Immunity Article

Gain-of-Function Pyrin Induce NLRP3 -Independent Interleukin-1b Activation and Severe Autoinflammation in Mice

Jae Jin Chae,1,5,* Young-Hun Cho,2,5 Geun-Shik Lee,1 Jun Cheng,3 P. Paul Liu,3 Lionel Feigenbaum,4 Stephen I. Katz,2 and Daniel L. Kastner1,* 1Medical Genetics Branch, National Genome Research Institute (NHGRI) 2Dermatology Branch, Center for Cancer Research, National Cancer Institute (NCI) 3Genetics and Molecular Biology Branch, NHGRI National Institutes of Health, Bethesda, MD 20892, USA 4Laboratory Animal Sciences Program, Science Applications International Corporation-Frederick, NCI-Frederick Cancer Research and Development Center, Frederick, MD 21702, USA 5These authors contributed equally to this work *Correspondence: [email protected] (J.J.C.), [email protected] (D.L.K.) DOI 10.1016/j.immuni.2011.02.020

SUMMARY phonuclear leucocytes into the affected tissues, neutrophilia, and a rapid acute phase response. The responsible for Missense mutations in the C-terminal B30.2 domain FMF, designated MEFV, is composed of 10 exons encoding of pyrin cause familial Mediterranean fever (FMF), a 781 protein known as pyrin (also known as mare- the most common Mendelian autoinflammatory nostrin) (French FMF Consortium, 1997; The International FMF disease. However, it remains controversial as to Consortium, 1997). So far more than 50 FMF-associated muta- whether FMF is due to the loss of an inhibitor of tions have been identified within MEFV (Touitou et al., 2004). inflammation or to the activity of a proinflammatory The carrier frequencies for FMF in the Middle East are as high as 1:3, suggesting that pyrin mutations may confer a selective molecule. We generated both pyrin-deficient mice advantage, possibly related to their effects on innate immunity. and ‘‘knockin’’ mice harboring mutant human B30.2 Pyrin is expressed predominantly in innate immune cells such domains. Homozygous knockin, but not pyrin-defi- as , , and dendritic cells, but not in lym- cient, mice exhibited spontaneous bone marrow- phocytes. In monocytes, the expression is variable and upregu- dependent inflammation similar to but more severe lated by proinflammatory cytokines, interferon-g (IFN-g) or tumor than human FMF. Caspase-1 was constitutively acti- necrosis factor-a (TNF-a)(Centola et al., 2000; Diaz et al., 2004). vated in knockin macrophages and active IL-1b was Pyrin is composed of at least five domains and each domain has secreted when stimulated with lipopolysaccharide a distinct role in interactions with several that are related alone, which is also observed in FMF patients. The to inflammation (Chae et al., 2009). The most notable domain of inflammatory phenotype of knockin mice was com- pyrin is the N-terminal PYRIN domain (also called PYD, PAAD, or pletely ablated by crossing with IL-1 receptor-defi- DAPIN). The PYRIN domain is also found in a large number of proteins implicated in the control of inflammation. PYRIN cient or adaptor molecule ASC-deficient mice, but domain-containing proteins play a key role in the inflamma- not NLRP3-deficient mice. Thus, our data provide somes, cytoplasmic multiprotein complexes mediating the evidence for an ASC-dependent NLRP3-indepen- proteolytic activation of caspase-1, which are essential for the dent inflammasome in which gain-of-function pyrin production of biologically active interleukin-1b (IL-1b). mutations cause autoinflammatory disease. We have reported a mouse model generated by targeted trun- cation of pyrin, which demonstrated that pyrin plays a role in the regulation of the inflammasome, mediated through PYRIN INTRODUCTION domain interactions (Chae et al., 2003). Nevertheless, the pyrin truncation mice did not show an overt phenotype related to Familial Mediterranean fever (FMF; MIM 249100) is the prototype FMF, suggesting that FMF may be caused by a gain of function and the most common of a class of disorders, termed Mendelian induced by disease-associated missense changes in pyrin. autoinflammatory diseases, that are characterized by recurring Indeed, no null FMF mutations have been identified to date. spontaneous episodes of fever and localized inflammation Most FMF-associated mutations are clustered in the C-terminal without high-titer autoantibodies or antigen-specific T cells (Ga- B30.2 domain of pyrin, and therefore this domain has been the lon et al., 2000; Kastner et al., 2010; Masters et al., 2009). There focus of current thinking on the pathogenesis of FMF. The is a growing body of data supporting the view that these condi- B30.2 domain has a role in attenuating the activation of IL-1b tions represent genetic lesions of the innate immune system. through direct interaction with caspase-1 as well as other inflam- FMF inflammation is mediated by a massive influx of polymor- masome components, such as the NACHT, LRR, and pyrin

Immunity 34, 755–768, May 27, 2011 ª2011 Elsevier Inc. 755 Immunity Molecular Pathogenesis of FMF

domain-containing proteins (NLRP or NALP) 1, 2, and 3 (Chae (Table S1). Although MefvM680I/M680I and MefvM694V/M694V mice et al., 2006; Papin et al., 2007). Possibly because of technical could not be distinguished from their WT or heterozygous differences, transfection studies have yielded contradictory littermates for several weeks after birth, milder but detectable results regarding the roles of wild-type and mutant pyrin in IL-1 illness developed after 4 to 8 weeks of age for MefvM680I/M680I activation and inflammation more broadly (Chae et al., 2006, mice (Figure 1D) and after more than 20 weeks after birth 2008; Papin et al., 2007; Yu et al., 2006, 2007). for MefvM694V/M694V mice, with growth retardation, wasting, In order to study the pathogenesis of FMF induced by muta- dermatitis, and arthritis, though less severe than that seen with tions in the B30.2 domain in vivo, we have generated various homozygous MefvV726A/V726A mice. Thus, the insertion of gene insertion ‘‘knockin’’ (KI) mouse models with three frequent FMF-associated mutant B30.2 domains into mouse pyrin FMF-associated B30.2 mutations (M680I, M694V, and V726A) induces spontaneous inflammation that is similar to or more by inserting the B30.2 domain of human pyrin into mouse pyrin, severe than the inflammation observed in FMF patients. Severity, which lacks a B30.2 orthologous domain. In contrast with the as assessed by the age of onset, varies inversely with the severity aforementioned truncation model, the KI mice showed severe of human phenotypes (MefvV726A/V726A > MefvM680I/M680I > spontaneous inflammatory phenotypes comparable to the MefvM694V/M694V). inflammation of FMF patients. With these KI mice, we explored the pathogenesis of FMF not only at the physiological level but The KI Mice Have Increased CD11b+ Cells also at the molecular level and found that the inflammatory in Lymph Node, Spleen, and Blood phenotype of these mice is mediated by the ASC-dependent, The homozygous KI mice also showed lymphadenopathy and NLRP3-independent production of IL-1b by bone marrow- splenomegaly (Figure S1D). Histological examination of spleens derived cells. from MefvV726A/V726A mice revealed the loss of normal germinal center architecture and the infiltration of inflammatory cells (Fig- RESULTS ure 1E). Indeed, flow cytometric analysis of MefvV726A/V726A sple- nocytes revealed that CD11b+ cells were markedly expanded Insertion of FMF-Associated Mutant B30.2 Domains whereas T and B cells were substantially decreased (Figure 2A). Induces Inflammation in Mice The expansion in numbers of CD11b+ cells was also observed in Although exon 10 of mouse Mefv has substantial homologies at lymph node (LN) (Figure S2A), BM, and peripheral blood as early the nucleotide level with exon 10 of human MEFV, which as 4 weeks and increasing with age. Subsequent analysis encodes the B30.2 domain, the mouse pyrin protein does not showed that the vast majority of the increased CD11b+ cells have the B30.2 domain because of frame-shift variations (Chae were Ly-6G+ neutrophils (granulocytes) (Figure 2A; Figure S2A), et al., 2000). Thus, to insert FMF-associated B30.2 mutations and the absolute number of F4/80+ cells (monocytes or macro- into mouse pyrin, exons 7–10 of mouse Mefv were replaced phages) was also increased. In MefvM680I/M680I (Figure S2B) with exons 7–10 of human MEFV encoding wild-type (WT) or and MefvM694V/M694V (data not shown) mice, the numbers of FMF-associated mutant B30.2 domains. The resultant fusion CD11b+ cells were also substantially increased, although not proteins were expected to express WT or one of three mutant to the degree seen in MefvV726A/V726A mice. human B30.2 domains (M680I, M694V, or V726A) (Figure 1A; Figures S1A and S1B available online). Heterozygotes of all The Inflammation of KI Mice Is Induced by a Gain mutant KI mice (MefvM680I/+, MefvM694V/+, and MefvV726A/+) of Function in the Mutant B30.2 Domain were born from chimeric males and appeared normal without with Dosage Effect evidence of pathology. However, heterozygous WT B30.2 Because, in the homozygous KI mice, the disease was induced domain KI mice (MefvB30.2/+) were not born. A total of 44 chimeric by the modified pyrin with insertion (KI) of the FMF-associated WT B30.2 males were produced from five different positive mutant B30.2 domain rather than disruption of WT pyrin, we embryonic stem (ES) cell clones, which had been screened by hypothesized that FMF-associated pyrin mutations might lead Southern blot and confirmed by long-range PCR. In spite of to a gain of function. However, heterozygotes showed normal high amounts of coat color chimerism, none of the chimeric growth without any inflammatory symptoms whereas homozy- males produced heterozygotes. Homozygous mutant KI mice gous KI mice were diseased, which is consistent with the reces- (MefvM680I/M680I, MefvM694V/M694V, and MefvV726A/V726A) were sive mode of inheritance for FMF and a loss-of-function model. born from interbred heterozygotes in Mendelian ratios. To help clarify this issue, pyrin-null (Mefv/) mice were gener- All homozygous mutant KI mice spontaneously developed ated (Figure S1E) and crossed with MefvV726A/+ mice to produce inflammatory phenotypes, with MefvV726A/V726A the most severe. hemizygotes (MefvV726A/) that expressed only mutant pyrin MefvV726A/V726A pups were runted with scales on the skin as early from a single allele. Mefv/ mice were grossly normal, and as 1 week of age. Some MefvV726A/V726A pups died around week 2 MefvV726A/ mice also showed normal features of growth and to 3, and their growth was severely retarded (Figure 1B; Fig- no signs of disease (Figures 2B and 2C). These data indicate ure S1C). Gross and histological examination of MefvV726A/V726A that missense mutations on both alleles are necessary for the mice revealed severe spontaneous inflammation in multiple induction of inflammation in KI mice, implying the importance tissues such as skin, liver, joints, and bone marrow (BM) (Fig- of the amount of mutant pyrin in inducing the inflammation. ure 1C). As in human FMF, a substantial infiltration of leukocytes, Indeed, in macrophages, the LPS- and IL-4-induced expression mainly neutrophils and macrophages, was also observed in all of pyrin from both alleles (Mefv+/+) was about two times the pyrin inflamed tissues. Moreover, homozygous KI mice showed expression from only a single allele (Mefv+/)(Figure 2D, right), anemia, lymphopenia, and granulocytosis in the peripheral blood and the expression of mutant pyrin in homozygotes was

