| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Genetics, Vol. 169, 917-929, February 2005, Copyright © 2005
doi:10.1534/genetics.104.035089
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,3,
,
,**,4
* Interdepartmental Genetics Program, Cellular and Developmental Biology Program
Molecular, Cellular and Developmental Biology Program
Department of Biochemistry, Biophysics and Molecular Biology
Department of Agronomy, Iowa State University, Ames, Iowa 50011
** Center for Plant Genomics, Iowa State University, Ames, Iowa 50011
4 Corresponding author: 2035B Roy J. Carver Co-Laboratory, Iowa State University, Ames, IA 50011.
E-mail: schnable{at}iastate.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Recombination events are of two types: reciprocal crossovers (COs) or nonreciprocal noncrossovers (NCOs), such as gene conversions. Detailed analyses of gene conversion events in Drosophila and fungi have established that: (1) conversion tract lengths are relatively short and continuous [e.g., they average 
350 bp in Drosophila (HILLIKER et al. 1994)] and (2) gene conversions exhibit a phenomenon termed polarity [e.g., DNA sequences near one end of the rosy locus of Drosophila and the ARG4 and HIS4 loci of yeast (SCHULTES and SZOSTAK 1990; DETLOFF et al. 1992) exhibit higher rates of gene conversion than do sequences at the other end of these loci]. In most cases, higher frequencies occur at the 5' ends of these genes.
In most plants, gene conversion and double-crossover events cannot be distinguished. Even so, in the absence of strong negative interference, intragenic double crossovers are expected to occur only very rarely. Although putative gene conversions were reported in maize as early as 1986 (DOONER 1986), only a few conversion tracts have been molecularly characterized (XU et al. 1995; DOONER and MARTINEZ-FEREZ 1997b; MATHERN and HAKE 1997; LI et al. 2001; YAO et al. 2002).
Several models have been proposed to explain the mechanism responsible for meiotic recombination (HOLLIDAY 1964; RESNICK 1976; SZOSTAK et al. 1983). The most widely accepted of these are based upon the double-strand break (DSB) repair model (SZOSTAK et al. 1983; SUN et al. 1991) in which differential resolution of double-Holliday junction (DHJ) intermediates results in COs or NCOs. Evidence in support of this model has been obtained from the yeast Saccharomyces cerevisiae and other simple model organisms. The meiosis-induced DSBs predicted by this model to initiate meiotic recombination have been observed at a number of loci (SUN et al. 1989, 1991; GAME 1992; ZENVIRTH et al. 1992; DE MASSY and NICOLAS 1993; FAN et al. 1995; LIU et al. 1995). Moreover, several studies have demonstrated the existence of recombination intermediates of the types postulated in the model, i.e., joint molecules (COLLINS and NEWLON 1994; SCHWACHA and KLECKNER 1994; SCHWACHA and KLECKNER 1995) and heteroduplex DNA (WHITE et al. 1985; LICHTEN et al. 1990; GOYON and LICHTEN 1993; NAG and PETES 1993). More recently, ALLERS and LICHTEN (2001) proposed a modified DSB repair model. In this model, COs and NCOs are similarly initiated by DSBs but following resection and single end invasion (SEI) only some result in DHJs that can resolve as COs; the remainder are processed via the synthesis-dependent strand annealing pathway and resolve as NCOs. The identification of SEI intermediates in meiosis provides physical evidence for this modified model (HUNTER and KLECKNER 2001). This model is further supported by the identification of meiotic-related mutants that disrupt SEI formation and drastically reduce the production of COs but not NCOs (BORNER et al. 2004).
It is thought that meiotic recombination in plants shares at least some mechanistic aspects with yeast (XU et al. 1995; PUCHTA et al. 1996; PUCHTA and HOHN 1996). Due to technical barriers, less is known about the molecular nature of meiotic recombination in plants than in simple model organisms. Meiotic recombination hotspots in S. cerevisiae correspond to nearby DSB hotspots (reviewed in LICHTEN and GOLDMAN 1995) that have been recently mapped throughout the genome (GERTON et al. 2000). Available techniques have thus far limited our ability to map DSBs in plants. Therefore, the possible association between DNA breaks and meiotic recombination hotspots in plants has not yet been established. Instead, meiotic recombination hot- and coldspots are identified by physically mapping recombination breakpoints.
Although DSBs have not yet been mapped in plants, there is evidence that DNA breaks play a role in meiotic recombination. For example, processes that enhance the rate of DSB formation stimulate recombination in plants. This conclusion is based on three findings. First, agents that can physically (e.g., X rays and UV irradiation) or chemically (e.g., methylmethanesulfonate and mitomycin C) induce DSBs are able to stimulate intrachromosomal recombination (reviewed in PUCHTA and HOHN 1996). Second, the expression of the site-specific endonuclease I-SceI in tobacco protoplasts (PUCHTA et al. 1993, 1996) and HO in somatic cells of Arabidopsis (CHIURAZZI et al. 1996) increases the rates of mitotic recombination. Third, autonomous transposons have the ability to increase the rates of recombination-like losses of duplicated regions surrounding corresponding nonautonomous transposons in Arabidopsis and maize (ATHMA and PETERSON 1991; LOWE et al. 1992; STINARD et al. 1993; XIAO et al. 2000; XIAO and PETERSON 2000). Because these events either occurred in the absence of meiosis (ATHMA and PETERSON 1991; XIAO et al. 2000; XIAO and PETERSON 2000) or did not involve the exchange of flanking markers (LOWE et al. 1992) and therefore did not involve meiotic crossovers, they do not address the question as to whether DSBs stimulate meiotic recombination. Indeed, DOONER and MARTINEZ-FEREZ (1997a) have reported that meiotic recombination at the bz1 locus in maize is not stimulated by the germinal excisions of the Ac transposon, which would be expected to introduce DSBs within bz1. The relationship between transposon excision and the stimulation of repair by recombination has not yet been elucidated in other transposon systems in maize.
