Carboxymethyl-lysine is a Prominent Target of Circulating IgA in IgA Nephropathy

Abstract

Introduction: IgA nephropathy (IgAN), the most prevalent glomerular disease worldwide, carries a significant risk of kidney failure. Its pathogenesis involves the presence of elevated levels of IgA and galactose-deficient IgA (Gd-IgA), deposition of these antibodies in the kidney mesangium, and complement-mediated glomerular injury, leading to progressive renal function loss. The source and specificity of IgA in this disease remain unclear. We hypothesized that pathogenic IgA results from an immune response to abnormal protein modifications.

Methods: We used an advanced enzyme-linked immunosorbent assay (ELISA) platform to assess serum IgA and Gd-IgA reactivity to 93 posttranslational modifications and other chemical adducts in 28 patients with biopsy-proven IgAN and 22 healthy controls.

Results: Carboxymethyl-lysine (CML) was identified as the dominant target of IgA and Gd-IgA, but not IgG in patients with IgAN. This finding was validated in an independent cohort of 15 IgAN cases and 15 controls. In addition, a positive correlation was found between serum CML concentration and IgA reactivity in patients with IgAN, alongside albuminuria. Lastly, immunofluorescence staining observed elevated CML deposition in glomeruli of patients with IgAN than in controls.

Conclusion: Our studies identify CML as a primary target of circulating IgA in IgAN, suggesting that aberrant responses to this modified self-antigen contribute to disease pathophysiology.

Keywords: IgA nephropathy; biomarker; biopsy; chemical adduct; kidney disease; proteinuria.

Graphical abstract

Figure 1

Serum level of total IgA, Gd-IgA1 and reactivity against adducts in patients with IgAN and controls. Serum level of (a) total IgA and (b) Gd-IgA1 in patients with IgAN and healthy controls. (c) Heat map of OD value of IgA reactivity against the 93 adducts (rows) in every sample of the training cohort (column). (d) Box plots of importance distribution of each adduct in Boruta algorithm. Blue boxplots represent a shadow attribute’s minimum, average, and maximum Z scores. Z scores of qualities that were rejected and confirmed, respectively, are represented by red and green boxplots. Box plot of (e) IgA, (f) Gd-IgA1, and (g) IgG reactivity (OD) toward CML in patients with IgAN and HCs. CML, N(6)-carboxymethyl-lysine; Gd-IgA1, galactose-deficient IgA1; HCs, healthy controls; IgAN, IgA nephropathy; OD, optical density. (∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001).

Figure 2

Anti-CML IgA in an independent validation cohort and ROC for anti-CML IgA, serum total IgA and Gd-IgA1 in the overall cohort. (a) Anti-CML IgA levels in patients with IgAN and healthy donors of the validation cohort. (b) ROC curve of anti-CML IgA, serum total IgA and Gd-IgA1 in the combined training and validation cohort. (c) Anti-CML IgA levels in patients with IgAN compared with DN, MN, and healthy donors, showing a significantly higher IgA reactivity against CML in patients with IgAN compared with other glomerular disease. CML, N(6)-carboxymethyl-lysine; DN, diabetic nephropathy; Gd-IgA1, galactose-deficient IgA1; IgAN, IgA nephropathy; MN, membranous nephropathy; ROC, receiver operating characteristic. (∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001).

Figure 3

Titration of IgA against CML and correlation with serum CML levels. (a) 2-fold titration of IgA against CML in both training and validation cohort. (b) higher serum dilution which IgA anti-CML is still detectable. (c) box plot of serum CML levels in both the cohort. (d) correlation between serum CML levels and IgA reactivity against CML. (e) correlation between urinary albumin and anti CML IgA (f) correlation between urinary albumin and serum CML. CML, N(6)-carboxymethyl-lysine. (∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001).

Figure 4

CML expression in tubules and glomeruli of patients with IgAN and controls. (a) Immunofluorescence staining showing higher expression of CML (green, top panels) in renal tubules and glomeruli from patients with IgAN compared with controls. DAPI nuclear staining: blue. Magnification, ×10. (b) Quantification of CML glomerular deposition. Fluorescence intensity was measured in all glomeruli of the available kidney sections. CML expression was then divided by the average MFI of the area stained (normalized data for the area) using ImageJ software. Statistical values determined by t test. (∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001). (c) Immunofluorescence staining at 20× includes 4 patients with IgA, 3 with DN, and 1 healthy control, showing DAPI (blue), CML (green), and IgA (red), with increased CML accumulation in patients with IgAN and IgA staining surrounding CML deposits. In contrast, patients with DN exhibit minimal CML staining. CML, N(6)-carboxymethyl-lysine; DN, diabetic nephropathy; IgAN, IgA nephropathy; MFI, mean fluorescence intensity.