Emerging Targets in Type 2 Diabetes and Diabetic Complications

Diabetes mellitus is a chronic, metabolic disorder characterized by abnormally high blood glucose levels known as hyperglycemia. The Greek word diabetes means to siphon or to pass through and the Latin word mellitus means sweet, referring to high sugar levels in the urines of patients with diabetes. The earliest mention of diabetes dates back to 1552 BC written on an Egyptian papyrus, making it one of the oldest diseases described in human history. Initial attempts for treating diabetes mainly focused on herbal extracts and dietary interventions. Patients with diabetes had very poor prognosis with very low quality of life and particularly it used to be a death sentence for children. It was not until the discovery of insulin in 1921 by Frederik G. Banting and Charles Best at the University of Toronto, when life‐saving treatments started to take off. [ 1 , 2 ] Later in 1923, Banting and John Macleod received the Nobel Prize in Physiology or Medicine for their discovery of insulin. Banting shared his winnings with his assistant Best, Macleod, on the other hand, shared it with James Collip, with whose help insulin was successfully purified. We can describe diabetes as a disease of insulin insufficiency or impaired insulin action. Mainly, two main types of diabetes exist: type 1 and type 2. Type 1 diabetes develops at early stages of life due to an auto‐immune disorder where the cells of the immune system attack the insulin producing β cells of the pancreas. Type 2 diabetes, however, develops later in life, due to systemic dysfunctions in metabolic homeostasis. Genetic background plays a critical role in predisposing individuals to type 2 diabetes, where unhealthy eating habits and sedentary life style act as powerful triggers. [ 3 , 4 ] Unlike type 1 diabetes, type 2 diabetes is relatively heterogeneous and very complex, involving too many pathophysiological mechanisms that not only affect pancreas but also the metabolic organs, making effective treatment very challenging. In 2018, Groop and colleagues stratified patients with type 2 diabetes into five different subgroups based on six variables: age at diagnosis, body‐mass index (BMI), insulin resistance, beta cell function, Hb1Ac levels, and glutamate decarboxylase antibodies. Each cluster represented a specific subset of patients with differing risk for particular diabetic complications, which were: 1) severe autoimmune diabetes; 2) severe insulin‐deficient diabetes; 3) severe insulin‐resistant diabetes; 4) mild obesity‐related diabetes; and 5) mild age‐related diabetes. [ 5 ] Similar to type 2 diabetes patients, individuals that are not yet diagnosed but are at a high risk of developing it, were also stratified into six different subgroups that could predict the complications such as diabetic kidney disease without rapid progression to overt type 2 diabetes. [ 6 ] These findings indicate that the pathophysiological variation between individuals already exists before type 2 diabetes develops. These findings by independent groups once again provide the evidence for heterogeneity and complexity of type 2 diabetes most likely due to aberrant regulation of different signaling pathways in different target tissues. For instance, it is possible that in severe autoimmune diabetes, defective immune system is responsible for development of type 2 diabetes, whereas in mild age‐related diabetes, pathways that play role in aging and cell senescence in β ‐cells might play a role. Dissecting these tissue‐specific signaling pathways and identifying novel targets that contribute to type 2 diabetes will definitely, in the future, improve the current taxonomy of diabetes and contribute to precision medicine. Although the precise definition of sub‐clusters is still a matter of debate and it may take some time to establish protocols and categorize the patients, such stratification will certainly contribute to identify patients that are at risk for developing type 2 diabetes and diabetic complications, which will lead to personalized diabetes therapies, which unfortunately do not exist yet. Diabetes is a global endemic. In 2019, 463 million of adults (20–79 years old) were living with diabetes; and again–only in–2019, diabetes caused 4.2 million deaths. The number of patients with diabetes is increasing at a very high rate, estimated to reach 700 million by 2045. Diabetes is not only about high blood glucose levels. Patients with diabetes also suffer from a number of complications, which are sometimes already present when diabetes is diagnosed such as diabetic retinopathy; or they develop later during the course of the disease. [ 7 , 8 ] These complications involve dysfunctions in many vital organs all over the body; mainly kidney, cardiovascular system, retina, and the nervous system. Fibrosis of the liver and fibrosis of the lungs as well as cognitive dysfunction are also emerging as novel pathologies that develop secondary to diabetes. In this review, we will introduce the nove