The a1 locus of maize is an excellent system for the study of meiotic recombination because: (1) intragenic recombination events can be easily identified by their visible nonparental phenotypes (i.e., colored vs. colorless kernels); (2) transposon-tagged a1 alleles have been cloned and characterized, e.g., a1-mum2 and a1::rdt alleles (O'REILLY et al. 1985; BROWN et al. 1989a), and a substantial degree of DNA sequence polymorphism exists between these alleles (XU et al. 1995) thereby facilitating the high-resolution mapping of recombination breakpoints; and (3) genetic markers flanking the a1 locus are available for distinguishing between NCO and CO events.
The a1 locus was used to test the effects of the trans-acting regulatory transposon MuDR (SCHNABLE and PETERSON 1986; CHOMET et al. 1991; HERSHBERGER et al. 1991; QIN et al. 1991; HSIA and SCHNABLE 1996) on the frequency and distribution of intragenic meiotic recombination events in the vicinity of the Mu1 nonautonomous transposon insertion. Rates of meiotic COs in the vicinity of a Mu1 insertion increase in the presence of MuDR, thereby demonstrating that MuDR is active during meiosis. We hypothesize that this stimulation of meiotic COs occurs via the introduction of DNA breaks generated by MuDR at the Mu1 insertion and that these DNA breaks stimulate meiotic COs in plants.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
0.1 cM centromere distal from the a1 locus (CIVARDI et al. 1994). Mutations at this locus condition a shrunken kernel phenotype. Two stocks were derived from a maize line obtained from D. S. Robertson that carried a1-mum2 and many genetically active copies of MuDR. Sibling spotted and nonspotted kernels derived from the same ears were used as the a1-mum2 with and without MuDR stocks, respectively. Consequently, these stocks differ by only the presence or absence of MuDR. The a1-dl stock is as described by XU et al. (1995).
Isolation of genetic recombinants:
Crosses 1 and 2 were conducted by planting the indicated parents in plots isolated from other maize pollen sources during the summer of 1994.
The female parents (listed first) were detasseled prior to anthesis to ensure that they would be pollinated by only the a1::rdt sh2 male parent. As expected, most of the progeny from these crosses had colorless, round (or spotted round when MuDR was present; class I, Figure 1) or colorless, shrunken (class II, Figure 1) kernel phenotypes. However, if intragenic recombination occurred in a1, four recombinant classes (Figure 1, classes III–VI) could also result; three of these can be identified via their nonparental phenotypes. Class IV recombinants condition a parental phenotype and therefore could not be identified. NCOs initiated from the a1-mum2 Sh2 chromosome (class VI) would condition colored round kernels. In the current study, this class could not be analyzed due to the difficulty in distinguishing colored kernels from the very heavily spotted kernels that contain MuDR. The two remaining recombinant classes (III and V) produce colored, shrunken kernels. These kernels were putative recombinants arising from either COs between the Mu1 and rdt transposon insertion sites in the a1-mum2 and a1::rdt alleles (class III, Figure 1) or gene conversion events in which the rdt transposon sequence and its flanking sequences were replaced by the corresponding sequences from the a1-mum2 allele (class V, Figure 1). In either case, the genotype of these kernels was designated A1' sh2/a1::rdt sh2. To verify and purify the putative recombinant alleles, colored shrunken kernels from crosses 1 and 2 were planted and subjected to cross 3. Colored, round kernels from cross 3 were then planted and the resulting plants were self-pollinated (cross 4).
|
Generating plasmid clones for sequencing the 3' regions of a1-mum2 and a1::rdt:
Within the 1.2-kb interval defined by the insertion sites of Mu1 and rdt, 20 DNA sequence polymorphisms exist between the a1-mum2 (GenBank accession no. AF347696) and a1::rdt alleles (GenBank accession no. AF072704). To sequence the region proximal to the rdt insertion site, the 10-kb a1::rdt clone pE10 (XU et al. 1995) was digested with SalI and a 2.6-kb fragment that contains the 3' region of the a1::rdt allele was subcloned into pBKS (Stratagene, La Jolla, CA) to create pSL2.6. Clone pSL2.6 was subsequently digested with SacI or KpnI to generate two overlapping subclones: 1.5-kb pSC1.5 and 1.4-kb pKN1.4, respectively. Clones pSL2.6, pSC1.5, and pKN1.4 were used as templates for sequencing the proximal region of the a1::rdt allele. A 1.0-kb fragment resulting from the SacI digestion of the 3.0-kb a1-mum2 subclone pYEN1 (XU et al. 1995) was subcloned into pBKS to generate pSC1.0. Clones pYEN1 and pSC1.0 were used as templates for sequencing the proximal region of the a1-mum2 allele.
Oligonucleotides:
Because the genic sequence of the a1-mum2 allele (AF347696) is identical to the A1-LC allele (YAO et al. 2002), oligonucleotides used as primers for PCR and sequencing were designed from the existing sequence of the A1-LC allele (GenBank accession nos. X05068, AF363390, and AF363391). With the exception of the a1-mum2-specific primer QZ1543, these oligonucleotides could also amplify the a1::rdt allele due to the high degree of sequence identity (98%) between a1-mum2 and a1::rdt.
