JOURNAL OF VIROLOGY, Oct. 2010, p. 10661–10670 0022-538X/10/$12.00 doi:10.1128/JVI.01081-10 Copyright © 2010, American Society for Microbiology. All Rights Reserved.
Vol. 84, No. 20
Papillomavirus Infection Requires ␥ Secretase䌤 Balasubramanyam Karanam,1 Shiwen Peng,1 Tong Li,1 Christopher Buck,2 Patricia M. Day,2 and Richard B. S. Roden1,3,4* Departments of Pathology,1 Oncology,3 and Gynecology and Obstetrics,4 The Johns Hopkins University, Baltimore, Maryland 21231, and Laboratory of Cellular Oncology, National Cancer Institute, Bethesda, Maryland 208922 Received 19 May 2010/Accepted 1 August 2010
The mechanism by which papillomaviruses breach cellular membranes to deliver their genomic cargo to the nucleus is poorly understood. Here, we show that infection by a broad range of papillomavirus types requires the intramembrane protease ␥ secretase. The ␥-secretase inhibitor (S,S)-2-[2-(3,5-difluorophenyl)acetylamino]-N-(1-methyl-2-oxo-5-phenyl-2,3-dihydro-1H-benzo[e][1,4]diazepin-3-yl)-propionamide (compound XXI) inhibits infection in vitro by all types of papillomavirus pseudovirions tested, with a 50% inhibitory concentration (IC50) of 130 to 1,000 pM, regardless of reporter construct and without impacting cellular viability. Conversely, XXI does not inhibit in vitro infection by adenovirus or pseudovirions derived from the BK or Merkel cell polyomaviruses. Vaginal application of XXI prevents infection of the mouse genital tract by human papillomavirus type 16 (HPV16) pseudovirions. Nicastrin and presenilin-1 are essential components of the ␥-secretase complex, and mouse embryo fibroblasts deficient in any one of these components were not infected by HPV16, whereas wild-type and -secretase (BACE1)-deficient cells were susceptible. Neither the uptake of HPV16 into Lamp-1-positive perinuclear vesicles nor the disassembly of capsid to reveal both internal L1 and L2 epitopes and bromodeoxyuridine (BrdU)-labeled encapsidated DNA is dependent upon ␥-secretase activity. However, blockade of ␥-secretase activity by XXI prevents the BrdU-labeled DNA encapsidated by HPV16 from reaching the ND10 subnuclear domains. Since prior studies indicate that L2 is critical for endosomal escape and targeting of the viral DNA to ND10 and that ␥ secretase is located in endosomal membranes, our findings suggest that either L2 or an intracellular receptor are cleaved by ␥ secretase as papillomavirus escapes the endosome. The necessary causal association of persistent infection by an “oncogenic” type of human papillomavirus (HPV) with cervical cancer is firmly established (52, 53). HPV is the most prevalent sexually transmitted infection, and although the majority of patients clear their infection, HPV is directly responsible for 5% of all cancer deaths worldwide (30). HPV is also associated with multiple other anogenital cancers and oropharyngeal cancers. The life cycle of HPV is closely linked to epithelial differentiation within stratified squamous epithelia (16). Initial infection occurs within the undifferentiated proliferative basal cell layer in which only the viral early proteins are expressed, whereas production of the late proteins and, thus, progeny virus is restricted to the terminally differentiated suprabasal compartment (53). The exquisite dependence of virion production upon epithelial differentiation and lack of a rapid phenotype in culture can be circumvented by ectopic expression of the capsid proteins L1 and L2 in cells maintaining viral genome or reporter constructs as episomes, resulting in “quasivirions” or “pseudovirions,” respectively, whose infectivity can be readily and rapidly quantified in vitro or in vivo (6, 11, 35, 41). The completion of the entire papillomavirus life cycle is species specific. However, studies with bovine papillomavirus (BPV) in horses and hamsters, HPV pseudovirions in mouse * Corresponding author. Mailing address: Department of Pathology, Johns Hopkins University, Room 308, Cancer Research Building 2, 1550 Orleans St., Baltimore, MD 21231. Phone: (410) 502-5161. Fax: (443) 287-4295. E-mail:
[email protected]. 䌤 Published ahead of print on 11 August 2010.
challenge models, and quasivirions in rabbits suggest that virion internalization and delivery of the encapsidated DNA to the nucleus are promiscuous and that tropism is determined at a later stage of the life cycle (11, 27, 29, 39). Although significant progress has been made in understanding the HPV life cycle and virion structure, many of the molecular events of virus internalization and infection are poorly defined (43). Both the L1 (major) and L2 (minor) capsid proteins provide essential functions during infection (41) (8). L1 is sufficient to form empty capsids, termed virus-like particles (VLPs) (25), which bind to basement membrane and to the cell surface and which also form the basis of the licensed HPV vaccines (10). Glycosaminoglycans (GAGs), most notably heparan sulfate (HS), play a critical role in virion binding and infection, both in vitro and in the murine vaginal challenge model, although differences between HPV types and target cells in vitro have been described (14, 19, 20), for example, between HPV16 and HPV31 (4, 34, 42). Once bound to the basement membrane, the virions undergo a conformation change resulting in the surface display of the amino terminus of L2 and its cleavage by a proprotein convertase (PC), furin and/or PC5/PC6, and the transfer of virions to the cell surface (24). The uptake of the virions is apparently slow as late addition of neutralizing antibodies several hours after initial cell surface binding prevents infection in vitro (9). The endocytic mechanisms reported for various papillomavirus types are diverse, but furin cleavage of L2 and endosomal acidification are critical shared steps (15, 38). In a late endosomal compartment, the L1 capsid disassembles, releasing L2 associated with the previously encapsidated DNA to gain access to the nucleus
10661
10662
KARANAM ET AL.
