The Plant Cell, Vol. 7, 2151-2161, December 1995 O 1995 American Society of Plant Physiologists Meiotic Recombination Break Points Resolve at High Rates at the 5' End of a Maize Coding Sequence Xiaojie Xu,' An-Ping Hsia,a Lei Zhang,a Basil J. Nikolau,b and Patrick S. Schnable a7c2' a Department of Zoology and Genetics, lowa State University, Ames, lowa 50011 Department of Biochemistry and Biophysics, lowa State University, Ames, lowa 50011 Department of Agronomy, lowa State University, Ames, lowa 50011 Sequence analysis of recombination break points has defined a 377-bp recombination hot spot within the anthocyaninl (al) gene. One-fifth of all recombination events that occurred within the 140-kb al-shrunken2 intenral resolved within this 377-bp hot spot. In yeast, meiotic double-strand breaks in chromosomal DNA are thought to initiate recombination and are generally located 5' of coding regions, near transcription promoter sequences. Because the a l recombination hot spot is located within the 5' transcribed region of the al gene, the sites at which recombination events initiate and resolve appear to be different, but both appear to be regulated in relation to transcribed sequences. Although transposon insertions are known to suppress recombination and alter the ratio of crossovers to apparent gene conversions, the Mutatorl transposon insertion in the a%mum2 allele does not alter the sites at which recombination events resolve. INTRODUCTION Meiotic recombination generates chromosomes with nove1 al- lelic arrays relative to their progenitor chromosomes. This is one of the processes that yields the genetic diversity on which evolutionary selection can act. However, recombination must be counterbalanced by the advantage of maintaining coevolved blocks of favorable alleles as a unit during meiosis. Indeed, recombination is highly regulated, as evidenced by the con- trol of the frequency (Schuchert et al., 1991; Korol and Iliad, 1994) and the position (Roberts, 1965; Lambie and Roeder, 1986; Tanksley et al., 1992; Moore et al., 1993) of recombina- tion events on chromosomes. Thus, meiotic recombination events do not occur randomly, and certain large intervals (e.g., those surrounding the centromeres) exhibit low rates of recom- bination per physical length. In addition, some relatively small chromosomal segments in a number of organisms are recom- binationally hyperactive, that is, they serve as recombination hot spots (reviewed in Lichten, 1995; also, see Nelson, 1968; Freeling, 1978; Stahl, 1979; Dooner et al., 1985, 1991; Wessler and Varagona, 1985; Dooner, 1986; Sachs et al., 1986; Uematsu et al., 1986; Steinmetz et al., 1987; Ponticelli et al., 1988; Dorer et al., 1989; Brown and Sundaresan, 1991; Janson et al., 1991; Symington et al., 1991; Shiroishi et al., 1993; Civardi et al., 1994). These recombination hot spots exhibit ratios between physical and genetic distances (a ratio that has been termed p; Civardi et al., 1994) that are threefold to 100-fold lower than the genome average. In many instances, chromosome regions that exhibit low values of p (i.e., recombination hot spots) oc- cur at or in the vicinity of genes. To whom correspondence should be addressed. The favored mechanistic model of recombination (Szostak et al., 1983) predicts that recombination is initiated by the gener- ation of a double-stranded break (DSB) on one chromosome of an allelic pair (Figure 1). After strand invasion, DNA synthe- sis, and the creation of a Holliday Junction (which can undergo branch migration), the recombination event resolves, leading to an exchange of strands and thus "crossing over." As pre- dicted by this model, in yeast, DSBs are associated with recombination hot spots (Nicolas et al., 1989; Sun et al., 1989; Cao et al., 1990; Schultes et al., 1990; Goldway et al., 1993; Nag et al., 1993). Analyses of the distribution of DSBs along yeast chromosomes suggest that recombination events initiate in the vicinity of transcription promoters (Wu and Lichten, 1994). Much less is known about where recombination events resolve subsequent to their initiation. To determine the physical rela- tionship between the position at which DSBs occur and the final recombination break points, we have used DNA sequence polymorphisms to map recombination break points to high reso- lution within a maize gene (the anrhocyaninl [a71 locus) that serves as a recombination hot spot (Brown and Sundaresan, 1991; Civardi et al., 1994). RESULTS Mutants at the a7 and shrunken2 (sh2) loci confer visible pheno- types that can be scored on kernels. This feature, in conjunction with the ease with which large numbers of kernels from an F, generation can be produced in maize, makes these genes and the chromosomal interval defined by them ideal models
Transcript
The Plant Cell, Vol. 7, 2151-2161, December 1995 O 1995 American Society of Plant Physiologists
Meiotic Recombination Break Points Resolve at High Rates at the 5' End of a Maize Coding Sequence
Xiaojie Xu,' An-Ping Hsia,a Lei Zhang,a Basil J. Nikolau,b and Patrick S. Schnable a7c2'
a Department of Zoology and Genetics, lowa State University, Ames, lowa 50011 Department of Biochemistry and Biophysics, lowa State University, Ames, lowa 50011 Department of Agronomy, lowa State University, Ames, lowa 50011
Sequence analysis of recombination break points has defined a 377-bp recombination hot spot within the anthocyaninl (a l ) gene. One-fifth of all recombination events that occurred within the 140-kb al-shrunken2 intenral resolved within this 377-bp hot spot. In yeast, meiotic double-strand breaks in chromosomal DNA are thought to initiate recombination and are generally located 5' of coding regions, near transcription promoter sequences. Because the a l recombination hot spot i s located within the 5' transcribed region of the al gene, the sites at which recombination events initiate and resolve appear t o be different, but both appear to be regulated in relation to transcribed sequences. Although transposon insertions are known to suppress recombination and alter the ratio of crossovers to apparent gene conversions, the Muta tor l transposon insert ion in the a%mum2 allele does not alter the sites at which recombination events resolve.
INTRODUCTION
Meiotic recombination generates chromosomes with nove1 al- lelic arrays relative to their progenitor chromosomes. This is one of the processes that yields the genetic diversity on which evolutionary selection can act. However, recombination must be counterbalanced by the advantage of maintaining coevolved blocks of favorable alleles as a unit during meiosis. Indeed, recombination is highly regulated, as evidenced by the con- trol of the frequency (Schuchert et al., 1991; Korol and Iliad, 1994) and the position (Roberts, 1965; Lambie and Roeder, 1986; Tanksley et al., 1992; Moore et al., 1993) of recombina- tion events on chromosomes. Thus, meiotic recombination events do not occur randomly, and certain large intervals (e.g., those surrounding the centromeres) exhibit low rates of recom- bination per physical length. In addition, some relatively small chromosomal segments in a number of organisms are recom- binationally hyperactive, that is, they serve as recombination hot spots (reviewed in Lichten, 1995; also, see Nelson, 1968; Freeling, 1978; Stahl, 1979; Dooner et al., 1985, 1991; Wessler and Varagona, 1985; Dooner, 1986; Sachs et al., 1986; Uematsu et al., 1986; Steinmetz et al., 1987; Ponticelli et al., 1988; Dorer et al., 1989; Brown and Sundaresan, 1991; Janson et al., 1991; Symington et al., 1991; Shiroishi et al., 1993; Civardi et al., 1994). These recombination hot spots exhibit ratios between physical and genetic distances (a ratio that has been termed p; Civardi et al., 1994) that are threefold to 100-fold lower than the genome average. In many instances, chromosome regions that exhibit low values of p (i.e., recombination hot spots) oc- cur at or in the vicinity of genes.
To whom correspondence should be addressed.
