Molecular Cloning and Characterization of the Mousetag7 Gene Encoding a Novel Cytokine*

Cloning of the mouse tag7 gene encoding a novel cytokine is described. The Tag7 protein consists of 182 amino acids. Genomic organization of the tag7 gene and its promoter region remind those of the genes of the tumor necrosis factor locus, although the tag7 gene is not linked to this locus. The gene is located on chromosome 7 at the area that corresponds to band 7A3, which has genetic linkage with lupus-like disease in mouse models. tag7 transcription is essential for lymphoid organs. It is also detected in certain areas of lungs, brain, and intestine and in some tumors. Tag7 protein is detectable in both cell-associated and soluble forms. The soluble form of Tag7 triggers apoptosis in mouse L929 cells in vitro and does not involve NF-κB activation. The relationship between Tag7 and tumor necrosis factor family of ligands is discussed.

The term "cytokine" has come to be used for a diverse group of growth factors, inflammatory mediators, and hematopoietic regulators that are distinct from more classical hormones secreted by the glands of the endocrine system. Cytokines are proteins that act as a soluble cell to cell messengers, distinguished by their high interactivity, and can act in both autocrine and paracrine manners.
Neoplastic cells themselves characteristically produce cytokines spontaneously, and even in the same lineage, they can have distinct cytokine expression and/or secretion profiles (1)(2)(3)(4). Cytokines secreted by tumors could be used either as autocrine factors, to recruit and suppress reactive leukocytes, or to modulate the activity of endothelial and stromal cells. The ability to release cytokine in an autocrine manner could be a key factor in the promotion of neoplastic transformation and in permitting tumor growth in vivo (5). In their turn, cytokines produced by the host in response to a tumor modify the scenario created by tumor growth (6). The analysis of expression of cytokine genes by tumors indicated that cells constitutively produced both autostimulatory and inhibitory cytokines. Ex-pression of several cytokines, including interleukin-1␤ (IL-1␤), 1 IL-6, IL-8, IL-10, tumor necrosis factor-␣, lymphotoxin-␣, and granulocyte-macrophage colony-stimulating factor by some tumors has been described. Tumor necrosis factor (TNF) and lymphotoxin-␣ (LT-␣, also known as TNF-␤) are related cytokines involved in many regulatory activities (7,8), but their roles in the immune system although suggesting very critical functions (9) are still enigmas. TNF and LT-␣ are released by a number of tumor cells originating in mouse fibrosarcoma, epithelial human cell lines, and T-cell leukemia (10 -12).
In this study, we used a pair of related mouse transplanted tumors with the opposite metastatic properties to identify genes overexpressed in one of them. As a result, the tag7 gene was cloned. It was expressed in tumor with high metastatic potential, although no correlation with metastatic potential was detected after study of many different tumors. The tag7 gene is preferentially transcribed in normal lymphoid and hematopoietic cells. Tag7 is a secreted protein possessing a significant cytotoxicity realized through apoptosis.

Cells and Cell
Cultures-VMR-0, VMR-L, CSML-0, and CSML-100 cell lines were established from their respective mammary adenocarcinoma transplanted tumors described in Ref. 13. Cell lines saved their ability to produce tumors in syngenetic mice with low (VMR-0 and CSML-0) and high (VMR-L and CSML-100) metastatic potential when injected subcutaneously. Mouse splenocytes, thymocytes, monocytes, and peritoneal macrophages were isolated as described previously (14). Cells were cultured in RPMI 1640 medium containing 5% FBS 100 units/ml penicillin, 100 units/ml streptomycin, 10 mM HEPES. Cells were activated with LPS (5 g/ml) in serum-free medium for different time. VMR-0 cells were transfected with the pM5Gneo tag7 construct or mock-transfected using Lipofectin reagent (Life Technologies, Inc.) according to the manufacturer's recommendations, and clones were selected on G418 resistance and maintained.
