Content uploaded by Jian Tan
Author content
All content in this area was uploaded by Jian Tan on Jun 26, 2016
Content may be subject to copyright.
Article
Dietary Fiber and Bacterial SCFA Enhance Oral
Tolerance and Protect against Food Allergy through
Diverse Cellular Pathways
Graphical Abstract
Highlights
dDietary fiber with vitamin A increases the potency of
tolerogenic CD103
+
DCs
dHigh-fiber diet protects mice against peanut allergy via gut
microbiota and SCFA
dHigh-fiber effects rely on epithelial GPR43 and immune cell
GPR109a
dDietary fiber promotes T
FH
and IgA responses
Authors
Jian Tan, Craig McKenzie,
Peter J. Vuillermin, ..., Reina E. Mebius,
Laurence Macia, Charles R. Mackay
Correspondence
laurence.macia@sydney.edu.au (L.M.),
charles.mackay@monash.edu (C.R.M.)
In Brief
Tan et al. examine the beneficial roles of
dietary fiber in peanut allergy using mice.
The authors find that this effect involves
reshaping of the gut microbiota as well as
increased levels of short-chain fatty acids
and activity of their receptors GPR43 and
GPR109a. High-fiber feeding also
increased tolerogenic CD103
+
DCs
activity, leading to increased Treg cell
differentiation.
Tan et al., 2016, Cell Reports 15, 2809–2824
June 21, 2016 ª2016 The Author(s).
http://dx.doi.org/10.1016/j.celrep.2016.05.047
Cell Reports
Article
Dietary Fiber and Bacterial SCFA Enhance
Oral Tolerance and Protect against Food Allergy
through Diverse Cellular Pathways
Jian Tan,
1
Craig McKenzie,
1
Peter J. Vuillermin,
2
Gera Goverse,
3
Carola G. Vinuesa,
4
Reina E. Mebius,
3
Laurence Macia,
1,5,6,7,
*and Charles R. Mackay
1,5,6,7,
*
1
Infection and Immunity Program, Monash Biomedicine Discovery Institute and Department of Biochemistry and Molecular Biology, Monash
University, Clayton, VIC 3800, Australia
2
School of Medicine, Deakin University, Waurn Ponds, VIC 3216, Australia
3
Department of Molecular Cell Biology and Immunology, VU University Medical Center, 1081 HZ Amsterdam, the Netherlands
4
Department of Pathogens and Immunity, John Curtin School of Medical Research, Australian National University, Building 131, Garran Road,
Canberra, ACT 0200, Australia
5
Charles Perkins Centre, The University of Sydney, Sydney, NSW 2006, Australia
6
Department of Physiology, Faculty of Medicine, The University of Sydney, Sydney, NSW 2006, Australia
7
Co-senior author
*Correspondence: laurence.macia@sydney.edu.au (L.M.), charles.mackay@monash.edu (C.R.M.)
http://dx.doi.org/10.1016/j.celrep.2016.05.047
SUMMARY
The incidence of food allergies in western countries
has increased dramatically in recent decades. Toler-
ance to food antigens relies on mucosal CD103
+
den-
dritic cells (DCs), which promote differentiation of
regulatory T (Treg) cells. We show that high-fiber
feeding in mice improved oral tolerance and protected
from food allergy. High-fiber feeding reshaped gut mi-
crobial ecology and increased the release of short-
chain fatty acids (SCFAs), particularly acetate and
butyrate. High-fiber feeding enhanced oral tolerance
and protected against food allergy by enhancing
retinal dehydrogenase activity in CD103
+
DC. This
protection depended on vitamin A in the diet. This
feeding regimen also boosted IgA production and
enhanced T follicular helper and mucosal germinal
center responses. Mice lacking GPR43 or GPR109A,
receptors for SCFAs, showed exacerbated food
allergy and fewer CD103
+
DCs. Dietary elements,
including fiber and vitamin A, therefore regulate
numerous protective pathways in the gastrointestinal
tract, necessary for immune non-responsiveness to
food antigens.
INTRODUCTION
Food allergy is now a major public health issue, due to its
increasing incidence over the past 20 years, particularly in west-
ern countries (Wang and Sampson, 2011). Food allergy develops
following loss of oral tolerance, which allows allergic sensitiza-
tion. However, the exact mechanisms whereby oral tolerance
is maintained, or lost, remain unclear. Excessive hygiene has
been evoked to explain increased allergy incidence (Strachan,
1989), but more recently, alterations in gut microflora composi-
tion have been suggested as an alternative explanation (Maslow-
ski and Mackay, 2011; Noverr and Huffnagle, 2005). This is
evident in germ-free mice, which tend to develop more-severe
allergies (Cerf-Bensussan and Gaboriau-Routhiau, 2010), and
given that specific probiotic treatment can alleviate food allergy
symptoms (Stefka et al., 2014).
Diet, especially consumption of dietary fiber, appears to be a
critical determinant for gut bacterial ecology, diversity, and func-
tion (De Filippo et al., 2010; Le Chatelier et al., 2013; Ou et al.,
2013; Turnbaugh et al., 2008). The gut microbiota promotes
epithelial integrity and regulatory T (Treg) cells, both critical for
mucosal homeostasis (Ahern et al., 2014; Atarashi et al., 2011;
Faith et al., 2014; Geuking et al., 2011; Macia et al., 2012). Die-
tary-fiber-derived metabolites have been implicated in gut ho-
meostasis and Treg cell biology (Arpaia et al., 2013; Furusawa
et al., 2013; Macia et al., 2015; Maslowski et al., 2009; Singh
et al., 2014; Smith et al., 2013). Dietary fiber is fermented in
the colon by anaerobic bacteria into short-chain fatty acids
(SCFAs), mainly acetate, butyrate, and propionate. These SCFAs
bind metabolite-sensing G-protein coupled receptors (GPCRs)
GPR43, GPR109A, and GPR41 with varying affinities. These
GPCRs are expressed on epithelial cells as well as immune cells.
Acetate has been shown to promote epithelial integrity in a
model of enteropathogenic infection, and SCFAs enhance gut
integrity in vitro (Fukuda et al., 2011). Western diets, typically
high in fat but also low in fiber, may therefore be associated
with changes in gut bacterial ecology, epithelial integrity, and
Treg cell development, and this may compromise oral tolerance
and allow for the development of food allergies.
Oral tolerance is a process through which systemic unrespon-
siveness to oral antigens arises. CD103-expressing dendritic
cells (DCs) (CD103
+
DCs) are present at high frequency in the
small intestine and migrate to the mesenteric lymph node
Cell Reports 15, 2809–2824, June 21, 2016 ª2016 The Author(s). 2809
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
(MLN) to initiate oral tolerance (Jaensson et al., 2008; Scott et al.,
2011). CD103
+
DCs express the enzyme retinaldehyde dehydro-
genase-2 (RALDH2) (encoded by Aldh1a2). RALDH2 converts
vitamin A to retinoic acid, which promotes the differentiation of
naive T cells into Treg cells (Coombes and Powrie, 2008; Iwata
et al., 2004; Jaensson et al., 2008) and imprints the gut homing
receptor CCR9 on them. Alteration of RALDH activity is associ-
ated with impaired oral tolerance (Hall et al., 2011). The matura-
tion and maintenance of CD103
+
DCs and their tolerogenic
phenotype is highly dependent on environmental conditioning
factors present locally in the small intestine (Pabst and Mowat,
2012). Disruption of epithelial integrity and subsequent release
of interleukin-33 (IL-33) and excess TSLP skews DCs toward
a Th2-like phenotype that elicits allergic sensitization (Iijima
et al., 2014; Paul and Zhu, 2010). Moreover, intestinal inflamma-
tion can abrogate the ability of CD103
+
DCs to promote Treg cell
differentiation (Laffont et al., 2010).
Immunoglobulin A (IgA) production also contributes to mucosal
immunity and oral tolerance. The gut microbiota promotes host
IgA responses, which in turn select for a microbiota composition
that promotes host mucosal homeostasis in a reciprocal positive
feedback loop (Kawamoto et al., 2014). This process is regulated
by T follicular helper (T
FH
) and T follicular regulatory cells (T
FR
)
(Kawamoto et al., 2014). IgA deficiency has been linked to exac-
erbated colitis (Cao et al., 2012) and systemic inflammation
(Kawamoto et al., 2012) and is also likely to be associated with
allergic diseases (Berin, 2012; Kukkonen et al., 2010).
In the present study, we report that dietary fiber together with
vitamin A plays a key role in promoting CD103
+
DC function, oral
tolerance, and protection from food allergy. These findings sup-
port the notion that diets deficient in fiber, typical of many west-
ern countries, could underlie the rise of food allergies in recent
decades.
RESULTS
High-Fiber Feeding Enhances CD103
+
DC Activity
Mucosal CD103
+
DCs are described as master regulators of im-
mune tolerance through their capacity to promote the differenti-
ation of naive T cells into Treg cells in the MLN (Scott et al., 2011).
We fed mice for at least 2 weeks on diets either depleted or en-
riched in fiber, which we have shown previously to alter SCFA
levels (Macia et al., 2015). We found that mice fed on a high-fiber
(HF) diet had a similar proportion and total number of CD103
DCs (Figure S1A) but a significantly higher proportion of
CD103
+
DCs (Figure 1A) in the MLN compared to zero-fiber
(ZF)-diet-fed mice. No differences in total cell number of
CD103
+
DCs were observed (Figure S1B). Increased proportion
of CD103
+
DCs also correlated with increased expression of key
tolerogenic genes Aldh1a2 as well as Ido in the MLN (Figure S1C),
both associated with immune tolerance, mediated by mucosal
CD103
+
, but not CD103
DCs (Feng et al., 2010; Matteoli
et al., 2010; van der Marel et al., 2007).
