Microbiota-derived acetylcholine can promote gut motility in Drosophila melanogaster

Abstract

The gut microbiota is crucial for intestinal health, including gastrointestinal (GI) motility. How commensal bacterial species influence GI motility has not been fully elucidated. A major factor of GI motility is the gut contraction promoting the propulsive movement of orally ingested materials. Here, we developed a method to monitor and quantify gut contractions in living Drosophila melanogaster larvae. We found that the culture medium of an isolated strain Lactiplantibacillus plantarum Lsi promoted gut contraction in vivo, which was not observed in Leuconostoc sp. Leui nor Acetobacter persici Ai culture medium. To identify bacteria-derived metabolites, we performed metabolome analysis of the culture media by liquid chromatography-tandem mass spectrometry (LC–MS/MS). Of the 66 metabolites detected, we found that some metabolites changed in a species-specific manner. Among them, acetylcholine was specifically produced by L. plantarum. Feeding exogenous acetylcholine increased the frequency of gut contractions, which was blocked by D-tubocurarine, an inhibitor of nicotinic acetylcholine receptors. In this study, we propose a mechanism by which the gut microbiota influences Drosophila gut motility.

This article is part of the theme issue ‘Sculpting the microbiome: how host factors determine and respond to microbial colonization’.

1. Introduction

Gastrointestinal (GI) motility is characterized as propulsive movement of food through the GI tract. Ingested food, as well as bacteria in the gut, is known to affect GI motility in animals. Microbes initiate or inhibit host signalling pathways, which can affect a wide range of physiological responses, including GI motility [1,2]. In mammals, varieties of nonpathogenic commensal bacteria, such as Lacticaseibacillus rhamnosus [3] and Lacticaseibacillus casei [4] have been found to affect GI motility [5]. Such host–microbe communication is primarily mediated by a direct response to bacterial metabolites and cell wall components, or by an indirect response to mucosal immune activation by bacterial immunogenic molecules [6]. However, our understanding of microbiota-dependent regulation of GI motility is far from complete.

Drosophila melanogaster is a useful model organism, with abundant genetic tools, to study the intricate interplay between the gut and microbiota at a molecular level. The Drosophila larval gut comprises a monolayer of epithelial cells, consisting of mainly nutrient absorptive enterocytes (ECs) and enteroendocrine cells (EECs) [7]. The gut epithelia is surrounded by visceral longitudinal and circular striated muscle that regulates gut motility [8]. The Drosophila microbiome is relatively simple, primarily consisting of bacteria from Lactobacillaceae and Acetobacteraceae families, which are aerotolerant and easy to culture [9]. It has been shown that an infection of opportunistic bacteria in adult Drosophila promotes gut contractions via a neuropeptide called diuretic hormone 31 (DH31) and enhances bacterial clearance [10]. However, no commensal bacteria have been reported to influence gut motility in Drosophila to date.

One regulatory mechanism behind gut motility is the secretion of neurotransmitters from EECs. For example, the secretion of serotonin from a subset of prevalent EECs called enterochromaffin cells is thought to modulate peristaltic reflexes in mice via the sensory nervous system [11]. Enterochromaffin cells were found to respond to short-chain fatty acids, a well-known bacterial metabolite, demonstrating a mechanism behind bacterial EEC activation [12]. Acetylcholine is another well-known neurotransmitter that also induces gut peristalsis in mammals by directly stimulating the visceral muscle expressing muscarinic acetylcholine receptor in the neuromuscular junction [13]. A recent report has even shown that gut dysmotility due to stress is mediated by decreases in colonic acetylcholine content and that pharmacological stimulation of nicotinic acetylcholine receptor (nAChR) can ameliorate the effects [14]. By contrast, Drosophila neuromuscular junctions are known to express ionotropic glutamate receptors, which mediate synaptic transmission [15]. However, involvement of acetylcholine in regulating Drosophila gut contraction has been suggested by LaJeunesse et al. where they noted that nAChRs were expressed in a subtype of EECs that regulate gut contractions via DH31 [16]. Whether microbiota-derived acetylcholine promotes gut contraction in the Drosophila midgut is yet to be explored. Here, we provide evidence that a commensal member Lactiplantibacillus plantarum produces acetylcholine, which can upregulate gut contractions in Drosophila larvae.

2. Material and methods

(a) . Drosophila stocks

Flies were raised on a yeast-based diet containing 4.5% cornmeal (Nippn Corporation), 6% brewer's yeast (Asahi Breweries, HB-P02), 6% glucose (Nihon Shokuhin Kako), and 0.8% agar (Ina Food Industry S-6) with 0.4% propionic acid (Wako 163-04726) and 0.15% butyl p-hydroxybenzoate (Wako 028-03685). Flies were maintained at 25°C.

