1. Art and Diseño

Supporting Information
Sharma et al. 10.1073/pnas.1420217112
SI Experimental Procedures
Generation and Characterization of Autoimmune STING-Deficient
Mice. For mortality assessment, additional STING+/−lpr/+ mice
were intercrossed to produce a second cohort of STING/lpr mice
(n = 15, with 12 deaths during the observation period) and WT/
lpr mice (n = 9, with five deaths), which were observed without
intervention until the time of death. A second F2 cross also was
established for fully backcrossed C57BL/6 STING−/− (kindly
provided by D. Stetson, University of Washington, Seattle) and
MRL.Faslpr mice. Identical results were obtained, and data are
cumulative from both crosses. A similar cross was set up for
IRF3−/− and MRL/lpr mice to generate cohorts of experimental
F2 mice and analyzed as above for STING/lpr mice (n ≥10 per
group). In both cohorts severely moribund animals were killed for
humane concerns and were included as deaths in the analysis. For
the TMPD-induced peritonitis, 6-wk-old C57BL/6, Unc93b3d/3d,
and STING−/− mice were injected i.p. with TMPD as described
previously (1) and were evaluated at day 14. The University of
Massachusetts Medical School (UMMS) and Yale Institutional
Animal Care and Use Committees approved all animal work.
Determination of Autoantibody Profiles. Images were captured on
a Nikon E600 at 200× magnification and constant excitation and
were processed in Adobe Photoshop. Autoantibody reactivity
was examined further using an autoantigen proteomic array
containing 88 autoantigens and 10 control proteins (2). Serum
samples, diluted 1:100, were detected with Cy3-labeled anti-mouse
IgG and Cy5-labeled anti-mouse IgM (Jackson ImmunoResearch),
and Tiff images were generated. Genepix Pro-6.0 software was used
to analyze the images and generate GenePix Results format (.GPR)
files. Net fluorescence intensities were defined as the spot intensity
minus background fluorescence intensity; data obtained from duplicate spots are averaged. The signal-to-noise ratio (SNR) was used
as a quantitative measure of the ability to resolve true signal from
background noise. A higher SNR indicates higher signal over
background noise. SNRs equal to or higher than 3 are considered
true signal-to-background noise ratios. The autoantibody levels
were normalized to internal controls, all values were scaled relative
to the minimum value [x − min(x)], and a pseudocount of 1 was
added to each value. This operation sets the smallest value in the
dataset to 1. All values were log10-transformed, and a heatmap was
generated using open-source R-based software at UMMS. Hierarchical clustering of data as noted in the R program is reported.
Gene-Expression Analysis. RNA isolated from total spleen was used
for quantitative RT-PCR using SYBR Green PCR Master Mix (BioRad) with the following primer pairs: β-actin, sense, 5′-TGGCATAGAGGTCTTTACGGA-3′, antisense, 5′-TTGAACATGGCATTGTTACCAA-3′; TLR7, sense, 5′-ATGTGGACACGGAAGAGACAA-3′, antisense 5′-GGTAAGGGTAAGATTGGTGGTG;
TLR9, sense, 5′-ATGGTTCTCCGTCGAAGGACT-3′, antisense
5′-GAGGCTTCAGCTCACAGGG-3′; TLR3, sense, 5′-GTGAGATACAACGTAGCTGACTG-3′, antisense 5′-TCCTGCATCCAAGATAGCAAGT-3′; GAPDH, sense, 5′-AGGTCGGTGTGAACGGATTTG-3′, antisense 5′-TGTAGACCATGTAGTTGAGGTCA-3′; A20 (TNFAIP3), sense 5′-ACAGTGGACCTGGTAAGAAAACA-3′, antisense 5′-CCTCCGTGACTGATGACAAGAT-3′; SOCS1, sense 5′-CTGCGGCTTCTATTGGGGAC-3′, antisense 5′-AAAAGGCAGTCGAAGGTCTCG-3′; SOCS3, sense
5′-ATGGTCACCCACAGCAAGTTT-3′, antisense 5′-TCCAGTAGAATCCGCTCTCCT-3′. For purified cell populations, BMMs
were prepared as described (3). Activated cells were harvested
Sharma et al. www.pnas.org/cgi/content/short/1420217112
into RLT buffer containing 2-mercaptoethanol for subsequent
processing with the RNeasy Mini kit (Qiagen). Each RNA sample
was adjusted to contain the same quantity by using the Nanodrop
ND-1000 spectrophotometer (Thermo Scientific). RNA then was
hybridized and quantified with the NanoString nCounter analysis
system (NanoString Technologies) per the manufacturer’s protocol.
