Genome Enablement of the Notothenioidei: Genome Size Estimates from 11 Species and BAC Libraries from 2 Representative Taxa

Abstract

The perciform suborder Notothenoidei provides a compelling opportunity to study the adaptive radiation of a marine species flock in the cold Southern Ocean surrounding Antarctica. To enable genome-level studies of these psychrophilic fishes, we estimated the sizes of the genomes of 11 Antarctic species and generated high-quality BAC libraries for 2, the notothen Notothenia coriiceps and the icefish Chaenocephalus aceratus. Our results indicate that evolution of phylogenetically derived notothenioid families, [e.g., the icefishes (Channichthyidae)], was accompanied by genome expansion. Species (n = 6) of the basal family Nototheniidae had C values that ranged between 0.98 and 1.20 pg, whereas those of the icefishes, the notothenioid crown group, were 1.66–1.83 pg (n = 4 species). The BAC libraries VMRC-19 (N. coriiceps) and VMRC-21 (C. aceratus) comprised 12X and 10X coverage of the respective genomes and had average insert sizes of 138 and 168 kb. Greater than 60% of paired BAC ends sampled from each library (~0.1% of each genome) contained repetitive sequences, and the repetitive element landscapes of the 2 genomes (13.4% of the N. coriiceps genome and 14.5% for C. aceratus) were similar. The representation and depth of coverage of the libraries were verified by identification of multiple Hox gene contigs: six discrete Hox clusters were found in N. coriiceps and at least five Hox clusters were found in C. aceratus. Given the unusual anatomical and physiological adaptations of the notothenioids, the availability of these BAC libraries sets the stage for expanded analysis of the psychrophilic mode of life.

The modern ichthyofauna of the Southern Ocean is uniquely depauperate with respect to species diversity (Eastman, 2005). Although the Southern Ocean constitutes approximately 10% of our planet’s oceanic waters, the number of fish species is quite small (~300 species vs. a global diversity of ~25,000 species), and almost half belong to a single perciform suborder, the Notothenioidei. The origin of this taxonomic restriction results from the dramatic paleoclimatic and paleogeographic history of Antarctica (Eastman, ’93) during the Cenozoic Era.

The opening of the Drake Passage, which separates Antarctica from South America, occurred via tectonic rifting ~40–34 million years (Myr) ago (Kennett, ’77; Livermore et al., 2005; Scher and Martin, 2006). The onset of widespread glaciation of Antarctica ~34 Myr ago (Zachos et al., 2001), which is attributable to declining atmospheric CO₂ (DeConto and Pollard, 2003), the establishment of the Antarctic Circumpolar Current and the development of the Antarctic Polar Front (between ~50 and 60°S), decoupled the Southern Ocean from warmer, subtropical waters to the north (Kennett, ’77). A cooling trend that began in the late Paleocene/early Eocene (~55 Myr ago), when water temperatures were 15–20°C, accelerated during the Miocene until the Southern Ocean approached the freezing point of seawater (−1.9°C) ~14–5 Myr ago (Clarke, ’90).

As the Southern Ocean cooled, the shallow water, cosmopolitan, and temperate fish fauna of late Eocene (38 Myr ago) became largely extinct owing to the destruction of inshore habitat and changes in trophic structure caused by repeated ice sheet scouring of the continental margin (Eastman, 2005). Thus, fish diversity was reduced and new ecological niches became available to taxa, such as the notothenioid fishes, which were diversifying in situ. Today, species of the notothenioid suborder constitute 46% of all the fish species of the Southern Ocean. At the highest latitudes, notothenioid species dominance is particularly clear, encompassing 77% of species diversity and 90% of biomass (Eastman, 2005).

The ancestral notothenioid stock probably arose as a sluggish, bottom-dwelling perciform species some 40–60 Myr ago in the cool temperate shelf waters of the Antarctic continent (DeWitt, ’71; Eastman, ’93; Eastman and Clarke, ’98). With the elimination of most of the Eocene ichthyofauna, the notothenioids diversified to fill the many vacated niches. Lacking swim bladders, the Notothenioidei evolved toward pelagic or partially pelagic zooplanktivory and piscivory by reduction of skeletal mineralization and enhancement of lipid deposition. Termed pelagization, the tailoring of morphology for life in the water column is the hallmark of the notothenioid radiation, has arisen independently several times in different clades (Eastman, ’97, ’99; Near et al., 2007), and reflects the retention of larval characteristics in the adult (pedomorphism) (Eastman, ’97). In the dominant family Nototheniidae, ~50% of the Antarctic species are semipelagic, epibenthic, cryopelagic, or pelagic (Eastman, 2005). On the basis of their rapid speciation (average time for speciation of 0.76–2.1 Myr), geographical restriction, and high endemism, the Notothenioidei are the best described example of a marine species flock (Eastman, 2000; Eastman and McCune, 2000).

