Transforming Growth Factor-Beta1 in Diabetic Kidney Disease

1Diabetic kidney disease (DKD), the most common cause of end-stage renal disease (ESRD) worldwide, accounts for about 40% of new cases of ESRD each year in the United States and China. With the increasing incidence of diabetes, there is a heightened need for therapy to delay progression of DKD. Existing therapies have had limited success. Renin-angiotensin-aldosterone system (RAAS) inhibitors, such as losartan and irbesartan, have been effective in reducing the risk of ESRD for patients with DKD. However, patients exhibited great variation in their responses to RAAS blockades. In the past two decades, there has been a decline in the rate of acute myocardial infarction and death from hyperglycemic crisis, but no change has occurred in the rate of ESRD. Although sodium-glucose co-transporter 2 (SGLT2) inhibitors have conferred cardiovascular and renal protection, effective therapy for DKD is still unavailable. An epidemiological study revealed that the 5-year mortality rate of DKD was approximately 40%, as high as many cancers. Transforming growth factor-β1 (TGF-β1) signaling contributes to DKD progression, and inhibiting TGF-β1 signaling has shown potential renoprotective properties in animal and human studies. In this mini-review, we discuss the pleiotropic and the potential therapeutic effects of TGF-β1 in DKD.

2TGF-βs exist as five isoforms, but only TGF-β1, TGF-β2, and TGF-β3 are present in mammals; the three isoforms elicit similar responses in vitro. TGF-β1, the most abundant isoform, is synthesized by all types of resident renal cells and infiltrating inflammatory cells. TGF-β1 is secreted into the extracellular matrix (ECM) in an inactive complex (latent TGF-β1) with TGF-β-latency-associated peptide (LAP) and latent TGF-β binding proteins (LTBP). The activation of latent TGF-β1 is mediated by proteolytic cleavage in the presence of the serine protease plasmin, reactive oxygen species (ROS), thrombospondin-1 (TSP-1), or integrins. Integrins bind to the arginine-glycine-aspartic acid sequence in LAP. This binding appears to change the conformation of the latent TGF-β1 complex by tractional force. This conformational change presents the latent TGF-β1 complex to transmembrane metalloproteinases, such as membrane-type-1-matrix metalloproteinase (MT-1-MMP), which cleave the latent TGF-β1 complex and release active TGF-β1. Active TGF-β1 interacts with its receptors to activate Smad-dependent and Smad-independent downstream signaling.

3Active TGF-β1 binds to a Type II membrane receptor, TGF-β Type II receptor (TβRII). This binding results in the phosphorylation and recruitment of the TGF-β Type I receptor (TβRI). The activated complex of TGF-β1-TβRI-TβRII phosphorylates Smad2 and -3. Then, the phosphorylated Smad2 and -3 bind to Smad4 to form the Smad complex. This Smad complex translocates into the nucleus and binds to Smad-binding elements (SBEs) or Smads-containing complexes, in turn, regulating transcription of genes encoding, e.g., collagen, fibronectin, α-smooth muscle actin, and Smad7.

4Smad proteins are classified into three subgroups. Smad2 and -3 comprise the receptor-regulated Smads (R-Smads) for TGF-β1 signaling, and Smad1, -5, and -8 for bone morphogenetic protein (BMP) signaling. Smad2 and -3 are key downstream mediators of TGF-β1, and they are highly activated in animal renal tissues in DKD. Smad2 and -3 may have distinct functions in renal fibrosis. Either a Smad3 knockout or a Smad3-specific inhibitor delayed de-differentiation of proximal tubular cells and alleviated renal fibrosis in a streptozotocin-induced model of diabetes. These findings suggested that TGF-β1/Smad3 signaling has critical activities in renal fibrosis. Conversely, unlike Smad3, the function of Smad2 in DKD is unclear. Overexpression of Smad2 attenuated TGF-β1-induced phosphorylated Smad3 and collagen expression, whereas deletion of Smad2 promoted renal fibrosis via substantially enhanced Smad3 signaling. Although Smad2 interacts with Smad3 physically, Smad2 and -3 may compete for phosphorylation in response to TGF-β1 stimulation. Thus, Smad2 may competitively inhibit phosphorylation of Smad3 in response to TGF-β1. Besides TGF-β1 signaling, Smad2 nuclear translocation and phosphorylation can also be mediated by advanced glycation end-products in DKD. Thus, the activity of Smad2 is complicated in DKD.

