Posts Tagged ‘Exosomes’

The Role of Exosomes in Metabolic Regulation

Author: Larry H. Bernstein, MD, FCAP


On 9/25/2017, Aviva Lev-Ari, PhD, RN commissioned Dr. Larry H. Bernstein to write a short article on the following topic reported on 9/22/2017 in


We are publishing, below the new article created by Larry H. Bernstein, MD, FCAP.



During the period between 9/2015  and 6/2017 the Team at Leaders in Pharmaceutical Business Intelligence (LPBI)  has launched an R&D effort lead by Aviva Lev-Ari, PhD, RN in conjunction with SBH Sciences, Inc. headed by Dr. Raphael Nir.

This effort, also known as, “DrugDiscovery @LPBI Group”  has yielded several publications on EXOSOMES on this Open Access Online Scientific Journal. Among them are included the following:


QIAGEN – International Leader in NGS and RNA Sequencing, 10/08/2017

Reporter: Aviva Lev-Ari, PhD, RN


cell-free DNA (cfDNA) tests could become the ultimate “Molecular Stethoscope” that opens up a whole new way of practicing Medicine, 09/08/2017

Reporter: Aviva Lev-Ari, PhD, RN


Detecting Multiple Types of Cancer With a Single Blood Test (Human Exomes Galore), 07/02/2017

Reporter and Curator: Irina Robu, PhD


Exosomes: Natural Carriers for siRNA Delivery, 04/24/2017

Reporter: Aviva Lev-Ari, PhD, RN


One blood sample can be tested for a comprehensive array of cancer cell biomarkers: R&D at WPI, 01/05/2017

Curator: Marzan Khan, B.Sc


SBI’s Exosome Research Technologies, 12/29/2016

Reporter: Aviva Lev-Ari, PhD, RN


A novel 5-gene pancreatic adenocarcinoma classifier: Meta-analysis of transcriptome data – Clinical Genomics Research @BIDMC, 12/28/2016

Curator: Tilda Barliya, PhD


Liquid Biopsy Chip detects an array of metastatic cancer cell markers in blood – R&D @Worcester Polytechnic Institute, Micro and Nanotechnology Lab, 12/28/2016

Reporters: Tilda Barliya, PhD and Aviva Lev-Ari, PhD, RN


Exosomes – History and Promise, 04/28/2016

Reporter: Aviva Lev-Ari, PhD, RN


Exosomes, 11/17/2015

Curator: Larry H. Bernstein, MD, FCAP


Liquid Biopsy Assay May Predict Drug Resistance, 11/16/2015

Curator: Larry H. Bernstein, MD, FCAP


Glypican-1 identifies cancer exosomes, 10/31/2015

Curator: Larry H. Bernstein, MD, FCAP


Circulating Biomarkers World Congress, March 23-24, 2015, Boston: Exosomes, Microvesicles, Circulating DNA, Circulating RNA, Circulating Tumor Cells, Sample Preparation, 03/24/2015

Reporter: Aviva Lev-Ari, PhD, RN


Cambridge Healthtech Institute’s Second Annual Exosomes and Microvesicles as Biomarkers and Diagnostics Conference, March 16-17, 2015 in Cambridge, MA, 03/17, 2015

Reporter: Aviva Lev-Ari, PhD, RN


The newly created think-piece on the relationship between regulatory functions of Exosomes and Metabolic processes is developed conceptually, below.


The Role of Exosomes in Metabolic Regulation

Author: Larry H. Bernstein, MD, FCAP

We have had more than a half century of research into the genetic code and transcription leading to abundant work on RNA and proteomics. However, more recent work in the last two decades has identified RNA interference in siRNA. These molecules may be found in the circulation, but it has been a challenge to find their use in therapeutics. Exosomes were first discovered in the 1980s, but only recently there has been a huge amount of research into their origin, structure and function. Exosomes are 30–120 nm endocytic membrane-bound extracellular vesicles (EVs)(1-23) , and more specifically multiple vesicle bodies (MVBs) by a budding process from invagination of the outer cell membrane that carry microRNA (miRNA), and have structures composed of protein and lipids (1,23-27 ). EVs are the membrane vesicles secreted by eukaryotic cells for intracellular communication by transferring the proteins, lipids, and RNA under various physiologic conditions as well as during the disease stage. EVs also act as a signalosomes in many biological processes. Inward budding of the plasma membrane forms small vesicles that fuse. Intraluminal vesicles (ILVs) are formed by invagination of the limiting endosomal membrane during the maturation process of early endosome.

EVs are the MVBs secreted that serve in intracellular communication by transferring a cargo consisting of proteins, lipids, and RNA under various physiologic conditions (4, 23). Exosome-mediated miRNA transfer between cells is considered to be necessary for intercellular signaling and exosome-associated miRNAs in biofluids (23). Exosomes carry various molecular constituents of their cell of origin, including proteins, lipids, mRNAs, and microRNAs (miRNAs) (. They are released from many cell types, such as dendritic cells (DCs), lymphocytes, platelets, mast cells, epithelial cells, endothelial cells, and neurons, and can be found in most bodily fluids including blood, urine, saliva, amniotic fluid, breast milk, hydrothoracic fluid, and ascitic fluid, as well as in culture medium of most cell types.Exosomes have also been shown to be involved in noncoding RNA surveillance machinery in generating antibody diversity (24). There are also a vast number of long non-coding RNAs (lncRNAs) and enhancer RNAs (eRNAs) that accumulate R-loop structures upon RNA exosome ablation, thereby, resolving deleterious DNA/RNA hybrids arising from active enhancers and distal divergent eRNA-expressing elements (lncRNA-CSR) engaged in long-range DNA interactions (25). RNA exosomes are large multimeric 3′-5′ exo- and endonucleases representing the central RNA 3′-end processing factor and are implicated in processing, quality control, and turnover of both coding and noncoding RNAs. They are large macromolecular cages that channel RNA to the ribonuclease sites (29). A major interest has been developed to characterize of exosomal cargo, which includes numerous non-randomly packed proteins and nucleic acids (1). Moreover, exosomes play an active role in tumorigenesis, metastasis, and response to therapy through the transfer of oncogenes and onco-miRNAs between cancer cells and the tumor stroma. Blood cells and the vascular endothelium is also exosomal shedding, which has significance for cardiovascular,   neurologicological disorders, stroke, and antiphospholipid syndrome (1). Dysregulation of microRNAs and the affected pathways is seen in numerous pathologies their expression can reflect molecular processes of tumor onset and progression qualifying microRNAs as potential diagnostic and prognostic biomarkers (30).

Exosomes are secreted by many cells like B lymphocytes and dendritic cells of hematopoietic and non-hematopoietic origin viz. platelets, Schwann cells, neurons, mast cells, cytotoxic T cells, oligodendrocytes, intestinal epithelial cells were also found to be releasing exosomes (4). They are engaged in complex functions like persuading immune response as the exosomes secreted by antigen presenting cells activate T cells (4). They all have a common set of proteins e.g. Rab family of GTPases, Alix and ESCRT (required for transport) protein and they maintain their cytoskeleton dynamics and participate in membrane fusion. However, they are involved in retrovirus disease pathology as a result of recruitment of the host`s endosomal compartments in order to generate viral vesicles, and they can either spread or limit an infection based on the type of pathogen and its target cells (5).

Upon further consideration, it is understandable how this growing biological work on exosomes has enormous significance for laboratory diagnostics (1, 3, 5, 6, 11, 14, 15, 17-20, 23,30-41) . They are released from many cell types, such as dendritic cells (DCs), lymphocytes, platelets, mast cells, epithelial cells, endothelial cells, and neurons, and can be found in most bodily fluids including blood, urine, saliva, amniotic fluid, breast milk, thoracic and abdominal effusions, and ascitic fluid (1). The involvement of exosomes in disease is broad, and includes: cancer, autoimmune and infectious disease, hematologic disorders, neurodegenerative diseases, and cardiovascular disease. Proteins frequently identified in exosomes include membrane transporters and fusion proteins (e.g., GTPases, annexins, and flotillin), heat shock proteins (e.g., HSC70), tetraspanins (e.g., CD9, CD63, and CD81), MVB biogenesis proteins (e.g., alix and TSG101), and lipid-related proteins and phospholipases. The exosomal lipid composition has been thoroughly analyzed in exosomes secreted from several cell types including DCs and mast cells, reticulocytes, and B-lymphocytes (1). Dysregulation of microRNAs of pathways observed in numerous pathologies (5, 10, 12, 21, 27, 35, 37) including cancers (30), particularly, colon, pancreas, breast, liver, brain, lung (2, 6, 17-20, 30, 33-36, 38, 39). Following these considerations, it is important that we characterize the content of exosomal cargo to gain clues to their biogenesis, targeting, and cellular effects which may lead to identification of biomarkers for disease diagnosis, prognosis and response to treatment (42).

We might continue in pursuit of a particular noteworthy exosome, the NLRP3 inflammasome, which is activated by a variety of external or host-derived stimuli, thereby, initiating an inflammatory response through caspase-1 activation, resulting in inflammatory cytokine IL-1b maturation and secretion (43).
Inflammasomes are multi-protein signaling complexes that activate the inflammatory caspases and the maturation of interleukin-1b. The NLRP3 inflammasome is linked with human autoinflammatory and autoimmune diseases (44). This makes the NLRP3 inflammasome a promising target for anti-inflammatory therapies. The NLRP3 inflammasome is activated in response to a variety of signals that indicate tissue damage, metabolic stress, and infection (45). Upon activation, the NLRP3 inflammasome serves as a platform for activation of the cysteine protease caspase-1, which leads to the processing and secretion of the proinflammatory cytokines interleukin-1β (IL-1β) and IL-18. Heritable and acquired inflammatory diseases are both characterized by dysregulation of NLRP3 inflammasome activation (45).
Receptors of innate immunity recognize conserved moieties associated with either cellular damage [danger-associated molecular patterns (DAMPs)] or invading organisms [pathogen-associated molecular patterns (PAMPs)](45). Either chronic stimulation or overwhelming tissue damage is injurious and responsible for the pathology seen in a number of autoinflammatory and autoimmune disorders, such as arthritis and diabetes. The nucleotide-binding domain leucine-rich repeat (LRR)-containing receptors (NLRs) are PRRs are found intracellularly and they share a unique domain architecture. It consists of a central nucleotide binding and oligomerization domain called the NACHT domain that is located between an N-terminal effector domain and a C-terminal LRR domain (45). The NLR family members NLRP1, NLRP3, and NLRC4 are capable of forming multiprotein complexes called inflammasomes when activated.

The (NLRP3) inflammasome is important in chronic airway diseases such as asthma and chronic obstructive pulmonary disease because the activation results, in pro-IL-1β processing and the secretion of the proinflammatory cytokine IL-1β (46). It has been proposed that Activation of the NLRP3 inflammasome by invading pathogens may prove cell type-specific in exacerbations of airway inflammation in asthma (46). First, NLRP3 interacts with the adaptor protein ASC by sensing microbial pathogens and self-danger signals. Then pro-caspase-1 is recruited and the large protein complex called the NLRP3 inflammasome is formed. This is followed by autocleavage and activation of caspase-1, after which pro-IL-1β and pro-IL-18 are converted into their mature forms. Ion fluxes disrupt membrane integrity, and also mitochondrial damage both play key roles in NLRP3 inflammasome activation (47). Depletion of mitochondria as well as inhibitors that block mitochondrial respiration and ROS production prevented NLRP3 inflammasome activation. Futhermore, genetic ablation of VDAC channels (namely VDAC1 and VDAC3) that are located on the mitochondrial outer membrane and that are responsible for exchanging ions and metabolites with the cytoplasm, leads to diminished mitochondrial (mt) ROS production and inhibition of NLRP3 inflammasome activation (47). Inflammasome activation not only occurs in immune cells, primarily macrophages and dendritic cells, but also in kidney cells, specifically the renal tubular epithelium. The NLRP3 inflammasome is probably involved in the pathogenesis of acute kidney injury, chronic kidney disease, diabetic nephropathy and crystal-related nephropathy (48). The inflammasome also plays a role in autoimmune kidney disease. IL-1 blockade and two recently identified specific NLRP3 inflammasome blockers, MCC950 and β-hydroxybutyrate, may prove to have value in the treatment of inflammasome-mediated conditions.

Autophagosomes derived from tumor cells are referred to as defective ribosomal products in blebs (DRibbles). DRibbles mediate tumor regression by stimulating potent T-cell responses and, thus, have been used as therapeutic cancer vaccines in multiple preclinical cancer models (49). It has been found that DRibbles could induce a rapid differentiation of monocytes and DC precursor (pre-DC) cells into functional APCs (49). Consequently, DRibbles could potentially induce strong innate immune responses via multiple pattern recognition receptors. This explains why DRibbles might be excellent antigen carriers to induce adaptive immune responses to both tumor cells and viruses. This suggests that isolated autophagosomes (DRibbles) from antigen donor cells activate inflammasomes by providing the necessary signals required for IL-1β production.

The Hsp90 system is characterized by a cohort of co-chaperones that bind to Hsp90 and affect its function (50). The co-chaperones enable Hsp90 to chaperone structurally and functionally diverse client proteins. Sahasrabudhe et al. (50) show that the nature of the client protein dictates the contribution of a co-chaperone to its maturation. The study reveals the general importance of the cochaperone Sgt1 (50). In addition to Hsp90, we have to consider Hsp60. Adult cardiac myocytes release heat shock protein (HSP)60 in exosomes. Extracellular HSP60, when not in exosomes, causes cardiac myocyte apoptosis via the activation of Toll-like receptor 4. the protein content of cardiac exosomes differed significantly from other types of exosomes in the literature and contained cytosolic, sarcomeric, and mitochondrial proteins (21).

A new Protein Organic Solvent Precipitation (PROSPR) method efficiently isolates the EV repertoire from human biological samples. Proteomic profiling of PROSPR-enriched CNS EVs indicated that > 75 % of the proteins identified matched previously reported exosomal and microvesicle cargoes. In addition lipidomic characterization of enriched CNS vesicles identified previously reported EV-specific lipid families and novel lipid isoforms not previously detected in human EVs. The characterization of these structures from central nervous system (CNS) tissues is relevant to current neuroscience, especially to advance the understanding of neurodegeneration in amyotrophic lateral sclerosis (ALS), Parkinson’s disease (PD) and Alzheimer’s disease (AD)(15). In addition, study of EVs in brain will enable characterization of the degenerative posttranslational modifications (DPMs) occurring in those proteins.
Neurodegenerative disease is characterized by dysregulation because of NLRP3 inflammasome activation. Alzheimer’s disease (AD) and Parkinson’s disease (PD), both neurodegenerative diseases are associated with the NLRP3 inflammasome. PD is characterized by accumulation of Lewy bodies (LB) formed by a-synuclein (aSyn) aggregation. A recent study revealed that aSyn induces synthesis of pro-IL-1b by an interaction with TLR2 and activates NLRP3 inflammasome resulting in caspase-1 activation and IL-1b maturation in human primary monocytes (43). In addition mitophagy downregulates NLRP3 inflammasome activation by eliminating damaged mitochondria, blocking NLRP3 inflammasome activating signals. It is notable that in this aberrant activation mitophagy downregulates NLRP3 inflammasome activation by eliminating damaged mitochondria, blocking NLRP3 inflammasome activating signals (43).


