Solubilization Processes In The Gastrointestinal Tract

Guide To Creating Spa Products

Handcrafter's Companion Guide

Get Instant Access

While the maximum solubility of a drug in the GIT is probably the most critical parameter controlling its absorption, decades of study have failed to identify a reliable method for consistent and accurate prediction of in vivo drug solubility. The in vivo dissolution rate and maximum solubility of a specific drug substance in the GIT is determined not only by the physicochemical characteristics of the drug substance (e.g., pKa, hydrophilicity, crystal structure, particle size, etc.), but also by the complex interaction of multiple GI physiological factors, each of which is subject to considerable intra- and intersubject variation. Solubilization by BS mixed-micelles, formed during lipid digestion, is thought to play a pivotal role in the absorption of poorly water-soluble hydrophobic drugs. Lipid digestion begins in the stomach, where gastric lipase hydrolyzes approximately 10% to 20% of ingested triglyceride (TG) to FFA and diglyceride (DG), both of which possess surfactant properties and greater water solubility than the parent TG (18). The crude lipid digestate enters the duodenum as a coarse emulsion which, primarily under the influence of pancreatic TG lipase, is almost completely hydrolyzed to 2-monoglyceride (2-MG) and additional FFA, which combine with BS to form mixed micelles. The hydrolytic activity of pancreatic TG lipase is dependent on the presence of colipase, which is essential for promoting its attachment to the relatively hydrophilic surface of the crude TG emulsion formed in the stomach (19-21). In vivo lipolysis is a dynamic process resulting in the formation of multiple colloid phases (22-24). During lipolysis, a lamellar liquid-crystalline phase, which is comprised of a relatively high ratio of LP to BS, forms continuously at the surface of the TG droplets; upon subsequent enrichment in BS, mul-tilamellar vesicles are formed and leave the surface of the TG droplets. Progressive enrichment of these vesicles with BS leads to their transformation into unilamellar vesicles and finally, mixed micelles, which form when the LP:BS ratio decreases to less than unity. The mixed micelles transport the hydrophobic LPs across the intestinal unstirred water layer, where upon arrival at the surface of the intestinal brush border membrane, the acidic microclimate results in micel-lar disintegration, releasing the hydrophobic LP in close proximity to the lipophilic intestinal epithelium, where they are rapidly absorbed. The remnant BS is returned to the intestinal lumen, where it is incorporated into nascent mixed micelles. Following the completion of lipid absorption, which takes place largely in the duodenum and upper jejunum, the BS is reabsorbed in the ileum by the process of enterohepatic recirculation.

The small intestinal concentrations of BS that control the solubilization of lipids and many hydrophobic drug substances, are not only influenced by dietary status (fed vs. fasted), and by various disease states, but are normally subject to considerable intra- and intersubject variation, as well. Human intestinal BS concentrations reported in the literature are dependent on the method of determination (e.g., site and sampling time with regard to meals and degree of dilution in vivo). But in general, mean fasted state BS concentrations typically range from 1.5 mM to 6mM (25-28), while mean postprandial concentrations typically range from 8 mM to 20 mM, with values as high as 40 mM having been reported (25,26,29,30). In a recent study, the level of human intestinal BS was found to be 1.8 mM and 8mM in fasted and fed state respectively (30).

Small intestinal lipase activity in healthy human subjects varies with dietary status and is assumed to be in excess since approximately 95% of ingested dietary lipid is absorbed. It should also be emphasized, however, that lipase activity varies not only with the specific lipid substrate (e.g., compared to long chain TG, short chain, and medium chain TG are better substrates for pancreatic lipase), but also with the physical presentation of the lipid substrate to the GIT. For instance, the lipolysis rate is inversely proportional to the lipid droplet size (31), whereas the presence of certain surfactants used in lipid formulations (32,33), as well as certain drugs themselves (33), have been shown to directly inhibit the rate of lipolysis. And the excipient, Cremophor RH40 (hydrophilic ethoxylated triglyceride surfactant), completely inhibited the lipolysis of medium chain triglyceride (MCT) oil over a 90 minute period when the oil and surfactant were present in equal amounts (32). Subsequent addition of Imwitor 988 (which is a mixture of medium chain mono- and diglycerides) allowed lipolysis to occur, presumably due to a change in the orientation of the surfactant at the oil-water interface. Finally, the lipid excipients Peceol and Gelucire 44/14 have been demonstrated to inhibit pancreatic lipase activity in a concentration-dependent manner (34). The number and complexity of these factors make it difficult to accurately estimate the actual lipase activity in the small intestine (25). However, Armand et al. (35) has reported a 15-fold increase in lipase activity, relative to the fasted state, upon intragas-tric administration of a test meal containing 48 g triglyceride to healthy human volunteers. In summary, accurate and quantitative in vitro assessment of the solubilization, trafficking, and intestinal absorption of hydrophobic drug substances has remained elusive, emphasizing the need for further work in this area.

