P94fer Facilitates Cellular Recovery of Gamma Irradiated Pre-T Cells

S Halachmy, O Bern, L Schreiber, M Carmel, Y Sharabi, J Shoham and U Nir

Department of Life Sciences, Bar-Ilan University, Ramat-Gan 52900, Israel p94fer is a ubiquitous, nuclear and cytoplasmic tyrosine These three kinases encompass a group of the known, kinase, whose accumulation has been demonstrated in all SH2 containing mammalian nuclear tyrosine kinases mammalian cell lines analysed. In the present work, the (Yates et al., 1995; Wilks and Kurban, 1988; Van Etten p94fer expression pro®le was determined in cell lines et al., 1989; Hao et al., 1989, 1991) which are not src which were not tested before. While being present in related. While p94fer and c-Fes share similar structures, several hematopoietic and non hematopoietic cell lines containing a SH2 domain which is ¯anked by an including thymic stromal cells, the p94fer kinase could not extended N-terminal tail (Wilks and Kurban, 1988; be detected in pre-T and T cell lines. p94fer was also Hao et al., 1989), the c-Abl kinase contains both SH2 absent in pre-B line, but accumulated in these cells upon and SH3 domains and an extended C-terminal portion their induced development to antibody producing cells. (Ren et al., 1994). Several functional motifs have been This is in agreement with the absence of p94fer in primary de®ned in the Abl c-terminus, including a nuclear thymic and splenic T lymphocytes and its induced localization signal (NLS) (Van Etten et al., 1989), accumulation in stimulated B cells. Relatively high DNA binding domain (Kipreos and Wang, 1992), actin p94fer levels were detected in primary thymic and splenic binding domain (Van Etten et al., 1994; McWhirter stromal cells. Ectopic expression of p94fer in pre-T cells and Wang, 1993) and binding sites for the adapter slightly aected their cell cycle pro®le but it did not proteins Crk, Grb-2 and Nck (Ren et al., 1994). The aect their apoptotic death which was induced by N-terminal tail of p94fer exerts a coiled-coil structure ionizing radiation. However, p94fer facilitated dramati- which mediates the interaction of that kinase with cally, the cellular recovery of gamma irradiated pre-T other cellular proteins (Kim and Wang, 1995). Unlike cells which have escaped the apoptotic death. The c-Fes whose expression is con®ned to hematopoietic enhanced recovery of the irradiated, p94fer expressing progenitor cells (MacDonald et al., 1985; Care et al., pre-T cells, resulted most probably from their increased 1996), mature cells of the myeloid lineage (Hanazono survival, rather than from a prominent change in their et al., 1993; Greer et al., 1990; Feldman et al., 1985; proliferation rate. The absence of p94fer from pre-B and Care et al., 1996) and vascular endothelial cells (Greer pre-T cells, may thus contribute to the relative sensitivity et al., 1994), p94fer (Letwin et al., 1988; Hao et al., of these cells to ionizing radiation and to their 1989; Feldman et al., 1986) and c-Abl (Wang and dependence on the functioning of other nuclear tyrosine Baltimore, 1983; Van Etten et al., 1989; Renshaw et al., kinasese. 1988; Muller et al., 1982; Boulter and Wagner, 1988; Bernards et al., 1988) were reported to be ubiquitously Keywords: ionizing-radiation; nuclear; p94fer; T cell; expressed in all tissues analysed. Although diering in tyrosine kinase their tissue distribution, the subcellular localization of c-Fes and p94fer is similar and they accumulate both in the cytoplasm and nucleus of the expressing cells (Yates et al., 1995; Hao et al., 1991). While the cellular Introduction role of the c-Fes nuclear fraction is not understood, its cytoplasmic fraction can promote proliferation of p94fer is an evolutionary conserved (Pawson et al., vascular endothelial cells (Greer et al., 1990) and has 1989), nuclear and cytoplasmic tyrosine kinase (Hao et been implicated in cytokine signal transduction in al., 1991) whose presence has been documented in all hematopoietic cells (Izuhara et al., 1994; Hanazono et mammalian cell lines analysed (Letwin et al., 1988; al., 1993). The cytoplasmic fraction of p94fer interacts Hao et al., 1989; Feldman et al., 1986). A truncated with the catenin-like substrate pp120 in ®broblastic form of p94fer, termed p15ferT, accumulates solely in cells (Kim and Wang, 1995) and its nuclear fraction, meiotic primary spermatocytes (Hazan et al., 1993; was found to associate with the cell chromatin (Hao et Fischman et al., 1990) and is encoded by a testis al., 1991). The kinase activity of the cytoplasmic p94fer, speci®c FER transcript (Keshet et al., 1990; Fischman is elevated in growth factors stimulated ®broblastic cell et al., 1990). p52ferT and p94fer dier in their N-termini lines (Kim and Wang, 1995). However, no direct link but they do share a common SH2 and kinase domains of p94fer to regulation of cell growth has been shown, (Hao et al., 1989; Fischman et al., 1990). The p94fer nor has a p94fer substrate been characterized, thus kinase domain (Letwin et al., 1988; Hao et al., 1989) is leaving the cellular role of p94fer elusive. The 70% and 50% homologous to the kinase domains of ubiquitous expression of p94fer suggest that it may two other known nuclear tyrosine kinases c-Fes (Yates have a function that is fundamental to all cells. The et al., 1995; Wilks and Kurban, 1988; Roebroek et al., cellular levels of p94fer dier however among various 1985) and c-Abl (Van Etten et al., 1989), respectively. cell types (Feldman et al., 1986). While ®broblastic (Hao et al., 1989, 1991), epithelial (Hao et al., 1989, Correspondence: U Nir 1991) and macrophage (U Nir, unpublished data) cell Received 8 August 1996; revised 11 March 1997; accepted lines accumulate high p94fer levels, the expression of 11 March 1997 that tyrosine kinase in some promyelocytic cell lines is Cellular recovery of irradiated pre-T cells SHalachmyet al 2872 relatively low (Hao et al., 1989). In an attempt to et al., 1980). The 70Z/3 dierentiation system was thus elucidate the cellular functions of p94fer, the determina- adopted for checking the dependence of p94fer tion of its accumulation pro®le has been extended to accumulation, on the development of pre-B cells to cell types which have not been analysed before. In this antibody producing B cells. 70Z/3 cells were exposed to work we show that p94fer is not detected in pre-B, pre-T 10 mg/ml LPS and their p94fer accumulation pro®le was and T cell lines and that ectopic expression of p94fer in followed for up to 68 h of LPS treatment. Exposing the pre-T cells, dramatically improved their recovery after 70Z/3 pre-B cells to LPS treatment, led to gradual being exposed to ionizing radiation. The absence of accumulation of p94fer in the maturing cells (Figure 2). p94fer from pre-B and pre-T cells, may thus contribute The kinetic of p94fer accumulation lagged however, to their relative sensitivity to ionizing radiation and to behind the known kinetic of the k light chain the dependence of these cells on the functioning of production. While the k protein level reaches a other ubiquitous nuclear tyrosine kinases, like c-Abl (Tybulewicz et al., 1991; Schwarzberg et al., 1991).

