Perivascular localization of macrophages in the intestinal mucosa is regulated by Nr4a1 and the microbiome

Abstract

While the ontogeny and recruitment of the intestinal monocyte/macrophage lineage has been studied extensively, their precise localization and function has been overlooked. Here we show by imaging the murine small and large intestines in steady-state that intestinal CX3CR1⁺ macrophages form an interdigitated network intimately adherent to the entire mucosal lamina propria vasculature. The macrophages form contacts with each other, which are disrupted in the absence of microbiome, monocyte recruitment (Ccr2^-/-), or monocyte conversion (Nr4a1^-/-). In dysbiosis, gaps exist between the perivascular macrophages correlating with increased bacterial translocation from the lamina propria into the bloodstream. The recruitment of monocytes and conversion to macrophages during intestinal injury is also dependent upon CCR2, Nr4a1 and the microbiome. These findings demonstrate a relationship between microbiome and the maturation of lamina propria perivascular macrophages into a tight anatomical barrier that might function to prevent bacterial translocation. These cells are also critical for emergency vascular repair.

Fig. 1. Intestinal CX3CR1⁺ macrophages and CCR2⁺ monocytes constitute the close network that surrounds vasculature.

a Representative stitched images (from 12 different fields of view) of small intestinal villi (left), small intestinal lamina propria (LP) (middle), and colonic LP (right) in Cx3cr1^GFP/+ mice. Scale bars, 100 μm. b Representative three-dimensional (3D) image of CX3CR1⁺ macrophages (green) in colonic LP and c quantification of their distribution. n = 5. d Representative 3D image of colonic LP CX3CR1⁺ macrophages and CCR2⁺ cells (red) at basal condition in Cx3cr1^GFP/+Ccr2^RFP/+ mouse. Arrows indicate orange/yellow cells. Scale bar, 50 μm. (e) Localization of colonic LP CCR2⁺ cells and their positional relationships with CX3CR1⁺ cells. n = 5. (f) Representative 3D images of small intestinal villi (left) and LP (right). Scale bars, 50 μm. g Localization of small intestinal CX3CR1⁺ and CCR2⁺ cells in villi and LP and their positional relationships. n = 5. Data represent mean ± SEM. Source data are provided as a Source Data file.

Fig. 2. Turnover of intestinal macrophage is dependent on monocyte conversion.

a Representative images of colonic LP in C57BL/6 mice 2, 4, 6, and 8 weeks after BMT from Cx3cr1^GFP/+Ccr2^RFP/+ mice. Scale bars, 50 μm. b Analysis of monocyte hues at the indicated time points. n = 3−5 in each time points. c Flow cytometry analysis of colonic LP CCR2⁺ and CX3CR1⁺ cells 2 weeks after BMT. Cells were pregated on size, viability, and CD45⁺. Data are representative of three independent experiments. d Time-lapse images demonstrate the change in color and shape of CX3CR1⁺CCR2⁺ cell 4 weeks after BMT (white arrow heads; round yellow to elongated green). White dotted lines exhibit the outline of indicated cell. Blue dotted lines indicate CD31⁺ vasculature. Scale bar, 20 μm. e Analysis of monocyte hues of the indicated cell in d at 0 and 45 min. f Representative images of colonic LP of C57BL/6 mice 6 weeks after BMT from Cx3cr1^GFP/+Ccr2^RFP/+, Cx3cr1^GFP/+Ccr2^RFP/RFP, or Nr4a1^−/−Cx3cr1^GFP/+Ccr2^RFP/+ mice. Scale bars, 50 μm. g Quantification of monocyte hues 6 weeks after BMT in each group. n = 3–5 per group. h Quantification of the number of CX3CR1^int and CX3CR1^hi macrophages by flow cytometry at 6 weeks after BMT in each group. Cells were gated on size, viability, CD45⁺, CD103⁻, CD11b⁺, F4/80⁺ and CX3CR1^int/hi. n = 4–8 per group. Data represent mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, NS not significant. Source data are provided as a Source Data file.

Fig. 3. Gut microbiota affects the distribution and morphology of intestinal CX3CR1⁺ macrophages.

