Microvascular Endothelial Cells Poised to Take Center Stage in Experimental Renal Fibrosis

Peritubular microvascular insufficiency leading to tubulointerstitial hypoxia, capillary rarefaction, and tubular atrophy has long been considered an important pathomechanism promoting renal fibrosis as a final common pathway in CKD.¹ Hence, molecular mechanisms and consequences of impairment of renal oxygenation have been the focus of intense research interest in both acute and chronic kidney injury, and opportunities for therapeutic targeting of hypoxia to retard or even prevent kidney disease progression have been proposed.² However, insights into how a broad spectrum of renal injuries may cause peritubular microvascular rarefaction and renal hypoxia remain remarkably limited, considering the postulated preeminent role of tissue hypoxia in particular in tubulointerstitial fibrosis and kidney disease progression.

The pivotal role of TGF-β in tissue repair (wound healing) and fibrosis was first characterized over a quarter of a century ago,³ but whether and in what way TGF-β may act on microvascular endothelia in the kidney after development in homeostasis and injury remains poorly understood. This knowledge gap is noteworthy because TGF-β was widely identified soon after this initial report as a ubiquitous culprit in experimental renal fibrosis and human kidney disease, and was labeled the relentlessly profibrotic protagonist of the “dark side of tissue repair.”⁴ However, TGF-β’s fibrogenic activity is only one of many essential biologic activities that this extraordinarily complex and evolutionarily highly conserved cell signaling system mediates by acting on virtually every cell type in development, homeostasis, and injury.

In the kidney, the activities of the TGF-β cytokine signaling system range from mesangial proliferation and matrix expansion, glomerular capillary development and endothelial dysfunction, tubular epithelial cell transition and fibrosis, and podocyte apoptosis and depletion (reviewed in Böttinger⁵). Although a few reports suggest a role for TGF-β in endothelial-mesenchymal transition (EndoMT) in cardiac development⁶ and cardiac fibrosis,⁷ the biologic function(s) and implications of TGF-β signaling in peritubular microvascular endothelia in homeostasis and disease remain poorly understood.

In this issue of JASN, a study by Xavier and colleagues examined whether endothelial TGF-β signaling contributes to renal fibrosis.⁸ Mice with cre/lox-mediated inducible genetic ablation of one copy of the TGF-β receptor type II kinase (TbRII) gene in vascular endothelial cells (TβRII^endo+/−) were viable and showed no phenotypic differences compared with controls. Although TbRII expression levels and downstream signaling were only moderately reduced in endothelial cell culture isolated from TβRII^endo+/− compared with control TβRII^endo+/+, the consequences in vivo were striking. When renal injury was induced by standard folic acid nephropathy or unilateral ureteral obstruction experimental models, pronounced tubulointerstitial fibrosis and collagen deposition characteristic of control TβRII^endo+/+ mice were remarkably reduced in TβRII^endo+/− animals. In a series of elegant experiments, the authors then explored how reduced endothelial TGF-β signaling protected TβRII^endo+/− animals against tubulointerstitial fibrosis. They demonstrated not only one, but two possible mechanisms. First, Cd31/α-smooth muscle actin (αSMA) double-positive blood vessels, a hallmark of EndoMT, were significantly reduced, and this was associated with improved renal perfusion and reduced tissue hypoxia in TβRII^endo+/− compared with TβRII^endo+/+ mice. Second, three-dimensional cultures of aortic rings from TβRII^endo+/− mice manifested impressively increased “sprouting” compared with TβRII^endo+/+, suggesting increased angiogenic potential. This phenomenon was possibly related to the observation that Smad1/Smad5-mediated angiogenic signaling was not impaired compared with antiangiogenic Smad2/Smad3 signaling in TβRII^endo+/− tissue.

