PYPAF3, a PYRIN-containing APAF-1-like Protein, Is a Feedback Regulator of Caspase-1-dependent Interleukin-1 (cid:1) Secretion* □ S

PYPAF3 is a member of the PYRIN-containing apoptotic protease-activating factor-1-like proteins (PYPAFs, also called NALPs). Among the members of this family, PYPAF1, PYPAF5, PYPAF7, and NALP1 have been shown to induce caspase-1-dependent interleukin-1 (cid:1) secretion and NF- (cid:2) B activation in the presence of the adaptor molecule ASC. On the other hand, we recently discovered that PYNOD, another member of this family, is a suppressor of these re-sponses. Here, we show that PYPAF3 is the second member that inhibits caspase-1-dependent interleukin-1 (cid:1) secretion. In contrast, PYPAF2/NALP2 does not inhibit this response but rather inhibits the NF- (cid:2) B activation that is induced by the combined expression of PYPAF1 and ASC. Both PYPAF2 and PYPAF3 mRNAs are broadly expressed in a variety of tissues; however, neither is expressed in skeletal muscle, and only PYPAF2 mRNA is expressed in heart and brain. They are also expressed in many cell lines of both hematopoietic and non-hematopoietic lineages. Stimulation of monocytic THP-1 cells with lipopolysaccharide , interleukin-1 (cid:1) ; Ipaf, IL-1-con-verting enzyme-protease-activating factor; CARD, caspase recruitment domain; Apaf, apoptotic protease-activating factor; PYPAF, PYRIN-containing Apaf-1-like protein; TNF, tumor necrosis factor; LRR, leucine-rich repeat; RIP-DD, receptor-interacting protein death domain; HA, human influenza virus hemagglutinin epitope; LPS, lipo- polysaccharide; RT, reverse transcription; ELISA, enzyme-linked im-munosorbent assay.

Interleukin (IL) 1 -1␤ is produced upon microbial infection mainly from monocytes/macrophages and neutrophils. IL-1␤ plays a fundamental role in protecting animals against infectious agents by activating immune cells, inducing a variety of cytokines and acute phase proteins and acting as a pyrogen (1,2). On the other hand, excessive production of IL-1␤ induces septic shock and is fatal for animals. Therefore, the regulation of IL-1␤ production is an important issue in infectious diseases from both the pathophysiological and pharmacological points of view. However, the regulatory mechanisms of IL-1␤ production are poorly understood.
Recently, we discovered that PYNOD, a novel member of the PYPAFs, has the potential to inhibit caspase-1-dependent IL-1␤ secretion as well as ASC-mediated NF-B activation (10). This was the first example of a PYPAF member playing a negative regulatory role in IL-1␤ secretion. This finding prompted us to investigate whether other ill-characterized members of this family have activities similar to PYNOD. Here we investigated the effects of PYPAF2 and PYPAF3 expression on NF-B activation and IL-1␤ secretion. These two proteins are highly homologous and likely to be products of a very recent gene duplication event. We found that PYPAF3 inhibits caspase-1-dependent IL-1␤ secretion but not ASC-mediated NF-B, while PYPAF2 inhibits the latter but not the former. Further study suggested that PYPAF3 is a feedback regulator of caspase-1-dependent IL-1␤ secretion.
Plasmids-A cDNA encoding human PYPAF2 (IMAGE:2821428) was purchased from Open Biosystems (Huntsville, AL). A cDNA encoding human PYPAF3 was amplified by RT-PCR from human peripheral blood lymphocyte poly(A)(ϩ) RNA. The cDNAs encoding PYPAF2, PYPAF3, and their leucine-rich repeat (LRR)-truncated mutants (PYPAF2dLRR (amino acids 1-683) and PYPAF3dLRR (amino acids 1-685), respectively) were cloned into the pEF-Bos mammalian expression vector (11). Plasmids expressing human PYPAF1, PYPAF1dLRR (amino acids 1-739), a dominant negative mutant of TRADD (amino * This work was supported in part by grants-in-aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology; the Japanese Government; and a grant from Novartis Foundation (Japan) for the Promotion of Science. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Figs. S1 and S2.
