Requirement for the E1 Helicase C-Terminal Domain in Papillomavirus DNA Replication In Vivo

Abstract

The papillomavirus (PV) E1 helicase contains a conserved C-terminal domain (CTD), located next to its ATP-binding site, whose function in vivo is still poorly understood. The CTD is comprised of an alpha helix followed by an acidic region (AR) and a C-terminal extension termed the C-tail. Recent biochemical studies on bovine papillomavirus 1 (BPV1) E1 showed that the AR and C-tail regulate the oligomerization of the protein into a double hexamer at the origin. In this study, we assessed the importance of the CTD of human papillomavirus 11 (HPV11) E1 in vivo, using a cell-based DNA replication assay. Our results indicate that combined deletion of the AR and C-tail drastically reduces DNA replication, by 85%, and that further truncation into the alpha-helical region compromises the structural integrity of the E1 helicase domain and its interaction with E2. Surprisingly, removal of the C-tail alone or mutation of highly conserved residues within the domain still allows significant levels of DNA replication (55%). This is in contrast to the absolute requirement for the C-tail reported for BPV1 E1 in vitro and confirmed here in vivo. Characterization of chimeric proteins in which the AR and C-tail from HPV11 E1 were replaced by those of BPV1 indicated that while the function of the AR is transferable, that of the C-tail is not. Collectively, these findings define the contribution of the three CTD subdomains to the DNA replication activity of E1 in vivo and suggest that the function of the C-tail has evolved in a PV type-specific manner.

Importance: While much is known about hexameric DNA helicases from superfamily 3, the papillomavirus E1 helicase contains a unique C-terminal domain (CTD) adjacent to its ATP-binding site. We show here that this CTD is important for the DNA replication activity of HPV11 E1 in vivo and that it can be divided into three functional subdomains that roughly correspond to the three conserved regions of the CTD: an alpha helix, needed for the structural integrity of the helicase domain, followed by an acidic region (AR) and a C-terminal tail (C-tail) that have been shown to regulate the oligomerization of BPV1 E1 in vitro. Characterization of E1 chimeras revealed that, while the function of the AR could be transferred from BPV1 E1 to HPV11 E1, that of the C-tail could not. These results suggest that the E1 CTD performs multiple functions in DNA replication, some of them in a virus type-specific manner.

FIG 1

Schematic representation of the papillomavirus E1 helicase highlighting the conservation of the C-terminal domain. (A) Diagram of the E1 protein showing the locations of the N-terminal regulatory region, OBD, Oligo, and ATPases associated with diverse cellular activities (AAA+) SF3 helicase/ATPase domain [ATP(SF3)] and of the ∼45-amino-acid-long CTD, which encompasses both the last alpha helix resolved in the available E1 crystal structures (helix α9) and the region previously called the flexible brace or C-terminal module, itself comprised of an acidic region (Acidic) and a C-tail subdomain separated by a linker sequence (10, 16, 17). (B) Amino acid sequence alignment of the E1 C termini from the indicated papillomavirus types (amino acid boundaries are indicated in parentheses). A large alignment of all of the E1 reference sequences deposited in the PAVE database was generated, but only those from BPV1 and 15 selected HPV types are presented for brevity. Residues are highlighted with increasingly darker shades of gray according to their degrees of conservation. The bottom of the alignment indicates the last region of the E1 C terminus that is resolved in the crystal structures of the E1 helicase domain from BPV1 (up to aa 577 and 579 in PBD structures 2V9P and 2GXA, respectively) and from HPV18 in complex with the E2 transactivation domain (PDB 1TUE) (18–20). While the major contacts between HPV18 E1 and E2 involve residues located in the ATPase domain, two potential minor contact points were identified within the C terminus (arrowheads) (20). (C) Amino acid sequence logo of the E1 C terminus generated from the alignment of all E1 reference sequences in the PAVE database (33).

