Phenylalanine Hydroxylase in Cultured Hepatocytes

The presence of phenylalanine hydroxylase as a constitutive enzyme in two clonal cell lines (H4-II-E-C3 and MH&) derived from rat hepatomas and cultured in vitro is described; eight other clonal cell strains originating from rat or rabbit liver did not contain detectable amounts of this enzyme. The levels of phenylalanine hydroxylase in the two hepatoma cell lines were stimulated by hydrocortisone, corticosterone, dexamethasone or N6-02’-dibutyryl-3’,5’cyclic adenosine monophosphate. The stimulation of the levels of the enzyme by hydrocortisone in the H4-II-E-C3 line was shown to depend on de novo protein synthesis as it was inhibited by cycloheximide. It is postulated that the hepatoma cell lines are probably good models for studying the factors regulating the expression of the phenylalanine hpdroxylase gene.

and MH&) derived from rat hepatomas and cultured in vitro is described; eight other clonal cell strains originating from rat or rabbit liver did not contain detectable amounts of this enzyme.
The stimulation of the levels of the enzyme by hydrocortisone in the H4-II-E-C3 line was shown to depend on de novo protein synthesis as it was inhibited by cycloheximide. It is postulated that the hepatoma cell lines are probably good models for studying the factors regulating the expression of the phenylalanine hpdroxylase gene.
The genetic and epigenetic factors governing the regulation of phenylalanine hydroxylase activity in the normal differentiated mammalian hepatocyte have not yet been defined; nor have the molecular mechanisms responsible for the apparent absence of this enzyme in the liver of "classical" phenylketonuric phenotype been determined. This is the first report on the presence of phenylalanine hydrosylase in two clonal cell lines of hepatic origin, out of 10 cell strains examined, and on the effect of corticosteroids on the levels of this enzyme in the cultured cells. METHOTIS ANY MATERIALS Cell Culture-The H4-II-E-C3, HTC, and RLC cells used in these studies were generously supplied by Drs. Van R. Potter (I), E. Brad Thompson (a), and Lazaro E. Gerschenson (3); respectively.
The basal medium used for the H4-II-E-C3 cells consisted of the same modification of Swim's medium except that it contained 585 mg of glutamine per liter ((S-77, 2G) medium).
The H4 cells were grown at first on this basal medium supplemented with 5% fetal bovine and 20 y0 horse serum (5F, 20H (2.77,2G) medium); but in order to avoid these inconveniently high serum levels, several different basal media and a variety of serum combinations were tested in order to define the conditions giving the highest terminal cell densities and the greatest proportion of healthy cells. Since the best growth media were the 5F,5C(S-77,2G) and 15F-(S-77,2G), strains of cells were initiated on these. Initial examination of the H4 cells for basal enzyme activity and steroid sensitivity (see below) after a single culture passage on either of these media yielded results that were qualitatively and quantitatively similar to those obtained with the cells grown continuously on 5F, 20H(S-77,ZG) medium.
During these studies the parameters examined were stable and reproducible from experiment to experiment.
The RL and MHlCl cells were grown on 15F(S-77, 2G) and 5F, 20H(S-77,2G) media, respectively; whereas the C-15 cells were maintained on Parker's medium 199 (8) supplemented with 5% each of fetal bovine and calf serum. Penicillin, streptomycin, and polymyxin B sulfate were included in all culture media at levels of 100,000 units, 100 mg, and 50,000 units per liter, respectively.
Fresh batches of growth media were prepared every 2 to 3 weeks and were sterilized by passage through an Ertel bacterial filter (Ertel Engineering Corporation, 8-14 N. Front St., Kingston, N.Y. 12401) after the asbestos pad was washed with 1 liter of citrate-saline solution (NaCl, sodium citrate, and phenol red; 6.0,3.0, and 0.01 g per liter, respectively). 1 The meaning of the letters and numerals in front of the code for t.he modified Swim's medium is the following: the letters F, C, and H denote fetal bovine, calf, and horse sera, respectively, and the numerals in front of these letters show the percentage of these sera (v/v) contained in the media.
The code for the H4-II-E-C3 hepatoma cells will be abbreviated to H4.
All cells were grown in surface culture in Falcon T-30 or T-250 flasks, the medium being changed thrice weekly.
Samples of all cell lines were frozen and stored in a liquid-nitrogen freezer in their respective growth media supplemented with 10% (v/v) dimethylsulfoxide.
Subcultivations were carried out as follows.
After two washings with citrate-saline solution followed by a single rapid washing with the appropriate trypsin solution, the cells of all cell lines, except H4, were detached from the growth surface of the flasks by a 5-min incubation at 37" with a 0.1% (w/v) solution of trypsin in citrate-saline solution.
