Stimulation of Facilitated [‘HIUridine Transport by Thyroid Hormone in GHI Cells EVIDENCE FOR REGULATION BY THE THYROID HORMONE NUCLEAR RECEPTOR*

We have previously shown that 3,5,3‘-triiodo-~-thy- ronine ( L - T ~ ) stimulates cell growth and a 4- to 8-fold increase in growth hormone mRNA in GH, cells. These effects appear to be mediated by a thyroid hormone nuclear receptor with an equilibrium dissociation con-stant for L-T3 of 0.2 nM and an abundance of about 10,000 receptors per cell nucleus. In this report, we show that L-T3 exerts a pleiotypic effect on GH, cells to rapidly (within 2 h) stimulate [3H]uridine uptake to a maximal value of 2.5- to %fold after 24 h. This results from an increase in the number of functional uridine “transport sites” as shown by studies docu- menting an increase in the apparent Vmax with no change in the Km, 17 p ~ . Although the labeling of the cellular uridine pool and pools of all phosphorylated uridine derivatives was increased by L - T ~ , there was no change in the relative amounts of the individual pools in cells incubated with or without hormone. The intracellular concentration of [3H]uridine was esti- mated to be similar to that of the medium, suggesting that facilitated transport mediates [3H]uridine uptake. That this increase in r3H]uridine transport was nuclear receptor-mediated is supported by the excellent cor- respondence of the L-T3 dose-response curve for [3H] uridine The cell monolayers were then chilled to 4 "C and harvested to isolate the nuclei as previously described (2, 18). The radioactivity in the nuclear pellet was quantitated using a refrigerated Packard y spectrometer at 55% efficiency followed by DNA determination. Nonspecific binding of L-['*~I]T~ to GHI cell nuclei was estimated using 5 pM nonradioactive L-T~. This value, always less than 5% of total binding, was subtracted from results obtained using radioactive hormone alone.


Stimulation of Facilitated ['HIUridine Transport by Thyroid Hormone in GHI Cells
EVIDENCE FOR REGULATION BY THE THYROID HORMONE NUCLEAR RECEPTOR* (Received for publication, November 4, 1985) Frederick Stanley, Jir S. Tsai, and Herbert H. Samuels From the Division of Molecular Endocrinology, The Rose F. Tishman Laboratories for Geriatric Endocrinology, Department of Medicine, New York University Medical Center, New York. New York 10016 We have previously shown that 3,5,3'-triiodo-~-thyronine ( L -T~) stimulates cell growth and a 4-to 8-fold increase in growth hormone mRNA in GH, cells. These effects appear to be mediated by a thyroid hormone nuclear receptor with an equilibrium dissociation constant for L-T3 of 0.2 nM and an abundance of about 10,000 receptors per cell nucleus. In this report, we show that L-T3 exerts a pleiotypic effect on GH, cells to rapidly (within 2 h) stimulate [3H]uridine uptake to a maximal value of 2.5-to %fold after 24 h. This results from an increase in the number of functional uridine "transport sites" as shown by studies documenting an increase in the apparent Vmax with no change in the K m , 17 p~. Although the labeling of the cellular uridine pool and pools of all phosphorylated uridine derivatives was increased by L -T~, there was no change in the relative amounts of the individual pools in cells incubated with or without hormone. The intracellular concentration of [3H]uridine was estimated to be similar to that of the medium, suggesting that facilitated transport mediates [3H]uridine uptake. That this increase in r3H]uridine transport was nuclear receptor-mediated is supported by the excellent correspondence of the L-T3 dose-response curve for [3H] uridine uptake and that for L-T3 binding to receptor. Finally, inhibition of protein synthesis by cycloheximide and RNA synthesis by actinomycin D demonstrated that the L-T3 effect required continuing protein and RNA synthesis. These results are consistent with an effect of the ~-T3-nuclear receptor complex to increase uridine uptake in GH, cells by altering the expression of gene(s) essential for the transport process.
