Rituximab’s New Therapeutic Target: The Podocyte Actin Cytoskeleton

KIDNEY FUNCTION AND FSGS

The kidney is responsible for controlling blood volume, pressure, and pH as well as maintaining serum electrolyte, metabolite, and glucose homeostasis. The nephron, its functional unit, consists of a glomerulus responsible for blood filtration and a tubule specialized for reabsorption and secretion of electrolytes and metabolites. The glomerulus consists of three layers: fenestrated endothelial cells that line the glomerular capillaries, a glomerular basement membrane (GBM) enriched in negatively charged glycosaminoglycans, and podocytes that form a tight interdigitating network of foot processes to generate a slit diaphragm (Fig. 1). This complex network permits selective passage of metabolites, electrolytes, and small proteins into the tubular system and provides a barrier to prevent large proteins and cells from traversing from the bloodstream into Bowman’s space and subsequently the urine. Disruption of the endothelium, GBM, or podocytes from myriad causes results in loss of the kidney’s ability to retain protein within the bloodstream and subsequent proteinuria—an excess of serum proteins in the urine. When urinary protein loss exceeds 3.5 g/day, patients develop nephrotic syndrome, which is characterized by loss of serum proteins and results in decreased intravascular osmotic pressure and edema, hyperlipidemias, thrombosis, and infections. Many patients with nephrotic syndrome proceed to end-stage renal disease (ESRD) and kidney failure, requiring dialysis or organ transplantation.

• Download high-res image
• Open in new tab
• Download Powerpoint

FSGS represents one of four major histological classifications of podocyte pathology and accounts for 10 to 15% of cases of nephrotic syndrome (2, 3). FSGS is characterized by focal and segmental scarring of the glomerulus, loss of podocytes (podocytopenia), and compromise of slit-diaphragm integrity. The mechanisms of disease pathogenesis are unclear, although familial cases of FSGS have provided clues that podocyte cell adhesion and cytoskeletal architecture play important roles in normal glomerular function. Circulating factors that enhance capillary permeability (possibly the soluble urokinase receptor or cardiotrophin-like cytokine 1) are present in some FSGS patients and may contribute to disease (4). Injection of FSGS sera into rats induces proteinuria, and incubation of FSGS sera with isolated rat glomeruli alters albumin permeability (5, 6). The importance of serum factors is also supported by the high percentage of FSGS patients that relapse after successful kidney transplantation and the efficacy of plasmapheresis in some patients.

An immune component is thought to contribute to FSGS disease pathogenesis, as the mainstay of therapy is immunosuppression, and modulation of immune regulators prevents the development of proteinuria in experimental disease models (7). Immune-modulating therapies for FSGS include corticosteroids, DNA alkylating agents (cyclophosphamide and chlorambucil), and inhibitors of T lymphocyte activation [cyclosporine (CsA) and tacrolimus (FK506)] and de novo lymphocyte generation (mycophenolate mofetil); plasmapheresis is also used for recurrent FSGS after kidney transplantation. Biotherapeutics have been used recently to treat FSGS, but the experience has been limited to case reports and small noncontrolled series (8). A phase I safety and tolerability study of 10 patients was completed for the anti-inflammatory agent adalimumab, which binds to and inhibits tumor necrosis factor–α (9). A growing number of case reports and noncontrolled series have reported beneficial effects of the biologic rituximab (8, 10).

RITUXIMAB AND FSGS

Rituximab is a chimeric mAb approved for treatment of non-Hodgkin’s lymphoma (NHL), chronic lymphocytic leukemia, and rheumatoid arthritis (RA) with inadequate responses to other therapies (11). Rituximab recognizes CD20, a glycoprotein with four transmembrane domains that is expressed on B-lineage cells and a small population of T cells of unknown function. Upon binding to CD20, rituximab depletes CD20⁺ cells through the Fcγ immunoglobulin receptor (FcγR), complement, and apoptotic pathways.

The early promise of rituximab in FSGS exhibited in noncontrolled studies suggests a potential role for B cells in FSGS disease pathogenesis. At odds with this hypothesis is the paucity of B cells in the affected glomeruli and lack of serum autoantibodies detected in this disease. In the current work, Fornoni and colleagues provide data to support a nonimmune mechanism by which rituximab may exert its therapeutic effects in recurrent FSGS after renal transplantation (1).

