Interaction of Adeno-associated Virus Rep78 with p53: Implications in Growth Inhibition¹ Free

AAV,⁴ a 4.7-kb single-stranded DNA virus, belongs to the family of parvoviridae and can infect various species, including humans (1). However, it is not known to be associated with any disease. In fact, it has been shown to inhibit the transforming potential of various agents (2). AAV latently infects the cell and integrates into a specific site, AAVS1, on chromosome 19 (3). It generally requires coinfection with helper viruses for its productive replication (1).

Several reports have demonstrated that tumors generated by adenovirus in chicks (4) and in hamsters (5, 6) were inhibited in the presence of AAV. Coinfection of AAV with adenovirus leads to the generation of a reduced number of plaques when compared with infection by adenovirus alone (7, 8). It has been observed that in the presence of AAV, the transforming potential of adenovirus in various cell lines is significantly reduced (7, 9, 10). These tumor-suppressive and antiproliferative properties have been mapped to the left half of the AAV genome, which codes for the multifunctional regulatory protein Rep78 (11, 12).

Although the tumor-suppressive and antiproliferative properties of AAV Rep78 have been well documented, nothing is known about the molecular mechanisms behind this phenomenon. To induce cell proliferation, the adenovirus first has to remove the cell cycle block. An important cell cycle checkpoint in mammalian cells that acts in the G₁ phase is mediated by the p53 tumor suppressor gene (13). Adenoviral E1B, an early gene product, has been shown to form a complex with p53 (14). This interaction of adenoviral E1B with p53 induces ubiquitin-mediated degradation of p53 (15). Because AAV inhibits the cell proliferation potentials of adenovirus, this work was undertaken to study whether AAV interferes with the adenoviral-mediated abrogation of p53 functions. As these antioncogenic properties of AAV are confined to its regulatory protein, AAV Rep78, we have investigated whether it interacts with p53, potentially protecting it from adenoviral-mediated degradation. Here we document that p53 levels are stabilized by AAV in adenovirus-infected cells. Furthermore, we document the interaction of AAV Rep78 with p53 that may be responsible for the observed protection of p53.

The HeLa and 293 cell lines and fibroblasts cells were procured from the American Type Culture Collection (Manassas, VA) and maintained at 37°C under a humidified atmosphere containing 5% CO₂ in air in DMEM supplemented with 10% fetal bovine serum, 1% penicillin G-streptomycin, and 2 mm l-glutamine (Life Technologies, Inc., Grand Island, NY). AAV-2 was prepared from HeLa cells infected with adenovirus-2 and AAV-2. At 24 h after plating (6 × 10⁶ cells; twenty 100-mm dishes), semiconfluent HeLa cultures were inoculated with 10 multiplicity of infection units of both AAV-2 and adenovirus-2 in DMEM without serum or antibiotics for 1 h at 37°C. Infectious medium was replaced with DMEM supplemented with 5% FCS, 1% penicillin G-streptomycin, and 2 mm l-glutamine (Life Technologies, Inc.). Cultures showing adenovirus-induced cytopathic effects observed after approximately 48 h were washed with DMEM without serum or antibiotics, and the cells were harvested in 10 ml of DMEM and subjected to three freeze/thaw cycles. Cells were spun at 10,000 rpm at 4°C for 20 min, and the supernatant was adjusted to 5 mm manganese chloride and incubated at 37°C for 30 min with 5,000 units of DNase I and RNase (0.2 mg/ml). This preparation was spun again at 10,000 rpm at 4°C for 20 min, and the supernatant was incubated at 56°C for 1 h to inactivate adenovirus. The absence of adenovirus was confirmed by inoculating HeLa cells with this preparation and probing the total cell extracts for adenoviral proteins. This preparation was adjusted to 5% glycerol, and the aliquots were stored in liquid nitrogen. All of the experiments were carried out with 10 multiplicity of infection units of adenovirus-2 and/or AAV-2 on semiconfluent HeLa or 293 cells, and the cells were harvested at the desired time intervals.

AAV Rep78 was PCR-amplified using linearized plasmid pSub201 (16). Primers designed to amplify Rep78 were as follows: (a) primer 1, 5′-CCGGGGTTTTACGAGATTGT-3′; and (b) primer 2, 5′-TTGTTCAAAGATGCAGTCAT-3′.

