Transcriptome dynamics of Deinococcus radiodurans recovering from ionizing radiation -- Data Supplement - Supporting Information -- Proceedings of the National Academy of Sciences

Supporting information for Liu et al. (2003) Proc. Natl. Acad. Sci. USA, 10.1073/pnas.0630387100

Supporting Text
Materials and Methods
Deinococcus radiodurans
R1 (DEIRA) Mutant Generation and Characterization. Genes shown in Tables 3 and 4 are good candidates for disruption based on their response to acute irradiation. Disruption vectors were formed by cloning PCR-generated internal segments (400–900 bp) of candidate ORFs into position 295 of the Escherichia coli plasmid pCR2.1 (Invitrogen) according to the manufacturer’s protocol. Purified disruption vectors were then transformed into DEIRA with kanamycin (Km) selection (8 µg/ml) yielding Km-resistant (Km^R) transformants (1). The possibility of heterozygosity (ESP DEIRA contains four haploid genomic copies per cell (2), where some of DEIRA’s haploid chromosomes contain a disruption while others do not, can be overcome by several rounds of selection on nutrient agar [1% tryptone/0.1% glucose/0.5% yeast extract (TGY)] containing Km (25 µg/ml) if the gene is nonessential; permanent heterozygosity indicates that a gene is essential. For construction of the DR0070 DEIRA mutant, a 575-bp PCR-generated internal segment of ORF DR0070 (600 bp) was first cloned into pCR2.1, yielding pMD890. Purified pMD890 was then transformed into DEIRA with Km selection, giving rise to Km^R transformants (yielding DEIRA mutant MD891). Homozygosity of a disruption was confirmed by detailed mapping of a transformant’s integration site. Total DNA from representative transformants was subjected to restriction endonuclease analysis, electrophoresis, and Southern blotting with diagnostic ³²P-radiolabeled probes as shown in Fig. 7. Stationary-phase cultures (OD₆₀₀ ≈ 1.2) of confirmed mutants were irradiated [⁶⁰Cobalt gamma cell irradiation unit (Model 109, J. L. Shepherd and Associates, San Fernando, CA)] to increasing doses extending to 20 kGy and compared to the survival of DEIRA strain MD68 (ref. 1; MD68 is the product of wild-type DEIRA transformed with the autonomously replicating Km^R-encoding plasmid pMD66). Following irradiation, cell viability was determined by appropriate dilution and selective plating for cfu on agar plates containing Km (25 µg/ml), as described (1, 3, 4). To investigate the effect of chronic irradiation, cells grown on rich medium (TGY with or without antibiotic selection) were exposed to continuous γ-radiation in a ¹³⁷Cs irradiator (50 Gy/h) at 30ºC (Atomic Energy, Ottawa). Control cultures were incubated in the absence of γ-radiation at the same temperature. Comparison of Expression Patterns Produced in Microarray and RT-PCR Experiments for a Set of Seven DEIRA Genes.
Real-time quantitative PCR protocol for Table 1: Relative transcript abundances displayed under nonirradiated and irradiated conditions by seven selected DEIRA were measured by real-time quantitative PCR. These experiments were performed on the same RNA samples used for microarray analysis. First-strand cDNA synthesis was carried out in 10-µl reactions containing 1 µg of total RNA, 3 µg of random hexamers (Invitrogen), 10 mM DTT, 500 µM dNTP mix, and 200 units of Superscript II RNase H-reverse transcriptase (Invitrogen) incubated at 42ºC for 60 min. The real-time quantitative PCR amplification was performed in 50-µl reaction volumes containing 0.5-µl aliquots of synthesized first-strand cDNA, gene-specific primer pairs, and 20,000× diluted SYBR Green I dye (Molecular Probes), using Bio-Rad’s iCycler according to their protocol. The PCR cycle parameters were set at 96ºC for 15 s, 55ºC for 30 s, and 72ºC for 30 s; a total of 45 cycles. The fluorescent intensity of SYBR Green I was monitored at the end of each extension step; the copy number of the target cDNA was estimated by the threshold cycle number according to the standard curve.

Total cellular RNA was prepared from DEIRA cells at the indicated time postirradiation (hours). The first-strand cDNA was synthesized with random hexamers using Superscript II RNase H^– reverse transcriptase (GIBCO/BRL) and quantified by microarray hybridization (see Materials and Methods in main text) or by PCR amplification with gene-specific primers (Table 2), using the iCycler according to the manufacturer’s instructions. Amounts of target cDNAs in the irradiated samples are expressed as folds of the nonirradiated control cells. The values of the microarray data are averages of three independent experiments of four replicates, whereas those of real-time quantitative PCR data are averages of two independent experiments of three replicates.

