease recurrence. By the time a typical cancer is clinically diagnosed, however, it has already found means to survive a prolonged period of potential immune attack. The mechanisms by which tumors evade immune surveillance are beginning to be elucidated,15,16 and include both direct suppression of effector cells and progressive editing of the host’s immune repertoire to disfavor future attack. It is inherently difficult to interfere with these defenses directly, due to the selection pressures in genetically heterogeneous neoplastic tissue. Much effort is thus being focused on methods for rendering therapeutically delivered immune cells resistant to their effects. The cytokine TGF-b is paradoxically known to function as both a tumor suppressor in healthy tissue and as a tumorderived species associated with multiple cancer-promoting activities, including enhanced immune evasion. This work identifies the pathway by which TGF-b compromises cytotoxic T cell function in the tumor microenvironment, and demonstrates an effective method for blocking this signal. In many clinical cases, however, editing of the patient’s immune repertoire has already removed or rendered anergic those immune cells able to recognize their cancer. Thus, the finding that blocking TGF-b signaling also appears to enhance the effectiveness of CAR-modified T cells— engineered with an antibody fragment targeting them with high affinity to a particular tumor-associated epitope—is a welcome addition to these already promising results.
Novel Fibonacci and non-Fibonacci structure in the sunflower: results of a citizen science experiment
Jonathan Swinton, Erinma Ochu, The MSI Turing’s Sunflower Consortium
This citizen science study evaluates the occurrence of Fibonacci structure in the spirals of sunflower (Helianthus annuus) seedheads. This phenomenon has competing biomathematical explanations, and our core premise is that observation of both Fibonacci and non-Fibonacci structure is informative for challenging such models. We collected data on 657 sunflowers. In our most reliable data subset, we evaluated 768 clockwise or anticlockwise parastichy numbers of which 565 were Fibonacci numbers, and a further 67 had Fibonacci structure of a predefined type. We also found more complex Fibonacci structures not previously reported in sunflowers. This is the third, and largest, study in the literature, although the first with explicit and independently checkable inclusion and analysis criteria and fully accessible data. This study systematically reports for the first time, to the best of our knowledge, seedheads without Fibonacci structure. Some of these are approximately Fibonacci, and we found in particular that parastichy numbers equal to one less than a Fibonacci number were present significantly more often than those one more than a Fibonacci number. An unexpected further result of this study was the existence of quasi-regular heads, in which no parastichy number could be definitively assigned.
Introduction
Fibonacci structure can be found in hundreds of different species of plants [1]. This has led to a variety of competing conceptual and mathematical models that have been developed to explain this phenomenon. It is not the purpose of this paper to survey these: reviews can be found in [1–4], with more recent work including [5–10]. Instead, we focus on providing empirical data useful for differentiating them.
These models are in some ways now very mathematically satisfying in that they can explain high Fibonacci numbers based on a small number of plausible assumptions, though they are not so satisfying to experimental scientists [11]. Despite an increasingly detailed molecular and biophysical understanding of plant organ positioning [12–14], the very parsimony and generality of the mathematical explanations make the generation and testing of experimental hypotheses difficult. There remains debate about the appropriate choice of mathematical models, and whether they need to be central to our understanding of the molecular developmental biology of the plant. While sunflowers provide easily the largest Fibonacci numbers in phyllotaxis, and thus, one might expect, some of the stronger constraints on any theory, there is a surprising lack of systematic data to support the debate. There have been only two large empirical studies of spirals in the capitulum, or head, of the sunflower: Weisse [15] and Schoute [16], which together counted 459 heads; Schoute found numbers from the main Fibonacci sequence 82% of the time and Weise 95%. The original motivation of this study was to add a third replication to these two historical studies of a widely discussed phenomenon. Much more recently, a study of a smaller sample of 21 seedheads was carried out by Couder [17], who specifically searched for non-Fibonacci examples, whereas Ryan et al. [18] studied the arrangement of seeds more closely in a small sample of Helianthus annuus and a sample of 33 of the related perennial H. tuberosus.
Neither the occurrence of Fibonacci structure nor the developmental biology leading to it are at all unique to sunflowers. As common in other species, the previous sunflower studies found not only Fibonacci numbers, but also the occasional occurrence of the double Fibonacci numbers, Lucas numbers and F4 numbers defined below [1]. It is worth pointing out the warning of Cooke [19] that numbers from these sequences make up all but three of the first 17 integers. This means that it is particularly valuable to look at specimens with large parastichy numbers, such as the sunflowers, where the prevalence of Fibonacci structure is at its most striking.
Neither Schoute nor Weisse reported their precise technique for assigning parastichy numbers to their samples, and it is noteworthy that neither author reported any observation of non-Fibonacci structure. One of the objectives of this study was to rigorously define Fibonacci structure in advance and to ensure that the assignment method, though inevitably subjective, was carefully documented.
This paper concentrates on the patterning of seeds towards the outer rim of sunflower seedheads. The number of ray florets (the ‘petals’, typically bright yellow) or the green bracts behind them tends to have a looser distribution around a Fibonacci number. In the only mass survey of these, Majumder & Chakravarti [20] counted ray florets on 1002 sunflower heads and found a distribution centred on 21.
This citizen science study evaluates the occurrence of Fibonacci structure in the spirals of sunflower (Helianthus annuus) seedheads. This phenomenon has competing biomathematical explanations, and our core premise is that observation of both Fibonacci and non-Fibonacci structure is informative for challenging such models. We collected data on 657 sunflowers. In our most reliable data subset, we evaluated 768 clockwise or anticlockwise parastichy numbers of which 565 were Fibonacci numbers, and a further 67 had Fibonacci structure of a predefined type. We also found more complex Fibonacci structures not previously reported in sunflowers. This is the third, and largest, study in the literature, although the first with explicit and independently checkable inclusion and analysis criteria and fully accessible data. This study systematically reports for the first time, to the best of our knowledge, seedheads without Fibonacci structure. Some of these are approximately Fibonacci, and we found in particular that parastichy numbers equal to one less than a Fibonacci number were present significantly more often than those one more than a Fibonacci number. An unexpected further result of this study was the existence of quasi-regular heads, in which no parastichy number could be definitively assigned.
Incorporation of irregularity into the mathematical models of phyllotaxis is relatively recent: [17] gave an example of a disordered pattern arising directly from the deterministic model while more recently the authors have begun to consider the effects of stochasticity [10,21]. Differentiating between these models will require data that go beyond capturing the relative prevalence of different types of Fibonacci structure, so this study was also designed to yield the first large-scale sample of disorder in the head of the sunflower.
The Fibonacci sequence is the sequence of integers 1,2,3,5,8,13,21,34,55,89,144… in which each member after the second is the sum of the two preceding. The Lucas sequence is the sequence of integers 1,3,4,7,11,18,29,47,76,123… obeying the same rule but with a different starting condition; the F4 sequence is similarly 1,4,5,9,14,23,37,60,97,…. The double Fibonacci sequence 2,4,6,10,16,26,42,68,110,… is double the Fibonacci sequence. We say that a parastichy number which is any of these numbers has Fibonacci structure. The sequencesF5=1,5,6,11,17,28,45,73,… and F8=1,8,9,17,26,43,69,112… also arise from the same rule, but as they had not been previously observed in sunflowers we did not include these in the pre-planned definition of Fibonacci structure for parsimony. One example of adjacent pairs from each of these sequences was, in fact, observed but both examples are classified as non-Fibonacci below. A parastichy number which is any of 12,20,33,54,88,143 is also not classed as having Fibonacci structure but is distinguished as a Fibonacci number minus one in some of the analyses, and similarly 14,22,35,56,90,145 as Fibonacci plus one.
When looking at a seedhead such as in figure 1 the eye naturally picks out at least one family of parastichies or spirals: in this case, there is a clockwise family highlighted in blue in the image on the right-hand side.
Figure 5 plots the individual pairs observed. On the reference line, the ratio of the numbers is equal to the golden ratio so departures from the line mark departures from Fibonacci structure, which are less evident in the more reliable photoreviewed dataset. It can be seen from table 3 that Fibonacci pairings dominate the dataset.
Observed pairings of Fibonacci types of clockwise and anticlockwise parastichy numbers. Other means any parastichy number which neither has Fibonacci structure nor is Fibonacci ±1. Of all the Fibonacci ±1/Fibonacci pairs, only sample 191, a (21,20) pair, was not close to an adjacent Fibonacci pair.
One typical example of a Fibonacci pair is shown in figure 6, with a double Fibonacci case infigure 1 and a Lucas one in figure 7. There was no photoreviewed example of an F4 pairing. The sole photoreviewed assignment of a parastichy number to the F4 sequence was the anticlockwise parastichy number 37 in sample 570, which was relatively disordered. The clockwise parastichy number was 55, lending support to the idea this may have been a perturbation of a (34,55) pattern. We also found adjacent members of higher-order Fibonacci series. Figures 8 and 9 each show well-ordered examples with parastichy counts found adjacent in the F5 and F8 series, respectively: neither of these have been previously reported in the sunflower.
Sunflower 095. An (89,55) example with 89 clockwise parastichies and 55 anticlockwise ones, extending right to the rim of the head. Because these are clear and unambiguous, the other parastichy families which are visible towards the centre are not counted here.
Sunflower 667. Anticlockwise parastichies only, showing competing parastichy families which are distinct but in some places overlapping.
Our core results are twofold. First, and unsurprisingly, Fibonacci numbers, and Fibonacci structure more generally, are commonly found in the patterns in the seedheads of sunflowers. Given the extent to which Fibonacci patterns have attracted pseudo-scientific attention [33], this substantial replication of limited previous studies needs no apology. We have also published, for the first time, examples of seedheads related to the F5 and F8 sequences but by themselves they do not add much to the evidence base. Our second core result, though, is a systematic survey of cases where Fibonacci structure, defined strictly or loosely, did not appear. Although not common, such cases do exist and should shed light on the underlying developmental mechanisms. This paper does not attempt to shed that light, but we highlight the observations that any convincing model should explain. First, the prevalence of Lucas numbers is higher than those of double Fibonacci numbers in all three large datasets in the literature, including ours, and there are sporadic appearances of F4, F5 and F8 sequences. Second, counts near to but not exactly equal to Fibonacci structure are also observable: we saw a parastichy count of 54 more often than the most common Lucas count of 47. Sometimes, ambiguity arises in the counting process as to whether an exact Fibonacci-structured number might be obtained instead, but there are sufficiently many unambiguous cases to be confident this is a genuine phenomenon. Third, among these approximately Fibonacci counts, those which are a Fibonacci number minus one are significantly more likely to be seen than a Fibonacci number plus one. Fourth, it is not uncommon for the parastichy families in a seedhead to have strong departures from rotational symmetry: this can have the effect of yielding parastichy numbers which have large departures from Fibonacci structure or which are completely uncountable. This is related to the appearance of competing parastichy families. Fifth, it is common for the parastichy count in one direction to be more orderly and less ambiguous than that in the other. Sixth, seedheads sometimes possess completely disordered regions which make the assignment of parastichy numbers impossible. Some of these observations are unsurprising, some can be challenged by different counting protocols, and some are likely to be easily explained by the mathematical properties of deformed lattices, but taken together they pose a challenge for further research.
It is in the nature of this crowd-sourced experiment with multiple data sources that it is much easier to show variability than it is to find correlates of that variability. We tried a number of cofactor analyses that found no significant effect of geography, growing conditions or seed type but if they do influence Fibonacci structure, they are likely to be much easier to detect in a single-experimenter setting.
We have been forced by our results to extend classifications of seedhead patterns beyond structured Fibonacci to approximate Fibonacci ones. Clearly, the more loose the definition of approximate Fibonacci, the easier it is to explain away departures from model predictions. Couder [17] found one case of a (54,87) pair that he interpreted as a triple Lucas pair 3×(18,29). While mathematically true, in the light of our data, it might be more compellingly be thought of as close to a (55,89) ideal than an exact triple Lucas one. Taken together, this need to accommodate non-exact patterns, the dominance of one less over one more than Fibonacci numbers, and the observation of overlapping parastichy families suggest that models that explicitly represent noisy developmental processes may be both necessary and testable for a full understanding of this fascinating phenomenon. In conclusion, this paper provides a testbed against which a new generation of mathematical models can and should be built.
Switching on genes, Volume 2 (Volume Two: Latest in Genomics Methodologies for Therapeutics: Gene Editing, NGS and BioInformatics, Simulations and the Genome Ontology), Part 1: Next Generation Sequencing (NGS)
Switching on genes
Curator: Larry H. Bernstein, MD, FCAP
LPBI
UPDATED 3/17/2020
Gene Expression Controls Revealed
Researchers have modelled every atom in a key part of the process for switching on genes, revealing a whole new area for potential drug targets.
Proteins are essential for processes that sustain life. They are created in cells through a process called gene expression, which uses instructions from stretches of DNA called genes to build proteins. Sometimes genes are faulty and create proteins that contain errors, preventing the cell from functioning properly. These lead to genetic diseases like cystic fibrosis and haemophilia.
Gene expression is controlled by molecules called transcription factors, which bind to the start of a gene sequence at its ‘basal machinery’ and tell it to switch on and start creating certain proteins.
The way transcription factors bind to the basal machinery is a ‘fuzzy’ process, meaning the exact sequence of events is unknown because the steps do not exist for long enough to be captured by traditional imaging techniques.
But now, by creating a computer simulation of all of the tens of thousands of atoms making up the process and modelling their movements in 50 million separate steps, researchers at Imperial College London have been able to determine the sequence of events that lead to genes being switched on.
