TRICHOTILLOMANIA (or hair pulling) is a condition characterised by excessive grooming and strong, repeated urges pull out one’s own hair. It is classified as an obsessive-compulsive disorder (OCD), and is relatively common, affecting about 2 in 100 people. Sufferers normally feel an increasing sense of tension before pulling out their scalp hair, facial hair, and even pubic hair, eyelashes or eyebrows. This provides gratification, but only briefly.
Hair pulling is usually thought of as being psychological in origin, but an intruiging new study now suggests that it occurs as a result of defects in the immune system. The study, which is published in the journal Neuron, shows that excessive grooming and hair pulling occur in mice because of reduced numbers of microglial cells, which are critical for the brain’s immune response. It also suggests – very unexpectedly – that bone marrow transplants may be an effective treatment for trichotillomania in humans.
Shau-Kwaun Chen of the Department of Human Genetics at the University of Utah School of Medicine and his colleagues investigated mice with mutations in the Hoxb8 gene. Hox genes are well known to be involved in establishing the body plan: most organisms express numerous Hox genes, and unique combinations expressed in different regions of the two major body axes constitute a code which instructs the embryo to form each organ and body part in the correct position.
Mice carrying Hoxb8 mutations exhibit excessive and pathological grooming, leading to hair loss. Grooming is an innate behaviour in mammals, and follows a stereotyped pattern, in which the head is groomed first, followed by the body and, finally, the genitals and tail. The Hoxb8 mutants spend twice as much time grooming as their healthy littermates. The chain of events is normal, but occurs more frequently, and each bout of grooming lasts longer, leading to hair removal and self-inflicted open sores at the grooming sites. This behaviour is very similar to that of humans with trichotillomania, so the Hoxb8 mutants serve as a useful model for the condition.
Researchers have been studying these mutants ever since they were first created in 2002, but still aren’t sure why they show excessive grooming. One hypothesis is that the behaviour is associated with reduced sensitivity to pain – the mice exhibit altered responses to noxious and thermal stimuli, which is thought to occur because of disrupted connections between neurons in the spinal cord which transmit this information up to the brain. But Hox gene mutations normally produce numerous defects, and the subtler ones often go unnoticed. And although excessive grooming involves numerous brain structures, Hoxb8 is present at very low levels in the brain, making it difficult to establish which cells express it.
Chen and his colleagues started off with an analysis of Hoxb8 expression in the brain, by generating a strain of mutant mice in which cells that express the gene also express yellow fluorescent protein as a “reporter” gene. Dissection of the animals’ brains revealed that small numbers of cells expressing yellow fluorescent protein were distributed throughout the organ, but were predominant in the cerebral cortex, striatum, olfactory bulb and brainstem. On closer inspection, the cells were identified as microglia, non-neuronal cells which constitute the main form of the brain’s immune defence. Yellow fluorescence was only seen in a subpopulation of the cells, however, which constitutes about 40% of the total number of microglia in the brain. And when the researchers looked at brain slices from Hoxb8 mutants, they saw a significant reduction of microglial cells.
During embryonic development, one subpopulation of microglia is generated in the bone marrow, from white blood cells called monocytes, before migrating through blood vessels and into the brain soon after birth. In newborn healthy mice, the researchers observed yellow fluorescent cells in several sites, including the choroid plexus and the meninges (the membranes enveloping the brain), at which migrating microglia first gain entry into the brain. They also found that the number of fluorescent cells decreased with distance from these sites.
These findings suggest that Hoxb8 is required for specifying the subpopulation of microglial cells that is derived from the bone marrow. The researchers therefore took blood samples from their genetically engineered mice, and found that all the blood cell types – which are produced in the bone marrow – expressed yellow fluorescent protein, confirming that they also express Hoxb8. Most bone marrow cells were also positive for yellow fluorescent protein.
Next, the researchers took bone marrow cells from healthy and Hoxb8 mutant mice and transplanted into the mutants. They then monitored the animals’ grooming behaviour, as well as the regrowth of hairless patches of skin, for a period of 5 months after the transplants. Bone marrow transplants from healthy mice, but not those lacking Hoxb8, were found to abolish excessive grooming in the mutants. In this group of animals, the hairless skin patches grew back and the open wounds healed (above). By 3 months, most were fully recovered, and could not be distinguished from their healthy litter mates on the basis of either grooming behaviour or appearance.
Finally, the researchers created another strain of mutant mice, in which the Hoxb8 gene is selectively deleted from haematopoietic stem cells, which give rise to all blood cells during development and in adults. (Senior author Mario Capecchi is one of pioneers of this “conditional knock-out” technology, and was awarded the 2007 Nobel Prize for Physiology or Medicine for his contribution to developing the technique.) These animals also exhibited pathological grooming, despite lacking spinal cord defects. And another mutant strain, in which Hoxb8 was selectively deleted from spinal cord cells, was insensitive to pain but did not exhibit excessive grooming. This rules out the possibility that pathological grooming is a result of an altered pain response.
This study therefore provides strong evidence that pathological grooming in Hoxb8 mutant mice occurs because of a reduction in the number microglial cells in the brain. Behaviours such as grooming reduce the number of pathogens on the skin surface, minimizing the risk that they will enter the body and cause disease. Pathogens on the skin can be irritating, and the natural response is to try to brush them off. Microglial cells are also involved in clearing pathogens – they are the brain’s first line of defence, which are deployed to clear up damaged cells or to destroy microbes. In evolutionary terms, it might be advantageous to link behaviours with the cellular responses that perform similar functions.
The mechanism by which a microglial deficiency leads to pathological grooming is as yet unknown, but the researchers offer an hypothesis. Microglia are known to be closely associated with synapses in the brain, and to regulate signalling between nerve cells by releasing chemical messengers called cytokines. Chen and his colleagues note that they observed microglia in many of the brain regions linked to OCD. A reduction in their numbers is, therefore, likely to disrupt activity in the neural circuits involved in pathological grooming. Regardless of the underlying mechanism, the surprise finding that bone marrow transplants abolish compulsive grooming behaviour in the Hoxb8 mutant mice has obvious implications for the treatment of obsessive-compulsive disorders in humans.
Chen, S., et al. (2010). Hematopoietic Origin of Pathological Grooming in Hoxb8 Mutant Mice. Cell 141: 775-785 DOI: [PDF].
Huey, E.D., et al. (2008). A psychological and neuroanatomical model of obsessive-compulsive disorder. J. Neuropsychiatry Clin. Neurosci. 20: 390-40. [PDF]