A comparative approach to elucidate chloroplast genome replication

Abstract

Background: Electron microscopy analyses of replicating chloroplast molecules earlier predicted bidirectional Cairns replication as the prevalent mechanism, perhaps followed by rounds of a rolling circle mechanism. This standard model is being challenged by the recent proposition of homologous recombination-mediated replication in chloroplasts.

Results: We address this issue in our current study by analyzing nucleotide composition in genome regions between known replication origins, with an aim to reveal any adenine to guanine deamination gradients. These gradual linear gradients typically result from the accumulation of deaminations over the time spent single-stranded by one of the strands of the circular molecule during replication and can, therefore, be used to model the course of replication. Our linear regression analyses on the nucleotide compositions of the non-coding regions and the synonymous third codon position of coding regions, between pairs of replication origins, reveal the existence of significant adenine to guanine deamination gradients in portions overlapping the Small Single Copy (SSC) and the Large Single Copy (LSC) regions between inverted repeats. These gradients increase bi-directionally from the center of each region towards the respective ends, suggesting that both the strands were left single-stranded during replication.

Conclusion: Single-stranded regions of the genome and gradients in time that these regions are left single-stranded, as revealed by our nucleotide composition analyses, appear to converge with the original bi-directional dual displacement loop model and restore evidence for its existence as the primary mechanism. Other proposed faster modes such as homologous recombination and rolling circle initiation could exist in addition to this primary mechanism to facilitate homoplasmy among the intra-cellular chloroplast population.

Figure 1

Locations of replication origins mapped on the tobacco chloroplast genome. The figure depicts the relative positions of known replication origins (A1, A2, B and R) and their complementary copies (origin type appended with '-C'), on the circular tobacco chloroplast genome, mapped using SimVector 4.22 . The six regions between each pair of replication origins are annotated respectively as 1 (between A2 and A1 [A1-C]), 2 (between A1 [A1-C] and R [B-C]/B), 3 (between R [B-C]/B and B/R [B-C]), 4 (between B/R [B-C] and A1), 5 (between A1 and A2) and 6 (between the A2s on either strand).

Figure 2

Mapping of replication origin-like sequence homologues in first category of chloroplast genomes. The figures show matches of replication-origin sequences mapped on the tobacco chloroplast genome with those in other Viridiplantae chloroplast genome categories, as found using NCBI pair wise Blast tool (E = 0.0001). Chloroplast genomes have been partitioned into seven categories based on the number of homologues found to known replication origins in tobacco. The number of regions between each pair of replication origins, analyzed for the presence of deamination gradients is indicated within parentheses next to each category. Representative genomes from first category are depicted here, with the name of the genome indicated within the circular map. Several categories have more than one representative genome, owing to their overall similar number of matches but differences in the type of matches (see Methods). Genomes of plant species other than those indicated among the representative maps have been tabulated in Table 1, according to their respective fits to match categories.

Figure 3

Mapping of replication origin-like sequence homologues in second category of chloroplast genomes. Representative genomes from the second category are depicted here. Categorization of chloroplast genomes were performed as described in Figure 2.

Figure 4

Mapping of replication origin-like sequence homologues in third category of chloroplast genomes. Representative genomes from the third category are depicted here. Categorization of chloroplast genomes were performed as described in Figure 2.

Figure 5

Mapping of replication origin-like sequence homologues in fourth category of chloroplast genomes. Representative genomes from the fourth category are depicted here. Categorization of chloroplast genomes were performed as described in Figure 2.

Figure 6

Mapping of replication origin-like sequence homologues in fifth category of chloroplast genomes. Representative genomes from the fifth category are depicted here. Categorization of chloroplast genomes were performed as described in Figure 2.

Figure 7

Mapping of replication origin-like sequence homologues in sixth category of chloroplast genomes. Representative genomes from the sixth category are depicted here. Categorization of chloroplast genomes were performed as described in Figure 2.

Figure 8

Mapping of replication origin-like sequence homologues in seventh category of chloroplast genomes. Representative genomes from the seventh category are depicted here. Categorization of chloroplast genomes were performed as described in Figure 2.

Figure 9

Minimalist Interpretation of Deamination Gradients in Chloroplast Genomes. The minimalist model that can be inferred based on our analyses of various regions between replication origins in all chloroplast genomes suggests presence of A → G gradients in the smaller (Type 2) and larger (Type 4) regions between A1 and A1-C origin copies and A2 and A2-C origin copies on either inverted repeats, which incidentally overlap with the Small Single Copy (SSC) and Large Single Copy (LSC) regions, respectively. These gradients are bi-directional, increasing from the center of each region towards the origins themselves. The directions of these gradients are shown on the representative genome (Zea mays).

Figure 10

Generation of single-strandedness and thereby, deamination gradients during the course of chloroplast genome replication. The cartoon features steps in the course of replication for the region between replication origin pairs on each inverted repeat. The two bubbles represent the displaced parental strands at origins A2 and A1 of each inverted repeat. These bubbles expand towards each other respectively, on each inverted repeat to form a Cairns replication intermediate. Replication forks move in the respective directions to synthesize two complete daughter strands. The parental strands are represented by complete lines, while the daughter strands are represented as dotted lines. The arrows indicate the direction of new strand synthesis. The growing windows of single-strandedness and therefore, A → G deamination gradients are indicated at each step, in parallel (Figures 10A and 10B respectively). The representation in Figure 10C shows the deamination gradients as we observe from analyzing nucleotide compositions of all chloroplast genomes for the region between A1 and A1-C origin copies on each inverted repeat, if we assume it to be overlapping the SSC or the region between A2 and A2-C origin copies, if we assume that it overlaps the LSC. This signature is possible only if replication proceeds as per steps indicated in Figure 10A.

Figure 11

Mechanism of Replication in the Chloroplast Genome. This figure shows the steps (1–5) highlighted in Figure 10, on the complete chloroplast genome. The parental strands are represented by complete lines, while the daughter strands are represented as dotted lines. The arrows indicate the direction of new strand synthesis.

Figure 12

Hypothesis for inferring the mechanism of replication in chloroplast genomes. The schematic shows extension of insights developed in animal mitochondrial genome systems, where deamination gradients have been related to single-strandedness during replication, to the plant chloroplast system. The hypothesis developed here is to test for the presence of local deamination gradients in regions spanning between replication origins (1–6), from which the direction in which the DNA is left single-stranded during replication could be inferred, to arrive at a model for the mechanism of chloroplast replication. The left portion of the image in the top rectangle, demonstrating mitochondrial replication is modified from Figure 1 in [60], and the graph on the right side of this image depicting genome-wide C → T and A → G deamination gradients is modified from Figure 4 in [23], respectively, such as to reflect heavy strand notations.

Figure 13

Relationship between number of homologues to tobacco chloroplast replication origins and length of the genome. The lengths are averaged across genomes within each category. The data point corresponding to category 1 does not fit in this positive trend, and is therefore excluded from this plot.

Figure 14

Relationship between numbers of replication origins homologous to those in tobacco and the extent of homology to a Chlamydomonas D-Loop sequence in Viridiplantae chloroplast genomes. The extents of homology are measured by the size of the region (nucleotides), in Viridiplantae chloroplast genomes that is homologous to a Chlamydomonas D-Loop sequence [17].