756 Immunity 34, 755–768, May 27, 2011 ª2011 Elsevier Inc. Immunity Molecular Pathogenesis of FMF

A Mouse pyrin Human pyrin

PYRIN B α-Helixa-Helix PYRIN B α-Helixa-Helix B30.2

PYRIN B α-Helixa-Helix B30.2

KI pyrin

BD V726A/V726A 25 25 Mefv M680I/M680I V726A/+ Mefv Mefv M680I/+ +/+ Mefv Mefv 20 20 Mefv+/+ p<0.001 15 15 Body weightBody (g)

10 weightBody (g) 10

5 5 1 2 3 4 5 6 7 1 2 3 4 5 6 7 8 9 10 11 12 13 Time (weeks) Time (weeks) C Mefv+/+ MefvV726A/V726A E +/+ Skin (ear) Mefv Liver V726A/V726A Mefv Ankle joint

Figure 1. Pyrin with an FMF-Associated Mutant B30.2 Domain Induces Inflammation in Mice (A) Comparison of the schematic structure of pyrin proteins of human, mouse, and KI mice in which mouse pyrin is fused with the human B30.2 domains. (B and D) Growth curves for (B) V726A mutant KI males (WT, n = 5; heterozygotes, n = 9; homozygotes, n = 4) and (D) M680I mutant KI females (WT, n = 9; heterozygotes, n = 10; homozygotes, n = 6). Data are as means ± SD. (C) Hematoxylin and eosin (H&E)-stained ear sections (2003) and liver sections (2003) from 8-week-old WT and MefvV726A/V726A mice, and H&E-stained section (1003) (bottom) of ankle joints of 4-week-old MefvV726A/V726A mice. (E) H&E-stained spleen sections (403) from 8-week-old WT and MefvV726A/V726A mice. Additional information is provided in Figure S1 and Table S1. substantially higher than heterozygotes both before (Figure 2D, BM-Derived Cells Are Necessary and Sufficient left) and after (Figure 2D, middle) stimulation of mouse pyrin for the Induction of FMF Inflammation expression with LPS and IL-4. Taken together, these results Because pyrin is expressed primarily in myeloid cells, and KI suggest that this mouse model of FMF can be explained by the mice showed the expansion in numbers of CD11b+ cells in gain-of-function and gene-dosage effect of mutant pyrin. LNs, spleen, and blood, we hypothesized that BM-derived cells

Immunity 34, 755–768, May 27, 2011 ª2011 Elsevier Inc. 757 Immunity Molecular Pathogenesis of FMF

A Week 4 Week 12 Mefv+/+ MefvV726A/+ MefvV726A/V726A Mefv+/+ MefvV726A/+ MefvV726A/V726A 4 4 4 104 104 104 10 10 10 18.8 0.61 23.7 0.63 13.4 0.26 18 0.84 18.7 0.62 10.4 0.14 3 3 3 103 103 103 10 10 10

2 2 2 102 102 102 10 10 10 CD3

1 1 1 101 101 101 10 10 10

25.6 55 20 55.7 63.9 22.5 0 10.4 70.7 0 12 68.7 0 75.1 14.3 100 100 100 10 10 10 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 CD19 1000 1000 1000 1000 1000 1000 10.2 10.8 50.4 4.01 4.58 60.8 800 800 800 800 800 800

C 600 600 600 600 600 600

SS 400 400 400 400 400 400

200 200 200 200 200 200

0 0 0 0 0 0 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 CD11b 4 4 4 104 104 104 10 10 10

3 3 3 3 3 3 10 78.3 10 82.7 10 25.1 10 76 10 64.2 10 23.3

16.3 9.25 66.1 2 13.4 2 26.2 2 56.9 102 102 102 10 10 10 F4/80

1 1 1 101 101 101 10 10 10

0 0 0 100 100 100 10 10 10 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 Ly-6G B MefvV726A/+ X MefvV726A/+ X Mefv-/-

Mefv+/+ MefvV726A/+ MefvV726A/V726A MefvV726A/- Mefv+/-

Healthy Healthy Sick Healthy Healthy Healthy 1000 1000 1000 1000 1000 1000 14.7 13.8 83.3 13.2 14.4 12.2 800 800 800 800 800 800

600 600 600 600 600 600

SSC 400 400 400 400 400 400

200 200 200 200 200 200

0 0 0 0 0 0 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 CD11b Peripheral blood, week 8

C 35 D Untreated LPS/IL-4 LPS/IL-4 p<0.001 30 25 20 +/+ - /- +/+ +/+ M680I/+ +/ - M680I/M680I M680I/+ 15 M680I/M680I 10 Mefv Mefv Mefv Mefv Mefv Mefv Mefv Mefv Mefv 5 Body weightBody (g) 0 Pyrin -/- +/+ V726A/- V726A/+

