Supporting Information
Stefka et al. 10.1073/pnas.1412008111
SI Methods
Intragastric Sensitization with Peanut/Cholera Toxin.
In all experi-
ments, specific pathogen-free mice were sensitized by intragastric
gavage with 6 mg of peanut (PN) plus 10
μ
g cholera toxin (CT,
List Biologicals) on days 0 and 2 and with 6 mg of PN plus 15
μ
g
CT on days 7, 14, 21, and 28. The dose of CT was reduced to
10
μ
g for each sensitization of gnotobiotic mice. Control groups
received CT in Tris buffer as a vehicle. When gnotobiotic or
Abx-treated mice were colonized, sensitization began 1 d after
fecal gavage. On day 35, all mice were challenged twice by ga-
vage with 20 mg of PN 30 min apart. Rectal temperature was
recorded before and 30 min after each challenge; the reported
Δ
T is the temperature 30 min after the second challenge minus
the temperature prechallenge. Serum, feces, and tissues were
harvested from each mouse on day 36.
Bacterial Colonization.
For
Bacteroides
isolation, full-length 16S
rDNA was PCR amplified and sequenced from individual colo-
nies from SPF feces after growth on Anaerobic Neomycin Blood
Agar plates (Remel). A strain with
>
99% sequence identity to
Bacteroides uniformis
(NR_040866) was further propagated in
Schaedler
’
s media. For Clostridia isolation, a chloroform-treated
fecal/cecal suspension was prepared with slight modifications
from ref. 1. A single fecal pellet plus an equivalent volume of
cecal material from the same donor was suspended in 18 mL of
prereduced PBS in an anaerobic chamber. Chloroform was added
to 3% (vol/vol) concentration; the sample was shaken vigorously
and then incubated at 37 °C for 1h. Chloroform was removed by
percolation with CO
2
from a compressed cylinder. This suspension
was then administered to an adult germ-free (GF) mouse as a re-
pository for all subsequent colonizations, which were performed as
described in the main text. Conventionalization of GF or antibiotic
(Abx)-treated mice was performed by suspending a fecal pellet and
an equivalent volume of cecal material from a young donor SPF
mouse in 1.5 mL sterile PBS and administering 100
μ
Lofthe
liquid phase to weanling C57BL/6 or C57BL/6 Foxp3
gfp
recipients
by gavage. A second gavage was given 2 wk after weaning in Abx
conventionalized and Abx Clostridia sensitization experiments. For
Abx recovery experiments, mice received 1 wk of Abx by gavage
before weaning and then were left unmanipulated; no additional
bacteria were administered. Whenever possible, littermates were
used to minimize variation in the composition of the microbiota
pretreatment. If littermates were not used, all pups were related by
maternal lineage.
Ig Detection.
For antigen-specific assays, plates were coated with
PN, and antigen-specific serum antibodies were detected with
goat anti-mouse IgG1-HRP (Southern Biotech), rat anti-mouse
IgE-AP (23G3, Southern Biotech), or goat anti-mouse IgG-HRP
(Southern Biotech). PN-specific standards were prepared from
the serum of sensitized mice by affinity purification on a
PN-conjugated CNBr-Sepharose column (GE Healthcare). For
PN-IgE standard, IgG was depleted with Protein G Sepharose
(GE Healthcare) before affinity purification. Total IgE plates were
coated with purified rat anti-mouse IgE (R35-72, BD Pharmin-
gen), and antibodies were detected with biotin rat anti-mouse IgE
(R35-118, BD Pharmingen) and streptavidin-HRP using a mouse
IgE standard (BD Biosciences). For fecal IgA, fecal pellets were
suspended in PBS containing 0.01% NaN
3
(1 mL/0.1 g feces) and
centrifuged to remove debris. Fecal supernatants were applied to
plates coated with goat anti-mouse polyclonal IgA (Southern
Biotech), and antibodies were detected with goat anti-mouse
IgA-AP (Southern Biotech). OD was normalized to a commer-
cial mouse IgA standard (BD Biosciences).
Preparation of 16S rRNA-Based Amplicon Library and Data Analysis.
