The ade6-M26 mutation of Schizosaccharomyces pombe created a meiotic recombination hotspot. Previous analyses indicated that the heptamer 5'-ATGACGT-3' was necessary and sufficient for hotspot activity; the Atf1-Pcr1 transcription factor binds to this sequence and activates M26. After finding cases in which the M26 heptamer in ade6 was, surprisingly, not active as a hotspot, we used an in vitro selection method (SELEX) that revealed an 18-bp consensus sequence for Atf1-Pcr1 binding, 5'-GNVTATGACGTCATNBNC-3', containing the M26 heptamer at its core. Using this consensus sequence as a guide, we made mutations on each side of the heptamer at two separate sites in ade6. These mutations increased the intracellular hotspot activity of the heptamer, in some cases by >15-fold. These results show that M26, the eukaryotic recombination hotspot with the most precisely defined nucleotide sequence, is larger than previously thought, and they provide valuable information for clarifying the role of M26, and perhaps other hotspots, in meiotic recombination.

---

MEIOSIS is a special form of cell division in sexually reproducing organisms in which one round of DNA replication is followed by two successive cell divisions, resulting in four haploid products (gametes or spores). The first division of meiosis is termed a reductional division, because homologous chromosomes (i.e., maternal and paternal homologs) are segregated to opposite poles, halving the number of chromosomes in each daughter cell. Prior to the first division, homologous chromosomes recombine with each other at a greatly elevated frequency compared to that during mitosis (ESPOSITO and KLAPHOLTZ 1981). This recombination frequently results in crossovers, or chiasmata, which are important in most organisms for the proper segregation of homologs at the first meiotic division (BAKER et al. 1976). Crossovers also result in the shuffling of genes between homologs and are thereby an important mechanism of increasing genetic diversity within a species.

Meiotic recombination events are not distributed evenly throughout the genome of most organisms. Rather, there are hotspots of recombination, "short segment[s] of chromosome with a conspicuously higher than average rate of recombination" (STAHL 2002, p. 976). In the two distantly related yeasts Saccharomyces cerevisiae and Schizosaccharomyces pombe, hotspots of recombination coincide with sites of programmed double-strand DNA breaks (DSBs) (SUN et al. 1989; FAN et al. 1995; PETES 2001; STEINER et al. 2002). The broken ends of the DNA initiate recombination by invading intact homologous DNA to form joint molecules, which can then be resolved to produce gene conversions (nonreciprocal exchanges) and crossovers (reciprocal exchanges) (for a review see PâQUES and HABER 1999).

The nucleotide sequences determining meiotic recombination hotspots are incompletely understood, but two classes of hotspots have been recognized (PETES 2001). -Hotspots require the binding of transcription factors, whereas ß-hotspots do not require the binding of known transcription factors. Most or all ß-hotspots are generated by insertions of foreign DNA (KIRKPATRICK et al. 1999a,b), which are likely to disrupt the native chromatin structure. GERTON et al. (2000) showed that meiotic DNA break sites in S. cerevisiae are associated with regions of elevated G + C content—at least 3% above the genomic average when viewed in moving 5-kb windows. Viewing smaller 1-kb windows, BLUMENTAL-PERRY et al. (2000) reported a correlation of DNA break sites with a 50- to 250-bp degenerate motif [termed the common homology region (CoHR)] that is low in G + C content and contains a central poly(A) tract. Current evidence suggests, however, that the CoHR motif does not accurately predict the location or occurrence of DSBs (HARING et al. 2004).

The M26 hotspot of S. pombe is the meiotic recombination hotspot whose sequence requirements have been most precisely defined at the nucleotide level. This hotspot was first identified as an allele of the ade6 gene that produced an unusually high frequency of adenine prototrophs in meiotic crosses with other ade6 alleles (GUTZ 1971). This hotspot results from a single G T nonsense mutation near the 5'-end of ade6 (SZANKASI et al. 1988 and Figure 1). A systematic mutational analysis of nucleotides neighboring this mutation indicated a 7-bp sequence, 5'-ATGACGT-3' (M26 mutation underlined), to be necessary for hotspot activity (SCHUCHERT et al. 1991). The ade6-M26 mutation creates a binding site for a heterodimeric transcription factor, Atf1-Pcr1, that is also essential for its hotspot activity; in atf1 or pcr1 mutants the M26 hotspot is inactive (WAHLS and SMITH 1994; KON et al. 1997).

View larger version (33K): In this window In a new window Download PPT slide   FIGURE 1.— Allele positions within ade6. Arrows indicate the orientation of the M26 and CRE sequences when applicable. The expanded portion shows the nucleotide sequence of the palindromic wt-CRE site (site 2; CRE sequences boxed). Each strand of the DNA at this site is numbered equivalently (5' 3') relative to the CRE sequence in the strand.

 
To test if the M26 heptamer sequence is sufficient for activity, site-directed mutagenesis of 1–4 bp was used to create the heptamer at five locations within the ade6 and ura4 genes (FOX et al. 1997). A hotspot was created in each case, suggesting that the M26 heptamer sequence was sufficient for hotspot activity. Importantly, this study also showed that hotspot activity was independent of the orientation of the M26 heptamer, since sequence inversions at two of the sites were also hotspots.

Although SCHUCHERT et al. (1991) found that any single-base-pair change within the M26 heptamer abolished activity of the hotspot, FOX et al. (2000) showed that a partially overlapping sequence, 5'-TGACGT^A/_C-3', termed cyclic AMP response element (CRE), at the same position in ade6 (Figure 1, site 1) is also an Atf1-Pcr1-dependent hotspot. Thus, the first A of the M26 heptamer is not essential provided that the heptamer is followed by an A or C (normally a G in the wild-type sequence at site 1). These data showed that the Atf1-Pcr1 recognition sequence is more flexible than previously thought and that base pairs outside the M26 heptamer can influence hotspot activity. Here, we identify an 18-bp consensus sequence for binding of the Atf1-Pcr1 transcription factor to purified DNA. This sequence contains the M26 heptamer at its core plus other significantly overrepresented bases to either side of the heptamer. Using the consensus sequence as a guide, we made mutations that showed that some of these additional bases are necessary for intracellular hotspot activity at a site within the ade6 gene (Figure 1, site 2), where the M26 heptamer alone is essentially inactive. The more complete definition of the M26 hotspot nucleotide sequence reported here allows a more accurate assessment of its role in meiotic recombination.