Mefv Actin Mefv V726A/V726A Mefv Mefv Mefv

758 Immunity 34, 755–768, May 27, 2011 ª2011 Elsevier Inc. Immunity Molecular Pathogenesis of FMF

might be critical for the pathogenesis of KI mice. When BM cells is mediated by soluble factors, such as cytokines, abnormally from MefvM680I/M680I mice were transferred to lethally irradiated secreted from mutant cells rather than a direct effect of mutant WT mice, the recipient mice started losing body weight and pyrin in the cells expressing it. developed dermatitis with spontaneous infiltration of inflamma- tory cells, whereas the control mice that had received WT BM The Disease of KI Mice Is Induced by IL-1b cells showed normal growth without an inflammatory phenotype In order to find the soluble factors that mediate inflammation in (Figures 3A and 3B). Moreover, the expansion in numbers of the the KI mice, various cytokines were measured from the sera of CD11b+ myeloid cells was also observed in the blood (Figure 3C), MefvV726A/V726A mice. We observed a significant elevation of LNs, and spleen of the recipient mice as seen in unmanipulated multiple cytokines, chemokines, and hematopoietic factors in MefvM680I/M680I mice, indicating that BM cells of KI mice trans- the sera of MefvV726A/V726A mice compared to the WT or hetero- ferred the KI phenotype into WT mice. Conversely, we also trans- zygous littermates (Figure 4A; Figure S3A). Of the elevated ferred congenic CD45.1 WT BM cells to CD45.2 MefvV726A/V726A cytokines, we primarily focused on IL-1b because it has been mice suffering from severe inflammation. Most of the recipients demonstrated that pyrin has a role in the regulation of IL-1b secre- showed normalized numbers of CD11b+ cells with gradual tion (Chae et al., 2003, 2006, 2008; Papin et al., 2007; Yu et al., body weight gain to the normal body weight range of WT adult 2006). The functionally active 17 kDa IL-1b is secreted mainly mice (Figures 3D and 3E). In contrast, we also observed that from innate immune cells after proteolytic cleavage of 34 kDa a few recipients failed to gain body weight and still had expanded pro-IL-1b by activated caspase-1. For the activation of cas- CD11b+ cells in their blood. Subsequent cytometric analyses pase-1, the 45 kDa pro-caspase-1 must be cleaved into 22 and showed that the rescued mice were well reconstituted with WT 10 kDa catalytic subunits (p20 and p10, respectively). Therefore, donor cells as expected, but there was no reconstitution of donor the activation and secretion of both IL-1b and caspase-1 was cells in the sick recipients (Figure 3F). Because of a tendency of examined by immunoblot from the cell culture supernatants of early death and small body size, we had given WT BM cells to 3- the BM CD11b+ cells and BM-derived macrophages (BMDMs). week-old MefvV726A/V726A mice intraperitoneally rather than by WT and MefvV726A/+ cells secreted IL-1b and p10 only after stim- intravenous injection and had irradiated the recipients with two ulation with LPS followed by ATP treatment, whereas the cells doses of 3.5 Gy. As evidenced by the flow cytometric data, stimulated with LPS alone did not secrete IL-1b and p10 despite this regimen did not result in complete chimerism in all recipi- high pro-IL-1b expression. In contrast, IL-1b and p10 were highly ents. However, the data indicate a strong correlation between secreted from BM CD11b+ cells and BMDMs of MefvV726A/V726A establishment of chimerism and phenotypic rescue. mice by LPS stimulation alone without ATP treatment (Figure 4B; Figure S3B). Moreover, we could also observe the secretion of The Expansion of Myeloid Cell Numbers Is Induced p10 from the untreated MefvV726A/V726A BM CD11b+ cells, indi- by Soluble Factors cating that the inflammasome is constitutively activated in the Next we investigated whether the expansion of CD11b+ cell myeloid cells of homozygous KI mice. numbers in KI mice was due to the direct involvement of the To further investigate the role of IL-1b in the pathogenesis of KI FMF-associated mutant B30.2 domain in the proliferation of the mice in vivo, we crossed MefvV726A/V726A mice with IL-1 receptor- inflammatory cells or initiated by soluble factors secreted from deficient (Il1r1/) mice to generate MefvV726A/V726AIl1r1/ the cells expressing mutant pyrin. We used mixed BM transplants mice. The MefvV726A/V726AIl1r1 / mice were grossly normal in which CD45.1+ WT BM cells were premixed with CD45.2+ without any inflammatory phenotypes such as myeloid cell MefvM680I/M680I or CD45.2+ WT BM cells and transferred into number expansion (Figures 4C and 4D). However, the inflamma- lethally irradiated CD45.1+CD45.2+ heterozygous WT recipients. some of the BM CD11b+ cells was still activated and IL-1b was We observed that within both recipients of CD45.1+ WT-CD45.2+ secreted from the cells with LPS alone (Figure 4E), suggesting WT BM cells and of CD45.1+ WT-CD45.2+MefvM680I/M680I BM that constitutive inflammasome activation may be a primary cells, each donor accounted for between 40% and 50% of total event in FMF KI mice. Taken together, these results demonstrate blood cells at 6 weeks after BM transplant. Subsequent analysis that IL-1b is a key soluble factor for inducing the inflammation in of blood cell lineages differentiated from each donor showed that these mice and that IL-1b blockade ameliorates the disease. CD11b+ cells comprised more than 50% of both CD45.1+ and CD45.2+ cells in the recipients of mixed CD45.1+ WT-CD45.2+ The Inflammasome Is Highly Activated MefvM680I/M680I BM cells, whereas CD11b+ cells comprised only in the Macrophages of FMF Patients about 20% of each lineage in mixed WT chimeras (Figure 3G). Consistent with several case reports demonstrating the marked These data suggest that the expansion of myeloid cell numbers amelioration of symptoms in FMF patients upon treatment with

Figure 2. Expansion of CD11b+ Cells and Gain-of-Function B30.2 Mutations with a Gene Dosage Effect (A) Flow cytometry analyses of spleens from 4- and 12-week-old MefvV726A/V726A mice. Total splenocytes were analyzed for the relative frequencies of lymphoid (CD3 and CD19 for T cells and B cells, respectively, first row) and myeloid (CD11b, second row) cells. Within CD11b+-gated myeloid cells, Ly-6G+ neutrophils and F4/80+ monocytes or macrophages were analyzed (third row). Numbers indicate percentage of total cells in respective quadrants or gates. Data are repre- sentative of three independent experiments. (B) Generation of hemizygous (MefvV726A/) KI mice for the mutant B30.2 domain. Peripheral blood cells were analyzed for CD11b+ cells from their 8-week-old offspring. Data are representative of three independent experiments. (C) Body weights of six 8-week-old males of Mefv+/+, MefvV726A/+, MefvV726A/V726A, MefvV726A/, and Mefv/. (D) Expression of pyrin in peritoneal macrophages from M680I-KI and pyrin-deficient mice after 24 hr of culture. Additional information is provided in Figure S2.

Immunity 34, 755–768, May 27, 2011 ª2011 Elsevier Inc. 759 Immunity Molecular Pathogenesis of FMF

AB CMefv+/+BM Mefv M680I/M680I BM +/+ +/+

+/+ into Mefv into Mefv 1000 1000 BM M680I/M680I 9.78 65 140 800 800 Mefv donor +/+ Mefv Mefv+/+ donor 600 600 SSC Mefv 120 into 400 400

200 200

0 0

100 BM 100 101 102 103 104 100 101 102 103 104

+/+ CD11b

4 4 80 10 10 Mefv

Relative body weight (%) body weight Relative 0 3 6 12 3 103 10 M680I/M680I

into 37.8 13 Time (weeks) 2 83 102 54 10 F4/80 Mefv 1 101 10

0 100 10 0 1 2 3 4 100 101 102 103 104 10 10 10 10 10 Ly-6G D Mefv+/+ into MefvV726A/V726A V726A/V726A Mefv E into MefvV726A/V726A Healthy-rescued Sick-no chimerism 1000 1000 1000 16.4 91.6 77.6 800 800 800 250 600 600 600

400 400

SSC 400 200 200 200 200

0 0 0 0 1 2 3 4 0 1 2 3 4 100 101 102 103 104 10 10 10 10 10 10 10 10 10 10 150 CD11b

4 4 104 10 10 100 +/+ V726A/V726A 3 3 Mefv into Mefv Healthy-rescued 103 10 10 37.4 2.86 9.59 Mefv+/+ into MefvV726A/V726A Sick-no chimerism 2 2 72.7 2 78.8 10 54.3 10 10 50 V726A/V726A V726A/V726A Relative body weight (%) body weight Relative Mefv into Mefv F4/80 1 1 101 10 10 0 4 8 12 16

0 0 100 10 10 0 1 2 3 4 0 1 2 3 4 Time (weeks) 100 101 102 103 104 10 10 10 10 10 10 10 10 10 10 Ly-6G

G CD11b CD3 CD19 19 16.5 64.5 F 104 Mefv+/+ Mefv+/+ into MefvV726A/V726A 103 CD45.2

102 52.1 Healthy-rescued 4 Sick-no chimerism 104 10 50 31.6 0.052 0.28 19.6 12.5 67.8 CD45.2 1 3 10 103 10 Mefv+/+ 40.2 2 0 102 10 10 CD45.1 100 101 102 103 104

1 6 Week 101 10 CD45.1

18 0.41 0 18.1 81.6 100 10 0 1 2 3 4 100 101 102 103 104 10 10 10 10 10 65.5 8.83 26 4 4 104 10 10

CD45.1 72.5 15.7 0.031 3.33 MefvM680I/M680I 3 3 10 10 103 CD45.2

2 102 10 102 49.5 1 101 10

Week 10 Week 51.9 10.1 37.4

CD45.2 1 88.9 10 10.1 1.65 0 7.7 +/+ 100 10 Mefv 0 1 2 3 4 100 101 102 103 104 10 10 10 10 10 41.1 100 CD45.1 CD11b 100 101 102 103 104 CD45.1

