Merkel cell carcinoma: histopathologic and prognostic features according to the immunohistochemical expression of Merkel cell polyomavirus large T antigen correlated with viral load.

Merkel cell carcinoma (MCC) is a neuroendocrine skin malignancy frequently associated with Merkel cell polyomavirus (MCPyV), which is suspected to be oncogenic. In a series of MCC patients, we compared clinical, histopathologic, and prognostic features according to the expression of viral large T antigen (LTA) correlated with viral load. We evaluated the LTA expression by immunohistochemistry using CM2B4 antibody and quantified viral load by real-time polymerase chain reaction. We analyzed formalin-fixed, paraffin-embedded (FFPE) tissue samples (n = 36) and corresponding fresh-frozen biopsies when available (n = 12), of the primary tumor and/or metastasis from 24 patients. MCPyV was detected in 88% and 58% of MCC patients by real-time polymerase chain reaction and immunohistochemistry, respectively. The relevance of viral load measurements was demonstrated by the strong consistency of viral load level between FFPE and corresponding frozen tissues as well as between primary tumor and metastases. From FFPE samples, 2 MCC subgroups were distinguished based on a viral load threshold defined by the positivity of CM2B4 immunostaining. In the LTA-negative subgroup with no or low viral load (nonsignificant), tumor cells showed more anisokaryosis (P = .01), and a solar elastosis around the tumor was more frequently observed (P = .03). LTA-positive MCCs with significant viral load had a lower proliferation index (P = .03) and a longer survival of corresponding patients (P = .008). Depending on MCPyV involvement, 2 MCC subgroups can be distinguished on histopathologic criteria, and the CM2B4 antibody is able to differentiate them reliably. Furthermore, the presence of a significant viral load in tumors is predictive of better prognosis.

Polyomavirus (BK)-associated pleomorphic giant cell carcinoma of the urinary bladder: a case report.

This report describes the morphological features of a pleomorphic giant cell carcinoma with focal trophoblastic differentiation of the urinary bladder in a male, 12 years post living related donor renal transplant. The voided urine cytology demonstrated rare decoy cells admixed with markedly atypical urothelial cell clusters, papillae and giant cells. Cystoprostatectomy demonstrated a nodular mass involving the trigone and right lateral-posterior wall, adjacent to the ureteral orifice. Hematoxylin-eosin stained sections showed two synchronous malignancies: (a) pleomorphic giant cell carcinoma with focal trophoblastic differentiation of the urinary bladder, metastatic to the omentum and (b) prostatic adenocarcinoma, Gleason score 3+4=7, involving the right prostate lobe. Strong diffuse expression of polyomavirus large T antigen was demonstrated in the primary and metastatic pleomorphic giant cell carcinoma, supporting a possible role for polyomavirus (BK) in the oncogenetic pathway. The prostatic adenocarcinoma was negative for polyomavirus large T antigen. Our findings of p63, CK7 and CK903 expression in pleomorphic giant cell carcinoma suggest that the tumor is of urothelial derivation. This is the first report describing the morphological features of urinary bladder pleomorphic giant cell carcinoma with trophoblastic differentiation, positive for polyomavirus large T antigen, arising in the background of BKV reactivation.

Homo- and heterotypic cell-cell contacts in Merkel cells and Merkel cell carcinomas: heterogeneity and indications for cadherin switching.