The cell suspension was then layered on top of 10 ml of growth medium and the cells, sedimented by centrifuging through the medium, were redispersed in a convenient volume of fresh medium and inoculated into new culture vessels. The H4 cells were treated with a 0.025% trypsin solution for 5 min at 37". 811 cell lines were tested, periodically, for bacterial or fungal contamination by inoculation of spent culture media into thioglycolate broth.
In addition, the HTC, RLC, H4, and MHrCr cells were assayed for the presence of mycoplasma with NABI (North American Biologicals, Inc., 13323 S. Normandie Ave., Cardena, Calif. 90249), or GIBCO (Grand Island Biologicals Co., 2323 Fifth St., Berkeley, Calif. 94710) mycoplasma broth and agar plates. All tests for microorganisms were negative? In order to reduce the likelihood of inadvertent contamination with mycoplasma through normal handling procedures, stock cultures of HTC, RLC, H4, and MHrCl cells were routinely maintained and passaged continuously in T-30 flasks by disposable plastic syringes without pipetting by mouth.
Unless otherwise indicated, H4 cells were plated and cultivated for experiments as follows. Confluent populations of cells, grown in T-250 flasks, were subcultured for the experiments at a 1: 10 or 1: 15 dilution and seeded into fresh flasks each containing 20 ml of growth medium; a fresh change of medium was given after 3 to 4 days; experimental medium was added on the following day. All experimental groups were harvested simultaneously 16 to 20 hours later.
Hormones, antimetabolites, and other components8 of experimental media were dissolved in the same batch of medium as was to be added to control cultures.
Sterile st,ock solutions of radioactive amino acids in distilled water 2 Electron microscopic examination of the cells revealed, however, that these cell lines, with t,he exception of the MHlCl line, contained C-type virions (12) which seem structurally to be identical wit,h the "microbodies" described recently by Gerschenson et al. (13) in cultured RLC cells. We are unable to tell, in all instances, whether this viral contamination of the cells occurred before or after the receipt of the prototype cultures in our laborat,ory. The viral infection is not relat.ed to the presence or absence of phenylalanine hydroxylase in the cultured cells: one of the two cell lines found to contain phenylalanine hydroxylase was infected with the virus particles (t,he H4-cells), and the other (MHIC1) was not.
Preparation of Extracts of H&II-E-C3 Cells and Rat Liver for Phenylalanine Hydroxylase Assay-Each experimental group consisted of 10 T-250 flasks of cells grown to confluency.
The experimental media were removed by aspiration, and the cell sheets in each flask were washed twice with 10 ml of citratesaline solution.
Three milliliters of cold Ca++-and Mg+f-free phosphate-buffered saline (NaCl, KCl, Na2HPOd. 7Hz0, and KHzPO,; 8.0, 0.2, 2.17, and 0.2 g per liter) were then added to each flask and the cells were detached mechanically by passing a rotating 2-cm long cylindrical Teflon-coated magnetic stirring bar back and forth over the vessel growth surface. The suspended cells from each experimental group were combined and transferred to a plastic conical centrifuge tube with two addi-tionaI IO-ml portions of phosphate-buffered saline, and the cells were sedimented by centrifuging for 5 min at 300 x g at room temperature.
All further steps were carried out at 4". After removal of the supernatant solution by aspiration, the cell pellet was dispersed in 5 ml of Tris-KC1 buffer (0.15 M KCl; 0.01 M Tris-HCl, pH 7.5; and 0.050/, (v/v) mercaptoethanol) and was transferred to a tared graduated conical plastic centrifuge tube with another 5 ml of buffer. The suspension was then centrifuged for 10 min at 270 X g; the supernatant solution was removed as completely as possible; and the cell pellet was weighed. Confluent cultures from 10 T-250 flasks usually gave a cell pellet of 500 to 600 mg. The cells were next rcdispersed in Tris-KC1 buffer, and the volume of the resulting suspension was adjusted to 4.0 ml. At this point, a drop of cell suspension was examined for plasma membrane integrity by means of trypan blue dye exclusion: usually fewer than 10% of the cells became stained. The cells were broken by treatment wit,h a 15-s burst of 20 kc s-1 ultrasound at a peak-to-peak amplitude of 8 pm in an MSE 100 watt ultrasonic disintegrator (Measuring & Scientific Equipment Ltd., London, England) at 0". All cells were disrupted as verified by phase-contrast microscopy.
The suspension of disrupted cells was next centrifuged at 100,000 X g for 60 min. The portion of the resulting particle-free supernatant solution lying below the superficial lipid layer was removed with a Pasteur pipette.
These particle-free protein solutions from all experimental groups were assayed for phenylalanine hydroxylase activity within 2 to 3 hours of this step; their protein content was measured by the method of Lowry et al. (9) with bovine serum albumin as a standard.