kinetics of binding of L -[ "~I ] T~ to nuclear receptors (2). After a lag period, L-T3 also stimulates cell replication (6). Hershko et al. (7) proposed that cells exhibit pleiotypic responses in which a set of unrelated metabolic events (e.g. amino acid, glucose, and nucleoside transport) change in concert with cell replication. If these changes are coordinately related to L-T3 stimulation of cell growth, these processes should be mediated by the same nuclear receptor which controls cell replication. Segal and Gordon (8,9) reported that L-T3 can rapidly stimulate deoxyglucose and uridine uptake in cultured chick embryo fibroblasts. Stimulation of deoxyglucose uptake during the first 6 h of L-T3 incubation was not blocked by inhibitors of RNA or protein synthesis. This suggested that L-T3 stimulation of membrane-related events could occur independently of a transcriptionally related process (8). However, the reported potency of different iodothyronine analogues paralleled that reported for responses thought to be mediated by the nuclear receptor (10). A similar iodothyronine analogue potency was subsequently described for 3-O-methylglucose uptake by rat thymocytes ( l l ) , a response which also was not blocked by cycloheximide. Although this occurred at concentrations several orders of magnitude greater than physiologic, iodothyronine binding studies suggested the existence of a membrane component having similar relative affinities for certain analogues as the nuclear receptor (11). These and other observations (12,13) imply that the thyroid hormones can interact with plasma membrane receptors which directly regulate membrane transport events.
In this study we have examined the effects of L-T3 on [3H] uridine uptake by GHI cells. L-T3 stimulates [3H]uridine uptake after a l-h lag period, and the rate becomes maximal within 24 h of incubation. Analysis of the intracellular acidsoluble radiolabeled uridine, UMP, UDP, and UTP pools indicates that L-T3 stimulates the transport of uridine which is then phosphorylated to generate the uridine nucleotides. Stimulation of [3H]uridine transport by L-T3 in GH, cells is totally dependent on RNA and protein synthesis. This observation along with the excellent parallelism between L-T3 nuclear receptor occupancy and [3H]uridine uptake indicates that this pleiotypic response reflects a membrane transport event that is modulated by the thyroid hormone nuclear receptor.

GHl Cell Culture Conditions and Measurement of [3HIUridine
Uptake-GH1 cells were cultured with Ham's F-10 medium supplemented with 2.5% fetal calf serum and 15% horse serum as previously described (3, 6). To study the effects of L-T3 on uridine uptake, the media was exchanged with Ham's F-10 media containing 10% (v/v) thyroid hormone-depleted calf serum, and the cells were incubated for 48 h. This process was repeated one additional time. L-T3 was added to the cell cultures in 10-20 pl of the same medium to achieve the fmal concentrations indicated in the text. At various times after L-T3 incubation, uridine uptake was assessed by incubating the cells with 1-3 pCi/ml of [3H]uridine for 0.5-1 h as indicated. In some experiments the cells received cycloheximide (0.5 mM) or actinomycin D (1 pg/ml) at the times indicated prior to [3H]uridine incubation. These compounds were prepared as 50-fold-concentrated stock solutions and added directly to the medium. After the incubation with [3H]uridine, the media were then removed and the cell monolayers were rapidly washed three times at 4 "C with chilled isotonic saline. One ml of 8% (w/v) trichloroacetic acid was then added to each monolayer to determine the trichloroacetic acid-soluble and precipitable uridine-radiolabeledpools. The trichloroacetic acid was removed and the residual cell material, which remains attached to the plastic surface, was solubilized with 0.4 N NaOH.
Aliquots of the trichloroacetic acid-soluble fraction and the 0.4 N NaOH fractions were analyzed using Aquasol and a liquid scintillation counter. A sample of the trichloroacetic acid-soluble fraction was also used for analysis of DNA (15) or protein (16). One million GH1 cells contain 10 pg of DNA and 100 pg of protein. To exclude the possibility that the radiolabeled trichloroacetic acid precipitable fraction represented incorporation into DNA, the fractions solubilized with 0.4 N NaOH were incubated for 2 b at 30 "C and then reprecipitated with 10% trichloroacetic acid (v/v). Less than 5% of the original trichloroacetic acid-insoluble radiolabeled material was recovered in the second trichloroacetic acid precipitate indicating that over 95% of the radiolabel was incorporated into RNA.
Separation of the Intracellular Radiolobeled Uridine Fractions-In some experiments the trichloroacetic acid-soluble radiolabeled material was separated into uracil, uridine, UMP, UDP, and UTP using a Bio-Rad AG 1X-10 resin column as described by Lindsay et al. (17). The trichloroacetic acid was removed by four sequential extractions with 4 volumes of water saturated diethyl ether. The aqueous phase (0.2 ml) was reacted with 0.3 ml of 0.1 M K2B4O7 and 0.08 ml of 1.0 N NH,OH to form a borate complex with uridine. The sample was then applied to a 0.7 X 2.5-cm column of Ag 1X-10 resin and the column was eluted with 8 ml of the same buffer. Uracil, uridine, UMP, UDP, and UTP were then sequentially eluted with increasing concentrations of NH&l as described (17).
