The human papillomavirus type 11 E1^E4 protein is a transglutaminase 3 substrate and induces abnormalities of the cornified cell envelope

Virology 345 (2006) 290 – 298 www.elsevier.com/locate/yviro

The human papillomavirus type 11 E1^E4 protein is a transglutaminase 3 substrate and induces abnormalities of the cornified cell envelope Darron R. Brown a,b,*, Douglas Kitchin a, Brahim Qadadri a, Nicole Neptune a, Teresa Batteiger a, Aaron Ermel a b

a Department of Medicine, Indiana University School of Medicine, Indianapolis, IN 46077, USA Department of Microbiology and Immunology, Indiana University School of Medicine, Indianapolis, IN 46077, USA

Received 24 August 2005; accepted 19 September 2005 Available online 28 October 2005

Abstract The human papillomavirus (HPV) E1^E4 protein is detected in the cytoplasm of differentiated keratinocytes, near the cornified cell envelope. HPV does not induce lysis of the infected keratinocyte, and the normally durable cornified cell envelope that forms during keratinocyte differentiation would seemingly inhibit viral egress. HPV infection induces abnormalities of the cornified cell envelope, but the exact mechanisms involved are not well understood. We tested whether the HPV 11 E1^E4 protein, which co-localizes the cell envelope and co-purifies with cell envelope fragments, could serve as an in vitro substrate for transglutaminases. We found evidence of E1^E4 cross-linking by endogenous transglutaminases in an in situ assay using frozen sections of human foreskin, and in addition, E1^E4 protein was cross-linked by recombinant transglutaminase 3 (but not transglutaminase 1) in an in vitro cross-linking assay. We also tested whether expression of E1^E4 in differentiated keratinocytes would induce morphologic alterations of cornified cell envelopes. Differentiated keratinocytes expressing E1^E4 were disorganized and pleomorphic compared to control cells, and cell envelopes purified from E1^E4-expressing cells were small, fragmented, and rough bordered compared to the round, smooth bordered cell envelopes from control cells. We conclude from these in vitro experiments that the E1^E4 protein is cross-linked by transglutaminase 3, and that E1^E4 expression in differentiated keratinocytes induces morphologic abnormalities of the cornified cell envelope. D 2005 Elsevier Inc. All rights reserved. Keywords: Human Papillomavirus; E1^E4; Cornified cell envelope

Introduction Human papillomaviruses (HPVs) infect epithelial tissues, including those of the genitalia. HPV types 6 and 11 cause condylomata acuminata, or genital warts (Brown et al., 1993; Ferenczy, 1995; Gissmann et al., 1982). These proliferative lesions often produce virions that are detected in differentiated keratinocytes, cells that have developed a cornified cell envelope (Ferenczy et al., 1981; Firzlaff et al., 1988; Morin et al., 1981; Toki et al., 1986). The cornified cell envelope is a dense matrix of proteins and lipids covalently cross-linked by the enzymatic activity of keratinocyte transglutaminases (Eck-

* Corresponding author. Department of Medicine, 545 Barnhill Drive, Indianapolis, IN 46202, USA. Fax: +1 317 274 1587. E-mail address: [email protected] (D.R. Brown). 0042-6822/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.virol.2005.09.048

ert et al., 1997; Kalinin et al., 2002; Reichert et al., 1993; Rice and Green, 1977; Steinert and Marekov, 1999). The endpoint of keratinocyte differentiation is a layer of flattened, durable, enucleated cornified cell envelopes that provides the host with barrier protection. In contrast to many other virus infections, HPV does not induce lysis of the target cell (Lehr and Brown, 2002). The virus must be able to escape the infected keratinocyte for transmission of infection, and the mechanisms for this process are not well understood. The cornified cell envelope in its normal, durable state would seemingly hinder virion release from desquamated keratinocytes. We have theorized that a defect in the cornified cell envelope could facilitate release of virions. We showed that natural HPV 11 infection induces cell envelope abnormalities, and that desquamated, cornified cells are effective transmitters of virus (Brown and Bryan, 2000; Bryan and Brown, 2001).

D.R. Brown et al. / Virology 345 (2006) 290 – 298

We have also previously shown that the HPV 11 E1^E4 protein localizes to the region of the cell envelope in immunohistochemical assays, co-purifies with cell envelope fragments, and is associated with the cell envelope structure in immunoelectron microscopic studies (Bryan and Brown, 2000). In the current study, we performed experiments to determine if keratinocyte transglutaminases could cross-link the HPV 11 E1^E4 protein. We next performed experiments to determine if expression of E1^E4 in differentiated human keratinocytes induced alterations in the morphologic characteristics of the cornified cell envelope. Results Expression of recombinant proteins Recombinant viral and cellular proteins were expressed and purified for these experiments. Hexa-histidine-tagged (histagged) HPV 11 E1^E4 protein was purified from bacterial using nickel columns, then analyzed in SDS-PAGE gels (Fig. 1, Panel A, Lane 1) and immunoblots (Fig. 1, Panel A, Lane 3). Freshly purified, his-tagged E1^E4 appeared in the gel as a single band of approximately 15 kDa, and as a doublet of approximately 14 and 15 kDa in the immunoblots using an antibody against the his-tag. His-tagged loricrin was expressed in SF21 insect cells, then purified using nickel columns and analyzed in SDS-PAGE gels (Fig. 1, Panel A, Lane 2) and immunoblots (Fig. 1, Panel A, Lane 4). Bands at 65 and 45 kDa were visible in the SDSPAGE gel, while only the 45-kDa protein was identified in the immunoblot using an anti-his-tag antibody. Thus, the 65-kDa protein seen in the SDS-PAGE gel may have been a contaminant protein. His-tagged transglutaminases were expressed in SF21 insect cells, purified using nickel columns and analyzed in SDSPAGE gels and immunoblots (Fig. 1, Panel B). For transglutaminase 1, bands at approximately 100 kDa and 65 kDa were visible in the SDS-PAGE gel (Fig. 1, Panel B, Lane 1), while only the 100-kDa protein was identified in the immunoblot using an anti-his-tag antibody (Fig. 1, Panel B, Lane 3). Thus,

