Background
Human Papillomaviruses (HPVs) are non-enveloped DNA viruses that can infect the skin and mucous membranes. HPVs are known to cause cutaneous, cervical, and respiratory warts and lesions [
5‐
7]. The capsid of HPVs is made of two virally encoded structural proteins L1 and L2 [
8‐
10]. The major capsid protein L1 is primarily involved in attachment of the virus to the plasma membrane, while the minor capsid protein L2 functions in viral genome trafficking and encapsidation [
11‐
15].
The infectious process begins via virion attachment to the cell surface through breaks in the skin. Although the virion-cell binding process is still unclear it is thought to occur by initial binding of the L1 protein on the virion capsid to heparan sulfate (a cell surface proteoglycan), followed by binding to a secondary receptor, putatively an integrin complex [
16‐
18]. α6β4 has been shown to be able to mediate cell binding in studies showing that antibodies against α6 could block virion binding to the epithelial cells CV-1 and HaCaT keratinocytes [
19]. However, α6β4 integrin may not be a necessary requirement for infection since studies also indicate that some PVs can infect cells such as BO-SV keratinocytes that lack this complex [
20]. After attachment to the cell surface the HPV16 virion is internalized via a mechanism that begins with clathrin mediated endocytosis [
2,
21,
22]. N-terminus cleavage of L2 by furin, a calcium dependent serine endoprotease found at the plasma membrane, Golgi and endosomes, has been suggested to be required for infection [
23,
24]. Our data suggests that after trafficking to the endosome, the reporter-virions may follow either an infectious route or a noninfectious route ([
3,
4]). In the infectious route, reporter-virions are moved to the caveolin-1 intracellular sorting pathway. This caveolin-1 pathway was shown to be necessary for infection, as infection is blocked in cells where caveolin-1 protein levels were reduced using siRNA against caveolin-1 ([
3,
4]). After entering the caveosome, the virion was shown to traffic in an L2-mediated event to a region where it colocalized with the endoplasmic reticulum (ER) t-SNARE syntaxin 18 and the ER chaperone calnexin and ERp29 ([
3,
4,
11,
14]). The non-infectious pathway results in trafficking from the endosome to the lysosome where reporter-virions may be processed for degradation by the cell. This latter pathway was shown using a non-infectious L2 mutant virus and neutralizing antibodies [
3]. It has been shown that ammonium chloride (NH
4Cl) blocks infection of Bovine Papillomavirus Type 1 (BPV1), a PV with similar kinetics to HPV16 [
2]. NH
4Cl neutralized the acidic endo-lysosome compartments suggesting that pH was responsible for PV capsid conformational changes leading to viral genome release. However, our data presented in this manuscript suggested that ammonium chloride blocked infection by preventing the movement of viral particles from the early endosome to the caveosome as was also shown for JC virus [
1]. In this manuscript we show that cysteine proteases and not pH may be responsible for changes leading to infection.
Cysteine proteases function as intracellular and extracellular molecules [
25]. The cysteine protease cathepsin B is associated with caveolae. Caveolae are defined as small invaginations of the plasma membrane associated with lipid rafts that contain caveolin-1 [
26,
27]. Similar to caveolae, endo-lysosomal compartments within cells contain cathepsin B but in addition have cathepsin L. Both of these cathepsins are zymogens (pro-forms) that are cleaved into their active form [
28,
29]. The exact mechanism of activation is not well understood however, activation of pro-cathepsin B may occur by S100A10, a protein found in caveolae, while activation of pro-cathepsin L may occur by heparan sulfate, a possible receptor for PV [
27,
30]. In addition to caveolae and endosomes, cathepsins have been found to be associated with lipid rafts, suggesting that cathepsins may be needed upon internalization to break apart matrices on viral surfaces [
25].
