Identification of an IL-4-Inducible Gene Expressed in Differentiating Lymphocytes and Male Germ Cells

Interleukin 4 (IL-4) is a cytokine that is involved in the differentiation of B and T lymphocytes. In this report, we describe the identification of a novel gene, N.52, which was cloned from the murine pre-B cell line R8205 grown in the presence of IL-4 for 48 hr. Although N.52 expression is detectable at low levels in unstimulated R8205 cells, the level of N.52 dramatically increases after only .4 hr exposure to IL-4 and remains at a high .level up to 48 hr. Although N.52 expression is low or absent in normal spleen B and T cells, its expression can be induced by the differentiation signals delivered by LPS in B cells and by Con A in T-cell hybrids. While N.52 mRNA is absent in all highly differentiated organs, it is detectable in stem cell harboring lymphoid tissues such as bone marrow, fetal liver, and thymus. Furthermore, N.52 mRNA is expressed at strikingly high levels in the testis, specifically in differentiating male germ cells. It is induced by differentiation signals triggered by the combination of cyclic AMP and retinoic acid in teratocarcinoma F9 cells. Taken together, these data suggest that N.52 is a developmentally regulated gene whose expression in cells of the immune and reproductive systems may be controlled by stimuli that induce differentiation.


INTRODUCTION
Cellular differentiation involves a complex pattern of gene expression in response to extracellular and intracellular signals. External signals are delivered via binding of chemical mediators to specific receptors on the cell surface, and this binding results in the activation of gene programs in a selective, tissue-specific manner (Kelly et al., 1983;Kr6nke et al., 1985;Lau and Nathans, 1987;Ryder et al., 1988). Cells of the immune system are known to respond to a myriad of chemical mediators called cytokines or interleukins. One such cytokine, interleukin 4 (IL-4), is a T-cell and mast-cell derived peptide that regulates a broad spectrum of biological activities in *Corresponding author. several cell types (reviewed by Paul and Ohara, 1987). Receptors for IL-4 are expressed on most cells of hematopoietic lineage Park et al., 1987) as well as some nonhematopoietic cells (Lowenthal et al., 1988). In particular,  induces growth and differentiation of pre-B cells (Hofman et al., 1988), the hyperexpression of class II major histocompatibility complex (MHC) molecules (Noelle et al., 1984;Roehm et al., 1984;Polla et al., 1986) and Fc receptors for IgE on resting B lymphocytes Defrance et al., 1987), and, together with bacterial lipopolysaccharide (LPS), induces the differentiation of mature B cells into immunoglobulin-secreting plasma cells Rothman et al., 1988). It is now clear that IL-4 also affects the development of T lymphocytes. IL-4 causes proliferation of antigen-stimulated T helper lymphocytes (Fernandez-Botran et al., 1986) and, together with phorbol esters (PMA), induces the differentiation of .cytotoxic T lympocytes from intrathymic precursors (Palacios et al., 1987). IL-4 also synergyizes with another cytokine, interleukin 3, to generate mast cells from precursors in the bone marrow (Mosmann et al., 1986).
Similar to differentiation of the hematopoetic cells, development of the germ-cell lineage also involves progression through a highly specific series of differentiation events, in response to a poorly understood network of signals. The involvement of specific hormones and growth factors in these processes has been implicated, but the molecular mechanisms involved are unclear. Similarly; the molecular mechanisms by which IL-4 mediates the proliferative and differentiative programs of cells of the immune system are not well understood. In this report, we describe the isolation of a gene that is induced by IL-4 in a murine pre-B cell line. This gene, designated N.52, is also inducible in both B and T lymphocytes by differentiative and activating signals delivered by polyclonal activators such as LPS and Con A. The preferential expression of N.52 in bone marrow and thymus, lymphoid organs containing rapidly proliferating and differentiating cells, as well as its strikingly high level of expression in differentiating germ cells of the testis, suggest that N.52 is a developmentally regulated gene whose expression contributes to cellular growth and differentiation, particularly within the hematopoietic and germ-cell lineages.

