|Home | About | Journals | Submit | Contact Us | Français|
The recently discovered Canis familiaris papillomavirus (PV) type 2 (CfPV2) provides a unique opportunity to study PV gene functions in vitro and in vivo. Unlike the previously characterized canine oral PV, CfPV2 contains an E5 open reading frame and is associated with progression to squamous cell carcinoma. In the current study, we have expressed and characterized the CfPV2-encoded E5 protein, a small, hydrophobic, 41-amino-acid polypeptide. We demonstrate that, similar to the E5 protein from high-risk human PV type 16, the CfPV2 E5 protein is localized in the endoplasmic reticulum (ER) and that its expression decreases keratinocyte proliferation and cell life span. E5 expression also increases the percentage of cells in the G1 phase of the cell cycle, with a concomitant decrease in the percentage of cells in S phase. To identify a potential mechanism for E5-mediated growth inhibition from the ER, we developed a real-time PCR method to quantify the splicing of XBP1 mRNA as a measure of ER stress. We found that the CfPV2 E5 protein induced ER stress and that this, as well as the observed growth inhibition, is tempered significantly by coexpression of the CfPV2 E6 and E7 genes. It is possible that the spatial/temporal regulation of E6/E7 gene expression during keratinocyte differentiation might therefore modulate E5 activity and ER stress.
Papillomaviruses (PVs) are a large group of DNA tumor viruses that infect differentiated cutaneous and mucosal epithelia in a wide variety of mammalian species. There are nearly 200 types of human PVs (HPVs) (61), some of which are termed high risk (e.g., HPV type 16 [HPV-16]) and have the potential to immortalize primary cells and facilitate malignant progression to cervical cancer (52). An estimated 20 million cases of HPV infection occur each year in the United States alone, and cervical cancer is the second most common cause of cancer deaths among women worldwide. In general, PV infections are species specific, making it impossible to study the in vivo life cycle of HPV and the roles of its encoded proteins in viral replication and tumorigenesis. However, a few animal models do exist and the canine oral PV (COPV) has been helpful in mimicking certain biological properties of the high-risk mucosatropic HPVs, leading to the development of highly effective prophylactic vaccines (39, 49, 56). Although COPV mimics the mucosal tropism of the high-risk HPVs, it rarely progresses to cancer and lacks one of the early viral genes that may play an important role in tumorigenesis, E5. Recently, a new canine PV (Canis familiaris PV type 2 [CfPV2]) was isolated from the footpads of dogs (43). Unlike COPV, CfPV2 induces epidermal tumors and, when persistent, these benign infections progress to squamous cell carcinoma and metastasize widely. CfPV2 also encodes an E5 protein. In general, PV E5 proteins are small hydrophobic oncoproteins that localize to the endoplasmic reticulum (ER) or Golgi membranes (11, 16) but have limited amino acid sequence homology. Numerous cellular binding partners have been described for HPV-16 E5 proteins, including the V-ATPase 16-kDa subunit (1, 16), the nuclear import protein karyopherin beta 3 (25), the ER-resident protein Bap31 (40), proteins involved in zinc transport (ZnT1, EVER1, and EVER2) (27, 35), erbB4 (24), and HLA I (2). The HPV-16 E5 protein alters signaling pathways, predominantly the epidermal growth factor receptor (EGFR) pathway (17, 21, 46, 58); induces koilocytosis in cooperation with the E6 protein (26); and alters the plasma membrane expression of caveolin (47), HLA (3), and ganglioside GM1 (47). The last two changes might explain the ability of HPV-16-infected cells to circumvent detection by the host immune response and initiate tumor formation (3, 4, 21, 36, 46, 47).
To provide a foundation for future in vivo studies, we initiated a series of in vitro experiments to define the intracellular localization and biological activity of CfPV2 E5. The current study demonstrates that CfPV2 E5 exhibits several properties of the HPV-16 E5 protein, including ER localization and inhibition of cell proliferation. A novel finding is that CfPV2 E5 activates the ER stress-signaling pathway, which may explain some of E5's growth-related activities.
