The Prostaglandin Transporter PGT Transports PGH₂

Abstract

Prostaglandin H₂ not only serves as the common precursor of all other PGs, but also directly triggers signals (eg. platelet aggregation), depending on its location and translocation. The prostaglandin carrier PGT mediates the transport of several prostanoids, such as PGE₂, and PGF_{2 α}. Here we used PGT in the plasma membrane as a model system to test the hypothesis that PGT also transports PGH₂. Using wild type and PGT-expressing MDCK cells, we show that PGH₂ uptake is mediated both by simple diffusion and by PGT. The PGH₂ influx permeability coefficient for diffusion is (5.66 ± 0.63) × 10⁻⁶ cm/s. The kinetic parameters of PGH₂ transport by PGT are K_m = 376 ± 34 nM and V_max = 210.2 ± 11.4 fmoles/mg protein/s. PGH₂ transport by PGT can be inhibited by excess PGE₂ or by a PGT inhibitor. We conclude that PGT may play a role in transporting PGH₂ across cellular membranes.

Introduction

Prostaglandin H₂ (PGH₂) is the common precursor of all other PGs (PGE₂, PGI₂, PGD₂ and PGF₂ ) and thromboxane A₂ (TxA₂). It is converted to these PGs by the specific terminal synthases in the cytoplasm (1-6). Besides serving as a common precursor, extracellular PGH₂ can directly trigger signals including Ca²⁺ release, serotonin release, vasoconstriction, and platelet aggregation (7-10). The signals induced by extracellular PGH₂ can therefore oppose the signals resulting from intracellular PGH₂, such as formation of PGI₂, the most potent endogenous inhibitor of platelet aggregation (11, 12). Thus the net result of PGH₂ depends on its localization of PGH₂. Of interest, Kent et al. (9) reported that endothelial cells exposed to PGH₂ produced PGI₂, suggesting that PGH₂ might be taken up into endothelial cells and converted to PGI₂ by PGI₂ synthase. This raises the question of how PGH₂ translocation occurs.

The prostaglandin transporter (PGT) has been shown to transport several prostaglandins, including PGE₂ (13, 14). PGT mediates the active uptake of PGE₂ into the cell against a concentration gradient (15). Based on the structural similarity between PGH₂ and PGE₂, as shown in Fig. 2C and D, we hypothesized that PGT might also transport PGH₂. Using MDCK cells stably expressing PGT, as previously reported (16), we conducted a detailed investigation of the kinetics of PGH₂ transport by PGT. The competition of transport by PGT between PGH₂ and PGE₂, and the inhibition of PGH₂ transport by our newly identified PGT inhibitor, TGBz T34 (15), confirmed that PGH₂ is, indeed, a substrate of PGT.

Fig. 2.

Plots of initial velocity of PGH₂ (A) or PGE₂ (B) uptake by PGT-expressing MDCK cells in the presence (squares) or absence (circles) of 25 μM TGBz T34 as a function of PGH₂ (A) or PGE₂ (B) concentration. Circles represent the total uptake and squares represent diffusional influx. The diamonds, which represent [total - diffusional] could be fit by the Michaelis-Menton equation (V_i = V_max [S] / ([S] + K_m)). These are representative plots of three individual sets of experiments. In each set of experiment, the error bars are S.D. values of triplicates. The kinetic parameters were generated in each of 3 sets of experiments and presented as mean values ± S.D. in Table 1. C and D, chemical structures of PGE₂ and PGH₂, respectively

Materials and methods

MDCK cells were stably transfected with the GFP-tagged PGT in our laboratory (16). Tritium labeled PGE₂ ([³H)PGE₂) and PGH₂ ([³H)PGH₂) were purchased from PerkinElmer and CaymanChem, respectively. Unlabeled PGE₂ was obtained from CaymanChem.