RESULTS
Isolation of the N.52 cDNA Earlier data from our laboratory demonstrated that IL-4 induces transcription of class IIMHC genes in an Abelson virus transformed, IL-4 receptor-bearing pre-B cell line, R8205 (Polla et al., 1986). We sought to identify other genes induced by IL-4 using this IL-4-responsive cell line as a model system. A cDNA library was constructed from R8205 cells grown in the presence of IL-4 for 48 hr and approximately 60,000 independent recombinant phage plaques were screened in duplicate with a subtracted probe. The probe was obtained by synthesis of [32p]-labeled cDNA from R8205 cells grown in the presence of IL-4, and subtraction of the cDNA with RNA derived from unstimulated R8205 cells. Positive clones (approximately 0.8%) identified by the first screening were rescreened by differential hybridization using cDNA probes prepared from R8205 cells grown in the presence or absence of IL-4. Close to 64% of the clones from this screening showed equivalent hybridization to both probes, 1% showed no hybridization to either probe, and 35% hybridized preferentially to the IL-4 stimulated probe. This last group contains genes that are specifically induced by IL-4. One clone from this group, N.52, hybridized selectively and strongly only to the induced probe. Northern blot analysis contirmeci that 0.95-kb transcripts for N.52 are expressed at significantly higher levels in R8205 cells grown in IL-4 for 48 hr than in uninduced cells where only a variable low to faint signal was detected (Fig. 1A). The kinetics of induction of N.52 in R8205 cells by IL-4 is shown in Fig. lB. N.52 levels increase within 4 hr of exposure to IL-4, continue to rise up to 24 hr and remain at a high level after 48 and 72 hr (not shown) growth in culture medium containing IL-4. Two additional higher molecular-weight transcripts are occasionally detected (see Fig. 5) and may represent partially spliced forms of N.52 nuclear RNA.
Sequence Analysis of N.52 cDNA The initially isolated partial cDNA clone contained an insert of 0.45kb. Several additional clones obtained from screening four different cDNA libraries with the original insert or with a 5' probe were sequenced and all found to contain at most 0.6 kb of 3' sequence. Since it was possible that the secondary structure of the N.52 RNA was impeding the reverse transcriptase synthesis of cDNA, we next attempted to obtain more 5' sequence by performing reverse transcription at 50C to unfold secondary structure, followed by anchored polymerase chain reaction (APCR) using heat-stable Taq polymerase (Loh et al., 1989). Although this approach allowed us to obtain an additional 120 nucleotides that were enriched in G +C residues, we did not obtain a fulllength cDNA clone. The nucleotide sequence of this partial cDNA contains a single open-reading frame, followed by a stop codon at nucleotide 361, and appears to extend to the 3' end of mRNA since it contains a classic AAAAA polyadenylation signal (Birnstiel et al., 1985) f6llowed by a poly A tail (Fig. 2). This cDNA encodes a polypeptide of 120 amino acids (Fig. 2). A comparison of the DNA and amino acid sequence of clone N.52 with sequences in the GenBank (version 60, July 21, 1989) and Swiss  left). Subsequently, the filters were hybridized with li cDNA probe (Polla et al., 1986) to confirm the stimulatory effects of IL-4 (middle panel), and A50 cDNA (Nguyen et al., 1983) as a control for the amount of RNA loaded in each lane (right panel). (B) Kinetics of induction of N.52 mRNA by IL-4 in R8205 cells. Total cellular RNA was obtained from R8205 cells cultured in the absence or presence of IL-4 for varying time periods. Ten #g of total cellular RNA were loaded in each lane, and the blot was hybridized with N.52 cDNA probe. The probe was then stripped from the filter and hybridized to the A50 control cDNA probe to determine the level of RNA in each lane. In addition to the 0.95-kb N.52 message, higher molecular-weight transcripts are occasionally seen, and may represent nuclear RNA or partially spliced forms of N.52 transcript. protein (version 10, July 21, 1989) databases, using computer programs, revealed no significant similarities. Further attempts to obtain additional 5' sequence will focus on the analysis of a recently obtained genomic clone (Nabari, unpublished data).

