RESEARCH ARTICLE

Changes in the Mucosa-Associated Microbiome and

Transcriptome across Gut Segments Are Associated with
Obesity in a Metabolic Syndrome Porcine Model
Song-Song Xu,a,b Nan Wang,b Lei Huang,a Xiu-Ling Zhang,b,c Shu-Tang Feng,b Sha-Sha Liu,b Yue Wang,b Zhi-Guo Liu,b
Bing-Yuan Wang,b Tian-Wen Wu,b Yu-Lian Mu,b Shao-Hua Hou,b Kui Lia,b

a Shenzhen Branch, Guangdong Laboratory of Lingnan Modern Agriculture, Genome Analysis Laboratory of the Ministry of Agriculture and Rural Affairs, Agricultural
Genomics Institute at Shenzhen, Chinese Academy of Agricultural Sciences, Shenzhen, China
b State Key Laboratory of Animal Nutrition, Key Laboratory of Animal Genetics Breeding and Reproduction of Ministry of Agriculture and Rural Affairs of China, Institute of
Animal Sciences, Chinese Academy of Agricultural Sciences, Beijing, China
c College of Animal Science and Technology, Nanjing Agricultural University, Nanjing, China

Song-Song Xu, Nan Wang, and Lei Huang contributed equally to this work. Author order was determined by the corresponding author after negotiation.

ABSTRACT Several studies have suggested a role for gut mucosa-associated micro-
biota in the development of obesity, but the mechanisms involved are poorly
defined. Here, the impact of the gut mucosa-associated microbiota on obesity and
related metabolic disorders was evaluated in a metabolic syndrome (MetS) porcine
model. Body composition was determined among male Wuzhishan minipigs consum-
ing a high-energy diet (HED) and compared to that of those consuming a normal diet
(ND), and gut segments (duodenum, jejunum, ileum, cecum, colon, and rectum) were
sampled for paired analysis of mucosa-associated microbiota and transcriptome signa-
tures with 16S rRNA gene and RNA sequencing, respectively. Our data indicated that
long-term HED feeding significantly increased body weight and visceral fat deposition
and aggravated metabolic disorders. Specially, HED feeding induced mucosa-associ-
ated microbiota dysbiosis and selectively increased the abundance of the families
Enterobacteriaceae, Moraxellaceae, and Lachnospiraceae in the upper intestine. The associa-
tion analysis indicated that specific bacteria play key roles in adiposity, e.g., Lactobacillus
johnsonii in the duodenum, Actinobacillus indolicus in the jejunum, Acinetobacter johnsonii
in the ileum, Clostridium butyricum in the cecum, Haemophilus parasuis in the colon, and
bacterium NLAEzlP808, Halomonas taeheungii, and Shewanella sp. JNUH029 in the rectum.
Transcriptome data further revealed intestinal lipid metabolism and immune dysfunction
in the MetS individuals, which may be associated with obesity and related metabolic dis-
orders. Our results indicated that gut mucosa-associated microbiota dysbiosis has the
potential to exacerbate obesity, partially through modulating systemic inflammatory
responses.
IMPORTANCE Obesity is a major risk factor for metabolic syndrome, which is the Editor Wei-Hua Chen, Huazhong University of
Science and Technology
most common cause of death worldwide, especially in developed countries. The link
Copyright © 2022 Xu et al. This is an open-
between obesity and gut mucosa-associated microbiota is unclear due to challenges access article distributed under the terms of
associated with the collection of intestinal samples from humans. The current report the Creative Commons Attribution 4.0
International license.
provides the first insight into obesity-microbiome-gut immunity connections in a
Address correspondence to Kui Li,
metabolic syndrome (MetS) porcine model. The present results show that dysbiosis of [email protected], or Shao-Hua Hou,
mucosal microbiota along the entire digestive tract play a critical role in the proinflam- [email protected].
matory response in the host-microbial metabolism axis, resulting in obesity and related The authors declare no conflict of interest.
metabolic disorders in the MetS model. Received 28 February 2022
Accepted 4 June 2022
KEYWORDS obesity, mucosa-associated microbiota, 16S rRNA gene sequencing, Published 7 July 2022
transcriptome, Inflammatory responses

July/August 2022 Volume 10 Issue 4 10.1128/spectrum.00717-22 1

Obesity by Modulating Host-Gut Microbiota Interaction Microbiology Spectrum

O besity has been identified as a modifiable risk factor for death and loss of produc-
tive life years worldwide (1). Specifically, excessive body weight and visceral lipid
accumulation have been recognized as major characteristics of obesity which further
contribute to disturbed glucose and lipid metabolism (2). Increasing evidence supports
the role of the gut microbiota as a crucial player in the pathogenesis of diet-induced
obesity and related metabolic complications (3, 4). Development of obesity has been
associated with specific microorganisms and metabolites that lead to inflammatory and
immune reactions in the intestine (4). For example, obese patients consuming a high level
of processed and animal-derived foods have an increased abundance of Erysipelotrichaceae,
Ruminococcaceae species of the Blautia genus, and Streptococcus species (5). However, the
underlying obesity-microbiome-gut immunity interactions remain largely unknown due to
the challenges associated with such studies, such as the ethical and logistical constraints
involved in obtaining human intestinal tissues.
The bacterial community shows a distinct distribution along the mammalian gastro-
intestinal (GI) tract, both longitudinally (proximal to distal) and radially (mucosa to
lumen) (6). However, few studies have analyzed microbiota profiles in the gut proximal
regions or those living within the outer mucosal layer, which may be dissimilar to the
fecal microbiota. The microbiota colonizing the outer mucosal layer, which can inter-
face with the epithelial layer, may play a pivotal role in GI immune cell composition (7).
A recent study showed that gut mucosa-associated microbes including Bacteroidetes
and Erysipelotrichaceae in nonobese diabetic mice could be used as biomarkers for
type 1 diabetes development (8). Thus, characterization of the gut mucosal microbial
community may contribute to our understanding of host-microbiome interactions in
both healthy and disease states.
Moreover, long-term dietary interventions can provide a constant source of substrates
to continuously shape the gut ecosystem (5). For example, the long-term consumption of
a high-energy diet (HED) has been extensively studied as a major cause of obesity and
related metabolic diseases such as diabetes (types 1 and 2), Crohn’s disease, ulcerative coli-
tis, liver cirrhosis, and atherosclerosis (9–12). However, the long-term effects of specific
diets in humans remain largely unknown due to challenges such as controlling actual nu-
trient intake (13). Identification of appropriate animal models would therefore aid in
understanding the potential impact of dietary effects on the gut microbiome. Some stud-
ies have used a porcine model to investigate these interactions, because pigs and humans
have highly similar physiological activities, such as gut microbiome colonization and meta-
bolic and immune functions (14, 15). A previous study revealed that an intergenerational
pig model of dietary restriction provided an opportunity to understand which features in
the developing pig microbiome were causally linked to regulation of various growth pa-
rameters (16). Recently, a study of the metabolic syndrome (MetS) porcine model by our
group indicated that a long-term HED altered the microbiome of gut contents, decreasing
levels of butyrate-producing bacteria, including the genus Bacteroides and the families
Lachnospiraceae and Ruminococcaceae (17).
This study was conducted in the MetS porcine model to investigate alterations to the
gut mucosa-associated microbiota and their effects on obesity and related metabolic dis-
orders. Here, we reported changes in the gut mucosa-associated microbiota and intestinal
transcriptome in different segments of the GI tract and the correlation of microbiota with
metabolic parameters (body weight; liver, heart, and spleen weight; visceral lipid accumu-
lation; serum cholesterol levels). This study identified specific gut mucosa-associated
microbes that may influence obesity and related metabolic disorders.

