Genetics, Vol. 154, 1125-1139, March 2000, Copyright © 2000
An mre11 Mutant of Coprinus cinereus Has Defects in Meiotic Chromosome Pairing, Condensation and Synapsis
Erin E. Gereckea and
Miriam E. Zolana
a Department of Biology, Indiana University, Bloomington, Indiana 47405
Corresponding author:
Miriam E. Zolan, Department of Biology, Indiana University, Jordan Hall 142, 1001 E. 3rd St., Bloomington, IN 47405-3700., mzolan{at}bio.indiana.edu (E-mail)
Communicating editor: P. J. PUKKILA
 | ABSTRACT |
|---|
The rad11 gene of the basidiomycete Coprinus cinereus is required for the completion of meiosis and for survival after gamma irradiation. We have cloned the rad11 gene and shown that it is a homolog of MRE11, a gene required for meiosis and DNA repair in numerous organisms. The expression of C. cinereus mre11 is induced during prophase I of meiosis and following gamma irradiation. The gene encodes a predicted polypeptide of 731 amino acids, and the mre11-1 (rad11-1) mutation is a single base pair change that results in a stop codon after amino acid 315. The mre11-1 mutant shows enhanced sensitivity to ionizing radiation, but no enhanced sensitivity to UV radiation. It shows a delay in fruitbody formation and a reduction in the number of mushrooms formed per dikaryon. The mre11-1 mutant also has several meiotic defects. Pachytene chromatin condensation is disrupted, and although some meiotic cells appear to achieve metaphase I condensation, no further meiotic progression is observed. The mre11-1 mutant also fails to undergo proper chromosome synapsis; neither axial elements nor mature synaptonemal complexes are complete. Finally, meiotic homolog pairing is reduced in the mre11-1 mutant. Thus, in C. cinereus, Mre11 is required for meiotic DNA metabolism.
THE processes involved in meiotic chromosome behavior and recombination bear striking similarity to events involved in certain types of DNA repair. In both meiosis and DNA double-strand break (DSB) repair, homologous sequences are identified and DNA breaks are repaired, often with an exchange of genetic material. In Saccharomyces cerevisiae, DSBs initiate meiotic recombination and are generated by a type II topoisomerase-like protein, Spo11 (KEENEY et al. 1997 
), in conjunction with the Mre11/Rad50/Xrs2 complex of proteins; this complex also is required for processing of these meiotic DSBs (JOHZUKA and OGAWA 1995 ; NAIRZ and KLEIN 1997 ; TSUBOUCHI and OGAWA 1998 ). Mutants of SPO11 in Caenorhabditis elegans (DERNBURG et al. 1998 ) and Drosophila melanogaster (MCKIM and HAYASHI-HAGIHARA 1998 ) are deficient in meiotic recombination, implying that certain aspects of meiotic DSB activity may be conserved. Mutations in S. cerevisiae MRE11, RAD50, or XRS2 result in defects in meiotic recombination and spore formation, as well as an increased sensitivity to ionizing radiation (reviewed in HABER 1998 ).
We have chosen to study meiosis and DNA repair in the basidiomycete Coprinus cinereus. This fungus grows as a monokaroytic mycelium, and two individuals with compatible mating types can mate to form a dikaryon. Under the proper light and temperature conditions, the dikaryon will form a fruitbody, or mushroom (ZOLAN et al. 1988 ). Each mushroom contains approximately ten million meiotic cells that undergo meiosis in a naturally synchronous manner. This feature of the biology of C. cinereus provides an ideal opportunity to examine the behavior of meiotic chromosomes cytologically. The dynamics of meiotic chromosomes during prophase and metaphase I have been documented thoroughly for wild-type mushrooms (RAJU and LU 1970 ; HOLM et al. 1981 ; SEITZ et al. 1996 ).
The rad3, rad9, rad11, and rad12 complementation groups of C. cinereus were identified initially in screens for radiation-sensitive, meiotic mutants (ZOLAN et al. 1988 ; VALENTINE et al. 1995 ). Mutant alleles of these genes exhibit meiotic defects and show enhanced sensitivity to ionizing radiation; the four genes comprise a single epistasis group for survival following gamma irradiation (VALENTINE et al. 1995 ). The phenotypes of rad3, rad9, and rad12 mutants have been described in detail (ZOLAN et al. 1988 , ZOLAN et al. 1994 , ZOLAN et al. 1995 ; PUKKILA et al. 1992 ; RAMESH and ZOLAN 1995 ; SEITZ et al. 1996 ). These mutants show defects in chromatin condensation and synapsis, and spore production and viability are decreased relative to the levels observed in wild-type mushrooms.
We had shown previously that the rad11-1 mutant has low spore production and fails to complete the meiotic divisions, and we mapped the rad11 gene to chromosome 6 of C. cinereus (ZOLAN et al. 1992 ; VALENTINE et al. 1995). We wanted to determine the identity of the rad11 gene and to explore further the meiotic defects of the rad11-1 mutant. In this study, we describe the cloning of rad11 and its identification as the C. cinereus homolog of MRE11. We also demonstrate that rad11-1/mre11-1 nuclei display defects in chromatin condensation, homolog pairing, and synaptonemal complex (SC) structure.
| MATERIALS AND METHODS |
|---|
Strains and culture conditions:
The dikaryotic wild-type C. cinereus strain (J6;5-5 x J6;5-4) used for transformation, gene expression, fruiting, and microscopy experiments has been described previously (VALENTINE et al. 1995 ). The rad11-1/mre11-1 mutant was generated by UV mutagenesis of Java-6 (BINNINGER et al. 1987 ) and then was crossed and backcrossed a total of five times to the Rad+ strain Okayama-7 (WU et al. 1983 ; VALENTINE et al. 1995 ). Four mre11-1 dikaryons congenic with the wild-type strain were generated by crossing sibling isolates from the fifth generation of backcrossed mre11-1 strains: (1) mre11-1;5-2 x mre11-1;5-5, (2) mre11-1;5-101 x mre11-1;5-110, (3) mre11-1;5-2 x mre11-1;5-7, and (4) mre11-1;5-108 x mre11-1;5-111. The transformation strain was constructed by mating an mre11-1 strain to strain 218 (trp1-1;1-6; BINNINGER et al. 1987 ) and isolating a Rad-Trp- progeny isolate. This strain was also used to examine fruitbody development in the mre11-1 mutant by crossing it to mre11-1;5-2.
Culture conditions, matings, and fruiting conditions were as described previously (ZOLAN et al. 1988 ), except for fruitbody development experiments, in which matings were performed directly on 25-mm slant cultures used to induce fruiting.
Gel electrophoresis and Southern hybridization:
Genomic DNA was isolated as described in ZOLAN and PUKKILA 1986 . Restriction digests were performed as directed by the manufacturer (New England Biolabs, Beverly, MA). Gel electrophoresis was performed using 0.7–1.2% agarose gels in 1x TBE (Tris-HCl, boric acid, EDTA). Extraction of DNA from agarose gels was performed using the QiaexII Gel extraction kit (QIAGEN, Valencia, CA). Southern blots were performed using standard techniques (ZOLAN et al. 1992 ), and gels were transferred to Magnagraph nylon membrane (MSI). Probes were radiolabeled using the random priming method (YEAGER STASSEN et al. 1997 ), unless stated otherwise. Hybridizations were performed as described in YEAGER STASSEN et al. 1997 , except that after 10–12 hr of hybridization, three 30-min washes at 65° using 0.2x SSC/0.1% SDS were performed, and then blots were exposed to X-ray film at -80°. The film was developed on a Kodak X-Omat automatic developer.
