Upload
vuongminh
View
223
Download
0
Embed Size (px)
Citation preview
Graphical Abstract. IFNγ and LPS can activate pro-inflammatory gene expression in
macrophages via NF-κB and STAT signaling. Calcium influx also activates calcium-dependent
PLA2
thus releasing poly-unsaturated fatty acids from esterified
lipids. Following inflammation-
induced expression, COX-2 oxidizes poly-unsaturated fatty acids which are further converted to
EFOX by a constitutively active hydroxy-dehydrogenase. EFOX adducts of GSH are exported,
while EFOX adduction to proteins may have effects on protein activity or subcellular
localization. EFOX can also modulate the transcription of pro-inflammatory genes and activate
PPARγ and
the Keap1/Nrf2-dependent gene expression.
IFNγ
PLA2
STAT
R2R1 R2R1
R2R1OH
R2R1OH
R2R1O
R2R1O
R2R1OH
R2R1OH
R2R1O
R2R1O
Dehydrogenase
COX-2
RRO
RRO
RRO
RRO
S
RRO
RRO
RRO
RRO
S
Glutathione
Ca++
PLA2
MRP
NF- Bκ
Proteinadduction
Nucleus
SIGNALING
PPAR activationγ
Ca++
LPS
TLR4
IFNRCd14
KEAP1RR
ORR
ORR
ORR
O
S
Ubiquitylation/proteosomaldegradation
HO-1NQO1
Nrf2
MCP-1IL-6NOS-2
PPARγ
Nrf2Nrf2
KEAP1
COX-2
?
Change in subcellular distribution
EFOX
EFOX GS-EFOX
?INFLAMMATION
GS-EFOX
Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Fig. 1. EFOX produced by THP-1 cells coelute with those produced by RAW264.7
cells. THP-1 cells were differentiated with PMA (86 nM) for 16 h, activated with Kdo2 (0.5 µg/ml) and
IFNγ (200 U/ml), and EFOX levels were detected 8 h post activation. MRM scans following the neutral
loss of 78 were used to detect EFOX-BME adducts.
Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Fig. 2. EFOX are formed in activated primary murine macrophages. Bone marrow
derived macrophages were activated with Kdo2 (0.5 µg/ml) and IFNγ (200 U/ml) and EFOX were
detected 10 h post activation. Enlarged chromatograms are reported in the insets.
Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Fig. 3. BME adducts with α,β-unsaturated oxo-derivatives yield the most reliable
calibration curves for quantification by MS/MS. (a) The compounds 15-OxoETE, 12-OxoETE, 15-
OxoEDE and 9-OxoOTrE were reacted with BME for 2 h and calibration curves were prepared by serial
dilution in the presence of 5-OxoETE-d7 as internal standard. (b-c) Serial dilution of 15-OxoETE, 12-
OxoETE, 15-OxoEDE and 9-OxoOTrE were quantified by MRM in the presence of internal standard (5-
oxoETE-d7) following the neutral loss of CO2 (b) or by SIM (c), following parent mass. All peak areas
corresponding to the compounds were normalized to the internal standard and plotted against their
concentrations.
0.01 0.1 1 10 100
0.01
0.1
1
10
15-Oxo-ETE
[Analyte]/[IS]
9-Oxo-OTrE 12-Oxo-ETE
15-Oxo-EDE
a
0.1 1
0.01
0.1
1
[Analyte]/[IS]
9-Oxo-OTrE 12-Oxo-ETE
15-Oxo-EDE
15-Oxo-ETE
b
0.01 0.1 1
0.01
0.1
1
15-Oxo-ETE
[Analyte]/[IS]
9-Oxo-OTrE 12-Oxo-ETE
15-Oxo-EDE
c
BME-based quantification
CO -Neutral Loss-based quantification2
SIM-based quantification
Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Fig. 4. EFOX production is dependent on RAW264.7 cell activation. RAW264.7
cells were activated with the indicated compounds and EFOX levels were quantified 20 h post activation.
Compound concentrations are as follows: LPS (0.5 µg/ml), Kdo2-Lipid A (0.5 µg/ml), IFNγ (200 U/ml),
PMA (3.24 µM), fMLP (1 µM). Data are expressed as mean ± S.D. (n=4), where * = significantly
different (p<0.01) from “PMA + IFNγ + LPS” and # = a significant difference (p<0.01) between LPS and
“Kdo2 + IFNγ” (one way ANOVA, post-hoc Tukey’s test). Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Fig. 5. EFOX production is time-dependent. RAW264.7 cells were activated with
Kdo2 (0.5 µg/ml) and IFNγ (200 U/ml) and EFOX levels were quantified at indicated times post
activation. Nature Chemical Biology: doi: 10.1038/nchembio.367
17-EFOX-D5 : y = 0.364 x + 0.00309 (r = 0.9996)
[Analyte] / [ IS]
17-E
FOX-
D5
Are
a / I
S A
rea
0 1 2 3 4 5 6 7 8 9 10 11 12 130
1
2
3
4
5
[Analyte] / [IS]
17-E
FOX-
D6
Are
a / I
S A
rea
0 1 2 3 4 5 6 7 8 9 10 110
1
2
3 17-EFOX-D6 : y = 0.339 x + 0.000777 (r = 0.9986)
Supplementary Fig. 6.
