Journal of Nutrition Animal Diets/Enrichment Products...

Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lamers, R.-J. A. N.
Right arrow Articles by van Nesselrooij, J. H. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lamers, R.-J. A. N.
Right arrow Articles by van Nesselrooij, J. H. J.

© 2003 The American Society for Nutritional Sciences J. Nutr. 133:1776-1780, June 2003

Metabolomics

Identification of Disease- and Nutrient-Related Metabolic Fingerprints in Osteoarthritic Guinea Pigs

Robert-Jan A. N. Lamers2, Jeroen DeGroot*, Elly J. Spies-Faber, Renger H. Jellema, Virginia B. Kraus{dagger}, Nicole Verzijl*, Johan M. TeKoppele*, Gerwin K. Spijksma, Jack T.W.E. Vogels, Jan van der Greef and Joop H. J. van Nesselrooij

TNO Netherlands Organization for Applied Scientific Research, Nutrition and Food Research and Pharma, 3700 AJ Zeist, The Netherlands; * TNO Prevention and Health, Gaubius Laboratory, 2301 CE Leiden, The Netherlands; and {dagger} Duke University Medical Center, Department of Medicine, Division of Rheumatology, Allergy and Clinical Immunology, Durham, NC 27710

2To whom correspondence should be addressed. E-mail: r.lamers{at}voeding.tno.nl.


   ABSTRACT
TOP
 ABSTRACT
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
Osteoarthritis (OA), one of the most common diseases among the elderly, is characterized by the progressive destruction of joint tissues. Its etiology is largely unclear and no effective disease-modifying treatment is currently available. Metabolic fingerprinting provides a novel tool for the identification of biomarkers. A metabolic fingerprint consists of a typical combination of metabolites in a biological fluid and is identified by a combination of 1H NMR spectroscopy and multivariate data analysis (MVDA). The current feasibility study was aimed at identifying a metabolic fingerprint for OA and applying this in a nutritional intervention study. Urine samples were collected from osteoarthritic male Hartley guinea pigs (n = 44) at 10 and 12 mo of age, treated from 4 mo onward with variable vitamin C doses (2.5–3, 30 and 150 mg/d) and from healthy male Strain 13 guinea pigs (n = 8) at 12 mo of age, treated with 30 mg vitamin C/d. NMR measurements were performed on all urine samples. Subsequently, MVDA was carried out on the data obtained using NMR. An NMR fingerprint was identified that reflected the osteoarthritic changes in guinea pigs. The metabolites that comprised the fingerprint indicate that energy and purine metabolism are of major importance in OA. Metabolic fingerprinting also allowed detection of differences in OA-specific metabolites induced by different dietary vitamin C intakes. This study demonstrates the feasibility of metabolic fingerprinting to identify disease-specific profiles of urinary metabolites. NMR fingerprinting is a promising means of identifying new disease markers and of gaining fresh insights into the pathophysiology of disease.

KEY WORDS: • metabolic fingerprinting • 1H NMR spectroscopy • multivariate data analysis • osteoarthritis • vitamin C

Biomedical research is continuously facing the challenge of elucidating the relationship between health, disease and metabolism on one hand and the effects of nutrition or pharmaceuticals on the other hand. Genomics will contribute to clarifying the etiology of most common genetic diseases and provide approaches for therapeutic intervention. However, knowledge of genomics is not the single universal tool for predictive medicine and nutritional strategies (1,2). A person’s phenotype results from the interaction of the genotype with the environment, in which nutrition plays a major role. Metabolites are the quantifiable molecules that best reflect phenotype (3) and are attractive candidates for biomarker fingerprints in nutritional intervention studies.

Biological fluids, such as urine and blood, contain a large number of metabolites that may provide valuable information on the metabolism of an organism, and thus about its health status (4). Metabolic fingerprinting, also referred to as metabolomics, metabonomics (4), metanomics (1) or related terms, is a technique that enables quantification and identification of metabolites in biological fluids. The technology has emerged from the profiling of body fluid approaches that were developed many decades ago for the study of inborn errors of metabolism and the effects of nutrition. The early work in this area was driven mainly by GC-MS, which allows low concentration components to be measured in single profiles. In the 1980s, especially in combination with multivariate data analysis (MVDA), MS profiling became a powerful fingerprinting methodology (5).

