BIOCHEMICAL
Vol. 175, No. 3, 1991 March 29, 1991
AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 880-885
CORRELATION BETWEEN ORAL DRUG ABSORPTION IN HUMANS AND APPARENT DRUG PERMEABILITY COEFFICIENTS IN HUMAN INTESTINAL EPITHELIAL (CACO-2) CELLS P. Artursson* and J. Karlsson Department of Pharmaceutics, Uppsala University, Box 580, S-751 23 Uppsala, Sweden Received
February
7,
1991
Monolayers of a well differentiated human intestinal epithelial cell line, Caco-2, were used as a model to study passive drug absorption across the intestinal epithelium. Absorption rate constants (expressed as apparent permeability coefficients) were determined for 20 drugs and peptides with different structural properties. The permeability coefficients ranged from approximately S x 10-S to 5 x 10-S cm/s. A good correlation was obtained between data on oral absorption in humans and the results in the Caco-2 model. Drugs that are completely absorbed in humans had permeability coefficients > 1 x 10-h cm/s. Drugs that are absorbed to > 1% but < 100% had permeability coefficients of 0.1-1.0 x 10-6 cm/s while drugs and peptides that are absorbed to < 1% had permeability coefficients of 11 x 10-7 cm/s. The results indicate that Caco-2 monolayers can be used as a model for studies on intestinal drug absorption. 0 1991 Academic Press, Inc. Almost all drugs that are given orally are absorbed across the intestinal mucosa by passive diffusion. In order to be absorbed, a drug has to diffuse across a series of separate barriers. These include, from the mucosal side, the mucus gel layer, the intestinal epithelial cells, the lamina propria and the endothelium of the capillaries. Among these, the single layer of epithelial cells is the most significant barrier to drug absorption (1). It should therefore be possible to use monolayers of intestinal epithelial cells to study passive drug absorption in viva This hypothesis was tested using a human intestinal epithelial cell line, Caco-2. The Caco-2 cell line is derived from a human colorectal carcinoma (2). It differs from other cell lines of the same origin in that it spontaneously differentiates into monolayers of polarized enterocytes under conventional cell culture conditions. After 2-3 weeks in cell culture, the monolayers express high levels of several brush border hydrolases and have well developed junctional complexes (3). The monolayers have a transepithelial electrical resistance of approximately 300 0 x cm2 which is similar to that found in the colon (4). In addition, several intestinal transport systems such as those for large neutral amino acids, bile acids, cobalamin and dipeptides have been identified (5-9). In this study the passive absorption of a number of different drugs and peptides was studied in Caco-2 monolayers grown on permeable supports and the results were compared with absorption data from humans. * To whom correspondence should be addressed. 0006-291X/91 Copyright Ail rights
$1.50
0 I991 by Academic Press, Inc. of reproduction in any form reserved.
880
Vol.
175, No. 3, 1991
BIOCHEMICAL
Materials
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
and Methods
Drags and radiolabelled markers. )3H\ - or 114Ci -1abelled and unlabelled
metoprolol, practolol, atenolol, alprenolol and felodipine were obtained from Dr Kurt-Jijrgen Hoffman, Hgssle AB, Sweden; olsalazine and sulphasalazine were obtained from Dr Peter Edman, Pharmacia AB, Sweden. PHI -1abelled and unlabelled argininevasopressin and 1-deamino-8-D-arginine-vasopressin were obtained from Dr Christina Ahlm, Ferring AB, Sweden and Dr Anna-Lena Ungell, Htissle AB, Sweden. /3H] terbutaline was obtained from Dr Lars BorgstrGm, AB Draco, Sweden. All other I3H I and 114CI-labelled compounds were from New England Nuclear, Boston, MA or Amersham, Arlington Heights, IL and all other unlabelled compounds were from Sigma Chemical Co, St Louis, MO. The radiolabelled compounds had a radiochemical purity of 97-99%.