Calibration curves of BME-17-EFOX-D6
(3) and BME-17-EFOX-D5 (4). For quantification purposes, calibration curves were made using BME-17-EFOX-D6
(3) (a) and BME-17-EFOX-D5
(4) (b) as analytes
and BME-5-oxoETE-d7 as internal standard. Quantification of EFOX was performed by using analyte/internal standard peak area ratios. The intracellular concentrations of EFOX-D6
and EFOX-D5
were calculated using the calibration lines made with their relative synthetic standards. For the remaining EFOX, quantification was performed by using linear coefficients obtained averaging BME-17-EFOX-D6 (3) and BME-17-
EFOX-D5
(4) derived coefficients.
a
b
Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Fig. 7. EFOX-D6 and EFOX-D4 are derived from the ω-3 series of fatty acids.
RAW264.7 cells were grown for 3 days in DMEM and 10% FBS supplemented with 32 μM of the
indicated fatty acid. On the third day cells were activated with Kdo2 0.5 µg/ml) and IFNγ (200 U/ml) and
EFOX-D6 and EFOX-D4 levels were quantified 20 h post activation.
Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Fig. 8. EFOX formation is dependent on PLA2 and COX-2 activity. RAW264.7 cells
were activated with Kdo2 (0.5 µg/ml) and IFNγ (200 U/ml) in the presence of the indicated inhibitors and
EFOX levels were quantified 20 h post activation. Inhibitor concentrations were as follows: genistein (25
µM), MAFP (25 µM), MK886 (500 nM), ETYA (25 µM) and OKA (50 nM). Data are expressed as mean
+ S.D. (n=4), where * = significantly different (p<0.01) from “Kdo2 + IFNγ” (one-way ANOVA, post-
hoc Tukey’s test). Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Fig. 9. EFOX formation is dependent on COX-2 activity. RAW264.7 cells were
activated with Kdo2 (0.5 µg/ml) and IFNγ (200 U/ml) in presence of the indicated inhibitors and EFOX
levels were quantified 20 h post activation. COX inhibitor concentrations were as follows: ASA (200
µM), indomethacin (25 µM), ibuprofen (100 µM), diclofenac (1 µM) and NS-398 (4 µM). Data are
expressed as mean ± S.D. (n=4), where * = significantly different (p<0.01) from “Kdo2 + IFNγ” (one way
ANOVA, post-hoc Tukey’s test). Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Figure 10. ASA increases production rate of EFOX precursor OH-DPA.
Temporal formation of hydroxy-precursors (MRM 345/327) of EFOX-D5
synthesized in vitro by
purified ovine COX-2 in presence of 10 μM DPA ± 0.2 mM
ASA.
0
5
10
15
20
25
30
0 20 40 60 80 100Time (min)
DPADPA + ASA
0
4
8
12
1 2
(pea
kar
eax1
05)
OH
-DPA
Nature Chemical Biology: doi: 10.1038/nchembio.367
10 20 30 40017-OH-DHA (12)
13-OH-DHA (11)
IFNγ
+ Kdo2
Cox-2 + ASA
Cox-2
RT (min)
IFNγ
+ Kdo2 + ASA
19.519.018.5
10 20 30 400
13-EFOX-D6 (16)
IFNγ
+ Kdo2
IFNγ
+ Kdo2 + ASA
a
RT (min)
b
15.0 16.0 17.0 18.0 19.0
Supplementary Fig. 11 Groeger et al.
17-EFOX-D6 (8)
Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Figure 11. ASA-acetylated COX-2 produces 17-OH-DHA (12), rather than 13-
OH-DHA (11), both in vitro and in activated cells. a, Chromatographic profiles of OH-DHA from
cell lysates of activated RAW264.7 cells (Kdo2 + IFNγ) ±
ASA were compared with OH-DHA
synthesized in vitro by COX-2 in presence of 10 μM DHA ±
ASA and with 13-OH-DHA and 17-
OH-DHA synthetic standards. b, Comparison of chromatographic profiles of EFOX-D6
generated by
activated RAW264.7 cells (Kdo2
+ IFNγ) ±
ASA with 13-EFOX-D6
(16) and 17-EFOX-D6
(8)
synthetic standards. The standards were enzymatically synthesized by incubating 20 µM 13-
or 17-
OH-DHA with 3α-hydroxysteroid
dehydrogenase, in presence of 100µM NAD+
for 10 min at 37°C.
Samples were reacted with BME according to protocol prior to chromatography.
Enlarged
chromatograms are reported in the insets. c, Structures and scheme of the fragmentation pattern and
mass spectra of 13-
and 17-OH-DHA (11, 12) synthesized in vitro by COX-2 in presence of 10 μM
DHA ± ASA compared with synthetic standards.
343.2
281.4
193.2
325.4299.4221.2177.2149.1
203.1
343.2
281.3
201.2
325.4245.2
299.3229.2173.2163.2 227.2
150 200 250 300 350150 200 250 300 350
343.2
281.4
193.2
325.4299.4221.2177.2149.1
203.1
343.2
281.3
201.2
325.4245.2
299.3229.2173.2163.2 227.2
c
OH
O
OH CH3
OH
O
OH CH3
O
OH CH3
OH
O
OH CH3
OH
O
OH CH3
273.1
245.2
221.1
193.1
17-OH-DHA
13-OH-DHA
5x
17-OH-DHA
13-OH-DHA
COX-2 + ASA
COX-2
m/z
273.1
273.1
(6)
(2)
Nature Chemical Biology: doi: 10.1038/nchembio.367
COX-2Actin
COX-2 siRNA
NTG siRNA
Kdo2 +
IFNγ
NT
*
Per
cent
Act
ivat
ed (%
)
0
3.5
nq nq nq nq
EFOX-D6 EFOX-D5 EFOX-D4 EFOX-D3 EFOX-E3 EFOX-E2
100
200
300
a
b
*
*
*
** * *
*
Nor
mal
ized
inte
nsity
(arb
itrar
y un
it)
*
Supplementary Figure 12. Knocking-down COX-2 expression by targeting siRNA
reduces
EFOX formation. RAW264.7 cells were transfected
with COX-2 targeting siRNA
or with the non-
targeting control. At 46h, transfection medium was substituted with activation medium
(SMEM/2%FBS + Kdo2 + IFNγ,
as described in Methods) and cells were harvested 22 h post-
activation. Non activated/non transfected
cells were used as control for basal levels of EFOX and
COX-2, while activated/non transfected
cells were used as control for levels of EFOX and COX-2
following standard activation procedure. Cell lysates
were analyzed for COX-2 expression by western
blot (a) and intracellular EFOX (b). Data are expressed as mean ±S.D. (n=5), where *=significantly
different (p<0.05) compared to Non targeting siRNA
control (NTG siRNA).