DCs analyzed from the MLN of HF-fed mice exhibited greater
enzymatic RALDH activity compared to ZF-fed mice as deter-
mined by ALDEFLUOR assay (Figure S1D), and this was
CD103
+
DC specific (Figure 1B). These did not correlate to
increased Aldh1a2 gene expression in purified CD103
+
DCs (Fig-
ure S1E), suggesting post-transcriptional effect of fiber on
RALDH activity. To study the tolerogenic properties of CD103
+
DCs in HF-fed mice, we sorted CD103
+
DCs from mice fed on
a ZF or HF diet and co-cultured them with CD4
+
CD25
CD62L
+
naive T cells derived from OT-II mice in the presence of oval-
bumin (OVA) peptide. CD103
+
DCs derived from HF-diet-fed
mice were more potent in converting naive T cells to FoxP3
+
Treg cells, as well as inducing greater expression of the gut
homing receptor CCR9, and these phenotypes were abrogated
in the presence of retinoic acid receptor (RAR) signaling inhibitor
LE540 (Figures 1C and 1D). Thus, HF diet feeding enhances the
tolerogenic activity of CD103
+
DCs, which is dependent on the
retinoic acid signaling pathway.
We next determined whether these effects could translate to an
antigen-specific tolerogenic phenotype in vivo. We adoptively
transferred carboxyfluorescein succinimidyl ester (CFSE)-labeled
OT-II CD4
+
T cells into mice fed on either a ZF or HF diet and then
orally challenged them with OVA. T cell proliferation was analyzed
by flow cytometry 72 hr later. CFSE-stained OT-II CD4
+
T cells
isolated from HF-challenged mice proliferated significantly less
compared to ZF-fed mice (Figure1E). This decreased proliferation
was associated with an increased proportion of antigen-specific
Treg cells in the MLN (Figure 1F). Naive T cells isolated from ZF-
and HF-fed mice had similar profiles of proliferation when stimu-
lated with anti-CD3 and anti-CD28 in vitro, suggesting that the
anti-proliferative effect of HF feeding was likely not through a
direct inhibition of naive T cell proliferation (Figure S1F). Thus,
HF feeding can influence the activity of CD103
+
DCs by upregu-
lating RALDH activity and enhancing antigen-specific Treg cell
responses.
HF Feeding Enhances Oral Tolerance and Protects
against Food Allergy
To examine whether the effects of dietary fiber on CD103
+
DC
activity and their capacity to induce Treg cells could enhance
tolerance in vivo, we subjected ZF- or HF-diet-fed mice to a
model of oral tolerance involving multiple challenges with peanut
extract (Shan et al., 2013) as depicted in Figure 2A. Compared
to ZF-diet-fed mice, antigen-challenged HF-diet-fed mice con-
tained a greater proportion of CD103
+
DCs in the MLN (Fig-
ure 2B), as well as an increased proportion of Treg cells that
expressed higher levels of CCR9 (Figure 2C). Antigen challenge
promoted the migration of CD103
+
DCs to the MLN in both diets
(as determined by the proportion and total number of CD103
+
DCs between two different time points at days 4 and 7 of
the model; Figure S2A), suggesting that increased Treg cell in-
duction was directly due to DCs themselves. Total cell numbers
in the MLN of HF-diet-fed mice were also significantly lower
when compared to ZF-fed mice (Figure S2B), suggesting a
diminished inflammatory response. Positive effects of dietary
fiber on total cell number, Treg cell induction, and increased
CCR9 expression appeared to be mucosal specific as no differ-
ences were observed in the spleen (Figures S2C–S2E).
Food allergy is one of the consequences of a breakdown of
oral tolerance. We subjected mice to an established model of
peanut allergy (Li et al., 2000), as depicted in Figure 2D. Mice
fed on HF diet showed significantly reduced clinical symptoms
of anaphylaxis, which correlated with lower levels of serum IgE
2810 Cell Reports 15, 2809–2824, June 21, 2016
A
B
C
E
D
F
Figure 1. HF Feeding Enhances CD103
+
DCs Activity
(A) Representative flow cytometry plot of CD11c
+
CD103
+
DCs of MHCII
+
cells in the MLN and graph representing the percentage of these cells in ZF- and HF-fed
mice.
(B) Representative flow cytometry plot of ALDEFLUOR expression in CD103
+
DCs in the MLN and corresponding graph between ZF- and HF-fed mice as
determined by ALDEFLUOR assay kit. ALDH inhibitor diethylaminobenzaldehyde (DEAB) was used to determine baseline background fluorescence.
(C and D) Representative flow cytometry plot of (C) FoxP3
+
cells after 5 days of co-culture of CD4
+
CD25
CD62L
+
naive OT-II T cells with MHCII
+
CD11c
+
CD103
+
DCs sorted from MLN of ZF- or HF-fed mice. (D) Corresponding mean fluorescence intensity (MFI) of CCR9 expression on gated cells is shown. Results are
representative of pooled triplicates and of at least three different cultures.
(E and F) 10
6
CD4
+
OT-II cells were adoptively transferred into ZF- or HF-diet-fed mice and mice challenged with 20 mg OVA orally 24 hr later. (E) Representative
histogram plot of CFSE and (F) FoxP3 expression in adoptively transferred CD4
+
OT-II T cells isolated from the MLN of ZF- or HF-diet-fed mice as determined by
flow cytometry 3 days after OVA challenge are shown.
All data are representative of at least two independent experiments. All data are represented as mean ±SEM of at least n = 5 mice per group; *p < 0.05; **p < 0.01;
***p < 0.005; ****p < 0.001.
Cell Reports 15, 2809–2824, June 21, 2016 2811
B
AC
E
D
FG
HI
J
Figure 2. HF Feeding Enhances Oral Tolerance and Protects against Food Allergy
(A) Experimental model of oral tolerance.
(B and C) Proportion of (B) MHCII
+
CD11c
+
CD103
+
DCs and (C) CD4
+
CD25
+
FoxP3
+
Treg cells (left) and mean fluorescence intensity of CCR9 on Treg cells (right)
in the MLN of ZF- and HF-fed mice 3 days after tolerance induction.
(D) Experimental model of food allergy.
(legend continued on next page)
2812 Cell Reports 15, 2809–2824, June 21, 2016
when compared to ZF-fed mice (Figure 2E). These improved fea-
tures were paralleled with a higher proportion of CD103
+
DCs as
well as Treg cells in the MLN (Figure 2F), which also correlated
with lower total cell numbers in the MLN (Figure 2G). Protection
from food allergy under HF diet feeding conditions was also
associated with decreased production of Th2 cytokines IL-4,
IL-5, and IL-13 from lymphocytes isolated from the MLN stimu-
lated with peanut extract in vitro (Figure S2F). A greater propor-
tion and total number of Treg cells were also observed in the
small intestine lamina propria of HF-fed mice compared to
ZF-fed mice (Figure S2G).
We next depleted Treg cells using anti-CD25-depleting anti-
body before subjecting mice to the food allergy model. Treg
cell depletion was confirmed by flow cytometry (Figure S2H).
Absence of Treg cells abrogated the protective effects of HF
diet feeding in food allergy, with increased anaphylaxis scores
and total serum IgE in mice treated with anti-CD25 compared
to isotype controls, both fed on a HF diet (Figure 2H). This result
is consistent with the notion that Treg cells are necessary for
protection against food allergy and other allergic diseases (Palo-
mares et al., 2010).
The intestinal epithelium may play a role in dictating tolerance
by conditioning mucosal DCs toward a tolerogenic or inflamma-
tory phenotype. Epithelial production of cytokines and other fac-
tors, such as IL-25, IL-33, TSLP, and granulocyte macrophage
colony-stimulating factor (GM-CSF), is critical in balancing toler-
ance versus Th2 responses by DCs. We examined whether
altered SCFA production could impair epithelial homeostasis
and promote Th2 skewing. qPCR analysis of small intestinal
jejunum revealed that ZF-fed mice exhibited greater gene
expression of Tslp and Il33 compared to HF-diet-fed mice (Fig-
ure 2I), whereas no differences in gene expression of Il25,Gmcsf,
and Muc2 were observed (data not shown). Gut epithelial perme-
ability appeared to be increased under ZF-fed conditions as
shown by significantly greater infiltration of bacteria to the
MLN in ZF-diet-fed mice (Figure 2J). We did not observe
morphological changes in the small intestine between ZF- versus
HF-diet-fed mice in cholera-toxin-induced inflammation (Fig-
ure S2I). Altogether, these results establish a role for dietary fiber
in intestinal homeostasis and epithelial integrity, both of which
are important for maintenance of oral tolerance.
