ArticlePDF Available

A developmental gene regulatory network for C. elegans anchor cell invasion

Authors:

Abstract and Figures

Cellular invasion is a key part of development, immunity, and disease. Using the in vivo model of C. elegans anchor cell invasion, we characterize the gene regulatory network that promotes cell invasion. The anchor cell is initially specified in a stochastic cell fate decision mediated by Notch signaling. Previous research has identified four conserved transcription factors, fos-1a (Fos), egl-43 (EVI1/MEL), hlh-2 (E/Daughterless) and nhr-67 (NR2E1/TLX), that mediate anchor cell specification and/or invasive behavior. Connections between these transcription factors and the underlying cell biology that they regulate are poorly understood. Here, using genome editing and RNA interference, we examine transcription factor interactions before and after anchor cell specification. Initially, these transcription factors function independently of one another to regulate LIN-12 (Notch) activity. Following anchor cell specification, egl-43, hlh-2, and nhr-67, function largely parallel to fos-1 in a type I coherent feed-forward loop with positive feedback to promote invasion. Together, these results demonstrate that the same transcription factors can function in cell fate specification and differentiated cell behavior, and that a gene regulatory network can be rapidly assembled to reinforce a post-mitotic, pro-invasive state.
Content may be subject to copyright.
RESEARCH ARTICLE
A developmental gene regulatory network for C. elegans anchor
cell invasion
Taylor N. Medwig-Kinney
1,
*, Jayson J. Smith
1,
*, Nicholas J. Palmisano
1
, Sujata Tank
1,2
, Wan Zhang
1
and David Q. Matus
1,
ABSTRACT
Cellular invasion is a key part of development, immunity and disease.
Using an in vivo model of Caenorhabditis elegans anchor cell invasion,
we characterize the gene regulatory network that promotes cell invasion.
The anchor cell is initially specified in a stochastic cell fate decision
mediated by Notch signaling. Previous research has identified four
conserved transcription factors, fos-1 (Fos), egl-43 (EVI1/MEL), hlh-2
(E/Daughterless) and nhr-67 (NR2E1/TLX), that mediate anchor cell
specification and/or invasive behavior. Connections between these
transcription factors and the underlying cell biology that they regulate are
poorly understood. Here, using genome editing and RNA interference,
we examine transcription factor interactions before and after anchor cell
specification. Initially, these transcription factors function independently
of one another to regulate LIN-12 (Notch) activity. Following anchor cell
specification, egl-43,hlh-2 and nhr-67 function largely parallel to fos-1
in a type I coherent feed-forward loop with positive feedback to
promote invasion. Together, these results demonstrate that the
same transcription factors can function in cell fate specification and
differentiated cell behavior, and that a gene regulatory network can be
rapidly assembled to reinforce a post-mitotic, pro-invasive state.
KEY WORDS: EGL-43, FOS-1, HLH-2, NHR-67, Gene regulatory
network, Cell invasion, Cell cycle arrest
INTRODUCTION
Invasion through basement membranes (BMs) is a cellular behavior that
is integral to embryo placentation, tissue patterning during embryonic
development, and immune response to infection and injury (Medwig and
Matus, 2017; Rowe and Weiss, 2008). Increased cellular invasiveness is
also a hallmark of metastatic cancer (Hanahan and Weinberg, 2011).
Previous research has identified several cell-autonomous mechanisms
that are highly conserved across different contexts of BM invasion
(Kelley et al., 2014; Medwig and Matus, 2017). These include
localization of F-actin and upregulation of matrix metalloproteinases
(MMPs) to chemically degrade the BM (Hagedorn et al., 2009, 2013;
Kelley et al., 2019; Levitan and Greenwald, 1998; Morrissey et al., 2014;
Sherwood et al., 2005). There is also growing evidence that cells must
undergo cell cycle arrest in order to invade the BM (Kohrman and Matus,
2017; Matus et al., 2015). How these tightly coordinated programs are
transcriptionally regulated is not well understood.
As many contexts of cellular invasion occur deep within tissue
layers and are thus difficult to visualize, we utilize morphogenesis
of the Caenorhabditis elegans uterine-vulval connection as a
genetically tractable and visually amenable model for examining
cell invasion in vivo (Sherwood and Sternberg, 2003). During the
mid-L3 stage, a specialized uterine cell called the anchor cell (AC),
invades the underlying BM, connecting the uterus to the vulval
epithelium to facilitate egg laying (Fig. 1A,B) (Sherwood and
Sternberg, 2003). The AC itself is specified in a cell fate decision
event earlier in development, during the L2 stage, when two initially
equipotent cells diverge via stochastic Notch asymmetry, giving rise
to the terminally differentiated AC and a proliferative ventral uterine
(VU) cell (Fig. 1A,B) (Wilkinson et al., 1994).
Prior research has identified four pro-invasive transcription
factors (TFs) that function cell-autonomously to regulate AC
invasion (Fig. 1A). These include the basic leucine zipper TF fos-1
(Fos), the basic helix-loop-helix TF hlh-2 (E/Daughterless), the nuclear
hormone receptor nhr-67 (NR2E1/Tailless/TLX) and the zinc-finger
TF egl-43 (EVI1/MEL) (Hwang et al., 2007; Matus et al., 2010;
Rimann and Hajnal, 2007; Schindler and Sherwood, 2011; Sherwood
et al., 2005; Verghese et al., 2011). NHR-67 functions upstream of the
cyclin-dependent kinase inhibitor CKI-1 (p21/p27), to induce G1/G0
cell cycle arrest, which is necessary for AC invasion (Matus et al.,
2015). Independently of NHR-67 activity, FOS-1A regulates the
expression of the long isoform of egl-43 (EGL-43L) and downstream
effectors including multiple MMPs (zmp-1,-3 and -6) and a cadherin
(cdh-3) (Hwang et al., 2007; Kelley et al., 2019; Matus et al., 2010;
Rimann and Hajnal, 2007). HLH-2 independently regulates cdh-3
along with other targets and cytoskeletal polarity (Schindler and
Sherwood, 2011). Prior work has also suggested that EGL-43 and
HLH-2 may regulate NHR-67 based on conserved HLH-2 binding
motifs present in the nhr-67 promoter as well as perturbation
experiments using nhr-67 transgenic reporters (Bodofsky et al.,
2018; Verghese et al., 2011). Interestingly, three of these TFs
(egl-43,hlh-2 and nhr-67) also play key roles in the specification of
the AC and VU cell fates during the L2 stage (Attner et al., 2019;
Hwang et al., 2007; Karp and Greenwald, 2004; Sallee et al., 2017;
Verghese et al., 2011). How these four TFs function, independently
and together, to regulate both specification and invasive activity of the
differentiated AC is poorly understood.
Here, using new highly efficient RNA interference (RNAi) vectors
(Sturm et al., 2018), we identify novel roles for egl-43 and hlh-2 in
maintaining the AC in a cell cycle-arrested state. Additionally, we
generated GFP knock-in alleles for each pro-invasive TF using
CRISPR/Cas9-mediated genome engineering. These tools allow for
quantitative measures of endogenous protein in the context of native
chromatin architecture and potential distal enhancer elements that
may be misrepresented using traditional transgenic reporters
(Dickinson and Goldstein, 2016; Dickinson et al., 2013; Kim et al.,
2014). Using these alleles and improved RNAi constructs, we report
Received 2 July 2019; Accepted 25 November 2019
1
Department of Biochemistry and Cell Biology, Stony BrookUniversity, Stony Brook,
NY 11794-5215, USA.
2
Science and Technology Research Program, Smithtown
High School East, St. James, NY 11780-1833, USA.
*These authors contributed equally to this work
Author for correspondence (david.matus@stonybrook.edu)
T.N.M.-K., 0000-0001-7989-3291; N.J.P., 0000-0002-7992-4462; D.Q.M., 0000-
0002-1570-5025
1
© 2020. Published by The Company of Biologists Ltd
|
Development (2020) 147, dev185850. doi:10.1242/dev.185850
DEVELOPMENT
the onset of relative expression of pro-invasive TFs in the lineages
leading to the AC and dissect their molecular epistatic interactions.
We find that, prior to AC/VU specification, egl-43,hlh-2 and nhr-67
function independently to regulate an endogenously tagged lin-12::
GFP allele. This network appears to be assembled following
specification, with the addition of a fourth node, fos-1,and
activation of TF interactions. Furthermore, we characterized both a
cell cycle-independent circuit and a cell cycle-dependent feed-
forward regulatory circuit with positive feedback that is crucial for
maintaining the AC in an invasive, post-mitotic state. These findings
provide new insights into how a set of reiteratively used TFs can be re-
purposed from regulating cell fate specification to coordinating a
differentiated cellular behavior.
RESULTS
EGL-43 or HLH-2 RNAi depletion results in extra ACs
Our initial goal was to determine interactions between the four pro-
invasive TFs fos-1,egl-43,hlh-2 and nhr-67. To accomplish this, we
first generated new RNAi targeting constructs in the improved RNAi
vector, T444T, which includes T7 termination sequences to prevent
the generation of non-specific RNA fragments from the vector
backbone, increasing the efficacy of gene silencing over the original
RNAi targeting vector, L4440 (Fig. S1A) (Sturm et al., 2018). As
three of four pro-invasive TFs (egl-43,hlh-2 and nhr-67)also
function during AC specification (Attner et al., 2019; Hwang et al.,
2007; Karp and Greenwald, 2004; Sallee et al., 2017; Verghese et al.,
2011), we assessed TF depletions in a genetic background in which
only the AC and neighboring uterine cells are sensitive to RNAi
following the AC/VU decision (Matus et al., 2010). The uterine-
specific RNAi-sensitive strain was generated through tissue-specific
restoration of the RDE-1 Piwi/Argonaut protein in an rde-1 mutant
background using the fos-1 promoter, which is expressed specifically
in the somatic gonad during the late L2/early L3 stage of larval
development (Haerty et al., 2008; Hagedorn et al., 2009; Matus et al.,
2010, 2015). AC invasion was quantified as the presence or absence
of a BM gap, visualized by laminin::GFP, at the P6.p four-cell stage, a
Fig. 1. RNAi depletion of pro-invasive TFs leads to
defects in invasion and proliferation. (A) Schematic
depicting the reiterative use of TFs in AC specification
and invasion. (B) Micrographs depicting AC (magenta,
cdh-3
1.5
>mCherry::PLCδ
PH
) specification and BM
(green, lam-1>LAM-1::GFP) invasion over
developmental time. Brackets indicate 1° VPCs. Yellow
and white arrowheads indicate the BM breach and ACs,
respectively. Asterisks indicate the AC/VU precursors,
Z1.ppp and Z4.aaa. (C) Representative empty vector
control and TF-RNAi depletion phenotypes, with
multiple ACs from additional confocal z-planes shown
as insets. Arrowheads indicate ACs.
(D,E) Stacked bar charts showing the penetrance of
invasion defects, binned by number of ACs, following
TF-RNAi depletion (P<0.00001, Fishers exact test,
empty vector compared with each TF RNAi, n50
animals examined for each treatment). Because of the
penetrance of the zero AC phenotype in early hlh-
2(RNAi)-treated animals (D), an L2-plating strategy was
used to assess post-specification AC invasion
phenotypes (E). Scale bars: 5 μm.
2
RESEARCH ARTICLE Development (2020) 147, dev185850. doi:10.1242/dev.185850
DEVELOPMENT
developmental window in which 100% of wild-type ACs are invaded
(Fig. 1C, Table S1). We also assessed invasion several hours later,
following vulval morphogenesis, allowing us to distinguish between
delays in invasion and complete loss of invasive capacity (Fig. 1E,
Fig. S1D, Table S1). Depletion of each TF resulted in invasion
defects ranging from moderate penetrance (38%; fos-1)tohighly
penetrant (69-94%; hlh-2,egl-43 and nhr-67), when synchronized
L1-stage animals were plated on RNAi bacteria and scored at the mid-
L3 stage (P6.p four-cell stage) and early L4 stage (P6.p eight-cell
stage) (Fig. 1C-E, Fig. S1D). Consistent with previous work,
depletion of fos-1 resulted in single ACs that failed to breach the
BM. Depletion of nhr-67 resulted in proliferative, non-invasive ACs
at a high penetrance. Depletion of egl-43 phenocopied nhr-67,witha
highly penetrant defect of multiple, non-invasive ACs. Lastly,
depletion of hlh-2 resulted in pleiotropic phenotypes in the AC,
from animals with zero, one, two or even, in several cases, three or
more ACs (Fig. 1D, Fig. S1C). In all cases, the penetrance of AC
invasion defects was significantly increased using a T444T-based
RNAi targeting vector compared with L4440-based vectors
(Fig. S1A). Thus, use of the improved RNAi vector revealed
unreported phenotypes resulting from depletion of egl-43 and hlh-2
during invasion, phenocopying depletion of nhr-67 resulting in
multiple non-invasive ACs (Matus et al., 2015).
