Animal Models of Polyomavirus Nephropathy: Hope and Reality
Version of Record online: 15 JUN 2006
American Journal of Transplantation
Volume 6, Issue 7, pages 1507–1509, July 2006
How to Cite
Nickeleit, V. (2006), Animal Models of Polyomavirus Nephropathy: Hope and Reality. American Journal of Transplantation, 6: 1507–1509. doi: 10.1111/j.1600-6143.2006.01400.x
- Issue online: 15 JUN 2006
- Version of Record online: 15 JUN 2006
- Received 2 March 2006, revised 15 April 2006 and accepted for publication 16 April 2006
The polyomavirus family consists of multiple virus strains that are highly adapted to specific host organisms. Relevant for humans are JC-, BK- and SV40-viruses, which cause different diseases despite approximately 70% gene homology. Viral illness including polyomavirus nephropathy (PVN) is generally due to the re-activation of latent viral infections (1).
Over the past 8 years, PVN, typically produced by the replication of BK viruses, has become the most important infectious complication affecting kidney transplants. Prior to the mid-1990s, it was practically never seen, even with retrospective searching (2). Thus, PVN represents ‘a new epidemic.’ Viral nephropathy currently has a prevalence of 1%–9% worldwide. At present, specific and effective anti-viral treatment strategies are poorly defined and PVN, especially when diagnosed in the florid (‘B’) and sclerosing (‘C’) stages of the disease, often results in chronic allograft dysfunction and failure (3–6). In kidney transplant recipients, PVN occurs on an average 10–12 months after grafting, is typically limited to the allograft and lacks clinical signs of a generalized infection. Other organs including the native kidneys are only very rarely affected, even in immunosuppressed patients, pointing to specific risk constellations that differ from other viral pathogens, such as cytomegalo- or adenovirus.
PVN must be diagnosed histologically in a kidney biopsy. It is characterized by viral replication in epithelial cell elements, most significantly in renal tubular cells. A productive infection results in intranuclear viral inclusion bodies and virally induced tubular epithelial cell lysis. Interstitial inflammation is largely absent in early disease stages (stage A), but develops during disease progression in association with virally induced epithelial cell injury (disease stages B and C) (6–8). PVN can on occasion be associated with histological signs of rejection (5,9–11). Risk factors promoting viral nephropathy are poorly understood; important for the development of disease appear to be new immunosuppressive drug regimens that can provide the right window of opportunity for viral replication (2,5).
Our current understanding of PVN is primarily based on clinical observations. In animal studies, PVN has either been incidentally detected (12), or the observed polyomavirus-induced injury only vaguely resembles human disease (13–15). In this regard, the unique ‘mouse model for polyomavirus-associated nephropathy of kidney transplants’ designed by Han Lee and colleagues, and recently reported in part 1 of the May issue of American Journal of Transplantation (16) seems to be an exciting prospect. The authors report that their model ‘mimics’ PVN seen in humans and accurately reproduces accelerated kidney failure.
Excitement about this new mouse model may, however, be premature. An animal model, defined by Merriam-Webster's Medical Dictionary, is a pathological condition similar to a human condition in an animal sufficiently like humans in its anatomy, physiology or response to a pathogen (17). Data provided hitherto by Han Lee and collaborators does not unquestionably fulfill these criteria.
- 1The authors postulate that similar to human PVN, polyomaviruses in their mouse model ‘…preferentially replicate in the allogeneic kidney grafts….’ This claim is only partially supported by data. Han Lee and colleagues showed that subsequent to active inoculation of study animals, renal allografts contained high viral DNA and infectious virion loads. These findings, however, simply demonstrate that viral particles were present in the kidney transplants; they do not necessarily prove ‘massive intrarenal viral replication.’ Replication should more clearly be demonstrated by measuring viral RNA levels, by demonstrating the expression of the large T antigen using immunohistochemistry, by demonstrating intranuclear viral inclusion bodies by light microscopy or by demonstrating intranuclear aggregates of viral particles employing electron microscopy.
- 2Marked viral replication in tubular epithelial cells and tubular damage is a sine qua non for any model of PVN. In the mouse, cytopathic changes including intranuclear viral inclusion bodies, acute tubular injury and epithelial cell lysis, i.e. the hallmarks of PVN in humans, have not been described. Therefore, Han Lee and colleagues have to design additional experiments to pinpoint the exact site of viral replication. Interestingly, in some experimental animals, viral replication can, on occasion, be observed in unexpected compartments. For example, in squirrel monkeys, BK virus replication was not found in tubular cells (as anticipated), but rather in endothelial cells (13). If the described mouse model showed any polyomavirus replication in graft infiltrating leukocytes or endothelial cells rather than epithelial cells, then the findings would not mirror human disease.
- 3The described study animals showed marked inflammation of the allografts in the absence of virally induced cytopathic changes; this finding differs from human PVN.
- 4Human PVN typically occurs 1 year after transplantation; it is likely caused by the re-activation of a latent BK virus infection in the graft. In patients, graft failure due to viral nephropathy usually develops over many months. ‘End-stage’ PVN histologically shows marked sclerosis, similar to late ‘chronic allograft nephropathy’ (PVN stage C) (6–8). In contrast, the described mouse model requires active inoculation of the study animals, resulting in very rapid graft failure due to severe inflammation. Consequently, model conditions most resemble a primary infection and do not ‘accurately reproduce…kidney failure that is observed clinically in patients,’ as postulated by Han Lee.
- 5All polyomaviruses are highly selective and major biological differences can be expected among different strains and host organisms. Observations made in mice may be quite different from human clinical scenarios.
Much work remains to be done by Han Lee and colleagues to further characterize and validate their model ‘in order to investigate the pathogenesis of polyomavirus-associated nephropathy.’ Initial observations are, however, intriguing, in particular the virally augmented alloreactive CD8+ T cell response. Could this finding explain the occasionally marked inflammatory reaction seen in human cases of PVN disease stage B, which can be rich in T cells and tubulitis (3,4,6,8,18)? Should this finding be interpreted and treated as acute cellular rejection (11)? Patients with PVN and concurrent graft rejection benefit from anti-rejection therapy based on anecdotal evidence (9,10). However, if infections with polyomaviruses, as suggested by Han Lee, accentuate an anti-donor T cell response, then why is allograft rejection in humans not significantly increased in patients with persistent PVN even when the immunosuppression is lowered (19,20)? Why does viral nephropathy in disease stage A lack marked inflammation (6,20)?
Potentially, some alterations of the study design might render the mouse model clinically more relevant, e.g. what can be learned by transplanting a donor kidney containing polyomaviruses (subsequent to an active inoculation) into an immunosuppressed recipient mouse? What type and level of immunosuppression in the host will promote viral replication and injury in the graft, i.e. PVN?
Hopefully, in the future, the described ‘mouse model’ (16) will provide new insights into the pathogenesis of human PVN.
- 6Polyomavirus nephropathy in native kidneys and renal allografts: An update on an escalating threat. Transpl Int. In press., .
- 7Morphological spectrum of polyoma virus disease in renal allografts: Diagnostic accuracy of urine cytology. Am J Transplant 2001; 1: 373–381., , et al.Direct Link:
- 8Latent and productive polyomavirus infections of renal allografts: Morphological, clinical, and pathophysiological aspects. In: AhsanN, ed. Polyomaviruses and Human Diseases, 1st Ed. New York , NY ; Georgetown , TX : Springer Science + Business Media, Landes Bioscience/Eurekah.com, 2006: 190–200., , .
- 17Merriam-Webster's Medical Desk Dictionary, 2nd (revised) Ed. Thomson Delmar, 2002.