Gene regulation and genetic susceptibility to neoplastic transformation: AP-1 and p80 expression in JB6 cells.

The mouse epidermal JB6 cell system consists of clonal genetic variants that are sensitive (P+) or resistant (P-) to the promotion of neoplastic transformation by phorbol esters and other tumor-promoting agents. P+ cells display AP-1-dependent phorbol-ester-inducible transactivation of gene expression, whereas P- cells have a defect in transactivation. Transfection of promotion sensitivity gene pro-1 into P- cells reconstituted both P+ phenotype and AP-1-dependent phorbol-ester-inducible transactivation. P- and P+ cells exhibited induction of c-jun and c-fos messenger RNA levels by phorbol ester, but P- cells had significantly lower basal and induced levels of jun mRNA than P+ cells. Basal and induced levels of c-jun protein were significantly lower in P- cells as well. Differences in levels the 80-kDa pI 4.5 protein p80 were also observed in JB6 cells as a function of preneoplastic progression; high levels of p80 protein and mRNA were observed in P- cells, intermediate levels in P+ cells, and negligible levels were observed in transformed derivatives of JB6 cells. Phorbol ester treatment induced phosphorylation but not synthesis of p80. These data are consistent with the hypotheses that AP-1 is required in the signal transduction pathway for promotion of neoplastic transformation by tumor promoter, that pro genes may control AP-1 activity, that threshold levels of Jun mRNA and protein may play a role in transactivation and promotion sensitivity, and that the p80 protein in JB6 cells may behave in vivo as a suppressor of cellular transformation. ImagesFIGURE 4.FIGURE 5.


Introduction
Recent progress in understanding the genetics of susceptibility to tumor promotion has come from in vivo studies in the mouse by Drinkwater (1), Di-Giovanni (2), Gould et al. (3), and Malkinson (4) and from studies with mouse epidermal JB6 cell variants in our laboratory and those of others. The JB6 cell lines were derived from untreated primary BALB/c mouse epidermal cell cultures that gave rise at a very low frequency to immortalized cell lines (5). The immortalized JB6 cells underwent further change to stably acquire sensitivity to induction of anchorage independence and tumorigenicity by phorbol esters such as tetradecanoyl phorbol-13-acetate (TPA) and other tumor promoters (5,6). Nonselective cloning soon after observation of this change yielded clonal lines that were stably sensitive (P+) or resistant (P-) to (7)(8)(9). The percentage of cells in agar that display TPA-induced anchorage independence is typically in the range of 20% for P+ cells and 0.2% or less for Pcells. These JB6 variants sensitive to tumor-promoter-induced transformation appear to undergo a transition analogous to second-stage tumor promotion in vivo since second-stage tumor promoters such as mezerein induce transformation, and second-stage inhibitors, such as retinoids, but not first-stage inhibitors, such as antiproteases, block the induced transformation (10). Table 1 summarizes the results of a number of studies that have used the P+ and P-variants to identify steps that may be required in the signal transduction pathway for promoter-induced transformation. Candidates for required events would be expected to show a P4/Pdifferential in some or all clonal variants tested; responses not showing a P+/Pdifferential may or may not be required ones. P+ and P-cells showed similar responses to mitogenic stimulation by phorbol esters, and displayed similar induction of protease activity. They also showed similar levels of protein kinase C activity. Whether there are protein kinase C subtype differences has not yet been established. Mitogenic stimulation from quiescence (11) Decreased synthesis of collagen, fibronectin (21-23) Induction of proteases: major excreted protein and plasminogen activator (N. H. Colburn and K. Hirano, unpublished) Increased glucose uptake and ornithine decarboxylase activity (24,25) Similar protein kinase C activation and substrates (10,12,13) What distinguishes P+ from P-cells?
