Proteinuria in diabetic kidney disease: a mechanistic viewpoint.

Proteinuria is the hallmark of diabetic kidney disease (DKD) and is an independent risk factor for both renal disease progression, and cardiovascular disease. Although the characteristic pathological changes in DKD include thickening of the glomerular basement membrane and mesangial expansion, these changes per se do not readily explain how patients develop proteinuria. Recent advances in podocyte and glomerular endothelial cell biology have shifted our focus to also include these cells of the glomerular filtration barrier in the development of proteinuria in DKD. This review describes the pathophysiological mechanisms at a cellular level which explain why patients with DKD develop proteinuria.

Accelerated apoptosis characterizes cyclosporine-associated interstitial fibrosis.

Recently we developed a model of cyclosporine nephropathy in rats characterized by tubulointerstitial (TI) injury, macrophage infiltration, and progressive interstitial fibrosis [1, 2]. To determine if the TI injury accompanying cyclosporine A (CsA) nephropathy was associated with accelerated apoptosis and ischemia, we treated rats for five weeks with CsA with or without losartan (to block angiotensin II type 1 receptor), or hydralazine/furosemide (H/F) (protocol #1). In protocol #2, rats received CsA with or without L-NAME (to block nitric oxide) or L-arginine (to provide a precursor to nitric oxide formation). Cyclosporine A treated rats had increased apoptosis of tubular and interstitial cells documented by PAS, propidium iodide staining, TUNEL assay, and electron microscopy compared to vehicle treated controls. Macrophages containing apoptotic cells could be confirmed by TUNEL/ED-1 doublestaining and colocalized in areas of TI injury. Animals treated with CsA + losartan had a statistically significant decrease in apoptosis (TUNEL + cells/mm2) when compared to CsA treated animals (6.0 vs. 19.9, P < or = 0.0001). The decrease in apoptosis in the CsA + H/F group was not statistically significant. Animals treated with CsA + L-NAME had a statistically significant increase in apoptosis compared to the CsA treated animals (12.3 vs. 6.4, P = 0.001). L-arginine administration with CsA resulted in a decrease in tubulointerstitial apoptosis versus CsA treated animals, however, this did not reach statistical significance. The addition of L-arginine did result in a significant reduction in interstitial fibrosis (P < 0.0001). Regression analysis revealed a significant correlation between apoptosis and interstitial fibrosis in both protocols. (CsA vs. CsA + losartan r = 0.63, P < 0.0001; CsA vs. CsA + L-NAME r = 0.83, P < 0.0001). We conclude that CsA nephropathy is associated with a marked increase in apoptosis of tubular and interstitial cells. Cyclosporine A induced apoptosis is partially mediated by angiotensin II and nitric oxide inhibition, suggesting a role for renal ischemia in this process, and CsA induced apoptosis correlates with interstitial fibrosis.

Thrombospondin 1 precedes and predicts the development of tubulointerstitial fibrosis in glomerular disease in the rat.

Tubulointerstitial fibrosis is one of the most important histologic features that predicts progression in kidney disease. Thrombospondin 1 is an extracellular matrix protein that can activate latent TGF-beta, a cytokine implicated in the pathogenesis of tubulointerstitial fibrosis. We examined the expression of thrombospondin 1 in several animal models of glomerulonephritis (anti-Thy1 model, aminonucleoside nephrosis, passive Heymann nephritis) that are associated with tubulointerstitial disease. Thrombospondin 1 mRNA and protein were transiently increased in tubular cells, myofibroblasts and some macrophages in areas of tubulointerstitial injury. Thrombospondin 1 expression always preceded the development of tubulointerstitial fibrosis, and correlated quantitatively and spatially with the later development of interstitial fibrosis. Thrombospondin 1 expression predicted the severity of tubulointerstitial fibrosis better than the degree of macrophage or myofibroblast accumulation. Thrombospondin 1 expression was associated with increased expression and activation of TGF-beta1 and decreased expression of LAP-TGF-beta in areas of tubulointerstitial injury. We conclude that thrombospondin 1 is an early marker predicting the development of tubulointerstitial kidney disease. De novo expression of thrombospondin 1 is associated and colocalized with increased expression of TGF-beta1 and decreased expression of LAP-TGF-beta during the development of tubulointerstitial disease in vivo. These data are consistent with the possibility that thrombospondin 1 may be an endogenous activator of TGF-beta.