The sequences of the oligonucleotides used as primers for PCR and sequencing were as follows: QZ1543, 5' AAA CAT AAA AAC AAT ACG TAA TCC AG 3' (a1-mum2-specific primer); XX907, 5' GTG TCT AAA ACC CTG GCG CA 3'; QZ1003, 5' ATA ATA GTA GCC TCC CGA ATA A 3'; XX231, 5' GCC AAA CTC TGA TTC GCT CCG TG 3'; XX390, 5' TCG GCT TGA TTA CCT CAT TCT 3'; XX025, 5' GGT AGG GCA GCG TGT GGT GTT 3' (XU et al. 1995); and XX026, 5' GAG GTC GTC GAG GTG GAT GAG CTG 3' (XU et al. 1995). The positions of the primers within the a1 gene are illustrated in Figure 2A.
|
Mapping the breakpoints of COs and the distal (5') endpoints of conversion tracts relative to a diagnostic PstI site:
On the basis of the strategy used to select recombinants, the breakpoints of all COs and the distal endpoints of all conversion tracts of NCOs recovered in this experiment were expected to fall within the 1.2-kb interval defined by the Mu1 and rdt insertion sites in each of the parental a1 alleles (Figure 2A). Previously, XU et al. (1995) identified a diagnostic PstI site within this 1.2-kb interval that can distinguish between a1-mum2 and a1::rdt derived sequences. This site is present in the a1::rdt allele, but absent from the a1-mum2 allele. Thus, PstI digestion of PCR-amplified recombinant alleles can be used to map the position of the breakpoint of each CO or the distal endpoint of each gene conversion tract relative to this site.
Primers (XX025 and XX026) were used to PCR amplify this 1.2-kb interval from each recombinant allele (Figure 2A). The 1.2-kb PCR products were fractionated by electrophoresis, purified by binding to NA45 DEAE membrane (Schleicher & Schuell, Keene, NH), and subjected to PstI digestion. If a given PCR product was digested by PstI, the breakpoint of the CO or the distal endpoint of the conversion tract associated with the corresponding A1' allele did not extend 5' of this diagnostic PstI site. Alternatively, if the PCR product was resistant to PstI digestion, the breakpoint of the CO or the distal endpoint of the conversion tract was between the diagnostic PstI site and the Mu1 insertion site. Thirty-six of the 40 A1' alleles were successfully amplified and mapped relative to the diagnostic PstI site.
PCR-based sequencing of the recombinant A1' alleles:
Once a CO breakpoint or conversion tract endpoint was mapped relative to the diagnostic PstI site, the region of the corresponding A1' allele that contained the recombination endpoint was PCR amplified, purified, and sequenced. Purified PCR products amplified using primers XX025 and XX026 from each of the 36 A1' alleles were used as templates for sequencing. For those 30 alleles that had breakpoints or endpoints distal to the polymorphic PstI site, primers XX390 and XX025 were used for sequencing. For those 6 alleles with breakpoints or endpoints proximal to the polymorphic PstI site, primers XX231 and XX026 were used for sequencing (Figure 2A). The positions of each CO breakpoint and the conversion tract endpoints proximal or distal to the polymorphic PstI site were identified by comparing the DNA sequence polymorphisms present in a given recombinant A1' allele to those in a1-mum2 and a1::rdt.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
To test the effect of the trans-acting regulatory element MuDR on CO frequency, colored shrunken kernels (Figure 1, classes III and V) were selected from parallel crosses without (cross 1) and with (cross 2) MuDR. The colored round class (class VI) was not analyzed for two reasons. First, a previous Mu transposition study in a1 (LISCH et al. 1995) demonstrated that Mu1 excision and repair from a1-mum2 generates fully restored alleles at a low rate, <10–4. Second, coupled with the extremely low rate of reversion at a1-mum2, it is very difficult to distinguish the colored round class (class VI) from the very heavily spotted parental phenotype obtained from cross 2.
Recombinant A1' alleles isolated from crosses 1 and 2 that were in coupling with the closely linked sh2 mutant allele could have arisen via either COs that resolved between the Mu1 and rdt insertion sites or NCOs having conversion tracts that span the rdt transposon insertion site. These putative colored shrunken recombinants were analyzed as described in MATERIALS AND METHODS. The validity of 8 cross 1 and 32 cross 2 recombinants were confirmed via testcrosses and/or RFLP analysis (Table 1) using marker php10080 as described by XU et al. (1995).
|
|
The rates of recombination at the a1 locus in the presence or absence of MuDR are shown in Table 2. The genetic distances (CO + NCO) associated with the 1.2-kb interval derived from cross 1 (without MuDR) and cross 2 (with MuDR) are significantly different (0.008 vs. 0.02 cM). Although the rate of class V NCOs was unaffected by the presence of MuDR, the rate of CO was four times higher in the presence of MuDR than in its absence (0.02 vs. 0.005 cM). On the basis of the DSB repair model, NCOs analyzed in this study (class V) must have been initiated by DNA breaks on the a1::rdt-containing homolog. Consistent with MuDR's ability to interact with Mu1 but not rdt, the rate of the class V NCOs was unaffected by the presence (cross 2) or absence (cross 1) of MuDR. Class VI events (NCOs initiated from a1-mum2; Figure 1) would be colored and round. Because such kernels can be extremely difficult to distinguish from heavily spotted a1-mum2 kernels, and no germinal revertants of a1-mum2 were identified in a previous screen of 10,000 kernels (LISCH et al. 1995), the rates at which class VI events occur were not initially determined in this study. We subsequently attempted to isolate class VI events from the a1-mum2 source used in this study to determine the frequency at which MuDR-induced DNA breaks are repaired via conversion. Round kernels that appeared to be fully colored and that carried a1-s were selected from the progeny of a1-mum2 Sh2/a1 sh2 plants that carried MuDR and that had been crossed by an a1-s pollen source (a cross similar to cross 2). Only one excision event was confirmed from a population of 46,000 spotted kernels (data not shown). Therefore, the rate of reversion of a1-mum2 to A1' (class VI events), i.e., 10–5, did not differ significantly from the rates of NCO from the a1::rdt chromosome (class V events) in the presence or absence of MuDR (Table 2). Hence, the increased rate of recombination that occurs in the presence of MuDR is due to an increased rate of COs.