Activated versus inactive pro 1 and pro 2 (14,15,26) Levels of an 80-kDa/pI 4.5 phosphoprotein (10,13,20) Ganglioside synthesis response to TPA (16,27) Induction of AP-1/jun-dependent transactivation of gene expression (19) and jun mRNA and protein levels Induction of 15-kDa and 16-kDa nuclear proteins (K. Hirano and N. H. Colburn, in preparation) DNA damage and poly ADP ribosylation responses (18) JB6 P+ but not P-cells possess activated DNA sequences called pro 1 and pro 2 that confer sensitivity to tumor-promoter-induced transformation when transferred into P-cells (14). Pro 1 appears to encode a transcript whose synthesis is catalyzed by RNA polymerase III (15). The mode of activation of pro 1 to a P+ active structure is not yet known. P+/Pdifferences in TPA modulated ganglioside synthesis (16) and induced synthesis of 15 and 16 kDa nuclear proteins (K. Hirano, B. Smith, and N.H. Colburn, in preparation) have also been observed. Nakamura et al. have reported that treatment of JB6 P+ cells with xanthine-xanthine oxidase, which generates superoxide anion and subsequently other active oxygen species, promotes neoplastic transformation (1 7). Cerutti and co-workers have found that JB6 P-cells show a greater DNA damage and ADP ribosylation response to transformation-promoting xanthine-xanthine oxidase treatment than do P+ cells (18). These results suggest that greater oxidant defense may be an important component of the P+ phenotype. Finally, Table 1 shows two other sets of genes whose expression and/or activity is differential in P+ and P-cells. These are the AP-1 transactivating complex composed of members of the jun and fos multigene families (19), and an 80-kDa pI 4.5 protein (p80). The present communication focuses on AP-1 and p80 and their roles in preneoplastic progression in JB6 cells.
AP-1 /jun-Mediated Transactivation of Gene Expression Is Differentially Induced by Tumor-Promoting Agents in P+ Cells and P-Cells TPA treatment of cells induces the expression of a number of genes, some of which encode proteins thought to be key participants in implementing neoplastic transformation (28,29). The list of such TPA-inducible genes includes several proto-oncogenes, including c-myc and c-fos, proteases, including collagenase, stromelysin and plasminogen activator, numerous virally encoded genes, and other sequences (28,29). Investigation of the identities of trans-regulatory factors that would be predicted to exist and to control cis enhancer elements in the promoter regions of these sequences led to the discovery of AP-1 transactivating protein (30,31). The complex consists of a heterodimeric species containing products of the jun and fos multigene families (32)(33)(34)(35); homodimeric Jun protein complexes have also been detected (35).
TPA induces AP-1 binding to a consensus upstream regulatory enhancer sequence (TGACTCA) in several genes thought to be involved in oncogenesis (30). The binding of AP-1 to its enhancer is likely to regulate transcription of these genes. We therefore hypothesized that AP-1 function is specifically required for the promotion phase of neoplastic transformation. If AP-1 controls a set of effector genes required for tumor-promoter-induced transformation, then some promotion-resistant variants may owe their resistance to a defect in tumor promoter inducibility of AP-1 function.
To investigate this hypothesis, mouse JB6 P+ and Pvariants were treated with TPA after transient transfection with plasmid 3XTRE-CAT, a construct that has three tandem TPA-responsive cis-enhancer elements attached to the Herpes simplex virus thymidine kinase (HSV-TK) promoter and a gene encoding chloramphenicol acetyltransferase (CAT) (30). Induced CAT gene expression in this system depends upon tumor-promoter-mediated activation of cellular AP-1 activity. This system enabled us to test the prediction that cellular genetic variants resistant to the promotion of neoplastic transformation by TPA would possess defective AP-1 transactivation function (19).
As shown in Figure 1A, P+ Cl 41 cells displayed significant inducible APl-dependent CAT protein synthesis within 3.5 hr of TPA treatment (but not within 1.5 hr). TPA-induced expression of CAT reached a maximum of 900 to 1000 units of enzyme activity in the P+ Cl 41 cells after 48 hr (5to 6-fold induction) and was persistent over at least 100 hr. In contrast, P-Cl 30 cells showed little inducibility by TPA at any time point tested from 0 to 100 hr. Uptake of the transfected 3XTRE-CAT plasmid in the R+ and Pcells was equalized as measured by Southern hybridization analysis of transient transfectants (19) (not shown). Furthermore, equal levels of CAT activity were observed in P+ Cl 41 cells and P-Cl 30 cells transfected with the constitutively expressed plasmid pRSVCAT at doses such that DNA uptake (copies per cell) was equalized. APi-dependent CAT synthesis was also induced in P+ cells but not in P-cells by epidermal growth factor (EGF) and by high concentrations of serum, two additional transformation promoting agents for JB6 cells (19) (not shown). These data point to a specific regulatory defect at the level of AP-1 function in the promotion resistant Cl 30 cells.  