Original Article Cytogenet Genome Res 2016;148:292–304 Accepted: April 4, 2016 DOI: 10.1159/000447478 by M. Schmid Published online: July 16, 2016 Cytogenetic Insights into the Evolution of Chromosomes and Sex Determination Reveal Striking Homology of Turtle Sex Chromosomes to Amphibian Autosomes a a a a Eugenia E. Montiel Daleen Badenhorst Ling S. Lee Robert Literman b a Vladimir Trifonov Nicole Valenzuela a b Department of Ecology, Evolution and Organismal Biology, Iowa State University, Ames, Iowa , USA; Institute of Molecular and Cellular Biology SB RAS, Novosibirsk , Russian Federation Key Words niotes revealed that the sex chromosomes of several turtles, 18S rDNA · C-bands · Evolution · Homology · Interstitial as well as mammals and some lizards, are homologous to telomeres · Karyotype evolution · NOR · Reptile · Sex components of Xenopus tropicalis XTR1 (carrying Dmrt1 ). chromosome · Sex determination · Turtle Other turtle sex chromosomes are homologous to XTR4 (car- rying Wt1 ). Interestingly, all known turtle sex chromosomes, except in Trionychidae, evolved via inversions around Dmrt1 Abstract or Wt1 . Thus, XTR1 appears to represent an amniote proto- Turtle karyotypes are highly conserved compared to other sex chromosome (perhaps linked ancestrally to XTR4) that vertebrates; yet, variation in diploid number (2n = 26–68) gave rise to turtle and other amniote sex chromosomes. reflects profound genomic reorganization, which correlates © 2016 S. Karger AG, Basel with evolutionary turnovers in sex determination. We evalu- ate the published literature and newly collected compara- tive cytogenetic data (G- and C-banding, 18S-NOR, and telo- The compartmentalization of genomes into chromo- mere-FISH mapping) from 13 species spanning 2n = 28–68 somes alters genome function, because the regulatory to revisit turtle genome evolution and sex determination. environment of genes is influenced by their synteny or Interstitial telomeric sites were detected in multiple lineages lack thereof [Ahituv et al., 2005], and co-expressed genes that underwent diploid number and sex determination turn- tend to be clustered in the genome [de Wit and van overs, suggesting chromosomal rearrangements. C-banding Steensel, 2009]. Thus, discerning the causes and conse- revealed potential interspecific variation in centromere com- quences of genome compartmentalization is key to un- position and interstitial heterochromatin at secondary con- derstanding genome function, because chromosome fu- strictions. 18S-NORs were detected in secondary constric- sion or fission events alter the number of chromosomes tions in a single chromosomal pair per species, refuting pre- and the synteny of the genes they contain. Reptilian ge- vious reports of multiple NORs in turtles. 18S-NORs are linked nomes vary in their size, organization, and composition to ZW chromosomes in Apalone and Pelodiscus and to X (not Y) in Staurotypus . Notably, comparative genomics across am- E.E.M. and D.B. contributed equally to this work. © 2016 S. Karger AG, Basel Nicole Valenzuela 1424–8581/16/1484–0292$39.50/0 Department of Ecology, Evolution and Organismal Biology Iowa State University E-Mail [email protected] Ames, IA 50011 (USA) www.karger.com/cgr E-Mail nvalenzu @ iastate.edu Downloaded by: Iowa State University 129.186.138.35 - 4/11/2017 12:17:39 AM Fig. 1. Ancestral reconstruction of diploid chromosome numbers duced at intermediate temperatures and females at lower and high- for turtle species with known diploid numbers and sex determina- er values. Three-letter acronyms are shown for the 13 target species tion mechanisms. Evolutionary transitions between temperature- whose cytogenetic data were collected during this study. Branch dependent sex determination (TSD) and genotypic sex determina- color denotes ancestral state reconstruction by parsimony follow- tion (GSD) are indicated in the branches where they were recon- ing Valenzuela and Adams [2011], whereas numbers at key nodes structed to have occurred by maximum likelihood [Valenzuela denote reconstruction by maximum likelihood using the ace func- and Adams, 2011]. GSD = GSD with uncharacterized sex chromo- tion of ape version 3.4 [Popescu et al., 2012] in R [R Core Develop- some system. TSDIa = TSD where males are produced at lower and ment Team, 2012]. females at higher temperature. TSDII = TSD where males are pro- Turtle Sex Chromosome Homology to Cytogenet Genome Res 2016;148:292–304 293 Amphibian Autosomes DOI: 10.1159/000447478 Downloaded by: Iowa State University 129.186.138.35 - 4/11/2017 12:17:39 AM Fig. 2. G-banded karyotypes of 13 target species examined in this study and their phylogenetic relationships. Taxonomic families are indicated at the corresponding nodes and sex-determining mechanisms are