Cells. Caco-2 cells were obtained from the American
Cell Culture Collection Rockville, MD. The cells were cultivated on polycarbonate filters (Nuclepore or Costar Transwell cell culture inserts; mean pore diameter 0.45 ctm) as described elsewhere (10). Cells of passage number 85-95 were used throughout. The cell culture procedure is comparable to that recently reported for cultivation on cellulose filters (6) but differs from that described for polycarbonate filters (ll), since the extracellular matrix components, such as collagen, have been excluded. The integrity of the monolayers was routinely checked by measurements of transepithelial electrical resistance and by determination of the permeability of hydrophilic marker 13Hl- mannitol(10). The transepithelial electrical resistance of 2- to 4-week-old monolayers was approximately 300 n x cm2. Drag absorption studies.All absorbtion studies were performed in Hank’s balanced salt
solution containing 25 mM N-2-Hydroxyethylpiperazine-N’-2-ethanesulfonic acid. The radiolabelled and unlabelled compounds were used at a final concentration of 1 x 10-4 or 1 x 10-5 M. The drug solutions were added to the apical side of the monolayers (2.0 ml). The monolayers were incubated in air at 37°C and 95% relative humidity. At regular intervals, samples were withdrawn from the basal (receiving) side. The radioactive samples were counted in a liquid scintillation counter. The resistance of the monolayers was checked at the end of each experiment, Apparent permeability coefficients (Papp) were calculated according to: PaPP=dt
dQ
1 AC,
where dQ/dt is the permeability rate, C, the initial concentration in the donor chamber and A the surface area of the monolayer (10). Results and Discussion Initial studies on the passive absorption of a homologous series of drugs across Caco-2 cells indicated that Caco-2 monolayers may be a useful model for drug absorption studies. A good correlation between drug absorption rate constants in the Caco-2 model and in a rat intestinal in situ model was obtained for a series of a-blocking agents (10). However, studies on narrow homologous series of compounds are of limited value since the good correlations may break down when structural diversity is introduced (12).
Therefore a larger number of compounds with different structural properties was used, Table 1. The lipophilicity of the drugs (expressed as octanol/water partition coefficients) spanned more than 8 orders of magnitude and the molecular weights differed by up to 29-fold. Uncharged as well as positively and negatively charged drugs 881
Vol.
175, No. 3, 1991
BIOCHEMICAL
AND BIOPHYSICAL
TABLE SUMMARY
Corticosterone Testosterone Propranolol Alprenonol W&al-in Metoprolol Felodipine Hydrocortisone Dexamethasone Salicylic acid
Acetylsalicylicacid Practolol Terbutaline Atenolol Mannitol AVP Sulphasalazine dDAVP Olsalazine PEG
OF STRUCTURAL CHARACTERISTICS
RESEARCH COMMUNICATIONS
1
PROPERTIES OF THE
AND ABSORPTION DRUGS
Pawa
1ogDb
% Abs. C
M.W.
Net charge d
54.5 51.8 41.9 40.5 38.3 27.0 22.1 21.5 12.5 11.9 2.4 0.90 0.38 0.20 0.18 0.14 0.13 0.13 0.11 0.052
1.89 3.31 1.54 1.00 0.12 0.07 3.48 1.53 1.74 -2.14 -2.57 -1.4 -1.4 -2.14 -3.10 -2.73 -0.13
100 100 90 93 98 95 loo 89 100 100 100 100 73 50 16 0 13
346 288 259 249 308 267 384 362 392 138 180 266 225 266 182 1084 398 1071 302 4000
0 0 +
-1.95 -4.5 -5.1
2 0
0
Abbreviations:AVP = arginine-vasopressin, dDAVP = 1-deamino-g-D-a&ninevasopressin, PEG= polythyleneglycol. 0 Apparentpermeabilitycoefficients(lO-6 cm/s).Thestandard deviationsweregenerallyless than 10% (n=3). b OctanoVwater partition coefficient at pH 7.4 (ref. 28-33 and personal communications with Dr. Kurt-Jiirgen Hoffman, Hlsde AB, Gtiteborg, Sweden; Dr. Peter Edman, Pharmacia AB, Uppsala, Sweden and Dr. Lars Borgstrtim, Draco AB, Lund,Sweden). c Percent absorbed of an orally administered dose (ref. 13-27). d Net charge at pH 7.4.
were studied. The absorption of the drugs after oral administration in humans varied from O-100%. All drugs were absorbed through the Caco-2 monolayers, Table 1. Thus, it was
possible to obtain permeability coefficients even for compounds that are not absorbed to a measurable extent in humans. The Papp values ranged from 54.5 x 10m6cm/s (corticosterone) to 0.052 x 10-b cm/s (polyethylene glycol4,OOO). The most common way to predict drug absorption is to determine the lipophilicity of the drug. The prediction relies on the assumption that more lipophilic drugs will partition faster into the lipid cell membranes. However, it is well known that lipophilicity (as defined by the pH-partition hypothesis) only gives an approximate indication of drug absorption (34). A coarse correlation between octanol/water partition coefficients and absorption rates was also found in this study, Fig. 1. 882
Vol.