The fly lines used in this study were Canton-S, DJ752-Gal4 (Bloomington Drosophila stock centre (BDSC) 8182), Dh31R-Gal4 [17], UAS-mCD8::GFP (BDSC 32186), nAChRα1-T2A-Gal4 [17], nAChRα2-LexA (BDSC 84413), nAChRα3-LexA (BDSC 84414), nAChRα5-LexA (BDSC 84415), nAChRα6-LexA (BDSC 84416), nAChRα7-Gal4 (BDSC 77828), nAChRβ2-LexA (BDSC 84417), nAChRβ3-Gal4 (BDSC 84418) and LexAop-GFP (BDSC 66687).

(b) . Bacterial stocks

Acetobacter persici Ai (A.p), L. plantarum Lsi (L.p) [18] and Leuconostoc sp. Leui (Leu) [19] were isolated previously from flies maintained in our laboratory. 16S rRNA sequence of Leu has 99% identity with L. mesenteroides or L. pseudomesenteroides. L. mesenteroides/pseudomesenteroides is generally not reported to be in laboratory stocks, suggesting it is uniquely present in some of our fly stocks [20].

(c) . Bacterial culture

Glycerol stocks of the isolated bacteria were preserved at −80°C. The stocks were cultured in approximately 3–5 ml of MRS broth (Oxoid, CM0359) overnight at 30°C. Bacterial proliferation was measured by checking its absorbance (OD600) using a DEN-600 Photometer (Funakoshi, BS-050109-AAK).

(d) . Quantification of metabolites in bacterial culture supernatant

To identify the metabolites produced by A.p, L.p and Leu, metabolome analysis of bacterial culture supernatant was performed using liquid chromatography-tandem mass spectrometry (LC–MS/MS) (LCMS–8060NX, Shimadzu). The glycerol stocks of the isolated bacteria were plated on an MRS agar plate for two days at 30°C. Colonies were picked and transferred to 3 ml of MRS broth, incubated at 30°C and cultured until OD600 = 0.5. This concentration was chosen because the OD600 is more reliable in lower concentration. The cultures were centrifuged at 25°C and 8000g for 5 min and the supernatant was collected. A portion of the culture supernatants (10 µl) were mixed with 150 µl of 80% methanol containing 10 µM internal standards (methionine sulfone and 2-morpholinoethanesulfonic acid), followed by deproteinization by mixing with 75 µl of acetonitrile. The sample was transferred into a 10 kDa centrifugal device (Pall, OD010C35). Using a centrifugal concentrator (TAITEC, VC-96R) and cold trap (TAITEC, VA-500R), the flowthrough of the sample was completely evaporated and then resolubilized in ultrapure water. The resolubilized samples were then injected into the LC–MS/MS with a PFPP column (Discovery HS F5 (2.1 mm × 150 mm, 3 µm), Sigma-Aldrich) in the column oven at 40°C. The samples were separated using a gradient from solvent A (0.1% formic acid, water) to solvent B (0.1% formic acid, acetonitrile) for 20 minutes. Primary metabolites package v.2 (Shimadzu) was modified and optimized for our system to measure the metabolites. Optimization of multiple reaction monitoring (MRM) parameters was performed by injecting the standard solution, then performing peak integration and parameter optimization with software (LabSolutions, Shimadzu). Quantification was performed using a standard curve drawn from serial dilutions (0.01–10 µM) of each metabolite.

(e) . Generation of germ-free Drosophila

To generate germ-free (GF) animals for the peristalsis experiment using the bleach method, Drosophila embryos were collected in a small cage by agar plates (2.3% agar, 1% sucrose, and 0.35% acetic acid) with live yeast paste for up to three hours. The embryos were then collected into a basket and sterilized by 500 ml of 70% ethanol twice. Then the embryos were dipped in a 3% sodium hypochlorite solution for 5 min, swiftly washed with 1 l of tap water, and transferred to a UV sterilized vial containing yeast-based diet. The vials were kept in a sterile plastic container. The bacterial contamination was checked by plating the food on MRS agar right before the peristalsis experiment.

(f) . Ex vivo peristalsis assay

To assess the effect of bacterial metabolites and acetylcholine on gut contraction, ex vivo peristalsis assay was performed. Briefly, third instar larvae reared in an axenic condition were fed with food supplemented with bacterial culture supernatant or acetylcholine for at least 5 min, and the gut was gently dissected to count the gut contraction manually. The detailed method was as follows.

