Lipid mediators of insulin signaling in diabetic kidney disease

Abstract

Diabetic kidney disease (DKD) affects ∼40% of patients with diabetes and is associated with high mortality rates. Among different cellular targets in DKD, podocytes, highly specialized epithelial cells of the glomerular filtration barrier, are injured in the early stages of DKD. Both clinical and experimental data support the role of preserved insulin signaling as a major contributor to podocyte function and survival. However, little is known about the key modulators of podocyte insulin signaling. This review summarizes the novel knowledge that intracellular lipids such as cholesterol and sphingolipids are major determinants of podocyte insulin signaling. In particular, the implications of these lipids on DKD development, progression, and treatment will be addressed.

INTRODUCTION

The increasing prevalence of type 2 diabetes mellitus worldwide has resulted in a rise in diabetic kidney disease (DKD). DKD that is due to either type 1 (T1D) or type 2 diabetes mellitus (T2D) is the most common cause of end-stage kidney disease (ESKD) in the world (165). The natural history of DKD is characterized by progressive albuminuria and/or gradual decline in estimated glomerular filtration rate (eGFR) even in the absence of albuminuria.

Multifactorial intervention trials targeting glycemic control, blood pressure, and lifestyle interventions have demonstrated that current treatments slow but do not halt the progression of DKD in both T1D (73) and T2D (55). One of the major obstacles for a definitive cure for DKD resides in the fact that the pathogenesis of DKD remains largely unknown. In fact, key glomerular cell constituents of the renal glomerular filtration barrier are affected in diabetes not only by high glucose and altered hemodynamics but also by a large variety of endocrine and autocrine factors (42, 67).

Insulin resistance is a condition in which cells fail to respond to the normal action of insulin. Insulin resistance can occur as a consequence of a mutation of the insulin receptor (IR) and downstream mediators or it can be caused by a multiplicity of factors affecting proper phosphoinositide 3-kinase (PI3K)/Akt pathway signaling responsible for glucose transporters trafficking among other metabolic pathways. Although the glomerular filtration barrier consists of three cellular components and of a glomerular basement membrane, the focus of this review will be on podocytes, which are terminally differentiated cells that have been shown to express functional IRs (38). Podocytes can become insulin resistant before the development of DKD (145, 160). Podocytes can also take up glucose in response to insulin via glucose transporter (GLUT)1 (185) and GLUT4 (62), both being major contributors to podocyte function. We will review both the clinical and experimental evidence supporting the importance of podocyte insulin signaling in the preservation of the glomerular filtration barrier. We will also review the contribution to insulin signaling of two insulin receptor (IR) subtypes (A and B) with specific roles in the modulating of cell growth, differentiation, apoptosis, and metabolism. A major focus of our review will be on lipid mediators of insulin signaling in podocytes, as we have reported a link between podocyte cholesterol content and insulin signaling (111). Furthermore, the contribution of certain sphingolipid species to insulin signaling via different IR isoforms will be discussed.

INSULIN RESISTANCE CONTRIBUTES TO DKD: CLINICAL EVIDENCE

Insulin resistance leads to important hemodynamic changes in the kidney, including increased sympathetic nervous tone, hypertension, and atherosclerosis of the renal microvasculature. Insulin resistance is a risk factor for DKD and may directly contribute to the development of the disease (175). Insulin resistance is frequently observed in patients with mild, advanced, or end-stage chronic kidney disease (for a review, see Ref. 163), and many studies have demonstrated a correlation between the presence of insulin resistance and albuminuria. A recent study on adult patients with newly diagnosed diabetes (with no differentiation between T1D and T2D) revealed that a higher risk of DKD is associated with insulin resistance and poor metabolic control (high HbA_1c) (3). Few cross-sectional studies of patients with T2D have demonstrated a positive correlation between measured or calculated insulin resistance and levels of albuminuria (16, 134, 135). Longitudinal studies have demonstrated that even in normalbuminuric patients with T2D or nondiabetic siblings of patients with T2D, insulin resistance predicts the development of albuminuria development (52, 126).

