Physiol Genomics 9: 1–4, 2002. First published February 19, 2002; 10.1152/physiolgenomics.00105.2001.
brief communication Microsatellite variation associated with prolactin expression and growth of salt-challenged tilapia J. T. STREELMAN AND T. D. KOCHER Hubbard Center for Genome Studies, University of New Hampshire, Durham, New Hampshire 03824 Received 6 November 2001; accepted in final form 11 February 2002
Streelman, J. T., and T. D. Kocher. Microsatellite variation associated with prolactin expression and growth of salt-challenged tilapia. Physiol Genomics 9: 1–4, 2002. First published February 19, 2002; 10.1152/physiolgenomics. 00105.2001.—Biologists have long argued that runs of alternating purines and pyrimidines could form alternative DNA structures, which might regulate transcription. Here, we report that simple sequence repeat polymorphisms in the tilapia prolactin 1 (prl 1) promoter are associated with differences in prl 1 gene expression and the growth response of salt-challenged fishes. Individuals homozygous for long microsatellite alleles express less prl 1 in fresh water but more prl 1 in half-seawater than fishes with other genotypes. Our work provides the first in vivo evidence that differences in microsatellite length among individuals may indeed affect gene expression and that variance in expression has concomitant physiological consequences. These results suggest that dinucleotide microsatellites represent an under-appreciated source of genetic variation for regulatory evolution. dinucleotide repeats; gene expression; promoter; evolution
the world’s most important aquacultural finfishes (13). A multitude of studies have addressed means to grow bigger tilapia faster, experimenting with water temperature, salinity, hybridization, and administration of growth-promoting hormones. Notably, tilapiine species differ in salt tolerance and growth response in different salinities (20). A great deal is known about the role of peptide hormones in fish osmoregulation. Prolactin is a member of the growth hormone (GH)/Prl gene family whose “freshwater adapting” role is to increase plasma osmolality by reducing gill Na⫹-K⫹-ATPase activity (15). The tilapiine pituitary produces two forms of prolactin, encoded by different genes (prl 1 and prl 2) that differ
TILAPIA ARE AMONG
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org). Address for reprint requests and other correspondence: J. T. Streelman, Hubbard Center for Genome Studies, Univ. of New Hampshire, 35 Colovos Road, Durham, New Hampshire 03824 (E-mail:
[email protected]).
in both molecular mass (24 and 20 kDa) and number of amino acid residues (188 and 177; Ref. 25). These two forms are roughly 70% identical at the amino acid level and appear to have differential osmoregulatory (1) and somatotropic (17) actions. Experiments suggest that as fishes move into saline environments, both prolactin mRNAs and serum levels decrease; this effect is more dramatic for prl 1 (2). The tilapia prl 1 promoter (21) shares noncanonical elements with that of rat. Both promoters have two (CA/GT)n microsatellites interspersed among putative binding sites for the pituitary-specific transcription factor Pit-1 (Fig. 1). Naylor and Clark (12) demonstrated in rat that these repeat sequences formed lefthanded Z-DNA in vitro and repressed prl 1 expression. We reasoned that similar regulatory effects might exist in the tilapia prl 1 promoter. Individual differences in microsatellite length might affect prl 1 expression in vivo and might contribute to known variance in salt tolerance and growth at different salinities. Despite the textbook interpretation that noncoding simple sequence repeats evolve in a neutral fashion, recent reports have indicated otherwise (11, 24). In fact, the abundance and position of dinucleotide microsatellites in eukaryotic genomes coupled with a high rate of mutation suggest a potential and pervasive role in gene regulation (9). Here, we used a natural system to evaluate the association between dinucleotide microsatellite variation, quantitative differences in gene expression and the physiological response to contrasting environments. METHODS
We initially concentrated on the microsatellite closest to the start of prl 1 transcription (⫺200 bp). We crossed females of Oreochromis mossambicus (salt-adapted) with an O. niloticus (freshwater-adapted) male carrying microsatellite alleles that differed by 17 repeat units (CA31 vs. CA14). O. mossambicus dams were homozygous for long alleles and the O. niloticus sire was a heterozygote. F1 individuals were selected for breeding such that there were three genotypes to follow in the F2 generation: homozygotes for short alleles
1094-8341/02 $5.00 Copyright © 2002 the American Physiological Society
1
2
MICROSATELLITES AND GENE EXPRESSION
Fig. 1. Schematic representation of the tilapia prl 1 promoter region. Identification of putative transcription factor binding sites (e.g., Pit 1) follows Ref. 21.
