INAUGURAL ARTICLE Microchromosomes are building blocks of bird, reptile, and mammal chromosomes  Paul D. Watersa,1 , Hardip R. Patelb,1 , Aurora Ruiz-Herrerac,d , Lucıa Alvarez-Gonz lezc,d, Nicholas C. Listera , a Oleg Simakove , Tariq Ezazf , Parwinder Kaurg , Celine Frereh, Frank Gru € tzneri , Arthur Georgesf , and Jennifer A. Marshall Gravesf,j,2 a School of Biotechnology and Biomolecular Science, Faculty of Science, UNSW Sydney, Sydney, NSW 2052, Australia; bThe John Curtin School of Medical Research, Australian National University, Canberra, ACT 2601, Australia; cDepartament de Biologia Cellular, Fisiologia i Immunologia, Universitat Auto  noma de Barcelona, Cerdanyola del Valle  s 08193, Spain; dGenome Integrity and Instability Group, Institut de Biotecnologia i Biomedicina, Universitat Auto  noma de Barcelona, Cerdanyola del Valle s 08193, Spain; Department of Neurosciences and Developmental Biology, University of Vienna, 1010 Vienna, Austria; e f Institute for Applied Ecology, University of Canberra, Canberra, ACT 2601, Australia; gUWA School of Agriculture and Environment, The University of Western Australia, Crawley, WA 6009, Australia; hGlobal Change Ecology Research Group, University of the Sunshine Coast, Sippy Downs, QLD 4556, Australia; iSchool of Biological Sciences, University of Adelaide, Adelaide, SA 5000, Australia; and jSchool of Life Sciences, La Trobe University, Bundoora, VIC 3068, Australia This contribution is part of the special series of Inaugural Articles by members of the National Academy of Sciences elected in 2019. Contributed by Jennifer A. Marshall Graves, August 27, 2021 (sent for review July 9, 2021; reviewed by Darren Griffin and Josquin F. G. Tibbits) Microchromosomes, once considered unimportant shreds of the Microchromosomes are GC-rich and gene-dense with a low con- chicken genome, are gene-rich elements with a high GC content tent of repetitive sequence (10) and have high rates of recombi- and few transposable elements. Their origin has been debated for nation. They replicate early and are hyperacetylated compared decades. We used cytological and whole-genome sequence com- to macrochromosomes, suggesting they are highly transcribed. parisons, and chromosome conformation capture, to trace their At the cytological level, most birds have extremely conserved origin and fate in genomes of reptiles, birds, and mammals. We karyotypes, including 9 pairs of macrochromosomes and 30 to find that microchromosomes as well as macrochromosomes are 32 pairs of microchromosomes (3), defined by relative sizes EVOLUTION highly conserved across birds and share synteny with single small where there is no abrupt size discontinuity. Chromosome con- chromosomes of the chordate amphioxus, attesting to their origin stitutions of birds are listed in ref. 11. Although there are some as elements of an ancient animal genome. Turtles and squamates spectacular exceptions, especially in the highly rearranged fal- (snakes and lizards) share different subsets of ancestral microchro- con and the parrot genomes, even distantly related birds such mosomes, having independently lost microchromosomes by fusion as chicken and emu share nine macrochromosome pairs with other microchromosomes or macrochromosomes. Patterns of fusions were quite different in different lineages. Cytological observations show that microchromosomes in all lineages are spa- Significance tially separated into a central compartment at interphase and dur- ing mitosis and meiosis. This reflects higher interaction between Genomes of birds and reptiles, but not mammals, consist of microchromosomes than with macrochromosomes, as observed by a few large chromosomes and many tiny microchromo- chromosome conformation capture, and suggests some functional somes. Microchromosomes are gene-rich and highly con- coherence. In highly rearranged genomes fused microchromo- served among birds and reptiles and share homology with somes retain most ancestral characteristics, but these may erode one or more of the tiny chromosomes of an invertebrate over evolutionary time; surprisingly, de novo microchromosomes that diverged from the vertebrate lineage 684 Ma. Micro- have rapidly adopted high interaction. Some chromosomes of chromosomes interact strongly and crowd together at the early-branching monotreme mammals align to several bird micro- center of cells, suggesting functional coherence. Many micro- chromosomes, suggesting multiple microchromosome fusions in a chromosomes have been lost independently in turtles, mammalian ancestor. Subsequently, multiple rearrangements snakes, and lizards as they have fused with each other or fueled the extraordinary karyotypic diversity of therian mammals. with larger chromosomes. In mammals they have completely Thus, microchromosomes, far from being aberrant genetic ele- disappeared, yet some chromosomes of the basal platypus ments, represent fundamental building blocks of amniote chromo- line up with several microchromosomes, suggesting that somes, and it is mammals, rather than reptiles and birds, that are they are the building blocks of the atypically variable chro- atypical. mosomes of mammals. vertebrate chromosome evolution j whole-genome alignment j Author contributions: J.A.M.G. designed research; P.D.W., H.R.P., A.R.-H., L.A.-G., T.E., chromosome conformation j microchromosome origin j amphioxus C.F., F.G., and A.G. performed research; H.R.P., A.R.-H., P.K., C.F., and F.G. contributed new reagents/analytic tools; P.D.W., H.R.P., A.R.-H., N.C.L., O.S., T.E., A.G., and J.A.M.G. analyzed data; P.D.W., H.R.P., and J.A.M.G. wrote the paper; and P.K. C performed genome assembly of DNA Zoo species using high-throughput chromosome lassic cytological studies described mammalian chromo- conformation capture. somes of a size easily visible under the microscope. Bird Reviewers: D.G., University of Kent; and J.F.G.T., Agriculture Victoria. and reptile karyotypes were strikingly different, with a size dis- Competing interest statement: Four authors (J.A.M.G., A.R.-H., T.E., and F.G.), as well continuity between macrochromosomes, with sizes (3 to 6 μm) as one of the referees, made contributions to a review of the continued usefulness of in the range of mammalian chromosomes and microchromo- cytogenetics in the age of genomics, solicited and edited by Janine Deakin (https:// somes (<0.5 μm) which looked more like specks of dust (e.g., www.mdpi.com/2073-4425/10/8/627). refs. 1–3). These microchromosomes stained oddly and occu- This open access article is distributed under Creative Commons Attribution License 4.0 (CC BY). pied a central position at mitosis (4). See Profile, e2116850118, in vol. 118, issue 43. An early view of microchromosomes as inconstant hetero- 1 P.D.W. and H.R.P. contributed equally to this work. chromatic elements (5), or even not chromosomes at all, was 2 To whom correspondence may be addressed. Email: j.graves@latrobe.edu.au. thoroughly debunked (1, 6–8). Like macrochromosomes, they This article contains supporting information online at https://www.pnas.org/lookup/ possess a centromere and telomeres at each end (with extralong suppl/doi:10.1073/pnas.2112494118/-/DCSupplemental. subtelomeric repeats) (9) and segregate regularly at mitosis. Published November 1, 2021. PNAS 2021 Vol. 118 No. 45 e2112494118 https://doi.org/10.1073/pnas.2112494118 j 1 of 11 identified by banding patterns, chromosome painting, and gene microchromosomes of several reptiles, including snakes mapping (8, 12–15). (python and tigersnake), lizards (spiny tailed monitor and our Microchromosomes are too small to distinguish morphologi- model, the central bearded dragon Pogona vitticeps), and a tur- cally, let alone by G-band patterns, and pairing them is mostly tle (eastern long-neck). We found that microchromosomes guesswork. However, their number is usually constant and were less strongly stained and tended to clump centrally in even, as expected for paired autosomes in diploids. Cytological mitotic and meiotic cells (e.g., in bearded dragon and turtle; examination, using specific DNA probes, suggests conservation Fig. 2 A and B). Several repetitive sequences specifically hybrid- of microchromosomes across 22 avian species (16), and com- ize to most or all the microchromosomes, allowing us to detect parative gene mapping and whole-genome analysis attests to their central position also within the interphase nucleus considerable conservation among distantly related bird groups (Fig. 2C). (17, 18). A few bird species have more microchromosomes, but We also used DNA paints that specifically hybridized to two chromosome painting reveals their recent origin from fission bearded dragon microchromosome pairs (Fig. 2D) to explore (14). Genome sequencing of many bird species now provides their conservation in other species. We found that in the east- unprecedented detail sufficient to compare microchromosomes ern long-neck turtle the two paints hybridized together at the across avian species (19). tip of a macrochromosome, implying that these two sequences Fewer comparative studies of microchromosome conserva- are fused, and fused to a macrochromosome in the turtle line- tion have been done in reptiles, but their genome structures are age; significantly, the paints hybridized to central regions of similar to those of birds, many with an abrupt distinction interphase nuclei in turtle as well as bearded dragon (Fig. 2E). between a few macrochromosomes and many microchromo- Together our observations support the view that microchro- somes (reviewed in refs. 20–22). However, turtles and snakes mosomes are differentiated from macrochromosomes, not