- Email: [email protected]
Expression of thevavOncogene in Somatic Cell Hybrids
EXPERIMENTAL CELL RESEARCH ARTICLE NO.
232, 388–394 (1997)
EX973535
Expression of the vav Oncogene in Somatic Cell Hybrids Diane J. Denkinger and Rodney S. Kawahara1 Department of Pharmacology, University of Nebraska Medical Center, 600 South 42nd Street, Omaha, Nebraska 68198-6260
The vav oncogene is expressed primarily in tissues of hematopoietic origin. While much effort has been focused on determining the role of vav in various signal transduction pathways, little is known about the mechanism by which vav is regulated in a tissue-selective manner. This issue was examined by developing somatic cell hybrids between human U937 cells, which express vav, and mouse Balb/c 3T3 cells, which do not. If vav is primarily regulated by the presence of positive acting transcription factors, then vav expression should be maintained in hybrid cells. In contrast, if the regulation of vav is primarily negative in nature, then vav expression should be extinguished in most of the somatic cell hybrids. Of the hybrid cells that were obtained, 64% were positive by reverse transcriptase– polymerase chain reaction for the expression of the vav oncogene. Differences in the pattern of restriction enzyme cleavage sites between the mouse and human PCR products were used to determine that 6 of 11 of the positive clones expressed the normally dormant mouse gene. The other positive clones were found to express the human vav gene. In all cases, the hybrid cells preferentially retained the chromosomes and the cellular morphological appearance of the mouse Balb/c 3T3 fusion partner, which does not express the vav oncogene. Since vav is able to be transiently expressed by hybrid cells with a predominately mouse phenotype, these results support the hypothesis that vav is regulated primarily by the presence of transactivating factors which stimulate transcription, rather than by a gene silencing mechanism. q 1997 Academic Press
INTRODUCTION
vav is a transforming oncogene that was discovered when genomic DNA from an esophageal carcinoma was transfected with pSV2neo into NIH3T3 cells [1]. The fortuitous recombination of the vav gene with pSV2neo led to its subsequent activation and identification. The isolation and sequence analysis of a full-length vav proto-oncogene [2–4] clone demonstrated the presence 1 To whom correspondence and reprint requests should be addressed. Fax: (402) 559-7495.
388
0014-4827/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
ECR 3535
/
6i1f$$$201
of protein domains with sequence similarity to helixloop-helix transcription factors, Dbl and pleckstrin homology domains [4–6], a zinc-finger-like cysteine motif similar to the phorbol ester binding site in protein kinase C [3, 6, 7], and two basic amino acid stretches that may function as nuclear localization signals. At the C-terminus, there is a single SH2 domain flanked by two SH3 domains [1, 2]. These findings suggest that vav might be a key component of the cell surface to nuclear signal transduction pathways and may function as a guanine nucleotide exchange factor for ras or the rho/rac family of ras-like molecules [8–11]. There is evidence that vav is able to translocate into the nucleus [12]. vav associates with ENX-1, a transcriptional regulator of homeobox gene expression [13], and Ku-70 [14], the DNA binding subunit of DNA-dependent protein kinase. ENX-1 interacts with the Nterminal region of vav and is homologous to the enhancer of zeste gene, which regulates homeobox expression in Drosophila [13]. Overexpression of Bmi-1, a member of the homeobox gene regulator family, results in B- and T-cell lymphomas [15], whereas the deletion of this gene causes defects in hematopoietic system development [16]. A similar role for ENX-1 may be possible, with vav acting as a messenger that relays cell surface signals by translocating into the nucleus. vav may also function as an integrator of signals to the cytoskeleton. Tyrosine phosphorylation of vav increases following activation of the T-cell receptor–CD4 complex [17, 18], the Fc receptor for IgE [17], B-cell IgM receptor [19], c-kit (the receptor for steel factor, stem cell factor, mast cell growth factor or c-kit ligand) [20], Flt3/Flk-2 receptor [21], and receptors for interleukin 2, interleukin 3, GM-CSF, insulin, prolactin, and interferon-a [22–26]. The expression of antisense RNA in embryonic stem (ES) cells disrupts the in vitro development of hematopoietic lineages and suggests that vav may have an important role in hematopoietic cell differentiation [27]. However, in vav-negative ES cells developed by homologous recombination, normal differentiation of myeloid, erythroid, and mast cell linages were observed [28, 29]. This evidence suggests that vav is not necessary for hematopoietic development. However, other members of the vav family (vav2 [30] or family mem-
04-02-97 01:34:33
eca
vav EXPRESSION IN SOMATIC CELL HYBRIDS
bers yet to be discovered) could have provided redundant pathways that compensated for the absence of vav in the knockout experiments. In the antisense experiments, there could have been cross-hybridization between the antisense probe and the mRNAs for all vav family members that resulted in the loss of all redundant pathways and the loss of hematopoietic differentiation. The absence of vav causes a decrease in T- and B-cell lineages and results in defective proliferative signaling by both the T- and B-cell antigen receptors. vav is likely to be necessary for the expansion of T- and Bcell lineages and in the early phases of antigen receptor signaling [31–33]. The expression of vav appears to be limited to cells of hematopoietic origin, trophoblasts, pancreas, and the ameloblasts of the developing tooth bud [34, 35, 49]. vav is not expressed in epithelial, muscle, brain, intestine, or kidney cells [1, 2, 4]. vav maps to human chromosome region 19p12–19p13.2 [36]. This region has been shown to be an important site of karyotypic abnormalities in leukemias, lymphomas, melanomas, and small cell lung carcinomas. However, there is no direct evidence that vav is important in these disease processes. The mechanism which regulates the tissue-specific expression of the vav oncogene is hypothesized to occur by one of two generalized processes. Either the gene is selectively silenced in cells where it is not expressed or cell-specific transcription factors stimulate the expression of vav in cell types where it is expressed. To distinguish between these mechanisms, cells which express the vav gene (human, U937 cells) were fused to cells which do not express vav (mouse, Balb/c 3T3 cells). If silencing is the predominant mechanism, the expression of the human vav gene is expected to be extinguished, whereas if vav is controlled by positive acting cell-specific transcription factors, it is anticipated that the expression of vav in the hybrid cells would be maintained and possibly activate the normally dormant mouse gene. MATERIALS AND METHODS Cell lines, electroporation, and cell fusion. The human leukemia cell line, U937 (CRL 1539), and mouse fibroblast, Balb/c 3T3 clone A31 (CCL 163), were obtained from American Type Culture Collection (Rockville, MD). The S49cyc0 cells were from the UCSF cell culture facility. HeLa cells were a kind gift from Dr. G. Stanley Cox (University of Nebraska Medical Center). HeLa cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 44 mM sodium bicarbonate, 15 mM Hepes, 50 units/ml penicillin– streptomycin, and 10% calf serum. Balb/c 3T3 and U937 cells were grown in RPMI medium 1640 supplemented with 10% fetal bovine serum (FBS). S49cyc0 cells were maintained in DMEM supplemented with 44 mM sodium bicarbonate, 15 mM Hepes, 50 units/ml penicillin–streptomycin, 10% FBS, and nonessential amino acids. The U937 cells were electroporated in a 2-mm gap BTX disposable cuvette with 40 mg pRSVneo, the generous gift of Dr. Timothy Ley,
AID
ECR 3535
/
6i1f$$$202
04-02-97 01:34:33
389