  1. Lin J, Li J, Huang B, Liu J, Chen X. Exosomes: Novel Biomarkers for Clinical Diagnosis. Scie World J 2015; Article ID 657086, 8 pages
  2. Kahlert C, Melo SA, Protopopov A, Tang J, Seth S, et al. Identification of Double-stranded Genomic DNA Spanning All Chromosomes with Mutated KRAS and p53 DNA in the Serum Exosomes of Patients with Pancreatic Cancer. J Biol Chem 2014; 289: 3869-3875. doi: 10.1074/jbc.C113.532267.
  3. Lässer C, Eldh M, Lötvall J. Isolation and Characterization of RNA-Containing Exosomes. J. Vis. Exp. 2012; 59, e3037. doi:10.3791/3037(2012).
  4. Kaur A, Leishangthem GD, Bhat P, et al. Role of Exosomes in Pathology – A Review. Journal of Pathology and Toxicology 2014; 1: 07-11
  5. Hosseini HM, Fooladi AAI, Nourani MR and Ghanezadeh F. The Role of Exosomes in Infectious Diseases. Inflammation & Allergy – Drug Targets 2013; 12:29-37.
  6. Ciregia F, Urbani A and Palmisano G. Extracellular Vesicles in Brain Tumors and Neurodegenerative Diseases. Front. Mol. Neurosci. 2017;10:276. doi: 10.3389/fnmol.2017.00276
  7. Zhang B, Yin Y, Lai RC, Lim SK. Immunotherapeutic potential of extracellular vesicles. Front Immunol (2014)
  8. Kowal J, Tkach M, Théry C. Biogenesis and secretion of exosomes. Current Opin in Cell Biol 2014 Aug; 29: 116-125.
  9. McKelvey KJ, Powell KL, Ashton AW, Morris JM and McCracken SA. Exosomes: Mechanisms of Uptake. J Circ Biomark, 2015; 4:7   DOI: 10.5772/61186
  10. Xiao T, Zhang W, Jiao B, Pan C-Z, Liu X and Shen L. The role of exosomes in the pathogenesis of Alzheimer’ disease. Translational Neurodegen 2017; 6:3. DOI 10.1186/s40035-017-0072-x
  11. Gonzales PA, Pisitkun T, Hoffert JD, et al. Large-Scale Proteomics and Phosphoproteomics of Urinary Exosomes. J Am Soc Nephrol 2009; 20: 363–379. doi: 10.1681/ASN.2008040406
  12. Waldenström A, Ronquist G. Role of Exosomes in Myocardial Remodeling. Circ Res. 2014; 114:315-324.
  13. Xin H, Li Y and Chopp M. Exosomes/miRNAs as mediating cell-based therapy of stroke. Front. Cell. Neurosci. 10 Nov, 2014; 8(377) doi: 10.3389/fncel.2014.00377
  14. Wang S, Zhang L, Wan S, Cansiz S, Cui C, et al. Aptasensor with Expanded Nucleotide Using DNA Nanotetrahedra for Electrochemical Detection of Cancerous Exosomes. ACS Nano, 2017; 11(4):3943–3949 DOI: 10.1021/acsnano.7b00373
  15. Gallart-Palau X, Serra A, Sze SK. (2016) Enrichment of extracellular vesicles from tissues of the central nervous system by PROSPR. Mol Neurodegener 11(1):41.
  16. Simpson RJ, Jensen SS, Lim JW. Proteomic profiling of exosomes: current perspectives. Proteomics. 2008 Oct; 8(19):4083-99. doi: 10.1002/pmic.200800109.
  17. Sandfeld-Paulsen R, Aggerholm-Pedersen N, Bæk R, Jakobs KR, et al. Exosomal proteins as prognostic biomarkers in non-small cell lung cancer. Mol Onc 2016 Dec; 10(10):1595-1602.
  18. Li W, Li C, Zhou T, et al. Role of exosomal proteins in cancer diagnosis. Molecular Cancer 2017; 16:145 DOI 10.1186/s12943-017-0706-8
  19. Zhang W, Xia W, Lv Z, Xin Y, Ni C, Yang L. Liquid Biopsy for Cancer: Circulating Tumor Cells, Circulating Free DNA or Exosomes? Cell Physiol Biochem 2017; 41:755-768. DOI: 10.1159/00045873
  20. Thakur BK ,…, Williams C, Rodriguez-Barrueco R, Silva JM, Zhang W, et al. Double-stranded DNA in exosomes: a novel biomarker in cancer detection. Cell Research 2014 June; 24(6):766-769. doi:10.1038/cr.2014.44.
  21. Malik ZA, Kott KS, Poe AJ, Kuo T, Chen L, Ferrara KW, Knowlton AA. Cardiac myocyte exosomes: stability, HSP60, and proteomics. Am J Physiol Heart Circ Physiol 304: H954–H965, 2013. doi:10.1152/ajpheart.00835.2012.
  22. De Toro J, Herschlik L, Waldner C and Mongini C. Emerging roles of exosomes in normal and pathological conditions: new insights for diagnosis and therapeutic applications. Front. Immunol. 2015; 6:203. doi: 10.3389/fimmu.2015.00203
  23. Chevilleta JR, Kanga Q, Rufa IK, Briggs HA, et al. Quantitative and stoichiometric analysis of the microRNA content of exosomes. PNAS 2014 Oct 14; 111(41): 14888–14893.
  24. Basu U, Meng F-L, Keim C, Grinstein V, Pefanis E, et al. The RNA Exosome Targets the AID Cytidine Deaminase to Both Strands of Transcribed Duplex DNA Substrates. Cell 2011; 144: 353–363, DOI 10.1016/j.cell.2011.01.001
  25. Pefanis E, Wang J, …, Rabadan R, Basu U. RNA Exosome-Regulated Long Non-Coding RNA Transcription Controls Super-Enhancer Activity. Cell 2015; 161: 774–789.
  26. Kilchert C,Wittmann S & Vasiljeva L. The regulation and functions of the nuclear RNA exosome complex. In RNA processing and modifications. Nature Reviews Molecular Cell Biology 17, 227–239 (2016) doi:10.1038/nrm.2015.15
  27. Guay C, Regazzi R. Exosomes as new players in metabolic organ cross-talk. Diabetes Obes Metab. 2017;19(Suppl. 1):137–146. DOI: 10.1111/dom.13027.
  28. Abramowicz A, Widlak P, Pietrowska M. Proteomic analysis of exosomal cargo: the challenge of high purity vesicle isolation. Molecular BioSystems MB-REV-02-2016-000082.R1
  29. Hopfner K-P, Hartung S. The RNA Exosomes. In Nucleic Acids and Molecular Biology. 2011. Ribonucleases pp 223-244.
  30. Fuessel S, Lohse-Fischer A, Vu Van D, Salomo K, Erdmann K, Wirth MP. (2017) Quantification of MicroRNAs in Urine-Derived Specimens. In Urothelial Carcinoma, Methods Mol Biol 1655:201-226.
  31. Street JM, Barran PE, Mackay CL, Weidt S, et al. Identification and proteomic profiling of exosomes in human cerebrospinal fluid. Journal of Translational Medicine 2012; 10:5.
  32. Pisitkun T, Shen R-F, and Knepper MA. Identification and proteomic profiling of exosomes in human urine. PNAS 2004, Sept 7; 101(36): 13368–13373.
  33. Duijvesz D, Burnum-Johnson KE, Gritsenko MA, Hoogland AM, Vredenbregt-van den Berg MS, et al. Proteomic Profiling of Exosomes Leads to the Identification of Novel Biomarkers for Prostate Cancer. PLoS ONE 2013; 8(12): e82589. doi:10.1371/journal.pone.0082589
  34. Welton JL, Khanna S, Giles PJ, Brennan P, et al. Proteomics Analysis of Bladder Cancer Exosomes. Molecular & Cellular Proteomics 2010; 9:1324–1338. DOI 10.1074/mcp.M000063-MCP201
  35. Lee S, Suh G-Y, Ryter SW, and Choi AMK. Regulation and Function of the Nucleotide Binding Domain Leucine-Rich Repeat-Containing Receptor, PyrinDomain-Containing-3 Inflammasome in Lung Disease. Am J Respir Cell Mol Biol 2016 Feb; 54(2):151–160. DOI: 10.1165/rcmb.2015-0231TR.
  36. Zhang X, Yuan X, Shi H, Wu L, Qian H, Xu W. Exosomes in cancer: small particle, big player. J Hematol Oncol (2015)
  37. Zhao X, Wu Y, Duan J, Ma Y, Shen Z, et al. Quantitative Proteomic Analysis of Exosome Protein Content Changes Induced by Hepatitis B Virus in Huh-7 Cells Using SILAC Labeling and LC–MS/MS. J. Proteome Res.; 2014, 13 (12):5391–5402. DOI: 10.1021/pr5008703
  38. Liang B, Peng P, et al. Characterization and proteomic analysis of ovarian cancer-derived exosomes. J Proteomics. 2013 Mar; 80:171-182.
  39. Beckler MD, Higginbotham JN, Franklin JL,…, Li M, Liebler DC, Coffey RJ. Proteomic analysis of exosomes from mutant KRAS colon cancer cells identifies intercellular transfer of mutant KRAS. Mol. Cell Proteomics. 2013 Feb 12; (2).
  40. Alvarez-Llamas G, Díaz J, Zubiri I. Proteome of Human Urinary Exosomes in Diabetic Nephropathy. In Biomarkers in Kidney Disease. Vinood B. Patel, Ed. Springer Science 2015; pp 1-21. DOI 10.1007/978-94-007-7743-9_22-1
  41. Simpson RJ, Jensen SS, Lim JW. Proteomic profiling of exosomes: current perspectives. Proteomics. 2008 Oct; 8(19):4083-99. doi: 10.1002/pmic.200800109.
  42. Scheya JKL, Luther M, Rose KL. Proteomics characterization of exosome cargo. Methods 2015 Oct; 87(1): 75-82.
  43. Kim M-J, Yoon J-H & Ryu J-H. Mitophagy: a balance regulator of NLRP3 inflammasome Activation. BMB Rep. 2016; 49(10): 529-535.
  44. Eun-Kyeong Jo, Kim JK, Shin D-M and C Sasakawa. Molecular mechanisms regulating NLRP3 inflammasome activation. Cell Molec Immunol 2016; 13: 148–159. doi:10.1038/cmi.2015.95
  45. Leemans JC, Cassel SL, and Sutterwala FS. Sensing damage by the NLRP3 inflammasome. Immunol Rev. 2011 Sept; 243(1): 152–162. doi:10.1111/j.1600-065X.2011.01043.x.
  46. Hirota JA, Im H, Rahman MM, Rumzhum NN, Manetsch M, Pascoe CD, Bunge K, Alkhouri H, Oliver BG, Ammit AJ. The nucleotide-binding domain and leucine-rich repeat protein-3 inflammasome is not activated in airway smooth muscle upon toll-like receptor-2 ligation. Am J Respir Cell Mol Biol. 2013 Oct; 49(4):517-24. doi: 10.1165/rcmb.2013-0047OC.
  47. Zhong Z, Sanchez-Lopez E, Karin M. Autophagy, NLRP3 inflammasome and auto-inflammatory immune diseases. Clin Exp Rheumatol. 2016 Jul-Aug; 34(4 Suppl 98):12-6. Epub 2016 Jul 21.
  48. Hutton HL, Ooi JD, Holdsworth SR, Kitching AR. The NLRP3 inflammasome in kidney disease and autoimmunity. Nephrology (Carlton). 2016 Sep; 21(9):736-44. doi: 10.1111/nep.12785
  49. Xing Y, Cao R and Hu H-M. TLR and NLRP3 inflammasome-dependent innate immune responses to tumor-derived autophagosomes (DRibbles). Cell Death and Disease (2016) 7, e2322; doi:10.1038/cddis.2016.206
  50. Sahasrabudhe P, Rohrberg J, Biebl MM, Rutz DA, Buchner J. The Plasticity of the Hsp90 Co-chaperone System. Molecular Cell 2017 Sept; 67:947–961.



Read Full Post »

mRNA data survival analysis

Curators: Larry H. Bernstein, MD, FCAP and Aviva Lev-Ari, PhD, RN



SURVIV for survival analysis of mRNA isoform variation

Shihao ShenYuanyuan WangChengyang WangYing Nian Wu & Yi Xing
Nature Communications7,Article number:11548
 Feb 2016      doi:10.1038/ncomms11548

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.

TABLE 1 (not shown)

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.

Network of survival-associated alternative splicing events


Figure 6: Splicing factor regulatory network of survival-associated alternative splicing events in IDC.

(ac) 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.


Alternative splicing predictors of cancer patient survival


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 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.


Read Full Post »

Lipids link to breast cancer

Larry H. Bernstein, MD, FCAP, Curator



Lipids Found Critical to Breast Cancer Cell Proliferation


Scientists in Spain report finding that breast cancer cells need to take up lipids from the extracellular environment so that they can continue to proliferate. The main protein involved in this process is LIPG, an enzyme found in the cell membrane and without which tumor cell growth is arrested. Analyses of more than 500 clinical samples from patients with various kinds of breast tumors reveal that 85% have high levels of LIPG expression.

The research (“FoxA and LIPG Endothelial Lipase Control the Uptake of Extracellular Lipids for Breast Cancer Growth”) is published in Nature Communications.

In Spain, breast cancer is the most common tumor in women and the fourth most common type in both sexes (data from the Spanish Society of Medical Oncology, 2012), registering more than 25,000 new diagnoses each year. According to figures from the World Health Organization, every year 1.38 million new cases of breast cancer are diagnosed and 458,000 people die from this disease (International Agency for Research on Cancer Globocan, 2008).

It was already known that cancer cells require extracellular glucose to grow and that they reprogram their internal machinery to produce greater amounts of lipids. The relevance of this study is that it reveals for the first time that tumor cells must import extracellular lipids to grow.

“This new knowledge related to metabolism could be the Achilles heel of breast cancer,” explains ICREA researcher and Institute for Research in Biomedicine–Barcelona group leader Roger Gomis, Ph.D., co-leader of the study together with Joan J. Guinovart, Ph.D., director of IRB Barcelona and professor at the University of Barcelona. Using animal models and cancer cell cultures, the scientists have demonstrated that blocking of LIPG activity arrests tumor growth.

“What is promising about this new therapeutic target is that LIPG function does not appear to be indispensable for life, so its inhibition may have fewer side effects than other treatments,” explains the first author of the study, Felipe Slebe, a Ph.D. Fellow at IRB Barcelona.

According to Dr. Guinovart, “because LIPG is a membrane protein, it is potentially easier to design a pharmacological agent to block its activity.”

