The Arabidopsis genome contains a single orthologue of MND1 (AtMND1) corresponding to the annotated gene ID No. At4g29170 and supported by a cDNA (Accession No. AA063855). The encoded protein shows 26% identity (47% similarity) to Mnd1 and is 230 aa in length which is close to that of Mnd1 (219 aa). Expression of AtMND1 was compared between rosette leaves and inflorescence using real time PCR (Fig. 1B). The results indicated a 9-fold higher expression of AtMND1 in reproductive tissues over leaves consistent with a possible role in reproductive development.

To examine the function of the AtMND1 gene we obtained bulked T4 seeds of an insertion line SALK_110052 29 carrying a T-DNA insertion in AtMND1 and identified 2 plants that were homozygous for the insertion. Both plants were found to be sterile (Fig. 2) whereas 14/14 plants that were not homozygous for the insertion were fertile. No defects in vegetative development were observed. The progeny of a single plant that was heterozygous for the insertion segregated 84:23 fertile:sterile consistent with a single gene recessive trait (0.5 > p > 0.25 for a single gene model; 0.001 > p for a two gene model). 31 plants at random were genotyped with respect to the presence of the insertion and all 4 that were found to be homozygous for the insertion were also sterile indicating that the phenotype was closely linked to the insertion (0.001 > p for an unlinked gene model). The mutant showed a greater than 100-fold reduction in AtMND1 expression suggesting that AtMND1 function is severely reduced and is probably null (data not shown). To determine whether sterility was caused by mutation of AtMND1, a 3.9 kb genomic fragment comprising the AtMND1 gene and its promoter region was cloned into a plant binary vector pCAMBIA1300 30 and transformed into Arabidopsis plants that were heterozygous for the Atmnd1 insertion allele, using in-planta transformation. 4 T1 transformant plants that were homozygous for the Atmnd1 insertion allele were identified out of 24 screened and all 4 were fertile, whereas among untransformed plants in a segregating population, 12 were identified to be homozygous for the Atmnd1 insertion allele and all were sterile (Fig. 2B) which demonstrates that the sterile phenotype is caused by mutation of AtMND1.

Reciprocal crosses were carried out between wild type and mutant to identify the developmental defects underlying sterility, and the results in Table 1 indicate that the mutant shows both male and female sterility. Observation of anthers showed that the pollen grains in the mutant were shrunken (Fig. 2F,G) and inviable (Fig. 2K,O) as determined by Alexander staining 31. Also the anther lobes remained below the level of the stigma as a result of reduced elongation of the anther filament. Likewise, examination of cleared ovules showed that in the mutant, postmeiotic development of the female gametophyte was arrested at the functional megaspore stage followed by degeneration of the functional megaspore (Fig. 2H–J,L–N; Table 2). About 0.6% of the developed ovules contained a mature embryo sac, which is consistent with the small number of seeds produced in the mutant. These results indicated that the mutant is male and female sterile due to both pollen and embryo sac formation being defective.

To obtain more detailed information on the expression pattern of AtMND1 in relation to the mutant phenotype we examined expression in inflorescence tissue sections by RNA in-situ hybridization using antisense RNA complementary to AtMND1 cDNA (Fig. 3). A basal level of expression was observed throughout reproductive tissues. In addition strong expression of AtMND1 was found within anther lobes at meiotic stages (Fig. 3A–E). The earliest increase in expression was detected in sporogenous cells at anther stage 4 32. Strong expression was observed in stage 5 anthers within microspore mother cells and also the tapetum. A high level of expression continued to be observed within meiotic cells during anther stage 6 and declined after meiosis. An increase in expression of AtMND1 was also detected within the megaspore mother cell in ovules although not as strongly as observed in microsporocytes (Fig. 3G).

