Isolation of a Novel Interleukin-1-inducible Nuclear Protein Bearing Ankyrin-repeat Motifs*

We isolated a novel gene termed interleukin (IL)-1-in-ducible nuclear ankyrin-repeat protein (INAP), of which expression was specifically induced by IL-1 in OP9 stromal cells. The INAP has ankyrin-repeat motifs and shares weak amino acid sequence homology with Bcl-3 and other I k B family members. The human genomic INAP gene found in the NCBI data base is located at chromosome 3q3.11. Northern blot analyses revealed that INAP was not expressed in any examined tissues without stimulation, but INAP expression was rapidly and transiently induced by IL-1 although not by tumor necrosis factor a nor by phorbol 12-myristate 13-acetate in OP9 cells. Im-munoblots with anti-INAP-specific antibody demonstrated that INAP was rapidly and specifically produced by IL-1 stimulation and was predominantly localized in the nucleus. Immunofluorescence stainings showed that the INAP newly synthesized by IL-1 stimulation was promptly translocated into the nucleus, and FLAG-tagged INAP forcibly expressed in NIH/3T3 cells was also specifically localized in the nucleus. The possible interaction of INAP with RelA/p65, NF- k B1/p50, NF- k B2/p52, C/EBP b , and retinoid X receptor was examined, but we could detect none of these interactions in the nuclear extracts of IL-1-stimulated cells. Unlike Bcl-3 and other I k B family

IB␣ is degraded in response to the NF-B inducers TNF␣, IL-1, lipopolysaccharide (LPS), phorbol 12-myristate 13-acetate (PMA), and double-stranded RNA. In contrast, IB␤ is degraded only when cells are stimulated with IL-1 or LPS, both of which cause persistent long term activation of NF-B (4 -7). Following degradation of the initial pool of IB␤ in response to IL-1 or LPS, newly synthesized IB␤ accumulates as an unphosphorylated protein that forms a stable complex with NF-B and prevents it from binding to newly synthesized IB␣ (4 -7), resulting in the prolonged activation of NF-B (4,8). This unphosphorylated IB␤ cannot block the nuclear localization signal of NF-B, thus this NF-B⅐IB␤ complex translocates into the nucleus. The function of this complex in the nucleus is yet to be elucidated, and the mechanism by which only IL-1 and LPS can degrade IB␤ remains to be resolved.
Here we identified a novel IB family member, termed IL-1inducible nuclear ankyrin-repeat protein (INAP), of which expression is specifically induced by IL-1. INAP was found to be weakly homologous to Bcl-3 and localized in the nucleus like Bcl-3. We discuss here the possible function of this novel IB family member.

Isolation of INAP cDNA-OP9 cells were cultured in ␣-minimum
Eagle's medium supplemented with 20% fetal calf serum (FCS) with or without 10 ng/ml of mouse IL-1␣ (Genzyme/Techne) in the presence of 10 g/ml of cycloheximide for 1 h, and total mRNA was isolated. The PCR-Select cDNA subtraction kit (CLONTECH) was used for cDNA synthesis and suppressive subtractive hybridization, according to the manufacturer's instructions. The cDNA from OP9 cells incubated with IL-1␣ was used as the tester sample, and that from untreated cells was used as the driver sample. The 5Ј end of INAP cDNA was confirmed by the rapid amplification of the cDNA ends (5Ј RACE) method.
Northern Blot Analysis-Total mRNAs from OP9 cells stimulated * 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.

