- Led Zeppelin – “Black Dog”

I knew as soon as I saw this paper in Science this week that I wouldn’t be able to resist. The upcoming, and most likely unfortunate, Led Zeppelin reunion has been in the news in Britain and as a dog owner I’m always a sucker for canine genetics. In fact, one of the most interesting scientific talks I’ve ever heard was by Robert Wayne of UCLA at the Genetics Society in London a couple of years ago. His premise at the time was that the domestic dog evolved from gray wolves in different places and at different times but always under the same evolutionary pressure. It is easier, if you’re a dog, to live as a scavenger with humans than to do the hunting yourself. Dogs aren’t dumb.

That’s neither here nor there; the research published this week in Science by Greg Barsh’s group at Stanford University focuses on coat color in dogs. Specifically they describe the characterization and cloning of a dominant mutation in dogs that renders a black coat – creatively named dominant black. Mammalian coat color is an excellent system for the study of basic genetics because, like Mendel’s peas, phenotypes are easy to follow and they tend to involve only a handful of genes in well behaved pathways. All of the diversity that is present in coat color in mammals, including humans, stems from variation in the levels of two pigments – yellow or red pheomelanin and brown or black eumelanin. The genetic regulation of these pigments’ synthesis is fairly well characterized and in most mammals is under the control of two genes – Agouti and Melanocortin 1 receptor (Mcr1). The latter gene encodes a protein that sits in the cell membrane and acts as a receptor for the protein encoded by Agouti. Typically the ligand Agouti binds to the receptor Mcr1 resulting in the production of the yellow/red pheomelanin.

The family dog has always been an exception to this neat two gene model. Researchers have known for some time, by using classical genetic analysis, that dominant black (KB*) is not an allele of either Agouti or Mcr1. This mutation is one of three alleles of the K locus – the dominant black KB (rendering black coat color) is dominant to kbr (brindle color) which is dominant to ky (yellow). The recent sequencing of the dog genome allowed Barsh’s group to clone the K locus – and reveals a curious candidate.

To get at the gene responsible for dominant black, Barsh’s group used genetic mapping to roughly localize the mutation in the genome. They were able to exploit the fact that most dog breeds were created in the last 200 years from a very small population to localize the gene more specifically. Their assumption – that mutations within a breed are identical – allowed them to use association mapping to map dominant black to one of 16 genes, an impressive mapping experiment. This was followed by sequencing of the relevant regions to confirm that KB results from a small deletion in CBD103, one of a cluster of B-defensin genes. These genes encode small antibiotic proteins that are secreted from the skin and are important in immunity.

A mutation in a immune system gene is an odd candidate for a factor in coat pigmentation, so Barsh’s group did a number of experiments to confirm that they had the right candidate. These included looking at expression of CBD103 in different mutant backgrounds, expressing it in a mouse model system and establishing the binding affinity of the protein encoded by CBD103 to components of the Mcr1-Agouti pathway. All of these experiments provide support to the hypothesis that the dominant KB allele results from a the deletion in CBD103 and an increase in the amount of the mutated protein. The binding experiments show that the CBD103 protein can, like Agouti, bind Mcr1.

These result lead Barsh and his colleagues to propose that CBD103 can also act as a ligand for Mcr1. When the ky allele is present, not enough CBD103 is produced to affect the Agouti-Mcr1 interaction. However, in the dominant KB mutants, CBD103 is produced at a higher lever and competes with Agouti for binding to Mcr1. Under these circumstances, the darker eumelanin pigments are synthesized – and, voila, a black dog. But why would a gene involved in protection from microbes also play a role in coat color. The authors propose, rather unsatisfyingly, that beyond the role that they play in immunity, B-defensins may have been important during mammalian evolution to provide camouflage by producing novel, environment specific coat patterns and colors.

That just leaves me to include a gratuitous picture of my dog, Timmins. He is a white – not albino – Siberian husky. I was curious how, with only yellow and black pigment to work with, you can get white coats in dogs. You may have been looking at your pooch wondering the same thing – what explains the massive diversity in coat colors and patterns in domestic dogs. Unfortunately, it’s a little more complicated than the story told above. There are parallel genetic pathways that control different patterns of color and presence or absence of pigments.

Just a little editorializing, if I may. I’m not sure why this paper made it into such a high impact journal. The research is well designed and controlled, but ultimately it is just the description of the mapped based cloning of a gene responsible for a previously described mutation. Lower impact journals like Genetics are full of these sorts of papers. Yes, the link between immune response and coat color is new but it is a stretch to see this kind of paper in Science. Perhaps what attracted me to this story is the same thing that attracted the editors of Science – a soft spot for our faithful companions.

Image credits:

Yellow & black labs

Segregating litter

* I have not learned how to do superscript in HTML yet, so apologies – to those of you in the know – for the dodgy notation on these alleles. I did my best to make clear dominance and color.