acids 196 -312 with alanine substitution at amino acids 296 -299), procaspase-1, pro-IL-1␤, ASC, MyD88, and receptor-interacting protein death domain (RIP-DD, amino acids 558 -671) with or without an N-terminal FLAG-tag were described previously (10). The human influenza virus hemagglutinin epitope (HA) was introduced as needed using pHA-EAK and pEAK-HA, which are pEAK8-derived mammalian expression vectors in which the sequence for HA is inserted at the Nand C-terminal side of the multicloning site, respectively. pEAK-HA was kindly provided by Dr. Hisashi Miyamori (Kanazawa University, Cancer Research Institute). Human Bcl-XL DNA was a kind gift from Dr. Shigeomi Shimizu (Osaka University Medical School), and it was subcloned into pHA-EAK.
For luciferase reporter assays and IL-1␤ secretion assays, HEK293 cells were plated onto 48-well plates at a density of 4 ϫ 10 4 cells/well, and transfection was carried out 18 h later using linear polyethyleneimine (M r about 25,000, Polysciences Inc., Warrington, PA) as described previously (12). For immunoprecipitation assays, 293T cells were plated onto 6-well plates at density of 5 ϫ 10 5 cells/well, and transfection was carried out 18 h later using Lipofectamine Plus (Invitrogen). To generate stable transfectants, THP-1 cells were transfected with pHA-EAK-PYPAF3 by electroporation. After selection with puromycin for 4 weeks, stable transfectants expressing HA-tagged PYPAF3 were selected by Western blot analysis.
Luciferase Reporter Assays for NF-B Activity-HEK293 cells in 48-well plates were transfected with plasmids including 15 ng/well of pNF-B-Luc (carrying a firefly luciferase cDNA driven by 5ϫ NF-Bbinding sites; Stratagene, La Jolla, CA) and 3 ng/well of pRL-TK (carrying Renilla luciferase cDNA driven by the HSV-TK promoter; Promega, Madison, WI). The total amount of DNA in each transfection was kept constant (300 ng/well) by the addition of empty vector (pEF-BOS).
After 24 h, cell lysates were prepared and the firefly, and Renilla luciferase activities were measured using the dual-luciferase reporter assay system (Promega). The relative luciferase activity was calculated as follows: relative luciferase activity ϭ firefly luciferase activity/renilla luciferase activity.
IL-1␤ Secretion Assays-HEK293 cells in 48-well plates were transfected with plasmids including those expressing pro-IL-1␤ (80 ng/well) and procaspase-1 (2 ng/well, unless otherwise specified). The total amount of DNA in each transfection was kept constant (300 ng/well) by the addition of empty vector (pEF-BOS). After 26 h, the culture supernatants were collected and subjected to ELISA for IL-1␤ using the Human IL-1␤ OptEIA ELISA Set (Pharmingen).
Immunoprecipitation and Western Blot Analysis-Immunoprecipitation and Western blotting were carried out as previously described (10), except that transfection and cell culture were performed in 6-well plates and that rabbit polyclonal and mouse monoclonal anti-HA antibodies (Sigma) were used for immunoprecipitation and Western blots, respectively. PYPAF-3-specific antisera were prepared from mice immunized with purified glutathione S-transferase-fused PYPAF-3 PYRIN domain (amino acids 1-93).
Assays for mRNA Expression-Panels of first-strand cDNA from multiple human tissues (Human I and Human II) were purchased from Clontech (Palo Alto, CA). Total RNA from cultured cell lines was prepared using the RNeasy mini kit (Qiagen, Hilden, Germany). RT-PCR analysis was performed as described previously (13). The primers used to detect human PYPAF2 mRNA were as follows: sense, 5Ј-gcaaaggatgaagtcagaga-3Ј; antisense, 5Ј-ccctccacccttatgtagat-3Ј, while those for PYPAF3 mRNA were as follows: sense, 5Ј-gcagtgggctgaattcttct-3Ј; antisense, 5Ј-gcctgacagagaatccacaa-3Ј. The amount of template cDNA was adjusted so that a similar amount of a PCR fragment of ␤-actin was generated within the linear range of the PCR.

