Genetics: Published Articles Ahead of Print, published on February 16, 2005 as 10.1534/genetics.104.035022
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45
Revised version submitted to GENETICS December 23, 2004
A Second Generation Genetic Linkage Map of Tilapia (Oreochromis spp.) Bo-Young Lee1, Woo-Jai Lee2, J. Todd Streelman1,5, Karen L. Carleton1, Aimee E. Howe1, Gideon Hulata3, Audun Slettan2, Justin E. Stern1, Yohey Terai4 and Thomas D. Kocher1*
1
Hubbard Center for Genome Studies University of New Hampshire Durham, NH 03824 USA 2
GenoMarASA Oslo Research Park Gaustadalleen 21 Oslo, 0349 Norway 3
Department of Aquaculture Institute of Animal Science Agricultural Research Organization Volcani Center, PO Box 6 Bet Dagan 50250 Israel 4
Faculty of Bioscience and Biotechnology Tokyo Institute of Technology 4259 Nagatsuta-cho, Midori-ku, Yokohama 226-8501 Japan 5
Present address: School of Biology Georgia Institute of Technology 310 Ferst Drive Atlanta GA 30332 USA
* Corresponding author
1
46
Sequence data from this article have been deposited with the EMBL/GenBank data
47
libraries under accession nos. G68180-G68324 and BV005269-BV005594.
48
2
48
Running head: Linkage map of tilapia
49 50
Keywords: microsatellite marker, teleost, sex determination, cichlid, sex linkage
51 52
Corresponding author:
53
Thomas D. Kocher
54
Hubbard Center for Genome Studies
55
University of New Hampshire
56
Suite 400 ETB. Gregg Hall
57
35 Colovos Road, Durham, NH 03824
58
E-mail:
[email protected]
59
TEL 603-862-2115; FAX 603-862-2940
60
3
60
ABSTRACT
61 62 63
We constructed a second generation linkage map of tilapia from the F2 progeny of an
64
interspecific cross between O. niloticus and O. aureus. The map reported here contains 525
65
microsatellite and 21 gene-based markers. It spans 1311 cM in 24 linkage groups, for an average
66
marker spacing of 2.4 cM. We detected associations of sex and red color with markers on
67
linkage group 3. This map will enable mapping and selective breeding of quantitative traits
68
important to the economic culture of tilapia as a food fish, and will contribute to the study of
69
closely related cichlids which have undergone explosive adaptive radiation in the lakes of East
70
Africa.
71
4
INTRODUCTION
71 72 73
Bony fishes have been spectacularly successful over the past 200 million years.
The
74
23,000 species of teleosts make up almost half of all living vertebrates (HELFMAN et al. 1997).
75
The largest order of vertebrates (~9300 species) are the perciform fishes, which includes more
76
than 1300 species in the family Cichlidae (NELSON 1994). Cichlid fishes are native to tropical
77
areas of the Americas, Africa and Asia, and have undergone spectacular adaptive radiations in
78
the lakes of East Africa (FRYER and ILES 1972).
79
The family Cichlidae also includes the tilapias, a group of 40-50 nominal species which are
80
important food fishes in their native habitats in Africa and the Middle East. Several species of
81
tilapia have been introduced to tropical and subtropical countries around the world as a means of
82
increasing supplies of animal protein. Tilapias have become one of the most important groups of
83
foodfish, with worldwide production of about two million metric tons in 2002. About 75 percent
84
of this production is from aquaculture (FAO 1999). The most important species in aquaculture is
85
the Nile tilapia, Oreochromis niloticus. Tilapias are also an important model for studies of fish
86
physiology (WRIGHT and LAND 1998) and endocrinology (MELAMED et al. 1998; SEALE et al.
87
2002).
88
We are developing genetic and physical maps of tilapia to support the isolation of genes
89
controlling economically important traits in these species. Our previous map of the tilapia
90
genome (KOCHER et al. 1998) documented the segregation of 62 microsatellite and 112 AFLP
91
markers in 41 haploid embryos derived from a single female O. niloticus. We identified 30
92
linkage groups spanning 704cM and estimated a total map length of ~1200cM. In this report we
93
present a significantly improved map based on an F2 intercross between O. niloticus and its sister
94
species, the blue tilapia, O. aureus.
