Description
The
Caenorhabditis elegans
vulva is one of the most comprehensively studied organs, work on which has revealed how multiple signaling networks including Wnt/β-catenin, Notch, and EGF/Ras/MAPK cooperate to specify then sculpt a structure that enables egg-laying in adult hermaphrodite worms. The adult vulva is an organ derived from the epidermis that is comprised of 22 cells that assemble into seven stacked rings forming a bridge between the worm's uterus and the external environment (Schindler and Sherwood 2013). These develop from 3 of 6 equipotent vulval precursor cells (VPCs) that are induced by an epidermal growth factor (EGF/
LIN-3
) signal from the anchor cell which activates its receptor (EGFR/
LET-23
) in the closest VPCs, inducing them to undergo 3 rounds of cell division to form the vulva (Hill and Sternberg 1992). In contrast, uninduced VPCs only divide once before fusing with the epidermis, except for P3.p which fuses before dividing in 50% of worms. In the induced VPCs, EGFR/
LET-23
activates the Ras homolog
LET-60
, which then activates the MAP Kinase ERK/
MPK-1
through Raf/
LIN-45
and MEK/
MEK-2
(
Fig. 1A
).
MPK-1
phosphorylates the ETS transcription factor ELK1/
LIN-1
, inactivating it and inducing vulval differentiation. Ablation of any MAPK component prevents vulval development, whereas gain of function (gf) mutations in the pathway lead to ectopic induction of VPCs, leading to more than 3 VPCs adopting vulval fates and development of additional vulva-like structures, a phenotype termed Multivulva (Muv).
Notably, mutations activating the MAPK pathway are partially penetrant, with
let-60
(
n1046
gf)
worms displaying ~70% penetrance, the
let-60
(ga89gf)
temperature sensitive allele displaying ~55% penetrance at restrictive temperatures (Ferguson and Horvitz 1985; Eisenmann and Kim 1997; Yoder et al. 2004), and
lin-45
gf
overexpressing worms displaying ~77% penetrance of Muv (Yoder et al. 2004). In contrast, loss of function of the MAPK target
lin-1
leads to a 100% penetrance of Muv (Beitel et al. 1995). Various gene ablations can enhance or suppress the penetrance of Muv in
let-60
(gf)
worms without affecting wild type vulval development. Collectively, these suggest that from Ras to
MPK-1
, a signaling threshold must be met for a VPC to adopt an induced fate. This is similar to cancer-causing Ras
gf
mutations in mammalian models, where not every mutant cell adopts a cancerous fate and not every carrier individual develops cancer (Muñoz-Maldonado, Zimmer, and Medová 2019). Understanding how thresholds regulating MAPK output are established may enable better prediction of cell fates as well as elucidate how organs normally develop robustly in wild-type organisms, with relatively little variation in overall form.
The Muv phenotype of Ras/
let-60
(gf) mutants depends on normal insulin-like growth factor (IGF) signaling (
Fig. 1A
). Loss of the IGF receptor
daf-2
or the downstream kinases
akt-1
and 2
or
sgk-1
leads to near-complete suppression of the Ras/
let-60
Muv phenotype, but does not affect normal vulval development in wild type worms (Battu, Hoier, and Hajnal 2003; Hall 2014). This suppression at least partially depends on the downstream FOXO transcription factor
DAF-16
(Hall 2014; Subramanian et al. 2021). However,
let-60
(gf);
daf-2
;
daf-16
triple mutants are not fully rescued to
let-60
(gf)
single mutant levels, indicating that a
daf-16
independent effector of IGF signaling may be responsible for the remaining suppression of
let-60
(gf)
by loss of
daf-2
. Furthermore, preliminary evidence indicates that this
daf-16
independent arm may be downstream of
sgk-1
specifically, as
daf-16
RNAi restores Muv in
let-60
(gf);
daf-2
worms but not in
let-60
;
sgk-1
worms (
Fig. 1B
). Likewise, loss of
daf-16
rescues the resistance of
daf-2
worms to triggering DNA damage induced germ cell apoptosis but does not affect a similar
sgk-1
phenotype (Perrin et al. 2013).
