Gabriel J. Starrett

BK polyomavirus (BKPyV) and chronic latent DNA virus infections frequently reactivate after hematopoietic stem cell transplantation (HSCT), where they contribute to major complications in humans with limited options for prevention or treatment. Mauritian cynomolgus macaques (MCMs; Macaca fascicularis ) serve as an important animal model for transplantation research due to their limited major histocompatibility complex diversity. Similar viral reactivations and disease complications occur in MCMs and humans post-transplantation, including polyomavirus-associated hemorrhagic cystitis and nephropathy. To define the polyomaviruses associated with these outcomes and characterize the broader urinary DNA virome, we sequenced rolling circle-amplified urine DNA from healthy and HSCT recipient cynomolgus macaques, allowing for comprehensive genomic characterization of the detected polyomaviruses. De novo assembly and annotation identified three polyomaviruses with apparent specific host tropism to cynomolgus macaques: Macaca fascicularis polyomavirus 2 (MafaPyV2) and Macaca fascicularis polyomavirus 3 (MafaPyV3), which are closely related to human polyomaviruses, and a newly identified strain of simian virus 40 (SV40 type IIB). All three viruses were detected in both HSCT recipients and healthy animals, but were shed at significantly higher relative loads in HSCT recipients. Multiple polyomaviruses were frequently detected within individual hosts, and all three were identified in HSCT recipients that developed urologic disease. Together, these findings further characterize polyomavirus diversity, shedding, and disease associations in hematopoietic stem cell–transplanted Mauritian cynomolgus macaques. IMPORTANCE: Reactivation of latent viruses after hematopoietic stem cell transplantation is a major cause of illness in transplant recipients, yet options for prevention and treatment remain limited. Progress has been hindered by the lack of animal models that accurately reflect human disease. In this study, we show that Mauritian cynomolgus macaques harbor multiple polyomaviruses that reactivate at high levels following transplantation and are associated with urinary tract disease, closely paralleling complications seen in humans. We define the genetic identity of these viruses and demonstrate that immunosuppression strongly enhances viral shedding and disease severity. By establishing clear similarities between viral reactivation in macaques and humans, this work validates the Mauritian cynomolgus macaque transplantation model as a powerful platform for studying virus-driven disease mechanisms and testing antiviral therapies and vaccines relevant to human transplantation.

Hemorrhagic cystitis (HC) caused by BK polyomavirus (BKPyV) reactivation is a serious complication in immunocompromised patients especially in hematopoietic stem cell transplant recipients, yet the effects of conditioning regimens on viral replication remain undefined. Current therapies for HC are largely supportive, and no specific antiviral drugs are approved for BKPyV reactivation, underscoring an urgent need for new interventions. Here, using a cellular model of HC, we provide the first mechanistic evidence that physiologic concentrations of genotoxic chemotherapy, immune-conditioning, and antiviral exposures commonly enhance BKPyV replication by reprogramming host chromatin remodeling machinery independently of immune suppression. Transcriptomic and epigenomic profiling of urothelial cells revealed that these treatments induce host and virus chromatin remodeling, specifically, the SWI/SNF ATP-dependent remodeling complex is activated and creates a regulatory landscape enriched for AP-1 transcription factor motifs, implicating the AP-1/SWI-SNF axis as a key regulator of viral replication. Functional inhibition of SWI/SNF with AU-15330 and FHD-286 consistently suppressed BKPyV replication across all tested therapeutic conditions. Patient urine samples and cellular experiments demonstrated these effects are independent of viral genomic changes. These findings suggest chromatin remodeling as the mechanistic driver of therapy-associated BKPyV reactivation and propose SWI/SNF inhibition as a translational therapeutic strategy to mitigate hemorrhagic cystitis.

Despite extensive sequencing, the genetic etiology of sporadic angiosarcoma remains poorly defined. Maffucci syndrome, characterized by vascular tumors and elevated cancer risk, is driven by mosaic gain-of-function mutations in IDH1/2, though these have not been reported in sporadic angiosarcoma. We identify recurrent, low-variant allele frequency hotspot mutations in IDH1/2 in over half of sporadic angiosarcomas. Mutations were validated by Sanger sequencing and immunohistochemistry. Mutant IDH1 endothelial cells promote tumorigenesis through non-cell-autonomous mechanisms, secreting 2-hydroxyglutarate (2-HG) to increase growth factor and endothelial-to-mesenchymal transition gene expression, activate pAkt/pERK signaling, induce DNA methylation changes, and promote anchorage-independent growth, which are reversed by the mutant IDH1 inhibitor ivosidenib. Patients with mosaic IDH1 mutations show reduced serum 2-HG and marked tumor regression following ivosidenib treatment. The clinical efficacy of ivosidenib in vascular tumors with subclonal IDH1 mutations suggests that low VAF IDH1/2 mutations may be a targetable vulnerability in sporadic angiosarcoma.