175,
No.
3, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
l
l
l
e
a
.G
1o-8,
01
-6
-4
-2
0
2
1 4
log D
Figure 1. Apparent permeability coefficients in the Caco-2 model as a function of
lipophilicity. Figure 2. Absorption in humans after oral administration as a function of apparent
permeability coefficients in the Caco-2 model. A better correlation was obtained when percent absorbed drug after oral administration in humans was expressed as a function of the Papp values, Fig. 2. Drugs that are completely absorbed in humans had Papp values of approximately > 1 x 10-6 cm/s. Drugs that are absorbed to < 100% had Papp values of < 1 x 10-6 cm/s. Many incompletely absorbed compounds are not absorbed across the cell membranes, i.e. they are excluded from the transcellular pathway, Instead, they are absorbed by the alternative paracellular pathway (across the tight junctions between the cells). The paracellular pathways in colonic epithelium are tighter than in small intestinal epithelium. The mean aqueous pore radius of human jejunum is 0.8 nm while the corresponding value for the colon is 0.3 nm (4). The electrical (ionic) permeability (which is an alternative measure of paracellular permeability) is higher in the small intestine than in the large intestine and Caco-2 monolayers (4). Therefore, many incompletely absorbed compounds are excluded from the paracellular pathway in the colon but not in the small intestine. Among these are several of the slowly absorbed compounds used in this study, including atenolol, polyethylene glycol and l-deamino-8D-arginine-vasopressin (12,27,35). The steep decrease in extent of drug absorption (from 100 to approximately 1%) for drugs with Papp values of 0.1-l x 10-6 cm/s probably reflects the differences in drug absorption profiles between the small and the large intestines, Fig. 2. Results showing a good correlation between drug absorption in humans and in rat small intestine have recently been published for a small number of compounds (22). The correlation presented in this study is at least as good as that found with rat intestine. However, it should be noted that Caco-2 is a transformed cell line with properties from different parts of the intestinal epithelium (36). Further, the absorption experiments were performed under conditions were drug metabolism could be excluded. Thus, arginine-vasopressin which is known to be extensively metabolized by pancreatic and brush border enzymes in viva is not metabolized by the Caco-2 cells (32). Moreover, it is not yet known if the size exclusion profile of the tight junctions of Caco-2 cells is similar to or different from that of normal colonic epithelium. In summary, many incompletely absorbed compounds are excluded from the transcellular route. Permeation through the alternative paracellular route is usually 883
Vol.
175,
No.
3, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
possible, mainly in the small intestine. The paracellular pathway is tighter and therefore more discriminating in colonic epithelia including Caco-2 monolayers. This indicates that the Caco-2 model can be used to differentiate drugs that are not absorbed in the large intestine. Such information will be valuble, for example, in the design of oral drug delivery systems. The results also suggest that Caco-2 monolayers are suitable for screening purposes. We have recently used the Caco-2 model to predict drug absorption after chemical modification of different drugs (37). Acknowledgments This work was supported by grants from The Swedish Medical Research Council (B9104X-09478OlA), The Swedish Fund for Scientific Research without Animal Experiments, and Centrala FiirsBksdjursn&nnden (FN L90-04). We thank MS Christina Stabeck for technical assistance. References 1. Jackson, M.J. ,(1987) In Physiology of the Gastrointestinal Tract (L.R. Johnson, Ed.), 2nd Edition, pp. 1597-1621. Raven Press, New York. 2. Fogh, J., Fogh, J.M. and Orfeo, T. (1977) J. Natl. Cancer Inst. 59,221-226. 