[For clarity, the terms M26 and CRE refer here to the heptamer sequences 5'-ATGACGT-3' and 5'-TGACGT^C/_A-3', respectively, while ade6-M26 refers only to the ade6 allele in which the M26 heptamer was originally identified (GUTZ 1971).]

S. pombe strains, growth media, and meiotic crosses:

The ade6 alleles used in this study are listed in Table 1; the strain numbers and genotypes are available on request. Solid and liquid growth media were made as described by GUTZ et al. (1974). Yeast extract agar (YEA) was supplemented with 100 µg/ml each of adenine, uracil, leucine, and lysine and 50 µg/ml of histidine (YEA + 5S). Yeast extract liquid medium (YEL) was supplemented with 100 µg/ml of adenine and uracil (YEL + 2S). For selection of Ade⁺ recombinants, guanine (80 µg/ml) was substituted for adenine in YEA + 5S (YEA + 4SG) (GRIMM et al. 1991).

View this table: In this window In a new window   TABLE 1 ade6 alleles

 
For crosses, single colonies were picked to YEL + 2S and grown for 2 days at 30° to saturation. Cells from 0.5 ml of culture of each parent strain were mixed in microfuge tubes, washed twice in 1 ml of H₂O, resuspended in 0.1 ml of H₂O, spread on synthetic sporulation agar (GUTZ et al. 1974) supplemented with 100 µg/ml each of adenine and leucine and 50 µg/ml of uracil, and incubated for 2 days at 25°. The resulting spores were harvested and treated with glusulase to kill remaining vegetative cells as previously described (PONTICELLI and SMITH 1989) with the following exception: after glusulase treatment, 0.45 ml of 100% ethanol was added to the 1 ml of spore suspension, mixed, and incubated for 20 min at room temperature. Spores were washed twice and resuspended in 1 ml of H₂O. Total and ade6⁺ spore yields were determined by plating appropriate dilutions on YEA + 5S and YEA + 4SG, respectively. For the experiments shown in Figures 2 and 8, repeat crosses of the same strains were performed on different days using different batches of growth and sporulation media. For the experiments shown in Figures 7 and 9, all crosses with a given test strain were performed in parallel under identical growth and sporulation conditions to minimize day-to-day fluctuations in recombinant frequencies.

View larger version (18K): In this window In a new window Download PPT slide   FIGURE 2.— Apparent orientation dependence of the M26 heptamer. The frequency of recombination was measured in crosses between ade6-M375 and the indicated alleles of ade6. The frequency of recombination in these crosses was not significantly affected by mutations abolishing CRE (ade6-3043) or creating M26 in the forward orientation (ade6-3047). However, mutation to M26 in the opposite orientation (ade6-3086) produced a large increase in recombination. Bars show the frequency of ade+ recombinants from each pair of alleles (mean ± SEM; n = 3–13).

 

View larger version (31K): In this window In a new window Download PPT slide   FIGURE 7.— Control of hotspot activity by nucleotides flanking the M26 heptamer. Multiple alleles of ade6 were constructed containing one or more base pair changes at the wt-CRE site of ade6 (site 2). The frequency of recombinants (mean ± SEM; n = 3) was measured in crosses between these alleles and ade6-469 (A) or ade6-M375 (B). The M26 heptamer, ade6-3047, has little hotspot activity in these crosses, but activity is increased by sequence changes (boldface type) to the left or right of the heptamer (boxed). See RESULTS and DISCUSSION for further details. The dashed line demarcates sequences written 5' 3' from those written 3' 5', the "top" and "bottom" strands of ade6, respectively.

 

View larger version (21K): In this window In a new window Download PPT slide   FIGURE 9.— Only a single additional base change is sufficient for optimal hotspot activity at site 1. Hotspot activities of three of the most active ade6-alleles (Figure 8) were determined in side-by-side comparisons with each other and ade6-M26. Bars show the frequency of ade6+ recombinants (mean ± SEM; n = 3) in crosses between these alleles and ade6-M216 (A) or ade6-469 (B).

 

Generation of new ade6 alleles:

Generation of ade6 alleles for this study was by site-directed mutagenesis using an overlap extension polymerase chain reaction (PCR; VALLEJO et al. 1995). These PCR products were used directly for lithium acetate-mediated transformation (BäHLER et al. 1998) of strains GP4258 (h^– ade6-3059 leu1-32 ura4-D18) or GP4261 (h^– ade6-3060 leu1-32 ura4-D18), each containing a 3.2-kb insertion of a ura4⁺-kanMX6 construct at sites 1 and 2 of the ade6 gene, respectively (Figure 1). Following transformation, cells were allowed to grow for 2 days at 32° in 100 ml of YEL + 2S to dilute the remaining ura4⁺ gene product in transformed cells. Samples (0.1 ml) of these cultures were plated on NBA (0.67% Difco yeast nitrogen base without amino acids, 1% glucose, 2% agar) supplemented with required nutrients at 100 µg/ml and 1 mg/ml of 5-fluoroorotic acid (5-FOA) to select for Ura^– transformants. Alternatively, some ade6 mutants were generated using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) on pJF63 [pBluescript KS+ (Stratagene) containing a 2.9-kb PvuII-SpeI ade6 insert; FARAH et al. 2002)]. The mutagenized plasmid was digested with StuI and SpeI and used for linear transformation of strain GP3162 (h^– ade6-3037 leu1-32 ura4-D18) or GP3163 (h⁺ ade6-3037 leu1-32 ura4-D18). All new ade6 alleles were verified by nucleotide sequencing and Southern blot hybridizations.

