Gut regulates brain synaptic assembly through neuroendocrine signaling pathway

The gut-brain axis plays an essential role in regulating neural development in response to environmental stimuli, such as microbes or nutrients. Defects in gut-brain communication can lead to various neurological disorders. However, it remains unknown whether gut plays any intrinsic role in regulating neuronal development. Through a genetic screen in C. elegans, we uncovered that an intrinsic Wnt-endocrine pathway in gut regulates synaptic development and neuronal activity in brain. Specifically, the Wnt signaling upregulates the expression of the neuropeptide NLP-40 in the gut, which then facilitates presynaptic assembly through the neuronal expressed GPCR AEX-2 receptor during development. The NLP-40 acts most likely through modulating neuronal activity and promoting synaptic protein trafficking. Therefore, this study reveals a novel role of gut in synaptic development in the brain.


INTRODUCTION
The gastrointestinal (GI) tract plays a critical role in regulating brain development and function. It not only provides the nutrient, but also senses the gut environmental stimuli. The gut-brain communication is recognized as gutbrain axis. Through this axis, diets and gut microbes regulate neurogenesis, microglia and astrocyte activation, myelination, blood brain barrier permeability, synaptic pruning and plasticity (Cryan et al., 2019). Dysregulation in the gutbrain axis is correlated with various neurodevelopmental or neurodegenerative disorders (Chidambaram et al., 2020;Kowalski and Mulak, 2019;Li and Zhou, 2016;Mulak and Bonaz, 2015;Quigley, 2017). Although accumulative evidence indicates that gut signaling affects neuronal development and function in the brain, it remains largely unknown if it plays any intrinsic role in brain development.
Gut-brain communication is mainly mediated through three pathways: afferent nervous system, immune system and endocrine system (Wang and Wang, 2016). Among those, the endocrine system is the most complex and highly conserved in both vertebrates and invertebrates (Bernhard Kleine and Rossmanith;Campbell et al., 2004;Dimaline and Dockray, 1994;Nassel et al., 2019). Upon stimulation by nutrients or gut microbiota, enteroendocrine cells (EECs) release neuropeptides or hormones such as cholecystokinin (CCK), ghrelin, leptin, peptide tyrosine-tyrosine (PYY), and glucagon-like peptide-1 (GLP-1) which can travel a long distance as endocrine factors through binding GPCR or other types of receptors to regulate the feeding behavior and the physiological function of the nervous system (Dockray, 2013;Wang et al., 2013).
However, it is largely undetermined whether the intestinal endocrine system is involved in neurodevelopment or synaptic formation.
In this study, we uncovered that a canonical Wnt pathway in the intestine regulates synaptic development in the nerve ring, which is the nematode's brain.
We further demonstrated that the Wnt signaling functions through upregulating the neuroendocrine molecule NLP-40, which then promotes the synaptic assembly most likely through modulating the neuronal activity. Therefore, our study reveals a novel gut-brain interaction underlying synaptogenesis in brain.

