Patterns of microchromosome organization remain highly conserved throughout avian evolution

The gross structure and organization of the genome of any individual species (both at a karyotypic level and in interphase nuclei) have broad functional significance. Specifically, the number and shape of chromosomes as well as the order of genes thereon can have an impact on the evolution, variation and phenotype of that species. In this regard, microchromosomes, present in many terrestrial vertebrates (but not mammals nor crocodilians), remain largely uncharacterized. Birds have a highly distinctive, ‘signature’ avian genome structure (karyotype) which is typically divided into ~ 10 macrochromosomes and ~ 30 evenly sized, morphologically indistinguishable microchromosomes (Christidis 1990; Masabanda et al. 2004; Griffin et al. 2007). The number and morphological similarity of avian microchromosomes are near-unique in nature, and they are impossible to distinguish using classical cytogenetic approaches. In fact, to date, although around 1000 karyotypes have been published for birds (a phylogenetic class that with ~ 10,500 extant species), these are all partial, with only 5–10 pairs of chromosomes readily identifiable. Nonetheless, this apparent degree of stability across a group with enormous phenotypic diversity is, at best, only inferred through a comparison of chromosome number from one bird to another. Exceptions to this rule include the Falconiformes (falcons) and the Psittaciformes (parrots), both of which have reduced diploid numbers relative to the norm, along with fewer microchromosomes. While this suggests evidence of chromosomal fusion (Nishida-Umehara et al. 2007; Nanda et al. 2007), the nature of these fusions is only recently being uncovered. Comparative genomics of avian macrochromosomes (up to chromosome 9) has been refined beyond karyotyping through the development of chromosome paints derived from amplification and fluorescence labelling of chicken macrochromosomes isolated by flow-cytometry (Griffin et al. 1999). Cross-species analysis on over 70 avian species from 15 different orders has revealed a remarkable lack of inter-macrochromosomal rearrangements. Some success has been achieved using microchromosomal paints (Lithgow et al. 2014), but this is limited to analysis of ‘pooled’ chromosomes (where each pool contains between 2 and 10 pairs of microchromosomes) due to the technical difficulties of separating individual microchromosomes by flow cytometry. A degree of success using a cross-species BAC (bacterial artificial chromosome) mapping approach has been reported, although until recently, this has been limited to closely related species, with 70% success rates reported using chicken BACs on turkey (Mealeagris gallopavo) (Griffin et al. 2008) reducing to 40–50% when tested on duck (Anas platyrynchos) (Fillon et al. 2007; Skinner et al. 2009). Investigation of the microchromosomes of the chicken and the zebra finch using a BAC mapping approach has also been reported by Völker et al. (2010), but this was achieved using the same specie hybridization with the comparative genomics performed in silico. As such, this approach can only be performed when both species have characterized BAC libraries, and even then, the very smallest microchromosomes (the so called ‘D-group’) have not been sequenced and would therefore not be amenable to this (to a BAC mapping) approach. We recently reported an approach to isolate BACs that hybridize in a universal manner across all bird species (Damas et al. 2017); however, to date, this has only been applied to five species (Damas et al. 2017; O’Connor et al. 2018b). To date, therefore, a meaningful comparative analysis of the majority of avian microchromosomes, in a range of species, has yet to be performed.

At a functional level, despite microchromosomes accounting for only 23% of the average avian genome size, they contain around 50% of genes and are thus highly GC-rich (average 48%) and gene-dense (McQueen et al. 1998; Smith et al. 2000; Habermann et al. 2001; Burt 2002; Warren et al. 2017). In the chicken (the most characterized of avian species), macrochromosomes range in size from ~ 23 to 200 Mb, but microchromosomes are, on average, only 12 Mb in length, the smallest being ~ 3 Mb (Hillier et al. 2004). Microchromosomes have been demonstrated to have a significantly higher recombination rate than macrochromosomes (Hillier et al. 2004; Backström et al. 2010) and, in the interphase nucleus, they appear to cluster in a central position, with the macrochromosomes occupying the nuclear periphery (Habermann et al. 2001; Federico et al. 2005; Skinner et al. 2009). Whether this is a function of their small physical size, or the fact that microchromosomes have a greater gene density (thereby more able to access the transcriptional machinery), is still not clear. Skinner et al. (2009) attempted to address this statistically, analysing chicken chromosomes that were outliers to the ‘small=more gene dense rule’. An alternative approach, however, is to analyse ‘former’ microchromosomes (i.e. those that have since fused in evolution to become part of a larger chromosome) such as those seen in Falconiformes (Damas et al. 2017; O’Connor et al. 2018b), which largely retain their inherent microchromosomal properties such as gene density, GC content, and recombination rate (Hillier et al. 2004).