Type 2 diabetes is a very heterogenous and complex disease that develops due to aberrant regulation of many signaling pathways. In this section, we will describe how insulin resistance develops in metabolic organs and what glucagon does in return. 3.1 Insulin Resistance in Liver Under conditions of over nutrition, high blood glucose levels oblige pancreas to produce and secrete more insulin. Constitutive activation of insulin signaling pathway at target tissues due to increased and sustained insulin levels, initiates several negative feedback loops, putting brakes on the initial steps of insulin signaling, contributing to pathological condition known as insulin resistance. One of the well described negative feedback loops takes place in response to overnutrition‐induced constitutive mTORC1 activation which leads to inhibitory phosphorylations on IRS‐1 by S6K1. [ 20 ] mTORC1 itself also phosphorylates IRS1 to promote its proteasome‐dependent degradation. [ 21 ] IRS‐1, indeed, acts as a critical target where independent signaling pathways merge on to establish an insulin resistant state ( Figure 2 ). The effects of S/T phosphorylations on IRS1 function are multifactorial: First, these phosphorylations might impair the IRS1–IR interaction and impair IR induced IRS1 tyrosine phosphorylations. Indeed, hepatocyte specific deletion of IQGAP1 scaffolding protein, which enables IR and IRS‐1 to interact, induces insulin resistance and glucose intolerance in vivo. [ 22 ] Second, S/T phosphorylations on IRS1 also promote its ubiquitin dependent proteasomal degradation. [ 21 ] Hyperactivated mTORC1 signaling also contributes to insulin resistance by phosphorylating and stabilizing Grb10 adaptor protein, which impairs IR‐IRS1 interaction. [ 23 , 24 , 25 ] Similarly, suppression of cytokine signaling (SOCS) scaffolding proteins impair the IR‐IRS‐1 interaction and promote degradation of IRS‐1. [ 26 ] Figure 2 Insulin resistance at IRS‐1/2. Insulin receptor substrate‐1/2 (IRS‐1/2) is a critical target that can be phosphorylated by various kinases to regulate its interaction with insulin receptor (IR). As Grb10, SOCS, or IQGAP1 proteins impair IR‐IRS1/2 interaction; JNK, IKK β , S6K1, mTORC1, and PIM kinases phosphorylate IRS‐1/2 to promote its proteasomal degradation. Recently our lab has identified transforming growth factor‐ β stimulated clone 22 D4 (TSC22D4) as a novel regulator of insulin resistance and hyperglycemia in mouse models of type 2 diabetes. Interestingly hepatic TSC22D4 expression positively correlates with insulin resistance in obese patients and liver specific TSC22D4 knockdown in diabetic mice improves glucose homeostasis and insulin resistance. [ 27 ] In addition, inflammatory signals initiated by tumor necrosis factor α (TNF α ), interleukin (IL)‐1ß, and IL6 cytokines merge on inhibitor of nuclear factor kappa‐B kinase subunit beta (IKKß) and jun N‐terminal kinase (JNK) signaling pathways to induce inhibitory phosphorylations on IRS‐1. [ 28 , 29 ] PIM kinases are also emerging as novel kinases responsible for IRS1 S1101 phosphorylation, yet the implications of these in metabolic regulation still remains elusive. [ 30 ] Chronic accumulation of unfolded proteins results in inflammatory responses in the cells leading to metabolic diseases such as type 2 diabetes and obesity. Endoplasmic reticulum (ER) is essentially responsible for protein synthesis and protein folding. In response to accumulation of unfolded or misfolded proteins in ER lumen, ER stress leads to unfolded protein response (UPR) to prevent additional injury to the cell. [ 31 , 32 , 33 ] In liver, ER stress results in insulin resistance by impairing regulation of gluconeogenesis and lipogenesis. Each of the UPR proteins has distinct