In order to gain evidence that the recoveries of phenylalanine hydroxylase enzyme by these procedures were complete, in several experiments treatment with trypsin (0.025% (w/v), 37", 5 min) and Triton X-100 (0.1 y0 (v/v), 4", 15 min) were chosen as alternatives to mechanical harvesting and cell disruption by ultrasound. 4 Since these methods resulted in no significant difference in the enzyme yield, we concluded that the mechanical harvesting and ultrasonic disruption of cells gave a satisfactory estimate of total cellular phenylalanine hydroxylase content; all data reported were obtained on preparations made by the mechanical processes.
In all experiments in which the phenylalanine hydroxylase levels of cultured cells were compared with those of extracts of adult rat liver, the animals were killed by cervical dislocation followed by decapitation; 1.0 g of liver was immediately excised and homogenized in 3.5 ml of Tris-KC1 buffer with a motordriven Potter-Elvehjem type smooth walled homogenizer having a tolerance of 0.10 to 0.15 mm. Homogenization routinely consisted of 8 complete up-and-down strokes of 5 to 7 s duration each and was carried out with constant cooling in an ice bath. The homogenate was then transferred to a graduated conical plastic centrifuge tube, and its volume was adjusted to 4.0 ml. For all subsequent steps, including ultrasonic disruption, the rat liver homogenate was treated as the suspensions of cultured cells.
In the assays (final volume 1.0 ml), the sample cuvette contained 0.1 ml of 1 M Tris-HCl buffer, pH 7.4; 0.05 ml of a 20 mM phenylalanine solution; the volume of the protein solution to be assayed; and sufficient distilled water to make 0.90 ml. The reference cuvette contained all of these components except phenylalanine.
After a 5-to lo-min equilibration at 27", the reaction was started by the simultaneous addition of 0.1 ml of a 1.7 mM solution of 6,7-dimethyl-2-amino-4-hydroxytetrahydropteridine hydrochloride to the sample and reference cuvette. Light absorbance was measured in a Unican model SP-1800 double-beam recording spectrophotometer.
The stoichiometry of the phenylalanine hydroxylase reaction under the conditions of this assay has been demonstrated with extracts of cultured cells as well as with preparations from various animal tissues to be 1 mole of tyrosine formed per mole of enzymatically oxidized tetrahydropteridine (cf . Table III). For determination of tyrosine the spectrophotometric reactions were terminated by the rapid addition of 0.25 ml of 30% trichloroacetic acid solution to the cuvettes. Tyrosine was then determined in the deproteinized solutions spectrofluorometrically with a Farrand Spectrofluorometer (Mark I; Farrand Optical Co., Inc., New York) by the methods of Udenfriend and Cooper (lla) and Waalkes and Cdenfriend (llb).
Standard curves for the tyrosine determinations were prepared by the inclusion of 2 to 30 nmoles of tyrosine, in addition to the usual amount of phenylalanine and pteridine cofactor, in incubation mixtures in which the enzyme was first killed by the addition of trichloroacetic acid. One enzyme unit is defined as the phenylalanine-dependent oxidation of I nmole of tetrahydropteridine to the dihydro form, or the formation of 1 nmole of tyrosine per min. Under the conditions of the spectrophotometric assay and with the amounts of protein used, substitution of either tyrosine or tryptophan (1 mM) in place of phenylalanine produced no detectable reaction either in t,he extracts of cultured cells or of liver, except in one cell extract which gave a very slight reaction with tyrosine.
Except when high initial velocities were recorded, the phenylalanine hydroxylase reaction rates of cell extracts were linear with respect to time for several minutes.
For rat liver and cell extracts, the initial reaction rate was linear with respect to the amount of extract assayed only up to 0.6 to 0.7 mg of protein ( Fig. 1). The extracts were routinely assayed at several concentrations of protein and all calculations of phenylalanine hydroxylase-specific activities and total enzyme levels were based on only those initial reaction rates that were proportional to the amount of protein assayed.
E$ect of Cycloheximide on the Incorporation of [G-sH]Phenylalanine into Cellular Proteins-All experimental media contain- Initial reaction rates of phenylalanine hydroxylase in extracts of H4-II-E-C3 cells and adult rat liver as a function of the amount of protein assayed. Thirty replicate flasks of H4-II-C3 cells were cultivated to confluency in 5F,20H(S-77, 2G) medium.
Eighteen hours before harvesting, one half of the cultures were given fresh 5F,20H(S-77,2G) medium alone, while the other half were given the same medium supplemented with 10-b M hydrocortisone sodium succinate. The cells from the 15 flasks in each experimental group were harvested and pooled. Extra&s from the cultured cells and from 1.5 g of rat liver were prepared as described in the text in 6.0 ml of Tris-KC1 buffer. A, 0.06 ml (0.28 mg of protein) to 0.75 ml (3.5 mg of protein) of extract from control cells ( l ) and 0.06 ml (0.34 mg of protein) to 0.75 ml (4.3 mg of protein) of extract from hydroeortisone-treated cells (0) were assayed. B, 0.01 ml (0.29 mg of protein) to 0.32 ml (9.3 mg of protein) of rat liver extract were assayed in duplicate; individual values are shown.
ing [G-3H]phenylalanine (specific activity, 406 Ci per mole), with or without cycloheximide at varying concentrations, were warmed to 37" before addition to cell cultures.
