Vitamin A and Retinoid Signaling in the Kidneys

Abstract

Vitamin A (VA, retinol) and its metabolites (commonly called retinoids) are required for the proper development of the kidney during embryogenesis, but retinoids also play key roles in the function and repair of the kidney in adults. Kidneys filter 180-200 liters of blood per day and each kidney contains approximately 1 million nephrons, which are often referred to as the ‘functional units’ of the kidney. Each nephron consists of a glomerulus and a series of tubules (proximal tubule, loop of Henle, distal tubule, and collecting duct) surrounded by a network of capillaries. VA is stored in the liver and converted to active metabolites, most notably retinoic acid (RA), which acts as an agonist for the retinoic acid receptors ((RARs α, β, and γ) to regulate gene transcription. In this review we discuss some of the actions of retinoids in the kidney after injury. For example, in an ischemia-reperfusion model in mice, injury-associated loss of proximal tubule (PT) differentiation markers occurs, followed by re-expression of these differentiation markers during PT repair. Notably, healthy proximal tubules express ALDH1a2, the enzyme that metabolizes retinaldehyde to RA, but transiently lose ALDH1a2 expression after injury, while nearby myofibroblasts transiently acquire RA-producing capabilities after injury. These results indicate that RA is important for renal tubular injury repair and that compensatory mechanisms exist for the generation of endogenous RA by other cell types upon proximal tubule injury. ALDH1a2 levels also increase in podocytes, epithelial cells of the glomeruli, after injury, and RA promotes podocyte differentiation. We also review the ability of exogenous, pharmacological doses of RA and receptor selective retinoids to treat numerous kidney diseases, including kidney cancer and diabetic kidney disease, and the emerging genetic evidence for the importance of retinoids and their receptors in maintaining or restoring kidney function after injury. In general, RA has a protective effect on the kidney after various types of injuries (eg. ischemia, cytotoxic actions of chemicals, hyperglycemia related to diabetes). As more research into the actions of each of the three RARs in the kidney is carried out, a greater understanding of the actions of vitamin A is likely to lead to new insights into the pathology of kidney disorders and the development of new therapies for kidney diseases.

1). Introduction: An Overview of Retinoid Signaling

Retinoids are defined as chemicals structurally and functionally related to vitamin A (VA, retinol). The term ‘retinoid’ includes both biologically active metabolites of vitamin A and synthetic agonists with actions in cells that are similar to those of the endogenous agonist and metabolite of vitamin A, all-trans retinoic acid (RA) (L. J. Gudas, 2022b). Most, but not all of the actions of retinoids (Al Tanoury, Piskunov, & Rochette-Egly, 2013) are carried out by the three retinoic acid receptors (RARs) α, β, and γ, which bind the endogenous ligand/agonist RA, the major endogenous ligand reported in the literature (Petkovich & Chambon, 2022). These RARs are members of the nuclear receptor family of transcription factors, and the genes that encode these three receptors are well conserved across vertebrates (Weikum, Liu, & Ortlund, 2018).

To bind to DNA, these RARs form heterodimers with one of the three RXRs α, β, and γ (Brand, et al., 1988). The RARs and RXRs are comprised of multiple domains. The DNA binding domains of these RARs and RXRs are responsible for the specificity of binding of these RAR/RXR heterodimers to DNA response elements called retinoic acid response elements (RAREs) near their primary target genes (Langston, Thompson, & Gudas, 1997; Penvose, Keenan, Bray, Ramlall, & Siggers, 2019; Rastinejad, 2001). The 12-helical ligand binding domains of these RARs and RXRs bind lipophilic signaling molecules, such as RA, in hydrophobic pockets. They are also involved in the recruitment of transcriptional coregulatory proteins in a ligand-dependent manner (Chandra, et al., 2017; Cheung & Kraus, 2010; Glass & Rosenfeld, 2000). Such coregulatory proteins include coactivator proteins that provide surfaces for interactions with histone modifying enzymes so that the RAREs of RAR target genes in chromatin can be accessed by the RARs (Fig. 1).

Figure 1.

Retinoic acid receptors (RARs) α, β, γ bind as heterodimers with RXRs α, β, γ. In the presence of an agonist, RAR/RXR heterodimers bind to and transcriptionally activate or repress genes with RAREs (retinoic acid response elements/enhancers). Shown here is transcriptional activation in the presence of retinoic acid (RA), an endogenous RAR agonist. HDAC, histone deacetylase; HAT, histone acetyl-transferase; SRC-1 (steroid receptor coactivator-1); pol II, RNA polymerase II; other proteins shown are part of the transcription activation complex. RARβ2, one isoform of RARγ, possesses a DR5 RARE in its promoter so one of the RARs is itself a primary target gene activated by RA.

If all three RARs can use RA as an endogenous agonist, then why are there three RARs rather than just one? The RAR α, β, and γ genes are surrounded by different DNA regulatory sequences so that each RAR gene can be expressed in various cell types and at different times during development and in the adult animal. Notably, the functions of specific RXR/RAR α, β, and γ heterodimers are needed at numerous developmental stages for proper embryonic patterning and organogenesis to take place during development (Mark, Ghyselinck, & Chambon, 2006). Furthermore, retinoid signaling through different receptors can have distinct outcomes, so the receptors are not redundant (Mark, et al., 2006).

Of course, the molecular actions of RA in cells, while a key feature of RA signaling, are only part of this signaling process. The uptake and metabolism of VA are complex and are just covered briefly here. VA must be obtained in the diet, and after being taken up by intestinal epithelial cells and esterified with long chain fatty acids to retinyl esters in the intestinal epithelia, much of the ingested VA is delivered to its target cells throughout the body via chylomicrons that contain other neutral lipid esters. These chylomicrons that contain retinyl esters move primarily through the lymphatics, which play an important role in the immune system. ~60% of the VA from the diet absorbed in the intestine is released by the intestinal enterocytes and secreted into intestinal lymph and ~30% is released into the hepatic portal vein, respectively (Hollander, 1981). After lipolytic processing to chylomicron remnants in the blood, these chylomicron remnants are taken up by hepatocytes in the liver, where much of the body’s VA is stored (Harrison, 2005). Notably, ApoE is required for the liver to take up these chylomicron remnants (Shirakami, Lee, Clugston, & Blaner, 2012). Because most cells of the body continuously require vitamin A, VA is then released from the liver as needed and is carried through the blood stream bound to serum retinol binding protein, RBP4. RBP4 maintains a constant level of VA in the blood (Harrison, 2005). During homeostasis, RBP4 circulates in a complex with transthyretin, a 55 kDalton molecular mass protein, to prevent loss of RBP4 and vitamin A from the body (Henze, et al., 2008). Vitamin A bound to the serum retinol binding protein, RBP4, is reabsorbed in the proximal tubules via the major endocytic receptor localized at the apical surfaces of proximal tubules, megalin (Lrp2) (Christensen, et al., 1999; Jing, et al., 2016; Raila, Willnow, & Schweigert, 2005). VA is then transported into at least some types of VA target cells via STRA6, a transmembrane protein with very high expression in the eye and placenta (Berry, et al., 2013; Y. Chen, et al., 2016; Kelly & von Lintig, 2015; Laursen, Kashyap, Scandura, & Gudas, 2015).

