Introduction

The members of the C/EBP family are nuclear transcription factors implicated in a variety of cell differentiation processes (Scott et al., 1992; Ness et al., 1993; Zhang et al., 1997; Lekstrom-Himes and Xanthopoulos, 1998). C/EBP- in particular is essential for normal granulocytic differentiation (Antonson et al., 1996; Chumakov et al., 1997; Morosetti et al., 1997), as mice with genetic disruption of the C/EBP- gene lack mature granulocytes expressing secondary granule proteins. C/EBP--deficient mice develop normally, except that they fail to produce functional neutrophils and eosinophils. Neutrophils from these C/EBP--/- mice have impaired chemotaxis, bactericidal activity and typically die of infections between 3 and 5 months of age (Yamanaka et al., 1997a, 1997b). The lack of secondary granule proteins in granulocytes from these mice results in the impairment of responses to inflammatory signals (Masera et al., 1996; Lekstrom-Himes et al., 1999). We used C/EBP- knockout mice to identify myelomonocytic genes regulated by C/EBP- (Kubota et al., 2000). Polymerase chain reaction (PCR)-based subtractive hybridization method RDA (representational difference analysis) was used to find genes that are not expressed in neutrophils and macrophages of C/EBP- knockout mice. Approximately 10% of clones identified using this technique were derived from the same novel gene, encoding a putative cysteine-rich protein named mXCP1 (for murine ten-cysteine protein 1). Here, we report cloning, chromosomal localization and identification of mXCP1 and its human homologue, hXCP1, as C/EBP--dependent members of the XCP/FIZZ/Resistin gene family. We further report that XCP1 directly interacts with human neutrophil -defensin and plays a role in myeloid cell chemotaxis.

Results

We have previously used a PCR-driven RDA assay to detect genes that are directly regulated by C/EBP- (Kubota et al., 2000). Briefly, cDNA derived from the myeloid cells (peritoneal exudate following intraperitoneal (i.p.) administration of thioglycolate (TG)) of the wild-type mice was subtracted from identically prepared cDNA of C/EBP--knockout mice. After several subtraction rounds, individual cDNA clones were obtained, sequenced and analysed by searching GenBank and EST databases. Out of 174 sequenced clones, 17 of them did not match previously known genes and were identified as encoding segments of a novel murine protein, mXCP1.

Comparison of putative translation reading frames of mXCP1 cDNA fragments with the nonredundant (nr) GenBank-EMBL database has yielded no match. However, a search of the dynamically translated EST database suggested the presence of several novel related genes, mXCP2, 3, 4, as well as at least two human (hXCP1 and hXCP2), rat (rXCP2) and Xenopus (xXCP1) homologues. These genes encode a family of short cysteine-rich proteins (synthesized as 105–138 amino-acid precursors) that share a characteristic 'signature' in their primary sequence (Figure 1a). The mXCP2, mXCP3 and mXCP4 were subsequently identified as FIZZ1, FIZZ2 and FIZZ3, respectively (Holcomb et al., 2000); mXCP4/FIZZ3 was also discovered independently by another group as the hormone resistin (Steppan et al., 2001a, 2001b).

Figure 1.

(a) Alignment of murine and human XCP-FIZZ-resistin family members. Putative N-terminal transmembrane/signal sequences are shaded. Consensus XCP-FIZZ family signature is shown at the bottom. (b) Sequence of the murine XCP1 gene. The exon structure was determined by comparing the cDNA and genomic sequences of mXCP1 and primer extension analysis. The position of the four exons of mXCP1 gene is underlined. The mRNA polyadenyation signal AATAAA, ATG start codon and the TAA stop codon are shown in bold

Full figure and legend (315K)

Identification of full-length mXCP1 cDNA

We have used sequence information obtained by analysis of RDA-derived short cDNA fragments of mXCP1 to devise primers to clone a full-length cDNA by a modified 5'RACE. The total cDNA from myeloid cells of a wild-type mouse was reverse-transcribed using an antisense mXCP1 primer. In order to enrich full-length cDNAs, Clontech 'CAP-finder' SMARTII oligonucleotide was then used as a 5'-specific oligonucleotide in PCR amplification of mXCP1-specific cDNA. The resulting amplification products were sized by agarose gel electrophoresis, cloned and sequenced. MXCP1 encodes a putative 117-residue precursor protein, containing a characteristic ten-cysteine (XCP) pattern in its C-terminal domain, and a 20-amino-acid-residue transmembrane/signal sequence at its N-terminus. Computer analysis of putative translation products of other XCP genes shows that they share this structure with mXCP1 (Figure 1a).

Identification of mXCP genes and chromosomal localization of XCP genes

A 550-bp cDNA fragment encoding the full-length mXCP1 was used as a probe to screen a murine genomic DASHII library (Clontech). Restriction enzyme mapping indicated that positive clones span a region of approximately 50 kb, and contain mXCP1, mXCP2 and mXCP3 genes. We found that mXCP1 and mXCP2 genes are arranged in a head-to-tail orientation less than 25 kb apart from each other. Human XCP-related clones were isolated from the human genomic phage library (Clontech). The sequencing of mXCP1, mXCP2, mXCP3 and their human analogues hXCP1 and hXCP2 was performed (Figure 1b, data for all XCP genes deposited in GenBank under accession numbers BE847000, AF510097, AF510098, AF510099, AF510100, AF352730, AF352731). Comparison of these genomic sequences to available cDNA ESTs and cDNA clones obtained by PCR has revealed that each gene spans approximately 2 kb. Murine genes and hXCP1 are comprised of four exons, while human XCP2 lacks the most 5' nontranslated exon.

Research Genetics mouse–hamster radiation hybrid (RH) library was used to map mXCP1, mXCP2, mXCP3 and mXCP4 genes. The mXCP1, mXCP12 and mXCP13 genes were found to form a cluster on murine chromosome 16, while mXCP4 is located on murine chromosome 8. Human genes encoding hXCP1 and hXCP2 were identified in the HTGS database as mapping to human chromosomes 3 and 19, respectively. Specifically, the best-fit location for mXCP1 and mXCP2 is 49.6 cR proximal to D16Mit15 and 13.5 cR distal to D16Mit85. The mXCP3 gene was found at a location just slightly proximal to mXCP1 and mXCP2 21.9 cR proximal to D16Mit15 and 36.2 cR distal to D16Mit85. The mXCP4 is located at chromosome 8 between D8Mit155 and D8Mit149.

Tissue specificity of mXCP expression

We have analysed the expression pattern of the murine XCP genes in wild-type and C/EBP- -/- (knockout) mice. In wild-type mice (Figure 2), the mXCP1 gene was expressed mainly in the bone marrow, spleen and, to some extent, colon (Figure 2, lanes 3, 10 and 5, respectively). The mXCP2 was expressed predominantly in the pancreas, tongue and skin (Figure 2, lanes 8, 9 and 1, respectively); mXCP3 expression was strictly limited to colon (with or without mucosa) and small intestine (Figure 2, lanes 5, 6, 7 and 11), while mXCP4 was strongly expressed in white adipose tissue (Figure 2, lane 17). The expression of murine C/EBP- (Figure 2, lanes 3, 5 and 10, top row) showed striking coincidence with that of mXCP1. We have also analysed mXCP1 expression in several murine tissues at different stages of development (Figure 3a). MXCP1 was strongly expressed in embryonic liver between day 18 postconception and at birth (lanes 4 and 3, respectively), after which time expression rapidly disappeared in this organ. The gene was also strongly expressed in the neonatal pancreas (lane 23), decreasing approximately 10-fold within the next 4 days; neonatal gut (lane 30), lung (lane 13) and heart (lane 8) also had transient expression. XCP1 expression was also sometimes observed in lungs of adult mice (Figure 3a, lane 11); however, this was not always the case and may be the result of inflammation (data not shown).