Because of the pleiotropy associated with hlh-2 depletion, which
affects AC specification and invasion (Schindler and Sherwood,
2011), we sought to rule out the possibility that the extra cdh-3-
expressing ACs are due to defects in AC specification. Depletions
using the traditional RNAi vector L4440 in the uterine-specific
RNAi-sensitive background is sufficient to bypass AC fate
specification defects that occur in genetic backgrounds in which all
cells are sensitive to RNAi (Matus et al., 2010). Specifically, loss of
HLH-2 prior to AC specification disrupts the lin-12(Notch)/lag-
2(Delta)-dependent signaling cascade, resulting in both pre-AC/VU
cells adopting a VU fate and therebyproducing a zero AC phenotype.
During AC specification, loss of HLH-2 results in upregulation of
pro-AC fate by activating lag-2/Delta resulting in both AC/VU cells
adopting an AC fate (Karp and Greenwald, 2004; Schindler and
Sherwood, 2011). Notably, hlh-2 depletion using the T444T vector
resulted in a greater occurrence of a zero AC phenotype compared
with the L4440 hlh-2 RNAi clone (Fig. 1D, Fig. S1A). This is likely
due to the improved efficiency of RNAi-mediated knockdown using
the enhanced T444T vector (Martinez et al., 2019), as the time period
between restoration of RDE-1 function and the timing of the AC/VU
decision appears decreased. Thus, to bypass these potential
specification defects due to increased RNAi penetrance for hlh-2
using the T444T vector, we depleted hlh-2 and the other TFs by
RNAi using an L2-plating strategy (Schindler and Sherwood 2011),
growing animals to the time of the L1/L2 molt on OP50 Escherichia
coli and then transferring them to TF-RNAi plates. L2 platings
resulted in a lower overall penetrance of AC invasion defects for all
TF-RNAi treatments (Fig. S1B), including hlh-2 (43% at the P6.p
four-cell stage; Fig. 1D) but all animals possessed an AC. Strikingly,
some animals possessed multiple cdh-3-expressing ACs at the P6.p
four-cell stage (Fig. 1C), and 10% of animals in the early L4 stage
possessed multiple ACs (Fig. 1C-E, Fig. S1C,D). As the number of
ACs increased over developmental time (Fig. 1D,E), these results
suggest that the multiple ACs observed could be arising through loss
of G1/G0-cell cycle arrest and nhr-67 activity (Matus et al., 2015).
Characterization of GFP-tagged alleles and RNAi penetrance
Next, we sought to examine the relationship of these four TFs in the
AC prior to and during invasion by combining gene knockdown with
endogenous GFP-tagged alleles. To accomplish this, we used
CRISPR/Cas9 genome-editing technology to knock in a codon-
optimized GFP tag into the endogenous locus of each TF (Fig. S2,
Tables S4-S6) (Dickinson and Goldstein, 2016). First, we examined a
minimum of 50 L1/L2-stage animals to determine the onset of
expression for each TF in the somatic gonad, which is derived from
two foundercells, Z1 and Z4 (Kimble and Hirsh, 1979). The two cells
that can stochastically give rise to the AC are the proximal great-
granddaughters of Z1 and Z4 (Z1.ppp and Z4.aaa). Examination of
L1- and L2-stage somatic gonads revealed that egl-43::GFP::egl-43
turns on first, in Z1 and Z4, and remains on in their descendants
through the time of AC/VU specification (Fig. 2A). As reported
recently (Attner et al., 2019), GFP::hlh-2 turns on at the second
division of Z1/Z4 cells in Z1.pp and Z4.aa. We also detected onset of
nhr-67::GFP in these cells in some animals, though it appears that
their expression is weaker in Z1.pp/Z4.aa compared with GFP::hlh-2,
but becomes robust in the AC/VU cells prior to specification. Finally,
we detected weak expression of GFP::fos-1 in L1 and L2 Z1/Z4
descendants that increased rapidly in the somatic gonad following AC
specification (Fig. 2A). In summary, consistent with previously
reported roles for egl-43,hlh-2 and nhr-67 in AC/VU specification
(Hwang et al., 2007; Karp and Greenwald, 2004; Verghese et al.,
2011), we detected the presence of endogenous GFP-tagged TFs in
Z1/Z4 descendants.
Next, we examined the expression patterns of TF::GFP alleles
following AC specification. All four TFs showed robust GFP
localization in the AC nucleus prior to and during invasion (Fig. 2B).
Additionally, the GFP-tagged strains had similar expression domains
in the somatic gonad during the L3 larval stage as previously reported
by traditional multi-copy array transgenes (Hwang et al., 2007;
Rimann and Hajnal, 2007; Sherwood et al., 2005; Verghese et al.,
2011). After examining expression patterns, we then synchronized TF
GFP-tagged strains and collected a developmental time course
quantifying the expression of GFP-tagged protein in the AC during
uterine/vulval development (Fig. 2B). All GFP-tagged TFs were
imaged using uniform acquisition settings on a spinning disk
confocal microscope using an EM-CCD camera, allowing us to
compare relative expression levels of endogenous GFP-tagged
proteins in the AC. Interestingly, all four TFs followed the same
general trend in expression levels, with a gradual increase in levels
prior to invasion, peaking at or just before the P6.p four-cell stage
when 100% of wild-type animals have generated a gap in the BM.
Expression levels then decreased following invasion, during vulval
morphogenesis, in the early L4 stage (P6.p eight-cell stage) (Fig. 2C,D).
To visualize invasion, we first crossed a BM reporter (laminin::
GFP) and an AC marker (cdh-3>mCherry:moeABD; see Materials
and Methods for transgene nomenclature) into each endogenously
GFP-tagged TF strain. We then performed a series of RNAi depletion
experiments targeting each TF and examining synchronized animals
at the P6.p 4-cell stage. As the insertion of GFP tags into native
genomic loci can potentially interfere with gene function, we
examined a minimum of 50 animals treated with control (T444T)
empty vector RNAi (Fig. S3). All four GFP-tagged strains showed
100% BM breach at the normal time of invasion in control animals
(empty vector) (Fig. S3) and the animals all appeared superficially
wild type. Additionally, all alleles generated did not appear to affect
viability or fertility. Although this does not rule out other more subtle
hypomorphic phenotypes, for the purpose of determining TF
regulatory relationships during AC invasion, we consider all alleles
generated to be equivalent to wild-type AC invasion.
As RNAi penetrance can be difficult to measure at the single cell
level across a population, we next assessed the efficacy of our newly
3
RESEARCH ARTICLE Development (2020) 147, dev185850. doi:10.1242/dev.185850
DEVELOPMENT
generated TF-RNAi vectors quantitatively. We examined a
minimum of 50 animals following targeted TF-RNAi depletions,
collecting spinning disk confocal z-stacks using the same
acquisition settings as used for the developmental series (Fig. 2B)
in all experiments. Following image quantification, our results were
consistent with the phenotypes identified in our original uterine-
specific RNAi screen (Fig. 1, Fig. S3). However, by being able to
quantify depletion of endogenous GFP-tagged TF targeted by
RNAi, we were able to correlate phenotype with mRNA depletion
more accurately. The nhr-67 and egl-43 improved RNAi constructs
strongly knocked down their GFP-tagged endogenous targets,
with depletions averaging 88% and 75%, respectively (Fig. S3E).
This strong knockdown of nhr-67 and egl-43 was also correlated
with highly penetrant AC invasion defects (61% and 100%,
respectively). The completely penetrant defect from egl-43(RNAi)
segregated into 26% of animals exhibiting a single AC that failed to
invade and all other cases phenocopying loss of nhr-67, with
multiple cdh-3-expressing non-invasive ACs (Fig. S3E). Depletion
of GFP::fos-1 by RNAi was also highly penetrant, with a mean
depletion of 92% of GFP-tagged endogenous protein and 76% of
treated animals with a defect in AC invasion (Fig. S3E). Although
only loss of hlh-2 prior to AC specification can affect AC fate
directly resulting in zero ACs (Karp and Greenwald, 2004), nhr-67
and egl-43 also regulate the AC/VU decision (Hwang et al., 2007;
Verghese et al., 2011). Thus, to ensure that the invasion defects we
were observing were the result of post-AC specification defects
following TF knockdown, we performed L2 platings and scored for
invasion. Because L2 platings result in less time (12 h at 25°C)
exposed to TF-RNAi, we were not surprised to observe a lower
penetrance of AC invasion defects compared with L1 platings
targeting the other TFs (Fig. S3F). As all L2 platings phenocopied
L1 platings, although at lower penetrance and with only depletion of
hlh-2 at the L1 stage directly affecting AC fate, we utilized L1
platings for their increased penetrance for nhr-67,egl-43 and fos-1
for the remainder of our experiments. Together, these results suggest
that following strong TF depletion, loss of either egl-43 or hlh-2
Fig. 2. Onset and expression of GFP-tagged alleles in the AC over developmental time. (A) Maximum intensity projections of GFP-tagged TFs (bottom) in
the context of the BM-lined uterus (laminin::GFP), in relation to the number of somatic cells in the gonad (top; ckb-3>H2B::mCherry, inverted). Dashed green
boxes indicate onset of TF expression. Dashed magenta line encircles the AC. Asterisk highlights weak TF expression. (B) Schematic illustrating the order
of onset of pro-invasive TFs in Z1/Z4 descendants over time (n25 for each stage). (C,D) Visualization (C) and quantification (D) of GFP-tagged TF levels in the
AC in relation to the division of P6.p. Brackets indicate 1° VPCs. Yellow and white arrowheads indicate the breach in the BM and ACs, respectively. In this
and all other figures, circles and error bars denote mean±s.d. (n25 for each stage). A.U., arbitrary units; M.G.V., mean gray value. Scale bars: 5 μm.
4
RESEARCH ARTICLE Development (2020) 147, dev185850. doi:10.1242/dev.185850
DEVELOPMENT
phenocopies nhr-67 depletion, generating multiple, non-invasive
ACs.
Identification of a feed-forward loop controlling NHR-67
activity
Next, we examined the relationship between the four pro-invasive
TFs during invasion. To accomplish this, we performed a series of
RNAi depletion experiments using synchronized L1 or L2 (for hlh-
2) stage animals and quantified the amount of GFP-tagged
endogenous TF in the AC in animals with defects in invasion at
the P6.p four-cell stage. Whereas depletion of fos-1 failed to
significantly downregulate levels of nhr-67::GFP, depletion of
egl-43 resulted in a strong reduction (65%) of nhr-67::GFP in the
AC (Fig. 3A,B). Intriguingly, in animals with a single non-invasive
AC following hlh-2(RNAi) we saw only a partial reduction (19%) of
nhr-67::GFP levels, whereas in animals with multiple cdh-3-
expressing cells we saw a significant reduction (49%) in nhr-67::
GFP (Fig. 5A,B). These results suggest that both egl-43 and hlh-2
co-regulate nhr-67 during invasion.
We repeated these TF-RNAi molecular epistasis experiments with
the remaining three TF-GFP-tagged strains and quantified depletion
of GFP in animals with invasion defects. Consistent with previous
studies using transcriptional and translational reporters (Hwang et al.,
2007; Rimann and Hajnal, 2007), we found that depletion of fos-1
regulated levels of egl-43::GFP::egl-43 (44% depletion) (Fig. 3C,D).
No other TF depletion significantly regulated the levels of egl-43 in
our experiments (Fig. 3C,D). hlh-2 is predicted to regulate egl-43
activity based on the presence of two conserved E-box binding motifs
Fig. 3. Regulatory interactions
among pro-invasive TFs at
endogenous loci.
(A,C,E,G) Sinaplots of GFP-tagged TF
levels, defined as the mean gray value
(M.G.V.) of individual AC nuclei,
following RNAi perturbation (n25
animals per treatment, P<1×10
6
,
Studentst-test of TF-RNAi treated
animals compared with empty vector).
(B,D,F,H) Representative phenotypes
(single versus multi AC, as noted
bottom left of each image) following
TF-RNAi depletion. Yellow and white
arrowheads indicate the breach in the
BM and ACs, respectively. Insets show
additional ACs in other focal planes.
Scale bars: 5 μm.
5
RESEARCH ARTICLE Development (2020) 147, dev185850. doi:10.1242/dev.185850
DEVELOPMENT
in the egl-43 promoter (Hwang et al., 2007). However, we did not
detect regulation of egl-43::GFP::egl-43 in either the AC (Fig. 3C,D)
or neighboring VU cells (Fig. S4). Next, we examined GFP::hlh-2
levels following TF-RNAi depletions. Intriguingly, we found a
similar pattern of regulation based on AC phenotype. Animals with
single non-invasive ACs showed partial depletion of GFP::hlh-2
following fos-1,egl-43 or nhr-67 depletion (35%, 20% and 14%,
respectively). However, animals with multiple cdh-3-expressing ACs
following egl-43(RNAi) or nhr-67(RNAi) showed stronger depletion
of GFP::hlh-2 (66% and 73%, respectively) (Fig. 3E,F). Finally, we
examined levels of GFP::fos-1 following TF depletions. Depletion
of either nhr-67 or egl-43 resulted in partial reduction of GFP::fos-1
levels in the AC (52% and 49%, respectively) (Fig. 3G,H). Together,
these results suggest that egl-43,hlh-2 and nhr-67 may function
together in a feed-forward regulatory loop with positive feedback to
maintain the AC in a post-mitotic, pro-invasive state.