24 30 48 Time, hours FIGURE 1. TPA induces APi-dependent CAT synth but not in Pcells. (A) Differential CAT inductio cells and P-Cl 30 cells. P+ Cl 41 cells and P-Cl plated, TPA treated in 2% fetal calf serum, harveste for CAT enzyme activity as described (19). TPA a vent DMSO treatments were conducted at each time showing differential inducibility were obtained in t dent experiments. Data from a representative e: plotted as units of CAT enzyme activity per 4 x 10 cells divided by the activity for 4 x 104 DMSO-t: Differential APl-dependent CAT gene r( P+ and P-cells was also observed in tw dently derived clonal JB6 P+ and P-cell is shown in Figure 1B. The independent I is designated Cl 25, and its time course f pendent CAT synthesis as a function of' ment reveals nonresponsiveness, as was c the P-Cl 30 cell line. The P+ clonal variant, designated pNP-20, was derived from parental P-Cl 30 cells by ring cloning of G418-selected transfectants > C141 ,TPA generated by introduction of a plasmid construct harboring the mouse promotion sensitivity gene pro 1 (28) and a neomycin resistance marker. This variant displayed anchorage-independent colony formation upon In contrast, the P-recipient cells did not display induced expression. Note that transfectants harboring the neo resistance marker have shown no ability to activate AP-1 function (36). Observation of defective AP-1 function in two in-pNP-20 (P+) dependent P-clonal variants and of competent AP-1 function in two independent P+ clonal variants demonstrates an association between AP-1 function and promotion of transformation. It is consistent with the hypothesis that AP-1 function is required along the signal transduction pathway for promotion of neoplastic transformation by TPA. Furthermore, the fact that introduction of a gene that confers sensitivity to promotion of transformation by TPA reconstitutes 0230 (P-) AP-1 function to a defective P-cell supports the hypothesis that pro genes can execute control over the activity of AP-1.
Measurement of jun and fos mRNA Levels in P+ and P-Cells TPA stimulates the accumulation of c-jun mRNA esis in P+ es in murine and human fibroblasts (42,43) and induces )ns in P+ cells transcription of the c-fos gene in a number of systems 130 cells were (44). Therefore, the possibility that differential AP-1 ,d, and assayed dependent transactivation in P+ and P cells was due nd control solto differences in expression of c-jun or c-fos was expoint. Results amined. Total RNA was isolated from JB6 cells at xperiment are various times after treatment with 10 ng/mL TPA, }4 TPA-treated and the relative levels of c-jun or c-fos mRNA were reated cells at determined by Northern blotting and densitometry ylor depfihendicol scanning analyses of the resulting autoradiographs TPA-inducible (Fig. 2, upper panel). TPA-induced c-jun mRNA lev-P-recipients els were higher in P+ cells than in P-cells at all times ted pNP cells, after addition of TPA, and basal levels of this message re treated and were also 5-to 10-fold higher in the P+ cells. In several lescribed in A, of TPA treat-experiments c-jun mRNA was undetectable in untreated P-cells. After 24 hr, the amount of c-jun mRNA returned to basal levels in the P+ cells, but remained slightly elevated in P-cells. egulation in In contrast to the results obtained for c-jun, TPAro indepen-stimulated levels of c-fos mRNA were essentially variants, as equal in P+ and P-cells (Fig. 2, lower panel). Basal P-cell line c-fos mRNA levels were 2.5to 10-fold higher in the or APl-de-P-cells. The degree of induction of c-fos message by TPA treat-TPA was much greater than that observed for c-jun, )bserved in and the levels of c-fos mRNA declined much more rapidly than those of c-jun mRNA, returning to nearbasal amounts after 2 hr of TPA treatment.
The above results suggest that the differential transactivation observed in response to TPA in promotionsensitive and resistant JB6 cells may be accounted for at least in part by differences in TPA-stimulated accumulation of c-jun mRNA. However, differential TPAinduced expression of c-fos, at least at the message level, can be ruled out as a contributor to the observed differences in AP-1-dependent transactivation. While AP-1-dependent transactivation via the 3XTRE in re-lp+ sponse to TPA is virtually undetectable in P-cells, TPA still stimulates c-jun mRNA accumulation to a significant degree in these cells. It is possible that a threshold iPlevel of c-jun mRNA and protein must be reached in order to stimulate AP-1 dependent transactivation above basal levels. 8 subjected to autoradiography and X-ray films were Time of TPA Treatment (hr) scanned by densitometry analysis. Figure 3 shows a representative time course experiment of TPA treatment in P+ Cl 41 and P-Cl 30 cells over 24 hr. As was the case forjun mRNA, Jun protein was observed at significantly lower levels in the P-Cl RE 2. TPA differentially induces c-jun but not c-fos mRNA 30 cells than in the P+ Cl 41 cells at time points ex-,cumulation in P+ and P-cells. Promotion-resistant (P-) and amined. Basal levels were approximately 5-fold lower romotion-sensitive (P-) JB6 cells were grown to near-conflu-in the P-cells than in the P+ cells, and induced levels ice in 5% serum in T150 flasks and then switched to 2% serum ranged from 2-to 10-fold lower during the time course.