effects on metabolic gene regulation. cAMP‐responsive element binding protein hepatocyte specific, for instance stimulates gluconeogenesis. X‐binding protein 1 (XBP1), on the other hand, suppresses FoxO1 activation hence indirectly promotes sterol regulatory element binding protein 1c (SREBP1c) to activate lipogenesis. Protein kinase RNA‐like endoplasmic reticulum kinase (PERK)–eukaryotic initiation factor 2 (eIf2 α ) branch also activates SREBP1c, and additionally promotes activating transcription factor 4 (ATF4), which in turn stimulates hepatic lipogenesis. [ 34 ] Fibroblast growth factor 21 (Fgf21) expression is also upregulated upon ER stress via PERK‐eiF2 α ‐ATF4 branch of UPR. [ 33 ] Fgf21 counteracts ER stress; and by inhibiting lipogenic program, it stimulates glucose uptake in the cells and alleviates hyperglycemia. Chronic UPR may also result in apoptotic cell death via upregulation of C/EBP homologous protein (CHOP) which is regulated by ATF4 in liver. [ 35 ] CHOP reduces B‐cell lymphoma 2 (Bcl2) anti‐apoptotic mitochondrial protein expression leading to cytochrome c release and caspase‐3 activation. [ 36 ] Very recent findings have shown that vitamin D receptor (VDR) blunts ER stress and UPR in the li

In addition to liver, skeletal muscle also plays a vital role in reducing blood glucose levels by promoting its uptake upon insulin stimulation. [ 42 ] In skeletal muscle, glucose transporter 4 (GLUT4) is identified as the most abundant glucose transporter isoform. Although a small portion of GLUT4 can be found on the cellular membrane, around 80% of GLUT4 is located in GLUT4 storage vesicles (GSV). In the presence of insulin or upon exercise, GLUT4 translocates from GSVs to muscle cell surface to promote glucose uptake. [ 42 , 43 ] Upon glucose uptake, skeletal muscle cells either direct the glucose to glycolysis or use it for glycogen synthesis depending on their metabolic needs. GLUT4 contains specific motifs in the amino cytoplasmic domain (FQQI) and in carboxyl cytoplasmic domain (LL and TELEY) which modulate GLUT4 trafficking from GSVs. Although the roles of certain proteins (GGA, retromer, AP1, etc.) that are interacting with these specific motifs in this trafficking are well known, there are still gaps to complete for GLUT4 translocation machinery including GSV regulations. [ 44 ] In type 2 diabetes, reduction in insulin's ability to stimulate glucose uptake from peripheral tissues occurs due to the disruption of GLUT4 translocation to the cell surface. [ 45 ] Muscle GLUT4 emerges as a specific target upon insulin action because exercise‐modulated GLUT4 translocation remains unchanged in type 2 diabetes. [ 44 ] Since exercise—induced glucose uptake remains preserved in insulin resistant—skeletal muscle, exercise is suggested as a key therapy against metabolic diseases such as type 2 diabetes. Although PI3K‐Akt signaling axis is one of the major pathways activated upon insulin engagement, defects in Akt phosphorylation on both phosphorylation sites (S‐473 and Thr‐308) showed only minor effects toward phosphorylation of Akt downstream targets. [ 44 ] Other than the canonical PI3K‐Akt‐Rab axis, non‐canonical PI3K‐Ras‐related C3 botulinum toxin substrate 1 (Rac1) – p21‐activated kinase (Pak1) actin remodeling pathway emerges as an alternative axis for GSV translocation upon insulin engagement. [ 46 ] Accumulation of plasma free fatty acid also causes insulin resistance in skeletal muscle. [ 47 ] Palmitic acid and stearic acid are some examples of saturated long chain fatty acids (FAs) causing mitochondrial dysfunction and insulin resistance. In mitochondrial dysfunction, not only mitochondrial density decreases in insulin resistant people but also rate of ATP synthesis and oxygen consumption decrease. Elevated reactive oxygen species (ROS) levels resulting from accumulation of FA‐derived metabolites (i.e., diacylglycerol and ceramides), impair mitochondrial biogenesis and activate stress kinases impairing glucose uptake and insulin tolerance. [ 47 ] Hepatokines, liver derived hormones, are essential for liver‐muscle trafficking; and Apolipoprotein J (ApoJ) emerges as a novel hepatokine targeting muscle glucose metabolism and insulin sensitivity via low density lipoprotein‐related protein 2 (LRP2). [ 48 ] LRP2 is required for insulin‐induced insulin receptor internalization in skeletal muscle. Muscle specific LRP2 deficiency