Fecal samples were mixed with 0.1 mm zirconia/silica beads in 1.4
mL ASL buffer (Qiagen) in a Mini-Beadbeater (Biospec); DNA
was extracted with the QIAamp DNA Stool Mini Kit (Qiagen)
using 1/2 inhibitEX tablet and 100
μ
L Buffer AE for elution.
Fecal DNA template was amplified using the 515F/806R region
of the 16S rRNA gene with primers and cycling conditions
modified slightly from ref. 2, specifically adapted for the Illumina
MiSeq by adding nine extra bases in the adapter region of the
forward amplification primer to support paired-end sequencing.
Briefly, the V4 region of the 16S rRNA gene was amplified with
region-specific primers that
included the Illumina flowcell
adapter sequences and 12-base barcodes on the reverse primer.
PCR reactions were completed in triplicate, and products were
pooled. Each pool was then quantified using Invitrogen
’
s Pico-
Green, pooled with equal amounts of DNA per sample, and
cleaned using the UltraClean PCR Clean-Up Kit (MoBIO).
Amplicons were then sequenced on an Illumina MiSeq at the Next-
Generation Sequencing Core at Argonne National Laboratory
using custom sequencing primers and procedures described in the
supplementary methods of ref. 3. Paired-end Illumina reads were
joined using fastq-join (E. Aronest
y, code.google.com/p/ea-utils).
All analysis of high-throughput sequencing data was performed
using QIIME v.1.6.0 as previously de
scribed (4), except that uclust
was used for operational taxonomic unit (OTU) selection (with
97% identity threshold). Where the UniFrac metric was used,
principal coordinates analysis of t
he unweighted pairwise distance
matrix is shown; even sampling was performed at a depth of 7,500
(Fig. S2
F
)or13,500(Fig.5
I
) sequences per sample. Phylogenetic
reconstruction of Clostridia OTUs (Fig. S1
C
) includes sequences
described in ref. 5.
Calculation of Bacterial Load.
Bacterial load was determined by
quantitative real-time PCR using a protocol modified from ref. 6.
Fecal DNA was extracted as for sequencing, and bacterial load was
quantified against a standard curve derived from a pCR4TOPO-TA
vector containing a nearly full-length copy of the 16S rRNA gene
from a member of
Porphyromonadaceae
. Bacterial DNA was am-
plified with universal primers 8F and 338R using the iQ SYBR
greensupermix(Bio-Rad Life Science) andthe StepOnePlus system
(Applied Biosystems). The results were normalized to fecal weight.
Isolation of Lymphocytes and Flow Cytometry.
Singlecellsuspensions
from spleen, mesenteric lymph node, and lamina propria (LP)
were stained with anti-CD4-PE (RM 4.4, eBioscience). In-
tracellular staining with anti-Foxp3-FITC (FJK-16s, eBioscience)
was performed after permeabilization with Foxp3 Fix/Perm buffer
(eBioscience). In experiments with gnotobiotic mice, Tregs were
analyzed 2 wk after colonization. For innate lymphoid cell (ILC)
cytokine analysis, cells were isolated and stained as described in ref.
7. Briefly, mononuclear cells from the colon were isolated using
mechanical/enzymatic digestion and Percoll density centrifugation.
The isolated cells were incubated for 4 h with 50 ng/mL phorbol
myristate acetate and 500 ng/mL ionomycin before fixation and
permeabilization using Foxp3 Fix/Perm buffer (eBioscience) and
staining for CD3, CD4, TCR
β
, NKp46, ROR
γ
t, and IL-22. All
samples were acquired on FACSCanto or LSRII, and data were
analyzed with FlowJo software (TreeStar).
Stefka et al.
www.pnas.org/cgi/content/short/1412008111
1of7
Microarray Analysis.
Data were normalized using Illumina software
and then analyzed with dChip. Genes were considered induced if
the detection
P
value for each sample was
>
0.05 and there was
≥
1.5-fold increase in expression in either colonized group com-
pared with GF. For verification of results, cDNA was produced
as described in
Methods
followed by quantitative real-time PCR
using Taqman primer/probe sets and Master Mix (Applied Bio-
systems). Results were normalized to
Gapdh
.