For full coverage of a complex mixture of metabolites, a combination of analytical techniques is desirable. However, for global screening, 1H NMR is an attractive approach because a wide range of metabolites can be quantified at the same time without extensive sample preparation. More in-depth studies can subsequently elucidate the metabolic pathways involved, especially when metabolite information is integrated with gene expression and proteomic data in Systems Biology strategies (6).

NMR spectra of biological fluids are very complex due to the mixture of numerous metabolites present in these fluids (4). Therefore, variations among samples are often too small to be recognized by eye. To increase the comparability of NMR spectra and thereby maximize the power of the subsequent data analysis, we developed a Partial Linear Fit algorithm (7) in the past. This algorithm adjusts minor shifts in the spectra while maintaining the resolution. To find significant differences, MVDA is required to explore recurrent patterns in a number of NMR spectra (4,5). MVDA is a powerful tool for the analysis of data sets with a large number of variables. It visualizes the correlation between variables in complex or large data sets (e.g., thousands of signals in NMR spectra) in relation to a target variable such as disease status. MVDA falls into two general classes, i.e., unsupervised and supervised techniques. Unsupervised methods such as principal component analysis (PCA) determine patterns within data sets, without prior knowledge, and visualize the data in such a way as to emphasize their similarities and differences. With such methods, a direct comparison of NMR spectra is made and subsets of data are formed, solely on the basis of NMR spectral similarities (4).

In PCA, data are transformed from a large set of related variables (e.g., NMR signals) to a smaller set of uncorrelated variables. The newly created variables are called principal components (PC) and aim at the expression of maximum variation in the original variables. Each PC forms an axis in multidimensional space and the calculated distance of an object (e.g., a complete NMR spectrum of a guinea pig urine sample) to this axis is termed a score. The contribution of each variable (e.g., a single NMR signal) to a PC can also be calculated, giving a loading. A high loading indicates a strong contribution of the original NMR signal to the investigated PC. Loadings can be displayed in a factor spectrum. Loading vectors are described as lines then, with a position equal to the position of the variables in the original spectra. The height of the lines indicates the contribution of the variables to the investigated direction.

Supervised methods such as partial least squares and principal component discriminant analysis (PCDA) are more powerful tools, which use additional information on the data set such as biochemical, histopathological or clinical data to identify differences between predefined groups (8). In PCDA, the scores from PCA are used as a starting point for linear discriminant analysis. Discriminant analysis works by combining the PC in such a way that differences between predefined groups are maximized.

Osteoarthritis (OA), the most common form of arthritis, is a multifactorial, chronic joint disease that is characterized by the progressive destruction of articular cartilage, resulting in impaired movement, pain and, ultimately, disability (9). A variety of systemic and local risk factors have been identified that predispose to the development of OA, including, but not limited to, age, gender, bone density, obesity, joint injury and nutritional factors (10). Despite the growing body of information on the pathogenesis of OA, its etiology is far from clear, and effective disease-modifying treatment is lacking. Diagnosis of OA is currently based on clinical symptoms in combination with imaging techniques such as radiology or MRI, to visualize the degenerative changes in the joint. These changes can be observed only in an advanced stage of the disease, in which joint tissue damage is considered irreversible. Alternative methods are therefore required to detect osteoarthritic changes in the joints in an early stage of the disease in a quantitative, reliable and sensitive manner (11). By measuring a combination of relevant metabolites in biological fluids, metabolic fingerprinting potentially meets these criteria.

Hartley outbred strain guinea pigs develop spontaneous progressive knee OA, with features similar to the human disease (12,13) and were chosen to investigate the potential of metabolic fingerprinting as a tool for diagnosing and monitoring a nutritional intervention.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 LITERATURE CITED
 
Animal handling procedure.

