The Collagen Repeat Sequence Is a Determinant of the Degree of Herpesvirus Saimiri STP Transforming Activity

Abstract

Herpesvirus saimiri (HVS) is divided into three subgroups, A, B, and C, based on sequence divergence at the left end of genomic DNA in which the saimiri transforming protein (STP) resides. Subgroup A and C strains transform primary common marmoset lymphocytes to interleukin-2-independent growth, whereas subgroup B strains do not. To investigate the nononcogenic phenotype of the subgroup B viruses, STP genes from seven subgroup B virus isolates were cloned and sequenced. Consistent with the lack of oncogenic activity of HVS subgroup B viruses, STP-B was deficient for transforming activity in rodent fibroblast cells. Sequence comparison reveals that STP-B lacks the signal-transducing modules found in STP proteins of the other subgroups, collagen repeats and an authentic SH2 binding motif. Substitution mutations demonstrated that the lack of collagen repeats but not an SH2 binding motif contributed to the nontransforming phenotype of STP-B. Introduction of the collagen repeat sequence induced oligomerization of STP-B, resulting in activation of NF-κB activity and deregulation of cell growth control. These results demonstrate that the collagen repeat sequence is a determinant of the degree of HVS STP transforming activity.

Herpesvirus saimiri (HVS) is the prototypic and best-characterized gamma-2-herpesvirus (rhadinovirus) (26). The only known human gamma-2-herpesvirus, human herpesvirus 8 or Kaposi's sarcoma-associated herpesvirus (KSHV), is highly homologous with HVS and has a similar genomic organization (45, 48). In addition, several herpesviruses isolated from rhesus monkeys, called rhesus rhadinovirus (1, 13, 50) and retroperitoneal fibromatosis herpesvirus (46), are also highly similar to KSHV and HVS. HVS infects most squirrel monkeys without apparent disease (16). In other nonhuman primates, however, HVS induces rapidly progressing fatal T-cell lymphoproliferative diseases (17, 26). Sequence divergence among HVS isolates is most extensive at the left end of the unique L-DNA of the viral genome and is the basis for classification of HVS into subgroups A, B, and C (5, 12, 39). Variation in this region is correlated with differences in the capacity of these viruses to immortalize T lymphocytes in vitro and to produce lymphoma in nonhuman primates (4, 12, 14, 32). Both subgroup A and C viruses immortalize common marmoset T lymphocytes to interleukin-2 (IL-2)-independent proliferation (14, 53). However, none of the subgroup B viruses tested were capable of immortalizing common marmoset T lymphocytes (53). Furthermore, highly oncogenic subgroup C strains immortalize human, rabbit, and rhesus monkey lymphocytes and can produce fulminant lymphoma in rhesus monkeys as well as in rabbits (2, 4, 7, 17, 38, 42).

HVS subgroup A strain 11 mutants with deletions in the first open reading frame at the left end of the genome are capable of replication but fail to immortalize common marmoset T lymphocytes in vitro and to induce lymphoma in vivo (12, 14, 44). This open reading frame is designated the saimiri transforming protein (STP) of HVS subgroup A (STP-A) (44). HVS subgroup C contains a divergent form of the STP gene (STP-C) along with an additional, apparently unrelated open reading frame, called Tip, in the leftmost position (5, 19). Both STP-C and STP-A are sufficient to transform rodent fibroblast cells in culture, but STP-C is considerably more potent (30). Similarities between STP-A11 and STP-C488 include highly acidic amino termini, the presence of collagen repeats in the central parts of the proteins, and hydrophobic membrane anchoring regions at the carboxyl termini (30). STP-C has 18 direct repeats of a collagen motif (Gly-Pro-Pro or Gly-Pro-Gln) that comprise more than 50% of the protein and are predicted to have triple α-helical structure (5, 19). A mutation that disrupts the collagen repeats has been shown to disrupt the transforming activity of STP-C488 (28).

STP-C is the only virus-encoded protein, to our knowledge, that has been found to associate with cellular Ras in oncogenic transformation (27). Interruption of the association between STP and ras interferes with the transforming activity of STP-C488 in culture (27). STP-A contains a highly conserved YAEV/I motif at amino acid residues 115 to 118 preceded by negatively charged glutamic acid residues, which matches very well with the consensus sequence for binding to SH2 domains of Src family kinases (36). Indeed, STP-A associates with cellular Src and is an in vitro substrate for Src kinase through its YAEV/I motif. Furthermore, the STPs of subgroups A and C are found to be stably associated with tumor necrosis factor (TNF) receptor-associated factors (TRAFs) (35). Mutational analyses demonstrate that the PXQ/EXT/S residues in STP are critical for TRAF association and that an interaction of STP-C with TRAFs contributes to the transformation of human lymphocytes and rodent fibroblasts (35).

