- Methodology article
- Open Access
Biologically meaningful expression profiling across species using heterologous hybridization to a cDNA microarray
© Renn et al; licensee BioMed Central Ltd. 2004
- Received: 16 June 2004
- Accepted: 06 July 2004
- Published: 06 July 2004
Unravelling the path from genotype to phenotype, as it is influenced by an organism's environment, is one of the central goals in biology. Gene expression profiling by means of microarrays has become very prominent in this endeavour, although resources exist only for relatively few model systems. As genomics has matured into a comparative research program, expression profiling now also provides a powerful tool for non-traditional model systems to elucidate the molecular basis of complex traits.
Here we present a microarray constructed with ~4500 features, derived from a brain-specific cDNA library for the African cichlid fish Astatotilapia burtoni (Perciformes). Heterologous hybridization, targeting RNA to an array constructed for a different species, is used for eight different fish species. We quantified the concordance in gene expression profiles across these species (number of genes and fold-changes). Although most robust when target RNA is derived from closely related species (<10 MA divergence time), our results showed consistent profiles for other closely related taxa (~65 MA divergence time) and, to a lesser extent, even very distantly related species (>200 MA divergence time).
This strategy overcomes some of the restrictions imposed on model systems that are of importance for evolutionary and ecological studies, but for which only limited sequence information is available. Our work validates the use of expression profiling for functional genomics within a comparative framework and provides a foundation for the molecular and cellular analysis of complex traits in a wide range of organisms.
- Nile Tilapia
- Phylogenetic Distance
- cDNA Array
- Cichlid Fish
- Cichlid Species
The expression activities of all the genes represented in an organism's genome at any given time constitute a complex phenotype that is closely connected with, but not solely dependent upon, the genotype. In fact, gene expression profiles represent the primary level of integration between environmental factors and the genome, providing the basis for protein synthesis which ultimately guides the implementation of complex macro-phenotypes such as morphology and behaviour. Therefore, by comparing gene expression profiles of different strains, populations, or even species, one can directly study the molecular basis of phenotypic variation. This comparative approach has recently been employed in expression profiling of model organisms such as Drosophila and S. cerevisiae, unveiling surprising patterns of sex-driven evolution and providing insights into the genetics of population-level variation in adaptive traits ([1–5]). Similarly, across-species comparisons of expression profiles between human and non-human primates have been used to test hypotheses about the functional complexity of the human brain ([6–8]).
Although it has become increasingly clear that genomics research benefits from such a comparative outlook, whole genome sequences are available for only a few of the less traditional model systems (e.g. Fugu (; Anopheles ; dog ; honey bee ). Given the current DNA microarray technologies (long or short oligonucleotide arrays and cDNA arrays), which allow simultaneous monitoring of thousands of genes, the PCR amplified cDNA array is the most accessible for studies in non-model organisms (but see  for oligonucleotide strategies). The construction of a cDNA microarray is not limited by the need for probe design based on advanced bioinformatics analysis of genome sequences only available for genetic "model organisms" (e.g., [12–14]). Due to the length of the probes (which are often full-length), it has been suggested that cDNA microarrays can also be used in heterologous hybridizations across strains and even relatively closely related species as long as sequence divergence is limited for a given gene . This approach has been used successfully, if only occasionally, to study the molecular basis of traits not present in traditional model species (e.g. hibernation, ). Heterologous hybridization offers a promising approach to study molecular mechanisms in species for which a wealth of ecological data as well as natural phenotypic and genetic variation are already available. As research areas such as ecophysiology ([17–19]) as well as ecology and evolution ([20–22]) are now merging with functional genomics (; reviewed in ) the technique of heterologous hybridization will become more prominent. While the feasibility of this technique has been indirectly suggested,  its potential has not been systematically tested using a biologically meaningful approach over a wide range of species.
In this paper we quantify the utility of a microarray constructed from an African cichlid fish cDNA library to study expression profiling in other fish species through the use of heterologous hybridization. We not only test RNA hybridization from closely related cichlids but also from more distantly related species of teleost fishes, a group that represents more than half of living vertebrates . Our results demonstrate that heterologous microarray hybridization can yield biologically meaningful data even in relatively distantly related species and establish this technique as a tool for comparative functional genomics in organisms not previously open to an integrative molecular analysis.
The rapid, recent, and repeated radiation of the cichlid fishes in East Africa's Great Lakes has produced a system especially amenable to this approach. Each of the Great Lakes -Victoria, Tanganyika, and Malawi- boasts hundreds of phenotypically diverse endemic cichlid species, evolved from one or a few common ancestors in very recent evolutionary history ([26–30]) and thus sharing highly similar genomic sequence. The astonishing variation in phenotype at the morphological and behavioural level ([31–38]), both within and among cichlid species has likely contributed to the evolutionary success of this group ([39, 40]). Therefore, through heterologous hybridization, a genomic approach can be used to uncover the molecular basis and the evolution of these complex traits.
Characterization of the cichlid fish cDNA array
As our study presents the first use of this cDNA microarray to determine gene expression in the brain of multiple cichlids species, we first characterized this tool. This array was originally built using anonymous clones from a brain cDNA library for the African cichlid Astatotilapia burtoni. Sequencing is being done in parallel with initial experiments. The gel electrophoretic analysis of PCR products performed during array construction identified reactions that either had failed to produce a product or had produced more than one product. From this we concluded that 4570 amplified cDNAs were reliably represented on the array. For these, the initial 5' sequencing returned sequence for eighty-six percent (3936) of the clones (Genbank CN468542 – CN472211 for clones > 150 bp). Sixty-four percent (2462) of the EST's represented unique sequences ("singletons") and the remaining 37% could be aligned to 575 EST clusters, predicting that up to 3000 unique genes may be represented on this array. Approximately 40% of the different genes showed significant homology to proteins predicted from the Fugu genome (, Salzburger et al., in prep.)