N.52 Expression Is Induced by Activating and Differentiating Stimuli
We have shown that IL-4 induces the expression of N.52 mRNA in a pre-B cell line known to be sensitive to IL-4-mediated signals. While IL-4 is known to provide a differentiation signal to B cells, it alone is a relatively weak activator of normal splenic B cells and requires costimulatory factors (Rabin et al., 1985;O'Garra et al., 1986;Snapper and Paul, 1987).
It was therefore not surprising that IL-4 failed to increase N.52 transcripts in normal splenic B cells (not shown). In contrast, treatment of splenic B cells with bacterial lipopolysaccharide (LPS), a potent polyclonal B-cell activator, induced N.52 transcripts in athymic spleen cells, which consist predominantly of B cells (Fig. 3A). N.52 transcripts were induced after 4 hr of exposure to LPS and remained high after 18 hr, the latest time point tested. Because of the effect of LPS stimulation of B cells on levels of N.52 transcripts, it was of interest to determine whether stimulation of another lymphoid population, T cells, also affected N.52 expression. Northern analysis demonstrated that two antigen-specific T-cell hybrids (Glimcher and Shevach, 1982), created by fusing antigen-specific normal T cells with the BW5147 thymoma, la:ked N.52 transcripts at baseline (Fig. 3B). The two T-cell hybrids were therefore treated with concanavalin A (Con A), which has been shown to induce the differentiation of splenic T cells into lymphokine-secreting cells (Farrar et al., 1980;Granelli-Piperno et al., 1984). Con A treatment of the two T hybrids induces high levels of N.52 mRNA (Fig. 3B), whereas similar treatment of splenic T cells failed to induce N.52 (not shown). These results suggest that splenic T cells may have passed the responsive stage for N.52 induction.
Since our data suggested that N.52 expression in lymphoid cells was inducible by differentiation stimuli, we were interested to see whether or not differentiation signals in nonlymphoid cells also induce N.52 expression. The F9 teratocarcinoma cell line is one such system in which differentiation can be triggered in vitro by a combination of cAMP and retinoic acid treatment (Strickland et al., 1980). Figure 3C demonstrates that these differentiation events coincide with .the appearance of N.52 transcripts. Therefore, the induction of N.52 mRNA in response to differentiation signals is not. limited to cells of the lymphoid lineage. Furthermore, these results suggest that N.52 expression accompanies active differentiation in different cell types. N.52 Expression in Normal Mouse Tissue Hox.-1.4 probe, a homeobox-containing gene expressed abundantly in the adult mouse testis RNA extracted from freshly isolated mouse tissue (Wolgemuth et al., 1986, 1.987) (Fig. 4, right panel). was examined by Northern blot analysis (Fig. 4). The results shown in Fig. 4 demonstrate that N.52 N.52 expression was undetectable in highly differentranscripts are significantly more abundant in testis tiated organs such as brain, heart, lungs, liver, and than Hox-l.4 transcripts. Subsequent hybridization kidney. In contrast, N.52 mRNA was present in of these blots to control probe, A50, indicated that thymus, bone marrow, and day-17 fetal liver, organs comparable levels of RNA were present in each harboring actively developing and differentiating lane. The expression of N.52 in testis is parallel to cells, the expression of the homeobox gene  Most striking, however, was the extremely high (Wolgemuth et al., 1986), except that the level of level of expression of N.52 in testis, an organ comexpression of N.52 is much higher than Hox-l.4. posed both of differentiating germ cells and somatic This may be due to the greater potency of the N.52 cells. N.52 mRNA was not detectable in ovary (not promoter/enhancer sequences in germ cells, shown). To investigate the relative level of expresalthough other possibilities such as RNA stability sion of N.52, testis RNA was hybridized with the effects cannot be ruled out.  For comparison of the level of N.52 mRNA with Hox-l.4, a gene with known expression in testis, the lane containing testis RNA was subsequently hybridized with the Hox-l.4 (0.8 kb Pst/Hind III genomic fragment (Wolgemuth et al., 1987). Exposure was 3 days at -70C. The autoradiographs are not lined up, as the Hox-l.4 mRNA is 1.3 to 1.4 kb in length. The blots were subsequently hybridized to A50 control probe.