by an unknown mechanism and to accumulate at the subnuclear domain, ND10 (13). Although L2 contains a C-terminal nuclear localization signal (17), entry to mitosis, which is associated with the dissolution of the nuclear membrane, is required for infection, suggesting that the complex with the viral nucleohistone core is unable pass through nuclear pores (36). It is unclear how the L2-genome complex escapes the endocytic compartment, but the carboxy terminus of L2 also contains both DNA binding and a membrane-destabilizing peptide (21). ␥ Secretase is an intramembranously cleaving protease (ICliP) linked to Alzheimer’s disease through its cleavage of amyloid precursor protein (APP) (1). It is a multicomponent complex, and presenilin (PS) is the catalytic unit whose active site contains two aspartate residues. In addition to the ninepass transmembrane protein PS, ␥ secretase requires nicastrin (NCT), anterior pharynx defective-1 (APH-1), and presenilin enhancer-2 in an equimolar ratio for proteolytic activity (28). The subcellular localization of ␥ secretase is controversial but includes the endoplasmic reticulum (23), endosome (26), lysosome (31), and plasma membrane (37), all of which are subcellular locales possibly traversed by papillomavirus during infection (43). By analogy to the cleavage of L2 by furin that is critical for exit from the endosomes (38), we hypothesized that I-CLiP might contribute to papillomavirus infection. Here, we report that a ␥-secretase inhibitor prevents HPV infection both in vitro and in the mouse vaginal challenge model and that cell lines lacking essential components of ␥ secretase are refractory to HPV infection.
MATERIALS AND METHODS Chemicals and reagents. Compound XXI [(S,S)-2-[2-(3,5-difluorophenyl)acetylamino]-N-(1-methyl-2-oxo-5-phenyl-2,3-dihydro-1H-benzo[e][1,4]diazepin-3yl)-propionamide] and other inhibitors were purchased from Calbiochem, with the exception of TAPI-0 (tumor necrosis factor alpha [TNF-␣] protease inhibitor 0], which was purchased from Biomol International, and marimastat, batimastat and DAPT, which were purchased from Tocris Biosciences. Cell lines null for PS1 (12), NCT (28), and -site APP-cleaving enzyme (BACE) (7) were described previously. BK and Merkel cell polyomaviruses (BKV and MCV, respectively) were generated as described previously (33), and green fluorescent protein (GFP)-recombinant adenovirus was kindly provided by Gary Ketner of Johns Hopkins University. HPV inhibition assays. HPV16, -18, -31, and -45 and cottontail rabbit papillomavirus (CRPV) pseudovirions with encapsulated secreted alkaline phosphatase (SEAP) or GFP reporter were generated by cotransfection of 293TT cells with plasmids encoding codon-modified L1 and L2 and a SEAP reporter plasmid, as described previously (32). Cells collected after transfection were matured overnight in Brij 58 (0.5%) and benzonase (0.5%) and purified by centrifugation on an OptiPrep step gradient (27, 33, and 39%) at 40,000 rpm for 4.5 h. Pseudovirus neutralization assays were carried out as outlined previously (18). Briefly, the pseudovirus with or without inhibitors was incubated with 293TT cells. At 68 to 72 h postinfection, the supernatants were collected, and SEAP activity in the supernatants was measured by a colorimetric assay. The 50% inhibitory concentration (IC50) was defined as the concentration that caused a 50% reduction in SEAP activity compared to activity in the control. Cell viability assay. Cell viability was determined by a Cell Titer 96 Aqueous nonradioactive cell proliferation assay using MTS [3,4-(5-dimethylthiazol-2-yl)5-(3-carboxymethoxy phenyl)-2-(4-sulfophenyl)-2H-tetrazolium salt] (Promega Corp., Madison, WI). Cells seeded at 1,000 cells/well in 100 l of medium in 96-well plates were treated with ␥-secretase inhibitors at specified concentrations (0 to 250 M) for 72 h. Cells were incubated according to the manufacturer’s protocol with the MTS mixture for the final 4 h. Formazan dye was quantified using a spectrophotometric plate reader to measure the absorbance at 490 nm
J. VIROL. (Bio-Rad enzyme-linked immunosorbent assay [ELISA] reader). All experiments were performed in triplicate. Flow cytometry. 293TT, HaCaT, A549, and NCT⫹/⫹ or NCT⫺/⫺, PS1⫺/⫺, or BACE⫺/⫺ mouse embryo fibroblasts were incubated with HPV pseudovirus containing GFP reporter plasmid for infection with and without inhibitor XXI at 500 nM. The cells were harvested by using trypsin at 36 h after infection, collected, washed in 1 ml, and resuspended in fluorescence-activated cell sorter (FACS) buffer (0.5% bovine serum albumin [BSA] in phosphate-buffered saline [PBS], pH 7.4; 200 l). GFP expression by the cells was analyzed by flow cytometry with a Becton Dickinson FACSCalibur. Data analysis was performed using CellQuest software (Becton Dickinson Immunocytometry System). In vivo analysis of HPV infection. Female BALB/c mice (6 to 8 weeks old) were purchased from the National Cancer Institute (Frederick, MD) and kept in the animal facility of the Johns Hopkins University (Baltimore, MD). Mouse experiments were approved by the Johns Hopkins University Animal Care and Use Committee. Mice received 3 mg of medroxyprogesterone (Depo-provera; Pfizer) diluted in 100 l of sterile PBS in a subcutaneous injection 4 days prior to HPV16 pseudovirus challenge. The pseudovirus inoculum was a 20-l dose composed of purified HPV16 pseudovirus carrying the luciferase reporter gene with a titer of ⬃5 ⫻ 109 IU/ml mixed in 2% carboxymethyl cellulose (CMC) (Sigma C5013). The virus was delivered in two doses. Half was deposited into mouse vagina by using an M50 positive-displacement pipette (Gilson). A cytobrush cell collector was inserted in the vagina and twirled clockwise and counter-clockwise 10 times, and the remaining 10 l was introduced in the presence/absence of various amounts of compound XXI (0 and 100 nM, 1 M, and 10 M) or 1% carrageenan. Three days later, the mice were anesthetized, luciferin (40 l at concentration of 7 mg/ml) was deposited intravaginally, and their images were acquired for 10 min using Xenogen IVIS 200. Immunofluorescence studies. Pseudovirions that contained packaged bromodeoxyuridine (BrdU)-labeled plasmid were prepared by addition of 20 M BrdU to the 293TT producer cells at 6 h posttransfection, as previously described (13). For microscopy cells were seeded onto glass coverslips in a 24-well plate at a density of 1 ⫻ 105 cells/well. Pseudoviruses were added for the time indicated in the text. BrdU was detected with BrdU Labeling and Detection Kit I (Roche) according to the manufacturer’s directions. Promyelocytic leukemia protein (PML) was detected with a rabbit polyclonal antiserum (Chemicon). The mouse monoclonal anti-Lamp-1 (clone H4A3) was purchased from the Developmental Studies Hybridoma Bank, Iowa City, IA. Detection of intact internalized capsids was performed with a polyclonal antiserum raised against HPV16 VLPs. Detection of an internal, linear L1 epitope was performed with the L1-7 monoclonal antibody, a kind gift from Martin Sapp. To detect L2 following uncoating, pseudovirions were made containing a carboxyl hemagglutinin (HA)-tagged L2 using the plasmid p16LlwCHA. The HA tag was detected with an anti-HA monoclonal antibody (Covance) as previously described (13). All images were acquired with a Zeiss LSM 510 confocal system interfaced with a Zeiss Axiovert 100 M microscope. Images were collated with Adobe Photoshop software. To delineate cellular peripheries, the fluorescence levels were increased in Adobe Photoshop. The outlines were traced in Adobe Photoshop, and the fluorescence levels were subsequently returned to the original levels.