SELEX:

The systematic evolution of ligands by exponential enrichment (SELEX) procedure selects for nucleic acid sequences among a random pool of sequences capable of binding to a particular target (TUERK and GOLD 1990; AUSUBEL et al. 2003). In this procedure, sequences that bind the target are purified and amplified by a PCR, followed by additional cycles of binding, purification, and amplification. For the experiments described here, the starting library contained a randomized 30-bp central region. A 70-base oligonucleotide, 5'-CAAGAATCTAGACGTAGGTG-(N)₃₀-CGAATCACCTAAGCTTGGTA-3' (Integrated DNA Technologies), was made double stranded by annealing it with a 3'-complementary primer and incubation at 37° with the Klenow fragment of DNA polymerase I and dNTPs in Klenow reaction buffer (Fisher, Pittsburgh). One hundred micrograms of this double-stranded (ds) oligonucleotide (10¹⁵ molecules) was incubated with gentle rotation overnight at 4° with 25 µl of protein A-conjugated Dynabeads (Dynal, Great Neck, NY) and 4 µg of anti-HA antibody (clone 12CA5; Roche, Indianapolis) in 0.8 ml of binding solution [12% glycerol, 12 mM HEPES (pH 7.9), 4 mM Tris-HCl (pH 7.9), 60 mM KCl, 1 mM EDTA, and 1 mM DTT]. This "preclearing" step was used to eliminate oligonucleotides that bound directly to protein A or anti-HA antibody.

Extracts of strain GP2436 (h⁺ ura4-D18 leu1-32 atf1::HA6His; SHIOZAKI and RUSSELL 1996) were prepared from frozen cell pellets by grinding with a mortar and pestle under liquid N₂. A solution (0.5 ml) containing 50 mM HEPES (pH 7.9), 500 mM NaCl, 20% glycerol, 1 mM DTT, 1 mM EDTA, and 2x complete protease inhibitors (Roche) was added to 2 ml of loosely packed frozen cell powder and gently rotated at 4° for 10 min. The resulting slurry was adjusted to 500 mM NaCl by addition of 5 M NaCl, and rotation was continued for an additional 10 min. The slurry was centrifuged twice (16,000 x g, 4°, 30 min) to remove insoluble cell debris. The protein concentration of the clarified cell extract was measured (Bio-Rad protein assay, Richmond, CA) and adjusted to 10 mg/ml. The extracts were frozen in 100-µl aliquots in liquid N₂ and stored at –80°.

The SELEX was performed by mixing 0.1 ml of cell extract, 0.1 ml of poly(dI-dC) (1 mg/ml; Sigma, St. Louis), 40 µl of 25x complete protease inhibitors (Roche), and 0.8 ml of the precleared oligonucleotide library (above); the mixture was rotated gently at 4° for 30 min. Protein A-conjugated Dynabeads (25 µl) and 4 µg of anti-HA antibody were added, and rotation continued for 3 hr to immunoprecipitate Atf1::HA6His-Pcr1 and associated oligonucleotides. The Dynabeads were washed once in binding solution (1 min, 4° with rotation) and boiled 2 min in 0.1 ml of H₂O. The supernatant was distributed equally into eight 100-µl PCR mixtures using primers to the constant region of the starting 70-bp oligonucleotide (5' sense and 3' complementary). Thirty cycles of PCR were performed using Platinum Taq DNA polymerase (Invitrogen, San Diego) and the manufacturer's recommended conditions with 4 mM MgCl₂. PCR products were purified on a 4% Metaphor agarose (Cambrex) gel, and the above procedures were repeated five times with the following exceptions: (1) the enriched oligonucleotide libraries were precleared with only 10 µl of Dynabeads and 2 µg of anti-HA antibody, and (2) after immunoprecipitation, one additional wash was added per cycle to a maximum of four washes.

After six cycles of SELEX, the remaining oligonucleotides were digested with XbaI and HindIII, restriction enzymes that cut at sites in the constant region, and cloned into pBluescript KS(+) (Stratagene).

Gel-mobility shift assays:

These assays were performed as described by FOX et al. (2000) with the following modifications. For the experiment shown in Figure 3, 215-bp DNA fragments were generated by a PCR using genomic DNA as template and primers with sequences 5'-TGCATCTTTATGGTAAAGCTG-3' and 5'-CACCTTGAATTCATCTAAAATGACGG-3'. PCR products were gel purified and end labeled using [-³²P]ATP and T4 polynucleotide kinase. Extracts were made from strains GP2429 (h⁺ atf1::ura4⁺ leu1-32 his7-366 ura4-D18 ade6-M210) and GP2435 (h^– pcr1::his7⁺ leu1-32 his7-366 ura4-D18), described by SHIOZAKI and RUSSELL (1996) and FOX et al. (2000). Complete protease inhibitor mixture (Roche) was used in place of individual protease inhibitors and was also included in binding reactions. For the experiment shown in Figure 5, 70-bp probes were generated by PCR using the pBluescript clones described in the preceding paragraph as template and 5'-sense and 3'-complementary primers to the constant regions of the 70-bp oligonucleotide used for SELEX.

View larger version (80K): In this window In a new window Download PPT slide   FIGURE 3.— The inactive wt-CRE and M26 heptamers bind the Atf1-Pcr1 transcription factor less strongly than an active heptamer at the same site. Gel mobility shift assays were conducted with 215-bp DNA fragments containing the indicated ade6 allele. DNA was end labeled and mixed with S. pombe protein extracts of cells with the indicated genotype. The percentage of Atf1-Pcr1-dependent shift is shown below each lane. Similar results were obtained in repeat experiments. The atf1 and pcr1 mutant extracts also produced a distinct mobility shift in these experiments that was 2% of the total probe in each case; the basis of this band is not clear.

