Physiological mechanism for enhancement of paracellular drug transport

Journal of Controlled Release 62 (1999) 141–148 www.elsevier.com / locate / jconrel

Physiological mechanism for enhancement of paracellular drug transport a, a a b a M. Hayashi *, T. Sakai , Y. Hasegawa , T. Nishikawahara , H. Tomioka , A. Iida a , N. Shimizu b , M. Tomita b , S. Awazu b a

Department of Pharmaceutics, Faculty of Pharmaceutical Sciences, Science University of Tokyo, Shinjuku-ku, Tokyo 162 -0826, Japan b Department of Biopharmaceutics, School of Pharmacy, Tokyo University of Pharmacy and Life Science, Hachioji, Tokyo 192 -0392, Japan

Abstract We examined the action mechanisms of enhancers that improve paracellular drug transport. For sodium caprate (C10), the increase in the intracellular calcium level was considered to induce the contraction of calmodulin-dependent actin filaments, followed by dilation of the paracellular pathway. Although decanoylcarnitine (DC) also increased the intracellular calcium level, the action was independent of calmodulin and thus, the action mechanism of acylcarnitines was considered to differ from that of C10. Other acylcarnitines, lauroylcarnitine (LC) and palmitoylcarnitine (PC) and organic acids, tartaric acid (TA) and citric acid (CA) decreased the intracellular ATP level and the intracellular pH. From these results, it was considered that one of the action mechanism of acylcarnitines and organic acids is that the intracellular acidosis increases the calcium level through the decrease in ATP levels, followed by opening the tight junction. Membrane dysfunction which was expected from the above mechanism was assessed by the transport function of electrolytes. Membrane conductance, which was increased by C10, LC and PC, returned to the control value during a 3- to 6-h recovery period. On the other hand, Cl 2 ion secretion, which was obtained from short-circuit current (Isc ), was decreased by these enhancers, but was normalized by C10 but not by LC and PC. Accordingly, C10 can be considered a safer enhancer than acylcarnitines.  1999 Elsevier Science B.V. All rights reserved. Keywords: Paracellular pathway; Absorption enhancer; Intracellular ATP; Intracellular acidosis; Cl 2 ion secretion

1. Introduction The paracellular pathway is a useful absorption pathway for many water-soluble and poorly lipidsoluble drugs, ionized drugs and high-molecularweight compounds [1]. Since the contribution of the paracellular pathway to the intestinal absorption is generally considered small due to the presence of a tight junction between the epithelial cells, the ab*Corresponding author.

sorption of the above drugs is poor. To increase the paracellular transport of less-absorbable drugs, the use of enhancers has been attempted. However, since it is difficult to determine whether the action mechanism of an enhancer is physiological or pathological, the practical use of the enhancers is limited. If the action is physiological, the membrane dysfunction induced by an enhancer would recover to the original condition. If the action is pathological, the action may induce membrane damage and thus make the use of the enhancer ill-advised.

0168-3659 / 99 / $ – see front matter  1999 Elsevier Science B.V. All rights reserved. PII: S0168-3659( 99 )00031-0

142

M. Hayashi et al. / Journal of Controlled Release 62 (1999) 141 – 148

In this in vitro study, the action mechanisms of enhancers that improve paracellular drug transport were examined in the rodent colon. As the enhancers, sodium caprate (C10) [2–4], acylcarnitines such as decanoylcarnitine (DC) [3,4], lauroylcarnitine (LC) and palmitoylcarnitine (PC) [5], and organic acids such as tartaric acid and citric acid were used. Their action mechanisms were compared, and physiological vs. pathological mechanisms are contrasted. Recovery from the membrane dysfunction induced by C10, PC and LC was examined by assessing the transport function of electrolytes in order to evaluate the practical use of enhancers.

2. Materials and methods

2.1. Materials Sodium caprate (C10), decanoylcarnitine chloride (DC), DL-lauroyl-carnitine chloride (LC), palmitoylcarnitine chloride (PC), amiloride, tetraethylammonium chloride, bumetanide, betanechol chloride, forskolin, fluoresceinisothiocyanate dextran 4000 (FD-4), and ATP disodium salt were obtained from Sigma (St. Louis, MO, USA). Tartaric acid (TA) and citric acid (CA) were purchased from Kanto (Tokyo, Japan). 39-O-acetyl-29,79-bis(carboxyethyl)-4 or 5-carboxyfluorescein diacetoxymethyl ester (BCECF-AM) was purchsed from CalbiochemNovabiochem (La Jolla, CA, USA).

