This webpage is part of a series on the Genetics of Dog Coat Color and Type and mounted on August 26, 2006 by Sheila Schmutz. It was last updated on Oct. 9, 2010.
Although C. C. Little's book is the considered the classic work on color, he did not address hair length or type. For a discussion of these, one must look at Frederick B. Hutt's book "Genetics for Dog Breeders" (1979, W. H. Freeman and Company, San Franciso) Chapter 8 or to the earlier work by by Ojvind Winge, "Inheritance in Dogs with Special Reference to the Hunting Breeds" (1950, Comstock Publishing, Ithaca, N.Y.).
Crawford and Loomis (1978) studied 1,216 St. Bernards from 221 litters and showed that short coat (L/-) was dominant to long coat (l/l). They stated that the type of coat is clear by 6 weeks of age. Interestingly they observed that short coated pups had bluish skin and the hair was white at the base, whereas long haired pups had pink skin and that the hair was buff at the base.
In 2006, almost 30 years later, Donna Housley and Patrick Venta published that the gene causing hair length in several breeds of dogs was fibroblast growth factor 5 (FGF5). They studied this gene because it causes "angora" in mice. They initially began their study with Cardigan Welsh Corgi in which the long haired variety is called "fluffy".
They have shown that the mutation in FGF5 causes hair length differences in:
Shadow, at the left, is a Rough Collie and Prima, at the right, is a Smooth Collie. Since Prima had pups with a rough coat, we know she is L/l in genotype. Different breeds use different names for longhaired and shorthaired.
At the Third International Conference on Advances in Canine and Feline Genomics held in Davis, CA from August 2-6, 2006, Dr. Housley said they were currently studying other breeds to determine if their hair length differences are also caused by this mutation..
In 2010, a paper by Edouard Cadieu in Elaine Ostrander's group at NIH published a paper that expanded the breeds studied to 80. Among these, many breeds were either all long-haired (35) or all short-haired (33). Some additional breeds had individuals of both long and short coat varieties and a few breeds (11) had only 1 or 2 individuals tested.
They indicate that the mutation causing long is recessive also, but the length of the "long" is variable within individuals and across the breeds. They classified some dog breeds as having meidum length hair and either or both genotypes were present in some of these.
Most dogs grow their coat to a specific length and then the hair growth stops. Pups of long coated breeds do not achieve their full length coats for several months.
In other breeds, the hair keeps growing and growing. Some people refer to these breeds as "non-shedding". Lisa Peterson of AKC was quoted in Pets magazine (Jan/Feb 2006 issue) as listing the following breeds in this category, but there are probably others also.
In late fall of 2009 a comprehensive study of dog coat length and type was published. This study involved the FGF5 gene previoulsy discovered to affect coat length but expanded the number of breeds tested, as discussed above. In addition two new genes were described which affect wire and curly coats in dogs.
Wirehaired dogs have coarser textured hair that is more bristley than smooth. Many wirehaired dogs also have "facial furnishings" or a beard. It was the "furnishings" characteric or "beard" that was used by Cadieu et al. 2009 to discover the gene involved is R-spondin-2 (RSPO2). It is quite likely that shorthaired wire coats feel more wirey than longhaired wire coats. The Wirehaired Pointing Griffon is an example of a dog with a wirehaired coat but there were many breeds with this genotype, including most of the Terriers.
Winge (1950) postulated that wire (W) is dominant to non-wire (w). This was supported in the stud by Cadieu et al. (2009). The dogs with furnishings had at least one mutant allele that had an insertion in the region past the stop codon of RSPO2.
Some breeds of dogs have "curly" coats, the Curly Coated Retriever being one example. The Irish Water Spaniel, shown at the right is another example. Cadieu et al. 2009 showed that a single base pair mutation in a keratin gene, KRT71 (previously called K6irs1, Kb34 and K71) is responsible for the curly coat trait in dogs. Arginine is changed to trypotophan at amino acid 151 in the mutation responsible for curly. Curly is recessive to straight.
A detailed study of how the genes for length and curly interact to cause various coat types in Portuguese Water dogs has also been conducted by Parker et al.
The following breeds were reported to have curly hair based on their genotype:
The following breeds were reported to have curly hair, or not, based on their genotype:
Corded coats are probably coats that are both curly and continuously growing. The Puli is an example of a breed with a corded coat, as illustrated by Raz on the right. The Komondor is another breed with a corded coat. It is said that this coat protected sheep herding dogs from being bitten by wolves since it was so thick the teeth didn't penetrate it.
A few breeds of dogs do not grow hair and are sometimes called "hairless". The Xoloitzcuintli, often called the Mexican Hairless is a breed that does not grow hair. The Chinese Crested grows hair on its extremities only in one variety, as shown by Melogy on the left. In the other variety, the dog has long hair all of its body and this is called "powder puff".
The literature states that at least some types of hairlessness are lethal when the dominant allele is homozygous. One copy of the dominant allele would result in a hairless dog. A dog inheriting two copies of the homozygous recessive allele would produce hair.
The mutation causing the hairless phenotype was discovered using a genome scan by a collaboration of the labs of Kirsten Lindblad-Toh and Tosso Leeb.
The American Hairless Terrier is always hairless. A breeding study by Sponenberg et al. suggests the trait in this breed is inherited as a recessive.
back to Dog Coat Color Main Page