Absence of niche competition alone would not have guaranteed the radiation of the notothenioids had they not evolved biochemical and physiological adaptations to their frigid environment. One key evolutionary adaptation of the notothenioids was the de novo acquisition of antifreeze glycoproteins, which bind to the surfaces of ice crystals that form in these fishes to inhibit ice propagation in their hyposmotic fluids (Cheng, ’98; Cheng and Detrich, 2007). The evolution of the antifreeze glycoprotein genes from a pancreatic, trypsinogen-like gene (Chen et al., ’97; Cheng and Chen, ’99) is a remarkable exemplar of a critical genetic innovation that fostered an extensive species radiation (Montgomery and Clements, 2000).

The notothenioids are notable not only for their resistance and compensatory adaptations to the extreme Antarctic marine environment, but also for their regressive evolutionary changes or losses of function [the “disaptations” of Montgomery and Clements (2000) and references therein]. Antarctic notothenioids are cold stenotherms—these high latitude species typically live at temperatures between −1.9 and +2°C and have incipient lethal temperatures near 5°C that cannot be raised by acclimation in the laboratory (Somero and DeVries, ’67). This stenothermality results in part from adaptive alterations to the biochemical, cellular, and physiological systems of these fishes. Major examples of compensatory adaptation include efficient microtubule assembly (Williams et al., ’85; Detrich et al., ’89, ’92, 2000; Paluh et al., 2004) and protein translocation at cold temperatures (Römisch et al., 2003), homeoviscous adaptation of membrane lipids to preserve membrane fluidity (Logue et al., 2000), and cold-stable lens crystallins that prevent cataract formation (Kiss et al., 2004). Consistent with their long isolation in a constantly cold environment, the Antarctic notothenioids have lost the inducible heat shock response (Hofmann et al., 2000; Buckley et al., 2004), which seems to have been recruited to an adaptive, constitutive status to deal with elevated denaturation of proteins caused by cold stress (Place et al., 2004; Place and Hofmann, 2005; Todgham et al., 2007). Other striking regressive changes include the loss of erythrocytes (Ruud, ’54) and the respiratory transport protein hemoglobin (Cocca et al., ’95, 2000; Zhao et al., ’98; Near et al., 2006) by all species of the icefish family and independent losses of cardiac myoglobin in a subset of icefish species (Sidell et al., ’97; O’Brien and Sidell, 2000; Sidell and O’Brien, 2006).

Recently, Chen et al. (2008) reported a large-scale transcriptomic up-regulation of 177 gene families in the Antarctic toothfish, Dissostichus mawsoni (Nototheniodei: Nototheniidae), with respect to their orthologs in temperate and tropical fish. At the genomic level, they found that 118 gene families in several notothenioid species have undergone substantial gene duplication relative to the temperate/tropical species. Overlap of the duplicated gene set with the transcriptionally up-regulated gene set suggested that gene family expansion may contribute to cold adaptation of cellular and physiological function, a conclusion supported by the augmentation of tubulin genes in the bullhead notothen, Notothenia coriiceps (Parker and Detrich, ’98). It is notable that 17 families of LINEs showed Antarctic-specific duplication, which suggests that LINE expansion may have facilitated gene duplication in Antarctic notothenioids by retrotransposition-mediated gene transduction (Xing et al., 2006).

The above findings suggest that the genomes of notothenioids are evolutionarily dynamic and that this plasticity may have contributed to the overall success of the group. The development of genomic resources from these difficult-to-obtain specimens is thus important, well justified, and valuable to a wide community of genomic, evolutionary, and ecological biologists and biomedical scientists. In 2003, we collected high molecular weight (HMW) DNA from a wide assortment of notothenioid fishes, with the aim of enabling genomic studies in this group, and subsequently submitted a white paper to the BAC library network at the National Human Genome Research Institute (NHGRI) to request construction of high-quality and high-representation BAC genomic libraries from selected notothenioid species. We were informed that BAC libraries would be generated from two of the species: Chaenocephalus aceratus (blackfin or Scotia Arc icefish) and N. coriiceps (bullhead notothen or yellowbelly rockcod). The selection of these two species was largely strategic in that C. aceratus represents a lineage (icefishes) devoid of erythrocytes and lacking hemoglobin genes, whereas N. coriiceps represents an erythrocyte-producing species in which hemoglobin genes are intact and functional. Thus, these two species, although not representative of the entire gamut of notothenioids, are extremely useful for genomic comparisons of processes that involve differences in their respective physiologies and anatomies as a result of hemoglobin loss by the icefishes. In this article, we describe the generation of BAC libraries from these 2 representative notothenioid species and further provide genome size estimates for 11 species of notothenioid fishes. Quality assurance data are provided for both BAC libraries that validate the utility of these resources for genomic analyses.

