Email updates

Keep up to date with the latest news and content from BMC Genomics and BioMed Central.

Open Access Highly Accessed Research article

Genome wide analysis of common and specific stress responses in adult drosophila melanogaster

Fabrice Girardot12, Véronique Monnier1 and Hervé Tricoire1*

Author Affiliations

1 Institut Jacques Monod, 2 place Jussieu, 75251 Paris, France

2 Present address: Equipe de Biologie Virtuelle, UMR 6543, Universilé de Nice, Parc Valrose, 06108 Nice Cedex 2, France

For all author emails, please log on.

BMC Genomics 2004, 5:74  doi:10.1186/1471-2164-5-74

Published: 30 September 2004

Additional files

Additional file 1:

For each of the 1368 probe sets identified as stress responsive in our data analysis, we calculated and reported in this table, for each stress condition, the mean ratio <Stress condition> ≤ (AVstressi / AVrefj)>i,j where AVstressi and AVrefj correspond to the average value measured for the ith sample in the stress condition and the jth sample respectively in the reference condition. To facilitate visual inspection, we used a color code (red corresponding to upregulation, green to downregulation) with thresholds corresponding to fold changes of 1.8 (dark colors), 1.5 (medium) and 1.25 (light). The standard error for each measurement is given in parenthesis. For each probe set, the mean detection p-value from MAS5 analysis of reference samples is reported in column 4 and cluster assignment in column 9. .

Format: XLS Size: 362KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional file 2:

For stresses induced by a) paraquat (5mM and 15mM experiments), b) H2O2 or c) tunicamycin we analyzed the distribution in biological processes (as defined by the Gene Ontology (GO) database) of the genes selected by the SAM analysis (responsive genes) and compared it to the same distribution for all the genes significantly detected on our microarrays (analysed genes). We report here the significantly over- or under-represented (P < 0.005) biological process and the number of analysed and responsive genes found inside these classes, for the different stress conditions. The p-value P associated to the null hypothesis of no association with a binomial distribution hypothesis is given for each class, (only classes with P < 0.005 were retained). For clarity of the figure some redundant branches of the tree were removed. Color codes for the classes: dark blue: classes present in the 3 stress responses; yellow: classes present in the two oxidative stress responses; green: classes present in paraquat and tunicamycin stress responses; light cyan: classes present in H2O2 and tunicamycin stress responses. 34 Color code for statistical analysis: orange: underrepresented class, blue: over-represented class. .

Format: XLS Size: 33KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional file 3:

For over or under-represented molecular functions classes we report here the number of analyzed (column 3) and responsive genes found inside the 7 groups of clusters A to G (columns 4 to 10, see text for details on the definition of these groups). A schematic response to oxidative and ER stress of the genes included in these groups is given in the first two lines. The number of genes inside each group is given in line 3. A color code identifies cases when the number of genes differs statistically (p<0.005) from a random distribution: orange: under-represented class, blue: over-represented class.

Format: XLS Size: 34KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional file 4:

For over or under-represented biological process classes we report here the number of analyzed (column 3) and responsive genes found inside the 7 groups of clusters A to G (columns 4 to 10, see text for details on the definition of these groups). A schematic response to oxidative and ER stress of the genes included in these groups is given in the first two lines. The number of genes inside each group is given in line 3. A color code identifies cases when the number of genes differs statistically (p<0.005)from a random distribution: orange: under-represented class, blue: over-represented class.

Format: XLS Size: 32KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional file 5:

Stress response for the peptidases

Format: XLS Size: 140KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional file 6:

List of stress responsive genes detected in aging and hyperoxia, immune or starvation stress experiments were compared with our data. Lines 1 to 3 indicate the number of genes found to be repressed (-), induced (+) or either (total) in the different experiments. Lines 5 to 7 indicate the number of genes in each of these categories found among our 1397 35 stress responsive genes (classes A to F) and the corresponding percentage from the initial number. In lines 8, 9 (respectively 10, 11) the same analysis is reported for genes included in the A (respectively B) classes defined as common stress responsive classes in our analysis. O2, old, infection, starvation: expression data from the different experiments compared to our data. All: List of 26 genes which are responsive to at least 4 stresses in these independent experiments

Format: XLS Size: 696KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional file 7:

Relationships to other stresses

Format: DOC Size: 43KB Download file

This file can be viewed with: Microsoft Word Viewer

Open Data

Additional file 8:

Table 1: Stress response characteristics of clusterized genes.The 1368 probe sets retained after statistical analysis were submitted to a divisive clustering algorithm (SOTA) which predicted 19 clusters. For each probe set k inside a cluster we calculated, for each stress condition, the mean ratio Rk = <Ln2 (AVstressi / AVref j )>i,j where AVstressi and AVref j denote the average value measured for the ith sample in the stress condition and the jth sample respectively in the reference condition. The mean of the Rk values provides a measurement of the mean intensity of variation for the genes inside a cluster, which is reported in this table. To facilitate visual inspection, we used a color code (red colors corresponding to upregulation, green colors to downregulation) with thresholds corresponding to fold changes of 1.8 (dark colors), 1.5 (medium) and 1.25 (light). The number N of probe sets in each cluster is also reported. From these values we identified groups of clusters (named from A to G) which present close behavior and were used for statistical functional analysis. Clusters corresponding to the common long term stress response (CLTSR) are outlined in red.

Format: XLS Size: 19KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional file 9:

Table 2: Functional analysis of stress responsive genes.For stresses induced by a) paraquat (5mM and 15mM experiments), b) H2O2 or c) tunicamycin we analyzed the distribution in functional classes (as defined by the Gene Ontology (GO) database) of the genes selected by the SAM analysis (responsive genes) and compared it to the same distribution for all the genes significantly detected on our microarrays (analysed genes). We report here the significantly over- or under-represented (P<0.005) molecular functions and the number of analysed and responsive genes found inside these classes, for the different stress conditions. The p-value P associated to the null hypothesis of no association with a binomial distribution hypothesis is given for each class, (only classes with P<0.005 were retained). For clarity of the figure some redundant branches of the tree were removed. Color codes for the classes: dark blue: classes present in the 3 stress responses; yellow: classes present in the two oxidative stress responses; green: classes present in paraquat and tunicamycin stress responses; light cyan: classes present in H2O2 and tunicamycin stress responses. Color code for statistical analysis: orange: under-represented class, blue: over-represented class.

Format: XLS Size: 35KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional file 10:

Table 3: Analysis of stress responses for members of some functional classes.From the 1368 stress responsive probe sets we extracted the subsets associated with genes annotated in the GO database as a) glutathione transferases (GO:0004364), b) P450 cytochromes (from list at http://p450.antibes.inra.fr/) and c) proteasome component (GO:0004299). For each probe set k within one of these subsets, we calculated and reported in this table, for each stress condition, the mean ratio <Stress condition >k =<(AVstressi / AVref j)>i,j where AVstressi and AVrefj correspond to the average value measured for the ith sample in the stress condition and the jth sample respectively in the reference condition. To facilitate visual inspection, we used a color code (red corresponding to upregulation, green to downregulation) with thresholds corresponding to fold changes of 1.8 (dark colors), 1.5 (medium) and 1.25 (light). The standard error for each measurement is given in parenthesis. For each probe set, the mean detection p-value from MAS5 analysis of reference samples is reported in column 3 and cluster assignment in column 8. In column 9 additional information is reported for each class: in a) we indicate the GT class deduced from sequence comparison with human and mouse GTs and from [44] (D: delta, O: omega, T: theta, T2: distantly related to theta, Z: zeta); in b) the name of the genes are reported; in c) we indicate the proteasome subunit to which the genes defined in column 2 belong. Note that in c) a large number of genes not retained by SAM analysis (without cluster number) seem to be upregulated in P15 condition. Genes used for comparison between microarray and quantitative RT-PCR (Fig. 3) are outlined in bold character.

Format: XLS Size: 41KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional file 11:

Table 4: Analysis of mutant flies' resistance to paraquat-induced oxidative stress. Paraquat resistance of 29 mutant lines was assayed in three independent experiments as described in Fig. 4. The survival data were submitted to a log-rank statistical analysis by comparison with w/+ reference flies. The results are presented in this table. Column 1 contains the tested genotype (same conventions as in Fig. 4a : Bloomington line numbers). The corresponding genotypes are described in column 6. The symbol for the gene affected is reported in column 2. Information from FlyBase about the allele used in this study is given in column 7 with a one character code: A: amorph; H: hypomorph; N: non complementation of deficiency; L: letal; R: recessive mutation; 5: insertion in the 5' regulatory region, 5'UTR or intron; C: insertion in the coding region. Column 8 indicates whether the tested line was outcrossed or not before the test. Column 4 is the result of a log-rank analysis of the second survival experiments shown in figure 4. A confidence index which refers to the mean of log10 (p-log-rank) for the three experiments is given in column 5. We considered that a strain had a significant effect on survival under oxidative stress conditions when this confidence index was lower than -3 and at least two experiments presented p-log-rank < 0.001. Under these stringent conditions 13 genotypes are shown to confer resistance (R) or sensitivity (S) to paraquat as indicated in column 3.

Format: XLS Size: 19KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data