 

View larger version (58K): In this window In a new window Download PPT slide   FIGURE 5.— In vitro binding of the SELEX clones to Atf1-Pcr1. The 70-bp clones indicated were end labeled, incubated with protein extracts, and assayed by gel mobility shift, as described in MATERIALS AND METHODS. Clone 11 was used as a standard for comparison in all such gels. All bands within the bracketed region were used for quantitation, and the percentage of bound probe is indicated above each lane. Lanes marked with an asterisk contained cell extracts from a pcr1 mutant strain.

 

Apparent orientation dependence of an M26 hotspot:

FOX et al. (2000) showed that the CRE sequence (5'-TGACGT^C/_A-3'), which partially overlaps the M26 sequence (5'-ATGACGT-3'), has significant Atf1-Pcr1-dependent hotspot activity when located at the same position as the original ade6-M26 mutation (Figure 1, site 1). This sequence, unlike M26, is found in the wild-type ade6 gene, at nucleotides 1131–1138 of the ade6 ORF (Figure 1, site 2). Given that both the M26 and CRE sequences are Atf1-Pcr1-dependent hotspots, and given that M26 generated a hotspot at every site where it was created within ade6 (FOX et al. 1997), it seemed probable that this wild-type CRE sequence (wt-CRE) would act as a natural hotspot within the gene.

To test this hypothesis, a translational stop mutation (ade6-3044, G1150T in the ORF) was introduced next to wt-CRE to serve as a marker for gene conversion at that site. In crosses with a strain containing the ade6-M375 allele, the mean frequency of recombinants was 400 ± 40 (SEM) ade⁺ per million viable spores (Figure 2). We then abolished the wt-CRE sequence by a C T substitution at nucleotide position 12 (ade6-3043) in both strains used in the cross. [A comparable C T mutation abolishes the CRE hotspot at site 1 (FOX et al. 2000)]. However, the ade6-3043 mutation resulted in no significant decrease in recombination (Figure 2), indicating that the wt-CRE sequence is not a hotspot.

Since all previous M26 creations within ade6 were hotspots, we next tested if the wt-CRE sequence (site 2) could be activated by mutation to an M26 sequence. In contrast to previous observations, this change produced no significant increase in recombination (Figure 2, ade6-3047). However, since the CRE sequence at site 2 is palindromic, it was also possible to create an M26 heptamer by a comparable mutation on the complementary strand. Remarkably, this mutation created a strong hotspot (Figure 2, ade6-3086), >10 times hotter than ade6-3047 with the M26 heptamer in the opposite orientation. Given that these two M26 heptamers are at essentially the same position within the ade6 gene (overlapping at 4 of 7 bp), it seemed unlikely that the difference in activity could be due to an effect of local chromatin structure as had been previously observed with transplacements of the entire ade6-M26 allele (PONTICELLI and SMITH 1992; VIRGIN et al. 1995). A priori, these data imply that hotspot activity is dependent on the orientation of the M26 heptamer at site 2, in contrast to the previously observed orientation independence (FOX et al. 1997).

In vitro selection for Atf1-Pcr1-binding sequences:

An alternative explanation to a strict orientation dependence of M26 hotspot activity at site 2 is that nucleotides adjacent to the heptamer are necessary for hotspot activity. If so, this might be reflected in differential binding of purified DNA with these sequences to the Atf1-Pcr1 transcription factor. Thus, we tested the wt-CRE sequence and both the active and the inactive orientations of M26 for their ability to bind Atf1-Pcr1 in vitro by gel mobility shift assays. Although all of these sequences bound Atf1-Pcr1, the active allele, ade6-3086, showed greater binding than either of the two inactive alleles (Figure 3), supporting the hypothesis that nucleotides adjacent to the M26 (or CRE) sequence influence binding of the transcription factor. Binding of Atf1-Pcr1 to purified ade6-3086 DNA was 3–4 times greater than to wt-CRE or ade6-3047 DNA, but intracellular hotspot activity of ade6-3086 was more than 10 times greater than that of wt-CRE (ade6-3044) or ade6-3047 (Figure 2). This quantitative difference may reflect the influence of intracellular chromatin structure on hotspot activity (see DISCUSSION).

To determine which nucleotides in addition to the M26 heptamer are needed for binding to purified DNA, we sought a wide range of sequences capable of strongly binding the Atf1-Pcr1 transcription factor. To this end we utilized the SELEX procedure (TUERK and GOLD 1990; AUSUBEL et al. 2003) on a ds-oligonucleotide library containing a 30-bp randomized central region. This technique selects for nucleic acid sequences among a random pool of sequences capable of binding to a particular target. After six cycles of SELEX, we sequenced 30 randomly chosen clones. All 30 clones were unique, but a 6-bp sequence, 5'-TGACGT-3', common to both the M26 and CRE hotspots, was found in all 30 (Figure 4). To derive a consensus Atf1-Pcr1-binding sequence, we aligned the 30 clones relative to this 6-bp common region. For those clones in which the common region occurred as a palindrome (5'-TGACGTCA-3'), we included both strands in the alignment. The M26 heptamer was apparently the preferred recognition sequence for Atf1-Pcr1 in these experiments, as it was found on at least one strand in 27 of the 30 clones analyzed, and in 12 clones M26 occurred as a perfect 10-bp palindrome (5'-ATGACGTCAT-3'). In addition, there were nonrandom distributions of bases extending four nucleotides to the left and four nucleotides to the right of the central 10-bp M26 palindrome. The distribution of bases at positions outside of this region was not significantly different from random (P > 0.05, ²-test, 3 d.f.).

View larger version (79K): In this window In a new window Download PPT slide   FIGURE 4.— The sequence of 30 clones derived from SELEX and the deduced consensus sequence. Sequences are aligned relative to the common region, 5'-TGACGT-3' (asterisks at bottom). Color code is as follows: gray, nucleotides from constant (nonrandom) region of the 70-bp oligonucleotide library; black or color, 30-bp randomized region; red, M26 heptamers; blue, CRE heptamers; green, matches to the consensus sequence for Atf1-Pcr1 binding outside the M26 or CRE heptamer (a few of these matches are in the constant region). Clones 18 and 19 were isolated as two inserts in a single plasmid but were counted as separate clones. Both strands of a clone were included when the common region 5'-TGACGT-3' occurred within a palindrome. The frequency of a given nucleotide at each position is indicated in the box at the bottom, and the most frequent (or significantly underrepresented) nucleotide is shown in boldface type. The probability that the observed distribution derives from a random distribution (25% of each nucleotide) for relevant columns is shown in the bottom row (2-test, d.f. = 3). V represents A, C, or G and B represents C, G, or T. N represents any nucleotide. The double-headed arrow spans the 10-bp M26 palindrome.

