Absorption sites of orally administered drugs in the small intestine

Introduction: In pharmacotherapy, drugs are mostly taken orally to be absorbed systemically from the small intestine, and some drugs are known to have preferential absorption sites in the small intestine. It would therefore be valuable to know the absorption sites of orally administered drugs and the influencing factors.

Areas covered:In this review, the author summarizes the reported absorption sites of orally administered drugs, as well as, influencing factors and experimental techniques. Information on the main absorption sites and influencing factors can help to develop ideal drug delivery systems and more effective pharmacotherapies.

Expert opinion: Various factors including: the solubility, lipophilicity, luminal concentration, pKa value, transporter substrate specificity, transporter expression, luminal fluid pH, gastrointestinal transit time, and intestinal metabolism determine the site-dependent intestinal absorption. However, most of the dissolved fraction of orally administered drugs including substrates for ABC and SLC transporters, except for some weakly basic drugs with higher pKa values, are considered to be absorbed sequentially from the proximal small intestine. Securing the solubility and stability of drugs prior to reaching to the main absorption sites and appropriate delivery rates of drugs at absorption sites are important goals for achieving effective pharmacotherapy.     

Griseofulvin absorption from different sites in the human small intestine

The site-dependent small-intestinal absorption pattern of griseofulvin was investigated in man. Griseofulvin was chosen as a model substance having extremely low water solubility and moderate lipid solubility. A conventional steady-state perfusion technique (triple-lumen tubing system with a 20 cm test segment) was applied. Dissolved griseofulvin (10.0 mg L^?1) was perfused (10 mL min^?1) during 160 min into different parts of the small intestine with the middle of the test segment between 85 cm and 270 cm beyond the teeth. Each of the ten healthy volunteers was examined twice with the test segment localized in different regions to allow for intraindividual comparisons. Mean drug absorption rates calculated from intestinal aspirate concentrations were similar in the two intestinal parts (proximal, 15.0 ± 5.9 μg (20cm min)^?1; distal, 16.2 ± 4.3 μg (20 cm min)^?1; mean ± SD). Absorption rate was strongly correlated to the amount of griseofulvin offered to the test segment per unit time. Extrapolating these findings it follows that an amount of griseofulvin, once dissolved, would be absorbed completely (>99%) along 100 cm of the small intestine. A significant, positive correlation between the rate of transmucosal fluid transport and the absorption rate of griseofulvin was observed in the distal parts investigated.     

Paracetamol absorption from different sites in the human small intestine.

Site-specificity in the small intestinal absorption of paracetamol was investigated using a segmental intestinal steady state perfusion technique (triple-lumen tubing system) combined with simultaneous measurements of serum drug concentrations. Dissolved paracetamol was perfused over 160 min into different parts of the small intestine (65-210 cm beyond the teeth). Each of the four healthy subjects was studied twice with a proximal and a more distal site of perfusion. Serum drug concentrations were similar after proximal and distal perfusions. Mean drug absorption rates calculated from intestinal aspirate concentrations were similar in both parts of the intestine--proximal: 869 micrograms 30 cm-1 min-1 (95% CI: 659-1079) vs distal: 941 micrograms 30 cm-1 min-1 (794-1088). The absorption rate was related directly to the amount of paracetamol perfused per unit time as well as to the rate of transmucosal water fluxes.     

Characterization of ethiofos absorption in the rat small intestine.

