Cynomolgus Monkey as Host for the Prediction of Drug Transporter Interactions

written by Marton Jani, SOLVO Biotechnology

The following monkey drug transporter proteins have been cloned and characterized in vitro as of mid 2014: OAT1, OAT3, OATP1B1, OATP1B3, OATP2B1, MRP1, MRP2, P-gp.

Tissue mRNA profiles of monkey transporter and drug metabolizing genes are available (Nishimura et al 2009). A landmark study on the absolute, age dependent quantification of BBB transporter proteins in monkey has also been conducted (Ito et al 2011).

Functionality of P-gp was compared across human and preclinical species with LLC-PK1 monolayers and a panel of probes and inhibitors. Excellent cross species correlation was obtained in direct transport studies. Inhibition studies revealed slight differences in IC₅₀ (2-3 fold) between human and monkey when applying verapamil and quinidine as inhibitors, digoxin and cyclosporine A as probes. (Takeuchi et al 2006, Suzuyama et al 2007)

Primary hepatocytes have been applied to study the inducibility of P-gp, MRP1 and MRP2 mRNA levels with reference inducers dexamethasone, omeprazol and rifampicin. MRP2 and P-gp mRNA were more responsive to induction in the monkey hepatocytes compared to human whereas MRP1 response was leveled (Nishimura et al 2008).

MRP2/Mrp2, BCRP/Bcrp and BSEP/Bsep mRNA and protein levels were assessed in hepatocytes and liver samples from human, rat, dog and monkey. Relative MRP2 mRNA levels were significantly lower in the monkey, however absolute protein quantification revealed an inverse relation. Among the studied species rat protein quantities were highest (approximately five-fold of human and three-fold of monkey) (Li et al 2009a).

BCRP protein amounts were well comparable, however BSEP protein expression is significantly higher (2-3 fold) in the monkey liver compared to human (Li et al 2009b).

Fluorescent probes for P-gp, MRP and BCRP mediated transport showed high similarity between monkey and human primary hepatocytes in efflux experiments (Li et al 2008).

Inside-out canalicular membrane vesicles are highly efficient in the study of apical efflux in hepatocytes. In this cross species study estradiol-glucuronide and LTC4 have been applied as MRP2/P-gp and MRP2 specific probes. Obtained K_M parameters consistently ranked the monkey preparations as higher affinity (approximately two-fold lower)(Shilling et al 2006) in accordance with other reports (Ninomiya et al 2006).

An atlas containing mRNA based tissue expression profiling of 50 xenobiotics transporters in preclinical species and human was published in 2006 (Bleasby et al 2006).

A recent review focuses on interspecies differences in drug transport on the molecular and system level (Chu et al 2013) and highlights that rodent data are far more abundant, and that significant transporter groups are still entirely lacking in the molecular level description. Also, transgenic animals have been instrumental in revealing the role and contribution of single transporters to active PK processes. This solution will never be practical with monkeys.

Systemic Level PK Interspecies Correlations

The most comprehensive interspecies study compares the predictive power of rat, dog and monkey i.v. PK profiles using a set of 103 drugs with available i.v. clinical data. Clearance rates, mean residence times and volumes of distribution have been assessed, and overall it was determined that the monkey i.v. PK provides the most accurate prediction (Ward and Smith 2004a, b).

Oral bioavailability in monkeys appears to be significantly lower than in human however, and the discrepancy primarily stems from lower intestinal availability (F_g). Compounds that are excreted largely unchanged and are not efflux transporter substrates show good cross species correlation. However compounds that undergo CYP3A metabolism and/or are efflux transporter substrates show poor correlation. This observation is solidified by tissue mRNA data that demonstrate significantly higher P-gp, MRP2, BCRP and CYP3A expression in the monkey small intestine compared to human or rodent. As such the monkey model may not be ideal as model of human oral bioavailability and PK (Akabane et al 2010, Takahashi et al 2008).

Further comparative monkey PK studies are available on a number of individual, mostly developmental compounds.

Barrier Specific In Vivo Tools in Monkey

The DDI apparent between probenecid and famotidine and also other H2 receptor antagonists cannot be detected in rodents. This was largely attributed to interspecies differences in the basolateral uptake activity of the proximal tubular epithelium, specifically OAT1, OAT3 and OCT isoforms. With a combination of in vivo and in vitro studies it was shown, that the monkey is closer to human in terms of both the kinetic parameters of OAT1/3 uptake activity and also OCT isoform expression pattern, as compared to rats, and that the DDI present in human can be reproduced in the monkey. Consequently it has been proposed that the monkey is a better model of human active renal clearance in comparison with murine species (Tahara et al 2006).

The brain is a safe haven for HIV particles and for successful therapy antiviral concentrations must reach therapeutic levels in the CNS. Attaining successful levels is hampered by P-gp mediated efflux at the BBB, that may be challenged with the coadministration of P-gp inhibitors.

Zosuquidar (LY335979) has been successfully employed to significantly elevate the brain concentration of nelfinavir (Kaddoumi et al 2007) in monkeys, an approach that has been validated in rodents beforehand (Anderson et al 2006). It was also shown that CSF levels did not correlate with brain levels, thus CSF concentrations need to be handled cautiously when acting as brain level surrogate.

DDI studies have been designed to replicate clinical DDI events with pitavastatin as victim and rifampicin and cyclosporin A as perpetrators, that act on hepatic uptake transporters. The kinetic profile of the probe in the monkey experiment was well comparable to the clinical observations, and pitavastatin was suggested as a probe candidate for hepatic DDI in vivo studies in the monkey (Takahashi et al 2013).

In order to set up a preclinical P-gp/CYP3A sensitive screening tool the PK of ketoconazole has been established in the monkey. Ketoconazole is a potent inhibitor of both enzymes and different profiles of repeated doses with and without a concomitant ketoconazole dose will single out interactions with these enzymes (Ward et al 2004). This method has been validated with midazolam and fexofenadine as victims in intestinal DDI detection (Ogasawara et al 2007) and also with simvastatin as victim in active hepatic clearance (Ogasawara et al 2009)

PET based experiments have also been successfully applied to study the effect of P-gp in the BBB on brain penetration using three 11C or 18F labeled P-gp substrate compounds. Individual K_p levels and also the effect of cyclosporin A were well in line with expectations, however significant interspecies differences have been observed between rat, minipig and monkey. The variations have been attributed to species dependent transport by P-gp and effect of CsA (Syvanen et al 2009).

Conclusions

Even though it can be assumed that the vast majority of data generated with the application of monkeys never reaches public domain, open information allows reaching the expected conclusion, that the cynomolgus monkey will generally return PK results with the highest predictive power among preclinical species, including drugs that interact with transporters.

Due to species specific circumstances at the small intestine (significantly higher expression of efflux transporters and CYP3A metabolic enzymes at the brush border membrane) the above statement is weakened by the observation that compounds that interact with either enzyme will usually generate an underestimation of human bioavailability for oral dosing.

A decent coverage of monkey drug transporters is already available in recombinant systems, albeit several important proteins (prominently BCRP and organic cation transporters) remain uncovered. The existing data show very high commonality in terms of transport activity and inhibition profile with human orthologs. To date MRP2 is the only transporter that can be singled out with relatively mild interspecies differences. This high commonality in fact could undermine the utility of the monkey in vitro methods in the long term, as the abundant human transporter based in vitro tools would be able to generate data that provide the same quality for scale up or translation.

The per data point expense entailed in monkey experiments will ensure that the monkeys remain the last line of proof of concept before going into human trials, and in the long term systemic PK studies will probably be replaced with in silico models populated with specific in vitro data.