Intestinal Wnt signaling regulates AIY presynaptic formation
C. elegans AIY neurons are a pair of bilateral symmetric interneurons located in the nerve ring, which is a structure analogous to vertebrate's brain ( Fig.   1A) (White et al., 1986). AIY neurites form presynapses with a stereotypic and highly reproducible pattern ( Fig. 1A') (White et al., 1986).
Wnt signaling pathways are highly conserved and play complicated roles in regulating synaptic formation, plasticity and maintenance (Dickins and Salinas, 2013;McLeod and Salinas, 2018;Yang and Zhang, 2020). We previously found that components in the canonical Wnt signaling pathway including Wnt/CWN-2, Frizzled/CFZ-2, Disheveled/DSH-2, b-catenin/SYS-1 and TCF/POP-1 are required for the AIY presynaptic morphogenesis (Shi et al., 2018). To address if the Wnt signaling pathway regulates the AIY presynaptic formation, we first quantified the fluorescent intensity of the AIY synaptic vesicle marker GFP::RAB-3 in cwn-2(ok895), cfz-2(ok1201), pop-1(hu9) loss-of-function mutant and sys-1 RNAi animals. We found that loss of this Wnt signaling resulted in a robust reduction of AIY synaptic marker GFP::RAB-3 (GFP intensity was reduced by 43.94%, 59.32%, 47.10% and 38.54% in cwn-2(ok895), cfz-2(ok1201), pop-1(hu9) and sys-1 RNAi. P<0.001, 0.0001 and 0.001 compared with WT for mutants, respectively, and P<0.01 for comparison between sys-1 and control RNAi; Fig. 1B-H). We then examined whether the regulatory role of Wnt was specific to synaptic vesicle, so we examined the active zone marker GFP::SYD-1. Similar to GFP::RAB-3, GFP::SYD-1 was significantly reduced in these Wnt mutants . To exclude the possibility that the reduction of the presynaptic marker was specific to the GFP fluorophore, we quantified the mCherry::RAB-3 and obtained similar results (Supplementary Fig. 1A-G). These results collectively suggest that the canonical Wnt signaling is required for AIY presynaptic assembly.
To determine the site of action for the Wnt signaling, we knocked down the Wnt receptor and the downstream signaling components through neural or intestinal specific RNAi. Interestingly, we found robust reduction of AIY 1I'-N'). Consistently, the cfz-2(ok1201) mutants could only be rescued by the intestine-specific, but not by the neuronal-specific, expression of cfz-2 cDNA (Supplementary Fig. 2D-F). These data collectively suggest that CWN-2 acts through canonical Wnt components CFZ-2/SYS-1/POP-1 in the gut to regulate AIY presynaptic formation.

Intestinal neuropeptide/NLP-40 is required for synaptogenesis
Gut can communicate with brain through neuroendocrine signaling pathway (Campbell et al., 2004;Dimaline and Dockray, 1994;Nassel et al., 2019). To test whether Wnt signaling in the gut regulates AIY synaptogenesis through neuroendocrine or neuropeptide signaling, we first examined if EGL-3, a key proprotein convertase (PC2) involved in neuropeptide maturation (Li and Kim, 2008), was required for the AIY presynaptic assembly. Indeed, the fluorescence intensity of AIY presynaptic markers was robustly reduced in egl-3(n589) mutants (P<0.01, Fig. 2A), suggesting that neuropeptides are required for the AIY presynaptic assembly. There are 115 neuropeptide encoding genes identified in C. elegans genome (Li and Kim, 2008;Nathoo et al., 2001). We screened most neuropeptide genes enriched in the intestine through RNAi for AIY synaptic regulators (Fig. 2B) (Li and Kim, 2008;Pauli et al., 2006;Spencer et al., 2011;Wang et al., 2013). Interestingly, nlp-40 knockdown displayed a dramatic reduction of the synaptic marker, suggesting that it is required for the AIY presynaptic assembly (Fig. 2C).
To confirm if AIY synaptogenesis requires nlp-40, we first examined the AIY synaptic phenotype in two independently isolated nlp-40 mutant alleles, tm4085 and vj3 (Wang et al., 2013). nlp-40(tm4085) harbors an in-frame deletion that deletes the neuropeptide P1 and part of P2, an A70 to V mutation in the P3 and an early stop deleting the last three amino acid at the C-terminus of P4 ( Fig. 2D and Supplementary Fig. 3). The vj3 deletes the nlp-40 promoter, the first exon and first intron ( Fig. 2D) (Wang et al., 2013). Similar to the RNAi results, we observed a robust reduction of the AIY synaptic vesicle marker GFP::RAB-3 in both nlp-40 mutant alleles (the GFP intensity was reduced by 79.01% and 67.49% respectively. P<0.001 compared to WT for both alleles; suggesting that the role of nlp-40 in AIY presynaptic assembly is not simply dosage dependent (Supplementary Fig. 5A-C). Based on these data, we conclude that that nlp-40 is required for AIY synaptogenesis.
Next, we wanted to determine the NLP-40 action site. First, we examined the nlp-40 expression pattern with both transcription and translational reporters and found nlp-40 was high expressed in the intestine beginning at early embryonic stages, and the protein was observed both in the intestine and the coelomocytes, which is consistent with previous report (Supplementary Fig. 6A-E) (Wang et al., 2013). Then, using intestinal specific RNAi to knock down nlp-40, we observed a robust reduction of AIY presynaptic marker GFP::RAB-3, phenocopying the nlp-40 mutants (the GFP::RAB-3 was reduced by 54.48%. P < 0.01; Fig. 2K-M). These data collectively suggest that nlp-40 is expressed in the intestine and secreted to pseudocoelom. We posit intestinal NLP-40 likely travels to the nervous system and promotes AIY presynaptic assembly.
To address when nlp-40 is required for the AIY presynaptic formation, we quantified the fluorescent intensity of AIY GFP::  mutants at the larval L1, L4 and adult day 1 stage. We found the GFP intensity was dramatically reduced in the nlp-40(tm4085) mutants as of the newly hatched L1 stage throughout the adult stage (Supplementary Fig. 6F-H, 6F'-H', 6I). These data suggest that intestinal nlp-40 is required for AIY presynaptic assembly during embryonic development.

