Gene names: the approaching end of a century‐long dilemma

The trouble started, of course, with Thomas Hunt Morgan. Yet, it would be unfair to blame him. Who amongst us, faced with the same problem, would not have done precisely the same? Nevertheless, he began a tradition whose legacy has been nearly a century's worth of subtle confusion and misunderstanding. Morgan's problem seemed simple enough. He had to choose an appropriate name for the first discovered mutant gene. The mutation caused the production of white-eyed fruit flies, either when in single copy (hemizygotes) in males or, when present in two copies (homozygotes) in females. (The gene is sex-linked, that is carried on the X chromosome.) What should he call the mutant gene? Well, the choice was obvious: white or, as abbreviated, w. The corresponding normal form or allele, the ``wild-type'', Morgan designated w‡ . Thus began the long tradition of naming genes with respect to their mutant phenotype. In effect, the role of the wild-type gene, the form fashioned by evolution, comes to be defined by what happens when its activity is not available to the orgenism. This, as anyone can see, is a highly indirect, and therefore dubious, way to characterize a gene. Perhaps the back-to-front character of this procedure is not too serious when the mutant phenotype is a direct consequence of the absence of the wild-type gene product, as in the case of w. It becomes highly problematical, however, when the mutant phenotype is the outcome of a long and complex series of events. An analogy, from a tale attributed to Benjamin Franklin, illustrates why. The story recounts a simple chain of events: for want of a nail, a horseshoe was lost; for want of a horseshoe, a horse was lost; for want of a horse, a rider was lost; for want of a rider, a message was lost; for want of a message, a battle was lost; for want of a victory, a kingdom was lost. The idea that big consequences can follow from small events is absolutely valid. But to conclude from the story that the function of horseshoe nails is to prevent the loss of kingdoms would be silly. Many gene names, however, have equivalent implications. There is another kind of confusion that creeps in. Many genes are named for their first discovered mutant phenotype. Long use of that gene name subtly reinforces the notion that the function of the wild-type gene is to participate in the process designated by the name. Apart from the fact of indirect consequences described above, two difficulties arise. First, many gene products have multiple uses. Second, where there are different rolesÐeither from different usage or from multiple biochemical rolesÐthese can be differentially affected by different mutations. When one considers these additional factors, it is easy to see just how large the potential for confusion associated with a particular gene name is. This situation is illustrated in the ``Problems & Paradigms'' article by Justin Courcelle, Philip Hanawalt and Ann Ganesan, in this issue. They discuss how the naming of the recA gene of Escherichia coli has shaped perceptions of the function of this gene product and why some rather different interpretations of its roles may now be in order. The fundamental problem is that newly-identifed genes do require names, to permit communication about them, but that these names have tended to do double-duty. They have been picked to act as both signifiers and as descriptors. As a signifier, a gene name has to identify unambiguously a particular gene, such that it is distinguished from all previously identified genes. As a descriptor, the gene name is intended to tell you something about the properties of the gene. When mutant phenotypes were all that there was to go on, these were, inevitably, drawn upon to provide quick, shorthand descriptors. Yet, for all the reasons enumerated above, such descriptive elements are both incomplete and misleading. In a sense, this situation has been bad enough for biologists, in creating subtle distortions in the ways that they think about the functions of particular genes. But it has had an additional spill-over effect in popular perceptions about genes and gene actions, which may be even more serious. In the early `90s, in particular, there was a host of claims, based on statistical evidence, that there were specific genes ``for'' particular traits, ranging from shyness and thrill-seeking to gender-preference and even criminality. The implication was always that they were special single genes ``for'' these traits. In virtually none of the press reports was it made clear that the scientists were talking about allelic variants of particular normal (wild-type) genes that might, indirectly and subtly, cause some alteration. The result is that many members of the public apparently think that there are specific aberrant genes that exist only inparticular individuals rather thanvariant alleles that might affect certain behaviors in certain circumstances. In most of these cases, the neuro-developmental subtleties involved in these situations were completely swept under the carpet, at least in the journalistic accounts. This stuff is not only pernicious in its support of rigid genetic deterministic thinking but it has considerably increased public confusion about genes and the basis of mutant effects. Hence, the tradition of gene-naming that Morgan unintentionally founded has a lot to answer for. The good news, however, is that this era of confusion is probably coming to a close. The agent of this change is genomics. Ideally, in future,