760 Immunity 34, 755–768, May 27, 2011 ª2011 Elsevier Inc. Immunity Molecular Pathogenesis of FMF

recombinant IL-1 receptor antagonist (Belkhir et al., 2007; lineage is probably critical for the pathogenesis of FMF. However, Bilginer et al., 2009; Calligaris et al., 2008; Chae et al., 2006; Kuijk in KI mice, we could also observe an increased fraction of acti- et al., 2007; Moser et al., 2009; Roldan et al., 2008), the aforemen- vated memory T cells (Figure S5), though the overall numbers tioned mouse data suggest that the inflammatory symptoms of of lymphocytes were relatively low in the LN and spleen of homo- human FMF may also be triggered by IL-1b. However, there zygous KI mice. Moreover, the cytokines IL-2, IL-9, IL-17, and has been no direct evidence showing the constitutive activation IFN-g, which are known to be primarily produced by T cells, of the inflammasome and subsequent high IL-1b production were also significantly increased in the sera of KI mice (Fig- from myeloid cells of FMF patients. Therefore, we first examined ure S3A). Therefore, to test whether T cells have a role in the path- IL-1b secretion from peripheral blood mononuclear cells ogenesis of the KI phenotype, we produced MefvV726A/V726A mice (PBMCs) of FMF patients who had mutations in the B30.2 domain on the RAG1-deficient background (MefvV726A/V726ARag1/). of pyrin. Similar to the FMF KI mice, we could observe that, MefvV726A/V726ARag1/ mice were runted with inflammatory without ATP treatment, various Toll-like receptor (TLR) ligands phenotypes similar to MefvV726A/V726ARag1+/+ mice and showed could induce IL-1b secretion from PBMCs of most of the FMF constitutive inflammasome activation in their myeloid cells com- patients tested (Figures S4A–S4C). However, some of the healthy parable to what was observed in cells from MefvV726A/V726A controls also secreted IL-1b in response to LPS alone. Moreover, Rag1+/+ mice (Figures 6A–6D). Nevertheless, activated T cells PBMCs from a few FMF patients secreted less IL-1b than may still amplify the knockin phenotype, because several cyto- a healthy control, though most FMF patients secreted higher kines were significantly decreased in MefvV726A/V726ARag1/ amounts of IL-1b than healthy controls (Figure S4D). This incon- mice as compared to MefvV726A/V726ARag1+/+ mice, though the sistency may be due to differences in pyrin expression that may amounts were still higher than in WT mice (Figure S3). in turn result from genetic or environmental factors, including treatment (Booty et al., 2009; Chae et al., 2008). Inflammation in FMF KI Mice Is ASC Dependent In order to minimize the variation coming from treatments and but NLRP3 Independent environmental factors, the PBMCs were first differentiated into Finally, we investigated the underlying mechanism for the inflam- macrophages in vitro, and IL-1b secretion was then measured masome activation of the FMF KI mice in vivo. Because (Figures 5A and 5B). In these experimental conditions, marked apoptosis-associated speck-like protein with a CARD (ASC) is IL-1b and p20 secretion were observed in FMF patients but not the essential adaptor molecule among the components of healthy controls when the macrophages were stimulated with most inflammasomes, and pyrin itself can interact with ASC LPS followed by ATP treatment, whereas they were not secreted through cognate PYRIN domain interactions, we crossed our or secreted at low amounts by LPS alone. We also examined MefvV726A/V726A mice with ASC-deficient mice (Asc/). inflammasome activation from the macrophages in which pyrin MefvV726A/V726AAsc/ mice showed normal growth without expression was highly induced by IFN-g treatment for 16 hr prior any inflammatory phenotype or myeloid cell expansion (Figures to LPS stimulation. The increased expression of pyrin was veri- 7A and 7B). We could also observe the complete ablation of fied in the IFN-g-treated macrophages from both patients and IL-1b secretion and caspase-1 activation from the BM CD11b+ healthy controls, but, similar to the inflammasome activation in cells of MefvV726A/V726AAsc/ mice after LPS stimulation, even the FMF KI mice, IL-1b and p20 were secreted only from patients’ with ATP treatment, as seen with the cells from Asc/ mice macrophages by LPS stimulation alone without ATP treatment. (Figure 7C), indicating that ASC is critical for the caspase-1 These results suggest that mutant pyrin is involved in inflamma- activation mediated by FMF-associated pyrin mutations. some activation to a much greater degree than WT pyrin. Given the association of ASC with NLRP3 in the NLRP3 inflam- masome, we asked whether the deficiency of NLRP3 would The Adaptive Immune System Has No Major Role affect the FMF KI phenotype. As a control, we observed that in the KI Phenotype the CD11b+ cells from NLRP3-deficient (Nlrp3/) mice did not Figures 2 and 3 indicate that the inflammatory phenotype of KI secrete IL-1b in response to the LPS and ATP stimulation mice is initiated by hematopoietic stem cells and that the myeloid in vitro (Figure 7D). However, the NLRP3-deficient KI mice

Figure 3. The Inflammatory Phenotypes of KI Mice Are Induced from Cells of the Hematopoietic Lineage (A–C) BM cells from KI mice induce inflammatory phenotypes in WT mice. WT mice were lethally irradiated and given BM cells from either MefvM680I/M680I or WT mice intravenously. (A) Body weight after bone marrow transplantation (BMT). Data are shown in percent relative body weight changes as means ± SD of duplicate measurements from four of each recipient. (B) H&E-stained ear sections (1003) from both recipients (WT BM cells into WT and MefvM680I/M680I BM cells into WT) after 6 weeks of BMT. (C) Flow cytometry analyses of peripheral blood cells from both recipients after 18 weeks of BMT. Data are representative of three independent experiments. (D–F) BM cells from congenic CD45.1 WT mice cure the inflammatory phenotypes of CD45.2 KI mice. MefvV726A/V726A mice were irradiated and given BM cells from either WT or MefvV726A/V726A mice intraperitoneally. (D) Flow cytometry analyses of peripheral blood cells after 10 weeks of BMT. (E) Body weight after BMT. Data are shown in percent relative body weight changes of individual recipients as indicated. (F) The degree of chimerism measured concomitantly by staining the congenic marker of donor cells (CD45.1). (G) The expansion of CD11b+ cell numbers is mediated by soluble factors. Congenic CD45.1+ WT BM cells were premixed with either CD45.2+ WT BM cells (top) or CD45.2+ MefvM680I/M680I BM cells (bottom) and injected into lethally irradiated CD45.1+ and CD45.2+ WT mice. Flow cytometry analyses of peripheral blood after 6 weeks of BMT for the chimerism (dot plot) and relative CD11b+, CD3+, and CD19+ cells according to the origin of each cell population (histogram). Data are representative of three independent experiments.

Immunity 34, 755–768, May 27, 2011 ª2011 Elsevier Inc. 761 Immunity Molecular Pathogenesis of FMF

AB C 600 ** BM CD11b+ 35 p<0.00001 400 +/+ V726A/+ V726A/V726A Mefv Mefv Mefv 30 (pg/ml)

β *** 200 25

IL-1 20 0 Untreat LPS ATP LPS + Untreat Untreat LPS ATP LPS + LPS ** ATP LPS + 15 3000 Sup. β Pro-IL1 10 2000 (g) weight Body 5 * IL1β 1000 0 - -/- -/ IL-6 (pg/ml) IL-6 +/+ +/+ 0 Il1r1 Il1r1 Il1r1 3000 ** Casp1 Il1r1 +/+ p10 +/+ 2000 Mefv Mefv (pg/ml) V726A/V726A V726A/V726A α ** Lys. Pro-IL1β V726A/V726A 1000 Mefv Actin Mefv TNF- 0 - /- -/ - +/+ +/+ +/+ Rag1 Rag1 Rag1 Rag1 Rag1 +/+

+/+ E V726A/+ Mefv - Mefv - - / +/+ +/+ V726A/V726A V726A/V726A +/+ -/ +/+ V726A/V726A V726A/V726A V726A/V726A V726A/V726A Mefv Il1r1 Il1r1 Mefv Mefv Mefv Il1r1 Il1r1 Mefv Mefv Mefv D Peripheral blood

+/+ V726A/V726A +/+ V726A/V726A

Mefv Mefv Mefv Mefv Untreat LPS ATP LPS + Untreat LPS ATP LPS + Untreat LPS ATP LPS + Untreat LPS ATP LPS + +/+ +/+ -/- -/- Il1r1 Il1r1 Il1r1 Il1r1 Sup. 1000 1000 1000 1000 β 10.7 63.5 9.63 12.5 Pro-IL1 800 800 800 800

600 600 600 600

400 400

SSC 400 400 IL1b

200 200 200 200

0 0 0 0 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 10 10 10 10 10 10 10 10 10 10 100 101 102 103 104 10 10 10 10 10 CD11b Casp1 4 4 4 10 10 104 10 p10

3 3 3 10 10 103 10 24.1 7.92 31.1 20.3 2 2 2 2 Lys. β 10 12 10 79.3 10 9.78 10 9.27 Pro-IL1

F4/80 1 1 1 10 10 101 10 Actin 0 0 0 10 10 100 10 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 10 10 10 10 10 10 10 10 10 10 100 101 102 103 104 10 10 10 10 10 Ly-6G

Figure 4. Inflammation in KI Mice Is Induced by IL-1b (A) Luminex analysis of serum cytokines of 8-week-old Mefv+/+Rag1+/+, MefvV726A/+Rag1+/+, MefvV726A/V726ARag1+/+, Mefv+/+Rag1/, and MefvV726A/V726A Rag1/ mice. *p < 0.05; **p < 0.0001; ***p < 0.000001. (B) CD11b+ cells from BM of WT, MefvV726A/+, and MefvV726A/V726A mice were stimulated with LPS or no stimulus for 3 hr followed by treatment with or without ATP. Cell culture supernatants (Sup.) and cell lysates (Lys.) were analyzed by immunoblotting. (C–E) Analyses of MefvV726A/V726A mice on the Il1r1/ background. (C) Body weights of 8-week-old males. (D) Flow cytometry analyses of peripheral blood cells from 8-week-old mice. Data are representative of three independent experiments. (E) Inflammasome activation in MefvV726A/V726A mice with Il1r1/ background. Culture supernatants and lysates from BM CD11b+ cells were collected and subjected to immunoblot as described in (B). Additional information is provided in Figure S3.

