Society: AGA

This session will focus on intercellular signaling mechanisms between neurons and non-neuronal cells that control gut motor functions. New data show that the activity and output of enteric neurons is regulated by multiple non-neuronal cell types including glia, interstitial cells of Cajal, fibroblast like cells, macrophages, and enterochromaffin cells. Understanding how these cells shape enteric neuromuscular function and how perturbations to these processes lead to motility disorders is important for ultimately treating gut motility disorders.

BACKGROUND: Gut resident macrophages play a critical role in diabetic gut dysmotility. Our previous study showed that miR-10b-5p regulates diabetes and gut dysmotility.¹ However, the molecular mechanism of miR-10b-5p in regulating gut resident macrophage polarization has yet to be fully explored. Here, we aimed to establish whether gut immune dysfunction, mediated by macrophage polarization, a core pathology present with diabetes and gut dysmotility, is regulated by miR-10b-5p. METHODS: To explore the regulatory mechanism of miR-10b-5p in gut macrophage polarization, we used a global mir-10b knockout mice line, generated using the Mb3Cas12a/Mb3Cpf1 endonuclease, as well as a high-fat, high-sucrose diet (HFHSD) induced diabetic mice with gut dysmotility. Macrophages were isolated from the colonic muscularis propria and analyzed through flow cytometry. M1 pro-inflammatory and M2 anti-inflammatory macrophage phenotypes were determined. Enrichment of M1 or M2 macrophage population was further evaluated by measuring pro-inflammatory and anti-inflammatory cytokines. Loss-of-function and gain-of-function studies were conducted to characterize diabetic gut dysmotility in mice. RESULTS: A deficiency of mir-10b in mice led to a shift in macrophage polarization from M2 anti-inflammatory dominant population to M1 pro-inflammatory dominant population, leading to increased pro-inflammatory TNF-α and IL-6 in mir-10b KO mice. The M1 macrophage polarization was associated with the development of diabetes and gut dysmotility in the mice. A similar pattern of M1 macrophage dominant population with increased pro-inflammatory TNF-α and IL-6 was observed in HFHSD-induced diabetic gut dysmotility mice. Injection of a miR-10b-5p mimic in diabetic gut dysmotility mice reversed the macrophage polarization to M2 anti-inflammatory dominant population with increased anti-inflammatory TGF-β and IL-10, rescuing diabetes and gut dysmotility phenotypes. CONCLUSIONS: miR-10b-5p positively regulates the gut resident macrophage polarization in mice with diabetic gut dysmotility. Restoration of miR-10b-5p may provide therapeutic benefits for gut immune dysfunction associated with diabetic gut dysmotility.

Reference:
Singh R, Ha SE, Wei L, Jin B, Zogg H, Poudrier SM, Jorgensen BG, Park C, Ronkon CF, Bartlett A, Cho S, Morales A, Chung YH, Lee MY, Park JK, Gottfried-Blackmore A, Nguyen L, Sanders KM, Ro S. miR-10b-5p Rescues Diabetes and Gastrointestinal Dysmotility. Gastroenterology. 2021 Apr;160(5):1662-1678.e18.

Background: The phenotype of muscularis macrophages (MMs) changes with the development of diabetic gastroparesis (DG), a complication of diabetes characterized by delayed gastric emptying. MMs phenotype depends partially on their origin. Both monocyte- and embryonic-derived cells contribute to the tissue resident MM population. The heterogeneity of resident MMs has been studied in the small intestine and colon but not for gastric MMs, of particular interest for DG. Aim: To assess gastric MMs heterogeneity and whether diabetes differentially affects MMs populations of different origin. Methods: C57Bl6 CX3CR1^gfp mice were made diabetic by STZ injection (160mg/kg) for 2 weeks of diabetes. Sorted CD45⁺ cells from dissociated gastric muscularis propria (n=3 WT) were analyzed using 10X Genomics Cellranger Single Cell Software Suite (v3.1.0) and Seurat package (v3.1). Gene sets of each cluster were analyzed using Gene Set Enrichment Analysis to identify enriched functional pathways. Results: Unbiased clustering analysis identified six clusters of tissue resident MMs (Fig.1A). Each cluster was enriched by a unique gene set (Fig.1B), whose function was determined using GSEA analysis (Fig.1C). Clusters 0 and 1 were enriched in genes previously associated with embryonic derived macrophages, such as Cx3Cr1, Lyve1 and Timd4. IHC confirmed Cx3Cr1⁺ and Lyve1⁺ MMs co-labeled with MHCII, a general MM marker, and populated specific tissue niches. Cx3Cr1⁺ MMs were found primarily within the myenteric region (MR) compared to the muscle layers (ML) (MR:34.5 ± 1.4 cells/field; ML:13.4 ± 1.6 cells per field, p<0.01), whereas Lyve1⁺MMs were more abundant within the submucosal plexus (SP) than the muscularis propria (MP), in proximity to large blood vessels (SP: 45.3± 2.7cells/field; MP: 12.6± 1.3cells per field, p<0.01). The expression levels of Cx3Cr1 and Lyve1, two genes enriched respectively in clusters 0 and 1, were changed in mice with diabetes (db) compared to controls (Cx3Cr1; CTRL: 2.45± 0.7; db: 0.75± 0.3, P<0.05) (Lyve1; CTRL 3.54± 0.8; db: 6.45± 1.1, P<0.05). Combination of IHC and flow cytometry data, in line with gene expressions, showed less Cx3Cr1 MMs (CTRL:52.4± 4.4 cells/field; db:23.4 ± 1.6 cells per field, p<0.01) and more Lyve1 MMs (CTRL57.9± 3.8 cells/field; db:76.7 ± 5.6 cells per field, p<0.05) after the onset of diabetes. Conclusions: We identified two subpopulations of gastric MMs suggestive of embryonic origin, which share a protective, anti-inflammatory phenotype and are located within specific tissue niches. In addition to genes previously identified in embryonic MMs of other gut regions, these populations expressed unique genes whose expression may depend on the gastric tissue environment. Early diabetes increases the number of Lyve1⁺ MMs and reduces the number of Cx3Cr1⁺MMs. Funding: NIH R01 DK127992 and DK129297.