Cathepsins B and L have been implicated in the mechanism of binding, entry and disassembly of several enveloped viruses. In the case of binding, treating Ebolavirus reporter-virions with cathepsin L enhanced infectivity by cleaving and removing a highly glycosylated mucin domain in the Ebolavirus glycoprotein and resulted in increased binding suggesting that cathepsins are indeed present on the cell membrane [
31]. Fusion of enveloped viruses such as Nipah, Hendra, SARS Coronavirus and Murine Coronavirus Mouse Hepatitis Virus has been shown to be dependent on cathepsins B and L. In Nipah virus, both cathepsin B and L were shown to cleave a membrane fusion protein required for virus-cell and cell-cell membrane fusion. Cleavage of the viral membrane fusion protein into the correct size was necessary for maturation into a fusogenic form. Cathepsin B was shown to cleave the fusion protein in a cell-free system into two fragments but the smaller of these fragments migrated slower than fragments produced during the cleavage that occurs in infection, suggesting that the fusion protein was not cleaved at the correct size. Cathepsin L was able to cleave the fusion protein into fragments of the correct size in the cell-free system suggesting that although cathepsin B and L are catalytically similar, they may have distinct target/sequence specificity [
32]. Similar to Nipah virus, cathepsin L was involved in cleaving the Hendra virus fusion protein into an active heterodimer [
33]. In SARS Coronavirus infection, cathepsin L was needed to cleave the spike protein, one of four major structural proteins, into two subunits: one having a high binding affinity to the receptor and the other mediating fusion of viral and cellular membranes [
34]. The requirement of cathepsins for fusion was also shown in Murine Coronavirus Mouse Hepatitis Virus (MHV), where proteolysis by cathepsin B and L was necessary for cleavage of the MHV-2 spike protein [
35]. In addition, proteolysis by cathepsins was shown to be important for disassembly of Reovirus. Using mouse embryonic fibroblasts derived from cathepsin B or L deficient and wild type mice, studies show that Reovirus disassembly was prohibited in the absence of cathepsins B and L [
36].
In this manuscript, we show a role for cathepsin B that may be important for HPV16 infection. It is, to our knowledge, the first description of the role of cysteine proteases on a non-enveloped virion.
Discussion
Cathepsin proteases have been shown to be able to modify the binding and entry of many enveloped viruses, and thus influence infection efficiency [
31‐
36]. In this manuscript, we broadened the involvement of cathepsin proteases to the non-enveloped virus, HPV16. We showed that cathepsin proteases play a role in the infection of HPV16 into human embryonic 293 cells (HEK 293), and into mouse embryonic fibroblasts (MEFs).
Data previously obtained by our laboratory has demonstrated that BPV1 and HPV16 infection follow a 'non-classical' trafficking route post-clathrin-mediated endocytosis [
3,
4]. Our data demonstrated that after reporter-virions have endocytosed using clathrin-coated pits, the reporter-virions are found in an early endosome (EEA1 positive vesicle), and are sorted to a caveolin-1 positive organelle putatively the caveosome before colocalizing with endoplasmic reticulum (ER) marker. Caveosome were originally identified as a necessary component of SV-40 trafficking [
38] and subsequent work has demonstrated that reporter-virions seem to traffic through caveosomes on their way to the endoplasmic reticulum. In order for BPV1 and HPV16 to reach the caveosome, there would have to be cross-talk between endosomes and caveosomes. This cross-talk has now been described for JC virus [
1] and for HPV 31 [
39]. A commonly used technique that is used to determine the role of pH and endo-lysosomes in viral trafficking involves using the lysosome pH neutralizing chemical ammonium chloride (NH
4Cl) to prevent the acidification of vesicles, unfortunately this treatment also results in the loss of fusion of intracellular vesicles. In fact, it was shown that NH
4Cl prevented the movement of JC virus from endosome to caveosome.
The observed loss of viral infection using NH
4Cl posed two hypotheses: 1) that the PV infection was preventing the fusion of vesicles, and 2) that NH
4Cl was preventing the function of endo-lysosome proteases by preventing their conversion from "pro-inactive" form to "active" form. Our data shown in Figure
2 confirmed that indeed there was loss of movement of HPV 16 reporter-virions from endosomes to caveosome that could account for the loss of infection observed using NH
4Cl (shown in Fig
1). Regarding the role of endo-lysosome proteases, we focused on the cysteine proteases cathepsin B and L, two highly abundant proteases in the endo-lysosome compartments, and as mentioned above, cathepsin B and L have been previously shown to be involved in viral infections. Our data showed that broad cysteine protease inhibitors and specific cathepsin B or L inhibitors were able to decrease infection, thus suggesting that cysteine proteases were in part mediating HPV16 infection in 293 cells.