only have fewer microchromosomes than birds. There is G-band by their much smaller size but also by their different staining and chromosome painting homology between the macrochro- properties (denoting different sequence makeup and chromo- mosomes of birds and turtles (23), and a close relationship some conformation) and their location together in the center of between the chromosomes of birds and squamates (snakes and the interphase nucleus and dividing cells. lizards) was noted early (24). Gene mapping and sequence Genome Sequence Comparisons. We performed pairwise whole- comparisons reveal many homologous synteny blocks (8, 25), and sequence comparisons show that several microchromo- genome alignments using LastZ (33) to identify syntenic blocks somes are conserved at the sequence level (26). Lizard karyo- (reciprocally best aligned chains) and show conservation of types are more variable; some species have clearly demarcated chromosomes from birds, turtles, and squamates within and macro- and microchromosomes, whereas others show no clear between lineages. The relationships of the species we used are presented in Fig. 1. We define microchromosome according to distinction. published karyotypes (see detailed information about species There are exceptional reptile and bird clades in which no and their chromosomes in SI Appendix, Table S1). In birds and abrupt size difference defines microchromosomes, and the size snakes all assembled macrochromosomes were smaller than 35 range of microchromosomes can also vary between clades. For Mb. In other clades the threshold size of the largest micro- example, eagle and parrot genomes have few microchromosomes chromsome was greater: turtles (45 Mb), lizards with rear- (27) and crocodilians have five very large macrochromosomes and ranged genomes (50 Mb), tegu (75 Mb), and alligator (96 Mb). few chromosomes in the microchromosome size range (28). Sequence comparisons of bird and turtle genomes. Of the bird The origin of microchromosomes has been debated for deca- species with chromosome-scale assembled genomes, we chose des. Initially they were thought to represent some sort of break- emu, chicken, pigeon, jackdaw, and hummingbird, species with down product of “normal” mammalian-like macrochromosomes conserved karyotypes that captured the deepest avian divergen- that existed in amniote, even tetrapod, ancestors (29), and this ces (Fig. 1). Genomes of these five birds display striking homol- view is still expressed (e.g., ref. 30). The alternative view is that at ogy, aligning over >87% of their length (Fig. 3A). The nine least some of them represent the small chromosomes of a verte- macrochromosomes (eight autosomes and the Z sex chromo- brate ancestor 400 Ma, retained intact by several vertebrate clades some) are almost invariant. We observed no fusions of macro- (8, 26, 31). Similarities with the small chromosomes of amphioxus chromosomes and only two macrochromosomes in humming- (the lancelet, an early branching chordate) now suggest a much bird and one in jackdaw that have undergone fission. We earlier origin (32), dating back to at least 684 My since they last conclude that macrochromosomes are highly conserved shared a common ancestor with vertebrates. in birds. With the availability of several chromosome-scale assemblies of There was little variation (28–31) in the number of micro- bird and reptile genomes (10, 19), it is now possible to trace the chromosomes across these five birds (Fig. 1 and SI Appendix, origin and fate of microchromosomes in birds, reptiles, and mam- Tables S1 and S2). Sequence comparisons of assembled micro- mals. We compared the genomes of 7 birds and 10 reptiles with chromosomes show that they too are highly conserved, nearly chromosome-level assemblies, as well as three mammals and an all showing a 1:1 correspondence between all five bird species amphioxus (Fig. 1). These comparisons provide evidence that, (Fig. 3A). The most prominent exception is a microchromo- indeed, microchromosomes represent a set of highly conserved some in all other bird lineages that aligns to chicken chromo- ancient animal chromosomes, whereas macrochromosomes, which some 4p, as previously noted (8, 12), and is significant because are considered “normal” because of their ubiquity in mammals, it also has homology to the conserved region of the mammalian have undergone multiple lineage-specific rearrangements, espe- X chromosome (34). Uniquely, the hummingbird genome con- cially in mammals. We gather evidence that microchromosomes tains two chromosomes in the microchromosome size range retain a high frequency of interchromosome interaction inside the that are parts of macrochromosomes in other birds, implying nucleus and regularly locate together at interphase and division, an origin by fission of macrochromosomes. Conserved micro- suggesting retention of an ancestral functional coherence between chromosomes are GC-rich and gene-dense in all species (mean a set of small ancestral chromosomes. of 40 genes per Mb compared to 17 genes per Mb on macro- chromosomes in chicken) (Fig. 4 and SI Appendix, Fig. S1). Results We conclude that microchromosomes and macrochromo- Cytological Observations of Microchromosomes. To broaden our somes are highly conserved between even the most distantly knowledge of microchromosome cytology we examined the related bird lineages separated by ∼110 My. 2 of 11 j PNAS Waters et al. https://doi.org/10.1073/pnas.2112494118 Microchromosomes are building blocks of bird, reptile, and mammal chromosomes INAUGURAL ARTICLE Chr Macro Micro Genome Common name number assembled (n) assembled (n) size (Gb) (n) % of genome % of genome Common lizard 19 11 (68.7%) 8 (22.5%) 1.34 Common wall 19 15 (87.9%) 4 (9.7%) 1.48 lizard Argentine black 19 5 (77.5%) 14 (15.8%) 1.93 & white tegu Eastern water ? 6 (82.8%) 19 (15.8%) 1.79 dragon Burmese python 18 8 (84.4%) 10 (12.7%) 1.4 Prairie 19 8 (86.4%) 11 (10.4%) 1.3 rattlesnake Indian cobra 19 8 (80.6%) 11 (13.7%) 1.67 Emu 40 9 (72.8%) 31 (26.2%) 1.18 Greater rhea 40 9 (73.2%) 21 (25.8%) 1.15 Chicken 39 9 (75.3%) 24 (23.3%) 1.05 Anna's 37 9 (71.6%) 26 (27.5%) 1.05 hummingbird Jackdaw 40 10 (75.4%) 21 (24%) 1.03 Band-tailed 40 9 (74.2%) 21 (23%) 1.06 pigeon Golden eagle 31 13 (71.4%) 14 (28%) 1.19 American alligator 16 5 (68.2%) 10 (30.6%) 2.14 Green sea turtle 28 11 (77.7%) 17 (17.8%) 2.11 Red-eared 25 11 (82.2%) 14 (17%) 2.11 slider turtle Platypus 26 26 (97.6%) 0 1.81 EVOLUTION Human 23 23 (97.8%) 0 3.21 Koala 8 8 (98.7%) 0 3.15 Florida lancelet 19 0 19 (94.2%) 0.51 684 200 0 Time (MYA) Fig. 1. Phylogenetic relationships of reptiles, birds, mammals, and amphioxus genome assemblies compared in this study. Cytological chromosome num- bers (n) are shown, along with the number of assembled macrochromosomes and microchromosomes (their percentage of the anchored genome) and genome size. Species names and full common names are given in SI Appendix, Table S1; in the text they are referred to by abbreviated common names. We then broadened our comparison to include turtle genomes. of microchromosomes was higher than for the macrochromo- Chromosome-level assemblies of two turtle species were available, somes in all these species (SI Appendix, Fig. S1). the green sea turtle and the red-eared slider turtle, with almost Sequence comparisons of snake genomes (Fig. 3B) showed identical karyotypes. As in birds, turtle microchromosomes have a that macro- and microchromosomes are almost wholly conserved higher GC content than macrochromosomes (SI Appendix, Fig. between python, rattlesnake, and cobra. There were relatively few S1A). Sequence comparison shows 1:1 correspondence between macrochromosome fusion/fission events between these species. two turtle genomes (Fig. 3A) except for two green sea turtle The 10 microchromosomes, too, were all conserved between the microchromosomes that align to parts of large macrochromo- three snake species, apart from some minor rearrangements. somes in the slider. However, notably, a cobra-specific microchromosome aligns to Turtle macrochromosomes align to emu macrochromosomes part of a small macrochromosome in other snake species. with some intrachromosomal rearrangements. However, turtles The water dragon, which is more closely related to snakes have seven fewer microchromosomes. There is an apparent than the other lizards studied (36), shares the snake macro- fusion of two microchromosomes conserved between sea turtle and microchromosome structure except for two microchromo- and emu that results in a larger microchromosome in the slider somes that align to termini of the largest snake chromosome. turtle. Four emu microchromosomes are present as two macro- Four water dragon microchromosomes had no alignments to chromosomes in both turtle species. Two emu microchromo- any of the snake or lizard genomes. Chromosomes of the water somes are fused to the termini of different macrochromosomes dragon and tegu were almost identical (Fig. 3B). in sea turtle and slider, and one is fused to a macrochromo- The common lizard and the wall lizard both had more and some in the sea turtle only. smaller macrochromosomes than tegu, water dragon, or snakes; Thus, genomes in turtles, as well as birds, are highly conserved. 