The favored mechanistic model of recombination (Szostak et al., 1983) predicts that recombination is initiated by the gener- ation of a double-stranded break (DSB) on one chromosome of an allelic pair (Figure 1). After strand invasion, DNA synthe- sis, and the creation of a Holliday Junction (which can undergo branch migration), the recombination event resolves, leading to an exchange of strands and thus "crossing over." As pre- dicted by this model, in yeast, DSBs are associated with recombination hot spots (Nicolas et al., 1989; Sun et al., 1989; Cao et al., 1990; Schultes et al., 1990; Goldway et al., 1993; Nag et al., 1993). Analyses of the distribution of DSBs along yeast chromosomes suggest that recombination events initiate in the vicinity of transcription promoters (Wu and Lichten, 1994). Much less is known about where recombination events resolve subsequent to their initiation. To determine the physical rela- tionship between the position at which DSBs occur and the final recombination break points, we have used DNA sequence polymorphisms to map recombination break points to high reso- lution within a maize gene (the anrhocyaninl [a71 locus) that serves as a recombination hot spot (Brown and Sundaresan, 1991; Civardi et al., 1994).
RESULTS
Mutants at the a7 and shrunken2 (sh2) loci confer visible pheno- types that can be scored on kernels. This feature, in conjunction with the ease with which large numbers of kernels from an F, generation can be produced in maize, makes these genes and the chromosomal interval defined by them ideal models
2152 The Plant Cell
A
E
F
h i j
H I I J
5' + 3'
3'
1
1
5'
- 3'
3
4-4 G
H i ' i j
H I J h I J
Figure 1. The Double-Strand DNA Break Repair Model Explains Reciproca1 Recombination and Gene Conversion.
(A) After premeiotic DNA replication, a cell contains four DNA duplexes carrying markers h, i, and j. In subsequent panels, only the two duplexes that recombine are illustrated. (8) A double-strand break occurs in one duplex. (C) A 5' to 3' exonuclease attacks the exposed 5' ends. (D) Strand invasion. The free 3' ends are used as primers for DNA synthesis. The uncut homolog (black) serves as the template in these synthesis reactions. (E) DNA synthesis (dashed lines) results in a four-strand intermediate with two Holliday Junctions. This structure can be resolved by cutting and religating at two sites involving the inner strands (resulting in [F]) or at one site involving the inner strands and another site involving the outer strands (resulting in [G]). (F) Gene conversion in the absence of reciprocal recombination oc- curs if cuts and religation occur at positions 1 and 2 as depicted in (E). In this example, marker i from the gray duplex has been converted to I from the black duplex. Markers h and j, which flank the conversion tract, have not exchanged. The maximum extent of the gene conver- sion tract is indicated by the positions of the vertical arrows. The actual extent of the conversion tract depends on the results of mismatch repair. (G) Gene conversion in combination with reciprocal recombination. Molecules recombinant for markers h and j would result if cuts and religation occur at positions 2 and 3, as depicted in (E). The recombi- nation break point is adjacent to a gene conversion tract. This conversion tract would usually not be detectable in plants. Instead, the apparent position of the recombination break point would be shifted from the position indicated by the right arrow to that indicated by the
---I
--I - - -
A
B
C E N ~ pkpinnxn a1 sk2
I cM 0.1 eM 3L -o//
140 kb
ol-mum2
I a1::rdr
a1::rdt n1::rdt
In the absense af recombination
(Mos1 pmgeny) (Rare pmgeny)
AI ' sh2 sh2
Mul rdl _I
a1::df
a1::rdt n1::rdt a1::rdf
colorless mund colorless shrunken colored shrunken
Figure 2. lsolation of lntragenic Recombinant Al' Alleles from Cross 1.
(A) Genetic and physical distances on the long arm of chrornosome 3. CEN 3 indicates the centromere. (6) lsolation of intragenic recombinants from cross 1.
for the study of meiotic recombination. Another importam fea- ture, the al-sh2 interval, has been cloned as a yeast artificial chromosome (Civardi et al., 1994).
lsolation of Recombinants
As depicted in Figure 26, intragenic recombinants at the a7 locus could be selected directly from the progeny of cross 1 (see Methods) on the basis of their phenotype. The a7-mum2 and a7::rdf mutant alleles contain transposon insertions in the a7 sequence that disrupt gene function. al-mum2 and a7::rdt behave as stable, recessive alleles when carried by kernels with the genotypes used in this study (see Methods). Thus, kernels that carry both of these mutant alleles are colorless. Most of the gametes generated from plants derived from such kernels carried one of these two mutant alleles. However, rare recombination events between the two transposons (which are 1.2 kb apart) generated chimeric dominant alleles (A77 that condition colored aleurones, as shown in Figure 3. These ker-
left arrow. If the Holliday Junctions migrate away from the DSB site after DNA synthesis (E), the apparent recombination break point would be further removed from the DSB site at which this recombination event initiated. Modified from Szostak et ai. (1983).
Crossovers Resolve at the 5' End of the a 7 Gene 2153
nels that carry AT alleles were shrunken because the 5' endof the a1 gene is closest to the sh2 gene (Civardi et al., 1994).Twenty-four such recombinants were isolated from a total popu-lation of 742,100 kernels generated over two pollinating seasons(Table 1). Hence, as shown in Figure 4B, the genetic distanceassociated with this 1.2-kb interval is 0.0065 ± 0.0009 cen-timorgans (cM) (see Methods). Of the 24 recombinants, 17 wereanalyzed. The validity of each of these 17 recombinants wasconfirmed via genetic crosses and DNA gel blotting experi-ments using a closely linked restriction fragment lengthpolymorphisms (RFLP) marker (see Methods).
Because certain transposons can suppress rates of recom-bination (Dooner, 1986) or alter the ratio of crossovers to putativegene conversions (Dooner and Kermicle, 1986; but see Doonerand Ralston, 1990), we were concerned that the Mutatorl (Mu1)transposon in the a1-mum2 allele might affect the distributionof recombination break points in the al locus. To control forany effects that the Mu1 transposon might have on recombi-nation, intragenic recombinants at the a1 locus also wereisolated from another heterozygous genotype (cross 2), as de-scribed in Figure 5A. Cross 2 is identical to cross 1, exceptthat in the former cross, the a1-mum2 allele has been replacedby the wild-type A1-LC allele. The A1-LC allele is identical insequence to a1-mum2, except that the former allele lacks aMul insertion (see Methods). Because the A1-LC allele condi-tions a colored aleurone, it was not possible to identifyintragenic recombinants from cross 2 directly. Instead, therecombination break points associated with 21 A1 sh2 (and35 a1 Sh2) recombinant chromosomes isolated from 67,000gametes produced by A1-LC Sh2/a1::rdt sh2 heterozygotes weremapped relative to the a1 locus. These recombinants shouldcarry recombination break points somewhere within the 140-kb interval defined by the a7 and sh2 loci (Figure 5A). To iden-tify those chromosomes carrying recombination break points
Table 1. Rate of Intragenic Recombination at the a1 Locus
Year1991"1993Pooled0
No. ColoredShrunkenKernels9
1524°
PopulationSize215,300526,800742,100
cMa
0.00840.00570.0065
± 0.0009
Figure 3. Phenotype Conferred by Chimeric AT Alleles.