tr 24 hr. The cells were treated with 10 ng/mL TPA (16 nM), + Lrvested by trypsinization and centrifugation, and total RNA In the P cells, TPA iduced a rapid accumulation of as extracted at the indicated times as described previously Jun protein within 30 min, followed by a decline over 5). RNA (10 ,ug/sample) was subjected to electrophoresis and the duration of the experiment; in the P cells a delay orthern blot analysis (45) using v-jun or v-fos cDNA probes in the onset of induction was observed In both cell lines ifts of P. Vogt and T. Curran, respectively) labeled with 32P y the random priming method (Pharmacia). The relative levels Jun protei levels declied by 24 hr of TPA treatment. Ic-jun mRNA (upper panel) and c-fos mRNA (lower panel) These data support the hypothesis that control ofJun ere determined by densitometric analysis of the resulting au-protein levels in JB6 P' and P-cells is pretranslational radiographs. c-jun mRNA levels are expressed as values rel-and is most likely caused by differential accumulation  Figure  2. Nuclei were prepared as described by Bos et al. (39) with minor modifications (37). Per sample, 1 x 106 nuclei were run in 10% Laemmli SDS polyacrylamide gels (39,40), transferred by Western blotting onto BA85 nitrocellulose filters (Schleicher and Schuell), and blotted with 5 ,ug/mL affinity purified rabbit anti PEP-2 antiserum, according to the methods of Towbin et al. (41), with modifications (37), followed by 5 x 10' cpm/mL 126I protein A (PRI/FCRF). Filters were exposed overnight to Kodak XAR film and resulting autoradiograms were scanned in an LKB Ultroscan XL Densitometer. The representative experiment in Figure 3 shows results from densitometric analysis of P+ and P-cells treated with TPA over a 24-hr time course. Data are plotted as relative optical intensity, using a value of 1.0 for the TPA-untreated 0 time control in the P+ Cl 41 cells.
tion and possibly neoplastic transformation. Since transactivation responses appear to persist beyond the time course of Jun induction it is possible that, while a Jun threshold may be necessary, it may not be sufficient for transactivation or promotion in this system. The precise role and mechanisms of control of cellular responsiveness by the AP-1 complex continue to be investigated.

Differential Expression of an 80-kDa/pI 4.5 Protein during Preneoplastic Progression
Previous investigations in this laboratory demonstrated a differential basal and induced levels of phosphorylated of p80 during preneoplastic progression in JB6 cells (46). Two-dimensional gel electrophoresis of proteins labeled in vivo with 2P-orthophosphate showed high levels of a phosphorylated 80 kDa/pI 4  (Bottom panel) P-, P+, and transformed total cellular RNAs were isolated according to the procedure of Deeley et al. (51) and subjected to Northern analysis as described in Simek et al. (20). Each lane contained 10 ,ug of total cellular RNA. The ifiters were exposed to Kodak XAR film for 2 days. and essentially none in neoplastically transformed derivatives of JB6 cells (46). Exposure to TPA caused elevated p80 phosphorylation in Pand P+ cells, but not in transformed cells. These results raised the question of whether this differential regulation was occurring at the pretranslational or posttranslational level. To determine whether the regulation was at the  Figure 4. Samples were loaded onto a 10% polyacrylamide gel. Panel A was dried and exposed to Kodak XAR film for 24 hr. Lanes 1, 3, 5, 7, and 9 preimmune; lanes 2, 4, 6, 8, and 10, p80 peptide antiserum.