Male Hartley guinea pigs (n = 46), which develop OA during aging, were purchased at 2 mo of age from Charles River Laboratories (Wilmington, MA) and fed standard guinea pig feed. Beginning at 4 mo of age, the guinea pigs were divided randomly into three treatment groups with varying doses of vitamin C provided in 30 mg of feed daily as follows: low dose (2.5–3 mg/d) which exceeded the minimum amount necessary to prevent scurvy, medium dose (30 mg/d), and high dose (150 mg/d). Guinea pigs also consumed ad libitum a standard Purina Lab Diet 5025 (Purina Mills, LLC, St. Louis, MO) without vitamin C. Further, Strain 13 guinea pigs (n = 8; obtained from Crest Caviary, Prundale, CA), which develop OA to a much lesser extent than the Hartley strain (14), were housed individually in solid bottom cages and fed 30 mg vitamin C/d. Metabolic cages (PLAS-LABS, Lansing, MI) suitable for guinea pigs were used to collect 24-h urine samples at 10 and 12 mo of age for the Hartley guinea pigs and at 12 mo of age for the Strain 13 guinea pigs. The collected urine samples were centrifuged at 1000 x g for 10 min to remove debris and stored at -80°C until analyses. For all experiments, "Principles of laboratory animal care" were followed and American guidelines and laws were followed.

NMR analysis of urine samples.

Before NMR spectroscopic analysis, 200-µL urine samples were lyophilized and reconstituted in 1 mL sodium phosphate buffer (0.1 mmol/L, pH 6.0, made up with D2O), to minimize spectral variance arising from differences in urinary pH. Sodium trimethylsilyl-[2,2,3,3,-2H4]-1-propionate (TMSP; 0.025 mmol/L) was added as an internal standard. NMR spectra were recorded in random order and in triplicate on a fully automated Varian UNITY 400 MHz spectrometer (Palo Alto, CA) using a 1H NMR set-up operating at a temperature of 293K.

Free induction decays (FID) were collected as 64,000 data points with a spectral width of 8.000 Hz; 45° pulses were used with an acquisition time of 4.10 s and a relaxation delay of 2 s. The spectra were acquired by accumulation of 128 FID. The signal of the residual water was removed by a presaturation technique in which the water peak was irradiated with a constant frequency for 2 s before the acquisition pulse. The spectra were processed using standard Varian software. An exponential window function with a line broadening of 0.5 Hz and a manual baseline correction were applied to all spectra. After referring to the internal NMR reference (TMSP {delta} = 0.0), line listings were prepared using standard Varian NMR software. To obtain these listings, all lines in the spectra above a threshold corresponding to ~3 times the signal-to-noise ratio were collected and converted to a data file suitable for multivariate data analysis applications.

NMR data processing and multivariate data analysis.

The NMR data reduction file was imported into Winlin (V1.10, TNO, The Netherlands). Minor variations from comparable signals in different NMR spectra were adjusted and lines were fitted without loss of resolution, based on the Partial Linear Fit algorithm (7). To correct for urinary dilution, the data were autoscaled so that small and large peaks contributed similarly to the final study result. Where needed, endogenous and exogenous metabolites of vitamin C were eliminated from the NMR spectra, leading to more universal OA-related changes, and PCDA was performed (15).

Age, strain or dose was used as additional input information. For each PCDA analysis, the respective data set was randomly divided into a training data set and a test data set. PCDA models were built upon the training data set. The resulting discriminants were quantified for each of the urinary NMR spectra and the first discriminant (D1) was plotted vs. the second discriminant (D2) to visualize clustering. In a 2-category discriminant approach, all variance (100%) is explained by the first discriminant axis D1. Therefore, only variations explained by D1, the horizontal direction, were examined. The test data set was used to test the accuracy of the PCDA model by passing it through this model to obtain the model’s prediction of classification of the test data into the clusters from the PCDA training model. Predictions were within the 95% CI.

Factor spectra were used then to correlate the position of clusters in the score plot to the original NMR features in the spectra. The factor spectra, or metabolic fingerprints, were prepared in directions of maximum separation of one category vs. the other category, to provide insight into the type of metabolites responsible for the separation between the categories.


   RESULTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 LITERATURE CITED
 
Metabolic fingerprinting in disease diagnosis.

Hartley outbred stock albino guinea pigs spontaneously develop an osteoarthritic condition that closely resembles its human counterpart. The earliest histological signs of the disease appear at 3 mo of age in the medial tibial plateau and gradually progress to extensive cartilage degeneration in guinea pigs >= 12 mo old (16).