One-half hour before the beginning of the experiment, the growth medium was removed from replicate confluent cultures of H4 cells in T-250 flasks, and the cell layers were washed once with 10 ml per flask of serumless S-77,2G medium.
The cells were then incubated at 37' in 10 ml of fresh serumless medium for another 15 min. Upon removal of this second wash, the experimental medium was added. After a 60-min incubation at 37", the experimental medium was aspirated from each flask, and the cell sheet was washed twice with 10 ml of cold citrate-NaCl solution. Three milliliters of cold phosphate-buffered NaCl solution were then added, and the cells were detached from the substratum mechanically with a magnetic stirring bar as described earlier.
The cell suspension from each T-250 flask was transferred to a graduated glass conical centrifuge tube with two 3-ml portions of phosphate-buffered NaCl solution and the cells were centrifuged for 5 min at 270 X g. After resuspension of the resulting cell pellet in 2.0 ml of distilled water, the cells were treated with ultrasound for 15 s, and 2.0 ml of a 12% (w/v) aqueous trichloroacetic acid solution were added. The precipitate was sedimented by centrifuging for 10 min at 1500 X g and was washed twice with 3.0 ml of cold By0 trichloroacetic acid solution. The last drops of the final supernatant solution were removed with a cotton swab. The washed trichloroacetic acid-insoluble material was resuspended in 1.0 ml of a 2 N NaOH solution and was left overnight at room temperature.
The sample was then heated at 60" until a clear solution was obtained; aliquots were removed for radioactive counting in a Packard Tri-Carb scintillation spectrometer, model 3320, and determination of protein.

Comparison
between Phenylalanine Hydroxylase Activity of Conjluent H4-II-E-C3 Cells after Continuous Cultivation on 5F, 6C(S-77,2G) Medium and that of Adult Rat Liver-The activity of phenylalanine hydroxylase was calculated from measurements of initial reaction velocities with the high speed supernatant fractions of either disrupted cells or rat liver homogenates.
Since the activities were normalized per mg of soluble protein in the ,extracts, it was recoguized that any factor that either altered the initial enzymatic rates of extracts or changed their soluble protein contents would necessarily affect the calculated values for specific activity.
Because t.he cultured cells were routinely disrupted by sonication, the effect of ultrasound on the yield of phenylalanine hydroxylase and sohlble protein from rat liver homogenates was first determined.
Sonica,tion, applied after homogenizing the liver, neither increased nor decreased the total enzyme units in the soluble supernatant, but it consistently increased the concentration of the soluble protein in the extract.
Therefore, in all experiments in which the phenylalanine hydroxylase activities of H4 cell and rat liver extracts were compared, the rat liver homogenates were sonicated in parallel along with the cell suspensions. Total enzyme yields per g wet weight of packed cells or tissue were calculat,cd from the measured values for the initial reaction rates, soluble protein concentrations, and sample wet weights. Table I gives the mean values for the phenylalanine hydrolase specific activities and total enzyme contents of confluent H4 eelIs cultured on 5F, 5C(S-77,2G) medium and adult rat liver, calculated from the data pooled from several experiments.5 Whereas the specific activity of the cell extracts was 5 The values of pheaylalsnine hydroxylase act,ivities measured by us are not maxima because the concentration of phenylalanine in the assays was only 1 mM and that of the 6,7-dimethyltetrahydropt,erin was 0.17 mM. The K,,, value for phenylalanine with the rat liver enzyme, when assayed with 6,7-dimethyltetrahy-  I Mean values for specific activity of phenylalanine hydroxylase and total enzyme content of basal conjluent H.&II-E-B cells and adult rat liver Extracts from confluent, slowly replicating cell cultures grown continuously on 5F,5C(S-77,2G) medium and from livers of 150-g male rats were prepared and assayed for phenylalanine hydroxylase activity as described under "Methods and Materials." The mean values and their standard deviations were calculated from the data of eight and nine separate experiments for the cultured cells and for rat liver, respectively. usually about one-third of that of the rat liver extracts, the total enzyme content per g wet weight of the cells was only about one-tenth of that of liver. This discrepancy results from the fact that the soluble (and total) protein content of rat liver is much greater than that of the cells. Effect of Adrenocorticosteroid Hormones on Phenylalanine Hydroxylase Activity of H4-II-E-C3 Cells-When H4 cells were cultivated continuously on 5F, 5C(S-77,2G) medium and confluent, slowly replicating populations were placed on either of these media supplemented with 10e5 M hydrocortisone sodium succinate for 16 to 20 hours, a 2-to 2.5-fold increase in total phenylalanine hydroxylase activity over basal enzyme 1eveIs was obtained.