color coded. Green arrowheads denote the chromosomal location of the NORs. [Olmo, 2008]. Genome evolution has received increased changed for over 200 million years [Bickham, 1981]. Yet, attention as new genomic resources are developed for the wide variation in chromosome number across turtles reptiles, a group that holds an important key to under- ( fig. 1 ) point to drastic changes in genome organization standing vertebrate evolution [Wang et al., 2006; Janes that took place during this time, and these rearrange- et al., 2008; Alfoldi et al., 2011; Castoe et al., 2013; Shaf- ments are curiously associated with evolutionary transi- fer et al., 2013; Vonk et al., 2013; Wang et al., 2013; tions in sex-determining mechanisms [Valenzuela and Green et al., 2014]. Adams, 2011]. Here, we provide new cytogenetic data Within reptiles, turtles have been considered to pos- from selected species on the evolution of turtle karyo- sess highly conserved karyotypes [Olmo, 2008; Pokorná types and review our knowledge of chromosomal sex de- et al., 2011] as some chromosomes appear to be un- termination in turtles. 294 Cytogenet Genome Res 2016;148:292–304 Montiel/Badenhorst/Lee/Literman/ DOI: 10.1159/000447478 Trifonov/Valenzuela Downloaded by: Iowa State University 129.186.138.35 - 4/11/2017 12:17:39 AM Methods ronym hereafter. A summary of new and previously pub- lished data is presented in table 1 . While standardized sys- Cell Culture, Chromosome Preparation, and Chromosome tems of chromosome nomenclature exist for some organ- Banding Chromosome preparations were obtained from fibroblast cell isms, including humans, domestic animals, and some cultures established from embryonic or muscle tissue following plants [e.g., Linde-Laursen et al., 1997; Cribiu et al., 2001; standard procedures as previously described [Valenzuela, 2009; Simons et al., 2013], no such system exists for turtles. This Badenhorst et al., 2013, 2015; Valenzuela et al., 2014; Montiel et al., is partly due to the scarcity of uniform banding informa- 2016]. New samples were obtained from private collections or pet trade specimens. Briefly, fibroblast cell cultures were established tion across taxa, which limits our understanding of turtle from collagenase (Sigma) digests and cultured using a medium genome macroevolution [Olmo, 2005]. Here, we adopted which was composed of 50% RPMI 1640 and 50% Leibowitz media the organization of chromosomes by decreasing size and supplemented with 15% fetal bovine serum, 2 m M L-glutamine, centromere position as is commonly used today. Namely, and 1% antibiotic-antimycotic solution (Sigma). Cultures were in- G-banded karyotypes ( fig. 2 ) were ordered by approxi- cubated at 30 ° C with no CO2 supplementation [Badenhorst et al., 2013]. Four hours prior to harvesting, 10 μg/ml colcemid (Karyo- mate size. In general, smaller microchromosomes were MAX ® , Invitrogen) was added to the cultures, which were har- nearly indistinguishable from each other by shape and vested after hypotonic exposure and fixed in 3: 1 methanol:acetic centromere position, and thus, they were ordered by ap- acid following standard procedures [Ezaz et al., 2006; Martinez et proximate size and G-banding pattern where possible. al., 2008; Badenhorst et al., 2013]. G- and C-banding followed con- The diploid number in turtles ranges from 2n = 26 in ventional protocols [Seabright, 1971; Sumner, 1972]. Peltocephalus dumerilianus [Killebrew, 1975a; Ventura et NOR Mapping by 18S rRNA FISH al., 2014] to 2n = 68 in Carettochelys insculpta [Bickham The distribution of NORs was investigated by FISH using a et al., 1983] with an average of 2n = 52 given our current turtle-specific 18S DNA fragment labeled by nick-translation and knowledge of turtle chromosomes [Olmo and Signorino, co-precipitated with salmon sperm DNA [Badenhorst et al., 2013]. 2005] ( fig. 1 ). Our comparative cytogenetic data ( fig. 2 ) Telomere FISH Mapping illustrate a general characteristic of turtle karyotypes that A telomeric probe containing the repeat motif (TTAGGG)n is shared by birds and squamates, namely, the presence of was constructed and labelled with biotin or DIG-dUTP by PCR 2 distinct components, macro- and microchromosomes following Ijdo et al. [1991]. The probe was hybridized to meta- [Olmo et al., 2002]. The presence of microchromosomes phase chromosomes of all target species following Badenhorst et appears to be the ancestral condition to amniotes as it is al. [2013]. Briefly, chromosome spreads and the DNA probe were also found in ancestral fish and amphibian lineages, and denatured together in a sealed slide on a hot plate at 65 ° C for 3 min. Hybridization took place in a humid chamber at 37 ° C for they appear to have fused among themselves in crocodil- 2 nights. Post-hybridization