RESULTS
Isolation and Structure of INAP-We attempted to isolate the genes of which transcriptions were induced by IL-1 in mouse stromal OP9 cells. Subtraction of mRNAs of OP9 cells before and after IL-1 stimulation led to the isolation of a number of cDNA fragments encoding factors related to the IL-1 response. Among them, we focused on a novel gene, termed INAP. The isolated mouse full-length INAP cDNA (2339 bp) contained an open reading frame, which encoded a polypeptide of 728 amino acids with a calculated molecular mass of 79,007 daltons and a predicted pI of 6.45. A search in the NCBI data base using the BLAST program revealed that a human genomic sequence (map element NT_022504) of chromosome 3 contained the human INAP gene, which consists of 12 exons and is located at chromosome 3q13.11. Human INAP constituted of 718 amino acids has a predicted molecular mass of 78,061 daltons. Human INAP has 82% amino acid identity and 85% similarity with mouse INAP and has a 10-amino acid deletion at amino acid 301-310 compared with mouse sequences. Fig. 1A shows the schematic drawings of the isolated mouse INAP cDNA and of the human genomic INAP gene found in the NCBI data base. Mouse and human INAP were found to be weakly homologous to the IB family and the Rel family. The most striking feature of mouse and human INAP is that both INAP contain five highly conserved ankyrin-repeat motifs in carboxyl-terminal regions. Using PESTFIND software, it was also found that both INAP have PEST (P, E, D, S, and T residue-rich)-like sequences, which are implicated in the rapid turnover of proteins (19). Although IB␣ and IB␤ have PEST sequences in carboxyl-terminal ends, both mouse and human INAP have them at amino-terminal regions (amino acids 11-84 and 185-203 in mouse and 184 -204 in human). A serine-rich region was also found in amino-terminal regions (amino acids 53-77 in mouse and 51-86 in human), and a glutamine-rich region was found in the middle regions (amino acids 245-308 in mouse and 247-299 in human). However, the glycine-rich region, Rel homology domain, or obvious nuclear localization signal, which commonly exist in the Rel family, was not found in INAP. Reinhardt's method (20) for cytoplasmic or nuclear discrimination predicted that there is a 94% possibility that INAP is localized in the nucleus.
To examine the relationship of INAP to other members of the IB and Rel families, a phylogenetic tree was constructed using the amino acid sequences of all known mammalian IB and Rel families (Fig. 1B)  IL-1-inducible Nuclear Protein with Ankyrin Repeats 12486 stimulation (Fig. 2, right panel). Although no INAP mRNA was detected in unstimulated OP9 cells (Fig. 2, right panel, lane 1), a single hybridized band was weakly detected 15 min after IL-1␣ stimulation (lane 2). The level of INAP mRNA increased and reached the maximum at 1 h after IL-1␣ stimulation (lane 4) and then decreased thereafter (lanes 5, 6), indicating that the INAP gene was rapidly and transiently transcribed after IL-1␣ stimulation in OP9 cells. IL-1␤ also exhibited the same effect on INAP expression (data not shown). Similarly, INAP expression after TNF␣ or PMA stimulation was examined, but no transcript was detected (Fig. 2, left and middle panels) although OP9 cells are responsive to TNF␣ and PMA, indicating that INAP transcription was specifically induced by IL-1 stimulation.
The expression of mouse INAP in the protein level was examined by immunoblot analyses with a purified anti-INAPspecific rabbit antibody. As shown in Fig. 3A, mouse INAP of 79 and 82 kDa were clearly detected in IL-1␣ (right panel)-but not TNF␣ (left panel)-or PMA (middle panel)-stimulated OP9 cells, confirming that INAP expression was specifically induced by IL-1␣. The double bands were clearly detected within 30 min after IL-1␣ stimulation (Fig. 3A, right panel, lane 2), although they were not seen in the cells without stimulation (lane 1). Although INAP transcripts were transiently expressed, the protein level increased until 1 h after IL-1␣ stimulation and retained its level even 24 h after stimulation (lanes 3-5), suggesting that the newly synthesized INAP is relatively stable and accumulates in the cells. We also detected human INAP in IL-1␣-stimulated HeLa cell extracts by the same antibody (data not shown).
The double bands were clearly recognized by anti-INAP antibody, and thus we speculated that the upper band might be the phosphorylated form of the lower band. The INAP was therefore immunoprecipitated with its specific antibody from the nuclear extracts, as well as the cytosolic extracts, which were prepared from OP9 cells treated with or without IL-1␣. The immunoprecipitates were treated with calf intestine alkaline phosphatase (CIP), separated by SDS-PAGE, and immu-noblotted with anti-INAP antibody. The results demonstrated that INAP was predominantly detected in the nuclear extracts and that the double bands were not affected by phosphatase treatment (Fig. 3B), indicating that INAP was not phosphorylated and that subcellular localization of INAP was not affected by its phosphorylation status. We therefore concluded that the upper band was not the phosphorylated form of the lower band. We further noticed that the INAP has a second Met codon at amino acid 26, and thus the lower band may be the protein product translated from this second Met codon.
