The human polyomavirus BK: Potential role in cancer



In human cancer, a role has been suggested for the human polyomavirus BK, primarily associated with tubulointerstitial nephritis and ureteric stenosis in renal transplant recipients, and with hemorrhagic cystitis in bone marrow transplant (BMT) recipients. After the initial infection, primarily unapparent and without clinical signs, the virus disseminates and establishes a persistent infection in the urinary tract and lymphocytes. There is correlative evidence regarding potential role of polyomavirus BK in cancer. In fact, the BK virus (BKV) DNA (complete genome and/or subgenomic fragments containing the early region) is able to transform embryonic fibroblasts and cells cultured from kidney and brain of hamster, mouse, rat, rabbit, and monkey. Nevertheless, transformation of human cells by BKV is inefficient and often abortive. Evidence supporting a possible role for BKV in human cancer has accumulated slowly in recent years, after the advent of polymerase chain reaction (PCR). BKV is known to commonly establish persistent infections in people and to be excreted in the urine by individuals who are asymptomatic, complicating the evaluation of its potential role in development of human cancer. Therefore, there is no certain proof that human polyomavirus BK directly causes the cancer in humans or acts as a cofactor in the pathogenesis of some types of human cancer. © 2005 Wiley-Liss, Inc.

The BK virus (BKV) was isolated for the first time in 1971 from the urine of an immunocompromised renal transplant patient (39 years old) whose initials were B. K. This patient was a Sudanese male who underwent a living, related, renal transplant and presented with graft dysfunction and features of ureteric stenosis 4 months post-transplant (Gardner et al., 1971). The human polyomavirus BK is closely related to human polyomavirus JC (JCV), also recovered in 1971 from the brain of an American patient, whose initials were J. C., with Hodgkin's disease and progressive multifocal leucoenkephalopathy (PML). BKV and JCV are antigenically different from one another and from all other animal polyomaviruses although, cross-reactions occur by using hyperimmune antisera. BKV, JCV, and simian virus 40 (SV40) share a minor, common antigenic determinant which is exposed on the surface of virions (Penney and Narayan, 1973; Takemoto and Mullarkey, 1973; Dougherty and DiStefano, 1974; Padgett and Walker, 1976; Jin and Gibson, 1996). BKV is ubiquitous in human populations worldwide (Padgett and Walker, 1976) except in some segregated populations living in isolated regions of Brazil, Paraguay, and Malaysia (Brown et al., 1975). Usually, the human polyomavirus BK infects young children and the seroprevalence is 70%–80% in adults (Knowles et al., 2003). Serologic surveys of populations using hemagglutination inhibition assays for the detection of antibodies indicate that seroconversion takes place early in life, at 5–7 years of age (Taguchi et al., 1982). Primary infection by BKV is usually inapparent and only occasionally may be accompanied by mild respiratory illness or urinary tract disease. During primary infection viremia occurs and the virus spreads to several organs of the infected individual where it remains in a latent state. After the initial infection, the virus disseminates and establishes a persistent infection in the urinary tract and lymphocytes (Dorries et al., 1994; Degener et al., 1997; Imperiale, 2000). Virus isolation and Southern blot analysis established that the kidney is the main site of BKV latency in healthy individuals (Barbanti-Brodano et al., 1998; Li et al., 2002). By various technical approaches, such as virus isolation, Southern blot, and polymerase chain reaction (PCR), BKV sequences were also detected in other organs: liver, stomach, lungs, parathyroid glands, and lymph nodes (Israel et al., 1978; Pater et al., 1980). Moreover, BKV and JCV can be detected in tonsillar tissue from both pediatric and adult donors (Goudsmit et al., 1982), and it has been reported that JCV can replicate in tonsillar B lymphocytes and stromal cells (Monaco et al., 1996), suggesting that the respiratory tract may be the initial site of viral infection. In fact, JCV is also able to infect tonsillar stromal cells productively in culture with efficiency nearly comparable to human glial cells. BKV has not been shown to infect stromal cells in culture. Since JCV and BKV viruses are not stable under conditions consistent with oral ingestion of virus, the inhalation of virus or a respiratory pathway for initial infection seems most likely. Therefore, the BKV use of lymphoid tissue for initiation of infection needs further study (Major, 2001).


