Advertisements
Feeds:
Posts
Comments

Archive for the ‘Leloir pathway’ Category


Summary, Metabolic Pathways

Author: Larry H. Bernstein, MD, FCAP 

 

This portion of a series of chapters on metabolism, proteomics and metabolomics dealt mainly with carbohydrate metabolism. Amino acids and lipids are presented more fully in the chapters that follow. There are features on the

  • functioning of enzymes and proteins,
  • on sequential changes in a chain reaction, and
  • on conformational changes that we shall also cover.

These are critical to developing a more complete understanding of life processes.

I needed to lay out the scope of metabolic reactions and pathways, and their complementary changes. These may not appear to be adaptive, if the circumstances and the duration is not clear. The metabolic pathways map in total
is in interaction with environmental conditions – light, heat, external nutrients and minerals, and toxins – all of which give direction and strength to these reactions. A developing goal is to discover how views introduced by molecular biology and genomics don’t clarify functional cellular dynamics that are not related to the classical view.  The work is vast.

Carbohydrate metabolism denotes the various biochemical processes responsible for the formation, breakdown and interconversion of carbohydrates in living organisms. The most important carbohydrate is glucose, a simple sugar (monosaccharide) that is metabolized by nearly all known organisms. Glucose and other carbohydrates are part of a wide variety of metabolic pathways across species: plants synthesize carbohydrates from carbon dioxide and water by photosynthesis storing the absorbed energy internally, often in the form of starch or lipids. Plant components are consumed by animals and fungi, and used as fuel for cellular respiration. Oxidation of one gram of carbohydrate yields approximately 4 kcal of energy and from lipids about 9 kcal. Energy obtained from metabolism (e.g. oxidation of glucose) is usually stored temporarily within cells in the form of ATP. Organisms capable of aerobic respiration metabolize glucose and oxygen to release energy with carbon dioxide and water as byproducts.

Carbohydrates are used for short-term fuel, and even though they are simpler to metabolize than fats, they don’t produce as equivalent energy yield measured by ATP.  In animals, the concentration of glucose in the blood is linked to the pancreatic endocrine hormone, insulin. . In most organisms, excess carbohydrates are regularly catabolized to form acetyl-CoA, which is a feed stock for the fatty acid synthesis pathway; fatty acids, triglycerides, and other lipids are commonly used for long-term energy storage. The hydrophobic character of lipids makes them a much more compact form of energy storage than hydrophilic carbohydrates.

Glucose is metabolized obtaining ATP and pyruvate by way of first splitting a six-carbon into two three carbon chains, which are converted to lactic acid from pyruvate in the lactic dehydrogenase reaction. The reverse conversion is by a separate unidirectional reaction back to pyruvate after moving through pyruvate dehydrogenase complex.

Pyruvate dehydrogenase complex (PDC) is a complex of three enzymes that convert pyruvate into acetyl-CoA by a process called pyruvate decarboxylation. Acetyl-CoA may then be used in the citric acid cycle to carry out cellular respiration, and this complex links the glycolysis metabolic pathway to the citric acid cycle. This multi-enzyme complex is related structurally and functionally to the oxoglutarate dehydrogenase and branched-chain oxo-acid dehydrogenase multi-enzyme complexes. In eukaryotic cells the reaction occurs inside the mitochondria, after transport of the substrate, pyruvate, from the cytosol. The transport of pyruvate into the mitochondria is via a transport protein and is active, consuming energy. On entry to the mitochondria pyruvate decarboxylation occurs, producing acetyl CoA. This irreversible reaction traps the acetyl CoA within the mitochondria. Pyruvate dehydrogenase deficiency from mutations in any of the enzymes or cofactors results in lactic acidosis.

PDH-rxns The acetyl group is transferred to coenzyme A

PDH-rxns The acetyl group is transferred to coenzyme A

http://guweb2.gonzaga.edu/faculty/cronk/biochem/images/PDH-rxns.gif

Typically, a breakdown of one molecule of glucose by aerobic respiration (i.e. involving both glycolysis and Kreb’s cycle) is about 33-35 ATP. This is categorized as:

Glycogenolysis – the breakdown of glycogen into glucose, which provides a glucose supply for glucose-dependent tissues.

Glycogenolysis in liver provides circulating glucose short term.

Glycogenolysis in muscle is obligatory for muscle contraction.

Pyruvate from glycolysis enters the Krebs cycle, also known as the citric acid cycle, in aerobic organisms.

Anaerobic breakdown by glycolysis – yielding 8-10 ATP

Aerobic respiration by Kreb’s cycle – yielding 25 ATP

The pentose phosphate pathway (shunt) converts hexoses into pentoses and regenerates NADPH. NADPH is an essential antioxidant in cells which prevents oxidative damage and acts as precursor for production of many biomolecules.

Glycogenesis – the conversion of excess glucose into glycogen as a cellular storage mechanism; achieving low osmotic pressure.

Gluconeogenesis – de novo synthesis of glucose molecules from simple organic compounds. An example in humans is the conversion of a few amino acids in cellular protein to glucose.

Metabolic use of glucose is highly important as an energy source for muscle cells and in the brain, and red blood cells.

The hormone insulin is the primary glucose regulatory signal in animals. It mainly promotes glucose uptake by the cells, and it causes the liver to store excess glucose as glycogen. Its absence

  1. turns off glucose uptake,
  2. reverses electrolyte adjustments,
  3. begins glycogen breakdown and glucose release into the circulation by some cells,
  4. begins lipid release from lipid storage cells, etc.

The level of circulatory glucose (known informally as “blood sugar”) is the most important signal to the insulin-producing cells.

  • insulin is made by beta cells in the pancreas,
  • fat is stored n adipose tissue cells, and
  • glycogen is both stored and released as needed by liver cells.
  • no glucose is released to the blood from internal glycogen stores from muscle cells.

The hormone glucagon, on the other hand, opposes that of insulin, forcing the conversion of glycogen in liver cells to glucose, and then release into the blood. Growth hormone, cortisol, and certain catecholamines (such as epinepherine) have glucoregulatory actions similar to glucagon.  These hormones are referred to as stress hormones because they are released under the influence of catabolic proinflammatory (stress) cytokines – interleukin-1 (IL1) and tumor necrosis factor α (TNFα).

Net Yield of GlycolysisThe preparatory phase consumes 2 ATP

The pay-off phase produces 4 ATP.

The gross yield of glycolysis is therefore

4 ATP – 2 ATP = 2 ATP

The pay-off phase also produces 2 molecules of NADH + H+ which can be further converted to a total of 5 molecules of ATP* by the electron transport chain (ETC) during oxidative phosphorylation.

Thus the net yield during glycolysis is 7 molecules of ATP*
This is calculated assuming one NADH molecule gives 2.5 molecules of ATP during oxidative phosphorylation.

Cellular respiration involves 3 stages for the breakdown of glucose – glycolysis, Kreb’s cycle and the electron transport system. Kreb’s cycle produces about 60-70% of ATP for release of energy in the body. It directly or indirectly connects with all the other individual pathways in the body.

The Kreb’s Cycle occurs in two stages:

  1. Conversion of Pyruvate to Acetyl CoA
  2. Acetyl CoA Enters the Kreb’s Cycle

Each pyruvate in the presence of pyruvate dehydrogenase (PDH) complex in the mitochondria gets converted to acetyl CoA which in turn enters the Kreb’s cycle. This reaction is called as oxidative  decarboxylation as the carboxyl group is removed from the pyruvate molecule in the form of CO2 thus yielding 2-carbon acetyl group which along with the coenzyme A forms acetyl CoA.

The PDH requires the sequential action of five co-factors or co-enzymes for the combined action of dehydrogenation and decarboxylation to take place. These five are TPP (thiamine phosphate), FAD (flavin adenine dinucleotide), NAD (nicotinamide adenine dinucleotide), coenzyme A (denoted as CoA-SH at times to depict role of -SH group) and lipoamide.

Acetyl CoA condenses with oxaloacetate (4C) to form a citrate (6C) by transferring its acetyl group in the presence of enzyme citrate synthase. The CoA liberated in this reaction is ready to participate in the oxidative decarboxylation of another molecule of pyruvate by PDH complex.

Isocitrate undergoes oxidative decarboxylation by the enzyme isocitrate dehydrogenase to form oxalosuccinate (intermediate- not shown) which in turn forms α-ketoglutarate (also known as oxoglutarate) which is a five carbon compound. CO2 and NADH are released in this step. α-ketoglutarate (5C) undergoes oxidative decarboxylation once again to form succinyl CoA (4C) catalysed by the enzyme α-ketoglutarate dehydrogenase complex.

Succinyl CoA is then converted to succinate by succinate thiokinase or succinyl coA synthetase in a reversible manner. This reaction involves an intermediate step in which the enzyme gets phosphorylated and then the phosphoryl group which has a high group transfer potential is transferred to GDP to form GTP.

Succinate then gets oxidised reversibly to fumarate by succinate dehydrogenase. The enzyme contains iron-sulfur clusters and covalently bound FAD which when undergoes electron exchange in the mitochondria causes the production of FADH2.

Fumarate is then by the enzyme fumarase converted to malate by hydration(addition of H2O) in a reversible manner.

Malate is then reversibly converted to oxaloacetate by malate dehydrogenase which is NAD linked and thus produces NADH.

The oxaloacetate produced is now ready to be utilized in the next cycle by the citrate synthase reaction and thus the equilibrium of the cycle shifts to the right.

The NADH formed in the cytosol can yield variable amounts of ATP depending on the shuttle system utilized to transport them into the mitochondrial matrix. This NADH, formed in the cytosol, is impermeable to the mitochondrial inner-membrane where oxidative phosphorylation takes place. Thus to carry this NADH to the mitochondrial matrix there are special shuttle systems in the body. The most active shuttle is the malate-aspartate shuttle via which 2.5 molecules of ATP are generated for 1 NADH molecule. This shuttle is mainly used by the heart, liver and kidneys. The brain and skeletal muscles use the other shuttle known as glycerol 3-phosphate shuttle which synthesizes 1.5 molecules of ATP for 1 NADH.

Glucose-6-phosphate Dehydrogenase is the committed step of the Pentose Phosphate Pathway. This enzyme is regulated by availability of the substrate NADP+. As NADPH is utilized in reductive synthetic pathways, the increasing concentration of NADP+ stimulates the Pentose Phosphate Pathway, to replenish NADPH. The importance of this pathway can easily be underestimated.  The main source for energy in respiration was considered to be tied to the high energy phosphate bond in phosphorylation and utilizes NADPH, converting it to NADP+. The pentose phosphate shunt is essential for the generation of nucleic acids, in regeneration of red cells and lens – requiring NADPH.

NAD+ serves as electron acceptor in catabolic pathways in which metabolites are oxidized. The resultant NADH is reoxidized by the respiratory chain, producing ATP.

The pyridine nucleotide transhydrogenase reaction concerns the energy-dependent reduction of TPN by DPNH. In 1959, Klingenberg and Slenczka made the important observation that incubation of isolated liver mitochondria with DPN-specific substrates or succinate in the absence of phosphate acceptor resulted in a rapid and almost complete reduction of  the intramitochondrial TPN. These and related findings led Klingenberg and co-workers (1-3) to postulate the occurrence of a ATP-controlled transhydrogenase reaction catalyzing the reduction of TPN by DPNH.  (The role of transhydrogenase in the energy-linked reduction of TPN.  Fritz Hommes, Ronald W. Estabrook, The Wenner-Gren Institute, University of Stockholm, Stockholm, Sweden. Biochemical and Biophysical Research Communications 11, (1), 2 Apr 1963, Pp 1–6. http://dx.doi.org:/10.1016/0006-291X(63)90017-2/).

Further studies observed the coupling of TPN-specific dehydrogenases with the transhydrogenase and observing the reduction of large amounts of diphosphopyridine nucleotide (DPN) in the presence of catalytic amounts of triphosphopyridine nucleotide (TPN). The studies showed the direct interaction between TPNHz and DPN, in the presence of transhydrogenase to yield products having the properties of TPN and DPNHZ. The reaction involves a transfer of electrons (or hydrogen) rather than a phosphate. (Pyridine Nucleotide Transhydrogenase  II. Direct Evidence for and Mechanism of the Transhydrogenase Reaction* by  Nathan 0. Kaplan, Sidney P. Colowick, And Elizabeth F. Neufeld. (From The Mccollum-Pratt Institute, The Johns Hopkins University, Baltimore, Maryland) J. Biol. Chem. 1952, 195:107-119.) http://www.JBC.org/Content/195/1/107.Citation
Notation: TPN, NADP; DPN, NAD+; reduced pyridine nucleotides: TPNH (NADPH2), DPNH (NADH).

Note: In this discussion there is a detailed presentation of the activity of lactic acid conversion in the mitochondria by way of PDH. In a later section there is mention of the bidirectional reaction of lactate dehydrogenase.  However, the forward reaction is dominant (pyruvate to lactate) and is described. This is not related to the kinetics of the LD reaction with respect to the defining characteristic – Km.

Biochemical Education Jan 1977; 5(1):15. Kinetics of Lactate Dehydrogenase: A Textbook Problem.
K.L. MANCHESTER. Department of Biochemistry, University of Witwatersrand, Johannesburg South Africa.

One presupposes that determined Km values are meaningful under intracellular conditions. In relation to teaching it is a simple experiment for students to determine for themselves the Km towards pyruvate of LDH in a post-mitochondrial supernatant of rat heart and thigh muscle. The difference in Km may be a factor of 3 or 4-fold.It is pertinent then to ask what is the range of suhstrate concentrations over which a difference in Km may be expected to lead to significant differences in activity and how these concentrations compare with pyruvate concentrations in the cell. The evidence of Vesell and co-workers that inhibition by pyruvate is more readily seen at low than at high enzyme concentration is important in emphasizing that under intracellular conditions enzyme concentrations may be relatively large in relation to the substrate available. This will be particularly so in relation to [NADH] which in the cytoplasm is likely to be in the ~M range.

A final point concerns the kinetic parameters for LDH quoted by Bergmeyer for lactate estimations a pH of 9 is recommended and the Km towards lactate at that pH is likely to be appreciably different from the quoted values at pH 7 — Though still at pH 9 showing a substantially lower value for lactate with the heart preparationhttp://onlinelibrary.wiley.com/doi/10.1016/0307-4412%2877%2990013-9/pdf

Several investigators have established that epidermis converts most of the glucose it uses to lactic acid even in the presence of oxygen. This is in contrast to most tissues where lactic acid production is used for energy production only when oxygen is not available. This large amount of lactic acid being continually produced within the epidermal cell must be excreted by the cell and then carried away by the blood stream to other tissues where the lactate can be utilized. The LDH reaction with pyruvate and NADH is reversible although at physiological pH the equilibrium position for the reaction lies very far to the right, i.e., in favor of lactate production. The speed of this reaction depends not only on the amount of enzyme present but also on the concentrations of the substances involved on both sides of the equation. The net direction in which the reaction will proceed depends solely on the relative concentrations of the substances on each side of the equation.
In vivo there is net conversion of pyruvate (formed from glucose) to lactate. Measurements of the speed of lactate production by sheets of epidermis floating on a medium containing glucose indicate a rate of lactate production of approximately 0.7 rn/sm/
mm/mg of fresh epidermis.Slice incubation experiments are presumably much closer to the actual in vivo conditions than
the homogenate experiments. The discrepancy between the
two indicates that in vivo conditions are far from optimal for the conversion of pyruvate to lactate. Only 1/100th of the maximal activity of the enzyme present is being achieved. The concentrations of the various substances involved are not
optimal in vivo since pyruvate and NADH concentrations are
lower than lactate and NAD concentrations and this might explain the in vivo inhibition of LDH activity. (Lactate Production And Lactate Dehydrogenase In The Human Epidermis*. KM. Halprin, A Ohkawara. J Invest Dermat 1966; 47(3): 222-6.)
http://www.nature.com/jid/journal/v47/n3/pdf/jid1966133a.pdf

Advertisements

Read Full Post »


Introduction to Metabolic Pathways

Author: Larry H. Bernstein, MD, FCAP

 

Humans, mammals, plants and animals, and eukaryotes and prokaryotes all share a common denominator in their manner of existence.  It makes no difference whether they inhabit the land, or the sea, or another living host. They exist by virtue of their metabolic adaptation by way of taking in nutrients as fuel, and converting the nutrients to waste in the expenditure of carrying out the functions of motility, breakdown and utilization of fuel, and replication of their functional mass.

There are essentially two major sources of fuel, mainly, carbohydrate and fat.  A third source, amino acids which requires protein breakdown, is utilized to a limited extent as needed from conversion of gluconeogenic amino acids for entry into the carbohydrate pathway. Amino acids follow specific metabolic pathways related to protein synthesis and cell renewal tied to genomic expression.

Carbohydrates are a major fuel utilized by way of either of two pathways.  They are a source of readily available fuel that is accessible either from breakdown of disaccharides or from hepatic glycogenolysis by way of the Cori cycle.  Fat derived energy is a high energy source that is metabolized by one carbon transfers using the oxidation of fatty acids in mitochondria. In the case of fats, the advantage of high energy is conferred by chain length.

Carbohydrate metabolism has either of two routes of utilization.  This introduces an innovation by way of the mitochondrion or its equivalent, for the process of respiration, or aerobic metabolism through the tricarboxylic acid, or Krebs cycle.  In the presence of low oxygen supply, carbohydrate is metabolized anaerobically, the six carbon glucose being split into two three carbon intermediates, which are finally converted from pyruvate to lactate.  In the presence of oxygen, the lactate is channeled back into respiration, or mitochondrial oxidation, referred to as oxidative phosphorylation. The actual mechanism of this process was of considerable debate for some years until it was resolved that the mechanism involve hydrogen transfers along the “electron transport chain” on the inner membrane of the mitochondrion, and it was tied to the formation of ATP from ADP linked to the so called “active acetate” in Acetyl-Coenzyme A, discovered by Fritz Lipmann (and Nathan O. Kaplan) at Massachusetts General Hospital.  Kaplan then joined with Sidney Colowick at the McCollum Pratt Institute at Johns Hopkins, where they shared tn the seminal discovery of the “pyridine nucleotide transhydrogenases” with Elizabeth Neufeld,  who later established her reputation in the mucopolysaccharidoses (MPS) with L-iduronidase and lysosomal storage disease.

This chapter covers primarily the metabolic pathways for glucose, anaerobic and by mitochondrial oxidation, the electron transport chain, fatty acid oxidation, galactose assimilation, and the hexose monophosphate shunt, essential for the generation of NADPH. The is to be more elaboration on lipids and coverage of transcription, involving amino acids and RNA in other chapters.