2.2. Determination of physiological parameters Male Wistar rats (200–250 g) and male golden hamsters (130–170 g) were used after they were fasted overnight. The colon segment was removed and set to the Ussing-type chamber by the method described in the previons report [2–4]. The physiological parameters, membrane resistance (R m ), membrane conductance (Gt ), and short-circuit current (Isc ) of the rat or hamster colon, were determined in the Ussing-type chamber [2]. For the examination of the transport function of electrolytes in the epithelial cells, the absorption function of Na 1 ion and secretory function of K 1 and Cl 2 ions were assessed from the change of Isc (DIsc ) using the respective inhibitors [6,7]. First, with the addition of 0.1 mM

amiloride on the mucosal side of the colonic membrane in the chamber, the absorptive transport of Na 1 ions is inhibited, and thus, the decrease in Isc can be regarded as an index of Na 1 ion absorption [6]. Second, 10 mM tetraethylammonium chloride is added to the mucosal side, and the secretory transport of K 1 ions is inhibited; thus the increase in Isc can be regarded as an index of K 1 ion secretion [6]. Finally, with the addition of 0.1 mM bumetanide to the serosal side of the colonic membrane in the chamber, all ion transport, i.e. all ion fluxes based on Na 1 ion absorption, K 1 ion secretion, and Cl 2 ion secretion are inhibited, resulting in a decrease in Isc [7]. Finally, the assessment of secretory dysfunction of Cl 2 ions is possible using the Na 1 and K 1 ion data.

2.3. Preparation of epithelial cells of the hamster colon The preparation of isolated colonic epithelial cells of hamsters was conducted according to the method of Brasitus [8]. The hamster colon was used since the yield of the epithelial cells is higher and the viability is maintained at a higher level compared to the rat colon.

2.4. Determination of intracellular ATP level and pH ( pHi ) Intracellular ATP levels were determined by a HPLC method [9] (mobile phase: 0.5 M ammonium dihydrogen phosphate solution, pH 5.5; column: SSC-ODS-262 (Senshu Kagaku, Tokyo, Japan; flowrate: 1.0 ml / min; detector: UV 254 nm). The intracellular pH (pH i ) was determined in the epithelial cell membrane of the hamster colon using BCECFAM [10], which penetrates into the cells and is hydrolyzed by the esterase to emit fluorescence corresponding to pH i . The fluorescence ratio was obtained for two excitation wavelengths (439 and 490 nm) and at 526 nm for emission.

2.5. Assessment of transport function of electrolytes after the enhancer-treatment First, after the injection of the enhancer solution into the rat colon, the colon was left as it was for 30

M. Hayashi et al. / Journal of Controlled Release 62 (1999) 141 – 148

min on the operation board. The solution was then washed out, and the colon was allowed to recover. At 3, 6 and 24 h after the start of this recovery, the Gt and Isc of the colon were measured in the Ussingtype chamber. The transport functions of Na 1 , K 1 and Cl 2 ions were also determined at the above time point.

143

nism of the absorption-enhancing effect of C10 is shown in Fig. 1. C10 increases the intracellular calcium levels through interaction with phospholipase C in the membrane, and enhances the permeability by opening the tight junction through activation of calmodulin-dependent contraction of the actin microfilament by the released calcium.

2.6. Response of Cl 2 ion secretory function to an activating agent after C10 -treatment

3.2. Enhancing mechanism of acylcarnitines

Following the addition 0.1 mM betanechol chloride, a protein kinase C activating agent [11], or 0.01 mM forskolin, an acylcyclase activating agent [12], to the serosal side of the colon which was let to stand for 24 h after the C10-treatment, in the Ussing-type chamber, DIsc was obtained and compared with that for the control colon.

Regarding the action of decanoylcarnitine, DC, the increases in the intracellular calcium level and the CL of FD-4 were also already reported; they are smaller than those in the case of C10 [3,4]. The effects of DC were not inhibited but rather were increased by both W7 and compound 48 / 80. H7 did not show any effect. The results suggest that DC operates by a mechanism different from that of C10.

2.7. Assay of FD-4 and its permeation clearance The assay of the paracellular marker FD-4 and the calculation of the permeation clearance (CL) of FD-4 followed the method described in previous reports [3,4].

2.8. Statistical analysis Statistical comparisons were made by Student’s t-test. Values of P,0.05 were taken as significant.