10 of these align to the five largest tegu and water dragon mac- Hereafter we use the emu genome (35) as a representative of the rochromosomes, suggesting they are the products of fission. ancestral bird state and the green sea turtle to represent the turtle Both had few chromosomes in the microchromosome size ancestral state, for interclade comparisons. range, and their alignments to multiple microchromosomes Sequence comparisons of squamate genomes. Several genomes conserved between snakes, water dragon, and tegu suggest of snakes (python, rattlesnake, and cobra) and lizards (common fusion of two or three microchromosomes. There was also evi- lizard, wall lizard, the distantly related tegu, and the outgroup dence of a triply fused microchromosome fused with another water dragon) have been recently sequenced and have microchromosome in the wall lizard to form a chromosome out chromosome-scale assembled microchromosomes to enable of the microchromosome range. detailed comparisons and definition of ancestral states. The size There is evidence, too, of microchromosomes arising de differences between macro- and microchromosomes were obvious novo from chromosome fission. Four novel microchromosomes in the snakes, water dragon, and tegu but less pronounced in the in the wall lizard have homology to two macrochromosomes in common and wall lizards. However, GC content and gene density the common lizard, which have homology to parts of two larger Waters et al. PNAS j 3 of 11 Microchromosomes are building blocks of bird, reptile, and mammal chromosomes https://doi.org/10.1073/pnas.2112494118 A B C (micro- or macrochromosome-scale) entities or are fused to bird macrochromosomes. Our representative squamates also have fewer microchromo- somes than birds. Python has 10 and tegu 12 (plus 2 that align to a single macrochromosome in python and birds). The 10 snake microchromosomes (and their tegu counterparts) all align to bird microchromosomes. Other bird microchromo- somes align to autosomes in python and tegu, one to a terminal D E position and seven others make up two chromosome arms. Thus squamates, too, share a subset of bird micro- chromosomes. Importantly, although nearly all of the microchromosomes in turtle and squamate assemblies are homologous to bird micro- chromosomes they represent different subsets. Of 21 emu microchromosomes with sufficient homology to regions of both tegu and turtle genomes, only eight are microchromosomes in Fig. 2. Cytological characterization of microchromosomes in reptiles. (A) both (Fig. 3C). Three are incorporated into different macro- Mitotic chromosomes of the bearded dragon, showing extreme size differ- chromosomes in tegu and turtle. Seven emu microchromo- ence, different staining, and central location of microchromosomes. Image somes have homology to turtle microchromosomes but credit: Jason Dobry (La Trobe University, Melbourne, Australia). (B) Polar squamate macrochromosomes, and three emu microchromo- view of diakinesis in bearded dragon spermatocyte, showing extreme size and staining difference and central location of macro and micro bivalents. somes have homology to squamate microchromosomes but tur- (C) Fluorescence in situ hybridization using microchromosome-specific repet- tle macrochromosomes. itive sequence [AGAT]n showing central clustering of labeled dragon micro- The simplest explanation of the pattern of microchromo- chromosomes in interphase nuclei. (D) Two probes (green and red) light up some retention is that birds represent the ancestral amniote two pairs of microchromosomes in dragon cells. (E) The same probes retain condition (31 microchromosome pairs) and micromicro and their central location in interphase nuclei, although they colocate to the ter- micromacro fusions reduced the numbers of microchromo- minus of a macrochromosome in the long-neck turtle (Inset). somes independently in turtles and squamates. Inspection of the bird microchromosomes with homology to macrochromosomes in the tegu. Another microchromosome macrochromosomes in turtles or squamates (or both) reveal shared by both lizards has homology to another region of the different patterns of microchromosome fusion and fission. same tegu macrochromosome. Among the bird microchromosomes with homology to regions Thus, microchromosomes are very conserved among snakes, of turtle macrochromosomes, there are fusions of two or three water dragon, and tegu and must represent the ancestral state. microchromosomes that generate larger (macrochromosome- However, the wall lizard and common lizard lineages have sized) chromosomes. For instance, the two smallest turtle undergone fusions (largely micromicro) and a few fissions. We macrochromosomes are each homologous to two bird micro- use the python and tegu to represent the ancestral state for liz- chromosomes, implying fusions of ancestral microchromo- ards and snakes in across-clade comparisons. somes. Other microchromosomes, or micromicro fusions, have Sequence comparisons between birds and reptiles suggest that fused to terminal regions of macrochromosomes in turtles or bird microchromosomes best represent the ancestral amniote squamates. Several appear to be Robertsonian (centric) fusions, condition. We used pairwise genome alignments of the repre- in which an ancestral microchromosome or micromicro fusion sentative bird, turtle, snake, and lizard for across-clade compar- has become a macrochromosome arm (e.g., the turtle macro- isons (Fig. 3C and SI Appendix, Fig. S2). There was striking chromosome pair pictured in Fig. 2E). Almost all other fusions homology between the macrochromosomes and also between are terminal, with very few examples of internal integration into the microchromosomes of emu, green turtle, python, and tegu. a macrochromosome. The genome sizes of turtles, snakes, and lizards were larger While the predominant evolutionary pattern is of continued than those of birds (Fig. 1), suggesting that the sequence was loss of microchromosomes by fusion, a few novel microchromo- expanded by the insertion and retention of repetitive elements. somes have been created by fission from macrochromosomes, Comparisons of scaled genomes (Fig. 3C) suggest that this for example independently in hummingbird, cobra, and com- expansion affected macrochromosomes and microchromosomes mon lizard. equally. Notably, a few emu microchromosomes expanded into To ensure that our inference about microchromosome con- the range of macrochromosomes in turtles or squamates; for servation is not an artifact of selected pairwise comparisons, we instance, one microchromosome in tegu has 1:1 homology to a calculated the percentage of microchromosome aligned regions small chromosome in the common and wall lizards, but because in each query species, which aligned to microchromosomes of it is larger than 50 Mb (our threshold for microchromosome another target species (Fig. 3D). We observed that 72 to 100% size in squamates; see SI Appendix, Table S1) it is classified as a of microchromosome alignments for all species (except wall liz- macrochromosome in the wall (but not the common) lizard ard and golden eagle) are to bird microchromosomes. When (Fig. 3C). chicken (fusion of a microchromosome to chromosome 4) and Turtles have fewer microchromosomes than birds. Of the 22 hummingbird (fission of macrochromosomes that result in new assembled emu microchromosomes, only 15 are conserved as microchromosomes) are excluded, >98% are microto-micro microchromosomes in the green turtle. A microchromosome alignments between birds. It is intriguing to note that the heat present in the two turtle assemblies is missing from the emu map (Fig. 3D) is not symmetrical. This is because when birds and chicken assemblies. Four emu microchromosomes are are the target genome most microchromosomes of other spe- present as micromicro fusions. The other two emu microchro- cies align to bird microchromosomes. However, many of their mosomes align to arms of macrochromosomes in turtles, sug- microchromosomes have been fused into macrochromosomes gesting macromicro fusion. Thus, turtles retain a subset of in other clades, so when birds are the query genome the microchromosomes homologous to those of emu. Others are microto-micro alignment proportion is reduced. This demon- fused either with another microchromosome to form larger strates that the plot order of species in Fig. 3 has no influence 4 of 11 j PNAS Waters et al. https://doi.org/10.1073/pnas.2112494118 Microchromosomes are building blocks of bird, reptile, and mammal chromosomes INAUGURAL ARTICLE A Green sea Birds and turtles B Lizards and snakes Common turtle lizard Red−eared Common slider turtle wall lizard Argentine Emu black and white tegu Water Chicken dragon Burmese Jackdaw python Prairie Pigeon rattlesnake Anna's Indian cobra hummingbird C Across clade representatives D Microchromosome conservation Common lizard 29 30 63 2 47 47 43 48 50 44 45 20 18 21 24 38 37 Argentine black Wall lizard 63 73 79 1 63 77 72 80 82 75 75 47 45 54 70 68 82 and white tegu Tegu 52 55 76 23 52 79 77 80 75 77 78 61 55 58 62 22 31 Water dragon 61 66 95 31 65 98 95 98 89 97 97 92 99 32 32 Python 59 65 96 34 57 98 95 98 90 97 97 99 95 99 99 24 25 Rattlesnake 57 63 94 34 56 96 93 96 88 95 96 97 96 97 23 25 Query species Cobra 52 58 89 31 53 90 87 90 83 89 89 83 86 92 20 25 Burmese python Emu 49 52 97 40 54 99 98 99 92 98 53 51 57 61 72 15 27 75% Rhea 49 52 97 41 52 99 98 99 92 99 54 52 58 62 71 15 25 50% Chicken 51 54 96 38 50 98 95 98 98 98 53 50 56 60 70 17 28 25% Hummingbird 41 44 85 40 46 89 84 82 95 89 49 47 51 57 66 14 23 Jackdaw 47 50 98 39 50 98 99 89 99 99 56 53 57 65 72 15 25 Pigeon 47 48 97 45 56 98 99 85 99 99 56 53 58 64 72 15 26 Emu Cormorant 51 55 92 34 76 72 76 70 75 75 40 37 45 50 62 17 39 EVOLUTION Eagle 8 10 38 23 34 36 45 30 41 41 24 22 26 25 31 0 2 Alligator 40 44 36 61 87 85 85 80 86 86 47 43 49 51 62 14 34 Sea turtle 97 93 22 77 99 97 99 100 97 97 54 50 57 57 75 27 29 Slider turtle 96 93 18 75 98 94 98 99 97 97 49 48 55 57 72 