(hereafter SS), homozygotes for long alleles (LL), and heterozygotes (SL). Seventy-five F2 fishes from the same brood were raised in three salinity treatments in separate 200-liter tanks: freshwater [⬃0 parts per thousand (ppt) NaCl], 16 ppt, and seawater (32 ppt). Two-week-old fry were acclimated to treatments over a period of 3 days and maintained until fish mass averaged 15 g (⬃6 mo). Within salt treatments, all three genotypes were grown in the same tank. Fishes were killed, sexed, and weighed. Pituitaries were harvested and stored at ⫺20°C in RNAlater (Ambion) until RNA extraction. Pituitary Prl 1 gene expression was measured from F2 individuals (cDNA preparations) using real-time PCR (8) with B-actin as a reference. Primers and probes were designed using Primer Express (version 1.5, Applied Biosystems). Probe binding sites were located at intron-exon boundaries to prevent amplification of genomic DNA. Probes were fluorescently labeled at the 3⬘ end. Fluorescence was monitored during 40 cycles of PCR on a GeneAmp 5700 sequence detection system (Applied Biosystems; 95°C for 15 s, 55°C for 30 s, and 65°C for 1 min). Critical cycle number was determined when fluorescence exceeded a threshold value set close to the background. Prl 1 relative gene expression was determined according to the formula 2B-actin CT ⫺ Prl 1 CT where CT is the critical cycle
number. Real-time PCR primer and probe sequences are: F primer, cccatcaacgaactgttcga; R primer, gcatgatcaccctgcctatagg; probe, ctcacccaggagctggactctcacttccctc. F2 individuals were genotyped as described (10) or by a Sac II restriction fragment length polymorphism (RFLP) in prl 1 cDNA, which distinguishes parental alleles. Differences among genotypes per treatment and treatments per genotype were evaluated by one-way ANOVA and appropriate post hoc multiple comparison tests. RESULTS AND DISCUSSION
Females were larger than males across all salinity treatments but did not differ from males in prl 1 expression. Mean fish mass increased with increasing salinity (P ⫽ 8.4 ⫻ 10⫺06). Expression differed among genotypes at 0 ppt (P ⫽ 0.006); SS individuals expressed approximately two times and seven times as much prl 1 as SL and LL fishes, respectively (Fig. 2A; Table 1). Expression differed among genotypes at 16 ppt (P ⫽ 0.0005) with a nearly opposite pattern to that seen in freshwater; individuals of genotype LL expressed twice as much prl 1 as did fishes of genotypes
Fig. 2. Prl 1 expression and fish mass differ across salt treatments. Graphs in A–D depict means ⫾ 1SE. The y-axes of A, B, and E are normalized Prl 1 expression levels determined according to the formula 2B-actin CT ⫺ Prl 1 CT where CT is the critical cycle number. *P ⬍ 0.05 and **P ⬍ 0.005, levels of significance from Tukey’s post hoc multiple comparison tests (see Table 1 for explanation).