The subchapters are as follows:

1.1      Carbohydrate Metabolism

1.2      Studies of Respiration Lead to Acetyl CoA

1.3      Pentose Shunt, Electron Transfer, Galactose, more Lipids in brief

1.4      The Multi-step Transfer of Phosphate Bond and Hydrogen Exchange Energy

Complex I or NADH-Q oxidoreductase

Complex I or NADH-Q oxidoreductase

Fatty acid oxidation and ETC

Fatty acid oxidation and ETC

Read Full Post »

Compilation of References in Leaders in Pharmaceutical Intelligence about proteomics, metabolomics, signaling pathways, and cell regulation


Compilation of References in Leaders in Pharmaceutical Intelligence about
proteomics, metabolomics, signaling pathways, and cell regulation

Curator: Larry H. Bernstein, MD, FCAP

 

Proteomics

  1. The Human Proteome Map Completed
    Reporter and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/08/28/the-human-proteome-map-completed/
  1. Proteomics – The Pathway to Understanding and Decision-making in Medicine
    Author and Curator, Larry H Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/06/24/proteomics-the-pathway-to-understanding-and-decision-making-in-medicine/
  1. Advances in Separations Technology for the “OMICs” and Clarification of Therapeutic Targets
    Author and Curator, Larry H Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2012/10/22/advances-in-separations-technology-for-the-omics-and-clarification-of-therapeutic-targets/
  1. Expanding the Genetic Alphabet and Linking the Genome to the Metabolome
    Author and Curator, Larry H Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2012/09/24/expanding-the-genetic-alphabet-and-linking-the-genome-to-the-metabolome/
  1. Synthesizing Synthetic Biology: PLOS Collections
    Reporter: Aviva Lev-Ari
    https://pharmaceuticalintelligence.com/2012/08/17/synthesizing-synthetic-biology-plos-collections/

 

Metabolomics

  1. Extracellular evaluation of intracellular flux in yeast cells
    Larry H. Bernstein, MD, FCAP, Reviewer and Curator
    https://pharmaceuticalintelligence.com/2014/08/25/extracellular-evaluation-of-intracellular-flux-in-yeast-cells/ 
  2. Metabolomic analysis of two leukemia cell lines. I.
    Larry H. Bernstein, MD, FCAP, Reviewer and Curator
    http://pharmaceuticalintelligence.com/2014/08/23/metabolomic-analysis-of-two-leukemia-cell-lines-_i/ 
  3. Metabolomic analysis of two leukemia cell lines. II.
    Larry H. Bernstein, MD, FCAP, Reviewer and Curator
    https://pharmaceuticalintelligence.com/2014/08/24/metabolomic-analysis-of-two-leukemia-cell-lines-ii/ 
  4. Metabolomics, Metabonomics and Functional Nutrition: the next step in nutritional metabolism and biotherapeutics
    Reviewer and Curator, Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/08/22/metabolomics-metabonomics-and-functional-nutrition-the-next-step-in-nutritional-metabolism-and-biotherapeutics/ 
  5. Buffering of genetic modules involved in tricarboxylic acid cycle metabolism provides homeomeostatic regulation
    Larry H. Bernstein, MD, FCAP, Reviewer and curator
    https://pharmaceuticalintelligence.com/2014/08/27/buffering-of-genetic-modules-involved-in-tricarboxylic-acid-cycle-metabolism-provides-homeomeostatic-regulation/

 

Metabolic Pathways

  1. Pentose Shunt, Electron Transfer, Galactose, more Lipids in brief
    Reviewer and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/08/21/pentose-shunt-electron-transfer-galactose-more-lipids-in-brief/
  2. Mitochondria: More than just the “powerhouse of the cell”
    Reviewer and Curator: Ritu Saxena
    https://pharmaceuticalintelligence.com/2012/07/09/mitochondria-more-than-just-the-powerhouse-of-the-cell/
  3. Mitochondrial fission and fusion: potential therapeutic targets?
    Reviewer and Curator: Ritu saxena
    https://pharmaceuticalintelligence.com/2012/10/31/mitochondrial-fission-and-fusion-potential-therapeutic-target/ 
  4. Mitochondrial mutation analysis might be “1-step” away
    Reviewer and Curator: Ritu Saxena
    https://pharmaceuticalintelligence.com/2012/08/14/mitochondrial-mutation-analysis-might-be-1-step-away/
  5. Selected References to Signaling and Metabolic Pathways in PharmaceuticalIntelligence.com
    Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/08/14/selected-references-to-signaling-and-metabolic-pathways-in-leaders-in-pharmaceutical-intelligence/
  6. Metabolic drivers in aggressive brain tumors
    Prabodh Kandal, PhD
    https://pharmaceuticalintelligence.com/2012/11/11/metabolic-drivers-in-aggressive-brain-tumors/ 
  7. Metabolite Identification Combining Genetic and Metabolic Information: Genetic association links unknown metabolites to functionally related genes
    Author and Curator: Aviva Lev-Ari, PhD, RD
    https://pharmaceuticalintelligence.com/2012/10/22/metabolite-identification-combining-genetic-and-metabolic-information-genetic-association-links-unknown-metabolites-to-functionally-related-genes/
  8. Mitochondria: Origin from oxygen free environment, role in aerobic glycolysis, metabolic adaptation
    Author and curator:Larry H Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2012/09/26/mitochondria-origin-from-oxygen-free-environment-role-in-aerobic-glycolysis-metabolic-adaptation/
  9. Therapeutic Targets for Diabetes and Related Metabolic Disorders
    Reporter, Aviva Lev-Ari, PhD, RD
    https://pharmaceuticalintelligence.com/2012/08/20/therapeutic-targets-for-diabetes-and-related-metabolic-disorders/
  10. Buffering of genetic modules involved in tricarboxylic acid cycle metabolism provides homeomeostatic regulation
    Larry H. Bernstein, MD, FCAP, Reviewer and curator
    https://pharmaceuticalintelligence.com/2014/08/27/buffering-of-genetic-modules-involved-in-tricarboxylic-acid-cycle-metabolism-provides-homeomeostatic-regulation/
  11. The multi-step transfer of phosphate bond and hydrogen exchange energy
    Curator:Larry H. Bernstein, MD, FCAP,
    https://pharmaceuticalintelligence.com/2014/08/19/the-multi-step-transfer-of-phosphate-bond-and-hydrogen-exchange-energy/
  12. Studies of Respiration Lead to Acetyl CoA
    Author and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/08/18/studies-of-respiration-lead-to-acetyl-coa/
  13. Lipid Metabolism
    Author and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/08/15/lipid-metabolism/
  14. Carbohydrate Metabolism
    Author and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/08/13/carbohydrate-metabolism/
  15. Prologue to Cancer – e-book Volume One – Where are we in this journey?
    Author and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/04/13/prologue-to-cancer-ebook-4-where-are-we-in-this-journey/
  16. Introduction – The Evolution of Cancer Therapy and Cancer Research: How We Got Here?
    Author and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/04/04/introduction-the-evolution-of-cancer-therapy-and-cancer-research-how-we-got-here/
  17. Inhibition of the Cardiomyocyte-Specific Kinase TNNI3K
    Author and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2013/11/01/inhibition-of-the-cardiomyocyte-specific-kinase-tnni3k/
  18. The Binding of Oligonucleotides in DNA and 3-D Lattice Structures
    Author and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2013/05/15/the-binding-of-oligonucleotides-in-dna-and-3-d-lattice-structures/
  19. Mitochondrial Metabolism and Cardiac Function
    Author and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2013/04/14/mitochondrial-metabolism-and-cardiac-function/
  20. How Methionine Imbalance with Sulfur-Insufficiency Leads to Hyperhomocysteinemia
    Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2013/04/04/sulfur-deficiency-leads_to_hyperhomocysteinemia/
  21. AMPK Is a Negative Regulator of the Warburg Effect and Suppresses Tumor Growth In Vivo
    Author and Curator: SJ. Williams
    https://pharmaceuticalintelligence.com/2013/03/12/ampk-is-a-negative-regulator-of-the-warburg-effect-and-suppresses-tumor-growth-in-vivo/
  22. A Second Look at the Transthyretin Nutrition Inflammatory Conundrum
    Author and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2012/12/03/a-second-look-at-the-transthyretin-nutrition-inflammatory-conundrum/
  23. Overview of Posttranslational Modification (PTM)
    Writer and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/07/29/overview-of-posttranslational-modification-ptm/
  24. Malnutrition in India, high newborn death rate and stunting of children age under five years
    Writer and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/07/15/malnutrition-in-india-high-newborn-death-rate-and-stunting-of-children-age-under-five-years/
  25. Update on mitochondrial function, respiration, and associated disorders
    Writer and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/07/08/update-on-mitochondrial-function-respiration-and-associated-disorders/
  26. Omega-3 fatty acids, depleting the source, and protein insufficiency in renal disease
    Larry H. Bernstein, MD, FCAP, Curator
    https://pharmaceuticalintelligence.com/2014/07/06/omega-3-fatty-acids-depleting-the-source-and-protein-insufficiency-in-renal-disease/ 
  27. Late Onset of Alzheimer’s Disease and One-carbon Metabolism
    Reporter and Curator: Dr. Sudipta Saha, Ph.D.
    https://pharmaceuticalintelligence.com/2013/05/06/alzheimers-disease-and-one-carbon-metabolism/
  28. Problems of vegetarianism
    Reporter and Curator: Dr. Sudipta Saha, Ph.D.
    https://pharmaceuticalintelligence.com/2013/04/22/problems-of-vegetarianism/

 

Signaling Pathways

  1. Introduction to e-Series A: Cardiovascular Diseases, Volume Four Part 2: Regenerative Medicine
    Larry H. Bernstein, MD, FCAP, writer, and Aviva Lev- Ari, PhD, RN  https://pharmaceuticalintelligence.com/2014/04/27/larryhbernintroduction_to_cardiovascular_diseases-translational_medicine-part_2/
  2. Epilogue: Envisioning New Insights in Cancer Translational Biology
    Series C: e-Books on Cancer & Oncology
    Author & Curator: Larry H. Bernstein, MD, FCAP, Series C Content Consultant
    https://pharmaceuticalintelligence.com/2014/03/29/epilogue-envisioning-new-insights/
  3. Ca2+-Stimulated Exocytosis:  The Role of Calmodulin and Protein Kinase C in Ca2+ Regulation of Hormone and Neurotransmitter  Writer and Curator: Larry H Bernstein, MD, FCAP and Curator and Content Editor: Aviva Lev-Ari, PhD, RN
    https://pharmaceuticalintelligence.com/2013/12/23/calmodulin-and-protein-kinase-c-drive-the-ca2-regulation-of-hormone-and-neurotransmitter-release-that-triggers-ca2-stimulated-exocy
  4. Cardiac Contractility & Myocardial Performance: Therapeutic Implications of Ryanopathy (Calcium Release-related Contractile Dysfunction) and Catecholamine Responses
    Author, and Content Consultant to e-SERIES A: Cardiovascular Diseases: Justin Pearlman, MD, PhD, FACC
    Author and Curator: Larry H Bernstein, MD, FCAP and Article Curator: Aviva Lev-Ari, PhD, RN
    https://pharmaceuticalintelligence.com/2013/08/28/cardiac-contractility-myocardium-performance-ventricular-arrhythmias-and-non-ischemic-heart-failure-therapeutic-implications-for-cardiomyocyte-ryanopathy-calcium-release-related-contractile/
  5. Role of Calcium, the Actin Skeleton, and Lipid Structures in Signaling and Cell Motility
    Author and Curator: Larry H Bernstein, MD, FCAP Author: Stephen Williams, PhD, and Curator: Aviva Lev-Ari, PhD, RN
    https://pharmaceuticalintelligence.com/2013/08/26/role-of-calcium-the-actin-skeleton-and-lipid-structures-in-signaling-and-cell-motility/
  6. Identification of Biomarkers that are Related to the Actin Cytoskeleton
    Larry H Bernstein, MD, FCAP, Author and Curator
    https://pharmaceuticalintelligence.com/2012/12/10/identification-of-biomarkers-that-are-related-to-the-actin-cytoskeleton/
  7. Advanced Topics in Sepsis and the Cardiovascular System at its End Stage
    Author and Curator: Larry H Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2013/08/18/advanced-topics-in-Sepsis-and-the-Cardiovascular-System-at-its-End-Stage/
  8. The Delicate Connection: IDO (Indolamine 2, 3 dehydrogenase) and Cancer Immunology
    Demet Sag, PhD, Author and Curator
    https://pharmaceuticalintelligence.com/2013/08/04/the-delicate-connection-ido-indolamine-2-3-dehydrogenase-and-immunology/
  9. IDO for Commitment of a Life Time: The Origins and Mechanisms of IDO, indolamine 2, 3-dioxygenase
    Demet Sag, PhD, Author and Curator
    https://pharmaceuticalintelligence.com/2013/08/04/ido-for-commitment-of-a-life-time-the-origins-and-mechanisms-of-ido-indolamine-2-3-dioxygenase/
  10. Confined Indolamine 2, 3 dioxygenase (IDO) Controls the Homeostasis of Immune Responses for Good and Bad
    Author and Curator: Demet Sag, PhD, CRA, GCP
    https://pharmaceuticalintelligence.com/2013/07/31/confined-indolamine-2-3-dehydrogenase-controls-the-hemostasis-of-immune-responses-for-good-and-bad/
  11. Signaling Pathway that Makes Young Neurons Connect was discovered @ Scripps Research Institute
    Reporter: Aviva Lev-Ari, PhD, RN
    https://pharmaceuticalintelligence.com/2013/06/26/signaling-pathway-that-makes-young-neurons-connect-was-discovered-scripps-research-institute/
  12. Naked Mole Rats Cancer-Free
    Writer and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2013/06/20/naked-mole-rats-cancer-free/
  13. Amyloidosis with Cardiomyopathy
    Writer and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2013/03/31/amyloidosis-with-cardiomyopathy/
  14. Liver endoplasmic reticulum stress and hepatosteatosis
    Larry H Bernstein, MD, FACP
    https://pharmaceuticalintelligence.com/2013/03/10/liver-endoplasmic-reticulum-stress-and-hepatosteatosis/
  15. The Molecular Biology of Renal Disorders: Nitric Oxide – Part III
    Curator and Author: Larry H Bernstein, MD, FACP
    https://pharmaceuticalintelligence.com/2012/11/26/the-molecular-biology-of-renal-disorders/
  16. Nitric Oxide Function in Coagulation – Part II
    Curator and Author: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2012/11/26/nitric-oxide-function-in-coagulation/
  17. Nitric Oxide, Platelets, Endothelium and Hemostasis
    Curator and Author: Larry H Bernstein, MD, FACP
    https://pharmaceuticalintelligence.com/2012/11/08/nitric-oxide-platelets-endothelium-and-hemostasis/
  18. Interaction of Nitric Oxide and Prostacyclin in Vascular Endothelium
    Curator and Author: Larry H Bernstein, MD, FACP
    https://pharmaceuticalintelligence.com/2012/09/14/interaction-of-nitric-oxide-and-prostacyclin-in-vascular-endothelium/
  19. Nitric Oxide and Immune Responses: Part 1
    Curator and Author:  Aviral Vatsa PhD, MBBS
    https://pharmaceuticalintelligence.com/2012/10/18/nitric-oxide-and-immune-responses-part-1/
  20. Nitric Oxide and Immune Responses: Part 2
    Curator and Author:  Aviral Vatsa PhD, MBBS
    https://pharmaceuticalintelligence.com/2012/10/28/nitric-oxide-and-immune-responses-part-2/
  21. Nitric Oxide and iNOS have Key Roles in Kidney Diseases – Part II
    Curator and Author: Larry H Bernstein, MD, FACP
    https://pharmaceuticalintelligence.com/2012/11/26/nitric-oxide-and-inos-have-key-roles-in-kidney-diseases/
  22. New Insights on Nitric Oxide donors – Part IV
    Curator and Author: Larry H Bernstein, MD, FACP
    https://pharmaceuticalintelligence.com/2012/11/26/new-insights-on-no-donors/
  23. Crucial role of Nitric Oxide in Cancer
    Curator and Author: Ritu Saxena, Ph.D.
    https://pharmaceuticalintelligence.com/2012/10/16/crucial-role-of-nitric-oxide-in-cancer/
  24. Nitric Oxide has a ubiquitous role in the regulation of glycolysis -with a concomitant influence on mitochondrial function
    Curator and Author: Larry H Bernstein, MD, FACP
    https://pharmaceuticalintelligence.com/2012/09/16/nitric-oxide-has-a-ubiquitous-role-in-the-regulation-of-glycolysis-with-a-concomitant-influence-on-mitochondrial-function/
  25. Nitric Oxide and Immune Responses: Part 2
    Author and Curator: Aviral Vatsa, PhD, MBBS
    https://pharmaceuticalintelligence.com/2012/10/28/nitric-oxide-and-immune-responses-part-2/
  26. Mitochondrial Damage and Repair under Oxidative Stress
    Author and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2012/10/28/mitochondrial-damage-and-repair-under-oxidative-stress/
  27. Is the Warburg Effect the cause or the effect of cancer: A 21st Century View?
    Curator and Author: Larry H Bernstein, MD, FACP
    https://pharmaceuticalintelligence.com/2012/10/17/is-the-warburg-effect-the-cause-or-the-effect-of-cancer-a-21st-century-view/
  28. Targeting Mitochondrial-bound Hexokinase for Cancer Therapy
    Curator and Author: Ziv Raviv, PhD, RN 04/06/2013
    https://pharmaceuticalintelligence.com/2013/04/06/targeting-mitochondrial-bound-hexokinase-for-cancer-therapy/
  29. Ubiquinin-Proteosome pathway, autophagy, the mitochondrion, proteolysis and cell apoptosis
    Curator and Author: Larry H Bernstein, MD, FACP
    https://pharmaceuticalintelligence.com/2012/10/30/ubiquinin-proteosome-pathway-autophagy-the-mitochondrion-proteolysis-and-cell-apoptosis/
  30. Ubiquitin-Proteosome pathway, Autophagy, the Mitochondrion, Proteolysis and Cell Apoptosis: Part III
    Curator and Author: Larry H Bernstein, MD, FACP
    https://pharmaceuticalintelligence.com/2013/02/14/ubiquinin-proteosome-pathway-autophagy-the-mitochondrion-proteolysis-and-cell-apoptosis-reconsidered/
  31. Biochemistry of the Coagulation Cascade and Platelet Aggregation – Part I
    Curator and Author: Larry H Bernstein, MD, FACP
    https://pharmaceuticalintelligence.com/2012/11/26/biochemistry-of-the-coagulation-cascade-and-platelet-aggregation/

 

Genomics, Transcriptomics, and Epigenetics

  1. What is the meaning of so many RNAs?
    Writer and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/08/06/what-is-the-meaning-of-so-many-rnas/
  2. RNA and the transcription the genetic code
    Larry H. Bernstein, MD, FCAP, Writer and Curator
    https://pharmaceuticalintelligence.com/2014/08/02/rna-and-the-transcription-of-the-genetic-code/
  3. A Primer on DNA and DNA Replication
    Writer and Curator: Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/07/29/a_primer_on_dna_and_dna_replication/
  4. Pathology Emergence in the 21st Century
    Author and Curator: Larry Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/08/03/pathology-emergence-in-the-21st-century/
  5. RNA and the transcription the genetic code
    Writer and Curator, Larry H. Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/08/02/rna-and-the-transcription-of-the-genetic-code/
  6. Commentary on Biomarkers for Genetics and Genomics of Cardiovascular Disease: Views by Larry H Bernstein, MD, FCAP
    Author: Larry H Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/07/16/commentary-on-biomarkers-for-genetics-and-genomics-of-cardiovascular-disease-views-by-larry-h-bernstein-md-fcap/
  7. Observations on Finding the Genetic Links in Common Disease: Whole Genomic Sequencing Studies
    Author an Curator: Larry H Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2013/05/18/observations-on-finding-the-genetic-links/
  8. Silencing Cancers with Synthetic siRNAs
    Larry H. Bernstein, MD, FCAP, Reviewer and Curator
    https://pharmaceuticalintelligence.com/2013/12/09/silencing-cancers-with-synthetic-sirnas/
  9. Cardiometabolic Syndrome and the Genetics of Hypertension: The Neuroendocrine Transcriptome Control Points
    Reporter: Aviva Lev-Ari, PhD, RN
    https://pharmaceuticalintelligence.com/2013/12/12/cardiometabolic-syndrome-and-the-genetics-of-hypertension-the-neuroendocrine-transcriptome-control-points/
  10. Developments in the Genomics and Proteomics of Type 2 Diabetes Mellitus and Treatment Targets
    Larry H. Bernstein, MD, FCAP, Reviewer and Curator
    https://pharmaceuticalintelligence.com/2013/12/08/developments-in-the-genomics-and-proteomics-of-type-2-diabetes-mellitus-and-treatment-targets/
  11. CT Angiography & TrueVision™ Metabolomics (Genomic Phenotyping) for new Therapeutic Targets to Atherosclerosis
    Reporter: Aviva Lev-Ari, PhD, RN
    https://pharmaceuticalintelligence.com/2013/11/15/ct-angiography-truevision-metabolomics-genomic-phenotyping-for-new-therapeutic-targets-to-atherosclerosis/
  12. CRACKING THE CODE OF HUMAN LIFE: The Birth of BioInformatics & Computational Genomics
    Genomics Curator, Larry H Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/08/30/cracking-the-code-of-human-life-the-birth-of-bioinformatics-computational-genomics/
  13. Big Data in Genomic Medicine
    Author and Curator, Larry H Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2012/12/17/big-data-in-genomic-medicine/
  14.  From Genomics of Microorganisms to Translational Medicine
    Author and Curator: Demet Sag, PhD
    https://pharmaceuticalintelligence.com/2014/03/20/without-the-past-no-future-but-learn-and-move-genomics-of-microorganisms-to-translational-medicine/
  15.  Summary of Genomics and Medicine: Role in Cardiovascular Diseases
    Author and Curator, Larry H Bernstein, MD, FCAP
    https://pharmaceuticalintelligence.com/2014/01/06/summary-of-genomics-and-medicine-role-in-cardiovascular-diseases/

Read Full Post »


Extracellular evaluation of intracellular flux in yeast cells

Larry H. Bernstein, MD, FCAP, Reviewer and Curator

Leaders in Pharmaceutical Intelligence

This is the fourth article in a series on metabolomics, which is a major development in -omics, integrating transcriptomics, proteomics,  genomics, metabolic pathways analysis, metabolic and genomic regulatory control using computational mapping.  In the previous two part presentation, flux analysis was not a topic for evaluation, but here it is the major focus.  It is a study of yeast cells, and bears some relationship to the comparison of glycemia, oxidative phosphorylation, TCA cycle, and ETC in leukemia cell lines.  In the previous study – system flux was beyond the scope of analysis, and explicitly stated.  The inferences made in comparing the two lymphocytic leukemia cells was of intracellular metabolism from extracellular measurements.  The study of yeast cells is aimed at looking at cellular effluxes, which is also an important method for studying pharmacological effects and drug resistance.

