Description
Embryonic elongation of the
C. elegans
embryo is a powerful system for exploring the actin regulatory mechanisms that facilitate coordinated cell shape change. During embryonic elongation, the
C. elegans
embryo increases in length four-fold while reducing its diameter three-fold; this dramatic morphological change is accomplished by actomyosin-mediated epidermal cell shape changes, without any contribution from cell division or migration (reviewed in Labouesse et al., 2026). Proteins containing LIM domains are known to bind actin filaments under tension and have been implicated in repair of strained actin bundles (Smith et al., 2010; Smith et al., 2014; Sun et al., 2020). In
C. elegans
, LIM domain proteins have been investigated in striated body wall muscle (Lecroisey et al., 2013; Lecroisey et al., 2008; Qadota et al., 2007), synapses (Luo et al., 2014), and gastrulation (Slabodnick et al., 2023). We recently showed that two LIM domain containing repeat (LCR) proteins,
ZYX-1
and
TES-1
, play roles at cell-cell junctions in elongating
C. elegans
embryos (Lynch et al., 2022), and so we sought to examine other
C. elegans
LIM domain containing proteins in the context of embryonic elongation. One such protein that has not been characterized during elongation is
LIM-8
, despite available single cell RNAseq data indicating it is expressed in the epidermis during elongation (Packer et al., 2019).
We endogenously tagged all known
LIM-8
isoforms with mNeonGreen using CRISPR/Cas9 genome editing technology (
Fig. 1A
).
LIM-8
::mNG signal is first visible during early elongation, though it remains dim until the two-fold stage (
Fig. 1B,
1C). During the second, muscle contraction-dependent phase of elongation,
LIM-8
::mNG signal is bright and localized to circumferential F-actin filament bundles (CFBs; Priess and Hirsh, 1986) in dorsal and ventral epidermal cells, the mesh-like actin network in seam epidermal cells (Costa et al., 1997; Gally et al., 2009), and proximal actin networks associated with adherens junctions (Costa et al., 1998;
Fig. 1D,
1E). These observations are consistent with scRNA seq data for
lim-8
from a published dataset, which shows transcriptional expression of
lim-8
in AB- and C-derived epidermis and seam lineages 390-510 min after the first cell division (Packer et al., 2019). However, it is unclear from these data and ours whether the change in expression results from uniform expression during this time window, from increasingly specific localization, or both. Future experiments will be required to adjudicate these possibilities. During the muscle contraction-dependent phase of elongation, the
LIM-8
::mNG knock-in may have utility as an endogenously encoded reporter of F-actin organization, particularly as an alternative to LifeAct (Havrylenko et al., 2015) or the F-actin-binding fragment of the spectraplakin
VAB-10
(Gally et al., 2009) for assessing the organization and alignment of CFBs in the embryonic epidermis.
In adult worms,
LIM-8
::mNG signal is visible in the pharynx, as expected from previously reported promoter::GFP fusion experiments (Qadota et al., 2007;
Fig. 1F
).
LIM-8
::mNG is also expressed in the uterus, the rectal and phasmid systems, and the spermatheca (
Fig. 1G,
1H, 1I). Unlike in the epidermis of elongating embryos,
LIM-8
in the spermatheca does not accumulate on aligned actin filaments but instead has a speckled appearance and localizes basal to adherens junctions (
Fig. 1I
). This localization pattern was observed in both occupied and unoccupied spermathecae, suggesting that the tension state of the actin filaments in the spermatheca does not determine whether
LIM-8
is recruited. In the spermatheca,
LIM-8
also localizes to apical adherens junctions (
Fig. 1I
'). While
LIM-8
has been shown by antibody staining to localize to the M line in adult body wall muscle (Qadota et al., 2007), we were unable to detect
LIM-8
::mNG in body wall muscle under our imaging conditions.
