Regional organization of gene expression in Streptomyces coelicolor

Gene 353 (2005) 53 – 66 www.elsevier.com/locate/gene

Regional organization of gene expression in Streptomyces coelicolor Nitsara Karoonuthaisiria, David Weavera, Jianqiang Huangb, Stanley N. Cohenb, Camilla M. Kaoa,T a

Department of Chemical Engineering, MC 5025, Stanford University, Stanford, CA 94305, USA b Department of Genetics, MC 5120, Stanford University, Stanford, CA 94305, USA

Received 10 October 2004; received in revised form 4 February 2005; accepted 14 March 2005 Received by A. Travers

Abstract Based on the chromosomal locations of genes inferred from sequence analysis to be essential for the viability of Streptomyces coelicolor, Bentley et al. [Bentley, S.D., et al. 2002. Complete genome sequence of the model actinomycete Streptomyces coelicolor A3(2), Nature 417, 141 – 147.] have suggested that a 4.9 Mb central region of the linear S. coelicolor chromosome encodes Fcore_ functions expressed during vegetative growth of this species, while 1.5 Mb and 2.3 Mb chromosomal DNA segments lateral to this core encode auxiliary functions proposed to be required under other growth conditions. To examine this hypothesis and experimentally identify genes expressed during vegetative growth of S. coelicolor cultures, we used DNA microarrays to measure globally the abundance of S. coelicolor transcripts in cells growing in liquid medium. We found that, overall, genes corresponding to the 4.9 Mb core region of the S. coelicolor M145 chromosome were more highly expressed under non-limiting growth conditions than genes in the 1.5 Mb left and 2.3 Mb right chromosome arms, supporting the notion of the core versus auxiliary organization of genes on the chromosome. To examine how this chromosomal distribution of transcripts changes under other growth conditions, we also measured gene expression changes during stationary phase and several stress conditions. During stationary phase, the composition of S. coelicolor transcripts appears to shift from large quantities of growth-related transcripts encoded in the core region to those of less characterized genes, which may be essential for differentiation and other physiological responses, encoded throughout the chromosome. After temperature and osmotic upshifts, we found that S. coelicolor transiently induces a set of several hundred genes located throughout the chromosome, which may function in response mechanisms common to the two stress conditions. D 2005 Elsevier B.V. All rights reserved. Keywords: S. coelicolor; Gene expression; Microarrays

1. Introduction Streptomyces bacteria live in the soil and must adapt to a wide range of terrestrial environments (Hodgson, 2000). These bacteria grow as hyphal masses which efficiently colonize soil and differentiate into resting spores when nutrients are depleted. Streptomycetes possess enzymatic pathways to degrade complex debris from dead plants, animals and fungi and synthesize a broad range of bioactive Abbreviations: kb, kilobase(s); Mb, megabase(s); min, minute(s); h, hours(s); RT-PCR, reverse transcription polymerase chain reaction. * Corresponding author. Tel.: +1 650 736 0547; fax: +1 650 723 9780. E-mail address: [email protected] (C.M. Kao). 0378-1119/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.gene.2005.03.042

secondary metabolites, which have made these bacteria a rich source of natural products for the pharmaceutical and agricultural industries. The genetic adaptability of Streptomyces bacteria is evident in the genome sequences of the model organism S. coelicolor M145 and the avermectin-producing S. avermitilis (Bentley et al., 2002; Ikeda et al., 2003). The 8.7 Mb chromosome of S. coelicolor M145 is linear, with an origin of replication located 61 kb left of center and 22 kb inverted repeat sequences at the chromosome ends (terminal inverted repeats; TIRs). The central 4.9 Mb region encodes many genes hypothesized to be essential for growth, including those predicted by sequence analysis to be involved in cell division and DNA replication. This

54

N. Karoonuthaisiri et al. / Gene 353 (2005) 53 – 66

suggested Fcore_ region of the S. coelicolor chromosome is flanked by 1.5 Mb left and 2.3 Mb right arm regions, which contain many genes supposed to encode non-essential proteins, such as hydrolytic and secondary metabolism enzymes. The chromosome arms have been proposed to encode auxiliary functions for specialized circumstances, and the genetic variability of these regions supports the notion that S. coelicolor can propagate itself in the absence of expression of genes in these regions (Chen et al., 2002). The internal regions of the S. avermitilis and S. coelicolor M145 chromosomes are highly conserved (Ikeda et al., 2003), suggesting that the Fcore_ versus Fauxiliary_ genome organization is a general feature of streptomycetes. Previous work in other microorganisms has measured abundances of transcripts under different environmental and physiological conditions as well as examined spatial patterns of transcriptional activity of chromosomes. Transcription in yeast has been analyzed globally by microarrays, serial analysis of gene expression (SAGE), RT-PCR and reporter fusions (Horak and Snyder, 2002), and numerous studies have characterized bacterial expression programs activated in response to various growth, stress, developmental and pathogenic conditions. In S. coelicolor M145, gene expression changes during the transition to stationary phase in liquid cultures have been measured using microarrays (Huang et al., 2001), during growth on solid medium at 42 -C and in a heat shock mutant strain (Bucca et al., 2003). Recent studies in Escherichia coli have applied

genetic and expression methods to identify periodic patterns of gene expression of the E. coli chromosome (Jeong et al., 2004; Kepes, 2004; Valens et al., 2004). To obtain information about the relative abundance of transcripts in S. coelicolor growing vegetatively in liquid medium, we compared the hybridization of genomic DNA and total RNA samples at single gene resolution using DNA microarrays. These experiments assessed the copy numbers of different mRNA transcripts relative to the copy numbers of the corresponding genes in the chromosome, providing information on which genes are highly or poorly expressed during vegetative growth. Overall, genes in the 4.9 Mb core region were found to be more highly expressed under nonlimiting culture conditions than those in the chromosome arms, consistent with the Fcore_ versus Fauxiliary_ organization predicted by Bentley et al. (2002) on the basis of sequence analysis. We also measured gene expression changes during stationary phase and several stress conditions to examine how the chromosomal distribution of transcripts varies in other growth conditions.

2. Materials and methods 2.1. Strains and growth conditions Table 1 lists the microarrays experiments conducted in this study. The S. coelicolor strains M600 and M145 were

Table 1 Microarray experiments performed in this study Experiment

Strain(s)

Medium

Gene expression levels during vegetative growth A M600 SMM B M600 R5 C M145, M600 R5 D M145 R5 E M600 R5, SMM F M145, M600 R5 Gene expression changes during G,H M145, M600 G M600 H M145

Green

Red

Description

DNA DNA DNA—M145 DNA—M145 RNA—R5 RNA—M145

RNA RNA RNA—M600 RNA—M145 RNA—SMM RNA—M600

Vegetative Vegetative Vegetative Vegetative Vegetative Vegetative

RNA 10 h—M600 RNA—12, 16, 20, 24, 34, 45 h RNA—16, 20, 24, 34, 40, 45 h

Initial time comparison Stationary phase time course Stationary phase time course

RNA—15, 30, 45, 60, 90 RNA—15, 30, 45, 60, 90 RNA—15, 30, 45, 60, 90 RNA—15, 30, 45, 60, 90 RNA—15, 30, 45, 60, 90 RNA—15, 30, 45, 60, 90 RNA—15, 30, 45, 60, 90 RNA—15, 30, 45, 60, 90 RNA—15, 30, 45, 60, 90 RNA—15, 30, 45, 60, 90 RNA—15, 30, 45, 60, 90 RNA—15, 30, 45, 60, 90 RNA—15, 30, 45, 60, 90 RNA—5, 10, 15, 30 min

