Selective modulation of murine intestinal M₁ and M₃ muscarinic receptor expression has divergent effects on specialized epithelial cells and body weight

Abstract

Background:

M₁ and M₃ muscarinic receptors encoded by CHRM1 and CHRM3 mediate neuronal and non-neuronal cholinergic signaling. Mice with global M₃R deficiency reportedly weigh less than controls, but the cell type(s) involved are unknown. As the intestinal epithelium modulates nutrient absorption, we asked if deleting M₁R and M₃R only from intestinal epithelial cells would alter the distribution of specialized small intestinal epithelial cells or body weight.

Methods:

We reviewed reports of global M₁R and M₃R deficiency and body weight, used single cell RNA sequencing (scRNA-Seq) to assess Chrm1 and Chrm3 expression by small intestinal epithelial cells, created mice with conditional intestinal epithelial cell M₁R and M₃R deletion (CKO mice), and compared the distribution of specialized intestinal epithelial cells and body weights of CKO and control mice, and the development of enteroids.

Results:

Prior weight comparisons commonly used only male mice, frequently without comparison to littermate controls. scRNA-Seq analysis of tissues from M₁R and M₃R floxed mice revealed robust Chrm1 and Chrm3 expression by enteric goblet cells. CKO mice with selective mucosal depletion of Chrm1 and Chrm3 RNA were viable, fertile, and had fewer small intestinal goblet cells than controls. M₃R CKO mice had more tuft cells than controls. Although female mice weighed ~20% less than males, we detected no weight differences between M₁R and M₃R CKO and control mice; enteroids derived from these mice developed at the same pace.

Conclusions:

Intestinal epithelial cell M₁R and M₃R deficiency impacts the distribution of specialized intestinal epithelial cells but not murine body weight.

Graphical Abstract

New and Noteworthy:

Small intestinal goblet cells robustly express both Chrm1 and Chrm3; mucosal immune cells express primarily Chrm3. Mice with intestinal epithelial cell (IEC) Chrm3/M₃R deletion have fewer small intestinal goblet cells but more tuft cells. The increase in tuft cells, which produce acetylcholine, may represent a feedback mechanism to compensate for reduced muscarinic receptor (MR) expression. IEC-selective deletion of MRs does not impact murine weight or enteroid development, suggesting MRs beyond the intestinal epithelium regulate weight.

Introduction

Weight regulation is a major focus of scientific inquiry. Prompted by a global obesity epidemic (1), there is great interest on the part of the scientific community, the pharmaceutical industry, and the general population to develop drugs as effective as bariatric surgery at promoting weight loss (2). In this regard, recent advances in the use of glucagon-like peptide-1 (GLP-1) analogues are a notable success (3, 4). Moreover, augmented weight loss when GLP-1 analogues are combined with glucose-dependent insulinotropic polypeptide (GIP) analogues and other agents, suggests that combinations of such agents may be therapeutically advantageous (5, 6); the authors of one study speculated that ‘there may be additive benefit in targeting multiple endogenous nutrient-stimulated hormone pathways that have been implicated in energy homeostasis’ (5). In this context, we wondered whether intestinal epithelial cholinergic muscarinic receptors could provide novel therapeutic targets to modulate body weight.

Muscarinic receptors, comprised of five subtypes designated M₁ - M₅ and encoded by CHRM1 – CHRM5, respectively, are expressed by most, if not all, mammalian organs and cells (7). Within the central and enteric nervous systems, M₁ and M₃ muscarinic receptors (M₁R and M₃R) are particularly important targets of brain and gut cholinergic neurotransmission which is largely mediated by acetylcholine. In addition to its neuronal origins, acetylcholine is also produced and released by normal and neoplastic non-neuronal cells in the gut (8, 9). M₃R expressed by intestinal epithelial cells play a particularly important role in regulating cell proliferation and mucosal homeostasis (8). The role of M₁R expression in regulating these functions is less certain and there is limited information regarding the relative distribution of M₁R and M₃R expression in intestinal epithelial cells.

Mice with global deficiency of selected muscarinic receptor subtypes have proven very useful for elucidating the biological actions of these receptors (7). For example, using global knockout mice, M₃R was identified as a key player in intestinal bacteria and worm expulsion (10, 11) and the progression of colon neoplasia (12, 13). Whereas they can provide preliminary insights into the role of muscarinic receptor subtypes in modulating normal and pathological organ function, the use of global knockout mice cannot precisely identify the relevant cell types. An intriguing example is the role of muscarinic receptor subtypes in regulating food intake and body weight. Notably, when mice with global deficiency of M₃R were initially created, they were described as ‘hypophagic and lean,’ and expressed drastically lower levels of orexigenic hormones (e.g., melanin-concentrating hormone, insulin) and anorexigenic hormones (e.g., leptin), suggesting a role for M₃R in the regulation of weight and appetite (14). Indeed, in our own studies using this mouse strain, we found that 6- to 20-week-old male mice with global M₃R deficiency and unlimited access to standard rodent chow weighed approximately 25% less than control mice (12, 15). Although less is known about the role that M₁R plays in weight regulation, we found M₁R-deficient male mice were also leaner than control mice (15).

As the intestinal epithelium plays a dynamic role in digestion and nutrient absorption, we wondered if M₁R and M₃R expression in gut epithelial cells plays a role in regulating murine body weight. Although the intestinal epithelium is comprised primarily of absorptive and secretory enterocytes, more than 30 other epithelial cell types can directly and indirectly impact food intake and nutrient absorption (16). Fatty acid oxidation by enterocytes can alter glucose homeostasis and influence eating behavior, whereas specialized cells such as enteroendocrine cells, which constitute ~1% of intestinal epithelial cells, can produce regulatory peptides (e.g., ghrelin, GIP, and GLP-1) that modulate eating behavior and weight regulation (17–19). A recent study used villin-Cre to create mice with conditional epithelial ablation of M₃R, and although baseline weight data were not reported, investigators found that M₃R-ablated male mice, but not female mice, were more susceptible to dextran sodium sulfate (DSS)-induced colitis and weight loss compared to controls (20). Interestingly, the lack of intestinal injury and weight loss in female mice was related to the estrogen-mediated upregulation of colonic epithelial stem cells, which are known to co-express M₁R.