a S C2 R1 NIH TEC 127 B Results

p94fer is not detected in pre-B, pre-T and T lymphocyte fer cell lines p94 ¨ To extend the previous analysis of the p94fer expression pro®le in various cell types (Letwin et al., 1988; Hao et al., 1989), the relative cellular levels of p94fer were determined in ®ve mouse cell lines which have not been 1234567 analysed before. These included two hematopoietic cell lines: R1 ± a friend erythroleukemia cell line (Marks et al., 1983) and cell line No 127 ± a mouse double positive pre-T thymoma (Irlin and Peled, 1992). The three non hematopoietic cell lines tested were: C2 ± a b 70Z TEC 127 T8 T4 NIH mouse myogenic cell line (Yae et al., 1996), STAC and TEC ± two thymic stromal epithelial cell lines p94fer ¨ (Glimcher et al., 1983). The p94fer levels in these cells were compared to those found in the previously analysed NIH3T3 mouse ®broblastic cell line and the human B cell line DAKIKI (Hao et al., 1989). Whole cell protein extracts prepared from the cultured cells, were fractionated in a SDS ± PAGE, Western blotted and the presence of p94fer was determined with the C1 p94fer antibodies. All the cell lines analysed except for 123456 127 (Figure 1a lane 6), were found to accumulate fer varying levels of the p94fer kinase. The absence of p94fer Figure 1 Accumulation of p94 in mammalian cell lines. (a) Whole cell proteins extracts from: lane 1, STAC cells; lane 2, C2 from the 127 pre-T cells, prompt us to check whether it myogenic cells; lane 3, R1 erythroleukemia cells; lane 4, NIH3T3; is expressed in pre-B and mature T cell lines. The lane 5, TEC cells; lane 6, pre-T 127 cells; and lane 7, DAKIKI B presence of p94fer was thus analysed in three other cells; were resolved in a 9% SDS ± PAGE, Western blotted and reacted with the C1 p94fer antibodies. Arrow on the left indicates mouse lymphocytic cell lines. These included the fer fer + the migration distance of p94 .(b) p94 accumulation in: lane 1, 2B4.11 cell line ± a CD4 T-cell hybridoma (Yang et 70Z/3 cells; lane 2, TEC cells; lane 3, 127 cells; lane 4, CTLL cells; al., 1993), CTLL ± a CD8+ IL-2 dependent cytotoxic T lane 5, 2B4.11 cells; lane 6, NIH3T3. Arrow on the left indicates derived cell line and the 70Z/3 mouse pre-B cell line the migration distance of p94fer (Sokaguchi et al., 1980). p94fer could not be detected in the pre-B and pre-T cell lines as well as in the two T cell lines analysed (Figure 1b, lanes 1, 3, 4 and 5). p94fer was clearly detected however in the TEC and NIH3T3 4h 24h 36h 48h 68h K NIH 70Z cells (Figure 1b lanes 2 and 6). Two additional bands of lower molecular weights were occasionally detected p94fer ¨ in the analysed samples (Figure 1b). These bands were not detected with antibodies directed against the unique N-terminus of either p94fer or p51ferT (data not shown). These bands seem thus to represent non FER proteins which cross react with the C1 antibodies. The 12345678 fer absence of p94 from pre-B cells (Figure 1b, lane 1) Figure 2 Induced accumulation of p94fer in 70Z/3 cells. Proteins and its presence in B cells lines (Hao et al., 1989 and were extracted from 70Z/3 cells treated with LPS for: lane 1, 4 h; Figure 1a lane 7), indicate the possible accumulation of lane 2, 24 h; lane 3, 36 h; lane 4, 48 h; lane 5, 68 h; and from: p94fer concomitantly with the maturation of pre-B cells lane 6, K562 cells; lane 7, NIH3T3 cells; lane 8, untreated 70Z/3 cells. 30 mg proteins extracted from the same number of cells were to antibody producing B cells. LPS was shown resolved in each lane of 9% SDS ± PAGE, Western blotted and previously to induce the production of immunoglobu- reacted with p94fer antibodies. Arrow on the left indicates the lin k light chain in the 70Z/3 pre-B cell line (Sokaguchi migration distance of p94fer Cellular recovery of irradiated pre-T cells S Halachmy et al 2873 maximum after 18 ± 20 h of LPS treatment (Sen and mature T cells were prepared from a spleen of a 4 Baltimore, 1986 and our unpublished data), the week-old mouse and were further puri®ed on a CD90 induced accumulation of p94fer in 70Z/3 cells was adsorption column. A fraction of these mature T cells observed after 24 h of LPS treatment and reached its was stimulated with ConA (Wadsworth et al., 1995) maximal level only after 48 h (Figure 2). and was further propagated in culture. Splenic B cells To verify the lack of the FER gene expression in the were also puri®ed from the splenic cells population by pre-T lymphocytic cell line, Northern blot analysis was being absorbed on a CD45 column. The B cells were carried out. RNA was extracted from two mouse either left untreated or stimulated with 25 mg/ml LPS ®broblastic cell lines: NIH3T3 and Ltk7 and from the (Sen and Baltimore, 1986). Enriched thymic and mouse pre-T cell line No. 127. The RNA was splenic stromal cell populations were obtained from fractionated in a formamide/formaldehyde gel, North- irradiated mice after lymphocytes removal (Eshel et al., ern blotted and reacted with the mouse fer cDNA 1990). Whole cell protein extracts were prepared from probe. While a 3 kb long fer transcript (Letwin et al., equal numbers of T cells before and after stimulation 1988; Hao et al., 1989) was clearly detected in NIH3T3 and from thymic and splenic stromal cells. Similarly, and Ltk7-cells, no transcript of that size could be equal numbers of B cells were taken for protein detected in the 127 pre-T cells (Figure 3) even after extraction, before and after LPS stimulation. Proteins long exposure of the nitrocellulose membrane (data not were separated in SDS ± PAGE, Western blotted and shown). reacted with the C1 p94fer antibodies. While relatively high p94fer levels were detected in enriched thymic and splenic stromal cells (Figure 4a), no 94 kD protein p94fer is detected in stromal but not in nonstromal reacted with the p94fer antibodies in extracts prepared primary thymic cells from untreated thymocytes (Figure 4b, lane 11). Nor The absence of p94fer from the pre-T and T cell lines was the 94 kD tyrosine kinase detected in T cells and its presence in the STAC and TEC thymic stromal cell lines, predicted the lack of its expression in primary thymocytes and the accumulation of that tyrosine kinase in thymic stromal cells. The relative p94fer levels were therefore determined in primary, immature and a mature T-cells and in thymic stromal cells. Thymocytes T S were prepared from 4 week-old mice and were puri®ed on a CD90 adsorption column. 90% of the cells which comprise the thymocyte population, are CD4, CD8 p94fer ¨ double positive premature T cells. Single positive

