Divergent effects of low versus high dose anti-TGF-β antibody in puromycin aminonucleoside nephropathy in rats

Background. Transforming growth factor-beta (TGF-beta) modulates immune/inflammatory cells, promotes extracellular matrix (ECM) accumulation, and is increased in fibrotic organs. Here we report the effects of administering a puromycin aminonucleoside nephropathy (PAN)-specific TGF-beta neutralizing antibody on glomerulosclerosis in vivo. Methods. Adult male Sprague-Dawley rats underwent uninephrectomy (Nx) followed by intraperitoneal PAN at weeks 2, 6, 7 and 8. Rats were treated with either high (5 mg/kg body weight) (N = 9) or low (0.5 mg/kg body weight) (N = 7) dose TGF-beta antibody intraperitoneally three times weekly until sacrifice at week 10. A PAN untreated control group (N = 7) was dosed with an isotype specific, null antibody. The nephrectomy samples were studied as normal kidney control (NL) (N = 5). Rats undergoing left kidney Nx (N = 5) only were also included as age-matched control. Renal function and morphology were assessed, and molecular studies performed. Results. Systolic blood pressure was increased in parallel over time in all groups (at 10 weeks, control 137 +/- 10 mm Hg; high 129 +/- 4 mm Hg; low 137 +/- 3 mm Hg) (P = NS). Both TGF-beta antibody treatments decreased renal cortex mRNA expressions similarly for TGF-beta1, TGF-beta2, and collagen III (TGF-beta1, control 0.36 +/- 0.02 mm Hg; high 0.19 +/- 0.01 mm Hg; low 0.19 +/- 0.02 mm Hg; P < 0.01 low and high vs. control; TGF-β2, control 0.38 +/- 0.03 mm Hg; high 0.19 +/- 0.02 mm Hg; low 0.20 +/- 0.03 mm Hg; P < 0.01 low and high vs. control; and collagen III, control 0.33 +/- 0.01 mm Hg; high 0.14 +/- 0.01 mm Hg; low 0.19 +/- 0.01 mm Hg; P < 0.01 low and high vs. control; P < 0.05 low vs. high, data expressed as mRNA normalized density units vs. 18S RNA). However, only low dose TGF-beta antibody improved renal function and sclerosis measured by serum creatinine and creatinine clearance (serum creatinine, control 2.3 +/- 0.5 mg/dL; high 2.5 +/- 0.5 mg/dL; low 0.8 +/- 0.1 mg/dL; P < 0.05 low vs. control and high; creatinine clearance, control 0.44 +/- 0.11 mL/min; high 0.70 +/- 0.26 mL/min; low 1.34 +/- 0.30 mL/min; P < 0.05 low vs. control, P = NS vs. high). In parallel, sclerosis index (0 to 4 + scale) was improved in low dose (control 2.67 +/- 0.27; high 2.37 +/- 0.30; low 1.78 +/- 0.24; P < 0.05 low vs. control). This improved function and structure was linked to decreased glomerular infiltrating macrophages (0 to 4 + score, control 2.3 +/- 0.2; high 1.8 +/- 0.4; low 0.8 +/- 0.1; P < 0.01 low vs. control; P < 0.05 low vs. high; P = NS high vs. control). Further, plasminogen activator inhibitor-1 (PAI-1) mRNA expression in renal cortex was attenuated after low dose TGF-β antibody treatment compared to control and high dose group (PAI-1/glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA ratio, NL 0.18 +/- 0.003; control 0.45 +/- 0.03; high 0.40 +/- 0.04; low 0.23 +/- 0.01; P < 0.05 low vs. control and high). Matrix metalloproteinase-9 (MMP-9) activity was maintained at higher levels in kidneys of the low dose TGF-beta antibody-treated group. Conclusion. These results show an in vivo dose-response with an agent that blocks the biologic activity of TGF-beta. Higher dose of TGF-beta antibody was without beneficial effect, suggesting that TGF-beta-mediated effects on PAI-1 and macrophage influx are bimodal and closely regulated. Given that both antibody doses reduced the expression of TGF-beta isoforms and collagen III production, but only low dose ameliorated histologic sclerosis, it appears that pharmacologic effects of anti-TGF-beta antibody on matrix synthesis and degradation are not equivalent.