Mucus Staining and Goblet Cell Quantification.
Tissue from the
center of the distal colon was collected from GF,
B. uniformis
, and
Clostridia-colonized mice 6 d after colonization or from sensi-
tized Abx-treated, Clostridia-colonized mice with or without
anti-IL-22 treatment 24 h after challenge. Tissue was fixed for
4 h at 4 °C in Carnoy
’
s fixative and then transferred to 70% EtOH
for paraffin embedding. Cross-sections of 5-
μ
m thickness were
stained with periodic acid Schiff as previously described (8).
Slides were imaged on a CRi Panoramic Scan Whole Slide
Scanner and analyzed with Pano
ramic Viewer. Goblet cells
were quantified in at least 7 crypts per mouse, average 11 crypts
per mouse.
Cell Culture and Cytokine Measurement.
Single-cell suspensions
were prepared from spleens harvested 24 h after challenge from
Abx-treated, Clostridia-colonized mice treated with anti-IL-22 or
an isotype control and sensitized with PN and CT. Cells were
plated at 2
×
10
5
cells per well with media alone, 1
μ
g/mL anti-
CD3 (clone 2C11), or 200
μ
g/mL PN and incubated at 37 °C for
72 h as previously described (9). After 72 h, plates were frozen at
−
20 °C. Cytokine concentrations in supernatants were measured
using Milliplex MAP Mouse Cytokine magnetic bead panel
(Millipore) and read on a Bio-Plex machine (BioRad) as pre-
viously described (10).
1. Itoh K, Mitsuoka T (1985) Characterization of clostridia isolated from faeces of limited
flora mice and their effect on caecal size when associated with germ-free mice.
Lab
Anim
19(2):111
–
118.
2. Caporaso JG, et al. (2011) Global patterns of 16S rRNA diversity at a depth of millions
of sequences per sample.
Proc Natl Acad Sci USA
108(Suppl 1):4516
–
4522.
3. Caporaso JG, et al. (2012) Ultra-high-throughput microbial community analysis on the
Illumina HiSeq and MiSeq platforms.
ISME J
6(8):1621
–
1624.
4. Devkota S, et al. (2012) Dietary-fat-induced taurocholic acid promotes pathobiont
expansion and colitis in Il10
−
/
−
mice.
Nature
487(7405):104
–
108.
5. Krogius-Kurikka L, et al. (2009) Sequence analysis of percent G
+
Cfractionlibraries
of human faecal bacterial DNA reveal
s a high number of Actinobacteria.
BMC
Microbiol
9:68.
6. Buffie CG, et al. (2012) Profound alterations of intestinal microbiota following
a single dose of clindamycin results in sustained susceptibility to
Clostridium difficile
-
induced colitis.
Infect Immun
80(1):62
–
73.
7. Qiu J, et al. (2012) The aryl hydrocarbon receptor regulates gut immunity through
modulation of innate lymphoid cells.
Immunity
36(1):92
–
104.
8. Sugimoto K, et al. (2008) IL-22 ameliorates intestinal inflammation in a mouse model
of ulcerative colitis.
J Clin Invest
118(2):534
–
544.
9. Bashir ME, Louie S, Shi HN, Nagler-Anderson C (2004) Toll-like receptor 4 signaling by
intestinal microbes influences susceptibility to food allergy.
J Immunol
172(11):
6978
–
6987.
10. Tjota MY, et al. (2013) IL-33-dependent induction of allergic lung inflammation by
Fc
γ
RIII signaling.
J Clin Invest
123(5):2287
–
2297.