Non transfected / Non activatedNon transfectedCOX-2 siRNANTG siRNA
Kdo2/IFNγ
Activated
*
**
Nature Chemical Biology: doi: 10.1038/nchembio.367
OH-DPA as substrate DPA as substrate
0 20 40 60 80 100
4
Time (min)
3
2
1
00 20 40 60 80 100
Time (min)
OH
-DPA
(p
eak
area
A.U
./ m
g pr
otei
n)
3
2
1
00 20 40 60 80 100
Time (min)
N/A
N/A N/A
IFNγ+ Kdo2 IFNγ+ Kdo2
IFNγ+ Kdo2
13-E
FOX-
D
(pea
k ar
ea A
.U./p
rote
in (m
g))
5
No substrate No substrate
DPA as substrate
No substrate
OH-DPA 13-EFOX-D5
(IFNγ+Kdo )2and (N/A) DPA
(IFNγ+Kdo )2 DPA(IFNγ+Kdo )2
DPA(N/A)
DPA(N/A)
4
3
2
1
0
13-E
FOX-
D
(pea
k ar
ea A
.U./p
rote
in (m
g))
5
13-EFOX-D5 13-EFOX-D5
Supplementary Figure 13. A hydroxy-dehydrogenase
activity is constitutive in RAW264.7 cells.
Production of EFOX-D5
and OH-DPA by RAW264.7 cell lysates
supplemented with OH-DPA, DPA,
or vehicle alone. Full and empty symbols indicate activated and non-activated cell lysates
respectively.
Nature Chemical Biology: doi: 10.1038/nchembio.367
17-EFOX-D6 (8)
Time(s)100 200 300 400
Abs
orba
nce
(AU
)
0.025
0.03
0.035
0.04
15-d-PGJ2
Time(s)0 100 200 300 400
Abs
orba
nce
(AU
)
0.035
0.04
0.045
0.05
0.055
Abs
orba
nce
(AU
)
0.034
0.038
0.042
0.046
0.05
Time(s)100 200 300 400
Abs
orba
nce
(AU
)0.03
0.035
0.04
0.045
0.05
0
0
Time(s)100 200 300 4000
15-oxo-ETE
Kobs = 3.92 x 10 -3 s-1 +/- 0.37 x 10 -3
Kobs = 3.49 x 10 -3 s-1 +/- 0.19 x 10 -3
Kobs = 2.95 x 10 -3 s-1 +/- 0.42 x 10 -3
Kobs = 4.77 x 10 -3 s-1 +/- 0.41 x 10 -3
Supplementary Fig. 14a Groeger et al.
a17-EFOX-D6 (7)
Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Fig. 14. EFOX reactivity: (a) EFOX have similar reactivity towards BME compared to
other α,β-unsaturated carbonyl FAs. The pseudo first order reaction rates between BME (50mM) and
different α,β-unsaturated carbonyl FAs (2.9 μM) were measured spectrophotometrically using a Agilent
8453 diode array. (b) Time response of EFOX-D5 reaction with activated RAW264.7 cell extracts. A fast
reaction with free EFOX-D5 was followed by a slower exchange of protein adducted EFOX-D5.
140
5 15 45 98 120 240
Perc
ent M
ax (%
)
Time (Min)
120
100
80
60
40
20
0
b
Nature Chemical Biology: doi: 10.1038/nchembio.367
VPTPNVSVVDLTC244R
VPTPNVSVVDLTC244R
17- EFOX-D5
aa b y b y
V 100.08 -100.08 -
P 197.13 1457.74197.13 1745.02
T 298.18 1360.69298.18 1647.97
P 395.23 1259.64395.23 1546.92
N 509.27 1162.59509.27 1449.87
V 608.34 1048.55608.34 1335.82
S 695.37 949.48695.37 1236.76
V 794.44 862.45794.44 1149.72
V 893.51 763.38893.51 1050.66
D 1008.54 664.311008.54 951.59
L 1121.62 549.281121.62 836.56
T 1222.67 436.21222.67 723.48
C* (244) 1382.7 335.151669.98 622.43
R - 175.12- 175.12
17-EFOX-D5 -adducted non adductedPeptide 1
Supplementary Fig. 15a Groeger et al.
a
Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Fig. 15b Groeger et al.