3. Pinto, M., Robine-Leon, S., Appay, M.-D., Kedinger, M., Triadou, N., Dussaulx, E., Lacroix, B., Simon-Assmann, P., Haffen, K., Fogh, J. and Zweibaum, A. (1983) Biol. Cell 47,323-330. 4. Powell, D.W. (1987) In Physiology of the Gastrointestinal Tract (L.R. Johnson, Ed.), 2nd Edition, pp. 1267-1305. Raven Press, New York. 5. Hidalgo, I.J. and Borchardt, T. (1990) Biochim. Biophys. Acta 1028,25-30. 6. Wilson, G., Hassan, I.F., Dix, C.J., Williamson, I., Shah, R., Mackay, M. and Artursson, P. (1990) J. Controlled Release 11,25-40. 7. Hidalgo, I.J. and Borchardt, T. (1990) Biochim. Biophys. Acta 1035,97-103. 8. Dix, C.J., Hassan, I.F., Obray, H.Y., Shah, R. and Wilson, G. (1990) Gastroenterology 98, 1272-1279. 9. Dantzig, A.H. and Bergin, L. (1990) Biochim. Biophys. Acta 1027,211-217. 10. Artursson, P. (1990) J. Pharm. Sci. 79,476-482. 11. Hidalgo, I.J., Raub, T.J. and Borchardt, R.T. (1989) Gastroenterology 96,736-749. 12. Taylor, D.C., Lynch, J. and Leahy, D.E. (1990) In Drug delivery to the gastrointestinal tract (J.G. Hardy, S.S. Davis. and C.G. Wilson, Eds.), pp. 133-145. Ellis Horwood Ltd, Chichester, U.K. 13. Dressman, J.B., Arnidon, G.L. and Fleicher, D. (1985) J. Pharm. Sci. 74,588-589. 14. Harvey, S.C. and Withrow, C.D. (1985) In Remington’s Pharmaceutical Sciences (A.R. Germaro, Ed.), 17th Edition, pp. 951-1001. Mack Publishing Company, Easton, PA. 15. Ruffalo, R.L., Garabedian-Ruffalo, S.M. and Garret, B.N. (1986) Cardiovasc. Rev. Rep. 7,692-702. 16. Johansson, R., Regardh, C.G. and SjBgren, J. (1971) Acta Pharm. Suet. 8,59-70. 17. Holford, N.H.G. (1986) Clin. Pharmacokinet. 11,483-504. 18. Edgar, B., Reg%rdh, C.G., Johnsson, G., Johansson, L., Lundborg, P., Liifberg, I. and Rii~, 0. (1985) Clin. Pharmacol. Ther. 38,205-211. 19. Mlynaryk, P. and Kirsner, J.B. (1963) Gastroenterology 44,257-260. 20. Lui, C.Y. Oberle, R., Fleischer, D. and Amidon, G.L. (1986) J. Pharm. Sci. 75,469474. 21. Rowland, M. and Riegelman, S. (1968) J. Pharm. Sci. 57, 1313-1319. 22. Amidon, G.L., Sinko, P.J. and Fleischer, D. (1988) Pharm. Res. 5,651-654. 23. Bodem, G. and Chidsey, C.A. (1974) Clin. Pharmacol. ‘I’her. 14,26-29. 884
Vol.
175, No. 3, 1991
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
24. Borgstrom, L., Nyberg, L., Jonsson, S., Lindberg, C. and Paulson, J. (1989) Br. J. Clin, Pharmac. 27,49-56. 25. Ukabam, SO., Clamp, J.R. and Cooper, B.T. (1983) Digestion 27,70-74. 26. Ryde, M. (1988) Acta Universitatis Upsaliensis, Comprehensive Summaries of Uppsala Dissertations from the Faculty of Pharmacy 39. 27. Olaison, G., Sjodahl, R., Leandersson, P. and Tagesson, C. (1989) Stand. J. Gastroenterol. 24,571-576. 28. Komiya, I., Park, J.Y., Kamani, A., Ho, N.F.H. and Highuchi, W.I. (1980) Int. J. Pharm. 4,249-262. 29. Tavaloki-Saberi, M.R. and Audus, K.L. (1989) Int. J. Pharm. 56,135-142. 30. Hansch, L.A. and Elkins, D. (1971) Chem. Rev. 21,525-616. 31. Grass, G.M. and Sweetana, S.A. (1988) Pharm. Res. 5,372-376. 32. Lundin, S. and Artursson, P. (1990) Int. J. Pharm. 64, 181-186. 33. Hollander, D., Koyama, S., Dadufalza, V., Quang Tran, D., Krugliak, P., Ma, T. and Ling. K.-Y. (1989) J. Lab. Clin. Med. 113,505-515. 34. Martin, Y.C. (1981) J. Med. Chem. 24,229-237. 35. Lundin, S. and Vilhardt, H. (1986) Acta Endocrinol. 112,457-460. 36. Neutra, M. and Louvard, D. (1989) In Modern Cell Biology (K.S. Matlin and J.D. Valentich, Eds.), Vol. 8, pp. 363-398. Alan R. Liss, New York. 37. Artursson, P. (unpublished results).