Stefka et al.
www.pnas.org/cgi/content/short/1412008111
2of7
0
2,000
4,000
6,000
8,000
7 weeks
27 weeks
SPF
Germ free
**
Total IgE (ng/mL)
0.0
0.2
0.4
0.6
0.8
1.0
Abundance
Other/Unclassified Phyla
All Members
All Members
Lachnospiraceae
Ruminococcaceae
Others
Others
Firmicutes
C
l
o
s
t
r
i
d
i
a
Bacteroidetes
Proteobacteria
#42 0.25%
#17 0.96%
#52 0.14%
Clostridium nexile
Oscillospira guilliermondii
#39 0.27%
#14 1.15%
#5 3.17%
#21 0.8%
#6 2.16%
#66 0.1%
#2 6.7%
#41 0.26%
#34 0.35%
#26 0.67%
#33 0.37%
#46 0.18%
Peptostreptococcus micros
#49 0.16%
#54 0.12%
#7 2.1%
#61 0.1%
#62 0.1%
Clostridium litorale
Tissierella praeacuta
Bacteroides uniformis
#45 0.19%
Clostridium propionicum
#16 1%
#68 0.1%
#35 0.34%
#12 1.53%
#4 4.08%
#53 0.13%
#51 0.16%
#64 0.1%
#56 0.11%
#57 0.11%
#32 0.39%
#43 0.25%
#23 0.69%
#69 0.1%
Clostridium colinum
#55 0.12%
#11 1.68%
#10 1.71%
#20 0.82%
#28 0.5%
#47 0.17%
#24 0.69%
#65 0.1%
#44 0.21%
#50 0.16%
#18 0.87%
#25 0.67%
Clostridium orbiscindens
Ruminococcus obeum
#9 2.04%
#70 0.1%
Clostridium xylanolyticum
#58 0.11%
Clostridium aminophilum
#13 1.17%
#1 30.51%
#31 0.42%
#19 0.83%
#30 0.44%
#27 0.54%
Eubacterium rectale
#38 0.29%
Roseburia cecicola
#67 0.1%
#22 0.72%
#36 0.34%
#36 0.34%18
#29 0.49%
Eubacterium limosum
Butyrivibrio fibrisolvens
#59 0.11%
#63 0.1%
#48 0.17%
Clostridium coccoides
#8 2.08%
#15 1.12%
#40 0.26%
#3 5.95%
#37 0.33%
Faecalibacterium prausnitzii
Cluster IV
Cluster XIVb
Cluster XIVa
D
Initial
Colonization
4 Months
10 Months
15 Months
OTU #1
OTU #70
A
B
C
0
>0%
>0.1%
>0.25%
>1%
>2.5%
Fig. S1.
(
A
) IgE concentrations determined by ELISA.
n
=
7
–
8 mice per group. (
B
) Average taxonomic classifications from paired-end Illumina sequencing of
fecal 16S rDNA amplicons from 41 Clostridia-colonized mice over a 15-mo period. (
C
) Phylogenetic reconstruction of OTUs present in feces from the same mice
as in
B
. Among reads classified as Clostridia, all OTUs (defined as 97% identical) with
>
0.1% abundance were considered for alignment with representative
sequences (Red text; pynast). OTUs are numbered 1
–
70, and their abundance is shown as a percentage. (
D
) Abundance of the 70 OTUs depicted in
C
in in-
dividual mice over time. **
P
<
0.01 by one-way ANOVA with Tukey posttest.
Stefka et al.
www.pnas.org/cgi/content/short/1412008111
3of7
0
100
200
300
400
**
*
NT
Abx
Abx.
conv.
Abx
Clost.
Abx
Recov.
**
0
200
400
2,000
4,000
NT
Abx
Abx.
conv.
Abx
Clost.
Abx
Recov.
0
2,000
4,000
6,000
8,000
**
*
NT
Abx
Abx.
conv.
Abx
Clost.
Abx
Recov.
Total IgE (ng/mL)
C
D
E
B
9
10
11
12
13
14
NT
Abx
Abx
Clost.
Abx
conv.
***
***
***
***
Abx
Recov.
Log Copies
16S rRNA gene / gfeces
F
A
0.0
0.2
0.4
0.6
0.8
1.0
Abx conv.
Abx Clost.
S24-7
Prevotellaceae
Rickinellaceae
Bacteroidaceae
Others
Ruminococcaceae
Lachnospiraceae
Others
Lactobacillaceae
Others
Others
All Members
Others
Desulfovibrionales
All Members
All Members
Bacteroidetes
Firmicutes
Proteobacteria
Bacilli
Clostridia
Other Phyla/Unclassified Bacteria
Actinobacteria
Erysipelotrichi
Abx Recov.
