Molecular Evolution and Functional Diversification of Replication Protein A1 in Plants

Abstract

Replication protein A (RPA) is a heterotrimeric, single-stranded DNA binding complex required for eukaryotic DNA replication, repair, and recombination. RPA is composed of three subunits, RPA1, RPA2, and RPA3. In contrast to single RPA subunit genes generally found in animals and yeast, plants encode multiple paralogs of RPA subunits, suggesting subfunctionalization. Genetic analysis demonstrates that five Arabidopsis thaliana RPA1 paralogs (RPA1A to RPA1E) have unique and overlapping functions in DNA replication, repair, and meiosis. We hypothesize here that RPA1 subfunctionalities will be reflected in major structural and sequence differences among the paralogs. To address this, we analyzed amino acid and nucleotide sequences of RPA1 paralogs from 25 complete genomes representing a wide spectrum of plants and unicellular green algae. We find here that the plant RPA1 gene family is divided into three general groups termed RPA1A, RPA1B, and RPA1C, which likely arose from two progenitor groups in unicellular green algae. In the family Brassicaceae the RPA1B and RPA1C groups have further expanded to include two unique sub-functional paralogs RPA1D and RPA1E, respectively. In addition, RPA1 groups have unique domains, motifs, cis-elements, gene expression profiles, and pattern of conservation that are consistent with proposed functions in monocot and dicot species, including a novel C-terminal zinc-finger domain found only in plant RPA1C-like sequences. These results allow for improved prediction of RPA1 subunit functions in newly sequenced plant genomes, and potentially provide a unique molecular tool to improve classification of Brassicaceae species.

Keywords: DNA repair; RPA; meiosis; replication.

Figure 1

Meiotic defective A. thaliana RPA1 mutant lines. (A) Siliques harvested from ~7 weeks old wild-type (WT) and RPA1 mutant plants. (B) Number of seeds per silique. The mutant plants have the following percentage of fertility reduction: rpa1a: ~70%; rpa1a (−/−) rpa1c (+/−): ~92%; rpa1a rpa1c: 100%. rpa1a (+/−) rpa1c (−/−) has similar fertility level as WT. Data are mean ± SE (n > 10). To analyze statistical difference F-test (ANOVA) and LSD were carried out at P ≤ 0.05. Bars with different letters indicate significant differences.

Figure 2

Evolutionary relationships of RPA1 proteins. (A) RPA1A group, (B) RPA1B group, (C) RPA1C group. The evolutionary history was inferred using the Maximum-Likelihood method performed with MEGA5.2 software package. Amino acid sequences were aligned using ClustalW and used to produce phylogenetic trees using the Jones-Taylor-Thornton (JTT) amino acid substitution model. Numbers next to the branches are bootstrap values (1000 replicates). Branches that show less than 70% bootstrap support were collapsed. Some RPA1 sequences denoted by asterisk (^*) [sorghum RPA1C-3, P. patens RPA1B, and V. cateri RPA1B] contain DBD-F deletion. Yeast RPA1 was used as an outgroup to root the tree. Except for bootstrapping and choice of model, all other parameters were left at default settings.

Figure 3

Evolutionary relationships of unicellular green algae RPA1 proteins. The evolutionary history was inferred using the Maximum Likelihood method performed with MEGA5.2 software package. Amino acid sequences were aligned using ClustalW and used to produce phylogenetic trees using the Jones-Taylor-Thornton (JTT) amino acid substitution model. Some RPA1 sequences denoted by asterisk (^*) [C. reinhardtii RPA1A, M. pussila RPA1B, and V. cateri RPA1B] contain DBD-F deletion. Numbers next to the branches are bootstrap values (1000 replicates). Scale represents number of amino acid substitutions per site. Except for bootstrapping and choice of model, all other parameters were left at default settings.

Figure 4

Schematic diagram of the structure and functional domains of RPA1 proteins. At, Arabidopsis thaliana; Sc, Saccharomyces cerevisiae; Hs, Homo sapiens. Blue inset boxes represent Generic Binding Surface I. Pink inset boxes represent Binding Surface I (Basic Cleft). Yellow inset boxes represent C5 (in At RPA1B, D), C6 (in At RPA1A, C, E), and C4 (in Sc RPA1and Hs RPA1)—type zinc-finger motifs. Green inset boxes represent CCHC-type zinc-finger motif. Bars above At RPA1C and At RPA1E indicate C-terminal extension region.

Figure 5

Proposed model of RPA1 evolution in plants and algae. Gray boxes from right to left are DBD-F and DBD-C, respectively, dark boxes from right to left are DBD-A and DBD- B, respectively, red boxes are Binding Surface I (BS-I), blue boxes are Generic Binding Surface I (GBS-I), yellow boxes are C4/C5/C6—type zinc-finger motifs (ZFM), green boxes are CCHC-type ZFM. Line bars indicate C-terminal extension (tail) region, broken lines and strips indicate poorly conserved domains.

Figure 6

Level and pattern of RPA1 expression at different developmental stages and root zones. (A) A. thaliana; (1) Germinated seed, (2) Seedling, (3) Young rosette leaf, (4) Developing rosette leaf, (5) Bolting, (6) Young flower, (7) Developed flower, (8) Flowers and siliques, (9) Mature siliques, (10) Senescence. (B) AtRPA1 expression in different root zones of 7-day-old Arabidopsis seedlings. (C) Soybean; (1) Germination, (2) Main shoot growth, (3) Flowering, (4) Fruit formation, (5) Bean development. (D) Rice; (1) Germination, (2) Seedling, (3) Tillering stage, (4) Stem elongation stage, (5) Booting stage, (6) Heading stage, (7) Flowering stage, (8) Milk stage, (9) Dough stage. (E) Psychomitrela patens; (1) Germination [protenema development], (2) Gametophore growth, (3) Gametangia development. Data were collected from genevestigator (A,C–E; Hruz et al., 2008) and Arabidopsis eFP Browser (B; Winter et al., 2007). Error bars indicate standard error.

Figure 7

Number of introns in plant RPA1 genes. Twenty plants (Supplementary Table S1) are included in the analysis. Data are mean ± SE. To analyze statistical difference F-test (ANOVA) and LSD were carried out at P ≤ 0.05. Bars with different letters indicate significant differences.

Figure 8

Frequency of optimal codon (F_OP) values for RPA1 genes of Arabidopsis and rice. Data are mean ± SE. To analyze statistical difference F-test (ANOVA) and LSD were carried out at P ≤ 0.05. Bars with different letters indicate significant differences.

Figure 9

Natural selection (ω) values for RPA1 genes. (A) ω-values for Arabidopsis and A. lyrata RPA1 genes. C. rubella RPA1 genes were used as a reference for ortholog pairwise sequence distance measurement (dN and dS). (B) ω-values for RPA1 genes of eight plants (Tomato, Cucumber, Strawberry, Castor oil plant, Grape, Cacao, Peach, and California Poplar). Rice RPA1 genes were used as a reference for ortholog pairwise sequence distance measurement (dN and dS). dN and dS analyses were conducted in MEGA5 using the Nei-Gojobori model. Data are mean ± SE. To analyze statistical difference F-test (ANOVA) and LSD were carried out at P ≤ 0.05. Bars with different letters indicate significant differences.