Chromosome walking:
A chromosome walk was initiated from cosmid 4A6 from an Okayama-7 chromosome-6-specific library (Fig 1A; ZOLAN et al. 1992 ). 4A6 DNA was obtained using Wizard columns (Promega, Madison, WI) and homemade solutions prepared as specified by the manufacturer. The DNA was digested with PstI, and T3 and T7 primers were used independently to generate primer extension products of insert DNA immediately adjacent to the vector. [32P]dATP was used in the reaction to create radioactively labeled fragments that were used to probe the chromosome-6-specific library using standard colony hybridization procedures (SAMBROOK et al. 1989 ). In this way, clones were isolated that showed hybridization with each end of 4A6; this process was repeated for the next steps in the walk. Ultimately, a single step was taken in one direction (21F1; ~17 kb of new sequence; Fig 1A), while two steps were taken in the other direction (4B12 and 21F11; ~50 kb of new sequence); in total, ~100 kb of DNA was covered.
DNA from cosmids 21F1 and 21F11 was digested with EcoRI, and subclones representing each end of the walk were ligated into pBluescript SK+ (Stratagene, La Jolla, CA), transformed into the XL1-Blue strain of Escherichia coli using the method of POPE and KENT 1996 , sequenced, translated, and compared to all polypeptide sequences in the NCBI database using BLAST. A 2.4-kb EcoRI subclone of 21F1 showed sequence similarity to the Mre11 family of proteins (Fig 1A). Ultimately, a 4.3-kb NotI-PstI subclone was generated from the same end of 21F1; this fragment contained the entire coding sequence of C. cinereus mre11 (Fig 1A).
Restriction fragment length polymorphism (RFLP) mapping:
C. cinereus Okayama-7 and a fourth-generation backcrossed rad11-1 strain (rad11-1;4-8) were mated, the resulting dikaryon was induced to fruit, and spore progeny were isolated. Genomic DNA from parental and progeny isolates was digested with HindIII, subjected to gel electrophoresis and Southern transfer, and probed with the radiolabeled insert of the 4.3-kb NotI-PstI subclone of cosmid 21F1 (Fig 1A).
Transformation rescue assays:
Transformations of C. cinereus oidial protoplasts were performed as described in BINNINGER et al. 1987 and ZOLAN et al. 1992 . A rad11-1;trp1-1,1-6 strain was transformed with either 3 µg of cosmid 21F1 (which contains the trp1 marker) or 10 µg of a plasmid construct (pEG1) containing the mre11 gene. To generate pEG1, a 4.1-kb fragment containing mre11 plus 689 bases upstream of the initiating Met and 705 bases downstream of the terminal stop codon was amplified from cosmid 21F1 using Vent DNA polymerase (New England Biolabs). Amplifications were performed in a Gene Amp PCR system (model 2400; Perkin-Elmer, Norwalk, CT). The product was cloned into pCR2.1 (Invitrogen, Carlsbad, CA), and the C. cinereus trp1 gene (SKRZYNIA et al. 1989 ) was cloned into the only HindIII site of the construct.
For each transformation, tryptophan prototrophs were selected on regeneration agar (BINNINGER et al. 1987 ) and screened for enhanced sensitivity to ionizing radiation by a chunk test (ZOLAN et al. 1988 ). Rescue of the meiotic defects of rad11-1 was assayed by mating transformants to a compatible rad11-1 strain and observing mushroom caps macroscopically for spore production. Six rescued transformants were analyzed for the presence of one or more additional copies of mre11. Genomic DNA was isolated and digested with HindIII, which does not digest within mre11. The DNA was subjected to gel electrophoresis and Southern transfer. The blot was hybridized with a radiolabeled 1.8-kb genomic fragment of mre11 (Fig 1A), which was generated by amplifying the region of mre11 using Taq polymerase (Fisher Scientific/Promega) and labeling the product with [32P]dATP.
DNA sequencing and analysis:
Plasmids containing subclones were isolated from E. coli using either a Qiagen Plasmid Mini kit (QIAGEN) or Perfect Prep Plasmid DNA kit (5 PRIME-3 PRIME, Inc.). Automated DNA sequencing was performed at the Indiana Institute for Molecular and Cellular Biology using a LI-COR model 4000L DNA Sequencer (LI-COR, Inc.) with a SequiTherm Long-Read cycle sequencing kit (Epicentre Technologies Corp.) or an ABi Prism model 377 DNA sequencer (Applied Biosystems, Foster City, CA) with a Thermo Sequenase cycle sequencing kit (Amersham Life Sciences, Inc., Piscataway, NJ). Internal oligonucleotide primers were made by Genosys, Inc. Sequencing data were assembled and edited using Sequencher ver. 3.0 (Gene Codes Corp.), and were translated using DNASIS ver. 2.0 (Hitachi Software Engineering Co., Ltd.). Predicted polypeptides were identified using blastx in BLAST (ALTSCHUL et al. 1990 ), and alignments among amino acid sequences were performed using CLUSTAL W ver. 1.74 (THOMPSON et al. 1994 ). Prediction of nuclear localization sequences was performed using PSORT (NAKAI and KANEHISA 1992 ).
The genomic sequence of mre11 from C. cinereus was determined from cosmid 21F1 (made from strain Okayama-7) and subclones derived from it (see above). The DNA sequences obtained were translated in all six reading frames and compared to all nonredundant polypeptides in the translated NCBI database (GenBank, Bethesda, MD) to identify similar sequences. The mutation in the rad11-1/mre11-1 mutant was determined by amplifying overlapping regions of the mre11 gene from backcrossed strain rad11-1;4-8 or the original, unbackcrossed, rad11 isolate in the Java-6 background and sequencing the PCR products directly, using the PCR primers as sequencing primers. Where polymorphisms between rad11-1 and Okayama-7 sequences were noted, the relevant regions of mre11 were sequenced from Java-6 genomic DNA to identify unique base changes in mre11-1. The sequence of a fifth-generation isolate of rad11-2 was determined in the same manner used for rad11-1. Additionally, putative rad11-2 isolates that represented each generation of backcrossing were screened for the mutation present in rad11-1 by amplifying the affected region by PCR and digesting the product with MfeI, which recognizes only the nonmutated sequence. Each strain used in phenotypic characterization (see above) was checked for the presence of the rad11-1 mutation.
Isolation of mre11 cDNA:
A C. cinereus cDNA 
ZAP library constructed from cap tissue isolated 1 hr before and 1 hr after karyogamy (YEAGER STASSEN et al. 1997 ) was screened for mre11. To obtain the 3' end of the mre11 cDNA, primers were designed based on the genomic DNA that encodes the predicted open reading frame of C. cinereus mre11 and a priming site within the vector (T7 promoter region). Two sequential PCR amplifications were performed. In the first, a primer from the 3' region of mre11 was used in conjunction with the T7 primer to amplify from the library: 3 µl of the library was diluted with 2 µl of nanopure water, heated to 100° for 3 min, and then quick-cooled. Samples were centrifuged for 3 min, and 2 µl of the supernatant was used as the template in a PCR reaction. A total of 1 µl of this reaction was used in a second PCR amplification, wherein a nested primer within the mre11 sequence was used in combination with the T7 primer. PCR products were separated by gel electrophoresis, transferred, and probed with a portion of mre11 (Fig 1A). Positive products were separated again on agarose gels, excised, purified, ligated into pCR2.1 (Invitrogen), and transformed into E. coli XL-1 Blue competent cells. A 1.1-kb clone that hybridized with an mre11 probe (Fig 1A) was identified and sequenced.
