The nuclear retention signal of HPV16 L2 protein is essential for incoming viral genome to transverse the trans-Golgi network

Virology 458-459 (2014) 93–105

Contents lists available at ScienceDirect

Virology journal homepage: www.elsevier.com/locate/yviro

The nuclear retention signal of HPV16 L2 protein is essential for incoming viral genome to transverse the trans-Golgi network Stephen DiGiuseppe 1, Malgorzata Bienkowska-Haba 1, Lydia Hilbig, Martin Sapp n Department of Microbiology and Immunology, Center for Molecular and Tumor Virology, Feist-Weiller Cancer Center, LSU Health Shreveport, Shreveport, LA, USA

art ic l e i nf o

a b s t r a c t

Article history: Received 10 January 2014 Returned to author for revisions 12 February 2014 Accepted 17 April 2014

The Human papillomavirus (HPV) capsid is composed of the major and minor capsid proteins, L1 and L2, respectively. Infectious entry requires a complex series of conformational changes in both proteins that lead to uptake and allow uncoating to occur. During entry, the capsid is disassembled and host cyclophilins dissociate L1 protein from the L2/DNA complex. Herein, we describe a mutant HPV16 L2 protein (HPV16 L2-R302/5A) that traffics pseudogenome to the trans-Golgi network (TGN) but fails to egress. Our data provide further evidence that HPV16 traffics through the TGN and demonstrates that L2 is essential for TGN egress. Furthermore, we show that cyclophilin activity is required for the L2/DNA complex to be transported to the TGN which is accompanied by a reduced L1 protein levels. & 2014 Elsevier Inc. All rights reserved.

Keywords: HPV Entry L2 L1 Cyclophilin Trans-Golgi network

Introduction Human Papillomaviruses (HPV) are non-enveloped epitheliotropic viruses, which are associated with malignancies of the epithelium including cervical, penile, anogenital, and oropharyngeal carcinoma. The viral capsid is composed of 360 copies of the major capsid protein, L1, organized into 72 pentamers, which are also called capsomeres. Sixty capsomeres are pentavalent, i.e. have five nearest neighbors, whereas 12 capsomeres are hexavalent (Chen et al., 2000; Finch and Klug, 1965). Capsomeres are interconnected by invading Cterminal arms, which are stabilized by intercapsomeric disulfide bonds resulting in covalently linked L1 dimers and trimers (Li et al., 1998; Mejia et al., 2006; Sapp et al., 1998; Wolf et al., 2010). In addition, up to 72 copies of the minor capsid protein, L2, are present within the capsid, mostly hidden inside with only a short N-terminal portion accessible (Buck et al., 2008; Liu et al., 1997; Modis et al., 2002). The protein shell encapsidates a circular, double-stranded DNA viral genome of about 8000 bp, which is organized into chromatin (Favre et al., 1977). Recent advances in understanding the HPV16 internalization process was based on the development of systems to generate pseudovirions, which mimic authentic particles (Buck et al., 2004, 2005a) that also encapsidate a pseudogenome. Using these particles, n Correspondence to: Department of Microbiology and Immunology, 1501 Kings Highway, Shreveport, LA, 71130, USA. Tel.: þ1 318 675 5760. E-mail address: [email protected] (M. Sapp). 1 Contributed equally.

http://dx.doi.org/10.1016/j.virol.2014.04.024 0042-6822/& 2014 Elsevier Inc. All rights reserved.

it was demonstrated that HPV16 particles attach to the extracellular matrix via primary heparin sulfate proteoglycan receptors using L1 lysine residues 278 and 361 present on capsomeres (Knappe et al., 2007). Attachment triggers conformational changes in both capsid proteins allowing transfer to unknown secondary receptors and infectious internalization (Day et al., 2008; Richards et al., 2013; Selinka et al., 2007; Spoden et al., 2008). These processes are extremely slow and asynchronous, taking several hours to complete (Christensen et al., 1995; Giroglou et al., 2001; Smith et al., 2007). Infectious internalization of HPV16 occurs via a clathrin- and caveolae-independent pathway (Spoden et al., 2008, 2013; Schelhaas et al., 2012). Acidification of the endocytic compartment is essential for successful infection (Day et al., 2003; Selinka et al., 2002) and is accompanied or followed by uncoating and egress from endosomes (Kamper et al., 2006), and subsequent accumulation of incoming pseudogenomes at PML nuclear bodies (Day et al., 2004). Successful delivery of the pseudogenome to the nucleus depends on the activity of numerous host cell factors. One such factor, cell surface cyclophilin B, a chaperone peptidyl-prolyl cis/trans isomerase, facilitates exposure of a hidden cleavage site on the N-terminus of the L2 protein that is subsequently cleaved by furin convertase, both events that have been shown to be required for pseudogenome delivery (Bienkowska-Haba et al., 2009; Richards et al., 2006). Our lab has demonstrated that host cell cyclophilins (CyPs) are also required at a secondary, post-internalization step which mediates the dissociation of the L1 protein from the L2/DNA complex prior to nuclear entry (Bienkowska-Haba et al., 2012). The L2 protein mediates egress of the pseudogenome from endosomes (Kamper et al., 2006) and

94

S. DiGiuseppe et al. / Virology 458-459 (2014) 93–105

retrograde transport of pseudogenome along microtubules to the nucleus (Florin et al., 2006; Schneider et al., 2011). Recently, a genome-wide siRNA screening identified cellular factors involved in retrograde trafficking of HPV. Among these factors, components of the retromer complex and Rab GTPases were found to be required for retrograde transport of HPV16 pseudovirions to the trans-Golgi network (TGN) (Day et al., 2013; Lipovsky et al., 2013). It was demonstrated that HPV16 virions enter the TGN in a retromerdependent manner during infectious entry. Additionally, it was also shown that TGN localization is dependent on furin-cleaved L2 protein (Day et al., 2013). Following TGN egress, the L2 protein accompanies the viral genome to PML nuclear bodies (PML NBs) (Day et al., 2004). Evidence indicates that nuclear import of viral DNA requires nuclear envelope breakdown (Pyeon et al., 2009) suggesting independence of active nuclear import pathways despite the existence of redundant nuclear localization signals (NLS) in L2 (Becker et al., 2003). The L2 protein has two canonical nuclear localization signals on both the N- and C-terminal ends. Recently, a third element was identified located between amino acids 291–315, which we named the mNLS, that contain a stretch of arginine residues that were shown to mediate nuclear retention in vitro (Mamoor et al., 2012). To further investigate the role of the mNLS in nuclear delivery, our lab has generated NLS mutants with exchanges in bases within mNLS and deletions in the N- and C-terminal NLSs. Using transfection and overexpression, we show that neither NLS is required for nuclear translocation. Additionally, we confirm that the mNLS functions as a nuclear retention signal in vitro, results that are consistent with a previous report (Mamoor et al., 2012). In the context of infectious entry, we show that pseudovirions with mutations in the mNLS accumulate in the TGN. Our data presented herein provide further evidence that HPV genomes traffic through the TGN and confirm that CyPs are essential for transport to the TGN.