MATERIALS AND METHODS

Collection of Specimens

Specimens of the 11 notothenioid species (see Table 1 for names) used in this study (including the Antarctic blackfin icefish, C. aceratus, and the bullhead notothen, N. coriiceps) were collected by bottom trawls or via baited fish traps deployed from the ARSV Laurence M. Gould in Dallmann Bay near Brabant Island in the Palmer Archipelago. They were transported alive to Palmer Station, Antarctica, where they were maintained in seawater aquaria at −1.5 to +0.5°C.

Table 1.

Estimated genome sizes of notothenioid fishes relative to outgroups.

Family	Species	No. examined	C value (pg)
Gasterosteidae	Gasterosteus aculeatusa,b,c	NA	0.58–0.70
Percidae	Perca flavescens,c,d P. fluviatilis,a,e Sander luciopercaa	NA	0.90–1.22
Serranidae	Mycteroperca interstitialis,c M. tigris,b,c,f Siniperca chuatsig	NA	0.80–1.30
Nototheniidae	Dissostichus mawsoni	3	1.20±0.04
	Gobionotothen gibberifrons	1	0.98
	Lepidonotothen nudifrons	1	1.12
	Notothenia coriiceps	2	1.13±0.21
	Trematomus larseni	2	1.16±0.001
	Trematomus newnesi	1	1.01
Bathydraconidae	Parachaenichthys charcoti	2	1.34±0.05
Channichthyidae	Chaenocephalus aceratus	2	1.79±0.14
	Champsocephalus gunnari	2	1.83±0.22
	Chionodraco rastrospinosus	2	1.66±0.04
	Pseudochaenichthys georgianus	2	1.79±0.09

NA, not applicable.

^aVinograd (’98);

^bHinegardner (’68);

^cHinegardner and Rosen (’72);

^dHardie and Hebert (2004);

^eVialli (’57);

^fMirsky and Ris (’51);

^gCui et al. (’91).

Sample Collection for Genome Size Estimates and High Molecular Weight DNA

All specimens used in this study were anesthetized with MS222 (0.1% (w/v)). Red-blooded species (non-icefishes) were bled by collecting blood through heparinized microhematocrit tubes from small incisions in the caudal sinus. For each specimen, we collected approximately 100 μL of blood. Blood was transferred to siliconized 1.5 mL microfuge tubes and brought to 1 mL with 1 × phosphate buffered saline (PBS). A 50 μl aliquot (for genome size determination) was fixed in 70% ethanol for 30 min and frozen at −70°C. The remainder was centrifuged briefly at 3,000 rpm and the supernatants removed. Cells were resuspended in 500 μL of PBS, counted with a hemacytometer, brought to a cell concentration of ~10⁵ cells/mL in 1% InCert agarose (FMC) in 1 × PBS, and then embedded in Bio-Rad plug molds (80 μL per plug). HMW DNA was prepared in situ in the agarose plugs using methods described earlier (Amemiya et al., ’96).

For icefishes, it was necessary to collect significant amounts of blood (>25 mL) in order to isolate sufficient numbers of leucocytes for flow cytometry and agarose embedding owing to the relative low numbers of cells per hematocrit volume (Hureau et al., ’77). Ten-milliliter heparinized syringes were used to collect blood via cardiac puncture of anesthetized specimens using 18-gauge needles. Blood samples were transferred to 15 mL conical centrifuge tubes, cells were collected by centrifugation for 10 min in a refrigerated table top centrifuge at 2,000 rpm, and the buffy coats were removed. Cells were transferred to siliconized 1.5 mL microfuge tubes, and samples were handled as described above for the red-blooded species.

Flow Cytometric Determination of Genome Size

Samples for flow cytometry were thawed on ice, centrifuged at 3,000 rpm for 30 sec in a microfuge, and the respective supernatants removed. Cells were resuspended in 1 mL of 1 × PBS and allowed to equilibrate for at least 30 min on ice. Just before flow cytometry, cells were pelleted as above, resuspended in propidium iodide staining solution (Arumuganathan and Searle, ’91), and incubated on ice for 15 min. Flow cytometric measurements were performed using a FACSort flow cytometer (Becton-Dickinson, San Jose, CA). An internal standard, chicken erythrocyte nuclei (2.33 pg/2C nucleus), was included in each run. Several thousand cells were measured per sample and each experiment consisted of four replicates. The genome sizes of the 11 species were calculated via comparison to the internal standard.

