Supporting information for Beltran et al. (2003) Proc. Natl. Acad. Sci. USA, 10.1073/pnas.0538075100
Supporting Materials and Methods
Drosophila stocks.
WT and mutant strains were maintained on standard medium and all experiments were performed at 25ºC. The WT stock used was Canton-S (Bloomington Stock Center, Indiana University, Bloomington). ash2I1/TM6C mutant line was obtained from an excision of a P-lacW element inserted in the ash2 gene (1), and homozygous ash2I1 larvae could be selected due to their Tubby+ phenotype.RNA analysis.
Total RNA was extracted from larvae by using Trizol (GIBCO/BRL) and poly(A) by using the polyA-TRACT system (Promega). mRNA (0.5 µg) was used for the Northern blot analysis. For the developmental Northern blot, we used poly(A) RNA isolated from different stages (kindly provided by S. Campuzano, Consejo Superior de Investigaciones Científicas, Universidad Autónoma, Madrid). The ash2 probe used for this analysis was a 659-bp PCR product (exons 5–8) obtained from the 3' end of the coding region, and a 300-bp fragment of the rp49 gene was used as a loading control. All probes were labeled with [32P]dCTP by random priming.5'-RACE was carried out by using the SMART RACE cDNA Amplification Kit (CLONTECH) according to the manufacturer’s instructions. The gene-specific primers were ash2- c1292 (5'-CGTGGCAGCTCCGTCGGGCA TCTCTTC-3') for the first 5'-RACE PCR and ash2-2R (5'-CTGACGAAATGG AGCATGTG-3') for the nested PCR. The resulting fragments of expected size were isolated from agarose gel by using the QIAquick Gel Extraction Kit (Qiagen, Chatsworth, CA) and automatically sequenced. In silico analysis was performed by using GENEID at /www1.imim.es/software/geneid/index.html (28) and GENSCAN at genes.mit.edu/GENSCAN.html (329).
Microarray construction.
cDNA microarrays were constructed by using the ESTs from the Drosophila Gene Collection 1.0 (DGC1.0) set (www.fruitfly.org/DGC/index.htlm). cDNA clones were PCR amplified directly from the bacteria glycerol stocks in 96-well format by using specific primers from pOT2 and pBS vectors. All PCR reactions were analyzed by agarose gel electrophoresis, and DNA was ready to spot after purification of the PCR products by Ethanol precipitation. Concentration of DNA was around 1 µg/µl. Purified PCR products (1–7 kb long) were robotically spotted (Cartesian PS5500) on poly-lysine coated microscope slides (PL-25C, CEL Associates, Houston). Slides were UV cross-linked at 60 mJ in an UV Stratalinker (Stratagene) and made ready for prehybridization. To verify clone identity, some cDNAs were randomly chosen and sequenced by using an ABI Prism 377 DNA sequencer (Perkin–Elmer).Probe preparation and hybridization.
Total RNA and poly(A) were purified as described above. One to three micrograms of poly(A) RNA were labeled by reverse transcription incorporation of Amino-allyl dUTP and coupling to NHS–cyanine dye using the Atlas Glass fluorescent labeling kit (CLONTECH) following the manufacturer’s specifications, with the exception of the probe purification that was performed by sodium acetate/EtOH precipitation. Ten microliters of Cy3- and Cy5-labeled probes resuspended in water were mixed with 5 µl of formamide/5 µl of 20´ SSC/0.3 µl of 10% SDS. Probe was warmed at 90ºC for 5 min before being added to the slide. For the hybridization, Corning Hybridization Chamber (no. 2551) and Erie Scientific (Portsmouth, NH) cover slips (no. 22IX25-2-4635) were used, and slides were hybridized overnight at 42ºC, washed, and scanned by using a ScanArray 4000 (Perkin–Elmer) laser scanner.Data acquisition and analysis.
Fly Base gene number for all possible cDNAs was found by comparing the Berkeley Drosophila Genome Project Collection Table with "external-databases.txt (10-23-01)" from FLYBASE (4). Data were acquired with GENEPIX PRO 3.0 (Axon Instruments, Foster City, CA) and filtered with the aid of CONVERT DATA 3.33 (www.le.ac.uk/cmht/microarray_lab/Microarray_Softwares/Microarray_Softwares.htm) default settings. For each experiment, the log (mF635-mB635) was plotted against the log (mF532-B532) for the filtered data, giving a mean slope for all experiments of 1.01 and a mean correlation coefficient (R2) of 0.84; therefore, no dye bias correction was applied. Our four chips came from two different batches and did not have exactly the same genes. To assess the reproducibility of data between chips coming from the same batch and, therefore, having exactly the same set of genes, we calculated the Pearson’s correlation coefficient between raw (before normalization of data) fluorescence intensities (at 532 and 635 nm) of all genes flagged as Good. The mean of these four correlation coefficients was 0.74. To estimate variability of the experimental procedure, we labeled two samples independently with Cy3 or Cy5 and hybridized a cDNA microarray. The correlation coefficient between raw fluorescent intensities F635-B635 and F532-B532 of the 3,819 spots flagged as okay was 0.97. The standardized ratios had a SD of 0.22, and 99.76% of the spots were below the 1.75-fold change, rendering this value as a good threshold to use in the One Class Response Significance Analysis of Microarrays (SAM) (5) conducted with an ≈0.025 false discovery rate (FDR).Gene ontology (GO) classification.