In contrast, after bacterial infection, a very different VA delivery mechanism is activated. Serum amyloid A (SAA) proteins, produced in large quantities by the liver, bind VA and transport VA to VA target cells, including intestinal myeloid cells (dendritic cells and macrophages). The VA, bound to SAA proteins, binds LRP1 (LDL receptor-related protein 1) and is then endocytosed and further metabolized by these myeloid cells to RA (Bang, et al., 2021). This increased production of SAA proteins by the liver is part of the acute phase response which occurs after local or systemic inflammation from bacterial infections, and even occurs in the context of chronic inflammation from tissue damage or tissue dysfunction (Mantovani & Garlanda, 2023).

As discussed above, RA, rather than vitamin A itself, is the major, endogenous agonist for the RARs. RA can additionally be further metabolized to 4-oxo-RA and 4-OH-RA. The metabolism of vitamin A within cells is complex and is not the focus of this review. However, a diagram of vitamin A metabolism is shown (Fig. 2). Recent reviews of vitamin A metabolism are provided (Czuba, et al., 2021; L. J. Gudas, 2022a; Kedishvili, 2013; O'Connor, Varshosaz, & Moise, 2022).

Figure 2.

Metabolism of retinol (vitamin A) to retinyl esters via LRAT (lecithin:retinol acyltransferase), and conversely, metabolism of retinyl esters to retinol by retinyl ester hydrolases (REHs). Metabolism of retinol to retinaldehyde by retinol dehydrogenases, primarily RDH10. Conversely, retinaldehyde can be converted to retinol via DHRS3 + RDH10. Retinaldehyde is metabolized to retinoic acid (RA) by aldehyde dehydrogenases 1a1, 1a2, and 1a3 (formerly called RALDH1,2,3). This is an irreversible reaction. RA can then be further oxidized to 4-hydroxy-RA and 4-oxo-RA by cytochrome P450 26a1, b1, and c1.

2). Functions of Retinoids in the Normal Kidney

a). Development of the Kidney

During development, vitamin A is obtained from the maternal circulation through the placenta (Quadro & Spiegler, 2020). Vitamin A is required for the proper development of the kidney. In 1953 Wilson et al. noted that maternal vitamin A deficiency resulted in renal hypoplasia in the gestating pups (Wilson, Roth, & Warkany, 1953). When rodents are given modestly limited vitamin A (a 50% decrease in circulating concentrations) during embryonic development, a 20% reduction in nephron number results (Bhat & Manolescu, 2008; Gilbert & Merlet-Bénichou, 2000; Lelièvre-Pégorier, et al., 1998). Kidney agenesis (bilateral absence of kidney tissue) is present in the majority of RARα^−/−; RARγ^−/− knock out’ murine mutants (Mendelsohn, et al., 1994). Moreover, knock out of all three RARs at embryonic day 10.5 in mice results in small, ectopic kidneys at embryonic day 12.5 (Mark, et al., 2021). The enzyme, ALDH1a2, which converts retinaldehyde to RA (Fig. 2) is highly expressed throughout the kidney nephrogenic zone during development (Niederreither, Fraulob, Garnier, Chambon, & Dollé, 2002), also indicating the importance of RA signaling in this zone. Notably, a human ALDH1a2 gene variant with greater activity has been associated with greater kidney size and higher serum RA levels (El Kares, et al., 2010).

The nephron is the functional unit of the kidney. RARα^−/−;RARβ2^−/− knock out mutants, when born, possess small kidneys with reduced numbers of nephrons and they lack the nephrogenic zone where new nephrons are generated. RA stimulates stromal cells that express RARα and β, which increase c-Ret expression in the ureteric bud (Batourina, et al., 2001; Mendelsohn, Batourina, Fung, Gilbert, & Dodd, 1999). GDNF (Glial cell ine-derived neurotrophic factor, produced in the metanephric mesenchyme, drives ureteric bud branching via the tyrosine kinase receptor c-Ret which is expressed in cells of the ureteric bud and confers sensitivity to the GDNF signal (Shakya, Watanabe, & Costantini, 2005). Thus, c-Ret expression controls renal branching morphogenesis and ultimately, the final number of nephrons in the kidney. Expression of c-Ret in the ureteric bud is very sensitive to signals dependent on RA that come from adjacent mesenchymal cells (Batourina, et al., 2001). In RARα^−/−;RARβ2^−/− knock out mutants, c-Ret is expressed at a reduced level, leading to defective branching morphogenesis of the ureteric bud (Batourina, et al., 2001; Rosselot, et al., 2010). A RA-responsive, secreted branching morphogen, late gestation lung protein one (LGL1), is involved in this branching process (Quinlan, Kaplan, Sweezey, & Goodyer, 2007), and other RA-inducible transcripts in the embryonic kidney have been identified (Takayama, Miyatake, & Nishida, 2014). In humans, dominant mutations in the gene nuclear receptor interacting protein 1 (NRIP1), a RA-inducible, transcriptional cofactor that directly interacts with RARs, can cause CAKUT (congenital anomalies of the kidney and urinary tract) by interfering with RA signaling during development (Vivante, et al., 2017). Mutations in the gene ‘growth regulation by estrogen in breast cancer 1-like’ (GREB1L), a target of RA signaling, can have a similar effect (De Tomasi, et al., 2017). Mutations in GREB1L can even result in kidney agenesis in mouse models (De Tomasi, et al., 2017). These molecular signaling processes explain, in part, why low vitamin A levels during embryogenesis result in a reduction in nephron number in the adult. More research is needed, though, to define the branching process in more detail in terms of the various morphogens involved.

Vitamin A deficiency in pregnant women may be related to an increased risk of developing primary hypertension later in life because of reduced nephron numbers (Hoy, et al., 2008). In humans, nephron production is completed by birth, so that, after birth, the nephrons can only enlarge (Hinchliffe, Sargent, Howard, Chan, & van Velzen, 1991). In humans, nephron numbers vary greatly, from 0.3-1.3 million nephrons/kidney (A. T. Clark & Bertram, 1999; Nyengaard & Bendtsen, 1992). In various animal models and in humans, maternal protein restriction can have serious consequences later in life for the offspring, including hypertension and greater susceptibility to renal injury, in part because of a reduction in nephron numbers in the offspring (Langley-Evans, Welham, & Jackson, 1999; Poladia, et al., 2006; Zimanyi, et al., 2006). Maternal protein restriction in rats is associated with an ~30% reduction in nephron numbers in their pups (Makrakis, Zimanyi, & Black, 2007). Remarkably, one bolus dose of RA at embryonic day 11.5 restored nephron number to the normal range in pups exposed to maternal protein restriction (Makrakis, et al., 2007). Taken together, these studies in rodents and humans indicate that adequate maternal vitamin A during gestation is critical for optimal numbers of nephrons to be formed and suggest that maternal protein restriction may also reduce the amount of vitamin A delivered to the developing embryo.

b). Actions of Retinoids in the Adult Kidney

While we have some knowledge of the roles of RA signaling during kidney development, the actions of RA in post-birth tissue maintenance and repair are less well understood. The kidneys act as the filtration system of the body by removing waste and toxins from blood while reabsorbing water and important solutes back into circulation via the nephrons (Bertram, Douglas-Denton, Diouf, Hughson, & Hoy, 2011; McMahon, 2016). Each nephron consists of a glomerulus and a series of tubules (proximal tubule, loop of Henle, distal tubule, and collecting duct) surrounded by a network of capillaries (McMahon, 2016) (Fig. 3).

Figure 3.