Flies were reared in a cage for up to three hours for egg laying on an agar plate. The embryos were then incubated on the agar plate for 24 h at 25°C. Newly hatched first instar larvae were picked up by forceps and transferred to the standard yeast-based diet containing 6 ml of 10% nipagin (Wako, 132-02635), 100 mg of rifampicin (TCl, R0079), 25 mg of tetracycline (Wako, 203-08592) and 250 mg of ampicillin (Wako, 014-23302) per litre, referred to as the nRTA diet. To prevent overcrowding, 50–60 larvae were transferred to each vial.

For the inoculation of bacterial metabolites and acetylcholine, the melted antibiotic (nRTA) diet was mixed with 0.5% bromophenol blue (Wako, 029-02912) and then supplemented with 50 µl culture supernatant or 1 µl of 1 mM acetylcholine (TCl, A0084) solution per 1 ml of nRTA diet. For the control, uncultured MRS broth or Milli-Q water was mixed into the nRTA food with bromophenol blue. Bromophenol blue was added to identify the gut region by visualising the copper cell region where the dye turns yellow. L.p and Leu were cultured until the absorbance (OD600) was approximately 2.0. Each supplemented diet was allocated onto a small plate, and its surface was scratched prior to feeding to soften the food. The third instar larvae were floated using 30% glycerol at approximately 85 h after egg laying (AEL) and transferred to a drop of the respective diet. After feeding for at least 5 min, the larvae were dissected in Schneider's medium (Gibco, 21720024) at room temperature on a silicon dissection dish, and the gut contractions were counted manually. The dissection was performed carefully so as not to touch or stretch the anterior midgut or to separate the tracheae from the gut. Contractions were characterized as either a single or consecutive contraction of the gut. In terms of peristaltic waves, only those that transmitted from the anterior to posterior were counted. As a form of quality control, we removed larvae with less than five contractions per minute. Such larvae showed signs of irregularity, mainly immobility, which may indicate that ecdysis is nearing. Furthermore, the dissection process may have been technically imperfect, such as unintentionally stretching or touching the gut. This may have led to other irregularities such as arhythmic gut contractions, including a cluster of contractions occurring only once during the minute, or no movement for the first 30 s. These larvae were removed since such contractions may be deemed unusual in normal larval gut. The videos were edited using ImageJ (version 2.9.0/1.53t) and annotated using an additional package [21].

(g) . In vivo peristalsis assay

For less invasive and stable observation of gut contractions, we developed in vivo peristalsis assay. Briefly, second instar larvae reared in axenic conditions were fed with food supplemented with KCl, acetylcholine or bacterial culture supernatant for at least five minutes. To assess the involvement of nAChR, D-tubocurarine (nAChR inhibitor) was inoculated before acetylcholine treatment. Then the larvae were attached to double-sided tape to count the gut contraction manually without dissection. The detailed method was as follows.

Newly hatched first instar larvae were picked up by forceps and transferred to the nRTA diet as described in (f). Alternatively, bleached GF larvae were generated as mentioned in (e). At approximately 60 h AEL, second instar larvae were floated using 30% glycerol and transferred to a drop of 1 ml supplemented diet containing 0.5% bromophenol blue. Larvae at an earlier stage was used for in vivo experiment because accumulation of lipids in the fat body during the later stages decreased the transparency of the larvae, which made it harder to observe the contractions.

To assess the effect of compounds on peristalsis, the nRTA diet was mixed with KCl (Wako 166-22112), acetylcholine, or bacterial culture supernatant. The final concentration was 5 mM for KCl and 1 µM for acetylcholine. For the supplementation of bacterial culture supernatant, L.p and Leu were cultured until the absorbance (OD600) was approximately 2.0, and A.p was cultured until the absorbance (OD600) was 1.3. A.p was not able to proliferate until the absorbance reached 2.0 most likely due to the acetic acid that accumulates in the culture medium. In an attempt to match the amount of metabolites in the supernatant, 50 µl of L.p or Leu culture, or 75 µl of A.p culture was supplemented per 1 ml of nRTA diet. For treatment with the nAChR inhibitor D-tubocurarine, the larvae were fed an nRTA diet supplemented with 20 µM D-tubocurarine (Wako 201-13581) for 1 hour prior to acetylcholine treatment. For experiments on bleached GF flies, the supplemented diet was prepared as mentioned above without the antibiotics. The food was UV sterilized for at least 20 min before use. These diets were kept at 4°C for up to five days.

