Live long and persist: polyomavirus immune evasion in the brain and kidney

Abstract

Polyomaviruses (PyVs) are widespread commensals among vertebrates, including humans, where they silently persist lifelong in healthy hosts. Polyomavirus infection in immunocompromised individuals can cause life-threatening diseases. Of the 14 human polyomaviruses discovered to date, resurgent infections by the JC and BK PyVs are responsible for high morbidity and mortality in individuals with certain inherited or acquired immune perturbations. JCPyV causes several brain disorders, the most fully characterized and of highest (albeit rare) incidence being Progressive Multifocal Leukoencephalopathy (PML). BKPyV infection elicits a diffuse interstitial nephritis in up to 10% of allograft kidneys, and approximately 10% of allogeneic hematopoietic stem cell transplant recipients develop BKPyV-associated hemorrhagic cystitis. No clinically efficacious anti-PyV agents are available. Because PyVs are species-specific, determinants of pathogenesis by human PyVs are inferred from infection of cells in tissue culture. Insights into viral and immunological factors that enable PyVs to persist and cause central nervous system (CNS) and kidney disease in vivo have emerged from recent studies using mouse PyV (MuPyV), a natural murine pathogen. In this perspective, we discuss recent findings using the MuPyV-mouse model to understand early immunovirologic events of CNS and kidney infection, the development of PyV antiviral agents, and promising research directions for polyomavirology.

1. Introduction

JCPyV infects over 70% of the population worldwide, with initial exposure occurring in childhood via fecal-oral or respiratory routes. Despite low-level replication of JCPyV in the urinary tract, healthy carriers remain asymptomatic [1]. If a person with JCPyV becomes immunocompromised, the virus levels can surge in the kidneys, enter the bloodstream, and potentially invade the brain [2]. Infection of oligodendrocytes and astrocytes results in several encephalopathies having significant mortality [3]. Infected oligodendrocytes, where LT expression results in G2 arrest and subsequent apoptosis, lead to the white matter lesions pathognomonic for PML [4,5].

During the AIDS epidemic, PML cases of this rare but often fatal brain disease rose from approximately 0.1% to 8% [6]. Depending on the location of the white matter lesions, PML patients manifest different cognitive and motor dysfunctions, including vision defects, loss of muscle control, and personality changes. With the advent of HAART (highly active antiretroviral therapy), the incidence of PML dramatically declined [7]. In recent years, however, PML has reemerged with the introduction of an arsenal of immunomodulatory therapies to treat autoimmune diseases (e.g., multiple sclerosis (MS), rheumatoid arthritis) and inflammatory disorders (e.g., ulcerative colitis and Crohn’s disease) [8–10]. Natalizumab, a humanized α4 integrin monoclonal antibody, blocks trafficking of T lymphocytes into the CNS and profoundly benefits patients with relapsing-remitting multiple sclerosis (MS); however, natalizumab may also disrupt marginal zone (MZ) and memory B cell responses [11]. The anti-CD20 IgG biologic, Rituxamab, depletes B cells and is a mainstay treatment for chronic lymphocytic leukemia, rheumatoid arthritis, and non-Hodgkin lymphoma. Rituximab and second-generation CD20 monoclonal antibodies (e.g., ocrelizumab and ofatumumab) are highly effective disease-modifying therapies for MS. The Janus kinase (JAKs) 1 and 2 inhibitor used to treat myelofibrosis, Ruxolitinab, has also been associated with the development of PML and granule cell neuropathy [12,13]. The variety of targets for these immunomodulatory therapies – disrupting both innate and adaptive arms of host immunity – demonstrates the difficulty of teasing out the conditions that predispose a patient to PML development. With newly diagnosed PML, current clinical protocols involve tapering off immunomodulatory therapies or reversing immunosuppression, leading to the potential rejection of transplanted organs or the high-mortality immune reconstitution inflammatory syndrome (IRIS) in the CNS [14].

