JOURNAL OF VIROLOGY, Aug. 2006, p. 8081–8088 0022-538X/06/$08.00⫹0 doi:10.1128/JVI.00065-06 Copyright © 2006, American Society for Microbiology. All Rights Reserved.
Vol. 80, No. 16
Global Changes in Cellular Gene Expression during Bacteriophage PRD1 Infection† Minna M. Poranen,1,2 Janne J. Ravantti,1,2 A. Marika Grahn,1,2‡ Rashi Gupta,1 Petri Auvinen,1 and Dennis H. Bamford1,2* Institute of Biotechnology1 and Department of Biological and Environmental Sciences,2 Viikki Biocenter, P.O. Box 56 (Viikinkaari 5), 00014 University of Helsinki, Helsinki, Finland Received 10 January 2006/Accepted 30 May 2006
Virus-induced changes in cellular gene expression and host physiology have been studied extensively. Still, there are only a few analyses covering the entire viral replication cycle and whole-host gene pool expression at the resolution of a single gene. Here we report changes in Escherichia coli gene expression during bacteriophage PRD1 infection using microarray technology. Relative mRNA levels were systematically measured for over 99% of the host open reading frames throughout the infection cycle. Although drastic modifications could be detected in the expression of individual genes, global changes at the whole-genome level were moderate. Notably, the majority of virus-induced changes took place only after the synthesis of virion components, indicating that there is no major reprogramming of the host during early infection. The most highly induced genes encoded chaparones and other stress-inducible proteins. genome (21, 41), or transcriptional profiles of hosts parasitized by lysogenic phages (13). In spite of the potential of the array technology, to our knowledge, only a single study (42) of virusinduced changes on host gene expression covers the entire host genome. To undertake a genome-wide analysis of a bacterial host under phage infection at a resolution from a single gene to the entire genome, we studied changes in Escherichia coli gene expression during bacteriophage PRD1 infection using microarray technology. PRD1 was selected because it represents a structurally complex virus composed of a protein capsid, a lipid membrane, and a DNA genome (1, 15); it shares similarities with viruses infecting animals (adenovirus, family Adenoviridae), algae (PBCV-1, Phycodnaviridae) and insects (Chilo iridescent virus, Iridoviridae) (5, 7, 9, 10); and its life cycle is well characterized (6, 24). Furthermore, PRD1 can infect E. coli, for which gene functions and transcriptional control are characterized in detail. PRD1 is a lytic bacteriophage of the Tectiviridae family (6). During a PRD1 infection, 24 phage-encoded protein species are synthesized from two early and three late promoters (8, 23). Eighteen of these proteins are incorporated into the virion in a process assisted by both phage- and host-encoded proteins (8, 26, 46). Towards the end of the phage reproduction cycle, virus-encoded holin and endolysin proteins lyse the host cell, releasing the progeny virions (53, 68). Under optimal conditions, the entire PRD1 life cycle takes about an hour. Here we report changes in E. coli gene expression during bacteriophage PRD1 infection. The aim of the present study was to analyze changes in E. coli cells under the stress from lytic phage infection. We have specifically analyzed PRD1induced effects by comparing samples from infected cells to that from a noninfected control at 5, 10, 15, 30, and 50 min postinfection (p.i.). Simultaneously, we have analyzed temporal changes in the noninfected E. coli culture to reveal changes that take place during the ageing of the culture. To our knowledge, this report is the first on microarray analysis of bacterial
The reproduction of a virus is dependent on the host organism, which provides substrates and synthesis machineries for viral progeny production. Most viruses interfere with the host’s normal physiology to promote viral progeny production by downregulating the expression of cellular genes. This phenomenon was discovered in the 1940s after an infection of bacterial cells with T-even bacteriophages (for a review, see reference 35) and has been proven to also apply to animal and plant virus infections (for a review, see reference 4). An extensive shutoff of host gene expression early in the viral infection is not, however, necessarily a universal mechanism (56); other strategies may also exist. In addition to redirecting cellular recourses for viral reproduction, viruses may also change the cellular physiology to prevent subsequent infections of an already infected host cell. This phenomenon, known as superinfection exclusion, operates during infection by a variety of viruses (14, 31, 36, 63, 65) and bacteriophages (29, 34, 38, 43, 60). Recently, the effect of virus infection on host mRNA synthesis has been broadly studied using microarray technology. While the earlier-applied technologies either provided information at the whole-genome level, without knowledge of the specific genes affected, or covered only a select set of host genes, the array technology can potentially offer information about individual genes in the entire host gene pool. Currently, array technology has been used to analyze animal virus infections or, in a few cases, study changes in infected plant (42, 57, 66) or fungal cells (2, 3). So far, the microarray-based analyses of phage infections have focused on the detection of phagespecific transcripts in infected cells (19, 39), changes in the host
* Corresponding author. Mailing address: Viikki Biocenter, P.O. Box 56 (Viikinkaari 5), 00014 University of Helsinki, Helsinki, Finland. Phone: (358) 9 191 59100. Fax: (358) 9 191 59098. E-mail: dennis
[email protected]. † Supplemental material for this article may be found at http://jvi .asm.org/. ‡ Present address: Roal Oy, P.O. Box 57, 05201 Rajama¨ki, Finland. 8081
8082
PORANEN ET AL.