Assessment of Thyroid Hormone Nuclear Receptor Occupancy by L- [1261]T3-Experiments to estimate thyroid hormone nuclear receptor occupancy were performed by culturing cells in 25-cm2 flasks with F-10 medium containing 10% (v/v) hormone-free calf serum for 24 h. The flasks were then incubated with the indicated concentrations of L -( ' *~I ] T~ for 24 h at 37 "C. The cell monolayers were then chilled to 4 "C and harvested to isolate the nuclei as previously described (2,18). The radioactivity in the nuclear pellet was quantitated using a refrigerated Packard y spectrometer at 55% efficiency followed by DNA determination. Nonspecific binding of L-['*~I]T~ to GHI cell nuclei was estimated using 5 pM nonradioactive L -T~. This value, always less than 5% of total binding, was subtracted from results obtained using radioactive hormone alone.

RESULTS AND DISCUSSION
Kinetics of Stimulation of Uridine Uptake and Incorporation into RNA by L-T3 in Cultured GH, Cells-The coordinate relationship of the cell uptake of amino acids, glucose, and nucleosides, and stimulation of growth has been recognized for over a decade (7,(19)(20)(21)(22)(23)(24)(25)(26)(27). Stimulation of replication by a variety of distinct factors increases the cell accumulation of [3H]uridine and has been extensively studied in a variety of cell lines (7,19,20,23,26,27). We have previously demonstrated that L-T3 increases the growth rate of GH1 cells (6). Therefore, we attempted to determine if L-T3 had a pleiotypic effect on the transport and/or metabolism of [3H]uridine. Fig.   1 illustrates the effect of 5 nM L-T3 on stimulating [3H] uridine uptake in GH1 cells into the trichloroacetic acidsoluble pool over 31 h of incubation. L-T3 increased the rate of accumulation of [3H]uridine into the trichloroacetic acidsoluble pool within 3 h, and this increased progressively to a maximal rate which was achieved within 24 h. Incorporation of [3H]uridine into RNA paralleled the change in accumulation of [3H]uridine into the trichloroacetic acid-soluble pool, and at each time point in the ~-T3-incubated and the control cells approximately 30% of the total cell 3H radioactivity was incorporated into RNA (not illustrated). Therefore, when corrected for the change in the trichloroacetic acid-soluble [3H]uridine nucleotide pool, L-T3 does not appear to stimulate total cell RNA synthetic rates.
Analysis of the Trichloroacetic Acid-soluble [3H]Uridine and Phosphorylated 13H] Uridine Pools-Evidence in a number of cell lines indicates that the rate-limiting step in the cellular accumulation of [3H]uridine is due to an increase in the rate of facilitated transport and not the phosphorylation of uridine (19,22,26,27). That this is the case in GH1 cells is indicated in Table I. L-T3 stimulated an approximate %fold increase in total cell [3H]uridine accumulation and incorporation into RNA represented approximately 40% of the total cell radioactivity. The trichloroacetic acid-soluble radiolabeled material was separated into fractions representing uracil, uridine, UMP, UDP, and UTP. With the exception of uracil, the picomoles of 3H radioactivity in each fraction of the L -T~incubated cells was approximately 2-fold greater than the control cells. In both the ~-T3-incubated and the control cells each radiolabeled fraction represents the same percentage of the total trichloroacetic acid-soluble radiolabeled material (uridine, 2%; UMP, 22%; UDP, 16%; UTP, 60%). Although the levels of [3H] in the pools of uridine and in all of the phosphorylated uridine derivatives increase with L-T3 incubation, the ratio of the L-T3 to the control cell cultures remains the same for each fraction. This strongly argues that metabolism of uridine to UMP, UDP, and UTP is not responsible for L-T3 mediated increase in uridine uptake by these cells.
Additionally, GHl cells have an average diameter of 16 pm which can be used to calculate an average cell volume of about 2 x mm3. Based on the picomoles of [3H]uridine/106 cells, and assuming that the cells are 80% water, we estimate that the intracellular [3H]uridine concentration in the L-T3 incubated cells is about 25-30 nM. This value is virtually identical to the extracellular [3H]uridine concentration used in the experiment of Table I (29 nM) suggesting that [3H] uridine enters the cell by facilitated transport.