Physical mapping of recombination breakpoints of COs and the 5' (distal) endpoints of conversion tracts:
The conversion endpoints of 8 NCO events and the recombination breakpoints of 28 CO events were physically mapped. Because recombinants were selected on the basis of their colored, shrunken phenotypes, the breakpoints of all the COs and the distal endpoints of the conversion tracts must have resolved within the 1.2-kb interval defined by the Mu1 and rdt transposon insertion sites (Figure 2).
Digestion with PstI revealed that the distal endpoints of six of the eight conversion tracts map 5' of the diagnostic PstI site that is polymorphic between a1-mum2 and a1::rdt. By virtue of the selection scheme used in the experiment, the conversion tracts cannot contain Mu1. Hence, the distal endpoints of these six conversion tracts must lie between the PstI site and the Mu1 insertion site. The 1.2-kb interval between the Mu1 and rdt insertion sites exhibits 20 polymorphisms between the a1-mum2 (GenBank accession no. AF347696) and a1::rdt alleles (GenBank accession no. AF072704). Regions containing the CO breakpoints or conversion tract endpoints associated with each of 36 A1' alleles were PCR amplified and the purified PCR products were sequenced. The sequence derived from each recombinant A1' allele was then compared to the sequences of the a1-mum2 and a1::rdt alleles. The switchpoint of sequence polymorphisms within each recombinant allele established, at the highest possible resolution, the position of each CO breakpoint or NCO conversion tract endpoint. The distributions of the CO breakpoints and the distal endpoints of the NCO conversion tracts are illustrated in Figure 2B. Whereas recombination hotspots in yeast are defined by regions of high DSB frequency, recombination hotspots in this and other plant studies are defined as regions with elevated rates of recombination resolution endpoints. The distal endpoints of 6 of 8 NCO events and the CO breakpoints of 21 of 28 crossover events mapped to the previously defined 377-bp recombination hotspot at the 5' end of the a1 coding sequence (XU et al. 1995).
Physical mapping of the 3' (proximal) endpoints of conversion tracts:
On the basis of the genetic screen used to isolate recombination events, we believe that the proximal endpoints for all of the conversion tracts must have resolved proximal to the rdt insertion site. Only two DNA sequence polymorphisms exist between the a1-mum2 and a1::rdt alleles in the 1.6 kb proximal to the rdt insertion site (Figure 2A). One of these polymorphisms is a 32-bp insertion/deletion that is present in the a1-mum2 allele, but absent from the a1::rdt allele. By using an a1-mum2-specific primer (QZ1543) that anneals to the 32-bp polymorphic sequence in combination with a primer (XX907) that amplifies both a1-mum2 and a1::rdt alleles, it was possible to map the proximal endpoints of the conversion tracts relative to this 32-bp polymorphic site (Figure 2C). The ability of these primers to PCR amplify a given A1' allele indicated that the conversion tract contained the 32-bp polymorphic sequence. Such a result would indicate that the proximal conversion tract endpoint was proximal to the 32-bp polymorphic site. The proximal conversion tract endpoints of four A1' alleles derived from NCOs mapped proximal to the 32-bp polymorphic site. The 1059 bp proximal to this polymorphism are identical between the two a1 alleles. No a1-specific RFLPs were detected between a1-mum2 and a1::rdt that mapped between the 32-bp polymorphism and php10080 when the 1.0-kb SacI/EcoRI fragment from the a1-mum2 subclone pSC1.0 was used as a hybridization probe in a DNA gel blot experiment involving genomic DNA (data not shown). As a result, the proximal endpoints of these four conversion tracts could not be determined with higher precision.
A negative PCR result using primers QZ1543 and XX907 demonstrated that the proximal endpoint of an NCO was distal to the 32-bp polymorphic site. The proximal endpoints of four NCO A1' alleles mapped distal to the 32-bp polymorphism. To map these proximal endpoints to higher resolution, the corresponding A1' alleles were PCR amplified using primers XX907 and QZ1003 (Figure 2A). By comparing the sequences of the resulting 0.9-kb PCR product to the a1 parental alleles used to generate the A1' alleles, the proximal endpoints were mapped at the highest possible resolution afforded by the sequence polymorphisms present between the parental alleles. The proximal endpoints of all four of these conversion tracts mapped to interval XXIII (Figure 2C). Although three of the conversion tracts are indistinguishable, they must have arisen via independent events because they were recovered from separate female parents in crosses 1 and 2.
Lengths of conversion tracts:
Because the positions of both the distal and the proximal endpoints of four conversion events were established, it was possible to calculate the sizes of the conversion tracts associated with each of the resulting A1' alleles. As depicted in Figure 2C, three of the four conversion tracts were between 621 and 1318 bp in length. The fourth is between 31 and 683 bp. For the remaining four conversion events, although their distal endpoints were mapped at the highest possible precision, it was not possible to map their proximal endpoints because of a lack of polymorphisms between the parental a1 alleles (a1-mum2 and a1::rdt). Therefore, the absolute sizes of these four conversion tracts could not be determined. However, on the basis of the positions of the corresponding distal endpoints, the lengths of these conversion tracts must be in excess of 534 bp (1 allele), 1124 bp (1 allele), 1320 bp (1 allele), and 1603 bp (1 allele).