Both primers were designed to have BglII/BamHI overhangs. The PCR product was gel-purified using the Qiaex II gel extraction kit (Qiagen, Chatsworth, CA). The resulting open reading frame of Rep78 was cloned into pQE16 (QIAexpressionist; Qiagen) vector as follows. First, the pQE16 vector was digested with BamHI and BglII, and the resulting linearized plasmid, which has a 6 His tag at the COOH-terminal end, was gel-purified. PCR-amplified Rep78 was ligated to linearized pQE16 overnight at 16°C. Clones were selected after bacterial transformation and analyzed for protein expression. Briefly, small overnight cultures were diluted 1:50 and grown until the absorbance (A⁶⁰⁰) reached 0.7. Isopropyl-β-d-thiogalactoside was added to a final concentration of 1 mm, and the incubation was continued for 4 h after the induction. Purification of His-Rep78 was carried out under native conditions, essentially following the protocol provided by Qiagen. The purity of the protein was confirmed by 8% SDS-PAGE.

The His-Rep78 protein was expressed as described previously, and protein was adsorbed to Ni-NTA spin columns (Qiagen) according to the instructions given in the manual. p53 protein (Santa Cruz Biotechnology, Santa Cruz, CA) was chromatographed on the Rep78 affinity column by incubating at 4°C for 30 min. His-Rep78 was then eluted with 250 mm imidazole and subjected to 8% SDS electrophoresis. Some blots were also transferred to nitrocellulose membrane as mentioned earlier for probing with antibodies.

After various treatments, 100-mm cell culture plates were washed three times with PBS and incubated with 1× lysis buffer [50 mm Tris (pH 7.4), 150 mm NaCl, 0.5% Triton X-100, 0.5% NP40, 1 mm EDTA, and 1 mm EGTA] for 30 min with a protease inhibitor mixture from Boehringer Mannheim. The cell pellet was subjected to three freeze/thaw cycles and left on ice at 4°C with constant shaking for 30 min. Cell debris was removed by refrigerated centrifugation for 5 min at 12,000 rpm. Supernatants were collected, and the protein content was estimated using the Micro BCA kit (Pierce, Rockford, IL). Protein contents of all of the samples were normalized to 10 mg/ml with lysis buffer, aliquoted, and stored at −70°C. Immunoprecipitations were conducted essentially as described previously (17).

After different virus treatments, aliquots of total protein extracts (100 μg) from cells were suspended in Laemmli’s sample buffer [0.1 m Tris-Cl (pH 6.8) containing 1% SDS, 0.05% β-mercaptoethanol, 10% glycerol, and 0.001% bromphenol blue]; boiled for 2 min; applied on either 8%, 12%, or 8–12% (18) glycerol gradient SDS-acrylamide along with a M_r 10,000 protein ladder from Life Technologies, Inc.; and electrophoresed for 16 h using the Bio-Rad PROTEIN II system at 60 V. Gels were electroblotted overnight onto nitrocellulose paper (Trans-Blot transfer membrane, 0.2 μm; Bio-Rad; Hercules, CA) at 40 V for 3 h in a Tris-glycine buffer system. The transfer was confirmed by Ponceu S staining of the blot. Nonspecific sites on the blots were blocked with PBST. Incubation with various antibodies [AAV Rep78 (American Research Products), p53 (Santa Cruz Biotechnology), or β-actin (Sigma, St. Loius, MO)] was performed for 2 h in PBST containing 1% BSA with constant rocking. Blots were washed three times with PBST and incubated in either antirabbit or antimouse horseradish peroxidase conjugates for 2 h in PBST. After washing, specific proteins were detected using ECL according to the instructions provided in the manual (Amersham Life Sciences, Inc., Arlington Heights, IL).