The single, colored kernel on this ear derived from cross 1 has thegenotype AT sh2la1::rdt sh2.
a Intragenic recombination was assayed using the scheme outlinedin Figure 2B. Because only one of the two possible classes of recom-bination events could be identified (see Figure 2B), the genetic distancewas calculated by doubling the rate at which colored, shrunken recom-binants were recovered. This calculation assumes that two events occurat equal frequencies. SE was calculated according to the equation(pq/n)">.b Data from Civardi et al (1994). Subsequent genetic tests estab-lished that one of the 10 colored, shrunken kernels originally reportedby Civardi et al. (1994) did not carry a legitimate intragenic recombinant.c The homogeneity x2 value (0.839, with one degree of freedom) in-dicates that there is no significant difference between the rate ofintragenic recombination in 1991 and 1993. The data for these 2 yearswere therefore pooled." In addition to the colored, shrunken kernels, a number of colored,round kernels were recovered in both years. Apparently, most of thesearose via pollen contamination of the experimental plot from a nearbycommercial cornfield. This conclusion is based on results obtainedby self-pollinating several dozen plants derived from colored, roundkernels. All the resulting selfed ears were distinct phenotypically fromthe a1-mum2 and a 1: :rdt stocks and segregated genetic markers pres-ent in commercial maize but absent from the experimental stocks (datanot shown). However, some fraction of the colored, round kernels mayhave arisen via gene conversion events that removed the Mu1 ele-ment from the a1-mum2 chromosome.
within or near the a1 locus, plants carrying the 21 indepen-dently derived Al sh2 recombinant chromosomes (and the 35a1 Sh2 chromosomes) were analyzed via DNA gel blots. ThoseA1 sh2 chromosomes that arose via crossovers within the 6-kbinterval defined by the rdt insertion in the aT.:rdt allele andthe first EcoRI site distal to the a7 locus (indicated with an aster-isk in Figure 5A) would be expected to display a novel RFLPpattern (Figure 5A; see Methods).
All of the progeny from cross 2 carry a nonrecombinant aT.:rdtsh2 chromosome from their paternal parent. Within the a1-sh2interval, this chromosome is identical to the a1::rdtsh2 paren-tal chromosome shown in Figure 5A. Recombinant progenyalso carry a chimeric chromosome derived from their maternalparent. Chimeric chromosomes that arose via recombinationbreak points occurring within the 134-kb interval defined bythe indicated EcoRI site and the sh2 locus have the genotypea1::rdt Sh2 or A1-LC sh2 and carry parental al alleles (Figure5B, lanes 3 and 5, respectively). In contrast, those chimericchromosomes that arose via recombination break points withinthe 6-kb interval can be identified because they carry novela7-hybridizing EcoRI restriction fragments (Figure 5B, lanes
2154 The Plant Cell
4 and 6, respectively). These novel a1 alleles are designatedA1' and al". The novel restriction fragments associated withAT alleles are not present in nonrecombinant siblings (i.e.,plants grown from colorless, shrunken, or colored round ker-nels from cross 2) of several AT sh2 recombinants (data notshown). By this RFLP assay, eight of the 21 Al sh2 (and eightof 35 a1 Sh2) chromosomes were established to have arisenvia crossovers within a 6-kb interval 5' of position +1083 (therdt insertion site in the a1 locus). The genetic distance as-sociated with this 6-kb interval is therefore 0.026 cM (seeMethods and Figure 4A).
Mapping Recombination Break Points
By virtue of their origin, the chimeric AT and A1' alleles gener-ated via crosses 1 and 2 include portions of the aT.-.rdt anda1-mum2 (or A1-LC) alleles. As shown in Figure 6, the aT.:rdtallele contains a centrally located Pstl recognition site (indi-cated with an asterisk) that is absent from the correspondinginterval of the a1-mum2 and A1-LC alleles. Fifteen of the ATand all eight of the AT alleles were polymerase chain reac-tion (PCR) amplified, gel purified, and subjected to Pstldigestion. As shown in Figures 7A and 7B, these analyses
Figure 4. Relationships between Physical and Genetic Distances.
The average value of this ratio (p) in the maize genome is 1456 kb/cM(Civardi et al., 1994).(A) Physical distances, genetic distances, and values of p within thea1-sh2 interval. EcoRI* indicates the first EcoRI restriction endonucleaserecognition site centromere distal (left in this figure) of the a1 locus.Other EcoRI sites are not shown.(B) Physical distance, genetic distance, and value of p within the atlocus.
EcoRI
I——
^
^H
a/probe
Parental Chromosomes (Maternal Parent of Cross 2)
EcoRI Al-LC 5' Hindlll EcoRI*
Sh2
700bp
Chimeric (Recombinant) Chromosomes Derived From the Maternal Parent of Cross 2
EcoRI Alt EcoRI sh2
700bp
B
10kb
9kb
1 2 3 4 5 6
Figure 5. AT Alleles That Arise via Recombination within a 6-kb In-terval That Includes the a? Locus Can Be Identified via Novel RFLPs.
(A) The physical distances associated with the indicated intervals ofparental chromosomes from cross 2 are shown in the top three lines.The fourth line indicates the extent of the a1 probe. Lines five andsix show the positions of relevant restriction enzyme recognition siteson the parental chromosomes relative to the a7 and sh2 loci. The po-sition of the 700-bp rdt transposon insertion is indicated. (The Xindicates a crossover that would generate a progeny chromosome thathas a novel aJ-hybridizing RFLP.) Lines seven and eight representschematics of recombinant chromosomes derived from cross 2 viaa crossover in the position indicated by the X between lines five and six.(B) DMA samples from plants homozygous for Al-LC or aT.:rdt (lanes1 and 2) or from plants grown from colorless, round (lanes 3 and 4)or colored, shrunken (lanes 5 and 6) kernels carrying recombinantchromosomes derived from cross 2 were digested with EcoRI, trans-ferred to nylon membrane, and hybridized with the a1 probe indicatedin (A). Numbers at left indicate the positions of molecular size stan-dards in kilobases.
demonstrated that in all 23 cases, the recombination breakpoints associated with the AT and AT alleles resolved 5' ofthe diagnostic Pstl site. Because the recombination breakpoints associated with the 15 AT alleles must have resolved3' of the Mu1 insertion (position -97; Figure 2B), this experi-ment mapped these break points to a 643-bp interval. Similarly,
Crossovers Resolve at the 5' End of the a7 Gene 2155
the eight A7* alleles must have resolved within this 643-bp in- terval or 5'of position -97 but 3'of the centromere distal EcoRl site that is indicated with an asterisk in Figure 5A.
Because a7-mum2 (and its apparent progenitor, A7-LC) ex- hibits abundant DNA sequence polymorphisms relative to al::rdf, it was possible to map more precisely the positions of recombination break points by sequencing diagnostic por- tions of the 643-bp interval from each of the 15 A7' and eight A7* alleles. These analyses demonstrate that recombination break points do not resolve uniformly across the a7 locus (Fig- ure 6); rather, they cluster within a 377-bp interval of the a7 gene. Fourteen of the 15 recombination break points associated with A7'alleles and four of the eight recombination break points associated with A7' alleles resolved within this 377-bp inter- val (the remaining four recombination break points associated with A7' alleles resolved at undetermined locations 5'of posi- tion -97). Hence, this 377-bp interval exhibits values of p equal to 62 kb/cM and 29 kb/cM in the two experiments (see Methods and Figure 48). In contrast, the overall value of p in the maize genome is 1456 kb/cM (Civardi et al., 1994). Hence, as mea- sured by crosses 1 and 2, this 377-bp recombination hot spot is 23 and 50 times more recombinant than the genome as a whole.