RESULTS
Long-term HED exacerbates obesity and related metabolic disorders in the
MetS model. After 64 months of dietary interventions, fat accumulation in the viscera
(liver, heart, and spleen) was significantly greater (P , 0.05) in the HED feeding group
than in the ND feeding group (Table 1). Coupled with our previous report (17), these
results demonstrated that long-term HED feeding aggravated visceral fat accumula-
tion, serum lipid profiles, and systemic inflammation in the MetS porcine model.

July/August 2022 Volume 10 Issue 4 10.1128/spectrum.00717-22 2

Obesity by Modulating Host-Gut Microbiota Interaction Microbiology Spectrum

TABLE 1 Metabolic parameters by diet group

ND feeding group HED feeding group
Metabolic parametera (mean ± SEM) (mean ± SEM)b P valuec
Body wt (kg) 50.44 6 4.18B 78.49 6 5.35A 0.002
Perirenal fat (g) 75.98 6 10.92B 142.61 6 19.37A 0.022
Omentum (g) 52.75 6 7.33B 177.42 6 32.91A 0.010
Leaf fat (g) 200.02 6 45.33B 966.24 6 227.57A 0.018
Liver (g) 656.67 6 73.24B 970.00 6 83.52A 0.021
Heart (g) 177.77 6 13.41B 236.58 6 15.31A 0.018
Spleen (g) 67.27 6 11.88B 153.64 6 20.38A 0.007
TC (mmol/l) 1.79 6 0.15B 3.63 6 0.27A 0.001
HDL (mmol/l) 1.04 6 0.06B 1.94 6 0.15A 0.002
LDL (mmol/l) 1.30 6 0.10B 2.44 6 0.21A 0.004
TGs (mmol/l) 0.32 6 0.03 0.37 6 0.07 0.653
aTC, total cholesterol; HDL, high-density lipoprotein; LDL, low-density lipoprotein; TGs, triglycerides.
bCapital letters indicate statistically significant differences between the HED and ND groups (P , 0.05). SEM,
standard error of the mean.
cP values were calculated with Student’s t test (n = 6 to 9 per group). HED, high-energy diet; ND, normal diet.

Gut mucosa-associated microbial diversity was decreased across different gut

segments in the MetS model. We sequenced a total of 92 intestinal mucosal speci-
mens along the entire digestive tract (including duodenum, jejunum, ileum, cecum, co-
lon, and rectum), using bacterial 16S rRNA gene sequencing, in the HED group (n = 11)
and the ND group (n = 6). The mucosa-associated microbial diversity was analyzed in
the six individual gut segments (Fig. 1A). Rarefaction curves revealed that the esti-
mated operational taxonomic unit (OTU) richness nearly reached saturation in each
region of the GI tract (see Fig. S1A in the supplemental material). Generally, the HED
feeding group had lower alpha (within-sample) diversity at the OTU level than the ND
feeding group as measured by the Shannon and Simpson indices (Fig. 1B). We calcu-
lated the bacterial copy number per gram of gut mucosa-associated contents using
quantitative real-time PCR (qPCR) in the duodenum, jejunum, ileum, cecum, and colon.
The HED feeding group showed highly significantly (P , 0.01) lower copy numbers per
gram of total bacteria compared to the ND feeding group (see Fig. S1B). Nonmetric
multidimensional scaling (NMDS) with taxonomic information showed a clear separa-
tion of microbial community structure between the two feeding groups (Fig. 1C; see
also Fig. S2 in the supplemental material). As shown in the NMDS plot, the small intes-
tine samples clustered closely together, whereas large intestine samples clustered less
closely on NMDS1 (Fig. 1C), indicating more diverse bacterial composition in the large
intestine. Furthermore, 27.57% (1,851 OTUs) were shared among the small intestine
samples, whereas 23.69% (1,913 OTUs) were shared among the large intestine samples
(Fig. 1D).
Gut mucosa-associated microbiota dysbiosis across different gut segments in
the MetS model. Analysis at the phylum level indicated that the gut mucosa-associ-
ated microbiota was dominated by five major phyla among the different gut segments:
Firmicutes, Proteobacteria, Bacteroidetes, Actinobacteria, and Fusobacteria, together con-
stituting up to 93% of the OTUs in each gut segment, on average (Fig. 1E). Notably, lev-
els of the major bacterial phyla Firmicutes and Proteobacteria were drastically different
from the proximal to distal regions of the GI tract in the MetS model (see Fig. S3). In
the rectum, MetS animals showed a significant decrease in Firmicutes (P = 0.028) and
increase in Proteobacteria (P = 0.009) compared with the ND feeding group. The aver-
age bacterial community compositions at the family level are shown in Fig. 1F. We also
compared differences in the microbial communities at the family, genus, and species
levels across the GI tract between the groups. The HED-fed pigs had increased coloni-
zation of several families, including Lactobacillaceae in the duodenum and cecum,
Enterobacteriaceae in the jejunum, Moraxellaceae in the ileum, and Lachnospiraceae in
the colon (Table 2). However, 12 families were significantly enriched in the ND feeding

July/August 2022 Volume 10 Issue 4 10.1128/spectrum.00717-22 3

Obesity by Modulating Host-Gut Microbiota Interaction Microbiology Spectrum

FIG 1 Mucosa-associated microbiome analysis of the metabolic syndrome porcine model with 16S rRNA gene sequencing. (A) Outline of the pig
gastrointestinal tract with delineation of gut segments sampled for analyses. (B) The mucosa-associated microbial alpha diversity analysis at the OTU level,
illustrated by Shannon and Simpson indices. (C) Beta diversity was evaluated with nonmetric multidimensional scaling (NMDS) using Bray-Curtis
dissimilarity. (D) Venn diagrams showing shared OTUs in the small and large intestine, respectively. (E and F) Taxonomic summary of top 10 phyla (E) and
top 15 families (F) (by abundance), represented by mean values per group and gut section. HED, high-energy diet; ND, normal diet.

group compared with the HED feeding group (P , 0.05): Bacteroidaceae in the jejunum,
Erysipelotrichaceae and Fusobacteriaceae in the ileum, Bacteroidales S247 group and
Streptococcaceae in the cecum, Ruminococcaceae, Spirochaetaceae, and Rikenellaceae in the
colon, and Lactobacillaceae, Ruminococcaceae, Lachnospiraceae, Erysipelotrichaceae, and
Halomonadaceae in the rectum.
At the genus level, only 23 genera were significantly different in small intestinal regions
between the treatment groups (P , 0.05) (Fig. 2). In the duodenal region, the genera
Bacillus, Faecalibacterium, Lactobacillus, Ruminococcaceae NK4A214 group, Ruminococcus 1,
and Stenotrophomonas were more abundant in the HED feeding group, whereas the gen-
era Allobaculum and Blautia were less abundant. In the jejunal region, Escherichia-Shigella
and Mitsuokella were more abundant and Actinobacillus, Bacteroides, and Comamonas
were less abundant in the HED feeding group. Additionally, Acinetobacter, Pasteurella, and
Ruminococcaceae UCG-014 were more abundant in the ileum region of the HED feeding
group than the ND feeding group. However, Anaerotruncus, Coprococcus 3, Desulfovibrio,
Faecalibacterium, Fusobacterium, Lachnoclostridium, Ruminococcus gauvreauii group, and
Solobacterium were less abundant in the HED feeding group. In contrast to the small intes-
tine, 39 genera were altered across the large intestinal regions. In the cecal luminal region
of the HED feeding group, Erysipelotrichaceae UCG-001, Lactobacillus, and Prevotella 9
were significantly more abundant and 17 genera (including Turicibacter, Streptococcus,