Expression of N.52 in Testis Is Limited to Germ Cells
The extremely high level of N.52 transcripts in testis prompted experiments to determine which testicular cell type(s) is involved in this expression. Two different and complementary approaches were used. Both neonatal and adult testes have a full complement of somatic cells, which include Leydig, Sertoli, peritubular, and macrophagelike cells. Germ-cell development in the testis is not complete, however, until approximately 4 weeks of age. We compared the level of N.52 transcripts in testes from athymic mice and from normal C3H mice, day 2, day 14, and adult (4 and 8 weeks) animals (Fig. 5A). In the athymic mice, N.52 transcripts were not detected in immature testes (day 2 postnatal), were present at very low levels in developing testes (day 14), and at very high levels in adult testes. In C3H mice, levels of N.52 transcript were already almost maximal by day 14 from three different spermatogenic cell populations including pachytene spermatocytes, early spermatids, and the residual bodies and cytoplasmic fragments. Samples A and B refer to two different concentrations of RNA ("30 and 60 g respectively).
Subsequently, the blot was hybridized with an actin cDNA probe. Exposure 7 days. The relatively low level of signal is presumably due to the fact that this blot had been hybridized and stripped of probe several times. normal animals, and this developmental retardation may extend to germ-cell maturation. This experiment suggested that the expression of N.52 was likely to be correlated with differentiating germ cells, and that the appearance of N.52 transcripts coincided with the beginning of meiosis. To test this more rigorously, RNAs were prepared from testes of a mutant strain of mice, W/Wv, whose testes are germ-cell-deficient. Gonads from mice homozygous for the W alleles (W/Wv) are virtually devoid of germ cells, and histological examination of W/W testes reveals normal, somatic cells but few or no identifiable germ cells (Coulumbre and Russell, 1954). RNA was also isolated from testes of sexually mature heterozygous siblings W//, and from adult Swiss Webster mice. Total and poly A / RNA prepared from testicular tissue of the fertile animals contained N.52 transcripts (Fig. 5B, first four lanes). In contrast, no N.52 transcripts were detected in either total or poly A / RNA from the germ-celldeficient W/W testes (Fig. 5B, last two lanes). These results indicate that the expression of N.52 transcripts in testes is germ-cell-specific since all of the somatic cell types are present in the testes of homozygous W/W animals.
The quantitative differences in expression of N.52 transcript between immature and mature testis (Fig.  5A, compare day 2 to 4 weeks) suggest that .its appearance correlates with meiotic stages of spermatogenesis. To determine the developmental stage at which the N.52 transcript appears, cells from adult testis were purified by sedimentation at unit gravity . RNAs were isolated from three different cellular populations, including the meiotic prophase spermatocytes (predominantly in pachytene stage), the early postmeiotic spermatids, and a fraction that included residual bodies and cytoplasmic fragments of elongating spermatids, and were probed for expression of N.52 mRNA (Fig. 5C, first four lanes). N.52 was not expressed in testes that contain only premeiotic germ cells (see Fig. 4A, 2d postnatal). N.52 transcript was present in germ cells in meiotic prophase (pachytene) and in ceils that were further advanced (spermatids) in the developmental pathway of spermatogenesis. No striking differences in levels of N.52 were seen in the different stages, including cytoplasmic fragments and residual bodies (Fig. 5C), when RNA loading in each lane was normalized. These data suggest that N.52 expression is likely to be induced by differentiation signals at the onset of meiosis and remains high during all the stages of germ-cell development.