Distinguishing between Reciproca1 Recombination and Gene Conversion
Recombination events can resolve via reciprocal recombina- tion or the related process of gene conversion. Gene conversion occurs via the nonreciprocal transfer of DNA sequences from one nonsister chromatid to another (Figure 1). Therefore, some of the A7' alleles isolated in this study could have arisen via the nonreciprocal transfer of sequence information from one a7 allele to the other so that the region containing a transposon was replaced. Such converted alleles would confer a colored kernel phenotype. However, because in these experiments only A7' alleles that were in coupling with the closely linked sh2 mutant were selected (Figure 2B), only those gene conversions that removed the rdf transposon from the al::rdtsh2 chromo- some would be expected to be recovered.
The genetic markers php70080 and sh2, which are located ~2 and 0.1 cM proximal and distal of the a7 locus, respectively (Figure 2A), were used to distinguish between reciprocal recom- bination events (i.e., single crossovers) and putative gene conversion events. The former would be expected to exhibit an exchange of flanking markers, whereas the latter would not. The a7::df sh2 stock used in crosses 1 and 2 (Figures 2 and 5)
a (position +1162) b
Al- LC or ol-mrmr2 aUele EcoRl -
5' untranscribed
I I 7 ,I 7 11111 I C I O S 1 VI1 V I1 iniervak
I 62kblcM , I I 2 I1 2 1 1 1 I I C I O S 2
VI1 V intewals , 29kbleM I
l4-377-bp h o t s p o t u I bp , Figure 6. Locations of Recombination Break Points Associated with Chimeric Al' and A7' Alleles.
The AI-LC allele is identical in sequence to al-mum2, except for the presence of lhe Mu7 insertion (indicated by dashes) in the latter. The vertical bars on the schematic diagrams of the a7 alleles represent sequence polymorphisms between the a7-mum2 (or A1-LC) and a1::rdt alleles that were used to define the intervals to which recombination break points of Al' and Al ' alleles were mapped. The bar widths are proportional to the number of base pairs associated with lhe indicated sequence polymorphisms. The locations and sequence coordinates of Pstl sites are shown. The Pstl site that is present in a1::rdt but absent in a7-mum2, and A1-LC is indicated by an asterisk. As depicted here, lhe centromere and the sh2 locus lie to the left and right of lhe a7 locus, respectively (see Figure 2A). The number of break points from crosses 1 and 2 associated with A l ' and A l ' alleles that map to the various intervals are indicated on the horizontal bars labeled cross 1 and cross 2.
2156 The Plant Cell
B
2.0kb
1.3kb
0.8 kb
0.63 kb
0.38 kb
0.22 kb
Figure 7. Mapping Recombination Break Points within AT and A1'Alleles Using the Diagnostic Pstl Site.(A) Recombinant alleles were PCR amplified (using primers XX026and XX025; Figure 6) from genomic DMA isolated from plants with thegenotype AT/a1::rdtor A1'/aT.:rdt. Agarose gel electrophoresis revealedone to three PCR products from each reaction. The expected 1.3-kba7-hybridizing product, derived from/U'or/U* alleles, was detectablein all reactions (represented by alleles AT-276, AT-102, and AT-275).The 2.0-kb a7-hybridizing product, derived from aT.-.rdt, was detect-able in most reactions (represented by allele AT-276). Some reactionsyielded an 0.8-kb nonspecific PCR product that did not cross-hybridizewith an a7-specific probe (represented by alleles AT-27B and AT-102).The DNA gel blots involving the a7-specific hybridization probe arenot shown. Numbers at right indicate the size of each individual frag-ment in kilobases.(B) The 1.3-kb PCR products derived from each of the recombinantalleles were gel purified and subjected to Pstl digestion and electropho-resis. In each of the 23 AT and AT alleles, the diagnostic 631-bpfragment was recovered (see Methods).
double crossovers. The probability of isolating twop/)p700SO-a7crossovers from a sample of 17 gametes is small (even ignor-ing interference). Therefore, it is most likely that these two ATalleles arose via gene conversion events. In these two recom-bination events, the centromere distal ends of the conversiontracts map to intervals V and VII (Figure 6) and extend the cen-tromere proximally to an undetermined location beyond therdt insertion at position +1083 of the a? gene.
DISCUSSION
Recombination Break Points Cluster in the 5' End of aCoding Region
Recombinants were isolated within a recombination hot spotvia two approaches. In the first approach (cross 1), it was pos-sible to select intragenic recombination events directly (ATalleles) by virtue of the fact that they conditioned a nonparen-tal phenotype at the a1 locus. Based on the results of cross1, a 1.2-kb interval of the a1 locus has a genetic distance of0.0065 cM (±0.0009 cM) and has a value of p equal to 184
I
was segregating for two alleles of the RFLP marker php10080.The inbred lines Ac1068 and Ac1069 carry these two allelesof phpWOSO. The 4.5- and 16-kb p/7p70080-hybridizing frag-ments shown in Figure 8 represent the php10080-Ac1068 andphplo080-Ac1069 alleles, respectively. The a1-mum2 stock car-ries a phpWOSO allele (the 9-kb hybridizing fragment) that isdistinguishable from both of those carried in the a1::rdt sh2stock. Hence, a1-mum2laT.:rdt heterozygotes have one oftwo possible RFLP patterns (Figure 8). Among AT alleles, anRFLP pattern that includes the 9-kb fragment indicates thata particular AT allele arose via reciprocal recombination (seeFigure 8, AT-275 and AT-279). In contrast, an RFLP patternthat lacks the 9-kb hybridizing fragment indicates that a par-ticular/47' allele arose via gene conversion or a very rare doublecrossover.
Two of the three possible RFLP configurations are illustrated(see Figure 8, Ar-274 and AT-276). The third possible RFLPconfiguration (only the 16-kb fragment) was not recovered inthis small sample. These analyses demonstrated that two ofthe 17X47'alleles (A1'-274 and XW-276) are in coupling with thea7::rdf-derived alleles of both php10080 and sh2 (Figure 8).Thus, these two AT alleles arose via gene conversion events or
5<N
R<N
P ' + «•
16kb
9kb
4.5kb
Figure 8. Distinguishing between Reciprocal Recombination and GeneConversion Events via RFLP Analysis.DNA samples from plants with the indicated genotypes or stocks weredigested with Hindlll, transferred to nylon membrane, and hybridizedwith a php70080-specific probe. Similar results were obtained in DNAgel blots after digestion with EcoRI. Numbers at right indicate the sizeof each individual fragment in kilobases.
Crossovers Resolve at the 5‘ End of the a7 Gene 2157
kb/cM (Figure 46). In the second approach (CrOSS 2), initially progeny were preselected that carried a recombination break point between the a1 and sh2 loci, a distance of 140 kb (Civardi et al., 1994). Subsequently, recombination break points that map to a 6-kb interval that includes the a7 locus (A7’ alleles) were identified by virtue of the fact that recombination within this interval generates novel RFLPs that hybridize to al. Based on this experiment, this 6-kb interval has a genetic distance of 0.026 cM and a value of p equal to 231 kb/cM (Figure 4A). Hence, the rates of recombination per kilobase within this 6-kb interval are sixfold higher than the genome average of 1456 kb/cM (Civardi et al., 1994). In contrast, the 134-kb interval that composss the remainder of the a7-sh2 interval has a value of p that is higher than the genome average (Figure 4A). These results extend an earlier study (Civardi et al., 1994), which sug- gested that recombination hot spots are discrete intervals adjacent to intervals that are substantially less recombination- ally active.