July/August 2022 Volume 10 Issue 4 10.1128/spectrum.00717-22 4

Obesity by Modulating Host-Gut Microbiota Interaction Microbiology Spectrum

TABLE 2 Relative abundances of mucosa-associated bacterial families detected in the HED

and ND groups
Mean % abundance ± SEMa

Intestinal segment Family HED ND P valueb

Duodenum Lactobacillaceae 22.92 6 2.16A 12.83 6 0.75B 0.002
Jejunum Enterobacteriaceae 23.33 6 2.91A 5.51 6 1.08B 0.014
Bacteroidaceae 0.20 6 0.10B 4.93 6 3.34A 0.027
Ileum Erysipelotrichaceae 4.33 6 0.65B 7.90 6 1.27A 0.019
Moraxellaceae 2.83 6 1.61A 0.20 6 0.09B 0.028
Fusobacteriaceae 0.30 6 0.17B 3.20 6 2.41A 0.042
Caecum Lactobacillaceae 44.44 6 7.73A 18.73 6 3.89B 0.020
Bacteroidales S247 group 1.51 6 0.55B 4.36 6 1.32A 0.028
Streptococcaceae 1.21 6 0.62B 3.69 6 1.06A 0.039
Colon Ruminococcaceae 14.72 6 0.78B 16.97 6 0.43A 0.047
Lachnospiraceae 12.35 6 0.38A 9.51 6 0.91B 0.016
Spirochaetaceae 0.68 6 0.09B 2.11 6 0.35A 0.016
Rikenellaceae 0.80 6 0.13B 1.69 6 0.14A 0.009
Rectum Lactobacillaceae 5.38 6 3.08B 20.00 6 4.46A 0.016
Ruminococcaceae 5.69 6 0.77B 11.53 6 1.83A 0.028
Lachnospiraceae 4.27 6 0.31B 7.11 6 0.91A 0.028
Erysipelotrichaceae 0.93 6 0.11B 3.37 6 0.83A 0.009
Halomonadaceae 0.45 6 0.10B 2.03 6 0.41A 0.028
letters indicate statistically significant differences between HED and ND feeding groups (P , 0.05).
aCapital

bP
values were determined by Mann-Whitney U test (n = 5 to 11 per diet group). HED, high-energy diet; ND,
normal diet.

Ruminococcus 2, Ruminiclostridium 6, and Halomonas) were less abundant in the HED

feeding group (P , 0.05). In the colonic regions, Lachnospiraceae NK4A136 group,
Roseburia, Haemophilus, Desulfovibrio, Campylobacter, and Ruminiclostridium 9 were sig-
nificantly more abundant in the HED feeding group than in the ND feeding group
(P , 0.05). However, Rikenellaceae RC9 gut group, Treponema 2, Ruminococcaceae UCG-
010, Ruminiclostridium 5, Ruminococcaceae UCG-013, and Ruminococcaceae UCG-004
were less abundant in the HED feeding group. In the rectal region of the HED feeding
group, Planomicrobium and Burkholderia were more abundant (P , 0.05) than in the ND
feeding group. Additionally, 13 genera (including Lactobacillus, Turicibacter, Halomonas,
Terrisporobacter, and Streptococcus) were less abundant in the HED feeding group than
in the ND feeding group.
The two treatment groups also exhibited marked changes in mucosa-associated micro-
biota composition at the species level across all luminal regions (Table 3). Lactobacillus
johnsonii showed the largest difference in abundance in the luminal regions of the duode-
num, ileum, and rectum. Interestingly, Lactobacillus johnsonii was most abundant in the
duodenal lumen of the HED feeding group (12.60%) and then steadily decreased through
the GI tract to the rectal lumen, where it reached 1.9%. Furthermore, several groups were
more abundant in the HED feeding group, namely, Stenotrophomonas maltophilia in the
duodenum, Shigella flexneri and Shigella sonnei in the jejunum, Acinetobacter johnsonii and
bacterium NLAE-zl-P808 in the ileum, Lactobacillus reuteri in the cecum, Haemophilus para-
suis and Actinobacillus porcinus in the colon, and bacterium NLAE-zl-P808, Enterobacteriaceae
bacterium_X2/SB59, Klebsiella oxytoca, Pantoea sp. FSGSA15, Planococcus sp. BM-G6,
Planomicrobium okeanokoites, and Pseudomonas putida in the rectum. Collectively, these
results indicated that the long-term HED modulated the mucosa-associated microbial
community structure, resulting in a bacterial composition dissimilar from that of the ND
group.
Correlations between gut mucosa-associated microbiota composition and obesity
and related metabolic disorders. We next focused on the association between mucosa-
associated bacterial species and the markers of obesity and related metabolic

July/August 2022 Volume 10 Issue 4 10.1128/spectrum.00717-22 5

Obesity by Modulating Host-Gut Microbiota Interaction Microbiology Spectrum

FIG 2 Changes in mucosa-associated microbes at the genus level in different regions of the GI tract in
the metabolic syndrome (MetS) porcine model. *, P , 0.05 (Mann-Whitney U test).

complications. Analysis based on Spearman’s correlation revealed that the abundance of

Lactobacillus johnsonii in the duodenum, Shigella sonnei in the jejunum, Acinetobacter john-
sonii in the ileum, Lactobacillus reuteri in the cecum, Haemophilus parasuis in the colon,
and bacterium_NLAEzlP808 and Pantoea sp. FSGSA15 in the rectum were positively associ-
ated with the measured metabolic parameters (Fig. 3). However, Actinobacillus indolicus in
the jejunum, Lactobacillus johnsonii in the ileum, Clostridium butyricum in the cecum, and
Halomonas taeheungii, Lactobacillus johnsonii, and Shewanella sp. JNUH029 in the rectum
were negatively associated with obesity and related metabolic disorders (Fig. 3). These
results suggested that HED-induced obesity was associated with significant changes in the
mucosa-associated microbiota composition.
Intestinal functions were altered across different gut segments in the MetS
model. We analyzed the RNA sequencing (RNA-seq) data from porcine intestinal tis-
sues, which were generated for each sample of the HED (n = 4) and ND (n = 4) feeding
groups. RNA-seq-based transcriptome profiling revealed a total of 4,889,868,618 raw
reads and a total of 4,587,375,920 clean reads after quality control (see Table S1 in the
supplemental material). For each sample, the clean reads comprised between 87.10%
and 96.72% of the raw reads. Among the clean reads, the percentage of reads with
Q30 values ranged from 87.11% to 95.09%, and the average GC content was 45.79%.
These results suggested that the RNA-seq data were of good quality.

July/August 2022 Volume 10 Issue 4 10.1128/spectrum.00717-22 6

Obesity by Modulating Host-Gut Microbiota Interaction Microbiology Spectrum

TABLE 3 Relative abundances of mucosa-associated bacterial species detected in the HED and ND groups
Mean % abundance ± SEMa