Multiple ACs derive from proliferation
As depletion of egl-43 and hlh-2 both regulate levels of nhr-67 and
their depletion results in multiple ACs that fail to invade, we next
wanted to assess whether their depletion was functionally
phenocopying depletion of nhr-67. To confirm that the presence
of multiple cells expressing the cdh-3-driven AC reporter were due
to defects in proliferation, we performed static imaging using a full-
length translational cdt-1>CDT-1::GFP reporter that indicates cell
cycle progression (Fig. 4A). As CDT-1 must be removed from
origins of replication during DNA licensing, the transgene localizes
to the nucleus during G1/G0 and is largely cytosolic at the onset of S
phase (Matus et al., 2014, 2015). As expected, T444T (empty
vector)- and fos-1(RNAi)-treated animals consistently exhibited
nuclear localization of CDT-1::GFP [Fig. 4A; 100%, T444T, n=15;
100%, fos-1(RNAi),n=19 animals examined], indicating cell cycle
arrest. However, following depletion of egl-43,hlh-2 and nhr-67 we
identified multiple animals (Fig. 4A) possessing non-invasive ACs
lacking nuclear CDT-1::GFP, indicative of cycling ACs [67%,
egl-43(RNAi),n=12/18; 58%, hlh-2(RNAi),n=11/19; 71%, nhr-
67(RNAi),n=10/14 animals examined]. Together, these results
strongly support our molecular epistasis data suggesting that egl-43
and hlh-2 function upstream of nhr-67 to maintain the post-mitotic
state of the AC during invasion.
After examining the role of pro-invasive TFs in regulating cell
cycle progression, we then sought to examine other downstream
targets involved in AC invasion. Using an endogenously tagged GFP
reporter for the MMP zmp-1 (Kelley et al., 2019), we assessed the
ability of ACs to produce MMPs following TF(RNAi) depletion. We
confirmed previously published findings (Kelley et al., 2019) that
fos-1 depletion results in reduction of zmp-1 expression (Fig. 4B).
Interestingly, in the cases of the other TF(RNAi) treatments, a
significant reduction of zmp-1 expression was only observed when
multiple cdh-3-expressing ACs were present (Fig. 4B), suggesting
that cell cycle progression may downregulate production of MMPs.
EGL-43 and HLH-2 play roles in cell cycle-dependent
and independent pathways
We have shown previously that the invasive activity of the AC
following depletion of NHR-67 can be completely rescued by
restoring the AC to a post-mitotic G1/G0 state through upregulation
of the cyclin-dependent kinase inhibitor CKI-1 ( p21/p27) (Matus
et al., 2015). As our epistasis experiments revealed that egl-43
and hlh-2 positively co-regulate NHR-67 activity, we induced
AC-specific expression of CKI-1 using a cdh-3>CKI-1::GFP
integrated array and assessed invasion following RNAi depletions
Fig. 4. Pro-invasive TFs regulate the cell cycle and MMPs. (A) Localization of the cell cycle state reportercdt-1>CDT-1::GFP (green) and BM (lam-1>LAM-1::
mCherry, magenta) in empty vector control (left) compared with TF-RNAi. White and yellow fractions (bottom right of each micrograph) indicate the
number of animals exhibiting the phenotype shown. Brackets indicate 1° VPCs and arrowheads indicate ACs. Dashed box indicates single AC and multi-AC
phenotypes resulting from hlh-2 depletion. (B,C) Visualization (B) and quantification (C) of endogenous MMP (zmp-1::GFP) expression in individual ACs
following RNAi perturbations (n20 animals per treatment, P<0.01, Studentst-test TF-RNAi treated animals compared with empty vector). Insets show additional
ACs in different focal planes. M.G.V., mean gray value. Scale bars: 5 μm.
6
RESEARCH ARTICLE Development (2020) 147, dev185850. doi:10.1242/dev.185850
DEVELOPMENT
(Fig. 5A,B). As expected, induced expression of CKI-1::GFP in the
AC completely rescued nhr-67(RNAi)-treated animals [Fig. 5A,B;
100% invaded (n=42) compared with 45% of control animals
lacking induced CKI-1::GFP (n=62)]. Additionally, cdh-3>CKI-1::
GFP partially rescued hlh-2 depletion in L2 RNAi-feeding
experiments by restoring AC invasion in most animals examined
[Fig. 5A,B; 87% invaded (n=77), compared with 56% of control
animals (n=59)]. Inducing G1/G0 arrest in the AC failed to rescue
fos-1(RNAi)-treated animals [Fig. 5A,B; 20% invaded (n=35),
compared with 51% of control animals (n=53)]. AC-specific
expression of CKI-1::GFP blocked AC proliferation following
egl-43(RNAi); however, this induced arrest failed to rescue invasion
[Fig. 5A,B; 10% invaded (n=31), compared with 19% of control
animals (n=37)]. This result suggests that egl-43 has pro-invasive
functions outside of the cell cycle-dependent pathway. In support of
a role for egl-43 outside of cell cycle control, induced CKI-1::GFP
restored the expression of a reporter of MMP activity (zmp-
1>mCherry) in nhr-67-depleted animals, but neither fos-1 nor
egl-43-depleted animals showed restoration of zmp-1 transcription
(Fig. 5C). Together, these results reveal the presence of two
integrated sub-circuits necessary for invasive activity, with egl-43
and hlh-2 likely having a crucial role in both circuits.
We next set out to test our putative network topology. To do this, we
generated two new single copy transgenic lines expressing either BFP-
tagged codon-optimized HLH-2 or the genomic region of NHR-67
under the transcriptional control of a heat shock promoter. Induced
expressionof HLH-2 failed to rescue AC invasion following depletion
of egl-43,fos-1 or nhr-67, consistent with hlh-2 functioning upstream
of nhr-67 and the cell cycle-independent roles of egl-43 and fos-1 in
regulating AC invasion (Fig. S5A). Similarly, induced expression of
NHR-67 did not rescue invasion in egl-43(RNAi)-andfos-1(RNAi)-
treated animals (Fig. S5B). Furthermore, induced expression of
NHR-67 failed to rescue AC invasion in hlh-2(RNAi)-treated animals
(Fig. S5B). This result, paired with the prevalence of single non-
invasive ACs (Fig. 1C) and only partial rescue from induced CKI-1
(Fig. 5A,B) suggests that hlh-2 regulates invasive targets
independently of its role in regulating nhr-67 and cell cycle arrest.
As our induced TF experiments failed to rescue invasion defects,
we examined an nhr-67 hypomorphic allele [nhr-67(pf88)], which
we have previously shown results in an incomplete penetrance of
mitotic, non-invasive ACs (Matus et al., 2015). Notably, nhr-
67(RNAi) using the enhanced T444T vector did not significantly
increase the invasion defect of the hypomorphic allele (Fig. S6),
supporting previous results using the L4440-based nhr-67 RNAi
vector (Matus et al., 2015). Thus, the nhr-67( pf88) allele may
represent a putative null phenotype for AC invasion. Similar to our
results with nhr-67(RNAi), depletion of hlh-2 also failed to enhance
the nhr-67(pf88) AC invasion defect significantly (Fig. S6).
Together, these results suggest that hlh-2 functions upstream of
nhr-67 in maintaining the AC in a post-mitotic, pro-invasive state.
EGL-43 isoforms function redundantly and in an
autoregulative manner
Two functional isoforms of egl-43 are encoded in the C. elegans
genome (Fig. S7A). Previous research has suggested that the
longer isoform functions downstream of fos-1 to modulate MMP
expression and other fos-1 transcriptional targets, including cdh-3
(Hwang et al., 2007; Rimann and Hajnal, 2007). The shorter
isoform has been predicted to function in Notch/Delta-mediated AC
specification and later Notch/Delta-mediated patterning of the ventral
uterus (Hwang et al., 2007) or as a potential competitive inhibitor for
long isoform binding of downstream targets (Rimann and Hajnal,
2007). Additionally, fos-1 and egl-43 have been shown to function
in an incoherent feed-forward loop with negative feedback, with fos-1
Fig. 5. HLH-2 depletion is partially rescued by induced G1/G0 arrest andEGL-43 has acell cycle-independent pro-invasive role. (A) Micrographs depicting
cdh-3>CKI-1::GFP (left), DIC overlay (middle), and zmp-1>mCherry expression (right) in empty vector control (top) and TF-RNAi depletions. Brackets indicate
1° VPCs. Yellow and white arrowheads indicate the breach in the BM and ACs, respectively. Inset shows the same image with enhanced exposure to show dim
expression. Scale bars: 5 μm. (B) Quantification of AC invasion defectsin control and CKI-1(+) animals(n27 animals per treatment,*P<0.00001, Fishers exact test
CKI-1(+) animals compared with those without the rescuing transgene). (C) Quantification of zmp-1>mCherry reporter levels in control and CKI-1(+) animals
(n27 animals per treatment, P<0.01, Studentst-test CKI-1(+) animals compared with those without the rescuing transgene). M.G.V., mean gray value.
7
RESEARCH ARTICLE Development (2020) 147, dev185850. doi:10.1242/dev.185850
DEVELOPMENT
positively regulating and egl-43 negatively regulating the levels of
mig-10/lamellipodin, a key adhesion protein important for stabilizing
the attachment of the AC to the BM. Titrating levels of MIG-10 is
crucial, as overexpression of MIG-10 results in AC invasion defects
(Wang et al., 2014). Although it is readily apparent that levels of
EGL-43 are also important, it is unclear from previous studies
whether the two isoforms have divergent functions during invasion.
Thus, we next decided to explore egl-43 isoform function.
First, we generated a knock-in allele of GFP at the egl-43
N-terminus to tag the long isoform (GFP::egl-43L) specifically.
This allowed us to compare expression patterns of the long isoform
to the internally GFP-tagged allele that should dually label both
isoforms (Fig. S7A,B). We examined animals during uterine-vulval
development and found overlapping expression patterns between
both isoforms with strong expression throughout the somatic gonad
(Fig. S7B). Next, we generated an egl-43L-specific improved RNAi
targeting vector in T444T and examined egl-43L depletion in the
uterine-specific RNAi-sensitive genetic background (Matus et al.,
2010). Depletion of egl-43L resulted in a penetrant AC invasion
defect, as expected, but, notably, we detected the presence of
animals with both a single non-invasive AC (19%) and animals with
multiple ACs (39%) (Fig. 6A,B), similar to depletion of both
isoforms (Fig. 1C,E). We then assessed whether the long isoform
has the same quantitative relationship to the other TFs as
observed when targeting both isoforms. Indeed, egl-43(RNAi)
and egl-43L(RNAi) exhibited comparable levels of depletion of
egl-43::GFP::egl-43 (75%, 78%), GFP::hlh-2 (66%, 60%) and
nhr-67::GFP (66%, 64%, respectively) (n25 animals examined per
treatment) (Fig. S7C).
To characterize loss-of-function phenotypes further and examine
transcriptional output of the two egl-43 isoforms, we next examined
endogenous transcriptional reporters that were created as an
intermediate step in generating GFP knock-in alleles using the
self-excising cassette (SEC) method (Dickinson and Goldstein
2016). Specifically, prior to Cre-Lox recombination, the insertion of
GFP at the N-terminus of a locus with a 6 kb SEC generates a
transcriptional reporter at the endogenous locus and often interferes
with the native transcriptional machinery at the locus where it is
inserted, generating a strong loss-of-function or putative null allele
(Dickinson and Goldstein, 2016). Examination of pre-floxed
versions of both edited knock-in alleles of egl-43 was performed in
the presence of an mT1 balancer, as homozygous animals are non-
fertile. Similar to the putative null allele, egl-43(tm1802) (Rimann
and Hajnal, 2007), the allele resulting from the SEC insertion of the
internal GFP tag had an extremely low frequency of escapers, and we
were unable to obtain L3-stage animals to examine for AC invasion
defects. However, similar to our results with RNAi (Fig. 6A,B), an
allele resulting from SEC insertion with the long isoform [egl-
43(bmd135)] displayed a 31% AC invasion defect (n=11/36), with
animals containing either single or multiple non-invading ACs (Fig.
S7D,E). Finally, as previous reports based on transgenic
transcriptional reporters have suggested that egl-43 may be
autoregulatory (Matus et al., 2010; Rimann and Hajnal, 2007;
Wang et al., 2014), we examined the mT1-balanced pre-floxed allele
of the internally tagged egl-43 targeting both isoforms (egl-43::
GFP^SEC^::egl-43). In support of these previous studies, we found
strong evidence of autoregulation at the endogenous locus, as egl-
43(RNAi) reduced the expression of GFP by 65% (n25 animals)
(Fig. 6C,D). Taken together, these results suggest that both isoforms
of egl-43 function redundantly to regulate multiple transcriptional
sub-circuits that are crucial for invasion.
Activation of the pro-invasive gene regulatory network
occurs post-specification
Our results here identify a putative feed-forward loop between
egl-43,hlh-2 and nhr-67 to maintain levels of NHR-67, which we
have previously shown to be crucial for keeping the AC in a post-
mitotic, pro-invasive state (Matus et al., 2015). As these same TFs
also have been shown to function during the AC/VU decision
(Hwang et al., 2007; Karp and Greenwald, 2004; Verghese et al.,
2011), we investigated whether the same regulatory relationships
were utilized during cell fate specification and invasion.