Moreover, similar findings have been reported for patients with T1D (46, 182). The EURODIAB study demonstrated that other conditions, such as diabetes duration, glycemic control, or waist-to-hip ratio, predict microalbuminuria prevalence in patients with T1D (25). In a study on 652 patients with T1D (including 6.2 yr of followup), insulin resistance predicted the development of albuminuria (17, 44). Smaller studies with euglycemic clamps confirmed these findings in patients with T1D as well as in their family members (181, 182).

The Study of Glucose and Insulin in Renal Disease showed that participants with moderate to severe CKD have lower insulin sensitivity and impaired glucose tolerance (39). Hyperinsulinemia and abnormal insulin resistance were associated with kidney failure in nondiabetic adults with CKD from the United States (29). Interventional studies have suggested that attenuating of insulin resistance (for example, using thiazolidinediones) may have renoprotective effects. In randomized trials, normalization of insulin sensitivity reduced the incidence and severity of albuminuria in patients with T2D (117, 149). Another recent trial demonstrated that use of metformin for 36 mo resulted in a preservation of eGFR (141). The evidence that insulin resistance may contribute to chronic liver disease also supports a potential causative role in other organs (21). Whether insulin resistance is the cause or the consequence of CKD remains to be established, but several experimental studies have suggested that insulin resistance contributes to DKD development and progression, as reviewed in more detail below.

INSULIN RESISTANCE AND THE KIDNEY: EXPERIMENTAL EVIDENCE

The development of insulin resistance does not include defects in the binding of insulin to its receptor on cell membranes. Instead, insulin resistance results in “postreceptor defects” that interfere with intracellular signaling processes. The insulin signaling cascade is divided into major pathways such as the PI3K/Akt pathway and MAPK/MEK pathway; in insulin resistance, these pathways are not equally impaired. Recently, it has been suggested that impairment of insulin signaling also occurs in the kidney (113). The kidney is unique in that insulin can access its target cells both from the lumen (in the case of podocytes) and from the basolateral side (in the case of tubular cells) (76). Despite compelling evidence that the kidney is an insulin-responsive organ, whether the kidney in general or in part is affected by insulin resistance similarly to the classical target organs, such as muscles, the liver, and adipose tissue, is not clear. While in adipose tissue both insulin receptor substrate 1 (IRS)1- and IRS2-dependent signals are impaired in the setting of insulin resistance, renal proximal tubule insulin signaling is impaired via IRS1 but not via IRS2 (113, 123). The human kidney abundantly expresses IR isoform B similarly to the classical target organs; glucose uptake is insulin dependent in podocytes only (38, 62, 101, 150) but not in tubules or mesangial cells (7, 170). In diabetic rats, only the PI3K/Akt pathway, but not the MAPK/ERK pathway, was affected in isolated glomeruli (113). Moreover, protein expression of IRS1 in glomeruli was reduced but could be restored by inhibition of PKC-β (113). The impairment of IRS2 signaling in podocyte in the onset of diabetic nephropathy has recently been suggested (147). Together, these data indicate that glomeruli, but not tubuli, can develop insulin resistance and that insulin resistance is selective (for a review, see Ref. 72).

Overall, insulin resistance has a lot of implications for podocyte biology, and all factors associated with systemic insulin resistance have been shown to disrupt podocyte insulin signaling, as has been recently reviewed (97).