Washington University, St. Louis. Settings for the electroporation were 2950 mF (capacitance), 186 ohm (resistance), and 100 V (voltage) using a BTX 600 (BTX, San Diego, CA). The actual values for field strength and pulse length were 0.39 kV/cm and 38 ms, respectively. Following electroporation, the cells were grown in media with 400 mg/ml Geneticin (G418) (Life Technologies, Inc., Grand Island, NY). Balb/c 3T3 cells were electroporated in a 2-mm BTX disposable cuvette with pRSV hygro [32] at settings of 900 mF (capacitance), 13 ohm (resistance), and 200 V (voltage). The actual values for field strength and pulse length were 0.83 kV/cm and 7.08 ms, respectively. Hygromycin B-resistant cells were selected in media containing 200 mg/ml Hygromycin B (Calbiochem, La Jolla, CA). G418-resistant U937 and hygromycin-resistant Balb/c 3T3 cells were harvested, washed once with PBS, resuspended, and counted. Approximately 3.0 1 106 cells from each cell line were combined and centrifuged and the supernatant was removed. While continuously agitating the tube, 1 ml PEG 1500 (Boehringer Mannheim, Indianapolis, IN), prewarmed to 427C, was added dropwise to the cell pellet. After further gentle agitation, 5 ml of serum-free RPMI 1640 was added dropwise over a 5-min period, followed by 15 ml over a 1-min period. The cells were centrifuged for 10 min and the supernatant was removed and discarded. The cell pellet was resuspended in RPMI 1640 with 10% FBS and transferred to a flask. Antibiotics (400 mg/ ml G418 and 200 mg/ml Hygromycin B) were added to the flask the next day. During the first week, multinucleated adherent cells were observed along with cells in suspension. Nonadherent cells were removed and maintained separately from adherent cells, but did not survive the selection process. One month later, the hybrid cells were cloned by limiting dilution method (approximately 1 cell per every 4 wells in 96-well plates) in selective media. The cloned hybrid cells were transferred to 24-well plates and frozen at 0767C in RPMI 1640 with 10% FBS and 7% DMSO for later analysis. Preparation of RNA. RNA extractions were prepared according to the method described by Chomzynski and Sacchi [38] with modifications. Hybrid clones were revived and allowed to grow to confluence in 100 1 20 mm tissue culture plates. Adherent hybrid clones and HeLa and Balb/c 3T3 cells were harvested by pouring off the media and adding 0.5 ml of solution D (4 M guanidinium thiocyanate, 25 mM sodium citrate, pH 7, 0.5% Sarkosyl, 0.1 M 2-mercaptoethanol). The cells were removed from the surface of the plate using a cell scraper. U937 and S49cyc0 were harvested by centrifugation and disrupted in 0.5 ml of solution D. Adherent cells were phenol extracted three times, while U937 and S49cyc0 cells were extracted four times. Occasionally, additional solution D was added after the first two to three extractions in order to maintain a volume of 400– 500 ml. This improved the recovery of the extracted total RNA. The RNA was precipitated with isopropanol, dried in a speed vac for 1– 3 min, and stored at 0767C. Reverse transcriptase PCR analysis. RNA samples were digested with RNase-free DNase I (RQ1 DNase I, Promega Corp., WI) by resuspending the RNA pellets in a solution composed of 6 mM MgCl2 , 10 mM KCl, 40 mM Tris, pH 8, and 5 U RQ1 DNase I. Reactions were incubated for 15–20 min at 377C. The RNA was extracted using the method of Chomzynski and Sacchi [38], except that only one phenol:chloroform extraction was conducted. The precipitated pellets were dissolved in 10–20 ml DEPC water and stored at 0767C. RT– PCR analysis was performed using the standard and EZ protocols (Perkin–Elmer, Roche Molecular Systems, Inc., Branchburg, NJ). Reactions contained 10–20 mg of RNA and 250–500 ng of each primer. Primer pairs consisted of primers TGCACCCACTGGCTGATCCAGTGTCGGG (VA-7) with TAAGACAAAGGAACTGGGACATCTGGGG (VA-8) or VA-7 with either GAAGGTTCGAATGTTCTTAAGACAAAGG (VA-8M), from the mouse vav, or GAAGGTTCTAATGTTCTTAAGGCACAGG (VA-8H), from the human vav gene. The RNA was denatured for 2 min at 957C, reverse transcribed for 15 min at 707C, denatured at 957C for 2 min, and amplified for 35–45 cycles (957C for 1 min, 607C for 1 min) followed by a final extension
eca
390
DENKINGER AND KAWAHARA
at 607C for 7 min. Alternatively, for the EZ protocol, RNA was reverse transcribed at 607C for 30 min, followed by amplification for 40 cycles (947C for 45 s, 607C for 45 s), and a final extension for 7 min at 607C. PCR products were electrophoresed on a 4% NuSieve 3:1 agarose gel (FMC Bioproducts, Rockland, ME) in 11 TAE and 0.14 mg/ml of ethidium bromide. Gels were illuminated with a UV lamp and photographed. Southern blot analysis. Total RNA samples (20 mg) from hybrid cells were undigested or digested with 20 mg RNase A at 377C for 30 min and amplified by RT–PCR using vav primers. The reaction product was electrophoresed on a 4% NuSieve agarose gel, excised, and reamplified to obtain sufficient material for restriction digest analysis. The PCR products were pooled, phenol:chloroform extracted, ethanol precipitated, resuspended, and divided between three tubes. One tube was digested with FokI, another with VspI, and the third was undigested. The samples were electrophoresed on a 4% NuSieve gel with an aliquot from each of the original RT–PCR products that were untreated or treated with RNase A. The gel was denatured in 0.5 M NaOH/1.5 M NaCl, neutralized with 1 M Tris/1.5 M NaCl, and transferred to a nylon membrane overnight in 101 SSC. The DNA was UV crosslinked at 0.12 J/cm2 in a UV Stratalinker 1800 (Stratagene, La Jolla, CA). Blots were prehybridized at 567C in BLOTTO (0.25% nonfat dry milk in 11 SSPE) for a minimum of 6 h. The probe used for hybridization was a 1039-bp PstI vav fragment obtained from the plasmid pSK114 (the kind gift of Dr. Shulamit Katzav). The fragment was labeled by nick translation and separated from the unincorporated [32P]dCTP on a Nick Column (Pharmacia Biotech AB, Uppsala Sweden). The probe was boiled, cooled, and added to the prehybridization solution for overnight hybridization at 567C. Blots were washed twice at room temperature for 15 min in 21 SSC/ 0.1% SDS. A final wash in 0.21 SSC/0.1% SDS at 567C for 15 min was completed prior to autoradiography.
RESULTS
Cell-specific expression of the vav oncogene. A RT – PCR assay to detect the expression of both the mouse and human vav oncogene was developed. Total RNA from cell lines S49cyc0 and U937, which express the vav oncogene, and nonhematopoietic cell lines (Balb/c 3T3 and HeLa) were amplified with primer pairs VA-7/VA-8M or with VA-7/VA-8H and electrophoresed on a 4% NuSieve agarose gel. As shown in Fig. 1A, a PCR band of 216 bp was detected in the U937 and S49cyc0 cell lines, but not in the Balb/c 3T3 and HeLa cell lines. As a control, 5 mg of total RNA from each cell line is shown in Fig. 1B to confirm the integrity and quantity of RNA used in the RT – PCR reaction. The size of the PCR product is consistent with the predicted size for vav. To determine whether the PCR products were amplified from either mouse or human vav mRNA, DNase Itreated total RNA samples from U937 or S49cyc0 cells were amplified by RT–PCR, digested with either FokI or Tth111I, and electrophoresed on a 4% NuSieve agarose gel. As shown in Fig. 2, the RT–PCR product amplified from total RNA from human U937 cells was sensitive to the enzyme FokI, but not to Tth111I, since the human vav gene contains two FokI sites and cannot be cleaved by Tth111I. The 199-bp PCR product was digested into three fragments (90, 70, 39 bp), of which
AID
ECR 3535
/
6i1f$$$202
04-02-97 01:34:33
FIG. 1. Cell type-specific expression of the vav oncogene. (A) Total RNA was prepared from U937, human histiocytic lymphoma; HeLa, human cervical epitheloid carcinoma; S49cyc0, mouse lymphoma; and Balb/c 3T3, mouse embryonic fibroblast. vav was detected by RT – PCR using primer pairs VA-7/VA-8M (S49cyc0, Balb/c 3T3) or VA7/VA-8H (U937, HeLa). The PCR products were electrophoresed on a 4% NuSieve agarose gel and photographed. HaeIII-digested fX174 RF DNA was used as a size marker in lane (M). (B) Equivalent quantities of total RNA from each sample were electrophoresed on a 1.5% agarose gel to determine RNA quality and quantity.
only the 90-bp band was visible. The other two bands were obscured by nonspecific bands of a similar size. In contrast, the mouse vav gene was sensitive to Tth111I, but not FokI, since the mouse gene contains a Tth111I site and is not sensitive to cleavage by FokI. In many experiments, there was a minor fraction of the amplified product from S49cyc0 cells that was resistant to the action of Tth111I. A similar resistant fraction was also observed using VspI, another restriction enzyme which was used to discriminate between mouse and human vav transcripts (Fig. 5A). The PCR product from S49cyc0 cells which was resistant to VspI was treated with T4 DNA polymerase to create blunt ends, cloned into the HincII site of pGEM4z, and sequenced. Direct sequence analysis demonstrated that the PCR product was indeed amplified from the vav oncogene, but contained different point mutations at the VspI site. It is not known if this represents transcripts from other vav family members or represents errors made by the PCR amplification process, but we feel that the latter is more likely than the former. As a further control, DNase I-treated total RNA was either left undi-
eca
vav EXPRESSION IN SOMATIC CELL HYBRIDS
391
FIG. 2. Differential sensitivity of human and mouse vav RT–PCR products to FokI and Tth111l. RT–PCR was conducted on total RNA from U937 (lanes 1–3, 7, 8) and S49cyc0 (lanes 4–6, 9, 10) cells. The RT–PCR products were undigested (lanes 1 and 4) or digested with either FokI (lanes 2 and 5) or Tth111I (lanes 3 and 6). Total RNA was untreated (lanes 7, 9) or treated with 10 mg DNase-free RNase A for 30 min at 377C (lanes 8, 10) prior to starting the RT–PCR reaction. The marker lane (M) contains fragments of HaeIII-digested fX174 RF DNA.