Fig. 6. Both isoforms of egl-43 function redundantly to regulate AC invasion. (A,B) Visualization (A) and quantification (B) of AC invasion and proliferation
defects resulting from RNAi depletion of the long isoform of egl-43 (egl-43L). (C,D) Visualization (C) and quantification (D) of egl-43>GFP expression in individual AC
nuclei, assessed using a balanced endogenous egl-43 transcriptional reporter [egl-43(bmd87)/mT1], following egl-43(RNAi) treatment compared with control (n25
animals per treatment, P<1×10
15
, Studentst-test). Arrowheads and brackets indicate ACs and 1° VPCs, respectively. M.G.V., mean gray value. Scale bars: 5 μm.
8
RESEARCH ARTICLE Development (2020) 147, dev185850. doi:10.1242/dev.185850
DEVELOPMENT
Intriguingly, our examination of TF onset establishes a hierarchy of
egl-43,hlh-2 and nhr-67 reaching steady-state levels in the
descendants of Z1 and Z4, although all three TFs are robustly
expressed in the AC/VU cells (Fig. 2A). To examine TF regulatory
relationships prior to specification, we performed L1 TF RNAi
platings and determined expression levels of TF-GFP in the AC/VU
cells in L2-stage animals. Although each TF RNAi robustly
depleted its target GFP allele, we failed to detect any molecular
epistatic interactions between TFs in the AC/VU cells (Fig. 7A,B).
In support of previous research demonstrating a role for these TFs in
regulating the AC/VU decision (Hwang et al., 2007; Karp and
Greenwald, 2004; Verghese et al., 2011), we detected significant
depletion of an endogenously GFP-tagged allele of lin-12 (Notch)
(Fig. 7C,D). Together, these results suggest that, despite the re-
iterative use of the same set of TFs in specification and invasion, the
gene regulatory network (GRN) we characterized here is specific for
the post-specification pro-invasive behavior of the AC.
DISCUSSION
Cellular invasion requires coordination of extrinsic cues from the
surrounding microenvironment and orchestration of intrinsic
regulatory circuits (Sherwood and Plastino, 2018). We focus here
on exploring the intrinsic pathways that autonomously regulate
C. elegans AC invasion. Combining improved RNAi targeting
vectors with GFP-tagged alleles enabled us to correlate phenotype
with depletion of the endogenous mRNAs and compare regulatory
relationships in a quantitative framework. Our experiments reveal
new roles for egl-43 and hlh-2 in regulating nhr-67-mediated cell
cycle arrest and suggest that these TFs function in a coherent feed-
forward loop with positive feedback. Additionally, we confirm
previous work demonstrating that egl-43 is autoregulatory and
demonstrate that it functions in both the cell cycle-dependent sub-
circuit as well as in a cell cycle-independent manner, through fos-1,
to orchestrate invasion (Fig. 8). Together, building upon previous
work using transgenic approaches (Bodofsky et al., 2018; Hwang
et al., 2007; Rimann and Hajnal, 2007; Schindler and Sherwood,
2011; Verghese et al., 2011), our results using endogenously tagged
alleles provide the first description of the native regulatory
relationships between the four TFs that promote invasion during
C. elegans uterine-vulval attachment.
A key function of nhr-67 is to maintain the post-mitotic state of
the AC (Matus et al., 2015). Here, we demonstrate that egl-43 and
Fig. 7. Regulatory relationships do not exist prior to AC specification. (A,B) Visualization (A) and quantification (B) of GFP-tagged TF levels in the AC/VU
precursors (Z1.ppp and Z4.aaa) following RNAi perturbations (n20 animals per treatment, P<1×10
6
, Studentst-test TF-RNAi treated animals compared
with empty vector). Dashed circles indicated location of AC/VU precursor cells in experiments where TF::GFP is strongly depleted by the targeting RNAi. (C,D)
Visualization (C) and quantification (D) of GFP-tagged lin-12 (asterisks) in the AC/VU precursors following RNAi perturbations (n20 animals per treatment,
P<0.01, Studentst-test TF-RNAi treated animals compared with empty vector). M.G.V., mean gray value. Scale bars: 5 μm.
9
RESEARCH ARTICLE Development (2020) 147, dev185850. doi:10.1242/dev.185850
DEVELOPMENT
hlh-2 both regulate nhr-67, and that nhr-67-depletion also reduces
HLH-2 levels. Network inference suggests these three TFs function
in a classic type I coherent feed-forward loop regulatory circuit with
positive feedback (Mangan and Alon, 2003) (Fig. 8). Our model fits
with data by others showing that hlh-2 may directly bind to the nhr-
67 promoter through two canonical E-box motifs found in a 164 bp
window in a functional promoter element deleted in several
hypomorphic alleles of nhr-67 (Bodofsky et al., 2018; Matus
et al., 2015; Verghese et al., 2011). Further, deletion of these hlh-2
conserved binding sites results in loss of AC-specific GFP
expression in transgenic lines (Bodofsky et al., 2018). Although
others have also reported positive regulation by EGL-43 on nhr-67
activity via transgenic reporters (Bodofsky et al., 2018), this
interaction, and many of the other TF interactions, may be indirect,
as no known binding sites exist for EGL-43 in the nhr-67 promoter.
Alternatively, EGL-43 and NHR-67 have been predicted, based on
yeast two-hybrid experiments, to interact at the protein-protein level
(Reece-Hoyes et al., 2013).
We have also examined the activityof the two previously identified
isoforms of egl-43 (Hwang et al., 2007; Rimann and Hajnal, 2007).
As both mutant analyses and RNAi depletion of the long isoform
phenocopy depletion of both isoforms, our data suggest that either the
long isoform of egl-43 functions redundantly with the short isoform
or it functions as the dominant isoform at the intersection of multiple
regulatory circuits. Of these two possible hypotheses, the latter is
supported by a recent report demonstrating that the long isoform of
egl-43 is a key upstream regulator of the post-mitotic state of the AC
as mutation of the transcriptional start methionine of the short isoform
of egl-43 has no observable AC invasion defect and an AC-specific
genetic knockout of the long isoform results in mitotic, non-invasive
ACs (Deng et al., 2019 preprint).
Together, our data and corroborating evidence from the literature
support the existence of a coherent feed-forward circuit with positive
feedback among egl-43,hlh-2 and nhr-67 in maintaining the post-
mitotic state of the AC (Fig. 8). We have previously shown that nhr-
67 positively regulates transcripts of cki-1 specifically in the AC
(Matus et al., 2015). Thus, to test our putative regulatory circuit, we
induced expression of cki-1 in the AC, which prevents the AC from
inappropriately entering the cell cycle. This forced G1/G0 arrest fully
rescues nhr-67 depletion (Matus et al., 2015). Induced cki-1 partially
rescued hlh-2 depletion but failed to rescue egl-43 depletion. Heat
shock-induced expression of nhr-67 failed to rescue either hlh-2 or
egl-43,suggestingthathlh-2,likeegl-43, has a c ell cycle-independent
function. These results suggest that egl-43 and hlh-2 function in
multiple regulatory circuits. For egl-43,thisissupportedbyrecent
work demonstrating a type I incoherent feed-forward loop between
fos-1,egl-43L and the BM adhesion protein MIG-10/lamellipodin
(Wang et al., 2014). Finally, our observation that egl-43 is
autoregulatory, also shown previously by promoter>GFP fusion
experiments (Matus et al., 2015; Rimann and Hajnal, 2007; Wang
et al., 2014), supports a model in which egl-43 occupies a key node at
the intersection of multiple pro-invasive circuits (Fig. 8).
Feed-forward regulatory loops are likely the most well-described
network motif occurring in all GRNs (Cordero and Hogeweg, 2006;
Davidson, 2010; Mangan and Alon, 2003). From E. coli (Shen-Orr
et al., 2002) and yeast (Lee et al., 2002; Milo et al., 2002) to a myriad
of examples across the Metazoa (reviewed by Davidson, 2010), feed-
forward loops are thought to function as filters for transient inputs
(reviewed by Alon, 2007; Hinman, 2016). The addition of positive
feedback, generating a coherent feed-forward loop, provides stability
to a sub-circuit and is often found in differentiation gene batteries
coincident with autoregulation (Davidson, 2010). These network
motifs have been described in many developmental contexts,
including MyoD-driven vertebrate skeletal muscle differentiation
(Penn et al., 2004), patterning of the Drosophila egg shell via the TF
Broad and interactions with EGFR and Dpp signaling (Yakoby et al.,
2007), Pax6-dependent regulation of Maf and crystallin expression in
the mouse embryonic lens (Xie and Cvekl, 2011) andterminal selector
neuronal differentiation in C. elegans and mammals (reviewed by
Hobert, 2008). During C. elegans embryonic development, a series of
coherent feed-forward loops utilizing SKN-1/MED-1,2 and then a
suite of reiteratively used GATA factors (END-1,-3) are required to
pattern endomesoderm development successfully (reviewed by
Maduro, 2009). Thus, the coherent feed-forward loop identified
here maintaining the post-mitotic state in the C. elegans AC likely
evolved to coordinate AC cell cycle exit, both as a by-product of
terminal differentiation and a morphogenetic requirement for invasive
behavior (Matus et al., 2015). AC invasion is necessary for egg laying,
Fig. 8. Summary model of the GRNs coordinating AC
specification and invasion. Left: the Notch-mediated AC/VU cell
fate decision appears to be independently regulated by egl-43,
hlh-2 and nhr-67. Right: network inference predicts that the GRN
regulating AC invasion contains cell cycle-independent and
-dependent sub-circuits, the latter entailing a type I coherent feed-
forward loop with positive feedback to maintain the AC in a post-
mitotic, invasive state. Dotted lines represent predicted feedback.
10
RESEARCH ARTICLE Development (2020) 147, dev185850. doi:10.1242/dev.185850
DEVELOPMENT
and defects in the process results in penetrant Protruding vulva/Egg-
laying defective (Pvl/Egl) phenotypes, reducing fecundity in
individual animals by nearly ten-fold. Thus, redundant control of
the sub-circuit regulating invasion and cell cycle arrest may have been
under strong selection, providing an explanation for the regulatory
relationships between egl-43,hlh-2 and nhr-67 we characterize here.
As egl-43,hlh-2 and nhr-67 reach steady-state levels prior to the
time of AC/VU specification and individually have been previously
shown to regulate different aspects of the stochastic cell fate decision
between the AC and VU (Hwang et al., 2007; Karp and Greenwald,
2004; Verghese et al., 2011), we performed RNAi depletion
experiments and examined TF::GFP levels in the AC/VU cells
during the L2 stage. Surprisingly, we found noevidence of regulation
between the TFs at this stage, though, as expected, all three TFs
regulated levels of endogenously tagged lin-12::GFP (Notch). This
suggests that the feed-forward loop functions specifically in the post-
specification AC to maintain the post-mitotic, invasive state (Fig. 8).
Transient TF-DNA binding and GRN rewiring can result from
responses to diverse stimuli (Luscombe et al., 2004; Swift and
Coruzzi, 2017). One plausible explanation for this temporal change in
network topology is that cell cycle-dependent changes in chromatin
remodeling and TF activity likely changes the genomic landscape
between the bipotential AC/VU cells and the post-mitotic,
differentiated AC (Ma et al., 2015). Additionally, there may be
differentially expressed co-factor(s) functioning during specification
or invasion that facilitate interactions between TFs. Reiterative use of
the same TFs to program different cell behaviors has been examined
during neuronal differentiation (Guillemot, 2007) and neural crest
specification and migration, as the winged-helix TF FoxD3 was
recently shown to regulate multiple independent chromatin-
organizing circuits, functioning as a pioneer factor during zebrafish
neural crest specification and a transcriptional repressor during later
events, including epithelial-mesenchymal transition and migration
(Lukoseviciute et al., 2018). Whether or not egl-43,nhr-67 or hlh-2
function as pioneer factors during AC/VU specification is an open
question, and one that could soon be tackled through isolation by
fluorescence-activated cell sorting and the application of newer
next-generation sequencing tools such as ATAC-seq (assay for
transposase-accessible chromatin using sequencing) or CUT&RUN
(cleavage under targets and release using nuclease) (Skene and
Henikoff, 2017).
We are just beginning to understand the connections between
regulatory circuits and morphogenetic behaviors (Christiaen et al.,
2008; Martik and McClay, 2015; Saunders and McClay, 2014).
Hopefully, the ease of genome editing and protein perturbation
strategies will facilitate the kind of analyses we have performed here
in other metazoan systems. In summary, in this study we characterize
the complex relationships between the four pro-invasive TFs that
function in the AC to program invasive behavior. We show that
although the TFs are reiteratively used in both cell fate specification
and in a differentiated cell behavior, invasion, they only interact with
each other following specification. We identify a classic type I feed-
forward loop regulating mitotic exit and controlling a switch between
proliferative and invasive behavior. Whether or not similar circuit
architecture is utilized to regulate invasive and proliferative cell
biology in other developmental invasive contexts, including
mammalian trophoblast implantation and placentation (Carter et al.,
2015; Red-Horse et al., 2004) and epithelial to mesenchymal
transition events during gastrulation (Vega et al., 2004), is still poorly
understood. Finally, as there appear to be many cancer subtypes that
may switch between proliferative and invasive fates (reviewed by
Kohrman and Matus, 2017), improving our understanding of the
GRN architecture of invasive cells may provide new therapeutic
nodes to target in reducing the lethality associated with cancer
metastasis.
