Mutation tendency of mutator Plasmodium berghei with proofreading-deficient DNA polymerase δ

Abstract

In this study, we investigated the mutation tendency of a mutator rodent malaria parasite, Plasmodium berghei, with proofreading-deficient DNA polymerase δ. Wild-type and mutator parasites were maintained in mice for over 24 weeks, and the genome-wide accumulated mutations were determined by high-throughput sequencing. The mutator P. berghei had a significant preference for C/G to A/T substitutions; thus, its genome had a trend towards a higher AT content. The mutation rate was influenced by the sequence context, and mutations were markedly elevated at TCT. Some genes mutated repeatedly in replicate passage lines. In particular, knockout mutations of the AP2-G gene were frequent, which conferred strong growth advantages on parasites during the blood stage but at the cost of losing the ability to form gametocytes. This is the first report to demonstrate a biased mutation tendency in malaria parasites, and its results help to promote our basic understanding of Plasmodium genetics.

Figure 1. Schematic diagram showing the parallel passage lines of parasites.

Three lines of PbWT parasites and four lines of PbMut parasites were maintained by serial passage in mice. Excluding the Md line, 10⁶ parasitised red blood cells (pRBCs) were injected intravenously (iv) into mice every 3 or 4 days in order to maintain the parasites. The Md line was maintained by weekly passage via intraperitoneal (ip) injection of 100 or 1,000 pRBCs into mice. The samples in double-lined squares were subjected to genome sequencing analysis.

Figure 2. Distribution of mutations detected in PbWT and PbMut parasites.

The outermost line shows the 14 chromosomes of P. berghei. The distributions of mutations detected in each sample are represented as follows: base substitutions in intergenic regions (light blue circles), base substitutions in introns (gray squares), synonymous mutations (light blue triangles), missense mutations (orange triangles), nonsense mutations (magenta triangles), indels in intergenic regions (light blue rhombuses), an indel in introns (a light purple rhombus), frameshift mutations (magenta rhombuses), and a codon deletion (a dark purple rhombus). Ma25P and Ma28, Mb29 and Mb25P, and Mc30 and Mc26P are grouped and displayed, where mutations from the population sample are represented by hollow glyphs and those from the clone sample are represented by solid glyphs. Bar charts outside the tracks show the distribution of mutations to indicate the frequencies of mutations detected from population samples. The AP2-G gene (PBANKA_143750) region is highlighted in light blue.

Figure 3. Relative mutation rates of six possible base substitutions in PbMut.

This chart was generated with the data obtained from Ma28, Mb29, Mc30, and Md45A. There was a statistically significant difference between base substitution types with different letters (Holm-adjusted P < 0.05; Student’s t-test). Error bars represent standard errors.

Figure 4. Sequence context-dependent mutation rate in PbMut.

The sequence context-dependent mutation rates for intergenic (light blue circles) and genic (orange squares) regions were calculated with the data obtained from Ma28, Mb29, Mc30, and Md45A. The orange and light blue dashed lines indicate the overall mutation rates in intergenic (11.2 ± 0.7 × 10⁻⁹ base⁻¹ day⁻¹) and genic (15.0 ± 2.3 × 10⁻⁹ base⁻¹ day⁻¹) regions, respectively. A statistically significant elevated mutation rate was observed at TCT compared with the average mutation rate in each region (*p < 0.05; Welch’s t-test). Error bars represent the bootstrap-estimated 95% confidence intervals.

Figure 5. Growth of ancestral and evolved parasites.

(A) Parasite growth was monitored for two mutator clones (Mut0 and Md45A) and two wild-type clones (WT0 and Wc29). Parasitemia was derived based on the mean of 10 mice (±SE). (B) Maximal growth rates (% day⁻¹) were estimated between days 3 and 6. Individual circle dots represent the maximal growth rate of parasites observed in individual mice, and yellow rhombus dots represent the mean. Boxplots show the distribution of data, where the middle, top, and bottom lines on a box represent the median and 75th and 25th percentiles of the data, respectively. There was a statistically significant difference between clones with different letters (Holm-adjusted P < 0.05; Wilcoxon rank-sum test).

Figure 6. Detrended correspondence analysis based on RNA-Seq data for P. berghei genes in four developmental stages.

(A) Circle dots are P. berghei genes, and cross marks denote four developmental stages (ring, trophozoite, schizont, and gametocyte). Mutated genes in PbMut are highlighted as follows: synonymous-mutated genes (green), missense-mutated genes (yellow), and nonsense-mutated genes (red). The names of nonsense-mutated genes are also shown in red letters. (B,C) Kernel density plots showing the distributions of dots along the DCA1 and DCA2 axes. Green, yellow, red, and black lines denote the densities of synonymous-mutated, missense-mutated, nonsense-mutated, and all genes, respectively.