3. Results and discussion

3.1. Enhancing mechanism of C10 The action mechanism of C10 was already described by our research group [3,4] and Lindmark and coworkers [13,14]. C10 increased the intracellular calcium level and enhanced the CL of FD-4. The C10 effect was inhibited by W7 (a myosin light chain kinase inhibitor), N-(6-aminohexyl)-5-chloro1-naphthalenesulfonamide hydrochloride, and by a phospholipase C inhibitor, compound 48 / 80. No inhibition was achieved by H7 (a proteinkinase C inhibitor), 1-(5-isoquinolinesulfonyl)-2-methylpiperazide dihydrochloride. A proposed detailed mecha-

3.3. Effect of LC and PC on membrane resistance, permeability and intracellular ATP levels It has been reported that the enhancing effect of PC relates to the decrease in the intracellular ATP level [5]. Thus, the effects of 0.5% LC and PC on the R m and CL of FD-4 in the rat colon were examined in detail and are shown in Fig. 2. LC and PC reduced the R m and increased the CL. Both compounds decreased the ATP levels in the hamster colonic epithelial cells, as shown in Fig. 2. Those results suggested that acylcarnitines open the tight junction (the gate of the paracellular pathway) by inducing a reduction in the intracellular ATP level.

3.4. Effects of TA and CA on R m , CL and intracellular ATP levels In Fig. 3, the effects of the organic acids TA and CA on R m , CL and the intracellular ATP levels are shown. Both TA and CA decreased the R m and increased the CL, and these effects were greater at pH 3 than at pH 4. Hydrogen chloride (HCl) had similar effects, though these were less than those of the organic acids, probably due to a buffering effect in the microclimate region on the membrane surface. The intracellular ATP levels were significantly reduced by TA, and the change was almost in-

144

M. Hayashi et al. / Journal of Controlled Release 62 (1999) 141 – 148

Fig. 1. Proposed absorption-enhancing mechanism of C10. PIP2 : phosphatidylinositol-4,5-diphosphate; PLC: phospholipase C; DAG: diacylglycerol; IP3 : inositol-1,4,5-triphosphate; CaM: calmodulin; MLCK: myosin light chain kinase; My: myosin; Ac: actin.

Fig. 2. Effects of 0.5% lauroylcarnitine (LC) and 0.5% palmitoylcarnitine (PC) on the membrane resistance (R m ), the permeation clearance (CL) of FD-4 and the intracellular ATP level in hamster colon. Each datum represents the mean6S.E. at least five experiments. **P,0.01, *0.01,P,0.05 vs. the control.

M. Hayashi et al. / Journal of Controlled Release 62 (1999) 141 – 148

145

Fig. 3. Effects of tartaric acid (TA), citric acid (CA) and HCl at pH 3 and pH 4 on the R m , the CL of FD-4 in rat colon and the intracellular ATP level in hamster colon. Each datum represents the mean6S.E. of at least five experiments. **P,0.01, *0.01,P,0.05 vs. the control.

dependent of the pH of the TA solution. It is likely that the pH-dependent effect of TA on the R m and the CL of FD-4 is due to the uptake of TA from the apical side, but the pH-independent effect on the intracellular ATP levels is due to the use of epithelial cells, where the polarity of the apical and basolateral sides cannot be differentiated. Regarding the effects of acylcarnitines and organic acids, similar reductions in the intracellular ATP level were found. Here, the pH of the 0.5% PC and LC solution which had an absorption-enhancing action was about 3 (Fig. 2), almost equal to the pH of the organic acid solution. Accordingly, the lowmedium pH condition is thought to be related to the decrease in the ATP level.

effects of acylcarnitine on pH i are also examined, but the decrease in pH i could be predicted from the very similar decrease in the intracellular ATP levels and the low-medium pH.

3.5. Effect of TA on pHi Compared to the control pH i , which was 7.2–7.4, TA decreased pH i by 0.5 at pH 4 and by 0.6 at pH 3 (Fig. 4). This phenomenon is called intracellular acidosis. Accordingly, the result suggested that organic acids and acylcarnitines permeate the cell membrane at the low-medium pH and release protons inside the cells. This leads to the intracellular acidosis and finally opens the tight junction. The

Fig. 4. Effects of tartaric acid (TA) at pH 3 and pH 4 on intracellular pH (pH i ) in hamster colon. Each datum represents the mean6S.E. of at least four experiments. *P,0.01 vs. the control.