“If a drug were found to block its activity, it could be used to develop more efficient chemotherapy treatments that are less toxic than those currently available,” adds Dr. Gomis.

The scientists are now looking into international collaborations for developing LIPG inhibitors.

FoxA and LIPG endothelial lipase control the uptake of extracellular lipids for breast cancer growth

Felipe SlebeFederico RojoMaria Vinaixa,…Joan AlbanellJoan J. Guinovart & Roger R. Gomis

Nature Communications7,Article number:11199

The mechanisms that allow breast cancer (BCa) cells to metabolically sustain rapid growth are poorly understood. Here we report that BCa cells are dependent on a mechanism to supply precursors for intracellular lipid production derived from extracellular sources and that the endothelial lipase (LIPG) fulfils this function. LIPG expression allows the import of lipid precursors, thereby contributing to BCa proliferation. LIPG stands out as an essential component of the lipid metabolic adaptations that BCa cells, and not normal tissue, must undergo to support high proliferation rates. LIPG is ubiquitously and highly expressed under the control of FoxA1 or FoxA2 in all BCa subtypes. The downregulation of either LIPG or FoxA in transformed cells results in decreased proliferation and impaired synthesis of intracellular lipids.

FoxA1 and FoxA2 in BCa growth

The importance of FoxA1 in BCa cells differentiation and its contribution to controlling the expression of metabolic genes in several other tissues makes this transcription factor a highly attractive target to explain the metabolic alterations reported in BCa. For these reason, we decided to ascertain the metabolic processes controlled by FoxA1 in BCa. We first confirmed the association between high FoxA1 expression (mRNA and protein) and luminal subtype (Fig. 1a). To this end, we used two cohorts of primary breast tumours with annotated clinical features and follow-up. The MSKCC/EMC BCa data set is based on gene expression profiles from an original series of 560 cases10, whereas the Spanish BCa data set (n=439) is a tissue microarray of formalin-fixed paraffin-embedded stage I–III breast tumour specimens11 (details provided in Methods Section). High FoxA1 gene expression significantly correlated with high expression of well-established luminal markers, such as GATA3 and ESR1, in primary tumours (Supplementary Fig. 1a). Next we explored FoxA1 expression beyond the luminal subtype. Lower FoxA1 expression was observed in non-luminal tumours (Fig. 1a,b); however, a subset also expressed higher FoxA1 levels (Supplementary Fig. 1b and Supplementary Table 1). Given that FoxA2, in conjunction with FoxA1, is also involved in the regulation of several metabolic pathways, we determined the expression of this factor in BCa samples. Unfortunately, no FoxA2 probes in the Affymetrix platform used in the MSKCC/EMC data set provided a reliable interpretation. To overcome this limitation, we used tissue arrays of early BCa samples (Spanish BCa set). Histological examination of FoxA2-stained tissue microarray slides from the Spanish BCa set revealed the expression of this factor in six non-luminal samples, which were scored as FoxA1 (examples in Fig. 1b and summarized inSupplementary Table 1). Collectively, the number of FoxA+ BCa samples detected by immunohistochemistry accounted for 81.3% of all samples in the Spanish BCa set (Supplementary Table 1), which represent a significant proportion of BCa and point to the participation of FoxA in this disease, beyond to its involvement in differentiation and control of hormonal responses.

Figure 1: FoxA1 and FoxA2 in BCa growth.

(a, top) FoxA1 mRNA expression in the MSKCC/EMC set. BCa samples were stratified in Luminal A, Luminal B, Her2, triple negative and unknown subgroups. The unknown group represents specimens that were not classified in any group. (bottom) FoxA1 protein levels by IHC staining in Luminal, Her2 and triple negative samples in the Spanish BCa set (cohort of 439 BCa patients). Data is average±s.d. (b) FoxA1 and FoxA2 IHC staining in FFPE human specimens representative of the different BCa subtypes. Six independent cases are depicted. FoxA1 and FoxA2 are expressed mainly in the nuclei of tumour cells. Scale bar, 50μm. (c) FoxA1 and FoxA2 mRNA expression analysis by qRT-PCR and protein expression by western blot in human BCa cell lines compared with HMECs. T-test was used. Data are average±s.e.m.; n= 3. Of note, MDA435 are of melanoma origin. (d) FoxA1 and FoxA2 expression in MCF7, MDA231 and their derivatives cells by qRT-PCR and western blot. FoxA1 and FoxA2 depletion was achieved with a doxycycline-inducible short hairpin vector. FoxA-depleted cells were rescued by expression of FoxA2 in MCF7 cells or FoxA1 in MDA231 cells. Cell populations were cultured in the presence or absence of doxycycline for 6 days. P value is the result of T-test. Data are average±s.e.m.;n=3. *P≤0.05, ***P≤0.001 (e, left) Schematic representation of MDA231 and MCF7 cells grown without doxycycline and inoculated in Balb/c nude mice treated with or without doxycycline to induce the expression of the indicated FoxA short hairpins. All tumour cell lines have GFP constitutive expression, and tRFP concomitantly with the short hairpin were expressed in doxycycline treated tumours. (right) Tumour growth of the indicated cell populations inoculated in Balb/c nude mice are determined at the indicated time points. P value is the result of T-test. Data are average±s.e.m.; n= 5–8 tumours. *P≤0.05,**P≤0.01, ***P≤0.001. FFPE, formalin-fixed paraffin-embedded.

Next, we extended our analysis to BCa cell lines for further mechanistic studies. We compared FoxA1 and FoxA2 mRNA expression in four estrogen receptor positive (ER+) (MCF7, T47D, BT474 and ZR75) and four estrogen receptor negative (ER−) (SKBR3, MDA468, BT20 and MDA231) BCa cell lines, a cell line of melanoma origin (MDA435), and human mammary epithelial cells (HMECs). Of note, two of the BCa lines tested were HER2+ (BT474 and SKBR3) (Fig. 1c). All ER+ BCa cells (MCF7, T47D, BT474 and ZR75), the ER−/HER2+ SKBR3 and both triple negative-like MDA468 and BT20 cell lines expressed FoxA1. Interestingly, MDA231 triple negative-like cells expressed high levels of FoxA2 but not FoxA1, and the non-tumour HMECs did not express these factors (Fig. 1c). No BCa cells co-expressed these two proteins (Fig. 1c). Our results suggest that the expression of FoxA transcription factors is a common feature of breast tumours, as well as of BCa cell lines. This notion implies that FoxA factors play a major role in BCa growth, independently of luminal fate specification.

To examine the molecular basis of the contribution of FoxA1 and FoxA2 to BCa growth, we engineered constitutive GFP-luciferase-expressing MCF7 and MDA231 cells with a doxycycline-inducible short-hairpin RNA (sh-RNA) vector targeting either FoxA1 or FoxA2. Doxycycline addition to the cell culture media decreased FoxA expression in both cell lines compared with control cells (ShControl (Dox+) and Sh FoxA1 or Sh FoxA2 (Dox−))(Fig. 1d), with the concomitant expression of tRFP (Supplementary Fig. 1c). Of note, there was no gain of expression of FoxA2 in FoxA1-depleted cells or vice versa (Fig. 1d). Interestingly, cancer cell proliferation was impaired in vitroupon depletion of either FoxA1 or FoxA2 in MCF7 and MDA231 cells, respectively (Supplementary Fig. 1d,e). Similarly, when Balb/c nude mice implanted with xenograft tumours from the above described cellular populations were treated with doxycycline and the short hairpins were induced, striking differences in tumour growth were observed. FoxA1-depleted MCF7 and FoxA2-depleted MDA231 tumour growth was blunted (Fig. 1e and additional controls in Supplementary Fig. 1f. Experimental details in the Supplementary Methods Section). Collectively, these observations confirm that FoxA1 or FoxA2 expression is required for BCa growth.

Previous studies indicate that FoxA1 and FoxA2 transcriptionally regulate common genes in the liver and pancreas that are central to development and metabolism. We therefore hypothesized that crossed expression of FoxA factors could rescue tumour growth by restoring the expression of essential metabolic genes. To this end, we engineered doxycycline-driven shFoxA1 MCF7 cells to express exogenous FoxA2 and doxycycline-driven shFoxA2 MDA231 cells to express exogenous FoxA1 (Fig. 1d). Interestingly, when these BCa modified cells were implanted in Balb/c nude mice and FoxA depletion was induced with doxycycline, the sustained expression of another FoxA factor (FoxA2 in MCF7 and FoxA1 in MDA231 cells) was sufficient for tumours to continuously grow (Fig. 1e and additional controls in Supplementary Fig. 1f). Quantitative real-time PCR (qRT-PCR) analysis confirmed FoxA expression in the distinct tumour populations ex-vivo (Supplementary Fig. 1g). These results showed that retention of minimal levels of FoxA1 or FoxA2 expression is necessary for BCa cell growth.

FoxA1- and FoxA2-regulated transcripts for BCa growth

Figure 2: A genomic approach to identify FoxA1- and FoxA2-regulated transcripts in MCF7 and MDA231 cells.

(a) FACS profiling of MCF7 and MDA231 cells derived from tumours isolated from mice on the basis of the expression of GFP+ and RFP− (control group) or GFP+ and tRFP+ (knockdown and rescue groups). (b) Representation of the transcripts up- and downregulated by FoxA in MCF7 and MDA231 cells isolated from tumours. Up- and downregulated transcripts present a Bayesian false discovery rate below 5% and fold change >2.5. (c) LIPG, Bcl2 and Cdh11OB mRNA levels of the indicated genetically modified MCF7 and MDA231 tumour xenografts analysed by qRT-PCR. P value is the result of T-test. Data are average±s.e.m.; n= 5–8 tumours. *P≤0.05, ***P≤0.001. (d) LIPG protein expression in constitutive shFoxA1 MCF7 or shFoxA2 MDA231 cells. (e) Promoter reporter assay in HEK 293 cells. Cells were transfected with LIPG promoter reporter and FoxA1 or FoxA2 expressing vectors when indicated. P value is the result of T-test. Data are average±s.e.m.; n=3. ****P≤0.0001.

LIPG expression in BCa

Next, we showed that LIPG expression in primary tumours was specific to BCa tumour cells and not to other stroma cellular entities (Fig. 3a). Subsequently, we tested LIPG expression in normal breast epithelia and interrogated 20 samples from mammoplasty reductions. Normal breast epithelial cells showed a lower expression of LIPG than cells from tumour specimens (Fig. 3b). Similar results were obtained for LIPG protein levels in a panel from BCa lines compared with HMEC cells. Of the cellular populations tested, the eight BCa cell lines expressing FoxA1 or FoxA2 had very high levels of LIPG protein compared with the melanoma MDA435 cell line and the human epithelial cell (Fig. 3c). Consistent with this observation, 83.8% of BCa samples in the Spanish tumour cohort were LIPG+ (Fig. 3d and Supplementary Table 3), and LIPG expression correlated with FoxA expression (Spearman correlation; r=0.477, P=0.000001; Fig. 3e). Further analysis showed that LIPG expression levels in primary tumours do not have the capacity to stratify patients for differential risk of overall or disease-free survival (Supplementary Fig. 2a) and are not dependent on estrogen signalling (Supplementary Fig. 2b), thus reinforcing the notion that LIPG is essential for BCa growth.

Figure 3: LIPG contributes to BCa growth.

a) Representative LIPG IHC staining on primary BCa tissues (cohort of 439 BCa patients). LIPG is expressed in the cytoplasm of tumour cells. Faint staining is also detected in the extracellular area. Scale bar, 50μm. (b) Representative LIPG IHC staining in normal breast tissue from mammoplasty reductions. Weak LIPG expression occurs in epithelial cells from ducts and lobuli. Scale bar, 50μm. (c) LIPG protein expression in human cancer cell lines compared with HMECs. Actin was used as loading control.*Unspecific band. Of note, MDA435 are of melanoma origin. (d) LIPG protein levels by IHC staining in Luminal, Her2, and triple negative samples in the Spanish BCa set (cohort of 439 BCa patients). Data is average±s.d. (e) Spearman correlation (P=0.000001) between FoxA and LIPG IHC staining intensities in Spanish BCa set (cohort of 439 BCa patients). (f) Left panel, in vitro proliferation curves of MCF7 and MDA231 cells transduced with a control or a LIPG short hairpin. Data are average±s.e.m.; n=3. (right) LIPG protein expression in shLIPG MCF7 and shLIPG MDA231 cells. The blot shown is representative of three independent experiments. P value is the result of T-test.**P≤0.01, ***P≤0.001. (g) Tumour growth of the indicated cell populations inoculated in Balb/c nude mice are determined at the indicated time points.P value is the result of T-test. Data are average±s.e.m.; n= 6–8 tumours. *P≤0.05.

LIPG is a phospholipase located in the cytosol and cellular membrane and has been shown to hydrolyse extracellular phospholipids from high-density lipoprotein that are afterwards incorporated into intracellular lipid species thus providing lipid precursors of cell metabolism17, 18. Thus we questioned whether LIPG regulates essential lipid intake in BCa and whether it is necessary for proliferation. To validate this hypothesis, we genetically downregulated the expression of this protein in MCF7 and MDA231 cells by means of sh-RNA (Fig. 3f and Supplementary Fig. 2c). LIPG depletion blunted BCa cell capacity to proliferate in vitro (Fig. 3f), as previously observed in FoxA-depleted cells (Supplementary Fig. 1d,e), and caused a reduction in invasion and self-renewal properties (Supplementary Fig. 3a–d). Similarly, LIPG-depleted cells were unable to grow tumours in vivo (Fig. 3g).

LIPG induces BCa cells lipid metabolic reprograming

Figure 4: LIPG regulates the uptake of lipids in BCa cells inducing a lipid metabolic reprograming.

LIPG regulates the uptake of lipids in BCa cells inducing a lipid metabolic reprograming.