Intestinal segment Species HED ND P valueb

Duodenum Lactobacillus johnsonii 0.126 6 0.011A 0.068 6 0.007B 0.020
Stenotrophomonas maltophilia 0.003 6 0.0005A 0.001 6 0.0005B 0.020
Jejunum Shigella flexneri 0.182 6 0.036A 0.033 6 0.009B 0.014
Shigella sonnei 0.007 6 0.003A 0.001 6 0.0006B 0.014
Actinobacillus indolicus 0.0003 6 0.00003B 0.002 6 0.001A 0.014
Actinobacillus rossii 0.0002 6 0.0001B 0.002 6 0.0005A 0.014
Ileum Acinetobacter johnsonii 0.002 6 0.0006A 0.0004 6 0.0002B 0.028
Fusobacterium ulcerans 0.0004 6 0.0002B 0.002 6 0.001A 0.034
Lactobacillus johnsonii 0.040 6 0.009B 0.083 6 0.0147A 0.042
Bacterium NLAE-zl-P808 0.001 6 0.0002A 0.0006 6 0.0001B 0.042
Cecum Shewanella algae 0.002 6 0.0007B 0.005 6 0.0006A 0.014
Halomonas sp. Betam1 0.0009 6 0.0004B 0.003 6 0.0003A 0.014
Helicobacter rappini 0.0006 6 0.0005B 0.002 6 0.0002A 0.020
Lactobacillus reuteri 0.103 6 0.025A 0.027 6 0.007B 0.039
Clostridium butyricum 0.0005 6 0.0002B 0.002 6 0.001A 0.045
Colon Haemophilus parasuis 0.002 6 0.0005A 0.0003 6 0.0002B 0.028
Actinobacillus porcinus 0.002 6 0.0004A 0.0005 6 0.0002B 0.028
Rectum Bacterium NLAE-zl-P808 0.001 6 0.0007A 0.00006 6 0.00002B 0.008
Enterobacteriaceae bacterium X2/SB59 0.003 6 0.002A 0.0001 6 0.00009B 0.028
Halomonas taeheungii 0.001 6 0.0002B 0.005 6 0.001A 0.028
Klebsiella oxytoca 0.004 6 0.0009A 0.001 6 0.0006B 0.047
Lactobacillus fermentum 0.0001 6 0.00004B 0.001 6 0.0003A 0.016
Lactobacillus johnsonii 0.019 6 0.012B 0.114 6 0.029A 0.016
Pantoea sp. FSGSA15 0.006 6 0.005A 0.0002 6 0.0002B 0.035
Planococcus sp. BM-G6 0.001 6 0.0008A 0.00005 6 0.00005B 0.031
Planomicrobium okeanokoites 0.003 6 0.002A 0.0002 6 0.0002B 0.047
Pseudomonas putida 0.014 6 0.012A 0.0005 6 0.0005B 0.045
Shewanella sp. JNU-H029 0.002 6 0.0005B 0.010 6 0.002A 0.009
aCapital letters indicate statistically significant differences between HED and ND groups (P , 0.05).
bP values were determined by Mann-Whitney U test (n = 5 to 11 per diet group). HED, high-energy diet; ND, normal diet.

We identified long noncoding RNAs (lncRNAs) and mRNAs in both the HED and ND
feeding groups. The majority of identified lncRNAs were located in intergenic noncod-
ing regions (see Fig. S4A in the supplemental material). The transcript length, exon
length, open reading frame (ORF) length, and conservation score of lncRNAs (including
both annotated and novel lncRNAs) and mRNAs were compared (see Fig. S4B, C, and
D). The novel lncRNAs were found to be significantly shorter in transcript length and
ORF length and to have fewer exons than mRNAs, consistent with general characteris-
tics of known lncRNAs. The distributions of exon numbers and ORF lengths showed
similar patterns in the annotated and novel lncRNAs. However, the novel lncRNAs were
less conserved than protein-coding transcripts, as determined with phastCon (see Fig.
S4E); this finding is consistent with a previous report (18).
We defined differentially expressed genes (DEGs) and differentially expressed lncRNAs
(DELs) as those with a false-discovery rate (FDR) of ,0.05 and jlog2(fold change)j of .0.6.
Compared to the ND feeding group, there were several hundred DEGs in the small intes-
tine for the HED feeding group; these comprised 132, 698, and 200 upregulated and 461,
288, and 231 downregulated genes in the duodenum, jejunum, and ileum, respectively
(Fig. 4A, B, and C; see also Tables S2 to S4 in the supplemental material). There were fewer
DEGs in the large intestine for the HED feeding group, including 112, 37, and 616 upregu-
lated and 124, 154, and 225 downregulated genes in the cecum, colon, and rectum,
respectively (Fig. 4D, E, and F; see also Tables S5 to S7). Individuals in the HED feeding
group displayed significant changes in several biological signaling pathways. In the small
intestine, the top pathways associated with the DEGs were related to fat and protein me-
tabolism, bile secretion, pathogen recognition, and inflammatory pathways (Fig. 4G, H,

July/August 2022 Volume 10 Issue 4 10.1128/spectrum.00717-22 7

Obesity by Modulating Host-Gut Microbiota Interaction Microbiology Spectrum

FIG 3 Spearman’s correlation between mucosa-associated microbes and obesity-related metabolic parameters. *, P , 0.05; **, P , 0.01 (two-
tailed Student's t test).

and I). In the large intestine, the top pathways associated with DEGs in the HED feeding
group were primarily related to fatty acid, protein, and carbohydrate metabolism and
immune system pathways (Fig. 4J, K, and L).
We also identified DELs in all of the luminal regions in the HED feeding group: 104
in the duodenum, 1,664 in the jejunum, 58 in the ileum, 50 in the cecum, 34 in the co-
lon, and 141 in the rectum (Fig. 5A to F). To investigate the potential functions of the
DELs, we performed lncRNA-mRNA coexpression pair analysis. After filtering, the coex-
pressed lncRNA-mRNA pairs with high (.0.9) correlation coefficients were selected,
including 20 DELs and 1,186 DEGs in the duodenum, 615 and 1,714 in the jejunum, 6
and 1,072 in the ileum, 7 and 110 in the cecum, 4 and 33 in the colon, and 23 and 99
in the rectum (see Tables S8 to S13 in the supplemental material). To understand the
main gene functions at a more global level, we functionally annotated the DEGs identi-
fied from DELs in the six intestinal segments. The top pathways among all tissues were
mainly associated with lipid and steroid metabolism and with immune and inflamma-
tory responses (Fig. 5G to L).

DISCUSSION
The gut contains a complex and dynamic bacterial community with great potential
to influence host health (2, 5). Specifically, the microbiota colonizing the outer mucus
layer has critical roles in bacterial-triggered host immune activation and metabolic dis-
orders (15, 19). Thus, the functional characteristics of mucosa-associated microbial
communities are key to our understanding of host-microbiome interactions in both
healthy and disease states. In our study, all animals were housed in same environmen-
tal conditions to ensure that changes in gut microbiota composition could be attrib-
uted to diet-specific effects rather than environmental influences. After 64 months of
high-energy diet, we successfully established a MetS porcine model characterized by
increased body weight, serum lipid, and proinflammatory cytokine levels, visible ather-
omatous plaque on abdominal aorta, accumulated lipid droplets and enhanced apo-
ptosis in hepatocytes, and impaired intestinal epithelial integrity (17). In our previous
study (17), we showed that diet composition impacted gut microbiota structure and

July/August 2022 Volume 10 Issue 4 10.1128/spectrum.00717-22 8

Obesity by Modulating Host-Gut Microbiota Interaction Microbiology Spectrum

FIG 4 Differentially expressed mRNAs (DEGs) in intestinal tissues of individuals in the HED group compared with those in the ND group. (A to F) The
volcano plot shows downregulated (blue), upregulated (red), and unchanged (gray) DEGs in the duodenum (A), jejunum (B), ileum (C), cecum (D), colon (E),
and rectum (F). (G to L) Top canonical KEGG pathways in the duodenum (G), jejunum (H), ileum (I), cecum (J), colon (K), and rectum (L). Red represents
enriched pathways and blue represents depleted pathways.

July/August 2022 Volume 10 Issue 4 10.1128/spectrum.00717-22 9

Obesity by Modulating Host-Gut Microbiota Interaction Microbiology Spectrum

FIG 5 Differentially expressed lncRNAs (DELs) in intestinal tissues of individuals in the HED group compared to those in tissues from the ND group. (A to F)
The volcano plot shows downregulated (blue), upregulated (red), and unchanged (gray) DELs in the duodenum (A), jejunum (B), ileum (C), cecum (D),
colon (E), and rectum (F). (G to L) Top canonical KEGG pathways of dysregulated mRNAs in coexpressed lncRNA-mRNA pairs with high correlation
coefficients in the duodenum (G), jejunum (H), ileum (I), cecum (J), colon (K), and rectum (L). Red represents enriched pathways and blue represents
depleted pathways.