DISRUPTING DETRIMENTAL GENES
The simulated process revealed ‘pockets’ in the gene basal machinery, which the transcription factors move in and out of during binding. Knowing how these structures fit together could lead to the design of molecules that interfere with or disrupt the process, potentially tackling diseases.
Lead researcher Dr Robert Weinzierl from Imperial’s Department of Life Sciences said: “For the first time, we can fill in the dynamic landscape of interaction between transcription factors and basal machinery. This is a central mechanism for gene expression – the interactions here determine whether a gene gets switched on and creates proteins.”
“Gene regulation is a completely new drug target that has previously been too challenging to explore,” added Dr Weinzierl. “This process influences biology on a really fundamental level, and could allow us to prevent the expression of detrimental genes.”
FASTER DRUG SCREENING
The researchers’ new technique predicts the movements of all the atoms in order to build up a picture of the structures involved changing every couple of femtoseconds – quadrillionths of a second. The results of the first trial of the technique are reported today in PLOS Computational Biology.
Dr Weinzierl has submitted a patent application for his computer-based approach to studying gene expression interactions. Using this, compounds could be screened for possible fit into the basal machinery pockets.
“With computer simulation, it becomes easy to identify candidate compounds that could target these interactions without the need to test them first in real life, cutting down the time required to sift for new drugs,” said Dr Weinzierl.
Steps that lead to genes being switched on revealed in atomic simulation
Researchers have modelled every atom in a key part of the process for switching on genes, revealing a whole new area for potential drug targets.
Proteins are essential for processes that sustain life. They are created in cells through a process called gene expression, which uses instructions from stretches of DNA called genes to build proteins. Sometimes genes are faulty and create proteins that contain errors, preventing the cell from functioning properly. These lead to genetic diseases like cystic fibrosis and haemophilia.
Gene regulation is a completely new drug target that has previously been too challenging to explore.
Gene expression is controlled by molecules called transcription factors, which bind to the start of a gene sequence at its ‘basal machinery’ and tell it to switch on and start creating certain proteins.
The way transcription factors bind to the basal machinery is a ‘fuzzy’ process, meaning the exact sequence of events is unknown because the steps do not exist for long enough to be captured by traditional imaging techniques.
But now, by creating a computer simulation of all of the tens of thousands of atoms making up the process and modelling their movements in 50 million separate steps, researchers at Imperial College London have been able to determine the sequence of events that lead to genes being switched on.
Transcriptional activation domains (ADs) are generally thought to be intrinsically unstructured, but capable of adopting limited secondary structure upon interaction with a coactivator surface. The indeterminate nature of this interface made it hitherto difficult to study structure/function relationships of such contacts. Here we used atomistic accelerated molecular dynamics (aMD) simulations to study the conformational changes of the GCN4 AD and variants thereof, either free in solution, or bound to the GAL11 coactivator surface. We show that the AD-coactivator interactions are highly dynamic while obeying distinct rules. The data provide insights into the constant and variable aspects of orientation of ADs relative to the coactivator, changes in secondary structure and energetic contributions stabilizing the various conformers at different time points. We also demonstrate that a prediction of α-helical propensity correlates directly with the experimentally measured transactivation potential of a large set of mutagenized ADs. The link between α-helical propensity and the stimulatory activity of ADs has fundamental practical and theoretical implications concerning the recruitment of ADs to coactivators.Author Summary
The regulated transcription of eukaryotic genes is governed by gene-specific transcription factors that contain activation domains to stimulate the expression of nearby genes. Activation domains are unable to take up a defined three-dimensional conformation. Nevertheless, as we demonstrate in our study, molecular dynamics simulations reveal that the key docking point of such domains (centered around several large hydrophobic amino acid sidechains) folds into fluctuating α-helical conformations. Analysis of published data shows that this tendency of adopting such local structures correlates directly with stimulation activity. We also investigate the interaction of these structurally unstable domains with a coactivator interaction partner. Computational simulations are ideally suited for analysing the rapidly changing, “fuzzy” interactions occurring between these protein partners. We gained new insights into the competitive nature of the key hydrophobic sidechains in binding to a pocket on the coactivator surface and documented for the first time the rapidly changing movements of an activation domain during these interactions.
The last decade has seen an incredible breakthrough in technologies that allow histones, transcription factors (TFs), and RNA polymerases to be precisely mapped throughout the genome. From this research, it is clear that there is a complex interaction between the chromatin landscape and the general transcriptional machinery and that the dynamic control of this interface is central to gene regulation. However, the chromatin remodeling enzymes and general TFs cannot, on their own, recognize and stably bind to promoter or enhancer regions. Rather, they are recruited to cis regulatory regions through interaction with site-specific DNA binding TFs and/or proteins that recognize epigenetic marks such as methylated cytosines or specifically modified amino acids in histones. These “recruitment” factors are modular in structure, reflecting their ability to interact with the genome via one region of the protein and to simultaneously bind to other regulatory proteins via “effector” domains. In this chapter, we provide examples of common effector domains that can function in transcriptional regulation via their ability to (a) interact with the basal transcriptional machinery and general co-activators, (b) interact with other TFs to allow cooperative binding, and (c) directly or indirectly recruit histone and chromatin modifying enzymes.
Transcriptional activation is a stepwise process that requires (a) creating and maintaining an open chromatin structure, (b) assembly of the preinitiation complex, and (c) transition to productive elongation (Fig. 12.1). Successful completion of each of these steps involves a diverse group of proteins, some of which function in a relatively promoter-specific manner whereas others regulate large sets of genes. Recent advances in molecular and computational biology allow histone and DNA modifications, TFs, and RNA polymerases to be precisely mapped throughout the genome, relative to active or silent promoters (see [1–3] for reviews). From this research, it is becoming clear that there is a complex interaction between the chromatin landscape and the transcriptional machinery and that the dynamic relationship of this interface is central to biological control over gene expression [4]. It is now recognized that regulatory factors can exert their influence on transcriptional activation either via co-localization with other proteins that are bound at or near core promoter regions or they can be recruited to distal enhancer regions and interact with promoter-bound proteins via looping mechanisms. However, generally speaking, the chromatin remodeling enzymes and the general transcription factors involved in initiation and elongation cannot, on their own, recognize and stably bind to the promoter or enhancer regions.
Regulation of transcription. Shown is a schematic representing the three steps needed for productive transcription, including Step 1: the creation of open chromatin, which involves interactions between DNA-bound proteins and histone modifying enzymes …
One way in which chromatin remodeling enzymes and general transcription factors are recruited to cis-regulatory regions is through interaction with site-specific DNA binding TFs (Fig. 12.2a). The three largest classes of site-specific DNA binding proteins in mammals contact the genome via conserved DNA binding domains called zinc fingers, homeodomains, and helix–loop–helix domains [5] (Chapter 3 of this volume provides a catalog of eukaryotic DNA binding domains, and Chapters 4 and 5 specifically review C2H2 zinc fingers and homeodomains). Each of these classes of site-specific DNA binding factors contains many different proteins; for example, in humans there are over 650 zinc finger proteins, ~ 250 homeodomain proteins, and ~80 helix-loop-helix proteins [5]. Within each class, individual TFs can bind to and regulate hundreds to thousands of different genes. Site-specific TFs are modular in their structure reflecting their ability to bind to DNA via their DNA binding domains and simultaneously bind to other transcriptional regulatory proteins via so-called effector domains. The modular nature of site-specific TFs has been repeatedly demonstrated using in vitro and in vivo reporter assays. In these experiments, effector domains are separated from their natural DNA binding domains and then engineered to be part of a fusion protein having a heterologous DNA binding domain. Numerous studies have shown that simply bringing such effector domains to promoter regions can modulate transcription [6–8].
Modular structure of effector domain-containing proteins. Effector domains can be recruited to specific genomic regions via a DNA binding domains that recognize short DNA sequence motifs (TFBS), b recognition of a methylated cytosine (shown as a black…
Another way in which chromatin remodeling enzymes and general transcription factors can be brought to the genome is via effector domains that reside in proteins that can recognize epigenomic marks. Similar to recognition of a short nucleotide motif by a DNA binding protein, other proteins can distinguish distinctively modified DNA and histone protein “motifs”. For example, methylated cytosine in the 5′-CpG-3′ dinucleotide sequence is specifically recognized by members of a family of proteins containing a conserved methyl-CpG binding domain (MBD). MBD-containing proteins, which include MeCP2, MBD1, MBD2 and MBD4, bind specifically to methyl-CpG motifs located throughout the genome [9]; see Fig. 12.2b. MBD-containing proteins function by recruiting various co-regulators to methyl-CpG sites. For example, MeCP2 simultaneously binds promoter regions containing methyl-CpG motifs and the Sin3-containing histone deacetylase complex via a transcriptional repression domain (TRD), resulting in histone deacetylation and transcriptional silencing [10, 11]. Likewise, MBD1 and MBD2 copurify with distinct cellular complexes which link DNA methylation with chromatin modification and transcriptional repression. Similarly, posttranslational modifications of the amino termini of core histones are correlated to transcriptional states and are recognized by relevant chromatin-associated proteins (Fig. 12.2c). Several different histone modifications have been identified, including acetylation, phosphorylation, and methylation, and specific protein domains have evolved to recognize several of these different modifications. For example, different methylation states of histone H3 at lysine 4 can be recognized by tudor, chromo, and plant homeodomains (PHD), by malignant brain tumor (MBT) domains, and by WD40 repeat domains (many of these domains are structurally related and are collectively referred to as the “royal family” [12], reviewed [13, 14]). Other examples of this family include the chromodomain of HP1, which interacts with lower (mono- and di-) methylation states of lysine 9 of histone H3 but preferentially binds to the trimethylated state [15, 16] and the tudor domain of 53BP1, which can discriminate between the diand tri-methyl state of H4K20, preferring the dimethyl form [17, 18]. Acetylated lysine is also recognized by specific protein modules called the bromodomain [19], which is found in many chromatin-associated proteins and in nearly all known nuclear histone acetyltransferases (HATs). Of course, epigenetic marks such as DNA methylation and histone modifications are located at specific genomic regions (which can vary in different cell types), indicating that DNA methylases and histone modifying enzymes must be recruited to the genome by sequence-specific mechanisms such as site-specific TFs or RNAs. For example, KRAB-ZNFs can recruit the KAP1/SETDB1 histone methylating complex and long non-coding RNAs can recruit the PRC2 histone methylation complex [20–23].
The focus of this chapter is on the effector domains that are brought to specific sites of the genome by DNA binding proteins, methyl-CpG binding proteins, or histone binding proteins. (The interaction of TFs with chromatin more generally is discussed in Chapter 11). We provide examples of common effector domains that can function in transcriptional regulation via their ability to influence each of the steps outlined in Fig. 12.1. Specifically, we discuss effector domains that can: (a) interact with the basal transcriptional machinery and general co-activators, (b) interact with other TFs to allow cooperative binding, and (c) directly or indirectly recruit histone and chromatin modifying enzymes.
Eukaryotic transcriptional dynamics: from single molecules to cell populations
Transcriptional regulation is achieved through combinatorial interactions between regulatory elements in the human genome and a vast range of factors that modulate the recruitment and activity of RNA polymerase. Experimental approaches for studying transcription in vivo now extend from single-molecule techniques to genome-wide measurements. Parallel to these developments is the need for testable quantitative and predictive models for understanding gene regulation. These conceptual models must also provide insight into the dynamics of transcription and the variability that is observed at the single-cell level. In this Review, we discuss recent results on transcriptional regulation and also the models those results engender. We show how a non-equilibrium description informs our view of transcription by explicitly considering time-and energy-dependence at the molecular level.
Transcriptional regulation in the nucleus is the culmination of the actions of a diverse range of factors, such as transcription factors, chromatin remodellers, polymerases, helicases, topoisomerases, kinases, chaperones, proteasomes, acetyltransferases, deacetylases and methyltransferases. Determining how these molecules work in concert in the eukaryotic nucleus to regulate genes remains a central challenge in molecular biology. Dynamics lie at the heart of this mystery. Megadalton complexes assemble and disassemble on genes within seconds1,2; nucleosome turnover ranges from minutes to hours3; and gene activity demonstrates complex temporal patterns such as oscillation and transcriptional bursting4,5. Exciting new experimental advances have enabled the study of dynamic transcriptional regulation at the single-molecule6 and genome-wide7levels, thus enhancing our understanding of transcriptional regulation in vivo. These approaches also necessitate new models for describing gene expression. In this Review, we discuss recent in vivo results and the quantitative models that are motivated by those results.
Chromatin immunoprecipitation (ChIP) provides genome-wide occupancy profiles for chromatin-interacting factors at near base-pair resolution in populations of cells8,9. Using this approach on a genome-wide level has generated comprehensive maps of regulation on a gene-by-gene basis7,8,10. This population approach has been complemented by single-cell imaging techniques. Almost all factors that have been studied by live-cell microscopy exhibit dwell times on chromatin on the order of seconds11, and single-cell studies demonstrate a great variability in gene expression among cells in a population, owing in part to the stochastic nature of transcription12. Despite these tremendous advances in understanding the behaviour of individual factors, both methods fall short of capturing the sequence of events that is required to activate or repress a gene in vivo. Ideally, the occupancy of many factors that are coincident on a single stretch of DNA would be measured to obtain a sense of the complexes and intermediates that assemble in vivo. However, this experimental challenge is a daunting one. Current re-ChIP (also known as sequential ChIP) experiments usually look at two factors4,13 but it would be necessary to look at an order of magnitude more factors to begin to capture the combinatorial complexity of transcriptional regulation in metazoans4,14–16.
