2023 Volume 70 Issue 8 Pages 771-776
Diabetic nephropathy is a public health problem worldwide. Our understanding of the molecular machinery, as well as the clinical therapies for diabetic nephropathy, has evolved dramatically in recent years. However, even with this progress, there are residual risks of kidney failure and cardiovascular events in patients with diabetes. Rho-associated, coiled-coil-containing protein kinase (ROCK) is activated in response to various pathologic stimuli in the context of diabetes. The contribution of ROCK has been investigated in vivo using gene deletion rodent models and specific inhibitors, which are providing key insights into the pathologic function of ROCK in diabetic nephropathy. ROCK has two isoforms, ROCK1 and ROCK2. Both isoforms are expressed in the kidney, including mesangial cells, podocytes, and endothelial cells. ROCK1 blunts AMP-activated protein kinase (AMPK), while ROCK2 negatively regulates peroxisome proliferator-activated receptor α (PPARα) to inhibit fatty acid oxidation, both of which lead to structural and functional impairment of glomeruli in diabetes. Of note, ROCK signaling is activated in the kidney of animal models and patients with diabetes. In addition, an observational study has shown that fasudil hydrochloride, an ATP-competitive selective ROCK inhibitor, significantly attenuated proteinuria among patients with diabetes. These findings highlight the promising prospects for the development of a ROCK-centered approach against the progression of diabetic nephropathy.
Diabetic nephropathy is a challenging public health issue due to its growing prevalence, as well as the extensive morbidity it causes. Moreover, the global impact and costs of dialysis are expected to increase substantially, affecting the most disadvantaged sectors of the population [1, 2]. The denouement of insulin resistance is progressive kidney injury, which is characterized by mesangial expansion, podocyte loss, and thickening of the glomerular basement membrane, culminating in structural and functional derangements. Numerous studies have investigated the activated renin-angiotensin system and glomerular hemodynamic changes as possible mediators that link to kidney function decline. Indeed, the administration of renin-angiotensin system inhibitors and sodium-coupled glucose transporter 2 inhibitors is the cornerstone of the current therapy to reduce the risk of kidney failure in patients with diabetes [3, 4]. Even with these agents, however, real-world data show a high residual risk of kidney death among people with diabetes [5]. As such, a detailed understanding of the molecular underpinnings of disease pathogenesis is critically important for designing specific and efficacious forms of therapy.
Early pioneering work identified dyslipidemia as a risk factor for abnormal kidney function [3]. It was later shown that potential contributing mechanisms of lipid nephrotoxicity include the deposition of cholesterol into kidney tissue, and an increase of oxidative stress derived from mesangial cells transformed into foam cells by uptake of oxidized low-density lipoprotein. Indeed, findings from meta-analysis revealed that statins could reduce both albuminuria and the progression of diabetic nephropathy [3]. This renoprotective action of statin is partially explained by the reduced low-density lipoprotein levels and subsequent lipotoxicity. Of note, we have previously shown that the administration of pravastatin significantly blunts the increase of urinary albumin excretion and oxidative stress in rodent models of diabetes. Importantly, the effects of statin on cholesterol levels differ depending on the species and it has been demonstrated that administration of statin to rodents (i.e., mice, rats) does not modify serum cholesterol levels [6]. Given the difference in cholesterol metabolism between species, it is reasonable to suggest that cholesterol-independent pathways are likely operative, and that the elucidation of such pathways may provide the foundation for novel therapeutic approaches.