Moreover, if confluent cells, cultured continuously on 5F, 5C(S-77,2G) medium, were washed free of serum-containing medium and incubated for 16 to 20 hours on serumless S-77,2G medium supplemented with lo-" M hydrocortisone, the hormone evoked a 4-to 5-fold stimulation of cellular phenylalanine hydroxylase activity over basal enzyme levels in the absence of serum (Table II, see also Fig. 1). This apparent enhancement of the hydrocortisone effect was caused by decreased basal enzyme activities and not by increased cellular phenylalanine hydrolase levels in the presence of the hormone since hormonestimulated enzyme activities under these conditions were equal to or even somewhat lower than the phenylalanine hydroxylase levels evoked by the steroid in serum-containing medium.
Cultivation of the cells on 5F,5C(S-77,2G) medium supplemented with lop5 M sodium succinate under the same conditions produced no stimulation of cellular enzyme activity, nor did the addition of hydrocortisone to the enzyme assay mixture in vitro increase the initial reaction rates of extracts from basal cells.
In order to ascertain that the increased reaction rates seen in extracts of hydroeortisone-treated cells resulted from a true increase of enzymic activity and were not artifacts, the amount of tyrosinc formed was determined at the termination of the spectrophotometric assay as described under "Methods and Materials." The data of Table III demonstrate the satisfactory  stoichiometric agreement between the phenylalanine-dependent oxidation of the tetrahydropteridine and the amount of tyrosine formed not only in the rat liver extract, but also in extracts of dropterin, has been reported to range from 0.76 mM to 1.25 rnM (10; 14, Ii): the K, value for the 6,?'-dimethyltetrahydropterin, measured in our laboratorv (10).   the presence and the absence of serum. As is shown in Fig. 3, the cells responded faster to hydrocortisone in the absence of serum than in its presence.
Dependence of Hydrocortisone Effect on Protein Synthesis--The effects of varying concentrations of cycloheximide on genera1 protein synthesis in H4 cells are illustrated in Fig. 4. In this experiment, the incorporation of [3H]phenylalanine into total trichloroaeetic acid-precipitable material was measured over a l-hour period at 37" in the absence of serum. Although this brief exposure to the drug at a concentration of 10P4 M had no effect on subsequent cell viability, a 1%hour incuba.tion with only 10-j M cycloheximidc, even in the presence of serum, resulted in pronounced cytotoxic effects. Therefore, since no such cytotoxicity was apparent after as much as a 17-hour exposure to 1Om6 M cycloheximide-a level of the drug that had been found to inhibit general cellular protein synthesis by 70%-this concentration of the anti-metabolite was chosen for the experiment shown in Fig. 5 cells as function of time of exoosure to hvdrocortisone in presence and absence of serum.-Curve A, SO replicate flasks of cells, plated at the same time in 5F,5C(S-77, 2G) medium, were grown continuously to confluency on that same medium. Two days before harvesting, all SO flasks were given fresh medium.
Then, at the time interval before harvesting indicated in the figure, the medium in each of the six decades of experimental flasks was replaced by 5F,5C(S-77.2G) medium supplemented with lO+ M hydrocortisone.
The 80 flasks were harvested in two separate heats, and cell extracts were assayed for phenylalanine hydroxylase. Each heat contained a zero time group of control flasks that were given medium without hydrocortisone.
The total enzyme cont.ent per g wet weight of packed cells for each experimental group is expressed as a percentage of the value obtained for the cells cultured in the absence of hormone.
Curve B, 70 replicate flasks of cells were plated and grown as described above. Sixteen hours before harvesting, the cells in the 60 experimental flasks were washed once with 10 ml per flask of serumless S-77,2G medium, while the ce!ls in the 10 control flasks were washed twice in the same manner. At in H4-II-E-C3 cells. The cultures, grown to confluency on 5F,5C(S-77,2G) medium, were rinsed with S-77,2G medium and were then incubated for 1 hour at 37" with 15 ml of serumless S-77,2G medium containing 11 X 106 dpm of [3H]phenylalanine (12.1 nmoles + 1.5 rmoles of unlabeled amino acid) and varying concentrations of cvcloheximide.
Two flasks containing cells killed at 100" were "the controls.