291

as above, the 65-kDa protein seen in the SDS-PAGE gel may have been a contaminant protein. His-tagged transglutaminase 3 proenzyme (i.e., the transglutaminase 3 protein prior to activation by digestion with dispase) appeared in the gel (Fig. 1, Panel B, Lane 2) and immunoblot (Fig. 1, Panel B, Lane 4) as a single band of approximately 90 kDa. Following digestion of transglutaminase 3 with dispase, a second immunoblot was performed that demonstrated the 90-kDa his-tagged transglutaminase 3 proenzyme prior to proteolytic cleavage (Fig. 1, Panel C, Lane 1), then as a 60-kDa activated, his-tagged transglutaminase 3 enzyme band after dispase digestion (Fig. 1, Panel C, Lane 2). The C-terminal 30 kDa fragment of the transglutaminase proenzyme did not appear on the immunoblot, most likely because the hexa-histidine tag was fused at the N-terminus of the protein. E1^ E4 cross-linking by endogenous transglutaminases in tissue sections Sections of neonatal foreskin tissue were analyzed for tissue histology (Fig. 2, Panel A) and for the presence of transglutaminase 3 in immunohistochemical assays using a mouse monoclonal antibody (Fig. 2, Panel B). Transglutaminase 3 was detected in the suprabasal layers of epithelium. To our knowledge, a monoclonal antibody against transglutaminase 1 is not available; therefore we could not test specifically for the location of transglutaminase 1 in the foreskin epithelium. We next performed an experiment using an in situ assay to provide evidence for incorporation of E1^E4 protein into the cell envelope by endogenous transglutaminases in neonatal foreskin tissue. Photomicrographs of Panels C through F were taken at the same exposure time, and a high degree of fluorescence was present only in the assay in which his-tagged E1^E4 protein and calcium were present (Fig. 2, Panel C). The highest degree of fluorescence was present in the most differentiated keratinocytes, in the region of the cornified cell envelope. No significant fluorescence was noted in the assay in which EDTA was added and calcium omitted (Fig. 2, Panel D). The higher abundance of E1^E4 in the more differentiated

Fig. 1. Analysis of recombinant proteins used in the in vitro cross-linking assay. Panel A: analysis of recombinant proteins. Lanes 1 and 3: SDS-PAGE and immunoblot analysis, respectively, of his-tagged E1^E4 protein. Lanes 2 and 4: SDS-PAGE and immunoblot analysis, respectively, of his-tagged loricrin. Panel B: analysis of his-tagged transglutaminases. Lanes 1 and 3: SDS-PAGE and immunoblot analysis, respectively, of his-tagged transglutaminase 1. Lanes 2 and 4: SDSPAGE and immunoblot analysis, respectively, of his-tagged transglutaminase 3. Panel C: Lane 1: immunoblot analysis of his-tagged transglutaminase 3. Lane 2: dispase-digested transglutaminase 3. Antibody used for all immunoblots was a rabbit polyclonal antibody against the hexa-histidine tag. Molecular weight markers (in kDa) are indicated on the left side of each panel.

292

D.R. Brown et al. / Virology 345 (2006) 290 – 298

transglutaminase-mediated cross-linking (Fig. 3, Panel A). Loricrin appeared as a 45-kDa monomer in immunoblots of cross-linking assays in which transglutaminase (in this case, transglutaminase 3) was omitted (Fig. 3, Panel A, Lane 1), and as ‘‘fast loricrin monomers’’ of approximately 35 kDa and ‘‘fast loricrin dimers’’ of approximately 60 kDa in assays containing transglutaminase (Fig. 3, Panel A, Lane 2). The rapid migration of cross-linked loricrin, a glycine- and serine-rich protein, has been described previously by Candi et al., and is thought to be caused by intramolecular cross-linking (Candi et al., 1995, 2001). In assays using recombinant transglutaminase 1, similar results to the transglutaminase 3 experiments were found using loricrin as a substrate (data not shown). Recombinant his-tagged HPV 11 E1^E4 protein was used in cross-linking experiments with recombinant transglutaminase 1 or transglutaminase 3. No evidence of cross-linking was found with transglutaminase 1 (data not shown). In contrast, reactions in which transglutaminase 3 was present produced the appearance of high molecular weight, immunoreactive complexes and a marked reduction in monomeric E1^E4 protein indicated that cross-linking between molecules had occurred (Fig. 4, Panel B, Lanes 1 and 2). Cross-linking assays of E1^E4 protein in which calcium was not added appeared similar to

Fig. 2. Analysis of foreskin epithelium and the in situ cross-linking assay. Panel A: histologic analysis of foreskin epithelium, stained with hematoxylin and eosin. Panel B: immunohistochemical analysis of epithelium using a mouse monoclonal antibody to detect transglutaminase 3. Panel C: in situ cross-linking assay. HPV 11 E1^E4 protein and substrate buffer (Tris – HCl 100 mM, pH 8.4, calcium chloride 5 mM) added. Panel D: in situ cross-linking assay. E1^E4 protein and control buffer (Tris – HCl 100 mM, pH 8.4, EDTA 5 mM) added. Panel E: in situ cross-linking assay. Substrate buffer with no added E1^E4 protein. Panel F: in situ cross-linking assay. E1^E4 protein and substrate buffer, but no primary antibody against the hexa-histidine tag added. Arrows in each panel indicate the basal layer of epithelium. All photomicrographs were taken at the same time exposure. Original magnification: 200.

layers of epithelium, only in the experiments in which calcium was added (Fig. 2, Panel C), suggested that non-specific binding of E1^E4 to keratins had not occurred, since keratins are abundant in all keratinocytes, including those in the basal layer of epithelium. No fluorescence was noted in the assay in which E1^E4 was omitted (Fig. 2, Panel E). Similarly, no significant fluorescence was observed in the assay in which anti-hexa-histidine antibodies were omitted (Fig. 2, Panel F). E1^ E4 is cross-linked in vitro by recombinant transglutaminase 3 We next performed in vitro cross-linking experiments using recombinant transglutaminases 1 and 3. Recombinant histagged loricrin was used as a substrate in control assays for