Figure 2.

Northern blot analysis of murine XCP gene expression: mXCP1 expression is coincident with that of C/EBP- in adult murine tissues. Northern blot hybridization was carried out using 20 g of total cellular RNA per lane. Gel loading was normalized by EtBr stained ribosomal RNA and hybridization with -actin specific probe (not shown). Following electrophoresis, RNA was blotted onto a nylon membrane and sequentially hybridized with EasyStrip-labeled cDNA probes representing murine XCP genes and C/EBP-. After each consecutive hybridization under strict conditions, washing and exposure to Kodak X-ray film, the probe was removed according to the kit manufacturer's instructions and the blots were exposed again for 12 h at -70°C to verify the complete removal of the radioactive probe

Full figure and legend (135K)

Figure 3.

(a) Changes in the expression of mXCP1 mRNA in different organs during murine development. Total RNA (20 g per lane) was separated on a 1.25% agarose gel in the presence of formaldehyde and subjected to Northern blotting. mXCP1 expression was maximal in murine neonatal pancreas (lanes 21–24) and liver (lanes 1–5) at 1 day after birth (lanes 23 and 3, respectively). RNA was prepared from murine organs. The age of mice used is shown as number of days after conception (postconception, pc) or days after birth (postpartum, ppt). (b) Northern blot analysis of mXCP1 and mXCP2 gene expression in tissues of wild-type (lanes 2, 4, 6 and 8) and C/EBP- knockout mice (lanes 1, 3, 5 and 7). Northern blot analysis was performed as described in panel a. Following hybridization and exposure, the mXCP1 EasyStrip-labeled probe was removed and the membrane was rehybridized with the mXCP2 probe. (c) Expression of C/EBP- results in the induction of the expression of the endogenous mXCP1 gene in transfected NIH3T3 cells and the bone marrow cells of C/EBP--deficient mice. NIH3T3 cells stably transfected with C/EBP- and C/EBP- zinc-inducible expression plasmids, as well as an empty pMTCB6 vector were kindly provided by Dr A Gombart. These cells were either treated for 6 h with 100 M of ZnSO₄ (lanes 2, 4 and 6) or left untreated (lanes 1, 3 and 5). C/EBP- knockout murine bone marrow cells were electroporated with 20 g per 10⁷ cells of the following expression plasmids: C/EBP-, lane 7; C/EBP-, lane 8; pcDNA3 empty expression vector, lane 9. Following transfection, cells were returned to culture media and grown for 16–24 h prior to RNA extraction. The mRNA was amplified using RT–PCR. Each PCR reaction included primers specific for the mXCP1 gene and primers for -actin as an internal control. PCR conditions yielded amplification in the linear range. PCR products were analysed by electrophoresis in 1.5% agarose and visualized by ethidium bromide staining. The specificity of mXCP1 cDNA amplification was also confirmed by blotting and hybridization using the mXCP1-specific cDNA probes and autoradiography (not shown)

Full figure and legend (125K)

XCP1 is a strictly C/EBP--dependent gene

Our previous experiments (Kubota et al., 2000, also unpublished observations) indicated that both mXCP1 and mXCP2 were strongly expressed in peritoneal exudates derived from wild-type mice following TG injection and were completely absent from exudates of similarly treated C/EBP- -/- mice. To confirm that mXCP1 expression is indeed regulated by C/EBP-, we have analysed mRNA derived from the spleen, bone marrow, colon and skin of wild-type and C/EBP- knockout mice following TG administration using Northern blotting and hybridization with mXCP1- and mXCP2-specific probes (Figure 3b). Several other genes, including the household -actin gene, were analysed in parallel and were used as controls (not shown). No expression of mXCP1 was found in any tissue derived from C/EBP- knockout mice, while its expression was quite robust in myeloid cells of the wild-type mice. The XCP2 expression was lacking in the skin of C/EBP- -/- mice (Figure 3b, compare lanes 7 and 8), but expression levels of this gene were unaffected in the colon (lanes 5 and 6) as well as in the pancreas and tongue of these mice (not shown). The mXCP2 expression thus appears to be dependent on C/EBP- in some tissues (skin), but not in others (colon, pancreas and tongue). In addition, peritoneal TG administration appears to induce mXCP2 expression strongly in the colon of wild-type and C/EBP- -/- mice (compare Figure 2a, lanes 5–7 and Figure 3b, lanes 5 and 6), again suggesting a C/EBP--independent mechanism of mXCP2 expression in the colon.

Since the mXCP1 expression pattern showed striking similarity to that of C/EBP- in wild-type mice and was lacking in C/EBP- -/- mice, we used C/EBP- expression vectors in an attempt to induce endogenous mXCP1 expression in murine cells lacking C/EBP- (Figure 3c). We studied mXCP1 induction in NIH3T3 cells that were stably transfected with either C/EBP-, C/EBP--inducible expression constructs or an empty vector pMT (cell lines provided by Dr A Fritz Gombart, Cedars-Sinai Research Institute) by a sensitive RT–PCR assay, with the amplification of -actin cDNA serving as an internal control. The induction of the C/EBP- gene following addition of zinc sulfate resulted in the specific expression of mXCP1 mRNA (Figure 3c, lanes 5 and 6). No expression was detected upon induction of either C/EBP- (lane 4) or the addition of inducer to cells containing an empty vector (lane 2). Likewise, we were able to stimulate endogenous mXCP1 gene expression in bone marrow-derived cells of C/EBP- -/- mice (Figure 3c, lane 7).

C/EBP-dependent expression of human XCP1

In wild-type mice, the expression of mXCP1 closely coincided with the expression of C/EBP-, prompting us to search for a possible regulatory homologue in human cells. We prepared oligonucleotide probes for hXCP1 and hXCP2 genes that we identified previously as possible homologues of mXCP1 and obtained full-length cDNA clones by RT–PCR. The cDNA probes were then radiolabeled and hybridized with Clontech multiple-tissue (MTN) blots. The hXCP1 was strongly expressed only in normal human bone marrow, with low-level expression in the lungs (Figure 4a, lanes 17 and 4, respectively), while hXCP2 expression was restricted to the colon and small intestine (Figure 4a, lanes 10 and 11). The expression pattern of hXCP1 was thus identical to that of C/EBP- in normal human tissues, while the expression of hXCP2 closely resembled that of murine XCP3/FIZZ2 gene. C/EBP- is strongly upregulated following treatment with 9-cis retinoic acid (RA) in NB4 cells (Chumakov et al., 1997; Park et al., 1999). We used an hXCP1 probe to analyse mRNA of NB4 cells prior to and after RA treatment and in several other cell lines with a known expression pattern of C/EBP- (Figure 4b). Human XCP1 was strongly induced in NB4 promyelocytic cells following exposure to RA (lanes 6 and 7). The hXCP1 was also strongly expressed in the Kcl22 myeloblastic cell line expressing high levels of C/EBP-. The hXCP1 was not induced by RA in U937 myelomonoblastic cells (lanes 9 and 10). As shown before (Poli et al., 1990), C/EBP- gene is not inducible by RA in U937 cells and expression patterns of hXCP1 and C/EBP- are identical.

