dt hm under sink conditions dM Dm SKpCd Dm Kp
dt hm hm where dM/dt is the flux of material per time, Dm the membrane diffusion coefficient, S the cross-sectional surface area, Kp the partition coefficient, hm the membrane thickness, Cd the concentration of donor chamber, Cr the concentration of receiver chamber, Cd — Cr the concentration gradient at time t, P the permeability, and /un is the fraction unionized.
Based on the pH-partition hypothesis, an in vitro method was widely utilized to predict drug absorption by measuring a compound's ability to partition (log P) between a fairly immiscible lipophilic solvent (octanol) and water or buffers at different pHs (Leo et al., 1971). While this method does provide adequate predictive power within series of compounds, its broad utility is often limited by observed deviations between in vitro and in vivo permeability. These deviations were observed in early literature, but a correlative explanation was often incorrectly proposed to rationalize a fit to the pH-partition hypothesis (Shore et al., 1957).
In its most basic sense, under the partitioning model absolute bioavailability should increase linearly with increasing lipophilicity (log P) due to the lipophilic nature of biological membranes, where permeation is a function of diffusion. Expansion of the model realizes that highly hydrophilic compounds (log P < 1-2) would not be absorbed transcellularly due to their polar/ionic character, just as highly hydrophobic compounds (log P > 4-5) would accumulate in the interior aliphatic portion of the cellular membrane due to their lipophilicity. For example, an absorption model that is depended solely on partitioning would result in a parabolic relationship between the fraction absorbed (log 1/concentration) and lipophilicity (log P), which would only occur in unusual cases in vivo (Higuchi and Davis, 1970). Numerous deviations in the predictive power of log P values for assessing the permeability of a compound have been observed, and therefore, other predictive tools were advanced to better address these inconsistencies. One of the more recognized tools, "The Rule of Five" was proposed by Lipinski et al. (1997, 2000, 2001) to estimate the permeability of compounds in silico based on molecular descriptors at the early stages of drug discovery. Lipinski's Rule of Five states: ".. ..poor absorption or permeability is more likely when there are more than 5 H-bond donors, 10 H-bond acceptors, the molecular weight is greater than 500 and the calculated log P is greater than 5." (Lipinski et al., 1997). The rule of five has found broad utility to predict the developability and optimization of NCEs in industry. However, several confounding factors, most importantly the role of drug transporters, have acted to limit the predictability and applicability of this and many analogous techniques.
The complex nature of membrane physiology and the lack of predictive absorption methodologies is better understood when one considers the numerous roles elucidated for different drug transporters in mediating transcellular influx and efflux of xenobiotics. There are numerous classes of transporter proteins that have been identified to date, each with different, sometimes overlapping, substrate specificity, capacity and affinity, as well as specific tissue, cellular and temporal expression patterns. Transporter proteins are integral proteins that function via either a facilitated diffusion, or active, energy-dependent mechanisms to mediate transcellular flux of xenobiotics and nutrients (Oh and Amidon 1999). Not surprisingly, a compound's physicochemical properties greatly influence its interactions with transporters and lipophilic character (i.e., partitioning) plays a major role in determining these interactions. As such, not only is there great overlap in substrate selectivity of many transporters, but there is also variability in uptake due to lipophilicity and solubility differences and resulting membrane interactions. Much of this variability may also be due to interactions with other potential substrates present in biological fluids, stearic influences of transporter binding, possible membrane interactions, such as changes in fluidity, the rate of transport (Vmax), or even competitive binding of substrates to other transporters. Thus researchers have developed various in vitro and in vivo models to delineate the role of individual transporter activities in mediating xenobiotic uptake (Stewart et al., 1997; Ekins et al., 2000, 2005; Kimura et al., 2002; van de Waterbeemd, 2002; Harrison et al., 2004; Kassel, 2004; Sun et al., 2004).
A number of cell models have been used to evaluate the intestinal permeability of drugs, with the Caco-2 cell model, derived from colorectal adenocarcinoma, being the most widely used. Caco-2 cells exhibit much of the barrier functionality of the normal endogenous intestinal epithelium. Under the right conditions, Caco-2 cells not only grow in a tight-knit monolayer and exhibit tight junctions, but they express many of the same receptors and transporters of the intestinal epithelium. The simplicity of this model makes in vitro permeability and uptake measurements relatively straightforward and conducive to automation. However, despite the obvious benefits of in vitro drug transport studies using cell lines, there are also many disadvantages and hurdles to overcome when using cell line models. For instance, even though transepithelial electrical resistance (TEER) can be determined to verify the integrity of the monolayer, leakiness is always of concern when performing these studies and can be a source of artifacts. Additionally, cellular energy requirements dictate the endogenous expression of a multitude of transporters, many with overlapping substrate specificities. This not only provides a source of variation, but also makes it difficult to assess the transport properties due to a single transporter. Moreover, culture conditions can greatly affect the genetic regulation of a multitude of functional proteins, resulting in intra-and interlaboratory variations and error. In short, assessing drug transport function via various cell line models can be a useful tool, however, it also presents many technological hurdles, and the techniques employed do raise some issues of applicability to the physiologic model.
In addition to those limitations mentioned above, cell lines do not adequately address the issue of transcriptional and translational variations within any particular patient population. Single nucleotide polymorphisms (SNPs) and the regulatory mechanisms associated with these genetic variations have broadened the pharmacogenomics field quite extensively. The biochemical architecture of the human intestine, including drug transporters and their variants, must be viewed in the light of evolution of mammalian nutritional requirements. For the majority of mammalian evolution, the intestine served as the primary gatekeeper that compounds must traverse when entering the mammalian organism. As such, the in vivo model of drug absorption and permeation with respect to drug transporters is a function of all those transporter proteins expressed at any given time and location within the vicinity of the drug product as it moves along the intestine, working in concert to facilitate the acquisition of nutrients. While useful for permeability prediction and screening of large databases of compounds, cell culture models fall short of simulating the actual in vivo conditions and should be rationalized in context.
Due to the innate limitations in studying drug permeability in vitro, the pharmaceutical industry has been ever evolving in its quest to further understand and enhance the intestinal absorption of pharmaceuticals. Current approaches include formulation design to either exploit, or inhibit transporter function through the use of various excipients. Surfactants, for example, are well known to alter membrane fluidity, thus altering potential substrate interactions and transporter interactions. Another approach to maximize bioavailability is to tailor NCE design to improve substrate affinity for a particular transporter. A classic example utilizing this approach is the addition of a peptidic valine moiety to the antiviral acyclovir to produce valacyclovir. While the free compound does absorb well through the GI tract, the valyl addition increases the compound's transporter affinity for the oligopeptide transporter, PepTl, thus drastically increasing its oral bioavailability, as will be discussed in greater detail later.
A fundamental understanding of drug transporters is essential when analyzing a drug's PK/PD behavior after oral administration. Difficulties in characterizing the intestinal transepithelial transport of drugs underscores the need for a complete understanding of the biophysical and biochemical barriers that are present in the GI tract. This chapter aims to delineate the role of several selected transporter families likely involved in the intestinal bioavailability of orally administered drugs. These transporter families have been selected due to the breadth of literature demonstrating their respective importance in intestinal drug absorption. The chapter will highlight the interplay of the molecular and functional characteristics for each of the different transporters and summarized with respect to the impact of these characteristics on altered bioavailability and pharmacokinetics.
Was this article helpful?