(a) Schematic representation of LIPG action. (b) Heat map representation of the downregulated (blue) lipids identified by MS/MS in the cell homogenates of MCF7 or MDA231 LIPG-depleted cells compared with shControl cells. Depicted lipids have a fold change >1.5 and P value<0.05 using the Welch’s t-testn=5. (c) Downregulated lipid species (previously identified in b) that are common to LIPG-depleted MCF7 and LIPG-depleted MDA231 cells. ShControl cells (red box), and shLIPG (blue box). P values are <0.05 and calculated using Welch’s t-test, n=5. Whiskers extend to a maximum of 1.5 × IQR beyond the box. (d) Heat map representation of the upregulated (red) lipids identified by MS/MS in the media of MCF7 or MDA231 LIPG-depleted cells compared with the corresponding shControl cells. Characterized lipids have a fold change >1.5 and P value<0.05 using the Welch’s t-test n=5. (e) Upregulated lipid species in the media (previously identified in d) that are common to LIPG-depleted MCF7 and LIPG-depleted MDA231 cells. ShControl cells (red box), and shLIPG cells (blue box). P values are <0.05 and calculated using Welch’s t-test, n=5. Whiskers extend to a maximum of 1.5 × IQR beyond the box. (f) Heat map representation of the MS/MS downregulated (blue) lipids in the cell media of MCF7/MDA231 LIPG-depleted or shControl cells (as described in d) compared with fresh medium (without cell incubation). Depicted lipid species have a log2 fold change>1.5 and P value<0.05 using the Welch’s t-test n=5. (g) MDA231 and MCF7 cell growth for 48h in complete medium: medium containing 10% FBS 10%); lipoprotein-free medium: medium containing 10% free lipoprotein FBS; and LPC (18:0): medium containing 10% free lipoprotein FBS and 20μM of LPC (18:0). P value is the result of T-test. Data are average±s.e.m.; n=3. **P≤0.01, ***P≤0.001, ****P≤0.0001. (h) Above, schematic representation of the experimental protocol used. (bottom) Tumour growth of the indicated cell populations inoculated in Balb/c nude mice treated with high-fat diet (HFD) are determined at the indicated time points. P value is the result of T-test. Data are average±s.e.m.; n= 6–8 tumours. *P≤0.05, **P≤0.01. Inside graph, plasma cholesterol levels of animals treated with standard diet (SD) or HFD. P value is the result of T-test. Data are average±s.e.m.; n= 4 animals per group. **P≤0.01, ***P≤0.001.

LIPG location has been shown to be functional on the outer face of the cellular membrane (Fig. 4a)18, thus we postulated the possibility that BCa cells are dependent on LIPG function to access extracellular lipids to support their growth needs. To test this notion, we profiled the media of control and LIPG-depleted MCF7 and MDA231 cells following the same liquid chromatography-mass spectrometry-based untargeted lipidomic approach as for cell homogenates. LIPG depletion prevented the absorption of particular lipids from the media (Supplementary Fig. 4a). The structural identification of the lipids by MS/MS confirms the absence of degradation of glycerophospholipids belonging to the LPC class in both MCF7 and MDA231 cells, which is depicted by higher levels in the media of these species in LIPG-depleted when compared with control cells (Fig. 4d,e). Interestingly when we analysed the LPCs species in the media of control and LIPG-depleted cells and compared with fresh media (without cells), all LPC species from control cell media were decreased. This reduction was weaker in the media of Sh LIPG cells, indicating that LIPG-depleted cells have a defect in processing and importing of pre-existing lipid species from the medium (Fig. 4f).

Finally, we evaluated which of the commonly identified potential substrates of LIPG sustains BCa cell proliferation. Initially, we confirmed that the growth of MCF7 and MDA231 cells is impaired when grown in vitro in lipoprotein-depleted media (Fig. 4g). Next we tested the capacity of LPC (18:0) to rescue BCa cell growth in the absence of lipoproteins and confirmed that this lysophosphatidylcholine was able to restore the cells’ capacity to proliferate (Fig. 4g). In accordance, this process was dependent on LIPG expression (Fig. 4g). Similarly, LIPG-depleted cells were not able to grow in vivo in animals fed with high-fat diet (Fig. 4h) indicating that LIPG is indispensable to process the extracellular lipids and mediate their uptake by the cells, irrespectively of the concentration of lipid substrates in circulation, a phenotype also observed in FoxA-depleted cells (Fig. 4h).

LIPG activity supports BCa growth

Figure 5: LIPG activity is essential for BCa growth.

LIPG activity is essential for BCa growth.

(a, top) Homology 3D structural model of LIPG (backbone coloured according to the QMEANlocal parameter values; red residues with low error). The heavy atoms of the three catalytic residues are shown explicitly and the residue mutated in this study is shown in green (Asp 193). (b) FoxA1, FoxA2 and LIPG protein expression in MCF7, MDA231 and their derivative cells determine by western blot. FoxA1 and FoxA2 depletion was achieved with a doxycycline-inducible short hairpin vector. FoxA-depleted cells were rescued by expression of a WT or Inactive LIPG. Cell populations were cultured in the presence or absence of doxycycline for 6 days. *blots represent different exposition times. (c) Tumour growth of the indicated cell populations inoculated in Balb/c nude mice are determined at the indicated time points. Pvalue is the result of T-test. Data are average±s.e.m.; n=5–8 tumours. *P≤0.05, **P≤0.01. (d) MDA231 and MCF7 cell growth for 48h treated with DMSO (control), FAS inhibitor (C75) and/or lipase inhibitor (Orlistat). For MDA231 cells C75 was used at a final concentration of 10μgml−1 and for MCF7 cells 8μgml−1. Orlistat was used at a final concentration of 30 or 10μgml−1 in MCF7 or MD231 respectively. Pvalue is the result of T-test. Data are average±s.e.m.; n=3.*P≤0.05, **P≤0.01, ***P≤0.001. (e) Forty-eight hours cell growth of MDA231 or MCF7 cells overexpressing exogenous WT or Inactive LIPG. Cells were treated with DMSO (control) and FAS inhibitor (C75) at a final concentration of 20μgml−1. P value is the result of T-test. Data are average±s.e.m.; n=3.***P≤0.001, ****P≤0.0001 (f) Schematic representation showing how FoxA controls LIPG and lipid metabolism to support tumour growth.

As previous reports showed that de novo lipid metabolism is necessary for BCa growth3, 22, we next questioned whether this lipid synthesis was sufficient or, instead, whether exogenous sources are also required to support BCa cell growth and proliferation, as suggested by our experimental data. To this end, we inhibited the activity of fatty acid synthase (FAS) in BCa cells by means of the chemical inhibitor C75 (ref. 23). FAS activity is crucial for de novo lipid synthesis in cancer cells3,22. To test the complementarity of both de novo and/or exogenous lipid supplies, we used a C75 concentration causing a 50% reduction in BCa cell growth in vitro 48h post incubation (Fig. 5d andSupplementary Fig. 5d). Similarly, we tested the contribution of LIPG inhibition by means of treatment with a lipase inhibitor, Orlistat21. A specific dose causing a 50% reduction in the growth of each BCa cell line was further used (Fig. 5d and Supplementary Fig. 5d). Interestingly, concomitant treatment with FAS and LIPG inhibitors caused an additive effect, blunting BCa cell growth (Fig. 5d). Next, we evaluated whether LIPG activity was sufficient to rescue the chemical inhibition of FAS. To this end, we overexpressed WT and inactive LIPG and grew MCF7 and MDA231 cells in the presence or absence of a high dose of C75 (20mgml−1), which blocks cell growth (Supplementary Fig. 5d). Complete blockade of FAS was not rescued by LIPG (Fig. 5e). Collectively, our results suggest that both exogenous lipid precursors provided by means of LIPG activity and de novo lipid synthesis mediated by FAS are necessary for BCa cell growth.


Here we reveal that FoxA factors provide a central metabolic growth function by specifically regulating LIPG expression, thereby allowing the acquisition of indispensable extracellular lipids for BCa tumour proliferation. FoxA family of transcription factors are expressed in the vast majority of BCa and FoxA1 is expressed across various BCa subtypes. Moreover we show that, in some cases, its absence is associated with the expression of FoxA2. Interestingly, in addition of FoxA1 contribution to luminal commitment24, 25, 26, 27 the factor may drive BCa growth by specifically regulating LIPG levels.

The catalytic activity of LIPG generates extracellular lipid precursors that are imported to fulfill the intracellular production of lipid species (Fig. 5f). LIPG downregulation blocks BCa cell growth, thereby indicating that the import of extracellular lipid precursors is important for the proliferation of these cells. This is a striking observation given that it is generally believed that de novo fatty acid synthesis is the main driver of tumour growth22. Indeed, our experimental data with LIPG-depleted BCa cells revealed a massive decrease of most intracellular glycerolipid intermediates in the synthesis of TG (PC, PE, PG and DG) and their derivatives (LPC and LPE). Accordingly, certain lipid species (LPC) in the media were not decreased in LIPG-depleted cells as much as in control cells, thus indicating that extracellular lipids are the substrates for intracellular lipid production. In particular, we demonstrate the relevance of extracellular LPC (18:0) for BCa cell proliferation in a lipoprotein-depleted medium, a process dependent on LIPG. In this context, a high-fat diet was shown to rescue the absence of a critical intracellular lipase, Monoacylglycerol lipase, for cancer pathogenesis given cancer cells ability to uptake lipids from the extracellular compartment was functional19. Herein, we showed that this rescue mechanism is not functional in BCa cells in the absence of FoxA2 or LIPG. In support of this notion, it is worth noting that extracellular LIPG activity releases fatty acids from high-density lipoprotein phospholipids and these acids are further employed for intracellular lipid production in the human hepatic cell line HepG2 (refs 28, 29).

In conclusion, BCa cells are dependent on a mechanism to supply precursors derived from extracellular sources for intracellular lipid production, and LIPG fulfills this function. Therefore, LIPG stands out as an important component of the lipid metabolic adaptations that BCa cells, and not normal tissue, must undergo to support high proliferation rates. Our results also suggest thatde novo lipid synthesis is necessary but not sufficient to support lipid production for BCa tumour growth. Accordingly, recent clinical studies demonstrate the association between lipids and lipoproteins in circulation and risk of BCa in women with extensive mammographic density. This observation implies that interventions aimed to reduce them may have effect on BCa risk30. All together, these observations make LIPG activity an Achilles heel of luminal and, more importantly, of triple negative/basal-like breast tumours, for which limited therapeutic options are currently available.

In normal cells, the glucose carbon flow is directed into a de novo lipogenic pathway that is regulated, in part, via phosphoinositide-3 kinase (PI-3K)-dependent activation of ATP citrate lyase (ACL), a key rate-limiting, enzyme in de novo lipogenesis. ACL is a cytosolic enzyme that catalyzes the generation of acetyl CoA from citrate. Inhibition of ACL results in a loss of B-cell growth and cell viability [10] .
The plasma membrane and its constituent phosphoinositides form the basis of the phosphatidylinositol 3-kinase (PI3-K) signaling pathway, which is crucial for cell proliferation and survival. Phosphatase and tensin-homolog deleted on chromosome 10 (PTEN) is a tumor-suppressor protein that regulates phosphatidylinositol 3-kinase (PI3-K) signaling by binding to the plasma membrane and hydrolyzing the 3′ phosphate from phosphatidylinositol (3,4,5)-trisphosphate (PI(3,4,5)P3) to form phosphatidylinositol (4,5)-bisphosphate (PI(4,5)P2). Several loss-of-function mutations in PTEN that impair lipid phosphatase activity and membrane binding are oncogenic, leading to the development of a variety of cancers. Of these three residues, R335 was observed to interact with the membrane to the greatest extent across all of the simulations. R335L, in common with several other germline mutations, has been associated with the inherited cancer [11] .
ACLY is up-regulated or activated in several types of cancers, and its inhibition is known to induce proliferation arrest in cancer cells both in vitro and in vivo. The last studies were showed that BCR-mediated signaling is regulated in part by the amount of membrane cholesterol. It was observed that statins (Lovostatin), the pharmacological inhibitors of cholesterol synthesis, induce apoptosis of CLL cells in vitro and in vivo. Also the ectopic expression of CD5 in a B-cell line stimulates the transcription of genes involved in the synthesis of cholesterol [12] .

[10] Zaidi N, Swinnen JV, Smans K. ATP-citrate lyase: a key player in cancer metabolism Cancer Res; 2012 (11): 3709-14.

[11] Craig N, Mark S.P. Sansom. Defining the Membrane-Associated State of the PTEN Tumor Suppressor Protein. Biophys J 2013; 5; 104(3: 613–21.

[12] Tomowiak C, Kennel A, Gary-Gouy, Hadife N. High Membrane Cholesterol in CLL B-Cells and Differential Expression of Cholesterol Synthesis Genes in IG GENE Unmutated vs Mutated Cells. British Journal of Medicine & Medical Research 2012; 2(3): 313-26.


Cancer’s Vanguard

Exosomes are emerging as key players in metastasis.

By Catherine Offord | April 1, 2016

PREPARING THE TURF: Before tumor cells arrive at their metastatic destination, part of the site is readied for them. One recent study of liver metastasis in mice found that resident macrophages called Kupffer cells take up exosomes from the original tumor (1). Additionally, macrophages from the bone marrow show up upon the release of fibronectin by other liver cells called stellate cells (2). A current proposal for additional steps in metastatic niche development includes the recruitment of epithelial cells and fibroblasts, which contribute to angiogenesis, and, finally, the arrival of tumor cells themselves (3).© IKUMI KAYAMA/STUDIO KAYAMA

In 2005, David Lyden noticed something unexpected. He and his colleagues at Weill Cornell Medical College had been researching metastasis—the spread of cancer from one part of the body to another. The team had shown that bone marrow–derived cells (BMDCs) were recruited to future metastatic sites before the arrival of tumor cells, confirming that metastasis occurred after a habitable microenvironment, or “premetastatic niche,” had been prepared.1

But carefully studying images of this microenvironment in the lung tissue of mice, Lyden saw something else. Amongst the BMDCs, the micrographs showed tiny specks, far too small to be cells, gathering at the future site of metastasis. “I said, ‘What are these viruses doing here?’” recalls Lyden. “I had no idea about exosomes, microvesicles, and microparticles.”

Those specks, Lyden would come to realize, were in fact primary tumor–derived exosomes. These membrane-enclosed vesicles packed full of molecules are now attracting growing attention as important mediators of intercellular communication, particularly when it comes to cancer’s insidious capacity to spread from one organ to another.

Preparing the ground

Tumors require a community of support cells, including fibroblasts, BMDCs, and endothelial cells, to provide functional and structural assistance and to modulate immune system behavior. Bringing together the first members of this community before the arrival of tumor cells is all part of cancer’s survival strategy, says Joshua Hood, a cancer researcher at the University of Louisville.

“It wouldn’t be efficient for tumor cells to strike out on their own, and just say, ‘Oh, here we are!’” he says. “They would run the risk of being destroyed.” Preparing a “nest” in advance makes the process much safer. “Then the tumor can just efficiently come along and set up shop without ever having to fight much of a battle with the immune system.”

But although Lyden’s group had shown that this preparation was taking place, it remained unclear how such a process might be regulated. For the next few years, many cancer researchers believed that tumor cells must communicate with the premetastatic niche primarily through tumor-secreted signaling molecules such as cytokines.

Meanwhile, research into extracellular vesicles, previously considered biological garbage bags, was revealing new modes of intercellular communication. In 2007, a group of scientists in Sweden discovered that exosomes, tiny vesicles measuring just 30 nanometers to 100 nanometers across, transport mRNA and microRNAs intercellularly, with the potential to effect changes in protein synthesis in recipient cells.2 A new means for tumors to regulate distant cellular environments came into focus, and research on exosomes exploded. In 2011, Hood and his colleagues showed that exosomes facilitate melanoma metastasis through the lymphatic system.3 The following year, Lyden’s group demonstrated that tumor-derived exosomes can direct BMDCs to one of melanoma’s most common sites of metastasis, the lung.4 Exosomes, it seemed, had been underestimated.