MATERIALS AND METHODS
C. elegans strains and culture conditions
Animals were reared under standard conditions and cultured at 25°C, with the
exception of temperature-sensitive strains including those containing heat
shock-inducible transgenes or the rrf-3(pk1426) allele (conferring RNAi
hypersensitivity), which were maintained at 15°C and 20°C (Brenner, 1974;
Simmer et al., 2002). Heat shock-inducible transgenes were activated by
incubating animals on plates at 32°C for 1 h following AC specification and
again just prior to AC invasion. Animals were synchronized for experiments
through alkaline hypochlorite treatment of gravid adults to isolate eggs (Porta-
de-la-Riva et al., 2012). In the text and figures, we designate linkage to a
promoter through the use of the>symbol and fusion of a proteins by a ::
annotation. The following transgenes and alleles were used in this study:
qyIs102[fos-1>RDE-1;myo-2>YFP] LG I hlh-2(bmd90[hlh-2>LoxP::GFP::
HLH-2]),qyIs227[cdh-3>mCherry::moeABD],bmd121[LoxP::hsp>NHR-
67::2xBFP],bmd142[hsp>HLH-2::2xBFP];LG II egl-43(bmd87[egl-
43>SEC::GFP::EGL-43]),egl-43(bmd88[egl-43>LoxP::GFP::EGL-43]),
egl-43(bmd136[egl-43L>LoxP::GFP::EGL-43]) rrf-3( pk1426);qyIs17
[zmp-1>mCherry]LG III unc-119(ed4),LG IV nhr-67(syb509[nhr-
67>NHR-67::GFP]),qyIs10[lam-1>LAM-1::GFP] LG V fos-1(bmd138
[fos-1>LoxP::GFP::FOS-1]),qyIs225[cdh-3>mCherry::moeABD],rde-
1(ne219),qyIs24[cdh-3
1.5
>mCherry::PLCδPH],qyIs266[cdh-3>CKI-1::
GFP] LG X qyIs7[lam-1>LAM-1::GFP]. See Table S3 for additional
details of strains used and generated in this study.
Molecular biology and microinjection
TFs were tagged at their endogenous loci using CRISPR/Cas9 genome
editing via microinjection into the hermaphrodite gonad (Dickinson and
Goldstein, 2016; Dickinson et al., 2013). Repair templates were generated as
synthetic DNAs from either Integrated DNA Technologies (IDT) as gBlocks
or Twist Biosciences as DNA fragments and cloned into ccdB compatible
sites in pDD282 by New England Biolabs Gibson assembly (Dickinson et al.,
2015). Homology arms ranged from 690 to 1200 bp (see Tables S4-S6 for
additional details). sgRNAs were constructed by EcoRV and NheI digestion
of the plasmid pDD122. A 230 bp amplicon was generated replacing the
sgRNA targeting sequence from pDD122 with a new sgRNA, and NEB
Gibson assembly was used to generate new sgRNA plasmids (see Tables S4-
S6 for additional details). Hermaphrodite adults were co-injected with guide
plasmid (50 ng/μl), repair plasmid (50 ng/μl) and an extrachromosomal array
marker (pCFJ90, 2.5 ng/μl), and incubated at 25°C for several days before
carrying out screening and floxing protocols associated with the SEC system
(Dickinson et al., 2015).
RNA interference
An RNAi library of the pro-invasive TFs was constructed by cloning 950-
1000 bp of synthetic DNA (663 bp for the egl-43 long-specific isoform)
based on cDNA sequences available on WormBase (www.wormbase.org)
into the highly efficient T444T RNAi vector (Grove et al., 2018; Sturm
et al., 2018). Synthetic DNAs were generated by IDT as gBlocks or Twist
Biosciences as DNA fragments and cloned into restriction-digested T444T
using NEB Gibson Assembly (see Tables S4, S5 and S7 for additional
details). For most experiments, synchronized L1-stage animals were directly
exposed to RNAi through feeding with bacteria expressing dsRNA (Conte
et al., 2015). Because early hlh-2 RNAi treatment perturbs AC specification,
animals were initially placed on empty vector RNAi plates and then
transferred to hlh-2 RNAi plates following AC specification, approximately
12 h later at 25°C or 24 h later at 15°C (Schindler and Sherwood, 2011).
Live-cell imaging
All micrographs included in this manuscript were collected on a Hamamatsu
Orca EM-CCD camera mounted on an upright Zeiss AxioImager A2 with a
Borealis-modified CSU10 Yokagawa spinning disk scan head (Nobska
Imaging) using 488 nm and 561 nm Vortran lasers in a VersaLase merge
11
RESEARCH ARTICLE Development (2020) 147, dev185850. doi:10.1242/dev.185850
DEVELOPMENT
and a Plan-Apochromat 100×/1.4 (NA) Oil DIC objective. MetaMorph
software (Molecular Devices) was used for microscopy automation. Several
experiments were scored using epifluorescence visualized on a Zeiss
Axiocam MRM camera, also mounted on an upright Zeiss AxioImager A2
and a Plan-Apochromat 100×/1.4 (NA) Oil DIC objective. Animals were
mounted into a drop of M9 on a 5% Noble agar pad containing
approximately 10 mM sodium azide anesthetic and topped with a coverslip.
Scoring of AC invasion
AC invasion was scored at the P6.p four-cell stage, when 100% of wild-type
animals exhibit a breached BM (Sherwood and Sternberg, 2003). Presence
of green fluorescence under the AC, in strains with the laminin::GFP
transgene, or presence of a phase dense line, in strains without the transgene,
was used to assess invasion. Wild-type invasion is defined as a breach as
wide as the basolateral surface of the AC, whereas partial invasion indicates
the presence of a breach smaller than the footprint of the AC (Sherwood and
Sternberg, 2003). Raw scoring data is available in Tables S1 and S2.
Image quantification and statistical analyses
Images were processed using Fiji/ImageJ (v.1.52q) (Schindelin et al., 2012).
Expression levels of TFs were measured by quantifying the mean gray value
of AC nuclei, defined as somatic gonad cells strongly expressing the
cdh-3>mCherry::moeABD transgene, subtracting the mean gray value of a
background region of an equal area to account for EM-CCD camera noise,
which we use as a proxy for background fluorescence, as measurements of
the mean gray value of unlabeled ACs approximately correspond to camera
noise (means=2514.2 and 2243.5, respectively). Levels of zmp-1>mCherry
were quantified in control and cdh-3>CKI-1::GFP animals by measuring
the mean gray value of the entire AC, selected either by a hand-drawn region
of interest or using the threshold and wand tools in Fiji/ImageJ. For
characterization of downstream targets (Fig. 4) and molecular epistasis
experiments (Fig. 3, Figs S3 and S7) at the L3 stage, only TF-RNAi animals
exhibiting defects in invasion were included in the analysis. Data
was normalized to negative control (empty vector) values for the plots in
Figs 3-7 and Figs S5-S7, and to both negative control and positive control
values for determining the interaction strengths represented in Fig. 8. Images
were overlaid and figures were assembled using Adobe Photoshop
(v. 20.0.6) and Illustrator CS (v. 10.14), respectively. Statistical analyses
and plotting of data were conducted using RStudio (v. 1.1.463). Individual
data points for each experiment were collected over multiple days. Statistical
significance was determined using either a two-tailed Studentst-test or
Fishers exact probability test. Figure legends specify when each test was
used and the P-value cut-off.
Acknowledgements
We are grateful to R. Adikes, B. Kinney, A. Kohrman, M. Martinez and B. Martin for
comments on the manuscript. We also thank I. Greenwald for providing the
endogenously tagged lin-12>LIN-12::GFP strain used in Fig. 7C. Some strains were
provided by the Caenorhabditis Genetics Center, which is funded by the NIH Office
of Research Infrastructure Programs [P40 OD010440].
Competing interests
The authors declare no competing or financial interests.
Author contributions
Conceptualization: T.N.M., D.Q.M.; Methodology: T.N.M., D.Q.M.; Validation:
T.N.M., D.Q.M.; Formal analysis: T.N.M., J.J.S., N.J.P., D.Q.M.; Investigation:
T.N.M., J.J.S., N.J.P., D.Q.M.; Resources: T.N.M., J.J.S., N.J.P., S.T., W.Z., D.Q.M.;
Data curation: T.N.M., J.J.S., D.Q.M.; Writing - original draft: T.N.M., D.Q.M.; Writing
- review & editing: T.N.M., J.J.S., N.J.P., D.Q.M.; Visualization: T.N.M., J.J.S.,
N.J.P., D.Q.M.; Supervision: W.Z., D.Q.M.; Project administration: D.Q.M.; Funding
acquisition: D.Q.M.
Funding
This work was funded by the National Institutes of Health (NIH) National Cancer
Institute [5R00CA154870-05 to D.Q.M.] and National Institute of General Medical
Sciences (NIGMS) [1R01GM121597-01 to D.Q.M.]. D.Q.M. is also a Damon
Runyon-Rachleff Innovator supported (in part) by the Damon Runyon Cancer
Research Foundation [DRR-47-17]. T.N.M.-K. is supported by the NIH Eunice
Kennedy Shriver National Institute of Child Health and Human Development
[F31HD100091-01]. J.J.S. is supported by NIGMS [3R01GM121597-02S1] and
N.J.P. is supported by the American Cancer Society [132969-PF-18-226-01-CSM].
Deposited in PMC for release after 12 months.
Supplementary information
Supplementary information available online at
http://dev.biologists.org/lookup/doi/10.1242/dev.185850.supplemental
References
Alon, U. (2007). Network motifs: theory and experimental approaches. Nat. Rev.
Genet. 8, 450-461. doi:10.1038/nrg2102
Attner, M. A., Keil, W., Benavidez, J. M. and Greenwald, I. (2019). HLH-2/E2A
expression links stochastic and deterministic elements of a cell fate decision
during C. elegans gonadogenesis. Curr. Biol. 29, 3094-3100. doi:10.1016/j.cub.
2019.07.062
Bodofsky, S., Liberatore, K., Pioppo, L., Lapadula, D., Thompson, L.,
Birnbaum, S., McClung, G., Kartik, A., Clever, S. and Wightman, B. (2018).
A tissue-specific enhancer of the C. elegans nhr-67/tailless gene drives
coordinated expression in uterine stem cells and the differentiated anchor cell.
Gene Expr. Patterns 30, 71-81. doi:10.1016/j.gep.2018.10.003
Brenner, S. (1974). The genetics of Caenorabditis elegans.Genetics 77, 71-94.
Carter, A. M., Enders, A. C. and Pijnenborg, R. (2015). The role of invasive
trophoblast in implantation and placentation of primates. Philos. Trans. R. Soc. B
Biol. Sci. 370, 20140070-20140070. doi:10.1098/rstb.2014.0070
Christiaen, L., Davidson, B., Kawashima, T., Powell, W., Nolla, H., Vranizan, K.
and Levine, M. (2008). The transcription/migration interface in heart precursors of
Ciona intestinalis.Science 320, 1349-1352. doi:10.1126/science.1158170
Conte, D., Jr., MacNeil, L. T., Walhout, A. J. M. and Mello, C. C. (2015). RNA
interference in Caenorhabditis Elegans.Curr. Protoc. Mol. Biol. 109,
26.3.1-26.3.30. doi:10.1002/0471142727.mb2603s109
Cordero, O. X. and Hogeweg, P. (2006). Feed-forward loop circuits as a side effect
of genome evolution. Mol. Biol. Evol. 23, 1931-1936. doi:10.1093/molbev/msl060
Davidson, E. H. (2010). Emerging properties of animal gene regulatory networks.
Nature 468, 911-920. doi:10.1038/nature09645
Deng, T., Daube, M., Hajnal, A. and Lattmann, E. (2019). The C. elegans homolog
of the Evi1 proto-oncogene, egl-43, coordinates G1 cell cycle arrest with pro-
invasive gene expression during anchor cell invasion. bioRxiv, 1-37. doi:10.1101/
802355
Dickinson, D. J. and Goldstein, B. (2016). CRISPR-based methods for
Caenorhabditis Elegans genome engineering. Genetics 202, 885-901. doi:10.
1534/genetics.115.182162
Dickinson, D. J., Ward, J. D., Reiner, D. J. and Goldstein, B. (2013). Engineering
the Caenorhabditis elegans genome using Cas9-triggered homologous
recombination. Nat. Methods 10, 1028-1034. doi:10.1038/nmeth.2641
Dickinson, D. J., Pani, A. M., Heppert, J. K., Higgins, C. D. and Goldstein, B.
(2015). Streamlined genome engineering with a self-excising drug selection
cassette. Genetics 200, 1035-1049. doi:10.1534/genetics.115.178335
Grove, C., Cain, S., Chen, W. J., Davis, P., Harris, T., Howe, K. L., Kishore, R.,
Lee, R., Paulini, M., Raciti, D. et al. (2018). Using WormBase: a genome biology
resource for Caenorhabditis elegans and related nematodes. Methods Mol. Biol.