Representative images of a small intestinal villi and b colonic lamina propria (LP), submucosa and muscularis in Cx3cr1^GFP/+Ccr2^RFP/+ mice (The left column is SPF control, the right column is Abx-treated mice). Scale bars, 50 μm. c Quantification of CX3CR1⁺ cells or CCR2⁺ cells per field of view (FOV) in each layer of the colon in control and Abx-treated mice. n = 6 per group. d Quantification of CD31⁺ area per FOV in colonic LP in control and Abx-treated mice. n = 6 per group. e Representative high-magnification and f 3D reconstructed images of CX3CR1⁺ macrophages and vasculature in LP. Scale bars, 20 μm. g Representative 3D reconstructed images of CX3CR1⁺ macrophages in LP in control and germ-free Cx3cr1^GFP/GFP mice. Quantification of CD31⁺ area per FOV in colonic LP in control and germ-free mice (right). n = 6 per group. h Flow cytometry analysis of the proportion of CX3CR1^int and CX3CR1^hi macrophages in the colonic LP in control and Abx-treated mice at steady state. Cells were pregated on size, viability, CD45⁺, CD103⁻, CD11b⁺, and F4/80⁺. i Absolute number of CX3CR1⁺, CD80⁺CD206⁻, and CD80⁻CD206⁺ macrophages in the total colon of control and Abx-treated mice. Cells were pregated on size, viability, CD45⁺, CD103⁻, CD11b⁺, and F4/80⁺. n = 4–8 per group. Data represent mean ± SEM. *p < 0.05, **p < 0.01, NS, not significant. Source data are provided as a Source Data file.

Fig. 4. Depletion of gut microbiota impairs the turnover of intestinal macrophages.

a Representative intravital images of colonic LP in SPF control and Abx-treated mice at 2, 4, 6, and 8 weeks after bone marrow transplantation (BMT) from Cx3cr1^GFP/+Ccr2^RFP/+ mice. Scale bars, 50 μm. b Analysis of monocyte hues in control and Abx-treated mice at the indicated time points. n = 3–5 in each time points. c The number of CX3CR1⁺CCR2^hi cells, CX3CR1^intCCR2^lo cells, and CX3CR1^hiCCR2^lo cells per colon at 2 and 6 weeks after BMT quantified by flow cytometry. Cells were pregated on size, viability, and CD45⁺. n = 4–5 per group. d Quantification of the number of CX3CR1^int and CX3CR1^hi macrophages at 2 and 6 weeks after BMT. Cells were pregated on size, viability, CD45⁺, CD103⁻, CD11b⁺, and F4/80⁺. n = 4–8 per group. e, f Quantification of monocyte hues of control and Abx-treated mice at 6 weeks after BMT from Cx3cr1^GFP/+Ccr2^RFP/+, Cx3cr1^GFP/+Ccr2^RFP/RFP or Nr4a1^−/−Cx3cr1^GFP/+Ccr2^RFP/+ mice. n = 3–5 per group. Data represent mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, NS not significant. Source data are provided as a Source Data file.

Fig. 5. CX3CR1⁺ macrophages in intestinal lamina propria of antibiotic-treated mice do not cover the vasculature and result in bacterial dissemination.

a Representative images of small intestinal villi and colonic lamina propria in control and Abx-treated Cx3cr1^GFP/+ mice at 24 h after the gavage of S. aureus MW2. Scale bars, 20 μm. b Quantification of area of vasculature covered by CX3CR1⁺ macrophages in control and Abx-treated Cx3cr1^GFP/+ mice with or without S. aureus infection. c CFU counts at 6 and 24 h after S. aureus MW2 gavage in control and Abx-treated mice. n = 9–10 per group. d Still image of the liver in control and Abx-treated mice at 24 h after gavage of mCherry-MW2. Scale bars, 50 μm. e Representative images of CX3CR1⁺ macrophage (green) catching S. aureus (red) in colonic epithelium at 24 h after the gavage of S. aureus mCherry-MW2. Higher magnification of the indicated area was shown in middle and 3D image was shown in right. Arrows indicate S. aureus. Scale bars, 20 μm. f Representative images of colonic lamina propria in control and Abx-treated Cx3cr1^GFP/+ mice at 6 hrs after the intraluminal injection of mCherry-Salmonella typhimurium. Scale bars, 50 μm. g Quantification of mCherry⁺ area in control and Abx-treated Cx3cr1^GFP/+ mice. h Images (right; 3D reconstruction) of mCherry-S. typhimurium caught by intestinal CX3CR1⁺ macrophages in SPF control mice. Transparency adjustment was applied (upper) to allow for visualization of S. typhimurium inside CX3CR1⁺ macrophages. Scale bar, 50 μm. 1 Unit = 23.3 μm. i Time-lapse images of colonic lamina propria in Abx-treated Cx3cr1^GFP/+ mice at 6 h after the intraluminal injection of mCherry-S. typhimurium. White arrows indicate S. typhimurium disseminating to blood flow. Blue dotted lines indicate CD31⁺ vasculature. Scale bar, 20 μm. j Flow cytometry analysis of the proportion of CX3CR1⁺ macrophages in the colonic LP in Csf1r^LsL-DTR/LysM^cre mice with or without diphtheria toxin treatment. Cells were pregated on size, viability, and CD45⁺. k CFU counts at 24 h after S. typhimurium (1 × 10⁹ CFU) gavage in control and macrophage-depleted mice. n = 6−8 per group. Data were pooled from two independent experiments. Data represent mean ± SEM. *p < 0.05, **p < 0.01, NS not significant. Source data are provided as a Source Data file.