HF-Mediated Protection against Food Allergy Relies on
Vitamin A Metabolism
The activity of RALDH is dependent on the availability of its sub-
strate, vitamin A, and thus determines the tolerogenic capability
of CD103
+
DCs (Molenaar et al., 2011). Therefore, we deter-
mined whether the tolerogenic effects of HF feeding relied on
vitamin A metabolism. To achieve this, mice were placed on
either a control (AIN93G) or vitamin-A-deficient diet (VAD) for at
least 12 weeks and then switched to a HF diet or a vitamin-A-
deficient HF diet (VAD-HF), respectively, for at least 2 weeks
prior to experimentations. To validate vitamin A availability on
mucosal DC function, we performed the ALDEFLUOR assay on
these mice. As expected, HF-derived DCs exhibited significantly
greater RALDH activity compared to control diet, as well as to
VAD and VAD-HF diet (Figure S3A). Both HF and VAD-HF diet
feeding were also linked to greater SCFA production compared
to control and VAD diet feeding, demonstrating the ability of
VAD-HF diet to still promote SCFA production (Figure S3B).
Unlike mice fed on a HF diet, mice fed on a VAD-HF diet (and
also control and VAD diet) showed significantly exacerbated
clinical symptoms of anaphylaxis as well as increased, but not
significant, total serum IgE levels (Figure 3A). These results also
correlated with a reduced proportion of CD103
+
DCs in VAD-
HF-fed mice when compared to HF-fed mice (Figure 3B). Surpris-
ingly, despite differences in proportion of CD103
+
DCs between
HF- and VAD-HF-diet-fed mice,both had similarly greater propor-
tions of Treg cells than mice fed on a control or VAD diet (Fig-
ure 3C), suggesting a proportion of Treg cell induction by HF
occurred independently of CD103
+
DCs. Despite similar numbers
of MLN Treg cells in VAD-HF- and HF-fed groups, total cell
numbers were significantlydecreased in MLN of HF-fed mice (Fig-
ure 3D). To determine whether the Treg cells were functional in
VAD-HF mice, we performed an in vivo proliferation assay utilizing
OT-II T cells as performed previously. Adoptively transferredCD4
+
T cells derived from OT-II mice proliferated significantly more in
VAD-HF compared to HF diet feeding (Figure 3E), suggesting
that Treg cells from VAD-HF-fed mice might be defective.
Unweighted principal coordinate analysis of UNIFRAC dis-
tances (PCoA) revealed that microbiota composition of mice
fed on a HF diet most closely resembled microbiota composition
of mice fed on a VAD-HF diet (which is identical to HF diet but
deficient in vitamin A), suggesting that bacterial composition
was largely dependent on the dietary fiber content rather than
on vitamin A (Figure 3F). Consistent with this, microbiota compo-
sition of mice fed on a control diet was similar to mice fed on a
VAD diet (which is compositionally similar to control diet but defi-
cient in vitamin A). Interestingly, mice fed on a ZF diet had the
most unique microbiota but were closer to microbiota of mice
fed on a control and VAD diet (Figure 3F). Supporting microbiota
analyses are presented in Figures S3C and S3D.
HF Feeding Promotes IgA Responses
One indicator of gut dysbiosis is the increase of IgA-coated bac-
teria. This likely represents an effort by the host immune system
(E–G) ZF- or HF-diet-fed mice were induced food allergy and (E) anaphylaxis score determined after PE challenge on day 28 (left) and total serum IgE determined
by ELISA (right). (F) Proportion of MLN MHCII
+
CD11c
+
CD103
+
DCs and CD4
+
CD25
+
FoxP3
+
Treg cells of ZF- and HF-fed mice determined by flow cytometry is
shown. (G) Numeration of total cells in the MLN is shown.
(H) Intraperitoneal (i.p.) administration of either 200 mgofa-CD25-depleting antibody or isotype control in HF-fed mice at days 3 and 4 of the peanut allergy
model. Anaphylaxis score (left) and total serum IgE determined by ELISA (right) after PE challenge on day 28 are shown.
(I) Relative gene expression of Tslp and Il33 to housekeeping gene Rpl13a in the jejunum of ZF- and HF-fed mice under basal conditions determined by qPCR.
(J) Bacteria count in MLN organ (colony-forming unit [CFU]/mg) from ZF- or HF-fed mice under basal conditions after overnight culture.
All data are representative of at least two independent experiments. All data are represented as mean ±SEM of n = at least 3 (B and C) and n = 6–8 (E–G) mice per
group; *p < 0.05; **p < 0.01; ***p < 0.005; ****p < 0.001.
Cell Reports 15, 2809–2824, June 21, 2016 2813
A
BCD
E
F
Figure 3. HF-Mediated Protection against Food Allergy Relates to Vitamin A Metabolism
(A–D) Control, VAD-, HF-, or VAD-HF-diet-fed mice were induced food allergy. (A) Anaphylaxis score determined after PE challenge on day 28 (left) and total
serum IgE concentration determined by ELISA (right) are shown. (B) Proportions of MLN MHCII
+
CD11c
+
CD103
+
DCs and (C) CD4
+
CD25
+
FoxP3
+
Treg cells
between ZF- and HF-fed mice were determined by flow cytometry. (D) Numeration of total cell numbers in MLN is shown.
(E) 10
6
CD4
+
OT-II cells were adoptively transferred into HF- or VAD-HF-diet-fed mice and challenged with 20 mg OVA orally 24 hr later. Proliferation of CD4
+
OT-II
T cells by CFSE staining was analyzed by flow cytometry 3 days after OVA challenge.
(legend continued on next page)
2814 Cell Reports 15, 2809–2824, June 21, 2016
to exclude unfavorable bacteria and is a phenomenon often
observed in patients with inflammatory bowel disease (Palm
et al., 2014). We consistently observed a greater percentage of
IgA-coated bacteria in feces isolated from ZF compared to HF-
fed mice (Figure 4A), suggesting the occurrence of dysbiosis in
ZF-fed mice. In addition to differences in microbiota composition
in colon, metagenomics analyses also revealed contrasting dif-
ferences in small intestinal microbiota composition, as demon-
strated by unweighted PCoA (Figure 4B). In accordance with
beneficial effects of HF diet feeding on gut homeostasis, mice
fed on a HF diet had greater levels of serum IgA compared to
ZF-fed mice (Figure 4C). As T
FH
and T
FR
responses have been
implicated in microbiota-induced IgA responses (Hirota et al.,
2013; Kawamoto et al., 2014), we evaluated whether increased
production of IgA also related to changes in T
FH
and T
FR
responses. Increased IgA in HF-fed mice was associated with
a significantly higher proportion of T
FH
cells, as defined by
CXCR5 and PD-1 expression on CD4
+
FoxP3
cells compared
to ZF-diet-fed mice (Figure 4D). This was also associated with
a higher proportion of T
FR
cells (Figure S4A), as defined by
CXCR5 and PD-1 expression on CD4
+
FoxP3
+
cells. These re-
sponses were also elevated in the MLN of HF-fed mice (Figures
S4B and S4C). Consistent with these data, mice fed on a HF diet
had a greater germinal center response, with a greater propor-
tion of B220
+
GL7
+
CD95
+
germinal center B cells in the Peyer’s
patches (PPs) (Figure 4E) and also the MLN (Figure S4D).
Germinal center reactions are required for generating IgA
+
plas-
mablasts, which mature into IgA-producing plasma cells that
migrate to the small intestinal lamina propria (Pabst, 2012).
Increased germinal center reactions in HF-fed mice indeed re-
sulted in a significantly higher proportion of IgA
+
B220
IgA-pro-
ducing plasma cells compared to ZF-fed mice (Figure 4F).
To confirm whether changes to microbiota composition due
to HF diet feeding could be responsible for the increase in IgA
production, we inoculated germ-free mice with microbiota
from ZF- versus HF-diet-fed mice. Mice reconstituted with a
HF microbiota (HFM) had significantly increased levels of IgA
compared to mice reconstituted with a ZF microbiota (ZFM) (Fig-
ure 4G), despite the fact that both groups of mice were fed on
normal chow post-reconstitution. To exclude a possible role of
SCFAs, we measured and found no observable differences in
SCFA levels in feces at 4 weeks post-reconstitution (Figure S4E).
Increased IgA in HFM mice was paralleled by increased pro-
portions of both T
FH
and T
FR
cells in the PP when compared to
ZFM mice (Figure S4F). Finally, to confirm that increased IgA
observed in HF-fed animals were linked to changes to the gut mi-
crobiota, we studied IgA levels in offspring, inheriting mother’s
microbiota at birth, in mothers either fed on ZF or HF diets. Vagi-
nally born offspring from HF-fed mothers also had higher levels
of serum IgA compared to offspring of mothers that were fed a
ZF diet (Figure 4H) whereas no differences were observed
when offspring were born by Cesarean section and fostered by
mothers fed on a control diet (Figure S4G). Effects of HF are
mediated by microbiota rather than SCFAs, as treatment with
acetate, propionate, or butyrate failed to induce increased IgA
levels in mice fed on a control diet (Figure S5C). Collectively,
the above data show that a HF diet promoted mucosal homeo-
stasis by inducing T
FH
-IgA pathways and that this related to
gut microbiota composition.
Gut Microbiota Composition and SCFAs Protect against
Food Allergy
To determine whether changes to gut microbiota composition
could account for the protective role of HF feeding in food
allergy, we reconstituted germ-free mice with either a ZF or
HFM. We found that, while fed on a normal chow, HFM mice
had significantly better clinical anaphylaxis scores as well as a
trend of decreases in total serum IgE levels (Figure 5A). HFM
mice had a greater proportion of Treg cells compared to ZFM
mice (Figure 5B). More surprisingly, no differences in proportion
of CD103
+
DCs or total cell numbers in the MLN were observed
under allergic conditions (Figures 5B and 5C), suggesting that
changes to microbiota itself were not fully capable of replicating
the full effects of HF diet feeding. To gain further insight into the
mechanisms involved, we analyzed feces from recipient mice
4 weeks post-reconstitution and compared them to feces
collected from donor mice. Results from unweighted PCoA re-
vealed contrasting differences. HFM inoculated into recipient
germ-free mice were much more stable and closely represented
those of donor mice, whereas the microbiota composition of ZF-
recipient mice differed markedly from donor mice (Figure 5D).
