Dexamethasone-mediated regulation of 3-methylcholanthrene-induced cytochrome P-450d mRNA accumulation in primary rat hepatocyte cultures.

We have previously demonstrated that cytochrome P-450c and P-450d mRNAs can be induced by 3-methylcholanthrene (MCA) in primary cultures of rat hepatocytes grown in serum-free hormonally defined medium (HDM) plus minimal salts (Silver, G., and Krauter, K. S. (1988) J. Biol. Chem. 263; 11802-11807). Such cultures were used to investigate the role of the individual hormonal components present in the medium in the hydrocarbon-mediated induction process. Replacing HDM with minimal salts plus 10% fetal bovine serum (FBS) resulted in a 4-fold reduction in the accumulation of P-450d mRNA in response to MCA. In contrast, no effect was seen on induced levels of P-450c mRNA. Mixing experiments, in which primary cultures of hepatocytes were grown in medium containing HDM plus 10% FBS, indicated that there was no negative acting component present in FBS, but rather there was a positive acting component present in the mixture of hormones in HDM which permitted P-450d induction by MCA. Testing the effects of singly deleting each of the 10 components in HDM on MCA-induced P-450d expression demonstrated that dexamethasone was the only factor which affected the induction of P-450d. Deletion of this component from HDM resulted in a 4-fold decrease in the maximum MCA induced expression of P-450d mRNA. Moreover, supplementation of minimal salts plus 10% FBS with dexamethasone restored full P-450d inducibility by MCA. Deletion of the other components from HDM had no effects on P-450d mRNA accumulation. Although substratum clearly contributed to the quality of primary hepatocyte cultures, we were unable to demonstrate any role of the substratum on MCA induction of P-450d. In vitro nuclear run-on experiments revealed that dexamethasone had little effect on the rate of transcription of the P-450d genes. Therefore, the effect of dexamethasone on induction must be at the posttranscriptional level.

We have previously demonstrated that cytochrome P-450~ and P-450d mRNAs can be induced by 3-methylcholanthrene (MCA) in primary cultures of rat hepatocytes grown in serum-free hormonally defined medium (HDM) plus minimal salts (Silver, G., and Krauter, K. S. (1988) J. Biol. Chem. 263;11802-11807 The primary response of most organisms to exposure to aryl hydrocarbons, including 3-methylcholanthrene (MCA),' P-napthoflavone, and 2,3,7,8-tetrachlorodibenzo-p-dioxin, is the induction of a set of genes whose products act to metabolize these xenobiotics (1, 2). In rat liver, cytochromes P-450~ and P-450d are the major MCA-induced proteins. Although both genes are expressed after carcinogen administration in uiuo, in cultured cell lines, only P-450~ and not P-450d is induced by MCA (1, 3-6). The failure to observe induction of P-450d in vitro has been postulated to be due to hypermethylation of the gene in cultured cells or the absence of required truns-acting factors in the cell types tested (1, 7).
Recently, we were among the first to report the MCAinduced expression of rat cytochrome P-450d mRNA in a cultured hepatocyte cell system (8,9). Accumulation of P-450d mRNA in response to the hydrocarbon was rapid, and reached levels of expression similar to those seen in rat liver in vivo. In addition, we showed that P-450d induction was regulated at a post-transcriptional level both in rat liver and in the cultured hepatocyte system (8). In our studies, we had utilized culture conditions which included hormonally defined medium, collagen substratum, and the absence of serum. Work by others had demonstrated that components of fetal bovine serum can extinguish expression of certain liver-specific genes in culture (10). We wished to explore the relationship between the cell growth conditions and the induction of P-450d by MCA. In this report, we investigated the contributions of each component in the defined medium with respect to MCA induction, and found that induction of P-450d mRNA by MCA was dependent on the presence of only one component of HDM, dexamethasone, in the culture medium. We further showed that serum, a component of the growth medium used by many other investigators, has no effect on the ability of primary cultured cells to induce P-450d in response to MCA. Finally, using in vitro nuclear run-on experiments, we demonstrated that the dexamethasone effect is entirely post-transcriptionally mediated. Of the 1080 bases in the fragment, 910 (84%) are identical to the rat P-450d mRNA sequence (l&16) with most of the mismatch in the last 250 bases of the sequence.