Subsequent mapping using DNA sequence polymorphisms established that recombination break points isolated from the two crosses did not resolve randomly across the a7 gene; 14 of 15 recombination events from cross 1 resolved within a 377- bp interval of the a7 gene. Similarly, approximately one-fifth (four of 21) of all recombination events from cross 2 that oc- curred within the 140-kb a7-sh2 interval mapped to this same 377-bp recombination hot spot. Within this 377-bp interval, the value of p was measured as 62 and 29 kb/cM in crosses 1 and 2, respectively (Figure 46). This interval is therefore 23 and 50 times more recombinant than the genome as a whole (as measured via crosses 1 and 2, respectively).
This 377-bp recombination hot spot is located near the 5’ end of the transcribed region of thea l gene. While this article was under review, Patterson et al. (1995) and Eggleston et al. (1995) published restriction mapping data that demonstrate that recombination events resolve at high rates in the 5’and 3’ends of the booster (b7) and r l loci of maize, respectively. In yeast, the meiotic DSBs that are thought to initiate recombination are generally located 5’ of genes, near transcription promoter sequences (Wu and Lichten, 1994). If this is true in plants, it therefore appears that although the sites at which recombina- tion events initiate and resolve may be different, both appear to be spatially regulated in relation to transcribed sequences.
The Mul Transposon Does Not Alter the Spatial Distribution of Recombination Break Points
These results extend to Mu7 the finding that the Dissociation (Ds) transposons (Dooner, 1986) have an inhibitory effect on recombination rates. In the present experiment, the 1.4-kb Mu7 insertion present in al-mum2 (used in cross l), but absent in A7-LC (used in cross 2). reduced recombination rates in a nearby 377-bp interval by 60%. The results from crosses 1 and 2 are directly comparable, because theA7-LC and al-mum2 alleles are identical in sequence except for the presence of the Mu7 insertion in the latter allele. However, because most
recombination break points isolated via the two approaches mapped to the same 377-bp interval, it can be concluded that although the Mo7 transposon insertion in the a7-mum2 allele suppresses recombination rates, it does not alter the distribu- tion of recombination break points within the a7 gene. Because all crosses in which recombination was measured involved the a7::rdt allele, we cannot draw conclusions regarding the ef- fects the rdt transposon insertion may have on recombination.
In yeast, it has been shown that DNA synthesis associated with recombination results in mutations (Strathern, 1995). How- ever, all sequenced regions of the A7‘ and A7* alleles were identical to either al-mum2 (orA7-LC) or akrdt, that is, recom- bination was precise (data not shown). However, given the low rate at which unrepaired recombination-related nucleotide mis- incorporation errors occur in yeast (between 10-5 and 10-6 per base; Strathern, 1995), the failure to uncover any novel se- quence alterations in the vicinity of the 19 maize recombination break points analyzed in this study may be the result of inade- quate sampling.
Gene Conversions and Reciproca1 Recombination Events Resolve in the Same Physical lnterval
Two of the A7’ recombinants isolated from cross 1 arose in the absence of exchange of flanking markers @hp70080 and sh2), suggesting that they arose via gene conversion. Gene conver- sions can occur in conjunction with @e., flanking) a crossover or independent of a crossover (Figure 1). Conversions occur- ring in conjunction with crossovers would not be detectable in plants (because in plants it is not usually possible to recover as a unit all the products of a single meiosis); such conver- sions would merely shift the apparent position of the associated recombination break point. In contrast, plant gene conversion events that are not associated with crossovers would have all of the characteristics of a DCO. Although it is not possible to prove formally that these two recombination events represent gene conversions, the probability that these alleles arose via DCOs is remote.
The 377-bp Hot Spot May Block Holliday Junction-Associated Branch Migration or 5’ to 3’ Exonuclease Activity
Based on the current models of recombination that explain gene conversion (Szostak et al., 1983), the DSBs that initiated these two gene conversion events must have occurred within the region of the a7 locus that was converted, that is, 3’ of in- terval V (Figure 6). In contrast, data from yeast indicate that the bulk of the DSBs that initiate recombination events occurs 5’of gene coding regions (Wu and Lichten, 1994). If the DSBs that initiate maize recombination events also occur 5’ of gene coding regions, some of the DSBs that led to the reciproca1 recombination events reported in this study probably occurred 5’ of the a7 recombination hot spot. Hence, because the two
2158 The Plant Cell
conversion events initiated via DSBs 3' of the recombination hot spot in the a7 gene, these data suggest that DSBs that occur in spatially distinct intervals can resolve within the same recombination hot spot. In addition, these results suggest that the recombination hot spot identified in this study has some feature that terminates the process of branch migration or 5' to 3'exonuclease activity (Figure 1) from both directions and thereby constrains the sites at which Holliday Junctions resolve.
METHODS
Gene Symbols, Allele Descriptions, and Genetic Stocks
The a7 locus encodes the enzyme dihydroquercetin reductase (EC 1.1.1.219), which is required for the biosynthesis of anthocyanin pig- ments that color the aleurone layer of maize kernels (Reddy et al., 1987). The closely linked shrunken2 (sh2) locus (Figure 2A) encodes ADP-glu- cose pyrophosphorylase (EC 2.7.7.27), which is involved in starch biosynthesis (Tsai and Nelson, 1966). As shown in Figure 3, kernels that do not carry wild-type alleles at these loci exhibit colorless and shrunken phenotypes, respectively.
The recessiveal-mum2 allele contains a 1.4-kb Mutatorl (Mul) trans- poson insertion at nucleotide -97, and the recessive a7::rdt allele contains a 0.7-kb rdf transposon insertion at nucleotide +1083 (OReilly et al., 1985; Shepherd et al., 1988; Brown et al., 1989; positions are relative to the start of transcription in the A7-LC allele). The recessive a7::rdt and a7-mum2 alleles both condition a colorless phenotype in the absence of trans-acting regulatory transposons (Dotted [Dt] and MuDR, respectively). The stocks used in this study do not carry Dt or MuDR. The al-mum2 stock without MuDR is derived from that de- scribed by Schnable and Peterson (1989). The a7::rdt and sh2 marker alleles used in this study were extracted from a commercially avail- able Asgrow sweet corn line (Civardi et al., 1994) and the F, hybrid Sweet Belle. This hybrid and its two parenta1 inbreds, Ac1068 and Ac1069, were gifts from the Asgrow Seed Company (Kalamazoo, MI). Ac1068 and Ac1069 carry distinguishable alleles at the php70080 lo- cus @hp7008O-Ac7068 and php7008O-Ac7069). This fact explains the observation that our Sweet Belle-derived a7::rdt sh2 stock is hetero- geneous for these two php70080 alleles. The a7-dl allele is a stable recessive allele obtained in coupling with etchedl (et7) from I? Peterson (lowa State University, Ames).
DNA Isolation, DNA Gel Blot Analyses, and DNA Probes
Maize DNA was isolated from immature plant leaves, as described by Saghai-Maroof et al. (1984). DNA gel blots were conducted via stan- dard protocols (Sambrook et al., 1989). The php70080-, a7-, and sh2-specific probes were isolated from p-php10080, pALC2 (Schwarz- Sommer et al., 1987), and pSh2-850 (or pSh2-1000) derived from a sh2 cDNA clone (Bhave et al., 1990). These clones were gifts from Pioneer Hi-Bred lnternational Inc. (Johnston, IA), Alfons Gierl (Tech- nische Universitat, Munich, Germany), and Curt Hannah (University of Florida, Gainesville), respectively.
