![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
,2
* Research Institute for Health Fundamentals and Institute of Life Sciences, Ajinomoto Co., Inc., Tokyo 210-8681, Japan
2 To whom correspondence should be addressed. E-mail: takeshi_kimura{at}ajinomoto.com.
 |
ABSTRACT |
|---|
 TOP ABSTRACT LITERATURE CITED |
|---|
KEY WORDS: • amino acid • metabolome • safety • excess • cluster analysis • correlation • multivariate • metabolism • animal • rat
Following the rapid developments in genome, transcriptome and proteome technology, there is a growing interest in metabolome research as can be seen by the rapidly growing number of publications over the last few years (Fig. 1). The basic idea in metabolomics is to characterize and understand the behavior of all metabolites in an organism, and to this end, a number of approaches are being pursued.
|
In contrast, modeling complex multicellular organisms is not a simple task and other more empirical approaches have been attempted. Apart from microorganisms, the use of genetic engineering for commercial use is most prevalent in plants, and metabolome analysis has become an important tool in characterizing the effects of the genetic modifications that have been made (4,5). The potential for using metabolomics for assuring the safety of genetic modifications made in plants has also been pointed out (6,7).
In animals and humans, metabolic profiling of blood and urine components to characterize toxicity and disease states such as inborn errors of metabolism has been ongoing since the introduction of gas-chromatography mass-spectrometry (GC-MS) in the mid-1960s (8). It could perhaps be said that some of the groundwork has already been done on the animal and human metabolome, although the data have not been analyzed from a bioinformatics perspective. Work on metabolic serotypes in rats is in progress (9), and it has been pointed out that metabolome research in animals and humans may offer very important insights into the biology of animals and humans especially in the field of nutrition (10).
Unlike the other -omic technologies, there is no single universal method of metabolite identification yet, and this is one of the hurdles to be overcome for metabolome research to develop at the same rapid pace as the other -omic technologies. Currently much of the analysis is dependent on mass spectrometry (MS)3 (11) and nuclear magnetic resonance (NMR) (12), but developments in the laboratory on a chip-type miniaturization technology may also yield important breakthroughs (13).
Analyzing the amino acid metabolomic subset
Data analysis methods for metabolomics are in the early stages of development. Various approaches have been suggested and many analyses and visualizations used in the DNA microarray area have been applied to metabolomics (14). Among the methods used to analyze metabolomic data are principal component analysis and clustering. We describe below some concepts and methods that we have been developing to utilize metabolomic data in understanding the relationships among plasma amino acid levels in animals.
As described above, one of the approaches to metabolome research is to determine all, or as many as possible, metabolites and to elucidate their relationships. We have taken the view that useful information may be gained by the analysis of a metabolomic subset, such as the set of amino acids, to detect changes in certain relationships. The study of another metabolomic subset, the lipid metabolome, has been shown to yield useful results (15). The amino acid metabolome seemed to be especially relevant in studying the effects of excessive intakes of protein and amino acids, and was a convenient model to focus on how to analyze metabolomic data.
The change in plasma amino acid concentrations over time, after the ingestion of a meal, has been well studied (16,17,18). The changes in postprandial or postabsorptive levels of amino acids after adaptation to various levels of protein in a meal have also been studied, and the paradoxical fall in plasma concentrations of certain amino acids with increasing protein intake has been noted (19,20). It is thus clear that plasma amino acid levels do not simply reflect the net intake of dietary amino acids and that the final plasma level for each amino acid is determined by a multitude of factors including catabolic rates (and anabolic rates for certain amino acids) and transport rates into and out of various organs. It is also clear that there is a metabolic adaptation that takes place after the ingestion of various levels of protein and amino acids, and recent DNA microarray data comparing the expression of several thousand genes from the liver of animals fed different diets suggest that changes in gene expression are associated with this adaptation process (21). Changes in gene expression and enzyme activities affecting metabolism in a particular organ should therefore be reflected in the relationships among the various metabolite concentrations in plasma.
When considering the relationships among metabolite levels, it is important to note that the various metabolites constitute networks within cells, tissues or the whole organism. At the single-cell level (i.e., at the intracellular level), each amino acid is influenced by several biochemical reactions that occur not only between amino acids, but also between amino acids and nonamino acid metabolites produced by sugar, lipid and nucleotide metabolism. In addition, at the multicell level (i.e., at the tissue level), amino acids are transported between cells via specific transporters. At the level of the entire organism, amino acids are continuously transported via blood flow to various tissues. The regulation of each biochemical reaction occurs at a variety of levels (e.g., transcriptional, posttranscriptional, enzymatic and substrate), but regardless of the level of the network, they exist within an individual living organism, and not between them, except possibly in the case of parasitic or symbiotic relationships. Therefore, it would be expected that much information regarding the network of relationships among various metabolites would be lost by the traditional approach of averaging results from different individuals within a treatment group. This is especially so when the potential for genetic variability and variations in gene expression levels among individuals within a treatment group is considered.