The relative distribution of M₁R and M₃R amongst small intestinal epithelial cells is uncertain, as is their role in mediating the weight differential noted between mice with global M₁R and M₃R deficiency and control mice. As described in the following work, we addressed these gaps in knowledge in several ways. First, we reviewed the published literature regarding reported effects of global M₁R and M₃R deletion on murine body weight. Then, given the critical role of intestinal epithelial cells in body weight homeostasis, we employed single cell RNA sequencing (scRNA-Seq) to compare Chrm1 and Chrm3 RNA expression in murine small intestinal epithelial cells. Next, we created female and male mice with conditional intestinal epithelial cell deficiency of M₁R and M₃R (designated here as M₁R and M₃R CKO mice) and assessed the impact of muscarinic receptor subtype deletion on phenotypic characteristics including murine body weight, the distribution of specialized intestinal epithelial cells, and the development and maturation of enteroids derived from the small intestines of M₁R and M₃R CKO mice compared to enteroids derived from sex-matched littermate control mice.

Methods

Chemicals and reagents.

If not otherwise specified, chemicals and reagents were obtained from Sigma Aldrich, St. Louis, MO, USA.

Literature search strategy.

We searched the existing literature for publications that reported or referred to weight data of mice with global knockout of either the M₁R or M₃R muscarinic receptors. Our search strategy is outlined in Supplemental Figures 1 and 2. We searched PubMed from January 1, 2001 through November 8, 2025 using the keywords ["M1R" AND ("mouse" OR "mice" OR "murine" OR "rodent")] and ["M3R" AND ("mouse" OR "mice" OR "murine" OR "rodent")]. Studies were excluded if they were not in English, or if the study did not involve muscarinic signaling, (e.g., M1R can also refer to an antigen of the Mpox virus). Studies were further screened, and those that did not study live M₁R or M₃R knockout mice were excluded. References of included studies were reviewed, and those relevant to this work were considered for inclusion; the references of those publications were reviewed iteratively to identify additional studies. We also searched PubMed and laboratory websites for publications from laboratories that developed mice with floxed Chrm1 and Chrm3 [i.e., the laboratories of Susumu Tonegawa (Massachusetts Institute of Technology), Jürgen Wess (NIDDK), J. Josh Lawrence (Texas Tech University), Jonah Chan (University of California, San Francisco), Neil M. Nathanson (University of Washington, Seattle), and Minoru Matsui (University of Tokyo)].Finally, studies were excluded if they did not weigh mice (e.g., if knockout mice were euthanized and experiments performed on ex vivo tissues, or if mice were only weighed after they had been exposed to a potentially confounding variable, such as treatment with a drug that might differentially affect knockout and wild-type mice). If data were provided as graphs or alluded to but did not include means ± error, we contacted the corresponding author to request additional details.

Animals.

All animal experiments were approved by the Office of Animal Welfare Assurance at the University of Maryland School of Medicine and by the Research and Development Committee at the VA Maryland Health Care System. We used the villin-Cre method to create C57BL/6 mice with intestinal epithelial cell-selective deletion of M₁R and M₃R. We obtained mice with floxed M₁R from the laboratory of Susumu Tonegawa at the Massachusetts Institute of Technology (Cambridge, MA). To achieve mice with intestinal epithelial cell deletion of M₁R, we crossed M₁R floxed mice with transgenic mice expressing Cre-recombinase under control of the Villin1 promoter [B6.Cg-Tg(Vil1-cre)1000 Gum/J (stock number 021504, Jackson Labs)] (4). Then we intercrossed the resulting heterozygous Villin-Cre CKO mice to obtain homozygous Chrm1/ heterozygous Villin-Cre CKO and littermate control mice. We obtained mice with floxed M₃R from the laboratory of Jurgen Wess at the National Institutes of Health (Bethesda, MD). To achieve mice with intestinal epithelial cell deletion of M₃R, we crossed M₃R floxed mice with transgenic mice expressing Cre-recombinase under control of the Villin1 promoter [B6.Cg-Tg(Vil1-cre)1000 Gum/J (stock number 021504, Jackson Labs)] (4). Then we intercrossed the resulting heterozygous Villin-Cre CKO mice to obtain homozygous Chrm3 /heterozygous Villin-Cre CKO and littermate control mice. The C57BL/6 M₃R floxed mice were derived from the N substrain and C57BL/6 M₁R floxed mice were derived from the J substrain. Villin-Cre mice were obtained from Jackson Labs (J substrain), so M₃R CKO mice used in the present work were on a mixed J/N substrain and the M₁R CKO mice were on a J/J substrain. Importantly, all comparisons were made to respective M₁R or M₃R littermate control mice.

To achieve microbiome equivalence, we cohoused CKO and littermate control mice in a specific pathogen-free facility and, unless stated otherwise, male and female mice were used between five and 30 weeks of age. To avoid confounders that could alter basal metabolic rate and energy expenditure (21), CKO and littermate control mice were maintained at the same ambient temperature and light/dark phases and fed the same standard rodent chow.

Organoid and cell culture.

Small intestinal organoids (enteroids) were generated from one male CKO and one male littermate control mice per genotype utilizing the StemCell Technologies (Vancouver, BC, Canada) protocol according to manufacturer’s instructions. Briefly, immediately after euthanasia, the murine small intestine was isolated, flushed with cold PBS, opened longitudinally, cut into ~1 mm² fragments to isolate crypts, and washed vigorously 25 times in PBS + 0.1% BSA. After washing, small intestinal pieces were resuspended in Gentle Cell Dissociation Reagent (StemCell; catalog number 100-1077) and incubated for 15 min on a rocking platform at 20 rpm. The supernatant was removed, and the pieces washed four times with cold PBS containing 1% BSA. After each washing step, the supernatant was filtered through a 70-μm nylon mesh. The fourth and purest filtered sample was centrifuged, washed, and used to seed the initial enteroids utilizing Matrigel (Corning; catalog number 356237) and IntestiCult Organoid Growth Medium (StemCell; catalog number 06005). Enteroids were maintained in a 37°C and 5% CO₂ incubator, media was changed three times per week, and enteroids were passaged weekly. Experiments were performed in triplicate. Enteroid cultures were imaged every two days utilizing a Nikon Eclipse Ti microscope (Melville, NY, USA). Enteroids were counted and maturity was assessed utilizing ImageJ software by an investigator masked to mouse genotype. Organoids were considered mature if they contained three or more buds.

Tissue preparation for scRNA-Seq.