Overview of nephron. Blood travels via the afferent arteriole into a bundle of capillaries called the glomerulus. Materials that are smaller than 100nm are filtered out of the glomerulus and directed into the proximal tubules. The filtered fluid and materials travel through the loop of Henle, consisting of a thin descending loop, a think ascending loop, and a thick ascending loop, the distal tubule, and finally the collecting duct. Many materials, such as sodium, potassium, bicarbonate, and water, are reabsorbed throughout this process. The mesangial cells and podocytes of the glomerulus, and proximal tubule epithelial cells are all known to be responsive to retinoic acid (RA).

In the glomerulus, fluid and proteins that are less than 100 nm in diameter can pass through slit diaphragms between podocytes into the Bowman’s Capsule, a membrane surrounding the glomerulus, where they are then directed into the proximal tubule (Arif & Nihalani, 2013; Deen, Lazzara, & Myers, 2001; Poliak, Quaggin, Hoenig, & Dworkin, 2014). Of the three RARs, podocytes primarily express RARα, though RARβ and γ are also expressed in podocytes (Agrawal, He, & Tharaux, 2021; X. P. Chen, et al., 2014; Ratnam, et al., 2011a).

Proximal tubules (PTs) are the most abundant cell type in the mammalian nephron, constituting 80 % of the cells of the renal cortex, and are made up of three segments that vary in ultrastructure (i.e., brush border thickness and mitochondrial content) (J. Z. Clark, et al., 2019; Andrew M. Hall, Polesel, & Berquez, 2021; Zhuo & Li, 2013). In adult mice, RAR α, β, and γ are normally expressed in the kidney proximal tubules (Nakamura, et al., 2019) (DiKun, Tang, Fu, Choi, Lu, Gudas, submitted for publication).

3). Retinoids in Kidney Diseases

a). Diabetic Kidney Disease (DKD)

Diabetic kidney disease (DKD) (also known as diabetic nephropathy) is the major microvascular problem associated with diabetes mellitus (DM). DKD occurs in about 40% of people with diabetes and often results in kidney failure, cardiovascular dysfunction, and early death (Oshima, et al., 2021). Since DM now affects ~ 11% (~35 million people) of the US population (Alicic, Rooney, & Tuttle, 2017; Koye, Magliano, Nelson, & Pavkov, 2018), DKD is a major medical issue, often resulting in kidney transplantation and dialysis. Glomerular hyperfiltration, glomerular and tubular hypertrophy, increased lipid production, lipotoxicity, mitochondrial dysfunction, and thickening of the glomerular basement membrane lead to clinical symptoms such as macroalbuminuria and a major decline in the glomerular filtration rate (GFR) in humans and in mouse models of DKD (Deji, et al., 2009; Andrew S. Levey, Grams, & Inker, 2022).

Because the kidneys of mammals have no portal blood supply, any vasoconstriction in the glomeruli may reduce blood supply throughout the kidney. Proximal tubule (PT) cells are quite sensitive to ischemia (Berger, et al., 2014), a consequence of this lack of portal blood supply and of the fact that blood flow in the kidney is high, constituting ~25% of cardiac output. Extensive tubulointerstitial hypoxia occurs in both early and late stage DKD (Mimura & Nangaku, 2010; Vinovskis, et al., 2020).

Although hypoxia is an important factor in DKD, hyperglycemia is the main factor that causes DKD (Oshima, et al., 2021). In addition, in DKD, proximal tubule cells must reabsorb increased amounts of glucose, which lowers ATP and reduces oxygen, generating relative intracellular hypoxia; this hypoxia can lead to both proliferation and metabolic reprogramming via activation of the transcription factor HIF1α (Edwards, Palm, & Layton, 2020; Mohandes, et al., 2023; Vallon & Thomson, 2020). There is also a genetic component to DKD (Mohandes, et al., 2023; Seaquist, Goetz, Rich, & Barbosa, 1989).

A class of drugs called sodium-glucose cotransporter 2 (SGLT2) inhibitors (gene name, SLC5a2) positively impacts the pathology of DKD in humans and many members of this class of drugs have been approved by the FDA (Bhatt, et al., 2020; Heerspink, et al., 2020; Perkovic, et al., 2019). SGLT2 inhibitors both lower fasting and postprandial (after eating) plasma glucose and prevent the use of ATP by these transporters, mechanisms by which these drugs presumably improve outcomes in DKD (DeFronzo, Norton, & Abdul-Ghani, 2017).

One piece of the puzzle in terms of the molecular causes of DKD comes from the protein kidney injury molecule-1 (KIM-1), a product of the HAVCR1 gene, which is highly increased in PTs of injured and diseased kidneys (Humphreys, et al., 2013; Ichimura, et al., 1998). KIM-1 is highly expressed early in DKD and the blood level of KIM-1 predicts progression of DKD independent of % hemoglobin A1C, estimated glomerular filtration (eGFR) rate, and ACR (albumin/creatinine ratio) in patients with Type 1 diabetes (Panduru, et al., 2015; Sabbisetti, et al., 2014). The function of KIM-1 was not understood until recently, when Mori et al. reported that KIM-1 mediates the uptake of palmitic acid-bound albumin by PT cells, which leads to DNA damage, PT cell cycle arrest, reduced mitochondrial transmembrane potential, and major inflammation and fibrosis (Mori, et al., 2021). Moreover, an experimental small molecule inhibitor of KIM-1 could reduce the tubule abnormalities in a mouse model (Mori, et al., 2021). We also observe a massive increase in KIM-1 expression in mice with RARα knocked out specifically in proximal tubule cells of adult mice (DiKun, Tang, Fu, Choi, Lu, Gudas, submitted for publication), demonstrating an involvement of retinoid signaling. Notably, transcriptomic analyses have also shown that transcripts relevant to inflammation are elevated in human DKD tubulointerstitial samples compared to normal control kidney samples (Woroniecka, et al., 2011).

Alterations in vitamin A metabolism and signaling occur in DKD, and RA at pharmacological doses can alleviate some of the pathology in the kidneys associated with DKD. For instance, Starkey et al. demonstrated that endogenous RA levels in the kidney were lower in renal cortical tissue from db/db genetically obese mice compared to controls (Starkey, et al., 2010). However, biopsies from patients with early DKD compared to control patients without DKD exhibited elevated levels of several transcripts involved in RA signaling (eg. RDH8, RDH12, RBP4) (Fan, et al., 2019), suggesting a protective effect since exogenously added RA is renoprotective for DKD. In an ob/ob genetic model of obesity-related diabetes, RA given daily by stomach intubation for 16 days lowered basal blood glucose concentrations, improved glucose tolerance, and increased insulin sensitivity, though kidney function was not explored in this study (Manolescu, Sima, & Bhat, 2010). Exogenously added RA reduced proteinuria and urinary albumin/creatinine ratio in a rat model in which streptozotocin was employed to generate diabetes (Han, et al., 2004). Moreover, in a rat model of type 1 diabetes in which rats were given a single dose of streptozotocin, RA, provided by oral gavage from day 3-21 after the streptozotocin injection, attenuated the TLR4/NF-kB pathway signaling and notably, prevented NF-kB from translocating to the nucleus in both glomeruli and PTs (Sierra-Mondragon, et al., 2018). In the same model, RA at pharmacological doses reduced fibrogenesis by lowering the signaling through TGFβ1/Smad3 signaling (Sierra-Mondragon, et al., 2019). Most of the rodent studies of RA in diabetes have utilized the streptozotocin model, which mimics type 1 diabetes. Less research is available in rodent models in which diabetes is induced by a high-fat diet, so this is an area in which researchers should focus.

