Rewriting Renal Fate

Kidney diseases are mechanistically heterogenous pathologies affecting multiple different cell types in the kidney. Most therapies for kidney diseases are noncurative, seeking to prolong function and replace critical roles in homeostasis. Gene therapies have strong potential to address many kidney diseases as they offer intervention into the pathogenic processes driving kidney damage and its progressive nature. Furthermore, with the explosion of developments in gene therapy technologies, including the recent utilization of individualized gene editing in humans, the challenge of how to use gene therapy to address kidney disease hinges on effective delivery of genetic payloads. The kidney poses a challenging target for delivering gene therapy as renal diseases can affect isolated or unique combinations of anatomical and functional kidney components.

Adeno-associated virus (AAV) vectors are proven delivery vehicles in gene therapy, with treatments for diseases affecting different organs already being approved for use in clinical practice or in stage 3 clinical trials (https://clinicaltrials.gov/search?intr=AAV). Their small size (approximately 25 nm) and favorable safety profile at moderate systemic doses makes them an attractive choice for gene transfer to different portions of the nephron. The purpose of this article was to offer perspective on recent developments in advancing AAV for gene therapy of kidney disease.

AAV Capsid Toolkit for Kidney Gene Delivery

Recent efforts in AAV capsid engineering have focused on identifying vectors with improved kidney tropism using library-based screening and directed evolution. Wu et al. developed AAV2-GEC, a capsid variant efficient gene transfer to murine glomerular endothelial cells (GECs), by performing in vivo screening of an AAV2 peptide display library in multiple rounds of selection from whole kidney to isolated glomeruli in mice.¹ This variant robustly transduced GECs in both healthy and diseased rodent models, although its translational potential remains unclear. Rosales et al. used a cross-species directed evolution strategy that yielded AAV.k13 and AAV.k20 capsids by cycling AAV9-based capsid libraries through mouse, pig, human kidney organoids, and ex vivo perfused nonhuman primate kidneys.² This approach enriched for several variants with consistent and robust kidney tropism across species, highlighted by AAV.k20—a vector that outperformed AAV9 in both transduction efficiency and specificity in proximal tubule (PT) cells by eight- to 12-fold. Importantly, these variants exhibited compatibility with self-complementary and single-stranded genomes, making them excellent candidates for developing potential kidney gene therapies. By applying evolutionary pressure across species and tissue contexts, this strategy prioritizes capsids likely to translate across both preclinical models and clinical trials (Figure 1).

Figure 1.

Developments in and opportunities for advancing AAV gene therapy for kidney disease. Top left, only certain cellular targets are currently targetable by AAVs. Some desirable targets are listed. Top right, AAV has packaging limitations. Several strategies have been proposed to overcome cargo limitations. In addition, one can target the pathologic process by delivering a therapeutic gene independent of the disease-causing mutation(s). Bottom, capsid cycling and enrichment have been performed in healthy models resulting in AAVs capable of cross-species transduction of PT cells and one targeting glomeruli in mice. Cycling in diseased models may offer more disease-targeted AAVs. AAV, adeno-associated virus; DCT, distal convoluted tubule; PCT, proximal convoluted tubule.

Expanding Scope through Vector Design

Another challenge facing the broad application of AAV vectors for kidney-targeted gene therapies is their limited packaging capacity (4.7 kb single stranded or 2.3 kb self-complementary), which is insufficient for delivering many large genes implicated in kidney disease. However, recent advances have shown promise in addressing this limitation (Figure 1). Specifically, trans-splicing dual AAV systems enable reconstitution of large genes in vivo, and recent development of the REVeRT system allows for more flexible splice site selection when splitting large genes across two vectors, expanding the AAV toolbox for kidney diseases with sizable genetic etiologies.³ Multivector AAV strategies are typically less efficient because successful expression requires cotransduction of each cell by all vectors rather than delivery of a single cassette. As opposed to using split intein-mediated protein trans-splicing, Peek et al. used another cross-species evolved variant (AAV.cc47)⁴ for delivery of split transgenes to PT cells in mice using hybrid homologous recombination-mRNA splicing to achieve 40%–74% transduction of PT cells with split AAVs recombining to express a protein compared to 64%–91% with one AAV expressing the full-length protein.⁵

Beyond gene replacement, the work by Wu et al. demonstrates how AAV-mediated delivery can target and disrupt a pathogenic process without correcting the underlying mutation.¹ Their engineered AAV2-GEC vector efficiently transduced GECs to deliver the IgG-degrading enzyme IdeS, ameliorating antibody-mediated GN in mice. Similarly, direct therapeutic targeting can circumvent gene size constraints, as illustrated in a recent study using cystic fibrosis transmembrane conductance regulator to deliver a compact but disease-modifying transgene in a model of polycystic kidney disease.⁶ This study also assessed expression approximately 60 days after a single intraperitoneal dose and detected vector genomes, transgene mRNA, and CFTR/green fluorescence protein protein in cyst epithelium, along with sustained biologic effect on cyst size, but it did not evaluate longer-term durability. Recently, Alonso et al. used AAV9 to deliver vascular endothelial growth factor C to podocytes in mice to ameliorate albuminuria and glycocalyx damage in a mouse model of type 1 diabetes.⁷ Collectively, these strategies underscore an expanding toolkit for gene therapy in the kidney, where interventions can include not only gene correction and replacement but also interference with downstream pathogenesis (Figure 1).