26 31 a tle Al rtle C Ea r m le Pi nt H Ja eon in w C bird R n a R C u sn a W P ake dr on on om l gu liz d d o e he Em tle br on ar ar m da or g Green sea a at Se tur ck er th ag C Wa Te tu or at o m liz g g lig um ck at y hi er l turtle id Sl Target species E Emu to amphioxus F Amphioxus to emu Chr 1 Chr 5 Chr 14 Chr 16 G Rearranged species H Mammals Golden Koala eagle Human Emu Platypus American alligator Emu Macro to macro homology Macro to micro homology Micro to micro homology Fig. 3. Sequence homology plots within and between birds, reptiles, and mammals and comparison to the chordate amphioxus. Relationships between species and genome sizes are shown on the right. Genomes are resized, with chromosome sizes depicted as a proportion of genome length. Some chro- mosomes have been reordered in the plots to better show homologies within and between clades (see SI Appendix, Table S2). Macrochromosomes are shown in red and microchromosomes in blue; changes between these states are shown in green. Chained and netted alignments were filtered for a mini- mum length of 100 kb for vertebrate species and 5 kb for amphioxus. Sequence homology between macro- and microchromosomes within (A) birds and turtles and (B) squamates. (C) Sequence homology between macro and microchromosomes of a representative lizard (tegu), snake (python), bird (emu), and the green sea turtle. (D) Heat map showing the fraction (as a function of alignment chain length) of microchromosomes in the query species (y axis) that align to microchromosomes in the target species (x axis). (E and F) Comparisons of emu and amphioxus chromosomes: Single emu microchromosomes have homology to one (or two) amphioxus chromosomes (E), single amphioxus chromosomes detect strong homology to one (or more) emu microchro- mosomes (from left to right DNAzoo scaffolds 9, 10, 29 and 8), as well as to macrochromosome regions (F). (G) Sequence comparisons between emu and the rearranged alligator and eagle genomes. (H) Sequence comparisons between emu and mammals: eutherian (human), marsupial (koala), and mono- treme (platypus) Waters et al. PNAS j 5 of 11 Microchromosomes are building blocks of bird, reptile, and mammal chromosomes https://doi.org/10.1073/pnas.2112494118 Macro vs. Micro Ancestral states 52 60 52 60 Chr types 55 48 Macro % GC % GC Chr types Chromosomes 55 50 48 Micro Macro 44 % GC New Macro 45 Micro % GC New Micro 50 40 40 44 Chromosome type Chromosome type 45 40 40 0 25 50 75 0 20 40 60 Length Mb Length Mb Chromosomes Fig. 4. GC content versus size and distance-dependent interaction probabilities (Ps) of golden eagle macro- and microchromosomes in the present derived state, and after partitioning the genome into ancestral micro- and macrochromosomes and new micro and macrochromosomes. The interchromo- some interaction (with scaffolds sorted from largest to smallest, blue is low interaction and red high) heat map is for the present derived state. on our interpretation of microchromosome conservation and ancient chromosome have retained synteny (physical linkage), ancestral state. with a single copy equating to one or more microchromosomes, as previously observed (32). Origin of Microchromosomes: Homology with Amphioxus Unsurprisingly, given their homology to bird microchromo- Chromosomes. If we accept that the division of the bird genome somes, most turtle, snake, and lizard microchromosomes also into macro- and microchromosomes represents the ancestral equated to single or fused amphioxus chromosomes. The most amniote condition, the occurrence of microchromosomes in parsimonious explanation is that reptile/bird microchromo- some fish (32, 37) suggests this characteristic might be ancestral somes represent ancestral chordate chromosomes (or fused to all vertebrates. Indeed, ancestral vertebrates may have pos- chromosomes) that have retained synteny. sessed even more microchromosomes which fused before the We conclude that reptile and bird microchromosomes are divergence of the sauria. relics of an original animal genome with tiny, gene-rich chro- However, did microchromosomes arise by fission of larger mosomes, represented today by amphioxus. Since turtle and chromosomes in an ancient chordate ancestor, or did larger squamate microchromosomes are different subsets of bird vertebrate chromosomes arise by fusion of ancestral microchro- microchromosomes, this implies that these ancestral microchro- mosomes? Or did both processes occur (8)? To address this mosomes have been progressively lost by fusion at different question, we compared conserved bird and reptile microchro- rates independently in different lineages. Fusion has evidently mosomes with the small chromosomes of a distant chordate rel- been slower in birds than in other reptile lineages. ative, the amphioxus (Florida lancelet, Branchiostoma floridae), which last shared a common ancestor with vertebrates 684 Ma. Fusion of Microchromosomes into Macrochromosomes. To examine Amphioxus has a small (520 Mb) genome divided into 19 the process by which ancient microchromosomes became incor- tiny chromosomes that range in size from 17 Mb to 35 Mb. porated into macrochromosomes in vertebrates, we analyzed These chromosomes are very gene-dense (60 genes per Mb the genomes of exceptional reptile and bird species in which compared to 10 per Mb in mammals; SI Appendix, Fig. S1H). fusion has removed many or most microchromosomes. We Comparison of the amphioxus sequence with those of garfish aligned the genome of emu (representing ancestral birds) with and chicken revealed two genome doublings: an autotetra- alligator (representing crocodilians that diverged from birds ploidization in the Cambrian ∼500 Ma and allotetraploidy by 240 Ma) and eagle (which shared a common ancestor with fusion of genomes that had diverged in a fish ancestor ∼460 other birds with a conserved karyotype only about 80 Ma). Ma, followed by extensive loss of duplicate genes (32). Con- We found that emu macro- and microchromosome sequence siderable sequence blocks shared synteny with the chicken had obvious homologs in the alligator and eagle genomes, genome, some of which represented 1:1 relationship with although they were considerably rearranged (Fig. 3G). Strong chicken microchromosomes. The Australian lungfish genome, homology between emu and alligator macrochromosome arms though much expanded with repetitive sequence, also pos- shows that the five very large alligator chromosomes are all sesses many microchromosomes with homology to amphioxus fusions and rearrangements of ancestral macrochromosomes, chromosomes (37). as was demonstrated by chromosome painting between the We aligned the bird (emu) and amphioxus genomes (Fig. 3E chicken and the Nile crocodile (21, 38), which shares many and SI Appendix, Fig. S3). Of the 21 emu microchromosomes chromosomes with alligator (28). An alligator macrochromo- with sufficient shared sequence to detect homology, nine bird some and the largest microchromosome were each generated microchromosomes each represented a single amphioxus chro- by fusions of an ancestral macro- and microchromosome. How- mosome, suggesting that ancestral chromosomes have been ever, the 10 smallest alligator chromosomes each comprise retained intact. Another six emu microchromosomes each con- either single or fused ancestral microchromosomes; one small tained sequences from two amphioxus chromosomes, implying alligator chromosome represents a single microchromosome, fusions in ancestral vertebrates. five alligator chromosomes represent fusions of two, and four We then aligned single amphioxus chromosomes to the emu alligator chromosomes fusions of three ancestral microchromo- genome, demonstrating that most have strong homology to one somes. These changes must have occurred in the crocodilian or two microchromosomes, as well as to two (sometimes three) lineage in the ∼240 My since they shared a common ancestor regions of macrochromosomes (Fig. 3F and SI Appendix, Fig. with birds. It is striking that all but two rearrangements are S4). These probably represent the four copies of the chordate either a macromacro or a micromicro fusion. genome that resulted from two rounds of doubling. It is notable The rearranged eagle genome, in contrast, contains many that one or two of these homologies (usually including at least fusions between macro- and microchromosomes and many fis- one microchromosome) are strong and focused, whereas the sions of ancestral macrochromosomes into smaller chromo- others are more dispersed, suggesting rearrangement, deletion, somes (Fig. 3G). Only three macrochromosomes have been and repeat expansion. This suggests that all copies of each retained intact; two have undergone centric rearrangements of 6 of 11 j PNAS Waters et al. https://doi.org/10.1073/pnas.2112494118 Microchromosomes are building blocks of bird, reptile, and mammal chromosomes INAUGURAL ARTICLE chromosome arms and two have each fused with an ancestral alligator, water dragon, python and eagle resulted in changes of microchromosome. The other four ancestral macrochromo- properties of ancestral and de novo microchromosomes. somes have undergone multiple fissions, the products of which The alligator has chromosomes out of the microchromosome have fused to other macrochromosome arms or ancestral size range that were formed by micromicro fusions (Fig. 3G). microchromosomes. The largest emu chromosome has under- We found that these fused microchromosomes retain their gone fission into seven regions, six of which are present as small higher GC ratio and still interact strongly with each other and chromosomes (five in the micro range). In