Time Course of PGH₂ or PGE₂ Transport

MDCK cells were seeded at 15-20% confluence on 24-well plates. The day on which the cells were seeded was considered day one. PGE₂ uptake experiments were conducted on day 4. All of the PGH₂ or PGE₂ uptake experiments were conducted at room temperature. On day 4, cells were washed twice with Waymouth buffer (135 mM NaCl, 13 mM H-Hepes, 13 mM Na-Hepes, 2.5 mM CaCl₂, 1.2 mM MgCl₂, 0.8 mM MgSO₄, 5 mM KCl, and 28 mM D-glucose). Then 200 μL of Waymouth buffer containing 1 μM [³H]PGH₂ or 1 nM [³H]PGE₂ were added to each well. At the designated time, the uptake of [³H]PGH₂ or [³H]PGE₂ was stopped by aspiration of uptake buffer, followed immediately by two washes with 500 μL of chilled Waymouth buffer. Cells were then lysed with 100 μL lysis buffer containing 0.25% SDS and 0.05 N NaOH. 1.5 mL of scintillation solution was added to each well, and intracellular [³H]PGH₂ or [³H]PGE₂ was counted by MicroBeta scintillation Counter.

To obtain the time course of PGE₂ transport in the presence of PGH₂ or TGBz T34, MDCK cells were incubated in Waymouth buffer containing 1 nM [³H]PGE₂ at room temperature for 9 minutes to reach the maximum intracellular level. Thereafter, either PGH₂ or TGBz T34 was added to the uptake buffer at a final concentration of 2.5 μM or 25 μM. At the designated time, the uptake of [³H]PGE₂ was stopped and uptake was assessed as described above. For the time course of PGH₂ uptake in the presence of TGBz T34, MDCK cells were incubated in Waymouth buffer containing 1 μM [³H]PGH₂ at room temperature for 120 minutes to reach the maximum intracellular level. At 120 minutes, TGBz T34 was added to the uptake buffer at a final concentration of 25 μM. Then the transport was stopped at varied time points and uptake was quantified as above.

Measurements of Kinetic Parameters of PGE₂ and PGH₂

We repeated the time courses of PGE₂ or PGH₂ uptake in the presence or absence of 25 μM TGBz T34 at various initial extracellular concentrations of PGE₂ (0, 20, 40, 80, 120, 160, and 300 nM) or PGH₂ (0, 10, 25, 50, 100, 250, 500, and 1000 nM). In the case of PGE₂, at low concentrations, the extracellular concentrations were taken as ³H labeled PGE₂, which has a specific activity of 500 μCi/mol. At high concentrations of PGE₂, we made a mixture of ³H labeled and unlabeled PGE₂ to a final specific activity of 25 μCi/mol.

The initial velocities at various extracellular concentrations of PGE₂ or PGH₂ were determined from the PGE₂ uptake in the first 2 minutes or the PGH₂ uptake in the first 3 minutes, in the presence or absence of TGBz T34; these were linear over the early time course of PGE₂ or PGH₂ uptake. The permeability coefficient of PGH₂ or PGE₂ influx (P_in) was obtained by linear regression fit of the initial rate in the presence of TGBz T34 versus extracellular PGE₂ or PGH₂ concentration.

We subtracted the initial velocities in the presence of TGBz T34 from those in the absence of TGBz T34. These resulted initial velocities were used to obtain K_m and V_max values by nonlinear regression fit of the initial rate versus extracellular PGE₂ or PGH₂ concentration to the Michaelis-Menten equation (V_i = V_max [S] / ([S] + K_m)).

Results

Time Course of PGH₂ Uptake

The time course of PGH₂ uptake by MDCK cells expressing PGT is shown in Fig. 1A. The circles represent the total PGH₂ uptake when the cells were incubated with 1 μM [³H]PGH₂. Squares indicate experiments in which cells were incubated with 1 μM [³H]PGH₂ in the presence of 25 μM TGBz T34, our newly identified PGT inhibitor (15). The line with diamonds represents PGH₂ total uptake minus PGH₂ influx when the PGT inhibitor was applied, which is equivalent to PGT-mediated uptake. The “overshoot” is characteristic of PGT (17).

Fig. 1.

A, time course of PGH₂ uptake by PGT-expressing MDCK cells in the presence (squares) and absence (circles) of 25 μM TGBz T34. The diamond line is formed by subtracting intracellular PGH₂ in the absence of TGBz T34 (circles) by intracellular PGH₂ in the presence of TGBz T34 (squares). B, time course of PGH₂ uptake by wild type MDCK cells. In both A and B, the values are presented as mean ± S.D. of three individual replicates.

