Chapitre IV : Increased homeologous recombination and microsatellite instability in a Arabidopsis thaliana msh2 mutant

Table des matières

Increased homeologous recombination and microsatellite instability in a Arabidopsis thaliana msh2 mutant.

Martine Jean, Liangliang Li, Aïda Azaiez, and François Belzile*

Département de phytologie, Université Laval, Québec, Qc, Canada, G1K 7P4

*Corresponding author ; tel : +1 418 656-2131 ext. 5763 ; fax : +1 418 656-7176

e-mail : fbelzile@rsvs.ulaval.ca

Key words :

Arabidopsis thaliana , MSH2 , Homologous recombination, Homeologous recombination, Microsatellite instability.

Les systèmes de réparation de l’ADN jouent un rôle essentiel dans le maintien de la stabilité du génome. Le système de correction des mésappariements (système MMR) permet non seulement d’éviter les mutations mais aussi de contrôler la recombinaison entre des séquences où la divergence entraîne des mésappariements de bases. Dans le but d’élucider le rôle du système MMR dans la recombinaison et dans la stabilité du génome chez Arabidopsis thaliana , nous avons étudié la fonction du gène AtMSH2 , une composante centrale du système MMR. À cette fin, le gène GUS a été utilisé dans cette étude comme système rapporteur de la recombinaison et de l’instabilité des microsatellites afin de caractériser une lignée portant une insertion à T-DNA dans le gène AtMSH2 . L’impact de la mutation msh2 sur l’instabilité des microsatellites endogènes a aussi été déterminé dans des lignées à descendance mono-graine. Une augmentation par 9,7 fois de la recombinaison homéologue et une diminution de 30 % de la recombinaison homologue ont été détectées chez le mutant msh2 . L’instabilité du microsatellite artificiel G7 a augmenté de 3,8 fois dans les tissus somatiques du mutant msh2 . L’impact de l’inactivation d’ AtMSH2 a été plus important dans la transmission des mutations germinales, où une augmentation de 52 fois de l’instabilité du microsatellite synthétique G16 a été décelée. La création et la fixation de nouveaux allèles dans divers microsatellites dinucléotidiques endogènes ont été observées suite à la propagation en série des lignées mutantes msh2 . Nos résultats confirment le rôle majeur du système MMR dans la stabilité du génome durant la reproduction sexuée chez Arabidopsis.

DNA repair systems play an essential role for the maintenance of genome stability. In addition to its function in mutation avoidance, the DNA mismatch repair (MMR) system is involved in regulating recombination between sequences where divergence leads to mismatched bases. With the aim of deciphering the role of the MMR system in recombination and genome stability in Arabidopsis thaliana , we are investigating the function of AtMSH2 , a key component of the MMR system. To achieve this goal, GUS -based reporter assays for recombination and microsatellite instability were used to characterize a line carrying a T-DNA insertion into the AtMSH2 gene. The impact of this msh2 mutation on the stability of endogenous microsatellites in lines propagated by single-seed descent was also determined. A 9.7-fold increase in homeologous recombination was detected in the msh2 mutant, as well as a 30% decrease in homologous recombination. The instability of a G7 synthetic microsatellite was elevated 3.8-fold in somatic tissues from the msh2 mutant but the impact of the msh2 mutation was much more important on the germinal transmission of mutations at a G16 synthetic microsatellite, where a 52-fold increase was detected. Similarly, the frequent creation and fixation of new alleles at various endogenous dinucleotide microsatellites was noted following serial propagation of msh2 lines. Our results thus support a major role for the MMR system in ensuring genome stability during sexual reproduction in Arabidopsis .

Since DNA is submitted to various attacks from the environment, DNA repair systems play an essential role to preserve genome integrity in all living organisms. Efficient DNA repair systems are particularly important in plants since their reproductive organs are formed from somatic meristems late during development, and unrepaired somatic mutations can potentially be passed on to subsequent generations.

One of the best-characterized DNA repair system is the DNA mismatch repair (MMR) system (Harfe and Jinks-Robertson 2000a, Schofield and Hsieh 2003). The MMR system is widely conserved among both prokaryotic and eukaryotic organisms, and plays two important roles in the maintenance of genome stability. First, it interacts with the replication complex to assist in correcting replication errors such as misincorporated bases that had escaped the proofreading function of the polymerase as well as small indels created by polymerase slippage. In addition to this anti-mutator role, the MMR system also regulates recombination between sequences where divergence leads to mismatched bases in recombination intermediates. In fact, a single mismatch has been shown to be enough to induce a detectable MMR-dependent decrease in recombination in yeast (Datta et al ., 1997). The MMR system therefore plays a prominent anti-recombination role in preventing ectopic (non-allelic) recombination as well as recombination between homeologous chromosomes in hybrids from interspecific crosses (Rayssiguier et al ., 1989, Hunter et al., 1996). Finally, some components of the MMR system also play a role in promoting homologous recombination (Saparbaev et al ., 1996, Sugawara et al ., 1997).

One of the key components of the E. coli MMR system is the MutS protein which, as an homodimer, recognizes and binds to base:base mismatches and small indels. To initiate DNA repair, the MutS homodimer then recruits a MutL homodimer. The latter is thought to act as a molecular matchmaker to activate the latent endonuclease activity of the MutH protein that makes a nick in the freshly replicated, still unmethylated, DNA strand. This strand bearing the incorrect nucleotide is then degraded and re-synthesized correctly.

In eukaryotes, there has been a multiplication of MutS and MutL homologues, which now work as heterodimers with specialized functions. For example in Saccharomyces cerevisiae , the best-characterized eukaryotic MMR system, there are six MutS homologues (MSH1-6). Three of the MutS homologues have no role in MMR; MSH1 in instance, is involved in the stability of the mitochondrial genome while the MSH4/MSH5 complex functions exclusively in meiotic recombination. MSH2, the main eukaryotic MutS homologue, works in complex with MSH6 to recognize base:base mutations and small indels, while the MSH2/MSH3 complex is specialized in the detection of indels from 1 to 16 bp. In plants, an additional MutS homologue, MSH7, has been identified (Adé et al ., 1999; Culligan and Hays 2000), and the MSH2/MSH7 complex was shown to have a specificity different from the two other MSH2-containing complexes (Culligan and Hays 2000, Wu et al ., 2003).

