Olfaction plays a pivotal role for many organisms. It is used to detect food, conspecifics, mates, as well as threats. However, the sense of smell is obviously more important to certain animals than others. Birds (with the exception of tube-nosed seabirds, Procellariiformes, and a few members of other groups such as New World vultures or the flightless and nocturnal kiwis) have traditionally been thought to possess an olfactory system inferior to, e.g., mammals, reflecting their overriding reliance on visual and sound cues, exemplified by elaborate and colourful feather plumages (e.g., birds of paradise) and complex song patterns (e.g., nightingales) [for reviews, see [1
]]. Recent findings have questioned this view, as it turns out that many groups of birds, including songbirds, Passeriformes, apparently have an acute sense of smell and accordingly are likely to also rely on olfaction in their daily life [4
]. It is important to note that these findings do not imply that all birds have similar olfactory systems and abilities. Instead, one can expect that even closely related species experienced different selection pressures on olfactory abilities. This notion was recently supported by the finding that the olfactory receptor gene repertoire of the nocturnal kakapo, Strigops habroptilus
, that is known to rely heavily on olfactory cues is larger than that of its related, but diurnal relatives (kea, Nestor notabilis
, and kaka, Nestor meridionalis
). This result strongly suggests that e.g. ecological niche adaptations such as daily activity patterns can lead to different olfactory abilities in close relatives [8
Among vertebrates, the sense of smell is mediated by olfactory receptors (ORs) expressed in sensory neurons within the olfactory epithelium [9
]. The coding region of OR genes is small (approximately 1000 bp) and intronless [9
]. Surprisingly, in-silico mining of the first draft of the chicken (Gallus gallus
) genome (release February 2004) revealed a very large gene family encoding putative ORs [12
]. Interestingly, the vast majority of these OR genes are confined to a large expansion of a single branch (the so called γ-c clade), in contrast to other vertebrates, where the OR genes are scattered across multiple clades [13
]. This organization of the chicken OR subgenome was also hinted at in another study on a diversity of nine bird species from seven different orders (Anseriformes, Apterygiformes, Cuculiformes, Galliformes, Passeriformes, Procellariiformes, Psittaciformes), including the chicken [15
]. However, the previous study was based on PCR and degenerate primers. Due to limitations of the PCR based method (e.g., primer bias), the previous study may have provided a biased representation of the OR repertoire with respect to the family composition. Thus, it is likely that the results obtained from whole genome searches are more accurate. Genome searches are also needed to show whether the large expansion of OR genes is indeed a common feature of birds and whether it is exclusive to birds within the sauropsid lineage. Finally, a genome-wide approach allows the mapping of the identified OR genes to specific chromosomal locations.
The availability of a second bird genome, that of the zebra finch (Taeniopygia guttata)
, provides the opportunity for a genome-wide comparative investigation into the structure of the bird OR family. To our knowledge, the OR gene repertoire of this avian model species has not yet been investigated. A genomic analysis of the OR genes from the zebra finch will reveal whether its OR gene repertoire is similar to the previously estimated OR gene repertoires from two other songbirds, that of the blue tit (Cyanistes caeruleus
) and the canary (Serinus canaria
]. Furthermore, the recent sequencing of the first reptilian genome (the Anolis
lizard, Anolis carolinensis
) allows for a comparison across the sauropsid lineage.
We here report on an exhaustive search for candidate OR genes from the draft genomes of the chicken (release May 2006), zebra finch (release July 2008) and green anole (release February 2007). The use of the recently improved chicken genome assembly improved previous estimates of the OR gene repertoire in the chicken. We identified a larger number and a substantially higher proportion of intact OR gene homologs in the chicken genome than previously reported [13
]. We show that the expanded γ-c clade found in chicken is also present in the zebra finch genome. This expansion appears to be bird-specific and not sauropsid-specific, as it is lacking from the lizard genome. We also demonstrate that the γ-c clade has been subjected to adaptive evolution. In addition, and surprisingly, the γ-c expansions of the two bird species do not intermix, i.e., they are lineage-specific. Finally, we show that the green anole has a comparatively small OR gene repertoire compared with other terrestrial vertebrates. Our findings raise the question why birds have evolved a special clade of species-specific OR genes. The function of these genes in relation to the birds' reliance on smell remains unknown.