The gulf between actual mechanisms of transcriptional regulation and experimental capabilities could be bridged by using quantitative models of transcription. Decades of biochemical, structural and genetic data have spawned multiple models of transcriptional regulation, several of which we discuss below (FIG. 1). Even though these views are not mutually exclusive and boundaries between them are not clear, they reflect fundamental differences regarding the mechanisms of the underlying molecular processes. Currently, most quantitative theoretical models describe transcriptional regulation as an equilibrium thermodynamic phenomenon — an assumption that allows model building without explicitly considering the dynamics. Here we explain how this description is fundamentally inconsistent with the canonical view of gene regulation based on a sequential, ordered recruitment of factors, which is an example of a non-equilibrium model. In the context of a non-equilibrium model, the transcriptional dynamics can exhibit a form of molecular memory so that the future behaviour of the system depends on its history. We will outline this gap between the molecular biologist’s canonical view of transcription and the quantitative approaches that are often used to describe it. We argue for a non-equilibrium view of transcriptional regulation that is informed and constrained by single-cell observations. With the ability to observe single transcription factors17 and single transcribing genes18 in living cells, new experimental and modelling possibilities are emerging for understanding transcription dynamics in vivo.
A model of how twist defects, stimulated by the remodeler ATPase, shift DNA around the nucleosome . From the Science article G.D. Bowman and S. Deindl. Remodeling the Genome. Science (2020); 166(6461:p35-36.
In complex organisms such as humans, a single genetic blueprint can give rise to a multitude of different cell types, from nerve to liver to muscle. Such cellular diversity relies on restricting which portions of genomic DNA are accessible and therefore can be read by cellular machinery. Ultimately, access to DNA depends on placement of a repetitive, spool-like structure called the nucleosome, the basic packaging unit of chromosomes. The nucleosome occludes two tight loops of DNA and thus represents a fundamentally repressive element. When and where nucleosomes are positioned can affect complex transcriptional programs, and therefore disruptions in the factors responsible for nucleosome positioning often result in cancers and multisystem developmental diseases. Although the mechanism of shifting nucleosomes along DNA has long proved elusive, a recent flurry of structural, biophysical, and biochemical work has revealed a core mechanistic framework explaining how nucleosomes are actively repositioned throughout the genome.
Nucleosomes are the most ubiquitous protein-DNA complexes in all eukaryotic cells. The core of each nucleosome is a symmetric, disk-like structure made of histone proteins that provides a scaffold around which two loops of the DNA helix are snugly wrapped (1). Histones are often modified through, for example, acetylation, methylation, and phosphorylation, which add an additional layer of information on top of the genetic code. This epigenetic information demarcates functionally distinct regions of the genome—for instance, whether a gene is active or designated to remain silent—for each cell type.
Owing to their extensive protein-DNA interface, nucleosomes are relatively stable structures. Active placement and reorganization of nucleosomes depend on chromatin remodelers. As the gatekeepers of nucleosome packaging, these enzymes participate both in activating and repressing gene expression. Remodelers can assemble, disassemble, and exchange histones within the nucleosome, as well as shift the position of the histone core along DNA. Acting on either face of the nucleosome disk, remodelers can move the histone core back and forth on DNA, changing which parts of DNA are exposed and which are wrapped up in the nucleosome. Increased exposure of DNA occurs when remodelers shift adjacent nucleosomes into each other, resulting in histone ejection (2).
The ability of remodelers to manipulate nucleosome structure stems from a highly conserved adenosine triphosphatase (ATPase) motor that belongs to a larger superfamily of helicase-like ATPases called superfamily 2 (SF2).
Consistent with an inchworm-type translocation mechanism, nucleosomal DNA is shifted by remodelers with an elementary step of a single base pair (4). Given the spiral structure of the duplex, nucleosomal DNA must shift around the histone core in a corkscrew fashion. Remodeler ATPases remain in a fixed location on the histone core during DNA translocation, which means that DNA all around the nucleosome must shift in response to localized action at the ATPase binding site. When poised for hydrolysis, the ATPase in its closed state eliminates this bulge, corkscrewing DNA toward the nucleosome midpoint (dyad) on the other side. This creation and elimination of a DNA bulge is equivalent to altering DNA twist.
These snapshots of remodelers altering DNA twist are complemented by the observation of active movement of nucleosomal DNA by remodelers during three-color single-molecule FRET (Förster resonance energy transfer) experiments, where DNA movements were coupled to ATP hydrolysis (13).
The authors suggest some questions that might direct future research. For example, in addition to DNA geometry and energetics, to what extent is twist diffusion dependent on other characteristics of the nucleosome?
Histone proteins come in a variety of flavors, and an important goal is identifying whether distinct biophysical properties of histone variants and other epigenetic signatures can determine or bias the outcomes of remodeling reactions.
Disease related changes in proteomics, protein folding, protein-protein interaction, Volume 2 (Volume Two: Latest in Genomics Methodologies for Therapeutics: Gene Editing, NGS and BioInformatics, Simulations and the Genome Ontology), Part 1: Next Generation Sequencing (NGS)
Disease related changes in proteomics, protein folding, protein-protein interaction
Curator: Larry H. Bernstein, MD, FCAP
LPBI
Frankenstein Proteins Stitched Together by Scientists
The Frankenstein monster, stitched together from disparate body parts, proved to be an abomination, but stitched together proteins may fare better. They may, for example, serve specific purposes in medicine, research, and industry. At least, that’s the ambition of scientists based at the University of North Carolina. They have developed a computational protocol called SEWING that builds new proteins from connected or disconnected pieces of existing structures. [Wikipedia]
Unlike Victor Frankenstein, who betrayed Promethean ambition when he sewed together his infamous creature, today’s biochemists are relatively modest. Rather than defy nature, they emulate it. For example, at the University of North Carolina (UNC), researchers have taken inspiration from natural evolutionary mechanisms to develop a technique called SEWING—Structure Extension With Native-substructure Graphs. SEWING is a computational protocol that describes how to stitch together new proteins from connected or disconnected pieces of existing structures.
“We can now begin to think about engineering proteins to do things that nothing else is capable of doing,” said UNC’s Brian Kuhlman, Ph.D. “The structure of a protein determines its function, so if we are going to learn how to design new functions, we have to learn how to design new structures. Our study is a critical step in that direction and provides tools for creating proteins that haven’t been seen before in nature.”
Traditionally, researchers have used computational protein design to recreate in the laboratory what already exists in the natural world. In recent years, their focus has shifted toward inventing novel proteins with new functionality. These design projects all start with a specific structural “blueprint” in mind, and as a result are limited. Dr. Kuhlman and his colleagues, however, believe that by removing the limitations of a predetermined blueprint and taking cues from evolution they can more easily create functional proteins.
Dr. Kuhlman’s UNC team developed a protein design approach that emulates natural mechanisms for shuffling tertiary structures such as pleats, coils, and furrows. Putting the approach into action, the UNC team mapped 50,000 stitched together proteins on the computer, and then it produced 21 promising structures in the laboratory. Details of this work appeared May 6 in the journal Science, in an article entitled, “Design of Structurally Distinct Proteins Using Strategies Inspired by Evolution.”
“Helical proteins designed with SEWING contain structural features absent from other de novo designed proteins and, in some cases, remain folded at more than 100°C,” wrote the authors. “High-resolution structures of the designed proteins CA01 and DA05R1 were solved by x-ray crystallography (2.2 angstrom resolution) and nuclear magnetic resonance, respectively, and there was excellent agreement with the design models.”
Essentially, the UNC scientists confirmed that the proteins they had synthesized contained the unique structural varieties that had been designed on the computer. The UNC scientists also determined that the structures they had created had new surface and pocket features. Such features, they noted, provide potential binding sites for ligands or macromolecules.
“We were excited that some had clefts or grooves on the surface, regions that naturally occurring proteins use for binding other proteins,” said the Science article’s first author, Tim M. Jacobs, Ph.D., a former graduate student in Dr. Kuhlman’s laboratory. “That’s important because if we wanted to create a protein that can act as a biosensor to detect a certain metabolite in the body, either for diagnostic or research purposes, it would need to have these grooves. Likewise, if we wanted to develop novel therapeutics, they would also need to attach to specific proteins.”
Currently, the UNC researchers are using SEWING to create proteins that can bind to several other proteins at a time. Many of the most important proteins are such multitaskers, including the blood protein hemoglobin.
Histone Mutation Deranges DNA Methylation to Cause Cancer
In some cancers, including chondroblastoma and a rare form of childhood sarcoma, a mutation in histone H3 reduces global levels of methylation (dark areas) in tumor cells but not in normal cells (arrowhead). The mutation locks the cells in a proliferative state to promote tumor development. [Laboratory of Chromatin Biology and Epigenetics at The Rockefeller University]
They have been called oncohistones, the mutated histones that are known to accompany certain pediatric cancers. Despite their suggestive moniker, oncohistones have kept their oncogenic secrets. For example, it has been unclear whether oncohistones are able to cause cancer on their own, or whether they need to act in concert with additional DNA mutations, that is, mutations other than those affecting histone structures.
While oncohistone mechanisms remain poorly understood, this particular question—the oncogenicity of lone oncohistones—has been resolved, at least in part. According to researchers based at The Rockefeller University, a change to the structure of a histone can trigger a tumor on its own.
This finding appeared May 13 in the journal Science, in an article entitled, “Histone H3K36 Mutations Promote Sarcomagenesis Through Altered Histone Methylation Landscape.” The article describes the Rockefeller team’s study of a histone protein called H3, which has been found in about 95% of samples of chondoblastoma, a benign tumor that arises in cartilage, typically during adolescence.
The Rockefeller scientists found that the H3 lysine 36–to–methionine (H3K36M) mutation impairs the differentiation of mesenchymal progenitor cells and generates undifferentiated sarcoma in vivo.
After the scientists inserted the H3 histone mutation into mouse mesenchymal progenitor cells (MPCs)—which generate cartilage, bone, and fat—they watched these cells lose the ability to differentiate in the lab. Next, the scientists injected the mutant cells into living mice, and the animals developed the tumors rich in MPCs, known as an undifferentiated sarcoma. Finally, the researchers tried to understand how the mutation causes the tumors to develop.
The scientists determined that H3K36M mutant nucleosomes inhibit the enzymatic activities of several H3K36 methyltransferases.
“Depleting H3K36 methyltransferases, or expressing an H3K36I mutant that similarly inhibits H3K36 methylation, is sufficient to phenocopy the H3K36M mutation,” the authors of the Science study wrote. “After the loss of H3K36 methylation, a genome-wide gain in H3K27 methylation leads to a redistribution of polycomb repressive complex 1 and de-repression of its target genes known to block mesenchymal differentiation.”
Essentially, when the H3K36M mutation occurs, the cell becomes locked in a proliferative state—meaning it divides constantly, leading to tumors. Specifically, the mutation inhibits enzymes that normally tag the histone with chemical groups known as methyls, allowing genes to be expressed normally.
In response to this lack of modification, another part of the histone becomes overmodified, or tagged with too many methyl groups. “This leads to an overall resetting of the landscape of chromatin, the complex of DNA and its associated factors, including histones,” explained co-author Peter Lewis, Ph.D., a professor at the University of Wisconsin-Madison and a former postdoctoral fellow in laboratory of C. David Allis, Ph.D., a professor at Rockefeller.
The finding—that a “resetting” of the chromatin landscape can lock the cell into a proliferative state—suggests that researchers should be on the hunt for more mutations in histones that might be driving tumors. For their part, the Rockefeller researchers are trying to learn more about how this specific mutation in histone H3 causes tumors to develop.
“We want to know which pathways cause the mesenchymal progenitor cells that carry the mutation to continue to divide, and not differentiate into the bone, fat, and cartilage cells they are destined to become,” said co-author Chao Lu, Ph.D., a postdoctoral fellow in the Allis lab.
Once researchers understand more about these pathways, added Dr. Lewis, they can consider ways of blocking them with drugs, particularly in tumors such as MPC-rich sarcomas—which, unlike chondroblastoma, can be deadly. In fact, drugs that block these pathways may already exist and may even be in use for other types of cancers.
“One long-term goal of our collaborative team is to better understand fundamental mechanisms that drive these processes, with the hope of providing new therapeutic approaches,” concluded Dr. Allis.
Histone H3K36 mutations promote sarcomagenesis through altered histone methylation landscape
Missense mutations (that change one amino acid for another) in histone H3 can produce a so-called oncohistone and are found in a number of pediatric cancers. For example, the lysine-36–to-methionine (K36M) mutation is seen in almost all chondroblastomas. Lu et al. show that K36M mutant histones are oncogenic, and they inhibit the normal methylation of this same residue in wild-type H3 histones. The mutant histones also interfere with the normal development of bone-related cells and the deposition of inhibitory chromatin marks.
Several types of pediatric cancers reportedly contain high-frequency missense mutations in histone H3, yet the underlying oncogenic mechanism remains poorly characterized. Here we report that the H3 lysine 36–to–methionine (H3K36M) mutation impairs the differentiation of mesenchymal progenitor cells and generates undifferentiated sarcoma in vivo. H3K36M mutant nucleosomes inhibit the enzymatic activities of several H3K36 methyltransferases. Depleting H3K36 methyltransferases, or expressing an H3K36I mutant that similarly inhibits H3K36 methylation, is sufficient to phenocopy the H3K36M mutation. After the loss of H3K36 methylation, a genome-wide gain in H3K27 methylation leads to a redistribution of polycomb repressive complex 1 and de-repression of its target genes known to block mesenchymal differentiation. Our findings are mirrored in human undifferentiated sarcomas in which novel K36M/I mutations in H3.1 are identified.