5The second Smad subgroup is the common-partner Smad (co-Smad), Smad4, which forms a heterotrimeric complex with phosphorylated R-Smads. The Smad4-containing complex translocates into the nucleus and regulates expression of the genes indicated earlier. Furthermore, Smad4 is implicated in suppressing nuclear factor-κB (NF-κB)-driven inflammation by inducing Smad7 expression.

6The third Smad subgroup is the inhibitory Smads (I-Smads). Members of this Smad family have a conserved carboxy-terminal MH2 domain. I-Smads inhibit TGF-β1 family signaling via interaction with type I receptors, and I-Smads compete with R-Smads for receptor activation. Smad7, one of the most investigated I-Smad in DKD, can cause degradation of TβRI and Smads activity in a negative feedback process. Smad7 inhibits Smad2/3 during renal fibrosis. In chronic kidney disease, TGF-β1 signaling upregulated the Smurfs and caused ubiquitin-dependent degradation of Smad7, which led to a decrease in Smad7 protein level. Smad7 knockout mice progressed to more severe interstitial fibrosis and enhanced inflammation, and overexpression of Smad7 in kidney was effective in reducing collagen matrix expression and in alleviating inflammatory infiltration in DKD. These findings revealed anti-fibrotic and anti-inflammatory functions of Smad7 in DKD.

7In addition to Smad-mediated transcription, TGF-β1 directly activates other signal transduction pathways in the pathophysiology of kidney disease. These other pathways include the mitogen-activated protein kinases (MAPK) pathway, growth and survival kinases phosphatidylinositol-3-kinase (PI3K)/Akt, small GTP-binding proteins such as Ras, RhoA, Rac1, and Cdc42, the Notch signaling pathway, Integrin-linked kinase (ILK), and the Wnt/β-catenin pathway. These non-canonical, non-Smad pathways can indirectly participate in de-differentiation of proximal tubular cells, apoptosis, and matrix formation, thereby mediating signaling responses either as stand-alone pathways or as pathways that converge onto Smads to control Smad activities.

8Diabetic kidney disease pathology is characterized by thickening of the glomerular basement membrane, mesangial expansion, segmental glomerulosclerosis or global glomerulosclerosis, tubulointerstitial fibrosis, and afferent and efferent arteriole hyalinosis. The TGF-β1 signaling pathway is activated in DKD, and the inhibition of TGF-β1 attenuates fibrosis in animal models of diabetes. Pathogenic stimuli in DKD activate TGF-β1 signaling. Angiotensin-II, which was elevated in mesangial cells and glomerular endothelial cells, has been implicated in activating TGF-β1 by generation of ROS from nicotinamide adenine dinucleotide phosphate oxidases or by activating protein kinase C- and p38 MAPK-dependent pathways. Hyperglycemia, mechanical stretch, and advanced glycation end products were found to upregulate TGF-β1 in DKD. TSP-1, a prototypic matricellular ECM protein, was heavily deposited in glomeruli of patients with DKD. TSP-1 binds to the latent TGF-β1 complex, and, by a non-proteolytic mechanism, converts latent TGF-β1 to the active form, which leads to upregulation of TGF-β1 signaling. Direct evidence for the importance of TSP-1 in regulating TGF-β signaling in DKD comes from two different models of type 1 diabetes. Streptozotocin-treated TSP-1 knockout mice showed decreased glomerular TGF-β signaling as measured by phosphorylated Smad2, and attenuated glomerulosclerosis. In another type 1 diabetic animal model, uninephrectomized Akita mice treated with TSP-1 blocking peptide LSKL were protected from tubulointerstitial fibrosis and had reduced phosphorylation of Smad2 and -3.

