Abstract
Transport of several monocarboxylic acids across the lipid bilayer was examined in liposomes consisting of egg yolk phosphatidylcholine and cholesterol. In the presence of inward proton gradient, salicylic acid (SA) was taken up rapidly by liposomes showing overshoot, saturation and competitive inhibition phenomena. These carrier-mediated like profiles of SA uptake can be explained by assuming a very high permeability through the liposomal membrane of protonated SA. Protonated SA in the extraliposomal solution (pH 5.8) was taken up by liposomes rapidly, followed by a redissociation to anion according to the intraliposomal pH (pH 7.5). The concentration gradient of protonated SA across the liposomal membrane is maintained until the intraliposomal pH decreased to the extraliposomal level, which facilitates the uptake of SA into liposomes. The permeability of the lipid bilayer to several compounds was estimated from the inhibitory effects of those compounds on SA uptake by liposomes. Good linear relationships were observed between their inhibitory effects on the liposomal uptake of SA and the permeability of the intestinal membrane to them determined both in vivo and in vitro. These results clearly indicate that the carrier-independent transport mechanism of monocarboxylic acids observed in liposomes significantly contributes to their absorption from the intestinal tract under physiological conditions.
The lipid bilayer of biomembranes works as a barrier against the diffusion of low-lipophilic molecules into cells. In the field of biopharmacy, however, it is well known that some organic acids, such as SA and acetylsalicylic acid (aspirin), are absorbed quickly from the small intestine after oral administration, which indicates a rapid diffusion across the enterocytes. Because those acids are ionized almost completely in the intestinal tract, the pH-partition theory cannot explain their extensive absorption. Many reports have been presented to answer the question, “Why can they easily permeate the enterocyte?” They include effect of microclimate-pH at the surface of the cell membrane (Daniel et al., 1985) and contribution of the diffusive or convective flux through the paracellular pore route (Barnet et al., 1978; Jackson et al., 1981).
Recently, Tsuji and Tamai (1996) reported a series of studies which suggest the participation of the carrier-mediated transport system in the intestinal absorption of several monocarboxylic acids. They usedin vitro experimental techniques with the BBMVs prepared from enterocytes (Tsuji et al., 1990; Simanjuntak et al., 1990; Tamai et al., 1995a) as well as the cultured cell monolayers (Caco-2 monolayers) (Takanaga et al., 1994;Tsuji et al., 1994). In the presence of pH gradient, many kinds of monocarboxylic acids including benzoic acid, acetic acid and salicylic acid were taken up rapidly by the BBMVs showing the overshoot, concentration dependence and mutual inhibition phenomena. Moreover, they isolated the cDNA clone that encodes MCT1, a proton/monocarboxylate cotranspoter, from the rat small intestinal cDNA library and suggested that MCT1 recognizes and transports some monocarboxylic acids in the small intestine (Takanaga et al., 1995; Tamai et al., 1995b). Although theirin vitro and molecular identification studies were well organized and gave very clear results, the contribution of this carrier-mediated system to the in vivo intestinal absorption is still unclear. Also, a question arises about why the small intestine equips such an energy-dependent and broad substrate-recognizing transport system to absorb xenobiotics into the body. The presence of a similar transport system was suggested for the absorption of monocarboxylic acids from the oral cavity by means of the in vitro uptake experiment with primary cultured cells of rabbit oral mucosa (Utoguchi et al., 1997). In this case, the meaning of the carrier-mediated absorption of such compounds from the oral mucosain vivo, which consists of multilayers of flatter epithelial cells with partially keratinized, seems more difficult to understand.
In this report, we have determined the uptake of monocarboxylic acids by liposomes and showed almost the same results with those obtained in the BBMV study. Because the liposomal membrane used here is composed of simple lipids, our results suggest a new interpretation of the mechanisms of transport of those acids, not carrier-mediated, and give the evidence to explain their rapid absorption from the intestine.
Materials and Methods
Chemicals.
[14C]Salicylic acid (54 Ci/mol) was purchased from New England Nuclear (Boston, MA). Egg-PC was kindly donated by the Research Institute of Nippon Fine Chemical Co., Ltd. (Hyogo, Japan). Cholesterol was purchased from Sigma Chemicals (St. Louis, MO). Dulbecco’s modified Eagle’s medium, nonessential amino acids, fetal bovine serum, l-glutamate, trypsin (0.25%)-ethylenediaminetetraacetic acid (1 mM) and antibiotic-antimycotic mixture (10,000 U/ml penicillin G, 10,000 mg/ml streptomycin sulfate and 25 mg/ml amphotericin B in 0.85% saline) were purchased from Gibco Laboratories (Lenexa, KS). All other reagents used in this study were of the highest purity.