-Barenaked Ladies – “Alcohol”

Alcoholism is a disease with major world health implications. There are over 15 million people in the U.S. suffering from alcoholism resulting in over 100,000 deaths a year and health care costs approaching 200 billion dollars. In Britain there are an estimated 36,000 hospital admissions a year that are directly related to alcohol and the public health cost in Australia is similar in scale to that in the USA. There is certainly a genetic component to alcoholism, but inheritance of the disease is complex and complicated by environmental factors. Alcoholism is thus a difficult disease to detect at an early stage and to treat. Isolating the genetic components – the genes that predispose a person to alcoholism – is also a complex process, but could prove valuable for detection and treatment.

For most genetic diseasez, scientists rely on animal models for the disease in which they are interested in studying. This is generally because one can neither selectively breed humans to establish inheritance patterns nor use experimental treatments. For diseases, like alcoholism, that have an environmental component there is a limit to how much control scientists can have over their subjects’ environments. Animal models allow researchers latitude for these and other genetic manipulations. The two species most commonly used for the study of alcoholism are mice and fruit flies (Drosophila). The latter may seem a strange choice but offer a number of advantages to geneticists. They are small and it is easy to rear thousands of flies in a small space. Their genome has been sequence and they have served as a model organism for the study of inheritance for over a century. Two-thirds of human disease genes have orthologues (versions) in Drosophila. Importantly for the study of alcohol abuse – flies get drunk.

A new study published in the latest issue Genome Biology exploits some of these advantages in a search for some of the genes that may be involved in alcoholism. The authors of this paper, a group from North Carolina State University, use whole genome expression analysis to identify candidate genes that affect alcohol sensitivity. The advantage that Drosophila provides in this study is in the experimental set-up. The NC State researchers established two artificially selected populations over 35 generations, one sensitive to alcohol and one resistant. This type of artificial selection involves selecting inidividuals that exhibit the trait of interest and breeding them together. The scale of these populations is only possible in an organism that requires little space and has a short generation time – like the humble fruit fly.

Using these two artificially selected populations, the researchers then looked for genes that varied in their expression level using a technique that allows each Drosophila gene to be analyzed. Gene expression refers to the production of an enzyme, a functional protein, from a gene, a stretch of DNA. Not surprisingly, considering the number of generations of selection, there were thousands of genes that were expressed at a different level in the two populations. Most of these changes, however, were modest and they focused on a few dozen genes with at least a two-fold change in the level of expression.

The most impressive thing about this paper is that they carry their research one step further. One of the main problems with whole genome transcriptional profiling by microarray is that the accuracy of the results is often questionable. This group confirms their microarray results by measuring alcohol sensitivity in mutants of each of the genes that were expressed differently in the two selected populations. Using these functional tests, mutants in 73% or 32 of the candidate genes were found to have effects on alcohol sensitivity. Of these, only three have been previously identified and 23 have human orthologues.

I try to avoid getting to much into the technical aspects of research in my Science Tuesday posts, but couldn’t resist this week. To establish whether or not flies are sensitive or resistant to alcohol they are placed in an inebriometer (really) and then subjected to a battery of locomoter tests including how high they can climb before falling and how long they remain mobile immediately after an overdose of alcohol. They also measure how often the flies fight after exposure to alcohol – not sure how flies fight. But if these types of tests sound fun to you, well… I don’t like to judge.

Basic conclusions are that there are a number of genes that may play a role in alcoholism in humans that have not been studied. There is no real pattern to be found in the type of genes affected – they play roles in a number of cellular processes. There were a surprising number of genes involved in sensitivity to odors that were differentially expressed in alcohol sensitive flies. However, the most exciting conclusions here is that this type of artificial selection experiment can provide researchers with dozens of candidate genes in humans to explore when studying complex genetic diseases.

There are a few flaws with these experiments. The researchers themselves point out that this study only really looks at one developmental time point, so it’s providing a picture of changes in gene expression at one time in the flies’ life cycle. Additionally, there may be changes in gene expression that are regulated by means other than transcriptional control. However, the biggest issue is that while flies can get drunk, they do not demonstrate the addictive propensity of human alcoholics. Thus, this study provides little information about the disease’s most deadly aspect. Most of the readers of this post have probably had a night in which they indulged a bit heavily – but only about 5% of you have felt the compulsion to carry on the next day. The current study provides a lot of genetic information about how the body, and by extension the brain, deals with an excess of alcohol. But it provides no information about the real issue in the disease of alcoholism, that next day syndrome – the addiction.

Image Credits:

Flies in a glass

Microarray

Buying music supports artists. This song is up for listening purposes only for a short time. If you’re an artist or label who wants the song taken down, please e-mail me: cdv[at]chrisdellavedova[dot]com.