GHB
GHB is a naturally occurring short-chain fatty acid formed from γ-aminobutyric acid (GABA) that is found in the mammalian brain, heart, liver and kidney (
77). It acts potentially as a neuromodulator through binding to the GABA(B) receptor (
78). In addition, GHB is formed from the precursors γ-butyrolactone and 1,4-butanediol (
79). Therapeutically, GHB is approved in the US to treat narcolepsy (marketed as Xyrem®) (
80) and in Europe for the treatment of alcohol withdrawal (
81). However, abuse of GHB is widespread; it is used by body builders for its growth hormone releasing properties (
82), by drug abusers as a recreational drug for its euphoric effects (
83), and in drug-facilitated sexual assault due to its sedative/hypnotic effects (
84). The increased abuse of GHB has lead to a rise in associated overdoses and fatalities (
82). Adverse events associated with GHB overdose are principally characterized by central nervous system and respiratory depression as well as cardiovascular and gastrointestinal effects with symptoms including seizures, dizziness, nausea, vomiting and unconsciousness potentially leading to coma and death. (
82) Currently, the treatment of GHB overdose is limited to supportive care; physostigmine and naloxone have been tried as antidotes with minimal success (
79).
GHB pharmacokinetics have been demonstrated to be nonlinear in humans (
85–
88) and rats (
89,
90), with total clearance decreasing as a function of increasing dose. Several mechanisms contribute to the observed nonlinear pharmacokinetics including capacity-limited metabolism (
85,
87,
89,
90), saturable absorption (
91), and nonlinear renal clearance (
6). While metabolic clearance represents the predominant elimination pathway for GHB (
77), renal clearance becomes increasingly important in overdose situations with high urinary concentrations reported in humans (
92,
93). In contrast to the observed changes in total clearance with increasing dose, renal clearance increases in a dose-dependent manner in rats (
6). Furthermore, the fraction of GHB excreted in urine increases tenfold (3% to 30%) over the dose range of 108–208 mg/h per kilogram (
6). These dose-dependent increases suggest the involvement of active renal reabsorption which is saturated at high concentrations.
In vitro studies have characterized the renal transport mechanisms of GHB and elucidated the MCT isoforms contributing to GHB reuptake. Studies were conducted in rat kidney membrane vesicles, a human kidney cell line (HK-2 cells) and rat MCT1 transfected MDA-MB231 cells. Studies conducted in rat brush border (BBM) and basolateral (BLM) membrane vesicles isolated from rat kidney cortex characterized the renal transport mechanism (
12). GHB and
l-lactate both undergo pH- and sodium-dependent uptake in BBM vesicles and pH-dependent uptake in BLM vesicles, suggesting the involvements of proton-dependent and sodium-dependent MCTs (
12). MCT1 is expressed at both membranes, although there is greater expression at the BLM; MCT2 is expressed only at the BLM (
12). HK-2 cells express MCT1, MCT2 and MCT4 at both the mRNA and protein level, which agrees with expression patterns in the human kidney cortex (
62). GHB uptake in HK-2 cells was driven by a pH-gradient, and was inhibited by CHC suggesting that MCTs, but not SMCTs, were responsible for its uptake in HK-2 cells (
11). Similar uptake parameters and similar inhibitory effects were observed for GHB and lactate suggesting transport by the same or similar transporters (
11). Additionally, GHB uptake was inhibited by pCMB indicating that MCT2 may not be an important transporter in GHB uptake (
11). Silencing RNA for MCT1, 2 and 4 in HK-2 cell studies suggested that GHB is predominantly transported by MCT1 (
10,
11), among the proton-dependent MCTs. Further studies, conducted in MDA-MB231 cells (endogenous expression of MCT2 and MCT4) and MDA-MB231 cells transfected with rat MCT1, provided further evidence regarding the specific MCT isoforms involved in GHB renal uptake (
10,
12). GHB was found to be a substrate for MCT1, 2 and 4 (
10,
11). However, based on the expression patterns of MCTs in the kidney, MCT1 is likely the primary isoform responsible for GHB renal uptake.