E1B-mediated active degradation of p53 protein in adenovirus-infected cells is well known (19). After adenovirus infection, normal levels of p53 transcript are observed. However, ubiquitin-mediated protein degradation leads to lower levels of the p53 protein, which is essential for adenoviral propagation and cell proliferation (15, 19). Because AAV inhibits cell proliferation initiated by the adenovirus, we examined the effect of AAV expression on p53 levels with regard to adenovirus infection. Total cell lysates were prepared at various time points after viral infections and probed for p53 by Western blot analysis. As seen in Fig. 1, no appreciable change in the protein levels is observed before 48 h in HeLa cells. However, there is a significant reduction in p53 levels at 48 and 72 h with adenovirus infection. Coinfection of AAV with adenovirus protects against adenoviral-mediated degradation of p53 (Fig. 1,A). Probing the same blot with β-actin confirmed equal protein loading in each lane. As seen in Fig. 1,B, the quantitation of p53 levels in comparison with the β-actin levels shows a significant block in the adenoviral-mediated degradation of the p53 by AAV. In an effort to correlate the inhibition of p53 degradation and AAV expression, we probed the cell lysates for AAV Rep protein expression at various time points. As seen in Fig. 1,C, Rep protein expression starts only at 30 h after AAV infection and reaches a plateau after 48 h, correlating with the observed inhibition of adenoviral-mediated p53 degradation by AAV at 48 h. The effect of adenovirus infection was also examined in 293 cells because, unlike HeLa cells, these cells do not contain HPV genes. Decreased levels of p53 after adenovirus infection, apparent after 24 h in this cell line, were also partially blocked by AAV (Fig. 2,A). In the 293 cell line, we observe a significant expression of AAV Rep proteins by 24 h, which is consistent with the observed inhibition at that time point (Fig. 2 B).

Next, we evaluated whether protein-protein interactions are involved in the stabilization of p53 protein as it has been reported that Rep78, is responsible for onco-suppressive properties of AAV (11, 12). To investigate the interaction between AAV Rep78 and p53, we produced the Rep78 affinity column as described in “Materials and Methods” and incubated it with p53. After several washings, eluted proteins were electrophoresed, transferred onto nitrocellulose blots, and probed with the p53 antibody. As seen in Fig. 3,A, p53 was eluted along with Rep78, indicating a physical interaction between both proteins in vitro. The same blot was also reprobed with both AAV Rep78 and p53 antibodies together. Fig. 3 B shows that the p53 is eluted with Rep78, confirming the identity and the interaction of the proteins.

Although previous experiments have shown the interaction of p53 and Rep78 in vitro, we wanted to confirm whether the two proteins interact in vivo. Cells were lysed 36 h after infection with the viruses, and one set was immunoprecipitated with Rep78 antibody, and the other set was immunoprecipitated with p53 antibody. These extracts were subjected to electrophoresis and Western blotting and probed for p53. As expected, when Rep78 precipitates were probed with p53 antibodies, we observed the presence of p53 in the cells infected with both adenovirus and AAV (Fig. 4,A). AAV is a helper-dependent virus, and the expression of AAV proteins occurs only with adenovirus superinfection in humans. Therefore, Rep78 expression is not expected in either mock-transfected cells, cells infected with adenovirus alone, or cells infected with AAV alone, as observed in this figure. When these extracts were immunoprecipitated with p53 and probed with p53 antibody (Fig. 4,B), adenovirus-infected cells did not show the presence of p53, confirming the degradation of p53 after adenovirus infection. In the presence of AAV superinfection, the p53 band is seen just below the IgG heavy chain, which is around M_r 55,000. To confirm the specificity of these in vivo interactions, we probed the p53 immunoprecipitations with Rep78 antibodies after electrophoresis and Western blot transfer. As seen in Fig. 5 AAV Rep78 and Rep52 proteins coimmunoprecipitated with p53. AAV produces four regulatory proteins, Rep78, Rep68, Rep52, and Rep40. All of them are produced by differential splicing of a transcript from a single open reading frame. The monoclonal antibody we used for probing AAV Rep78 recognizes all four regulatory proteins; however, the amounts of all four proteins vary to a significant extent in both HeLa (Fig. 1,C) and 293 cells (Fig. 2 B). Rep78 and Rep52 are overexpressed compared to the other two regulatory proteins, Rep68 and Rep40, thus these two proteins were seen on the blot.

Adenovirus depends on host cell mechanisms to replicate the viral genome. In consequence, it encodes early gene products capable of activating host cell cycle progression and proliferative processes (14). The transition from G₁ to S phase is an important regulatory point in cell cycle progression. p53, a well-characterized tumor suppressor protein, is antithetical to growth by blocking the cell in the G₁ phase of the cell cycle (13). Adenovirus large E1B protein, an early gene product, has been shown to form a complex with p53 (14). This interaction not only inactivates p53 but targets it for ubiquitination, leading to its active degradation (15). These reports indicate that the inactivation and degradation of p53 is essential for adenoviral-mediated cell proliferation.