The larvae were fed with each diet for at least five minutes and gently picked up from the food, lightly placed on a moist tissue for approximately three seconds to remove excess food from the larval surface. Then the larvae were placed on an adhesive double-sided tape (Nitto J1410) attached to a petri dish with the ventral side up, and immediately observed under a stereo microscope to count the number of contractions manually. Larvae that showed no contraction for the first 30 s or exhibited less than five contractions per minute were excluded since such larvae showed an arrhythmic contraction pattern, such as a cluster of contractions occurring at once. The videos were edited using ImageJ (version 2.9.0/1.53t) and annotated using an additional package [21].

(h) . Immunostaining

To analyse the localization of nAChR-expressing cells and the expression of possible nAChR involved in peristalsis regulation, immunostaining was performed as described below. The third instar larvae were dissected in PBS, and the guts were fixed with 4% PFA for 30 min or an hour at room temperature. The samples were washed with PBST (0.1% Triton X-100) three times, blocked with 5% normal donkey serum (NDS) in PBST for at least one hour at room temperature and then incubated overnight at 4°C with anti-rat GFP antibody (NACALAI 04404-26) in PBST solution with 5% NDS. The samples were washed three times with PBST and incubated at room temperature for at least two hours with anti-rat Alexa Fluor 488 secondary antibody (Molecular Probe A-21208) and phalloidin-iFluor 647 (Abcam ab176759). The tissues were washed three times and mounted in SlowFade Gold antifade reagent (Invitrogen S36936). The tissues were observed using a Leica TCS SP8 confocal microscope. The images were analysed using ImageJ (version 2.9.0/1.53t).

(i) . Statistical analysis

Statistical analysis for metabolomics was performed using Metaboanalyst 5.0 [22]. The values for compounds that were not detected were replaced by LoDs, 1/5 of the minimum value of each compound. The values were then log transformed, and their Z scores were calculated using the auto scale option. Principal component analysis (PCA) was calculated and plotted. One-way ANOVA and Fisher's least significant difference (LSD) were performed to filter compounds using an ANOVA p value (FDR) cut-off of 0.01, and the compounds considered to exhibit significant differences were plotted onto a heatmap.

Statistical analysis was performed using GraphPad Prism 9 or 10. Biological replicates are represented by data points. Unpaired and two-sided Student's t tests were used to compare two groups. One-way ANOVA with Holm-Šídák's multiple comparison test was used to compare multiple groups. Trends across multiple experiments were consistent, but due to variability in experimental condition, only representative experiments are shown in each figure. Full data are available in the raw data file.

3. Results and discussion

(a) . Development of a method to count gut contraction in vivo

Several studies have shown that gut contractions in Drosophila can be monitored ex vivo using a dissected gut in a medium [16,23]. We identified gut regions by feeding the larvae food containing bromophenol blue dye, which turns yellow in acidic regions, visualising the copper cell region [24]. To quantify the gut contractions in the larval gut, we first attempted to expose the gut by making an incision on the larval body surface using forceps in Schneider's insect medium. However, the observed gut contractions were weak, and the number of contractions was highly variable among individuals. The gut contractions increasingly weakened or strengthened during the observation, and in most of the former cases, the contractions eventually weakened to the point of being unrecognizable within 2–5 min. Furthermore, we observed that in the anterior midgut, not all contractions moved smoothly from the anterior to posterior side of the gut. In most cases, many gut contractions seemed to occur in specific regions and skip the rest (electronic supplementary material, movie S1). These unnatural features of gut contractions may be attributed to our technical immatureness or variability between samples since this method is highly invasive and risks damaging tissues involved in regulating peristalsis, such as the visceral muscle or neurons. It is also conceivable that the medium composition is inappropriate to mimic the in vivo situation.

Therefore, taking advantage of the transparent nature of Drosophila larvae, we developed a method to count gut contractions without dissection (figure 1a). To do this, larvae 60 h after egg laying (AEL) were fed a specific diet for at least 5 min and then placed on adhesive tape to fix the animal (figure 1a). Using this method, we observed more pronounced and frequent gut contractions compared to the ex vivo method. One caveat of this method is that a part of the anterior midgut is hidden by the fat body or other tissues and cannot be observed. To quantify gut contractions, we focused on peristalsis from the region between the proventriculus and the first fold of the gut (figure 1b), which was visible, definable, and actively contracting. We were able to characterize three main types of gut contractions in the area (electronic supplementary material, movie S2). The first type was a peristaltic wave that glided from the anterior to posterior direction of the gut (anterograde peristalsis). The second type was a similar peristaltic movement that travelled in the opposite direction, i.e. from the posterior to the anterior (retrograde peristalsis). In most cases, these two types of peristaltic movement occurred consecutively. The third one was a short contraction that occurred between the first and second type of contractions (electronic supplementary material, movie S2). Retrograde peristalsis is observed in humans [25] as well as in fish [26], suggesting that it is a part of the widely conserved physiological process of gut motility.