July/August 2022 Volume 10 Issue 4 10.1128/spectrum.00717-22 10

Obesity by Modulating Host-Gut Microbiota Interaction Microbiology Spectrum

that the high-energy diet pattern decreased the abundance of short-chain fatty acid-
producing bacteria, including Bacteroides, Lachnospiraceae, and Ruminococcaceae. In
this study, we found that the high-energy diet also decreased gut mucosa-associated
microbial diversity and disrupted its structure along the entire digestive tract. Notably,
the metabolic parameters were strongly correlated with the abundance of gut mu-
cosa-associated microbiota in the MetS model. Coupled with gut transcriptome analy-
sis, these findings indicated that gut mucosa-associated microbiota dysbiosis might
exacerbate obesity and related metabolic disorders in the MetS model.
Recently, several studies have implicated mucosa-associated microbiota dysbiosis in
the pathogenesis of obesity and related metabolic disorders. A study of the duodenal mu-
cosal microbiota of patients with intestinal metaplasia showed an inverse relationship
between microbial diversity and metabolic diseases (20). Another study suggested that
patients with morbid obesity had lower jejunal mucosa-associated microbial diversity than
healthy controls (21). We here confirmed previous reported results demonstrating that
HED-induced obese minipigs had decreased mucosa-associated microbial diversity and
richness in the GI tract. Our results showed that the mucosa-associated microbiota along
the longitudinal axis of the small and large intestine was dominated by Firmicutes and
Proteobacteria. The abundance level of Firmicutes changed drastically along the length of
the intestine, reaching the lowest levels in the rectal lumen of the HED feeding group. In
contrast, Proteobacteria were presented at relatively low levels in the cecum and colon,
steadily increased in abundance along the length of the large intestine, and reached the
highest levels in the rectum. Interestingly, reduced abundance of Firmicutes and high prev-
alence of Proteobacteria have previously been linked to high-polysaccharide diets and lipo-
polysaccharide production in humans and other mammals (13, 19, 22).
We found that specific Firmicutes and Proteobacteria were significantly affected by
dietary interventions across the GI tract. Consistent with a previous report showing
greater Lactobacillaceae abundance in Heligmosomoides polygyrus-infected C57BL/6
mice (23), we observed elevated levels of Lactobacillus and Lactobacillus johnsonii in
the duodenum of HED-fed animals. Lactic acid bacteria are representative probiotics
that have demonstrated beneficial effects, such as improvement in epithelial barrier
function, immunity enhancement, and anti-inflammatory activities (24). Interestingly, a
higher number of Lactobacillaceae species (Lactobacillus johnsonii and Lactobacillus
reuteri) colonized the upper intestine (duodenum and cecum) of the HED feeding
group, which led to mucosal protection and anti-inflammatory effects (5). However,
Lactobacillaceae species (Lactobacillus johnsonii and Lactobacillus fermentum) were less
abundant in the rectal region of the HED feeding group individuals, consistent with an
increase in proinflammatory metabolites. Additionally, the HED feeding group showed
elevated levels of both jejunal Enterobacteriaceae (Escherichia-Shigella, Shigella flexneri,
and Shigella sonnei) and ileal Moraxellaceae (Acinetobacter and Acinetobacter johnsonii)
compared with the ND feeding group. Enterobacteriaceae is a family of Gram-negative
bacteria that has previously been linked to obesity and hepatic damage (25).
Moraxellaceae, a family of Proteobacteria, colonizes mucosal membranes and is associ-
ated with hepatic steatosis (26). Moreover, the abundance of the Halomonadaceae
family (Halomonas and Halomonas taeheungii) was lower in the HED feeding group, a
result that was also observed in nursery pigs on a high-fat diet (27). Our findings sug-
gest that the mucosa-associated microbiota serves as a link between diet and disease
risk by modulating pro- and anti-inflammatory responses.
Previous studies have reported that obesity was associated with chronic systemic
inflammation (28). In the current study, Lactobacillus johnsonii in the duodenal region
showed significant positive correlations with obesity and associated metabolic disor-
ders, which was likely related to anti-inflammatory responses. However, we found that
Lactobacillus johnsonii was inversely correlated with metabolic parameters in the cecal
and rectal luminal regions of the HED feeding group. Lactobacillus johnsonii has been
shown to have an antiobesity effect by preventing inflammation and mucosal barrier
disruption in the gut (29). Moreover, anti-inflammatory-associated bacteria, including

July/August 2022 Volume 10 Issue 4 10.1128/spectrum.00717-22 11

Obesity by Modulating Host-Gut Microbiota Interaction Microbiology Spectrum

jejunal Actinobacillus indolicus, cecal Clostridium butyricum, and rectal Halomonas tae-
heungii and Shewanella sp. JNUH029, were negatively correlated with visceral fat
deposition and serum lipid levels, indicating stimulation of proinflammatory metabo-
lites following long-term HED feeding. Further, inflammatory bowel disease-associ-
ated bacteria, including ileal Acinetobacter johnsonii, colonic Haemophilus parasuis,
and rectal bacterium NLAEzlP808, were positively correlated with visceral fat levels.
Consequently, in the MetS model, enrichment of intestinal mucosa-associated patho-
genic bacteria induced systemic inflammation, which was related to obesity and
related metabolic disorders.
Here, transcriptome analysis demonstrated that, compared to results with the ND feed-
ing group, HED feeding altered expression of genes primarily involved in lipid metabolism
and inflammatory responses in intestinal tissues. For example, pathway analysis of upregu-
lated genes in the small intestine revealed that fat digestion and absorption, the comple-
ment system, and inflammatory pathways were enriched in the HED feeding group. The
microbiota colonizing the small intestinal outer mucus layer is mainly responsible for lipid
metabolism and IgA production, which are associated with metabolic disorders (30).
Therefore, mucosa-associated microbiota dysbiosis may influence intestinal gene expres-
sion (7). Previous studies have demonstrated that the complement system plays a key role
in maintaining host immunosurveillance and tissue homeostasis by regulating the elimina-
tion of pathogens (31). Additionally, enriched signaling pathways among upregulated
genes in the small intestine, including cytochrome P450, proliferator-activated receptor
(PPAR), and T-cell receptor signaling pathways, highlighted the elevated inflammatory
state in the MetS model. Furthermore, large intestine transcriptome data revealed the acti-
vation of pathways involved in pyruvate and fatty acid metabolism and inflammation,
including AMP-activated protein kinase and PPAR signaling, in the MetS model. Of note,
both the small and large intestine showed upregulation of genes in the PPAR signaling
pathway, which regulates expression of a large variety of genes involved in lipid and car-
bohydrate metabolism (32). The intestinal PPAR signaling pathway has also been consid-
ered an important molecular pathway for shaping gut immune responses to bacterial load
and diet by regulating the recruitment and activity of various cell populations in both the
innate and the adaptive immune systems (33). We previously reported decreased expres-
sion of genes related to the intestinal tight junction in the MetS model (17), leading to
increased permeability of the gut barrier to endotoxins, pathogenic bacteria, and other
antigens. The results of the current study suggest that mucosal microbiota dysbiosis is
likely to exacerbate obesity and related metabolic disorders by regulating organic nutrient
metabolism and host inflammatory pathways (Fig. 6).
In conclusion, we showed here striking differences in the mucosa-associated micro-
biome and transcriptome profiles along the entire length of the GI tract that were asso-
ciated with obesity and related metabolic disorders in the MetS porcine model.
Interactions between the mucosal microbiota and metabolic parameters were ana-
lyzed to interpret the etiological mechanism of obesity. We identified differentially
expressed genes in intestinal tissues that were potentially reflective of metabolic disor-
ders and higher inflammation in the HED feeding group. Thus, our results indicated
that mucosal microbiota dysbiosis along the entire digestive tract promoted obesity,
which might occur partially through amplification of systemic inflammatory responses.