To create a single cell suspension from small intestinal tissue of one M₁R and one M₃R male littermate control mice, the proximal third of the small intestine was harvested and gently irrigated in 5 ml of cold DPBS (Corning; catalog number 21-040-CV) containing 1% BSA. The tissue was cut longitudinally and sliced into small fragments measuring 2 mm. These tissue fragments were incubated in 10 ml of ice-cold 20 mM RNase-free EDTA-DPBS (Thermo Fisher Scientific; catalog number AM9260G) for 60 min. During incubation, the tissue fragments were mixed every 15 min by pipetting through wide-bore tips. After the tissues fragments were incubated in EDTA-DPBS, the villus-rich supernatant was collected and filtered through a 30 μm MACS strainer (Miltenyi Biotec; catalog number 130-098-458). The filtrate was then centrifuged at 300 g for 5 min at 4°C in a swing bucket centrifuge, the supernatant was decanted, and the pellet washed twice in 10 ml DPBS containing 1% BSA. After the last wash, the pellet was set aside on ice. Concurrently, the tissue fragments were incubated in 5 ml TrypLE (Life Technologies; catalog number 12605) for 10 min at 37°C. After the TrypLE incubation, the tissue fragments were resuspended by pipetting through wide-bore tips; the crypt-rich supernatant was retrieved and filtered through a 30 μm MACS strainer. The sample was centrifuged at 300 g for 5 min at 4°C in a swinging bucket centrifuge, and the pellet washed twice in 10 ml DPBS containing 1% BSA. After the last wash, villus-rich and crypt-rich pellets were combined in 5 ml cold 0.04% BSA-PBS. Cell viability was immediately assessed using the Countess Automated Cell Counter; only samples with at least 80% viability and at least 1,000 cells/μl were used.

RNA preparation, library generation, and scRNA-Seq.

Single-cell libraries were generated using the Chromium Single cell 3’ GEM library (10X Genomics) according to the manufacturer’s instructions. Each capture targeted 10,000 cells per sample; sequencing targeted 20,000 reads per cell. Generated cDNA libraries were sequenced on a NovaSeq 6000 sequencing system at the Maryland Genomic Sequencing Center. 7109 to 16284 cells were captured on the 10x Chromium chip in separate small intestinal samples from M₁R and M₃R male littermate control mice.

scRNA-Seq data analysis.

Cell Ranger 6.1 (10X Genomics) was used to align sequences to the Ensemble mouse MM10 genome. Raw scRNA-seq matrix was analyzed using standard pipelines in Seurat V4.3.0. Low-quality cells were removed using the following options: 200 < nFeatures<10000, percent.mt < 15%, and nCount_RNA < 30000. Variance stabilizing normalization of the filtered counts was performed using the SCTransform function. FindIntegrationAnchors and IntegrateData commands of Seurat were used to integrate samples and correct for batch effects. Thirty principal components were used for dimension reduction and Uniform Manifold Approximation and Projection (UMAP) were used for visualization in the dimensionally reduced space. Clusters were annotated based on the public marker databases using R package (clustermole) (Supplemental Table 1). Raw data were deposited in the NCBI’s Gene Expression Omnibus database (GEO).

Quantitative RT-PCR (qPCR).

qPCR was performed on RNA obtained from tissue sections and intestinal mucosal scrapings stored in RNAlater. First-strand cDNAs were synthesized from 5 μg RNA (Superscript III First Strand Synthesis System for RT-PCR, Invitrogen). qPCR was performed using 50 ng cDNA, the SYBR Green PCR Master Mix (Applied Biosystems), and forward and reverse primers (final concentration 0.5 μM in 20-μl sample volumes). The Chrm1-5 primer sequences we used were validated previously by the Wess laboratory at the National Institutes of Health (Bethesda, MD) (22). Chrm1-5 and Gapdh primer sequences (all 5’-3’):

Chrm1 Forward: AGTCCCAACATCACCGTCTTG

Reverse: CAGGTTGCCTGTCACTGTAGC

Chrm2 Forward: TGGAGCACAACAAGATCCAGAAT

Reverse: CCCCTGAACGCAGTTTTCA

Chrm3 Forward: ACCTGTTCACGACCTACATCA

Reverse: AGTGAGTGGCCTGGTAATAGAAA

Chrm4 Forward: GTGACTGCCATCGAGATCGTAC

Reverse: CAAACTTTCGGGCCACATTG

Chrm5 Forward: GGCCCAGAGAGAACGGAAC

Reverse: TTCCCGTTGTTGAGGTGCTT

Gapdh Forward ACAACTTTGGCATTGTGGAA

Reverse GATGCAGGGATGATGTTCTG

qPCR was performed using Step One (Applied Biosystems; Waltham, MA) with Power SYBR Green Master Mix (ABI). PCR conditions included 5 min at 95°C followed by 40 cycles at 95°C for 15 seconds, 60°C for 20 seconds, and 72°C for 40 seconds and a final cycle at 95°C for 15 seconds, 60°C for 15 seconds, and 95°C for 15 seconds. PCR data were analyzed using ABI instrument software SDS 2.1. Expression of genes was normalized to Gapdh. Quantitative qPCR data were evaluated using the comparative C_T (2^−ΔΔCT) method.

Immunohistochemistry.

To visualize goblet cells, we stained six small intestinal tissue slides per genotype with Alcian blue-periodic acid Schiff (Sigma; catalog numbers B8438 and 395B). Anti-DCAMKL1 antibody (Abcam; catalog number 31704), and cell junction stain (Abcam; catalog number 397581) were used to identify tuft cells and e-cadherin, respectively. Chromogranin A was used to identify enteroendocrine cells, and hematoxylin and eosin staining was used to visualize Paneth cells. ImageScope (Aperio) was used to quantify goblet, enteroendocrine, and Paneth cells. ImageJ was used to quantify tuft cells within five high-power field images per tissue slide taken at 20x magnification. Tuft cell count is reported as tuft cell number/crypt-villus unit as they were detected both at the +4 crypt position and along the epithelial lining of villi. Researchers conducting the analysis were masked to mouse genotype.

Statistical analysis.

Unless indicated otherwise, data are presented as means ± SEM of at least three independent experiments. We used the two-tailed unpaired Student's t-test to compare continuous outcomes that were approximately normally distributed. For not normally distributed data, we used the Kolmogrov-Smirnov test for analysis. We considered p <0.05 statistically significant.

Results

Reported effects of M₁R and M₃R deletion on murine body weight.

We searched public databases for papers relevant to the effects of M₁R and M₃R deficiency on weight. The results of this literature search, summarized in Tables 1 and 2, revealed several common themes. First, there is more published information regarding the effects of global M₃R (Table 1) vs M₁R (Table 2) deficiency on body weight. Second, with few exceptions, prior studies reporting weight comparisons provided data only for male mice. Third, most mice were on 129SvEv/CF1 or C57BL/6 genetic backgrounds. Fourth, in many studies it was difficult to ascertain whether comparisons were to littermate control mice or whether consideration was given to possible differences in the gut microbiome, e.g., information regarding whether mice were co-housed. These findings raise questions about possible confounding variables that could impact murine body weight.

Table 1.

Published data reporting the effects of M₃R deficiency on murine body weight.

