SUPPLEMENTARY INFORMATION

Transcription

1 SUPPLEMENTARY INFORMATION doi: /nature11351 a 3.. b 3.. #$%&%&'()(%$ #:=>,3 > / ,9:5 9:+&' 9:+&' ;555< c 3.. #$%&%&'()(%$ /.. *+ #$%,-, &'$(,-, &)*+,,-, #$%&%&'()(%$ /. #$%&%&'()(%$ /...?..3.?.3.? :+&';@A-B9< /. 32. /0. +(BC;B(8< e /. f 3.. G6,3;8A-B9< 3F 3. F.. D. 1. E. 3/. 3F. 32. /3. /0. +(BC;B(8< HIJK(K: @A-A /@A-A 3@A-A /..?F@A-A.?/F@A-A.?3/F@A-A /.. D F.. L(8I%C=>&=%(8MC)%(&8 1.. Supplementary Figure 1. Characteriza6on of FlaTox. (a d) Lysis of bone marrow derived macrophages (BMDM) was quan=fied by lactate dehydrogenase release assay. (a) Wild type (B6) 4 h with 4 µg/ml PA, 2 µg/ml all others. Secreted proteins were probed with CASP1 p10 an=body. (b) Indicated genotypes 4 h with FlaTox. (c) Wild type (B6) 4 h with FlaTox at indicated doses. (d) Wild type (B6) with FlaTox for indicated =mes. (e) Wild type (B6) pre treated 4 h with 0.5 µg/ml PAM3CSK4 then FlaTox for indicated =mes. Processed IL 1β released into the supernatant was quan=fied by enzyme linked immunosorbent assay (ELISA) as previously described 4. (f) Wild type (B6) mice injected intravenously (tail vein) with FlaTox and monitored for survival (n=14). Indicated doses are LFn FlaA. PA dose was 2x LFn FlaA. Data shown (± s.e.m.) are pooled from mul=ple experiments (f) or representa=ve of two (e) or three (a d) independent experiments. 1

2 RESEARCH SUPPLEMENTARY INFORMATION Supplementary Figure 2 PBS # 400x FlaTox $ 1200x Supplementary Figure 2. No histological changes detected in the intes6ne 30 minutes post FlaTox injec6on. (a d) Wild type (B6) mice were injected intraperitoneally with PBS (a b) or FlaTox (4 µg/g body weight PA + 2 µg/g body weight LFn FlaA) (c d). A]er 30 minutes, intes=nal =ssues (cecum + small intes=ne + large intes=ne) were collected and fixed by gentle injec=on of 10% buffered formalin. Tissues were stored in 10% buffered formalin before being embedded in paraffin, sec=oned, and stained with hematoxylin and eosin (University of Michigan Pathology Core). Sec=ons were analyzed by board cer=fied pathologist (I. Bergin) blinded to experimental groups. No abnormali=es were detected in any sec=ons. Duodenal sec=ons shown in figure are representa=ve of en=re intes=nal =ssue and experiment was repeated twice (n=1 2). a, c: x400, bar = 50µm; b, d: x W W W. N A T U R E. C O M / N A T U R E

3 SUPPLEMENTARY INFORMATION RESEARCH E HH #%& HH /,?(+0),@A.BA*(=.9:5; D C $ F*+.6$.30)+,)+.9G5; #% #$ '()*+,-../0-+..1)2,3+(0) %& #$ '()*+,-../0-+..1)2,3+(0) # %#& $ % 4*)5.6$.80)+,)+.9:5; %# %& <(=),>.6$.80)+,)+.9:5; %$ %# % #$ '()*+,-../0-+..1)2,3+(0) % #$ '()*+,-../0-+..1)2,3+(0) Supplementary Figure 3. FlaTox induced fluid loss in the peritoneal cavity and gut. (a d) Wild type (B6) mice were injected intraperitoneally with FlaTox (4 µg/g body weight PA + 2 µg/g body weight LFn FlaA) and analyzed at indicated =me. (a) Fluid in peritoneal cavity collected with syringe and quan=fied by weight. (bd) Fluid in intes=nal =ssues (b), lung (c) and kidneys (d) were determined by the difference between wet and dry organ weight (dried overnight at 3 C). Data shown are representa=ve of three independent experiments. ** p < (Mann Whitney t test) 3