Preparation of liposomes.
MLV composed of egg-PC and cholesterol (1:1 as a molar ratio) were prepared in buffered solution containing 270 mM mannitol and 25 mM Tris-Hepes buffer (pH 7.5) or 25 mM Tris-Mes buffer (pH 5.8) according to the procedures described previously (Oku et al., 1980). The final concentration of liposomes in the suspension was adjusted to 6 mg/ml as egg-PC concentration.
Uptake of [14C]salicylic acid by liposomes.
The uptake of 14C-SA by liposomes was measured at 25°C by the rapid filtration method according to the procedures generally used in the uptake study with BBMVs. The uptake was initiated by adding 90 μl of the uptake medium containing14C-SA (11.11 μM) and 270 mM mannitol, 25 mM Tris-Hepes or Tris-Mes buffer (pH = 7.5 or 5.8) to 10 μl of liposome suspension. (Thus, the final concentration of14C-SA was 10 μM.) The uptake reaction was stopped at the selected time by adding 1 ml of ice-cold stop solution (25 mM Tris-Mes, 270 mM mannitol, pH = 7.5 or 5.8) and the mixture was filtrated immediately through a Millipore filter (HAWP, 0.45 μm, Millipore Corporation, Bedford, MA). Then the filter was washed twice with ice-cold stop solution and the radioactivity of14C-SA remaining on the filter was counted in a liquid scintillation counter. Nonspecific binding of the radioactivity to the filter was determined by filtering the reaction mixture without liposome. When investigating the inhibition of14C-SA uptake by other compounds, the defined concentrations of unlabeled compounds were added to the uptake medium before initiating the uptake reaction.
Permeability of the Caco-2 monolayer in vitro.
The Caco-2 cell line was purchased from American Type Culture Collection (Rockville, MD) at passage 17. Caco-2 monolayers were obtained by culturing Caco-2 cells in Transwell cell culture system (Costar, Cambridge, MA) by the same procedure as reported previously (Tanaka et al., 1995). The culture medium (1.5 ml in the insert and 2.6 ml in the well) was replaced every 2 to 3 days for the first 5 days and every day thereafter until the time of usage, which was between the 15th and 18th day after the seeding. In this series of studies, passage numbers of Caco-2 cells were between 34 and 54.
Transport of various compounds across the Caco-2 monolayer was studied in the Transwell system at 37°C. TM was composed of 136.89 mM NaCl, 5.36 mM KCl, 0.34 mM Na2HPO4, 0.44 mM KH2PO4, 0.41 mM MgSO4·7H2O, 19.45 mM glucose, 1.26 mM CaCl2, 0.49 mM MgCl2·6H2O, 4.17 mM NaHCO3 and 10 mM Hepes, and adjusted the pH to 5.8 for mucosal solution and to 7.5 for serosal solution. After preincubation of Caco-2 monolayers with TM (1.5 ml on the mucosal side and 2.6 ml on the serosal side) for 25 min, TM containing 0.1 mM of each compound was loaded onto the mucosal side. Thereafter, 100-μl samples were taken from the serosal side every 10 min for 60 min. The volume of the serosal solution was maintained constant by adding the fresh TM. The mucosal-to-serosal permeability of each compound was calculated from its flux rate estimated as a rate of increase in the serosal concentration.
Permeability of rat intestinal membrane in vivo.
The permeability of rat jejunum to each compound was evaluated by single-pass perfusion method as described by Hu et al. (1988). Male Wistar rats weighing 200 to 250 g were anesthetized with pentobarbital, then the abdominal cavity was opened and an intestinal loop (length, 10–15 cm) was made at the upper portion of the jejunum (beginning 2–4 cm below the ligament of Treitz) by cannulation with a silicone tube (internal diameter, 3 mm). After removing the intestinal contents by a slow infusion of saline and air, perfusion medium (140 mM NaCl, 5 mM KCl, 10 mM Hepes, adjusted to pH 6.5) containing each compound (0.1 mM) and FITC-dextran (MW 4000; 1 μM) was perfused with an infusion pump (Harvard Apparatus, model 944, S. Natick, MA) at a flow rate of 1.0 ml/min. The effluent was corrected from 30 min after starting the perfusion to 90 min at 10-min intervals, because steady-state absorption usually was achieved by 30 min under these conditions. The in vivo drug permeability was calculated from the ratio of outlet/inlet drug concentration where the effect of water transport during perfusion was corrected using the concentration ratio of a nonabsorbable marker (FITC-dextran) (Yamashitaet al., 1997).