gested or digested with RNase A prior to RT–PCR amplification. As shown in Fig. 2, samples treated with RNase A did not produce the 199-bp band detected in untreated samples. These results further validate the RT–PCR method used and its ability to accurately discriminate between human and mouse vav oncogene transcripts. The development of U937 and Balb/c 3T3 somatic cell hybrids. Figure 3 shows a schematic diagram of the method employed to obtain the human 1 mouse somatic cell hybrids. Human U937 cells which express the vav oncogene were transfected with pRSVneo and G418-resistant cells were obtained. Mouse Balb/c 3T3 cells which do not express the vav oncogene were transfected with pRSVhygro and selected with hygromycin to obtain resistant cells. Both resistant cell lines were fused with PEG 1500 and selected in the presence of both G418 and hygromycin. Control cells which contained either resistant gene alone did not survive the selection procedure with both antibiotics. After 1 week of selection, hybrid adherent cells were separated from the cells in suspension. None of the cells in suspension resulted in viable somatic cell hybrids that were resistant to both antibiotics. However, adherent cells, some of which were visibly multinucleated, resulted in viable somatic cell hybrids. Hybrid cells were cloned in 96well plates by limiting dilution method and frozen at 0767C. Small batches of hybrid cells were revived and analyzed for vav mRNA expression by RT–PCR. Of the 89 cloned hybrid cells, 39 were analyzed and 64% of them were positive for the expression of either the mouse or the human vav gene. The remainder of the clones were nonviable and did not survive the storage process. Cell fusion activates the normally inactive mouse vav gene. Total RNA from 11 positive clones was ampli-
AID
ECR 3535
/
6i1f$$$203
04-02-97 01:34:33
fied by RT–PCR and digested with either FokI or Tth111I. In Fig. 4, the RT–PCR product from a representative hybrid clone (F8) was not obtained if the RNA
FIG. 3. Fusion of the U937 human histiocytic lymphoma with Balb/c 3T3 mouse fibroblasts. Human U937 cells resistant to neomycin were fused with hygromycin-resistant Balb/c 3T3 cells and selected with both geneticin and hygromycin. The fused cells that were resistant to both antibiotics were cloned and subjected to RT–PCR.
eca
392
DENKINGER AND KAWAHARA
in cells which normally do not express the gene or by tissue-specific transcription factors which activate vav expression in cells that do express the gene. To determine which of the two mechanisms was responsible, human leukemia cell line U937 was fused to mouse Balb/c 3T3 cells, which do not express the vav oncogene. The majority of the hybrid cells that were obtained expressed vav and in six distinct cases, the expression of the normally dormant mouse gene was clearly detectable. All hybrid cells examined retained predominately mouse chromosomes and were morphologically fibro-
FIG. 4. Hybrid cell F8 expresses the mouse vav oncogene. Resistant U937 and Balb/c 3T3 cells were fused and cloned as described in Fig. 3. Total RNA was extracted and the RT–PCR products were undigested (lane 1 and 3) or digested with FokI (lane 4) or Tth111I (lane 5). Prior to amplification, the total RNA was treated with RNase A (lane 2). The marker lane (M) contains fragments of HaeIII-digested fX174 RF DNA.
was pretreated with RNase A. The RT–PCR product could be cleaved by Tth111I, but not by FokI. This pattern of restriction enzyme sensitivity was consistent with the RT–PCR product originating from the mouse vav mRNA transcript. RT–PCR reaction products derived from hybrid cell total RNA were also examined for differential sensitivity to digestion by FokI and VspI (Fig. 5). The restriction enzyme digestion pattern of the amplified products from five hybrid clones matched the digestion pattern obtained from U937 cells. This indicated that these five hybrid clones expressed the human vav gene (Fig. 5B). Other clones, such as F70, clearly demonstrated an identical pattern of digestion that matched the pattern found in S49cyc0 cells and expressed the normally dormant mouse vav gene. The identity of the amplification products was further confirmed by southern blotting with a probe to the vav oncogene (Fig. 5D). Of 11 clones examined, 6 were positive for the expression of the mouse vav gene, while the others expressed the human vav gene. None of the vav-positive clones led to the establishment of cell lines which stably expressed the vav gene. Invariably, the expression of vav was found to decrease with the length of time in culture and appeared to correlate with the random loss of identifiable human chromosomes (observed through karyotype analysis of the hybrid cells). The remainder of the original 25 positive clones could not be definitively analyzed due to the transient instability of the hybrid cells. DISCUSSION
The tissue-specific expression of the vav oncogene was hypothesized to be regulated either by silencing
AID
ECR 3535
/
6i1f$$$203
04-02-97 01:34:33
FIG. 5. Differential sensitivity of human and mouse vav RT– PCR products to FokI and VspI in hybrid cells. (A) RT–PCR was conducted on total RNA from U937 and S49cyc0 cells and the product was undigested or digested with either FokI or VspI. (B) RT–PCR product from hybrid cell clone F24 was undigested or digested with FokI and VspI. This pattern of sensitivity indicates the expression of the human vav gene. In C, total RNA from clone F70 was untreated or treated with 10 mg DNase-free RNase A for 30 min at 377C prior to starting the RT–PCR reaction. The RT–PCR products were undigested or digested with FokI and VspI. The pattern of sensitivity indicates the expression of the normally dormant mouse vav gene in hybrid cell clone F70, in contrast with the clone shown in B. The gel shown in C was Southern blotted with a vav probe in D. The marker lane (M) contains fragments of HaeIII-digested fX174 RF DNA.
eca
vav EXPRESSION IN SOMATIC CELL HYBRIDS
blast-like. These factors would favor silencing if it were the mechanism. Since the majority of the hybrid clones were found to express vav, this finding does not support silencing as the major mechanism which regulates vav expression. The activation of the normally dormant mouse vav gene further supports this point of view. If silencing were the mechanism, we would have expected vav expression to be extinguished in the hybrid clones; however, this was not the case. In the minority of clones where vav could not be detected, karyotypic analysis showed either the absence of recognizable human chromosomes or a small number of them. If vav expression was regulated by silencing, it would imply that negative regulatory factors were lost or inactivated in hybrid cells and thus allowed vav to be expressed. vav expression could only be observed transiently and no stable hybrid cell lines which express vav could be established. It is more likely that hybrid cells will lose, rather than gain, genetic material over time to approach a stable karyotype. Therefore, the transient nature of vav expression argues against silencing as the predominant mechanism, since it is unlikely that a negative factor could be lost and then regained to modulate vav expression. The expression of vav is more likely to be regulated by the presence of positive factors which were contributed by the U937 cells. It is not possible to definitively exclude the possibility that the cells which express the vav gene did not lose important genetic material which could have mediated a silencing event or any of the more complicated mechanisms that can be envisioned. Examination of the vav promoter will be useful in designing further experiments to test these hypotheses. Since the cells were cloned and had undergone multiple rounds of cell division, it is unlikely that original human proteins which were transferred at the time of fusion were responsible for the transcriptional activation of the mouse vav gene. It is more likely that the genes which encode the transactivating factors responsible for the tissuespecific expression of the vav gene remained active following cell fusion. If this is true, several implications follow. First, the mechanism is sufficiently conserved to activate the dormant mouse gene and we should expect the DNA recognition sites of the transactivators to be conserved between the human and the mouse species. Since vav is first detected in the fetal liver at a relatively early stage of development [34], a period where there are more similarities than differences between the mouse and the human embryo, this conservation is not unexpected. This would imply that the transgenic mouse may be a valid model for the study of the regulation of the human vav gene. Second, in the hybrid cellular environment, which should not favor vav expression because it more closely resembles the mouse phenotype, both the mouse and human vav
AID
ECR 3535
/
6i1f$$$203
04-02-97 01:34:33
393
genes are able to be maintained in an active state. This would imply that the genes of the transcription factors which regulate vav are active in the hybrid cellular environment. Since multiple rounds of cell division have likely occurred, this would mean that the genes were able to maintain their expression either by autoregulation or as a network of complementing transactivators. In either case, this situation was not ultimately stable, since all of the clones were observed to lose their ability to express vav with time. The exact cause of the loss of vav expression has not been determined, but it is likely that multiple factors or gene products are involved in expression of the final phenotype and subsequent loss of these factors leads to the loss of vav expression. Third, the mouse gene is not irreversibly inactivated. This has been observed in other systems as well [39]. Analysis of gene expression in somatic cell hybrids and multinucleate heterokaryons have shown that the differentiated state is reversible and can be reprogrammed by the action of transacting factors. Fusion of K562 cells with mouse nonerythroid cells activates the mouse embryonic e-globin gene [40]. Transient heterokaryons derived from the fusion of human nonerythroid cells with mouse erythroleukemia cells results in the activation of human b globins [41]. Mouse zinc finger transcription factor GATA-1, which is important in the regulation of the b-globin locus, is activated in heterokaryons derived from human K562 cells and mouse ES cells, which do not express GATA-1 or globin genes [42]. The existence of complex reprogramming schemes is illustrated by the fusion of mouse myoblasts with human nonmuscle cells. These fusions resulted in the activation of the human muscle regulators followed by the activation of the mouse muscle genes [43]. The regulators that are known to be activated in muscle heterokaryons include members of the helixloop-helix family of myogenic transcription factors [44, 45]. Other examples where genes that are normally dormant in one fusion partner and are activated by factors from the other include the galecitin-1 (L-14-I) gene [46], the mouse a-fetoprotein gene [47], and murine macrophage class II genes [48]. These experiments support the hypothesis that the vav oncogene is regulated by the presence of one or more transacting factors. Analysis of the promoter region of the vav gene is under way in our laboratory to further test the hypothesis suggested by this work. The study of the transcriptional regulation of vav will likely provide new information that will contribute to our understanding of its role in the developmental regulation of the hematopoietic system. We gratefully acknowledge the assistance of Renee Fordyce-Boyer and her staff at the Hattie B. Munroe Center for Human Genetics for their kind assistance with the karyotyping of some of the somatic cell hybrids used in this work. This work was supported by Grant GM46665 from the National Institutes of Health.