Physiol Genomics • VOL
9•
www.physiolgenomics.org
3
MICROSATELLITES AND GENE EXPRESSION
Table 1. Summary of data discussed in the text Salt
Genotype
Expression
Freshwater (0 ppt)
SS SL LL SS SL LL SS SL LL
14.38 ⫾ 3.56 7.96 ⫾ 1.47 2.75 ⫾ 0.38 0.66 ⫾ 0.08 0.66 ⫾ 0.07 1.37 ⫾ 0.13 0.22 ⫾ 0.11 0.15 ⫾ 0.04 0.12 ⫾ 0.04
16 ppt 32 ppt
Significance
P ⬍ 0.05 P ⬍ 0.005
Mass, g
7.75 ⫾ 1.47 10.56 ⫾ 1.42 9.94 ⫾ 1.19 16.69 ⫾ 2.3 15.26 ⫾ 1.8 8.77 ⫾ 0.22 15.0 ⫾ 2.38 16.74 ⫾ 1.04 14.32 ⫾ 1.72
Significance
P ⬍ 0.05
Values for “Expression” and “Mass” are means ⫾ 1SE, respectively. Significance values are the results of Tukey’s post hoc multiple comparison tests. For instance, in the freshwater treatment, genotype LL exhibits significantly less Prl 1 expression than the other two genotypes, which do not differ significantly from one another. SS, short alleles; LL, long alleles; SL, heterozygotes; ppt, parts per thousand.
SS and SL (Fig. 2B, Table 1). Notably, LL individuals were only one-half as big as those of other genotypes at 16 ppt (Fig. 2D, Table 1). There were no differences among genotypes in either prl 1 expression or fish mass at 32 ppt (Table 1). We detected significant genotype ⫻ environment interactions. For all genotypes, both prl 1 expression and mass differed across salt treatments (SS, P ⫽ 1.5 ⫻ 10⫺05 and P ⫽ 0.03; SL, P ⫽ 3.8 ⫻ 10⫺06 and P ⫽ 0.008; LL, P ⫽ 0.0008 and P ⫽ 0.06; for expression and mass, respectively). Individuals of genotype SS and SL showed dramatic decreases (⬎10 fold) in prl 1 expression from 0 to 16 ppt. This response was not nearly as apparent for genotype LL (Fig. 2E). Each genotype grew best at a different salinity treatment (Fig. 2F), and mean growth of genotypes was inversely correlated with mean prl 1 expression across salinities (SS, r2 ⫽ 0.957; SL, r2 ⫽ 0.971; LL, r2 ⫽ 0.530). Figure 2, E and F, demonstrates divergent norms of reaction for each genotype in response to external salt concentration. In natural populations, these “crossover interactions” are prima facie evidence for the adaptive maintenance of genetic variation in fluctuating environments (5). The possibility that runs of alternating purines and pyrimidines might affect gene expression was suggested nearly 20 years ago (6, 7, 14). We speculate that repeats of varying length can induce promoter conformations that differ in their ability to bind transcriptional regulators. There is good evidence that tilapia prl 1 expression is regulated in response to plasma osmolality (18), perhaps via interactions with Pit-1. Long prl 1 microsatellite alleles may thus act as insulators, restricting access of enhancers (freshwater) and repressors (saltwater) to Pit-1 sites. The similarity between our results and those of Naylor and Clark (12) are both unexpected and uncanny, suggesting either the conservation of a regulatory function for CA/GT microsatellites in the prl promoter over ⬃400 million years of vertebrate evolution or the convergent acquisition of this biological role. Numerous reports have confirmed that variably sized simple sequence repeats in eukaryotic promoters can elicit differential expression when cloned into vectors (7, 12, 19, 22). Shimajiri et al. (19) indicate that the matrix metalloproteinase gene is downregulated by shortening of a promoter microsatellite. Interestingly, Tae et al. (22) Physiol Genomics • VOL
9•
report that the negative effect of a dinucleotide repeat in the acetyl-CoA carboxylase gene is CAAT box dependent and can be overcome by the presence of CAAT enhancer binding protein. These studies suggest that there is no general trend in the direction of effects induced by microsatellite mutation (e.g., enhancer or repressor) and that such effects are likely to be complicated by the cellular milieu. Because F2 fishes are not isogenic lines segregating for different microsatellite alleles, we cannot formally exclude the possibility that the alleles we followed are in linkage disequilibrium with an unmeasured, causal polymorphism. However, sequence from 1.2 kb of the prl 1 promoter