Chapitre III : Involvement of the Arabidopsis thaliana AtPMS1 gene in somatic repeat instability

Table des matières

[Plant Molecular Biology, 2004, Vol. 56. No 3: 339-349]

Involvement of the Arabidopsis thaliana AtPMS1 gene in somatic repeat instability

A. H. Alou1,2,†, A. Azaiez1, †, M. Jean1, and F. J. Belzile1*

1 Département de Phytologie, Pavillon Marchand, Université Laval, Québec, Canada G1K 7P4

2 Institut National de Recherches Agronomiques du Niger (INRAN), B.P. 429, Niamey, Niger

A.H.A. and A.A. contributed equally to this work.

Key words : DNA repair, MMR, MutL homolog, PMS1 , microsatellite instability.

Les gènes de correction des mésappariements de bases assurent la stabilité génétique de tous les organismes sur Terre. Le gène AtPMS1 est un membre clef de la famille de gènes impliqués dans la correction des bases malencontreusement incorporées par la polymérase, ou résultant d'homéologie de séquences. Les objectifs de cette recherche visaient à inactiver le gène AtPMS1 , et à étudier son implication dans le maintien de la stabilité du génome. Pour le premier objectif, nous avons surexprimé une forme altérée (ou une autre tronquée) du gène AtPMS1 sauvage. Dans la première forme allélique nous avons changé deux (2) acides aminés dans la boîte MutL (GFRGEAL), de façon à générer une mutation faux-sens. Dans la deuxième forme allélique, nous avons amputé 87 acides aminés autour de la boîte MutL, tronquant ainsi la protéine codée par le gène normal. Les deux formes alléliques maintenaient le cadre de lecture ouvert et retenaient toute leur capacité d'interaction protéine-protéine dans la partie C-terminale. Sur le plan morphologique, les transformants issus de ces constructions n’étaient pas différents des plantes sauvages. Nous avons toutefois observé un léger jaunissement des transformants. Pour le deuxième objectif, nous avons inactivé le gène gusA en lui insérant un microsatellite synthétique de 7 ou 16 guanines. Les dérapages de la polymérase dans ces séquences devaient engendrer la réversion en cadre du gène gusA. En comparant le nombre de secteurs bleus induits par les transformants à ceux des plantes contrôles, nous avons obtenu une mesure indirecte de l'inactivation du gène AtPMS1 sur la stabilité des microsatellites. Nos résultats montrent qu'il est possible d'inactiver le gène AtPMS1 par cette approche, et que les transformants induisent 5 à 28 fois plus d'instabilité que les plantes témoins, et ce en fonction du type de microsatellite utilisé. Cette approche peut permettre à l'amélioration variétale d'augmenter le taux de recombinaison génétique entre espèces génétiquement éloignées afin d'accroître la performance des plantes cultivées en puisant les caractères désirables dans les espèces sauvages.

Mismatch repair (MMR) genes participate in the maintenance of genome stability in all organisms. Based on its high degree of sequence conservation, it seems likely that the AtPMS1 gene of Arabidopsis thaliana is part of the MMR system in this model plant. To test this hypothesis, we aimed to disrupt AtPMS1 function by over-expressing mutated alleles expected to result in a dominant negative effect. To create one mutant allele we substituted two amino acids in the MutL-box, and for the other mutant allele we deleted 87 amino acids comprising the whole MutL-box. Contrary to published reports in some eukaryotes, transgenic plants expressing these alleles did not exhibit a decrease in fertility nor any other visible phenotype. To examine the impact of these mutations on microsatellite instability, the phenotype most often observed in organisms defective in MMR, reporter lines containing a uidA (GUS) gene inactivated by the insertion of a synthetic microsatellite (G7 or G16) were used. GUS gene function in these lines can be restored following the loss of one base or the gain of two bases in the repetitive tract. This results in the observation of blue sectors on a white background following histochemical staining. In a subset of the transformants, a significant increase (2- to 28-fold) in microsatellite instability was observed relative to wild-type. This report shows that MMR function can be disrupted via a dominant negative approach. It is the first report to describe the phenotypic consequence of disrupting the function of a MutL homolog in plants.

The mismatch repair (MMR) system is specialized in the correction of mispaired bases. Mismatched bases can arise as a result of DNA replication errors, exposure to various natural and man-made mutagens, or as a consequence of sequence divergence between homoeologous chromosomes which undergo recombination (Modrich, 1991). Perhaps because of the universality of such damages, the MMR system has been highly conserved throughout all kingdoms.

The MMR system is best characterized in Escherichia coli , where the major components for base-base mismatches and the correction of short insertion/deletion loops are products of the MutHLS system (Harfe and Jinks-Robertson, 2000). A MutS homodimer recognizes and binds to mismatched DNA upon which a MutL homodimer is recruited. The latter facilitates the interaction of MutS with MutH, an endonuclease that proceeds to attack the strand of DNA bearing the incorrect base(s). Following the removal of the erroneous sequence, the single stranded template is once again replicated to yield a corrected DNA duplex.