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
1400 bp (GLOOR et al. 1991; PRESTON and ENGELS 1996). In yeast, the average meiotic conversion tract lengths range from 0.4 to 1.6 kb in a 9-kb interval (BORTS and HABER 1989). Even so, longer conversion tracts of 9 and 12 kb have also been observed in yeast (BORTS et al. 2000). Conversion tracts >5 kb have also been observed in Neurospora (YEADON and CATCHESIDE 1998). Few plant conversion tracts have been characterized. In maize, two a1 conversion tracts were in excess of 621 and 815 bp (XU et al. 1995) and two bz1 conversion tracts were between 965 and 1165 bp and between 1.1 and 1.5 kb (DOONER and MARTINEZ-FEREZ 1997b). The conversion tracts of two NCO-like events isolated from the maize Kn1-O tandem duplication (MATHERN and HAKE 1997) are 1.7 and 3 kb. YAO et al. (2002) identified two putative NCOs from the maize a1-sh2 interval that have conversion tracts that are at least 17 kb. The eight conversion tracts characterized in this study (crosses 1 and 2) ranged in size from >31 bp to >1603 bp. These NCO events resulted from conversion of the rdt insertion and its surrounding sequences and were therefore initiated by DSBs on the a1::rdt-containing chromosome. Conversion tracts on the a1::rdt-containing chromosome that extended 5' of the Mu1 insertion site would not have been recovered.
The single revertant allele isolated from a1-mum2 in this study exhibited 100% identity to the wild-type progenitor of a1-mum2 (i.e., A1-LC). This revertant may therefore have arisen via a class VI NCO event (Figure 1) within the interval 51 bp upstream and 165 bp downstream of the Mu1 insertion site that lacks polymorphisms between the parental alleles of crosses 1 and 2. This would be consistent with the finding that in mice and yeast, gene conversion tracts can be <100 bp (SWEETSER et al. 1994; ELLIOTT et al. 1998; PALMER et al. 2003).
Factors affecting CO/NCO ratios:
In yeast heterozygotes, the presence of only a few nucleotide polymorphisms can drastically affect the frequencies at which COs (BORTS and HABER 1987; BORTS et al. 1990) and NCOs (CHEN and JINKS-ROBERTSON 1999; NICKOLOFF et al. 1999) are recovered. For example, 0.09% heterology between alleles at the MAT loci in yeast reduced COs by twofold and increased the rate of NCOs by threefold (BORTS et al. 1990). Similarly, in the maize bz1 gene, the degree of sequence similarity between the parental alleles affects the CO/NCO ratio (DOONER 2002). In Dooner's study the CO/NCO ratio observed between bz1 "heteroalleles" was >20. In the current study, the CO/NCO ratio observed in plants that lacked MuDR was 1.8. This dramatic difference in the CO/NCO ratio between the two studies cannot be attributed to differences in the degree of DNA sequence polymorphism, because Dooner's heteroalleles exhibited approximately the same degree of DNA sequence polymorphism (1.5%) as the parental a1 alleles used in the current study (1.8%). Instead, the 10-fold difference in the CO/NCO ratios between the two studies could be a consequence of locus-specific differences in the relative rates of conversions and crossovers or differences caused by genetic background. Although it is not possible to exclude locus-specific effects, this study does establish that CO/NCO ratios can be influenced by the differences in genetic background. Specifically, this study establishes that MuDR affects CO/NCO ratios. For example, the CO/NCO ratios differ by more than threefold between crosses 1 and 2 (compare CO/NCO ratios, Table 2) for which the genetic backgrounds are identical except for the absence (cross 1) or presence (cross 2) of MuDR. We cannot exclude the possibility that the relative impacts of MuDR on rates of COs and NCOs may differ depending upon the level of sequence heterology present in a heterozygote. For example, it is possible that in heterozygotes exhibiting a degree of heterology lower than that present in crosses 1 and 2, MuDR might increase the rate of NCOs. This, however, seems unlikely given that reversions of Mu-induced alleles of all loci studied are quite rare in diverse genetic backgrounds, which would be expected to exhibit varying levels of heterology at the target loci.
MuDR increases the rate of COs at a1:
According to the most widely accepted recombination models (SZOSTAK et al. 1983; SUN et al. 1991; ALLERS and LICHTEN 2001), which are well supported by data from yeast, meiotic recombination is initiated by DSBs and the subsequent repair and resolution of these DSBs results in COs or NCOs. Although in plants the mechanisms underlying meiotic recombination are not as well understood, it is thought that they are at least similar to those that occur in yeast. Support for this view comes from the isolation of plant homologs of many of the yeast genes involved in DSB processing and meiotic recombination (reviewed in BHATT et al. 2001; SCHWARZACHER 2003). Also, agents that introduce DSBs into plant chromosomes increase rates of mitotic recombination and intrachromosomal recombination.
It has not yet, however, been determined in plants whether the stimulation of DNA breaks increases the rate of homologous meiotic recombination. To address this question, recombinants were isolated from a1-mum2/a1::rdt heterozygotes that carried or did not carry MuDR. MuDR encodes a transposase (CHOMET et al. 1991; HERSHBERGER et al. 1991; QIN et al. 1991; HSIA and SCHNABLE 1996) required for the transposition of Mu elements. This transposition must involve some type of DNA breaks. The recovery of chromosomes that contain deletions of sequences adjacent to Mu elements and internally deleted Mu elements is consistent with the existence of MuDR-catalyzed DNA breaks in the vicinity of Mu elements (LEVY et al. 1989; LEVY and WALBOT 1991; LISCH et al. 1995; HSIA and SCHNABLE 1996; ASAKURA et al. 2002; KIM and WALBOT 2003). Because the mechanism by which MuDR catalyzes transposition has not been determined, these MuDR-generated breaks could be either DSBs or single-strand nicks. It has long been assumed that these breaks were DSBs, but recent evidence suggests that single-strand nicks can stimulate homologous recombination in the V(D)J regions of immunoglobulin genes (LEE et al. 2004). Hence, MuDR-generated single-strand nicks could affect recombination directly or could be converted into DSBs during DNA replication (KUZMINOV 2001). Alternatively, MuDR may directly catalyze DSB formation.
