Description
Organisms adapt to their dynamic internal and external environments by leveraging systemic physiological responses using signaling pathways involving multiple tissues and extensive cross-talk. Recently, the gut has emerged as an important signaling center keenly attuned to an animal's feeding state, food quality, and pathogenic exposure. Within
C. elegans
, gut-derived secreted peptide signals influence feeding, stress response, waste release and avoidance behavior by acting on neurons of the enteric and central nervous systems (Artan et al., 2016; Jia et al., 2024; Sural et al., 2025). For example,
INS-7
and
INS-35
regulate the NSM neuron to adjust pharyngeal pumping rates (Sural et al., 2025). Additional insulin family members such as
INS-3
and
INS-11
function in gut-to-brain signaling pathways (Lee and Mylonakis, 2017; Veuthey et al., 2025). Several non-insulin peptides are also involved.
FLP-2
(FMRFamide-like peptide) acts via the interneuron AIY to induce responses to oxidative stress within the animal (Jia et al., 2024).
NLP-40
(Neuropeptide-like protein) is released to instruct the posterior enteric neurons, AVL and DVB, to elicit steps of the digestive/defecation motor program; its release is tied to a periodic calcium wave that occurs in well-fed animals with a regular frequency of 45-55 seconds (Choi et al., 2023; Mahoney et al., 2008; Wang et al., 2013).
This regular calcium wave may trigger release of additional peptides that influence nervous system behavior more broadly. Several mutants with poor digestive waste release originally identified due to their constipated phenotypes turn out to encode integral components of peptide processing and release such as
egl-3
,
egl-21
,
aex-1
,
aex-4
, and
aex-5
(Husson et al., 2007; Khawand et al., 2026; Mahoney et al., 2008). At least one peptide,
NLP-40
, is released in time with the digestive motor program, via activation of synaptotagmin-2 (
snt-2
)
, a calcium-sensing protein that triggers DCV fusion (Wang et al., 2013). The tyramine response pathway offers additional evidence of calcium-triggered peptide release (Veuthey et al., 2025). This enticing evidence suggests that the rhythmic calcium flux that triggers
NLP-40
release, and other steps of the digestive/defecation motor program, may be used more broadly to influence behavioral choices where feeding and nutritional status are relevant. With this notion in mind, we explored an observation of social feeding behavior in a mutant that alters intestinal calcium wave dynamics.
The
innexin-16
(
inx-16
) gap junction subunit is critical for periodic intestinal calcium waves associated with the digestive motor program; loss of function
inx-16
mutants disrupt the intestinal calcium wave and greatly reduce the digestive motor program steps that require
NLP-40
release (Peters et al., 2007). During routine maintenance of
inx-16
(
tm1589
)
, we observed increased aggregation of well-fed worms and became intrigued. To explore this further, we assayed for social feeding behavior quantitatively. For comparison, we included
N2
, a solitary strain, and a social feeding strain,
npr-1
(
ad609
).
The gene,
npr-1
, encodes a
C. elegans
receptor related to neuropeptide Y and is neuronally expressed. When active,
NPR-1
inhibits the tendency of animals to aggregate in groups, suppresses oxygen level preferences, and affects bordering (de Bono and Bargmann, 1998; Gray et al., 2004). The
npr-1
allele we used,
ad609
, has been characterized to be a strong loss of function allele, equivalent to a genetic null (de Bono and Bargmann, 1998). Our results show that
inx-16
mutant animals maintain contact with other animals more frequently than wild-type (
N2
) solitary worms; young adult
inx-16
(
tm1589
)
animals demonstrated a significant difference in the number of
inx-16
animals found in direct contact with their peers compared to
N2
(~0% vs. ~15%;
Figure 1A
). Yet, the
inx-16
phenotype is much weaker than the
npr-1
(
ad609
)
allele (~80% vs. ~15%;
Figure 1A
)
.
To verify that
INX-16
functions within the intestine to cause the observed tendency to aggregate, we employed tissue specific rescue of
inx-16
mutants. Our and others'
inx-16
expression analyses indicated that
inx-16
was only expressed in the intestine, yet two studies have provided phenotypic evidence suggesting additional areas of function (Bhattacharya et al., 2019; Liu et al., 2013; Sojka et al., 2025). Intestine-specific rescue was achieved using a construct with a short, well-defined fragment, regulatory sequence of the egg yolk protein, vitellogenin-2, (
vit-2
) gene, placed upstream of an
INX-16
translational fusion (MacMorris et al., 1992). Prior work using this construct led to full rescue of the defecation defects in adults (Peters et al., 2007). Likewise, solitary feeding behavior was restored in the
inx-16
+
p
vit-2
intestinal rescue line (
Figure 1A
). By contrast, driving expression of
inx-16
in body wall muscles, using the
myo-3
promoter, was unable to restore solitary (wild-type) feeding behavior (
Figure 1A
; Liu et al., 2013). These data suggest that intestinal
INX-16
function contributes to the animal's feeding behaviors.