All polyomaviruses have a closed circular dsDNA genome. The complete sequences of BKV contain 5,153 bp for strain DUN (Seif et al., 1979b), 4,963 bp for strain MM (Yang and Wu, 1979; Tavis et al., 1989), and 5,098 bp for strain AS (Yoshiike and Takemoto, 1986). The predicted amino acid sequences of the BKV share 73% homology with those if SV40 (Yoshiike and Takemoto, 1986) and 75% with those of JCV (Wolker and Frisque, 1986). The genome of BKV is functionally divided into three regions: the early, the late, and the non-coding control region (Fig. 1). The first region codes for the small and large T-antigens (t-Ag and T-Ag), the second region codes for the viral capsid proteins VP1–VP2–VP3 and agno-protein, and the last region (NCCR) contains the transcriptional control elements for both “early” and “late” gene expression (Seif et al., 1979a,b; Yoshiike and Takemoto, 1986). The direction of early and late transcription is divergent, with opposite DNA strands participating in these processes (Lednicky and Butel, 1999). The primary transcript from the early region is alternatively spliced in two mRNAs encoding large and small T-antigen (T-Ag and t-Ag). T-Ag is a large nuclear phosphoprotein and it is an essential factor for viral DNA replication. It binds to the viral origin of replication region (ori), where it promotes unwinding of the double helix and recruitment of cellular proteins required for DNA synthesis, including DNA polymerase-α and replication protein A (Stahl et al., 1986; Dean et al., 1987; Melendy and Stillman, 1993). Another major function of T-Ag is to support S-phase induction accounting for the ability of T-Ag to transform cells. It is known that T-Ag stimulates the cell cycle machinery by binding to several cellular proteins involved in crucial signal transduction pathways that control cell cycle progression and apoptosis (White and Khalili, 2004). The role of the t-Ag in the polyomavirus life cycle is less clear. Analysis of SV40 deletion mutants revealed that t-Ag is not essential for lytic infection in culture (Shenk et al., 1976). However, t-Ag cooperates with T-Ag in the transformation of cells by SV40 (Sleigh et al., 1978; Martin et al., 1979) and increases virus yield in permissive cell infections (Rundell and Parakati, 2001). As viral replication proceeds, the late genes start to be expressed. T-Ag acts to stimulate transcription from the late promoter and repress transcription from the early promoter. The gene products of the late region are the capsid proteins VP1, VP2, and VP3. These proteins assemble with the replicated viral DNA to form virions and are released upon cell lysis (White and Khalili, 2004).

Figure 1.

Schematic representation of the gene organization in the BK virus (BKV) genome. The double circle represents the double stranded DNA genomes. The genome is divided into three regions. The early region encodes three regulatory proteins (Agt, AgT, T'). The late region specifies four structural proteins and agnoprotein (VP1, VP2, VP3, VPx). The non-coding control region contains the elements for the control of viral DNA replication (ori) and viral gene expression. The arrows indicate the positive and negative strands according to the direction of viral transcription.

The NCCR contains the origin of replication that marks the point of commencement of a bi-directional process of DNA replication during the viral life cycle. Also, it contains sequence blocks arbitrarily referred to by the alphabetical designations P, Q, R, and S. These sequence blocks serve as regulatory regions, or enhancer elements believed to contain several transcription factor binding sites involved in the modulation of viral transcription (Moens et al., 1995). There is “in vitro” evidence that NCCR variants determine host cell permissiveness and rate of viral replication (Johnsen et al., 1995; Daniel et al., 1996). Actually, it is not known if genetic alterations are essential for the pathogenesis of BK virus allograft nephropathy (BK-VAN) after kidney transplantation, but BK-strains with rearranged NCCR have been described in nephropathy associated with other immunosuppressed states (Smith et al., 1998; Stoner et al., 2002; Randhawa et al., 2003). In the non-permissive infection, some portion of the BKV DNA can be randomly integrated into the host chromosomal DNA (Chenciner et al., 1980).


