Transcription factors do a lot of things for us, like make arms and legs in the right places.  It turns out they may also make us go bald.
We all know that being bald runs in families.  Both my Dad and I are bald, as are five of my eight uncles and both Grandfathers.  Based on twin studies, the heritability of male pattern baldness is 81% .  This means that if your father or uncle is bald, you have a pretty good chance of going bald, too.   Even though the presence of a genetic component to baldness seems obvious, until this decade nobody had any idea what genes are responsible for this terrible affliction. 

The first gene, the Androgen Receptor (AR), was identified in 2001.  This gene was found because it made sense:  it is a receptor that regulates a cell's sensitivity to androgens (male sex hormones).   It is well established that balding regions of the scalp have increased levels of both the androgen DHT (a derivative of testosterone) and the androgen receptor.  Elimination of androgens (such as by castration, but I don't recommend it) can eliminate balding.  [This is why I have always argued that balding is a very manly trait.]

For the next few years, no new genes were found even though AR can't be the whole story, since some men are much balder than others.  On October 12, two studies were published in Nature Genetics (one German and one British) that have nearly identical results, and both implicate a new region of the genome in male pattern baldness.  Both studies examined various European populations and found that balding people tended to inherit the same genetic variants on a region of chromosome 20 called 20p11. 

So how big of a deal is it if you have the balding variant of this region?  It depends on your AR gene.  The British study found that if someone carries an allele of AR that doesn't promote balding, then your odds of becoming bald don't even double compared to the rest of the population (odd ratio ~ 1.6).  But, for those poor bastards who have balding alleles at both AR and 20p11 (about 14% of the population), your odds ratio shoots up to 7.12, meaning that you are seven times more likely to become bald than the more fortunate souls around you. 

So what does this genomic region do?  The short answer is, absolutely nothing.   What is striking about the region of 20p11 identified is that it is over 350,000 base pairs (letters of DNA code) away from the nearest gene.  That's a long way as far as DNA is concerned.  The only way this region could affect balding, as far as anyone can tell, is by changing how strongly a nearby gene is expressed (turned on or off).  Regions of the genome that affect the expression of nearby genes are called cis-regulatory regions.  350,000 bases is a long way, but it is not unheard of for cis-regulatory regions to affect gene expression at that distance.

So what nearby genes are affected?  The two closest genes are PAX1 and FOXA2.  What is important about these genes is that they are both transcription factors.  Transcription factors are genes responsible for turning other genes on and off (they bind to the cis-regulatory regions I just mentioned).  Since all of the cells in your body start with all the same genes, the reason you have a liver, heart, limbs, and so on is because transcription factors turn certain genes on and off at specific places (for example, the transcription factor MyoD instructs cells to become muscle).  Of course, it's possible that the effect of 20p11 has nothing to do with PAX1 or FOXA2, but the current best explanation of the data is that one of those two genes is involved.  Coincidentally, mutations in FOXI3,  a gene similar to FOXA2, cause several breed of hairless dogs. 

What I like about these data is that they provide evidence in humans for a trend that is growing in popularity among evolutionary geneticists; the trend that transcription factors will often be a source of variation.  Because they turn entire groups of genes on and off,   transcription factors are poised to have huge effects (adding just four transcription factors to a skin cell can turn it back into a stem cell).  My first instinct is that transcription factors are just too imporant to mess with in evolution (mutating one transcription factor in flies creates this monster). But, as we learn more and more about genetic variation, geneticists are finding that there are ways for a transcription factor to have a big effect on a trait without totally screwing up the organism.  Transcription factor changes are now known to underlie domestication traits in corn, wheat, rice, and tomatoes.  Transcription factor changes are also responsible for adaptive evolution in the threespine stickleback fish, and for differences between the larvae of fly species (ironically, whether or not they have hair).

If it turns out that the transcription factors PAX1 or FOXA2 really do affect male pattern baldness, then I may see some gene therapy in my future.  I think it's too late for Dad.