A  The N.52 Gene Is Highly Conserved Genomic Southern blot analysis of mouse DNA digested with a panel of restriction endonucleases probed with a 0.6-kb cDNA insert (Fig. 6) revealed a fairly simple pattern consistent with the presence of a single copy gene. A cross-hybridizing species in human DNA remained easily detectable even under stringent hybridization and wash conditions (Fig. 6), suggesting some degree of evolutionary conservation of N.52 across species.

DISCUSSION
Despite the multiplicity of effects of IL-4 on various cell types, very little is known about the molecular mechanisms by which IL-4 exerts these effects. Those genes that are known to be induced by IL-4 include class II MHC (Polla et al., 1986), Fc epsilon receptor (Hudak et al., 1987), Thy-1 (Snapper et al., 1988), and mou,se pancreatic lipase (Grusby, unpublished data). In this report, we describe the isolation of an IL-4 inducible gene, designated N.52, by differential and subtractive hybridization screening of a murine pre-B cell cDNA library.
N.52 is a novel gene that demonstrates several interesting features. First, it is inducible by IL-4, which is known to initiate a differentiation program in pre-B cells (Hofman et al., 1988) and by other polyclonal activators such as Con A and LPS that trigger differentiation in lymphoid cells (Granelli-Piperno et al., 1984;Rothman et al., 1988). Second, it is expressed in bone marrow, thymus, testis, and fetal liver, organs harboring rapidly growing and differentiating stem cells. Third, the very high level of N.52 expression in developing germ cells, and its inducibility in response to differentiation signals in F9 cells in lymphoid cells demonstrate that N.52 expression accompanies differentiation events in widely disparate cell types.
Mammalian cells respond to mitogenic and differentiating stimuli by sequential activation of selected genes. Immediate early response genes such as the c-fos, c-jun, and myc protooncogenes are transcriptionally activated within minutes after exposure to the relevant stimuli (Kelly et al., 1983;Greenberg and Ziff, 1984;Ryder et al., 1988). Activation of this set of genes is followed by activation of a second set (Kwon et al., 1987) (such as transferrin and interleukin 2 receptor genes) (Kr6nke et al., 1985) whose transcripts accumulate later and remain high for more extended periods. The kinetics of N.52 induction in cells of B lineage in response to external stimuli resemble the latter group. Although N.52 transcripts are detectable in unstimulated R8205 pre-B cells, exposure to IL-4 begins to increase steady-state mRNA levels by 4 hr. N.52 transcripts increase and are maintained at maximal level for up to 72hr after IL-4 exposure. These kinetics are similar to the induction of the class II MHC genes by IL-4 in both the R8205 pre-B cell line and in normal splenic B cells (Noelle et al., 1986;Polla et al., 1986). Another similarity between N.52 and class II MHC expression in the B lineage is the absence of expression of both these gene products in myelomas, a terminally differentiated B cell state (Nabavi, unpublished data).
Activation and differentiation of lymphoid cells is modulated through complex interactions of numerous extracellular stimuli, which may synergize or interfere with IL-4 mediated signals, depending on cell type and developmental stage Paul and Ohara, 1987)! These stimuli include LPS, which causes blast transformation of resting B cells and. drives them to Ig production (Snapper and Paul, 1987;Lutzker et al., 1988;Rothman et al., 1988), and Con A, which activates expression of some genes, including lymphokines and lymphokine receptors in resting T cells (Farrar et al., 1980;Kr6nke et al., 1985;Burd et al., 1987;Kwon et al., 1987). Some other stimuli include PMA, cIgM, and lymphokines (Farrar et al., 1980;Granelli-Piperno et al., 1984;Kr6nke et al., 1985;Rabin et al., 1985). N.52 expressionis induced by a subset of these activation and differentiation signals, including IL-4, LPS, and Con A, but not by PMA (not shown). The ability of a prticular stimulus to induce N.52 expression varies with cell type and developmental stage, as LPS, but not IL-4, can induce N.52 in resting B cells. IL-4, however, can induce N.52 in R8205 cells, which are transformed cells in the pre-B cell stage of development.