Since RARα is important in maintaining homeostasis in proximal tubule cells (DiKun, Tang, Fu, Choi, Lu, Gudas, submitted for publication), a number of RARα selective agonists have been designed so that the effects of RARα can be separated from those of RARβ and γ; some of these RARα agonists have been tested in diabetes animal models (Y. Zhong, et al., 2011). However, we’ve shown that an RARβ2 agonist also can ameliorate the pathology associated with DKD (Trasino, Tang, Shevchuk, Choi, & Gudas, 2018). The use of RAR selective agonists should be developed further for the treatment of DKD, given the positive results in the literature discussed here.

b). Acute Kidney Injury (AKI)

Acute kidney injury (AKI) is characterized by an abrupt decrease in kidney function persisting up to 7 days and is classified by a decrease in urinary output by <0.5mL/kg/h for at least 6 hours or a 1.5-fold increase in serum creatinine (SCr) compared to baseline (Kellum, et al., 2021; Lopes & Jorge, 2013; Ostermann, et al., 2020; Yaqub, et al., 2022). Incidence rates of kidney injury vary widely between setting (hospital-acquired vs. community-acquired) (Lewington, Cerdá, & Mehta, 2013). Hospital-acquired AKI occurs more commonly in elderly patients and is often due to post-surgical complications, sepsis, or nephrotoxic interventions. Conversely, community-acquired AKI more often occurs in low to middle income regions in a younger population from complications of regional infections (dengue, malaria, etc.), diarrhea and hypovolemia caused by toxins in contaminated water, and septic pregnancies (Cerdá, Bagga, Kher, & Chakravarthi, 2008; Hoste, et al., 2018; Lewington, et al., 2013).

As previously mentioned, the kidney is comprised of various cell types, most of which can be affected in some way to cause injury. Many overlapping events may trigger AKI, so for simplicity, we will group the most prevalent causes of kidney injury into the following three categories: (1) pre-renal, (2) intra-renal (intrinsic), and (3) post-renal (Devarajan, 2006; D. Patschan & Müller, 2015).

Pre-Renal Injury

The maintenance of optimal blood pressure and cardiac function plays a crucial role in kidney functionality. Hemodynamic alterations resulting in decreased blood flow or volume fall under the scope of ‘pre-renal injury’ and are considered the most common cause of AKI (Devarajan, 2006). Pre-renal injury affects the function of glomeruli without impacting the morphology, meaning normal function can be reattained if blood flow and volume are restored before structural damage occurs. Various factors can lead to decreased blood flow or hypotension, the most common being inflammatory shock from sepsis, heart failure, and cardiac surgery. Certain medications, such as non-steroidal anti-inflammatory drugs (NSAIDs), also impact glomerular filtration rate (GFR) by influencing constriction or dilation of afferent and efferent arterioles (Drożdżal, et al., 2021; Daniel Patschan, Patschan, & Müller, 2012).

The Renin-Angiotensin-Aldosterone System (RAAS) utilizes a complex feedback mechanism to control blood pressure, cardiovascular function, and fluid balance by allowing crosstalk between blood vessels and kidneys. A drop in blood pressure stimulates juxtaglomerular renal cells to release renin, which converts angiotensinogen from the liver into angiotensin (ANG) I. Interestingly, the mouse renin gene has an enhancer region containing a retinoic acid response element (RARE) (Itani, Liu, Pratt, & Sigmund, 2007; Shi, Gross, & Sigmund, 2001). Shi, et al. showed that renin mRNA levels were increased in kidneys after RA administration via subcutaneous injection in mice fed a vitamin A deficient diet (VAD). RA treatment also increased renin mRNA levels in cultured As4.1 cells, which are renin-expressing juxtaglomerular cells derived from mice (Shi, et al., 2001). Angiotensin-converting enzyme I (ACE I) further metabolizes ANG I to ANG II, which acts as a vasoconstrictor to increase blood pressure (Mirabito Colafella, Bovée, & Danser, 2019; Patel, Rauf, Khan, & Abu-Izneid, 2017). RA is a negative regulator of ACE I, ANG I, ANG II, and angiotensin II type 1 receptor (AT₁-R) (Takeda, Ichiki, Funakoshi, Ito, & Takeshita, 2000; T. B. Zhou, Ou, Rong, & Drummen, 2014). Zhong, et al. observed decreased ACE II expression in a spontaneously hypertensive rat (SHR) model that was alleviated via intraperitoneal injection of RA (J. C. Zhong, et al., 2004).

Intra-renal (Intrinsic) Injury

Intra-renal (intrinsic) injury refers to physical damage affecting various structures within the kidney itself, such as the glomeruli, tubules, interstitium, and vasculature (Goyal, Daneshpajouhnejad, Hashmi, & Bashir, 2023; D. Patschan & Müller, 2015; Daniel Patschan, et al., 2012). The reduction in blood flow characteristic of pre-renal AKI can lead to oxygen deprivation and cell death in renal tissue; this is referred to as ischemia and is an excellent example of how prolonged pre-renal injury can develop into intra-renal injury (Bonventre & Yang, 2011; Daniel Patschan, et al., 2012). Depending on the affected renal cell type, ischemia will cause damage via varied mechanisms. For example, ischemia damages podocytes, as well as endothelial cells lining the blood vessels of glomeruli, resulting in protein leakage into the urine (proteinuria), and glomerular inflammation (glomerulonephritis) (Y. Chen, et al., 2019; M. C. Wagner, et al., 2008). Ischemia also leads to renal tubular epithelial cell (RTEC) death by causing a reduction in ATP production, a critical feature needed to maintain tubule health and function (Eleftheriadis, Pissas, Antoniadi, Liakopoulos, & Stefanidis, 2018; D. Patschan & Müller, 2015). This condition is referred to as acute tubular necrosis (ATN) and is the most common cause of intra-renal injury (Hanif, Bali, & Ramphul, 2023; Malek & Nematbakhsh, 2015). Tubular death during ATN causes infiltration of reactive oxygen species (ROS)-releasing inflammatory cells, which continue to damage RTECs in a vicious cycle leading to further cell death (Devarajan, 2006; Malek & Nematbakhsh, 2015; Nath & Norby, 2000).

The proximal tubules (PTs) are the nephron segment most damaged during AKI (Bhargava & Schnellmann, 2017; Naved, et al., 2022). After AKI, the tubules of the kidney can undergo a remarkable amount of regeneration within a few days (Berger, et al., 2014). By lineage tracing experiments in mice, differentiated kidney proximal tubule epithelial cells were shown to be capable of repairing/regenerating these tubules by transiently dedifferentiating and expressing stem-like molecular markers (Kusaba, Lalli, Kramann, Kobayashi, & Humphreys, 2014). In a bilateral ischemia-reperfusion model in mice, injury-associated loss of PT differentiation markers occurred, followed by re-expression of these differentiation markers during PT repair. In addition, even 6 weeks after injury increased numbers of macrophages and lymphocytes interacted with the PT cells (Dixon, Wu, Muto, Wilson, & Humphreys, 2022).