RESULTS ␥-Secretase inhibitors block papillomavirus infection. To identify the possible relevance of ␥ secretase to HPV16 infection, we tested known inhibitors (Table 1) for their ability to prevent the delivery of the secreted alkaline phosphatase (SEAP) reporter gene to 293TT cells by HPV16 pseudovirions. The ␥-secretase inhibitors significantly reduced the delivery of SEAP to 293TT cells by HPV16 pseudovirions with IC50s similar to those previously described for their inhibition of ␥ secretase. However, this inhibition was also associated with visible toxicity for several inhibitors in the M range (inhibitors 565755, 565760, 565762, 565763, and 565766) (Table 1), suggesting possible nonspecific effects on infection. Therefore, we utilized the MTS-based assay to measure cell viability in the presence of titrated ␥-secretase inhibitors. Notably, the highest-activity inhibitor, compound XXI, was able to inhibit HPV infection with an IC50 that is 106-fold lower than the concentration associated with cellular toxicity (Table 1). XXI inhib-
HPV INFECTION REQUIRES ␥ SECRETASE
VOL. 84, 2010
10663
TABLE 1. Inhibitors tested in this study IC50 for:b
Targeta
Inhibitor name(s)
Target
XXI (关565790兴 (S,S)- 2-关2-(3,5-difluorophenyl)-acetylamino兴-N(1-methyl-2-oxo-5-phenyl-2,3-dihydro-1Hbenzo关e兴关1,4兴diazepin-3-yl)-propionamide) DAPT (N-关(3,5-difluorophenyl)acetyl兴-l-alanyl-2phenyl兴glycine-1,1-dimethylethyl ester) 565755 565760 (N-benzyloxycarbonyl-Leu-leucinal) 565762 (N-benzyloxycarbonyl-Leu-phenylalaninal) 565763 关1-(S)-endo-N-(1,3,3)-trimethylbicyclo关2,2,1兴hept-2-yl)4-fluorophenyl sulfonamide)兴 565766 (N-tert-butyloxycarbonyl-Gly-Val-valinal) TAPI-0 Batimastat Marimastat a b
Cell toxicity
HPV16
␥ Secretase
300 pM
250 M
300 pM
␥ Secretase
200 nM
ND
125 nM
␥ ␥ ␥ ␥
Secretase Secretase Secretase Secretase
13 M 35 M 5 M 1.8 M
125 M 125 M 31 M 7 M
3 M 15 M 3 M 3 M
␥ Secretase TACE, MMPs MMP MMP
3 M 8.8 nM, 0.2–68 nM 3–20 nM 3–13 nM
250 M ND ND ND
125 M 125 M 31 M 62 M
TACE, TNF-␣-converting enzyme; MMP, matrix metalloprotease. ND, not done.
ited the delivery of the encapsidated SEAP-encoding plasmid to 293TT cells by HPV16 pseudovirions with an IC50 of 300 pM, consistent the 300-pM IC50 described for its activity against ␥ secretase. Importantly, inhibition of HPV16 pseudoviral infection by XXI occurred without visible toxicity or measurable loss of cell viability. A second inhibitor, DAPT, inhibited HPV16 infection with an IC50 of 125 nM in the absence of visible toxicity (Table 1). Furthermore XXI failed to inhibit expression of SEAP when the reporter construct was directly transfected into 293TT cells, suggesting that XXI blocks infection rather than reporter expression. To further exclude nonspecific effects relating to the SEAP reporter, we tested the impact of XXI on infection of 293TT cells by HPV16 pseudovirions carrying a GFP reporter. Analysis by flow cytometry revealed that XXI potently inhibits the expression of the GFP reporter to 293TT cells by HPV16 pseudovirions (Fig. 1). Notably, the GFP reporter is driven by a different promoter to the SEAP reporter. To generalize this observation, we tested the ability of XXI to inhibit infection by pseudovirions derived from other papillomavirus genotypes. XXI inhibited infection by all of the diverse papillomavirus types tested, including HPV16, HPV18 (Fig. 1) (with 97% and 98% inhibition, respectively), HPV31, HPV45, and HPV58 with IC50s in the range of 130 pM to 1 nM (Table 2). It is noteworthy that XXI inhibited infection by HPV31 since its uptake pathway shows some differences from that of HPV16 (4, 34, 47, 48). Likewise, XXI also inhibited skin-tropic papillomavirus types (HPV5) and also papillomaviruses that infect other species (cottontail rabbit papillomavirus [CRPV] and bovine papillomavirus type 1 [BPV1]). The ability of XXI to inhibit HPV16 infection of other cell types was also examined. XXI blocked HPV16 pseudovirion infection of A549 cells (data not shown) as well as the spontaneously immortalized keratinocyte line, HaCaT, suggesting that the inhibitor is not acting upon T antigen (T-Ag) to block infection (Fig. 1). Substrates of ␥ secretase are typically type I membrane proteins that multimerize and undergo a prerequisite precleavage of the majority of the ectodomain by sheddases leaving a residual of ⬎30 amino acids that is recognized by NCT (46, 49).
