Geminivirus C3 protein: replication enhancement and protein interactions

Abstract

Most dicot-infecting geminiviruses encode a replication enhancer protein (C3, AL3, or REn) that is required for optimal replication of their small, single-stranded DNA genomes. C3 interacts with C1, the essential viral replication protein that initiates rolling circle replication. C3 also homo-oligomerizes and interacts with at least two host-encoded proteins, proliferating cell nuclear antigen (PCNA) and the retinoblastoma-related protein (pRBR). It has been proposed that protein interactions contribute to C3 function. Using the C3 protein of Tomato yellow leaf curl virus, we examined the impact of mutations to amino acids that are conserved across the C3 protein family on replication enhancement and protein interactions. Surprisingly, many of the mutations did not affect replication enhancement activity of C3 in tobacco protoplasts. Other mutations either enhanced or were detrimental to C3 replication activity. Analysis of mutated proteins in yeast two-hybrid assays indicated that mutations that inactivate C3 replication enhancement activity also reduce or inactivate C3 oligomerization and interaction with C1 and PCNA. In contrast, mutated C3 proteins impaired for pRBR binding are fully functional in replication assays. Hydrophobic residues in the middle of the C3 protein were implicated in C3 interaction with itself, C1, and PCNA, while polar resides at both the N and C termini of the protein are important for C3-pRBR interaction. These experiments established the importance of C3-C3, C3-C1, and C3-PCNA interactions in geminivirus replication. While C3-pRBR interaction is not required for viral replication in cycling cells, it may play a role during infection of differentiated cells in intact plants.

FIG. 1.

TYLCV C3 and TGMV AL3 are functional homologs. (A) The entire genome of TYLCV and the replication-competent A component of TGMV are shown. TYLCV is a 2.8-kb DNA with six open reading frames, while TGMV A is a 2.5-kb DNA with five open reading frames (arrows). Both DNAs have a 5′ intergenic region (white box), which contains the rolling circle replication origin. (B) A linear diagram of TYLCV or TGMV A complementary sense open reading frames shows the region deleted in the ΔC3 and ΔAL3 replicons. (C) Tobacco protoplasts were electroporated with the indicated replicon and expression cassette, and double-stranded and single-stranded viral DNA accumulation were monitored and quantified on DNA gel blots at 72 h posttransfection by phosphorimage analysis. The replicons were wild-type TYLCV (TY, lane 1), ΔC3 (lanes 2 to 4), wild-type TGMV A (TG, lane 5), and ΔAL3 (lanes 6 to 8). The expression cassettes were an empty 35S cassette (lanes 1, 2, 5, and 6), a TYLCV C3 cassette (lanes 3 and 8), and a TGMV AL3 cassette (lanes 4 and 7).

FIG. 2.

A consensus sequence for the geminivirus C3 protein family. The amino acid sequences of 113 C3 proteins were compared and scored for identical or functionally homologous residues. Amino acid positions that displayed ≥90% conservation are shown as bold capital letters. The uppercase letters designate conservation at the 70 to 90% level, while the lowercase letters indicate 50 to 70% conservation. Positions displaying <50% conservation are marked by an “x.” The sequences of TYLCV C3 and TGMV AL3 are shown below the consensus, with identical or similar residues designated by white letters on a black background. Each of the 30 mutated proteins is shown and labeled below the amino acid sequences. Mutations to alanine are prefixed with an “a,” followed by a center amino acid number within the mutated group. Mutations of two or three amino acids to alanine are indicated as diamonds. Extended mutations with at least four alanine substitutions are indicated by the triangles. Mutations that reversed an amino acid charge are prefixed with a “c” and designated by a circle.

FIG. 3.