a 1 2 fer ¨

b 123 B cells: +++++ Stim: – 24h 48h 72h 96h FIB NIH b ¨

1234567

28S ¨ T cells: ++++ Stim: – 24h 48h THY FIB NIH ¨

18S ¨ 8 9 10 11 12 13 Figure 4 Accumulation of p94fer in thymic and splenic cells. (a) 30 mg proteins from: lane 1, thymic stroma; lane 2, splenic stroma were resolved in 9% SDS ± PAGE, Western blotted and reacted with p94fer antibodies. Arrow on the left marks the migration distance of p94fer.(b) Proteins were extracted from: lane 1, splenic 123 B cells; lane 2, B cells stimulated with LPS for 24 h; lane 3, B cells Figure 3 Tissue speci®c expression of the fer mRNA. (a) Whole stimulated for 48 h; lane 4, B cells stimulated for 72 h; lane 5, B cell RNA was extracted from: lane 1, Ltk7 cells; lane 2, NIH3T3 cells stimulated for 96 h; lane 6, Primary new born mouse cells; and lane 3, pre-T 127 cells. 30 mg RNA were resolved in a ®broblasts; lane 7 NIH3T3 cells; lane 8, splenic T cells; lane 9, T 1.2%, formamide/formaldehyde agarose gel, Northern blotted cells stimulated with ConA for 24 h; lane 10, T cells stimulated and reacted with 32P-labeled fer cDNA probe. Arrow on the left for 48 h; lane 11, non stromal thymic cells; lane 12, mouse marks the migration distance of the 3 kb for mRNA. (b) primary ®broblasts; lane 13, NIH3T3 cells. The proteins were Ethidium bromide staining of the RNA samples presented in Western blotted and reacted with p94fer antibodies. Arrow on the part (a) left marks the migration distance of p94fer Cellular recovery of irradiated pre-T cells SHalachmyet al 2874 stimulated with ConA for 24 or 48 h (Figure 4b, lanes p94fer facilitates cellular recovery of gamma irradiated 8 ± 10). p94fer was also absent in puri®ed splenic B cells pre-T cells (Figure 4b, lane 1) but accumulated in cells which were stimulated with LPS for 48 h (Figure 4b, lane 3). The The association of p94fer with chromatin (Hao et al., p94fer level peaked at 96 h of LPS treatment (Figure 4b, 1991), indicate its possible involvement in cellular lane 5), but it was still signi®cantly lower than the pathways which are linked to the inspection of p94fer levels detected in primary new born mouse chromosomal DNA functioning and integrity. Since ®broblasts (Figure 4b, lane 6). That time point (96 h) pre-T cells are characterized by their apoptotic could not be followed in stimulated T cells since these response to certain genotoxic stress (Strasser et al., cells underwent extensive cell death after residing in 1994), we sought to check whether p94fer could aect culture for more than 72 h. the cellular response of 127 cells, to that kind of treatment. L9, L2/3 and NEO cells were exposed to 950 rad g radiation, which induces apoptotic death in fer in pre-T 127 cells Ectopic expression of p94 up to 90% of 127 cells, within 4 days (Schreiber et al., The availability of cell lines devoid of p94fer,or 1996). The viability and apoptotic death of the three expressing very low levels of that enzyme, oer clones, was followed for up to 14 days post irradiation. valuable tool for exploring the cellular functions of All the three cell lines tested, exerted similar apoptotic that tyrosine kinase. We therefore attempted to follow response, during the ®rst 4 days which followed the the eects of ectopically expressed p94fer, on the growth irradiation exposure time. More than 80% of the cells characteristics of transfected pre-T cells. The mouse fer in all clones, underwent apoptosis by the 4th day after cDNA was inserted into the pECE vector (Ellis et al., irradiation (Figure 8). The number of viable L9, L2/3 1986) in which it is constitutively transcribed under the and NEO cells diered however at the 4th day after control of the SV40 early promoter. Pre-T 127 cells were cotransfected with the newly constructed plasmid, termed pECEfer and the SVneo plasmid which can a confer G418 resistance to the transfected cells (South- ern and Berg, 1982). G418 resistant clones were selected and their cellular p94fer levels were determined. Several pre-T 127 clones diering in their p94fer production levels were isolated and further propagated. The ectopically expressed p94fer exerted in vitro autokinase activity when immunoprecipitated from the transfected cells (data not shown). The cellular p94fer levels in two 127 clones, is shown in Figure 5. These clones termed L9 and L2/3 were independently isolated from two separated rounds of transfection. The two clones exerted relatively high and stable p94fer expression levels (Figure 5, lanes 3 and 4). The L9 and L2/3 clones were thus selected for further analysis of the p94fer cellular eects, on the transfected pre-T 127 cells. In an attempt to follow the cellular eects of p94fer on 127 cells, basic growth parameters of the L9 clone were compared to those of the NEO 127 cells which harbor the G418 resistance gene but lack the exogenous fer cDNA. The cell cycle pro®les of both L9 and NEO cells were found to be similar (Figure 6a). While 40% of the proliferating NEO cells resided in