Sheddases responsible for the precleavage of ␥-secretase substrates include the disintegrin and metalloprotease (ADAM) family members 9, 10, and 17 as well as aspartyl proteases like BACE1 that constitute the ␣- and -site APP-cleaving enzymes, respectively (1). None of the three ␣-secretase inhibitors tested, TAPI-0, batimastat, or marimastat, impacted the delivery of the SEAP reporter until well above their IC50 value for their known targets and until toxicity was visible in the 293TT cells, suggesting that these effects were nonspecific (Table 1). In contrast, an inhibitor of furin activity, decanoyl-RVKR-chloromethylketone (CMK), is a potent inhibitor of HPV16 infection, as previously reported (38). ␥ Secretase is not required for infection by polyomaviruses BKV or MCV or by adenovirus. It is possible that ␥-secretase activity is also required for infection by other structurally related nonenveloped viruses, such as the polyomaviruses BKV and MCV or adenovirus. To address this possibility, we examined the impact of XXI upon infection of A549 (data not shown) or 293TT cells by pseudoviruses derived from BKV and MCV polyomaviruses and adenovirus carrying a GFP reporter (Fig. 2). XXI consistently inhibited infection by HPV pseudovirions but not by the BKV or MCV pseudovirions or adenovirus (Fig. 2). This suggests that XXI is blocking a pathway required by the papillomaviruses but not by adenovirus or these polyomaviruses and that XXI is unlikely to prevent HPV infection via nonspecific, global effects on endocytic trafficking. Presenilin-1 and nicastrin are required for HPV infection. To further exclude the possibility that off-target drug effects are responsible for the ability of XXI to prevent HPV infection, we sought a genetic approach to validate the requirement for ␥-secretase activity. Presenilin (PS), an aspartyl protease, is the catalytic subunit of ␥-secretase activity, and its major isoform is encoded by PS1. Nicastrin (NCT) is a critical component of ␥-secretase activity that is required for substrate recognition. Some substrates of ␥ secretase are precleaved proximal to the membrane by sheddases such as the -site APP-cleaving enzyme, encoded by BACE1, to facilitate their recognition by NCT. Therefore, we examined whether HPV pseudovirions carrying a GFP reporter can infect PS1⫺/⫺, NCT⫺/⫺,or BACE1⫺/⫺ murine embryo fibroblasts. Fi-
10664
KARANAM ET AL.
J. VIROL.
FIG. 1. ␥-Secretase inhibitor XXI blocks HPV16 and HPV18 infection of 293TT and HaCaT cells. (A) 293TT cells (untreated in A1) were incubated with HPV16 pseudovirus containing a GFP reporter either without (A2) or with (A3) 500 nM XXI. (B) As for panel A, but using HPV18 pseudovirions. (C) As for panel A, but using HaCaT cells. (D) 293TT cells (untreated in D1) were incubated with HPV16 pseudovirus containing a GFP reporter either without (D2) or with the L2-specific and neutralizing monoclonal antibody RG-1 (D3).
broblasts derived from either PS1⫺/⫺ or NCT⫺/⫺ mice were not susceptible to HPV infection, whereas BACE1⫺/⫺ cells exhibited robust HPV pseudovirion infectivity (Fig. 3). XXI protects mice against vaginal challenge with HPV16. Entry and infection of cultured cell lines may not mimic all of the events of HPV infection in vivo. To examine whether ␥-secretase activity is also required for HPV infection in vivo, we used a murine challenge model in which HPV16 pseudovi-
TABLE 2. XXI inhibits infection of 293TT cells by pseudovirions derived from by diverse papillomavirus types Virus
HPV16........................................................................................ CRPV ......................................................................................... HPV31........................................................................................ HPV45........................................................................................ BPV1 .......................................................................................... HPV58........................................................................................ HPV5..........................................................................................
IC50 of XXI
300 pM 130 pM 500 pM 130 pM 1 nM 500 pM 1 nM
rus carrying a luciferase reporter was delivered to the vaginal vault alone or with titrated doses of XXI. A similar approach has demonstrated that carrageenan, when administered with HPV16 at the time of challenge, potently inhibits infection (19, 39). Mice in which compound XXI (10 M) was coadministered exhibited similarly strong protection against HPV16 pseudoviral challenge as with carrageenan (Fig. 4). XXI blocks genome localization at ND10 but not capsid uptake or disassembly. We sought to determine at which step during the infectious process ␥-secretase activity is required. There was no obvious effect upon the efficiency of capsid binding and endocytic uptake in the presence of XXI, as determined by immunofluorescence analysis with antibody reactive with the virion surface (data not shown). Upon HPV16 pseudovirion uptake, disassembly of the capsid reveals otherwise buried epitopes, including one recognized by monoclonal antibody 33L1-7 that is revealed only as virions disassemble in a late endosome/lysosome compartment during infection (2). Treatment with XXI did not impact the uptake of
VOL. 84, 2010
HPV INFECTION REQUIRES ␥ SECRETASE
10665
FIG. 2. XXI inhibits infection by HPV but not other small DNA tumor viruses. 293TT cells were incubated with HPV16, BKV, or MCV pseudoviruses or with recombinant adenovirus containing the GFP reporter in the presence or absence of XXI (500 nM).