Furin protease has been shown to be necessary for viral infection by allowing the escape of the viral particle from an endosome (Richards RM, Lowy DR, Schiller JT, Day PM [
24]). Richards and colleagues theorized that furin allowed the escape of reporter-virions from an endosome as observed by the loss of endosome marker (EEA1) staining overlap with reporter-virions. Our data supports a loss of EEA1 overlap with reporter-virions but show that reporter-virions are moved to a caveolin-1 vesicle. It is unclear where Richards and colleagues propose viral particles escape to. Because furin cleavage was shown to occur after capsid conformation changes, we addressed if cathepsins B or L were playing a role in capsid structural changes that aided the furin cleavage event. To our surprise the pre-treatment of purified viral particles with cathepsin B, but not cathepsin L, was able to overcome the block of infection observed in the presence of furin inhibitor. The significance of this finding needs further work. It is possible that HPV16 utilizes cathepsin B as a "backup" mechanism for furin in order to establish infection.
In a recent study addressing the role of endosome proteases in the disassembly of HPV16 [
24], the authors negated the role of cathepsin B and L in the HPV16 infectious process. The differences between our studies and theirs may be due to the variation in inhibitors, cell lines, and quantity of viral particles used. In the paper by Richards and colleagues the cathepsin B inhibitor used was CA-074, a membrane impermeable inhibitor that would not have shown an effect on an intracellular process (we used both a permeable and non-permeable inhibitor); the cathepsin L inhibitor used was only described as "cathepsin L inhibitor" and no further conclusions can be drawn as to the specificity of the inhibitor [
24]. In addition, the observations described by Richards and colleagues were seen in the HPV 18 positive cervical carcinoma HeLa cells while our studies were performed in the adenovirus E1A transformed human embryonic kidney 293 cell line and in MEFs derived from cathepsin B-deficient mice and cathepsin L-deficient mice [
24]. HeLa cells were recently shown to be infected by a non-clathrin mediated endocytosis event, a finding that may also explain the differences in both studies [
40]. Finally the experiments performed in the paper by Richards and colleagues, used a minimum of 7.5 ng of PV while our experiments were carried out using less than 0.33 ng of HPV16 (1 ng of VLPs has 30 million particles); a difference that may also contribute to the differences in results.
Methods
Cells, antibodies, proteases and inhibitors
293 cells, a human embryonic kidney cell line (HEK), HaCaT cells, spontaneously immortalized human keratinocytes, and cathepsin B deficient (-/-), wild type (+/+), cathepsin L deficient (-/-) and wild type (+/+) mouse embryonic fibroblasts (MEFs) were grown in Dulbecco's Modified Eagle's medium (DMEM) supplemented with 10% Fetal Bovine Serum (DMEM-10), and 100 IU/mL penicillin/streptomycin. MEF cells were gifts from Dr. T. S. Dermody (Vanderbilt University School of Medicine, Nashville, TN). Goat anti-EEA1 (which recognizes early endosomes) antibody was purchased from Santa Cruz Biotechnology, Santa Cruz, CA. Rabbit anti-caveolin 1 antibody was purchased from Cell Signaling Technology (Danvers, MA). The monoclonal HPV16 L1 antibody, H16 V.5 was a generous gift from Dr. Neil Christensen (Penn State University, Hershey, PA). The following inhibitors were obtained from Calbiochem (Gibbstown, NJ) and used at the following non-toxic concentration: CA-074Me (6 μM), Z-FF-FMK an irreversible cell permeable cathepsin L inhibitor (10 μM), E64 (10 μM), E64-d (10 μM). Furin inhibitor DEC-RVKR-CMK was obtained from Biomol International (Plymouth Meeting, PA) and used at the non-toxic concentration of 60 μM. Ammonium chloride was used at a 20 mM concentration as per Day et al., (Sigma, St. Louis, MO) [
2]. Cathepsin B protease (Calbiochem) and cathepsin L protease (R&D Systems Minneapolis, MN) were used at a concentration of 37 pM, a lower concentration compared to Ebert et al., [
36].
Flow cytometry
Inhibitors were added to 293 cells and allowed to incubate on cells at 37°C with 5% CO2 for 2 hours prior to infection. After the 2 hr incubation period, the cells were placed on ice. HPV 16 reporter-virions containing a DsRed or GFP transgene were added to cells on ice for 2 hours to allow for binding. At 2 hours, inhibitors and unbound virus were removed by washing with DMEM-10 and replaced with 500 μl of warm DMEM-10 plus inhibitor. MEFs (not treated with inhibitors) were placed on ice for 2 hours in the presence of HPV16 reporter-virions to allow for binding. Unbound reporter-virions were removed by washing with DMEM-10 and replaced with 500 μl of warm DMEM-10. The cells were incubated at 37°C with 5% CO2 for 48 hours. Cells were harvested using trypsin. The cells were spun for 1 min at 16,100 × G, the pellet was washed 5× in 1× PBS and resuspended in 300 ul of 1× PBS. 10,000 cells were counted on a fluorescence activated cell sorter (FACS) and the number of DsRed or GFP positive cells was used to determine the percent of infected cells (FACS performed at RFUMS Flow cytometry core). All experiments were repeated using reporter-virions made from different preparations.