MATERIALS
The first 830 bases have 777 (93%) identical with the rat mRNA with no insertions or deletions in the alignment.
The Ps-450 fragment cross-hybridizes weakly with rat P-450~ due to its 60% homology over the length of the clone. Thus the probe detects both P-450d and P-450~ mRNAs on Northern blots. Control experiments indicate that the levels of expression of P-450~ mRNA estimated by the cross-hybridization with the Pa-450 probe are directly proportional to the levels seen using the P-450c-specific probe, and thus the cross-hybridization can be used to estimate the relative induction of P-450~ by MCA in any experiment (data not shown).
As an internal loading control we have used a cDNA clone to the liver-abundant mouse al-antitrypsin mRNA (12,17 Hybridization and washing of filters were carried out as previously described (8).

RESULTS
Fetal Bovine Serum Does Not Contain an Inhibitor of P-450 mRNA Induction-The study of the molecular basis of gene induction by aryl hydrocarbons has been greatly facilitated by the ability to obtain gene induction in cultured cells.
Thus, a considerable body of information has been obtained describing the regulation of mouse PI-450 and rat P-450~ genes, which are induced by hydrocarbons in many cell types tested (18)(19)(20)(21)(22)(23)(24)(25). On the other hand, very little is known about the molecular basis of the hydrocarbon induction of rat P-450d or mouse P3-450 genes, mainly because induction of P-450d in cultured cells has only recently been attained (8,9). We have previously shown that P-450~' could be induced in primary hepatocyte cultures when cells were grown in hormonally defined medium on collagen substratum (8). However, it was not clear from those experiments whether or not defined medium was required to obtain MCA induction. It was important to determine if our success at obtaining MCAinduced expression of P-450d was the result of the growth conditions employed or rather was related to a less tangible part of the cell preparation protocol we used. In the studies described below, we determine the precise culture components which are required to permit P-450d induction in primary hepatocytes, and by simple mixing experiments demonstrate that negative effecters of P-450d inducibility are not present in serum.
One hypothesis to explain the failure of others to obtain MCA induction of P-450d in cultured hepatocytes is that serum used in the medium inhibits the expression of P-450d by an unknown mechanism. To address this issue, we compared the induction by MCA of both P-450~ and P-450d mRNAs in cells cultured in HDM (RPM1 1640 minimal salts plus hormones) to cells cultured in RPM1 minimal salts supplemented with 10% FBS, both grown on type I collagen substrata. As shown in Fig. 1, cells grown in HDM and treated with MCA for 24 h produce -35-fold more P-450d than do cells which are not exposed to MCA (see Table I for quantitation). In contrast, cells cultured in the presence of RPM1 plus serum showed only an G-fold induction of P-450d by MCA (see Fig. 1). This represented a 4-fold reduction in the response to MCA. Surprisingly, P-450~ mRNA, which is inducible in many cell types, showed essentially no difference (1.2-fold) in its level of induction in the presence or absence of serum. The difference in P-450~ and P-450d response to the growth medium are no doubt related to their different mechanisms of induction by MCA (8)  The data were obtained by computerized densitometric scanning of Northern blots as previously described (32). The data represent the average of three replicates which agreed within 5%. Similar data were obtained in at least three indeDendent exaeriments.
Relative" level of induction  To distinguish between these alternatives, a mixing experiment was performed. FBS was added to complete HDM to a final concentration of 10%. Cultures were maintained in the mixed medium for 24 h, and then MCA induction of P-450 expression was assayed by Northern blot analysis. As seen in Fig. 1, the levels of induction of P-450~ and P-450d mRNAs in the serum-supplemented HDM were indistinguishable from the levels seen in HDM alone. The simplest interpretation of these data are that no inhibitor is present in serum which prevents induction of P-450d mRNA, but rather a component of HDM is permissive for induction of this mRNA. However, a synergistic effect of serum plus HDM resulting in an observed recovery of P-450d inducibility by a mechanism which is independent of the effects seen in cells grown in HDM alone cannot be ruled out. Only Dexamethasone Is Required for High Levels of Accumulation of P-450d mRNA Following MCA Treatment-Since HDM is a completely defined cell culture medium, it is possible to delete individual components to determine which factors are necessary for MCA inducibility. In separate experiments, each of the 10 hormones and trace elements present in HDM was singly deleted, and then the inducibility of P-450d mRNA by MCA was compared to the inducibility seen in complete HDM. As demonstrated in Fig. 2 and summarized in Table I, only the deletion of dexamethasone resulted in a significant reduction of P-450d mRNA inducibility. Its omission from the medium resulted in -4-fold loss of induction of P-450d mRNA and no effect on P-450~ mRNA induction (1.5fold). This effect paralleled the observations of MCA induction in the RPM1 plus serum medium described above. None of the other factors had significant effects, when deleted singly, on the induction of either P-450~ or P-450d (Table I).
It is worth noting that the absolute levels of expression of o(!antitrypsin, the loading control, and the basal expression of P-450d was affected by the particular composition of the growth medium (for example, see  Table I, complete inducibility was restored by the addition of dexamethasone to cultures grown in RPM1 plus serum. Thus by the criterion that the deletion of dexamethasone from complete HDM results in a loss of inducibility of P-450d, and by the criterion that the single addition of dexamethasone to minimal salts plus serum restores inducibility of P-450d, it appears that dexamethasone is a key element in the observed MCA induction of P-450d mRNA in cultured rat hepatocytes. Interactions between multiple components in either defined medium or between dexamethasone and serum cannot be rigorously excluded as contributing to the observed phenotypes. Substratum Has Little Effect on Induction of P-450 by MCA-It has been clearly shown that substratum can have profound effects on the levels of expression of some P-450 genes, as well as on the expression of certain liver-specific functions (5, 10,26). In the experiments described above, cells were cultured on type I collagen gels applied to plastic dishes. Experiments were performed as described in the legend to Fig. 1. The specific growth conditions are indicated over each panel. The presence (+) or absence (-) of the type I collagen substratum is indicated over each panel. al-AT, cul-antitrypsin.
We therefore tested the effects of cell growth on plastic or collagen plates with HDM. As can be seen in Fig. 3 and Table  1, MCA induction of P-450d and P-450~ mRNA was unaffected by the presence or absence of substratum.
High levels of MCA induction were seen in the presence or absence of collagen gels.