The underlying hypothesis of the present study is that OA will disturb metabolism, which will be reflected in an aberrant urinary metabolic composition. Using metabolic fingerprinting, such OA-induced abnormal urinary composition may be quantified. However, aging may also cause disturbances in metabolism that are independent of a pathological change (17). Therefore, a suitable control group was essential for the construction of a representative metabolic fingerprint for OA. To exclude the possibility that metabolic differences caused by aging interfered with those caused by OA, NMR analysis was performed on urine samples of Hartley guinea pigs that were followed longitudinally. Samples were collected from the same guinea pigs treated with 30 mg vitamin C/d at 10 and 12 mo of age. This approach minimized age effects on the metabolite profile, i.e., at 10 mo of age, guinea pigs are fully grown, whereas the OA severity is expected to increase in the guinea pigs from 10 mo onward (14). Comparison of two urinary spectra of a single guinea pig at 10 and 12 mo of age (Fig. 1) showed that, at first glance, the differences between the respective NMR spectra were small.



View larger version (12K):
[in this window]
[in a new window]
  		FIGURE 1 NMR spectra of urine taken at 10 mo (left panel) and 12 mo (right panel) from one Hartley guinea pig. Initially, no differences appeared to exist between the two spectra.

 
Because the NMR spectra contain a range of different signals, combinations of which represent the different metabolites, PCDA was used to visualize disease-related differences in these spectra (15), using age (10 or 12 mo) as additional input information. The resulting score plot (Fig. 2A) clearly showed the grouping of samples per age group (and thus possibly per OA severity) that could not be distinguished by eye on the basis of the individual NMR spectra. Thus, PCDA resulted in a visible and significant separation between groups that is based on ratios of the concentrations of metabolites which were characteristic for each group. A metabolic fingerprint was derived from this score plot, which was likely to reflect OA (Fig. 2B). In this fingerprint, NMR signals of guinea pig urinary metabolites are visualized which increased or decreased in association with age and thus OA.



View larger version (25K):
[in this window]
[in a new window]
  		FIGURE 2 Principal component discriminant analysis (PCDA) score plots visualizing disease-related differences in NMR spectra. (A) PCDA score plot of urinary NMR spectra of Hartley guinea pigs at 10 vs. 12 mo of age (the points represent the complete urinary NMR spectra of the guinea pigs). At 10 mo of age, guinea pigs are fully grown, whereas at 12 mo of age their osteoarthritis (OA) severity has increased substantially (14). This increase in OA severity is reflected by the urinary composition because there is a clear difference between the positions of the two groups. (B) Factor spectrum, or metabolic fingerprint, of urinary NMR spectra of 10 vs. 12 mo old Hartley guinea pigs that is typical for OA. Peaks (representing NMR signals) in the positive direction indicate metabolites that are more abundant in urine of guinea pigs with severe OA (at 12 mo) than in urine of guinea pigs with milder disease (at 10 mo). Consequently, metabolites that are more abundant in urine of these healthier guinea pigs are presented as peaks in the negative direction.

 
With this approach, it is possible that the small age difference (2 mo), rather than a different severity of OA, could be responsible for the clustering of the data and thus the metabolic fingerprint. Therefore, to ascertain that the metabolic fingerprint obtained reflected OA differences and not age differences, a complementary approach was used. NMR spectra were obtained from urine samples of 12-mo-old Strain 13 guinea pigs, treated with 30 mg vitamin C/d, that had only very limited OA changes (14). These NMR spectra were subjected to PCDA together with the data obtained from 12-mo-old Hartley guinea pigs, treated with 30 mg vitamin C/d. Using this design, only age-matched samples were compared, thereby completely eliminating age-related changes as confounders of the differences in the metabolite profile. For the PCDA, an ideal situation was assumed in which, at 12 mo of age, Hartley guinea pigs had OA, whereas Strain 13 guinea pigs did not (14). The score plot resulting from this analysis, in which "strain" (i.e., Hartley vs. Strain 13) was used as additional input data for PCDA, again showed clear separation of the two groups (Fig. 3A). A metabolic fingerprint was used to assign the NMR signals that explained differences between the two groups, and thus health status (Fig. 3B). In this analysis, both OA as well as strain differences could in principle be responsible for the separation of the groups.