762 Immunity 34, 755–768, May 27, 2011 ª2011 Elsevier Inc. Immunity Molecular Pathogenesis of FMF

A Figure 5. Inflammasome Activation in FMF LPS LPS ATP Patients Macrophages differentiated from PBMCs of (A) Untreat IFN-γ (16 hr) Untreat IFN-γ (16 hr) two healthy donors (controls 8 and 9), two FMF patients (FMF 9, homozygote of M694V; FMF 10, compound heterozygote of V726A/E148Q), (B) two healthy donors (controls 10 and 11), and two FMF patients (FMF 11, heterozygote of V726A; FMF 12, FMF 9 FMF 10 FMF 10 FMF 9 FMF 10 FMF 9 FMF 10 FMF 9 Control 9 Control 8 Control 9 Control 8 Control 8 Control 9 Control 8 Control 9 compound heterozygote of V726A/P369S) were Sup. IL1β treated with IFN-g or no stimulus for 16 hr. Macrophages were treated as described in Fig- ure 4B, and cell culture supernatants (Sup.) and cell Casp1 lysates (Lys.) were analyzed by immunoblotting. p20 Additional information is provided in Figure S4.

Lys. Pro-IL1β

these short-term assays was similar Pyrin among WT, pyrin-deficient, and FMF KI mice, which suggests that the NLRC4 and AIM2 inflammasomes do not mark- Actin edly contribute to caspase-1 activation of the FMF KI mice. Conversely, data from pyrin-deficient mice demonstrate B γ that pyrin is dispensable for NLRC4 and Untreated IFN- (16 hr) AIM2 inflammasome activation as well LPS LPS ATP LPS LPS ATP as NLRP3 activation by LPS and ATP. Nevertheless, macrophages from pyrin- deficient mice produced somewhat higher amounts of mature IL-1b com- pared with WT mice when stimulated FMF 12 FMF 11 FMF 12 FMF 11 FMF 12 FMF 11 FMF 12 FMF 11 Control 10 Control 11 Control 10 Control 11 Control 10 Control 11 Control 10 Control 11 with LPS alone for 1 to 3 days, as seen in our previous study of pyrin-truncation Sup. IL1β mice (Figure S6D; Chae et al., 2003). Taken together with previously reported Casp1 findings on the direct interaction of pyrin p20 with ASC, our results provide evidence for an ASC-dependent, pyrin inflamma- Lys. Pro-IL1β some, which is activated in the presence of the FMF-associated mutations of pyrin. Pyrin DISCUSSION Actin In this manuscript we described the profound phenotype established by intro- ducing FMF-associated mutations into (MefvV726A/V726ANlrp3/) were runted and showed similar in- the mouse germline. Consistent with the conception of FMF as flammatory phenotypes to the symptoms of KI mice expressing an autoinflammatory disorder, the cells and molecules of the Nlrp3 (MefvV726A/V726ANlrp3+/+)(Figures 7B and 7E). Moreover innate immune system are strongly implicated in the pathogen- we observed that the LPS-stimulated CD11b+ cells of esis of disease in these animals. BM-derived cells are both MefvV726A/V726ANlrp3/ actively secreted IL-1b and p10, even necessary and sufficient to confer the phenotype, which includes without ATP treatment, whereas the cells of Mefv+/+Nlrp3/ a marked expansion of granulocyte numbers in the blood and did not secrete IL-1b after LPS stimulation with ATP treatment peripheral lymphoid organs that persists on a Rag1/ back- (Figure 7D). These results demonstrate that NLRP3 is not neces- ground. Extending earlier data implicating IL-1b in the pathogen- sary for the caspase-1 activation of the FMF KI mice. esis of FMF, the disease phenotype is markedly attenuated in We also examined other NLRP3-independent inflammasomes mice deficient for the IL-1 receptor, and additional crosses by introducing Salmonella typhimurium or double-stranded DNA provided genetic evidence for an ASC-dependent, NLRP3-inde- into BMDMs to interrogate NLRC4 (IPAF, ICE protease-acti- pendent pyrin inflammasome. vating factor) or AIM2 (absent in melanoma 2) inflammasome Characterization of these FMF KI mice provides yet another activation, respectively (Figures S6A–S6C). IL-1b secretion in step in a paradigm shift in our understanding of the genetics of

Immunity 34, 755–768, May 27, 2011 ª2011 Elsevier Inc. 763 Immunity Molecular Pathogenesis of FMF

A Blood C p<0.0001 30 MefvV726A/V726A Mefv+/+ MefvV726A/V726A Rag1-/- Rag1-/- Rag1+/+ 1000 1000 1000 94 75.2 62.7 800 800 800 20

600 600 600

400 400 400 SSC 200 200 200 10 0 0 0 0 1 2 3 4 0 1 2 3 4 10 10 10 10 10 10 10 10 10 10 100 101 102 103 104 Body weight (g) weight Body CD11b

4 4 10 104 10 0

3 3 3 -/-

10 10 +/+ 8.82 10 23.5 6.33 /-

2 2 2 - 10 82.7 10 37.8 10 69 +/+ Rag1 Rag1 F4/80 1 1 1 10 10 10 V726A/V726A V726A/V726A +/+ +/+ Rag1 Rag1

0 0 10 100 10 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 Mefv Mefv Mefv Mefv Ly-6G D +/+ -/- +/+ +/+ V726A/V726A V726A/V726A B Spleen Mefv Rag1 Mefv Rag1 Rag1 Mefv MefvV726A/V726A Mefv+/+ MefvV726A/V726A Rag1-/- Rag1-/- Rag1+/+ 1000 1000 1000 89.7 76.6 58.1 800 800 800 LPS + ATP LPS + LPS + ATP LPS + Untreat LPS Untreat LPS Untreat LPS ATP LPS + 600 600 600

400 400 400 SSC Sup. Pro-IL-1β 200 200 200

0 0 0 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 β CD11b IL-1

4 10 104 104 42.1 8.95 43.4 3 10 103 103 29.1 73.9 27.3 Casp1 2 10 102 102 p10 F4/80 1 10 101 101 Lys. Pro-IL-1β 0 10 100 100 0 1 2 3 4 10 10 10 10 10 100 101 102 103 104 100 101 102 103 104 Ly-6G Actin

Figure 6. Analyses of MefvV726A/V726A Mice on a Rag1–/– Background (A and B) Flow cytometry analyses of peripheral blood cells (A) and splenocytes (B) from 8-week-old mice analyzed as described in Figure 3C. Data are representative of three independent experiments. (C) Body weights of 8-week-old males. (D) Inflammasome activation in MefvV726A/V726A mice with Rag1/ background. Culture supernatants and lysates from BM CD11b+ cells were analyzed by immunoblotting. Additional information is provided in Figure S5.

FMF. Based on segregation analysis in Sephardi Jewish families known autoinflammatory (Booty et al., 2009; Marek-Yagel with severe disease (Sohar et al., 1967), FMF has long been et al., 2009; Ozen, 2009). This observation, taken together with considered an autosomal-recessive illness, and in fact the posi- the fact that nearly all FMF-associated mutations are missense tional cloning studies leading to the identification of MEFV were rather than null mutations, has led to a reconsideration of the based on this assumption. However, the availability of genetic simple loss-of-function recessive model of FMF inheritance. testing has led both to the recognition of a biochemical pheno- The data presented here add substantially to the case for a gain- type in asymptomatic heterozygotes (Lachmann et al., 2006) of-function model. In the current study, we have shown that the and to the diagnosis of FMF in patients with relatively mild clinical insertion of three different FMF-associated mutant human B30.2 phenotypes. As many as 30% of the patients diagnosed with domains into mouse pyrin induced inflammatory phenotypes FMF have only a single demonstrable despite similar to or more severe than those seen in FMF patients, whereas sequencing of the entire MEFV genomic region and several other deleting mouse pyrin produced no overt inflammatory phenotype.