Analytical methods.
In the study of the drug permeability of rat intestine and Caco-2 monolayer, the amount of drugs in each sample was determined by means of high-performance liquid chromatography (LC-6A Shimadzu Co., Kyoto, Japan) with a reversed phase column (Lichrospher 100 RP-18, Gaskuro-kogyo, Tokyo, Japan) equipped with a variable wave length ultraviolet detector (SPD-10A, Shimadzu Co., Kyoto Japan) or a fluorescence spectromonitor (RP-530, Shimadzu Co.). The analytical conditions for each drug were as follows. Salicylic acid: mobile phase, water containing 0.01 M citric acid/methanol (40:60 by volume); flow rate, 1.0 ml/min; wave length, 300 nm for excitation and 408 nm for emission. p-Hydroxybenzoic acid: mobile phase, water containing 0.01 M citric acid/methanol (70:30 by volume); flow rate, 1.0 ml/min; wave length (UV), 250 nm. m-Hydroxybenzoic acid: mobile phase, water containing 0.01 M citric acid/methanol (65:35 by volume); flow rate, 1.0 ml/min; wave length (UV), 230 nm. Benzoic acid: mobile phase, water containing 0.01 M citric acid/methanol (55:45 by volume); flow rate, 1.0 ml/min; wave length (UV), 230 nm. Nicotinic acid: mobile phase, water containing 0.05 M NaH2PO4 (pH = 3) with 5 mM Na-1-Octanesulfonate/methanol (85:15 by volume); flow rate, 0.8 ml/min; wave length (UV), 260 nm. Benzenesulfonic acid/mobile phase, water containing 0.1% NaH2PO4; flow rate, 1.0 ml/min; wave length (UV), 262 nm. In all analysis, the temperature of the column was kept at 45°C. The concentration of FITC-dextran was determined fluorometrically (495 nm for excitation and 514 nm for emission) with a spectrofluoro-photometer (RF-5300PC, Shimadzu Co., Kyoto Japan). The concentration of sulfanilic acid was estimated spectrophotometrically as described by Kimura et al. (1981).
Results
Characterization of liposomal uptake of SA.
Figure1 demonstrates the time course of liposomal uptake of 14C-SA measured under various conditions. When the pH gradient was present across the liposomal membrane which produces an inward proton gradient ([pH]out = 5.8 and [pH]in = 7.5), 14C-SA was taken up rapidly and the uptake profile showed overshoot. Only a small amount of14C-SA was taken up when both sides of the liposomal membrane were adjusted to the same pH {[pH]out = [pH]in = 7.5 or [pH]out = [pH]in = 5.8 (data not shown)}. Addition of the high concentration (1 mM) of unlabeled benzoic acid to the extraliposomal solution dramatically reduced the initial uptake of 14C-SA even in the presence of pH gradient. Because these findings are exactly the same to those reported as the carrier-mediated uptake by BBMVs, we undertook further studies to characterize the liposomal uptake of14C-SA.
The effect of pH on the uptake of 14C-SA was examined by changing the [pH]out across a range from 5.5 to 7.5 with a fixed [pH]in at 7.5. Figure 2 shows the initial uptake (30 sec) of 14C-SA. The uptake decreased with minimizing the pH gradient across the liposomal membrane, which suggests the importance of the inward proton gradient as a driving force of 14C-SA uptake.
The carrier-mediated uptake of drugs by BBMVs shows concentration and temperature dependence. Therefore, it is interesting to investigate the effect of drug concentration and incubation temperature on the uptake by liposomes. Figure 3 shows the initial uptake of SA at various concentrations. SA concentration in the extraliposomal solution was adjusted by adding unlabeled SA. The uptake of SA showed the typical pattern of the saturable process at the high SA concentration. The concentration-dependent uptake of SA can be described well with the Michaelis-Menten equation, giving aKm and Vmax as 0.34 mM and 2.94 nmol/30 sec, respectively. The effect of temperature also is shown in figure 3. The uptake of SA apparently was reduced at the lower temperature (4°C).
Effect of various compounds on liposomal uptake of SA.