View larger version (26K):
[in this window]
[in a new window]
  		FIGURE 3 Principal component discriminant analysis (PCDA) score plots visualizing disease-related differences in NMR spectra. (a) PCDA score plot of urinary NMR spectra of 12 mo old Hartley vs. Strain 13 guinea pigs. Strain 13 guinea pigs are healthy, whereas Hartley guinea pigs suffer from osteoarthritis (OA) (14). The difference in urinary composition is reflected by the clear separation into two groups in the graph. (b) OA specific metabolic fingerprint of urinary NMR spectra of 12 mo old Hartley vs. Strain 13 guinea pigs. Peaks in the positive direction indicate metabolites that are more abundant in urine of Hartley guinea pigs than in urine of Strain 13 guinea pigs. Consequently, metabolites that are more abundant in urine of Strain 13 guinea pigs are presented as peaks in the negative direction.

 
Combining the results from both analyses allowed elimination of the confounding factor from each of the approaches, i.e., age and strain, respectively. The metabolic fingerprint obtained from the comparison of urinary NMR spectra of 10- and 12-mo-old Hartley guinea pigs was similar to the metabolic fingerprint obtained from the age-matched strain comparison (compare Figs. 2b and 3b). Thus, metabolites that were present in both fingerprints are likely to be OA specific. Using this approach, NMR regions in the metabolic fingerprint that emerged as strongly associated with OA are 2.0–2.9, 3.0–4.7 and 6.3–7.5 ppm. Signals at 1.35, 2.42, 2.45, 2.88, 4.03 and 4.33 ppm in a positive direction and 1.45, 1.48 and 4.13 in a negative direction were especially abundant when OA was present. Signals in these regions represent, among others, lactic acid, malic acid, hypoxanthine and alanine. These metabolites are consistent with the hypothesis suggesting increased energy utilization and altered purine metabolism that exist in OA (18, 19).

Metabolic fingerprinting in efficacy studies.

To investigate the suitability of metabolic fingerprinting for monitoring the efficacy of a treatment on a disease, the effect of the nutrient vitamin C on the development of OA was studied. Vitamin C has been reported to have a beneficial effect on the progression of human OA (20). A longitudinal intervention study was carried out in which Hartley guinea pigs received a low (2.5–3 mg/d), medium (30 mg/d) or high (150 mg/d) dose of vitamin C. Urine samples were collected at 12 mo of age and were subjected to NMR analysis. PCDA was carried out on the NMR data, with vitamin C dose as additional input data, and a score plot was drawn (Fig. 4). The urinary NMR spectra of the guinea pigs treated with variable vitamin C doses were different, as is clear from the differences in their group positions on the score plot. This implies that the urinary metabolic composition, and thus OA metabolism, was affected by low vs. medium vs. high vitamin C dose. Hence, in examining the NMR data (i.e., the change of the metabolic fingerprint), vitamin C had a noticeable effect on the development of OA, assuming that the guinea pigs did not suffer from any comorbidity. Studies are in progress to assess the relationship between the NMR spectra and identified metabolic fingerprints and histological OA grading of the knee joint tissues.



View larger version (16K):
[in this window]
[in a new window]
  		FIGURE 4 Principal component discriminant analysis (PCDA) score plot of urinary NMR spectra of Hartley guinea pigs at 12 mo of age, treated with variable vitamin C doses [2.5–3, 30 or 150 mg/(kg · d)]. Vitamin C treatment affects the urinary metabolic composition, which is visualized by the different position of the three groups.

 

   DISCUSSION
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
LITERATURE CITED
 
In the present study, a novel method was employed to identify a metabolic fingerprint for OA, using NMR in combination with MVDA on urine samples of outbred male Hartley guinea pigs that spontaneously develop knee OA. An NMR fingerprint for OA was identified that was independent of age or strain effects and therefore can be used as a diagnostic tool for OA in guinea pigs. Also, in an efficacy study, metabolic fingerprinting detected differences in OA-specific metabolites due to treatment with vitamin C. In this study it was essential to use samples from a well-defined animal study, rather than human material because differences in OA severity, medication, diet and habits, for example, create additional variability that would greatly hinder de novo identification of an OA-specific urinary fingerprint. However, once identified amid the thousands of other metabolites, the urinary fingerprint for OA may also be quantified in humans. Such a study is forthcoming.