Because of the central role of MSH2 in MMR, a mutation in the MSH2 gene can be very detrimental to genome stability. Indeed, msh2 mutants usually display a dramatic increase in base:base mutations as well as in microsatellite instability, and homeologous recombination is also elevated (Harfe and Jinks-Robertson 2000a). In humans, mutations in the MSH2 gene have been associated with a predisposition to certain sporadic and familial colorectal cancers that display as "signature" phenotype an increase in microsatellite instability (Liu et al . 1996).

Recently, a MSH2 homologue has been identified and characterized in Arabidopsis (Culligan and Hays 1997; Adé et al ., 1999; Adé et al ., 2001; Culligan and Hays 2000). Mutations in this gene have recently been associated with microsatellite instability (Leonard et al ., 2003). The effects on homologous and homeologous recombination, as well as on the long-term genome stability were however not determined. With the aim of deciphering the role of the MMR system in recombination and genetic stability in Arabidopsis thaliana , GUS -based reporter assays for recombination and microsatellite instability were used to characterize a line carrying a T-DNA insertion into the AtMSH2 gene. The impact of a msh2 mutation on the stability of endogenous microsatellites in lines propagated by single-seed descent was also determined.

For PCR analyses, DNA was extracted from leaves by a rapid microextraction protocol (Edwards et al ., 1991). PCR reactions were carried out on a Perkin Elmer DNA thermal cycler in a 20 µl volume containing 1 µl of the DNA-containing solution, 0.5 µM of each primer, 0.2 mM of each dNTP, 2% DMSO and 0.5 U of Taq DNA polymerase (NEB) in its associated buffer. Cycling conditions began by a hot start of 5 min at 94°C followed by 35 cycles of 30 s at 94°C, 30 s at 60°C, 1 min at 72°C, and concluded by 10 min at 72°C. Sequences of the various primers used in the present study were as follows: a: 5'-CAACTATAAGGCTTCCCTTCATC-3'; b: 5'-ATGCTCACATATAGCCCAAGCTAAACC-3'; c: 5'-CCGCTCCTTTCGCTTTCTTCCCTTCC-3'; d: 5'-GGCGACTTTTGAACGCGCAATAATGGTTTC-3'; e: 5'-TTTGCCCTCATTTAGTTTCATCCCAAGCA-3'; f: 5'-ACGAGAGCTGTTAGATTCTTTG-3'; g: 5'-CTGATGAGTGATGCGAATTCTCCG-3'. It can be noted that the T-DNA-specific primer that was used in the present study was located internally to those used by the Salk Institute to characterize the T-DNA insertion lines.

For reverse transcription mediated PCR (RT-PCR), mature leaves from plants grown in soil were frozen in liquid N2 and ground to a fine powder from which total RNA was extracted as described in Verwoerd et al . (1989). About 1.5 µg of total RNA was reverse-transcribed with the First-Strand cDNA Synthesis kit (Amersham Biosciences) using poly-d(T) oligonucleotides as primers. The resulting cDNAs were amplified by PCR as described above, using primers for the AtMSH2 (primer f and g, see above) and AtMLH1 genes (MLH3: 5'-CTTATTGCTGGAGTTGACAGCTGC-3'; MLH4: 5'-AGTTCTTCTACAATCAGCGGTTTC-3').

Two recombination reporter lines carrying recombination substrates with either 0 or 1.9% of sequence divergence (lines R3L66 and R3M11L38, Li et al ., 2004) were used to allow an easy detection of recombination events as blue sectors (Figure 4.1). These lines harbor an intron-containing partially inverted GUS gene. Recombination between two 589 bp overlapping fragments of the intron create a functional GUS gene. Sequence divergence between the two overlapping fragments is 0% when assessing homologous recombination, and 1.9% (11 mismatches) for homeologous recombination.

Two microsatellite instability (MSI) reporter lines carrying either a G7 or G16 synthetic microsatellite (lines 131B and 121A, Azaiez et al ., in preparation) were used to allow easy detection of mutations as blue sectors. These lines harbor a microsatellite-containing out-of-frame GUS gene that was constructed by introducing a synthetic mononucleotide microsatellite (either G7 or G16) in the Msc I site of the GUS gene. Mutations that change the number of repeats in the microsatellite to one that restores the correct reading-frame (for example, a deletion of 1 bp) create a functional GUS gene.

The msh2 mutant was crossed to these reporter lines and F2 plants homozygous for both the reporter construct and either the wild-type or msh2 mutant allele were identified by PCR with primers specific for either the AtMSH2 gene (primers a and b or a and c for the disrupted or wild-type allele, respectively) or the reporter construct (GUS-F: 5'-GTTTCGATGCGGTCACTCATTAC-3'; GUS-R: 5'-CACACTGATACTCTTCACTCCAC-3'; 5'r3l66: 5'-GTCTGGTGCTTGATTCCAATCTTCAGG-3'; 3'r3l66: 5'-GGAAAAAGGATTACGGAGAAAAGGCCAC-3'; 5'r3m11l38: 5'-GGTCGTGGGTTCCATTAGGCTTTCTCTG-3'; 3'r3m11l38: 5'-GCTCAGCGACGGCGATGACTATCCACC-3'), or by the segregation ratio obtained on selective medium. F3 or F4 progenies homozygous for both the reporter construct and either the wild-type or msh2 mutant allele were compared to evaluate the effect of the msh2 mutation.