Fig. 3. In vitro cross-linking assays. Antibody used for all immunoblots in the cross-linking assay was a rabbit polyclonal antibody against the hexa-histidine tag. Panel A: in vitro cross-linking assay using his-tagged loricrin as a substrate. Lane 1: loricrin cross-linking assay in substrate buffer (Tris – HCl 10 mM, pH 8.4, calcium chloride 5 mM) with no added transglutaminase 3. Lane 2: loricrin cross-linking assay in substrate buffer with transglutaminase 3 added. Large arrows indicate ‘‘fast migrating dimer’’; small arrow indicates the ‘‘fast migrating monomer’’. Panel B: in vitro cross-linking assay using his-tagged E1^E4 as a substrate. Lane 1: his-tagged E1^E4 protein in Tris – HCl 10 mM, pH 8.4, shown as an indicator for the amount of E1^E4 added to cross-linking reactions. Lane 1 does not depict a cross-linking reaction. Lane 2: cross-linking assay of his-tagged E1^E4 protein in substrate buffer (Tris – HCl 10 mM, pH 8.4, calcium chloride 5 mM) with added transglutaminase 3. Lane 3: E1^E4 protein in Tris – HCl 10 mM, pH 8.4 with transglutaminase 3 added, but no added calcium chloride in the reaction. Lane 4: cross-linking assay of histagged E1^E4 protein in control buffer (Tris – HCl, 10 mM, pH 8.4, EDTA 1 mM) with transglutaminase 3 added. Large arrow indicates the position of the loading wells and the small arrow indicates the E1^E4 monomer. Panel C: E1^E4 protein, in amounts equal to the in vitro cross-linking reaction shown in Panel B, were incubated overnight at 37 -C without the addition of transglutaminase 3, in either distilled water (Lane 1) or Tris – HCl 10 mM, pH 8.4 (Lane 2). Large arrow indicates the position of the loading wells and the small arrow indicates the E1^E4 monomer. Molecular weight markers (in kDa) are indicated on the left side of the figure.

D.R. Brown et al. / Virology 345 (2006) 290 – 298

293

Fig. 4. Immunofluorescence microscopic analysis of expression of GFP (Panel A), GFP-E1^E4 (Panel B), or non-fused E1^E4 protein (Panel C) in undifferentiated NIKS cells transduced with recombinant retroviruses and grown in selective medium. All photomicrographs were taken at the same time exposure. Original magnification 400.

reactions containing added calcium (Fig. 3, Panel B, Lane 3). High molecular weight multimers were present, although monomeric E1^E4 was present in greater abundance than in the assay containing added calcium. Aggregates of approximately 30, 45, and 60 kDa were present in this reaction, which may have resulted from E1^E4 self-aggregation. In the reaction containing EDTA, a further reduction of cross-linking occurred, as evidenced by the increased presence of E1^E4 monomeric protein (Fig. 3, Panel B, Lane 4). Aggregates of 30, 45, and 60 kDa were also present in this reaction. Because E1^E4 can self-aggregate (Ashmole et al., 1998; Bryan et al., 1998; Roberts et al., 1994), E1^E4 protein, in amounts equal to the in vitro cross-linking reaction, were incubated overnight at 37 -C without the addition of transglutaminase, in either distilled water (Fig. 3, Panel C, Lane 1) or Tris –HCl 10 mM, pH 8.4 (Fig. 3, Panel C, Lane 2). We found no reduction in the abundance of E1^E4 monomer in these reactions. However, aggregates of approximately 30, 45, and 60 kDa were formed, similar to the reaction containing transglutaminase 3 and EDTA (Fig. 3, Panel B, Lane 4). Thus, the complexes at 30, 45, and 60 kDa were likely due to selfaggregation of E1^E4 protein rather than by cross-linking mediated by transglutaminase 3. E1^ E4 alters morphology of differentiated keratinocytes and cell envelopes If E1^E4 is cross-linked by keratinocyte transglutaminases, then the morphologic properties of cells transduced by retroviruses expressing E1^E4, and cornified cell envelopes from these cells may differ from control cells. We therefore

expressed E1^E4 in NIKS cells, induced the cells to differentiate by growth in medium containing 2 mM calcium chloride, isolated cell envelopes, and examined them for morphologic features. Undifferentiated NIKS cells transduced with retroviruses expressing GFP and selected with G418 demonstrated diffuse green fluorescence in nearly all cells (Fig. 4, Panel A). Cells transduced with retroviruses expressing GFP-E1^E4 demonstrated fluorescence in an intermediate filament distribution in nearly all cells (Fig. 4, Panel B). Cells transduced with retroviruses expressing non-fused E1^E4 were probed with an antibody raised against the E1^E4 protein. These cells demonstrated fluorescence in an intermediate filament distribution in nearly all cells, similar to, but less intense than that seen for GFP-E1^E4 expression (Fig. 4, Panel C). Differentiated NIKS cells expressing GFP, GFP-E1^E4, or E1^E4 were prepared by grown in medium containing 2 mM calcium chloride as described in Methods. Cells were harvested using a cell scraper, and a portion was fixed in zinc formalin, embedded in paraffin, and sections were prepared. One section from each preparation was stained with hematoxylin and eosin. Differentiated cells expressing GFP demonstrated appeared organized and uniform in size and shape (Fig. 5, Panel A). The nuclei of these cells were also uniform in size and shape. In contrast, differentiated cells expressing GFP-E1^E4 or nonfused E1^E4 were varied in size and shape (Fig. 5, Panels B and C). The nuclei of these cells were also varied in size and shape. Next, cornified cell envelopes were prepared from differentiated NIKS cells by heating to 80 -C in 1% SDS for 4 h, concentration by centrifugation, and analysis by phase contrast

Fig. 5. NIKS cells transduced with recombinant retroviruses and induced to differentiate. Light microscopic analysis. Panel A: cells transduced with retroviruses expressing GFP. Panel B: cells transduced with retroviruses expressing GFP-E1^E4. Panel C: cells transduced with retroviruses expressing non-fused E1^E4. Original magnification 200.