INSULIN SIGNALING IN PODOCYTES

Insulin exerts control over cell metabolism by binding to its cell surface receptor, which has been characterized in great detail (34, 98, 128, 129, 176). The occupied receptor is autophosphorylated on tyrosine residues and can thereby tyrosine phosphorylate other cellular proteins, such as the IRS family of proteins, to transduce the insulin signal into the signal network of a cell. The further downstream events involve the generation of second messengers. Among them, PI3K has a major role in insulin function, mainly via phosphorylation of Akt/PKB and PKC (PKC-ζ) cascades. Activated Akt induces glycogen synthesis through inhibition of glycogen synthase kinase 3 (GSK-3), protein synthesis via mammalian target of rapamycin (mTOR), and downstream elements and cell survival through inhibition of several proapoptotic agents (Bad, FoxO transcription factors, GSK3, and macrophage stimulating 1). Insulin signaling also has growth and mitogenic effects (Akt cascade and the Ras/MAPK pathway), inhibits autophagy (via Unc-51-like autophagy activating kinase; see Ref. 119), stimulates glucose uptake (for a review, see Ref. 174), inhibits gluconeogenesis in the liver [through disruption of cAMP response element-binding protein (CREB)/CREB-binding protein (CBP)/mTOR complex (mTORC)2 binding; see Ref. 172], induces fatty acid and cholesterol synthesis (via the regulation of sterol regulatory element-binding protein transcription factors; see Refs. 102, 132, 140, and 180), and promotes fatty acid synthesis (through activation of upstream transcription factor 1 and liver X receptor; see Refs. 142, 156, 173, and 177).

The glomerulus is composed of three cell types: podocytes, endothelial glomerular cells, and mesangial cells. All of these cells have been shown to respond to insulin stimulation. However, podocytes have the highest IR and IRS1 expression levels compared with endothelial and mesangial cells (113). Podocytes constantly express all components of insulin signaling (38, 160) and modulate glucose uptake via GLUT1 (185) and GLUT4 (62), both of which have been shown to be involved in the pathogenesis of DKD (for a review, see Ref. 37). Loss of insulin-stimulated Akt phosphorylation has been reported in podocytes of both T1D (43) and T2D (160) mouse models. The inability to signal through Akt2 was associated with increased podocyte susceptibility to cell death (23). More evidence for the critical role of insulin in podocyte function came from mice with podocyte-specific deletion of the IR (175). In these animals, albuminuria developed, along with effacement of the podocyte foot process, apoptosis, thickening of the glomerular basement membrane, and increased glomerulosclerosis. Additionally, studies have reported that expression levels of the IR and IRS1 were decreased in the glomeruli of insulin-resistant and diabetic rats (113, 164). IRS2 has also been reported to affect podocyte insulin sensitivity. Podocytes depleted of IRS2 were unable to phosphorylate Akt and failed in GLUT4-mediated glucose uptake in association with marked reduction of cortical F-actin remodeling and increased cell motility (147). In addition to these functions, insulin signaling may also regulate podocyte function by modulation of VEGF-A production (65), and it can modify podocyte contractility by regulating Ca²⁺ influx (via transient receptor potential C6 and calcineurin-dependent pathways) (178). Finally, insulin signaling in podocytes can regulate the activity of equilibrative nucleoside transporters responsible for controlling extracellular levels of adenosine (5) and the adaptive endoplasmic reticulum stress response (via p85-X box-binding protein 1) (108).

In a recent study (23), a link between activation of Akt and mTOR in podocytes has been suggested. Activation of mTORC1 downstream of Akt has been shown to induce podocyte injury in DKD (58, 77), whereas inhibition of mTORC2 has been shown to promote insulin resistance in Drosophila (148).