Microchromosomes are thought to have originated ~ 400 mya (million year ago) in the ancestral vertebrate karyotype (Burt 2002). Bioinformatic reconstructions demonstrated that avian microchromosomes corresponded directly with gnathostome ancestor protochromosomes (Nakatani et al. 2007), suggesting that they have remained remarkably unchanged throughout evolution. Indeed, in a recent study, we suggested that the typical avian-like pattern became mostly established before birds and turtles diverged and was present in the theropod dinosaur lineage (O’Connor et al. 2018a). However, the lack of cytogenetic tools to characterize the microchromosomes described has impeded confirmation of this by direct evidence. The purpose of this study was therefore to test the hypothesis that inter-microchromosomal rearrangement is a rare phenomenon in avian evolution, i.e. that the genomic structure (karyotype) is widely stable (unchanged) across most species. Further, we hypothesized that fused ancestral former microchromosomes occupy a central nuclear position i.e. that their nuclear organization is a function of their genomic properties, not their physical size.

Results presented here demonstrate that the extraordinary degree of genome stability evident in the macrochromsomes of birds also extends to the (previously intractable) microchromosomes, illustrating a hitherto undiscovered overall genome stability (i.e. consistency of karyotype) that is rarely seen in other classes. As demonstrated in Fig. 2, our analysis of over 22 species across 10 avian orders reveals that 8 of the 10 orders demonstrate no change from the microchromosomal pattern seen in the chicken. Of the two orders where interchromosomal rearrangements had previously been identified using macrochromosome paints (Falconiformes and Psittaciformes (Nishida-Umehara et al. 2007; Nanda et al. 2007)), we have now been able to identify the ancestral microchromosomes involved in those rearrangements.

Of the avian species that exhibit microchromosomal rearrangements, the three representatives of the Falconiformes tested here (the saker, gyr and peregrine falcons) share the same pattern of fusion. This would suggest that the common ancestor of the falcons had the same karyotypic structure and that there has been little interchromosomal change since. Of the other highly rearranged order, the Psittaciformes, the microchromosomal fusions exhibited in each of the species tested differ from one another, suggesting that karyotypic evolution has continued from their common ancestor and that there are species-specific rearrangements. However, in all of these cases, it appears that there is a pattern of microchromosomes remaining as discrete units even when fused into highly complex karyotypes. Interestingly, this same pattern is also evident in the chicken, where the p-arm of chromosome four is a microchromosome in most other species, whilst retaining its uniquely microchromosomal characteristics such as high GC and gene content (Hillier et al. 2004).

Even taking into account the lineage-specific rearrangements, there appear to be five microchromosomes that across all birds tested, remain as microchromosomes with no signs of apparent fusion. In the chicken, these are five of the smallest sequenced chromosomes (GGA22, 24, 25, 26, 27) with sizes ranging from 4 to 6.5 Mb. Further sequence analysis may reveal signature features of these chromosomes that indicate a biological reason as to why these chromosomes are left intact. If there is any correlation with the size of the chromosomes and their lack of interchromosomal rearrangement, then this would suggest that the very smallest D-group chicken microchromosomes (Masabanda et al. 2004) which do not to date have sequences associated with them are also less prone to chromosomal fusion.

The remarkable conservation of genome organization at a karyotypic level, and even more so at a nuclear level in birds (and possibly other dinosaurs (O’Connor et al. 2018a)) is a situation rarely reported in nature. It is certainly not the case in mammals, the most studied of clades, which display great variation in karyotypic and nuclear organization. Conservation for such a long time period implies evolutionary success, and we are aware of no other animal group for which this applies to such a degree.