The missing link

Comparative genome analysis has changed theories about evolution and disease. Andrew West examines how it may also change ideas about drug development.

The genomes of rodents, primates and humans have recently been compared for the first time, and the results may have some far-reaching consequences for the world of drug discovery.

The search for new drugs involves a labyrinth of trials where 99 per cent of the paths are a dead end. Despite millions spent on research and the involvement of some of the world’s top scientists, the successful commercialisation of drugs remains a complicated affair marred by low compound discovery and high failure rates.

Now, as well as answering some fundamental questions about mankind, genome sequencing projects are being used in conjunction with current technologies to help revolutionise the way potential cures for human diseases are evaluated. Genome sequencing involves determining the position of every chemical base in the DNA of the species being examined. This data can be used to establish the positions within the DNA strands of genes that have a specific function in the organism. By comparing animal genomes to the human genome, important genetic features, such as genes responsible for disease and immunity, can be identified and contrasted. Then, rather than discovering new drugs by trial and error, drugs can be specifically designed to target a particular defective gene. Treatments can be tailor-made for individuals and adverse reactions to drugs will be easier to predict and avoid, saving time, the high cost of unsuccessful clinical trials and the lives of thousands of laboratory animals.

But rather than validating current testing techniques as hoped, the publication of the mouse and rat genome sequences and comparisons to the human genome have muddied the waters. With a draft chimpanzee genome sequence available and a high-quality version due soon, the genetic differences between rodents and humans have become more apparent, raising further questions about the validity of mice and rats as human models in drug trials.

Comparing the mouse and human genomes produces some dramatic conclusions. The mouse genome contains 2.5bn DNA letters, around 14 per cent fewer than the human genome, but the human genome contains more repeat sequences than the former. Almost all the 30 000 to 40 000 human and mouse genes are related and have 80 per cent similarity. However, this number could be misleading; some genes are almost identical, while others are barely recognisable as being the same. The genes responsible for smell, immunology and reproduction are most different between humans and mice. These observed differences have allowed scientists to investigate why mice are immune to certain human diseases. They also give a genetic basis to why mice produce many more offspring than humans and have a better sense of smell.

However, the rat genome sequence proved to be even more suprising. Again, almost all rat genes have a related human counterpart, but the human and rat genes are 90 per cent similar. Further, only six of the human genes known to cause disease when defective do not have 1:1 analogues in the rat sequence. Theoretically, this makes the rat a better human disease model.

Similarities and differences
Comparing the three genomes shows around one third of the genome sequences can be lined up for all three species, showing regions that code for important mammalian characteristics. Regions where the sequences line up for the mouse and rat but not for the human genome have also identified where rodent-specific genes are located. Finally, there are clear regions where the genome sequences cannot be aligned. These regions contain species-specific information: regions, in effect, which contain the code to make us human.

With the mouse and rat genomes proving to be so similar to the human genome, it may seem surprising that drug development is hampered by so many failed trials. The problems arise for two reasons. Rodent physiology is quite different to human physiology; the way rodents absorb and excrete drug molecules differs from humans. For example, rats have no gall bladder and can excrete bile very readily. If a drug is excreted via bile, its half-life will be severely reduced in the rat model, possibly suggesting that the drug would be unsuccessful in human trials.

However, the most significant difference in drug molecule activity in humans, rats and mice is the way the chemicals are metabolised. The genomes show humans, mice and rats each produce different amounts and types of detoxifying chemicals. For example, cytochrome P450s (CYPs) have been identified as key substrates for breaking down and removing drugs from the system. In humans, CYP2D6 is an important molecule for metabolising a range of drugs like Prozac and codeine. People deficient in this CYP enzyme are very likely to have adverse drug reactions.

The genetic sequence in humans contains the biological code to produce 57 CYPs; the mouse genome contains the code for 102 and rats have fewer at 53. In addition to the different numbers, rats and mice also produce different types of CYPs to humans. CYP2D6 is our only active 2D enzyme, while in mice there are nine versions; humans have four CYP2c enzymes while mice have 15. These factors suggest that, even using genetically modified mice, which express specific diseases, or inbred rats with engineered mutations, drugs that may combat the disease in humans could be ineffective or toxic in rodents and never make it past the first animal testing phase.

David Nelson, an expert in CYP analysis from the University of Tennessee, Memphis, US, agrees. ’The number of P450 genes in mice is much higher than humans,’ he states. ’It seems unlikely that the substrate specificities and regulation of these additional genes will be the same as in humans.’

So would chimpanzees, our closest ancestral relatives, be better test subjects? The low-quality draft genome sequence was produced in December 2003, with the high-quality draft sequence due soon. Physiologically, the chimpanzee is much closer to humans than rodents are. Comparing the human and chimp genomes has shown our genes to be 98 per cent similar; much higher than the mouse or rat. Theoretically this makes the chimpanzee a better test study for human disease. So far, genetic differences have been identified between the species in the genes responsible for sense of smell, hairiness, digestion and hearing, among others. The latter gives insight into why chimps are difficult to train to understand human speech. Mutations in the human gene alpha tectorin cause poor frequency response in the ear and lead to congenital deafness. Chimp alpha tectorin behaves significantly differently to the human variety, suggesting chimp hearing may not be as well developed as ours.

Large-scale searching for comparisons between chimp and human disease genes is on-going and, as yet, incomplete. It is already known that chimpanzees do not suffer from human diseases such as malaria and AIDS, while they do suffer from diseases that do not affect humans. Scientists are keen to discover the reasons for the species barrier. Nelson has already begun to investigate CYP coding regions in the draft chimpanzee genome and concludes they are almost identical to human CYP coding regions. The number of CYP enzymes produced is also likely to be very similar.

However, even though chimpanzees are likely to be good models for human disease, they are unlikely to become the animal of choice for drug testing and research. ’Chimps are too rare and too precious to use as experimental animals for drug studies,’ says Nelson. Coupled with the public feeling on animal testing, as shown by the demonstrations against a planned primate research facility in Cambridge, UK, it seems realistic to assume chimpanzees will not replace mice as the most common test animal, even if they are more appropriate.

So, where does the information gained from comparative genome analysis leave the world of drug development? Mice are unlikely to be replaced by other animals in testing laboratories because their genes can be easily manipulated and they can be reared and handled simply and cheaply, but comparative genome analysis shows that drug development will continue to be complicated by human/rodent differences.

’An alternative is to "humanise" the mice by replacing the mice CYP gene clusters with human gene clusters,’ says Nelson. This would overcome the problem of mouse CYP enzymes giving different responses to drug molecules than human CYP enzymes, but the physiological differences would still remain. Research into this area is continuing, but this approach may prove to be the best compromise.

There is little doubt that comparative genome analysis is a powerful technique for determining species similarities and differences. With the sequencing of over 200 genomes completed and 1000 on-going, insights into evolution will also be gained. Knowing how our genome differs to others will help to explain what makes us human and give scientists more chance in the fight against disease.

Acknowledgements

Andrew West is a PhD student at Leicester University, UK, and recipient of the RSC Marriott studentship