Tiny terraformers

Armed with the knowledge that exosomes are involved in multiple stages of melanoma metastasis, Lyden’s lab went searching for the vesicles’ potential role in the metastasis of other cancers. Turning to pancreatic ductal adenocarcinoma (PDAC)—one of the most lethal cancers in humans—postdoctoral researcher Bruno Costa-Silva led a series of exhaustive in vitro and in vivo experiments in mouse models to detail the process of premetastatic niche formation in the liver, PDAC’s most common destination. The team’s results, published last May, reveal an intricate series of sequential steps—mediated by PDAC-derived exosomes (Nature Cell Biol, 17:816-26, 2015).

Using fluorescence labeling, Lyden’s group observed that PDAC-derived exosomes are taken up by Kupffer cells, specialized macrophages lining the outer walls of blood vessels in the liver. There, the exosomes trigger the cells’ secretion of transforming growth factor β (a type of cytokine involved in cell proliferation), plus the production of fibronectin by neighboring hepatic stellate cells, and the recruitment of BMDCs.

The researchers also showed that this cascade of events could be inhibited by depleting exosomal macrophage migratory inhibitory factor (MIF), an abundant protein in PDAC exosomes. “If you target the specific proteins of exosomes, you can reduce metastasis,” explains coauthor Héctor Peinado, leader of the microenvironment and metastasis group at the Spanish National Cancer Research Center.

For Hood, the findings add to a developing picture of exosomes’ vital role as “vanguard” in the progression of cancer. “It’s like the colonization of a new planet,” he says. “They’re terraforming the environment to make it hospitable.”



  • B. Costa-Silva et al., “Pancreatic cancer exosomes initiate pre-metastatic niche formation in the liver,”Nature Cell Biol, 17:816-26, 2015.
  • A. Hoshino et al., “Tumour exosome integrins determine organotropic metastasis,” Nature, 527:329-35, 2015.
  • L. Zhang et al., “Microenvironment-induced PTEN loss by exosomal microRNA primes brain metastasis outgrowth,”Nature, 527:100-04, 2015.

Internal mail

Although research was revealing the steps involved in forming premetastatic sites, it was less clear how these sites were being selected. “This has always been a great mystery in cancer,” says Ayuko Hoshino, a research associate in Lyden’s lab. “Why do certain cancers metastasize to certain organs?”

One theory, proposed in 1928 by pathologist James Ewing, suggested that anatomical and mechanical factors explained organ specificity in metastasis. The premetastatic niche, then, might form wherever exosomes are likely to land. But this couldn’t be the whole story, says Hoshino. “For instance, there’s eye melanoma. Thinking about that site, you could imagine it metastasizing to the brain. But actually, it almost only metastasizes to the liver.”

Because exosomes arrive at metastatic sites before tumor cells, the team reasoned, perhaps the exosomes themselves were organotropic (i.e., attracted to particular organs or tissues). Sure enough, Lyden says, when Hoshino and Costa-Silva began injecting tumor-derived exosomes into mice, “their preliminary findings were that wherever they injected the exosomes, the pancreatic cancer ones were ending up in the liver and the breast metastasis exosomes would end up in the lung.”

Using mass spectrometry, the researchers analyzed the protein content of exosomes from lung-tropic, liver-tropic, and brain-tropic tumors. They found that the composition of exosomes’ integrins—membrane proteins involved in cell adhesion—was destination-specific (Nature, 527:329-35, 2015). Exosomes bearing integrin α6β4, for example, were directed to the lung, where they could prepare a premetastatic niche potent enough even for normally bone-tropic tumor cells to colonize. Integrin αvβ5, meanwhile, directed metastasis to the liver.

The researchers also showed that exosomal integrins didn’t necessarily correspond to the parent-cell proteins, making exosomes potentially better indicators of where a cancer will spread than the tumor cells themselves. “We can show that an integrin that’s high in the tumor cell might be completely absent in the tumor exosome or vice versa,” says Lyden, adding that, taken together, the results point to a role for exosomes in “dictating the future sites of metastasis.”

“It’s a beautiful story,” says Dihua Yu, a molecular and cellular oncologist at the University of Texas MD Anderson Cancer Center. “This is a very novel finding that gives really good indicators for potential strategies to intervene in metastasis.”

Metastatic crosstalk

In the same month that Lyden’s group published its work on organotropism, Yu’s own lab published a different exosome study—one that told another side of the story.

Yu and her colleagues had found that when tumor cells in mice metastasized to the brain, they downregulated expression of a tumor suppressor gene called PTEN, and became primed for growth at the metastatic site. When the tumor cells were taken out of the microenvironment and put in culture, however, they restored normal PTEN expression.

The researchers demonstrated that a microRNA from astrocytes—star-shape glial cells in the brain—reversibly downregulated the levels of PTEN transcripts in the tumor cells, but they couldn’t figure out how the microRNA was getting into the tumor. Blocking “obvious signaling pathways,” such as gap junctions, failed to have an effect, Yu says.

Scrutinizing astrocyte-conditioned media using electron microscopy, the researchers identified spherical vesicles between 30 nanometers and 100 nanometers in diameter—the defining size of exosomes. Exposing mouse tumor cells to these vesicles increased cell microRNA content and reduced PTENexpression (Nature, 527:100-04, 2015). The study revealed yet another role for exosomes in the communication between tumors and their microenvironment.

The findings were a surprise, says Yu, not least because they showed a different perspective from the bulk of recent research. “We’re talking about astrocytes in the brain secreting exosomes to give welcome help to the cancer cells,” she says.

“I find it an extremely interesting paper because it shows that the astrocytes can change the whole phenotype of the tumor in the brain,” says Lyden. He adds that the results underline the importance of studying the mutational status of tumors at various sites. “All this work in exosomes, it adds to the complexity,” he says. “We can’t just target tumor cells at the primary site. We’ll have to understand all the details of metastasis if we’re really going to tackle it.”

What’s next?

The discovery of multiple roles for exosomes in metastasis has generated excitement about the potential for their use in diagnostics and treatment. As protective containers of tumor-derived genetic material, exosomes could provide information about the status of cancer progression. And as mediators of premetastatic niche formation, they make obvious targets for inhibition. (See “Banking on Blood Tests,”here.)

Exosomes might even be useful as vehicles to deliver drugs because they’re patient-matched and “naturally designed to function in a biocompatible way with living systems,” says Hood. “You could take them out of people, and at some point down the road try to have patients be their own nanofactory, using their own particles for treatment purposes.”

Pancreatic cancer exosomes initiate pre-metastatic nihe formation in the liver

Bruno Costa-SilvaNicole M. AielloAllyson J. Ocean, et al.   Nature Cell Biology 2015; 17,816–826

Pancreatic ductal adenocarcinomas (PDACs) are highly metastatic with poor prognosis, mainly due to delayed detection. We hypothesized that intercellular communication is critical for metastatic progression. Here, we show that PDAC-derived exosomes induce liver pre-metastatic niche formation in naive mice and consequently increase liver metastatic burden. Uptake of PDAC-derived exosomes by Kupffer cells caused transforming growth factor β secretion and upregulation of fibronectin production by hepatic stellate cells. This fibrotic microenvironment enhanced recruitment of bone marrow-derived macrophages. We found that macrophage migration inhibitory factor (MIF) was highly expressed in PDAC-derived exosomes, and its blockade prevented liver pre-metastatic niche formation and metastasis. Compared with patients whose pancreatic tumours did not progress, MIF was markedly higher in exosomes from stage I PDAC patients who later developed liver metastasis. These findings suggest that exosomal MIF primes the liver for metastasis and may be a prognostic marker for the development of PDAC liver metastasis.

Ayuko HoshinoBruno Costa-SilvaTang-Long ShenGoncalo RodriguesAyako HashimotoMilica Tesic Mark, et al. Nature Nov 2015; 527,329–335

Ever since Stephen Paget’s 1889 hypothesis, metastatic organotropism has remained one of cancer’s greatest mysteries. Here we demonstrate that exosomes from mouse and human lung-, liver- and brain-tropic tumour cells fuse preferentially with resident cells at their predicted destination, namely lung fibroblasts and epithelial cells, liver Kupffer cells and brain endothelial cells. We show that tumour-derived exosomes uptaken by organ-specific cells prepare the pre-metastatic niche. Treatment with exosomes from lung-tropic models redirected the metastasis of bone-tropic tumour cells. Exosome proteomics revealed distinct integrin expression patterns, in which the exosomal integrins α6β4 and α6β1 were associated with lung metastasis, while exosomal integrin αvβ5 was linked to liver metastasis. Targeting the integrins α6β4 and αvβ5 decreased exosome uptake, as well as lung and liver metastasis, respectively. We demonstrate that exosome integrin uptake by resident cells activates Src phosphorylation and pro-inflammatory S100 gene expression. Finally, our clinical data indicate that exosomal integrins could be used to predict organ-specific metastasis.

  1. Paget, S. The distribution of secondary growths in cancer of the breast. 1889. Cancer Metastasis Rev. 8, 98101 (1989)
  2. Hart, I. R. & Fidler, I. J. Role of organ selectivity in the determination of metastatic patterns of B16 melanoma. Cancer Res. 40, 22812287 (1980)
  3. Müller, A. et al. Involvement of chemokine receptors in breast cancer metastasis. Nature410, 5056 (2001)
  4. Weilbaecher, K. N., Guise, T. A. & McCauley, L. K. Cancer to bone: a fatal attraction. Nature Rev. Cancer 11, 411425 (2011)
  5. Zhou, W. et al. Cancer-secreted miR-105 destroys vascular endothelial barriers to promote metastasis. Cancer Cell 25, 501515 (2014)
  6. Chang, Q. et al. The IL-6/JAK/Stat3 feed-forward loop drives tumorigenesis and metastasis.Neoplasia 15, 848862 (2013)
  7. Lu, X. & Kang, Y. Organotropism of breast cancer metastasis. J. Mammary Gland Biol. Neoplasia 12, 153162 (2007)


Microenvironment-induced PTEN loss by exosomal microRNA primes brain metastasis outgrowth

Lin ZhangSiyuan ZhangJun YaoFrank J. LoweryQingling ZhangWen-Chien Huang, et al.  Nature  Nov 2015; 527,100–104

The development of life-threatening cancer metastases at distant organs requires disseminated tumour cells’ adaptation to, and co-evolution with, the drastically different microenvironments of metastatic sites1. Cancer cells of common origin manifest distinct gene expression patterns after metastasizing to different organs2. Clearly, the dynamic interaction between metastatic tumour cells and extrinsic signals at individual metastatic organ sites critically effects the subsequent metastatic outgrowth3, 4. Yet, it is unclear when and how disseminated tumour cells acquire the essential traits from the microenvironment of metastatic organs that prime their subsequent outgrowth. Here we show that both human and mouse tumour cells with normal expression of PTEN, an important tumour suppressor, lose PTEN expression after dissemination to the brain, but not to other organs. The PTEN level in PTEN-loss brain metastatic tumour cells is restored after leaving the brain microenvironment. This brain microenvironment-dependent, reversible PTEN messenger RNA and protein downregulation is epigenetically regulated by microRNAs from brain astrocytes. Mechanistically, astrocyte-derived exosomes mediate an intercellular transfer of PTEN-targeting microRNAs to metastatic tumour cells, while astrocyte-specific depletion of PTEN-targeting microRNAs or blockade of astrocyte exosome secretion rescues the PTEN loss and suppresses brain metastasis in vivo. Furthermore, this adaptive PTEN loss in brain metastatic tumour cells leads to an increased secretion of the chemokine CCL2, which recruits IBA1-expressing myeloid cells that reciprocally enhance the outgrowth of brain metastatic tumour cells via enhanced proliferation and reduced apoptosis. Our findings demonstrate a remarkable plasticity of PTEN expression in metastatic tumour cells in response to different organ microenvironments, underpinning an essential role of co-evolution between the metastatic cells and their microenvironment during the adaptive metastatic outgrowth. Our findings signify the dynamic and reciprocal cross-talk between tumour cells and the metastatic niche; importantly, they provide new opportunities for effective anti-metastasis therapies, especially of consequence for brain metastasis patients.

  1. Quail, D. F. & Joyce, J.A. Microenvironmental regulation of tumor progression and metastasis. Nature Med. 19, 14231437 (2013)
  2. Park, E. S. et al. Cross-species hybridization of microarrays for studying tumor transcriptome of brain metastasis. Proc. Natl Acad. Sci. USA 108, 1745617461 (2011)
  3. Joyce, J. A. & Pollard, J. W. Microenvironmental regulation of metastasis. Nature Rev. Cancer 9, 239252 (2009)
  4. Vanharanta, S. & Massagué, J. Origins of metastatic traits. Cancer Cell 24, 410421 (2013)
  5. Gray, J. Cancer: genomics of metastasis. Nature 464, 989990 (2010)
  6. Friedl, P. & Alexander, S. Cancer invasion and the microenvironment: plasticity and reciprocity. Cell 147, 9921009 (2011)

Banking on Blood Tests

How close are liquid biopsies to replacing current diagnostics?

By Jyoti Madhusoodanan | April 1, 2016

No matter where a tumor lurks in the body, its secrets circulate in the blood. Stray tumor cells begin metastatic migrations by slipping into the vasculature. Vesicles secreted by cancer cells and free-floating DNA are also released into the bloodstream. Because these bits of cellular debris are a grab-bag of biomarkers that could both signal a cancer’s presence and predict its progression and response to treatment, the use of blood-based tests, or liquid biopsies, to detect and evaluate them is now drawing significant commercial interest.

Last year, San Diego–based Pathway Genomics began advertising a screen “for the early detection of up to 10 different cancer types in high-risk populations.” But the screen had only been tested in already-diagnosed patients, not in at-risk individuals, and within weeks of making it commercially available, the company received an FDA notice to provide more information about their promotional claims before further marketing. “We . . . have not found any published evidence that this test or any similar test has been clinically validated as a screening tool for early detection of cancer in high risk individuals,” the agency wrote.

The Forces of Cancer

A tumor’s physical environment fuels its growth and causes treatment resistance.

By Lance L. Munn and Rakesh K. Jain | April 1, 2016

Ahelium balloon tugs gently at the end of its string. The tension in the string resists the buoyant force of the helium, and the elastic nature of the balloon’s rubber contains the helium gas as it tries   to expand. Cutting the string or poking the rubber with a pin reveals the precarious balance between the forces, upsets the equilibrium, and sets the system into motion.

Some biological tissues also exist in such a state of offsetting forces. The most familiar example is the balance between blood pressure and the elastic tension in the cardiovascular system that contains and conveys blood without bursting or collapsing. And in tumors, both solid and fluid forces are generated that make the cancerous tissue a lot like that helium balloon: cut a tumor with a scalpel and it rapidly swells and deforms as pent-up forces break free from structural elements that are severed.1

One force that is notably higher in tumors than in healthy tissues is fluid pressure, resulting from hyperpermeable, leaky blood vessels and a dearth of draining lymphatic vessels. Researchers have known since the 1950s that tumors exhibit elevated fluid pressure, but the implications for tumor progression and drug delivery were not realized until the late 1980s. That was when we (R.K.J. and colleagues) used a mathematical model to predict—and subsequently validate in animal and human tumors—that a precipitous drop in fluid pressure at the tumor–normal tissue interface causes interstitial fluid to ooze out of the tumor.2 This seeping fluid pushes drugs, growth factors, and cancer cells into the surrounding tissue and lymphatics, reducing drug delivery and facilitating local tumor invasion and distant metastasis.