The 5' end of the mre11 cDNA was amplified using random amplification of cDNA ends (5' RACE; FROHMAN et al. 1988 ). A total of 7 µg of wild-type total meiotic RNA collected from cap tissue at karyogamy was used in a reverse transcription (RT) reaction (SuperScriptII; Gibco BRL Life Technologies, Grand Island, NY). cDNA was purified using the GlassMax system (Gibco BRL Life Technologies), tailed with dCTP (using terminal deoxynucleotidyl transferase; Promega), and 2 µl of the tailing reaction was used to amplify the 5' end of the mre11 cDNA, using a poly-G anchor primer and mre11 primer. Products were separated by gel electrophoresis, and the gel was blotted and probed with a radiolabeled fragment of mre11 genomic DNA (Fig 1A). Positive products were cloned and sequenced as above, and the 5' end of the mre11 cDNA transcript was identified by locating the first base that followed the poly-G tail.
RT-PCR was used to isolate the remainder of the mre11 cDNA sequence. Poly(A)+ RNA, purified from wild-type caps collected at 6 hr after karyogamy (YEAGER STASSEN et al. 1997 ), was used in the RT reaction with an oligo(dT) primer to generate cDNA (SuperScript II; Gibco BRL Life Technologies). To amplify the central 2.1 kb of mre11 cDNA, 2 µl of the 20 µl RT reaction was used as the template in a 100-µl PCR reaction. The products were separated by gel electrophoresis, blotted, and hybridized with a genomic fragment of mre11 (Fig 1A). The amplified product was cloned and sequenced. Finally, a second RT-PCR reaction using primers in the 5' region of mre11 was used to amplify the remainder of the mre11 cDNA. The product was cloned and sequenced as described above.
Northern analysis:
Approximately 0.5 g of cap tissue from 0, 6, and 12 hr after karyogamy was isolated from the wild-type cross J6;5-4 x J6;5-5. Java-6 vegetative dikaryon, Okayama-7 unirradiated vegetative monokaryon, and Okayama-7 vegetative monokaryon isolated 4 hr after irradiation with 40 krad also were collected. Total RNA was isolated as described in YEAGER STASSEN et al. 1997 , except that the extraction buffer and ethanol solutions did not contain diethylpyrocarbonate (DEPC), and RNA samples were resuspended in RNase-free H2O. RNA samples were separated by gel electrophoresis on 1.2% agarose gels containing formaldehyde (MANIATIS et al. 1982 ) and were blotted onto nylon membrane. Blots were hybridized with a 1.8-kb genomic fragment of mre11 (Fig 1A) using QuikHyb (Stratagene). Blots were exposed either to radiographic film or to a phosphorimager screen (Molecular Dynamics) and were visualized using a PhosphorImager scanner and ImageQuant ver. 3.3 software (Molecular Dynamics). Gels were stained with ethidium bromide to confirm equal loading, based on the quantity of rRNA present.
Radiation sensitivity assays:
Chunk tests were performed as described by ZOLAN et al. 1988 in a 137Cs irradiator (Mark-1 model 68-A, J. L. Shepard and Associates).
UV sensitivity of mre11-1 oidia was assayed as described in RAMESH and ZOLAN 1995 , with a dose rate of 2 J/m2/sec at a distance of 45 cm. Doses used were 0, 50, 100, and 150 J/m2.
Microscopy:
For DAPI (4',6-diamidino-2'-phenylindole dihydrochloride) staining of basidia, a sliver of cap tissue approximately the thickness of one gill layer was taken from a fruitbody from which the veil cells had been removed. A drop of DAPI (1 µg/ml) was applied to the tissue on a glass slide, a coverslip was added, and the sample was flattened by gently tapping the coverslip with a rubber mallet. Tissue was examined under UV fluorescence on a Nikon Microphot-FXA microscope equipped with appropriate epifluorescence filters.
Surface spreads of C. cinereus meiotic chromosomes were performed as described by PUKKILA et al. 1992 . Acridine orange and silver nitrate staining of meiotic nuclei were performed as described in SEITZ et al. 1996 . Analysis of chromatin condensation was performed on a Nikon Microphot-FXA fluorescence microscope. Data were collected on 35-mm color slide film. For examination of silver-nitrate-stained synaptonemal complexes, electron microscopy grids were floated and then analyzed using a JEOL-1010 electron microscope.
Fluorescence in situ hybridization (FISH) was performed as described in LI et al. 1999 . Meiotic nuclei from mre11-1 caps were examined at 0, 6, and 10 hr after karyogamy. DNA probes were prepared as described in LI et al. 1999 . Each probe was composed of two overlapping cosmids from interstitial regions of either chromosome 8 or 13. The chromosome 8 probe consisted of cosmids 2G7 and 3F2 from the chromosome-8-specific library (~61 kb in complexity), and the chromosome 13 probe consisted of cosmids 1B2 and 2B9 from the chromosome-13-specific library (42 kb).
Spore production assay:
Spore production was determined using the drop method described by RAMESH and ZOLAN 1995 .
| RESULTS |
|---|
Isolation of mre11 from C. cinereus:
The C. cinereus rad11 gene is required for successful completion of meiosis and for survival after exposure to ionizing radiation (VALENTINE et al. 1995 ). We had shown previously, using RFLP and chromosome length polymorphism mapping, that rad11 is located on chromosome 6 of C. cinereus (ZOLAN et al. 1994 ). Using a C. cinereus chromosome-6-specific cosmid library, we had isolated a cosmid (4A6; Fig 1A), which is 1.6 map units from rad11 (ZOLAN et al. 1992 ). Because a single map unit in C. cinereus can be as small as 30 kb (DUTTA 1974 ; HOLM et al. 1981 ), a bidirectional chromosome walk was initiated from this cosmid. Cosmids that hybridized with either end of 4A6 were identified and the process was repeated for subsequent steps in the walk. DNA fragments from the distal ends of the walk were subcloned and sequenced, and the DNA sequence from the distal end of cosmid 21F1 (Fig 1A) was found to encode a polypeptide with similarity to Mre11 homologs from Homo sapiens, Mus musculus, S. cerevisiae, Schizosaccharomyces pombe, and C. elegans.
mre11 mutants in S. cerevisiae exhibit enhanced radiation sensitivity and meiotic defects (e.g., AJIMURA et al. 1993 ). Therefore, several experiments were performed to determine whether mre11 is rad11. First, RFLP mapping was used to ask whether mre11 maps to the rad11 locus. No recombination between an RFLP pattern specific to mre11 and the radiation-sensitivity phenotype was revealed in 50 progeny of a cross between a wild-type strain (Okayama-7) and rad11-1 (data not shown). This indicated that the region of DNA containing mre11 is two or fewer map units from the locus encoding rad11.