AIMS: Merkel cell carcinomas (MCCs) are rare but aggressive tumours associated recently with Merkel cell polyomavirus (MCV). As development and progression of several types of carcinomas can be promoted by changes in cell adhesion proteins, the aim of this study was to examine homo- and heterotypic cell contacts of Merkel cells and MCCs. METHODS AND RESULTS: Merkel cells of healthy glabrous epidermis and 52 MCCs were analysed by double-label immunostaining, immunofluorescence and confocal microscopy. Merkel cells were connected to keratinocytes by E- and P-cadherin, desmoglein 2 and desmocollin 2. In contrast, the vast majority of MCCs (90%) contained N-cadherin, but only 67% and 65% contained E- and P-cadherin, respectively. Interestingly, P-cadherin was absent significantly more frequently in lymph node metastases than in primary tumours and by trend in more advanced clinical stages. Moreover, major subsets of MCCs synthesized desmoglein 2 and, surprisingly, tight junction proteins. No significant differences were observed upon stratification for MCV DNA, detected in 84% of tumours by real-time polymerase chain reaction. CONCLUSIONS: Assuming that MCCs originate from Merkel cells, our data indicate a switch from E- and P-cadherin to N-cadherin during tumorigenesis. Whether the unexpected heterogeneity of junctional proteins can be exploited for prognostic and therapeutic purposes will need to be examined.

Detection of polyomavirus SV40 in tonsils from immunocompetent children.

BACKGROUND: BK virus (BKV), JC virus (JCV) and simian virus 40 (SV40) are nonenveloped DNA viruses, members of the family Polyomaviridae. BK and JC viruses establish persistent infections in humans, and evidence suggests that SV40 can infect humans, as well. Whether persistence occurs in the lymphoid system is unknown. METHODS: Paraffin-embedded tonsils from 220 immunocompetent children (mean age 9.3 years) were examined by polymerase chain reaction (PCR) to detect viral DNA of BKV, JCV, SV40, and Epstein-Barr virus (EBV). RESULTS: Polyomavirus-specific DNA sequences were detected in 8.3% (29/351) of specimens collected from 220 children. Twenty-one (9.5%) children had polyomavirus DNA present in at least one tonsil, with sequences identified as SV40 (n=20) and BKV (n=1). Polyomavirus JCV was not detected. Among patients positive for SV40, 8 of 14 (57%) contained viral DNA in both available tonsils. EBV DNA was detected in 99 (28.2%) samples from 67 (30.5%) patients. Eleven samples (3.1%) from 8 (3.6%) children were positive for both polyomavirus and EBV. SV40-positive children were significantly older than the SV40-negative subjects (P<0.001). T-antigen expression was detected in an SV40 DNA-positive tonsil by immunohistochemistry. CONCLUSIONS: These results suggest that SV40 can infect tonsils, that lymphoid tissue may represent a site for polyomavirus persistence, and that immunohistochemistry is not a useful detection assay when there are very few virus-infected cells in a tissue.

Is there a role for SV40 in human cancer?

The question of whether Simian Virus 40 (SV40) can cause human tumors has been one of the most highly controversial topics in cancer research during the last 50 years. The longstanding debate began with the discovery of SV40 as a contaminant in poliovirus vaccine stocks that were used to inoculate approximately 100 million children and adults in the United States between 1955 and 1963, and countless more throughout the world. Concerns regarding the potential health risk of SV40 exposure were reinforced by studies demonstrating SV40's potential to transform human cells and promote tumor growth in animal models. Many studies have attempted to assess the relationship between the potential exposure of humans to SV40 and cancer incidence. Reports of the detection of SV40 DNA in a variety of cancers have raised serious concerns as to whether the inadvertent inoculation with SV40 has led to the development of cancer in humans. However, inconsistent reports linking SV40 with various tumor types has led to conflicting views regarding the potential of SV40 as a human cancer virus. Several recent studies suggest that older detection methodologies were flawed, and the limitations of these methods could account for most, if not all, of the positive correlations of SV40 in human tumors to date. Although many people may have been exposed to SV40 by polio vaccination, there is inadequate evidence to support widespread SV40 infection in the population, increased tumor incidence in those individuals who received contaminated vaccine, or a direct role for SV40 in human cancer.

Purified JC virus T and T' proteins differentially interact with the retinoblastoma family of tumor suppressor proteins.