addition, two small the smaller microchromosomes (SI Appendix, Fig. S1). eagle macrochromosomes derive from micromicro fusions. In the water dragon, several ancestral microchromosomes Thus, eagle microchromosomes include only four ancestral are fused to form four new regions of macrochromosomes. microchromosomes and one micromicro fusion. Three ances- These are also present in the python genome (SI Appendix, Fig. tral microchromosomes have fused to form macrochromosome S1I) so must have occurred in a common squamate ancestor arms or terminal regions. What is striking is that six eagle about 180 Ma. When we plotted GC content and contact prob- microchromosomes were derived from regions (mostly termi- abilities according to the ancestral state (SI Appendix, Fig. S1F) nal) of ancestral macrochromosomes so represent de novo we found that these ancestral microchromosomes fused to mac- microchromosomes. rochromosomes had a lower GC content but retained their Thus, in both alligator and eagle, macrochromosome arms high contact probability. However, two ancestral microchromo- and microchromosomes have been fused with each other, and somes that were more recently fused to either end of the largest with the termini of macrochromosomes. However, the patterns scaffold only in python still retain their elevated GC content of fusions and fissions are quite different in the eagle and alli- (SI Appendix, Fig. S1G). This suggests that the GC-richness of gator, attesting to independent rearrangement facilitated by dif- microchromosomes is retained on rearrangement but erodes ferent mechanisms. over time. The much-rearranged eagle genome, for which there is both Microchromosome Interactions. Our cytogenetic studies (dis- chromosome-level assembly and Hi-C data, allowed us to assess cussed above and see Fig. 2) confirm and extend previous features of both microchromosomes which fused to form mac- observations (4) that microchromosomes are spatially segre- rochromosomes, and macrochromosomes that were split into microchromosomes, in the ∼80 My since the eagle last shared a EVOLUTION gated within cells, occupying a central location at interphase, and during mitosis and meiosis, in turtles and squamates as common ancestor with condors, which retain an ancestral bird well as birds. We also showed that microchromosomes may karyotype (44). retain this central position even after fusion to macrochromo- Of the 30 ancestral microchromosomes present in other somes, implying that size alone does not determine microchro- birds, only four remain in the eagle; two others are fused into a mosome location. larger microchromosome and 12 are fused with macrochromo- Data from high-throughput chromosome conformation cap- somes. Nine new microchromosomes represent regions of ture (Hi-C) now provide a molecular description of this spatial ancestral macrochromosomes which underwent fission (Fig. segregation. Hi-C data incorporated in the new emu assembly 3G). (35) and the rattlesnake assembly (39) reveal that microchro- Unlike other bird species (SI Appendix, Fig. S1), the eagle mosomes interact with each other more than with macrochro- micro- and macrochromosomes showed no abrupt differences in GC content or contact probabilities with size (Fig. 4), reflect- mosomes in these species, as has been observed also in other ing their recent reshuffling. The microchromosomes still birds, snakes, and turtles (40). We confirmed and extended retained elevated interchromosomal interactions (Fig. 4), these observations to other birds and reptiles (green sea turtle, although this was not as well correlated with chromosome size alligator, python, eagle, tegu, emu, greater rhea, and water as for other species. dragon), using DNA Zoo Hi-C data (41) (SI Appendix, Fig. S1 To explore the origin of these differences, we examined the A–G). These species show various degrees of rearrangement of characteristics of eagle sequence according to ancestral state as ancestral macro- and microchromosomes, permitting us to deduced from comparison with the emu genome (Fig. 3G). determine whether microchromosome fusion or macrochromo- Regions of synteny with the emu genome were classified as some fission can alter the GC ratio and interaction macro (macro in both species), micro (micro in both species), characteristics. new macro (micro in emu but macro in eagle), or new micro On average, pairs of loci borne on the same chromosome (macro in emu but micro in eagle). We found that GC content present higher interaction probabilities (intrachromosomal reflected the ancestral state (Fig. 4). The microchromosomes interactions) than between loci on different chromosomes incorporated into macrochromosomes (new macros) in eagle (interchromosomal interactions). At the genome-wide level, maintained high GC content, and macrochromosomes broken this cis/trans interaction pattern reflects chromosomal territori- down to microchromosome size (new micros) retained their ality (42). As expected, the interaction between neighboring low GC content. loci on the same chromosome decreases as genomic distance In contrast, the distance-dependent contact probabilities increases. This reduction in interactions can be represented as [P(s)] of these four classes of regions did not consistently reflect genomic distance-dependent contact probabilities [P(s)], repre- their ancestral state. As expected, ancestral macrochromosomes senting the level of chromosome compaction (43). displayed lesser and microchromosomes greater distance- Genome-wide heat-map plots of representative species (SI dependent interaction probabilities (Fig. 4). Microchromo- Appendix, Fig. S1 A–G) all show that, as well as a high GC somes incorporated into macrochromosomes also maintained ratio, there is a higher interchromosomal interaction and higher relatively high contact probabilities at genomic distances below distant dependent contact probabilities [P(s)] between micro- 1 Mbp, albeit slightly lower than the ancestral microchromo- chromosomes than between macrochromosomes or between somes. Surprisingly, however, the new microchromosomes macro- and microchromosomes. More intense interaction is derived recently from macrochromosomes had almost the same therefore an intrinsic feature of all reptile/bird microchromo- elevated distance-dependent interaction probabilities as the somes, reflected by their arrangement within the nucleus during ancestral microchromosomes. This cannot be a consequence of interphase (Fig. 2). altered base ratio, since percent GC was not increased, and Given that this interaction pattern is an ancestral feature of may reflect different levels of chromatin compaction or microchromosomes, we asked whether rearrangements in compartmentalization. Waters et al. PNAS j 7 of 11 Microchromosomes are building blocks of bird, reptile, and mammal chromosomes https://doi.org/10.1073/pnas.2112494118 We conclude that the high interaction probabilities of ances- demonstrable homology reflect paralogous sequences gener- tral microchromosomes are maintained after their incorpora- ated by the two genome doublings early in vertebrate evolution tion into a macrochromosome but erode over time. However, (32, 46). For most of these paralogous regions one or two of novel microchromosomes rapidly adopt high interaction with the strongest are located on bird microchromosomes, probably other microchromosomes. representing paralogous sequences that retain the gene richness and low repetitive sequence content of the original chromo- The Fate of Microchromosomes in Mammals. To explore the fate somes, whereas other genome copies have suffered deletion, of ancient microchromosomes in therian mammals (eutherian transposable element invasion, and rearrangement. It is inter- and marsupial mammals), we compared the genomes of emu esting to speculate that microchromosomes may be protected with those of koala (a marsupial) and human (eutherian) (Fig. from rearrangement and insertion of repetitive elements by 3H). We observed regions with some homology to bird micro- their longer subtelomere regions, their spatial isolation, and chromosomes, though these were weak and dispersed. There high interaction. was one region of the human genome (chromosome 17) and Microchromosomes as well as macrochromosomes have two regions of the koala genome that had homology to two or been proportionately lengthened by insertion of transposable more microchromosomes but no evidence that microchromo- elements, as observed over many vertebrates (47), occasionally somes are retained intact in either species. moving out of the microchromosome range. However, most Monotremes (egg-laying mammals such as the platypus) microchromosomes have been progressively lost by fusion, and have a karyotype somewhat resembling those of reptiles, with 6 only very occasionally gained by fission of macrochromosomes, pairs of large chromosomes and 20 pairs of much smaller chro- in all bird and reptile lineages (Fig. 5). mosomes (45). Although the small platypus chromosomes are Loss of microchromosomes in birds and turtles by fusion not in the size range of reptile microchromosomes, we asked with macrochromosomes has been documented in several cyto- whether these were the vestiges of ancestral microchromo- logical studies (e.g., ref. 48). Patterns of microchromosome loss somes. We found that most platypus chromosomes, large and (Fig. 5) were examined by comparing the genomes of conserved small, had contributions from multiple chicken macro- and representatives of birds, turtles, and squamates. Almost all loss microchromosomes, and there was no obvious enrichment of can be attributed to fusion of