Fig. S1. Example 2D gels at the two temperatures considered: (A) 14°C, (B) 28°C. The areas boxed in (A) are shown in detail in Fig. S2.

Fig. S2. Example spots identified via the differential proteomic analysis. The gel regions are those highlighted in Fig. S1A; arrows indicate two differentially expressed spots.

Fig. S3. Transcript and protein expression at 14°C and 28°C. Each arrow goes from the expression at 28°C to the expression at 14°C. Fine lines are standard error bars. Transcription is given in units of log to the base 10 Miller units measured using chromosomally integrated promoterless ′lacZ fusions. The protein expression is given in arbitrary units based on the spot density on 2D gels. Arrows are labelled with the gene/protein name in Table 1.

Fig. S4. Fitness of strains with temperature-responsive genes knocked out (A) in vitro, (B) in the plant environment. The fitness given is the selection rate constant (r, units h−1). Values above zero indicate increased fitness relative to the marked wild-type strain and values below zero indicate decreased fitness. Each value is an average of 4–10 replicates with standard error bars. The values for the wild-type controls (WT) were insignificantly different from zero (P > 0.1, Wilcoxon signed rank test in each treatment separately or all treatments combined). For ease of comparison in each treatment the scale has been offset such that the control appears equal to zero [observed control values ± standard error (A) 14°C = −0.10 ± 0.19; 28°C = −0.13 ± 0.17; (B) shoot = −0.18 ± 0.12; root = −0.11 ± 0.07]. Strains are labelled according to the gene deleted (Table 1/Table S2).

Table S1. Protein identifications. All the data for a differentially expressed spot are given in one row; in each case the information given is: locus ID from the Pseudomonas fluorescens SBW25 genome; the identified peptides corresponding to those loci; the score of the match (from Mascot); the putative identity of the protein; the observed mass and pI in the gels and the calculated masses and pIs of the identified proteins. Where multiple identifications were obtained from a single spot, these are given in separate columns (up to 5). These correspond both to the same spot from different gels (in which case the same locus may be identified) or multiple proteins found within a single spot (in which case this is indicated in the 'multiple proteins?' column). A threshold protein score of 30 was used to indicate identification, however matches below that threshold are also shown. Table 1 principally lists identifications with a score > 50.

Table S2. Proteins with increased expression at 14°C and 28°C. Highlighted rows correspond to proteins targeted for genetic manipulation and further characterized in Table S3. Cases where a single protein was identified from a differentially expressed spot with high confidence (see Table S1) are given. All identifications from differentially expressed spots and supporting evidence are given in Table S1.

Table S3. Homology of temperature-responsive genes genetically characterized in this study

Table S4. anova for transcription and fitness of knockout strains. Shown are the analyses for transcription (A), in vitro fitness (selection rate constant) (B) and in vivo fitness (selection rate constant) (C). 'Temperature' is 14°C versus 28°C, 'Location' is shoot versus root. Significant effects (P < 0.05) are highlighted in bold. In (C), qualitatively the same results are obtained if the identity of the plant from which the root and shoot measurements were taken is included as a random effect or if the model is sequentially reduced to a minimal adequate model by removing the interaction and 'Location' effects.

Table S5. Correlations among variables. Pairwise Pearson correlations (−1 = perfect inverse correlation, 0 = no correlation, 1 = perfect positive correlation) are given among the averages of each trait for each gene. Fold changes use a log scale. For fitness in vivo, the average of the root and shoot values is used. For Tee1, where two protein spots were identified, the average of the two protein fold changes is used. Genetic fitness effect in vitro at 28°C is not given since this is only distinguishable from zero for one gene (mucD, see Fig. S1).

Table S6. Bacterial strains and plasmids used in this study.

Table S7. Oligonucleotide primers used in this study.

Appendix S1. Experimental procedures.

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

EMI4_102_sm_Fig_S1-4_and_Table_S2-7.doc1468KSupporting info item
EMI4_102_sm_Table_S1.xls62KSupporting 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.