Influence of L-T3 on the V,,, and the K, for Uridine Transport- Table I suggests that L-T3 stimulates the accumulation of [3H]uridine into GH1 cells which is then rapidly phosphorylated to UMP, UDP, and UTP. This supports the notion that L-T3 increases the rate of [3H]uridine uptake either by stimulating an increase in the K, for uridine transport or by increasing the number of transport sites. The K , and V, . for uridine uptake in control cells and cells incubated with L-T3 for 6 h were examined (Fig. 2). L-T3 stimulated a 1.7-fold increase in the V,,, while K , for uridine uptake (17 p M ) was identical to control values. This suggests that L-T3 stimulates [3H]uridine uptake by increasing in the number of functional transport units. Furthermore, the K, for facilitated transport reported for a number of cell lines ranges from 6 to 80 pM (19,(26)(27)(28)(29). These values are in good agreement with the K, of 17 p~ which we have observed in GHl cells f L-T3.
Uridine transport in some but not all cell lines can be uridine for 45 min prior to the times indicated in the figure. The cells were then chilled to 4 "C and the acidsoluble radioactivity determined. Each point reflects the average of three cell cultures which showed less than +5% variation.

FIG. 2 (center). Estimation of the K,,, and the V , , for uridine uptake in L-T3 incubated and control
cell cultures. GHI cells were incubated with (0) or without (0) 5 nM L-T3 for 6 h. [3H]Uridine was then added (1 pCi/ml) along with various concentrations of nonradioactive uridine. After a 1-h incubation a sample of the medium was counted to determine the actual uridine concentrations which were calculated to be 8, 11, 15,22,40,85, and 160 p~. The cells were washed and processed as described under "Experimental Procedures" to determine the acid-soluble radioactive fraction. V, pmol/h/106 cells. Each point represents the average of three cell cultures which showed less than +5% variation.  GH1 cells were incubated +5 nM L-T3 for 6 h followed by a 1-h incubation with 3 pCi/ml [3H]uridine. The cells were then washed and the acid-soluble and -insoluble fractions prepared using 8% trichloroacetic acid (w/v). After removing the trichloroacetic acid with diethyl ether, the acid-soluble radioactivity was then separated into fractions containing uracil, uridine, UMP, UDP, and UTP using AG 1-XlO resin as described under "Experimental Procedures." The amount of incorporation of [3H]uridine into RNA was taken to reflect the acid-insoluble radioactivity. Each point represents the average of three cell cultures which showed less than 28% variation.   [3H]Uridine Transport by L -T~-A previous study in cultured chick embryo heart cells suggested that L-T3 could stimulate sugar transport even when RNA or protein synthesis was inhibited by actinomycin D or cycloheximide (8). To explore this we examined the influence of cycloheximide (0.5 mM) and actinomycin D (1 pg/ml) on uridine uptake in the trichloroacetic acid-soluble pool i R cells first incubated for 24 h with L-T3 to stimulate steady state levels of [3H]uridine uptake (Fig. 3). These concentrations of cycloheximide and actinomycin D inhibit protein synthesis by 95 and 20%, respectively (33), while actinomycin D inhibits RNA synthesis by greater than 95% (33). Cycloheximide incubation caused a first-order decay in the rate of [3H]uridine uptake with a halflife of approximately 3 h in the ~-TB-cultured cells. The relative amounts of UTP, UDP, UMP, and uridine in the radiolabeled trichloroacetic acid-soluble pool was not altered by cycloheximide incubation (data not shown).

FIG. 3 (right). Influence of cycloheximide and actinomycin D on the ~-T3-stimulated and basal ['HI uridine uptake in GHI cells. GHl cells were incubated
Actinomycin D also elicited a decrease in the rate of uridine uptake in ~-T3-incubated cells which after a 1-2-h lag period also decayed with a half-life of about 3 h. The decay of uridine uptake in the control cell cultures incubated with cycloheximide or actinomycin D (Fig. 3) both showed a half-life of greater than 15 h. The reason for the different half-lives in the control and L-T3 cultured cells are unclear but may indicate that L-T3 stimulates the synthesis of a distinct uridine transport system which has a similar K , as the control cells. Alternatively, the transport system stimulated by L-T3 may be incorporated into a different membrane "environment" and therefore turns over more rapidly than in the control cell cultures.