BAC Cloning

HMW DNAs were analyzed from two specimens each of C. aceratus and N. coriiceps via pulsed field gel (PFG) analysis on a CHEF DRIII apparatus (Bio-Rad, Hercules, CA). The DNAs were of generally good quality based on their size distribution in the PFG runs (Amemiya et al., ’96). BAC library construction closely followed earlier published methods (Osoegawa et al., ’98; Danke et al., 2004). In short, HMW DNA from one specimen of each species was subjected to a series of partial digest titrations using EcoRI and EcoRI methylase, visually assessed via analytical PFG electrophoresis, then subjected to preparative scale-up using optimized digestion conditions. Restriction fragments were run in preparative PFGs using a CHEF–Mapper apparatus (Bio-Rad), fragments of desired sizes excised and electroeluted (Strong et al., ’97), and purified fragments ligated with EcoRI-digested and dephosphorylated pCCBACE1 vector (Epicentre Technologies, Madison, WI). The ligation products were then transformed into DH10B (T1 resistant) electro-competent cells (Invitrogen) and plated onto LB-agar media supplemented with chloramphenicol (12.5 μg/mL). Crude analysis of the transformant insert sizes was performed with the Colony Fast-Screen system (Epicentre Technologies); PFG analysis of NotI-digested transformants was subsequently used to more accurately analyze the insert sizes and to determine which size fraction would ultimately be utilized for construction of the libraries. Once selected, the appropriate size fractions were ligated in scaled-up reactions and bulk transformations were carried out. Transformants were flash frozen (liquid nitrogen bath) in SOC medium containing 7.5% glycerol and stored at −70°C. Before arraying, frozen transformants were thawed at room temperature and ~400 colonies were plated per 10 cm² LB agar plate containing 12.5 μg/mL chloramphenicol, 120 μg/mL X-gal, and 0.1 mM IPTG. Colonies were cultured overnight at 37°C. White (recombinant) colonies were picked into 384-well microtiter dishes (Genetix, Boston, MA) containing LB media with 12.5 μg/mL chloramphenicol and 7.5% glycerol using a colony picking robot (Norgren Systems, Ronceverte, WV) and then cultured overnight at 37°C and stored at −70°C. Both libraries encompass at least 10X coverage based on the average insert sizes, the percentage of empty clones (Table 2), and the respective genome size as estimated by flow cytometry (described above). Replicas of the library as well as high-density colony filters (22 cm²) were made using the Total Array System (BioRobotics, Boston, MA).

Table 2.

Summary statistics for BAC libraries constructed in this study.

	C. aceratus (VMRC-21)	N. coriiceps (VMRC-19)
Vector used	pCC1BAC	pCC1BAC
Restriction enzyme used	EcoRI	EcoRI
DNA source	Leucocytes	Erythrocytes
Number of 384-well plates	336	288
E. coli strain used	DH10B T1 resistant	DH10B T1 resistant
Empty wells	0	0
Estimated non-recombinant clones and vector artifact clones	6,451 (5%)	8,847 (8%)
Estimated insert-containing clones	122,573 (95%)	101,745 (92%)
Average insert size	168 kb	138 kb
Median insert size	168 kb	155 kb
Estimated genome coverage	10X	12X
Restriction enzyme used	EcoRI	EcoRI

Quality Control Experiments

Average insert sizes were determined from ~120 randomly selected clones per library via restriction analysis as earlier described (Danke et al., 2004). To establish insert cloning efficiency and obtain estimates of repetitive element densities in the two notothenioid fishes, paired reads of BAC clones representing ~0.1% of each genome were obtained by automated DNA sequencing (The Institute for Genomic Research, now the J. Craig Venter Institute, Rockville, MD). Sequence from a total of 2.0 × 10⁶ bp of the C. aceratus genome, representing 2,733 ends, and of 1.0 × 10⁶ bp of the N. coriiceps genome (1,388 ends) was obtained. The end sequence data were also analyzed for repetitive elements using RepeatMasker v3.0 (Smit et al., 2004), CENSOR (Kohany et al., 2006), and the Repbase Update (Jurka et al., 2005). To assure that the library encompassed expected genomic regions we hybridized high density colony filters with a generic Hox homeobox probe and restriction fingerprinted all resultant clones using an automated fingerprinting method (Schein et al., 2004).

Animal Care

Research involving Antarctic fishes was approved by Northeastern University’s Institutional Animal Care and Use Committee and conformed to NIH guidelines.