eca
394
DENKINGER AND KAWAHARA
REFERENCES 1. Katzav, S., Martin-Zanca, D., and Barbacid, M. (1989) EMBO J. 8, 2283–2290. 2. Katzav, S., Cleveland, J. L., Heslop, H. E., and Pulido, D. (1991) Mol. Cell. Biol. 11, 1912–1920. 3. Coppola, J., Bryant, S., Koda, T., Conway, D., and Barbacid, M. (1991) Cell Growth Differ. 2, 95–105. 4. Adams, J. M., Houston, H., Allen, J., Lints, T., and Harvey, R. (1992) Oncogene 7, 611–618. 5. Galland, F., Katzav, S., and Birnbaum, D. (1992) Oncogene 7, 585–587. 6. Boguski, M. S., Bairoch, A., Attwood, T. K., and Michaels, G. S. (1992) Nature 358, 113. 7. Quest, A. F. G., Bloomenthal, J., Bardes, E. S. G., and Bell, R. M. (1992) J. Biol. Chem. 267, 10193–10197. 8. Gulbins, E., Coggeshall, K. M., Langlet, C., Baier, G., BonnefoyBerard, N., Burn, P., Wittinghofer, A., Katzav, S., and Altman, A. (1994) Mol. Cell. Biol. 14, 906–913. 9. Katzav, S., Packham, G., Sutherland, M., Aroca, P., Santos, E., and Cleveland, J. L. (1995) Oncogene 11, 1079–1088. 10. Khosravi-Far, R., Chrzanowska-Wodnicka, M., Solski, P. A., Eva, A., Burridge, K., and Der, C. J. (1994) Mol. Cell. Biol. 14, 6848–6857. 11. Bustelo, X. R., Suen, K.-L., Leftheris, K., Meyers, C. A., and Barbacid, M. (1994) Oncogene 9, 2405–2413. 12. Clevenger, C. V., Ngo, W., Sokol, D. L., Luger, S. M., Gewirtz, A. M. (1995) J. Biol. Chem. 270, 13246–13253. 13. Hobert, O., Jallal, B., and Ullrich, A. (1996) Mol. Cell. Biol. 16, 3066–3073. 14. Romero, F., Dargemont, C., Pozo, F., Reeves, W. H., Camonis, J., Gisselbrecht, S., and Fischer, S. (1996) Mol. Cell. Biol. 16, 37–44. 15. van Lohuizen, M., Verbeek, S., Scheijen, B., Wientjens, E., van der Gulden, H., and Berns, A. (1991) Cell 65, 737–752. 16. van der Lugt, N. M. T., Domen, J., Linders, K., van Roon, M., Robanus-Maandag, E., te Riele, H., van der Walk, M., Deschamps, J., Sofroniew, M., van Lohuizen, M., and Berns, A. (1994) Genes Dev. 8, 757–769. 17. Margolis, B., Hu, P., Katzav, S., Li, W., Oliver, J. M., Ullrich, A., Weiss, A., and Schlessinger, J. (1992) Nature 356, 71–74. 18. Bustelo, X. R., Ledbetter, J. A., and Barbacid, M. (1992) Nature 356, 68–71. 19. Bustelo, X. R., and Barbacid, M. (1992) Science 256, 1196–1199. 20. Alai, M., Mui, A. L. F., Cutler, R. L., Bustelo, X. R., Barbacid, M., and Krystal, G. (1992) J. Biol. Chem. 267, 18021–18025. 21. Rottapel, R., Turck, C. W., Casteran, N., Liu, X., Birnbaum, D., Pawson, T., and Dubreuil, P. (1994) Oncogene 9, 1755–1765. 22. Evans, G. A., Howard, O. M. Z., Erwin, R., and Farrar, W. L. (1993) Biochem. J. 294, 339–342. 23. Matsuguchi, T., Inhorn, R. C., Carlesso, N., Xu, G., Drucker, B., and Griffin, J. D. (1995) EMBO J. 14, 257–265. 24. Uddin, S., Katzav, S., White, M. F., and Platanias, L. C. (1995) J. Biol. Chem. 270, 7712–7716.
25. Clevenger, C. V., Ngo, W., Sokol, D. L., Luger, S. M., and Gewirtz, A. M. (1995) J. Biol. Chem. 270, 13246–13253. 26. Platanias, L. C., and Sweet, M. E. (1994) J. Biol. Chem. 269, 3143–3146. 27. Wulf, G. M., Adra, C. N., and Lin, B. (1993) EMBO J. 13, 5065– 5074. 28. Zmuidzinas, A., Fisher, K. D., Lira, S. A., Forrester, L., Bryant, S., Bernstein, A., and Barbacid, M. (1995) EMBO J. 14, 1–11. 29. Zhang, R., Tsai, F. Y., and Orkin, S. H. (1994) Proc. Natl. Acad. Sci. USA 91, 12755–12759. 30. Henske, E. P., Short, M. P., Jozwiak, S., Bovey, C. M., Ramlakhan, S., Haines, J. L., and Kwiatkowski, D. J. (1995) Ann. Hum. Genet. 59, 25–37. 31. Tarakhovsky, A., Turner, M., Schaal, S., Mee, P. J., Duddy, L. P., Rajewsky, K., and Tybulewicz, V. L. J. (1995) Nature 374, 467–470. 32. Zhang, R., Alt, F., Davidson, L., Orkin, S. H., and Swat, W. (1995) Nature 374, 470–473. 33. Fischer, K. D., Zmuldzinas, A., Gardner, S., Barbacid, M., Bernstein, A., and Guidos, C. (1995) Nature 374, 474–477. 34. Bustelo, X. R., Rubin, S. D., Suen, K. L., Carrasco, D., and Barbacid, M. (1993) Cell Growth Differ. 4, 297–308. 35. Higuchi, T., Kanzaki, H., Fujimoto, M., Hatayama, H., Watanabe, H., Fukumoto, M., Kaneoko, Y., Higashitsuji, H., Kishishita, M., Mori, T., and Fujita, J. (1995) Biol. Reprod. 53, 840– 846. 36. Martinerie, C., Cannizzaro, L. A., Croce, C. M., Huebner, K., Katzav, S., and Barbacid, M. (1990) Hum. Genet. 86, 65–68. 37. Bonapace, I. M., Sanchez, M., Obici, S., Gallo, A., Garofalo, S., Gentile, R., Cocozza, S., and Avvedimento, E. V. (1990) Mol. Cell. Biol. 10, 1033–1040. 38. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156–159. 39. Baron, M. H. (1993) Curr. Opin. Cell. Biol. 5, 1050–1056. 40. Baron, M. H., and Maniatis, T. (1986) Cell 46, 591–602. 41. Baron, M. H., and Maniatis, T. (1991) Mol. Cell. Biol. 11, 1239– 1247. 42. Baron, M. H., and Farrington, S. M. (1994) Mol. Cell. Biol. 14, 3108–3114. 43. Hardeman, E. C., Chiu, C. P., Minty, A., and Blau, H. M. (1986) Cell 47, 123–130. 44. Rudnick, M. A., and Jaenisch, R. (1995) BioEssays 17, 203–209. 45. Thayer, M. J., and Weintraub, H. (1990) Cell 63, 23–32. 46. Chiariotti, L., Benvenuto, G., Zarrilli, R., Rossi, E., Salvatore, P., Colantuoni, V., and Bruni, C. B. (1994) Cell Growth Differ. 5, 769–775. 47. Spear, B. T., and Ellis, A. W. (1995) Somat. Cell. Mol. Genet. 21, 19–31. 48. Sartoris, S., Scupoli, M. T., Scarpellino, L., Paiola, F., Jotterand-Bellomo, M., Tridente, G., and Accolla, R. S. (1990) J. Immunol. 145, 1960–1967. 49. Schuebel, K., Bustelo, X. R., Nielsen, D. A., Song, B.-J., Barbacid, M., Goldman, D., and Lee, I. J. (1996) Oncogene 13, 363– 371.