in SS vs. LL individuals revealed no consistent substitutions in known regulatory element binding sites. In fact, additional sequence differences in this region are found in the second simple sequence repeat 800 bp upstream of the start codon (Fig. 1). Individuals of genotype SS had shorter alleles (homozygous for GT14) than LL fishes (homozygous for GT39). Taken together, both microsatellites in the prl 1 promoter of SS fishes are less than half the size of the corresponding sequence in LL individuals, a pattern of positive covariance contrasting with data from other bony fishes (4). It is possible that the simple sequence repeats in the prl 1 promoter physically interact with one another to elicit transcriptional change. The negative correlation between prl 1 expression and fish mass is more difficult to explain. Prolactin and GH are activated by Pit-1 in distinct pituitary cell lineages (3), and these hormones have antagonistic affects on fish osmoregulation (16). Our data do not discern whether growth effects are mediated via osmoregulatory differences or transcriptional inhibition/interconversion of GH-expressing somatotropes by Prlexpressing lactotropes. From an evolutionary perspective, our results run counter to the textbook interpretation that dinucleotide microsatellite variation lacks functional consequences. These loci are rapidly evolving and ubiquitous components of eukaryotic genomes. Found preferentially in noncoding DNA (23), dinucleotide repeats provide a cellular means to alter the amount of gene product among individuals without changing amino acid sequence. The association of microsatellites with a long list of transcription factors and environmentally www.physiolgenomics.org
4
MICROSATELLITES AND GENE EXPRESSION
regulated genes suggests an under-appreciated role in the fine-scale modulation of gene expression. Research was supported by NSF/Alfred P. Sloan Foundation Grant DBI-9803946 and USDA/NRICGP Grant 00352059267 to J. T. Streelman. REFERENCES 1. Auperin B, Rentier-Delrue F, Martial JA, and Prunet P. Evidence that two tilapia prolactins have different osmoregulatory functions during adaptation to a hyperosmotic environment. J Mol Endocrinol 12: 13–24, 1994. 2. Ayson FG, Kaneko T, Hasegawa S, and Hirano T. Differential expression of two prolactin and growth hormone genes during early development of tilapia (Oreochromis mossambicus) in fresh water and seawater: implications for possible involvement in osmoregulation during early life stages. Gen Comp Endocrinol 95: 143–152, 1994. 3. Dasen JS and Rosenfeld MG. Combinatorial codes in signaling and synergy: lessons from pituitary development. Curr Opin Genet Dev 9: 566–574, 1999. 4. Dermitzakis ET, Clark AG, Batargias C, Magoulas A, and Zouros E. Negative covariance suggests mutation bias in a two-locus microsatellite system in the fish Sparus aurata. Genetics 150: 1567–1575, 1998. 5. Gillespie JH and Turelli M. Genotype-environment interactions and the maintenance of polygenic variation. Genetics 121: 129–138, 1989. 6. Hamada H, Petrino MG, and Kakunaga T. A novel repeated element with Z-DNA-forming potential is widely found in evolutionary diverse eukaryotic genomes. Proc Natl Acad Sci USA 79: 6465–6469, 1982. 7. Hamada H, Seidman M, Howard BH, and Gorman CM. Enhanced gene expression by the poly(dT-dG)-poly(dC-dA) sequence. Mol Cell Biol 4: 2622–2630, 1984. 8. Heid CA, Stevens J, Livak KJ, and Williams PM. Real time quantitative PCR. Genome Res 6: 986–994, 1996. 9. Kashi Y, King D, and Soller M. Simple sequence repeats as a source of quantitative genetic variation. Trends Genet 13: 74–78, 1997. 10. Lee WJ and Kocher TD. Microsatellite mapping of the prolactin locus in the tilapia genome. Anim Genet 29: 68–69, 1998. 11. Majewski J and Ott J. GT repeats are associated with recombination on human chromosome 22. Genome Res 10: 1108–1114, 2000. 12. Naylor LH and Clark EM. d(TG)n-d(CA)n sequences upstream of the rat prolactin gene form Z-DNA and inhibit gene transcription. Nucleic Acids Res 18: 1595–1601, 1990.