CfPV2 E5 codons were modified for optimum expression in mammalian cells (see Current Protocols in Molecular Biology, Table A.1.4 ). The constructs for CfPV2 E5 were made without an epitope tag (E5) or with a hemagglutinin (HA) epitope tag (YPYDVPDYA) at either the C-terminal or N-terminal ends (E5HA and HAE5, respectively) and were cloned into the following vectors: pJS55 for transient expression from the SV40 promoter (modified from pSG5; Stratagene) (42) or pLXSN for stable expression in cell lines utilizing the LTR promoter (Clontech).
COS-1 and NIH 3T3 cells were obtained from American Type Culture Collection (Manassas, VA) and maintained in Dulbeco's modified Eagle's medium (Invitrogen) supplemented with 10% fetal bovine serum, 2 mM glutamine, penicillin (100 units/ml), and streptomycin (100 μg/ml). Human ectocervical cells (HECs), immortalized with the HPV-16 E6 and E7 genes (6), and primary human foreskin keratinocyte (HFK) cells were maintained in keratinocyte growth medium (Invitrogen) supplemented with 50 μg/ml bovine pituitary extract, 26 ng/ml recombinant epidermal growth factor, and 10 μg/ml gentamicin. NIH 3T3, HEC, and HFK stable cell lines were prepared by transduction with Phoenix retrovirus encoding LXSN, E5, E5HA, HAE5, or HPV6b E5 (with or without CfPV2 E6/E7) and selected with G418 selection at 1 mg/ml (3T3 cells), 100 μg/ml (HEC cells), or 50 μg/ml (HFK cells).
COS-1 cells were seeded onto sterile glass coverslips (22 by 22 mm) in six-well plates at 1 × 105 cells per well and transfected with JS55 constructs using Lipofectamine 2000 (Invitrogen) according to the manufacturer's protocol.
HEC lines were grown to subconfluency in 150-mm2 tissue culture dishes and washed with 10 ml phosphate-buffered saline (PBS) at 4°C. Five-ml cold Triton X 100 (TX-100) extraction buffer (25 mM HEPES-NaOH [pH 7.4], 150 mM NaCl, 1% [wt/vol] TX-100) containing protease inhibitors (2 μg/ml aprotinin, 2 μg/ml leupeptin, 2 μg/ml pepstatin, and 1 mM phenylmethylsulfonyl fluoride) was added to the cells at 4°C, and the cells were rocked gently for 20 min to solubilize detergent-sensitive membrane proteins. Detergent-resistant proteins (DRM) remaining on the dish were solubilized in 2 ml RIPA buffer (0.04 M MOPS [morpholinepropanesulfonic acid], pH 7.2, containing 0.15 M NaCl, 0.001 M EDTA, 1% NP-40, 1% sodium deoxycholate, and 0.1% sodium dodecyl sulfate) plus protease inhibitors. For total cell lysates, the cells were solubilized in 5 ml RIPA buffer. The protein concentration of the lysates was determined using the Bio-Rad DC protein assay kit. Immunoprecipitations were performed using 4 μl of 12CA5 anti-HA mouse ascites fluid (a gift from J. Bolen, Millennium Pharmaceuticals, Cambridge, MA) and 800 to 1,000 μg of cellular protein. Flotillin-2 and the transferrin receptor were detected in RIPA lysates that were concentrated by precipitation with 20% (wt/vol) trichloracetic acid (TCA) for 30 to 60 min at 4°C. Precipitates were collected by centrifugation for 5 min at 10,000 rpm, washed in cold 95% acetone containing 0.1 N HCl, and centrifuged again. After removal of the supernatant, pellets were resuspended in 2× Laemmli sample buffer containing 10% β-mercaptoethanol and heated for 10 min at 110°C. The immunoprecipitates and TCA precipitates were separated electrophoretically using Tris-glycine minigels (Invitrogen) and analyzed by Western blotting. The HA epitope tag was detected using 12CA5 ascites fluid diluted 1:5,000 in PBS containing 2% bovine serum albumin and 0.05% Tween 20. Antibodies recognizing flotillin-2 (0.05 μg/ml; BD Biosciences no. 612124) and transferrin receptor (0.08 μg/ml; BD Biosciences no. 610383) were diluted in 2% bovine serum albumin/0.5% (wt/vol) TX-100, 140 mM NaCl, and 10 mM Na3PO4, pH 7.4.