<b>Figure: A) </b>Identification of the 6 different populations of muscularis macrophages obtained from unbiased-cluster analysis. <b>B) </b>Heat map of set of different genes enriched in each cluster. <b>C) </b> Main functional pathways associated to each cluster identified by GSEA analysis.

Figure: A) Identification of the 6 different populations of muscularis macrophages obtained from unbiased-cluster analysis. B) Heat map of set of different genes enriched in each cluster. C) Main functional pathways associated to each cluster identified by GSEA analysis.

Background: In the central nervous system (CNS), Complement 1q (C1q) proteins are secreted by microglia to mediate synaptic pruning. In the GI tract, a population of macrophages called muscularis macrophages (MMs) share a similar phenotype to microglia and is distributed across the gut muscularis propria participating in the regulation of GI motility. Despite the prominent role in the CNS, the distribution of C1q proteins within the gut MMs and its contribution to GI motility is unknown.
Aim: This study aimed to determine the distribution of C1qa MMs and assess the impact of C1qa MMs on enteric neurons and GI motility.
Methods: CD45⁺ cells were isolated from dissociated human and mouse colon muscularis propria and analyzed using 10X Genomics Cellranger Single Cell Software Suite (v3.1.0) and Seurat package (v3.1). Whole-mount immunohistochemistry and RT-qPCR were performed on gut muscularis propria to determine C1qa distribution and expression. To conditionally knockout (KO) C1qa from immune cells, we generated a LysM^C1qa mouse model to compare with LysM^cre- control mice. GI transit was determined by measuring the whole gut and small bowel transit and gastric emptying (GE) of solids.
Results: Single-cell sequencing of human and mouse CD45⁺ cells identified a macrophage cluster enriched for C1qa. Whole-mount immunohistochemistry showed that C1qa MMs are closely associated with enteric neurons (ENs) (Fig.1A). Immunofluorescence staining and expression of two macrophages’ markers (Cd68 and Adgre1) demonstrated no changes in MMs number per field between the experimental groups (LysM^C1qa:123±23.33; LysM^cre-:128±34.76, p=NS). LysM^C1qa mice had a faster whole-gut transit time than control mice (LysM^C1qa:250±23 min; LysM^cre-:320±32 min, p<0.05) (Fig.1B). GE was not different (LysM^C1qa:85±12.45 min; LysM^cre:83±10.21 min), while small bowel transit time, reported as percentage of bowel crossed by the die, was changed between the groups (LysM^C1qa:98±1.45%; LysM^cre-:65±3.4%, p<0.05). Real-Time PCR and immunofluorescence revealed that LysM^C1qa mice had higher expression levels of synaptic markers, such as PSD95 and SYP in the stomach (LysM^C1qa:1.08±0.05; LysM^cre-:1.31±0.18, p<0.05), small intestine (LysM^C1qa:1.07±0.04; LysM^cre-:3.36±0.23, p<0.001) and colon (LysM^C1qa:1.03±0.01; LysM^cre- 1.84±0.12, p<0.01) compared to controls.
Conclusions: We identified a novel population of MMs expressing C1qa in the GI tract that affects GI transit potentially by regulating synapse formation. Ongoing studies focus on dissecting the effect of macrophage-released C1qa on synapse formation and pruning and C1qa MMs’ role in aged-associated GI dysmotility.
Funding: Supported by DK129297, DK127992-1, DK115255, ANMS Young Investigator grant and AGA grant#36 and P30DK084567.

<b>Figure1:C1qa MMs’</b> <b>distribution in the myenteric plexus and their effect on GI motility</b>.<br /> <b>A</b>) Whole mount immunofluorescence of C1qa in relation to MHCII (macrophage marker) and Beta3-Tub (nerves marker) within the myenteric plexus of the gut muscularis propria. <b>B</b>) Whole gut transit between the two experimental groups.

Figure1:C1qa MMs’ distribution in the myenteric plexus and their effect on GI motility.
A) Whole mount immunofluorescence of C1qa in relation to MHCII (macrophage marker) and Beta3-Tub (nerves marker) within the myenteric plexus of the gut muscularis propria. B) Whole gut transit between the two experimental groups.

Gianluca Cipriani

Appearances