Mutations in several unusual dpy genes appear to affect the general level of X-chromosome expression, implicating these genes in control of X chromosome dosage compensation (Hodgkin, 1983, Mol. Gen. Genet. 192:452 Meneely and Wood, 1984, C. elegans Newsletter, Vol. 8, #1, p. 6; Meneely and Wood, 1985 C. elegans Meeting Abstracts, p. 90; De Long et al., ibid., p. 99; Meyer and Champness, ibid. p. 101). For example, dpy-21 mutations appear to increase and dpy-22 mutations to decrease X expression by genetic tests. To determine whether these effects are occurring at the RNA level, we have employed a quantitative RNA dot blot procedure to compare ratios of X-linked to autosomal transcripts between males and hermaphrodites and between dpy- 21 or dpy-22 and wild-type hermaphrodites, using gene-specific probes for the X linked genes act-4 and myo-2 and the autosomal genes act-1, p. 100). We have previously reported that adult males and hermaphrodites have the same amount of act-4 mRNA relative to act-1 or unc-54 mRNA (ibid.). We have now shown that the same is true for myo-2 mRNA relative to myo-1 mRNA levels. Expression of these two genes would seem unlikely to differ for physiological reasons between males and hermaphrodites, because both code for myosins of the pharynx, which is not a sexually dimorphic structure. Therefore, the equivalence of these ratios in the two sexes supports the view that dosage compensation of these genes operates at the RNA level. Conceivably, our results could be explained by special regulatory controls on act-4 and myo-2, operating outside of, or in addition to, the general X-chromosome dosage compensation mechanism. To test this possibility, we are attempting to determine whether these genes, like other X-linked genes analyzed genetically, show dose-sensitive expression when present in extra copies because of a duplication or fewer copies because of a deficiency. We have preliminary evidence from quantitative Southern blot analysis that act-4 is covered by the duplication mnDp57 (X;I), given to us by R. Herman. If we confirm this result, we shall assay act-4 expression in mnDp57-containing strains. We are testing known duplications for extra copies of myo-2; so far we have found that this gene is not covered by mnDp10 (X;I). We previously reported that in L1 animals a dpy-21 mutation did not affect the ratio of act-4 mRNA to unc-54 mRNA (ibid.). We have now shown that in adult hermaphrodites, the dpy-21(e428) mutation increases the ratio of act-4 mRNA to unc-54 mRNA by a factor of 1.53 +- .08 relative to the same ratio in N2. We are presently testing the effect of dpy-21 in adult animals on myo-2 mRNA levels. One possible reason for the difference in results between L1's and adults with act- 4 could be that the dpy-21 gene is not active prior to the L1 stage of development. A more trivial possibility is that many of the animals in our L1 preparations were starved and could have degraded some of their RNA's by the time they were harvested. We are attempting to test these and other possible reasons for the apparent difference.