The specific activity-of the proteins was determined as described under "Experimental Procedure." The mean protein content of the 16 experimental cultures was 4.15 f 0.36 (s.D.) mg. The data shown in closed circles were obtained on two cultures that had been exposed to 10-G M hydrocortisone continuously for 24 hours in the absence of cycloheximide.
of 10-C M cycloheximide over a 17-to M-hour period in serumless medium blocked the hydrocortisone effect by 60% to 65'%, but did not significantly decrease basal cellular enzyme levels. Furthermore, a 24-hour preincubation with low5 M hydrocortisone had only a slight effect on general protein synthesis in the cells as judged by an increase of approximately 10 y0 in the incorporation of t3H]phenylalanine into cellular proteins. It is, therefore, concluded (a) that hydrocortisone stimuiates phenylalanine hydroxylase activity in H4 cells by a mechanism dependent upon de TWVO protein synthesis and (b) that the hydrocortisone effect on cellular protein synthesis is specific to a relatively small number of cellular proteins, one of which may well be the phenylalanine hydroxylase enzyme itself. E$ect of Steroid Horwwrm other than Hydrocortisom and Dibutyryl Cyclic Adenosine 3',6'-Monophosphate on Phenylalanine Hydroxyluse Activity of H.&II-E-C3 Cells--As is indicated in Table IV, a 16-to 19-hour exposure to either dexamethasone sodium phosphate or corticosterone at a concentration of lo+ M in the presence of serum produced an elevation of cellular phenylalanine hydroxylase activity essentially equivalent to a maximum stimulation by hydrocortisone under the same conditions. In addition, preliminary experiments have also indicated that cellular phenylalanine hydroxylase levels are altered by the presence of cyclic nucleotides.
A 16-to 19-hour cultivation on serum-containing medium in the presence of 1OP M dibutyry1 cyclic adenosine 3',5'-monophosphate, but in the absence of this t,ime, after a second washing, S-77,2G medium supplemented with 10-L M hydrocort,isone was added to the 10 flasks of the 16-methyl xanthines, stimulated phenylalanine hydroxylase activity hour group, but S-77,2G medium alone was given to the 60 re-1.5-to 2-fold over basal levels. A more thorough examination maining flasks. Then, at the time interval before harvesting of the influence of cyclic nucleotides on the regulation of this enindicated in the figure, the medium in each of the remaining five zyme in H4 cells is being carried out. decades of experimental flasks was replaced by S-77,2G medium Examination of Additional Cell Lines of Hepatic Origin for supplemented with KY5 M hydrocortisone. The zero time group of control flasks was harvested without any further medium Phenylalanine Hydroxylase Activity-Several cell lines of hepatic change. The cultures were harvested and the cell extracts were origin were examined for the presence of phenylalanine hydroxylprepared and assayed for phenylalanine hydroxylase activity. ase activity under conditions essentially identical with those The data are plotted as for Curve A.
described for the H4 cells. In every instance, extracts from the 400r HC  IV   2  RADIOACTIVITY  3  -1 FIG. 5. The effects of cycloheximide on phenylalanine hydroxylase levels and on protein synthesis in presence and absence of hydrocortisone.
Forty cultures were washed twice with 8 ml of serumless S-77,2G medium and were given 20 ml of S-77,2G medium alone or the same medium supplemented with 10m5 M hydrocortisone, 10-S M cycloheximide, or lwK M hydrocortisone plus 10-e M cycloheximide as indicated in the figure. The cells were harvested, and the cell extracts were assayed for phenylalanine hydroxylase.
The total enzyme content per g wet weight of oacked cells is expressed as a percentage of the value (120 uniis per g) obtained for the cells incubated for 18 hours on serrumless medium in the absence of hydrocortisone and cycloheximide.
The eight remaining flasks were washed with the serumless medium as described above and a pair of flasks was incubated under each of the experimental conditions except that the media also contained 22 X lo6 dpm of [G-aH]phenylalanine. Two flasks containing cells killed at 100" were the controls. After 18 hours of incubation the incorporation of 3H into proteins was measured and is expressed (black columns) as a percentage of the specific activity of the proteins found in the pair of flasks in the absence of hormone and cycloheximide (926,400 dpm per mg). Comnonents of incubations in addition to the basal S-77,2G i&l[irn are indicated above the columns: 5F,6C, 5yo fetal 'bovine and calf sera; CX, cycloheximide; HC, hydrocortisone.
proband cells were prepared and compared with extracts of H4 cells and/or of rat liver. With the single exception of the C-15 cells and the four epithelioid subclones derived from them, all cell lines were assayed both under basal conditions and after a 16-to 20-hour exposure to lop5 M hydrocortisone in the presence of serum.
The following cell lines showed no detectable phenylalanine hydroxylase activities when the assays were made on 1.2 to 9.0 mg of protein: the parental clone and four subclones of C-15 cells, derived from normal rabbit liver; HTC cells, originating from a Morris rat hepatoma; and the RLC and RL cells, both derived from normal rat liver. The only cell type in which phenylalanine hydroxylase activity was detected in addition to the H4 cells was the MHlCl line. The phenylalanine hydroxylase of these cells had the following characteristics in common with the enzyme from H4 cells: (a) proportionality between initial reaction rate and amount of cell extract present in the assay was maintained only for the lowest soluble protein concentrations examined (up to 0.7 mg per assay); (b) an 1% to 24.hour exposure to serum-containing medium supplemented with low5 M hydrocor- Forty replicate flasks of cells were plated in 5F,20H(S-77,2G) (Experiment 1) or 5F,5C(S-77,2G) (Experiment 2) medium and grown to confluency.