Injury-induced cell death can take place via several pathways, including apoptosis, necrosis, ferroptosis, and autophagy, all of which release pathway-specific damage-associated molecular patterns (DAMPs) to alert the immune system that damage has occurred (Garg, et al., 2015; Rayego-Mateos, et al., 2022; Sun, et al., 2015). A common DAMP signal observed in renal injury is circulating mitochondrial DNA (mtDNA) released from damaged mitochondria. In the early stages of AKI, mitochondria homeostasis is markedly altered due to an imbalance of proteins involved in fusion and fission, which are crucial processes for the maintenance of mitochondrial health; this is most prominent in proximal tubules, as they have the highest mitochondrial content compared to any other kidney cell type (Funk & Schnellmann, 2012; A. M. Hall, Rhodes, Sandoval, Corridon, & Molitoris, 2013; J. Liu, Jia, & Gong, 2021; Mulay, Linkermann, & Anders, 2016). Chidipi, et al. observed an increase in the level of the mitochondrial fission protein dynamin-related protein 1 (DRP1) in response to exogenously added RA in cultured human embryonic kidney cells (HEK293), suggesting that RA can modulate mitochondrial homeostasis (Chidipi, et al., 2021).

DAMP signaling activates the proapoptotic protein and member of the larger nuclear receptor family of transcription factors, nuclear receptor 4A1 (Nur77, NGFI-B) post-injury. Balasubramanian, et al. showed that activation of retinoic acid receptors (RARs) through administration of 9-cis-retinoic acid, an isomer of RA, repressed Nur77 and ameliorated AKI in mice with ischemia-reperfusion injury (IRI) (Balasubramanian, Jansen, Valerius, Humphreys, & Strom, 2012). Additionally, administration of RA to wild type mice prior to IRI blunted renal injury by preventing the activation of Nur77 (Balasubramanian, et al., 2012). RA also protects against cis-platin-induced apoptosis and autophagy in RTECs both in-vivo and in-vitro, which is in part mediated by RARβ (Junxia Wu, et al., 2019; J. Wu, et al., 2020; Yago-Ibanez, Garcia-Pastor, Lucio-Cazana, & Fernandez-Martinez, 2020). In addition to its role in programmed cell death, Nur77 can modulate inflammation via a complex relationship with nuclear factor kappa-light-chain-enhancer of activated B cells (NF-kB). This interaction is complex because Nur77 can bind to the p65 subunit of NF-kB to promote inflammation, but Nur77 also has anti-inflammatory functions by modulating the expression of NF-kB inhibitors (X. M. Li, et al., 2015; Mulay, et al., 2016; Pei, Castrillo, & Tontonoz, 2006).

Retinoic acid-inducible gene I (RIG-I) is another important DAMP sensor that is expressed in glomeruli and RTECs during renal injury. Similar to Nur77, RIG-I can induce both pro- and anti-inflammatory responses; RIG-I can bind to the p65 subunit of NF-kB to induce inflammation and can also inhibit kappaB kinase epsilon (IKBKE), an upstream regulator of NF-kB (H. X. Zhang, et al., 2013). Wang, et al. observed in a rat model of crush-syndrome (CS)-induced AKI that myoglobin released from injured skeletal muscle bound to RIG-I in renal tissue and promoted inflammation via an NF-kB/caspase 3 pathway (P. T. Wang, et al., 2021; Z. Zhou, et al., 2020). Inflammation regulation by RIG-I also indirectly affects renal fibrosis, which will be addressed in more detail when we discuss chronic kidney diseases (CKD) later in this review.

During early stages of injury, macrophages recognize DAMPs from injured renal tissue and polarize into (M₁) macrophages, which promote tissue damage by regulating the expression of pro-inflammatory mediators, such as tumor necrosis factor alpha (TNF-α) and inducible nitric oxide synthase (iNOS) (H. Chen, Liu, & Zhuang, 2022; Lee, Fessler, Qu, Heymann, & Kopp, 2020; Xing Li, et al., 2018; Tian, et al., 2015). M₁ macrophages are later replaced by alternatively activated (M₂) macrophages which ameliorate inflammation by producing anti-inflammatory cytokines, such as interleukin 10 (IL-10), and promote wound healing via secretion of transforming growth factor beta (TGFβ) (H. Chen, et al., 2022; Tian, et al., 2015). Macrophages expressing aldehyde dehydrogenases ALDH1a2 and 3 (ALDH1a2/3, formerly Raldh2/3), RA-synthesizing enzymes, are recruited to the kidney insterstitium post-injury, resulting in RA signaling and the direct suppression of inflammatory M₁ macrophages (Chiba, et al., 2016). This locally synthesized RA further activates repair-associated M₂ macrophages indirectly by promoting RA signaling in proximal tubule epithelial cells (PTECs) (Chiba, et al., 2016).

We have placed some emphasis on tubular injury thus far as it is the most common type of intra-renal injury. However, we will now briefly shift to what is currently understood of retinoids regarding other renal structures and cell types during AKI. As previously mentioned, glomerulonephritis (GN) occurs when glomeruli become inflamed and damaged (Bienholz, Wilde, & Kribben, 2015). A key characteristic of GN is injury to podocytes resulting in apoptosis, foot process effacement/detachment, and loss of podocyte-specific markers (Sakhi, et al., 2019). RA administration to primary cultured mouse podocytes causes expression of markers specific to podocyte-differentiation; and the same has been shown in mice with anti-glomerular antibody-induced glomerulonephritis (Oseto, et al., 2003; Vaughan, et al., 2005). Furthermore, several retinoids, including RA, and 13-cis-RA, reduce the expression of pro-inflammatory factors in-vivo using a mouse model of HIV-associated nephropathy and in-vitro in cultured cytokine-stimulated mouse mesangial cells (Dai, et al., 2017; Datta & Lianos, 1999; Xuezhu Li, Dai, Chuang, & He, 2014). Moreover, administration of RA by oral gavage attenuated glomerulonephritis through the activation of RXR and RARα in an anti-Thy1.1 glomerulonephritis (Thy-GN) mouse model (Schaier, et al., 2004; J. Wagner, et al., 2000). Our lab has also demonstrated the importance of RARα for proximal tubule homeostasis; we showed that genetic deletion of RARα selectively in proximal tubules resulted in acute injury, which later progressed to chronic injury (DiKun, Tang, Fu, Choi, Lu, Gudas, submitted for publication).

Additionally, there is emerging evidence implicating collecting ducts as protectors against tubulointerstitial injury (H. Huang, et al., 2020; M. Huang, et al., 2019; Xu, 2019). Wong, et al. observed RA expression in collecting ducts of healthy adult mouse kidneys, as well as the expression of RARβ2 (Y. F. Wong, et al., 2011). Collecting ducts cultured from human patients with polycystic kidney disease displayed lower RA signaling via RAR, which was normalized with administration of the synthetic RARα agonist RA-568 (Papadimitriou, et al., 2020). Thus, RA appears to have a protective role in the collecting ducts (Xu, 2019). It is clear that retinoids and RARs exhibit injury-repairing capabilities in a variety of renal cell types. Much effort continues to be made to further elucidate the molecular mechanisms by which retinoids in AKI exhibit therapeutic potential.