two, or sometimes three, micro- microchromosomes in the small platypus chromosomes (Fig. chromosomes to form larger chromosomes, as well as fusion of 3H). microchromosomes to macrochromosomes. These fusions are However, homology with ancestral microchromosomes was almost always to terminal locations or via centric fusion to con- not distributed randomly within the platypus genome. Four stitute a chromosome arm, as was implied by early observations platypus chromosomes appear to be entirely composed of of interstitial and centromeric telomere sequence (49). regions with homology to one to four emu microchromosomes, Patterns of microchromosome loss were also examined by suggesting that they evolved from multiple microchromosome following the chromosome changes in exceptional species (alli- fusions. Other regions with strong homology lay at the ends or gator and golden eagle) with more rearranged chromosomes as comprised arms of platypus chromosomes. shown by cytogenetic and sequencing methods (50). Our exten- This suggests that in the reptile-like ancestor of mammals sive analysis showed that these genomes differ from the stan- micromicro fusions were common (as they were also in the dard simply by more fusions of the same types that distinguish crocodilian ancestor). These blocks of fused microchromo- turtle and squamate genomes, suggesting that the same pro- somes have been preserved in platypus macrochromosomes but cesses occurred, but at much higher rates. were disrupted in therian mammals, so that little vestige The onset of instability in a lineage could be quite sudden. remains. The high rate of genome reshuffling in mammals con- For instance, extensive genome remodeling in the eagle lineage trasts with the stability of genomes in other amniote lineages. must have occurred in the last 80 My since eagles shared a common ancestor with condors, which retain a near-ancestral Discussion genome arrangement (44). Cormorants diverged from the same Our analysis of sequence data confirms that reptile and bird conserved lineage a short time later. Our analysis of the cormo- microchromosomes, as well as macrochromosomes, are highly rant assembly (51) (Fig. 3D) revealed a rearranged genome; conserved within bird, turtle, and squamate lineages, and even however, few of the rearrangements are shared with eagles. between these lineages. Microchromosomes are most numerous The high rate of independent changes in cormorants or eagles and almost invariant among birds, many species having 31 well- suggests that instability was introduced to the genome of a conserved microchromosomes. Different microchromosome subsets have been retained in turtles and squamates, implying that the 31 microchromosomes were present in the common Stepwise evolutionary change Karyotype Ancestral ancestor of birds and reptiles about 300 Ma, expanding on the chromosomes 20 considered by Burt (8) to be ancestral. Previous work using Micro-micro older and more limited sequence datasets implied that chicken fusion microchromosomes represented ancestral birds but suggested Expansion that a dinosaur ancestor of birds and crocodiles and an archelo- saur ancestor of birds and turtles >250 Ma had fewer and Micro-macro larger chromosomes that underwent fission (26). Our demon- fusion stration that squamates, too, share many microchromosomes Macro rearrangement with emu implies an older origin. Strong homology with the small chromosomes of amphioxus implies that the conserved set of bird microchromosomes rep- Fig. 5. Evolutionary loss and gain of microchromosomes has reshaped the resent retention of ancient chromosomes of a common ancestor amniote kartyotype. Fusion between microchromosomes, expansion by that lived 684 Ma. Comparisons of single amphioxus chromo- the insertion of transposable elements, fusion with macrochromosomes, somes with the bird genome reveal homology to four regions, and macrochromosome rearrangement have led to a reduction of micro- one or two with bird microchromosomes and two or three with chromosomes. Rarely macrochromosome fissions result in new bird macrochromosomes. Presumably these four regions of microchromosomes. 8 of 11 j PNAS Waters et al. https://doi.org/10.1073/pnas.2112494118 Microchromosomes are building blocks of bird, reptile, and mammal chromosomes INAUGURAL ARTICLE common ancestor but was expressed independently in the two identical-sized (3 Gb) genome is packaged as anything between lineages. One explanation of the sudden onset of instability 3 pairs of enormous chromosomes in the Indian muntjac to 51 after millions of years of extreme conservation is that the pairs of small chromosomes in the red viscacha rat (59). Not genome of an eagle ancestor (∼80 Ma) was invaded by a trans- only is the highly rearranged eutherian genome subdivided into posable element which facilitated interchromosome rearrange- a few large or many small chromosomes but sequences have ments (52). been scrambled in many lineages. Large synteny blocks are Intriguingly, the patterns of microchromosome fusions are shared by some species (e.g., humans, cats, and even sloths), quite different in the two exceptional species we examined. enabling reconstruction of ancestral eutherian karyotypes (60, Rearrangements in the alligator are almost all confined to 61), but even these blocks of shared synteny are wildly different micromicro and macromacro fusions, whereas in the eagle they in other species, especially the rodents. The marsupial karyo- are largely micromacro fusions. Assortative fusions (micromi- type, in contrast, is highly conserved between all 260 species cro and macromacro) may be favored by the spatial compart- and derives from a n = 7 basal karyotype largely by Robertso- mentalization (53, 54). However, it is likely that rearrangements nian translocations (62). are driven by the location and type of repetitive elements We found that in eutherians and marsupials microchromo- inserted, as seems to be the case for parrots (27). Transposable somes have completely disappeared, visible only as broken up elements are ubiquitous in reptiles and birds (55) but are not patches of homology peppering the genome. randomly distributed in the genome (56). For instance, there Even the early-branching monotreme mammals, which have a are several repetitive sequences that are shared only by micro- rather reptile-like karyotype with six large and many small chro- chromosomes (57) (e.g., Fig. 2C), which may favor micromicro mosomes, retain none of the ancestral microchromosomes. Most rearrangements, such as those that predominate in alligator. of the small platypus chromosomes have homology to several Other transposable elements are observed to cluster at centro- regions of macro- and microchromosomes. However, two large meres of all chromosomes (58) and may be involved in Robert- and two small platypus chromosomes seem to be completely sonian fusions between macro- and microchromosomes. made up of fusions of several ancestral microchromosomes, sug- Loss of microchromosomes has, rarely, been offset in some gesting that the process of amalgamation may have started from lineages by the creation of novel microchromosomes from mac- many micromicro fusions in a mammalian ancestor. rochromosome fragments. Several appear in the eagle genome, We propose that microchromosome fusion occurred in the EVOLUTION and there are one or two in other birds (e.g., hummungbird) ancestor of all mammals after divergence from the reptile/bird lin- and squamates (e.g., common lizard). eage 310 Ma and before the divergence of monotremes from ther- Reptile and bird microchromosomes are distinguished by ians 188 Ma. These microchromosome blocks must have already their high gene density and high GC content, and particularly undergone major sequence reshuffling before the marsupial– by their compartmentalization in the center of interphase and eutherian divergence 168 Ma because microchromosome-homolo- dividing cells. We have documented differential staining gous sequences are split up and distributed all over the genome in (denoting differences in sequence and conformation) and spa- both lineages. tial segregation in a variety of squamates and turtles as well as The sequence shuffling and size variation among eutherian birds consistent with many earlier cytological observations (4). chromosomes would require some event that loosened the con- The molecular underpinnings of these properties were revealed straints on rearrangement in a mammalian ancestor, probably by our analysis of chromosome conformation capture data, invasion and amplification of particular retrotransposons, which which shows that microchromosomes in bird, turtle, and squa- provide sites for crossing over between nonhomologs (reviewed mate lineages all show high interaction within this compart- in ref. 63). ment. These intense interchromosome relationships have been Thus, mammal genomes are spectacularly atypical among thought to denote different chromatin organization of micro- amniotes, displaying variation that has been sometimes credited chromosomes that reflects some functional coherence. (e.g., ref. 64) with their successful colonization of varied We analyzed rearranged genomes to examine properties of (terrestial and underground, arboreal, aqueous, even aerial) regions whose status has changed between micro- and macro- environments. We need to understand what effects these rear- chromosome. We found that high interaction, as well as high rangements between ancestral macro- and microchromosomes GC content, is retained by ancestral microchromosomes that had on genome function. The high gene density, atypical base