Role of the glucocorticoid receptor in glomerular disease

Abstract

Glucocorticoids are potent anti-inflammatory agents that are commonly used in the treatment of various glomerular diseases. Data from in vitro and in vivo studies, in both animals and humans, convincingly demonstrate that glucocorticoids have many beneficial direct effects on glomeruli, including podocytes, suggesting that, in theory, systemic administration is not necessary to achieve therapeutic benefit. Indeed, it is increasingly recognized that systemic steroids often have an unfavorable risk-to-benefit ratio. As we move into an age of personalized medicine, strategies to develop targeted steroid delivery systems and individualized risk assessment algorithms are desirable in clinicians’ efforts to “first, do no harm.”

GLUCOCORTICOID RECEPTORS

Nuclear receptors are transcription factors that are key modulators of diverse biological processes including metabolism, inflammation, and circadian rhythm. The glucocorticoid (GC) receptor (GR) is a member of the nuclear receptor superfamily, which comprises 48 entities including steroid receptors, retinoid X receptor heterodimers, dimeric orphan receptors, and monomeric orphan receptors. The steroid receptor family includes GR, mineralocorticoid receptor, progesterone receptor, androgen receptor, and estrogen receptor (27). Like all nuclear receptors, GR possesses a distinct structural organization that includes an NH₂-terminal domain, DNA-binding domain (DBD), hinge region, ligand-binding domain (LBD), and COOH-terminal domain (Fig. 1) (7). The NH₂-terminal domain is highly variable and contains at least one active transactivation region (AF-1). The DBD is the most highly conserved region and allows for weak dimerization of either homodimers or heterodimers. The hinge region lies between the DBD and LBD and contains the nuclear localization signal. The LBD, which is the largest component, allows for the binding of receptor-specific ligands, allows for strong dimerization, and contains another active transactivation region (AF-2). Finally, the COOH-terminal F domain may or may not be present, and its function is unknown (22). The GR is ubiquitously found in most cell types including in human podocytes, epithelial cells, and endothelial cells (30).

Fig. 1.

Linear structural organization of nuclear receptors. Nuclear receptors share a common structure. The A/B domain (NH₂-terminal region) contains at least one active transactivation region (AF-1) but can be highly variable. The C domain (DNA-binding domain) is the most highly conserved and is involved in homodimeric and heterodimeric dimerization of nuclear receptors. The D domain, which is less highly conserved, acts as a hinge region between the C and E domains and contains the nuclear localization signal (NLS). The largest domain is the E domain, which is the ligand-binding domain. In the case of glucocorticoid receptor, steroids such as endogenous cortisol or synthetic dexamethasone would bind here, as shown. The E domain also contains another transactivation region (AF-2) and enables strong dimerization. The final F domain is highly variable and may not always be present; its function is unknown.

GCs enact their anti-inflammatory and immunosuppressive actions primarily by genomic effects. In humans, GR has two main isoforms: α and β. It is the α-isoform that can bind ligand and mediate these genomic effects, whereas the β-isoform is unable to bind GC (21). In its inactive form, GR resides in the cell cytoplasm and is bound to stabilizing cofactors such as heat-shock protein 90 and the immunophilins FK506-binding protein (FKBP)51 and FKBP52. Binding of either endogenous or synthetic GC to the LBD of the α-isoform results in dissociation of these cofactors. This, in turn, allows an allosteric change in the ligand-receptor complex, thereby allowing translocation to the nucleus. In the nucleus, this complex can bind to specific GC response elements (GREs), resulting in either 1) transactivation with resulting protein synthesis of anti-inflammatory genes or 2) transrepression with resulting inhibition of protein synthesis of proinflammatory genes (Fig. 2). There are also nongenomic effects that are less well understood that are mediated by plasma membrane-bound GR or nonspecific interactions with membrane-bound GR (28).

Fig. 2.

Classic glucocorticoid (GC) metabolism. In its inactive form, the GC receptor (GR) resides in the cell cytoplasm complexed to immunophilins (IP) and heat shock proteins (HSP), which stabilize the receptor. The GR has three exposed domains: the ligand-binding domain (LBD), the DNA-binding domain (DBD), and an immunogenic domain (ID). Upon binding its ligand, IP and HSP dissociate from the receptor and allow it to translocate to the nucleus, where it binds to specific GC response elements (GREs), which results in either transactivation or transrepression of target genes.

GR is widely expressed in the human kidney, including in the glomerulus (30), and GCs are used clinically in a variety of proteinuric glomerular diseases, including membranoproliferative glomerulonephritis (11), membranous nephropathy (2), IgA nephropathy (12), crescentic glomerulonephritis of many etiologies (14, 20, 23), and antiglomerular basement membrane disease (29). However, it is from the treatment of nephrotic syndrome (NS), both in pediatric and adult patients, for which GCs are a mainstay of therapy, that most of the current knowledge about GC effects on glomeruli and clinical practice patterns have evolved.