Cloning of the tag7 Gene-A fragment of the tag7 cDNA was isolated by differential display technique essentially as described by Liang and Pardee (15). T 12 AC oligonucleotide was used as the anchoring primer, and AATCGGGCTG was used as the arbitrary primer. A fragment of 390 bp was used as a probe for Northern blot hybridization and cDNA library screening (16). cDNA library was prepared on poly(A) ϩ RNA isolated from VMR-L tumors using ZAP-cDNA Gigapack Cloning kit (Stratagene) according to the manufacturer's recommendation. Positive clones were purified, and the inserts were excised as pBluescript clones using helper phage as described by the manufacturer. Few clones were sequenced using Sequenase Version 2.0 sequencing kit (Amersham Pharmacia Biotech) and synthetic oligonucleotide primers (Applied Biosystems 391 DNA synthesizer). Genomic library was constructed in the FIX II vector (Stratagene) and screened according to a standard procedure (16). The inserts were subcloned in pGEM7Z vector and partially sequenced as described above.
RNA Hybridization-Total RNA from tumors and cell lines was * This work was supported by International Association for the Promotion of Cooperation with Scientists from the New Independent States of the Former Soviet Union (INTAS) Grant N1010-CT93-0029, Pharmaceutical European Community Organization Grant ERB3530PL941128, the Russian Foundation for Basic Research, and the Moscow Anticancer Program. 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.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) X86374 (cDNA) and Y12088 (genomic DNA).
§ These authors contributed equally to this work. ¶ To whom correspondence should be addressed. Tel.: 7-095-1359970; Fax: 7-095-1354105; E-mail: slk@mx.ibg.rssi.ru. isolated by guanidine thiocyanate procedure, resolved on a 1.2% agarose-formaldehyde gel, and blotted onto Hybond-N as recommended by the manufacturer (Amersham Pharmacia Biotech). The EcoRI/XhoI fragment from the longest cDNA clone was labeled by random priming and used as a probe. To equalize an amount of RNA loaded on each lane, hybridization with glyceraldehyde-3-phosphate dehydrogenase DNA probe was performed. In situ hybridization was performed as described previously (17).
DNAs-Oligonucleotides were synthesized corresponding to the 5Ј and 3Ј ends of the coding regions of the mouse tag7 gene, with BamHI and HindIII restriction sites appended to the ends of oligonucleotides. The coding region of the gene was amplified by standard polymerase chain reaction techniques, cut with BamHI and HindIII, and inserted in frame in the BamHI and HindIII sites of the pQE30 expression vector (Qiagen). For eucaryotic expression, full-sized tag7 cDNA was subcloned in NheI-XhoI sites of pBK-CMV (Stratagene) and EcoRI-BamHI sites of pM5Gneo vectors.
Chromosomal Mapping of the tag7 Gene-Fluorescence in situ hybridization on metaphase mouse chromosomes was performed by Genome Systems, Inc.
Immunological Methods-Escherichia coli recombinant Tag7 protein was expressed in M15[pREP4] (Qiagen) and purified on Ni-NTA agarose (Qiagen) as recommended by the manufacturer. Rabbit antibodies raised against recombinant Tag7 were affinity-purified on a Sepharose (Amersham Pharmacia Biotech) column with the immobilized recombinant Tag7, as recommended by the manufacturer. SDS-polyacrylamide gel electrophoresis, immunoprecipitation, and immunoblotting were performed according to standard procedures (16). The approximate amount of secreted Tag7 by VMRSX8 clone was determined by immunoblotting.
Analysis of Tag7 Multimeric Structure-Mouse splenocytes (1.5 ϫ 10 8 cells) were treated with LPS for 18 h on two 100-mm dishes as described above. After that, Complete Mini EDTA-free protease inhibitor mix (Boehringer Mannheim) and EDTA to final concentration 1 mM were added to conditioned medium. Conditioned medium was concentrated and fractionated on Superdex 75 HR10/30 column (Amersham Pharmacia Biotech) in 20 mM Tris-HCl, pH 7.5, 100 mM NaCl according to the manufacturer's recommendation. For column calibration, bovine milk ␣-lactalbumin (14.2 kDa), bovine erythrocyte carbonic anhydrase (29 kDa), and chicken egg ovalbumin (45 kDa) (Sigma) were used. Proteins from collected fractions were precipitated with trichloroacetic acid, resolved by 15% SDS-polyacrylamide gel electrophoresis, and transferred to membrane. The membrane was then probed with anti-Tag7 polyclonal antiserum.