virus-host interaction, focusing on the changes in host gene expression during infection. MATERIALS AND METHODS Sample collection. E. coli K-12 strain JE2571 (leu thr thi lacY thy pil) (11), carrying conjugative plasmid RP4 (17), was used as a host for bacteriophage PRD1 (49). Cells were grown at 37°C in MOPS (morpholinepropanesulfonic acid) (pH 7.4)-buffered synthetic rich medium (47) supplemented with 4 mM leucine and 2 mM threonine. An overnight culture was diluted into fresh medium (1.5 ml to 50 ml) and grown with aeration at 37°C to an optical density at 600 nm (OD600) of 1.6. The dilution was repeated, now 12-fold (to an OD600 of 0.14 to 0.16), and the culture was split in half. The aeration was continued at 37°C until the OD600 reached 1.6. One portion of the cell culture was infected with PRD1 with a multiplicity of infection of 30; the other portion was a noninfected control culture. Cultures were continuously aerated at 37°C. Samples for RNA isolation were taken from infected and noninfected cultures at time points of 5, 10, 15, 30, and 50 min p.i. A sample at the zero time point was collected from the noninfected culture. In addition, a sample was collected from the infected culture 2 h p.i. to verify the productivity of the infection and to calculate the number of viral progenies. All of the samples were immediately further processed. The experiment was repeated four times (three times for microarray analysis and once for quantitative real-time PCR [qPCR]). Isolation of total RNA. Total RNA was isolated from ⬃1 ⫻ 109 cells (10 ml of cell culture) using two phenol and three ether extractions. In the first extraction, the phenol (Sigma P-4682, pH 4.3) was preheated to 95 to 100°C and the sample was supplemented with 1 ml of lysis buffer (0.5 M Tris-HCl, pH 7.5, 0.2 M EDTA, 10% sodium dodecyl sulfate [SDS]) for rapid cell lysis. The nucleic acids were precipitated from the aqueous phase by NaOAc and ethanol. Further purification was performed using an RNeasy midikit (QIAGEN; catalog no. 75144) and on-column DNase (QIAGEN; catalog no. 79254) digestion. The purified RNA samples were stored at ⫺80°C. Concentrations and qualities of the RNA samples were determined by measuring absorbance values at 260 and 280 nm, and the integrity of the total RNA was conformed by gel electrophoresis and ethidium bromide staining. The A260-to-A280 ratio was ⱖ2.1 in all RNA samples. mRNA enrichment and synthesis of fluorescent cDNA probes. The 16S and 23S rRNAs were removed from the total RNA samples by a capture hybridization approach using the MicroExpress bacterial mRNA purification procedure (Ambion; catalog no. 1905). The synthesis of Cy3- and Cy5-labeled cDNA for microarray hybridization was carried out according to the Amersham Biosciences protocol for CyScribe (catalog no. RPN5660) cDNA postlabeling. mRNA from 12 g of total RNA was used as a template for cDNA synthesis primed by random nonamers (Amersham Biosciences). After cDNA synthesis, the template RNA was degraded by RNase H (USB) at 37°C for 15 min. The amino-allyl-labeled cDNA was purified with a QIAquick PCR purification kit (QIAGEN; catalog no. 28104) using phosphate buffers (28). The cDNA was precipitated with NaAc and ethanol and then labeled with a freshly made suspension of Cy3 or Cy5 dye (in NaHCO3, pH 9) according to the protocol for CyScribe postlabeling (Amersham Biosciences; catalog no. RPN5660). The labeled cDNA was purified using the QIAGEN PCR purification procedure. Samples were stored at ⫺20°C in the dark. Test samples were labeled with Cy3, and reference samples were labeled with Cy5. Reverse labeling of the RNA was performed with one of the three samples to verify transcriptional induction rations. Hybridization. Differently labeled test and reference samples were combined together with nonspecific unlabeled E. coli rRNA (Boehringer Mannheim; catalog no. 206938) and single-stranded DNA (ssDNA) from herring sperm (Sigma; catalog no. D7290). The mixed samples were denatured and hybridized on glass DNA microarrays obtained from the Institute of Food Research (Norwich, United Kingdom; http://www.ifr.bbsrc.ac.uk/Safety/Microarrays/default.html). The hybridization was performed in 50% formamide, 6⫻ SSC (1⫻ SSC is 0.15 M NaCl plus 0.015 M sodium citrate), 0.5% SDS, and 5⫻ Denhardt’s solution (0.1% Ficoll 400, 0.1% polyvinylpyrrolidone, 0.1% bovine serum albumin) at 45°C for ⬃16 h. After the hybridization, the slides were washed once with 0.1⫻ SSC-0.1% SDS (for 10 min), twice with 0.1⫻ SSC (for 5 min), and three times with H2O (for 2 min). Data collection. Each microarray contained 7,680 spots arranged in 16 blocks, with each block containing 20 by 24 spots. A total of 4,266 of the probes were unique and specific for E. coli K-12 (according to E. coli genome annotation U00096). Every glass carried a duplicate printed microarray. The microarray glasses were scanned with ScanArray 5000 (Perkin Elmer) using 5-m resolution. Spot segmentation and intensity calculations were performed using GenePix image analysis software (Axon Instruments, Inc., 1999
J. VIROL. version). The local background was subtracted from the raw spot intensity to produce background-subtracted intensities, and the median intensity value for each spot was taken as a representative intensity value. Negative median values were set to zero prior to further processing. Data processing. The data contained three types of values: E. coli zero-minute expression intensities (E0); E. coli expression intensities for 5, 10, 15, 30, and 50 min p.i. (Ex, where x ⫽ 5, 10, 15, 30, or 50); and PRD1-infected E. coli expression intensities for 5, 10, 15, 30, and 50 min p.i. (Px, where x ⫽ 5, 10, 15, 30, or 50). There were 34 repeats of E0 (including self-hybridization), 12 repeats of Ex, and 6 repeats of Px at each time point. The total hybridization intensities summed over all of the E. coli K-12-specific spots within each array were set to a constant value, and a median value for an open reading frame (ORF) was taken as the representative intensity value at each time point and condition. Finally, different experiments were brought back to register by their mean total ratios, which were calculated from the original spot intensities by first scaling the Ex/E0 pair of expressions and then, with the scaled Ex, the pair of expressions Px/Ex. When the mean values were used (instead of the median values) throughout the procedure described above, the differences in global scale were negligible, as was the effect of a dye swap. After the spot intensities were normalized, the data were made to reflect the current annotation of the E. coli K-12 genome (GenBank accession no. U00096.2; version 24 June 2004) according to the information at http://biocyc .org/ecocyc/release-notes.shtml. As a result, the number of E. coli K-12-specific ORFs in our analysis was 4,207. Genes with signal values of less than 15% (0.15) of the mean value from all of the ORFs (mean, 1.0) were considered to be not expressed at a given time point. This cutoff value is based on median standard deviation values calculated from the E0 data set for each ORF. Over 98% of the calculated median minus standard deviation values were ⬎0, if genes with signal values less that 0.15 were omitted. Based on this criterion, 1,672 ORFs, corresponding to ⬃39% of all E. coli ORFs, were not expressed in the noninfected E. coli cells at the zero time point. This is in good accordance with data previously reported for an E. coli culture grown in rich medium (1,776 silent genes) (64). Of the 1,672 nonexpressed ORFs in the E0 sample, only 552 have established functions. Differences in gene expression were considered significant if the expression values differed more than threefold between two conditions. Using this criterion, 99% of the calculated [(SD ⫹ median)/median] ratios for each ORF within the E0 data set were ⬍3 (calculated using signal values of ⬎0.15). The same criterion was used to detect the induction of genes which were considered nonexpressed (signal values of ⬎[3 ⫻ 0.15]). In order to detect more subtle changes in expression levels than a threefold cutoff would reveal, a t test was performed using the R program (http://www.r-project.org) for all Ex/E0 and Px/Ex conditions. A P value of ⬍0.01 was used to indicate a statistically significant change between conditions. Operon organization and gene functions are based on information on the EcoCyc (Encyclopedia of Escherichia coli K12 Genes and Metabolism; http: //ecocyc.org/) website (32). Validation of transcript levels using qPCR. Total RNA samples for qPCR were isolated from a cell culture independent of those used for microarray sample preparation. The isolated RNA was subjected to additional DNase treatment (RQ1 DNase; Promega) and purification with the RNeasy MinElute cleanup kit (QIAGEN, catalog no. 74204). The purified RNA was used as a template for cDNA production using random hexamer primers and MultiScribe reverse transcriptase (TaqMan reverse transcription reagents; Applied Biosystems; catalog no. N8080234). The resulting cDNA samples were used as templates for qPCR. Primer pairs specific to several genes of interest were designated using PrimerExpress software (Applied Biosystems) (data not shown), and the PCR was carried out using SYBR green PCR master mix (Applied Biosystems). Thermal cycling and real-time monitoring of SYBR green fluorescence were performed with an ABI Prism 7000 sequence detector system (Applied Biosystems). Changes in expression levels were analyzed using ABI Prism 7000 SDS software (Applied Biosystems) and ftsZ (a cell division gene) as a reference. Western blot analysis. Western blotting was carried out as described previously (54) using primary antibody to Hsp60 (groL gene product). Polyclonal anti-Hsp60 serum was raised by immunizing a rabbit with purified Hsp60 (GroEL) protein (Boehringer) as described previously (27).