IR ISOFORMS IN HEALTH AND DISEASE

Alternative splicing of the IR adds additional complexity to the understanding of insulin signaling and may have implications in the development of DKD. The IR exists in two isoforms, which are formed due to the absence (isoform A) or presence (isoform B) of exon 11 of the IR gene. The expression of IR isoforms is tissue specific. IRA is predominantly expressed in fetal and cancer tissues (54), whereas IRB is predominantly expressed in insulin target organs such as the liver, muscles, adipose tissue, and, interestingly, kidney (see Refs. 120 and 169; for a review, see Ref. 12). IR splicing is a conserved mechanism in mammals responsible for the specificity in insulin signaling. A predominant expression of IRA is associated with a decrease in metabolic signaling of insulin and in an increase in the signaling of IGFs (i.e., in developmental and fetal growth). In contrast, increased expression of IRB is associated with increased metabolic actions of insulin. Additionally, the use of IR isoform-selective insulin analogs has demonstrated that IRA expression more than IRB is relevant to efficient glycogen synthesis, whereas IRB more than IRA exerted a stronger effect on glycogen accumulation and lipogenesis (169). Dysregulation of the balance between expression of IRA and IRB in adult life plays an important role in different pathological processes (12). The first evidence that IR isoforms are differentially expressed in disease was obtained in cancer fibroblasts and breast cancer cells (54, 153) and later was shown in a wide variety of cancers (86, 154, 166). In murine hepatocytes, it has been shown that IRA has a stronger effect than IRB in favoring basal glucose uptake by binding GLUT1/2 and inducing GSK-3α/β phosphorylation (41). It also been shown that patients with diabetes have an increased IRA-to-IRB ratio in skeletal muscle, which is linked to insulin resistance (139). While the importance of podocyte IR signaling in DKD has been elegantly described (175), the specific contribution of different IR isoforms in the pathogenesis of DKD has not yet been established.

CAVEOLAE AND INSULIN SIGNALING IN PODOCYTES

Caveolin-1, a critical regulator of insulin receptor expression, is highly expressed in podocytes, where it binds the key slit diaphragm proteins nephrin and CD2-associated protein (157). Cholesterol and sphingolipids together with caveolin are required for caveolae to form and for the different components of the slit diaphragm to function (50, 151, 167). Caveolae are involved in numerous cellular processes such as receptor-mediated uptake, receptor-mediated signaling, and vesicular trafficking (see Refs. 66 and 104; for a review, see Ref. 22). Interestingly, binding of IRs to caveolin-1 can be dissociated by plasma membrane molecules such as monosialodihexosylganglioside (GM3) (85) as well as by circulating TNF-α (155). GM3 can also interfere with raft clustering in podocytes (81). In a recent study (171) on zebrafish podocytes, it was shown that overexpression of caveolin-1 causes significant proteinuria and podocyte injury, suggesting that caveolin-1 could be another therapeutic target in nephrotic syndrome and podocyte injury. Whether and how this would interfere with insulin signaling remains to be established. In fact, some recent studies have demonstrated that IR signaling occurs in a particular subset of membrane microdomains and requires caveolin-2 (93, 94), and thus further studies are needed to elucidate the link between caveolin-2, insulin signaling, and podocyte function.

CHOLESTEROL AND INSULIN SIGNALING

Among several lipids involved in the modulation of insulin signaling, cholesterol has been extensively studied as a key modulator of skeletal muscle cell function (10). We have recently identified cholesterol as a major mediator of podocyte function (111, 115, 137). In fact, sequestration of cholesterol with cyclodextrin was found to protect podocytes in DKD and to restore proper insulin signaling (111). In rats fed a high-cholesterol diet, activity of IRS1 and Akt was impaired, and tyrosine phosphorylation of caveolin-1 failed in response to insulin stimulation in the liver (63). In rat adipocytes, inhibition of downstream insulin signaling via IRS1 has been shown to be in a state of cholesterol depletion (133), suggesting an important role of cholesterol in the metabolic control of insulin signaling. A further link between intracellular cholesterol and insulin signaling is suggested in mice with hepatocyte-specific knockout of ATP-binding cassette transporter A1 (ABCA1), where impairment of ABCA1-dependent cholesterol efflux was associated with reduced hepatic insulin-stimulated Akt phosphorylation (87). More recently, treatment with small-molecule ABCA1 inducers was found to be sufficient to protect from experimental DKD (45). Interestingly, a 40% decrease in insulin-stimulated glucose uptake was reported in skeletal muscle of mice on a high-fat diet (106), suggesting that cholesterol may be a novel regulator of GLUT4 trafficking to the plasma membrane (for a review, see Ref. 10). Our own findings also suggested that accumulation of triglyceride-enriched lipid droplets via increased uptake of free fatty acids by CD36 (82) or esterified cholesterol-enriched lipid droplets via suppressed activity of ABCA1 (137) may both contribute to podocytes injury (Fig. 1).