The gene-expression data first were normalized to an internal
positive control set, then to an internal negative control set, and
then to seven housekeeping genes, i.e., GAPDH, β-glucuronidase,
β-actin, hypoxanthine phosphoribosyltransferase 1, tubulin β, phosphoglycerate kinase 1, and clathrin H chain 1. All values were log10transformed, and a heatmap was generated using the open-source
R-based software at UMMS as stated above.
Flow Cytometry. The following conjugated anti-mouse Abs were
used: peridinin chlorophyll (PerCP)-Cy5.5 anti-CD19 (1D3), phycoerythrin (PE)-Cy7 anti–Ly-6G (1A8), allophycocyanin (APC)-e780
anti-CD11b (M1/70), PerCP-Cy5.5 anti-TCRβ (Η57.597), and PECy7 anti-CD95 (Jo-2) (BD Biosciences); PerCP-Cy5.5 anti-F4/80
(BM8), V500 anti-CD4 (GK1.5), eFluor605NC anti-CD8α (53-6.7),
and efluor450 anti-CD45R (RA3-6B2) (eBioscience); efluor450
anti-CD11c (3.9) and APC anti–Ly-6C (ER-MP20) (Serotec); FITC
PNA (Vector Labs); APC anti-CD93 AA4.1, PE anti-CD138 (281-2),
APC anti-NK1.1 (PK136), FITC anti-CD86 (GL1), efluor450 anti–
IFN-γ (XMG1.2), APC-Cy7 anti-CD44 (IM7), PE-Cy7 anti-CD69
(H1.2F3), APC anti–Siglec-H (eBio440c), and PE anti-CD317 (Bst-2,
ebio927) (eBioscience); and FITC anti–IFN-α (RMMA-1; PBL IFN
Science). Treg cells were stained using APC-efluor780 anti-CD4
(RM 4–5; eBioscience), APC anti-CD25 (PC61; eBioscience), and
FITC anti-FOXP3 (FJK-16s; eBioscience). Live-Dead Blue and
Aqua dead cell stains (Invitrogen) were used to distinguish between live and dead cells.
Cells were incubated in either CD16/32 (Fc Block; BD Biosciences) or 2.4G2 supernatant (2.4G2 hybridoma, ATCC) before
staining. Cellular DNA was determined by flow cytometry using
propidium iodide or 7-aminoactinomycin D (BD Biosciences).
Intracellular staining of IFN-γ and IFN-α was carried out on
splenic single-cell suspensions preincubated with Monensin (BD
Biosciences) for 2 h and then with Fc block (2.4G2) for 15 min.
Cells were permeabilized and fixed with BD Cytofix-Cytoperm
buffer and subsequently incubated with FITC-conjugated rat
monoclonal antibody to IFN-α (Clone RMMA-1; PBL) or e450conjugated monoclonal antibody to IFN-γ (eBioscience). Cells
were acquired on a LSR II (BD Biosciences) and analyzed with
FlowJo software (Tree Star).
Cell Cultures. BMMs were derived as described previously (3).
Briefly bone marrow extracts were expanded in vitro in the
presence of L929 supernatants for 7 d. The rested BMMs then
were stimulated with TLR ligands, CpG B (1826 ODN; 5 μM;
IDT), CLO97 (300 ng; InvivoGen), pI:C (100 μg/mL; InvivoGen),
CpG A (2336 ODN; 1 μM IDT), Sendai virus (Cantrell strain;
Charles River Laboratories; 200 high activity units), and poly(dA-dT)
(1 μg/mL; Sigma), ISD as previously described (4) (3μM IDT)
for 20 h or for the times specified. Transfection of intracellular
ligands was done with Lipofectamine 2000 from Invitrogen. Fulllength murine STING or a scrambled sequence was cloned into the
pRetro-Zeocin (pRZ) vector and tagged with HA and citrine. The
production of viral particles and transduction of target cells were
conducted according to the protocols at www.broad.mit.edu/
genome_bio/trc/publicProtocols.html. Zeocin selection (encoded
1 of 6
in the pRZ vector) was used until stable expression was obtained.