Fig. 1B shows PGH₂ uptake by wild-type MDCK cells. The inset of Fig. 1B shows the time course of PGH₂ uptake by wild-type MDCK cells for the first 15 minutes. It is evident that PGH₂ influx in PGT-expressing cells plus the inhibitor TGBZ T34 (squares in Fig. 1A) is almost identical to the data of Fig. 1B; each of these represents PGH₂ influx by simple diffusion. Since the pattern of total PGH₂ uptake was similar to that of PGT-mediated uptake, we conclude that PGT mediates the majority of PGH₂ uptake.

Kinetics of PGH₂ Uptake in Comparison with PGE₂ Uptake

To obtain detailed kinetic parameters of PGH₂ influx, we measured initial rates of PGH₂ uptake at various extracellular PGH₂ concentrations in the presence and absence of TGBz T34. In the presence of PGT inhibitor, the plot of the initial rates of PGH₂ uptake versus concentration could be fitted with a straight line (squares in Fig. 2A), indicating that this part of influx was caused by simple diffusion. We calculated the permeability coefficient of PGH₂ influx (P_in) by diffusion by dividing the slope of this linear line by the total cell surface. P_in for PGH₂ was (5.66 ± 0.63) × 10⁻⁶ (Table 1).

Table 1.

Kinetic parameters of PGH2 and PGE2 influx by both PGT and diffusion.

Kinetic Parameters	PGH2	PGE2
Km (nM)	376 ± 34	91.0 ± 8.2
Vmax (fmoles/mg protein/s)	210.2 ± 11.4	78.9 ± 6.9
Influx permeability coefficient (Pin) (cm/s)	(5.66 ± 0.63) × 10−6	(1.06 ± 0.13) × 10−6

All kinetic parameters are presented as mean ± S.D. (n = 3).

In the absence of TGBZ T34, the plot with circles depicts the total PGH₂ uptake (Fig. 2A). The PGT mediated rates (diamonds) from Fig. 2A were obtained by subtracting the influx by diffusion (squares) from the total influx (circles). When plotted against PGH₂ concentration, they could be fitted by the Michaelis-Menton equation (Fig. 2A). The binding constant of PGH₂ to PGT, K_m, and the maximum velocity, V_max, are listed in Table 1. PGT-mediated transport constituted 80% of PGH₂ influx; the remainder was mediated by diffusion.

Using the same method, we conducted a similar investigation of the kinetics of PGE₂ uptake. As shown in Fig. 2B, circles depict the total PGE₂ uptake in the absence of TGBz T34, and squares depict the PGE₂ influx in the presence of TGBz T34. The plot of the latter is linear and shows the component of PGE₂ influx caused by diffusion. P_in for PGE₂, was (1.06 ± 0.13) × 10⁻⁶ cm/s (Table 1). Using the same methods as for analyzing PGH₂ kinetics, we generated the PGT-mediated uptake component (diamonds), which could be fitted by the Michaelis-Menton equation. The binding constant of PGE₂ to PGT, K_m, and the maximum velocity, V_max, are listed in Table 1. As opposed to the case with PGH₂, PGT-mediated PGE₂ transport was almost identical to total PGE₂ uptake, suggesting that PGE₂ influx is mainly mediated by PGT. Comparison of the kinetic parameters of PGE₂ to PGH₂ uptake shows that P_in for PGH₂ was 5-fold that for PGE₂; K_m of PGH₂ was 4.1-fold that of PGE₂; and V_max of PGH₂ was 2.7-fold that of PGE₂.

To address the possibility that ³H-PGH₂ was metabolized to ³H-PGE₂ and that the measured tracer influx therefore represented ³H-PGE₂, and not ³H-PGH₂, we added 0, 0.5, or 1.0 μM unlabeled PGH₂ to the medium overlaying WT MDCK cells at 37°C, waited 3 minutes, and determined the resulting PGE₂ concentrations by enzyme-linked immunoassay. The resulting medium PGE₂ concentrations were 0.159 ± 0.014, 0.166 ± 0.014, and 0.156 ± 0.017 nM (n = 3), respectively. These trivial levels of PGE₂ are well below the affinity of PGT (Table 1), indicating that conversion of PGH₂ to PGE₂ cannot account for our uptake results.