While
lim-8
loss of function alone is not associated with overt phenotypes (Qadota et al., 2007), the strong actin localization of
LIM-8
during elongation, combined with previously identified roles for LIM-domain containing proteins
TES-1
and
ZYX-1
using sensitized backgrounds (Lynch et al., 2022), led us to consider whether
LIM-8
might be functionally redundant with other actin-binding proteins during elongation. Since both
tes-1
and
zyx-1
loss of function enhance the lethality of the hypomorphic α-catenin mutant
hmp-1
(
fe4
)
, we tested whether
lim-8
might also be a
hmp-1
(
fe4
)
enhancer. We observed no significant difference in embryonic lethality between offspring of
hmp-1
(
fe4
)
mothers and
hmp-1
(
fe4
)
mothers injected with
lim-8
dsRNA (
Fig. 1J
). We also tested whether a STOP-IN null allele (Wang et al., 2018) for
lim-8
would synergize with mutant alleles for other genes involved in elongation. The β
H
-spectrin
sma-1
is essential for efficient elongation, and loss of
sma-1
function, while not lethal, results in hatchlings that reach just over two-fold, rather than complete four-fold elongation (McKeown et al., 1998; Praitis et al., 2005). A
lim-8
;
sma-1
double null line did not show any embryonic lethality, and no obvious additional morphological defects were observed on plates (
Fig. 1J
). To ask whether loss of
LIM-8
might synergize with other LIM domain-containing proteins during elongation, we crossed the
lim-8
null allele with
tes-1
and
zyx-1
null alleles. Since the resulting line likewise did not display embryonic lethality, we considered whether another LIM domain containing protein,
ALP-1
, might be supporting elongation in this background through functional redundancy (
Fig. 1J
). As the existing Ahringer library clone targeted to
alp-1
covers sequence unique to the d isoform and the clone used to knock down
alp-1
expression in previous studies is targeted to
T11B7.5
, a gene found in an intron of
alp-1
, and therefore is not specific to
alp
-1 (McKeown et al., 2006), we generated a new RNAi clone targeting all known isoforms of
alp-1
but only sequences specific to that gene. Injection of
alp-1
dsRNA generated using this clone into the
lim-8
(sy1989);
tes-1
(
ok1036
);
zyx-1
(
cp419
)
line did not result in any embryonic lethality (
Fig. 1J
).
From these results, we conclude that
LIM-8
does not make major contributions to
C. elegans
embryonic elongation. Any role it may play is dispensable for proper elongation; it is possible that additional proteins act redundantly with
LIM-8
that we did not address in this study. Nevertheless, the differences in
LIM-8
localization between the epidermis and the spermatheca suggest that
LIM-8
may be a useful probe for interrogating the mechanisms regulating the association of LIM domain-containing proteins with actin filaments, which may be of biological relevance in other contexts.
Methods
Strains
Hermaphrodite
C. elegans
were maintained at 20°C on nematode growth media (NGM) plates seeded with
E. coli
strain
OP50
according to standard protocols.
N2
Bristol was used as wildtype (Brenner, 1974).
Strains used:
N2
:
Wildtype,
from
C. elegans
Genetics Center (CGC)
SU1073
:
zyx-1
(
cp419
)II;
tes-1
(
ok1036
)IV
from (Lynch et al., 2022)
PS9957
:
lim-8
(
sy1989
) III
from CGC (Wang et al., 2018)
SU1164
:
zyx-1
(
cp419
) II;
lim-8
(
sy1989
) III;
tes-1
(
ok1036
) IV
from this study
SU1168
:
lim-8
(
jc116
[
lim-8
::FLAG::mNG+LoxP]) III 2xOC
from this study
AZ30
:
sma-1
(
ru18
) V
from (Praitis et al., 2005)
SU1170
:
sma-1
(
ru18
) V 4xOC
from this study
SU1349
:
curIs11
[plin-26::LifeAct::mCherry];
lim-8
(
jc116
[
lim-8
::FLAG::mNG+LoxP]) III
from this study. Note: the LifeAct integrant likely resides on
LG X
hmp-1
(
fe4
) V
isolated from balanced line
SU287
(
hmp-1
(
fe4
)/
lon-3
(
e2175
) V) (Lynch et al., 2012) on 6/20/2026
Confocal imaging
Embryos were mounted on 10% agar pads in M9 buffer according to standard protocols (Hardin et al., 2022). Embryos were surrounded with
OP50
during mounting to paralyze them by hypoxia (Gally et al., 2009). Confocal images were collected at 20°C using a Dragonfly 500 spinning disc confocal microscope (Oxford Instruments) mounted on a Leica DMi8 microscope equipped with a 63X glycerol immersion objective and a Zyla camera and controlled by Fusion software (Oxford Instruments) using a slice spacing of 0.3 μm. Adult worms were mounted on 9% agarose pads in M9 buffer and imaged using the same settings as for embryos, except that Z stacks were collected with a step size of 0.5 μm.