Influence of Kidney Environment on Gene Transfer

Emerging evidence suggests that the diseased state of the kidney can significantly influence susceptibility to AAV-mediated gene transfer (Figure 1). Recently, Lin et al. demonstrated that mouse Alport podocytes could be transduced by AAV9 in vivo.⁸ They observed transduction of 1.8%–26% of podocytes in Alport mice that correlated with proteinuria when using doses of 3×10¹³ to 9×10¹³ vg/kg. No transduction of wild type mice podocytes was observed with AAV9. The increased permeability of the glomerular basement membrane in Alport syndrome was proposed as a key factor facilitating this heightened AAV penetration into podocytes. A parallel study in a CKD model further argues that pathologic changes in the kidney can affect AAV transduction efficiency in both PT cells and podocytes.⁹ This study by Nakai et al. found that intravenous injection of AAV9 showed no transduction of wildtype kidney while showing 35% in podocytes and 9.3% in the PT in CKD.⁹ Specifically, local delivery of a previously derived capsid obtained by cycling in human pancreatic islet cells (AAV-KP1), but not AAV9, through renal vein or pelvic dosing transduces PTs; while systemic AAV9, but not AAV-KP1, enhances PT and podocyte transduction in a CKD model.

In another recent preprint study, Rubin et al. demonstrate that AAV transduction of renal epithelial cells can be significantly enhanced by inducing proteinuria in mice via lipopolysaccharide treatment, suggesting this disruption of the glomerular filtration barrier allowed viral vectors, which would normally be excluded, to reach and transduce tubular epithelial cells, especially when using AAV8 and AAV1.¹⁰ Notably, this increased kidney transduction occurred without corresponding increases in liver transduction when using AAV1. The clinical relevance and translatability of such studies that rely heavily on murine disease models in the context of improving kidney gene transfer remains to be determined. Collectively, these studies suggest that the microenvironment of diseased kidneys may affect AAV access, uptake, and/or transduction efficiency. This raises the possibility that performing capsid selection within disease models (mouse, larger animal models, or human explants) could yield AAV vectors optimized for targeted gene delivery within the altered cellular landscape of kidney pathology.

Despite these advances, significant challenges remain. Although different AAVs capable of transducing renal blood vessels (endothelial cells), mesangial cells, PT cells, and podocytes may now be available for preclinical and clinical testing, other key disease relevant targets such as the distal convoluted tubule, thick ascending limb, and collecting ducts have yet to be targeted (Figure 1). Furthermore, while the altered physiology of diseased kidneys may hinder or benefit AAV tropism, it may also prove to make prophylactic treatment before disease progression more challenging to pursue. In addition, the durability of AAV-mediated transgene expression in dividing renal cell types, particularly in the context of different disease states is currently unknown. Immunogenicity and preexisting anti-AAV immunity can limit the effectiveness of AAV-based therapies, and capsids selected directly on human tissues are no exception to this limitation.⁸ This is particularly important given that potential for dose-dependent toxicities with AAV vectors have been reported, with consequences as severe as death. It is worth noting that routes other than systemic intravenous injection, such as the renal pelvis or rectourethral injection, may provide alternate delivery paths, while affording possible protection against vector preimmunity as well as limiting systemic exposure⁹

Summary

AAV has already demonstrated substantial utility as a gene therapy vector in other organ systems, and recent technologic advances may enable therapeutic access to the kidney. The identification of AAV variants with enhanced renal tropism combined with emerging insights into how the diseased kidney microenvironment can significantly influence AAV transduction profiles and may present opportunities to design capsids that exploit these pathologic changes for improved and selective gene transfer efficiency to the kidney (Figure 1). Continued innovation in dual-vector systems, promoter-enhancer toolkits for kidney cell type-specific gene expression, genome editing tools, and strategies to disrupt pathogenic pathways are likely to complement these advances. Taken together, AAV-based gene transfer could translate into effective therapies for both monogenic and complex renal diseases. Although the role of AAV vectors in kidney gene therapy continues to evolve from demonstrated preclinical efficacy to clinical application, several challenges including vector-associated toxicity, immune responses that limit redosing, and other safety and durability concerns indicate that there remains space for further optimization.

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/KN9/B459.

Author Contributions

Conceptualization: Aravind Asokan, Peyton G. Hickman, Matthew H. Wilson.

Writing – original draft: Aravind Asokan, Peyton G. Hickman, Matthew H. Wilson.

Writing – review & editing: Aravind Asokan, Peyton G. Hickman, Matthew H. Wilson.

Funding

None.