Control (no shift) Control (shift to fresh medium) 30 -CY14 -C 30 -CY35 -C 30 -CY37 -C 30 -CY42 -C Sucrose upshift (0Y1 M) Ethanol upshift (0Y5%) Phosphate upshift (1Y30 mM) Control (no shift) 30 -CY37 -C, biological duplicates Ethanol upshift (0Y5%) Sucrose downshift (0.3Y0 M) 30 -CY37 -C

stationary phase R5 RNA 12 h—M145 R5 RNA—10 h (OD¨1) R5 RNA—12 h (OD¨1)

Gene expression changes during stress conditions I M600 SMM RNA—0 J M600 SMM RNA—0 K M600 SMM RNA—0 L M600 SMM RNA—0 M M600 SMM RNA—0 N M600 SMM RNA—0 O M600 SMM RNA—0 P M600 SMM RNA—0 Q M600 SMM RNA—0 R M600 R5 RNA—0 S M600 R5 RNA—0 T M600 R5 RNA—0 U M600 R5 RNA—0 V M600 SMM RNA—0

min min min min min min min min min min min min min min

(OD¨0.5) (OD¨0.5) (OD¨0.5) (OD¨0.5) (OD¨0.5) (OD¨0.5) (OD¨0.5) (OD¨0.5) (OD¨0.5) (OD¨0.5) (OD¨0.5) (OD¨0.5) (OD¨0.5) (OD¨0.5)

min min min min min min min min min min min min min

growth, growth, growth, growth, growth, growth,

biological biological biological biological biological biological

triplicates triplicates triplicates triplicates triplicates triplicates

N. Karoonuthaisiri et al. / Gene 353 (2005) 53 – 66

used (Kieser et al., 2000). For experiments A –D, genomic DNA was harvested from shake flask cultures grown at 30 -C in YEME medium (Kieser et al., 2000). For experiments A – E, total RNA was isolated from shake flask cultures grown at 30 -C in rich medium (R5) or supplemented minimal medium (SMM); the latter which contained 0.5% instead of 0.2% casamino acids (Kieser et al., 2000). These cultures were grown to an optical density of OD450 ¨ 1 for R5 and OD450 ¨ 0.5 for SMM. Cell samples were harvested by filtration at room temperature (Whatman no. 1002 055; 15 –20 mm diameter) and total RNA was prepared using the modified Kirby mix protocol (Kieser et al., 2000) and modifications previously described (Weaver et al., 2004). For experiments I – V, 120 mL of medium was placed in a 500 mL shake flask and inoculated with 5  107 spores/mL. Each culture was grown to early exponential phase (OD450 ¨ 0.5) and 20 mL was harvested as the reference sample. A shift then was applied and 10 mL samples were harvested after 15, 30, 45, 60 and 90 min (experiments I –U) or after 5, 10, 15 and 30 min (experiment V). Total RNA was prepared as above. 2.1.1. Controls Experiments I and R were performed without applying a shift after the reference sample was taken. After the reference sample was taken for experiment J, the culture was collected by centrifugation at 3000  g for 3 min and resuspended in an equal volume of fresh medium. 2.1.2. Temperature shifts (experiments K – N and S) After the reference samples were taken, each culture was collected by centrifugation at 3000g for 3 min, resuspended in an equal volume of medium preequilibrated at the shift temperature, and returned to the shift temperature for growth. 2.1.3. Ethanol shocks (experiments P and T) After the reference samples were taken, ethanol was added to a final volume concentration of 5%. 2.1.4. Osmotic shifts Experiment O was performed in SMM, which contains no sucrose. After the reference sample was taken, the culture was collected by centrifugation and resuspended in an equal volume of SMM with 1 M sucrose. Experiment U was performed in R5, which contains 300 mM sucrose. After the reference sample was taken, the culture was collected by centrifugation and resuspended in an equal volume of R5 without sucrose.

55

2.2. Microarray hybridizations DNA microarrays were fabricated as previously described (Huang et al., 2001). 97% of the predicted genes in S. coelicolor were represented on the microarrays. Random hexamers with 72% G – C composition were used for all the labeling reactions. For labeling of genomic DNA, 4 Ag genomic DNA and 12 Ag random hexamers (72% G – C composition) were mixed with water to obtain a volume of 23 AL. The mixture was heated at 100 -C for 2 min and then left on the benchtop for 10 min. The following components were added to a final volume of 50 AL at the given final concentrations: Cy3- or Cy5-dCTP (750 AM; Amersham Pharmacia Biotech); dATP, dTTP (40 AM each); dGTP (100 AM); dCTP (2 AM); Klenow buffer (1); and Klenow fragment (20 U; New England Biolabs catalog no. 212). The reaction was incubated at 37 -C for 2 h. 500 AL of TE (10 mM Tris pH 7.4, 1 mM EDTA pH 8.0) was added to the reaction, and the mixture was concentrated in a microcon-30 filter (Amicon) to a final volume of ¨ 6 Al. The labeled sample was applied to a microarray as described below. For labeling of RNA, 15 Ag total RNA and 5 Ag random hexamers (72% G – C composition) were mixed with water to obtain a volume of 7 AL. The mixture was heated at 75 -C for 15 min and cooled immediately on ice. The following components were added to a final volume of 15 AL, at the given final concentrations: Cy3- or Cy5-dCTP (1.7 mM; Amersham Pharmacia Biotech); dATP, dTTP (400 AM each); dGTP (1 mM); dCTP (20 AM); SuperScript II buffer (1); DTT (10 mM); and SuperScript II (400 U; Gibco BRL). The reaction was incubated at room temperature for 10 min, and then 42 -C for 2.5 h. 500 AL of TE (10 mM Tris pH 7.4, 1 mM EDTA pH 8.0) was added to the reaction, and the mixture was concentrated in a microcon-10 filter (Amicon) to a final volume of ¨ 6 Al. (This wash step was repeated for RNA labelings.) The labeled sample was applied to a microarray as described next. For each microarray experiment, the Cy3 and Cy5 samples were mixed together. SSC (final concentrations 0.6 M NaCl, 0.06 M trisodium citrate dehydrate), SDS (0.2% final concentration) and polyA (10 Ag) were added. The hybridization mixture was heated at 100 -C for 2 min and applied to a microarray. The microarray was incubated at 65 -C for 10 – 12 h and then scanned with a GenePix 4000B (Axon Instruments). Microarray data was visualized using the program Treeview (Eisen et al., 1998). 2.3. Actinorhodin measurements

2.1.5. Phosphate upshift Experiment Q was performed in SMM, which contains 0.5 mM each of NaH2PO4 and K2HPO4. After the reference sample was taken, phosphate buffer with equimolar NaH2PO4 and K2HPO4 was added to a final concentration of 30 mM.

Actinorhodin titers were measured using a previously described assay (Kieser et al., 2000). Combined cell and supernatant samples were treated with potassium hydroxide (1 N final concentration), the mixture was centrifuged and the A 640 of the supernatants was measured.

