Experimental Study of the Mechanism of Recombination
In the middle of the 1930s most of the evidence for the chiasmatype theory had been obtained. (The final stroke to complete the picture was added by Taylor’s experiments; see 4.8.) However, since it was restricted to the level of cytological observations, it could not answer the question of the causes of the breaks in the chromatids which it postulated and which lead to the formation of chiasmata and of crossing-over. Since there were no direct experimental approaches to the study of the mechanism of crossing-over at that time, there was plenty of opportunity for various types of speculation. The hypothesis which enjoyed the greatest popularity was that of Darlington, which claimed to give a complete and final explanation of all the phenomena connected with crossing-over (synapsis, chiasma formation, and interference) on the basis of the assumption that the cell embarks prematurely (compared with mitosis) on meiotic division, as a result of which chromosomes which have not yet divided conjugate because of their inherent “pairing urge.” Replication of chromosomes, which begins after cyto-logically detectable synapsis, creates mechanical stresses in the helically coiled homologues, leading to breakages (1.13). According to Belling, recombination of genes is also caused by replication, but the recombinant chromatids arise, not mechanically, but through late replication of the chro-moneme compared with replication of the chromomeres (1.14).
Unable to display preview. Download preview PDF.
- This concept was developed original in order to explain a number of special features observed in the structure and behavior of lampbrush chromosomes (see below). Briefly it can be described as follows: each unit of a genetic function in the chromosome consists of one master-copy. To apply this concept to such well established facts as the existence of not more than two alleles of one gene in the diploid organism, and the fact that rf values between sites in neighboring genes may be less than between sites in the genes (Calef, 1957), it is postulated that: 1) any mutations arising in the slave-repeats are quickly corrected back to the original nucleotide sequence of the master-copy; 2) the slave-repeats do not take part in recombinations and they are excluded from the synaptincmal complex (4.6), but they form chromomeres and, subsequently, lampbrush loops. This hypothesis is supported by the fact that chiasmata are never observed in the loops (Callan, 1960). Any changes arising as the result of recombinations or mutations in the master-copy arc later reproduced in the slave-repeats (for the mechanism of correction, see 5.6) (Callan and Lloyd, 1960; Callan, 1967; Whitehouse, 1967b).Google Scholar
- The nature of injuries in DNA induced by these agents was established at the same time (Szybalski and Lorkiewicz, 1962; Strauss et al., 1968; survey: Zhestyanikov, 1969). Repair of single-stranded breakages takes place with the participation of the same enzymes as in the case of UV injuries, with the exception of those which operate at the excision stage (Howard-Flanders and Boyce, 1966; Cleaver, 1971). For this reason it is virtually impossible to distinguish between uvr cells and the wild type as regards their sensitivity to x-rays. Most breaks are healed by polymerase III (Youngs and Smith, 1973) and polymerase I (Town et al., 1971; Laipis and Ganesan, 1972; Pauling et al., 1972; Jacobs et al., 1972) so quickly that they cannot be detected by the use of the technique of McGrath and Williams (1966). This process takes place even in a buffer and it is accompanied by the insertion of 1 to 3 nucleotides (Painter and Young, 1972; Worthy and Elper, 1972). Any residual single-stranded breaks are repaired by the slow recA/exrA system (Kapp and Smith, 1970; Sedgwick and Bridges, 1972), either by inducing exchanges (see below) or without them (Bridges, 1971).Google Scholar
- * Conventionally lethal mutations of two types are known. The first type includes “opal,” “amber,” and “ochre” mutations. They are connected with the appearance of nonsense codons in the genetic text, leading to termination of the translation process (synthesis of the polypeptide chain). In some strains of E. coli there are suppressor genes which code a modified transfer RNA enabling translation to be resumed. These mutants can reproduce normally on these (permissive) Su (am) and Su (ochre) strains. The second type of conventionally lethal mutations includes thermosensitive mutants (ts). These ts mutations have a protein structure which is so modified that they are unable to function at raised temperatures (42 C). At ordinary temperatures the ts mutants of the phages develop normally. The rII mutations which can grow only on rex~ mutants of E. coli B can also be included among the conventionally lethal group (Gussin and Peterson, 1971).Google Scholar
- *The common nature of the mechanisms of formation of chromosomal aberrations and crossing-over is emphasized by the observations of Schacht (1958). In his experiments on Drosophila males (in which crossing-over does not normally take place) irradiation of the cells at presynaptic stages induced crossing-over, but at postsynaptic stages it induced chromosomal aberrations (translocations). The breakages arising in the chromosomes evidently lead to homologous exchanges (crossing-over) if synapsis is present, but to nonhomologous exchanges (translocations) in its absence. It is also remarkable that spontaneous breakages take place in the region of chiasma formation in species with localized chiasmata (Walters, 1956).Google Scholar
- † Unfortunately it is difficult to use UV-irradiation in experiments on Drosophila and similar objects because of its low penetrating power (Browning and Altenburg, 1964; Proust and Prudhomme, 1968).Google Scholar
- *Besides the experiments of Schacht and of Howard-Flanders mentioned above, other evidence of the reparative function of crossing-over is given by results obtained by Sparrow (1952). His experiments on Trillium showed that the number of fragments counted in metaphase of haploid mitosis of micro-sporocytes is increased tenfold after irradiation in postpachytene stages of meiosis compared with that observed after irradiation of cells at the pachytene stage. With respect to the two types of reparative recombination in yeasts (postreplicative and “dark,” i.e., repairing single-stranded breaks), the papers by Hunnable and Cox (1971) and Fabre (1971,1972) should be consulted. Multiple reactivation in phages evidently also takes place on account of recombination repair (Rayssiguier, 1972).Google Scholar
- † Conclusions regarding the polarity of chromatid subunits, reflecting the polarity of the DNA strands, are drawn from the study of spontaneous (4.3) and x-ray induced (Brewen and Peacock, 1969) exchanges between titium-labeled chromosomes.Google Scholar
- †Nonhomologous reciprocal exchanges of any type, leading to chromosomal aberrations evidently take place at regions of accidental homology. Under these circumstances hybrid genes must be formed and the products of the exchange will be nonviable or mutant. Since not a single mutant was found among hundreds of translocations and inversions, these workers (Berg and Curtiss, 1967) concluded that nonhomologous exchanges take place at the boundaries of the genes (compare Fan, 1969). In any case they do not depend on rec-systems (see the detailed survey by Franklin, 1971a). the frequency of deletions is increased in polA - (Coukell and Yanofsky, 1971).Google Scholar