Worm Breeder's Gazette 13(2): 44 (February 1, 1994)
These abstracts should not be cited in bibliographies. Material contained herein should be treated as personal communication and should be cited as such only with the consent of the author.
The ratio of X chromosomes to sets of autosomes (X/A ratio) is the primary signal for sex determination in C. elegans. Diploid animals with one X chromosome are male while those with two X chromosomes are hermaphrodite. Is the sex-determination signal the dose of a small set of specific X-linked genes, as in Drosophila, or are there a multitude of signal elements dispersed throughout the X chromosome? Analysis of XO animals with a duplication of a portion of the X chromosome, or of XX animals heterozygous for a deficiency, may reveal regions of the X chromosome whose dosage strongly affects sex determination: these regions should harbor signal elements. Because the X/A ratio is also the signal for the essential process of X-chromosome dosage compensation, we expect that changing the dosage of signal elements could result in sex-specific lethality.
Duplications and/or deficiencies have been isolated for much of the X chromosome; their analysis leads to the conclusion that for the regions covered, their dosage can be changed from one to two in XO animals and/or from two to one in XX animals with no deleterious effects. Therefore either the signal is dispersed, or the signal elements lie outside the duplicated or deleted regions. Most of these duplications and deficiencies were isolated however, in ways that would necessarily bias against including regions whose dosage strongly contributes to the sex-determination signal. Because of this bias, we decided to isolate new X-chromosome duplications, using methods that would allow us to isolate them regardless of their effect on the sex or viability of XO animals. We isolated new duplications in the unc-9 region and in the unc-2 region of X. Two different methods were used: for the unc-9 region, we first isolated a very large duplication, yDp4 ,in an XO animal. The XO yDp4 animal was a self-fertile intersex. We then isolated smaller derivatives of yDp4 that still retained unc-9 by screening for Lon non-Unc animals in a lon-2 unc-9 XX background. For the unc-2 region of X we used an alternate approach. We reasoned that a null mutation in the sdc-2 gene - a gene required for the XX mode of sex determination and dosage compensation - must be epistatic to any change in the upstream sex-determination signal; hence, an sdc-2 (null)XO animal should be male regardless of its dosage of signal elements. Therefore we could use Bob Herman's original method of isolating X-chromosome duplications in males, but in an sdc-2 (null)mutant background. We mutagenized wild-type males, mated them to unc-2 sdc-2 (null)hermaphrodites (which were kept alive by an unstable extrachromosomal array carrying the wild-type sdc-2 gene), and screened for non-Unc males in the F1 cross-progeny. In this way we isolated 7 free duplications of the unc-2 region that behave identically and appear to extend to the left end of X. yDp13 is one of these (see Figure). These duplications, like previously isolated duplications that contain the left end of X, recombine frequently with the X chromosome, making them difficult to maintain and characterize. Therefore we isolated an additional set of duplications that lack the left end of X, and hence the pairing site, using the same screen, except that we mutagenized mnDp66 (X;I); meDÄ5 males (see Figure) rather than wild-type males. We isolated 4 duplications from this screen; three free Dps, and one, yDp14 ,attached to an autosome. These duplications all include dpy-3 as well as unc-2 ,and presumably end at the meDÄ5 breakpoint.
The unc-9 region duplications differ greatly from the unc-2 region duplications in their effects on males. In the unc-9 region, duplications that include from sma-5 to unc-9 ( yDp5 ,9,11, 6 and 12) have no effect on male viability; however, males with the larger duplications appear less fertile than males with the smaller duplications. The three largest unc-9 duplications ( yDp7 ,8 and 10), with left endpoints between xol-1 and sdc-2 ,are deleterious to both males and hermaphrodites. Males with one copy of these duplications are infertile and dumpy, and hermaphrodites with two copies of these duplications are dumpy, scrawny, and sick. Therefore, these duplications probably include regions of X that must be present in the proper dosage in both sexes, rather than sex-determination
signal elements. The region near unc-2 ,however, is likely to contain a signal element. The duplications of unc-2 that extend to the left end of X are XO-specific lethal. Only 5-10% of XO animals with such a duplication become adults, and these "escaper" animals are invariably thin, small, slow-growing, but fertile males. The male-lethality is completely rescued by an sdc-2 (null)mutation, or by an array carrying the xol-1 gene fused to a constitutive promoter. yDp14 and the other duplications of unc-2 isolated from the meDÄ5 X chromosome, are much less lethal to males. Approximately 40% of yDp14 /+; unc-2 XO males survive, and they range in phenotype from nearly wild-type to thin, small, and slow growing males. We were able to reconstruct the severe XO-specific lethality typical of yDp13 by combining mnDp66 and yDp14 in the same strain. mnDp66 alone has no effect on viability of XO animals. Conversely we were able to rescue yDp13 lethality with meDf6 .That is, meDÄ6; yDpÄ3 males are wild type. Finally, yDp14 /+males are completely rescued to wild-type by sdc-2 (null).This region of X therefore is likely to contain a sex-determination signal element, as previously suggested by Jonathan Hodgkin (1993 Intl. C. elegans Mtg.). The requirement for mnDp66 in addition to yDpl4 for severe effects on viability of XO animals suggests either that there may be two or more signal elements near the left end of X, or that XO-specific lethality may require the duplication of a large non-specific region of X in addition to duplication of a signal element. We will determine which model is correct by combining yDp13 and yDp14 with duplications and deficiencies in other regions of X, and by suppressing the male-specific lethality of these duplications, thereby isolating mutants defective in the sex-determination signal.