Total RNA was prepared from subconfluent cell cultures using the RNaqueous-4 PCR kit (Ambion) with DNase treatment according to the manufacturer's protocol. Reverse transcriptase PCR (RT-PCR) was performed with the RETROScript kit (Ambion) with the addition of a separate RNA denaturation step of 3 min at 80°C with oligo(dT). The RT step was 60 min at 50°C followed by 10 min at 92°C, and the PCR step was 95°C for 4 min; 25 cycles of 94°C for 30 s, 55°C for 30 s, 72°C for 1 min; and final extension at 72°C for 10 min. Samples of the RNA were also subjected to PCR to confirm that they were free of DNA. The following CfPV2 E5-specific primers were used to detect untagged, codon-modified E5: RTCME5-F, 5′-ATGTACATGCTGGAGCTGCTGCAC-3′; and RTCME5-R, 5′-TCACAGGAAGTACTGCAGGATGAAG-3′. The following CfPV2 E6/E7-specific primers were used: E6-F, 5′-GGCCACTGAGCATACGCAGT-3′; and E7-R, 5′-GCCCGTCTAAGAGAAGCTGC-3′. PCR products were separated on a 2% agarose gel.
Cells were grown to approximately 80% confluence on 22- by 22-mm glass coverslips in six-well plates (BD Falcon) and transduced or transfected with the designated gene constructs. After 24 h, the cells were washed with PBS and fixed with 4% (wt/vol) paraformaldehyde for 20 min at room temperature. Cells were then washed four times with PBS, permeabilized with 0.1% (wt/vol) saponin in PBS for 10 min at room temperature, and washed two times with PBS. Coverslips were transferred to a humidity chamber and blocked for 20 to 30 min in Pgel-S (PBS containing 0.2% gelatin and 0.1% saponin or 5 μg/ml digitonin) supplemented with 20% normal donkey serum. The coverslips were washed three times with PBS (in six-well plates) and returned to the humidity chamber for incubation with the primary antibodies for 1 hour at room temperature. These were then washed three times in PBS and incubated with secondary antibodies (diluted in Pgel-S) for 1 hour at room temperature in the dark. Coverslips were then washed three times in PBS containing 0.2% gelatin and once with PBS. Nuclei were stained for 3 minutes at room temperature with 0.5 μg/ml Hoescht (no. 33342) in PBS and washed three times with PBS. Coverslips were mounted on glass slides using ProLong antifade mounting medium (Invitrogen; no. P7481) for 1 hour at room temperature and were stored at 4°C. A Zeiss Axioskop microscope (Carl Zeiss, Inc., Thornwood, NY) equipped with a ×63 objective, Hammamutsu charge-coupled-device camera, and Openlab 3.0.7 software was used to image the cells.
The following primary antibodies were used for immunofluorescence labeling: 5 to 7 μg/ml mouse HA.11 (Covance; no. MMS-101R), 2.7 μg/ml rabbit anti-calnexin (Santa Cruz; no. sc-11397), 0.6 μg/ml rabbit anti-GM130 (Calbiochem; no. CB1008), 1.3 μg/ml rabbit anti-karyopherin β-3 (Santa Cruz; no. sc-11369), and 1 μg/ml mouse anti-EGFR (Chemicon, Temecula, CA; no. CBL416). The following secondary antibodies were used at 5 μg/ml: Alexa Fluor 488 donkey anti-mouse immunoglobulin G (IgG) (Invitrogen; no. A-21202), Alexa Fluor 488 donkey anti-rabbit IgG (Invitrogen; no. A-21206), Alexa Fluor 555 donkey anti-mouse IgG (Invitrogen; no. A-31570), and Alexa Fluor 555 donkey anti-rabbit IgG (Invitrogen; no. A-31572).
