Leaders in Pharmaceutical Business Intelligence (LPBI) Group

Summary of Metabolomics

Author and Curator: Larry H. Bernstein, MD, FCAP 

This concludes the series on metabolomics, a rapidly developing science that is interconnected with a group termed – OMICS: proteomics, transcriptomics, genomics, and metabolomics.  This chapter is most representative of the many important studies being done in the field, which ranges most widely because it has opened doors into nutrition and nutritional supplements, plant biochemistry, agricultural crops and breeding, animal breeding, worldwide malnutrition, diabetes, cancer, neurosciences, circulatory, respiratory, and musculosletal disorders, infectious diseases and immune system disorders.  Obviously, it is not possible to cover the full range of activity, but metabolomics is most comprehensive in exploring the full range of metabolic changes that occur in health during the full age range from development to the geriatric years.  It can be integrated well with gene expression, proteomics studies, and epidemiological investigations.

The subchapters are given here:

7.1   Extracellular evaluation of intracellular flux in yeast cells  

 http://pharmaceuticalintelligence.com/2014/08/25/extracellular-evaluation-of-intracellular-flux-in-yeast-cells/

 7.2    Metabolomic analysis of two leukemia cell lines. I.  

         http://pharmaceuticalintelligence.com/2014/08/23/metabolomic-analysis-of-two-leukemia-cell-lines-_i/

  7.3   Metabolomic analysis of two leukemia cell lines. II.

           http://pharmaceuticalintelligence.com/2014/08/24/metabolomic-analysis-of-two-leukemia-cell-lines-ii/

  7.4   Buffering of genetic modules involved in tricarboxylic acid cycle metabolism provides homeostatic
regulation  

           http://pharmaceuticalintelligence.com/2014/08/27/buffering-of-genetic-modules-involved-in-tricarboxylic-acid-cycle-metabolism-provides-homeomeostatic-regulation/

  7.5   Metabolomics, Metabonomics and Functional Nutrition: the next step in nutritional metabolism and
biotherapeutics

 http://pharmaceuticalintelligence.com/2014/08/22/metabolomics-metabonomics-and-functional-nutrition-the-next-step-in-nutritional-metabolism-and-biotherapeutics/

    7.6    Isoenzymes in cell metabolic pathways

 http://pharmaceuticalintelligence.com/2014/10/06/isoenzymes-in-cell-metabolic-pathways/

7.7   A Brief Curation of Proteomics, Metabolomics, and Metabolism

http://pharmaceuticalintelligence.com/2014/10/03/a-brief-curation-of-proteomics-metabolomics-and-metabolism/

   7.8   Metabolomics is about Metabolic Systems Integration

     http://pharmaceuticalintelligence.com/2014/10/13/metabolomics-is-about-metabolic-systems-integration/

 7.9  Mechanisms of Drug Resistance

   http://pharmaceuticalintelligence.com/2014/10/09/mechanisms-of-drug-resistance/

7.10  Development Of Super-Resolved Fluorescence Microscopy

    http://pharmaceuticalintelligence.com/2014/10/12/development-of-super-resolved-fluorescence-microscopy  

7.11  Metabolic Reactions Need Just Enough

 http://pharmaceuticalintelligence.com/2014/10/14/metabolic-reactions-need-just-enough/

7.12  Metabolomics Summary and Perspective

   This chapter will be followed by an exploration of disease and pharmaceutical directed studies using these methods  8. Impairments in pathological states: endocrine disorders; stress
hypermetabolism; cancer.

Networking metabolites and diseases

P Braun, E Rietman, and M Vidal
Center for Cancer Systems Biology and Department of Cancer Biology, Dana-Farber Cancer Institute, and Department of Genetics, Harvard Medical School,  Boston, MA; and Physical Sciences Inc., Andover, MA 01810
PNAS July 22, 2008; 105(29): 9849–9850.    http://pnas.org/cgi/doi/10.1073/pnas.0805644105

Biological systems are increasingly viewed and analyzed as

• highly complex networks of interlinked macromolecules and metabolites.

Network analysis has been applied to

• interactome maps of protein–protein, protein–DNA, and protein–RNA interactions
• as well as transcriptional, metabolic, and genetic data.

Such network views of biological systems should facilitate the detection of

• nonlinear long-range effects of perturbations, for example, by mutations, and
• help identification of unanticipated indirect causal connections.

Diseasome and Drug-Target Network

Recently, Goh et al. (1) constructed a ‘‘diseasome’’ network in which

• two diseases are linked to each other if
• they share at least one gene, in which mutations are associated with both diseases.

In the resulting network, related disease families cluster tightly together, thus

• phenotypically defining functional modules.

Importantly, for the first time this study applied concepts from network biology to human diseases,

• thus opening the door for discovering causal relationships between
• disregulated networks and resulting ailments.

Subsequently Yilderim et al. (2) linked drugs to protein targets in a drug–target network,

• which could then be overlaid with the diseasome network.

One notable finding was the recent trend toward the development of

• new compounds directly targeted at disease gene products, whereas previous drugs,
• often found by trial and error, appear to target proteins only indirectly related to
• the actual disease molecular mechanisms.

An important question that remains in this emerging field of network analysis consists of

• investigating the extent to which directly targeting the product of mutated genes is an efficient approach or
• whether targeting network properties instead, and
• thereby accounting for indirect nonlinear effects of system perturbations by drugs, may prove more fruitful.