Received November 4, 1996 Revised version received January 31, 1997
AID
ECR 3535
/
6i1f$$$204
04-02-97 01:34:33
eca
232, 388–394 (1997)
EX973535
Expression of the vav Oncogene in Somatic Cell Hybrids Diane J. Denkinger and Rodney S. Kawahara1 Department of Pharmacology, University of Nebraska Medical Center, 600 South 42nd Street, Omaha, Nebraska 68198-6260
The vav oncogene is expressed primarily in tissues of hematopoietic origin. While much effort has been focused on determining the role of vav in various signal transduction pathways, little is known about the mechanism by which vav is regulated in a tissue-selective manner. This issue was examined by developing somatic cell hybrids between human U937 cells, which express vav, and mouse Balb/c 3T3 cells, which do not. If vav is primarily regulated by the presence of positive acting transcription factors, then vav expression should be maintained in hybrid cells. In contrast, if the regulation of vav is primarily negative in nature, then vav expression should be extinguished in most of the somatic cell hybrids. Of the hybrid cells that were obtained, 64% were positive by reverse transcriptase– polymerase chain reaction for the expression of the vav oncogene. Differences in the pattern of restriction enzyme cleavage sites between the mouse and human PCR products were used to determine that 6 of 11 of the positive clones expressed the normally dormant mouse gene. The other positive clones were found to express the human vav gene. In all cases, the hybrid cells preferentially retained the chromosomes and the cellular morphological appearance of the mouse Balb/c 3T3 fusion partner, which does not express the vav oncogene. Since vav is able to be transiently expressed by hybrid cells with a predominately mouse phenotype, these results support the hypothesis that vav is regulated primarily by the presence of transactivating factors which stimulate transcription, rather than by a gene silencing mechanism. q 1997 Academic Press
INTRODUCTION
vav is a transforming oncogene that was discovered when genomic DNA from an esophageal carcinoma was transfected with pSV2neo into NIH3T3 cells [1]. The fortuitous recombination of the vav gene with pSV2neo led to its subsequent activation and identification. The isolation and sequence analysis of a full-length vav proto-oncogene [2–4] clone demonstrated the presence 1 To whom correspondence and reprint requests should be addressed. Fax: (402) 559-7495.
388
0014-4827/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
ECR 3535
/
6i1f$$$201
of protein domains with sequence similarity to helixloop-helix transcription factors, Dbl and pleckstrin homology domains [4–6], a zinc-finger-like cysteine motif similar to the phorbol ester binding site in protein kinase C [3, 6, 7], and two basic amino acid stretches that may function as nuclear localization signals. At the C-terminus, there is a single SH2 domain flanked by two SH3 domains [1, 2]. These findings suggest that vav might be a key component of the cell surface to nuclear signal transduction pathways and may function as a guanine nucleotide exchange factor for ras or the rho/rac family of ras-like molecules [8–11]. There is evidence that vav is able to translocate into the nucleus [12]. vav associates with ENX-1, a transcriptional regulator of homeobox gene expression [13], and Ku-70 [14], the DNA binding subunit of DNA-dependent protein kinase. ENX-1 interacts with the Nterminal region of vav and is homologous to the enhancer of zeste gene, which regulates homeobox expression in Drosophila [13]. Overexpression of Bmi-1, a member of the homeobox gene regulator family, results in B- and T-cell lymphomas [15], whereas the deletion of this gene causes defects in hematopoietic system development [16]. A similar role for ENX-1 may be possible, with vav acting as a messenger that relays cell surface signals by translocating into the nucleus. vav may also function as an integrator of signals to the cytoskeleton. Tyrosine phosphorylation of vav increases following activation of the T-cell receptor–CD4 complex [17, 18], the Fc receptor for IgE [17], B-cell IgM receptor [19], c-kit (the receptor for steel factor, stem cell factor, mast cell growth factor or c-kit ligand) [20], Flt3/Flk-2 receptor [21], and receptors for interleukin 2, interleukin 3, GM-CSF, insulin, prolactin, and interferon-a [22–26]. The expression of antisense RNA in embryonic stem (ES) cells disrupts the in vitro development of hematopoietic lineages and suggests that vav may have an important role in hematopoietic cell differentiation [27]. However, in vav-negative ES cells developed by homologous recombination, normal differentiation of myeloid, erythroid, and mast cell linages were observed [28, 29]. This evidence suggests that vav is not necessary for hematopoietic development. However, other members of the vav family (vav2 [30] or family mem-
04-02-97 01:34:33
eca
vav EXPRESSION IN SOMATIC CELL HYBRIDS
bers yet to be discovered) could have provided redundant pathways that compensated for the absence of vav in the knockout experiments. In the antisense experiments, there could have been cross-hybridization between the antisense probe and the mRNAs for all vav family members that resulted in the loss of all redundant pathways and the loss of hematopoietic differentiation. The absence of vav causes a decrease in T- and B-cell lineages and results in defective proliferative signaling by both the T- and B-cell antigen receptors. vav is likely to be necessary for the expansion of T- and Bcell lineages and in the early phases of antigen receptor signaling [31–33]. The expression of vav appears to be limited to cells of hematopoietic origin, trophoblasts, pancreas, and the ameloblasts of the developing tooth bud [34, 35, 49]. vav is not expressed in epithelial, muscle, brain, intestine, or kidney cells [1, 2, 4]. vav maps to human chromosome region 19p12–19p13.2 [36]. This region has been shown to be an important site of karyotypic abnormalities in leukemias, lymphomas, melanomas, and small cell lung carcinomas. However, there is no direct evidence that vav is important in these disease processes. The mechanism which regulates the tissue-specific expression of the vav oncogene is hypothesized to occur by one of two generalized processes. Either the gene is selectively silenced in cells where it is not expressed or cell-specific transcription factors stimulate the expression of vav in cell types where it is expressed. To distinguish between these mechanisms, cells which express the vav gene (human, U937 cells) were fused to cells which do not express vav (mouse, Balb/c 3T3 cells). If silencing is the predominant mechanism, the expression of the human vav gene is expected to be extinguished, whereas if vav is controlled by positive acting cell-specific transcription factors, it is anticipated that the expression of vav in the hybrid cells would be maintained and possibly activate the normally dormant mouse gene. MATERIALS AND METHODS Cell lines, electroporation, and cell fusion. The human leukemia cell line, U937 (CRL 1539), and mouse fibroblast, Balb/c 3T3 clone A31 (CCL 163), were obtained from American Type Culture Collection (Rockville, MD). The S49cyc0 cells were from the UCSF cell culture facility. HeLa cells were a kind gift from Dr. G. Stanley Cox (University of Nebraska Medical Center). HeLa cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 44 mM sodium bicarbonate, 15 mM Hepes, 50 units/ml penicillin– streptomycin, and 10% calf serum. Balb/c 3T3 and U937 cells were grown in RPMI medium 1640 supplemented with 10% fetal bovine serum (FBS). S49cyc0 cells were maintained in DMEM supplemented with 44 mM sodium bicarbonate, 15 mM Hepes, 50 units/ml penicillin–streptomycin, 10% FBS, and nonessential amino acids. The U937 cells were electroporated in a 2-mm gap BTX disposable cuvette with 40 mg pRSVneo, the generous gift of Dr. Timothy Ley,
AID
ECR 3535
/
6i1f$$$202
04-02-97 01:34:33
389
