The clinical chemome: a tool for the diagnosis and management of chronic disease.

Garrod was among the first to recognize the importance of genetics as a determinant of disease. In his Croonian Lecture in 1908, he introduced the term “inborn errors of metabolism”, long before the nature of genes, proteins, and enzymes was understood. Numerous recent articles in Clinical Chemistry have focused on the application of modern mass spectrometry to the analysis of metabolites in plasma or urine for diagnosis of genetic diseases resulting from defects in single enzymes or transport systems. Novel liquid chromatography–mass spectrometry (LC/MS) methodologies, supported by bioinformatics and pattern-recognition algorithms, are also being developed for “omic” analysis and diagnosis of disease, including “genomics” (genetic complement), “transcriptomics” (gene expression), “proteomics” (protein synthesis and signaling), “metabolomics” (concentrations and fluxes of cellular metabolites), and “metabonomics” (systemic profiling by analysis of biological fluids) (1 ). The more common, noninfectious diseases in the modern world, such as cancer, diabetes, and cardiovascular and neurodegenerative diseases, do not fit into Garrod’s construct—that a deficiency in a single enzyme leads to a single disease entity. These and other chronic, age-related diseases are multifactorial and multigene in origin and often have a substantial nonenzymatic, nonmetabolic, chemical component. Chemical modification of proteins, often as a result of oxidative stress mediated by reactive oxygen and nitrogen species, is considered an important factor in the pathogenesis of these diseases (2 ). Chronic tissue damage is detectable in all of these diseases in the form of oxidized nucleotides and amino acids as well as in the form of chemical modifications of proteins by products of oxidation of sugars and lipids, known respectively as advanced glycoxidation and lipoxidation end-products (AGE/ALEs) (3 ). These compounds accumulate in tissues and on plasma proteins and are excreted in urine as a result of normal turnover processes. The complement of nonenzymatic, chemical modifications of biomolecules in the body constitutes the “chemome”, and the analysis of these nonenzymatic products, such as glycohemoglobin, isoprostanes, malondialdehyde, nitroand ortho-tyrosine, and AGE/ALEs, can provide a unique insight into health, disease, risk for disease, and response to therapy. Changes in constituents of the chemome also appear with age; some modified amino acids accumulate naturally with age in long-lived proteins, such as crystallins and collagens, and the rate of their accumulation is accelerated in chronic disease (3 ). Compared with other analytical methods, LC-tandem MS (LC/MS/MS), particularly with electrospray ionization (ESI) and in the multiple-reaction monitoring mode, is particularly suited for “omic” analyses. The latest generation of instruments provides exceptional sensitivity and specificity for detection and quantification of trace analytes and allows for simultaneous analysis of a wide range of analytes by both positiveand negative-ionization methods, e.g., for the measurement of amino and carboxylic acids in urine (4 ). In contrast to gas chromatography–mass spectrometry (GC/MS) analysis, LC/MS does not typically require derivatization of the analyte, and it is widely applicable to the analysis of highermolecular-weight nonvolatile analytes, including peptides and gene fragments. With decreases in the cost of instruments and improvements in software, LC/MS/MS will become increasingly important and available for clinical diagnostics. In this issue of Clinical Chemistry, Teerlink et al. (5 ) describe the analysis of two AGE/ALEs, N -(carboxymethyl)lysine (CML) and N -(carboxyethyl)lysine (CEL), in human plasma protein by stable-isotope-dilution LC/ MS/MS. They demonstrate increases in CML and CEL in plasma proteins from patients on hemodialysis and peritoneal dialysis. They treat CML and CEL, which are increased twoto threefold in plasma proteins from dialysis patients, as “biomarkers of oxidative stress resulting from sugar and lipid oxidation”, supporting the hypothesis of Miyata et al. (6 ) on the increase in carbonyl stress in uremia. It will be important to determine whether these compounds will be useful as predictors of risk for development of cardiovascular disease and other complications of chronic dialysis therapy, and, of course, whether therapeutic intervention will suppress the concentrations of these compounds and the risk for development of uremic complications. The LC/MS/MS method applied by Teerlink et al. (5 ) is a first step—only two analytes are measured—that demonstrates an approach that can be used in a much more global approach to analysis of the chemome. The authors used reversed-phase (RP) HPLC with an aqueous–acetonitrile gradient containing nonafluoropentanoic acid (NFPA) as an ion-pairing agent. NFPA yielded good retention (retention time, 6–8 min) and peak shape for the polar amino acids CML and CEL. The method is a substantial improvement over current methodology: it does not require derivatization, has high throughput, and has sensitivity comparable to current GC/MS and GC/ MS/MS methods. Chaimbault et al. (7 ) compared several perfluoroalkanoic acid homologs and concluded that NFPA and tridecafluoroheptanoic acid yield optimal separation of protein amino acids by RP-HPLC on graphitic carbon columns. Ahmed and Thornalley (8 ) used a similar column with an acidic ammonium formate–acetonitrile gradient for analysis of 16 AGE/ALEs and amino acid oxidation and nitration products in tissue proteins and physiologic fluids. One of the unique features of their work was the use of exhaustive proteolytic digestion, as described by Glomb and Pfahler (9 ), to minimize degradation of acidand/or base-labile AGE/ALEs during hydrolysis of tissue proteins. They were able to detect increased AGEs in patients with vascular complications in Editorial