Primary HFKs were transduced with Phoenix retrovirus (32, 44) and selected with 50 μg/ml G418. Controls for XBP1 splicing were prepared from primary HFKs that were either treated with tunicamycin (3 μg/ml; Sigma) for 2 h at 37°C or untreated. RNA was prepared from subconfluent HFK cells in keratinocyte growth medium without supplements using an RNaqueous-4PCR kit (Ambion) according to the manufacturer's protocol with DNase treatment. cDNA was generated with a RETROScript kit (Ambion) per the manufacturer's protocol using oligo(dT). Approximately 20 to 30 ng cDNA was used to perform real-time PCR using iQ SYBR green supermix (Bio-Rad) with the following primer sets: total XBP1-F, 5′ CCTTGTAGTTGAGAACCAGG 3′; total XBP1-R, 5′ GACTCAGCAGACCCGGCCAC 3′; XBP1s-F, 5′ GGTCTGCTGAGTCCGCAGCAGG 3′; XBP1u-F, 5′ GGTCTGCTGAGTCCGCAGCACTC 3′; and XBP1s/u-R, 5′ TCCAGAATGCCCAACAGGAT 3′.
The samples were run on Bio-Rad's My iQ single-color real-time PCR detection system (95°C, 3 min; 95°C, 10 s; 60°C 30 s; and 55°C, 10 s [40 times] followed by a melting curve up to 95°C at 0.5°C intervals) and analyzed using Bio-Rad iQ5 software. The relative ratio of spliced to unspliced XBP1 mRNA was normalized to 1 for the controls, and results for the experimental groups were calculated as change relative to those for the control groups. Statistical analysis was performed with a paired t test.
Primary HFKs were transduced with Phoenix retrovirus and selected with G418 as described. After selection, the cells were maintained in T75 tissue culture flasks and split 1:4 when they reached 80 to 90% confluence. The average growth rates (population doublings per day) were calculated and normalized against the LXSN vector control. Also, the average life span was measured (total population doublings). Statistical analysis was performed using the paired t test. These experiments were repeated a minimum of three times with cells isolated from different donors.
After trypsinizing HFK cells at no more than 90% confluence from a T75 flask, at least half of the cells were washed with 2 ml PBS and centrifuged at low speed (1,000 rpm) at 4°C for 5 minutes to generate a soft pellet. The cell pellet was resuspended in 0.5 ml cold PBS and vortexed at low speed. Ice-cold 100% ethanol (0.5 ml) was added sequentially three times to a final concentration of 75%. Fixed cells were stored for up to 1 week at −20°C before being stained with propidium iodide (50 mg/ml propidium iodide, 50 mg/ml RNase A, and 0.01% TX-100 in PBS) at the Lombardi Comprehensive Cancer Center Flow Cytometry and Cell Sorting core facility. Fluorescence-activated cell sorter analysis was performed with a BD FACSort cell sorter (BD Biosciences) and FCS Express software (De Novo Software). DNA histograms were modeled using ModFit LT software (Verity Software House). Statistical analysis was performed with a paired t test.
CfPV2 E5 was codon modified from the wild-type sequence (NCBI accession no. NC_006564) to enhance expression in cultured cells. A similar approach has been used to optimize expression of the HPV-16 E5 protein (19) and the HPV-16 L1 protein (15). The amino acid sequence of CfPV2 E5 is aligned with the wild-type and codon-modified nucleotide sequences shown in Fig. Fig.1A.1A. The codon-modified E5 gene was expressed in both epitope-tagged and nontagged forms. The epitope-tagged versions contained the HA epitope at either the N or C terminus. Expression of CfPV2 E5 tagged at its C terminus (E5HA) was confirmed by combined immunoprecipitation (IP)/immunoblotting in transiently transfected COS-1 cells (Fig. (Fig.1B,1B, E5HA lanes), as well as in NIH 3T3 murine fibroblasts and human foreskin and cervical cells (data not shown). The epitope-tagged E5 protein migrated as a 5-kDa protein, consistent with its predicted amino acid composition. While E5 tagged at its N terminus was detectable by fluorescence, it was not detected efficiently by IP; therefore, it was not used for biochemical studies. Since antibodies are not available to detect the untagged canine E5 protein, its expression was verified by RT-PCR in HEC cells (Fig. (Fig.1C),1C), as well as in HFK, NIH 3T3, and COS-1 cells (data not shown). Although the canine E5 protein contains two cysteine residues that could potentially participate in dimerization (as observed with the BPV-1 E5 protein), omission of the reducing agent in analytical gels failed to reveal disulfide-dependent dimers or higher-order oligomers (Fig. (Fig.1B).1B). To further exclude the possibility that CfPV2 E5 was forming dimers, we constructed an E5 mutant in which the two cysteine residues were replaced by alanines (mcys). The molecular weights of wild-type E5 and its cysteine mutant were identical in reducing and nonreducing electrophoretic conditions, further verifying the lack of disulfide bond-mediated dimerization by the canine E5 protein.