Post-Renal Injury

Post-renal injury is the least common type of AKI, accounting for an estimated 5% of cases (D. Patschan & Müller, 2015). This type of injury occurs via urinary tract obstruction by entities, such as tumors, blockages, hematomas, and enlarged prostate (Needham, 2005). Damaged cells from intra-renal injury can also become detached and form casts that cause blockages (Basile, Anderson, & Sutton, 2012). There is not much data available connecting retinoids to post-renal injury. However, both RARα and RARβ2 are known to be crucial in the branching of the ureteric bud during embryogenesis, indicating an important role for retinoid signaling in the structural formation of the urinary tract (Burrow, 2000; Mendelsohn, et al., 1999).

c). Chronic Kidney Disease (CKD)

Chronic Kidney Disease (CKD) affects 13.4% of the global population and is defined as kidney damage or loss of function lasting more than 3 months (A. S. Levey, Becker, & Inker, 2015). CKD is divided into five stages primarily based on glomerular filtration rate (GFR) and excess urinary excretion of albumin, with stage 5 indicating end stage renal disease (ESRD) (Webster, Nagler, Morton, & Masson, 2017). Management of comorbidities, such as obesity, diabetes, and high blood pressure via medications that lower blood pressure and glucose are often the first line of treatment used to slow CKD progression (Cavanaugh, 2007). While early stages are reversible, they often go undiagnosed, with an estimated 90% of people remaining unaware until the kidneys have progressed to ESRD (Alfego, et al., 2021). Interstitial fibrosis is a hallmark of CKD alongside extracellular matrix (ECM) accumulation, tubular atrophy, and inflammation (Webster, et al., 2017). Due to the complex nature of fibrosis, involving crosstalk of multiple interstitial cell types and cellular mechanisms, there are currently no FDA-approved drugs directly targeting fibrotic processes. However, some medications demonstrate secondary anti-fibrotic effects, and their marketing has been adjusted accordingly (Heerspink, et al., 2020; Ruiz-Ortega, Lamas, & Ortiz, 2022). For example, the SGLT2 inhibitor, Dapagliflozin, originally marketed as a type 2 diabetes drug to prevent reabsorption of glucose, attenuates fibrosis in cardiac and kidney tissue and, in 2021, the FDA approved dapagliflozin to be used against CKD progression (Yuyuan Liu, et al., 2022; Xuan, et al., 2021). While promising, there is still much work to be done to alleviate interstitial fibrosis and prevent the development of ESRD. Many advances have been made in the retinoid field indicating their importance in kidney homeostasis and protection against injury (Lorraine J. Gudas, 2022). In this section, we will discuss the various mechanisms that lead to CKD development and the promising potential of retinoids as therapeutics.

Prolonged acute kidney injury (AKI) of proximal tubular epithelial cells (PTECs) often precedes the development of CKD (Leung, Tonelli, & James, 2013; Takaori, et al., 2016). In response to acute injury, surviving PTECs attempt to rebuild the tubule by de-differentiating, proliferating, and re-differentiating to replace lost tubular cells (Kramann, Kusaba, & Humphreys, 2015; Kusaba, et al., 2014). During this repair cycle, short-lived secretion of TGF-β occurs and induces extracellular matrix (ECM) secretion to promote wound healing (Y. Liu, 2011; Rayego-Mateos, et al., 2022; Takaori, et al., 2016). However, repeated injury causes PTECs undergoing repair to become arrested in the G2/M phase of the cell cycle, resulting in large amounts of TGF-β secretion and subsequent priming of residential fibroblasts into activated myofibroblasts (Wynn, 2010). Increased ECM secretion from myofibroblasts further leads to collagen deposition and interstitial fibrosis characteristic of AKI to CKD transition (Leung, et al., 2013; Strutz & Zeisberg, 2006; Zeisberg & Neilson, 2010). Interestingly, injury-induced myofibroblasts express ALDH1a2 (Raldh2), an enzyme that converts retinaldehyde to retinoic acid (RA), and thus transiently acquire the ability to secrete RA (Nakamura, et al., 2019). Healthy proximal tubules also express ALDH1a2 but lose this ability after injury, suggesting that myofibroblasts acquire RA-producing capabilities as a compensatory mechanism and that RA is involved in renal tubular injury repair (Nakamura, et al., 2019).

In two mouse models of CKD, unilateral ischemia-reperfusion injury (IRI) and unilateral ureteral obstruction (UUO), two distinct states of proximal tubules (PTs) were identified shortly after injury by single-cell RNA sequencing. One state showed dysregulated lipid metabolism, and the other showed dysregulated amino acid metabolism (H. Li, Dixon, Wu, & Humphreys, 2022). In related single-cell RNA sequencing research, a proinflammatory and profibrotic PT cell was identified that failed to repair after bilateral ischemia-reperfusion injury to mice (Kirita, Wu, Uchimura, Wilson, & Humphreys, 2020). The expression of proteins involved in ECM deposition, specifically TGF-β, collagen 1, fibronectin, and α-smooth muscle actin, were increased in rats with UUO and alleviated upon RA administration by oral gavage (Z. Y. Li, et al., 2013). Long, et al. corroborated these findings and showed that RARα and RARβ expression was negatively correlated with ECM deposition in the same rat model (Kishimoto, et al., 2011; Long, Qin, Zhou, & Lei, 2012). There is evidence that the anti-fibrotic effects of RA are partially modulated through the activation of angiopoietins-1 (Angpt-1), which protects against ECM deposition and is downregulated in renal fibrosis, as well as alpha-1-acid glycoprotein (AGP), which has anti-inflammatory and anti-fibrotic capabilities (Watanabe, et al., 2020; Z. Zhong, et al., 2021). Furthermore, Watanabe, et al. showed that Am80, a synthetic RAR agonist that binds to both RARα and RARβ, ameliorated fibrosis in mice that received UUO to the same degree as RA, a pan-RAR agonist (Watanabe, et al., 2020). Our lab has recently shown that loss of RARα in PTs causes tubular injury, upregulation of TGF-β, and interstitial fibrosis (DiKun, Tang, Fu, Choi, Lu, Gudas, submitted for publication). These data collectively suggest that RA signaling indirectly alleviates fibrosis and that these effects are modulated through RARα and RARβ activation.

As previously discussed in the ‘Intra-renal AKI’ section, DAMPs released from damaged renal tubules recruit and polarize macrophages into the pro-inflammatory M₁ type, followed by the anti-inflammatory M₂ type (H. Chen, et al., 2022; Xing Li, et al., 2018). However, M₂ macrophages also secrete profibrotic growth factors, meaning prolonged inflammation is another source of TGF-β secretion and fibroblast priming in renal fibrosis (H. Chen, et al., 2022; Y. Liu, 2011; Nakamura, et al., 2019). In acute renal injury, retinoic acid-inducible gene I (RIG-I) expression is elevated in renal epithelial tubular cells (RTECs) and promotes the release of pro-inflammatory cytokines IL-6 and IL-1β via NF-kB activation (H. X. Zhang, et al., 2013; Z. Zhou, et al., 2020). Activation of RIG-I, an RNA sensor, is likely not induced by RA in AKI, but rather is triggered by cytosolic mitochondrial RNA release (Doke, et al., 2023; Tigano, Vargas, Tremblay-Belzile, Fu, & Sfeir, 2021). Importantly, NAD⁺ levels are lower in human CKD patients than in healthy controls (Ralto, Rhee, & Parikh, 2020), and supplementation with NAD⁺ precursors such as nicotinamide riboside after AKI induced by cis-platin in mice limits RIG-I dependent inflammation from cytosolic mitochondrial RNA release (Doke, et al., 2023). Utilizing a UUO mouse model, Zhou, et al. demonstrated that RIG-I-induced cytokine release from RTECs enhanced expression of c-myc in fibroblasts, which subsequently activated the TGF-β/Smad3 signaling pathway and promoted myofibroblast activation (Z. Zhou, et al., 2020).