The expression pattern of AtMND1 and the phenotype of the AtMND1 mutant are together suggestive of a defect in meiosis. We therefore compared meiotic prophase stages from the mutant and wild type using spread preparations of meiotic chromosomes following the method of Ross et al., 1997 33. The initial stages of meiotic prophase corresponding to early leptotene were seen to occur in the Atmnd1 mutant and thread-like chromosomes were apparent (Fig. 4). Association of the nucleolar heterochromatin present on chromosomes 2 and 4 as well as the synizetic knot, which is formed during late leptotene concomitant with pairing and the start of synapsis 34 could also be observed (Fig. 4A,F). However, abnormalities could be detected starting at the zygotene stage, both with respect to the appearance of the chromosomes and the organization of pericentromeric heterochromatin. Chromosomes in the mutant appeared less compact when compared to wild type and pericentromeric heterochromatin regions were more extended and unpaired than at the corresponding stage in wild type (Fig. 4B,G). During the course of zygotene, the differences became more pronounced, and synapsis was defective. The thickening of chromosomes along segments of their length representing synapsed regions, that characteristically appears during zygotene and is complete by pachytene, did not take place in the mutant (Fig. 4C,H). At diplotene the chromosomes appeared as a diffuse and fragmented mass with 10–12 separate spots of condensed pericentromeric heterochromatin (Fig. 4D,I). At diakinesis, separated chromosomes and fragments could be clearly distinguished (Fig. 4E,J). The sorting of these fragments at anaphase I and II was irregular and bridges could be observed (Fig. 4K–N,P–S). Following meiosis polyads that contained a variable number of fragmented chromosomes were formed (Fig. 4O,T). These gave rise to defective spores that did not form viable pollen. Female meiosis in the mutant was also defective and chromosome fragmentation was observed (Fig. 4U–W). These data suggested the possibility that the mutant is defective in homologous pairing and synapsis and the accumulation of fragments may arise from defects in repair of DSBs.

To examine homologous pairing we carried out FISH experiments (Fig. 5) using a telomere repeat based oligonucleotide probe that hybridizes strongly to the centromere of chromosome 1 but not to centromeres of the remaining chromosomes 35. In wild type, two well separated signals were observed at leptotene indicating that chromosomes were unpaired (Fig. 5A). A single signal was observed at late zygotene (46/46 nuclei) and pachytene (23/23 nuclei) indicative of pairing and synapsis having taken place (Fig. 5C,D). At diplotene, twin signals close to each other (6/6 nuclei) were seen indicative of centromere regions having desynapsed (Fig. 5E). These twin signals again merged and at metaphase I only a single signal was seen (data not shown). The Atmnd1 mutant showed two widely separated signals starting from leptotene and throughout all subsequent stages (58/61 nuclei for zygotene and 42/43 for pachytene) indicating that homologous pairing as well as synapsis was defective (Fig. 5F–J).

The chromosome fragmentation phenotype together with the absence of homologous pairing suggested that the Atmnd1 mutant was defective in recombination possibly due to defects in the repair of meiotic DSBs. AtSPO11-1 is one of three SPO11 homologues in Arabidopsis and is specifically required for recombination and synapsis during meiosis 3637. To test whether fragmentation was dependent upon Atspo11-1, a plant that was heterozygous for Atmnd1 was crossed to a SALK insertion line (SALK_146172) that was heterozygous for a T-DNA insertion in the seventh intron of Atspo11-1 29. F1 plants that carried both insertions were identified in the F1 and homozygous Atmnd1 Atspo11-1 double mutants were obtained in the F2. Analysis of male meiotic chromosome spreads of the double mutant and comparison to the Atspo11-1 single mutant indicated that the chromosome fragmentation phenotype of Atmnd1 was suppressed by Atspo11-1 (Fig. 6). Chromosomes in the double mutant did not undergo fragmentation after diplotene (Fig. 6B,C) and remained as intact univalents that segregated randomly at the first meiotic division (Fig. 6C,D,H,I). The double mutant phenotype resembled that of the Atspo11-1 mutant. Atspo11-1 is therefore epistatic to Atmnd1.