Sixteen (Experiment 1) or 19 (Experiment 2) hours before harvesting, the medium in each experimental group of 10 flasks was replaced by either the respective growth medium alone or the same medium supplemented with lO+ M steroid hormone or lo-& M dibutyryl cyclic adenosine 3', 5'-monophosphate (cyclic AMP).
The cultures were then harvested, and the cell extracts were assayed for phenylalanine hydroxylase.  tisone evoked a 2-fold increase in total cellular enzyme content; and (c) low5 M dexamethasone produced a stimulation of phenylalanine hydroxylase activity comparable to that of hydrocortisone. DISCUSSION In an evaluation of the validity of the use of these cultured hepatoma cells as an experimental model for studying the regulatory mechanisms governing the expression of the phenylalanine hydroxylase gene in mammalian liver, one needs first to compare the levels of phenylalanine hydroxylase found in the cultured cells with those of the liver in viva. The phenylalanine hydroxylase activity of H4 cells per g wet weight was about 10% of that of adult rat, liver. One gram of packed H4 cells grown under our experimental conditions contains 4.5 X lo8 cells. According to a recent report 1 cm3 (or -1 g) of adult rat liver contains 1.0 x 108 parenchymal cells (16). It is calculated from these figures and from the data of Table I, and on the assumption that the enzyme in the liver is localized exclusively in the parenchymal cells, that the phenylalanine hydroxylase content of the H4 cells under basal conditions is only about 2.1 Y. of that of the liver cells (2.0 x 10F7 units versus 96 X lo-' units per cell). However, the difference between the enzyme content of the two cell types becomes much less once the calculations are related to cytoplasmic volume.
One of the main morphological characteristics of the H4 cells is their small cytoplasm as compared to the cytoplasm of the liver parenchymal cells; the size of the nuclei of the two cells, on the other hand, is very similar (Fig. 6). We have determined in IO-,um thick sections, made from pellets of H4 cells and adult rat liver fixed in Bouin's solution and stained with hematoxylin and eosin, the ratio of the volume of cytoplasm to that of the nucleus by the method of Chalkley (17). These measurements showed that the cytoplasmic volume of the H4 cells was only about one-fourth of that of the liver cells. It follows, therefore, that the phenylalanine hydroxylase content of basal stimulated enzyme activities can be augmented by a further factor of 2 when the cells are cultivated continuously in the presence of horse serum under some conditions.
The sensitivity of the cells to other hormones and humoral factors has not yet been investigated, nor have steroids and cyclic nucleotides been tested in combination in order to examine additive or possible synergistic effects. It seems, therefore, probable that, by a combination of specific cell-culture media and supplemental hormones, the levels of phenylalanine hydroxylase in the cultured cells, expressed per unit volume of cytoplasm, can be raised to values comparable to those found in the parenchymal cells of rat liver.
Current experiments are being directed at a more complete delineation of the various factors necessary to evoke a maximum expression of the phenylalanine hydroxylase gene in these cultured hepatoma cells.
Studies on whole animals and perfused organs have demonstrated that a number of enzymes characteristic of the differentiated liver are under hormonal control in Vito (18). Moreover, the activities of several of these hepatic marker enzymes are influenced specifically by alterations in t'he levels of circulating adrenocorticosteroids (19-21).
In concordance with these studies, the adaptation of differentiated hepatic tissues to t'he conditions of cell culture has recently yielded clonal cell populations that possess a number of liver-specific functions (l-3, 5, 6,22-27), many of which still retain much, if not most of their original in vivo responsiveness to hormones and other hunloral f&ors (l-3, 13, 22, 25, 28-34).
Nevertheless, in spit.e of t,he growing number of these examples, the phenylalanine hydrosylase enzyme has not been reported previously to be present in any permanent cell line of liver origin, and only little evidence has been presented for the regulation of this enzyme by hormones in the whole animal (14,35,36). Eagle et al. (37) noted, however, the conversion of phenylalanine into tyrosine in a strain of HeLa cells cultured in the absence of tyrosine, and Cartwright and Danks (38) reported phenylalanine hydroxylase in primary cult'ures of human fetal muscle fibroblasts.
Our work demonstrates, first, that phenylalanine hydroxylase is present in two established cell lines derived from transplantable rat hepatomas and, second, that the activity of t,his enzyme is modified in a similar manner in both cell types by physiological concentrations of adrenocorticosteroid hormones.