BKPyV is the major infectious cause of dysfunction and loss of renal allografts [15]. Like JCPyV, most humans are seropositive for BKPyV and asymptomatic [16]. Polyomavirus-associated nephropathy (PVAN) occurs in up to 10% of kidney transplant patients [17–19]. Analysis of BKPyV serotypes and genotypes in kidney donors and recipients revealed that the source of BKPyV causing PVAN originates from the donor kidney [20]. BKPyV infection can also cause hemorrhagic cystitis, an inflammatory bladder condition, and bladder carcinoma after organ transplantation [21]. Primary human renal proximal tubule epithelial cells infected with BKPyV fail to upregulate expression of interferon type I (IFN-I) stimulated genes, while those of JCPyV do, suggesting that BKPyV interdicts innate antiviral immunity to facilitate its persistence in the kidney [22]. JCPyV’s inability to thwart innate immunity may account for the rarity of JCPyV-associated nephropathy [23].

JCPyV and BKPyV are non-enveloped “small” DNA viruses of approximately 50 nm from the family Polyomaviridae and the genus Orthopolyomavirus [24]. The double-stranded covalently closed DNA genome encodes 2–3 nonstructural T antigens that regulate host cell cycle, viral DNA replication, and viral transcription, and three proteins, designated VP1–3, that form the T = 7 icosahedral capsid composed of 72 VP1 pentamers. MuPyV is a member of the same genus as JCPyV and BKPyV [25]. MuPyV infection in wild-type (WT) mice shows similar transmission routes and establishes a multi-organ silent persistent infection, including in the brain and kidney [26]. Due to its approximately 5-kb genome and overlapping open reading frames (ORFs), manipulating the PyV genome without handicapping viral viability is challenging. In addition, the rarity of PML, its late stage at the time of diagnosis, and the invasive nature of brain-based sampling limit the availability of clinical samples to interrogate mechanisms of pathogenesis. These limitations showcase the expanded translational applications of the MuPyV-mouse model.

How JCPyV traffics to the brain and enters the CNS is unknown, as is how differences in underlying immunodeficiencies may affect the magnitude and pattern of systemic viral spread [1]. Moreover, in non-barrier tissues (e.g., brain and kidney), our understanding of chemokine receptor-chemokine axes guiding T cell entry, T cell differentiation, functionality, and maintenance in response to persistent microbial infections is incomplete. Our group has shown that MuPyVs with mutations in VP1 emerge under conditions of CD4⁺ T cell deficiency and low antibody (Ab) coverage of viral epitopes [26]. Compared to WT MuPyV, these mutations impair infection in the kidney while preserving the ability to infect the brain. These findings align with those showing that JCPyV isolates from the blood and CNS, but not the urine, of PML patients have non-synonymous VP1 mutations within its external loops that mediate attachment to host cell sialylated glycolipids and glycoproteins. Together, these VP1 mutations impair binding by neutralizing Ab (nAb) and can concomitantly shift viral tropism [27–29]. Thus, the MuPyV CNS infection model shows that CD4⁺ T cell insufficiency leads to a shortfall in epitope coverage by VP1 nAbs, resulting in outgrowth of VP1 mutant viruses with altered tropism. These findings support a mechanism connecting CD4⁺ T cell immunodeficiency with VP1 mutations that confer neurovirulence in JCPyV-PML. nAb-escape VP1 mutations have also been found in BKPyV sequenced from the kidneys of renal transplant patients [30,31]. In addition, JCPyV-PML isolates have rearrangements in the approximately 500-bp noncoding region that contains the origin of replication, binding sites for host transcription factors, and early and late promoters for the early primary transcript encoding the T antigens and a late primary transcript encoding the capsid proteins. These noncoding region rearrangements facilitate replication of the JCPyV DNA in glial cells [32].

The work from our laboratory and other polyomavirus laboratories seeks to develop a fundamental understanding of immunovirology to benefit patients afflicted with JCPyV- and BKPyV-associated diseases (Figure 1). We foresee use of the MuPyV model serving as the basis for uncovering mechanisms linking particular immunomodulatory agents with PML, defining cell types infected by the virus that transport virus between organs and within tissues, and how innate and adaptive immune effector cells find and contain infectious foci in the brain and kidney. Answers to these questions will be instrumental in modifying immunomodulatory regimens to minimize the risk for developing PyV-associated diseases and identifying targets for antiviral agents.