Results Nuclear translocation of HPV16 L2 protein in vivo Two sequence elements at the N- and C-terminus of HPV16 L2 protein termed nNLS and cNLS, respectively, have been extensively characterized as putative nuclear localization signals using a variety of in vitro techniques (Bordeaux et al., 2006; Darshan et al., 2004; Klucevsek et al., 2006; Mamoor et al., 2012; Sun et al., 1995). In order to identify NLSs that might be used in vivo during infection and after synthesis, we deleted these elements and tested the intracellular localization of mutant L2 following transfection of expression plasmids. The truncations were kept short to minimize potential effects on L2 protein folding. All transfected mutant L2 constructs were expressed in HeLa cells (Fig. 1F). As shown in Fig. 1A and G, deletion of both signals did not abrogate nuclear import of HPV16 L2 protein suggesting that neither NLS is essential for L2 nuclear translocation. We therefore concentrated on a third element, which was shown to influence intracellular localization of HPV6b (Sun et al., 1995) and HPV33 L2 (Becker et al., 2003). This region is highly conserved among members of the Papillomaviridae family Alphapapillomavirus genus (Fig. 1B). Located between HPV16 L2 amino acid residues 291 and 315, this region contains a number of conserved arginine residues. We introduced point mutations in this region in the context of full length and terminally truncated L2 protein focusing on but not restricted to basic amino acid residues. Replacement of arginine residues 297, 298, 302, and 305 for alanine affected nuclear localization when analyzed at 24 hours post-transfection (hptx) of full length and truncated L2. At this time, a significant percentage of mutant L2 protein was found in the cytoplasm. In addition, a varying fraction of mutant L2 localized to the nucleus in a punctate pattern co-localizing with PML protein (Fig. 1C). A quantification of the results is provided

in Fig. 1G. Additional deletions of the terminal nNLS and cNLS only had a minor impact on subcellular localization as shown for 16L2-13-455-R297/8-302/5A. These data suggest that the contribution of nNLS and cNLS to nuclear import is rather negligible under our conditions. The strongest effect was observed when all four arginine residues were replaced. However, pronounced reductions in nuclear import of L2 protein were also found for double mutants R297/8A and R302/5A (Fig. 1D and G). Exchange of arg-291, lys-309, arg-315, ser-304, gly-307, or ser-295 and thr-296 for alanine did not affect nuclear accumulation of L2 protein (data not shown). We also exchanged arginine residues at positions 295 and 298 of HPV18 L2, which are homologous to arg-302 and arg-305 of HPV16 L2, and found that these mutant proteins mainly localize to the cytoplasm as well at 24 hptx (Fig. 1E and G). Looking at earlier times posttransfection, we observed wild-type HPV16 L2 localize in the nucleus as early as 6 hptx and remained nuclear (Fig. 2A) whereas HPV16 mutant L2 protein localized in the nucleus at earlier times posttransfection (Fig. 2B), but was found to be relocated to the cytoplasm after 12 hptx. The cytoplasmic relocalization can be abrogated by inhibiting CRM1-dependent nuclear export with leptomycin B (LMB) treatment (Fig. 2C). Taken together, these results confirm previous observations by others (Mamoor et al., 2012) that mutations in this motif did not abrogate nuclear import but rather affected nuclear retention. Characterization of mutant pseudoviruses In order to investigate the role of this sequence element in context of infectious HPV entry, pseudoviruses harboring mNLS mutant L2 protein were generated using the 293TT production cell line (Buck et al., 2004). All tested mutant L2 proteins were efficiently incorporated into pseudoviruses, and all pseudoviruses encapsidated the pEGFP-C1 pseudogenome used for our studies with comparable efficiency (Fig. 3A). We determined the infectivity of the L2 mutants by infecting 293TT and HaCaT cells with mutant pseudovirus for 72 h at 37 1C and quantified GFP transcript by real-time PCR (Fig. 3B). The exchange of arginine 302 and 305 with alanine in the L2 protein rendered the pseudovirions non-infectious (Fig. 3B). Mutants 16L2-R291A, -S304A, -G307A, -K309A, and -R315A fully supported infection (data not shown). Similarly, mutant HPV18 pseudovirus harboring 18L2-R295/ 8A was also strongly defective (data not shown) suggesting that the importance of this region may be conserved among genital high-risk Papillomaviruses. We assessed the integrity of the capsid structure of intact virions by measuring reactivity with L1-specific antibodies by the enzyme-linked immunosorbent assay (ELISA). Pseudoviruses were bound for 1 h at 37 1C to wells of an ELISA plate and incubated with the HPV16 VLP-specific polyclonal antibody K75, monoclonal antibody (MAb) 33L1-7 that recognizes a linear epitope of the L1 protein at amino acids 303 to 313 indicative of disassembly, and MAb H16.V5 that recognizes a conformationally-dependent epitope of the FG loop of the L1 protein when assembled into capsids (Sapp et al., 1994; White et al., 1999). We observed comparable reactivity with all three antibodies between wild-type and mutant pseudoviruses suggesting that these mutations do not disrupt the capsid structure (Fig. 3C). Reduced infectivity is not due to deficiency in cell binding or uncoating To show that the reduction in infectivity is not due to a defect in cell binding, we bound pseudovirus particles to HaCaT cells grown on coverslips for 1 h at 37 1C, then fixed and stained using MAb 33L1-7 following the Click-iT reaction cocktail, which has previously been shown to denature the capsid protein to make the linear L1-specific epitope accessible (Bienkowska-Haba et al., 2012). We did not observe any difference in the binding of mutant L2 pseudoviruses compared to the binding of wild-type

S. DiGiuseppe et al. / Virology 458-459 (2014) 93–105

95

Fig. 1. in vivo nuclear import of HPV16 L2. (A) Intracellular localization of wt and terminally truncated HPV16 L2 at 24 hptx of HeLa cells. (B) Schematic diagram of L2 and result of multiple alignment analysis of L2 derived from 93 HPV types. Y axis gives percent conservation. Asterisks indicate fully conserved residues. ((C)–(E)) Intracellular localization of indicated HPV16 ((C) and (D)) and HPV18 (E) L2 mutants at 24 hptx of HeLa cells. The L2 protein (green), PML protein (red), and actin (blue) were stained. Note the cytoplasmic accumulation of mNLS point mutant L2 protein. (F) Western blot of indicated wild-type and mutant L2 proteins. (G) Quantitative analysis of intracellular localization of HPV16 wild-type (A) and indicated L2 mutants ((C) and (D)) with between 50 and 100 cells counted per group. Quantification is from one representative experiment from three biological replicates. N: exclusive nuclear localization; C/N: mainly cytoplasmic and additionally subnuclear PML NB localization.