Sequencing the al-mum2 and al::rdt Alleles
An "old lysate of the phage clOMu, which contains a 7.8-kb EcoRl frag- ment that spans the a7-mum2 allele sequence (OReilly et al., 1985),
was generously provided by Alfons Gierl (Technische Universitat). Un- fortunately, this lysate no longer contained viable phage. To rescue the al-mum2 allele, phage DNA was extracted from the lysate, digested with Notl and EcoRI, and subcloned into pUC9 (Vieira and Messing, 1982). The 3.0- and 4.8-kb subclones were termed pYENl and pYEN2, respectively. The a7::rdt allele was obtained as a 10-kb EcoRl frag- ment in the pUC9 vector and was a generous gift from N. Shepherd (DuPont, Wilmington, DE). This clone was termed pElO. The DNA se- quence of the 1.2-kb interval of the a7-mum2 allele that is defined by the two transposon insertion sites was found to be identical to the cor- responding interval of the existing sequence of the A7-LC allele (GenBank accession number X05068). The sequence of the corre- sponding interval from the al::rdt allele was obtained and submitted to GenBank (accession number U23161). All sequences are based on at least three sequencing reactions. Sequencing was performed at the lowa State University Nucleic Acid Facility on an ABI 373A auto- mated DNA sequencer (Applied Biosystems, Foster City, CA) using dideoxy terminators.
Genetic Crosses
Cross 1 is al-mum2 Sh2lal::rdt sh2 x a7::rdt sh2lal::rdt sh2. Cross 2 is A7-LC Sh2lal::rdt sh2 x a7::rdt sh2lal::rdt sh2.
Generating lntragenic Recombinants
Rare intragenic a7 recombination events were selected from crosses 1 and 2. Cross 1 was conducted by planting the two parents in near isolation from other maize pollen sources at the Ross and lden re- search farms (Ames, IA) during the 1991 and 1993 summer seasons. To conduct this cross, the female parent (which is listed first) of cross 1 was detasseled (emasculated) before anthesis. Therefore, the seed produced by the female parents was generated via cross 1. Most ker- nels isolated from this cross would be expected to have the genotype of either al-mum2 Sh2lal::rdt sh2 or a7::rdt sh2lal::rdt sh2. These geno- types would condition colorless, round and colorless, shrunken kernel phenotypes, respectively. However, rare intragenic recombination events at the a7 locus that occur within the 1.2-kb interval between the two transposon insertion sites in the a7 gene can generate chimeric A7' alleles (Figure 26) that condition colored kernels (Figure 3). The intra- genic recombinants isolated from the 1991 experiment were reported by Civardi et al. (1994).
The recombination events between the A7-LC Sh2 and a7::rdt sh2 chromosomes analyzed in this report were isolated by Civardi et al. (1994). Briefly, cross 2 was conducted via controlled hand pollinations performed in our genetics nursery in 1992 at the Curtiss research farm (Ames, IA). Putative recombinants were selected as colored, shrunken and colorless, round kernels.
The recombination break point associated with each recombinant was mapped relative to the first EcoRl site 5'of the A7-LC allele, using DNA gel blot analyses in which an al-specific probe was utilized. This EcoRl site is marked with an asterisk in Figure 5. Recombination break points within the 134-kb interval defined by the marked EcoRl site and the sh2 locus would be expected to yield al-hybridizing restriction frag- ments identical to those of Al-LC or a7::rdt. In contrast, break points within the 6-kb interval defined by the 700-bp rdt insertion in a7::rdt (position +1083) and the marked EcoRl site would be expected togener- ate nove1 restriction fragments. The al-hybridizing EcoRl restriction fragments associated with the A l ' sh2 and al'Sh2 chromosomes would be expected to be 700 bp smaller than those associated with al::rdt and 700 bp larger than those associated with Al-LC, respectively. Using
Crossovers Resolve at the 5' End of the a7 Gene 2159
this method, we identified those recombinants that potentially arose via intragenic recombination within the a7 locus.
traction: 0.09 - 0.026 = 0.064 cM. The genetic distance of the 377-bp interval was calculated as 418 (number of Al' alleles with break points in the 377-bp interval divided by the number tested) x 0.026 (the genetic distance of the 6-kb interval) = 0.013 cM.
Confirmation of Al' and Al' Alleles
Polymerase Chain Reaction Amplification of Recombinant Al ' and Al' Alleles Genetic crosses and DNA gel blotting experiments were used to con-
firm the validity of most of the putative Al' intragenic recombinants from cross 1 (seven colored, shrunken kernels did not germinate and therefore could not be tested). Colored, shrunken kernels from cross 1 (presumed genotype being AI' sh2/al::rdt sh2) were crossed by al- dl Sh2/al-d/ Sh2. Half of the progeny kernels from this cross were ex- pected to be colored and round. The other half were expected to be colorless and round. Plants derived from colored, round kernels were self-pollinated. Self-pollinated progenies that segregated in a t2: l ra- tio of colored, shrunken:colored, round:colorless, round kernels were deemed to carry legitimate Al' alleles. As a further test, plants carry- ing the A7' alleles were subjected to DNA gel blot analysis with the RFLP marker phpl0080 (data not shown). Because the php70080 lo- cus is c\ose\y Iinked to the a l locus (Figure 2A), only those A?' alleles that were in coupling with one of the parenta1 alleles of phpl0080 were considered to be valid recombinants.
The confirmation procedure for the progeny of cross 2 was described previously by Civardi et al. (1994). As a further confirmation, putative recombinants from cross 2 were subjected to DNA gel blot analyses using php70080-specific and sh2-specific probes (data not shown). Only those progeny that exhibited appropriate alleles of these loci were considered to be valid recombinants.
Calculations of Genetic Distances
Al'alleles arise in cross 1 via one of the two possible intragenic recom- bination events (see Figure 26). The reciproca1 recombination event generates a7 alleles that carry both transposon insertions and thus do not condition distinctive phenotypes, that is, they condition color- less kernels. Because only one of the two possible classes of recombination events could be identified in the case of cross 1, the genetic distance associated with the 1.2-kb interval was calculated by doubling the rate at which colored, shrunken kernels were recovered (2 x 24/742,100 = 0.0065 cM). This calculation assumes that the two events occur at an equal frequency. Standard errors were calculated according to the equation (pq/n)1'2, where p is the map distance in centimorgans (cM), q is (100 - p), and n is the number of kernels from the testcross that were scored. To calculate the genetic distance associated with the 377-bp interval as measured by cross 1. the propor- tion Of Al'alleles whose recombination break point mapped to the 377-bp interval was multiplied by the genetic distance of the 1.2-kb interval
The collection of recombinants derived from cross 2 has been used to estimate the genetic distance between the a7 and sh2 loci as 0.09 -c 0.01 cM (Civardi et al., 1994). Because not all of these recombinants were used in the present study, estimates of genetic distances for phys- ical intervals smaller than the 140-kb al-sh2 interval were determined by multiplying the proportion of tested recombinants with a recombi- nation break point in the interval of interest by the genetic distance associated with the next larger interval. To illustrate, the genetic dis- tance of the 6-kb interval was calculated as 16/56 (number of Al' and al' alleles divided by the number tested) x 0.09 cM (the genetic dis- tance of the entire al-sh2 interval) = 0.026 cM. The genetic distance of the remainder of the 140-kb al-sh2 interval was calculated by sub-
(14/15 x 0.0065 CM = 0.0061 cM).