The amino termini of polyomavirus T antigens contain LXCXE and J domains, which are necessary for binding and inactivating the retinoblastoma family of tumor suppressors. Both of these motifs are found in the JC virus (JCV) early proteins T'(135), T'(136), and T'(165), leading to the suggestion that these recently discovered proteins complement the cell-cycle-deregulating function of the JCV large T antigen (TAg). To investigate this hypothesis, the three JCV T' proteins were produced in a baculovirus expression system and purified by immunoaffinity chromatography. To facilitate purification, hybridomas that secrete antibodies recognizing amino-terminal epitopes of JCV early proteins were produced. Potential interactions between the early viral proteins and the cellular proteins pRB, p107, and p130 were investigated by incubating purified JCV TAg and T' proteins with extracts of MOLT-4 cells, a human T cell line. The four viral proteins preferentially bound hypophosphorylated species of the cellular proteins and exhibited the highest binding affinity to p107 and the lowest affinity to pRB. TAg and T'(165) bound more pRB and less p107 than did T'(135) and T'(136); T'(165) also bound less p130 than the other three early proteins. Results of these in vitro interactions were compared to those obtained in vivo using POJ cells, a transformed human glial cell line that expresses JCV early proteins, relatively high levels of pRB and p107, and low levels of p130. Most of the pRB in POJ cells is hyperphosphorylated, and only a fraction of the hypophosphorylated form(s) of pRB is bound by the viral proteins. In contrast, only hypophosphorylated p130 is detected in the transformed cells, and most of this protein was found in complex with the viral proteins. Finally, nearly all of the p107 in POJ cells is bound by the JCV proteins.

Enhanced binding to origin DNA at low pH enables easy detection of polyomavirus large T antigen by gel mobility shift assay of unfixed complexes.

Enhanced, stable binding by polyomavirus large T antigen to the viral DNA replication origin at pH 6 allowed the development of a gel mobility shift assay for the detection of large T antigen. Such assays were not possible at pH 7.6 without previous fixation, due to instability of the complexes. We demonstrated that the gel mobility shift assay at pH 6 is very sensitive, allowing the detection of as little as 5 ng large T antigen, and is highly specific for DNA containing G(A/G)GGC target sequences. This method was used to detect large T antigen in crude cell lysates from transformed yeast cell lines or nuclear extracts from infected insect cells. Large T antigen-DNA complexes remained at or near the loading well in 5% acrylamide or 1.5% agarose gels, indicating that these complexes are very large. Glycerol gradient analysis showed that protein-DNA complexes formed at pH 6 were massive, and that large T antigen also formed large complexes when incubated at low pH in the absence of DNA. These results show that pH has a major effect on binding of large T antigen to its multiple target sites in the viral origin of DNA replication, presumably by affecting protein-protein interactions that are important for the stability of large T antigen-DNA complexes.

Interaction between the N-terminus of human topoisomerase I and SV40 large T antigen.

We have attempted to identify human topoisomerase I-binding proteins in order to gain information regarding the cellular roles of this protein and the cytotoxic mechanisms of the anticancer drug camptothecin, which specifically targets topoisomerase I. In the course of this work we identified an interaction between the N-terminus of human topoisomerase I and the SV40 T antigen that is detectable in vitro using both affinity chromatography and co-immunoprecipitation. Additional results indicate that this interaction does not require intermediary DNA or stoichiometric quantities of other proteins. Furthermore, the interaction is detectable in vivo using a yeast two-hybrid assay. Two binding sites for T antigen are apparent on the topoisomerase I protein: one consisting of amino acids 1-139, the other present in the 383-765 region of the protein. Interestingly, nucleolin, which binds the 166-210 region of topoisomerase I, is able to bind an N-terminal fragment of topoisomerase I concurrently with T antigen. Taken together with our prior identification of nucleolin as a topoisomerase I-binding protein, the current results suggest that helicase-binding is a major role of the N-terminus of human topoisomerase I and that the resultant helicase-topoisomerase complex may function as a eukaryotic gyrase.

SV40 large T antigen binds to the TBP-TAF(I) complex SL1 and coactivates ribosomal RNA transcription.