VIH163DHFGIVEGLMTTVHAITATQK
VIH163DHFGIVEGLMTTVHAITATQK
17-EFOX-D5
baa b y b y
V 100.08 - 100.08 -
I 213.16 2863.61 213.16 2519.31
H* 694.52 2750.52 350.22 2406.22
D 809.55 2269.16 465.25 2269.16
H 946.60 2154.14 602.30 2154.14
F 1093.67 2017.08 749.37 2017.08
G 1150.69 1870.01 806.39 1870.01
I 1263.78 1812.99 919.48 1812.99
V 1362.85 1699.90 1018.55 1699.90
E 1491.89 1600.84 1147.59 1600.84
G 1548.91 1471.79 1204.61 1471.79
L 1661.99 1414.77 1317.69 1414.77
M 1793.04 1301.69 1448.74 1301.69
T 1894.08 1170.65 1549.78 1170.65
T 1995.13 1069.60 1650.83 1069.60
V 2094.20 968.55 1749.90 968.55
H 2231.26 869.48 1886.96 869.48
A 2302.30 732.43 1958.00 732.43
I 2415.38 661.39 2071.08 661.39
T 256.43 548.30 2172.13 548.30
A 2587.46 447.26 2243.16 447.26
T 2688.51 376.22 2344.21 376.22
Q 2816.57 275.17 2472.27 275.17
K 147.11 147.11
non adductedPeptide 2
17-EFOX-D5 -adducted
Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Fig. 15c Groeger et al.
VSNASC149TTNC153LAPLAK
IAMIAM
VSNASC149TTNC153LAPLAK
IAM17-EFOX-D5
aa b y b y
I 114.09 - 114.09 -
V 213.60 1994.10 213.16 1706.82
S 300.19 1895.03 300.19 1607.75
N 414.23 1808.00 414.23 1520.72
A 485.27 1693.96 485.27 1406.68
S 572.30 1622.92 572.30 1335.64
C# 1019.61 1535.89 732.33 1248.61
T 1120.66 1088.58 833.38 1088.58
T 1221.71 987.53 934.43 987.53
N 1335.75 886.48 1048.47 886.48
C* 1495.78 772.44 1208.50 772.44
L 1608.87 612.41 1321.59 612.41
A 1679.90 499.32 1392.62 499.32
P 1776.96 428.29 1489.68 428.29
L 1890.04 331.23 1602.76 331.23
A 1961.08 218.15 1673.80 218.15
K - 147.11 - 147.11
non adductedPeptide 3
c17-EFOX-D5 -adducted
Nature Chemical Biology: doi: 10.1038/nchembio.367
VVDLMVH328MASK
VVDLMVH328MASK
17-EFOX-D5
Supplementary Fig. 15. Mass spectrometric analysis of in vitro reaction of GAPDH with 17-EFOX-D5
(16). Four residues were detected and confirmed as being targets for 17-EFOX-D5 (16) in treated rabbit
GAPDH: Cys 244 (a), His 163 (b), Cys 149 (c) and His 328 (d). Upper panels show 17-EFOX-D5 modified
peptides and lower panels show spectra from corresponding modified and native peptide.
daa b y b y
V 100.80 - 100.80 -
V 199.14 1474.87 199.14 1130.57
D 314.17 1375.80 314.17 1031.50
L 427.26 1260.77 427.26 916.47
M 558.30 1147.69 558.30 803.39
V 657.36 1016.65 657.36 672.35
H# 1138.72 917.58 794.42 573.28
M 1269.76 436.22 925.46 436.22
A 1340.80 305.18 996.50 305.18
S 1427.83 234.14 1083.53 234.14
K - 147.11 - 147.11
non adductedPeptide 4
17-EFOX-D5 -adducted
Nature Chemical Biology: doi: 10.1038/nchembio.367
0
1
2
3
4
0 5 10 15 20Time (min)
GS-
17-E
FOX-
D5
(pea
k ar
ea X
104
)
GST (20 pg/ml)GST (2 pg/ml)GST (0.2 pg/ml)No GST
Supplementary Fig. 16. 17-EFOX-D5 (7) is substrate of glutathione S-transferase (GST). 17-EFOX-
D5 (7) (20 µM) was incubated at 37° C in phosphate buffer at pH 8 with 0.1 mM glutathione and the
indicated concentration of GST. Samples were collected at the indicated time points, excess GST was
removed by acetonitrile precipitation followed by centrifugation. Formation of GS-17-EFOX-D5 (9) was
monitored by nanoLC-MS/MS.
Nature Chemical Biology: doi: 10.1038/nchembio.367
200 300 400 500 600 700
308.2
345.3
420.2
505.3523.3
634.4
200 300 400 500 600 700200 300 400 500 600 700200 300 400 500 600 700 200 300 400 500 600 700
308.2345.3
420.2505.3
523.3
634.4
200 300 400 500 600 700200 300 400 500 600 700
200 300 400 500 600 700
308.2
345.3
420.2
505.3
634.4
523.3
200 300 400 500 600 700200 300 400 500 600 700200 300 400 500 600 700
Rel
ativ
eab
unda
nce
Time (min)0 10 20 30 40
200 300 400 500 600 700
308.2345.3
327.2
420.3233.1
634.4
505.3523.3
200 300 400 500 600 700200 300 400 500 600 700
200 300 400 500 600 700
420.2505.3
634.4
523.3
308.2345.3
200 300 400 500 600 700200 300 400 500 600 700
Time (min)
N/A
0 10 20 30 40
N/A
N/A
+ ASA
N/A
GS-13-EFOX-D5
5.1e2
1.1e4 1.1e4
5.1e2
In vitro synthesized
Medium
Pellet
b
+ ASA
GS-17-EFOX-D5
Supplementary Fig. 17a Groeger et al.