Osteopontin in chronic puromycin aminonucleoside nephrosis.

Increased expression of osteopontin (OPN) associated with interstitial monocyte infiltration has been demonstrated in the early phase of a variety of experimental renal diseases. Whether these changes occur in the chronic phase of progressive glomerular disease is unknown. Chronic puromycin aminonucleoside nephrosis (PAN) was induced in 16 rats by the injection of a single bolus of PA into the internal jugular vein, which results in a triphasic disease characterized by minimal glomerular change and marked proteinuria, peaking at about 10 to 14 d and subsiding by 28 d, followed by a quiescent 4-wk period of no or minimal proteinuria and then the development of progressive focal glomerulosclerosis (FGS) and increasing proteinuria. Fifteen rats injected similarly with normal saline served as controls. At 11 d after injection, PA rats demonstrated significantly greater urinary protein excretion (P = 0.0107), cortical tubular OPN expression (P = 0.0086), and intraglomerular (P = 0.0009) and interstitial (P = 0.0212) monocyte infiltration than did the controls. At 42 d, no significant differences between the two groups with respect to the above parameters were detected. At 98 d, PA rats had FGS and showed a definite trend to increased proteinuria, cortical tubular OPN, and intraglomerular monocyte infiltration. Although the cortical interstitial monocyte count was not elevated in PA rats compared with controls, there were significantly more monocytes around OPN-positive cortical tubules than around OPN-negative ones (P = 0.0011). Cortical tubular OPN expression correlated well with urinary protein excretion (r = 0.932, P < 0.0001), cortical tubular proliferating cell nuclear antigen (r = 0.796, P < 0.0001), and intraglomerular monocyte count (r = 0.552, P = 0.0013). The results are consistent with a monocyte chemoattractant role for OPN and suggest that OPN is upregulated in the chronic phase of PAN and that this increase in expression is a result of glomerular events.

Osteopontin regulation in cultured rat renal epithelial cells.

Osteopontin is a secreted, arginine-glycine-aspartate (RGD)-containing phosphoprotein that is up-regulated in kidney cortical tubular epithelial cells in many experimental models of tubulointerstitial fibrosis. Its close association with infiltrating macrophage in this disease and its ability to directly stimulate macrophage migration has made it a key target as a molecule likely to be important in mediating renal inflammation. The mechanism responsible for osteopontin up-regulation in kidney disease is unknown, but may involve induction by specific cytokines released by damaged glomeruli or other parts of the kidney, prior to the onset of interstitial disease. We have investigated this hypothesis by testing the effects of angiotensin II, bFGF, TGF beta 1, EGF, and IGF, important renal cytokines, on osteopontin regulation in cultured NRK52E cells, a rat renal epithelial cell line. Using Northern blot, Western blot, and ELISA analyses, we find that NRK52E cells constitutively express low levels of osteopontin mRNA and protein. TGF beta 1 and EGF are potent inducers of osteopontin mRNA and protein in those cells. mRNA stability and nuclear run on assays suggest that induction of osteopontin expression by TGF beta 1 and EGF is via increased transcription of the osteopontin gene. In contrast, IGF-1, angiotensin II, and PDGF BB did not significantly modulate osteopontin expression in NRK52E cells. These studies are consistent with the hypothesis that release of potent cytokines by the injured kidney might be one mechanism whereby elevated levels of osteopontin are synthesized by cortical tubular epithelial cells early in tubulointerstitial disease.

SPARC is expressed in renal interstitial fibrosis and in renal vascular injury.