294

D.R. Brown et al. / Virology 345 (2006) 290 – 298

Fig. 6. Cell envelopes from NIKS cells transduced with recombinant retroviruses and induced to differentiate. Phase contrast analysis. Panel A: cell envelopes from cells transduced with retroviruses expressing GFP. Panel B: cell envelopes from cells transduced with retroviruses expressing GFP-E1^E4. Panel C: cell envelopes from cells transduced with retroviruses expressing non-fused E1^E4. Original magnification 400.

microscopy. Cell envelopes made from differentiated NIKS cells expressing GFP were consistently round or cuboidal (Fig. 6, Panel A). In contrast, cell envelopes from cells expressing GFP-E1^E4 or non-fused E1^E4 were small and rough bordered, and appeared fragmented in many cases (Fig. 6, Panels B and C). Discussion The mechanisms that facilitate release of newly formed HPV from the differentiated keratinocyte are not completely understood. HPV infection is not a lytic process, and therefore some other method must exist for newly formed virions to escape the infected cell. We previously showed that HPV infection weakens the cornified cell envelope, and that the composition of the cell envelope matrix is altered in HPV 11 infection; most notably with a reduction in loricrin and an increase in the small proline-rich proteins (Brown and Bryan, 2000; Lehr et al., 2002). Could E1^E4 have a role in release of mature HPV particles from the differentiated keratinocyte? The E1^E4 protein and the L1 major capsid protein are expressed from a bicistronic transcript (Brown et al., 1996; Chow et al., 1987; Nasseri et al., 1987). Interestingly, nearly all cells in natural lesions such as condylomata acuminata and low grade dysplastic lesions of the cervix containing detectable amounts of the L1 major capsid protein also contain detectable E1^E4 protein (Brown et al., 1994; Doorbar et al., 1997). In addition, the HPV 11 E1^E4 protein localizes to the region of the cell envelope, co-purifies with cell envelopes from these tissues, and is associated in vivo with the cell envelope structure in immunoelectron microscopic studies (Bryan and Brown, 2000). We show here that the HPV 11 E1^E4 protein can serve as a substrate for keratinocyte transglutaminases, the enzymes that cross-link precursor proteins including loricrin, involucrin, small proline rich proteins (SPRs), and cytokeratins to form the dense, insoluble cornified cell envelope (Aeschlimann and Paulsson, 1994; Candi et al., 1999; Friedrich et al., 1999; Kim et al., 1995; Steinert and Marekov, 1999; Steven and Steinert, 1994). Transglutaminases are expressed in differentiated epithelium and are dependent on the presence of calcium for optimal activity. Transglutaminases-mediated cross-linking results in the formation of large macromolecules, formed by inter- and intramolecular isopeptide bonding between glutamine and lysine residues (Kalinin et al., 2002; Nemes and Steinert, 1999; Steinert and Marekov, 1999). Transglutaminase

1 is membrane bound and is involved in early cross-linking events such as cross-linking of involucrin, a cell envelope ‘‘scaffold’’ protein. Transglutaminase 3 is soluble in the cytoplasm and is involved in later cross-linking events, such as cross-linking of loricrin. Transglutaminase substrates contain lysine or glutamine residues and are cytoplasmic in location. The HPV 11 E1^E4 protein is 90 amino acids in length, containing three lysine and three glutamine residues. E1^E4 proteins are detected in the cytoplasm of infected keratinocytes, and become increasingly abundant as the keratinocyte differentiates (Brown et al., 1994, 2004; Doorbar et al., 1986, 1997). Lysine and glutamine residues are present in E1^E4 proteins of HPV types 6, 11, 16, 18, 59, and 83. HPV 16 E1^E4 contains six lysine and five glutamine residues. Several of these lysine and glutamine residues are highly conserved in E1^E4 proteins of different HPV types. Thus, E1^E4 proteins appears to meet the basic requirements of a potential transglutaminase substrate. We also show here that expression of E1^E4 protein in differentiated human keratinocytes is associated with morphologic changes in the cells themselves, and in the cell envelopes that are derived from these cells. These morphologic differences were similar to those between healthy and HPV-infected genital epithelium, suggesting that E1^E4 expression in differentiated human keratinocytes may have had a specific effect on development of the cornified cell envelope. However, our studies do not yet provide definitive proof that E1^E4 protein is incorporated into the cornified cell envelope in natural infection. Further work needs to be done to prove that an isopeptide bond is formed with loricrin, an SPR, or other cell envelope protein, and which E1^E4 residues are involved. This definitive proof will require isolation, purification, and sequencing of small, cross-linked peptides isolated from trypsin-digested cross-linking reaction products, or purified cornified cell envelopes from infected tissues (Steinert and Marekov, 1997; Steinert et al., 1998). We have begun studies using these methodologies. Our studies do not prove that E1^E4 protein causes the abnormalities of cell envelopes seen in HPV infection. Such a cytopathic effect is likely to be multifactorial. In addition, it is likely that E1^E4 has numerous functions, based on prior studies of E1^E4 function (Davy et al., 2002; Doorbar et al., 1991, 2000; Nakahara et al., 2002; Roberts et al., 1994; Wang et al., 2004). The HPV 11 E1^E4, a protein expressed in differentiated cells of infected epithelium, appears to be a satisfactory substrate for transglutaminase 3, the enzyme