nlp-40 acts downstream of the Wnt signaling pathway to promote AIY presynaptic assembly
Thus far, we demonstrated that AIY presynaptic assembly requires both Wnt signaling and nlp-40 from the intestine. To determine the nature of their interaction, we examined the genetic interaction between the genes encoding Wnt signaling components and nlp-40 by building three sets of double mutants: cwn-2(ok895);nlp-40(tm4085), cfz-2(ok1201);nlp-40(tm4085) and pop- 1(hu9);nlp-40(tm4085). We found that the fluorescent intensity of the AIY presynaptic marker GFP::RAB-3 in all double mutants was similar to that in nlp-40(tm4085) single mutants, but slightly weaker than in Wnts single mutants (Fig.   3A). The epistatic effects of nlp-40 to Wnt genes suggest that nlp-40 acts downstream of the Wnt signaling pathway to regulate AIY presynaptic assembly.
To test the genetic interaction between nlp-40 and aex-2, we quantified the AIY presynaptic marker GFP::RAB-3 in the nlp-40(tm4085);aex-2(sa3) double mutants. We found that the GFP intensity in the double mutants was similar to that in either single mutants ( Fig. 5B-E, I), indicating that nlp-40 and aex-2 act in the same genetic signaling pathway to regulate AIY presynaptic assembly.
The genetic interaction data is consistent with a model that AEX-2 acts as the NLP-40 receptor to regulate the AIY presynaptic assembly.
Next, we wanted to determine the aex-2 action site. First, we built an aex-2 transcription reporter, which showed that aex-2 was mainly expressed in the nervous system with some in the posterior intestine ( Fig. 5K), which is consistent with the previous report (Wang et al., 2013). When colabeled with the AIY cytoplasmic marker Pttx-3::mCherry, we noted that the aex-2 reporter was co-labeled with the AIY marker ( Fig. 5L-N'), indicating that aex-2 is expressed in AIY interneurons. To address where aex-2 acts, we expressed aex-2 cDNA with AIY-or intestine-specific promoter in the aex-2(sa3) mutants, and found that only AIY-specific expression rescued the presynaptic defect ( Fig.  5G, H, J). These data suggest that aex-2 acts cell-autonomously in the AIY interneurons to regulate presynaptic formation.
These data suggest that Wnt-NLP-40 signaling pathway promotes AIY synaptic assembly probably through modulating neuronal activity.

NLP-40 regulates synaptic protein trafficking
Neuropeptides and neurosteroids can regulate postsynaptic receptor trafficking (Lorenz-Guertin and Jacob, 2018). To test if NLP-40 regulates AIY presynaptic protein transports, we performed time-lapse to analyze the GFP::RAB-3 trafficking in the zone 1 region (Fig. 7A). We found that in the nlp-40 mutants, the number of the GFP::RAB-3 puncta and the transport velocity of anterograde, but not retrograde, were significantly reduced ( Fig. 7B-C"). Interestingly, the fluorescent intensity of GFP::RAB-3 in the soma was also significantly reduced, which is probably due to post transcriptional regulation since the promoter activity was not affected in nlp-40 mutants ( Fig. 7D and 7E). The decrease of GFP::RAB-3 could be due to the side effect of the inefficient trafficking. Collectively, those data suggest that NLP-40 regulates synaptic assembly most likely through modulating synaptic protein trafficking.