 
We confirmed the ability of the SELEX-derived sequences to bind Atf1-Pcr1 in vitro by gel mobility shift assays as above. All of the sequences showed substantial mobility shifts when mixed with wild-type (but not atf1 or pcr1 mutant) cell extracts (Figure 5). In general, those sequences most closely resembling the consensus sequence showed the strongest mobility shifts (Table 2), with one curious exception, clone 7, which contained a single CRE sequence.

View this table: In this window In a new window   TABLE 2 Relative Atf1-Pcr1-binding efficiencies of SELEX-derived clones

 
As described below, we used the consensus sequence for Atf1-Pcr1 binding to purified DNA as a guide for mutagenesis to determine more precisely the nucleotide sequences necessary and sufficient for intracellular M26 hotspot activity.

Apparent orientation dependence of the M26 heptamer is due to adjacent nucleotides:

We noted earlier the large difference in hotspot activity between the active and inactive orientations of the M26 heptamer at site 2 (ade6-3086 and ade6-3047, respectively; Figure 2). Could this difference be explained by their flanking nucleotides? Comparison of each of these sequences with the consensus sequence revealed two bases adjacent to the first A of each heptamer that could explain the difference in activities (Figure 6, positions 3 and 4). These bases matched the consensus sequence in the active, but not the inactive, allele. In addition, the data from Figure 4 suggested that the A at position 1 of the active sequence is preferred over the C found at the same position of the inactive sequence (Figure 6). Therefore, we tested whether the inactive orientation of M26 (ade6-3047) could be activated by mutating each of the nucleotides at positions 1, 3, and 4 to match those found in the active orientation. Figure 7 (bottom set of sequences) shows that mutation of any one of these bases resulted in a 2- to 3-fold increase in the frequency of ade⁺ recombinants observed in test crosses (t-test, P < 0.05). The double mutant T3 C, C4 T (ade6-3090) produced an 6-fold increase, and the triple mutant (ade6-3091) produced an 15-fold increase or about the same activity as ade6-3086 (the active orientation).

View larger version (19K): In this window In a new window Download PPT slide   FIGURE 6.— Comparison of the active (ade6-3086) and inactive (ade6-3047) M26 sequences with the consensus sequence. Matches or close matches to the consensus sequence are underlined. M26 heptamers are boxed, and numbering is as in Figure 1.

 
The data above imply that at least three additional nucleotides to the 5'-side of the M26 heptamer (positions 1, 3, and 4) at site 2 are required for optimal hotspot activity. If so, it should also be possible to inactivate the active M26 orientation, ade6-3086, by mutating the equivalent bases to match those of the inactive heptamer, ade6-3047. As predicted, this allele (ade6-3094) showed a significant reduction in recombination (Figure 7, t-test, P < 0.001). However, recombination with this allele was not reduced to the level of ade6-3047 (the inactive orientation allele). Thus, we made additional mutations to the other side of the active M26 orientation (ade6-3093, equivalent to an 18-bp inversion of the entire inactive sequence), but these changes failed to eliminate the residual activity. Hence, there may be additional nucleotides outside of the inverted segment affecting hotspot activity, although some degree of orientation or chromatin structure dependence cannot be completely excluded.

The consensus sequence (Figure 4) and the data from Table 2 suggested that the 10-bp M26 palindrome was the preferred target for Atf1-Pcr1 binding. Therefore, we also tested this sequence (ade6-3084) as well as the full consensus sequence (ade6-3070; shown in Figure 7 as the complement of the consensus sequence in Figure 4) to see whether recombination at site 2 could be further increased over that already observed. ade6-3070 showed a marginally significant increase of hotspot activity relative to ade6-3086 in crosses with one, but not the other test allele (Figure 7). Thus, the single-base-pair change of ade6-3086 may be sufficient (or nearly so) for optimal hotspot activity at site 2.

Nucleotides outside the M26 heptamer at a second site also affect hotspot activity:

The preceding results suggested that M26 hotspot activity is not an all-or-nothing phenomenon. Rather, there is a continuum between inactivity and full activity. Consistent with this view, we observed an 10-fold range in activities among previous M26 creations in the ade6 gene (FOX et al. 1997; STEINER et al. 2002; and our unpublished results). Thus, we wished to know whether the activity of other M26 heptamers was also affected by their flanking sequences. SCHUCHERT et al. (1991) had previously tested the sequence requirements for hotspot activity in the original ade6-M26 allele (site 1) and concluded that only the heptamer 5'-ATGACGT-3' was necessary for hotspot activity. We conducted a similar analysis at that site by testing every nucleotide outside of the ade6-M26 heptamer that showed a nonrandom distribution in the SELEX consensus sequence (Figure 4). A total of 13 new ade6 alleles containing mutations at site 1 were generated and crossed with 2 test alleles: ade6-M216 and ade6-469. Substantial differences in recombinant frequencies were observed among these alleles (Figure 8), showing that nucleotides near, but outside of, the M26 heptamer also affected hotspot activity at site 1.

View larger version (57K): In this window In a new window Download PPT slide   FIGURE 8.— Nucleotides flanking the M26 heptamer at a second site also affect hotspot activity. Crosses were performed with the indicated ade6 allele and ade6-M216 (A) or ade6-469 (B). Each nucleotide outside the M26 heptamer (boxed) at site 1 (Figure 1) showing a nonrandom distribution in the consensus sequence (Figure 4) was tested for its effect on hotspot activity. If the nucleotide present in the wild-type sequence was different from the consensus, it was changed to match the consensus; otherwise, it was changed not to match (ade6-3072 and ade6-3075). Bars show the frequency of ade+ recombinants (mean ± SEM; n = 3 in A and 5 in B).