Figure 4.

(a) Expression of hXCP1 and hXCP2 genes in normal human tissues. The expression of human XCP genes was determined by Northern hybridization with specific cDNA probes. Human multiple tissue blots (MTN, MTNII and MTNIII; Clontech) containing 2 g poly(A)⁺ RNA from each tissue were sequentially hybridized and washed under conditions of high stringency according to the protocol recommended by the manufacturer, in ExpressHyb (Clontech) hybridization solution. Blots were exposed for 4 days at -70°C. (b) Expression of hXCP1 closely parallels the expression of C/EBP- in human cell lines. Northern blot analysis was carried out with 30 g/lane of total RNA of various myeloid and nonmyeloid cell lines. U937PR-9 myelomonoblastic cells induced to express PML/RAR fusion gene were treated with 9-cis RA (10^-8 mol/l, lanes 2 and 4; 10^-6 mol/l, lanes 3 and 5). NB4 promyleocytic and U937 cell lines were treated with 10^-8 mol/l of 9-cis RA or left untreated (compare lanes 6 and 7). Myeloblastic Kcl22 cells (lane 8) were not treated with RA, since they already expressed C/EBP- at maximal levels. Cells were harvested at 4 and 24 h. The top panel shows the hybridization with the full-length hXCP1 cDNA probe. The middle panel shows hybridization with the hC/EBP- cDNA probe, and the bottom panel shows the hybridization with the -actin probe as assessment of RNA quantities on each lane. (c) Expression of C/EBP- induces endogenous hXCP1 gene expression in transfected human HEK293 cells. Human epithelial HEK293 cells were transiently transfected with either C/EBP-, c-Myb, C/EBP- or ATF4 expression plasmids. At 24 h after transfection, the expression of endogenous hXCP1 gene was detected by Northern blotting and hybridization with the hXCP1-specific probe. Untransfected HEK293 cells or HEK293 cells transfected with an empty expression vector DNA (first and last lanes, respectively) served as negative controls

Full figure and legend (181K)

Since hXCP1 appears to be regulated in parallel with C/EBP- in these cell lines, we hypothesized the direct regulation of this gene by C/EBP- in a manner similar to murine mXCP1, making hXCP1 a likely functional homologue of mXCP1. To test this hypothesis, we transfected human embryonic kidney 293 (HEK293) cells with expression constructs for C/EBP-, c-Myb, C/EBP- or ATF4 (Figure 4c) and used Northern blotting and hybridization with an hXCP1-specific probe to detect the induced expression of hXCP1. As predicted, the expression of C/EBP- directed by a strong CMV promoter in HEK293 cells resulted in the induction of hXCP1 (lane 2). Cotransfection of C/EBP- with synergistically acting c-Myb and ATF4 expression constructs (Gombart A.F., personal communication) resulted in an even greater induction of hXCP1 (lane 6), while the transfection of these genes separately did not induce hXCP1 (lanes 3, 4 and 5, respectively). The expression of hXCP1 was undetectable in either mock-transfected HEK293 cells or in HEK293 cells transfected with an empty pcDNA3.1 vector (Figure 4c, lanes 1 and 7, respectively). These data suggest that regulation of the hXCP1 gene is homologous to that of murine XCP1.

The only other human member of the XCP/FIZZ/Resistin gene family found in the human genome is hXCP2, which is colon-specific and thus represents a homologue of mXCP3/FIZZ2. The expression of hXCP2 gene was not detected in human adipose tissue (data not shown); however, chromosomal localization of hXCP2 on human chromosome 19, homologous to murine chromosome 8 (where we have localized the mXCP4/FIZZ3/Resistin gene) suggested similarity to mXCP4/Resistin.

Cellular localization of mXCP1

A number of structurally similar cysteine-bridged proteins, such as defensins and trefoil factors, are active upon secretion and processing, which preserves only the cysteine-bridged C-terminal domain (Selsted et al., 1992; Ganz, 1999; Taupin et al., 2000). Our computer analysis of the XCP protein sequence has identified the presence of a 20-amino-acid N-terminal transmembrane region. However, no putative processing cleavage sites have been found downstream, leaving it unclear as to whether the XCP proteins remain attached to the membrane through an N-terminal 'anchor', or whether they are enzymatically cleaved and secreted either extracellularly, or into the granule compartment. We analysed transfected cells expressing mXCP1 from a CMV-promoter-driven pcDNA3.1 vector by immunoprecipitation and then Western blotting using anti-mXCP1 antipeptide antibodies. As shown in Figure 5, mXCP1 is secreted into the media by the transfected COS-1 cells, and this is accompanied by processing that reduces the mXCP1 protein size from approximately 13 kDa to approximately 9 kDa (Figure 5, compare lanes 2 and 3). Normal murine bone marrow was also found to contain mXCP1 protein species in the range between 8 and 10 kDa (lane 4).

Figure 5.

Western blotting analysis of mXCP1 localization using rabbit anti-mXCP1 antipeptide antibodies. COS-1 cells were transiently transfected with 5 g of either mXCP1 expression plasmids (lanes 2 and 3) or an empty expression vector (lane 1). Cell media and cell pellets were collected 24 h after transfection and analysed by immunoprecipitation with anti-XCP1 antibodies (1 g) and 10–20% gradient SDS–PAGE. Each lane represents either 1 ml of cell culture media or an equivalent amount of transfected COS1 cell pellet (5 10⁴ cells) (lanes 1–3) or 5 10⁴ of normal murine bone marrow cells (lane 4). Arrows on the right-hand side of the figure indicate the position of standard protein size markers

Full figure and legend (92K)

Identification of XCP1 activity

Since the majority of genes identified in the course of our original RDA subtraction experiments were either chemokines or chemokine receptors (Kubota et al., 2000), perhaps mXCP1 also has a role in cell migration, activation and chemotaxis. To test this hypothesis, we used a Boyden cell migration chamber to compare cell motility either in the presence or absence of XCP1 in the media. To achieve this, murine and human XCP1 expression constructs were transfected into COS-1 cells. The expression of XCP1 in transfected COS-1-conditioned media was confirmed by immunoblotting using rabbit antipeptide XCP1-specific antibodies (data not shown). We then tested the effect of XCP1 protein expressed by transfected COS1 cells on cell migration and chemotaxis of bone marrow cells of wild-type and C/EBP- -/- mice. The results of this assay are summarized in Table 1. On average, the presence of murine or human XCP1 in the culture media resulted in a four- to fivefold increase in the number of migrating bone marrow-derived myeloid cells.

Table 1 - Chemotactic response of murine bone marrow cells to XCP1-containing conditioned media.