Table1 summarizes the inhibitory effect of various compounds on the initial uptake of14C-SA. The presence of several monocarboxylic acid compounds (1 mM) significantly reduced the uptake of14C-SA. The degree of inhibition was the highest by benzoic acid, valproic acid and SA itself, and the lowest byl- and d-lactic acid among the monocarboxylic acid compounds. The structural analogs of SA,p-hydroxybenzoic acid and m-hydroxybenzoic acid showed the difference in the inhibitory effect. Dicarboxylic acid compound, phthalic acid, only slightly inhibited the uptake of14C-SA. In contrast, compounds not containing the carboxylic moiety, benzenesulfonic acid and sulfanilic acid, had no effect on SA uptake.
Relationship between the liposomal uptake and the intestinal transport.
The permeabilities of the intestinal membrane to several compounds were measured in vivo and in vitro and are summarized in table 2. As already reported, in vivo permeability of the intestinal membrane, measured by means of a single-pass perfusion method in rat small intestine, correlated well with in vitro permeability of the Caco-2 monolayer (Yamashita et al., 1997). When the intestinal permeability of these compounds is compared with their inhibitory effects on the liposomal uptake of14C-SA (table 1), it is obvious that the compound having a high permeability through the intestinal membrane shows a high extent of inhibition, and vice versa. As demonstrated in figures 4 and5, the inhibitory effect (represented as % inhibited) and the permeability of the intestinal membrane bothin vivo and in vitro showed good linear relationships with quite high regression coefficients (r = 0.969 in fig. 4 and 0.925 in fig. 5, respectively).
Discussion
“Overshoot uptake” or “competitive inhibition” is considered to be one of the criteria for the carrier-mediated transport. Because these phenomena were observed clearly in the uptake of some monocarboxylic acids by BBMVs, Tsuji et al. (1990)first suggested the presence of a specific transporter in their intestinal absorption. The driving force of this transport system is defined as a proton gradient across the membrane, because the presence of an inward proton gradient dramatically promoted uptake by BBMVs. Inin vivo conditions, the sodium/proton antiporter at the brush-border membrane could serve the proton gradient enough to facilitate the transport of these compounds (Shiau et al., 1985). Recently, a proton/monocarboxylic acid transporter, MCT1, which mediates transport of lactic acid and pyruvic acid, was isolated from Chinese hamster ovary cells (Garcia et al., 1994). By means of molecular characteristic techniques, Tamai et al. (1995b)found the expression of MCT1-related proteins in rat and rabbit intestinal epithelial cells as well as in Caco-2 cells, and suggested their contribution to the carrier-mediated absorption of monocarboxylic acid compounds.
However, our results presented here strongly indicate the necessity to reconsider the mechanisms of monocarboxylic acid transport. It is obvious that overshoot uptake, pH dependence, concentration dependence and also competitive inhibition cannot be sufficient proof for the carrier-mediated transport, because the liposomal membrane consists of only egg-PC and cholesterol, without any carrier proteins. A rather slow process of 14C-SA uptake by liposomes in figure 1, where a maximum uptake was observed at 5 min, might be caused by the multilamellar structure of liposomes used here. In fact, when the unilamellar liposomes were prepared instead of MLV, a sharper peak was observed within 1 min followed by a rapid decrease in the uptake amount of 14C-SA (data not shown).
The classical idea of simple passive diffusion could not explain these carrier-mediated like phenomena in SA uptake. We, therefore, propose a new interpretation of monocarboxylic acid transport across the lipid bilayer, which is driven by a proton gradient across the membrane. At pH 5.8, 0.14% of SA is expected to be protonated in the solution (pKa of SA = 3.0). Assuming that the protonated SA passes through the lipid membrane very quickly, it will redissociate to the ionized form after being taken up by liposomes according to the [pH]in (initial [pH]in = 7.5). The concentration gradient of protonated SA across the liposomal membrane is thus maintained until the [pH]in decreases to the extraliposomal level (pH = 5.8), which facilitates the uptake of SA into liposomes. Without pH gradient, SA initially taken up as a protonated form quickly reaches the equilibrium state with the extraliposomal SA leading to the very low level of SA uptake. As illustrated in figure 6, the most important point in this mechanism is that the uptake rate of the protonated SA must be faster than the inward diffusion of proton into liposomes. In such a situation, proton is supplied into liposomes mainly by the dissociation of SA, and the uptake of SA could last until the pH gradient is eliminated. This should be the reason for the overshoot uptake of SA. If the uptake rate is slower than that of proton, the pH gradient is eliminated by the inward flux of proton itself before the uptake of the substrate progresses.