Figure 1.

Recent discoveries about the immune response to MuPyV in the spleen, kidney, and brain and outstanding questions in the field. Created in https://BioRender.com.

2. How does CD4⁺ T cell immunodeficiency promote outgrowth of nAb-escape PyV variants?

CD4⁺ T cell deficiency, either inherited (e.g., idiopathic CD4 lymphopenia) or acquired (e.g., AIDS, immunomodulatory monoclonal (mAb) therapies, chemotherapies), is a major predisposing condition for PML [26,33–35]. Additionally, the reconstitution of CD4⁺ T cells following kidney transplant is correlated with the decrease of BKPyV viral loads. The CD4⁺ T cell response to BKPyV has its own cytotoxic abilities as well as maintains an anti-BKPyV CD8⁺ T cell response, but it could also affect the anti-BKPyV antibody (Ab) response [36–38]. As the helper cells of the adaptive immune response, one of CD4⁺ T cells’ primary responsibilities is to provide signals to B cells to form germinal centers (GCs). This is called the T-dependent (TD) B cell response [39]. The TD B cell response is initiated when subset of CD4⁺ T cells termed follicular helper cells (T_FH) engage B cells to form and enter the GCs micro-environments where B cell receptors (BCR) undergo somatic hypermutation (SHM) to increase Ig diversity and selection of high-affinity BCRs. GC B cells are the primary source of memory B cells and Ab-producing long-lived plasma cells (LLPCs) [40]. Traditionally, it was believed that class-switched recombination (CSR) to determine Ab effector function of BCR genes also occurred within the GC, but recent studies indicated that CSR happens prior to B cells seeding or entering GCs [41,42]. B cells also have an innate-like role. B cells can proliferate prior to the availability of T_FH cells or in situations lacking T_FH cells [the T-independent (TI) response]. In response to bacteria, TI B cells quickly expand into short-lived PCs that primarily secrete IgM. First described as a response to bacteria, amassing literature points to a virus-specific TI response as well. Influenza, SARS-CoV-2, and PyVs all mount a TI B cell response [43–47]. This response undergoes CSR and possibly SMH in high-inflammation environments [48]. MuPyV generates a virus-specific TI IgG response primarily comprised of IgG2 and IgG1 that controls virus infection as well as the TD response during early infection. TI IgG response, however, fails to sustain protection during MuPyV persistent infection [49]. TI IgG is detected throughout persistent infection, but loss of virus control is due to TI IgGs being both lower in avidity and of lower concentration than the TD anti-MuPyV IgG response – setting the stage for emergence of variants with VP1 mutations that enable evasion of nAbs.

These nAb-escape VP1 mutations occur under “Goldilocks” conditions. If there is a robust, diverse, and high-avidity VP1 IgG nAb response, as found in a healthy person, the virus is controlled and maintained as a smoldering, persistent infection. Conversely, we have observed that uMT mice lacking mature B cells, and therefore have no Abs, are chronically viremic with WT MuPyV [26,50]. VP1 mutations arise in individuals with a limited, oligoclonal, or near-monoclonal IgG VP1 epitope repertoire. This “just right” amount of nAb is insufficient to fully control the virus, but enough to put pressure on the virus to select those with VP1 nAb escape mutations. The TI IgG antiviral response constitutes this type of condition, which is conducive to the outgrowth of MuPyVs carrying VP1 mutations conferring resistance to nAbs.