fused with macrochromosomes in eagle. This is consistent with ratio, spatial segregation, and high interaction between micro- our cytological observation that two ancestral microchromo- chromosomes suggests a functional coherence of this part of somes fused to a macrochromosome in a turtle retain their the genome, which survives in the subsets of microchromo- central position (Fig. 2E), as was also observed for the micro- somes retained in birds, turtles, and squamates. Gene-dense chromosome uniquely fused to chromosome 4 in chickens (10) and active chromosome regions are also located centrally in and fused microchromosomes in the rearranged genomes of mammalian cells (65), but these do not equate with ancestral falcons and parrots (50). However, high GC content does microchromosomes. It is difficult to unscramble cause-and- appear to erode with time, as shown by a lower GC content of effect relationships between chromosome size, location, gene fused microchromosomes in snakes. density, GC content, and distance-dependent interaction We discovered, unexpectedly, that novel microchromosomes probabilities. derived by fission of ancestral macrochromosomes rapidly Our overall conclusion is that bird microchromosomes repre- acquired an increased distance-dependent interaction probabil- sent remnants of the original building blocks of vertebrate ity, although their GC content stayed low. This suggests that genome. They retain high gene density and low content of repeti- intense interaction may be the result not of sequence composi- tive sequence and share conserved features across all reptile and tion but of chromosome compaction or location in the cell, bird clades. Their progressive fusion with each other, and with which may be influenced by epigenetic factors or chromosome macrochromosomes, occurred conservatively and gradually in size. most reptile lineages, but more rapidly in a few clades. Multiple The extreme conservatism of the bird/reptile genomes, and microchromosome fusions occurred early in mammal evolution even the modest rearrangements that characterize the crocodi- and were followed by lineage-specific rearrangement and a huge lians, expose the flagrant chromosome shuffling in mammals as variety of fusions and fissions that disrupted the relationship a glaring exception among amniotes. Mammals have extraordi- between microchromosomes. Among amniotes, even vertebrates, narily variable karyotypes. In eutherian mammals, a near- mammal genomes are the true exceptions. Waters et al. PNAS j 9 of 11 Microchromosomes are building blocks of bird, reptile, and mammal chromosomes https://doi.org/10.1073/pnas.2112494118 Materials and Methods chromosomes based on the karyotype information (SI Appendix, Table S1). Each square of the matrix represents the mean of normalized interaction val- Cytology. Mitotic and meiotic chromosomes preparations and chromosome ues between the extracted scaffolds at 50-kb bin resolution. paint preparation and painting were performed following protocol described in ref. 63. Repeat mapping was performed following protocol described in ref. 64. Briefly, 200 ng 50 -Cy3–labeled (AAGG)8 probe was mixed with 15 μL hybrid- Distance-Dependent Contact Probability P(s). The Hi-C matrices were exported ization buffer [50% formamide, 10% dextran sulfate, 2× saline sodium citrate with HiCExplorer (v3.6) to h5 format. The scaffolds for each species were classi- (SSC), and 40 mmol/L sodium phosphate, pH 7.0, and 1× Denhardt’s solution] fied as either macro- or microchromosomes (SI Appendix, Table S2). Using the added to slides containing fixed chromosome preparation and denatured at hicAdjustMatrix function, independent matrices were created for macrochro- 68 °C for 5 min. After denaturation, slides were incubated overnight in a moist mosomes and microchromosomes. Distance-dependent contact probabilities hybridization chamber at 37 °C. Slides were then washed once at 60 °C in 0.4× [P(s)] were calculated using hicPlotDistVsCounts (from the HiCExplorer pack- SSC and 0.3% Igepal for 2 min, followed by another wash at room tempera- age), which were plotted with a maximum distance of 1 × 108 bp. ture in 2× SSC and 0.1% Igepal for 1 min and air-dried. Slides were then coun- terstained with DAPI with Vectashield. Image analysis was performed using a Golden Eagle Analysis. Regions of the golden eagle genome were classified Zeiss Axioplan epifluorescence microscope equipped with a charge-coupled according to homology with the chicken genome. These were macro (macro device camera (RT-Spot; Zeiss). in both species), micro (micro in both species), new macro (micro in chicken but macro in eagle), or new micro (macro in chicken but micro in eagle). These Whole-Genome Alignments. One-way all-versus-all LastZ (Release 1.02) (31) newly created scaffolds were then used to calculate GC content and distance- alignments were performed for 23 reptile (including birds) species and platy- dependent contact probabilities as described above. pus. Human, Tasmanian devil, and amphioxus genomes were aligned against selected genomes. Alignments were chained and netted using the UCSC Tool- Data Availability. Genome data used in the publication are all publicly avail- kit (http://hgdownload.soe.ucsc.edu/admin/exe/linux.x86_64). Workflow was able (see SI Appendix, Table S1). All coordinates of pairwise alignment chains automated using scripts available at https://github.com/kango2/tiny. are available for download at https://cloudstor.aarnet.edu.au/plus/s/ Briefly, the LastZ alignment parameters were K = 2400 L = 3000 Y = 9400 H twyXXR0kfpRZJiu. Due to large file sizes (∼3 TB total), raw alignments can be = 2000–ambiguous = iupac. Chaining was performed with axtChain using provided upon request. All code and metadata used for automating whole- -minScore = 3000 -linearGap = medium as parameters. Chains were sorted genome alignments and plotting of figures are available at https://github. with chainSort, prenetting was performed with chainPreNet, and netting was com/kango2/tiny. performed with chainNet, each using default parameters. Syntenic blocks were calculated with netSyntenic. ACKNOWLEDGMENTS. P.K. gratefully acknowledges the resources provided Homology and statistics were plotted in R using the tidyverse package by University of Western Australia, the Aiden laboratory at Baylor College of (v1.3.0) with custom scripts available at the GitHub repository. Microchromo- Medicine, DNA Zoo consortium partners, and additional computational some and macrochromosome labels, when not available, were assigned to resources and support from the Pawsey Supercomputing Centre with funding assembled sequences based on published karyotype data summarized in SI from the Australian Government and the Government of Western Australia. Appendix, Tables S1 and S2. Unpublished genome assemblies and sequencing data for tegu, eastern water dragon, burmese python, greater rhea, band-tailed pigeon, golden eagle, GC Content and Gene Density. GC content of each scaffold was calculated American alligator, green sea turtle, koala, and emu are used with permission using BBMap (v38. 9) (65) for DNAzoo data or obtained from the relevant from the DNA Zoo Consortium (https://www.dnazoo.org/). P.D.W., A.G., J.A. NCBI genome information website. Gene density per megabase was calcu- M.G., F.G., T.E., and C.F. are supported by grants from the Australian Research lated by dividing the number of annotated genes on a chromosome by Council (DP170101147, DP180100931, DP210103512, and FT200100192). H.R.P. is supported by an Australian National University research fellowship. This its length. work was supported by computational resources provided by the Australian Government through National Computational Infrastructure under the Interscaffold Interactions. The Hi-C data were obtained from DNAzoo (SI National Computational Merit Allocation Scheme and Australian National Appendix, Table S1). Hi-C matrices were exported to 50-kb GInteractions bins University Merit Allocation Scheme. A.R.-H. is supported by the Spanish Minis- with HiCExplorer (v3.6) (66). This format consisted of seven columns: origin try of Economy and Competitiveness (CGL2017-83802-P) and by the Spanish scaffold, origin scaffold start, origin scaffold end, target scaffold, target scaf- Ministry of Science and Innovation (PID2020-112557GB-I00). L.A.-G. is sup- fold start, target scaffold end, and number of interactions between bins. The ported by an FPI predoctoral fellowship from the Spanish Ministry of Economy largest scaffolds were extracted, equal to the expected number of and Competitiveness (PRE-2018-083257). 1. J. M. Van Brink, L’expression morphologique de la digame tie chez les Sauropside s et 14. R. Kretschmer, M. A. Ferguson-Smith, E. H. C. de Oliveira, Karyotype evolution in les Monotre mes. Chromosoma 10, 1–72 (1959). birds: From conventional staining to chromosome painting. Genes (Basel) 9, 181 2. L. Christidis, Animal Cytogenetics 4: Chordata 3 B, Aves (Gebr€ uder Borntraeger, Ber- (2018). lin, 1990). 15. L. G. Kiazim et al., Comparative mapping of the macrochromosomes of eight avian 3. N. Takagi, M. Sasaki, A phylogenetic study of bird karyotypes. Chromosoma 46, species provides further insight into their phylogenetic relationships and avian kar- 91–120 (1974). yotype evolution. Cells 10, 362 (2021). 4. F. A. Habermann et al., Arrangements of macro- and microchromosomes in chicken 16. R. E. O’Connor et al., Patterns of microchromosome organization remain highly con- cells. Chromosome Res. 9, 569–584 (2001). served throughout avian evolution. Chromosoma 128, 21–29 (2019). 