Washington University, St. Louis. Settings for the electroporation were 2950 mF (capacitance), 186 ohm (resistance), and 100 V (voltage) using a BTX 600 (BTX, San Diego, CA). The actual values for field strength and pulse length were 0.39 kV/cm and 38 ms, respectively. Following electroporation, the cells were grown in media with 400 mg/ml Geneticin (G418) (Life Technologies, Inc., Grand Island, NY). Balb/c 3T3 cells were electroporated in a 2-mm BTX disposable cuvette with pRSV hygro [32] at settings of 900 mF (capacitance), 13 ohm (resistance), and 200 V (voltage). The actual values for field strength and pulse length were 0.83 kV/cm and 7.08 ms, respectively. Hygromycin B-resistant cells were selected in media containing 200 mg/ml Hygromycin B (Calbiochem, La Jolla, CA). G418-resistant U937 and hygromycin-resistant Balb/c 3T3 cells were harvested, washed once with PBS, resuspended, and counted. Approximately 3.0 1 106 cells from each cell line were combined and centrifuged and the supernatant was removed. While continuously agitating the tube, 1 ml PEG 1500 (Boehringer Mannheim, Indianapolis, IN), prewarmed to 427C, was added dropwise to the cell pellet. After further gentle agitation, 5 ml of serum-free RPMI 1640 was added dropwise over a 5-min period, followed by 15 ml over a 1-min period. The cells were centrifuged for 10 min and the supernatant was removed and discarded. The cell pellet was resuspended in RPMI 1640 with 10% FBS and transferred to a flask. Antibiotics (400 mg/ ml G418 and 200 mg/ml Hygromycin B) were added to the flask the next day. During the first week, multinucleated adherent cells were observed along with cells in suspension. Nonadherent cells were removed and maintained separately from adherent cells, but did not survive the selection process. One month later, the hybrid cells were cloned by limiting dilution method (approximately 1 cell per every 4 wells in 96-well plates) in selective media. The cloned hybrid cells were transferred to 24-well plates and frozen at 0767C in RPMI 1640 with 10% FBS and 7% DMSO for later analysis. Preparation of RNA. RNA extractions were prepared according to the method described by Chomzynski and Sacchi [38] with modifications. Hybrid clones were revived and allowed to grow to confluence in 100 1 20 mm tissue culture plates. Adherent hybrid clones and HeLa and Balb/c 3T3 cells were harvested by pouring off the media and adding 0.5 ml of solution D (4 M guanidinium thiocyanate, 25 mM sodium citrate, pH 7, 0.5% Sarkosyl, 0.1 M 2-mercaptoethanol). The cells were removed from the surface of the plate using a cell scraper. U937 and S49cyc0 were harvested by centrifugation and disrupted in 0.5 ml of solution D. Adherent cells were phenol extracted three times, while U937 and S49cyc0 cells were extracted four times. Occasionally, additional solution D was added after the first two to three extractions in order to maintain a volume of 400– 500 ml. This improved the recovery of the extracted total RNA. The RNA was precipitated with isopropanol, dried in a speed vac for 1– 3 min, and stored at 0767C. Reverse transcriptase PCR analysis. RNA samples were digested with RNase-free DNase I (RQ1 DNase I, Promega Corp., WI) by resuspending the RNA pellets in a solution composed of 6 mM MgCl2 , 10 mM KCl, 40 mM Tris, pH 8, and 5 U RQ1 DNase I. Reactions were incubated for 15–20 min at 377C. The RNA was extracted using the method of Chomzynski and Sacchi [38], except that only one phenol:chloroform extraction was conducted. The precipitated pellets were dissolved in 10–20 ml DEPC water and stored at 0767C. RT– PCR analysis was performed using the standard and EZ protocols (Perkin–Elmer, Roche Molecular Systems, Inc., Branchburg, NJ). Reactions contained 10–20 mg of RNA and 250–500 ng of each primer. Primer pairs consisted of primers TGCACCCACTGGCTGATCCAGTGTCGGG (VA-7) with TAAGACAAAGGAACTGGGACATCTGGGG (VA-8) or VA-7 with either GAAGGTTCGAATGTTCTTAAGACAAAGG (VA-8M), from the mouse vav, or GAAGGTTCTAATGTTCTTAAGGCACAGG (VA-8H), from the human vav gene. The RNA was denatured for 2 min at 957C, reverse transcribed for 15 min at 707C, denatured at 957C for 2 min, and amplified for 35–45 cycles (957C for 1 min, 607C for 1 min) followed by a final extension
eca
390
DENKINGER AND KAWAHARA
at 607C for 7 min. Alternatively, for the EZ protocol, RNA was reverse transcribed at 607C for 30 min, followed by amplification for 40 cycles (947C for 45 s, 607C for 45 s), and a final extension for 7 min at 607C. PCR products were electrophoresed on a 4% NuSieve 3:1 agarose gel (FMC Bioproducts, Rockland, ME) in 11 TAE and 0.14 mg/ml of ethidium bromide. Gels were illuminated with a UV lamp and photographed. Southern blot analysis. Total RNA samples (20 mg) from hybrid cells were undigested or digested with 20 mg RNase A at 377C for 30 min and amplified by RT–PCR using vav primers. The reaction product was electrophoresed on a 4% NuSieve agarose gel, excised, and reamplified to obtain sufficient material for restriction digest analysis. The PCR products were pooled, phenol:chloroform extracted, ethanol precipitated, resuspended, and divided between three tubes. One tube was digested with FokI, another with VspI, and the third was undigested. The samples were electrophoresed on a 4% NuSieve gel with an aliquot from each of the original RT–PCR products that were untreated or treated with RNase A. The gel was denatured in 0.5 M NaOH/1.5 M NaCl, neutralized with 1 M Tris/1.5 M NaCl, and transferred to a nylon membrane overnight in 101 SSC. The DNA was UV crosslinked at 0.12 J/cm2 in a UV Stratalinker 1800 (Stratagene, La Jolla, CA). Blots were prehybridized at 567C in BLOTTO (0.25% nonfat dry milk in 11 SSPE) for a minimum of 6 h. The probe used for hybridization was a 1039-bp PstI vav fragment obtained from the plasmid pSK114 (the kind gift of Dr. Shulamit Katzav). The fragment was labeled by nick translation and separated from the unincorporated [32P]dCTP on a Nick Column (Pharmacia Biotech AB, Uppsala Sweden). The probe was boiled, cooled, and added to the prehybridization solution for overnight hybridization at 567C. Blots were washed twice at room temperature for 15 min in 21 SSC/ 0.1% SDS. A final wash in 0.21 SSC/0.1% SDS at 567C for 15 min was completed prior to autoradiography.
RESULTS
Cell-specific expression of the vav oncogene. A RT – PCR assay to detect the expression of both the mouse and human vav oncogene was developed. Total RNA from cell lines S49cyc0 and U937, which express the vav oncogene, and nonhematopoietic cell lines (Balb/c 3T3 and HeLa) were amplified with primer pairs VA-7/VA-8M or with VA-7/VA-8H and electrophoresed on a 4% NuSieve agarose gel. As shown in Fig. 1A, a PCR band of 216 bp was detected in the U937 and S49cyc0 cell lines, but not in the Balb/c 3T3 and HeLa cell lines. As a control, 5 mg of total RNA from each cell line is shown in Fig. 1B to confirm the integrity and quantity of RNA used in the RT – PCR reaction. The size of the PCR product is consistent with the predicted size for vav. To determine whether the PCR products were amplified from either mouse or human vav mRNA, DNase Itreated total RNA samples from U937 or S49cyc0 cells were amplified by RT–PCR, digested with either FokI or Tth111I, and electrophoresed on a 4% NuSieve agarose gel. As shown in Fig. 2, the RT–PCR product amplified from total RNA from human U937 cells was sensitive to the enzyme FokI, but not to Tth111I, since the human vav gene contains two FokI sites and cannot be cleaved by Tth111I. The 199-bp PCR product was digested into three fragments (90, 70, 39 bp), of which
AID
ECR 3535
/
6i1f$$$202
04-02-97 01:34:33
FIG. 1. Cell type-specific expression of the vav oncogene. (A) Total RNA was prepared from U937, human histiocytic lymphoma; HeLa, human cervical epitheloid carcinoma; S49cyc0, mouse lymphoma; and Balb/c 3T3, mouse embryonic fibroblast. vav was detected by RT – PCR using primer pairs VA-7/VA-8M (S49cyc0, Balb/c 3T3) or VA7/VA-8H (U937, HeLa). The PCR products were electrophoresed on a 4% NuSieve agarose gel and photographed. HaeIII-digested fX174 RF DNA was used as a size marker in lane (M). (B) Equivalent quantities of total RNA from each sample were electrophoresed on a 1.5% agarose gel to determine RNA quality and quantity.
only the 90-bp band was visible. The other two bands were obscured by nonspecific bands of a similar size. In contrast, the mouse vav gene was sensitive to Tth111I, but not FokI, since the mouse gene contains a Tth111I site and is not sensitive to cleavage by FokI. In many experiments, there was a minor fraction of the amplified product from S49cyc0 cells that was resistant to the action of Tth111I. A similar resistant fraction was also observed using VspI, another restriction enzyme which was used to discriminate between mouse and human vav transcripts (Fig. 5A). The PCR product from S49cyc0 cells which was resistant to VspI was treated with T4 DNA polymerase to create blunt ends, cloned into the HincII site of pGEM4z, and sequenced. Direct sequence analysis demonstrated that the PCR product was indeed amplified from the vav oncogene, but contained different point mutations at the VspI site. It is not known if this represents transcripts from other vav family members or represents errors made by the PCR amplification process, but we feel that the latter is more likely than the former. As a further control, DNase I-treated total RNA was either left undi-
eca
vav EXPRESSION IN SOMATIC CELL HYBRIDS
391
FIG. 2. Differential sensitivity of human and mouse vav RT–PCR products to FokI and Tth111l. RT–PCR was conducted on total RNA from U937 (lanes 1–3, 7, 8) and S49cyc0 (lanes 4–6, 9, 10) cells. The RT–PCR products were undigested (lanes 1 and 4) or digested with either FokI (lanes 2 and 5) or Tth111I (lanes 3 and 6). Total RNA was untreated (lanes 7, 9) or treated with 10 mg DNase-free RNase A for 30 min at 377C (lanes 8, 10) prior to starting the RT–PCR reaction. The marker lane (M) contains fragments of HaeIII-digested fX174 RF DNA.