Here, we sought to identify putative FOXO/
daf-16
independent regulators of MAPK signaling downstream of the IGFR/
daf-2
pathway. We began our analysis by RNAi knockdown of genes that were experimentally shown or computationally predicted to interact genetically or physically with
sgk-1
as annotated on WormBase (Sternberg et al. 2024). These knockdowns were carried out in several mutant backgrounds (
Fig. 1B
): (1)
let-60
(
n1046
gf)
worms to look for enhancement or suppression of the Muv phenotype; (2)
let-60
(
n1046
gf);
daf-2
(
e1370
lf)
worms to look for rescue of suppression by
daf-2
loss
;
and (3)
let-60
;
daf-2
;
sgk-1
(e538lf)
to look for rescue of suppression by
sgk-1
loss
.
Only genes present in the Ahringer lab RNAi library were screened (Kamath et al. 2003). Consistent with prior reports, among the modifiers we observed suppression of
let-60
(gf)
Muv by knockdown of
akt-1
,
akt-2
and
age-1
as well as enhancement of Muv by
daf-18
(Hall 2014; Nakdimon et al. 2012). Likewise, while
daf-16
RNAi restored Muv in
let-60
(gf);
daf-2
(lf)
worms, it had no influence on
let-60
(gf);
sgk-1
(lf)
worms
.
Notably, RNAi knockdown of the homeodomain interacting protein kinase HIPK/
hpk-1
significantly rescued Muv in
let-60
(gf);
sgk-1
(lf)
worms (
Fig. 1C
) but had no effect on
let-60
(gf)
single mutant worms,
let-60
(gf);
daf-2
(lf)
worms, or
let-60
(gf);
daf-2
(lf);
daf-16
(lf)
worms, indicating that an
sgk-1
specific pathway regulating Ras/MAPK signaling may involve
hpk-1
.
We thus predict that one function of
sgk-1
is to suppress
hpk-1
activity or expression, parallel to negative regulation of
daf-16
by
akt-1
and
2
. However,
hpk-1
is unlikely to explain the incomplete rescue of Muv in
let-60
(gf);
daf-2
(lf);
daf-16
(lf)
worms, as RNAi of
hpk-1
in this triple mutant background did not yield a further increase in Muv penetrance to
let-60
(gf)
single mutant levels; or, that its effect still requires active
daf-16
.
Factors which are inhibited by IGF signaling, like
daf-16
, should rescue Muv in the suppressed
let-60
(gf);
daf-2
(lf)
or
let-60
(gf);
sgk-1
(lf)
backgrounds when knocked down by RNAi. However, those that are positively regulated by IGF signaling would only show an effect in
let-60
(gf)
single mutant worms when depleted. We found several genes that phenocopied
daf-2
(lf)
and
sgk-1
(lf)
when knocked down in
let-60
(gf)
worms by RNAi. Among these were the
skn-1
/Nrf transcription factor and its target
gcs-1
/
GCLC (Tullet et al. 2008; An and Blackwell 2003), as well as the paraquat responsive transcription factor
pqm-1
which was previously shown to complement
daf-16
in regulation of longevity by
daf-2
and is regulated by
sgk-1
through phosphorylation (Tepper et al. 2013; Dowen et al. 2016). We found that RNAi knockdown or genetic ablation of
pqm-1
modestly suppressed Muv of
let-60
(gf)
worms (
Fig. 1D
and E
). One enhancer of the
let-60
(gf)
Muv phenotype, the ubiquitously transcribed on X homolog
utx-1
had no effect on
let-60
(gf);
daf-2
(lf)
or
let-60
(gf);
sgk-1
(lf)
worms, indicating that its effect also depends on having functional IGF signaling.