This study demonstrated the feasibility of metabolic fingerprinting to identify metabolite profiles in (pre)clinical research. As shown here, this technique can distinguish a disease from a nondisease state. In addition to their contribution to the fingerprint, the individual metabolites may provide additional insight into the pathogenesis of OA. Lactic acid, malic acid, hypoxanthine and alanine contributed heavily to the fingerprint, suggesting their involvement in the osteoarthritic process. Increased energy demand and altered purine metabolism may thus play an important role in OA. Further identification of the metabolites involved and combining the current metabolic data with proteomic and genomic approaches to form a holistic, integrated picture of the metabolic pathways implicated in OA, may provide new insights into OA pathogenesis and thereby identify new disease targets. This approach also has the potential to promote the development of new biomarkers for OA.

Metabolic fingerprinting has the ability to distinguish disease-specific metabolites. Metabolic fingerprints, as demonstrated in this study of OA, can also provide a sensitive outcome measurement tool that can be used to evaluate the effects of a nutrient or drug intervention on the incidence and progression of disease. Therefore, this powerful technique has broad applicability in the field of clinical nutritional and pharmaceutical research.


   ACKNOWLEDGMENTS
 
The authors gratefully acknowledge Janet L. Huebner at Duke University Medical Center, Durham, NC, for technical expertise and histopathological scoring.


   FOOTNOTES
 
1 Funded in part by a grant from the Arthritis Foundation (to V.B.K.). Back

3 Abbreviations: FID, free induction decay; MVDA, multivariate data analysis; OA, osteoarthritis; PCA, principal components analysis; PCDA, principal component discriminant analysis; TMSP, sodium trimethylsilyl-[2,2,3,3,-2H4]-1-propionate. Back

Manuscript received 17 December 2002. Initial review completed 21 February 2003. Revision accepted 17 March 2003.


   LITERATURE CITED
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 

1. Watkins, S. M. (2000) Comprehensive lipid analysis: a powerful metanomic tool for predictive and diagnostic medicine. Isr. Med. Assoc. J. 2:722-724.[Medline]

2. van Ommen, B. & Stierum, R. (2002) Nutrigenomics: exploiting systems biology in the nutrition and health area. Curr. Opin. Biotechnol. 13:517-521.[Medline]

3. Watkins, S. M., Hammock, B. D., Newman, J. W. & German, J. B. (2001) Individual metabolism should guide agriculture toward foods for improved health and nutrition. Am. J. Clin. Nutr. 74:283-286.[Abstract/Free Full Text]

4. Nicholson, J. K., Lindon, J. C. & Holmes, E. (1999) "Metabonomics": understanding the metabolic responses of living systems to pathophysiological stimuli via multivariate statistical analysis of biological NMR spectroscopic data. Xenobiotica 29:1181-1189.[Medline]

5. van der Greef, J. (2002) The role of metabolomics in Systems Biology, a new vision for drug discovery and development. Harrigan, G. G. Goodacre, R. eds. Metabolic Profiling: Its Role in Biomarker Discovery and Gene Function Analysis 2002 Kluwer Academic Publishing Boston, MA. .

6. Ideker, T., Galitski, T. & Hood, L. (2001) A new approach to decoding life: systems biology. Annu. Rev. Genomics Hum. Genet. 2:343-372.[Medline]

7. Vogels, J.T.W.E., Tas, A. C., Venekamp, J. & van der Greef, J. (1996) Partial linear fit: a new NMR spectroscopy preprocessing tool for pattern recognition applications. J. Chemom. 10:425-438.

8. Lindon, J. C., Nicholson, J. K., Holmes, E. & Everett, J. R. (2000) Metabonomics: metabolic processes studies by NMR spectroscopy of biofluids combined with pattern recognition. Concepts Magn. Reson. 12:289-320.