Mitochondria? We Don’t Need No Stinking Mitochondria!
Diagram comparing typical eukaryotic cell to the newly discovered mitochondria-free organism. [Karnkowska et al., 2016, Current Biology 26, 1–11]
The organelle that produces a significant portion of energy for eukaryotic cells would seemingly be indispensable, yet over the years, a number of organisms have been discovered that challenge that biological pretense. However, these so-called amitochondrial species may lack a defined organelle, but they still retain some residual functions of their mitochondria-containing brethren. Even the intestinal eukaryotic parasite Giardia intestinalis, which was for many years considered to be mitochondria-free, was proven recently to contain a considerably shriveled version of the organelle.
Now, an international group of scientists has released results from a new study that challenges the notion that mitochondria are essential for eukaryotes—discovering an organism that resides in the gut of chinchillas that contains absolutely no trace of mitochondria at all.
“In low-oxygen environments, eukaryotes often possess a reduced form of the mitochondrion, but it was believed that some of the mitochondrial functions are so essential that these organelles are indispensable for their life,” explained lead study author Anna Karnkowska, Ph.D., visiting scientist at the University of British Columbia in Vancouver. “We have characterized a eukaryotic microbe which indeed possesses no mitochondrion at all.”
Mysterious Eukaryote Missing Mitochondria
Researchers uncover the first example of a eukaryotic organism that lacks the organelles.
Monocercomonoides sp. PA203VLADIMIR HAMPL, CHARLES UNIVERSITY, PRAGUE, CZECH REPUBLIC
Scientists have long thought that mitochondria—organelles responsible for energy generation—are an essential and defining feature of a eukaryotic cell. Now, researchers from Charles University in Prague and their colleagues are challenging this notion with their discovery of a eukaryotic organism,Monocercomonoides species PA203, which lacks mitochondria. The team’s phylogenetic analysis, published today (May 12) in Current Biology,suggests that Monocercomonoides—which belong to the Oxymonadida group of protozoa and live in low-oxygen environments—did have mitochondria at one point, but eventually lost the organelles.
“This is quite a groundbreaking discovery,” said Thijs Ettema, who studies microbial genome evolution at Uppsala University in Sweden and was not involved in the work.
“This study shows that mitochondria are not so central for all lineages of living eukaryotes,” Toni Gabaldonof the Center for Genomic Regulation in Barcelona, Spain, who also was not involved in the work, wrote in an email to The Scientist. “Yet, this mitochondrial-devoid, single-cell eukaryote is as complex as other eukaryotic cells in almost any other aspect of cellular complexity.”
Charles University’s Vladimir Hampl studies the evolution of protists. Along with Anna Karnkowska and colleagues, Hampl decided to sequence the genome of Monocercomonoides, a little-studied protist that lives in the digestive tracts of vertebrates. The 75-megabase genome—the first of an oxymonad—did not contain any conserved genes found on mitochondrial genomes of other eukaryotes, the researchers found. It also did not contain any nuclear genes associated with mitochondrial functions.
“It was surprising and for a long time, we didn’t believe that the [mitochondria-associated genes were really not there]. We thought we were missing something,” Hampl told The Scientist. “But when the data kept accumulating, we switched to the hypothesis that this organism really didn’t have mitochondria.”
Because researchers have previously not found examples of eukaryotes without some form of mitochondria, the current theory of the origin of eukaryotes poses that the appearance of mitochondria was crucial to the identity of these organisms.
“We now view these mitochondria-like organelles as a continuum from full mitochondria to very small . Some anaerobic protists, for example, have only pared down versions of mitochondria, such as hydrogenosomes and mitosomes, which lack a mitochondrial genome. But these mitochondrion-like organelles perform essential functions of the iron-sulfur cluster assembly pathway, which is known to be conserved in virtually all eukaryotic organisms studied to date.
Yet, in their analysis, the researchers found no evidence of the presence of any components of this mitochondrial pathway.
Like the scaling down of mitochondria into mitosomes in some organisms, the ancestors of modernMonocercomonoides once had mitochondria. “Because this organism is phylogenetically nested among relatives that had conventional mitochondria, this is most likely a secondary adaptation,” said Michael Gray, a biochemist who studies mitochondria at Dalhousie University in Nova Scotia and was not involved in the study. According to Gray, the finding of a mitochondria-deficient eukaryote does not mean that the organelles did not play a major role in the evolution of eukaryotic cells.
To be sure they were not missing mitochondrial proteins, Hampl’s team also searched for potential mitochondrial protein homologs of other anaerobic species, and for signature sequences of a range of known mitochondrial proteins. While similar searches with other species uncovered a few mitochondrial proteins, the team’s analysis of Monocercomonoides came up empty.
“The data is very complete,” said Ettema. “It is difficult to prove the absence of something but [these authors] do a convincing job.”
To form the essential iron-sulfur clusters, the team discovered that Monocercomonoides use a sulfur mobilization system found in the cytosol, and that an ancestor of the organism acquired this system by lateral gene transfer from bacteria. This cytosolic, compensating system allowed Monocercomonoides to lose the otherwise essential iron-sulfur cluster-forming pathway in the mitochondrion, the team proposed.
“This work shows the great evolutionary plasticity of the eukaryotic cell,” said Karnkowska, who participated in the study while she was a postdoc at Charles University. Karnkowska, who is now a visiting researcher at the University of British Columbia in Canada, added: “This is a striking example of how far the evolution of a eukaryotic cell can go that was beyond our expectations.”
“The results highlight how many surprises may await us in the poorly studied eukaryotic phyla that live in under-explored environments,” Gabaldon said.
Ettema agreed. “Now that we’ve found one, we need to look at the bigger picture and see if there are other examples of eukaryotes that have lost their mitochondria, to understand how adaptable eukaryotes are.”
Karnkowska et al., “A eukaryote without a mitochondrial organelle,” Current Biology,doi:10.1016/j.cub.2016.03.053, 2016.
•Monocercomonoides sp. is a eukaryotic microorganism with no mitochondria
•The complete absence of mitochondria is a secondary loss, not an ancestral feature
•The essential mitochondrial ISC pathway was replaced by a bacterial SUF system
The presence of mitochondria and related organelles in every studied eukaryote supports the view that mitochondria are essential cellular components. Here, we report the genome sequence of a microbial eukaryote, the oxymonad Monocercomonoides sp., which revealed that this organism lacks all hallmark mitochondrial proteins. Crucially, the mitochondrial iron-sulfur cluster assembly pathway, thought to be conserved in virtually all eukaryotic cells, has been replaced by a cytosolic sulfur mobilization system (SUF) acquired by lateral gene transfer from bacteria. In the context of eukaryotic phylogeny, our data suggest that Monocercomonoides is not primitively amitochondrial but has lost the mitochondrion secondarily. This is the first example of a eukaryote lacking any form of a mitochondrion, demonstrating that this organelle is not absolutely essential for the viability of a eukaryotic cell.
This method catches a bait protein together with its associated protein partners in virus-like particles that are budded from human cells. Like this, cell lysis is not needed and protein complexes are preserved during purification.
With his feet in both a proteomics lab and an interactomics lab, VIB/UGent professor Sven Eyckerman is well aware of the shortcomings of conventional approaches to analyze protein complexes. The lysis conditions required in mass spectrometry–based strategies to break open cell membranes often affect protein-protein interactions. “The first step in a classical study on protein complexes essentially turns the highly organized cellular structure into a big messy soup”, Eyckerman explains.
Inspired by virus biology, Eyckerman came up with a creative solution. “We used the natural process of HIV particle formation to our benefit by hacking a completely safe form of the virus to abduct intact protein machines from the cell.” It is well known that the HIV virus captures a number of host proteins during its particle formation. By fusing a bait protein to the HIV-1 GAG protein, interaction partners become trapped within virus-like particles that bud from mammalian cells. Standard proteomic approaches are used next to reveal the content of these particles. Fittingly, the team named the method ‘Virotrap’.
The Virotrap approach is exceptional as protein networks can be characterized under natural conditions. By trapping protein complexes in the protective environment of a virus-like shell, the intact complexes are preserved during the purification process. The researchers showed the method was suitable for detection of known binary interactions as well as mass spectrometry-based identification of novel protein partners.
Virotrap is a textbook example of bringing research teams with complementary expertise together. Cross-pollination with the labs of Jan Tavernier (VIB/UGent) and Kris Gevaert (VIB/UGent) enabled the development of this platform.
Jan Tavernier: “Virotrap represents a new concept in co-complex analysis wherein complex stability is physically guaranteed by a protective, physical structure. It is complementary to the arsenal of existing interactomics methods, but also holds potential for other fields, like drug target characterization. We also developed a small molecule-variant of Virotrap that could successfully trap protein partners for small molecule baits.”
Kris Gevaert: “Virotrap can also impact our understanding of disease pathways. We were actually surprised to see that this virus-based system could be used to study antiviral pathways, like Toll-like receptor signaling. Understanding these protein machines in their natural environment is essential if we want to modulate their activity in pathology.“
Trapping mammalian protein complexes in viral particles
Cell lysis is an inevitable step in classical mass spectrometry–based strategies to analyse protein complexes. Complementary lysis conditions, in situ cross-linking strategies and proximal labelling techniques are currently used to reduce lysis effects on the protein complex. We have developed Virotrap, a viral particle sorting approach that obviates the need for cell homogenization and preserves the protein complexes during purification. By fusing a bait protein to the HIV-1 GAG protein, we show that interaction partners become trapped within virus-like particles (VLPs) that bud from mammalian cells. Using an efficient VLP enrichment protocol, Virotrap allows the detection of known binary interactions and MS-based identification of novel protein partners as well. In addition, we show the identification of stimulus-dependent interactions and demonstrate trapping of protein partners for small molecules. Virotrap constitutes an elegant complementary approach to the arsenal of methods to study protein complexes.
Proteins mostly exert their function within supramolecular complexes. Strategies for detecting protein–protein interactions (PPIs) can be roughly divided into genetic systems1 and co-purification strategies combined with mass spectrometry (MS) analysis (for example, AP–MS)2. The latter approaches typically require cell or tissue homogenization using detergents, followed by capture of the protein complex using affinity tags3 or specific antibodies4. The protein complexes extracted from this ‘soup’ of constituents are then subjected to several washing steps before actual analysis by trypsin digestion and liquid chromatography–MS/MS analysis. Such lysis and purification protocols are typically empirical and have mostly been optimized using model interactions in single labs. In fact, lysis conditions can profoundly affect the number of both specific and nonspecific proteins that are identified in a typical AP–MS set-up. Indeed, recent studies using the nuclear pore complex as a model protein complex describe optimization of purifications for the different proteins in the complex by examining 96 different conditions5. Nevertheless, for new purifications, it remains hard to correctly estimate the loss of factors in a standard AP–MS experiment due to washing and dilution effects during treatments (that is, false negatives). These considerations have pushed the concept of stabilizing PPIs before the actual homogenization step. A classical approach involves cross-linking with simple reagents (for example, formaldehyde) or with more advanced isotope-labelled cross-linkers (reviewed in ref. 2). However, experimental challenges such as cell permeability and reactivity still preclude the widespread use of cross-linking agents. Moreover, MS-generated spectra of cross-linked peptides are notoriously difficult to identify correctly. A recent lysis-independent solution involves the expression of a bait protein fused to a promiscuous biotin ligase, which results in labelling of proteins proximal to the activity of the enzyme-tagged bait protein6. When compared with AP–MS, this BioID approach delivers a complementary set of candidate proteins, including novel interaction partners7, 8. Such particular studies clearly underscore the need for complementary approaches in the co-complex strategies.
The evolutionary stress on viruses promoted highly condensed coding of information and maximal functionality for small genomes. Accordingly, for HIV-1 it is sufficient to express a single protein, the p55 GAG protein, for efficient production of virus-like particles (VLPs) from cells9, 10. This protein is highly mobile before its accumulation in cholesterol-rich regions of the membrane, where multimerization initiates the budding process11. A total of 4,000–5,000 GAG molecules is required to form a single particle of about 145 nm (ref. 12). Both VLPs and mature viruses contain a number of host proteins that are recruited by binding to viral proteins. These proteins can either contribute to the infectivity (for example, Cyclophilin/FKBPA13) or act as antiviral proteins preventing the spreading of the virus (for example, APOBEC proteins14).
We here describe the development and application of Virotrap, an elegant co-purification strategy based on the trapping of a bait protein together with its associated protein partners in VLPs that are budded from the cell. After enrichment, these particles can be analysed by targeted (for example, western blotting) or unbiased approaches (MS-based proteomics). Virotrap allows detection of known binary PPIs, analysis of protein complexes and their dynamics, and readily detects protein binders for small molecules.