Somatic regions where a recombination event has created a functional intron in the GUS gene (for the recombination reporters) or where a mutation has created a in-frame GUS gene (for the MSI reporters) were detected as blue sectors after staining with X-Gluc and bleaching with ethanol as described in Li et al . (2004). Germinally-transmitted mutations in the G16 synthetic microsatellite were detected as plants (termed "germinal revertant") that were completely blue after staining.

Figure 4.1. T-DNA carrying a GUS -based reporter system for recombination.

A reporter construct based on the GUS gene was constructed by Li et al. (2004) to detect recombination events. Recombination between two 589-bp overlapping fragments of an intron (striped rectangle) restore a functional GUS gene (grey rectangle) containing a complete copy of the intron (bordered by GT-AG consensus splice sites). Two versions of the reporter were used. When assessing homologous recombination, the two copies of the intron were identical (0% sequence divergence), whereas a modified construct with 1.9% divergence between the copies of the intron (11 mismatches) was used to assess homeologous recombination. P: CaMV35S promoter; T: Nos terminator.

The number of sectors/plants was determined on mature plants (at the 6-8 or 10-14 leaf stage for the recombination and MSI assay, respectively) grown in soil in growth chambers. The number of plants scored per line ranged from 119 to 137 for the recombination assay. For the MSI assay with the G7 synthetic microsatellite, 29 and 62 plants were scored for the msh2 mutant and wild-type, respectively. To quantify the germinal transmission of MSI at the G16 synthetic microsatellite, 1356 and 2240 plants were tested for the msh2 mutant and wild-type, respectively. To estimate the level of somatic MSI of the G16 synthetic microsatellite, the number of sectors/plant was scored on two F2 homozygous wild-type and heterozygous plants. For these lines, the number of sectors/leaf was obtained by dividing the total number of sectors by the total number of leaves on the plants. For the homozygous msh2 mutant, the number of sectors/leaf was obtained by scoring two leaves from two plants (four leaves total). A Student's t test was performed to determine if the differences observed between the msh2 mutant and the wild-type were significant.

For the complementation experiment, an AtMSH2 cDNA clone (Genbank accession AF026549) was digested with Not I, blunted and subcloned into a blunted Nco I - BstE II restriction-digested pCAMBIA1302 binary vector (Genbank accession AF234298, obtained from CAMBIA, Canberra, Australia). F3 plants homozygous for both the reporter construct and either the wild-type or msh2 mutant allele were transformed by the dipping method (Clough and Bent 1998) using Agrobacterium tumefaciens, strain GV3101. The intact pCAMBIA vector was used as control. Seeds from the Agrobacterium -treated plants were plated on selective medium and hygromycin-resistant T1 plants were transferred to soil and their genotype characterized. Plants that do not contain tandem repeats (as indicated by the absence of a PCR product from a LB/RB junction) and that carry a single locus insertion (as determined by a 3:1 segregation ratio of T2 seeds on hygromycin) were identified. The number of sectors/"plant" was determined by scoring six mature leaves from each of two flowering T1 plants hemizygous for either the AtMSH2 cDNA construct or the pCAMBIA vector control.

To investigate the function of the AtMSH2 gene in Arabidopsis , we first searched for an insertional mutant by performing a PCR-based screen with insert-specific and AtMSH2 -specific primers on four populations of Arabidopsis insertional mutant lines (Feldmann and Thomas Jack collections of T-DNA lines, AMAZE collection of En-1 lines, and Dubois collection of Ac lines). No insertional mutant for the AtMSH2 gene was found among the 15,990 lines that compose these collections (data not shown).

We then screened the Salk Institute T-DNA Express Database by performing a BLAST search on their web site using the AtMSH2 genomic sequence as query. Three lines (SALK_002707, SALK_002708 and SALK_020225) were identified as putatively containing a T-DNA insertion in the open reading frame (ORF) of the AtMSH2 gene. Three PCR amplifications were performed to characterize the T-DNA insertion present in these lines: one with primers specific for either the wild-type or disrupted AtMSH2 alleles, one with primers specific for the NPTII gene and one with primers specific for the left border (LB) and right border (RB) of the T-DNA to detect possible tandem insertions by their RB/LB junctions (Figure 4.2A). These analyses indicated that lines SALK_002707 and SALK_020225 are not AtMSH2 T-DNA insertional mutants (Figure 4.2B). They also revealed that line SALK_002708 is homozygous for a T-DNA insertion in the MSH2 gene and suggested that this insertion is composed of a tandem of at least two T-DNAs (Figure 4.2B).

Sequencing of the AtMSH2 /T-DNA junctions confirmed that the LB of the T-DNA insert is located near the end of exon 7, but revealed that the AtMSH2 genomic DNA adjacent to the RB of the T-DNA insertion corresponds to a 153-bp insertion of filler DNA directly fused to exon 13, only 37 bp upstream from the stop codon (Figure 4.2A). Line SALK_002708 thus appears to carry a 1510-bp deletion of the AtMSH2 gene, in addition to a T-DNA insertion predicted to alter the AtMSH2 ORF by creating a stop codon immediately after amino acid 645. Any transcript produced by this line would therefore result, at best, in the synthesis of a truncated MSH2 protein that would be lacking the last 292 C-terminal amino-acids. To determine whether residual transcripts are present in line SALK_002708, RT-PCR analyses were performed with AtMSH2 primers located upstream from the T-DNA insertion. No band corresponding to AtMSH2 transcripts was detected in line SALK_002708 while a band of the expected size was observed in wild-type Columbia (Figure 4.2C). Successful amplification of transcripts from the AtMLH1 gene was achieved under the same experimental conditions, thus demonstrating that this assay is sensitive enough to detect transcripts of low abundance. Line SALK_002708 therefore appears to be a null msh2 mutant and was used to study the effects of a loss of function of the AtMSH2 gene on the stability of Arabidopsis genome.

Figure 4.2 (next page): Molecular characterization of the Arabidopsis msh2 T-DNA insertional mutant line SALK_002708.

(A) Schematic representation of the AtMSH2 genomic region illustrating the T-DNA insertion in mutant line SALK_002708.