Full table

Members of the XCP family of proteins are structurally similar to several groups of short proteins with diverse activities. These are defensins, -defensins and trefoil proteins that are predominantly implicated as having antimicrobial activities. In our initial observation, glutathione S-transferase (GST) - mXCP1 fusion protein was extremely toxic to bacteria, as E. coli growth ceased immediately upon addition of the IPTG inducer. However, a series of tests for microbicidal activity using a multifold dilutional assay in liquid media against a broad range of agents at physiological salt concentration were negative when we used the purified GST-mXCP1 fusion protein, (His)₆-tagged mXCP1 or COS1-derived mXCP1 (data not shown). We were also unable to detect any significant effect of purified mXCP1 or anti-mXCP1 antibodies on the colony formation of murine bone marrow-derived hematopoietic stem cells grown in a variety of cytokines in soft agar (data not shown).

Identification of XCP1-interacting proteins

We attempted to identify possible XCP1 receptors and/or interacting proteins using a yeast two-hybrid system. We cloned sequences encoding murine and human XCP1 proteins (lacking the signal sequences) into yeast pGBT8 'bait' vector and used Gal4DB-XCP1-fusion-expressing clones to screen a pretransformed Matchmaker (Clontech) human leukocyte cDNA library. High stringency screening and sequencing of human cDNA found 72 strongly interacting clones, of which 12 clones encoded human neutrophil -defensin (at least eight clones out of 12 were independently derived). Other clones included fatty-acid coenzyme A ligase (FACL2) and several fragments of genes containing immunoglobulin (Ig)-like domains (MYBC, Ig light chain, IgH, IgHA1), each of these represented by at least two independent clones. Other positive clones included MRP-8, calgranulin, APK antigen, palmitoylated 55 kDa erythroid membrane protein, COX7C, CD68, RP42 (NIF) and lysozyme. Defensin-containing clones were found in independent screenings using both the murine and human XCP1 as 'baits'.

To verify the specificity of a physical interaction between mXCP1, hXCP1 and -defensin, we carried out an in vitro pulldown assay using [³⁵S]methionine-labeled XCP1 protein produced by transcription and translation in rabbit reticulocyte lysates together with purified GST-defensin fusion protein. The full-length XCP1 protein was specifically pulled down by GST-defensin (Figure 6a, lanes 1, 3 and 5). As controls, we used rabbit reticulocyte lysates programmed with antisense RNA (Figure 6a, lanes 2, 4 and 6) and free GST protein used in excess (Figure 5, lanes 9 and 10). The specific interaction between defensin and XCP1 was also verified using cotransfection of both XCP1 and -defensin into COS1 cells, followed by immunoprecipitation and immunoblotting with anti-XCP1 antibodies (Figure 6b).

Figure 6.

(a) XCP1 binds -defensin. GST-pulldown assays were performed essentially as described in Materials and methods. The in vitro synthesized ³⁵S-labeled murine XCP1 (lanes 1, 3 and 5) or the in vitro translation reaction programmed with an empty pCNDA3 vector (lanes 2, 4 and 6) was incubated with increasing amounts of GST--defensin fusion proteins (lanes 1–6) or GST alone (lanes 9 and 10). The samples were analysed by SDS–PAGE through 12.5% gels and subjected to autoradiography. The position of mXCP1 is indicated on the right-hand side of the figure by an arrow. (b) XCP1–Defensin interaction in vivo. COS-1 cells were transfected with either eucaryotic GST expression vector (Chumakov and Koeffler, 1993) (lane 1) or vectors expressing the GST–defensin fusion protein (lanes 3 and 4), hXCP1 (lanes 2 and 5), GST-defensin and hXCP1 (lane 6). Approximately 200 g of whole cell lysate protein was immunoprecipitated with an anti-GST antibody (Santa Cruz), lanes 4–6, separated on 10–20% gradient SDS–PAGE, electroblotted onto an Immobilon P membrane and blotted with an anti-XCP1 antibody. In all, 35 g of total whole-cell extract protein from transfected cells were run on lanes 1–3 without immunoprecipitation. Positions of GST–defensin fusion protein (GST-DEF) and XCP1 are indicated by arrows

Full figure and legend (133K)

Discussion

Identification of mXCP1 gene

The C/EBP- gene is exclusively expressed in a strictly lineage-specific manner in myeloid cells (Antonson et al., 1996; Chumakov et al., 1997; Morosetti et al., 1997). The expression is highest at the promyelocyte stage of differentiation, but it is also readily detectable in monocytes/macrophages and granulocytes (Yamanaka et al., 1997a, 1997b). C/EBP--deficient mice have functionally defective neutrophils and macrophages; except for the myeloid hematopoietic compartment, all other cells of these mice appear to be normal. Previous studies indicated that C/EBP- could transactivate lactoferrin and gelatinase in neutrophils (Lekstrom-Himes and Xanthopoulos, 1999; Lekstrom-Himes et al., 1999 and the M-CSF receptor and a number of chemokines in macrophages; Williams et al., 1991).

In our previous study, we used PCR-based RDA analysis of peritoneal cells from C/EBP- knockout mice to identify genes that are specifically absent in C/EBP- -/- mice and are likely targets of C/EBP- regulation (Kubota et al., 2000). Approximately 10% of identified clones were derived from a novel gene that we named mXCP1 (mouse ten-cysteine containing protein 1). Computer analysis suggested the existence of a total of four related genes in mice and two in humans. Proteins encoded by these genes share a conserved ten-cysteine (XCP) signature in their C-terminal portion. We used this sequence information and cDNA probes to clone the four murine XCP genes (mXCP1, mXCP2, mXCP3 and mXCP3) as well as the two human homologues (hXCP1 and hXCP2). With the exception of hXCP2, all these genes are comprised of four exons, with the most 5' exon being noncoding (Figure 1b). While this work was in progress, the mXCP2 gene was identified as FIZZ1 (found in lung inflammation zone), and the mXCP4 gene was identified as the adipocyte-specific hormone resistin (Holcomb et al., 2000; Steppan et al., 2001a, 2001b).

We cloned and sequenced murine and human XCP1-homologous genes using cDNA-derived probes and murine and human genomic -phage libraries. Murine XCP1, XCP2/FIZZ1 and XCP3 genes were clustered within the region of approximately 50 kb, mXCP1 and mXCP2 genes being arranged in head-to-tail tandem less than 25 kb apart. The murine-hamster RH panel (Research Genetics, Inc.) was then used to identify the chromosomal location of these genes. The three murine XCP1-3 genes mapped to chromosome 16 in an area homologous to human chromosome 3q13.1 locus where the human hXCP1 gene is located. mXCP4/Resistin is not linked to this gene cluster; chromosomal mapping with the murine-hamster RH panel indicated that the mXCP4 gene is found on murine chromosome 8, syntenic to human chromosome 19, where hXCP2 is located.

Tissue specificity of expression of the XCP genes

We have used Northern blot hybridization to identify the pattern of tissue-specific expression of murine and human XCP genes (Figures 2 and 4a, b, d; data summarized in Table 2). In wild-type mice, mXCP1 was found to be expressed in the same tissues where C/EBP- is expressed, while mXCP1 expression was completely absent in C/EBP- -/- mice, suggesting regulation of the mXCP1 gene by C/EBP-. The murine embryonic fibroblast cell line NIH3T3 does not express mXCP1. Upon stable transfection with inducible C/EBP- and C/EBP- expression vectors, the cells qthat expressed C/EBP- also expressed mXCP1 (Figure 3c). This experiment indicated that the expression of C/EBP- alone was sufficient to induce mXCP1 in the nonmyeloid cell line. We also carried out in vitro reconstitution experiments, restoring C/EBP- gene expression in bone marrow-derived cells of C/EBP- -/- mice (Figure 3b, right panel). Taken together, these results strongly suggest that mXCP1 is strictly dependent on C/EBP- for its expression.

