Cathepsin S–Dependent Protease–Activated Receptor-2 Activation: A New Mechanism of Endothelial Dysfunction

Endothelial dysfunction plays an important role in the development of diabetic nephropathy.¹ Although the loss of endothelial nitric oxide synthase is well recognized as a cause of hypertension and endothelial dysfunction, other mechanisms are likely to contribute to endothelial damage. One such possible mechanism is endothelial damage caused by infiltrating macrophages. Although macrophages play a role in the induction and progression of human and experimental diabetic nephropathy,^2–4 there is no evidence that macrophage-derived factors are a cause of endothelial damage. A study by Kumar et al.⁵ in this issue of the Journal of the American Society of Nephrology has established a new pathologic mechanism in diabetic nephropathy in which macrophage–derived cathepsin S activates protease–activated receptor-2 (PAR-2) on endothelial cells to cause substantial endothelial damage, giving rise to vascular albumin leakage, inflammation, and glomerulosclerosis.

Cathepsins are a family of cysteine proteases. One member of this family, cathepsin S, has been implicated in the pathogenesis of a range of disease states.⁶ Cathepsin S contributes to protein degradation in the endosomal/lysosomal pathway, including proteolytic degradation of the li chaperone protein, which allows peptide antigens to bind to MHC class 2 molecules and be presented to T lymphocytes. In addition to its well defined role in the adaptive immune response, cathepsin S can also be secreted. Cathepsin S enzymatic activity operates across a relatively wide range of pH values, which enables activity in acidic endosomes as well as the more neutral conditions of the extracellular space, in which it can cleave substrates, such as elastin, E-cadherin, secretory leukoprotease inhibitor, junctional adhesion molecule-B, and PAR-2.⁶ One distinct feature of cathepsin S is that its expression is largely restricted to leukocyte subsets, particularly macrophages, although cathepsin S expression can be induced in a variety of nonleukocyte cell types. Experimental studies using cathepsin S gene–deficient mice or cathepsin S inhibitors have defined a role for this enzyme in autoimmune diseases, atherosclerosis, airway hyper-responsiveness, cancer metastasis, neovascularization, and neuropathic pain.⁶ In addition, elevated serum levels of cathepsin S have been described in a number of diseases and correlate with insulin resistance, diabetes, atherosclerosis, and heart disease.^7,8

The study by Kumar et al.⁵ proposes that macrophage–derived cathepsin S induces endothelial cell damage by activation of PAR-2 on the endothelial cell surface. This is on the basis of several key findings. Injection of recombinant cathepsin S into normal mice induced marked endothelial damage, resulting in vascular leakage and albuminuria; however, Par-2 gene–deficient mice were protected from these injurious effects, and studies of cultured endothelial cells showed that cathepsin S acts through PAR-2 and not through other members of the PAR family. A key finding was that treatment with a selective cathepsin S inhibitor, RO5461111, in established type 2 diabetic nephropathy in uninephrectomized db/db mice reduced endothelial damage, albuminuria, and glomerulosclerosis as well as albumin leakage in the retina.⁵ These beneficial effects were associated with protection from endothelial cell damage and loss and protection from podocyte damage and loss as well as a reduction in macrophage infiltration and markers of inflammation without any apparent effect on body weight or blood glucose levels. Importantly, administration of a PAR-2 inhibitor, GB83, gave parallel findings with reduced endothelial cell loss and glomerulosclerosis, providing the first direct evidence that PAR-2 plays a pathogenic role in diabetic nephropathy. Interestingly, combined drug treatment showed no added benefit, implying that cathepsin S and PAR-2 activations operate through the same mechanism to cause diabetic kidney injury. Finally, CD68+ macrophages were identified as the main source of cathepsin S synthesis in both human and experimental diabetic nephropathy.⁵

Studies that make a substantial contribution to a field inevitably raise additional questions. For example, does cathepsin S promote diabetic renal injury through actions on other cell types? This is a difficult issue to address, because studies of PAR-2 expression in the kidney have been plagued by problems of a lack of specificity of antibodies using for immunohistochemistry. Tubular epithelial cells, mesangial cells, and podocytes all express PAR-2 in culture, and activation of PAR-2 in these cell types induces a variety of proinflammatory and profibrotic responses, including CCL2 and TGF-β1 production.⁹ It is also known that the kidney expresses relatively high levels of PAR-2 mRNA, suggesting that PAR-2 may be widely expressed in the kidney; however, defining which renal cell types do, in fact, express PAR-2 in situ remains unresolved. It may require conditional Par-2 gene deletion in endothelial cells to formally address this question.

Another question raised by the study by Kumar et al.⁵ relates to the source of cathepsin S that causes PAR-2 activation and consequent endothelial cell damage in the diabetic kidney. Although macrophages were identified as the main source of local cathepsin S production within the diabetic kidney, serum cathepsin S levels are increased in many diseases, including diabetes.⁸ Thus, damage to the endothelium could be mediated through circulating cathepsin S, the source of which is unclear. Although serum levels of cathepsin S can be elevated in a range of diseases, there is lack of data on the actual level of cathepsin S enzymatic activity in these conditions. This is a particularly important issue in CKD, in which there are increased blood levels of the cathepsin inhibitor, cystatin C. These data provide impetus for measuring circulating levels of cathepsin S activity as both a biomarker and a pathogenic mechanism in progressive kidney disease.

Is this mechanism of cathepsin S/PAR-2–induced endothelial damage specific to diabetic renal injury, or is it of general importance in progressive kidney disease? We lack data on cathepsin S expression in other forms of kidney disease. The only well characterized inducer of cathepsin S synthesis in macrophages is IFN-γ,¹⁰ suggesting production by M1–type polarized proinflammatory macrophages. This is consistent with the presence of macrophages expressing the archetypal M1–type marker, inducible nitric oxide synthase, within the diabetic kidney.¹¹ Indeed, M1–type proinflammatory macrophages are a common feature in biopsies from patients with kidney disease, especially rapidly progressive forms of disease,¹² suggesting that macrophage–derived cathepsin S may be a common mechanism of renal injury.

One implication of the study by Kumar et al.⁵ is that we have two new therapeutic targets for diabetic nephropathy: cathepsin S and PAR-2. However, which should we target given the lack of added benefit with combined blockade over that of individual blockade? There are several pharmaceutical companies that have cathepsin S inhibitors in clinical trials in areas such as neuropathic pain, rheumatoid arthritis, psoriasis, and osteoporosis.⁶ In addition, PAR-2 inhibitors are in preclinical studies.¹³ In the case of diabetic nephropathy, cathepsin S may be the more attractive target, because cathepsin S has also been implicated in related comorbidities, such as insulin resistance, atherosclerosis, and aortic stiffening.⁶ However, it may also be the case that combined cathepsin S and PAR-2 inhibition will provide added benefit in other types of kidney diseases. For example, experimental models of rapidly progressive GN have identified a role for cathepsin S in the adaptive immune response leading to lupus nephritis,¹⁴ whereas PAR-2 promotes glomerular thrombosis in antiglomerular basement membrane disease.¹⁵ In addition, Par-2 gene–deficient mice exhibit reduced interstitial fibrosis in the unilateral ureteric obstruction model.¹⁶

In conclusion, this exciting study has identified a specific approach to targeting endothelial dysfunction in diabetic nephropathy. These findings may also be highly relevant in nondiabetic kidney disease.

Disclosures

None.