37
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 PLA 2 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. IFN γ PLA 2 STAT R 2 R 1 R 2 R 1 R 2 R 1 OH R 2 R 1 OH R 2 R 1 O R 2 R 1 O R 2 R 1 OH R 2 R 1 OH R 2 R 1 O R 2 R 1 O Dehydrogenase COX-2 R R O R R O R R O R R O S R R O R R O R R O R R O S Glutathione Ca ++ PLA 2 MRP NF- B κ Protein adduction Nucleus SIGNALING PPAR activation γ Ca ++ LPS TLR4 IFNR Cd14 KEAP1 R R O R R O R R O R R O S Ubiquitylation/ proteosomal degradation HO-1 NQO1 Nrf2 MCP-1 IL-6 NOS-2 PPARγ Nrf2 Nrf2 KEAP1 COX-2 ? Change in subcellular distribution EFOX EFOX GS-EFOX ? INFLAMMATION GS-EFOX Nature Chemical Biology: doi: 10.1038/nchembio.367

Nature Chemical Biology: doi: 10.1038/nchembio Chemical Biology: doi: 10.1038/nchembio.367. Supplementary Fig. 1. EFOX produced by THP-1 cells coelute with those produced by RAW264.7

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