TGF-β/Smad3 signaling also promotes the transition of M₂ macrophages to myofibroblasts in a process known as macrophage-myofibroblast transition (MMT) (S. Wang, et al., 2016). Bhatia, et al. demonstrated that MMT is prevalent in cells lacking the mitophagy regulators PARKIN and mitofusin 2 (MFN2), which are involved in the removal of dysfunctional mitochondria, and that both regulators were downregulated in an adenine diet-induced mouse model of renal fibrosis (D. Bhatia, et al., 2019). There is a positive correlation between mitochondrial dysfunction and AKI to CKD transition; failure to remove damaged mitochondria due to an impaired mitophagy pathway may be contributing to this relationship (Divya Bhatia, Capili, & Choi, 2020; Finsterer & Scorza, 2017). An indicator of mitochondrial damage is the accumulation of reactive oxygen species (ROS) (H. Zhang, et al., 2008). Interestingly, RA administered to mice with UUO-induced fibrosis had increased renal expression of prohibitin, a protein located in the mitochondrial membrane that protects against ROS-induced damage (T. B. Zhou, et al., 2012). Our lab has shown that PT-specific deletion of RARα results in a disruption of the mitochondrial matrix visible through electron microscopy (DiKun, Tang, Fu, Choi, Lu, Gudas, submitted for publication). This suggests that one of the ways RA protects against renal injury is by upholding mitochondrial integrity.

Patients with CKD show changes in vitamin A homeostasis, including an inverse correlation between GFR and plasma retinol binding protein 4 (RBP4) levels and increased plasma RA (Jing, et al., 2016). Our lab has shown dysregulated retinoid metabolism in mice with PT-specific deletion of RARα. We observed decreases in retinol and RA and an increase in retinyl palmitate in fibrotic kidney cortices; the opposite was observed in serum (DiKun, Tang, Fu, Choi, Lu, Gudas, submitted for publication).

d). Human Immunodeficiency Virus (HIV)-Associated Nephropathy (HIVAN)

HIV-related nephropathy (HIVAN) is the major cause of end-stage renal failure in HIV-1 positive patients (Carbone, D'Agati, Cheng, & Appel, 1989). While HIVAN can affect all cell types in the kidney, podocytes are the most dramatically affected by HIV infection (Lu, He, & Klotman, 2007). HIV-infected podocytes express reduced levels of p27 and p57, two proteins involved in cell proliferation regulation, and they lose their differentiation markers, such as Wilms tumor 1 (WT1), nephrin, and synaptopodin (Barisoni, Bruggeman, Mundel, D'Agati, & Klotman, 2000; Bruggeman, et al., 2000; Shankland, et al., 2000). In podocytes, expression of the HIV gene Nef induces Stat3 and mitogen-activated protein kinase (MAPK) 1 and 2 activity in podocytes, resulting in the dedifferentiation and proliferation of podocytes (He, et al., 2004). In some studies, a transgenic murine model of HIV-associated nephropathy (Tg26) is used in which some HIV genes, but not the actual virus, are expressed. In this Tg26 model, pharmacological doses of RA lower proteinuria and glomerulosclerosis, inhibit cell proliferation, and restore differentiation markers. Furthermore, in this mouse model the effects of a specific RARα agonist, Am580, are blocked by a selective RARα antagonist. In a related genetic study, RAR knockout mice were crossed with the Tg26 mice and the knockout of RARα (a whole-body knockout) in the Tg26 model caused even greater proteinuria, glomerulosclerosis, and podocyte injury. Intriguingly, RARβ protein exhibited greatly reduced expression in the Tg26 glomeruli compared to glomeruli from wild type mice (Mallipattu & He, 2015; Ratnam, et al., 2011b). These studies suggest that drugs targeting the RARs might have benefit in HIVAN.

e). Puromycin, Adriamycin, and Cis-Platin Injury

Podocyte injury during puromycin aminoglycoside (PAN)-induced nephrosis can be reversed by RA treatment. RA treatment can induce podocyte differentiation and increase nephrin expression and daily delivery of RA can reduce proteinuria and improve the effacement of foot processes (Vaughan, et al., 2005). In contrast, both an RA antagonist and a vitamin A deficient (VAD) diet delayed recovery from puromycin aminoglycoside-induced nephrosis. Notably, ALDH1a2 was greatly increased in podocytes with PAN-nephrosis (Suzuki, et al., 2003), suggesting that ALDH1a2 activity is important for normal podocyte differentiation. Suzuki et al. also showed that there are RAREs in the nephrin promoter and that RA activated transcription of nephrin in an animal model of podocyte injury (Suzuki, et al., 2003). These data indicate that endogenous RA produced by ALDH1a2 in podocytes is involved in their repair after injury and that endogenous RA regulates the expression of some genes in differentiated podocytes.

Cis-platin associated AKI develops in many cancer chemotherapy regimens. Reactive oxygen scavengers can be protective against cis-platin associated AKI in murine models (Williams, et al., 2021). RARβ actions are also protective against cis-platin induced proximal tubule injury (Yago-Ibanez, et al., 2020) and Wu et al. demonstrated that pharmacological doses of RA reduced cis-platin acute kidney injury by autophagy activation in a murine model (J. Wu, et al., 2020). Thus, as discussed in prior sections of this review, both endogenous and/or exogenous RA can improve podocyte repair and promote podocyte differentiation after injury.

f). Polycystic Kidney Disease (PKD)

A common cause of end-stage renal disease is cystic kidneys. Polycystic kidney disease (PKD) is generally an inherited disorder and autosomal dominant PKD is the most common form; the prevalence of the autosomal dominant form is ~ 1/400 (Bergmann, et al., 2018; Bergmann, et al., 2005; Harris & Torres, 2014). Mutations in polycystin 1 (PKD1) and polycystin 2 (PKD2) are the most frequent causes of the autosomal dominant form, and these mutations cause progressive expansion of renal cysts (Bergmann, et al., 2018). Notably, these polycystin proteins are predominantly located in the primary cilium, an antenna-like structure that extends out from the apical surface of the epithelial cells in the lumen of the nephron (Bergmann, et al., 2005). Evidence suggests that the function of these primary cilia in tubule epithelial cells is to measure and sense urine flow, though this is still not settled in the literature. Primary cilia lengthen after acute tubular injury and cilia length appears to reflect tubule cell proliferation (Kim, et al., 2013; Verghese, et al., 2009).