Figure 1.

A novel method for counting larval gut contractions. (a) Experimental scheme of the novel gut larval contraction assay. At 60 h AEL, larvae were placed on food containing 1% bromophenol blue dye for at least 5 min. Larvae were gently picked up by forceps and then placed on black double-sided tape to observe intestinal contractions. (b) A representative image of a larva fed food containing 1% bromophenol blue and immobilized on double-sided tape. The yellow arrows represent the overall movement of the food from the mouth (termed mouth hook) through the anterior midgut. The final yellow arrow represents the food travelling towards the dorsal side. The orange square represents the area referred to as the anterior region, which was observed to count gut contractions. Scale bar = 500 µm. (c) Representative snapshots of one anterograde peristaltic contraction taken every 175 ms. Yellow arrowheads represent propulsive movement of the food. Scale bar = 500 µm. (d) Gut contraction count of larvae fed 5 mM KCl using in vivo assay. n = 10. Mean and data points indicating biological replicates are shown. The p value was determined by unpaired two-tailed Student's t test.

It is noteworthy that, even using this less invasive method, the frequency and the pattern of the contractions differed greatly between larvae. For simplicity, we quantified anterograde gut contractions that originated from the anterior region (figure 1b,c). The average number of gut contractions was approximately 12 times per minute in the wild-type larvae fed with Milli-Q water supplemented control food (figure 1d). To test whether the larval gut can respond to a contraction stimulant in this model, we fed KCl, which induces muscle contraction by Ca²⁺ influx and myosin light chain phosphorylation [27]. KCl ingestion indeed increased the contraction count (figure 1d). Together, we developed a less invasive method to monitor gut contractions in vivo, which is applicable to test the effect of bacterial metabolites.

(b) . L. plantarum culture medium specifically increased gut contractions

To test whether bacterial culture supernatant can alter gut contractions, we quantified the anterograde gut contractions of larvae that were fed fly food supplemented with bacterial culture supernatant. We tested three isolated strains, Acetobacter persici Ai (A.p) [18], Lactiplantibacillus plantarum Lsi (L.p) [18] and Leuconostoc sp. Leui (Leu) that has 99% identity with L. mesenteroides/pseudomesenteroides [19]. This bacterium is identified specifically in our lab. We found that the gut contractions in the anterior region tended to increase when larvae were fed with L.p culture supernatant but the effect was not as pronounced as the larvae fed with Leu culture supernatant, which is also in the Lactobacillaceae family (figure 2a). Furthermore, the A.p culture supernatant did not increase gut contractions (figure 2b).

Figure 2.

L.p specifically promotes gut contraction in the anterior midgut. (a, b) Gut contraction count of larvae fed food supplemented with L.p, Leu (a), or A.p (b) culture supernatant using the in vivo assay. L.p, Lactiplantibacillus plantarum Lsi; Leu, Leuconostoc sp. Leui; A.p, Acetobacter persici Ai. (c) A representative image of the larval gut with bromophenol blue-supplemented food. The yellow circle shows the junction region observed for contractions using the contraction assay requiring dissection. Scale bar = 500 µm. (d) Gut contraction count of larvae fed food supplemented with L.p culture supernatant using ex vivo assay. For all graphs, the mean and data points indicating biological replicates are shown. p values were determined by one-way ANOVA with Holm-Šídák's multiple comparison test (a) or unpaired two-tailed Student's t test (b, d).

Since we found an increase in gut contractions in the anterior region upon feeding with L.p supernatant using the in vivo method (figure 2a), we validated whether this could be reproduced ex vivo. Gut contractions in the anterior region in the cultured condition were not frequent in our experimental setting, but we noticed that the number of contractions was constant in the junction region where the gut is folded right before the acidic zone (figure 2c). The contraction of this area was almost synchronous with that at the beginning of the acidic zone (electronic supplementary material, movie S1). We found that the basal number of contractions observed was lower in the ex vivo assay compared to the in vivo assay.