56

N. Karoonuthaisiri et al. / Gene 353 (2005) 53 – 66

3. Results 3.1. Regional gene expression during vegetative growth To measure relative levels of different S. coelicolor transcripts during vegetative growth in liquid cultures, we compared genomic DNA samples to total RNA samples using microarrays (Table 1, experiments A –D). Two S. coelicolor laboratory strains, M600 and M145, and two growth media, SMM (a supplemented minimal medium) and R5 (a rich medium), were used (Kieser et al., 2000). In experiments A – D, genomic DNA was labeled with a green fluorescent dye (Cy3) and total RNA with a red fluorescent dye (Cy5). Therefore red and green microarray spots represented highly and poorly expressed genes, respectively, relative to the average transcript level. Experiments A and B compared copy numbers of M600 transcripts in SMM and R5 medium, respectively, to gene copy numbers in the M600 chromosome. Experiments C and D compared copy numbers of M600 and M145 transcripts, respectively, in R5 medium relative to gene copy numbers in the M145 chromosome. In two control experiments, total RNA samples were directly compared using microarrays to measure relative levels of the same transcripts between two different samples (Table 1, experiments E – F). Experiment E directly compared transcript levels of M600 growing in R5 versus SMM medium. Experiment F directly compared transcript levels of M145 versus M600 growing in R5 medium. These experiments were conducted as biological triplicates (different cultures). DNA microarrays with 97% of the 7825 predicted genes of S. coelicolor M145 were used (Huang et al., 2001). The data were stored and analyzed as quantitative ratios of red:green fluorescence intensities (Gollub et al., 2003). During vegetative growth in either supplemented minimal medium or rich medium, S. coelicolor M600 and M145 expressed genes in the middle of the chromosome more highly than those in the chromosome arms (Fig. 1A and B). On average, genes in the middle and ends of the chromosome had expression differences of approximately four-fold (Fig. 1C). The regions of lower gene expression in the left and right chromosome arms correlated with the uneven 1.5 Mb and 2.3 Mb arm lengths proposed by Bentley et al. (2002). We observed this expression pattern for genomic DNA isolated from both vegetative and stationary phase cultures, indicating that higher copy numbers of DNA near the origins of replicating chromosomes do not account for higher gene expression of the core region. We also have recently shown that certain S. coelicolor strains, including M600, have a second copy of a 1.04 Mb segment of DNA from the left chromosome end inserted in the right TIR, increasing the TIRs from 22 kb to 1.06 Mb (Weaver et al., 2004). As reported, M145 and M600 expressed genes in the 1.06 Mb TIRs of M600 similarly, despite the two-fold higher copy number of this region in M600 (Fig. 1D) (Weaver et al., 2004). The

different TIR lengths (and similar gene expression) between M600 and M145 also were evident in experiments A – B, with M600 possessing lower red:green ratios for the first 1000 genes (Fig. 1A, ‘‘DNA duplicated in M600’’). In general, S. coelicolor M600 and M145 expressed few genes differently during vegetative growth in rich medium, with no genes consistently up- or down-regulated in one strain relative to the other in the biological triplicate experiments. During vegetative growth in liquid cultures, S. coelicolor expressed many genes involved in growth and primary metabolism at high levels (see Supplementary material, Figs. 2, 3A and 4). These genes encoded RNA polymerase subunits, ribosomal proteins and translation factors, nucleotide metabolism enzymes, chaperone and secretion proteins, stress response proteins (e.g., heat shock, cold shock, oxidative stress), cell division and differentiation proteins, and enzymes for nitrogen assimilation, amino acid biosynthesis, fatty acid biosynthesis and energy generation (glycolysis, TCA cycle, electron transport, ATP synthesis). In general, genes encoding known and putative regulatory genes had lower transcript levels in vegetatively growing cells than genes involved in growth and primary metabolism, although several sigma factor genes had transcript levels four- to eight-fold higher than the average (the principal sigma factor hrdB (SCO5820); the extracytoplasmic function sigma factors SCO5147, sigR (SCO5216) and sigQ (SCO4908); and the stress sigma factor sigJ/sigB (SCO0600)). Small regions of highly expressed genes, which may represent either operons or groups of functionally related genes, typically were interspersed in larger regions of genes with significantly lower expression levels. S. coelicolor expressed few genes differently during vegetative growth in rich versus supplemented minimal media. This observation is consistent with both media containing essential amino acids and trace elements and the isolation of total RNA from cultures before major nutrients were depleted. However, supplemented minimal medium contains 67 mM glycine and S. coelicolor expressed genes involved in glycine utilization more highly in this medium. S. coelicolor also expressed genes involved in fatty acid biosynthesis more highly in supplemented minimal medium, suggesting that the bacterium represses de novo fatty acid biosynthesis in rich medium. In addition, S. coelicolor expressed several genes involved in the synthesis of cofactor molecules, in particular biotin and thiamine, more highly in supplemented minimal medium relative to rich medium. 3.2. Changes in the distribution of genes expressed during stationary phase To examine how the distribution of transcript levels across the S. coelicolor chromosome varies in other growth conditions, we measured gene expression changes during the transition from vegetative growth to stationary phase. These experiments extend the microarray study by Huang et

N. Karoonuthaisiri et al. / Gene 353 (2005) 53 – 66

57

DNA: M600 M600 M145 M145 v. v. v. v. M600 M600 M600 M145

RNA: Medium: SMM R5

R5

R5

Expt.: A B C D Biol. triplicate: 123 123 123 123

A

DNA Duplicated in M600

S. coelicolor M145 chromosome

Left arm B

(8.7 Mb)

(1.5 Mb) Left arm (1.5 Mb)

C

Right arm (2.3 Mb)

Ori

DNA v. RNA - Expt. D1 (M145)

log2 (Red / Green)

6

Core oriC

(4.9 Mb)

4 2 0 -2 -4 Left arm 1.5 Mb -6 1

Right arm 2.3 Mb

Core 4.9 Mb 2

3

4

5

6

7

8

Chromosome Position (Mb)

D

M145 v. M600 - Expt. F1 (RNA)

Right arm (2.3 Mb)

log2 (Red / Green)

6 4 2 0 -2 -4 -6

Genes present in two copies in M600 1

2

3

4

5

6

7

8

Chromosome Position (Mb)

<1/8 fold

>8 fold

Fig. 1. Transcript levels across the S. coelicolor chromosome during vegetative growth in liquid medium. Two strains, S. coelicolor M145 and M600, and two growth media, a rich medium (R5) and supplemented minimal medium (SMM), were examined. (A) Genomic DNA versus total RNA comparisons. Experiments A – D compare DNA and RNA from different strains and growth media (see Table 1). F1_, F2_ and F3_ are biological triplicate experiments. The microarray data are shown on a color scale, with green, black and red representing transcript levels less than, equal to or greater than the average transcript level, respectively. Genes corresponding to the 4.9 Mb core region of the S. coelicolor M145 chromosome were found to be more highly expressed under nonlimiting growth conditions than genes in the 1.5 Mb left and 2.3 Mb right chromosome arms, supporting the core versus auxiliary organization of essential and nonessential genes proposed by Bentley et al. (2002) on the basis of sequence analysis. (B) A scheme of the linear S. coelicolor M145 chromosome. (C) A plot of relative levels of M145 transcripts in R5 medium. Microarray data from experiment D1 is shown. Red:green ratios are plotted on a log2 scale. The average transcript level is set to log2(Red/Green) = 0. On average, genes in the middle and ends of the chromosome had expression differences of approximately fourfold. (D) A plot of relative levels of the same transcripts in M145 and M600 in R5 medium. Microarray data from experiment F1 is shown. M145 and M600 strains expressed genes in the 1.06 Mb TIRs of M600 similarly, even though M600 contains this DNA at a two-fold higher copy number than M145 does (Weaver et al., 2004).

al. (2001) from S. coelicolor M145 to S. coelicolor M600. Total RNA samples from both strains were analyzed using microarrays (Table 1, experiments G – H). Cultures of M600 and M145 were grown in R5 medium to a cell density of OD450 ¨ 1, which occurred approximately 10 h and 12 h

after inoculation, respectively. From each culture, total RNA was harvested as the initial sample. From the M600 culture, additional samples were harvested at 12, 16, 20, 24, 34 and 45 h. From the M145 culture, RNA samples were harvested at 16, 20, 24, 34, 40 and 45 h. The two strains had similar