Results
Expression Pattern Correlation Between Genes in Predicted Operons and Quality of Microarray Analysis.
Operons are the principal form of gene coregulation in prokaryotes (5), and the expression patterns of genes within an operon are expected to be strongly correlated. To test this prediction, we compared the expression profiles of genes in predicted DEIRA operons with random gene groups.

Conserved operons in DEIRA were identified using the approach described (6). We considered all conserved operons common for DEIRA and at least one other completely sequenced prokaryotic genome. Only those operons that contained unidirectional genes separated by <100 bp were further analyzed. In total, 141 predicted operons containing 435 genes were predicted. Pearson’s linear correlation coefficient (CC) was used to measure the correlation between expression patterns of any two given genes. The significance of a correlation (Pcc) was computed using the STATISTICA program. The number of genes that had a significant correlation (SC) with at least one gene from the same predicted operon was calculated for each of the predicted operons. To test the hypothesis that genes within one operon show a significant correlation between their expression profiles with a higher frequency compared to genes located at random positions in the genome, a Monte Carlo simulation was used. To create a random set of gene groups "pseudo-operons," 435 genes were randomly chosen from the 2,976 analyzed genes and each gene was randomly assigned to one of 141 gene groups. The number of genes within each gene group was the same as in the analysis of the actual data, but all genes included in one random gene group were separated by at least 10 genes in the genome. Correlation analysis was performed for 10,000 sets of random gene groups and the number of genes that had a significant correlation with at least one gene from the same "pseudo-operon" (SC*, SC values observed in random gene groups) was calculated for each set.

Among 435 genes within 141 predicted operons, 297 genes showed significant correlation (CC ³ 0.67, Pcc £ 0.05), with at least one other gene from the same operon (SC = 68%). Analysis of the 10,000 SC* values for random gene groups showed that they formed a normal distribution, with a mean value of 45% (196 genes), and maximal and minimal SC values of 52% and 39%, respectively (SD 1.7%). It should be noted that the results for random gene groups do not necessarily represent a background of noncorrelated expression because genes involved in unrelated processes might have similar expression profiles because they could be independently responding to γ-radiation. Nevertheless, the probability that the observed SC value for genes in predicted DEIRA operons (68%) belongs to the SC* distribution was 6.9 × 10- ²⁰ (Zelterman’s statistics).

Poorly Characterized Genes That Play a Role in Cell Recovery After Irradiation.
DEIRA has many expanded families of paralogous genes that have been explored in detail previously and are suggested to play some role in radiation resistance (7, 8). Indeed, many genes from expanded families are up-regulated, e.g., those of the TerEDXZ/CABP/SCP2 family (DR2220–DR2225 and DRA0057) and the PR1 family gene (DR1548). The most notable example of induction of genes from the same family are the members of the YfiT/DinB family, which includes homologs of DinB-like proteins that belong to the SOS regulon of Bacillus subtilis (9). DEIRA has 12 members of this family (8), and four of them are strongly induced in a recA-like manner (DR0053, DR0841, DR1642, and DR1899), supporting an important role in the irradiation response (Table 3).

1. Daly, M. J., Ouyang, L., Fuchs, P. & Minton, K. W. (1994) J. Bacteriol. 176, 3508–3517.

2. Hunter, M. T. (1978) J. Bacteriol. 134, 71–75.

3. Daly, M. J. & Minton, K. W. (1995) J. Bacteriol. 177, 5495–5505.

4. Daly, M. J. & Minton, K. W. (1996) J. Bacteriol. 178, 4461–4471.

5. Miller, J. H. & Reznikoff, W. S. E. (1978) The Operon (Cold Spring Harbor Lab. Press, Plainview, NY).

6. Wolf, Y. I., Rogozin, I. B., Kondrashov, A. S. & Koonin, E. V. (2001) Genome Res. 11, 356–72.

7. Makarova, K. S., Aravind, L., Daly, M. J. & Koonin, E. V. (2000) Genetica 108, 25–34.

8. Makarova, K. S., Aravind, L., Wolf, Y. I., Tatusov, R. L., Minton, K. W., Koonin, E. V. & Daly, M. J. (2001) Microbiol. Mol. Biol. Rev. 65, 44–79.

9. Yasbin, R. E., Cheo, D. & Bayles, K. W. (1991) Res. Microbiol. 142, 885–892.