pseudoviruses (Fig. 4A and C). To assess whether there was a defect in pseudovirus uncoating that could explain our reduction in infectivity, we infected HaCaT cells with wild-type and mutant pseudovirus for 18 h at 37 1C. Next, we fixed, permeabilized, and stained the L1 protein with MAb 33L1-7 without the Click-iT reaction cocktail and quantified uncoating relative to wild-type pseudovirus. The L2 mutants and wild-type pseudovirions exhibited comparable reactivity therefore suggesting that uncoating was not impacted by these mutations (Fig. 4B and D). Mutant pseudovirus fails to deliver encapsidated genome to the nucleus During infection, the L2 protein delivers the viral genome to the nucleus where it co-localizes with PML nuclear bodies (Day et al., 2004). In line with previous findings, our results from Fig. 1 indicate deletions in the mNLS element of the L2 protein abrogate

nuclear retention of the pseudogenome (Mamoor et al., 2012). Our infectivity data shows that this region is critical for pseudoinfection as point mutations in this region render these virions noninfectious. We next sought to investigate whether virions harboring this mutant L2 protein are reaching the nucleus. Therefore, we infected HaCaT cells with wild-type and mutant pseudovirus carrying EdU-labeled pseudogenome for 24 and 40 h at 37 1C. The cells were subsequently fixed, permeabilized, and stained following the Click-iT reaction cocktail. We observed that after 24 and 40 h, the wild-type pseudovirus successfully delivered the pseudogenome to the nucleus where it co-localizes with PML protein (Fig. 5A and C). In agreement with previous reports, pseudogenome was also detected in the cytoplasm indicating that not all pseudogenomes are delivered to the nucleus (Day et al., 2004). We were unable to detect any 16L2-R302/5A mutant pseudogenome in the nucleus at 24 or 40 h (Fig. 5B and D). In addition, we observed a cytosolic localization of pseudogenome

96

S. DiGiuseppe et al. / Virology 458-459 (2014) 93–105

Fig. 2. Intracellular localization of L2 protein at early times post-transfection. (A) Intracellular localization of HPV16 wild-type L2 and (B) indicated mutant L2 at early times post-transfection of HeLa cells. Note exclusive nuclear localization at early time points, but cytoplasmic distribution at later time points. (C) intracellular localization of HPV16 L2 wild-type and indicated mutants after 20 h post-transfection of HeLa cells in the presence or absence of 10 ng/ml LMB added to the cells at time 0 hptx.

delivered by 16L2-R302/5A mutant at both indicated times postpseudoinfection. Pseudogenome passes through the trans-Golgi network during infectious entry Our results from Fig. 4 and 5 show that pseudoviral particles with mutations in the L2 mNLS do not impact cell binding or uncoating, yet fail to deliver the pseudogenome to the nucleus. Therefore, we sought to determine where the pseudogenome delivered by 16L2-R302/5A mutant particles was accumulating late in the entry process. It was recently reported that the viral

genome passes through the TGN prior to entry into the nucleus (Day et al., 2013; Lipovsky et al., 2013). To test delivery of pseudogenome to the TGN, we infected HaCaT cells with wildtype and mutant pseudovirus carrying EdU-labeled pseudogenome for 18 and 36 h at 37 1C, fixed, permeabilized, and stained the TGN following the Click-iT reaction cocktail. We quantified the number of pseudogenome puncta co-localized with TGN46 (Fig. 6E) and observed that at 18 h, there was comparable colocalization between mutant and wild-type pseudogenome in the TGN (Fig. 6A and B). After 36 h, we observed that the majority of the wild-type pseudogenome has passed through the TGN and is localized in the nucleus (Fig. 6C). In contrast, pseudogenome

S. DiGiuseppe et al. / Virology 458-459 (2014) 93–105

97

Fig. 3. Characterization of mutant pseudoviruses. (A) The L1 (55 kDa) and L2 (75 kDa) proteins were detected with MAb HPV16 312F and 33L2-1 respectively by Western blot analysis. The pseudogenome of purified pseudovirus was detected using real-time PCR and CT values are reported. (B) 293TT and HaCaT cells were infected with HPV16 wild-type, 16L2-R302/5A, and 16L2-R297/8-302/5A pseudoviruses for 72 h and infectivity was measured by quantifying GFP transcript by real-time PCR. (C) A standard ELISA was performed on HPV16 wild-type (wt), 16L2-R302/5A (302/5), and 16L2-GP-N-R302/5A (GPN-302/5) pseudoviruses against the reactivity of polyclonal antibody anti-L1 K75, MAb anti-L1 33L1-7, and MAb anti-L1 V5 and absorbance was measured at 450 nm.

delivered by the 16L2-R302/5A mutant pseudovirus accumulated in the TGN at 36 h (Fig. 6D). These data support the previous findings that HPV16 traffics through the TGN and suggests that mutating arginine residues within the mNLS element of the L2 protein results in the accumulation of the pseudogenome in the TGN. L1 protein accompanies the L2/DNA complex to the TGN at reduced levels It is well-established that L2 but not L1 protein accompanies viral genome to the nucleus. Since this region was previously shown to bind promiscuously to DNA in vitro (Klucevsek et al., 2006), we were concerned that mutations in the mNLS element of the L2 protein may result in a reduced affinity for the pseudogenome. Therefore, we wanted to test whether the L2 protein is dissociating from the pseudogenome in the TGN. We grew HaCaT cells on coverslips and infected them with mutant 16L2-R302/5A pseudovirions carrying EdU-labeled pseudogenome for 36 h and co-stained for L2, EdU, and TGN46 following the Click-iT reaction cocktail. We observed pseudogenome co-localizes with mutant 16L2-R302/5A L2 protein in the TGN at 36 h (Fig. 7A) whereas wild-type L2 was co-localized with pseudogenome in the nucleus (data not shown). It has been recently shown that L1 protein colocalizes with TGN makers (Lipovsky et al., 2013). Therefore, next we asked whether we could also detect L1 protein co-localized with the pseudogenome in the TGN. To this end, we infected HaCaT cells with 16L2-R302/5A mutant pseudovirus carrying an EdU-labeled pseudogenome and stained for L1 and TGN46 at 36 h

following the Click-iT reaction cocktail to denature the capsid proteins. This was followed by staining with the 33L1-7 antibody that recognizes a linear epitope of the L1 capsid protein. In order to allow for quantification, we also measured the L1/viral pseudogenome signal strength of viral pseudoparticles bound for 1 h to ECM and compared EdU-derived signal strength of ECM and TGNresident pseudogenome (Fig. 7B). At 1 h, the L1 protein strongly co-localized with the pseudogenome for both the wild-type (data not shown) and mutant 16L2-R302/5A pseudovirus (Fig. 7F). After 36 h, the wild-type pseudogenome was predominately localized to the nucleus where L1 was completely undetectable as expected (data not shown). However, the TGN-resident pseudogenome delivered by mutant pseudovirions showed a 75% loss of L1 signal (Fig. 7C and G) compared to the ECM-resident pseudogenome. To test whether this loss in signal is due to limitations in the accessibility of our detection reagents, we measured the peak signal intensity of EdU puncta at 1 h compared to 36 h within the TGN. We found no significant loss in peak signal intensity (Fig. 7D) suggesting that we are losing L1 signal within the TGN. Additionally, we observed a complete absence of L1 signal in 38% of TGNresident EdU puncta (Fig. 7E and H). Taken together, these data suggest that the L1 protein is accompanying the pseudogenome to the TGN in both wild-type and mutant infections, but at greatly reduced levels. Cyclophilin activity is required prior to trafficking to the TGN We have previously shown that cyclophilins are required at two distinct steps in the entry process. First, cyclophilins mediate