Fig. 1.

Proposed mechanisms of podocyte susceptibility to lipid-mediated injury. 1) Upregulation of free fatty acids (FFA). The CD36 pathway leads to the accumulation of triglyceride (TG)-enriched lipid droplets and activation of inflammasome pathways, which causes mitochondrial damage and podocyte injury. 2) Overexpression of sphingomyelin phosphodiesterase acid-like 3b (SMPDL3b) leads to the accumulation of ceramide (Cer)-1-phosphate (C1P) and disruption of insulin receptor (IR)/caveolin-1 (Cav1) interactions, which results in decreased phosphorylation of PKB (Akt) and increased podocyte injury. 3) Local TNF-α causes activation of nuclear factor of activated T cells (NFAT) and its translocation to the nucleus, where it blocks transcription of ATP-binding cassette transporter A1 (ABCA1), resulting in the accumulation of cholesterol and esterified cholesterol-enriched lipid droplets, leading to mitochondrial damage and podocyte injury. NLRP3, nucleotide-binding oligomerization domain, leucine rich repeat and pyrin domain containing 3; ROS, reactive oxygen species; TNFR, TNF receptor; SOAT1, sterol O-acyltransferase 1.

SPHINGOLIPIDS AND INSULIN SIGNALING

Bioactive sphingolipids such as ceramide, ceramide-1-phosphate (C1P), and sphingosine-1-phosphate (S1P) play numerous roles in cell survival, proliferation, differentiation, and multiple aspects of stress responses. They also interact with intracellular signaling pathways (6), bind transmembrane domains of signaling proteins within the lipid bilayer (36), or even create membrane pores in mitochondria (35). Moreover, sphingolipids play a pivotal role in glomerular disorders of genetic and nongenetic origins (for a review, see Ref. 112).

Ceramide has attracted the most attention due to its roles not only in cell death and senescence but also in cell differentiation, membrane fluidity, protein anchoring, immune activation, and insulin resistance. Plasma ceramide levels are elevated in patients with T2D (18, 68) and are inversely associated with insulin sensitivity (68). Thus, both the liver and skeletal muscle of mice lacking dihydroceramide desaturase 1, which converts metabolically inactive dihydroceramide into active ceramide, have been shown as a target of ceramide-induced insulin resistance (70). Ceramide inhibits activation of Akt/PKB (27) but does not affect early signaling events such as IRS1 phosphorylation or PI3K activity (159). Moreover, patients with Gaucher’s disease, a lysosomal storage disease caused by failure to degrade glycosylated ceramides, are insulin resistant (95). Inhibition of glucosylceramide synthase improves insulin sensitivity in obese rodents. It has been shown that mice lacking GM3 synthase, which causes production of glycosphingolipids, showed improved insulin sensitivity in adipocytes (179), whereas in polycystic kidney disease loss of GM3 synthase has protective effects against cyst formation (125). However, ceramide and GM3 seems to act at different loci in different tissues, as glucosylceramide synthase inhibitors increased ceramide levels and antagonized signaling to Akt/PKB in myotubes (28). In addition, a recent study (90) showed that accumulation of ceramide in sphingomyelin synthase 2 knockout mouse embryonic fibroblasts fails to impair insulin signaling.