 
To quantitate this effect more rigorously, three of the most active alleles in these experiments (ade6-3074, ade6-3079, and ade6-3083) and ade6-M26 were analyzed side-by-side in crosses with two test alelles (Figure 9). Each of these very active alleles showed comparable frequencies of recombination, and each was significantly more active than ade6-M26 (t-test, P < 0.05; Figure 9). Surprisingly, addition of the single G C mutation at position 12 (ade6-3074) to the ade6-M26 sequence was enough to stimulate recombination three- to fourfold in these crosses. Additional mutations, including mutation to the full consensus sequence, showed little if any further stimulation, suggesting that only the additional C is necessary for optimal hotspot activity at site 1. SCHUCHERT et al. (1991) also observed increased levels of recombination caused by the G C mutation at position 12 (referred to as 16C in their article) but concluded that this level was not significantly higher than that of ade6-M26 itself. However, in those experiments the strains used for comparison were derived in different ways. The ade6-M26 allele was created in a strain exposed to X-ray mutagenesis (GUTZ 1971) and, hence, could have been associated with other mutations affecting recombination frequencies. In fact, in the experiments of SCHUCHERT et al. (1991), when the ade6-M26 allele was recreated by site-directed mutagenesis in the same strain background as that of the 16C mutant, recombination frequencies were significantly lower compared to those of 16C (t-test, P < 0.05). Thus, their results are compatible with our results showing that nucleotides outside of the M26 heptamer are necessary for optimal hotspot activity.

The ade6-M26 hotspot was previously characterized as a unique 7-bp sequence, 5'-ATGACGT-3', thought to be necessary and sufficient for the elevated frequency of recombination observed with that allele (SCHUCHERT et al. 1991; FOX et al. 1997). Here, we report that the M26 heptamer is not sufficient for hotspot activity, at least at one site and in one orientation within ade6. The inactivity of M26 at that site is probably due, at least in part, to its relatively weak binding of the Atf1-Pcr1 transcription factor (Figure 3). Hence, we used an in vitro selection and amplification procedure (SELEX) to identify the nucleotide sequences of purified DNA that bound strongly to Atf1-Pcr1 and found an 18-bp consensus sequence containing M26 as a 10-bp palindrome at the center. Positions on each side of the central palindrome also showed significant over- or underrepresentation of particular bases (Figure 4). We used this consensus sequence as a guide to test nucleotides flanking the M26 heptamer for a role in intracellular hotspot activity. Many mutations altering these nucleotides were capable of either increasing or decreasing recombination by as much as 15-fold relative to the M26 heptamer without the additional mutations (Figures 7–9). Thus, more than the M26 heptamer is necessary for optimal hotspot activity.

The most striking feature of the consensus sequence shown in Figure 4 is the M26 palindrome at the central 10 bp. This is not necessarily a remarkable result—if one M26 heptamer is sufficient for some binding of Atf1-Pcr1, then two overlapping heptamers could bind even more strongly and apparently do, since 9 of the 10 strongest binding sequences shown in Table 2 contain the 10-bp M26 palindrome, and the remaining 14 do not. Surprisingly, this palindrome, or even the full consensus sequence, was not necessarily more active as an intracellular hotspot than some other sequences at either of the two sites tested within ade6. A possible explanation for this apparent difference is that the in vitro binding conditions used in these experiments may not accurately reflect the conditions within the cell, where transcription factors must interact with chromatin rather than naked DNA. Indeed, ade6-M26 is often inactive as a hotspot when transplaced on 3- to 5-kb DNA fragments into novel sites in the genome, presumably due to an alteration in chromatin structure (PONTICELLI and SMITH 1992; VIRGIN et al. 1995). Alternatively, factors other than the quality of a protein-binding site may limit recombination even when a stronger binding site is created. Nevertheless, the most active hotspots we found contained at a minimum overlapping M26 and CRE sequences, i.e., 5'-ATGACGTCA-3' (Figures 7 and 9, ade6-3086 and ade6-3074, respectively). The most active hotspot we previously found, ade6-3049, also contains overlapping M26 and CRE sequences and shows activity comparable to that of ade6-3074 (our unpublished results). [ade6-3049 is identical to ade6-3011 but lacks the extraneous mutation found with ade6-3011 (FOX et al. 1997; STEINER et al. 2002).]

Our mutational analysis of M26 hotspot sequences was based on the consensus sequence derived in Figure 4; we focused only on those bases showing a distribution significantly different from random and at fixed positions relative to the central common region. (None of the full 30-bp sequences found by SELEX were directly tested for intracellular hotspot activity.) We assumed that those nucleotides were the ones most likely to affect binding of Atf1-Pcr1 and, hence, hotspot activity. While this assumption was largely supported by our experiments, we do not purport to have identified every nucleotide affecting M26 activity. However, nucleotides outside of the consensus sequence are likely to have only small effects on hotspot activity unless they exert those effects from nonfixed positions or by some means other than binding of the Atf1-Pcr1 transcription factor. The sequence requirements for optimal activity could also vary from site to site depending on the influence of local chromatin structure. Meiosis-specific changes in chromatin structure occur at site 1 with either ade6-M26 or a CRE sequence (ade6-3013) and may determine in part the level of hotspot activity (MIZUNO et al. 1997; FOX et al. 2000; YAMADA et al 2004). An influence of chromatin structure may account for the quantitative differences between Atf1-Pcr1 binding to purified DNA (Figure 3) and intracellular hotspot activity (Figure 2) of certain sequences studied here.