FIG 2

Luciferase reporter cell-based assay for HPV11 DNA replication. (A) Principle of the HPV11 DNA replication assay. Expression vectors for GFP-tagged HPV11 E1 (p11E1) and triple-Flag-tagged HPV11 E2 (p11E2) are transfected into C33A cells, together with a plasmid containing the HPV11 minimal origin of replication (ori) linked in cis to an FLuc reporter gene (pFLORI11). A nonreplicating plasmid (pRL) encoding RLuc is used as an internal control (ctl). In this assay, replication of the origin-containing plasmid pFLORI11 is detected as an E1- and E2-dependent increase in FLuc activity relative to the level of RLuc expression from the control plasmid, pRL. DNA replication levels are presented as FLuc/RLuc activity ratios measured using a dual-luciferase assay. (B) Dependence of the HPV11 DNA replication assay on E1 and E2 expression as a function of time. DNA replication levels are presented as FLuc/RLuc ratios, measured in C33A cells that were cotransfected with 2.5 ng of pFLORI11 and 0.5 ng of pRL, together with (+) or without (−) 10 ng of p11E1 (E1) and 10 ng p11E2 (E2), as indicated. DNA replication levels were measured at different time points posttransfection (24, 48, and 72 h). Standard deviations are indicated by the error bars. Control assays performed in the absence of p11E1, p11E2, or both expression vectors showed no viral DNA replication. (C) Validation of the assay with replication-defective E1 mutant proteins. Shown are DNA replication activities supported by the indicated E1 proteins carrying deleterious amino acid substitutions in the OBD (K286A/R288A), Oligo (F393A), Walker A box (ATP) (K484A), and β-hairpin (β-HP) (K551A/H552A). DNA replication levels were measured from cells transfected with three different amounts of E1 expression plasmid (2.5, 5, and 10 ng) or with an empty vector as a negative control (No E1). Replication activity was measured 72 h posttransfection and is reported as a percentage of the FLuc/RLuc ratio obtained with 10 ng of wild-type E1 expression plasmid (WT), which was set to 100%. Standard deviations are indicated. Statistical significance was assessed by comparing the DNA replication activity of each E1 mutant protein to that of wild-type E1 (white bars), using one-way ANOVA followed by Dunnett's post hoc analysis. Significant differences are indicated (****, P ≤ 0.0001). (D) Expression levels of the indicated E1 mutant proteins compared to wild-type E1. Extracts from transfected cells were separated on an SDS-12% PAGE gel prior to immunoblotting with an anti-GFP antibody. Tubulin (Tub) was used as a loading control.

FIG 3

The C terminus of E1 contains three subdomains that play roles in HPV11 DNA replication. (A) Schematic representation and amino acid sequence of the C terminus of HPV11 E1. The boxes indicate the amino acid residues retained in the C-terminally truncated E1 proteins; 1 to 649 refers to the wild-type protein. The boxes are shaded according to the levels of DNA replication (rep) supported by each E1 protein (as shown in panel B), with the darkest shade indicating the highest level of DNA replication and an open box indicating background levels of activity (<10%). (B) The DNA replication activities of the indicated E1 proteins were evaluated as described in the legend to Fig. 2, using three amounts of E1 expression vector (2.5, 5, and 10 ng). DNA replication levels were measured 72 h posttransfection and are reported as percentages of the levels obtained with 10 ng of wild-type E1 expression plasmid (WT). Statistical significance was assessed by comparing the DNA replication activity of each truncated E1 protein to that of the preceding (i.e., 5-amino-acid-longer) deletion using one-way ANOVA followed by Bonferroni's post hoc analysis. Significant differences are indicated (*, P ≤ 0.05; ****, P ≤ 0.0001). (C) Expression of GFP-tagged wild-type E1 and truncated derivatives. Extracts from transfected cells were separated on an SDS-10% PAGE gel prior to immunoblotting with an anti-GFP antibody. Tubulin was used as a loading control.

FIG 4

The DNA replication activity of E1 is reduced by C-terminal deletions that remove the AR and abolished by those that encroach into helix 9. (A) Representation and amino acid sequence of a portion of the E1 C terminus encompassing the end of helix α9, the linker region, and the acidic region. The C-terminal boundaries of the different truncated E1 proteins are indicated. (B) Deletions into the AR. Shown are the DNA replication activities of E1 proteins ending between residues 629 and 624. In addition to lacking the C-tail, these proteins lack increasingly larger portions, in single-amino-acid increments, of the AR. The activity of each protein was determined using three amounts of E1 expression vector (2.5, 5, and 10 ng) 72 h posttransfection and is reported as a percentage of the activity obtained with 10 ng of wild-type E1 expression plasmid (WT), as described in the legend to Fig. 3B. Statistical significance was assessed by comparing the DNA replication activity of each E1 protein to that of E1(1–629) (white bars) using one-way ANOVA followed by Dunnett's post hoc analysis. Significant differences are indicated (**, P ≤ 0.01; ****, P ≤ 0.0001). (C) Expression of GFP-tagged wild-type E1 and truncated derivatives. Extracts from transfected cells were separated on an SDS-10% PAGE gel prior to immunoblotting with an anti-GFP antibody. Tubulin was used as a loading control. (D) Deletions into the linker region and helix 9. Shown are the DNA replication activities of truncated E1 proteins lacking increasingly larger portions of the linker region and helix 9. DNA replication levels and statistical significance relative to E1(1–624) (white bars) were determined as described for panel A. (E) Expression of the indicated GFP-E1 proteins was determined as described for panel C.