Supporting microbiota analyses are presented in Figure S5A.
Because dietary fiber is fermented by the gut microbiota into
SCFAs, we next determined whether SCFAs could directly
mediate the beneficial effects of fiber in our food allergy model.
We fed mice with acetate, butyrate, or propionate in drinking wa-
ter for 3 weeks prior to initial sensitization and throughout the
experiment. Protection against food allergy was observed in
mice treated with acetate and butyrate, but not propionate,
with a reduction seen in anaphylaxis clinical scores as well as
total serum IgE levels (Figure 5E). Similar to HF-diet feeding, pro-
tection against food allergy in acetate- and butyrate-treated
mice correlated with the induction of CD103
+
DCs and Treg
cell responses in the MLN (Figure 5F). These results also corre-
lated with lower total cell numbers in the MLN (Figure 5G). We
performed metagenomics studies on mice treated with SCFAs
in drinking water. Unweighted PCoA revealed no apparent
change to microbiota composition when mice were supple-
mented with SCFAs in drinking water (Figure S5B), establishing
the direct effect of SCFAs on food allergy. SCFA supplementa-
tion in drinking water did not alter IgA levels or T
FH
/T
FR
responses
(Figures S5C and S5D). To further distinguish the beneficial ef-
fects of SCFAs versus changes to gut microbiota composition
on food allergy, we depleted microbiota from control mice using
(F) Fecal microbiota composition was analyzed by 16S rRNA metagenomic sequencing from feces of mice fed on either a control (AIN93G), VAD, HF, VAD- HF, or
ZF diet. Relative abundance of bacteria is presented at the family level (left), and comparison of microbial community diversity is prese nted as unweighted PCoA
plots (right). Supporting microbiota analyses are presented in Figures S3C and S3D.
All data are representative of at least two independent experiments. All data are represented as mean ±SEM of n = 5–7 mice per group; *p < 0.05; **p < 0.01;
***p < 0.005; ****p < 0.001.
Cell Reports 15, 2809–2824, June 21, 2016 2815
AB
DC
E
F
GH
Figure 4. HF Feeding Promotes IgA Responses
(A) Proportion of IgA-coated bacteria was determined by flow cytometry of IgA-stained bacteria isolated from feces of ZF- or HF-diet-fed mice.
(B) Fecal microbiota composition was analyzed by 16S rRNA metagenomic sequencing from small intestine content of mice fed on either a ZF or HF diet. Relative
abundance of bacteria is presented at the family level (left), and comparison of microbial community diversity is presented as unweighted PCoA plots (right).
(legend continued on next page)
2816 Cell Reports 15, 2809–2824, June 21, 2016
antibiotics and treated mice with a mixture of SCFAs containing
acetate, propionate, and butyrate to mimic HF-diet feeding. We
found that, despite supplementation with SCFAs at levels known
to have beneficial effects on the immune system, they did not
protect against food allergy in the absence of a gut microbiota
(Figures S5E–S5G).
One pathway by which the gut microbiota can protect against
food allergy involves direct signaling of bacterial products in the
host via the adaptor protein MYD88 (Pawar et al., 2015). HF
feeding in Myd88
/
did not protect against food allergy, as
shown by higher clinical anaphylaxis scores as well as serum
IgE levels compared to wild-type (WT) mice fed on a HF diet
(Figure 5H). Accordingly, the proportion of CD103
+
DCs were
significantly lower in Myd88
/
mice compared to WT mice (Fig-
ure 5I), although no differences in proportion of Treg cells or total
cell numbers in the MLN were observed (Figure 5J). Altogether,
these results demonstrate that the beneficial effects of fiber on
food allergy are mediated by SCFAs, but other elements of the
gut microbiota are necessary, such as ability to elicit MYD88
signaling.
GPR43 and GPR109a Are Required for Dietary-Fiber-
Mediated Protection against Food Allergy
We compared allergic responses in WT, Gpr43
/
, and
Gpr109a
/
mice fed on a HF diet. Under basal conditions,
both Gpr43
/
and Gpr109a
/
mice had lower expression of
Aldh1a2 transcript as well as RALDH activity when compared
to WT mice (Figures S6A and S6B). In the food allergy model,
both Gpr43
/
and Gpr109a
/
mice showed exacerbated clin-
ical anaphylaxis scores as well as total serum IgE levels, when
compared to WT mice, even under conditions of HF feeding
(Figure 6A). These results also correlated with impaired induction
of CD103
+
DC and Treg cell responses in the MLN (Figure 6B).
There were also greater total cell numbers in the MLN of
Gpr43
/
and Gpr109a
/
mice, compared to WT mice (Fig-
ure 6C). Consistent with our propionate data (Figure 5E), HF
diet in Gpr41
/
mice protected against food allergy equally
well as in WT mice (Figures S6C–S6E).
Epithelial integrity is linked to allergic disease. We observed
that both Gpr43
/
and Gpr109a
/
mice exhibited a signifi-
cantly greater infiltration of bacteria to the MLN when compared
to WT mice under basal conditions (Figure 6D). This suggested
impaired epithelial integrity, which may contribute to the pro-in-
flammatory phenotype in Gpr43
/
and Gpr109a
/
mice. We
measured intestinal permeability via fluorescein isothiocyanate
(FITC)-dextran permeability assay and found that Gpr43
/
mice exhibited significantly greater and Gpr109a
/
mice ex-
hibited a strong trend to an increase in small intestinal perme-
ability than WT mice (Figure 6E). We also examined TSLP, an
epithelial-derived factor that may condition mucosal DC to adopt
a Th2-skewing phenotype. We found that Gpr43
/
mice had
higher expression of Tslp over both WT and Gpr109a
/
mice,
whereas no differences in Il33 expression were found between
all groups of mice (Figure 6F). Additionally, we did not detect
morphological changes to the small intestine in histological anal-
ysis (Figure S6F).
We found that CD103
+
DCs as well as the small intestine
epithelium expressed transcripts for both Gpr43 and Gpr109a;
Gpr109a transcripts were much more highly expressed in
CD103
+
DCs whereas, for Gpr43, it was the epithelium (Fig-
ure S6G). We employed Gpr43-conditional gene-deficient mice
and found that expression of GPR43 on intestinal cells rather
than on immune cells was important for HF-mediated protec-
tion against food allergy (Figure 6G). Conversely, bone-marrow
chimera experiments showed that GPR109A expression on im-
mune cells was much more important for HF-mediated protec-
tion against food allergy than expression on non-immune cells
(including epithelium; Figure 6H). Consistent with this notion,
we found that butyrate was the most-potent SCFA in promoting
RALDH activity (Figure S6H). Lastly, we compared microbiota
composition between WT, Gpr43
/
, and Gpr109a
/
mice.
Deficiency in either GPR43 or GPR109A altered microbiota
composition greatly (Figure 6I). Unweighted pair-group method
of analysis (UPGMA) hierarchy illustrated that Gpr43
/
mice
shared greater similarity to WT mice, whereas Gpr109a
/
had
the most-unique microbiota composition (Figure S6I). Thus,
Gpr43
/
and Gpr109a
/
mouse phenotypes support the
fundamental role that SCFAs play in protection against food
allergy by influencing CD103
+
DC responses and gut epithelial
integrity.
DISCUSSION
Oral tolerance prevents the host from mounting an inappropriate
immune response to innocuous antigens. Instrumental in this are
CD103
+
DCs, which constitutively uptake luminal antigens and
initiate a tolerogenic response by promoting the differentiation
of Treg cells in the MLN (Hall et al., 2011; Pabst and Mowat,
2012). How this system might be disrupted in food allergies is un-
known. In the present work, we show that dietary fiber/SCFAs
together with vitamin A and a healthy gut microbiota maintain a
tolerogenic mucosal environment and protect against the devel-
opment of food allergy. This is achieved principally through
enhancement to tolerogenic CD103
+
DC functions. These re-
sponses were dependent on epithelial GPR43 and immune
GPR109A signaling as the full tolerogenic effect of dietary fiber
(C) Total serum IgA as measured by ELISA of mice fed on either a ZF or HF diet.
(D) Representative flow cytometry plot of CXCR5
+
PD1
hi
cells that are CD4
+
FoxP3
in the PP between ZF- and HF-diet-fed mice.
(E) Representative flow cytometry plot of CD95
+
GL-7
+
cells that are B220
+
in the PP between ZF- and HF-diet-fed mice.
(F) Representative flow cytometry plot of B220
IgA
+
plasma B cells in the small intestine lamina propria between ZF- and HF-diet-fed mice.
(G) Germ-free mice were inoculated with fecal suspension from mice fed a ZF or HF diet, and total serum IgA was tracked for a total of 4 weeks determined
by ELISA.
(H) Pregnant mother was fed on either a control or HF diet at estrous cycle 13 (E13), and offspring were born by Cesarean section at E20 and transferred to foster
mother, which were fed on a control diet. Serum IgA was measured by ELISA when offspring were at least 6 weeks of age.