gested or digested with RNase A prior to RT–PCR amplification. As shown in Fig. 2, samples treated with RNase A did not produce the 199-bp band detected in untreated samples. These results further validate the RT–PCR method used and its ability to accurately discriminate between human and mouse vav oncogene transcripts. The development of U937 and Balb/c 3T3 somatic cell hybrids. Figure 3 shows a schematic diagram of the method employed to obtain the human 1 mouse somatic cell hybrids. Human U937 cells which express the vav oncogene were transfected with pRSVneo and G418-resistant cells were obtained. Mouse Balb/c 3T3 cells which do not express the vav oncogene were transfected with pRSVhygro and selected with hygromycin to obtain resistant cells. Both resistant cell lines were fused with PEG 1500 and selected in the presence of both G418 and hygromycin. Control cells which contained either resistant gene alone did not survive the selection procedure with both antibiotics. After 1 week of selection, hybrid adherent cells were separated from the cells in suspension. None of the cells in suspension resulted in viable somatic cell hybrids that were resistant to both antibiotics. However, adherent cells, some of which were visibly multinucleated, resulted in viable somatic cell hybrids. Hybrid cells were cloned in 96well plates by limiting dilution method and frozen at 0767C. Small batches of hybrid cells were revived and analyzed for vav mRNA expression by RT–PCR. Of the 89 cloned hybrid cells, 39 were analyzed and 64% of them were positive for the expression of either the mouse or the human vav gene. The remainder of the clones were nonviable and did not survive the storage process. Cell fusion activates the normally inactive mouse vav gene. Total RNA from 11 positive clones was ampli-
AID
ECR 3535
/
6i1f$$$203
04-02-97 01:34:33
fied by RT–PCR and digested with either FokI or Tth111I. In Fig. 4, the RT–PCR product from a representative hybrid clone (F8) was not obtained if the RNA
FIG. 3. Fusion of the U937 human histiocytic lymphoma with Balb/c 3T3 mouse fibroblasts. Human U937 cells resistant to neomycin were fused with hygromycin-resistant Balb/c 3T3 cells and selected with both geneticin and hygromycin. The fused cells that were resistant to both antibiotics were cloned and subjected to RT–PCR.
eca
392
DENKINGER AND KAWAHARA
in cells which normally do not express the gene or by tissue-specific transcription factors which activate vav expression in cells that do express the gene. To determine which of the two mechanisms was responsible, human leukemia cell line U937 was fused to mouse Balb/c 3T3 cells, which do not express the vav oncogene. The majority of the hybrid cells that were obtained expressed vav and in six distinct cases, the expression of the normally dormant mouse gene was clearly detectable. All hybrid cells examined retained predominately mouse chromosomes and were morphologically fibro-
FIG. 4. Hybrid cell F8 expresses the mouse vav oncogene. Resistant U937 and Balb/c 3T3 cells were fused and cloned as described in Fig. 3. Total RNA was extracted and the RT–PCR products were undigested (lane 1 and 3) or digested with FokI (lane 4) or Tth111I (lane 5). Prior to amplification, the total RNA was treated with RNase A (lane 2). The marker lane (M) contains fragments of HaeIII-digested fX174 RF DNA.
was pretreated with RNase A. The RT–PCR product could be cleaved by Tth111I, but not by FokI. This pattern of restriction enzyme sensitivity was consistent with the RT–PCR product originating from the mouse vav mRNA transcript. RT–PCR reaction products derived from hybrid cell total RNA were also examined for differential sensitivity to digestion by FokI and VspI (Fig. 5). The restriction enzyme digestion pattern of the amplified products from five hybrid clones matched the digestion pattern obtained from U937 cells. This indicated that these five hybrid clones expressed the human vav gene (Fig. 5B). Other clones, such as F70, clearly demonstrated an identical pattern of digestion that matched the pattern found in S49cyc0 cells and expressed the normally dormant mouse vav gene. The identity of the amplification products was further confirmed by southern blotting with a probe to the vav oncogene (Fig. 5D). Of 11 clones examined, 6 were positive for the expression of the mouse vav gene, while the others expressed the human vav gene. None of the vav-positive clones led to the establishment of cell lines which stably expressed the vav gene. Invariably, the expression of vav was found to decrease with the length of time in culture and appeared to correlate with the random loss of identifiable human chromosomes (observed through karyotype analysis of the hybrid cells). The remainder of the original 25 positive clones could not be definitively analyzed due to the transient instability of the hybrid cells. DISCUSSION
The tissue-specific expression of the vav oncogene was hypothesized to be regulated either by silencing
AID
ECR 3535
/
6i1f$$$203
04-02-97 01:34:33
FIG. 5. Differential sensitivity of human and mouse vav RT– PCR products to FokI and VspI in hybrid cells. (A) RT–PCR was conducted on total RNA from U937 and S49cyc0 cells and the product was undigested or digested with either FokI or VspI. (B) RT–PCR product from hybrid cell clone F24 was undigested or digested with FokI and VspI. This pattern of sensitivity indicates the expression of the human vav gene. In C, total RNA from clone F70 was untreated or treated with 10 mg DNase-free RNase A for 30 min at 377C prior to starting the RT–PCR reaction. The RT–PCR products were undigested or digested with FokI and VspI. The pattern of sensitivity indicates the expression of the normally dormant mouse vav gene in hybrid cell clone F70, in contrast with the clone shown in B. The gel shown in C was Southern blotted with a vav probe in D. The marker lane (M) contains fragments of HaeIII-digested fX174 RF DNA.
eca
vav EXPRESSION IN SOMATIC CELL HYBRIDS
blast-like. These factors would favor silencing if it were the mechanism. Since the majority of the hybrid clones were found to express vav, this finding does not support silencing as the major mechanism which regulates vav expression. The activation of the normally dormant mouse vav gene further supports this point of view. If silencing were the mechanism, we would have expected vav expression to be extinguished in the hybrid clones; however, this was not the case. In the minority of clones where vav could not be detected, karyotypic analysis showed either the absence of recognizable human chromosomes or a small number of them. If vav expression was regulated by silencing, it would imply that negative regulatory factors were lost or inactivated in hybrid cells and thus allowed vav to be expressed. vav expression could only be observed transiently and no stable hybrid cell lines which express vav could be established. It is more likely that hybrid cells will lose, rather than gain, genetic material over time to approach a stable karyotype. Therefore, the transient nature of vav expression argues against silencing as the predominant mechanism, since it is unlikely that a negative factor could be lost and then regained to modulate vav expression. The expression of vav is more likely to be regulated by the presence of positive factors which were contributed by the U937 cells. It is not possible to definitively exclude the possibility that the cells which express the vav gene did not lose important genetic material which could have mediated a silencing event or any of the more complicated mechanisms that can be envisioned. Examination of the vav promoter will be useful in designing further experiments to test these hypotheses. Since the cells were cloned and had undergone multiple rounds of cell division, it is unlikely that original human proteins which were transferred at the time of fusion were responsible for the transcriptional activation of the mouse vav gene. It is more likely that the genes which encode the transactivating factors responsible for the tissuespecific expression of the vav gene remained active following cell fusion. If this is true, several implications follow. First, the mechanism is sufficiently conserved to activate the dormant mouse gene and we should expect the DNA recognition sites of the transactivators to be conserved between the human and the mouse species. Since vav is first detected in the fetal liver at a relatively early stage of development [34], a period where there are more similarities than differences between the mouse and the human embryo, this conservation is not unexpected. This would imply that the transgenic mouse may be a valid model for the study of the regulation of the human vav gene. Second, in the hybrid cellular environment, which should not favor vav expression because it more closely resembles the mouse phenotype, both the mouse and human vav
AID
ECR 3535
/
6i1f$$$203
04-02-97 01:34:33
393
genes are able to be maintained in an active state. This would imply that the genes of the transcription factors which regulate vav are active in the hybrid cellular environment. Since multiple rounds of cell division have likely occurred, this would mean that the genes were able to maintain their expression either by autoregulation or as a network of complementing transactivators. In either case, this situation was not ultimately stable, since all of the clones were observed to lose their ability to express vav with time. The exact cause of the loss of vav expression has not been determined, but it is likely that multiple factors or gene products are involved in expression of the final phenotype and subsequent loss of these factors leads to the loss of vav expression. Third, the mouse gene is not irreversibly inactivated. This has been observed in other systems as well [39]. Analysis of gene expression in somatic cell hybrids and multinucleate heterokaryons have shown that the differentiated state is reversible and can be reprogrammed by the action of transacting factors. Fusion of K562 cells with mouse nonerythroid cells activates the mouse embryonic e-globin gene [40]. Transient heterokaryons derived from the fusion of human nonerythroid cells with mouse erythroleukemia cells results in the activation of human b globins [41]. Mouse zinc finger transcription factor GATA-1, which is important in the regulation of the b-globin locus, is activated in heterokaryons derived from human K562 cells and mouse ES cells, which do not express GATA-1 or globin genes [42]. The existence of complex reprogramming schemes is illustrated by the fusion of mouse myoblasts with human nonmuscle cells. These fusions resulted in the activation of the human muscle regulators followed by the activation of the mouse muscle genes [43]. The regulators that are known to be activated in muscle heterokaryons include members of the helixloop-helix family of myogenic transcription factors [44, 45]. Other examples where genes that are normally dormant in one fusion partner and are activated by factors from the other include the galecitin-1 (L-14-I) gene [46], the mouse a-fetoprotein gene [47], and murine macrophage class II genes [48]. These experiments support the hypothesis that the vav oncogene is regulated by the presence of one or more transacting factors. Analysis of the promoter region of the vav gene is under way in our laboratory to further test the hypothesis suggested by this work. The study of the transcriptional regulation of vav will likely provide new information that will contribute to our understanding of its role in the developmental regulation of the hematopoietic system. We gratefully acknowledge the assistance of Renee Fordyce-Boyer and her staff at the Hattie B. Munroe Center for Human Genetics for their kind assistance with the karyotyping of some of the somatic cell hybrids used in this work. This work was supported by Grant GM46665 from the National Institutes of Health.