In 2006, Perosa and colleagues reported a novel rituximab-binding amino acid sequence, WPxWLE (12), that is not found in CD20 or any other human protein (13). However, the reverse sequence, ELWKPW, is found within the acid sphingomyelinase-like phosphodiesterase 3b protein (SMPDL-3b), a putative acid-sphingomyelinase (ASMase). Both forward and reverse sequences bind rituximab but not other anti-CD20 mAbs.

Fornoni and colleagues followed 27 high-risk patients with primary FSGS who received kidney transplantation for ESRD (1). Patients were treated with one dose of rituximab (375 mg/m², which is 25% of the approved dose for treatment of NHL), thymoglobulin (a T cell immunomodulatory gamma globulin), and daclizumab (an anti-IL2Rα mAb) and thereafter with maintenance immunosuppression to prevent organ rejection. Biopsies of donor kidneys were obtained ~2 hours after reperfusion, but prior to rituximab infusion, in 20 of 27 patients. Patients were followed for 12 months after transplantation. As a study limitation, a historical comparator group that did not receive rituximab was used as the control. The historical group contained more living donors, which may skew the data favorably to the rituximab-treated group because organs from living donors may give rise to worse outcomes in FSGS transplant patients (14). This caveat notwithstanding, patients treated with rituximab demonstrated improved renal function as measured by estimated glomerular filtration rate (eGFR) at 3 and 6, but not 12, months after transplantation when compared with the historical control group.

To investigate how rituximab improved renal function, the authors examined CD20 expression in kidney sections and podocytes (1). Although rituximab bound podocytes in normal and posttransplant kidneys, this reactivity did not result from CD20 because another CD20 mAb (1F5) did not demonstrate staining, and the podocytes expressed neither CD20 mRNA nor protein. However, consistent with previous phage-display results (13), rituximab appeared to bind SMPDL-3b expressed on podocytes, as rituximab immunoprecipitated SMPDL-3b, and an SMPDL-3b–derived peptide competed with rituximab for binding to podocytes (1). Kidney biopsies from patients with recurrent (rec) FSGS had fewer SMPDL-3b–positive cells per glomerulus when compared with nonrecurrent (non-rec) FSGS. Because sera from FSGS patients can induce proteinuria when injected into rats and alter the podocyte actin cytoskeleton in vitro (5), the authors analyzed the effects of FSGS sera on podocytes. Human podocytes incubated in vitro with sera from rec FSGS patients showed decreased SMPDL-3b protein and mRNA expression and decreased ASMase protein expression and activity. Addition of rituximab reversed these down-regulatory effects of rec FSGS sera, as did expression of a green fluorescent protein (GFP)–SMPDL-3b fusion protein. No effects were observed with sera from non-rec FSGS patients or normal individuals in the presence or absence of rituximab (1).

The authors then analyzed whether alterations in SMPDL-3b and ASMase activity induced by rec FSGS sera affected podocyte cytoskeletal architecture. Incubation of podocytes with rec, but not non-rec, FSGS sera resulted in disruption of actin stress fiber formation that was reversed with the addition of rituximab. Importantly, loss of podocyte actin stress fibers in vitro correlated with FSGS recurrence and worsened kidney function in patients. The effects of rituximab on actin stress fiber formation and podocyte viability in vitro were dependent on SMPDL-3b expression, as knockdown of SMPDL-3b in podocytes negated the effects of rituximab in the rec FSGS sera–treated samples.

These data are quite provocative because they support a model of an adventitious off-target mechanism for rituximab in posttransplantation FSGS (Fig. 2). Rather than modulation of B cells in FSGS pathogenesis, these data suggest that rituximab binds podocyte SMPDL-3b and attenuates the pathogenic effects of rec FSGS sera on SMPDL-3b expression and ASMase activity. In turn, podocyte architecture and viability are preserved and may improve clinical outcome.

• Download high-res image
• Open in new tab
• Download Powerpoint

Of note, the benefit of rituximab on ΔeGFR was lost 12 months after transplantation, which suggests that the effects of one dose of rituximab were neither permanent nor durable. On the basis of rituximab pharmacokinetics in other diseases, the in vivo activity of the single dose of rituximab administered in these patients is likely to have waned. In RA, recovery of peripheral blood B cell number typically begins 6 months after rituximab administration at doses that are two- to fourfold greater than those used in this study. Additional analyses of B cell recovery, rituximab serum concentrations, and ΔeGFR are needed to decipher rituximab regimens that may provide better clinical outcome and graft survival.