The B cell source of TI IgG is unclear. Two candidates are B1b cells or MZ B cells. B1b cells, however, are predominantly situated in the peritoneal and pleural cavities, although they are present in low numbers in secondary lymphoid organs (e.g., spleen and lymph nodes) and bone marrow. While B1b cells are known to be involved in TI responses to bacteria, they have only been reported to generate a “natural” Ab response consisting of IgM, IgG3, and IgA [51,52]. The inbred mouse strain, MA/MyJ, generates a natural Ab response that protects against MuPyV. While Carroll et al. did not specify which B cell subset was responsible for this anti-MuPyV natural Ab response, it is possible that B1b cells are involved in generating these anti-MuPyV Abs [53]. B1b cells, however, seem unlikely to be the source of the anti-MuPyV TI response due to their primary location and preference to generate an IgM response. Option 2, the MZ B cells, are vital in the early response to blood-borne antigens and quickly respond to antigens found in blood filtered by the spleen [54]. There has been a case report showing the development of PML in a patient with MZ B-cell lymphoma [55]. Additionally, adoptive transfer of MZ B cells to SCID mice was sufficient to prevent a lethal MuPyV-induced myeloproliferative disease [46]. Collectively, these findings favor the likelihood that MZ B cells are the source of anti-MuPyV TI IgGs. Advancing this possibility, antibody-mediated blockade of α4β1 integrin very late antigen-4 (VLA-4) adhesion molecule and lymphocyte function-associated antigen-1 (LFA-1) disrupts MZs in the spleen [56]. The PML-associated mAb biologic natalizumab targets α4 integrins and is highly associated with the development of PML. By depleting MZ B cells, it is conceivable that natalizumab may dampen the anti-JCPyV nAb TI response and thereby contribute to the pathogenesis of PML in patients with relative/absolute CD4⁺ T cell deficiency [11].

Understanding the signals provided to MZ B cells that promote the TI IgG response could provide a potential therapeutic target. It has been reported, and we have confirmed, that the adapter protein MyD88 is necessary for the long-term maintenance of the anti-MuPyV TI IgG response [47,57]. The MyD88 adaptor protein is best characterized for mediating signal transduction from most toll-like receptors (TLRs). TLRs recognize conserved molecular patterns and are highly involved in the innate immune responses to highly repetitive structures, such as the 72 VP1 pentamers forming the PyV capsid. It is reasonable to hypothesize that TLRs on MZ B cells may be involved in the anti-MuPyV TI IgG response, as TLRs are highly expressed on MZ B cells, and that some TLRs (i.e., TLR4, TLR7, and TLR9) are involved in the TI responses to bacteria and other viruses [58–60]. Additional literature suggests that MuPyV and BKPyV engage TLR4 and that the loss of TLR4 signaling increases susceptibility to infection [61–65]. TLR4 recognizes LPS from Gram-negative bacteria, raising into question how TLR4 can bind PyV virions. Another possibility is that MyD88 mediates signaling by the cell surface receptor transmembrane activator and CAML interactor (TACI). Like TLRs, TACI is also known to be involved in TI IgG response and is highly expressed on MZ B cells [66,67]. Additionally, TACI signaling through MyD88 has been linked to CSR, matching that seen in the anti-MuPyV TI IgG response [68]. TACI, however, would not directly bind MuPyV virions. Instead, the ligands for TACI are B Cell-Activating Factor (BAFF) and A Proliferation-Inducing Ligand (APRIL) [69]. BAFF and APRIL are secreted by myeloid cells. A plausible model is that PyV in the blood infects splenic MZ macrophages that secrete either BAFF/APRIL to activate TACI-expressing MZ B cells; these activated MZ B cells then differentiate into TI PCs that secrete neutralizing anti-PyV IgG.

We found consistent TI IgG production during persistent MuPyV infection, despite TI anti-MuPyV IgG-secreting PCs being short-lived [49]. These findings indicate that TI PCs must be continuously resupplied to maintain the anti-PyV IgG response. By extension, these data suggest that MuPyV persistent infection is required to maintain the TI IgG. In a CD4⁺ T cell-deficient patient, JCPyV or BKPyV persistent infection of MZ macrophages could drive differentiation of MZ B cells into TI PCs, which in turn elaborate PyV-specific TI IgG. This low-titer, low-avidity IgG response facilitates the emergence of VP1 variants that evade nAb-recognition, and with rare variants acquiring changes in their tropism that incur neuro-/nephro-virulence.