Tubulointerstitial inflammation and fibrosis are critical determinants for renal function and prognosis in a variety of human nephropathies. Yet, the pathophysiology of the injury remains obscure. We investigated the expression of SPARC (secreted protein acidic and rich in cysteine) by immunohistochemistry and in situ hybridization in experimental models characterized by tubulointerstitial fibrosis and matrix expansion in rats. SPARC is a secreted glycoprotein that has been demonstrated to affect cellular interaction with matrix proteins, modulate cell proliferation, bind to and/or inhibit growth factors such as PDGF and bFGF, and regulate angiogenesis. Interstitial expression of SPARC was most prominent in passive Heyman nephritis (PHN), chronic cyclosporine A (CsA) nephropathy, and the remnant kidney model and, to a lesser extent, in angiotensin II (Ang II)-infused animals. SPARC protein and mRNA were substantially increased at sites of tubulointerstitial fibrosis/matrix expansion. In the PHN model, SPARC protein was expressed by interstitial fibroblasts that also produced alpha-smooth muscle actin ("myofibroblasts") and correlated both temporally (r = 0.97) and spatially with sites of type I collagen deposition. Interstitial cell proliferation preceded the development of interstitial fibrosis, and maximal SPARC expression (d15) coincided with the initial decline in interstitial proliferation. In the Ang II-infusion model, which is characterized by arteriolopathy and tubulointerstitial injury, an increase in SPARC protein and mRNA was also seen in injured blood vessels. SPARC was shown to be expressed by vascular smooth muscle cells and also by cells in the adventitia of hypertrophied arteries. In summary, SPARC was transiently expressed by interstitial fibroblasts at sites of tubulointerstitial injury and fibrosis, and by smooth muscle cells and cells in the adventitia of injured arteries in the Ang II-model. In addition to its proposed role in extracellular matrix deposition. the antiproliferative properties of SPARC might contribute to the resolution of interstitial fibroblast proliferation in the PHN model.

Changes in cell-cycle protein expression during experimental mesangial proliferative glomerulonephritis.

A characteristic response to mesangial cell injury is proliferation, which is closely linked to mesangial matrix accumulation and the progression of glomerular disease. Cell proliferation in non-renal cells in vitro is regulated at the level of the cell-cycle by specific cyclins and their catalytic partners, cyclin dependent kinases (CDK). Cyclin kinase inhibitors (CKI) prevent proliferation by inhibiting cell-cycle progression. However, the expression of cell-cycle regulatory proteins in the kidney and in renal disease is unknown. To determine this we studied the expression of cell-cycle proteins in vivo in normal rats and rats with experimental mesangial proliferative glomerulonephritis (Thy1 model). Normal quiescent rat glomeruli have a differential expression for CKI's, where p27Kip1 is highly expressed, and the levels for p21 (Cip1, Waf1, Sdi1, Cap20) (p21) are low. The onset of mesangial cell proliferation in Thy1 glomerulonephritis is associated with a reduction in p27Kip1 levels when mesangial cell proliferation is maximal. Mesangial cell proliferation in vivo is also associated with an increase in glomerular expression of cyclin A, and an increase in expression and activity for CDK2. The resolution of mesangial cell proliferation was associated with a return to baseline levels for p27Kip1, while the expression for p21 increased substantially. Furthermore, mesangial cell p21 expression was maintained following the resolution of proliferation. These results provide evidence for a complex interplay of cell-cycle regulatory proteins during the glomerular response to injury in vivo. The marked increase in CDK2 expression during mesangial cell proliferation and the sustained increase in p21 expression following the resolution of mesangial cell proliferation suggests that the in vivo expression of certain cell-cycle proteins may differ from that described in non-renal cells in vitro.

Differential expression of transforming growth factor-beta isoforms and receptors in experimental membranous nephropathy.

In membranous nephropathy (MN) overproduction of matrix by glomerular epithelial cells (GEC) is believed to be responsible for glomerular basement membrane thickening and spikes. We studied experimental MN in rats (passive Heymann nephritis, PHN) at 5, 10 and 30 days. PHN rats exhibited a marked increase in GEC immunostaining for TGF-beta 2 at all time points. TGF-beta 3 staining was increased at day 10 only, and TGF-beta 1 was unchanged. Glomerular mRNA for TGF-beta 2 and -beta 3 was increased by day 5 when urine protein increased, whereas TGF-beta 1 was not. TGF-beta 2 bioactivity was increased at day 5. There was also a marked increase in GEC immunostaining for TGF-beta receptor type I (T beta IR) and TGF-beta receptor type II (T beta IIR) at all time points in PHN. mRNA levels for both receptors increased at day 5. Increases in protein expression and mRNA levels for the TGF-beta 2 and -beta 3 isoforms, and T beta IR and T beta RII were prevented by complement depletion. We conclude that complement-mediated injury to the GEC in vivo is associated with the up-regulation of TGF-beta 2 and -beta 3 isoforms, an increase in TGF-beta 2 bioactivity, and an increase in T beta RI and T beta RII expression. This contrasts with changes in TGF-beta 1 reported in mesangial disease, suggesting that TGF-beta 2 and -beta 3 may be important in diseases of the GEC. The differential expression of TGF-beta isoforms and receptors may be important determinants of the GEC response to injury.