RESULTS AND DISCUSSION Dissection of virus-induced changes. To distinguish the virus-induced changes from the E. coli growth-dependent changes, we constructed two experimental setups: one to analyze the noninfected E. coli culture under the experimental
VOL. 80, 2006
GENE EXPRESSION CHANGES IN PRD1-INFECTED E. COLI
FIG. 1. Experimental setup. E. coli K-12 JE2571 (RP4) cells, cultured in synthetic rich medium with aeration to the exponential growth phase, were infected with PRD1. A high multiplicity of infection (30 infective viruses/cell) was used to ensure that practically all of the cells were infected. (A) Turbidity values (OD600) from three independent experiments plotted as a function of time (dashed lines, infected cells; solid lines, noninfected cells). The addition of the virus is indicated by an arrow. Samples (⬃109 cells) were collected for total RNA isolation at five time points (circles, 5, 10, 15, 30, and 50 min p.i.) from both infected and noninfected cultures. In addition, a sample from the zero time point was collected from the noninfected control culture. (B and C) Phase-contrast images (Olympus; model no. BX50) of PRD1-infected (B) and noninfected (C) cells around 40 min p.i. The frequency of cells within different stages of cell division was equal in both samples. Bar, 10 m.
conditions and time scale of the infection (noninfected x-min sample versus noninfected 0-min sample; Ex/E0) and another to evaluate the difference between the infected and noninfected cultures within each experimental time point (infected x-min sample versus noninfected x min sample; Px/Ex). The results of these two measurements establish the mRNA content of the infected cell in comparison to that before the infection (infected x-min sample versus noninfected 0-min sample; Ex/E0 ⫻ Px/Ex ⫽ Px/E0). We made six replicates of each condition and time point, three biological replicates (Fig. 1A) (RNA isolated from three independent cultures) and two technical replicates from each of them (altogether 124 measurements). Our analysis covers over 99% (4,207 ORFs) of the currently known E. coli ORFs (4,242 ORFs, GenBank version U00096.2). Initially, the selection of differentially expressed genes was based on differences (n-fold) calculated from the normalized expression values of two different conditions (as described in Materials and Methods). A threefold cutoff value was applied throughout the analysis, and this discussion is focused on genes with known physiological functions. The similar expression patterns for genes within the same operons and regulons strengthen the significance of the results obtained. The microarray expression values were verified for selected genes using qPCR and Western blot analyses. In addition, statistical analysis was applied to determine the significance of the results and to allow the detection of genes with lower changes (n-fold) but with high confidence levels in terms of P value. Gene expression changes in the noninfected E. coli culture. The physiological state and the transcriptional activity of the
8083
noninfected E. coli cells were not constant during the measuring period that is optimal for virus production. The growth rate of the culture started to decrease at the time point equivalent to 15 min p.i. (Fig. 1A), indicating that the bacteria were facing nutrient exhaustion and preparing for the stationary phase. Previous studies have identified the stimulation of 215 E. coli genes during the transition into the stationary phase (62). We detected the induction and repression of 263 and 99, respectively, ORFs in the noninfected E. coli culture during the 50-min experimental period (Fig. 2A). Most notably, the entire cysteine regulon (cysB, cysDNC, cysJIH, cysPUWAM, cysK, sbp, cbl, and tauABCD) was upregulated as well as many genes responding to phosphate starvation (phoH, phoBR, phoA, pstSCAB, and phoU) or involved in the uptake and biosynthesis of methionine (metA, metBL, metC, metF, metNIQ, metJ, metK, metE, and metR) and arginine (argC, argE, argF, argI, and carAB), indicating that some of the essential components of the growth medium became limited. The aging of the culture resulted also in the downregulation of several genes functioning in energy metabolism (atpBFAGC, cyoCDE, and nuoEFGHIJLMN) and transport of small molecules (e.g., dppBCDF, potABCD, and rbsDACB). Virus-induced changes on host gene expression. The growth rate of the E. coli culture was not affected by the PRD1 infection; the turbidity (OD600) of PRD1-infected and noninfected cultures rises with equal rates until the phage induces lysis of the infected cells at ⬃50 min p.i (Fig. 1A). Also, the cell division rate of the infected and noninfected cultures was the same (Fig. 1B). Surprisingly, the PRD1-induced changes in host gene expression were relatively moderate and the total number of ORFs affected due to the infection (327) (Fig. 2A) was less than the number of differentially expressed ORFs (462) in the noninfected cells during the experimental period (Fig. 2A). Genes that are involved in cell growth and division were not affected during infection, which is in accordance with the data shown in Fig. 1. The majority of the virus-induced changes occurred after the production of virion components, during the assembly of progeny virions and the expression of the lysis components (30 to 50 min p.i.) (Fig. 2) (46, 68). Thus, there was no major reprogramming of the cell physiology early in the infection cycle to promote the production of viral components. Unlike what was expected, pronounced transcriptional activation of genes involved in the biosynthesis of amino acids, phospholipids, or nucleotides (the building blocks for the virion synthesis) was not observed. Instead, the upregulation of several genes involved in the biosynthesis of different amino acids was delayed or suppressed in infected cells (see below) (see Fig. 4N and P). Owing to this unexpected result, we calculated the quantity of substrate molecules utilized for the production of progeny virions and compared it to the total supply in the replicating E. coli cell (Table 1). We observed that the PRD1 progeny production utilizes only a fraction of the host capacity to synthesize DNA, proteins, and phospholipids (5 to 15%). This observation rationalizes the invariability in the expression of biosynthetic enzymes during infection. Early phase of PRD1 infection—induction of uptake pathways. The changes detected in the host transcriptosome early in the infection (5 to 15 min p.i.) included genes involved in osmotic adaptation (osmB, osmY, and treR), anaerobic respiration (fdoHI, fdnG, hypC, and nuoK), uptake of organic com-
8084
PORANEN ET AL.