Exons are indicated as open boxes and the ATG and TGA codons are shown. The AtMSH2 genomic DNA missing in the mutant allele (about 1.5 kb, from exon 7 up to exon 13) is indicated by a grey box. Dashed lines represent the 153-bp filler DNA found adjacent to the right border of the T-DNA insertion. Small arrows indicate the approximate location of the various primers used in the present study. (a: 5'-CAACTATAAGGCTTCCCTTCATC-3'; b: 5'-ATGCTCACATATAGCCCAAGCTAAACC-3'; c: 5'-CCGCTCCTTTCGCTTTCTTCCCTTCC-3'; d: 5'-GGCGACTTTTGAACGCGCAATAATGGTTTC-3'; e: 5'-TTTGCCCTCATTTAGTTTCATCCCAAGCA-3'; f: 5'-ACGAGAGCTGTTAGATTCTTTG-3'; g: 5'-CTGATGAGTGATGCGAATTCTCCG-3')

(B) Characterization of the T-DNA insertion from lines SALK_002707, SALK_002708 and SALK_0020225.

PCR amplification was performed with primers specific for either the disrupted (primers a and b in Figure 4.2A) or the wild-type (primers a and c in Figure 4.2A) AtMSH2 alleles, as well as for the NPTII gene (NPT.5P: 5'-ACTGAAGCGGGAAGGGACTGGCTGCTATTG-3'; NPT.3P: 5'-GATACCGTAAAGCACGAGGAAGCGGTCAG-3'). 1: SALK_002707, 2: SALK_002708 and 3: SALK_0020225.

(C) RT-PCR analysis of AtMSH2 expression in mature leaves.

Amplification with primers specific for the AtMSH2 (primers f and g in Figure 1A) and AtMLH1 genes (MLH3: 5'-CTTATTGCTGGAGTTGACAGCTGC-3'; MLH4: 5'-AGTTCTTCTACAATCAGCGGTTTC-3') was performed on reversed-transcribed RNA extracted from wild-type Columbia (Col) and line SALK_002708 (2708). The arrows indicate the position of the expected bands.

The msh2 mutant was crossed with two reporter lines that carry recombination substrates with either 0 or 1.9% of sequence divergence (Li et al ., 2004) and allow for the detection of recombination events as blue sectors (Figure 4.1). F3 plants homozygous for both the reporter construct and either the wild-type or mutant allele were compared to evaluate the effect of a msh2 mutation on homologous and homeologous recombination (Table 4.1).

The number of sectors was scored on plants at the 6- to 8-leaf stage. The number of plants scored per genotype ranged from 119 to 137. The corrected ratio (last column on the right) was calculated by dividing the "relative to wild-type" ratio by 0.7, thereby ‘factoring out’ the decrease in recombination attributed to the loss of AtMSH2 activity.

Homologous recombination decreased by about 30% in the msh2 mutant (P = 0.01), with 9.4 and 13.4 sectors/plant being detected in the mutant and wild type, respectively. Recombination between homeologous substrates with 1.9% of sequence divergence was decreased to 2.6 sectors/plant in wild-type plants, indicating that sequence divergence significantly reduced the frequency of recombination. In contrast, it was 6.8-fold higher in the msh2 mutant, with 17.7 sectors/plant. This, however, confounds two opposing effects. Indeed, if AtMSH2 played no role in homeologous recombination, we would have expected to see approximately 1.8 sector/plant due to the decreased frequency of homologous recombination described previously (0.7 x 2.6 sectors/plant = 1.82 sector/plant). In fact, the results observed (17.7 sectors/plant) correspond to a 9.7-fold increase in homeologous recombination when the data were normalized to compensate for the impact of the msh2 mutation on homologous recombination. AtMSH2 thus appears to play a role both in promoting homologous recombination and in inhibiting homeologous recombination in Arabidopsis .

To confirm that the disruption of the AtMSH2 gene was directly responsible for this decrease in homeologous recombination, a genetic complementation experiment was conducted. A 3051-bp cDNA covering the entire AtMSH2 ORF under the control of a CaMV35S promoter was introduced into msh2 mutant plants carrying the homeologous recombination reporter. T1 lines hemizygous for either the AtMSH2 cDNA construct or a control T-DNA (unmodified pCAMBIA) were compared to evaluate the effect of the AtMSH2 transgene on homeologous recombination in a msh2 mutant background (Table 4.1). If the inactivation of the AtMSH2 gene were indeed responsible for the effects described above, we would expect the complementation of the mutant (with a full-length cDNA) to restore the reduced frequency of recombination seen in the wild-type line with the homeologous substrate. This is indeed what was observed as 3.8 sectors/plant were seen in T1 transformants, a result that is very close to the 2.6 sectors/plant seen in the wild-type line. When transformed with the empty vector, 2.5-fold more sectors were seen again, reflecting the increased rate of homeologous recombination in a msh2 background. In relative terms, again taking into account the effect of AtMSH2 on homologous recombination, homeologous recombination in a complemented (+/+) line was increased 3.6-fold relative to a non-complemented (-/-) mutant. This result confirms that the absence of a functional AtMSH2 gene is responsible for the increase in homeologous recombination observed in the msh2 insertional mutant.

The msh2 mutant was crossed with two reporter lines that carry either a G7 or G16 synthetic microsatellite (Azaiez et al ., in preparation) to allow the measurement of microsatellite instability (MSI) as blue sectors. F3 plants homozygous for both the reporter construct and either the wild-type or mutant allele were compared to evaluate the effect of a msh2 mutation on microsatellite instability.