Primary infection with BKV in healthy children is usually asymptomatic. The secondary infection may be due to reactivation of latent virus or re-infection with a new strain (Lin et al., 2001). Clinical studies carried out in immunosuppressed and immunocompetent patients indicate that the reactivation of BKV from latency is mainly associated with immunological impairment (Padgett and Walker, 1976; Tognon et al., 2003).

BKV infection may be transmitted via the donor organ or may be acquired in the community. Mild pyrexia, malaise, vomiting, respiratory illness, pericarditis, and transient hepatic dysfunction have been reported with primary BKV infection.

Difficulties may arise in distinguishing between BKV infection with asymptomatic viral shedding in the urine and BKV disease which is associated with tubulointerstitial nephritis (Gardner et al., 1984) and ureteric stenosis in renal transplant recipients (Gardner et al., 1971; Hirsch et al., 2002), and hemorrhagic cystitis in bone marrow transplant (BMT) recipients (Arthur et al., 1986; Cottler-Fox et al., 1989; Leung et al., 2001). BKV infection does not always progress to disease, and the histopathological changes in the presence of BKV are not always associated with functional impairment (Lin et al., 2001). As matter of fact, although BKV is routinely found in urine of renal allograft recipients and patients with other kidney diseases (Arthur and Shah, 1989), this virus is also found in the urine of normal, healthy individuals. Asymptomatic viruria with BKV occurs in 0.3% of non-immunosuppressed patients, 3% of pregnant women, 10%–45% of renal transplant recipients, and 50% of bone marrow recipients (Flaegstad et al., 1986; Demeter, 2000; Randhawa and Demetris, 2000). Even if great advances were found in recent years, many questions regarding BKV-induced syndromes in human remain unanswered, and the exact nature, clinical course, and treatment of these syndromes are still challenging issues for future studies.


The early region of the BKV genome encodes the two viral oncoproteins: T-Ag and t-Ag. BKV T-Ag displays multiple functions that alter the normal physiological metabolism of cells, ultimately leading to immortalization and neoplastic transformation (Imperiale, 2000, 2001). T-Ag interacts with tumor suppressor proteins in the cell in order to optimize the environment for viral replication. Moreover, T-antigen transforms rodent cells in culture and can also immortalize human cells in the presence of activated oncogenes such as ras or myc (Shah et al., 1976; Portolani and Borgatti, 1978; Purchio and Fareed, 1979; Takemoto et al., 1979; Grossi et al., 1982; Pater and Pater, 1986). BKV T-Ag binds to pRb family proteins and p53 (Bollag et al., 1989; Dyson et al., 1989; Harris et al., 1996). The binding of BKV T-Ag to p53 stabilizes the last one but interferes with its response to DNA damaging agents (Harris et al., 1998b). In cells expressing BKV T-Ag, it is difficult to detect significant amounts of complexes composed of T-antigen and pRb, p107, or p130, because the low expressed levels of TAg from the BKV early promoter. Nevertheless, BKV T-Ag stimulates the release of the E2F activity from pRb family complexes allowing serum-independent growth. Mutational analysis has demonstrated that the activation of E2F requires both the pRb-binding domain and a newly defined N-terminal domain of the molecule called the J domain. The name of this domain is based on structural and functional homology to the DnaJ family of molecular chaperones. This domain mediates a number of important T-antigen functions related to transcription, replication, and transformation (Campbell et al., 1997; Kelley and Georgopoulos, 1997; Stubdal et al., 1997; Brodsky and Pipas, 1998; Harris et al., 1998a; Sock et al., 1999). In order to explain how this domain functions in oncogenic transformation, it has been proposed that T-Ag interacts with the pRb family protein, causing release of E2F and subsequent J domain-mediated degradation of the pRB family protein (Imperiale, 2000). More, BKV T-Ag induce chromosomal aberrations in cultured cells (Theile and Grabowski, 1990; Tognon et al., 1996; Trabanelli et al., 1998), and increased chromosomal damage correlates with high titers of antibodies against the viruses (Lazutka et al., 1996). T-Ag induces inappropriate cell growth through its interactions with pRb family proteins; it is mutagenic and interferes with the ability of the cell to respond to damaged DNA through its interaction with p53. It is clear that BKV T-Ag has the molecular characteristics to contribute to the oncogenic process. A more detailed molecular analysis will be required before a role for BKV in human cancer can be confirmed or refused, but proofs might be complicated by the fact that a large percentage of the population is infected with this virus. Only tumor cells must be examined in tumor samples, not contaminating normal cells from surrounding tissue, especially when using sensitive techniques such as PCR or RT-PCR.