Abx (1 week)
Abundance
PC1 (23% of var. expl.)
+0.45
-0.17
PC2 (13% of var. expl.)
+0.18
-0.36
PC3 (12% of var. expl.)
-0.24
+0.29
Abx conv.
Abx Clost.
Abx Recov.
Abx (1 week)
Fig. S2.
(
A
–
D
) Groups of mice were not treated (NT, white) or treated with Abx by gavage before weaning and then in the drinking water afterward for 5 wk
(as in Fig. 1; Abx, dark gray). Alternatively, groups of mice received Abx by gavage preweaning and then were colonized at weaning with conventional SP
F
feces and cecal contents (Abx. conv., blue), or with a consortium of Clostridia (Abx Clost., green), or were allowed to recover without colonization (
Abx Recov.,
black). Mice in each treatment group were sensitized with PN/CT and challenged on day 35. (
A
) Concentration of PN-specific IgE, (
B
) IgG1, and (
C
) total IgE in
the serum of sensitized mice, collected 24 h after challenge, as determined by ELISA.
n
=
4
–
10 mice per group. (
D
) Bacterial load at killing.
n
=
3
–
5 mice per
group. (
E
) Taxonomic classifications and (
F
) UniFrac analysis for Illumina-derived sequences of fecal 16S rDNA from mice in
A
–
D
.
n
=
4
–
5 mice per group. Each
group was housed in two cages, separated by sex. Data shown are pooled from two independent experiments. Bar graphs depict mean and SEM; in
A
–
D
and
F
,
each circle represents an individual mouse. *
P
<
0.05, **
P
<
0.01, ***
P
<
0.001 by one-way ANOVA with Tukey posttest (
A
–
D
).
Stefka et al.
www.pnas.org/cgi/content/short/1412008111
4of7
0
1
2
3
4
5
GF
Clost.
%LTi4 cells
0
2
4
6
8
10
GF
Clost.
%LTi0 cells
0
1
2
GF
Clost.
%NK22 cells
0
2
4
6
8
10
12
GF
Clost.
%CD4
+
TCR
+
cells
Gated on CD3
-
ROR
cells
Germ free
Clostridia
Gated on CD4
+
TCR
+
cells
Germ free
Clos tridia
SSC
IL-22
SSC
IL-22
A
B
Fig. S3.
(
A
) Representative flow cytometry plots for data in Fig. 4
B
.(
B
) Percentage of Lti0 (CD3
−
CD4
−
ROR
γ
t
+
NKp46
−
), Lti4 (CD3
−
CD4
+
ROR
γ
t
+
NKp46
−
), and
NK-22 (CD3
−
ROR
γ
t
+
NKp46
+
) ILC subsets among CD3
−
TCR
β
−
cells and the percentage of CD4
+
TCR
β
+
cells in the colonic LP of GF (red) and Clostridia-colonized
(green) mice.
n
=
3 mice per group. Data are representative of three independent experiments.
Stefka et al.