98

S. DiGiuseppe et al. / Virology 458-459 (2014) 93–105

Fig. 4. Reduced infectivity is not due to deficiency in cell binding or uncoating. ((A) and (C)) HPV16 wild-type (wt), 16L2-R302/5A (302/5), 16L2-GP-N-R302/5A (GPN302/5) pseudoviruses were bound to HaCaT cells for 1 h and detected using MAb 33L1-7 following Click-iT reaction cocktail. The L1 protein (green), cell surface marker caveolin-1 (red), and nucleus (blue) were stained and pseudovirus binding was quantified relative to wild-type. ((B) and (D)) HaCaT cells were infected with HPV16 wild-type (wt), 16L2-R302/5A (302/5), 16L2-GP-N-R302/5A (GPN-302/5) and stained using MAb 33L1-7 without Click-iT reaction cocktail after 18 h to assess uncoating. The L1 protein (green), actin (purple), and nucleus (blue) were stained. ((C) and (D)) Quantification of binding and uncoating was calculated by counting 100 cells (n¼ 100) in one representative experiment from two biological replicates.

the isomerization of the L2 protein on the cell surface to expose the previously hidden region of L2 encompassing a furin cleavage site (Bienkowska-Haba et al., 2009). Second, cyclophilins are required at a post-internalization step to mediate the dissociation of the L1 protein from the L2/DNA complex prior to trafficking to

the nucleus (Bienkowska-Haba et al., 2012). Since our data from Fig. 7 showed that L1 protein is present in the TGN at reduced levels, we asked whether cyclophilin activity was required for trafficking to the TGN. For this, we used a pseudovirus with mutations in a putative cyclophilin binding domain at glycine 99

S. DiGiuseppe et al. / Virology 458-459 (2014) 93–105

99

Fig. 5. Nuclear delivery of encapsidated DNA by mutant pseudovirus. HaCaT cells were infected with HPV16 wild-type (wt) and 16L2-R302/5A (302/5) pseudovirus and stained after 24 h ((A) and (B)) or 40 h ((C) and (D)). PML protein (green), EdU-labeled pseudogenome (red), NUP153 (grey), and the nucleus (blue) were stained following Click-iT reaction cocktail. Note the cytoplasmic localization of pseudogenome delivered by mutant 16L2-302/5 pseudovirus.

and proline 100 that can bypass the surface cyclophilin requirement however infect cells comparable to wild-type (BienkowskaHaba et al., 2012). Therefore, we introduced the point mutations arg-302 and arg-305 to generate mutant 16L2-GP-N-R302/5A, which bypasses the surface cyclophilin requirement, yet accumulates in the TGN. This mutant is non-infectious (data not shown). We infected HaCaT cells with wild-type and mutant pseudoviruses carrying EdU-labeled pseudogenomes at 37 1C for 36 h in the presence or absence of cyclosporine A, a potent inhibitor of cyclophilins, stained the TGN following the Click-iT reaction cocktail, and quantified the images (Fig. 8D). As expected, in the absence of cyclosporine A, we found the majority of wild-type pseudogenome is localized in the nucleus (Fig. 8A) whereas pseudogenomes delivered by both 16L2-R302/5A (Fig. 8B) and 16L2-GP-N-R302/5A (Fig. 8C) pseudovirions localized in the TGN. In the presence of cyclosporine A, neither the wild-type nor mutant pseudogenomes were detectable in the TGN (Fig. 8A–C) demonstrating that cyclophilin activity is required for successful trafficking to the TGN. Additionally, we observed noticeably less pseudogenome present after 36 h in cells treated with cyclosporine A compared to untreated cells. We suspect this loss of EdU signal is due to degradation of the pseudogenome.

Discussion Herein, we investigated the role of a highly conserved central arginine-rich motif within the L2 protein with regard to nuclear delivery and cytoplasmic trafficking. Using transfection and overexpression, we showed that deletions of both terminal NLSs did not significantly affect intracellular localization of HPV16 L2 protein. In contrast, mutations in a third arginine-rich stretch of

residues changed the intracellular localization of L2 protein resulting in a mainly cytoplasmic localization. We selected to mutate this region because it has homology to the β-like import receptor binding (BIB) domains, often found in ribosomal proteins, which are recognized by a number of more recently identified import receptors like importin 4, importin 9a and 9b as well as importin 11. Despite the homology of all three putative L2 NLSs with known NLSs and their capability to serve as NLSs when fused to GFP, our data suggest that none of them are being used for L2 nuclear translocation following transfection and over-expression. That is because we found exclusive localization in the nucleus at early times post-transfection of an expression cassette encoding L2 protein harboring mutations in all NLSs. We conclude that either L2 protein harbors an additional unconventional NLS, which does not require extended stretches of basic amino acid residues, or it partners with a cellular protein, which provides the NLS. However, we cannot completely rule out the possibility of passive diffusion into the nucleus (Mamoor et al., 2012). Since L2 protein again accumulates in the cytoplasm at later times post-transfection and inhibitors of the nuclear export factor, CRM1, result in nuclear rather than cytoplasmic localization of mutant L2 protein, our data suggest that the mNLS rather functions as a nuclear retention signal. This is consistent with previously published findings (Mamoor et al., 2012). Since mutations in this central arginine-rich motif did not negatively affect pseudovirus assembly, we were able to investigate the effect of these mutations on entry and infectivity. It has been previously demonstrated that this region of 16L2 is essential for infectivity in both HaCaT cells and in the vaginal mouse model (Mamoor et al., 2012). We found that exchanges of basic amino acids within the mNLS did not impair the early steps of HPV16 pseudoinfection like attachment, internalization, and uncoating as

100

S. DiGiuseppe et al. / Virology 458-459 (2014) 93–105

Fig. 6. Pseudogenome passes through the trans-Golgi network during infectious entry. HaCaT cells were infected with wild-type (wt) and 16L2-R302/5A (302/5) for 18 h ((A) and (B)) and 36 h ((C) and (D)). TGN marker TGN46 (cyan), EdU-labeled pseudogenome (red), and nucleus (grey) were stained following Click-iT reaction. (E) Colocalization of TGN46 and EdU was quantified as a percentage of total EdU puncta. Statistical significance was determined by Student's t-test (p o 0.0001) where 50 cells were counted (n¼50). Cells were quantified from one representative experiment from three biological replicates. Note the accumulation of the pseudogenome co-localized with the TGN after 36 h.