Similarly, Con A induced N.52 expression in T hybrids, but not in splenic T cells (not shown). It is possible that splenic T cells have already passed the responsive stage for N.52 induction. Alternatively, Con A may induce N.52 in the T hybrids possibly by acting upon the BW5147 thymoma fusion partner. It will be interesting to identify the critical stimuli necessary for N.52 induction at the various stages of Band T-cell development.
The overlapping expression of N.52 within the lymphoid and reproductive compartments is not surprising since this has been shown for several gene products. Class II MHC antigens and CD4-1ike molecules have been reported on male germ cells (Ashida and Scofield, 1987;Bishara et al., 1987). In addition, an IL-lc-like soluble factor has been found in testicular interstitial fluid (Gustafsson et al., 1988). Other lines of evidence also link the immune and reproductive systems physiologically, via direct or indirect hormonal interactions. Differences in humoral and cell-mediated immune responses in male and female animals have been attributed to the effects of the sex hormones on the immune system (Paavonen et al., 1981). The effects of sex hormones in thymic development have been demonstrated; for example, castration of both male and female animals results in spleen and thymus hyperplasia and an increase in peripheral lymphocyte counts (Eidinger and Garrett, 1972;Allen et al., 1984). Reciprocal effects of the thymus on reproductive development have also been noted (Strich et al., 1985). Congenitally athymic mice show reproductive defects, and these defects can be corrected by injection of thymosin or by thymus grafts (Strich et al., 1985). In view of the thymus-reproductive system interaction, our observation of the lag time in the expression of N.52 in testes of 2-week-old athymic mice compared to the age-matched normal animals is interesting. However, the significance of this observation is not clear at present. The process of spermatogenesis involves a series of cellular differentiation steps in which spermatogenic stem cells undergo functional and morphological specialization to form spermatozoa. Identification of stage-specific gene products is of particular interest in understanding this complex pathway of cellular differentiation. Some of these gene products are expressed at specific stages and thus presumably involved in basic cellular functions underlying morphological changes that occur during spermatogenesis, whereas others, such as the protooncogenes c-abl , c-mos (Goldman et al., 1987), pim-1 (Sorrentino et al., 1988), and int-1 (Schackleford and Varmus, 1987), have been implicated in postmeiotic stages of spermatogenic differentiation. Several nuclear protooncogenes such as c-myc and c-fos, which are activated during germ-cell differentiation, are also induced by different stimuli during lymphocyte differentiation (Klemsz et al., 1989;Wolfes et al., 1989). Protooncogenes c-mos and c-raf are members of the serine and threonine kinase families, respectively, with overlapping expression in somatic and germ cells (Wolfes et al., 1989). Although the function of the N.52 gene product is not yet known, its induction by different stimuli (sex hormones, lymphokines, polyclonal activators) that act through different signal transduction pathways suggests that it may function in a common final pathway of cellular growth and clifferentiation. Further, its presence in such different cell types suggests that phenotypically and ontologically distinct cells may utilize overlapping molecular pathways for differentiation.

MATERIALS AND METHODS
cDNA Synthesis and Cloning R8205 cells were grown in the presence of 800 U of rIL-4/ml (Immunex Corp., Seattle, Washington) for 48 hr. R8205 cells stimulation by IL-4 was confirmed by surface class II MHC expression, as determined by flow cytometry analysis (Polla et al., 1986). Poly A / RNA was prepared by oligo dT cellulose chromatography of total RNA, isolated by the guanidine isothiocyanate method (Chirgwin et al., 1979). Firststrand cDNA was synthesized from 4/g of poly A /