RESULTS AND DISCUSSION

Genome Size Estimates

Figure 1 illustrates the relationships between the 11 notothenioid species examined in this study. The six notothens (family Nototheniidae, n = 50 species total), one dragonfish (Bathydraconidae, n = 16), and four icefishes (Channichthyidae, n = 16) represent a cross section of red- and white-blooded species. [See Eastman (2005) and his web site http://www.oucom.ohiou.edu/dbms-eastman for a complete listing of valid notothenioid species.] Percids (perches) are thought to be the sister group to the notothenioids, and the Serranidae (sea basses) and Gasterosteidae (sticklebacks) are also closely related (Dettaï and Lecointre, 2004).

Figure 1.

Phylogenetic relationships among the 11 notothenioid species examined in this study. Six notothens of a basal, red-blooded family (Nototheniidae) and four icefishes of the white-blooded notothenioid crown group (Channichthyidae) are shown with the size ranges of their genomes (see Table 1 for species names and genome estimates). One species (P. charcoti) of the red-blooded dragonfish (Bathydraconidae), the sister group to the icefish, is also presented. The perches (Percidae) are the probable sister group to the Notothenioidei (see Table 1 for the species used to obtain the genome range estimate). The positions of the notothenioid families Bovichtidae, Eleginops, Artedidraconidae, and Harpagiferidae (gray font; not examined herein) are shown for comparison. The closed circles indicate key evolutionary events during the diversification of the Notothenioidei: the acquisition of antifreeze proteins; the decline in hemoglobin multiplicity and performance; and the losses of hemoglobins and of red blood cells by the icefishes. The tree is based on the studies of Lecointre et al. (’97), Balushkin (2000), Near et al. (2004), and Near and Cheng (2008).

Genome size determinations were estimated for the notothenioids by flow cytometry using erythrocytes from non-icefish specimens and leukocytes from icefishes, with chicken erythrocyte nuclei as an internal standard. Figure 2 shows graphical results from representative flow cytometry experiments for two red-blooded notothens, Gobionotothen gibberifrons and N. coriiceps (panels A, B), and for two white-blooded icefishes, C. aceratus and Chionodraco rastrospinosus (panels C, D). Genome size estimates for all species examined in this study, and for several related groups, are summarized in Figure 1 and Table 1.

Figure 2.

Representative flow cytometric histograms of four notothenioid specimens: (A) Gobionotothen gibberifrons; (B) Notothenia coriiceps; (C) Chaenocephalus aceratus; (D) Chionodraco rastrospinosus. The Y-axis represents the number of cells counted and the X-axis represents propidium iodide fluorescence in arbitrary units (AU). The peak labeled “C” in each panel was obtained from chicken erythrocyte nuclei (internal control, 2.33 pg per 2C nucleus); relative DNA contents were determined by extrapolation with this internal control and are given in Table 1. The peaks for the two nototheniids (A and B) fall to the left of the internal standard, whereas those for the two channichthyids (C and D) fall to the right of the internal standard and indicate the larger genome sizes for the two icefishes. The plots shown in A–D are from single flow cytometric runs; an additional three runs were performed for each specimen in order to assure consistency and accuracy. Thus, the final determination of genome size was based on several thousands of cell measurements.

Although our taxonomic sampling of the Notothenioidei is incomplete, there is a trend toward increased genome size as one moves from the phylogenetically basal Nototheniidae to the derived families Bathydraconidae and Channichthyidae. The basal Nototheniidae, by contrast, have genomes whose sizes are comparable to those of outgroups Percidae, Serranidae, and Gasterosteidae, suggesting that a C value of ~1 pg is ancestral for the notothenioid suborder. Our genome size estimates for the 11 notothenioids fall in a range that is consistent with the limited number of earlier measurements (Morescalchi et al., ’92, ’96; Hardie and Hebert, 2003, 2004; Chen et al., 2008). Furthermore, comparison of data for the two species common to our study and earlier reports, Antarctic toothfish D. mawsoni and blackfin icefish C. aceratus, indicate that our estimates are robust. We calculate that the genome of D. mawsoni contains 1.20 pg of DNA, which is slightly greater than values (0.97, 1.03 pg) reported earlier (Hardie and Hebert, 2003, 2004; Chen et al., 2008), whereas our result for C. aceratus (1.79 pg) is slightly smaller than the 1.86 pg reported by Chen et al. (2008). The small sample size of our study precludes a definitive analysis of the patterns of genome evolution, but our estimates do support the hypothesis that genome expansion accompanied the evolution of phylogenetically derived notothenioids. In particular, the icefishes (channichthyids), which are considered the most phylogenetically derived group within the notothenioids, have notably larger genomes (Figs. 1 and 2; Table 1).