level of p80 synthesis, proteins from cell lysates of JB6 P-, P+ and transformed cells were analyzed by immunoblotting with a peptide antiserum (47) specific for p80. As shown in the top panel of Figure 4, differential expression ofp80 protein was observed, with high levels of expression in P-cells, intermediate levels in P+ cells, and little or no expression in neoplastically transformed cells. Similar results were observed when a second set of independently derived JB6 P, P+ and transformed clonal variants were analyzed. A p80 cDNA which had been cloned by p80 peptide antibody screening (20) was used to analyze JB6 cellular p80 mRNA expression and to determine the extent to which p80 protein levels might be limited by p80 mRNA concentration. As shown in the bottom panel of Figure 4, when this p80 clone was used as a probe against P-and P+ total cellular RNA, a single 2.6-kb band was observed, but little or no hybridization was seen with RNA from transformed cells. Densitometric analysis from three experiments showed the mean value for the hybridizing band in P+ RNA was 50 + 2% and transformed RNA was 2.5 + 0.4% of the P-RNA value. This pattern was nearly identical to that observed for the differential expression of p80 protein in these cells, indicating that intracellular p80 protein concentration is regulated by the levels of p80 mRNA.
To determine whether TPA induces p80 phosphorylation and thereby test the hypothesis that p80 is a PKC substrate, TPA treated JB6 P-cell lysates were immunoprecipitated with p80 peptide antiserum. Fig-ure 5A shows the pattern of p80 phosphorylation in JB6 P-cells treated with TPA for 0, 2, 5, 8, and 24 hr. This experiment showed an increase in p80 phosphorylation with a 6-fold maximum at 2 hr after initiation of TPA treatment (lane 4) that persisted for 5 hr (lane 6) and returned to basal levels by 24 hr (compare lanes 10 and 2). This time course was comparable to studies done in this laboratory (13) and by others (48) not using p80 antibody. The decrease in p80 phosphorylation occurred after 24 hr of TPA treatment was paralleled in JB6 cells by a decrease in protein kinase C activity and concentration (data not shown). This result correlated with findings demonstrating that treatment of cells with phorbol esters leads to progressive downmodulation of phorbol ester receptors (49) followed by disappearance of protein kinase C activity (50). Thus, phosphorylation of p80 in JB6 cells is dependent on protein kinase C, and p80 may or may not be a direct protein kinase C substrate.
To determine whether the observed increase in p80 phosphorylation reflected an increase in synthesis or was controlled posttranslationally, JB6 P-cells were exposed to TPA for 0.5, 1, 4, and 24 hr. Cell lysates were then analyzed by immunoblotting for levels of p80, using p80 peptide antiserum. The results of this experiment are shown in Figure 5B. The level of p80 did not increase after tumor promoter treatment but actually appeared to decrease after prolonged TPA exposure (24 hr). Shorter TPA exposure times were also tested and again showed no increase in p80 synthesis. The results from this experiment confirmed that p80 synthesis was not increased by exposure to TPA. In addition, total cellular RNA, isolated from P-cells after TPA treatment for 0, 4, and 24 hr, showed no difference in the level of p80 hybridizing RNA (data not shown). Therefore, this study indicates that TPA treatment specifically induces the phosphorylation and not the synthesis of the p80 protein.

Conclusions
Regulation of c-jun expression in response to TPA and the subsequent activation ofa specific set oftarget genes in response to AP-1 may play a part in the tumor promotion process. A number of phorbol ester and growth factor inducible genes have been found to contain TREs in their promoter regions, and their expression is believed to be regulated via binding of AP-1 to this promoter element. These genes include stromelysis/transin (30,53) and collagenase (30,54), proteases that may play a role in tumor invasiveness and metastasis, metallothionein IIA (30), and interleukin-2 (55). In addition, AP-1 is thought to be involved in the positive autoregulation of c-jun (56) and in both positive and negative autoregulation of c-fos (57)(58)(59)(60). Thus, differences in tumor-promoter-induced gene activation by the AP-1 transcription factor between P+ and P-cells may lead to different patterns of expression of key effector genes. The defective APl-dependent transactivation observed in P-cells may account for their promotion-resistant phenotype. Relevant target genes for the promotion process in these cells remain to be identified.
In addition to transiently regulated genes, certain genes are constitutively switched on or switched off during preneoplastic progression (61,62). The 80 kDa protein observed in JB6 cells appears to be such a negatively regulated protein. The fact that p80 levels decrease as cells progress toward the neoplastic end point is compatible with its postulated role as a tumor suppressor. Whether p80 is coordinately regulated by, or with, the AP-1 protein via signals transduced by protein kinase C is currently unknown. Further investigations are underway to elucidate whether and by what mechanisms AP-1 and p80 modulate susceptibility to transformation promotion and progression to the neoplastically transformed cellular phenotype.