Concept of the Virotrap system
Classical AP–MS approaches rely on cell homogenization to access protein complexes, a step that can vary significantly with the lysis conditions (detergents, salt concentrations, pH conditions and so on)5. To eliminate the homogenization step in AP–MS, we reasoned that incorporation of a protein complex inside a secreted VLP traps the interaction partners under native conditions and protects them during further purification. We thus explored the possibility of protein complex packaging by the expression of GAG-bait protein chimeras (Fig. 1) as expression of GAG results in the release of VLPs from the cells9, 10. As a first PPI pair to evaluate this concept, we selected the HRAS protein as a bait combined with the RAF1 prey protein. We were able to specifically detect the HRAS–RAF1 interaction following enrichment of VLPs via ultracentrifugation (Supplementary Fig. 1a). To prevent tedious ultracentrifugation steps, we designed a novel single-step protocol wherein we co-express the vesicular stomatitis virus glycoprotein (VSV-G) together with a tagged version of this glycoprotein in addition to the GAG bait and prey. Both tagged and untagged VSV-G proteins are probably presented as trimers on the surface of the VLPs, allowing efficient antibody-based recovery from large volumes. The HRAS–RAF1 interaction was confirmed using this single-step protocol (Supplementary Fig. 1b). No associations with unrelated bait or prey proteins were observed for both protocols.
Figure 1: Schematic representation of the Virotrap strategy.
Expression of a GAG-bait fusion protein (1) results in submembrane multimerization (2) and subsequent budding of VLPs from cells (3). Interaction partners of the bait protein are also trapped within these VLPs and can be identified after purification by western blotting or MS analysis (4).
Virotrap for the detection of binary interactions
We next explored the reciprocal detection of a set of PPI pairs, which were selected based on published evidence and cytosolic localization15. After single-step purification and western blot analysis, we could readily detect reciprocal interactions between CDK2 and CKS1B, LCP2 and GRAP2, and S100A1 and S100B (Fig. 2a). Only for the LCP2 prey we observed nonspecific association with an irrelevant bait construct. However, the particle levels of the GRAP2 bait were substantially lower as compared with those of the GAG control construct (GAG protein levels in VLPs; Fig. 2a, second panel of the LCP2 prey). After quantification of the intensities of bait and prey proteins and normalization of prey levels using bait levels, we observed a strong enrichment for the GAG-GRAP2 bait (Supplementary Fig. 2).
…..
Virotrap for unbiased discovery of novel interactions
For the detection of novel interaction partners, we scaled up VLP production and purification protocols (Supplementary Fig. 5 and Supplementary Note 1 for an overview of the protocol) and investigated protein partners trapped using the following bait proteins: Fas-associated via death domain (FADD), A20 (TNFAIP3), nuclear factor-κB (NF-κB) essential modifier (IKBKG), TRAF family member-associated NF-κB activator (TANK), MYD88 and ring finger protein 41 (RNF41). To obtain specific interactors from the lists of identified proteins, we challenged the data with a combined protein list of 19 unrelated Virotrap experiments (Supplementary Table 1 for an overview). Figure 3 shows the design and the list of candidate interactors obtained after removal of all proteins that were found in the 19 control samples (including removal of proteins from the control list identified with a single peptide). The remaining list of confident protein identifications (identified with at least two peptides in at least two biological repeats) reveals both known and novel candidate interaction partners. All candidate interactors including single peptide protein identifications are given in Supplementary Data 2 and also include recurrent protein identifications of known interactors based on a single peptide; for example, CASP8 for FADD and TANK for NEMO. Using alternative methods, we confirmed the interaction between A20 and FADD, and the associations with transmembrane proteins (insulin receptor and insulin-like growth factor receptor 1) that were captured using RNF41 as a bait (Supplementary Fig. 6). To address the use of Virotrap for the detection of dynamic interactions, we activated the NF-κB pathway via the tumour necrosis factor (TNF) receptor (TNFRSF1A) using TNFα (TNF) and performed Virotrap analysis using A20 as bait (Fig. 3). This resulted in the additional enrichment of receptor-interacting kinase (RIPK1), TNFR1-associated via death domain (TRADD), TNFRSF1A and TNF itself, confirming the expected activated complex20.
Figure 3: Use of Virotrap for unbiased interactome analysis
Lysis conditions used in AP–MS strategies are critical for the preservation of protein complexes. A multitude of lysis conditions have been described, culminating in a recent report where protein complex stability was assessed under 96 lysis/purification protocols5. Moreover, the authors suggest to optimize the conditions for every complex, implying an important workload for researchers embarking on protein complex analysis using classical AP–MS. As lysis results in a profound change of the subcellular context and significantly alters the concentration of proteins, loss of complex integrity during a classical AP–MS protocol can be expected. A clear evolution towards ‘lysis-independent’ approaches in the co-complex analysis field is evident with the introduction of BioID6 and APEX25 where proximal proteins, including proteins residing in the complex, are labelled with biotin by an enzymatic activity fused to a bait protein. A side-by-side comparison between classical AP–MS and BioID showed overlapping and unique candidate binding proteins for both approaches7, 8, supporting the notion that complementary methods are needed to provide a comprehensive view on protein complexes. This has also been clearly demonstrated for binary approaches15 and is a logical consequence of the heterogenic nature underlying PPIs (binding mechanism, requirement for posttranslational modifications, location, affinity and so on).
In this report, we explore an alternative, yet complementary method to isolate protein complexes without interfering with cellular integrity. By trapping protein complexes in the protective environment of a virus-like shell, the intact complexes are preserved during the purification process. This constitutes a new concept in co-complex analysis wherein complex stability is physically guaranteed by a protective, physical structure. A comparison of our Virotrap approach with AP–MS shows complementary data, with specific false positives and false negatives for both methods (Supplementary Fig. 7).
The current implementation of the Virotrap platform implies the use of a GAG-bait construct resulting in considerable expression of the bait protein. Different strategies are currently pursued to reduce bait expression including co-expression of a native GAG protein together with the GAG-bait protein, not only reducing bait expression but also creating more ‘space’ in the particles potentially accommodating larger bait protein complexes. Nevertheless, the presence of the bait on the forming GAG scaffold creates an intracellular affinity matrix (comparable to the early in vitro affinity columns for purification of interaction partners from lysates26) that has the potential to compete with endogenous complexes by avidity effects. This avidity effect is a powerful mechanism that aids in the recruitment of cyclophilin to GAG27, a well-known weak interaction (Kd=16 μM (ref. 28)) detectable as a background association in the Virotrap system. Although background binding may be increased by elevated bait expression, weaker associations are readily detectable (for example, MAL—MYD88-binding study; Fig. 2c).
The size of Virotrap particles (around 145 nm) suggests limitations in the size of the protein complex that can be accommodated in the particles. Further experimentation is required to define the maximum size of proteins or the number of protein complexes that can be trapped inside the particles.
….
In conclusion, Virotrap captures significant parts of known interactomes and reveals new interactions. This cell lysis-free approach purifies protein complexes under native conditions and thus provides a powerful method to complement AP–MS or other PPI data. Future improvements of the system include strategies to reduce bait expression to more physiological levels and application of advanced data analysis options to filter out background. These developments can further aid in the deployment of Virotrap as a powerful extension of the current co-complex technology arsenal.
New Autism Blood Biomarker Identified
Researchers at UT Southwestern Medical Center have identified a blood biomarker that may aid in earlier diagnosis of children with autism spectrum disorder, or ASD
In a recent edition of Scientific Reports, UT Southwestern researchers reported on the identification of a blood biomarker that could distinguish the majority of ASD study participants versus a control group of similar age range. In addition, the biomarker was significantly correlated with the level of communication impairment, suggesting that the blood test may give insight into ASD severity.
“Numerous investigators have long sought a biomarker for ASD,” said Dr. Dwight German, study senior author and Professor of Psychiatry at UT Southwestern. “The blood biomarker reported here along with others we are testing can represent a useful test with over 80 percent accuracy in identifying ASD.”
ASD1 – was 66 percent accurate in diagnosing ASD. When combined with thyroid stimulating hormone level measurements, the ASD1-binding biomarker was 73 percent accurate at diagnosis
A Search for Blood Biomarkers for Autism: Peptoids
Autism spectrum disorder (ASD) is a neurodevelopmental disorder characterized by impairments in social interaction and communication, and restricted, repetitive patterns of behavior. In order to identify individuals with ASD and initiate interventions at the earliest possible age, biomarkers for the disorder are desirable. Research findings have identified widespread changes in the immune system in children with autism, at both systemic and cellular levels. In an attempt to find candidate antibody biomarkers for ASD, highly complex libraries of peptoids (oligo-N-substituted glycines) were screened for compounds that preferentially bind IgG from boys with ASD over typically developing (TD) boys. Unexpectedly, many peptoids were identified that preferentially bound IgG from TD boys. One of these peptoids was studied further and found to bind significantly higher levels (>2-fold) of the IgG1 subtype in serum from TD boys (n = 60) compared to ASD boys (n = 74), as well as compared to older adult males (n = 53). Together these data suggest that ASD boys have reduced levels (>50%) of an IgG1 antibody, which resembles the level found normally with advanced age. In this discovery study, the ASD1 peptoid was 66% accurate in predicting ASD.
….
Peptoid libraries have been used previously to search for autoantibodies for neurodegenerative diseases19 and for systemic lupus erythematosus (SLE)21. In the case of SLE, peptoids were identified that could identify subjects with the disease and related syndromes with moderate sensitivity (70%) and excellent specificity (97.5%). Peptoids were used to measure IgG levels from both healthy subjects and SLE patients. Binding to the SLE-peptoid was significantly higher in SLE patients vs. healthy controls. The IgG bound to the SLE-peptoid was found to react with several autoantigens, suggesting that the peptoids are capable of interacting with multiple, structurally similar molecules. These data indicate that IgG binding to peptoids can identify subjects with high levels of pathogenic autoantibodies vs. a single antibody.
In the present study, the ASD1 peptoid binds significantly lower levels of IgG1 in ASD males vs. TD males. This finding suggests that the ASD1 peptoid recognizes antibody(-ies) of an IgG1 subtype that is (are) significantly lower in abundance in the ASD males vs. TD males. Although a previous study14 has demonstrated lower levels of plasma IgG in ASD vs. TD children, here, we additionally quantified serum IgG levels in our individuals and found no difference in IgG between the two groups (data not shown). Furthermore, our IgG levels did not correlate with ASD1 binding levels, indicating that ASD1 does not bind IgG generically, and that the peptoid’s ability to differentiate between ASD and TD males is related to a specific antibody(-ies).
ASD subjects underwent a diagnostic evaluation using the ADOS and ADI-R, and application of the DSM-IV criteria prior to study inclusion. Only those subjects with a diagnosis of Autistic Disorder were included in the study. The ADOS is a semi-structured observation of a child’s behavior that allows examiners to observe the three core domains of ASD symptoms: reciprocal social interaction, communication, and restricted and repetitive behaviors1. When ADOS subdomain scores were compared with peptoid binding, the only significant relationship was with Social Interaction. However, the positive correlation would suggest that lower peptoid binding is associated with better social interaction, not poorer social interaction as anticipated.
The ADI-R is a structured parental interview that measures the core features of ASD symptoms in the areas of reciprocal social interaction, communication and language, and patterns of behavior. Of the three ADI-R subdomains, only the Communication domain was related to ASD1 peptoid binding, and this correlation was negative suggesting that low peptoid binding is associated with greater communication problems. These latter data are similar to the findings of Heuer et al.14 who found that children with autism with low levels of plasma IgG have high scores on the Aberrant Behavior Checklist (p < 0.0001). Thus, peptoid binding to IgG1 may be useful as a severity marker for ASD allowing for further characterization of individuals, but further research is needed.
It is interesting that in serum samples from older men, the ASD1 binding is similar to that in the ASD boys. This is consistent with the observation that with aging there is a reduction in the strength of the immune system, and the changes are gender-specific25. Recent studies using parabiosis26, in which blood from young mice reverse age-related impairments in cognitive function and synaptic plasticity in old mice, reveal that blood constituents from young subjects may contain important substances for maintaining neuronal functions. Work is in progress to identify the antibody/antibodies that are differentially binding to the ASD1 peptoid, which appear as a single band on the electrophoresis gel (Fig. 4).
……..
The ADI-R is a structured parental interview that measures the core features of ASD symptoms in the areas of reciprocal social interaction, communication and language, and patterns of behavior. Of the three ADI-R subdomains, only the Communication domain was related to ASD1 peptoid binding, and this correlation was negative suggesting that low peptoid binding is associated with greater communication problems. These latter data are similar to the findings of Heuer et al.14 who found that children with autism with low levels of plasma IgG have high scores on the Aberrant Behavior Checklist (p < 0.0001). Thus, peptoid binding to IgG1 may be useful as a severity marker for ASD allowing for further characterization of individuals, but further research is needed.
Titration of IgG binding to ASD1 using serum pooled from 10 TD males and 10 ASD males demonstrates ASD1’s ability to differentiate between the two groups. (B)Detecting IgG1 subclass instead of total IgG amplifies this differentiation. (C) IgG1 binding of individual ASD (n=74) and TD (n=60) male serum samples (1:100 dilution) to ASD1 significantly differs with TD>ASD. In addition, IgG1 binding of older adult male (AM) serum samples (n=53) to ASD1 is significantly lower than TD males, and not different from ASD males. The three groups were compared with a Kruskal-Wallis ANOVA, H = 10.1781, p<0.006. **p<0.005. Error bars show SEM. (D) Receiver-operating characteristic curve for ASD1’s ability to discriminate between ASD and TD males.