Self-hybridization experiments, in which two samples of the platform species A. burtoni genomic DNA were competitively hybridized, revealed that 93% (mean = 4264 + SD 229; n = 6) of the features on the array could be hybridized at a fluorescent intensity at least two standard deviations above mean background intensity. Signal intensities for the 635 nm and 532 nm channels were highly correlated in these experiments (r = 0.991 same DNA isolation, r = 0.941 same animal but different DNA isolation, r = 0.974 two different animals, all p < 0.0001) indicating that technical variability – due to differences in DNA source, isolation, and fluorescent labelling – was very low. These results suggest that the majority of the A. burtoni library spots will provide reliable data.
Self-hybridization experiments, where two samples of RNA from a pool of several A. burtoni brains were competitively hybridized, revealed that ~94% of the features on the array (4316+SD 431; n = 10) represent genes that are expressed at detectable levels using a whole brain strategy. Again, signal intensities for the 635 nm and 532 nm channels were highly correlated in these experiments (r = 0.974 – 0.997 all p < 0.0001) indicating that technical variability due to Cy3 or Cy5 labelling is very low.
Establishing a reference profile in A. burtoni
In order to evaluate the sensitivity and consistency of heterologous hybridization for several species on this array platform, we first devised a direct comparison experiment to maximize expression differences within A. burtoni, the species from which the array was built. These results provide a reference for the evaluation of the heterologous hybridization results presented below. As a large percentage of a genome's protein-coding genes are expressed in the brain ([42, 43]), we expected this microarray to be useful for expression profiling in other tissues. Therefore, brain-derived RNA was competitively hybridized against mixed-tissue RNA (skin, muscle, blood vessels) extracted from the same individual. The distinct nature of these tissues predicts a dramatic difference in the expression of a large number of genes. Four replicates from each of two individual A. burtoni (one adult and one juvenile) were used to identify consistent gene expression differences between the two tissues, independent of age and reproductive status. 88% of the A. burtoni microarray probes hybridized above the background cut-off (4165+SD 421, n = 8).
The inclusion of replicate spots in microarray construction provides a standard for internal control, and increased confidence in the estimation of differential regulation for these spots. Based on the sequence information (available for 656 of the 804 spots) we estimated that this reference set contains a total of 472 genes. 41 of these genes are represented by three or more cDNAs (173 spots total). We find greater than 75% concordant regulation across representative spots (at least three out of four clones or all three for those clusters which contain only three clones) for 27 of these genes. Concordance smaller than 75% could be caused by (i) hybridization failure (due to low probe concentration), (ii) improper assignment of clones to genes, (iii) sequencing errors, (iv) alternative splicing, (v) chance (false negatives).
We used the same brain vs. body intra-species experimental design to test the hypothesis that hybridization of heterologous RNA to the array can yield biologically meaningful results. We chose two other cichlid species endemic to Lake Tanganyika, Enantiopus melanogenys and Neolamprologus brichardi, and a more distantly related African cichlid, the Nile tilapia, Oreochromis niloticus. Based on their wide use in behavioural, evolutionary, developmental, and genetic studies, we also selected four more distantly related species that are not part of the order perciform . Two poeciliid fish: platyfish (Xiphophorus sp. ) and guppy (Poecilia reticulata ; 65 MY divergence time ); one salmonid: Atlantic salmon (Salmo salar ), and one cyprinid: zebrafish (Danio rerio ; 200 MY divergence time ). We first quantified the extent of hybridization for a given species with a simple threshold cut-off, and then compared these heterologous expression profile results to the reference set of the 804 brain-enriched genes from A. burtoni in order to quantify the performance of heterologous hybridization for detection of hybridization, regulation, and biologically meaningful results.
Detection of hybridization
Correlation of expression profiles across species
In poeciliids, regression coefficients were r2 = 0.45 (Xiphophorus sp.) and r2 = 0.37 (P. reticulata). However, in the more distantly related taxa (salmonids and cyprinids) regression coefficients were r2 = 0.21 and r2 = 0.11, demonstrating that the A. burtoni data explained less of the fold-change variation. This reflects the fact that more spots showed different expression in these distantly related species compared to A. burtoni. The drop in the regression slope with phylogenetic distance (Figure 3) suggests that although gene expression in A. burtoni and the other species were in the same direction, the magnitude of change in expression that was detected was much lower in these species. The effect of phylogenetic distance on both our ability to detect subtle gene regulation and its magnitude shows that the sensitivity of our array is very good for cichlids and even species that are not members of the perciformes (e.g., platyfish and guppy), but is lower for distantly related species such as salmon and zebra fish.
Detection of regulated gene expression
Many genes are regulated in each species. Even in more distantly related fish species, the spots determined to be up-regulated in the brain encompass a great number of unique spots, and are not simply a few genes represented multiple times on the array (n refers to the number of array hybridizations). The estimated number unique genes up-regulated in brain is calculated independently for each species based on the sum of singleton ESTs and the number of genes predicted by clustered ESTs relative to the sequence information available. This result indicates that the redundancy rate for spots determined to be up-regulated in the brain is in agreement with the overall array redundancy and that differences in expression can be detected for mRNAs of low copy number.