AAV has been shown to possess a tumor suppressor property against adenovirus in various systems (4, 5, 6, 7). This property has been mapped to the left half of the AAV genome, which codes for a multifunctional regulatory protein, Rep78 (11, 12). When the adenovirus-infected HeLa cells were challenged with AAV, p53 degradation by adenovirus was substantially inhibited (Fig. 1, A and B). We chose the time points to observe p53 protein levels by exploring the timing of AAV expression after the initial infection (Fig. 1,C). AAV expression starts 36 h after the viral infection, and we observed the inhibition of adenoviral-mediated p53 degradation at 48 h. Similarly, in 293 cells, adenoviral-mediated degradation of p53 protein was substantially reduced in the presence of AAV at 24 and 36 h, when 293 cells express the maximum amounts of AAV Rep proteins (Fig. 2 B). The reduction in p53 after adenoviral infection is less marked in 293 cells, possibly due to the presence of adenoviral early genes in this cell line. The AAV-mediated p53 protection in 293 cells is also less marked compared to that in the HeLa cells. This may be due to observation at earlier time points, the presence of adenoviral early genes in this cell line, or a high background of p53 expression in control cells. Interestingly, the AAV-mediated protection of p53 becomes more obvious at 36 h than at 24 h.

Whereas these data indicate that AAV rescues p53 from adenovirus-mediated degradation, they do not provide a possible mechanism. To address this question, we have investigated the role of Rep78 in exerting growth-inhibitory effects on primary human cells (20) and the transforming potentials of various DNA tumor viruses (2). With this background, we checked whether AAV Rep78 has any binding affinity with p53 that may be responsible for its protection from adenovirus-mediated degradation. The binding of p53 with the Rep78 affinity column in a dose-dependent manner confirmed its in vitro interaction with AAV Rep78 (Fig. 3,A). Neither glutathione S-transferase moiety nor p53 moiety of the fusion protein has any affinity toward the Ni-NTA affinity column (data not shown), and further coelution of Rep78 with p53 was confirmed by probing the same blot with both antibodies (Fig. 3,B). These protein-protein interactions were confirmed in vivo by immunoprecipitating 293 cells with Rep78, followed by p53 probing. We observed a strong p53 signal in coimmunoprecipitates of Rep78 when it is expressed in the cells. Also the absence of p53 signal in either control, with AAV or with adenovirus infection, shows that this interaction is specific for Rep78 (Fig. 4 A). It is important to observe that Rep78 is expressed only when adenovirus is coinfected with AAV, because AAV expression is dependent on the adenovirus helper functions in humans (1). In view of the onco-suppressive nature of Rep78, its interaction with p53 both in vitro and in vivo leading to elevated amounts of p53 suggests a possible stabilization of p53 from ubiquitin-mediated degradation by adenovirus. By protecting p53 from ubiquitin-mediated degradation, the function of p53 as a cell cycle blocking agent is restored. Protein-protein interactions leading to protection from ubiquitin-mediated degradation are not uncommon. For example, the binding of retinoblastoma, another onco-suppressor protein, to E2F-1, which is a ubiquitous cell cycle-related transcription factor, confers protection against ubiquitin-mediated degradation (21).

Interestingly, of the four spliced versions of AAV Rep78, we detected only Rep78 and Rep52 in coimmunoprecipitations with p53 (Fig. 5,A). It is possible that the lower expressions of Rep68 and Rep40, compared to the other two Rep proteins (Figs. 1 C and 2B), resulted in the much lower coprecipitations and the inability to detect them on the blot. At present, we are trying to determine how the interaction between AAV Rep proteins and p53 protects it from adenoviral-mediated degradation.

This report suggests for the first time the possible molecular mechanism behind the onco-suppressive nature of AAV, which has been widely reported over the years. Insights into the molecular mechanisms underlying the onco-suppressive nature of AAV not only contribute to our basic understanding about oncogenesis but may also suggest a possible use of AAV Rep78 as an antiproliferative agent. It is worthwhile to note that AAV infects humans without any known pathological consequences (1), and the sero-epidmeological data clearly suggest a negative correlation between the occurrence of cervical cancer and the presence of AAV (22, 23). In summary, we present evidence that the onco-suppressive properties of AAV against adenovirus are mediated, at least in part, by the protection of p53 degradation, presumably by the interaction with AAV Rep78.

We thank Jeana Cromer for editorial assistance.