The absorption characteristics of ethiofos were studied using the rat in situ intestine circulating perfusion technique. Slow absorption kinetics were observed for ethiofos with varying rates of absorption and metabolism/degradation in situ as a function of buffer and absorption enhancers. In most cases less than 10 per cent of the radiolabeled compound is lost from the circulating perfusate in 90 min. In addition, over the same time period greater than 40 per cent of the intact parent compound was lost by degradation. Much of the difference can be accounted for in the formation of the free thiol metabolite. WR-1065, suggesting ester hydrolysis or metabolic activity. Good stability was observed in all perfusate systems ex vivo indicating that the degradation occurs in situ. The disodium salt of ethylenediaminetetraacetic acid (EDTA) was shown to be an effective absorption enhancer of ethiofos. The enhancement of intestinal absorption by EDTA was dose-dependent resulting in a 20-fold increase in blood levels of ethiofos in the portal blood. Follow-up studies in the rhesus monkey confirm this observation. Salicylate and dimethylsulfoxide (DMSO) also resulted in absorption enhancement although to a lesser degree than that seen after EDTA treatment. Addition of several alkaline phosphatase inhibitors did not significantly improve absorption of ethiofos in the rat small intestine. Proposed mechanism(s) for intestinal absorption and absorption enhancement of ethiofos are discussed.     

Characterization of WR-1065 absorption in the rat small intestine

A circulating in situ rat small intestine absorption model was used to study the lumenal metabolism and absorption of [14C]WR-1065. WR-1065 was found to be more tissue reactive and toxic than its phosphorylated form, ethiofos, at equimolar perfusate concentrations. The disappearance profiles of the radiolabeled drug and free WR-1065 indicate that WR-1065 is extensively metabolized in the intestinal lumen prior to absorption. Coadministration of disodium ethylenediaminetetraacetic acid enhances the absorption of the free thiol although not to the same extent as seen with ethiofos. Perfusion of WR-1065 in citrate buffer decreased lumenal degradation of the drug but resulted in decreased absorption. The total material converted to WR-1065 portal blood profiles following ethiofos and WR-1065 perfusion were altered possibly due to distribution and metabolism differences. This study coupled with earlier work completed on ethiofos have increased our understanding of the significant barriers to absorption observed following oral administration of these compounds.     

Characterization and solubilization of the specific binding sites for d-α-tocopherol from human erythrocyte membranes

Previous work from our laboratory has demonstrated the presence of specific binding sites for d-α-tocopherol (vitamin E) in intact human erythrocytes [A. E. Kitabchi and J. Wimalasena, Biochim. biophys. Acta684, 300 (1982)]. The binding was time, temperature and cell concentration dependent. To localize the binding sites, red blood cells were further fractionated; greater than 90% of the tocopherol binding sites were localized on membranes. The washed membrane fraction from normal human erythrocytes has specific binding sites for d-α-tocopherol with properties suggestive of protein receptors. Two binding sites with K_a values of 3.31 × 10⁷ M^?1 and 1.51 × 10⁶ M^?1 were demonstrated, and solubilized d-α-tocopherol binding site complexes were resolved to major component with an M_r of 65,000 and a minor component with an M_r of 125,000.     

Misoprostol attenuates acetic acid-induced damage in rabbit distal small intestine.

The influence of misoprostol pretreatment (100 micrograms/loop intraluminally) on small intestinal damage induced by acetic acid was evaluated in anesthetized rabbits. In this model injury was induced by intraluminal administration into loops of distal small intestine, of a solution of calcium gluconate (50 mg/ml) and acetic (200 mM). After 3 hr damage was associated with increase in loop fluid volume, loop fluid protein levels and epithelial permeability to 51Cr-EDTA, all of which were attenuated by misoprostol pretreatment. Similar protective effects were noted 45 min after the insult, suggesting that misoprostol may be therapeutically useful in conditions where the epithelial barrier is compromised.     

Presence of specific platelet-activating factor binding sites in the rat retina

The specificity of the effects of platelet-activating factor (PAF) on the electrophysiology of the retina suggests the existence of specific sites for PAF in this tissue. In this study, we report the presence of tritiated PAF ([3H]PAF) specific binding sites in membrane preparations of the retina of albino rats. The binding of [3H]PAF was saturable, specific, time-dependent and reversible. Scatchard analysis of the data revealed that the high-affinity retinal binding site possessed a Kd of 2.9 +/- 0.4 nM and Bmax of 0.85 +/- 0.16 pmol/mg protein. These values are comparable with those found for the membranous PAF receptor sites in platelets, neutrophils, lung tissue and brain. We have recently reported that PAF dose dependently modulates the b-wave of the electroretinogram (ERG) obtained from the isolated rat retina. The results of the present study suggest that such PAF-induced disturbances of the ERG may be mediated via specific receptors located in the retina.     