885
Vol. 175, No. 3, 1991 March 29, 1991
AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 880-885
CORRELATION BETWEEN ORAL DRUG ABSORPTION IN HUMANS AND APPARENT DRUG PERMEABILITY COEFFICIENTS IN HUMAN INTESTINAL EPITHELIAL (CACO-2) CELLS P. Artursson* and J. Karlsson Department of Pharmaceutics, Uppsala University, Box 580, S-751 23 Uppsala, Sweden Received
February
7,
1991
Monolayers of a well differentiated human intestinal epithelial cell line, Caco-2, were used as a model to study passive drug absorption across the intestinal epithelium. Absorption rate constants (expressed as apparent permeability coefficients) were determined for 20 drugs and peptides with different structural properties. The permeability coefficients ranged from approximately S x 10-S to 5 x 10-S cm/s. A good correlation was obtained between data on oral absorption in humans and the results in the Caco-2 model. Drugs that are completely absorbed in humans had permeability coefficients > 1 x 10-h cm/s. Drugs that are absorbed to > 1% but < 100% had permeability coefficients of 0.1-1.0 x 10-6 cm/s while drugs and peptides that are absorbed to < 1% had permeability coefficients of 11 x 10-7 cm/s. The results indicate that Caco-2 monolayers can be used as a model for studies on intestinal drug absorption. 0 1991 Academic Press, Inc. Almost all drugs that are given orally are absorbed across the intestinal mucosa by passive diffusion. In order to be absorbed, a drug has to diffuse across a series of separate barriers. These include, from the mucosal side, the mucus gel layer, the intestinal epithelial cells, the lamina propria and the endothelium of the capillaries. Among these, the single layer of epithelial cells is the most significant barrier to drug absorption (1). It should therefore be possible to use monolayers of intestinal epithelial cells to study passive drug absorption in viva This hypothesis was tested using a human intestinal epithelial cell line, Caco-2. The Caco-2 cell line is derived from a human colorectal carcinoma (2). It differs from other cell lines of the same origin in that it spontaneously differentiates into monolayers of polarized enterocytes under conventional cell culture conditions. After 2-3 weeks in cell culture, the monolayers express high levels of several brush border hydrolases and have well developed junctional complexes (3). The monolayers have a transepithelial electrical resistance of approximately 300 0 x cm2 which is similar to that found in the colon (4). In addition, several intestinal transport systems such as those for large neutral amino acids, bile acids, cobalamin and dipeptides have been identified (5-9). In this study the passive absorption of a number of different drugs and peptides was studied in Caco-2 monolayers grown on permeable supports and the results were compared with absorption data from humans. * To whom correspondence should be addressed. 0006-291X/91 Copyright Ail rights
$1.50
0 I991 by Academic Press, Inc. of reproduction in any form reserved.
880
Vol.
175, No. 3, 1991
BIOCHEMICAL
Materials
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
and Methods
Drags and radiolabelled markers. )3H\ - or 114Ci -1abelled and unlabelled
metoprolol, practolol, atenolol, alprenolol and felodipine were obtained from Dr Kurt-Jijrgen Hoffman, Hgssle AB, Sweden; olsalazine and sulphasalazine were obtained from Dr Peter Edman, Pharmacia AB, Sweden. PHI -1abelled and unlabelled argininevasopressin and 1-deamino-8-D-arginine-vasopressin were obtained from Dr Christina Ahlm, Ferring AB, Sweden and Dr Anna-Lena Ungell, Htissle AB, Sweden. /3H] terbutaline was obtained from Dr Lars BorgstrGm, AB Draco, Sweden. All other I3H I and 114CI-labelled compounds were from New England Nuclear, Boston, MA or Amersham, Arlington Heights, IL and all other unlabelled compounds were from Sigma Chemical Co, St Louis, MO. The radiolabelled compounds had a radiochemical purity of 97-99%.