Both sphingosine and ceramide can be phosphorylated into S1P and C1P, respectively. The prosurvival activity of S1P highlights its role as a potential therapeutic target in neurodegenerative disorders (40, 83, 121), rheumatoid arthritis (32), cancer (15, 47, 122, 186), lung disorders (for a review, see Ref. 118), heart disorders (30, 79), renal oncology (4), acute kidney injury (9, 138), and nephrotic syndrome (143). Interestingly, the plasma S1P level is increased in the genetically obese ob/ob mouse model (92, 146) and in rodent models of T1D (53). In murine adipocytes, deficiency of serine palmitoyltransferase C2, which catalyzes the first step of de novo sphingolipid synthesis, causes markedly reduced adipose tissue mass and systemic insulin resistance and hyperglycemia (100). In vitro studies using C₂C₁₂ myoblasts have shown that S1P can stimulate basal glucose uptake by transphosphorylation of the IR (144), and in rat primary adipocytes, S1P has been shown to stimulate lipolysis (84). In vivo studies on transgenic mice overexpressing sphingosine kinase 1 showed improved insulin sensitivity compared with wild-type mice after 6 wk of a high-fat diet (19) and glucose tolerance, insulin uptake, and Akt phosphorylation through ceramide reduction in muscles (20). In primary hepatocytes, activated expression of sphingosine kinase 2 was associated with elevated expression of Akt, with no alteration of IRS phosphorylation, thus ameliorating glucose intolerance and insulin resistance (99). It has been suggested that S1P contributes to hepatic insulin resistance in primary rat and human hepatocytes as well as in the liver of high-fat diet-fed New Zealand obese mice (48). Overall, high-fat diet-fed mice demonstrated increased amounts of ceramide species, sphingomyelin species, sphingosine, and S1P in the liver, skeletal muscles, adipose tissue, heart, and plasma (91).

The role of sphingolipids in the kidney has remained largely unknown. It has been demonstrated that renal levels of S1P were increased in the streptozotocin mouse model, a phenomenon that was prevented by intraperitoneal insulin injections (127). Other studies in humans revealed that polymorphism in the SGPL1 gene, which encodes S1P lyase 1, is associated with reduced enzymatic activity of S1P lyase 1 and with the development of nephrotic syndrome (103, 107). In mice, lack of the Sgpl1 gene causes progression of foot process effacement, resulting in severe proteinuria (107). The use of unselective S1P receptor agonist FTY720, acting on S1P receptors 1, 3, 4, and 5 but not on S1P receptor 2, showed nephroprotection in streptozotocin-induced diabetic nephropathy in rats (8), whereas berberine improved renal injury in DKD via downregulation of S1P receptor 2 (74).

In contrast, much less known about C1P, as its pro- or anti-inflammatory role is still a matter of debate (24, 59, 61, 64). C1P and its precursor ceramide have opposite effects on cell survival, and a correct balance between the concentrations of C1P and ceramide is essential for cell and tissue homeostasis. In contrast to S1P, C1P is most likely not secreted by intact cells but is released by leaky or damaged cells (88). Although C1P has been claimed to stimulate cell proliferation in C₂C₁₂ myoblasts (13, 56) and macrophages (130, 131), not all studies have proven this concept. Thus, in leukemic cells, C1P did not affect cell proliferation, but it was described as an important factor mediating cell migration (1). C1P has been shown to modulate Akt phosphorylation in skin fibroblasts, hematopoietic cells (88), macrophages (57, 60), and adipocytes (26), consistent with the observation that bioactive sphingolipids are major modulators of insulin signaling (80, 110, 158). In renal mesangial cells, it has been shown that knockout of ceramide kinase, an enzyme that generates C1P from ceramide, causes proliferation of PGE₂ and is efficient for the treatment of mesangioproliferative glomerular diseases (136). Taken together, although the role of S1P and C1P on insulin resistance and PI3K/Akt signaling in the liver, muscles, and pancreatic β-cells is not fully understood, data on kidneys are mostly unavailable.