58

N. Karoonuthaisiri et al. / Gene 353 (2005) 53 – 66

growth rates and actinorhodin titers, entering stationary phase between 16 and 20 h (Fig. 2A). For both time courses, the initial RNA sample was labeled with the green fluorescent dye (Cy3) and subsequent time points with the red fluorescent dye (Cy5). Therefore red and green microarray spots represented induced and repressed genes, respectively, relative to transcript levels of the same culture at OD450 ¨ 1. As a control, transcript levels between the M145 and M600 cultures at OD450 ¨ 1 were directly compared in experiment G and H (Table 1). During the transition to stationary phase in liquid rich medium, S. coelicolor M600 and M145 showed similar changes in gene expression throughout the chromosome. For example, 22 known and predicted secondary metabolite gene clusters have been annotated in the M145 genome sequence (Bentley et al., 2002), 16 located in the chromosome arms (Fig. 2B and D) and six located in the chromosome core region (Fig. 2C). Both strains expressed these gene clusters at low levels during vegetative growth (Fig. 2, experiments C – D). During stationary phase, the strains induced the gene clusters of the three known chromosomally encoded antibiotics (CDA, actinorhodin and prodiginines), the nonribosomal peptide coelichelin (Challis and Ravel, 2000), desferrioxamines (Challis and Hopwood, 2003) and the butyrolactones (Takano et al., 2001), as well as gene clusters of a putative deoxysugar synthase and a type I polyketide synthase (PKS) (Fig. 2, experiments G –H). Thus S. coelicolor may produce up to eight secondary metabolites during stationary phase in liquid rich medium. The lack of expression of the other clusters may indicate functions during growth conditions other than stationary phase, e.g., the putative unsaturated fatty acid eicosapentaenoic acid for conditions that require remodeling of the cell membrane, the putative carotenoid isorenieratene for conditions that require protection from radiation or the coelibactin genes for iron limitation (Challis and Hopwood, 2003). The lack of expression of the whiE spore pigment cluster (Yu and Hopwood, 1995) is consistent with the inability of S. coelicolor liquid cultures to sporulate. The secondary metabolite gene clusters showed similar times of induction between S. coelicolor M600 and M145, with the exception of the actinorhodin gene cluster (Fig. 2B, C and D, graphs). M600 expressed the actinorhodin genes after 12 h of growth, whereas M145 expressed these genes after 24 h of growth. The cause of this difference in actinorhodin gene regulation remains to be determined. Fig. 3 shows a comparison of vegetative transcript levels and stationary phase expression changes for segments of the left chromosome arm, chromosome core region and right chromosome arm. The segment shown of the left chromosome arm contains few highly expressed genes during vegetative growth, but contains genes induced during stationary phase, e.g., two secondary metabolite clusters, hypothetical proteins and putative enzymes (Fig. 3A). Similarly, the segment shown of the right chromosome arm contains few abundant vegetative

transcripts and genes induced during stationary phase, e.g., one conservon (cvn) cluster (see below) and a cluster with a putative regulator and hypothetical proteins (Fig. 3C). In contrast, the segment shown of the chromosome core region contains many transcripts abundant during vegetative growth, such as those involved in transcription, translation and primary metabolism; S. coelicolor repressed these genes during stationary phase (Fig. 3B). (Note that, after gene repression, some highly transcribed genes, in particular those encoding ribosomal proteins, still had high transcript levels relative to other induced genes.) This chromosome segment also contains genes with low transcriptional activity during vegetative growth, which then were induced during stationary phase, e.g., several clusters of putative enzymes and hypothetical proteins, a nitrite reductase cluster and the gene cluster for the S. coelicolor antibiotic actinorhodin (Fig. 3B). In general, genes repressed during stationary phase had chromosomal positions in the core region, whereas induced genes had positions in both the chromosome core and arm regions. Of several hundred genes poorly expressed during vegetative growth and induced during stationary phase, approximately 40% had positions in the chromosome arms (see Supplementary material, Fig. 1A and C). Over half of these genes had unknown putative functions, and this class of genes was enriched with putative secreted proteins. Of several hundred repressed genes highly expressed during vegetative growth and repressed during stationary phase, approximately 8% had positions in the chromosome arms (see Supplementary material, Fig. 2A and C). Less than a third of these genes had unknown putative functions, and the remainder had roles in growth and metabolism. In contrast to these genes, S. coelicolor M600 and M145 induced few genes during stationary phase entry that already had high expression levels during vegetative growth in liquid rich medium (see Supplementary material, Fig. 3A) and repressed few genes that already had low expression levels during vegetative growth (see Supplementary material, Fig. 3B). Genes with abundant transcripts during both vegetative growth and stationary phase (i.e., unchanged expression after stationary phase entry) also primarily had chromosome core positions and included genes involved in cell maintenance (the major vegetative sigma factor hrdB, groES/EL chaperones, ferredoxin, carbon and amino acid metabolism) (see Supplementary material, Fig. 4). Different transcript levels were observed for homologs within gene families, e.g., ribosomal protein genes (see Supplementary material, Fig. 5) and the 13 uncharacterized conservon (cvn) clusters (see Supplementary material, Fig. 6) (Bentley et al., 2002). 3.3. A gene expression response common to temperature and osmotic upshifts To examine the chromosomal distribution of genes induced under different stress conditions, we measured

N. Karoonuthaisiri et al. / Gene 353 (2005) 53 – 66 Fig. 2. Gene expression changes of 22 known and putative secondary metabolite gene clusters (Bentley et al., 2002) during stationary phase in liquid rich medium. (A) Growth curves and actinorhodin production of the S. coelicolor M600 and M145 cultures. Secondary metabolite clusters in the (B) left chromosome arm, (C) chromosome core region and (D) right chromosome arm. In experiments C – D, red, green and black represent genes with high, low and average transcript levels during vegetative growth (see Table 1 and Fig. 1). In experiment ‘‘G,H’’, red, green and black represent genes with higher, lower and similar transcript levels, respectively, in 12 h cultures (OD450 ¨1) of S. coelicolor M600 relative to S. coelicolor M145. In experiments G – H, red, green and black represent gene induction, gene repression and unchanged expression, respectively, relative to transcript levels in the same culture at 12 h for S. coelicolor M600 (G) and S. coelicolor M145 (H) (see Table 1). Gray represents no microarray data. The black bars denote genes with adjacent positions in the chromosome. For each gene, F+_ and F F denote the DNA strand encoding the open reading frame. During stationary phase, the strains induced the gene clusters of the three known chromosomally encoded antibiotics (CDA, actinorhodin and prodiginines), the nonribosomal peptide coelichelin, desferrioxamines, the butyrolactones, a putative deoxysugar synthase and a type I polyketide synthase (PKS). Thus S. coelicolor may produce up to eight secondary metabolites during stationary phase in liquid rich medium. 59

60

N. Karoonuthaisiri et al. / Gene 353 (2005) 53 – 66

Fig. 3. A comparison of vegetative transcript levels and stationary phase expression changes for segments of the chromosome core and arm regions. 680 genes from the (A) left chromosome arm (SCO0126 – SCO0805), (B) chromosome core region (SCO4643 – SCO5322) and (C) right chromosome arm (SCO6608 – SCO7287) are shown. See Fig. 2 legend for details. During stationary phase entry, S. coelicolor down-regulates transcripts highly abundant during vegetative growth and induces genes poorly expressed during vegetative growth. These data are analogous to those reported by Huang et al. (2001).