3. What recent advancements in polyomavirus research can inform future PML treatments and diagnostics?

While a number of inflammatory and/or infectious CNS diseases have treatment options, PML does not. The diagnosis itself can be hard to confirm, as an invasive brain biopsy is usually required. Overall, developing treatments for CNS disorders can be extremely difficult due to the variety in nuances of host response to each individual pathogen, as seen in studies examining the recruitment of T cells to the brain after West Nile Virus infection [70–72]. With West Nile virus, both the CXCR4-CXCL12 and the CXCR6-CXCL16 axes can individually regulate T cell entry; however, in our MuPyV system, CXCR4 and CXCR6 dually affect virus-specific T cell recruitment [73]. CXCR6 has been implicated as being critical to T cell and monocyte infiltration in multiple organs, such as the lungs, kidneys, and retina [74–79]. We find that the majority of MuPyV-specific CD8⁺ T cells in the brain co-express CXCR4 and CXCR6, with CXCR6 upregulated during infection. Spatial transcriptomics of acutely infected mice reveals that the ependyma constitutively produces CXCL12 and CXCL16 regardless of the state of infection.

Infiltrating T cells are vital for the clearance of CNS pathogens; however, their numbers must be regulated to limit tissue-damaging inflammation. If CXCR4 or CXCR6 is inhibited alone, there are no changes in virus level and lymphocyte numbers. When CXCR6^−/− mice are treated with AMD3100, a FDA-approved CXCR4 antagonist, an increase of virus-specific CD8⁺ T cells and CD4⁺ T cells in the brain following intracranial IC) inoculation is seen, but this is not the case in WT mice given AMD3100 [80,81]. This result points toward the high stringency needed to balance viral control against neuroinflammation. A treatment consisting of a combination of CXCR4 and CXCR6 blocking therapeutics may be an innovative treatment for PML, with the caveat that such intervention may elevate CNS injury.

The route and mechanism of JCPyV neuroinvasion are major open questions. The blood–brain barrier (BBB) has historically been considered the access portal via which JCPyV invades the brain parenchyma, although recent studies by Atwood and colleagues support the likelihood that the blood cerebrospinal fluid (BCSF) barrier is the major route of entry into the CNS. JCPyV may breach these barriers as free infectious virus, infected myeloid cell “Trojan horses,” or JCPyV encased within extracellular vesicles (EVs) released from infected cells. EVs shield the virus from nAbs and may serve as a vehicle for trafficking virus from the periphery to the CNS [82]. A recent study demonstrates that JCPyV EVs may cross polarized brain endothelial monolayers; while not infected themselves, a small amount of JCPyV EVs transit apically-to-basally to infect glial cells [83]. Using choroid plexus epithelial cell (CPE) monolayers in transwell culture, O’Hara et al. showed that free virions and virus-laden EVs productively infect CPE and allow JCPyV to efficiently infect glia [83]. Inhibitors of clathrin and lipid raft-dependent transcytosis revealed that JCPyV-EVs and free virions use different pathways and directions to cross CPE. Infection of the basolateral surface of the CPE by free virus with release of JCPyV-EVs at the apical surface could seed the CSF with both EVs and free virions. The implication of JCPyV-EVs and free virus in neuro-invasion represents another target for new therapeutic development, including inhibiting EV formation or antivirals to block virus binding or replication, and at-risk patient monitoring plans.