While the role of MutS and MutH in the repair process is well documented, the role of MutL has been more difficult to elucidate. MutL protein presumably plays the role of a "molecular matchmaker" between several proteins involved in MMR and other cellular functions. It apparently plays its multiple bridging roles through induced conformational and structural changes resulting from various interactions involving its N- and C-terminal regions (Schofield and Hsieh, 2003).

Crystallography has suggested that dimerization of the N-terminal portion of the MutL protein occurs as a result of ATP binding (Ban et al. , 1999), and this has been speculated to play a role in the MMR initiation function. The C-terminal portion of the MutL protein has been linked to protein-protein interactions with MutS (Galio et al. , 1999; Wu and Marinus, 1999), MutH (Ban and Yang, 1998b; Hall and Matson, 1999), and also the helicase UvrD (Dao and Modrich, 1998; Hall et al. , 1998; Yamaguchi et al. , 1998).

In eukaryotes, MutL functions are carried out by four homologues (MLH1, MLH2, MLH3 and PMS1) that form heterodimers known as MutLα (MLH1-PMS1), MutLβ (MLH1-MLH3) or MutLγ (MLH1-MLH2) (Jiricny, 1998; Harfe and Jinks-Robertson, 2000; Jiricny, 2000). Each of these dimers exhibits a preference for the type of mismatch it is most efficient at repairing. MLH1 obviously plays a central role as it is involved in each heterodimer, but PMS1 is also very important as it forms, with MLH1, the major dimer involved in the repair of single mispaired bases (MutLα). Mutations in either protein have similar phenotypic consequences, with symptoms being usually less severe for a defect in PMS1 relative to MLH1 as some functional overlap exists between the various dimers (Flores-Rozas and Kolodner, 1998; Yao et al. , 1999b).

Mutations in the PMS1 gene produce both somatic and germinal effects. The phenotype most often documented in somatic cells is a dramatic increase in microsatellite instability (MSI) (Boland et al. , 1998; Lipkin et al. , 2000; 2001; Leonard et al. , 2003). Microsatellites are simple tandem repeats of one to five or six nucleotides. Microsatellites are inherently unstable as a consequence of a high level of DNA polymerase slippage on repetitive DNA during the replication process (Sia et al. , 1997; Tran et al. , 1997; Sia et al. , 2001). Failure of the MMR system to correct the ensuing mismatch results in novel alleles that are either longer or shorter. For example, it has been found that mononucleotide runs of 9 to 14 adenines were 1,700 to 10,000 times more unstable in PMS1-deficient yeast cells than in PMS1-proficient cells (Tran et al. , 1997). In addition, germinal effects attributed to mutations in the PMS1 gene include male sterility in mouse (Baker et al. , 1995; Prolla et al. , 1998) and higher spore lethality, increased post-meiotic segregation, and an increased meiotic recombination rate between homoeologous chromosomes in yeast (Chambers et al. , 1996).

Certain mutations in PMS1 are particularly interesting as they confer a dominant negative phenotype. Such alleles, when overexpressed in a wild-type cell, result in a phenotype similar to that produced by the deletion of the PMS1 gene. Mutations F126A in the Saccharomyces cerevisiae PMS1 gene and R95F in the MutL gene of E. coli are located in a highly conserved N-terminal domain call the MutL-box (GFRGEAL) and both have been demonstrated to act in such a dominant negative fashion (Aronshtam and Marinus, 1996; Peltomaki and de la Chapelle, 1997; Ban and Yang, 1998a, 1998b; Ban et al. , 1999). Such alleles would be particularly interesting to investigate MMR function in plants because, for most species, MMR mutants are not readily available and cannot be obtained using gene knockout strategies available in other species.

In this work, we used a dominant negative mutation approach to investigate the role of the AtPMS1 gene on genome stability in Arabidopsis . Transgenic plants harboring either of two mutant alleles of AtPMS1 were produced and characterized phenotypically, with a particular emphasis being placed on the assessment of microsatellite instability.