Metabolomic series

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

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

2.  Metabolomic analysis of two leukemia cell lines. I

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

3.  Metabolomic analysis of two leukemia cell lines. II.

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

4.  Extracellular evaluation of intracellular flux in yeast cells

Q1. What is efflux?

Q2. What measurements were excluded from the previous study that would not allow inference about fluxes?

Q3. Would this study bear any relationship to the Pasteur effect?

Q4 What is a genome scale network reconstruction?

Q5 What type of information is required for a network prediction model?

Q6. Is there a difference between the metabolites profiles for yeast grown under aerobic and anaerobuc conditions – under the constrainsts?

Q7.  If there is a difference in the S metabolism, would there be an effect on ATP production?

 

 

Connecting extracellular metabolomic measurements to intracellular flux
states in yeast

Monica L Mo1Bernhard Ø Palsson1 and Markus J Herrgård12*

Author Affiliations

1 Department of Bioengineering, University of California, San Diego, La Jolla, CA 92093, USA

2 Current address: Synthetic Genomics, Inc, 11149 N Torrey Pines Rd, La Jolla, CA 92037, USA

For all author emails, please log on.

BMC Systems Biology 2009, 3:37  doi:10.1186/1752-0509-3-37

 

The electronic version of this article is the complete one and can be found online at: http://www.biomedcentral.com/1752-0509/3/37

 

Received: 15 December 2008
Accepted: 25 March 2009
Published: 25 March 2009

© 2009 Mo et al; licensee BioMed Central Ltd.

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Background

Metabolomics has emerged as a powerful tool in the

  • quantitative identification of physiological and disease-induced biological states.

Extracellular metabolome or metabolic profiling data, in particular,

  • can provide an insightful view of intracellular physiological states in a noninvasive manner.

Results

We used an updated genome-scale

  • metabolic network model of Saccharomyces cerevisiae, iMM904, to investigate
  1. how changes in the extracellular metabolome can be used
  2. to study systemic changes in intracellular metabolic states.

The iMM904 metabolic network was reconstructed based on

  • an existing genome-scale network, iND750,
  • and includes 904 genes and 1,412 reactions.

The network model was first validated by

  • comparing 2,888 in silico single-gene deletion strain growth phenotype predictions
  • to published experimental data.

Extracellular metabolome data measured

  • of ammonium assimilation pathways 
  • in response to environmental and genetic perturbations

was then integrated with the iMM904 network

  • in the form of relative overflow secretion constraints and
  • a flux sampling approach was used to characterize candidate flux distributions allowed by these constraints.

Predicted intracellular flux changes were

  • consistent with published measurements
  • on intracellular metabolite levels and fluxes.

Patterns of predicted intracellular flux changes

  • could also be used to correctly identify the regions of
  • the metabolic network that were perturbed.

Conclusion

Our results indicate that

  • integrating quantitative extracellular metabolomic profiles
  • in a constraint-based framework
  • enables inferring changes in intracellular metabolic flux states.

Similar methods could potentially be applied

  • towards analyzing biofluid metabolome variations
  • related to human physiological and disease states.

Background

“Omics” technologies are rapidly generating high amounts of data

  • at varying levels of biological detail.

In addition, there is a rapidly growing literature and

  • accompanying databases that compile this information.

This has provided the basis for the assembly of

  • genome-scale metabolic networks for various microbial and eukaryotic organisms [111].

These network reconstructions serve

  • as manually curated knowledge bases of
  • biological information as well as
  • mathematical representations of biochemical components and
  • interactions specific to each organism.

genome-scale network reconstruction is

  • structured collection of genes, proteins, biochemical reactions, and metabolites
  • determined to exist and operate within a particular organism.

This network can be converted into a predictive model

  • that enables in silico simulations of allowable network states based on
  • governing physico-chemical and genetic constraints [12,13].

A wide range of constraint-based methods have been developed and applied

  • to analyze network metabolic capabilities under
  • different environmental and genetic conditions [13].

These methods have been extensively used to

  • study genome-scale metabolic networks and have successfully predicted, for example,
  1. optimal metabolic states,
  2. gene deletion lethality, and
  3. adaptive evolutionary endpoints [1416].

Most of these applications utilize

  • optimization-based methods such as flux balance analysis (FBA)
  • to explore the metabolic flux space.

However, the behavior of genome-scale metabolic networks can also be studied

  • using unbiased approaches such as
  • uniform random sampling of steady-state flux distributions [17].

Instead of identifying a single optimal flux distribution based on

  • a given optimization criterion (e.g. biomass production),

these methods allow statistical analysis of

  • a large range of possible alternative flux solutions determined by
  • constraints imposed on the network.

Sampling methods have been previously used to study

  1. global organization of E. coli metabolism [18] as well as
  2. to identify candidate disease states in the cardiomyocyte mitochondria [19].

Network reconstructions provide a structured framework

  • to systematically integrate and analyze disparate datasets
  • including transcriptomic, proteomic, metabolomic, and fluxomic data.

Metabolomic data is one of the more relevant data types for this type of analysis as

  1. network reconstructions define the biochemical links between metabolites, and
  2. recent advancements in analytical technologies have allowed increasingly comprehensive
  • intracellular and extracellular metabolite level measurements [20,21].

The metabolome is

  1. the set of metabolites present under a given physiological condition
  2. at a particular time and is the culminating phenotype resulting from
  • various “upstream” control mechanisms of metabolic processes.

Of particular interest to this present study are

  • the quantitative profiles of metabolites that are secreted into the extracellular environment
  • by cells under different conditions.

Recent advances in profiling the extracellular metabolome (EM) have allowed

  • obtaining insightful biological information on cellular metabolism
  • without disrupting the cell itself.

This information can be obtained through various

  • analytical detection,
  • identification, and
  • quantization techniques

for a variety of systems ranging from

  • unicellular model organisms to human biofluids [2023].

Metabolite secretion by a cell reflects its internal metabolic state, and

  • its composition varies in response to
  • genetic or experimental perturbations
  • due to changes in intracellular pathway activities
  • involved in the production and utilization of extracellular metabolites [21].

Variations in metabolic fluxes can be reflected in EM changes which can

  • provide insight into the intracellular pathway activities related to metabolite secretion.

The extracellular metabolomic approach has already shown promise

  • in a variety of applications, including
  1. capturing detailed metabolite biomarker variations related to disease and
  2. drug-induced states and
  3. characterizing gene functions in yeast [2427].

However, interpreting changes in the extracellular metabolome can be challenging

  • due to the indirect relationship between the proximal cause of the change
    (e.g. a mutation)
  • and metabolite secretion.

Since metabolic networks describe

  • mechanistic,
  • biochemical links between metabolites,

integrating such data can allow a systematic approach

  • to identifying altered pathways linked to
  • quantitative changes in secretion profiles.

Measured secretion rates of major byproduct metabolites

  • can be applied as additional exchange flux constraints
  • that define observed metabolic behavior.

For example, a recent study integrating small-scale EM data

  • with a genome-scale yeast model
  • correctly predicted oxygen consumption and ethanol production capacities
  • in mutant strains with respiratory deficiencies [28].

The respiratory deficient mutant study

  • used high accuracy measurements for a small number of
  • major byproduct secretion rates
  • together with an optimization-based method well suited for such data.

Here, we expand the application range of the model-based method used in [28]

  • to extracellular metabolome profiles,
  • which represent a temporal snapshot of the relative abundance
  • for a larger number of secreted metabolites.

Our approach is complementary to

  • statistical (i.e. “top-down”) approaches to metabolome analysis [29]
  • and can potentially be used in applications such as biofluid-based diagnostics or
  • large-scale characterization of mutants strains using metabolite profiles.

This study implements a constraint-based sampling approach on

  • an updated genome-scale network of yeast metabolism
  • to systematically determine how EM level variations

are linked to global changes in intracellular metabolic flux states.

By using a sampling-based network approach and statistical methods (Figure 1),

  • EM changes were linked to systemic intracellular flux perturbations
    in an unbiased manner
  • without relying on defining single optimal flux distributions
  • used in the previously mentioned study [28].

The inferred perturbations in intracellular reaction fluxes were further analyzed

  • using reporter metabolite and subsystem (i.e., metabolic pathway) approaches [30]
  • in order to identify dominant metabolic features that are collectively perturbed (Figure 2).

The sampling-based approach also has the additional benefit of

  • being less sensitive to inaccuracies in metabolite secretion profiles than
  • optimization-based methods and can effectively be used – in biofluid metabolome analysis.
integration of exometabolomic (EM) data

integration of exometabolomic (EM) data

Figure 1. Schematic illustrating the integration of exometabolomic (EM) data with the constraint-based framework.

(A) Cells are subjected to genetic and/or environmental perturbations to secrete metabolite patterns unique to that condition.
(B) EM is detected, identified, and quantified.
(C) EM data is integrated as required secretion flux constraints to define allowable solution space.
(D) Random sampling of solution space yields the range of feasible flux distributions for intracellular reactions.
(E) Sampled fluxes were compared to sampled fluxes of another condition to determine

  • which metabolic regions were altered between the two conditions (see Figure 2).

(F) Significantly altered metabolic regions were identified.

http://www.biomedcentral.com/content/figures/1752-0509-3-37-1.jpg

 

sampling and scoring analysis to determine intracellular flux changes

sampling and scoring analysis to determine intracellular flux changes

Figure 2. Schematic of sampling and scoring analysis to determine intracellular flux changes.

(A) Reaction fluxes are sampled for two conditions.
(B & C) Sample of flux differences is calculated by selecting random flux values from each condition

  • to obtain a distribution of flux differences for each reaction.

(D) Standardized reaction Z-scores are determined, which represent

  • how far the sampled flux differences deviates from a zero flux change.

Reaction scores can be used in

  1. visualizing perturbation subnetworks and
  2. analyzing reporter metabolites and subsystems.

http://www.biomedcentral.com/content/figures/1752-0509-3-37-2.jpg

This study was divided into two parts and describes:

(i) the reconstruction and validation of an expanded S. cerevisiae metabolic network, iMM904; and
(ii) the systematic inference of intracellular metabolic states from

  • two yeast EM data sets using a constraint-based sampling approach.

The first EM data set compares wild type yeast to the gdh1/GDH2 (glutamate dehydrogenase) strain [31],

  • which indicated good agreement between predicted metabolic changes
  • of intracellular metabolite levels and fluxes [31,32].

The second EM data set focused on secreted amino acid measurements

  • from a separate study of yeast cultured in different
    ammonium and potassium concentrations [33].

We analyzed the EM data to gain further insight into

  • perturbed ammonium assimilation processes as well as
  1. metabolic states relating potassium limitation and
  2. ammonium excess conditions to one another.

The model-based analysis of both

  • separately published extracellular metabolome datasets
  • suggests a relationship between
  1. glutamate,
  2. threonine and
  3. folate metabolism,
  • which are collectively perturbed when
    ammonium assimilation processes are broadly disrupted
  1. either by environmental (excess ammonia) or
  2. genetic (gene deletion/overexpression) perturbations.

The methods herein present an approach to

  • interpreting extracellular metabolome data and
  • associating these measured secreted metabolite variations
  • to changes in intracellular metabolic network states.

Additional file 1. iMM904 network content.

The data provided represent the content description of the iMM904 metabolic network and
detailed information on the expanded content.

Format: XLS Size: 2.7MB Download file

This file can be viewed with: Microsoft Excel Viewer

Additional file 2. iMM904 model files.

The data provided are the model text files of the iMM904 metabolic network
that is compatible with the available COBRA Toolbox [13]. The model structure
can be loaded into Matlab using the ‘SimPhenyPlus’ format with GPR and compound information.

Format: ZIP Size: 163KB Download file

Conversion of the network to a predictive model

The network reconstruction was converted to a constraint-based model using established procedures [13].

Network reactions and metabolites were assembled into a stoichiometric matrix 

  • containing the stoichiometric coefficients of the reactions in the network.

The steady-state solution space containing possible flux distributions

  • is determined by calculating the null space of S= 0,

where is the reaction flux vector.

Minimal media conditions were set through constraints on exchange fluxes

  • corresponding to the experimental measured substrate uptake rates.

All the model-based calculations were done using the Matlab COBRA Toolbox [13]

  • utilizing the glpk or Tomlab/CPLEX (Tomopt, Inc.) optimization solvers.

Chemostat growth simulations

The iMM904 model was initially validated by

  1. simulating wild type yeast growth in aerobic and anaerobic
    carbon-limited chemostat conditions
  2. and comparing the simulation results to published experimental data

on substrate uptake and byproduct secretion in these conditions [34].

The study was performed following the approach taken to validate the iFF708 model in a previous study [35].

The predicted glucose uptake rates were determined

  1. by setting the in silico growth rate to the measured dilution rate,
    – equivalent under continuous culture growth,
  2. and minimizing the glucose uptake rate.

The accuracy of in silico predictions of

  • substrate uptake and byproduct secretion by the iMM904 model
  • was similar to the accuracy obtained using the iFF708 model
  • and results are shown in Figure S1 [see Additional file 3].

Additional file 3. Supplemental figures. 

The file provides the supplemental figures and descriptions of S1, S2, S3, and S4.

Format: PDF Size: 513KB Download file

This file can be viewed with: Adobe Acrobat Reader

Genome-scale gene deletion phenotype predictions

The iMM904 network was further validated by

  • performing genome-scale gene lethality computations
  • following established procedures to determine growth phenotypes
  1. under minimal medium conditions and
  2. compared to published data.

A modified version of the biomass function used in previous iND750 studies

  1. was set as the objective to be maximized and
  2. gene deletions were simulated by

setting the flux through the corresponding reaction(s) to zero.

The biomass function was based on the experimentally measured

  1. composition of major cellular constituents
  2. during exponential growth of yeast cells and
  3. was reformulated to include trace amounts of
  4. additional cofactors and metabolites
  5. with the assumed fractional contribution of 10-.

These additional biomass compounds were included

according to the biomass formulation used in the iLL672 study

  • to improve lethality predictions through
  • the inclusion of additional essential biomass components [3].

The model was constrained by limiting

  1. the carbon source uptake to 10 mmol/h/gDW
  2. and oxygen uptake to 2 mmol/h/gDW.

Ammonia, phosphate, and sulfate were assumed to be non-limiting.

The experimental phenotyping data was obtained

  • using strains that were auxotrophic for
  1. methionine,
  2. leucine,
  3. histidine, and
  4. uracil.

These auxotrophies were simulated

  1. by deleting the appropriate genes from the model and
  2. supplementing the in silico strain with the appropriate supplements
  3. at non-limiting, but low levels.

Furthermore, trace amounts of essential nutrients that are present

  • in the experimental minimal media formulation
  1. 4-aminobenzoate,
  2. biotin,
  3. inositol,
  4. nicotinate,
  5. panthothenate,
  6. thiamin)
  • were supplied in the simulations [3].

Three distinct methods to simulate the outcome of gene deletions were utilized:

  1. Flux-balance analysis (FBA) [36-38],
  2. Minimization of Metabolic Adjustment (MoMA) [39], and
  3. a linear version of MoMA (linearMoMA).

In the linearMoMA method, minimization of the quadratic objective function
of the original MoMA algorithm

  • was replaced by minimization of the corresponding 1-norm objective function
    (i.e. sum of the absolute values of the differences of wild type FBA solution
    and the knockout strain flux solution).

The computed results were then compared to growth phenotype data
(viable/lethal) from a previously published experimental gene deletion study [3].

The comparison between experimental and in silico deletion phenotypes involved

  • choosing a threshold for the predicted relative growth rate of
  • a deletion strain that is considered to be viable.

We used standard ROC curve analysis

  • to assess the accuracy of different prediction methods and models
  • across the full range of the viability threshold parameter,
    results shown in Figure S2 [see Additional file 3].

The ROC curve plots the true viable rate against the false viable rate

  • allowing comparison of different models and methods
  • without requiring arbitrarily choosing this parameter a priori [40].

The optimal prediction performance corresponds to

  • the point closest to the top left corner of the ROC plot
    (i.e. 100% true viable rate, 0% false viable rate).

Table 1

Table 1 Comparison of iMM904 and iLL672 gene deletion predictions and experimental data under minimal media conditions
Media Model Method True viable False viable False lethal True lethal True viable % False viable % MCC
Glucose iMM904 full FBA 647 10 32 33 95.29 23.26 0.6
iMM904 full linMOMA 644 10 35 33 94.85 23.26 0.58
iMM904 full MOMA 644 10 35 33 94.85 23.26 0.58
iMM904 red FBA 440 9 28 33 94.02 21.43 0.61
iMM904 red linMOMA 437 9 31 33 93.38 21.43 0.6
iMM904 red MOMA 437 9 31 33 93.38 21.43 0.6
iLL672 full MOMA 433 9 35 33 92.52 21.43 0.57
Galactose iMM904 full FBA 595 32 36 59 94.29 35.16 0.58
iMM904 full linMOMA 595 32 36 59 94.29 35.16 0.58
iMM904 full MOMA 595 32 36 59 94.29 35.16 0.58
iMM904 red FBA 409 12 33 56 92.53 17.65 0.67
iMM904 red linMOMA 409 12 33 56 92.53 17.65 0.67
iMM904 red MOMA 409 12 33 56 92.53 17.65 0.67
iLL672 full MOMA 411 19 31 49 92.99 27.94 0.61
Glycerol iMM904 full FBA 596 43 36 47 94.3 47.78 0.48
iMM904 full linMOMA 595 44 37 46 94.15 48.89 0.47
iMM904 full MOMA 598 44 34 46 94.62 48.89 0.48
iMM904 red FBA 410 20 34 46 92.34 30.3 0.57
iMM904 red linMOMA 409 21 35 45 92.12 31.82 0.56
iMM904 red MOMA 412 21 32 45 92.79 31.82 0.57
iLL672 full MOMA 406 20 38 46 91.44 30.3 0.55
Ethanol iMM904 full FBA 593 45 29 55 95.34 45 0.54
iMM904 full linMOMA 592 45 30 55 95.18 45 0.54
iMM904 full MOMA 592 44 30 56 95.18 44 0.55
iMM904 red FBA 408 21 27 54 93.79 28 0.64
iMM904 red linMOMA 407 21 28 54 93.56 28 0.63
iMM904 red MOMA 407 20 28 55 93.56 26.67 0.64
iLL672 full MOMA 401 13 34 62 92.18 17.33 0.68
MCC, Matthews correlation coefficient (see Methods). Note that the iLL672 predictions were obtained directly from [3] and thus the viability threshold was not optimized using the maximum MCC approach.
Mo et al. BMC Systems Biology 2009 3:37  http://dx.doi.org:/10.1186/1752-0509-3-37

 

The values reported in Table 1 correspond to selecting

  • the optimal viability threshold based on this criterion.

We summarized the overall prediction accuracy of a model and method

  • using the Matthews Correlation Coefficient (MCC) [40].