or hepatic specific ApoJ deficiency promotes glucose intolerance and insulin resistance. ApoJ KO mice have defective insulin signaling with reduced phosphorylation on its canonical targets such as insulin receptor, IRS1/2, Akt and Akt substrate of 160 kDa (AS160) in skeletal muscle. FGF21 and selenoprotein B are other examples of hepatokines directly affecting glucose and lipid metabolism in liver, muscle, and adipose tissue. LRP2 also binds to selenoprotein B and promotes its uptake in kidney. [ 48 ] Liver kinase B1 ( Lkb1 ) suppresses amino acid induced gluconeogenesis in the liver. Hepatocyte specific Lkb1 deletion showed increased levels of hepatic amino acid catabolism, inducing gluconeogenesis. Although Lkb1 deficiency increased levels of amino acids in liver, it decreased the levels of amino acids in plasma. This metabolic impairment disrupts protein homeostasis in skeletal muscle and contributes to metabolic disorders such as cachexia and sarcopenia. [ 49 ] Cachexia is a metabolic syndrome that involves extreme body weight loss and muscle wasting. Usually cachexia emerges as a complication of certain diseases such as cancer, AIDS, or chronic kidney disease. [ 50 ] In very rare cases, cachexia also represents itself as a complication of diabetes also known as diabetic neuropathic cachexia (DNC). The underlying molecular mechanisms that lead to DNC remain elusive. Unlike DNC, sarcopenia is more prevalent among patients with type 2 diabetes. Sarcopenia involves age‐related loss in muscle mass and function due to impaired protein metabolism, mitochondrial dysfunction, and cell death, causing inflammation and impairing skeletal muscle's ability to uptake glucose. Hence, sarcopenia has a bidirectional relationship with type 2 diabetes, that is, while it promotes pathogenesis of typ

When Banting and Best discovered insulin, they also noted that the pancreatic extracts contained hyperglycemic properties as well. In 1922, Kimball and Murlin successfully isolated the fraction that had only hyperglycemic effect and named it glucagon. [ 70 ] In 1948, Sutherland and de Duve showed that, it is the alpha cells of the pancreatic islets that produce glucagon. [ 70 ] Although skeletal muscle, heart, kidney, stomach, and small intestine are among the organs that express the glucagon receptor, glucagon exerts its metabolic effects mainly by liver. Glucagon receptor is a seven transmembrane receptor that belongs to G‐protein coupled receptor family. Binding of glucagon to hepatic glucagon receptor activates adenylate cyclase (AC) 5 and AC 6 increasing cellular cAMP levels, which act as a second messenger to activate protein kinase A (PKA) signaling ( Figure 4 ). [ 71 , 72 ] cAMP action in liver is very critical during fasting state to ensure glucose‐dependent tissues such as brain and red blood cells have sufficient glucose supply provided by liver. Figure 4 Glucagon signaling. Upon glucagon binding, GCGR activates adenylate cyclase that increases cAMP levels in the cytoplasm. cAMP activates PKA which phosphorylates CREB and leads to its translocation to nucleus. CREB forms a complex with CBP and CRTC2 to regulate gluconeogenic gene expression and fatty acid oxidation via targets such as PGC1 α , FoxO1, hepatic HNF4 α , FXR, and LXR. ECM: Extracellular matrix. One of the well‐characterized substrates of PKA is cAMP responsive element binding protein (CREB). CREB acts a transcription factor (binding to promoter regions of genes) that plays key role in gluconeogenesis such as pyruvate carboxylase (PC), phosphoenolpyruvat‐carboxykinase (PEPCK), glucose‐6‐phosphatase (G6Pase), and peroxisome proliferator‐activated receptor‐gamma coactivator‐1alpha (PGC1 α ) (Figure 4 ). [ 73 , 74 , 75 , 76 , 77 , 78 , 79 , 80 ] Recently, Krüppel like factor 9 (Klf9) was identified as a novel upstream regulator of PGC1 α . Glucocorticoids (i.e., dexamethasone) and fasting upregulates Klf9 gene expression and Klf9 itself binds to the promoter region of PGC1 α and acts a transcriptional activator. Interestingly liver specific Klf9 deficiency alleviates dexamethasone‐induced hyperglycemia, potentially revealing one of the mechanisms explaining how glucocorticoids might promote diabetes. [ 81 ] 4.1 Glucagon Signaling in Diabetes Unlike in healthy individuals where glucagon levels elevate under conditions of hypoglycemia, some patients with diabetes present increased