Genetic and Molecular Characterization of the w Locus

In the mid-1970s, ideas about the mechanisms of genetic recombination were progressing. The "polarity" gradient affecting the region of the CR/CS and ER/ Es allelic pairs of mitochondrial DNA was interpreted as a site-specific conversion event at the locus in the crosses between and strains (Du-jon et al. 1974). It was postulated that a DNA break at the locus was responsible for the initiation of the gene conversion tract (see Fig. 1A). The direct demonstration of this break took 10 more years, however (see Sect. 5, below). For one thing, the molecular techniques were not available. For another, double-strand breaks were not popular at that time. The initiation of genetic recombination had just been explained by a single-strand repair mechanism originating at a single-strand nick, not a double-strand break (Mesel-son and Radding 1975).

During these years most of the progress on the a locus came from the genetic analysis of mutants. Back in 1970, Piotr Slonimski's lab had already noted the existence of CR mitochondrial mutants, isolated from strains, that showed no or low "polarity" in crosses (Coen et al. 1970). Additional mutants which, after extensive genetic analysis, appeared to be second site suppressors of another and specific chloramphenicol mutation, played a critical role in our understanding of the phenomenon (Dujon et al. 1976) and, subsequently, in the discovery of the first homing endonuclease (see Sect. 6, below). They were also arising only from parental strains, never from the ones. It was as if the allele could be inactivated by mutation but not the allele. Mutants that had lost the a locus were called mn for neutral. Despite the extensive genetic analysis of the mn mutants, the nature of these mutations could never have been solved without the help of molecular methods (see Fig. 2).

Introns were discovered in 1977; unfortunately, not in yeast mitochondria where, retrospectively, they are so obvious! It took less than a year, using the then novel molecular mapping methods, to understand that the difference between and mitochondrial DNA resulted from the presence of a ca. 1.1kb insert in the former, absent from the latter (Bos et al. 1978; Faye et al. 1979). This insert, located within the large rRNA coding gene, had the characteristics of an intron as judged from electron micrographs of RNA/DNA hybrids.

I sequenced the intron of the yeast mitochondrial rRNA gene in Walter Gilbert's lab at Harvard University where I had just arrived in the fall of 1978 as a young visiting scientist, eager to learn molecular methods. Alan Maxam and Walter Gilbert had just developed a new technique to sequence DNA, and this a HETEROSEXUAL PAIRING EVENT

SEQUENTIAL DISSYMETRICAL MOLECULAR

GENE CONVERSION PROCESS PRODUCTS

Polar region

laboratory was very interested in introns. The a intron was appealing in that it did not follow the emerging theories of the time. It was associated with this strange phenomenon of "polarity", which it seemed responsible for, and during which it was propagated. The intron is 1143 bp long and, together with its flanking exons and the corresponding regions of the intron-less gene in an strain and an mn mutant, I sequenced less than 3000 bp in total. This, however, took me nearly a year of work, given the cloning and sequencing methods of the time. Nucleotides were spelled out one by one from gel electrophoresis lanes, sequenced fragments were assembled and verified manually. However, my efforts were extremely rewarding (Dujon 1980). First, the nature of the and CR mutations was solved (Fig. 2). Each corresponds to a single base-pair substitution in a highly significant location of the rRNA molecule. The proximity of the mn mutation to the intron insertion site was remarkable (only 3 nt away), explaining why such mutations never appeared from the mitochondria due to their possible interference with RNA splicing. Second, the intron was inserted in a region of the rRNA gene corresponding to one of the most highly conserved regions of the peptidyl transferase center of the ribos-ome, according to the secondary structure of the E. coli large rRNA molecule (for review, see Dorner et al. 2002). But my best reward was elsewhere. Within the a intron was a 708-bp open reading frame, able to encode a 235 ami-no-acid protein. This was unprecedented. If this reading frame were translated, it meant that an intron of an RNA gene would have a protein product. At this time, the sequence had no similarity in the meager databases of the time. What was this strange protein?

Fig. 1. From the polarity of recombination to the mechanism of intron homing, a The first model proposed to explain the "polarity" gradient affecting the mitochondrial markers closely linked to the w locus. Figure redrawn from Dujon et al. (1974). Each bar represents a double-stranded DNA molecule (solid bar w+, open bar or). During a cross, a hypothetical break was postulated to occur at the or site, initiating a gene conversion process that extends over the flanking genetic markers. All genetic markers were believed to lie on the same side of the w locus (compare to the reality shown in Fig. 2). b Schematic representation of the mechanism of intron homing. Each bar represents a DNA strand (solid bar in-tron-plus gene, striped bar intron-minus gene). The intronic sequence is symbolized by a thinner bar and the reading frame encoding the homing-endonuclease is shaded. The in-tron-encoded endonuclease (here, I-Scel) recognizes and cleaves the intron-less gene (arrowheads), hence initiating double-strand break repair that uses the uncleaved intron-plus gene as the donor of genetic information. The conversion of flanking exon sequences (and the genetic markers lying in them, letters) is explained by the degradation of the free DNA ends prior to or during the repair. The probability of exon conversion diminishes with distance from the break

Fig. 2. Discovery of the w locus and recognition site of I-Scel. a Pertinent mitochondrial mutations. Part of the conserved sequence of the yeast mitochondrial 21S rRNA molecule in the region of the peptidyl transferase center corresponds to the intron-less or RNA. The position of the 1143 nt group I intron in the w+ gene is indicated by an arrowhead. The "guide" RNA necessary for intron splicing (PI stem) is shown by a bar. Coordinates (italics) refer to positions in the mature rRNA molecule. Mutations conferring resistance to chloramphenicol (C^) are shown in gray circles. The mutation conferring resistance to erythromycin (ER) is circled. Data taken from Dujon (1980), Michel et al. (1982), Sor and Fukuhara (1982), and Dujon et al. (1985). b I-Scel recognition site. ^

Fig. 2. Discovery of the w locus and recognition site of I-Scel. a Pertinent mitochondrial mutations. Part of the conserved sequence of the yeast mitochondrial 21S rRNA molecule in the region of the peptidyl transferase center corresponds to the intron-less or RNA. The position of the 1143 nt group I intron in the w+ gene is indicated by an arrowhead. The "guide" RNA necessary for intron splicing (PI stem) is shown by a bar. Coordinates (italics) refer to positions in the mature rRNA molecule. Mutations conferring resistance to chloramphenicol (C^) are shown in gray circles. The mutation conferring resistance to erythromycin (ER) is circled. Data taken from Dujon (1980), Michel et al. (1982), Sor and Fukuhara (1982), and Dujon et al. (1985). b I-Scel recognition site. ^

The mitochondrial gene in the w+ strain (intron borders are shown in lowercase) is shown. The extension of the recognition site of the intron-encoded I-Scel endonuclease is shaded. Cleavage of each DNA strand is indicated by arrowheads. Coordinates are relative to the intron insertion site. Data taken from Dujon (1980), Dujon et al. (1985) and Colleaux et al. (1988)

0 0

Post a comment