Chen et al. (2008) recently reported that 118 gene families are substantially expanded in polar notothenioids with respect to temperate representatives, and that five notothenioid species show a trend toward increased genome size that correlates with a derived phylogenetic position. They suggest that evolutionary adaptation of Antarctic notothenioids to chronic and extremely cold temperature has led to expansion of gene families involved in metabolic processes (e.g., protein biosynthesis, folding and degradation, lipid metabolism, etc.) that are critical to the physiological fitness of these organisms. Our results provide a complementary, non-exclusive rationale, namely that genome expansion may have been driven in part to offset reduced hemoprotein function in polar notothenioids. Relative to temperate and tropical fishes, basal notothenioid clades (i.e., Nototheniidae) have low hematocrits and red cell hemoglobin concentrations (Kunzmann et al., ’91). Furthermore, hemoglobin multiplicity, functionality, and concentration tend to decline in red-blooded species the more derived their notothenioid group (D’Avino and di Prisco, ’88; Kunzmann et al., ’91; Bargelloni et al., ’98; di Prisco, ’98; di Prisco et al., 2007). Loss of hemoprotein expression by icefishes was clearly maladaptive as shown by their numerous compensatory cardiovascular adaptations, including increases in mitochondrial densities in heart muscle, cardiac output, and capillary density (Near et al., 2006; Sidell and O’Brien, 2006; Cheng and Detrich, 2007; Detrich, 2007). Thus, we hypothesize that the genomes of icefishes also expanded to provide additional gene templates to support the observed changes in cardiovascular physiology. The two BAC libraries described in this report provide the opportunity to examine these and other hypotheses in detail.

BAC Library Construction

BAC libraries were constructed for two notothenioids, the red-blooded notothen N. coriiceps and the white-blooded channichthyid C. aceratus. These taxa were selected by the NIH BAC Resource Network (http://www.genome.gov/10001844) as representative of two disparate lineages of the Antarctic notothenioids that, by comparative genomics, could reveal new information regarding the genetics of evolutionary adaptation to extreme environments. Construction of the libraries was performed under the auspices of the NHGRI and followed rigid quality control standards that mandated: (1) 10X genomic coverage; (2) average insert size of ~150 kb; (3) assurance that the library contained a minimum of 95% insert-containing clones; and (4) little or no detectable contamination by bacterial DNA sources.

The single most important factor in generating large insert BAC clones is the preparation of HMW DNA of impeccable integrity and quality. Agarose embedding of cells and subsequent processing of DNA needed to be accomplished while in the field in Antarctica under less than ideal laboratory conditions. Nonetheless, efforts were made to process the specimens immediately after collection and to apply fastidious technique. Because a PFG apparatus was not available in the field in Antarctica, the quality of the isolated DNAs could be assessed only after the samples were transported back to Seattle. Based on visual inspection of these DNAs by PFG electrophoresis (not shown), it was concluded that most of the DNAs prepared in the field (including those of C. aceratus and N. coriiceps) were of relatively HMW and seemed to be of suitable quality for BAC library construction.

In pilot experiments, agarose-embedded DNAs from both species were subjected to partial digestions with EcoRI. After PFG verification, optimized reaction conditions were scaled up for preparative runs (Osoegawa et al., ’98). Large-scale digests were separated on preparative pulsed field gels and fractions were taken. The quality of EcoRI partial digests was assessed by analytical pulsed field gels (Fig. 3A). Fractions containing fragments in the desired size range (roughly between 140 and 200 kb) were used to scale up the ligations and transformations (which were subsequently archived in bulk at −70°C). Figure 3B shows a gel of 20 randomly selected C. aceratus clones that were digested with NotI in order to determine insert sizes. In total, more than 100 individual BAC clones were analyzed in this way and more than 200 additional clones were analyzed using the less robust Epilyse^TM method (Epicentre Technologies). Once it was determined that the insert sizes and cloning efficiency met the NHGRI quality standards, clones were thawed, plated, and robotically picked into 384-well microtiter dishes. Summary statistics for the resultant libraries are given in Table 2. The libraries have been designated VMRC-19 (N. coriiceps) and VMRC-21 (C. aceratus) and comprise 12X and 10X coverage of the respective genomes based on the total number of clones in the library, the average insert sizes, and the genome sizes determined for the taxa (Table 1).

Figure 3.

Analytical pulsed field gels for Chaenocephalus aceratus BAC library construction. (A) Samples from successive partial digest fractions (1–8) excised from a preparative pulsed field gel of bulk digested C. aceratus HMW DNA. Fractions 5 and 6 were eluted from the preparative gel and pooled for construction of the BAC library. (B) Analysis of C. aceratus BAC clones generated from fractions 5 and 6. Miniprep DNAs from 20 individual clones were digested with NotI (which excises the insert); the vector band is indicated by the arrow. Insert sizes in this batch of clones ranges from ~80 kb (clone 2) to ≥170 kb (clones 3 and 12). The doublet appearance of many of the bands is owing to a gel loading artifact. Molecular weight standards were run on the outer lanes of the gels; sizes are indicated in kb.