Characterization of specific binding sites for [3H](d)-N-allylnormetazocine in rat brain membranes

Binding of [3H](d)-N-allylnormetazocine ([3H](d)-NANM) to rat brain membranes is stereospecific, reversible, and saturable (Bmax = 260 fmol/mg of protein) and manifests moderately high affinity (Kd = 20 nM). The rank order of potency among opioidbenzomorphans and phencyclidine (PCP) analogs for competition for [3H](d)-NANM-binding sites is as follows: (d)-NANM = PCP-3-OH greater than (d)-cyclazocine greater than N-ethylphenylcyclohexylamine greater than PCP greater than (l)-cyclazocine = dextrorphan greater than (d/l)-ethylketocyclazocine greater than (d/l)-bremazocine greater than (1)-NANM greater than 1-phenylcyclohexylamine greater than levorphanol. Other opioid ligands, relatively selective for each of the types of opioid binding sites other than sigma, such as morphine (mu), H-Tyr-D-Ala(Me)Phe-NH-CH2-OH (mu), D-Ala2-D-Leu5-enkephalin (delta), tifluadom (kappa), and U 50488 (kappa) as well as etorphine and naloxone were all unable to compete with [3H](d)-NANM for specific binding even at a concentration of 1 microM. Regional distribution studies of [3H](d)-NANM-binding sites show high density in the hippocampus, thalamus, hypothalamus, and amygdala and low density in cerebellum and nonfrontal neocortex membranes of the rat brain. These binding sites are very sensitive to protein-modifying enzymes and reagents such as trypsin and N-ethylmaleimide and to heat denaturation. These results provide direct biochemical evidence for the existence of distinct (d)-NANM-binding sites in rat brain. In addition, this study supports the view that PCP and several of its analogues and the dextrorotatory isomers of psychotomimetic benzomorphans may act at a common recognition site in rat central nervous system.     

Partial characterization of specific cantharidin binding sites in mouse tissues

The mode of action of cantharidin, the natural vesicant of blister beetles, is examined by radioligand binding studies with mouse tissues. [3H]Cantharidin undergoes specific and saturable binding with the liver cytosol, which is characterized as follows: Kd and Bmax values of 30 nM and 1.8 pmol/mg of protein, respectively; linearity with respect to protein concentration; pH optimum of 6.5 to 7.5; association and dissociation half-times of 20 min and 12 hr, respectively; and 50% inhibition by Mg2+ at 70 microM, Ca2+ at 224 microM, pyrophosphate at 27 microM, and nucleotide triphosphates at 52-81 microM. The binding site undergoes a loss of activity at 45 degrees or higher. The toxicological relevance of this specific [3H]cantharidin binding site of mouse liver cytosol is established in three ways. First, the potency of 15 active cantharidin analogs for inhibiting [3H]cantharidin binding is correlated with their acute toxicity to mice (r = 0.829). Second, 26 related compounds that are inactive in inhibiting [3H]cantharidin binding are also of little or no toxicity to mice. Finally, the binding of [3H] cantharidin to liver cytosol from mice poisoned with increasing amounts of unlabeled cantharidin is inhibited in a dose-dependent manner. [3H]Cantharidin also specifically binds to cytosol fractions of blood, brain, heart, kidney, lung, pancreas, skin, spleen, and stomach. The characteristics of the specific binding site in brain are very similar to those determined in liver with respect to Kd, Bmax, association/dissociation kinetics, and sensitivity to inhibitors. It therefore appears that the toxicity of cantharidin and related oxabicycloheptanes, including the herbicide endothal, is attributable to binding at a specific site in liver and possibly other tissues.     