Table 2 - Summary of mRNA expression data for members of the XCP gene family.

Full table

We have also analysed the expression of mXCP homologues, mXCP2, 3 and 4. Like mXCP1, mXCP2 expression was found lacking in peritoneal exudate and the skin of C/EBP- -/- mice, but the mXCP2 expression was unaffected by C/EBP- status in the pancreas and tongue. These results suggest that the expression of mXCP2 is dependent on C/EBP- in some cell types, but not in others. We can hypothesize that this is a result of close physical proximity of mXCP1 and mXCP2 genes in the murine genome, therefore the C/EBP--responsive elements of the mXCP1 gene may also be affecting mXCP2. Alternatively, the mXCP2 gene may itself possess C/EBP--responsive elements, making it subject to complex regulation in different cell types. Putative upstream promoter regions of mXCP1 and mXCP2 genes have regions of a high degree of homology (over 80%). While this work was in progress, Holcomb et al. (2000) identified mXCP2 (FIZZ1) as an inflammation-inducible gene in lungs of mice with induced asthma.

The expression of mXCP3 gene was found exclusively in the colon (with or without mucosa) and small intestine, in agreement with the data presented by another group for FIZZ2 (Holcomb et al., 2000), while mXCP4/Resistin was found to be highly expressed in murine white fat tissue.

Our results suggesting the direct regulation of mXCP1 by C/EBP- prompted us to search for a possible regulatory and functional homologue of mXCP1 in human cells. We were able to identify and sequence only two human genes, suggesting that XCP genes have diverged relatively recently in mice. The extent of protein homology between these genes did not offer any clues as to regulatory or functional homology. This is not uncommon for multiple gene families, for example, chicken NFM (Ness et al., 1993) appears to be a homologue of C/EBP- in mice, yet its regulation and function corresponds more closely to that of mammalian C/EBP- (Chumakov et al., 1997). The expression patterns of hXCP1 and hXCP2 were found to be very similar to those of mXCP1 and mXCP3/FIZZ2, respectively. While this work was in progress, cloning of resistin (FIZZ3, mXCP4) cDNA in mice resulted in misidentification of hXCP1 as human resistin (Steppan et al., 2001a, 2001b). However, expression patterns of hXCP1 and mXCP4/Resistin/FIZZ3 show nothing in common (this study, see also Savage et al., 2001). Human resistin mRNA is undetectable in preadipocytes, endothelial cells, but is detectable in circulating mononuclear cells (Savage et al., 2001). The same study indicated that results linking murine resistin to obesity, insulin resistance and PPAR- action are not translatable to humans.

The expression of hXCP1 appeared to be regulated in human cells in parallel to C/EBP-. The hXCP1 was strongly expressed only in human bone marrow (Figure 4), strongly upregulated in the promyelocytic cell line NB4 in response to RA (in parallel with C/EBP- upregulation), but it was not upregulated in U937 cells, where C/EBP- is not inducible by RA. The expression of hXCP1 was found to be absent in cell lines lacking C/EBP- expression. Previous studies have suggested that the C/EBP- protein interacts with several transcriptional factors, including c-Myb (Mink et al., 1996; Verbeek et al., 1999a, 1999b), cAMP response element-binding protein (CREB) (LeClair et al., 1992), C/EBP- (Chih D., personal communication) and C/EBP- (CHOP) (Gombart A.F., unpublished). Similar to mXCP1, the introduction of C/EBP- expression alone was sufficient to induce endogenous hXCP1 gene expression, and this induction was amplified by cotransfection with c-myb and ATF4 transcription factors in HEK293 cells (Figure 4c).

The only other known human member of the XCP/FIZZ/Resistin gene family is hXCP2. It appears to be a regulatory homologue of mXCP3/FIZZ2, since its expression is limited to the colon and small intestine (Holcomb et al., 2000; Steppan et al., 2001a, 2001b). Taken together, our results suggest that hXCP1 is a regulatory homologue of the C/EBP--dependent mXCP1 gene, while hXCP2 is homologous to mXCP3/FIZZ2, in agreement with previously published observations (Holcomb et al., 2000; Steppan et al., 2001a, 2001b; Hartman et al., 2002).

XCP1 is a secreted protein with chemotactic activity

The majority of genes identified by our RDA subtraction experiments (Kubota et al., 2000) were either cytokines or extracellular cytokine receptors. The predicted structure of mXCP1 and hXCP1 proteins was also reminiscent of cysteine-bridged family of proteins that appear to share a general architecture, such as -defensins (Selsted et al., 1992; Ganz, 1999), -defensins (Huttner et al., 1994; Yount et al., 1995; Yang et al., 1999), trefoil factors (Kinoshita et al., 2000; Taupin et al., 2000), prostate stem cell antigen (PSCA), RIG-E and several other proteins that have a signal sequence at their amino terminus and a cysteine-crosslinked carboxy terminal-secreted domain (Figure 1). Our computer analysis failed to predict protease cleavage sites next to the signal sequences, leaving it uncertain whether XCP1 protein detaches from the cell membrane (secreted) or remains anchored there. Immunoprecipitation with antibodies raised against the XCP1 peptide and Western blotting showed that mXCP1 is secreted into the media as a processed 9 kDa protein by transfected COS1 cells (Figure 5).

In an attempt to identify a possible function of XCP1 protein, we tested whether XCP1 has a role in either cell migration, chemotaxis, antimicrobial activity or colony formation. Since bone marrow cells of wild-type mice express endogenous mXCP1, we compared the chemotactic response of bone marrow cells derived from untreated C/EBP- -/- and wild-type mice. For cells derived from the wild-type mice, the presence of media containing either hXCP1 or mXCP1 did not result in a consistent statistically significant chemotactic response. However, bone marrow cells derived from C/EBP- -/- mice showed a marked increase (four- to fivefold) in the number of migrating cells, suggesting that both hXCP1 and mXCP1 have a role in chemotaxis. Of note, we have previously identified several other known chemokines, such as cathepsin L, MIP-1-, MCP-3, Gal/GalNAc-specific lectin as significantly downregulated in the C/EBP- knockout mice (Kubota et al., 2000). We have previously found that consistent with downregulation of these chemokines, as well as mXCP1, neutrophil migration was impaired in granulocytes lacking C/EBP- (Kubota et al., 2000).

Since the protein structure of members of the XCP gene family is structurally reminiscent of -defensins, corticostatin and -defensins, we tested COS1-derived mXCP1 as well as bacterially expressed His-tagged XCP1 and GST-tagged mXCP1 proteins for microbicidal activity against a range of microbial pathogens using a multifold dilutional assay in liquid media. We did not find any microbicidal, bacteriostatic or antiviral activity of XCP1 at physiological salt concentrations (data not shown). Likewise, we were unable to detect a significant effect of either purified His-tagged mXCP1 or anti-mXCP1 antipeptide antibodies on colony formation by murine bone marrow-derived hematopoietic stem cells grown in soft agar containing myeloid cytokines (data not shown).