Statins inhibit the conversion of HMG-CoA to mevalonate, resulting in decreased intracellular concentrations of isoprenoids (i.e., farnesyl diphosphate and geranylgeranyl diphosphate) to attenuate the intracellular activity of the small G protein Rho. This mechanical concept is the necessary framework for understanding the cholesterol-independent effects of statins. Rho belongs to the Ras superfamily along with Ras and Rab, which are stimulated by cytokines, growth factors, and glucose. Inactive Rho exists in the cytoplasm in a guanosine diphosphate-bound form. However, when Rho is modified with isoprenoids, it binds to the cell membrane and is converted to an active form bound to guanosine triphosphate. Rho-associated, coiled-coil-containing protein kinase (ROCK) is a serine-threonine kinase that is regulated by guanosine triphosphate-bound Rho. The Rho/ROCK signaling pathway has recently been the subject of intense investigation in human health and disease. Studies have shown that various cellular functions such as cell contraction, migration, and gene expression are orchestrated through a mechanism involving ROCK modification [7]. Moreover, in investigations highlighting the devastating toll of ROCK activation on human health, ROCK is becoming increasingly recognized as a significant component of pathology in cancer, atherosclerosis, and kidney disease [8]. Several contributing mechanisms have been proposed for ROCK pathogenesis, including angiogenesis, vascular inflammation, and fibrotic reactions. For example, ROCK induces gene expression of platelet-derived growth factor, transforming growth factor β, and extracellular matrix production, which are known to accelerate the progression of glomerulosclerosis. As stated above, statins suppress the activation of Rho/ROCK signals through the modulation of isoprenoids, thereby inhibiting this process. However, statins have broad actions on small G proteins other than Rho (i.e., Cdc42, Rac1). Gene targeting approaches have shown the essential roles of GTPases in maintaining the normal structure of the glomerular filtration barrier. For example, podocyte-specific ablation of Cdc42 leads to massive proteinuria and kidney failure, resulting in death within two weeks after birth [9]. With respect to Rac1, glomerulosclerosis is accelerated by the salt load or adriamycin administration in Rac1-deficient mice [10], indicating that Rac1 plays protective roles for the glomerular epithelium. Unlike statins, these small GTPases are not modulated by ROCK inhibition (Fig. 1). Treatment targeting ROCK may therefore be more effective and safer when the primary goal is to suppress the initiation or progression of kidney disease. Importantly, we and others have previously demonstrated that ROCK is involved not only in nephropathy but also in the pathogenesis of retinopathy, neuropathy, and atherosclerosis. For instance, retinal microvascular damage and neovascularization are prevented by ROCK inhibition [11, 12]. Moreover, ROCK inhibitor attenuates diabetes-induced suppression of motor nerve conduction velocity through the regulation of E-cadherin distribution [13]. The involvement of ROCK in vascular inflammation and remodeling has also been reported in recent studies [14, 15]. The published evidence to date suggests that ROCK is a promising disease-modifying target for the comprehensive management of diabetic vascular complications.
Small GTPases in kidney disease
The Rho/ROCK signaling pathway has been implicated in a variety of proteinuric kidney disease. On the other hand, gene deletion studies have revealed the indispensable roles of Cdc42 and Rac1 for maintaining normal kidney function. It is necessary to avoid eliminating these small GTPases from therapeutic targets in the management of kidney disease.
In an effort to better understand the role of ROCK signaling in diabetic nephropathy, a series of animal experiments were performed. Intraperitoneal administration of streptozotocin to wild-type mice induces kidney damage, resulting in increased urinary albumin excretion and glomerulosclerosis. Immunostaining of myosin phosphatase targeting subunit 1 (MYPT1), a substrate of ROCK, revealed an increase in phosphorylated MYPT1 in the kidney of streptozotocin-injected rat, suggesting activation of ROCK signaling in diabetes [16]. These findings were recapitulated in high-fat diet-fed mice and db/db mice [17]. In addition, the administration of fasudil hydrochloride, an ATP-competitive selective ROCK inhibitor, significantly attenuates glomerular histological changes and urinary albumin excretion in type 1 and type 2 diabetes models.
The above studies suggested that ROCK activation is a fundamental process in the development of nephropathy regardless of the type of diabetes. However, the supporting evidence for the benefits of ROCK blockade was based mainly on animal studies. To confirm these concepts in patients, we extended our investigation to human samples and applied observational study. In support of the obtained experimental findings, kidney ROCK activation was detected among patients with diabetes [18]. In postmortem kidney tissues from patients with diabetic nephropathy, the phosphorylated form of MYPT1 was significantly increased in the glomeruli and tubular interstitium. As these results indicate the importance of ROCK in human diabetic nephropathy, we next conducted a small-scale observational study to verify the renoprotective effect of ROCK inhibitors in patients with diabetes. Fasudil hydrochloride is clinically approved for up to 14 days to prevent vasospasm after cerebral hemorrhage. We extracted the clinical information of 15 patients who received fasudil hydrochloride from among 23,241 patients with diabetes who visited our university hospital, and then analyzed their kidney parameters. Among patients receiving fasudil hydrochloride, urinary albumin excretion was significantly reduced after treatment independent of blood pressure or blood glucose levels. Taken together, these results support the notion that ROCK is an important therapeutic target against the progression of diabetic nephropathy.