Tag7 Cytotoxicity Assay and Neutralization of Cytotoxic Activity-As a source of the native form of the Tag7 protein, the medium conditioned by the VMR-0 pM5Gneo tag7 (VMRSX8)-transfected cells was used. L929 cells were cultured in 96-well plates at a density of 3 ϫ 10 4 cells/well. After overnight incubation, cells were treated with actinomycin D (1 g/ml) for 2 h at 37°C in serum-free medium. After that, 100 l/well of the VMRSX8-conditioned medium or VMR-0-conditioned medium were added. As indicated, recombinant human TNF (rhTNF) (Sigma) was added at a concentration of 10 ng/ml in a volume of 100 l/well. To determine cell death, CytoTox 96 Assay kit (Promega) was used or cells were stained with trypan blue and the coded samples were counted under the microscope in a blind fashion, with a minimum of 100 cells scored for each group.
Neutralization of the cytotoxic effect of the Tag7 protein or rhTNF was performed using affinity-purified polyclonal rabbit anti-Tag7 and polyclonal anti-hTNF (Sigma) antibodies. Polyclonal antibodies were added to the VMRSX8-conditioned medium at a final concentration of 2 g/ml, and the cytotoxic effect was determined as described above; rabbit IgG in the same concentration was used as a control.
DNA Fragmentation Analysis-DNA fragmentation analysis was performed as described (18), with modifications. In brief, 2 ϫ 10 6 L929 cells were preincubated with actinomycin D (1 g/ml) for 2 h at 37°C in serum-free medium and subsequently incubated with the VMRSX8conditioned medium or rhTNF for 5 h. Cells were harvested and lysed in 20 mM Tris-HCl, pH 8.0, 0.8% Triton X-100, 10 mM EDTA, pH 8.0. After centrifugation, DNA from the supernatant was precipitated at -20°C by isopropanol in the presence of NaCl. DNA was resuspended in TE buffer (10 mM Tris-Cl, pH 7.4, 1 mM EDTA, pH 8.0) and treated with RNase A, and fragments were resolved in 1.8% agarose gel in TBE buffer (90 mM Tris-borate, 90 mM boric acid, 2 mM EDTA).
NF-B Activation Assay-Nuclear extracts were obtained as described in Ref. 19, and EMSA was performed according to Refs. 16 and 20. NF-B consensus oligonucleotide (Promega) was used for gel shift assay.

RESULTS
Isolation and Structure of the Murine tag7 Gene-A system of two related mouse transplanted tumors with opposite metastatic potentials was used to isolate mRNA molecules specifically expressed in one of them. A fragment of complementary DNA was isolated by the differential display technique from the highly metastatic mouse mammary adenocarcinoma VMR-L as overexpressed mRNA in comparison with the nonmetastatic tumor VMR-0. The fragment gave rise to the 0.7-kb mRNA transcript specific for highly metastatic VMR-L tumor. This gene was designated tag7.
A cDNA library was constructed with mRNA from VMR-L tumor in a ZAP II vector, and the fragment was used as a probe to isolate complementary DNA. Most of the cDNA clones were sequenced. The tag7 cDNA is predicted to encode a protein 182 amino acids long with a molecular mass of 20.2 kDa (Fig. 1). Hydropathy computer analysis suggested the presence of a signal sequence (amino acids 1-21) for translation into endoplasmic reticulum followed by extracellular region (amino acids 22-181). The existence of signal peptide indicates that the protein product may be secreted. Protein data base searches (FASTA program from IntelliGenetics and BLAST from the National Center for Biotechnology Information) revealed no significant similarity between the Tag7 polypeptide and known gene products.
To obtain a genomic copy of the tag7 gene, a mouse genomic library was constructed in FIXII vector and screened, and the genomic structure of tag7 was established. The murine tag7 gene spans 8 kb, consists of three exons, and contains a nonconsensus TATA box (Fig. 1). The mRNA start site was mapped using primer extension (data not shown), revealing two initiation starts separated by four nucleotides 22 bp downstream of the TATA box, probably due to the nonconsensus structure of the latter.
Computer analysis of approximately 200 bp of the upstream sequence (Signal Scan) revealed a number of potential binding sites for some known transcription factors, such as Ets-1, NF-B, Sp-1, and MyoD (Fig. 1). The order of binding sites and the distance from the mRNA start site are very close to the control region of mouse LT-␤ gene (21). Furthermore, the 5Ј-untranslated end of the mRNA was short, which is also typical for lymphotoxin genes.