146

M. Hayashi et al. / Journal of Controlled Release 62 (1999) 141 – 148

3.6. Proposed enhancing mechanism of organic acids and acylcarnitines The two phenomena of intracellular acidosis and reduction of the intracellular ATP level can be combined physiologically [15]. The enhancing mechanism of organic acids and acylcarnitines is summarized in Fig. 5. Intracellular acidosis increases the calcium level directly or through the ATP depletion. The increase in the calcium level activates the actomyosin contraction by the activation of cytoskeletal destabilization or through other processes leading to the opening of the tight junction. The activation of phospholipase which was found in the action mechanism of C10 can be also included among these processes. The difference in the action mechanisms of C10 and that of organic acids and acylcarnitines is that the C10-effect is dependent on calmodulin and myosin light chain kinase, whereas the effect of the latter enhancers is not.

3.7. Reversibility of the effects of C10 and acylcarnitine by membrane conductance, Gt The ratios of Gt immediately (0 h) and 3 and 6 h

Fig. 6. Effects of 0.5% C10, 0.5% PC and 0.5% LC on membrane conductance (Gt ) in the rat colon. Each datum represents the ratio of Gt value just (0 h), 3 and 6 h after the treatment with an enhancer to the control ratio without enhancer treatment and is the mean6S.E. of at least eight experiments.

after the treatments with 0.5% C10, PC and LC compared to the control ratio (treatment with enhancers) are shown in Fig. 6. The use of each of these enhancers produced a significant increase in the ratio just after the treatment. However, the ratios decreased as the recovery time proceeded, and finally were not significantly different from the control ratio. Accordingly, the membrane barrier dysfunction induced by the enhancers can return to the normal

Fig. 5. Proposed absorption-enhancing mechanism of organic acids. PKC: protein kinase C.

M. Hayashi et al. / Journal of Controlled Release 62 (1999) 141 – 148

147

condition, at least regarding the membrane conductance.

3.8. Reversibility of the C10, PC and LC effects by Isc and DIsc The ratios of Isc and DIsc after the treatment with enhancers to the control are shown in Fig. 7. The Isc was decreased by all of the enhancers. The decrease induced by C10 was smaller than that by acylcarnitine. The Isc in the C10-treated colons did not return to the control level by 6 h of recovery, but at 24 h, the values almost returned to the control level. Recovery from PC treatment and that from LC treatment was not observed at 6 h. These results suggested that it takes more time for the recovery of membrane barrier dysfunction induced by PC or LC than by C10. As shown in Fig. 7, the DIsc produced by C10, PC and LC, which is an index for the secretory function of Cl 2 ions, is shown as the ratio to the control values. The decrease in the ratio of DIsc induced by C10 was smaller than those for PC and LC. The recovery patterns of DIsc , i.e., the secretory function of Cl 2 ions were very similar to those of Isc . Accordingly, it was shown that the action of the enhancers correlated well with the secretory function of Cl 2 ions. It is well known that the secretion of Cl 2 ions is related to the occurrence of diarrhea [16], and thus, as one of the pathological effects of enhancers, diarrhea is thought to be induced by

Fig. 8. Effects of 0.1 mM betanechol chloride (A) and 0.1 mM forskolin (B) on DIsc as an index of the transport function of Cl 2 ions in the rat colon. Each datum represents the absolute Isc value in the rats 24 h after the treatment with an enhancer and in the control preparation, and is the mean6S.E. of at least four experiments.

membrane damage. The secretory function of Cl 2 ions reduced by C10 returned to the normal level faster than that reduced by PC or LC, suggesting that C10 is a safer enhancer than PC and LC.

3.9. Response of the colon to activating agents of Cl 2 ion secretion 24 h after C10 -treatment The response of the colon to two activators of Cl 2 ion secretion, betanechol and forskolin, which induce an increase in protein kinase C and acylcyclase activities, respectively, was examined 24 h after C10 treatment. In both cases, the Cl 2 ion secretory function was enhanced to almost the same degree as was the control (Fig. 8), suggesting that the transport function of Cl 2 ions returned to almost the control level within the 24-h recovery time.

4. Conclusions

Fig. 7. Effects of 0.5% C10, 0.5% PC and 0.5% LC on Isc (A) and DIsc (B) in the rat colon. Each datum represents the ratio of Isc and the ratio of DIsc to the respective control, as shown in Fig. 6, and is the mean6S.E. of at least eight experiments.