The MCC ranges from -1 (all predictions incorrect) to +1 (all predictions correct) and

  • is suitable for summarizing overall prediction performance

in our case where there are substantially more viable than lethal gene deletions.

ROC plots were produced in Matlab (Mathworks, Inc.).

 

Table 1. Comparison of iMM904 and iLL672

  • gene deletion predictions and
  • experimental data

Inferring perturbed metabolic regions based on EM profiles

The method implemented in this study is shown schematically in Figures 1 and 2

Constraining the iMM904 network 

Relative levels of quantitative EM data were incorporated into the constraint-based framework

  • as overflow secretion exchange fluxes to simulate the required low-level production of
  • experimentally observed excreted metabolites.

The primary objective of this study is to associate

  • relative metabolite levels that are generally measured for metabonomic or biofluid analyses
  • to the quantitative ranges of intracellular reaction fluxes required to produce them.

However, without detailed kinetic information or dynamic metabolite measurements available,

  • we approximated EM datasets of relative quantitative metabolite levels
  • to be proportional to the rate in which they are secreted and detected
  • (at a steady state) – into the extracellular media.

This approach is analogous to approximating uptake rates based

  • on metabolite concentrations from a previous study performing sampling analysis
  • on a cardiomyocyte mitochondrial network
  • to identify differential flux distribution ranges

for various environmental (i.e. substrate uptake) conditions [19].

The raw data was normalized by the raw maximum value of the dataset
(thus the maximum secretion flux was 1 mmol/hr/gDW) with

  • an assumed error of 10%
  • to set the lower and upper bounds and thus
  • inherently accounting for sampling calculation sensitivity.

The gdh1/GDH2 strains were flask cultured under minimal glucose media conditions; thus,

  • glucose and oxygen uptake rates were set at 15 and 2 mmol/hr/gDW, respectively,
  • for the gdh1/GDH2 strain study.

In the anaerobic case the oxygen uptake rate was set to zero, and

  • sterols and fatty acids were provided as in silico supplements as described in [35].

For the potassium limitation/ammonium toxicity study

  • the growth rate was set at 0.17 1/h, and
  • the glucose uptake rate was minimized
  • to mimic experimental chemostat cultivation conditions.

These input constraints were constant for each perturbation and comparative wild-type condition

  • such that the calculated solution spaces between the conditions
  • differed based only on variations in the output secretion constraints.

FBA optimization of EM-constrained networks

A modified FBA method with minimization of the 1-norm objective function

  • between two optimal flux distributions was used
  • to determine optimal intracellular fluxes
  • based on the EM-constrained metabolic models.

This method determines two optimal flux distributions simultaneously

  • for two differently constrained models (e.g. wild type vs. mutant) –
  • these flux distributions maximize biomass production in each case and
  • the 1-norm distance between the distributions is as small as possible
  • given the two sets of constraints.

This approach avoids problems with

  • alternative optimal solutions when comparing two FBA-computed flux distributions
  • by assuming minimal rerouting of flux distibution between a perturbed network and its reference network.

Reaction flux changes from the FBA optimization results were determined

  • by computing the relative percentage fold change for each reaction
  • between the mutant and wild-type flux distributions.

Random sampling of the steady-state solution space

We utilized artificial centering hit-and-run (ACHR) Monte Carlo sampling [19,41]

  • to uniformly sample the metabolic flux solution space
  • defined by the constraints described above.

Reactions, and their participating metabolites, found to participate in intracellular loops [42]

  • were discarded from further analysis as these reactions can have arbitrary flux values.

The following sections describe the approaches used for the analysis of the different datasets.

Sampling approach used in the gdh1/GDH2 study

Due to the overall shape of the metabolic flux solution space,

  • most of the sampled flux distributions resided close to the minimally allowed growth rate
    (i.e. biomass production) and
  • corresponded to various futile cycles that utilized substrates but
  • did not produce significant biomass.

In order to study more physiologically relevant portions of the flux space

  • we restricted the sampling to the part of the solution space
  • where the growth rate was at least 50% of the maximum growth rate
  • for the condition as determined by FBA.

This assumes that cellular growth remains an important overall objective by the yeast cells

  • even in batch cultivation conditions, but
  • that the intracellular flux distributions
  • may not correspond to maximum biomass production [43].

To test the sensitivity of the results to the minimum growth rate threshold,

  • separate Monte Carlo samples were created for each minimum threshold
  • ranging from 50% to 100% at 5% increments.

We also tested the sensitivity of the results

  • to the relative magnitude of the extracellular metabolite secretion rates
  • by performing the sampling at three different relative levels

(0 corresponding to no extracellular metabolite secretion, maximum rate of 0.5 mmol/hr/gDW,
and maximum rate of 1.0 mmol/hr/gDW).

For each minimum growth rate threshold and extracellular metabolite secretion rate,

  • the ACHR sampler was run for 5 million steps and
  • a flux distribution was stored every 5000 steps.

The sensitivity analysis results are presented in Figures S3 and S4 [see Additional File 3], and

  • the results indicate that the reaction Z-scores (see below) are not significantly affected by
  1. either the portion of the solution space sampled or
  2. the exact scaling of secretion rates.

The final overall sample used was created by combining the samples for all minimum growth rate thresholds

  • for the highest extracellular metabolite secretion rate (maximum 1 mmol/hr/gDW).

This approach allowed biasing the sampling towards

  • physiologically relevant parts of the solution space
  • without imposing the requirement of strictly maximizing a predetermined objective function.

The samples obtained with no EM data were used as control samples

  • to filter reporter metabolites/subsystems whose scores were significantly high
  • due to only random differences between sampling runs.

Sampling approach used in the potassium limitation/ammonium toxicity study

Since the experimental data used in this study was generated in chemostat conditions, and

  • previous studies have indicated that chemostat flux patterns predicted by FBA are
  • close to the experimentally measured ones [43],
  • we assumed that sampling of the optimal solution space was appropriate for this study.

In order to sample a physiologically reasonable range of flux distributions,

  • samples for four different oxygen uptake rates
    (1, 2, 3, and 4 mmol/hr/gDW with 5 million steps each)
  • were combined in the final analysis.

Standardized scoring of flux differences between perturbation and control conditions

Z-score based approach was implemented to quantify differences in flux samples between two conditions (Figure 2).
First, two flux vectors were chosen randomly,

  • one from each of the two samples to be compared and
  • the difference between the flux vectors was computed.

This approach was repeated to create a sample of 10,000 (n) flux difference vectors

  • for each pair of conditions considered (e.g. mutant or perturbed environment vs. wild type).

Based on this flux difference sample, the sample mean (μdiff,i) and standard deviation (σdiff,i)

  • between the two conditions was calculated for each reaction i. The reaction Z-score was calculated as:

 

reaction Z-score

reaction Z-score

which describes the sampled mean difference deviation

  • from a population mean change of zero (i.e. no flux difference
    between perturbation and wild type).

Note that this approach allows accounting for uncertainty in the

  • flux distributions inferred based on the extracellular metabolite secretion constraints.

This is in contrast to approaches such as FBA or MoMA that would predict

  • a single flux distribution for each condition and thus potentially
  • overestimate differences between conditions.

The reaction Z-scores can then be further used in analysis

  • to identify significantly perturbed regions of the metabolic network
  • based on reporter metabolite [44] or subsystem [30] Z-scores.

These reporter regions indicate, or “report”, dominant perturbation features

  • at the metabolite and pathway levels for a particular condition.

The reporter metabolite Z-score for any metabolite can be derived from the reaction Z-scores

  • of the reactions consuming or producing j (set of reactions denoted as Rj) as:

 

reporter z-score for any metabolite j

reporter z-score for any metabolite j

where Nis the number of reactions in Rand mmet,is calculated as

 

distributional correction for m_met,j SQRT

distributional correction for m_met,j SQRT

To account and correct for background distribution, the metabolite Z-score was normalized

  • by computing μmet,Nj and σmet,,Nj corresponding to the mean mmet and
  • its standard deviation for 1,000 randomly generated reaction sets of size Nj.

Z-scores for subsystems were calculated similarly by considering the set of reactions R

  • that belongs to each subsystem k.

Hence, positive metabolite and subsystem scores indicate a significantly perturbed metabolic region

  • relative to other regions, whereas
  • a negative score indicate regions that are not perturbed
  • more significantly than what is expected by random chance.

Perturbation subnetworks of reactions and connecting metabolites were visualized using Cytoscape [45].

Results and discussion

  1. Reconstruction and validation of iMM904 network iMM904 network content 

A previously reconstructed S. cerevisiae network, iND750,

  • was used as the basis for the construction of the expanded iMM904 network.
  • Prior to its presentation here, the
    iMM904 network content was the basis for a consensus jamboree network that was recently published
  • but has not yet been adapted for FBA calculations [46].

The majority of iND750 content was carried over and

  • further expanded on to construct iMM904, which accounts for
  1. 904 genes,
  2. 1,228 individual metabolites, and
  3. 1,412 reactions of which
  •                       395 are transport reactions.

Both the number of gene-associated reactions and the number of metabolites

  • increased in iMM904 compared with the iND750 network.

Additional genes and reactions included in the network primarily expanded the

  • lipid,
  • transport, and
  • carbohydrate subsystems.

The lipid subsystem includes

  • new genes and
  • reactions involving the degradation of sphingolipids and glycerolipids.

Sterol metabolism was also expanded to include

  • the formation and degradation of steryl esters, the
  •                      storage form of sterols.

The majority of the new transport reactions were added

  • to connect network gaps between intracellular compartments
  • to enable the completion of known physiological functions.

We also added a number of new secretion pathways

  • based on experimentally observed secreted metabolites [31].

A number of gene-protein-reaction (GPR) relationships were modified

  • to include additional gene products that are required to catalyze a reaction.

For example, the protein compounds

  • thioredoxin and
  • ferricytochrome C

were explicitly represented as compounds in iND750 reactions, but

  • the genes encoding these proteins were not associated with their corresponding GPRs.

Other examples include glycogenin and NADPH cytochrome p450 reductases (CPRs),

  1. which are required in the assembly of glycogen and
  2. to sustain catalytic activity in cytochromes p450, respectively.

These additional proteins were included in iMM904 as

  • part of protein complexes to provide a more complete
  • representation of the genes and
  • their corresponding products necessary for a catalytic activity to occur.

Major modifications to existing reactions were in cofactor biosynthesis, namely in

  • quinone,
  • beta-alanine, and
  • riboflavin biosynthetic pathways.

Reactions from previous S. cerevisiae networks associated with

  • quinone,
  • beta-alanine, and
  • riboflavin biosynthetic pathways

were essentially inferred from known reaction mechanisms based on

  • reactions in previous network reconstructions of E. coli [2,47].

These pathways were manually reviewed

  • based on current literature and subsequently replaced by
  • reactions and metabolites specific to yeast.

Additional changes in other subsystems were also made, such as

  1. changes to the compartmental location of a gene and
  2. its corresponding reaction(s),
  3. changes in reaction reversibility and cofactor specificity, and
  4. the elucidation of particular transport mechanisms.

A comprehensive listing of iMM904 network contents as well as

  • a detailed list of changes between iND750 and iMM904 is included
    [see Additional file 1].

Predicting deletion growth phenotypes

The updated genome-scale iMM904 metabolic network was validated

  • by comparing in silico single-gene deletion predictions to
  • in vivo results from a previous study used
  • to analyze another S. cerevisiae metabolic model, iLL672 [3].

This network was constructed based on the iFF708 network [22],

  • which was also the starting point for
  • reconstructing the iND750 network [2].

The experimental data used to validate the iLL672 model consisted of

3,360 single-gene knockout strain phenotypes evaluated

  • under minimal media growth conditions with
  1. glucose,
  2. galactose,
  3. glycerol, and
  4. ethanol

as sole carbon sources. Growth phenotypes for the iMM904 network were predictedusing

  1. FBA [3234],
  2. MoMA [35], and
  3. linear MoMA methods

as described in Methods and subsequently compared to the experimental data (Table 1).

Each deleted gene growth prediction comparison was classified as

  1. true lethal,
  2. true viable,
  3. false lethal, or
  4. false viable.

The growth rate threshold for considering a prediction viable was chosen

  • for each condition and method separately
  • to optimize the tradeoff between true viable and false viable predictions
    (maximum Matthews correlation coefficient, see Methods).

Since iMM904 has 212 more genes than iLL672 with experimental data, we also present results

  • for the subset of iMM904 predictions with genes included in iLL672 (reduced iMM904 set).

When the same gene sets are compared, iMM904 improves gene lethality predictions under

  • glucose,
  • galactose, and
  • glycerol conditions

over iLL672 somewhat, but is less accurate

  • at predicting growth phenotypes under the ethanol condition.

It should be noted that the iLL672 predictions were obtained directly from [3]

  • thus the growth rate threshold was not optimized similarly to iMM904 predictions.

Overall, when viability cutoff is chosen

  • as indicated above for each method separately,
  • the three prediction methods perform similarly
  1. FBA,
  2. MOMA, and
  3. linear MOMA) .

While the full gene complement in iMM904 greatly increased

  • the number of true viable predictions,
  • the full model also made significantly more false viable predictions
  • compared with reduced iMM904 and iLL672 predictions.

However, it is important to note that 143 reactions involved in dead-end biosynthetic pathways were actually

  • removed from iFF708 to build the iLL672 reconstruction [3].

These dead-ends are considered “knowledge gaps” in pathways

  • that have not been fully characterized and, as a result,
  • lead to false viable predictions when determining gene essentiality
  • if the pathway is in fact required for growth under a certain condition [2,26].

As more of these pathways are elucidated and

  • included in the model to
  • fill in existing network gaps,
  • we can expect false viable prediction rates to consequently decrease.

Thus, while a larger network has a temporarily reduced capacity to accurately predict gene deletion phenotypes,

  • it captures a more complete picture of currently known metabolic functions and
  • provides a framework for network expansion as new pathways are elucidated [48].

 

Inferring intracellular perturbation states from metabolic profiles – Aerobic and anaerobic gdh1/GDH2 mutant behavior

The gdh1/GDH2 mutant strain was previously developed [49,50]

  • to lower NADPH consumption in ammonia assimilation, which would
  • favor the NADPH-dependent fermentation of xylose.

In this strain, the NADPH-dependent glutamate dehydrogenase, Gdh1, was

  • deleted and the NADH-dependent form of the enzyme, Gdh2,
  •                     was overexpressed.

The net effect is to allow efficient assimilation of ammonia

  • into glutamate using NADH instead of NADPH as a cofactor.

While growth characteristics remained unaffected,

  • relative quantities of secreted metabolites differed between the wild-type and mutant strain
  • under aerobic and anaerobic conditions.

We analyzed EM data for the gdh1/GDH2 and wild-type strains reported

  • in [31] under aerobic and anaerobic conditions separately using
  • both FBA optimization and
  • sampling-based approaches as described in Methods.

43 measured extracellular and intracellular metabolites from the original dataset [31],

  • primarily of central carbon and amino acid metabolism,
  • were explicitly represented in the iMM904 network [see Additional file 4].

Extracellular metabolite levels were used

  • to formulate secretion constraints and
  • differential intracellular metabolites were used
  • to compare and validate the intracellular flux predictions.

Perturbed reactions from the FBA results were

  • determined by calculating relative flux changes, and
  • reaction Z-scores were calculated from the sampling analysis
  • to quantify flux changes between the mutant and wild-type strains,
  • with Z reaction > 1.96 corresponding to a two-tailed p-value < 0.05 and
  • considered to be significantly perturbed [see Additional file 4].

Additional file 4. Gdh mutant aerobic and anaerobic analysis results. 

The data provided are the full results for the exometabolomic analysis of aerobic and anerobic gdh1/GDH2 mutant.

Format: XLS Size: 669KB Download file

This file can be viewed with: Microsoft Excel Viewer

To validate the predicted results, reaction flux changes from both FBA and sampling methods were compared to differential intracellular metabolite level data measured from the same study. Intracellular metabolites involved in highly perturbed reactions (i.e. reactants and products) predicted from FBA and sampling analyses were identified and
compared to metabolites that were experimentally identified as significantly changed (< 0.05) between mutant and wild-type. Statistical measures of recall, accuracy, and
precision were calculated and represent the predictive sensitivity, exactness, and reproducibility respectively. From the sampling analysis, a considerably larger number of
significantly perturbed reactions are predicted in the anaerobic case (505 reactions, or 70.7% of active reactions) than in aerobic (394 reactions, or 49.8% of active reactions). The top percentile of FBA flux changes equivalent to the percentage of significantly perturbed sampling reactions were compared to the intracellular data. Results from both analyses are summarized in Table 2. Sampling predictions were considerably higher in recall than FBA predictions for both conditions, with respective ranges of 0.83–1
compared to 0.48–0.96. Accuracy was also higher in sampling predictions; however, precision was slightly better in the FBA predictions as expected due to the smaller
number of predicted changes. Overall, the sampling predictions of perturbed intracellular metabolites are strongly consistent with the experimental data and significantly
outperforms that of FBA optimization predictions in accurately predicting differential metabolites involved in perturbed intracellular fluxes.

Table 2. Statistical comparison of the differential intracellular metabolite data set (< 0.05) with metabolites involved in perturbed reactions predicted by FBA optimization and sampling analyses for aerobic and anaerobic gdh1/GDH2 mutant.

 

Table 2 Statistical comparison of the differential intracellular metabolite data set (p < 0.05)
with metabolites involved in perturbed reactions predicted by FBA optimization and
sampling analyses for aerobic and anaerobic gdh1/GDH2 mutant.
                           Aerobic                         Anaerobic                             Overall
FBA Sampling FBA Sampling FBA
Recall 0.48 0.83 0.96 1 0.71 0.91
Accuracy 0.55 0.62 0.64 0.64 0.6 0.63
Precision 0.78 0.69 0.64 0.63 0.68 0.66
Overall statistics indicate combined results of both conditions.
Mo et al. BMC Systems Biology 2009 3:37   http://dx.doi.org:/10.1186/1752-0509-3-37


Figure 3.
 Perturbation reaction subnetwork of gdh1/GDH2 mutant under aerobic conditions.

The network illustrates a simplified subset of highly perturbedPerturbation subnetworks can be drawn to visualize predicted significantly perturbed intracellular reactions and illustrate their connection to the observed secreted metabolites in the aerobic and anaerobic gdh1/GDH2 mutants.

Perturbation reaction subnetwork of gdh1.GDH2 mutant under aerobic conditions.

Perturbation reaction subnetwork of gdh1.GDH2 mutant under aerobic conditions.

Figure 3 shows an example of a simplified aerobic perturbation subnetwork consisting primarily of proximal pathways connected directly to a subset of major secreted
metabolites

  • glutamate,
  • proline,
  • D-lactate, and
  • 2-hydroxybuturate.

Figure 4 displays anaerobic reactions with Z-scores of similar magnitude to the perturbed reactions in Figure 3. The same subset of metabolites is also present in the
larger anaerobic perturbation network and indicates that the NADPH/NADH balance perturbation induced by the gdh1/GDH2 manipulation has widespread effects
beyond just altering glutamate metabolism anaerobically.

Interestingly, it is clear that the majority of the secreted metabolite pathways involve connected perturbed reactions that broadly converge on glutamate.

Note that Figures 3 and 4 only show the subnetworks that consisted of two or more connected reactions  for a number of secreted metabolites no contiguous perturbed pathway could be identified by the sampling approach. This indicates that the secreted metabolite pattern alone is not sufficient to determine which specific
production and secretion pathways are used by the cell for these metabolites.

Reactions connected to aerobically-secreted metabolites predicted from the sampling analysis of the gdh1/GDH2 mutant strain.
The major secreted metabolites

  • glutamate,
  • proline,
  • D-lactate, and
  • 2-hydroxybuturate

were also detected in the anaerobic condition. Metabolite abbreviations are found in Additional file 1.

Figure 4.

Perturbation reaction subnetwork of gdh1/GDH2 mutant under anaerobic conditions.

Perturbation reaction subnetwork of gdh1.GDH2 mutant under anaerobic conditions

Perturbation reaction subnetwork of gdh1.GDH2 mutant under anaerobic conditions

Subnetwork illustrates the highly perturbed anaerobic reactions of similar Z-reaction magnitude to the reactions in Figure 3.

