![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Biomedical Genomics Center and Department of Veterinary Pathobiology, University of Minnesota, St. Paul, Minnesota 55108
Received 23 October 2001/ Accepted 22 March 2002
 |
ABSTRACT |
|---|
 Top Abstract TextReferences |
|---|
|
TEXT |
|---|
|
TopAbstractTextReferences |
|---|
One of the first observed regulatory responses to nutrient limitation in Escherichia coli was a decrease in RNA accumulation upon amino acid starvation. It was later found that the nucleotides ppGpp (guanosine 3',5'-bispyrophosphate) and pppGpp (guanosine 3'-diphosphate, 5'-triphosphate) accumulated under starvation conditions (3). Control of ppGpp and pppGpp levels appears to be mediated by the enzymes RelA and SpoT, which are themselves regulated by the ratio of charged to uncharged tRNAs (2). Previous research on the response of E. coli to conditions of limited nutrients has shown that the transcription of genes involved in amino acid biosynthesis and transport, sulfur fixation, and stress tolerance increased in minimal medium (20, 22). In contrast, the expression of genes responsible for small-molecule degradation, ATP generation, and protein synthesis and folding was generally higher in rich medium (20, 22).
Pasteurella multocida is a gram-negative, nonmotile, rod-shaped, facultative anaerobe that has been isolated from a wide range of mammals and birds throughout the world. This organism is the etiologic agent of a variety of economically significant diseases, including fowl cholera in poultry, hemorrhagic septicemia in cattle and buffalo, atrophic rhinitis in swine, and snuffles in rabbits (15, 18). The organism's global distribution, the severity of the diseases it causes, and the wide variety of livestock affected by P. multocida account for considerable economic losses caused by this pathogen worldwide (18).
The elucidation of the complete genome sequence of P. multocida (14) has allowed researchers to utilize new strategies towards the identification and characterization of genes involved in the pathogenesis of this bacterium. Microarrays are especially well suited for monitoring the complex changes in bacterial gene expression that occur in response to different environmental conditions. Recent work by our lab has utilized microarray technology to identify genes in P. multocida that are transcriptionally altered in response to iron limitation (17). In the current study, we have used microarrays consisting of all 2,014 identified P. multocida open reading frames (ORFs) to monitor gene expression in response to growth in rich and minimal media. This approach has enabled us to identify the repertoire of genes whose expression is altered in response to a change in nutrient availability.
Bacterial growth, RNA isolation, and microarray analyses.
Chemically defined medium (CDM) was prepared as described previously (8) with the following modifications: CaCl2 and MgCl2 were filter sterilized and added after the medium was treated with 5 g of Chelex-100 (Bio-Rad, Hercules, Calif.) per liter for 2 h on ice, and FeSO4 was omitted. P. multocida Pm70 was grown to log phase in a flask of brain heart infusion (BHI) medium (Becton Dickinson, Franklin Lakes, N.J.) at 37°C. The culture was split into two 200-ml volumes, briefly centrifuged at 4°C, washed with full-strength phosphate-buffered saline (pH 7.0), and centrifuged again. One pellet was resuspended in 175 ml of BHI, and the other was resuspended in 175 ml of CDM. The resuspended cultures were incubated on a rotary shaker at 37°C, and 8-ml volumes were removed at 1, 5, 10, 15, 30, 60, and 90 min after resuspension. These samples were briefly centrifuged at 4°C, and the pellets were flash frozen in dry ice and ethanol. Total-RNA extractions were performed with RNeasy mini columns (QIAGEN, Valencia, Calif.), with DNase digestions being done on column by adding 82 Kunitz units of enzyme (QIAGEN) and incubating the columns at room temperature for 15 min. Gene expression analysis with DNA microarrays was performed as described at our website(http://www.agac.umn.edu/microarray/protocols/protocols.htm). In brief, a library of targets representing all 2,014 ORFs from P. multocida Pm70 (GenBank accession no. AE004439) was constructed with primers designed to amplify fragments of 
500 bp from each ORF from genomic DNA. Two successive rounds of PCR were performed to minimize genomic-DNA contamination in the products of amplification, and the final 100-µl reaction mixtures were checked for quality on agarose gels and purified by using MultiScreen PCR plates (Millipore, Bedford, Mass.). All 2,014 (100%) ORF segments were successfully amplified and printed in triplicate onto poly-L-lysine-coated slides by a Microgrid II robot (BioRobotics, Boston, Mass.). RNA from P. multocida grown in BHI alone or in CDM was indirectly labeled with Cy5 or Cy3, respectively, by incorporating an amino-allyl dUTP during reverse transcription followed by chemical coupling to the dyes. The labeled cDNAs from two independent experiments were then competitively hybridized with the printed microarrays. Thus, a minimum of six data points were collected for each of the 2,014 ORFs for the seven time points examined in order to obtain a data set amenable to robust statistical analyses (see below). Images of the hybridized arrays were obtained with a Scanarray 5000 microarray scanner (GSI Lumonics, Watertown, Mass.). Individual spots on the arrays were flagged for quality during the analysis of the scanned images. The local background fluorescence was subtracted from each spot, and the fluorescent intensities for individual spots were normalized based on the median intensities of fluorescence in the Cy3 and Cy5 channels. Replicate spots that had Cy3-to-Cy5 ratios beyond 1.5 standard deviations from the average ratio were discarded, and values for the remaining replicates were then averaged to determine the final ratio reported for each ORF. The statistical significance of the data collected was determined with the Significance Analysis of Microarray data program (21). Data visualization and analysis were performed using DecisionSite 6.1 software (Spotfire, Somerville, Mass.).