The Wnt/nlp-40/aex-2 signaling pathway is required for the presynaptic assembly in the nerve ring in general
Thus far, we have demonstrated that Wnt-NLP-40 signaling in the intestine is required for AIY presynaptic assembly. To determine if this regulation is general for presynaptic assembly or AIY specific, we first quantified GFP::RAB-3 in the nerve ring region in Wnt signaling mutants. Interestingly, the synaptic marker GFP::RAB-3 was significantly reduced in cwn-2(ok895), cfz-2(ok1201) and pop-1(hu9) mutants (the intensity was reduced by 33.98%, 45.59% and 52.64% respectively, P<0.05, 0.01 and 0.001 as compared to the WT for mutants, respectively; Fig. 8A-E, H). Furthermore, we found that the reduction of the presynaptic marker in cwn-2(ok895) was rescued by a wild type transgene (Fig.   8I). Those data collectively suggest that the Wnt signaling is required for the presynaptic assembly in the nerve ring.
Next, we examined the presynaptic marker in nlp-40 and aex-2 mutants. We found that the GFP::RAB-3 intensity is dramatically reduced in nlp- 40(tm4085) or aex-2(sa3) mutants (the intensity is reduced by 44.89% and 49.65% respectively, P<0.01 and 0.001 compared to WT; Fig. 8F, G and H). The reduction of the presynaptic marker was rescued by the corresponding wild type transgene (Fig. 8I). The results indicate that nlp-40 and aex-2 are required for presynaptic assembly in nerve ring.

DISCUSSION
Overall, we demonstrate that a canonical Wnt signaling pathway that includes Wnt/CWN-2, Frizzled/CFZ-2, b-catenin/SYS-1 and TCF/POP-1 acts in the intestine to promote synaptic assembly in the C. elegans nerve ring, a structure analogous to mammalian brain. Wnt signaling exerts this effect by upregulating the expression of nlp-40, which encodes a conserved neuropeptide. NLP-40 promotes synaptic assembly through the neuronal expressed GPCR receptor, AEX-2 ( Fig. 9). This gut Wnt-neuroendocrine signaling promotes synaptic assembly most likely through modulating neuronal activity. We therefore uncover a previously unidentified molecular mechanism in which the Wntneuroendocrine signaling pathway in the intestine regulates synaptic assembly.
These results highlight the intrinsic role of gut in the neurodevelopment, and reveal a critical effect of novel gut-brain communication.

Wnt-Neuropeptide axis regulates synaptic assembly
The genes encoding Wnt signaling pathway components and their roles in synaptogenesis are highly conserved in C. elegans (He et al., 2018;Park and Shen, 2012). Wnt signaling plays critical roles in synaptic development.
However, the underlying mechanisms are not completely understood. In this study, we demonstrated that intestinal Wnt signaling promotes synaptic assembly in the nerve ring. This Wnt signaling promotes synaptic assembly by upregulating the expression of the intestinal secreted neuropeptide, NLP-40, which acts through the neuronal expressed GPCR, AEX-2.
Wnt ligands act either in the nervous system or in non-neuronal tissues to regulate synaptic assembly or plasticity. However, the receptors or downstream components have only been shown to act in neurons or their target cells. For example, the Wnt ligands, LIN-44 and EGL-20, expressed in the tail suppress presynaptic assembly through the frizzled receptor LIN-17 or MIG-1 in DA8, DA9 and DB7 neurons (Klassen and Shen, 2007;Mizumoto and Shen, 2013).
Presynaptic CWN-2 and LIN-44 regulate activity-dependent postsynaptic localization of the acetylcholine receptor ACR-16 mediated by the LIN-17 and CAM-1 receptors in the postsynaptic muscle cells (Jensen et al., 2012;Pandey et al., 2017). Through b-catenin and TCF/LEF transcription factors, the canonical Wnt signaling regulates the abundance of the postsynaptic glutamate receptor GLR-1 (Dreier et al., 2005). Here, we found that a canonical Wnt signaling in the intestine promotes synaptic assembly in the C. elegans brain through the neuropeptide-like peptide NLP-40.
The role of Wnt in synaptic development is highly conserved in vertebrates (Budnik and Salinas, 2011;He et al., 2018;Park and Shen, 2012). For example, Wnt7a and Wnt5a play critical roles in pre-or postsynaptic development in mammals (He et al., 2018). Although the role of Wnt-neuropeptide in synaptic development was previously not reported, The Wnt/b-catenin has been shown to regulate the expression of gut hormone/neuropeptides (Garcia-Jimenez, 2010;Garcia-Martinez et al., 2009;Kim et al., 2015), suggesting that Wntneuropeptide interactions are evolutionally conserved. In addition, hormones/neuropeptides are also involved in synaptic assembly and function in mammals (Corvino et al., 2015;D'Ercole et al., 2002;Garcia et al., 2014;Wang et al., 2015). Collectively, these studies suggest that the function of Wntneuropeptide axis in synaptic assembly is most likely conserved in vertebrates.

