Gut microbiome-derived butyrate inhibits the immunosuppressive factors PD-L1 and IL-10 in tumor-associated macrophages in gastric cancer

Figures & data

Figure 1. Increased levels of immunosuppressive markers in the PBMCs of AGC patients.

PBMCs were isolated from HC, EGC, and AGC patients, and stimulated with 25 ng/mL PMA and 250 ng/mL ionomycin for 4 h. After stimulation, cells were stained with CD68, CD11c, PD-L1, and IL-10 antibodies and analyzed by flow cytometry. (a) Representative FACS plots show the population of CD68+PD-L1+, CD68+IL-10+, CD11c+PD-L1+, and CD11c+IL-10+. (b) Bar graphs showing mean percentages of CD68+PD-L1+ (left) and CD68+IL-10+ (right) cells in HC, EGC (n = 11) and AGC (n = 29) patients. (c) Bar graphs showing mean percentages of CD11c+PD-L1+ (left) and CD11c+IL-10+ (right) cells in EGC (n = 11) and AGC (n = 29) patients. Data are mean ± SD (*p < .05, **p < .01, ***p < .001).

Figure 2. Increased levels of immunosuppressive markers in the tumor mucosa of AGC patients.

Isolated tumor mucosa tissues were stained with CD68, CD11c, PD-L1, and IL-10 antibodies during surgery and analyzed using confocal microscopy. (a) Representative images show CD68+PD-L1+ and IL-10+ cells in the tumor mucosa of EGC (top) and AGC (bottom) patients. (b) Bar graphs showing the mean number of CD68+PD-L1+ (left, n = 3) and CD68+ IL-10+ (right, n = 5) cells in the tumor mucosa of EGC and AGC patients. Data are mean ± SD (****p < .001).

Table 1. Clinicopathological characteristics of the patients with early gastric cancer and advanced gastric cancer.

		Patient with EGC	Patient with AGC	P value
		N = 14	N = 34
Age, mean (SD)		59.5(±12.61)	62.8235(±10.65)	<0.0001
Sex (M:F ratio)	M	10(71.4)	22(64.71)	0.340114
	F	4(28.6)	12(35.29)
BMI, mean (SD)		23.92(±5.369)	24.34(±3.699)	<0.0001
ECOG	0	12(85.71)	27(79.41)	0.050037
	1	2(14.29)	6(17.65)
	2	0	1(2.94)
Smoking (%)	Never	3(21.42)	16(47.06)	0.011447
	Quit	7(50)	7(20.59)
	Active	4(28.57)	11(32.35)
Alcohol (%)	Never	8(57.14)	23(67.65)	0.167044
	Social	5(35.71)	8(23.53)
	Heavy	1(7.14)	3(8.82)
H. pylori infection	Don’t know	9	20
	Negative	3	11
	Positive	2	3
Comorbidities	Non	4(28.57)	11(32.35)	0.762884
	Hypertension	5(35.71)	13(38.24)
	DM	1(7.14)	2(5.88)
	Cardiovascular	1(7.14)	1(2.94)
	Pulmonary	1(7.14)	2(5.88)
	others	2(14.29)	2(5.88)
pT stage	T1	14
	T2–4	0	34
pN stage	N0	14	6(17.65)	0.050044
	N1–3	0	28(82.35)
pM stage	M0	14	30(88.24)	0.001523
	M1	0	4(11.76)
pStage 7^th	I	14	7(20.59)	0.00023
	II	0	10(29.41)
	III	0	13(38.24)
	IV	0	4(11.76)
Differentiation	Differentiated	5(35.71)	10(29.41)	0.197
	Undifferentiated	9(64.29)	24(70.59)
Lauren	Intestinal	6(42.86)	8(23.53)	0.08618
	Diffuse/Mixed	8(57.14)	22(64.71)
	Indeterminated	0	4(11.76)
Lymphatic invasion	Negative	14	10(29.41)	0.012211
	Positive	0	24(70.59)
Vascular invasion	Negative	14	28(82.35)	0.00645
	Positive	0	6(17.65)
Neural invasion	Negative	14	16(47.06)	0.015401
	Positive	0	18(52.94)

Data are numbers (percentages) or means (± SD). The chi-squared test was used to test for between-group differences in categorical variables; P < .05 was deemed indicative of statistical significance. BMI; body mass index; ECOG, Eastern Cooperative Oncology Group score.

Figure 3. Gut microbiota profile of patients with GC.

Fecal samples were collected from HCs and GC patients, and analyzed for the gut microbiome. (a) Bar graphs showing observed OTUs (left) and Chao1 diversity (right). (b) Representative plots show principal coordinates analysis (left) and Bray-Curtis distance (right) between HCs and GC patients.

Figure 4. Differential microbial abundance in GC patients.

Gut microbiome of HCs and GC patients were analyzed at the phylum (a), family (b), and genus (c) levels. (d) Histogram showing the linear discriminant analysis (LDA) scores for bacterial abundances in HCs and GC patients. (e) Bar graphs showing relative abundances of Faecalibacterium, Collinsella, Bifidobacterium, f_ Ruminococcus, and Ruminococcus in HC and GC. (f) Bar graphs showing relative abundances of Faecalibacterium, Collinsella, Bifidobacterium, f_ Ruminococcus, and Ruminococcus in EGC and AGC.

Figure 5. Reduction of PD-L1 and IL-10 levels by butyrate.

Isolated PBMCs from patients with GC were cultured with 100 ng/mL LPS in the absence or presence of 0.5 mM of butyrate or 10 μg/mL Faecalibacterium for 72 h. After 72 h, the supernatant was harvested for ELISA and cells were stimulated with 25 ng/mL PMA and 250 ng/mL ionomycin for 4 h. After stimulation, cells were stained with antibodies against CD68, PD-L1, and IL-10 for flow cytometry. (a) Bar graphs showing mean percentages of PD-L1 (left) and IL-10 (right) of CD68⁺ cells in the indicated conditions. (b) Bar graph showing the mean IL-10 level in the supernatant of the indicated conditions. (c) Bar graphs showing the mean percentage of PD-L1 (left) and IL-10 (right) of CD68⁺ cells in the indicated conditions. Data are mean ± SEM (*p < .05, **p < .01, ***p < .001).

Figure 6. Inhibition of tumor cell growth by butyrate.

(a) AGS cells were cultured with different concentration of butyrate for 48 h. Cell growth was measured using CCK8 cell counting kit. (b) 5 × 10⁶ PBMCs from GC patients were injected into NSG mice. Seven days after PBMC injection, 5 × 10⁶ AGS cells were subcutaneously injected into mice. Fourteen days after the injection of AGS cells, blood samples were collected for flow cytometry. Then, 200 mg/kg of butyrate was administered orally to mice every day. The mice were euthanized at 54 days after the experiment initiation. (c) Graphs (top) showing tumor size in the Veh (PBMCs only), butyrate (PBMCs, AGS cells, and butyrate), and AGS (AGS only) groups. Bar graph (bottom) showing mean tumor size in the indicated group. (d) NF-_KB, IL-10, PD-L1, STAT3, VEGF, and GDF-15 expression levels in the tumor tissues of the indicated groups were assessed using immunohistochemistry. (e) Bar graphs showing the mean percentages of NF-_KB-, IL-10-, PD-L1-, and STAT3-positive area. Data are mean ± SEM (*p < .05, **p < .01).

Data availability statement

The authors confirm that the data supporting the findings of this study are available within the article. Gut microbiome analysis was deposited in the BioProject and Sequence Read Archive (SRA; https://www.ncbi.nlm.nih.gov/sra) under the accession numbers PRJNA937852.