To assess whether L-T3 directly activates the uridine transport system or if this regulation is mediated by stimulation of RNA and protein synthesis, we examined the influence of cycloheximide (0.5 mM) and actinomycin D (1 pg/ml) on the induction process. Cycloheximide (Fig. 4) completely inhibited the induction of uridine uptake by L-T3 to the same value as in the control cells with cycloheximide. Furthermore, addition of cycloheximide 4 h after L-T3 stimulation of uridine transport resulted in a reduction in the uridine transport rate to the same value as the control cells. Similar results were also obtained in a separate experiment performed with actinomycin D (Fig. 5).
In a study using a related cell line (GH4C1), Martin et al. (29) reported that thyrotropin-releasing hormone (TRH) induced a rapid 2-to 3-fold increase followed by a prolonged 40-50% decrease in [3H]uridine uptake. These changes were attributed to an effect of TRH on [3H]uridine phosphorylation (34). Although the rapid effect of TRH on [3H]uridine uptake was due to a change in the apparent V,,, rather than the K,,, (34), there are two significant differences between our observations with L-T3 and those of Martin et al. (29). First, they observed a maximal effect within 30 min of TRH incubation. This is much more rapid than the effect of L-T3 in GH,cells (Figs. 1,4,and 5). Second, [3H]uridine uptake stimulated by TRH was not altered by cycloheximide incubation, whereas the effect of L-T3 in GH, cells is inhibited (Figs. 3 and 4).

Complexes to Stimulation of r3H]Uridine Uptake by L -T~-
The above studies indicate that induction of uridine transport by L-T3 is mediated by a mechanism which is dependent upon RNA and protein synthesis. This supports the notion that this transport function is regulated by the thyroid hormone nuclear receptor which has been shown to be rate-limiting for transcriptional control of genes regulated by thyroid hormone (2). Further support for this conclusion is illustrated in Fig.  6. This study shows an excellent agreement between the rate of uridine transport 24 h after incubation with 0.1-5 nM L-T3 and the level of ~-['~~I]T3-receptor complexes present 24 h after incubation with L -[ "~I ] T~ over the same concentration range.
Although abundant evidence indicates that the primary action of thyroid hormone is mediated at the transcriptional level by its nuclear receptor (2-5), several reports have suggested that L-T3 can increase membrane transport processes by a direct interaction with plasma membrane binding components (8-11). While this study was in progress Halpern and Hinkle (35) reported that L-T3 rapidly stimulated the accumulation of [3H]uridine in GH4Cl cells. The response, 1.5-to 2-fold, was maximal within 2 h and did not increase with further incubation. However, basal total cell [3H]uridine uptake was increased 100% by cycloheximide or decreased by 50% with 10 pg/ml actinomycin D. Therefore, it was not possible to resolve whether stimulation of [3H]uridine uptake by L-T3 was dependent on RNA and protein synthesis or to exclude that L-T3 acted at the cell surface as suggested for TRH stimulation of [3H]uridine uptake in GH4C, cells (35).
Several lines of evidence suggest that L-T3 stimulation of [3H]uridine transport in GH, cells is mediated at the nuclear level and does not reflect a direct action at the plasma membrane. First, Fig. 6 shows an almost exact correspondence between the dose-response curve for ~-T3-receptor binding and L-T3 stimulation of [3H]uridine transport. Second, ongoing protein synthesis is necessary for the effect since the L-T3 response is rapidly inhibited by cycloheximide (Figs. 3 and  4). Supporting this observation are the results of Surks et al. received unlabeled L-T3 for 24 h were incubated for 1 h with 1 pCi/ ml of 13H]uridine and the acid-soluble radioactive fraction (0) determined as described under "Experimental Procedures." Each point represents the average of three cell cultures which showed less than +7% variation. The extent of "nonspecific" binding of L-['*~I]T~, as determined by co-incubation with 5 FM L -T~, represented less than 5% of the total nuclear bound ' *' I radioactivity.
acid transport in GC cells was dependent on protein synthesis.
Additionally, iodothyronine analogue potency was found to parallel the relative analogue affinity for the thyroid hormone nuclear receptor (24). Finally, the actinomycin D-mediated inhibition of increased [3H]uridine transport (Figs. 3 and 5), under conditions where protein synthesis is only minimally inhibited, suggests that RNA synthesis is necessary for L -T~mediated effect. Our results, along with the demonstrated ability of L-T3 to rapidly stimulate growth hormone gene transcription (2), suggest that the hormone-mediated increase in [3H]uridine uptake is likely due to the regulated expression of genes involved in the transport process.