Impact of supplementation with a food-derived microbial community on obesity-associated inflammation and gut microbiota composition
Roselli et al. Genes & Nutrition
Impact of supplementation with a food-derived microbial community on obesity-associated inflammation and gut microbiota composition
Marianna Roselli 0
Chiara Devirgiliis 0
Paola Zinno 0
Barbara Guantario 0
Alberto Finamore 0
Rita Rami 0
Giuditta Perozzi 0
0 Food and Nutrition Research Centre, Council for Agricultural Research and Economics (CREA) , Via Ardeatina 546, 00178 Rome , Italy
Background: Obesity is a complex pathology associated with dysbiosis, metabolic alterations, and low-grade chronic inflammation promoted by immune cells, infiltrating and populating the adipose tissue. Probiotic supplementation was suggested to be capable of counteracting obesity-associated immune and microbial alterations, based on its proven immunomodulatory activity and positive effect on gut microbial balance. Traditional fermented foods represent a natural source of live microbes, including environmental strains with probiotic features, which could transiently colonise the gut. The aim of our work was to evaluate the impact of supplementation with a complex foodborne bacterial consortium on obesity-associated inflammation and gut microbiota composition in a mouse model. Methods: C57BL/6J mice fed a 45% high fat diet (HFD) for 90 days were supplemented with a mixture of foodborne lactic acid bacteria derived from the traditional fermented dairy product “Mozzarella di Bufala Campana” (MBC) or with the commercial probiotic GG strain of Lactobacillus rhamnosus (LGG). Inflammation was assessed in epididymal white adipose tissue (WAT) following HFD. Faecal microbiota composition was studied by next-generation sequencing. Results: Significant reduction of epididymal WAT weight was observed in MBC-treated, as compared to LGG and control, animals. Serum metabolic profiling showed correspondingly reduced levels of triglycerides and higher levels of HDL cholesterol, as well as a trend toward reduction of LDL-cholesterol levels. Analysis of the principal leucocyte subpopulations in epididymal WAT revealed increased regulatory T cells and CD4+ cells in MBC microbiotasupplemented mice, as well as decreased macrophage and CD8+ cell numbers, suggesting anti-inflammatory effects. These results were associated with lower levels of pro-inflammatory cytokines and chemokines in WAT explants. Faecal bacterial profiling demonstrated increased Firmicutes/Bacteroidetes ratio in all mice groups following HFD. Conclusions: Taken together, these results indicate a protective effect of MBC microbiota supplementation toward HFD-induced fat accumulation and triglyceride and cholesterol levels, as well as inflammation, suggesting a stronger effect of a mixed microbial consortium vs single-strain probiotic supplementation. The immunomodulatory activity exerted by the MBC microbiota could be due to synergistic interactions within the microbial consortium, highlighting the important role of dietary microbes with yet uncharacterised probiotic effect.
Chronic inflammation; Fermented dairy; Foodborne microbiota; White adipose tissue; High fat diet
Obesity is a chronic, multifactorial disorder reaching
epidemic proportions globally, affecting persons of
virtually all ages in both developed and developing countries
]. Promoted by a combination of genetic
predisposition, nutritional excess, and sedentary lifestyle, obesity
is primarily characterised by increased fat mass,
accompanied by development of related disorders [
Expansion of the adipose organ, mainly affecting white
adipose tissue (WAT), results in adipocyte dysfunction.
WAT has been increasingly considered not only a
metabolic organ, but also an active endocrine tissue, as it
secretes a large number of peptide hormones called
adipokines, such as leptin and adiponectin, that operate
in a complex network and actively communicate with
other organs [
]. Secretion by the adipose organ is
disturbed in obesity, as adipokine release is dysregulated
and associated with production of several inflammation
mediators. For this reason, the adipose tissue is considered
to be a major contributor to obesity-linked low-grade
chronic inflammation [
]. The inflammatory process
involves impairment of both the innate and adaptive
immune system and is triggered by local secretion of
inflammatory cytokines and chemokines such as tumour
necrosis factor-α (TNF-α), interleukin-6 (IL-6), monocyte
chemoattractant protein (MCP)-1, and Regulated on
Activation Normal T cell Expressed and Secreted (RANTES).
These mediators recruit immune cells from blood vessels,
such as lymphocytes and macrophages, which in turn
massively infiltrate the adipose tissue [
]. Indeed, high
levels of inflammatory cells such as T CD8+ lymphocytes
and activated M1 macrophages are found in obese WAT,
accompanied by decreased levels of CD4+CD25+Foxp3+
regulatory T (Treg) cells, a key population in
maintaining immunological tolerance and immune
]. This inflammatory status, arising locally and
then becoming systemic, triggers the onset of other
diseases frequently associated with obesity such as the
metabolic syndrome, characterised by visceral obesity,
high blood pressure, insulin resistance, high
circulating triglyceride levels, and low HDL cholesterol,
leading in turn to increased risk of cardiovascular
The gut microbiota has recently attracted much
attention as a crucial factor associated with obesity
]. Alterations of intestinal microbial composition,
in terms of bacterial phyla and classes associated with
improved energy extraction from the undigested
dietary carbohydrate component, were identified in obese
human subjects and animal obesity models, with
consequent impact on host metabolism and energy
]. Both diet- and genetically induced
obesity were shown to associate with imbalance in the
relative proportion of Gram-negative Bacteroidetes
and Gram-positive Firmicutes, the two major phyla of
gut bacteria, with the latter prevailing in obese
]. However, imbalance in these two bacterial
phyla is not sufficient by itself to determine the
obesity phenotype. Other factors, such as diet, pre- and
probiotic supplementation, antibiotics, surgery, and
faecal transplantation, can impact the overall
metabolic capacity of the gut microbiome [
]. Within this
context, dietary interventions aimed at promoting
selection of beneficial intestinal microbes could
represent a powerful strategy to counteract
obesityassociated intestinal dysbiosis. There is growing
evidence that probiotic and/or prebiotic supplementation
can positively modulate gut microbiota, thus
representing important assets in the management of
]. The probiotic component of the gut
microbiota can confer health benefits to the host
mainly acting on immunomodulation and positively
influencing intestinal microbial balance [
supplementation was therefore suggested to be able
to counteract obesity-associated immune alterations
and microbial imbalance [
]. As an alternative to
commercially available probiotic strains, a natural
source of live bacteria is represented by fermented
foods, which also confer the advantage of providing
the host with a complex microbiota containing several
environmental strains with potential probiotic
features, such as the capability to transiently colonise
animal and human gut and interact with the resident
gut microbiota, mainly at a trophic level [
Increasing scientific interest in fermented foods was
also recently boosted by their possible use as models
for more complex microbiota such as the gut [
The most relevant foodborne lactic acid bacteria (LAB)
belong to the Lactobacillus, Lactococcus, Streptococcus,
Pediococcus, and Leuconostoc genera. Several LAB species
are also highly represented within the resident gut
microbiota of healthy humans. Lactobacillus species, in
particular, are abundant both in food and in the gut [
The aim of our work was to evaluate the impact of
supplementation with a complex foodborne bacterial
community on obesity-associated inflammation, as well
as on gut microbiota composition. For this purpose, we
used a mouse model of high fat diet (HFD)-induced
obesity, comparing the effect of supplementation with a
mixture of natural LAB strains derived from the
traditional fermented dairy product “Mozzarella di Bufala
Campana” (MBC) [
] and with the well-characterised
probiotic GG strain of Lactobacillus rhamnosus (LGG).