Regardless of the mechanism by which MuDR generates DNA breaks, crosses harboring MuDR would be expected to have an increased rate of breaks in the vicinity of the Mu1 insertion in the a1-mum2 allele. At least three processes could repair MuDR-induced DNA breaks at a1-mum2. These include COs between the two homologs (class III and IV events, Figure 1); conversion of the Mu1-containing homolog, using as template the homolog that does not contain Mu1 (class VI NCO events, Figure 1); or DSB repair, using as template the sister chromatid. This latter process would not generate recombinant chromosomes and would in fact regenerate the parental a1-mum2 allele or either internal or adjacent deletions of Mu1 if gap repair is interrupted (LISCH et al. 1995; HSIA and SCHNABLE 1996; ASAKURA et al. 2002; KIM and WALBOT 2003).
MuDR is required for the transposition of Mu transposons, a process that requires the introduction of DNA breaks. According to accepted recombination models, COs are also initiated by DNA breaks. Hence, we interpret our observation that plants that carry MuDR exhibit four times more class III COs (Figure 1) than do plants that do not carry MuDR (Table 2) to indicate that during meiosis at least some MuDR-induced DNA breaks are repaired via the CO pathway. Hence, our results strongly suggest that DNA breaks stimulate meiotic COs in plants.
Although MuDR stimulates meiotic COs, the Ac transposon does not (DOONER and MARTINEZ-FEREZ 1997a). This suggests that Ac-induced DSBs are separated temporally or spatially from meiotic recombination (DOONER and MARTINEZ-FEREZ 1997a) or that Ac-induced DSBs are repaired by another pathway, for example via the formation of hairpins followed by nonhomologous end joining (NHEJ) repair at sites of microhomology (WEIL and KUNZE 2000; YU et al. 2004).
We cannot rule out the possibility that the increased rates of CO observed in this study are not a direct consequence of an increased rate of breaks at Mu1 but are instead a consequence of potential changes in the chromatin architecture at a1-mum2 that occur in the presence of MuDR. Even if the transposase per se is not generating breaks at the Mu1 insertion site, the transposase must at least be creating a local environment that is more conducive to the formation of endogenous DNA breaks. If this alternative model is correct, a MuDR-encoded transposase that can bind to Mu terminal inverted repeats, but that cannot catalyze transposition, should increase the rate of recombination in the vicinity of a Mu insertion. Regardless of whether the breaks are caused directly or indirectly by MuDR, it is clear that MuDR stimulates the formation of DNA breaks at the a1-mum2 allele, resulting in increased rates of meiotic CO.
Does MuDR affect the rate of gene conversion?
There are two classes of NCOs (classes V and VI, Figure 1). Although MuDR increased the rate of COs it did not increase the rate of class V NCOs in the a1-mum2/a1::rdt heterozygote. This is as expected because MuDR would not be predicted to interact with a1::rdt. In contrast, the existence of somatic excision events in MuDR-containing stocks demonstrates that MuDR interacts (directly or indirectly) with Mu insertions. Even so, germinal reverants of Mu-insertion alleles (including class VI NCOs) are rare (reviewed in BENNETZEN 1996; LISCH 2002; WALBOT and RUDENKO 2002); the frequencies of germinal revertants from the bronze locus are 8 x 10–5 (BROWN et al. 1989b) and between 4.9 x 10–6 and 2.3 x 10–3 (SCHNABLE et al. 1989). Consistent with these results, we and others (LISCH et al. 1995) have shown that the rate of class VI NCOs at a1-mum2 is low. It is puzzling that in this heterozygote, although MuDR increases the rate of CO fourfold, the rate of class VI NCOs is low in the presence of MuDR.
In yeast, several meiotic mutants that affect SEI and DHJ formation drastically reduce the frequency of COs but not NCOs. This suggests that recombination outcomes (CO vs. NCO repair) are determined prior to stable strand exchange (i.e., SEI; BORNER et al. 2004). If this is also true in plants, the MuDR-generated breaks might be designated prior to strand exchange to be repaired by the CO pathway. In V(D)J site-specific recombination, the RAG recombinases act as molecular shepherds that allow repair of the RAG-generated DSB by the NHEJ machinery and not other repair pathways (LEE et al. 2004). Similarly, during meiosis the MuDR transposase and/or protein(s) involved in the meiotic recombination machinery may remain bound to the MuDR-generated DNA breaks and thereby channel repairs to the CO pathway. It is also possible that some repairs of MuDR-induced DNA breaks could be channeled to pathways that were not detected in this study because they do not yield COs or germinal revertants (e.g., repair using the sister chromatid as template or NHEJ). The high somatic and low germinal reversion rates observed in the Mu system could be explained if these "molecular shepherds" differ between the mitotic and meiotic cellular programs.
MuDR does not affect the distribution of recombination breakpoints:
Insertion/deletion polymorphisms (IDPs) and transposon insertions suppress recombination in nearby regions of the bz1 locus (DOONER and MARTINEZ-FEREZ 1997b). By doing so these IDPs change the distribution of recombination breakpoints across the bz1 locus, creating apparent recombination hotspots. In contrast, the distribution of recombination breakpoints across the a1 gene is not affected by the Mu1 insertion at position –97 in a1-mum2 (XU et al. 1995; YAO et al. 2002). Hence, the recombination hotspot reported by XU et al. (1995) is not a consequence of the Mu1 insertion in one of the parental alleles used by that study.