M3R KO weight, g	Control weight, g	p-value	n	Age, wks	Sex	Genetic background	Key Findings	Ref
ND; graph available	ND; graph available	< 0.001	12-19	12	M, F	C57BL/6J × 129SvEv	Global KO mice weighed ~22% < WT littermates	(14)
ND; graph available	ND; graph available	< 0.001	8-15	14-24	M	129SvEv	Global KO mice weighed ~25% < WT littermates	(14)
30.7 ± 1.5	35.7 ± 1.6	< 0.05	3-9	15-27	M	C57BL/6J × 129SvEv	Global KO mice weighed < age-matched controls (unclear if littermates)	(38)
26.1 ± 1.2	33.8 ± 0.5	< 0.001	ND	11-23	M	129SvEv/CF1	Global KO mice weighed < age-matched controls (not littermates)	(39)
35.4 ± 0.5	42.3 ± 2.0	< 0.05	9-11	Adult	M	129SvEv/CF1	Global KO mice weighed < age-matched controls (unclear if littermates)	(40)
ND; graph available	ND; graph available	< 0.05	6	4-24	M	129SvEv/C57BL/6J	Global KO mice weighed < WT and M3R+/− littermates.	(41)
20.7 ± 0.5 - week 0; 27.9 ± 0.5 - week 20	28.3 ± 0.5 - week 0; 33.4 ± 1.0 - week 20	< 0.001 at 0 and 20 weeks	16 KO, 22 WT	ND	M	129S6/SvEvTac x CF1	Global KO weighed ~16% < WT mice (unclear if littermates)	(12)
ND	ND	< 0.01 at each time point	29 KO, 14 WT	6-12	M	C57BL/6 (Apcmin/+) x SvEv:CF1 (Chrm3−/−)	Global KO mice weighed ~20% < littermate controls	(13)
22.8 ± 0.5	27.5 ± 0.6	< 0.05	8-9	10-12	M	C57Bl/6NTac x 129 Sv/J	Global KO mice weighed < WT controls (not littermates)	(42)
28.3 ± 1.2	33.8 ± 0.9	< 0.05	13	9-12	M	129SvEv/CF1	M3R global KO weighed less than WT mice (unclear if littermates)	(43)
24.06 ± 0.54	28.70 ± 0.51	< 0.05	20 KO, 25 WT	6	M	129S6/SvEvTac x CF1	Global KO weighed 15-20% < WT mice (unclear if littermates)	(15)
28.5 ± 1.1	31.3 ± 1.4	0.14	6 KO, 6 WT	10-14	M	ND	Global KO mice tended to weigh less than WT littermates	(44)
ND; graph available	ND; graph available	< 0.05	6-10	0-24	F	C57BL/6J	Global KO mice weighed less than WT littermates, had less food intake yet similar metabolic rates, and shorter body and femur lengths	(32)
ND; graph available	ND; graph available	< 0.05	5-7	0-15	M, F	129/SvJ	Global KO weighed ~50% less than WT littermates; KO male weight was nearly the same as WT littermates by 15 weeks, but KO female weights remained ~20% less than WT females; KO mice fed paste food weighed more than those fed standard dry food	(33)
25.2 ± 0.7	25.5 ± 2.9	> 0.05	4 KO, 6 WT	8	M	ND	Global KO mice weighed the same as WT littermates	(45)
14.7 ± 0.3	19.8 ± 0.4	< 0.01	5 KO, 6 WT	7	M	C57BL/6NTac	Global KO mice weighed less than WT littermates at 7 weeks and 10 months; KO mice also had less lean mass and food intake (p < 0.01), and a trend towards less fat mass	(46)
M: 24.5 ± 1.4 F: 17.3 ± 1.4	M: 28.8 ± 1.6 F: 19.4 ± 0.3	< 0.05	6 KO M, 4 KO F, 7 WT M, 3 WT F	ND	M, F	C57BL	M3R global KO weighed less than WT mice (unclear if littermates)	(47)
18.2 ± 0.4	19.6 ± 0.2	< 0.05	18	12	M	C57BL/6	M3R global KO weighed less than WT littermates	(48)
ND; graph available	ND; graph available	< 0.001	6	10	M	B6.129	KO mice weighed less than WT littermates	(49)
ND; unpublished graph provided	ND; unpublished graph provided	< 0.001	5-6	3-13	ND	BALB/c	KO mice weighed less than WT littermates	(50)
ND	ND	ND	10 KO, 10 WT	12	M, F	C57BL/6NTac	ND	(51)
ND	ND	ND	3-6	12-36	M	C57BL/6J x 129Sv/J	ND	(52)
ND	ND	ND	5 KO, 5 WT	ND	F	ND	ND	(53)

KO, knockout; n, number of mice; ND, no data provided in publication; Ref, references.

Table 2.

Published data reporting the effects of M₁R deficiency on murine body weight.

M1R KO weight, g	Control weight, g	p	n	Age, wks	Sex	Genetic background	Key Findings	Ref
34.0 ± 0.8	35.7 ± 0.5	> 0.05	ND	11-23	M	129SvEv/CF1	Global KO mice weighed the same as age- and sex- matched controls (not littermates)	(39)
26.3 ± 0.5	27.5 ± 0.6	> 0.05	8-9	10-12	M	C57Bl/6NTac x 129 Sv/J	Global KO mice weighed the same as WT controls (not littermates)	(42)
29.5 ± 0.6	32.7 ± 0.9	< 0.01	25 KO, 21 WT	11	M	C57BL/6J × 129SvEv	Global KO mice weighed ~10% < WT littermate controls	(54)
24.95 ± 0.48	28.70 ± 0.51	< 0.05	23 KO; 25 WT	5-6	M	129S6/SvEvTac x CF1	Global KO mice weighed < WT littermate controls	(15)
21.5 ± 1.5	24.2 ± 1.2	p < 0.0001	12 KO, 12 WT	8	M	C57BL/6J	Global KO mice weighed < WT littermate controls	(55)
16.3-22.9	16.3-22.9	p > 0.05	15 KO, 13 WT	6-8	F	C57BL/6	Global KO mice weighed the same as littermate controls	(56)
ND	ND	p > 0.05	ND	ND	ND	ND	Global KO mice weighed the same as WT controls (unclear if littermates)	(57)
ND	ND	p > 0.05	41 KO, 46 WT	8-36	ND	C57BL/6 x 129SvJ	Global KO mice weighed the same as WT controls (unclear if littermates)	(58)
ND	ND	p > 0.05	ND	ND	ND	CF1 x 129SvEv	Global KO mice weighed the same as littermate controls	(59)
ND	Range 15-34	ND	5 KO,5 WT	ND	M	C57BL/6J	ND	(60)
ND	ND	ND	33KO, 33 WT	8-10	M	129S6/SvEv x CF1	ND	(61)
ND	ND	ND	9 KO, 6 WT	3-5	M	C57BL/6 x 129/SvEv x CF1 [Bartko, 2014 Cohort 1]	ND	(62)
ND	ND	ND	9 KO, 7 WT	12-16	M	C57BL/6 x 129/SvEv x CF1 [Bartko, 2014 Cohort 2]	ND	(62)
ND	ND	ND	5 KO, 6 WT	12-16	M	C57BL/6 x 129/SvEv x CF1 [Bartko, 2014 Cohort 3]	ND	(62)
ND	ND	ND	18 KO, 18 WT	12-24	M	C57BL/6	ND	(63)
ND	ND	ND	5 KO, 4 WT	~24	ND	ND	ND	(64)
ND	ND	ND	ND	6-8	ND	C57BL/6J	ND	(65)
ND	ND	ND	16 KO, 13 WT	12-13	M	C57BL/6	ND	(66)
26-34	26-34	ND	6 KO, 6 WT	ND (“adult”)	M	C57BL/6	ND	(67)
ND	ND	ND	15 KO, 14 WT	4-8	M	C57BL/6NTac	ND	(68)
ND	ND	ND	ND	4-8	M	C57BL/6NTac	ND	(69)
ND	ND	ND	ND	4-8	M	C57BL/6NTac	ND	(70)