We have recently reported that the expression of sphingomyelin phosphodiesterase acid-like 3b (SMPDL3b), a lipid-raft associated protein (51) that regulates plasma membrane fluidity (69), is increased in glomeruli from patients with DKD as well as in glomeruli of db/db diabetic mice (183). Upre gulation of SMPDL3b was also found in normal human podocytes exposed to the sera of patients with DKD (183) and was found to be associated with an inability of DKD sera-treated human podocytes to phosphorylate Akt in response to insulin (111). Because SMPDL3b is a protein with homology to acid sphingomyelinase (31% amino acid identity and 48% overall amino acid similarity), we hypothesized that SMPDL3b may activate the sphingomyelin metabolic pathway, leading to the accumulation of sphingolipids other than sphingomyelin. We demonstrated that high SMPDL3b expression negatively affects the availability of C1P and demonstrated that exogenous administration of C1P in SMPDL3b-overexpressing podocytes is sufficient to restore insulin signaling and protects podocytes from injury (114). In the same study, we showed that human podocytes express both IR isoforms, IRA and IRB, and that SMPDL3b may facilitate IRB subtype signaling, thus favoring Akt phosphorylation in response to insulin by modulating IRA/IRB binding to caveolin. In additional in vivo experiments, we demonstrated that diabetic mice with podocyte-specific Smpdl3b deficiency are protected from the development of DKD, along with reduced albuminuria and preservation of podocyte numbers. This occurred in association with the restoration of the C1P content in kidney cortexes and, most importantly, with the restoration of podocyte-specific Akt phosphorylation (114). Overall, these in vitro and in vivo data reveal SMPDL3b as a master modulator of insulin signaling in podocytes (Fig. 1).

IR SIGNALING AS A THERAPEUTIC TARGET FOR DKD

Current DKD therapies are focused mainly on achievement of specific targets for glycemia and for blood pressure. Among the emerging new treatment strategies, the use of oral hyperglycemic agents [dipeptidyl peptidase-4 and Na⁺-glucose cotransporter (SGLT)2 inhibitors], mineralocorticoid receptor and endothelin receptor antagonists, selective NADPH oxidase 1/4 inhibitors, PKC inhibitors, advanced glycation end-product inhibitors, and agents that interfere with the VEGF pathway and with inflammatory mediators have shown promising results (for a review, see Ref. 89).

Studies with older insulin sensitizers of the class of thiazolidinediones have demonstrated that insulin sensitizers may further reduce albuminuria when achieving the same HbA_1c targets as standard of care (149). In addition, the REMOVAL trial demonstrated that use of metformin is associated with significant preservation of eGFR at 36 mo (141). The question of how insulin sensitizers may protect the kidneys has been of interest during the past 15 yr. Interestingly, metformin has been described as an inhibitor of the respiratory chain complex 1 in the mitochondria via AMP-activated kinase, leading to reduced reactive oxygen species production (49, 124). It would be very interesting to know whether improvement of renal insulin sensitivity and improved mitochondrial respiration also contribute to the very prominent renoprotective effect observed with SGLT2 inhibitors (31, 109, 168).

Many ways to selectively regulate IR activity and downstream signaling have been discovered. With regard to targeting specific IR isoforms, it has been shown that proliferative effects of long-acting insulin analogs may occur predominantly via IRA (152), whereas mitogenic responses occur preferentially via IGF-1 receptors (161, 162). The use of IR ligands that might be able to separate metabolic from mitogenic IR actions has been actively studied in the past years. Based on studies of autoantibodies to the IR in patients with rare disease and severe insulin resistance, novel anti-IR monoclonal antibodies were designed to induce distinct structural states of the receptor and, therefore, different signals (11, 78). These antibodies demonstrated the possibility of differently stimulating the pleiotropic biological signals of IR activation and the advantage being specific to IR isoforms and not to the IGF-1 receptor (14). In addition to antibodies, aptamers (184) and small synthetic peptides (96) were obtained to modify IR activity.