J. VIROL.
FIG. 2. Global changes in E. coli gene expression during the PRD1 infection. (A) The number of up- and downregulated ORFs in noninfected (Ex/E0) and PRD1-infected (Px/Ex) E. coli cells. Changes in expression levels were considered significant when the relative expression value between two conditions was greater than 3.0 (upregulated) or less than 0.33 (downregulated). Percentages of up- or downregulated ORFs from all of the measured ORFs are shown in parentheses. “Changed” indicates the total number of genes which have changed expression during the experiment. (B) Time scale of viral life cycle. PRD1 recognizes a specific receptor on the host cell outer membrane and delivers its linear dsDNA genome into the host cytoplasm using its internal membrane as a genome delivery device (25). This process induces changes in the permeability and energetic state of the host plasma membrane (18). Early viral proteins involved in viral genome replication are synthesized immediately after infection (⬃5 min p.i.) (45), leading to phage DNA replication (10 to 20 min p.i.) (52). Protein components of the virions are produced at about 15 min p.i (45). Viral membrane proteins associated with the host plasma membrane interact with soluble viral capsid proteins to form empty procapsids at about 30 min p.i. (6, 46). The viral genome is packaged into empty particles, and at about 40 min p.i., DNA containing viral particles can be detected inside infected cells (46, 59). Already at about 35 min p.i., viral lysis components are present within infected cells and become active around 55 min p.i. due to decreases in the cellular ATP levels (68). Approximately 200 progeny virions are released from each infected cell.
pounds (dctA, galP, glpTQ, gntT, manY, nupG, and treB), and export of cations (zntA). PRD1 genome delivery into the cell (Fig. 2B) induces transient permeability changes in the host’s outer and plasma membranes and the release of cytoplasmic compounds into the external milieu (18). We propose that the activation of uptake pathways and anaerobic respiration systems early in the PRD1 infection is aimed toward compensating these effects and restoring normal cytoplasmic conditions. Cellular stress responses. Many of the most highly induced genes during the assembly of progeny virions (30 to 50 min p.i.) (Fig. 2) were chaperonins, proteases, and other stress-inducible genes of the 32-dependent heat shock regulon (as shown in Table 2 for dnaKJ, groSL, grpE, hslVU, htpG, clpB, ibpAB,
rfaFCL and Fig. 3A for groL and ibpB). Also, the hslO gene for the molecular chaperone Hsp33 (30) was induced in infected cells (Table 2). A general signal for the 32-dependent transcription is the stress-induced unfolding of proteins in the cytoplasm (67). Apparently, the amount of unfolded polypeptides within the cell increases due to infection, inducing the 32-regulon. Previous studies have revealed that PRD1 reproduction is actually dependent on several cellular chaperonins; the correct folding of the PRD1 major capsid proteins P3 and P5 is dependent on the Hsp60 (groL) and Hsp10 (groS) chaperonins (26). The morphogenesis of other bacteriophages, such as lambda and T4, also relies on host chaperonins (22, 61). One of the best-characterized bacterial stress responses is
TABLE 1. Quantity of substrate molecules utilized for the production of PRD1 virions and the total supply in a replicating E. coli cell No. of molecules or molecular mass consumed for Production of a PRD1 virion
Replication of an E. coli cell
Proportion of E. coli replication capacity consumed for PRD1 reproductiond
29,854 nucleotidesa 26,000 lipid molecules (15) 46.2 MDa protein mass (6)
21,342,505 nucleotidesb 22,000,000 lipid molecules (48) 94,000 MDa protein massc
0.14 0.12 0.05
Product
dsDNA Membrane Protein a
The PRD1 genome is 14,925 bp long (accession no. 001421). An E. coli genome is 4,639,675 bp long (U00096.2), and an actively dividing cell contains about 2.3 genomes (48). An E. coli cell contains about 2,350,000 protein molecules with an average molecular mass of 40 kDa (48). d Calculated for 100 virions. This is based on the number of progeny PRD1 virions produced (mean of the three experiments, 1.0 ⫻ 1011 PFU/ml) and the number of E. coli cells at 15 min p.i. when the assembly of progeny virions is initiated (mean of the three experiments, 9.8 ⫻ 108 cells/ml). (E. coli cells are dividing during the infection.) b c
GENE EXPRESSION CHANGES IN PRD1-INFECTED E. COLI
VOL. 80, 2006
8085
TABLE 2. Regulation of selected stress-inducible genes during PRD1 infection Gene type and name Heat shock-inducible genes dnaK dnaJ groS groL
b no.a
b0014 b0015 b4142 b4143
Change in expression (n ⫽ fold) after infection timeb
Description and/or product 5 min
10 min
15 min
30 min
50 min
0.91 (0.1748) 1 (0.0566) 0.8 (0.3963) 1.21 (0.0035)
0.72 (0.0044) 0.78 (0.0111) 0.54 (0.0022) 1.12 (0.4847)
1.51 (0.0227) 1.11 (0.2021) 1.01 (0.4168) 1.58 (0.0622)
3.74 (0.0004) 3.04 (0.0196) 7.07 (0.0073) 3.61 (0.0003)
5.51 (1E-05) 8.01 (4E-07) 13.17 (3E-07) 6.73 (1E-05)
1.16 (0.0088)
0.69 (0.0019)
1.02 (0.444)
1.56 (0.166)
5.29 (3E-05)
grpE
b2614
hslU
b3931
hslV htpG clpB ibpB ibpA rfaF rfaC rfaL
b3932 b0473 b2592 b3686 b3687 b3620 b3621 b3622
Chaperone Hsp70; DNA biosynthesis Chaperone with DnaK GroES, chaperone binds to Hsp60 GroEL, chaperone Hsp60, peptidedependent ATPase Phage lambda replication; host DNA synthesis; protein repair hslVU, ATPase subunit, homologous to chaperones hslVU, peptidase subunit chaperone Hsp90 Heat shock protein Heat shock protein Heat shock protein Lipopolysaccharide core biosynthesis Lipopolysaccharide core biosynthesis Lipopolysaccharide core biosynthesis
Redox regulated gene hslO
b3401
Hsp33 chaperone
1.2 (0.0509)
0.69 (0.0037)
1.18 (0.1931)
1.46 (0.0023)
5.57 (6E-05)
Phage shock regulon pspA
b1304
Phage shock protein, inner membrane protein Phage shock protein Phage shock protein: activates Psp expression Phage shock protein Phage shock protein Orf, hypothetical protein
1.08 (0.004)
2.12 (3E-06)
2.95 (4E-06)
3.84 (0.0002)
11.18 (0.0004)
1.17 (0.2752) 1.52 (0.0845)
0.96 (0.3561) 1.42 (0.3491)
2.96 (0.0001) 2.12 (0.0058)
1.77 (0.0395) 1.96 (0.0018)
7.01 (0.0003) 5.61 (0.0006)
1.81 (0.291) 1.26 (0.2027) 2.06 (0.0079)
0.83 (0.3261) 1.69 (0.1682) 0.98 (0.3919)
3.1 (0.0153) 2.14 (0.0074) 1.56 (0.2452)
2.35 (0.236) 2.1 (0.0403) 1.22 (0.256)
3.94 (0.0002) 3.48 (0.0009) 9.87 (0.0003)
pspB pspC
b1305 b1306
pspD pspE yjbO
b1307 b1308 b4050
a b
0.93 (0.1321)
0.67 (0.0043)
1.02 (0.4553)
3.76 (0.0012)
6.16 (3E-07)
0.89 (0.2008) 0.91 (0.1493) 0.8 (0.3996) 0.73 (0.4396) 0.49 (0.0007) 1.39 (1E-07) 1.28 (5E-06) 0.9 (0.2103)
0.59 (0.0005) 0.64 (0.0014) 0.58 (0.0044) 0.71 (0.2882) 0.46 (0.0196) 1.13 (0.4303) 1.22 (0.101) 1.16 (0.4005)
1.44 (0.0186) 0.97 (0.3677) 0.88 (0.2795) 1.92 (0.2814) 2.19 (0.0292) 1.18 (0.0104) 0.87 (0.1439) 0.64 (0.0407)
2.95 (0.0011) 4.48 (0.0002) 3.92 (5E-07) 6.41 (0.001) 9.01 (0.0034) 6.34 (9E-06) 8.22 (9E-05) 4.47 (8E-07)
8.3 (2E-07) 14.04 (1E-06) 7.02 (8E-06) 57.35 (4E-08) 32.3 (8E-10) 1.1 (0.3521) 1.07 (0.3932) 1.48 (0.0744)
b no., E. coli gene identification number (Blattner number). Values in parentheses are P values.