PGH₂ serves as a PGT anion exchange substrate

In the absence of an extracellular substrate for PGT, PGE₂ efflux occurs exclusively by diffusion (15, 17). However, when a PGT substrate is present in the extracellular medium, PGE₂ efflux is accelerated via PGT acting as an obligate anion exchanger (17). In the following experiments, we examined whether unlabeled extracellular PGH₂, like unlabeled extracellular PGE₂, can serve as a “trans” exchange partner to drive tracer PGE₂ efflux.

As shown in Fig. 3 (circles), under control conditions the cells could be rapidly loaded with ³H-PGE₂ such that intracellular ³H-PGE₂ reached a peak at 9 min and leveled off thereafter. This biphasic pattern results from a “pump-leak” system in which PGT-mediated uptake (“pump”) initially exceeds the diffusion-mediated efflux (“leak”) rate (18). As expected from our prior work (15), application of 25 μM TGBz T34 at 9 minutes abruptly stopped further PGT-mediated ³H-PGE₂ uptake. The subsequent rate of ³HPGE₂ efflux represents the background diffusional leak (dots in Fig. 3).

Fig. 3.

Time Course of PGE₂ uptake by PGT-expressing MDCK cells in the absence (circles) and presence of 2.5 μM PGH₂ (squares) or 25 μM TGBZ T34 (closed circles). PGE₂ uptake as a function of time was first established by allowing uptake to various time points, washing, and counting intracellular PGE₂ (open circles). In separate experiments, either 2.5 μM PGH₂ or 25 μM TGBZ T34 was added at 10 minutes when intracellular PGE₂ had reached the peak level. Thereafter, PGE₂ uptake was stopped at various points by washing and intracellular PGE₂ was measured. The values are presented as mean ± S.D. of triplicates.

We repeated the experiment, but this time added 2.5 μM unlabeled extracellular PGH₂ during the ³H-PGE₂ uptake. Intracellular ³H-PGE₂ fell to baseline within 5 minutes and remained at that level for the rest of the time course. Of note, extracellular PGH₂ induced a faster efflux of intracellular ³H-PGE₂ than TGBz T34, indicating that PGH₂ exchanges with ³H-PGE₂ on PGT.

Diffusional efflux of PGH₂

Because PGH₂ influx can occur to some extent by diffusion (Fig. 2 and Table 1), we examined the degree to which PGH₂ also effluxes from cells by diffusion. Toward this end, we loaded cells via PGT for 120 minutes with ³H- PGH₂ to a peak level, and then stopped further uptake with 25 μM TGBz T34. As shown in Fig. 4, this block of carrier-mediated PGH₂ uptake was followed by a rapid depletion of intracellular ³H-PGH₂.This pattern is qualitatively the same as for PGE₂ (Fig. 3), and indicates that pre-loaded ³H-PGH₂ exits cells by diffusion in the same fashion as pre-loaded PGE₂.

Fig. 4.

Diffusional efflux of PGH₂. As a baseline, ³H-PGH₂ uptake was allowed to proceed to various time points and stopped by washing, and then intracellular ³H-PGH₂ was measured (open circles). In separate experiments, 25 μM TGBZ T34 was added at 120 minutes when intracellular PGH₂ had reached peak level. Thereafter PGH₂ uptake was stopped at various time points by washing and intracellular PGH₂ was measured (closed circles). The values are presented as mean ± S.D. of triplicates.

Discussion

PGH₂ serves as the common precursor of all other prostaglandins inside cells. It is also rapidly released into the extracellular compartment (Fig. 3) where it can trigger Ca²⁺ release and platelet aggregation by activating its receptor (7-10, 19). These signals could be reduced as PGH₂ was taken back up into cells and converted into other PGs, including PGI₂, which inhibits platelet aggregation. It appears that PGH₂ plays different roles and may trigger opposing signals, directly or indirectly, from different cellular compartments. Therefore, its translocation is critical for its level in either the intracellular or extracellular compartments. Here, we show that PGH₂ efflux is mainly mediated by diffusion and its influx is mediated by both passive diffusion and by a carrier, PGT.