PYPAF2 Inhibits the NF-B Activation Induced by the Combined Expression of PYPAF1 and ASC-
We first investigated the effect of PYPAF2 and PYPAF3 expression on ASC-mediated NF-B activation in HEK293 cells using luciferase reporter assays. Consistent with a previous report (14), PYPAF1 promoted this response, and truncation of the LRRs (PYPAF1dLRR) enhanced this activity (Fig. 1A). In contrast, PYPAF2 and PYPAF3 and even their LRR-truncated mutants failed to show such an activity (Fig. 1A). Instead, PYPAF2 dose-dependently inhibited the NF-B activation induced by PYPAF1dLRR plus ASC (Fig. 1B). PYPAF3 showed little or no such activity. It was previously shown that PYPAF4 inhibits NF-B activation induced by TNF-␣ or IL-1␤ (15). In contrast, PYPAF2 did not inhibit TNF-␣-induced NF-B activation, while a dominant negative mutant of TRADD inhibited this response, serving as a positive control (Fig. 1C). In this context, PYPAF2 is more similar to PYNOD (10) than PYPAF4.
PYPAF3 Inhibits Caspase-1-dependent IL-1␤ Secretion-Next, we investigated whether PYPAF2 and PYPAF3 promotes caspase-1-dependent IL-1␤ secretion in the presence of ASC in HEK293 cells. Previous reports lack agreement on whether or not PYPAF2 promotes this response (5,9). In our hands, IL-1␤ secretion was not induced by PYPAF2, PYPAF3, or their LRRdeletion mutants, irrespective of the presence or absence of ASC ( Fig. 2A and supplemental Fig. S1, A and B), while it was efficiently induced by PYPAF1 and PYPAF1dLRR in the presence of ASC as reported previously (4,10). We then investigated the effect of PYPAF2 or PYPAF3 expression on the caspase-1-dependent IL-1␤ secretion induced by PYPAF1dLRR plus ASC. In contrast to its lack of inhibitory activity against ASC-mediated NF-B activation described above, PYPAF3 strongly and dose-dependently inhibited the IL-1␤ secretion, while PYPAF2, which successfully inhibited the ASC-mediated NF-B activation, did not (Fig. 2B). To further examine the mechanism by which PYPAF3 inhibits caspase-1-dependent IL-1␤ secretion, we investigated the effect of PYPAF3 on the IL-1␤ secretion induced by caspase-1 overexpression. In the absence of ASC and PYPAF1, caspase-1 induced IL-1␤ secretion dose-dependently (Fig. 2C). The expression of PYPAF3 but not PYPAF2 inhibited this response dose-dependently, indicating that PYPAF3 affects caspase-1 activation or downstream events.
PYPAF3 Interacts with Both Procaspase-1 and Pro-IL-1␤-We next investigated whether there are physical interactions between PYPAF3 and caspase-1 or IL-1␤. These proteins were transiently expressed in 293T cells, and the cell lysates were subjected to immunoprecipitation analysis. Immunoprecipitation of caspase-1 or IL-1␤ but not RIP-DD resulted in the coprecipitation of PYPAF3 (Fig. 4, A and B, left panels). In addition, the immunoprecipitation of PYPAF3 but not Bcl-XL resulted in the coprecipitation of caspase-1 and IL-1␤ (Fig. 4, A  and B, right panels). These results indicate that PYPAF3 physically interacts with both caspase-1 and IL-1␤.
If caspase-1 or IL-1␤ was endogenously expressed in 293T cells, it was possible that caspase-1 or IL-1␤ indirectly interacted with PYPAF3 through the other one. Therefore, We examined the expression of caspase-1 and IL-1␤ mRNA in 293T cells. No RT-PCR product corresponding to caspase-1 cDNA was detected (see supplemental Fig. S2A), indicating that caspase-1 is not expressed in this cell line. On the other hand, we detected a faint band corresponding to IL-1␤ cDNA (see supplemental Fig. S2A). Then we investigated whether a significant amount of IL-1␤ is present in 293T cells. We could not detect any IL-1␤ protein in the culture supernatants or cell

FIG. 1. PYPAF2 inhibits ASC-mediated NF-B activation.
