Yesterday's post on the cancers likely to be induced by CT (sometimes called "CAT') scans in medicine raised the feline image, and we'll pursue it a bit further, but in a totally different way: about the biology of copy cats.
One of the recent flurries in genetics, which is using up a lot of ink, paper, publicity, and grant funding, has to do with what is called copy number variation (CNV). People are finding that genomes are loaded with additions or deletions of segments of DNA such that in some human genomes there will be one copy of a segment with whatever genes or other functional units it contains, and other human genomes will have that segment duplicated (or even sometimes multiplied more than that) or deleted.
The fact that we vary either between the two copies of the genome that we each carry (one inherited from each parent), or between the genomes of different people, explains the 'V' in CNV--for 'variation'. The realization that there's all this variation is very different from the standard, stereotypical view of 'the' human genome, which doesn't actually exist: each of us has two sets of chromosomes, no two of which are alike. 'The' human genome is an arbitrarily accepted reference sequence, not even of any one particular individual, that we use for comparative and other purposes. Sequences that differ from this reference sequence are then called variants.
CNVs are potentially important discoveries and of course, for various reasons, have become the focus of searches for genetic causes of diseases that refuse to be mapped the old fashioned way (by variation in specific gene sequences). Some evidence for this has been found, and it may prove useful, at least in relation to some disorders.
But the ballyhoo over CNVs reveals a few things that are telling about the way science works, and in a sense, about the not-so-profound level of understanding that is widespread even in much of the biomedical community. It's the idea that we have a particular genome structure that varies by single nucleotide changes (the standard idea of 'mutation'), or else by unusual one-off chromosomal anomalies (such as in the triplicated chromosome 21 found in Down Syndrome).
That simple idea of inheritance was the classical one, largely due to Mendel's work in peas that showed that variation is due to altered states of specific entities we now call 'genes'. It seemed valid, more or less, until a bit past the middle of the last century, and it is still widely assumed and taught today.
Thus, when CNVs are discovered, much ado is raised as if this is some sort of paradigm-shifting change in our thinking. But that's not true. Even by around 1970 it had become clear that most new genes arise as duplications of various sorts of existing genes. The highly structured nature of genes (with coding and non-coding parts and so on) would be very unlikely to arise by accumulations of single nucleotide mutations. What happens is genes change by mutation, but arise by duplication (we discuss this and some of its nuances in our book, because regulatory regions seem often to arise by the latter, classical form).
In fact, it appears that in many living lineages including our own vertebrate ancestry, there have been duplications of the entire genome.
We've known for decades about the gene 'families' that result from duplication and, essentially, that all genes are members of gene families with this kind of history. For the reason of that universality, there's no point in citing an 'example'.
So what's the big deal? Nobody can be surprised that gene copy numbers vary, since we've known for 50 years that that is how genomes evolve. This obviously means that copy numbers must be variable in their population until the ones that eventually become fixed and hence part of their species' stereotype genome sequence do become fixed.
Unless you don't believe in evolution, or believe that it stopped in ancient times, which would almost have to be explained by some kind of made-up theological view, you'd have no reason not to expect that the same must still be going on today. CNVs would have to have some kind of frequency and population distribution that, like those of any single-nucleotide mutations, reflected their effect, population size, and other things that control the change of frequency of genetic variation.
And how can it be a surprise that CNVs have effects on traits? After all, the members of gene families that we have long known of, like the hemoglobin genes and countless others, have different functions: the more ancient their duplication origin sometimes the more different those functions are.
Nor can it be a surprise that a sudden copy number change might alter various cellular balances that were established over long time periods, and hence could be related to disease.
CNVs might not have been specifically predicted in their current biomedical obsession, but they have long been entirely predictable. We knew gene duplication occurred, and their subsequent dynamics necessarily means they are variable in their respective populations.
CNVs may or may not turn out to be highly important in a public health sense, but they seem surely to be so in some specific instances. Technology has enabled their discovery to accelerate and provides ways to look for CNVs. But they are in no way surprising or conceptually new. The conceptual issue is how, in our thirst for simplicity, anyone could be surprised. Indeed, if genetics and evolution are taught properly, CNVs are simply part of the way that evolution works.
The copy cats have been those books and instructors (yes, including yours truly) and classes that somehow ignored these things. We taught that once genes were duplicated they could accrue different functional variation, but we simply ignored or didn't think about the fact that each one that appears to be different between species had to have a long sojourn of polymorphism on the way to becoming fixed. And the new copy cats are copying the fad and treating CNVs as if they were some sort of astounding new finding in our understanding of life. That's mainly grantsmanship.