1757, 399-470. doi:10.1007/978-1-4939-7737-6_14
Guillemot, F. (2007). Spatial and temporal specification of neural fates
by transcription factor codes. Development 134, 3771-3780. doi:10.1242/
dev.006379
Haerty, W., Artieri, C., Khezri, N., Singh, R. S. and Gupta, B. P. (2008).
Comparative analysis of function and interaction of transcription factors in
nematodes: Extensive conservation of orthology coupled to rapid sequence
evolution. BMC Genomics 9, 1-16. doi:10.1186/1471-2164-9-399
Hagedorn, E. J., Yashiro, H., Ziel, J. W., Ihara, S., Wang, Z. and Sherwood, D. R.
(2009). Integrin acts upstream of Netrin signaling to regulate formation of the
anchor cells invasive membrane in C. elegans. Dev. Cell 17, 187-198. doi:10.
1016/j.devcel.2009.06.006
Hagedorn, E. J., Ziel, J. W., Morrissey, M. A., Linden, L. M., Wang, Z., Chi, Q.,
Johnson, S. A. and Sherwood, D. R. (2013). The netrin receptor DCC focuses
invadopodia-driven basement membrane transmigration in vivo. J. Cell Biol. 201,
903-913. doi:10.1083/jcb.201301091
Hanahan, D. and Weinberg, R. A. (2011). Hallmarks of cancer: the next generation.
Cell 144, 646-674. doi:10.1016/j.cell.2011.02.013
Hinman,V.F.(2016). Conservation and evolution of gene networks driving
development. Encycl. Evol. Biol. 110-116. doi:10.1016/B978-0-12-800049-6.
00130-X
Hobert, O. (2008). Regulatory logic of neuronal diversity: terminal selector genes
and selector motifs. Proc. Natl. Acad. Sci. USA 105, 20067-20071. doi:10.1073/
pnas.0806070105
Hwang, B. J., Meruelo, A. D. and Sternberg, P. W. (2007). C. elegans EVI1 proto-
oncogene, EGL-43, is necessary for Notch-mediated cell fate specification and
regulates cell invasion. Development 134, 669-679. doi:10.1242/dev.02769
12
RESEARCH ARTICLE Development (2020) 147, dev185850. doi:10.1242/dev.185850
DEVELOPMENT
Karp, X. and Greenwald, I. (2004). Multiple roles for the E/Daughterless ortholog
HLH-2 during C. elegans gonadogenesis. Dev. Biol. 272, 460-469. doi:10.1016/j.
ydbio.2004.05.015
Kelley, L. C., Lohmer, L. L., Hagedorn, E. J. and Sherwood, D. R. (2014).
Traversing the basement membrane in vivo: A diversity of strategies. J. Cell Biol.
204, 291-302. doi:10.1083/jcb.201311112
Kelley, L. C., Chi, Q., Cáceres, R., Hastie, E., Schindler, A. J., Jiang, Y., Matus,
D. Q., Plastino, J. and Sherwood, D. R. (2019). Adaptive F-actin polymerization
and localized ATP production drive basement membrane invasion in the absence
of MMPs. Dev. Cell 48, 313-328.e8. doi:10.1016/j.devcel.2018.12.018
Kim, H., Ishidate, T., Ghanta, K. S., Seth, M., Conte, D., Shirayama, M. and Mello,
C. C. (2014). A Co-CRISPR strategy for efficient genome editing in Caenorhabditis
elegans. Genetics 197, 1069-1080. doi:10.1534/genetics.114.166389
Kimble, J. and Hirsh, D. (1979). The postembryonic cell lineages of the
hermaphrodite and male gonads in Caenorhabditis elegans.Dev. Biol. 70,
396-417. doi:10.1016/0012-1606(79)90035-6
Kohrman, A. Q. and Matus, D. Q. (2017). Divide or conquer: cell cycle regulation of
invasive behavior. Trends Cell Biol. 27, 12-25. doi:10.1016/j.tcb.2016.08.003
Lee, T. I., Hannett, N., Harbison, C., Thompson, C., Simon, I., Zeitlinger, J.,
Jennings, E., Murray, H., Gordon, D., Ren, B. et al. (2002). Transcriptional
regulatory networks in Saccharomyces cerevisiae.Science 298, 799-804. doi:10.
1126/science.1075090
Levitan, D. and Greenwald, I. (1998). LIN-12 protein expression and localization
during vulval development in C. elegans.Development 125, 3101-3109.
Lukoseviciute, M., Gavriouchkina, D., Williams, R. M., Hochgreb-Hagele, T.,
Senanayake, U., Chong-Morrison, V.,Thongjuea, S., Repapi, E., Mead, A. and
Sauka-Spengler, T. (2018). From pioneer to repressor: bimodal foxd3 activity
dynamically remodels neural crest regulatory landscape in vivo. Dev. Cell 47,
608-628.e6. doi:10.1016/j.devcel.2018.11.009
Luscombe, N. M., Babu, M. M., Yu, H., Snyder, M., Teichmann, S. A. and
Gerstein, M. (2004). Genomic analysis of regulatory network dynamics reveals
large topological changes. Nature 431, 308-312. doi:10.1038/nature02782
Ma, Y., Kanakousaki, K. and Buttitta, L. (2015). How the cell cycle impacts
chromatin architecture and influences cell fate. Front. Genet. 5, 1-18. doi:10.3389/
fgene.2015.00019
Maduro, M. F. (2009). Structure and evolution of the C. elegans embryonic
endomesoderm network. Biochim. Biophys. Acta Gene Regul. Mech. 1789,
250-260. doi:10.1016/j.bbagrm.2008.07.013
Mangan, S. and Alon, U. (2003). Structure and function of the feed-forward loop
network motif. Proc. Natl. Acad. Sci. USA 100, 11980-11985. doi:10.1073/pnas.
2133841100
Martik, M. L. and McClay, D. R. (2015). Deployment of a retinal determination gene
network drives directed cell migration in the sea urchin embryo. eLife 4, e08827.
doi:10.7554/eLife.08827
Martinez, M. A. Q., Kinney, B. A., Medwig-Kinney, T. N., Ashley, G., Ragle, J. M.,
Johnson, L., Aguilera, J., Hammell, C. M., Ward, J. D. and Matus, D. Q. (2019).
Rapid degradation of Caenorhabditis elegans proteins at single-cell resolution
with a synthetic auxin. G3 (Bethesda) 9, g3.400781.2019. doi:10.1534/g3.119.
400781
Matus, D. Q., Li, X.-Y., Durbin, S., Agarwal, D., Chi, Q., Weiss, S. J. and
Sherwood, D. R. (2010). In vivo identification of regulators of cell invasion across
basement membranes. Sci. Signal. 3, ra35. doi:10.1126/scisignal.2000654
Matus, D. Q., Chang, E., Makohon-Moore, S. C., Hagedorn, M. A., Chi, Q. and
Sherwood, D. R. (2014). Cell division and targeted cell cycle arrest opens and
stabilizes basement membrane gaps. Nat. Commun. 5, 4184. doi:10.1038/
ncomms5184
Matus, D. Q., Lohmer, L. L., Kelley, L. C., Schindler, A. J., Kohrman, A. Q.,
Barkoulas, M., Zhang, W., Chi, Q. and Sherwood, D. R. (2015). Invasive cell
fate requires G1 cell-cycle arrest and histone deacetylase-mediated changes in
gene expression. Dev. Cell 35, 162-174. doi:10.1016/j.devcel.2015.10.002
Medwig, T. N. and Matus, D. Q. (2017). Breaking down barriers: the evolution of cell
invasion. Curr. Opin. Genet. Dev. 47, 33-40. doi:10.1016/j.gde.2017.08.003
Milo, R., Shen-Orr, S., Itzkovitz, S., Kashtan, N., Chklovskii, D. and Alon, U.
(2002). Network motifs: simple building blocks of complex networks. Science 298,
824-827. doi:10.1126/science.298.5594.824
Morrissey, M. A., Keeley, D. P., Hagedorn, E. J., McClatchey, S. T. H., Chi, Q.,
Hall, D. H. and Sherwood, D. R. (2014). B-LINK: a Hemicentin, Plakin, and
integrin-dependent adhesion system that links tissues by connecting adjacent
basement membranes. Dev. Cell 31, 319-331. doi:10.1016/j.devcel.2014.08.024
Penn, B. H., Bergstrom, D. A., Dilworth, F. J., Bengal, E. and Tapscott, S. J.
(2004). A MyoD-generated feed-forward circuit temporally patterns gene
expression during skeletal muscle differentiation. Genes Dev. 18, 2348-2353.
doi:10.1101/gad.1234304
Porta-de-la-Riva, M., Fontrodona, L., Villanueva, A. and Cerón, J. (2012). Basic
Caenorhabditis elegans methods: synchronization and observation. J. Vis. Exp.
64, e4019. doi:10.3791/4019
Red-Horse, K., Zhou, Y., Genbacev, O., Prakobphol, A., Foulk, R., McMaster, M.
and Fisher, S. J. (2004). Trophoblast differentiation during embryo implantation
and formation of the maternal-fetal interface. J. Clin. Invest. 114, 744-754. doi:10.
1172/JCI200422991
Reece-Hoyes, J. S., Pons, C., Diallo, A., Mori, A., Shrestha, S., Kadreppa, S.,
Nelson, J., DiPrima, S., Dricot, A., Lajoie, B. R. et al. (2013). Extensive rewiring
and complex evolutionary dynamics in a C. elegans multiparameter transcription
factor network. Mol. Cell 51, 116-127. doi:10.1016/j.molcel.2013.05.018
Rimann, I. and Hajnal, A. (2007). Regulation of anchor cell invasion and uterine cell
fates by the egl-43 Evi-1 proto-oncogene in Caenorhabditis elegans.Dev. Biol.
308,187-195. doi:10.1016/j.ydbio.2007.05.023
Rowe, R. G. and Weiss, S. J. (2008). Breaching the basement membrane: who,
when and how? Trends Cell Biol. 18, 560-574. doi:10.1016/j.tcb.2008.08.007
Sallee, M. D., Littleford, H. E. and Greenwald, I. (2017). A bHLH code for sexually
dimorphic form and function of the C. elegans somatic gonad. Curr. Biol. 27,
1853-1860.e5. doi:10.1016/j.cub.2017.05.059
Saunders, L. R. and McClay, D. R. (2014). Sub-circuits of a gene regulatory
network control a developmental epithelial-mesenchymal transition. Development
141, 1503-1513. doi:10.1242/dev.101436
Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pietzsch,
T., Preibisch, S., Rueden, C., Saalfeld, S., Schmid, B. et al. (2012). Fiji: an
open-source platform for biological-image analysis. Nat. Methods 9, 676-682.
doi:10.1038/nmeth.2019
Schindler, A. J. and Sherwood, D. R. (2011). The transcription factor HLH-2/E/
Daughterless regulates anchor cell invasion across basement membrane in
C. elegans. Dev. Biol. 357, 380-391. doi:10.1016/j.ydbio.2011.07.012
Shen-Orr, S. S., Milo, R., Mangan, S. and Alon, U. (2002). Network motifs in the
transcriptional regulation network of Escherichia coli. Nat. Genet. 31, 64-68.
doi:10.1038/ng881
Sherwood, D. R. and Plastino, J. (2018). Invading, leading and navigating cells in
Caenorhabditis elegans: Insights into cell movement in vivo. Genetics 208, 53-78.
doi:10.1534/genetics.117.300082
Sherwood, D. R. and Sternberg, P. W. (2003). Anchor cell invasion into the vulval
epithelium in C. elegans.Dev. Cell 5, 21-31. doi:10.1016/S1534-5807(03)00168-0
Sherwood, D. R., Butler, J. A., Kramer, J. M. and Sternberg, P. W. (2005). FOS-1
promotes basement-membrane removal during anchor-cell invasion in C.
elegans.Cell 121, 951-962. doi:10.1016/j.cell.2005.03.031
Simmer, F., Tijsterman, M., Parrish, S., Koushika, S. P., Nonet, M. L., Fire, A.,
Ahringer, J., Plasterk, R. H. A., Louis, S., Road, T. C. et al. (2002). Loss of the
putative RNA-directed RNA polymerase RRF-3 makes C. elegans hypersensitive
to RNAi. Current 12, 1317-1319. doi:10.1016/S0960-9822(02)01041-2
Skene, P. J. and Henikoff, S. (2017). An efficient targeted nuclease strategy for
high-resolution mapping of DNA binding sites. eLife 6, e21856. doi:10.7554/eLife.
21856
Sturm, Á., Sasko
̋i, É., Tibor, K., Weinhardt, N. and Vellai, T. (2018). Highly
efficient RNAi and Cas9-based auto-cloning systems for C. elegans research.
Nucleic Acids Res. 46, e105. doi:10.1093/nar/gky516
Swift, J. and Coruzzi, G. M. (2017). A matter of time how transient transcription
factor interactions create dynamic gene regulatory networks. Biochim. Biophys.
Acta Gene Regul. Mech. 1860, 75-83. doi:10.1016/j.bbagrm.2016.08.007
Vega, S., Morales, A. V., Ocaña, O. H., Valdés, F., Fabregat, I. and Nieto, M. A.