Having shown that
inx-16
'
s actions in the intestine can rescue the social feeding phenotype, we queried whether aggregated feeding was a common feature of mutants with impaired digestive waste release (Thomas, 1990). Defecation mutants induce constipation. One constipated mutant with defects in the expulsion step,
exp-1
, disrupts an excitatory GABA-ergic receptor in enteric muscles and has been shown to mildly increase social feeding behavior (Beg and Jorgensen, 2003; Bendesky et al., 2012). We assayed this mutant, and others, selecting a set of mutants that affect each defecation motor step and all tissue types involved in the motor program (Branicky and Hekimi, 2006; Thomas, 1990). Two genes required for the first motor step (the posterior body contraction), the intestinal sodium proton exchanger,
pbo-4
/nhx-7
, and the downstream proton-gated ion channel,
pbo-5
, localized to body wall muscles, were tested (Beg et al., 2008; Pfeiffer et al., 2008). Neither of these Pbo
mutants showed increases in social feeding in comparison to
N2
(
Figure 1B
). The later steps of the program, anterior body contraction and enteric muscle contraction, require
NLP-40
release from the intestine to stimulate GABA release from enteric neurons, AVL and DVB (Wang et al., 2013).
NLP-40
binding activates
AEX-2
, a g-protein coupled receptor, on these neurons (Allman et al., 2016; Mahoney et al., 2008). Like the Pbo mutants,
aex-2
(
sa3
)
showed no significant increase over wild-type in our aggregation assay (
Figure 1B
). In our assays,
exp-1
(
sa6
)
did not increase aggregation, which differs from prior findings that reported a small but significant increase (Bendesky et al., 2012). Though our defecation mutant panel is not exhaustive, these data suggest that the social feeding phenotype of
inx-16
is not broadly shared with mutants suffering from constipation; disrupting defecation per se does not alter social feeding behavior in
C. elegans
.
Two key gene pathways control the choice between solitary and social feeding behavior:
npr-1
and
daf-7
. Therefore, we used genetic assays to elucidate potential connections between
inx-16
and these known pathways (de Bono and Bargmann, 1998; de Bono et al., 2002). Double mutants of
inx-16
with
daf-7
(
e1372
)
and
npr-1
(
ad609
)
were constructed and tested alongside single mutants to determine whether the double mutant lines had enhanced aggregation. The combination of
inx-16
and
daf-7
(
e1372
)
enhanced aggregation substantially, with the
inx-16
;
daf-7
double mutant exhibiting increased aggregation compared to the
daf-7
mutant alone, approximately 23% for the double compared to ~13% for either single mutants (
Figure 1C
). This additive effect is consistent with
inx-16
functioning at least partly independently of
daf-7
. By contrast, the combination of
inx-16
and the strong loss of function
npr-1
allele,
ad609
, did not augment the social feeding level relative to that of
npr-1
alone. The
inx-16
;
npr-1
double mutants exhibit no increase in aggregation compared to the
npr-1
mutant alone (60% for the double compared to 58% for the single mutant,
Figure 1C
). The results of a two-way ANOVA analysis support a highly significant genetic interaction between
inx-16
and
npr-1
(P < 0.003; F = (1, 20)=19). These findings are consistent with
inx-16
functioning via the
npr-1
pathway to mediate social feeding behavior.
Since our findings placed
inx-16
within the
npr-1
pathway rather than the
daf-7
pathway, we searched for
NPR-1
ligands with evidence of intestinal expression. Of the two identified
NPR-1
ligands,
FLP-18
and
FLP-21
,
flp-21
showed intestinal expression in addition to a subset of neurons (Kubiak et al., 2003; Rogers et al., 2003). The
flp-21
mutation causes a significantly increased propensity for animals to aggregate, enhancing aggregation to ~15% (
Figure 1D
). Interestingly,
flp-21
's
magnitude of impact on feeding behavior is similar to that of
inx-16
mutation, ~15% for
flp-21
vs. ~14% for
inx-16
(
Figure 1D
; Rogers et al., 2003). Thus, we hypothesized that
FLP-21
could be released from the intestine following calcium waves in a manner analogous to
NLP-40
release. To begin to investigate this model, we explored epistasis between
inx-16
and
flp-21
. The double mutant line exhibited an aggregation profile very similar to that of either individual mutant strain (
inx-16
;
flp-2
1 = ~
13%;
Figure 1D
). Statistical analysis for gene interaction upholds this epistatic relationship (P < 0.0001; F = (1, 28) = 89). Our results support a model placing
FLP-21
release downstream of
INX-16
function, and by inference the rhythmic intestinal calcium wave that directs the digestive motor program (Dal Santo et al., 1999; Espelt et al., 2005; Teramoto and Iwasaki, 2006).
Due to
flp-21
's reported intestinal localization, we suggest that
inx-16
's
social feeding phenotype may be due to reduced
FLP-21
release from the intestine. Since
NPR-1
signaling dampens social feeding, loss of
FLP-21
would increase aggregation, as seen in both
inx-16
and
flp-21
mutants. Our model is supported by the previously reported epistasis analysis of
flp-21
(
pk1601
) and
npr-1
(
ad609
);
this
flp-21
allele did not enhance
npr-1
(
ad609
)'s social feeding level, matching our results (
Figure 1C
; Rogers et al., 2003). Given that
flp-18
, the other known
NPR-1
ligand, is not expressed in the intestine,
flp-18
presumably remains active in the
inx-16
mutant, allowing ample residual suppression of social feeding to result in the modest aggregation seen in both the
inx-16
and
flp-21
mutants (
Figure 1D
; Kapahi et al., 2010).
In summary, our findings demonstrate a behavioral feeding phenotype associated with disrupted intestinal calcium wave dynamics. We propose that this phenotype is due to disruption of
FLP-21
mediated activation of
NPR-1
. Moreover, the social feeding phenotype of
inx-16
suggests that the periodic intestinal calcium waves serve to regulate peptide release broadly, influencing a host of behaviors beyond defecation and adding to the growing body of evidence illustrating the importance of gut-to-brain signaling for animal behavior.