Although the ultimate confirmation that the sensitivity of phenylnlanine hydroxylase in the cultured cells to regulation by st,eroid hormones or other humoral factors is a true represent'at'ion of the in vivo situation must await further whole-animal and liver-perfusion studies, the initial observation by Freedland (35) of steroidevoked stimulation of phenylalanine hydroxylase activity in adult rats coupled with the recent report by McGee ef al. (14) describing a 2-fold enhancement of hepatic phenylalanine hydroxylase levels after hydrocortisone inject'ion during a brief period in the developing rat would suggest that' adrenocorticosteroids probably play an important role in the maintenance of normal levels of this enzyme in vivo. Moreover, by analogy with the tyrosine aminotransferase system in H4 and HTC cells, the sensitivity of which to steroids, cyclic nucleotides, and insulin does substantiate the findings from earlier studies with perfused livers (2&34), it is tempting to speculate that the regulatory mechanisms governing the expression of the phenylalanine hydroxylase gene in these cultured cells also exist in the mammalian liver in vivo.
The similarities between the behavior of the phenylalanine hydroxylase and tyrosine aminotransferase systems of the H4 cells are striking. Reel  medium: X 740. The I-14 cells h&e a much smaller cytoplasmic volume than the liver cells. The cytological appearance of H4 cells grown in the presence of 10+~ M hydrocortisone could not be distinguished in such sections from that of cells grown without the hormone.
H4 cells per unit of cytoplasmic volume is about 8.4% of that of the liver parenchymal cell. The steroid hormones, which caused no measurable change in the ratio of cytoplasmic to nuclear volumes, were shown to increase 2-to 2.5-fold the phenylalanine hydroxylase levels of the H4 cells. Thus, adrenocortical steroids can elevate cellular enzyme activities per unit volume of cytoplasm to values that are up to 21% of those of liver parenchymal cells in Z&JO. Furthermore, recent experiments, to be reported later, on the influence of the nature of the sera used in the culture media on cellular enzyme levels have shown that hydrocortisone-maximum "induction" of tyrosine aminotransferase activity by hydrocortisone after a 24.hour preincubation of the cells on a medium free of serum and hormone.
In our studies, the maximum stimulation of phenylalanine hydroxylase activity obtained with hydrocortisone was 4-to 5-fold on a serumless medium (Table II).
However, in our experiments the cultures were placed on the experimental media directly from serum-containing medium without any preincubation in the absence of serum. Under these conditions, basal enzyme activities were seen to decrease to about one-half of their former values by 16 to 20 hours. If this decay of basal phenylalanine hydroxylase activity were to have continued with the same half-life for an additional 24 hours, the degree of stimulation by hydrocortisone would have been at least 8-to IO-fold.
We found, in an experiment not reported here, that a 24.hour preincubation of the cells on serumless S-77, 2G medium in the absence of hydrocortisone did not affect the maximum enzyme levels attained upon subsequent incubation with the hormone.
Hence, the degree of stimulation of phenylalanine hydroxylase by hydrocortisone appears to be comparable to that of tyrosine aminotransferase in the H4 cells. The time course of the effect of hydrocortisone on these two enzymes is also similar, with maximum stimulation of activity occurring by 12 hours in absence of serum in both systems (Reference 31 :and Fig. 3B). However, the response of the phenylalanine hydroxylase and tyrosine aminotransferase enzymes of H4 cells to varying dosages of hydrocortisone in the absence of serum was somewhat different.
The stimulation of tyrosine aminotransferase took place over a broader range of hydrocortisone concentrations (1 X 1Oms M to 5 X 1OV M) (28) than did that of phenylalanine hydroxylase (1 X 10ms M to 1 X 1OV M) (Fig. 2B), and, in addition, Harnett and Wicks (32) observed that concentrations of dexametha.sone (and hydrocortisone) greater than lo+ M "produced variable and sometimes inhibitory effects" on the induction of the aminotransferase.
In contrast to our findings with phenylalanine hydroxylase (Table II), basal tyrosine iaminotransferase activities appear to be stable in the absence of serum (34,39), whereas, at least in HTC cells, the presence of .serum potentiates the induction of the transferase by dexamethssane (39). The effect of insulin, glucagon, or other hormones on the basal and hydrocortisone-stimulated phenylalanine hydroxyl-.ase activity of H4 cells has not yet been investigated.
In a manner analogous to the tyrosine aminotransferase systems of H4 (39), HTC (2) and RLC (3) cells, hydrocortisone stimulates the phenylalanine hydrosylase activity of H4 cells by a mechanism dependent upon protein synthesis (Fig. 5B).
In two experiments, hydrocortisone had only a slight effect, if .any, on general protein synthesis in the cells, as judged by incorporation of [3H]1)l~tnylalalline into protein, and the addition of the hormone directly to rell extracts in vitro did not enhance the phenylalanine hydroxylase activity. The simplest interpreta.tion of these observations is that hydrocortisone stimulates cellular enzyme activities by inducing the synthesis of new phenylalaninc hydroxylasc protein; however, the alternative possibility that the hormone induces the synthesis of a second enzyme which, in turn, activates phenylalanine hydroxylase cannot be ruled out.