All the plants described in this study were Arabidopsis ecotype Col-0. The T-DNA insertion lines SALK_110052 and SALK_146172 used in this study were obtained from the Arabidopsis Biological Resources Centre, Ohio State University. Plants were grown as described previously 48.

Genomic DNA was extracted from the SALK T-DNA insertion lines using the method of Dellaporta et al.,1983 49. Presence of the T-DNA insert in Atmnd1 was confirmed by PCR using a left border outwardly directed primer (LB1) in combination with a gene-specific primer (AtMND1F1) flanking the site of insertion. Based on our analysis, there are at least two tandemly placed T-DNA inserts placed next to each other such that the orientation in the genome is LB-RB-LB. Junction fragments on either ends of T-DNA were amplified using primer combinations AtMND1F1 and LB1 and AtMND1R1 and LB2. Sequencing of the AtMND1R1-LB2 product revealed the presence of a 86 bp deletion within intron 7. Primers AtMND1F1 and AtMND1R1 were utilized to amplify the wild type allele. For the Atspo11-1 line the presence of the T-DNA insert was confirmed using the primer TSPO11R in combination with LB1 and wild type copy using TSPO11F in combination with TSPO11R. Homozygous insertion lines showed a phenotype that was the same as that of Atspo11-1-1 36.

A full-length genomic clone spanning 3891 bp (AGI coordinates 14382003–14385894) was amplified using primers MndF1 and MndR1 that incorporates restriction sites, BamH1 at 5' and EcoR1 at 3' end respectively. The amplification was done using TripleMaster PCR system (Eppendorf) as per the manufacturer's instructions. The resulting 3.89 kb fragment was cloned into the pGEM-T vector (Promega) followed by sequencing of the fragment. Restriction digestion with enzymes BamH1 and EcoR1 was performed to release the fragment, which was sub-cloned into the binary vector pCAMBIA1300. The fragment was mobilized into Agrobacterium strain AGL1 by tri-parental mating. The in planta transformation was carried out on heterozygous Atmnd1 plants by vacuum infiltration as reported earlier 50. The transformants were selected on medium comprising of 1% bacto agar, 1% sucrose, 1 mM KNO3 and 50 μg/ml of hygromycin B and genotyped by PCR using Atmnd_out_FG_Rev1 and Atmnd_out_FG_For1 primers.

Total RNA was isolated using Trizol (Sigma) as per the manufacturers protocol. cDNA was synthesized from 5–7 μg of RNA using the Superscript first strand synthesis system (Invitrogen) with Oligo dT primers. Real Time PCR reactions were done in a 10 μl volume comprising of primer, cDNA template and 1× SYBER Green PCR master mix (Applied Biosystems). GAPC was used as the internal normalization control. PCR was performed on the ABI Prism 7900 HT Sequence Detection System (Applied Biosystems) in a 384 well reaction plate according to the manufacturer's recommendations. Primers were MNDRTF1 and MNDRTR1 for AtMND1 and GAPRTF1 and GAPRTR2 for GAPC. Cycling parameters consisted of 2 minutes incubation at 50°C, 10 minutes at 95°C and 40 cycles of 95°C for 15 seconds, 57°C for 30 seconds and 67°C for 30 seconds. Each PCR reaction was performed in triplicate and the experiment were repeated twice. Specificity of the amplifications was verified at the end of each PCR run using ABI prism dissociation curve analysis software. Results from the ABI Prism 7900 HT Sequence Detection System were analyzed further using Microsoft Excel. Quantification of mRNA was calculated from threshold points (Ct values) located in the log-linear range of real time PCR amplification plots.

In situ hybridizations were carried out as described earlier 48. Anti-sense RNA specific to the AtMND1 gene was used as probe along with sense control. We used the full-length AtMND1 cDNA amplified from the cDNA clone using Nco1F and EcoR1R primers and subcloned into pGEM-T vector for strand specific probe synthesis. Floral stages are according to 51.