An intrinsic role of gut in synaptic development
Using tissue-specific RNAi, tissue-specific rescue and transgenic GFP reporter experiments, we demonstrated that the Wnt-NLP-40 axis acts in the gut to regulate presynaptic assembly of nerve ring neurons. Further, we showed that the effect of the Wnt-NLP-40 axis on synaptic assembly at the newly hatched L1 stage, suggesting it acts during embryogenesis before feeding. We propose that the Wnt-NLP-40 axis acts through a stimulus-independent, intrinsic developmental program.  (Chikka et al., 2016). However, the intrinsic role of gut in the brain development is largely unexplored, and here we provide a molecular mechanism underlying this novel gut-brain interaction.
In vertebrates, the gut-brain-microbiota axis plays a vital role in mediating brain development and brain function, such as locomotion, exploratory and risktaking behaviors and spatial memory (Bercik et al., 2011;Diaz Heijtz et al., 2011;Gareau et al., 2011;Neufeld et al., 2011). Endocrine is a highly conserved signaling pathway mediated the gut-brain communication (Farzi et al., 2018;Matafome et al., 2017;Sun et al., 2018). Although it is unknown if gut-brain axis plays any intrinsic role in neural circuit formation in vertebrates, we posit that the signal pathway regulating synaptic formation is probably conserved given the conservation of gut-brain axis, Wnt signaling, neuroendocrine signaling.

Spontaneous neural activity regulates synaptic assembly
It is well known that neuroactivity regulates synaptic formation and plasticity (Hooks and Chen, 2020;Pan and Monje, 2020). The neuronal activity can regulate the trafficking of synaptic components (Lorenz-Guertin and Jacob, 2018; Sears and Broadie, 2017;Tiruchinapalli et al., 2003). Interestingly, here we found that Wnt-neuroendocrine signaling pathway is required for AIY spontaneously firing and synaptic vesicle trafficking, suggesting that the intrinsic gut signaling pathway most likely regulates synaptic formation through modulating neuronal activity.
Neuronal spontaneous activity is common during embryogenesis and plays critical roles in neuronal development including neurogenesis and circuit formation (Blankenship and Feller, 2010;Moody and Bosma, 2005). Based on our finding and studies from other groups, we speculate that the role of neuronal spontaneous activity in synaptogenesis is conserved.
Collectively, our findings not only provide mechanistic insights into the synaptic assembly, but also uncovered a novel gut-brain interaction.

Strains and maintenance
All C. elegans strains were grown with E. coli OP50 on standard NGM plates at Table   S1. Animals were used in this study at adult day 1 stage unless specified.

RT and qRT-PCR
qRT-PCR was performed as the previous study (Shen et al., 2012). In brief, for mRNA, total RNA was prepared by traditional Trizol extraction methods from ~100 worms. cDNA was subsequently generated by GoScript TM Reverse Transcription system for RT-qPCR (Bio-Rad). qRT-PCR was performed with 2xNovoStart®SYBR qPCR SuperMix Plus (Novoprotein) on a CFX384 Touch TM Real-Time PCR Detection System (Bio-Rad). Three technical replicates were performed in each reaction. nlp-40 was used as reference for mRNA quantification, the results were from at least three biological replicates. qRT-PCR primers for nlp-40 were designed and was listed in Table S2.

RNA interference
RNAi constructs were transformed into HT115. RNAi feeding experiments were performed based on the previous study (Fraser et al., 2000). For tissue-specific RNAi experiment, we used the worm strain FDU2851(sid-1(qt9); alxIs9[Pvha- laser. Individual animals were imaged for 60s at a rate of 2 Hz. Data were analyzed using custom written scripts in R Studio, and the images rotation and brightness/contrast were processed with Adobe photoshop CC. The 60s video was exported with Imaris 4.0.