SPARC is expressed by mesangial cells in experimental mesangial proliferative nephritis and inhibits platelet-derived-growth-factor-medicated mesangial cell proliferation in vitro.

Mesangial cell proliferation is a characteristic feature of many glomerular diseases and often precedes extracellular matrix expansion and glomerulosclerosis. This study provides the first evidence that SPARC (secreted protein acidic and rich in cysteine) could be an endogenous factor mediating resolution of experimental mesangial proliferative nephritis in the rat. SPARC is a platelet-derived-growth-factor-binding glycoprotein that inhibits proliferation of endothelial cells and fibroblasts. We now show that SPARC is synthesized by mesangial cells in culture and that SPARC mRNA levels are increased by platelet-derived growth factor and basic fibroblast growth factor. Recombinant SPARC or the synthetic SPARC peptide 2.1 inhibited platelet-derived-growth-factor-induced mesangial cell DNA synthesis in vitro. In a model of experimental mesangioproliferative glomerulonephritis, SPARC mRNA was increased 5-fold by day 7 and was identified in the mesangium by in situ hybridization. Similarly, SPARC was increased in glomerular mesangial cells and visceral epithelial cells by day 5 and reached maximal expression levels by day 7. Mesangial cell proliferation increased by 36-fold on day 5 and decreased abruptly on day 7. Maximal expression of SPARC was correlated with the resolution of mesangial cell proliferation. We propose that SPARC functions in part as an endogenous inhibitor of platelet-derived-growth-factor-mediated mesangial cell proliferation in glomerulonephritis and that it could account for the resolution of cellular proliferation in this disease.

Pathogenesis of cyclosporine nephropathy: roles of angiotensin II and osteopontin.

Low-salt-diet, cyclosporine (CsA; 15 mg/kg per day)-treated rats develop striped interstitial fibrosis, arteriolar hyalinosis, and azotemia similar to the chronic nephropathy observed in humans. To examine the role of angiotensin II in this model, rats on a low-salt diet were given CsA, CsA and the angiotensin II receptor Type I antagonist Losartan (10 mg/kg per day), CsA and hydralazine/furosemide, or vehicle. At Day 35, CsA-treated rats had tubular injury, arteriolopathy of the afferent arteriole, increased expression of the monocyte-macrophage adhesive protein osteopontin, interstitial macrophage infiltration, increased interstitial transforming growth factor-beta expression, and interstitial fibrosis. This study provides new insight in both pathogenic and therapeutic aspects of CsA nephropathy. The pathogenesis of CsA nephropathy involves the expression of osteopontin by tubular epithelial cells, the level of which closely correlates with the degree of macrophage infiltration and interstitial fibrosis in all groups (r = 0.79 and 0.74, respectively; P < 0.001). Therapeutic conclusions can be drawn from the observation that both losartan and hydralazine/furosemide reduced osteopontin expression, macrophage infiltration, transforming growth factor-beta expression, and interstitial fibrosis, but did not prevent the decrease in GFR. Treatment with losartan, but not with hydralazine and furosemide, markedly reduced arteriolopathy. It was concluded that angiotensin II contributes to the vasculopathy (hyalinosis) induced by CsA. In contrast, the interstitial fibrosis mediated by CsA can be partially prevented by both an angiotensin II Type I receptor antagonist or by hydralazine and furosemide. This suggests that the interstitial fibrosis can be dissociated from the vascular effects of CsA. The beneficial effects of lowering blood pressure or vasodilation per se may be difficult to distinguish from the specific effects of angiotensin II receptor blockade.