Replication enhancement activities of mutated C3 proteins. (A) Tobacco protoplasts were transfected with the TYLCV ΔC3 replicon and wild-type or mutated C3 expression cassettes. Double-stranded viral DNA accumulation was quantified on DNA gel blots by phosphorimage analysis. Relative replication activity was calculated as the ratio of viral DNA in the presence of the mutated expression cassette to that in the presence of the wild-type C3 expression cassettes and plotted against the central position of the mutated motif. Open symbols indicate proteins that displayed wild-type activity. Filled symbols above the dashed line represent proteins that complemented ΔC3 replication significantly better than wild-type C3, while filled symbols below the dashed line represent mutants that showed significantly reduced complementation. Each data point represents at least three independent experiments, with the bars showing 2 standard errors. The statistical significance of each sample relative to the wild type was determined using Student's t test and a cutoff of P < 0.05. The diamonds, triangles, and circles are defined in the Fig. 2 legend. (B) DNA gel blots showing replication complementation for severely impaired C3/AL3 mutants. Wild-type (wt) TYLCV and TGMV A accumulation are shown in lanes 1 and 9, respectively. Lanes 2 to 8 correspond to the ΔC3 replicon, and lanes 10 to 16 contain the ΔAL3 replicon. The expression cassettes in the first panel were an empty 35S cassette (lanes 1 and 2), a wild-type C3 cassette (lane 3), or the mutant C3 cassettes a29 (lane 4), a53 (lane 5), a69 (lane 6), a86 (lane 7), and a93 (lane 8). The expression cassettes in the second panel were an empty 35S cassette (lanes 9 and 10), a wild-type AL3 cassette (lane 11), or the mutant AL3 cassettes t29 (lane 12), t53 (lane 13), t69 (lane 14), t86 (lane 15), and t93 (lane 16). (C) Immunoblot showing TGMV AL3 protein expression in baculovirus-infected SF9 cells. Wild-type AL3 (lane 1), t29 (lane 2), t53 (lane 3), t69 (lane 4), t86 (lane 5), and t93 (lane 6) were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and detected using an anti-AL3 antibody.

FIG. 4.

TYLCV C3 protein interactions. (A) C3 protein interactions were measured in yeast two-hybrid assays. Wild-type or mutated C3 sequence was fused to the GAL4 activation domain (indicated by the first protein listed in each protein pair) or the GAL4 DNA binding domain (second protein listed). Transformants were selected for protein interactions on medium lacking adenine, histidine, leucine, and tryptophan (−AHLW). Medium lacking leucine and tryptophan (−LW) selected for the input plasmids only. Yeast cotransfected with a wild-type C3 cassette and a cassette corresponding to C3, C1, PCNA, or pRBR grew on both −AHLW and −LW media. Yeast cotransfected with certain mutant C3 cassettes and a C3, C1, PCNA, or pRBR cassette grew only on −LW medium. (B) The strength of the interactions shown for wild-type C3 in panel A was quantified in growth assays. Yeast growth on −AHLW medium was normalized to growth on −LW medium as a measure of strength. (C to F) Growth assays for yeast cotransfected with indicated C3 mutants and (C) wild-type C3, (D) C1, (E) PCNA, and (F) pRBR. Growth in the presence of mutated C3 proteins was normalized to growth in the presence of wild-type C3, which was set at 100. Proteins that were not significantly different than wild type, determined as in Fig. 3, are shown as open symbols. Significantly different proteins are marked with filled symbols. Each data point reflects at least four independent experiments, with the bars representing 2 standard errors. The diamonds and triangles are defined in the Fig. 2 legend.

FIG. 5.

Model of C3 protein interactions. TYLCV C3 consists of 134 amino acids containing three hydrophobic clusters shown as filled boxes above the rectangular diagram. pRBR interaction with C3 was decreased by mutations clustered in the first 13 and final 9 amino acids of C3 (shown as solid regions). C3 homo-oligomerization was adversely affected by mutations in the middle of the protein, shown shaded within the rectangular region. PCNA interaction with C3 (dashed line) was impacted by mutations near the beginning of C3 to amino acid 95. C1 interaction with C3 (dotted line) was sensitive to mutations between amino acids 28 and 128.