Total number of spots
No sequence available
Number of ESTs
Number of clusters
Estimated number of unique genes
A. burtoni (adult) (n = 4)
A. burtoni (juvenile) (n = 4)
N. brichardi (n = 3)
E. melanogenys (n = 3)
O. niloticus (n = 2)
Xiphophorus sp. (n = 2)
P. reticulata (n = 2)
S. salar (n = 2)
D. rerio (n = 2)
Detection of consistent gene regulation
Detection of subtle gene regulation
In this paper we describe the first systematic analysis of heterologous microarray hybridizations across a range of vertebrate species. This work validates the use of expression profiling for functional genomics within a comparative framework and provides a foundation for the molecular and cellular analysis of complex traits at the organismal, population, and ecological levels . We clearly show the utility of the array for heterologous hybridization across a range of fish species for which there is no other microarray available. We can detect array features (though reduced in number) that hybridize above background as well as spots that show tissue-specific regulation, many of which correspond to those regulated in A. burtoni.
The variation in brain-specific gene expression between individual fish of different Tanganyikan cichlid species is comparable to the variation observed between adult and juvenile individuals of A. burtoni. The slight decrease in our ability to identify the A. burtoni genes of subtle regulation (i.e., low fold-change classes) in other cichlid species may be due to either the smaller number of replicates performed for these species or the increased individual variation in these fold-change classes. Alternatively, this result could also reflect real species-specific differences in gene expression. Even in distantly related species 26% – 53% of the significantly up-regulated A. burtoni genes could still be re-identified. In this analysis we assume that the increasing number of genes that failed to hybridize with increased phylogenetic distance was due mainly to sequence divergence. This assumption provides a conservative guideline regarding the utility of heterologous hybridization. Tissue-specific gene regulation is obviously not expected to be identical in all species. Therefore, it is possible that more than 26% – 53% of the array spots are informative for distantly related species. Heterologous hybridization experiments on any microarray are of limited use for genes that have undergone rapid evolutionary change in coding regions, large rearrangements, and duplication (e.g., functional divergence of paralogous genes). Our regression analysis across species demonstrates that gene regulation is robust and identifiable, although its magnitude decreases with phylogenetic distance. Our results suggest that with sound statistical analysis and additional replicates ([49–51]) even subtly regulated genes can be identified in the distantly related species. Given our results using species that have diverged more than 65 million years ago (guppies and platyfish), it is clear that this array will perform splendidly in the > 12,000 species within the large order Perciformes, to which cichlids belong (e.g., gouramis, mackerels, blennies, wrasses, bass, sunfish, perch, gobies, and damselfish).
Future detailed studies focusing on multiple species will benefit from inter-species genomic DNA hybridizations in order to determine spots that are most affected by sequence divergence . Such experiments will differentiate between genes whose regulation is different (genomic-hybridization ratios equal to 1) and genes whose sequence has diverged considerably (genomic-hybridization ratios significantly different from 1). We explored this strategy by competitively hybridizing to the array A. burtoni genomic DNA against genomic DNA from either the Nile tilapia (ca. 10 million years divergence time) or the zebrafish (more than 200 million years of divergence). As we had previously determined which of the 804 reference spots were significantly regulated in either of these two species (see Figure 4), we divided the genomic hybridization results for the reference spots into two classes depending on whether they were also brain-enriched in the other species or not. Interestingly, the mean ratios of these two classes were not different in the O. niloticus/A. burtoni genomic DNA hybridization (Student t-test: t = -1.6, p = 0.1). However, when genomic DNA from A. burtoni and the distantly related D. rerio was competitively hybridized to the array, we not only found many spots that hybridized preferentially with A. burtoni genomic DNA; we also found a significant difference for the mean hybridization ratios (t = -9.4, p < 0.001) between the two reference spot classes (i.e., those spots that did and those that did not show significant brain-specific regulation in D. rerio). These results suggest that the difference in gene regulation observed between A. burtoni and the Nile tilapia may be due to real functional differences while the small number of re-identified reference spots observed in zebrafish may be largely due to sequence divergence. Sequence divergence hinders accurate hybridization at these spots during heterologous hybridization experiments, indicating that these spots cannot be used for functional analysis within this species. In conclusion, genomic DNA hybridization experiments can be used to estimate the false negative rate for a within-species RNA experiment and may be essential for distinguishing between variation due to sequence divergence and variation due to transcript abundance in across-species RNA experiments. Two general rules can be derived from this analysis: First, identify the phylogenetically closest existing array platform; second, before initiating an extensive expression profiling experiment utilizing heterologous hybridization to any array, conduct a statistical analysis of genomic hybridization results. These steps will maximize the number of useful spots and assure the disqualification of those spots whose DNA hybridization ratios are significantly different from 1.
The great number of ecological, evolutionary, aquaculture and conservation studies in widely divergent fish species will be greatly enhanced by the development of genomic resources. Because natural variation is fundamentally polygenic and arises from complex interactions within the genome as well as with the environment, a multiple-gene approach to the study of phenotypic regulation will provide new insights. The combination of diverse ecological characteristics in African cichlid fishes and their ability to reproduce a full behavioural repertoire in captivity provides a powerful framework for studies both in the field and in the laboratory. Their astonishing phenotypic diversity, despite minimal genetic divergence, the result of a uniquely rapid and recent radiation (e.g. [52–54]), allows us to utilize a single cichlid microarray to study the more than 2000 different East African cichlid species. We foresee the utility of this array for examining natural variation of gene expression as it relates to phenotypic plasticity, adaptation, and speciation, and population studies central to organismal and evolutionary biology. Both within and across species this microarray can be used to study the molecular basis of species-specific characters such as jaw morphology, male colour patterns, brain anatomy, reproduction, and behaviour, as well as the mechanisms underlying phenotypic plasticity, which may contribute to the rapid rate of speciation (reviewed in ).
While the cichlid fish cDNA microarray will greatly facilitate the comparative functional genomic approach for an important group of fishes, we expect that the results of our systematic heterologous hybridization studies presented here will encourage researchers in many fields to utilize existing cDNA arrays for diverse groups of teleosts and other taxa.