Second, transformation experiments were performed to determine whether mre11 from C. cinereus is able to rescue the meiotic defects and radiation sensitivity of the rad11-1 mutant. A rad11-1;trp1-1,1-6 strain was transformed with either cosmid 21F1 or a plasmid containing the mre11 coding sequence plus ~0.7 kb upstream and downstream of the gene (pEG1); both constructs contained the trp1 marker. Tryptophan prototrophs (184 from the cosmid transformation and 97 from the pEG1 transformation) were screened for radiation sensitivity and gross spore production levels. Transformation with either construct resulted in rescue of both defects of rad11-1. A total of 79% of cosmid transformants and 82% of pEG1 transformants were resistant to 40 krad of 
-radiation as measured by growth of a small mycelial patch (chunk test; ZOLAN et al. 1988 ). Of these transformants, 83% of cosmid transformants and 76% of subclone transformants were competent for spore production; complementation of the sporeless phenotype was observed in crosses of transformants to a compatible rad11-1 strain. To confirm that the phenotypic rescue of rad11-1 was due to introduction of mre11, six transformants exhibiting rescue of the radiation-sensitivity phenotype were analyzed for the presence of additional copies of mre11 in their genomes. In each, at least one new band was detected using a probe specific to mre11 (Fig 1A; data not shown). Thus, the rescue of both phenotypes of rad11-1 in these transformants likely was due to the presence of introduced mre11 DNA.
In a third approach to determine whether mre11 is rad11, we asked if the mre11 gene is altered in the rad11-1 mutant. The genomic sequence of mre11 from the Okayama-7 strain of C. cinereus was determined (GenBank accession no.
AF178433). A single base change (C to T transition) was identified in the rad11-1 mutant 1256 bases downstream of the predicted translational start site of mre11. This mutation was identified in a fourth-generation backcrossed isolate (see MATERIALS AND METHODS), and the original, unbackcrossed rad11-1 mutant also carries the same C to T transition. The mre11 gene also was sequenced in the wild-type strain (Java-6) that was mutagenized to create rad11-1. The base change in rad11-1 is unique to the rad11-1 strain and is not a polymorphism present between Java-6 and Okayama-7. Thus, three lines of evidence, RFLP mapping, transformation rescue, and direct sequencing, have shown that C. cinereus rad11 is mre11.
Previous studies in C. cinereus had indicated that our collection of mutants contained two mutant alleles of rad11. Fifth-generation backcrossed isolates of two strains in our collection failed to complement one another, and their mutations mapped to the same locus on chromosome 6; these mutants were named rad11-1 and rad11-2 (VALENTINE et al. 1995 ). However, analysis of a fifth-generation isolate of rad11-2 showed that the sequence of mre11 in this strain is identical to that of mre11 in rad11-1, including the C to T mutation at the same position. Examination of the presumptive unbackcrossed progenitor of this strain showed no defect in the mre11 gene. We conclude that rad11-1 was duplicated inadvertently during the backcrossing procedure. Thus, all fifth-generation isolates of rad11-1 and rad11-2 have been collectively renamed mre11-1. All strains used for the experiments presented in this work were examined and have the identical mutation in mre11.
Gene structure and expression of mre11 in C. cinereus:
To determine the structure of mre11 in C. cinereus, the cDNA sequence was obtained and aligned to the genomic sequence for strain Okayama-7. The mre11 coding sequence consists of 2193 bases interrupted by 10 introns ranging in size from 46 to 61 bp, and the average size is 53 bp (Table 1; Fig 1A). The number and average size of the introns are consistent with those of introns from other C. cinereus genes [e.g., trp1 (SKRZYNIA et al. 1989 ), rad9 (SEITZ et al. 1996 ), and rad51 (YEAGER STASSEN et al. 1997 )]. The 5' splice sites of the introns correlate well with the consensus sequence determined for filamentous fungi (GTRNGT, Table 1; GURR et al. 1987 ; EDELMANN and STABEN 1994 ). The 3' splice sites are also similar to the consensus sequence, YAG; in introns 2 and 8, the Y was replaced by an A, which is not uncommon for fungal genes (GURR et al. 1987 ; SEITZ et al. 1996 ). Each intron was examined for a branchpoint consensus sequence (CTRAY) derived from the examination of various multicellular eukaryotes, including fungi (Neurospora crassa; e.g., KINNAIRD and FINCHAM 1983 ; WOUDT et al. 1983 ), plants (BROWN 1986 ), and mammals (KELLER and NOON 1984 ). Each mre11 intron contained at least one sequence that corresponded well to this consensus sequence (underlined in Table 1). In many cases, a T was present in the first position, which is common in the introns of many organisms (e.g., KELLER and NOON 1984 ; BROWN 1986 ). The consensus branchpoint sequence that can be derived from these 10 introns is YTNAY. In every intron, one consensus sequence was found that is predicted to initiate between 12 and 20 bases from the 3' splice junction (Table 1).
The transcriptional start site of C. cinereus mre11 was determined by 5' RACE. The first base following the G-tail sequence used in the procedure was located 108 nt upstream of the predicted translational start site (Fig 1A); this methionine codon was selected because it was the only one present in the proper reading frame in the first 165 amino acids (aa) of the predicted Mre11 polypeptide sequence. The 3' end was determined by sequencing a cDNA clone obtained by amplification from a C. cinereus cDNA library (see MATERIALS AND METHODS) and identifying the last base before the poly(A) tract. This was found to be 123 nt from the translational stop site (Fig 1A).
Expression of mre11 during meiosis and following gamma irradiation was examined by Northern analysis. Meiosis is naturally synchronous in C. cinereus; therefore, we examined the expression of mre11 in wild-type cap tissue from different meiotic stages. A 2.4-kb transcript was observed at karyogamy (K; Fig 2A, lane 1) and, of the timepoints examined, was most abundant at 6 hr after karyogamy (K + 6; the pachytene stage of prophase I; Fig 2A, lane 2). Analysis of phosphorimager data indicated a fourfold induction of mre11 transcript levels between K and K + 6. A twofold reduction in transcript was observed between K + 6 and K + 12 (Fig 2A, lanes 2 and 3). A transcript also was observed in vegetative monokaryon tissue, and a twofold induction of transcript level was observed in vegetative tissue harvested 4 hr after irradiation (Fig 2A, lanes 4 and 5). The gel was stained with ethidium bromide prior to blotting to confirm equal loading of the samples, based on the quantity of rRNA present (Fig 2B).
View larger version (100K):
In this window
In a new window
Download PPT slide
|
Figure 2.