However, to answer such questions it is important to have a good understanding of the various influences that can lead to diseases.

Metabolic Connections

One group of diseases that was very poorly connected in the original diseasome network was the family of metabolic diseases.

In this issue of PNAS, Lee et al. (3) hypothesize that metabolic diseases may instead be connected

• via metabolites and common reactions.

To investigate this hypothesis Lee et al. first constructed a metabolic network from data available in

• two manually curated databases detailing well known

1. metabolic reactions,
2. the involved metabolites, and
3. catalyzing enzymes.

In addition, gene–disease associations were identified by using the Online Mendelian Inheritance in Man (OMIM) database (http://ncbi.nlm.nih.gov/sites/
entrez?dbomim&itooltoolbar). In a last step,

• a metabolic disease network (MDN) was constructed by connecting
• two diseases if their associated genes are linked in the metabolic network
• by a common metabolite or metabolites used in a common reaction.

Metabolites are not only linked by common reactions, but

• on a larger scale by coupled fluxes within a metabolic network,
• which may also influence disease phenotypes.

An increase in the concentration of one metabolite may increase several fluxes

• across reaction pathways that use this compound, which
• may lead to diverse phenotypes and distinct diseases.

The fluxes within the metabolic network are calculated by using

• the Flux Coupling Finder method described by Nikolaev et al. (4) and Burgard et al. (5),
• which is based on the assumption that pools of metabolites are conserved.

To functionally validate the network, coexpression correlations are measured for genes

• linked by adjacent reactions and those linked by fluxes.

Interestingly, the average coexpression correlation for flux-coupled genes (0.31)

• is higher than that for genes simply catalyzing adjacent reactions (0.24)
  (compared with 0.10 for all gene pairs in the network).

If the links between diseases identified in the MDN are functionally and causally relevant

• it should be expected that linked diseases occur more frequently in the same individual.

To test this hypothesis, Lee et al. (3) measured the co-occurrence of diseases in patients by using detailed Medicare information

• of 13 million patients and 32 million hospital visits within a 3-year period.

A comorbidity index was computed to measure the degree to which one disease

• will increase the likelihood of a second disease in the same patient.

The average comorbidity for all genes is 0.0008 (Pearson correlation coefficient),

• which increases 3-fold to 0.0027 when disease pairs that are metabolically linked are analyzed,
• which is highly statistically significant (P 108).

When diseases are analyzed that are directionally coupled by a flux (see ref. 3 for details),

• the correlation increases to 0.0062.

Thus, whereas 17% of all diseases in the network show significant comorbidity, this fraction

• nearly doubles to 31% for metabolically linked diseases.

Further analysis reveals that comorbidity effects can be detected up to three links (metabolites, reactions)

• apart from each other with statistical significance, but not farther away.

In the MDN, several highly connected hubs, e.g., hypertension and hemolytic anemia, are

• linked to many different co-occurring diseases not unexpected for such complex diseases
• that can result from many different genetic alterations or variants.

Importantly, though, most of the connections to the different linked diseases

• are mediated by diverse connections in the metabolic network.

Thus, in the future such insights may be helpful for finer classification of the complex hub disease.

Furthermore, depending on the onset of the complex (hub) disease in relation to the associated diseases,

• such relationships may potentially be used to systematically
• stratify patients and develop targeted treatments acting on
• the underlying metabolic links.

Returning to the starting point of their study, Lee et al. (3) next investigated

• whether metabolic diseases are better linked through the metabolic network
• than they are in the previously described gene–disease network.

When purely metabolic diseases are considered, the comorbidity is, in fact,

• best predicted by metabolic links.

Interestingly, when all diseases linked to metabolic enzymes are considered,

• which involves many diseases that are merely related to metabolic diseases through multifunctional enzymes,
• the gene and metabolic networks are nearly equally predictive of comorbidity,
• indicating that as a general approach information from
• many different biological dimensions should be integrated to identify the most relevant connections.

Together, all these findings support the initial hypothesis that metabolic diseases are linked by metabolic networks.

Practically, alteration of one metabolite or one reaction can have numerous repercussions in the network,

• each of which can manifest as different diseases that frequently occur together in affected patients.

Radoslav Bozov

1. Glycine, as the only amino acid having no isomer driven central carbon allowing for hing occupancy of ‘free’
   motifs, where quark (proton) ‘fluxes’ play at, is a one – step away observable (1) from synthesis of pyrimidines
   to glyoxylate mitochondrial ‘shunt’ entangling at least two differential compartments longly objected by
   Japanese metabolomics study groups.
2. One carbon systems emerge out of a glycoprotein ‘complex’, pyrimidine synthase pathway, that possesses
   significant similarity to  BRCA2 and most other transcription factors suggesting that protein allocation is
   coorchestrated by modifications and spatially transforming construes as an outcome of energy processing.
   Directly deduced by TCS, life cannot exist without mutations, as mutations and chromatin states appear to
   be a sort of energy hold and release ‘gates’.
3. Phosphorylations and small molecules as such as cGMP, cAMP play a role of decompression machinery
   for amplifying bio signal processing C-S, C-N, C-O, interference open systems.  By decompression
   of one relative discrete space, another one becomes compressed, which gets uncertainty of absolute energy
   processing within space scalar wise into vector objected space represented by chromatin remodeling processes,
   possibly seen as network identities information.
4. Unifying network and quantum theory possess implications to relativity concepts and energy relevant computational methodology.