If you own a cat, you'll know how much time they spend preening themselves. Rather than a bad case of narcissism, this represents a type of grooming behaviour, which is essential to the well-being of most animals. But, like most things, it can go wrong. In humans, defective grooming behaviour ranges from the mild, such as nail-biting, to the severe, such as the obsessive–compulsive disorder trichotillomania (compulsive hair pulling that results in hair loss). Little is known about the genetics that underlie grooming behaviours, but Greer and Capecchi now report that mice with null mutations in the homeobox gene Hoxb8 groom excessively, potentially providing researchers with a valuable model for studying abnormal grooming behaviour in humans.

The authors constructed a null mutation in Hoxb8 by inserting a frameshift mutation in exon 1 and a neo gene, surrounded by loxP elements, in exon 2. They then removed the neo gene by inducing Cre-mediated recombination, to produce an allele with nonsense codons in both exons. The removal of neo turned out to be very important because when neo was retained, a skeletal phenotype was seen in homozygous mutant mice. This was most likely because the presence of neo causes the aberrant expression of nearby genes in the Hoxb cluster.

In the absence of neo, bald patches (sometimes accompanied by skin lesions) were the only other phenotype in Hoxb8−/− mice. There was no obvious skin defect to account for this, so the authors videoed the mutant mice for 24 h to see whether they showed any behavioural abnormalities. Strikingly, they found that the mutants spend twice as long grooming themselves as their wild-type littermates. This excessive licking and biting leads to the loss of body hair and to skin lesions. Furthermore, the mutants also excessively groom normal mice in the same cage, indicating that the behaviour is not the result of some abnormality in the skin of the mutant mice — rather it is a defect in grooming behaviour itself.

Greer and Capecchi also studied the expression of Hoxb8 and found high levels of expression in several parts of the brain that had been previously implicated in grooming behaviour in several mammals. Imaging studies have also detected abnormalities in the same brain regions in patients with obsessive–compulsive disorder, which fits with the similarity in phenotype between trichotillomania and the Hoxb8 mutants. This work therefore provides a valuable way into the biology that underlies grooming behaviour. The search will now be on to see whether variation in HOXB8 itself accounts for any of the variation in grooming behaviour in humans.