All data are representative of at least two independent experiments. All data are represented as mean ±SEM; n = 4–7 mice per group; *p < 0.05; **p < 0.01.
Cell Reports 15, 2809–2824, June 21, 2016 2817
A
BC
D
E
FG
IHJ
(legend on next page)
2818 Cell Reports 15, 2809–2824, June 21, 2016
could not be replicated in full and partial GPR43 and GPR109a-
deficient mice. Both the presence and composition of the gut-
microbiota-affected allergic responses and MYD88 signaling
were additional important elements. HF feeding and associated
changes to the microbiota also boosted host IgA production,
T
FH
, and germinal center responses. These studies show that
diet, bacterial composition, and metabolites collectively account
for food allergy in mice and operate through several pathways.
Similar mechanisms may underlie the increased incidence of
food allergies in certain human populations.
Previous studies have suggested that an altered gut microbiota
may underlie allergies (Atarashi et al., 2011; Hill et al., 2012; Huff-
nagle, 2010; Noval Rivas et al., 2013; Noverr and Huffnagle, 2005;
Stefka et al., 2014). Regardless of the precise bacterial species
responsible for protection or susceptibility, the findings presented
here raise the possibility that a healthy gut microbiota and its
manipulation to efficient SCFA production may protect against
food allergies. Dysbiosis during infancy, particularly altered levels
of the SCFA producer Bifidobacterium, is associated with
increased risk of allergy (Bjo
¨rkste
´n et al., 2001). This bacterium
produces high levels of SCFAs and promotes epithelial integrity
(Fukuda et al., 2011). Bifidobacterium alleviates food allergy in
mice (Kim et al., 2016) and its administration to pregnant women
significantly reduced allergy in offspring, and this correlated with
changes in gut microbiota composition (Enomoto et al., 2014),
possibly through SCFA elevation. Oral immunotherapy with the
probiotic Lactobacillus Rhamnosus, a high-SCFA producer, led
to peanut unresponsiveness in 82% of allergic children (Tang
et al., 2015). Success rates might be even higher by combining
this treatment with dietary fiber as a prebiotic to increase SCFA
production and reshape gut microbiota composition.
HF feeding and associated changes to the microbiota
boosted host IgA production, and this correlated with enhanced
T
FH
/T
FR
numbers and germinal center responses. Diversified and
selected IgAs originating from mucosal germinal center reac-
tions contribute to maintenance of a diversified and balanced
microbiota (Kawamoto et al., 2014). This in turn facilitates the
expansion of Foxp3
+
T
FR
cells, induction of germinal centers,
and IgA responses in a symbiotic regulatory loop (Kawamoto
et al., 2014). It may be that this symbiotic loop, which is likely
influenced by intake of dietary fiber, is highly relevant to mucosal
tolerance to food antigens. It is not known exactly how IgA plays
a role in oral tolerance but may protect against food allergy by
neutralizing food antigens and limiting their access to the im-
mune system (Berin, 2012). Whereas there is no direct evidence
as yet that IgA plays a beneficial role in food allergy, deficiency in
IgA has been linked to various allergic disorders, particularly
food allergy in humans (Janzi et al., 2009; Kukkonen et al.,
2010; Yel, 2010). Children deficient in IgA develop significantly
more food allergies by the age of 4 (Tuano et al., 2015), and prev-
alence of celiac disease (which shares similarities with IgE-medi-
ated food allergy) is 15 times higher in IgA-deficient subjects
(Meini et al., 1996). Higher levels of IgA may be indicative of a
healthy mucosal immune system and be beneficial in protection
against food allergies. HF feeding during pregnancy, and promo-
tion of mucosal IgA responses, could protect against food allergy
development in offspring, as fecal IgA levels may be vertically
transmitted (Moon et al., 2015).
Our findings suggest that deficiency of dietary fiber by altering
gut microbiota and SCFAs may explain the increase in food al-
lergies in western countries. Daily consumption of dietary fiber
in the western world is well below the recommended levels, and
we have argued previously that insufficient intake of dietary fiber
may contribute to a range of ‘‘western lifestyle’’ inflammatory dis-
eases (Macia et al., 2012; Maslowski and Mackay, 2011; Thorburn
et al., 2014). However,other factors may contribute, such as emul-
sifiers used in processed foods, which adversely affect the
composition of the gut microbiota (Chassaing et al., 2015), or
the widespread use of antibiotics. Vitamin A is another dietary
element that contributes to oral tolerance, and we demonstrate
that lack of dietary vitamin A exacerbates foodallergy, even under
HF feeding. Vitamin A deficiency is mostly observed in developing
countries, and indeed, incidence of food allergies in these coun-
tries is higher than first appreciated (Gray et al., 2014). Whereas
vitamin A deficiencyappears rare in western countries, population
studies indicate an inadequate intake of vitamin A against recom-
mended levels in up to 5% of children (age 1–3) and up to 35% of
adults (age > 19) in Canada and the US (Fulgoni et al., 2011;Wilson
et al., 2013). Healthy adultscan mitigate short-term vitamin A defi-
ciency from vitamin A stores in their liver; however, infants and
children have a much-lower capacity for this. Food allergy is
most common among children (Koplin et al., 2011), and short-
term vitamin A deficiency at this age could have a significant
impact on the child’s oral tolerance. Oral tolerance is abrogated
in neonatal mice by physiological vitamin A deficiency (Turfkruyer
et al., 2016). Moreover, the pro form of vitamin A is abundant in
Figure 5. Gut Microbiota Composition and SCFAs Relate to Protection against Food Allergy
(A–C) Germ-free mice inoculated with ZF (ZFM) or HF (HFM) microbiota were induced food allergy. (A) Anaphylaxis score after PE challenge on day 28 (left) and
total serum IgE concentration measured by ELISA (right) are shown. (B) Proportions of MLN MHCII
+
CD11c
+
CD103
+
DCs and CD4
+
CD25
+
FoxP3
+
Treg cells of
ZFM and HFM mice determined by flow cytometry are shown. (C) Numeration of total cell numbers in MLN is shown.
(D) Fecal microbiota composition analyzed by 16S rRNA metagenomic sequencing in donor mice fed on either a ZF or HF diet and recipient germ-free mice
reconstituted with donor feces 4 weeks post-reconstitution (ZFM and HFM). Relative abundance of bacteria is presented at the family level (left), and compariso n
of microbial community diversity is presented as unweighted PCoA plots (right). Supporting microbiota analyses are presented in Figures S5A and S5B.
(E–G) Mice were administered sodium acetate, propionate, or butyrate in drinking water or drinking water alone for 3 weeks and induced food allergy. (E)
Anaphylaxis score determined after PE challenge on day 28 and total serum IgE concentration s determined by ELISA are shown. (F) Proportions of MLN
MHCII
+
CD11c
+
CD103
+
DCs and CD4
+
CD25
+
FoxP3
+
Treg cells determined by flow cytometry are shown. (G) Numeration of total cell numbers in MLN is shown.
(H–J) WT and Myd88
/
mice were fed on a HF diet and induced food allergy. (H) Anaphylaxis score was determined after PE challenge on day 28 and total serum
IgE concentration measured by ELISA. (I) Proportions of MHCII
+
CD11c
+
CD103
+
DCs and (J) CD4
+
CD25
+
FoxP3
+
Treg cells in the MLN were determined by flow
cytometry and numeration of total cell numbers in MLN.
All data are representative of at least two independent experiments. All data are represented as mean ±SEM of n = 4–6 mice; *p < 0.05; **p < 0.01; ***p < 0.005;
****p < 0.001.
Cell Reports 15, 2809–2824, June 21, 2016 2819
A
BC
DEF
GH
I
(legend on next page)
2820 Cell Reports 15, 2809–2824, June 21, 2016
vegetables, which are also a source of fiber. Deficiencies of fiber
or vitamin A even for short periods of time may predispose to
food allergy, if they occur at critical stages of development of
the mucosal immune system.
Collectively, our data support the notion that dietary elements,
including fiber and vitamin A, are essential for the tolerogenic
function of CD103
+
DCs and maintenance of mucosal homeo-
stasis, including proper IgA responses and epithelial barrier
function. The practical outcome of this is the promotion of oral
tolerance and protection from food allergy. The findings here
together with cellular and molecular mechanisms support the
use of dietary and probiotic approaches to prevent or treat
food allergies in humans.
EXPERIMENTAL PROCEDURES
Animals
Gpr43
/
mice (Deltagen) were crossed to a C57BL/6 background greater
than 13 generations. Gpr109a
/
, Villin-Cre.Gpr43
/
, Vav-Cre.Gpr43
/
mice were generated in house. Myd88
/
, C57BL/6-Tg(TcraTcrb)425Cbn/Crl
(OT-II), and germ-free mice were purchased from WEHI Institute. All mice
were on C57BL/6 background and maintained under specific pathogen-free
conditions. Female mice between 6 and 8 weeks were used for all experi-
ments. All experimental procedures involving mice were approved by the
Monash Animal Ethics Committee.
Diets and SCFA Treatments
Custom dietswere all based on modification to the AIN93G control diet and were
purchasedfrom Specialty Feeds. HF diet (SF11-029) is enriched in guar gum and
cellulose (35% crude fiber)and ZF diet (SF11-028) devoid of fiber or starch. Mice
were fed on these diets for 2 weeks prior and throughout experiments. VAD diet
(SF08-014)or VAD-HF diet (SF14-129) was fed to mice for at least 12 weeks prior
and throughout experiments. Sodium acetate, propionate, or butyrate (Sigma-
Aldrich) was administered in drinking water at 200 mM, 100 mM, and 100 mM,
respectively, for 3 weeks prior and throughout experiments.