Chromosomal Mapping of tag7-To determine chromosomal localization of tag7 in mouse genome, metaphase chromosomes were analyzed by fluorescence in situ hybridization. A total of 80 metaphase cells were analyzed; 62 exhibited specific labeling. This experiment resulted in the specific labeling of the centromeric region of chromosome 7 (data not shown). Measurements of 10 specifically hybridized chromosomes 7 demonstrated that signal is located at a position that is 9% of the distance from the centromere, an area that corresponds to band 7A3.
Transcription of tag7 Is Specific for Lymphoid and Some Other Cells-Northern hybridization was used to estimate the correlation between tag7 mRNA expression and metastatic properties of tumors. However, no correlation was observed upon hybridization with total RNA of CSML-0 and CSML-100 tumors (Fig. 2). Moreover, the expression of the tag7 gene appeared to be specific in this pair for the tumor with a low metastatic capacity, CSML-0. There was also no correlation between the metastatic potential and tag7 mRNA expression in other tested murine tumor cell lines (data not shown). The tag7 mRNA level dramatically altered after the establishment of the tumors as a cell culture (Fig. 2). In an established cell line from VMR-L tumors, the level of tag7 transcription dropped down, whereas in the CSML-0 cell line, obtained from corresponding tumors, it appeared to be up-regulated. Thus, tag7 in transplanted tumor is regulated by the host factors.
To determine the tissue-specific pattern of tag7 transcription, Northern blot analyses and in situ hybridization were performed. Northern hybridization of several adult mouse tissues (Fig. 3A) revealed the highest level of tag7 transcription in the lungs and spleen, a detectable level in brain and thymus, and no mRNA content or just a very low level in the other tissues tested. However, in situ hybridization performed on sections of selected organs allowed us to detect a high level of tag7 transcripts in certain areas of brain and intestine (Fig. 3,  B-E). Only single, randomly distributed cells were labeled in the thymus. In the lungs, the label was concentrated in the intraalveolar space, where, presumably, alveolar macrophages are located (data not shown). Very characteristic pictures were observed in the brain, where the distribution of tag7-expressing cells was extremely nonrandom. In the cerebellum, Purkinje cells were specifically labeled. Similarly, only certain layers of neurons were positive in hippocampus (Fig. 3, B and  C). Finally, strong expression of the tag7 gene could be detected among the cells filling the space within the intestinal villus (Fig. 3D).
Considering a similarity in distribution of cis-regulatory elements of the tag7 and LT-␤ genes, we analyzed tag7 expression in some mouse lymphoid cells. Northern blot analysis was performed with total RNA isolated from mouse circulating monocytes, thymocytes, splenocytes, and resident peritoneal macrophages (Fig. 4). All cells except for monocytes displayed a constitutively high level of tag7 mRNA.
Stimulation with IL-2 or phytohemagglutinin did not lead to a significant increase of the mRNA content in thymocytes and macrophages in the time course studied (data not shown). In mouse splenocytes cultured with LPS, the level of tag7 transcription dropped down within the first hours of activation (Fig.  4). After the first 5 h of LPS treatment, the level of tag7 mRNA decreased twice; thereupon, induction took place, but the increase of the mRNA level was relatively low. The highest observed level of tag7 mRNA was observed after 24 h of stimulation (Fig. 4). However, Northern blot analysis of murine B cell lymphoma cell line WEHI-231 and T cell lymphoma LBRM-33 did not reveal any detectable level of tag7 transcripts (data not shown) and furthermore, no tag7 transcripts were detected in the above cell lines stimulated with phorbol ester, LPS, and calcium ionophore for 20 h (data not shown).
Tag7 Is a Secreted Protein and Secretion Can Be Activated in Lymphoid Cells-To study Tag7 expression at the protein level, rabbit polyclonal antibodies raised against the E. coli Tag7 recombinant protein were used for Western blotting analysis of VMR-L and CSML-0 cells (Fig. 5A). The cells were separated from the cultivation medium and lysed. Proteins from both cellular lysates and cultivation medium were immunoprecipitated with anti-Tag7 antibodies, separated in polyacrylamide gel, and transferred to membrane. Most of the Tag7 protein was detected in conditioned medium of VMR-L cells (Fig. 5A). A similar pattern of gene expression was observed for another Tag7-expressing cell line, CSML-0 (data not shown). The level of tag7 transcription in CSML-0 cell line was higher than in VMR-L cells (Fig. 2); however, the level of protein synthesis remained low, and we were unable to detect Tag7 protein without immunoprecipitation with antibodies.