BK polyomavirus (BKPyV) and chronic latent DNA virus infections frequently reactivate after hematopoietic stem cell transplantation (HSCT), where they contribute to major complications in humans with limited options for prevention or treatment. Mauritian cynomolgus macaques (MCM; Macaca fascicularis ) serve as an important animal model for transplantation research due to their limited major histocompatibility complex (MHC) diversity. Similar viral reactivations and disease complications occur in MCMs and humans post-transplantation, including polyomavirus-associated hemorrhagic cystitis and nephropathy. To define the polyomaviruses associated with these outcomes and characterize the broader urinary DNA virome, we sequenced rolling circle-amplified urine DNA from HSCT recipient and healthy cynomolgus macaques, allowing for comprehensive genomic characterization of the detected polyomaviruses. De novo assembly and annotation identified three polyomaviruses with apparent specific host tropism to cynomolgus macaques: Macaca fascicularis polyomavirus 2 (MafaPyV2) and Macaca fascicularis polyomavirus 3 (MafaPyV3), which are closely related to human polyomaviruses, and a newly identified strain of simian virus 40 (SV40 type IIB). All three viruses were detected in both HSCT recipients and healthy animals but were shed at significantly higher relative loads in HSCT recipients. Multiple polyomaviruses were frequently detected within individual hosts, and all three were identified in HSCT recipients that developed urologic disease. Together, these findings further characterize polyomavirus diversity, shedding, and disease associations in hematopoietic stem cell–transplanted Mauritian cynomolgus macaques. Importance: Reactivation of latent viruses after hematopoietic stem cell transplantation is a major cause of illness in transplant recipients, yet options for prevention and treatment remain limited. Progress has been hindered by the lack of animal models that accurately reflect human disease. In this study, we show that Mauritian cynomolgus macaques harbor multiple polyomaviruses that reactivate at high levels following transplantation and are associated with urinary tract disease, closely paralleling complications seen in humans. We define the genetic identity of these viruses and demonstrate that immunosuppression strongly enhances viral shedding and disease severity. By establishing clear similarities between viral reactivation in macaques and humans, this work validates the Mauritian cynomolgus macaque transplantation model as a powerful platform for studying virus-driven disease mechanisms and testing antiviral therapies and vaccines relevant to human transplantation.

Carcinogenesis in human urothelium is driven by a high burden of mutations caused by the antiviral “APOBEC3” (apolipoprotein B mRNA editing enzyme, catalytic subunit–like 3) cytosine deaminase enzymes; however, there is no established viral etiology. BK polyomavirus (BKPyV) is a ubiquitous childhood infection that persists in the kidney during adulthood and is frequently detected in urine. Chronic BKPyV infections of normal human urothelium induced an innate response, including apical extrusion of infected cells. Local paracrine interferon signaling induced APOBEC3 expression in both infected and juxtaposed bystander cells, leading to acquisition of hallmark APOBEC3-mediated mutational signatures that recapitulated the variation in mutational character found in patients with muscle-invasive bladder cancer. In our model for urothelial carcinogenesis, uninfected bystander cells witnessing BKPyV infection become APOBEC3 damaged, escape extrusion, and acquire hypermutable advantage. “Transmutagenesis” explains how cells proximal to infected neighbors acquire cancer-initiating mutations. This hypothesis for how urothelial cancers can develop as APOBEC3 signature rich, while remaining virus-negative, suggests that a large proportion of urothelial carcinomas may be preventable by antiviral intervention.

Table S1: Patient summary of demographic characteristics and mutation status in individual sets and the combined BC set

Table S2: Accession numbers of viruses evaluated in RNA-seq of bladder tumors from UROMOL

Table S7: Relationship of PyV status with FGFR3/PIK3CA-mutation status and cigarette smoking

Table S4: Top recurrent mutations in FGFR3 and PIK3CA within NMIBC (1,573 patients) and MIBC (1,243 patients)