the G1 phase, 44% of the L9 cells resided in that stage of the cell cycle (Figure 6b). These values re¯ect the almost identical growth curves of the two clones (S Halachmy, data not shown).

NEO NIH L9 L2/3 b

p94fer ¨

Figure 6 Cell cycle pro®les of 127 cells. (a) 127 (1), NEO (2) and 1234 L9 (3) cells were exposed to ¯ow cytometry analysis. Black areas denote G0/G1 (left) and G2/M (right) populations. Grey area Figure 5 Ectopic expression of p94fer in pre-T 127 cells. 30 mg represents the population of cells residing in S phase. (b) protein from: lane 1, NEO cells; lane 2, NIH3T3 cells; lane 3, L9 Percentage of cells residing in each one of cell cycle stages. The cells; Lane 4 L2/3 cells, were Western blotted and reacted with values were derived from the pro®les presented in (a) by using the p94fer antibodies. Arrow on the left indicates the migration Multicycle program for evaluation of ¯ow cytometry cell cycle distance of p94fer pro®les Cellular recovery of irradiated pre-T cells S Halachmy et al 2875 irradiation (Figure 7a), suggesting that the NEO cells and non expressing 127 cells. The augmented recovery underwent a necrotic as well as an apoptotic death. of the p94fer expressing cell lines, could result from The viability of the L9, L2/3 and NEO cultures was either an increased proliferation rate of the two clones, further followed, for monitoring the recovery pro®le of or from a higher viability of the p94fer expressing cells, the irradiated cells. While no signi®cant dierence at the post irradiation recovery stage. In an attempt to could be observed in the apoptotic death of irradiated discriminate between these two possibilities, the cell pre-T cells which either express or do not express p94fer cycle pro®les of the NEO, L2/3 and L9 cells were (Figure 8), these cells diered dramatically in their determined at dierent times after irradiation by using delayed recovery pro®le. The number of viable L9, L2/ ¯ow cytometry analysis of cells stained with PI and 3 and NEO cells started to increase at the 5th day after BrdU. The distribution of the recovering NEO and L9 4 4 irradiation (Figure 7a). While 3.1610 and 2.5610 cells, between the G1 and S phase did not vary viable cells/cc were counted in the recovering L9 and signi®cantly until the 7th day after irradiation (Figure L2/3 cultures respectively, the NEO culture contained 8). The cell cycle pro®le of the NEO and L9 cells only 46103 viable cells/cc, at the 5th day after slightly diered however at the late recovery stage. irradiation (Figure 7a). Since recovering L9 and L2/3 While 50% of the NEO cells resided in S phase at the cells reached their stationary growth phase 12 days 8th day after irradiation, 59% of the L9 cells resided in after irradiation (data not shown), the numbers of the same cell cycle stage at that day (Figure 8). At the viable cells in the three tested clones, were followed up 11th post irradiation day, 58% of the NEO cells and to the 11th day post irradiation (Figures 7 and 8). At 67% of the L9 cells, resided in S phase, respectively that day, the L9 and L2/3 cultures contained 6.256106 (Figure 8). Similar results were obtained when the cell and 4.146106 viable cells/cc respectively, and the NEO cycle pro®les of L2/3 and NEO cells were compared cells contained only 36104 viable cells/cc (Figure 7a). (data not shown). The viability percentage of the The same results were obtained when the recovery irradiated-NEO and L9 or L2/3 clones, diered pro®les of the L9 and L2/3 cells were compared to the signi®cantly at all recovery stages (Figure 7b). While recovery pro®le of another independently isolated NEO 90% of the L9 and L2/3 cells excluded Trypan blue at clone (data not shown). These results were obtained in the 10th post irradiation day, only 30% of the NEO ®ve out of ®ve independent experiments thus establish- cells were viable at that stage (Figure 7b). The ing the dierent recovery pro®les of p94fer expressing decreased viability of the NEO cells did not re¯ect their entrance into late apoptotic death, since the percentage of NEO and L9 cells which underwent apoptosis at days 6, 8 and 11 after irradiation, were similar (Figure 8).