Experimental and clinical studies have shown that ROCK activation is an essential factor that exacerbates abnormal kidney structure and function. To better understand the mechanism underlying the action of ROCK, we focused on the hypoxic environment of the diabetic kidney. The published evidence to date suggests that diabetic kidneys are in a hypoxic state due to renin-angiotensin system-mediated vasoconstriction and increased oxygen consumption associated with sodium reabsorption in tubules [19]. This is in agreement with reports describing the expression of hypoxia-inducible factor 1α (HIF-1α) in the kidney of rodent models and patients with diabetes [20]. Multiple lines of evidence strongly implicate HIF-1α as an important regulator of fibrosis in health and disease, such as wound healing and kidney fibrosis. Interestingly, pharmacological inhibition of ROCK significantly attenuates glomerular HIF-1α activation and protects against glomerular sclerosis in mice [17].
These results prompted us to analyze the precise mechanism of ROCK-mediated HIF-1α regulation. The expression level of HIF-1α is regulated at the transcriptional, translational, and post-translational levels. Under aerobic conditions, HIF-1α is hydroxylated by proline hydroxylase, which interacts with the von Hippel-Lindau protein that leads to polyubiquitination and proteasomal degradation. On the other hand, under anaerobic conditions, stabilized HIF-1α translocates to the nucleus, forms a dimer with the constitutively expressed HIF-1β, to become transcriptionally active. We have demonstrated that ROCK negatively regulates glomerular HIF degradation through the inhibition of prolyl hydroxylase expression, which promotes HIF-1 stabilization and subsequent gene transcription of profibrotic factors such as connective tissue growth factor and plasminogen activator inhibitor 1. However, the effects of HIF activation on the kidney may differ depending on the cell type, primary disease, and disease stage, and therefore careful accumulation of data is required. Further studies in our laboratory are underway to dissect the detailed function and regulation of ROCK in kidney tubules and interstitium.
Several studies provide compelling evidence supporting the importance of inflammation in diabetic nephropathy. In diabetic kidneys, induction of cell adhesion molecules such as intercellular adhesion molecule 1 (ICAM-1) and macrophage infiltration are observed [21]. The most cogent evidence implicating ICAM-1 in the pathogenesis of diabetic nephropathy is derived from loss-of-function experiments [22]. In contrast to wild type mice, ICAM-1 deficient mice exhibit a marked reduction in albuminuria and histological abnormalities. In addition, chemokines such as monocyte chemoattractant protein 1 and macrophage colony-stimulating factor promote tissue injury by the infiltration of inflammatory cells into the kidney. Of note, ROCK positively regulates the expression of these chemokines in the glomerulus to promote monocyte migration and macrophage proliferation [23, 24]. Among the processes of chemokine production associated with inflammation, the main cellular machinery for gene regulation is the nuclear factor κB (NF-κB) signaling pathway. NF-κB forms a complex with inhibitor IκB and exists in the cytoplasm, but when IκB is degraded by extracellular stimulation, it translocates into the nucleus where it exerts transcriptional activity. The mechanistic insight gleaned from mesangial cells provided evidence of ROCK-mediated NF-κB regulation. ROCK is involved in the inflammatory process by promoting the translocation of NF-κB into the nucleus to induce transcription of target genes. When considered alongside previous observations, these findings implicate a novel transcriptional circuitry involving HIF-1α and NF-κB that coordinates glomerular damage.
ROCK has two isoforms, ROCK1 and ROCK2. Both isoforms are expressed in almost all organs. In the kidney of humans and other mammalian species, ROCK1 and ROCK2 are distributed in glomerular mesangial cells, podocytes, tubules, and endothelial cells. Previous studies have revealed important insights regarding the biological significance of ROCK, but are also limited since the role of each isoform remains incompletely understood, representing an active research area.