THE DYNAMIC LIPOLYSIS MODEL Description of the Model

Based on the foregoing discussion of intestinal lipid digestion and absorption, the reader should have gained an appreciation for the number and complexity of the parameters which can influence gastrointestinal (GI) absorption of BCS 2 compounds, many of which are absorbed via the same pathways as lipids. The in vitro dynamic lipolysis model developed in our laboratories (Fig. 1), which incorporates many of these parameters, has been used to study the processes governing the absorption of BCS 2 drugs as well as the interactions of formulation and food on these processes. The pH of the media in the lipolysis model was chosen based on the small intestinal pH in the fed (pH 6.5) and fasted (pH 5.5) states (18,36), the pH range for optimal pancreatic lipase activity (between pH 6 and 10) (37) and the average pKa of FFAs (pH 6.4) (38).

Due initially to economic considerations, in vitro lipolysis models have employed crude porcine bile (39-41) and pancreatic extracts (which contain approximately equimolar concentrations of colipase and lipase) (42) in lieu of the more costly purified BSs (43-48) or purified colipase and lipase enzymes, respectively (39,40,49,50). While use of purified BSs and pancreatic lipases allows the researcher to study the mechanism of lipolysis under more controlled conditions, the use of crude extracts will produce a system closer to the in vivo situation due

Figure 1 The dynamic lipolysis apparatus consists of a thermostatically controlled, double wall reaction vessel, a computer controlled pH-stat with an auto-burette for the addition of NaOH, and a peristaltic pump for the continuous, controlled addition of calcium chloride. All experiments are performed under continuous agitation via magnetic stirring.

Figure 1 The dynamic lipolysis apparatus consists of a thermostatically controlled, double wall reaction vessel, a computer controlled pH-stat with an auto-burette for the addition of NaOH, and a peristaltic pump for the continuous, controlled addition of calcium chloride. All experiments are performed under continuous agitation via magnetic stirring.

to the presence of phospholipase and other enzymes present in the normal physiological secretions. The pancreatic lipase solution used in our experiments is prepared by dissolving in water, an amount of the crude extract appropriate to the enzymatic activity desired followed by centrifugation to remove insoluble particulate matter. The lipase activity of the resulting solution can be assayed by several different methods; for example, the USP (51) describes a method based on the hydrolysis of a gum arabic/olive oil-in-water emulsion but another assay, based on tributyrin hydrolysis, is also widely used (52). It should be noted that the measured lipase activity is dependent not only on the lipid substrate but also on the oil droplet surface area presented to the lipase enzyme. Therefore, results generated by different experimental methods should not be compared.

The approximate composition of the crude porcine bile extract used in our experiments is glycochenodeoxycholic acid (42.0%), glycocholic acid (23.6%), glycohyocholic acid (16.4%), taurochenodeoxycholic acid (5.2%), chenodeoxy-cholic acid (5.1%), glycohyodeoxycholic acid (3.2%), taurocholic acid (3.0%), and hyocholic acid (1.4%). The total concentration of BS in the extract is determined colorimetrically and is expressed as the amount of 3a-hydroxy bile acids present (40). Due to a very low PL content in bile extract, egg phosphatidylcholine (PC) is added to the media in a BS:PC ratio of 5:1, to approximate the normal physiological conditions (10). It should be noted that regardless of the source of BS employed, PL must always be added to achieve a physiological BS:PL ratio.

