When considering the use of macrocyclic lactones such as ivermectin or moxidectin in dogs, veterinarians have followed the adage, “white feet, don’t treat.” This refers to the known sensitivity of Scotch collies (both rough and smooth) to neurotoxicity when administered these drugs at higher than label doses. But the adage has also been applied to many other herding breeds and has prevented veterinarians from using these drugs in situations where they would have been ideal. The neurotoxicity was attributed to a leaky “blood-brain barrier” in susceptible dogs. Recent developments in the molecular mechanisms of this phenomenon have opened a new frontier in the area of pharmacogenetics — drug disposition determined by the animal’s genotype. So “white feet, don’t treat” is no longer the practice standard; now it is “white feet, test to see if you can treat.”
This story started serendipitously in the 1990s with some genetically deficient mice. The mice were multidrug resistant (mdr) knockout mice. The mdr gene codes for a P-glycoprotein. Compared with wild-type mice, mdr knockout mice have altered central nervous system (CNS) penetration, enhanced oral absorption, and altered excretion (both urinary and biliary) of some P-glycoprotein substrate drugs. Thinking that this would cause some significant problems for the knockout mice, researchers were disappointed to find that the mice were healthy, fertile, and lived a normal life span. They had no abnormalities in anatomy, nor were there any abnormalities in any physiologic parameters. The researchers sadly concluded that under normal laboratory conditions, P-glycoprotein was not essential for basic functions. So the knockout mice languished in the laboratory, until the room they were housed in developed a mite infestation. Laboratory protocol for mite infestation included a topical spray of ivermectin. The next day, nearly all the knockout mice in the room were dead, but the wild-type mice were perfectly healthy. The knockout mice had brain concentrations of ivermectin 100 times higher than did the wild-type mice. This accidental discovery led the way to hundreds of studies on the role of the P-glycoprotein drug transporter system (1).
Recognizing that the collies and other herding breeds that were sensitive to ivermectin were similar to the mdr knockout mice, Dr. Katrina Mealey at Washington State University, College of Veterinary Medicine, demonstrated that a deletion mutation of the mdr gene was present in ivermectin-sensitive collies (2). The deletion mutation produces a frame shift that generates a premature stop codon in the mdr gene, resulting in a severely truncated, nonfunctional P-glycoprotein. P-glycoprotein is a large protein that functions as a transmem-brane efflux pump; transporting chemicals from the inside the cell to outside the cell. It is normally expressed in the apical border of intestinal epithelial cells, brain capillary endothelial cells, biliary canalicular cells, renal proximal tubular epithelial cells, placenta, and testes. Adenosine triphosphate (ATP) hydrolysis provides the energy for active drug transport, so the transporter can function against steep concentration gradients. P-glycoprotein transports a wide variety of drugs with diverse chemical structures, including chemotherapy drugs, immunosuppressants, antiparasitic agents, HIV-1 protease inhibitors, and corticosteroids (Table 1). How the P-glycoprotein transporter can recognize and transport such structurally diverse compounds is not known. Whether or not a drug will be a P-glycoprotein substrate cannot be based simply on its chemical structure. Many P-glycoprotein substrates are natural compounds, or synthetic derivatives of natural compounds, so this seems to be an evolutionary advantage as a protective mechanism to decrease exposure to toxic xenobiotics. Microbial pathogens and cancer cells make use of it for drug resistance (3).
Table 1.
Currently known P-glycoprotein substrate drugs
| Chemotherapeutics | Cardiac Drugs |
|---|---|
| Doxorubicin | Digoxin |
| Mitoxantrone | Diltiazem |
| Paclitaxel | Losartan |
| Vinblastine | Quinidine |
| Vincristine | Verapamil |
| Antimicrobials/Antifungals | Steroids |
| Doxycycline | Aldosterone |
| Erythromycin | Cortisol |
| Itraconazole | Dexamethasone |
| Ketoconazole | Estradiol |
| Rifampin | Hydrocortisone |
| Tetracycline | Methylprednisolone |
| Immunosuppressants | Miscellaneous |
| Cyclosporin A | Amitriptyline |
| Tacrolimus | Butorphanol |
| Antiemetics | Ivermectin |
| Domperidone | Morphine |
| Ondansetron | Moxidectin |
| H1-antihistamines | Phenothiazines |
| Fexofenadine | Phenytoin |
| Terfenadine | Selamectin |
| H2-antihistamines | |
| Cimeditine | |
| Ranitidine |
P-glycoprotein does not have intrinsic metabolic functions, but it is an important component of intestinal drug metabolism. Cytochrome P450 3A (CYP 3A) is the major phase 1 drug metabolizing enzyme family in mammals. Cytochrome P450 3A and P-glycoprotein are expressed at high levels in the villus tip of enterocytes in the gastrointestinal tract. They work in concert to prevent oral absorption of many drugs, as substrates of P-glycoprotein are often also substrates for CYP 3A. When a substrate drug is present in the intestinal tract, it is absorbed by passive processes into the enterocyte. Once inside the enterocyte, 3 things can happen: the drug may be metabolized by CYP 3A, the drug may enter the systemic circulation, or the drug may be extruded by P-glycoprotein back into the intestinal lumen, where it may enter another enterocyte at a more distal site along the digestive tract, thus allowing further access to CYP 3A. So non-P-glycoprotein substrate drugs pass through the enterocyte only once, while P-glycoprotein substrate drugs may continuously cycle between the enterocyte and the intestinal lumen, resulting in either repeated access of CYP 3A to the drug molecule or fecal excretion of the drug because of repeated P-glycoprotein efflux. Because so many drugs are substrates for both P-glycoprotein and CYP 3A, it is difficult to discern the individual contributions of each protein to reduced oral drug absorption.