Differential detergent extraction (10, 14) was utilized to determine if CfPV2 E5 was present in DRMs, similar to what we have observed for the HPV-16 E5 protein (47). Membrane proteins that reside in DRMs are not soluble in TX-100 detergent at 4°C. Figure Figure2A2A demonstrates that the CfPV2 E5 protein is virtually completely solubilized with cold TX-100, with little E5 residing in DRMs in either stable HEC lines (Fig. (Fig.2A)2A) or NIH 3T3 cell lines (data not shown), indicating that it is not present in DRMs. To verify our detergent fractionation technique, the same detergent extracts were probed for the presence of flotillin-2, which is present predominantly in DRMs (7), and the transferrin receptor, which is present primarily in detergent-sensitive membrane domains (14) (Fig. (Fig.2B).2B). The data demonstrate the both flotillin-2 and the transferrin receptor are present in the expected membrane fractions (DRM fraction for flotillin-2 and TX-100-soluble fraction for the transferrin receptor). This analysis has been performed with both HEC and NIH 3T3 cell lines with similar results.
The 83-amino-acid HPV-16 E5 protein resides predominantly in ER membranes, whereas the smaller 44-amino-acid BPV-1 E5 protein resides in membranes of the Golgi apparatus (18, 46-48). To determine the intracellular localization of the 41-amino-acid CfPV2 E5 protein, we performed immunofluorescence microscopic analysis of stable CfPV2 E5-expressing HEC and NIH 3T3 cell lines, as well as transiently transfected COS-1 cells (Fig. (Fig.3).3). Cells were coimmunostained for the HA epitope and the ER-specific marker calnexin (51) or Golgi-specific marker golgin GM130 (30). CfPV2 E5 immunofluorescence merged with that of calnexin, indicating that E5 is present in the ER in the three cell types evaluated. CfPV2 E5 did not colocalize with the Golgi marker, GM130, in either stable HEC or NIH 3T3 cell lines. However, when canine E5 was transiently overexpressed in COS-1 cells, it was detected in both the ER and Golgi compartments. This result has been described previously for the HPV-16 E5 protein, which accumulates in the Golgi and the ER when overexpressed in COS-1 cells (16, 22). Thus, it appears that CfPV2 E5 is localized predominantly in the ER membranes when the protein is stably expressed in cell lines.
Differential membrane permeabilization was employed to deduce the orientation of the N- and C-terminal epitope tags on the CfPV2 E5 protein within the ER membrane. Fig. Fig.44 shows both stable HEC lines and transiently transfected COS-1 cells that were either permeabilized with digitonin (which makes the cytoplasmic compartment accessible to antibodies) (38) or saponin (which makes all cell compartments accessible to antibody) (37). Two antibodies were used to validate the permeabilization technique: an EGFR antibody that recognizes the extracellular domain of the receptor, which is localized in the Golgi lumen during processing before it is transported to the cell surface (51), and an antibody that recognizes the cytoplasmic karyopherin beta-3 (KNβ3) protein (53). As expected, the EGFR antibody reacted only with the cell surface EGFR in digitonin-permeabilized cells, whereas in saponin-permeablized cells it showed additional bright staining in the Golgi compartment. In contrast, the KNβ3 antibody showed the same staining pattern with either permeabilization method (Fig. (Fig.4).4). Cells expressing either N- or C-terminally tagged E5 (HAE5 or E5HA, respectively) were similarly detectable with the HA.11 antibody in cells partially permeabilized by digitonin or fully permeabilized with saponin, indicating that the HA epitope was exposed to the cytoplasm whether appended to the N or C terminus. This suggests that the canine E5 protein forms a loop in the membrane with its termini oriented toward the cytoplasm. However, an important caveat is that the HA epitope is hydrophilic and might potentially distort the orientation of the highly hydrophobic E5 protein, resulting in the cytoplasmic location of the E5 termini. It remains a possibility, therefore, that the native, nontagged E5 protein is entirely embedded in ER membranes, with both its N and C termini inaccessible to aqueous environments. Regardless, the E5HA protein retains the biologic activities of the nontagged protein (see below), indicating that it is not inactivated by epitope addition.