pms1-FR AF allele

We used the QuikChange kit (Stratagene, Vancouver, Canada) to make specific mutations in the GFRGEAL motif of the AtPMS1 gene. We changed codon F108 into A108, and codon R109 into F109 (Figure 3.1A). The mutagenesis was performed on a plasmid (pPMS500) containing a 500bp Sac I fragment from the AtPMS1 cDNA cloned into pBlueScript (Stratagene, Vancouver, Canada). pPMS500 was amplified using a high fidelity DNA polymerase (Pfu-Turbo; Stratagene, Vancouver, Canada) and two mutagenic primers, each specific for one strand: PmsMutF (5’-CTACTTATGGTGCTTTTGGAGAAGCTTTGAGCTCCAC-3') and PmsMutR (5'-GTGGAGCTCAAAGCTTCTCCAAAAGCACCATAAGTAG-3’). A Hind III site was introduced in the gene by changing the third nucleotide of codon GCC (A112) into GCT (Figure 3.1A). This additional Hind III site differentiates the pms1-FR AF allele from wild type. Amplification was performed in a Perkin Elmer thermocycler (model 480) with a 2 min denaturation at 95°C, followed by 17 cycles of 1 min at 95°C, 1 min at 45°C, and 10 min at 68°C. The dam-methylated template was digested with Dpn I and the newly synthesized DNA was extracted from 1% agarose gel using the GFX kit (Amersham Biosciences, Piscataway, USA). Plasmid DNA was transferred into E. coli DH5α. Transformant clones were selected on the basis of the presence of the Hind III restriction site and sequenced with the M13 forward primer (5'-CGCCAGGGTTTTCCCAGTCACGAC-3') to confirm the absence of random mutations within the cloned fragment. A clone with the intended mutations was digested with Sac I, and the altered 500bp fragment was substituted for the wild-type sequence in the AtPMS1 cDNA (Figure 3.1B). The altered versions of the AtPMS1 gene (on BamH I/ EcoR V fragments) were cloned into the binary vector pBI121 ( BamH I/ Ecl 136) and transferred into Agrobacterium tumefaciens strain LBA4401 for plant transformation.

pms1-∆87 allele

We used restriction enzymes Ehe I and Bcl I to generate the pms1-∆87 allele by deleting a 261bp fragment from the AtPMS1 cDNA (Figure 3.1B). The Bcl I end of the digested plasmid was filled in using the Klenow fragment, and the truncated plasmid was re-circularized through a blunt-end ligation. Following mutagenesis, the pms1-∆87 allele was cloned as above described for plant transformation.

Figure 3.1: Schematic diagram of the plasmid constructs harboring the mutant pms1 alleles.

(A) pms1-FR AF allele. The lower portion of the diagram shows the overall composition of the T-DNA. The location and orientation of primers AtPMS4 (4), AtPMS5 (5), AtPMS6 (6), and M13 forward (MF) used to genotype pms1 transformant plants are indicated. The location and orientation of primers NPR and NTF used for the detection of the T - DNA tandem insertions are also shown. The upper portion of the diagram illustrates the sequence differences between the wild-type and pms1-FR AF alleles. Highlighted in bold are the two amino acids mutated in the MutL-box (GFRGEAL) as well as the novel Hin dIII site engineered into the pms1-FR AF allele. (B) pms1-Δ87 allele. As above, the lower portion of the diagram indicates the position of relevant primers and restriction sites along the T - DNA. The upper portion shows the Ehe I- Bcl I region (comprising the MutL-box) deleted from the AtPMS1 gene to generate this allele.

Plants were transformed by dipping inflorescences in a solution containing Agrobacterium tumefaciens strain LBA4401 carrying the appropriate construct (Bent, 2000). Transformants were selected by plating T1 seeds on germination medium (Valvekens et al. , 1988) supplemented with 50 µg/ml of kanamycin. Resistant T1 plants were thereafter transplanted into soil (Promix), and transformants were confirmed by PCR using the primers AtPMS4 (5'-GGTGTCATATCT-TCCTTTGAGATG-3') and AtPMS6 (5'-CTGCTCGCCAAATTGGTACCACTG-3'). Amplifications were done in a 20 µl reaction volume containing 0.2 mM dNTPs, 1x Taq buffer (Amersham Biosciences, Piscataway, USA), 500 nM of each primer, and 1.25 U of Taq DNA polymerase (Amersham Biosciences, Piscataway, USA) in presence of 1 µl of genomic DNA. The DNA was extracted using a rapid DNA extraction protocol (Edwards et al. , 1991). Cycling was done in a Perkin Elmer Thermocycler (Model 480) in the following conditions: denaturation (94°C, 30 sec), annealing (50°C, 30 sec), and extension (72°C, 1 min) for 30 rounds of amplification. Each PCR amplification was initiated by a 5 min hot start at 94°C, and completed with a 10 min extension period at 72°C. T2 seeds were collected from selected T1 plants and plated on kanamycin medium. Lines showing a Mendelian 3R:1S ratio were kept for further analysis. T3 seeds were produced from 12 individual T2 plants from each line and these were plated on kanamycin to identify homozygous lines. Homozygous T3 plants were then transplanted to soil for seed production and further molecular characterization. A total of seven pms1-FR AF independent homozygous plants and three pms1-∆87 independent homozygous plants were advanced to the T4 generation.