The P-glycoprotein system can be knowingly manipulated. For example, ketoconazole inhibits P-glycoprotein efflux activity and CYP 3A metabolic activity; when administered concurrently with cyclosporin, it increases the oral bioavailability of cyclosporine. Concurrent administration of P-glycoprotein substrate drugs and inhibitor drugs must be done very carefully, or toxicity can occur. But this effect may be utilized for effective oral administration of drugs that normally would have poor bioavailability.
P-glycoprotein is also expressed on renal tubular cells and bile canalicular cells. It appears to play a role in drug excretion. Decreased renal or biliary excretion may be involved in the susceptibility of collies and related breeds to toxicity from chemotherapy drugs, such as doxorubicin and vincristine, and digoxin (4,5).
The blood-brain barrier limits the passage of drugs into the CNS. P-glycoprotein is normally expressed on brain capillary endothelial cells and functions as part of the blood-brain barrier to pump drugs out of the CNS. Dogs with the gene deletion have increased brain concentrations of drugs including ivermectin, moxidectin, loperamide, and corticosteroids. Once the gene defect was understood, a retrospective study from the Illinois Poison Control Center revealed that collies were over represented for cases of loperamide CNS toxicity when given “normal” doses (6). In normal dogs, loperamide does not cross the blood-brain barrier and even in heterozygote dogs, normal doses of loperamide do not cause toxicity. The blood-brain barrier of homozygous dogs is also more permeable to exogenous and endogenous steroid hormones. Collies are often considered to be “poor doers” and there is evidence that homozygous dogs have continuous suppression of the hypothalamic-pituitary axis. According to Dr. Mealey, such “atypical Addisonian” dogs require exogenous corticosteroid supplementation when stressed or ill.
Genetic studies have documented the mdr gene deletion in 10 breeds (7) (Table 2). The gene deletion is widespread in collies, with 30% being homozygous and 40% being heterozygous. Its frequency is much lower in other herding breeds of collie lineage, such as the Shetland sheepdog (8.4% of dogs tested carried the mutation), Old English sheepdog (3.6%), and Australian sheepdog (16.6%). Because of different lineage, the gene deletion has not been found in Border collies, bearded collies, or Australian cattle dogs. The deletion also occurs in 2 sighthound breeds, the longhaired whippet and the silken windhound, with suspicion that the mutation was introduced with Shetland sheepdog crosses. It has also been identified recently in white German shepherd dogs. Dogs that are homozygous for the gene deletion readily show adverse effects from ivermectin and other P-glycoprotein substrate drugs at dosages that cause no adverse effects in normal dogs. Heterozygote dogs may show toxicity at increased doses of substrate drugs, such as daily ivermectin administration for the treatment of demo-decosis. Rather than avoiding the use of ivermectin and other P-glycoprotein substrate drugs in collies and other affected breeds, the genotype of a dog can be determined before treatment. A simple cheek swab is all that is required and samples can be sent to:
Table 2.
Breeds Known to Have the MDR-1 Gene Deletion
| Australian shepherds |
| Miniature Australian shepherds |
| Collies |
| English shepherds |
| Longhaired whippets |
| McNabs |
| Old English sheepdogs |
| Shetland sheepdogs |
| Silken windhounds |
| White German shepherd dogs |
Dr. Katrina Mealey, Veterinary Clinical Pharmacology Laboratory, College of Veterinary Medicine, Washington State University, Pullman, WA 99164–6610, (509) 335–2988 (phone), (509) 335–0880 (fax), www.vetmed.wsu.edu/depts-vcpl
References
- 1.Schinkel AH, Smit JJ, van Tellingen O, et al. Disruption of the mouse mdr1a P-glycoprotein gene leads to a deficiency in the blood-brain barrier and to increased sensitivity to drugs. Cell. 1994;77:491–502. doi: 10.1016/0092-8674(94)90212-7. [DOI] [PubMed] [Google Scholar]
- 2.Mealey KL, Bentjen SA, Gay JM, Cantor GH. Ivermectin sensitivity in collies is associated with a deletion mutation of the mdr1 gene. Pharmacogenetics. 2001;11:727–733. doi: 10.1097/00008571-200111000-00012. [DOI] [PubMed] [Google Scholar]
- 3.Mealey KL. Therapeutic implications of the MDR-1 gene. J Vet Pharmacol Ther. 2004;27:257–264. doi: 10.1111/j.1365-2885.2004.00607.x. [DOI] [PubMed] [Google Scholar]
- 4.Henik RA, Kellum HB, Bentjen SA, Mealey KL. Digoxin and mexiletine sensitivity in a Collie with the MDR1 mutation. J Vet Intern Med. 2006;20:415–417. doi: 10.1892/0891-6640(2006)20[415:damsia]2.0.co;2. [DOI] [PubMed] [Google Scholar]
- 5.Mealey KL, Northrup NC, Bentjen SA. Increased toxicity of P-glycoprotein-substrate chemotherapeutic agents in a dog with the MDR1 deletion mutation associated with ivermectin sensitivity. J Am Vet Med Assoc. 2003;223:1453–1455. 1434. doi: 10.2460/javma.2003.223.1453. [DOI] [PubMed] [Google Scholar]
- 6.Hugnet C, Cadore JL, Buronfosse F, et al. Loperamide poisoning in the dog. Vet Hum Toxicol. 1996;38(1):31–33. [PubMed] [Google Scholar]
- 7.Neff MW, Robertson KR, Wong AK, et al. Breed distribution and history of canine mdr1-1Delta, a pharmacogenetic mutation that marks the emergence of breeds from the collie lineage. Proc Natl Acad Sci USA. 2004;101:11725–11730. doi: 10.1073/pnas.0402374101. [DOI] [PMC free article] [PubMed] [Google Scholar]