Table S1 Comparison of microarray and qPCRresults. (A) Significance values are shown for the fixed factors(treatment, family and family-by-treatment interaction) in theanalysis for microarray and qPCR for the 14 genes (T1-T14) thatwere validated. The qPCR was analyzed with different sample sets ofvarying size and relatedness to the microarray samples; Microarraysamples: the same samples as in the microarray experiment(= 35); Microarray families: additionalsamples from the same three families that were used in themicroarray analysis (= 57);Non-microarray samples: samples not included in the microarrayanalysis from all six families (= 52);Non-microarray families: samples from the three other families thatwere not included in the microarray analysis(= 30); All samples: all the samples inthe thermal experiment (including the microarray samples,= 87). (B) The number of individualsmaking up these samples sets by family and treatment are shownbelow the results.

Table S2 The primers used in the qPCR analysis.For each primer pair the corresponding names of the microarrayprobes, genes (unique 454 contigs), the best matching Bombyxmori and Drosophila melanogaster homologs are indicated. The primer ID describes whether the gene was used as an endogenous control (C) or a target (T) in the analysis.

Table S3 Principal component analysis of phenotypic measurements of field collected larvae reared in two temperatures (standard and warm), nested ANOVA analysis of these components by individual within family, family within population, population and residual, and the estimated broad-sense heritabilities.

Table S4 Pearson’s correlation estimatesfor consecutive measures of development time and body weight acrosspre-adult developmental stages and correlations of body weight anddevelopment time at 7th instar and pupae in the field collected larvae in two rearing conditions (standard and warm).

Table S5 Mixed model analysis of body weightand development time in the second generation lab-reared larvae (6families with 10 larvae in each treatment per family) before andafter the 7th instar larvae were exposed to the temperature treatments (cold, standard and hot).

Table S6 Lists of probes significant for thefactors: family, treatment or family by treatment interaction withtwo statistical cut-off levels; Subset 1: probes with Pvalue < 0.05, Subset2: probes containing less than one expectedfalse positive (based on Q value estimation), and a list of candidate genes for family by treatment interactions (candidate gene list).

Table S7 Results from the gene set enrichmentanalysis (Babelomics – FatiScan; Al-Shahrour et al. 2006) for pair-wise comparisons of treatments (cold, standard and hot), families (N170, N74 and O171), families within the same treatment and treatments with the same family.

Fig. S1 Bi-plots of pre-treatment differencesin larval weight and last molt date among families and sexes in thesecond-generation lab-reader larvae. (A) Using samples with knownsex. (B) Centering these values by family to highlight thedifferences between the sexes. (C) Combining the samples with knownand inferred sex. Error bars in (A) and (C) indicate standard errorof the means. Female larvae were on average 16.5% heavier thanmales (R= 0.321;F= 18.84;< 0.0001) and reached the last instarstage 1.7 days later than the males(R= 0.514;F= 64.36;< 0.0001). Assuming these samerelationships for our entire dataset, we were able to assignindividuals of unknown sex (34% of entire dataset,= 61/179) using the following criteria:males were required to have pre-treatment phenotype values forlarval weight and last molt date less than the correspondingminimum observed among the females within each given family.

Fig. S2 The distributions of the coefficient of variation (CV) for (A) the microarray probes after pre-processing and (B) qPCR crossing point values (DCt) after filtering.

Fig. S3 P value distributions of the microarray probes for the fixed factors (family, treatment and family by treatment interaction) in the mixed model analysis.

Fig. S4 Direction and magnitude of expressionchange for the probes belonging to the GO and KEGG categories(= 57) identified in the gene setenrichment analysis. Standardized least squared mean expression ineach family plotted in the three temperature treatments. See Table3 for the descriptions of the categories.

Fig. S5 K-means clusters of the probesbelonging to the enriched category response to unfolded protein(GO:0006986). Clusters 1 and 3 represent the majority of probes(= 15 in both) and have a near lineartrend (either decreasing or increasing) in respect to elevation oftemperature. Clusters 2 and 4 contain the remaining probes (2:= 5; 4: = 4)and have nearly identical expression at the temperature extremes(hot and cold), but differ in respect to the standardcondition.

Fig. S6 qPCR validation showing the expressiondifference between hot and cold treatment across 14 genes (T1-T14).Results from the microarray experiment(= 35) and qPCR with multiple samplesets; Microarray samples: the same samples as in the microarrayexperiment (= 35); Microarray families:additional samples from the same three families that were used inthe microarray analysis (= 57);Non-microarray samples: samples not included in the microarrayanalysis from all six families (= 52);Non-microarray families: samples from the three other families thatwere not included in the microarray analysis(= 30); All samples: all the samples inthe thermal experiment (including the microarray samples,= 87). The different sample sets used inthe qPCR analysis are indicated in the inset.

Fig. S7 Correlations of expression between (A)all probes belonging to the category larval serum protein(LSP), (B) two probes belonging to unique LSP genes and (C) thesame LSP probe and one probe belonging to the category:cuticle-binding protein. Each dot represents a uniquebiological sample.

mec5521_sm_TableS1-S5-FigureS1-S7.pdf443KSupporting info item
mec5521_sm_TableS6.xlsx11640KSupporting info item
mec5521_sm_TableS7.xls7768KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.