A-C, HEK293 cells were transfected with the indicated amounts (ng) of expression plasmids encoding ASC, FLAG-or HA-tagged PYPAFs, and/or a dominant negative mutant of TRADD (TRADD-DN) together with pNF-B-Luc and pRL-TK. Cells were cultured for 24 h before cell lysates were prepared. To induce NF-B activation by TNF-␣, cells were stimulated with 1 ng/ml TNF-␣ during the last 6 h of the culture. The expression of PYPAFs and/or ASC was visualized by Western blot analysis using anti-FLAG and anti-HA antibodies or an anti-ASC antibody (upper panels). The amount of ASC in the cell lysate decreased when it was coexpressed with PYPAF1, because ASC formed insoluble aggregates (14). NF-B activity was evaluated by luciferase assays (lower panels). Error bars represent the range of duplicate samples. Data are representative of at least three independent experiments. A, neither PYPAF2 nor PYPAF3 promoted NF-B activation in the presence of ASC. B, the expression of PYPAF2 but not PYPAF3 inhibited the NF-B activation induced by the simultaneous expression of ASC and PYPAF1dLRR. C, PYPAF2 did not inhibit TNF-␣-induced NF-B activation.

FIG. 2. PYPAF3 inhibits caspase-1-dependent IL-1␤ secretion.
A and B, HEK293 cells were transfected with the indicated amounts (ng) of expression plasmids encoding ASC and/or FLAG-or HA-tagged PYPAFs together with expression plasmids for pro-IL-1␤ and procaspase-1. After 26 h, culture supernatants were collected, and cell lysates were prepared. Cell lysates were subjected to Western blot analysis using anti-FLAG and anti-HA antibodies or an anti-ASC antibody (upper panels), while culture supernatants were subjected to ELISA for IL-1␤ (lower panels). C, HEK293 cells were transfected with the indicated amounts (ng) of expression plasmids encoding procaspase-1 and HA-tagged PYPAF2 or PYPAF3 together with an expression plasmid for pro-IL-1␤. Cell lysates and culture supernatants were prepared and analyzed as described above. A-C, error bars represent the range of duplicate samples. Data are representative of at least three independent experiments. lysates from 293T cells using ELISA irrespective of the presence or absence of exogenous caspase-1, unless IL-1␤ was exogenously expressed by transfection (see supplemental Fig.  S2B). Thus, the expression level of endogenous IL-1␤ protein in 293T cells was very low, if any. These results suggest that caspase-1 and IL-1␤ interacted with PYPAF3 independent of each other.
Expression of PYPAF2 and PYPAF3 mRNA-The expression of PYPAF2 and PYPAF3 mRNA in human tissues and cell lines was investigated by RT-PCR. Both PYPAF2 and PYPAF3 mRNA were detected in various tissues (Fig. 5A). However, skeletal muscle expressed neither PYPAF2 nor PYPAF3 mRNA, and heart and brain expressed low levels of PYPAF2 and no PYPAF3 mRNA. PYPAF2 and PYPAF3 mRNA were detected in many but not all tested cell lines, including hematopoietic and non-hematopoietic cell lines (Fig. 5B). However, the expression levels of these mRNAs in the cell lines were not well correlated with each other.

Lipopolysaccharide (LPS) and IL-1␤ Induce PYPAF3 Expression in Monocytic THP-1 Cells and Peripheral Blood
Mononuclear Cells-We next investigated whether inflammatory stimuli induce PYPAF3 expression in hematopoietic cell lines, using HL-60 promyelocytic leukemia, Jurkat acute T cell leukemia, U937 histiocytic lymphoma, and THP-1 acute monocytic leukemia cells. Interestingly, LPS potently enhanced the expression level of PYPAF3 but not PYPAF2 mRNA in THP-1 cells (Fig.  6A). Such induction was not observed in HL-60, Jurkat, or U937 cells. IL-1␤ also induced PYPAF3 mRNA expression in THP-1 cells in a stimulation time-and dose-dependent manner (Fig. 6B). Furthermore, LPS and IL-1␤ enhanced PYPAF3 mRNA expression in peripheral blood mononuclear cells as revealed by RT-PCR analyses (Fig. 6C). These results suggest that elevation of the IL-1␤ concentration results in the induction of PYPAF3, which has the potential to inhibit IL-1␤ production, in monocytes.