Discussion

Three SH2 containing non src related nuclear tyrosine kinases, have been described in mammalian cells. These include the p94fer (Hao et al., 1991), c-Fes (Yates et al., 1995) and the c-Abl (Van Etten et al., 1989) tyrosine kinase. The three tyrosine kinases dier in their tissue distribution but no mammalian cell devoid of all three nuclear kinases, has been described. To date, only p94fer (Letwin et al., 1988; Hao et al., 1989; Feldman et al., 1986) and c-Abl (Boulter and Wagner, 1988; Bernards et al., 1988) were shown to accumulate in all mammalian cell types analysed. p94fer was thus detected in various hematopoietic, as well as non hematopoietic cell lines (Pawson et al., 1989; Letwin et al., 1988; Hao et al., 1989, 1991). The hematopoietic cell lines tested for the expression of p94fer, represent most hematopoietic lineages. These include: from the erythropoietic lineage ± the erythroleukemia cell line R1 (Figure 1a lane 3) and the chronic myelogenous leukemia cell line K562 (Figure 2 lane 6 and Hao et al., 1989), from the granulocyte/monocyte lineage ± the promyelocytic leukemia cell line HL60 (Hao et al., 1989) and the macrophage cell line J774 (U Nir, unpublished data) and from the lymphopoietic lineage the human B- lymphocyte cell line DAKIKI (Figure 1a lane 7 and Figure 7 Number of viable cells in irradiated cultures. NEO, L9 Hao et al., 1989). Our present ®ndings, implying that and L2/3 clones were exposed to 9.5 Gy g radiation. The number p94fer is absent, or accumulates at very low levels in (a) and percentage (b) of viable cells was determined for up to 11 days post irradiation. The numbers of viable cells in each clone, pre-B, pre-T and T cells, indicate the possible following irradiation, are plotted in logarithmic scale. Viabilities restriction of p94fer expression, to speci®c cells or were not determined at the 3rd day after irradiation developmental stages in the four hematopoietic Cellular recovery of irradiated pre-T cells SHalachmyet al 2876