SV40 large T antigen is a multifunctional regulatory protein that plays a key role in the viral life cycle and can stimulate cell proliferation. To accomplish this, large T antigen has to control the expression of cellular genes involved in cell cycle progression and cell growth. rRNA synthesis by RNA polymerase I (Pol I) is tightly associated with cell growth and proliferation, and previous studies indicated that large T antigen up-regulates RNA Pol I transcription in SV40-infected cells. How this process occurs is currently unclear. To investigate the mechanisms of large T antigen stimulation of RNA Pol I transcription, we have established an in vitro transcription system that is responsive to large T antigen. Here, we show that recombinant large T antigen stimulates Pol I transcription reconstituted with purified RNA Pol I, UBF, and the TBP/TAF complex SL1. Immunoprecipitation experiments revealed that large T antigen directly binds to SL1, in vitro, as well as in SV40-infected cells. In addition, our data indicate that this interaction occurs by direct association with three SL1 subunits, namely TBP, TAF(I)48, and TAF(I)110. Transcription studies with large T antigen deletion mutants show that the 538-amino-acid amino-terminal domain is necessary for full stimulation of Pol I transcription. Importantly, mutants that no longer bind to SL1 are also unable to stimulate Pol I transcription. This indicates that recruitment of large T antigen to the rRNA promoter by SL1 constitutes a crucial step in the activation process. Taken together with recent studies on large T antigen activation of RNA Pol II transcription, these results suggest that viral modulation of genes involved in cell proliferation involves direct targeting of promoter-specific TBP/TAF complexes (i.e., SL1 or TFIID) by large T antigen.

Production of active polyomavirus large T antigen in yeast Pichia pastoris.

The coding region of polyomavirus large T antigen was engineered into the genome of the methylotrophic yeast Pichia pastoris by use of the vector pHIL-D2. Expression of large T antigen was induced by methanol under the control of the strong alcohol oxidase (AOX1) promoter. Large T antigen was purified by immunoaffinity chromatography. We showed that yeast-derived large T antigen bound specifically to a DNA fragment that contains the polyomavirus replication origin, protected the four known major binding sites in the origin against DNase I digestion, and could unwind the strands of an origin-containing DNA fragment in an ATP-dependent manner. This system therefore provides a convenient and inexpensive source of biologically active polyomavirus large T antigen for in vitro studies.

Association of p300 and CBP with simian virus 40 large T antigen.

p300 and the CREB-binding protein CBP are two large nuclear phosphoproteins that are structurally highly related. Both function, in part, as transcriptional adapters and are targeted by the adenovirus E1A oncoprotein. We show here that p300 and CBP interact with another transforming protein, the simian virus 40 large T antigen (T). This interaction depends on the integrity of a region of T which is critical for its transforming and mitogenic properties and includes its LXCXE Rb-binding motif. T interferes with normal p300 and CBP function on at least two different levels. The presence of T alters the phosphorylation states of both proteins and inhibits their transcriptional activities on certain promoters. Although E1A and T show little sequence similarity, they interact with the same domain of p300 and CBP, suggesting that this region exhibits considerable flexibility in accommodating diverse protein ligands.

Simian virus 40 large T antigen alters the phosphorylation state of the RB-related proteins p130 and p107.