the pseudogenome as well as the L2 protein was easily detected within the cell by confocal microscopy. However, some of the mutant pseudovirions were completely defective for infecting 293TT and HaCaT cells. Non-infectious pseudovirus harboring the L2 protein with mutations in arg-302 and arg-305 curiously showed a cytoplasmic distribution of the pseudogenome whereas wild-type pseudogenome was localizing in the nucleus at PML NB after 24 h. We noticed that much of the 16L2-R302/5A pseudogenome was observed in close proximity to the nucleus. Upon further investigation, we found that wild-type 16L2 pseudogenome passed through the TGN whereas pseudogenome carried by 16L2-R302/5A mutant pseudovirus trafficked to but was unable to escape the TGN even after 36 hpi. In line with previous reports by Day et al. (2013) and Lipovsky et al. (2013), these results provide

further evidence that HPV16 traffics through the TGN during infectious entry. It has been demonstrated that the mNLS of L2, abundant in positively charged arginine residues, promiscuously binds to DNA presumably through interacting with the negatively charged phosphate backbone of DNA (Klucevsek et al., 2006). Therefore, we thought that it was plausible that mutations in arginine residues may have resulted in a loss of affinity for the pseudogenome and thus it was being lost in the TGN. However, we did not detect any mutant L2 protein in the nucleus, which was instead co-localized with the pseudogenome in the TGN. Furthermore, quantification did not indicate partial loss of L2 protein from the pseudoviral genome in the TGN (data not shown). Therefore, these data suggest that the pseudogenome and the mutant L2

S. DiGiuseppe et al. / Virology 458-459 (2014) 93–105

101

Fig. 7. L1 accompanies the L2/DNA complex to the TGN. HaCaT cells were infected with HPV 16L2-R302/5A (302/5) for 1 h (B) or 36 h ((A) and (B)). The L2 protein (green) (A) or L1 protein (green) (B), EdU-labeled pseudogenome (red), TGN46 (cyan), and nucleus (grey) were stained following Click-iT reaction cocktail. (C) Peak signal intensity ratios of L1/EdU at 36 h co-localized with the TGN were quantified as a percentage of 1 h. Statistical significance was determined by Student's t-test (po0.0001). (D) The average peak EdU signal intensity of 1 h compared to peak EdU signal intensity at 36 h within the TGN as a percentage of 1 h. Note that there is no loss of EdU signal at 1 h compared to 36 h within the TGN. (E) The number of particles that had no detectable L1 signal intensity at 1 h compared 36 h co-localized with the TGN as a percentage of total particles counted. ((F)–(H)) Histograms of fluorescent signal intensity reported as AU from representative pseudovirions selected from 1 and 36 h. Note strong co-localization of L1 and pseudogenome after 1 h (F) binding whereas there is reduced L1 signal intensity ((G) and (H)) co-localized with pseudogenome within the TGN after 36 h. Statistical significance was determined by Student's t-test (po0.0001) where 100 particles were analyzed (n¼100).

102

S. DiGiuseppe et al. / Virology 458-459 (2014) 93–105

Fig. 8. Cyclophilin activity is required prior to trafficking to the TGN. ((A)–(C)) HaCaT cells were infected with HPV16 wild-type (wt), 16L2-R302/5A (302/5), 16L2-GP-NR302/5A (GPN-302/5) for 36 h in the presence or absence of 10 mM cyclosporine A (Cs). TGN46 (cyan), EdU-labeled pseudogenome (red), and nucleus (grey) were stained following the Click-iT reaction cocktail. (D) Co-localization of EdU/TGN46 was quantified as a percentage of total EdU puncta. Statistical significance was determined by Student's t-test (p o 0.001) where 50 cells were counted (n¼ 50). Cells were quantified from one representative experiment from two biological replicates. Note that pseudogenome does not co-localize with the TGN in cyclosporine A treated cells after 36 h.

protein are retained in the TGN together, presumably in a complex, unable to escape. One explanation for this could be that the mNLS region of L2 is required for interacting with a TGN egress factor that is required to carry the pseudogenome to a subsequent compartment, such as the endoplasmic reticulum as previously proposed (Bossis et al., 2005; Lipovsky et al., 2013). Another explanation is that the mNLS is important for interacting with a nuclear factor that carries the L2/DNA complex into the nucleus. Further experiments are required to provide mechanistic insight into how the L2 protein and viral DNA escape from the TGN. In any case, our observations clearly demonstrate that an intact mNLS of the L2 protein is important to establish the pseudoviral infection. Our lab has previously shown that cyclophilins are required for the dissociation of the L1 protein from the L2/DNA complex (Bienkowska-Haba et al., 2009, 2012). We presumed that this dissociation occurred within the endosome prior to nuclear delivery as L1 is undetectable in the nucleus. Herein, we found that inhibiting cyclophilins blocked trafficking of the pseudogenome to the TGN. Therefore, we delineate that cyclophilin activity

facilitates the dissociation of the L1 protein from the L2/DNA complex prior to trafficking to the TGN. This may be explained by a couple reasons. First, the L1 protein may have a strong putative sorting signal which prevents trafficking of the L2/DNA complex to the TGN until the majority of L1 is removed. Second, the L1 protein sterically blocks the L2 protein from engaging factors required for retrograde trafficking. Such engagements may include the interaction of the L2 protein with components of the retromer complex, the dynein light chain, and sorting nexin 17, all of which play roles in retrograde trafficking (Bergant and Banks, 2013; Florin et al., 2006; Lipovsky et al., 2013; Schneider et al., 2011). However, Lipovsky et al. demonstrated by the highly sensitive proximity ligation assay that the L1 protein is detectable in the TGN. We therefore compared the L1 signal strength of intact pseudovirions on the ECM to pseudovirions localized in the TGN. In line with previous findings, we also found L1 protein in the TGN colocalized with the pseudogenome but at reduced levels. A significant percentage of pseudogenomes detected within the TGN had no detectable L1 protein suggesting that L1 has been

S. DiGiuseppe et al. / Virology 458-459 (2014) 93–105

completely removed from these pseudoviral genomes. Therefore, these data confirm our previous findings that CyPs facilitate dissociating the majority of L1 protein from the L2/DNA complex, presumably in the endosome, an event that according to data presented herein are required for trafficking to the TGN. Taken together, we can speculate the following scenarios. First, cyclophilins present in the endosome remove the bulk of the L1 protein from the L2/DNA complex prior trafficking to the TGN, while the remainder of the L1 protein is removed by TGN resident cyclophilins (Price et al., 1994). Second, we see an incomplete L1 dissociation in our system because we are depleting cyclophilins by infecting with too many virions. Lastly, the presence of L1 protein physically blocks hidden L2 sorting signals, interactions with egress factors, or a combination of both that would normally result in TGN egress. Therefore, only the pseudovirions that have L1 removed may progress to the nucleus, which may explain why we see a subset of pseudovirions without L1 protein present in the TGN for the L2-R302/5A mutant pseuodoviral infection and no detectable L1 protein in the nucleus for wild-type infection. Clearly many questions about how HPV traffics to and escapes the TGN to deliver its viral genome to the nucleus remain unanswered. However, the data presented here confirm the results from previous findings. Furthermore, we demonstrate that cyclophilin activity is required for the transport to the viral genome of HPV16 to the TGN.