Physiol Genomics • VOL
9•
13. Naylor RL, Goldburg RJ, Primavera JH, Kautsky N, Beveridge MC, Clay J, Folke C, Lubchenco J, Mooney H, and Troell M. Effects of aquaculture on world fish supplies. Nature 405: 1017–1024, 2001. 14. Nordheim A and Rich A. The sequence (dC-dA)n-(dG-dT)n forms left-handed Z-DNA in negatively supercoiled plasmids. Proc Natl Acad Sci USA 80: 1821–1825, 1983. 15. Sakamoto T, Shepherd BS, Madsen SS, Nishioka RS, Siharath K, Richman NH III, Bern HA, and Grau EG. Osmoregulatory actions of growth hormone and prolactin in an advanced teleost. Gen Comp Endocrinol 106: 95–101, 1997. 16. Seidelen M and Madsen SS. Prolactin antagonizes the seawater adaptive effect of cortisol and growth hormone in anadromous brown trout. Zoo Sci 14: 249–256, 1997. 17. Shepherd BS, Sakamoto T, Nishioka RS, Richman NH III, Mori I, Madsen SS, Chen TT, Hirano T, Bern HA, and Grau EG. Somatotropic actions of the homologous growth hormone and prolactins in the euryhaline teleost, the tilapia Oreochromis mossambicus. Proc Natl Acad Sci USA 94: 2067–2072, 1997. 18. Shepherd BS, Sakamoto T, Hyodo S, Nishioka RS, Ball C, Bern HA, and Grau EG. Is the primitive regulation of pituitary prolactin (tPRL177 and tPRL188) secretion and gene expression in the euryhaline tilapia (Oreochromis mossambicus) hypothalamic or environmental? J Endocrinol 161: 121–129, 1999. 19. Shimajiri S, Nobuyuki A, Tanimoto A, Murata Y, Hamada T, Wang KY, and Sasaguri Y. Shortened microsatellite d(CA)21 sequence down-regulates promoter activity of matrix metalloproteinase 9 gene. FEBS Lett 455: 70–74, 1999. 20. Suresh AV and Lin CK. Tilapia culture in saline waters: a review. Prog Fish Cult 448: 161–167, 1992. 21. Swennen D, Poncelet AC, Sekkali B, Rentier-Delrue F, Martial JA, and Belayew A. Structure of the tilapia (Oreochromis mossambicus) prolactin I gene. DNA Cell Biol 11: 673–684, 1992. 22. Tae HJ, Luo X, and Kim KH. Roles of CCAAT/enhancer binding protein and its binding sites on repression and derepression of acetyl-CoA carboxylase gene. J Biol Chem 269: 10475– 10484, 1994. 23. To´th G, Ga´spa`ri Z, and Jurka J. Microsatellites in different eukaryotic genomes: survey and analysis. Genome Res 10: 967– 981, 2000. 24. Turpeinen T, Tenhola T, Manninen O, Nevo E, and Nissila¨ E. Microsatellite diversity associated with ecological factors in Hordeum spontaneum populations in Israel. Mol Ecol 10: 1577– 1591, 2001. 25. Yamaguchi K, Specker JL, King DS, Yokoo Y, and Nishioka RS. Complete amino acid sequence of a pair of fish (tilapia) prolactins, tPRL177 and tPRL188. J Biol Chem 263: 9113–9121, 1988.
www.physiolgenomics.org
