Worm Breeder's Gazette 14(3): 27 (June 1, 1996)

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.

Does a Retinoic Acid Receptor Function as an X Signal Element?

Ilil Carmi, Jennifer Kopczynski, Barbara Meyer

Department of Molecular and Cell Biology U.C. Berkeley, CA 94720

        We are interested in understanding the primary sex determination
signal of C. elegans (the X/A ratio) and how it regulates its likely
target xol-1.   xol-1 is expressed at high levels in XO males in
response to a low X/A ratio and at lower levels in XX hermaphrodites in
response to a high X/A ratio.   To identify upstream regulators of
xol-1, and possibly components of the primary signal, JK screened for
mutations that increase the level of expression of a xol-1::lacZ
translational fusion in XX embryos (IWM abstract, 1993 and 1995).   y263
was one mutation isolated in this screen, and we consequently obtained a
second allele of the same gene, gm41.  We are provisionally calling the
gene defined by these mutations rox-1 (regulator of xol).  rox-1 is
X-linked, and in addition to increasing the level of expression of a
xol-1::lacZ transcriptional fusion in XX embryos, it also causes
XX-specific lethality.  Escapers of this lethality are Dpy, Egl and
masculinized, revealing defects in dosage compensation and sex
determination, as might be expected for a mutation that acts upstream of
xol-1 in the sex determination and dosage compensation pathways.  rox-1
mutations appear to be recessive.  rox-1/+ hermaphrodites are wild type,
and a duplication of the region, stDp2, rescues rox-1 hermaphrodites. 
In addition, a deficiency of the region, nDf19, behaves like the rox-1
mutant in that it increases the expression of the xol-1::lacZ reporter
gene in XX embryos.  The rox-1 phenotype is fully suppressed by a
xol-1(y9) loss-of-function mutation, providing further evidence that
rox-1acts upstream of xol-1.  We are conducting a genetic analysis of
rox-1 to determine if it is a counted element that forms part of the X/A
ratio, or instead a signal transduction gene that acts between the X/A
ratio and xol-1.  If rox-1 is a signal element, we expect that varying
its dose would change the perceived X/A ratio:  increasing the dose of a
signal element should increase the perceived X/A ratio and favor
hermaphrodite development, while decreasing its dose would favor male
development.  In addition, since the primary signal is multigenic, the
effects of a signal elements may be additive.  As indicated by the
experiments described below, the dose sensitivity of rox-1 and its
interactions with other components of the signal are consistent with its
being a signal component.
        Since loss-of-function mutations in rox-1 can kill
hermaphrodites, we tested if the converse is also true.  That is, can
extra copies of rox-1 cause XO-specific lethality?  We found that two
copies of stDp2 causes XO-specific lethality.  To determine if this
lethality is due to dosage compensation defects, we tested whether
mutations in sdc-2, a downstream gene in the dosage compensation
pathway, can suppress this lethality and found that the males are
rescued.  Subsequently, we found that rox-1 mutations suppress the
XO-specific lethality caused by stDp2.  This result demonstrates that
the lethality is due in part to the presence of extra copies of rox-1.  
        Since dose sensitivity is one characteristic of signal elements,
we tested whether rox-1 has other properties of a counted element.  We
already know of at least three signal elements at the left end of X and
therefore we asked if rox-1 acts additively with these elements to
determine the X/A ratio.  First, we tested whether mutations in rox-1
can counteract the effect of increasing the dose of other signal
elements.  yDp14 and yDp13 are duplications of signal elements at the
left end of X.  These duplications increase the perceived X/A ratio and
hence cause XO-specific lethality.  rox-1 mutations partially suppress
this male lethality.   Conversely, increasing the dose of other signal
elements can partially suppress the phenotype of rox-1 hermaphrodites.  
The duplications yDp14 and mnDp66 can suppress the XX masculinization
caused by rox-1 mutations.  In addition, the effects of rox-1 mutations
are synergistic with those of other signal elements.  yDf20 is a
deficiency that uncovers at least two signal elements at the left end of
X.  While yDf20/+  and rox-1/+ XX hermaphrodites are wild type,
rox-1/yDf20 hermaphrodites are Dpy, Egl and masculinized.  Similarly,
mutations in the signal element fox-1 cause no detectable phenotype in
XX animals, but are completely lethal to hermaphrodites in combination
with rox-1 mutations.
        We have initiated a molecular characterization of rox-1.  Based
on genetic mapping and RFLP analysis, we mapped rox-1 to a narrow region
on the X chromosome near unc-115.  Using transformation rescue
experiments, we identified two cosmids from this region, F44A6 and
CO6F4, that rescue the rox-1 mutant phenotype.   According to the C.
elegans genome sequencing project, the region of overlap between these
two cosmids contains four open reading frames.  We have tested
individual subclones containing each of these open reading frames for
their ability to rescue rox-1 mutants.  One of the subclones, a 10Kb
fragment encoding a predicted nuclear hormone receptor homologue of the
retinoic acid receptor type rescues rox-1 hermaphrodites.  We are
pursuing experiments to verify that this open reading frame is
responsible for rescue.