Since HPV-16 E5 causes growth inhibition and life span reduction of primary HFK cells (19), we examined whether CfPV2 E5 had similar activities. Primary HFK cells that were stably expressing either the nontagged or epitope-tagged E5 protein exhibited significantly decreased growth rates compared to control HFK cells transduced with vector (LXSN) (Fig. 5A and B). A similar inhibitory effect was observed when monitoring cell culture life span (Fig. (Fig.5C).5C). Control cells ceased proliferating after approximately 10 population doublings, whereas E5-expressing cells stopped dividing after approximately five population doublings. Finally, E5 induced a change in cell cycle distribution (Fig. (Fig.5D).5D). E5-expressing cells had an increase of cells in the G1 phase with a concomitant decrease in the S phase compared to the control cells. There were approximately 50% fewer E5 cells in S phase than those in the LXSN control. It is noteworthy that the HA-tagged E5 protein did not have significant differences for any of these activities compared to the untagged E5 protein.
Interestingly, when E5 was coexpressed with the CfPV2 E6/E7 genes, the growth inhibition, decreased life span, and cell cycle profile changes were minimized (Fig. (Fig.6).6). For example, there were no significant differences in growth rate or cell cycle distribution in the cells coexpressing E5, E6, and E7. There was still a significant inhibition of cell culture life span in cells expressing nontagged E5 protein, although it was much less prominent than the observations represented in Fig. Fig.5C,5C, in which E6/E7 was absent. Importantly, E5 mRNA expression levels were equivalent when expressed alone or in the presence of E6/E7 (Fig. (Fig.7).7). Thus, inhibition of E5 biological activity by the coexpression of E6/E7 is not mediated by inhibition of E5 gene expression.
Given its location in the ER membranes and its inhibition of cell growth, we postulated that E5 might be inducing ER stress. Therefore, we developed a real-time PCR method for quantifying ER stress via analysis of XBP1 mRNA splicing. XBP1 mRNA is spliced in response to ER stress through activation of IRE1 in the unfolded protein response (UPR) (12, 28, 54). The validity of this method for detecting spliced XBP1 mRNA was confirmed with primary HFK cells treated with tunicamycin (Tm), which induces ER stress by preventing glycosylation of proteins (50) (Fig. (Fig.8A).8A). Cells treated with Tm showed a threefold increase in the spliced form of XBP1 mRNA compared to untreated HFK cells. In stably transduced HFK lines, CfPV2 E5 and E5HA protein induced a 2.5- and 2.0-fold increase, respectively, in XBP1 mRNA splicing compared to the vector control (Fig. (Fig.8B).8B). However, when E5 was coexpressed with CfPV2 E6/E7, the change in XBP1 mRNA splicing was greatly reduced (Fig. (Fig.8C).8C). Again, there was no significant difference in the levels of XBP1 splicing between the tagged and untagged E5 proteins. These results demonstrate that E5 alone induces the UPR and that its effect is minimized by coexpression with E6/E7.
It was a possibility that the induction of ER stress by the canine E5 protein might be the simple consequence of expressing a hydrophobic protein in the ER. However, the HPV 6b protein did not induce ER stress (Fig. (Fig.8D),8D), despite its known localization in the ER (16, 18, 19).
CfPV2 is a new PV that may be a useful tool for understanding the life cycle of PVs. One of the most notable differences between CfPV2 and COPV is the presence of an E5 ORF. In general, E5 proteins are small hydrophobic proteins that reside in either the ER or Golgi membranes (11, 16). HPV E5 proteins have been implicated in a variety of biological processes, including growth factor receptor activation (20, 43, 45), immune evasion (3, 46, 57), and koilocytosis (26). In order to augment the expression of the canine E5 protein, we employed the technique of codon modification, which has been utilized successfully to express the HPV-16 E5 (19) and L1 proteins (15). The molecular basis for our observed increase in CfPV2 E5 protein expression by codon-modified sequences could arise from increases in mRNA stability, altered mRNA structure and translation, altered tRNA species abundance, or changes in protein stability, although none of these has been proven or disproven. Studies of HPV-16 E5 (18) suggest that there is no increase in E5 mRNA, suggesting that mRNA stability is not a factor. This is not the case for HPV-16 L1, in which mRNA stability appears to be responsible for regulation (14).