To shed light on the role of ROCK isoforms, several loss-of-function studies were conducted in mice. With regard to ROCK1, most of the mice harboring systemic deletion of ROCK1 die early postnatally [25]. However, heterozygous mice are viable into adulthood and resulted in the attenuation of urinary albumin excretion caused by streptozotocin [26] or high-fat diet [27]. To gain insight into how ROCK1 modifies kidney function, we profiled gene expression levels in the kidney. Interestingly, the mechanism of ROCK1-mediated renoprotection included the correction of fatty acid metabolism which is governed by AMP-activated protein kinase (AMPK).
The metabolic kidney is an omnivore that utilizes glucose and fatty acids to meet its energy demands for sustained function. Energy homeostasis in regulating kidney disease has garnered significant interest. It has been demonstrated that substrates used for kidney metabolism differ depending on cell type. For instance, it is known that both glucose and fatty acids are used in the glomeruli. On the other hand, fatty acids are the main source of energy production in tubules. In the setting of kidney disease, blockage of glomerular glycolysis by pyruvate kinase M2 deficiency has been shown to exacerbate nephropathy. Moreover, inhibition of fatty acid transport into mitochondria reduces energy production in glomeruli to increase oxidative stress. We have previously demonstrated that ROCK1 is expressed in glomeruli to suppress mitochondrial respiration and enhance oxidative stress in diabetes. Furthermore, ROCK1 has been shown to regulate mitochondrial fission in the state of hyperglycemia [28]. Based on these findings, future studies aimed at elucidating the clinical effects of ROCK1 inhibition, through the generation of ROCK1 inhibitor, will therefore be beneficial.
Since ROCK2 has protein features distinct from those of ROCK1, we next generated podocyte-specific ROCK2-deficient mice to examine its functional relevance [29]. These mice were viable and were born at a normal Mendelian ratio. ROCK2 deficiency restricted to the glomerular epithelium does not cause structural or functional abnormalities in the kidney but prevents the progression of nephropathy in both insulin-deficient and insulin-resistant animal models. From a mechanistic standpoint, gene expression analysis by RNA-sequencing and quantitative PCR (qPCR) revealed that ROCK2 negatively regulates peroxisome proliferator-activated receptor α (PPARα) to inhibit fatty acid utilization. As loss of PPARα is associated with impaired ability to access energy source and impaired tubular protein reabsorption, when taken in the context of our current findings, it is likely that the activation of PPARα by ROCK2 inhibition is nephroprotective. Supporting this model are our recent observations that PPARα induction occurs in mesangial cells treated with ROCK inhibitor [30]. These findings provide a molecular bridge between ROCK isoforms and cellular energy homeostasis in the diabetic kidney (Fig. 2). Given that there are multiple cell types found within the kidney (mesangial cells, podocytes, endothelial cells, tubular cells, immune cells), defining target genes that are both induced and suppressed by ROCK isoforms in these cells will provide valuable insights into the pathological process. The overlapping and divergent function of ROCK isoforms in each cell type is also an important area of investigation.
ROCK-mediated energy metabolism
ROCK1 regulates AMPK to inhibit mitochondrial respiration. Meanwhile, ROCK2 blunts PPARα expression, which leads to impaired fatty acid utilization. The inhibition of both ROCK isoforms may have additional benefits for kidney protection.
ROCK is emerging as a critical regulator of diabetic micro- and macrovascular complications with great pertinence to the treatment of diabetic nephropathy. The experimental findings thus far obtained constitute an important step forward in making the promise of the ROCK-centered approach a reality. ROCK inhibition, taken together with other interventions, can be significant disrupters of the unrelenting progression of diabetic nephropathy which can enable patients to lead long, healthy, and productive lives. This now needs to be tested in clinical trials as part of the effort to combat the ongoing pandemic of diabetes.
I wish to thank all members of our laboratories and our respected colleagues in the ROCK field for critical input and discussions. This work was supported by JSPS KAKENHI grant number 23K07709; the Astellas Foundation for Research on Metabolic Disorders; the Japan Diabetes Foundation; the Mochida Memorial Foundation for Medical and Pharmaceutical Research; and the Ichiro Kanehara Foundation.
The author declares no conflicts of interest associated with this research.