Germ-free Mice Reconstitution with Microbiota
Total colonic and fecal contents from ZF-or HF-diet-fed mice wereresuspended
in sterile cold PBS at a concentration of 100 mg/ml and homogenized. Suspen-
sions werefiltered through a 70 mm cell strainerand 200 ml adm inistered intraga s-
trically to germ-free mice. Mice were used for experimentations 4 weeks later.
Bone Marrow Chimeras
Recipient 6- to 8-week-old mice were lethally irradiated with two doses of
4.5 Gy 4 hr apart. Ten million bone marrow cells isolated from sex-matched
6- to 10-week-old donor mice were injected intravenously 6 hr later. Mice
were used for experimentations 6 weeks later.
Bacteria 16S rRNA Sequencing and Bioinformatics
Mouse fecal samples were collected sterilely and stored at 80C. Fecal
DNA were extracted with the QIAamp DNA Stool Mini Kit (QIAGEN) and
sequenced using tagged amplicon spanning the V3 to V4 region of bacterial
16S rRNA of approximately 540 bp using the Illumina MiSeq sequencer at
the Micromon facility of Monash University. Raw data were quality filtered
and trimmed of any sequencing adapters using the trimmomatic software
(Bolger et al., 2014) and sequences assembled and error corrected using
the PEAR software (Zhang et al., 2014). Processed data are then analyzed
using QIIME 1.9.1 software (http://qiime.org/index.html) using default param-
eter settings.
Metabolites Measurement
Fecal acetate, propionate, and butyrate concentrations were measured by gas
chromatography mass spectrometry at the Bio21 Institute facility of Melbourne
University.
Tolerance Induction and Food Allergy Model
Tolerance was induced by intragastric administration of 1 mg peanut extract
(PE) daily for 5 consecutive days. The murine model of food allergy has been
previously described (Li et al., 2000); briefly, mice were sensitized by adminis-
tration of 1 mg PE with 10 mg of cholera toxin (List Biologicals) in 200 ml of PBS
intragastrically on days 0 and 7, followed by a booster challenge with 10 mg PE
on day 21. Mice were challenged with 1 mg PE on day 28 and symptoms of
clinical anaphylaxis monitored for 40–60 min: 0, no clinical symptoms; 1, repet-
itive mouth/ear scratching and ear canal digging with hind legs; 2, decreased
activity, self-isolation, and puffiness around eyes and/or mouth; 3, periods of
motionless for more than 1 min, lying prone on stomach, and decreased activ-
ity with increased respiratory rate; 4, no response to whisker stimuli and
reduced or no response to prodding; 5, tremor, convulsion, and death.
Detailed protocol for the extraction of PE can be found in the Supplemental
Experimental Procedures.
In Vivo Gut Permeability Assay
For assessment of in vivo gut permeability, 4 hr fasted mice received
500 mg/kg FITC-Dextran 4000 (Sigma) intragastrically and blood collected
1 hr later. Serum concentration of FITC-dextran was determined by fluor-
ometry (excitation 485 nm; emission 520 nm on a FLUOstar Optima micro-
plate reader; BMG Labtech) relative to standard curve generated from
serial dilution of FITC-dextran in control (non-treated) serum.
In Vivo Proliferation Assay
For assessment of antigen-specific proliferation responses in vivo, mice
were injected intravenously with 10 million CFSE-stained naive OT-II CD4
+
Figure 6. GPR43 and GPR109A Are Required for Dietary-Fiber-Mediated Protection against Food Allergy
(A–C) WT, Gpr43
/
, and Gpr109a
/
were fed on a HF diet for at least 2 weeks and induced food allergy. (A) Anaphylaxis score determined after PE challenge on
day 28 (left) and total serum IgE determined by ELISA (right) are shown. (B) Proportions of MLN MHCII
+
CD11c
+
CD103
+
DCs and CD4
+
CD25
+
FoxP3
+
Treg cells
were determined by flow cytometry. (C) Numeration of total cell numbers in MLN is shown.
(D) Bacteria count in MLN organ (CFU/mg) from WT, Gpr43
/
, and Gpr109a
/
mice under basal conditions after overnight culture.
(E) Small intestinal permeability in WT, Gpr43
/
, and Gpr109a
/
mice fed on HF diet as determined by FITC-dextran translocation into blood 1 hr after oral
gavage of FITC-dextran.
(F) Relative gene expression of Tslp and Il133 to housekeeping gene Rpl13a in the jejunum of WT, Gpr43
/
, and Gpr109a
/
mice under basal conditions
determined by qPCR.
(G) Anaphylaxis score of WT, Villin-Cre.Gpr43
/
, and Vav-Cre.Gpr43
/
mice fed on a HF diet in a model of peanut allergy determined after PE challenge on
day 28.
(H) Anaphylaxis score of bone marrow chimera mice with WT mice reconstituted with Gpr109a
/
bone marrow cells (WT/Gpr109a BM) or Gpr109a
/
mice
reconstituted with WT bone marrow cells (Gpr109a/WT BM) fed on HF diet in a model of peanut allergy after PE challenge on day 28.
(I) Fecal microbiota composition analyzed by 16S rRNA metagenomic sequencing in WT, Gpr43
/
, and Gpr109a
/
mice fed on HF diet. Relative abundance of
bacteria presented at the family level (left) and comparison of microbial community diversity presented as unweighted PCoA plots (right) are shown. Supporting
microbiota analyses are presented in Figure S5I.
All data are representative of at least two independent experiments. All data are represented as mean ±SEM of at least n = 5 mice per group; *p < 0.05; **p < 0.01;
***p < 0.05; ****p < 0.001.
Cell Reports 15, 2809–2824, June 21, 2016 2821
T cells and challenged with 20 mg of OVA (grade V; Sigma) intragastrically
24 hr later. Cell proliferation was analyzed by flow cytometry analysis 3 days
later.
Treg Cell Conversion Assay
5310
4
MHCII
+
CD11c
hi
CD103
+
DCs were sorted from pooled MLN of ZF- or
HF-diet-fed mice and co-cultured with 2 310
5
splenic CD4
+
CD25
CD62L
+
naive OT-II T cells in medium supplemented with 1 mg/ml purified mouse
anti-CD3 (no azide; low endotoxin; eBioscience), 3 ng/ml recombinant-h uman
transforming growth factor b(TGF-b) (Peprotech), and 10 mg/ml OVA
323–329
peptide (Genscript) for 5 days.
Quantitative Real-Time PCR
Total RNA extraction was performed using Trizol (Ambion) following manufac-
turer’s instructions. cDNA was generated using tetro cDNA synthesis kit
(Bioline) following manufacturer’s instructions. qPCR was conducted using
AccuPower GreenStar qPCR Master Mix (Bioneer) on an Applied Biosystems
7500 Real-Time PCR machine. A full list of primer sequences is listed in the
Supplemental Experimental Procedures.
Cell Isolation and Analysis
For isolation of small intestinal lymphocytes cells, Peyer’s patches were
excised from small intestine, washed, and incubated with 20 ml Hank’s
balanced salt solution (HBSS) solution containing 5 mM EDTA and 5% fetal
calf serum (FCS) for 30 min with mild agitation. Epithelial cells were removed
by vortexing briefly afterward. Small intestine pieces were then transferred
to a solution of HBSS containing Collagenase type IV (GIBCO), DNase I
(Roche), and 10% FCS and digested for 90 min before passing through a
70-mm cell strainer. Lymphocytes were enriched via Percoll gradient of 40%
and 80% (GE Healthcare).
Flow Cytometry
Treg cell identification was made with the Foxp3/Transcription Factor Staining
Buffer kit (eBioscience) and RALDH activity assay with the ALDEFLUOR kit
(STEMCELL Technologies) according to manufacturer’s instructions. All cells
were sorted on a BD Influx cell sorter with >98% purity. A full list of antibodies
used can be found in Supplemental Experimental Procedures.
Statistics
A two-tailed Student’s t test was used for analysis of the differences between
mean of groups. p values < 0.05 were considered statistically significant.
ACCESSION NUMBERS
The accession number for the representative 16S rRNA sequences reported in
this paper is SRA: SRP073413.
SUPPLEMENTAL INFORMATION
Supplemental Information includes Supplemental Experimental Procedures
and six figures and can be found with this article online at http://dx.doi.org/
10.1016/j.celrep.2016.05.047.
AUTHOR CONTRIBUTIONS
J.T. performed most of the experiments and wrote the manuscript. C.M. per-
formed experiments. P.J.V. contributed suggestions and edited the manu-
script. G.G., R.E.M., and C.G.V. discussed mechanistic concepts and edited
the manuscript. L.M. initiated and supervised the study and also wrote the
manuscript. C.R.M. supervised the study and wrote the manuscript.
ACKNOWLEDGMENTS
This work was supported by the National Health and Medical Research Coun-
cil (NHMRC) of Australia (1068890). We thank Sidonia Fagarasan for review of
the manuscript, Linda Mason for her help with mouse work, and the staff of
Monash FlowCore facility for assistance in cell sorting experiments. We thank
Professor Malcolm McConville, Dedreia Tull, David Peter De Souza, and Kon-
stantinos Andreas Kouremenos at Bio21 for metabolomics measurements.