Cells. Caco-2 cells were obtained from the American
Cell Culture Collection Rockville, MD. The cells were cultivated on polycarbonate filters (Nuclepore or Costar Transwell cell culture inserts; mean pore diameter 0.45 ctm) as described elsewhere (10). Cells of passage number 85-95 were used throughout. The cell culture procedure is comparable to that recently reported for cultivation on cellulose filters (6) but differs from that described for polycarbonate filters (ll), since the extracellular matrix components, such as collagen, have been excluded. The integrity of the monolayers was routinely checked by measurements of transepithelial electrical resistance and by determination of the permeability of hydrophilic marker 13Hl- mannitol(10). The transepithelial electrical resistance of 2- to 4-week-old monolayers was approximately 300 n x cm2. Drag absorption studies.All absorbtion studies were performed in Hank’s balanced salt
solution containing 25 mM N-2-Hydroxyethylpiperazine-N’-2-ethanesulfonic acid. The radiolabelled and unlabelled compounds were used at a final concentration of 1 x 10-4 or 1 x 10-5 M. The drug solutions were added to the apical side of the monolayers (2.0 ml). The monolayers were incubated in air at 37°C and 95% relative humidity. At regular intervals, samples were withdrawn from the basal (receiving) side. The radioactive samples were counted in a liquid scintillation counter. The resistance of the monolayers was checked at the end of each experiment, Apparent permeability coefficients (Papp) were calculated according to: PaPP=dt
dQ
1 AC,
where dQ/dt is the permeability rate, C, the initial concentration in the donor chamber and A the surface area of the monolayer (10). Results and Discussion Initial studies on the passive absorption of a homologous series of drugs across Caco-2 cells indicated that Caco-2 monolayers may be a useful model for drug absorption studies. A good correlation between drug absorption rate constants in the Caco-2 model and in a rat intestinal in situ model was obtained for a series of a-blocking agents (10). However, studies on narrow homologous series of compounds are of limited value since the good correlations may break down when structural diversity is introduced (12).
Therefore a larger number of compounds with different structural properties was used, Table 1. The lipophilicity of the drugs (expressed as octanol/water partition coefficients) spanned more than 8 orders of magnitude and the molecular weights differed by up to 29-fold. Uncharged as well as positively and negatively charged drugs 881
Vol.
175, No. 3, 1991
BIOCHEMICAL
AND BIOPHYSICAL
TABLE SUMMARY
Corticosterone Testosterone Propranolol Alprenonol W&al-in Metoprolol Felodipine Hydrocortisone Dexamethasone Salicylic acid
Acetylsalicylicacid Practolol Terbutaline Atenolol Mannitol AVP Sulphasalazine dDAVP Olsalazine PEG
OF STRUCTURAL CHARACTERISTICS
RESEARCH COMMUNICATIONS
1
PROPERTIES OF THE
AND ABSORPTION DRUGS
Pawa
1ogDb
% Abs. C
M.W.
Net charge d
54.5 51.8 41.9 40.5 38.3 27.0 22.1 21.5 12.5 11.9 2.4 0.90 0.38 0.20 0.18 0.14 0.13 0.13 0.11 0.052
1.89 3.31 1.54 1.00 0.12 0.07 3.48 1.53 1.74 -2.14 -2.57 -1.4 -1.4 -2.14 -3.10 -2.73 -0.13
100 100 90 93 98 95 loo 89 100 100 100 100 73 50 16 0 13
346 288 259 249 308 267 384 362 392 138 180 266 225 266 182 1084 398 1071 302 4000
0 0 +
-1.95 -4.5 -5.1
2 0
0
Abbreviations:AVP = arginine-vasopressin, dDAVP = 1-deamino-g-D-a&ninevasopressin, PEG= polythyleneglycol. 0 Apparentpermeabilitycoefficients(lO-6 cm/s).Thestandard deviationsweregenerallyless than 10% (n=3). b OctanoVwater partition coefficient at pH 7.4 (ref. 28-33 and personal communications with Dr. Kurt-Jiirgen Hoffman, Hlsde AB, Gtiteborg, Sweden; Dr. Peter Edman, Pharmacia AB, Uppsala, Sweden and Dr. Lars Borgstrtim, Draco AB, Lund,Sweden). c Percent absorbed of an orally administered dose (ref. 13-27). d Net charge at pH 7.4.
were studied. The absorption of the drugs after oral administration in humans varied from O-100%. All drugs were absorbed through the Caco-2 monolayers, Table 1. Thus, it was
possible to obtain permeability coefficients even for compounds that are not absorbed to a measurable extent in humans. The Papp values ranged from 54.5 x 10m6cm/s (corticosterone) to 0.052 x 10-b cm/s (polyethylene glycol4,OOO). The most common way to predict drug absorption is to determine the lipophilicity of the drug. The prediction relies on the assumption that more lipophilic drugs will partition faster into the lipid cell membranes. However, it is well known that lipophilicity (as defined by the pH-partition hypothesis) only gives an approximate indication of drug absorption (34). A coarse correlation between octanol/water partition coefficients and absorption rates was also found in this study, Fig. 1. 882
Vol.