For G16 containing plants, GUS activity was revealed using the first three fully-developed leaves from 24 two- to three-week old F1 seedlings grown in vitro . For F1 plants with the G7 construct, an entire rosette (~10 leaf stage) from 18 to 24 plants grown in potting mix (Promix) was stained. In both cases, staining was done by incubating the plant parts on ice in 5 ml of buffer solution (100 mM NaPO4 pH7; 0.2 mM X-gluc [5-bromo-4-chloro-3-indolyl-ß-D-glucuronide (Rose Scientific Ltd, Alberta, Canada)]; 0.2% (v/v) Triton X-100). Samples were vacuum-infiltrated (10 min) on ice twice and then incubated in the dark at 37°C for 48 h. The tissues were cleared in 70% ethanol at 37°C for at least 5 h. To ensure complete removal of the chlorophyll, the ethanol was changed several times, then the cleared tissues were stored in 70% ethanol. For each plant the average number of blue sectors per leaf was determined using a binocular microscope. Data collected on each set of F1 progeny were tested for normality prior to means comparisons with the control plants. Where normally distributed (F1 of crosses to G16-B), the mean number of blue sectors per leaf of each pms1 line was compared to the appropriate control line using Student's t-test. In the other two sets of data (crosses to G16-C and G7-A), data were not normally distributed and means comparisons were done using a non-parametric method, Wilcoxon’s rank sum test. All statistical analyses were done with the Statistical Analysis System (SAS Institute 2003, Cary, NC).

To study the function of the PMS1 gene, mutations expected to result in a dominant negative phenotype were produced by either changing two specific amino acids in the highly conserved GFRGEAL motif, or by completely deleting the region of the protein containing this motif. A first mutant allele ( pms1- FR→AF) was obtained by site-specific mutagenesis and resulted in two substitutions (F108A and R109F) (Figure 3.1A), while the second mutant allele ( pms1- ∆87) was obtained by removing a 261bp Ehe I- Bcl I segment coding for 87 amino acids (Figure 3.1B). In the pms1- FR→AF allele, an additional silent mutation was made to introduce a Hind III site within the PMS1 gene by changing the third base of codon 112 (GCC into GCT) (Figure 3.1A). This silent mutation allowed us to distinguish the transformant and wild-type alleles. These constructs were used to transform A. thaliana (ecotype Columbia) and a number of independent transformants were obtained for each allele.

To rapidly eliminate transformants with multi-locus insertions, the T2 progeny were tested for a 3R:1S segregation ratio on kanamycin selection. To eliminate transformants with multiple tandem copies of the inserted T - DNA, transformants were further screened by PCR to identify plants displaying a left and right border (LB/RB) junction. Indeed, PCR with primers NTF/NPR can reveal only the LB/RB junction, leaving LB/LB and RB/RB border junction undetected because such junctions result in formation of a hairpin structure difficult to amplify. Finally, we performed a Southern analysis on those transformants known to have a single T - DNA insertion locus and no tandem insertions to identify individuals with, typically, 1 or 2 copies of the T-DNA (data not shown). In this fashion, seven pms1- FR→AF transformants and three pms1- ∆87 transformants were selected for use in subsequent work. Upon visual examination, these transformants displayed a normal phenotype.

To assess transgene expression in the different transformants, a northern analysis was performed using total RNA from seedlings of the ten transformants and one untransformed control. As can be seen in Figure 3.2A, transcripts hybridizing with the AtPMS1 cDNA were detected in all transformants, whereas no transcript could be detected in the control due to the low level of expression of the endogenous copy of AtPMS1 . Upon probing the same membrane with an actin probe, a uniform signal was obtained in all transformants and in the control, indicating that relatively equal amounts of RNA were loaded and that the control RNA was not degraded (Figure 3.2B).

To further investigate the expression of AtPMS1 gene in the mature leaf tissue and to assess its abundance relative to the transcripts of the mutant alleles, we performed a RT-PCR analysis using total RNA extracted from mature leaves. As can be seen in Figure 3.3, results revealed that the mutant alleles were expressed in all transgenic lines at a higher level than the wild-type allele in control plants. Indeed, while the wild-type AtPMS1 transcript was absent or barely detectable in the control plants (Figure 3.3, lanes 1-4), it was not visible in the pms1 -Δ87 transformants (Figure 3.3, lanes 13-20). This is thought to result from the high abundance of the over-expressed mutant allele that out-competes the normal transcript for the primers. Since the transcript of the pms1- FR→AF allele (Figure 3.3, lanes 5-12) is identical in size to the wild-type allele, the amplicons from these lines were digested with Hind III to allow us to distinguish products from these two alleles. Again, this analysis showed that most, if not all, of the amplification product was from the transgene rather than the wild-type allele (data not shown).

Figure 3.2: Northern analysis of the pms1 transformants.