In freshly isolated mouse splenocytes, tag7 mRNA was present at a rather high level, but Tag7 protein was also detected in these cells only after immunoprecipitation with anti-Tag7 antibodies. Mouse splenocytes (Fig. 5B) contained a significant amount of cellular Tag7 protein, and LPS stimulation increased its content 5-7-fold with a maximum at about 24 h. Although during the first hours of activation most of Tag7 protein was found in a splenocyte-associated form, a long ex-posure of cells to LPS resulted not only in increasing the overall amount of the protein but also in its secretion. To determine the quaternary structure of the protein, cultivation medium from LPS-activated mouse splenocytes was concentrated and applied on Superdex 75HR column. Fractions were collected and concentrated, and Western blot analysis with anti-Tag7 antibodies was performed (Fig. 5C). About half of the total amount of the protein was detected in a 50-kDa fraction; the rest of the protein remained in low molecular weight fractions. This fact allowed us to suggest that secreted native Tag7 protein could also exist as a multimeric complex.
Secreted Tag7 Induces Cell Death in an Apoptotic Manner-Recombinant E. coli Tag7 protein was isolated as a denatured protein from inclusion bodies. All of our attempts to refold the protein after purification on Ni-NTA column failed. During dialysis, the purified from the E. coli system material irreversibly precipitated. To perform functional studies with Tag7 protein, we constructed cell line constitutively producing Tag7.
The VMR-0 cell line, which does not express tag7 (Fig. 6A), was stably transfected with the constructions expressing the tag7 gene. For transfections, both pBK-CMV and pM5Gneo eucaryotic expression vectors were used; however, the pM5Gneo vector allowed us to obtain significantly higher level of Tag7 expression. We analyzed 15 G418-resistant clones on the level of Tag7 expression. In most of them, the level of the exogenous tag7 transcription was relatively high, but Tag7 led not only to protein synthesis activation but also to the secretion of the protein. Tag7 protein was also immunoprecipitated before polyacrylamide gel electrophoresis from cellular lysates (10 7 cells) or their conditioned medium. Sizes of molecular mass markers are indicated on the right. C, native Tag7 presumably exists in two major forms: monomer and trimer. A Superdex 75 column was calibrated using molecular mass standards (upper curve). Then, concentrated conditioned medium from mouse splenocytes was applied on the column, fractions were collected and concentrated, and Western analysis was performed.

FIG. 6. Soluble Tag7 protein is a cytolytic agent that induces cell death in the apoptotic manner.
A, conditioned medium of 10 7 of mock-transfected VMR-0 and VMRSX8 cells were immunoprecipitated with anti-Tag7 antibodies and were used for Western blotting. B, L929 cells were cultured for 5 h in the presence of mock-transfected VMR-0and VMRSX8-conditioned medium or rhTNF (10 ng/ml) under the indicated conditions. Cell death was assayed by trypan blue staining (conditioned medium and rhTNF) or lactate dehydrogenase release (rhTNF). Data represent cell death minus spontaneous cell death and in presence of rabbit IgG. Error bars represent the S. D. of measurements from five independent experiments, and the values are the average of these measurements. C, oligonucleosomal DNA fragmentation induced by supernatant of VMRSX8 cells. L929 cells were cultured for 5 h with conditioned medium of mock-transfected VMR-0, VMRSX8, or rhTNF as indicated, and then fragmented DNA was recovered and resolved by 1.8% agarose gel electrophoresis in TBE. Positions of size markers are shown on the left. protein synthesis remained hardly detectable. The clone with the highest level of Tag7 secretion (VMRSX8) (less than 10 ng/ml) was selected for further studies of soluble Tag7 protein.
For this, L929 cells were treated with either conditioned supernatant from VMRSX8 cells or from the control mock-transfected VMR-0 cells for different time intervals. It was found that conditioned medium from VMRSX8 cells caused death in target L929 cells. Maximum cytotoxicity was observed at 5 h and did not increase significantly after 24 h incubation (Fig.  6B). The results obtained demonstrated that both TNF and Tag7 killed L929 cells in the presence of actinomycin D. Addition of anti-Tag7 polyclonal antibodies specifically blocked cell death caused by VMRSX8 supernatant and did not affect TNFinduced apoptosis. At the same time, anti-TNF antibodies did not block Tag7-induced cell death. The supernatant from VMR-0 cells did not possess any cytotoxic activity. Human breast adenocarcinoma cell line MCF7 was also susceptible to Tag7 killing (data not shown).