Which of those two possible mechanisms best explains the protective effect of limiting endothelial TGF-β signaling in renal injury? The authors do not yet answer this intriguing question; however, mitigation of EndoMT seems immediately plausible because this effect was directly demonstrated in injured kidneys. By contrast, the increased angiogenic potential was only assessed in aortic ring cultures and was not directly demonstrated in injured kidneys. Thus, more work will be needed to clarify whether limiting endothelial TβRII signaling in renal injury models protects against hypoxia and ultimately fibrosis by curbing EndoMT and associated microvascular loss, or by permitting increased compensatory angiogenesis. Eventual resolution of this unanswered question will be important because the underlying downstream signaling pathways, and thus potential therapeutic targets, are different. EndoMT may be mediated by canonical TβRII/activin-like kinase ALK5 (also TβRI) receptor complex activation of Smad2/Smad3 pathways, whereas proangiogenic activity can be exerted by TGF-β selectively in endothelial cells by ALK1 receptor recruitment to multimeric TβR complexes, resulting in activation of the Smad1/Smad5-mediated bone morphogenetic protein pathway.⁹

Pending future resolution of these issues, the elegant studies by Xavier and colleagues add EndoMT to an already dizzying repertoire of TGF-β–induced pathomechanisms that ultimately damage and destroy glomeruli and nephrons in renal disease.⁵ Indeed, EndoMT was previously linked with tubulointerstitial fibrosis¹⁰; however, TGF-β has not been implicated in those reports. What do we know about signaling mediators acting downstream of TβR complexes and perhaps in concert with canonical Smad2/Smad3 proteins in endothelial cells? In pancreatic microvascular endothelial cell cultures, TGF-β–induced mesenchymal αSMA expression was mediated by nuclear accumulation of myocardin-related transcription factor A, which required activation of Smad2/Smad3 and RhoA pathways.¹¹ In the heart, NAD phosphate oxidase-2 activation by transgenic overexpression promotes EndoMT and proinflammatory effects and causes endothelial dysfunction, aggravating cardiac fibrosis and diastolic dysfunction.¹² Interestingly, endothelin-1/endothelin receptor type A–mediated glomerular endothelial cell dysfunction is required for podocyte depletion and progression of glomerulosclerosis in several podocyte injury models.¹³ Endothelial fenestrations and surface layer are characteristic hallmarks of differentiated capillary endothelial cells. A reduced number of glomerular endothelial fenestrations in patients with type 1 and type 2 diabetes has been documented.¹⁴ In addition, the endothelial surface layer composed of glycocalyx and endothelial cell coat has only recently been recognized as an important barrier to protein permeability, and studies assessing endothelial glycocalyx demonstrated a severe reduction of the endothelial glycocalyx in patients with type 2 diabetes and in animal models.¹⁵ In addition, endothelial phenotypes, including endothelial dysfunction by mitochondrial oxidative stress, endothelial senescence, and nitric oxide reduction have emerged as a key pathogenic mechanism in kidney injury associated with diabetes, hypertension, obesity, dyslipidemia, and aging.

Together with the present work by Xavier and colleagues, these observations collectively point to an emerging central role of endothelial cell injury in initiation and/or progression of renal disease in both glomerular and tubulointerstitial compartments. Elucidation of the molecular mechanisms of endothelial cell injury, including Edn1 and TGF-β stimulation among others, will be essential to understand mechanisms of renal hypoxia and fibrosis and kidney disease progression. It is intriguing to speculate whether and in what way the multiple manifestations of endothelial injury, including endothelial glycocalyx damage, endothelial dysfunction, or EndoMT, are interdependent or hierarchically ordered in a common pathway. Resolution of the endothelial cell response to injury will likely be of critical importance because the nature of the endothelial cell injury phenotype may initiate/control mediators of cross-talk among different renal cell types, specifying renal repair or disease progression. In music, the beautiful sound of a symphonic orchestra depends on the coordinated performance of its sections and on the unimpeded and unambiguous signaling between sections and conductor. As the single cell type reaching both glomerular and postglomerular sections of the nephron, future investigations may well find that renal microvascular endothelial cells serve as “conductors” in the kidney, directing and coordinating cell–cell signaling in renal homeostasis and disease.

Disclosures

None.