IN VITRO GC EFFECTS

In vitro studies have shown that GCs can protect podocytes from injury by increasing the expression and phosphorylation of the podocyte gene nephrin, which is critical in maintaining the slit diaphragm (6, 18, 31). GCs have been shown to downregulate two key regulators of cell motility in podocytes, Rac1 and RhoA (15); they have also been shown to protect podocyte stress fibers from drug-induced injury via stabilization of the cytoskeletal protein α-actinin-4 (10, 32). GCs can stabilize downstream targets of GR, including Krüppel-like factor 15, a kidney-enriched zinc finger transcription factor that is required for restoring podocyte differentiation markers under conditions of cell stress (13). Of note, Krüppel-like factor 15 expression has been found to correlate with GC responsiveness in patients with a biopsy proven diagnosis of focal and segmental glomerulosclerosis or minimal change disease (13).

IN VIVO GC EFFECTS

Additional understanding of the renoprotective role of GCs in proteinuric renal disease has been derived from animal models. Recently, the creation of a mouse model with tissue-specific podocyte GR knockout (pGR KO) has provided further evidence for a critical cell-specific role of this receptor. Mice with pGR KO have no renal phenotype at baseline but are more susceptible to both systemic and renal-limited injury with lipopolysaccharide (LPS) and nephrotoxic serum, respectively, as demonstrated by massive proteinuria compared with podocyte GR-replete controls (33). In this study (33), the authors also showed that GR-deficient podocytes possessed a migration defect, were unable to activate focal adhesion kinase, and demonstrated actin cytoskeleton abnormalities after LPS treatment but not during control conditions. In another elegant in vivo mouse study, Kuppe et al. (9) demonstrated that inactivation of GR, specifically in kidney epithelial cells, through the use of Pax8 Cre, conferred renal benefit, including mitigation of proteinuria and attenuation of morphological disease, in a model of crescentic glomerulonephritis. Taken together, both of these studies strongly suggested a beneficial role for kidney-specific, and even cell-specific, administration of GC. The advantage of such delivery systems would be being able to provide targeted therapy with potentially all the therapeutic benefit and none of the undesirable side effects that often limit systemic steroid use.

Indeed, investigation of more targeted steroid administration has begun. One promising strategy is the use of liposomal encapsulation of steroids, which allows drug to be released at sites of inflammation. In a mouse model of lupus nephritis, subcutaneous delivery of liposome-based steroidal nano-drug was shown to have superior efficacy compared with free (non-liposomal) GC in terms of overall survival, kidney pathological injury score, and blood urea nitrogen concentrations (17). A liposomal encapsulation approach of prednisolone has also been used to study the response of human endothelial cells and macrophages subjected to LPS treatment. In this work, liposomal uptake by macrophages was complete 8 h after administration, and cells showed decreased levels of the proinflammatory cytokine IL-6 as well as evidence of nuclear translocation of GR, at least as efficaciously as free prednisolone alone (25). These authors went on to show that, using a rat model of ischemia-reperfusion injury, liposomes accumulated in the injured kidney and that liposomal prednisolone administration was effective in reducing the proinflammatory milieu in the kidney (25). In another study, Colombo et al. (5) developed drug-loaded nanocarriers that can successfully deliver dexamethasone (DEX) to glomerular cells, and this work has been validated both in vitro and in vivo in mice (3).

GC USE IN HUMAN GLOMERULAR DISEASE

Conditionally immortalized podocytes isolated from healthy kidney donors and subjected to DEX treatment followed by RNA extraction and sequencing (RNA-Seq) have demonstrated that GCs can effect myriad nonimmune pathways in humans. Some pathways, including Ras protein signal transduction, Rho motility signaling, and collagen fibril organization, are induced, whereas other pathways, such as cell adhesion, extracellular matrix organization, and regulation of cell migration, are simultaneously inhibited (8). Further analysis of these podocyte RNA-Seq data compared with transcriptome data from patients with minimal change NS revealed a subset of particular genes that were downregulated in NS and tended to be significantly upregulated by DEX; most notably, this list included FKBP51, the immunophilin that regulates GR sensitivity (8). These data suggest that steroid responsiveness in NS may be, at least in part, dictated by gene expression states that allow optimal steroid sensitivity.

