Abstract
Nurminsky and Hartl reply — One hallmark of persistent strong background selection is a severe diminution of codon usage bias1,3. An example is the Drosophila gene rolled (gene 1 in Fig. 1), which is located in the centromeric heterochromatin of chromosome 2, where recombination is severely restricted. Other genes shown in Fig. 1are located near the base of the X chromosome. Genes 10 and 11 are AnnX and Cdic, respectively, which flank the Sdic gene4. As pointed out by Charlesworth and Charlesworth, gene 2, which is su(f), has much less DNA sequence variation than gene 17 (Zw). This difference is consistent with the discordant levels of codon usage bias and suggests strong background selection at su(f) but not at Zw. The degree of codon usage bias shows an extremely sharp increase as the gene locations progress outwards from su(f). The transition is near genes 5 (sol) and 6 (slgA), which are proximal to the Cdic-AnnXregion.
Main
The result is that AnnX and Cdic are located in a region of codon usage bias similar to that of Zw. The cytological region 19DE might therefore support a level of DNA sequence variation of Sdic and Cdic comparable to that of Zw. However, based on the amount of polymorphism observed for Zw, the probability of obtaining a value as low or lower than that for Sdic and Cdic is about 0.043 and 0.008, respectively. These estimates are based on 10,000 simulations using Watterson's formula5 for pairwise mismatches in the infinite-alleles model with no recombination, so they should be conservative. There seems to be a statistically significant difference between Zw and the other two genes.
The evolution of Sdic required an initial duplication and gene fusion accompanied or followed by three deletions, two more insertions/deletions (including one that created a new splice junction), 11 nucleotide substitutions (including reversal of a chain-terminating codon), and a tenfold tandem reiteration of the Sdic coding sequence. Although all these changes may have occurred shortly after the divergence of D. melanogasterand D. simulans, the similarity in degree of codon usage bias of Cdic and Zw, contrasted with the significant discrepancy in their levels of nucleotide diversity, provides independent evidence in support of our original inference of at least one relatively recent selective sweep. The negative values of Tajima's D statistic6 for Sdic and Cdic also support this idea, notwithstanding their lack of statistical significance.
Charlesworth and Charlesworth are correct in pointing out that all the genes in the region 19DE might have limited DNA sequence variation as a result of background selection, despite the higher than predicted level of codon usage bias indicated in Fig. 1. We agree that a much more complete characterization of the levels of genetic diversity and recombination in this region would be informative.
References
Kliman, R. M. & Hey, J. Mol. Biol. Evol. 10, 1239–1258 (1994).
Munte, A., Aguadé, M. & Segarra, C. Genetics 147, 165–175 (1997).
3.Kreitman, M. E. & Antezana, M. in Evolutionary Genetics from Molecules to Morphology (eds Singh, R. S. & Krimbas, C.) (Cambridge Univ. Press, New York, in the press).
Nurminsky, D. I., Nurminskaya, M. V., De Aguiar, D. & Hartl, D. L. Nature 396, 572–575 (1998).
Watterson, G. A. Theor. Popul. Biol. 7, 256–276 (1975).
Tajima, F. Genetics 123, 585–595 (1989).
Wright, F. Gene 87, 23–29 (1990).
Shields, D. C., Sharp, P. M., Higgins, D. G. & Wright, F. Mol. Biol. Evol. 5, 704–716 (1988).
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Nurminsky, D., Hartl, D. reply: How was the Sdic gene fixed?. Nature 400, 520 (1999). https://doi.org/10.1038/22924
Issue Date:
DOI: https://doi.org/10.1038/22924
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.