Because the process of apoptosis is known to rapidly induce DNA fragmentation, the ability of the secreted form of Tag7 to trigger apoptosis was examined utilizing a DNA fragmentation assay. Conditioned supernatant of VMRSX8 cells was added to target L929 cells. As controls, we used mock-transfected VMR-0-conditioned medium and recombinant TNF. After a 5-h incubation, fragmented DNA in the cytoplasm was recovered and resolved by agarose gel electrophoresis. Murine L929 cells were triggered to undergo apoptosis by the soluble form of the Tag7 protein and recombinant TNF but not with supernatant from VMR-0 cells as, evidenced from intranucleosomal DNA fragmentation (Fig. 6C).
NF-B Accumulation Is Not Activated in Target Cells by Tag7-TNF has been shown to induce apoptosis and NF-B activation, two of the most important activities signaled by TNFR-1 (22). We investigated whether Tag7 does also activate NF-B. VMR-0 cells stably transfected with tag7-expressing construct (VMRSX8) did not show changed viability and proliferation. To investigate the NF-B activation in sensitive to Tag7 cells, L929 cells were treated with either conditioned supernatant from cells transfected with tag7 (VMRSX8) or recombinant human TNF in concentration 10 ng/ml. Nuclear extracts were prepared 2 h later, and reacted with end-labeled NF-B specific probe and subjected to EMSA (Fig. 7). Tag7 did not induce detectable NF-B activation in L929 cells, although cytotoxicity of Tag7 was significantly higher than that of TNF during short-term of exposure of the cells to these cytotoxic agents. DISCUSSION During the last decade, a number of TNF-related cytokines were identified and cloned. Some of them are produced by tumor cells (11,23). Using the "differential display" technique, we have isolated a novel mouse gene with a remote homology to the TNF family of ligands. It came out first from the genomic organization of the tag7 gene. It is similar to the mouse LT-␤ three-exon structure and the noncanonical TATA box with multiple transcription initiation sites. The promoter regions of TNF, lymphotoxins, and tag7 genes show similarity in regulatory elements, although their combination is unique for each gene.
The tissue specificity of tag7 expression also resembles that of mouse LT-␤ expression with quite a few exceptions. In situ hybridization analysis for tag7 on sections of selected organs also demonstrated the pattern of the signal distribution close to LT-␤ mRNA. The total level of tag7 mRNA in brain was very low; however, specific cell types gave a strong hybridization signal upon in situ hybridization. Specific cortical areas of the brain (the hippocampus and Purkinje cells of the cerebellum) (Fig. 3, B and C) showed a high level of tag7 mRNA content. The same pattern of gene transcription was observed for the mouse LT-␤ (21) gene. Only minor differences in the patterns of tag7 and LT-␤ mRNA distribution were observed. Lymphoid and hematopoietic organs contained Tag7-expressing cells, although in the thymus, the level of tag7 transcription was very low. On the other hand, the level of tag7 transcription in spleen was very high, and on sections, tag7 mRNA was detected almost everywhere (data not shown), whereas LT-␤ mRNA was predominantly expressed in the white pulp. This implies that two genes may still be differentially regulated and activated through different signaling pathways. A significant level of tag7 transcription in lungs could be attributed to the large amount of alveolar macrophages.
The presence in the regulatory region of the tag7 gene of NF-B binding site supposed very fast response on well-known stimuli, e.g. LPS or phorbol 12-myristate 13-acetate, which could result in increased transcription of the gene. We detected down-regulation of tag7 transcription in lymphocytes at the early stages of activation. A similar type of down-regulation of TNF mRNA expression in macrophages is mediated through the regulation of NF-B activation (24,25). The reduction in tag7 transcription indicates that additional nuclear factors may be missing or that silencers may be activated in such a way that transcription of the tag7 gene is prevented. Activation of the tag7 gene expression was relatively low in splenocytes and tumor cell lines expressing the gene: VMR-L and CSML-0. Alteration in the level of tag7 transcription after establishment transplanted tumors as a cell lines points to the participation of host factors in tag7 transcription. It was further supported by the observation that in B and T cell lines, we were unable to detect tag7 mRNA, although its transcription was abundant in freshly isolated splenocytes. However, we can not rule out the possibility that splenocytes became activated due to isolation procedure.