The MBC bacterial consortium was dominated by
Lactobacillus delbrueckii, Lactobacillus fermentum, and
Leuconostoc lactis [
]. LGG was used as probiotic
control on the basis of its proven beneficial effects in the
prevention of obesity [
Experimental design, animals, and diets
Six-week-old C57BL/6J male mice, obtained from
Charles River Laboratories (Como, Italy), were kept at
23 °C with a 12-h light-dark cycle and fed ad libitum
with a standard laboratory diet (4RF21, Mucedola,
Milano, Italy, www.mucedola.it). Mice had free access to
food and water throughout the experiments. Food intake
and body weight were recorded every other day. After
1 week of adaptation, animals were randomly divided
into three groups (five mice per group) and orally
supplemented for 15 days with 1 × 109 CFU/day of a
mixture of natural LAB strains extracted from MBC [
or with the probiotic strain LGG. Phosphate-buffered
saline (PBS) supplementation was used as control (CTRL).
After 15 days, all mice were shifted to HFD (http://
total calories from fat, designed with similarities to
Research Diets, Inc., formula D12451 and provided by
Mucedola) while continuing to receive bacterial
supplementation for 90 additional days. Due to logistic reasons
related to the number of animals that could be handled
at the same time, the experimental design envisioned
two rounds of treatment, 2 weeks apart from each other,
in which the two groups of mice, of the same age, were
fed the same batches of diets. Therefore, the second
group of mice was not aimed at testing reproducibility,
but rather at increasing the number of treated animals.
Statistical analysis of the results included all animals
subjected to the same supplementation protocol,
irrespective of their treatment within experimental period 1
or 2. At the end of the experimental period, mice were
anaesthetised by intraperitoneal injection of
pentobarbital (10 mg/kg) following overnight fasting, blood was
drawn via cardiac puncture, and epididymal WAT was
excised, weighed, and immediately placed in ice-cold
PBS under sterile conditions. Serum was prepared from
blood and stored at − 80 °C until further analysis. Faeces
were collected and stored at − 80 °C for microbiological
analysis at the following time points: t0 (beginning of
bacterial treatments), t15 (shift to HFD) and t105
(90 days on HFD). The experimental protocol and
sampling times are summarised in Fig. 1.
MBC is a traditional Italian fermented cheese with PDO
designation (Product of Designated Origin, EEC
Regulation no. 1107). It is consumed fresh, within 2 weeks
from production, and it contains high titres of live
]. To prepare the MBC microbiota, 10 g of
cheese samples were diluted in 90 ml sodium citrate
solution (2% w/v) and homogenised in a BagMixer400
(Interscience, France), as previously described [
standardise the bacterial inoculum to be administered to
mice, the MBC homogenate was entirely used as a single
inoculum in 2 l of De Man Rogosa Sharpe (MRS)
medium (Oxoid Ltd., England) and incubated at 37 °C
for 48 h under anaerobic conditions (Anaerocult A,
Merck, Germany) to obtain a final bacterial titre of
about 1.5 × 109 CFU/ml. The resulting bacterial
suspension was divided in aliquots containing 1 × 109 CFU
each, stored at − 80 °C in 20% (v/v) glycerol, and thawed
daily for oral administration to mice, following washing,
resuspension in 1× PBS, and mixing with small amounts
of minced feed.
The LGG strain ATCC53103 was grown, prepared,
and orally given to mice as described above for MBC
Serum metabolic measurements
The following plasma parameters were analysed: glucose
(Glucose Liquid kit, Sentinel Diagnostics, Milan, Italy),
HDL and LDL cholesterol (Max Discovery HDL and
LDL Cholesterol Assay Kits, Bioo Scientific, Austin, TX),
and triglycerides (Triglycerides Liquid kit, Sentinel
Diagnostics). The adiponectin was quantified by ELISA
(Biorbyt, Cambridge, UK). Analyses were conducted on
a subset of five samples for each treatment, due to
technical issues related to serum withdrawal or hemolysis.
Immune cell isolation and staining
Macrophages and lymphocytes were isolated from the
epididymal WAT stromal vascular fraction (SVF),
according to [
], as several populations of immune cells
are well known to reside in the SVF. The following
monoclonal antibodies, purchased from eBioscience
(San Diego, CA), were used in this study: FITC anti-CD3
(clone 500A2), PE anti-CD8 (clone 53-6.7), PE-Cy5
antiCD4 (clone RM4-5), FITC anti-CD11b (clone M1/70),
PE anti-F4/80 (clone BM8), PerCP-Cy5.5 anti-CD45
(clone 30-F11), and anti-CD16/CD32 (clone 93). Briefly,
1 × 106 cells, resuspended in FACS labelling buffer (PBS
with 2 mM EDTA and 1% foetal calf serum), were
preincubated for 20 min with anti-CD16/CD32 to avoid
non-specific binding, then washed and labelled with the
appropriate mixture of antibodies for 30 min,
centrifuged, and resuspended in FACS labelling buffer. Flow
cytometry analysis was performed using a FACSCalibur
flow cytometer (BD Biosciences, Milan, Italy). To
exclude dead/dying cells that could non-specifically bind
antibodies, leukocytes were gated according to forward
and side scatter. The percentage of T helper and
cytotoxic cells was calculated on lymphocyte gate (CD3+),
whereas the CD11b+ and F4/80+ cell subsets were
calculated on the leukocyte gate (CD45+). Treg cell (CD4
+CD25+Foxp3+) analysis was performed with a specific
kit (eBioscience, San Diego, CA) staining CD4 (FITC),
CD25 (PE) and transcription factor Foxp3 (PE-Cy5),
according to the manufacturer’s instructions. The
percentage of CD25+Foxp3+ cells was calculated on lymphocyte
CD4+ gate. For all analyses, at least 10.000 events were
acquired and analysed using the CellQuest software (BD
Biosciences, Milan, Italy).