764 Immunity 34, 755–768, May 27, 2011 ª2011 Elsevier Inc. Immunity Molecular Pathogenesis of FMF

AEPeripheral blood Peripheral blood

Mefv+/+ MefvV726A/V726A Mefv+/+ MefvV726A/V726A Mefv+/+ MefvV726A/V726A Mefv+/+ MefvV726A/V726A Asc+/+ Asc+/+ Asc-/- Asc-/- Nlrp3+/+ Nlrp3+/+ Nlrp3-/- Nlrp3-/- 1000 1000 1000 1000 1000 12.1 1000 67.6 1000 16.1 1000 73 14.3 75 13.6 19.2 800 800 800 800 800 800 800 800

600 600 600 600 600 600 600 600

400 400 400 400 400 400 400 400 SSC SSC

200 200 200 200 200 200 200 200

0 0 0 0 0 0 0 0 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 CD11b CD11b 4 4 104 10 10 104 104 104 104 104

3 3 103 10 10 103 103 103 103 103 43.8 7.11 43 34.9 33.7 11.3 27.7 9.14 2 2 83.2 2 2 2 2 2 2 10 33.6 10 10 40.5 10 53.4 10 33.4 10 61.8 10 36 10 72.6 F4/80 F4/80 1 1 101 10 10 101 101 101 101 101

0 0 100 10 10 100 100 100 100 100 0 1 2 3 4 0 1 2 3 4 100 101 102 103 104 10 10 10 10 10 10 10 10 10 10 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 Ly-6G Ly-6G

B C D

p<0.00001 p<0.0001 -/- +/+ -/- +/+ +/+ +/+ +/+ V726A/V726A V726A/V726A +/+ V726A/V726A V726A/V726A - 30 / +/+ - -/- +/+ Nlrp3 Mefv Mefv Nlrp3 Mefv Asc Mefv Asc Mefv Asc Mefv Asc Mefv Nlrp3 Mefv 25 Nlrp3 20 15 10 Untreat LPS Untreat LPS ATP LPS + Untreat LPS ATP LPS + LPS + ATP LPS + Untreat LPS ATP LPS + Untreat LPS ATP LPS + Untreat LPS ATP LPS + Untreat LPS ATP LPS + Untreat LPS ATP LPS + Body weight (g) weight Body 5 Sup. Pro-IL1β 0 -/- -/- +/+ +/+ +/+ +/+ IL1β Nlrp3 Nlrp3 Nlrp3 Nlrp3 Nlrp3 Nlrp3 - -/ +/+ +/+ -/- +/+ +/+ Asc Asc Asc Asc Asc Asc Casp1 +/+ +/+ +/+ p10 Mefv Mefv Mefv Lys.

V726A/V726A V726A/V726A Pro-IL1β V726A/V726A Mefv Mefv

Mefv Actin

Figure 7. FMF-Associated B30.2 Mutations Activate an ASC-Dependent but NALP3-Independent Pyrin Inflammasome (A–C) Analyses of the FMF KI (MefvV726A/V726A) mice with Asc/ background. (A) Flow cytometry analyses of peripheral blood cells from 8-week-old mice analyzed as in Figure 3C. Data are representative of three independent experiments. (B) Body weights of indicated 8-week-old males. (C) Inflammasome activation in MefvV726A/V726A mice with Asc/ background. Culture supernatants and lysates from BM CD11b+ cells were analyzed by immunoblotting. (D and E) Analyses of the FMF KI (MefvV726A/V726A) mice with Nlrp3/ background. (D) Inflammasome activation in MefvV726A/V726A mice with Nlrp3/ background. Culture supernatants and lysates from BM CD11b+ cells were analyzed by immunoblotting. (E) Flow cytometry analyses of peripheral blood cells from 8-week-old mice analyzed as in Figure 3C. Data are representative of three independent experiments. Additional information is provided in Figure S6.

Neither heterozygous nor hemizygous KI mice exhibited the B30.2 pyrin mutations but rather to the expression of any pyrin cardinal features seen in homozygotes, indicating a gene dosage that chimerizes the N-terminal mouse pyrin with the C-terminal effect. The fact that pyrin expression is cytokine inducible human B30.2 domain. Absent the generation of knockin mice suggests a possible feed-forward mechanism in which homozy- for the wild-type human B30.2 domain, this will always remain gous KI cells may express more pyrin at baseline, which in turn formally possible. However, against this hypothesis, there is leads to inflammatory cytokine production and further increases a genotype-phenotype relationship in the mice, with the severity in pyrin expression. Such a gene dosage effect may also be oper- of the phenotype in the mice inversely proportionate to the ative in , in that higher pyrin protein expression has been severity of the phenotype associated with the corresponding observed in the leukocytes of FMF patients, compared with genotype in humans. A similar observation has recently been healthy controls (Booty et al., 2009; Chae et al., 2008). reported for NLRP3 KI mice (Brydges et al., 2009). Of possible It is possible that the disease phenotype observed in FMF KI relevance in the case of pyrin, mice do not express a C-terminal mice is due not to the expression of the specific FMF-associated B30.2 domain (Chae et al., 2000). Thus, it is not totally surprising

Immunity 34, 755–768, May 27, 2011 ª2011 Elsevier Inc. 765 Immunity Molecular Pathogenesis of FMF