The M26 hotspot was discovered as a single-base-pair mutation that fortuitously created a binding site for a transcription factor (GUTZ 1971; SZANKASI et al. 1988; WAHLS and SMITH 1994; KON et al. 1997). Identification of this mutation has since allowed a detailed dissection of the nucleotide sequence required for its activity. Although we are aware of no other eukaryotic recombination hotspot whose nucleotide sequence has been defined at this level, M26 is similar to other hotspots that also require transcription-factor binding (-hotspots), for example, the HIS4 hotspot of S. cerevisiae (WHITE et al. 1993). In fact, most hotspots in S. cerevisiae appear to occur in gene promoters (BAUDAT and NICOLAS 1997; GERTON et al. 2000), where transcription factors and other proteins often bind. The SELEX procedure we describe here could be utilized to characterize other -hotspots in similar detail. Simple sequences found by this type of analysis may be active at many sites in the genome.

The M26 heptamer is found at 300 sites in the sequenced S. pombe genome (http://www.sanger.ac.uk/Projects/S_pombe/; our unpublished observation). Although it seems plausible that this sequence acts as a hotspot in at least some of those locations, any such natural M26 hotspot has yet to be identified. The results reported here provide a more complete picture of the nucleotide sequence requirements for M26 activity and, therefore, a tool with which to search for natural M26 hotspots in the S. pombe genome. Identifying such natural hotspots will help to elucidate further the control of meiotic recombination.

We thank members of our lab for helpful comments on the manuscript and Joe Farah and Jeff Virgin for plasmids and strains used in this study. This work was supported by National Institutes of Health (NIH) research grant GM31693 to G.R.S. W.W.S. was supported during part of this research by National Research Service Award GM20190 from the NIH and Special Fellowship 3230-05 from the Leukemia and Lymphoma Society.

¹ Present address: Niagara University, Department of Biology, Lewiston, NY 14109.

AUSUBEL, F. M., R. BRENT, R. E. KINGSTON, D. D. MOORE, J. G. SEIDMAN et al. (Editors), 2003 Current Protocols in Molecular Biology. John Wiley & Sons, New York.

BäHLER, J., J. Q. WU, M. S. LONGTINE, N. G. SHAH, A. MCKENZIE, III et al., 1998 Heterologous modules for efficient and versatile PCR-based gene targeting in Schizosaccharomyces pombe. Yeast 14: 943–951.[CrossRef][Medline]

BAKER, B. S., A. T. C. CARPENTER, M. S. ESPOSITO, R. E. ESPOSITO and L. SANDLER, 1976 The genetic control of meiosis. Annu. Rev. Genet. 10: 53–134.[CrossRef][Medline]

BAUDAT, F., and A. NICOLAS, 1997 Clustering of meiotic double-strand breaks on yeast chromosome III. Proc. Natl. Acad. Sci. USA 94: 5213–5218.[Abstract/Free Full Text]

BLUMENTAL-PERRY, A., D. ZENVIRTH, S. KLEIN, I. ONN and G. SIMCHEN, 2000 DNA motif associated with meiotic double-strand break regions in Saccharomyces cerevisiae. EMBO Rep. 1: 232–238.[CrossRef][Medline]

ESPOSITO, R. E., and S. KLAPHOLTZ, 1981 Meiosis and ascospore development, pp. 211–287 in The Molecular Biology of the Yeast Saccharomyces: Life Cycle and Inheritance, edited by J. N. STRATHERN, E. W. JONES and J. R. BROACH. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

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]

FARAH, J. A., E. HARTSUIKER, K.-I. MIZUNO, K. OHTA and G. R. SMITH, 2002 A 160-bp palindrome is a Rad50•Rad32-dependent mitotic recombination hotspot in Schizosaccharomyces pombe. Genetics 161: 461–468.[Abstract/Free Full Text]

FOX, M. E., J. B. VIRGIN, J. METZGER and G. R. SMITH, 1997 Position- and orientation-independent activity of the Schizosaccharomyces pombe meiotic recombination hot spot M26. Proc. Natl. Acad. Sci. USA 94: 7446–7451.[Abstract/Free Full Text]

FOX, M. E., T. YAMADA, K. OHTA and G. R. SMITH, 2000 A family of CRE-related DNA sequences with meiotic recombination hotspot activity in Schizosaccharomyces pombe. Genetics 156: 59–68.[Abstract/Free Full Text]

GERTON, J. L., J. DERISI, R. SHROFF, M. LICHTEN, P. O. BROWN et al., 2000 Global mapping of meiotic recombination hotspots and coldspots in the yeast Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 97: 11383–11390.[Abstract/Free Full Text]

GRIMM, C., P. SCHäR, P. MUNZ and J. KOHLI, 1991 The strong ADH1 promoter stimulates mitotic and meiotic recombination at the ADE6 gene of Schizosaccharomyces pombe. Mol. Cell. Biol. 11: 289–298.[Abstract/Free Full Text]

GUTZ, H., 1971 Site specific induction of gene conversion in Schizosaccharomyces pombe. Genetics 69: 317–337.[Free Full Text]

GUTZ, H., H. HESLOT, U. LEUPOLD and N. LOPRIENO, 1974 Schizosaccharomyces pombe, pp. 395–446 in Handbook of Genetics, edited by R. C. KING. Plenum Press, New York.

HARING, S. J., L. J. LAUTNER, J. M. COMERON and R. E. MALONE, 2004 A test of the CoHR motif associated with meiotic double-strand breaks in Saccharomyces cerevisiae. EMBO Rep. 5: 41–46.[CrossRef][Medline]

KIRKPATRICK, D. T., Q. FAN and T. D. PETES, 1999a Maximal stimulation of meiotic recombination by a yeast transcription factor requires the transcription activation domain and a DNA-binding domain. Genetics 152: 101–115.[Abstract/Free Full Text]

KIRKPATRICK, D. T., Y.-H. WANG, M. DOMINSKA, J. D. GRIFFITH and T. D. PETES, 1999b Control of meiotic recombination and gene expression in yeast by a simple repetitive DNA sequence that excludes nucleosomes. Mol. Cell. Biol. 19: 7661–7671.[Abstract/Free Full Text]

KON, N., M. D. KRAWCHUK, B. G. WARREN, G. R. SMITH and W. P. WAHLS, 1997 Transcription factor Mts1/Mts2 (Atf1/Pcr1, Gad7/Pcr1) activates the M26 meiotic recombination hotspot in Schizosaccharomyces pombe. Proc. Natl. Acad. Sci. USA 94: 13756–13770.