MATERIALS AND METHODS

Study design and sample collection. Animal maintenance and experimental treatments were con-
ducted in accordance with the guidelines of the Animal Care and Use Committee of the Germplasm
Resource Center of Chinese Experimental Minipigs. The experimental design was previously described
by Xu et al. (17). Briefly, male Wuzhishan minipigs were randomly assigned to two groups: a normal diet
(ND) group, which was fed a normal chow diet, and a high-energy diet (HED) group fed a high level of
saturated fat and cholesterol. The HED formula comprised 3% cholesterol, 10% fat (lard), and 87% base
material (48% corn, 20% wheat, 15% soybean cake, 12% rice bran, and 5% fish meal). All randomly
selected individuals were euthanized at 64 months of age. We collected epithelium scrapings from GI
tracts (HED group, n = 11; ND group, n = 6) that were separated into six sections of tissue: duodenum, je-
junum, ileum, cecum, colon, and rectum. Additionally, the intestinal tissues (six sections) were collected

July/August 2022 Volume 10 Issue 4 10.1128/spectrum.00717-22 12

Obesity by Modulating Host-Gut Microbiota Interaction Microbiology Spectrum

FIG 6 Impact of gut mucosa-associated microbiota on local and distant organs contributes to obesity development and progression. A
long-term high-energy diet can cause dysbiosis of mucosa-associated microbiota across gut segments, which leads to obesity
development by altering gene expression in the small and large intestine, exacerbating hyperlipemia, and triggering low-grade chronic
inflammation.

from the HED and ND treatment groups. Samples were immediately snap-frozen in liquid nitrogen and
stored at 280°C prior to further analysis.
Metabolic characterization. Total body and liver weights were measured in the two treatment
groups as previously reported (17). Additionally, internal organ (heart and spleen) and visceral fat (peri-
renal fat, omentum, and leaf fat) weights were also measured at the endpoint of the dietary intervention
study (64 months) in the two groups. Pathological examination of the liver and serum lipid profiles were
also performed during month 64 of the treatment phase as previously described (17). All metabolic pa-
rameters are listed in Table 1.
Gut mucosa-associated bacterial DNA extraction and 16S rRNA sequencing. Gut mucosa-associ-
ated bacterial genomic DNA was extracted using a DNA extraction kit (TIANGEN Biotech, Beijing, China).
For each sample, total DNA concentration and purity were measured using the NanoDrop One spectropho-
tometer (Thermo Fisher Scientific, Waltham, MA, USA) at 260 and 280 nm. Extracted DNA was immediately
stored at 280°C. The V3-V4 hypervariable regions were amplified using universal primers with barcodes. 16S
DNA libraries were recovered using a GeneJET gel extraction kit (Thermo Fisher Scientific) and quantified
using the Qubit 2.0 fluorometer (Thermo Fisher Scientific). Purified DNA libraries were generated and index
codes added using the NextR Ultra DNA library prep kit for Illumina (New England Biolabs [NEB], Ipswich, MA,
USA) following the manufacturer’s instructions. Finally, the libraries were sequenced on an Illumina HiSeq
2500 platform with 250-bp paired-end reads using the standard protocol (Illumina, San Diego, CA, USA). Total
bacterial copy numbers were quantified by qPCR as described by Bi et al. (34).
Operational taxonomic unit clustering and microbial diversity and taxonomic analyses. The
raw reads were filtered to obtain clean reads by removing adapters pollution and low-quality reads with
Trimmomatic v0.36 (35). The sample numbers retained in the working data set for further analysis were
as follows: 16 duodenum (10 HED and 6 ND individuals); 9 jejunum (5 HED and 4 ND individuals); 12 il-
eum (7 HED and 5 ND individuals); 14 cecum (8 HED and 6 ND individuals); 10 colon (5 HED and 5 ND
individuals); and 10 rectum (5 HED and 5 ND individuals). Possible chimeras were identified with
UCHIME (http://www.drive5.com/usearch/manual/uchime_algo.html). Denoised sequences were clus-
tered using USEARCH v10.0 (http://www.drive5.com/usearch/manual/uchime_algo.html), and those with
similarity of $97% were classified as OTUs by using the mothur pipeline (36). Taxonomy was assigned
with uclust in QIIME (v1.9.1; http://qiime.org/index.html) and the Silva database. Nonmetric multidimensional
scaling (NMDS) analysis was conducted and NMDS plots, taxonomy, and heatmaps were visualized using R
v4.0.5 (https://cran.r-project.org/).
Gut tissue RNA extraction and transcriptome sequencing. RNA was isolated from intestinal tissues
using TRIzol reagent (Invitrogen, Shanghai, China) following the manufacturer’s standard instructions, then
treated with RNase-free DNase I (TaKaRa, Shanghai, China) to remove residual genomic DNA. The RNA

July/August 2022 Volume 10 Issue 4 10.1128/spectrum.00717-22 13

Obesity by Modulating Host-Gut Microbiota Interaction Microbiology Spectrum

concentration and purity were measured using a Qubit RNA HS assay kit on a Qubit 2.0 fluorometer (Life
Technologies, Grand Island, NY, USA), and rRNA was depleted using a Ribo-zero rRNA removal kit (Epicentre,
Madison, WI, USA). cDNA libraries were generated with the NEBNext Ultra directional RNA library prep kit for
Illumina (NEB) following the manufacturer’s protocol. The library cDNA was amplified by PCR and validated
for fragment size and quantity by using the Qubit 2.0 fluorometer (Life Technologies). Finally, the libraries
were sequenced on the Illumina HiSeq 2500 platform to obtain 150-bp paired-end reads.
Transcriptome sequence analysis. Raw sequencing reads were quality checked using FastQC
v.0.11.7 (37). The index adaptors and low-quality bases (Q , 20) were trimmed to a minimum of 36 bp
using Trimmomatic v0.36 (35). Clean reads were aligned to the Sus scrofa genome (https://www.ncbi.nlm
.nih.gov/genome/?term=pig) using TopHat v2.0.10 with default parameters (38), and only reads uniquely
aligned to known genes were used for further analysis. Cufflinks v2.1.1 (38) was used to assemble novel
lncRNAs, annotated lncRNAs, and annotated mRNA transcripts individually using the default parameters.
The coding probability was calculated for novel transcripts, which were retained if they met the following
criteria: coding probability score of ,0.5 in CPC v0.9-r2 (39) and CPAT (40) and identified as noncoding
with CNCI v2 (41). Novel lncRNAs were defined as those that met the above criteria, were longer than
200 bp, and had at least two exons. The conservation levels for lncRNAs and mRNAs were evaluated with
8-way PhastCons scores (42). The expression levels of lncRNAs and mRNAs were calculated in fragments
per kilobase of transcript per million mapped reads (FPKM) using Cuffquant v2.1.1 (38). The differentially
expressed lncRNAs (DELs) and mRNAs (DEGs) between HED and ND groups were identified using the
Bayes-regularized t test with FDR correction using Cyber-T bayesreg (43). Those with an FDR of ,0.05 and
jlog2(fold change)j of .0.6 were considered statistically significant. To identify significantly enriched bio-
logical pathways among DEGs, enrichment analyses were conducted using the KEGG Pathway database
(https://www.genome.jp/kegg/pathway.html).
Statistical analyses. Statistically significant differences between groups were evaluated using
Student’s t test. Only microbial taxa with a relative abundance higher than 0.1% in at least 50% of sam-
ples were included in analyses. OTU abundance between the HED and ND groups was evaluated with
the Mann-Whitney U test, with a P value of ,0.05 considered statistically significant.
Data availability. The raw sequencing data generated in this study have been deposited in the
NCBI SRA database with the accession numbers PRJNA831679 (gut mucosa-associated microbiota) and
PRJNA833901 (intestinal transcriptome).

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.
SUPPLEMENTAL FILE 1, PDF file, 0.6 MB.
SUPPLEMENTAL FILE 2, XLSX file, 2 MB.