KO, knockout; n, number of mice; ND, no data provided in publication; Ref, references.

Despite these caveats, several conclusions regarding the effects of global deletion of M₁R and M₃R on mouse weight can be drawn from the data depicted in Tables 1 and 2. First, male mice with global deletion of M₃R consistently weighed up to 25% less than control mice (Table 1). Second, the results in male mice with global deletion of M₁R are inconsistent; some studies reported that M₁R-deficient mice weighed less than control mice, but others observed no meaningful differences (Table 2). Third, and perhaps most importantly, because of the potential confounders mentioned above and the global knockout of these muscarinic receptor subtypes, it is not possible to determine whether deletion of these receptors from intestinal epithelial cells contributes to variations in body weight.

To pursue this line of investigation, our initial experiments were directed at analyzing the cellular distribution of Chrm1 and Chrm3, the genes respectively encoding M₁R and M₃R, among small intestinal epithelial cells. For this analysis, we used tissue obtained from the proximal small intestine of the two different parental floxed mouse strains we used to create M₁R and M₃R CKO mice.

Expression patterns of Chrm1 and Chrm3 RNA in murine small intestinal epithelial cells.

To assess the expression of the genes for M₁R and M₃R in the small intestinal epithelial cell population, we used scRNA-Seq to compare the distribution of Chrm1 and Chrm3 RNA in cells derived from the mouse proximal small intestine. A bubble plot (Fig. 1A) depicts the expression of representative marker genes associated with small intestinal epithelial cell types; we used these marker genes to identify different cell populations in subsequent analyses.

Figure 1. Distribution of Chrm1 and Chrm3 expression in murine small intestinal mucosal cells.

A: Bubble plot depicts the expression of representative intestinal mucosal cell-type marker genes used to identify epithelial cell populations by single cell-RNA sequencing. Cell name abbreviations: LEP, late enterocyte precursor cells, CD8⁺ HT, CD8⁺_high T cells, IS, intestinal stem cells, EEP, early enterocyte precursor cells, CD8⁺ LT, CD8⁺ _low T cells, EB, erythroblasts, M, macrophages, TAC, transit amplifying cells, EC, enteroendocrine cells. B-C: UMAP plots depict the distribution and identity of small intestinal epithelial cell subtypes in mucosal samples obtained from M₁R and M₃R parental mouse strains, and the relative expression of Chrm1 and Chrm3 within each cell population. Dashed circles highlight greater relative expression of Chrm1 and Chrm3 within specific cell populations. D-E: Violin plots depict the distribution of Chrm1 and Chrm3 RNA expressions within small intestinal epithelial cell populations derived from M₁R and M₃R control mice. Small intestinal tissue samples for scRNA-Seq analysis were derived from one male littermate control M₁R parental strain mouse and one male littermate control M₃R parental strain mouse. log2-fold change > 0.25, FDR < 0.05.

Since we were obligated to use two different floxed parental strains to create M₁R and M₃R CKO mice, it was important to confirm there were no meaningful differences in the cellular distribution of M₁R and M₃R expression between these two mouse strains. Reassuringly, as illustrated by the UMAP plots depicted in Figures 1B and 1C, the distribution of Chrm1 and Chrm3 expression amongst the small intestinal epithelial cell subpopulations from the two floxed parental mouse strains were indistinguishable.

As shown in Figure 1B and C and the violin plots in Figure 1D and E, these scRNA-Seq assays revealed that Chrm1 and Chrm3 RNA were highly expressed in the goblet cell populations of both floxed parental mouse strains. Moreover, Chrm3 RNA was modestly expressed in immune cells, e.g., T cells, natural killer cells, and regulatory T cells. Negligible Chrm1 and Chrm3 RNA expression were detected in the enteroendocrine cell population (Fig. 1). Importantly, we detected no meaningful differences in these patterns of expression when comparing small intestinal epithelial cells derived from the two floxed parental mouse strains used to create CKO mice.

Comparison of M₁R and M₃R CKO and littermate control mouse phenotypes.

Age- and sex-matched M₁R and M₃R CKO male and female mice on a C57BL/6 genetic background were viable, fertile, and grossly indistinguishable from their respective littermate controls (Fig. 2A). For all genotypes, 12-week-old female mice were shorter than males (p<0.005). M₁R CKO mouse vs control mouse body length (cm) was 9.0 ± 0.2 vs 8.9 ± 0.1 for male mice and 8.9 ± 0.2 vs 8.1 ± 0.2 for female mice; M₃R CKO mouse vs control mouse length was 9.3 ± 0.2 vs 9.0 ± 0 for male mice and 8.9 ± 0.3 vs 8.5 ± 0.3 for female mice (Supplemental Fig. 3A) – none of these differences between sex-matched CKO and control mice achieved statistical significance. Likewise, the gastrointestinal tracts of 12-week-old M₁R and M₃R CKO mice were grossly indistinguishable from those of sex- and age-matched littermate controls (Fig 2B). M₁R CKO mouse vs control mouse small intestine length (cm) was 36.5 ± 1.8 vs 37.2 ± 1.0 for male mice and 38.0 ± 1.9 vs 36.3 ± 2.0 for female mice; M₃R CKO mouse vs control mouse length was 40.2 ± 1.0 vs 40.7 ± 0.8 for male mice and 41.0 ± 0.8 vs 37.8 ± 1.5 for female mice (Supplemental Fig. 3B); again, comparisons of these values revealed no statistically significant differences.

Figure. 2. M₁R and M₃R CKO phenotypes and expression of muscarinic receptors in murine brain and stomach.