At the low temperature, decreased fluidity of the lipid bilayer would reduce the uptake of SA. Although the phase-transition temperature of egg-PC is around −10°C, the lipid layer, existing as the liquid crystal state, becomes more rigid at 4°C, which markedly decreases the diffusion constant of SA in the membrane.
Following this concept, the concentration-dependent uptake of SA and also the competitive inhibition by related compounds can be understood as the consumption of the proton gradient. The addition of a higher concentration of inhibitors to the extraliposomal solution would cause the diffusion of much greater amount of protonated moieties into the liposomes. Then, the pH gradient is eliminated quickly by their dissociation, which results in the suppressed uptake of14C-SA. The compound which shows a higher inhibitory effect, therefore, is considered to permeate through the liposomal membrane at a higher rate. In other words, the degree of inhibition may represent the rate of uptake by the liposomes, and from the inhibitory effect of each compound, we can evaluate its permeability through the lipid membrane.
The inhibitory effects of several compounds on the liposomal uptake of14C-SA correlate well with their permeabilities through the intestinal membrane in vivo and in vitro (figs. 4 and 5). Because the vicinity of the luminal surface of enterocytes is kept acidic (pH = 5.5–6.0), the initial condition in our liposomal uptake experiments can give almost the same pH gradient as that under the normal conditions in the intestinein vivo. The proton gradient across the liposomal membrane is dissipated rapidly by the uptake of SA, which causes the overshoot phenomenon. In contrast, the proton gradient across the apical membrane of enterocytes is maintained constant by the proton/sodium antiporter, the secondary active transporter which is driven by the sodium gradient across the cell membrane generated by the energy-dependent active pump (sodium/potassium exchange pump at the basolateral membrane). The steady-state absorption of SA in vivo, therefore, could be facilitated by the proton gradient generated by such energy-dependent mechanisms. This should be the reason for the good linear relationship between the initial uptake by liposomes (estimated from inhibitory effects) and the steady-state permeability of the intestinal membranein vivo. In this respect, we suggest strongly that the mechanism of monocarboxylic acid transport proposed here significantly contributes to their rapid intestinal absorption. Without assuming any specific transporters, it is possible to explain the pH- and the concentration-dependent absorption of monocarboxylic acid compounds from the intestinal tract. If large amounts of monocarboxylic acids exist at the surface of enterocytes and permeate the cellular membrane quickly, their permeation might diminish the proton gradient temporally even in vivo (it might be the abnormal condition) which leads to the inhibition of SA absorption.
We do not suggest that the possibility of carrier-mediated absorption of monocarboxylic acid compounds should be abolished completely. Although the role of MCT1 in the intestine is not fully understood yet, it seems obvious that MCT1 transports some of monocarboxylic acids, such as lactic acid or pyruvic acid, in the intestine. Stereoselective transport of lactic acids observed in Caco-2 cells (Ogihara et al., 1996) and MCT1-expressed Xenopus laevis oocytes (Takanaga et al., 1995) might represent the contribution of the specific transporter in the absorption of these compounds, because in our liposome system, both l- and d-lactic acids failed to inhibit the uptake of SA significantly. To clarify the whole process of the monocarboxylic acid absorption, therefore, the quantitative evaluation on the contribution of these two different mechanisms to the intestinal transport must be undertaken in vivo.
In conclusion, we have demonstrated the pH-dependent but not carrier- mediated uptake of salicylic acid and some other monocarboxylic acids compounds by liposomes. The new interpretation proposed here changes the basic rules of weak electrolyte transport through the lipid bilayer and also describes the mechanisms of intestinal absorption of drugs having a monocarboxyl moiety.
Footnotes
-
Send reprint requests to: Shinji Yamashita, Ph.D., Faculty of Pharmaceutical Sciences, Setsunan University, 45–1 Nagaotoge-cho, Hirakata, Osaka 573–01, Japan.
- Abbreviations:
- SA
- salicylic acid
- 14C-SA
- radiolabeled salicylic acid
- BBMVs
- brush-border membrane vesicles
- MCT1
- proton/monocarboxylate cotransporter
- egg-PC
- egg yolk phosphatidylcholine
- MLV
- multilamellar large liposome
- TM
- transport medium
- [pH]out
- extraliposomal pH
- [pH]in
- intraliposomal pH
- Mes
- 2-morpholinoethanesulfonic acid
- Hepes
- N-(2-hydroxyethyl)piperazine-N′-ethanesulfonic acid
- FITC
- fluorescein isothiocyanate
- Received October 6, 1997.
- Accepted February 24, 1998.
- The American Society for Pharmacology and Experimental Therapeutics