Template DNA was obtained from plants derived from colored, shrunken kernels that carried Al'and Al ' alleles and that were isolated via crosses 1 and 2. The Al' and A7* alleles from each sample were polymerase chain reaction (PCR) amplified using two primers (XX025 and XX026) that flank the 1.2-kb interval defined by the two transposon insertion sites. The 5' ends of primers XX025 and XX026 are located at posi- tions -122 and +1162 in the a7 gene, just upstream of the Mul insertion site in the al-mum2 allele and just downstream of the rdt insertion site in the al::rdr allele, respectively. The sequences of the two primers are as follows: XX025, 5'-GGTAGTTGCAGCGTGTGGTGTT-3'; XX026,
All PCR primers were synthesized on a 394 DNA/RNA synthesizer from Applied Biosystems at the lowa State University Nucleic Acid Fa- cility. PCR amplifications were conducted for 40 cycles as follows: 94OC for 2.5 min, 63OC for 1 min, and 72% for 1.5 min.
5'-GAGGTCGTCGAGGTGGATGAGCTG-3!
Mapping Al' Recombination Break Points Relative to the Diagnostic Pstl Site
DNAsequence analyses predict the presence and absence of a diag- nostic Pstl recognition site in the a7::rdf and the al-mum2 (and Al-LC) alleles, respectively (Figure 6). The presence of this diagnostic Pstl site in the a7::rdtallele used in this experiment was confirmed by PCR amplifying this allele from the sweet corn hybrid Sweet Belle (our source of the al::rdtallele), using primers XX025 and XX026, and then sub- jecting the resulting PCR product to Pstl digestion.
TheA7'and Al ' alleles were PCR amplified from DNA isolated from plants with the genotype Al'/al::rdt or Al*/al::rdt (two of the 17 Al'al- leles were not successfully amplified). Ths PCR primers XX025 and XX026 amplify the entire 1.2-kb interval between the Mul and rdt trans- poson insertion sites (Figure 6). Total PCR were subjected to electrophoresis through Tris acetate EDTA agarose gels (Figure 7A). DNA gel blot analyses using an al-specific hybridization probe con- firmed that the resulting 1.3-kb PCR product was derived from the a7 locus (data not shown). The PCR products derived from the A7' and Al' alleles were excised from duplicate gels and purified using the Gene-Clean kit (Bio-101, Inc., Midwest Scientific, St. Louis, MO). The 1.3-kb PCR products derived from each of the Al'and Al* alleles were gel purified and subjected to Pstl digestion and electrophoresis (Fig- ure 78).
Because the 1.2-kb interval of the a7::rdr allele contains a centrally located Pstl site that is absent from the corresponding interval of the a7-mum2 and Al-LC alleles (marked with an asterisk in Figure 6), it was possible to map each recombination break point relative to this restriction enzyme site. Three of the predicted Pstl digestion products (-381, 220, and 52 bp) are not informative. However, if a given recom- bination break point occurred between the Mul insertion and this diagnostic Pstl site, the resulting digest should include an -631-bp fragment. Alternatively, if the recombination break point was located between the rdt insertion and this Pstl site, then the Pstl restriction enzyme should cleave the 631-bp fragment into two fragments of ~ 3 8 9 and 242 bp.
2160 The Plant Cell
Sequencing the Al’ and Al’ Alleles
Purified PCR products (derived from primers XX026 and XX025; see Figures 6 and 7) were used directly for sequencing; sequencing results are based on at least two reactions. The primers used for sequencing (and their positions within the a7 gene) are as follows: XX653,5‘-CGA-
TGATTCGCTCCGTG-3’ (260 to 282); and XX390,5’-TCGGCTTGATTA- CCTCATTCT-3’ (604 to 584).
GCCAGGAGCCGACGAAG-3’ (163 to 144); XX231, 5‘-GCCAAACTC-
Sequence Analyses
Sequence analyses were performed using Version 7 of the Genetics Computer Group (Madison, WI) software.
ACKNOWLEDGMENTS
We thank members of the Schnable laboratory and Joel Hansen Of
the Nikolau laboratory for assistance with planting, detasseling, and harvesting the isolated crossing plots. We also thank Yiji Xia, John VanDiepen, Kevin Van Dee, and Weijun Chen for providing additional technical assistance; Homer Caton of AgriPro Seeds (Ames, IA) for providing access to maize ear-drying facilities; and John lmsande for critical review of the manuscript. P.S.S. and B.J.N. were supported by grants from the Midwest Plant Biotechnology Consortium, with match- ing funds from Pioneer Hi-Bred International, Inc. and Cargill Hybrid Seeds. X.J.X., L.Z., and A-P.H. are students in the lowa State Univer- sity Molecular, Cellular and Developmental Biology (X.J.X. and L.Z.), and lnterdepartmental Genetics (A-P.H.) graduate programs. This is Journal Paper No. J-16288 of the lowa Agriculture and Home Economics Experiment Station (Ames, IA). Project No. 3125.
Received July 27, 1995; accepted October 16, 1995.
REFERENCES
Bhave, M.R., Lawrence, S., Barton, C., and Hannah, L.C. (1990). ldentification and molecular characterization of shrunken-2 cDNA clones of maize. Plant Cell 2, 581-588.
Brown, J., and Sundaresan, V. (1991). A recombination hotspot in the maize A7 intragenic region. Theor. Appl. Genet. 81, 185-188.
Brown, J.J., Mattes, M.G., OReilly, C., andShepherd, N.S. (1989). Molecular characterization of rDt, a maize transposon of the “Dot- red” controlling element system. MOI. Gen. Genet. 215, 239-244.
Cao, L., Alani, E., and Kleckner, N. (1990). A pathway for genera- tion and processing of double-strand breaks during meiotic recombination in S . cerevisiae. Cell 61, 1089-1 101.
Civardi, L., Xia, Y.J., Edwards, K.J., Schnable, P.S., and Nikolau, B.J. (1994). The relationship between genetic and physical distances in the cloned al-sh2 interval of the Zea mays L. genome. Proc. Natl. Acad. Sci. USA 91, 8268-8272.
Dooner, H.K. (1986). Genetic fine structure of the bronze locus in maize. Genetics 113, 1021-1036.
Dooner, H.K., and Kermicle, J.L. (1986). The transposable element Ds affects the pattern of intragenic recombination at the Bz and R loci in maize. Genetics 113, 135-143.
Dooner, H.K., and Ralston, E.J. (1990). Effect of the Mul insertion on intragenic recombination at the bz locus in maize. Maydica 35, 333-337.
Dooner, H.K., Weck, E., Adams, S., Ralston, E., Favreau, M., and English, J. (1985). A molecular genetic analysis of insertions in the bronze locus in maize. MOI. Gen. Genet. 200, 240-246.
Dooner, H.K., Keller, J., Harper, E., and Ralston, E. (1991). Vari- able patterns of transposition of the maize element Activator in tobacco. Plant Cell 3, 473-482.
Dorer, D.R., and Christensen, A.C. (1 989). A recombination hotspot at the triplo-lethal locus of Drosophila melanogaster, Genetics 122,
Eggleston, W.B., Alleman, M., and Kermicle, J.L. (1995). Molecu- lar organization and germina1 instability of R-stippled maize. Genetics
Freeling, M. (1978). Allelic variation at the leve1 of intragenic recombi- nation. Genetics 89, 211-224.
Goldway, M., Sherman, A., Zenvirth, D., Arbel, T., and Simchen, G. (1993). A short chromosomal region with major roles in yeast chro- mosome 111 meiotic disjunction, recombination and double strand breaks. Genetics 133, 159-169.