A significantly larger number of reactions indicates mutant metabolic effects are more widespread in the anaerobic environment.
The network shows that perturbed pathways converge on glutamate, the main site in which the gdh1/GDH2 modification was introduced, which
suggests that the direct genetic perturbation effects are amplified under this environment. Metabolite abbreviations are found in Additional file 1.

To further highlight metabolic regions that have been systemically affected by the gdh1/GDH2 modification, reporter metabolite and subsystem methods [30] were used to
summarize reaction scores around specific metabolites and in specific metabolic subsystems. The top ten significant scores for metabolites/subsystems associated with more
than three reactions are summarized in Tables 3 (aerobic) and 4 (anaerobic), with Z > 1.64 corresponding to < 0.05 for a one-tailed distribution. Full data for all reactions,
reporter metabolites, and reporter subsystems is included [see Additional file 4].

Table 3. List of the top ten significant reporter metabolite and subsystem scores for the gdh1/GDH2 vs. wild type comparison in aerobic conditions.

Table 3
List of the top ten significant reporter metabolite and subsystem scores for the gdh1/GDH2 vs. wild type comparison in aerobic conditions.
Reporter metabolite Z-score No of reactions*
L-proline [c] 2.71 4
Carbon dioxide [m] 2.51 15
Proton [m] 2.19 51
Glyceraldehyde 3-phosphate [c] 1.93 7
Ubiquinone-6 [m] 1.82 5
Ubiquinol-6 [m] 1.82 5
Ribulose-5-phosphate [c] 1.8 4
Uracil [c] 1.74 4
L-homoserine [c] 1.72 4
Alpha-ketoglutarate [m] 1.71 8
Reporter subsystem Z-score No of reactions
Citric Acid Cycle 4.58 7
Pentose Phosphate Pathway 3.29 12
Glycine and Serine Metabolism 2.69 17
Alanine and Aspartate Metabolism 2.65 6
Oxidative Phosphorylation 1.79 8
Thiamine Metabolism 1.54 8
Arginine and Proline Metabolism 1.44 20
Other Amino Acid Metabolism 1.28 5
Glycolysis/Gluconeogenesis 0.58 14
Anaplerotic reactions 0.19 9
*Number of reactions categorized in a subsystem or found to be neighboring each metabolite
Mo et al. BMC Systems Biology 2009 3:37   http://dx.doi.org:/10.1186/1752-0509-3-37

Table 4. List of top ten significant reporter metabolite and subsystem scores for the gdh1/GDH2 vs. wild type comparison in anaerobic conditions.

 

Table 4
List of top ten significant reporter metabolite and subsystem scores for the gdh1/GDH2 vs. wild type comparison in anaerobic conditions.
Reporter metabolite Z-score No of reactions
Glutamate [c] 4.52 35
Aspartate [c] 3.21 11
Alpha-ketoglutarate [c] 2.66 17
Glycine [c] 2.65 7
Pyruvate [m] 2.56 7
Ribulose-5-phosphate [c] 2.43 4
Threonine [c] 2.28 6
10-formyltetrahydrofolate [c] 2.27 5
Fumarate [c] 2.27 5
L-proline [c] 2.04 4
Reporter subsystem Z-score No of reactions
Valine, Leucine, and Isoleucine Metabolism 3.97 15
Tyrosine, Tryptophan, and Phenylalanine Metabolism 3.39 23
Pentose Phosphate Pathway 3.29 11
Purine and Pyrimidine Biosynthesis 3.08 40
Arginine and Proline Metabolism 2.96 19
Threonine and Lysine Metabolism 2.74 14
NAD Biosynthesis 2.66 7
Alanine and Aspartate Metabolism 2.65 6
Histidine Metabolism 2.24 10
Cysteine Metabolism 1.85 10
Mo et al. BMC Systems Biology 2009 3:37   http://dx.doi.org:/10.1186/1752-0509-3-37
Open Data

Perturbations under aerobic conditions largely consisted of pathways involved in mediating the NADH and NADPH balance. Among the highest scoring aerobic subsystems
are TCA cycle and pentose phosphate pathway – key pathways directly involved in the generation of NADH and NADPH. Reporter metabolites involved in these
subsystems –

  • glyceraldehyde-3-phosphate,
  • ribulose-5-phosphate, and
  • alpha-ketoglutarate – were also identified.

These results are consistent with flux and enzyme activity measurements

  • of the gdh1/GDH2 strain under aerobic conditions [32],
  1. which reported significant reduction in the pentose phosphate pathway flux
  2. with concomitant changes in other central metabolic pathways.

Levels of several TCA cycle intermediates (e.g. fumarate, succinate, malate) were also elevated

  • in the gdh1/GDH2 mutant according to the differential intracellular metabolite data.

Altered energy metabolism, as indicated by

  • reporter metabolites (i.e. ubiquinone- , ubiquinol, mitochondrial proton)
  • and subsystem (oxidative phosphorylation),

is certainly feasible as NADH is a primary reducing agent for ATP production.

Pentose phosphate pathway and NAD biosynthesis also appears

  • among the most perturbed anaerobic subsystems, further suggesting
  • perturbed cofactor balance as a common, dominant effect under both conditions.

Glutamate dehydrogenase is a critical enzyme of amino acid biosynthesis as it acts as

  • the entry point for ammonium assimilation via glutamate.

Consequently, metabolic subsystems involved in amino acid biosynthesis were broadly perturbed

  • as a result of the gdh1/GDH2 modification in both aerobic and anaerobic conditions.

For example, the proline biosynthesis pathway that uses glutamate as a precursor

  • was significantly perturbed in both conditions,
  • with significantly changed intracellular and extracellular levels.

There were differences, however, in that more amino acid related subsystems were

  • significantly affected in the anaerobic case (Table 4),
  • further highlighting that altered ammonium assimilation in the mutant
  • has a more widespread effect under anaerobic conditions.

This effect is especially pronounced for

  • threonine and nucleotide metabolism,
  • which were predicted to be significantly perturbed only in anaerobic conditions.

Intracellular threonine levels were amongst the most significantly reduced

  • relative to other differential intracellular metabolites in the anaerobically grown gdh1/GDH2 strain
    (see [31] and Additional file 4), and
  • the relationship between threonine and nucleotide biosynthesis is further supported

by threonine’s recently discovered role as a key precursor in yeast nucleotide biosynthesis [51].

Other key anaerobic reporter metabolites are

  • glycine and 10-formyltetrahydrofolate,
  • both of which are involved in the cytosolic folate cycle (one-carbon metabolism).

Folate is intimately linked to biosynthetic pathways of

  • glycine (with threonine as its precursor) and purines
  • by mediating one-carbon reaction transfers necessary in their metabolism and
  • is a key cofactor in cellular growth [52].

Thus, the anaerobic perturbations identified in the analysis emphasize the close relationship

  • between threonine, folate, and nucleotide metabolic pathways as well as
  • their potential connection to perturbed ammonium assimilation processes.

Interestingly, this association has been previously demonstrated at the transcriptional level

  • as yeast ammonium assimilation (via glutamine synthesis) was found to be
  • co-regulated with genes involved in glycine, folate, and purine synthesis [53].

In summary, the overall differences in predicted gdh1/GDH2 mutant behavior

  • under aerobic and anaerobic conditions show that changes in flux states
  • directly related to modified ammonium assimilation pathway
  1. are amplified anaerobically whereas the
  2. indirect effects through NADH/NADPH balance are more significant aerobically.

Perturbed metabolic regions under aerobic conditions were predominantly

  • in central metabolic pathways involved in responding to the changed NADH/NADPH demand
  • and did not necessarily emphasize that glutamate dehydrogenase was the site of the genetic modification.

The majority of affected anaerobic pathways were involved directly

  • in modified ammonium assimilation as evidenced by

1) significantly perturbed amino acid subsystems,

2) a broad perturbation subnetwork converging on glutamate (Figure 4), and

3) glutamate as the most significant reporter metabolite (Table 4).

Potassium-limited and excess ammonium environments

A recent study reported that potassium limitation resulted in significant

  • growth retardation effect in yeast due to excess ammonium uptake
  • when ammonium was provided as the sole nitrogen source [33].

The proposed mechanism for this effect was that ammonium

  • could to be freely transported through potassium channels
  • when potassium concentrations were low in the media environment, thereby
  • resulting in excess ammonium uptake [33].

As a result, yeast incurred a significant metabolic cost

  • in assimilating ammonia to glutamate and
  • secreting significant amounts of glutamate and other amino acids
  • in potassium-limited conditions as a means to detoxify the excess ammonium.

A similar effect was observed when yeast was grown

  • with no potassium limitation,
  • but with excess ammonia in the environment.

While the observed effect of both environments (low potassium or excess ammonia) was similar,

  • quantitatively unique amino acid secretion profiles suggested that
  • internal metabolic states in these conditions are potentially different.

In order to elucidate the differences in internal metabolic states, we utilized

  • the iMM904 model and the EM profile analysis method to analyze amino acid secretion profiles
  • for a range of low potassium and high ammonia conditions reported in [33].

As before, we utilized amino acid secretion patterns as constraints to the iMM904 model,

  1. sampled the allowable solution space,
  2. computed reaction Z-scores for changes from a reference condition (normal potassium and ammonia), and
  3. finally summarized the resulting changes using reporter metabolites.

Figure 5 shows a clustering of the most significant reporter metabolites (Z ≥ 1.96 in any of the four conditions studied)

  • obtained from this analysis across the four conditions studied.

Interestingly, the potassium-limited environment perturbed only a subset of

  • the significant reporter metabolites identified in the high ammonia environments.

Both low potassium environments shared a consistent pattern of

  • highly perturbed amino acids and related precursor biosynthesis metabolites
    (e.g. pyruvate, PRPP, alpha-ketoglutarate)
  • with high ammonium environments.

The amino acid perturbation pattern (indicated by red labels in Figure 5) was present in

  • the ammonium-toxic environments, although the pattern was
  • slightly weaker for the lower ammonium concentration.

Nevertheless, the results clearly indicate that a similar

  • ammonium detoxifying mechanism that primarily perturbs pathways
  • directly related to amino acid metabolism
  • exists under both types of media conditions.

Figure 5.

Clustergram of top reporter metabolites - y in ammonium-toxic and potassium-limited conditions

Clustergram of top reporter metabolites – y in ammonium-toxic and potassium-limited conditions

Clustergram of top reporter metabolites (i.e. in yellow) in ammonium-toxic and potassium-limited conditions.

Amino acid perturbation patterns (shown in red) were shown to be consistently scored across conditions, indicating that potassium-limited environments K1 (lowest
concentration) and K2 (low concentration) elicited a similar ammonium detoxification response as ammonium-toxic environments N1 (high concentration) and N2
(highest concentration). Metabolites associated with folate metabolism (highlighted in green) are also highly perturbed in ammonium-toxic conditions. Metabolite
abbreviations are found in Additional file 1.

In addition to perturbed amino acids, a secondary effect notably appears at high ammonia levels in which metabolic regions related to folate metabolism are significantly affected. As highlighted in green in Figure 3, we predicted significantly perturbed key metabolites involved in the cytosolic folate cycle. These include tetrahydrofolate derivatives and other metabolites connected to the folate pathway, namely glycine and the methionine-derived methylation cofactors S-adenosylmethionine and S-adenosyl-homocysteine. Additionally, threonine was identified to be a key perturbed metabolite in excess ammonium conditions. These results further illustrate the close
connection between threonine biosynthesis, folate metabolism involving glycine derived from its threonine precursor, and nucleotide biosynthesis [51] that was discussed in
conjunction with the gdh1/GDH2 strain data. Taken together with the anaerobic gdh1/GDH2 data, the results consistently suggest highly perturbed threonine and folate
metabolism when amino acid-related pathways are broadly affected.

In both ammonium-toxic and potassium-limited environments, impaired cellular growth was observed, which can be attributed to high energetic costs of increased
ammonium assimilation to synthesize and excrete amino acids. However, under high ammonium environments, reporter metabolites related to threonine and folate
metabolism indicated that their perturbation, and thus purine supply, may be an additional factor in decreasing cellular viability as there is a direct relationship between
intracellular folate levels and growth rate [54]. Based on these results, we concluded that while potassiumlimited growth in yeast indeed shares physiological features with
growth in ammonium excess, its effects are not as detrimental as actual ammonium excess. The effects on proximal amino acid metabolic pathways are similar in both
environments as indicated by the secretion of the majority of amino acids. However, when our method was applied to analyze the physiological basis behind differences in
secretion profiles between low potassium and high ammonium conditions, ammonium excess was predicted to likely disrupt physiological ammonium assimilation processes,
which in turn potentially impacts folate metabolism and associated cellular growth.

Conclusion

The method presented in this study presents an approach to connecting intracellular flux states to metabolites that are excreted under various physiological conditions. We
showed that well-curated genome-scale metabolic networks can be used to integrate and analyze quantitative EM data by systematically identifying altered intracellular
pathways related to measured changes in the extracellular metabolome. We were able to identify statistically significant metabolic regions that were altered as a result of
genetic (gdh1/GD2 mutant) and environmental (excess ammonium and limited potassium) perturbations, and the predicted intracellular metabolic changes were consistent
with previously published experimental data including measurements of intracellular metabolite levels and metabolic fluxes. Our reanalysis of previously published EM data
on ammonium assimilation-related genetic and environmental perturbations also resulted in testable hypotheses about the role of threonine and folate pathways in mediating
broad responses to changes in ammonium utilization. These studies also demonstrated that the samplingbased method can be readily applied when only partial secreted
metabolite profiles (e.g. only amino acids) are available.

With the emergence of metabolite biofluid biomarkers as a diagnostic tool in human disease [55,56] and the availability of genome-scale human metabolic networks [1],
extensions of the present method would allow identifying potential pathway changes linked to these biomarkers. Employing such a method for studying yeast metabolism was possible as the metabolomic data was measured under controllable environmental conditions where the inputs and outputs of the system were defined. Measured metabolite biomarkers in a clinical setting, however, is far from a controlled environment with significant variations in genetic, nutritional, and environmental factors between different
patients. While there are certainly limitations for clinical applications, the method introduced here is a progressive step towards applying genome-scale metabolic networks
towards analyzing biofluid metabolome data as it 1) avoids the need to only study optimal metabolic states based on a predetermined objective function, 2) allows dealing with noisy experimental data through the sampling approach, and 3) enables analysis even with limited identification of metabolites in the data. The ability to establish potential
connections between extracellular markers and intracellular pathways would be valuable in delineating the genetic and environmental factors associated with a particular
disease.

Authors’ contributions

Conceived and designed the experiments: MLM MJH BOP. Performed experiments: MLM MJH. Analyzed the data: MLM MJH. Wrote the paper: MLM MJH BOP. All authors have read and approved the final manuscript.

Acknowledgements

We thank Jens Nielsen for providing the raw metabolome data for the mutant strain, and Jan Schellenberger and Ines Thiele for valuable discussions. This work was supported by NIH grant R01 GM071808. BOP serves on the scientific advisory board of Genomatica Inc.

 

References

  1. Duarte NC, Becker SA, Jamshidi N, Thiele I, Mo ML, Vo TD, Srivas R, Palsson BO: Global reconstruction of the human metabolic network based on genomic and bibliomic data. 

Proc Natl Acad Sci USA 2007, 104(6):1777-1782. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Duarte NC, Herrgard MJ, Palsson B: Reconstruction and Validation of Saccharomyces cerevisiae iND750, a Fully Compartmentalized Genome-Scale Metabolic Model. 

Genome Res 2004, 14(7):1298-1309. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Kuepfer L, Sauer U, Blank LM: Metabolic functions of duplicate genes in Saccharomyces cerevisiae. 

Genome Res 2005, 15(10):1421-1430. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Nookaew I, Jewett MC, Meechai A, Thammarongtham C, Laoteng K, Cheevadhanarak S, Nielsen J, Bhumiratana S: The genome-scale metabolic model iIN800 of Saccharomyces cerevisiae and its validation: a scaffold to query lipid metabolism. 

BMC Syst Biol 2008, 2:71. PubMed Abstract | BioMed Central Full Text | PubMed Central Full Text

  1. Edwards JS, Palsson BO: Systems properties of the Haemophilus influenzae Rd metabolic genotype. 

J biol chem 1999, 274(25):17410-17416. PubMed Abstract | Publisher Full Text

  1. Edwards JS, Palsson BO: The Escherichia coli MG1655 in silico metabolic genotype: Its definition, characteristics, and capabilities. 

Proc Natl Acad Sci USA 2000, 97(10):5528-5533. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Thiele I, Vo TD, Price ND, Palsson B: An Expanded Metabolic Reconstruction of Helicobacter pylori (IT341 GSM/GPR): An in silico genome-scale characterization of single and double deletion mutants. 

J Bacteriol 2005, 187(16):5818-5830. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Vo TD, Greenberg HJ, Palsson BO: Reconstruction and functional characterization of the human mitochondrial metabolic network based on proteomic and biochemical data. 

J Biol Chem 2004, 279(38):39532-39540. PubMed Abstract | Publisher Full Text

  1. Reed JL, Vo TD, Schilling CH, Palsson BO: An expanded genome-scale model of Escherichia coli K-12 (iJR904 GSM/GPR). 

Genome Biology 2003, 4(9):R54.51-R54.12. BioMed Central Full Text

  1. Feist AM, Henry CS, Reed JL, Krummenacker M, Joyce AR, Karp PD, V H, Palsson BO: A genome-scale metabolic reconstruction for Escherichia coli K-12 MG1655 that accounts for 1261 ORFs and thermodynamic information. 

Molecular Systems Biology 2007, 3:121. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Suthers PF, Dasika MS, Kumar VS, Denisov G, Glass JI, Maranas CD: A genome-scale metabolic reconstruction of Mycoplasma genitalium, iPS189. 

PLos Comp Biol 2009, 5(2):e1000285. Publisher Full Text

  1. Price ND, Reed JL, Palsson BO: Genome-scale models of microbial cells: evaluating the consequences of constraints. 

Nat Rev Microbiol 2004, 2(11):886-897. PubMed Abstract | Publisher Full Text

  1. Becker SA, Feist AM, Mo ML, Hannum G, Palsson BO, Herrgard MJ: Quantitative Prediction of Cellular Metabolism with Constraint-based Models: The COBRA Toolbox. 

Nature protocols 2007, 2(3):727-738. PubMed Abstract | Publisher Full Text

  1. Reed JL, Palsson BO: Genome-Scale In Silico Models of E. coli Have Multiple Equivalent Phenotypic States: Assessment of Correlated Reaction Subsets That Comprise Network States. 

Genome Res 2004, 14(9):1797-1805. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Fong SS, Palsson BO: Metabolic gene deletion strains of Escherichia coli evolve to computationally predicted growth phenotypes. 

Nature Genetics 2004, 36(10):1056-1058. PubMed Abstract | Publisher Full Text

  1. Ibarra RU, Edwards JS, Palsson BO: Escherichia coli K-12 undergoes adaptive evolution to achieve in silico predicted optimal growth. 

Nature 2002, 420(6912):186-189. PubMed Abstract | Publisher Full Text

  1. Schellenberger J, Palsson BØ: Use of randomized sampling for analysis of metabolic networks. 

J Biol Chem 2009, 284(9):5457-5461. PubMed Abstract | Publisher Full Text

  1. Almaas E, Kovács B, Vicsek T, Oltvai ZN, Barabási AL: Global organization of metabolic fluxes in the bacterium Escherichia coli. 

Nature 2004, 427(6977):839-843. PubMed Abstract | Publisher Full Text

  1. Thiele I, Price ND, Vo TD, Palsson BO: Candidate metabolic network states in human mitochondria: Impact of diabetes, ischemia, and diet. 

J Biol Chem 2005, 280(12):11683-11695. PubMed Abstract | Publisher Full Text

  1. Kell DB: Metabolomics and systems biology: making sense of the soup. 

Curr Opin Microbiol 2004, 7(3):296-307. PubMed Abstract | Publisher Full Text

  1. Kell DB, Brown M, Davey HM, Dunn WB, Spasic I, Oliver SG: Metabolic footprinting and systems biology: the medium is the message. 

Nat Rev Microbiol 2005, 3(7):557-565. PubMed Abstract | Publisher Full Text

  1. Goodacre R, Vaidyanathan S, Dunn WB, Harrigan GG, Kell DB: Metabolomics by numbers: acquiring and understanding global metabolite data. 