Overexpression of STING was verified by Western blot analysis.
Determination of Autoantibody Profiles. ANAs were detected by
immunofluorescence on HEp-2 slides (Antibodies, Inc.) as previously described (1).
Cytokine Assays. A Luminex screen was performed on the serum
using a Bio-Plex 23-plex kit (Bio-Rad) and a Bio-Plex 200 reader
(Bio-Rad) per the manufacturer’s instructions.
ELISA. Serum levels of IL-6, TNF-α, and IFN-γ protein levels
were determined by ELISA (eBioscience). The murine IFN-β kit
was as described (3). An anti-mouse albumin ELISA kit was used
to measure urine protein (Bethyl Laboratories).
The interstitial score was determined by examining 20 high-power
fields and scoring the interstitial inflammation on a scale from 0 to
4 as absent or involving <25, 25–50, or >50% of the interstitium.
IDO-1 Staining. Spleen tissue was snap-frozen in optimum cutting
temperature (OCT) medium. For IDO-1 staining cryostat sections were fixed in cold methanol and processed as previously
described (6). Anti-mouse IDO-1 (Rockland Immunochemicals)
and anti-mouse DyLight-488 (Sigma) were used to visualize
IDO-1 levels in the spleen. Nuclei were counterstained with
DAPI (Invitrogen). Images were captured on a Leica SP8 confocal microscope.
Determination of Clinical Disease. H&E-stained kidney sections
were scored in a blinded fashion by three independent observers
to determine a glomerular and interstitial inflammation score as
described (5). Briefly, a mean glomerular score was calculated for
each mouse by grading injury in 20 glomeruli. Glomeruli were
scored as follows: 0, normal; 1, mesangial expansion; 2, endocapillary proliferation; 3, capillaritis or necrotic changes; 4, crescents.
ELISPOT Assay. To detect antibody-forming cells (AFCs) that make
κ light-chain IgG antibodies, 96-well multiscreen filter plates
were coated overnight at 4 °C with 5 μg/mL polyclonal goat-anti
mouse κ (1050–01; Southern Biotech). Nonspecific binding was
blocked with 1% BSA in PBS, and serially diluted cell suspensions were plated and incubated at 37 °C for 8 h. Biotinylated
secondary antibodies to IgG (1030-08; Southern Biotech) were
detected with alkaline phosphatase-conjugated streptavidin and
bromo-4-chloro-3-indolyl phosphate substrate (Southern Biotech).
1. Bossaller L, et al. (2013) Overexpression of membrane-bound fas ligand (CD95L) exacerbates autoimmune disease and renal pathology in pristane-induced lupus. J Immunol 191(5):2104–2114.
2. Li QZ, et al. (2007) Protein array autoantibody profiles for insights into systemic lupus
erythematosus and incomplete lupus syndromes. Clin Exp Immunol 147(1):60–70.
3. Sharma S, et al. (2011) Innate immune recognition of an AT-rich stem-loop DNA motif
in the Plasmodium falciparum genome. Immunity 35(2):194–207.
4. Stetson DB, Medzhitov R (2006) Recognition of cytosolic DNA activates an IRF3dependent innate immune response. Immunity 24(1):93–103.
5. Richez C, et al. (2010) IFN regulatory factor 5 is required for disease development in the
FcgammaRIIB-/-Yaa and FcgammaRIIB-/- mouse models of systemic lupus erythematosus. J Immunol 184(2):796–806.
6. Ravishankar B, et al. (2012) Tolerance to apoptotic cells is regulated by indoleamine
2,3-dioxygenase. Proc Natl Acad Sci USA 109(10):3909–3914.