175,
No.
3, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
l
l
l
e
a
.G
1o-8,
01
-6
-4
-2
0
2
1 4
log D
Figure 1. Apparent permeability coefficients in the Caco-2 model as a function of
lipophilicity. Figure 2. Absorption in humans after oral administration as a function of apparent
permeability coefficients in the Caco-2 model. A better correlation was obtained when percent absorbed drug after oral administration in humans was expressed as a function of the Papp values, Fig. 2. Drugs that are completely absorbed in humans had Papp values of approximately > 1 x 10-6 cm/s. Drugs that are absorbed to < 100% had Papp values of < 1 x 10-6 cm/s. Many incompletely absorbed compounds are not absorbed across the cell membranes, i.e. they are excluded from the transcellular pathway, Instead, they are absorbed by the alternative paracellular pathway (across the tight junctions between the cells). The paracellular pathways in colonic epithelium are tighter than in small intestinal epithelium. The mean aqueous pore radius of human jejunum is 0.8 nm while the corresponding value for the colon is 0.3 nm (4). The electrical (ionic) permeability (which is an alternative measure of paracellular permeability) is higher in the small intestine than in the large intestine and Caco-2 monolayers (4). Therefore, many incompletely absorbed compounds are excluded from the paracellular pathway in the colon but not in the small intestine. Among these are several of the slowly absorbed compounds used in this study, including atenolol, polyethylene glycol and l-deamino-8D-arginine-vasopressin (12,27,35). The steep decrease in extent of drug absorption (from 100 to approximately 1%) for drugs with Papp values of 0.1-l x 10-6 cm/s probably reflects the differences in drug absorption profiles between the small and the large intestines, Fig. 2. Results showing a good correlation between drug absorption in humans and in rat small intestine have recently been published for a small number of compounds (22). The correlation presented in this study is at least as good as that found with rat intestine. However, it should be noted that Caco-2 is a transformed cell line with properties from different parts of the intestinal epithelium (36). Further, the absorption experiments were performed under conditions were drug metabolism could be excluded. Thus, arginine-vasopressin which is known to be extensively metabolized by pancreatic and brush border enzymes in viva is not metabolized by the Caco-2 cells (32). Moreover, it is not yet known if the size exclusion profile of the tight junctions of Caco-2 cells is similar to or different from that of normal colonic epithelium. In summary, many incompletely absorbed compounds are excluded from the transcellular route. Permeation through the alternative paracellular route is usually 883
Vol.
175,
No.
3, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
possible, mainly in the small intestine. The paracellular pathway is tighter and therefore more discriminating in colonic epithelia including Caco-2 monolayers. This indicates that the Caco-2 model can be used to differentiate drugs that are not absorbed in the large intestine. Such information will be valuble, for example, in the design of oral drug delivery systems. The results also suggest that Caco-2 monolayers are suitable for screening purposes. We have recently used the Caco-2 model to predict drug absorption after chemical modification of different drugs (37). Acknowledgments This work was supported by grants from The Swedish Medical Research Council (B9104X-09478OlA), The Swedish Fund for Scientific Research without Animal Experiments, and Centrala FiirsBksdjursn&nnden (FN L90-04). We thank MS Christina Stabeck for technical assistance. References 1. Jackson, M.J. ,(1987) In Physiology of the Gastrointestinal Tract (L.R. Johnson, Ed.), 2nd Edition, pp. 1597-1621. Raven Press, New York. 2. Fogh, J., Fogh, J.M. and Orfeo, T. (1977) J. Natl. Cancer Inst. 59,221-226. 