Immunofluorescence
Cells transfected on coverslips were washed 3× in 1× PBS and fixed in 4% paraformaldehyde for 20 min at 4°C. Paraformaldehyde was removed via 3 washes with 1× PBS. Cells were permeabilized with blocking buffer (0.2% fish skin gelatin (Sigma) and 0.2% Triton X-100 in PBS) for 5 min. The coverslips were washed 3× with 1× PBS and incubated with the appropriate primary antibody at 1:100 dilution in blocking buffer (1:25 dilution for LAMP1 antibody). Fluorescence labelled Alexa-flour donkey anti-mouse 488, goat anti-rabbit 594, chicken anti-goat 594, (Molecular Probes/Invitrogen, Eugene, OR) were used as secondary antibodies at 1:2,000 dilution in blocking buffer in a 30 minute incubation. Coverslips were incubated for 5 minutes with TOPRO-3 (Invitrogen, Carlsbad CA) at 1:1000 dilution for nuclear staining. The coverslips were mounted on glass slides using Prolong anti-fade mounting medium (Invitrogen). Fluorescence confocal microscopy and stereology (the quantification of the percentage of colocalization observed in the image, i.e. merged colors) were performed using an Olympus Fluoview 300 microscope, and analyzed with Fluoview and stereology software (Olympus, Melville, NY) at the microscopy core of Rosalind Franklin University of Medicine and Science (RFUMS) (North Chicago, IL). All images are shown with z stacks.
Cytotoxicity Assay
Cytotoxicity studies for the various inhibitors were carried out using the CellTiter-Glo Luminescent Cell Viability Assay Kit (Promega, Madison, WI). Cells were incubated with various concentrations of inhibitors for 48 hours and then the supernatant and trypsinized cells were collected. 100 μl of the harvested cell suspension was added to a well in a 96-well plate. The CellTiter-Glo substrate and buffer were combined, and 100 μl was added to each well containing sample. The reagent and cells are mixed for 5 min on a shaker at room temperature. The 96-well plate was then allowed to rest at room temperature for 10 minutes before being analyzed by a Bio-Tek Synergy HT Plate Reader using the KC4 V3.4 software (Bio-Tek, Winooski, VT). All samples were analyzed in triplicate.
Reporter-virion production and purification
Reporter-virions were made as described [
37]. In brief, 293TT cells were co-transfected with p16llwcha, a bicistronic HPV16 L1 and L2 plasmid and 8frb, the DsRed or 8fwb, the GFP cDNA containing packaging plasmid. Constructs and cells were gifts from Drs. Day and Schiller (National Cancer Institute, National Institute of Health, Bethesda, MD). Cells were harvested and lysed after 48 hours. Reporter-virions were allowed to mature at 37°C over night allowing for proper conformation of the capsid proteins. After a high salt extraction, reporter-virions were purified on an optiprep gradient (27%–39%) via ultracentrifugation. Titer of reporter-virions was determined by FACS for the percentage of DsRed or GFP positive cells 48 hours after infection. Tritium labelled reporter-virions were made with the addition of
3H 24 hours post transfection.
Binding of radioactive reporter-virions
Tritium labelled reporter-virions containing a DsRed transgene were added to 293 cells on ice for 2 hours to allow for binding without internalization. The unbound reporter-virions were removed by washing with 1× PBS. Cells were harvested in 30 μl 1× PBS, spotted on Whatmann paper and allowed to dry. The samples were fixed in 5% Trichloroacetic acid (TCA) for 20 minutes and precipitated in 95% ethanol for 20 minutes. The samples were analyzed with the LS 6500 Multipurpose Scintillation Counter (Beckman Coulter, Palatine, IL). All experiments were repeated using reporter-virions made from different preparations.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
SAD carried out the flow cytometric analysis, immunofluorescence analysis, radioactive studies, cytotoxicity analysis, and performed the statistical analysis. PIM and SAD participated in the design of the study and drafted the manuscript. All authors read and approved the final manuscript.