Sharma et al. www.pnas.org/cgi/content/short/1420217112
2 of 6
Total Cell Count (x108)
1
0
ST
E
Spleen
Lymph-node
n.s
p=0.0049
p=0.0003
p=0.0012
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.5
ST
W
T/
lp
r
IN
G
/lp
r
W
T/
lp
r
ST
IN
G
/lp
r
0.0
Females
Males
3.0
2.0
1.0
0
W
T/
lp
r
ST
IN
G
/lp
r
W
T/
lp
r
ST
IN
G
/lp
r
1.5
F
n.s
p=0.0492
2.0
gm
5
4
3
2
1
0
W
T/
lp
r
IR
F3
/lp
IR
r
F3
+/
-/l
pr
2.5
2.0
1.5
1.0
0.5
0.0
W
T/
lp
r
IR
F3
/lp
IR
r
F3
+/
-/l
pr
Total Cell Count (x108)
D
2
0.8
0.4
0
Females
IR
F
F 3 3/lp
+/ r
-/l
p
W r
T/
lp
r
0
6
2
IR
0.2
10
W
T/
lp
r
IN
G
/lp
r
W
T
ST / W T
IN
ST G/
IN WT
G
+/
-/l
pr
1
0.4
p=0.0068
ST
3
p=0.004
3
W
T/
lp
r
IN
G
/lp
r
W
T
ST / W T
IN
ST G/
IN WT
G
+/
-/l
pr
p=0.0077
5
ST
Axillary LN
C
DN B220+ T-cells
W
T/
lp
r
IN
G
/lp
r
W
T/
W
ST
T
IN
ST G
IN /WT
G
+/
-/l
pr
T/
lp
r
ST
IN
G
/lp
W
Spleen
Total Cell Count (x108)
B
r
A
Males
Fig. S1. STING deficiency but not IRF3 deficiency promotes disease severity in MRL/lpr mice (related to Figs. 1 and 2). (A) Representative images of spleen and
axillary lymph node from 16-wk-old WT/lpr and STING/lpr mice. (B) Total numbers of splenic B220+ TCRβ+ CD4− CD8− DN cells in 16- to 25-wk-old WT/lpr, STING/
lpr, WT/WT, STING/WT, and STING+/−/lpr mice. (C) Total numbers of B220+ TCRβ− B cells and B220− TCRβ+ T cells in the spleens of 16- or 25-wk-old STING/lpr mice
and littermates as determined by flow cytometry. (D) As in C; total cell numbers of splenic CD19+ TCRβ− B cells and CD19− TCRβ+ T cells in WT/lpr, IRF3/lpr, and
IRF3+/−/lpr mice (E) Weights of spleens and largest axillary lymph nodes) from F2 littermates segregated by sex. (F) Percentage of live TCRβ− CD44+ CD138+
CD22low AFCs in the spleens of WT/lpr, IRF3/lpr, and IRF3+/−/lpr mice. n = 16 or more mice per group. Wider horizontal bars represent the means of each group,
and error bars represent SEM. WT/lpr, IRF3/lpr, and IRF3+/−/lpr mice were analyzed at age 22 wk; WT/lpr and STING/lpr mice were analyzed at age 16 wk. P
values for WT/lpr vs. STING/lpr in B and C were calculated using two-way ANOVA.
Sharma et al. www.pnas.org/cgi/content/short/1420217112
3 of 6
CD11b+
p=0.0081
STING/lpr
wt/wt
p=0.0112
50
R1
40
6
30
4
R2
20
0
t
t
10
0
/w
t/w
G
ST
Ly6G
Neutrophils
Macrophages
8
10
8
6
4
2
0
6
6
4
4
2
2
F3
/lp
r
W
T/
lp
IN r
G
/lp
r
w
t/w
ST
t
IN
G
/w
t
W
T/
lp
r
0
0
IR
8
p=0.0007
F3
/lp
r
CD86+ F4/80+ CD11b+ Mø
ST
Total Cell Count x106
p=0.0019
W
T
ST /lpr
IN
G
/lp
r
w
t/w
ST
t
IN
G
/w
t
25
20
15
10
5
0
p=0.0015
10
8
6
4
2
0
W
T
ST /lpr
IN
G
/lp
r
w
t/w
ST
t
IN
G
/w
t
p=0.0071
W
T
ST /lpr
IN
G
/lp
r
w
t/
ST wt
IN
G
/w
t
25
20
15
10
5
0
E
F4/80+ CD11b+ Mø
Ly6Ghi Granulocytes
IR
D
Ly6Chi Monocytes
Total Cell Count x107
W
T/
lp
r
w
IN
ST
W
T
ST /lpr
IN
G
/lp
r
2
C
Total CD86+ Cell Count (x105)
WT/lpr
B
Ly6C
8
CD11c+
W
T/
lp
r
IN
G
/lp
r
w
t/w
ST
t
IN
G
/w
t
Total Cell Count (x107)
A
p=0.0472
500
200
100
0
0
0
0
/lp
r
t
ST /w
IN t
G
/w
t
w
G
50
0
r
/lp
G
IN
ST
W
T/
lp
r
/lp
G
200
r
200
/lp
200
100
300
G
400
IN
400
IN
p=0.0047
ST
400
W
T/
lp
r
ST
IN
G
/lp
r
600
ST
150
p=0.0245
400
600
W
T/
lp
r
MIP-1
W
T/
lp
r
800
MIP-1
r
800
p=0.0193
KC
600
r
(pg/ml)
800
p=0.0409
CCL5 (RANTES)
/lp
CCL2 (MCP-1)
G
G
CD11c
IN
F4/80
ST
F4/80+
n.s.