5. E. H. Newcomer, Accessory chromosomes in the domestic fowl. Genetics 40, 587–588 17. R. Kretschmer et al., Interspecies chromosome mapping in caprimulgiformes, pici- (1955). formes, suliformes, and trogoniformes (aves): Cytogenomic insight into microchro- 6. H. A. McQueen, G. Siriaco, A. P. Bird, Chicken microchromosomes are hyperacety- mosome organization and karyotype evolution in birds. Cells 10, 826 (2021). lated, early replicating, and gene rich. Genome Res. 8, 621–630 (1998). 18. J. Damas, J. Kim, M. Farre, D. K. Griffin, D. M. Larkin, Reconstruction of avian ances- 7. J. Smith et al., Differences in gene density on chicken macrochromosomes and micro- tral karyotypes reveals differences in the evolutionary history of macro- and micro- chromosomes. Anim. Genet. 31, 96–103 (2000). chromosomes. Genome Biol. 19, 155 (2018). 8. D. W. Burt, Origin and evolution of avian microchromosomes. Cytogenet. Genome 19. G. Zhang et al., Comparative genomics reveals insights into avian genome evolution Res. 96, 97–112 (2002). and adaptation. Science 346, 1311–1320 (2014). 9. I. Nanda, M. Schmid, Localization of the telomeric (TTAGGG)n sequence in chicken 20. E. Olmo, Trends in the evolution of reptilian chromosomes. Integr. Comp. Biol. 48, (Gallus domesticus) chromosomes. Cytogenet. Cell Genet. 65, 190–193 (1994). 486–493 (2008). 10. International Chicken Genome Sequencing Consortium, Sequence and comparative 21. J. E. Deakin, T. Ezaz, Understanding the evolution of reptile chromosomes through analysis of the chicken genome provide unique perspectives on vertebrate evolution. applications of combined cytogenetics and genomics approaches. Cytogenet. Nature 432, 695–716 (2004). Genome Res. 157, 7–20 (2019). 11. T. M. Degrandi et al., Introducing the bird chromosome database: An overview of 22. S. M. I. Alam, S. D. Sarre, D. Gleeson, A. Georges, T. Ezaz, Did lizards follow unique cytogenetic studies in birds. Cytogenet. Genome Res. 160, 199–205 (2020). pathways in sex chromosome evolution? Genes (Basel) 9, 239 (2018). 12. S. Shetty, D. K. Griffin, J. A. Graves, Comparative painting reveals strong chromosome 23. Y. Matsuda et al., Highly conserved linkage homology between birds and turtles: homology over 80 million years of bird evolution. Chromosome Res. 7, 289–295 Bird and turtle chromosomes are precise counterparts of each other. Chromosome (1999). Res. 13, 601–615 (2005). 13. D. K. Griffin, L. B. Robertson, H. G. Tempest, B. M. Skinner, The evolution of the avian 24. W. Becak, M. L. Becak, H. R. Nazareth, S. Ohno, Close karyological kinship between genome as revealed by comparative molecular cytogenetics. Cytogenet. Genome the reptilian suborder serpentes and the class aves. Chromosoma 15, 606–617 Res. 117, 64–77 (2007). (1964). 10 of 11 j PNAS Waters et al. https://doi.org/10.1073/pnas.2112494118 Microchromosomes are building blocks of bird, reptile, and mammal chromosomes INAUGURAL ARTICLE 25. Y. Uno et al., Inference of the protokaryotypes of amniotes and tetrapods and the 44. T. Raudsepp et al., Cytogenetic analysis of California condor (Gymnogyps california- evolutionary processes of microchromosomes from comparative gene mapping. nus) chromosomes: Comparison with chicken (Gallus gallus) macrochromosomes. PLoS One 7, e53027 (2012). Cytogenet. Genome Res. 98, 54–60 (2002). 26. R. E. O’Connor et al., Reconstruction of the diapsid ancestral genome permits chromo- 45. D. McMillan et al., Characterizing the chromosomes of the platypus (Ornithorhyn- some evolution tracing in avian and non-avian dinosaurs. Nat. Commun. 9, 1883 chus anatinus). Chromosome Res. 15, 961–974 (2007). (2018). 46. T. D. Lamb, Analysis of paralogons, origin of the vertebrate karyotype, and ancient 27. Z. Huang et al., Recurrent chromosome reshuffling and the evolution of neo-sex chromosomes retained in extant species. Genome Biol. Evol. 13, evab044 (2021). chromosomes in parrots. bioRxiv [Preprint] (2021). http://doi.org/10.1101/2021.03.08. 47. A. Canapa, M. Barucca, M. A. Biscotti, M. Forconi, E. Olmo, Transposons, genome 434498 (Accessed 22 June 2021). size, and evolutionary insights in animals. Cytogenet. Genome Res. 147, 217–239 28. V. C. S. Oliveira et al., Revisiting the karyotypes of alligators and caimans (Crocodylia, (2015). Alligatoridae) after a half-century delay: Bridging the gap in the chromosomal evolu- 48. P. E. Lithgow et al., Novel tools for characterising inter and intra chromosomal rear- tion of reptiles. Cells 10, 1397 (2021). rangements in avian microchromosomes. Chromosome Res. 22, 85–97 (2014). 29. S. R. Voss et al., Origin of amphibian and avian chromosomes by fission, fusion, and 49. I. Nanda et al., Distribution of telomeric (TTAGGG)(n) sequences in avian chromo- retention of ancestral chromosomes. Genome Res. 21, 1306–1312 (2011). somes. Chromosoma 111, 215–227 (2002). 30. A. Kapusta, A. Suh, Evolution of bird genomes-a transposon’s-eye view. Ann. N. Y. 50. R. E. O’Connor et al., Chromosome-level assembly reveals extensive rearrangement Acad. Sci. 1389, 164–185 (2017). in saker falcon and budgerigar, but not ostrich, genomes. Genome Biol. 19, 171 31. Y. Nakatani, H. Takeda, Y. Kohara, S. Morishita, Reconstruction of the vertebrate (2018). ancestral genome reveals dynamic genome reorganization in early vertebrates. 51. A. Burga et al., A genetic signature of the evolution of loss of flight in the Galapagos Genome Res. 17, 1254–1265 (2007). cormorant. Science 356, eaal3345 (2017). 32. O. Simakov et al., Deeply conserved synteny resolves early events in vertebrate evolu- 52. J. Zhang, C. Yu, L. Krishnaswamy, T. Peterson, Transposable elements as catalysts for tion. Nat. Ecol. Evol. 4, 820–830 (2020). chromosome rearrangements. Methods Mol. Biol. 701, 315–326 (2011). 33. R. S. Harris, “Improved pairwise alignment of genomic DNA,” PhD thesis, The Penn- 53. T. Marques-Bonet, A. Navarro, Chromosomal rearrangements are associated with sylvania State University, State College, PA (2007). higher rates of molecular evolution in mammals. Gene 353, 147–154 (2005). 34. M. Kohn, H. Kehrer-Sawatzki, W. Vogel, J. A. M. Graves, H. Hameister, Wide genome 54. Y. Zhang et al., Spatial organization of the mouse genome and its role in recurrent comparisons reveal the origins of the human X chromosome. Trends Genet. 20, chromosomal translocations. Cell 148, 908–921 (2012). 598–603 (2004). 55. D. Kordis, Transposable elements in reptilian and avian (sauropsida) genomes. Cyto- 35. J. Liu et al., A new emu genome illuminates the evolution of genome configuration genet. Genome Res. 127, 94–111 (2009). and nuclear architecture of avian chromosomes. Genome Res. 31, 497–511 (2021). 56. T. Sultana, A. Zamborlini, G. Cristofari, P. Lesage, Integration site selection by retrovi- 36. R. A. Pyron, F. T. Burbrink, J. J. Wiens, A phylogeny and revised classification of Squa- ruses and transposable elements in eukaryotes. Nat. Rev. Genet. 18, 292–308 (2017). mata, including 4161 species of lizards and snakes. BMC Evol. Biol. 13, 93 (2013). 57. F. Gr€utzner et al., Chicken microchromosomes are hypermethylated and can be iden- EVOLUTION 37. A. Meyer et al., Giant lungfish genome elucidates the conquest of land by verte- tified by specific painting probes. Cytogenet. Cell Genet. 93, 265–269 (2001). brates. Nature 590, 284–289 (2021). 58. G. Hartley, R. J. O’Neill, Centromere repeats: Hidden gems of the genome. Genes 38. F. Kasai, P. C. O’Brien, M. A. Ferguson-Smith, Reassessment of genome size in turtle (Basel) 10, 223 (2019). and crocodile based on chromosome measurement by flow karyotyping: Close simi- 59. A. S. Graphodatsky, P. L. Perelman, S. J. O’Brien, Eds., Atlas of Mammalian Chromo- larity to chicken. Biol. Lett. 8, 631–635 (2012). somes (Wiley & Sons, ed. 2, 2020). 39. D. R. Schield et al., The origins and evolution of chromosomes, dosage compensation, 60. J. Kim et al., Reconstruction and evolutionary history of eutherian chromosomes. and mechanisms underlying venom regulation in snakes. Genome Res. 29, 590–601 Proc. Natl. Acad. Sci. U.S.A. 114, E5379–E5388 (2017). (2019). 61. M. Svartman, G. Stone, R. Stanyon, The ancestral eutherian karyotype is present in 40. B. W. Perry, D. R. Schield, R. H. Adams, T. A. Castoe, Microchromosomes exhibit dis- Xenarthra. PLoS Genet. 2, e109 (2006). tinct features of vertebrate chromosome structure and function with underappreci- 62. J. E. Deakin et al., Reconstruction of the ancestral marsupial karyotype from compar- ated ramifications for genome evolution. Mol. Biol. Evol. 38, 904–910 (2021). ative gene maps. BMC Evol. Biol. 13, 258 (2013). 41. O. Dudchenko et al., De novo assembly of the Aedes aegypti genome using Hi-C 63. S. J. Klein, R. J. O’Neill, Transposable elements: Genome innovation, chromosome yields chromosome-length scaffolds. Science 356, 92–95 (2017). diversity, and centromere conflict. Chromosome Res. 26, 5–23 (2018). 42. B. R. Lajoie, J. Dekker, N. Kaplan, The Hitchhiker’s guide to Hi-C analysis: Practical 64. M. J. D. White, Modes of Speciation (W. H. Freeman, San Francisco, 1978). guidelines. Methods 72, 65–75 (2015). 65. T. Cremer, C. Cremer, Chromosome territories, nuclear architecture and gene regula- 43. J. Dekker, M. A. Marti-Renom, L. A. Mirny, Exploring the three-dimensional organiza- tion in mammalian cells. Nat. Rev. Genet. 2, 292–301 (2001). tion of genomes: Interpreting chromatin interaction data. Nat. Rev. Genet. 14, 66. T. Ezaz et al., The dragon lizard Pogona vitticeps has ZZ/ZW micro-sex chromosomes. 390–403 (2013). Chromosome Res. 13, 763–776 (2005). Waters et al. PNAS j 11 of 11 Microchromosomes are building blocks of bird, reptile, and mammal chromosomes https://doi.org/10.1073/pnas.2112494118