Advertisement
Research Article

Accelerated FoxP2 Evolution in Echolocating Bats

  • Gang Li,

    Affiliations: School of Life Science, East China Normal University, Shanghai, China, Institute of Zoology and Graduate University, Chinese Academy of Sciences, Beijing, China

    X
  • Jinhong Wang,

    Affiliation: School of Life Science, East China Normal University, Shanghai, China

    X
  • Stephen J. Rossiter mail,

    To whom correspondence should be addressed. E-mail: s.j.rossiter@qmul.ac.uk (SR); syzhang@bio.ecnu.edu.cn (SZ)

    Affiliation: School of Biological and Chemical Sciences, Queen Mary, University of London, London, United Kingdom

    X
  • Gareth Jones,

    Affiliation: School of Biological Sciences, University of Bristol, Bristol, United Kingdom

    X
  • Shuyi Zhang mail

    To whom correspondence should be addressed. E-mail: s.j.rossiter@qmul.ac.uk (SR); syzhang@bio.ecnu.edu.cn (SZ)

    Affiliation: School of Life Science, East China Normal University, Shanghai, China

    X
  • Published: September 19, 2007
  • DOI: 10.1371/journal.pone.0000900

Reader Comments (4)

Post a new comment on this article

From John D. Pettigrew, University of Queensland

Posted by PLoS_ONE_Group on 31 Dec 2007 at 15:08 GMT

FOXP2 Destroys the Microbat Paraphyly Hypothesis: A New Tool for Bat Phylogeny.

Phylogenetic Weakness: This study is weakened by its phylogenetic assumptions. Although it is stated in the paper that the phylogeny of bats is increasingly well resolved [1], a compositional artefact corrupts recent phylogenetic arrangements of bats [2, 3, 4].
Ironically, the present data on the amino acid substitutions in the FOXP2 gene provide their own excellent phylogenetic signals which explode these recent arrangements of bats, such as the paraphyly of microbats and the Yinpterochiroptera-Yangochiroptera dichotomy.

Microbat Paraphyly is a Compositional Artefact: According to a number of recent DNA sequence studies, microbats are not monophyletic, but instead are divided into two groups [5,6,7]:- A. Rhinolophoids: 3 families (Hipposideridae, Rhinolophidae, Megadermatidae) which are characterised by close affinity to the megabats, and B. an outlying group composed of the remaining 15 families of microbats. This “paraphyly of microbats” is regarded with scepticism by many, despite the apparent strong support from DNA sequence data, because there is not a single character from outside DNA that supports a link between rhinolophoids and megabats, and molecular studies with proteins contradict it; e.g.[8]. Instead, both these groups have the most biased genomes amongst vertebrates, with both sharing such a high AT content (75% in megabats and not much less in rhinolophoids) [9,10] that the phylogenetic distance between them is underestimated by a factor of 2 in DNA-DNA hybridisation studies [2,3]. An underestimate of this magnitude will affect all tree-building algorithms, whether parsimony, maximum likelihood, Bayesian or logdet, since all begin the tree building exercise using the two closest taxa.

Warnings about this potential artefact have been largely ignored, so that the new DNA trees of bats are widely accepted despite their lack of congruence with other disciplines. This is in contrast to other examples of DNA bias where accommodations were made in the DNA rather than in the phylogeny. For example, Amphioxus appears outside the echinoderms in DNA trees despite its well-accepted status as a cephalochordate [11], and the eucaryote, Dictyostelium, is strongly placed with the prokaryotes by DNA sequence data because of its high AT content [12]. In these cases, the corrections involved use of codons rather than individual bases, a relevant point to the phylogenetically-informative amino acid substitutions that are provided by the present FoxP2 genes.

FOXP2 Explodes Microbat Paraphyly: There is no support whatever for the paraphyly of microbats from the numerous unique amino acid substitutions of FoxP2. While megabats have a number of derived amino acid substitutions in FoxP2, not a single one is shared with any of the many that characterise the rhinolophoids. Worse still for paraphyly, the Megadermatidae, which are united by the paraphyletic idea to rhinolophids, hipposiderids and megabats, are instead joined by three different sets of substitutions to the Vespertilionoidea, demolishing both the paraphyletic arrangement as well as the Yinpterochiroptera-Yangochiroptera dichotomy (in the paraphyly proposal megadermatids are classified as Yin, as are megabats, while vespertilionids are Yang).

FOXP2 Substitutions Order Microbats Families: There are a number of unique substitutions that link the molossids (freetail bats) to the vespertilionids. This arrangement (Vespertilionoidea) has been put forward on a number of occasions on the basis of morphology and so sits well with many bat taxonomists, in contrast to the paraphyly arrangement. Similarly, there are links between megadermatids and vespertilionoids, as well as a possible link of phyllostomids to this same superfamily.

One case where phylogeny is not an issue is the substitution in Pteronotus parnellii that is absent from three sister taxa in the same genus. This could form a basis for the authors’ argument that FOXP2 is somehow connected to the sophisticated aspects of echolocation, since P. parnellii parnellii is the only New World microbat to have mastered Doppler shift compensation for the capture of insects in leaf clutter. On the other hand, a single substitution is not much to hang one’s hat on, particularly when one of the non-echolocating megabats has many. This substitution in P. parnellii is “I” (isoleucine), an amino acid codon in the FYMINK family of AT rich codons [9], so it might reflect the strong AT bias of bats, perhaps exaggerated in P. parnellii if it is more metabolically active on account of its new-and-improved echolocation

Ancient Origins of Microbats Could Explain the FOXP2 Diversity: An alternative to the vocal hypothesis that could account for the impressive divergence in bats of the highly-conservative FOXP2 gene is the likely ancient origin of microbats. The fossil record of bats is poor, but the oldest microbats, at around 50 MyA, have wings and cochleas like modern bats [13], so bat history is clearly much older than that. This inference is supported by circumstantial evidence in the form of Cretaceous inventions of insect ears for detecting the ultrasonic cries of microbats [14,15]. Dixie Pierson’s [16] excellent immunological studies of proteins in many families of microbats placed their divergence at around 100MyA, a figure that could be reconciled with the FOXP2 diversity, which is roughly 2.5X that of the whales, as predicted from relative ages of 40 vs 100 MyA [17]. Pierson’s figure is likely to be more accurate than uncorrected estimates from bat DNA, which is subject to the gross underes timates already mentioned.

So it is possible that chiropteran FOXP2 has had plenty of time to diverge rather experiencing accelerated evolution, given the limited present understanding of the origins of microbats, which neither have an unequivocal sister taxon that might provide a date from molecules, nor a fossil record that illuminates those origins.

References: see http://www.uq.edu.au/nuq/...

Prof. JD Pettigrew FRS
Vision Touch and Hearing Research Centre School of Biomedical Sciences
University of Queensland
www.uq.edu.au/nuq/jack/ja...