PYPAF-3 Inhibits LPS-induced IL-1␤ Secretion from THP-1 Cells-Finally, to obtain more insight into the biological significance of the ability of PYPAF3 to inhibit IL-1␤ secretion, we investigated whether PYPAF3 inhibits the LPS-induced endogenous IL-1␤ production in THP-1 cells. To this end, we estab- A, the tissue distribution of PYPAF2 and PYPAF3 mRNA was examined by PCR analysis using a panel of first-strand cDNA from normal human tissues. A sequence of glyceraldehyde-3phophate dehydrogenase (G3PDH) was amplified as an internal control. PBL, peripheral blood leukocytes. B, expression of PYPAF2 and PYPAF3 mRNA in the indicated human cell lines was examined by RT-PCR analysis using total RNA samples. A sequence of ␤-actin was amplified as an internal control. HeLa, epitheloid carcinoma; HepG2, hepatocellular carcinoma; ECV304, umbilical cord endothelial cell; PK-1, pancreatic cancer; SW480, colon adenocarcinoma; K562, chronic myelogenous leukemia; Jurkat, acute T cell leukemia; U937, histiocytic lymphoma; THP-1, acute monocytic leukemia. lished stable transfectants expressing N-terminal HA-tagged PYPAF3. Western blot analysis using anti-HA or anti-PYPAF3 antibodies failed to detect a protein corresponding to fulllength PYPAF3; however, we reproducibly detected a 35-kDa band with either anti-HA or anti-PYPAF3 antibodies specifically in PYPAF3-expressing but not in control transfectants ( Fig. 7A and data not shown). We established five independent clones of PYPAF3 transfectants, and all of them produced little or no IL-1␤ upon LPS stimulation (Fig. 7B, showing two representative clones), while control transfectants receiving empty vector produced a large amount of IL-1␤ under these conditions. These results indicate that PYPAF3 inhibits the endogenous IL-1␤ production in THP-1 cells. DISCUSSION In this study, we demonstrated that PYPAF3 inhibits caspase-1-dependent IL-1␤ processing, while PYPAF2 inhibits ASC-mediated NF-B activation. We recently found that a novel PYPAF member, which we named PYNOD, inhibits both of these activities (10). In addition, PYPAF4 was previously reported to inhibit the NF-B activation induced by TNF-␣ or IL-1␤ (15). Therefore, these members seem to constitute an anti-inflammatory subgroup of PYPAFs, in contrast to the proinflammatory members, which include PYPAF1, PYPAF5, PYPAF7, and NALP1. PYPAF3 interacted with both caspase-1 and IL-1␤ and inhibited the processing of both. At present, it is not clear whether PYPAF3 can inhibit the processing of IL-1␤ by already-activated caspase-1. On the other hand, PYPAF2, like PYNOD, inhibited the ASC-mediated but not TNF-␣-mediated NF-B activation. These results suggest that the molecular mechanisms by which PYPAF2 and PYNOD inhibit NF-B activation are different from that of PYPAF4, which directly binds to IKK␣ and inhibits NF-B activation induced by cytokines (15).
PYPAF2 and PYPAF3 are products of highly homologous genes that reside next to each other on chromosome 19q13.42 with a head-to-head orientation (7,8). In mouse, only one homolog corresponding to these two human genes has been found (16). Thus, it is likely that these two genes were generated as the result of a very recent gene duplication event.
Because PYNOD inhibits both caspase-1-dependent IL-1␤ processing and ASC-mediated NF-B activation, and separable regions are responsible for each of these functions, 2 it is possible that the common ancestor gene for PYPAF2 and PYPAF3 encoded a protein that possessed both of these activities, which might have been separately inherited by PYPAF2 and PYPAF3. In this context, it would be interesting to test whether the mouse protein representing these two human proteins has both functions. This structural similarity and functional difference between PYPAF2 and PYPAF3 may be useful for investigating the structure-function relationship of these proteins.