Regulation of P-450d mRNA by Dexamethasone Is at the
Post-transcriptional Level-We have previously shown that MCA induction of P-450d mRNA in hepatocytes grown in complete HDM was regulated primarily at a post-transcriptional level (8). In order to determine whether or not the dexamethasone enhancement of MCA induction which we have described is a transciptionally or post-transcriptionally mediated process, we performed in vitro nuclear run-on experiments on cultures grown in the presence or absence of dexamethasone.
Nascent ["'PIUTP-labeled RNA isolated from cells grown with or without dexamethasone were hybridized under conditions of DNA excess to single-stranded cDNA corresponding to either PI-450 or P:,-450 sense and antisense sequences as well as to actin, cu,-antitrypsin, and myosin heavy chain control DNAs. Under the stringent hybridization and washing conditions used, there is no detectable cross-hybridization of P-450~ RNA and P:(-450 probe DNA (8). As can be seen in Fig. 4 and Table II In vitro run-on experiments were performed on cells grown in the presence or absence of dexamethasone and treated with MCA for the times indicated. Labeled RNA (2 x 10' cpm, 0 h; 8 X 10' cpm, 12 and 24 h) was hybridized under conditions of DNA excess to either double-stranded cDNA inserts corresponding to y-actin (A), N,antitrypsin (L), and skeletal muscle myosin heavy chain (MHC) or to single-stranded cDNA in Ml3 vector corresponding to sense (-) or antisense (+) strands of PI-450 or PI,-450 gene-specific clones.

Single-stranded
Ml3 DNA was also included as an additional negative control.
Filters were washed stringently, treated with ribonuclease A, and subjected to autoradiography as described previously (8). Densitometric scans were performed on exposures in the linear range of the film sensitivity. asone on the level of transcription of either P-450~ or P-450d genes at any time following MCA induction.
Comparison of the level of transcription of the genes plus and minus dexamethasone at 0, 12, or 24 h shows that there were no significant transcriptional effects. Although dexamethasone increases the accumulation of P-450d in response to MCA, in the data shown, there may be a slight (-2-fold) decrease in the transcription of P-450d in response to dexamethasone. However, we do not consider this level of change to be significant over the variance from experiment to experiment. As we previously reported, there is a 5-fold MCA induction of P-450~ transcription and -2-fold MCA induction of P-450d in this culture system. It should be noted that levels of transcription of actin decrease over the time period of the experiment and is typical of freshly plated hepatocytes which undergo limited growth upon initial plating.
It is also significant that hybridization of nascent RNA is seen only to the antisense strand of the probe and not to the sense strand. This result eliminates the possibility that the hybridization signal observed at any time point might be obscured by hybridization from a cryptic transcription unit in the opposite orientation.