Developmental analysis of whole mount anthers and ovules was done after fixing and clearing the inflorescence in methyl benzoate as described previously 46. The slides were observed on a Zeiss Axioplan 2 Imaging microscope under DIC optics using a 40× oil immersion objective. Pollen viability was examined using the method of Alexander staining 31. Meiotic chromosome spreads were prepared, analyzed, and staged based on chromosomal morphology and with respect to the stage of the surrounding tapetal cells, according to Ross et al., 1996 33 with minor modifications as described in 52. Chromosomes were stained with DAPI (1 μg/ml) and observed on a Zeiss Axioplan 2 Imaging microscope using a 365 nm excitation and 420 nm long pass emission filter and a 100× oil objective. Images were captured on an Axioplan CCD camera using Axiovision (version 3.2) and processed using Adobe Photoshop 6.0.

Meiotic spreads were carried out as described above and FISH analysis was done according to 53 with incorporation of minor modifications. The hybridization mix was prepared with 5' FITC labeled probe FITC-(CCCTAAA)₆ 35 at a concentration of 5 μg/ml in 50% deionised formamide, 2× SSC and 10% dextran sulphate. The hybridization mix was denatured on a hot block for 3 minutes at 100°C and immediately cooled on ice. The slides were denatured separately with 100 μl denaturation mix comprising of 70% deionised formamide, 2× SSC and 50 mM sodium phosphate buffer pH 7.0 mounted under a 24 × 50 mm² cover slip and incubated at 80°C for 5 minutes. After the incubation, the slides were washed in ice-cold 70% ethanol for two minutes followed by dehydration in 70%, 90% and 100% ethanol respectively (2 minutes each) and air dried. Denatured probe (100 μl) was then applied to the slides and covered with a 24 × 50 mm² cover slip. Hybridization was carried out in a moist chamber for 18 hours at 37°C. Post-hybridization washes were performed in 2× SSC, pH 7.0 (two washes each for 5 minutes at room temperature) followed by 2× SSC for 5 minutes at 42°C. Chromosomes were counter stained with DAPI (1 μg/ml) in Vectashield (Vector Laboratories). Fluorescence detection was done on a Zeiss Axioplan 2 Imaging microscope equipped with epifluorescence illumination and distinct filters for DAPI and FITC using a 100× oil immersion objective. The images were captured with a Axioplan CCD camera using Axiovision software (Zeiss, version 3.2) and processed using Adobe Photoshop 6.0.

1. AtMND1F1 ACCGAAGAAGGGTGTAATTAGTCAGTC

2. AtMND1R1 ATTGTCGCAGTGTGAAGATGTTATCTG

3. MndF1 CAGGAGAATTCAAACCGAGAACATGAAACAGATCC

4. MndR1 GACGAGGATCCAATCATAGAAACAGACTTGGACC

5. Atmnd_out_FG_Rev1 CCTGGACCAGAAGAAGGTAAGGGTTTTG

6. Atmnd_out_FG_For1 GAGCTATTCACATGCTTAACAAGTTGCTAACAG

7. Nco1 F GCTCGCCCATGGCTATGTCTAAGAAACGGGGAC

8. EcoR1 R GCGGAGAATTCCTAAGCTTCATCTTGTACTAGCT

9. LB1 AACCAGCGTGGACCGCTTGCTGCAACTC

10. LB2 CAGGGCCAGGCGGTGAAGG

11. MNDRTF1 TCGATGATGATCTTGTTGCGAA

12. MNDRTR1 TCACACTGATCAACAAGTTCTGCt

13. GAPRTR2 CAGTCTTCTGAGTAGCAGTGATTGA

14. GAPRTF1 AGCACGAATACAAGTCCGACCT

15. TSPO11R ACTGTGATAACAATGCAGCGGTTCG

16. TSPO11F CAGCACAATCCATTGTGGACCGTGC