A: M₁R and M₃R CKO male and female 12-week-old mice were grossly indistinguishable from their respective littermate controls. B: M₁R and M₃R CKO and littermate control mouse small intestine and colon were similar in length and appearance. Representative photographs are shown of the resected intestines from control littermate (top) and M₁R and M₃R CKO mice (bottom). The stomach is labeled S and the cecum labeled C. C: Relative expression of Chrm1, 2, 3, 4, and 5 RNA in whole brain tissue extracts from M₁R and M₃R CKO male mice and their respective littermate controls. RNA was measured by qPCR using the muscarinic receptor subtype primers shown in Methods. n = 3 M₁R and M₃R CKO and littermate control male mice. Bars represent means ± SEM. Results were normalized to Gapdh RNA expression using the value for Chrm1 RNA in control male mouse brain as a comparator. D: Relative expression of Chrm1, 2, 3, 4, and 5 RNA in whole stomach tissue extracts from M₁R and M₃R CKO mice and their respective littermate controls. RNA was measured by qPCR using the muscarinic receptor subtype primers shown in Methods. n = 3 M₁R and M₃R CKO and littermate control male mice. Bars represent means ± SEM. Results were normalized to Gapdh RNA expression using the value for Chrm1 RNA in control male mouse brain as a comparator. Two-tailed unpaired Student’s t test were used to compare expression levels.

Comparison of M₁R and M₃R CKO and littermate control mouse expression of Chrm1 and Chrm3 RNA.

To explore expression patterns for the genes encoding the five muscarinic receptor subtypes, we measured relative Chrm1, 2, 3, 4, and 5 RNA expression in the small intestine and compared the results to the expression of these receptor subtypes in the brain, liver, and stomach of 12-week-old male and female mice. We selected these organs anticipating that the brain would abundantly express muscarinic receptors, and that the stomach and liver would have intermediate and negligible expression, respectively, as has been previously described (23, 24). Moreover, since the brain, stomach, and liver lack villin expression, we anticipated expression of M₁R and M₃R in these organs would be unaffected by villin-Cre conditional knockout in intestinal epithelial cells.

As anticipated, we detected abundant expressions of both Chrm1 and Chrm3 RNA in the brain (Fig. 2C). In all five murine genotypes tested, brain Chrm1 expression exceeded that of the other four muscarinic receptor subtypes, whereas Chrm5 expression was the lowest (Fig. 2C). Thus, to compare relative RNA expression in the other tissues tested, we assigned Chrm1 RNA expression in male M₁R littermate control mouse brain as the comparator, arbitrarily setting that value at 100%. Compared to the brain, the relative expression of all five muscarinic receptor subtypes was substantially lower in the stomach (Fig. 2D) and even more so in the liver (Fig. 3A). In the stomach, presumably due to abundant smooth muscle, Chrm2 expression levels were the highest, followed by approximately 10-fold lower levels of Chrm1 and Chrm3 expression (Fig. 2D). Importantly, we observed no significant differences in Chrm1 and Chrm3 expression levels in brain, liver, and stomach extracts from M₁R and M₃R CKO mice compared to their respective littermate controls (Fig. 2C and D and Fig. 3A).

Figure 3. Muscarinic receptor RNA expression in livers and small intestinal mucosa from CKO and control mice.

A: Relative expression of Chrm1, 2, 3, 4, and 5 RNA in liver extracts from Chrm1 and Chrm3 CKO male mice and their respective littermate controls. Expanded vertical axis scale reveals relative expression of Chrm1, 2, 3, 4, and 5 RNA in liver extracts from Chrm1 and Chrm3 CKO male mice and their respective littermate controls. n = 3 Chrm1 and Chrm3 CKO and littermate control mice. B: Relative expression of Chrm1, 2, 3, 4, and 5 RNA in extracts of small intestinal mucosal scrapings from M₁R and M₃R CKO male and female mice and their respective littermate controls. RNA was measured by qPCR using the muscarinic receptor subtype primers shown in Methods. n = 5 female and 6 male M₁R and 6 female and 6 male M₃R CKO and the same numbers of littermate control mice. Bars represent means ± SEM. Results were normalized to Gapdh RNA expression using the value for Chrm1 RNA in control male mouse brain as a comparator. Two-tailed unpaired Student’s t tests were used to compare expression levels. Unless otherwise denoted in the figure, all other comparisons did not achieve statistical significance. C: Relative expression of Chrm1, 2, 3, 4, and 5 RNA in small intestinal mucosal extracts from female and male Chrm1 and Chrm3 CKO mice and their respective littermate controls. n = 5 female and 6 male Chrm1 CKO and littermate control mice and 6 female and 6 male Chrm3 and littermate control mice. Bars represent means ± SEM. RNA was measured by qPCR using the muscarinic receptor subtype primers shown in Methods. Results were normalized to Gapdh RNA expression using the value for Chrm1 RNA in control male mouse brain as a comparator.

Selective Chrm1 and Chrm3 deletion in conditional knockout mice.

To confirm that the genes for M₁R and M₃R were selectively deleted from small intestinal epithelial cells, we compared RT-PCR results in mucosal extracts from the proximal small intestine. Analyses of these extracts revealed nearly complete deletion of Chrm1 and Chrm3 RNA in small intestinal mucosa from CKO but not control mice; relative levels of Chrm1 and Chrm3 RNA expression were approximately 100-fold less in CKO mouse small intestinal mucosa (Fig. 3B). We believe the token residual expression of Chrm1 and Chrm3 RNA in small intestinal mucosal extracts from CKO mice most likely represents Chrm1 and Chrm3 RNA expression in smooth muscle cells, neurons, immunocytes, and other non-epithelial cells that do not express villin. To exclude sex differences in RNA expression, we compared expression of RNA for the five muscarinic receptor subtypes in small intestinal mucosal extracts obtained from female and male CKO and littermate control mice. As shown in Figure 3C, we detected no differences in Chrm1-5 expression levels when comparing results for females, males, or the two sexes combined. Likewise, there was a similar reduction of Chrm1 and Chrm3 RNA levels in small intestinal mucosa obtained from the respective male and female CKO mice. Like our findings in brain, liver, and stomach tissue, there was minimal expression of Chrm5 RNA in the proximal small intestine mucosa (Fig. 3B and C).

Altered distribution of goblet and tuft cells in the small intestine of M1R and M3R CKO mice.