Observed alterations in gene expression. A complete set of all results is presented on our website at http://www.agac.umn.edu/Microarray/bhi-cdm/index.htm. Of the 2,014 genes present on the arrays, the levels of expression of 669 were significantly altered (as determined by observing a 1.5-fold change for at least two time points) when P. multocida cultures growing in BHI and CDM were compared. Although statistical analyses of replicates often enable us to detect significant differences that are considerably less than 1.5-fold, as evidenced by the data sets provided, in our experience and that of others (24), the 1.5-fold value provides an extra measure of confidence in ensuring that gene expression was altered.
Of these genes, 230 were expressed at elevated levels in CDM while 439 were expressed at elevated levels in BHI. The remaining 1,345 genes were either not well measured or their levels of expression were not significantly different between the two populations. We observed several general trends among the genes whose expression was altered (Fig. 1). The classification strategy we used was based on the system developed for the Haemophilus influenzae genome (http://www.tigr.org/tigr-scripts/CMR2/GenomePage3.spl?database=ghi). A large number of genes involved in energy metabolism, transport, protein synthesis, and binding were expressed at higher levels in BHI than in CDM. Genes in these categories included hbpA, involved in hemin binding (7), afuABC, an iron ABC transport system (4), the ATP synthase gene operons atpABCDFGH and tadABCEG (10), and several tRNA synthetases. Additionally, the transcription of genes involved in protein, nucleotide, and lipid synthesis generally increased in the rich medium. These results indicate that upon exposure to a limiting nutrient environment, P. multocida immediately begins to turn off many energy-intensive biosynthetic pathways.
|
|
|
Interestingly, several genes involved in tryptophan biosynthesis (trpABCE) were expressed at higher levels when arginine was added to the medium (Fig. 2B). These results complement observations made by Khodursky and coworkers (11) indicating that the regulatory elements controlling genes involved in tryptophan and arginine metabolism may interact to some degree.
Outer membrane proteins. Several genes coding for outer membrane proteins were transcriptionally activated under starvation conditions. ORF PM0059 codes for the filamentous hemagglutinin PfhB2 and is thought to be involved in virulence (13), while PM1570 is similar to the Haemophilus influenzae surface fibril protein Hsf (19). The expression of the nonspecific adherence genes tadA, -B, -C, -E, and -G was also higher in CDM. PM0331 is homologous to the immunogenic Vibrio cholerae outer membrane protein OmpW (9), and RfaE participates in the synthesis of lipopolysaccharide (12). PM0084 and PM0085 also showed increased expression and code for a type 4 fimbrial subunit and a protein transport protein, respectively (5). PM1518 and PM1926 are homologous to outer membrane lipoproteins, and both were expressed at higher levels in minimal medium than in rich medium. Additionally, the hypothetical proteins PM0674, PM1886, and PM1936 showed increased expression and have functional domains similar to lipoprotein lipid attachment sites. Alterations in the outer membrane composition that increase the rate of nutrient diffusion would benefit P. multocida under low-nutrient conditions and have been noted to occur in other bacteria (6). The increased expression of genes involved in attachment indicates that nutrient limitation may serve as one of the signals associated with the regulation of colonization factors.
The expression of a number of genes involved in protein synthesis, transport, and stabilization was altered. Genes related to the heat shock transcription factor RpoH (
32) underwent significant change during the experiment. The gene encoding RpoH was transcribed at higher levels by P. multocida growing in minimal medium. RpoH normally accumulates via increased translation and stabilization upon exposure to an increase in temperature (25), but it is unclear what environmental signals might lead to the observed rise in transcription. Additionally, the genes encoding FtsH and DnaK and -J were expressed at higher levels in CDM than in BHI. DnaK and -J are chaperone proteins that are thought to contribute to the degradation of RpoH by the FtsH protease (1). Thus, it appears that RpoH-mediated transcription was induced in cells shifted to minimal medium and repressed in cells in rich medium.