Cytokine and chemokine secretion in WAT explants
WAT explant cultures were established essentially as
described by [
]. Briefly, epididymal WAT was dissected,
weighed, minced, and placed into 12-well tissue culture
plates (Corning, Milan, Italy) at 120 mg/well, with either
1 ml T cell activation medium (complete DMEM
containing 3.7 g/l NaHCO3, 10% heat-inactivated foetal calf
serum, 4 mM glutamine, 1% non-essential amino acids,
105 U/l penicillin and 100 mg/l streptomycin, 5 ng/ml
phorbol 12-myristate 13-acetate (PMA), and 1 ng/ml
ionomycin) or control medium (complete DMEM
without ionomycin and PMA). All reagents were from
Euroclone (Milan, Italy), except for ionomycin and
PMA, which were from Sigma (Milan, Italy).
Conditioned media were collected after 24 h of culture at
37 °C in an atmosphere of 5% CO2/95% air at 90%
relative humidity and stored at − 80 °C until further
analysis. The levels of cytokines and chemokines were
analysed using Bio-plex/Luminex technology (mouse
magnetic Luminex screening assay, Labospace, Milan) or
ELISA assays (Affymetrix, eBioscience, San Diego, CA).
The following cytokines and chemokines were
simultaneously detected by Luminex technology in 50 μl undiluted
samples: interferon gamma-induced protein (IP)-10,
granulocyte macrophage-colony stimulating factor (GM-CSF),
Regulated on Activation-Normal T cell Expressed and
Secreted (RANTES), interleukin (IL)-23, IL-4, and IL-10.
The following cytokines were analysed by ELISA (100 μl
samples): tumour necrosis factor (TNF)-α, interferon
(IFN)-γ, IL-17A, and IL-6. For these latter two cytokines,
samples were diluted 1:500, as the readings by Luminex
assays for IL-17A and IL-6 were out of range.
DNA extraction from faecal samples
Total DNA was extracted from 80 mg faecal samples
with QIAamp DNA Stool Mini Kit (Qiagen, Hilden,
Germany) according to manufacturer’s instructions.
Qiagen DNA extraction method used in this work was
chosen as it was listed among the most reproducible
kits, ensuring minimal influence on next-generation
sequencing (NGS) data analysis [
NGS was performed on faecal DNA samples from four
animals for each of the three experimental groups, at the
three time points indicated in Fig. 1, namely t0, t15, and
t105 (total number of samples = 36). Partial 16S rRNA
gene sequences were amplified using primer pair
Probio_Uni and /Probio_Rev, which targets the V3 region of
the gene and sequenced at the DNA sequencing facility
of GenProbio srl (www.genprobio.com) using a MiSeq
(Illumina). Primers and protocols, including amplicon
checks, were as described in [
]. Individual sequence
reads were filtered with the Illumina software to remove
low quality and polyclonal sequences. All Illumina
quality-approved, trimmed, and filtered data were
exported as .fastq files and processed using a custom
script based on the QIIME software suite [
control retained sequences 140–400 bp long, with mean
sequence quality score > 20, and truncation at first base
if a low quality rolling 10-bp window was found.
Presence of homopolymers > 7 bp and sequences with
mismatched primers were omitted. To calculate
downstream diversity (alpha and beta diversity indices,
UniFrac analysis), 16S rRNA operational taxonomic
units (OTUs) were defined at ≥ 97% sequence homology
using uclust [
]. All reads were classified to the lowest
possible taxonomic rank using QIIME and a reference
dataset from the SILVA database [
between samples were calculated by unweighted UniFrac
]. The range of similarities is calculated between the
values 0 and 1. Principal Coordinate Analysis (PCoA)
was applied using the UniFrac program.
Statistical univariate analysis
Values in graphs and tables represent means ± SD. Prior
to analysis, normal distribution and homogeneity of
variance of all variables were assumed with Shapiro-Wilk’s
and Levene’s tests, respectively. Statistical significance
was evaluated by one-way ANOVA or by ANCOVA,
followed by post hoc Tukey honestly significant
difference (HSD) test. Differences with P values < 0.05 were
considered significant. Statistical univariate analysis was
performed with the “Statistica” software package (version
5.0; Stat Soft Inc., Tulsa, OK).
Statistical multivariate analysis
Non-supervised principal component analysis (PCA) of
WAT immunological profiles (leukocyte subpopulations
and cytokine/chemokine secretion) was performed with
Past software, version 2.17c [
]. Data were collected in
a matrix of 27 rows (number of animals) and 15
columns (number of variables) and were auto-scaled by
mean-centring and normalised by standard deviation.
Pearson’s correlation coefficients between variables and
principal components, as well as statistical significance
of the correlation, were also calculated.
Bacterial supplementation affects epididymal WAT weight and metabolic parameters
Body and WAT weight values in the three groups of
mice are shown in Table 1 in comparison with food and
energy intakes. As expected, HFD feeding induced
significant weight increase in all groups, leading to
comparable body weight and weight gain values by the
end of the experimental period. Nevertheless, significant
reduction of WAT weight (P < 0.05) was observed in
MBC-treated animals, as compared to LGG and CTRL
mice. Food and energy intake were similar in the three
mice groups. To account for a possible influence of food
intake on WAT weight, ANCOVA analysis was
performed, considering WAT weight as the dependent
variable, treatment as the independent variable, and food
intake as the covariate. The results confirmed that WAT
weight reduction in the MBC group as compared to
LGG and CTRL could not be attributable to differential
food intake. Supplementing with the foodborne MBC
microbiota also led to reduced serum levels of
triglycerides, coupled with higher levels of HDL
cholesterol (P < 0.05 and P < 0.001, respectively), and a trend
toward decreased LDL cholesterol (P = 0.05) as
compared to the CTRL group (Table 2). Serum metabolic
parameters of LGG-treated mice displayed a similar but
milder effect, with a trend toward reduced triglyceride
levels (P = 0.05) and increased HDL-cholesterol levels
(P < 0.05). No significant differences were detected
among the three groups of mice concerning fasting
glucose and adiponectin levels.
