P-gp is expressed on the apical membranes of brain capillary endothelial mono-
layers derived from rat (Nakagawa et al. 2009 ), man (Poller et al. 2008 ), cow (Culot
et al. 2008 ), and pig (Smith et al. 2007 ).P-gp actively reduces cerebrospinal fluid
concentrations of drugs before they reach the brain. In dogs, this is a mechanism
which prevents or reduces neurotoxicity of ivermectin in most breeds (Mealey 2008 ;
Mealey et al. 2008 ). This same protective mechanism is one which limits distribu-
tion of useful drugs into CSF. Furthermore, as these transporters act upon a range of
substrates, there is a consequent increased risk of drug–drug interactions at the level
of the transporter. Pekcec et al. ( 2008 ) observed over-expression and increased
function ofP-gp at the BBB of epileptic dogs, which may contribute to resistance
to and reduced efficacy of anti-epileptic drugs, such as phenobarbitone. Overcoming
P-gp efflux function with modifying agents may be useful and even necessary when
considering treatment strategies involvingP-gp substrates for veterinary and human
patients with CNS disorders. Specific examples where overcomingP-gp could be of
benefit include delivery of chemotherapeutic agents to brain tumours (Miller et al
2008 ), as well as for treatment of drug-resistant epilepsy (Luna-Torto ́s et al. 2008 ).
4.2 Potential Veterinary Drug Interactions Based on Transporters
The species-specific expression of transporters has been discussed. Significant
advances have been made in pharmacogenetics in human medicine as a result of
the influence of patient-specific target receptor expression and metabolic enzyme
activity on drug PK and PD (Martinez et al.2008a). Although this tailored approach
remains in the early stages of development in veterinary therapy, some indications
of “proof of principle” have been generated through knowledge ofP-gp expression
in canine breeds. The trigger for this research was the discovery of the idiosyncratic
neurological hypersensitivity of multi-drug resistant gene knock-out (ABCB1-1D)
herding dog breeds to ivermectin (reviewed in Mealey 2008 ). The phenomenon has
been observed in Collies, Australian Shepherds and Shetland sheepdogs (Martinez
et al.2008a), as well as inAbcb1a(/) mice (Schinkel et al. 1994 ) and in a herd of
Australian Murray Grey cattle (Eagleson et al. 1987 ). Normally, brain concentra-
tions of ivermectin are many times lower than those in plasma and the liver, but
failure of the BBB to prevent ivermectin permeation into the CNS results in
ivermectin neurotoxicity inABCB1-1DCollies (Mealey et al. 2003 ). An experiment
demonstrating that CNS levels of ivermectin were 100-fold greater inAbcb1a(/)
mice than in their wild-type counterparts indicated that this was due to reduced
efflux from the BBB (Schinkel et al. 1994 ). The prevalence of theABCB1-1D
homozygous genotype in Collies has been estimated to be approximately 30–40%
(Mealey et al. 2002 ). Ivermectin-sensitive Collies develop clinical signs after
administration of oral doses as low as 0.1 mg/kg (Paul et al. 1987 ), while non-
sensitive dogs can tolerate doses up to 2.5 mg/kg (Campbell and Benz 1984 ). The
dysfunctional phenotype is due to a spontaneous gene knock-out resulting in a
4-base pair deletion in theABCB1gene (Mealey 2004 ), which prevents the protein
from emerging from the golgi apparatus for routing to the plasma membrane of BBB
100 D.J. Brayden et al.