a
523.3 420.2345.3
308.2
( )18
3+
HO
O O
S
HN
NH
O CO2H
O
NH2
CO2H
(15)
Nature Chemical Biology: doi: 10.1038/nchembio.367
Time (min) Time (min)20 25 30 35 40
N/A N/A
N/A N/A
GS-13-EFOX-D6 GS-17-EFOX-D6
0.9e2 0.9e2
In vitro synthesized
Medium
Pellet
Rel
ativ
e ab
unda
nce
1.5e4 1.5e4
250 300 350 400 450 500 550m/z
250 300 350 400 450 500 550m/z
250 300 350 400 450 500 550m/z
250 300 350 400 450 500 550m/z
250 300 350 400 450 500 550m/z
20 25 30 35 40
308.2
343.3418.2 503.3
521.32X
308.2343.3
418.2 503.3
521.32X
308.2343.3
418.2
503.3 521.3
308.2
343.3
418.2
503.3521.3
2X
343.3 418.2503.3
521.32X
308.3
Supplementary Fig. 17. Detection of GS-EFOX-D5 and -D6 adducts in pellets and media of
activated RAW264.7 cells. (a) Structure and fragmentation pattern of GS-13-EFOX-D5 (15). (b, c)
Chromatographic profiles and mass spectra of glutathione adducts of 13- (15) and 17-EFOX-D5 (9) (b)
and -D6 (c) derived from in vitro enzymatic synthesis (upper panels), cell medium (middle panel) and cell
pellet (lower panel). Differences due to recovery efficiency were taken into account by correcting the
signal levels using externally added GS-5-oxoETE-d7. Fragments 345.3 (EFOX-D5) and 343.3 (EFOX-
D6), corresponding to the free lipids, were selected and monitored as the ones giving the best signal to
noise ratio in samples derived from cell media. Fragments 523.3 (EFOX-D5) and 521.3 (EFOX-D6),
corresponding to fragment y2, were selected and monitored as the ones giving the best signal to noise
ratio in samples derived from cell pellet.
IFNγ
+ Kdo2
IFNγ
+ Kdo2
IFNγ
+ Kdo2 + ASA
IFNγ
+ Kdo2 + ASA
c
Nature Chemical Biology: doi: 10.1038/nchembio.367
HO-1 GCLM NQO1*
*
*
*
*
* * *
5 15 25
17-EFOX-D 5
5 10 20 20 20
17-EFOX-D 6
13-OH-D
HA
17-OH-D
HANT 5 15 25
17-EFOX-D 5
5 10 20 20 20
17-EFOX-D 6
13-OH-D
HA
17-OH-D
HANT 5 15 25
17-EFOX-D 5
5 10 20 20 20
17-EFOX-D 6
13-OH-D
HA
17-OH-D
HANT
Actin
5 15 25
17-EFOX-D 5
5 10 20 20 20
17-EFOX-D 6
13-OH-D
HA
17-OH-D
HANT
Nor
mal
ized
Inte
nsity
(R
elat
ive
Uni
t)
14
0
0.9
0
2
0
Nor
mal
ized
Inte
nsity
(R
elat
ive
Uni
t)
40
0
Nuclear Nrf2
* *
Supplementary Figure 18.
a
b
c* *
d
5 15 25
17-EFOX-D 5
5 10 20 20 20
17-EFOX-D 6
13-OH-D
HA
17-OH-D
HANT K/I 5 15 25
17-EFOX-D 5
5 10 20 20 20
17-EFOX-D 6
13-OH-D
HA
17-OH-D
HANT K/I
25
0
iNOS COX-2
IFNγ + Kdo2 IFNγ + Kdo2
** *
*
Nor
mal
ized
Inte
nsity
(R
elat
ive
Uni
t)
Nature Chemical Biology: doi: 10.1038/nchembio.367
75100
37
50
COX-2 (Fig. 3c) Actin (Fig. 3c)
75100
Nrf2 (Fig. 5a)LaminB (Fig. 5a)
Actin (Fig. 5b)
75100
HO-1 (Fig. 5b)
Nqo1 (Fig. 5b) GCLM (Fig. 5b)
25
37
37
50
25
37
25
37
Supplementary Figure 18e.
e
Nature Chemical Biology: doi: 10.1038/nchembio.367
Actin (Fig. 5d) iNOS (Fig. 5d) COX-2 (Fig. 5d)
37
5075
150100
50
75
Supplementary Figure 18. Densitometry, statistical analysis of western blots, full gel images and
measurement of intracellular glutathione (GSH). RAW264.7 cells treated with the indicated
compounds at the indicated concentrations were harvested at 1 h for measurement of nuclear Nrf2 (a)
and 18 h for measurement of HO-1, GCLM and NQO-1 (b). (c) The levels of GSH in RAW264.7 cells
treated with increasing concentrations of EFOX for 22 h were quantified as described in the
supplementary methods. (d) RAW264.7 cells pretreated for 6 h with the indicated compounds
were
harvested 12 h post activation and iNOS
and COX-2 expression was measured by western blot. (e) Full
images of all western blots displayed in the present study. The software ImageJ
was used for
densitometric
measurements. Data are expressed as mean ±S.E.M. (n=3), where *=significantly
different (p<0.05) compared to non treated (“NT”, a-c) or to K/I ( “Kdo2+IFNγ”, d).