INAP Is Rapidly Translocated into the Nucleus-Subcellular localization of INAP in OP9 cells with or without IL-1␣ stimulation for 1 h was examined by indirect immunofluorescence microscopic analysis, and the fluorescent images were overlaid on difference interference contrast images (Fig. 4A). The INAP stained with purified anti-INAP-specific antibody in red were predominantly detected in the nuclei of OP9 cells treated with IL-1␣ for 1 h (Fig. 4A, left lower panel), whereas it was rarely seen in the cells prior to IL-1␣ stimulation (left upper panel). The chromosomes stained with DAPI in blue (right lower panel) were completely overlapped with INAP stained in red (left lower panel). These results clearly demonstrated that newly synthesized INAP was promptly translocated into the nucleus by IL-1␣ stimulation in OP9 cells. Furthermore, we found that IL-1 stimulated the production of INAP in various mouse organs including spleen, small intestine, lung, liver, heart, and kidney and that INAP was always localized in nucleus in these IL-1-stimulated tissues (data not shown). Thus, IL-1-specific INAP expression and its nuclear localization are not specific events observed only in OP9 cells.
To confirm these findings, FLAG-tagged INAP and HAtagged INAP were transiently expressed in mouse fibroblast NIH/3T3 cells. Fig. 4B shows that FLAG-tagged INAP was clearly detected only in the nucleus of the transfected cells by anti-FLAG antibody (stained in green; left panel). The chromosomes stained with DAPI in blue (right panel) were completely overlapped with INAP stained in green (left panel) in transfectants. Similarly, HA-tagged INAP was also localized in the nucleus (data not shown). Taken together, these results clearly indicate that INAP was promptly translocated into the nucleus after INAP protein synthesis was induced by IL-1 stimulation. DISCUSSION We isolated a novel IL-1-inducible nuclear factor, INAP, which is related to the IB family and the Rel family. It is well known that IB family members bind to the RelA/p65⅐NF-B1/ p50 complex and prevents the complex from activating and translocating into the nucleus. Therefore, possible interaction of INAP with the RelA/p65⅐NF-B1/p50 complex was examined by immunoprecipitation followed by immunoblot analysis. However, we failed to detect the direct and/or indirect binding of INAP to RelA/p65 (data not shown). Furthermore, one of the most important factors induced by IL-1 stimulation is IL6, of which gene expression is regulated by C/EBP␤ (NF-IL6), AP-1, and NF-B (21,22). Therefore, we also examined the possible interactions of INAP with C/EBP␤ and c-Fos/c-Jun in nuclear extracts prepared from IL-1␣-stimulated OP9 cells. Once again, we could not detect the interactions of INAP with C/EBP␤ or with c-Fos/c-Jun (data not shown). Moreover, the fact that Bcl-3, the protein most closely related to INAP, associates NF-B1/p50 or NF-B2/p52 homodimers and modulates their transactivation activities (11,16) motivated us to examine whether INAP associates with NF-B1/p50 or NF-B2/p52 in IL-1-stimulated nucleus. None of these interactions, however, was detected by immunoprecipitation followed by immunoblot analysis (data not shown). It has also been reported that Bcl-3 binds to RXR (17) or to AP-1 (18) and regulate specific gene expression, and thus we also examined the possible interactions of INAP with RXR or with AP-1 but failed to detect the bindings (data not shown). We concluded that INAP could not bind to any of the binding partners with which Bcl-3 has been reported to interact and that INAP is a very unique protein in the IB family and is clearly distinct even from the most closely related IB family member, Bcl-3. To determine the biological function of INAP on IL-1 signalings it is very important to identify the INAP-binding proteins in the IL-1-stimulated nucleus by other means such as yeast two-hybrid screening, pulldown experiments, and far-Western screening.
INAP was found to be a novel nuclear factor related to the IB family, but by IL-1 stimulation INAP was newly produced and accumulated in the nucleus, rather than being degraded as were other IB family members. From a gene expression point of view, INAP is quite a distinct protein from these family members. There exists no obvious nuclear localization signal in INAP, but it does exist in the nucleus. Furthermore, INAP was not phosphorylated no matter whether it was localized in the cytoplasm or nucleus, indicating that the phosphorylation status of INAP does not affect its subcellular localization.
A few potential NF-B binding sites were found around Ϫ340 bp upstream from the initiation codon in human INAP gene promoter regions. However, we demonstrated that INAP gene expression was rapidly induced by IL-1 but not by TNF␣ nor by PMA, all of which are known to activate the NF-B signaling pathway. Thus, INAP gene expression is not simply regulated by NF-B signaling. IL-1-and LPS-specific persistent activation of NF-B has been reported (4, 5), but rapid IL-1-(and LPS-) specific INAP gene expression cannot be explained by this mechanism. The mechanism is thus obscure at this moment, and further analyses are required.