Subgroup A and C strains immortalize common marmoset lymphocytes to IL-2-independent growth, but none of the subgroup B strains tested score positive in this assay (14). We hypothesized that differences in transforming activity in the three HVS subgroups are primarily due to the differences in oncogenic activity of the STP gene. To test this hypothesis, we have cloned and sequenced the STP gene from seven subgroup B virus isolates. Sequence comparisons reveal that STP-B is much closer to STP-A than to STP-C. Consistent with the lack of transforming ability of the subgroup B viruses, STP-B was deficient in the ability to transform rodent fibroblasts. Mutational analysis demonstrated that the lack of transforming activity of STP-B was due to the absence of collagen repeat sequences and that the introduction of collagen repeat sequences induced oligomerization of STP-B, NF-κB activation, and cell growth transformation. These results suggest that a membrane-associated, oligomerized STP may mimic a ligand-independent, constitutively active receptor, generating continuous signals for abnormal cell growth.

MATERIALS AND METHODS

Cell culture.

OMK cells were grown in minimum essential medium (MEM) supplemented with 10% fetal calf serum. 293T cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum. An electroporation procedure at 250 V and 960 μF in serum-free DMEM was used for transient expression in 293T cells. pBabe-STP-B was introduced into Rat-1 cells using a fusion transfection procedure (Boehringer Mannheim), and transfected cells were cultured with selection medium containing puromycin (5 μg/ml) for the next 5 weeks.

DNA sequence and plasmid constructions.

The leftmost 1.2-kb fragment of HVS strain SMHI, kindly provided by Peter Medvezky (39), was sequenced on both strands using an ABI PRISM 377 automatic DNA sequencer. DNA containing the STP-B open reading frame was amplified from pSBH1.2 by PCR using primers containing EcoRI and XbaI recognition sequences at the ends. Amplified DNA was ligated into the EcoRI and XbaI cloning sites of the pFJ vector for transient expression. For AU-1 tagging, the 5′ primer CGC GGA TCC ATG GAC ACC TAT CGC TAT ATA GCA AGA GGT CTA GGT GAA GGA was used for PCR amplification (29). Amino-terminal AU-1-tagged STP-B DNAs were completely sequenced to verify 100% agreement with the original sequence. To generate the pBabe-puro expression vector containing STP-B, the EcoRI-XhoI fragment containing STP-B was subcloned into the EcoRI and SalI sites of pBabe-puro. The chicken src gene was subcloned into vector pFJ for expression (36).

All mutations in STP-B were generated by PCR using oligonucleotide-directed mutagenesis (15). To facilitate mutagenesis, the STP-B gene was subcloned into the pSP72 vector (Promega Biotech, Madison, Wis.). PCR cycling for mutagenesis was accomplished with a DNA thermal cycler (Perkin-Elmer Cetus Instruments, Norwalk, Conn.) with the following conditions: 30 cycles of 2 min at 50°C for annealing, 5 min at 72°C for polymerization, and 1 min at 94°C for denaturation. Each STP-B mutant was completely sequenced to verify the presence of the mutation and the absence of any other changes. After confirmation of the sequence, DNA containing the desired STP-B mutation was recloned into the EcoRI and XbaI cloning sites of vector pFJ for gene expression.

Virus isolation and molecular cloning of STP genes.

To isolate virus, owl monkey kidney cells (OMK 637) were cocultivated with purified peripheral blood mononuclear cells from squirrel monkeys (16, 17). Coculture was maintained for 2 to 3 weeks until the appearance of cytopathic changes. Virus pellets were obtained by centrifugation of 5 ml of cell-free supernatant from infected cell cultures. Virions were suspended in 0.1 ml of 50 mM Tris hydrochloride (pH 7.5)–10 mM EDTA–50 mM NaCl; proteinase K and sodium dodecyl sulfate (SDS) were added to final concentrations of 1 mg/ml and 1%, respectively. After overnight incubation at 65°C, the viral DNA was extracted once with buffer-saturated phenol and once with chloroform-isoamyl alcohol. Virion DNA was precipitated with 2.5 volume of ethanol, pelleted by centrifugation, and suspended in Tris-EDTA (TE) buffer.