DISCUSSION
The work presented in this paper clearly demonstrates that a hormonal component of the medium, dexamethasone, is the major factor required for maximal MCA-induced accumulation of P-450d mRNA in cultured primary rat hepatocytes. The effects of other media components and extracellular matrix components were shown to be minimal. Addition of dexamethasone to cells cultured in only minimal salts and FBS, conditions which are relatively nonpermissive for the induction of P-450d mRNA, resulted in essentially complete restoration of the maximally inducible phenotype. This demonstrates that dexamethasone is most likely a positive acting factor in the induction process and that there are probably no negative acting factors present in serum. Interestingly, the differential effects of hormones and extracellular matrix on the expression of cytochrome P-450~ mRNA are negligible. P-450~ is MCA responsive under essentially all conditions tested. This implies that P-450~ regulation is affected by a fundamentally different mechanism than P-450d in cultured hepatocytes.
In vitro nuclear run-on experiments were also performed on cells grown in the presence and absence of dexamethasone. These clearly demonstrated that removal of dexamethasone from HDM had no significant effect on the transcription rates of either P-450 gene in the presence or absence of MCA. We have previously performed nuclear run-on experiments on cells grown in HDM on collagen substratum or on nuclei derived from rat liver tissue, and found that MCA treatment of cells had little effect on the rate of transcription of the P-450d gene. This demonstrated that MCA induction of P-450d is mediated primarily at the post-transcriptional level both in rat liver and in rat liver cell culture (8). Taken with the results presented above, it appears that dexamethasone must act as a potentiator of a post-transcriptional regulatory step, since removal of dexamethasone from the defined medium results in a reduction of MCA-induced P-450d mRNA accumulation. 1.

5.
Work by others has demonstrated that dexamethasone acts on a variety of genes at the transcriptional level by a receptormediated event that represents perhaps the best studied mammalian transcriptional regulatory process (27). It has also been shown that several genes appear to undergo post-transcriptional regulation in response to dexamethasone (28,29). Recent experiments indicate that in at least the case of one of these genes, al-acid glycoprotein, both a transcriptional and post-transcriptional step may be modulated by dexamethasone (28, 30). The P-450d mRNA accumulation is induced by MCA in the presence of dexamethasone, but is not induced by dexamethasone.
Thus dexamethasone is acting as a cofactor in this response and is not a primary mediator of the accumulation.
Perhaps, dexamethasone induces a factor which interacts with P-450d mRNA and can result in higher levels of mRNA accumulation.
Dexamethasone has been shown to post-transcriptionally induce at least one other form of cytochrome P-450, P-450b (31). Treatment of rats with the glucocorticoid results in an overall IO-12-fold accumulation of P-450b in the absence of any other treatment. In the case of P-450d, however, dexamethasone treatment alone fails to induce this mRNA. Rather, dexamethasone appears to enhance the effect of MCA on induction of P-450d. Whether the dexamethasone effects seen on these two different genes are related to each other awaits further characterization of the underlying mechanisms of their induction.

15.
It is not apparent why other investigators have previously failed to obtain MCA-induced expression of P-45Od in primary hepatocyte cultures (1, 3-6). We were unable to define cellular conditions in which P-450d mRNA was not induced by MCA. We observed less than a 5-fold difference between the conditions which produced the lowest levels of induction and those which produced the highest. Furthermore, the maximum levels of induction were similar to the levels of induction seen in rat liver, when data were normalized to total mRNA loaded. It is possible that a technical detail involving the perfusion technique, or cell handling after plating is involved. However, Pasco et al. (9) have reported similar success in obtaining high levels of MCA-induced expression of P-450d in cultured hepatocytes. In any case, the levels of expression of the P-450d mRNA are sufficient to permit a much more detailed analysis of the molecular basis of induction. Further studies are in progress to determine how MCA can selectively and rapidly induce accumulation of P-450d.