Northern analysis of mre11 expression. (A) Wild-type meiotic RNA hybridized with a 1.8-kb fragment of mre11 genomic DNA (Fig 1A). Lane 1, cap tissue isolated at karyogamy. Lane 2, cap tissue at 6 hr after karyogamy. Lane 3, cap tissue at 12 hr after karyogamy. Lane 4, total RNA isolated from wild-type vegetative monokaryon tissue. Lane 5, total RNA isolated from wild-type vegetative monokaryon tissue 4 hr after irradiation. (B) Gel that was blotted for the hybridization shown in A, stained with ethidium bromide.
|
|
Characteristics of C. cinereus Mre11:
The predicted C. cinereus Mre11 polypeptide (CcMre11) consists of 731 aa and has a molecular mass of 80.6 kD and a pI of 4.86, values that agree well with those of Mre11 from H. sapiens (81 k D, PETRINI et al. 1995 ; pI of 5.4) and S. cerevisiae (72 k D, JOHZUKA and OGAWA 1995 ; pI of 5.4). The polypeptide shows high levels of similarity with other homologs of Mre11. Like other members of the family, CcMre11 is well-conserved throughout the N-terminal portion of the polypeptide but is quite divergent in the C terminus (Fig 1B). Pairwise alignments were performed between aa 18–478 of CcMre11 and the equivalent regions of three Mre11 homologs. MUS23 from N. crassa (WATANABE et al. 1997 ) shares 50.5% identity with CcMre11, Mre11 from H. sapiens (PAULL and GELLERT 1998 ) shares 47.1% identity, and Mre11 from S. cerevisiae (JOHZUKA and OGAWA 1995 ) shares 44.5% identity with CcMre11. The well-conserved putative phosphoesterase motifs required for the nuclease activities of Mre11 are also present in CcMre11 (Fig 1B, shaded in black; SHARPLES and LEACH 1995 ). A region containing putative overlapping 7-aa nuclear localization sequences was identified in CcMre11 from aa 598–607 (Fig 1B, boxed). The C to T transition identified in the mre11-1 mutant (discussed above) changes the predicted glutamine residue (aa 316; Fig 1B, asterisk) into a stop codon. If the mutant gene is translated, this change should result in a truncated polypeptide less than half the length of the wild-type protein.
Mitotic phenotypes of the C. cinereus mre11-1 mutant:
The mre11-1 mutant was isolated in a screen for strains that showed both enhanced sensitivity to ionizing radiation and defects in meiosis (VALENTINE et al. 1995 ). The mutation in mre11-1 is recessive; full complementation of both phenotypes is achieved when mre11-1 is crossed to wild-type strains (VALENTINE et al. 1995 ). The mre11-1 mutant demonstrates enhanced sensitivity to ionizing radiation. Following irradiation with 40 krad of gamma rays, small patches of mre11-1 mycelia inoculated on complete medium show no new growth after 2 days, whereas the growth of wild-type cultures is not inhibited. In more sensitive assays, we have shown that uninucleate mitotic spores (oidia) have 0.1% survival following 40 krad of radiation treatment (VALENTINE et al. 1995 ). To investigate whether mre11-1 is sensitive to UV radiation, oidia from mre11-1 and wild-type strains were screened for survival following UV irradiation. The mre11-1 mutant did not show a level of survival lower than that of the wild-type strain; e.g., at 100 J/m2, both strains showed ~5% survival. These results are similar to those obtained for other members of the C. cinereus epistasis group that includes mre11-1 (ZOLAN et al. 1988 ; RAMESH and ZOLAN 1995 ).
A dikaryon of C. cinereus can be induced to produce mushrooms under certain light and temperature conditions (ZOLAN et al. 1988 ). In this process, the dikaryotic mycelium forms hyphal knots, some of which differentiate into fruitbody initials. These in turn grow in size to form a mature mushroom. The culminating event of fruitbody formation is the successful completion of meiosis and spore formation in the mushroom cap; however, fruitbody development is a mitotic event that occurs even in the absence of meiosis (ZOLAN et al. 1988 ; PUKKILA 1994 ).
We have found that fruitbody development is slower, and fewer mushrooms are formed, in the mre11-1 mutant than in a wild-type strain. The number of mushrooms generated per mating inoculum and the number of days required for fruitbody maturation were determined for 10 independent crosses between either congenic wild-type strains (J6;5-4 x J6;5-5) or compatible mre11-1 strains. In wild-type crosses, the number of mushrooms produced ranged from two to five, with half of the matings yielding more than three mushrooms each. In contrast, the majority of matings (80%) between mre11-1 strains produced one or two mushrooms each, and no mating produced more than three mushrooms. The time required for fruitbody maturation was longer in the mre11-1 mutant as well. For wild-type crosses, the first day mature fruitbodies (opened mushrooms) were observed occurred on day 10 or 11 postinoculation. In contrast, most mre11-1 mushrooms appeared between 12 and 14 days postinoculation, and one mating required 20 days to fruit. This effect was not due to obvious differences in growth rate between wild-type and mre11-1 monokaryons nor to differences in the kinetics of mating between the two types of crosses (data not shown). The delay of fruiting in mre11-1 appeared to occur in the initiation of fruitbody formation, as the maturation of a fruitbody initial to a full mushroom occurs with the same kinetics in mre11-1 as in wild-type cultures; once a fruitbody initial appears, it usually takes 2–3 days to develop into a mature, opened mushroom.
Meiotic progression and chromatin condensation in the mre11-1 mutant:
We have exploited the natural synchrony of meiosis in C. cinereus to examine the meiotic defects of the mre11-1 mutant. DAPI staining of mre11-1 basidia revealed that karyogamy, the beginning of meiosis, occurred at approximately the same time as in wild-type cells (SEITZ et al. 1996 ); roughly 50% of nuclei showed nucleolar fusion 1 hr prior to the lights coming on, on a 16-hr light/8-hr dark schedule.
Meiotic prophase I is characterized by chromatin condensation and SC formation; full SC structures characterize the pachytene stage (VON WETTSTEIN et al. 1994 ). Wild-type C. cinereus meiotic nuclei are essentially all in pachytene at K + 6 (RAJU and LU 1970 ; PUKKILA et al. 1992 ; SEITZ et al. 1996 ; LI et al. 1999 ); the chromosomes appear as highly condensed, visibly paired figures by acridine orange staining (Fig 3A; SEITZ et al. 1996 ). In the mre11-1 mutant, acridine-orange-stained nuclei exhibited variable, but always incomplete, levels of condensation at K + 6, as assessed by the visualization of individual chromosomes. The majority of nuclei showed limited condensation superimposed upon a diffuse background, and short regions of apparent pairing were occasionally visible (Fig 3B). However, in some nuclei (~25% of all nuclei examined), the chromosomes appeared distinct and condensed with limited visible pairing, reminiscent of the leptotene stage in wild-type nuclei (Fig 3C; SEITZ et al. 1996 ). The ratio in which these phenotypes occurred was not consistent; on some slides, most nuclei appeared to be of the diffuse type, while on others, the condensed yet well-spread phenotype was predominant. The less-condensed nuclei did not seem to spread as well as those that were more condensed; this has also been observed for alleles of rad12 (RAMESH and ZOLAN 1995 ). No nuclei achieved wild-type levels of chromatin condensation at K + 6.
View larger version (50K):
In this window
In a new window
Download PPT slide
|
Figure 3.
Chromatin condensation in wild-type and mre11-1 meiotic nuclei. Meiotic nuclei from various timepoints were spread and were stained with acridine orange. (A) A pachytene wild-type nucleus at 6 hr after karyogamy (K + 6). (B) An uncondensed mre11-1 nucleus at K + 6. (C) A condensed, but unpaired, mre11-1 nucleus at K + 6. (D) A wild-type metaphase nucleus at K + 9. (E) An mre11-1 metaphase-like nucleus at K + 12. Arrows in A–C indicate the nucleolus. Bars, 1 µm. Numbers of mre11-1 nuclei scored were as follows: K + 6, 173; K + 12, 14.
|
|
Metaphase occurs in wild-type nuclei by K + 9 (Fig 3D). This stage is characterized by a greater level of condensation than that observed at K + 6, and it is preceded by a brief "diffuse stage" at diplotene in which the chromatin appears less condensed than at the pachytene stage (LU and RAJU 1970 ; SEITZ et al. 1996 ). Between K + 8 and K + 10, acridine orange staining of mre11-1 nuclei revealed a slight increase in condensation from the levels observed at K + 6; by K + 11, nuclei appeared more diffuse than those at K + 8 to K + 10 (data not shown). By K + 12, many nuclei appeared to achieve a more highly condensed, metaphase-like state (Fig 3E). The fraction of nuclei in metaphase varied among samples; one preparation had 30% metaphase-like spreads, while a different slide contained virtually 100% metaphase-like spreads.