HPV16 pseudovirions by HeLa cells and their display of the 33L1-7 epitope in perinuclear vesicles at 24 h postinfection (Fig. 5). This suggests that the uptake of HPV16 pseudovirions and subsequent disassembly of L1 capsid structures are not dependent upon ␥-secretase activity. In contrast, exposure of this epitope is prevented by cyclophilin inhibition which likely acts earlier during the endocytic process (2). It is possible that treatment with XXI impacts the trafficking of HPV16 pseudovirions during infection. The dependence of HPV16 infection upon endosome acidification and escape from a Lamp-1-positive compartment has been previously documented (21, 42). Therefore, we compared the localization of HPV16 pseudovirions with Lamp-1 at 20 h postinfection in the presence or absence of XXI by immunofluorescence staining with L1 VLP-specific antiserum (Fig. 6). No significant difference in the uptake of HPV16 pseudovirions into perinuclear vesicles containing Lamp-1 was noted in the presence of XXI, suggesting that this ␥-secretase inhibitor does not block infection by rerouting the particles to another vesicular compartment or otherwise noninfectious pathway. Disassembly of the capsid during infection reveals the buried carboxy terminus of L2 and then releases L2 and the nucleohistone core. By an unknown process, the released pseudoviral DNA, accompanied by L2, exits the vesicular structures, traffics into the nucleus, and accumulates adjacent to PML in ND10 (13). We sought to examine the role of ␥-secretase activity in the delivery of the viral DNA to the nucleus utilizing a previ-
ously described uncoating assay (13). This assay established the procedure to monitor the exposure of the encapsidated genome or an internal L2 epitope based on accessibility to antibody staining. HeLa cells were incubated with HPV16 pseudovirions containing BrdU-labeled plasmid, and their uptake and disassembly followed either without treatment or in the presence of XXI. Immunofluorescence staining with an anti-BrdU monoclonal antibody demonstrated that, as previously described, the BrdU-labeled DNA encapsidated by HPV16 pseudovirus was found both in perinuclear vesicles and at ND10 at 24 h postinfection (Fig. 7). However, in the presence of XXI only the nonnuclear staining was detected. To ensure that we were not observing a simple delay in kinetics of nuclear delivery, we examined the localization of the genome at a later time point (40 h). At this time the genome from the untreated virus was very strongly localized at ND10, as indicated by colocalization with PML. Conversely, in the cells treated with the ␥-secretase inhibitor XXI, the genome was nearly undetectable in the endosomal compartment, and no detectable intranuclear signal was found (Fig. 7), suggesting that ␥-secretase activity is required for entry of the viral DNA into the nucleus. We also examined the localization of L2 utilizing a C-terminally HA-tagged L2 as described in previous work (13). At the 40-h postinfection time point, the detection of the HA-tagged L2 revealed some perinuclear vesicles containing uncoated pseudovirions, but the majority of staining was found within the nucleus (Fig. 8A to C). In contrast, when the infection was
10666
KARANAM ET AL.
J. VIROL.
performed in the presence of XXI, the staining for the HA tag was restricted to the perinuclear vesicles (Fig. 8D to F). Therefore, inhibition of ␥-secretase activity prevented the entry of L2 into the nucleus during HPV16 infection. DISCUSSION
FIG. 3. HPV infection requires ␥-secretase components PS1 and NCT but not -secretase (BACE1). Wild-type (⫹/⫹) or deficient (⫺/⫺) cell lines for the components of ␥ secretase, nicastrin (B panels) and presenilin 1 (C panels), or for BACE1 (D panels) were incubated with HPV16 pseudovirus containing GFP reporter plasmid (A2 to D2) or buffer (A1 to D1).
Inhibitor and genetic analyses show that ␥-secretase activity is necessary for papillomavirus infection in vitro and in vivo. However, ␥ secretase is not required for either adenovirus or pseudovirions of the polyomaviruses BK and MCV to infect cells, suggesting that this mechanism is not conserved and that the ␥-secretase inhibitors are not globally disrupting vesicular traffic. Inhibition of ␥ secretase does not impact the uptake or the disassembly of papillomavirus in the perinuclear late endosome/lysosomal compartment to reveal its nucleohistone core and the C terminus of L2. However, the ␥-secretase inhibitor XXI appears to prevent both the nucleohistone core and associated L2 from reaching the nucleus. XXI does not impact the delivery and expression of reporter constructs by liposome-mediated transfection, suggesting that the block of infection does not reflect either inhibition of DNA entry into the nucleus from the cytoplasm or reporter expression. It is unlikely that a component of ␥ secretase is a receptor/cellular entry factor because knockout of either PS1 or NCT eliminates infectivity. Rather, the data suggest that ␥-secretase activity might be required to cleave a cellular entry factor or papillomavirus capsid protein and that this process is necessary for the viral genome to exit the Lamp-1-positive late endosome/lyso-
FIG. 4. XXI inhibits HPV16 infection of the murine vaginal tract. XXI at 0 (A1), 100 nM (A2), 1 M (A3), or 10 M (A4) or 1% carrageenan (A5) was applied intravaginally in each mouse (5 per group) along with 5 ⫻ 109 HPV16 pseudovirions in 20 l of 2% carboxymethylcellulose. Three days later, the mice were anesthetized, luciferin (10 l at 7 mg/ml) was deposited intravaginally, and the images were acquired for 10 min with a Xenogen IVIS 200. Equally sized areas encompassing the vaginal site in control mice were analyzed using Living Image, version 2.20, software (A1 to A5). The bioluminescence data for each group are plotted in panel B.
VOL. 84, 2010
HPV INFECTION REQUIRES ␥ SECRETASE
10667
FIG. 5. XXI does not prevent the uptake of HPV16 and its disassembly to reveal an internal L1 epitope. Pseudovirions were added to cells for 17 h in the absence (A) or presence (B) of XXI. Disassembled L1 was detected with the monoclonal antibody L1-7 following methanol fixation, and staining in the absence of pseudovirions is shown in panel C.
somal compartment and reach the ND10 in the nucleus. Syntaxin-18 has been implicated as cellular entry factor for papillomavirus and, intriguingly, several other syntaxins have been shown to associate with ␥ secretase (3, 50, 51). Syndecans are important in HPV binding and uptake and are also ␥-secretase substrates (44, 45). The perinuclear BrdU signal was noticeably reduced at the 40-h time point in the presence of XXI. However, the L2-
associated HA staining was not likewise reduced. This result suggests two possibilities. Perhaps, in the absence of ␥-secretase activity, both L2 and viral DNA are retained in the late endosome/lysosome compartment, wherein the genome is subject to degradation solely by nucleases. However, previous studies have shown that prevention of endosome escape by either furin or alpha defensin inhibition does not result in any noticeable acceleration of genome degradation (5, 38). A sec-
FIG. 6. XXI does not prevent the uptake of HPV16 into perinuclear vesicles containing Lamp-1. Pseudovirions were added to cells for 20 h in the absence (A to C) or presence (D to F) of XXI. The cells were immunostained with a rabbit polyclonal antibody to HPV16 L1 VLP and Alexa Fluor 594-coupled donkey anti-rabbit IgG (A and D) and a monoclonal anti-Lamp-1 and Alexa Fluor 488-coupled donkey anti-mouse IgG (B and E). The merged signals are shown in panels C and F.