W
T/
lp
r
Ly6C
% Max
F4/80-
Inflammatory DCs
5
4
3
2
1
0
IN
wt/wt
ST
STING/lpr
W
T/
lp
r
WT/lpr
CD11b+
Total CD86+ Cell Count (x105)
F
Fig. S2. STING deficiency but not IRF3 deficiency promotes the expansion and activation of myeloid-derived cell populations (related to Fig. 4). (A) Total
number of cells positive for CD11b and CD11c. (B) Identification of Ly6Chi CD11b+ inflammatory monocytes (R1) and Ly6Cint Ly6Ghi CD11b+ N1 neutrophils (R2).
(C) Total CD86+ cell counts of CD11b+ Ly6Chi monocytes and CD11b+ Ly6Ghi granulocytes. (D) Counts of total F4/80+ and CD86+ F4/80+ macrophages in STING/
lpr mice and littermates. Mice were 16 wk old unless otherwise noted. (E) Total counts of splenic neutrophils and macrophages in 22-wk-old IRF3/lpr mice and
littermates. (F) Identification of CD11b+ F4/80− Ly6Chi CD11c+ iDCs and total CD86+ cell counts of iDCs. (G) Serum levels of chemokines determined by Luminexmultiplex array. P values were determined by Student’s t test.
Sharma et al. www.pnas.org/cgi/content/short/1420217112
4 of 6
C
B
CD11b+
6
4
WT/lpr
STING/lpr
Ly6C
8
4
2
2
IFN
/lp
W
T/
lp
r
G
/lp
r
N
TI
r
0
0
R1
F3
Bst2
Siglec-H
p=0.0134
6
IR
STING/lpr
Total numbers of
SiglecH+ cells (x106)
WT/lpr
CD11c+
W
T/
lp
r
A
% Max
S
R2
IFN
Fig. S3. STING deficiency but not IRF3 deficiency promotes the enrichment of IFN-α+ lymphoid-derived pDCs and Ly6C+ cells (related to Fig. 4). (A) Gating
strategy for cells positive for intracellular staining of IFN-α within the CD11c+ Bst-2+ SiglecH+ pDC subset. (B) Total numbers of SiglecH+ pDCs in the spleens of
16-wk-old WT/lpr and STING/lpr mice and in 22-wk-old IRF3/lpr mice and WT/lpr littermates. (C) Identification of cells positive for intracellular staining of IFN-α
within the Ly6Chi CD11b+ inflammatory monocyte and Ly6Clo CD11b+ monocyte populations. R1 and R2 gates were used to designate Ly6Chi and Ly6C+/lo cells.