9Mechanisms of TGF-β1 regulated fibrosis in DKD are multifactorial and involve (1) overexpression of ECM, (2) decreased degradation of ECM, (3) enhanced cross-linking between collagen and elastin fibers, and (4) overactivation of proximal tubular and endothelial cell de-differentiation. Both canonical TGF-β1/Smads-dependent signaling pathways and alternative signaling by TGF-β1 are involved in stimulating collagen expression and accumulation. Neutralizing all three mammalian TGF-β isoforms (-β1, -β2, and -β3) with antibodies reduced ECM gene (fibronectin and type IV collagen) expression and attenuated renal fibrosis in mice with type 1 or type 2 diabetes. Thus, TGF-β1 has a critical signaling function in ECM accumulation in DKD.

10TGF-β1 expression greatly inhibited ECM degradation by promoting the synthesis of plasminogen activator inhibitor-1 (PAI-1) which resulted in renal fibrosis. The abundance of matrix metalloproteinase-9 (MMP-9), an ECM-degradation MMP, was decreased in transgenic mice that overexpressed TGF-β1. In addition, TGF-β1 augmented the expression of tissue inhibitor of metalloproteinases-1, which inhibited the ECM-degrading MMPs.

11TGF-β1 promotes formation of the cross-linking between collagen and elastin fibers by upregulating lysyl oxidase. In vitro, TGF-β1 significantly increased (∼5 times) lysyl oxidase expression in tubular epithelial cells. In addition, TGF-β1 stimulated expression of procollagen lysyl hydroxylase 2, an enzyme that hydroxylates lysyl residues of collagen telopeptides and stabilizes collagen cross-linking. Crosslinking increases ECM resistance to degradation by MMPs.

12De-differentiation of the proximal tubular cells and endothelial cells contributes to renal fibrosis in diabetic mice. Extensive studies confirmed that TGF-β1 contributes to renal fibrosis by stimulating proximal tubular de-differentiation and endothelial de-differentiation. Hypoxia-inducible factor 1α (HIF-1α) accumulated in DKD and HIF-1α enhanced de-differentiation of murine proximal tubular epithelial cells in vitro. Conditional HIF-1α ablation decreased interstitial collagen deposition and inhibited the development of tubulointerstitial fibrosis. Although TGF-β1 stimulation increased HIF-1α expression, blocking TGF-β1 signaling inhibited HIF-1α activity, and, conversely, blocking HIF-1α activity decreased TGF-β1 signaling. These studies suggested cross-talk between TGF-β1 and HIF-1α signaling in regulating proximal tubular de-differentiation. As to endothelial de-differentiation, in animal models of folic acid nephropathy or unilateral ureteral obstruction, curtailed TGF-β signaling in the endothelium by endothelium-specific heterozygous TβRII knockout reduced endothelial de-differentiation and led to less tubulointerstitial fibrosis. The mechanism by which TGF-β1 regulates endothelial de-differentiation is unknown. TGF-β1 stimulated endothelial de-differentiation in mouse endothelial cells by activating Snail expression.

13In summary, the active TGF-β1 system promotes renal fibrosis, and it is involved in elevating collagen synthesis, suppressing ECM degradation, promoting collagen cross-linking, and fostering proximal tubular or endothelial cell de-differentiation (Figure 1).

14TGF-β1 is a critical factor in the pathophysiological progression of DKD, having both pro- and anti-inflammatory properties.

15TGF-β1 control of innate immune cells can have severe pathological consequences. Leukocytes and fibroblasts are recruited by the activation of resident kidney immune cells in DKD. This recruitment stimulates the expression of pro-inflammatory and chemotactic cytokines, which further drives the infiltration of monocytes and macrophages. TGF-β1 recruited macrophages and dendritic cells by stimulating the production of chemokines, including tumor necrosis factor-alpha (TNF-α), monocyte chemoattractant protein-1 (MCP-1), and inducible nitric oxide synthase. Furthermore, the secreted chemokines induced TGF-β1 expression in a positive feedback loop, which sustained the high levels of TGF-β1 in the microenvironment. TGF-β1 induced the expression and release of other proinflammatory cytokines such as interleukin-8 (IL-8) and MCP-1 in proximal tubular cells. In addition, TGF-β1 drove the differentiation of T helper 17 cells, which were activated in various proinflammatory conditions. In the presence of IL-6, TGF-β1 promoted the differentiation of naive T lymphocytes into proinflammatory T helper cells that produced IL-17 and augmented autoimmune conditions, which were enhanced by IL-1β and TNF-α. In this way, TGF-β1 propagates and amplifies the proinflammatory and profibrotic processes that contribute to renal insufficiency in DKD (Figure 1).