We have constructed a cDNA microarray with ~4500 features from a brain-specific cDNA library for the African cichlid fish Astatotilapia burtoni. We describe the first quantitative functional analysis of heterologous hybridization across a range of vertebrate species to a single cDNA microarray platform. We validate a genomic strategy that overcomes some of the restrictions imposed by systems for which only limited sequence information is available. Although most robust when sample RNA is derived from closely related cichlids, expression profiling results showed consistent hybridization for other closely related taxa (~65 million years divergence) and, to a lesser extent, even very distantly related species. This work represents a first step toward bringing genomics to bear in cichlids and other non-traditional model systems. Crucially, we demonstrate the feasibility of functional genomic studies in a comparative context for any organism.
Part I: Construction of a custom-made cDNA array
Library construction – A full-length, directional (EcoRI – XhoI) cDNA library was constructed in Lambda ZapII phage vector (Stratagene) with mRNA from A. burtoni brains (both sexes at all stages of development and social condition were included) and was generously provided by U. DeMarco and R. Fernald (Stanford University). The pBluescript phagemid, pBSIIsk, was excised from the Lambda ZAP vector, following protocol for transformation into XL1-Blue MRF' (Stratagene) E. coli strain for plating and picking.
Plating, selection, and amplification of bacterial colonies – Cells were plated on LB agar supplemented with ampicillin in 20 cm Q-bot trays (Genetix). 5755 Bacterial colonies were selected by the Q-bot (Genetix) and inoculated into 96-well plates with 150 μl LB+amp glycerol for overnight growth at 37°C in a humid incubator.
Replicated plates (without glycerol), produced a working set of 58 plates for PCR amplification. Plasmid inserts were amplified by colony PCR in Microseal 96-well plates (MJ Research) on MJ Tetrads (MJ Research) using custom vector primers for pBSIIsk- (CSVP2: TTCCCAGTCACGACGTTGTAAAA, 23mer, Tm = 60.9°C; CSVP3: AAGCGCGCAATTAACCCTCACTA, 23mer, Tm = 62.7°C). Reaction conditions were as follows: 1 × Taq Buffer + 2 mM MgCl2; 0.25 mM dNTP mix; 0,18 μM each primer; 1.5 units FastStart Taq (Roche). Samples were denatured for 5 min at 95°C followed by the 35 cycles of 95°C for 45 sec, 60°C for 20 sec, 72°C for 3 min. Samples were then held at 72°C for 5 min and stored at 4°C. PCR products were visualized on 1% agarose gels and scored for strong, single product (4570 passes = 79.4%). The plates were purified by vacuum filtration to remove excess dNTPs and primers using the MultiScreen-PCR 96-well Filtration System (Millipore); re-suspended in MilliQ-grade water to an average estimated concentration of 100–200 ng/μl; transferred to Costar 96-well V bottom polypropylene storage plates (Corning); and dehydrated for storage. After all inserts had been amplified, the products were re-suspended in nuclease-free de-ionized water and compressed into a 384-well plate format without reconfiguration using a BioMek FX liquid handling robot (Beckman Instruments) and sterile barrier tips (Beckman-Coulter). The plates were dehydrated for storage and re-hydrated in 10 μl of 3 × SSC for array printing.
Array production – All A. burtoni cDNA clones (including 1185 that failed the gel analysis above) and 120 control clones were spotted in duplicate arrays onto NaOH cleaned, poly-lysine (Sigma) coated slides using the 16-pin format on an OmniGrid-100 arrayer (GeneMachines). Yeast, Arabidopsis, mouse, and human clones were included as negative controls.
Slides were re-hydrated and UV cross-linked with 6000 mJ (Stratalinker). Slides were blocked with succinate anhydride, 1-Methyl-2 polypyrolidinone and sodium borate, then denatured in boiling water and spun dry according to standard protocol . Hydrated and blocked arrays were stored in light-proof containers in a desiccator until hybridization.
Part II: Characterization of cDNA array protocol and cross-species hybridizations
Fish species used – Male A. burtoni, Enantiopus melanogenys and Neolamprologus brichardi were randomly selected from a lab-reared stock. The Tilapia (O. niloticus) was obtained from aquaculture supplier. The other non-cichlid species were obtained from a local supplier (Xiphophorus sp., Poecilia reticulata) from the Harvard University zebrafish facility (Danio rerio) and the S.O. Conte Anadromous Fish Research Center (Atlantic salmon, S. salar).
Fish were killed with 0.03% tricaine methanesulfonate (Sigma) in accordance with the animal protocol (#22-22) approved by the Harvard University Institutional Animal Care & Use Committee, and brains and a mixture of "body tissues" containing muscle, skin, and bone, were dissected out immediately. The samples were minced in 1 ml of RNAlater solution (Ambion) and stored in 4° overnight followed by long term storage at -20°C.
DNA extraction- Genomic DNA was isolated from mixed tissue. Approximately 100 mg of tissue was homogenized and digested in buffer solution (60 mM Tris, pH 8.0, 100 mM EDTA, 0.5% SDS) containing proteinase K (0.5 mg/ml) at 37°C for 12 to 16 hours followed by phenol:chloroform:iso-amyl alcohol extraction (25:24:1) using the Phase Lock Gel light system (Eppendorf) for phase separation. Yield and quality was evaluated by gel analysis and standard spectrophotometric measurements.
DNA labelling- For each DNA probe 2 μg of genomic DNA was restriction-digested with Sau3aI (New England Biolabs) and labelled according to a standard Klenow protocol (Invitrogen) with direct incorporation of Cy3 or Cy5-dCTP (Amersham). Labelled DNA was purified and concentrated on a YM30 Amicon (Millipore) filter, salts were adjusted to 3XSSC and 1.5 % SDS. The denatured probe was applied beneath a lifter cover slip (Erie Scientific Corp.) and hybridized overnight in the dark at 65°C in a humidified chamber (Telechem) submerged in a water bath. Excess probe was removed by rinsing in 2 × SSC 0.01 % SDS at 65°C followed by two rinses at room temperature (1 × SSC and 0.2 × SSC) and centrifuged to dry.
