Cell Host & Microbe, Volume
12
Supplemental Information
Outer Membrane Vesicles of a Human Commensal
Mediate Immune Regulation and Disease Protection
Yue Shen, Maria Letizia Giardin
o Torchia, Gregory W. Lawson,
Christopher L. Karp, Jonathan D. Ashwell, and Sarkis K. Mazmanian
SUPPLEMENTAL INVENTORY
Figure S1, related to Figure 1. Wild-type
B. fragilis
and
B. fragilis
PSA deletion mutant
produce similar amounts of OMVs an
d have similar proteomic profiles.
Figure S2, related to Figure 2. PSA-containing OMVs given orally to animals suppress
pro-inflammatory cytokines and induce anti-in
flammatory cytokines in colon tissue during
TNBS colitis.
Figure S3, related to Figure 3. OMVs are internalized and localized in the cytoplasm of
DCs, and up-regulate co-stimulatory molecules; neutralization of IL-10R signaling
partially abrogates PSA activity
in vitro
.
Figure S4, related to Figure 4. TLR2-/- DCs internalize OMVs, and purified PSA elicits
IL-10 production from T cells in a TLR2-dependent manner, while PSA-containing OMVs
do not induce IL-10 directly from T cells.
Figure S5, related to Figure 5. Gadd45
-/- DCs internalize OMVs, and Gadd45
is
required in BMDCs to mediate PSA activi
ty during protection from intestinal
inflammation.
μ
Δ
!" # $"%
Δ
&$'()
*'
+ ,
(
(
(
&$%()
*%'
" -
)
(
'
')
&$(.(
*%%
+ ,
'%
(
.
%
&$%%
*%
+ ,
%'
(
(
&$)
*%).
" -
%)
%
%
%
&$%.(
*%(
" - " ##/
%%
%
%.
.
&$(.
*%
+ ,
%
%
&$%.()
*%
+ ,
%'
%
%()
%)
&$')
*('
" -
%
'%
&$(.
*%((
+ ,
%
%)
%%.
%
&$''
*.
+ ,
%%)
.
.
%%
&$'
*%%
" - 0 /1
%%)
%'
)
%
&$%%.
*%%%(
+ ,
%%)
.
%)
%
&$'%
*%'
" -
%)
%%
%
Figure S1, related to Figure 1. Wild-type
B. fragilis
and
B. fragilis
PSA deletion
mutant produce similar amounts of OMVs and have similar proteomic profiles.
(
A
)Wild-type
B. fragilis
and the PSA deletion mutant (
B. fragilis
PSA) produce similar
amounts of OMVs during
in vitro
culture. Total protein recovered from each OMV
preparation was normalized by OD
600
of the culture at the time of harvest. Error bars
indicate SEM. Result is shown from >10 combined experiments performed
independently.
p
value determined by Student’s t-test. NS: not significant. (
B
) OMVs
from wild-type or PSA deletion mutant
B. fragilis
show no significant difference in protein
composition. Proteome mass spectrometry
shows 100% overlap of the identified
proteins (>1 unique peptide identified for each protein) between WT-OMVs and
∆
PSA-
OMVs. Among all of the identified proteins, we semi-quantitatively compared the amount
of those relatively abundant proteins according to the number of unique peptides
identified. The majority of them show no difference between WT-OMV and
∆
PSA-OMV.
Last two columns represent number of unique peptides
±
SEM.
Results are shown from
3 combined experiments performed independently.
Δ
!" # $ %& $ '
(
(
) (
Δ
) (
*'#+ ,-
./
/
0
0/
/
/
*'#+1) .(2 3!(2
Δ
4
4/
4
4/
*'#+1) (2 3!(2
Δ
Δ
Δ
Δ
/
/
/
/
α
Δ
α
Δ
Δ
-
5677& 8,6$
Figure S2, related to Figure 2. PSA-containing OMVs given orally to animals suppress pro-inflammatory cytokines and
induce anti-inflammatory cytokines in colon tissue during TNBS colitis. (A)
Cytokine transcript analysis by qRT-PCR of RNA
recovered from purified CD4+ T cells from mesenteric lymph nodes
. Each symbol represents a single animal. Error bars indicate
SEM from 4 animals/group. Results are shown from
3 combined experiments, each performed independently.
(B)
Cytokine transcript
analysis by qRT-PCR from RNA recovered from whole colons of
each treatment group. Each symbol represents a single animal.
Error bars indicate SEM. Results are shown from
3 combined experiments, each performed independently.
(C)
Cytokine analysis by
intracellular cytokine staining (ICCS) on CD4+ T cells from colon
LPL preparations of each treatment group. Each symbol represe
nts
a single animal. Error bars indicate SEM. Results are representative of 2 independent trials.
(D)
Weight loss in animals (treated with
OMVs rectally) following the induction of TNBScolitis (day 0)
measured as reduction from initial weight until day of sacrifice
(day 3).
All groups contained 3-4 animals, with error bars indicating standard error(SEM). Results are representative of 2 independent t
rials.
(E)
Quantification of colon lengthfrom vehicle treated (EtOH) and TNBS groups. Error bars indicate SEM. ***
p
<0.001. NS:
notsignificant.Results are representative of 2 independent trials.
!"#$
#%&
'
'
(
(
(
())
*#
#%
+"
Δ
,-
+.-/%#
!"#/+"$
+.-/%#
!"#/
Δ
,-$
+"$
Δ
,-$
0
))
)
(
(
$!1$2
$
3
444
()
()
()
()
)
()
)
444
4
444
4
44
"
0
+"$
Δ
,-$
,& 5#%(
3
#6
3
"# '78
Figure S3, related to Figure 3. OMVs are internalized and localized in the cytoplasm of DCs, and up-regulate co-stimulatory
molecules; neutralization of IL-10R signaling partially abrogates PSA activity
in vitro
.
(
A
) Fluorescent micrographs of OMV
(WT or
∆
PSA) internalization by DCs. OMV were labeled with Fluore
scein isothiocyanate (FITC, green) and incubated with cultured
DCs for 2hrs. Cells were fixed and cell membrane was stained with Wheat Germ Agglutinin (WGA)-tetramethylrhodamine (red).
Scale bar: 7.5μm. (
B
) Actin polymerization is required for OMV uptake by DCs
. Flow cytometry analysis of OMV internalization by
DCs pre-treated with Cytochalasin D. OMVs were labeled with FI
TC and incubated with cultured DCs for various times (as indicate
d).
Cells were stained with anti-CD11c. Percentages show CD11c+OMV+ populations (compare to Figure 3A). (
C
) WT-OMVs and
∆
PSA-OMVs up-regulate the co-stimulatory molecule CD86 (B7.2)
for DC activation. FC plots of DCs incubated with WT-OMVs and
∆
PSA-OMVs for various times (as indicated) and stained with
anti-CD11c and anti-CD86. Percentages show CD86+ populations
among CD11c+ cells. (
D
) Quantification of percentage of CD4+ T cells in each proliferating peak (as is labeled in Figure 3H). Error
bars indicate SEM. Results are representative of 3 independent trials. *
p
<0.05; **
p
<0.01; ***
p
<0.001. (
E
) Neutralization of IL-10R
signaling partially abrogates PSA activity.
In vitro
suppression assay was set up as in Figure 3H except that CFSE labeled responder
cells (Teff) were incubated with 20μg/ml of anti-IL-10R (+) or isotype control (-) for 1 hour before addition of Tregs purified
from DC-T
culture under various conditions as indicated. Percentages show total proliferating cells. (Treg:Teff=1:4) Results are represen
tative of
2 independent trials.