16Nevertheless, TGF-β1 also possesses anti-inflammatory properties, which was suggested by the findings that targeted deletion of the TGF-β1 gene resulted in profound multifocal inflammatory disease in mice. Additionally, TGF-β1 knockout mice developed severe inflammatory responses that were evidenced by massive lymphocytes, macrophages, immunoblasts, and plasma cell infiltration in many organs. Tubular epithelial cell-specific TβRII knockout mice showed massive leukocytes or macrophages infiltration, increased proinflammatory cytokine release, and enhanced renal inflammation. Direct evidence for the importance of TGF-β1 in anti-inflammation comes from two studies. First, used animal studies to investigate the effect of different doses of TGF-β antibodies on glomerulosclerosis. Only low dose TGF-β antibody decreased macrophage infiltration, and reduced sclerosis, indicating that the amount of TGF-β may influence the inflammatory process. Second, regulatory T cells appeared to ameliorate DKD, and nude mice, which lacked all T-cell subtypes, had more severe DKD. In the presence of IL-2, TGF-β1 converted naive T cells into Foxp3 + regulatory T cells and inhibited the progression of DKD.

17Thus, the effects of TGF-β1 activation in renal inflammation may be protective or harmful depending on concentration or the presence of IL-6. However, the underlying mechanism by which TGF-β1 exerts its anti-inflammatory properties in DKD requires further investigation.

18Recent studies illustrated that TGF-β1 promoted autophagy. Autophagy, a system for removing protein aggregates and damaged organelles to maintain cellular homeostasis, is impaired in glomeruli and tubules in DKD. However, persistent activation of autophagy in kidney tubular epithelial cells induced tubular degeneration and promoted renal fibrosis. Overexpression of TGF-β1 in renal tubules induced the accumulation of autophagosomes and stimulated expression of autophagy-related genes. In proximal tubular cells, TGF-β1 promoted autophagy by generation of ROS, which contributed to the proapoptotic effect of TGF-β1. proposed TGF-β1-driven autophagy as a novel mechanism of tubular degeneration that led to renal interstitial fibrosis. On the contrary, TGF-β1 induced autophagy had positive effects. In a study by, TGF-β1 induced autophagy in mesangial cells, and autophagy enhanced cell survival by preventing mesangial cells from undergoing apoptosis. Whether TGF-β1 driven autophagy has protective or deleterious effects on kidney depending upon the amount. In the study by Koesters et al., TGF-β1 level was higher than its level in pathological disease states, which triggered violent autophagy and promoted kidney injury. Thus, we need further clarification of the functions of TGF-β1 signaling-induced autophagy in the pathogenesis of DKD.

19TGF-β1 also suppresses reabsorption of glucose by proximal epithelial cells. A dose-dependent increase in TGF-β1 expression by genetic manipulation increased urinary output of glucose in Akita mice, whereas genetic insufficiency of TGF-β1 decreased glucose output. Moreover, SGLT2 was directly regulated by TGF-β1 via Smad3 and TGF-β1 showed decreased expression of SGLT1 and SGLT2. Thus, these results support the notion that TGF-β1 suppresses urinary glucose reabsorption in proximal tubular epithelial cells (Figure 1).