blood glucagon levels despite hyperglycemia. Persistent exposure to glucagon creates an excessive burden on liver due to ongoing gluconeogenesis and glycogenolysis, which in turn exacerbates hyperglycemia and eventually creating a vicious cycle, contributing to pathological condition known as insulin resistance. Indeed, approaches antagonizing glucagon signaling as well as studies in glucagon receptor knockout mice lead to promising results in which blood glucose levels decreased while glucose tolerance and insulin sensitivity improved. [ 82 , 83 , 84 , 85 ] For instance, inhibition of glucagon receptor (GCGR) via overexpression of β ‐arrestin 2 alleviated metabolic defects in HFD fed mice. [ 86 ] β ‐arrestins bind to GCGRs when GCGRs undergo multiple phosphorylation events by glucagon receptor kinases (GRKs) and inhibition of GCGR signaling via β ‐arrestins involve at least two mechanisms. β ‐arrestins can either impair the interaction between GCGR and G proteins or promote the internalization of GCGRs via clathrin‐mediated endocytosis leading to a desensitization mode (Figure 4 ). [ 87 ] The barcode hypothesis of GCGR refers to the long‐standing idea that specific phosphorylation events on GCGR direct the interaction with corresponding β ‐arrestins, induce different conformational changes on β ‐arrestins and dictate which signaling molecules they will recruit to initiate corresponding signaling pathways. Recent studies with atomic‐level simulations and site‐directed spectroscopy showed that the barcode hypothesis might indeed be a valid one and point out that it is not the number of phosphorylation events per se but the position of phosphorylated residues that act as barcodes. [ 88 ] Although anti‐glucagon approaches in several independent labs alleviated diabetic symptoms in mouse models, developing therapies that target glucagon has been challenging due to its lipogenic potential. Very recently, a high throughput screening of 300.000 compounds led to the discovery of SRI‐37330, an orally bioavailable small molecule. SRI‐37330 treatment ameliorated diabetes both in type 1 and type 2 diabetes mouse models. SRI‐37330 impaired thioredoxin‐interacting protein (TXNIP) function in pancreas, which in turn impaired glucagon secretion from the alpha cells, contributed to lower blood glucose levels. [ 89 ] Unlike glucagon receptor antagonists, SRI‐37330 did not have any

“Science, has been built upon many errors; but they are errors which it was good to fall into, for they led to the truth” said once the ingenious and talented French novelist Jules Verne, who himself developed type 2 diabetes in his fifties and unfortunately suffered miserably due to the diabetic complications in his late years. [ 113 ] Diabetes is hardly a disease of mere elevation in blood glucose levels. In most cases, it brings along a plethora of complications in peripheral tissues such as kidneys, cardiovascular system, retina, the nervous system, and liver. Although the symptoms and the indications of these pathologies are quite well characterized, the underlying molecular mechanisms remain elusive. Hence, the existing therapies are not always as effective. The usual suspect leading to diabetic complications would be hyperglycemia; yet studies indicate that strict blood glucose control does not always prevent the progress of these pathologies let alone reversing it. [ 114 , 115 ] Large interventions trails such as UKPDS, VADT, ACCORD, and ADVANCED with glucose‐lowering approaches presented evidence for statistically significant reductions in relative risks for developing some of the diabetic complications, the rate of absolute risks however remained relatively small, such as a reduction of 0.28% for microvascular complications or 0.04% reduction in diabetic kidney disease in the UKPDS study. [ 115 , 116 ] Multifactorial interventions targeting hypertension, dyslipidemia, and microalbuminuria, along with hyperglycemia, on the other hand, were much more effective in reducing diabetic complications in Steno‐2 study. [ 117 , 118 ] In addition to strict glycemic control, insulin sensitizers were also associated with a reduction in diabetic complications but only with a 1.5% of absolute risk ratio for cardiovascular mortality and with a 1.8% absolute risk ratio for cardiovascular events in the case of pioglitazone. [ 119 ] These data also raise the question how much hyperglycemia