Association between peptoid binding and ADOS and ADI-R subdomains
Higher scores in any domain on the ADOS and ADI-R are indicative of more abnormal behaviors and/or symptoms. Among ADOS subdomains, there was no significant relationship between Communication and peptoid binding (z = 0.04, p = 0.966), Communication + Social interaction (z = 1.53, p = 0.127), or Stereotyped Behaviors and Restrictive Interests (SBRI) (z = 0.46, p = 0.647). Higher scores on the Social Interaction domain were significantly associated with higher peptoid binding (z = 2.04, p = 0.041).
Among ADI-R subdomains, higher scores on the Communication domain were associated with lower levels of peptoid binding (z = −2.28, p = 0.023). There was not a significant relationship between Social Interaction (z = 0.07, p = 0.941) or Restrictive/Repetitive Stereotyped Behaviors (z = −1.40, p = 0.162) and peptoid binding.
Computational Model Finds New Protein-Protein Interactions
Researchers at University of Pittsburgh have discovered 500 new protein-protein interactions (PPIs) associated with genes linked to schizophrenia.
Using a computational model they developed, researchers at the University of Pittsburgh School of Medicine have discovered more than 500 new protein-protein interactions (PPIs) associated with genes linked to schizophrenia. The findings, published online in npj Schizophrenia, a Nature Publishing Group journal, could lead to greater understanding of the biological underpinnings of this mental illness, as well as point the way to treatments.
There have been many genome-wide association studies (GWAS) that have identified gene variants associated with an increased risk for schizophrenia, but in most cases there is little known about the proteins that these genes make, what they do and how they interact, said senior investigator Madhavi Ganapathiraju, Ph.D., assistant professor of biomedical informatics, Pitt School of Medicine.
“GWAS studies and other research efforts have shown us what genes might be relevant in schizophrenia,” she said. “What we have done is the next step. We are trying to understand how these genes relate to each other, which could show us the biological pathways that are important in the disease.”
Each gene makes proteins and proteins typically interact with each other in a biological process. Information about interacting partners can shed light on the role of a gene that has not been studied, revealing pathways and biological processes associated with the disease and also its relation to other complex diseases.
Dr. Ganapathiraju’s team developed a computational model called High-Precision Protein Interaction Prediction (HiPPIP) and applied it to discover PPIs of schizophrenia-linked genes identified through GWAS, as well as historically known risk genes. They found 504 never-before known PPIs, and noted also that while schizophrenia-linked genes identified historically and through GWAS had little overlap, the model showed they shared more than 100 common interactors.
“We can infer what the protein might do by checking out the company it keeps,” Dr. Ganapathiraju explained. “For example, if I know you have many friends who play hockey, it could mean that you are involved in hockey, too. Similarly, if we see that an unknown protein interacts with multiple proteins involved in neural signaling, for example, there is a high likelihood that the unknown entity also is involved in the same.”
Dr. Ganapathiraju and colleagues have drawn such inferences on protein function based on the PPIs of proteins, and made their findings available on a website Schizo-Pi. This information can be used by biologists to explore the schizophrenia interactome with the aim of understanding more about the disease or developing new treatment drugs.
Schizophrenia interactome with 504 novel protein–protein interactions
(GWAS) have revealed the role of rare and common genetic variants, but the functional effects of the risk variants remain to be understood. Protein interactome-based studies can facilitate the study of molecular mechanisms by which the risk genes relate to schizophrenia (SZ) genesis, but protein–protein interactions (PPIs) are unknown for many of the liability genes. We developed a computational model to discover PPIs, which is found to be highly accurate according to computational evaluations and experimental validations of selected PPIs. We present here, 365 novel PPIs of liability genes identified by the SZ Working Group of the Psychiatric Genomics Consortium (PGC). Seventeen genes that had no previously known interactions have 57 novel interactions by our method. Among the new interactors are 19 drug targets that are targeted by 130 drugs. In addition, we computed 147 novel PPIs of 25 candidate genes investigated in the pre-GWAS era. While there is little overlap between the GWAS genes and the pre-GWAS genes, the interactomes reveal that they largely belong to the same pathways, thus reconciling the apparent disparities between the GWAS and prior gene association studies. The interactome including 504 novel PPIs overall, could motivate other systems biology studies and trials with repurposed drugs. The PPIs are made available on a webserver, called Schizo-Pi at http://severus.dbmi.pitt.edu/schizo-pi with advanced search capabilities.
Schizophrenia (SZ) is a common, potentially severe psychiatric disorder that afflicts all populations.1 Gene mapping studies suggest that SZ is a complex disorder, with a cumulative impact of variable genetic effects coupled with environmental factors.2 As many as 38 genome-wide association studies (GWAS) have been reported on SZ out of a total of 1,750 GWAS publications on 1,087 traits or diseases reported in the GWAS catalog maintained by the National Human Genome Research Institute of USA3 (as of April 2015), revealing the common variants associated with SZ.4 The SZ Working Group of the Psychiatric Genomics Consortium (PGC) identified 108 genetic loci that likely confer risk for SZ.5 While the role of genetics has been clearly validated by this study, the functional impact of the risk variants is not well-understood.6,7 Several of the genes implicated by the GWAS have unknown functions and could participate in possibly hitherto unknown pathways.8 Further, there is little or no overlap between the genes identified through GWAS and ‘candidate genes’ proposed in the pre-GWAS era.9
Interactome-based studies can be useful in discovering the functional associations of genes. For example,disrupted in schizophrenia 1 (DISC1), an SZ related candidate gene originally had no known homolog in humans. Although it had well-characterized protein domains such as coiled-coil domains and leucine-zipper domains, its function was unknown.10,11 Once its protein–protein interactions (PPIs) were determined using yeast 2-hybrid technology,12 investigators successfully linked DISC1 to cAMP signaling, axon elongation, and neuronal migration, and accelerated the research pertaining to SZ in general, and DISC1 in particular.13 Typically such studies are carried out on known protein–protein interaction (PPI) networks, or as in the case of DISC1, when there is a specific gene of interest, its PPIs are determined by methods such as yeast 2-hybrid technology.
Knowledge of human PPI networks is thus valuable for accelerating discovery of protein function, and indeed, biomedical research in general. However, of the hundreds of thousands of biophysical PPIs thought to exist in the human interactome,14,15 <100,000 are known today (Human Protein Reference Database, HPRD16 and BioGRID17 databases). Gold standard experimental methods for the determination of all the PPIs in human interactome are time-consuming, expensive and may not even be feasible, as about 250 million pairs of proteins would need to be tested overall; high-throughput methods such as yeast 2-hybrid have important limitations for whole interactome determination as they have a low recall of 23% (i.e., remaining 77% of true interactions need to be determined by other means), and a low precision (i.e., the screens have to be repeated multiple times to achieve high selectivity).18,19Computational methods are therefore necessary to complete the interactome expeditiously. Algorithms have begun emerging to predict PPIs using statistical machine learning on the characteristics of the proteins, but these algorithms are employed predominantly to study yeast. Two significant computational predictions have been reported for human interactome; although they have had high false positive rates, these methods have laid the foundation for computational prediction of human PPIs.20,21
We have created a new PPI prediction model called High-Confidence Protein–Protein Interaction Prediction (HiPPIP) model. Novel interactions predicted with this model are making translational impact. For example, we discovered a PPI between OASL and DDX58, which on validation showed that an increased expression of OASL could boost innate immunity to combat influenza by activating the RIG-I pathway.22 Also, the interactome of the genes associated with congenital heart disease showed that the disease morphogenesis has a close connection with the structure and function of cilia.23Here, we describe the HiPPIP model and its application to SZ genes to construct the SZ interactome. After computational evaluations and experimental validations of selected novel PPIs, we present here 504 highly confident novel PPIs in the SZ interactome, shedding new light onto several uncharacterized genes that are associated with SZ.
We developed a computational model called HiPPIP to predict PPIs (see Methods and Supplementary File 1). The model has been evaluated by computational methods and experimental validations and is found to be highly accurate. Evaluations on a held-out test data showed a precision of 97.5% and a recall of 5%. 5% recall out of 150,000 to 600,000 estimated number of interactions in the human interactome corresponds to 7,500–30,000 novel PPIs in the whole interactome. Note that, it is likely that the real precision would be higher than 97.5% because in this test data, randomly paired proteins are treated as non-interacting protein pairs, whereas some of them may actually be interacting pairs with a small probability; thus, some of the pairs that are treated as false positives in test set are likely to be true but hitherto unknown interactions. In Figure 1a, we show the precision versus recall of our method on ‘hub proteins’ where we considered all pairs that received a score >0.5 by HiPPIP to be novel interactions. In Figure 1b, we show the number of true positives versus false positives observed in hub proteins. Both these figures also show our method to be superior in comparison to the prediction of membrane-receptor interactome by Qi et al’s.24 True positives versus false positives are also shown for individual hub proteins by our method in Figure 1cand by Qi et al’s.23 in Figure 1d. These evaluations showed that our predictions contain mostly true positives. Unlike in other domains where ranked lists are commonly used such as information retrieval, in PPI prediction the ‘false positives’ may actually be unlabeled instances that are indeed true interactions that are not yet discovered. In fact, such unlabeled pairs predicted as interactors of the hub gene HMGB1 (namely, the pairs HMGB1-KL and HMGB1-FLT1) were validated by experimental methods and found to be true PPIs (See the Figures e–g inSupplementary File 3). Thus, we concluded that the protein pairs that received a score of ⩾0.5 are highly confident to be true interactions. The pairs that receive a score less than but close to 0.5 (i.e., in the range of 0.4–0.5) may also contain several true PPIs; however, we cannot confidently say that all in this range are true PPIs. Only the PPIs predicted with a score >0.5 are included in the interactome.
Computational evaluation of predicted protein–protein interactions on hub proteins: (a) precision recall curve. (b) True positive versus false positives in ranked lists of hub type membrane receptors for our method and that by Qi et al. True positives versus false positives are shown for individual membrane receptors by our method in (c) and by Qi et al. in (d). Thick line is the average, which is also the same as shown in (b). Note:x-axis is recall in (a), whereas it is number of false positives in (b–d). The range of y-axis is observed by varying the threshold from 1.0–0 in (a), and to 0.5 in (b–d).
SZ interactome
By applying HiPPIP to the GWAS genes and Historic (pre-GWAS) genes, we predicted over 500 high confidence new PPIs adding to about 1400 previously known PPIs.
Schizophrenia interactome: network view of the schizophrenia interactome is shown as a graph, where genes are shown as nodes and PPIs as edges connecting the nodes. Schizophrenia-associated genes are shown as dark blue nodes, novel interactors as red color nodes and known interactors as blue color nodes. The source of the schizophrenia genes is indicated by its label font, where Historic genes are shown italicized, GWAS genes are shown in bold, and the one gene that is common to both is shown in italicized and bold. For clarity, the source is also indicated by the shape of the node (triangular for GWAS and square for Historic and hexagonal for both). Symbols are shown only for the schizophrenia-associated genes; actual interactions may be accessed on the web. Red edges are the novel interactions, whereas blue edges are known interactions. GWAS, genome-wide association studies of schizophrenia; PPI, protein–protein interaction.
We have made the known and novel interactions of all SZ-associated genes available on a webserver called Schizo-Pi, at the addresshttp://severus.dbmi.pitt.edu/schizo-pi. This webserver is similar to Wiki-Pi33 which presents comprehensive annotations of both participating proteins of a PPI side-by-side. The difference between Wiki-Pi which we developed earlier, and Schizo-Pi, is the inclusion of novel predicted interactions of the SZ genes into the latter.
Despite the many advances in biomedical research, identifying the molecular mechanisms underlying the disease is still challenging. Studies based on protein interactions were proven to be valuable in identifying novel gene associations that could shed new light on disease pathology.35 The interactome including more than 500 novel PPIs will help to identify pathways and biological processes associated with the disease and also its relation to other complex diseases. It also helps identify potential drugs that could be repurposed to use for SZ treatment.
Functional and pathway enrichment in SZ interactome
When a gene of interest has little known information, functions of its interacting partners serve as a starting point to hypothesize its own function. We computed statistically significant enrichment of GO biological process terms among the interacting partners of each of the genes using BinGO36 (see online at http://severus.dbmi.pitt.edu/schizo-pi).
Protein aggregation and aggregate toxicity: new insights into protein folding, misfolding diseases and biological evolution
Massimo Stefani · Christopher M. Dobson
Abstract The deposition of proteins in the form of amyloid fibrils and plaques is the characteristic feature of more than 20 degenerative conditions affecting either the central nervous system or a variety of peripheral tissues. As these conditions include Alzheimer’s, Parkinson’s and the prion diseases, several forms of fatal systemic amyloidosis, and at least one condition associated with medical intervention (haemodialysis), they are of enormous importance in the context of present-day human health and welfare. Much remains to be learned about the mechanism by which the proteins associated with these diseases aggregate and form amyloid structures, and how the latter affect the functions of the organs with which they are associated. A great deal of information concerning these diseases has emerged, however, during the past 5 years, much of it causing a number of fundamental assumptions about the amyloid diseases to be reexamined. For example, it is now apparent that the ability to form amyloid structures is not an unusual feature of the small number of proteins associated with these diseases but is instead a general property of polypeptide chains. It has also been found recently that aggregates of proteins not associated with amyloid diseases can impair the ability of cells to function to a similar extent as aggregates of proteins linked with specific neurodegenerative conditions. Moreover, the mature amyloid fibrils or plaques appear to be substantially less toxic than the prefibrillar aggregates that are their precursors. The toxicity of these early aggregates appears to result from an intrinsic ability to impair fundamental cellular processes by interacting with cellular membranes, causing oxidative stress and increases in free Ca2+ that eventually lead to apoptotic or necrotic cell death. The ‘new view’ of these diseases also suggests that other degenerative conditions could have similar underlying origins to those of the amyloidoses. In addition, cellular protection mechanisms, such as molecular chaperones and the protein degradation machinery, appear to be crucial in the prevention of disease in normally functioning living organisms. It also suggests some intriguing new factors that could be of great significance in the evolution of biological molecules and the mechanisms that regulate their behaviour.