Although the rates of CO at a1 increase fourfold in the presence of MuDR, the distribution of recombination breakpoints is not altered by MuDR; 19 of the 24 characterized breakpoints cluster in the 377-bp hotspot at the 5' end of the a1 gene (Figure 2B) previously identified in stocks that lack MuDR (XU et al. 1995). Hence, the COs that are apparently initiated by MuDR-induced breaks at the Mu1 insertion at –97 resolve at the same positions within a1 as do the COs that are initiated by the DSBs that form in the absence of MuDR. COs resolve in this same hotspot even in an A1 allele that does not contain a Mu1 insertion (YAO et al. 2002). These findings are consistent with the hypothesis that the DSBs that initiate recombination in nonmutant a1 alleles occur in the vicinity of position –97. Even though the relationship between the positions of DSB hotspots and recombination hotspots has not yet been studied in plants, the observation that most a1 recombinants mapped within 700 bp of the Mu1 insertion site is consistent with the observation that most recombination hotspots in budding yeast are located within 1 or 2 kb of a DSB hotspot (SMITH 2001).
MuDR is active in meiotic cells:
Although the Mu transposase mudrA and mudrB transcripts are highly expressed in mature pollen and mudrB promoter::reporter constructs are expressed at levels 20-fold higher in pollen than in leaves (RAIZADA et al. 2001a), to date, there has not been conclusive evidence that the MuDR transposase is active and functions during meiosis. Instead, germinally transmissible Mu insertions could arise via late somatic excision in premeiotic cells (reviewed in WALBOT and RUDENKO 2002) or via postmeiotic transposition (ROBERTSON and STINARD 1993). In contrast, by demonstrating that MuDR increases rates of meiotic recombination at a1-mum2 fourfold, the current study provides the first direct evidence that MuDR is active during meiosis.
Evolutionary implications:
This study extends our understanding of how transposons can alter recombination rates. The insertion of a nonautonomous transposon into a gene typically reduces the rate of intragenic recombination. For example, the insertion of a Mu1 transposon into the A1-LC allele (which generated a1-mum2) reduced the rate of recombination approximately twofold (XU et al. 1995). In contrast and as revealed by this study, if a Mu-insertion allele is present in a genome that contains an active copy of MuDR, the rate of intragenic recombination can actually be higher (twofold in this case) than that of the original allele that lacked a transposon insertion. Because a variety of plant DNA transposons have an affinity for inserting into genes (BUREAU et al. 1996; RAIZADA et al. 2001b; JIANG et al. 2004) and intragenic recombination can generate new alleles, their ability to alter rates of intragenic recombination could have significant evolutionary implications.
APPENDIX Positions of the recombination breakpoints associated with the A1' alleles isolated from crosses 1 and 2
|
NA, not applicable.
a All of the A1' alleles from Table 1 that were confirmed by RFLP analysis are listed.
b Intervals as shown in Figure 2.
c Not determined because the associated A1' allele failed to PCR amplify.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
|
FOOTNOTES |
|---|
2 Present address: Department of Biological Sciences, University of Missouri, Columbia, MO 65211.
3 Present address: Department of Pathology, University of Wisconsin, Madison, WI 53706.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
ALLERS, T., and M. LICHTEN, 2001 Differential timing and control of noncrossover and crossover recombination during meiosis. Cell 106: 47–57.[CrossRef][Medline]
ASAKURA, N., C. NAKAMURA, T. ISHII, Y. KASAI and S. YOSHIDA, 2002 A transcriptionally active maize MuDR-like transposable element in rice and its relatives. Mol. Genet. Genomics 268: 321–330.[CrossRef][Medline]
ATHMA, P., and T. PETERSON, 1991 Ac induces homologous recombination at the maize P locus. Genetics 128: 163–173.[Abstract]
BENNETZEN, J. L., 1996 The Mutator transposable element system of maize. Curr. Top. Microbiol. Immunol. 204: 195–229.[Medline]
BETRAN, E., J. ROZAS, A. NAVARRO and A. BARBADILLA, 1997 The estimation of the number and the length distribution of gene conversion tracts from population DNA sequence data. Genetics 146: 89–99.[Abstract]
BHATT, A. M., C. CANALES and H. G. DICKINSON, 2001 Plant meiosis: the means to 1N. Trends Plant Sci. 6: 114–121.[CrossRef][Medline]
BORNER, G. V., N. KLECKNER and N. HUNTER, 2004 Crossover/noncrossover differentiation, synaptonemal complex formation, and regulatory surveillance at the leptotene/zygotene transition of meiosis. Cell 117: 29–45.[CrossRef][Medline]
BORTS, R. H., and J. E. HABER, 1987 Meiotic recombination in yeast: alteration by multiple heterozygosities. Science 237: 1459–1465.
BORTS, R. H., and J. E. HABER, 1989 Length and distribution of meiotic gene conversion tracts and crossovers in Saccharomyces cerevisiae. Genetics 123: 69–80.
BORTS, R. H., W. Y. LEUNG, W. KRAMER, B. KRAMER, M. WILLIAMSON et al., 1990 Mismatch repair-induced meiotic recombination requires the pms1 gene product. Genetics 124: 573–584.[Abstract]
BORTS, R. H., S. R. CHAMBERS and M. F. ABDULLAH, 2000 The many faces of mismatch repair in meiosis. Mutat. Res. 451: 129–150.[Medline]
BROWN, J. J., M. G. MATTES, C. O'REILLY and N. S. SHEPHERD, 1989a Molecular characterization of rDt, a maize transposon of the "Dotted" controlling element system. Mol. Gen. Genet. 215: 239–244.[CrossRef][Medline]
BROWN, W. E., D. S. ROBERTSON and J. L. BENNETZEN, 1989b Molecular analysis of multiple Mutator-derived alleles of the Bronze locus of maize. Genetics 122: 439–445.
BUREAU, T. E., P. C. RONALD and S. R. WESSLER, 1996 A computer-based systematic survey reveals the predominance of small inverted-repeat elements in wild-type rice genes. Proc. Natl. Acad. Sci. USA 93: 8524–8529.
CHEN, W., and S. JINKS-ROBERTSON, 1999 The role of the mismatch repair machinery in regulating mitotic and meiotic recombination between diverged sequences in yeast. Genetics 151: 1299–1313.