that mice harboring the wild-type human B30.2 domain may ment (Milledge et al., 2002). Although the morbidity of allogeneic express a phenotype even more severe (embryonic lethality) bone marrow transplantation generally makes this an unattrac- than the V726A homozygous genotype, which is associated tive alternative, the possibility of gene correction with inducible with the mildest FMF phenotype in humans. pleuripotent stem cells may eventually modify this calculus. We present several lines of evidence that IL-1 plays a crucial FMF KI mice may also shed light on the role of pyrin variants in role in the phenotype observed in mice homozygous for FMF- human history and evolution. The extraordinarily high carrier associated pyrin mutations. To date there are four different frequency of FMF-associated MEFV mutations in certain popula- macromolecular inflammasomes, nucleated by NLRP1, tions, coupled with the fact that different mutations predominate NLRP3, NLRC4, and AIM2. The NLRP3 inflammasome has in geographically proximate ethnic groups, strongly suggests been shown to be activated by a wide range of pathogen-asso- that heterozygous FMF mutations may confer a selective advan- ciated or danger-associated molecular patterns (PAMPs, tage against some as-yet-unknown pathogen. Moreover, DAMPs) and murine Nlrp3/ cells do not produce any detect- sequence comparisons of the C-terminal B30.2 doman of pyrin able amount of IL-1b in response to LPS and ATP in vitro (Maria- with nonhuman primates suggest episodic positive selection thasan et al., 2006; Sutterwala et al., 2006). Nevertheless, we (Schaner et al., 2001). The fact that several disease-associated have found that the NLRP3 inflammasome has no role in the mutations are clustered around a putative binding pocket in inflammation of FMF KI mice. In addition, we did not observe the C-terminal B30.2 domain provides a possible structural basis any differences in IL-1b secretion among WT, pyrin-deficient, for this hypothesis (Masters et al., 2009). It is thus possible that and KI macrophages induced by double-stranded DNA or one of the functions of pyrin is to serve as an intracellular S. typhimurium, suggesting that the AIM2 or NLRC4 inflamma- receptor for one or more PAMPs. FMF KI mice may serve as some is not involved in the inflammation of KI mice. The neces- the experimental platform to test such hypotheses, thereby al- sity of ASC for the inflammation of KI mice also excludes the lowing us to probe these most fundamental questions regarding NLRP1 inflammasome from the pathogenesis of FMF inflamma- the role of pyrin and its mutants in human health and disease. tion, because murine NLRP1 orthologs are predicted to be unable to interact with ASC because of lack of functional PYRIN EXPERIMENTAL PROCEDURES domains. Indeed, ASC is dispensable for NLRP1-dependent caspase-1 activation in mouse macrophages (Hsu et al., 2008). Mice Taken together, these data suggest a previously unrecognized Generation of FMF KI Mice ASC-dependent inflammasome in which mutant pyrin induces With pPNT-double loxP as a construct backbone, we generated the targeting constructs by inserting a 6.0 kb KpnI/PvuII genomic fragment encompassing IL-1b activation. 0 the end of exon 2 to exon 6 (5 arm) and a 2.5 kb PstI/SpeI genomic fragment An inhibitory role of pyrin in caspase-1 activation has been (30 arm) (Figure S1A). For the KI of WT human B30.2 domain, a 1.7 kb genomic proposed in our previous study with a mouse model expressing fragment of human MEFV encompassing exons 7 to 10 was inserted a truncated form of pyrin (Chae et al., 2003). However, this inter- upstream of the 30 arm. The mutant genomic fragments with M680I, M694V, pretation has been debated because the pyrin-truncation mice or V726A were also generated by site-directed mutagenesis with Quick- express a functional N-terminal PYRIN domain that can interact Change XL site-directed mutagenesis kit (Stratagene) and inserted in the upstream segment of the 30 arm. Linearized constructs were introduced into with ASC. In the current study, we have shown that pyrin-null 129/SvEvTac-derived TC-1 embryonic stem (ES) cells. G418- and gancyclo- macrophages produced higher amounts of mature IL-1b when vir-resistant clones were screened for homologous recombination by long- stimulated with LPS alone for 1 to 3 days, although they did range PCR and Southern blotting (Figure S1B). Heterozygotes generated not show evidence of increased IL-1b activation in short-term from chimeras were crossed with EIIa-Cre transgenic mice (Jackson Labora- assays of inflammasome activation. Moreover, neither pyrin- tory) to remove the neo cassette and backcrossed with C57BL/6 mice at least truncation nor -deficient mice exhibit a spontaneous inflamma- six generations. Generation of Pyrin-Deficient Mice tory phenotype. These results suggest a more subtle inhibitory As shown in Figure S1E, the targeting constructs were generated by inserting role of pyrin that may be limited to noninflammasome-mediated a 4.7 kb genomic fragment upstream of exon 1 (50 arm) and a 6.0 kb genomic IL-1b activation. fragment encompassing exon 3 to exon 7 (30 arm). By using the linearized To test the hypothesis that pyrin mutations cause IL-1b activa- construct, we generated pyrin KO mice as described in the generation of KI tion in FMF patients, we cultured patient or control PBMCs for mice and backcrossed them with C57BL/6 mice at least six generations. 7 days to minimize the effect of the patients’ colchicine treat- Other Mouse Strains / / / ment, and then induced pyrin with IFN-g (Centola et al., 2000). Il1r1 and Rag1 mice were from the Jackson Laboratory. Asc and Nlrp3/ mice were a gift from V.M. Dixit (Genentech). In 4/4 FMF patients, but 0/4 controls, we observed caspase-1 All animal studies were performed according to National Institutes of Health activation and IL-1b secretion even without an ATP second guidelines and were approved by the Institutional Animal Care and Use signal. These data provide a scientific basis for IL-1 inhibition Committee of National Institute of Arthritis and Musculoskeletal and Skin in the treatment of FMF, particularly in patients with life-threat- Diseases (NIAMS). ening amyloid deposition. Our data also establish a conceptual basis for the possible Flow Cytometry Analysis development of hematopoietic stem cell therapy for refractory RBC-lysed single-cell suspensions were obtained from LNs, spleen, BM, and FMF. In reciprocal BM transfer experiments, we found that cells peripheral blood and stained with fluorochrome-conjugated antibodies in PharMingen Stain Buffer (BSA) (BD PharMingen, San Diego, CA). CD3 APC, of the hematopoietic lineage are both necessary and sufficient CD19 FITC, CD11b PE, Ly-6G FITC, CD3 PerCP-Cy5.5, CD4 FITC, CD8a for the FMF KI phenotype. In a patient with FMF who underwent APC, CD25 PE, CD44 PE, CD62L PE, and CD69 PE were purchased from bone marrow transplantation for congenital dyserythropoietic BD PharMingen. F4/80 Alexa647 Ab was purchased from Invitrogen (Carlsbad, anemia, FMF symptoms completely resolved off colchicine treat- CA). CD45.1 (Alexa488, Alexa647) and CD45.2 (Alexa488, Alexa647) Abs were

766 Immunity 34, 755–768, May 27, 2011 ª2011 Elsevier Inc. Immunity Molecular Pathogenesis of FMF

purchased from Biolegend (San Diego, CA). Cells were analyzed on a REFERENCES FACSCaliber flow cytometer (BD Biosciences, San Jose, CA) and analyzed with FlowJo software (Three Star, Ashland, OR). Belkhir, R., Moulonguet-Doleris, L., Hachulla, E., Prinseau, J., Baglin, A., and Hanslik, T. (2007). Treatment of familial Mediterranean fever with anakinra. Cell Culture and Immunoblots Ann. Intern. Med. 146, 825–826. Peritoneal macrophages were obtained by washing the peritoneum with RPMI Bilginer, Y., Ayaz, N.A., and Ozen, S. (2009). Anti-IL-1 treatment for secondary medium. Mouse CD11b+ cells were isolated from BM with CD11b MicroBeads in an adolescent with FMF and Behcet’s disease. Clin. Rheumatol. (Miltenyi Biotech, Auburn, CA). BMDMs were obtained by differentiating BM 29, 209–210. cells with 10 ng/ml of M-CSF (Peprotech, Rocky Hill, NJ) for 7 days. Peritoneal Booty, M.G., Chae, J.J., Masters, S.L., Remmers, E.F., Barham, B., Le, J.M., m macrophages were treated with 1 g/ml of LPS (Invivogen, San Diego, CA), Barron, K.S., Holland, S.M., Kastner, D.L., and Aksentijevich, I. (2009). 12.5 ng/ml of IL-4 (Peprotech), or LPS/IL-4. After 24 hr, cell lysates were Familial Mediterranean fever with a single MEFV mutation: Where is the + m collected. CD11b cells or BMDMs were stimulated with LPS (1 g/ml) or no second hit? Arthritis Rheum. 60, 1851–1861. stimulus for 3 hr. Cell culture media were changed with fresh media containing Brydges, S.D., Mueller, J.L., McGeough, M.D., Pena, C.A., Misaghi, A., 5 mM ATP or media only. After 30 min, cell culture supernatants and cell Gandhi, C., Putnam, C.D., Boyle, D.L., Firestein, G.S., Horner, A.A., et al. lysates were collected. (2009). Inflammasome-mediated disease animal models reveal roles for innate Blood specimens from healthy controls and FMF patients were drawn after but not adaptive immunity. Immunity 30, 875–887. obtaining informed consent under a protocol approved by the NIAMS Institu- tional Review Board. PBMCs were isolated by Ficoll-Hypaque centrifugation Calligaris, L., Marchetti, F., Tommasini, A., and Ventura, A. (2008). The efficacy and also differentiated into macrophages with 800 units/ml of GM-CSF (Pepro- of anakinra in an adolescent with colchicine-resistant familial Mediterranean tech). Macrophages were treated with 20 ng/ml of IFN-g (Peprotech) or no fever. Eur. J. Pediatr. 167, 695–696. stimulus for 16 hr. Macrophages or PBMCs were stimulated with TLR ligands, Centola, M., Wood, G., Frucht, D.M., Galon, J., Aringer, M., Farrell, C., Kingma, LPS (1 mg/ml), Pam3CSK4 (1 mg/ml), or Gardiquimod (1 mg/ml) (Invivogen). After D.W., Horwitz, M.E., Mansfield, E., Holland, S.M., et al. (2000). The gene for 3 hr, cell culture media were changed with fresh media, cells were incubated familial Mediterranean fever, MEFV, is expressed in early leukocyte develop- for 30 min, and cell culture supernatants and cell lysates were collected. Cell ment and is regulated in response to inflammatory mediators. Blood 95, lysates or cell culture supernatants were subjected to SDS-PAGE for immuno- 3223–3231. blot analysis by primary antibodies: anti-mouse pyrin Ab (Chae et al., 2003); Chae, J.J., Aksentijevich, I., and Kastner, D.L. (2009). Advances in the under- anti-human pyrin Ab (Chae et al., 2006); anti-human IL-1b and mouse IL-1b standing of familial Mediterranean fever and possibilities for targeted therapy. Abs (R&D Systems, Minneapolis, MN); anti-mouse caspase-1, human Br. J. Haematol. 146, 467–478. caspase-1, and actin Abs (Santa Cruz Biotechnology, Santa Cruz, CA). Chae, J.J., Centola, M., Aksentijevich, I., Dutra, A., Tran, M., Wood, G., Nagaraju, K., Kingma, D.W., Liu, P.P., and Kastner, D.L. (2000). Isolation, BMT genomic organization, and expression analysis of the mouse and rat homologs For routine BMT experiments, 6- to 8-week-old female mice were lethally of MEFV, the gene for familial mediterranean fever. Mamm. Genome 11, (10 Gy) irradiated the day before BMT and reconstituted by intravenous injec- 428–435. tion of 2 3 106 (WT) or 5 3 106 (KI) RBC-lysed BM cells, unless otherwise specified. For the transfer of KI phenotypes into WT recipients, CD45.1+/+ Chae, J.J., Komarow, H.D., Cheng, J., Wood, G., Raben, N., Liu, P.P., and C57BL/6 mice were used as recipients. For the rescue of inflammatory pheno- Kastner, D.L. (2003). Targeted disruption of pyrin, the FMF protein, causes types of KI mice with WT BM cells, 10 3 106 BM cells from CD45.1+/+ C57BL/6 heightened sensitivity to endotoxin and a defect in macrophage apoptosis. mice were transferred intraperitoneally into 3- to 4-week-old CD45.2+/+ Mol. Cell 11, 591–604. MefvV726A/V726A mice that had been irradiated with two 3.5 Gy (2 hr interval, Chae, J.J., Wood, G., Masters, S.L., Richard, K., Park, G., Smith, B.J., and finished 2 hr before i.p. injection). For the mixed BMT, 2 3 106 CD45.1+/+ Kastner, D.L. (2006). The B30.2 domain of pyrin, the familial Mediterranean WT BM cells were premixed with either 2 3 106 CD45.2+/+ WT or 5 3 106 fever protein, interacts directly with caspase-1 to modulate IL-1b production. CD45.2+/+ KI (MefvM680I/M680I) BM cells immediately before injection and trans- Proc. Natl. Acad. Sci. USA 103, 9982–9987. ferred to recipients heterozygous for both CD45.1 and CD45.2 antigens. Chae, J.J., Wood, G., Richard, K., Jaffe, H., Colburn, N.T., Masters, S.L., Gumucio, D.L., Shoham, N.G., and Kastner, D.L. (2008). The familial Measurement of Cytokines, Chemokines, Mediterranean fever protein, pyrin, is cleaved by caspase-1 and activates and Hematopoietic Factors NF-kB through its N-terminal fragment. Blood 112, 1794–1803. Sera were obtained from 8-week-old mice and analyzed by Bio-Plex Mouse Diaz, A., Hu, C., Kastner, D.L., Schaner, P., Reginato, A.M., Richards, N., and Cytokine 23-plex Assay (Bio-Rad, Hercules, CA) according to the manufac- Gumucio, D.L. (2004). Lipopolysaccharide-induced expression of multiple turer’s instructions. alternatively spliced MEFV transcripts in human synovial fibroblasts: A prom- inent splice isoform lacks the C-terminal domain that is highly mutated in Statistical Analyses familial Mediterranean fever. Arthritis Rheum. 50, 3679–3689. Statistical analyses were performed with the Student’s t test. French FMF Consortium. (1997). A candidate gene for familial Mediterranean fever. Nat. Genet. 17, 25–31. SUPPLEMENTAL INFORMATION Galon, J., Aksentijevich, I., McDermott, M.F., O’Shea, J.J., and Kastner, D.L. (2000). TNFRSF1A mutations and autoinflammatory syndromes. Curr. Opin. Supplemental Information includes six figures and one table and can be found Immunol. 12, 479–486. with this article online at doi:10.1016/j.immuni.2011.02.020. Hsu, L.C., Ali, S.R., McGillivray, S., Tseng, P.H., Mariathasan, S., Humke, E.W., Eckmann, L., Powell, J.J., Nizet, V., Dixit, V.M., and Karin, M. (2008). A NOD2- ACKNOWLEDGMENTS NALP1 complex mediates caspase-1-dependent IL-1b secretion in response to Bacillus anthracis infection and muramyl dipeptide. Proc. Natl. Acad. Sci. This work was supported by the Intramural Research Program of NIAMS, NCI, USA 105, 7803–7808. and NHGRI, NIH. The International FMF Consortium. (1997). Ancient missense mutations in Received: July 27, 2010 a new member of the RoRet gene family are likely to cause familial Revised: December 10, 2010 Mediterranean fever. Cell 90, 797–807. Accepted: February 12, 2011 Kastner, D.L., Aksentijevich, I., and Goldbach-Mansky, R. (2010). Published online: May 19, 2011 Autoinflammatory disease reloaded: A clinical perspective. Cell 140, 784–790.