5
95 96
MATERIAL AND METHODS Mapping population: Breeding was performed by GH at the Volcani Center. We crossed
97
a red O. niloticus male derived from the University of Stirling stock (MCANDREW and
98
MAJUMDAR 1983) with a normally colored O. aureus female. Fingerlings were shipped to UNH,
99
where they were raised to sexual maturity (mean weight = 149 g). The F2 family consisted of
100 101
156 offspring, but for most loci we focused our genotyping on a subset of 70 individuals. Microsatellite markers: At GenoMar, a small insert genomic library was constructed from
102
O. niloticus DNA in pBluescriptKS vector and screened with a (CA)10 probe. A total of more
103
than 1500 clones were sequenced, and primer sets designed for 1319 putative loci. At UNH, an
104
enriched library was prepared from O. niloticus DNA using a magnetic bead-enrichment
105
protocol (CARLETON et al. 2002). A total of 240 clones were sequenced, and primer sets
106
designed for 165 putative loci using the Primer3 software (ROZEN and SKALETSKY 1998).
107
Genotyping: PCR amplification were carried out in 20 ul reaction volumes containing 1X
108
PCR buffer (10 mM Tris, 50 mM KCl, 0.1% TritonX, pH 9.0), 2.0 mM MgCl2, 200 uM of each
109
dNTP (Promega), 0.5 units Taq enzyme, 4 pmoles each of forward and reverse primer (Operon),
110
and 30-50 ng of genomic DNA. Forward primers were labeled with a fluorescent dye (6-FAM,
111
HEX, or TET). PCR amplification was performed in a Perkin-Elmer 9600 or 9700 using the
112
following conditions: one cycle of 2 min at 94°C, 28 cycles of 94°C for 30 sec, 50-55°C for 30
113
sec, 72°C for 1min and one cycle of 72°C for 7 min. PCR products for 2-3 loci were combined
114
with a mixture of ABI Genescan-500 TAMRA size standard, gel loading buffer, and deionized
115
formamide. Sizing of the PCR products was accomplished by electrophoresis in 6% or 4%
116
denaturing polyacrylamide gels on an ABI373 or ABI377 automated DNA sequencer for 8 hr or
117
2.5 hr respectively. All allele sizes were analyzed with ABI Genescan (Version 3.1) program
6
118
and genotypes saved in a FileMaker database (Microsoft Corp., Redmond WA) for further
119
analysis.
120
Type I markers: Polymorphisms in known genes were identified by one of two general
121
approaches. Some loci already deposited in GenBank contained simple-sequence motifs which
122
were polymorphic in our cross [c-ski class 1 (AJ012011); RasGRF2 (STREELMAN et al. 1998);
123
Clc 5 (AF182216); TH/Igf2 (AH006117); Insulin (AF038123); UVopsin (CARLETON et al. 2000);
124
Blue opsin (CARLETON and KOCHER 2001); Prolactin (LEE et al. 1998)]. For Bmp4, Dlx2, Dmo,
125
Dmrt1, Mhc, Otx2, Pax9, Pmch, Pomc, Tf, Trp1 and WT1_1, degenerate primers were used to
126
amplify genomic DNA or pooled cDNAs derived from tilapia embryos. Resulting amplicons
127
were cloned, sequenced and verified by BLAST. Gene-specific primers were designed and SNPs
128
distinguishing O. niloticus and O. aureus were identified. PCR-RFLP was then used to genotype
129
the F2. Partial sequences of these genes in O. aureus and O. niloticus have been deposited in
130
GenBank (Table 1).