a 1 7

2 8

3 9

4 10

5 11

6 12 Cellular recovery of irradiated pre-T cells S Halachmy et al 2877 lineages. This excludes p94fer from being de®ned as an p94fer along the B cells maturation process (Figures 2 ubiquitous tyrosine kinase. Since the expression of c- and 4b), may suggest however, that p94fer is not directly Fes is also restricted to de®ned tissues (MacDonald et involved in the promotion of that process. al., 1985; Hanazono et al., 1993; Greer et al., 1990, The identi®cation of a pre-T cell line devoid of p94fer, 1994; Feldman et al., 1985; Care et al., 1996), this oers a convenient tool for characterizing cellular leaves c-Abl as the only known, ubiquitous nuclear functions of that kinase. Ectopic expression of p94fer in tyrosine kinase, which is expressed in all mammalian pre-T cells did not signi®cantly aect their cell cycle cells (Wang and Baltimore, 1983; Renshaw et al., pro®le nor did it aect their apoptotic response to 1988; Muller et al., 1982; Boulter and Wagner, 1988; ionizing radiation (Figures 7 and 8). It did change Bernards et al., 1988). The absence of p94fer from however dramatically, the recovery pro®le of irradiated lymphoid progenitor cells and the proven presence of 127 cells (Figure 7). While those irradiated NEO c-Fes in T (Izuhara et al., 1994) and p94fer in B (Hao transfected pre-T 127 cells, which have escaped the et al., 1989) cell lines, apparently leaves the pre-T and apoptotic death, barely recovered from the late radiation pre-B cells as the only cells, known to harbor c-Abl eects, cells overexpressing p94fer, returned within 10 (Wang and Baltimore, 1983; Renshaw et al., 1988; days, to their normal growth rate. Interestingly, the p94fer Ben-Neriah et al., 1986) and not c-Fes or p94fer.We expressing cell lines ± L9 and L2/3 and the nonexpressing detected p94fer expression in the mouse teratocarcino- cell lines ± NEO, exhibited similar cell cycle and BrdU men cell line F9 (K Fischman, unpublished data) and incorporation pro®les at the early recovery stages (day 6, in the C2 myogenic (myoblast) cells (Figure 1). Figure 8). The two clone types diered however However the possibility that non lymphoid differen- signi®cantly in their viability during that phase (Figure tiating cells do not express p94fer at some stage of their 7b). These ®ndings strongly suggest that p94fer promotes development was not entirely excluded. What can be the early recovery of the irradiated cells by acting mainly clearly stated at the moment is that pre-B and pre-T as a survival factor. p94fer aected the cell cycle pro®le of cells, are dependent on the functioning of c-Abl. That the irradiated 127 cells at their late recovery stage (days 8 dependence is re¯ected in the intriguing observation of and 11, Figure 8) and led to 16% increase in the number thymic and splenic atrophy, accompanied by prefer- of S phase residing cells (Figure 8). This 16% increase ential loss of B and T lymphoid progenitors, in ablm/m could enhance the late recovery of the irradiated, 127 mice. These mice carry homozygous mutations in the cells. It could not solely account however, for the two c-Abl gene (Tybulewicz et al., 1991; Schwarzberg et order of magnitude dierence, in the late recovery al., 1991). Inactivation of the c-Abl kinase in these pro®les of p94fer expressing and non-expressing cells mice, leads thus to preferential de®ciencies in pre-T (Figure 7a). It seems thus, that p94fer aects the viability and pre-B cells, while all other cells which also express of the irradiated 127 cells, at their late recovery stages, as c-Abl, remain mostly unaected (Tybulewicz et al., well. p94fer did not aect the resistance of 127 cells to 1991; Schwarzberg et al., 1991). We are currently mitomycin c (U Nir unpublished results), thus suggesting checking the possible redundancy in some cellular that it does not activate general repair or resistance functions of c-Abl, p94fer and c-Fes and the possible pathways of these cells, to genotoxic stress. Genotoxic dependence of any proliferating mammalian cell, on stress was shown to activate growth factor receptors the functioning of at least one of these three nuclear tyrosine kinases (Coer et al., 1995), cellular tyrosine enzymes. kinases like p56/p53lyn (Kharbanda et al., 1994) and While being absent in pre-B and T cell lines, we and nuclear tyrosine kinase like c-Abl (Kharbanda et al., others (Figure 1a lane 7 and Hao et al., 1989) detect 1994). In addition, genotoxic stress regimes were found p94fer in B cells. The delayed accumulation pro®le of to activate various unidenti®ed tyrosine phosphorylation processes in treated lymphocytes (Uckun et al., 1993; Schieven et al., 1993). Tyrosine phosphorylation signal transduction events, seem thus to be involved in mammalian cell response to genotoxic stress. The activation of the cytoplasmic p94fer fraction, in growth b factors stimulated ®broblasts (Kim and Wang, 1995), demonstrate the possible integration of p94fer in signal transduction pathways and its envisaged ability to potentiate these processes. p94fer may thus aect the recovery of irradiated pre-T cells, by improving their response to growth and survival factors which are supplied in their growth medium. This notion is supported by our preliminary observations which indicate the impaired sensitivity of irradiated 127 NEO cells, that have escaped apoptotic death, to serum stimulation and their consequent exposure to necrotic Figure 8 Cell cycle pro®les of irradiated 127 clones. (a) Samples death (S Halachmy unpublished data). p94fer seems thus of 9.5 Gy g irradiated NEO 1)-6) and L9 7)-12) cultures, were stained with PI and BrdU at 5 h (1,7), 24 h (2,8) 4 days, (3,9), 6 to compensate this impaired sensitivity and consequently days (4,10), 8 days (5,11) and 11 days (6,12) after irradiation, and leads to the improved survival and recovery of the analysed by ¯ow cytometry. Black areas represent cells residing in irradiated cells. In accordance with that theme, p94fer was G0/G1 (left) and G2M (right). Dark grey represents cells residing also found to increase the viability of non irradiated 127 in S phase and white represents apoptotic cells. (b) Percentage of apoptotic (Apop.) cells and cells residing in the dierent cell cycle cells, under low serum growth conditions (S Halachmy stages were evaluated from ¯ow cytometry cell cycle pro®les (a), unpublished results). This argument is also supported by by using the Multicycle program the fact that exposure of irradiated 127 NEO cells to Cellular recovery of irradiated pre-T cells SHalachmyet al 2878 increased serum concentrations, signi®cantly improved cytes were prepared from 4 ± 6 week-old mice. Pre-T and T their cellular recovery and facilitated their return to cells were puri®ed on a CD90 magnetic cell sorting column normal growth rate (S Halachmy unpublished data). The according to the manufacturers instructions (Miltenyi impaired recovery of irradiated 127 NEO cells does not Biotec, Inc.). The cells were stimulated with 8 mg/ml 6 seem to result, however, from growth factor depletion in ConA at culture density of 2610 cells/ml. B lymphocytes were puri®ed on a CD45 column (Miltenyi Biotec, Inc.). the medium of the irradiated cells, since that medium did The cells were stimulated with 25 mg/ml LPS (Sen and not aect the growth rate of non irradiated 127 cells (S Baltimore, 1986) at culture density of 26106 cells/ml. The Halachmy unpublished data). p94fer seems thus to resuce purity (95%) and proliferation pro®les of the T and B cells the irradiated pre-T cells from late necrotic death, which were determined by ¯ow cytometry analysis. Enriched could result from their abrogated response to external stromal cells were prepared from mice, 48 h after being survival stimuli. An analogous functioning of c-Fes as a exposed to 4.5 Gy g-irradiation. Thymuses and spleens survival factor, was demonstrated in dierentiating were cut into small pieces in serum free DMEM. The HL60 cells. Prevention of c-Fes expression in these fragments were dispersed by being forced in and out of a cells, interferred with their induced granulocytic differ- 3 cc disposable syringe and the supernatant enriched with entiation and caused their apoptotic death (Manfredini thymocytes was discarded. The stromal fragments were washed three times with DMEM. Primary thymocytes and fer et al., 1993). p94 and c-Fes may thus exert, at least splenocytes, 127 cells, The R1-erythroleukemia cell line under de®ned physiological conditions, functions of (Marks et al., 1983), CTLL, 2B4.11 (Yang et al., 1993) and cellular survival factors. However, unlike c-Fes, p94fer 70Z/3 cells (Sokaguchi et al., 1980), were grown in RPMI does not exert anti apoptotic activity and it does not supplemented with 10% fetal calf serum, 1 mM sodium prevent apoptosis in irradiated cells. The ineciency of pyruvate, 2 mM glutamine, 1 mM non essential amino acids p94fer, in preventing the apoptotic death of 127 cells, did and 561072 mM 2-bME. k light chain production in 70Z/3 not result from enhanced degradation of p94fer in early cells was induced by exposing the cells to 10 mg/ml LPS stages of the apoptotic process, since the cellular levels of (Sigma), for up to 68 h. the ectopically expressed p94fer, did not vary signi®cantly between the early and late, post irradiation periods (data Preparation of cell extracts and Western blot analysis not shown). p94fer can enhance however the recovery of cells which have escaped the apoptotic death. The For whole cell protein extracts, cells were lysed with triple absence of p94fer from pre-B and pre-T cells, could detergent lysis buer as was previously described therefore lead to their impaired recovery after irradia- (Sambrook et al., 1989). Cellular extracts (30 mg) were resolved in 9% SDS ± PAGE. Proteins were electroblotted tion. This may contribute to the pronounced sensitivity into a nitrocellulose membrane and incubated with of these cells to ionizing radiation (Strasser et al., 1994). 1 : 10000 dilution of C1 p94fer antibodies. These antibodies In line with our present ®ndings, it will be interesting to were raised against a synthetic peptide derived from the ®nd out whether interference with the p94fer expression in last 15 C-terminal amino acids of the mouse p94fer various cells and tumors could render them more (Fischman et al., 1990) and they do not cross react with sensitive to ionizing radiation. This may oer new c-Fes or c-Abl. The reacting bands were visualized with the strategies for manipulating the sensitivity of cells to ECL Western blotting analysis system (RPN 2106, ionizing radiation. Identi®cation of putative cellular Amersham International). p94fer substrates, should enable one to follow the fer activation of the cytoplasmic and nuclear p94 RNA extraction and Northern blot analysis fractions, under de®ned physiological conditions. This should lead to a better understanding of the role played RNA was extracted from tissue culture cells with by p94fer in controlling cell survival and/or growth. Trireagent (Molecular Research Center, Inc.) according to the manufacturer's instructions. RNA was fractionated in 1.2% formamide/formaldehyde agarose gel and exposed to Northern blot analysis as described before (Fischman et Materials and methods al., 1990). The full length murine fer c-DNA (a kind gift from T Pawson) was used as a probe in that analysis. Cell cultures and cell lines TEC, a thymic epithelial cell line originating from C57B1/6 Plasmid construction mice, was kindly provided by Dr Kruisbeek, Cancer fer Institute, Amsterdam, The Netherlands (Glimcher et al., The mouse p94 cDNA clone (a kind gift from T Pawson) 1983). STAC, another epithelial cell line, was a kind gift of extendingfrom52bpupstreamofthefermRNAAUG Dr D Zippori, the Weizmann Institute of Science, Rehovot, translation initiation codon, to 30 bp past its UAG Israel. Cell line No. 127, a double positive murine termination codon, was inserted into an SpeI-EcoRI cut thymoma cell line (Irlin and Peled, 1992), was kindly and blunt ended pECE vector (Ellis et al., 1986). In the provided by Dr N Haran Gera, the Weizmann Institute of newly constructed plasmid termed pECEfer, the fer cDNA Sciences, Rehovot, Israel. C2 ± a murine myogenic cell line is transcribed under the control of the SV40 early promoter (Yae and Saxel, 1977) was kindly provided by Dr A and is ¯anked at its 3' end, by the SV40 early genes Shainberg, Bar-Ilan University, Ramat-Gan, Israel. 2B4.11 polyadenylation sequences (Ellis et al., 1986). is a murine CD4+ T-cell hybridoma, speci®c for the antigen