By K + 12, wild-type basidia have completed the two meiotic divisions and contain four nuclei (RAJU and LU 1970 ; VALENTINE et al. 1995 ). By K + 15, most mre11-1 basidia exhibited a single diffuse or condensed nucleus (data not shown), as was seen by VALENTINE et al. 1995 at K + 12 using propidium iodide staining. Some cells contained deformed or stretched nuclei, and a few contained two nuclei, suggestive of anaphase-like activity; however, no further meiotic progression was ever observed (data not shown; VALENTINE et al. 1995 ). The mre11-1 mutant also makes few spore initials as observed by electron microscopy, and often only a single spore is initiated instead of four (VALENTINE et al. 1995 ). Macroscopically, mushroom caps appear to be devoid of mature spores, and quantitation of spore production confirmed that essentially no mature spores are made in mre11-1 mushrooms.
Synaptonemal complex formation and homolog pairing in mre11-1:
Surface-spread nuclei of mre11-1 were stained with silver nitrate and observed by transmission electron microscopy to examine the SC, a proteinaceous structure that forms between homologs during prophase I. In wild-type C. cinereus, axial elements (AE), which are components of the SC, form completely and subsequently synapse in the mature SC (HOLM et al. 1981 ; SEITZ et al. 1996 ). Wild-type meiotic nuclei exhibit maximal synapsis at pachytene (K + 6; Fig 4A; SEITZ et al. 1996 ). In contrast, neither AEs nor mature SC was complete in the mre11-1 mutant at the same timepoint. Some synapsis was evident in the majority (62%) of nuclei at K + 6, but not all AEs present were synapsed (Fig 4B). Additionally, regions of synapsis involving three different AE fragments were occasionally observed (data not shown). In some cases, possible pairing partner switches or aberrant synapsis was observed (arrow, Fig 4B). Some nuclei (26%) showed no synapsis, exhibiting only limited AE formation against a diffuse background (Fig 4C); these nuclei resembled wild-type leptotene nuclei (SEITZ et al. 1996 ), but the AEs were generally less complete and were more wispy in appearance. A small percentage (13%) of nuclei had no discernible structure at K + 6 (Fig 4D). By K + 10, no SC was visible in any nucleus; only a diffuse haze was observed (Fig 4E). Thus, as in wild-type nuclei (SEITZ et al. 1996 ), any SC that forms in mre11-1 nuclei is dispersed or degraded by late prophase.
View larger version (69K):
In this window
In a new window
Download PPT slide
|
Figure 4.
SC formation in wild-type and mre11-1 nuclei. (A) A wild-type pachytene nucleus at 6 hr after karyogamy (K + 6). The arrow indicates the nucleolus. (B–D) mre11-1 nuclei at K + 6, as described in the text. The arrow in B indicates aberrant synapsis. (E) An mre11-1 nucleus at K + 10. Bars, 1 µm. Numbers of mre11-1 nuclei scored were: K + 6, 82; K + 10, 13.
|
|
Some synapsis occurs in more than half of all mre11-1 nuclei. However, because SC formation is not dependent upon homology (MCCLINTOCK 1933 ; LOIDL et al. 1991 ; MAGUIRE and RIESS 1994 ), the synapsis data do not show explicitly whether pairing associations between homologous chromosomes are defective in the mre11-1 mutant. FISH was used to assay homolog pairing in mre11-1. DNA probes from either chromosome 8 or 13 were hybridized to spread meiotic nuclei of mre11-1 and were labeled with FITC and rhodamine, respectively. The chromosomes were counterstained with DAPI, and the spreads were examined by fluorescence microscopy. In previous studies of homolog pairing in wild-type nuclei, we defined pairing for each probe as one observable focus of staining or two foci situated 1.1 µm or less from one another (LI et al. 1999 ). In wild-type nuclei, pairing occurs rapidly, such that chromosome 8 was paired in 93% of meiotic nuclei examined just after nuclear and nucleolar fusion (Fig 5A), and >95% of probes from either chromosome 8 or 13 were paired at K + 6 (Fig 5B; LI et al. 1999 ).
View larger version (112K):
In this window
In a new window
Download PPT slide
|
Figure 5.
FISH analysis of wild-type and mre11-1 meiotic chromosomes. Chromosome 8 is indicated in blue, and chromosome 13 is in red. (A) A wild-type nucleus at 1 hr prior to karyogamy (K - 1), just after nucleolar fusion. (B) A wild-type pachytene nucleus at 6 hr after karyogamy (K + 6). (C) An mre11-1 nucleus at K + 6 in which the locus on chromosome 13 is paired, but that on chromosome 8 is not. (D) An mre11-1 nucleus at K + 6 in which both chromosomes 8 and 13 exhibit paired signals. The nucleolar signal present in this sample is background we occasionally observe. Bar, 1 µm.
|
|
Pairing was determined for mre11-1 nuclei at K + 1, K + 6, and K + 10. At each timepoint, a given probe was paired in about half of all nuclei examined, although at each timepoint, the locus on chromosome 13 was more likely to be paired than that on chromosome 8 (Table 2). Approximately 75% of nuclei demonstrated pairing for at least one of the two chromosomes (Fig 5C; Table 2), but only 30% showed pairing for both chromosomes examined (Fig 5D; Table 2). These results demonstrate that homolog pairing interactions stable to spreading are reduced, but not abolished, in the mre11-1 mutant.
| DISCUSSION |
|---|
MRE11 homologs have been identified in a wide range of organisms, including S. cerevisiae (AJIMURA et al. 1993 ), S. pombe (TAVASSOLI et al. 1995 ), N. crassa (WATANABE et al. 1997 ), and humans (PETRINI et al. 1995 ). Mre11 shows sequence similarity to the SbcD endonuclease from E. coli (SHARPLES and LEACH 1995 ), and it has been shown recently to have several activities related to processes of DSB repair (primarily by the mechanism of nonhomologous end-joining), telomere maintenance, and meiotic recombination (reviewed in HABER 1998 ). In conjunction with Rad50 and Xrs2 (in S. cerevisiae) or p95, the product of the NBS1 gene implicated in Nijmegen breakage syndrome (in humans; CARNEY et al. 1998 ), Mre11 is thought to act as a nuclease and/or structural component in the repair of DSB lesions (BRESSAN et al. 1998 ; TRUJILLO et al. 1998 ; USUI et al. 1998 ). Mre11 also is required for the formation and processing of the DSBs that initiate meiotic recombination in S. cerevisiae (JOHZUKA and OGAWA 1995 ; NAIRZ and KLEIN 1997 ; TSUBOUCHI and OGAWA 1998 ; MOREAU et al. 1999 ).