gene expression changes after the application of shifts to vegetative cultures of S. coelicolor M600. Eight different stress conditions were examined in liquid minimal and rich medium (Table 1, experiments I –V). The stress conditions included a temperature downshift, three temperature upshifts, an osmotic upshift (with sucrose as an uncharged and noncatabolized osmolyte; Kieser et al., 2000), an osmotic downshift (with sucrose as the osmolyte), an ethanol upshift and a phosphate upshift. Three control experiments were conducted, in which no shift was applied (Table 1, experiments I and R) or the culture was transferred to fresh medium (Table 1, experiment J). In

experiments I –U, M600 cultures were grown to a cell density of OD450 ¨ 0.5 and total RNA samples were harvested before and 15, 30, 45, 60 and 90 min after the shift. In experiment V, RNA samples were harvested 5, 10, 15 and 30 min after the shift. For each time course, the initial RNA sample was labeled with the green fluorescent dye (Cy3) and later samples with the red fluorescent dye (Cy5). Therefore red and green microarray spots represented induced and repressed genes, respectively, relative to M600 transcript levels at OD450 ¨0.5. The following discussion focuses on the temperature and sucrose upshift experiments, given the number of replicates conducted for

N. Karoonuthaisiri et al. / Gene 353 (2005) 53 – 66

the temperature upshifts (three time courses in SMM and two time courses in R5) and the common expression changes observed for the two stress conditions. We did not observe a gene expression response shared by all the stress conditions examined here, which is consistent with a previous proteomic study that did not find a common set of S. coelicolor proteins induced by different stress conditions (Vohradsky et al., 2000). In contrast to the expression of common stress regulons by one or a few sigma factors in E. coli and Bacillus subtilis, the large number of S. coelicolor sigma factors (65) may control the expression of different genes under different stress conditions. The transient induction of several hundred genes located in the chromosome core and arm regions characterized the S. coelicolor M600 responses to the temperature and sucrose upshifts (Fig. 4). The transcript levels of these genes increased within 15 min after the temperature upshifts and returned to their previous levels by 30 min (Fig. 4; see experiment V in particular), consistent with a previous observation that S. coelicolor synthesizes certain heat shock proteins within the first hour after a 37 -C shift (Puglia et al., 1995). This induction profile resembled the transient induction of genes after exposure to oxidative stress (Paget et al., 2001) and the cell wall antibiotic vancomycin (Hong et al., 2002). S. coelicolor induced the genes later after the sucrose upshift, at approximately one hour (Fig. 4, experiment O). At subsequent growth stages, the transcript levels of these genes returned to previous levels (Karoonuthaisiri, unpublished data). S. coelicolor transiently induced a subset of these genes after the phosphate upshift (Fig. 4, experiment Q). The transiently induced genes had four general classes: (1) Genes with previously characterized roles in heat shock responses, including lon (SCO5285), which encodes the Lon protease, and a cluster that encodes the DnaK chaperone and ClpB protease (SCO3660 – SCO3671) (Fig. 4, blue text). These seven genes comprise three S. coelicolor transcripts previously identified as upregulated in a HspR mutant and during growth on solid medium at 42 -C (Bucca et al., 2003). The microarray data here suggested additional levels of regulation for these genes: S. coelicolor did not induce the seven heat shock genes after the phosphate upshift (Fig. 4B) and extended the expression of these genes after both the 42 -C and ethanol upshifts to ¨ 60 min (Fig. 4D), while still transiently expressing other genes after the phosphate and 42 -C shifts. A specific cellular defect common to the 42 -C and ethanol upshifts, such as extensive protein denaturation, may lengthen the expression of the heat shock genes. (2) Sigma factors, anti-sigma factors and anti-sigma factor antagonists (Fig. 4, red text). The transiently induced genes included five of nine S. coelicolor homologs of

61

the B. subtilis sigma factor sigB, the latter which regulates stress responses in B. subtilis: sigH (SCO5243), sigJ/sigB (SCO0600), sigK (SCO6520), sigL (SCO7278) and sigM (SCO7314). These sigma factors have previously been implicated in S. coelicolor stress responses (Kormanec et al., 2000; Cho et al., 2001; Kelemen et al., 2001; Viollier et al., 2003) and their collective induction after temperature and osmotic upshifts may reflect their participation in a regulatory system that prepares S. coelicolor for multiple stresses. The transiently induced genes included ssgB (SCO1541) but not gltB (SCO2026), genes which have promoters transcribed in vitro by RNA polmyerase containing jH (Kormanec and Sevcikova, 2002a,b). The transiently induced genes did not include catB (SCO0666), which has a promoter transcribed in vitro by RNA polmyerase containing jB (Cho et al., 2001). Thus RNA polymerase holoenzymes containing another sigma factor may transcribe ssgB, gltB and catB in vivo at short times after temperature and osmotic upshifts, as different sigma factors likely regulate promoters in the sigR regulon with multiple transcription start sites (Paget et al., 2001). After the sucrose upshift, S. coelicolor synthesizes transcripts of sigI (SCO3068), sigJ/sigB, sigK, sigM, sigH and sigL in a sequential order (Fig. 4E), which is consistent with the asynchronous synthesis of the SigI (SCO3068), SigJ/SigB and SigH proteins reported after an osmotic upshift (Fig. 4E) (Viollier et al., 2003). (Note that S. coelicolor does not increase sigI transcript levels after temperature upshifts.) Together, these data suggest that S. coelicolor may have an alternative sigma factor responsible for the transient induction of the genes identified here, and the bacterium may activate this sigma factor at early times after temperature upshifts and at later times after an osmotic upshift. Furthermore, an osmotic upshift may initiate a sigma factor cascade that ultimately activates the alternative sigma factor, perhaps analogous to the sigma factor cascade that governs sporulation in B. subtilis (Piggot and Losick, 2002). (3) Genes involved in carbohydrate metabolism, including sugar transferases, transporters and enzymes (Fig. 4, orange text). These genes may increase the synthesis of carbohydrates to stabilize proteins and increase intracellular osmolarity after temperature and osmotic upshifts, respectively (Singer and Lindquist, 1998). The microarray data suggested multiple levels of regulation for one or both of the highly similar gene clusters encoding trehalose synthase and alpha-amylase (SCO5442 –SCO5443, SCO7334 – SCO7335) (Schneider et al., 2000), which have sustained induction during stationary phase and transient induction after the temperature and sucrose upshifts (Fig. 4F).

62

N. Karoonuthaisiri et al. / Gene 353 (2005) 53 – 66

Shift Conditions

(RNA v. RNA)

Vegetative Growth

SMM

R5

SMM

(OD~1)

(DNA v. RNA) 35C

37C

42C

1 M 30 mM 37C

37C

37C

30C

30C

30C

0 M 0 mM 30C

30C

30C

L

M

N

D Expt. C Biol. trip. 123 123

O

Q

S

S

1530456090 1530456090 1530456090 1530456090 1530456090 1530456090 1530456090

V 5101530

<1/8 fold

>8 fold

Expt. Min. after shift

Left Chromosome Arm

A

SCO0230 + SCO0231 SCO0596 SCO0597 SCO0598 SCO0599 SCO0600 SCO0601

tetR-family regulator hypothetical protein

+ + + + -

dpsA DNA-binding protein conserved hypothetical protein arsJ anti-sigma factor antagonist prsJ anti-sigma factor sigJ sigma factor membrane protein

SCO0616 SCO0617 -

conserved hypothetical protein conserved hypothetical protein

SCO0678 SCO0679 + SCO0680 -

hypothetical protein hypothetical protein transmembrane efflux protein

SCO0697 + SCO0698 -

AraC-family regulator hypothetical protein

SCO0718 SCO0719 SCO0720 SCO0721 SCO0722 SCO0723

+ +

hypothetical protein conserved hypothetical protein membrane protein glycosyl transferase hydrolase conserved hypothetical protein