Purified virion DNA from six different strains of HVS subgroup B was used for PCR amplification using the 5′ primer CGG AAT TCA TGG CAA GAG GTC TAG GTG A, which corresponds to the amino-terminal sequence of STP-B-SMH1, and 3′ primer CGC CTC GAG TAA TTA CTA GCA TTA AAC C, which corresponds to the carboxyl-terminal sequence of STP-B-SMHI. Primers used for PCR contained EcoRI and XbaI sites (underlined), which were used for subsequent cloning. PCR-amplified DNA was digested with EcoRI and XbaI restriction enzymes and subcloned into the pSP72 vector. Independent clones were subsequently sequenced on both strands using an ABI PRISM 377 automatic DNA sequencer.

Immunoprecipitation and immunoblot.

Cells were harvested and lysed with lysis buffer (0.15 M NaCl, 1% Nonidet P-40, 50 mM HEPES buffer [pH 8.0]) or radioimmunoprecipitation assay buffer (0.15 M NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris [pH 7.5]) containing 0.1 mM Na₂VO₃, 1 mM NaF, and protease inhibitors (leupeptin, aprotinin, phenylmethylsulfonyl fluoride, and bestatin). Immunoprecipitated proteins from cleared cell lysates were separated by SDS-polyacrylamide gel electrophoresis (PAGE) and reacted in immunoblot assays. For protein immunoblots, polypeptides in cell lysates corresponding to 10⁵ cells were resolved by SDS-PAGE and transferred to a nitrocellulose membrane filter. Immunoblot detection was performed with a 1:1,000 dilution of primary antibody with an ECL kit (Amersham).

In vitro kinase assays.

For in vitro protein kinase assays, complexes prepared as described above were washed once more with kinase buffer and resuspended with 10 μl of the same buffer containing 5 μCi of [γ-³²P]ATP (6,000 Ci/mmol; NEN) for 15 min at room temperature.

Assays for growth properties.

For focus formation, 10⁶ cells were plated in 100-mm tissue culture dishes and maintained with DMEM plus 10% serum, changed every 4 days. At day 14, cells were photographed.

Reporter assays.

All transfections included 5 μg of pGKβgal, which expresses β-galactosidase from a phosphoglucokinase promoter, and 5 μg of 3X-kB-luc, which has three copies of the NF-κB binding site from the murine major histocompatibility complex class I promoter upstream of a minimal fos promoter and a luciferase gene (35). At 48 h posttransfection, cells were washed once in phosphate-buffered saline and lysed in 200 μl of reporter lysis buffer (Promega). Assays for luciferase were performed with a Luminometer using a luciferase assay (Promega). Values were normalized for β-galactosidase activity.

RESULTS

Amino acid sequence analysis of STPs from seven different strains of HVS subgroup B.

DNA sequence analysis revealed an open reading frame from the 1.2-kb leftmost fragment of HVS subgroup B strain SMHI, which is located at a position equivalent to STPs of HVS subgroups A and C (39). This open reading frame, referred to as STP-B-SMHI, STP of subgroup B strain SMHI, was predicted to encode 171 amino acids (Fig. 1A). Primary amino acid sequences of STPs from six additional subgroup B strains (349-78, 423-79, 77-5B, S295C, 29-76, and 24-76) were determined by cloning and sequencing (Fig. 1A). Amino acid substitutions in STP-B from different strains were assessed by comparison with the STP-B-SMHI sequence (Fig. 1A). The STP-Bs from strains SMHI, 349-78, 423-79, 77-5B, and S295C were well conserved and showed 99% identity (Fig. 1A). In contrast, the STP-Bs from strains 29-76 and 24-76 showed significant divergence from those of the other five strains; they showed only 76.5% identity to the STP-B of the other five strains (Fig. 1A).

FIG. 1.

(A) Amino acid sequence and structural motifs of STP-B isolates. The amino acid sequences of seven STP-B clones were aligned to demonstrate similarities. TRAF-B indicates the putative TRAF-binding motif, SH2-B indicates the putative SH2-binding motif, and the grey box at the carboxyl terminus indicates the potential membrane-anchoring region. Divergent amino acid sequences are indicated with bold letters. (B) Sequence comparison of STP-A and STP-B. The grey boxes indicate the TRAF-binding motif (PxQxT/S) and SH2-binding motif (EExxYAEI/V). The bars and dots indicate identity and similarity, respectively. (C) Alignment of the TRAF-binding motifs of STP-B-SMHI with those of herpesvirus papio (HVP) LMP1. The grey boxes indicate TRAF-binding motifs.