RNA extraction – Total RNA was extracted from brains and mixed tissue of males according to a standard Trizol protocol (Invitrogen), following tissue homogenization (Tissue Tearor, Biospec Products). The RNA was analyzed for quantity and quality on the Bioanalyzer (Agilent) and a standard spectrophotometer (Agilent).
RNA labelling – Two μg of total RNA was labelled for each sample ( by first annealing primer in a 15 μl reaction with 1 μl of primer solution (5 μg/μl each poly dT 12–18 or 5 μg/μl each poly dT 12–18 with 5 μg/μl random hexamer oligonucleotides). Reverse transcription reactions were prepared on ice: 5 μl 5 × 1st strand buffer (Invitrogen); 2 μl 0.1 M DTT; 0.6 μl 50 × amino-allyl-dUTP/dNTP mix (2.5 mM each dATP, dCTP, dGTP, 1.5 mM dTTP (Invitrogen) and 10 mM amino-allyl dUTP (Sigma)); and 2 μl (200 U/μl) SuperScript II (Invitrogen), and then incubated at 42°C for 2 hours. RNA was hydrolyzed, and the reaction was stopped by adding 10 μl of 1 N NAOH and 10 μl of 0.5 M EDTA and placed at 65°C for 7 min. The reaction was neutralized with 25 μl of 1 M HEPES pH 7.5 (GIBCO BRL). The cDNA was then rinsed and concentrated on a YM-30 filter (Millipore). The dye-coupling reaction required adding 1.5 μl of 1 M sodium bicarbonate pH 9.0 and the appropriate Cy3 or Cy5 CyDye Post-labeling reactive Dye Pack (Amersham) and placing it for 1 hour at room temperature in the dark. The labelled cDNA was then purified using a Qiagen PCR column, pooled with the appropriate sample for competitive hybridization and concentrated to 50 μl on a YM 30 filter. The appropriate hybridization buffer conditions were achieved by adding 6 μl 20 × SSC (Gibco), 3 μl poly (dA) poly(dT) (Sigma) and 0.96 μl 1 M HEPES and 0.9 μl 10% SDS to each combined labelled probe. Hybridizations and post-hybridization processing were performed as in the DNA hybridization procedure (see above). Note that Cy3 and Cy5 dyes were "swapped" between tissues when technical replicates were performed, such that brain RNA was labelled at least once in "green" (Cy3) and once in "red" (Cy5) in a given species to avoid gene-by-dye effects .
Analysis – Hybridized arrays were scanned with an Axon 4000B scanner (Axon Instruments) using Genepix 4.0 software (Axon Instruments) for initial spot finding. The data sets were filtered for spots flagged as "bad" because of irregularities on surface of array (dust, speckle, scratch). Intensity values of spots showing hybridization intensity two standard deviations above background intensity in both channels were used for spot counting and correlation analysis on technical replicates of A. burtoni genomic DNA.
Raw data from Genepix was imported into R and analyzed using the LIMMA library (Linear Models for Microarray Data,) for within-array print-tip loess normalization of intensities, identification of statistically significant regulation (moderated t-statistics using empirical Bayes shrinkage of the standard errors), and calculation of average fold-changes. Background subtracted intensities from unflagged spots were used in normalization and model fitting. The normalized and fitted data of intensities, number of significantly regulated spots and fold change were used for all remaining intra- and inter-species analysis.
The raw and analyzed data for the 24 microarray experiments used in this study have been submitted to Gene Expression Omnibus (SERIES ID = GSE975, available online ). The ESTs representing the cDNAs on the microarray have been submitted to NCBI GenBank.
All correlations analyses were performed using Pearson correlation coefficient tests. Linear regression analyses were used to estimate the amount of variation in fold change observed in a heterologous hybridization that could be explained by the fold change observed in A. burtoni and estimate the slope of the relationship between these two variables.
We are grateful to Uli DeMarco and Russell Fernald (Stanford University) for providing the cDNA library. We thank Sarah Annis, Claire Bailey, Christian Daly and Keith Morneau for expert technical assistance with the array construction and hybridization protocol, Christian R. Landry for bioinformatics support as well as Josiah Altschuler for animal husbandry. Colin Meiklejohn, Yuk Fai Leung, Ping Ma and Jon Wilkins provided valuable comments on earlier versions of the manuscript. This work was supported by a NIH National Research Service Award to SCPR, a FQRNT (Fonds québécois de la recherche sur la nature et les technologies) postdoctoral fellowship to NAH, and by the Bauer Center for Genomics Research.