10668
KARANAM ET AL.
J. VIROL.
FIG. 7. XXI prevents delivery of reporter DNA by HPV16 to the nucleus. HeLa cells were allowed to internalize pseudovirions that were assembled in the presence of BrdU. At either 24 or 40 h postentry, the cells were fixed and processed for BrdU detection coincident with detection of the PML protein. Panels A to C show the staining of untreated cells at 24 h. Panels D to F demonstrate the pattern of staining detected in cells that were treated with XXI during the 24-h incubation. In panels G to L, the cells were harvested at 40 h. Panels G to I are images of untreated cells, and panels J to L are images of XXI-treated cells. For all images, mouse anti-BrdU staining is shown in the green channel (A, D, G, and J), and rabbit anti-PML is shown in the red channel (B, E, H, and K). The merges are shown in the third column (C, F, I, and J). The outlines of the cells are delineated in panels C, F, I, and L.
VOL. 84, 2010
HPV INFECTION REQUIRES ␥ SECRETASE
10669
FIG. 8. Upon infection with HPV16 pseudovirus, XXI prevents the uptake of HPV16 L2 into the PML-containing subnuclear domain ND10 but not the display of the C terminus of L2 in perinuclear vesicles. Pseudovirions prepared with a carboxyl HA-tagged L2 using the plasmid p16LlwCHA were added to cells for 40 h in the absence (A to C) or presence (D to F) of XXI. The cells were immunostained with a monoclonal antibody to the HA tag and Alexa Fluor 488-coupled donkey anti-mouse IgG (A and D) and with rabbit polyclonal antibody to PML and Alexa Fluor 594-coupled donkey anti-rabbit IgG (B and E), and the merged signals are shown in panels C and F, respectively.
ond possibility is that an L2-genome complex is competent to escape from the endosome but is restricted in its ability to enter the nucleus. The continued detection of vesicular L2 in the presence of XXI could reflect uncoated L2 that is not associated with this proposed L2-genome complex as the stoichiometry is unknown. It is also possible that a papillomavirus capsid protein is the target of ␥ secretase. We note that L2 is required for the encapsidated DNA to reach the nucleus but not for uptake of virions. L2 contains a putative DNA in vitro binding domain (17, 40) and traffics to ND10 when overexpressed alone in cells (13). The substrates of ␥ secretase are typically precut with a proprotein convertase at a position less than 30 residues from the transmembrane domain to reveal a free amine to facilitate NCT recognition (1). Interestingly, L2 is cut with the proprotein convertase furin, and this is a critical event for infection (38). The cut site(s) for ␥ secretase are within transmembrane domains. No such transmembrane domain has been demonstrated in either L1 or L2. However, sequences at the C terminus of L2 have been demonstrated as membrane destabilizing and required for infection (21). Furthermore, transmembrane domain prediction software suggests that a highly conserved transmembrane domain might reside between residues 45 to 67 of HPV16 L2 (22), with equivalent regions in L2 sequences of all other types examined, but none in L1 or VP2 of the polyomaviruses MCV and BKV. Nevertheless, to date, we have failed to detect a consistent cleavage
of L2 or L1 upon infection of HaCaT cells with HPV16 pseudovirions that could be blocked by ␥-secretase (or furin) inhibitors. Clearly, further studies are needed to clarify the mechanistic role of ␥ secretase in papillomavirus infection. ACKNOWLEDGMENTS Funding for this study was provided by the National Institutes of Health, National Cancer Institute (grants R01CA118790 and R01 CA133749) and SPORE in Cervical Cancer (grant P50 CA098252 to R.B.S.R.). We thank Gary Ketner (Johns Hopkins University, Baltimore, MD) and Martin Sapp (LSUHSC, Shreveport, LA) for reagents. R.B.S.R. has served as a paid consultant of Merck & Co, Inc., and Knobbe Martens Olson & Bear LLC. R.B.S.R. has received unrestricted educational grant funding from GlaxoSmithKline. R.B.S.R. is a coinventor on L2 patents licensed to Shantha Biotechnics, Ltd., PaxVax, Inc., and Acambis, Inc., and to GlaxoSmithKline. The terms of these arrangements are being managed by Johns Hopkins University in accordance with its conflict of interest policies. REFERENCES 1. Beel, A. J., and C. R. Sanders. 2008. Substrate specificity of gamma-secretase and other intramembrane proteases. Cell Mol. Life Sci. 65:1311–1334. 2. Bienkowska-Haba, M., H. D. Patel, and M. Sapp. 2009. Target cell cyclophilins facilitate human papillomavirus type 16 infection. PLoS Pathog. 5:e1000524. 3. Bossis, I., R. B. Roden, R. Gambhira, R. Yang, M. Tagaya, P. M. Howley, and P. I. Meneses. 2005. Interaction of tSNARE syntaxin 18 with the papillomavirus minor capsid protein mediates infection. J. Virol. 79:6723–6731. 4. Bousarghin, L., A. Touze, P. Y. Sizaret, and P. Coursaget. 2003. Human papillomavirus types 16, 31, and 58 use different endocytosis pathways to enter cells. J. Virol. 77:3846–3850.
10670
KARANAM ET AL.