Mediation of the cardiac effects of forskolin by specific binding sites

The effects of forskolin and seven derivatives on cardiac functions were investigated by using the Langendorff technique and the results compared with the respective potencies obtained from adenylate cyclase and binding studies. In the isolated heart, forskolin increased all parameters measured in the same concentration range as demonstrated by identical EC50 values for increasing contractile force (270 nmol/L), heart rate (276 nmol/L), and coronary flow (249 nmol/L). Compared with its analogs, forskolin was the most potent agonist followed by 7-desacetylforskolin-7-ethylcarbonate, 7-desacetyl-7-propionylforskolin, 14,15-dihydroforskolin, and 7-desacetylforskolin. An identical order of potencies was obtained when these compounds were tested for their ability to inhibit [3H]forskolin binding and stimulate adenylate cyclase in a myocardial preparation. While the KI values derived from the binding experiments (386 nmol/L for forskolin) were similar to the respective EC50 values for the cardiac stimulatory effects, the ED50 values for adenylate cyclase stimulation were about 30-fold higher, e.g., 10 mumol/L for forskolin. Derivatives modified in position 9 of the molecule (1,9-dideoxyforskolin, 9,14-epoxy-15-hydroxyforskolin, and 7-desacetylforskolin-6,7:1,9-dicarbonate) had no effect on the isolated heart nor in adenylate cyclase or binding studies. This identical structure-activity profile observed in the three systems used suggests that the cardiac effects of forskolin are elicited by the one specific binding site described in this study.     

Effects and specific binding sites of endothelin in human lung preparations

Endothelin binding in human isolated lung membrane fractions revealed a single class of high affinity recognition sites with a K_d of 1.33±0.15 nM and a B_max of 9.61±1.44 pmol/mg protein. Endothelin inhibited [¹²⁵I]endothelin binding with a K_i of 1.90±0.15 nM whereas structurally unrelated compounds had no effect. Endothelin was a potent contractile agonist on human isolated pulmonary arterial (HPA) and venous (HPV) muscle preparations (pD₂ values: 9.64 and 10.36, respectively). Neither indomethacin (1 μM), nicardipine (0.01, 0.10, 1.0 μM) nor diltiazem (1, 10, 100 μM) altered the sensitivity of HPA to endothelin. Human isolated bronchial muscle preparations were less sensitive to endothelin than vascular tissues. These data suggest that pulmonary veins may be a major target for this constrictory peptide in the human lung.     

Characterization of peripheral benzodiazepine binding sites in human term placenta

Peripheral benzodiazepine binding sites were characterized in human term placental membranes using [3H]PK 11195, which is a ligand specific for peripheral benzodiazepine binding sites. Binding of [3H]PK 11195 to human term placental membranes was found to be saturable. Scatchard analysis revealed a single population of binding sites (r = 0.98). Equilibrium dissociation constant (KD) was 2.1 +/- 0.3 nM, and density of binding sites (Bmax) was 920 +/- 105 fmol/mg protein. The KD value calculated from kinetic experiments was 3.6 +/- 0.2 nM. The ability of various drugs to displace [3H]PK 11195 from human term placental binding sites was tested: the inhibition constants (KI) for PK 11195, Ro 5-4864, and diazepam were 2.9, 11.8, and 177 nM, respectively, whereas clonazepam, methyl-beta-carboline-3-carboxylate, Ro 15-1788, chlordiazepoxide, atropine, and estradiol were inefficient in displacing [3H]PK 11195 (KI greater than 10(-5) M).     