WAT immunological profiles highlight the antiinflammatory effect of MBC microbiota supplementation
Flow cytometry analysis of the main leukocyte
subpopulations in epididymal WAT (Fig. 2) revealed increased
numbers of the immune homeostasis regulator CD4+
CD25+ Foxp3+ Treg cells (Fig. 2a, P < 0.001 vs CTRL and
P < 0.01 vs LGG) and CD4+ T lymphocytes (Fig. 2b,
P < 0.001 vs CTRL) in MBC microbiota-supplemented
mice, accompanied by decreased pro-inflammatory CD8+
T lymphocytes (Fig. 2b, P < 0.001 vs CTRL), CD11b+
activated leukocytes and F4/80+ macrophages (Fig. 2c,
P < 0.001 and P < 0.01 vs CTRL, respectively), suggesting
that MBC supplementation associates with an overall
antiinflammatory effect. LGG treatment also positively affected
WAT leukocyte subpopulations in terms of increased
percentage of Treg (P < 0.05 vs CTRL) and CD4+ cells
(P < 0.001 vs CTRL) and decreased CD8+ cells (P < 0.001
vs CTRL) as well as activated leukocytes (P < 0.01 vs
Leukocyte profiling of MBC-treated animals was
associated in cultured WAT explants with decreased levels
of pro-inflammatory cytokines and chemokines, such as
IL-6, TNF-α and IFN-γ (P < 0.001 vs CTRL and LGG),
IL-17A (P < 0.001 vs LGG), IP-10 (P < 0.01 vs LGG and
P < 0.05 vs CTRL), GM-CSF, and RANTES (P < 0.05 vs
CTRL). Reduced levels were also observed in WAT
leukocytes of LGG-supplemented mice, but they related
to a smaller subset of pro-inflammatory cytokines,
namely IL-6 and IFN-γ (P < 0.001 vs CTRL), IL-17A,
*P < 0.05 versus CTRL (ANOVA); #P < 0.01 versus CTRL and LGG (ANCOVA)
*P < 0.05 versus CTRL; **P < 0.001 versus CTRL; #P = 0.05 versus CTRL
and RANTES (P < 0.001 and P < 0.01 vs CTRL,
respectively) (Fig. 3). No significant differences were observed
among mice groups for the two anti-inflammatory
cytokines IL-4 and IL-10 nor for pro-inflammatory IL-23
(data not shown).
Considering the dynamic and inherently multivariate
nature of the immune response, WAT immunological
profiles were further explored by principal component
analysis (PCA) (Table 3). The first three principal
components accounted for 64.15% of the overall variance,
with individual values of 33.81, 19.47, and 10.87% for
PC1, PC2, and PC3, respectively. The most informative
score plot was the PC1/PC2 shown in Fig. 4, where PC1
was responsible for clearly discriminating MBC samples
from LGG and CTRL samples. The variables mostly
contributing to such discrimination are identified by
higher loading values on PC1 (presented in italic
characters in Table 3), indicating significant correlation
between PC1 and the specific variable. In particular:
PC1 shows strong significant inverse correlation with
the pro-inflammatory markers CD3CD8+ (r = − 0.813),
CD11b+ (r = − 0.727), F4/80+ (r = − 0.804), IL-6 (r = −
0.669), TNF-α (r = − 0.660), and GM-CSF (r = − 0.544)
and significant direct correlation with the
antiinflammatory markers CD3CD4+ (r = 0.778) and
CD4CD25+ (r = 0.819). However, a tendency of the LGG
and CTRL samples to separate into two distinct clusters
is also observed (Fig. 4). PC2, on the other hand,
discriminates a subgroup of CTRL mice showing both
pro- and anti-inflammatory features. These features are
highlighted by the most discriminative variables: the
pro-inflammatory cytokines IP-10 (r = 0.842) and IFN-γ
(r = 0.587) and the anti-inflammatory markers IL-4
(r = 0.733) and IL-10 (r = 0.763) (Table 3).
Impact of bacterial supplementation on gut microbiota profiles
Next-generation sequencing (NGS) of 16S rDNA from
faecal samples of treated or control mice was used to
retrieve information on the bacterial relative abundance at
time points t0, t15, and t105. Taxonomical assignment
and read abundance estimates for all detected
operational taxonomic units (OTUs) are reported in Fig. 5 at
the phylum level, while the corresponding profiles at the
species level are listed in Additional file 1: Table S1. As
expected, Bacteroidetes and Firmicutes were detected as
predominant bacterial phyla, with different relative
proportions related to the time points analysed (Fig. 5).
Notably, all three experimental groups displayed
statistically significant increase in the Firmicutes/Bacteroidetes
ratio at the final time point as compared to the
beginning of the HFD treatment (t105 vs t15: P < 0.001 for
MBC and LGG; P < 0.05 for CTRL). These altered ratios
were also accompanied by decreased microbial
biodiversity, measured by the Chao1 and Shannon indices (data
not shown). Differences in the overall composition of
the faecal bacterial community were further analysed
using the UniFrac phylogeny-based metric [
Coordinates Analysis (PCoA) confirmed clustering of
bacterial species according to sampling time. The first
three principal components accounted for 41% of the
overall variance, with individual values of 23, 10, and 8%
for PC1, PC2, and PC3, respectively. The most
informative score plot was the PC1/PC2, shown in Fig. 6. A
clear difference was observed between the initial (t0,
t15) and final (t105) time points (Fig. 6a), while no
difference could be observed among the three experimental
conditions when samples were grouped according to
supplementation type (Fig. 6b). However, it is worth
noting that both L. delbrueckii and Leuc. lactis species,
representing two major components of the MBC
], were detected exclusively in faecal
samples of MBC-supplemented mice, although at very
low abundance (Additional file 1: Table S1).
In this work, we investigated the effects of a complex
foodborne bacterial community (MBC microbiota) on
obesity-associated inflammation and gut microbiota
composition in a HFD-induced obese mouse model. The
cultivable LAB component of MBC microbiota, selected
by growth in MRS medium, was extracted from a
fermented unripened cheese especially rich in live titres
of LAB species [
] dominated by L. fermentum, L.
delbrueckii, and Leuc. lactis [
] whose strains have often
been associated with probiotic features [
rationale for supplementing mice with the microbial
consortium was based on the highly biodiverse nature of
foodborne strains in fermented dairies, including several
LAB strains of environmental origin with beneficial,
although yet uncharacterised, features [
combined metabolic functions and metabolites have
been suggested to exert positive effects on host
physiology through synergistic mechanisms, more efficiently
than single strain supplementation [
]. However, the
probiotic capacity of mixed foodborne microbial
consortia has been gaining consideration only recently [
Moreover, most published work report supplementation
with single bacterial strains, and only few studies
compared multi-strain probiotic mixtures to investigate
possible synergistic interactions [
]. We chose to run a
parallel group of mice for comparison, supplemented
with the single probiotic strain GG of Lactobacillus
rhamnosus that was shown to exert positive effects on
obesity-related inflammation in mice and humans [
The obese phenotype was induced in C57BL/6J mice
by feeding a 45% HFD for 3 months, resulting in weight
gain in all experimental groups irrespective of bacterial
supplementation type. Many other studies report
decreased body weight gain following probiotic
]. Although we detected constant
weight gain in all mice groups, decreased epididymal
WAT weight was evident following oral administration
of MBC microbiota as compared to the other mice
groups, as well as a more pronounced anti-inflammatory
effect than LGG supplementation. Decreased
inflammation and amelioration of obesity-related metabolic
and immunological dysfunctions were previously
observed with bacterial supplementation of HFD-fed
], but they were not accompanied by WAT
weight reduction. WAT is considered the main
contributor to development of the obesity-associated low-grade
chronic systemic inflammatory state, which is
characterised by an imbalanced cytokine network with increased
production of several pro-inflammatory mediators.