e
Nature Chemical Biology: doi: 10.1038/nchembio.367
5 10 20 - - - -
- - - 5 15 25 -
-
-
17-EFOX-D6 (µM) 5 10 20 - - - -17-EFOX-D5 (µM) - - - 5 15 25 -
-
-
5 10 20 - - - -
- - - 5 15 25 -
-
-
IL-6 MCP-1 IL-10
ng/m
g pr
otei
n
250
500
750
25
50
75
4
8
12
0 0 0
Nitr
ite(n
mol
/mg
prot
ein)
17-EFOX-D6 (µM) 5 10 20 - - - -
17-EFOX-D5 (µM) - - - 5 15 25 -
-
-
iNOSCOX-2
Actin
20
40
60
0
a
b
Supplementary Fig. 19. 17-EFOX-D6 and 17-EFOX-D5 modulate inflammatory response in bone marrow-
derived macrophages. Cells were treated with increasing concentration of 17-EFOX-D6 (8) and 17-EFOX-D5 (7)
for 6h and Kdo2 and IFNγ were added. Samples were collected at 12 h. (a) Nitrite levels were measured in the cell
media and normalized by the total protein content; iNOS and COX-2 levels were measured in total cell lysates. (b)
IL-6, MCP-1 and IL-10 levels were measured in cell media and normalized by the total protein content. Data are
expressed as mean ±S.E.M. (n=3), where *=significantly different (p<0.05) compared to Kdo2+IFNγ.
+IFNγ
+ Kdo2 +IFNγ
+ Kdo2 +IFNγ
+ Kdo2
+IFNγ
+ Kdo2
**
*
* *
* * * *
Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Fig. 20. ω-6 derived fatty acids are produced by activated RAW264.7 cells.
RAW264.7 cells were activated with Kdo2
(0.5 µg/ml) and IFNγ (200 U/ml), and oxo-FA levels
were detected 22 h post activation. MRM scans (317/293, 395/317 and 393/315) following the
neutral loss of 78 were used to detect electrophile-BME adducts. The assignment of oxoEPA
to
the 395/315 species produced by activated RAW264.7 cells was based on its mass, adduction
with BME, and the fact that it did not co-elute with 15-dPGJ2
(synthetic oxoEPA
standards were
unavailable).Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Figure 21. Intracellular concentrations of EFOX-D6
, -D5
and -D4
in cells treated
with increasing concentration of 17-EFOX-D6
(8) and -D5 (7). RAW264.7 cells were treated with
the indicated concentrations of 17-EFOX-D6
(7) and -D5
(16) for 18 h and intracellular EFOX were
measured in cell extracts according to protocol. Extracts from activated cells (Kdo2+IFNγ) and non
treated cells (NT) were also analyzed as positive and negative control. Results are reported as
percentage of activated cells (primary y-axis) and as intracellular concentration (secondary y-axes),
estimated as reported in Supplementary Methods. Data are expressed as mean ±S.D. (n=3), where
*=significantly different (p<0.05) compared to “Kdo2+IFNγ”.
Per
cent
act
ivat
ed (%
)
100
0
EFOX-D6
EFOX-D5
EFOX-D4
NT Kdo2+IFNγ
EFOX-D6
5 μM 10 μM 20 μM 5 μM 15 μM 25 μM
EFOX-D5
IC C
once
ntra
tion
(nM
)
300
400 112 n
M49
2 nM76
0 nM
*
*, vs K/I
*
200
* * **
* * *
*
*
*
* * **
*
*
Nature Chemical Biology: doi: 10.1038/nchembio.367
17-EFOX-D6
17-EFOX-D5
-
-
5
-
25
-
-
5
-
25
4.5E5
p65
DN
A Bi
ndin
g (R
FU)
**
**
0
45’ treatment90’ treatment
17-EFOX-D 6(20 μ
M)
17-EFOX-D 5(25 μ
M)
13-OH-DHA (20 μ
M)
17-OH-DHA (20 μ
M)
17-EFOX-D 6(20 μ
M)
17-EFOX-D 5(25 μ
M)
13-OH-DHA (20 μ
M)
17-OH-DHA (20 μ
M)
+IFNγ +Kdo2+IFNγ +Kdo2
Supplementary Figure 22. EFOX inhibit p65 DNA binding activity in vitro. (a)
p65 DNA binding
activity was measured in nuclear cell extracts of RAW264.7 treated with the indicated compound for 6
h and then stimulated with Kdo2 + IFNγ
for the indicated time. P65 DNA binding activity was
measured by TransAM
NFκB
p65 Chemi
Kit (2μg/well, Active Motif, #40097), following
manufacturer’s instructions. (b) Recombinant p65 (Active Motif, #31102) was incubated with the
indicated concentration of 17-EFOX-D6
(8) and –D5
, (7) in phosphate buffer pH 8, at 42°C for 2h.
DNA binding activity was measured by TransAM
NFκB p65 Chemi
Kit (10 ng
p65/well), following
manufacturer’s instructions (the only difference was that DTT was not added).
Data are expressed as
mean ±S.D. (n=4), where *=significantly different (p<0.05) compared to
non treated control.
a
b
p65
DN
A Bi
ndin
g (R
FU)
**
**
**
Nature Chemical Biology: doi: 10.1038/nchembio.367
MS EFOX-BME- MS/MS (EFOX-BME)-BME-
theoretical m/z
measured m/z Δ
ppmtheoretical
m/zmeasured
m/z Δ
ppm
BME-EFOX-E3397.2418 397.2414 0.956596
EFOX-E3319.2279 319.2285
- 1.87953
BME-EFOX-E2 399.2575 399.2572 0.701302 EFOX-E2 321.2435 321.2445 -3.1129
BME-EFOX-D6 419.2262 419.2258 1.03365 EFOX-D6 n.d. n.d. n.d.