pigeon cytochrome c (Yang et al., 1993). CTLL ± a fer + Transfection of the 127 pre-T cell line and selection of p94 mitogen resistant, IL-2 dependent, mouse CD8 cytotoxic expressing clones T cell line, was kindly provided by Dr B Sredni, Bar-Ilan University, Ramat-Gan, Israel. Ltk7 and NIH3T3 are two 56106 127 cells, suspended in RPMI containing electro- mouse ®broblastic cell lines. The TEC, STAC, C2, Ltk7 poration buer (109 mM sucrose and 2.8 mM phosphate and NIH3T3 cell lines were cultured in DMEM supple- buer pH 7.4) were mixed with 50 mg pECEfer and 5 mgof mented with 100 U/ml penicillin, 100 mg/ml streptomycin the pSVneo plasmid (Southern and Berg, 1982). Electro- and 10% FCS. Freshly isolated thymocytes and spleno- poration was carried out with the BIO ± RAD gene pulser, Cellular recovery of irradiated pre-T cells S Halachmy et al 2879 by applying 500 mFa and 450 V. The electroporated cells viability was determined by using the Trypan blue were then divided to subgroups, containing 105 cells each. exclusion test. Each group of cells in 1 ml RPMI containing 400 mgG418 (Southern and Berg, 1982), was transferred to 24 well Flow cytometry analysis plates. Surviving clones were further propagated and grown in 50 ml ¯asks. Whole cell protein extracts were 105 cells were spun down, washed with PBS and prepared from the G418 resistant clones and the expression resuspended in 0.5 ml staining buer containing: 0.1% of p94fer was determined by Western blot analysis. Groups BSA, 0.01% Sodium Azide, 50 mg/ml Propidium Iodide of clones expressing p94fer, were subcloned by serial (PI), 0.1% Triton X100 and 0.3 mg/ml RNase. The cell limiting dilution in 96 well plates, for obtaining homo- suspension was left for 10 min at room temperature. Nuclei genous clones. were spun down, resuspended in 0.5 ml PBS and analysed for relative DNA content by a Coulter FACSort (Becton Dickinson) ¯ow cytometer. Animal and cell irradiation Mice were exposed to 450 rad g-radiation at room temperature, utilizing a 137Cs source (Cammacell 1000 ± Atomic Energy of Canada Limited) at a rate of 450 rad/ Acknowledgements min. Transfected 127 cells were seeded at 104 cells/ml in a We thank U Caro for running the ¯ow cytometry analysis freshly prepared medium containing 400 mg/ml G418. The and A Goldreich for typing the manuscript. This work was cells were irradiated at 950 rad and samples were taken supported by a grant from the Israel Academy Science daily for ¯ow cytometry and viability analysis. Cell Foundation.