Based on this insight, we suggested in 2001 that anti-angiogenic drugs could be used to lower a tumor’s fluid pressure and improve treatment outcome.3 This hypothesis changed the thinking about how existing anti-angiogenesis therapies actually work and spurred research into other physical forces acting in cancer.4 In the last 15 years, researchers have identified diverse sources of increased pressure in tumors, which may serve as possible targets for cancer therapy.5 For example, solid forces exerted by the extracellular matrix can be reduced by treatment with drugs approved by the US Food and Drug Administration (FDA) for controlling hypertension (angiotensin blockers) or diabetes (metformin). Retrospective clinical studies have found improved survival in cancer patients who were treated with these agents, which are now being tested in prospective trials for a variety of solid tumors.6,7

Tumors under pressure

In vitro experiments showing that cancer cells actively migrate in response to fluid flow have supported the hypothesis that fluid escaping from the boundary of a tumor may guide the invasive migration of cancer cells toward lymphatic or blood vessels, potentially encouraging metastasis. There remains controversy over how the fluid forces induce the migration; the cells may respond to chemical gradients created by the cells and distorted by the flowing fluid,8 or the fluid may activate cell mechanosensors.9 Because of the potential for new therapeutic interventions, the transduction of mechanical fluid forces into biochemical signals by cell mechanosensors is an active area of investigation. In a more direct manner, the fluid flow can physically carry cancer cells to lymph nodes.

Fluid forces may also promote tumor progression by recruiting blood vessels into the cancerous mass.10 Because tumor blood vessels are leaky, plasma can pass freely between vessels that have different pressures. When this happens at the periphery of a tumor, where angiogenic growth factors are prevalent, there can be synergistic induction of new vessel sprouts.


And fluid pressure is just one of the many forces in a tumor that can influence its development and progression. Tumors also develop increased solid pressure, as compared with normal tissue, stemming from the uncontrolled division of cancer cells and from the infiltration and proliferation of stromal and immune cells from the surrounding tissue and circulation. High-molecular-weight polysaccharides known as hydrogels found in the extracellular matrix (ECM) also add pressure on a tumor. The most well-studied of these hydrogels is hyaluronan; when the polysaccharide absorbs water, it swells, pressing on surrounding cells and structural elements of the tissue.

The ECM contains a highly interconnected network of collagen and other fibers and is normally very good at resisting and containing such tension. It also has support from infiltrating myofibroblasts, which detect areas where the ECM density or tension is not normal and initiate actomyosin-based contraction of collagen and elastin matrix structures to restore tensional homeostasis. But while this repair effort is typically effective in healthy tissues, uncooperative tumor cells interfere with these efforts, both by themselves generating pressure and by hyperactivating cancer-associated fibroblasts to produce more ECM and thus produce even more force.11

Because cell growth and ECM composition are not spatially uniform in cancer, tumors are subjected to multiple, dispersed sources of pressure associated with matrix “containers” of various sizes. This solid pressure from within the tumor deforms the surrounding normal tissue, potentially facilitating the metastatic escape of cancer cells. The physical forces also compress blood vessels and lymphatic vessels in the tumor and adjacent normal tissue,12 increasing the fluid pressure in the tumor13  and interrupting the delivery of nutrients, removal of waste, and entry of tumor-targeted drugs via the blood.4 Insufficient blood flow also results in poor oxygenation, which has been linked to immunosuppression, inflammation, invasion, and metastasis, as well as lowered efficacy of chemo-, radio-, and immunotherapies.4 These are all indirect consequences of solid stresses in and on tumors.

Such forces can also have direct effects on cancer cells, and may serve as independent triggers for tumor invasion. Mechanical forces are central to many of our sense systems, such as hearing, touch, and pain, and to tissue maintenance programs, such as bone regeneration and blood vessel remodeling. In these systems, mechanical forces are transduced by mechanosensors to activate downstream biochemical and genetic pathways. (See “Full Speed Ahead,” The Scientist, December 2009.) Cancer cells may similarly be able to sense and respond to dynamic forces in tumors. We have shown, for example, that metastatic cancer cells exposed to compressive stresses in a culture dish undergo a phenotypic transformation to become more invasive,14 and others have shown that compressive forces applied in vivo can also induce oncogenes in normal epithelium of the mouse colon.15

It is thus becoming quite clear that the physical environment can influence a tumor’s development and spread, and it may even be possible for physical forces to kick-start cancerous growth.


Full Speed Ahead

Physical forces acting in and around cells are fast—and making waves in the world of molecular biology.

By Jef Akst | December 1, 2009

When it comes to survival, few things are more important than being able to respond quickly to a change of circumstances. And when it comes to fast-acting indicators, it turns out that signals induced by physical forces acting in and around cells, appropriately dubbed biomechanical signals, are the champions of the cellular world.

“If you look at this mechanical signaling, it’s about 30 meters per second—that’s very fast,” says bioengineer Ning Wang of the University of Illinois at Urbana-Champaign. That’s faster than most family-owned speedboats, and second only to electrical (e.g., nerve) impulses in biological signaling. By comparison, small chemicals moving by diffusion average a mere 2 micrometers per second—a speed even the slowest row boater could easily top.

Indeed, when the two signal types were pitted against each other in a cellular race last year, the mechanical signals left chemical signals in their wake, activating proteins at distant sites in the cytoplasm in just a fraction of a second, at least 40 times faster than their growth factor opponent.1 Mechanical signals are so fast, Wang adds, they are “beyond our resolution,” meaning that current imaging techniques cannot capture the very first cellular changes that result from mechanical stress, which occur within nanoseconds.

For centuries, scientists have scrutinized the molecular inner workings of the body, with little or no regard to the physical environment in which these biological reactions take place. But the growing realization that physical forces have a pervasive presence in physiology (operating in a variety of bodily systems in thebone, blood, kidney, and ear, for instance), and act with astonishing speed, has caused many to consider the possibility that mechanical signaling may be just as important as chemical communication in the life of a cell.

“Biologists have traditionally ignored the role of mechanics in biology,” says biomechanical engineer Mohammad Mofrad of the University of California, Berkley, “[but] biomechanics is becoming increasingly accepted, and people are recognizing its role in development, in disease, and in general cellular and tissue function.”

The wave within: Mechanical forces acting inside the cell

Once believed to be little more than sacks of chemically active goop, cells didn’t seem capable of transmitting physical forces into their depths, and researchers largely limited their search for molecules or structures that respond to physical forces, or mechanosensors, to the plasma membrane.

Mechanical signaling may be just as important as chemical communication in the life of a cell.

In the late 1990s, however, closer examination revealed that the cell’s interior is in fact a highly structured environment, composed of a network of filaments.2 Pull on one side of the cell, and these filaments will transmit the force all the way to other side, tugging on and bumping into a variety of cellular structures along the way—similar to how a boat’s wake sends a series of small waves lapping up on a distant and otherwise peaceful shoreline. Scientists are now realizing the potential of such intracellular jostling to induce molecular changes throughout the cell, and the search for mechanosensing molecules has escalated dramatically in scope, including, for example, several proteins of the nucleus.

It’s a search that will likely last a while, predicts cell biologist Donald Ingber, director of the Wyss Institute for Biologically Inspired Engineering at Harvard University. “To try to find out what’s the mechanosensor is kind of crazy at this point,” he says. As scientists are now learning, “the whole cell is the mechanosensor.”

A key player, most agree, is the cytoskeleton, which is comprised of a variety of microfilaments, including rigid actin filaments and active myosin motors—the two principle components of muscle. Activation of the so-called nonmuscle myosins causes the cytoskeleton to contract, much like an arm muscle does when it lifts a heavy object.

The first intimation that the cytoskeleton could go beyond its established inner-cell duties (molecule transport and cell movement and division) came in 1997, when Ingber did the logical (in hindsight, at least) experiment of pulling on the cells to see what happened inside.2 Using a tiny glass micropipette coated in ligands, Ingber and his team gently probed the surface proteins known as integrins, which secure the cell to the extracellular matrix. When they quickly pulled the micropipette away, they saw an immediate cellular makeover: cytoskeletal elements turned 90 degrees, the nucleus distorted, and the nucleolus—a small, dense structure within the nucleus that functions primarily in ribosome assembly—aligned itself with the direction of the applied force.

“That kind of blew people away,” Ingber recalls. “It revealed that cells have incredible levels of structure not only in the cytoplasm but in the nucleus as well.”

Wang (once a postdoc in Ingber’s lab at the Harvard School of Public Health) and other collaborators combined a similar technique with fluorescent imaging technology to visualize how these forces were channeled within the cell’s interior. Upping the resolution and further refining these techniques, Wang began mapping these intracellular forces as they made their way through the cell. In 2005, the maps confirmed the physical connection between the cell-surface integrins and the nucleus, and showed that these external forces follow a nonrandom path dictated by the tension of the cytoskeletal elements.3

“Biomechanics is becoming increasingly accepted, and people are recognizing its role in development, in disease, and in general cellular and tissue function.”
–Mohammad Mofrad

The end point of these mechanical pathways is likely a mechanosensitive protein, which changes shape in response to the force, thereby exposing new binding areas or otherwise changing the protein’s function. In mitochondria, for example, mechanical forces may trigger the release of reactive oxygen species and activation of signaling molecules that contribute to inflammation and atherosclerosis.

Similarly, proteins on the nuclear membrane may pass mechanical signals into the nucleus by way of a specialized structure known as LINC (linker of nucleoskeleton and cytoskeleton), which physically links the actin cytoskeleton to proteins important in nuclear organization and gene function. To determine if mechanical forces directly affect gene expression, last year scientists began exploiting the increasingly popular fluorescence resonance energy transfer (FRET) technology,1 in which energy emitted by one fluorescent molecule can stimulate another, resulting in a visible energy transfer that can track enzymatic activities in live cells. By combining FRET technology with the techniques that apply physical forces to specific cell membrane proteins, scientists can visualize entire mechanochemical transduction pathways, Wang says.

“The big issue right now in the field of mechanotransduction is whether the genes in the nucleus can be directly activated by forces applied to the cell surface,” Wang explains. While the physical maps of the cytoskeleton tentatively sketch out a path that supports this possibility, confirmatory data is lacking. This combination of new technologies will be “tremendously” helpful in answering that question, he says, and “push the field” towards a more complete understanding of how mechanical forces can influence cellular life.

An early start: Mechanical forces in development

In the world of developmental biology, the cytoskeleton’s role in biomechanics really comes into its own. As the embryo develops, the cells themselves are the force generators, and by contracting at critical times, the cytoskeleton can initiate many key developmental steps, from invagination and gastrulation to proliferation and differentiation, and overall cellular organization.

The idea that physical forces play a role in development is not a new one. In the early 20th century, back when Albert Einstein was first developing the molecular basis of viscosity and scientists were realizing molecules are distinct particles, biologist and mathematician D’Arcy Thompson of the University of Dundee in Scotland suggested that mechanical strain is a key player in morphogenesis. Now, nearly a century later, biologists are finally beginning to agree.

Because Thompson “couldn’t measure [the forces] at that time, that kind of thinking got pushed to the wayside as genetic thinking took over biology,” says bioengineer Christopher Chen of the University of Pennsylvania. That is, until 2003, when Emmanuel Farge of the Curie Institute in France squeezedDrosophila embryos to mimic the compression experienced during early development and activated twist—a critical gene in the formation of the digestive tract.4 These results gave weight to Thompson’s idea that stress in the embryo stimulates development and growth, and inspired developmental scientists to begin considering mechanical effects, Chen says. “Now we’re at the stage where there’s a lot of interest and willingness to consider the fact that mechanical forces are not only shaping the embryo, but are linked to the differentiation programs that are going on.”

Again, the cytoskeleton is a key player in this process. In fruit flies and frogs, for example, nonmuscle myosins contract the actin filaments to generate the compressive forces necessary for successful gastrulation—the first major shape-changing event of development. Myosins similarly influence proliferation in the development of the Drosophila egg chamber, with increased myosin activity resulting in increased cell division.

Cytoskeleton contractility also appears to direct stem cell differentiation. In 2006, Dennis Discher of the University of Pennsylvania demonstrated that the tension of the substrate on which cells are grown in culture is important for determining what type of tissue the cells will form.5 Cells grown on soft matrices that mimic brain tissue tended to grow into neural cells, while cells grown on stiffer matrices grew into muscle cell precursors, and hard matrices yielded bone. In this case, it seems that stiffer substrates increased the expression of nonmuscle myosin, generating greater tension in the actin cytoskeleton and affecting differentiation. (Altering or inhibiting myosin contraction can also affect differentiation.)

“To try to find out what’s the mechanosensor is kind of crazy at this point. As scientists are now learning, the whole cell is the mechanosensor.”
–Donald Ingber
Shaping a tumor

In addition to the influence of physical forces on cancer growth and invasion, forces can alter a tumor’s mechanical properties, and vice versa. Tumors are more rigid, or stiffer, than surrounding tissues, usually because they contain excess collagen in the ECM,5 and this can contain and amplify local forces produced by proliferating cancer cells. On the other hand, tumor rigidity can be further enhanced if the cells exert tension on ECM collagen fibers by pulling on them, or by stretching them, as occurs when tumors grow uncontrollably. Fluid forces can also influence the assembly of collagen fibers within and around tumors,8potentially increasing stiffness.

Importantly, tumor stiffness tends to be associated with poor prognoses, though the reasons for this are not fully understood. Cells are known to differentiate into different lineages depending on the local rigidity;16 for example, stem cells differentiate into bone on stiff substrates, but make adipose (fat) cells on softer substrates. Similar mechanisms are thought to affect tumor progression when the ECM changes rigidity, inducing cancer cells to become more invasive as well as more likely to metastasize. Indeed, longer collagen fibers in the matrix are associated with increased invasion and metastasis, as well as reduced survival, in mice.17

In addition, the abnormal ECM in tumors can affect cancer progression by activating normal stromal cells, such as macrophages and fibroblasts, that accelerate tumor growth and treatment resistance. These activated stromal cells further strengthen and stretch the ECM, causing a snowball effect.

The biochemical composition and organization of the ECM also influences tumor biology. Dysregulation of normal matrix signals can lead to tumor progression, characterized by excessive cell proliferation, immortality, enhanced migration, changes in metabolism, and evasion of the immune response. More research is needed to dissect the relationships between the ECM’s mechanical properties, forces, and cell signaling pathways.