BAC End Sequencing—Quality Control

To determine the efficiency of insert cloning and to assess potential bacterial contamination of the two libraries, paired reads of BAC clones representing ~0.1% of each genome were generated. Sequence reads averaging 734 bp were obtained from 89% of the C. aceratus insert ends and comparable values were found for N. coriiceps (756 bp, 90%) (Table 3). Clones consisting only of vector sequences were rare (5.8%) and inserts of bacterial origin were not detected.

Table 3.

Summary statistics for BAC end sequence (BES) analyses.

Datum	C. aceratus	N. coriiceps
Number of BAC clones	1536	768
Total BESs (forward and reverse)	2733	1388
Non-successful BESs	339	148
Total Mb of sequences surveyed	2.0	1.0
Genome size (pg)	1.79	1.13
Percent genome sequenced	0.12	0.10
Minimum sequence length (bp)	63	35
Average sequence length (bp)	734	756
Maximum sequence length (bp)	985	961
Percent GC content	42	42
Percent BAC ends with repetitive sequence	68.2	62.3
Percent transposable elements (TEs)	13	12
% SSR and satellites	1.5	1.4

BAC End Sequencing—Genome Characteristics and Repetitive Element Analysis

Large-scale BAC end sequencing of C. aceratus and of N. coriiceps (Table 3) indicated that the GC contents of the two species are comparable (42%), in agreement with earlier estimates of notothenioid genomes (41–43%) obtained by analytical ultracentrifugation (Bucciarelli et al., 2002). Furthermore, their repetitive element landscapes [transposable elements (TEs), simple-sequence repeats (SSRs) and satellite DNAs] are similar (Table 3); 14.5% of the 2 Mb of C. aceratus sequence consisted of TEs, SSRs, and satellite DNA compared with the 13.4% found for 1Mb of N. coriiceps. sequence Greater than 60% of the BAC ends sampled from each library contained repetitive sequences (Table 3), with a somewhat higher proportion found in the icefish (68.2% for C. aceratus, 62.3% for N. coriiceps).

Figure 4 and Table 4 provide details of the repetitive elements of the two notothenioid genomes. When extrapolated to genome size, their repetitive element profiles were very similar (Fig. 4B). Examination of individual DNA transposons and LTR and non-LTR retrotransposons reinforced the similarity of the repetitive element landscapes of the two fishes, but some exceptions were found. For example, representation of the Mariner/Tc1 transposon, Gypsy and Copia LTR retrotransposons, and SINE elements differed between C. aceratus and N. coriiceps. We did not observe the LINE expansion (17 elements whose numbers increased 8- to >300-fold in more phylogenetically derived notothenioids) reported by Chen et al. (2008), most likely because our random sampling regime lacked the specificity of their EST-based approach.

Figure 4.

Repetitive element landscapes of the genomes of C. aceratus and of N. coriiceps. (A) Total genomic copy numbers of repetitive elements for C. aceratus and for N. coriiceps. (B) Repetitive element copy numbers normalized to genome size. Species: C. aceratus, white bars; N. coriiceps, red bars. Abbreviations: CR1, chicken repeat 1; DIRS1, Dictyostelium intermediate sequence repeat 1; ERV, endogenous retroviral sequence; hAT, hobo, Ac, and Tam3 family; L1, long interspersed element 1; LTR, long terminal repeat; SINE, short interspersed element; SSR, simple sequence repeat; TE, transposable element. Repetitive elements in the paired-end reads of the BAC inserts were evaluated by use of RepeatMasker v3.0 (Smit et al., 2004), CENSOR (Kohany et al., 2006), and the Repbase Update (Jurka et al., 2005). Error bars were calculated by the method of Shedlock et al. (2007).

Table 4.

Repetitive elements in BAC end sequences.

Repetitive element classa	C. aceratus	N. coriiceps
Transposable elements	1,390	618
DNA transposons	566 (40.9)	253 (41.1)
Mariner/Tc1	27 (1.9)	32 (5.2)
Kolobok	32 (2.3)	17 (2.8)
Harbinger	30 (2.2)	7 (1.1)
hAT	211 (15.2)	91 (15.0)
Other	266 (19.1)	106 (17.0)
LTR retrotransposons	258 (18.3)	80 (13.0)
Gypsy	156 (11.0)	49 (7.9)
Copia	25 (1.8)	1 (0.2)
DIRS	46 (3.3)	17 (2.8)
Other	31 (2.2)	13 (2.1)
Non-LTR retrotransposons	468 (33.4)	241 (38.6)
CR1	337 (24.0)	157 (25.0)
L1	46 (3.3)	31 (5.0)
SINE	52 (3.7)	37 (6.0)
Other	33 (2.4)	13 (2.6)
Endogenous retrovirus	98 (7.0)	44 (7.0)
SSRs and Satellites	474	246

^aSee legend to Figure 4 for abbreviations.