D.R. Brown et al. / Virology 345 (2006) 290 – 298

primarily responsible for incorporating loricrin and other cornified cell envelope components into the cornified cell envelope during the process of keratinocyte differentiation. This finding may explain the co-localization of the HPV 11 E1^E4 protein with the cell envelope in natural infection, and the co-purification of E1^E4 with cornified cell envelope fragments. As with all observations with E1^E4 proteins, further work is needed to determine the functional significance of our findings in natural HPV infection. Methods Cloning and expression of HPV 11 E1^ E4, loricrin, and transglutaminases The HPV 11 E1^E4 sequence was previously cloned by RTPCR using RNA purified from an HPV 11-infected human xenograft grown in an athymic mouse (Brown et al., 1994). The E1^E4 sequence was subcloned from pCR3.1 into pET28A (Novagen, Madison, WI). For expression of E1^E4 protein, competent BL21 (DE3) pLysS E. coli (Invitrogen, Carlsbad, CA) were transformed with the pET28A-E1^E4 11 according to manufacturer’s instructions. A single transformed colony was used to inoculate LB growth medium containing ampicillin (50 Ag/mL). The culture was grown in a shaking incubator at 37 -C to an optical density of 0.6 at 540 nm, at which time IPTG was added to a final concentration of 0.4 mM. The cells were grown for an additional 3 h in a shaking incubator at 37 -C. Cells were then harvested by centrifugation and expression verified by immunoblots using a rabbit polyclonal antibody directed against the hexa-histidine tag or antibodies specifically directed against the HPV 11 E1^E4 protein (Brown et al., 1994). E1^E4 protein was then purified from the cell extract under denaturing conditions using a nickel resin column (Ni-NTA His Bind) according to manufacturer’s instructions (Novagen), and dialyzed against 1 PBS. For loricrin cloning, a cDNA library was prepared from human neonatal foreskin tissue using the Smart PCR cDNA Synthesis Kit as directed by the manufacturer (Clontech, Palo Alto, CA). The loricrin gene was then amplified by from this library by PCR using the primers: 5V CAA GAT GTC TTC TCA GAA AAA GCA GCC CAC and 3V CTA TTT GGA CGG CCA GGT AGG CGC. Conditions used for the PCR were: 94 -C hold for 5 min, followed by 35 cycles of 94 -C for 1 min, 68 -C for two and one half minute, followed by a 5-min hold at 72 -C. The loricrin amplimer of approximately 1100 bp was purified from a gel fragment and cloned into the pCR3.1 vector (Invitrogen, Carlsbad, CA). The loricrin sequence was then subcloned into pET28A. The loricrin sequence and the inframe 5V hexa-histidine sequences were then subcloned from the pET28A vector into the pBacPAK9 vector (BD Biosciences, Palo Alto, CA). DNA sequencing was then performed on all clones in both directions to confirm correct amplification and cloning of the loricrin open reading frame. The transglutaminase 1 ORF was amplified by RT-PCR from the foreskin cDNA library using the primers: 5V GTCGACGTCCTTGCTCATCTGACTCC and 3V GTACTGCGGTTGCCAA-

295

CACT. The Expand Long Template PCR system (Roche Molecular Diagnostics, Carmel, IN) was used (with Buffer 1) to amplify transglutaminase 1 using the conditions: 94 -C hold for 2 min, followed by ten cycles of 94 -C for 20 s, 56 -C for 30 s, and 68 -C for 3 min, followed by three more rounds of ten cycles each, with an extra minute of extension at 68 -C for each of the three additional rounds. The transglutaminase 1 amplimer of approximately 2200 bp was purified from a gel fragment and cloned into the pXL-Topo II vector (Invitrogen). The transglutaminase 1 sequence was then subcloned into pET28A. The transglutaminase 1 sequence and the in-frame 5V hexa-histidine sequence were then subcloned into the pBacPAK9 vector (BD Biosciences, Palo Alto, CA). DNA sequencing was performed in both directions to confirm proper cloning of the transglutaminase 1 open reading frame. The transglutaminase 3 ORF was amplified by RT-PCR from the foreskin cDNA library using the primers: 5V TGGATCCGGATCCAACATGGCTGCTCTAGGAGT and 3V TGCGGCCGCGTCGACTGCTCCGTGTTGTCCAAGTT. The Expand Long Template PCR system (Roche Molecular Diagnostics) was used (with Buffer 1) to amplify transglutaminase 3 using the same conditions as above for transglutaminase 1. The transglutaminase 3 amplimer of approximately 2000 bp was purified from a gel fragment and cloned into the pXL-Topo II vector (Invitrogen). The transglutaminase 3 sequence were then subcloned into pET28A. The transglutaminase 3 sequence and the in-frame 5V hexa-histidine sequence was then subcloned from the pET28A vector into pBacPAK8 vector (BD Biosciences). DNA sequencing was performed in both directions to confirm proper cloning of the transglutaminase 3 open reading frame. Competent Sf21 insect cells (Clontech) were transformed with the pBacPAK9-loricrin, pBacPAK9-transglutaminase 1, or pBacPAK8-transglutaminase 3 plasmids as directed by the manufacturer. Cells were grown for 4 days at 27 -C and then harvested using PopCulture (with added Benzonase reagent at 0.4 AL/mL of original culture volume) according to the manufacturer’s instructions (Novagen). Protease inhibitor cocktail III (Calbiochem, San Diego, CA) was added at 0.1 AL/mL (original culture volume) to the lysates. Cells were then harvested by centrifugation and proteins purified in nondenaturing conditions using a nickel resin column as above according to manufacturer’s instructions (Novagen), then dialyzed against 1 PBS. Immunoblots were performed using a rabbit polyclonal antibody against the hexa-histidine tag to verify expression of E1^E4, loricrin, transglutaminase 1, and transglutaminase 3. In situ cross-linking assay The in situ cross-linking assay utilized frozen sections of human foreskin tissue derived from neonatal circumcisions. This assay has provided evidence of incorporation of numerous cellular proteins into genital epithelium, and presumably into the cornified cell envelope by the activity of transglutaminases (Akiyama et al., 2000; Raghunath et al., 1998). A portion of the