Epididymal WAT, like other intra-abdominal WAT
depots, is now recognised to have a more negative impact
on health than subcutaneous WAT [
], and its decreased
weight following MBC supplementation further highlights
a higher efficacy of this complex microbial community in
supporting healthy metabolism. The specific
antiinflammatory effects observed in our study involved
decreased levels of the pro-inflammatory cytokines IL-6 and
IFN-γ and of the chemokines IP-10 and RANTES in
cultured WAT explants of LGG-supplemented mice, while
MBC-treated animals displayed stronger decrease in the
expression of a broader panel of pro-inflammatory
cytokines and chemokines, namely IL-6, TNF-α, IL-17A,
IFNγ, IP-10, GM-CSF, and RANTES. Other studies using
single probiotic strains or multi-strain mixtures observed
decreased expression of some of these markers [
48, 50, 52
IL-6 and TNF-α are the main cytokines produced by
proinflammatory macrophages in obese adipose tissue,
whereas RANTES and IP-10 are important lymphocyte
and macrophage chemo-attractants [
]. IFN-γ is secreted
by infiltrating CD8+ T cells, thus contributing to the
critical events driving adipose tissue inflammation [
Regarding IL-17, it was suggested that obesity predisposes
to selective expansion of the Th17 subclass of T
lymphocytes, producing high levels of IL-17 in an IL-6-dependent
]. The cytokine GM-CSF, although not
frequently measured in studies addressing
probioticdependent immunomodulation in obesity, was reported to
increase in the serum of obese mice [
The positive effects exerted by MBC supplementation
on the overall profile of WAT inflammatory cytokines
and chemokines were also associated to improved
balance between the major sub-populations of immune
cells, as revealed by reduced percentage of
proinflammatory CD8+ T lymphocytes, activated leukocytes
and macrophages, and increased CD4+ T lymphocytes
and CD25+Foxp3+ Treg cells. Similar findings were
reported in other tissues following Bifidobacterium
pseudocatenulatum supplementation [
], in the adipose
tissue after Lactobacillus gasseri supplementation [
or using a probiotic mixture of L. rhamnosus and
Bifidobacterium animalis subsp. lactis [
]. Treg cells
are highly represented in the WAT of lean mice, and they
are essential for the maintenance of an anti-inflammatory
environment in the absence of obesity. Treg cell number
has been shown to decrease in the WAT of obese mice,
contributing to worsen the inflammatory state [
The increased Treg cell number that we observe after
MBC supplementation is a result of particular relevance,
considering that selective modulation of this population
was shown to be tightly related to the level of
obesityassociated inflammation .
The anti-inflammatory effects occurring with MBC
supplementation were even more evident following PCA
analysis of the datasets, which clearly discriminated MBC
samples from LGG and CTRL samples along the first
principal component axis. This confirms the key role of
the immune cell subpopulations, as well as of the
cytokines GM-CSF, IL-6, and TNF-α, as the most important
variables contributing to the discrimination. Separation of
the LGG and CTRL samples into two distinct clusters was
highlighted only as a trend. These effects were
accompanied by positive changes in the expression of lipid
metabolism biomarkers in the MBC-supplemented group, with
decreased circulating levels of triglycerides, increased
HDL-cholesterol levels, and a trend toward decreasing
LDL cholesterol. Higher levels of circulating HDL
cholesterol were also observed in the LGG mice group, in line
with previous reports on supplementation with single
probiotics or mixtures [
31, 48, 50
Interaction with the host metagenome is considered
an important aspect in probiotic-mediated immune
]. We analysed faecal microbiota
biodiversity in treated mice by NGS of 16S rDNA. Our
results confirmed that gut microbiota composition was
indeed affected by HFD, leading to the establishment
of an increased Firmicutes/Bacteroidetes ratio typical
of the obesity pattern . Bacterial supplementation
was not able to overcome HFD-induced effects on gut
microbial profile, as no substantial modifications in
faecal microbiota composition could be observed over
time by NGS. The overriding effect of HFD on
microbial biodiversity was also confirmed by advanced
multivariate statistical analysis, namely Principal
Coordinates Analysis (PCoA), revealing no specific
clustering of bacterial species according to supplementation
type, while highlighting a clear variation of microbial
composition at the end of the experimental period in
all mice groups. Other studies reported different
extent of alterations in resident gut microbiota profile
following probiotic treatment of HFD-fed mice [
], but the studies are not always comparable due
to different experimental designs (duration of
treatment, percent dietary fat, etc.) and experimental
approaches employed for microbial profiling (i.e. NGS,
qPCR). In our study, the high sensitivity of NGS
allowed to detect two of the three predominant species
characterising the MBC-derived microbiota, namely L.
delbrueckii and Leuc. lactis, although with low relative
abundance in the faecal microbiome of supplemented
mice. These two species may thus be able to colonise
the gut of supplemented mice more efficiently. Gut
colonisation capacity of some components of
MBCderived microbiota was also shown in the simple
model organism Caenorhabditis elegans . On the
other hand, the L. rhamnosus species that includes the
LGG strain was undetectable in faecal microbiomes of
LGG-treated mice. Conflicting results concerning LGG
colonisation capacity have been reported in the
literature. Park et al. recently observed decreased
Lactobacillus relative abundances in the murine gut,
including the LGG strain, following LAB
], while in another report of orally
administered LGG to knockout (ApoE−/−) mice fed HFD, L.
rhamnosus could be recovered by faecal dilution and
]. Nevertheless, several reports indicate that
oral administration of specific bacteria can exert
beneficial effects on the host even in the absence of
Taken together, our results suggest that
supplementation with a biodiverse foodborne bacterial consortium
can exert beneficial effects on obesity-associated
inflammation and health-related parameters more effectively
than single probiotic strain supplementation. A recent
report by Sonnenburg et al. clearly shows that dietary
perturbations can lead to permanent loss of specific gut
bacterial taxa, due to negative selection of metabolic
activities that become unnecessary under imbalanced
dietary regimens [
]. These results point at limitations in
microbiota resilience occurring under extreme
conditions, such as HFD-induced obesity, where the
alterations cannot be reversed by simple dietary intervention
if not accompanied by specific bacterial supplementation
aimed at restoring the lost taxa. Foodborne bacteria
could play a key role in this respect, and to the best of
our knowledge, this is among the very few reports
evaluating the impact of a complex microbial
consortium naturally occurring in a traditional fermented food
on host physiology.