FIG 5

C-terminal deletions that extend into helix 9 reduce interaction with E2. (A) LUMIER E1-E2 coimmunoprecipitation assay. Expression vectors for GFP-tagged wild-type E1 and truncated derivatives, as well as for the full-length L617D mutant protein, were cotransfected with a vector encoding E2 fused to Renilla luciferase (RLuc-E2). An empty vector encoding GFP alone (No E1) was used as a negative control. For each E1 protein, the amount of RLuc-E2 that was coprecipitated by GFP-E1 was determined by measuring the levels of Renilla luciferase activity and normalized to the amount of RLuc-E2 present in the input cellular extract. The amount of RLuc-E2 coprecipitated by wild-type GFP-E1 was set at 100%. Each bar represents the average of three independent experiments, each performed in duplicate. Standard deviations are indicated by the error bars. Statistical significance was assessed by comparing the DNA replication activity of each E1 protein to that of wild-type GFP-E1 (white bar) using one-way ANOVA followed by Dunnett's post hoc analysis. Significant differences are indicated (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001). (B) Intracellular localization of the indicated GFP-E1 proteins in the presence of E2 (unless otherwise indicated). C33A cells transiently expressing the wild-type or truncated E1 proteins were visualized by fluorescence confocal microscopy. DNA was stained with DAPI to visualize the cell nuclei. Note the presence of GFP-E1 in E2-induced nuclear foci and the absence of such foci for E1(1–615) and E1(1–614). (C) Diagram summarizing the DNA replication and E2 interaction activities of the 19 truncated E1 proteins characterized in this study.

FIG 6

Single-amino-acid substitutions in the E1 AR have little effect on HPV11 DNA replication. (A) Representation and amino acid sequence of the C terminus of HPV11 E1. The 5 residues within the AR that were subjected to mutagenesis are shaded in black. Also indicated is the fact that serine 626 is a putative CK2 phosphorylation site. (B) Mutant E1 proteins were evaluated for HPV11 DNA replication activity (as for Fig. 3B). Only the mutant proteins that had two or three amino acid substitutions showed a decrease in activity, albeit moderate (10 to 20%). Statistical significance was assessed by comparing the DNA replication activity of each E1 mutant protein to that of wild-type E1 (white bars) using one-way ANOVA followed by Dunnett's post hoc analysis. Significant differences are indicated (*, P ≤ 0.05; ***, P ≤ 0.001). Standard deviations are indicated by the error bars. (C) Expression of GFP-tagged wild-type E1 and mutant derivatives. Extracts from transfected cells were separated on an SDS-12% PAGE gel prior to immunoblotting with an anti-GFP antibody. Tubulin was used as a loading control.

FIG 7

Replacement of the highly conserved F638 and C640 residues in HPV11 E1 abrogates the function of the C-tail. (A) Representation and amino acid sequence of the C terminus of HPV11 E1 highlighting the two most highly conserved amino acid residues in the C-tail, F638 and C640. (B) Mutant E1 proteins carrying the indicated amino acid substitutions at F638 and/or C640 were evaluated for HPV11 DNA replication activity (as for Fig. 3B). The activity of E1(1–634), which lacks a functional C-tail, is shown for comparison. Statistical significance was assessed by comparing the DNA replication activity of each E1 protein to that of wild-type E1 (white bars) using one-way ANOVA followed by Dunnett's post hoc analysis. Significant differences are indicated (**, P ≤ 0.01). (C) Expression of GFP-tagged wild-type E1 and mutant derivatives. Extracts from transfected cells were separated on an SDS-12% PAGE gel prior to immunoblotting with an anti-GFP antibody. Tubulin was used as a loading control.