296

D.R. Brown et al. / Virology 345 (2006) 290 – 298

tissue was embedded in paraffin, and sections were prepared for histologic and immunohistochemical analysis as previously described, using a mouse monoclonal antibody against human transglutaminase 3 (Hitomi et al., 2003a, 2003b). For the cross-linking assay, cryostat sections (6 Am thick) were air dried for 10 min at room temperature, then incubated for 60 min at room temperature in blocking buffer (Tris – HCl 100 mM, pH 8.4, BSA 1%). The sections were washed three times in 1 phosphate-buffered saline (PBS), pH 8.4. The sections were incubated with substrate buffer (Tris –HCl 100 mM, pH 8.4, calcium chloride 5 mM) for 5 h at room temperature. The sections were then washed three times in 1 PBS. His-tagged HPV 11 E1^E4 protein was added to substrate buffer at an approximate concentration of 70 ng/AL, and 100 AL of substrate buffer with added E1^E4 protein was applied the tissue section. Following an overnight incubation at room temperature, the sections were incubated in stop buffer (1 PBS, EDTA 25 mM) for 10 min at room temperature. The sections were again washed three times with 1 PBS, and a primary rabbit polyclonal serum against the hexahistidine tag (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) was added (1:100 in 1 PBS) for 2 h at room temperature, followed by three additional washes with 1 PBS. A secondary swine anti-rabbit FITC-conjugated antibody (Dako Cytomation California Inc., Carpinteria, CA) was applied (1:100 dilution in 1 PBS) and the sections were incubated for 2 h at room temperature, followed by washing three times with 1 PBS. The sections were then mounted with Vectashield (Vector Laboratories, Burlingame, CA) and viewed under the fluorescent microscope at 488 nm wavelength. Photomicrographs were taken of all sections at the same exposure time. Because transglutaminases are dependent on calcium for activity, an experiment was performed using a control buffer (Tris –HCl 100 mM, pH 8.4, EDTA 5 mM) in place of the substrate buffer, which contained calcium chloride. To determine if the serum against the hexa-histidine tag bound nonspecifically to the tissue section, an additional control experiment was performed that included substrate buffer, but E1^E4 protein was omitted. Lastly, an experiment was performed that utilized E1^E4 protein in substrate buffer, omission of the serum against the hexa-histidine tag, followed by incubation in the FITC-conjugated secondary antibodies as indicated above. In vitro cross-linking assay The in vitro cross-linking assay was performed using nondenatured transglutaminase 1 or transglutaminase 3 to determine if HPV 11 E1^E4 protein could serve as a substrate for these enzymes. Loricrin was used as a control substrate for transglutaminase 1 or transglutaminase 3 cross-linking. This in vitro cross-linking assay has been used to provide evidence of incorporation of numerous cellular proteins into the cornified cell envelope by the activity of transglutaminases (Candi et al., 1995, 2001). The proenzyme of transglutaminase 3 (but not transglutaminase 1) requires activation by proteolytic cleavage before it effectively cross-links precursor molecules (Hitomi et al., 2003a, 2003b). Transglutaminase 3 was activated by adding

neutral dispase I (Roche Applied Science, Indianapolis, IN) to a concentration of 0.0015 units/AL and incubating 37 -C for 30 min. Immunoblots were then performed to demonstrate cleavage of transglutaminase 3 into active peptide fragments of the appropriate size. Transglutaminase 1 (or activated transglutaminase 3) and loricrin (or HPV 11 E1^E4) were added to a substrate buffer (Tris – HCl 10 mM, pH 8.4, calcium chloride 5 mM). Reactions were also performed using substrate buffer with no added calcium chloride, and using substrate buffer with EDTA 5 mM and no added calcium chloride. To provide further evidence that transglutaminase-mediated cross-linking was responsible for any observed reduced abundance of monomeric E1^E4 protein or the formation of high molecular weight complexes, E1^E4 protein in an amount equal to that in cross-linking reactions was incubated at 37 -C overnight in water or in Tris – HCl, 10 mM, pH 8.4. Equal volumes of all reactions were separated by SDSPAGE electrophoresis followed by transfer to nitrocellulose membrane Bio-Rad Laboratories, Hercules, CA) as previously described (Brown et al., 1994). Membranes were probed with polyclonal serum against the hexa-histidine tag to determine if a reduction in the abundance of monomeric loricrin or E1^E4 occurred, if a mobility shift could be detected, or if high molecular weight complexes formed. Retroviruses and expression of E1^ E4 in NIKS cells The GFP sequence, the GFP-E1^E4 sequence, or E1^E4 sequences (Bryan et al., 1998) were cloned into the pLNCX2 vector (Clontech, Palo Alto, CA). Plasmids were sequenced in both directions. Construction of vector packaging cell lines was performed at the Indiana University School of Medicine Vector Laboratory using established methods (Kotani et al., 1994; Markowitz et al., 1988). Briefly, each plasmid construct was transfected into the PG13 cell line (Miller et al., 1991). Supernatant containing the transiently expressed, GALV pseudotyped vector was collected at 24 h post-transfection media change. The harvested vector containing supernatant was then used to transduce the phoenix ampho packaging cell line (http://www.stanford.edu/group/nolan/MTAs/mtas.html). Transduction of the phoenix ampho cell line was carried out by doing three rounds of spinoculation (Kotani et al., 1994). The transduced phoenix ampho cell population (producer cells) was used to produce the retroviral vector used in the rest of the experiments reported here. Supernatants (containing retroviruses) from producer cells were frozen in growth medium (DMEM plus 10% calf serum). NIKS cells were grown to 60% confluence in Complete F medium in the presence of mitomycin C-treated J2 cells as previously described (Brown et al., 2004). Supernatants were thawed on ice, and 2 mL was added to 3 mL of complete F medium, then mixed well by inversion and added to NIKS cells. Polybrene (Sigma Aldrich Chemicals, St. Louis, MO) was added to the medium at a final concentration of 8 Ag/mL. Transduction was thus performed for 4 h at 37 -C in 5% CO2. The medium containing retroviruses was replaced after 4 h with

