Caspase-12 and Diabetic Nephropathy: From Mice to Men?

Since the seminal observation that high glucose induces renal cell apoptosis in culture and in vivo,¹ investigators have sought to identify the molecular mechanisms of renal cell apoptosis in diabetic nephropathy. In this issue of JASN, Brezniceanu et al.² explore proapoptotic genes that are upregulated differentially by reactive oxygen species (ROS) in renal proximal tubular cells of diabetic (db/db) mice. Expression of caspase-12 and other endoplasmic reticulum (ER) stress genes, such as HSPA5/GRP78/BiP and CHOP, are increased in the proximal tubules of these mice compared with nondiabetic and diabetic catalase transgenic mice. Reduction of ROS generation also inhibits albumin-stimulated expression and activity of caspase-12 in a human proximal tubule cell line (HK-2). Furthermore, knockdown of caspase-12 with small interfering RNAs reduces albumin-induced apoptosis in HK-2 cells. The authors of this article conclude that albuminuria may induce ROS-mediated ER stress and subsequent tubular apoptosis in diabetic kidneys. The involvement of caspase-12 in this process, especially in human cells, stands out as the most original part of the article; however, no information is provided on human diabetic nephropathy, and the key question is how relevant these data are to humans.

There is increasing evidence that renal cells, including tubular cells, are lost through apoptosis in experimental and human diabetic nephropathy. Recent efforts to identify novel changes in the human transcriptome have also added a host of participants to the process of renal cell apoptosis during diabetic nephropathy.^3,4 These studies have uncovered evidence of the presence of ER stress in renal cells in human diabetic nephropathy.⁵ Stimuli that increase the demand on the ER to synthesize proteins or degrade improperly folded proteins cause this stress.⁶ Several components of the diabetic milieu, such as high glucose, free fatty acids, albumin, oxidative activity, and inflammation, induce ER stress in renal cells.⁶

Cells respond to ER stress by an adaptive response that leads to upregulation of ER chaperone proteins HSPA5/GRP78/BiP and HYOU1/ORP150 and the prosurvival transcription factor XBP1, among others. These genes are upregulated in progressive human diabetic nephropathy.⁵ Despite this evidence of ER stress in human diabetic nephropathy, it is still unclear to what extent this stress contributes to cell loss. In this regard, expression of potentially lethal unfolded protein response genes, such as CHOP, is unchanged or repressed in human diabetic nephropathy.⁶

Caspase-12, a member of the caspase family of intracellular cysteine proteases, was first identified in mice.⁷ Murine caspase-12 is processed during ER stress–induced apoptosis, and caspase-12^−/− mice are protected from tubular injury induced by tunicamycin.⁷ Whereas degradation or proteolysis of caspase-12, as observed in acetaminophen-exposed tubular cells, is a well-established hallmark of ER stress, its central role in ER stress–induced apoptosis has been questioned.^8,9 Contrary to other caspases, caspase-12 proteolytic activity seems to be limited to autocleavage.¹⁰ Thus, caspase-12 may be incapable of efficiently processing cellular substrates and initiating apoptosis. The functional consequences of autocleavage are unknown, but it has been proposed to result in caspase-12 inactivation.¹⁰ Calpains may also process caspase-12. More recently, it has become evident that murine caspase-12 is an inhibitor of inflammation that dampens antibacterial responses by inhibiting caspase-1 and the NOD2/RIP2 pathways.^11,12 In mice, deletion of the gene encoding caspase-12 confers resistance to peritonitis and septic shock by enhancing bacterial clearance.¹¹

A common human single-nucleotide polymorphism (SNP) of caspase-12 results in a premature stop codon yielding a truncated protein that contains only the N-terminal CARD prodomain (caspase-12S). Thus, caspase-12S lacks the large and small subunits of the active mature enzyme, present in the full-length variant molecule.¹³ Truncated caspase-12 is inactive and lacks catalytic and anti-inflammatory activity. The caspase-12S variant is now present in 98% of humans.¹⁴ Only in certain sub-Saharan African populations (28% of individuals) is the full-length caspase-12 polymorphism present. The driving force for the positive selection of caspase-12S is thought to be increased resistance to sepsis, which seems more important in higher density European and Asian populations than in the original African population.^13,14

With this background, Brezniceanu et al.² observed increased expression of caspase-12 co-localization to apoptotic proximal tubular cells in murine diabetic nephropathy. Unexpected, because human HK-2 cells carry the caspase-12–inactivating SNP, the authors also observed the expression and activity of a full-length protein, which was upregulated by high glucose and, as previously reported in rat tubular cells, by albumin.¹⁵ Moreover, small interfering RNA targeting of human caspase-12 protected cells from albumin-induced tubular cell apoptosis.² These are provocative observations that add to the controversy surrounding the role of caspase-12 in cell death and in humans.

There was a discrepancy between caspase-12 activity and caspase-3 activity/apoptosis that depends on the nature of the stimulus and suggests that caspase-12 enzymatic activity is not the sole determinant of apoptosis. In this regard, additional mechanisms for the proapoptotic action of caspase-12, besides the proteolytic activity displayed by other caspases, should be explored. The recent observation that caspase-12 interferes with NF-κB activation and nuclear localization may provide clues for further study.¹²

The SNP encoding full-length caspase-12 is more prevalent in individuals of African descent, and the risk for progression to ESRD from diabetic nephropathy and other proteinuric kidney diseases is higher in black as compared with white individuals. It may be interesting to study the influence of caspase-12 SNPs on the course of proteinuric kidney disease in these populations. A different question is also the relevance of the findings by Brezniceanu et al. for wider human populations that lack full-length caspase-12. It is plausible that the observed expression of full-length caspase-12 in HK-2 cells is the result of the presence of aminoglycosides in the cell culture media. Aminoglycosides promote translational reading through premature stop codons and have been tested in clinical trials of genetic disease caused by single-nucleotide mutations resulting in a premature stop codon.¹⁶

In summary, although there is evidence of a role for caspase-12 and ER stress in albumin-induced tubular cell injury, there are significant differences between the in vivo human and experimental situations that preclude the direct extrapolation of animal model or cell culture results to the clinic at this point.

• Copyright © 2010 by the American Society of Nephrology