Our most recent study supports the possibility of using active sphingolipids for the cure of DKD (Fig. 1). Indeed, kidneys are among the most sensitive organs to sphingolipid alterations (for a review, see Ref. 2). For instance, rapamycin treatment of rats with streptozotocin-induced DKD significantly decreased formation of many sphingolipids species, which have been shown to be elevated in DKD (105). Our group previously showed that increased expression of SMPDL3b was associated with increased RhoA activity and apoptosis in db/db mice (183). Because ceramide, sphingosine, and S1P are known to accumulate in apoptotic cells, we determined the ceramide content in kidney cortexes of db/db mice and found it to be decreased (114). On the other hand, SMPDL3b expression is increased in DKD, and targeting of SMPDL3b would indeed protect the podocytes from damage (51, 183). Another study has revealed that specific sphingomyelin species (SM d18:1/16:0) are accumulated in the glomeruli of diabetic or high-fat diet-fed mice, which leads to suppressed AMP-activated protein kinase, to an elevated ATP-to-AMP ratio and, as a result, to reduced mitochondrial activity and biogenesis (116). Inhibition of sphingomyelin synthases using siRNA reverses these effects. It has been shown that adiponectin, which is known to take part in the regulation of energy metabolism, stimulates ceramidase activity (via its two receptors, AdipoR1 and AdipoR2) and enhances ceramide catabolism and formation of its antiapoptotic metabolite, S1P (71). AdipoRon, a synthetic adiponectin agonist, may ameliorate ceramide-induced lipotoxicity in DKD by decreasing ceramide, oxidative stress, and apoptosis in glomerular endothelial cells and podocytes (33) and might be considered a potential treatment therapy of DKD. More recently, we have reported that exogenous C1P administration was associated with increased Akt phosphorylation in vitro and with protection from albuminuria and mesangial expansion in db/db mice in vivo (114), suggesting exogenous C1P as a new therapeutic strategy to treat diabetic complications such as DKD. Although we have demonstrated that rituximab may target sphingolipid-associated disorders (51), novel therapeutic strategies specifically targeting proteins such as SMPDL3b remain to be developed. Therefore, it is becoming clear that the strategies of manipulating acid sphingomyelinases, SMPDL3b, sphingosine kinases, S1P lyase, glucosylceramide synthase, and GM3 synthase (as reviewed in Ref. 75) may translate into novel sphingolipid therapeutics for DKD and require further investigation.

CONCLUDING REMARKS

Here, we summarized the clinical and experimental evidence supporting a role of lipids in the modulation of renal cell insulin signaling and development of DKD. We also provide evidence of the complexity of IR signaling and of how specific isoform targeting may be needed in future studies. The complex role of IR isoforms A and B in insulin target organs is only partially understood. It is clear that IR isoforms are responsible for the complexity of insulin signaling diversification, which involves different ligand binding affinities, different membrane partitioning, different trafficking, and different abilities to interact with various molecular partners and to preferentially modify selected downstream signaling pathways. More evidence has appeared that dysregulation of the podocyte insulin signaling and perturbations in free fatty acids, cholesterol, and sphingolipid-related pathways play an important role in renal disorders and may contribute to the development and progression of DKD. Because of the knowledge that podocyte insulin signaling and lipid metabolism are connected, novel attractive therapeutic strategies can be developed. Our present understanding of these signaling cascades is incomplete, and future work in this area will be a step forward toward precision medicine in DKD.

GRANTS

A. Fornoni is supported by National Institutes of Health Grants DK-090316, DK-104753, U24-DK-076169, U54-DK-083912, UM1-DK-100846, and 1-UL1-TR-000460.

DISCLOSURES

A. Fornoni is an inventor on pending or issued patents (US10,183,038 and US10,052,345) aimed to diagnose or treat proteinuric renal diseases and stands to gain royalties from their future commercialization. A. Fornoni is also Chief Scientific Officer of L&F Health LLC and is a consultant for ZyVersa Therapeutics. ZyVersa Therapeutics has licensed worldwide rights to develop and commercialize hydroxypropyl-β-cyclodextrin for treatment of kidney disease from L&F Research. The patent associated with the use of hydroxypropyl-β-cyclodextrin is published under US10,195,227. A. Fornoni is Chief Medical Officer of LipoNexT, LLC. A. Mitrofanova and M. A. Sosa have no competing interests.

AUTHOR CONTRIBUTIONS

A.M. prepared figures; A.M. drafted manuscript; M.A.S. and A.F. edited and revised manuscript; A.F. approved final version of manuscript.