CHIURAZZI, M., A. RAY, J. F. VIRET, R. PERERA, X. H. WANG et al., 1996 Enhancement of somatic intrachromosomal homologous recombination in Arabidopsis by the HO endonuclease. Plant Cell 8: 2057–2066.[Abstract]
CHOMET, P., D. LISCH, K. J. HARDEMAN, V. L. CHANDLER and M. FREELING, 1991 Identification of a regulatory transposon that controls the Mutator transposable element system in maize. Genetics 129: 261–270.[Abstract]
CHOU, Q., M. RUSSELL, D. E. BIRCH, J. RAYMOND and W. BLOCH, 1992 Prevention of pre-PCR mis-priming and primer dimerization improves low-copy-number amplifications. Nucleic Acids Res. 20: 1717–1723.
CIVARDI, L., Y. XIA, K. J. EDWARDS, P. S. SCHNABLE and B. J. NIKOLAU, 1994 The relationship between genetic and physical distances in the cloned a1-sh2 interval of the Zea mays L. genome. Proc. Natl. Acad. Sci. USA 91: 8268–8272.
COLLINS, I., and C. S. NEWLON, 1994 Meiosis-specific formation of joint DNA molecules containing sequences from homologous chromosomes. Cell 76: 65–75.[CrossRef][Medline]
DE MASSY, B., and A. NICOLAS, 1993 The control in cis of the position and the amount of the ARG4 meiotic double-strand break of Saccharomyces cerevisiae. EMBO J. 12: 1459–1466.[Medline]
DETLOFF, P., M. A. WHITE and T. D. PETES, 1992 Analysis of a gene conversion gradient at the HIS4 locus in Saccharomyces cerevisiae. Genetics 132: 113–123.[Abstract]
DOONER, H. K., 1986 Genetic fine structure of the bronze locus in maize. Genetics 113: 1021–1036.
DOONER, H. K., 2002 Extensive interallelic polymorphisms drive meiotic recombination into a crossover pathway. Plant Cell 14: 1173–1183.
DOONER, H. K., and I. M. MARTINEZ-FEREZ, 1997a Germinal excisions of the maize transposon activator do not stimulate meiotic recombination or homology-dependent repair at the bz locus. Genetics 147: 1923–1932.[Abstract]
DOONER, H. K., and I. M. MARTINEZ-FEREZ, 1997b Recombination occurs uniformly within the bronze gene, a meiotic recombination hotspot in the maize genome. Plant Cell 9: 1633–1646.[Abstract]
EGGLESTON, W. B., M. ALLEMAN and J. L. KERMICLE, 1995 Molecular organization and germinal instability of R-stippled maize. Genetics 141: 347–360.[Abstract]
ELLIOTT, B., C. RICHARDSON, J. WINDERBAUM, J. A. NICKOLOFF and M. JASIN, 1998 Gene conversion tracts from double-strand break repair in mammalian cells. Mol. Cell. Biol. 18: 93–101.
FAN, Q., F. XU and T. D. PETES, 1995 Meiosis-specific double-strand DNA breaks at the HIS4 recombination hot spot in the yeast Saccharomyces cerevisiae: control in cis and trans. Mol. Cell. Biol. 15: 1679–1688.[Abstract]
GAME, J. C., 1992 Pulsed-field gel analysis of the pattern of DNA double-strand breaks in the Saccharomyces genome during meiosis. Dev. Genet. 13: 485–497.[CrossRef][Medline]
GERTON, J. L., J. DERISI, R. SHROFF, M. LICHTEN, P. O. BROWN et al., 2000 Inaugural article: global mapping of meiotic recombination hotspots and coldspots in the yeast Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 97: 11383–11390.
GLOOR, G. B., N. A. NASSIF, D. M. JOHNSON-SCHLITZ, C. R. PRESTON and W. R. ENGELS, 1991 Targeted gene replacement in Drosophila via P element-induced gap repair. Science 253: 1110–1117.
GOYON, C., and M. LICHTEN, 1993 Timing of molecular events in meiosis in Saccharomyces cerevisiae: stable heteroduplex DNA is formed late in meiotic prophase. Mol. Cell. Biol. 13: 373–382.
HERSHBERGER, R. J., C. A. WARREN and V. WALBOT, 1991 Mutator activity in maize correlates with the presence and expression of the Mu transposable element Mu9. Proc. Natl. Acad. Sci. USA 88: 10198–10202.
HILLIKER, A. J., G. HARAUZ, A. G. REAUME, M. GRAY, S. H. CLARK et al., 1994 Meiotic gene conversion tract length distribution within the rosy locus of Drosophila melanogaster. Genetics 137: 1019–1026.[Abstract]
HOLLIDAY, R., 1964 A mechanism for gene conversion in fungi. Genet. Res. 78: 282–304.
HSIA, A. P., and P. S. SCHNABLE, 1996 DNA sequence analyses support the role of interrupted gap repair in the origin of internal deletions of the maize transposon, MuDR. Genetics 142: 603–618.[Abstract]
HUNTER, N., and N. KLECKNER, 2001 The single-end invasion: an asymmetric intermediate at the double-strand break to double-Holliday junction transition of meiotic recombination. Cell 106: 59–70.[CrossRef][Medline]
JIANG, N., C. FESCHOTTE, X. ZHANG and S. R. WESSLER, 2004 Using rice to understand the origin and amplification of miniature inverted repeat transposable elements (MITEs). Curr. Opin. Plant Biol. 7: 115–119.[CrossRef][Medline]
KIM, S. H., and V. WALBOT, 2003 Deletion derivatives of the MuDR regulatory transposon of maize encode antisense transcripts but are not dominant-negative regulators of mutator activities. Plant Cell 15: 2430–2447.