Characterization of absorption process of taurine across rat small intestine

A mechanism of taurine transfer across the rat small intestine was elucidated by using thein situ recirculation perfusion or loop method. Taurine uptake was saturable, Km=39.9 mM, and energy dependent, and required sodium. The close structural analogues, aminomethane sulfonic acid, γ-aminobutyric acid, hypotaurine, and β-alanine, reduced significantly taurine uptake when present in 10-fold excess. the α-amino acid, glycine, did not inhibit uptake. Hence, all of these findings lead to a conclusion that a carrier-mediated transport system for taurine exists in the small intestine.     

Identification and characterisation of heterogeneous somatostatin binding sites in rat distal colonic mucosa

We have previously shown that the somatostatin (SRIF) sst₂ receptor-selective peptide, BIM-23027, is a potent antisecretory agent in rat isolated distal colonic mucosa (RDCM) and in radioligand binding studies in RDCM membranes, it only maximally inhibited approximately 40% of [¹²⁵I]-Tyr¹¹-SRIF-14 binding (McKeen ES, Feniuk W, Humphrey PPA (1995) Naunyn-Schmiedeberg's Arch Pharmacol 352:402–411). The aim of this study was to characterise the BIM-23027-sensitive and -insensitive SRIF binding sites in more detail and to compare their properties with those of the recombinant sst₂ receptor stably expressed in mouse fibroblast (Ltk^–) cells.SRIF-14, SRIF-28, CGP-23996 and D Trp⁸-SRIF-14 abolished [¹²⁵I]-Tyr¹¹-SRIF-14 binding (pIC₅₀ values, 8.7–9.7) but the competition curves had Hill slopes which were less than unity. Octreotide and L-362,855 inhibited binding over a wide concentration range (0.1 nM-1 M) and inhibition of binding was incomplete at the highest concentration studied. BIM-23056 (PIC₅₀ <6.5) was a weak inhibitor of [¹²⁵]-Tyr¹¹-SRIF-14 binding. GTPS decreased [¹²⁵I]-Tyr¹¹-SRIF-14 binding by 40%. Further binding experiments with [¹²⁵I]-Tyr¹¹-SRIF-14 were carried out in RDCM in the continuous presence of BIM-23027 (1 M). Under these conditions, seglitide had no effect on [¹²⁵I]-Tyr¹¹-SRIF-14 binding at concentrations up to 10 M, whilst SRIF-14 and SRIF-28 abolished specific [¹²⁵I]-Tyr¹¹-SRIF-14 binding in a manner which was consistent with the ligand binding to two sites. SRIF-14 and SRIF-28 displayed high affinity (pIC₅₀ values of 9.8 and 9.3 respectively) for approximately 70% of these binding sites and low affinity (pIC₅₀ values of 7.8 and 7.3) for the remaining sites. Octreotide, L-362,855 and BIM-23056 were weak inhibitors of [¹²⁵I]-Tyr¹¹-SRIF-14 binding (PIC₅₀ <6.5). [¹²⁵I]-BIM-23027 labelled a single population of SRIF binding sites in RDCM membranes and mouse fibroblast (Ltk^–) cells stably expressing the human recombinant sst₂ receptor. There was a significant correlation between the affinitestimates of a range of SRIF analogues at inhibiting [¹²⁵I]-BIM-23027 binding in RDCM membranes and binding to the recombinant sst₂ receptor in Ltk^– cells, suggesting that the sites labelled by [¹²⁵I]-BIM-23027 in RDCM are similar to the sst₂ receptor. GTPS (100 M) decreased [¹²⁵I]-BIM-23027 binding in RDCM by 60%.The results from these studies demonstrate that [¹²⁵I]-Tyr¹¹-SRIF-14 labels a heterogeneous population of high affinity SRIF binding sites in RDCM membranes. The majority of these sites are insensitive to GTPS and display negligible affinity for the cyclic hexapeptides, BIM-23027 and seglitide. The remaining high affinity binding sites can be selectively labelled with [¹²⁵I]-BIM-23027, are sensitive to GTPS and show similar characteristics to the recombinant sst₂ receptor which appears to mediate the antisecretory effects of SRIF in the mucosa (McKeen ES, Feniuk W, Humphrey PPA (1995) Naunyn-Schmiedeberg's Arch Pharmacol 352:402–411).     