4 RESEARCH SUPPLEMENTARY INFORMATION :+;2<=0=>? ( ' :+;2<=0=>? ( ' % # )*+,-'. #$% &'& ()*+, &'& )-+. &'& /0,'/0,1 &'& # $ % & ' / # $ % & ' / # % $ ( *4E64<>32<4+,FG. $( $' $% I4E>39<03+,:J-G. H ' & $# )*,-'. 02)'023'24 CDC % )*,-'. 02)'023'24 CDC Supplementary Figure 4. Time to morbidity and role of tumor necrosis factor. (a) Wild type (B6) mice were injected intravenously (tail vein) with FlaTox or FlaTox(AAA) (1.0 µg/g PA µg/g LFn FlaA or LFn FlaA(AAA)) and monitored 10 h for morbidity. Time to morbidity was defined as the earliest =me when mice failed to right themselves or had a rectal temperature < 25 C (n=5 6). (b) Mice of indicated genotypes were injected with FlaTox and monitored as in (a) (n=5 6). (c d) Mice were injected intraperitoneally with FlaTox (4 µg/g PA + 2 µg/g LFn FlaA) and rectal temperature (c) and hematocrit (d) were measured a]er 30 min (n=4 5). Data shown are representa=ve of one (d) or two (a c) independent experiments. 4

5 SUPPLEMENTARY INFORMATION RESEARCH /2 /% /1 F>?*G'A)*C#H6$E& 53 %3 43 /0 #$%& '()* +,-.+,- 13 #$%& '()* +,-.+,- # $ >?@>A*BA>C#DE& /2 /% /1 F>?*G'A)*C#H6$E& 53 %3 43 /0 #$%& 689: #$%& 689: 9.9 % & >?@>A*BA>C#DE& /2 /% /1 F>?*G'A)*C#H6$E& 53 %3 43 /0 ;$< =6: 13 ;$< =6: Supplementary Figure 5. Mast cells, lymphocytes, and neutrophils are not required for early hematopoie6c response to FlaTox. (a b) Wild type (B6) or mast cell deficient ckit wsh/wsh mice were injected intraperitoneally with FlaTox (4 µg/g body weight PA + 2 µg/g body weight LFn FlaA) and rectal temperature (a) and hematocrit (b) were measured 30 minutes post injec=on (n=5 6). (b c) Wild type (B6) or lymphocyte deficient Rag1 / mice were analyzed as above (n=5). (e f) Naïve or neutrophil depleted (a GR1) wild type (B6) mice were analyzed as above (n=5). Data shown are representa=ve of one (d; f) or at least two (a c; e) independent experiments. 5

6 RESEARCH SUPPLEMENTARY INFORMATION Supplementary Figure 6./0 F$3T 2KH #3$3T F3$T.?&PL&J(+; $T $./0 F$FT 2KH E5$5T /MIQ+Q;LM(-R&SMJ D5$T $DT 5$DT $DT 5$3T $66T >E1 >E1 Spleen Lamina Propria F $E5 F $D3 F E$ F $6 E E E E Diphtheria Toxin PBS./0./0 D 3 F E D 3 3 D E F >E1 $5E 3 D E F >E1./0./0 D 3 F E D 3 3 D E F./G $E 3 D E F./G./0./0 D 3 F E D 3 3 D E F >E1 E$6# 3 D E F >E1./0./0 D 3 F E D 3 $E5 3 D E F./G $5 $#5 3 D E F./G P Spleen Lamina Propria F $F F $#3 F E$ F E$F E E E E PBS./0 D $6./0 D $D5./0 D./0 D D$ D E F >E1 3 D E F./G 3 D E F >E1 3 D E F./G Clodronate./0 F E D 3 $D #$;1D./0 F E D 3 $ED $55./0 F E D 3 3$5./0 F E D 3 3$53 E$3 3 D E F >E1 3 D E F./G 3 D E F >E1 3 D E F./G # E E D6 D D5 D# HI?;;J K&J;-9(LL&@ DF D # 9MJ*+;N--2&N+--OJ%;G+M&J Supplementary Figure 6. Deple6on and transfer analysis. (ab) Resident peritoneal cells lavaged from macrophagedepleted wild type (B6) mice (a) or diphtheria toxin treated FVB Tg(CD11b DTR) mice (b) were labeled for CD11b and F4 80 and analyzed by flow cytometry. (c d) Splenic cells and isolated lamina propria lymphocytes were stained with CD11c, CD11b, and F4 80 an=bodies and analyzed by flow cytometry. (c) CD11b + cell depleted FVB Tg(CD11b/EGFP)34Lan/J (CD11b DTR) mice. (d) Macrophage depleted wild type (B6) mice. (d) Nlrc4 / mice were injected intravenously with bone marrow (2 femurs + 2 =bias) or spleen (en=re spleen) cells from a wildtype (B6) mouse. A]er 30 min, mice were injected intraperitoneally with FlaTox and rectal temperature was monitored over =me (n=2). Data shown (± s.e.m.) are representa=ve of at least two independent experiments. 6

