Several studies show an episodic acceleration of evolutionary rates in protein coding
genes. Acceleration-deceleration patterns were reported in the evolution of globin and
cytochrome c in vertebrates (Goodman 1981), the vesicular stomatitis virus in cattle
(Nichol et al. 1993), lysozymes in primates (Messier and Stewart 1997), and the GPDH
gene in the Diptera (Ayala et al. 1996). Bursts of rapid changes were also observed in the
growth hormone in primates (Liu et al. 2001; Wallis et al. 2001) and other protein
hormones in mammals (Wallis 2000, 2001). Episodic changes of substitution rates in
proteins are usually explained by positive adaptive selection, and often viewed as an
argument against the neutral allele theory and the molecular clock hypothesis (Gillespie
1984, 1993). However, the evidence for a functional impact of rapid changes is not
always clear, suggesting that positive selection may not be the only cause of episodic
acceleration of substitution rates. For example, the episodic evolution of xanthine
dehydrogenase was suggested to be related to the particularities of the genomes in which
the locus is embedded (Rodríguez-Trelles et al. 2001).
Episodic variations of substitution rates have also been reported in large- and small-
subunit ribosomal RNA (LSU and SSU rRNA) genes, the universally conserved function
of which excludes positive selection. A 10-fold acceleration of substitution rates, probably
caused by a change of DNA base composition, was reported in ribosomal genes of the
dipteran stem-lineage (Friedrich and Tautz 1997). Here we present the example of an
even faster (>30-fold) episodic change in evolutionary rates of the SSU rRNA gene in
Foraminifera.
Foraminifera as a tool to study evolutionary ratesForaminifera are the most important group of microfossils widely used in
micropalaeontology for the stratigraphic analysis of ancient sediments and for
palaeoecological and palaeoclimatic reconstructions (Culver 1993; Wilson and Norris
2001). Foraminiferan tests are abundant and widespread in sediment samples. Species
identification is sometimes difficult, but the recognition of genera and higher taxonomic
units is usually relatively easy. This makes Foraminifera an excellent tool for stratigraphic
purposes. The quality of the foraminiferan fossil record is particularly good for planktonic
species, in which occurrences in marine strata have been precisely calibrated by radiometric
dating and tested with independent biochronological datasets (Berggren et al. 1995).
Despite the abundance and diversity of Foraminifera, their use in molecular
evolutionary studies is obstructed by technical problems. Obtaining large amounts of
foraminiferan DNA is hindered by their slow growth and the difficulty of getting them to
reproduce in laboratory cultures. Moreover, because many shallow water Foraminifera
live in association with symbiotic algae and epiphytic micro-organisms, their DNA
extractions are often contaminated with foreign eukaryotic DNA. At present, the
foraminiferan DNA database is composed mainly of ribosomal genes, for which specific
foraminiferan PCR primers exist (Pawlowski 2000) and whose large number of copies
allows amplification even from single cells. The first sequences of foraminiferan protein-
coding genes, including actin, tubulin, and RNA polymerase, have been obtained only
recently (Pawlowski et al. 1999; unpublished data). For this reason, the present study is
limited to analyses of ribosomal data.
110 JAN PAWLOWSKI AND CÉDRIC BERNEY