Transcription of linked genes

Gene linkage may be a reflection of a requirement for co-ordinated cis transcriptional regulation, as has been shown for the globin locus and/or may be a sign of recent evolutionary divergence. The most well-known example of linkage within the GPCR family is provided by the olfactory receptor subfamily. Although there are only 47 human olfactory type GPCRs listed in the GPCR data base, initial estimates accounted for up to 906 olfactory receptor genes of which 60% were scored as pseudogenes (International Human Genome Sequencing Consortium). More recent genome mining studies (Zozulya et al. 2001) have revealed a total of 347 functional olfactory type GPCRs in the human genome. These can be further subdivided into subfamilies on the basis of similarity of coding regions. These subfamilies tend to exist as clusters of around 10 or so members spread over at least 25 chromosomal loci (Ben-Arie etal. 1994; Rouquier etal. 1998; Trask etal. 1998). Many of these clusters are located at the subteleomeric regions near the termini of the chromosomes. Subteleomeric regions of the chromosome are gene rich and show high levels of recombination resulting in increased gene duplication and it has been hypothesized that these regions act as 'olfactory receptor nurseries' (Trask et al. 1998). One consequence of elevated meiotic recombination would be the generation of high levels of pseudogenes—which is exactly what is seen in the olfactory gene subfamily.

In rodents, each olfactory receptor gene is expressed in one of four zones in the olfactory epithelium (Ressler 1994, 1993; Vassar etal. 1994, 1993) and each olfactory neuron is assumed to express only one (or very few) olfactory receptors (Malnic et al. 1999; Ressler, 1994, 1993; Vassar et al. 1994, 1993). Furthermore, allelic inactivation ensures that only one copy of each array is transcribed (Chess 1994). One further observation is that members of individual clusters tend to be expressed in the same region of the olfactory epithelium. Collectively, this would imply that within each cluster there is a mechanism to ensure co-ordinate transcriptional regulation. It was hypothesized some time ago (Chess 1994) that allelic inactivation was consistent with a stochastic cis mechanism of gene activation since without the allelic inactivation, stochastic activation would lead to two different genes within an array being expressed by a single olfactory neuron—a situation that is not seen. By analogy with the locus control region of the globin locus (Grosveld 1999; Grosveld et al. 1998), it was suggested that an equivalent region adjacent to an array of olfactory receptor genes could be responsible for the stochastic activation of a single gene within the locus. Although the mechanisms responsible for this restricted transcription are largely unknown, recent transgenic experiments have shed some light on the types of mechanisms that maybe operative.

Serizawa etal. (Serizawa etal. 2000) generated a line of mice that carried a tagged endogenous MOR28 gene and a 200 kb YAC containing a differentially tagged MOR28 transgene. The salient finding was that transgene and endogenous gene were almost never co-expressed in the same olfactory neuron, thereby demonstrating that mutually exclusive expression did not require cis linkage. The classical picture of gene-specific transcription envisages a combination of transcription factors that uniquely specify activation of that gene locus. One of the most important and least contentious outcomes of this study is the rejection of this notion since such a model would require co-expression of transgene and endogenous gene. Although gene re-arrangement is a formal possibility, it is unlikely. The analogous system here is provided by the antigen-receptor genes in lymphocytes which undergo re-arrangement and deletion in order to bring distal promoter and enhancer elements into proximity to allow transcription. However, no evidence for DNA arrangements is apparent at the olfactory receptor locus. In a commentary by Randall Reed on the work of Serizawa (Reed 2000), it was hypothesized that two events occurred. Stage one was a 'low-probability' event in which one allele is primed for potential expression. Stage two requires transcription of the selected allele. Some 'high-probability' event or product downstream of the transcription prevents priming of other alleles. Although attractive in dispensing with the need for cis linkage in order to maintain stochastic selection ofone gene within an array, the weakness is the absence of any direct supportive evidence.

Another subfamily of GPCRs that appears to be linked are the C-C chemokine receptors. In the human genome 6 C-C type chemokine receptor genes (CCXCR1, CCR1, CCR3, CCR2, CCR5, CCRL2) are arranged in a tandem array within a 400 kb region of chromosome 3 (3p21.3). This linkage group appears to be conserved in the mouse genome where mCCR1, mCCR2, mCCR3 and mCCR5 are all found in a cluster on chromosome 9 (9 72.0 cM). It remains to be seen if this linkage is accompanied by a co-ordinate transcriptional regulation.

0 0

Post a comment