SCO0775 SCO0776 SCO0777 SCO0778 SCO0779 SCO0780 SCO0781 SCO0782 SCO0785 SCO0786 SCO0787

+ conserved hypothetical protein membrane protein + onserved hypothetical protein + membrane protein conserved hypothetical protein + zinc-binding oxidoreductase + anti-sigma factor antagonist - prsA ribose-phosphate pyrophosphokinase membrane protein + membrane protein + secreted hydrolase

SCO0830 SCO0831 SCO0832 SCO0833 SCO0834 SCO0835 SCO0836

-

SCO0961 SCO1053 SCO1225 SCO1373 SCO1388

- glgC glucose-1-phosphate adenylyltransferase hypothetical protein + osmoprotectant transporter conserved hypothetical protein mannose-1-phosphate guanyltransferase

penicillin-binding protein conserved hypothetical protein conserved hypothetical protein hypothetical protein hypothetical protein methyltransferase conserved hypothetical protein

Chromosome Core Region

B

SCO1541 + ssgB regulator SCO2264 + membrane protein SCO2315 + membrane protein SCO2341 SCO2342 -

hypothetical protein secreted protein

SCO2493 -

membrane protein

SCO2649 SCO2650 -

4-alpha-glucanotransferase secreted protein

SCO2886 SCO2887 SCO2888 SCO2889

hydrolase membrane protein conserved hypothetical protein sugar transport protein

+ -

SCO3158 - ssgE hypothetical protein SCO3410 + membrane protein SCO3660 SCO3661 SCO3662 SCO3663 SCO3664 SCO3665 SCO3666 SCO3667 SCO3668 SCO3669 SCO3670 SCO3671

+ + -

SCO3802 SCO3857 SCO4093 SCO4118 SCO4290 SCO4693 SCO4799 SCO4846 SCO4903 SCO4982 SCO5243

+ + - otsA + + + - sigH

clpB

hspR dnaJ grpE dnaK

SCO5285 + lon SCO5286 -

conserved hypothetical protein protease ATP-binding subunit hypothetical protein membrane protein regulator regulator regulator secreted solute-binding protein heat shock protein molecular chaperone heat shock protein heat shock protein 70 membrane protein regulator membrane protein tetR-family regulator trehalose phosphate synthase membrane protein secreted lipase membrane protein membrane protein membrane protein sigma factor ATP-dependent protease secreted hydrolase

SCO5442 - treSI trehalose synthase SCO5443 - pep1I alpha-amylase

Fig. 4. Transient induction of genes after temperature and osmotic upshifts in the (A) left chromosome arm, (B) chromosome core region and (C) and right chromosome arm. In experiments C – D, red, green and black represent genes with high, low and average transcript levels during vegetative growth in liquid rich medium (see Table 1 and Fig. 1). In experiments L – V, red, green and black represent gene induction, gene repression and unchanged expression, respectively, relative to transcript levels at OD450 ¨0.5 of S. coelicolor M600 liquid cultures (see Table 1). Gray represents no microarray data. After temperature and osmotic upshifts, S. coelicolor induces a common set of genes transiently across the chromosome. (D) Transient induction of heat shock genes after 35 -C and 37 -C upshifts and prolonged induction of the same genes after 42 -C and ethanol upshifts. (E) Transient induction of sigma factor homologs of the B. subtilis sigma factor sigB, which regulates stress responses in B. subtilis, after temperature upshifts and serial induction of the same genes after a sucrose upshift. (F) Sustained induction of trehalose synthase and alpha-amylase genes after entry into stationary phase (see Fig. 2 legend for details) and transient induction of the same genes after temperature upshifts.

N. Karoonuthaisiri et al. / Gene 353 (2005) 53 – 66

63

Shift Conditions

(RNA v. RNA)

Vegetative Growth

SMM

R5

SMM

(OD~1)

(DNA v. RNA) 35C

37C

42C

1 M 30 mM 37C

37C

37C

30C

30C

30C

0 M 0 mM 30C

30C

30C

L

M

N

D Expt. C Biol. trip. 123 123

O

Q

S

S

1530456090 1530456090 1530456090 1530456090 1530456090 1530456090 1530456090

V 5101530

<1/8 fold Expt. Min. after shift

Right Chromosome Arm

C

SCO5998 - murA2 UDP-N-acetylglucosamine transferase SCO6014 + cationic amino acid transporter SCO6062 + SCO6063 + SCO6064 + SCO6494 SCO6495 SCO6496 SCO6497 SCO6498 SCO6499 SCO6500 SCO6501 SCO6502 SCO6503 SCO6504 SCO6505 SCO6506 SCO6507 SCO6508 SCO6509 SCO6510 SCO6511 SCO6512 SCO6513 SCO6514 SCO6515 SCO6516

D

>8 fold

SMM

35C

37C

42C

30C

30C

30C

L

M

N

1530456090 1530456090 1530456090

E

5% EtOH 0 O 1530456090

Expt. Min. after shift

-

Stationary Phase

SMM

H

12 16 20 24 34 45 16 20 24 34 40 45

G, H

G

35C

37C

42C

30C

30C

30C

L

M

N

1530456090 1530456090 1530456090

gvpJ gvpL gvpS gvpK

pfpI

SCO6520 + sigK

sigma factor

SCO6721 SCO6722 SCO6723 SCO6724

+ - ssgD + +

membrane protein regulator oxidoreductase hypothetical protein

SCO7041 SCO7104 SCO7186 SCO7205 SCO7210

+ + + + +

hypothetical protein ECF sigma factor membrane protein hydrolase conserved hypothetical protein

SCO7277 + SCO7278 - sigL

regulator sigma factor

SCO7287 SCO7288 +

hypothetical protein membrane protein

SCO7313 SCO7314 SCO7315 SCO7316 SCO7317 SCO7318 SCO7319

conserved hypothetical protein sigma factor conserved hypothetical protein membrane protein hypothetical protein membrane protein oxidoreductase

+ + sigM + + + -

SCO7325 SCO7328 SCO7331 -

anti-sigma factor antagonist regulator conserved hypothetical protein

SCO7334 - treSII SCO7335 - pep1II

trehalose synthase alpha-amylase

SCO7378 SCO7381 + SCO7393 -

hypothetical protein hypothetical protein lipoprotein

SCO7412 + SCO7413 + SCO7414 -

pyruvate dehydrogenase oxidoreductase membrane protein

SCO7442 + SCO7443 + pgm

hypothetical protein phosphoglucomutase

SCO7590 - katA2

catalase

SCO7646 + SCO7647 +

secreted protein calcium binding protein

SCO7747 SCO7748 SCO7749 SCO7750 SCO7752 SCO7753 SCO7754 SCO7755 SCO7756 SCO7757

conserved hypothetical protein conserved hypothetical protein hypothetical protein conserved hypothetical protein membrane protein hypothetical protein anti-sigma factor antagonist conserved hypothetical protein hypothetical protein transposase

+ + + + + + + -

SMM

35C

37C

42C

1 M Sucrose

30C

30C

30C

0M

L

M

N

upshift

O 1530456090

Expt. Min. after shift + + + + -

sigI sigJ / sigB sigK sigM sigH sigL

SCO4034 - sigN SCO4035 - sigF SCO7341 - sigG

(time course)

(RNA v. RNA)

gvpO gvpA gvpF gvpG

membrane protein dehydrogenase dehydrogenase transketolase conserved hypothetical protein gas vesical synthesis protein gas vesical synthesis protein gas vesical synthesis protein gas vesical synthesis protein hypothetical protein conserved hypothetical protein gas vesical synthesis protein gas vesical synthesis protein gas vesical synthesis protein gas vesical synthesis protein hydrophobic protein conserved hypothetical protein conserved hypothetical protein ABC transporter ATP-binding protein conserved hypothetical protein hypothetical protein protease hypothetical protein

SCO3068 SCO0600 SCO6520 SCO7314 SCO5243 SCO7278

hspR dnaJ grpE dnaK

SCO5285 + lon

F

tkt2

1530456090 1530456090 1530456090

SCO3660 SCO3661 - clpB SCO3668 SCO3669 SCO3670 SCO3671

+ + + + + + + + + + + + + + + + -

ABC transporter ATP-binding subunit ABC transporter permease ABC transporter permease

Expt. Min. after shift SCO5442 - treSI SCO5443 - pep1I SCO7334 - treSII SCO7335 - pep1II

Fig. 4 (continued).

