The minor capsid protein L2 contributes to two steps in the human papillomavirus type 31 life cycle

Abstract

Prior studies, which have relied upon the use of pseudovirions generated in heterologous cell types, have led to sometimes conflicting conclusions regarding the role of the minor capsid protein of papillomaviruses, L2, in the viral life cycle. In this study we carry out analyses with true virus particles assembled in the natural host cell to assess L2's role in the viral infectious life cycle. For these studies we used the organotypic (raft) culture system to recapitulate the full viral life cycle of the high-risk human papillomavirus HPV31, which was either wild type or mutant for L2. After transfection, the L2 mutant HPV31 genome was able to establish itself as a nuclear plasmid in proliferating populations of poorly differentiated (basal-like) human keratinocytes and to amplify its genome to high copy number, support late viral gene expression, and cause formation of virus particles in human keratinocytes that had been induced to undergo terminal differentiation. These results indicate that aspects of both the nonproductive and productive phases of the viral life cycle occur normally in the absence of functional L2. However, upon the analysis of the virus particles generated, we found an approximate 10-fold reduction in the amount of viral DNA encapsidated into L2-deficient virions. Furthermore, there was an over-100-fold reduction in the infectivity of L2-deficient virus. Because the latter deficiency cannot be accounted for solely by the 10-fold decrease in encapsidation, we conclude that L2 contributes to at least two steps in the production of infectious virus.

FIG. 1.

Establishment of human keratinocyte cell populations harboring wild-type or L2 mutant HPV31. (A) Mutation of the L2 open reading frame. AAA of codon 4 was mutated to TAA to introduce a stop codon. ATG of codon 51 was mutated to TTG to eliminate the possibility of reinitiation of translation. (B) Southern analysis of total genomic DNA from clonal NIKS populations transfected with HPV31 DNA. Equal amounts of total genomic DNA were loaded for each cell line. The three lanes labeled WT HPV31a are from three different wild-type clonal populations, and the lanes labeled L2 Mut HPV31a are from three different L2 mutant clonal populations. Left panel: DNA was loaded uncut. Right panel: DNA was cut with a restriction enzyme that linearized the HPV31 genome. Position of open circular, supercoiled, and linear HPV31 DNA bands are indicated based upon the migration pattern of the positive control DNA (CIN-612 9E) loaded in the second lane on each gel and based upon molecular weight standards (not shown). The negative control was total genomic DNA isolated from untransfected NIKS cells (lane 1 in each gel).

FIG. 2.

Immunohistochemical analyses of rafts. (A to D) Immunohistochemistry for BrdU incorporation. (E to H) FISH with fluorescein isothiocyanate-conjugated antibody specific for DIG-labeled HPV31 DNA probe. (I to L) Immunohistochemistry for L1 expression with H16.D9 antibody. (M to P) Immunohistochemistry for L2 expression with αHPV31-L2 antibody. For all panels displaying immunohistochemical staining (A to D and I to P), brown cells are positive and are indicated with blue arrows. Counterstain was done with hematoxylin. For all panels displaying immunofluorescent staining (E to H), cells positive for FISH signal have green nuclei (white arrows), whereas FISH-negative cells have blue nuclei (a consequence of the DAPI counterstain). Sections present in the first column (A, E, I, and M) are taken from the negative control NIKS rafts, those in the second column (B, F, J, and N) are from the positive control CIN-612 9E rafts, those in the third column (C, G, K, and O) are from wild-type (WT) HPV31a-harboring rafts, and those in the fourth column (D, H, L, and P) are from L2 mutant HPV31a-harboring rafts.

FIG. 3.

L2 is not needed to induce suprabasal DNA synthesis or for amplification of viral DNA. (A) Quantification of suprabasal BrdU incorporation. Bars show percent BrdU-positive nuclei that are supraparabasal, relative to the average value obtained for rafts harboring wild-type HPV31. (B) Quantification of the amplification of viral DNA. Bars represent percent nuclei (detected by DAPI) that are positive for amplified viral DNA relative to the average value obtained for rafts harboring wild-type HPV31. These data were quantified as described in Materials and Methods.

FIG. 4.