Ionized calcium has been shown to have a significant, positive influence on the lipolysis rate of triglycerides in the presence of BSs, presumably by diminishing the electrostatic repulsion that occurs between enzyme and substrate, thereby facilitating the binding of pancreatic lipase to the TG droplets (12,19,39,40,49,50). In addition, lipolysis is thought to depend on the formation of a catalytically active complex comprised of lipase, mixed micelles, and calcium (49,50). Accumulation of FFA on the surface of the TG droplets during lipolysis will sterically hinder attachment of pancreatic lipase, resulting in a progressive decline in the lipolysis rate (43,50). The FFA chain length has been found to influence the degree of inhibition, with long chain fatty acids (LCFA) being more potent inhibitors of lipolysis than either short or medium chain fatty acids (40). In vivo, solubilization of FFA in BS-PL micelles and subsequent absorption prevents accumulation of FFA in the intestinal lumen thereby preventing inhibition of lipolysis. In the dynamic lipolysis model, continuous addition of calcium chloride solution serves to control accumulation of FFA in the medium by forming insoluble calcium soaps, which precipitate, thus removing FFA from the system and preventing accumulation (32). During in vitro lipolysis experiments, addition of calcium initially results in a relatively high lipolysis rate, which subsequently declines and then stabilizes. This is presumably due to increased concentrations of FFA in the dissolution medium and possibly due to the loss of BS through the formation and precipitation of insoluble calcium-BS complexes (Fig. 2) (24,32,33,39,42,50).

Dynamic lipolysis experiments are conducted in a jacketed thermostatically-controlled reaction vessel maintained at 37°C and agitated continuously with a magnetic stirring device (Fig. 1) (39-41). The reaction medium consists of a mixture of BS, PL, buffer, and lipid substrate (e.g., dietary lipid or lipid-based formulation incorporating the drug substance). Lipolysis is initiated by addition of the lipase solution and the pH and free calcium concentration of the reaction mixture is maintained by the computer-controlled addition of sodium hydroxide and calcium chloride solutions, respectively.

Samples of the reaction medium are withdrawn immediately following addition of the lipase solution and at serial time points subsequent to the initiation of lipolysis. The lipolysis reaction is quenched by addition of the lipase inhibitor, 4-bromobenzene boronic acid, and the samples are subsequently ultracentrifuged, resulting in the formation of three distinct phases:

■ a pellet comprised largely of insoluble calcium soaps of fatty acids,

■ an intermediate aqueous layer, consisting of BS mixed-micelles and various lipid vesicles, and

■ an upper-most, oily layer comprised of DG, and unhydrolyzed TG.

The aqueous phase is of greatest interest in the study of the GI absorption of hydrophobic drugs. This phase has been extensively characterized with regard to its composition and content of LP and solubilized drugs as well as the identity and size of its component micellar and vesicular entities (39,40,53).

Figure 2 The lipolysis of triglycerides (TG) in the dynamic lipolysis model over 40 minutes, measured as mM fatty acids (FA) titrated, at three different rates of Ca2+ addition in the presence of 8mM bile salts. In this series, 18.3mM of titrated FA are equivalent to complete hydrolysis of the added TG into FA and monoglycerides. These experiments demonstrate a clear correlation between the lipolysis rate and the rate of Ca2+ addition. The initial, high rate of lipolysis declines and remains relatively constant at approximately five minutes after initiation of Ca2+ addition. Each line represents a different rate of Ca2+ addition, as defined below: (a) 0.072mmol Ca2+/min, final conc. = 9.4mM Ca2+ (- - - -);

(b) 0.135 mmol Ca2+/min, final conc. = 17.4mM Ca2+ (---); (c) 0.181mmol Ca2+/min, final conc. = 23.0mM Ca2+ (- -); (d) 0.0mmol Ca2+/min, final conc. = 0.0mM Ca2+ (- - );

(e) 17.4 mM Ca2+ added at the initiation of the experiment (-); this concentration of

Ca2+ corresponds to the final concentration in (b) and results in a similar rate of lipolysis. All experiments were conducted in duplicate. Source: From Ref. 39.

Figure 2 The lipolysis of triglycerides (TG) in the dynamic lipolysis model over 40 minutes, measured as mM fatty acids (FA) titrated, at three different rates of Ca2+ addition in the presence of 8mM bile salts. In this series, 18.3mM of titrated FA are equivalent to complete hydrolysis of the added TG into FA and monoglycerides. These experiments demonstrate a clear correlation between the lipolysis rate and the rate of Ca2+ addition. The initial, high rate of lipolysis declines and remains relatively constant at approximately five minutes after initiation of Ca2+ addition. Each line represents a different rate of Ca2+ addition, as defined below: (a) 0.072mmol Ca2+/min, final conc. = 9.4mM Ca2+ (- - - -);

(b) 0.135 mmol Ca2+/min, final conc. = 17.4mM Ca2+ (---); (c) 0.181mmol Ca2+/min, final conc. = 23.0mM Ca2+ (- -); (d) 0.0mmol Ca2+/min, final conc. = 0.0mM Ca2+ (- - );

(e) 17.4 mM Ca2+ added at the initiation of the experiment (-); this concentration of

Ca2+ corresponds to the final concentration in (b) and results in a similar rate of lipolysis. All experiments were conducted in duplicate. Source: From Ref. 39.