Based on primary amino acid sequence, the presence of potential structural motifs in STP-B was assessed by comparison with STP-A and STP-C. This assessment demonstrated that STP-B was much closer to STP-A than to STP-C (Fig. 1B). Five basic structural motifs, an acidic amino terminus, a TRAF-binding motif, an SH2-binding motif, a collagen repeat, and a hydrophobic carboxyl terminus were examined on the basis of previous analysis of STP-A and STP-C (30, 35, 36). The acidic amino terminus from amino acids 1 to 37 and the hydrophobic stretch at the carboxyl terminus from amino acids 145 to 166 appeared to be highly conserved in all seven strains that were examined. STP-B has PXQXT sequences similar to those in STP-A, STP-C, LMP1, CD30, CD40, and TANK (8, 18, 20, 35, 40, 49) (Fig. 1A and C). These proteins have been shown to employ this motif as a core sequence to interact with TRAFs. STP-Bs from strains SMHI, 349-78, 423-79, 77-5B, and S295C contain two PXQXT motifs, whereas STP-Bs from strains 29-76 and 24-76 have a single motif (Fig. 1A). Systematic searches for optimal SH2 domain sequences have found that the consensus sequence for Src family kinase SH2 binding is EExxYEEV/I (51, 52). STP-A has a highly conserved YAEV sequence preceded by two negatively charged glutamic acid residues (Fig. 1B), which is required for binding to the SH2 domain of Src (36). A potential SH2 binding motif for Src family kinases is observed at amino acid positions 118 to 121 (YAEI) of STP-B, that is also highly conserved in all strains examined (Fig. 1A). However, unlike STP-A, all STP-B isolates do not contain the negatively charged amino acids preceding the potential SH2 binding motif (Fig. 1B). Furthermore, STP-B appears to lack a collagen repeat sequence (Gly-X-Y, where X and/or Y is proline) (Fig. 1A). The primary amino acid sequence of STP-A11 has nine repeats of this motif, and STP-C488 has 18 direct repeats comprising more than 50% of the protein. These results indicate that STP-B is similar to but distinct from STP-A and STP-C.

Identification of STP of subgroup B strains SMHI and 29-76.

To examine STP-B expression, we selected STP-B sequences of strains SMHI and 29-76 to represent each of the distinct groups (Fig. 1A). STP-B-SMHI and STP-B29-76 genes were tagged with an AU-1 epitope at their amino termini, cloned into plasmid pFJ containing the SRα-0 promoter (54), and expressed in 293T cells. The AU-1 antibody reacted with a protein with an apparent molecular size of 26 kDa upon immunoblot analysis from 293T cells transfected with pFJ containing the AU1-tagged STP-B-SMHI or STP-B29-76 (Fig. 2). In contrast, these proteins were not detected in 293T cells transfected with an empty vector (Fig. 2).

FIG. 2.

Identification of STP-B protein. 293T cells were transfected with pFJ vector, pFJ-STP-B29-76 (29-76), or pFJ-STP-B-SMHI (SMHI). Cell lysates were fractionated by SDS-PAGE, transferred to nitrocellulose, and reacted with an anti-AU-1 antibody. Sizes are shown in kilodaltons in this and subsequent figures.

Lack of transforming ability of STPs of HVS subgroup B isolates.

Since STP-C488 and STP-A11 are sufficient to transform rodent fibroblast cells (30), we investigated the transforming potential of the STP-B genes. STP-B-SMHI was expressed in rodent fibroblast Rat-1 cells using the retroviral vector pBabe-puro. After selection with puromycin, expression of STP-B-SMHI in Rat-1 cells was detected by immunoblot assay with an anti-AU-1 antibody (data not shown). To investigate the consequence of STP-B-SMHI expression, the growth properties of Rat-STP-B-SMHI cells were compared with those of control cells. As shown in Fig. 3, the growth properties of the puromycin-resistant Rat-STP-B-SMHI cells did not differ significantly from those of Rat-babe cells. Rat-STP-B-SMHI cells grew in flat monolayers and formed less than 50 foci, as control Rat-babe cells did (Fig. 3). Expression of STP-B29-76 also did not induce transformation of Rat-1 cells (data not shown). Thus, consistent with the lack of transforming ability of HVS subgroup B virus, STP-B was unable to transform rodent fibroblasts.