XCP1 binds human neutrophil -defensin

Cytokines and chemokines exert their pleiotropic effects on target cells by interacting with cell surface receptors (Broxmeyer et al., 1999; Yang et al., 1999). We have, therefore, attempted to identify possible targets for XCP1 action by using the yeast two-hybrid assay system. Analysis of these sequences indicated that 12 clones contained human neutrophil -defensin among a total of 72 clones strongly interacting with both mXCP1 and hXCP1. The specific interaction of XCP1 with defensin was confirmed by using GST-pulldown assays (Figure 6) and GST--defensin fusion protein as well as -galactosidase assay (Table 3).

Table 3 - Growth on His(-) media and LacZ activity.

Full table

While this paper was in preparation, another group indicated that mXCP2/FIZZ1/RELM is the only member of this gene family that does not form cysteine-crosslinked dimers because it lacks the 11th cysteine residue closest to the amino terminus (Banerjee and Lazar, 2001). In view of the protein homology between mXCP1 and mXCP2/FIZZ1, as well as lack of XCP1 homodimers identified by yeast two-hybrid assay, we can hypothesize that mXCP1 will, likewise, be unable to form cysteine-crosslinked homodimers. Further studies are needed to determine whether XCP1--defensin dimers are formed by the bridging of cysteines.

Defensins are 3–4 kDa antimicrobial peptides that are stored in the cytoplasmic granules of neutrophils, some macrophages and intestinal Paneth cells (Selsted et al., 1992; Ganz, 1999). Highest defensin mRNA levels are found in promyelocytes, not in mature neutrophils (Haugen et al., 2001) and they are directly inducible by RA (Herwig et al., 1996). These properties suggest that the interaction of XCP1 and defensins is quite possible in vivo. Besides antimicrobial properties, defensins HP1 and HP4 significantly inhibit spontaneous and cytokine-induced NK activity and stimulate the corticol action on the peripheral blood mononuclear cell (PBMC) at 10^-8–10^-9 M, and reduce cytokine production (IFN- and IL-6) by PBMC (Masera et al., 1996). One of the defensins, known as corticostatin I, inhibits ACTH-stimulated steroidogenesis and enhances divalent cation-independent phagocytosis by peritoneal macrophages (Ichinose et al., 1996).

-Defensins and related -defensins and trefoil factors form a large and diverse group of small-molecular-weight polypeptides having a variety of activities, all related to innate immunity – including microbicidal activity (Yount et al., 1995; Ganz, 1999; Gropp et al., 1999), endocrine suppression of adrenocorticotropin hormone production (Masera et al., 1996; Yang et al., 1999), mast cell degranulation (Yang et al., 1999), regulation of calcium-ion channels (Selsted et al., 1992) and chemotaxis (Huttner et al., 1994; Ichinose et al., 1996; Broxmeyer et al., 1999; Kinoshita et al., 2000). Further experiments are needed to elucidate the role that the novel XCP1 proteins may have in conjunction with defensins.

Taken together, we have found that mXCP1 and its human homologue hXCP1 are C/EBP--dependent myeloid genes that code for a secreted protein that can stimulate chemotaxis of myeloid cells and directly interacts with -defensin. Further studies, including reporter assays with the XCP1 promoter, will be needed to appreciate fully the role of C/EBP- in the control of these genes.

Materials and methods

Cell lines and transfection

NB-4 cells (Lanotte et al., 1991) were a kind gift from M Lanotte (St Louis Hospital, Paris, France). U937PR9 cells (Grignani et al., 1993) were generously provided by PG Pelicci (Perugia University, Perugia, Italy). Other cell lines were either established by our group (KG-1) (Koeffler and Golde, 1978) or obtained from American Type Culture Collection (ATCC) and maintained in appropriate culture media, as specified by the supplier. U937PR9 were cultured either with or without 100 M ZnSO₄ for induction of PML/RAR expression in addition to ATRA as indicated in the figure legend. Transfections used 5 g of expression plasmids in COS-1 and HEK293 cells with Lipofectin (Gibco BRL). Cell lines and bone marrow cells were electroporated with 20 g of expression plasmids per 10⁷ cells by five pulses for 99 ms at 500 V on ice using GenePulser (Bio-Rad). Following transfection, cells were returned to culture media and grown for 16–24 h prior to analysis.

Mice and TG challenge

C/EBP- -/- mice were generously provided by Dr KG Xanthopoulos and Julie Lekstrom Himes (NIH) and wild-type mice (129SV) were obtained from Harlan Sprague–Dawley, Inc. (Indianapolis, IN, USA) and maintained in pathogen-free conditions. Their genotypes were evaluated, as previously reported (Yamanaka et al., 1997a, 1997b). C/EBP- -/- mice aged 1 month and their age-matched wild-type counterparts received 2 ml of 4% thioglycollate broth (Sigma Chemical Company, St Louis, MO, USA) by i.p. injection, as previously reported (Kubota et al., 2000). Mice were killed by cervical dislocation, and peritoneal exudate cells were harvested by lavage with 8 ml of Hanks' balanced salt solution (Life Technologies, Inc., Gaithersburg, MD, USA) at 4°C.

Chromosomal mapping

We used GeneBridge hamster/mouse RH panel (Research Genetics, Inc.) to map the chromosomal loci for murine XCP genes. Mapping was performed by PCR using gene-specific primers derived from the last two exons of XCP genes. Each PCR reaction contained 25 ng of genomic DNA of an individual RH clone, 10 pmol of each of the primers and Taq-polymerase buffer exactly as suggested by the manufacturer. Samples were amplified with 30–35 cycles in a programmable thermal controller (MJ Research Inc., Watertown, MA, USA). PCR results were electronically submitted to the Jackson Laboratory Mouse Radiation Hybrid Database (http://www.jax.org/resources/documents/cmdata). The chromosomal position of the human XCP1 and XCP2 genes was discovered by BLAST search analysis of HTGS contig database.

RNA extraction, Northern blot and RT–PCR analysis

Total RNA was isolated from cell cultures and tissues of wild-type and C/EBP- -/- mice using Trizol (Life Technologies, Inc.). For PCR analysis and synthesis of full-length DNA clones, complimentary DNA (DNA) was synthesized following DNase I (RQ1 RNase-free DNase, Promega) using either SuperScript II RNase H Reverse Transcriptase (Gibco BRL) or SMART PCR cDNA Synthesis kit (Clontech, Palo Alto, CA, USA) essentially as suggested by the manufacturers. Semiquantitative PCR analysis was performed using standard methods as previously described (Morosetti et al., 1997). PCR products were electrophoresed on a 1.5% agarose gel and transferred onto a nylon membrane by alkaline transfer. To confirm the specificity of the PCR product, hybridization of the membranes was carried out using [-³²P]ATP-labeled oligonucleotide probes. PCR primer and oligonucleotide probe sequences and their respective PCR conditions will be provided upon request. Internal standard and positive controls included oligonucleotides specific for household genes, such as -actin or GAPDH. Standard techniques were used to prepare RNA for Northern blots. Briefly, total RNA was prepared from a variety of mouse organs pretreated with the RNALater reagent (Ambion) followed by the Trizol (Gibco BRL) extraction method. In all, 20 g of wild-type and C/EBP- -/- (knockout) mice-derived RNA for each tissue was separated in agarose gels containing formaldehyde, blotted onto the nitrocellulose membrane and sequentially hybridized with cDNA probes for mXCP1, mXCP2, mXCP3, mXCP4 as well as C/EBP- and HGPRT. Probes were ³²P-labeled using the StripEasy reagent kit by Ambion; the same kit was used to remove probes between consecutive hybridizations.