The genetic information within a cell encodes not only the specific structures and functions of proteins but also the way these structures are attained through the process known as protein folding. In recent years many of the underlying features of the fundamental mechanism of this complex process and the manner in which it is regulated in living systems have emerged from a combination of experimental and theoretical studies [1]. The knowledge gained from these studies has also raised a host of interesting issues. It has become apparent, for example, that the folding and unfolding of proteins is associated with a whole range of cellular processes from the trafficking of molecules to specific organelles to the regulation of the cell cycle and the immune response. Such observations led to the inevitable conclusion that the failure to fold correctly, or to remain correctly folded, gives rise to many different types of biological malfunctions and hence to many different forms of disease [2]. In addition, it has been recognised recently that a large number of eukaryotic genes code for proteins that appear to be ‘natively unfolded’, and that proteins can adopt, under certain circumstances, highly organised multi-molecular assemblies whose structures are not specifically encoded in the amino acid sequence. Both these observations have raised challenging questions about one of the most fundamental principles of biology: the close relationship between the sequence, structure and function of proteins, as we discuss below [3].
It is well established that proteins that are ‘misfolded’, i.e. that are not in their functionally relevant conformation, are devoid of normal biological activity. In addition, they often aggregate and/or interact inappropriately with other cellular components leading to impairment of cell viability and eventually to cell death. Many diseases, often known as misfolding or conformational diseases, ultimately result from the presence in a living system of protein molecules with structures that are ‘incorrect’, i.e. that differ from those in normally functioning organisms [4]. Such diseases include conditions in which a specific protein, or protein complex, fails to fold correctly (e.g. cystic fibrosis, Marfan syndrome, amyotonic lateral sclerosis) or is not sufficiently stable to perform its normal function (e.g. many forms of cancer). They also include conditions in which aberrant folding behaviour results in the failure of a protein to be correctly trafficked (e.g. familial hypercholesterolaemia, α1-antitrypsin deficiency, and some forms of retinitis pigmentosa) [4]. The tendency of proteins to aggregate, often to give species extremely intractable to dissolution and refolding, is of course also well known in other circumstances. Examples include the formation of inclusion bodies during overexpression of heterologous proteins in bacteria and the precipitation of proteins during laboratory purification procedures. Indeed, protein aggregation is well established as one of the major difficulties associated with the production and handling of proteins in the biotechnology and pharmaceutical industries [5].
Considerable attention is presently focused on a group of protein folding diseases known as amyloidoses. In these diseases specific peptides or proteins fail to fold or to remain correctly folded and then aggregate (often with other components) so as to give rise to ‘amyloid’ deposits in tissue. Amyloid structures can be recognised because they possess a series of specific tinctorial and biophysical characteristics that reflect a common core structure based on the presence of highly organised βsheets [6]. The deposits in strictly defined amyloidoses are extracellular and can often be observed as thread-like fibrillar structures, sometimes assembled further into larger aggregates or plaques. These diseases include a range of sporadic, familial or transmissible degenerative diseases, some of which affect the brain and the central nervous system (e.g. Alzheimer’s and Creutzfeldt-Jakob diseases), while others involve peripheral tissues and organs such as the liver, heart and spleen (e.g. systemic amyloidoses and type II diabetes) [7, 8]. In other forms of amyloidosis, such as primary or secondary systemic amyloidoses, proteinaceous deposits are found in skeletal tissue and joints (e.g. haemodialysis-related amyloidosis) as well as in several organs (e.g. heart and kidney). Yet other components such as collagen, glycosaminoglycans and proteins (e.g. serum amyloid protein) are often present in the deposits protecting them against degradation [9, 10, 11]. Similar deposits to those in the amyloidoses are, however, found intracellularly in other diseases; these can be localised either in the cytoplasm, in the form of specialised aggregates known as aggresomes or as Lewy or Russell bodies or in the nucleus (see below).
The presence in tissue of proteinaceous deposits is a hallmark of all these diseases, suggesting a causative link between aggregate formation and pathological symptoms (often known as the amyloid hypothesis) [7, 8, 12]. At the present time the link between amyloid formation and disease is widely accepted on the basis of a large number of biochemical and genetic studies. The specific nature of the pathogenic species, and the molecular basis of their ability to damage cells, are however, the subject of intense debate [13, 14, 15, 16, 17, 18, 19, 20]. In neurodegenerative disorders it is very likely that the impairment of cellular function follows directly from the interactions of the aggregated proteins with cellular components [21, 22]. In the systemic non-neurological diseases, however, it is widely believed that the accumulation in vital organs of large amounts of amyloid deposits can by itself cause at least some of the clinical symptoms [23]. It is quite possible, however, that there are other more specific effects of aggregates on biochemical processes even in these diseases. The presence of extracellular or intracellular aggregates of a specific polypeptide molecule is a characteristic of all the 20 or so recognised amyloid diseases. The polypeptides involved include full length proteins (e.g. lysozyme or immunoglobulin light chains), biological peptides (amylin, atrial natriuretic factor) and fragments of larger proteins produced as a result of specific processing (e.g. the Alzheimer βpeptide) or of more general degradation [e.g. poly(Q) stretches cleaved from proteins with poly(Q) extensions such as huntingtin, ataxins and the androgen receptor]. The peptides and proteins associated with known amyloid diseases are listed in Table 1. In some cases the proteins involved have wild type sequences, as in sporadic forms of the diseases, but in other cases these are variants resulting from genetic mutations associated with familial forms of the diseases. In some cases both sporadic and familial diseases are associated with a given protein; in this case the mutational variants are usually associated with early-onset forms of the disease. In the case of the neurodegenerative diseases associated with the prion protein some forms of the diseases are transmissible. The existence of familial forms of a number of amyloid diseases has provided significant clues to the origins of the pathologies. For example, there are increasingly strong links between the age at onset of familial forms of disease and the effects of the mutations involved on the propensity of the affected proteins to aggregate in vitro. Such findings also support the link between the process of aggregation and the clinical manifestations of disease [24, 25].
The presence in cells of misfolded or aggregated proteins triggers a complex biological response. In the cytosol, this is referred to as the ‘heat shock response’ and in the endoplasmic reticulum (ER) it is known as the ‘unfolded protein response’. These responses lead to the expression, among others, of the genes for heat shock proteins (Hsp, or molecular chaperone proteins) and proteins involved in the ubiquitin-proteasome pathway [26]. The evolution of such complex biochemical machinery testifies to the fact that it is necessary for cells to isolate and clear rapidly and efficiently any unfolded or incorrectly folded protein as soon as it appears. In itself this fact suggests that these species could have a generally adverse effect on cellular components and cell viability. Indeed, it was a major step forward in understanding many aspects of cell biology when it was recognised that proteins previously associated only with stress, such as heat shock, are in fact crucial in the normal functioning of living systems. This advance, for example, led to the discovery of the role of molecular chaperones in protein folding and in the normal ‘housekeeping’ processes that are inherent in healthy cells [27, 28]. More recently a number of degenerative diseases, both neurological and systemic, have been linked to, or shown to be affected by, impairment of the ubiquitin-proteasome pathway (Table 2). The diseases are primarily associated with a reduction in either the expression or the biological activity of Hsps, ubiquitin, ubiquitinating or deubiquitinating enzymes and the proteasome itself, as we show below [29, 30, 31, 32], or even to the failure of the quality control mechanisms that ensure proper maturation of proteins in the ER. The latter normally leads to degradation of a significant proportion of polypeptide chains before they have attained their native conformations through retrograde translocation to the cytosol [33, 34].
….
It is now well established that the molecular basis of protein aggregation into amyloid structures involves the existence of ‘misfolded’ forms of proteins, i.e. proteins that are not in the structures in which they normally function in vivo or of fragments of proteins resulting from degradation processes that are inherently unable to fold [4, 7, 8, 36]. Aggregation is one of the common consequences of a polypeptide chain failing to reach or maintain its functional three-dimensional structure. Such events can be associated with specific mutations, misprocessing phenomena, aberrant interactions with metal ions, changes in environmental conditions, such as pH or temperature, or chemical modification (oxidation, proteolysis). Perturbations in the conformational properties of the polypeptide chain resulting from such phenomena may affect equilibrium 1 in Fig. 1 increasing the population of partially unfolded, or misfolded, species that are much more aggregation-prone than the native state.
Fig. 1 Overview of the possible fates of a newly synthesised polypeptide chain. The equilibrium ① between the partially folded molecules and the natively folded ones is usually strongly in favour of the latter except as a result of specific mutations, chemical modifications or partially destabilising solution conditions. The increased equilibrium populations of molecules in the partially or completely unfolded ensemble of structures are usually degraded by the proteasome; when this clearance mechanism is impaired, such species often form disordered aggregates or shift equilibrium ② towards the nucleation of pre-fibrillar assemblies that eventually grow into mature fibrils (equilibrium ③). DANGER! indicates that pre-fibrillar aggregates in most cases display much higher toxicity than mature fibrils. Heat shock proteins (Hsp) can suppress the appearance of pre-fibrillar assemblies by minimising the population of the partially folded molecules by assisting in the correct folding of the nascent chain and the unfolded protein response target incorrectly folded proteins for degradation.
……
Little is known at present about the detailed arrangement of the polypeptide chains themselves within amyloid fibrils, either those parts involved in the core βstrands or in regions that connect the various β-strands. Recent data suggest that the sheets are relatively untwisted and may in some cases at least exist in quite specific supersecondary structure motifs such as β-helices [6, 40] or the recently proposed µ-helix [41]. It seems possible that there may be significant differences in the way the strands are assembled depending on characteristics of the polypeptide chain involved [6, 42]. Factors including length, sequence (and in some cases the presence of disulphide bonds or post-translational modifications such as glycosylation) may be important in determining details of the structures. Several recent papers report structural models for amyloid fibrils containing different polypeptide chains, including the Aβ40 peptide, insulin and fragments of the prion protein, based on data from such techniques as cryo-electron microscopy and solid-state magnetic resonance spectroscopy [43, 44]. These models have much in common and do indeed appear to reflect the fact that the structures of different fibrils are likely to be variations on a common theme [40]. It is also emerging that there may be some common and highly organised assemblies of amyloid protofilaments that are not simply extended threads or ribbons. It is clear, for example, that in some cases large closed loops can be formed [45, 46, 47], and there may be specific types of relatively small spherical or ‘doughnut’ shaped structures that can result in at least some circumstances (see below).
…..
The similarity of some early amyloid aggregates with the pores resulting from oligomerisation of bacterial toxins and pore-forming eukaryotic proteins (see below) also suggest that the basic mechanism of protein aggregation into amyloid structures may not only be associated with diseases but in some cases could result in species with functional significance. Recent evidence indicates that a variety of micro-organisms may exploit the controlled aggregation of specific proteins (or their precursors) to generate functional structures. Examples include bacterial curli [52] and proteins of the interior fibre cells of mammalian ocular lenses, whose β-sheet arrays seem to be organised in an amyloid-like supramolecular order [53]. In this case the inherent stability of amyloid-like protein structure may contribute to the long-term structural integrity and transparency of the lens. Recently it has been hypothesised that amyloid-like aggregates of serum amyloid A found in secondary amyloidoses following chronic inflammatory diseases protect the host against bacterial infections by inducing lysis of bacterial cells [54]. One particularly interesting example is a ‘misfolded’ form of the milk protein α-lactalbumin that is formed at low pH and trapped by the presence of specific lipid molecules [55]. This form of the protein has been reported to trigger apoptosis selectively in tumour cells providing evidence for its importance in protecting infants from certain types of cancer [55]. ….
Amyloid formation is a generic property of polypeptide chains ….
It is clear that the presence of different side chains can influence the details of amyloid structures, particularly the assembly of protofibrils, and that they give rise to the variations on the common structural theme discussed above. More fundamentally, the composition and sequence of a peptide or protein affects profoundly its propensity to form amyloid structures under given conditions (see below).
Because the formation of stable protein aggregates of amyloid type does not normally occur in vivo under physiological conditions, it is likely that the proteins encoded in the genomes of living organisms are endowed with structural adaptations that mitigate against aggregation under these conditions. A recent survey involving a large number of structures of β-proteins highlights several strategies through which natural proteins avoid intermolecular association of β-strands in their native states [65]. Other surveys of protein databases indicate that nature disfavours sequences of alternating polar and nonpolar residues, as well as clusters of several consecutive hydrophobic residues, both of which enhance the tendency of a protein to aggregate prior to becoming completely folded [66, 67].
……
Precursors of amyloid fibrils can be toxic to cells
It was generally assumed until recently that the proteinaceous aggregates most toxic to cells are likely to be mature amyloid fibrils, the form of aggregates that have been commonly detected in pathological deposits. It therefore appeared probable that the pathogenic features underlying amyloid diseases are a consequence of the interaction with cells of extracellular deposits of aggregated material. As well as forming the basis for understanding the fundamental causes of these diseases, this scenario stimulated the exploration of therapeutic approaches to amyloidoses that focused mainly on the search for molecules able to impair the growth and deposition of fibrillar forms of aggregated proteins. ….