The mre11 gene of C. cinereus:
We isolated the mre11 gene from C. cinereus and showed that mre11 corresponds to the genetically defined rad11 locus. By identifying C. cinereus rad11 as mre11, we have linked rad11 and, by association, the other C. cinereus rad genes with meiotic function to pathways involved in DNA repair. The coding sequence of mre11 in C. cinereus is 2193 bases in length, and it is interrupted by 10 introns (Fig 1A). The C. cinereus mre11 transcript is 2.4 kb in size and is induced during meiosis and after treatment with ionizing radiation (Fig 2). During meiosis, a transcript is present at the K timepoint but is more abundant at K + 6. Defects in homolog pairing are observable at K in the mre11-1 mutant (e.g., Table 2), confirming that Mre11 is required early in meiosis. It is unclear whether the mre11 transcript levels reflect protein accumulation accurately; Mre11 protein could be more abundant prior to K + 6, even though transcript levels appear to peak at that timepoint. This idea is not without precedent, as it has been shown that transcript levels of S. cerevisiae RAD50 fluctuate independently of Rad50 protein levels (RAYMOND and KLECKNER 1993 ). Alternatively, while certain activities of Mre11 may be needed early in prophase, other functions, such as recombination initiation, could peak during pachytene in C. cinereus.
At 4 hr following gamma irradiation, the mre11 transcript level was induced only twofold over that of an unirradiated control culture (Fig 2A). The actual peak of induction following irradiation may occur at a different time; rad51 expression peaks at 2 hr after irradiation (YEAGER STASSEN et al. 1997 ). However, observed induction also depends on the baseline expression level, and an mre11 transcript was observed in the unirradiated monokaryon culture. This suggests that mre11 may be required for normal mitotic growth and, thus, is an essential gene in C. cinereus. If this is true, then mre11-1 likely is not a null mutant allele. Examination of both wild-type and mutant Mre11 protein levels will assist us in understanding how mre11 expression is regulated in meiosis and following irradiation.
In BLAST analysis, the C. cinereus Mre11 protein showed highest identity (50.5% over 461 aa) to an Mre11 homolog in N. crassa, MUS23 (WATANABE et al. 1997 ). However, MUS23 is truncated at the C terminus relative to other Mre11 homologs and may be functionally dissimilar to them (WATANABE et al. 1997 , and see below). Among other homologs, CcMre11 shows highest similarity to Mre11 from mammals. It shares the conserved features common to all Mre11 proteins: the four phosphoesterase motifs required for the nuclease activities of Mre11 and acidic residues at the C terminus (Fig 1B; e.g., SHARPLES and LEACH 1995 ; WATANABE et al. 1997 ).
The C. cinereus mre11-1 mutant likely does not make a full-length protein; it is predicted to truncate after aa 315 (Fig 1B). USUI et al. 1998 have shown in S. cerevisiae that a region at the C terminus (DNA-binding site B; Fig 1B) is required for meiotic DSB formation. A second region, DNA-binding site A (Fig 1B), is located between aa 410 and 420 in S. cerevisiae and is required for DSB processing. Both DNA-binding site A and functional phosphoesterase sequences appear to be necessary for nuclease activity in S. cerevisiae (USUI et al. 1998 ). Although the phosphoesterase motifs are not altered in the C. cinereus mre11-1 mutant, neither DNA-binding region is predicted to exist in the mutant polypeptide, based on alignments to ScMre11 (Fig 1B). Therefore, it is unlikely that mre11-1 is capable of forming or processing DSBs. It is also important to note that a region of CcMre11 containing a putative nuclear localization signal (Fig 1B) would be missing in the mre11-1 mutant. A similar sequence in the HD1 A mating-type protein of C. cinereus has been shown empirically to function in nuclear localization (SPIT et al. 1998 ). Therefore, Mre11 may not be targeted properly to the nucleus in mre11-1. However, USUI et al. 1998 have demonstrated that the first half of Mre11 in S. cerevisiae contains one or more regions that allow binding to Rad50. Because the first half of Mre11 should be present in the mre11-1 mutant, the mutant protein might achieve nuclear localization via an association with CcRad50. Thus, it is unclear whether the mutant Mre11 protein is capable of being targeted to the nucleus.
Mitotic phenotypes of the C. cinereus mre11-1 mutant:
The C. cinereus mre11-1 mutant has two demonstrated mitotic phenotypes, enhanced sensitivity to ionizing radiation and defects in fruitbody development. All S. cerevisiae MRE11 mutants identified show some degree of sensitivity to ionizing radiation or methyl methanesulfonate; where examined, these mutants have been shown to have defects in DSB repair (reviewed in HABER 1998 ). Therefore, the C. cinereus mre11-1 mutant likely is defective in repairing radiation-induced DSBs, although the mechanisms used for DSB repair in C. cinereus have not been characterized. The C. cinereus mre11-1 mutant does not show enhanced sensitivity to UV radiation. This renders mre11-1 more similar in phenotype to mre11 mutants from S. cerevisiae, which do not demonstrate UV sensitivity (e.g., AJIMURA et al. 1993 ; NAIRZ and KLEIN 1997 ; TSUBOUCHI and OGAWA 1998 ), than to UV-sensitive mutants of MRE11 homologs identified in two other fungi, N. crassa (mus-23; WATANABE et al. 1997 ) and S. pombe (rad32; TAVASSOLI et al. 1995 ). The mus-23 and rad32 genes are shorter than the other MRE11 homologs, including mre11 from C. cinereus. It has been proposed that, in addition to their roles in the repair of DSBs, mus-23 and rad32 may also function in a nucleotide excision repair pathway for UV-induced lesions that has been identified only in N. crassa and S. pombe (WATANABE et al. 1997 ; WILSON et al. 1999 ). Although this pathway has not been sought in C. cinereus, the phenotypic similarity of mre11-1 to the mre11 mutants of S. cerevisiae may imply that the mitotic repair function of CcMre11 is more similar to that of Mre11 in S. cerevisiae than to those of MUS23 in N. crassa, despite the higher sequence similarity of CcMre11 to MUS23.
Our data also indicate a role for C. cinereus Mre11 in fruitbody development. The mre11-1 mutant is a poor fruiter, requiring longer periods of incubation, which result ultimately in fewer mushrooms than are seen for wild-type crosses. Notably, only the initiation of fruiting, and not the process of fruitbody maturation from an initial, is affected in the mre11-1 mutant. Mre11 has been implicated in several general activities of vegetative cells. A rad32 mutant of S. pombe was demonstrated to have a 300-fold increase in minichromosome loss, relative to wild-type levels (TAVASSOLI et al. 1995 ). Mre11 also has been implicated in telomere maintenance (BOULTON and JACKSON 1998 ; NUGENT et al. 1998 ), detection of DNA damage (LEE et al. 1998 ; NELMS et al. 1998 ), and interactions between sister chromatids during G2 (MOORE and HABER 1996 ; HABER 1998 ). Although mre11-1 may not be a null mutant, deficiencies in any of these activities caused by the mre11-1 mutation might affect fruitbody formation in the mutant indirectly. For example, although no obvious differences in vegetative growth rate were observed between mre11-1 and wild-type cultures, the nuclear density or cell number may be lower in the mre11-1 mutant, resulting in a reduced ability of the mycelium to undergo differentiation into fruitbodies. Alternatively, if a vegetative culture must pass a DNA damage checkpoint before allowing fruitbody differentiation to begin, the mre11-1 mutant may require more time or may be less likely to pass such a checkpoint, resulting in fewer mushrooms that initiate more slowly than in wild-type cultures.