(A) Expression of the AtPMS1 transgene. Twenty µg of total RNA from transformants (lanes 1-7: pms1-FR AF allele; lanes 8-10: pms1- Δ 87 allele) or Columbia wild-type (lane 11, WT) were separated on a denaturing agarose gel. The complete AtPMS1 cDNA was used as probe for this analysis. (B) Expression of the ACT2 gene. The uniformity of loading was verified by hybridizing the same membrane used above with the ACT2 gene as probe. The size of each transcript (in kb) is indicated on the right.

Figure 3.3: RT-PCR analysis of the wild-type and mutant transcripts.

For each line, four independently-isolated samples of RNA were used for reverse transcription and amplified with gene-specific primers (AtPMS1 and AtPMS2). The RNAs used were as follows: wild-type Columbia (lanes 1–4), pms1-FR AF transformants (lanes 5-8 were from line 1 and lanes 9-12 were from line 2), pms1-Δ87 (lanes 13-16 were from line 1 and lanes 17-20 were from line 2). The size of each amplicon (in bp) is indicated on the right.

To measure microsatellite instability, reporter lines containing a stretch of 16 guanines (G16) or 7 guanines (G7) in the GUS gene were used to characterize the impact of the pms1 mutant alleles. A detailed description of this reporter system will be published elsewhere. The presence of these mononucleotide tracts within the GUS gene resulted in a +1 frameshift in the reading frame. Polymerase slippage resulting in a single base pair deletion, or insertion of two base pairs in the mononucleotide stretch are known to restore gene function (Leonard et al ., 2003). Homozygous T3 plants of the ten pms1 lines and wild-type Columbia were crossed to a homozygous T3 reporter line, G16 - B, carrying a G16 microsatellite. Sufficient F1 seed was obtained for eight of the transformants and the wild-type control. The F1 plants resulting from these crosses were stained and we determined the average number of blue sectors per leaf. As can be seen in Figure 3.4A, control F1 plants (wild-type x G16-B) averaged 12 sectors per leaf. In contrast, all three F1 progenies derived from crosses with transformants harboring the deletion allele ( pms1-Δ87 ) showed significantly more sectors per leaf: 27 (line 1), 31 (line 2) or 29 (line 3). This represents a 2.3- to 2.5-fold increase relative to the F1 control plants. Among the five F1 progenies derived from crosses with transformants harboring the pms1-FR AF , two showed a significant increase in the number of blue sectors per leaf: line 3 (36 sectors) and line 4 (41 sectors). This represents a 3- and 3.4-fold increase relative to the F1 control plants. As for the other F1 progenies with the pms1-FR AF allele, two (lines 2 and 7) exhibited more sectors per leaf than control progeny, but this increase was not statistically significant.

To further characterize the subset of transformants in which a significant increase in microsatellite instability had been observed with the G16-B reporter line, this subset was crossed to a second reporter line carrying the G16 microsatellite, G16 - C. This reporter line is known to generate approximately 10-fold fewer blue sectors than G16 - B (Azaiez et al ., in preparation). Crosses were made between the G16 - C reporter (homozygous T3 plants) and the four transformants that had shown the greatest increase in microsatellite instability (lines 2 and 3 for pms1-Δ87 ; lines 3 and 4 for pms1-FR AF ) as well as with wild-type Columbia. As anticipated based on prior experience with this reporter line, on average only one sector per leaf was observed in the control F1 (Figure 3.4B). In all F1 progenies harboring a mutant pms1 allele, we observed a significant increase in the number of blue sectors per leaf relative to the control plants. For the F1 progenies containing the pms1-Δ87 mutant allele, the number of blue sectors per leaf observed was 25 and 3 for lines 2 and 3, respectively. The average number of blue sectors per leaf observed in the F1 progenies carrying the other mutant allele ( pms1-FR AF ) was 6 and 5 for lines 3 and 4, respectively.

In examining the data obtained with the G16 reporter lines, it strucks us that the impact of the mutant pms1 alleles was greater when the basal rate of reversion in the microsatellite instability reporter was lower. To test if this trend could be extended, we used a third reporter line with a shorter mononucleotide tract (G7), that results in even fewer blue sectors. One line for each pms1 allele ( pms1-Δ87 line 2 and pms1-FR AF line 3) and wild-type Columbia were crossed to a G7 reporter line (G7 - A). The F1 control progeny yielded only 0.1 sector per leaf (Figure 3.4C), while in contrast the F1 progenies harboring the mutant allele showed a marked increase in the number of sectors per leaf: a 17.4-fold increase for pms1-Δ87 line 2, and a 37.9-fold increase for pms1-FR AF line 3 (Figure 3.4C).

Figure 3.4 (next page): Effect of mutant pms1 alleles on microsatellites instability.