We suggest that, by examining the correlations among plasma amino acid levels in individual animals, some information about the organization of the network can be obtained. We have found that, by plotting the plasma levels of pairs of amino acid against each other in each animal, high correlation can be seen between certain pairs of amino acids, whereas little or no correlation is seen between other pairs (Fig. 2). By calculating correlation coefficients between each pair of amino acids, a matrix of multivariate correlation coefficients can be generated. Alternately, these data can be further processed to compare the correlations of one amino acid with all other amino acids against the correlations of another amino acid with all other amino acids, respectively. This gives a correlation matrix representing the closeness of the patterns of correlations among all amino acids.
|
|
In many cases, metabolome research would produce many unidentified peaks and the data would be of pseudoquantitative nature as the identities of the metabolites would not be known and there would be no standards to compare. We have found correlation analysis to be insensitive to the quantitative scaling of data, so analysis of the unidentified peaks, which may be observed when the object sample has gone through chromatographic or mass spectrometry analysis, is possible. This is a great advantage over other methods of analysis that rely on absolute values, and makes it possible to treat the unknown peaks together with the known ones. This opens the way to the targeted selection of unidentified peaks that correlate with a chosen marker, greatly reducing the expenditure of time and energy in the identification of unnecessary peaks. We believe that the combination of arbitrarily scaled pseudoquantitative metabolic profiling and correlation-based analyses could serve as the basis of a metabolomic approach to assessing the safety of amino acid intake, as shown in Figure 4. Quantitative data of known metabolites, arbitrarily scaled data of unidentified substances, and data quantifying toxicity (or other effects) in animals given various levels of an amino acid including excess levels, could be analyzed together by CAMC to give clusters representing the closeness of behavior of the metabolic and toxicological parameters with increasing doses of amino acids. Metabolites that cluster together with parameters representing toxicity are likely to be related directly or indirectly with toxicity. By investigating further the identity of the unidentified metabolites that cluster with toxicity parameters, and whether the dosing of these compounds to animals leads to similar types of effects observed after the ingestion of excessive levels of an amino acid, it may be possible to determine the metabolite responsible for the toxicity or a surrogate marker for it.
|
The use of metabolic profiling for searching for suspected toxic compounds and metabolic markers of disease has a long history. Developments in metabolomics should aid this process by offering a more systematic framework for the collection and analysis of data, by allowing the detection of relationships that were previously undetected and by making it possible to understand the position of a particular metabolite within the network of all measured metabolites. In the short term, metabolomic techniques could help identify metabolites that are closely associated with the effects of extreme intakes of macronutrients and help define the adequate range of intakes for various macronutrients. In the long term, the study of metabolomics could lead to the understanding of what happens to metabolism as a whole after the excess or lack of a single nutrient such as an amino acid. In addition, it could lead to the development of paradigms that take into account the complexities of metabolism, culminating in the establishment of scientifically based assessment strategies for the adequate range of macronutrient intakes.
|
FOOTNOTES |
|---|
3 Abbreviations used: CAMC, cluster analysis of multivariate correlations ; GC-MS, gas-chromatography mass-spectrometry; MS, mass spectrometry; NMR, nuclear magnetic resonance.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTLITERATURE CITED |
|---|
1. Oliver, S. G. (2002) Functional genomics: lessons from yeast. Philos. Trans. R. Soc. Lond. B Biol. Sci. 357: 17–23.[Medline]
2. Forster, J., Gombert, A. K. & Nielsen, J. (2002) A functional genomics approach using metabolomics and in silico pathway analysis. Biotechnol. Bioeng. 79: 703–712.[Medline]
3. Tomita, M. (2001) Whole-cell simulation: a grand challenge of the 21st century. Trends Biotechnol. 19: 205–210.[Medline]
4. Kose, F., Weckwerth, W., Linke, T. & Fiehn, O. (2001) Visualizing plant metabolomic correlation networks using clique-metabolite matrices. Bioinformatics
17: 1198–1208.
5. Fiehn, O. (2002) Metabolomics—the link between genotypes and phenotypes. Plant Mol. Biol. 48: 155–171.[Medline]
6. Kuiper, H. A., Kleter, G. A., Noteborn, H. P. & Kok, E. J. (2001) Assessment of the food safety issues related to genetically modified foods. Plant J. 27: 503–528.[Medline]
7. Schilter, B. & Constable, A. (2002) Regulatory control of genetically modified (GM) foods: likely developments. Toxicol. Lett. 127: 341–349.[Medline]
8. Gates, S. C. & Sweeley, C. C. (1978) Quantitative metabolic profiling based on gas chromatography. Clin. Chem.
24: 1663–1673.
9. Shi, H., Vigneau-Callahan, K. E., Shestopalov, A. I., Milbury, P. E., Matson, W. R. & Kristal, B. S. (2002) Characterization of diet-dependent metabolic serotypes: primary validation of male and female serotypes in independent cohorts of rats. J. Nutr.
132: 1039–1046.
10. German, J. B., Roberts, M. A., Fay, L. & Watkins, S. M. (2002) Metabolomics and individual metabolic assessment: the next great challenge for nutrition. J. Nutr.
132: 2486–2487.