64

N. Karoonuthaisiri et al. / Gene 353 (2005) 53 – 66

(4) Genes with unknown functions, e.g., putative enzymes, membrane proteins, secreted proteins and hypothetical proteins. The transiently induced genes also included one of the two clusters of putative gas vesicle synthesis genes (SCO6494 – SCO6516) (Fig. 4C). The transiently induced genes did not include 10 additional genes previously found to be upregulated in a HspR mutant (Bucca et al., 2003), which may reflect differences in gene regulation at short times after a temperature upshift in a liquid medium (this study) compared to gene regulation in a mutant strain growing vegetatively on a solid medium at the normal physiological temperature (30 -C) (Bucca et al., 2003).

4. Discussion During vegetative growth in minimal and rich media, genes corresponding to the 4.9 Mb core region of the S. coelicolor M145 chromosome were found to be more highly expressed than genes in the 1.5 Mb left and 2.3 Mb right chromosome arms. This observation experimentally supports a genome organization, proposed by Bentley et al. (Bentley et al., 2002) from gene ontology considerations, in which the chromosome core encodes gene functions essential for growth while the chromosome arms encode Fauxiliary_ functions required under other growth conditions. This functional organization of gene expression contrasts with the periodic patterns of transcriptional activity observed for the E. coli chromosome, which have been proposed to be caused by DNA structures such as solenoids and supercoils (Jeong et al., 2004; Kepes, 2004; Valens et al., 2004). The spatial separation of Fcore_ and Fauxiliary_ genes in Streptomyces chromosomes may facilitate the evolution of new auxiliary functions through genetic instability, since core genes with essential functions would avoid amplifications and deletions that occur in the chromosome arms. The low levels of gene expression observed for the chromosome arms also are consistent with genetic instability (since large deletions and amplifications of unexpressed genes would not be expected to impair growth) and a lesser need for specialized cell functions of soil environments in the culture conditions examined (such as the secretion of hydrolytic enzymes to degrade complex debris and antibiotic production to defend against competing microorganisms). The microarray data have suggested several additional characteristics of transcriptional activity in S. coelicolor during vegetative growth. (a) The genes most highly expressed in rapidly growing cells are predicted to encode proteins primarily involved in growth and primary metabolism. Most regulatory genes have low expression levels, presumably because only low concentrations of the corresponding proteins are required for regulatory functions such as DNA binding. (b) S. coelicolor expresses many

contiguous sets of genes in its chromosome which likely are operons. These data provide additional evidence for operons or co-regulated genes predicted by computer calculations, biochemical experiments and previous microarray analysis (Huang et al., 2001; Bucca et al., 2003). (c) S. coelicolor expresses similar genes during vegetative growth in rich and supplemented minimal medium, indicating that both media are fully supplemented with essential nutrients and that the cultures had not exhausted major nutrients at the time of analysis. (d) S. coelicolor strains containing short (22 kb) versus long (1.06 Mb) TIRs (Weaver et al., 2004) express most genes to about the same extent during vegetative growth in rich medium, implying that S. coelicolor possesses mechanisms to maintain homeostasis in gene expression for large tracts of DNA present at higher copy numbers. Since subtle physiological differences are known to exist for these prototrophic strains, such as the timing of expression of the actinorhodin gene cluster (Bystrykh et al., 1996), more significant gene expression differences may occur under specific environmental conditions. The examination of other growth conditions revealed additional characteristics of gene expression in S. coelicolor. (a) During stationary phase entry, S. coelicolor down-regulates large quantities of growth-related transcripts encoded in the chromosome core and up-regulates transcripts of Fauxiliary_ genes encoded throughout the chromosome, which may be essential for differentiation and other physiological responses employed by streptomycetes. Thus the composition of S. coelicolor transcripts appears to shift from transcripts concentrated in the core region to transcripts distributed across the chromosome. (b) After temperature upshifts and an osmotic upshift with sucrose as the osmolyte, S. coelicolor transiently induces several hundred genes located in the chromosome core and arm regions. These genes encode heat shock proteins and carbohydrate metabolism enzymes, suggesting that the two stress conditions induce similar defects in S. coelicolor cells (e.g., protein denaturation) or can be countered by similar mechanisms (e.g., synthesis of carbohydrates to stabilize proteins or increase intracellular osmolarity) (Singer and Lindquist, 1998). The collective induction of sigma factors and their regulators suggests the existence of a regulatory system that prepares S. coelicolor for multiple stresses, and co-expression of the transiently induced genes may indicate the presence an alternative sigma factor dedicated to transcribing these genes. The S. coelicolor DnaK and HspR proteins have been shown to negatively regulate the operon encoding their cognate genes (Bucca et al., 2003) and may regulate additional genes identified in this study. Whether DnaK and HspR regulate the kinetics of the gene expression responses observed here after temperature upshifts, or control the induction of genes after an osmotic upshift, remains to be determined. Sigma factors and associated regulatory proteins have been characterized for several other S. coelicolor sensing

N. Karoonuthaisiri et al. / Gene 353 (2005) 53 – 66

systems (e.g., oxidative stress (Paget et al., 1999; Li et al., 2003; Bae et al., 2004), NADH/NAD+ levels (Brekasis and Paget, 2003), cell wall antibiotics (Hong et al., 2004)) and may provide models for the sensing system that detects and responds to the as-yet unknown signal that transiently appears after temperature and osmotic upshifts.

Acknowledgments We thank David Hopwood, Stephen Bentley and Robert McDaniel for helpful comments on the manuscript and Kevin Pan and Stephen Bentley for information on S. coelicolor gene positions. This work was supported in part by NSF grant BES-0093900-001 and NIH grant GM6547002 to C.M.K. and by NIH grant AI-08619 to S.N.C.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.gene.2005. 03.042.