The instability of the G7 synthetic microsatellite was increased 3.8-fold in somatic tissues of the msh2 mutant, with an average of 2.7 sectors/plant compared to 0.7 sectors/plant in wild type (Table 4.2). By contrast, the level of instability of the G16 synthetic microsatellite could not be estimated accurately in somatic tissues of the msh2 mutant because of the strikingly different staining pattern of plants carrying this reporter (Figure 4.3). Whereas the wild-type plants exhibited a large number of relatively small sectors (the largest occupying only a small proportion of the surface of a leaf), the mutants showed sectors with a much broader size distribution (some even covering half a leaf). This important difference suggests that mutations in the microsatellite are occurring earlier in development and that the frequency of mutation is increased. Estimates of the rate of somatic mutation at the G16 synthetic microsatellite were nonetheless made on a small sample of F2 wild-type homozygotes and heterozygotes (Table 4.2). Interestingly, the instability of the G16 synthetic microsatellite seemed to increase 2.3-fold in plants heterozygous for the mutant allele, with 1432.5 and 3300.0 sectors/plant being detected in wild-type homozygotes and heterozygotes, respectively. Since the instability of the G16 synthetic microsatellite is very high, germinally-transmitted mutations (detected as entirely blue plants) were present in the F3 progeny. The effect of a msh2 mutation on the instability of the G16 synthetic microsatellite was thus investigated by scoring germinal revertants in the progeny from the msh2 mutants and wild-type plants. Whereas the frequency of germinally-transmitted mutations was limited to 0.4% in wild-type, it reached 18.7% in the mutant, a 52-fold increase (Table 4.2). Together, these results indicate that AtMSH2 plays a significant role in ensuring the stability of mononucleotide microsatellites with a high number of repeats (such as G7 and G16 microsatellites) and in decreasing the germinal transmission of somatic mutations for these microsatellites. Our results further suggest that the msh2 mutant allele can have an impact on the mutation rate even in heterozygous plants.

The number of sectors was scored on plants at the 10- to 14-leaf stage. For assays with the G7 synthetic microsatellite, 29 mutant and 62 wild-type homozygotes were scored. For assays with the G16 synthetic microsatellite, two plants each were scored for the mutants and heterozygotes. The percentage of germinal revertants (completely blue plants carrying a germinally-transmitted mutation) for the G16 synthetic microsatellite was calculated by screening a total of 1356 and 2240 F3 or F4 plants for the msh2 mutant and wild-type lines, respectively. ND: not determined.

Figure 4.3. Comparison between plants carrying the G16 microsatellite construct in a msh2 mutant (M), wild-type (W) or heterozygous (H) genetic background.

Somatic regions where a mutation has created a functional (in-frame) GUS gene are detected as blue sectors after staining with X-Gluc and bleaching with ethanol.

To evaluate the long-term effect of a msh2 mutation on the variability of endogenous microsatellites with a high number of repeats, three T4 msh2 mutant plants (derived from a single T3 plant) and two wild-type Columbia plants were propagated by single seed descent for three generations before being screened for the presence of novel alleles at six endogenous dinucleotide microsatellites ranging in size from 21 to 41 repeats (Table 4.3). Nine to twenty T7 plants were examined for each of the five lines (3 mutant, 2 wild-type) and scored as carrying either the wild-type allele (A) or a novel allele (B, C, ...). At least ten distinct mutation events were detected among the 18 locus/line combinations in the mutant lines (3 lines x 6 loci), while only one mutation event was detected at the 12 possible locus/line combinations in the wild-type lines (2 lines x 6 loci). Novel alleles were present at four of the six microsatellite loci surveyed among the mutant progenies, but only at one locus (NGA1107) in the wild-type background.

Two types of distribution were noted for these new alleles: some were present in only one or a few individuals from a line, while others were present in most individuals from a line. Indeed, there were five cases where a new allele was detected in only a few individuals among mutant progeny (lines 1 and 3 for ICE3, lines 1 and 2 for NGA8, and line 3 for ICE10). In most cases, however, the new alleles were present in all or nearly all of the plants tested for a line. For example, the main allele present at locus NGA1107 in one of the wild-type lines was different from the allele present in all of the other lines. Similarly, for loci NGA172 and NGA8, the predominant allele present in one of the mutant lines was different from the wild-type Col allele while at locus ICE3, which contains the longest microsatellite tested (41 repeats), all three mutant lines were fixed or nearly fixed for an allele other than the Col allele, with two lines sharing the same allele (2 bp longer than the Col allele) and the other having a third allele which was 4 bp longer than the Col allele. Since this last allele is 2 bp (one additional repeat) longer that the other ICE3 mutant allele, it could possibly represent a secondary mutation of this allele. Our results thus support a crucial role for AtMSH2 in ensuring long-term stability at microsatellite loci during sexual reproduction in Arabidopsis .

Three T4 plants from the msh2 mutant line (line SALK_002708) and two wild-type Columbia plants were propagated by single seed descent for three generations. Nine to twenty plants per progeny were genotyped and scored as "A" when carrying the Columbia (Col) allele, and as "B" or "C" when carrying a novel allele (bold) corresponding to either the loss or gain of one repeat (2 bp) compared to the Col allele, except in ICE3 where C is two repeats (4 bp) longer than the Col allele.

In the present paper, a T-DNA insertional mutant was used to investigate the impact of the loss of AtMSH2 activity on recombination and MSI. We described evidence that AtMSH2 could play three roles in Arabidopsis : first in promoting homologous recombination, second in aborting recombination in the presence of sequence divergence, and third in preventing small insertion/deletion mutations that occur during the replication of microsatellites.

In the present study, homologous recombination was 30% higher in wild-type plants than in the msh2 mutant. A small stimulatory effect of MSH2 on homologous recombination has also been noted in yeast (Saparbaev et al ., 1996, Sugawara et al ., 1997). This role was, however, found to be MMR-independent and instead requires the cooperation of the MSH2/MSH3 complex with a RAD1/RAD10 complex to remove unpaired 3' tails in recombination intermediates.