Preparation of anti-mXCP1 antibodies

Using an antigenicity-predicting algorithm, we have designed and synthesized 20 amino-acid residue peptides to make antibodies against mXCP1 (mp1A) protein. Rabbit polyclonal antiserum was obtained from two rabbits immunized with each peptide by Lampire Biologicals Laboratories. Approximately 10–15 mg of antibodies specific against each peptide were purified using Protein A–Sepharose 4B affinity columns (Pharmacia).

Chemotaxis assay

Cell migration in response to XCP1 was evaluated using a modified chemotaxis microchamber technique as described (Bacon et al., 1988; Dieu et al., 1998). Briefly, 2 10⁵ COS-1 cells were transfected with either XCP1 expression plasmids, control vector plasmids or were mock-transfected. These cells were placed in the lower wells of the Boyden chamber. Bone marrow cells from either C/EBP- -/- or wild-type mice were dispersed and rinsed in RPMI1640 media and 10⁵ cells/well were applied to the upper wells and allowed to migrate to the underside of the filter. Each assay was performed in duplicate and the results are presented as the mean number of migrating cells per two randomly chosen low-power fields.

Identification of XCP1-interacting proteins

Proteins that can physically interact with XCP1 were identified using a yeast two-hybrid system. To construct the 'bait' expression plasmid for identification of XCP1-interacting proteins, PCR fragments encoding murine or human XCP1 proteins were cloned into the pGBT8 expression vector. The resulting 'bait' constructs encode fusion proteins of the mature XCP1 (lacking N-terminal signal sequences) and the GAL4 DNA-binding domain. These expression vectors were transfected into the MATa AH109 yeast strain and expression of the fusion proteins was verified by Western blotting. The AH109 cells carrying the GAL4-XCP1 fusion constructs were mated with pretransformed human leukocyte-derived MATCHMAKER library as described by the supplier (Clontech Laboratories, Inc., Palo Alto, CA, USA) and screened for interaction at high stringency. The -galactosidase filter assay was used to measure the strength of interactions and was carried out as described by the manufacturer (Clontech). Positive colonies were replated to test the phenotype and to confirm interaction in yeast. Clones were further analysed by plasmid DNA extraction, PCR amplification of inserts using matchmaker 5' and 3' AD LD-insert screening amplimers and sequencing was carried out by ABI PRISM Dye termination Cycle sequencing (Perkin-Elmer, Foster City, CA, USA) and analysed by BLAST searches against available databases (Altschul et al., 1994; Altschul and Gish, 1996). To confirm interaction for XCP1 with human neutrophil defensin, the GST-defensin fusion protein was prepared and purified using the pGEX5X-1 vector and glutathione–agarose beads using standard protocol provided by the manufacturer (Pharmacia). Sequences encoding human 1 defensin were inserted into the pGEX expression vector following amplification of the defensin DNA fragment with defensin-specific primers (5'EcoDef 5'-CTCGAATTCAGGCAAGAGCTGATGAG-3' and 3' BamDef 5'-CACGGATCCAAGCTCAGCTCAGCAGCAGAATGC-3').

In vitro GST-pulldown assay

Human and murine XCP1 proteins were synthesized in vitro in the presence of [³⁵S]methionine and cysteine by the TNT-reticulocyte lysate system (Promega, Madison, WI, USA) as described by the manufacturer. A measure of 4 l of radiolabeled proteins were mixed with different amounts of GST-fusion proteins loaded on glutathione–sepharose beads in 0.5 ml of binding buffer (20 mM Tris-Cl (pH 7.5), 150 mM NaCl, 0.5% NP40, 0.1 mM EDTA, 1 mM dithiothreitol (DTT) and complete protease inhibitor cocktail (Roche Molecular Biochemicals)). Proteins were allowed to interact with continuous mixing for 1 h at 4°C. Binding reactions were washed three times with 1.4 ml of binding buffer, denatured in a sample buffer and analysed by sodium dodecyl sulfate (SDS)–polyacrylamide gel electrophoresis. Gels were dried and exposed to a Kodak X-Omat film (Kodak, Rochester, NY, USA) at -80°C.