FIG. 3.

Growth properties of Rat-1 cells expressing the STP-B gene. Puromycin-resistant cells were obtained after transfection with the retroviral vector containing STP-B-SMHI or its mutants. Puromycin-resistant cells were plated at 10⁶ cells per 100-mm tissue culture dish. After 14 days of incubation, cells were photographed to show focus formation. Vector, Rat-babe; C488, Rat-STP-C488; SMHI, Rat-STP-B-SMHI; SMHI/EE, Rat-STP-B-SMHI/EE; SMHI/Col, Rat-STP-B-SMHI/Col; SMHI/EE/Col, Rat-STP-B-SMHI/EE/Col. Magnification, ×100.

Collagen repeats are a determinant of the transforming activity of STP.

Amino acid sequence comparison reveals that STP of subgroup B lacks two signal-transducing elements compared to the STPs of HVS subgroups A and C: the negatively charged amino acids preceding the SH2 binding motif and the collagen repeats. To investigate whether the absence of these elements may contribute to the lack of transforming ability of STP-B, these elements were substituted into the STP-B-SMHI singly or in combination. The asparagine at 114 (N₁₁₄) and the serine at 115 (S₁₁₅) preceding the putative YAEI SH2 binding motif of STP-B-SMHI were replaced with two glutamic acids (E₁₁₄ and E₁₁₅), which mimics the SH2 binding motif of STP-A11, creating STP-B-SMHI/EE. A 162-bp DNA fragment encoding 18 repeats of the collagen motif from STP-C488 was introduced between amino acid residues 71 and 72 of STP-B-SMHI, creating STP-B-SMHI/Col. Finally, STP-B-SMHI/EE/Col, containing substitutions of both elements, was also created.

The STP-B-SMHI/EE, STP-B-SMHI/Col, and STP-B-SMHI/EE/Col genes were expressed in rodent fibroblast Rat-1 cells using the retroviral vector pBabe-puro. The growth properties of Rat-1 cells expressing mutant forms of STP-B-SMHI were compared with those of Rat-STP-B-SMHI cells. Rat-1 cells transformed by STP-C488 (30) were included as a positive control. Unlike wild-type STP-B-SMHI, STP-B-SMHI/Col and STP-B-SMHI/EE/Col mutants strongly transformed Rat-1 cells, resulting in focus formation (Fig. 3). Foci were recognizable even before cells reached confluence. The numbers of foci observed for Rat-STP-B-SMHI/Col and Rat-STP-B-SMHI/EE/Col cells were over 1,000 per 100-mm tissue culture dish, which is equivalent to that of Rat-STP-C488 (Fig. 3). In contrast, STP-B-SMHI/EE did not induce transformation of Rat-1 cells (Fig. 3), indicating that an introduction of the negative charged amino acids to the putative SH2 binding motif did not increase the transforming activity of STP-B. When STP-B29-76 mutants, including STP-B29-76/EE, STP-B29-76/Col, and STP-B29-76/EE/Col, were used for the transformation assay, essentially the same results were observed (data not shown). These results suggest that the collagen repeat sequence is a determinant of the transforming activity of STP.

Oligomerization of STP-B by collagen repeats.

Cellular collagens have been extensively characterized to form a triple α-helix structure (6). To investigate whether introduction of the collagen repeat sequences induced oligomerization of the STP-B protein, 293T cells were transfected with expression vectors containing STP-B-SMHI, STP-B-SMHI/EE, STP-B-SMHI/Col, STP-B29-76, or STP-B29-76/Col. Since heat treatment dissociates oligomerization of collagens (3), cell lysates containing STP-B mutants were subjected to SDS-PAGE with and without heat treatment, followed by immunoblot analysis with an anti-AU-1 antibody. STP-B-SMHI and STP-B29-76 migrated at 26 kDa with or without heat treatment, whereas STP-B-SMHI/EE, STP-B-SMHI/Col, and STP-B29-76/Col migrated as larger sizes in SDS-PAGE than STP-B-SMHI and STP-B29-76 (Fig. 4). Similar to wt STP-B, the migration of STP-B-SMHI/EE was not significantly altered by heat treatment. In striking contrast, the migration of STP-B-SMHI/Col and STP-B29-76/Col in SDS-PAGE was drastically altered by heat treatment. STP-B-SMHI/Col and STP-B29-76/Col were detected as bands of approximately 100-200 kDa before heat treatment and as a 35-kDa band after heat treatment. This result indicates that introduction of the collagen repeat sequences but not the negative charged amino acids induces oligomerization of STP-B.