- Jin W, Riley RM, Wolfinger RD, White KP, Passador-Gurgel G, Gibson G: The contributions of sex, genotype and age to transcriptional variance in Drosophila melanogaster. Nat Genet. 2001, 29: 389-395. 10.1038/ng766.View ArticlePubMedGoogle Scholar
- Ranz JM, Castillo-Davis CI, Meiklejohn CD, Hartl DL: Sex-dependent gene expression and evolution of the Drosophila transcriptome. Science. 2003, 300: 1742-1745. 10.1126/science.1085881.View ArticlePubMedGoogle Scholar
- Meiklejohn CD, Parsch J, Ranz JM, Hartl DL: Rapid evolution of male-biased gene expression in Drosophila. Proc Natl Acad Sci U S A. 2003, 100: 9894-9899. 10.1073/pnas.1630690100.PubMed CentralView ArticlePubMedGoogle Scholar
- Bochdanovits Z, van der Klis H, de Jong G: Covariation of larval gene expression and adult body size in natural populations of Drosophila melanogaster. Mol Biol Evol. 2003, 20: 1760-1766. 10.1093/molbev/msg179.View ArticlePubMedGoogle Scholar
- Townsend JP, Cavalieri D, Hartl DL: Population genetic variation in genome-wide gene expression. Mol Biol Evol. 2003, 20: 955-963. 10.1093/molbev/msg106.View ArticlePubMedGoogle Scholar
- Enard W, Khaitovich P, Klose J, Zollner S, Heissig F, Giavalisco P, Nieselt-Struwe K, Muchmore E, Varki A, Ravid R, Doxiadis GM, Bontrop RE, Paabo S: Intra- and interspecific variation in primate gene expression patterns. Science. 2002, 296: 340-343. 10.1126/science.1068996.View ArticlePubMedGoogle Scholar
- Gu J, Gu X: Induced gene expression in human brain after the split from chimpanzee. Trends Genet. 2003, 19: 63-65. 10.1016/S0168-9525(02)00040-9.View ArticlePubMedGoogle Scholar
- Nahon JL: Birth of 'human-specific' genes during primate evolution. Genetica. 2003, 118: 193-208. 10.1023/A:1024157714736.View ArticlePubMedGoogle Scholar
- Aparicio S, Chapman J, Stupka E, Putnam N, Chia JM, Dehal P, Christoffels A, Rash S, Hoon S, Smit A, Gelpke MD, Roach J, Oh T, Ho IY, Wong M, Detter C, Verhoef F, Predki P, Tay A, Lucas S, Richardson P, Smith SF, Clark MS, Edwards YJ, Doggett N, Zharkikh A, Tavtigian SV, Pruss D, Barnstead M, Evans C, Baden H, Powell J, Glusman G, Rowen L, Hood L, Tan YH, Elgar G, Hawkins T, Venkatesh B, Rokhsar D, Brenner S: Whole-genome shotgun assembly and analysis of the genome of Fugu rubripes. Science. 2002, 297: 1301-1310. 10.1126/science.1072104.View ArticlePubMedGoogle Scholar
- Holt RA, Subramanian GM, Halpern A, Sutton GG, Charlab R, Nusskern DR, Wincker P, Clark AG, Ribeiro JM, Wides R, Salzberg SL, Loftus B, Yandell M, Majoros WH, Rusch DB, Lai Z, Kraft CL, Abril JF, Anthouard V, Arensburger P, Atkinson PW, Baden H, de Berardinis V, Baldwin D, Benes V, Biedler J, Blass C, Bolanos R, Boscus D, Barnstead M, Cai S, Center A, Chaturverdi K, Christophides GK, Chrystal MA, Clamp M, Cravchik A, Curwen V, Dana A, Delcher A, Dew I, Evans CA, Flanigan M, Grundschober-Freimoser A, Friedli L, Gu Z, Guan P, Guigo R, Hillenmeyer ME, Hladun SL, Hogan JR, Hong YS, Hoover J, Jaillon O, Ke Z, Kodira C, Kokoza E, Koutsos A, Letunic I, Levitsky A, Liang Y, Lin JJ, Lobo NF, Lopez JR, Malek JA, McIntosh TC, Meister S, Miller J, Mobarry C, Mongin E, Murphy SD, O'Brochta DA, Pfannkoch C, Qi R, Regier MA, Remington K, Shao H, Sharakhova MV, Sitter CD, Shetty J, Smith TJ, Strong R, Sun J, Thomasova D, Ton LQ, Topalis P, Tu Z, Unger MF, Walenz B, Wang A, Wang J, Wang M, Wang X, Woodford KJ, Wortman JR, Wu M, Yao A, Zdobnov EM, Zhang H, Zhao Q, Zhao S, Zhu SC, Zhimulev I, Coluzzi M, della Torre A, Roth CW, Louis C, Kalush F, Mural RJ, Myers EW, Adams MD, Smith HO, Broder S, Gardner MJ, Fraser CM, Birney E, Bork P, Brey PT, Venter JC, Weissenbach J, Kafatos FC, Collins FH, Hoffman SL: The genome sequence of the malaria mosquito Anopheles gambiae. Science. 2002, 298: 129-149. 10.1126/science.1076181.View ArticlePubMedGoogle Scholar
- Kirkness EF, Bafna V, Halpern AL, Levy S, Remington K, Rusch DB, Delcher AL, Pop M, Wang W, Fraser CM, Venter JC: The dog genome: survey sequencing and comparative analysis. Science. 2003, 301: 1898-1903. 10.1126/science.1086432.View ArticlePubMedGoogle Scholar
- Whitfield CW, Band MR, Bonaldo MF, Kumar CG, Liu L, Pardinas JR, Robertson HM, Soares MB, Robinson GE: Annotated expressed sequence tags and cDNA microarrays for studies of brain and behavior in the honey bee. Genome Res. 2002, 12: 555-566. 10.1101/gr.5302.PubMed CentralView ArticlePubMedGoogle Scholar
- Whitfield CW, Cziko AM, Robinson GE: Gene expression profiles in the brain predict behavior in individual honey bees. Science. 2003, 302: 296-299. 10.1126/science.1086807.View ArticlePubMedGoogle Scholar
- Gracey AY, Troll JV, Somero GN: Hypoxia-induced gene expression profiling in the euryoxic fish Gillichthys mirabilis. Proc Natl Acad Sci U S A. 2001, 98: 1993-1998. 10.1073/pnas.98.4.1993.PubMed CentralView ArticlePubMedGoogle Scholar
