Anti-
lin-14 antibodies detect a nuclear protein in specific somatic cells of late embryos and early to mid-L1 animals that quickly disappears during late L1 and later stages. Two major proteins of 76 and 67 kD are detected in immunoblots of protein extracts only from embryos and L1 larvae. These proteins do not correspond to the lin- 14a and
lin-14b gene activities (proposed by Ambros & Horvitz, 1987) nor to the
lin-14A and
lin-14B transcripts (see Burglin et. al., WBG this issue) because both proteins are detected in various
lin-14a-b+ and a+b- mutants. We have analyzed the temporal and cellular expression patterns of
lin-14 protein accumulation in various heterochronic mutants, i.e.
lin-4(
e912),
lin-28(
n719),
lin-29(
n333), and the double mutants
lin-4(
e912);
lin-28(
n719) and
lin-14(
n536sd);
lin-28(
n719).
lin-4, like lin- 14 semidominant mutants, exhibits a retarded phenotype, displays inappropriately high levels of the 76 and 67 kD
lin-14 proteins at all developmental stages, and shows inappropriate
lin-14 staining in nuclei at L2, L3, L4 and adult stages. Thus, the
lin-4 gene negatively regulates
lin-14 protein levels after larval stage 1. Northern blots and RNase protections suggest that this
lin-4 regulation of
lin-14 is post-transcriptional, as with the
lin-14 gain- of-function (gof) mutants; the temporal regulation of both
lin-14 transcripts is normal in a
lin-4 mutant (Wightman & Ruvkun, unpublished results). In contrast,
lin-28 animals have a steeper lin- 14 temporal gradient compared to wild-type; little to no
lin-14 protein is visible in the hypodermal and intestinal nuclei of early to mid-L1 animals while the ventral cord neurons, body wall muscle and nerve ring nuclei are comparable in intensity to wild-type. Thus, the
lin-28 gene activity positively regulates
lin-14 protein levels in specific cell lineages in L1 animals. We do not yet know if this effect is transcriptional or post-transcriptional. These data suggest that at least in the hypodermal and intestinal lineages,
lin-14 does not act downstream of
lin-28 but instead either up-regulates
lin-14 gene expression or controls the stability/activity of the
lin-14 protein. Because of the opposing effects of
lin-4/lin-14gof and lin- 28 gene activities on
lin-14, we examined the temporal regulation of
lin-14 expression in double mutants to characterize their epistasis relationship in molecular terms.
lin-4;
lin-28 and
lin-14gof;
lin-28 mutants show
lin-14 staining in late embryos to mid-L1 stages which fades to undetectable in late L1 and later stage animals. Therefore, while both
lin-4/lin-14gof and
lin-28 mutations affect
lin-14 protein levels, the
lin-28 mutation is epistatic to the effects of the
lin-4 and
lin-14gof mutations on the pattern of
lin-14 protein accumulation. This molecular epistasis is consistent with the genetic epistasis of
lin-28 to
lin-4 and
lin-14gof mutations.
lin-29 animals show no variation from the wild-type pattern of
lin-14 protein confirming that it is indeed downstream of
lin-14 in the epistasis pathway. We have also found that in wild-type animals, the down-regulation of
lin-14 protein levels is linked to actual developmental time, (not to GMT!). We have investigated this by allowing eggs to hatch and suspending animals at the L1 stage by starvation for up to 125 hours. There is no decline in
lin-14 protein levels in these arrested L1s by immunoblot or immunostaining analyses. Northern blots showed that lin- 14 transcript levels are similar in starved L1s compared to newly hatched L1s. However, in the presence of cycloheximide, starved L1s show a steep decline in
lin-14 staining at 24 to 48 hours. These data suggest that rather than being stable,
lin-14 protein is continually translated in these arrested L1s. If starved L1s are then fed and develop into L2s, the
lin-14 protein levels fade as occurs during normal development so that upon feeding, a signal to resume postembryonic development causes a decrease in
lin-14 protein levels. This decrease in
lin-14 is probably attained by negative post- transcriptional regulation of the
lin-14 3'UTR (see Burglin et. al., WBG this issue) The
lin-4 gene product is a good candidate for this negative regulatory activity. Our data predicts that the
lin-4 gene may become active after this L1 feeding switch. In starved L1s of lin- 28 and
lin-14(
n179ts25C) mutants, the
lin-14 protein staining fades faster than in wild-type suggesting that either a functional
lin-14 protein positively autoregulates
lin-14 gene expression or protein stability, and that the
lin-28 activity is necessary for this effect.