and insulin resistance play role in development of diabetic complications and whether they are just epiphenomena, that is, they might be symptoms of type 2 diabetes but might not necessarily contribute to its pathogenesis. In addition to hyperglycemia, deregulation of other cellular activities such as generation of the reactive metabolites may contribute to development of diabetic late complications. According to Brownlee hypothesis a.k.a unifying hypothesis, hyperglycemia elevates ROS levels, which modify and impair the glycolytic enzyme glyceraldehyde 3‐phosphate dehydrogenase (GAPDH). Inhibition of glycolysis via inhibition of GAPDH diverts the upstream metabolites from glycolysis to glucose overutilization pathways, which are the following: 1) the polyol pathway; 2) the protein kinase C (PKC) pathway; 3) advanced glycation end product (AGE) formation pathway; and 4) the hexosamine pathway. [ 120 ] These pathways lead to mitochondrial dysfunction and elevate ROS levels even further, contributing to disease progression and represent the root cause of diabetic complications. [ 120 ] A major drawback of this hypothesis is the fact that ROS have a very short half‐life and spatially very limited actions. Although there are some studies that show patients with diabetes have elevated ROS, ROS levels do not necessarily change between patients with and without diabetic complications. [ 121 ] Although mitochondrial impairment is relevant in terms of pathogenesis of diabetes and diabetic complications, the evidence for ROS‐induced mitochondrial dysfunction that leads to diabetic complications also remains elusive. As both experimental and clinical approaches fail to provide solid and consistent evidence to support Brownlee hypothesis, researchers are investigating alternative pathways or metabolites that might play role in diabetic complications. Methylglyoxal (MG) represents one of these reactive metabolites; the levels of which increase upon hyperglycemic flux and impaired detoxification. One of the enzymes that play role in MG detoxification is Glyoxalase 1 (Glo1). Glo1 knockout flies have elevated levels of MG, which induces type 2 diabetes like phenotype such as insulin resistance, obesity, and hyperglycemia. [ 122 ] Similarly, Glo1 knockout together with diet‐induced obesity elevates MG levels and induces type 2 diabetes like symptoms in zebrafish. [ 123 ] In support of these findings, MG is also sufficient to induce retinopathy like lesions in rat models without inducing hyperglycemia, [ 124 ] suggesting that accumulation of MG is creating a shortcut to develop diabetes‐like phenotype in the absence of hyperglycemia. In addition to Glo1, MG can also be metabolized either by aldo–keto reductases (AKR) to hydroxyacetone or by aldehyde dehydrogenase (ALDH) to pyruvate. Compensatory MG detoxification by increased AKR and ALDH activities is more relevant in mammals, as unlike in Drosophila and zebrafish, loss of Glo1 do not elevate MG l

Diabetic retinopathy is a common complication of diabetes. Almost 20% of the patients have diabetic retinopathy at the time of diagnosis with diabetes and overall 40–45% of the patients develop retinopathy during the course of the disease. Diabetic retinopathy involves dysfunction in two main cell types of the retina: endothelial cells of the retinal microvasculature and the pericytes that lie beneath the endothelial cells to support and regulate endothelial cell function. Briefly, hyperglycemia, oxidative stress and AGEs, impair the tight junctions between the endothelial cells and induce detachment and apoptosis of pericytes. Pericyte loss is one of the very early pathologies in diabetic retinopathy, which renders these cells an important target for early interventions to prevent the further progress of the disease. Hyperglycemia leads to detachment of pericytes from the endothelial cells, which eventually leads to apoptosis and increases the blood‐retina barrier permeability. Signaling pathways that contribute to pericyte loss include Notch 1, Notch 3, hypoxia inducible factor 1 α (HIF1 α ), and VEGF‐1 ( Figure 8 ). [ 181 , 182 ] Figure 8 Diabetic retinopathy. Endothelial cells and pericytes are the two regulators of diabetic retinopathy. Hyperglycemia and oxidative stress cause pericyte detachment from the endothelial cells via Notch1/3, HIF1 α , and VEGF‐1 signaling pathways. Anti‐VEGF‐1 therapies