Meiotic phenotypes of the C. cinereus mre11-1 mutant:
The C. cinereus mre11-1 mutant is defective in meiosis, as examined by light and electron microscopy and by FISH. Chromatin condensation, synapsis, and homolog pairing are all defective in the mutant (Fig 3 Fig 4 Fig 5; Table 2). Although no nuclei exhibit wild-type levels of any of these processes, the phenotypes are quite variable. Several explanations for this are plausible. First, the mre11-1 mutation may be leaky. Although the mutant protein is predicted to be truncated to approximately half the length of the wild-type product, we have no evidence to suggest that it is entirely nonfunctional. If some product is present, there may be variability from cell to cell in the amount of functional protein present. Also, stochastic variability in meiotic progression may occur among cells in the absence of wild-type Mre11. Second, some phenotypic variability may derive from the assays themselves. We have used two-dimensional microscopy to examine structures and processes that occur inside the three-dimensional cell. The effect of spreading forces on weak interactions may result in observable differences among spread nuclei, when in reality these may represent very similar conditions in an intact cell.
In spite of the variability of chromatin condensation observed during prophase I (Fig 3B and Fig C), the mre11-1 mutant is able to progress to an event resembling metaphase I (Fig 3E). Thus, in mre11-1 nuclei, as was seen for the C. cinereus rad9-1 mutant (SEITZ et al. 1996 ), the processes of prophase and metaphase condensation have been uncoupled. Though the majority of mre11-1 nuclei appear to achieve metaphase, the number within a given experiment is variable. This is not true for rad9-1 nuclei; though none of the nuclei condensed properly during prophase I, consistently half of all nuclei in each experiment appeared to achieve metaphase condensation (SEITZ et al. 1996 ). Both of these phenotypes are in contrast to those of C. cinereus rad12 mutants, in which significant prophase I condensation occurs, but all nuclei arrest in diffuse diplotene (RAMESH and ZOLAN 1995 ). If the observed postdiplotene condensation in mre11-1 nuclei truly represents metaphase, then the missing activity of Mre11 is not required for metaphase condensation.
Like chromatin condensation, synapsis is defective in the mre11-1 mutant. In most nuclei, some synapsis is evident, but it is never complete (Fig 4B). The SC is a protein structure that in wild-type nuclei forms between aligned homologous chromosomes during prophase I. However, SC can form between regions of chromatin that are nonhomologous (LOIDL et al. 1991 ; MAGUIRE and RIESS 1994 ). Our analysis did not address directly whether regions of synapsis, as observed by electron microscopy, occur between paired homologs, as analyzed by FISH. However, the presence of pairing partner switches and aberrant synapsis (Fig 4B) implies that at least some nonhomologous synapsis is occurring.
This interpretation is similar to that proposed by NAIRZ and KLEIN 1997 for the mre11S mutant in S. cerevisiae. The authors concluded that because more SC structures were visible than could be accounted for by the amount of homolog pairing, some SC must be forming between nonhomologous chromosomes. The mre11S mutant can form, but cannot process, DSBs (NAIRZ and KLEIN 1997 ). NAIRZ and KLEIN 1997 suggested that a lack of DSB processing during meiosis might uncouple synapsis from the homology search. Their model suggested that, in the absence of DSBs, no synapsis can occur. Unprocessed DSBs would lead to partial, nonhomologous synapsis (as in the case of mre11S), while fully homologous synapsis would require the processing of DSBs. If programmed DSBs also are features of meiosis in C. cinereus, then the mre11-1 mutant would be predicted to make no DSBs because it is a truncation (see above). However, some synapsis (in certain cases, quite extensive synapsis; Fig 4B) does take place. Therefore, either some meiotic DSBs are made in the mre11-1 mutant, or the model for the role of DSBs in synapsis proposed by NAIRZ and KLEIN 1997 does not apply to C. cinereus.
Our analysis showed that homolog pairing stable to spreading is defective, but not abolished, in C. cinereus mre11-1 nuclei; 30% of nuclei demonstrate pairing of single loci on two chromosomes (Table 2). Additionally, the level of pairing is constant throughout meiosis, from shortly after karyogamy (K + 1) until K + 10. For this study we elected to use probes representing single loci on two different chromosomes, chromosome 8, which is ~2.5 Mb in size, and chromosome 13, which is ~1 Mb. In nuclei in which only one chromosome was paired, it was more likely to be chromosome 13 (Table 1). This may reflect a difficulty in forming or maintaining pairing associations on the larger chromosome (chromosome 8), or it may be due to the nature of the region of DNA probed on each chromosome; however, in wild-type nuclei neither chromosome demonstrated a bias in pairing ability at these loci (LI et al. 1999 ).
By examining a single locus on a given chromosome, rather than employing a chromosome painting approach in which a large region or the entire length of a chromosome is probed, we were able to examine unambiguously the uncondensed chromatin of the mre11-1 mutant. However, by relying on a single locus per chromosome, we most likely have underestimated the total amount of pairing on a given chromosome (WEINER and KLECKNER 1994 ; KLECKNER 1996 ); this was shown to be the case in our analysis of wild-type nuclei (LI et al. 1999 ). Additionally, if pairing interactions occur in mre11-1 nuclei, but are weak, they may be susceptible to disruption by the spreading forces used in these experiments. This effect likely is not due to the uncondensed or partially condensed state of the mre11-1 chromatin, as pairing interactions in uncondensed wild-type nuclei at early meiotic timepoints appear to be stable to spreading (Fig 5A; LI et al. 1999 ). For these reasons, the percentage of mre11-1 nuclei examined that have portions of chromosome 8 and 13 paired may be >30%.
Our results for pairing in the C. cinereus mre11-1 mutant are generally consistent with those obtained for related mutants in S. cerevisiae. WEINER and KLECKNER 1994 showed that a rad50
mutant paired at ~25% of the level seen in a wild-type strain, and a rad50S mutant paired with ~37% efficiency. The authors suggested that homolog pairing and meiotic DSB activity likely are coordinated early in meiosis, and that meiotic recombination events may help to stabilize pairing interactions. By parallel analysis, DSBs may not be formed at all in the C. cinereus mre11-1 mutant (assuming that programmed DSBs are a feature of meiosis in C. cinereus), resulting in a decreased ability of the mutant to stabilize pairing interactions. An alternative explanation is that homolog pairing occurs with wild-type frequency in the C. cinereus mre11-1 mutant, but the interactions in the mutant are more susceptible to disruption by spreading.
Our pairing data for mre11-1 in C. cinereus differ, however, from those presented by NAIRZ and KLEIN 1997 for the S. cerevisiae mre11S mutant. They found that pairing in a homozygous mre11S mutant at a pachytene timepoint is reduced to ~4% of the level seen in wild-type cells, which is lower than the 30% pairing at K + 6 seen for the C. cinereus mre11-1 mutant. It is unclear why these values are so different. One possibility is that the two mutations confer differences in pairing efficiency. Alternatively, probe choice may reflect variations in the ability of different regions of the chromosomes examined to pair efficiently. A third possibility is that pairing interactions within nuclei from the two organisms may show different susceptibilities to spreading techniques. However, we cannot rule out the possibility that Mre11 function is more critical for homolog
TOP
ABSTRACT
MATERIALS AND METHODS