To measure the effects of selective intestinal epithelial cell deletion of M₁R and M₃R on the relative distribution of tuft and goblet cells in different regions of the small intestine of M₁R and M₃R control and CKO mice, we used a combination of anti-DCAMKL1 antibody and Alcian blue-periodic acid Schiff staining. Interestingly, compared to littermate controls, we detected fewer goblet cells in the proximal third of the small intestine (p<0.05) (Fig. 4A) of M₃R CKO mice and twice as many tuft cells per villus in the middle third of M₃R CKO mouse small intestine (p<0.005) (Fig. 4B). Comparing small intestinal segments obtained from M₁R and M₃R control and CKO mice, we detected no statistically significant differences in the numbers of either Paneth or enteroendocrine cells (Supplemental Figure 4).

Figure 4. Quantification of goblet and tuft cells in small intestinal segments from M₁R and M₃R CKO mice compared to littermate controls.

A: Comparison of the numbers of goblet cells per villus in the proximal, middle (Mid), and distal small intestine of M₁R and M₃R mice compared to littermate controls. B: Comparison of the numbers of tuft cells per villus in the proximal, middle, and distal small intestine of M₁R and M₃R mice compared to littermate controls. Symbols and error bars represent means ± SEM.

Comparison of M₁R and M₃R CKO and littermate control mouse weights.

For all genotypes, female mice weighed ~20% less than males (p<0.0005). Nonetheless, over the course of 30 weeks, compared to age- and sex-matched control mice there were no statistically significant differences in the weights of female and male M₁R (Fig. 5A; Supplemental Table 2) and M₃R (Fig. 5B; Supplemental Table 3) control vs CKO mice. At 30 weeks of age, the body weights (mean ± SEM) of M₁R control vs CKO mice were 30.5 ± 0.3 vs 29.1 ± 0.6 grams for male mice and 22.2 ± 0.5 vs 22.1 ± 0.4 grams for female mice; the body weights of M₃R control vs CKO mice were 33.6 ± 0.6 vs 32.4 ± 0.9 grams for male mice and 25.6 ± 0.8 vs 25.2 ± 0.6 grams for female mice. Although male mice weighed significantly more than female mice, analysis of sex-matched values within each revealed no statistically significant differences.

Figure 5. Impact of intestinal mucosal M₁R and M₃R expression on mouse body weight and enteroid development.

A and B: Comparison of body weights for male and female M₁R (A) and M₃R (B) littermate control and CKO mice measured at 5 to 30 weeks of age. The numbers of mice weighed at each time point are shown in Supplemental Tables 2 and 3. Symbols and error bars represent means ± SEM. Many error bars are not visualized in the graph because their span was smaller than the symbol size. Some symbols are not detected because they overlap. The nonparametric Kolmogrov-Smirnov test was used to analyze comparisons between mouse body weights. C: Light microscopic (10X) images of representative immature (day-1) and mature (day-7) enteroids cultured in domes of 1:1 Matrigel Matrix and IntestiCult media. Immature enteroids (i.e., spheroids) have less than two crypt protrusions from the pseudolumen. D: Comparison of enteroid growth over 7 days relative to the number of day-1enteroids. E. Comparison of enteroid maturity over 7 days relative to mature day-1 enteroids (mature enteroids contain ≥3 crypt buds).

Comparison of the development of enteroids derived from M₁R and M₃R CKO and littermate control mouse small intestines.

Enteroids can recapitulate the self-renewal and differentiation capacity of adult M₁R and M₃R CKO mouse intestines. Prior data has shown that M₃R deletion in the colon leads to changes in colonic Lgr5+ stem cells counts (20). As an additional measure of whether deletion of M₁R and M₃R expression in small intestinal epithelial stem cells altered the development of the absorptive mucosal surface, we created enteroids from the four murine genotypes by mincing surgically-resected proximal small intestine, and embedding the tissue in Matrigel for ex vivo culture in growth factor-enriched media (25). The development of enteroids from the four genotypes, observed and analyzed over 7 days until the next passage, was grossly indistinguishable (Fig. 5C). Relative to day-1 of culture, enteroids derived from the four mouse genotypes developed at the same pace (Fig. 5D). Likewise, relative to day-1 of culture, enteroids developed crypt protrusions and matured at similar rates (Fig. 5E).

Discussion

The intestinal epithelium impacts weight maintenance via its vital role in the absorption of nutrients and the secretion and elimination of waste products from the gastrointestinal tract. As shown in Tables 1 and 2, several research groups reported that global deletion of M₃R results in substantially leaner mice, and a few reported similar results following global deletion of M₁R. Nonetheless, the findings in most of these studies were mitigated by important limitations. Besides the frequent lack of appropriate controls for comparisons, the general restriction to male mice, and the failure to consider that differences in the gut microbiome could impact weight disparities, these studies did not explore whether intestinal epithelial cell-selective deletion of these two muscarinic receptor subtypes altered body weight. Indeed, prior to the current report, little was known regarding the general distribution of M₁R and M₃R in the different cell types comprising the small intestinal epithelium. Due to recent reports of contrasting outcomes following activation of M₁R and M₃R (24, 26), we considered it important to explore potential differences in the distribution of M₁R and M₃R among the epithelial cells comprising the small intestinal mucosa.

We found the highest levels Chrm3 RNA expression in goblet cells but also detected moderate Chrm3 RNA expression in immune cells of the intestinal epithelium. Chrm1 RNA was also expressed primarily in goblet cells, although to lower levels relative to Chrm3. Chrm3 was moderately expressed in enteroendocrine cells. Goblet and immune cells are not known to be key players in weight regulation or maintenance within the small intestinal epithelial layer. To address other limitations in prior reports, in the present communication we compared outcomes in CKO mice with selective intestinal epithelial cell deletion of M₁R and M₃R to those in littermate control mice. Moreover, we examined outcomes in both male and female mice and co-housed animals to control for potential differences in the gut microbiome. Intake of rodent chow was standardized amongst animals. It is unlikely that modified diets, such as a high-fat diet, would have different effects on CKO vs control mice.

Selective deletion of M₁R and M₃R altered the numbers of goblet and tuft cells in a region- dependent manner in the small intestine. Because tuft cells express choline acetyltransferase and are a major source of intestinal acetylcholine production (8), expansion of this cell type may constitute a feedback mechanism to compensate for the loss of muscarinic signaling due to MR deletion, thereby sustaining the cholinergic niche. Additionally, tuft cells are known to be involved in the regulation of epithelial and stromal cell differentiation in the wake of mucosal alterations (8). Future studies will explore the impact of the increased numbers of tuft cells on other cell types, such as Lgr5+ cells, and utilize organoids and other experimental models to investigate the compensatory mechanisms that strive to preserve cholinergic signaling. To characterize the impact of selective deletion of M₃R on immune cell subpopulations and host defense, we propose additional gene expression assays and studies using experimental models of colitis and cancer. Though published data demonstrate global M₃R deficiency worsens inflammation in the dextran sodium sulfate-induced colitis model (27), it is not known whether intestinal epithelial cell-selective deletion of M₃R sustains the same effect.