A constitutive level of tag7 expression in the lymphoid and hematopoietic tissues points to a role for this gene in immune system. Furthermore, it is evident that the main step in regulation cascade occurred at the posttranscriptional level. Even insignificant changes in the mRNA production result in a change of overall amount of protein synthesized and its secretion, although constitutive level of Tag7 protein in isolated lymphoid cells is rather high.
We did not find any significant homology for Tag7 with any known proteins. However, similarity in genomic organization and expression pattern with lymphotoxin-␤ motivated us for FIG. 7. Soluble Tag7 does not induce NF-B activation in mouse L929 cells. L929 cells were treated with rhTNF-, VMR-0-, or VMRSX8-conditioned medium and analyzed for NF-B activation. As a control, nuclear extracts from untreated L929 cells were used. Free DNA is not shown. more detailed amino acid sequence analysis. Low homology allowed multiple alignments, but this analysis revealed the existence of five regions of low homology in the extracellular domains of the Tag7 polypeptide and TNF family members (Fig. 8). In the first two regions, the level of identity with other TNF family ligands is relatively high, especially in residues buried in the ␤-sheet interior (26). The rest of the domains show less homology, but the buried residues are still conservative. Moreover, gel filtration data support the similarity in quaternary structure of Tag7 and TNF family ligands. Like lymphotoxins and TNF itself, native secreted Tag7 was detected as a multimer, and the molecular weight of the complex allows us to speculate that it forms trimers. However, whether Tag7 forms homo-or heterotrimers is unknown. Tag7 is rich in cysteine residues and, like LT-␣, contains 4 methionine residues (27).
The tag7 gene was mapped on mouse chromosome 7. Previously, it was observed that many of TNF-related genes are clustered in the genome. TNF/lymphotoxins genes are linked on mouse chromosome 17 and human chromosome 6 (28 -30). The tag7 gene does not map on the same chromosome with other known TNF ligand family members. Chromosome location offers a possible in vivo role of the tag7 gene. The cytogenetic band 7A3 has been genetically linked with lupus-like nephritis in the MRL and New Zealand hybrid models of systemic lupus erythrematosus (31,32). Although systemic lupus erythrematosus is unlikely to involve mutations with severe functional alterations, gene knockout experiments may provide insight into pathogenic process.
Tag7 released in conditioned medium possesses cytotoxicity and triggers intranucleosomal DNA fragmentation in target cells in the same way as many known members of the TNF family. Fragmentation of DNA is one of the characteristics of apoptosis. We can not exclude possibility that Tag7 acts via binding with known "death domain" receptors, but it is unlikely that TNF receptor is involved in the apoptotic cell death caused by Tag7. Cell lines that naturally produce Tag7 protein, such as CSML-0 or VMR-0 (which was engineered to produce it) (33), are easily susceptible to killing by TNF. Another piece of evidence that points to the existence of a specific Tag7 receptor is the existence of different pathways of signal transduction. In addition to inducing apoptosis, TNF receptor 1, CD95 (Apo1), DR3, DR4, and DR5 can activate the transcription factor NF-B (34 -36), which controls expression of multiple immunomodulatory genes (37). We did not detect up-regulation of nuclear NF-B in target L929 cells in the presence of secreted Tag7 protein. However, we can not rule out the possibility that in another system, Tag7 would induce NF-B activation, as observed for TRAIL (TNF-related apoptosis-inducing ligand) receptors (38).
A tumor can produce cytokine; this may have important consequences, which may be direct (promoting or inhibiting tumor growth) or indirect (changing such growth through interactions on the microenvironment). The effect of a single cytokine, however, cannot be readily predicted, because its presence induces other cytokine(s), which can significantly affect the primary action. Recent studies with the tumor cells transformed with tag7 demonstrated the important role of the gene in tumor growth. 2 FIG. 8. An amino acid sequence comparison of Tag7 and the TNF family of ligands. The alignment is arranged to show maximum similarity of Tag7 and the TNF-like cytokines that are most close to Tag7 . The similarity in the rest of the sequence is much lower than shown here. b indicates residues buried in the ␤-sheet interior of ligands of the TNF family that are similar for the proteins shown.