RAD1 controls the meiotic expansion of the human HRAS1 minisatellite in Saccharomyces cerevisiae

Abstract

Minisatellite DNA is repetitive DNA with a repeat unit length from 15 to 100 bp. While stable during mitosis, it destabilizes during meiosis, altering both in length and in sequence composition. The basis for this instability is unknown. To investigate the factors controlling minisatellite stability, a minisatellite sequence 3' of the human HRAS1 gene was introduced into the Saccharomyces cerevisiae genome, replacing the wild-type HIS4 promoter. The minisatellite tract exhibited the same phenotypes in yeast that it exhibited in mammalian systems. The insertion stimulated transcription of the HIS4 gene; mRNA production was detected at levels above those seen with the wild-type promoter. The insertion stimulated meiotic recombination and created a hot spot for initiation of double-strand breaks during meiosis in the regions immediately flanking the repetitive DNA. The tract length altered at a high frequency during meiosis, and both expansions and contractions in length were detected. Tract expansion, but not contraction, was controlled by the product of the RAD1 gene. RAD1 is the first gene identified that controls specifically the expansion of minisatellite tracts. A model for tract length alteration based on these results is presented.

FIG. 1.

The HRAS1 minisatellite sequence and diagram of the HIS4 locus. (a) Sequence of the HRAS1 minisatellite and flanking regions. HIS4 locus sequences are lowercase and in boldface. Lowercase, underlined sequences were inserted during the cloning steps (see Materials and Methods). Capitalized sequences are unique HRAS1 sequences. The capitalized, boldface, bracketed sequence is one HRAS1 minisatellite repeat (type 1)—the underlined letters indicate the two sites of variability between repeats at +7 and +15. As indicated, there are 30 repeats in the human A1a allele utilized for this study. (b) Arrangement of the HIS4 locus in various diploid strains. DNY26 is a diploid strain bearing the wild-type HIS4 promoter sequence and the his4-lopc heterozygous insertion (dark vertical bar) disrupting the HIS4 coding sequence on one chromosome. PD81 is isogenic with DNY26, except that the promoter region has been deleted, as indicated by the open parentheses. DTK314 is also isogenic but contains the HRAS1 minisatellite sequence (white boxes) inserted in place of the region deleted in PD81. Dark gray rectangles are BIK1 coding sequences; light gray rectangles are HIS4 coding sequences; ovals are the TATAA box for HIS4.

FIG. 2.

The HRAS1 minisatellite affects HIS4 transcription. (a) Strains containing the minisatellite insertion grow on medium lacking histidine. DNY26 is wild type for the HIS4 promoter. PD81 has a deletion of the promoter region, while DTK314 contains the minisatellite insertion into the site of the promoter deletion at HIS4. Strains were grown on solid media with or without histidine. (b) HIS4 mRNA production is stimulated by the minisatellite insertion. Total RNA was isolated from DNY26, PD81, and DTK314. Northern analysis was performed using probes to HIS4 and ACT1. ACT1 was used as a control to normalize for RNA levels in each lane.

FIG. 3.

DSB formation during meiosis at HIS4. (a) The minisatellite insertion stimulates DSB formation during meiosis. DNA from rad50S derivatives of DNY26 (wild-type promoter [FX3]), DTK314 (minisatellite insertion [DTK472]), and PD81 (promoter deletion [DTK490]) was isolated from strains prior to meiosis and 24 h after meiotic induction. Southern analysis was performed using a HIS4-specific probe. Asterisks mark the bands resulting from meiosis-specific DSBs. (b) Processing events following DSB formation during meiosis. A diagram of the HIS4 locus in DTK314 is shown. The asterisks mark the approximate locations of the meiosis-specific DSBs. The arrows (labeled A, B, C, and D) underneath the diagram represent the direction of processing events subsequent to DSB formation.

FIG. 3.

DSB formation during meiosis at HIS4. (a) The minisatellite insertion stimulates DSB formation during meiosis. DNA from rad50S derivatives of DNY26 (wild-type promoter [FX3]), DTK314 (minisatellite insertion [DTK472]), and PD81 (promoter deletion [DTK490]) was isolated from strains prior to meiosis and 24 h after meiotic induction. Southern analysis was performed using a HIS4-specific probe. Asterisks mark the bands resulting from meiosis-specific DSBs. (b) Processing events following DSB formation during meiosis. A diagram of the HIS4 locus in DTK314 is shown. The asterisks mark the approximate locations of the meiosis-specific DSBs. The arrows (labeled A, B, C, and D) underneath the diagram represent the direction of processing events subsequent to DSB formation.

FIG. 4.

Tract length of the minisatellite alters during meiosis at high frequency. Representative whole-cell PCRs across the minisatellite tract in spore colonies (a, b, c, and d) derived from individual diploid cells are shown. The horizontal black bars mark the band size of the unaltered minisatellite tract. Asterisks indicate the spore colonies exhibiting altered-length alleles. The size standards are 100-bp ladders; as indicated, the brighter lower band is the 600-bp band.

FIG. 5.

A model for tract length alteration during meiosis. Individual single-stranded DNA molecules are depicted, with one homolog in black and the other in gray. The his4-lopc allele is shown as a dark bar, while the minisatellite tract is shown as a series of diagonal bars. In panel a, a DSB is shown between the minisatellite tract and his4-lopc. The asterisk marks the relative site of the second DSB observed (Fig. 3). In panel b, the ends of the break have been resected. Strand invasion, heteroduplex formation, and the beginning of repair synthesis are depicted in panel c. The enlarged portion shows the potential misalignment of minisatellite repeats following strand invasion. One repeat unit has formed a loop on the invading strand; other configurations are possible. The DSBR model (64) predicts invasion of the second end followed by repair synthesis and Holliday junction resolution (d). The SDSA model (reviewed in reference 49) predicts a strand displacement event followed by repair synthesis (e).