D.R. Brown et al. / Virology 345 (2006) 290 – 298

fresh Complete F medium. The transduction procedure was repeated for 4 h the next day. NIKS cells were split and replated onto mitomycin-treated J2 cells after 96 h. A drop of cells was grown on a glass cover slip each time NIKS cells were split for analysis of expression. Transduced NIKS cells were then grown to confluence in Complete F medium containing G418 200 Ag per mL, at which time 2 mM calcium chloride was added to induce differentiation. Medium was changed every 48 h. Cells were grown for 21 additional days, then harvested with a cell scraper. A portion of the cells was fixed with zinc formalin for preparation of paraffin-embedded sections. Some of the cells were used to prepare cornified cell envelopes as previously described (Brown and Bryan, 2000). Acknowledgments This study was funded by NIH (NIAID) AI51524. We thank Drs. Kiyotaka Hitomi and Masatoshi Maki (Nagoya University, Nagoya, Japan) for proving the monoclonal antibody to human transglutaminase 3, and Dr. Janine Bryan (Merck Research Laboratories) for critical reading of the manuscript and for her helpful comments. References Aeschlimann, D., Paulsson, M., 1994. Transglutaminases: protein cross-linking enzyme in tissues and body fluids. Thromb. Haemostasis 71, 402 – 415. Akiyama, M., Smith, L., Shimizu, H., 2000. Expression of transglutaminase activity in developing human epidermis. Br. J. Dermatol. 142 (2), 223 – 225. Ashmole, I., Gallimore, P.H., Roberts, S., 1998. Identification of conserved hydrophobic C-terminal residues of the human papillomavirus type 1 E1^E4 protein necessary for E4 oligomerisation in vivo. Virology 240 (2), 221 – 231. Brown, D., Bryan, J., 2000. Abnormalities of cornified cell envelopes isolated from human papillomavirus type 11-infected genital epithelium. Virology 271 (1), 65 – 70. Brown, D.R., Bryan, J.T., Cramer, H., Fife, K.H., 1993. Analysis of human papillomavirus types in exophytic condylomata acuminata by hybrid capture and Southern blot techniques. J. Clin. Microbiol. 31 (10), 2667 – 2673. Brown, D.R., Fan, L., Jones, J., Bryan, J., 1994. Colocalization of human papillomavirus type 11 E1^E4 and L1 proteins in human foreskin implants grown in athymic mice. Virology 201 (1), 46 – 54. Brown, D.R., Pratt, L., Bryan, J.T., Fife, K.H., Jansen, K., 1996. Virus-like particles and E1^E4 protein expressed from the human papillomavirus type 11 bicistronic E1^E4^L1 transcript. Virology 222 (1), 43 – 50. Brown, D., Brown, C., Lehr, E., 2004. Intracellular expression patterns of the human papillomavirus type 59 E1/E4 protein in COS cells, keratinocytes, and genital epithelium. Intervirology 47 (6), 321 – 327. Bryan, J., Brown, D., 2000. Association of the human papillomavirus type 11 E1^E4 protein with cornified cell envelopes derived from infected genital epithelium. Virology 277, 262 – 269. Bryan, J., Brown, D., 2001. Transmission of human papillomavirus type 11 infection by desquamated cornified cells. Virology 281, 35 – 42. Bryan, J.T., Fife, K.H., Brown, D.R., 1998. The intracellular expression pattern of human papillomavirus type 11 E1^E4 protein correlates with its ability to self associate. Virology 241 (1), 49 – 60. Candi, E., Melino, G., Mei, G., Tarcsa, E., Chung, S., Marekov, L., Steinert, P., 1995. Biochemical, structural, and transglutaminase substrate properties of human loricrin, the major epidermal cornified cell envelope protein. J. Biol. Chem. 270 (3), 26382 – 26390. Candi, E., Taresa, E., Idler, W., Kartasova, T., Marekov, L., Steinert, P., 1999.

297

Transglutaminase cross-linking properties of the small proline-rich 1 family of cornified cell envelope proteins. J. Biol. Chem. 274 (11), 7226 – 7237. Candi, E., Oddi, S., Terrinoni, A., Paradisi, A., Ranalli, M., Finazzi-Agro, A., Melino, G., 2001. Transglutaminase 5 cross-links loricrin, involucrin, and small proline-rich proteins in vitro. J. Biol. Chem. 276 (37), 35014 – 35023. Chow, L.T., Nasseri, M., Wolinsky, S.M., Broker, T.R., 1987. Human papillomavirus types 6 and 11 mRNAs from genital condylomata acuminata. J. Virol. 61, 2581 – 2588. Davy, C., Jackson, D., Wang, Q., Raj, K., Masterson, P., Fenner, N., Southern, S., Cuthill, S., Millar, J., Doorbar, J., 2002. Identification of a G(2) arrest domain in the E1^E4 protein of human papillomavirus type 16. J. Virol. 76 (19), 9806 – 9818. Doorbar, J., Campbell, D., Grand, R.J.A., Gallimore, P.H., 1986. Identification of the human papilloma virus-la E4 gene products. EMBO J. 5, 355 – 362. Doorbar, J., Ely, S., Sterling, J., McLean, C., Crawford, L., 1991. Specific interaction between HPV-16 E1 – E4 and cytokeratins results in collapse of the epithelial cell intermediate filament network. Nature 352, 824 – 827. Doorbar, J., Foo, C., Coleman, N., Medcalf, L., Hartley, O., Prospero, T., Napthine, S., Sterling, J., Winter, G., Griffin, H., 1997. Characterization of events during the late stages of HPV16 infection in vivo using high-affinity synthetic Fabs to E4. Virology 238 (1), 40 – 52. Doorbar, J., Elston, R.C., Napthine, S., Raj, K., Medcalf, E., Jackson, D., Coleman, N., Griffin, H.M., Masterson, P., Stacey, S., Mengistu, Y., Dunlop, J., 2000. The E1^E4 protein of human papillomavirus type 16 associates with a putative RNA helicase through sequences in its C terminus. J. Virol. 74 (21), 10081 – 10095. Eckert, R., Crish, J., Robinson, N., 1997. The epidermal keratinocyte as a model for the study of gene regulation and cell differentiation. Physiol. Rev. 77 (2), 397 – 424. Ferenczy, A., 1995. Epidemiology and clinical pathophysiology of condylomata acuminata. Am. J. Obstet. Gynecol. 172 (4 Part 2), 1331 – 1339. Ferenczy, A., Braun, L., Shah, K.V., 1981. Human papillomavirus (HPV) in condylomatous lesions of cervix. A comparative ultrastructural and immunohistochemical study. Am. J. Surg. Pathol. 5, 661 – 670. Firzlaff, J.M., Kiviat, N.B., Beckmann, A.M., Jenison, S.A., Galloway, D.A., 1988. Detection of human papillomavirus capsid antigens in various squamous epithelial lesions using antibodies directed against the L1 and L2 open reading frames. Virology 164, 467 – 477. Friedrich, M., Villena-Heinsen, C., He, J., Mink, D., Reichrath, J., Reitnauer, K., Schmidt, W., 1999. Expression of transglutaminase K in normal cervix tissue and cervix carcinomas. Histochem. J. 31 (1), 13 – 18. Gissmann, L., de Villiers, E.-M., zur Hausen, H., 1982. Analysis of human genital warts (condylomata acuminata) and other genital tumors for human papillomavirus type 6 DNA. Int. J. Cancer 29, 143 – 146. Hitomi, K., Ikeda, N., Maki, M., 2003a. Immunological detection of proteolytically activated epidermal-type transglutaminase (TGase 3) using cleavage-sitespecific antibody. Biosci. Biotechnol. Biochem. 67 (11), 2492 – 2494. Hitomi, K., Presland, R., Nakayama, T., Fleckman, P., Dale, B., Maki, M., 2003b. Analysis of epidermal-type transglutaminase (transglutaminase 3) in human stratified epithelia and cultured keratinocytes using monoclonal antibodies. J. Dermatol. Sci. 32 (2), 95 – 103. Kalinin, A., Kajava, A., Steinert, P., 2002. Epithelial barrier function: assembly and structural features of the cornified cell envelope. BioEssays 24 (9), 789 – 800. Kim, S., Chung, S., Yoneda, K., Steinert, P., 1995. Expression of transglutaminase 1 in human epidermis. J. Invest. Dermatol. 104, 211 – 217. Kotani, H., Newton III, P., Zhang, S., Chiang, Y., Otto, E., Weaver, L., Blaese, M., Anderson, W., McGarrity, G., 1994. Improved methods of retroviral vector transduction and production for gene therapy. Hum. Gene Ther. 5 (5), 19 – 28. Lehr, E., Brown, D., 2002. How do human papillomaviruses escape the infected keratinocyte? Papillomavirus Rep. 13, 181 – 189. Lehr, E., Jarnik, M., Brown, D., 2002. Human papillomavirus type 11 alters transcription and expression of loricrin, the major cell envelope protein. Virology 298, 240 – 247. Markowitz, D., Goff, S., Bank, A., 1988. Construction and use of a safe and efficient amphotropic packaging cell line. Virology 167, 400 – 406. Miller, A., Garcia, J., von Suhr, N., Lynch, C., Wilson, C., Eiden, M., 1991.