ACKNOWLEDGMENTS
This work was supported by the National Key Research and Development Program of
China (2021YFF1000600), the National Natural Science Foundation of China (32072690), the
Major Scientific Research Tasks for Scientific and Technological Innovation Projects of the
Chinese Academy of Agricultural Sciences (CAAS-ZDRW202006), and the Agricultural
Science and Technology Innovation Program of the Chinese Academy of Agricultural
Sciences (ASTIP-IAS05).
K.L., S.-H.H., and Y.-L.M. conceived and designed the project. N.W., L.H., X.-L.Z., S.-T.F., S.-
S.L., Y.W., Z.-G.L., B.-Y.W., and T.-W.W. collected the samples. S.-S.X., N.W., and L.H. analyzed
the data and wrote the paper. All authors reviewed and approved the final manuscript.
We have no conflict of interest relevant to this study to declare.

REFERENCES
1. Arnoriaga-Rodriguez M, Mayneris-Perxachs J, Burokas A, Contreras-Rodriguez 4. Newman TM, Shively CA, Register TC, Appt SE, Yadav H, Colwell RR,
O, Blasco G, Coll C, Biarnes C, Miranda-Olivos R, Latorre J, Moreno-Navarrete Fanelli B, Dadlani M, Graubics K, Nguyen UT, Ramamoorthy S, Uberseder
JM, Castells-Nobau A, Sabater M, Palomo-Buitrago ME, Puig J, Pedraza S, Gich B, Clear KYJ, Wilson AS, Reeves KD, Chappell MC, Tooze JA, Cook KL. 2021.
J, Perez-Brocal V, Ricart W, Moya A, Fernandez-Real X, Ramio-Torrenta L, Diet, obesity, and the gut microbiome as determinants modulating meta-
Pamplona R, Sol J, Jove M, Portero-Otin M, Maldonado R, Fernandez-Real JM. bolic outcomes in a non-human primate model. Microbiome 9:100.
2020. Obesity impairs short-term and working memory through gut microbial https://doi.org/10.1186/s40168-021-01069-y.
metabolism of aromatic amino acids. Cell Metab 32:548–560. https://doi.org/ 5. Bolte LA, Vich Vila A, Imhann F, Collij V, Gacesa R, Peters V, Wijmenga C,
10.1016/j.cmet.2020.09.002. Kurilshikov A, Campmans-Kuijpers MJE, Fu J, Dijkstra G, Zhernakova A,
2. Zhao M, Jiang Z, Cai H, Li Y, Mo Q, Deng L, Zhong H, Liu T, Zhang H, Weersma RK. 2021. Long-term dietary patterns are associated with pro-
Kang JX, Feng F. 2020. Modulation of the gut microbiota during inflammatory and anti-inflammatory features of the gut microbiome. Gut
high-dose glycerol monolaurate-mediated amelioration of obesity in 70:1287–1298. https://doi.org/10.1136/gutjnl-2020-322670.
mice fed a high-fat diet. mBio 11:e00190-20. https://doi.org/10 6. Dovrolis N, Michalopoulos G, Theodoropoulos GE, Arvanitidis K, Kolios G,
.1128/mBio.00190-20. Sechi LA, Eliopoulos AG, Gazouli M. 2020. The interplay between mucosal
3. Ley RE, Turnbaugh PJ, Klein S, Gordon JI. 2006. Microbial ecology: human microbiota composition and host gene-expression is linked with inflixi-
gut microbes associated with obesity. Nature 444:1022–1023. https://doi mab response in inflammatory bowel diseases. Microorganisms 8:438.
.org/10.1038/4441022a. https://doi.org/10.3390/microorganisms8030438.