the SOS response, which is activated under different stress conditions, the main trigger being ssDNA (44). In the beginning of PRD1 infection, the viral linear double-stranded DNA (dsDNA) genome is replicated (Fig. 2B), potentially exposing ssDNA. However, none of the approximately 15 genes of the SOS response regulon were activated during the PRD1 infection. This suggests that the viral ssDNA binding proteins P12 (which is the most highly expressed viral protein during PRD1 infection) and P19 (50, 51) can efficiently shield any exposed ssDNA. Neither of the viral ssDNA binding proteins is needed for viral genome replication in vitro (55), suggesting that the evolutional driving force on P12- and P19-encoding genes may have been to prevent the SOS induction. Components of the phage shock regulon (pspABCDE and yjbO) (37) were among the most highly induced genes during late PRD1 infection (30 and 50 min p.i.) (Table 2). The phage shock protein (Psp) system responds to stress that reduces the energy status of the cell, including extreme temperatures and osmolarity, mislocalization of envelope proteins, and the presence of ionophores or ethanol (16). Initially, phage shock protein A (PspA) was detected at high concentrations in filamentous phage-infected E. coli cells (12). However, an induction of the Psp response during an infection of other coliphages has not been described so far. Apparently PspA has a role in maintaining the proton motive force under different conditions of cellular stress (33). The induction of the pspA gene during PRD1 infection could explain the increase in membrane voltage detected during infection (18). The upregulation of the
pspA gene and the whole psp operon also coincides with the production of PRD1 holin protein (about 30 min p.i.). The activation of the lysis functions by holin can be triggered prematurely by ATP depletion (68). The induction of PspA production may be a host defense mechanism for maintaining the cellular energy status in order to postpone the viral lysis functions. Alternatively, the phage has evolved a mechanism to prevent premature lysis by activating the expression of the psp operon. Induction of exopolysaccharide synthesis. E. coli cells have the potential to produce a number of exopolysaccharides on the cell surface. These exopolysaccharides often protect bacteria from harsh environments, chemicals, or bacteriophages. We detected a significant induction of several genes involved in the synthesis of the colonic acid capsule (wza, wzc, wcaABDE, gmd, wcaGHI, cpsBG, wcaJ, wzxC, wcaKL, and rcsA) (32, 58), starting with the stimulation of the rcsA gene (a positive regulator of capsule polysaccharide synthesis) at 10 min p.i. Exopolysaccharide induction can be viewed as a host defense mechanism for protecting the bacterial population from further phage infections. Alternatively, the synthesis of the host colonic acid capsule is a mechanism for superinfection immunity. Host protein synthesis shutoff. The expression of the guanosine-3⬘,5⬘-bis pyrophosphate, (p)ppGpp, synthetase II (spoT) was transiently increased at 30 min after PRD1 infection. (Fig. 3A). The increase in (p)ppGpp concentration is known to be connected to nutrient limitation and leads to a reduction in rRNA synthesis and changes in cellular protein
8086
PORANEN ET AL.
FIG. 3. Verification of microarray results by qPCR and Western blot analysis. (A) Comparison of mRNA levels in noninfected and PRD1-infected cells by microarray and qPCR for selected E. coli genes. Genes are indicated within the upper left corner of each panel. Solid line, noninfected x-min sample versus noninfected 0-min sample (Ex/E0); dashed line (black), infected x-min sample versus noninfected 0-min sample; (Px/E0); dashed line (gray), infected x-min sample versus noninfected x-min sample (Px/Ex). Notice the different scale for ibpB and metB. (B) Detection of the groL gene product Hsp60 from PRD1infected cells by Western blot analysis. Samples taken from PRD1infected cells at different time points of infection were analyzed by Western blotting using Hsp60-specific antibodies.
synthesis (40). Accordingly, a reduction in host protein synthesis has been detected in PRD1-infected, pulse-labeled cells beginning at 40 min p.i. (45). Host protein synthesis shutoff is considered to be a common phenomenon associated with animal and plant virus infections (for examples, see references 4 and 20). Host macromolecular synthesis is also drastically reduced during the first few minutes of bacteriophage T2, T4, and T6 infections (35). It has been suggested that the down-