Live image acquisition and Kymograph analysis
Animals at L3 stage were anesthetized on 3% agarose pads with 100 mM muscimol. All images were captured with Andor Dragonfly Spinning Disc Confocal Microscope with 60X objectives. The exposure time is 400ms and 800ms for wild-type and nlp-40 mutants respectively. Time-lapse movies were acquired continually with 1s interval for 5min. For quantification，the timelapse movies were processed by Fiji and KymographDirector2.1 (Fig. 7B, 7C).

Firstly, the movies were imported into Fiji to get the Kymograph images with
Kymograph clear 2.0. Each kymograph was processed with KymographDirector2.1 to generate the final data (GraphPad Software), and the intensity threshold was set based on the background.

Fluorescence microscope and confocal imaging
For general microscopy experiments, worms were anesthetized on 3% agarose pads with 50 mM muscimol. Images were captured with Andor Dragonfly Spinning Disc Confocal Microscope with 10X or 40X objectives, 488 nm (for GFP) or 561 nm (for mCherry) laser. The fluorescence intensity was quantified with Imaris 4.0, and the images rotation and brightness/contrast were processed with Adobe photoshop CC.

Quantification and statistical analysis
We quantified the fluorescence intensity using Imaris. The density obtained by cropping the whole AIY interneuron zone (showed in Fig. 1A dashed box) and the whole intestinal zone and NR zone (showed in Fig. 7A dashed box) in Imaris.
To monitor the temporal role of indicated genes, we quantified the F1 animals' phenotype when F0 day1 animals' eggs laid and were synchronized as 12, 20, 30, 42, 66 hr later, respectively. All quantified data were collected from at least three biological replicates. For data with transgenic animals, at least two independent lines were used. The statistical P values were measured using GraphPad Prism 6.0 (GraphPad Software). Statistical analysis between two groups were based on Student's t-test (two-tailed), among more than two groups were based on one-way ANOVA followed by Dunnett's test, as indicated in the Fig. legends. All quantitative data were collected blindly.

ACKNOWLEDGMENTS
We thank ZF Altun and DH Hall for help with schematic in Fig. 1A

COMPLIANCE WITH ETHICS GUIDELINES
The authors declare no competing interests. All institutional and national guidelines for the care and use of laboratory animals were followed.  Wang, M., Witvliet, D., Wu, M., Kang, L., and Shao, Z. (2021). Temperature regulates synaptic subcellular specificity mediated by inhibitory glutamate signaling. Plos Genet 17, e1009295. White, J.G., Southgate, E., Thomson, J.N., and Brenner, S. (1986). The Structure of the Nervous-System of the Nematode Caenorhabditis-Elegans. Philosophical Transactions of the Royal Society B-Biological Sciences 314, 1-340. Yang, Y., and Zhang, Z. (2020). Microglia and Wnt Pathways: Prospects for Inflammation in Alzheimer's Disease. Front Aging Neurosci 12, 110.   at least three biological replicates. The total number (N) of independent animals assayed is indicated in the bars. Statistical analyses are based on one-way ANOVA followed by Dunnett's test and two-tailed Student's t-test. ***p<0.001, ****p<0.0001, ns: not significant as compared to WT or control RNAi (p>0.05).
Error bars represent SEM.    Transgenic data are averaged from at least two independent lines. The total number (N) of independent animals assayed is indicated in the bars. Statistical analyses are based on one-way ANOVA followed by Dunnett's test and twotailed Student's t-test. *p<0.05, **p<0.01, ***p<0.001. Error bars represent SEM. Transgenic data are averaged from at least two independent lines. The total number (N) of independent animals assayed is indicated in the bars. Statistical analyses are based on one-way ANOVA followed by Dunnett's test. ns: not significant (p>0.05), **p<0.01, ***p<0.001, ****p<0.0001. Error bars represent SEM.
indicate the AIY soma. The scale bar in (F) applies to (F'), and the one in (G) applies to (G', H and H'). The length is 10μm. (I) Quantification of the fluorescent intensity of AIY RAB-3 for the indicated genotypes at the indicated stages. Data for each genotype are averaged from at least three biological replicates. The total number (N) of independent animals assayed is indicated in the bars. Statistical analyses are based on two-tailed Student's t-test. *p<0.05, ***p<0.001, ****p<0.0001. Error bars represent SEM.