103

HPV16 pseudovirus generation, viral genome equivalence, and infection assays Pseudovirus harboring encapsidated pEGFP-C1 plasmid was generated using the 293TT cell line as previously described (Buck and Thompson, 2007; Buck et al., 2005b) using the plasmids puf16L1 and puf16L2HA-30 . Virions were characterized by L1and L2-specific Western blot using mouse monoclonal antibodies 312F and 33L2-1, respectively. The viral DNA was isolated using the DNeasy Blood & Tissue kit according to manufacturer's instructions (69504, Qiagen) supplemented with 4 mM Ethylenediaminetetraacetic acid (EDTA) and Dithiothreitol (DTT). The viral genome equivalent (vge) was determined by real-time PCR. No significant differences were observed with regard to packaging, plasmid encapsidation, and L2 incorporation among the L2 mutant proteins characterized herein. For pseudogenome detection by immunofluorescence, pseudogenome was labeled by supplementing growth media with 100 mM 5-ethynyl-20 -deoxyuridine (EdU) at 6 hptx as described (11) during generation of pseudovirus. For infection assays, same amount of vge was used to infect 293TT cells and monitored by flow cytometry at 72 h (data not shown). Infectivity was measured by infecting HaCaT and 293TT cells with the same amount of vge. After 72 h, the total RNA was isolated using TRIzol, which was followed by the isolation of mRNA using GenElute™ mRNA Miniprep Kit (MRN70, Sigma) according to manufacturer's protocol. The cDNA was synthesized and the amount of GFP transcript was measured by real-time PCR.

Methods Cell lines

Transfection of expression plasmids and immunofluorescence

293TT and HeLa cells were cultured in DMEM supplemented with 10% FBS and antibiotics. HaCaT cells were grown in low glucose DMEM containing 5% FBS and antibiotics.

HeLa cells were grown on glass cover slips in 6-well plates and transfected with 3 mg of indicated L2 expression plasmid in serumfree medium using MATra reagent (#7-2001-100, IBA) according to manufacturer's protocol. After 24 hptx, whole-cell lysates were collected and L2 protein expression was detected by Western blot analysis. For the immunofluorescence in the presence of Leptomycin B (LMB) (L2913, Sigma), 10 ng/ml LMB was added at 0 hptx and cells were stained at 20 hptx. At the indicated times posttransfection, cells grown on coverslips were washed with PBS and fixed with ethanol. Next, slides were washed and blocked with 5% goat serum in PBS for 30 min, followed by 1 h incubation with primary antibodies. After extensive washing, cells were incubated with AlexaFluor-tagged secondary antibodies for 1 h. After extensive washing with PBS, cells were mounted with ProLongs Gold and SlowFadesAntifade. Groups of cells were counted between 50 and 100 cells in total on single-slice images captured by fluorescence microscopy using Images were captured by fluorescence microscopy using Leica DBMI6000 or Zeiss LSM 510. Images were all enhanced uniformly using Adobe Photoshop.

Antibodies HPV16 L1-specific mouse monoclonal antibody (MAb) 33L1-7, HPV16 312F, polyclonal rabbit antibody K75, and HPV16 L2specific antibody 33L2-1 have been described. For the detection of HPV18 L1 protein, L1-specific mouse MAb 28F was used. H16.V5 was a kind gift from N. D. Christensen, Hershey Medical Center. Click-iT EdU imaging kit, Alexa Fluor (AF)-labeled secondary antibodies, and phalloidin were used in immunofluorescent microscopy (Invitrogen). Peroxidase-conjugated AffiniPure goat anti-mouse and anti-rabbit antibodies were used for Western blot analysis (Jackson ImmunoResearch). The trans-Golgi network was detected using MAb anti-p230 (611280, BD Biosciences) and polyclonal sheep anti-TGN46 (AHP500GT, AbD Serotec). The PML protein was detected using polyclonal rabbit anti-PML (AB1370, Chemicon Interntional) and nuclear pore complex using MAb antiNUP153 (MMS-102P, Covance). The cell surface was detected using rabbit polyclonal anti-Caveolin-1 (ab2910, abcam) and Alexa Fluor 647-conjugated Phallodin toxin (A22287, Invitrogen). Generation of mutations Point mutations in L2 were generated by site-directed mutagenesis using pairs of complementary PCR primers specific to codon-optimized plasmid puf16L2HA-30 using the PfuUltra II Hotstart DNA Polymerase 2  Master Mix (#600630, Stratagene) following manufacturer's protocol. The entire plasmid is amplified during the PCR reaction, and the PCR products were digested with Dpn1 to remove methylated template DNA. The PCR products were transformed and the mutations were confirmed by sequencing (SeqWright).

Enzyme linked immunosorbent assay (ELISA) Pseudovirions were diluted in phosphate-buffered saline (PBS), added to a 96-well plate (Nunc-Immuno Module, Nunc) in triplicate, and bound for 1 h at 37 1C. Wells were washed with PBS—0.2% Tween 20 (PBST) and blocked with 0.01% bovine serum albumin (BSA) in PBST for 1 h at 37 1C. Wells were incubated with primary antibodies polyclonal rabbit antibody K75, MAb 33L1-7 and MAb H16.V5 for 30 min at 37 1C in a humidity chamber and subsequently washed with PBST. Wells were incubated with peroxidase-conjugated anti-mouse or anti-rabbit secondary antibody for 30 min at 37 1C in a humidity chamber and subsequently washed with PBST. The assays were developed with trimethylbenzidine (Promega) and the reaction was halted by adding 1 M HCl. Absorbance was measured at 450 nm using a FLUOstar-Omega