4. Can the MuPyV mouse model be extended to kidney pathology and infection?

One source of control of T cells, including their metabolism, inhibitory receptors, and activation state, is chemokines and their respective ligands. Both CXCR3 and CXCR6 have been connected to T cell maintenance and memory T cell generation, warranting further investigation in the context of MuPyV, especially in a non-barrier tissue like the kidney [77,84,85]. CXCR6 knockout mice confirmed the importance of the CXCR6-CXCL16 axis for response to MuPyV, presenting with lowered infiltrating T cell numbers and poor virus control compared to the WT mice. The retention of CD8⁺ T cells was impaired in the knockout mice, which also contributed to the increased kidney pathology. As a “reality check,” CD8⁺ T cells in a needle biopsy from a kidney transplant patient with PVAN also expressed CXCR6, and bioinformatic analysis of publicly available RNAseq datasets from PVAN allograft kidneys showed significant upregulation of CXCR6 and CXCL16 transcripts. This work validates the use of the MuPyV model for studying kidney pathology during a polyomavirus infection and is a great advance in multiple fields [86]. From the unraveling of the basis of T cell control, the issue of a lack of balance in immune response can be addressed; in transplant cases, the donor T cell response is responsible for the loss of graft as the CXCR axis may represent a focus for better targeted immune modulating medications to increase long-term survival.

This question has been explored in a MuPyV infection of murine kidney allograft system [87]. Full MHC-I and MHC-II mismatch between the donor kidney and recipient resulted in high levels of MuPyV replication in the transplanted kidney and a robust anti-donor T cell response leading to allograft loss [88,89]. Despite departures from the clinical scenario (i.e., recipient mice were not immunosuppressed and naïve mice were acutely infected), these studies represent important steps in the development of a PVAN mouse model.

5. What is the future for polyomavirus antiviral agents?

A litany of anti-PyV agents with tolerable cytotoxicity have been described, but all have been “lost in translation” in treating patients with PyV-associated diseases. A number of small-molecule compounds have been identified as inhibiting viral replication in some manner; most of their mechanism of action are unknown [90]. The compound GW-5074 decreases JCPyV infection in cell culture through disruption of MAPK-ERK signaling via phosphorylation inhibition [91]. Antagonists for α₂-adrenergic, 5HT_2A- and 5HT₃-receptors, like Mirtazapine, a treatment for depression, have been used off-label for treating JCPyV based on evidence that 5HT_2A is an uptake receptor where JCPyV binds [92,93]. This has been tested in HIV patients as well as in patients receiving immunomodulating therapy, resulting in improvement of PML symptoms. Serotonin-dopamine inhibitors (SDIs) drugs, such as Chlorpromazine, were then tested in vitro to a similar result, mitigating JCPyV infection via the blockade of clathrin-dependent endocytosis [93,94]. Drugs from this class are already widely in use and well studied in human subjects, but again, advancing therapies from in vitro virus infection systems may fail to translate into successful clinical trials. Designing decoy receptors to compete with susceptible host cells for JCPyV is another attractive approach. JCPyV binds to the (2,6)-linked sialic acid moiety of its attachment receptor, Lactotetrasaccharide series C (LSTc). Soluble LSTc reduces JCPyV infection in vitro [95]. A caveat here is that LSTc is not expressed by astrocytes, oligodendrocytesand glia infected by JCPyV. JCPyV-EVs enable attachment receptor-independent infection of these glia; EV-mediated infection would circumvent LSTc decoy receptors.

We recently explored the potential of Brincidofovir (BCV) as a PyV antiviral. BCV is currently FDA-approved to treat small-pox virus infections [96,97]. BCV is a lipid–cidofovir conjugate lacking cidofovir’s nephrotoxicity; its nucleotide analog replaces deoxycytidine triphosphate and is expected to inhibit replication of viral DNA genomes [98]. Three weeks of treatment with Brincidofovir before MuPyV infection was well tolerated by the mice and limited acute MuPyV infection both in the brain and the kidney. In B cell-deficient, chronically MuPyVemic mice, BCV decreased viremia in the blood and virus levels in the kidney. Future examination of similar nucleotide analogs may reveal other novel treatment avenues, even though some have already been examined. Cidofovir, for example, has been tried in a small number of PML patients but did not slow or reverse disease progression; however, Brincidofovir reopens this therapeutic door with its increased cellular uptake and higher bioavailability [99]. Another successful in vivo study was able to reduce MuPyV kidney infection through the use of Retro-2.1, which inhibits retrograde transport [100]. In an acutely infected immunocompromised mouse model, this treatment showed therapeutic promise without impacting the host T cell response or kidney function. Retrograde transport inhibition first showed promise in cell culture by limiting infection of BKPyV, JCPyV, and simian virus 40 (SV40) [101]. These studies further serve to strengthen the foundation of potential treatment options and MuPyV as a paradigm for polyomavirus research.