References

1. Altschul SF, Bogusky MS, Gish W and Wooton JC. (1994). Nat. Genet., 6, 119–129. | Article | PubMed | ISI | ChemPort |
2. Altschul SF and Gish W. (1996). Methods Enzymol., 266, 460–480. | Article | PubMed | ISI | ChemPort |
3. Antonson P, Stellan B, Yamanaka R and Xanthopoulos KG. (1996). Genomics, 35, 30–38. | Article | PubMed | ISI | ChemPort |
4. Bacon KB, Camp RDR, Cunningham FM and Woollard PM. (1988). Br. J. Pharmacol., 95, 966–974. | PubMed | ISI | ChemPort |
5. Banerjee RR and Lazar MA. (2001). J. Biol. Chem., 276, 25970–25973. | Article | PubMed |
6. Broxmeyer HE, Kim CH, Cooper SH, Hangoc G, Hromas R and Pelus LM. (1999). Ann. N.Y. Acad. Sci., 872, 142–162. | PubMed | ChemPort |
7. Chumakov AM and Koeffler HP. (1993). Gene, 131, 231–236. | Article | PubMed |
8. Chumakov AM, Grillier I, Chumakova E, Chih D, Slater J and Koeffler HP. (1997). Mol. Cell. Biol., 17, 1375–1386. | PubMed | ISI | ChemPort |
9. Dieu M-C, Vanbervliet B, Vicari A, Bridon J-M, Oldham E, Ait-Yahia S, Briere F, Zlotnik A, Lebecque S and Caux C. (1998). J. Exp. Med., 188, 373–386. | Article | PubMed | ISI | ChemPort |
10. Ganz T. (1999). Science, 286, 420–421. | Article | PubMed | ISI | ChemPort |
11. Grignani F, Ferrucci PF, Testa U, Talamo G, Fagioli M, Alcalay M, Mencarelli A, Peschle C and Nicoletti I. (1993). Cell, 74, 423–431. | Article | PubMed | ISI | ChemPort |
12. Gropp R, Frye M, Wagner TO and Bargon J. (1999). Hum. Gene Ther., 10, 957–964. | Article | PubMed | ISI | ChemPort |
13. Hartman HB, Hu X, Tyler KX, Dalal CK and Lazar MA. (2002). J. Biol. Chem., 277, 19754–19761. | Article | PubMed | ISI | ChemPort |
14. Haugen F, Jergensen A, Drevon CA and Trayhurn P. (2001). FEBS Lett., 507, 105–108. | Article | PubMed |
15. Herwig S, Su Q, Zhang W, Ma Y and Tempst P. (1996). Blood, 87, 350–364. | PubMed |
16. Holcomb IN, Kabakoff RC, Chan B, Baker TW, Gurney A, Henzel W, Nelson C, Lowman HB, Wright BD, Skelton NJ, Frantz DG, Tumas DB, Peale FV, Shelton DL and Hebert CC. (2000). EMBO J., 19, 4046–4055. | Article | PubMed | ISI | ChemPort |
17. Huttner KM, Selsted M and Ouellette AJ. (1994). Genomics, 19, 448–453. | Article | PubMed | ISI | ChemPort |
18. Ichinose M, Asai M, Imai K and Sawada M. (1996). Immunopharmacology, 35, 103–109. | Article | PubMed |
19. Kinoshita K, Taupin DR, Itoh H and Podolsky DK. (2000). Mol. Cell. Biol., 20, 4680–4690. | Article | PubMed | ISI | ChemPort |
20. Koeffler HP and Golde DW. (1978). Science, 200, 1153–1154. | Article | PubMed | ISI | ChemPort |
21. Kubota T, Kawano S, Chih DY, Hisatake Y, Chumakov AM, Taguchi H and Koeffler HP. (2000). Blood, 96, 3953–3957. | PubMed | ChemPort |
22. Lanotte M, Martin-Thouvenin V, Najman S, Muto A, Ueno H, Awaya N, Kawai Y, Asou H, Kamada N and Iakeda Y. (1991). Blood, 77, 1080–1086. | PubMed | ISI | ChemPort |
23. LeClair KP, Blanar MA and Sharp PA. (1992). Proc. Natl. Acad. Sci. USA, 89, 8145–8149. | Article | PubMed | ChemPort |
24. Lekstrom-Himes J, Dorman SE, Kopar P, Holland SM and Gallin J. (1999). J. Exp. Med., 189, 1847–1852. | Article | PubMed | ISI | ChemPort |
25. Lekstrom-Himes J and Xanthopoulos KG. (1998). J. Biol. Chem., 273, 28545–28548. | Article | PubMed | ISI | ChemPort |
26. Lekstrom-Himes J and Xanthopoulos KG. (1999). Blood, 93, 3096–3105. | PubMed | ISI | ChemPort |
27. Masera RG, Bateman A, Muscettola M, Solomon S and Angeli A. (1996). Regul. Pept., 62, 13–21. | Article | PubMed |
28. Mink S, Kerber U and Klempnauer KH. (1996). Mol. Cell. Biol., 16, 1316–1325. | PubMed | ISI | ChemPort |
29. Morosetti R, Park DJ, Chumakov AM, Grillier I, Shiohara M, Gombart AF, Nakamaki T, Weinberg K and Koeffler HP. (1997). Blood, 90, 2591–2600. | PubMed | ISI | ChemPort |
30. Ness SA, Kowenz-Leutz E, Casini T, Graf T and Leutz A. (1993). Genes Dev., 7, 749–759. | PubMed | ISI | ChemPort |
31. Park DJ, Chumakov AM, Vuong PT, Chih DY, Gombart AF, Miller Jr WH and Koeffler HP. (1999). J. Clin. Invest., 103, 1399–1408. | PubMed | ISI | ChemPort |
32. Poli V, Mancini FP and Cortese R. (1990). Cell, 63, 643–653. | Article | PubMed | ISI | ChemPort |
33. Savage DB, Sewter CP, Klenk ES, Segal DG, Vidal-Puig A, Considine R and O'Rahilly S. (2001). Diabetes, 50, 2199–2202. | Article | PubMed | ISI | ChemPort |
34. Scott LM, Civin CI, Rorth P and Friedman AD. (1992). Blood, 80, 1725–1735. | PubMed | ISI | ChemPort |
35. Selsted ME, Miller SI, Henschen AH and Ouellette AJ. (1992). J. Cell. Biol., 118, 929–936. | Article | PubMed | ISI | ChemPort |
36. Steppan CM, Bailey ST, Bhat S, Brown EJ, Banerjee RR, Wright CM, Patel HR, Ahima RS and Lazar MA. (2001a). Nature, 409, 307–312. | Article | PubMed | ISI | ChemPort |
37. Steppan CM, Brown EJ, Wright CM, Bhat S, Banerjee RR, Dai CY, Enders GH, Silberg DG, Wen X, Wu GD and Lazar MA. (2001b). Proc. Natl. Acad. Sci. USA, 98, 502–506. | Article | PubMed | ChemPort |
38. Taupin DR, Kinoshita K and Podolsky DK. (2000). Proc. Natl. Acad. Sci. USA, 97, 799–804. | Article | PubMed | ChemPort |
39. Verbeek W, Gombart AF, Chumakov AM, Muller C, Friedman AD and Koeffler HP. (1999a). Blood, 93, 3327–3337. | PubMed | ISI | ChemPort |
40. Verbeek W, Lekstrom-Himes J, Park DJ, Dang PM-C, Vuong PT, Kawano S, Babior BM, Xanthopoulos KG and Koeffler HP. (1999b). Blood, 94, 3141–3150. | PubMed | ChemPort |
41. Williams SC, Cantwell CA and Johnson PF. (1991). Genes Dev., 5, 1553–1567. | PubMed | ISI | ChemPort |
42. Yamanaka R, Barlow C, Lekstrom-Himes J, Castilla LH, Liu PP, Eckhaus M, Decker T, Wynshaw-Boris A and Xanthopoulos KG. (1997a). Proc. Natl. Acad. Sci. USA, 94, 13187–13192. | Article | PubMed | ChemPort |
43. Yamanaka R, Kim GD, Radomska HS, Lekstrom-Himes J, Smith LT, Antonson P, Tenen DG and Xanthopoulos KG. (1997b). Proc. Natl. Acad. Sci. USA, 94, 6462–6467. | Article | PubMed | ChemPort |
44. Yang D, Chertov O, Bykovskaia SN, Chen Q, Buffo MJ, Shogan J, Anderson M, Schroder JM, Wang JM, Howard OMZ and Oppenheim JJ. (1999). Science, 286, 525–528. | Article | PubMed | ISI | ChemPort |
45. Yount NY, Wang MS, Yuan J, Banaiee N, Ouellette AJ and Selsted ME. (1995). J. Immunol., 155, 4476–4484. | PubMed | ISI | ChemPort |
46. Zhang DE, Zhang P, Wang N, Hetherington CJ, Darlington GJ and Tenen DG. (1997). Proc. Natl. Acad. Sci. USA, 94, 569–574. | Article | PubMed | ChemPort |

Acknowledgements

We are grateful to all investigators who shared their valuable reagents with us. We especially thank Drs James O'Kelly and Steffen W Nichols for their invaluable help in colony formation and antimicrobial activity assays, and Kimberly Burgin for excellent secretarial assistance. We are grateful to Drs A Gombart and Carl W Miller for helpful discussions, and Dr Doris Chih and other members of the Koeffler lab for generously sharing their reagents.

This work was supported in part by NIH grants and the Parker Hughes Fund. HPK is the holder of the Mark Goodson Chair of Oncology Research at Cedars Sinai Medical Center and is a member of the Jonsson Cancer Center and the Molecular Biology Institute (UCLA).

Sequence data for all XCP genes were deposited in GenBank under accession numbers BE847000, AF510097, AF510098, AF510099, AF510100, AF352730 and AF352731.