Our experiments also demonstrated that 1.9% of sequence divergence decreases recombination by 5.2-fold in wild-type plants. We confirmed experimentally, by mutant analysis and genetic complementation, that AtMSH2 is responsible for this decrease in homeologous recombination in Arabidopsis . The anti-recombination effect of MSH2 on homeologous recombination has been well documented in other organisms where it is attributed to the ability of the MMR machinery to detect heterologies in recombination substrates and halt the recombination event by reversing the formation of the recombination intermediate (heteroduplex rejection) (Harfe and Jinks-Robertson 2000a). The 9.7-fold increase in homeologous recombination detected in the Arabidopsis mutant is similar to what has been reported in other species. For instance, a 9.6-fold increase in homeologous recombination was detected for a sequence divergence of 1% in a yeast msh2 mutant (Datta et al ., 1997). Similarly, substrates with 1.5% sequence divergence are 10-fold more likely to undergo recombination in msh2 mammalian cells than in wild-type cells (Elliot and Jasin 2001).

Since AtMSH2 was found to regulate homeologous recombination in Arabidopsis , this gene may play a role in reproductive isolation between closely related plant species, in a way similar to the barrier that has been documented between yeast species (Hunter et al ., 1996). In the absence of the anti-recombination properties of the MSH2 protein, the level of genetic exchanges between homeologous chromosomes in a hybrid formed between plant species with 1.9% of sequence divergence could be predicted to be at least 9.7-fold higher that it normally is. If such an increase could be harnessed, it could facilitate plant breeding by decreasing the number of generations needed to introgress valuable traits from wild relatives into elite cultivars.

There was a 3.8-fold increase in MSI at a G7 microsatellite in somatic tissues, and a 52-fold increase in MSI at a G16 microsatellite in germinal tissues in the msh2 mutant. The anti-mutation effect of MSH2 has been well characterized in other organisms and is attributed to the interaction of the MMR machinery with the replication complex to assist in correcting small indels created by polymerase slippage, especially in microsatellite loci containing a high number of mono or dinucleotide repeats (Harfe and Jinks-Robertson 2000a). However, the increases in MSI detected in the Arabidopsis mutant are much lower than those reported in yeast msh2 mutants, in which Sia et al . (1997) found that a G18 microsatellite was destabilized 6300-fold, while Tran et al . (1997) detected a 408 and 9756-fold increase in MSI for a A7 and A14 microsatellites, respectively.

Why then does the loss of MSH2 in Arabidopsis lead to such a modest increase in MSI relative to what has been reported in other species? There could be two explanations for this phenomenon. The first is that line SALK_002708 may not be a null msh2 mutant and could still produce a low amount of functional MSH2 protein, sufficient to sustain some MMR function. An alternative hypothesis would be that the MMR system may not contribute much to mutation avoidance in Arabidopsis so that its inactivation would have a low impact on the mutation rate.

As stated above, the presence of some functional MSH2 protein and, thus, of a residual MMR activity could explain the modest increase in MSI detected in line SALK_002708. However, many observations argue against this hypothesis. First, no transcripts from the mutant allele were detected by a RT-PCR assay sensitive enough to detect the low copy AtMLH1 transcripts. Furthermore, since this line carries a T-DNA insertion as well as a 1.5 kb deletion in the AtMSH2 gene, any residual transcripts would produce a truncated MSH2 protein lacking an essential N-terminal functional domain (Schofield and Hsieh 2003). Second, our results indicate that line SALK_002708 displayed an increase in homeologous recombination similar to those reported in msh2 mutants from other species. Similarly, the 2.3-fold increase in MSI that was detected in Arabidopsis plants heterozygous for this mutant allele is reminiscent of the 3-fold mutator phenotype that was observed in yeast diploids heterozygous for a msh2 deletion (Drotschmann et al., 1999). Third, the increase in MSI at endogenous dinucleotide microsatellite loci that was documented by Leonard et al . (2003) with a RNA-interference (RNAi) msh2 mutant was similar to the one found in line SALK_002708. These authors also detected a 5-fold increase in MSI at a G7 synthetic microsatellite in line SALK_002708, a result similar to ours (3.8-fold increase). Finally, in yeast, a mutation in the PSM1 gene (another component of the MMR system) is known to induce a level of MSI similar to that observed in a msh2 mutant (Harfe and Jinks-Robertson 2000b, Gragg et al ., 2002). However, increases in MSI similar to ours (2- to 28-fold) were recently reported for Arabidopsis lines in which MMR had been inactivated through the expression of dominant negative alleles of the Arabidopsis pms1 gene (Alou et al ., 2004). From all these observations, we conclude that line SALK_002708 is most likely a true null msh2 mutant and that the low increase in MSI detected in this line must really reflect what happens in Arabidopsis in absence of MMR activity.

Again one is left to wonder why the loss of MMR function has such a modest impact on MSI. This could easily be explained if the MMR system did not play a major role in mutation avoidance in Arabidopsis . Indeed, it has already been postulated that the MMR system may be more or less inactive in somatic tissues in Arabidopsis (Kovalchuk et al ., 2000, Leonard et al ., 2003). In support for this theory, Adé et al . (1999) were only able to detect AtMSH2 transcripts in actively dividing cell suspensions. Furthermore, a rate of somatic point mutations exceeding by at least 100-fold those estimated for yeast and other eukaryotes has been reported for wild-type Arabidopsis (Kovalchuk et al ., 2000). All these observations point toward the fact that DNA repair may not be very active in Arabidopsis somatic tissues. It would, however, be expected to put more efforts in protecting the meristematic tissues that later give rise to its reproductive organs and germline since unrepaired somatic mutations can potentially be passed on to subsequent generations. Indeed, the increase in germinally-transmitted mutations that was detected in the msh2 mutant was more important than in the somatic tissues (52-fold vs about 3.8-fold), thus supporting the hypothesis that MMR may be more active in these tissues. Nonetheless, the increase in MSI in germinal tissues is still modest compared to those observed in other species. This suggests that, even in wild-type plants, a lot of mutations may go unrepaired and potentially be transmitted to the next generation, thus promoting genetic variability among Arabidopsis natural populations.