Our results demonstrate a stronger effect of a mixed
microbial consortium vs single-strain probiotic
supplementation in ameliorating HFD-induced inflammation in the
WAT of obese mice. The present study highlights the
importance of considering complex foodborne microbial
consortia naturally occurring in fermented products for
human consumption as potential probiotic vectors. It
also points at the importance of coupling multivariate to
univariate statistical analysis for better understanding of
the key factors responsible for probiotic effects. The
observed immunomodulatory activity exerted by the
MBC-derived microbiota suggests synergistic
interactions of microbial strains of environmental origin,
present within the foodborne consortium. More studies
are needed to further investigate the role of dietary
microbes with yet uncharacterised probiotic effect, aimed
also at identifying novel, under-represented strains
which could be unique to the foodborne microbiota.
Additional file 1: Table S1. Complete results of the identification, at
the species level or, where not possible, at higher taxonomic ranks, of the
sequences obtained by next-generation sequencing of faecal mice
samples, expressed as percentage. Species representing the components
of MBC microbiota are highlighted in bold. (XLSX 96 kb)
CFU: Colony-forming units; CTRL: Control; GM-CSF: Granulocyte
macrophagecolony stimulating factor; HFD: High fat diet; IFN: Interferon; IL: Interleukin;
IP: Interferon gamma-induced protein; LAB: Lactic acid bacteria; LGG: L.
rhamnosus GG; MBC: Mozzarella di Bufala Campana; MRS: De Man Rogosa
Sharpe medium; NGS: Next-generation sequencing; OTUs: Operational
taxonomic units; PCA: Principal component analysis; PCoA: Principal
Coordinates Analysis; RANTES: Regulated on Activation-Normal T cell
Expressed and Secreted; TNF: Tumour necrosis factor; Treg: Regulatory T cells;
WAT: White adipose tissue
The Authors wish to thank Kariklia Pascucci for her kind support in daily lab
work, Dr. Andrea Ciolfi for his valuable assistance in managing the NGS data,
and Dr. Fausta Natella and Dr. Gianni Pastore for their helpful suggestions on
the statistical analysis.
This work was funded in part by the Italian Ministry of Agriculture, Food &
Forestry Policies (MiPAAF), with grant “MEDITO” (DM 12487/7303/11) and
with national support to the JPI-HDHL “ENPADASI” project.
Availability of data and materials
Raw NGS data are available at the EMBL-EBI European Nucleotide Archive
], under the study accession number PRJEB20801.
MR, GP, and CD conceived and designed the experiments. MR, PZ, BG,
AF, and CD performed the experiments. MR and CD analysed the data
and supervised all data analyses. GP contributed reagents/materials/
analysis tools. MR, GP, and CD wrote the paper. RR did the animal
experiments/treatments. MR and AF performed the immunological
analysis. PZ, BG, and CD performed the microbiological analysis. CD
analysed the microbiota sequencing data. All authors read and approved
the final manuscript.
All experimental procedures involving animals complied with the European
Guidelines for the Care and Use of Animals for Research Purposes (Directive
2010/63/EU), and protocols were approved by the Ethical Committee of the
Food and Nutrition Research Center and by the National Health Ministry,
General Direction of Animal Health and Veterinary Drugs (agreement no
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
1. World Health Organization. Obesity and overweight fact sheet N 311- updated March 2011 2011, http://www.who.int/mediacentre/factsheets/ fs311/en/index.html.
2. Ng M , Fleming T , Robinson M , Thomson B , Graetz N , Margono C , Mullany EC , Biryukov S , Abbafati C , Abera SF , et al. Global, regional, and national prevalence of overweight and obesity in children and adults during 1980-2013: a systematic analysis for the Global Burden of Disease Study 2013 . Lancet. 2014 ; 384 : 766 - 81 .
3. Serra-Majem L , Bautista-Castano I . Etiology of obesity: two “key issues” and other emerging factors . Nutr Hosp . 2013 ; 28 ( Suppl 5 ): 32 - 43 .
4. Crino M , Sacks G , Vandevijvere S , Swinburn B , Neal B. The influence on population weight gain and obesity of the macronutrient composition and energy density of the food supply . Curr Obes Rep . 2015 ; 4 : 1 - 10 .
5. Shabana HS . Obesity, more than a 'cosmetic' problem. Current knowledge and future prospects of human obesity genetics . Biochem Genet . 2016 ; 54 : 1 - 28 .
6. Cao H. Adipocytokines in obesity and metabolic disease . J Endocrinol . 2014 ; 220 : T47 - 59 .
7. Khan M , Joseph F. Adipose tissue and adipokines: the association with and application of adipokines in obesity . Scientifica (Cairo) . 2014 ; 2014 : 328592 .
8. Maury E , Brichard SM . Adipokine dysregulation, adipose tissue inflammation and metabolic syndrome . Mol Cell Endocrinol . 2010 ; 314 : 1 - 16 .
9. Huh JY , Park YJ , Ham M , Kim JB . Crosstalk between adipocytes and immune cells in adipose tissue inflammation and metabolic dysregulation in obesity . Mol Cells . 2014 ; 37 : 365 - 71 .
10. Feuerer M , Herrero L , Cipolletta D , Naaz A , Wong J , Nayer A , Lee J , Goldfine AB , Benoist C , Shoelson S , Mathis D . Lean, but not obese, fat is enriched for a unique population of regulatory T cells that affect metabolic parameters . Nat Med . 2009 ; 15 : 930 - 9 .
11. Kucharska AM , Pyrzak B , Demkow U . Regulatory T cells in obesity . Adv Exp Med Biol . 2015 ; 866 : 35 - 40 .
12. Gyllenhammer LE , Lam J , Alderete TL , Allayee H , Akbari O , Katkhouda N , Goran MI . Lower omental t-regulatory cell count is associated with higher fasting glucose and lower beta-cell function in adults with obesity . Obesity (Silver Spring) . 2016 ; 24 : 1274 - 82 .
13. Alberti KG , Zimmet P , Shaw J. The metabolic syndrome-a new worldwide definition . Lancet . 2005 ; 366 : 1059 - 62 .
14. Schaffler A , Muller-Ladner U , Scholmerich J , Buchler C . Role of adipose tissue as an inflammatory organ in human diseases . Endocr Rev . 2006 ; 27 : 449 - 67 .