104

S. DiGiuseppe et al. / Virology 458-459 (2014) 93–105

plate reader (BMG Labtech). This experiment was repeated twice in triplicate and error bars are reported in standard error. Immunofluorescent microscopy after infection HaCaT cells were grown on No. 01 coverslips to approximately 50% confluency and infected with HPV16 pseudovirus in the presence or absence of 10 μM cyclosporine (C988900; Toronto Research Chemicals). Cells were infected with approximately 1  106 to 1  107 vge/coverslip. At indicated times post-infection, coverslips were washed with pH 10.5 PBS for exactly 2 min to remove any unbound pseudovirus with the exception of Figs. 3 and 6, fixed with 4% paraformaldehyde (PFA) for 20 min at room temperature, washed with pH 7.5 PBS, permeabilized with 0.2% Triton X-100 in PBS for 7 min, washed, and blocked with 5% goat serum or 5% BSA in PBS for 30 min followed by the Click-iT reaction cocktail containing Alexa Fluor 555 for pseudogenome detection. After washing, cells were incubated with primary antibodies in 2.5% goat serum or 5% BSA for 30 min at 37 1C. We extensively washed and incubated with secondary antibodies in 2.5% goat serum or 5% BSA for 30 min at 37 1C. We extensively wash again in PBS and mounted with ProLongs Gold and SlowFades Antifade containing 40 ,60 -diamidino-2-phenylindole (DAPI) (P36931; Invitrogen). All images were acquired with a Leica TCS SP5 Spectral Confocal Microscope. Single slices were quantified by selecting regions of interest (ROIs) around the TGN and nucleus and counting co-localization as a percentage of total EdUspecific puncta where both channels show an overlapping peak fluorescent intensity with arbitrary units (AU) greater than 700 (Figs. 6 and 8). Error bars are reported in standard error (n ¼50). Quantifications are from one representative experiment from three (Fig. 6) and two (Fig. 8) biological replicates, respectively. Statistical significance was determined using a Student's t-test (p o0.0001) and (p o0.001) for Figs. 6 and 8, respectively. Ratio of L1/EdU was determined by taking the ratio of the peak AU of L1 and EdU at times indicated (Fig. 7). We set a threshold of 300 AU for all analyses. Any peaks below 300 AU were counted as L1 negative particles. Error bars are reported in standard error (n ¼250). Statistical significance was determined using a Student's t-test (po 0.0001). Images were all enhanced uniformly using Adobe Photoshop. Cell-binding and uncoating assays The cell-binding assay was performed by growing HaCaT cells on coverslips to 50% confluency and binding equal number of particles for 1 h at 37 1C. Cells were stained as described above without the pH 10.5 PBS wash and stained with MAb 33L1-7 following Click-iT reaction to denature capsid protein. Assay was quantified by taking pixel sum ratio of L1-specific signal to caveolin-1. The uncoating assay was performed by growing HaCaT cells to 50% confluency on coverslips then infected with equal number of particles for 18 h at 37 1C. Then, cells were stained as previously described above with MAb 33L1-7 and Alexa Fluor 647 conjugated phallodin without Click-iT reaction cocktail. Uncoating was quantified by measuring the pixel sum ratio of the L1- to actin-specific signal.

Acknowledgments We are grateful to members of the Sapp lab for helpful discussions and J.T. Schiller and N.D. Christensen for providing reagents. The project described was supported by R01AI081809 from the National Institute of Allergy and Infectious Diseases to M.S. This

project was also supported in part by grants from the National Center for Research Resources (5P20RR018724-10) and the National Institute of General Medical Sciences (8 P20 GM10343310) from the National Institutes of Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. References Becker, K.A., Florin, L., Sapp, C., Sapp, M., 2003. Dissection of human papillomavirus type 33 L2 domains involved in nuclear domains (ND) 10 homing and reorganization. Virology 314, 161–167. Bergant, M., Banks, L., 2013. SNX17 facilitates infection with diverse papillomavirus types. J. Virol. 87, 1270–1273. Bienkowska-Haba, M., Patel, H.D., Sapp, M., 2009. Target cell cyclophilins facilitate human papillomavirus type 16 infection. PLoS Pathog. 5, e1000524. Bienkowska-Haba, M., Williams, C., Kim, S.M., Garcea, R.L., Sapp, M., 2012. Cyclophilins facilitate dissociation of the human papillomavirus type 16 capsid protein L1 from the L2/DNA complex following virus entry. J. Virol. 86, 9875–9887. Bordeaux, J., Forte, S., Harding, E., Darshan, M.S., Klucevsek, K., Moroianu, J., 2006. The l2 minor capsid protein of low-risk human papillomavirus type 11 interacts with host nuclear import receptors and viral DNA. J. Virol. 80, 8259–8262. Bossis, I., Roden, R.B., Gambhira, R., Yang, R., Tagaya, M., Howley, P.M., Meneses, P.I., 2005. Interaction of tSNARE syntaxin 18 with the papillomavirus minor capsid protein mediates infection. J. Virol. 79, 6723–6731. Buck, C.B., Cheng, N., Thompson, C.D., Lowy, D.R., Steven, A.C., Schiller, J.T., Trus, B.L., 2008. Arrangement of L2 within the papillomavirus capsid. J. Virol. 82, 5190–5197. Buck, C.B., Pastrana, D.V., Lowy, D.R., Schiller, J.T., 2004. Efficient intracellular assembly of papillomaviral vectors. J. Virol. 78, 751–757. Buck, C.B., Pastrana, D.V., Lowy, D.R., Schiller, J.T., 2005a. Generation of HPV pseudovirions using transfection and their use in neutralization assays. Methods Mol. Med. 119, 445–462. Buck, C.B., Thompson, C.D., 2007. Production of papillomavirus-based gene transfer vectors. Current Protocols in Cell Biology/Editorial Board, Juan S. Bonifacino… [et al.] Chapter 26, Unit 26 21. Buck, C.B., Thompson, C.D., Pang, Y.Y., Lowy, D.R., Schiller, J.T., 2005b. Maturation of papillomavirus capsids. J. Virol. 79, 2839–2846. Chen, X.S., Garcea, R.L., Goldberg, I., Casini, G., Harrison, S.C., 2000. Structure of small virus-like particles assembled from the L1 protein of human papillomavirus 16. Mol. Cell 5, 557–567. Christensen, N.D., Cladel, N.M., Reed, C.A., 1995. Postattachment neutralization of papillomaviruses by monoclonal and polyclonal antibodies. Virology 207, 136–142. Darshan, M.S., Lucchi, J., Harding, E., Moroianu, J., 2004. The l2 minor capsid protein of human papillomavirus type 16 interacts with a network of nuclear import receptors. J. Virol. 78, 12179–12188. Day, P.M., Baker, C.C., Lowy, D.R., Schiller, J.T., 2004. Establishment of papillomavirus infection is enhanced by promyelocytic leukemia protein (PML) expression. Proc. Nat. Acad. Sci. U.S.A. 101, 14252–14257. Day, P.M., Gambhira, R., Roden, R.B., Lowy, D.R., Schiller, J.T., 2008. Mechanisms of human papillomavirus type 16 neutralization by l2 cross-neutralizing and l1 type-specific antibodies. J. Virol. 82, 4638–4646. Day, P.M., Lowy, D.R., Schiller, J.T., 2003. Papillomaviruses infect cells via a clathrindependent pathway. Virology 307, 1–11. Day, P.M., Thompson, C.D., Schowalter, R.M., Lowy, D.R., Schiller, J.T., 2013. Identification of a role for the trans-Golgi network in human papillomavirus 16 pseudovirus infection. J. Virol. 87, 3862–3870. Favre, M., Breitburd, F., Croissant, O., Orth, G, 1977. Chromatin-like structures obtained after alkaline disruption of bovine and human papillomaviruses. J. Virol. 21, 1205–1209. Finch, J.T., Klug, A., 1965. The structure of viruses of the papilloma–polyoma type 3. Structure of rabbit papilloma virus, with an appendix on the topography of contrast in negative-staining for electron-microscopy. J. Mol. Biol. 13, 1–12. Florin, L., Becker, K.A., Lambert, C., Nowak, T., Sapp, C., Strand, D., Streeck, R.E., Sapp, M., 2006. Identification of a dynein interacting domain in the papillomavirus minor capsid protein l2. J. Virol. 80, 6691–6696. Giroglou, T., Florin, L., Schafer, F., Streeck, R.E., Sapp, M., 2001. Human papillomavirus infection requires cell surface heparan sulfate. J. Virol. 75, 1565–1570. Kamper, N., Day, P.M., Nowak, T., Selinka, H.C., Florin, L., Bolscher, J., Hilbig, L., Schiller, J.T., Sapp, M., 2006. A membrane-destabilizing peptide in capsid protein L2 is required for egress of papillomavirus genomes from endosomes. J. Virol. 80, 759–768. Klucevsek, K., Daley, J., Darshan, M.S., Bordeaux, J., Moroianu, J., 2006. Nuclear import strategies of high-risk HPV18 L2 minor capsid protein. Virology 352, 200–208. Knappe, M., Bodevin, S., Selinka, H.C., Spillmann, D., Streeck, R.E., Chen, X.S., Lindahl, U., Sapp, M., 2007. Surface-exposed amino acid residues of HPV16 L1 protein mediating interaction with cell surface heparan sulfate. J. Biol. Chem. 282, 27913–27922. Li, M., Beard, P., Estes, P.A., Lyon, M.K., Garcea, R.L., 1998. Intercapsomeric disulfide bonds in papillomavirus assembly and disassembly. J. Virol. 72, 2160–2167.