After three generations of propagation by single seed descent, our results show that msh2 mutant lines display new alleles at many endogenous microsatellite loci but at only one novel allele at one locus in the wild-type lines. This is similar to what is reported in other species. Indeed, Drosophila melanogaster spel1 (the fly msh2 -homologue) and Caenorhabditis elegans msh2 mutants also display highly elevated MSI at endogenous dinucleotide microsatellite loci when analyzed after several generations of propagation (Flores and Engels 1999; Degtyareva et al ., 2002). All T7 plants carrying a new allele are expected to be at least heterozygous or possibly even homozygous for this allele since the level of sensitivity of our PCR assay would not allow the detection of somatic mutations (mutant sectors) (for a more complete discussion on the sensitivity of a similar PCR assay, see Leonard et al ., 2003). Consequently, the new alleles detected in the present study likely correspond to germinally-transmitted mutations.

Two types of distribution were noted for these new alleles: some were present in only one or a few individuals from a line, while others were present in most individuals from a line. When an allele is present in only a few individuals, it may represent a mutation that originated in the previous generation and was germinally transmitted by one of the few gametes carrying this mutated allele. The germinal transmission of mutated dinucleotide microsatellite alleles would be expected to be increased in the msh2 mutant, since an increase in the frequency of germinally transmitted mutations was detected with our G16 microsatellite reporter, probably as a results of the increase in the number and size of the mutated somatic sectors that was noted in homozygous mutant plants.

When all or nearly all of the T7 plants tested for a line carry the same allele, this strongly suggests that the plant used for propagation by single seed descent in the previous generation was at least heterozygous or more likely homozygous for this allele. The presence of a plant homozygous for a novel allele in one of the wild-type lines may be a random event or it could reflect the high mutation rate naturally present in Arabidopsis . On the other hand, the presence of homozygous plants among T6 progeny would not be unexpected since plants heterozygous for new microsatellite alleles were present as early as among T3 progeny in line SALK_002708. Indeed, Leonard et al . (2003) detected a 3:1 segregation ratio for new alleles at three microsatellite loci in T4 progenies derived from this msh2 line. This early appearance of new alleles could be a consequence of the mutator phenotype expected in plants homozygous or even heterozygous for the msh2 allele. An elevated mutation rate would thus be expected to be present in the hemizygous T1, the homozygous mutant and heterozygous T2 plants, as well as in all later generations since the T3 progeny from line SALK_002708 is known to be homozygous for the msh2 allele (see Results, as well as Leonard et al ., 2003). This mutator phenotype would increase the probability that individuals heterozygous for new alleles at endogenous microsatellite loci would be created and, in a selfing species like Arabidopsis , the reproduction of such individuals could lead to a rapid fixation of these alleles in natural populations.

Finally, serially-cultured msh2 nematodes were found to die out much more quickly than those from wild-type worms (Degtyareva et al ., 2002). Indeed, under their experimental conditions, 95% of the msh2 lines had become extinct after 20 generations compared to only 20% of the wild-type lines. These authors concluded that under the elevated mutation rates induced by the msh2 background, mutations affecting fertility or viability must rapidly accumulate with passing generations. Whether this will also be the case in the Arabidopsis mutant remains to be determined. Only a modest increase in MSI was, however, detected in the msh2 mutant. Since wild-type Arabidopsis are known to naturally harbor a high mutation rate (Kovalchuk et al., 2000), it may therefore have developed mechanisms to prevent any long-term deleterious side effect associated with such a phenotype. Furthermore, a small segregation bias against the germinal transmission of the msh2 mutant allele was detected in our experiments. This could suggest that, although gametes carrying the msh2 allele are capable of a level of fertilization sufficient to achieve an overall normal seed set in a homozygous mutant plant, they may be at a slight disadvantage in an heterozygous plant when competing against gametes carrying the normal AtMSH2 allele. This may be a way to regulate the spreading of such mutator mutations and to prevent the accumulation of deleterious mutations in natural populations. Since such a bias was not observed in other species, it could, however, reflect a problem during the replication or recombination of the msh2 allele present in line SALK_002708. Whether this segregation bias and the low increase in MSI detected in absence of MMR are traits characteristic of this mutant allele, of Arabidopsis , or of plants in general still remains to be determined.

Adé J, Belzile F, Philippe H, Doutriaux MP . (1999) Four mismatch repair paralogues coexist in Arabidopsis thaliana : AtMSH2 , AtMSH3 , AtMSH6-1 and AtMSH6-2 . Mol Gen Genet. 262 :239-49.

Adé J, Haffani Y, Belzile FJ . (2001) Functional analysis of the Arabidopsis thaliana mismatch repair gene MSH2 . Genome. 44 :651-7.

Alou AH. Azaiez A, Jean M, Belzile FJ . (2004) Involvement of Arabidopsis thaliana AtPMS1 gene in somatic repeat instability. Plant Mol Biol. (in press).

Baumann E, Lewald J, Saedler H, Schulz B, Wisman E . (1998) Successful PCR-based reverse genetic screens using an En-1 -mutagenised Arabidopsis thaliana population generated via single-seed descent. Theor Appl Genet. 97 :729-34.

Clauss MJ, Cobban H, Mitchell-Olds T . (2002) Cross-species microsatellite markers for elucidating population genetic structure in Arabidopsis and Arabis ( Brassicaeae ). Mol Ecol. 11 :591-601.

Clough SJ, Bent AF . (1998) Floral dip: a simplified method for Agrobacterium -mediated transformation of Arabidopsis thaliana . Plant J. 16 :735-43.

Culligan KM, Hays JB . (1997) DNA mismatch repair in plants. An Arabidopsis thaliana gene that predicts a protein belonging to the MSH2 subfamily of eukaryotic MutS homologs. Plant Physiol. 115 :833-9.