Radioligand binding and functional characterisation of tachykinin receptors in chicken small intestine

Tachykinin receptors in chicken intestine were studied using radioligand binding and functional techniques. Mechanisms of tachykinin-induced contraction were also investigated. Binding of [125I]Bolton-Hunter substance P ([125I]BH-SP) to chicken ileal membranes was rapid, saturable, of high affinity and to a single population of binding sites with Kd 0.72 nM and Bmax 0.48 fmol/ wet weight tissue. The rank order of agonists competing for [125I]BH-SP binding sites was [Sar9]SP > [Arg3]SP (natural tachykinin in chickens) > SP > [Pro9]SP > or = NKA > eledoisin > [Sar9,Met(O2)11]SP > [Lys5,MeLeu9,Nle10]-NKA(4-10) > senktide, suggesting similarities to the mammalian NK1 receptor. The NK1 receptor antagonist CP 99994, and NK2 receptor antagonist SR 48968 were weak competitors while spantide, RP 67580, GR 82334, GR 94800 and MEN 11420 were ineffective. The radioligand [125I]NKA showed no specific binding to ileal membranes. The potency order of most tachykinins in contacting isolated ileal longitudinal segments was in good agreement with that obtained from competition binding studies. Contractions to [Arg3]SP, NKA and senktide were greatly reduced by tetrodotoxin, suggesting that neurally-mediated responses were primarily involved. [Arg3]SP and NKA acted mainly by increasing release of acetylcholine, prostaglandins and probably tachykinins. Responses to [Arg3]SP were virtually abolished by nifedipine but were unaffected by NK1 receptor antagonists. Senktide-induced contraction was inhibited by the NK3 receptor antagonist, SR 142801, but was unaffected by atropine or L-NAME. The study provides evidence for a tachykinin receptor with similarities to the NK1 receptor in the chicken small intestine. In addition, senktide may act on a receptor similar to the mammalian NK3 receptor.     

Evidence for specific N-terminal galanin fragment binding sites in the rat brain.

The existence and widespread distribution of specific [125I]galanin-(1-15) fragment binding sites in the rat brain was demonstrated by using [125I]galanin-(1-15) as a radioligand in quantitative receptor autoradiographical studies. These binding sites were also present in several areas lacking or having very few [125I]galanin-(1-29) binding sites, such as the dorsal hippocampal formation, the neocortex and the neostriatum. [125I]Galanin-(1-15) binding sites showed a high selectivity for the fragment, since galanin-(1-15) could displace 80% of the binding whereas porcine galanin-(1-29) could only displace 30%. The binding was saturable with a Kd of 0.63 +/- 0.02 nM and a Bmax of 15.3 +/- 1.7 fmol/mg in sections from the dorsal hippocampal formation. Thus, a new type of galanin receptor selective for N-terminal galanin fragments may exist in the rat brain.     

Characterization of binding sites for diisopropyl phosphorofluoridate in spinal cord cytosol

Gel filtration chromatography was performed on cytosol preparation of hen spinal cord to find molecular target(s) for organophosphorus-induced delayed neurotoxicity (OPIDN). Three binding peaks of [(3)H]diisopropyl phosphorofluoridate (DFP), an organophosphate that induces OPIDN, were separated from the cytosol preparation. The activities of acetylcholinesterase (AChE) and neuropathy target esterase (NTE) that has been proposed as a screening method for OPIDN eluted in the fractions within these two DFP binding peaks. However, the other peak had none of the activities of AChE and NTE. Therefore, this DFP binding proteins in cytosol may be peculiar to the pathogenesis of OPIDN.