MIZUNO, K., Y. EMURA, M. BAUR, J. KOHLI, K. OHTA et al., 1997 The meiotic recombination hot spot created by the single-base substitution ade6–M26 results in remodeling of chromatin structure in fission yeast. Genes Dev. 11: 876–886.[Abstract/Free Full Text]

PâQUES, F., and J. E. HABER, 1999 Multiple pathways of recombination induced by double-strand breaks in Saccharomyces cerevisiae. Microbiol. Mol. Biol. Rev. 63: 349–404.[Abstract/Free Full Text]

PETES, T. D., 2001 Meiotic recombination hot spots and cold spots. Nat. Rev. Genet. 2: 360–370.[CrossRef][Medline]

PONTICELLI, A. S., and G. R. SMITH, 1989 Meiotic recombination-deficient mutants of Schizosaccharomyces pombe. Genetics 123: 45–54.[Abstract/Free Full Text]

PONTICELLI, A. S., and G. R. SMITH, 1992 Chromosomal context dependence of a eukaryotic recombinational hot spot. Proc. Natl. Acad. Sci. USA 89: 227–231.[Abstract/Free Full Text]

SCHUCHERT, P., M. LANGSFORD, E. KäSLIN and J. KOHLI, 1991 A specific DNA sequence is required for high frequency of recombination in the ade6 gene of fission yeast. EMBO J. 10: 2157–2163.[Medline]

SHIOZAKI, K., and P. RUSSELL, 1996 Conjugation, meiosis, and the osmotic stress response are regulated by Spc1 kinase through Atf1 transcription factor in fission yeast. Genes Dev. 10: 2276–2288.[Abstract/Free Full Text]

STAHL, F. W., 2002 Hot spot of recombination, pp. 976–977 in Encyclopedia of Genetics, edited by S. BRENNER and J. H. MILLER. Academic Press, San Diego.

STEINER, W. W., R. W. SCHRECKHISE and G. R. SMITH, 2002 Meiotic DNA breaks at the S. pombe recombination hotspot M26. Mol. Cell 9: 847–855.[CrossRef][Medline]

SUN, H., D. TRECO, N. P. SCHULTES and J. W. SZOSTAK, 1989 Double-strand breaks at an initiation site for meiotic gene conversion. Nature 338: 87–90.[CrossRef][Medline]

SZANKASI, P., W. D. HEYER, P. SCHUCHERT and J. KOHLI, 1988 DNA sequence analysis of the ade6 gene of Schizosaccharomyces pombe: wild-type and mutant alleles including the recombination hotspot allele ade6–M26. J. Mol. Biol. 204: 917–925.[CrossRef][Medline]

TUERK, C., and L. GOLD, 1990 Systematic evolution of ligands by exponential enrichment: RNA ligands to bacteriophage T4 DNA polymerase. Science 249: 505–510.[Abstract/Free Full Text]

VALLEJO, A. N., R. J. POGULIS and L. R. PEASE, 1995 Mutagenesis by PCR, pp. 603–612 in PCR Primer: A Laboratory Manual, edited by G. S. DVEKSLER. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

VIRGIN, J. B., J. METZGER and G. R. SMITH, 1995 Active and inactive transplacement of the M26 recombination hotspot in Schizosaccharomyces pombe. Genetics 141: 33–48.[Abstract]

WAHLS, W. P., and G. R. SMITH, 1994 A heteromeric protein that binds to a meiotic homologous recombination hotspot: correlation of binding and hotspot activity. Genes Dev. 8: 1693–1702.[Abstract/Free Full Text]

WHITE, M. A., M. DOMINSKA and T. D. PETES, 1993 Transcription factors are required for the meiotic recombination hotspot at the HIS4 locus in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 90: 6621–6625.[Abstract/Free Full Text]

YAMADA, T., K. MIZUNO, K. HIROTA, N. KON, W. P. WAHLS et al., 2004 Roles of histone acetylation and chromatin remodeling factor in a meiotic recombination hotspot. EMBO J. 23: 1792–1803.[CrossRef][Medline]

Communicating editor: A. NICOLAS

This article has been cited by other articles:

				J. Gao, M. K. Davidson, and W. P. Wahls Distinct regions of ATF/CREB proteins Atf1 and Pcr1 control recombination hotspot ade6-M26 and the osmotic stress response Nucleic Acids Res., May 1, 2008; 36(9): 2838 - 2851. [Abstract] [Full Text] [PDF]

				J. A. Farah, G. Cromie, L. Davis, W. W. Steiner, and G. R. Smith Activation of an Alternative, Rec12 (Spo11)-Independent Pathway of Fission Yeast Meiotic Recombination in the Absence of a DNA Flap Endonuclease Genetics, December 1, 2005; 171(4): 1499 - 1511. [Abstract] [Full Text] [PDF]

				W. W. Steiner and G. R. Smith Natural Meiotic Recombination Hot Spots in the Schizosaccharomyces pombe Genome Successfully Predicted from the Simple Sequence Motif M26 Mol. Cell. Biol., October 15, 2005; 25(20): 9054 - 9062. [Abstract] [Full Text] [PDF]

				E. A. Espeso, L. Cobeno, and H. N. Arst Jr. Discrepancies Between Recombination Frequencies and Physical Distances in Aspergillus nidulans: Implications for Gene Identification Genetics, October 1, 2005; 171(2): 835 - 838. [Abstract] [Full Text] [PDF]

International Access Link

Online ISS 1091-6490
Copyright 2008 by the Genetics Society of America
phone: 412-268-1812   fax: 412-268-1813   email: genetics-gsa{at}andrew.cmu.edu