This difference may be attributed to the various parameters of the experiments. First, exposing the dissected gut may have weakened the contraction profile due to factors such as our dissection techniques or inappropriate media as stated above. Alternatively, we used younger larvae in the in vivo assay (60 h AEL) than used in the ex vivo assay (84 h AEL), because at the later stages of development, lipid accumulates in the fat body which hides sections of the gut. This discrepancy in larval stages may have resulted in different gut contraction caused by various factors including food intake behaviour which is known to change after reaching critical weight [28]. Despite this basal difference observed between the in vivo and ex vivo assay, feeding L.p culture supernatant slightly increased the number of gut contractions in the ex vivo assay as well (figure 2d).

(c) . Acetylcholine is present in L.p culture supernatant

The specific effect of L.p suggested that this bacterial species specifically produces some metabolites that stimulate gut contractions. To identify such metabolites, we performed a metabolome analysis on the culture supernatants of the three bacterial strains (figure 3a–c). As expected, the two lactic acid bacterial strains produced high amounts of lactate (figure 3c,d), validating this analytical method. In addition, L.p and Leu, but not A.p, culture supernatant contained lower amounts of purine metabolites, such as adenosine (figure 3c,e), guanosine (figure 3c), and inosine (figure 3c), which is consistent with our previous observation in the bacteria-conditioned diet [29].

Figure 3.

Metabolome analysis of bacterial culture supernatant. (a) Experimental workflow. A colony from an MRS plate was used to culture bacteria in MRS broth until OD600 = 0.5. The culture was centrifuged, and its supernatant was collected and analysed using LC–MS/MS. The sample was analysed for 109 compounds, out of which 66 were detected. One-way ANOVA with LSD was performed, and 21 compounds were selected by ANOVA p value (FDR) < 0.01 cut off. L.p, Lactiplantibacillus plantarum Lsi; Leu, Leuconostoc sp. Leui; A.p, Acetobacter persici Ai. (b) Principal component analysis of the 66 compounds detected. (c) Heatmap of the Z score of 21 compounds selected by p < 0.01. (d, e) Quantification of lactate (d) and adenosine (e) by LC–MS/MS in the larval whole body. Adenosine level below 0.01 µM was defined as ‘not determined’. For all graphs, the mean and data points indicating biological replicates are shown. p values were determined by one-way ANOVA and LSD using Metaboanalyst 5.0 [22](d, e).

We noticed that a mean of 2.6 µM acetylcholine was detected in the L.p culture supernatant but the amount detected in A.p or Leu culture supernatant was below determination cut-off of 0.01 µM (figure 3c asterisk, 4a). This is consistent with previous reports from as early as 1947 which show that L. plantarum produces acetylcholine [30]. Acetylcholine production is affected by various culture conditions, such as temperature and composition of culture medium [30], suggesting that acetylcholine production may not be necessary for bacterial proliferation. Acetylcholine synthesis has been evolutionarily conserved in fungi, plants and some bacterial species, including Bacillus subtilis, Escherichia coli and Staphylococcus aureus, although to a lesser extent compared to L.p [31]. In bacteria, the exact role of acetylcholine as well as its biosynthesis pathway is unknown, but bacterial sensing of acetylcholine has been reported, suggesting the possibility of involvement in interorganismal communication [32].

Figure 4.

L.p-derived acetylcholine promotes gut contractions in the anterior midgut. (a) Quantification of acetylcholine (ACh) in bacterial culture supernatant. Acetylcholine level below 0.01 µM was defined as ‘not determined’. L.p, Lactiplantibacillus plantarum Lsi; Leu, Leuconostoc sp. Leui; A.p, Acetobacter persici Ai. (b, c) Contraction counts of larval guts that ingested food supplemented with 1 µM acetylcholine using the ex vivo assay (b) or in vivo assay (c). (d) Contraction count of larval guts using the in vivo assay that were exposed to 20 µM D-tubocurarine prior to 1 µM acetylcholine feeding. For all graphs, the mean and data points indicating biological replicates are shown. p values were determined by unpaired two-tailed Student's t test (a–d).

(d) . Gut contraction is increased in larvae fed acetylcholine

Acetylcholine is well known as a primary neurotransmitter in the central nervous system, but reports have also shown that peripheral acetylcholine plays an important role including in regulation of immunity [33] and gut barrier homeostasis [34]. Since acetylcholine promotes gut motility in mammals [35] and potentially in Drosophila [16], we hypothesized that acetylcholine mediates the effect of L.p culture supernatant on GI motility. One of the experiments to test this possibility is to observe the effect of bacterial culture supernatant from the L.p mutant, which cannot produce acetylcholine. However, it is impossible to generate the mutant since the biosynthetic pathway for acetylcholine production is not known. Therefore, we tested whether orally administered exogenous acetylcholine can promote gut contractions. In the dissected gut, we found that feeding on a 1 µM acetylcholine supplemented diet increased the number of contractions ex vivo and in vivo (figure 4b,c).