BME-EFOX-D5421.2418 421.2417 0.316524
EFOX-D5343.2279 343.2285
- 1.74811
BME-EFOX-D4423.2575 423.2573 0.567031
EFOX-D4345.2435 345.2438
- 0.86895
BME-EFOX-D3425.2731 425.2728 0.627048
EFOX-D3347.2592 347.2596
- 1.15188
Supplementary Table 1.
Theoretical and experimental mass values were determined for the
different precursor BME-EFOX adducts and after collision induced loss of BME.
Nature Chemical Biology: doi: 10.1038/nchembio.367
Supplementary Methods 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Luche Reaction. Samples were added to 1 ml of 200 mM CeCl3 and NaBH4 was added in
excess (~200 mM final). Samples were incubated for 1 h with shaking at RT. The reaction was
terminated by the addition of a few drops of glacial acetic acid and the lipids were extracted by
the method of Bligh and Dyer1. Samples were dried and reconstituted in methanol.
COX-2 reactions. Ovine placental COX-2 (20 U/ml) was preincubated in Tris/heme/phenol
(THP) buffer ± 2 mM ASA at 37°C. THP buffer, freshly prepared before each reaction, consisted
of Tris•Cl (100 mM, pH 8.1), hematin (1 μM), and phenol (1 μM)2,3.
Preparation of primary macrophages. Bone marrow derived macrophages were isolated from
C57BL/6 mice according to the protocol developed by Davies4. Femurs and tibiae were isolated
from euthanized mice, severed at both ends by scissors, and their contents flushed with 10 ml of
DMEM. Freshly isolated bone marrow cells were then differentiated to macrophages by MCSF
produced by L-cells and maintained in DMEM supplemented with 30% L-cell supernatant, 20%
FBS, L-glutamine (2 mM), sodium pyruvate (1 mM), penicillin, and streptomycin.
Western blot. Protein concentrations of samples were measured by BCA assay (Pierce). The
following primary antibodies were used: Nrf2 (Santa Cruz, sc-722), HO-1 (Assay Design, SPA-
896), GCLM (Proteintech, 14241-1-AP), NQO1 (Abcam, ab34173), Cox-2 (Santa Cruz, sc-
1745), iNOS (BD Transduction Lab, 610332), Lamin B1 (Abcam, ab16048), and actin (detected
by Sigma A2066). Secondary antibodies were purchased from Santa Cruz Biotechnology.
Pseudo first order rate calculation for the reaction of α,β-unsaturated carbonyl compounds
with BME. The pseudo first order reaction rates between BME (50mM) and different α,β-
unsaturated carbonyl compounds (2.9 µM) were measured spectrophotometrically using an
Agilent 8453 diode array. The absorbance changes (decrease) were followed at the maximum
absorbance peak for the different species and corresponded to the conjugation of the diene
1 Nature Chemical Biology: doi: 10.1038/nchembio.367
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
(absorbance peak at 234nm) and the carbonyl, resulting in a characteristic shift of maximal
absorbance to higher wavelengths (270-320 nm range). The reaction was carried out in
phosphate buffer pH 7.4 at 37°C and 450 spectra were recorded (at 1 spectrum per sec). The
decrease in absorbance was adjusted to a first order curve using UV-Vis Chemstation (Agilent).
PPARγ activation reporter assay. PPARγ UAS-bla HEK 293H cells (Invitrogen) were used in
the assay and contained a PPARγ ligand-binding domain/GAL4 DNA-binding domain chimera
stably integrated into a parental cell line already engineered with a beta lactamase reporter gene
under control of the UAS response element. PPARγ UAS-bla HEK 293H cells were maintained
in growth medium (DMEM with 10% dialyzed FBS, 500 μg/ml Geneticin, 100 μg/ml
Hygromycin, 1 mM sodium pyruvate, 0.1 mM non-essential amino acids, 25 mM HEPES, and
penicillin/streptomycin). Cells were plated in black-walled clear bottom, poly-D-Lysine coated
384-well plates at a density of 30,000 cells per well in assay medium (Phenol red-free DMEM
with 0.1% charcoal-stripped FBS (Invitrogen)). Cells were then treated with serial dilutions of
indicated ligands and incubated at 37°C and 5% CO2. After 18 h of incubation, LiveBLAzer™-
FRET B/G Loading solution (Invitrogen) was added to the cells and incubated at ambient
temperature for 2 h. A Synergy2 microplate reader (BioTek, Winooski VT) was used to measure
fluorescence intensity at 460 nm and 580 nm emission following excitation at 406 nm. Following
subtraction of the fluorescence background from cell-free wells, the ratio of fluorescence
intensity of 460 nm vs. 530 nm (460 nm/530 nm) was calculated. Data are expressed as mean ±
SD and EC50s were calculated using Log(agonist) vs response--Variable slope equation in
GraphPad PRISM®.
