Characterization of Rad51 from apicomplexan parasite Toxoplasma gondii: an implication for inefficient gene targeting

Abstract

Repairing double strand breaks (DSBs) is absolutely essential for the survival of obligate intracellular parasite Toxoplasma gondii. Thus, DSB repair mechanisms could be excellent targets for chemotherapeutic interventions. Recent genetic and bioinformatics analyses confirm the presence of both homologous recombination (HR) as well as non homologous end joining (NHEJ) proteins in this lower eukaryote. In order to get mechanistic insights into the HR mediated DSB repair pathway in this parasite, we have characterized the key protein involved in homologous recombination, namely TgRad51, at the biochemical and genetic levels. We have purified recombinant TgRad51 protein to 99% homogeneity and have characterized it biochemically. The ATP hydrolysis activity of TgRad51 shows a higher K(M) and much lower k(cat) compared to bacterial RecA or Rad51 from other related protozoan parasites. Taking yeast as a surrogate model system we have shown that TgRad51 is less efficient in gene conversion mechanism. Further, we have found that TgRad51 mediated gene integration is more prone towards random genetic loci rather than targeted locus. We hypothesize that compromised ATPase activity of TgRad51 is responsible for inefficient gene targeting and poor gene conversion efficiency in this protozoan parasite. With increase in homologous flanking regions almost three fold increments in targeted gene integration is observed, which is similar to the trend found with ScRad51. Our findings not only help us in understanding the reason behind inefficient gene targeting in T. gondii but also could be exploited to facilitate high throughput knockout as well as epitope tagging of Toxoplasma genes.

Figure 1. Alignment of TgRad51 with Plasmodium falciparum (PfRad51), human (HsRad51), Saccharomyces cerevisiae (ScRad51), Schizosaccharomyces pombe (SpRad51), Trypanosoma bricei (TbRad51) and Leishmania major (LmRad51) using Clustal method (Meg align, DNA star).

Shaded areas represent identical amino acids. The two DNA binding Walker motifs are boxed.

Figure 2. Purification and ssDNA dependent ATPase activity of recombinant TgRAD51 protein.

(A) M corresponds to molecular weight standards, the sizes of marker proteins in kDa are indicated, Lane 1 and 2 are uninduced and induced cell free extracts respectively, Lane 3 corresponds to the proteins in soluble fraction, Lane 4 being the loading flow through and Lane 5 is the washing flow through. Lane 6 and 7 correspond to the eluted fractions with increasing imidazole concentrations i.e. 250 mM and 400 mM respectively. B) Western blot analysis shows the presence of Histidine tagged TgRAD51 in lane 1 from the cells bearing pET-TgRAD51 and lane 2 shows the absence of the protein in the cell bearing the empty vector pET. C) The standard curve showing linear relationship between inorganic phosphate added and increase in absorbance at 360 nm in response to the addition to MESG as substrate and the enzyme PNP (EnzChek phosphate assay kit). D) ssDNA dependent ATP hydrolysis of TgRAD51 using EnZChek phosphate assay kit. At 300 µM ATP concentration, with 2 µM TgRAD51 and 60 µM φxssDNA, the rate of the reaction as calculated from the slope of the curve is 0.027 µM P_i produced/min (•) where as in absence of ssDNA (▪) no ATP hydrolysis occurs. E) Michaelis Menten curve has been plotted with ATP concentrations in the range of 20 µM to 600 µM and kinetic parameters are derived from the curve.

Figure 3. Gene conversion efficiency of TgRAD51 is poor compared to that of ScRAD51.

(A) Schematic diagram of DSB repair choice experiment. URA3 and KANMX represent wild type alleles. A HO endonuclease site is incorporated in ura3::HOcs mutant allele. Once the HO induced DSB is repaired by gene conversion (GC), the mutant ura3::HOcs allele is converted into wild type URA3 allele. A single strand annealing (SSA) event leads to deletion of the intervening sequence (containing KANMX gene) and merging of ura3::HOcs and URA3 alleles to create the wild type URA3 allele. (B) Bar diagram showing the percentage of cells survived after induction of DSB. The relevant genotypes are marked on the X-axis. The white bars represent fraction of the cells survived by repairing the DSB using GC mechanism, where as the hatched bars denote the fraction of the survivors that employed SSA mechanism. Each bars represent mean value ± SD from 4 different experiments. (C) Western blots showing the abundance of ScRad51 and TgRad51 proteins. Different lanes are marked with the respective genotypes. Actin is the loading control.

Figure 4. Gene targeting efficiency of TgRAD51 increases with increase in stretch of homologous sequences.

(A) Schematic diagram showing molecular events leading to targeted integration of ADE2 gene at ADH4 locus versus random integrations. KANMX is a second selectable marker retained only in case of random integration via non-homologous recombination mechanism. (B) Bar diagram showing efficiency of gene targeting at the ADH4 locus in cells harboring ScRad51 or TgRad51 as the sole recombinase. The mean value from four independent experiments is plotted with standard deviations. (C) Schematic diagram showing knockout strategy for SBA1 gene. Varying lengths of flanking homologous sequences on either side are indicated. (D) Efficiency of gene knockout with increasing flanking homology. The lengths of the flanking homologous stretches are indicated on the X-axis. These experiments are done at least 3 times and the mean values with standard deviations are plotted.

Figure 5. Gene targeting efficiency of TgRad51 is independent of Ku80 function.

(A) Schematic diagram showing knockout strategy for CHL1 gene. Varying lengths of flanking homologous sequences on either side are indicated. (B) Efficiency of gene knockout with increasing flanking homology in KU80 proficient (closed- circle, triangle or square) and KU80 deficient (open- circle, triangle or square) cells. The lengths of the flanking homologous stretches are indicated on the X-axis. These experiments are done at least 3 times and the mean values with standard deviations are plotted.