(A) Microsatellite instability in F1 progeny of crosses between the G16 - B reporter line and eight pms1 transformants (allele and line shown on the left) or wild - type Columbia (Control). The average number of blue sectors per leaf (number to the right) and a 95% confidence interval around this mean (stick) are shown for each set of F1 progeny. Results differing significantly (P≤0.05) from that seen in the Control are indicated with an asterisk .(B) Microsatellite instability in F1 progeny of crosses between the G16 - C reporter line and four transformants or wild-type Columbia. (C) Microsatellite instability in F1 progeny of crosses between the G7 - A reporter line and two transformants or wild-type Columbia.

3.5. Discussion

Mismatch repair genes constitute a highly conserved family of genes involved in maintaining genome integrity. Mutations that impede on the function of members of this family have been linked to greater instability of short repeat tracts (Sia et al. , 1997; Tran et al. , 1997). We have analyzed the effect of selective mutagenesis on the Arabidopsis AtPMS1 gene, a key member of the MMR family of genes. We specifically targeted for mutation a region of the AtPMS1 protein presumed to be involved in MMR activity. To create one pms1 altered allele, we removed a stretch of 87 amino acids spanning the entire MutL-box. To generate the other pms1 altered allele, we changed the amino acids F108 into A108, and R109 into F109 within the highly conserved MutL-box (GFRGEAL). We selectively made the change of FR→AF in the MutL-box because in E. coli , the change of R95 to F95 in this GFRGEAL motif considerably reduced the MutL affinity for ATP binding (Ban et al. , 1999). In S. cerevisiae , changing of F126 into A126 in the GFRGEAL motif of PMS1 increased the mutation rate 639 times relative to the wild type. Thus, on the basis of these results, we assumed that by combining the two specific mutations within the Arabidopsis AtPMS1 gene, or by deleting a large region around the MutL-box, we might impair the function of the MutLα dimers. Indeed, a functional MutLα complex requires not only protein-protein interaction between AtPMS1 and AtMLH1 through their C-terminal portions, but also the involvement of their N-terminal portion in the MMR activity. By over-expressing the altered forms of the PMS1 gene, we postulated that most if not all of the C-terminal binding sites of the AtMLH1 protein would be tied up by the altered PMS1 protein.

We produced ten independent transformants that displayed a morphology, which, upon visual examination, was not different overall from that of the wild type. As the introduction of these mutant alleles did not markedly affect the appearance of the plants, we proceeded to analyze the effect of the mutant alleles on microsatellite instability (MSI), a phenotype typically indicative of a defect in MMR function in all eukaryotes. We assessed this phenotype using a microsatellite instability reporter analogous to the one used by Leonard et al . (2003). We used synthetic microsatellites of two different lengths introduced into the GUS gene, to assess the effect of the inactivation of the AtPMS1 gene. Our data indicated that reversion events were significantly more frequent in the F1 progenies of some of the pms1 transformants compared to control F1. More than half of the F1 progenies containing the altered forms of the PMS1 gene showed a significant increase in MSI when the G16-B reporter line was used. Among the subset of these pms1 transformants that were crossed with the G16-C reporter, all four lines examined displayed a significant increase in MSI. Furthermore, maybe because the G16 microsatellite was much more stable in the G16-C reporter line, a greater increase in MSI was observed than with the G16-B reporter lines. The MSI observed in the F1 progenies of the lines that were crossed to a G7 reporter line was even higher.

This suggests that lines expressing either the pms1-Δ87 allele or the pms1-FR AF allele have indeed been impaired in their ability to carry out MMR. It appears as if those pms1 transformants showing a significant increase in MSI were also those whose transcript abundance seemed to be greatest (Figure 3.2). This might suggest that, in some cases, the transgene may not have been expressed sufficiently to allow the accumulation of enough modified AtPMS1 protein to perturb MMR activity. An alternative hypothesis is that a presumed increase in the abundance of AtPMS1 protein is itself the cause of the increased MSI and that such an effect would have been seen with a wild-type allele of AtPMS1 . Although formally possible, this hypothesis runs counter to the data from Saccharomyces cerevisiae . Shcherbakova et al . (2001) reported that while the over-expression of a wild-type MLH1 gene led to a 200-fold increase in the mutation rate (as measured using the his7-2 allele), the over-expression of PMS1 only led to a 8-fold increase (4% of the increase seen with MLH1 over-expression). Using another reporter allele ( lys2::InsEA14 ), overproduction of the PMS1 protein resulted in only a very modest increase in the mutation rate, corresponding to less than 2% of the increase seen when overproducing MLH1.