To further explore the influence of MCT1 on GHB renal reabsorption, studies were conducted to assess the modulation of MCT1-mediated GHB transport through the evaluation of potential inhibitors. Uptake of GHB in MDA-MB231 cells was inhibited by the classic MCT inhibitors CHC, phloretin and pCMB with uptake being approximately 60% of control cells (
10). In rat MCT1-transfected MDA-MB231 cells, GHB uptake was inhibited by phloretin, CHC and
d-lactate (
12). GHB uptake was also inhibited to a large extent by
l-lactate in rat BBM and BLM vesicles (
12). As the inhibition of MCTs results in altered GHB uptake in kidney cells, further studies investigated potential strategies for increasing the renal elimination of GHB by administrating the MCT substrates
l-lactate and pyruvate (
6). The administration of
l-lactate resulted in an approximately twofold increase in renal clearance (63 to 118 ml/h per kilogram) with a concomitant increase in total clearance and decrease in steady-state plasma concentrations (
6). These results suggest that administration of MCT inhibitors presents a possible clinical strategy for increasing GHB elimination in overdose cases. Furthermore, administration of
l-lactate was well tolerated and with no or minimal changes in blood electrolytes indicating a lack of toxicity (unpublished data). Additional
in vitro and
in vivo studies were conducted to evaluate novel MCT inhibitors and their potential to increase GHB renal clearance and overall pharmacokinetics. Quercetin, a naturally occurring flavonoid, has been demonstrated to inhibit MCT1 and MCT2-mediated
l-lactate uptake (
58), therefore, we assessed the inhibitory potential of a range of flavonoid compounds. All flavonoid aglycones that were evaluated inhibited GHB uptake in MCT1-transfected cells, with luteolin, morin, and phloretin resulting in the greatest reduction in GHB uptake (IC
50 values of 0.41, 6.21 and 2.57 μM, respectively) (
94). In contrast, the flavonoid glycosides had minimal effects on GHB uptake (
94). The high potency of luteolin for inhibiting GHB uptake suggested that this compound may be effective at increasing the renal clearance of GHB. Co-administration with luteolin (10 mg/kg) significantly altered the pharmacokinetic profile of GHB (1 g/kg) in rats; renal clearance of GHB increased more than threefold (1.36 to 4.28 ml/min per kilogram) with a concomitant decrease in the pharmacodynamic effect, return of righting reflex (
94). These data suggest that flavonoids such as luteolin are potent MCT inhibitors both
in vitro and
in vivo.
Additional studies have demonstrated that GHB transport occurs via a carrier-mediated processes in the intestine (
91), brain (
95), and at the blood–brain barrier (
5). MCT expression has been demonstrated along the length of the intestine with MCT1 being the predominant isoform (
19). Protein expression of MCT1, MCT2 and MCT4 has been demonstrated in human intestinal Caco-2 cells (unpublished data). Transport studies conducted in Caco-2 cells showed that GHB and D-lactate uptake occurred in a pH- and proton-dependent manner and similar uptake parameters were observed for GHB and
d-lactate which could be inhibited by MCT substrates and inhibitors (unpublished data). These results suggest that GHB’s absorption is mediated, at least in part, by MCTs in the human intestine.
In situ brain perfusion experiments conducted in rats demonstrated saturable uptake of GHB into various brain regions (
5). Furthermore, GHB uptake was inhibited by known MCT inhibitors including lactate and pyruvate, suggesting that GHB is a substrate for MCTs expressed at the blood–brain barrier (
5).
Influence of MCTs on the Disposition of Other Drugs
While GHB represents the best studied drug substrate of MCTs, a number of other drugs have been demonstrated to be MCT substrates or inhibitors in various
in vitro systems. Interestingly, the recent characterization of MCT6 showed that this isoform was involved in the transport of diuretics, such as bumetanide, and did not transport monocarboxylates (
3).
Investigations into MCT drug substrates have typically been conducted in tissue specific cell lines and membrane vesicles, including Caco-2 cells, NBL-2 cells, and MDCK cells as well as
in vivo. Caco-2 cells and intestinal membrane vesicles represent effective models to assess transporter-mediated intestinal absorption. Caco-2 cells have been demonstrated to express MCT1, 2 and 4 protein (unpublished data from our lab) with MCT1 expression being the predominant form consistent with expression in the human intestine (
19). MCT1 is expressed on brush border (apical) membrane of intestinal cells and thereby facilitates the intestinal cell uptake of its substrates. In these systems, MCTs have been demonstrated to transport the β-lactam antibiotics cefdinir (
96) and carindacillin (
97,
98), salicylic acid (
99), pravastatin (
100) and atorvastatin (
101). Furthermore, the role of MCT1 in the intestinal uptake of the β-lactam antibiotic prodrug, carindacillin, has been confirmed in rat intestinal brush border membrane vesicles. It is thought that the MCT-mediated uptake of carindacillin contributes to the improved exposure to carbenicillin (
97). This concept has also been employed in the design of a gabapentin prodrug, XP13512, which was designed to be a substrate for MCT1, specifically to overcome the poor intestinal absorption of gabapentin. Uptake of XP13512 was demonstrated to be MCT1-mediated in Caco-2, HEK-derived and MDCK cell monolayers (
38).
In vivo studies on XP13512 demonstrated increased oral absorption and bioavailability in rats and monkeys when compared to gabapentin suggesting that exploitation of MCT-mediated uptake may provide a novel strategy for improving intestinal drug absorption (
39).
MCT-mediated uptake of drugs, including statins, probenecid and GHB, has also been demonstrated at the blood-brain barrier. Tsuji
et al. (1993) (
102) demonstrated that the uptake of simvastatin acid in bovine brain capillary endothelial cells occurred via a pH-dependent carrier-mediated process. Furthermore, the authors showed that pravastatin had a lower affinity for the same transport process and in addition was able to inhibit the uptake of simvastatin acid (
102). Statins were recently identified to be inhibitors of MCT4 with the lipophilic statins (fluvastatin, atorvastatin, lovastatin acid, simvastatin acid and cerivastatin) demonstrating the greatest inhibitory potency (
65). Further studies need to be conducted to determine the influence of brain MCT expression on the overall disposition of MCT substrates and the potential therapeutic implications.