Diet-induced shifts in the gut microbiota influence anastomotic healing in a murine model of colonic surgery

Figures & data

Figure 1. (a) Schematic of experimental protocol. (b) Anastomotic scoring examples demonstrating healed anastomosis with local adhesions (top left), contained AL/perianastomotic abscess (bottom left), and anastomotic dehiscence with gross peritonitis (right). Arrows point to anastomosis.

Table 1. Survival contingency table: no significant differences in survival to necropsy on POD7 between groups.

	Survived (n)	Died (n)	Fisher’s exact test P-value
LD PBS	15	1	0.38
WD PBS	15	1
WD-ldFMT	16	0
LD-wdFMT	13	3

Table 2. AL contingency table: no significant difference in AL observed between control dietary groups or between LD mice that received PBS vs. FMT. Significant reduction in AL observed in WD mice that received FMT vs. PBS.

	No AL (n)	+AL (n)	Chi-square or Fisher’s exact test P-value*
LD PBS	9	6	0.71
WD PBS	10	5
LD PBS	9	6	0.46
LD-wdFMT	12	4
WD PBS	10	5	0.04
WD-ldFMT	13	0

*Fisher’s exact test used if contingency table contains cells with value < 5.

Figure 2. a) Mean weight on DOS varied by dietary group: LD: 30.6 ± 2.6 g, WD: 36.4 ± 4.0 g (ANOVA F = 47.2, P <.0001). b) abundant taxa for each dietary groups after 10-week feeding period prior to DietGel initiation (baseline). c) Bray-Curtis dissimilarity indices demonstrated significant clustering by diet at baseline (ANOSIM R = 0.15, P <.001). Genera significantly correlated to dietary axis position by Spearman correlation are plotted on the PCoA chart with vector length indicating magnitude of correlation. *Not further classified to genus.

Table 3. Alpha diversity. ANOVA performed for pairwise comparisons demonstrated that LD had significantly higher alpha diversity at baseline (by Chao1 index) and on POD4 (by Shannon and Chao1 index). On POD4, WD mice with AL had significantly lower alpha diversity compared to mice without AL in both PBS and FMT groups.

		Mean Shannon index ± SE (a, b)*					Mean Chao1 index ± SE (a, b)*
		LD		WD		ANOVA P-value	LD		WD		ANOVA P-value
Baseline		2.7 ± 0.06 (a)		2.8 ± 0.06 (a)		0.19	172.6 ± 4.8 (b)		150.6 ± 4.7 (a)		0.0022
DOS		2.8 ± 0.05 (a)		2.7 ± 0.05 (a)		0.27	180.0 ± 6.1 (a)		170.6 ± 6.5 (a)		0.32
POD4		3.2 ± 0.05 (b)		3.0 ± 0.05 (a)		0.014	202.2 ± 11.3 (b)		166.9 ± 12.3 (a)		0.039
POD7		3.0 ± 0.04 (a)		3.1 ± 0.05 (a)		0.36	185.6 ± 18.0 (a)		199.0 ± 18.9 (a)		0.61
		AL – No		AL – Yes		P-value	AL – No		AL – Yes		P-value
LD	Baseline	2.7 ± 0.08 (a)		2.6 ± 0.10 (a)		0.69	171.4 ± 6.5 (a)		174.6 ± 8.3 (a)		0.77
	DOS	2.7 ± 0.06 (a)		2.8 ± 0.08 (a)		0.47	177.6 ± 5.5 (a)		183.8 ± 8.0 (a)		0.53
	POD4	3.2 ± 0.06 (a)		3.0 ± 0.08 (a)		0.067	213.3 ± 18.3 (a)		179.0 ± 26.5 (a)		0.30
	POD7	3.0 ± 0.05 (a)		3.1 ± 0.08 (a)		0.25	190.5 ± 12.4 (a)		175.3 ± 17.9 (a)		0.49
WD	Baseline	2.7 ± 0.07 (a)		2.8 ± 0.13 (a)		0.52	150.0 ± 5.2 (a)		152.3 ± 9.6 (a)		0.84
	DOS	2.7 ± 0.05 (a)		2.8 ± 0.11 (a)		0.33	173.5 ± 8.9 (a)		157.6 ± 18.7 (a)		0.45
	POD4	3.0 ± 0.05 (b)		2.7 ± 0.10 (a)		0.0023	167.7 ± 3.3 (a)		163.8 ± 6.7 (a)		0.61
	POD7	3.1 ± 0.05 (a)		3.2 ± 0.10 (a)		0.27	208.2 ± 27.9 (a)		156.6 ± 59.9 (a)		0.44
POD4		PBS		FMT		P-value	PBS		FMT		P-value
LD		AL – No	AL – Yes	AL – No	AL – Yes		AL – No	AL – Yes	AL – No	AL – Yes
		3.2 ± 0.09 (a)	3.0 ± 0.11 (a)	3.2 ± 0.08 (a)	3.1 ± 0.13 (a)	0.25	229.2 ± 28.7 (a)	180.0 ± 35.1 (a)	201.4 ± 24.8 (a)	177.5 ± 43.0 (a)	0.66
WD		3.1 ± 0.08 (b)	2.7 ± 0.10 (a)	3.0 ± 0.07 (b)	-	0.0085	174.7 ± 4.8 (a)	163.8 ± 6.4 (a)	162.4 ± 4.1 (a)	-	0.15

*Pairwise comparisons are denoted by lower case letters, (a) <; (b). Bolded values in the table are statistically significant.