Supplementation of acetylcholine or L.p supernatant both showed roughly the same increase in gut contractions. We estimated that acetylcholine concentration in the food may be lower when L.p supernatant was administered than when acetylcholine reagent was used (final acetylcholine concentration is 1 µM in acetylcholine supplemented food and 0.2–1.0 µM in L.p supernatant supplemented food). Although we did not test the sufficiency of enhancing gut contractions by lower concentrations of acetylcholine, an additive or synergistic effect between acetylcholine and other factors in the bacterial culture supernatant is likely to exist. In mammals, bacterial components such as LPS are thought to directly induce gut contractions. Immune response elicited by bacteria can also indirectly affect gut peristalsis, which suggests the involvement of cell wall components in regulating gut contractions [6]. Taken together, the bacterial presence will most likely affect the gut movement via various mechanisms including acetylcholine. Untangling the intricate mechanisms behind the bacterial effect on gut contraction by diverse factors would be of great interest.

Given that acetylcholine produced by L. plantarum promotes gut motility, environmental factors affecting the acetylcholine production by L. plantarum can indirectly influence the bacterial effect on the host. For instance, temperature and the medium composition (the existence of glucose and choline) modify the acetylcholine production of L. plantarum while it is not critical for the bacterial growth per se [30]. Further analysis is necessary to elucidate the regulatory mechanisms of acetylcholine production in L. plantarum and how it influences the gut motility of the host.

To assess the involvement of nicotinic acetylcholine receptor (nAChR) in gut contraction, we fed larvae D-tubocurarine, an nAChR antagonist. When larvae were treated with D-tubocurarine, the increase in gut contraction by acetylcholine feeding was attenuated (figure 4d). Taken together, these results indicate that acetylcholine derived from L.p can promote gut contractions.

(e) . Acetylcholine-derived gut contraction promotion occurs without exposure to antibiotics

Both the in vivo and ex vivo assay were performed using larvae that were fed with an antibiotics cocktail to remove the effects of other bacteria or their metabolites that may affect gut contractions. To rule out the effect of antibiotics usage, we performed the in vivo assay using a GF larvae produced by bleaching the embryo, instead of larvae fed with antibiotics. We noticed that GF larvae showed higher average contraction counts when fed with bacterial culture vehicle (MRS medium): approximately 19 (GF, electronic supplementary material, S1a) versus 17 (antibiotics, figure 2a). This result suggests that there is a basal difference between the two conditions at least when MRS is added to the diet.

Feeding the GF larvae with L.p supernatant did not show as pronounced an effect as observed in the antibiotics-fed larvae (electronic supplementary material, figure S1a). However, when the GF larvae were fed with acetylcholine-supplemented food, their gut contractions increased (electronic supplementary material, figure S1b). This result suggests that (i) antibiotic exposure is necessary to observe gut contraction increase by L.p, but not by acetylcholine or (ii) bleaching the embryo to produce GF larvae may have impaired or altered the larval function, masking the increase in gut contractions caused by L.p supernatant, but not by acetylcholine. The latter seems to be the most likely when considering the basal difference as mentioned above. It is possible that the gut contraction reached its near maximum at around the control group fed with MRS medium (electronic supplementary material, S1a). Another possibility is that since the acetylcholine concentration was higher in acetylcholine-treated food (1 µM) than in L.p supernatant-treated food (approximately 0.05–0.25 µM), the negative effect of embryo bleaching might have been offset in this assay. In addition, we observed that the GF larvae showed higher variability in size than antibiotics-treated animals, suggesting that bleaching the embryo may have caused growth impairment of the larvae. Therefore, there is a high possibility that the damage of bleach, by affecting the gut physiology and larval developmental stages, may be masking the increase in gut contractions by L.p supernatant. To fully negate the impact of bleaching as well as antibiotic exposure, an authentic GF animal should be created, perhaps using an isolater.