Targeting the ECM

Because unchecked proliferation of cancer cells increases solid stress in the tumor, anticancer therapies should decrease the compressive forces in tumors and reopen collapsed blood and lymphatic vessels.11 This is exactly what happens when tumors are treated with certain doses of paclitaxel or docetaxel, two widely used cancer drugs. Shrinking tumors increases blood flow and allows more efficient fluid movement through the extravascular space, lowering the tumor interstitial fluid pressure in mouse models and in patients with breast cancer.5 However, cancer cells invariably develop resistance to treatment and begin to regrow, increasing solid stress again. As a result, other targets for reducing solid stresses are needed.

Because of its role in containing and concentrating the forces in a tumor, the collagen matrix within and around the tumor is another potential target for relieving tumor-related stresses. Indeed, solid stress in tumors can be reduced by drugs that selectively reprogram activated fibroblasts or modify the assembly of matrix components such as collagen and hyaluronan. In rodent studies, targeting these force-altering components in the tumor microenvironment has been shown to decrease solid stress, improve blood perfusion and drug delivery, and improve tumor response to chemotherapy and animal survival.6 We have found, for example, that injecting tumors with a collagen-digesting enzyme increases the diffusion of antibodies and viral particles and improves drug penetration in the tumor. Similarly, treatments that target transforming growth factor–beta (TGF-β), which controls the production of collagen by myofibroblasts, increase perfusion, improve the delivery of drugs of all sizes in mammary tumors, and improve treatment outcomes in mice.5

A class of drugs that is widely used to control blood pressure in hypertensive patients also blocks the TGF-β pathway. These drugs, known as angiotensin receptor 1 blockers, can reduce collagen production in and around the tumor by reducing the activity of TGF-β, as well as by blocking the function of connective tissue growth factor (CTGF), which is involved in stabilizing collagen and inducing resistance to chemotherapy.6Losartan and other angiotensin inhibitors reduce levels of collagen in various experimental models of fibrosis, and decrease renal and cardiac fibrosis in hypertensive patients. When given to mice with one of four different types of tumors characterized by high levels of cancer-associated fibroblasts (CAFs) and excess extracellular matrix—pancreatic ductal adenocarcinoma, breast cancer, sarcoma, and melanoma—losartan treatment caused a decrease in collagen content in a dose-dependent manner, enhanced penetration of nanoparticles into the tumor, and improved efficacy of diverse anticancer drugs. This is supported by a number of retrospective studies in patients with pancreatic, lung, and kidney cancers.6Researchers at Massachusetts General Hospital are now running a Phase 1/2 clinical trial to test losartan in pancreatic cancer patients.

THE TUMOR ENVIRONMENT: The extracellular matrix and stromal cells within a tumor’s microenvironment influence the physical forces a tumor experiences. Left: The immunofluorescent image shows stromal cells (red and green) surrounding tumor cells (red cluster with blue nuclei); the cells were isolated from a mouse model of lung adenocarcinoma. Right: In this immunofluorescent image of triple-negative breast cancer, tumor cells (blue) are in close contact with matrix collagen (purple). Immune cells are labeled in red and green.VASILENA GOCHEVA, JACKS LAB, KOCH INSTITUTE AT MIT; DONGMEI ZUO, LABORATORY DR. MORAG PARK

Another potential cancer treatment target is hyaluronan, which is abundant in 20 percent to 30 percent of human tumors, most notably breast, colon, and prostate cancers. In addition to its role as a pressure-creating gel, hyaluronan can sequester growth factors and inhibit interstitial fluid movement within the tumor. Hyaluronidase, an enzyme that digests hyaluronan, reduces mechanical stress in tumors grown in mice.1 And San Diego–based Halozyme Therapeutics’s PEGPH20, a formulation of hyaluronidase coated with polyethylene glycol to enhance bioavailability, can decompress blood vessels and improve treatment outcome in genetically engineered mouse models of pancreatic ductal adenocarcinoma. Based on these studies, Halozyme researchers are now testing PEGPH20 in a randomized clinical trial of pancreatic cancer patients. Another matrix-altering drug is the widely-prescribed antidiabetic drug metformin, which has been shown to decrease collagen and hyaluronan levels in pancreatic tumors in obese mice and patients.7 Metformin is currently being tested in more than 200 clinical trials worldwide as a treatment for different types of cancer.

Clearly, tumors should be studied not only in light of their biochemical processes and genetic underpinnings, but also for the specific physical forces and mechanical properties that may influence progression. Understanding the physical microenvironment of tumors, as well as its interplay with the biochemical environment, is necessary to improve cancer detection, prevention, and treatment.

  1. T. Stylianopoulos et al., “Causes, consequences, and remedies for growth-induced solid stress in murine and human tumors,” PNAS, 109:15101-08, 2012.
  2. R.K. Jain, L.T. Baxter, “Mechanisms of heterogeneous distribution of monoclonal antibodies and other macromolecules in tumors: Significance of elevated interstitial pressure,” Cancer Res, 48:7022-32, 1988.
  3. R.K. Jain, “Normalization of tumor vasculature: An emerging concept in antiangiogenic therapy,”Science, 307:58-62, 2005.
  4. R.K. Jain, “Antiangiogenesis strategies revisited: From starving tumors to alleviating hypoxia,”Cancer Cell, 26:605-22, 2014.
  5. R.K. Jain et al., “The role of mechanical forces in tumor growth and therapy,” Annu Rev Biomed Eng, 16:321-46, 2014.
  6. V.P. Chauhan et al., “Angiotensin inhibition enhances drug delivery and potentiates chemotherapy by decompressing tumour blood vessels,” Nat Commun, 4:2516, 2013.
  7. J. Incio et al., “Metformin reduces desmoplasia in pancreatic cancer by reprogramming stellate cells and tumor-associated macrophages,” PLOS ONE, 10:e0141392, 2015.
  8. M.A. Swartz, A.W. Lund, “Lymphatic and interstitial flow in the tumour microenvironment: linking mechanobiology with immunity,” Nat Rev Cancer, 12:210-19, 2012.
  9. H. Qazi et al., “Cancer cell glycocalyx mediates mechanotransduction and flow-regulated invasion,”Integr Biol, 5:1334-43, 2013.
  10. J.W. Song, L.L. Munn, “Fluid forces control endothelial sprouting,” PNAS, 108:15342-47, 2011.

Read Full Post »

Glypican-1 identifies cancer exosomes

Larry H. Bernstein, MD, FCAP, Curator


Glypican-1 identifies cancer exosomes and detects early pancreatic cancer

Sonia A. MeloLinda B. LueckeChristoph KahlertAgustin F. FernandezSeth T. GammonJudith Kaye, et al.

Nature (09 July 2015); 523: 177–182

Most cells shed so-called extracellular vesicles or exosomes consisting of proteins and nucleic acids enclosed in phospholipid bilayers. Exosomes derived from cancer cells can be isolated.

Exosomes are lipid-bilayer-enclosed extracellular vesicles that contain proteins and nucleic acids. They are secreted by all cells and circulate in the blood. Specific detection and isolation of cancer-cell-derived exosomes in the circulation is currently lacking. Using mass spectrometry analyses, we identify a cell surface proteoglycan, glypican-1 (GPC1), specifically enriched on cancer-cell-derived exosomes. GPC1+ circulating exosomes (crExos) were monitored and isolated using flow cytometry from the serum of patients and mice with cancer. GPC1+ crExos were detected in the serum of patients with pancreatic cancer with absolute specificity and sensitivity, distinguishing healthy subjects and patients with a benign pancreatic disease from patients with early- and late-stage pancreatic cancer. Levels of GPC1+ crExos correlate with tumour burden and the survival of pre- and post-surgical patients. GPC1+ crExos from patients and from mice with spontaneous pancreatic tumours carry specific KRAS mutations, and reliably detect pancreatic intraepithelial lesions in mice despite negative signals by magnetic resonance imaging. GPC1+ crExos may serve as a potential non-invasive diagnostic and screening tool to detect early stages of pancreatic cancer to facilitate possible curative surgical therapy.

Figure 1: GPC1 is present on cancer exosomes.

GPC1 is present on cancer exosomes.

a, Venn diagram of proteins from NIH/3T3 (blue), MCF10A (red), HDF (green), E10 (yellow) and MDA-MB-231 (purple) exosomes. In total, 48 proteins were exclusively detected in MDA-MB-231 exosomes (n = 3 protein samples,…

Figure 2: GPC1+ crExos are a non-invasive biomarker for pancreatic cancer.

GPC1+ crExos are a non-invasive biomarker for pancreatic cancer.

a, Percentage of GPC1+crExo beads in healthy donors, patients with breast cancer and patients with PDAC (analysis of variance (ANOVA), post-hoc Tamhane T2, ****P < 0.0001). b, Frequency ofKRAS mutation in 47 tumours…

Figure 3: Levels of circulating GPC1+exosomes inform pancreatic cancer resection outcome.

Levels of circulating GPC1+ exosomes inform pancreatic cancer resection outcome.

a, Longitudinal blood collection pre- and post-operatively (day 7). b, Percentage of GPC1+crExo beads from patients with BPD (n = 4), PCPL (n = 4) or PDAC (n = 29) (paired two-tailed Student’s t-test, **P < 0.01, ****P < 0.0001). Data a…

Cancer: Diagnosis by extracellular vesicles

Nature (09 July 2015); 523: 161–162.

The detection of a single molecule anchored to circulating extracellular vesicles allows late-stage pancreatic cancer to be identified from just one drop of a patient’s blood. See Article p.177

ReferencesAuthor information

  1. Melo, S. A. et al. Nature 523, 177182 (2015).    Article
  2. Kleeff, J. et al. J. Clin. Invest. 102, 16621673 (1998).

3.   Aikawa, T. et al. J. Clin. Invest. 118, 8999 (2008).

4.  Whipple, C. A., Young, A. L. & Korc, M. Oncogene 31, 25352544 (2012).

  • Kim, J. W. et al. Clin. Cancer Res. 11, 10101020 (2005).
  • Zhou, H. et al. Kidney Int. 70, 18471857 (2006).
  • Skog, J. et al. Nature Cell Biol. 10, 14701476 (2008).
  • Peinado, H. et al. Nature Med. 18, 883891 (2012).
  • Logozzi, M. et al. PLoS ONE 4, e5219 (2009).
  • Jakobsen, K. R. et al. J. Extracell. Vesicles 4, 26659 (2015).
  • Madhavan, B. et al. Int. J. Cancer 136, 26162627 (2015).
  • Trams, E. G., Lauter, C. J., Salem, N. Jr & Heine, U. Biochim. Biophys. Acta 645, 6370(1981).
  • Johnstone, R. M., Adam, M., Hammond, J. R., Orr, L. & Turbide, C. J. Biol. Chem. 262,94129420 (1987).
  • Gould, S. J. & Raposo, G. J. Extracell. Vesicles 2, 20389 (2013).

Oxidative stress inhibits distant metastasis by human melanoma cells

Elena PiskounovaMichalis AgathocleousMalea M. MurphyZeping HuSara E. HuddlestunZhiyu Zhao, et al.

Nature 14 Oct 2015

Solid cancer cells commonly enter the blood and disseminate systemically, but are highly inefficient at forming distant metastases for poorly understood reasons. Here we studied human melanomas that differed in their metastasis histories in patients and in their capacity to metastasize in NOD-SCID-Il2rg−/− (NSG) mice. We show that melanomas had high frequencies of cells that formed subcutaneous tumours, but much lower percentages of cells that formed tumours after intravenous or intrasplenic transplantation, particularly among inefficiently metastasizing melanomas. Melanoma cells in the blood and visceral organs experienced oxidative stress not observed in established subcutaneous tumours. Successfully metastasizing melanomas underwent reversible metabolic changes during metastasis that increased their capacity to withstand oxidative stress, including increased dependence on NADPH-generating enzymes in the folate pathway. Antioxidants promoted distant metastasis in NSG mice. Folate pathway inhibition using low-dose methotrexate, ALDH1L2 knockdown, or MTHFD1 knockdown inhibited distant metastasis without significantly affecting the growth of subcutaneous tumours in the same mice. Oxidative stress thus limits distant metastasis by melanoma cells in vivo.

Lymph node-independent liver metastasis in a model of metastatic colorectal cancer

Ida B. EnquistZinaida GoodAdrian M. JubbGermaine FuhXi WangMelissa R. JunttilaErica L. Jackson & Kevin G. Leong

Nature Communications  26 Mar 2014; 3530(5)

Deciphering metastatic routes is critically important as metastasis is a primary cause of cancer mortality. In colorectal cancer (CRC), it is unknown whether liver metastases derive from cancer cells that first colonize intestinal lymph nodes, or whether such metastases can form without prior lymph node involvement. A lack of relevant metastatic CRC models has precluded investigations into metastatic routes. Here we describe a metastatic CRC mouse model and show that liver metastases can manifest without a lymph node metastatic intermediary. Colorectal tumours transplanted onto the colonic mucosa invade and metastasize to specific target organs including the intestinal lymph nodes, liver and lungs. Importantly, this metastatic pattern differs from that observed following caecum implantation, which invariably involves peritoneal carcinomatosis. Anti-angiogenesis inhibits liver metastasis, yet anti-lymphangiogenesis does not impact liver metastasis despite abrogating lymph node metastasis. Our data demonstrate direct hematogenous spread as a dissemination route that contributes to CRC liver malignancy.

Comprehensive models of human primary and metastatic colorectal tumors in immunodeficient and immunocompetent mice by chemokine targeting

Huanhuan Joyce ChenJian SunZhiliang HuangHarry Hou JrMyra ArcillaNikolai RakhilinDaniel J JoeJiahn ChoiPoornima GadamsettyJeff MilsomGovind NandakumarRandy LongmanXi Kathy Zhou, et al.

Nature Biotechnology (2015);  33:656–660

Current orthotopic xenograft models of human colorectal cancer (CRC) require surgery and do not robustly form metastases in the liver, the most common site clinically. CCR9 traffics lymphocytes to intestine and colorectum. We engineered use of the chemokine receptor CCR9 in CRC cell lines and patient-derived cells to create primary gastrointestinal (GI) tumors in immunodeficient mice by tail-vein injection rather than surgery. The tumors metastasize inducibly and robustly to the liver. Metastases have higher DKK4 and NOTCH signaling levels and are more chemoresistant than paired subcutaneous xenografts. Using this approach, we generated 17 chemokine-targeted mouse models (CTMMs) that recapitulate the majority of common human somatic CRC mutations. We also show that primary tumors can be modeled in immunocompetent mice by microinjecting CCR9-expressing cancer cell lines into early-stage mouse blastocysts, which induces central immune tolerance. We expect that CTMMs will facilitate investigation of the biology of CRC metastasis and drug screening.