PYPAF2 and PYPAF3 mRNAs were broadly expressed in most tissues, with a few exceptions. Skeletal muscle expressed neither PYPAF2 nor PYPAF3, while heart and brain expressed PYPAF2 but not PYPAF3. Interestingly, PYNOD is mainly expressed in skeletal muscle, heart, and brain (10). Thus, it is likely that PYNOD and PYPAF2/PYPAF3 play similar functions in different tissues and/or different contexts. Despite the broad expression of PYPAF2 and PYPAF3 mRNA, their expression levels in a given tissue or cell line were not well correlated with each other. Furthermore, LPS stimulation induced the expression of PYPAF3 but not PYPAF2 mRNA in THP-1 cells. These results suggest that the expression of these mRNAs is differently regulated despite their genes being adjacent on the chromosome.
When we examined the N-terminal HA-tagged PYPAF3 protein expressed in stable transfectants of THP-1, we could detect only a 35-kDa fragment, which corresponds to the N-terminal fragment of PYPAF3 consisting of the PYRIN domain and the NAIP/CIITA/HET-E/TP1 (NACHT) domain, suggesting that PYPAF3 is quickly degraded in these cells. In addition, when we transfected HEK293 cells with a plasmid expressing the N-terminal portion of PYPAF3 corresponding to the p35 fragment together with expression plasmids for caspase-1 and pro-IL-1␤, it inhibited the IL-1␤ secretion significantly but less efficiently than a plasmid expressing full-length PYPAF3 (data not shown). We are now trying to determine the essential 2 Y. Wang and T. Suda, unpublished observation. Total RNA was prepared from the following cells, and the expression levels of PYPAF2 and PYPAF3 mRNA were examined by RT-PCR analysis. A sequence of ␤-actin was amplified as an internal control. A, HL-60 (promyelocytic leukemia), Jurkat, U937, and THP-1 cells were cultured with or without LPS (1 g/ml) for 6 h. B, THP-1 cells were treated with 10 ng/ml recombinant mouse IL-1␤ for the indicated period (upper panels) or treated with the indicated doses of mouse IL-1␤ for 6 h (lower panels). C, peripheral blood mononuclear cells were cultured with or without LPS (1 g/ml) or recombinant mouse IL-1␤ (5 ng/ml) for 14 h.

FIG. 7. PYPAF3 inhibits LPS-induced IL-1␤ secretion from THP-1 cells.
A, cell lysates were prepared from independent clones of THP-1-derived cells stably transfected with empty vector (vec #1 and vec #2) or an expression plasmid encoding HA-tagged PYPAF3 (PY3 #1 and PY3 #2). The cell lysates were subjected to Western blot analysis using an anti-HA antibody. B, the THP-1-derived stable transfectants described above were pretreated with phorbol myristate acetate (150 nM) for 3 h to induce differentiation into macrophage-like cells. The cells were cultured with or without 1 g/ml LPS for 12 h. Culture supernatants were then collected and subjected to ELISA for IL-1␤. region of PYPAF3 for its caspase-1-inhibitory activity.
Stimulation by LPS or IL-1␤ induced the expression of PYPAF3 mRNA in monocytic THP-1 cells. Furthermore, stable transfectants of this cell line expressing PYPAF3 were severely defective in the LPS-induced IL-1␤ production. Therefore, it is likely that the PYPAF3 is a feedback regulator for IL-1␤ secretion. Considering that an exaggerated production of IL-1␤ is highly toxic to animals, such a feedback regulation mechanism should be very important under pathophysiological conditions. ICEBERG, COP, and CARD8 are CARD-containing proteins that have been reported to interact with caspase-1 and inhibit caspase-1-dependent IL-1␤ secretion (17)(18)(19). However, the genes for these proteins have been found in humans but not in mice (16). In contrast, PYPAF3 and PYNOD have counterparts in mice (10,16). Therefore, these proteins may play fundamental roles as caspase-1 inhibitors in mammals. This notion should be investigated by gene-targeting experiments in the future.