7 SUPPLEMENTARY INFORMATION RESEARCH #$%&'()*+,)*+-.)./, #$%#&'%($)*+,*-)./.*01+,2 0%(&%&1*./+'.1(&1./ 3456*-0#8$9'9:9';)2 3456,*-'8<:0'=&)2,>?6145?6145 2*%+#$+3&' 4%+,5&6-&'/.'(8 4%+,5&6-&'/.'(98 :8;<4929 :=;<4929 :8;<929 :=;<929?6145 =;<4929 =;<929 >.)+3.', Supplementary Figure. Major eicosanoid pathways in mouse macrophages. Eicosanoid biosynthesis is ini=ated by release of arachidonic acid from cell membranes by phospholipases, such as cpla2. Free arachidonic acid is processed into thromboxanes and prostaglandins by the cyclooxygenase (COX, Ptgs) enzymes; into 12 HETE (12 hydroxyeicosatetraenoic acid) and 15 HETE by 12/15 LIPOXYGENASE (12/15 LOX, Alox15); and into 5 HETE and leukotrienes (e.g. LTB4) by 5 LIPOXYGENASE (5 LOX, Alox5). Lipoxins are synthesized from 15 HPETE (15 hydroperoxyeicosatetraenoic acid) by 5 LOX. Unlike the mixed 12/15 LOX ac=vity of the mouse Alox15, human macrophages, depending on their phenotype, express two dis=nct 15 LOX enzymes (ALOX15, ALOX15B) that primarily generate 15 HPETE.

8 RESEARCH SUPPLEMENTARY INFORMATION ' # 8 & % $ KK V).(0,1(&4 W=W,-+./ W=W DN T>:)4U 012&3 W=W 1,23456*,/,-8-./<=+>0 & %# % % $ K T>:)4U T>:)4U.QQQ0 DQ 1T/WT>:Q 1DN DN 1,23456*,/,-8-./<=+>0 # 8 & % $ KKK V).(0,1(&4 W=W DN T>:)4U T>:)4U.QQQ0,-+./ W=W 012&3 W=W 1,23456*,/,-8-./<=+>0 %# % $# $ # KK V).(0 DN T>:)4U #$%&'()*+(*),-+./ W=W # $ % & 1,23456*,/,-8-./<=+>0 $8 $% $ L GENO PD1Q%-R/S $ L ( ( 8 8 % % K 1)8 DA64M54;*;?-@,>>-1A;*; 1,23456*,/,-8-./<=+>0 $( $% L 8 GENO QD)QWQE $ L ( 8 % KK 1)8 DA64M54;*;?-@,>>-1A;*; 1,23456*,/,-8-.9:6;0./<=+>0 & %# % $# $ # J B,;*C,/5-D,6*54/,:> & ( I $% $# )*+,-.+*/0 $ '# # %#?-@,>>-1A;*;-.>*/,0 1,23456*,/,-8-.9:6;0./<=+>0 & %# % $# $ # J 4/,-E:664F-G,6*H,C J J J J & ( I $% $# )*+,-.+*/0 '# # %#?-@,>>-1A;*;-.>*/,0 Supplementary Figure 8. Inflammasome dependent leukotriene B 4 and cysteinyl leukotriene produc6on. (a f) Resident peritoneal macrophages were selected ex vivo by adherence to plas=c. (a) Indicated genotypes treated 30 minutes with FlaTox (10 µg/ml PA + 5 µg/ml LFn FlaA) and cysteinyl leukotrienes quan=fied by enzyme immunoassay. (b d) LTB 4 quan=fied by enzyme immunoassay (EIA). (b) 30 minutes with indicated proteins (10 µg/ml PA, 5 µg/ml all others), or lipopolysaccharide (1 µg/ml). (c) Indicated genotypes treated as in (a). (d) Infected with S. typhimurium (MOI=5) or treated with FlaTox for 180 minutes. (e f) Pre treated 45 minutes with DMSO, cpla2 inhibitor (e: 0.2 µm pyrrophenone) or calcium chelator (f: 10 µm BAPTA AM 10), followed by FlaTox (10 µg/ml PA + 5 µg/ml LFn FlaA). Cell lysis (120 minutes) was quan=fied by lactate dehydrogenase (LDH) release and LTB 4 (30 minutes) by EIA. (g h) Resident peritoneal macrophage (g) or bone marrow derived macrophages (h), were treated with FlaTox (10 µg/ml PA + 5 µg/ml LFn FlaA). Cell lysis (LDH release) and LTB 4 (EIA) were quan=fied at the indicated =mepoints. Data shown (± s.e.m.) are representa=ve of two (d, g, h) or at least three (a, b, c, e, f) independent experiments. * p < 0.05; ** <.009; *** <.000 (Student t test). # = not detected. 8