Analysis of crude preparations of virus. (A and B) Immunoblot (Western) analyses of equal volumes of crude virus preparations from rafts of three different clonal NIKS populations each harboring wild-type (WT; lanes 2 to 4) or L2 mutant (L2 Mut; lanes 5 to 7) HPV31a genomes, along with a sample prepared in parallel from rafts of the negative control (untransfected) NIKS cell line (lane 1). Numbers at left indicate molecular mass in kilodaltons. In panel A, L1 was detected using the monoclonal antibody H16.D9. Note the lower levels of L1 detected in lanes 2 and 7, which reflect these particular samples yielding less virus. In panel B, L2 was detected using a polyclonal antibody to HPV31 L2. The arrow indicates the L2-specific band. Note the presence of multiple nonspecific bands that migrated faster than the L2-specific band and are present in the negative control sample isolated from rafts of untransfected NIKS (lane 1). (C) TEMs of particles of about 55 nm in diameter. The side of each square is 100 nm long. Images are shown at 110,00× magnification. No particles of 55 nm in diameter were seen in the crude preps of NIKS cells.

FIG. 5.

Virus particles lacking L2 contain less DNase-resistant DNA. (A) Representative Southern blot of DNase-resistant DNA obtained from virus preps harvested from rafts of NIKS harboring wild-type (lanes 2 and 3) or L2 mutant (lanes 4 and 5) genomes. Lane 1, bacterially synthesized plasmid DNA that was spiked into the preparations of virus prior to DNase treatment to evaluate the completeness of DNase digestion. Note the absence of spike DNA in lanes 2 to 5. (B) Bar graph providing the quantification of the relative levels of DNase-resistant viral DNA in virus preps from rafts harboring wild-type or L2 mutant HPV31a genomes. Results are provided for three independent experiments. In each experiment, DNA was quantified from phosphorimager analyses of Southern analyses (an example of which is shown in panel A), evaluating levels of DNase-resistant DNA in virus preps from rafts of one or more independent clonal NIKS populations harboring either wild-type (dark bars) or L2 mutant (light bars) HPV31a genomes. For each experiment, the average level of DNase-resistant DNA in virus preps for rafts harboring wild-type HPV31a genomes was set to 100. Based upon the values obtained in these three independent analyses, the average relative amount of DNase-resistant DNA in the L2 mutant virus preps was 8.1% of that observed in the wild-type virus preps.

FIG. 6.

L2 is required for production of infectious particles. Shown are RT-PCR (A) and Taqman PCR (B) analyses of RNA isolated from HaCaT cells exposed to virus preps isolated from rafts of NIKS cells harboring wild-type (WT) or L2 mutant (L2 Mut) HPV31a genomes, untransfected NIKS cells harboring no viral genomes (NIKS), or the positive control cell line harboring HPV31b genomes (CIN-612 9E). (A) RT-PCR of spliced early viral transcripts, E1^E4 and E1*I, E2, and control PCR for β-actin transcripts. The three lanes of each genotype represent 10-fold serial dilutions of virus; the WT, L2 Mut, and CIN612 9E samples used had equivalent levels of L1 protein as judged by quantitative L1 Western blot analysis. For the negative control sample (NIKS), an equivalent volume of sample was that used for the other samples. A loading error accounts for the absence of the B-actin signal in one of the positive control (CIN612 9E) RT-PCR samples. (B) Quantitative RT-PCR. Tenfold dilutions of virus preps, quantified by Western blotting for L1, were used in independent infections of HaCaT cells. Equal amounts of total cellular RNA were used in RT, and triplicate cDNA samples were subjected to qPCR for E1^E4 transcripts. Relative levels of E1^E4 transcripts were determined by comparison with copy number standards. Values obtained from infections with wild-type virus are indicated by the box symbol, and for L2 mutant virus they are indicated by the triangle symbol; error bars represent standard errors of the means. Note that the level of RNA quantified in the cells infected with L2 mutant virus was not above background levels (i.e., they gave threshold values no greater than that seen with RNA from mock infections). Shown is one representative experiment. Similar results were obtained with other independent experiments using independent sources of wild-type and L2 mutant virus preps. In all cases wild-type virus gave robust RT-qPCR signals, whereas L2 mutant virus gave signals not above background levels seen with mock infections.