FIG. 4.

Oligomerization of STP-B by collagen repeats. 293T cells were transfected with pFJ-STP-B29-76 (lane 1), pFJ-STP-B29-76/Col (lane 2), pFJ-STP-B-SMHI (lane 3), pFJ-STP-B-SMHI/EE (lane 4), pFJ-STP-B-SMHI/Col (lane 5), and pFJ (lane 6). At 48 h after transfection, heat-treated and non-heat-treated cell lysates were fractionated by SDS-PAGE, transferred to nitrocellulose, and reacted with an anti-AU-1 antibody.

Association of STP-B with Src.

While STP-B from all isolates contains a highly conserved potential SH2 binding motif for Src family kinases, it lacks the negatively charged amino acids preceding this motif (Fig. 1). To investigate the potential association of STP-B with Src family kinases, 293T cells were cotransfected with an expression vector containing STP-B-SMHI or STP-B29-76 and Src tyrosine kinase. STP-A11, which has been shown to associate with and be phosphorylated by Src (36), was included as a control. After cotransfection, cell lysates were reacted with an anti-AU-1 antibody, and the immunoprecipitates were then resolved by SDS-PAGE after in vitro kinase reaction with [γ-³²P]ATP. It showed that the Src-binding activity of STP-B-SMHI and STP-B29-76 was significantly weaker than that of STP-A11 (Fig. 5).

FIG. 5.

Comparison of Src-binding activity between STP-A and STP-B. 293T cells were transfected with pFJ-Src with or without pFJ-AU1-STP-A11 (A11), pFJ-AU1-STP-B-SMHI (SMHI), and pFJ-AU1-STP-B29-76 (29-76). After 48 h, cell lysates were used for immunoprecipitation with anti-AU-1 antibody. AU-1 immune complexes were subjected to an in vitro kinase reaction. The expression level of Src, STP-A11, STP-B-SMHI, and STP-B29-76 in 293T cells was evaluated by immunoblot with anti-Src and anti-AU-1 antibodies (bottom two panels).

To investigate the potential role of negatively charged amino acids for efficient Src binding, the STP-B-SMHI/EE mutant containing the replacement of the asparagine at 114 (N₁₁₄) and the serine at 115 (S₁₁₅) with glutamic acids (E₁₁₄ and E₁₁₅) was compared with wild-type STP-SMHI for the Src-binding activity. STP-B-SMHI/Col and STP-B-SMHI/EE/Col mutants were also included in this assay. After immunoprecipitation with an anti-Src antibody, Src immune complexes were subjected to an in vitro kinase assay and an anti-AU-1 immunoblot or antiphosphotyrosine immunoblot. These experiments showed that the substitution of negatively charged amino acids at the putative SH2-binding motif enhanced the Src-binding activity of STP-B-SMHI by approximately fourfold (Fig. 6A and B, lanes 3 versus 4 and lanes 5 versus 6). In contrast, the insertion of collagen repeat sequences did not affect the Src-binding activity of STP-B-SMHI (Fig. 6A and B, lane 5). Essentially the same results were obtained with the STP-B29-76/EE mutant (data not shown). These results demonstrate that an attenuated Src-binding activity of STP-B is due to the absence of the negatively charged amino acids preceding the putative SH2-binding motif.

FIG. 6.

Enhanced Src-binding activity of STP-B by the substitution of negatively charged amino acids at the putative SH2-binding motif. 293T cells were transfected with pFJ-src alone (lane 2) or together with pFJ-STP-B-SMHI (lane 3), pFJ-STP-B-SMHI/EE (lane 4), pFJ-STP-B-SMHI/Col (lane 5), or pFJ-STP-B-SMHI/EE/Col (lane 6). After 48 h, cell lysates were used for immunoprecipitation (I.P.) with anti-Src 2 antibody. Src immune complexes were used for in vitro kinase assays (A). ³²P-labeled proteins were separated by SDS-PAGE followed by autoradiography in a Fuji Phospho Imager. Also, Src immune complexes were separated by SDS-PAGE, transferred to nitrocellulose, and reacted with an anti-AU-1 antibody (B). The asterisk indicates the heavy chain of immunoglobulin. The same nitrocellulose membrane described above was stripped with SDS and β-mercaptoethanol and reprobed with antiphosphotyrosine (P-Y) antibody (C). The expression level of Src, STP-B-SMHI, STP-B-SMHI/EE, STP-B-SMHI/Col, and STP-B-SMHI/EE/Col in 293T cells was evaluated by immunoblot (I.B.) with anti-Src and anti-AU-1 antibodies at the bottom of panel A.