Trends Biotechnol 2004, 22(5):245-252. PubMed Abstract | Publisher Full Text

  1. Lenz EM, Bright J, Wilson ID, Morgan SR, Nash AF: A 1H NMR-based metabonomic study of urine and plasma samples obtained from healthy human subjects. 

J Pharm Biomed Anal 2003, 33(5):1103-1115. PubMed Abstract | Publisher Full Text

  1. Allen J, Davey HM, Broadhurst D, Heald JK, Rowland JJ, Oliver SG, Kell DB: High-throughput classification of yeast mutants for functional genomics using metabolic footprinting. 

Nat Biotech 2003, 21(6):692-696. Publisher Full Text

  1. Nicholson JK, Connelly J, Lindon JC, Holmes E: Metabonomics: a platform for studying drug toxicity and gene function. 

Nat Rev Drug Discov 2002, 1(2):153-161. PubMed Abstract | Publisher Full Text

  1. Mortishire-Smith RJ, Skiles GL, Lawrence JW, Spence S, Nicholls AW, Johnson BA, Nicholson JK: Use of metabonomics to identify impaired fatty acid metabolism as the mechanism of a drug-induced toxicity. 

Chem Res Toxicol 2004, 17(2):165-173. PubMed Abstract | Publisher Full Text

  1. Sabatine MS, Liu E, Morrow DA, Heller E, McCarroll R, Wiegand R, Berriz GF, Roth FP, Gerszten RE: Metabolomic identification of novel biomarkers of myocardial ischemia. 

Circulation 2005, 112(25):3868-3875. PubMed Abstract | Publisher Full Text

  1. Cakir T, Efe C, Dikicioglu D, Hortaçsu AKB, Oliver SG: Flux balance analysis of a genome-scale yeast model constrained by exometabolomic data allows metabolic system identification of genetically different strains. 

Biotechnol Prog 2007, 23(2):320-326. PubMed Abstract | Publisher Full Text

  1. Bang JW, Crockford DJ, Holmes E, Pazos F, Sternberg MJ, Muggleton SH, Nicholson JK: Integrative top-down system metabolic modeling in experimental disease states via data-driven Bayesian methods. 

J Proteome Res 2008, 7(2):497-503. PubMed Abstract | Publisher Full Text

  1. Oliveira AP, Patil KR, Nielsen J: Architecture of transcriptional regulatory circuits is knitted over the topology of bio-molecular interaction networks. 

BMC Syst Biol 2008, 2:17. PubMed Abstract | BioMed Central Full Text | PubMed Central Full Text

  1. Villas-Boas SG, Moxley JF, Akesson M, Stephanopoulos G, Nielsen J: High-throughput metabolic state analysis: the missing link in integrated functional genomics of yeasts. 

Biochem J 2005, 388(Pt 2):669-677. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Moreira dos Santos M, Thygesen G, Kötter P, Olsson L, Nielsen J: Aerobic physiology of redox-engineered Saccharomyces cerevisiae strains modified in the ammonium assimilation for increased NADPH availability. 

FEMS Yeast Res 2003, 4(1):59-68. PubMed Abstract | Publisher Full Text

  1. Hess DC, Lu W, Rabinowitz JD, Botstein D: Ammonium toxicity and potassium limitation in yeast. 

PLoS Biol 2006, 4(11):e351. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Nissen TL, Schulze U, Nielsen J, Villadsen J: Flux distributions in anaerobic, glucose-limited continuous cultures of Saccharomyces cerevisiae. 

Microbiology 1997, 143(Pt 1):203-218. PubMed Abstract | Publisher Full Text

  1. Famili I, Forster J, Nielsen J, Palsson BO: Saccharomyces cerevisiae phenotypes can be predicted by using constraint-based analysis of a genome-scale reconstructed metabolic network. 

Proc Natl Acad Sci USA 2003, 100(23):13134-13139. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Bonarius HPJ, Schmid G, Tramper J: Flux analysis of underdetermined metabolic networks: The quest for the missing constraints. 

Trends in Biotechnology 1997, 15(8):308-314. Publisher Full Text

  1. Edwards JS, Palsson BO: Metabolic flux balance analysis and the in silico analysis of Escherichia coli K-12 gene deletions. 

BMC Bioinformatics 2000, 1:1. PubMed Abstract | BioMed Central Full Text | PubMed Central Full Text

  1. Varma A, Palsson BO: Metabolic Flux Balancing: Basic concepts, Scientific and Practical Use. 

Nat Biotechnol 1994, 12:994-998. Publisher Full Text

  1. Segre D, Vitkup D, Church GM: Analysis of optimality in natural and perturbed metabolic networks. 

Proc Natl Acad Sci USA 2002, 99(23):15112-15117. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Baldi P, Brunak S, Chauvin Y, Andersen CA, Nielsen H: Assessing the accuracy of prediction algorithms for classification: an overview. 

Bioinformatics 2000, 16(5):412-424. PubMed Abstract | Publisher Full Text

  1. Price ND, Schellenberger J, Palsson BO: Uniform Sampling of Steady State Flux Spaces: Means to Design Experiments and to Interpret Enzymopathies. 

Biophysical Journal 2004, 87(4):2172-2186. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Price ND, Thiele I, Palsson BO: Candidate states of Helicobacter pylori’s genome-scale metabolic network upon application of “loop law” thermodynamic constraints. 

Biophysical J 2006, 90(11):3919-3928. Publisher Full Text

  1. Schuetz R, Kuepfer L, Sauer U: Systematic evaluation of objective functions for predicting intracellular fluxes in Escherichia coli. 

Mol Syst Biol 2007, 3:119. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Patil KR, Nielsen J: Uncovering transcriptional regulation of metabolism by using metabolic network topology. 

Proc Natl Acad Sci USA 2005, 102(8):2685-2689. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T: Cytoscape: a software environment for integrated models of biomolecular interaction networks. 

Genome Res 2003, 13(11):2498-2504. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Herrgard MJ, Swainston N, Dobson P, Dunn WB, Arga KY, Arvas M, Bluthgen N, Borger S, Costenoble R, Heinemann M, et al.: A consensus yeast metabolic network reconstruction obtained from a community approach to systems biology. 

Nat Biotech 2008, 26:1155-1160. Publisher Full Text

  1. Forster J, Famili I, Fu PC, Palsson BO, Nielsen J: Genome-Scale Reconstruction of the Saccharomyces cerevisiae Metabolic Network. 

Genome Research 2003, 13(2):244-253. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Reed JL, Patel TR, Chen KH, Joyce AR, Applebee MK, Herring CD, Bui OT, Knight EM, Fong SS, Palsson BO: Systems Approach to Genome Annotation: Prediction and Validation of Metabolic Functions. 

Proc Natl Acad Sci USA 2006, 103(46):17480-17484. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Nissen TL, Kielland-Brandt MC, Nielsen J, Villadsen J: Optimization of ethanol production in Saccharomyces cerevisiae by metabolic engineering of the ammonium assimilation. 

Metab Eng 2000, 2(1):69-77. PubMed Abstract | Publisher Full Text

  1. Roca C, Nielsen J, Olsson L: Metabolic engineering of ammonium assimilation in xylose-fermenting Saccharomyces cerevisiae improves ethanol production. 

Appl Environ Microbiol 2003, 69(8):4732-4736. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Hartman JL IV: Buffering of deoxyribonucleotide pool homeostasis by threonine metabolism. 

Proc Natl Acad Sci USA 2007, 104(28):11700-11705. PubMed Abstract | Publisher Full Text | PubMed Central Full Text

  1. Gelling CL, Piper MD, Hong SP, Kornfeld GD, Dawes IW: Identification of a novel one-carbon metabolism regulon in Saccharomyces cerevisiae. 

J Biol Chem 2004, 279(8):7072-7081. PubMed Abstract | Publisher Full Text

  1. Denis V, Daignan-Fornier B: Synthesis of glutamine, glycine and 10-formyl tetrahydrofolate is coregulated with purine biosynthesis in Saccharomyces cerevisiae. 

Mol Gen Genet 1998, 259(3):246-255. PubMed Abstract | Publisher Full Text

  1. Hjortmo S, Patring J, Andlid T: Growth rate and medium composition strongly affect folate content in Saccharomyces cerevisiae. 

Int J Food Microbiol 2008, 123(1–2):93-100. PubMed Abstract | Publisher Full Text

  1. Kussmann MRF, Affolter M: OMICS-driven biomarker discovery in nutrition and health. 

J Biotechnol 2006, 124(4):758-787. PubMed Abstract | Publisher Full Text

  1. Serkova NJ, Niemann CU: Pattern recognition and biomarker validation using quantitative 1H-NMR-based metabolomics. 

Expert Rev Mol Diagn 2006, 6(5):717-731. PubMed Abstract | Publisher Full Text

 

Read Full Post »


Metabolomic analysis of two leukemia cell lines. I.

Larry H. Bernstein, MD, FCAP, Reviewer and Curator

Leaders in Pharmaceutical Intelligence

 

I have just posted a review of metabolomics.  In the last few weeks, the Human Metabolome was published.  I am hopeful that my decision has taken the right path to prepare my readers adequately if they will have read the articles that preceded this.  I pondered how I would present this massive piece of work, a study using two leukemia cell lines and mapping the features and differences that drive the carcinogenesis pathways, and identify key metabolic signatures in these differentiated cell types and subtypes.  It is a culmination of a large collaborative effort that required cell culture, enzymatic assays, mass spectrometry, the full measure of which I need not present here, and a very superb validation of the model with a description of method limitations or conflicts.  This is a beautiful piece of work carried out by a small group by today’s standards.

I shall begin this by asking a few questions that will be addressed in the article, which I need to beak up into parts, to draw the readers in more effectively.

Q 1. What metabolic pathways do you expect to have the largest role in the study about to be presented?

Q2. What are the largest metabolic differences that one expects to see in compairing the two lymphoblastic cell lines?

Q3. What methods would be used to extract the information based on external metabolites, enzymes, substrates, etc., to create the model for the cell internal metabolome?

 

 

Abstract

Metabolic models can provide a mechanistic framework to analyze information-rich omics data sets, and are increasingly being used

  • to investigate metabolic alternations in human diseases.

An expression of the altered metabolic pathway utilization is

  • the selection of metabolites consumed and released by cells.

However, methods for the inference of intracellular metabolic states from extracellular measurements in the context of metabolic models

  • remain underdeveloped compared to methods for other omics data.

Herein, we describe a workflow for such an integrative analysis

  • extracting the information from extracellular metabolomics data.

We demonstrate, using the lymphoblastic leukemia cell lines Molt-4 and CCRF-CEM, how

  • our methods can reveal differences in cell metabolism.

Our models explain metabolite uptake and secretion by

  • predicting a more glycolytic phenotype for the CCRF-CEM model and
  • a more oxidative phenotype for the Molt-4 model, which
  • was supported by our experimental data.

Gene expression analysis revealed altered expression of gene products at

  • key regulatory steps in those central metabolic pathways,

and literature query emphasized

  • the role of these genes in cancer metabolism.

Moreover, in silico gene knock-outs identified

  • unique control points for each cell line model, e.g., phosphoglycerate dehydrogenase for the Molt-4 model.

Thus, our workflow is well suited to the characterization of cellular metabolic traits based on

  • extracellular metabolomic data, and
  • it allows the integration of multiple omics data sets into a cohesive picture based on a defined model context.

Keywords Constraint-based modeling _ Metabolomics _Multi-omics _ Metabolic network _ Transcriptomics

 

Reviewer Summary:

  1. A model is introduced to demonstrate a lymphocytic integrated data set using to cell lines.
  2. The method is required to integrate extracted data sets from extracellular metabolites to an intracellular picture of cellular metabolism for each cell line.
  3. The method predicts a more glycolytic or a more oxidative metabolic framework for one or the othe cell line.
  4. The genetic phenotypes differ with a unique control point for each cell line.
  5. The model presents an integration of omics data sets into a cohesive picture based on the model context.

Without having seen the full presentation –

  1. Is the method a snapshot of the neoplastic processes described?
  2. Does the model give insight into the cellular metabolism of an initial cell state for either one or both cell lines?
  3. Would one be able to predict a therapeutic strategy based on the model for either or both cell lines?

Before proceeding further into the study, I would conjecture that there is no way of knowing the initial state ( consistent with what is described by Ilya Prigogine for a self-organizing system) because the model is based on the study of cultured cells that had an unknown metabolic control profile in a host proliferating bone marrow that is likely B-cell origin.  So this is a snapshot of a stable state of two incubated cell lines.  Then the question that is raised is whether there is not only a genetic-phenotypic relationship between the cells in culture and the external metabolites produced, but also whether differences can be discerned between the  internal metabolic constructions that would fit into a family tree.

 

Introduction

Modern high-throughput techniques

  • have increased the pace of biological data generation.

Also referred to as the ‘‘omics avalanche’’, this wealth of data

  • provides great opportunities for metabolic discovery.

Omics data sets contain a snapshot of almost the entire repertoire of

  • mRNA, protein, or metabolites at a given time point or
  • under a particular set of experimental conditions.

Because of the high complexity of the data sets,

  • computational modeling is essential for their integrative analysis.

Currently, such data analysis

  • is a bottleneck in the research process and
  • methods are needed to facilitate the use of these data sets, e.g.,
  1. through meta-analysis of data available in public databases
    [e.g., the human protein atlas (Uhlen et al. 2010)
  2. or the gene expression omnibus (Barrett  et al.  2011)], and
  3. to increase the accessibility of valuable information
    for the biomedical research community.

Constraint-based modeling and analysis (COBRA) is

  • a computational approach that has been successfully used
  • to investigate and engineer microbial metabolism through
    the prediction of steady-states (Durot et al.2009).

The basis of COBRA is network reconstruction: networks are assembled

  1. in a bottom-up fashion based on genomic data and
  2. extensive organism-specific information from the literature.

Metabolic reconstructions

  1. capture information on the known biochemical transformations
    taking place in a target organism
  2. to generate a biochemical, genetic and genomic knowledge base
    (Reed et al. 2006).

Once assembled, a metabolic reconstruction

  • can be converted into a mathematical model
    (Thiele and Palsson 2010), and
  • model properties can be interrogated using a great variety of methods
    (Schellenberger et al. 2011).

The ability of COBRA models to represent

  • genotype–phenotype and environment–phenotype relationships
  • arises through the imposition of constraints,
  • which limit the system to a subset of possible network states
    (Lewis et al. 2012).

Currently, COBRA models exist for more than 100 organisms, including humans
(Duarte et al. 2007; Thiele et al. 2013).

Since the first human metabolic reconstruction was described
[Recon 1 (Duarte et al. 2007)],

  • biomedical applications of COBRA have increased
    (Bordbar and Palsson 2012).

One way to contextualize networks is to

  • define their system boundaries
  • according to the metabolic states of the system,
    e.g., disease or dietary regimes.

The consequences of the applied constraints

  • can then be assessed for the entire network
    (Sahoo and Thiele 2013).

Additionally, omics data sets have frequently been used

  • to generate cell-type or condition-specific metabolic models.

Models exist for specific cell types, such as

  • enterocytes (Sahoo and Thiele2013),
  • macrophages (Bordbar et al. 2010), and
  • adipocytes (Mardinoglu et al. 2013), and
  • even multi-cell assemblies that represent
    the interactions of brain cells (Lewis et al. 2010).

All of these cell type specific models,

  • except the enterocyte reconstruction
  • were generated based on omics data sets.

Cell-type-specific models have been used

  • to study diverse human disease conditions.

For example, an adipocyte model was generated using

  • transcriptomic,
  • proteomic, and
  • metabolomics data.

This model was subsequently used to investigate

  • metabolic alternations in adipocytes
  • that would allow for the stratification of obese patients
    (Mardinoglu et al. 2013).

One highly active field within the biomedical applications of COBRA is

  • cancer metabolism (Jerby and Ruppin, 2012).

Omics-driven large-scale models have been used

  • to predict drug targets (Folger et al. 2011; Jerby et al. 2012).

A cancer model was generated using

  • multiple gene expression data sets and
  • subsequently used to predict synthetic lethal gene pairs
  • as potential drug targets selective for the cancer model,
  • but non-toxic to the global model (Recon 1),
  • a consequence of the reduced redundancy in the
    cancer specific model (Folger et al. 2011).

In a follow up study, lethal synergy between

  • FH and enzymes of the heme metabolic pathway
    were experimentally validated and
  • resolved the mechanism by which FH deficient cells,
    e.g., in renal-cell cancer cells
  • survive a non-functional TCA cycle (Frezza et al. 2011).

Contextualized models, which contain only 

  • the subset of reactions active in 
  • a particular tissue (or cell-) type,
  • can be generated in different ways
    (Becker and Palsson, 2008; Jerby et al. 2010).

However, the existing algorithms mainly consider

  • gene expression and proteomic data to define the reaction sets
  • that comprise the contextualized metabolic models.

These subset of reactions are usually defined based on

  • the expression or absence of expression of the genes or proteins
    (present and absent calls), or
  • inferred from expression values or differential gene expression.

Comprehensive reviews of the methods are available
(Blazier and Papin, 2012; Hyduke et al. 2013).

Only the compilation of a large set of omics data sets

  • can result in a tissue (or cell-type) specific metabolic model, whereas

the representation of one particular experimental condition is achieved through

  • the integration of omics data set generated from one experiment only
    (condition-specific cell line model).

Recently, metabolomic data sets

  • have become more comprehensive and using these data sets allow
  • direct determination of the metabolic network components (the metabolites).

Additionally, metabolomics has proven to be

  1. stable,
  2. relatively inexpensive, and
  3. highly reproducible
    (Antonucci et al. 2012).

These factors make metabolomic data sets

  •  particularly valuable for interrogation of metabolic phenotypes. 

Thus, the integration of these data sets is now an active field of research
(Li et al. 2013; Mo et al. 2009; Paglia et al. 2012b; Schmidt et al. 2013).

Generally, metabolomic data can be incorporated into metabolic networks as

  1. qualitative,
  2. quantitative, and
  3. thermodynamic constraints
    (Fleming et al. 2009; Mo et al. 2009).

Mo et al. used metabolites detected in the spent medium
of yeast cells to determine

  • intracellular flux states through a sampling analysis (Mo et al. 2009),
  • which allowed unbiased interrogation of the possible network states
    (Schellenberger and Palsson 2009)
  • and prediction of internal pathway use.

Such analyses have also been used

  • to reveal the effects of enzymopathies on red blood cells (Price et al. 2004),
  • to study effects of diet on diabetes (Thiele et al. 2005) and
  • to define macrophage metabolic states (Bordbar et al. 2010).

This type of analysis is available as a function in the COBRA toolbox
(Schellenberger et al. 2011).

 

 

 

In this study, we established a workflow for the generation and analysis of

  • condition-specific metabolic cell line models that
  • can facilitate the interpretation of metabolomic data.

Our modeling yields meaningful predictions regarding

  • metabolic differences between two lymphoblastic leukemia cell lines
    (Fig. 1A).
Differences in the use of the TCA cycle by the CCRF-CEM

Differences in the use of the TCA cycle by the CCRF-CEM

 

 

 

http://link.springer.com/static-content/images/404/art%253A10.1007%252
Fs11306-014-0721-3/MediaObjects/11306_2014_721_Fig1_HTML.gif

Fig. 1

A  Combined experimental and computational pipeline to study human metabolism.
Experimental work and omics data analysis steps precede computational modeling. Model

  • predictions are validated based on targeted experimental data.

Metabolomic and transcriptomic data are used for

  • model refinement and submodel extraction.

Functional analysis methods are used to characterize

  • the metabolism of the cell-line models and compare it to additional experimental
    data.

The validated models are subsequently 

  • used for the prediction of drug targets.

B Uptake and secretion pattern of model.
All metabolite uptakes and secretions that were mapped during model
generation are shown.
Metabolite uptakes are depicted on the left, and

  • secreted metabolites are shown on the right.

A number of metabolite exchanges mapped to the model

  • were unique to one cell line.

Differences between cell lines were used to set

  • quantitative constraints for the sampling analysis.

C Statistics about the cell line-specific network generation.