are used to inhibit detachment of pericytes. Glial cells express Sema4d during hypoxia and upon Sema4d binding to its receptor Plexin B1 in pericytes, mDia/Src pathway gets activated. Activated Src promotes VE‐cadherin internalization and loosens the tight junctions between endothelial cells. Ang1‐Tie2 binding also impairs Src function, while Ang2 inhibits Ang1‐Tie interaction. cPWWP2A circular RNA downregulates miR579, which in turn promotes Ang1 expression. Hyperactive VEGF‐1 signaling contributes significantly to the progression of diabetic retinopathy by inducing highly unstructured, disorganized neovascularization of endothelial cells. Anti‐VEGF‐1 therapies have been very effective to delay the disease progression. Yet, not all patients respond to VEGF‐1 treatment equally. Recent findings indicate that Semaphorin 4d (Sema4d) levels in the body fluids can successfully predict whether patients will respond to anti‐VEGF1 therapy or not. [ 183 ] Non‐ or little responders of the anti‐VEGF‐1 therapy have elevated levels of Sema4d in the aqueous fluid of the eye. Sema4d not only acts as a biomarker but also plays a significant role in progression of diabetic retinopathy. Indeed, a combination therapy of anti‐VEGF1 and anti‐Sema4d might be a better alternative compared to only anti‐VEGF1 treatments. Sema4d is a membrane bound protein whose expression elevated upon hypoxia in the retinal glial cells. Once shedded at the cell membrane by ADAM17, Sema4d binds to its receptor PlexinB1 at the surface of pericytes and endothelial cells, which activates downstream signaling of mDia1/Src pathway. Activation of Src contributes to phosphorylation and internalization of vascular endothelial cadherin (VE‐Cadherin) which loosens the tight junctions and contributes to vascular leakage exacerbating the diabetic retinopathy (Figure 8 ). [ 183 , 184 ] Hence, Sema4d sets a nice example of how the crosstalk between retinal glial cells and the pericytes holds critical function to maintain a healthy vasculature in the eye. Other upstream regulators of Src include Angiopoietin 1 (Ang1), which is expressed and secreted by pericytes and binds to its receptor Tie2 on endothelial cells. Activated Ang1/Tie2 signaling promotes TGF β and platelet‐derived growth factor (PDGF) signaling in endothelial cells, which stabilize the intercellular interactions. Ang1/Tie2 signaling also impairs Src function to promote the tightening of cell–cell junctions between epithelial cells blunting vascular hyperpermeability. [ 184 ] Ang2, on the other hand, acts as an antagonist to blunt the Ang1/Tie2 signaling and promote blood retina barrier permeability (Figure 8 ). [ 185 ] Similar to glial cell–pericyte crosstalk, pericyte–endothelial cell crosstalk is also essential for a healthy retinal vasculature. One of the newly identified mediators of pericyte–endothelial cell crosstalk is the circular RNA called cPWWP2A. Circular RNAs are a group of non‐coding RNAs with a closed loop structure and usually act as sponges to downregulate the action of their target microRNAs. High glucose levels upregulate the expression of cPWWP2A in pericytes, which acts as a sponge to impair miR‐579 function and upregulate Ang1/Occludin/SIRT1 expression. cPWWP2A is also packed in exosomes and secreted into the pericyte medium to regulate the proliferation, migration, and tube formation of retinal endothelial cells. [ 186 ] Similar to cPWWP2A, cZNF532 is another novel circular RNA that plays role in controlling pericyte function and vascularization. cZNF532 acts as a sponge to dow

Non‐alcoholic steatohepatitis (NASH) and liver fibrosis are also emerging as late complications of diabetes. Under normal conditions, liver is great at handling acute stress conditions and regenerate when required. Apoptosis of damaged cells is a crucial part of this regeneration process, which needs to be under control for healthy liver function. Constant exposure to hyperglycemia, insulin resistance, and excessive lipid accumulation, on the other hand, induce a chronic inflammatory state where lipotoxicity and oxidative stress contribute to development of NASH from a relatively benign state of non‐alcoholic fatty liver disease (NAFLD). Unlike NAFLD, NASH is hardly reversible which progresses further into fibrosis and eventually cirrhosis, when not managed properly. Currently there are no FDA approved therapies to treat NASH and/or