Structural basis and molecular features of amyloid toxicity
The presence of toxic aggregates inside or outside cells can impair a number of cell functions that ultimately lead to cell death by an apoptotic mechanism [95, 96]. Recent research suggests, however, that in most cases initial perturbations to fundamental cellular processes underlie the impairment of cell function induced by aggregates of disease-associated polypeptides. Many pieces of data point to a central role of modifications to the intracellular redox status and free Ca2+ levels in cells exposed to toxic aggregates [45, 89, 97, 98, 99, 100, 101]. A modification of the intracellular redox status in such cells is associated with a sharp increase in the quantity of reactive oxygen species (ROS) that is reminiscent of the oxidative burst by which leukocytes destroy invading foreign cells after phagocytosis. In addition, changes have been observed in reactive nitrogen species, lipid peroxidation, deregulation of NO metabolism [97], protein nitrosylation [102] and upregulation of heme oxygenase-1, a specific marker of oxidative stress [103]. ….
Results have recently been reported concerning the toxicity towards cultured cells of aggregates of poly(Q) peptides which argues against a disease mechanism based on specific toxic features of the aggregates. These results indicate that there is a close relationship between the toxicity of proteins with poly(Q) extensions and their nuclear localisation. In addition they support the hypotheses that the toxicity of poly(Q) aggregates can be a consequence of altered interactions with nuclear coactivator or corepressor molecules including p53, CBP, Sp1 and TAF130 or of the interaction with transcription factors and nuclear coactivators, such as CBP, endowed with short poly(Q) stretches ([95] and references therein)…..
Concluding remarks
The data reported in the past few years strongly suggest that the conversion of normally soluble proteins into amyloid fibrils and the toxicity of small aggregates appearing during the early stages of the formation of the latter are common or generic features of polypeptide chains. Moreover, the molecular basis of this toxicity also appears to display common features between the different systems that have so far been studied. The ability of many, perhaps all, natural polypeptides to ‘misfold’ and convert into toxic aggregates under suitable conditions suggests that one of the most important driving forces in the evolution of proteins must have been the negative selection against sequence changes that increase the tendency of a polypeptide chain to aggregate. Nevertheless, as protein folding is a stochastic process, and no such process can be completely infallible, misfolded proteins or protein folding intermediates in equilibrium with the natively folded molecules must continuously form within cells. Thus mechanisms to deal with such species must have co-evolved with proteins. Indeed, it is clear that misfolding, and the associated tendency to aggregate, is kept under control by molecular chaperones, which render the resulting species harmless assisting in their refolding, or triggering their degradation by the cellular clearance machinery [166, 167, 168, 169, 170, 171, 172, 173, 175, 177, 178].
Misfolded and aggregated species are likely to owe their toxicity to the exposure on their surfaces of regions of proteins that are buried in the interior of the structures of the correctly folded native states. The exposure of large patches of hydrophobic groups is likely to be particularly significant as such patches favour the interaction of the misfolded species with cell membranes [44, 83, 89, 90, 91, 93]. Interactions of this type are likely to lead to the impairment of the function and integrity of the membranes involved, giving rise to a loss of regulation of the intracellular ion balance and redox status and eventually to cell death. In addition, misfolded proteins undoubtedly interact inappropriately with other cellular components, potentially giving rise to the impairment of a range of other biological processes. Under some conditions the intracellular content of aggregated species may increase directly, due to an enhanced propensity of incompletely folded or misfolded species to aggregate within the cell itself. This could occur as the result of the expression of mutational variants of proteins with decreased stability or cooperativity or with an intrinsically higher propensity to aggregate. It could also occur as a result of the overproduction of some types of protein, for example, because of other genetic factors or other disease conditions, or because of perturbations to the cellular environment that generate conditions favouring aggregation, such as heat shock or oxidative stress. Finally, the accumulation of misfolded or aggregated proteins could arise from the chaperone and clearance mechanisms becoming overwhelmed as a result of specific mutant phenotypes or of the general effects of ageing [173, 174].
The topics discussed in this review not only provide a great deal of evidence for the ‘new view’ that proteins have an intrinsic capability of misfolding and forming structures such as amyloid fibrils but also suggest that the role of molecular chaperones is even more important than was thought in the past. The role of these ubiquitous proteins in enhancing the efficiency of protein folding is well established [185]. It could well be that they are at least as important in controlling the harmful effects of misfolded or aggregated proteins as in enhancing the yield of functional molecules.
Nutritional Status is Associated with Faster Cognitive Decline and Worse Functional Impairment in the Progression of Dementia: The Cache County Dementia Progression Study1
Nutritional status may be a modifiable factor in the progression of dementia. We examined the association of nutritional status and rate of cognitive and functional decline in a U.S. population-based sample. Study design was an observational longitudinal study with annual follow-ups up to 6 years of 292 persons with dementia (72% Alzheimer’s disease, 56% female) in Cache County, UT using the Mini-Mental State Exam (MMSE), Clinical Dementia Rating Sum of Boxes (CDR-sb), and modified Mini Nutritional Assessment (mMNA). mMNA scores declined by approximately 0.50 points/year, suggesting increasing risk for malnutrition. Lower mMNA score predicted faster rate of decline on the MMSE at earlier follow-up times, but slower decline at later follow-up times, whereas higher mMNA scores had the opposite pattern (mMNA by time β= 0.22, p = 0.017; mMNA by time2 β= –0.04, p = 0.04). Lower mMNA score was associated with greater impairment on the CDR-sb over the course of dementia (β= 0.35, p < 0.001). Assessment of malnutrition may be useful in predicting rates of progression in dementia and may provide a target for clinical intervention.
Shared Genetic Risk Factors for Late-Life Depression and Alzheimer’s Disease
Background: Considerable evidence has been reported for the comorbidity between late-life depression (LLD) and Alzheimer’s disease (AD), both of which are very common in the general elderly population and represent a large burden on the health of the elderly. The pathophysiological mechanisms underlying the link between LLD and AD are poorly understood. Because both LLD and AD can be heritable and are influenced by multiple risk genes, shared genetic risk factors between LLD and AD may exist. Objective: The objective is to review the existing evidence for genetic risk factors that are common to LLD and AD and to outline the biological substrates proposed to mediate this association. Methods: A literature review was performed. Results: Genetic polymorphisms of brain-derived neurotrophic factor, apolipoprotein E, interleukin 1-beta, and methylenetetrahydrofolate reductase have been demonstrated to confer increased risk to both LLD and AD by studies examining either LLD or AD patients. These results contribute to the understanding of pathophysiological mechanisms that are common to both of these disorders, including deficits in nerve growth factors, inflammatory changes, and dysregulation mechanisms involving lipoprotein and folate. Other conflicting results have also been reviewed, and few studies have investigated the effects of the described polymorphisms on both LLD and AD. Conclusion: The findings suggest that common genetic pathways may underlie LLD and AD comorbidity. Studies to evaluate the genetic relationship between LLD and AD may provide insights into the molecular mechanisms that trigger disease progression as the population ages.
Association of Vitamin B12, Folate, and Sulfur Amino Acids With Brain Magnetic Resonance Imaging Measures in Older Adults: A Longitudinal Population-Based Study
Importance Vitamin B12, folate, and sulfur amino acids may be modifiable risk factors for structural brain changes that precede clinical dementia.
Objective To investigate the association of circulating levels of vitamin B12, red blood cell folate, and sulfur amino acids with the rate of total brain volume loss and the change in white matter hyperintensity volume as measured by fluid-attenuated inversion recovery in older adults.
Design, Setting, and Participants The magnetic resonance imaging subsample of the Swedish National Study on Aging and Care in Kungsholmen, a population-based longitudinal study in Stockholm, Sweden, was conducted in 501 participants aged 60 years or older who were free of dementia at baseline. A total of 299 participants underwent repeated structural brain magnetic resonance imaging scans from September 17, 2001, to December 17, 2009.
Main Outcomes and Measures The rate of brain tissue volume loss and the progression of total white matter hyperintensity volume.
Results In the multi-adjusted linear mixed models, among 501 participants (300 women [59.9%]; mean [SD] age, 70.9 [9.1] years), higher baseline vitamin B12 and holotranscobalamin levels were associated with a decreased rate of total brain volume loss during the study period: for each increase of 1 SD, β (SE) was 0.048 (0.013) for vitamin B12 (P < .001) and 0.040 (0.013) for holotranscobalamin (P = .002). Increased total homocysteine levels were associated with faster rates of total brain volume loss in the whole sample (β [SE] per 1-SD increase, –0.035 [0.015]; P = .02) and with the progression of white matter hyperintensity among participants with systolic blood pressure greater than 140 mm Hg (β [SE] per 1-SD increase, 0.000019 [0.00001]; P = .047). No longitudinal associations were found for red blood cell folate and other sulfur amino acids.
Conclusions and Relevance This study suggests that both vitamin B12 and total homocysteine concentrations may be related to accelerated aging of the brain. Randomized clinical trials are needed to determine the importance of vitamin B12supplementation on slowing brain aging in older adults.
Notes from Kurzweill
This vitamin stops the aging process in organs, say Swiss researchers
A potential breakthrough for regenerative medicine, pending further studies
Improved muscle stem cell numbers and muscle function in NR-treated aged mice: Newly regenerated muscle fibers 7 days after muscle damage in aged mice (left: control group; right: fed NR). (Scale bar = 50 μm). (credit: Hongbo Zhang et al./Science) http://www.kurzweilai.net/images/improved-muscle-fibers.png
EPFL researchers have restored the ability of mice organs to regenerate and extend life by simply administering nicotinamide riboside (NR) to them.
NR has been shown in previous studies to be effective in boosting metabolism and treating a number of degenerative diseases. Now, an article by PhD student Hongbo Zhang published in Science also describes the restorative effects of NR on the functioning of stem cells for regenerating organs.
As in all mammals, as mice age, the regenerative capacity of certain organs (such as the liver and kidneys) and muscles (including the heart) diminishes. Their ability to repair them following an injury is also affected. This leads to many of the disorders typical of aging.
Mitochondria —> stem cells —> organs
To understand how the regeneration process deteriorates with age, Zhang teamed up with colleagues from ETH Zurich, the University of Zurich, and universities in Canada and Brazil. By using several biomarkers, they were able to identify the molecular chain that regulates how mitochondria — the “powerhouse” of the cell — function and how they change with age. “We were able to show for the first time that their ability to function properly was important for stem cells,” said Auwerx.
Under normal conditions, these stem cells, reacting to signals sent by the body, regenerate damaged organs by producing new specific cells. At least in young bodies. “We demonstrated that fatigue in stem cells was one of the main causes of poor regeneration or even degeneration in certain tissues or organs,” said Zhang.
How to revitalize stem cells
Which is why the researchers wanted to “revitalize” stem cells in the muscles of elderly mice. And they did so by precisely targeting the molecules that help the mitochondria to function properly. “We gave nicotinamide riboside to 2-year-old mice, which is an advanced age for them,” said Zhang.
“This substance, which is close to vitamin B3, is a precursor of NAD+, a molecule that plays a key role in mitochondrial activity. And our results are extremely promising: muscular regeneration is much better in mice that received NR, and they lived longer than the mice that didn’t get it.”
Parallel studies have revealed a comparable effect on stem cells of the brain and skin. “This work could have very important implications in the field of regenerative medicine,” said Auwerx. This work on the aging process also has potential for treating diseases that can affect — and be fatal — in young people, like muscular dystrophy (myopathy).
So far, no negative side effects have been observed following the use of NR, even at high doses. But while it appears to boost the functioning of all cells, it could include pathological ones, so further in-depth studies are required.
Abstract of NAD+ repletion improves mitochondrial and stem cell function and enhances life span in mice
Adult stem cells (SCs) are essential for tissue maintenance and regeneration yet are susceptible to senescence during aging. We demonstrate the importance of the amount of the oxidized form of cellular nicotinamide adenine dinucleotide (NAD+) and its impact on mitochondrial activity as a pivotal switch to modulate muscle SC (MuSC) senescence. Treatment with the NAD+ precursor nicotinamide riboside (NR) induced the mitochondrial unfolded protein response (UPRmt) and synthesis of prohibitin proteins, and this rejuvenated MuSCs in aged mice. NR also prevented MuSC senescence in the Mdx mouse model of muscular dystrophy. We furthermore demonstrate that NR delays senescence of neural SCs (NSCs) and melanocyte SCs (McSCs), and increased mouse lifespan. Strategies that conserve cellular NAD+ may reprogram dysfunctional SCs and improve lifespan in mammals.
Discriminating the gene target of a distal regulatory element from other nearby transcribed genes is a challenging problem with the potential to illuminate the causal underpinnings of complex diseases. We present TargetFinder, a computational method that reconstructs regulatory landscapes from diverse features along the genome. The resulting models accurately predict individual enhancer–promoter interactions across multiple cell lines with a false discovery rate up to 15 times smaller than that obtained using the closest gene. By evaluating the genomic features driving this accuracy, we uncover interactions between structural proteins, transcription factors, epigenetic modifications, and transcription that together distinguish interacting from non-interacting enhancer–promoter pairs. Most of this signature is not proximal to the enhancers and promoters but instead decorates the looping DNA. We conclude that complex but consistent combinations of marks on the one-dimensional genome encode the three-dimensional structure of fine-scale regulatory interactions.
. 
. Assuming that the exon-inclusion read count ICk follows a binomial distribution with the total read count nk
2012pharmaceutical
sjwilliamspa