Automatic functional annotation of the regulated genes in our microarray experiments has been obtained using the GO database (www.geneontology.org; ref. 6).A number of scripts were written to parse the GO database and the association file (gene_association.fb from 10/10/2001; ftp://ftp.geneontology.org/pub/go/gene-associations) to obtain, in a recursive fashion, GO definitions for the Drosophila genes. Pie charts have been used to display GO classifications for different data sets (see Fig. 3 and http://www.ub.es/epidd/arrays/index.htm for details on the whole process).
To produce Fig. 4, a table was constructed where genes (rows) that had a given GO term (columns) presented a positive value in the intersecting cell. This table was used to perform an average linkage clustering with CLUSTER and TREEVIEW programs (7) that grouped together similar GO terms in one axis and genes with similar GO descriptions in the other. The *.cdt file was loaded to Microsoft EXCEL to modify color of cells according to their ratio and add additional information.
Chromosomal localization and sequence analysis.
Genes from our microarray were mapped onto chromosome coordinates of the Drosophila genome. The National Center for Biotechnology Information distribution as of April 10, 2000, of the Drosophila genome was used. All annotated genes in the genomic scaffolds were mapped onto chromosomal coordinates, by matching the 120 bases at the 5' end of each gene against the assembled chromosomal sequences. We used BLASTN (8) to compare the gene to the genomic sequences.For the set of genes from our microarray whose chromosomal coordinates were annotated, Moran’s index (9) was computed to measure their aggregation in the Drosophila genome with respect to their level of expression in the microarray experiments. This index is a measure of correlation between the values of a spatial location attribute and a second attribute representing some specific property in a given dataset. Moran's index has been recently used to analyze the aggregation of genes in the human genome with respect to the observed mutational rates and GC content changes between human genes and their mouse orthologs (10). In our case, this index measures the correlation between the geographical localization of the genes along the chromosomes and their level of expression. As a correlation coefficient, the values for the index range from +1 meaning strong positive spatial correlation to - 1, indicating strong negative correlation, with 0 meaning a random pattern of distribution in the space. To assess the significance of the computed value for the index, 1,000 Monte Carlo tests with randomly shuffled data were performed. The significance level is the proportion of times that the value of the index for a random permutation was larger than the computed value for the original data.
RT-PCR.
Total RNA (750 ng) and poly(dT)-24 (100 ng) were used for cDNA synthesis. The reaction was carried out in a final volume of 25 µl by using 5 units of Avian myeloblastosis virus–RT (Promega) and 200 units of Moloney murine leukemia virus RT (GIBCO). One microliter of the RT reaction was used for the PCR reaction, and all specific primers were designed to amplify a product of about 500 bp.Generation of mitotic recombination clones.
In imaginal discs, clones were generated by FLP-mediated mitotic recombination (11). Clones were induced in second instar larvae (50 and 60 h after egg laying, AEL) by 25-min heat shock at 37ºC. Mutant clones for ash2I1 were detected by the absence of β-galactosidase (β-gal) immunostaining in clones generated in the genotypes hsFLP; FRT 82B arm lacZ/FRT82B ash2I1 for twin-spot analysis and hsFLP; FRT82B M(3)arm lacZ/FRT82B ash2I1 for clonal analysis in a Minute background to generate large clones. After incubation in PBS containing 1% BSA (Sigma) and 0.1% TritonX-100 to block nonspecific binding, the larvae were immunostained on whole-mount preparations with the rabbit anti-CYCA (1:150), kindly provided by D. Glover (Cambridge University, Cambridge, U.K.), mouse anti-UBX (1:20), kindly provided by R. White (Cambridge University), mouse anti-FASII (1:4) from the University of Iowa and anti-β-gal (Cappel) antibodies. Incubation was done in blocking serum overnight at 4ºC. Secondary antibodies conjugated to Rhodamine red and FITC (Jackson ImmunoResearch) were used. The specimens were analyzed under a Leica confocal laser-scanning microscope. Mitotic recombination was also induced by x-rays from a Philipps x-ray source operating at a dose of 10 Gy (100 kV, 15 mA, 2-mm Al filter). We used f36a, mwh f+87D M(3)w124/ash2I1 to generate M+ clones at 60-h AEL. Adult flies of appropriate genotype were dissected out and their wings mounted in lactic acid/ethanol (1:1) for microscopy.Additional data.
Additional data are available at www.ub.es/epidd/arrays/index.htm.1. Amorós, M., Corominas, M., Deak, P. & Serras, F. (2002) Int. J. Dev. Biol. 46, 321–324.
3. Burge, C. & Karlin, S. (1997) J. Mol. Biol. 268, 78–94.
4. Anonymous (1999) Nucleic Acids Res. 27, 85–88.
5. Tusher, V. G., Tibshirani, R. & Chu, G. (2001) Proc. Natl. Acad. Sci. USA 98, 5116–5121.
6. Ashburner, M., Ball, C. A., Blake, J. A., Botstein, D., Butler, H., Cherry, J. M., Davis, A. P., Dolinski, K., Dwight, S. S., Eppig, J. T., et al. (2000) Nat. Genet. 25, 25–29.
7. Eisen, M. B., Spellman, P. T., Brown, P. O. & Botstein, D. (1998) Proc. Natl. Acad. Sci. USA 95, 14863–14868.
8. Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W. & Lipman, D. J. (1997) Nucleic Acids Res. 25, 3389–3402.
9. Moran, P. A. P. (1950) Biometrika 37, 17–23.
10. Castresana, J. (2002) Nucleic Acids Res. 30, 1751–1756.
11. Xu, T. & Rubin, G. M. (1993) Development (Cambridge, U.K.) 117, 1223–237.