Hox Cluster Isolation

To confirm that the respective libraries contained authentic gene sequences at appropriate depth of coverage, we screened high density colony filters of each library using a cocktail of Hox probes. These probes were generated from C. aceratus and N. coriiceps genomic DNA using degenerate PCR (Kim et al., 2000). In total, we scored 125 hybridizing clones for C. aceratus and 81 for N. coriiceps. These clones were then assembled into contigs using an automated restriction fingerprinting procedure (Fjell et al., 2003; Schein et al., 2004), which tentatively identified at least 5 discrete Hox contigs in C. aceratus (average size ~21 clones) and 6 discrete contigs in N. coriiceps (average size ~10 clones). An example of this finger-printing procedure is given in Figure 5. All contigs generated by the automated fingerprinting procedure were, by sequence analysis, Hox gene-containing (data not shown). The reason for the discrepancy between the clone numbers in the C. aceratus vs. N. coriiceps Hox contigs is not known, but may be related to differences in probe labeling. Nonetheless, the depth of coverage (~10 and ~21 clones per Hox cluster for N. coriiceps and C. aceratus, respectively) is in line with the estimated coverage of these libraries (≥10X). In addition, the fact that >4 Hox clusters were found in these taxa clearly supports the existence of a whole genome duplication event at the base of the teleost lineage (Wittbrodt et al., ’98; Meyer and Schartl, ’99; Crow et al., 2006). Details of the Hox clusters of the respective species will be reported elsewhere.

Figure 5.

Restriction fingerprint and contig assembly of a Hox cluster from C. aceratus. The 16 clones in this contig were assembled from HindIII fingerprints of 125 putative Hox-containing clones that were selected on the basis of hybridization with a bulk Hox homeobox probe. (A) Fragments from 16 clones that comprised an assembled contig are depicted in a gel format output from the automated fingerprinting procedure; bars indicate fragments that were used for the assembly algorithm. (B) Tiling path of the 16 BAC clones of the assembled contig. DNA sequencing of Hox PCR amplicons verified that the contig comprised a Hox cluster (HoxDa).

Implications of the Data and Public Accessibility

Fishes of the suborder Notothenoidei present a unique opportunity to study the adaptive radiation of a marine species flock in an extremely cold environment, the Southern Ocean. To enable genome-level studies among these psychrophilic fishes, we determined genomic sizes for 11 notothenioid species and generated high-quality BAC libraries for two, N. coriiceps (VMRC-19) and C. aceratus (VMRC-21). The choice of species for library production was based primarily on their overt hematopoietic differences: N. coriiceps is a red-blooded species that forms all hematopoietic lineages, whereas C. aceratus is representative of the white-blooded icefishes that have lost their hemoglobin genes and capacity for erythropoiesis. However, the notothenioids possess many unusual anatomical and physiological adaptations, some of which mimic maladaptive human genetic diseases (Albertson et al., 2009). For example, many notothenioid species have evolved neutral (or near neutral) buoyancy in part through reduction of bone mineralization (De Vries and Eastman, ’78; Eastman and DeVries, ’81; Near et al., 2003, 2007), a condition that resembles human osteopenia (decreased bone mineral density) and can lead to osteoporosis. In this context, the skeleton of N. coriiceps is robustly ossified, whereas that of C. aceratus is poorly mineralized. Albertson et al. (2009) argued that identification of the genetic factors that underlie the evolved bone features of buoyant notothenioids is likely to improve our understanding of the mechanisms that regulate bone density in humans. Thus, the two notothenioid BAC libraries described herein will prove useful as resources to facilitate uncovering the genetic changes that adapt organisms to the psychrophilic mode of life and that cause some human genetic diseases.

The BAC libraries are ≥10X coverage in depth with median insert sizes >150 kb. The use of ever-improving next generation sequencing technology for obtaining high-quality contiguous sequence (Wicker et al., 2006; Dohm et al., 2007; Smith et al., 2008, 2009) should further increase the utility of these BAC resources. Our analysis of the repetitive element landscapes of N. coriiceps and of C. aceratus will contribute to a better understanding of the phylogenetic relationship of notothenioid fishes to other teleosts (cf. Shedlock et al., 2007). The BAC libraries for the two species have been archived at the BACPAC Resources Center in Children’s Hospital of Oakland Research Institute (http://bacpac.chori.org), which serves as a distribution center for high-density filters and individual clones.