15. Sam S , Haffner S , Davidson MH , D'Agostino RB Sr , Feinstein S , Kondos G , Perez A , Mazzone T. Relation of abdominal fat depots to systemic markers of inflammation in type 2 diabetes . Diabetes Care . 2009 ; 32 : 932 - 7 .
16. Rocha VZ , Folco EJ . Inflammatory concepts of obesity . Int J Inflam . 2011 ; 2011 : 529061 .
17. Bell DS . Changes seen in gut bacteria content and distribution with obesity: causation or association? Postgrad Med . 2015 ; 127 : 863 - 8 .
18. Cani PD , Osto M , Geurts L , Everard A . Involvement of gut microbiota in the development of low-grade inflammation and type 2 diabetes associated with obesity . Gut Microbes . 2012 ; 3 : 279 - 88 .
19. Ley RE , Backhed F , Turnbaugh P , Lozupone CA , Knight RD , Gordon JI . Obesity alters gut microbial ecology . Proc Natl Acad Sci U S A . 2005 ; 102 : 11070 - 5 .
20. John GK , Mullin GE . The gut microbiome and obesity . Curr Oncol Rep . 2016 ; 18 : 45 .
21. Dahiya DK , Renuka PM , Shandilya UK , Dhewa T , Kumar N , Kumar S , Puniya AK , Shukla P . Gut microbiota modulation and its relationship with obesity using prebiotic fibers and probiotics: a review . Front Microbiol . 2017 ; 8 : 563 .
22. Sanchez B , Delgado S , Blanco-Miguez A , Lourenco A , Gueimonde M , Margolles A. Probiotics, gut microbiota, and their influence on host health and disease . Mol Nutr Food Res . 2017 ; 61
23. Gauffin Cano P , Santacruz A , Moya A , Sanz Y. Bacteroides uniformis CECT 7771 ameliorates metabolic and immunological dysfunction in mice with high-fat-diet induced obesity . PLoS One . 2012 ; 7 : e41079 .
24. Ji YS , Kim HN , Park HJ , Lee JE , Yeo SY , Yang JS , Park SY , Yoon HS , Cho GS , Franz CM , et al. Modulation of the murine microbiome with a concomitant anti-obesity effect by Lactobacillus rhamnosus GG and Lactobacillus sakei NR28 . Benef Microbes . 2012 ; 3 : 13 - 22 .
25. Takemura N , Okubo T , Sonoyama K. Lactobacillus plantarum strain no. 14 reduces adipocyte size in mice fed high-fat diet . Exp Biol Med (Maywood) . 2010 ; 235 : 849 - 56 .
26. Shiby VK , Mishra HN . Fermented milks and milk products as functional foods-a review . Crit Rev Food Sci Nutr . 2013 ; 53 : 482 - 96 .
27. Wolfe BE , Dutton RJ . Fermented foods as experimentally tractable microbial ecosystems . Cell . 2015 ; 161 : 49 - 55 .
28. Tamang JP , Watanabe K , Holzapfel WH . Review: diversity of microorganisms in global fermented foods and beverages . Front Microbiol . 2016 ; 7 : 377 .
29. Devirgiliis C , Caravelli A , Coppola D , Barile S , Perozzi G . Antibiotic resistance and microbial composition along the manufacturing process of Mozzarella di Bufala Campana . Int J Food Microbiol . 2008 ; 128 : 378 - 84 .
30. Zanni E , Laudenzi C , Schifano E , Palleschi C , Perozzi G , Uccelletti D , Devirgiliis C . Impact of a complex food microbiota on energy metabolism in the model organism Caenorhabditis elegans . Biomed Res Int . 2015 ; 2015 : 621709 .
31. Kobyliak N , Conte C , Cammarota G , Haley AP , Styriak I , Gaspar L , Fusek J , Rodrigo L , Kruzliak P. Probiotics in prevention and treatment of obesity: a critical view . Nutr Metab (Lond) . 2016 ; 13 : 14 .
32. Kim B , Park KY , Ji Y , Park S , Holzapfel W , Hyun CK . Protective effects of Lactobacillus rhamnosus GG against dyslipidemia in high-fat diet-induced obese mice . Biochem Biophys Res Commun . 2016 ; 473 : 530 - 6 .
33. Orr JS , Kennedy AJ , Hasty AH . Isolation of adipose tissue immune cells . J Vis Exp . 2013 :e50707.
34. Strissel KJ , DeFuria J , Shaul ME , Bennett G , Greenberg AS , Obin MS. T-cell recruitment and Th1 polarization in adipose tissue during diet-induced obesity in C57BL/6 mice . Obesity (Silver Spring) . 2010 ; 18 : 1918 - 25 .
35. Wagner Mackenzie B , Waite DW , Taylor MW . Evaluating variation in human gut microbiota profiles due to DNA extraction method and inter-subject differences . Front Microbiol . 2015 ; 6 : 130 .
36. Milani C , Hevia A , Foroni E , Duranti S , Turroni F , Lugli GA , Sanchez B , Martin R , Gueimonde M , van Sinderen D , et al. Assessing the fecal microbiota: an optimized ion torrent 16S rRNA gene-based analysis protocol . PLoS One . 2013 ; 8 : e68739 .
37. Caporaso JG , Kuczynski J , Stombaugh J , Bittinger K , Bushman FD , Costello EK , Fierer N , Pena AG , Goodrich JK , Gordon JI , et al. QIIME allows analysis of high-throughput community sequencing data . Nat Methods . 2010 ; 7 : 335 - 6 .
38. Edgar RC . Search and clustering orders of magnitude faster than BLAST . Bioinformatics . 2010 ; 26 : 2460 - 1 .
39. Quast C , Pruesse E , Yilmaz P , Gerken J , Schweer T , Yarza P , Peplies J , Glockner FO . The SILVA ribosomal RNA gene database project: improved data processing and web-based tools . Nucleic Acids Res . 2013 ; 41 : D590 - 6 .
40. Lozupone C , Knight R. UniFrac : a new phylogenetic method for comparing microbial communities . Appl Environ Microbiol . 2005 ; 71 : 8228 - 35 .
41. Hammer Ø , Harper DAT , Ryan PD . PAST: paleontological statistics software package for education and data analysis . Palaeontol Electron . 2001 ; 4 : 9pp .
42. Tamang JP , Shin DH , Jung SJ , Chae SW . Functional properties of microorganisms in fermented foods . Front Microbiol . 2016 ; 7 : 578 .
43. Bordoni A , Danesi F , Dardevet D , Dupont D , Fernandez AS , Gille D , Nunes Dos Santos C , Pinto P , Re R , Remond D , et al. Dairy products and inflammation: a review of the clinical evidence . Crit Rev Food Sci Nutr . 2017 ; 57 : 2497 - 525 .