p130 and p107 are nuclear phosphoproteins related to the retinoblastoma gene product (pRb). pRb, p107, and p130 each undergo cell cycle-dependent phosphorylation, form complexes with the E2F family of transcription factors, and associate with oncoproteins of DNA tumor viruses, including simian virus 40 (SV40) large T antigen (TAg) and adenovirus ElA protein. The results of recent studies with mouse embryo fibroblasts (MEFs) lacking the retinoblastoma gene (Rb-1) have suggested that p130 and p107 may be important targets for SV40 large TAg-mediated transformation (J.B. Christensen and M.J. Imperiale, J. Virol. 65:3945-3948, 1995; J. Zalvide and J.A. DeCaprio, Mol. Cell. Biol. 15:5800-5810, 1995). In this report, we demonstrate that the expression of TAg affects the phosphorylation state of p130 and p107. In cells expressing wild-type TAg, only un(der)phosphorylated p130 and p107 were detected. To determine the domains within TAg that contribute to this effect on the phosphorylation of p130, we performed transient expression assays. While transiently expressed p130 was apparently phosphorylated normally, only un(der)phosphorylated p130 was detected when p130 was coexpressed with TAg. Using this assay, we found that the first 147 amino acids of TAg were sufficient to alter the phosphorylation state of p130. Within this region, the LXCXE domain of TAg, required for binding to the retinoblastoma family of proteins, was necessary but not sufficient to affect p130 phosphorylation. Residues within the first 82 amino acids of TAg were also required. TAg with mutations in the N terminus retained the ability to efficiently associate with p130 but did not affect its phosphorylation state. This demonstrates that the effect of SV40 TAg on p130 is not simply the result of binding and suggests that TAg has a novel effect on p130 and p107 that differs from its effect on pRb.

Modification of an N-terminal regulatory domain of T antigen restores p53-T antigen complex formation in the absence of an essential metal ion cofactor.

We have discovered that the ability of the tumor suppressor protein p53 to bind to the viral large T antigen (TAg) oncogene product is regulated by divalent cations. Both proteins were purified from an insect cell line infected with the appropriate baculovirus expression vector. In a two-site capture enzyme-linked immunosorbent assay, complex formation between the purified proteins is strictly dependent on the addition of specific concentrations of divalent metal ions, notably zinc, copper, cadmium, cobalt, manganese, and nickel. In the presence of zinc the pattern of proteolytic fragments obtained when TAg was subjected to proteolysis by endoproteinase Glu-C (V8) was strikingly different, supporting the idea that a conformational change in TAg associated with ion binding is required for it to complex with p53. Monoclonal antibody analysis provides supporting evidence for a conformational change. When TAg was captured onto an enzyme-linked immunosorbent assay plate coated with PAb 419 as opposed to many other anti-TAg antibodies, complex formation was completely independent of the presence of additional divalent cations. Our results suggest that the ability of p53 and TAg to form a stable complex in vitro is dependent upon a regulatory domain residing in the N terminus of TAg, zinc ions or the binding of a specific monoclonal antibody (PAb 419) provoking a conformational change in TAg that facilitates and supports complex formation.

The human breast cell DNA synthesome: its purification from tumor tissue and cell culture.

In this report, we describe for the first time the isolation and purification of a multiprotein complex for DNA replication from MDA MB-468 human breast cancer cells. This complex, which we designate the DNA synthesome, fully supports the in vitro replication of simian virus 40 (SV40) origin-containing DNA in the presence of the viral large T-antigen. Since the SV40 virus utilizes the host's cellular proteins for its own DNA replication, our results indicate that the DNA synthesome may play a role not only in viral DNA synthesis but in human breast cell DNA replication as well. Our studies demonstrate that the following DNA replication proteins constitute the DNA synthesome: DNA polymerase alpha, DNA primase, DNA polymerase delta, proliferating cell nuclear antigen, replication protein A, replication factor C, DNA topoisomerases I, II, and DNA polymerase epsilon. In addition, we successfully isolated the DNA synthesome from human breast tumor tissue as well as from xenografts from nude mice injected with the human breast cancer cell line MCF-7. The DNA synthesome purified from the breast cancer tissues fully supports SV40 DNA replication in vitro. Furthermore, our results obtained from a novel forward mutagenesis assay suggest that the DNA synthesome isolated from a nonmalignant breast cell line mediates SV40 DNA replication by an error-resistant mechanism. In contrast, the DNA synthesome derived from malignant breast cells and tissue exhibited a lower fidelity for DNA synthesis in vitro. Overall, our data support the role of the DNA synthesome as mediating breast cell DNA replication in vitro and in vivo.

Cooperation of simian virus 40 large and small T antigens in metabolic stabilization of tumor suppressor p53 during cellular transformation.