The early-region-encoded protein of polyomavirus, the middle T antigen, interacts with cellular Src, and this interaction stimulates Src kinase activity (9). In contrast, in vitro kinase assay (Fig. 6A) and antiphosphotyrosine immunoblot (Fig. 6C) showed that an interaction of STP-B-SMHI did not increase autophosphorylation and tyrosine phosphorylation of Src kinase. Conversely, the enhanced interaction of STP-B-SMHI/EE and STP-B-SMHI/EE/Col slightly reduced the level of autophosphorylation of Src kinase (Fig. 6A). These results demonstrated that unlike polyomavivus middle T antigen, STP-B does not activate Src kinase activity.

HVS subgroup B STPs interact with TRAFs.

A BLAST search analysis showed that STP-B contains sequence homology with the region surrounding the PXQXT TRAF-binding motifs of HVS STP-A and herpesvirus papio LMP1 (Fig. 1B and 1C). This motif of STP-A, STP-C, LMP1, CD30, CD40, and TANK has been shown to be a core sequence for TRAF binding (8, 18, 20, 49). The interaction of STP-B with TRAFs was therefore investigated by cotransfecting 293T cells with expression vectors containing the AU-1-tagged STP-B-SMHI or STP-B29-76 and FLAG-tagged TRAF1 or TRAF2 (35). After transfection, TRAF complexes were precipitated with an anti-FLAG antibody, and the presence of STP-B-SMHI or STP-B29-76 in TRAF immune complexes was examined by immunoblot with an anti-AU-1 antibody. STP-B-SMHI and STP-B29-76 were readily detected in the TRAF1 and TRAF2 precipitates (Fig. 7). Repeated experiments showed that while STP-B-SMHI and STP-B29-76 exhibited equivalent levels of TRAF1 interaction, STP-B29-76 had higher levels of TRAF2 interaction than did STP-B-SMHI (Fig. 7). In contrast, STP-B-SMHI and STP-B29-76 were not detected from precipitates from negative control cell lysates without FLAG-tagged TRAF expression under the same conditions (Fig. 7). Furthermore, STP-B-SMHI/EE and STP-B-SMHI/Col mutants had similar levels of interactions with TRAF1 and TRAF2 as wild-type STP-B-SMHI (data not shown). These experiments demonstrated that STP-B-SMHI and STP-B29-76 specifically interacted with TRAFs in 293T cells and that STP-B29-76 had a higher binding activity to TRAF2 than did STP-B-SMHI.

FIG. 7.

Interaction of STP-B with TRAFs. 293T cells were transfected with the STP-B-SMHI or STP-B29-76 expression vector together with FLAG-tagged TRAF1 or TRAF2 expression vector as shown in the figure. After 48 h, cell extracts were used for immunoprecipitations (I.P.) with an anti-FLAG antibody. The upper panels show anti-FLAG immune complexes that were subjected to immunoblot (I.B.) with an anti-AU-1 antibody to detect STP-B. The expression level of TRAFs and STP-B-SMHI or STP-B29-76 in 293T cells was evaluated by immunoblot with anti-FLAG or anti-AU-1 antibodies (bottom two panels). The asterisk indicates the heavy chain of immunoglobulin.

Introduction of the collagen repeat sequence into STP-B induces activation of NF-κB activity.

Since TRAF2 can mediate NF-κB activation by interactions with EBV LMP1, HVS STP-C488, and TNF receptors (24, 35, 41, 43, 47, 49), we investigated the effect of STP-B and its mutants on NF-κB activation in Rat-1 cells using an NF-κB-driven luciferase reporter plasmid, 3X-kB-L, and a control β-galactosidase expression plasmid, pGKβgal. Relative luciferase values were normalized to β-galactosidase activity for transfection efficiency. Two independent assays revealed that stable expression of STP-B-SMHI and STP-B29-76 in Rat-1 cells did not induce NF-κB activity (Fig. 8). In striking contrast, the stable expression of STP-B-SMHI/Col and STP-B2/Col increased NF-κB activity by approximately 7- to 10-fold (Fig. 8). Furthermore, consistent with the level of TRAF2 interaction (Fig. 7), STP-B29-76/Col induced slightly higher levels of NF-κB activation than STP-B-SMHI/Col (Fig. 8). These results demonstrated that the introduction of the collagen repeat sequence into STP-B resulted in a dramatic increase in NF-κB activity.