Figure 3. (a) Significant clustering was observed on POD4 in the LD mice based on presence or absence of AL (ANOSIM R = 0.49, P <.001) with genera significantly correlated to axis position by Spearman correlation shown. (b) clustering between AL and no AL approached significance in the WD group (ANOSIM R = 0.27, P =.05). (c) clustering based on AL was lost with in LD mice that received FMT (FMT_N vs. FMT_Y, ANOSIM R = 0.085, P = 0.26). (d) significant clustering between AL and no AL was demonstrated among WD mice (ANOSIM R = 0.15, P =.047). FMT treatment made mice without AL further from mice with AL than PBS (FMT_N vs. PBS_Y, ANOSIM R = 0.28, P = 0.017; PBS_N vs. PBS_Y, ANOSIM R = 0.27, P =.043, Bonferroni pair-wise error rate 0.017). Genera significantly correlated to axis position by Spearman correlation are shown. *Not further classified to genus.

Figure 4. Canonical correspondence analysis demonstrates directional relationships between macronutrients, AL, and genera on POD4. *Not further classified to genus.

Figure 5. (a) Principal coordinate analysis (r² = 0.67) of Bray-Curtis dissimilarities among all samples on POD4. Groups are differentiated by color. Ellipses indicate approximate separation of samples by ANOSIM. (b) genera significantly correlated to axis position by Spearman correlation. *Not further classified to genus.

Table 4. Median relative abundances of predominant genera with interquartile range in brackets. Bolded taxa indicate a significant difference way found in a Kruskal–Wallis pairwise comparison. Superscripts indicate which specific Kruskal–Wallis pairwise comparison was significant.

Taxa	AL	LD	LD-wdFMT	WD	WD-ldFMT
Acetatifactor	1.7 [1.32–2.96]	1.67 [1.37–1.93]	1.71 [1.58–2.12]	2.13 [1.91–3.29]	2.62 [1.8–3.33]
Akkermansia	17.72 [13.05–28.9]^b	9.93 [7.44–10.41]^b	13.21 [11.28–16.57]^b	17.67 [8.14–19.71]	16.39 [11.01–21.5]
Alistipes	4.39 [2.69–6]^abc	7.31 [6.35–10.35]^b	7.01 [6.33–8.61]^b	6.05 [3.58–7.4]	6.41 [5.03–7.72]^c
Bacteroides	9.9 [6.84–14.25]^ab	2.16 [1.85–4.25]^b	6.28 [5.58–8.43]^b	6.55 [4.37–9.97]	7.45 [6.32–9.02]
Bifidobacterium	0.52 [0.38–1.39]	1.06 [0.34–1.72]	1.79 [0.67–2.57]	0.48 [0.16–0.72]	0.35 [0.07–0.69]
Clostridiales*	3.42 [2.13–4.6]	4.23 [1.8–4.99]	3.58 [1.95–4.17]	2.33 [0.78–4.36]	4.41 [2.78–6.45]
Clostridium sensu stricto	1.93 [0.57–2.2]^abc	4 [3.83–4.99]^b	2.09 [0.05–2.8]^b	3.72 [2.37–4.88]	4.1 [2.13–4.45]
Dubosiella	14.97 [11.94–18.62]^c	22.24 [18.27–24.5]	16.54 [10.28–17.82]	12.32 [10.91–15.63]	10.14 [5.51–14.1]^c
Faecalibaculum	0 [0–0]^c	0 [0–0]	0 [0–0]	0 [0–0]	0.05 [0.02–0.57]^c
Lachnospiraceae*	9.72 [4.37–13.61]^ab	15.35 [14.05–17.94]^b	15.65 [13.4–18.85]^b	12.31 [10.14–16.2]	11.08 [8.4–12.62]
Lactococcus	0 [0–0]	0 [0–0]	0 [0–0]	0 [0–0]	0 [0–0]
Muribaculaceae*	9.14 [8.12–10.53]	8.92 [5.72–9.8]	8.27 [7.02–10.22]	8.14 [6.3–8.28]	8.83 [6.69–9.86]
Paramuribaculum	7.93 [6.58–8.49]^ac	8.15 [7.26–9.94]	8.44 [5.95–9.27]	11.3 [9.97–12.48]	10.95 [9.58–12.02]^c
Parasutterella	1.37 [0.98–1.47]	1.63 [1.41–1.75]	1.08 [0.93–1.59]	1.29 [1.02–1.43]	0.92 [0.76–1.5]
Ruminococcaceae*	4.71 [3.67–5.45]	4.06 [3.63–4.85]	4.92 [4.41–7.08]	4.58 [3.79–5.57]	5 [3.45–7.58]

^aComparisons between all samples without AL vs AL samples.

^bComparisons between LD and LD-wdFMT samples compared to the AL samples.

^cComparisons between WD-ldFMT samples and AL samples.

*Not further classified to genus.

Table 5. Median AL index scores among all groups with interquartile range in brackets.

Group	AL Index
LD	0.59 [0.56–0.71]^a
LD-wdFMT	0.47 [0.4–0.63]
WD	0.53 [0.4–1.01]
WD-ldFMT	0.55 [0.45–0.72]^a
AL	0.31 [0.2–0.43]^a

^aSignificant difference found in Kruskal-Wallis pairwise comparison.

Data availability statement

Sequence data were deposited in the NCBI SRA under accession number SRP408098 and can be found at https://www.ncbi.nlm.nih.gov/sra/?term=SRP408098.