The E1 proteins

Abstract

E1, an ATP-dependent DNA helicase, is the only enzyme encoded by papillomaviruses (PVs). It is essential for replication and amplification of the viral episome in the nucleus of infected cells. To do so, E1 assembles into a double-hexamer at the viral origin, unwinds DNA at the origin and ahead of the replication fork and interacts with cellular DNA replication factors. Biochemical and structural studies have revealed the assembly pathway of E1 at the origin and how the enzyme unwinds DNA using a spiral escalator mechanism. E1 is tightly regulated in vivo, in particular by post-translational modifications that restrict its accumulation in the nucleus. Here we review how different functional domains of E1 orchestrate viral DNA replication, with an emphasis on their interactions with substrate DNA, host DNA replication factors and modifying enzymes. These studies have made E1 one of the best characterized helicases and provided unique insights on how PVs usurp different host-cell machineries to replicate and amplify their genome in a tightly controlled manner.

Keywords: ATPase; DNA replication; E1; Episome; Helicase; Papillomavirus; Phosphoryaltion; Post-translational modifications.

Figure 1. Domain structure of E1

Diagram of the BPV1 E1 protein highlighting the functions and amino acid boundaries of its various domains. The N-terminal regulatory region is shown in light purple while the domains required for viral DNA replication in vitro are colored in deep purple. The diagram shows the locations of the bipartite nuclear localization signal (NLS), origin binding domain (DBD), minimal oligomerization domain (O), AAA+ ATP-binding domain and C-terminal brace. The latter three regions comprise the helicase domain (HD), as indicated. Below the HD are shown the positions of the Walker A (WA), Walker B (WB), β-hairpin (β), sensor 1, 2 and 3 (S1, S2 and S3, respectively), and arginine-finger (R). The locations of the four regions (A-D) of homology with the large T antigen of SV40 are also indicated. The bottom part of the figure summarizes the main functions of the DBD and HD.

Figure 2. Crystal structure of the E1 DBD

(A) Crystal structure of the BPV1 E1 DBD highlighting the locations of the DNA-binding loop (red), DNA binding helix (helix α4, yellow) and dimerization helix (helix α3, green). (B) Crystal structure of two BPV1 E1 DBD dimers bound to DNA (PDB accession number 1KSX) (Enemark et al., 2002). The structure shows the mode of DNA-binding and dimer-formation. It also shows that the two DBD dimers are bound on separate faces of the double-helix and do not interact with each other.

Figure 3. Crystal structure of the E1 helicase domain

Crystal structure of the hexameric BPV1 E1 HD bound to ADP and single-stranded DNA (PDB accession number 2GXA) (Enemark and Joshua-Tor, 2006). The six E1 monomers are colored in blue and green in an alternating fashion. ADP is shown in red. The location of the DNA is not shown in this figure for clarity. The figure shows both cartoon and surface representations of the HD. The front view depicts how the six monomers assemble to create a central channel. The side view shows the collar formed by the six minimal oligomerization domains and the larger subdomain formed by the AAA+ ATP-binding site. The bottom inset highlights the fact that each nucleotide-binding site is formed at the junction of two E1 monomers.

Figure 4. Location of the p-hairpins and single-stranded DNA within the E1 helicase central channel

Crystal structure of the hexameric BPV1 E1 HD bound to ADP and single-stranded DNA (PDB accession number 2GXA) (Enemark and Joshua-Tor, 2006). The E1 hexamer is colored as in Fig. 3 but with the six DNA binding β-hairpins (amino acids 504 to 508) highlighted in different colors. The single stranded DNA is omitted to allow for easier visualization of the β-hairpins.