When autosomal dominant polycystic kidney disease patient samples were analyzed by single cell RNA sequencing, most proximal tubule cells showed activation of proinflammatory, profibrotic pathways driven by a ‘failed repair’ transcriptomic signature when compared to healthy human kidney samples (Muto, et al., 2022). Collecting duct epithelial cells from these patients also showed increased binding motif availability for NF-kB, CREB, and RARs on the orphan G protein receptor GPCRC5A, a protein that is linked to the proliferation of cells in the cysts (Muto, et al., 2022). A 200 bp region of the PKD1 gene proximal promoter was shown to be involved in RA activation in cultured HEK293T cells (Islam, et al., 2008), but how this relates to the regulation of the PKD1 gene in these polycystic disease patients is not yet clear, especially given that RA and paclitaxel suppress the proliferation of autosomal dominant polycystic kidney disease epithelial cells in culture (Nguyen, Hoang, Ryu, Oh, & Kim, 2021). Clearly, more work is needed in in vivo murine models of polycystic kidney disease to determine the roles of the RARs in polycystic kidney disease because cultured cell models are probably too simplistic to accurately reflect the human disease.

g). Role of Retinoids in Cancer with a Focus on Kidney Cancer

In addition to their importance for normal development and stem cell differentiation (Gudas & Wagner, 2011; Mendoza-Parra, et al., 2016), RARs have many functions in adult tissues to limit various pathological processes. For example, RARs can act as tumor suppressors to limit the development of many types of cancer (Tang & Gudas, 2011). RARγ, the major RAR present in murine and human epidermis, acts as a tumor suppressor as loss of RARγ increased the abundance of skin tumors (C. F. Chen, Goyette, & Lohnes, 2004; Darwiche, et al., 1996; Z. Wang, Boudjelal, Kang, Voorhees, & Fisher, 1999). RARγ also plays a role in stimulating necroptotic cell death in mouse embryo fibroblasts induced by DNA damaging drugs such as cis-platin and etoposide, supporting a role for RARγ as a tumor suppressor (Kadigamuwa, et al., 2019). When overexpressed in oral cancer cell lines, RARβ can induce growth arrest and apoptosis (Hayashi, et al., 2001). Moreover, RARβ expression is often reduced or lost in oral and epidermal squamous cell carcinoma cell lines, again suggesting a tumor suppressor function for RARβ (Hu, Crowe, Rheinwald, Chambon, & Gudas, 1991).

Furthermore, abnormal gene translocations of these RARs can result in cancer. For example, acute myelocytic leukemia most commonly occurs when the RARα gene becomes abnormally fused with the PML gene, forming a PML-RARα fusion protein that contains both the DNA-binding and ligand-binding domains of RARα downstream of the N-terminal PML sequence. This fusion protein causes broad transcriptional activation, and the PML-RARα fusion protein can bind to more sequences than the canonical RAREs, allowing propagation of the leukemic cells (Tan, et al., 2021). High, pharmacological doses of RA can degrade this PML-RARα fusion protein, resulting in the cure of acute promyelocytic leukemia (Ablain, et al., 2013; Korsos & Miller, 2022; Zhu, et al., 1999). RARβ translocations have also been identified in patients with acute promyelocytic leukemias that lack RARα translocations (Osumi, et al., 2018).

Because of the exciting results showing that RA could cure acute promyelocytic leukemia, in the late 1990’s many researchers, including our group, began to test various formulations of RA and 13-cis-RA, an isomer of RA, alone or in combination with interferon-α or a histone deacetylase inhibitor, for the treatment of kidney cancer. The results of these trials were variable, with some positive clinical responses and some negative results (Aass, et al., 2005; Atzpodien, et al., 2002; Berg, et al., 1999; Boorjian, et al., 2007; Casali, Sega, Casali, Serrone, & Terzoli, 1998; Escudier, et al., 1998; Fosså, et al., 2004; Goldberg, et al., 2002; Hoffman, et al., 1996; Miller, et al., 2000; Molina, et al., 2020; R. J. Motzer, et al., 2000; Tavares, Nanus, Yang, & Gudas, 2008; Touma, et al., 2005; X. F. Wang, et al., 2005; M. Wong, Goldstein, Woo, Testa, & Gurney, 2002). However, more recent clinical trials have demonstrated greater efficacy in the treatment of kidney cancer through the use of immuno-oncology agents and combinations of vascular endothelial growth factor targeted therapies (Choueiri, et al., 2021; R. Motzer, et al., 2021; Rini & Powles, 2019). Thus, RA plus interferon-α is no longer used for kidney cancer treatment as better treatment options are available, though the use of retinoids in combination therapy should be kept in mind.

4). SUMMARY/CONCLUSIONS

Thus, retinoids and their receptors are involved in kidney development, kidney disease, and cell repair at multiple points. They are essential for normal kidney development, and failures in retinoid signaling can lead to abnormal kidney development that predisposes toward the development of kidney disease and even to the absence of functional kidneys. Retinoids orchestrate processes in both CKD and AKI and they may yet prove useful in combination chemotherapy for kidney cancer. There are multiple opportunities to develop retinoids, RAR agonists, and RAR antagonists as therapeutics to prevent or slow the development of kidney diseases.

Abbreviations

ACE I

Angiotensin-converting enzyme I

ACR

Albumin/creatinine ratio

AKI

Acute kidney injury

ARF

Acute renal failure

AKI

Acute Kidney Injury

AKIN

Acute Kidney Injury Network

ALDH

Aldehyde dehydrogenases

AGP

Alpha-1-acid glycoprotein

ANG

Angiotensin

Angpt-1

Angiopoietins-1

ARB

ANG II receptor blockers

AT1-R

Angiotensin II type 1 receptor

ATIN

Acute interstitial nephritis

ATN

Acute tubular necrosis

ATP

Adenosine triphosphate

CAKUT

Congenital anomalies of the kidney and urinary tract

CKD

Chronic Kidney Disease

CS

Crush-syndrome

DAMPs

Damage-associated molecular patterns

DKD

Diabetic kidney disease

DM

Diabetes mellitus

DRP1

Dynamin-related protein 1

eGFR

Estimated glomerular filtration rate

ESRD

End stage renal disease

ECM

Extracellular matrix

GFR

Glomerular filtration rate

GN

Glomerulonephritis

GREB1L

Growth regulation by estrogen in breast cancer 1-like

HEK293

Human embryonic kidney cells

HI

High income

HMGB1

High-mobility group box 1

HSP

Heat shock proteins

IKK

I-kappaB kinase

IGFBP7

Insulin like growth factor binding protein 7

iNOS

Inducible nitric oxide synthase

IL-10

Interleukin 10

Intrinsic

Intra-renal

IRI

Ischemia-reperfusion injury

KDIGO

Kidney Disease Improving Global Outcomes

KIM-1

Protein kidney injury molecule-1

LGL1

Late gestation lung protein one

LMI

Low to middle income

LRP1

LDL receptor-related protein 1

MMT

Macrophage-myofibroblast transition

mtDNA

Mitochondrial DNA

NF-kB

Nuclear factor kappa-light-chain-enhancer of activated B cells

NRIP1

Nuclear receptor interacting protein 1

NSAIDs

Non-steroidal anti-inflammatory drugs

PKD

Polycystic kidney disease

PTECs

Proximal tubular epithelial cells

PT

Proximal tubule

PTs

Proximal tubules

RA

Retinoic acid

RAAS

Renin-Angiotensin-Aldosterone System

RARE

Retinoic acid response element

RARs

Retinoic acid receptors

RBP

Retinol binding protein

ROS

Reactive oxygen species

RTECs

Renal tubular epithelial cells

RIG-I

Retinoic acid-inducible gene 1

RIFLE

Risk, Injury, Failure, Loss of kidney function and End-stage kidney disease

ROS

Reactive oxygen species

RTECs

Renal tubular epithelial cells

SAA

Serum amyloid A

SCr

Serum creatinine

SGLT2

Sodium-glucose cotransporter 2

SHR

Spontaneously hypertensive rat

STRA6

Stimulated by retinoic acid gene 6

Thy-GN

Thy1.1 glomerulonephritis

TIMP-2

Tissue inhibitor of metalloproteinases 2

TGFβ

Transforming growth factor beta

TNF-α

Tumor necrosis factor alpha

UUO

Unilateral ureteral obstruction

VA

Vitamin A

VAD

Vitamin A Deficient