References

Ben-Neriah Y, Bernards A, Poskind M, Daley GQ and MacDonald I, Levy J and Pawson T. (1985). Mol. Cell. Biol., Baltimore D. (1986). Cell, 44, 577 ± 586. 5, 2543 ± 2551. Bernards A, Paskind N and Baltimore D. (1988). Oncogene, Manfredini R, Grande A, Taglia®co E, Barbieri D, Zucchini 2, 297 ± 304. P,CitroG,ZupiG,FranceschiC,TorelliUandFerrariS. Boulter CA and Wagner EF. (1988). Exp. Cell Res., 179, (1993). J. Exp. Med., 178, 381 ± 389. 214 ± 221. Marks PA, Chen Z-X, Banks J and Rifkind RA. (1983). Proc. Care A, Mattio G, Montesoro, Parolini I, Russo G, Colombo Natl. Acad. Sci. USA, 80, 2281 ± 2284. MP and Peschle C. (1996). Oncogene, 9, 739 ± 747. McWhirter JR and Wang JYJ. (1993). EMBO J., 12, 1523 ± Coer PJ, Burgering BM, Peppelenbosch MP, Bos JL and 1546. Kruijer W. (1995). Oncogene, 11, 561 ± 569. Muller R, Salmon DJ, Trembley JM, Cline MR and Verma Ellis L, Clauser E, Morgan DO, Edery M, Roth RA and IM. (1982). Nature, 299, 640 ± 644. Rutter WJ. (1986). Cell, 45, 721 ± 732. Pawson T, Letwin K, Lee T, Hao Q-L, Heisterkamp N and Eshel I, Savion N and Shoham J. (1990). J. Immunol., 144, Groen J. (1989). Mol. Cell. Biol., 9, 5722 ± 5725. 1554 ± 1562. Ren R, Ye ZS and Baltimore D. (1994). Genes Dev., 8, 783 ± Feldman RA, Gabrilove JL, Tam JP, Moore MAS and 795. Hanafusa H. (1985). Proc. Natl. Acad. Sci. USA, 82, Renshaw MW, Capozza MA and Wang JYJ. (1988). Mol. 2379 ± 2383. Cell. Biol., 8, 4547 ± 4551. Feldman RA, Tam JP and Hanafusa H. (1986). Mol. Cell. Roebroek AJM, Schalken JA, Verbeek JS, Van der Ouwe- Biol., 6, 1065 ± 1073. land AMW, Onnekink C, Bloemers HPJ and Van de Ven Fischman K, Edman JC, Shakleford GM, Turner JA, Rutter WJM. (1985). EMBO J., 4, 2897 ± 2903. WJ and Nir U. (1990). Mol. Cell. Biol., 10, 146 ± 153. Sambrook J, Fritsch EF and Maniatis T. (eds). (1989). Glimcher LH, Kruisbeek AM, Paul WE and Green I. (1983). Molecular clone. A Laboratory Manual. Cold Spring Scand. J. Immunol., 17, 1±11. Harbor Laboratory Press. Greer P, Maltby V, Rossant J, Bernstein A and Pawson T. Schieven GL, Kirihara JM, Gilliland LK, Uckun FM and (1990). Mol. Cell. Biol., 10, 2521 ± 2527. Ledbetter JA. (1993). Mol. Cell. Biol., 4, 523 ± 530. Greer P, Haigh J, Mbamalu G, Khoo W, Bernstein A and Schreiber L, Sharabi Y, Schwartz D, Gold®nger N, Brodie C, Pawson T. (1994). Mol. Cell. Biol., 14, 6755 ± 6763. Rotter V and Shoham J. (1996). Scand. J. Immunol., (in HanazonoY,ChibaS,SasakiK,ManoH,MiyaJimaA,Arai press). K, Yazaki Y and Hiral H. (1993). EMBO J., 12, 1641 ± 1646. Schwarzberg PL, Stall AM, Hardin JD, Bowdish KS, Hao Q-L, Heisterkamp N and Groen J. (1989). Mol. Cell. Humaran T, Boast S, Harbison ML, Robertson EJ and Biol., 9, 1587 ± 1593. Go SP. (1991). Cell, 65, 1165 ± 1175. Hao Q-L, Ferris DK, White G, Heisterkamp N and Groen Sen R and Baltimore D. (1986). Cell, 47, 921 ± 928. J. (1991). Mol. Cell. Biol., 11, 1180 ± 1183. Sokaguchi N, Kishimoto T, Kikutani H, Watanabe T, Hazan B, Bern O, Carmel M, Lejbkowicz F, Goldstein RS Yoshida N, Shimizu A, Yamawaki-Kataoka Y, Honjo T and Nir U. (1993). Cell Growth Dier., 4, 443 ± 449. and Yamamuna Y. (1980). J. Immunol., 125, 2654 ± 2659. Irlin Y and Peled A. (1992). Immunol. Lett., 33, 233 ± 238. Southern PJ and Berg P. (1982). J. Mol. Appl. Genet., 1, 327 ± Izuhara K, Feldman RA, Greer P and Harada N. (1994). J. 341. Biol. Chem., 269, 18623 ± 18629. Strasser A, Harris AW, Jacks T and Cory S. (1994). Cell, 79, Keshet E, Itin A, Fischman K and Nir U. (1990). Mol. Cell. 329 ± 339. Biol., 10, 5021 ± 5025. Tybulewicz VLJ, Crawford CE, Jackson PK, Bronson RT Kharbanda S, Yuan ZM, Rubin E, Weichselbaum R and and Mulligan RC. (1991). Cell, 65, 1153 ± 1163. Kufe D. (1994). J. Biol. Chem., 269, 20739 ± 20743. Uckun FM, Schieven GL, Tuel-Ahlgren LM, Dibirdik I, Kim L and Wang TW. (1995). Mol. Cell. Biol., 15, 4553 ± Myers DE, Ledbetter A and Song CW. (1993). Proc. Natl. 4561. Acad. Sci. USA, 90, 252 ± 256. Kipreos ET and Wang JYJ. (1992). Science, 256, 382 ± 385. Van Etten RA, Jackson P and Baltimore D. (1989). Cell, 58, Letwin K, Yee S-P and Pawson T. (1988). Oncogene, 3, 621 ± 669 ± 678. 627. Cellular recovery of irradiated pre-T cells SHalachmyet al 2880 Van Etten RA, Jackson PK, Baltimore D, Sanders MC, Yae A, Schwarz Y, Hacohen D, Nir U and Salzberg S. Masudario PT and Jammey PA. (1994). J. Cell Biol., 124, (1996). Cell Growth Dier. (in press). 325 ± 340. Yae D and Saxel O. (1977). Nature, 270, 725 ± 727. Wadsworth SA, Chang AC, Hang UP, Halvorson MJ, Otto S Yang Y, Vacchio MS and Ashwell JD. (1993). Proc. Natl. and Coligan JE. (1995). J. Immunol., 154, 2125 ± 2133. Acad. Sci. USA, 90, 6170 ± 6174. Wang JYJ and Baltimore D. (1983). Mol. Cell. Biol., 3, 773 ± Yates KE, Lynch MR, Wang SG, Salmon DJ and Gasson 779. JC. (1995). Oncogene, 10, 1239 ± 1242. Wilks AF and Kurban RR. (1988). Oncogene, 3, 289 ± 294.