5. Buck, C. B., P. M. Day, C. D. Thompson, J. Lubkowski, W. Lu, D. R. Lowy, and J. T. Schiller. 2006. Human alpha-defensins block papillomavirus infection. Proc. Natl. Acad. Sci. U. S. A. 103:1516–1521. 6. Buck, C. B., D. V. Pastrana, D. R. Lowy, and J. T. Schiller. 2004. Efficient intracellular assembly of papillomaviral vectors. J. Virol. 78:751–757. 7. Cai, H., Y. Wang, D. McCarthy, H. Wen, D. R. Borchelt, D. L. Price, and P. C. Wong. 2001. BACE1 is the major -secretase for generation of A peptides by neurons. Nat. Neurosci. 4:233–234. 8. Campos, S. K., and M. A. Ozbun. 2009. Two highly conserved cysteine residues in HPV16 L2 form an intramolecular disulfide bond and are critical for infectivity in human keratinocytes. PLoS One 4:e4463. 9. Christensen, N. D., N. M. Cladel, and C. A. Reed. 1995. Postattachment neutralization of papillomaviruses by monoclonal and polyclonal antibodies. Virology 207:136–142. 10. Culp, T. D., L. R. Budgeon, and N. D. Christensen. 2006. Human papillomaviruses bind a basal extracellular matrix component secreted by keratinocytes which is distinct from a membrane-associated receptor. Virology 347:147–159. 11. Culp, T. D., N. M. Cladel, K. Balogh, L. R. Budgeon, A. Mejia, and N. D. Christensen. 2006. Papillomavirus particles assembled in 293TT cells are infectious in vivo. J. Virol. 80:11381–11384. 12. Davis, J. A., S. Naruse, H. Chen, C. Eckman, S. Younkin, D. L. Price, D. R. Borchelt, S. S. Sisodia, and P. C. Wong. 1998. An Alzheimer’s disease-linked PS1 variant rescues the developmental abnormalities of PS1-deficient embryos. Neuron 20:603–609. 13. Day, P. M., C. C. Baker, D. R. Lowy, and J. T. Schiller. 2004. Establishment of papillomavirus infection is enhanced by promyelocytic leukemia protein (PML) expression. Proc. Natl. Acad. Sci. U. S. A. 101:14252–14257. 14. Day, P. M., D. R. Lowy, and J. T. Schiller. 2008. Heparan sulfate-independent cell binding and infection with furin-precleaved papillomavirus capsids. J. Virol. 82:12565–12568. 15. Day, P. M., D. R. Lowy, and J. T. Schiller. 2003. Papillomaviruses infect cells via a clathrin-dependent pathway. Virology 307:1–11. 16. Doorbar, J. 2005. The papillomavirus life cycle. J. Clin. Virol. 32(Suppl 1):S7–S15. 17. Fay, A., W. H. Yutzy IV, R. B. Roden, and J. Moroianu. 2004. The positively charged termini of L2 minor capsid protein required for bovine papillomavirus infection function separately in nuclear import and DNA binding. J. Virol. 78:13447–13454. 18. Jagu, S., B. Karanam, R. Gambhira, S. V. Chivukula, R. J. Chaganti, D. R. Lowy, J. T. Schiller, and R. B. Roden. 2009. Concatenated multitype L2 fusion proteins as candidate prophylactic pan-human papillomavirus vaccines. J. Natl. Cancer Inst. 101:782–792. 19. Johnson, K. M., R. C. Kines, J. N. Roberts, D. R. Lowy, J. T. Schiller, and P. M. Day. 2009. Role of heparan sulfate in attachment to and infection of the murine female genital tract by human papillomavirus. J. Virol. 83:2067– 2074. 20. Joyce, J. G., J. S. Tung, C. T. Przysiecki, J. C. Cook, E. D. Lehman, J. A. Sands, K. U. Jansen, and P. M. Keller. 1999. The L1 major capsid protein of human papillomavirus type 11 recombinant virus-like particles interacts with heparin and cell-surface glycosaminoglycans on human keratinocytes. J. Biol. Chem. 274:5810–5822. 21. Kamper, N., P. M. Day, T. Nowak, H. C. Selinka, L. Florin, J. Bolscher, L. Hilbig, J. T. Schiller, and M. Sapp. 2006. A membrane-destabilizing peptide in capsid protein L2 is required for egress of papillomavirus genomes from endosomes. J. Virol. 80:759–768. 22. Karanam, B., S. Jagu, W. K. Huh, and R. B. Roden. 2009. Developing vaccines against minor capsid antigen L2 to prevent papillomavirus infection. Immunol. Cell Biol. 87:287–299. 23. Kim, S. H., Y. I. Yin, Y. M. Li, and S. S. Sisodia. 2004. Evidence that assembly of an active gamma-secretase complex occurs in the early compartments of the secretory pathway. J. Biol. Chem. 279:48615–48619. 24. Kines, R. C., C. D. Thompson, D. R. Lowy, J. T. Schiller, and P. M. Day. 2009. The initial steps leading to papillomavirus infection occur on the basement membrane prior to cell surface binding. Proc. Natl. Acad. Sci. U. S. A. 106:20458–20463. 25. Kirnbauer, R., F. Booy, N. Cheng, D. R. Lowy, and J. T. Schiller. 1992. Papillomavirus L1 major capsid protein self-assembles into virus-like particles that are highly immunogenic. Proc. Natl. Acad. Sci. U. S. A. 89:12180– 12184. 26. Lah, J. J., and A. I. Levey. 2000. Endogenous presenilin-1 targets to endocytic rather than biosynthetic compartments. Mol. Cell Neurosci. 16:111– 126. 27. Lancaster, W. D., C. Olson, and W. Meinke. 1977. Bovine papilloma virus: presence of virus-specific DNA sequences in naturally occurring equine tumors. Proc. Natl. Acad. Sci. U. S. A. 74:524–528. 28. Li, T., G. Ma, H. Cai, D. L. Price, and P. C. Wong. 2003. Nicastrin is required for assembly of presenilin/gamma-secretase complexes to mediate Notch signaling and for processing and trafficking of beta-amyloid precursor protein in mammals. J. Neurosci. 23:3272–3277.