 Quantitative constraints.
For the sampling analysis, an additional

  • set of constraints was imposed on the cell line specific models,
  • emphasizing the differences in metabolite uptake and secretion between cell lines.

Higher uptake of a metabolite was allowed in the model of the cell line

  • that consumed more of the metabolite in vitro, whereas
  • the supply was restricted for the model with lower in vitro uptake.

This was done by establishing the same ratio between the models bounds as detected in vitro.
X denotes the factor(slope ratio) that

  1. distinguishes the bounds, and
  2. which was individual for each metabolite.
  • (a) The uptake of a metabolite could be x times higher in CCRF-CEM cells,
    (b) the metabolite uptake could be x times higher in Molt-4,
    (c) metabolite secretion could be x times higher in CCRF-CEM, or
    (d) metabolite secretion could be x times higher in Molt-4 cells. LOD limit of detection.

The consequence of the adjustment was, in case of uptake, that  one model

  1. was constrained to a lower metabolite uptake (A, B), and the difference
  2. depended on the ratio detected in vitro.

In case of secretion,

  • one model had to secrete more of the metabolite, and again

the difference depended on

  • the experimental difference detected between the cell lines.

Q5. What is your expectation that this type of integrative approach could be used for facilitating medical data interpretations?

The most inventive approach was made years ago by using data constructions from the medical literature by a pioneer in the medical record development, but the technology was  not what it is today, and the cost of data input was high.  Nevertheless, the data acquisition would not be uniform across institutions, except for those that belong to a consolidated network with all of the data in the cloud, and the calculations would be carried out with a separate engine.  However, whether the uniform capture of the massive amount of data needed is not possible in the near foreseeable future.  There is no accurate way of assessing the system cost, and predicting the benefits.  In carrying this model forward there has to be a minimal amount of insufficient data.  The developments in the regulatory sphere have created a high barrier.

This concludes a first portion of this presentation.

 

Read Full Post »

Pentose Shunt, Electron Transfer, Galactose, more Lipids in brief


Pentose Shunt, Electron Transfer, Galactose, more Lipids in brief

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

Pentose Shunt, Electron Transfer, Galactose, and other Lipids in brief

This is a continuation of the series of articles that spans the horizon of the genetic
code and the progression in complexity from genomics to proteomics, which must
be completed before proceeding to metabolomics and multi-omics.  At this point
we have covered genomics, transcriptomics, signaling, and carbohydrate metabolism
with considerable detail.In carbohydrates. There are two topics that need some attention –
(1) pentose phosphate shunt;
(2) H+ transfer
(3) galactose.
(4) more lipids
Then we are to move on to proteins and proteomics.

Summary of this series:

The outline of what I am presenting in series is as follows:

  1. Signaling and Signaling Pathways
    https://pharmaceuticalintelligence.com/2014/08/12/signaling-and-signaling-pathways/
  2. Signaling transduction tutorial.
    https://pharmaceuticalintelligence.com/2014/08/12/signaling-transduction-tutorial/
  3. Carbohydrate metabolism
    https://pharmaceuticalintelligence.com/2014/08/13/carbohydrate-metabolism/

Selected References to Signaling and Metabolic Pathways published in this Open Access Online Scientific Journal, include the following: 

https://pharmaceuticalintelligence.com/2014/08/14/selected-references-to-signaling-
and-metabolic-pathways-in-leaders-in-pharmaceutical-intelligence/

  1. Lipid metabolism

4.1  Studies of respiration lead to Acetyl CoA
https://pharmaceuticalintelligence.com/2014/08/18/studies-of-respiration-lead-to-acetyl-coa/

4.2 The multi-step transfer of phosphate bond and hydrogen exchange energy
https://pharmaceuticalintelligence.com/2014/08/19/the-multi-step-transfer-of-phosphate-
bond-and-hydrogen-exchange-energy/

5.Pentose shunt, electron transfers, galactose, and other lipids in brief

6. Protein synthesis and degradation

7.  Subcellular structure

8. Impairments in pathological states: endocrine disorders; stress
hypermetabolism; cancer.

Section I. Pentose Shunt

Bernard L. Horecker’s Contributions to Elucidating the Pentose Phosphate Pathway

Nicole Kresge,     Robert D. Simoni and     Robert L. Hill

The Enzymatic Conversion of 6-Phosphogluconate to Ribulose-5-Phosphate
and Ribose-5-Phosphate (Horecker, B. L., Smyrniotis, P. Z., and Seegmiller,
J. E.      J. Biol. Chem. 1951; 193: 383–396

Bernard Horecker

Bernard Leonard Horecker (1914) began his training in enzymology in 1936 as a
graduate student at the University of Chicago in the laboratory of T. R. Hogness.
His initial project involved studying succinic dehydrogenase from beef heart using
the Warburg manometric apparatus. However, when Erwin Hass arrived from Otto
Warburg’s laboratory he asked Horecker to join him in the search for an enzyme
that would catalyze the reduction of cytochrome c by reduced NADP. This marked
the beginning of Horecker’s lifelong involvement with the pentose phosphate pathway.

During World War II, Horecker left Chicago and got a job at the National Institutes of
Health (NIH) in Frederick S. Brackett’s laboratory in the Division of Industrial Hygiene.
As part of the wartime effort, Horecker was assigned the task of developing a method
to determine the carbon monoxide hemoglobin content of the blood of Navy pilots
returning from combat missions. When the war ended, Horecker returned to research
in enzymology and began studying the reduction of cytochrome c by the succinic
dehydrogenase system.

Shortly after he began these investigation changes, Horecker was approached by
future Nobel laureate Arthur Kornberg, who was convinced that enzymes were the
key to understanding intracellular biochemical processes
. Kornberg suggested
they collaborate, and the two began to study the effect of cyanide on the succinic
dehydrogenase system. Cyanide had previously been found to inhibit enzymes
containing a heme group, with the exception of cytochrome c. However, Horecker
and Kornberg found that

  • cyanide did in fact react with cytochrome c and concluded that
  • previous groups had failed to perceive this interaction because
    • the shift in the absorption maximum was too small to be detected by
      visual examination.

Two years later, Kornberg invited Horecker and Leon Heppel to join him in setting up
a new Section on Enzymes in the Laboratory of Physiology at the NIH. Their Section on Enzymes eventually became part of the new Experimental Biology and Medicine
Institute and was later renamed the National Institute of Arthritis and Metabolic
Diseases.

Horecker and Kornberg continued to collaborate, this time on

  • the isolation of DPN and TPN.

By 1948 they had amassed a huge supply of the coenzymes and were able to
present Otto Warburg, the discoverer of TPN, with a gift of 25 mg of the enzyme
when he came to visit. Horecker also collaborated with Heppel on 

  • the isolation of cytochrome c reductase from yeast and 
  • eventually accomplished the first isolation of the flavoprotein from
    mammalian liver.

Along with his lab technician Pauline Smyrniotis, Horecker began to study

  • the enzymes involved in the oxidation of 6-phosphogluconate and the
    metabolic intermediates formed in the pentose phosphate pathway.

Joined by Horecker’s first postdoctoral student, J. E. Seegmiller, they worked
out a new method for the preparation of glucose 6-phosphate and 6-phosphogluconate, 
both of which were not yet commercially available.
As reported in the Journal of Biological Chemistry (JBC) Classic reprinted here, they

  • purified 6-phosphogluconate dehydrogenase from brewer’s yeast (1), and 
  • by coupling the reduction of TPN to its reoxidation by pyruvate in
    the presence of lactic dehydrogenase
    ,
  • they were able to show that the first product of 6-phosphogluconate oxidation,
  • in addition to carbon dioxide, was ribulose 5-phosphte.
  • This pentose ester was then converted to ribose 5-phosphate by a
    pentose-phosphate isomerase.

They were able to separate ribulose 5-phosphate from ribose 5- phosphate and demonstrate their interconversion using a recently developed nucleotide separation
technique called ion-exchange chromatography. Horecker and Seegmiller later
showed that 6-phosphogluconate metabolism by enzymes from mammalian
tissues also produced the same products
.8

Bernard Horecker

Bernard Horecker

http://www.jbc.org/content/280/29/e26/F1.small.gif

Over the next several years, Horecker played a key role in elucidating the

  • remaining steps of the pentose phosphate pathway.

His total contributions included the discovery of three new sugar phosphate esters,
ribulose 5-phosphate, sedoheptulose 7-phosphate, and erythrose 4-phosphate, and
three new enzymes, transketolase, transaldolase, and pentose-phosphate 3-epimerase.
The outline of the complete pentose phosphate cycle was published in 1955
(2). Horecker’s personal account of his work on the pentose phosphate pathway can
be found in his JBC Reflection (3).1

Horecker’s contributions to science were recognized with many awards and honors
including the Washington Academy of Sciences Award for Scientific Achievement in
Biological Sciences (1954) and his election to the National Academy of Sciences in
1961. Horecker also served as president of the American Society of Biological
Chemists (now the American Society for Biochemistry and Molecular Biology) in 1968.

Footnotes

  • 1 All biographical information on Bernard L. Horecker was taken from Ref. 3.
  • The American Society for Biochemistry and Molecular Biology, Inc.

References

  1. ↵Horecker, B. L., and Smyrniotis, P. Z. (1951) Phosphogluconic acid dehydrogenase
    from yeast. J. Biol. Chem. 193, 371–381FREE Full Text
  2. Gunsalus, I. C., Horecker, B. L., and Wood, W. A. (1955) Pathways of carbohydrate
    metabolism in microorganisms. Bacteriol. Rev. 19, 79–128  FREE Full Text
  3. Horecker, B. L. (2002) The pentose phosphate pathway. J. Biol. Chem. 277, 47965–
    47971 FREE Full Text

The Pentose Phosphate Pathway (also called Phosphogluconate Pathway, or Hexose
Monophosphate Shunt) is depicted with structures of intermediates in Fig. 23-25
p. 863 of Biochemistry, by Voet & Voet, 3rd Edition. The linear portion of the pathway
carries out oxidation and decarboxylation of glucose-6-phosphate, producing the
5-C sugar ribulose-5-phosphate.

Glucose-6-phosphate Dehydrogenase catalyzes oxidation of the aldehyde
(hemiacetal), at C1 of glucose-6-phosphate, to a carboxylic acid in ester linkage
(lactone). NADPserves as electron acceptor.

6-Phosphogluconolactonase catalyzes hydrolysis of the ester linkage (lactone)
resulting in ring opening. The product is 6-phosphogluconate. Although ring opening
occurs in the absence of a catalyst, 6-Phosphogluconolactonase speeds up the
reaction, decreasing the lifetime of the highly reactive, and thus potentially
toxic, 6-phosphogluconolactone.

Phosphogluconate Dehydrogenase catalyzes oxidative decarboxylation of
6-phosphogluconate, to yield the 5-C ketose ribulose-5-phosphate. The
hydroxyl at C(C2 of the product) is oxidized to a ketone. This promotes loss
of the carboxyl at C1 as CO2.  NADP+ again serves as oxidant (electron acceptor).

pglucose hd

pglucose hd

https://www.rpi.edu/dept/bcbp/molbiochem/MBWeb/mb2/part1/images/pglucd.gif

Reduction of NADP+ (as with NAD+) involves transfer of 2e- plus 1H+ to the
nicotinamide moiety.

nadp

NADPH, a product of the Pentose Phosphate Pathway, functions as a reductant in
various synthetic (anabolic) pathways, including fatty acid synthesis.

NAD+ serves as electron acceptor in catabolic pathways in which metabolites are
oxidized. The resultant NADH is reoxidized by the respiratory chain, producing ATP.

nadnadp

https://www.rpi.edu/dept/bcbp/molbiochem/MBWeb/mb2/part1/images/nadnadp.gif

Regulation: 
Glucose-6-phosphate Dehydrogenase is the committed step of the Pentose
Phosphate Pathway. This enzyme is regulated by availability of the substrate NADP+.
As NADPH is utilized in reductive synthetic pathways, the increasing concentration of
NADP+ stimulates the Pentose Phosphate Pathway, to replenish NADPH.

The remainder of the Pentose Phosphate Pathway accomplishes conversion of the
5-C ribulose-5-phosphate to the 5-C product ribose-5-phosphate, or to the 3-C
glyceraldehyde -3-phosphate and the 6-C fructose-6-phosphate (reactions 4 to 8
p. 863).

Transketolase utilizes as prosthetic group thiamine pyrophosphate (TPP), a
derivative of vitamin B1.

tpp

tpp

https://www.rpi.edu/dept/bcbp/molbiochem/MBWeb/mb2/part1/images/tpp.gif

Thiamine pyrophosphate binds at the active sites of enzymes in a “V” conformation.The amino group of the aminopyrimidine moiety is close to the dissociable proton,
and serves as the proton acceptor. This proton transfer is promoted by a glutamate
residue adjacent to the pyrimidine ring.

The positively charged N in the thiazole ring acts as an electron sink, promoting
C-C bond cleavage. The 3-C aldose glyceraldehyde-3-phosphate is released.
2-C fragment remains on TPP.

FASEB J. 1996 Mar;10(4):461-70.   http://www.ncbi.nlm.nih.gov/pubmed/8647345

Reviewer

The importance of this pathway can easily be underestimated.  The main source for
energy in respiration was considered to be tied to the

  • high energy phosphate bond in phosphorylation and utilizes NADPH, converting it to NADP+.

glycolysis n skeletal muscle in short term, dependent on muscle glycogen conversion
to glucose, and there is a buildup of lactic acid – used as fuel by the heart.  This
pathway accounts for roughly 5% of metabolic needs, varying between tissues,
depending on there priority for synthetic functions, such as endocrine or nucleic
acid production.

The mature erythrocyte and the ocular lens both are enucleate.  85% of their
metabolic energy needs are by anaerobic glycolysis.  Consider the erythrocyte
somewhat different than the lens because it has iron-based hemoglobin, which
exchanges O2 and CO2 in the pulmonary alveoli, and in that role, is a rapid
regulator of H+ and pH in the circulation (carbonic anhydrase reaction), and also to
a lesser extent in the kidney cortex, where H+ is removed  from the circulation to
the urine, making the blood less acidic, except when there is a reciprocal loss of K+.
This is how we need a nomogram to determine respiratory vs renal acidosis or
alkalosis.  In the case of chronic renal disease, there is substantial loss of
functioning nephrons, loss of countercurrent multiplier, and a reduced capacity to
remove H+.  So there is both a metabolic acidosis and a hyperkalemia, with increased
serum creatinine, but the creatinine is only from muscle mass – not accurately
reflecting total body mass, which includes visceral organs.  The only accurate
measure of lean body mass would be in the linear relationship between circulating
hepatic produced transthyretin (TTR).

The pentose phosphate shunt is essential for

  • the generation of nucleic acids, in regeneration of red cells and lens – requiring NADPH.

Insofar as the red blood cell is engaged in O2 exchange, the lactic dehydrogenase
isoenzyme composition is the same as the heart. What about the lens of and cornea the eye, and platelets?  The explanation does appear to be more complex than
has been proposed and is not discussed here.

Section II. Mitochondrial NADH – NADP+ Transhydrogenase Reaction

There is also another consideration for the balance of di- and tri- phospopyridine
nucleotides in their oxidized and reduced forms.  I have brought this into the
discussion because of the centrality of hydride tranfer to mitochondrial oxidative
phosphorylation and the energetics – for catabolism and synthesis.

The role of transhydrogenase in the energy-linked reduction of TPN 

Fritz HommesRonald W. Estabrook∗∗

The Wenner-Gren Institute, University of Stockholm
Stockholm, Sweden
Biochemical and Biophysical Research Communications 11, (1), 2 Apr 1963, Pp 1–6
http://dx.doi.org:/10.1016/0006-291X(63)90017-2

In 1959, Klingenberg and Slenczka (1) made the important observation that incubation of isolated

  • liver mitochondria with DPN-specific substrates or succinate in the absence of phosphate
    acceptor resulted in a rapid and almost complete reduction of the intramitochondrial TPN.

These and related findings led Klingenberg and co-workers (1-3) to postulate

  • the occurrence of an ATP-controlled transhydrogenase reaction catalyzing the reduction of
    mitochondrial TPN by DPNH. A similar conclusion was reached by Estabrook and Nissley (4).

The present paper describes the demonstration and some properties of an

  • energy-dependent reduction of TPN by DPNH, catalyzed by submitochondrial particles.

Preliminary reports of some of these results have already appeared (5, 6 ) , and a
complete account is being published elsewhere (7).We have studied the energy- dependent reduction of TPN by PNH with submitochondrial particles from both
rat liver and beef heart. Rat liver particles were prepared essentially according to
the method of Kielley and Bronk (8), and beef heart particles by the method of
Low and Vallin (9).

PYRIDINE NUCLEOTIDE TRANSHYDROGENASE  II. DIRECT EVIDENCE FOR
AND MECHANISM OF THE
 TRANSHYDROGENASE REACTION*

BY  NATHAN 0. KAPLAN, SIDNEY P. COLOWICK, AND ELIZABETH F. NEUFELD
(From the McCollum-Pratt Institute, The Johns Hopkins University, Baltimore,
Maryland)  J. Biol. Chem. 1952, 195:107-119.
http://www.jbc.org/content/195/1/107.citation

NO Kaplan

NO Kaplan

Sidney Colowick

Sidney Colowick

Elizabeth Neufeld

Elizabeth Neufeld

Kaplan studied carbohydrate metabolism in the liver under David M. Greenberg at the
University of California, Berkeley medical school. He earned his Ph.D. in 1943. From
1942 to 1944, Kaplan participated in the Manhattan Project. From 1945 to 1949,
Kaplan worked with Fritz Lipmann at Massachusetts General Hospital to study
coenzyme A. He worked at the McCollum-Pratt Institute of Johns Hopkins University
from 1950 to 957. In 1957, he was recruited to head a new graduate program in
biochemistry at Brandeis University. In 1968, Kaplan moved to the University of
California, San Diego
, where he studied the role of lactate dehydrogenase in cancer. He also founded a colony of nude mice, a strain of laboratory mice useful in the study
of cancer and other diseases. [1] He was a member of the National Academy of
Sciences.One of Kaplan’s students at the University of California was genomic
researcher Craig Venter.[2]3]  He was, with Sidney Colowick, a founding editor of the scientific book series Methods
in Enzymology
.[1]

http://books.nap.edu/books/0309049768/xhtml/images/img00009.jpg

Colowick became Carl Cori’s first graduate student and earned his Ph.D. at
Washington University St. Louis in 1942, continuing to work with the Coris (Nobel
Prize jointly) for 10 years. At the age of 21, he published his first paper on the
classical studies of glucose 1-phosphate (2), and a year later he was the sole author on a paper on the synthesis of mannose 1-phosphate and galactose 1-phosphate (3). Both papers were published in the JBC. During his time in the Cori lab,

Colowick was involved in many projects. Along with Herman Kalckar he discovered
myokinase (distinguished from adenylate kinase from liver), which is now known as
adenyl kinase. This discovery proved to be important in understanding transphos-phorylation reactions in yeast and animal cells. Colowick’s interest then turned to
the conversion of glucose to polysaccharides, and he and Earl Sutherland (who
will be featured in an upcoming JBC Classic) published an important paper on the
formation of glycogen from glucose using purified enzymes (4). In 1951, Colowick
and Nathan Kaplan were approached by Kurt Jacoby of Academic Press to do a
series comparable to Methodem der Ferment Forschung. Colowick and Kaplan
planned and edited the first 6 volumes of Methods in Enzymology, launching in 1955
what became a series of well known and useful handbooks. He continued as
Editor of the series until his death in 1985.

The Structure of NADH: the Work of Sidney P. Colowick

Nicole KresgeRobert D. Simoni and Robert L. Hill

On the Structure of Reduced Diphosphopyridine Nucleotide

(Pullman, M. E., San Pietro, A., and Colowick, S. P. (1954)

J. Biol. Chem. 206, 129–141)

Elizabeth Neufeld
·  Born: September 27, 1928 (age 85), Paris, France
·  EducationQueens College, City University of New YorkUniversity of California,
Berkeley

http://fdb5.ctrl.ucla.edu/biological-chemistry/institution/photo?personnel%5fid=45290&max_width=155&max_height=225

In Paper I (l), indirect evidence was presented for the following transhydrogenase
reaction, catalyzed by an enzyme present in extracts of Pseudomonas
fluorescens:

TPNHz + DPN -+ TPN + DPNHz

The evidence was obtained by coupling TPN-specific dehydrogenases with the
transhydrogenase and observing the reduction of large amounts of diphosphopyridine nucleotide (DPN) in the presence of catalytic amounts of triphosphopyridine
nucleotide (TPN).