More recently, Teeninga et al. (24) examined the relationship between outcomes of pediatric NS and genetic determinants of GC sensitivity and showed that patients possessing the GR-9β + TthIII-1 polymorphism (rs6198, A→G), which ultimately leads to increased expression and stability of the GR β-isoform, demonstrated impaired GC sensitivity and more unfavorable clinical outcomes. However, the factors and mechanisms linking this polymorphism to poorer clinical outcomes have not been investigated and are currently unknown. Interestingly, higher incidence of this polymorphism has been shown in patients in other populations with proinflammatory conditions such as heart failure (19) and rheumatoid arthritis (26).

Genome-wide transcriptional profiling studies in adult patients have demonstrated that early life low socioeconomic status is associated with resistance to GC signaling and exaggerated inflammatory responses (16), both of which are likely key mediators of a condition like NS, and perhaps other glomerular diseases, for which steroids are the mainstay of therapy. A recent observational study (1) in a small cohort of 16 adolescent patients with steroid-sensitive NS found that perceived stress recorded via online diary entries predicted proteinuria. However, to date, the relationship between objective metrics of stress, frequency of relapse, and response to GCs has not been investigated in the NS population.

Therapeutic benefits from the steroid treatment of glomerular diseases are not uniform across patients, even among those with the same diagnosis, and benefits are often tempered by unwanted side effects. For example, the Therapeutic Evaluation of Steroids in IgA Nephropathy Global (TESTING) randomized clinical trial assessed the effect of oral methylprednisolone therapy versus placebo on decline in glomerular filtration rate and development of end-stage renal disease in adult patients with biopsy-proven IgA nephropathy (12). Although the patients that took steroids clearly demonstrated a renal benefit as evidenced by a slower decline in estimated glomerular filtration rate and lower urinary protein excretion compared with those given placebo, the trial was halted early due to an increased incidence of adverse events, mainly infectious and gastrointestinal in nature, in the group that received steroids (12).

Recently, the effect of oral pulse DEX was studied in a small adult cohort with focal and segmental glomerulosclerosis as a potential therapeutic alternative to long-term oral prednisone, given the many adverse effects observed with the latter (4). DEX, a purely synthetic GC with no mineralocorticoid effect, is approximately six to seven times more potent than prednisone. Although the study was small and the treatment course short (4 daily doses for 4 days every 28 days for 32 wk), ~30% of participants experienced side effects ranging from mood and sleep disorders to tendon rupture and increased ocular pressure. However, the effects on improving proteinuria were generally favorable (4). This study highlights the observation that prolonged steroid administration is often not needed to achieve some therapeutic benefit in glomerular disease, although, by the same token, side effects may also be as common and severe with brief steroid courses compared with lengthy courses. It is not at all clear which patients will receive the maximal benefit (and why), nor is it clear which patients will experience the worst side effects (and why).

PERSPECTIVES

In thinking about the continued evolution of medicine and its increasingly personalized nature, one must equally consider the effect of a variety of factors; these include environmental factors, epigenetics, and lifespan variables in addition to pathophysiology and genetics. In the quest to understand both population-based and individual responses to steroid therapy for glomerular disease, and indeed all diseases that are treated with steroids, any or all of these factors may be relevant. As a medical community, we need to gain a better understanding of tissue-specific, and even cell-specific, effects of GCs if we hope to optimize therapeutic responses while minimizing side effects. Further investigation is required to uncover the molecular underpinnings of GC responsiveness in particular cell types, for although steroids have been in use for decades, there is much yet to learn about how best to optimize their use.

Could risk assessment models that would help predict favorable response to steroids be generated using a combination of demographic, social, and medical factors? Is genetic testing, which has become increasingly inexpensive and routine, warranted in any patient slated to receive steroid therapy for genetic variants that may identify patients that are unlikely to respond to corticosteroids? Is empiric systemic steroid use without such testing still justified?

In this author’s opinion, indiscriminate use of steroids, based only on the patient’s diagnosis and not any individual characteristics, is a haphazard therapeutic approach. While there exist occasional cases that demonstrate genetic podocytopathies can respond to immunosuppression, the vast majority will be resistant to such therapies and I suggest steroids be avoided in these scenarios. With regard to nongenetic podocytopathies, which can only be so-labeled after genetic testing is completed, I would encourage a trial of steroids while completing parallel investigations expediently. It may not be reasonable to delay treatment with steroids until further testing can be accomplished, although I would argue neither is it reasonable to initiate steroids and monitor for a therapeutic response for weeks to months without attempting to gather more information about whether such a course of action is likely to benefit an individual patient. As we continue to refine the practice of medicine in an increasingly data-driven, technology-based environment, it is our responsibility as physicians to use these available tools to optimize the outcome of each of our individual patients.

GRANTS

J. E. Goodwin is supported by the National Heart, Lung, and Blood Institute Grant R01-HL-131952.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

J.E.G. prepared figures; J.E.G. drafted manuscript; J.E.G edited and revised manuscript; J.E.G approved final version of manuscript.