FIG 8

Replacement of the highly conserved F594 and C596 residues in BPV1 E1 indicates an essential role for the C-tail in BPV1 DNA replication. (A) Depiction and amino acid sequence of the BPV1 E1 C terminus highlighting the two most highly conserved amino acid residues in the C-tail, F594 and C596, analogous to F638 and C640 in HPV11 E1 (Fig. 7). (B) The DNA replication activities of mutant BPV1 E1 proteins carrying the F594A and C596A substitutions were tested for the ability to support BPV1 DNA replication using a luciferase-based assay (25) similar to the one described in this study for HPV11 and using three amounts of BPV1 E1 expression vector (2.5, 5, and 10 ng) 72 h posttransfection. DNA replication levels are reported as percentages of the activity obtained with 10 ng of wild-type BPV1 E1 expression plasmid (WT). Note the more profound effect of the two C-tail substitutions on the activity of BPV1 E1 compared to HPV11 (Fig. 7B). Statistical significance was assessed by comparing the DNA replication activity of each BPV1 E1 mutant protein to that of wild-type E1 (white bars) using one-way ANOVA followed by Dunnett's post hoc analysis. Significant differences are indicated (****, P ≤ 0.0001). (C) Dominant-negative inhibition of BPV1 DNA replication by E1 F638A and E1 C640A. BPV1 DNA replication was performed using 10 ng of wild-type E1 expression plasmid and increasing amounts of expression vector (0, 3.13, 6.26, 12.5, 25, 50, 100, and 200 ng) for either E1 F638A or E1 C640A, as indicated. (D) Abilities of the WT and mutant BPV1 E1 proteins to support DNA replication using the origin and/or E2 from HPV11. HPV11 or BPV1 DNA replication assays were performed with 5 and 10 ng of BPV1 E1 expression plasmid (wild-type or mutant proteins), together with 2.5 ng of origin plasmid (Ori) and 10 ng of E2 expression plasmid from either BPV1 or HPV11, as indicated at the top. DNA replication levels were determined 72 h posttransfection and are reported as percentages of the activity measured with wild-type E1, E2, and the origin from BPV1. Statistical significance was assessed by comparing the DNA replication activity of each BPV1 E1 mutant protein to that of wild-type E1, using one-way ANOVA followed by Dunnett's post hoc analysis. Significant differences are indicated (****, P ≤ 0.0001). Standard deviations are indicated by the error bars.

FIG 9

The C terminus of HPV11 E1 can be partially replaced by the analogous domain from BPV1 E1. (A) Depiction and amino acid sequence of the C termini of the two HPV11/BPV1 E1 chimeras created in this study. The regions highlighted in black are from HPV11 E1, while those in gray are from BPV1 E1. Chimera 617 (Ch617) is comprised of residues 1 to 617 of HPV11 E1 fused to amino acids 572 to 605 of BPV1 E1. Ch620 contains residues 1 to 120 of HPV11 E1 fused to amino acids 575 to 605 of BPV1 E1. Mutant derivatives of both chimeras were also created by replacing the two conserved C-tail residues F594 and C596 (boxed) with alanines. (B) DNA replication levels supported by the Ch617 and Ch620 chimeras and mutant derivatives were measured using the HPV11 DNA replication assay essentially as described in the legend to Fig. 3B but using 2.5, 1.25, and 0.625 ng of E1 expression vector. DNA replication activities were measured 72 h posttransfection and are reported as percentages of the levels obtained with 2.5 ng of wild-type HPV11 E1 expression plasmid [E1(1–649)]. Also presented are the levels of DNA replication supported by HPV11 truncated E1 proteins ending at residues 617 and 620 to depict the “baseline” DNA replication value for each chimera in the absence of the BPV1 E1 C terminus. Statistical significance was assessed by comparing the DNA replication activity of each chimera to that of its parental truncated protein (white bars) using one-way ANOVA followed by Dunnett's post hoc analysis. Significant differences are indicated (***, P ≤ 0.001; ****, P ≤ 0.0001). (C) Expression of the indicated E1 proteins and chimeras. Extracts from transfected cells were separated on an SDS-10% PAGE gel prior to immunoblotting with an anti-GFP antibody. Tubulin was used as a loading control.

FIG 10

Summary of the data and homology models of the E1 helicase domain. (A) Schematic representation of the HPV11 E1 core domain (aa 1 to 617) fused to the linker sequence, AR, and C-tail domain. Helix 9, which is required for the proper folding of the helicase domain, is diagrammed within the E1 core region. Also shown is a summary of the results presented here for the effects of removing the C-tail, AR, and linker sequence and of deleting part of helix 9 on the ability of HPV11 E1 to support viral DNA replication and to interact with E2 (++++, like wild-type E1; ++, ∼50% reduction; +, ∼75% reduction; −, inactive). (B) Homology model of the HPV11 E1 HD based on the structures of the analogous regions from BPV1 and HPV18 (18–20). A cartoon representation of the AR (red) and C-tail (green), drawn to scale, is shown in an extended conformation to suggest how far in space these two regions could possibly reach. The arrows are meant to represent the flexible nature of the E1 C terminus. (C) Model of the HPV11 E1 HD in its hexameric form. ADP is colored magenta. (D) Model of the HPV11 E1 HD (blue) in complex with the E2 TAD (gray) and Brd4 helix (magenta). The model is based on the structures of the HPV18 E1-E2 complex and the HPV18 E2-Brd4 complex (35). The C-terminal alpha helix of Brd4, which binds on the opposite face of the E2 TAD (gray), is colored magenta. The inset shows the interaction of helix 9 (orange) with the E2 TAD. Also shown in orange is the side chain of arginine 616 that likely contacts the E2 TAD directly, as observed for the analogous residue (arginine 622) in the HPV18 E1-E2 TAD structure (20).