July/August 2022 Volume 10 Issue 4 10.1128/spectrum.00717-22 14

Obesity by Modulating Host-Gut Microbiota Interaction Microbiology Spectrum

7. Brink LR, Matazel K, Piccolo BD, Bowlin AK, Chintapalli SV, Shankar K, hepatic damage and subcutaneous fat accumulation in high-fat diet fed
Yeruva L. 2019. Neonatal diet impacts bioregional microbiota composi- mice. PLoS One 13:e0198262. https://doi.org/10.1371/journal.pone.0198262.
tion in piglets fed human breast milk or infant formula. J Nutr 149: 26. Chen YH, Wang YC, Chiu CC, Lee YP, Hung SW, Huang CC, Chiu CF, Chen
2236–2246. https://doi.org/10.1093/jn/nxz170. TH, Huang WC, Chuang HL. 2020. Housing condition-associated changes
8. Hu Y, Peng J, Li F, Wong FS, Wen L. 2018. Evaluation of different mucosal in gut microbiota further affect the host response to diet-induced nonal-
microbiota leads to gut microbiota-based prediction of type 1 diabetes in coholic fatty liver. J Nutr Biochem 79:108362. https://doi.org/10.1016/j
NOD mice. Sci Rep 8:15451. https://doi.org/10.1038/s41598-018-33571-z. .jnutbio.2020.108362.
9. Erridge C. 2011. Diet, commensals and the intestine as sources of patho- 27. Yang F, Zhang S, Tian M, Chen J, Chen F, Guan W. 2020. Different sources
gen-associated molecular patterns in atherosclerosis, type 2 diabetes and of high fat diet induces marked changes in gut microbiota of nursery
non-alcoholic fatty liver disease. Atherosclerosis 216:1–6. https://doi.org/ pigs. Front Microbiol 11:859. https://doi.org/10.3389/fmicb.2020.00859.
10.1016/j.atherosclerosis.2011.02.043. 28. Bouter KE, van Raalte DH, Groen AK, Nieuwdorp M. 2017. Role of the gut
10. Sears CL, Garrett WS. 2014. Microbes, microbiota, and colon cancer. Cell microbiome in the pathogenesis of obesity and obesity-related metabolic
Host Microbe 15:317–328. https://doi.org/10.1016/j.chom.2014.02.007. dysfunction. Gastroenterology 152:1671–1678. https://doi.org/10.1053/j
11. Shively CA, Register TC, Appt SE, Clarkson TB, Uberseder B, Clear KYJ, Wilson .gastro.2016.12.048.
AS, Chiba A, Tooze JA, Cook KL. 2018. Consumption of Mediterranean versus 29. Yang G, Hong E, Oh S, Kim E. 2020. Non-viable Lactobacillus johnsonii
western diet leads to distinct mammary gland microbiome populations. Cell
JNU3402 protects against diet-induced obesity. Foods 9:1494. https://doi
Rep 25:47–56. https://doi.org/10.1016/j.celrep.2018.08.078.
.org/10.3390/foods9101494.
12. Olsen I, Yamazaki K. 2019. Can oral bacteria affect the microbiome of the gut? J
30. Esterhazy D, Canesso MCC, Mesin L, Muller PA, de Castro TBR, Lockhart A,
Oral Microbiol 11:1586422. https://doi.org/10.1080/20002297.2019.1586422.
ElJalby M, Faria AMC, Mucida D. 2019. Compartmentalized gut lymph
13. Nagpal R, Shively CA, Appt SA, Register TC, Michalson KT, Vitolins MZ,
node drainage dictates adaptive immune responses. Nature 569:
Yadav H. 2018. Gut microbiome composition in non-human primates
consuming a western or Mediterranean diet. Front Nutr 5:28. https://doi 126–130. https://doi.org/10.1038/s41586-019-1125-3.
.org/10.3389/fnut.2018.00028. 31. Tatapudi VS, Montgomery RA. 2019. Therapeutic modulation of the com-
14. Heinritz SN, Mosenthin R, Weiss E. 2013. Use of pigs as a potential model plement system in kidney transplantation. Front Immunol 10:2306.
for research into dietary modulation of the human gut microbiota. Nutr https://doi.org/10.3389/fimmu.2019.02306.
Res Rev 26:191–209. https://doi.org/10.1017/S0954422413000152. 32. Berger JP, Akiyama TE, Meinke PT. 2005. PPARs: therapeutic targets for
15. Elolimy AA, Washam C, Byrum S, Chen C, Dawson H, Bowlin AK, Randolph CE, metabolic disease. Trends Pharmacol Sci 26:244–251. https://doi.org/10
Saraf MK, Yeruva L. 2020. Formula diet alters the ileal metagenome and tran- .1016/j.tips.2005.03.003.
scriptome at weaning and during the postweaning period in a porcine 33. Peyrin-Biroulet L, Beisner J, Wang G, Nuding S, Oommen ST, Kelly D,
model. mSystems 5:e00457-20. https://doi.org/10.1128/mSystems.00457-20. Parmentier-Decrucq E, Dessein R, Merour E, Chavatte P, Grandjean T,
16. Chang HW, McNulty NP, Hibberd MC, O'Donnell D, Cheng J, Lombard V, Bressenot A, Desreumaux P, Colombel JF, Desvergne B, Stange EF,
Henrissat B, Ilkayeva O, Muehlbauer MJ, Newgard CB, Barratt MJ, Lin X, Odle Wehkamp J, Chamaillard M. 2010. Peroxisome proliferator-activated receptor
J, Gordon JI. 2021. Gut microbiome contributions to altered metabolism in a gamma activation is required for maintenance of innate antimicrobial immu-
pig model of undernutrition. Proc Natl Acad Sci U S A 118:e2024446118. nity in the colon. Proc Natl Acad Sci U S A 107:8772–8777. https://doi.org/10
https://doi.org/10.1073/pnas.2024446118. .1073/pnas.0905745107.
17. Xu SS, Zhang XL, Liu SS, Feng ST, Xiang GM, Xu CJ, Fan ZY, Xu K, Wang N, 34. Bi Y, Tu Y, Zhang N, Wang S, Zhang F, Suen G, Shao D, Li S, Diao Q. 2021.
Wang Y, Che JJ, Liu ZG, Mu YL, Li K. 2022. Multi-omic analysis in a meta- Multiomics analysis reveals the presence of a microbiome in the gut of fe-
bolic syndrome porcine model implicates arachidonic acid metabolism tal lambs. Gut 70:853–864. https://doi.org/10.1136/gutjnl-2020-320951.
disorder as a risk factor for atherosclerosis. Front Nutr 9:807118. https:// 35. Bolger AM, Lohse M, Usadel B. 2014. Trimmomatic: a flexible trimmer for
doi.org/10.3389/fnut.2022.807118. Illumina sequence data. Bioinformatics 30:2114–2120. https://doi.org/10
18. Chen W, Zhang X, Li J, Huang S, Xiang S, Hu X, Liu C. 2018. Comprehensive .1093/bioinformatics/btu170.
analysis of coding-lncRNA gene co-expression network uncovers conserved 36. Liu YX, Qin Y, Chen T, Lu M, Qian X, Guo X, Bai Y. 2021. A practical guide
functional lncRNAs in zebrafish. BMC Genomics 19:112. https://doi.org/10 to amplicon and metagenomic analysis of microbiome data. Protein Cell
.1186/s12864-018-4458-7. 12:315–330. https://doi.org/10.1007/s13238-020-00724-8.
19. Kelly J, Daly K, Moran AW, Ryan S, Bravo D, Shirazi-Beechey SP. 2017. 37. Andrews S. 2010. FastQC: a quality control tool for high throughput
Composition and diversity of mucosa-associated microbiota along the sequence data. http://www.bioinformatics.babraham.ac.uk/projects/fastqc.
entire length of the pig gastrointestinal tract; dietary influences. Environ 38. Trapnell C, Roberts A, Goff L, Pertea G, Kim D, Kelley DR, Pimentel H,
Microbiol 19:1425–1438. https://doi.org/10.1111/1462-2920.13619. Salzberg SL, Rinn JL, Pachter L. 2012. Differential gene and transcript expres-
20. Gong J, Li L, Zuo X, Li Y. 2019. Change of the duodenal mucosa-associated sion analysis of RNA-seq experiments with TopHat and Cufflinks. Nat Protoc
microbiota is related to intestinal metaplasia. BMC Microbiol 19:275. 7:562–578. https://doi.org/10.1038/nprot.2012.016.
https://doi.org/10.1186/s12866-019-1666-5. 39. Kong L, Zhang Y, Ye ZQ, Liu XQ, Zhao SQ, Wei L, Gao G. 2007. CPC: assess
21. Gutierrez-Repiso C, Moreno-Indias I, Martin-Nunez GM, Ho-Plagaro A,
the protein-coding potential of transcripts using sequence features and
Rodriguez-Canete A, Gonzalo M, Garcia-Fuentes E, Tinahones FJ. 2020. Mu-
support vector machine. Nucleic Acids Res 35:W345–W349. https://doi
cosa-associated microbiota in the jejunum of patients with morbid obesity:
.org/10.1093/nar/gkm391.
alterations in states of insulin resistance and metformin treatment. Surg
40. Wang L, Park HJ, Dasari S, Wang S, Kocher JP, Li W. 2013. CPAT: Cod-
Obes Relat Dis 16:1575–1585. https://doi.org/10.1016/j.soard.2020.04.008.
ing-Potential Assessment Tool using an alignment-free logistic
22. Rizzatti G, Lopetuso LR, Gibiino G, Binda C, Gasbarrini A. 2017. Proteobac-
teria: a common factor in human diseases. Biomed Res Int 2017:9351507. regression model. Nucleic Acids Res 41:e74. https://doi.org/10.1093/
https://doi.org/10.1155/2017/9351507. nar/gkt006.
23. Reynolds LA, Smith KA, Filbey KJ, Harcus Y, Hewitson JP, Redpath SA, 41. Sun L, Luo H, Bu D, Zhao G, Yu K, Zhang C, Liu Y, Chen R, Zhao Y. 2013. Uti-
Valdez Y, Yebra MJ, Finlay BB, Maizels RM. 2014. Commensal-pathogen lizing sequence intrinsic composition to classify protein-coding and long
interactions in the intestinal tract: lactobacilli promote infection with, and non-coding transcripts. Nucleic Acids Res 41:e166. https://doi.org/10
are promoted by, helminth parasites. Gut Microbes 5:522–532. https://doi .1093/nar/gkt646.
.org/10.4161/gmic.32155. 42. Siepel A, Bejerano G, Pedersen JS, Hinrichs AS, Hou M, Rosenbloom K,
24. London LE, Kumar AH, Wall R, Casey PG, O'Sullivan O, Shanahan F, Hill C, Clawson H, Spieth J, Hillier LW, Richards S, Weinstock GM, Wilson RK,
Cotter PD, Fitzgerald GF, Ross RP, Caplice NM, Stanton C. 2014. Exopoly- Gibbs RA, Kent WJ, Miller W, Haussler D. 2005. Evolutionarily conserved
saccharide-producing probiotic Lactobacilli reduce serum cholesterol elements in vertebrate, insect, worm, and yeast genomes. Genome Res
and modify enteric microbiota in ApoE-deficient mice. J Nutr 144: 15:1034–1050. https://doi.org/10.1101/gr.3715005.
1956–1962. https://doi.org/10.3945/jn.114.191627. 43. Kayala MA, Baldi P. 2012. Cyber-T web server: differential analysis of high-
25. Keskitalo A, Munukka E, Toivonen R, Hollmen M, Kainulainen H, Huovinen P, throughput data. Nucleic Acids Res 40:W553–W559. https://doi.org/10
Jalkanen S, Pekkala S. 2018. Enterobacter cloacae administration induces .1093/nar/gks420.

July/August 2022 Volume 10 Issue 4 10.1128/spectrum.00717-22 15