eca
394
DENKINGER AND KAWAHARA
REFERENCES 1. Katzav, S., Martin-Zanca, D., and Barbacid, M. (1989) EMBO J. 8, 2283–2290. 2. Katzav, S., Cleveland, J. L., Heslop, H. E., and Pulido, D. (1991) Mol. Cell. Biol. 11, 1912–1920. 3. Coppola, J., Bryant, S., Koda, T., Conway, D., and Barbacid, M. (1991) Cell Growth Differ. 2, 95–105. 4. Adams, J. M., Houston, H., Allen, J., Lints, T., and Harvey, R. (1992) Oncogene 7, 611–618. 5. Galland, F., Katzav, S., and Birnbaum, D. (1992) Oncogene 7, 585–587. 6. Boguski, M. S., Bairoch, A., Attwood, T. K., and Michaels, G. S. (1992) Nature 358, 113. 7. Quest, A. F. G., Bloomenthal, J., Bardes, E. S. G., and Bell, R. M. (1992) J. Biol. Chem. 267, 10193–10197. 8. Gulbins, E., Coggeshall, K. M., Langlet, C., Baier, G., BonnefoyBerard, N., Burn, P., Wittinghofer, A., Katzav, S., and Altman, A. (1994) Mol. Cell. Biol. 14, 906–913. 9. Katzav, S., Packham, G., Sutherland, M., Aroca, P., Santos, E., and Cleveland, J. L. (1995) Oncogene 11, 1079–1088. 10. Khosravi-Far, R., Chrzanowska-Wodnicka, M., Solski, P. A., Eva, A., Burridge, K., and Der, C. J. (1994) Mol. Cell. Biol. 14, 6848–6857. 11. Bustelo, X. R., Suen, K.-L., Leftheris, K., Meyers, C. A., and Barbacid, M. (1994) Oncogene 9, 2405–2413. 12. Clevenger, C. V., Ngo, W., Sokol, D. L., Luger, S. M., Gewirtz, A. M. (1995) J. Biol. Chem. 270, 13246–13253. 13. Hobert, O., Jallal, B., and Ullrich, A. (1996) Mol. Cell. Biol. 16, 3066–3073. 14. Romero, F., Dargemont, C., Pozo, F., Reeves, W. H., Camonis, J., Gisselbrecht, S., and Fischer, S. (1996) Mol. Cell. Biol. 16, 37–44. 15. van Lohuizen, M., Verbeek, S., Scheijen, B., Wientjens, E., van der Gulden, H., and Berns, A. (1991) Cell 65, 737–752. 16. van der Lugt, N. M. T., Domen, J., Linders, K., van Roon, M., Robanus-Maandag, E., te Riele, H., van der Walk, M., Deschamps, J., Sofroniew, M., van Lohuizen, M., and Berns, A. (1994) Genes Dev. 8, 757–769. 17. Margolis, B., Hu, P., Katzav, S., Li, W., Oliver, J. M., Ullrich, A., Weiss, A., and Schlessinger, J. (1992) Nature 356, 71–74. 18. Bustelo, X. R., Ledbetter, J. A., and Barbacid, M. (1992) Nature 356, 68–71. 19. Bustelo, X. R., and Barbacid, M. (1992) Science 256, 1196–1199. 20. Alai, M., Mui, A. L. F., Cutler, R. L., Bustelo, X. R., Barbacid, M., and Krystal, G. (1992) J. Biol. Chem. 267, 18021–18025. 21. Rottapel, R., Turck, C. W., Casteran, N., Liu, X., Birnbaum, D., Pawson, T., and Dubreuil, P. (1994) Oncogene 9, 1755–1765. 22. Evans, G. A., Howard, O. M. Z., Erwin, R., and Farrar, W. L. (1993) Biochem. J. 294, 339–342. 23. Matsuguchi, T., Inhorn, R. C., Carlesso, N., Xu, G., Drucker, B., and Griffin, J. D. (1995) EMBO J. 14, 257–265. 24. Uddin, S., Katzav, S., White, M. F., and Platanias, L. C. (1995) J. Biol. Chem. 270, 7712–7716.
25. Clevenger, C. V., Ngo, W., Sokol, D. L., Luger, S. M., and Gewirtz, A. M. (1995) J. Biol. Chem. 270, 13246–13253. 26. Platanias, L. C., and Sweet, M. E. (1994) J. Biol. Chem. 269, 3143–3146. 27. Wulf, G. M., Adra, C. N., and Lin, B. (1993) EMBO J. 13, 5065– 5074. 28. Zmuidzinas, A., Fisher, K. D., Lira, S. A., Forrester, L., Bryant, S., Bernstein, A., and Barbacid, M. (1995) EMBO J. 14, 1–11. 29. Zhang, R., Tsai, F. Y., and Orkin, S. H. (1994) Proc. Natl. Acad. Sci. USA 91, 12755–12759. 30. Henske, E. P., Short, M. P., Jozwiak, S., Bovey, C. M., Ramlakhan, S., Haines, J. L., and Kwiatkowski, D. J. (1995) Ann. Hum. Genet. 59, 25–37. 31. Tarakhovsky, A., Turner, M., Schaal, S., Mee, P. J., Duddy, L. P., Rajewsky, K., and Tybulewicz, V. L. J. (1995) Nature 374, 467–470. 32. Zhang, R., Alt, F., Davidson, L., Orkin, S. H., and Swat, W. (1995) Nature 374, 470–473. 33. Fischer, K. D., Zmuldzinas, A., Gardner, S., Barbacid, M., Bernstein, A., and Guidos, C. (1995) Nature 374, 474–477. 34. Bustelo, X. R., Rubin, S. D., Suen, K. L., Carrasco, D., and Barbacid, M. (1993) Cell Growth Differ. 4, 297–308. 35. Higuchi, T., Kanzaki, H., Fujimoto, M., Hatayama, H., Watanabe, H., Fukumoto, M., Kaneoko, Y., Higashitsuji, H., Kishishita, M., Mori, T., and Fujita, J. (1995) Biol. Reprod. 53, 840– 846. 36. Martinerie, C., Cannizzaro, L. A., Croce, C. M., Huebner, K., Katzav, S., and Barbacid, M. (1990) Hum. Genet. 86, 65–68. 37. Bonapace, I. M., Sanchez, M., Obici, S., Gallo, A., Garofalo, S., Gentile, R., Cocozza, S., and Avvedimento, E. V. (1990) Mol. Cell. Biol. 10, 1033–1040. 38. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156–159. 39. Baron, M. H. (1993) Curr. Opin. Cell. Biol. 5, 1050–1056. 40. Baron, M. H., and Maniatis, T. (1986) Cell 46, 591–602. 41. Baron, M. H., and Maniatis, T. (1991) Mol. Cell. Biol. 11, 1239– 1247. 42. Baron, M. H., and Farrington, S. M. (1994) Mol. Cell. Biol. 14, 3108–3114. 43. Hardeman, E. C., Chiu, C. P., Minty, A., and Blau, H. M. (1986) Cell 47, 123–130. 44. Rudnick, M. A., and Jaenisch, R. (1995) BioEssays 17, 203–209. 45. Thayer, M. J., and Weintraub, H. (1990) Cell 63, 23–32. 46. Chiariotti, L., Benvenuto, G., Zarrilli, R., Rossi, E., Salvatore, P., Colantuoni, V., and Bruni, C. B. (1994) Cell Growth Differ. 5, 769–775. 47. Spear, B. T., and Ellis, A. W. (1995) Somat. Cell. Mol. Genet. 21, 19–31. 48. Sartoris, S., Scupoli, M. T., Scarpellino, L., Paiola, F., Jotterand-Bellomo, M., Tridente, G., and Accolla, R. S. (1990) J. Immunol. 145, 1960–1967. 49. Schuebel, K., Bustelo, X. R., Nielsen, D. A., Song, B.-J., Barbacid, M., Goldman, D., and Lee, I. J. (1996) Oncogene 13, 363– 371.
Received November 4, 1996 Revised version received January 31, 1997
AID
ECR 3535
/
6i1f$$$204
04-02-97 01:34:33
eca