9. Creamer, P. & Hochberg, M. C. (1997) Osteoarthritis. Lancet 350:503-508.[Medline]

10. Felson, D. T. (2000) Osteoarthritis: new insights. Part 1: The disease and its risk factors. Ann. Intern. Med. 133:635-646.[Abstract/Free Full Text]

11. DeGroot, J., Bank, R. A., Tchetverikov, I., Verzijl, N. & TeKoppele, J. M. (2002) Molecular markers for osteoarthritis: the road ahead. Curr. Opin. Rheumatol. 14:585-589.[Medline]

12. Huebner, J. L., Otterness, I. G., Freund, E. M., Caterson, B. & Kraus, V. B. (1998) Collagenase 1 and collagenase 3 expression in a guinea pig model of osteoarthritis. Arthritis Rheum 41:877-890.[Medline]

13. Bendele, A. M. (2001) Animal models of osteoarthritis. J. Musculoskel. Neuron Interact. 1:363-376.

14. Huebner, J. L., Hanes, M., Beekman, B., Tekoppele, J. M. & Kraus, V. B. (2002) A comparative analysis of bone and cartilage metabolism in two strains of guinea-pig with varying degrees of naturally occurring osteoarthritis. Osteoarthritis Cartilage 10:758.[Medline]

15. Hoogerbrugge, R., Willig, S. J. & Kistemaker, P. G. (1983) Discriminant analysis by double stage principal component analysis. Anal. Chem. 55:1710-1712.

16. Bendele, A. M., White, S. L. & Hulman, J. F. (1989) Osteoarthrosis in guinea pigs: histopathologic and scanning electron microscopic features. Lab. Anim. Sci. 39:115-121.[Medline]

17. Timbrell, J. (2000) Principles of Biochemical Toxicology 2000:117-171 Taylor and Francis London, UK.

18. Damyanovich, A. Z., Staples, J. R. & Marshall, K. W. (1999) 1H NMR investigation of changes in the metabolic profile of synovial fluid in bilateral canine osteoarthritis with unilateral joint denervation. Osteoarthritis Cartilage 7:165-172.[Medline]

19. Herbert, K. E., Scott, D. L. & Perrett, D. (1988) Nucleosides and bases in synovial fluid from patients with rheumatoid arthritis and osteoarthritis. Clin. Sci. (Lond.) 74:97-99.[Medline]

20. McAlindon, T. E., Jacques, P., Zhang, Y., Hannan, M. T., Aliabadi, P., Weissman, B., Rush, D., Levy, D. & Felson, D. T. (1996) Do antioxidant micronutrients protect against the development and progression of knee osteoarthritis?. Arthritis Rheum 39:648-656.[Medline]




This article has been cited by other articles:


Home page
 BioinformaticsHome page
'ANOVA-simultaneous component analysis (ASCA): a new tool for analyzing designed metabolomics data'
Bioinformatics, December 15, 2007; 23(24): 3415 - 3415.
[Full Text] [PDF]

Home page
 BioinformaticsHome page
A. K. Smilde, J. J. Jansen, H. C. J. Hoefsloot, R.-J. A. N. Lamers, J. van der Greef, and M. E. Timmerman
ANOVA-simultaneous component analysis (ASCA): a new tool for analyzing designed metabolomics data
Bioinformatics, July 1, 2005; 21(13): 3043 - 3048.
[Abstract] [Full Text] [PDF]

Home page
 Toxicol PatholHome page
W. H. M. Heijne, R.-J. A. N. Lamers, P. J. van Bladeren, J. P. Groten, J. H. J. van Nesselrooij, and B. van Ommen
Profiles of Metabolites and Gene Expression in Rats with Chemically Induced Hepatic Necrosis
Toxicol Pathol, June 1, 2005; 33(4): 425 - 433.
[Abstract] [Full Text] [PDF]

Home page
 J. Nutr.Home page
R.-J. A. N. Lamers, E. C. H. H. Wessels, J. J. M. van de Sandt, K. Venema, G. Schaafsma, J. van der Greef, and J. H. J. van Nesselrooij
A Pilot Study to Investigate Effects of Inulin on Caco-2 Cells through In Vitro Metabolic Fingerprinting
J. Nutr., October 1, 2003; 133(10): 3080 - 3084.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lamers, R.-J. A. N.
Right arrow Articles by van Nesselrooij, J. H. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lamers, R.-J. A. N.
Right arrow Articles by van Nesselrooij, J. H. J.

Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]