J. VIROL.
regulation of protein synthesis is specifically induced by the virus and is aimed to channel host recourses for the virus reproduction. The timing of the spoT upregulation and the host protein synthesis reduction in PRD1-infected cells suggests that the host protein synthesis downregulation in PRD1infected cells is rather a secondary effect of nutrient limitation not aimed at boosting viral protein synthesis. Similar to PRD1, bacteriophage ⌽29 has little impact on host transcription until late in infection (56). Verification of microarray data by qPCR and Western blot analysis. qPCR and Western blot analyses were performed for selected genes to confirm the microarray results (Fig. 3). Five genes (cysB, groL, ibpB, metB, and spoT) that showed different expression patterns in microarray analysis (Fig. 3A, left panel) were selected for qPCRs. The qPCR results shown in Fig. 3A (right panel) were in good agreement with the microarray data. Western blot analysis was performed to detect cellular levels of Hsp60 protein (encoded by the groL gene) in PRD1-infected E. coli cultures. Both microarray and qPCR analyses suggested that, in noninfected cells, the expression of the groL gene is reduced, but in PRD1-infected cells, this normally occurring downregulation is suppressed (Fig. 3A). Accordingly, the Hsp60 levels in infected cells were constant throughout the infection (Fig. 3B). Temporal changes in E. coli gene expression. To further analyze the dynamic behavior of the cellular transcriptosome during the viral infection, all of the E. coli ORFs included in our experiment were clustered based on their expression patterns in infected and noninfected cells. The specific clusters formed are presented in Fig. 4, and the genes within each cluster are listed in Table S1 in the supplemental material. The majority of the ORFs (3,803; corresponding to 90% of the genome) follow the same expression profile in both noninfected and infected cells throughout the infection cycle (Fig. 4A to G). Certain processes were delayed in infected cells; for example, the induction of genes involved in methionine uptake and biosynthesis (metA, metBL, metC, metF, metNIQ, metJ, metE, metR, and mmuP) occurred later in infected cells compared to that in noninfected cells (Fig. 4N). Also, the expression of several biosynthetic genes for lipopolysaccharide (rfaFC), membrane-associated oligosaccharide (mdoGH), and enterobacterial common antigen (rffH, wecB, and wecC) were repressed earlier in noninfected cells than in infected cells (Fig. 4O). Genes associated with taurine and sulfate transport and transcriptional regulation of the cysteine regulon (cysB, cysM, cbl, sbp, tauABCD, and fliY) were upregulated in noninfected cells, but this induction was suppressed in infected cells (Fig. 4P). In addition, the downregulation of several genes was suppressed in infected cells (Fig. 4Q). Interestingly, most of the induced (Fig. 4H to L) or repressed (Fig. 4M) ORFs during PRD1 infection were expressed at constant levels in the noninfected cells through the experimental period (Fig. 4H to J and L to M), whereas the majority of the ORFs that changed expression in the noninfected cells were not affected by infection (Fig. 4B to G). This indicates that the pathways connected to starvation and phage infection do not significantly overlap. Detection of additional genes with significantly changed expression. The use of a high cutoff (n-fold) may exclude genes with lower changes (n-fold) but with high confidence levels. To
VOL. 80, 2006
GENE EXPRESSION CHANGES IN PRD1-INFECTED E. COLI
8087
their own reproduction by efficient takeover and reprogramming of the host physiology or, as presented here, replicate within the cell without affecting major biosynthetic pathways during the period of progeny production. This latter strategy may ensure virus propagation without the induction of host defense mechanisms, while the previous strategy likely activates host responses. The reproduction of PRD1 phage within an actively dividing E. coli cell consumes only a fraction of the host’s biosynthetic capacity. The induction of host defense systems would probably have an adverse effect on phage progeny production and decrease phage fitness. Such overtaking activity would also consume the coding capacity of the viral genome and extend the replication time. The reprogramming of host cell physiology by viruses should not solely be considered as a way to redirect biosynthetic materials for viral reproduction. This assumption may be correct for the large phages for which such phenomenon was initially described (for a review, see reference 35). The modification of host cell physiology should also be considered as a potential mechanism of superinfection exclusion or host defense. This could be the case, especially among viruses that do not significantly utilize the host’s replication capacity. It would be intriguing to obtain results of the consumption of the host’s replication capacity in other viral systems to evaluate the role of host reprogramming. FIG. 4. Representatives of gene expression pattern clusters. E. coli ORFs were divided into three expression categories within each time point and experiment: downregulated (⫺1), upregulated (1), and not changed (0). The simplified expression profiles were clustered so that the ORFs having the same temporal expression pattern in both conditions (infected and noninfected) were placed in the same cluster. The number of possible clusters was 59,049 (three classes in five time points in two conditions). However, only 116 different clusters were obtained. All clusters containing more than 10 members are shown (17 clusters, panels A to Q). Solid line, noninfected x-min sample versus noninfected 0-min sample (Ex/E0); dashed line, infected x-min sample versus noninfected x-min sample (Px/Ex); dotted line, sum of the previous ones (infected x-min sample versus noninfected 0-min sample; Px/E0). The number of ORFs in each cluster is indicated within each panel. Clusters can be divided into seven classes based on the virus effect: (i) no changes in gene expression in comparison to the noninfected culture (A to G), (ii) upregulation of gene-expression in the infected cell (H to L), (iii) downregulation of gene expression in the infected cell (M), (iv) delayed upregulation of gene expression in infected cell in comparison to the noninfected culture (N), (v) delayed downregulation in the infected cell in comparison to the noninfected culture (O), (vi) no change in expression level during the infection due to the virus-induced suppression of upregulation (P), and (vii) no change in expression level during infection due to virus-induced suppression of downregulation (Q).
investigate this possibility, we applied statistical methods as described in Materials and Methods. The results are summarized in Table S2 in the supplemental material. The overall results of this analysis support our initial gene selection based on differences (n-fold) (see Table S2 in the supplemental material). By applying criteria of changes of ⬎2-fold and a P value of ⬍0.01, it was possible to identify one more heat shockregulated gene, hslJ, which was induced late during PRD1 infection. The hslJ protein is a membrane-associated heat shock protein. Conclusions. There seems to be several reproduction tactics among viruses. Viruses may either harness the host cell for
ACKNOWLEDGMENTS We thank Marja-Leena Pera¨la¨ for assistance with sample collection. This investigation was supported by the Finnish Centre of Excellence Program 2006-2011 (1213467 and 1213992) from the Academy of Finland to D.H.B. REFERENCES 1. Abrescia, N. G., J. J. Cockburn, J. M. Grimes, G. C. Sutton, J. M. Diprose, S. J. Butcher, S. D. Fuller, C. San Martin, R. M. Burnett, D. I. Stuart, D. H. Bamford, and J. K. Bamford. 2004. Insights into assembly from structural analysis of bacteriophage PRD1. Nature 432:68–74. 2. Allen, T. D., A. L. Dawe, and D. L. Nuss. 2003. Use of cDNA microarrays to monitor transcriptional responses of the chestnut blight fungus Cryphonectria parasitica to infection by virulence-attenuating hypoviruses. Eukaryot. Cell 2:1253–1265. 3. Allen, T. D., and D. L. Nuss. 2004. Specific and common alterations in host gene transcript accumulation following infection of the chestnut blight fungus by mild and severe hypoviruses. J. Virol. 78:4145–4155. 4. Aranda, M., and A. Maule. 1998. Virus-induced host gene shutoff in animals and plants. Virology 243:261–267. 5. Bamford, D. H., R. M. Burnett, and D. I. Stuart. 2002. Evolution of viral structure. Theor. Popul. Biol. 61:461–470. 6. Bamford, D. H., J. Caldentey, and J. K. Bamford. 1995. Bacteriophage PRD1: a broad host range dsDNA tectivirus with an internal membrane. Adv. Virus Res. 45:281–319. 7. Bamford, D. H., J. M. Grimes, and D. I. Stuart. 2005. What does structure tell us about virus evolution? Curr. Opin. Struct. Biol. 15:655–663. 8. Bamford, J. K., J. J. Cockburn, J. Diprose, J. M. Grimes, G. Sutton, D. I. Stuart, and D. H. Bamford. 2002. Diffraction quality crystals of PRD1, a 66-MDa dsDNA virus with an internal membrane. J. Struct. Biol. 139:103– 112. 9. Benson, S. D., J. K. Bamford, D. H. Bamford, and R. M. Burnett. 2004. Does common architecture reveal a viral lineage spanning all three domains of life? Mol. Cell 16:673–685. 10. Benson, S. D., J. K. Bamford, D. H. Bamford, and R. M. Burnett. 1999. Viral evolution revealed by bacteriophage PRD1 and human adenovirus coat protein structures. Cell 98:825–833. 11. Bradley, D. E. 1980. Morphological and serological relationships of conjugative pili. Plasmid 4:155–169. 12. Brissette, J. L., M. Russel, L. Weiner, and P. Model. 1990. Phage shock protein, a stress protein of Escherichia coli. Proc. Natl. Acad. Sci. USA 87:862–866. 13. Chen, Y., I. Golding, S. Sawai, L. Guo, and E. C. Cox. 2005. Population fitness and the regulation of Escherichia coli genes by bacterial viruses. PLoS Biol. 3:1276–1282.