- Rise ML, von Schalburg KR, Brown GD, Mawer MA, Devlin RH, Kuipers N, Busby M, Beetz-Sargent M, Alberto R, Gibbs AR, Hunt P, Shukin R, Zeznik JA, Nelson C, Jones SR, Smailus DE, Jones SJ, Schein JE, Marra MA, Butterfield YS, Stott JM, Ng SH, Davidson WS, Koop BF: Development and application of a salmonid EST database and cDNA microarray: data mining and interspecific hybridization characteristics. Genome Res. 2004, 14: 478-490. 10.1101/gr.1687304.PubMed CentralView ArticlePubMedGoogle Scholar
- Hittel D, Storey KB: Differential expression of adipose- and heart-type fatty acid binding proteins in hibernating ground squirrels. Biochim Biophys Acta. 2001, 1522: 238-243. 10.1016/S0167-4781(01)00338-4.View ArticlePubMedGoogle Scholar
- Gracey AY, Cossins AR: Application of microarray technology in environmental and comparative physiology. Annu Rev Physiol. 2003, 65: 231-259. 10.1146/annurev.physiol.65.092101.142716.View ArticlePubMedGoogle Scholar
- Hogstrand C, Balesaria S, Glover CN: Application of genomics and proteomics for study of the integrated response to zinc exposure in a non-model fish species, the rainbow trout. Comp Biochem Physiol B Biochem Mol Biol. 2002, 133: 523-535. 10.1016/S1096-4959(02)00125-2.View ArticlePubMedGoogle Scholar
- Ju Z, Dunham RA, Liu Z: Differential gene expression in the brain of channel catfish ( Ictalurus punctatus) in response to cold acclimation. Mol Genet Genomics. 2002, 268: 87-95. 10.1007/s00438-002-0727-9.View ArticlePubMedGoogle Scholar
- Oleksiak MF, Kolell KJ, Crawford DL: Utility of natural populations for microarray analyses: isolation of genes necessary for functional genomic studies. Mar Biotechnol (NY). 2001, 3: S203-11. 10.1007/s10126-001-0043-0.View ArticleGoogle Scholar
- Oleksiak MF, Churchill GA, Crawford DL: Variation in gene expression within and among natural populations. Nat Genet. 2002, 32: 261-266. 10.1038/ng983.View ArticlePubMedGoogle Scholar
- Nakazawa M, Cebria F, Mineta K, Ikeo K, Agata K, Gojobori T: Search for the evolutionary origin of a brain: planarian brain characterized by microarray. Mol Biol Evol. 2003, 20: 784-791. 10.1093/molbev/msg086.View ArticlePubMedGoogle Scholar
- Stearns SC, Magwene P: The naturalist in a world of genomics. Am Nat. 2003, 161: 171-180. 10.1086/367983.View ArticlePubMedGoogle Scholar
- Feder ME, Mitchell-Olds T: Evolutionary and ecological functional genomics. Nat Rev Genet. 2003, 4: 651-657. 10.1038/nrg1128.View ArticlePubMedGoogle Scholar
- Helfman GS, Collette BB, Facey DE: The diversity of fishes. 1997, Malden, Blackwell science, 528-Google Scholar
- Fryer G, Iles TD: The Cichlid Fishes of the Great Lakes of Africa. 1972, Edinburgh, Oliver & Boyd, 641-Google Scholar
- Streelman JT, Zardoya R, Meyer A, Karl SA: Multilocus phylogeny of cichlid fishes (Pisces: Perciformes): evolutionary comparison of microsatellite and single-copy nuclear loci. Mol Biol Evol. 1998, 15: 798-808.View ArticlePubMedGoogle Scholar
- Sturmbauer C: Explosive speciation in cichlid fishes of the African Great lakes: a dynamic model of adaptive radiation. Journal of Fish Biology. 1998, 53: 18-36. 10.1006/jfbi.1998.0808.View ArticleGoogle Scholar
- Albertson RC, Markert JA, Danley PD, Kocher TD: Phylogeny of a rapidly evolving clade: the cichlid fishes of Lake Malawi, East Africa. Proc Natl Acad Sci U S A. 1999, 96: 5107-5110. 10.1073/pnas.96.9.5107.PubMed CentralView ArticlePubMedGoogle Scholar
- Miya M, Takeshima H, Endo H, Ishiguro NB, Inoue JG, Mukai T, Satoh TP, Yamaguchi M, Kawaguchi A, Mabuchi K, Shirai SM, Nishida M: Major patterns of higher teleostean phylogenies: a new perspective based on 100 complete mitochondrial DNA sequences. Mol Phylogenet Evol. 2003, 26: 121-138. 10.1016/S1055-7903(02)00332-9.View ArticlePubMedGoogle Scholar
- Verheyen E, Salzburger W, Snoeks J, Meyer A: Origin of the superflock of cichlid fishes from Lake Victoria, East Africa. Science. 2003, 300: 325-329. 10.1126/science.1080699.View ArticlePubMedGoogle Scholar
- Taborsky M: Broodcare helpers in the cichlid fish Lamprologus brichardi: their costs and benefits. Animal Behavior. 1984, 32: 1236-1252.View ArticleGoogle Scholar
- Liem KF: Functional Morphology. Cichlid Fishes: Behaviour, ecology and evolution. Edited by: Keenleyside MHA. 1991, Dordrecht, Kluwer Academic PublishersGoogle Scholar
- Meyer A: Phylogenetic relationships and evolutionary processes in East African cichlid fishes. Evolutionary Ecology. 1993, 8: 279-284. 10.1016/0169-5347(93)90255-N.View ArticleGoogle Scholar
- Huysseune A: Phenotypic plasticity in the lower pharyngeal jaw dentition of Astatoreochromis allaudi (Teleostei: Cichlidae). Arch Oral Biol. 1995, 40: 1005-1040. 10.1016/0003-9969(95)00074-Y.View ArticlePubMedGoogle Scholar