While
pqm-1
has no obvious mammalian homolog, we reasoned that we may nonetheless identify conserved regulators of Ras signaling among its transcriptional targets, as a ChIP-seq dataset for
pqm-1
was available. We thus performed a second RNAi screen in
let-60
(gf)
as well as
let-60
(gf);
pqm-1
(lf)
worms targeting genes which were bound in their promoter regions by
PQM-1
.
Among the top enhancers of Muv in both backgrounds, we identified
nhl-2
as previously reported (Hammell et al. 2009) as well as
ceh-18
which has known vulval development phenotypes in several other genetic backgrounds (Parry, Xu, and Ruvkun 2007). Notably, we identified an uncharacterized gene
F22D6.2
, with no known phenotypes, that enhanced Muv in
let-60
(gf)
worms but not in
pqm-1
(lf)
worms. Furthermore,
F22D6.2
was well-conserved in its AN1 domain encoding region with mammalian
ZFAND3, 5
and
6
genes (van Kempen et al. 2023). An existing allele,
ok1685
,
which is a partial deletion of the 5' region encompassing the first exon (encoding the N-terminal A20 Zinc finger domain) and a small part of the second exon yielded no phenotype when crossed to
let-60
(gf)
worms (
Fig. 1
G-I
). We tagged both wild type
F22D6.2
as well as the
ok1685
allele with FLAG at the C-terminus and confirmed that a protein is translated from the
ok1685
allele (
Fig. 1H
), which likely explains the lack of phenotype observed. Indeed, generating a full coding sequence deletion of
F22D6.2
by CRISPR/Cas9 (
on33
allele) phenocopied the
F22D6.2
RNAi enhancement of the Muv phenotype of
let-60
(gf)
worms (
Fig. 1
I, J
). Upon further work with
F22D6.2
(
on33
)
worms, we noted an unusual multigenerational germline feminization phenotype at higher temperatures, leading us to name this gene and its paralog
F56F3.4
as multigenerational sterility and temperature regulated,
mstr-1
and
mstr-2
, respectively. This, we studied in detail (Eroglu et al. 2024).
Methods
Caenorhabditis elegans
worms were maintained at 20
o
C on nematode growth medium (NGM) plates, and fed
OP50
Escherichia coli
as previously described (Brenner 1974).
HT115
E. coli
clones carrying RNAi vectors for the indicated gene (Kamath et al. 2003) were grown overnight (16h at 37
o
C) in liquid LB medium with ampicillin (100 μg/mL) and tetracycline (10 μg/mL) selection, then induced for 6h with IPTG (0.5 mM) before being concentrated 10-fold and seeded onto NGM plates supplemented with 0.25 mM IPTG and 25 μg/ml carbenicillin. Embryos obtained by bleach synchronization were plated on RNAi bacterial lawns (50-100 embryos per treatment), grown for 3-4 days at 20
o
C and then scored as adults for penetrance of the Muv phenotype. Control refers to RNAi towards the non-expressed pseudogene
Y95B8A_84.g
(Klosin et al. 2017)
.
CRISPR/Cas9 was performed as described (Eroglu, Yu, and Derry 2023), by injection of pre-assembled Cas9-crRNA with single-stranded DNA repair templates. Vulval induction was scored as described (Subramanian et al. 2021), by quantifying the number of VPC daughter cells at L4, assigning a score of 0.5 for 3-4 daughters and 1.0 for 5-8 daughters. Western blot was performed by picking 20 adult worms directly into 1x sample buffer (62.5mM Tris pH 6.8, 2% SDS, 50mM DTT, 10% glycerol, with 0.002% bromophenol blue), incubating at 95
o
C for 15 minutes, followed by electrophoretic separation on BioRad TGX pre-cast gels before being transferred to PVDF membranes. Blocking was for 30 mins with 5% nonfat milk powder in TBS with 0.1% tween-20, and detection was with 1:1000
mouse
anti-FLAG (M2, Sigma F3165) primary antibody in blocking buffer and 1:5000 goat-anti-
mouse
IgG (H+L) HRP conjugated secondary antibody (ThermoFisher 31430) in blocking buffer.