11. Hall, R., Beale, M., Fiehn, O., Hardy, N., Sumner, L. & Bino, R. (2002) Plant metabolomics: the missing link in functional genomics strategies. Plant Cell
14: 1437–1440.
12. Reo, N. V. (2002) NMR-based metabolomics. Drug Chem. Toxicol. 25: 375–382.[Medline]
13. Khandurina, J. & Guttman, A. (2002) Bioanalysis in microfluidic devices. J. Chromatogr. A 943: 159–183.[Medline]
14. Mendes, P. (2002) Emerging bioinformatics for the metabolome. Brief. Bioinform.
3: 134–145.
15. Watkins, S. M., Reifsnyder, P. R., Pan, H. J., German, J. B. & Leiter, E. H. (2002) Lipid metabolome-wide effects of the PPARgamma agonist rosiglitazone. J. Lipid Res.
43: 1809–1817.
16. Ozalp, I., Young, V. R., Nagchaudhuri, J., Tontisirin, K. & Scrimshaw, N. S. (1972) Plasma amino acid response in young men given diets devoid of single essential amino acids. J. Nutr. 102: 1147–1158.
17. Ljungqvist, B. G., Svanberg, U. S. & Young, V. R. (1978) Plasma amino acid response to single test meals in humans. II. Healthy young adults given synthetic amino acid mixtures. Res. Exp. Med. (Berl.) 174: 13–28.[Medline]
18. Hegarty, J. E., Fairclough, P. D., Moriarty, K. J., Clark, M. L., Kelly, M. J. & Dawson, A. M. (1982) Comparison of plasma and intraluminal amino acid profiles in man after meals containing a protein hydrolysate and equivalent amino acid mixture. Gut
23: 670–674.
19. Moundras, C., Remesy, C. & Demigne, C. (1993) Dietary protein paradox: decrease of amino acid availability induced by high-protein diets. Am. J. Physiol. 264: G1057–G1065.[Medline]
20. Forslund, A. H., Hambraeus, L., van Beurden, H., Holmback, U., El-Khoury, A. E., Hjorth, G., Olsson, R., Stridsberg, M., Wide, L., Akerfeldt, T., Regan, M. & Young, V. R. (2000) Inverse relationship between protein intake and plasma free amino acids in healthy men at physical exercise. Am. J. Physiol. Endocrinol. Metab.
278: E857–E867.
21. Endo, Y., Fu, Z., Abe, K., Arai, S. & Kato, H. (2002) Dietary protein quality and quantity affect rat hepatic gene expression. J. Nutr. 132: 1632–1637.
22. Dybing, E., Doe, J., Groten, J., Kleiner, J., O'Brien, J., Renwick, A. G., Schlatter, J., Steinberg, P., Tritscher, A., Walker, R. & Younes, M. (2002) Hazard characterisation of chemicals in food and diet. dose response, mechanisms and extrapolation issues. Food Chem. Toxicol. 40: 237–282.[Medline]
23. Harper, A. E., Benevenga, N. J. & Wohlhueter, R. M. (1970) Effects of ingestion of disproportionate amounts of amino acids. Physiol. Rev.
50: 428–558.
This article has been cited by other articles:
|
|
 |
|
  Y. Noguchi, N. Shikata, Y. Furuhata, T. Kimura, and M. Takahashi Characterization of dietary protein-dependent amino acid metabolism by linking free amino acids with transcriptional profiles through analysis of correlation Physiol Genomics, August 14, 2008; 34(3): 315 - 326. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Toue, R. Kodama, M. Amao, Y. Kawamata, T. Kimura, and R. Sakai Screening of Toxicity Biomarkers for Methionine Excess in Rats J. Nutr., June 1, 2006; 136(6): 1716S - 1721S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Noguchi, Q.-W. Zhang, T. Sugimoto, Y. Furuhata, R. Sakai, M. Mori, M. Takahashi, and T. Kimura Network analysis of plasma and tissue amino acids and the generation of an amino index for potential diagnostic use Am. J. Clinical Nutrition, February 1, 2006; 83(2): 513S - 519S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. G. Robertson Metabonomics in Toxicology: A Review Toxicol. Sci., June 1, 2005; 85(2): 809 - 822. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Matsuzaki, H. Kato, R. Sakai, S. Toue, M. Amao, and T. Kimura Transcriptomics and Metabolomics of Dietary Leucine Excess J. Nutr., June 1, 2005; 135(6): 1571S - 1575S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. Milner Molecular Targets for Bioactive Food Components J. Nutr., September 1, 2004; 134(9): 2492S - 2498S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. M. Rand Variability in Planning and Assessment: Statistical Considerations and Models J. Nutr., June 1, 2004; 134(6): 1610S - 1616S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Sakai, M. Miura, M. Amao, R. Kodama, S. Toue, Y. Noguchi, and T. Kimura Potential Approaches to the Assessment of Amino Acid Adequacy in Rats: A Progress Report J. Nutr., June 1, 2004; 134(6): 1651S - 1655S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Kato and T. Kimura Evaluation of the Effects of the Dietary Intake of Proteins and Amino Acids by DNA Microarray Technology J. Nutr., June 1, 2003; 133(6): 2073S - 2077. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