There is a correlative evidence of the potential role of polyomavirus BK in cancer. In fact, BKV DNA (complete genome and/or subgenomic fragments containing the early region) is able to transform embryonic fibroblasts and cells cultured from kidney and brain of hamster, mouse, rat, rabbit, and monkey (Corallini et al., 2001; Tognon et al., 2003; White and Khalili, 2004). In these experiments, all cells in the culture express BKV TAg in their nuclei. A recombinant DNA containing the BKV TAg gene and the activated c-Ha-ras oncogene (pBK/c-rasA) induced neoplastic transformation of early-passage hamster embryo cells with greater efficiency than each of the two genes transfected independently, suggesting a synergic effect between BKV TAg and c-Ha-ras. Moreover, BKV induces tumors in experimental animals: transgenic mice expressing BKV develop hepatocellular carcinoma and renal tumors (Small et al., 1986; Dalrymple and Beemon, 1990; Imperiale, 2000). Transformation of human cells by BKV is inefficient and often abortive (Shah et al., 1976; Portolani and Borgatti, 1978; Tognon et al., 2003). BKV-infected or transfected human cells generally do not display a completely transformed phenotype characterized by immortalization, anchorage independence, and tumorigenicity in nude mice, although they show morphological alterations and an increased lifespan (Purchio and Fareed, 1979; Grossi et al., 1982). A fully transformed phenotype was observed in human embryo kidney (HEK) cells transfected with a recombinant plasmid containing BKV early region and the adenovirus 12 E1A gene (Vasavada et al., 1986). These cells grow as a continuous cell line suggesting that in human cells BKV T-Ag contributes only a partially transformed phenotype and must interact with other oncogene functions to induce a complete transformation. The recombinants pBK/c-rasA and pBK/c-myc induced morphologic transformation of human embryo fibroblasts and kidney cells, but transformed cells were not immortalized (Pater and Pater, 1986). Tumorigenic cell lines were established only from transformed human fetal brain cells persistently infected by BKV. Fetal brain cells had all the characteristics of transformed cells and retained viral DNA in an episomal state but they were negative for TAg expression (Takemoto et al., 1979).

Evidence supporting a possible role for BKV in human cancer has accumulated slowly in recent years. The advent of PCR technology allowed investigators to examine even small biopsy samples for the presence of viral sequences. BKV DNA sequences have been reported in a range of human tumors, many of which do not fit the normal spectrum of anatomical sites thought to be infected by the virus, including rhabdomyosarcoma, lung, Kaposi's sarcoma, pancreas, liver, brain, and various urinary tract neoplasms (Fiori and Di Mayorca, 1976; Pater et al., 1980; Caputo et al., 1983; Barbanti-Brodano et al., 1987; Corallini et al., 1987; Dorries et al., 1987; Negrini et al., 1990; De Mattei et al., 1995; Monini et al., 1995a,b; Flaegstad et al., 1999; Imperiale, 2000, 2001; Fioriti et al., 2003). Interestingly, other reports found no evidence for BKV in brain tumors, urothelial carcinoma of the bladder, and of the renal pelvis, medulloblastomas, meningiomas, and ependymomas (Arthur et al., 1994; Weggen et al., 2000; Knoll et al., 2003; Fischetti et al., 2004).


BKV is known to commonly establish persistent infections in people and to be excreted in the urine by individuals who are asymptomatic, complicating the evaluation of its potential role in development of human cancer. Future investigations of the molecular biology will be needed for the demonstration of significant molecular changes in the lifestyle of the virus between normal and abnormal cells. Therefore, there is no certain proof that human polyomavirus BK directly causes the cancer in humans or acts as a cofactor in the pathogenesis of some type of human cancer.


Grant sponsor: National programme on AIDS research and MIUR (Italy).