(2004). Snail blocks the cell cycle and confers resistance to cell death. Genes
Dev. 18, 1131-1143. doi:10.1101/gad.294104
Verghese, E., Schocken, J., Jacob, S., Wimer, A. M., Royce, R., Nesmith, J. E.,
Baer, G. M., Clever, S., McCain, E., Lakowski, B. et al. (2011). The tailless
ortholog nhr-67 functions in the development of the C. elegans ventral uterus.
Dev. Biol. 356, 516-528. doi:10.1016/j.ydbio.2011.06.007
Wang, L., Shen, W., Lei, S., Matus, D., Sherwood, D. and Wang, Z. (2014). MIG-
10 (lamellipodin) stabilizes invading cell adhesion to basement membrane and is
a negative transcriptional target of EGL-43 in C. elegans.Biochem. Bioph ys. Res.
Commun. 452, 328-333. doi:10.1016/j.bbrc.2014.08.049
Wilkinson, H. A., Fitzgerald, K. and Greenwald, I. (1994). Reciprocal changes in
expression of the receptor lin-12 and its ligand lag-2 prior to commitment in a C.
elegans cell fatedecision. Cell 79, 1187-1198. doi:10.1016/0092-8674(94)90010-8
Xie, Q. and Cvekl, A. (2011). The orchestration of mammalian tissue
morphogenesis through a series of coherent feed-forward loops. J. Biol. Chem.
286, 43259-43271. doi:10.1074/jbc.M111.264580
Yakoby, N., Lembong, J., Schupbach, T. and Shvartsman, S. Y. (2007).
Drosophila eggshell is patterned by sequential action of feedforward and
feedback loops. Development 135, 343-351. doi:10.1242/dev.008920
13
RESEARCH ARTICLE Development (2020) 147, dev185850. doi:10.1242/dev.185850
DEVELOPMENT
Development: doi:10.1242/dev.185850: Supplementary information
Development • Supplementary information
Fig. S1. Improved RNAi vector increases penetrance of TF depletion phenotypes.
(A) Stacked bar graph depicting the penetrance of AC invasion defects at the P6.p 4-cell
stage comparing L4440-based versus T444T-based RNAi depletions in a uterine-specific
RNAi-hypersensitive background. Asterisk (*) denotes a statistically significant difference
between the vectors and represents a p-value < 0.03 by Fisher’s exact test (n ≥ 30
animals per treatment). (B) Stacked bar graphs depicting AC invasion defects following
delayed TF-RNAi treatment at the L2 stage (n = 30 animals per treatment). (C) Single
plane of confocal z-stack depicting early depletion of hlh-2 at the L1 stage. The AC-
specific membrane marker (magenta, cdh-31.5>mCherry::PLCδPH) and BM marker
(green, lam-1>LAM-1::GFP) are overlaid in each micrograph. (D) Single plane of confocal
z-stacks during the L4 stage (P6.p 8-cell stage) depicting representative phenotypes
associated with TF-RNAi treatment.
Development: doi:10.1242/dev.185850: Supplementary information
Development • Supplementary information
Fig. S2. Schematic of endogenous GFP-tagged loci of pro-invasive TFs. Codon-
optimized GFP was integrated at the N-terminus of fos-1A and hlh-2, at the C-terminus
of nhr-67, and internally in the egl-43 locus (tagging both isoforms). This figure was made
using http://wormweb.org/exonintron. Scale bar, 100 bp.
Development: doi:10.1242/dev.185850: Supplementary information
Development • Supplementary information
Development: doi:10.1242/dev.185850: Supplementary information
Development • Supplementary information
Figure S3. AC invasion phenotypes do not necessarily correlate to TF expression
levels. (A-D) Single planes of confocal z-stack depicting representative phenotype (single
vs. multi AC, bottom left of each image) of fluorescence alone (left; AC, magenta,
expressing cdh-3>mCherry::moeABD and BM, green) and DIC overlay (right). (E) Sina
plots of GFP-tagged TF levels, defined as the mean gray value of individual AC nuclei at
the P6.p 4-cell stage following TF-RNAi knockdown. In this and all other figures, statistical
significance as compared to empty vector controls is denoted as an open black circle and
here represents a p-value of < 1×10-6 by Student’s t test (n ≥ 50 animals per treatment).
Development: doi:10.1242/dev.185850: Supplementary information
Development • Supplementary information
Fig S4. HLH-2 does not regulate EGL-43 expression in the ventral uterus. (A) Single
plane of confocal z-stack depicting levels of egl-43::GFP::egl-43 in the ventral uterine
(VU) cells (white arrowheads) in empty vector controls (top) as compared to hlh-2(RNAi)
depletion. BM indicated by green arrowheads. (B) Sina plots of egl-43::GFP::egl-43
levels, defined as the mean gray value of individual VU nuclei, following hlh-2 RNAi
treatment (n ≥ 10 animals, n 37 VU cells quantified per treatment; n.s., not significant,
p-value = 0.45, Student’s t-test).
Development: doi:10.1242/dev.185850: Supplementary information
Development • Supplementary information
Figure S5. Induced expression of hlh-2 and nhr-67 reveal cell cycle independent
roles for hlh-2. (A-B) Single planes of confocal z-stack depicting representative
phenotype (single vs. multi AC) as fluorescence overlays (AC, magenta, expressing cdh-
3>mCherry::moeABD, and BM, green) following no heat shock control (left) as compared
Development: doi:10.1242/dev.185850: Supplementary information
Development • Supplementary information
to heat shock induced expression of HLH-2::2xTagBFP (blue) (A) and NHR-
67::2xTagBFP (B). (C-D) Bar graphs depict the penetrance of invasion defects in control
(black) compared to heat shock induced HLH-2 (C) and NHR-67 (D) at the P6.p 4-cell
stage (n 30 animals examined for each RNAi treatment; n.s., not significant; *p-value <
10-6; Fisher’s exact test).
Development: doi:10.1242/dev.185850: Supplementary information
Development • Supplementary information
Figure S6. Depletion of hlh-2 does not significantly increase the invasion defect of
an nhr-67(pf88) hypomorph. Bar graph depicts the penetrance of AC invasion defects
at the P6.p 4-cell stage in control (empty vector) as compared to hlh-2(RNAiL2) or nhr-
67(RNAi) treatment (n 50 animals examined for each RNAi treatment, n.s. not
significant, Fisher’s exact, p-value = 0.5350 (empty vector vs. hlh-2), 0.1690 (empty
vector vs. nhr-67) and 0.6676 (hlh-2 vs. nhr-67)).
Development: doi:10.1242/dev.185850: Supplementary information
Development • Supplementary information
Development: doi:10.1242/dev.185850: Supplementary information
Development • Supplementary information
Figure S7. Both isoforms of egl-43 function redundantly to regulate AC invasion.
(A) Schematics (via http://wormweb.org/exonintron) of GFP insertion into the egl-43 locus
to tag the long (top) or both isoforms (bottom). Scale bar, 100 bp. (B) DIC overlaid with
single confocal planes of GFP::egl-43L (top) and egl-43::GFP::egl-43 (bottom). (C) Sina
plots of GFP-tagged TF levels, defined as the mean gray value of individual AC nuclei,
following RNAi perturbation (n 25 animals for each treatment; n.s., not significant)
between RNAi treatments targeting egl-43L and egl-43. (D) Single plane of confocal z-
stacks depicting representative micrographs of control (top) versus egl-43(bmd135)
animals expressing egl-43L>GFP (green) and an AC reporter (magenta). (E) Stacked bar
graph depicting penetrance of AC invasion defect. Asterisk (*) denotes statistical
significance between control and mT1(-) animals and represents a p-value < 10-3 by
Fisher’s exact test (n ≥ 25 animals per treatment).
Development: doi:10.1242/dev.185850: Supplementary information
Development • Supplementary information
Table S1: Scoring table *percentages may not necessarily sum to 100 due to rounding
Genotype RNAi Treatment
(gene, vector, L1/L2 plating) P6.p
stage % Invaded (#ACs) % Not invaded (#ACs) n
0 1 2 3+ 0 1 2 3+
rrf-3(pk1426) II; unc-
119(ed3) III; rde-1(ne219)
V; fos-1a>RDE-1, myo-
2>YFP; cdh-
3>PH::mCherry; lam-
1>LAM-1::GFP + unc-
119(+)
empty vector T444T L1 4-cell 0 100 0 0 0 0 0 0 50
egl-43(RNAi) T444T L1 4-cell 0 12 1 0 0 27 19 41 75
egl-43(RNAi) T444T L2 4-cell 0 63 0 0 0 37 0 0 30
egl-43L(RNAi) T444T L1 4-cell 0 41 1 0 0 19 15 24 80
fos-1(RNAi) T444T L1 4-cell 0 62 0 0 0 38 0 0 66
fos-1(RNAi) T444T L2 4-cell 0 73 0 0 0 27 0 0 30
hlh-2(RNAi) T444T L1 4-cell 0 30 0 0 21 37 6 5 115
hlh-2(RNAi) T444T L2 4-cell 0 57 0 0 0 36 6 1 83
nhr-67(RNAi) T444T L1 4-cell 0 7 0 0 0 13 0 81 72
nhr-67(RNAi) T444T L2 4-cell 0 43 0 0 0 37 17 3 30
empty vector T444T L1 8-cell 0 100 0 0 0 0 0 0 50
egl-43(RNAi) T444T L1 8-cell 0 40 5 5 0 0 4 45 55
egl-43L(RNAi) T444T L1 8-cell 0 78 9 5 0 7 0 2 58
fos-1(RNAi) T444T L1 8-cell 0 90 0 0 0 10 0 0 50
hlh-2(RNAi) T444T L1 8-cell 0 41 3 0 24 24 1 8 76
hlh-2(RNAi) T444T L2 8-cell 0 64 1 1 0 23 1 9 81
nhr-67(RNAi) T444T L1 8-cell 0 32 0 0 0 0 0 68 57
empty vector L4440 L1 4-cell 0 100 0 0 0 0 0 0 30
egl-43(RNAi) L4440 L1 4-cell 0 55 0 0 0 0 0 45 31
fos-1(RNAi) L4440 L1 4-cell 0 84 0 0 0 16 0 0 32
hlh-2(RNAi) L4440 L1 4-cell 0 69 0 0 9 16 0 6 32
hlh-2(RNAi) L4440 L2 4-cell 0 80 0 0 0 13 3 3 30
nhr-67(RNAi) L4440 L1 4-cell 0 40 0 0 0 7 0 53 30
cdh-3(5kb)>CKI-1::GFP;
zmp-1>mCherry
empty vector T444T L1 4-cell 0 100 0 0 0 0 0 0 51
egl-43(RNAi) T444T L1 4-cell 0 10 0 0 0 90 0 0 31
fos-1(RNAi) T444T L1 4-cell 0 20 0 0 0 80 0 0 35
hlh-2(RNAi) T444T L2 4-cell 0 92 0 0 0 8 0 0 77
nhr-67(RNAi) T444T L1 4-cell 0 100 0 0 0 0 0 0 42
laminin::GFP; zmp-
1>mCherry
empty vector T444T L1 4-cell 0 100 0 0 0 0 0 0 27
egl-43(RNAi) T444T L1 4-cell 0 19 3 0 0 30 22 27 37
fos-1(RNAi) T444T L1 4-cell 0 49 0 0 0 51 0 0 53
hlh-2(RNAi) T444T L2 4-cell 0 56 0 0 0 31 3 10 59
nhr-67(RNAi) T444T L1 4-cell 0 47 0 0 0 0 35 18 62
egl-43>LoxP::GFP::EGL-
43; cdh-3>mCherry::
moeABD; laminin::GFP
empty vector T444T L1 4-cell 0 100 0 0 0 0 0 0 68
egl-43(RNAi) T444T L1 4-cell 0 0 0 0 0 8 14 78 158
egl-43(RNAi) T444T L2 4-cell 0 16 0 0 0 78 0 6 32
hlh-2>LoxP::GFP::HLH-2;
cdh-3>mCherry::moeABD;
laminin::GFP
empty vector T444T L1 4-cell 0 100 0 0 0 0 0 0 83
hlh-2(RNAi) T444T L2 4-cell 0 24 0 0 0 76 0 0 50
fos-1>LoxP::GFP::FOS-1;
cdh-3>mCherry::moeABD;
laminin::GFP
empty vector T444T L1 4-cell 0 100 0 0 0 0 0 0 74
fos-1(RNAi) T444T L1 4-cell 0 18 0 0 0 9 5 68 120
fos-1(RNAi) T444T L2 4-cell 0 93 0 0 0 7 0 0 30
nhr-67>NHR-67::GFP;
cdh-3>mCherry::moeABD;
laminin::GFP
empty vector T444T L1 4-cell 0 100 0 0 0 0 0 0 70
nhr-67(RNAi) T444T L1 4-cell 0 15 1 0 0 0 11 73 130
nhr-67(RNAi) T444T L2 4-cell 0 76 0 0 0 6 0 18 34
egl-43L>SEC::GFP::EGL-
43; laminin::GFP; cdh-
3>mCherry::moeABD; N/A 4-cell 0 67 3 0 0 8 17 6 51
egl-43L>SEC::GFP::EGL-
43/mT1; laminin::GFP;
cdh-3>mCherry::moeABD N/A 4-cell 0 100 0 0 0 0 0 0 25
nhr-67(pf88); qyIs227[cdh-
3>mCherry::moeABD] I;
qyIs7[laminin::GFP] X.