20Blockade of TGF-β1 signaling as a therapeutic strategy has been achieved by gene technology and pharmacological drugs (Table 1). Inhibition of TGF-β with a pan-neutralizing monoclonal antibody (1D11) against all three isoforms ameliorated renal fibrosis and alleviated kidney structural changes in the rodent models of type 1 and type 2 diabetes mellitus. Pirfenidone is a low molecular weight synthetic molecule that has antifibrotic properties in animal models; it suppresses production of ROS and downregulates genes encoding profibrotic cytokines, such as α-SMA, collagen I, and collagen IV. Pirfenidone upregulates regulator of G-protein signaling 2. Moreover, found that pirfenidone decreased TGF-β promoter activity, blocked TGF-β1 production, and was effective in reducing mesangial matrix expansion and fibrosis in DKD. Switching TGF-β1 expression from low to high by genetic manipulation exacerbated renal injury in Akita mice, a result that further supported the idea that blockade of TGF-β1 was renoprotective for DKD.

21The success of TGF-β1 signaling inhibition in animal studies has promoted the strategy in clinical investigations with DKD. Pirfenidone significantly increased estimated glomerular filtration rates (eGFR) in a cohort of 77 diabetic patients with baseline eGFR of 20–75 ml/min/1.73 m2. However, a placebo-controlled, phase II study that used a humanized TGF-β1-specific neutralizing monoclonal antibody plus renin-angiotensin system blockades failed to slow the progression of DKD in diabetic patients who had eGFR of 20–60 ml/min/1.73 m2. Lack of improvement in clinical trials may be explained by the fact that rodent models of diabetes do not recapitulate tubulointerstitial fibrosis to the same degree observed in human disease. Also, inhibiting TGF-β1 fully and indiscriminately may not be wise because of its multiple physiological functions.

22Nevertheless, targeting the conversion of latent to active TGF-β1 holds promise as a DKD therapeutic intervention. Animal studies revealed that overexpression of an active form of TGF-β1 in liver led to progressive kidney fibrosis in mice, whereas overexpression of latent TGF-β1 in the skin displayed anti-inflammatory and anti-fibrosis effects in obstructive and crescentic glomerulonephritis. The distinct functions of active and latent TGF-β1 in renal fibrosis and inflammation suggest that a better therapeutic approach would be to block conversion of latent TGF-β to active TGF-β. showed that inhibiting conversion of latent to active TGF-β1 in human proximal tubular cells reduced matrix protein expression and inhibited fibrosis under hyperglycemia and hypoxia conditions. What is more, the αv-containing integrins with different β-subunits that interact with latent TGF-β1 and activate TGF-β1 have a critical function in kidney fibrosis. A pharmacologic inhibitor of αvβ1 integrin prevented activation of the latent TGF-β complex and ameliorated renal fibrosis in mice fed an adenine diet. The mechanisms of the distinct functions of latent versus active TGF-β1 may be related to the prevention of Smad7 from Smurf-mediated ubiquitination and degradation in response to higher levels of latent TGF-β1. Smad7 inhibits TGF-β signaling by promoting degradation of the TβRI and inhibiting Smad2/3/4 activity. But in chronic kidney disease, active TGF-β1 activates the Smurfs and arkadia-dependent ubiquitin-proteasome pathways, which degrades Smad7 protein by a post-transcriptional modification mechanism.

23On the basis of experimental and clinical studies, modulating TGF-β1, instead of directly inhibiting TGF-β1 ligands/receptors, may be a good antifibrosis tactic for DKD. TGF-β1 promotes wound healing, tissue regeneration, anti-inflammation, autophagy, and urinary glucose regulation. Nonetheless, the dose regimen must be considered carefully because a large dose of TGF-β blockade had severe toxicity and poor efficacy in animal experiments. A pan-neutralizing monoclonal antibody could also lead to undesired effects such as tumor formation, even though animal studies have not exhibited such events during prolonged TGF-β1 inhibition. What is more, developing molecules that suppress the activation of latent TGF-β1 would be a potential therapy. Given the central role of TGF-β1 in the pathophysiology of DKD, the TGF-β1 system is an attractive target to retard the progression of DKD, provided that the approach maintains an acceptable balance between renoprotective and harmful effects.

24LZ and FL conceptualized this review and decided on the content. LZ, YZ, and FL wrote and revised the manuscript. All authors approved the final version of the manuscript.

25The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.