Received: February 5, 2016
Revised: March 2, 2016
Accepted: May 10, 2016
Published: June 21, 2016
REFERENCES
Ahern, P.P., Faith, J.J., and Gordon, J.I. (2014). Mining the human gut micro-
biota for effector strains that shape the immune system. Immunity 40,
815–823.
Arpaia, N., Campbell, C., Fan, X., Dikiy, S., van der Veeken, J., deRoos, P., Liu,
H., Cross, J.R., Pfeffer, K., Coffer, P.J., and Rudensky, A.Y. (2013). Metabo lites
produced by commensal bacteria promote peripheral regulatory T-cell gener-
ation. Nature 504, 451–455.
Atarashi, K., Tanoue, T., Shima, T., Imaoka, A., Kuwahara, T., Momose, Y.,
Cheng, G., Yamasaki, S., Saito, T., Ohba, Y., et al. (2011). Induction of colonic
regulatory T cells by indigenous Clostridium species. Science 331, 337–341.
Berin, M.C. (2012). Mucosal antibodies in the regulation of tolerance and
allergy to foods. Semin. Immunopathol. 34, 633–642.
Bjo
¨rkste
´n, B., Sepp, E., Julge, K., Voor, T., and Mikelsaar, M. (2001). Allergy
development and the intestinal microflora during the first year of life.
J. Allergy Clin. Immunol. 108, 516–520.
Bolger, A.M., Lohse, M., and Usadel, B. (2014). Trimmomatic: a flexible
trimmer for Illumina sequence data. Bioinformatics 30, 2114–2120.
Cao, A.T., Yao, S., Gong, B., Elson, C.O., and Cong, Y. (2012). Th17 cells up-
regulate polymeric Ig receptor and intestinal IgA and contribute to intestinal
homeostasis. J. Immunol. 189, 4666–4673.
Cerf-Bensussan, N., and Gaboriau-Routhiau, V. (2010). The immune system
and the gut microbiota: friends or foes? Nat. Rev. Immunol. 10, 735–744.
Chassaing, B., Koren, O., Goodrich, J.K., Poole, A.C., Srinivasan, S., Ley, R.E.,
and Gewirtz, A.T. (2015). Dietary emulsifiers impact the mouse gut microbiota
promoting colitis and metabolic syndrome. Nature 519, 92–96.
Coombes, J.L., and Powrie, F. (2008). Dendritic cells in intestinal immune regu-
lation. Nat. Rev. Immunol. 8, 435–446.
De Filippo, C., Cavalieri, D., Di Paola, M., Ramazzotti, M., Poullet, J.B., Mas-
sart, S., Collini, S., Pieraccini, G., and Lionetti, P. (2010). Impact of diet in
shaping gut microbiota revealed by a comparative study in children from
Europe and rural Africa. Proc. Natl. Acad. Sci. USA 107, 14691–14696.
Enomoto, T., Sowa, M., Nishimori, K., Shimazu, S., Yoshida, A., Yamada, K.,
Furukawa, F., Nakagawa, T., Yanagisawa, N., Iwabuchi, N., et al. (2014).
Effects of bifidobacterial supplementation to pregnant women and infants in
the prevention of allergy development in infants and on fecal microbio ta. Aller-
gol. Int. 63, 575–585.
Faith, J.J., Ahern, P.P., Ridaura, V.K., Cheng, J., and Gordon, J.I. (2014). Iden-
tifying gut microbe-host phenotype relationships using combinatorial commu-
nities in gnotobiotic mice. Sci. Transl. Med. 6, 220ra11.
Feng, T., Cong, Y., Qin, H., Benveniste, E.N., and Elson, C.O. (2010). Genera-
tion of mucosal dendritic cells from bone marrow reveals a critical role of
retinoic acid. J. Immunol. 185, 5915–5925.
Fukuda, S., Toh, H., Hase, K., Oshima, K., Nakanishi, Y., Yoshimura, K., Tobe,
T., Clarke, J.M., Topping, D.L., Suzuki, T., et al. (2011). Bifidobacteria can pro-
tect from enteropathogenic infection through production of acetate. Nature
469, 543–547.
Fulgoni, V.L., 3rd, Keast, D.R., Bailey, R.L., and Dwyer, J. (2011). Foods, forti-
ficants, and supplements: Where do Americans get their nutrients? J. Nutr.
141, 1847–1854.
Furusawa, Y., Obata, Y., Fukuda, S., Endo, T.A., Nakato, G., Takahashi, D.,
Nakanishi, Y., Uetake, C., Kato, K., Kato, T., et al. (2013). Commensal
2822 Cell Reports 15, 2809–2824, June 21, 2016
microbe-derived butyrate induces the differentiation of colonic regulatory
T cells. Nature 504, 446–450.
Geuking, M.B., Cahenzli, J., Lawson, M.A., Ng, D.C., Slack, E., Hapfelmeier,
S., McCoy, K.D., and Macpherson, A.J. (2011). Intestinal bacterial colonization
induces mutualistic regulatory T cell responses. Immunity 34, 794–806.
Gray, C.L., Levin, M.E., Zar, H.J., Potter, P.C., Khumalo, N.P., Volkwyn, L., Fe-
nemore, B., and du Toit, G. (2014). Food allergy in South African children with
atopic dermatitis. Pediatr. Allergy Immunol. 25, 572–579.
Hall, J.A., Grainger, J.R., Spencer, S.P., and Belkaid, Y. (2011). The role of ret-
inoic acid in tolerance and immunity. Immunity 35, 13–22.
Hill, D.A., Siracusa, M.C., Abt, M.C., Kim, B.S., Kobuley, D., Kubo, M., Kam-
bayashi, T., Larosa, D.F., Renner, E.D., Orange, J.S., et al. (2012). Commensal
bacteria-derived signals regulate basophil hematopoiesis and allergic inflam-
mation. Nat. Med. 18, 538–546.
Hirota, K., Turner, J.E., Villa, M., Duarte, J.H., Demengeot, J., Steinmetz, O.M.,
and Stockinger, B. (2013). Plasticity of Th17 cells in Peyer’s patches is respon-
sible for the induction of T cell-dependent IgA responses. Nat. Immunol. 14,
372–379.
Huffnagle, G.B. (2010). The microbiota and allergies/asthma. PLoS Pathog. 6,
e1000549.
Iijima, K., Kobayashi, T., Hara, K., Kephart, G.M., Ziegler, S.F., McKenzie, A.N.,
and Kita, H. (2014). IL-33 and thymic stromal lymphopoietin mediate immune
pathology in response to chronic airborne allergen exposure. J. Immun ol. 193,
1549–1559.
Iwata, M., Hirakiyama, A., Eshima, Y., Kagechika, H., Kato, C., and Song, S.Y.
(2004). Retinoic acid imprints gut-homing specificity on T cells. Immunity 21,
527–538.
Jaensson, E., Uronen-Hansson, H., Pabst, O., Eksteen, B., Tian, J., Coombes,
J.L., Berg, P.L., Davidsson, T., Powrie, F., Johansson-Lindbom, B., and
Agace, W.W. (2008). Small intestinal CD103+ dendritic cells display unique
functional properties that are conserved between mice and humans. J. Exp.
Med. 205, 2139–2149.
Janzi, M., Kull, I., Sjo
¨berg, R., Wan, J., Mele
´n, E., Bayat, N., Ostblom, E., Pan-
Hammarstro
¨m, Q., Nilsson, P., and Hammarstro
¨m, L. (2009). Selective IgA
deficiency in early life: association to infections and allergic diseases during
childhood. Clin. Immunol. 133, 78–85.
Kawamoto, S., Tran, T.H., Maruya, M., Suzuki, K., Doi, Y., Tsutsui, Y., Kato,
L.M., and Fagarasan, S. (2012). The inhibitory receptor PD-1 regulates IgA se-
lection and bacterial composition in the gut. Science 336, 485–489.
Kawamoto, S., Maruya, M., Kato, L.M., Suda, W., Atarashi, K., Doi, Y., Tsutsui,
Y., Qin, H., Honda, K., Okada, T., et al. (2014). Foxp3(+) T cells regulate immu-
noglobulin a selection and facilitate diversification of bacterial species respon-
sible for immune homeostasis. Immunity 41, 152–165.
Kim, J.H., Jeun, E.J., Hong, C.P., Kim, S.H., Jang, M.S., Lee, E.J., Moon, S.J.,
Yun, C.H., Im, S.H., Jeong, S.G., et al. (2016). Extracellular vesicle-derived
protein from Bifidobacterium longum alleviates food allergy through mast
cell suppression. J. Allergy Clin. Immunol. 137, 507–516.e8.
Koplin, J.J., Martin, P.E., and Allen, K.J. (2011). An update on epidemiology of
anaphylaxis in children and adults. Curr. Opin. Allergy Clin. Immunol. 11,
492–496.
Kukkonen, K., Kuitunen, M., Haahtela, T., Korpela, R., Poussa, T., and Savi-
lahti, E. (2010). High intestinal IgA associates with reduced risk of IgE-associ-
ated allergic diseases. Pediatr. Allergy Immunol. 21, 67–73.
Laffont, S., Siddiqui, K.R., and Powrie, F. (2010). Intestinal inflammation abro-
gates the tolerogenic properties of MLN CD103+ dendritic cells. Eur. J. Immu-
nol. 40, 1877–1883.