Induction of the intestinal stem cell signature gene SMOC-2 is required for L1-mediated colon cancer progression

A Shvab, G Haase, A Ben-Shmuel, N Gavert, T Brabletz, S Dedhar and A Ben-Ze’ev

Oncogene , (27 April 2015) |

Overactivation of Wnt-β-catenin signaling, including β-catenin-TCF target gene expression, is a hallmark of colorectal cancer (CRC) development. We identified the immunoglobulin family of cell-adhesion receptors member L1 as a β-catenin-TCF target gene preferentially expressed at the invasive edge of human CRC tissue. L1 can confer enhanced motility and liver metastasis when expressed in CRC cells. This ability of L1-mediated metastasis is exerted by a mechanism involving ezrin and the activation of NF-κB target genes. In this study, we identified the secreted modular calcium-binding matricellular protein-2 (SMOC-2) as a gene activated by L1-ezrin-NF-κB signaling. SMOC-2 is also known as an intestinal stem cell signature gene in mice expressing Lgr5 in cells at the bottom of intestinal crypts. The induction of SMOC-2 expression in L1-expressing CRC cells was necessary for the increase in cell motility, proliferation under stress and liver metastasis conferred by L1. SMOC-2 expression induced a more mesenchymal like phenotype in CRC cells, a decrease in E-cadherin and an increase in Snail by signaling that involves integrin-linked kinase (ILK). SMOC-2 was localized at the bottom of normal human colonic crypts and at increased levels in CRC tissue with preferential expression in invasive areas of the tumor. We found an increase in Lgr5 levels in CRC cells overexpressing L1, p65 or SMOC-2, suggesting that L1-mediated CRC progression involves the acquisition of a stem cell-like phenotype, and that SMOC-2 elevation is necessary for L1-mediated induction of more aggressive/invasive CRC properties.

Global analysis of L1-transcriptomes identified IGFBP-2 as a target of ezrin and NF-κB signaling that promotes colon cancer progression

A Ben-Shmuel, A Shvab, N Gavert, T Brabletz and A Ben-Ze’ev

Oncogene 06 Aug 2012; Oncogene  (04 July 2013); 32: 3220-3230 |

L1, a neuronal cell adhesion receptor of the immunoglobulin-like protein family is expressed in invading colorectal cancer (CRC) cells as a target gene of Wnt/β-catenin signaling. Overexpression of L1 in CRC cells enhances cell motility and proliferation, and confers liver metastasis. We recently identified ezrin and the IκB-NF-κB pathway as essential for the biological properties conferred by L1 in CRC cells. Here, we studied the underlying molecular mechanisms and found that L1 enhances ezrin phosphorylation, via Rho-associated protein kinase (ROCK), and is required for L1–ezrin co-localization at the juxtamembrane domain and for enhancing cell motility. Global transcriptomes from L1-expressing CRC cells were compared with transcriptomes from the same cells expressing small hairpin RNA (shRNA) to ezrin. Among the genes whose expression was elevated by L1 and ezrin we identified insulin-like growth factor-binding protein 2 (IGFBP-2) and showed that its increased expression is mediated by an NF-κB-mediated transactivation of the IGFBP-2 gene promoter. Expression of a constitutively activated mutant ezrin (Ezrin567D) could also increase IGFBP-2 levels in CRC cells. Overexpression of IGFBP-2 in CRC cells lacking L1-enhanced cell proliferation (in the absence of serum), cell motility, tumorigenesis and induced liver metastasis, similar to L1 overexpression. Suppression of endogenous IGFBP-2 in L1-transfected cells inhibited these properties conferred by L1. We detected IGFBP-2 in a unique organization at the bottom of human colonic crypts in normal mucosa and at increased levels throughout human CRC tissue samples co-localizing with the phosphorylated p65 subunit of NF-κB. Finally, we found that IGFBP-2 and L1 can form a molecular complex suggesting that L1-mediated signaling by the L1–ezrin–NF-κB pathway, that induces IGFBP-2 expression, has an important role in CRC progression.


Exosome Scouts Help Tumors Populate Distant Organs

  • Click Image To Enlarge +
    This image shows exosomes (green) that have infiltrated the whole lung. [Ayuko Hoshino, David Lyden, Weill Cornell Medicine

    When certain types of cancer spread, they seem to prefer particular organs in the body, a choosiness that led Stephen Paget to propose the “seed and soil” hypothesis. This hypothesis, now more than 100 years old, suggests that different organs are somehow more receptive to certain types of cancer, just as different soils seem to allow some seeds, but not others, to find purchase.

    While this hypothesis is as expressive as ever, it still lacks detail. It doesn’t suggest what mechanisms might drive organ-specific metastasis, or organotropic metastasis. The hypothesis, however, is being taken farther by researchers based at Weill Cornell Medicine. These researchers suggest that the old seed-and-soil idea, which sounds as haphazard as the dispersal of seeds by uncultivated plants, could be updated to describe a process that is more directed.

    Essentially, a tumor metastasis may proceed the way settlers cultivate new land. First, scouts and pioneers are dispatched to identify fertile spots and develop basic infrastructure. Then, once the ground is prepared, settlers establish new communities.

    In this scenario, the scouts are tumor exosomes. These exosomes are released by tumors in the millions, and they carry samples of the tumors’ proteins and genetic content. They fuse preferentially with cells at specific locations, and they ensure that recipient organs are prepared to host the tumor cells they represent.

    Most important, this updated view of organotropic metastasis includes a mechanism to explain how exosomes are directed to specific organs. The exosomes, it turns out, are outfitted with particular sets of integrins, proteins that serve as a kind of destination label.

    Supportive findings appeared October 28 in the journal Nature, in an article entitled, “Tumour exosome integrins determine organotropic metastasis.” This article described how the Weill Cornell researchers, in collaboration with scientists from the Memorial Sloan Kettering Cancer center and the Spanish National Cancer Research Centre (CNIO), examined exosomes from mouse and human lung-, liver-, and brain-tropic tumor cells. These exosomes were seen to fuse preferentially with resident cells at their predicted destinations, namely, lung fibroblasts and epithelial cells, liver Kupffer cells, and brain endothelial cells.

    “Exosome proteomics revealed distinct integrin expression patterns, in which the exosomal integrins α6β4 and α6β1 were associated with lung metastasis, while exosomal integrin αvβ5 was linked to liver metastasis,” wrote the authors. “Targeting the integrins α6β4 and αvβ5 decreased exosome uptake, as well as lung and liver metastasis, respectively.”

    In other words, the study demonstrated the importance of integrins in metastatic nesting by blocking specific integrins in tumors that metastasize to specific organs. For example, when integrins were blocked in breast cancer, metastasis to lungs was reduced. Similarly, when integrins were blocked in pancreatic cancer, metastasis to liver was reduced.

    In addition, the study showed that a tumor could be “tricked” by changing the integrin destination code of its exosomes. For example, a tumor that would normally go to the bones could be directed to the lungs instead.

    “The integrin-specific signature that we identified may have significant value clinically, serving as a prognostic indicator for metastasis to specific organ sites,” said senior author David Lyden, M.D., Ph.D., the Stavros S. Niarchos Professor in Pediatric Cardiology and a professor of pediatrics and of cell and developmental biology at Weill Cornell Medicine. “Instead of waiting for late-stage metastasis, we can now initiate preventative strategies at an earlier point of disease progression with the hope of preventing its spread. This really changes the treatment paradigm.”


  • Using CRISPR as a High-Throughput Cancer Screening and Modeling Tool
  • Click Image To Enlarge +
    Using CRISPR/Cas9, scientists created a new high-throughput screening tool for studying the development and progression of liver cancer in mice. [Ernesto del Aguila III, NHGRI]

    A contingent of researchers from the UK, Germany, and Spain have recently developed a novel CRISPR/Cas9 system that they believe can be utilized as a multiplexed screening approach to study and model cancer development in mice. In the current study, the investigators directly mutated genes within adult mouse livers to elucidate their role in cancer development and progression—simultaneously uncovering the gene combinations that coordinate to cause liver cancer.

    “We reasoned that, by targeting mutations directly to adult liver cells using CRISPR/Cas9, we could better study and understand the biology of this important cancer,” explained co-author Mathias Friedrich, Ph.D., research scientist at the Wellcome Trust Sanger Institute. “Other approaches to engineer mutations in mice, such as stem cell manipulation, are limited by the laborious process, the long time frames and large numbers of animals needed. And, our method better mimics important aspects of human cancer biology than many “classic” mouse models: as in most human cancers, the mutations occur in the adult and only affect a few cells”.

    The findings from this study were published online recently in PNAS through an article entitled “CRISPR/Cas9 somatic multiplex-mutagenesis for high-throughput functional cancer genomics in mice.”

    This new approach is rapid, scalable, and extremely efficient, allowing the researchers to examine an array of genes or large regions of the genome concurrently. Moreover, this methodology affords scientists the ability to distinguish between cancer driver mutations and passenger mutations—those that occur as side-effects of cancer development.

    The research team developed a list of up to eighteen genes with known or unknown evidence for their importance in two forms of liver cancer. They then introduced the CRISPR/Cas9 molecules, targeting various combinations of these genes into mice, which subsequently developed liver or bile duct cancer within a few months.

    “Our approach enables us to simultaneously target multiple putative genes in individual cells,” noted co-author Roland Rad, Ph.D., project leader at the Technical University of Munich and the German Cancer Research Center Heidelberg. “We can now rapidly and efficiently screen which genes are cancer-causing and which ones are not. And, we can study how genes work together to cause cancers—a crucial piece of the puzzle we must solve to understand and tackle the disease.”

    The investigators were able to confirm that a set of DNA-binding proteins called ARID (AT-rich interactive domain), influence the organization of chromosomes and are important for liver cancer development. Furthermore, mutations in a second protein, TET2, were found to be causative in bile duct cancer: although TET2 has not been found to be mutated in human biliary cancers, the proteins that it interacts with have been, showing that the CRISPR/Cas9 method can identify human cancer genes that are not mutated, but whose function is disturbed by other events.

    “The new tools of targeting genes in combination and inducing insertions or deletions in chromosomes change our ability to identify new cancer-causing genes and to understand their role in cancer,” stated senior group leader and co-author Allan Bradley, Ph.D., director emeritus from the Sanger Institute. “Our results show that this approach is feasible and productive in liver cancer; we will now continue to study our new findings and try to extend the approach to other cancer types.”

    This CRISPR/Cas9 approach may also be favorable for an in-depth examination of genomic deserts —regions within the human genome that appear to be devoid of genes. Yet, recent data from the ENCODE Project suggests that deserts can be populated, if not by genes, then by DNA regulatory regions that influence the activity of genes.

    “Liver cancer has many DNA alterations in regions lacking genes: we don’t know which of these might be important for the disease,” said Dr. Rad. “However, we could show that it is now possible to delete such regions to systematically determine their role in liver cancer development.”


CRISPR Used to Create Mouse Models of Cancer

  • When scientists study the genetics of cancer, they often breed mice strains that carry selected cancer-associated mutations. But cultivating such strains, usually via transgenesis or gene targeting in embryonic stem cells, is often time-consuming and expensive. Could there be a better way—a faster, cheaper way—to create mice strains that carry particular genetic flaws?

    An alternative has been proposed by researchers from MIT. They have shown that the CRISPR gene editing system can introduce cancer-causing mutations into the livers of adult mice. The researchers anticipate that their method will allow for more rapid testing of any single genes or gene combinations that are suspected of being capable of initiating tumor formation in the liver.

    “The sequencing of human tumors has revealed hundreds of oncogenes and tumor suppressor genes in different combinations. The flexibility of this technology, as delivery gets better in the future, will give you a way to pretty rapidly test those combinations,” said Phillip Sharp, Ph.D., a professor at MIT’s Koch Institute for Integrative Cancer Research.

    Dr. Sharp was part of the MIT research team, which was led by Koch Institute director Tyler Jacks, Ph.D. Dr. Jacks noted that the CRISPR technique, which not only provides the ability to delete genes, but also to replace them with altered versions, “really opens up all sorts of new possibilities when you think about the kinds of genes that you would want to mutate in the future.” Both loss of function and gain of function, he explained, are possible.

    The MIT researchers presented their results August 6 in Nature, in an article entitled, “CRISPR-mediated direct mutation of cancer genes in the mouse liver.” It described how cancer models were generated using the CRISPR/Cas (clustered regularly interspaced short palindromic repeats/CRISPR-associated proteins) system in vivo in wild-type mice.

    “We used hydrodynamic injection to deliver a CRISPR plasmid DNA expressing Cas9 and single guide RNAs (sgRNAs) to the liver that directly target the tumor suppressor genes Pten and p53 (also known as TP53 and Trp53), alone and in combination,” wrote the authors. “CRISPR-mediated Pten mutation led to elevated Akt phosphorylation and lipid accumulation in hepatocytes, phenocopying the effects of deletion of the gene using Cre–LoxP technology. Simultaneous targeting of Pten and p53 induced liver tumors that mimicked those caused by Cre–loxP-mediated deletion of Pten and p53.”

    Studies of such genetically engineered mice have yielded many important discoveries, but the process, which requires introducing mutations into embryonic stem cells, can take more than a year and costs hundreds of thousands of dollars. Using Cas enzymes targeted to cut snippets of p53 and Pten, the researchers were able to disrupt those two genes in about 3% of liver cells, enough to produce liver tumors within three months.

    With traditional techniques, genetically engineering such models is “a very long process,” commented Dr. Jacks. “And the more genes you’re working with, the longer and more complicated it becomes.

    The researchers also used CRISPR to create a mouse model with an oncogene called beta catenin, which makes cells more likely to become cancerous if additional mutations occur later on. To create this model, the researchers had to cut out the normal version of the gene and replace it with an overactive form, which was successful in about 0.5% of hepatocytes.

    In the Nature article, the authors emphasized that simplified methods of testing the oncogenic properties of candidates in vivo are critical. In particular, they cited the need to somehow evaluate the thousands of candidate cancer genes that are being discovered through next-generation sequencing efforts.

    Already looking forward to refining their method of generating cancer models, the authors suggested that it could attain greater sensitivity if CRISPR/Cas9-mediated mutagenesis could be performed on a “sensitized” background carrying constitutive or conditional mutations in a tumor suppressor gene such as p53. “More efficient delivery techniques, such as adenovirus or adeno-associated virus, more potent sgRNAs, and longer homologous recombination templates,” they wrote, “might also improve the overall efficiency of this method and expand the range of tissue that could be targeted.”



Bioinformatics beyond Genome Crunching  

Flow Cytometry, Workflow Development, and Other Information Stores Can Become Treasure Troves If You Use the Right IT Tools and Services

  • Click Image To Enlarge +
    Shown here is the FlowJo platform’s visualization of surface activation marker expression (CD38) on live lymphocyte CD8+ T cells. Colors represent all combinations of subsets positive and negative for interferon gamma (IFN?), perforin (Perf), and phosphorylated ERK (pERK).

    Advances in bioinformatics are no longer limited to just crunching through genomic and exosomic data. Bioinformatics, a discipline at the interface between biotechnology and information technology, also has lessons for flow cytometry and experimental design, as well as database searches, for both internal and external content.

    One company offering variations on traditional genome crunching is DNAnexus. With the advent of the $1,000 genome, researchers find themselves drowning in data. To analyze the terabytes of information, they must contract with an organization to provide the computing power, or they must perform the necessary server installation and maintenance work in house.

Read Full Post »