References Bae, J.B., Park, J.H., Hahn, M.Y., Kim, M.S., Roe, J.H., 2004. Redoxdependent changes in RsrA, an anti-sigma factor in Streptomyces coelicolor: zinc release and disulfide bond formation. J. Mol. Biol. 335, 425 – 435. Bentley, S.D., et al., 2002. Complete genome sequence of the model actinomycete Streptomyces coelicolor A3(2). Nature 417, 141 – 147. Brekasis, D., Paget, M.S., 2003. A novel sensor of NADH/NAD+ redox poise in Streptomyces coelicolor A3(2). EMBO J. 22, 4856 – 4865. Bucca, G., Brassington, A.M., Hotchkiss, G., Mersinias, V., Smith, C.P., 2003. Negative feedback regulation of dnaK, clpB and lon expression by the DnaK chaperone machine in Streptomyces coelicolor, identified by transcriptome and in vivo DnaK-depletion analysis. Mol. Microbiol. 50, 153 – 166. Bystrykh, L.V., Fernandez-Moreno, M.A., Herrema, J.K., Malpartida, F., Hopwood, D.A., Dijkhuizen, L., 1996. Production of actinorhodinrelated ‘‘blue pigments’’ by Streptomyces coelicolor A3(2). J. Bacteriol. 178, 2238 – 2244. Challis, G.L., Hopwood, D.A., 2003. Synergy and contingency as driving forces for the evolution of multiple secondary metabolite production by Streptomyces species. Proc. Natl. Acad. Sci. U. S. A. 100, 14555 – 14561. Challis, G.L., Ravel, J., 2000. Coelichelin, a new peptide siderophore encoded by the Streptomyces coelicolor genome: structure prediction from the sequence of its non-ribosomal peptide synthetase. FEMS Microbiol. Lett. 187, 111 – 114. Chen, C.W., Huang, C.H., Lee, H.H., Tsai, H.H., Kirby, R., 2002. Once the circle has been broken: dynamics and evolution of Streptomyces chromosomes. Trends Genet. 18, 522 – 529. Cho, Y.H., Lee, E.J., Ahn, B.E., Roe, J.H., 2001. SigB, an RNA polymerase sigma factor required for osmoprotection and proper differentiation of Streptomyces coelicolor. Mol. Microbiol. 42, 205 – 214. Eisen, M.B., Spellman, P.T., Brown, P.O., Botstein, D., 1998. Cluster analysis and display of genome-wide expression patterns. Proc. Natl. Acad. Sci. U. S. A. 95, 14863 – 14868.

65

Gollub, J., et al., 2003. Microarray Database: data access and quality assessment tools. Nucleic Acids Res. 31, 94 – 96. Hodgson, D.A., 2000. Primary metabolism and its control in streptomycetes: a most unusual group of bacteria. Adv. Microb. Physiol. 42, 47 – 238. Hong, H.J., Paget, M.S., Buttner, M.J., 2002. A signal transduction system in Streptomyces coelicolor that activates the expression of a putative cell wall glycan operon in response to vancomycin and other cell wallspecific antibiotics. Mol. Microbiol. 44, 1199 – 1211. Hong, H.J., Hutchings, M.I., Neu, J.M., Wright, G.D., Paget, M.S., Buttner, M.J., 2004. Characterization of an inducible vancomycin resistance system in Streptomyces coelicolor reveals a novel gene (vanK) required for drug resistance. Mol. Microbiol. 52, 1107 – 1121. Horak, C.E., Snyder, M., 2002. Global analysis of gene expression in yeast. Funct. Integr. Genomics 2, 171 – 180. Huang, J., Lih, C.J., Pan, K.H., Cohen, S.N., 2001. Global analysis of growth phase responsive gene expression and regulation of antibiotic biosynthetic pathways in Streptomyces coelicolor using DNA microarrays. Genes Dev. 15, 3183 – 3192. Ikeda, H., et al., 2003. Complete genome sequence and comparative analysis of the industrial microorganism Streptomyces avermitilis. Nat. Biotechnol. 21, 526 – 531. Jeong, K.S., Ahn, J., Khodursky, A.B., 2004. Spatial patterns of transcriptional activity in the chromosome of Escherichia coli. Genome Biol. 5, R86. Kelemen, G.H., Viollier, P.H., Tenor, J., Marri, L., Buttner, M.J., Thompson, C.J., 2001. A connection between stress and development in the multicellular prokaryote Streptomyces coelicolor A3(2). Mol. Microbiol. 40, 804 – 814. Kepes, F., 2004. Periodic transcriptional organization of the E. coli genome. J. Mol. Biol. 340, 957 – 964. Kieser, T., Bibb, M.J., Buttner, M.J., Chater, K.F., Hopwood, D.A., 2000. Practical Streptomyces genetics. John Innes Foundation, Norwich, England. Kormanec, J., Sevcikova, B., 2002a. The stress-response sigma factor sigma(H) controls the expression of ssgB, a homologue of the sporulation-specific cell division gene ssgA, in Streptomyces coelicolor A3(2). Mol. Genet. Genomics 267, 536 – 543. Kormanec, J., Sevcikova, B., 2002b. Stress-response sigma factor sigma(H) directs expression of the gltB gene encoding glutamate synthase in Streptomyces coelicolor A3(2). Biochim. Biophys. Acta 1577, 149 – 154. Kormanec, J., Sevcikova, B., Halgasova, N., Knirschova, R., Rezuchova, B., 2000. Identification and transcriptional characterization of the gene encoding the stress-response sigma factor sigma(H) in Streptomyces coelicolor A3(2). FEMS Microbiol. Lett. 189, 31 – 38. Li, W., Bottril, A.R., Bibb, M.J., Buttner, M.J., Paget, M.S., Kleanthous, C., 2003. The Role of zinc in the disulphide stress-regulated anti-sigma factor RsrA from Streptomyces coelicolor. J. Mol. Biol. 333, 461 – 472. Paget, M.S., Leibovitz, E., Buttner, M.J., 1999. A putative two-component signal transduction system regulates sigmaE, a sigma factor required for normal cell wall integrity in Streptomyces coelicolor A3(2). Mol. Microbiol. 33, 97 – 107. Paget, M.S., Molle, V., Aharonowitz, Y., Buttner, M.J., 2001. Defining the disulphide stress response in Streptomyces coelicolor A3(2): identification of the sigmaR regulon. Mol. Microbiol. 42, 1007 – 1020. Piggot, P.J., Losick, R., 2002. Sporulation genes and intercompartmental regulation. In: Sonenshein, A.L., Hoch, J.A., Losick, R. (Eds.), Bacillus subtilis and its Closest Relatives: from Genes to Cells. ASM Press, Washington, D.C., pp. 483 – 518. Puglia, A.M., Vohradsky, J., Thompson, C.J., 1995. Developmental control of the heat-shock stress regulon in Streptomyces coelicolor. Mol. Microbiol. 17, 737 – 746. Schneider, D., Bruton, C.J., Chater, K., 2000. Duplicated gene clusters suggest an interplay of glycogen and trehalose metabolism during sequential stages of aerial mycelium development in Streptomyces coelicolor A3(2). Mol. Gen. Genet. 263, 543 – 553.

66

N. Karoonuthaisiri et al. / Gene 353 (2005) 53 – 66

Singer, M.A., Lindquist, S., 1998. Multiple effects of trehalose on protein folding in vitro and in vivo. Mol. Cell 1, 639 – 648. Takano, E., Chakraburtty, R., Nihira, T., Yamada, Y., Bibb, M.J., 2001. A complex role for the gamma-butyrolactone SCB1 in regulating antibiotic production in Streptomyces coelicolor A3(2). Mol. Microbiol. 41, 1015 – 1028. Valens, M., Penaud, S., Rossignol, M., Cornet, F., Boccard, F., 2004. Macrodomain organization of the Escherichia coli chromosome. EMBO J. 23, 4330 – 4341. Viollier, P.H., Kelemen, G.H., Dale, G.E., Nguyen, K.T., Buttner, M.J., Thompson, C.J., 2003. Specialized osmotic stress response systems involve multiple SigB-like sigma factors in Streptomyces coelicolor. Mol. Microbiol. 47, 699 – 714.

Vohradsky, J., et al., 2000. Developmental control of stress stimulons in Streptomyces coelicolor revealed by statistical analyses of global gene expression patterns. J. Bacteriol. 182, 4979 – 4986. Weaver, D., et al., 2004. Genome plasticity in Streptomyces: identification of 1 Mb TIRs in the S. coelicolor A3(2) chromosome. Mol. Microbiol. 51, 1535 – 1550. Yu, T.W., Hopwood, D.A., 1995. Ectopic expression of the Streptomyces coelicolor while genes for polyketide spore pigment synthesis and their interaction with the act genes for actinorhodin biosynthesis. Microbiology 141, 2779 – 2791.