Simulating Fed State Effects Using the Lipolysis Model

During lipid digestion, a drug dissolved in a substrate lipid vehicle will either (i) remain in the lipid, (ii) be solubilized in the aqueous phase in combination with LPs, or (iii) precipitate.

During lipolysis, the trafficking of a particular drug between these various phases will be controlled by multiple factors, many of which are poorly understood, but are thought to include the drug lipophilicity and affinity for the various lipolytic phases.

Zangenberg et al. (41) conducted dynamic lipolysis experiments on the poorly water-soluble drug substances, danazol (Log P 4.5) and probucol (cLog P 11) at up to 75% total lipolysis of the lipid content of model long chain triglyceride

(LCT) emulsion formulations of these drugs. For danazol, the relatively high drug concentration in the aqueous phase was found to be directly correlated with the concentrations of surfactants (BS, PL) and LP (FA, MG) present. In comparison, probucol, which had a relatively high Log P and poor solubility in the aqueous phase, remained largely solubilized in the oil phase, due to its relatively high cLog P. The inverse relationship between Log P of the drug substance and solubilization in the aqueous phase was replicated by Kaukonen et al. (43), who conducted dynamic lipolysis experiments over 30 minutes on five poorly soluble drug substances ranging in log P from 2 to 8.1. At the termination of the experiments, in which only 42% or 61% of the added TG had been hydrolyzed, a considerable amount of undigested oil remained in which the most lipophilic compounds (cinnarizine, log P 5.5 and halofantrine, log P 8.1) were preferentially retained. However, it should be emphasized that in these experiments the entire amount of calcium chloride was added to the lipolysis medium at the initiation of the experiment, in contrast with the controlled, continuous addition of calcium chloride in the dynamic lipolysis model, as it has been described by Zangenberg et al. (41) and this could have influenced the extent of lipolysis.

Christensen et al. (40) employed a lipolysis model to study the solubilization of probucol (Log P 10.9), LU28-179 (Log P 8.5), and flupentixol (Log P 4.5), in media containing 2.8% TG and 20/4 mM BS/PL; lipase-free media was included as a control. The solubility of flupentixol in the lipolysis medium was higher than the added amount, resulting in nearly complete solubilization of the drug prior to the initiation of lipolysis. At 30 minutes subsequent to the initiation of lipolysis, the solubility of flupentixol in the medium was observed to decrease (Fig. 3A). The most likely explanation for this observation is that the continuous addition of Ca2+ required to maintain the rate of lipolysis resulted in progressive neutralization of the negative surface charge of the mixed micelles weakening the electrostatic attraction, and subsequently decreasing the solubility of the protonated flupentixol molecules. In contrast, the aqueous concentrations of LU 28-179 and probucol increased as lipolysis progressed, presumably due to the formation of LP in which these drugs had relatively high solubility (Figs. 3B and 3C). During lipolysis, the average micel-lar size increases, most likely due to the incorporation of newly formed LP. Prior to the initiation of lipolysis, slight increases in micellar size were observed, possibly due to the high concentration of counterions in the media. Immediately following the initiation of lipolysis, a lag phase preceded the incorporation of the drug into the aqueous phase. This may have been due to an initial, preferential association of LP with the undigested oil phase immediately following the initiation of lipolysis or subsequent to the precipitation of insoluble drug-BS complexes prior to the accumulation of solubilizing mixed micelles as lipolysis progressed.

Studies carried out with fed dogs have shown enhanced absorption of the poorly soluble LU 28-179, relative to the fasted state (Internal report, H. Lundbeck). Similar findings have been demonstrated for probucol, administered to mini-pigs (54). The enhancing effect that food has on the absorption of many drugs is due, at least in part, to the generation of LPs with surface-active

Was this article helpful?

0 0
Lower Your Cholesterol In Just 33 Days

Lower Your Cholesterol In Just 33 Days

Discover secrets, myths, truths, lies and strategies for dealing effectively with cholesterol, now and forever! Uncover techniques, remedies and alternative for lowering your cholesterol quickly and significantly in just ONE MONTH! Find insights into the screenings, meanings and numbers involved in lowering cholesterol and the implications, consideration it has for your lifestyle and future!

Get My Free Ebook


Post a comment