FIG. 8.

Activation of NF-κB activity by oligomerization of STP-B. Rat-1 cells stably expressing STP-B or its mutants were transfected with 5 μg of an NF-κB-driven luciferase reporter (3XkB plasmid) together with 5 μg of the pGKβgal control plasmid to measure transfection efficiency. Forty-eight hours after transfection, cell lysates were used for luciferase and β-galactosidase assays. Luciferase activity (in relative light units) was determined and normalized to β-galactosidase activities. Vector, Rat-babe; SMHI, Rat-STP-B-SMHI; SMHI/Col, Rat-STP-B-SMHI/Col; 29-76, Rat-STP-B29-76; 29-76/Col, Rat-STP-B29-76/Col. Values represent the average of two independent experiments.

DISCUSSION

In this report, we demonstrate that consistent with the lack of oncogenic activity of HVS subgroup B virus, STP-B is deficient in the ability to transform rodent fibroblast cells. Sequence comparison of seven STP-B isolates revealed that STP-B lacks components of signal-transducing modules found in STPs of the other groups, the collagen repeats and an authentic SH2-binding motif. Substitution mutations demonstrated that the lack of collagen repeats but not an authentic SH2-binding motif contributes to the nontransforming phenotype of STP-B. Thus, these results demonstrate that the collagen repeat sequence is a determinant of the transforming activity of STP.

In addition to the absence of collagen repeat sequences, STP-B also lacks the negatively charged amino acids preceding the highly conserved YAEI sequence, which is the putative binding motif of the SH2 domain of Src family kinases. STP-A and most other SH2-binding proteins (51) contain these negatively charged amino acids preceding the YAEI/V motif. Substitution of negatively charged amino acids at the putative SH2-binding motif significantly enhances the Src-binding ability of STP-B. However, unlike the interaction of polyomavirus middle T antigen with Src kinase, which stimulates its kinase activity and induces cell growth transformation (9), the interaction of STP-B with Src neither activates its kinase activity nor contributes to the transforming activity of STP-B. In addition, an interaction of STP-A with c-Src has been shown to have no effect on Src kinase activity and transformation (36). Thus, an interaction of STP-A and STP-B with Src kinase appears to play a role in HVS biology other than transformation.

Human papillomaviruses (HPVs) can be divided into two subgroups based on disease association; high-risk HPV types are primarily associated with malignant tumors, and low-risk HPV types are exclusively associated with benign lesions (23). The major differences between high-risk and low-risk viruses in oncogenic ability are due to functional differences in the E6 and E7 oncoproteins (23). The E7 proteins derived from the high-risk HPVs bind to pRB with a higher affinity than the E7 proteins from the low-risk HPVs (25). The high-risk HPV E6 proteins associate with and target the p53 tumor suppressor protein for degradation at a higher rate than the low-risk HPV E6 proteins (33, 55). Similar to HPV, HVS subgroups exhibit differences in their capacity to induce pathogenesis, and the level of pathogenicity of HVS is strongly correlated to the oncogenic activity of the STP gene (4, 12, 14, 32). Our results demonstrate that the collagen repeat sequence is a determinant of the degree of STP transforming activity. This is also supported by the fact that a mutation that disrupts the collagen repeats disrupts the transforming activity of STP-C in culture (27). Furthermore, the transforming activity of STP-C, which has a higher number of collagen repeats, is stronger than that of STP-A.

Despite the absence of discernible homology among gammaherpesvirus transforming proteins, including Epstein-Barr virus LMP1, HVS STP, KSHV K1, and rhesus rhadinovirus R1, they all share the ability to self-oligomerize (10). Epstein-Barr virus LMP1 has been shown to aggregate through its membrane-spanning domains, mimicking a ligand-induced activated CD40 receptor (21, 22, 31, 56). KSHV K1 and rhesus rhadinovirus R1 have also been shown to oligomerize through disulfide bonding of their extracellular domain (11, 34, 37). Here, we demonstrate that the STP protein forms oligomers through its collagen repeats and the integrity of this domain is essential for the transforming activity of this protein. Thus, oligomerization of gammaherpesvirus transforming proteins may cluster the cellular signal-transducing molecules, mimicking a ligand-independent, constitutively active receptor, which generates continuous signals and ultimately results in cell growth transformation.