Figure 5. Model of the assembly of BPV1 E1 at the minimal origin of DNA replication

(A) Diagram of the BPV1 origin (ori) showing the locations of the two E2 binding sites (designated E2BS11 and E2BS12), of the six E1 binding sites (E1BS) and of the AT-rich region (AT-rich). Also shown is the structure of the minimal ori fragment that lacks E1BS11 and has been used for most in vitro studies on the assembly of E1 and E2 on DNA. The nucleotide sequence of the six E1BS is indicated underneath the minimal ori. Two pairs of sites, E1BS 1 and 3 and E1BS 2 and 4, can support the dimerization of E1. E1BS 5 and 6 are not paired and are shown as dashed lines. (B) Schematic representation of the initiation of BPV1 DNA replication on the minimal BPV1 ori. Replication is initiated by the recruitment of E1 (light purple), by E2 (light blue), to the ori resulting in the assembly of an E1-E2-ori ternary complex comprised of a dimer of E1 bound to one pair of E1BS (E1BS 1 and 4) and a dimer of E2 bound to E2BS11. This complex serves as a template for the recruitment of additional E1 molecules and assembly of the E1 double-trimer (DT) intermediate. The E1 DT may not accumulate in vivo but can be trapped in a stable form in vitro by performing the assembly reaction with a non-hydrolysable nucleotide. Upon ATP-hydrolysis, however, this E1 DT is rapidly converted into a double-hexamer in which each of the two hexamers encircles a separate DNA strand. This complex is capable of DNA unwinding and likely serves as a platform for the recruitment of the cellular DNA replication factors DNA polymerase α primase (Pol α-prim), topoisomerase I (Topo I) and replication protein A (RPA). Further details on the assembly of E1 at the origin are provided in the text.

Figure 6. Crystal structure of the E1 ATPase domain in complex with the E2 transactivation domain (TAD)

Structure of the complex formed between the E2 transactivation domain (TAD, cyan), and the C-terminal ATPase domain of E1 (light purple) of HPV18 E1 (PDB accession number 1TUE) (Abbate et al., 2004). The inset highlights the salt bridge formed between the highly conserved E45 of the TAD (orange) and R454 of E1 (purple). Also shown is the location of the loop 2 in E1, which is involved both in E2-interaction and E1 oligomerization, in a mutually exclusive manner.

Figure 7. Proposed model for the papillomavirus DNA replication fork

The E1 protein (in purple) is shown oriented with its N-terminal origin binding domain (DBD) facing towards the unwound dsDNA. For simplicity, only one of the two hexamers is shown. The E1 helicase encircles the ssDNA template for leading strand DNA synthesis, pumping the ssDNA template through the E1 complex from the E1-DBD side towards the E1 HD side. Interaction of topoisomerase I (Topo I) (in pink) with the DBD assists both in E1 origin binding/specificity and targets Topo I to the incoming dsDNA, where its action is necessary for progression of the replication fork. Interaction of replication protein A (RPA) (in blue) with the DBD is involved in loading RPA onto the lagging strand ssDNA template newly unwounded by the E1 helicase action. The interaction of the HD with polymerase α-primase (Pol α-prim) (in yellow) may stimulate primer synthesis. As each short RNA-DNA primer is synthesized by Pol α-prim, replication factor C (RFC) (in orange), in coordination with RPA, prevents Pol α-prim re-association and loads PCNA (in green) and DNA polymerase δ (in green) onto the recessed 3´ DNA end, assembling a processive DNA polymerase complex. The various interactions of E1 with Topo I, RPA and Pol α-prim, as well as the interactions between the cellular factors themselves, are coordinated physically and temporally in a highly-organized sequential manner necessary for replication fork assembly and function.

Figure 8. Nucleo-cytoplasmic shuttling motifs in the N-terminal regulatory region of E1

Sequence alignment of the E1 shuttling module from different PV types, as indicated. The position of the first amino acid in each sequence is indicated (number in parentheses). The location of the bi-partite nuclear localization signal (NLS) is indicated by a double-arrow and highlighted in yellow. The nuclear export signal (NES) and cyclin-binding motif (CBM) are shaded in light blue and purple, respectively. Putative and known Cdk-phosphorylation sites, [S/T]-P, are colored in pink. Note the absence of some of these nucleo-cytoplasmic shuttling motifs in several E1 proteins from high-risk HPV types and from BPV1.