298

D.R. Brown et al. / Virology 345 (2006) 290 – 298

Construction and properties of retrovirus packaging cells based on gibbon ape leukemia virus. J. Virol. 65, 2220 – 2224. Morin, C., Braun, L., Casas-Cordero, M., Shah, K.V., Roy, M., Fortier, M., Meisels, A., 1981. Confirmation of the papillomavirus etiology of condylomatous cervix lesions by the peroxidase – antiperoxidase technique. J. Natl. Cancer Inst. 66, 831 – 835. Nakahara, T., Nishimura, A., Tanaka, M., Ueno, T., Ishimoto, A., Sakai, H., 2002. Modulation of the cell division cycle by human papillomavirus type 18 E4. J. Virol. 76 (21), 10914 – 10920. Nasseri, M., Hirochika, R., Broker, T.R., Chow, L.T., 1987. A human papillomavirus type 11 transcript encoding an E1^E4 protein. Virology 159, 433 – 439. Nemes, Z., Steinert, P., 1999. Bricks and mortar of the epidermal barrier. Exp. Mol. Med. 31 (1), 5 – 19. Raghunath, M., Hennies, H., Velten, F., Wiebe, V., Steinert, P., Reis, A., Traupe, H., 1998. A novel in situ method for the detection of deficient transglutaminase activity in the skin. Arch. Dermatol. Res. 290 (11), 621 – 627. Reichert, U., Michel, S., Schmidt, R., 1993. The cornified envelope: a key structure of terminally differentiating keratinocytes. In: Darmon, M., Blumenberg, M. (Eds.), Molecular Biology of the Skin: The Keratinocyte. Academic Press, New York, NY. 291 pp. Rice, R., Green, H., 1977. The cornified envelope of terminally differentiated human epidermal keratinocytes consists of cross-linked protein. Cell 11, 417 – 422.

Roberts, S., Ashmole, I., Gibson, L.J., Rookes, S.M., Barton, G.J., Gallimore, P.H., 1994. Mutational analysis of human papillomavirus E4 proteins: identification of structural features important in the formation of cytoplasmic E4 cytokeratin networks in epithelial cells. J. Virol. 68 (10), 6432 – 6445. Steinert, P., Marekov, L., 1997. Direct evidence that involucrin is a major early isopeptide cross-linked component of the keratinocyte cornified cell envelope. J. Biol. Chem. 272 (17), 2021 – 2030. Steinert, P., Marekov, L., 1999. Initiation of assembly of the cell envelope barrier structure of stratified squamous epithelia. Mol. Biol. Cell 10 (12), 4247 – 4261. Steinert, P., Candi, E., Kartasova, T., Marekov, L., 1998. Small proline-rich proteins are cross-bridging proteins in the cornified cell envelopes of stratified squamous epithelia. J. Struct. Biol. 122, 76 – 85. Steven, A., Steinert, P., 1994. Protein composition of cornified cell envelopes of epidermal keratinocytes. J. Cell Sci. 107 (Pt. 2), 693 – 700. Toki, T., Oikawa, N., Tase, T., Satoh, S., Wada, Y., Yajima, A., 1986. Immunohistochemical demonstration of papillomavirus antigen in cervical dysplasia and vulvar condyloma. Gynecol. Obstet. Invest. 22, 97 – 101. Wang, Q., Griffin, H., Southern, S., Jackson, D., Martin, A., McIntosh, P., Davy, C., Masterson, P.J., Walker, P.A., Laskey, P., Omary, M.B., Doorbar, J., 2004. Functional analysis of the human papillomavirus type 16 E1^E4 protein provides a mechanism for in vivo and in vitro keratin filament reorganization. J. Virol. 78, 821 – 833.