empty vector T444T L1 4-cell 0 34 0 0 0 0 0 66 58
hlh-2(RNAi) T444T L1 4-cell 0 23 0 0 0 0 0 77 51
nhr-67(RNAi) T444T L1 4-cell 0 27 0 0 0 1 0 71 65
Development: doi:10.1242/dev.185850: Supplementary information
Development • Supplementary information
Table S2: Rescue experiments *percentages may not necessarily sum to 100 due to rounding
Genotype Condition RNAi Treatment
(gene, vector, L1/L2 plating) P6.p
stage % Invaded % Not Invaded n
bmd142[hsp>HLH-
2::2xBFP] I; qyIs225[cdh-
3>mCherry::moeABD] V;
qyIs7[laminin::GFP] X.
control empty vector T444T L1 4-cell 100 0 30
control egl-43(RNAi) T444T L1 4-cell 37 63 30
control fos-1(RNAi) T444T L1 4-cell 53 47 30
control hlh-2(RNAi) T444T L2 4-cell 56 44 32
control nhr-67(RNAi) T444T L1 4-cell 41 59 34
heat shocked empty vector T444T L1 4-cell 100 0 30
heat shocked egl-43(RNAi) T444T L1 4-cell 43 57 30
heat shocked fos-1(RNAi) T444T L1 4-cell 43 57 30
heat shocked hlh-2(RNAi) T444T L2 4-cell 81 19 31
heat shocked nhr-67(RNAi) T444T L1 4-cell 40 60 43
bmd121[LoxP::hsp>NHR-
67::2xBFP] I; qyIs227[cdh-
3>mCherry::moeABD] I;
qyIs7[laminin::GFP] X.
control empty vector T444T L1 4-cell 100 0 30
control egl-43(RNAi) T444T L1 4-cell 19 81 30
control fos-1(RNAi) T444T L1 4-cell 43 57 31
control hlh-2(RNAi) T444T L2 4-cell 64 36 33
control nhr-67(RNAi) T444T L1 4-cell 33 67 30
heat shocked empty vector T444T L1 4-cell 100 0 30
heat shocked egl-43(RNAi) T444T L1 4-cell 17 83 30
heat shocked fos-1(RNAi) T444T L1 4-cell 40 60 30
heat shocked hlh-2(RNAi) T444T L2 4-cell 53 47 30
heat shocked nhr-67(RNAi) T444T L1 4-cell 30 70 30
Development: doi:10.1242/dev.185850: Supplementary information
Development • Supplementary information
Table S3: Strains
Strain Genotype Description Figure(s) Source
NK1316
rrf-3(pk1426) II; unc-119(ed3) III; rde-1(ne219)
V; qyIs102[fos-1a>RDE-1, myo-2>YFP];
qyIs24[cdh-3>PH::mCherry]; qyIs10[lam-
1>LAM-1::GFP with unc-119(with)]
uterine-specific RNAi strain with AC
and BM markers 1, 6, S1 Matus et
al., 2015
GS9129 arTi145[ckb-3>mCherry::H2B] II; lin-
12(ar624[lin-12>LIN-12::GFP]) III. endogenous lin-12 GFP reporter
with somatic gonad marker 2, 7 Attner et
al., 2019
DQM337 egl-43(bmd88[egl-43>LoxP::GFP::EGL-43]) II;
qyIs225[cdh-3>mCherry::moeABD] V. endogenous egl-43 GFP reporter
with AC marker 2, S7 This study
DQM497 fos-1(bmd138[fos-1>LoxP::GFP::FOS-1]) V. endogenous fos-1 GFP reporter 2 This study
DQM352 hlh-2(bmd90[hlh-2>LoxP::GFP::HLH-2]) I;
qyIs225[cdh-3>mCherry::moeABD] V. endogenous hlh-2 GFP reporter with
AC marker 2 This study
DQM291 nhr-67(syb509[nhr-67>NHR-67::GFP]) IV;
qyIs227[cdh-3>mCherry::moeABD] I. endogenous nhr-67 GFP reporter
with AC marker 2 This study
DQM335 egl-43(bmd88[egl-43>LoxP::GFP::EGL-43]) II;
qyIs225[cdh-3>mCherry::moeABD] V;
qyIs7[laminin::GFP] X.
endogenous egl-43 GFP reporter
with AC and BM markers 2-3, 7, S3,
S4, S7 This study
DQM515 fos-1(bmd138[fos-1>LoxP::GFP::FOS-1]) V;
qyIs227[cdh-3>mCherry::moeABD] I;
qyIs7[laminin::GFP] X.
endogenous fos-1 GFP reporter with
AC and BM markers 2-3, 7, S3 This study
DQM350 hlh-2(bmd90[hlh-2>LoxP::GFP::HLH-2]) I;
qyIs225[cdh-3>mCherry::moeABD] V;
qyIs7[laminin::GFP] X.
endogenous hlh-2 GFP reporter with
AC and BM markers 2-3, 7, S3,
S7 This study
DQM368 nhr-67(syb509[nhr-67>NHR-67::GFP]) IV;
qyIs225[cdh-3>mCherry::moeABD] V;
qyIs7[laminin::GFP] X.
endogenous nhr-67 GFP reporter
with AC and BM markers 2-3, 7, S3,
S7 This study
DQM7 qyIs330 [laminin::mCherry]; qyIs232 [CDT-
1::GFP] G1 cell cycle phase reporter with BM
marker 4 Matus et
al., 2015
DQM533 zmp-1(qy17[zmp-1::LoxP::GFP]) III;
qyIs225[cdh-3>mCherry::moeABD] V endogenous zmp-1 GFP reporter
with AC marker 4 This study
DQM39 qyIs17 [zmp-1>mCherry] II; qyIs266[cdh-
3(5kb)>CKI-1::GFP] V zmp-1 mCherry reporter with AC-
specific CKI-1 over-expression 5 Matus et
al., 2015
NK272 qyIs7[laminin::GFP]; qyIs17[zmp-1>mCherry] zmp-1 mCherry reporter with BM
marker 5 Matus et
al., 2015
DQM503 egl-43(bmd87[egl-43>SEC::GFP::EGL-43]
II/mT1; qyIs227[cdh-3>mCherry::moeABD] I;
qyIs7[laminin::GFP] X.
endogenous egl-43 transcriptional
reporter (with SEC) balanced over
mT1 with AC and BM markers 6 This study
DQM444 bmd121[LoxP::hsp>NHR-67::2xBFP] I;
qyIs227[cdh-3>mCherry::moeABD] I;
qyIs7[laminin::GFP] X. heat shock inducible NHR-67 with
AC and BM markers S5 This study
DQM552 bmd142[hsp>HLH-2::2xBFP] I; qyIs225[cdh-
3>mCherry::moeABD] V; qyIs7[laminin::GFP]
X. heat shock inducible HLH-2 (codon-
optimized) with AC and BM markers S5 This study
DQM266 nhr-67(pf88); qyIs227[cdh-3>mCherry::
moeABD] I; qyIs7[laminin::GFP] X. nhr-67 hypomorphic mutant with AC
and BM markers S6 This study
DQM500 bmd135[egl-43L>SEC::GFP::EGL-43] II/mT1;
qyIs227[cdh-3>mCherry::moeABD] I;
qyIs7[laminin::GFP] X.
endogenous egl-43L transcriptional
reporter (with SEC) balanced with
mT1 balancer with AC and BM
markers
S7 This study
DQM494 egl-43(bmd136[egl-43L>LoxP::GFP::EGL-43])
II. endogenous egl-43L GFP reporter S7 This study
DQM300 egl-43(bmd88[egl-43>LoxP::GFP::EGL-43]) II. endogenous egl-43 GFP reporter Will be made available
through the CGC
(cgc.umn.edu)
DQM497 fos-1(bmd138[fos-1>LoxP::GFP::FOS-1]) V. endogenous fos-1 GFP reporter
DQM311 hlh-2(bmd90[hlh-2>LoxP::GFP::HLH-2]) I. endogenous hlh-2 GFP reporter
PHX509 nhr-67(syb509[nhr-67>NHR-67::GFP]) IV. endogenous nhr-67 GFP reporter
Development: doi:10.1242/dev.185850: Supplementary information
Development • Supplementary information
Table S4: Primers
Table S5: Plasmids
Plasmid Base vector Description
pTNM011 pDD122 egl-43 internal sgRNA
pTNM012 pDD282 egl-43::GFP::egl-43 repair template
pTNM013 pDD122 fos-1 N-terminal sgRNA
pTNM014 pDD282 GFP::fos-1 repair template
pTNM015 pDD122 hlh-2 N-terminal sgRNA
pTNM016 pDD282 GFP::hlh-2 repair template
pTNM046 pDD122 egl-43L N-terminal sgRNA
pTNM047 pDD282 GFP::egl-43L repair template
pTNM051 pAP088 heat shock inducible HLH-2 (codon-optimized)
pWZ172 T444T egl-43 RNAi
pWZ173 T444T egl-43L RNAi
pWZ174 T444T fos-1 RNAi
pWZ175 T444T hlh-2 RNAi
pWZ176 T444T nhr-67 RNAi
pWZ193 pAP088 heat shock inducible NHR-67
Primer Primer sequence (5’→3’) Primer
type Amplicon Template
DQM657 tcactatagggagaccggcaATG Forward hlh-2 and nhr-67 synthesized DNAs for
T444T
Twist Biosciences
gene fragments for
hlh-2, nhr-67
DQM658 attgggtaccgggcccc Reverse
DQM688 gagctcAGATCTatgagcatcgacacagacttc Forward BglII-egl-43L-XhoI for T444T Twist Biosciences
gene fragment for
egl-43L
DQM689 acgtacCTCGAGctgactttgacacgttgggc Reverse
DQM720 TCACTATAGGGAGACCGGCAATG Forward BglII-egl-43-SalI for T444T egl-43 IDT gBlock
DQM722 GCCCCCCCTCGAGGTCGAACTTTTGGCAC
CGGAAC Reverse
DQM708 ggtttcgccacctctgacttg Forward colony PCR screening of T444T
constructs T444T-based
constructs
DM191 gtaatacgactcactatagggcgaattgg Reverse
DQM433 ACGTTGTAAAACGACGGCCAG Forward amplify left homology arm (universal) Twist Biosciences
gene fragments
DQM434 CTCCAGTGAACAATTCTTCTCCTTTACTC Reverse amplify left homology arm (universal)
DQM435 GCGTGATTACAAGGATGACGATGAC Forward amplify right homology arm (universal)
DQM436 GAAACAGCTATGACCATGTTATCGATTTCC Reverse amplify right homology arm (universal)
DQM751 tcctattgcgagatgtcttGTCCACTCTCTTATATAG
CAGGTTTTAGAGCTAGAAATAGC Forward fos-1a sgRNA pDD122
DQM747 tcctattgcgagatgtcttGgatgctcatcctgaaaacttGTT
TTAGAGCTAGAAATAGC Forward egl-43L sgRNA pDD122
DQM438 tcctattgcgagatgtcttGaagtcagATGCCATCACA
AGGTTTTAGAGCTAGAAATAGC Forward egl-43 sgRNA pDD122
DQM440 tcctattgcgagatgtcttGAGTTTTCAGAACCTCAA
TGGGTTTTAGAGCTAGAAATAGC Forward hlh-2 sgRNA pDD122
DQM412 AGATTGTACTGAGAGTGCACCATATGCGG
TGTGAAATACCGCAC Reverse amplify sgRNA (universal) pDD122
Development: doi:10.1242/dev.185850: Supplementary information
Development • Supplementary information
Table S6: CRISPR reagents
Gene Guide Left homology arm sequence Right homology arm sequence
egl-43 (internal)
aagtcagatgccatcacaag
CTAAGATATGAGAACCCGTTTACAGAGTCCTTTTA
TAGAATTTTGGTTTTTATAATAGAGATGTATGGAA
ACCGGGCAAAGTTAATTAGGATTCTTACAGCCAA
CAAGAAAATTTTAAGATAAGTAACGACCAAAACTT
GAGCGGAGTTGAAAGTTCAGTGCATTACAATTGA
AGTTTTTATTTATTATTTATTTATCTTGGGCTAGAA
TAGATGGAGCATCTTCACAGGACTGGACTTATAT
AATGGCTATCTGCCTGCCTGCCTACCTGCCTTTC
GTTTTATTTATACTGTATTTGCGCAATATAAACTCA
TGCATTTCTATTTGTTAGAAGTTAAAAAAAACAAT
ATTTAGAAAACGTTCACATGTACTTTGATAGTTGG
CTCGCATGTTGGCAGAAGGCAGGCACAGGGAAC
CCTGAAGGCACTTAGGCAGGTTCGCTGGACAAA
ATACCTGCTGATTGTGCTGATTTATTTTCACTAAT
ATAATCAGATATGAAACATGGACAATTGGGTACA
ATACTAATAGAGGGTGTACTGCTATATAACTTCTC
CGAATCAAAACTTAACAGTCATTCACATTCACTAT
CACGTGTTATCAATCAATTTTTTTCAGCGGTCTCA
AACAACACTCTCATATTCATTGCTCCTTAAAGCCG
TTTCGGTGTCATTT