3. Pinto, M., Robine-Leon, S., Appay, M.-D., Kedinger, M., Triadou, N., Dussaulx, E., Lacroix, B., Simon-Assmann, P., Haffen, K., Fogh, J. and Zweibaum, A. (1983) Biol. Cell 47,323-330. 4. Powell, D.W. (1987) In Physiology of the Gastrointestinal Tract (L.R. Johnson, Ed.), 2nd Edition, pp. 1267-1305. Raven Press, New York. 5. Hidalgo, I.J. and Borchardt, T. (1990) Biochim. Biophys. Acta 1028,25-30. 6. Wilson, G., Hassan, I.F., Dix, C.J., Williamson, I., Shah, R., Mackay, M. and Artursson, P. (1990) J. Controlled Release 11,25-40. 7. Hidalgo, I.J. and Borchardt, T. (1990) Biochim. Biophys. Acta 1035,97-103. 8. Dix, C.J., Hassan, I.F., Obray, H.Y., Shah, R. and Wilson, G. (1990) Gastroenterology 98, 1272-1279. 9. Dantzig, A.H. and Bergin, L. (1990) Biochim. Biophys. Acta 1027,211-217. 10. Artursson, P. (1990) J. Pharm. Sci. 79,476-482. 11. Hidalgo, I.J., Raub, T.J. and Borchardt, R.T. (1989) Gastroenterology 96,736-749. 12. Taylor, D.C., Lynch, J. and Leahy, D.E. (1990) In Drug delivery to the gastrointestinal tract (J.G. Hardy, S.S. Davis. and C.G. Wilson, Eds.), pp. 133-145. Ellis Horwood Ltd, Chichester, U.K. 13. Dressman, J.B., Arnidon, G.L. and Fleicher, D. (1985) J. Pharm. Sci. 74,588-589. 14. Harvey, S.C. and Withrow, C.D. (1985) In Remington’s Pharmaceutical Sciences (A.R. Germaro, Ed.), 17th Edition, pp. 951-1001. Mack Publishing Company, Easton, PA. 15. Ruffalo, R.L., Garabedian-Ruffalo, S.M. and Garret, B.N. (1986) Cardiovasc. Rev. Rep. 7,692-702. 16. Johansson, R., Regardh, C.G. and SjBgren, J. (1971) Acta Pharm. Suet. 8,59-70. 17. Holford, N.H.G. (1986) Clin. Pharmacokinet. 11,483-504. 18. Edgar, B., Reg%rdh, C.G., Johnsson, G., Johansson, L., Lundborg, P., Liifberg, I. and Rii~, 0. (1985) Clin. Pharmacol. Ther. 38,205-211. 19. Mlynaryk, P. and Kirsner, J.B. (1963) Gastroenterology 44,257-260. 20. Lui, C.Y. Oberle, R., Fleischer, D. and Amidon, G.L. (1986) J. Pharm. Sci. 75,469474. 21. Rowland, M. and Riegelman, S. (1968) J. Pharm. Sci. 57, 1313-1319. 22. Amidon, G.L., Sinko, P.J. and Fleischer, D. (1988) Pharm. Res. 5,651-654. 23. Bodem, G. and Chidsey, C.A. (1974) Clin. Pharmacol. ‘I’her. 14,26-29. 884
Vol.
175, No. 3, 1991
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
24. Borgstrom, L., Nyberg, L., Jonsson, S., Lindberg, C. and Paulson, J. (1989) Br. J. Clin, Pharmac. 27,49-56. 25. Ukabam, SO., Clamp, J.R. and Cooper, B.T. (1983) Digestion 27,70-74. 26. Ryde, M. (1988) Acta Universitatis Upsaliensis, Comprehensive Summaries of Uppsala Dissertations from the Faculty of Pharmacy 39. 27. Olaison, G., Sjodahl, R., Leandersson, P. and Tagesson, C. (1989) Stand. J. Gastroenterol. 24,571-576. 28. Komiya, I., Park, J.Y., Kamani, A., Ho, N.F.H. and Highuchi, W.I. (1980) Int. J. Pharm. 4,249-262. 29. Tavaloki-Saberi, M.R. and Audus, K.L. (1989) Int. J. Pharm. 56,135-142. 30. Hansch, L.A. and Elkins, D. (1971) Chem. Rev. 21,525-616. 31. Grass, G.M. and Sweetana, S.A. (1988) Pharm. Res. 5,372-376. 32. Lundin, S. and Artursson, P. (1990) Int. J. Pharm. 64, 181-186. 33. Hollander, D., Koyama, S., Dadufalza, V., Quang Tran, D., Krugliak, P., Ma, T. and Ling. K.-Y. (1989) J. Lab. Clin. Med. 113,505-515. 34. Martin, Y.C. (1981) J. Med. Chem. 24,229-237. 35. Lundin, S. and Vilhardt, H. (1986) Acta Endocrinol. 112,457-460. 36. Neutra, M. and Louvard, D. (1989) In Modern Cell Biology (K.S. Matlin and J.D. Valentich, Eds.), Vol. 8, pp. 363-398. Alan R. Liss, New York. 37. Artursson, P. (unpublished results).
885