131
Linkage analysis: The Kosambi mapping function was used for all analyses. We initially
132
used Crimap v2.4 (LANDER and GREEN 1987) hosted at the UK Human Genome Mapping Project
133
Resource Centre (http://www.hgmp.mrc.ac.uk/Bioinformatics/) for linkage analysis and map
134
construction. Later, we used JoinMap 3.0 (VAN OOIJEN and VOORRIPS 2001), performing an
135
initial clustering of markers using a LOD threshold >1.00 and a recombination limit of 4.0, and the the order of
137
adjacent triplets of markers was repeatedly tested to optimize the marker order. The results of
138
the Crimap and Joinmap analyses are largely consistent, and here we present only the latest
139
analysis performed with Joinmap.
7
140
Phenotypic traits: The sex of each F2 animal was identified by dissection and gross
141
examination of the gonads. Sex linkage was analyzed using the non-parametric Kruskal-Wallis
142
rank sum test in MapQTL 4.0 (VAN OOIJEN et al. 2002). The family also segregated for the red
143
color mutation inherited from the O. niloticus parent. Each fish was classified as either red or
144
black, and linkage analyzed using the Kruskal-Wallis statistic.
145
RESULTS
146 147 148
Marker polymorphism: Of the 165 new UNH markers tested, 145 successfully amplified genomic DNA. 108 of
149
these were informative, while 11 were not variable in this family. The other 26 markers
150
produced amplification patterns that were judged too complex to score (CARLETON et al. 2002).
151
Of the 1319 GM markers tested, at least 355 produced informative genotypes for linkage
152
analysis. An additional 82 microsatellite markers were selected from our prior maps of tilapia
153
and Lake Malawi cichlids (KOCHER et al. 1998; ALBERTSON et al. 2003).
154
Genotype data on a total of 565 markers were passed forward into linkage analysis.
155
Thirteen of these markers were discarded because they exhibited unusual patterns of segregation
156
suggestive of technical problems in scoring the polymorphism. For the final set of 552 markers,
157
genotypes were obtained for an average of 65.29 individuals per marker (range 27-140).
158
Linkage map: We identified linkages among 525 microsatellite and 20 Type I (gene)
159
markers. Only 7 markers were not linked to another marker in this analysis. The proportion of
160
linked markers is therefore 98.7%. The markers fell into 22 large and two small linkage groups
161
(Figure 1). The number of markers per linkage group (LG) ranges from 4 to 51 (Table 2).
162
Despite repeated efforts, we were not able to join the markers into 22 linkage groups
8
163
corresponding to the 22 chromosomes expected from the karyotype (MAJUMDAR and
164
MCANDREW 1986). We expect that the two small linkage groups (8 and 24) will eventually
165
coalesce with other linkage groups to resolve the discrepancy.
166
Where possible, we established a correspondence with our previous map (KOCHER et al.
167
1998), with the intention of providing a stable nomenclature for the linkage groups. Markers on
168
LG 10, 14, 18, 19 and 30 in the previous map coalesced with LG 5, 3, 2, 14 and 19 respectively
169
in the new map. Only two markers showed unexpected linkage relationships. UNH193 was
170
expected to fall in LG 1, but now maps to LG7. UNH 201 was expected to lie in LG4, but now
171
maps to LG10. We were not able to score polymorphisms for any of the markers on LG15 in the
172
previous map, so the correspondence of these markers to the new map is therefore unknown.
173
The relationships of markers in the two maps are listed in Table 3.
174
Map length: The map spans 1311 cM, with an average marker density of 2.41 cM/marker.
175
Marker density varies among chromosomes, from a high of 1cM/marker on LG24 to a low of
176
3.96 cM/marker on LG3 (Table 2). Since the genome size of these species is ~1.1 gigabase, this
177
marker spacing corresponds to an average of 840kb/cM or about 1.7Mb/marker. A comparison
178
of 14 pairwise distances among markers also present on the previous map (KOCHER et al. 1998)
179
suggests approximately 30% more recombination in the new map, which may be related to the
180
sex-specific differences in recombination discussed below.
181
Segregation distortion: Thirty-three of the 546 markers (6%) showed evidence of
182
segregation distortion at p