8088
PORANEN ET AL.
14. Christen, L., J. Seto, and E. G. Niles. 1990. Superinfection exclusion of vaccinia virus in virus-infected cell cultures. Virology 174:35–42. 15. Cockburn, J. J., N. G. Abrescia, J. M. Grimes, G. C. Sutton, J. M. Diprose, J. M. Benevides, G. J. Thomas, Jr., J. K. Bamford, D. H. Bamford, and D. I. Stuart. 2004. Membrane structure and interactions with protein and DNA in bacteriophage PRD1. Nature 432:122–125. 16. Darwin, A. J. 2005. The phage-shock-protein response. Mol. Microbiol. 57:621–628. 17. Datta, N., R. W. Hedges, E. J. Shaw, R. B. Sykes, and M. H. Richmond. 1971. Properties of an R factor from Pseudomonas aeruginosa. J. Bacteriol. 108: 1244–1249. 18. Daugelavicˇius, R., J. K. Bamford, and D. H. Bamford. 1997. Changes in host cell energetics in response to bacteriophage PRD1 DNA entry. J. Bacteriol. 179:5203–5210. 19. Duplessis, M., W. M. Russell, D. A. Romero, and S. Moineau. 2005. Global gene expression analysis of two Streptococcus thermophilus bacteriophages using DNA microarray. Virology 340:192–208. 20. Ehrenfeld, E. 1982. Poliovirus-induced inhibition of host-cell protein synthesis. Cell 28:435–436. 21. Frye, J. G., S. Porwollik, F. Blackmer, P. Cheng, and M. McClelland. 2005. Host gene expression changes and DNA amplification during temperate phage induction. J. Bacteriol. 187:1485–1492. 22. Georgopoulos, C. P., R. W. Hendrix, A. D. Kaiser, and W. B. Wood. 1972. Role of the host cell in bacteriophage morphogenesis: effects of a bacterial mutation on T4 head assembly. Nat. New Biol. 239:38–41. 23. Grahn, A. M., J. K. Bamford, M. C. O’Neill, and D. H. Bamford. 1994. Functional organization of the bacteriophage PRD1 genome. J. Bacteriol. 176:3062–3068. 24. Grahn, A. M., S. J. Butcher, J. K. H. Bamford, and D. H. Bamford. 2006. PRD1: dissecting the genome, structure, and entry, p. 161–170. In R. Calendar (ed.), The bacteriophages, 2nd ed. Oxford University Press, New York, N.Y. 25. Grahn, A. M., R. Daugelavicˇius, and D. H. Bamford. 2002. Sequential model of phage PRD1 DNA delivery: active involvement of the viral membrane. Mol. Microbiol. 46:1199–1209. 26. Ha ¨nninen, A. L., D. H. Bamford, and J. K. Bamford. 1997. Assembly of membrane-containing bacteriophage PRD1 is dependent on GroEL and GroES. Virology 227:207–210. 27. Ha ¨nninen, A. L., D. H. Bamford, and J. K. Bamford. 1997. Probing phage PRD1-specific proteins with monoclonal and polyclonal antibodies. Virology 227:198–206. 28. Hegde, P., R. Qi, K. Abernathy, C. Gay, S. Dharap, R. Gaspard, J. E. Hughes, E. Snesrud, N. Lee, and J. Quackenbush. 2000. A concise guide to cDNA microarray analysis. BioTechniques 29:548–556. 29. Hutchison, C. A., III, and R. L. Sinsheimer. 1971. Requirement of protein synthesis for bacteriophage ⌽X174 superinfection exclusion. J. Virol. 8:121– 124. 30. Jakob, U., W. Muse, M. Eser, and J. C. Bardwell. 1999. Chaperone activity with a redox switch. Cell 96:341–352. 31. Karpf, A. R., E. Lenches, E. G. Strauss, J. H. Strauss, and D. T. Brown. 1997. Superinfection exclusion of alphaviruses in three mosquito cell lines persistently infected with Sindbis virus. J. Virol. 71:7119–7123. 32. Keseler, I. M., J. Collado-Vides, S. Gama-Castro, J. Ingraham, S. Paley, I. T. Paulsen, M. Peralta-Gil, and P. D. Karp. 2005. EcoCyc: a comprehensive database resource for Escherichia coli. Nucleic Acids Res. 33: D334–D337. 33. Kleerebezem, M., W. Crielaard, and J. Tommassen. 1996. Involvement of stress protein PspA (phage shock protein A) of Escherichia coli in maintenance of the protonmotive force under stress conditions. EMBO J. 15:162– 171. 34. Kliem, M., and B. Dreiseikelmann. 1989. The superimmunity gene sim of bacteriophage P1 causes superinfection exclusion. Virology 171:350–355. 35. Koerner, J. F., and D. P. Snustad. 1979. Shutoff of host macromolecular synthesis after T-even bacteriophage infection. Microbiol. Rev. 43:199–223. 36. Lee, Y. M., D. M. Tscherne, S. I. Yun, I. Frolov, and C. M. Rice. 2005. Dual mechanisms of pestiviral superinfection exclusion at entry and RNA replication. J. Virol. 79:3231–3242. 37. Lloyd, L. J., S. E. Jones, G. Jovanovic, P. Gyaneshwar, M. D. Rolfe, A. Thompson, J. C. Hinton, and M. Buck. 2004. Identification of a new member of the phage shock protein response in Escherichia coli, the phage shock protein G (PspG). J. Biol. Chem. 279:55707–55714. 38. Lu, M. J., and U. Henning. 1989. The immunity (imm) gene of Escherichia coli bacteriophage T4. J. Virol. 63:3472–3478. 39. Luke, K., A. Radek, X. Liu, J. Campbell, M. Uzan, R. Haselkorn, and Y. Kogan. 2002. Microarray analysis of gene expression during bacteriophage T4 infection. Virology 299:182–191. 40. Magnusson, L. U., A. Farewell, and T. Nystrom. 2005. ppGpp: a global regulator in Escherichia coli. Trends Microbiol. 13:236–242. 41. Manna, D., A. M. Breier, and N. P. Higgins. 2004. Microarray analysis of transposition targets in Escherichia coli: the impact of transcription. Proc. Natl. Acad. Sci. USA 101:9780–9785. 42. Marathe, R., Z. Guan, R. Anandalakshmi, H. Zhao, and S. P. DineshKumar. 2004. Study of Arabidopsis thaliana resistome in response to cucum-
J. VIROL.
43. 44. 45.