Perhaps surprisingly given their important functions, we found that selective deletion of M₁R and M₃R from intestinal epithelial cells did not significantly impact murine weight or the initiation and development of enteroids derived from those cells. Besides the known lower size and body weight of female compared to male mice, we detected no differences between either M₁R or M₃R CKO mice and their sex- and age-matched littermate controls. Given the lack of observable difference in density of intestinal enteroendocrine cells between the CKO and control groups, it is possible that signaling via enteroendocrine-derived peptides (e.g., GLP-1, GIP, ghrelin) is involved in weight regulation in a manner independent of local intestinal epithelial expression of muscarinic receptors. It is also possible that the observed differences in goblet cell density could lead to changes in the amount of mucus production which could affect nutrient reabsorption. Follow up studies utilizing Transwell chambers and other experimental models will aim to analyze how alterations in goblet cell number may impact the mucous layer. Although goblet cells and the mucin they produce are well known to play a role in the gut-immune interaction and protection of the mucosa from inflammation, less is known regarding causal relationships between goblet cells and weight maintenance (28, 29).

It is noteworthy that investigators have reported differences in weight gain between N and J substrains of C57BL/6 mice (30). A strength or our experimental approach was that we compared outcomes to M₁R and M₃R littermate control mice separately, we did not make comparisons between the two genotypes. Hence, we accounted for the J vs N substrain status of the mice as a potential variable and substrain status did not impact either our results or conclusions. Notably, as substrain information is not provided in other studies, many of which were published before this potential confounder was reported, this confounder may have impacted weight comparisons in other investigations using M₁R and M₃R floxed mice if appropriate controls were not employed.

Thus, we conclude that differences in the weights of global M₁R- and M₃R-deficient mice are more likely due to receptor deficiency in extra-intestinal organs, such as the brain, which has the greatest relative expression of these receptors in the body (Figure 2). This conjecture is in line with studies demonstrating the importance of the dorsal motor nucleus of the vagus nerve system in intestinal fat absorption, and with the increasingly appreciated role of the brain-gut axis in metabolic homeostasis (31). Nonetheless, it is also possible that M₁R and M₃R expressions in additional extra-intestinal organs play a role. For example, Shi et al. found that compared to littermate controls, male and female M₃R-deficient mice had lower bone mass at 6, 12, and 24 weeks, with higher levels of bone resorption and lower levels of bone formation biomarkers (32). These findings could not be attributed to differences in the levels of circulating leptin, insulin-like growth factor 1 (IGF-1), or growth hormone. The investigators then created CKO mice with selective neuronal deletion of M₃R – those animals had the same low bone mass as M₃R KO mice, but, unlike the findings in Yamada et al. (14), the CKO mice were neither hypophagic nor lean (32). Although the investigators did not directly examine body weight, these effects on bone metabolism provide another potential mechanism whereby cholinergic signaling via M₃R may impact weight regulation. Another possibility considering the work of Matsui et al. (33) and Gautam et al. (34) is that in global M₃ muscarinic receptor knockouts, disruption of salivary function impairs mouse feeding behavior.

We acknowledge several study limitations. These include the use of the villin-Cre approach to delete M₁R and M₃R selectively from murine intestinal epithelial cells. Extra-intestinal Cre activity in Tg(Vil1-cre)1000 Gum transgenic lines are reported in specific nuclei in the brain and in scattered regions of the pancreas (35). Though we observed no differences in Chrm1 and Chrm3 expression between control and CKO brain tissues, we did not analyze specific areas of the medulla and thalamus wherein nuclei are reported to have Cre recombinase activity (35). These limitations would have been more concerning if we had observed a meaningful weight difference between CKO and control mice.

Another limitation is our use of the proximal small intestine, not the entire small intestine, for scRNA sequencing to assess the distribution of Chrm1 and Chrm3 expression and for the preparation of enteroids. Cost constraints limited the number of mice we were able to employ for scRNA-Seq analysis. Yet, given the robust number of viable cells sequenced per sample and the quality controls employed in our scRNA-Seq analyses, we are confident we achieved an unbiased analysis of the distribution of Chrm1 and Chrm3 expression within the intestinal epithelium. Of course, we cannot exclude a role for M₁R and M₃R expression in components of the enteric nervous system, immunocytes, or other intestinal cell types that do not express villin. The analysis of enteroendocrine cell subtypes and alterations in corresponding hormone levels was not performed given the lack of differences in enteroendocrine cell density but does not exclude possible changes in subtype proportions. The animal chow was standardized, but we did not assess food intake. Despite these limitations, in concert with the data summarized in Tables 1 and 2, our findings strongly support the conclusion that extra-intestinal and not intestinal epithelial cell muscarinic receptor expression plays a role in regulating body weight. Lastly, we acknowledge the limitation regarding the differences between murine and human biology (21); this requires us to be circumspect regarding the potential broader implications of our findings to human physiology.

Despite these limitations and acknowledging that we currently do not know precisely in which cell types the lack of muscarinic receptor expression plays a role in murine weight regulation, we achieved the goals of our research. By excluding a role for intestinal epithelial cell expression of M₁ and M₃ muscarinic receptors in murine weight regulation, we believe this work provides an important advance in the field. Although GLP-1-based and other pharmaceutical approaches are efficacious at promoting human weight reduction, they have limitations. These include their high cost, adverse side-effects, and failure to surpass therapeutic plateaus (36). In the interest of identifying novel bariatric pharmaceuticals, future studies will address the role that muscarinic receptor expression by non-epithelial cells of the gut, including components of the enteric nervous system, play in weight regulation.

Conclusions

Our findings reveal differential expression of Chrm1 and Chrm3 RNA in subpopulations of small intestinal epithelial cells. Goblet cells highly express Chrm1 and Chrm3, whereas immune cells moderately express Chrm3. These differences are consistent amongst the two mouse models we used to generate conditional M₁R and M₃R knockout mice. Interestingly, CKO mice had fewer small intestinal goblet cells and only M₃R CKO mice had more tuft cells; as tuft cells are reported to produce acetylcholine, the latter finding suggests a possible compensatory feedback mechanism to compensate for reduced muscarinic receptor expression. Notably, unlike global deletion of M₁R and M₃R, intestinal cell-selective deletion of these receptors did not impact murine body weight or enteroid development. These findings suggest that extra-intestinal M₁ and M₃ receptor expression, such as in the brain, enteric nervous system, or salivary glands, is more likely to play a key role in weight regulation (37).

Supplementary Material

Supplemental materials are available via DOI: https://doi.org/10.17605/OSF.IO/V5T28. These are composed of Supplemental Tables 1, 2, and 3, and Supplemental Figures 1, 2, 3, and 4.