J. VIROL. 29. Olson, C., D. E. Gordon, M. G. Robl, and K. P. Lee. 1969. Oncogenicity of bovine papilloma virus. Arch. Environ. Health 19:827–837. 30. Parkin, D. M., and F. Bray. 2006. Chapter 2: The burden of HPV-related cancers. Vaccine 24(Suppl 3):S11–S25. 31. Pasternak, S. H., R. D. Bagshaw, M. Guiral, S. Zhang, C. A. Ackerley, B. J. Pak, J. W. Callahan, and D. J. Mahuran. 2003. Presenilin-1, nicastrin, amyloid precursor protein, and gamma-secretase activity are co-localized in the lysosomal membrane. J. Biol. Chem. 278:26687–26694. 32. Pastrana, D. V., C. B. Buck, Y. Y. Pang, C. D. Thompson, P. E. Castle, P. C. FitzGerald, S. Kruger Kjaer, D. R. Lowy, and J. T. Schiller. 2004. Reactivity of human sera in a sensitive, high-throughput pseudovirus-based papillomavirus neutralization assay for HPV16 and HPV18. Virology 321:205–216. 33. Pastrana, D. V., Y. L. Tolstov, J. C. Becker, P. S. Moore, Y. Chang, and C. B. Buck. 2009. Quantitation of human seroresponsiveness to Merkel cell polyomavirus. PLoS Pathog. 5:e1000578. 34. Patterson, N. A., J. L. Smith, and M. A. Ozbun. 2005. Human papillomavirus type 31b infection of human keratinocytes does not require heparan sulfate. J. Virol. 79:6838–6847. 35. Pyeon, D., P. F. Lambert, and P. Ahlquist. 2005. Production of infectious human papillomavirus independently of viral replication and epithelial cell differentiation. Proc. Natl. Acad. Sci. U. S. A. 102:9311–9316. 36. Pyeon, D., S. M. Pearce, S. M. Lank, P. Ahlquist, and P. F. Lambert. 2009. Establishment of human papillomavirus infection requires cell cycle progression. PLoS Pathog. 5:e1000318. 37. Ray, W. J., M. Yao, J. Mumm, E. H. Schroeter, P. Saftig, M. Wolfe, D. J. Selkoe, R. Kopan, and A. M. Goate. 1999. Cell surface presenilin-1 participates in the gamma-secretase-like proteolysis of Notch. J. Biol. Chem. 274: 36801–36807. 38. Richards, R. M., D. R. Lowy, J. T. Schiller, and P. M. Day. 2006. Cleavage of the papillomavirus minor capsid protein, L2, at a furin consensus site is necessary for infection. Proc. Natl. Acad. Sci. U. S. A. 103:1522–1527. 39. Roberts, J. N., C. B. Buck, C. D. Thompson, R. Kines, M. Bernardo, P. L. Choyke, D. R. Lowy, and J. T. Schiller. 2007. Genital transmission of HPV in a mouse model is potentiated by nonoxynol-9 and inhibited by carrageenan. Nat. Med. 13:857–861. 40. Roden, R. B., P. M. Day, B. K. Bronzo, W. H. t. Yutzy, Y. Yang, D. R. Lowy, and J. T. Schiller. 2001. Positively charged termini of the L2 minor capsid protein are necessary for papillomavirus infection. J. Virol. 75:10493–10497. 41. Roden, R. B., H. L. Greenstone, R. Kirnbauer, F. P. Booy, J. Jessie, D. R. Lowy, and J. T. Schiller. 1996. In vitro generation and type-specific neutralization of a human papillomavirus type 16 virion pseudotype. J. Virol. 70: 5875–5883. 42. Sapp, M., and M. Bienkowska-Haba. 2009. Viral entry mechanisms: human papillomavirus and a long journey from extracellular matrix to the nucleus. FEBS J. 276:7206–7216. 43. Sapp, M., and P. M. Day. 2009. Structure, attachment and entry of polyomaand papillomaviruses. Virology 384:400–409. 44. Schulz, J. G., W. Annaert, J. Vandekerckhove, P. Zimmermann, B. De Strooper, and G. David. 2003. Syndecan 3 intramembrane proteolysis is presenilin/gamma-secretase-dependent and modulates cytosolic signaling. J. Biol. Chem. 278:48651–48657. 45. Shafti-Keramat, S., A. Handisurya, E. Kriehuber, G. Meneguzzi, K. Slupetzky, and R. Kirnbauer. 2003. Different heparan sulfate proteoglycans serve as cellular receptors for human papillomaviruses. J. Virol. 77:13125– 13135. 46. Shah, S., S. F. Lee, K. Tabuchi, Y. H. Hao, C. Yu, Q. LaPlant, H. Ball, C. E. Dann III, T. Sudhof, and G. Yu. 2005. Nicastrin functions as a gammasecretase-substrate receptor. Cell 122:435–447. 47. Smith, J. L., S. K. Campos, and M. A. Ozbun. 2007. Human papillomavirus type 31 uses a caveolin 1- and dynamin 2-mediated entry pathway for infection of human keratinocytes. J. Virol. 81:9922–9931. 48. Smith, J. L., S. K. Campos, A. Wandinger-Ness, and M. A. Ozbun. 2008. Caveolin-1-dependent infectious entry of human papillomavirus type 31 in human keratinocytes proceeds to the endosomal pathway for pH-dependent uncoating. J. Virol. 82:9505–9512. 49. Struhl, G., and A. Adachi. 2000. Requirements for presenilin-dependent cleavage of notch and other transmembrane proteins. Mol. Cell 6:625–636. 50. Suga, K., A. Saito, T. Tomiyama, H. Mori, and K. Akagawa. 2005. Syntaxin 5 interacts specifically with presenilin holoproteins and affects processing of APP in neuronal cells. J. Neurochem. 94:425–439. 51. Vetrivel, K. S., H. Cheng, S. H. Kim, Y. Chen, N. Y. Barnes, A. T. Parent, S. S. Sisodia, and G. Thinakaran. 2005. Spatial segregation of gammasecretase and substrates in distinct membrane domains. J. Biol. Chem. 280: 25892–25900. 52. Walboomers, J. M., M. V. Jacobs, M. M. Manos, F. X. Bosch, J. A. Kummer, K. V. Shah, P. J. Snijders, J. Peto, C. J. Meijer, and N. Mun ˜ oz. 1999. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J. Pathol. 189:12–19. 53. zur Hausen, H. 2002. Papillomaviruses and cancer: from basic studies to clinical application. Nat. Rev. Cancer 2:342–350.
