An Alternative to the Distributive Pairing Hypothesis in Drosophila.

GRELL ( 1962a) has made the remarkable observations that, in Drosophila, R*chromosomes involved in nonhomologous (heterologous) pairing in a variety of configurations are derived entirely from the no-exchange class of tetrad and that the frequency of that class is not altered appreciably. This is contrary to the expectation that heterologous pairing, where possible, would generally compete with homologous pairing and so increase the overall frequency of no-exchange tetrads in those chromosomes which may pair in either of these two ways. To account for these results, it has been hypothesized (GRELL 1962a,b) that the pairing of chromosomes may be divided into two different phases occurring sequentially. First in time comes the synaptic process between homologs, which may lead to exchange. Following this the chromosomes may fall apart, particularly if they have not been involved in an exchange, and then undergo pairing once again, either homologously or heterologously. The first synaptic process, which has been called exchange pairing, involves homologs only and generally results in crossing over, whereas the second (distributive pairing) is concerned with the segregation of the chromosomes, homologs or not, to opposite poles. This ingenious hypothesis adequately accounts for the experimental observations, yet there are several reasons for examining it closely. In the first place, it seems inefficient for an organism, after having evolved an intricate mechanism which insures the synapsis of homologous chromosomes, to permit the synapsed chromosomes to fall apart freely, (thus losing what must surely be one of the main benefits of synapsis, the apposition of homologs for metaphase orientation and normal anaphase disjunction), and would develop a second and different system (distributive pairing) to bring the chromosomes together once more, thereby accomplishing no more than what had already been accomplished more effectively by synapsis. In the second place, the opportunity for pairing and separation of nonhomologous chromosomes provided by “distributive pairing” will, under natural conditions, give rise to a high frequency of lethal zygotes (see, for instance, the measurements of COOPER, ZIMMERING and KRIVSHENKO [ 19551 on D. melanogaster or those of TERZAGHI and KNAPP [1960] on D. pseudoobscura) . Thus, there should be considerable selective pressure against the incorporation of such a mechanism. This note represents an attempt to provide a simpler model which will on the one hand, account for the data and, on the other, be free of the objections men-