PODOCYTES: FUNCTION FOLLOWS FORM

The ability of rituximab to modulate the effects of rec FSGS sera on the podocyte cytoskeleton provides clues to FSGS disease pathogenesis. The genetics of familial forms of FSGS point to the importance of cellular polarity and regulation of the actin cytoskeleton in the podocyte. The slit diaphragm is stabilized by a transmembrane multiprotein scaffold (Fig. 1, inset) (2, 3), and mutations in the genes that encode these proteins have been described in familial and idiopathic FSGS (15–17). Hence, the genetics of familial forms of FSGS implicate the actin cytoskeleton as causative in podocyte pathology.

Podocytes continually adapt their actin cytoskeleton in response to mechanical stress induced by hydraulic pressure within the glomerulus. Mechanical stress is transmitted to podocytes through cell-matrix contacts, mediated at least in part by the α3β1 integrin, which is required to maintain glomerular structural integrity (18). Integrin-linked kinase (ILK), a β integrin–associated kinase known to regulate cell adhesion and anchorage-dependent growth, is increased in congenital nephrotic syndrome and in experimental models of proteinuric kidney disease (19). Addition of sera from a subset of rec FSGS patients to podocytes activates ILK and results in disassembly of focal contacts and detachment of podocytes from a laminin matrix (20). Conversely, mice with a podocyte-specific ILK deletion develop FSGS-like features and terminal renal failure (21). Hence, an altered actin cytoskeleton may not adequately adapt to mechanical stress, and such podocytes may thus be predisposed to dysfunction and cell death in familial and, potentially, acquired FSGS.

The nonimmune therapeutic effects of cyclosporine A (CsA) also suggest the importance of the podocyte cytoskeleton. Inhibition of calcineurin by CsA preserves synaptopodin:14-3-3β interaction and maintenance of podocyte stress fibers (22). Thus, existing therapies for FSGS may function through stabilization of the podocyte actin cytoskeleton.

CERAMIDE AND STRESS

How might the proposed ASMase alteration in the study reported by Fornoni and colleagues (1) relate to the actin cytoskeleton? First, nothing is known about SMPDL-3b function, although it contains an ASMase-like domain, and much of the speculation is inferred from what is known about ASMase (SMPD1). The highly regulated ASMases catalyze conversion of the membrane lipid sphingomyelin to the second-messenger ceramide in response to activated membrane receptors—such as the tumor necrosis factor–α, interleukin-1, and interferon-γ receptors—and nonreceptor cellular stress inducers, including radiation, pathogens, free radicals, and other cytotoxic agents (23). Ceramide plays important roles in cell growth, apoptosis, and cytoskeletal regulation. Cellular stress induced by the DNA-damaging agent cisplatin results in ASMase activation that, in turn, permits the serine/threonine phosphatase PP2A to dephosphorylate ezrin, a protein that links actin with the plasma membrane (24). Subsequent relocation of ezrin to the cytosol results in the dissociation of actin filaments from the plasma membrane and loss of cellular filopodia—membrane protrusions that probe the cellular surroundings. How rec FSGS sera deplete ASMase and how this affects the podocyte cytoskeleton remain unknown, but Fornoni et al. strongly suggest that these are related events.

Rituximab alters the ASMase/ceramide pathway in B cells (25) and induces ASMase translocation from the intracellular compartment to the external leaflet of the plasma membrane; here, ASMase is activated, ceramide is increased in membrane raft microdomains, and cell growth and clonogenic potential are inhibited. Further investigation is needed to determine whether these effects arise from rituximab binding to CD20, as has been suspected, or from the unexpected off-target effect described by Fornoni and colleagues (1).

The fortuitous circumstances of the genetics that implicate proper podocyte actin architecture in glomerular function, an emerging appreciation of how ceramide regulates actin, and the potential off-target therapeutic mechanism by which rituximab alters podocyte cytoskeletal structure and survival may have provided researchers with important clues for understanding FSGS pathogenesis at the molecular level. Undoubtedly, B cells play an important pathogenic role in other proteinuric kidney diseases. The hallmark of idiopathic membranous nephropathy (IMN), a common cause of nephrotic syndrome, is GBM deposition of immune complexes and complement (26). In addition, autoantibodies against M-type phospholipase A2 have been discovered in sera and kidney biopsies in IMN patients that appear to correlate with disease course and response after rituximab therapy (27–29). Dissection of the molecular and cellular etiologies of nephrotic syndromes should pave the way to improvements in disease classification and the development of a new generation of therapies directed toward immune and nonimmune etiologies of proteinuric diseases.