In this paper, data will be reported showing the direct

  • interaction between TPNHz and DPN, in thepresence of transhydrogenase alone,
  • to yield products having the propertiesof TPN and DPNHZ.

Information will be given indicating that the reaction involves

  • a transfer of electrons (or hydrogen) rather than a phosphate 

Experiments dealing with the kinetics and reversibility of the reaction, and with the
nature of the products, suggest that the reaction is a complex one, not fully described
by the above formulation.

Materials and Methods [edited]

The TPN and DPN used in these studies were preparations of approximately 75
percent purity and were prepared from sheep liver by the chromatographic procedure
of Kornberg and Horecker (unpublished). Reduced DPN was prepared enzymatically with alcohol dehydrogenase as described elsewhere (2). Reduced TPN was prepared by treating TPN with hydrosulfite. This treated mixture contained 2 pM of TPNHz per ml.
The preparations of desamino DPN and reduced desamino DPN have been
described previously (2, 3). Phosphogluconate was a barium salt which was kindly
supplied by Dr. B. F. Horecker. Cytochrome c was obtained from the Sigma Chemical Company.

Transhydrogenase preparations with an activity of 250 to 7000 units per mg. were
used in these studies. The DPNase was a purified enzyme, which was obtained
from zinc-deficient Neurospora and had an activity of 5500 units per mg. (4). The
alcohol dehydrogenase was a crystalline preparation isolated from yeast according to the procedure of Racker (5).

Phosphogluconate dehydrogenase from yeast and a 10 per cent pure preparation of the TPN-specific cytochrome c reductase from liver (6) were gifts of Dr. B. F.
Horecker.

DPN was assayed with alcohol and crystalline yeast alcohol dehydrogenase. TPN was determined By the specific phosphogluconic acid dehydrogenase from yeast and also by the specific isocitric dehydrogenase from pig heart. Reduced DPN was
determined by the use of acetaldehyde and the yeast alcohol dehydrogenase.
All of the above assays were based on the measurement of optical density changes
at 340 rnp. TPNHz was determined with the TPN-specific cytochrome c reductase system. The assay of the reaction followed increase in optical density at 550 rnp  as a measure of the reduction of the cytochrome c after cytochrome c
reductase was added to initiate the reaction. The changes at 550 rnp are plotted for different concentrations of TPNHz in Fig. 3, a. The method is an extremely sensitive and accurate assay for reduced TPN.

Results
[No Figures or Table shown]

Formation of DPNHz from TPNHz and DPN-Fig. 1, a illustrates the direct reaction between TPNHz and DPN to form DPNHZ. The reaction was carried out by incubating TPNHz with DPN in the presence of the
transhydrogenase, yeast alcohol dehydrogenase, and acetaldehyde. Since the yeast dehydrogenase is specific for DPN,

  • a decrease in absorption at340 rnp can only be due to the formation of reduced DPN. It can
    be seen from the curves in Fig. 1, a that a decrease in optical density occurs only in the
    presence of the complete system.

The Pseudomonas enzyme is essential for the formation of DPNH2. It is noteworthy
that, under the conditions of reaction in Fig. 1, a,

  • approximately 40 per cent of theTPNH, reacted with the DPN.

Fig. 1, a also indicates that magnesium is not required for transhydrogenase activity.  The reaction between TPNHz and DPN takes place in the absence of alcohol
dehydrogenase and acetaldehyde
. This can be demonstrated by incubating the
two pyridine nucleotides with the transhydrogenase for 4 8 12 16 20 24 28 32 36
minutes

FIG. 1. Evidence for enzymatic reaction of TPNHt with DPN.

  • Rate offormation of DPNH2.

(b) DPN disappearance and TPN formation.

(c) Identification of desamino DPNHz as product of reaction of TPNHz with desamino DPN.  (assaying for reduced DPN by the yeast alcohol dehydrogenase technique.

Table I (Experiment 1) summarizes the results of such experiments in which TPNHz was added with varying amounts of DPN.

  • In the absence of DPN, no DPNHz was formed. This eliminates the possibility that TPNH 2 is
    converted to DPNHz
  • by removal ofthe monoester phosphate grouping.

The data also show that the extent of the reaction is

  • dependent on the concentration of DPN.

Even with a large excess of DPN, only approximately 40 per cent of the TPNHzreacts to form reduced DPN. It is of importance to emphasize that in the above
experiments, which were carried out in phosphate buffer, the extent of  the reaction

  • is the same in the presence or absence of acetaldehyde andalcohol dehydrogenase.

With an excess of DPN and different  levels of TPNHZ,

  • the amount of reduced DPN which is formed is
  • dependent on the concentration of TPNHz(Table I, Experiment 2).
  • In all cases, the amount of DPNHz formed is approximately
    40 per cent of the added reduced TPN.

Formation of TPN-The reaction between TPNHz and DPN should yield TPN as well as DPNHz.
The formation of TPN is demonstrated in Table 1. in Fig. 1, b. In this experiment,
TPNHz was allowed to react with DPN in the presence of the transhydrogenase
(PS.), and then alcohol and alcohol dehydrogenase were added . This
would result in reduction of the residual DPN, and the sample incubated with the
transhydrogenase contained less DPN. After the completion of the alcohol
dehydrogenase reaction, phosphogluconate and phosphogluconic dehydrogenase (PGAD) were added to reduce the TPN. The addition of this TPN-specific
dehydrogenase results in an

  • increase inoptical density in the enzymatically treated sample.
  • This change represents the amount of TPN formed.

It is of interest to point out that, after addition of both dehydrogenases,

  • the total optical density change is the same in both

Therefore it is evident that

  • for every mole of DPN disappearing  a mole of TPN appears.

Balance of All Components of Reaction

Table II (Experiment 1) shows that,

  • if measurements for all components of the reaction are made, one can demonstrate
    that there is
  • a mole for mole disappearance of TPNH, and DPN, and
  • a stoichiometric appearance of TPN and DPNH2.
  1. The oxidized forms of the nucleotides were assayed as described
  2. the reduced form of TPN was determined by the TPNHz-specific cytochrome c reductase,
  3. the DPNHz by means of yeast alcohol dehydrogenase plus

This stoichiometric balance is true, however,

  • only when the analyses for the oxidized forms are determined directly on the reaction

When analyses are made after acidification of the incubated reaction mixture,

  • the values found forDPN and TPN are much lower than those obtained by direct analysis.

This discrepancy in the balance when analyses for the oxidized nucleotides are
carried out in acid is indicated in Table II (Experiment 2). The results, when
compared with the findings in Experiment 1, are quite striking.

Reaction of TPNHz with Desamino DPN

Desamino DPN

  • reacts with the transhydrogenase system at the same rate as does DPN (2).

This was of value in establishing the fact that

  • the transhydrogenase catalyzesa transfer of hydrogen rather than a phosphate transfer reaction.

The reaction between desamino DPN and TPNHz can be written in two ways.

TPN f desamino DPNHz

TPNH, + desamino DPN

DPNH2 + desamino TPN

If the reaction involved an electron transfer,

  • desamino DPNHz would be
  • Phosphate transfer would result in the production of reduced

Desamino DPNHz can be distinguished from DPNHz by its

  • slowerrate of reaction with yeast alcohol dehydrogenase (2, 3).

Fig. 1, c illustrates that, when desamino DPN reacts with TPNH2, 

  • the product of the reaction is desamino DPNHZ.

This is indicated by the slow rate of oxidation of the product by yeast alcohol
dehydrogenase and acetaldehyde.

From the above evidence phosphate transfer 

  • has been ruled out as a possible mechanism for the transhydrogenase reaction.

Inhibition by TPN

As mentioned in Paper I and as will be discussed later in this paper,

  • the transhydrogenase reaction does not appear to be readily reversible.

This is surprising, particularly since only approximately 

  • 40 per cent of the TPNHz undergoes reaction with DPN
    under the conditions described above. It was therefore thought that
  • the TPN formed might inhibit further transfer of electrons from TPNH2.

Table III summarizes data showing the

  • strong inhibitory effect of TPN on thereaction between TPNHz and DPN.

It is evident from the data that

  • TPN concentration is a factor in determining the extent of the reaction.

Effect of Removal of TPN on Extent of Reaction

A purified DPNase from Neurospora has been found

  • to cleave the nicotinamide riboside linkagesof the oxidized forms of both TPN and DPN
  • without acting on thereduced forms of both nucleotides (4).

It has been found, however, that

  • the DPNase hydrolyzes desamino DPN at a very slow rate (3).

In the reaction between TPNHz and desamino DPN, TPN and desamino DPNH:,

  • TPNis the only component of this reaction attacked by the Neurospora enzyme
    at an appreciable rate

It was  thought that addition of the DPNase to the TPNHZ-desamino DPN trans-
hydrogenase reaction mixture

  • would split the TPN formed andpermit the reaction to go to completion.

This, indeed, proved to be the case, as indicated in Table IV, where addition of
the DPNase with desamino DPN results in almost

  • a stoichiometric formation of desamino DPNHz
  • and a complete disappearance of TPNH2.

Extent of Reaction in Buffers Other Than Phosphate

All the reactions described above were carried out in phosphate buffer of pH 7.5.
If the transhydrogenase reaction between TPNHz and DPN is run at the same pH
in tris(hydroxymethyl)aminomethane buffer (TRIS buffer)

  • with acetaldehydeand alcohol dehydrogenase present,
  • the reaction proceeds muchfurther toward completion 
  • than is the case under the same conditions ina phosphate medium (Fig. 2, a).

The importance of phosphate concentration in governing the extent of the reaction
is illustrated in Fig. 2, b.

In the presence of TRIS the transfer reaction

  • seems to go further toward completion in the presence of acetaldehyde
    and 
    alcohol dehydrogenase
  • than when these two components are absent.

This is not true of the reaction in phosphate,

  • in which the extent is independent of the alcoholdehydrogenase system.

Removal of one of the products of the reaction (DPNHp) in TRIS thus

  • appears to permit the reaction to approach completion,whereas
  • in phosphate this removal is without effect on the finalcourse of the reaction.

The extent of the reaction in TRIS in the absence of alcohol dehydrogenase
and acetaldehyde
 is

  • somewhat greater than when the reaction is run in phosphate.

TPN also inhibits the reaction of TPNHz with DPN in TRIS medium, but the inhibition

  • is not as marked as when the reaction is carried out in phosphate buffer.

Reversibility of Transhydrogenase Reaction;

Reaction between DPNHz and TPN

In Paper I, it was mentioned that no reversal of the reaction could be achieved in a system containing alcohol, alcohol dehydrogenase, TPN, and catalytic amounts of
DPN.

When DPNH, and TPN are incubated with the purified transhydrogenase, there is
also

  • no evidence for reversibility.

This is indicated in Table V which shows that

  • there is no disappearance of DPNHz in such a system.

It was thought that removal of the TPNHz, which might be formed in the reaction,
could promote the reversal of the reaction. Hence,

  • by using the TPNHe-specific cytochrome c reductase, one could
  1. not only accomplishthe removal of any reduced TPN,
  2. but also follow the course of the reaction.

A system containing DPNH2, TPN, the transhydrogenase, the cytochrome c
reductase, and cytochrome c, however, gives

  • no reduction of the cytochrome

This is true for either TRIS or phosphate buffers.2

Some positive evidence for the reversibility has been obtained by using a system
containing

  • DPNH2, TPNH2, cytochrome c, and the cytochrome creductase in TRIS buffer.

In this case, there is, of course, reduction of cytochrome c by TPNHZ, but,

  • when the transhydrogenase is present.,there is
  • additional reduction over and above that due to the added TPNH2.

This additional reduction suggests that some reversibility of the reaction occurred
under these conditions. Fig. 3, b shows

  • the necessity of DPNHzfor this additional reduction.

Interaction of DPNHz with Desamino DPN-

If desamino DPN and DPNHz are incubated with the purified Pseudomonas enzyme,
there appears

  • to be a transfer of electrons to form desamino DPNHz.

This is illustrated in Fig. 4, a, which shows the

  • decreased rate of oxidation by thealcohol dehydrogenase system
  • after incubation with the transhydrogenase.
  • Incubation of desamino DPNHz with DPN results in the formation of DPNH2,
  • which is detected by the faster rate of oxidation by the alcohol dehydrogenase system
  • after reaction of the pyridine nucleotides with thetranshydrogenase (Fig. 4, b).

It is evident from the above experiments that

the transhydrogenase catalyzes an exchange of hydrogens between

  • the adenylic and inosinic pyridine nucleotides.

However, it is difficult to obtain any quantitative information on the rate or extent of
the reaction by the method used, because

  • desamino DPNHz also reacts with the alcohol dehydrogenase system,
  • although at a much slower rate than does DPNH2.

DISCUSSION

The results of the balance experiments seem to offer convincing evidence that
the transhydrogenase catalyzes the following reaction.

TPNHz + DPN -+ DPNHz + TPN

Since desamino DPNHz is formed from TPNHz and desamino DPN,

  • thereaction appears to involve an electron (or hydrogen) transfer
  • rather thana transfer of the monoester phosphate grouping of TPN.

A number of the findings reported in this paper are not readily understandable in
terms of the above simple formulation of the reaction. It is difficult to understand
the greater extent of the reaction in TRIS than in phosphate when acetaldehyde
and alcohol dehydrogenase are present.

One possibility is that an intermediate may be involved which is more easily converted
to reduced DPN in the TRIS medium. The existence of such an intermediate is also
suggested by the discrepancies noted in balance experiments, in which

  • analyses of the oxidized nucleotides after acidification showed
  • much lower values than those found by direct analysis.

These findings suggest that the reaction may involve

  • a 1 electron ratherthan a 2 electron transfer with
  • the formation of acid-labile free radicals as intermediates.

The transfer of hydrogens from DPNHz to desamino DPN

  • to yield desamino DPNHz and DPN and the reversal of this transfer
  • indicate the unique role of the transhydrogenase
  • in promoting electron exchange between the pyridine nucleotides.

In this connection, it is of interest that alcohol dehydrogenase and lactic
dehydrogenase cannot duplicate this exchange  between the DPN and
the desamino systems.3  If one assumes that desamino DPN behaves
like DPN,

  • one might predict that the transhydrogenase would catalyze an
    exchange of electrons (or hydrogen) 3.

Since alcohol dehydrogenase alone

  • does not catalyze an exchange of electrons between the adenylic
    and inosinic pyridine nucleotides, this rules out the possibility
  • that the dehydrogenase is converted to a reduced intermediate
  • during electron between DPNHz and added DPN.

It is hoped to investigate this possibility with isotopically labeled DPN.
Experiments to test the interaction between TPN and desamino TPN are
also now in progress.

It seems likely that the transhydrogenase will prove capable of

  • catalyzingan exchange between TPN and TPNH2, as well as between DPN and

The observed inhibition by TPN of the reaction between TPNHz and DPN may
therefore

  • be due to a competition between DPN and TPNfor the TPNH2.

SUMMARY

  1. Direct evidence for the following transhydrogenase reaction. catalyzedby an
    enzyme from Pseudomonas fluorescens, is presented.

TPNHz + DPN -+ TPN + DPNHz

Balance experiments have shown that for every mole of TPNHz disappearing
1 mole of TPN appears and that for each mole of DPNHz generated 1 mole of
DPN disappears. The oxidized nucleotides found at the end of the reaction,
however, show anomalous lability toward acid.

  1. The transhydrogenase also promotes the following reaction.

TPNHz + desamino DPN -+ TPN + desamino DPNH,

This rules out the possibility that the transhydrogenase reaction involves a
phosphate transfer and indicates that the

  • enzyme catalyzes a shift of electrons (or hydrogen atoms).

The reaction of TPNHz with DPN in 0.1 M phosphate buffer is strongly
inhibited by TPN; thus

  • it proceeds only to the extent of about40 per cent or less, even
  • when DPNHz is removed continuously by meansof acetaldehyde
    and alcohol dehydrogenase.
  • In other buffers, in whichTPN is less inhibitory, the reaction proceeds
    much further toward completion under these conditions.
  • The reaction in phosphate buffer proceedsto completion when TPN
    is removed as it is formed.
  1. DPNHz does not react with TPN to form TPNHz and DPN in the presence
    of transhydrogenase. Some evidence, however, has been obtained for
    the reversibility by using the following system:
  • DPNHZ, TPNHZ, cytochromec, the TPNHz-specific cytochrome c reductase,
    and the transhydrogenase.
  1. Evidence is cited for the following reversible reaction, which is catalyzed
    by the transhydrogenase.

DPNHz + desamino DPN fi DPN + desamino DPNHz

  1. The results are discussed with respect to the possibility that the
    transhydrogenase reaction may
  • involve a 1 electron transfer with theformation of free radicals as intermediates.

 

BIBLIOGRAPHY

  1. Coiowick, S. P., Kaplan, N. O., Neufeld, E. F., and Ciotti, M. M., J. Biol. Chem.,196, 95 (1952).
  2. Pullman, 111. E., Colowick, S. P., and Kaplan, N. O., J. Biol. Chem., 194, 593(1952).
  3. Kaplan, N. O., Colowick, S. P., and Ciotti, M. M., J. Biol. Chem., 194, 579 (1952).
  4. Kaplan, N. O., Colowick, S. P., and Nason, A., J. Biol. Chem., 191, 473 (1951).
  5. Racker, E., J. Biol. Chem., 184, 313 (1950).
  6. Horecker, B. F., J. Biol. Chem., 183, 593 (1950).

Section !II. 

Luis_Federico_Leloir_-_young

The Leloir pathway: a mechanistic imperative for three enzymes to change
the stereochemical configuration of a single carbon in galactose.

Frey PA.
FASEB J. 1996 Mar;10(4):461-70.    http://www.fasebj.org/content/10/4/461.full.pdf
PMID:8647345

The biological interconversion of galactose and glucose takes place only by way of
the Leloir pathway and requires the three enzymes galactokinase, galactose-1-P
uridylyltransferase, and UDP-galactose 4-epimerase.
The only biological importance of these enzymes appears to be to

  • provide for the interconversion of galactosyl and glucosyl groups.

Galactose mutarotase also participates by producing the galactokinase substrate
alpha-D-galactose from its beta-anomer. The galacto/gluco configurational change takes place at the level of the nucleotide sugar by an oxidation/reduction
mechanism in the active site of the epimerase NAD+ complex. The nucleotide portion
of UDP-galactose and UDP-glucose participates in the epimerization process in two ways:

1) by serving as a binding anchor that allows epimerization to take place at glycosyl-C-4 through weak binding of the sugar, and

2) by inducing a conformational change in the epimerase that destabilizes NAD+ and
increases its reactivity toward substrates.

Reversible hydride transfer is thereby facilitated between NAD+ and carbon-4
of the weakly bound sugars.

The structure of the enzyme reveals many details of the binding of NAD+ and
inhibitors at the active site
.

The essential roles of the kinase and transferase are to attach the UDP group
to galactose, allowing for its participation in catalysis by the epimerase. The
transferase is a Zn/Fe metalloprotein
, in which the metal ions stabilize the
structure rather than participating in catalysis. The structure is interesting
in that

  • it consists of single beta-sheet with 13 antiparallel strands and 1 parallel strand
    connected by 6 helices.

The mechanism of UMP attachment at the active site of the transferase is a double
displacement
, with the participation of a covalent UMP-His 166-enzyme intermediate
in the Escherichia coli enzyme. The evolution of this mechanism appears to have
been guided by the principle of economy in the evolution of binding sites.

PMID: 8647345 Free full text

Section IV.

More on Lipids – Role of lipids – classification

  • Energy
  • Energy Storage
  • Hormones
  • Vitamins
  • Digestion
  • Insulation
  • Membrane structure: Hydrophobic properties

Lipid types

lipid types

lipid types

nat occuring FAs in mammals

nat occuring FAs in mammals

Read Full Post »