- Balshine-Earn S, Neat FC, Reid H, Taborsky M: Paying to stay or paying to breed? Field evidence for direct benefits of helping behavior in a cooperatively breeding fish. Behavioral Ecology. 1998, 9: 432-438.View ArticleGoogle Scholar
- Barlow GW: The cichlid fishes: Nature's grand experiment in evolution. 2000, Cambridge, Perseus Publishing, 335-Google Scholar
- Hofmann HA, Fernald RD: Social status controls somatostatin neuron size and growth. J Neurosci. 2000, 20: 4740-4744.PubMedGoogle Scholar
- Seehausen O, Witte F, Alphen JJM, Bouton N: Direct mate choice maintains diversity among sympatric cichlids in Lake Victoria. Journal of Fish Biology. 1998, 53: 37-55. 10.1006/jfbi.1998.0812.View ArticleGoogle Scholar
- Danley PD, Kocher TD: Speciation in rapidly diverging systems: lessons from Lake Malawi. Mol Ecol. 2001, 10: 1075-1086. 10.1046/j.1365-294X.2001.01283.x.View ArticlePubMedGoogle Scholar
- DOE Joint Genome Institute: JGI Fugu v3.0. http://genome.jgi-psf.org/fugu6/fugu6.home.html.Google Scholar
- Datson NA, van der Perk J, de Kloet ER, Vreugdenhil E: Expression profile of 30,000 genes in rat hippocampus using SAGE. Hippocampus. 2001, 11: 430-444. 10.1002/hipo.1058.abs.View ArticlePubMedGoogle Scholar
- Evans SJ, Datson NA, Kabbaj M, Thompson RC, Vreugdenhil E, De Kloet ER, Watson SJ, Akil H: Evaluation of Affymetrix Gene Chip sensitivity in rat hippocampal tissue using SAGE analysis. Serial Analysis of Gene Expression. Eur J Neurosci. 2002, 16: 409-413. 10.1046/j.1460-9568.2002.02097.x.View ArticlePubMedGoogle Scholar
- Ryan MJ, Rosenthal GG: Variation and selection in swordtails. Model Systems in Behavioral Ecology: Integrating Conceptual, Theoretical, and Empirical Approaches. Edited by: Dugatkin LA. 2001, Princeton, Princeton University Press, 133-148.Google Scholar
- Reznick DA, Bryga H, Endler J: Experimentally induced life-history evolution in a natural population. Nature. 1990, 346: 357-359. 10.1038/346357a0.View ArticleGoogle Scholar
- Carroll RL: Vertebrate paleontology and evolution. 1987, New York, W.H. freeman and company, 698-Google Scholar
- Stearns SC, Hendry AP: Evolution illuminated: Salmon and their relatives. 2003, Oxford, Oxford university pressGoogle Scholar
- Westerfield M: The Zebrafish Book. 1995, Eugene, University of Oregon PressGoogle Scholar
- Churchill GA: Fundamentals of experimental design for cDNA microarrays. Nat Genet. 2002, 32 Suppl: 490-495. 10.1038/ng1031.View ArticlePubMedGoogle Scholar
- Cui X, Churchill GA: Statistical tests for differential expression in cDNA microarray experiments. Genome Biol. 2003, 4: 210-10.1186/gb-2003-4-4-210.PubMed CentralView ArticlePubMedGoogle Scholar
- Lee ML, Kuo FC, Whitmore GA, Sklar J: Importance of replication in microarray gene expression studies: statistical methods and evidence from repetitive cDNA hybridizations. Proc Natl Acad Sci U S A. 2000, 97: 9834-9839. 10.1073/pnas.97.18.9834.PubMed CentralView ArticlePubMedGoogle Scholar
- Salzburger W, Meyer A, Baric S, Verheyen E, Sturmbauer C: Phylogeny of the Lake Tanganyika cichlid species flock and its relationship to the Central and East African haplochromine cichlid fish faunas. Syst Biol. 2002, 51: 113-135. 10.1080/106351502753475907.View ArticlePubMedGoogle Scholar
- Seehausen O: Patterns in fish radiation are compatible with Pleistocene desiccation of Lake Victoria and 14,600 year history for its cichlid species flock. Proc R Soc Lond B Biol Sci. 2002, 269: 491-497. 10.1098/rspb.2001.1906.View ArticleGoogle Scholar
- Nishida M: Phylogenetic relationships and evolution of Tanganyikan cichlids: A molecular perspective. Fish Communities in Lake Tanganyika. Edited by: Kawanabe H, Hori M and Nagoshi M. 1997, Kyoto, Kyoto University Press, 1-23.Google Scholar
- Kocher TD: Adaptive evolution and explosive speciation: the cichlid model. Nature Reviews Genetics. 2004, 5: 288-298. 10.1038/nrg1316.View ArticlePubMedGoogle Scholar
- Hegde P, Qi R, Abernathy K, Gay C, Dharap S, Gaspard R, Hughes JE, Snesrud E, Lee N, Quackenbush J: A concise guide to cDNA microarray analysis. Biotechniques. 2000, 29: 548-50, 552-4, 556 passim.PubMedGoogle Scholar
- Smyth GK, Yang YH, Speed T: Statistical issues in cDNA microarray data analysis. Methods Mol Biol. 2003, 224: 111-136. 10.1385/1-59259-364-X:111.PubMedGoogle Scholar
- Gene Expression Omnibus: http://www.ncbi.nlm.nih.gov/geo/query/?acc=GSE975.Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article: verbatim copying and redistribution of this article are permitted in all media for any purpose, provided this notice is preserved along with the article's original URL.