Overall, the increases observed in MSI using a given reporter line (G16-B, G16-C or G7-A) are relatively similar. The sole exception to this is presented by the data from the pms1-∆87(2) x G16-C cross. Indeed, whereas the other three F1 populations show an average number of sectors per leaf ranging between 3.3 and 6.1, the F1 of pms1-∆87(2) x G16-C shows 24.6 sectors per leaf. If the exact same transformant (G16-C) was used, it cannot be argued that position effects are responsible for this difference. One explanation could be that a mix-up occurred and that these F1 were derived actually from a cross with the other reporter used in this work, G16-B. As shown in Figure 3.4A, the 95% confidence interval for the mean number of sectors per leaf in such a cross ranges between 16.8 and 45.0. The observed frequency of sectors reported in Figure 3.4B (24.6) falls well within this range. As both reporter lines contain the exact same reporter construct, we have no simple way of testing this hypothesis.

Although an increase in MSI was expected for plants impaired in MMR, the magnitude of the observed effect differed from one reporter to the next. In prior work in yeast, it had been shown that longer repetitive tracts are more highly destabilized in MMR mutants than shorter ones (Tran et al ., 1997; Umar et al ., 1998). Here, on the contrary, we observed a greater increase in MSI using a G7 reporter than G16 reporters. As this observation derives from only two crosses, however, we feel it would be premature to suggest such a trend in plants and more work will need to be done to determine if this trend is real. More importantly, we found that the increase in MSI was much lower than that found in other organisms (Tran et al. , 1997; Umar et al. , 1998). Indeed, microsatellite instability in the plants carrying the mutant pms1 alleles was on average six times higher than normal with the G16 reporter and 28 times higher for the G7 reporter. In yeast, the loss of PMS1 function results in a 1,500- to 3,100-fold increase in the instability of a tract of 18 Gs (Sia et al. , 1997).

It may be that Arabidopsis (and possibly all plants) differs in some important behavior with respect to DNA damage and repair as related to the MMR system. Indeed, very recently, Leonard et al . (2003) have reported in Arabidopsis the phenotype of an insertional T-DNA mutation in the AtMSH2 gene, another member of the MMR family of gene. On average, only a modest (5 - fold) increase was seen in the instability of a GUS reporter gene containing a synthetic G7 microsatellite. The magnitude of this effect is very similar to the one observed in our pms1 transformants, but again much smaller than that seen in other eukaryotes. This suggests that, in Arabidopsis, the complete loss of function of a key MMR gene in somatic tissues results in a less dramatic increase in MSI than in other eukaryotes.

It has been suggested that the relatively modest impact of MMR defects on MSI in Arabidopsis might be due to the fact that, in differentiated tissues, the fidelity of DNA replication might not be a critical factor (Leonard et al. , 2003). The basis for this hypothesis is that MMR genes are expressed at extremely low levels in differentiated tissues, often below the detection threshold by northern analysis (Ade et al. , 1999). As differentiated tissues do not contribute to the germ line and because DNA damage does not lead to a disease like cancer in plant cells, maybe MMR activity is already quite low in these tissues. Therefore, the loss of such a minor activity might not result in a strong phenotype. In support of this hypothesis, it has already been observed that in plants, the spontaneous mutation rate as measured using a reporter gene in somatic tissues, is 1,000 to 1,500 times higher than the spontaneous mutation rate found in bacteria, yeast and mouse (Kovalchuk et al. , 2000).

To further advance our understanding of the implication the AtPMS1 gene in genome stability, we are currently characterizing a pms1 T - DNA insertional mutant and RNAi plants in which PMS1 inactivation is being sought through gene silencing. The use of the same reporters to test MSI in these mutants should prove extremely informative to confirm and expand our knowledge of the function of the AtPMS1 gene.

Acknowledgements

A. H. Alou was supported by a graduate fellowship from the «Programme Canadien de Bourses de la Francophonie». This work was also supported by a research grant from the Natural Sciences and Engineering Research Council of Canada and RhoBio to F. J. Belzile.

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: 239-249

Aronshtam A, Marinus MG (1996) Dominant negative mutator mutations in the MutL gene of Escherichia coli . Nucleic Acids Res 24: 2498-2504

Baker SM, Bronner CE, Zhang L, Plug AW, Robatzek M, Warren G, Elliott EA, Yu J, Ashley T, Arnheim N (1995) Male mice defective in the DNA mismatch repair gene PMS2 exhibit abnormal chromosome synapsis in meiosis. Cell 82: 309-319

Ban C, Junop M, Yang W (1999) Transformation of MutL by ATP binding and hydrolysis: a switch in DNA mismatch repair. Cell 97: 85-97

Ban C, Yang W (1998a) Crystal structure and ATPase activity of MutL: implications for DNA repair and mutagenesis. Cell 95: 541-552

Ban C, Yang W (1998b) Structural basis for MutH activation in E. coli mismatch repair and relationship of MutH to restriction endonucleases. EMBO J 17: 1526-1534

Bent AF (2000) Arabidopsis in planta transformation. Uses, mechanisms, and prospects for transformation of other species. Plant Physiol 124: 1540-1547