Hypothetical proteins. Of the genes whose expression was significantly altered, 164 have unknown functions or are homologous only to hypothetical proteins. Especially of interest are those genes that were expressed at higher levels in organisms grown in minimal rather than in rich medium. One of the most intriguing of these genes was PM1623. Its expression was a minimum of 2-fold higher in organisms grown in CDM for the entire experiment, and its maximum increase was 4.8-fold at 10 min after the culture was transferred to minimal medium. No other genes were expressed at such consistently elevated levels in the presence of CDM. A search of publicly available databases revealed no significant matches to any known proteins or functional patterns. The expression profile of PM1623 was similar to those of artI, ompW, dsbA, and rsgA; however, none of these genes were transcriptionally elevated more than 2.9-fold in the presence of CDM during the first 30 min of the experiment. Based on the types of genes that were transcriptionally activated in CDM, we might expect PM1623 to play an important role in nutrient acquisition or biosynthesis. Two other hypothetical proteins of interest were PM1232 and PM1233, which had limited homology to the SNO and SNZ domains, respectively. Proteins containing these domains in Saccharomyces cerevisiae have been shown to respond to nutrient limitation (16). The genes encoding these proteins displayed very distinct expression profiles over the course of the experiment. Both genes were initially expressed at higher levels in CDM than in BHI, but after 15 min their expression rapidly decreased such that by 60 min, it was approximately sixfold higher in BHI. It is unclear why these genes underwent such a dramatic change in expression; one possible explanation is that a nutrient present at critically low levels was rapidly depleted in the minimal medium and triggered the observed change in expression. Further illustrating the limitation of our current understanding of bacterial metabolism, the expression of genes coding for eight other hypothetical proteins with no homologs was significantly increased in CDM.
Real-time RT-PCR analyses. We performed real-time reverse transcription-PCR (RT-PCR) on a selected subset of genes to verify the microarray results. Briefly, RNA from cells grown in BHI or CDM was extracted and reverse transcribed into cDNA as described for the microarray methods. Pairs of primers representing selected genes were then used to amplify the cDNA targets in the presence of SYBR green dye (Molecular Probes, Eugene, Oreg.). Results for the 10 genes examined (argA, artI, trpE, rsgA, PM1623, hbpA, atpA, afuC, dnaK, and tadA) were in accordance with the microarray data 90% of the time. In the one instance of discordance, PM1623 expression in organisms grown in CDM was noted to be elevated 4.85-fold at the 10-min time point according to the microarray analyses but was not significantly altered (17.43 versus 17.46 threshold cycle [Ct]) according to the RT-PCR. Given the large number of replicates carried out for the microarray data and the relatively small Ct number for PM1623, which makes the differentiation of expression levels more difficult, it is likely that greater weight should be given to the results of the microarray analyses.
Concluding comments. Profiling of global gene expression in bacteria can provide us with new insights into bacterial gene regulation. Two of the greatest advantages gained by studying gene expression by a whole-genome-based approach are the elucidation of novel regulatory networks and the identification of hypothetical genes that respond to defined experimental conditions. The results of this type of analysis can then be used to generate new hypotheses and identify genes that require further study and characterization. In our study, we compared levels of gene expression in P. multocida in response to rich and minimal media (Fig. 3). We found that artPIQM, an operon of genes homologous to an arginine transport system in E. coli, was transcribed at elevated levels in minimal medium compared to levels in rich medium or minimal medium supplemented with arginine. This seems to indicate that these genes are activated only in response to severe arginine limitation and may encode a high-affinity transport system that may also be part of a global nutrient stress response system in P. multocida.
|
Of the ORFs that have been identified in P. multocida and have unknown functions, PM1623 was identified as being transcribed at higher levels in response to nutrient limitation. The protein predicted to be encoded by PM1623 has no known function but is homologous to hypothetical proteins in Actinobacillus actinomycetemcomitans, Haemophilus ducreyi, and Neisseria meningitidis. Further investigations are warranted to elucidate the functions of this ORF and other hypothetical ORFs whose transcription was altered. The results from these experiments have provided us with several new insights into P. multocida gene expression in a nutrient-limited environment and many new gene targets for functional analyses.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TopAbstractTextReferences |
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Appl. Environ. Microbiol. | Infect. Immun. | Eukaryot. Cell |
|---|---|---|
| Mol. Cell. Biol. | J. Virol. | Microbiol. Mol. Biol. Rev. |
| ALL ASM JOURNALS |