9 9 SUPPLEMENTARY INFORMATION RESEARCH #$%&'()&$*$&'*+),-./ : 8; 8< 8= 8= 98 98= 9 : ; < 9 : ; < = 9 : ; < = > 8= 98 98= :8 :8= ;8 ;8= 5%(.?(6#)@*+),-./0 1(AB#,/#)'&)*C:*+),-./0 1(AB#,/#)'&)*D:*+),-./0 9:EFD5D*+),-./0 9=EFD5D*+),-./0 =EFD5D*+),-./0 G@HI(B&@)@*2<*+),-./0 ; > J 9: 9= 9K 9 : ; < = > a L LLL M838 LLL L LLL LLL LLL Supplementary Figure 9. FlaTox induced eicosanoid biosynthesis in vivo. Wild type (B6) mice were injected intraperitoneally with FlaTox or FlaTox (AAA) (4 µg/g PA + 2 µg/g LFn FlaA or LFn FlaA(AAA)) for 20 min. Peritoneal lavage was analyzed for eicosanoids by LC/MS/MS. Data shown are pooled from mul=ple experiments. N.S. = not significant; * p < 0.04; ** p < *** p < (Mann Whitney t test).

10 RESEARCH SUPPLEMENTARY INFORMATION % # % # Relative Fluo-4 Intensity ( ' % # # $ % & Time (min) Relative Fluo-4 Intensity ( ' % # # $ % & Time (min) Supplementary Figure 10. Ionomycin-induced calcium flux. (a-b) Wild-type (B6) (a) or Casp1 -/- resident peritoneal macrophages loaded with Fluo-4 were imaged for 5 min and ionomycin (1 µm) was added after 1-2 min. Each trace represents one cell over time. Data shown are representative of >30 cells analyzed in two independent experiments. 10

11 SUPPLEMENTARY INFORMATION RESEARCH % % Cell Lysis $# # # Prostaglandin E2 (ng/ml) % $&# #& &# - + glycine - + glycine # $ % Cell Lysis % $# # # Prostaglandin E2 (ng/ml) % $&# #& &# FlaTox Digitonin H2O2 FlaTox Digitonin H2O2 Supplementary Figure 11. Separa6on of cell lysis and prostaglandin E 2 biosynthesis. (a d) Resident peritoneal macrophages selected ex vivo by adherence to plas=c. (a b) Treated 60 min with FlaTox (10 µg/ml PA + 5 µg/ml LFn FlaA) with or without glycine (5 mm). (c d) Treated 60 min with FlaTox (10 µg/ml PA + 5 µg/ml LFn FlaA), digitonin (10 µg/ml), or H 2 O 2 (1%). Cell lysis (a, c) was quan=fied by lactate dehydrogenase release and PGE 2 biosynthesis (b, d) by enzyme immunoassay. Data shown (± s.e.m.) are representa=ve of three independent experiments. 11

12 RESEARCH SUPPLEMENTARY INFORMATION #$%&'()$'*&+,(-+.*)/+0( (,-.*-/0& 10& 5,6(& #$%& /234& F01G& 31,08& ' 8H& '($()*+& & & ' 8H& 0239#:;.#3&03#:& ' 8H& ' 8H& ' 8H& ' 8H& PP ( '<=&,<=& 12;(>?/.:#.(@&/)AB;>2#).)(@&>92;CD;E.)&& IJK6LMN,05LI(4'56(,'(89:8;8<( =9;>584?5;@(8A5BCD( E48>F34B(GF;@(3588( 6LMN,05LI&19C9BH;?H(;?I9>J5?<( E48>F34B(:9BC94K;3;L4J5?( ;CCF?9(>933(B9>BF;AC9?A( :B5A9>JE9(B98:5?89( Supplementary Figure 12. Model of inflammasome dependent eicosanoid biosynthesis. LFn FlaA and protec=ve an=gen (PA) are taken up into endosomes, where the PA channel inserts into the membrane and translocates LFn FlaA to the cytosol. Once in the cytosol, LFn FlaA is detected by the NAIP5/NLRC4 inflammasome leading to recruitment and ac=va=on of CASP1. Ac=vated CASP1 targets as yet uniden=fied substrates to cause an influx of cytosolic calcium (Ca 2+ ). The rise in Ca 2+ ac=vates cytosolic phospholipase A2 (cpla2), which processes membrane phospholipids to release arachidonic acid (AA). cpla2 ac=vity is further enhanced by phosphoryla=on downstream of TLR signaling. Cyclooxygenase (COX) and lipoxygenase (LOX) enzymes then process AA into the eicosanoids prostaglandins, thromboxanes, and leukotrienes. Outside the cell, these eicosanoids have potent inflammatory effects. When ac=vated systemically, this pathway leads to an eicosanoid storm with devasta=ng effects that include vascular collapse and possible death in mice. On the other hand, we predict that localized ac=va=on of eicosanoid biosynthesis in infected cells protects the host by opening the site of infec=on and recrui=ng an appropriate immune response. This inflammasome dependent induc=on of eicosanoid biosynthesis is absent in bone marrow derived macrophages, but may be a general phenotype of resident macrophages, such as those in the peritoneum, that express high levels of the COX and LOX enzymes. 12