Le Chatelier, E., Nielsen, T., Qin, J., Prifti, E., Hildebrand, F., Falony, G., Al-
meida, M., Arumugam, M., Batto, J.M., Kennedy, S., et al.; MetaHIT con-
sortium (2013). Richness of human gut microbiome correlates with metabolic
markers. Nature 500, 541–546.
Li, X.M., Serebrisky, D., Lee, S.Y., Huang, C.K., Bardina, L., Schofield, B.H.,
Stanley, J.S., Burks, A.W., Bannon, G.A., and Sampson, H.A. (2000). A murine
model of peanut anaphylaxis: T- and B-cell responses to a major peanut
allergen mimic human responses. J. Allergy Clin. Immunol. 106, 150–158.
Macia, L., Thorburn, A.N., Binge, L.C., Marino, E., Rogers, K.E., Maslowski,
K.M., Vieira, A.T., Kranich, J., and Mackay, C.R. (2012). Microbial influences
on epithelial integrity and immune function as a basis for inflammatory dis-
eases. Immunol. Rev. 245, 164–176.
Macia, L., Tan, J., Vieira, A.T., Leach, K., Stanley, D., Luong, S., Maruya, M.,
Ian McKenzie, C., Hijikata, A., Wong, C., et al. (2015). Metabolite-sens ing re-
ceptors GPR43 and GPR109A facilitate dietary fibre-induced gut homeostasis
through regulation of the inflammasome. Nat. Commun. 6, 6734.
Maslowski, K.M., and Mackay, C.R. (2011). Diet, gut microbiota and immune
responses. Nat. Immunol. 12, 5–9.
Maslowski, K.M., Vieira, A.T., Ng, A., Kranich, J., Sierro, F., Yu, D., Schilter,
H.C., Rolph, M.S., Mackay, F., Artis, D., et al. (2009). Regulation of inflamma-
tory responses by gut microbiota and chemoattractant receptor GPR43.
Nature 461, 1282–1286.
Matteoli, G., Mazzini, E., Iliev, I.D., Mileti, E., Fallarino, F., Puccetti, P.,
Chieppa, M., and Rescigno, M. (2010). Gut CD103+ dendritic cells express
indoleamine 2,3-dioxygenase which influences T regulatory/T effector cell
balance and oral tolerance induction. Gut 59, 595–604.
Meini, A., Pillan, N.M., Villanacci, V., Monafo, V., Ugazio, A.G., and Plebani, A.
(1996). Prevalence and diagnosis of celiac disease in IgA-deficient children.
Ann. Allergy Asthma Immunol. 77, 333–336.
Molenaar, R., Knippenberg, M., Goverse, G., Olivier, B.J., de Vos, A.F.,
O’Toole, T., and Mebius, R.E. (2011). Expression of retinaldehyde dehydroge-
nase enzymes in mucosal dendritic cells and gut-draining lymph node stromal
cells is controlled by dietary vitamin A. J. Immunol. 186, 1934–1942.
Moon, C., Baldridge, M.T., Wallace, M.A., Burnham, C.A., Virgin, H.W., and
Stappenbeck, T.S. (2015). Vertically transmitted faecal IgA levels determine
extra-chromosomal phenotypic variation. Nature 521, 90–93.
Noval Rivas, M., Burton, O.T., Wise, P., Zhang, Y.Q., Hobson, S.A., Garcia
Lloret, M., Chehoud, C., Kuczynski, J., DeSantis, T., Warrington, J., et al.
(2013). A microbiota signature associated with experimental food allergy pro-
motes allergic sensitization and anaphylaxis. J. Allergy Clin. Immunol. 131,
201–212.
Noverr, M.C., and Huffnagle, G.B. (2005). The ‘microflora hypothesis’ of
allergic diseases. Clin. Exp. Allergy 35, 1511–1520.
Ou, J., Carbonero, F., Zoetendal, E.G., DeLany, J.P., Wang, M., Newton, K.,
Gaskins, H.R., and O’Keefe, S.J. (2013). Diet, microbiota, and microbial
metabolites in colon cancer risk in rural Africans and African Americans. Am.
J. Clin. Nutr. 98, 111–120.
Pabst, O. (2012). New concepts in the generation and functions of IgA. Nat.
Rev. Immunol. 12, 821–832.
Pabst, O., and Mowat, A.M. (2012). Oral tolerance to food protein. Mucosal
Immunol. 5, 232–239.
Palm, N.W., de Zoete, M.R., Cullen, T.W., Barry, N.A., Stefanowski, J., Hao, L.,
Degnan, P.H., Hu, J., Peter, I., Zhang, W., et al. (2014). Immunoglobulin A
coating identifies colitogenic bacteria in inflammatory bowel disease. Cell
158, 1000–1010.
Palomares, O., Yaman, G., Azkur, A.K., Akkoc, T., Akdis, M., and Akdis, C.A.
(2010). Role of Treg in immune regulation of allergic diseases. Eur. J. Immunol.
40, 1232–1240.
Paul, W.E., and Zhu, J. (2010). How are T(H)2-type immune responses initiated
and amplified? Nat. Rev. Immunol. 10, 225–235.
Pawar, S., Feehley, T., and Nagler, C. (2015). Commensal bacteria-induced
MyD88 signaling regulates intestinal permeability to food allergen via anti-mi-
crobial peptide and mucin production (MUC9P.741). J. Immunol. 194, 205.5.
Scott, C.L., Aumeunier, A.M., and Mowat, A.M. (2011). Intestinal CD103+ den-
dritic cells: master regulators of tolerance? Trends Immunol. 32, 412–419.
Shan, M., Gentile, M., Yeiser, J.R., Walland, A.C., Bornstein, V.U., Chen, K.,
He, B., Cassis, L., Bigas, A., Cols, M., et al. (2013). Mucus enhances gut ho-
meostasis and oral tolerance by delivering immunoregulatory signals. Science
342, 447–453.
Cell Reports 15, 2809–2824, June 21, 2016 2823
Singh, N., Gurav, A., Sivaprakasam, S., Brady, E., Padia, R., Shi, H., Thangar-
aju, M., Prasad, P.D., Manicassamy, S., Munn, D.H., et al. (2014). Activation of
Gpr109a, receptor for niacin and the commensal metabolite butyrate, sup-
presses colonic inflammation and carcinogenesis. Immunity 40, 128–139.
Smith, P.M., Howitt, M.R., Panikov, N., Michaud, M., Gallini, C.A., Bohlooly-Y,
M., Glickman, J.N., and Garrett, W.S. (2013). The microbial metabolites, short-
chain fatty acids, regulate colonic Treg cell homeostasis. Science 341,
569–573.
Stefka, A.T., Feehley, T., Tripathi, P., Qiu, J., McCoy, K., Mazmanian, S.K.,
Tjota, M.Y., Seo, G.Y., Cao, S., Theriault, B.R., et al. (2014). Commensal bac-
teria protect against food allergen sensitization. Proc. Natl. Acad. Sci. USA
111, 13145–13150.
Strachan, D.P. (1989). Hay fever, hygiene, and household size. BMJ 299,
1259–1260.
Tang, M.L., Ponsonby, A.L., Orsini, F., Tey, D., Robinson, M., Su, E.L., Lic-
ciardi, P., Burks, W., and Donath, S. (2015). Administration of a probiotic
with peanut oral immunotherapy: a randomized trial. J. Allergy Clin. Immunol.
135, 737–744.e8.
Thorburn, A.N., Macia, L., and Mackay, C.R. (2014). Diet, metabolites, and
‘‘western-lifestyle’’ inflammatory diseases. Immunity 40, 833–842.
Tuano, K.S., Orange, J.S., Sullivan, K., Cunningham-Rundles, C., Bonilla, F.A.,
and Davis, C.M. (2015). Food allergy in patients with primary immunodefi-
ciency diseases: prevalence within the US Immunodeficiency Network
(USIDNET). J. Allergy Clin. Immunol. 135, 273–275.
Turfkruyer, M., Rekima, A., Macchiaverni, P., Le Bourhis, L., Muncan, V., van
den Brink, G.R., Tulic, M.K., and Verhasselt, V. (2016). Oral tolerance is ineffi-
cient in neonatal mice due to a physiological vitamin A deficiency. Mucosal
Immunol. 9, 479–491.
Turnbaugh, P.J., Ba
¨ckhed, F., Fulton, L., and Gordon, J.I. (2008). Diet-induced
obesity is linked to marked but reversible alterations in the mouse distal gut
microbiome. Cell Host Microbe 3, 213–223.
van der Marel, A.P., Samsom, J.N., Greuter, M., van Berkel, L.A., O’Toole, T.,
Kraal, G., and Mebius, R.E. (2007). Blockade of IDO inhibits nasal tolerance
induction. J. Immunol. 179, 894–900.
Wang, J., and Sampson, H.A. (2011). Food allergy. J. Clin. Invest. 121,
827–835.
Wilson, K., Ducharme, R., and Hawken, S. (2013). Association between socio-
economic status and adverse events following immunization at 2, 4, 6 and
12 months. Hum. Vaccin. Immunother. 9, 1153–1157.
Yel, L. (2010). Selective IgA deficiency. J. Clin. Immunol. 30, 10–16.
Zhang, J., Kobert, K., Flouri, T., and Stamatakis, A. (2014). PEAR: a fast and
accurate Illumina Paired-End reAd mergeR. Bioinformatics 30, 614–620.
2824 Cell Reports 15, 2809–2824, June 21, 2016
