The minimum replication origin of merkel cell polyomavirus has a unique large T-antigen loading architecture and requires small T-antigen expression for optimal replication

Abstract

Merkel cell polyomavirus (MCV) is a recently discovered human polyomavirus causing the majority of human Merkel cell carcinomas. We mapped a 71-bp minimal MCV replication core origin sufficient for initiating eukaryotic DNA replication in the presence of wild-type MCV large T protein (LT). The origin includes a poly(T)-rich tract and eight variably oriented, GAGGC-like pentanucleotide sequences (PS) that serve as LT recognition sites. Mutation analysis shows that only four of the eight PS are required for origin replication. A single point mutation in one origin PS from a naturally occurring, tumor-derived virus reduces LT assembly on the origin and eliminates viral DNA replication. Tumor-derived LT having mutations truncating either the origin-binding domain or the helicase domain also prevent LT-origin assembly. Optimal MCV replication requires coexpression of MCV small T protein (sT), together with LT. An intact DnaJ domain on the LT is required for replication but is dispensable on the sT. In contrast, PP2A targeting by sT is required for enhanced replication. The MCV origin provides a novel model for eukaryotic replication from a defined DNA element and illustrates the selective pressure within tumors to abrogate independent MCV replication.

FIG. 1.

Mapping of the MCV core origin. (A) Comparison of the origin sequences among polyomavirus family. The straight and dotted arrow heads indicate complete (GAGGC) or incomplete pentanucleotide sequences (GXGGC and GAGXC) serving for LT binding sites, respectively. The origin of SV40 subgroup (GenBank accession no. EF59667.1) consists of highly conserved region, including the AT-rich tract on the late gene side, site II GAGGC repeats, the early palindrome (EP), and early enhancer region on the early gene side. The MCV origin (GenBank accession no. EU375804) is closely related to the origin of Py (GenBank accession no. J02288). MCV origin constructs of different lengths (589 to 69 bp) were cloned into pCR2.1 vector (Invitrogen) to identify the core origin sequence. (B) Deletion analysis of MCV origin. The origin constructs were cotransfected into 293 cells with either TAg206.wt or empty vector as a negative control, and the replication from the MCV origin was analyzed by Southern blotting. The autoradiogram was quantified by using ImageQuant software (GE Healthcare). The DpnI-sensitive DNA band was used as an internal standard for the amount of input DNA.

FIG. 2.

Definition of MCV minimal core origin. Single-base-pair deletions from both sides of the origin sequence (late, nucleotide 5356; early, nucleotide 48) were performed to define a minimal core origin. The expression level of TAg was observed by Western blotting (data not shown), and an autoradiogram of replication was quantified as in Fig. 1B (middle panel).

FIG. 3.

Pentanucleotide requirements for MCV origin replication. (A) Mutation of each pentanucleotide in the 97-bp origin was generated by PCR-based direct mutagenesis (AT to GC) to define the essential pentanucleotides for MCV origin replication. (B) Replication efficiency was examined by Southern blotting.

FIG. 4.

Single mutational analysis of pentanucleotides. (A) Each sequence of pentanucleotides 1, 2, 3, 7, and 8 was mutated by AT-to-GC substitutions to disrupt the TAg binding site. To avoid reconstitution of the T-antigen binding site [5′-G(A/G)GGC-3′] by mutation, AG-to-T substitutions were introduced in S7-2, S8-2, S2-2, and S3-2. (B) Southern blot results demonstrate that single base mutations in pentanucleotides 1, 2, and 7 can abolish replication efficiently, indicating that these pentanucleotides sites are critical for MCV origin replication as a functional unit.

FIG. 5.

ChIP analysis. (A) The binding efficiency of both non-tumor- and tumor-derived TAg to MCV origin in vivo was examined by using a ChIP assay. The 97-bp origin-containing plasmid [Ori339(97) or Ori350(97)] and the TAg constructs from nontumor (TAg206.wt) or tumor (TAg350 and TAg339) tissues (50) were cotransfected into 293 cells, and the binding of LT to the origin was investigated by quantitative PCR. All experiments were repeated at least three times. Error bars indicate the standard deviations. (B) The relative expression levels of wild-type and mutant LT proteins are demonstrated by Western blotting.

FIG. 6.

Intrinsic activities of T antigens that influence MCV replication. (A) Alternative spliced products (T1 to T4) of the early region give rise to LT, sT, and 57kT. Arrows indicate the positions of siRNAs, siT, and sisT. siT targets all TAg transcripts, while sisT targets only sT transcripts. An asterisk indicates the epitope site for CM8E6, and a star indicates the epitope site of CM2B4 antibodies. These antibodies detect LT/57kT/sT or LT/57kT proteins, respectively. The predicted MCV LT gene sequence retains all major conserved features of other polyomavirus LTs, including DnaJ, Rb-targeting, origin-binding, and helicase/ATPase domains. (B) sT expression significantly contributes to the efficiency of origin replication. LT, 57kT, and sT expression plasmids were cotransfected in various combinations into 293 cell. For sT expression, a green fluorescent protein (GFP) fusion construct was used. Western blotting was performed to monitor the expression level of each transcript product using V5 antibody (Invitrogen) for LT and 57kT and anti-GFP (SantaCruz) antibody for sT. A representative experiment out of three repeats is shown. (C) siRNA knockdown of LT, 57kT, and/or sT expression confirms the contribution of sT to efficient origin replication (D). A nonsilencing siRNA (siCtrl) was used as a control. Three specific bands (120, 57, and 19 kDa) are detected by Western blotting with CM8E6 antibody in 293 cells transfected with genomic TAg expression constructs (TAg206.wt) (C, right panel). CM2B4 detected both ∼120- and ∼57-kDa bands corresponding to the full-length LT and 57kT protein products, respectively (49) (C, left panel). (E) To examine the interaction between TAg and Hsc70 and its role in origin replication, Myc-tagged Hsc70 construct (pCMV-myc-Hsc70) was transfected into 293 cells with either V5-tagged wild-type TAg (TAg206.wt) or Hsc70 mutant TAg (D44H, D44N). Lysates were immunoprecipitated with anti-myc or anti-V5 antibody and immunoblotted with anti-V5 or anti-myc antibody in a reciprocal way, respectively. The LacZ gene-expressing vector, pcDNA6/V5-His/LacZ (Invitrogen), was used as a negative control for binding and 2% of the lysate was used for input control. (F) Hsc70 binding to TAg promotes efficient viral replication. For the replication assay, TAg206.wt or mutants (Hsc70 binding mutant, D44N; Rb binding mutant, LXCXK; Rb/Hsc70 binding mutant, LXCXK/D44N) were cotransfected with MCV origin plasmid [Ori339(97)] into 293 cells. Replication efficiency was measured and quantified by using ImageQuant (GE Healthcare).

FIG. 7.

Model of the MCV OBDs on the MCV origin. A model of the MCV OBD was placed on each of the putative binding sites. OBDs positioned at the required PS1, PS2, PS4, and PS7 are shown in blue. The remainder are yellow. The GAGGC sequences are shown in cyan. The AT-rich tracts of DNA that flank site 1/2 are shown in light magenta.