Compared to PGE₂, PGH₂ is transported by PGT with a lower affinity, but at a higher rate. Using transport across the plasma membrane as a model system, we found that PGH₂ has a higher passive permeability than PGE₂. The time course of PGH₂ uptake across the plasma membrane in our assay system is different from that of PGE₂ uptake (Fig. 1A and 3). It took about two hours for PGH₂ to reach a peak intracellular level, whereas it took only 10 minutes for PGE₂ to do so. This is in spite of the fact that about 20% of PGH₂ influx is caused by diffusion, whereas PGE₂ influx by diffusion is negligible. Correspondingly, the efflux of PGH₂ caused by simple diffusion is much greater than that of PGE₂. Therefore, in the “pump-leak” system of PGT-mediated uptake and diffusional efflux, the latter component is much higher for PGH₂ than for PGE₂.

The differences in transport kinetics between PGH₂ and PGE₂ are likely due to chemical structure differences. PGH₂ has a 7-member ring with two oxygens at the head of the molecule, whereas PGE₂ has a five-member ring with a hydroxyl group. Although there are two oxygens in the 7-membrane ring, PGH₂ is more hydrophobic than PGE₂ because of its lack of a hydroxyl group. This greater hydrophobicity probably accounts for the higher passive diffusion of PGH₂ relative to PGE₂. The same structural difference probably accounts for the weaker binding of PGH₂ to PGT, as compared to PGE₂, implying that it is the head of the fatty acid that is interacting with PGT. The -OH group on the ring of PGE₂ is important for the interaction, and it is likely to be a proton donor instead of an acceptor, in order for PGE₂ to be differentiated from PGH₂.

Although PGH₂ binds to PGT more weakly than PGE₂, the V_max of PGH₂ is higher than that of PGE₂, indicating that PGT turns over PGH₂ faster than PGE₂. The V_max of PGH₂ is 2.7-fold that of PGE₂. Given the same amount of PGT, k_cat of PGH₂ will be 2.7 folds of that of PGE₂. The ratio of K_m of PGH₂ to that of PGE₂ is 4.1. The specificity of PGT for a substrate is defined by k_cat/K_m. Since k_cat/K_m for PGE₂ is 1.5 fold that for PGH₂, we conclude that PGT slightly prefers PGE₂ over PGH₂.

As shown in Fig. 3B and 4, TGBz T34 exerted full inhibition of PGE₂ transport, implying that almost 100% of PGE₂ influx was mediated by PGT. In the case of PGH₂ uptake, TGBz T34 inhibited 80% of PGH₂ transport, consistent with the conclusion that about 80% of PGH₂ uptake was mediated by PGT and about 20% by diffusion (Fig. 2A and 4).

PGH₂ is synthesized by two enzymes cyclooxygenase (COX)-1 or –2, which have been localized to the luminal surfaces of the endoplasmic reticulum (ER) (20), at specific activities of 25 – 40 nmoles/min/mg protein (21-23). PGH₂ is consumed in the cytoplasm by the prostaglandin synthases (PGDS, PGES, PGFS, PGIS), at specific activities of 1 – 4000 nmoles/min/mg protein (1; 2; 4-6; 24-30). The luminal surface of the ER is topologically similar to the exofacial surface of the plasma membrane. Thus, present studies shed light on the kinetics of PGH₂ transport across the ER membrane. The specific transport activity of PGT for PGH₂ is in the range of 0.012 - 1 nmoles/min/mg protein. If PGH₂ translocation from the ER lumen to the cytoplasm is rate-limiting for prostaglandin synthesis, PGT might control the translocation of PGH₂ across the ER, and therefore the accessibility of PGH₂ to the terminal prostaglandin synthases.

In any event, PGT-mediated uptake of PGH₂ may regulate the net signaling that PGH₂ triggers. PGT is strongly expressed in endothelial cells and its expression is induced when endothelial cells are under sheer stress (31-33). PGH₂-induced platelet aggregation would be counteracted by the anti-coagulation effect of PGI₂, formed after PGH₂ is transported into endothelial cells by PGT.

Abbreviations

PG

Prostaglandin

PGT

prostaglandin transporter

PGH₂

prostaglandin H₂

PGE₂

prostaglandin E₂

AA

arachidonic acid

TxA₂

thromboxane A₂

COX

cyclooxygenase

ER

endoplasmic reticulum