S. DiGiuseppe et al. / Virology 458-459 (2014) 93–105

Lipovsky, A., Popa, A., Pimienta, G., Wyler, M., Bhan, A., Kuruvilla, L., Guie, M.A., Poffenberger, A.C., Nelson, C.D., Atwood, W.J., DiMaio, D., 2013. Genome-wide siRNA screen identifies the retromer as a cellular entry factor for human papillomavirus. Proc. Nat. Acad. Sci. U.S.A. 110, 7452–7457. Liu, W.J., Gissmann, L., Sun, X.Y., Kanjanahaluethai, A., Muller, M., Doorbar, J., Zhou, J., 1997. Sequence close to the N-terminus of L2 protein is displayed on the surface of bovine papillomavirus type 1 virions. Virology 227, 474–483. Mamoor, S., Onder, Z., Karanam, B., Kwak, K., Bordeaux, J., Crosby, L., Roden, R.B., Moroianu, J., 2012. The high risk HPV16 L2 minor capsid protein has multiple transport signals that mediate its nucleocytoplasmic traffic. Virology 422, 413–424. Mejia, A.F., Culp, T.D., Cladel, N.M., Balogh, K.K., Budgeon, L.R., Buck, C.B., Christensen, N.D., 2006. Preclinical model to test human papillomavirus virus (HPV) capsid vaccines in vivo using infectious HPV/cottontail rabbit papillomavirus chimeric papillomavirus particles. J. Virol. 80, 12393–12397. Modis, Y., Trus, B.L., Harrison, S.C., 2002. Atomic model of the papillomavirus capsid. EMBO J. 21, 4754–4762. Price, E.R., Jin, M., Lim, D., Pati, S., Walsh, C.T., McKeon, F.D., 1994. Cyclophilin B trafficking through the secretory pathway is altered by binding of cyclosporin A. Proc. Nat. Acad. Sci. U.S.A. 91, 3931–3935. Pyeon, D., Pearce, S.M., Lank, S.M., Ahlquist, P., Lambert, P.F., 2009. Establishment of human papillomavirus infection requires cell cycle progression. PLoS Pathog. 5, e1000318. Richards, K.F., Bienkowska-Haba, M., Dasgupta, J., Chen, X.S., Sapp, M., 2013. Multiple heparan sulfate binding site engagements are required for the infectious entry of human papillomavirus type 16. J. Virol. 87, 11426–11437. Richards, R.M., Lowy, D.R., Schiller, J.T., Day, P.M., 2006. Cleavage of the papillomavirus minor capsid protein, L2, at a furin consensus site is necessary for infection. Proc. Nat. Acad. Sci. U.S.A. 103, 1522–1527. Sapp, M., Fligge, C., Petzak, I., Harris, J.R., Streeck, R.E., 1998. Papillomavirus assembly requires trimerization of the major capsid protein by disulfides between two highly conserved cysteines. J. Virol. 72, 6186–6189. Sapp, M., Kraus, U., Volpers, C., Snijders, P.J., Walboomers, J.M., Streeck, R.E., 1994. Analysis of type-restricted and cross-reactive epitopes on virus-like particles of

105

human papillomavirus type 33 and in infected tissues using monoclonal antibodies to the major capsid protein. J. Gen. Virol. 75 (12), 3375–3383. Schelhaas M., Schah B., Holzer M., Blattmann P., Kuhling L., Day P.M., Schiller J.T., Helenius A., Entry of human papillomavirus type 16 by actin-dependent and clathrin- and lipid raft-independent endocytosis, PLos Pathog. 8, 2012, e1002657. Schneider, M.A., Spoden, G.A., Florin, L., Lambert, C., 2011. Identification of the dynein light chains required for human papillomavirus infection. Cell. Microbiol. 13, 32–46. Selinka H.C., Giroglou T., and Sapp M., Analysis of the infectious entry pathway of human papillomavirus type 33 pseudovirions, Virology 299, 2002, 279–287. Selinka, H.C., Florin, L., Patel, H.D., Freitag, K., Schmidtke, M., Makarov, V.A., Sapp, M., 2007. Inhibition of transfer to secondary receptors by heparan sulfatebinding drug or antibody induces noninfectious uptake of human papillomavirus. J. Virol. 81, 10970–10980. Smith, J.L., Campos, S.K., Ozbun, M.A., 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. Spoden, G., Freitag, K., Husmann, M., Boller, K., Sapp, M., Lambert, C., Florin, L., 2008. Clathrin- and caveolin-independent entry of human papillomavirus type 16—involvement of tetraspanin-enriched microdomains (TEMs). PLoS One 3, e3313. Spoden, G., Kuhling, L., Cordes, N., Frenzel, B., Sapp, M., Boller, K., Florin, L., Schelhaas, M., 2013. Human papillomavirus types 16, 18, and 31 share similar endocytic requirements for entry. J. Virol. 87, 7765–7773. Sun, X.Y., Frazer, I., Muller, M., Gissmann, L., Zhou, J., 1995. Sequences required for the nuclear targeting and accumulation of human papillomavirus type 6B L2 protein. Virology 213, 321–327. White, W.I., Wilson, S.D., Palmer-Hill, F.J., Woods, R.M., Ghim, S.J., Hewitt, L.A., Goldman, D.M., Burke, S.J., Jenson, A.B., Koenig, S., Suzich, J.A., 1999. Characterization of a major neutralizing epitope on human papillomavirus type 16 L1. J. Virol. 73, 4882–4889. Wolf, M., Garcea, R.L., Grigorieff, N., Harrison, S.C., 2010. Subunit interactions in bovine papillomavirus. Proc. Nat. Acad. Sci. U.S.A. 107, 6298–6303.