liver fibrosis, which represent the most common cause of liver transplants worldwide. Mitochondrial non‐coding RNAs are recently identified as contributing factors to NASH development. Steatohepatitis‐associated circRNA ATP5B regulator (SCAR), which is located in mitochondria, inhibits mitochondrial ROS output and fibroblast activation. [ 207 ] AMPK; a central regulator of cell metabolism, also plays a crucial role in maintaining hepatic homeostasis. NASH development suppresses the function of AMPK, which under normal conditions phosphorylates and inhibits caspase 6 activity. Downregulation of AMPK during NASH leads to hyperactivation of Casp 6 causing excessive apoptosis in the liver tissue which exacerbates inflammation and liver injury. [ 208 ] One of the key mediators of liver inflammation and injury is JNK1. JNK1 induces the expression of pro‐inflammatory cytokines and chemokines such as IL‐6, MCP‐1 via its targets c‐jun and c‐fos. Apoptosis signal‐regulating kinase 1 (ASK1) represents one of the critical upstream activators JNK1. Ask1 homodimerization is indispensable for its activity, which is impaired by direct binding of Casp 8 and FADD like apoptosis regulator (CFLAR) protein. Interestingly hepatic CFLAR expression is reduced during metabolic syndrome and NASH. Adeno associated virus (AAV) mediated hepatic reconstitution of CFLAR improves glucose tolerance and alleviates liver fibrosis both in mice and monkeys, rendering it an attractive target for development of novel therapies against NASH. [ 209 ] In addition to hepatocytes, hepatic stellate cells also play an important role in development of NASH and liver fibrosis. When stimulated by inflammatory cytokines or growth factors such as TGF β 1, stellate cells get activated and undergo epithelial to mesenchymal transition, gain fibroblastic features and start to proliferate. Upon activation, stellate cells also increase the production of extracellular matrix components such as collagen contributing to liver stiffness and fibrosis. The crosstalk between hepatocytes and stellate cells exhibit a major factor that accelerates fibrotic process. TAZ protein, for instance, initiates a signaling cascade to activate the expression of secretory factor Indian Hedgehog (Ihh). Ihh secreted from hepatocytes binds to smoothened receptor on stellate cells and induces the expression of pro‐fibrogenic genes and promotes proliferation. TAZ silencing in hepatocytes delays the progression of NASH and alleviates inflammation. [ 210 ]

Although diabetes mellitus is one of the earliest described diseases of the human history, there is still no cure for it. Currently the existing therapies for type 2 diabetes target reducing blood glucose levels and alleviating the symptoms of accompanying complications. Although there are cases where bariatric surgeries, intermittent fasting, or certain diets such as ketogenic diet improve type 2 diabetes; these interventions are either highly invasive or the diet regimens are hard to follow up in the long run, respectively. Thus, main research efforts should aim for the development of novel preventive and therapeutic concepts. This will likely include the more thorough investigation of SGLT2 inhibitors and GLP1 receptor agonists, which show effective clinical outcomes not only in reducing blood glucose levels, but also in alleviating the diabetic complications particularly in cardiovascular system and kidney. [ 216 , 217 ] Furthermore, the clinical validation of uni‐molecular, dual agonists, islet cell replacement as well as novel RNA‐based therapies for tailored diabetes treatment will certainly contribute to more efficacious therapies in type 2 diabetes and related complications. Emerging insights into mechanisms of diabetic long‐term complications that go beyond the simple glucose‐centric view and incorporate as‐yet unexplored organ complications will be the basis for intensified research efforts to prevent or even reverse long‐term complications. In addition to a better understanding of mechanisms playing role in pathogenesis of type 2 diabetes, patient stratifications based on these very pathogenic mechanisms will pave the way to more effective treatments for type 2 diabetes and its complications. Major progress in these areas will eventually move us closer to our vision to make type 2 diabetes a livable and most importantly a curable disease in the future.