For the effective and safe use of absorption enhancers, the following physiological examinations are first necessary: (1) a clarification of the enhancing mechanism physiologically; (2) the prediction of the epithelial barrier dysfunction induced by the enhancing action; (3) an examination of the recovery from the membrane damage by a physiological assessment and (4) an assessment of the potential for the practical use of the enhancer. With these step-by-

148

M. Hayashi et al. / Journal of Controlled Release 62 (1999) 141 – 148

step examinations, the clinical use of an enhancer can be considered. [8]

Acknowledgements [9]

This work was supported in part by a Grant-in Aid for Scientific Research provided by the Ministry of Education, Science, Sports and Culture of Japan (No. 06672280).

References [1] M. Hayashi, M. Tomita, S. Awzu, Transcellular and paracellular contribution to transport processes in the colorectal route, Adv. Drug Deliv. Rev. 28 (1997) 191–204. [2] T. Sawada, T. Ogawa, M. Tomita, M. Hayashi, S. Awazu, Role of paracellular pathway in nonelectrolyte permeation across rat colon epithelium enhanced by sodium caprate and sodium caprylate, Pharm. Res. 8 (1991) 1365–1371. [3] M. Tomita, M. Hayashi, S. Awazu, Absorption-enhancing mechanism of sodium caprate and decanoylcarnitine in Caco-2 cells, J. Pharmacol. Exp. Ther. 272 (1995) 739–741. [4] M. Tomita, M. Hayashi, S. Awazu, Absorption-enhancing mechanism of EDTA, caprate and decanoylcarnitine in Caco-2 cells, J. Pharm. Sci. 85 (1996) 608–611. [5] J.H. Hochman, J.A. Fix, E.L. LeCluyse, In vitro and in vivo analysis of the mechanism of absorption enhancement by palmitoylcarnitine, J. Pharmacol. Exp. Ther. 269 (1994) 813–822. [6] S.J. Downing, S.D. Maguiness, A. Watson, H.J. Leese, Electrophysiological basis of human Fallopian tubal fluid formation, J. Reprod. Fertil. 111 (1997) 29–34. [7] I.M. Brzuszczak, J. Zhao, C. Bell, D. Stiel, I. Fielding, J.

[10]

[11]

[12]

[13]

[14]

[15]

[16]

Percy, R. Smith, E.V. OLoughlin, Cyclic AMP-dependent anion secretion in human small and large intestine, J. Gastroenterol. Hepatol. 11 (1996) 804–810. T.A. Brasitus, Lipid dynamics and protein-lipid interactions in rat colonic epithelial cell basolateral membrane, Biochim. Biophys. Acta 728 (1983) 20–30. M.W. Taylor, H.V. Hershey, R.A. Levine, K. Coy, S. Olivelle, Improved method of resolving nucleosides by reversed-phase high-performance liquid chromatography, J. Chromatogr. 219 (1981) 133–139. G. Bishcof, E. Cosentini, G. Hamilton, M. Riegler, J. Zacherl, B. Tereky, W. Feil, R. Schiessel, T.E. Machen, E. Wenzl, Effects of extracellular pH on intracellular pH regulation and growth in a human colon carcinoma cell-line, Biochim. Biophys. Acta 1282 (1996) 131–139. S.A. Przyborski, R.J. Levine, Cholinergic modulation of electrogenic ion transport in different regions of the rat small intestine, J. Pharm. Pharmacol. 49 (1997) 691–697. P.B. Bijlsma, A.J. Kiliaan, G. Scholten, M. Heyman, J.A. Groot, J.A.J.M. Taminiau, Carbachol, but not forskolin, increases mucosal-to-serosal transport of intact protein in rat ileum in vitro, Am. J. Physiol. 271 (1996) G147–G155. T. Lindmark, T. Nikkila, P. Artursson, Mechanism of absorption enhancement by medium chain fatty acids in intestinal epithelial Caco-2 cell monolayers, J. Pharmacol. Exp. Ther. 275 (1995) 958–964. T. Lindmark, Y. Kimura, P. Artursson, Absorption enhancement through intracellular regulation of tight junction permeability by medium chain fatty acids in Caco-2 cells, J. Pharmacol. Exp. Ther. 284 (1998) 362–369. M.J. Menconi, A.L. Salzman, N. Unno, R.M. Ezzell, D.M. Casey, D.A. Brown, Y. Tsuji, M.P. Fink, Acidosis induces hyperpermeability in Caco-2BBe cultured intestinal epithelial monolayers, Am. J. Physiol. 272 (1997) G1007–G1021. J. Tripuraneni, A. Koutsouris, L. Pestic, P.D. Lanerolle, G. Hecht, The toxin of diarrheic shellfish poisoning, okadaic acid, increases intestinal epithelial paracellular permeability, Gastroenterology 112 (1997) 100–108.