Fig. 6. CD11b⁺Ly6C⁺CX3CR1⁺CCR2^hi monocytes accumulate early within intestinal sterile injury site and change their phenotype overtime.

a Representative images taken from 6 to 96 h after focal intestinal injury and b quantification of their monocyte hues within injury in control Cx3cr1^GFP/+Ccr2^RFP/+ mice. White dotted line indicates injury area. Scale bars, 100 μm. n = 5 per group. c Flow cytometry analysis of CX3CR1⁺ and CCR2⁺ cells in injured colon isolated from Cx3cr1^GFP/+Ccr2^RFP/+ mice with or without burn injury (6 h). Cells were pregated on size, viability, and CD45⁺. d Quantification of the proportion of CX3CR1^-CCR2⁺, CX3CR1⁺CCR2^hi, and CX3CR1⁺CCR2^lo cells per CD45⁺ cells. n = 3 per group. e Quantification of the proportion of CD11b⁺Ly6C⁺ cells per CX3CR1⁺CCR2^hi cells. n = 3 per group. f Representative image of the colon 7 days after burn injury (left). High-magnification 3D image was shown in right. Scale bar, 100 μm. g Localization (adjacent or non- adjacent to reconstructed blood vessels) of CX3CR1⁺ cells 7 days after burn injury. n = 3 per group. Data represent mean ± SEM. **p < 0.01, ***p < 0.001, NS, not significant. Source data are provided as a Source Data file.

Fig. 7. Gut microbiota is involved in the intestinal repair process by regulating the conversion of monocytes.

a Representative images taken from 6 to 168 h after focal intestinal injury and b quantification of their monocyte hues within injury in control or Abx-treated Cx3cr1^GFP/+Ccr2^RFP/+ mice. Scale bars, 100 μm. n = 4–6 per group. c Flow cytometry analysis and d quantification of the proportion and number of CX3CR1⁺ macrophages (total, CD80⁺CD206⁻, or CD80⁻CD206⁺) in injured colon isolated from control and Abx-treated mice at indicated time points after burn injury. Cells were pregated on size, viability, CD45⁺, CD103⁻, CD11b⁺, and F4/80⁺. Data are representative of three independent experiments. n = 4–8 per group. e Representative images of necrotic cells (SYTOX orange, red) within injury at 48 h post injury. Scale bars, 100 μm. f Quantification of SYTOX orange⁺ area within injury at 24 and 48 h post injury in control and Abx-treated mice. n = 5 per group. g Representative images of injury site 4 days post injury in control and Abx-treated mice. Mice were administered anti-CD31 (red) antibody intravenously to visualize vasculature. White dashed line highlights original injury border. Scale bars, 100 μm. h Quantification of revascularization (CD31⁺ area within injury) at indicated time points in control and Abx-treated mice. n = 5 per group. i Representative images of intestinal injury site 4 days post injury. Time-lapse images were taken before and after intravenous TRITC-albumin (red) administration in control and Abx-treated mice. Scale bars, 50 μm. j Quantification of extravascular (outside of CD31⁺ vasculature) TRITC⁺ area/FOV at indicated time points in control and Abx-treated mice. n = 5 per group. Data represent mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, NS not significant. Source data are provided as a Source Data file.

Fig. 8. Gut microbiota affects the tissue repair process in DSS colitis.

a Macroscopic findings of the colon 7 days after the start of DSS water in control and Abx-treated mice (left) and quantification of colon length (right; n = 8–14 per group). Luminex assays of b pro-inflammatory cytokines, c chemokines in colon tissue samples at steady state and at 7 days of DSS-treatment (n = 5–7 per group). Quantification of d monocyte hues and e CD31⁺ area within injury site of lamina propria in control and Abx-treated Cx3cr1^GFP/+Ccr2^RFP/+ mice at 7 days of DSS-treatment. n = 3–5 per group. f Quantification of the number of CX3CR1⁺ macrophages (total, CD80⁺CD206⁻, or CD80^-CD206⁺) in colon isolated from control and Abx-treated mice after 7 days of DSS treatment. Cells were pregated on size, viability, CD45⁺, CD103⁻, CD11b⁺, and F4/80⁺. n = 4–5 per group. g Survival of control and Abx-treated mice in 2% DSS-treatment. n = 10 per group. Data were pooled from two independent experiments. h Colonic blood score 7 days after the start of DSS water (n = 8–14 per group). i Body weight change of control and Nr4a1^−/− mice in 2% DSS colitis. n = 10 per group. j Quantification of colon length and colonic blood score (n = 9–14 per group). k Representative stitch images of lamina propria and l quantification of monocyte hues and CD31⁺ area/injury area in Cx3cr1^GFP/+Ccr2^RFP/+ and Nr4a1^−/−Cx3cr1^GFP/+Ccr2^RFP/+ mice at 7 days of DSS-treatment. White dotted line indicates injury area. Scale bars, 100 μm. n = 3–5 per group. Data represent mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, NS not significant. Source data are provided as a Source Data file.