Due to the lack of an E5-specific antibody, an HA epitope tag was placed at the E5 C terminus to enable immunologic detection. To date, all of our studies indicate that the epitope does not affect the properties of the E5 protein, and this construct may prove useful in identifying E5-associated proteins.
CfPV2 E5 slows keratinocyte proliferation, and we propose that this may be caused by the induction of ER stress. ER stress can be induced by the accumulation of unfolded proteins in the ER, thereby activating the UPR and ultimately resulting in attenuated mRNA translation, promotion of protein folding, secretion, degradation, and potentially apoptosis (41, 59). The goal of this response is to allow the cell to reestablish a balance of the appropriate cellular machinery needed to make and process properly folded proteins (41). When a cell cannot manage unfolded or misfolded proteins accumulating in the ER, the ATF6, PERK, or IRE1 pathway is activated (41). With regard to the last pathway, IRE1 oligomerizes and transphosphorylates itself into an active endoribonuclease (12, 28). The RNase function of IRE1 then cleaves a 26-nucleotide portion of the XBP1 mRNA, which results in a more stable, spliced form of XBP1 mRNA, which is translated into a coactivator protein. The spliced XBP1 protein acts as a transcription factor and preferentially activates genes encoding ER chaperones, lipid metabolism enzymes, and ER-associated degradation genes, thereby allowing the cell to reestablish ER homeostasis (41, 59). We developed a real-time RT-PCR method to quantify the ratio of spliced to unspliced XBP1 mRNA and show that CfPV2 E5 is capable of significantly increasing the amount of the spliced form of XBP1 mRNA. Interestingly, this effect, as well as the observed growth inhibition, is abrogated by coexpression with CfPV2 E6/E7 genes. As previously noted, the UPR can result in cell cycle arrest in G1 c17 (c17 c17 8c17, c17 9c17 c17), which was also observed with cells that express CfPV2 E5. Again, these effects were negligible when CfPV2 E6/E7 was also expressed. It is important to note that growth inhibition by E5 is apparently not the result of cell death or apoptosis since there were no changes in cell morphology or the presence of a pre-G1 shoulder in the fluorescence-activated cell sorter profile of cellular DNA content, which is indicative of DNA fragmentation.
Flavivirus and hepatitus B viral proteins NS4B (31, 55, 60) and HBx (29), respectively, have both been reported to activate the UPR by splicing of XBP1 mRNA. Both NS4B and HBx are associated with internal membranes and are reported to be involved in regulation of the viral life cycles (23, 33). Their activation of the UPR has been implicated as a possible mechanism for controlling either viral replication or pathogenesis (29, 31, 55, 60). The activation of the UPR by CfPV2 E5 may have similar effects on the viral life cycle. Also, preliminary microarray analysis (unpublished data) indicates that CfPV2 E5 decreases expression of genes involved in lipid, amino acid, and carbohydrate metabolism. This further supports our current findings that canine E5 is activating the UPR. Our results do not reveal whether CfPV2 E5 induces an accumulation of unfolded proteins within the ER, whether it interacts directly with a component of the UPR or leads to ER membrane destabilization and calcium flux changes to activate the UPR.
We performed the growth inhibition and ER stress studies of primary HFKs since primary canine keratinocyte cultures were not available. With this caveat in mind, our results demonstrate that the canine E5 protein induces ER stress in keratinocytes and that the HA epitope tag does not affect this biological activity. In addition, coexpression of the canine E6/E7 proteins with E5 seems to either alter or mask the activities of E5 that are evident when E5 is expressed alone. In the future, it will be important to correlate the differential expression of the canine E5, E6, and E7 proteins with in vivo phenotypes.
This work was supported by NIH grants R01CA106440 and R01CA53371 to R.S.
Published ahead of print on 7 October 2009.