13 SUPPLEMENTARY INFORMATION RESEARCH a J(>K,+./(L(0401LM<)G6 BA& BA# BAC A8 A$ D4E FG*:,H 9:14$6 &'())*+#,-. ;<; /0 ;<; b :()=(.*+>.(01?56 8 & % $ # &'())*+#,-. #$% ;<; c '()*+,-./+0456 & % $ 9:14$6 &'())*+#,-. #$% ;<; Supplementary Figure 13. MYD88/TRIF dependent signaling enhances FlaTox pathology. (a) Resident peritoneal macrophages of indicated genotypes were selected ex vivo by adherence to plas=c and treated with FlaTox (10 µg/ml PA + 5 µg/ml LFn FlaA). LTB 4 was quan=fied by enzyme immunoassay a]er 30 minutes. (b c) Mice of indicated genotypes were injected intraperitoneally with FlaTox (4 µg/g PA + 2 µg/g LFn FlaA) and rectal temperature (b) and hematocrit (c) were measured a]er 30 minutes (n=5 ). Data shown (± s.e.m.) are representa=ve of at least two independent experiments. * p < 0.04; ** p < *** p < (a: Student t test; b c: Mann Whitney t test). 13

14 RESEARCH SUPPLEMENTARY INFORMATION & % '()*+,+,-./.+* $ # ),-0& 121 ),-0& 120 ),-0& 020 Supplementary Figure 14. Pyroptosis is COX 1 independent. BMDM from B6;129P2 Cox1 / mice and liwermate controls were incubated 4 h with FlaTox (4 µg/ml PA + 2 µg/ml LFn FlaA). Cell lysis was quan=fied by release of lactate dehydrogenase. Data shown (± s.e.m.) are representa=ve of three independent experiments. 14

15 SUPPLEMENTARY INFORMATION RESEARCH 8<AB<;=C;<1DE-F )# $ ) ' * * * * * * * *;<8;<= 23><?= *+, 5689: -./&0'4 5689: -./&4 5689: -./'4 5689: *+, * G<A=9?;H=1DIJ+-F $( $# (( (# # # # $# %# &'# &(# LH3C=<M1B9M=1H3><?=H93 )( *;<8;<=K 23><?=K *+, -./&0'4 -./&4 -./'4 *+, *+, 5689: # $ )# %# 8<AB<;=C;<1DE-F N $ ) G<A=9?;H=1DIJ+-F $# ' O8D+$F #$% 0 (# O8D+$F #$% 0 Supplementary Figure 15. COX 1 inhibi6on but not COX 2 inhibi6on or LOX 5 dele6on protects mice from FlaTox. (a b) Wild type (B6) mice were pretreated intraperitoneally for 30 minutes with 1 µg/g body weight of COX 1/2 non specific (indomethacin), COX 1 specific (SC 560) or COX 2 specific (Cay 10404) inhibitors. 5 µg of FlaTox (5 µg PA + 5 µg LFn FlaA) was delivered in 200 µl PBS by tail vein injec=on. (a) Rectal temperature was monitored over =me. (b) Hematocrit was measured 60 minutes post injec=on. (c d) Mice of indicated genotypes were injected intraperitoneally with FlaTox (4 µg/g PA + 2 µg/g LFn FlaA) and temperature (c) and hematocrit (d) were measured 30 m.p.i. Data shown (± s.e.m.) are representa=ve of at least two independent experiments. * p = (Mann Whitney t test). 15