P values are calculated using two-way ANOVA and are reported for STING/lpr vs. WT/lpr.
p<0.0001
p<0.0001
poly I:C ( g/ml)
0
0
30
10
30
3
10
1
30
10
0
30
0
0
3
10
C57Bl/6
STING-/-
C
(
C -)
pG
B
LO
97
C
(
C -)
p
C GB
LO
97
dT
IS
D
A
LPS(ng/ml)
50
40
30
20
10
0
C57Bl/6
STING-/-
STING-/-
poly I:C ( g/ml)
5000
3000
1000
0
10
30
10
0
30
0
C57Bl/6
pd
(
pd -)
A
dT
IS
D
STING-/-
1000
800
600
400
200
0
4
2
0
0
0
1
3
Pam2Cys4 ( g/ml)
30
20
10
2
0
IFN (IU/ml)
C57Bl/6
0
(
pd -)
A
dT
IS
D
(
C -)
p
C GB
LO
97
(
C -)
pG
C B
LO
97
(
pd -)
A
dT
IS
D
STING-/-
4
1
)
0
6
2
30
10
0
30
0
2
p<0.0001
8
(-
5
0
IL-6 ng/ml
IL-6 ng/ml
10
4
p<0.0001
3
15
6
B
p<0.0001
3
10
20
8
C57Bl/6
p<0.0001
p=0.0002
p<0.0001
1
BMMs
10
0
TNF ng/ml
A
Fig. S4. STING deficiency modulates TLR signaling strength (related to Fig. 6). (A) Levels of TNF-α and IL-6 secreted by BMMs from WT and STING−/− mice in
response to TLR9 (CpGB, 5 μM), TLR7 (CLO97, 300 ng/mL), poly (dA:dT) (1 μg/mL), and ISD (3 μM) ligands. Supernatants were collected 24 h after stimulation. (B)
Levels of IL-6 secreted by C57BL/6 and STING−/− BMMs treated with noted concentrations of various TLR ligands: Pam2Cys4 (TLR2 ), poly(I:C) (TLR3), and LPS
(TLR4) for 24 h. (C) Levels of IFN-β secreted by C57BL/6 and STING−/− BMMs treated with a concentration range of poly(I:C) for 24 h.
Sharma et al. www.pnas.org/cgi/content/short/1420217112
5 of 6
B
TLR3
TLR7
9
C 7
pG
B
)
(-
LO
(C
LO
(-
9
C 7
pG
B
C
0
B
0
pG
2
C
2
SOCS3
250
200
150
100
50
0
)
4
)
4
TLR9
D
0
l
)
g/
m
(30
30
C
LO
97
97
LO
C
IS
D
0
)
0
(-
10
M
50
5
20
1
B
30
100
pG
150
2
C
3
10
8
6
4
2
0
0n
pG
C
(-
LO
C
B
LO
pG
C
)
(C
40
IS
D
0
200
)
0
TNF
(-
0
4
()
g/
m
C
l
pG
B
5
M
200
IL-6
RAWS+EV
RAWS+STING
0n
100
B
5
pG
400
C
200
LO
10
C
600
B
300
)
15
97
800
97
400
20
)
SOCS3
(-
WT/lpr
STING/lpr
SOCS1
97
A20 (TNFAIP3)
Relative Expression (A.U.)
C
Relative Expression (A.U.)
SOCS1
6
97
C57Bl/6
STING-/-
LO
150
100
50
10
5
0
A20 (TNFAIP3)
6
C
Relative Expression (A.U.)
BMMs
ST WT
IN /lp
G r
S /lpr
TI w
N t
G
-/
ST WT IN /lp
G r
S /lpr
TI w
N t
G
ST WT /IN /lp
G r
S /lpr
TI w
N t
G
-/-
TLR expression/GAPDH A.U.
A
Fig. S5. STING expression regulates basal levels of key negative regulators and TLR-signaling strength but not TLR levels (related to Fig. 7). (A) Real-time PCR
analysis of resting levels of TLR3, -7, and -9 in BMMs derived from WT/lpr, STING/lpr, WT/WT, and STING/WT littermate mice relative to the housekeeping gene
GAPDH. (B and C) Protein levels of A20 and SOCS1 in C57BL/6 and STING−/− BMMs (B) or WT/lpr- and STING/lpr-derived BMMs (C) either left untreated or
treated with endosomal TLR ligands CpGB (5 μM) and CLO97 (300 ng/mL) for 3 h. (D) Secreted levels of IL-6 and TNF-α from the RAW 264.7 macrophage cell
line stably overexpressing STING in a retrovector (pRetroZeocin) or a control empty retrovector (EV), treated with CpG-B, CLO97, and ISD (3 μM), 3 h after
stimulation.
Sharma et al. www.pnas.org/cgi/content/short/1420217112
6 of 6