44. Rosa DD , Dias MM , Grzeskowiak LM , Reis SA , Conceicao LL , Peluzio MD . Milk kefir: nutritional, microbiological and health benefits . Nutr Res Rev . 2017 : 1 - 15 .
45. Kim DH , Kim H , Jeong D , Kang IB , Chon JW , Kim HS , Song KY , Seo KH . Kefir alleviates obesity and hepatic steatosis in high-fat diet-fed mice by modulation of gut microbiota and mycobiota: targeted and untargeted community analysis with correlation of biomarkers . J Nutr Biochem. 2017 ; 44 : 35 - 43 .
46. Choi JW , Kang HW , Lim WC , Kim MK , Lee IY , Cho HY . Kefir prevented excess fat accumulation in diet-induced obese mice . Biosci Biotechnol Biochem . 2017 ; 81 : 958 - 65 .
47. Foligne B , Parayre S , Cheddani R , Famelart MH , Madec MN , Ple C , Breton J , Dewulf J , Jan G , Deutsch SM . Immunomodulation properties of multispecies fermented milks . Food Microbiol . 2016 ; 53 : 60 - 9 .
48. Alard J , Lehrter V , Rhimi M , Mangin I , Peucelle V , Abraham AL , Mariadassou M , Maguin E , Waligora-Dupriet AJ , Pot B , et al. Beneficial metabolic effects of selected probiotics on diet-induced obesity and insulin resistance in mice are associated with improvement of dysbiotic gut microbiota . Environ Microbiol . 2016 ; 18 : 1484 - 97 .
49. Cano PG , Santacruz A , Trejo FM , Sanz Y. Bifidobacterium CECT 7765 improves metabolic and immunological alterations associated with obesity in high-fat diet-fed mice . Obesity (Silver Spring) . 2013 ; 21 : 2310 - 21 .
50. Moya-Perez A , Neef A , Sanz Y. Bifidobacterium pseudocatenulatum CECT 7765 reduces obesity-associated inflammation by restoring the lymphocytemacrophage balance and gut microbiota sstructure in high-fat diet-fed mice . PLoS One . 2015 ; 10 : e0126976 .
51. Wajchenberg BL , Giannella-Neto D , da Silva ME , Santos RF . Depot-specific hormonal characteristics of subcutaneous and visceral adipose tissue and their relation to the metabolic syndrome . Horm Metab Res . 2002 ; 34 : 616 - 21 .
52. Novotny Nunez I , Maldonado Galdeano C , de Moreno de LeBlanc A , Perdigon G . Lactobacillus casei CRL 431 administration decreases inflammatory cytokines in a diet-induced obese mouse model . Nutrition . 2015 ; 31 : 1000 - 7 .
53. Nishimura S , Manabe I , Nagasaki M , Eto K , Yamashita H , Ohsugi M , Otsu M , Hara K , Ueki K , Sugiura S , et al. CD8+ effector T cells contribute to macrophage recruitment and adipose tissue inflammation in obesity . Nat Med . 2009 ; 15 : 914 - 20 .
54. Ahmed M , Gaffen SL . IL-17 in obesity and adipogenesis . Cytokine Growth Factor Rev . 2010 ; 21 : 449 - 53 .
55. Boi SK , Buchta CM , Pearson NA , Francis MB , Meyerholz DK , Grobe JL , Norian LA . Obesity alters immune and metabolic profiles. New insight from obeseresistant mice on high-fat diet . Obesity (Silver Spring) . 2016 ; 24 : 2140 - 9 .
56. Ukibe K , Miyoshi M , Kadooka Y. Administration of Lactobacillus gasseri SBT2055 suppresses macrophage infiltration into adipose tissue in dietinduced obese mice . Br J Nutr . 2015 ; 114 : 1180 - 7 .
57. Le Barz M , Anhe FF , Varin TV , Desjardins Y , Levy E , Roy D , Urdaci MC , Marette A . Probiotics as complementary treatment for metabolic disorders . Diabetes Metab J . 2015 ; 39 : 291 - 303 .
58. Turnbaugh PJ , Backhed F , Fulton L , Gordon JI . Diet-induced obesity is linked to marked but reversible alterations in the mouse distal gut microbiome . Cell Host Microbe . 2008 ; 3 : 213 - 23 .
59. Park S , Ji Y , Jung HY , Park H , Kang J , Choi SH , Shin H , Hyun CK , Kim KT , Holzapfel WH . Lactobacillus plantarum HAC01 regulates gut microbiota and adipose tissue accumulation in a diet-induced obesity murine model . Appl Microbiol Biotechnol . 2017 ; 101 : 1605 - 14 .
60. Gauffin-Cano PG , Santacruz A , Trejo FM , Sanz Y. Bifidobacterium CECT 7765 improves metabolic and immunological alterations associated with obesity in high-fat diet-fed mice . Obesity (Silver Spring) . 2013 ; 21 : 2310 - 21 .
61. Chan YK , Brar MS , Kirjavainen PV , Chen Y , Peng J , Li D , Leung FC , El-Nezami H . High fat diet induced atherosclerosis is accompanied with low colonic bacterial diversity and altered abundances that correlates with plaque size, plasma A-FABP and cholesterol: a pilot study of high fat diet and its intervention with Lactobacillus rhamnosus GG (LGG) or telmisartan in ApoE−/− mice . BMC Microbiol . 2016 ; 16 : 264 .
62. Thomas CM , Versalovic J . Probiotics-host communication: modulation of signaling pathways in the intestine . Gut Microbes . 2010 ; 1 : 148 - 63 .
63. Hsiao EY , McBride SW , Hsien S , Sharon G , Hyde ER , McCue T , Codelli JA , Chow J , Reisman SE , Petrosino JF , et al. Microbiota modulate behavioral and physiological abnormalities associated with neurodevelopmental disorders . Cell . 2013 ; 155 : 1451 - 63 .
64. Bjerg AT , Sorensen MB , Krych L , Hansen LH , Astrup A , Kristensen M , Nielsen DS . The effect of Lactobacillus paracasei subsp. paracasei and L. casei W8(R) on blood levels of triacylglycerol is independent of colonisation . Benef Microbes . 2015 ; 6 : 263 - 9 .
65. Sonnenburg ED , Smits SA , Tikhonov M , Higginbottom SK , Wingreen NS , Sonnenburg JL . Diet-induced extinctions in the gut microbiota compound over generations . Nature . 2016 ; 529 : 212 - 5 .
66. Leinonen R , Akhtar R , Birney E , Bower L , Cerdeno-Tarraga A , Cheng Y , Cleland I , Faruque N , Goodgame N , Gibson R , et al. The European Nucleotide Archive. Nucleic Acids Res . 2011 ; 39 : D28 - 31 .