CRISPR/Cas9 genome editing
To endogenously tag the
C. elegans
lim-8
locus with FLAG and mNeonGreen, we used the self-excising selection cassette (SEC) method previously described by Dickinson et al. (Dickinson et al., 2015). The guide sequence was cloned into pJW1219 using blunt-end cloning to produce pARM1. A repair template for the C terminus of
LIM-8
was generated using SapTrap cloning (Dickinson et al., 2018; Schwartz and Jorgensen, 2016) to assemble homology arms, linker, mNG, and self-excising cassette into the pMLS257 backbone to form pARM2. The forward primer for the 3' homology arm introduced a mutation to remove the PAM. pARM1 at 50 ng/µl and pARM2 at 15 ng/µl were injected into the gonads of adult
C. elegans
hermaphrodites along with coinjection markers pGH8 (4ng/µl), pCFJ90 (2ng/µl), and pCFJ104 (4ng/µl). Hygromycin selection for worms with the SEC, screening, and removal of SEC by heat shock were performed as described in Dickinson et
al. (Dickinson et al., 2015). Oligonucleotides used for cloning, genotyping, and sequencing are as follows (Sequence complementary to
C. elegans
DNA indicated in ALL CAPS, mutation introduced to remove PAM in red):
pJW1219 F with guide sequence added: ACATTCTAGAAAGAACTTCTgtttaagagctatgctggaaacag
pJW1219 R: caagacatctcgcaataggag
pJW1219 gRNA sequencing: gtgtgaaataccgcacaga
5' homology arm forward: ggctgctcttcgtggGCTACGTCATCGTCGTTGAGG
5' homology arm reverse: gggtgctcttcgcgcGACTTTGCTCAATTGCACACGATC
3' homology arm forward: ggctgctcttcgggtTAAGCTCTAAATAGTCCAGTCgCAGAAGTTC
3' homology arm reverse: gggtgctcttcgtacGCGGTCGGTTTTGAGTTTCC
SEC sequencing reverse for repair template: ttaatgtgagttagctcactcattagg
lim-8
knock in genotyping forward: GCTACGTCATCGTCGTTGAGG
lim-8
knock in genotyping reverse: CCACCTCCGCCGTATAATGC
lim-8
knock in sequencing: GGACGTGGTGCTGCAATGAT
Cloning
Approximately 0.5kb of the
alp-1
coding region, common to all known isoforms and not sharing sequences with any other known gene, was amplified from
N2
worm cDNA and cloned into the L4440 feeding RNAi vector using standard restriction cloning techniques to produce pARM4. Oligonucleotides used for cloning are as follows (Sequence complementary to
C. elegans
DNA indicated in ALL CAPS):
Forward primer for
alp-1
cDNA with HindIII restriction site added: gcaagcttGGAGTGATGGCTCGTGCAG
Reverse primer for
alp-1
cDNA with NheI restriction site added: gcgctagcCATTTGAAGGTGGTGGATTTGACGG
RNAi and plate counts
dsRNA was generated using T7 transcription kit using a
lim-8
cDNA clone (courtesy of G. Benian) and pARM4 as templates. dsRNA was injected into L4/young adult worms at 2 µg/µl for
alp-1
and 3 µg/µl for
lim-8
. Uninjected young adults and injected worms 20-24 hours post-injection were singled and allowed to lay for 20-24 hours. For
hmp-1
(
fe4
)
and
sma-1
(
ru18
)
lines this time was extended up to 28 hours due to gonad defects in
hmp-1
(
fe4
)
and body morphology defects in both lines, which slowed egg laying. Adult worms were then removed; 20-24 hours after removal of the adult worm, the number of hatchlings and dead eggs on each plate was counted. Each experiment was performed twice.
Statistical analysis
Percent lethality data, with the offspring of each worm considered as a biological replicate, were compared using Welch's t test (where two groups were to be compared), and one-way Brown-Forsythe and Welch ANOVA (where three groups were to be compared) to assess the significance of differences
between groups using GraphPad Prism software. N values for each group are included in
Figure 1J
.
scRNAseq analysis
Transcript levels for
lim-8
in epidermal cells were obtained using the summarized expression tool available in the online VisCello viewer (https://cello.shinyapps.io/celegans_explorer/) for the
C. elegans
embryo scRNA seq atlas developed by Packer et al. (2019).