Proteomic analysis of alkylation of GAPDH by 17-EFOX-D5 (7). GAPDH (1.25 mg/ml) was
reacted for 2 h at RT with 100 µM 17-EFOX-D5. Samples were then reduced using 10 mM
2 Nature Chemical Biology: doi: 10.1038/nchembio.367
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
TCEP and free cysteines alkylated with 20 mM iodoacetamide (IAM). The native and 17-EFOX-
D5 modified proteins were trypsinized with sequencing grade trypsin in 50 mM phosphate buffer
pH 7.4, at 37°C overnight using an enzyme:substrate ratio of 1:50 (w/w) and subjected to
nanoLC/ESI MS/MS analysis. A trapping reversed-phase column (Waters Symmetry C18-20
mm-180 µm ID and 5 µm particle size) was loaded with 1 µl of digested sample at a 5 µl /min
flow rate for 3 min in 1 % solvent B (acetonitrile + 0.1 % Formic acid). Peptides were resolved
using a Waters BEH130 column (C18-100 mm-75 µm ID and 1.7 µm particle size) at 40 °C with
a linear gradient of solvent B (1–40% in 59 min) at a flow rate of 0.5 µl/min followed by a linear
gradient of solvent B (60–80% in 20 min). Electrospray voltage was 1.4 kV, and capillary
temperature was 200°C. Peptides were detected in the positive ion mode using a mass range of
100–2000. Instrument and source parameters were optimized by tuning for GS-17-EFOX-D5 (9)
adducts. Peptide identification was performed using Bioworks 3.0 with variable modifications
set as methionine +/- oxidation (16 Da), cysteine +/- IAM or +/- 17-EFOX-D5 (7).
Measurement of glutathione adducts. Cells were seeded in T175 flasks and were activated as
described above. Cell pellets and culture media were collected 21 h post-activation. Pellets were
resuspended in 200 µl phosphate buffer and exposed to freeze-thaw cycles using liquid nitrogen.
Cold acetonitrile was added, protein precipitate was removed by centrifugation and GSH adducts
were measured in the supernatant. GSH adducts were measured in cell culture media after C-18
solid phase extraction. GS-5-oxoETE-d7 was added as internal standard to cell pellets (before
protein precipitation) and to cell media (before C-18 extraction). In vitro enzymatic synthesis of
GSH adducts was performed by incubating the indicated FA with 0.1 mM glutathione in
presence of glutathione S-transferase (GST, 0.02 ng/ml) in 50 mM Tris-HCl buffer at pH 8.5, at
37°C for 15 min. The reaction was stopped by adding acetonitrile/0.1 % formic acid and GST
3 Nature Chemical Biology: doi: 10.1038/nchembio.367
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
was removed by centrifugation. GSH adducts were analyzed by nano-LC-MS/MS using
nanoACQUITY UltraPerformance LC coupled with Thermo-Fisher LTQ. Waters XBridge
BEH130 C18 NanoEase Column (3.5 µm, 100 µm x 100mm) was used. Chromatography was
performed using a binary flow system consisting of A (H2O/0.1% formic acid) and B
(ACN/0.1% formic acid) at 0.5 µl/min under the following conditions: hold at 1.5% B for 3 min,
then 1.5 to 30 % B in 10 min, then 30 to 70 % B in 27 min. The following parent ions, 633.3,
650.3 and 652.3, were monitored and fragmented for identification of GS-5-oxoETE-d7, GS-17-
EFOX-D5 (9) and GS-17-EFOX-D6 (10), respectively.
Theoretical and experimental mass values. Accurate mass of the different EFOX-BME
adducts were obtained from cell extracts using an HPLC-coupled LTQ/Orbitrap hybrid mass
spectrometer (Thermo Fisher Scientific, Bremen, Germany) equipped with a HESI probe. Sheath
and auxiliary gases were set at 65 and 35 respectively. Capillary and source heater temperatures
were set at 300°C and 475°C respectively. The source voltage was set at 2 kV and the S-lens RF
level at 40%. For MS/MS experiments in the Orbitrap, predefined ions were isolated with an
isolation width of 2.5 Da and collision-induced dissociation was conducted in the LTQ (CID, 35
%) or the c-trap (HCD, 25 %) with an activation q of 0.25, an activation time of 25 ms and the
resulting product ions were measured with resolving power ranging from 7.5 to 100K. M/z
values of the MS and MS/MS scans were determined from at least 3 independent experiments
and reported as the mean value. Standard deviation values for MS scan modes were lower than 1
ppm for all analyzed samples.
GSH determination. The levels of GSH were determined following the method reported by
Harwood et al. Labeled GSH (containing glycine 1,2-13C2,15N) was obtained from Cambridge
Isotope Laboratories (Andover, MA, USA). Briefly, RAW264.7 cells seeded in 12 well plates
4 Nature Chemical Biology: doi: 10.1038/nchembio.367
5
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 19 20
were supplemented with different concentrations of 17-EFOX-D6 (8) and 17-EFOX-D5 (7) for 22
h. Cells were then scraped, pelleted, resuspended in phosphate buffer containing NEM (10 mM)
and subjected to a freeze thaw cycle. After 30 min incubation, 1 µM labeled GS-NEM was added
and protein was precipitated using ACN. GS-NEM was quantified by reverse phase HPLC-MS-
MS (Thermo Hypercarb column) using the following transitions: 433.3/304.2 (GS-NEM) and
436.3/307.2 (labeled GS-NEM). Calibration curves were prepared for GS-NEM (0.01-100
pmoles), using labeled GS-NEM (10 pmoles) as standard.
REFERENCES
1. Bligh, E.G. & Dyer, W.J. A rapid method of total lipid extraction and purification. Can J Biochem Physiol 37, 911-7 (1959).
2. Gierse, J.K. & Koboldt, C.M. Current Protocols in Pharmacology. in Cyclooxygenase Assays (ed. S.J., E.) (John Wiley and Sons, Inc., 1998).
3. Serhan, C.N. et al. Resolvins: a family of bioactive products of omega-3 fatty acid transformation circuits initiated by aspirin treatment that counter proinflammation signals. J Exp Med 196, 1025-37 (2002).
4. Davies, J.Q. & Gordon, S. Isolation and culture of murine macrophages. Methods Mol Biol 290, 91-103 (2005).
Nature Chemical Biology: doi: 10.1038/nchembio.367