6. Conclusions

In vitro studies with JCPyV and BKPyV infection systems, analysis of virus infection and immune cell infiltrates in brain and kidney biopsies, coupled with in vivo work using MuPyV, are helping to fill in the picture of PyV-induced/-associated pathogenesis and antiviral defense/evasion. Recent advances show promise for chemokine-based interventions, as well as understanding how different types of immune compromise facilitate the emergence of pathogenic variants that escape viral immunity. PML patients may benefit from checkpoint blockade immunotherapy to harness host PyV-specific T cells as well as autologous/allogenic JCPyV-specific T cell therapy [33,102]. Inhibitors of JCPyV and BKPyV uptake and replication continue to be uncovered and validated in vivo using the MuPyV-mouse model.

7. Future perspectives

Although we pose a few potential treatment options, there is much room for improvement in terms of the specificity of these approaches. Modulating immune responses with a wide sweeping approach fosters the concern of off-target effects and over-response from host defenses. Knowing cell types harboring PyV infection, as well as the chemokine gradients that direct their infiltration into different organs and tissues will potentially inform the design of new antiviral therapeutics.

Clinical trials of adoptive HLA-matched, virus-specific T cell therapy for PML or PVAN are underway [103,104]. As some of the kidney studies highlighted in this article show the importance of CXCR6 expression on virus-specific CD8⁺ T cells for virus control, virus-specific T cells overexpressing CXCR6 could be given to PVAN patients with HLA-compatible kidney transplants to promote BKPyV control. This approach has already been taken in a pancreatic cancer model, improving the T cell anti-tumor response, demonstrating promise for use in renal allografts [105]. The caveat to this is that a virus under pressure from the immunological response may mutate epitopes for virus-specific CD4⁺ and CD8⁺ T cells, rendering the treatment ineffective [106]. Other viruses, like Hepatitis B virus, have been shown to do this by mutating T cell epitopes to escape detection [107]. There is also a large challenge in finding a balance so as not to induce more pathology by an excess immune response.

Interests in therapeutics may dwindle if a vaccine is developed for JCPyV and BKPyV and added to the childhood vaccination schedule. If the persistent smoldering infection can be eliminated from the general population, then PML or PVAN development is unlikely, removing one lingering concern of immunosuppression. A study in 2023 developed a virus-like particle-based vaccine for ByPyV and JCPyV that was tested with success in rhesus macaques [108].

As the arsenal of immunomodulatory therapies continue to find expanded clinical use, the number of PML and PVAN cases will continue to rise and command increased motivation to develop antiviral therapeutics, vaccines, and identify ways to harness virus-specific immunity while minimizing bystander tissue injury. The MuPyV mouse model will be instrumental in acquiring a comprehensive understanding of human PyV pathogenesis and immunity.

Article highlights.

• Polyomavirus, both JCPyV and BKPyV, contribute to devastating and fatal diseases in the kidney and CNS. A mouse model with MuPyV has been established to answer questions about the immune response and potential therapeutics in vivo.

• The anti-PyV antibody response is crucial for keeping the persistent infection “smoldering.” Patients having an impaired CD4⁺ T cell response due to underlying disease or mAb therapies will generate a low concentration and low avidity anti-PyV TI IgG response that opens the door for VP1-escape variants.

• Dual blockade of CXCR6 and CXCR4 may represent a new avenue of bolstering the immune response to PML, and other future therapeutic targets may need to target EVs.

• MuPyV is also a valuable animal model for PVAN, where the CXCR6-CXCL16 axis modulates virus control in the kidney.

• Brincidofovir and serotonin receptor blocking agents prevent polyomavirus replication in vivo and in vitro, respectively, and could become the first approved JCPyV antiviral treatments.