Metabolic stabilization of the tumor suppressor p53 is a key event in cellular transformation by simian virus 40 (SV40). Expression of the SV40 large tumor antigen (large T) is necessary but not sufficient for this process, as metabolic stabilization of p53 complexed to large T in abortively SV40-infected cells strictly depends on the cellular systems analyzed (F. Tiemann and W. Deppert, J. Virol. 68:2869-2878, 1994). Comparative analyses of various cells differing in metabolic stabilization of p53 upon abortive infection with SV40 revealed that metabolic stabilization of p53 closely correlated with expression of the SV40 small t antigen (small t) in these cells: 3T3 cells do not express small t and do not stabilize p53 upon infection with wild-type SV40. However, ectopic expression of small t in 3T3 cells provided these cells with the capacity to stabilize p53 upon SV40 infection. Conversely, precrisis mouse embryo cells express small t and mediate metabolic stabilization of p53 upon infection with wild-type SV40. Infection of these cells with an SV40 small-t deletion mutant did not lead to metabolic stabilization of p53. Small-t expression and metabolic stabilization of p53 correlated with an enhanced transformation efficiency by SV40, supporting the conclusion that at least part of the documented helper effect of small t in SV40 transformation is its ability to promote metabolic stabilization of p53 complexed to large T.

T antigens encoded by replication-defective simian virus 40 mutants dl1135 and 5080.

We present a preliminary biochemical characterization of two simian virus 40 mutants that affect different T antigen replication functions. SV40 T antigen mutants dl1135 (delta 17-27 amino acids) and 5080 (P-L) have been studied extensively with regard to their ability to transform cells in culture and induce tumors in transgenic mice. Both mutants are defective for viral DNA replication in vivo. In order to assess in more detail the molecular basis for the in vivo replication defects of 5080 and dl1135, we expressed the mutant proteins using the baculovirus system and purified them by immunoaffinity chromatography. With each of the purified proteins, we examined some of the biochemical activities of T antigen required for replication, viz. ATPase, binding to the origin of replication (ori) and assembly on ori, DNA helicase and unwinding, and replication in in vitro assays. Consistent with previous studies, we found that the 5080 protein is defective for multiple biochemical activities including ATPase, helicase, ori-specific unwinding, and ATP-induced hexamerization. However, this mutant retains some sequence-specific DNA binding activity. In contrast, the dl1135 protein exhibited significant levels of activity in all assays, including the ability to drive SV40 DNA replication in vitro. Thus, dl1135 is one of several mutants with an altered amino-terminal domain which can replicate DNA in vitro, but not in vivo. Thus, while the 5080 mutation affects a T antigen enzymatic function directly required for viral DNA synthesis, dl1135 may alter an activity required to prepare the cell for viral replication.

Cell fractionation in non-ionic detergent distinguishes sub-populations of polyoma virus middle T antigen and reveals a novel form.

The transforming function of polyoma virus, middle T antigen (MT), interacts with several cellular enzymes, essential to its oncogenic activity. We have used cell fractionation to study the various MT/cellular protein complexes. We demonstrate that MT can be separated into three sub-species, dependent upon extraction in two buffers that we designate A and B: Antigen extracted from whole cells by both buffers (called MT1) is associated with most of the phosphorylated phosphatidyl-inositol kinase 85 kD subunit, pp85, and protein phosphatase 2A. Antigen (MT2), associated with the greater portion of pp60c-src, is extracted by buffer B, but not buffer A. A third population (MT3), resistant to extraction by either buffer, is not detectably associated with protein phosphatase 2A or pp85. It is, however, associated with a low level kinase activity. The interaction between pp60c-src and MT appears to influence the formation of both MT2 and MT3. MT2 fractionates with the cellular microtubule network, but does not appear to be directly associated with it. MT3, a previously undescribed population, comprises about one third of MT in wild type antigen-containing cells. It is missing in mutants incapable of interacting with pp60c-src, but exists in the absence of an interaction with pp85. We suggest that MT3 may be an intermediate in, or product of, one of the MT/pp60c-src signalling pathways, distinct from that involving pp85.