Janson, M., Larsson, C., Werelius, B., Jones, C., Glaser, T., Nakamura, Y., Jones, C.P., and Nordenskjold, M. (1991). Detailed physical map of human chromosomal region llq12-13 shows high meiotic recombination rate around the MfN l locus. Proc. Natl. Acad. Sci. USA 88, 10609-10613.
Korol, A.B., and Iliad, K.G. (1994). lncreased recombination frequen- cies resulting from directional selection for geotaxis in Drosophila. Heredity 72, 64-68.
Lambie, E.J., and Roeder, G.S. (1986). Repression in meiotic cross- ing over by a centromere (CEN3) in Saccharomyces cerevisiae. Genetics 114, 769-789.
Lichten, M., and Goldman, A.S.H. (1995). Meiotic recombination hot- spots. Annu. Rev. Genet., in press.
Moore, G., Gale, M.D., Kurata, N., and Flavell, R.B. (1993). Molecu- lar analysis of small grain cereal genomes: Current status and prospects. Bio/Technology 11, 584-589.
Nag, D.K., and Petes, T.D. (1993). Physical detection of heteroduplexes during meiotic recombination in the yeast Sacchammyces cerevisiae. MOI. Cell. Biol. 13, 2324-2331.
Nelson, O.E. (1968). The waxy locus in maize. 11. The location of the controlling element alleles. Genetics 60, 507-524.
Nicolas, A., Treco, D., Schultes, N.P., and Szostak, J.W. (1989). An initiation site for meiotic gene conversion in the yeast Sacchammyces cerevisiae. Nature 338, 35-39.
OReilly, C., Shepherd, N.S., Pereira, A., Schwarz-Sommer, Z., Bertram, I., Robertson, D.S., Peterson, P.A., and Saedler, H. (1985). Molecular cloning of the a7 locus of Zea mays using the trans- posable elements f n and Mul. EMBO J. 4, 877-882.
Patterson, G.I., Kubo, K.M., Shroyer, T., and Chandler, V.L. (1995). Sequences required for paramutation of the maize b gene map to a region containing the promoter and upstream sequences. Genetics
Ponticelli, A.S., Sena, E.P., and Smith, G.R. (1988). Genetic and physical analysis of the M26 recombination hotspot of Schizosac- charomyces pombe. Genetics 119, 491-497.
397-401.
141, 347-360.
140, 1389-1406.
Crossovers Resolve at the 5' End of the a7 Gene 2161
Reddy, A.R., Britsch, L., Salamini, F., Saedler, H., and Rohde, W. (1987). The A1 (Anthocyaninl) locus in Zea mays encodes dihydro- quercetin reductase. Plant Sci. 52, 7-13.
Roberts, P.A. (1965). Difference in the behavior of Eu and hetero- chromatin: Crossing-over. Nature 205, 725-726.
Sachs, M., Dennis, E., Gerlach, W., and Peacock, W.J. (1986). Two alleles of the maize Adhl have 3'structural and poly(A) addition poly- morphisms. Genetics 113, 449-467.
Saghai-Maroof, M.A., Soliman, K.M., Jorgensen, R.A., and Allard, R.W. (1984). Ribosomal DNA spacer-length polymorphisms in bar- ley: Mendelian inheritance, chromosomal location, and population dynamics. Proc. Natl. Acad. Sci. USA 81, 8014-8018.
Sambrook, J., Fritsch, E.F., and Manlatis, T. (1989). Molecular Clon- ing: A Laboratory Manual, 2nd ed. (Cold Spring Harbor, NY Cold Spring Harbor Laboratory Press).
Schnable, P.S., and Peterson, P.A. (1989). Genetic evidence of a rela- tionship between two maize transposable element systems: Cy and Mutator: MOI. Gen. Genet. 215, 317-321.
Schuchert, P., Langsford, M., Kaslin, E., and Kohli, J. (1991). A spe- cific DNA sequence is required for high frequency of recombination in the ade6 gene of fission yeast. EMBO J. 10, 2157-2163.
Schultes, N.P., and Srostak, J.W. (1990). Decreasing gradients of gene conversion on both sides of the initiation site for meiotic recom- bination at the ARG4 locus in yeast. Genetics 126, 613-1322.
Schwarz-Sommer, Z., Shepherd, N., Tacke, E., Gierl, A., Rohde, W., Leclercq, L., Mattes, M., Berndtgen, R., Peterson, P.A., and Saedler, H. (1987). lnfluence of transposable elements on the struc- ture and function of the A7 gene of Zea mays. EMBO J. 6,287-294.
Shepherd, N.S., Sheridan, W.F., Mattes, M.G., and Deno, G. (1988). The use of Mofator for gene tagging: Cross-referencing between transposable element systems. In Plant Transposable Elements, O. Nelson, ed (New York: Plenum Press), pp. 137-148.
Shiroishi, T., Sagai, T., and Moriwaki, K. (1993). Hotspots of meiotic recombination in the mouse major histocompatibility complex. Genet- ica 88, 187-196.
Stahl, F.W. (1979). Special sites in generalized recombination. Annu. Rev. Genet. 13, 7-24.
Steinmetz, M., Uematsu, Y., and Lindahl, K.F. (1987). Hotspots Of
homologous recombination in mammalian genomes. Trends Genet.
Strathern, J.N., Shafer, B.K., and McGill, C.B. (1995). DNA synthe- sis errors associated with double-strand-break repair. Genetics 140,
Sun, H., Treco, D., Schultes, N.P., and Szostak, J.W. (1989). Double- strand breaks at an initiation site for meiotic gene conversion. Na- ture 338, 87-90.
Symington, L.S., Brown, A., Oliver, S.G., Greenwell, P., and Petes, T.D. (1991). Genetic analysis of a meiotic recombination hotspot On chromosome 111 of Sacchammyces cerevisiae. Genetics 128,717-727.
Szostak, J.W., Orr-Weaver, T.L., Rothstein, R.J., and Stahl, F.W. (1983). The double-strand break repair model for recombination. Cell
Tanksley, S.D., Ganal, M.W., Prince, J.P., devicente, M.C., Bonierbale, M.W., Broun, P., Fulton, T.M., Giovannoni, J.J., Grandillo, S., Martin, G.B., Messequer, R., Miller, J.C., Miller, L., Paterson, A.H., Pineda, O., Roder, M.S., Wing, R.A., Wu, W., and Young, N.D. (1992). High density molecular linkage maps of the tomato and potato genomes. Genetics 132, 1141-1160.
Uematsu, Y., Kiefer, H., Schulze, R., Lindahl, K.F., and Steinmetz, M. (1986). Molecular characterization of a meiotic recombinational hotspot enhancing homologous equal crossing-over. EMBO J. 5,
Vieira, J., and Messing, J. (1982). The pUC plasmids, an M13mp7- derived system for insertion mutagenesis and sequencing with syn- thetic universal primers. Gene 19, 259-268.
Wessler, S.R., and Varagona, M.J. (1985). Molecular basis of muta- tions at the waxy locus of maize: Correlation with fine structure genetic map. Proc. Natl. Acad. Sci. USA 82, 4177-4181.
Wu, TC, and Lichten, M. (1994). Meiosis-induced double-strand break sites determined by yeast chromatin structure. Science 263,515-517.
3, 7-10.
965-972.
33, 25-35.
2123-2129.
DOI 10.1105/tpc.7.12.2151 1995;7;2151-2161Plant Cell
X Xu, A P Hsia, L Zhang, B J Nikolau and P S SchnableMeiotic recombination break points resolve at high rates at the 5' end of a maize coding sequence.
This information is current as of February 13, 2020