Culligan KM, Hays JB . (2000) Arabidopsis MutS homologs-AtMSH2, AtMSH3, AtMSH6, and a novel AtMSH7-form three distinct protein heterodimers with different specificities for mismatched DNA. Plant Cell. 12 :991-1002.

Datta A, Hendrix M, Lipsitch M, Jinks-Robertson S. (1997) Dual roles for DNA sequence identity and the mismatch repair system in the regulation of mitotic crossing-over in yeast. Proc Natl Acad Sci U S A. 94 :9757-62.

Degtyareva NP, Greenwell P, Hofmann ER, Hengartner MO, Zhang L, Culotti JG, Petes TD . (2002) Caenorhabditis elegans DNA mismatch repair gene msh-2 is required for microsatellite stability and maintenance of genome integrity. Proc Natl Acad Sci U S A . 99 :2158-63.

Drotschmann K, Clark AB, Tran HT, Resnick MA, Gordenin DA, Kunkel TA . (1999) Mutator phenotypes of yeast strains heterozygous for mutations in the MSH2 gene. Proc Natl Acad Sci U S A. 96 :2970-5.

Dubois P, Cutler S, Belzile FJ . (1998) Regional insertional mutagenesis on chromosome III of Arabidopsis thaliana using the maize Ac element. Plant J. 13 :141-51.

Edwards K, Johnstone C, Thompson C . (1991) A simple and rapid method for the preparation of plant genomic DNA for PCR analysis. Nucleic Acids Res. 19 :1349.

Elliott B, Jasin M . (2001) Repair of double-strand breaks by homologous recombination in mismatch repair-defective mammalian cells. Mol Cell Biol. 21 :2671-82.

Flores C, Engels W . (1999) Microsatellite instability in Drosophila spellchecker1 (MutS homolog) mutants. Proc Natl Acad Sci U S A. 96 :2964-9.

Gragg H, Harfe BD, Jinks-Robertson S. (2002) Base composition of mononucleotide runs affects DNA polymerase slippage and removal of frameshift intermediates by mismatch repair in Saccharomyces cerevisiae . Mol Cell Biol . 22 :8756-62.

Harfe BD, Jinks-Robertson S . (2000a) DNA mismatch repair and genetic instability. Annu Rev Genet. 34 :359-399.

Harfe BD, Jinks-Robertson S . (2000b) Sequence composition and context effects on the generation and repair of frameshift intermediates in mononucleotide runs in Saccharomyces cerevisiae . Genetics. 156 :571-8.

Hunter N, Chambers SR, Louis EJ, Borts RH. (1996) The mismatch repair system contributes to meiotic sterility in an interspecific yeast hybrid. EMBO J. 15 :1726-33.

Jean M, Pelletier J, Hilpert M, Belzile F, Kunze R. (1999) Isolation and characterization of AtMLH1 , a MutL homologue from Arabidopsis thaliana . Mol Gen Genet. 262 :633-42.

Kovalchuk I, Kovalchuk O, Hohn B . (2000) Genome-wide variation of the somatic mutation frequency in transgenic plants. EMBO J. 19 :4431-8.

Leonard JM, Bollmann SR, Hays JB . (2003) Reduction of stability of Arabidopsis genomic and transgenic DNA-repeat sequences (microsatellites) by inactivation of AtMSH2 mismatch-repair function. Plant Physiol. 133 :328-38.

Li L, Santerre-Ayotte S, Boivin EB, Jean M, Belzile F (2004) A novel reporter for intrachromosomal homoeologous recombination in Arabidopsis thaliana . Plant J. 40 : 1007-1015.

Liu B, Parsons R, Papadopoulos N, Nicolaides NC, Lynch HT, Watson P, Jass JR, Dunlop M, Wyllie A, Peltomaki P, de la Chapelle A, Hamilton SR, Vogelstein B, Kinzler KW . (1996) Analysis of mismatch repair genes in hereditary non-polyposis colorectal cancer patients. Nat Med. 2 :169-74.

Rayssiguier C, Thaler DS, Radman M. (1989) The barrier to recombination between Escherichia coli and Salmonella typhimurium is disrupted in mismatch-repair mutants. Nature. 342 :396-401.

Saparbaev M, Prakash L, Prakash S. (1996) Requirement of mismatch repair genes MSH2 and MSH3 in the RAD1-RAD10 pathway of mitotic recombination in Saccharomyces cerevisiae . Genetics. 142 :727-36.

Schofield MJ, Hsieh P . (2003) DNA mismatch repair: molecular mechanisms and biological function. Annu Rev Microbiol. 57 :579-608.

Sia EA, Kokoska RJ, Dominska M, Greenwell P, Petes TD . (1997) Microsatellite instability in yeast: dependence on repeat unit size and DNA mismatch repair genes. Mol Cell Biol . 17 :2851-8.

Sugawara N, Paques F, Colaiacovo M, Haber JE. (1997) Role of Saccharomyces cerevisiae Msh2 and Msh3 repair proteins in double-strand break-induced recombination. Proc Natl Acad Sci U S A. 94 :9214-9.

Tran HT, Keen JD, Kricker M, Resnick MA, Gordenin DA . (1997) Hypermutability of homonucleotide runs in mismatch repair and DNA polymerase proofreading yeast mutants. Mol Cell Biol. 17 :2859-65.

Valvekens D, Van Montagu M, Van Lijsebettens M. (1988) Agrobacterium tumefaciens -mediated transformation of Arabidopsis thaliana root explants by using kanamycin selection. Proc Natl Acad Sci U S A. 85 :5536-40.

Verwoerd TC, Dekker BM, Hoekema A . (1989) A small-scale procedure for the rapid isolation of plant RNAs. Nucleic Acids Res. 17 :2362.

Wu SY, Culligan K, Lamers M, Hays J . (2003) Dissimilar mispair-recognition spectra of Arabidopsis DNA-mismatch-repair proteins MSH2*MSH6 (MutSalpha) and MSH2*MSH7 (MutSgamma). Nucleic Acids Res. 31 :6027-34.