A proton nuclear magnetic resonance-based metabolomic approach in IgA nephropathy urinary profiles


Immunoglobulin A nephropathy (IgAN) is a mesangial proliferative glomerulonephritis, the most common primary glomerulonephritis in many developed countries of the world where renal biopsy is widely used for diagnosis. It is an immune complex disorder characterized by abnormal production of deglycosylated IgA1, which causes formation of immune complexes and/or polymeric IgA1 to be deposited in the glomeruli Serino et al. 2012 [1] (full text); Cox et al. 2010 [2]). Nuclear magnetic resonance (NMR) spectroscopy is a commonly used analytical method to analyze the small molecule composition, that is, the metabolome, of body fluids such as urine and blood serum. Metabolite concentrations have been associated with the biochemical status of an organism and their variations reflect changes in metabolism arising from specific conditions Del Coco et al. [3] (full text), including disease and response to drug treatment.


Both one-dimensional NOESY and transverse-relaxation filter CPMG NMR spectra were recorded to investigate the urine metabolome of 24 IgAN patients and to detect altered metabolic profiles in comparison with 68 healthy matched controls (Figure 1). The spectral data were analyzed using multivariate statistical techniques. The analysis revealed that the NMR spectra of IgAN patients were statistically different from those of the controls (Figure 2; P = 4 9 10-7 for 1D-NOESY and P = 2 9 10-7 for CPMG). The robustness of the determined statistical model was confirmed by its predictive performance (for the 1D-NOESY dataset: sensitivity = 67 %, specificity = 95 %; for the CPMG dataset sensitivity = 60 %, specificity = 94 %). Of the original 375 variables per spectrum, 36 (34) PC were sufficient to describe 99.0 % of the variance of the 1D-NOESY (CPMG) dataset (Figure 3). The contribution of the original spectral variables to the separation of groups was then evaluated determining the global standardized loading coefficients of the PCA/CA analysis. Figure 4 (for 1DNOESY and CPMG datasets, respectively) shows the distribution of samples in the CA coordinate space, where only one dimension is meaningful, as there are only two classes involved.The global coefficients for the first canonical variable (the one explaining differences between groups) are displayed in Figure 5. Figure 6 better highlights additional signals corresponding to altered metabolite levels of IgAN patients. For the first time we found metabolites, including betaine and citrate, that are differentially modulated in IgAN patients compared to controls and that may be directly involved in the pathogenesis of IgAN. These metabolites may influence, directly or indirectly, the TNF-a, a regulating factor of the Th1/Th2 cell balance that is relevant in the pathology (Figure 7). The involvement of metabolites such as betaine and citrate in TNF-a regulation supports the power of the identified metabolic profiles to discern IgAN from controls.


In summary, the present study illustrates the application of NMR spectroscopy to investigate the urine metabolite levels of patients affected by IgAN in comparison with healthy matched controls, resulting in the identification of a metabolic profile specific for IgAN. The described results lead us to propose a link between a number of metabolites, the TNF-a regulatory pathway of Th1/Th2 cell balance, and IgA nephropathy.