www.pnas.org/cgi/content/short/1412008111
5of7
Pre
1 hr
3 hr
0
20
40
60
80
100
150
200
250
Abx PN+CT
AbxClostridia PN+CT
*
Ara h 6 in serum (ng/mL)
0
10
20
30
40
Abx
Abx+IL-22Fc
AbxClostridia
AbxClostridia+
IL-22
Spleen
MLN
Colonic
LP
***
**
% Foxp3
+
of CD4
+
0.0
0.5
1.0
1.5
Rag
-/-
AbxClostridia+isotype
Rag
-/-
AbxClostridia+
CD90.2
*
Small Intestine
Colon
Relative expression Il22
Pre
15 min
45 min
0
20
40
60
80
Rag
-/-
AbxClostridia+isotype PN
Rag
-/-
AbxClostridia+
CD90.2 PN
Ara h 6 in serum (ng/mL)
ABC
DE
FGH
Pre
1 hr
3 hr
0
20
40
60
80
100
200
300
400
500
Abx PN+CT
AbxClostridia PN+CT
Ara h 2 in serum (ng/mL)
Pre
15 min
45 min
0
50
100
150
200
Rag
-/-
AbxClostridia+isotype PN
Rag
-/-
AbxClostridia+
CD90.2 PN
Ara h 2 in serum (ng/mL)
I
AbxClostridia
+isotype
AbxClostridia
+
IL-22
0
5
10
15
20
*
AbxClostridia
+isotype
AbxClostridia
+
IL-22
Number of Goblet Cells / Crypt
J
0
7
14
21
28
36
0.0
0.2
0.4
0.6
0.8
1.0
Days after weaning
Bacteroidales abundance
Pre
1 hr
3 hr
0
10
20
30
40
100
200
300
400
Abx PN
AbxClostridia PN
Abx+IL-22Fc PN
NT PN
Ara h 2 in serum (ng/mL)
Pre
1 hr
3 hr
0
50
100
150
AbxClostridia+
IL-22 PN
AbxClostridia+isotype PN
Ara h 2 in serum (ng/mL)
36
0
7
14
21
28
0.0
0.1
0.2
0.3
AbxClostridia+
IL-22
AbxClostridia+isotype
Days after weaning
Clostridiales abundance
0
50
100
2000
4000
6000
CD3
IFN-
(pg/mL)
K
L
0
50
100
150
200
250
AbxClostridia+isotype
AbxClostridia+
IL-22
CD3
IL-13 (pg/mL)
Fig. S4.
(
A
) Serum Ara h 2 levels after PN gavage in NT or Abx-treated mice and Abx-treated mice colonized with Clostridia or given one i.p. injection of IL-22-
Fc (mice from Fig. 5
B
;
n
=
6
–
9 mice per group) and (
B
) Abx-treated Clostridia colonized mice treated with neutralizing antibody to IL-22 or an isotype control
(mice from Fig. 5
C
;
n
=
5
–
8 mice per group). (
C
) Flow cytometric analysis of Foxp3
+
Tregs among CD4
+
T cells in Abx mice with or without subsequent IL-22-Fc
treatment, Clostridia colonization, or Clostridia colonization plus neutralizing antibody to IL-22 (
n
=
3
–
7 mice per group). (
D
) Serum levels of Ara h 6 and (
E
)
Ara h 2 in Abx-treated mice or Abx-treated mice colonized with Clostridia and challenged with PN plus CT.
n
=
4 mice per group. (
F
) Serum Ara h 6 and (
G
)Ara
h 2 after PN gavage in
Rag
−
/
−
Abx-treated mice colonized with Clostridia and injected i.p. with anti-CD90.2 (ILC depleted) or isotype control.
n
=
8 mice per
Legend continued on following page
Stefka et al.
www.pnas.org/cgi/content/short/1412008111
6of7
group. (
H
) Quantitative real-time PCR analysis of IL-22 transcripts from lamina propria lymphocytes isolated after colonization from Abx-treated Clostrid
ia-
colonized
Rag
−
/
−
mice injected with anti-CD90.2 (ILC depleted) or isotype control (from
F
and
G
). Data are plotted relative to AbxClostridia
+
isotype and
normalized to
Hprt
.
n
=
8 mice per group. (
I
) Quantification and representative images of goblet cells in distal colon of sensitized Abx-treated Clostridia-
colonized mice treated with anti-IL-22 or an isotype control at 24 h after challenge (mice from Fig. 5
D
–
I
).
n
=
5 mice per group. (Scale bar, 100
μ
m.) (
J
)
Concentration of IL-13 and (
K
) IFN-
γ
in culture supernatants from splenocytes of mice in Fig. 5
D
–
I
.(
L
) Abundance of Clostridiales and Bacteroidales for mice
from Fig. 5
D
–
I
.*
P
<
0.05, **
P
<
0.01, ***
P
<
0.001 by two-way ANOVA with Bonferroni posttest (
C
,
D
,and
H
) or Student
t
test (
I
).
Stefka et al.
www.pnas.org/cgi/content/short/1412008111
7of7