(f) . icotinic acetylcholine receptor is expressed in both the junction region and anterior midgut

Since the effects of acetylcholine on the gut motility of the anterior region were mediated by nAChR, we attempted to identify the exact nAChR subunits involved in gut contraction among 10 known Drosophila nAChR subunits [36]. Since most nAChRs are highly expressed in the anterior region to the junction region, enteric neurons innervating the proventriculus or anterior midgut may also play a part in mediating the increase in gut peristalsis by L. plantarum-derived acetylcholine. From immunostaining of nAChR subunits, we found that nAChRα1, α2, α3, α5, α6, β2 and β3 were expressed in neurons, and nAChRα2, α3, α5 and α6 were also expressed in enterocytes in the anterior midgut (electronic supplementary material, figure S2). Although the mechanism by which enterocytes regulate gut peristalsis is unknown, neurons are known to regulate gut motility. Serotonergic enteric neurons have been found to innervate the larval anterior midgut and regulate feeding behaviour [37,38], but the role of cholinergic neurons has not been explored. In adult Drosophila, some neurons are known to project through the visceral muscle, innervating the basal side of the epithelial cell [39]. Proximity of the neuron to the gut lumen may allow exogenous acetylcholine to directly stimulate the neurons. The role of nAChR-positive neurons in gut contraction still needs to be tested to fully understand the mechanisms of gut contractions in the anterior midgut. We also noted that some nAChR subunits are expressed in the Malpighian tubules, the most striking being nAChRα5 (electronic supplementary material, figure S2).

We could not find any notable expression in the junction region using reporter flies (electronic supplementary material, figure S2), but there is a report of nAChR expression in a couple of EECs located in the junction region using α-bungarotoxin, which is known to bind to nAChR [16]. It is possible that the reporter lines we used to visualize nAChR expression were not enough to detect low levels of expression in EECs. LaJeunesse et al. identifies the EECs using a DJ752 Gal4 driver. They found that the cells coexpress DH31, and that its excitation is sufficient to induce gut contractions in Drosophila larvae [16]. LaJeunesse et al. used what they described as an ‘enigmatic muscle tether’, to identify the region. This muscle tether was described to be connected to the gastric caeca, which is located around the anterior region [16]. We confirmed that both the DJ752 driver-positive cells and the muscle tether were located around the junction region where gut contraction was observed (electronic supplementary material, figure S3a, b).

We observed that this muscle tether as well as the tip of the gastric caeca also expressed DH31 receptor (DH31R). Whether the muscle tether works as a link to communicate the movement of the junction region to a more anterior region of the gut needs to be explored. Additionally, we observed that DH31R was expressed in the circular visceral muscle of the anterior midgut, a region anterior to the junction region (electronic supplementary material, figure S3c). The circular visceral muscle expressing DH31R contracts in some larvae (electronic supplementary material, movie S3). DH31R was further expressed in enterocytes at the most anterior midgut, muscle at the gastric caeca, and neurons at the proventriculus (electronic supplementary material, figure S3c). Although the function of DH31R in these regions is unknown, DH31 secretion and its reception in these regions might also be related to the regulation of gut contraction. Taken together, we formed a hypothesis that L. plantarum-derived acetylcholine stimulates EECs to secrete DH31, which is received by DH31R expressed in various regions of the gut, directly inducing gut contraction. This hypothesis would be of interest to address in future studies.

4. Conclusion

In this paper, we established a novel and accessible method to observe and quantify gut contractions less invasively in vivo using Drosophila larvae. We further revealed that acetylcholine or acetylcholine-producing bacteria such as L. plantarum can increase gut contractions in Drosophila larvae. Our present study shows the possibility that acetylcholine derived from commensal bacteria may affect host physiology via gut motility promotion.

5. Limitations

Our observations revealed that at least in the case of Drosophila, intestinal peristalsis included different types of peristaltic movements. Of the three main types of gut contractions, only two were counted in this study. Furthermore, it is still unclear whether the bacterial metabolites altered the initiation or intensity of the gut contraction since both cases could lead to a higher number of gut contractions being observed. Quality as well as quantity needs to be addressed to gain a comprehensive understanding of peristaltic movement. In addition, to assess whether acetylcholine derived from L.p is sufficient for gut contraction, bacterial gene manipulation is further needed.

Ethics

This work did not require ethical approval from a human subject or animal welfare committee.

Data accessibility

Supporting data were uploaded as the electronic supplementary material. The remaining raw data are available upon reasonable request to the corresponding author.

Supplementary material is available online [40].

Declaration of AI use

We have not used AI-assisted technologies in creating this article.

Authors' contributions

Y.F.: formal analysis, investigation, methodology, visualization, writing—original draft, writing—review and editing; H.K.: conceptualization, investigation, methodology, supervision, writing—review and editing; F.O.: conceptualization, funding acquisition, supervision, writing—review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We have no competing interests.

Funding

This work was supported by AMED-PRIME to F.O. (grant no. JP20gm6310011), by grants from the Japan Society for the Promotion of Science to F.O. (grant no. 22H02769), and by grants from Yakult Bio-Science Foundation to F.O.