Boland CR, Thibodeau SN, Hamilton SR, Sidransky D, Eshleman JR, Burt RW, Meltzer SJ, Rodriguez-Bigas MA, Fodde R, Ranzani GN, Srivastava S (1998) A National Cancer Institute Workshop on Microsatellite Instability for cancer detection and familial predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer. Cancer Res 58: 5248-5257

Chambers SR, Hunter N, Louis EJ, Borts RH (1996) The mismatch repair system reduces meiotic homeologous recombination and stimulates recombination-dependent chromosome loss. Mol Cell Biol 16: 6110-6120

Church GM, Gilbert W (1984) Genomic sequencing. Proc Natl Acad Sci U S A 81: 1991-1995

Dao V, Modrich P (1998) Mismatch-, MutS-, MutL-, and helicase II-dependent unwinding from the single-strand break of an incised heteroduplex. J Biol Chem 273: 9202-9207

Dubois P, Cutler S, Belzile FJ (1998) Regional insertional mutagenesis on chromosome III of Arabidopsis thaliana using the maize Ac element. Plant J 13: 141-151

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

Flores-Rozas H, Kolodner RD (1998) The Saccharomyces cerevisiae MLH3 gene functions in MSH3-dependent suppression of frameshift mutations. Proc Natl Acad Sci U S A 95: 12404-12409

Galio L, Bouquet C, Brooks P (1999) ATP hydrolysis-dependent formation of a dynamic ternary nucleoprotein complex with MutS and MutL. Nucleic Acids Res 27: 2325-2331

Hall M, Jordan J, Matson S (1998) Evidence for a physical interaction between the Escherichia coli methyl-directed mismatch repair proteins MutL and UvrD. EMBO J 17: 1535 - 1541

Hall MC, Matson SW (1999) The Escherichia coli MutL protein physically interacts with MutH and stimulates the MutH-associated endonuclease activity. J Biol Chem 274: 1306-1312

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

Jiricny J (1998) Eukaryotic mismatch repair: an update. Mutat Res 409: 107-121

Jiricny J (2000) Mediating mismatch repair. Nat Genet 24: 6-8

Kovalchuk I, Kovalchuk O, Hohn B (2000) Genome-wide variation of the somatic mutation frequency in transgenic plants. EMBO J 19: 4431-4438

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: 328-338

Lipkin SM, Wang V, Jacoby R, Banerjee-Basu S, Baxevanis AD, Lynch HT, Elliott RM, Collins FS (2000) MLH3: a DNA mismatch repair gene associated with mammalian microsatellite instability. Nat Genet 24: 27-35

Lipkin SM, Wang V, Stoler DL, Anderson GR, Kirsch I, Hadley D, Lynch HT, Collins FS (2001) Germline and somatic mutation analyses in the DNA mismatch repair gene MLH3: Evidence for somatic mutation in colorectal cancers. Hum Mutat 17: 389-396

Modrich P (1991) Mechanisms and biological effects of mismatch repair. Annu Rev Genet 25: 229-253

Peltomaki P, de la Chapelle A (1997) Mutations predisposing to hereditary nonpolyposis colorectal cancer. Adv Cancer Res 71: 93-119

Prolla T, Baker S, Harris A, Tsao E, Yao X, Bronner C, Zheng B, Gordon M, Reneker J, Arnheim N, Shibata D, Bradley A, Liskay R (1998) Tumor susceptibility and spontaneous mutation in mice deficient in Mlh1, Pms1 and Pms2 DNA mismatch repair. Nat Genet 18: 276 - 279

Sambrook J, Russell DW (2001) Molecular cloning: a laboratory manual, Ed 3rd. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.

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

Sia EA, Dominska M, Stefanovic L, Petes TD (2001) Isolation and characterization of point mutations in mismatch repair genes that destabilize microsatellites in yeast. Mol Cell Biol 21: 8157-8167

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: 2851-2858

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: 2859-2865

Umar A, Risinger J, Glaab W, Tindall K, Barrett J, Kunkel T (1998) Functional overlap in mismatch repair by human MSH3 and MSH6. Genetics 148: 1637-1646

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: 5536-5540.

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

Wu TH, Marinus MG (1999) Deletion mutation analysis of the MutS gene in Escherichia coli . J Biol Chem 274: 5948-5952

Yamaguchi M, Dao V, Modrich P (1998) MutS and MutL activate DNA helicase II in a mismatch-dependent manner. J Biol Chem 273: 9197-9201

Yao X, Buermeyer AB, Narayanan L, Tran D, Baker SM, Prolla TA, Glazer PM, Liskay RM, Arnheim N (1999) Different mutator phenotypes in Mlh1- versus Pms2-deficient mice. Proc Natl Acad Sci U S A 96 : 6850-6855