Immunity 34, 755–768, May 27, 2011 ª2011 Elsevier Inc. 767 Immunity Molecular Pathogenesis of FMF

Kuijk, L.M., Govers, A.M., Frenkel, J., and Hofhuis, W.J. (2007). Effective Papin, S., Cuenin, S., Agostini, L., Martinon, F., Werner, S., Beer, H.D., Gru¨ tter, treatment of a colchicine-resistant familial Mediterranean fever patient with C., Gru¨ tter, M., and Tschopp, J. (2007). The SPRY domain of Pyrin, mutated in anakinra. Ann. Rheum. Dis. 66, 1545–1546. familial Mediterranean fever patients, interacts with inflammasome compo- nents and inhibits proIL-1beta processing. Cell Death Differ. 14, 1457–1466. Lachmann, H.J., Sengu¨ l, B., Yavuzsxen, T.U., Booth, D.R., Booth, S.E., Bybee, A., Gallimore, J.R., Soytu¨ rk, M., Akar, S., Tunca, M., and Hawkins, P.N. (2006). Roldan, R., Ruiz, A.M., Miranda, M.D., and Collantes, E. (2008). Anakinra: New Clinical and subclinical inflammation in patients with familial Mediterranean therapeutic approach in children with Familial Mediterranean Fever resistant to fever and in heterozygous carriers of MEFV mutations. Rheumatology colchicine. Joint Bone Spine 75, 504–505. (Oxford) 45, 746–750. Schaner, P., Richards, N., Wadhwa, A., Aksentijevich, I., Kastner, D., Tucker, Marek-Yagel, D., Berkun, Y., Padeh, S., Abu, A., Reznik-Wolf, H., Livneh, A., P., and Gumucio, D. (2001). Episodic evolution of pyrin in primates: Human Pras, M., and Pras, E. (2009). Clinical disease among patients heterozygous mutations recapitulate ancestral amino acid states. Nat. Genet. 27, 318–321. for familial Mediterranean fever. Arthritis Rheum. 60, 1862–1866. Sohar, E., Gafni, J., Pras, M., and Heller, H. (1967). Familial Mediterranean fever. A survey of 470 cases and review of the literature. Am. J. Med. 43, Mariathasan, S., Weiss, D.S., Newton, K., McBride, J., O’Rourke, K., Roose- 227–253. Girma, M., Lee, W.P., Weinrauch, Y., Monack, D.M., and Dixit, V.M. (2006). Cryopyrin activates the inflammasome in response to toxins and ATP. Sutterwala, F.S., Ogura, Y., Szczepanik, M., Lara-Tejero, M., Lichtenberger, Nature 440, 228–232. G.S., Grant, E.P., Bertin, J., Coyle, A.J., Gala´ n, J.E., Askenase, P.W., and Flavell, R.A. (2006). Critical role for NALP3/CIAS1/Cryopyrin in innate and Masters, S.L., Simon, A., Aksentijevich, I., and Kastner, D.L. (2009). Horror adaptive immunity through its regulation of caspase-1. Immunity 24, 317–327. autoinflammaticus: The molecular pathophysiology of autoinflammatory disease (*). Annu. Rev. Immunol. 27, 621–668. Touitou, I., Lesage, S., McDermott, M., Cuisset, L., Hoffman, H., Dode, C., Shoham, N., Aganna, E., Hugot, J.P., Wise, C., et al. (2004). Infevers: An Milledge, J., Shaw, P.J., Mansour, A., Williamson, S., Bennetts, B., Roscioli, T., evolving mutation database for auto-inflammatory syndromes. Hum. Mutat. Curtin, J., and Christodoulou, J. (2002). Allogeneic bone marrow transplanta- 24, 194–198. tion: Cure for familial Mediterranean fever. Blood 100, 774–777. Yu, J.W., Fernandes-Alnemri, T., Datta, P., Wu, J., Juliana, C., Solorzano, L., Moser, C., Pohl, G., Haslinger, I., Knapp, S., Rowczenio, D., Russel, T., McCormick, M., Zhang, Z., and Alnemri, E.S. (2007). Pyrin activates the ASC Lachmann, H.J., Lang, U., and Kovarik, J. (2009). Successful treatment of pyroptosome in response to engagement by autoinflammatory PSTPIP1 familial Mediterranean fever with Anakinra and outcome after renal transplan- mutants. Mol. Cell 28, 214–227. tation. Nephrol. Dial. Transplant. 24, 676–678. Yu, J.W., Wu, J., Zhang, Z., Datta, P., Ibrahimi, I., Taniguchi, S., Sagara, J., Ozen, S. (2009). Changing concepts in familial Mediterranean fever: Is it Fernandes-Alnemri, T., and Alnemri, E.S. (2006). Cryopyrin and pyrin activate possible to have an autosomal-recessive disease with only one mutation? caspase-1, but not NF-kB, via ASC oligomerization. Cell Death Differ. 13, Arthritis Rheum. 60, 1575–1577. 236–249.

768 Immunity 34, 755–768, May 27, 2011 ª2011 Elsevier Inc.