46.
47. 48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61. 62.
63.
64.
65. 66.
67. 68.
ber mosaic virus infection using whole genome microarray. Plant Mol. Biol. 55:501–520. McAllister, W. T., and C. L. Barrett. 1977. Superinfection exclusion by bacteriophage T7. J. Virol. 24:709–711. Michel, B. 2005. After 30 years of study, the bacterial SOS response still surprises us. PLoS Biol. 3:1174–1176. Mindich, L., D. Bamford, C. Goldthwaite, M. Laverty, and G. Mackenzie. 1982. Isolation of nonsense mutants of lipid-containing bacteriophage PRD1. J. Virol. 44:1013–1020. Mindich, L., D. Bamford, T. McGraw, and G. Mackenzie. 1982. Assembly of bacteriophage PRD1: particle formation with wild-type and mutant viruses. J. Virol. 44:1021–1030. Neidhardt, F. C., P. L. Bloch, and D. F. Smith. 1974. Culture medium for enterobacteria. J. Bacteriol. 119:736–747. Neidhardt, F. C., and H. E. Umbarger. 1996. Chemical composition of Escherichia coli, p. 13–16. In F. C. Neidhardt, R. Curtiss III, J. L. Ingraham, E. C. C. Lin, K. B. Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella typhimurium: cellular and molecular biology, 2nd ed. American Society for Microbiology, Washington, D.C. Olsen, R. H., J. S. Siak, and R. H. Gray. 1974. Characteristics of PRD1, a plasmid-dependent broad host range DNA bacteriophage. J. Virol. 14:689– 699. Pakula, T. M., J. Caldentey, C. Gutierrez, V. M. Olkkonen, M. Salas, and D. H. Bamford. 1993. Overproduction, purification, and characterization of DNA-binding protein P19 of bacteriophage PRD1. Gene 126:99–104. Pakula, T. M., J. Caldentey, M. Serrano, C. Gutierrez, J. M. Hermoso, M. Salas, and D. H. Bamford. 1990. Characterization of a DNA binding protein of bacteriophage PRD1 involved in DNA replication. Nucleic Acids Res. 18:6553–6557. Rydman, P. S., and D. H. Bamford. 2000. Bacteriophage PRD1 DNA entry uses a viral membrane-associated transglycosylase activity. Mol. Microbiol. 37:356–363. Rydman, P. S., and D. H. Bamford. 2003. Identification and mutational analysis of bacteriophage PRD1 holin protein P35. J. Bacteriol. 185:3795– 3803. Rydman, P. S., J. K. Bamford, and D. H. Bamford. 2001. A minor capsid protein P30 is essential for bacteriophage PRD1 capsid assembly. J. Mol. Biol. 313:785–795. Savilahti, H., J. Caldentey, K. Lundstrom, J. E. Syva ¨oja, and D. H. Bamford. 1991. Overexpression, purification, and characterization of Escherichia coli bacteriophage PRD1 DNA polymerase. In vitro synthesis of full-length PRD1 DNA with purified proteins. J. Biol. Chem. 266:18737–18744. Schachtele, C. F., C. V. De Sain, L. A. Hawley, and D. L. Anderson. 1972. Transcription during the development of bacteriophage 29: production of host- and 29-specific ribonucleic acid. J. Virol. 10:1170–1178. Senthil, G., H. Liu, V. G. Puram, A. Clark, A. Stromberg, and M. M. Goodin. 2005. Specific and common changes in Nicotiana benthamiana gene expression in response to infection by enveloped viruses. J. Gen. Virol. 86:2615– 2625. Stevenson, G., K. Andrianopoulos, M. Hobbs, and P. R. Reeves. 1996. Organization of the Escherichia coli K-12 gene cluster responsible for production of the extracellular polysaccharide colanic acid. J. Bacteriol. 178: 4885–4893. Stro ¨msten, N. J., D. H. Bamford, and J. K. Bamford. 2005. In vitro DNA packaging of PRD1: a common mechanism for internal-membrane viruses. J. Mol. Biol. 348:617–629. Susskind, M. M., A. Wright, and D. Botstein. 1971. Superinfection exclusion by P22 prophage in lysogens of Salmonella typhimurium. II. Genetic evidence for two exclusion systems. Virology 45:638–652. Takano, T., and T. Kakefuda. 1972. Involvement of a bacterial factor in morphogenesis of bacteriophage capsid. Nat. New Biol. 239:34–37. Tani, T. H., A. Khodursky, R. M. Blumenthal, P. O. Brown, and R. G. Matthews. 2002. Adaptation to famine: a family of stationary-phase genes revealed by microarray analysis. Proc. Natl. Acad. Sci. USA 99:13471–13476. Walters, K. A., M. A. Joyce, W. R. Addison, K. P. Fischer, and D. L. Tyrrell. 2004. Superinfection exclusion in duck hepatitis B virus infection is mediated by the large surface antigen. J. Virol. 78:7925–7937. Wei, Y., J. M. Lee, C. Richmond, F. R. Blattner, J. A. Rafalski, and R. A. LaRossa. 2001. High-density microarray-mediated gene expression profiling of Escherichia coli. J. Bacteriol. 183:545–556. Whitaker-Dowling, P., J. S. Youngner, C. C. Widnell, and D. K. Wilcox. 1983. Superinfection exclusion by vesicular stomatitis virus. Virology 131:137–143. Whitham, S. A., S. Quan, H. S. Chang, B. Cooper, B. Estes, T. Zhu, X. Wang, and Y. M. Hou. 2003. Diverse RNA viruses elicit the expression of common sets of genes in susceptible Arabidopsis thaliana plants. Plant J. 33:271–283. Yura, T., and K. Nakahigashi. 1999. Regulation of the heat-shock response. Curr. Opin. Microbiol. 2:153–158. ˇ iedaite¨, G., R. Daugelavicˇius, J. K. H. Bamford, and D. H. Bamford. 2005. Z The holin protein of bacteriophage PRD1 forms a pore for small-molecule and endolysin translocation. J. Bacteriol. 187:5397–5405.
