INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Actinobacillus pleuropneumoniae is the causative agent of an acute necrotizing hemorrhagic pleuropneumonia in swine (13, 19, 22). Unfortunately, there is not an abundance of knowledge about what gene products play an important role in A. pleuropneumoniae disease. Our laboratory has developed an in vivo expression technology (IVET) system to identify gene products that have a role in the pathogenesis of swine pleuropneumonia (6). This IVET system has been used to identify gene promoters that are specifically induced during infection.

This IVET approach is based on the complementation of a defined attenuated riboflavin-requiring auxotroph (Rib) by a promoter trap vector that contains promoterless copies of the genes necessary to complement the genetic lesion in riboflavin synthesis. If the fragment of A. pleuropneumoniae genomic DNA ligated into the vector contains a functional promoter, the rib genes are expressed and the auxotroph is able to survive and cause disease in experimentally infected pigs. The goal of our IVET studies is to recover clones from infected pigs, to characterize the promoters that are selected, and to determine what gene(s) lies downstream of each promoter to identify its role in A. pleuropneumoniae pathogenesis. A part of this work is to characterize these in vivo-expressed promoters and compare their structure to that of housekeeping gene promoters. However, the housekeeping or sigma-70 promoter structure in A. pleuropneumoniae is unknown.

The sigma-70 promoters in Escherichia coli are characterized by two nucleotide sequences that are centered at positions 35 and 10 relative to the transcriptional start site. The accepted consensus sequences are TTGACA and TATAAT for the 35 and 10 regions, respectively. These sequences are separated by 17 ± 1 nucleotides (10, 11, 16). There is a similar structure for other well-studied organisms such as Bacillus subtilis (12), but the research on promoter structures in pathogens such as A. pleuropneumoniae is limited.

Only a few attempts have been made to identify promoter elements in A. pleuropneumoniae, and none of the sigma factors have been identified or characterized (5, 9, 14). The results from these experiments show no clear similarity to E. coli promoters or to promoters from other eubacteria. This raises the question as to the structure of a sigma-70-like promoter in A. pleuropneumoniae and how it compares with other eubacterial promoters. Since several genes encoding antibiotic resistance markers that are readily expressed in E. coli are not expressed in A. pleuropneumoniae (25), it is likely that the A. pleuropneumoniae consensus promoter does differ in some way from that found in E. coli.

The goal of this study was to identify and characterize promoter sequences active in A. pleuropneumoniae under standard laboratory growth conditions. We identified DNA fragments with promoter activity by their ability to express promoterless lux genes and identified their transcriptional start sites by primer extension analysis. We have compared the DNA sequences of these active promoters and propose a consensus promoter sequence for A. pleuropneumoniae.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains, media, and growth conditions. The A. pleuropneumoniae strains used in this study were AP100, a virulent serotype 1 strain (ATCC 27088), and AP233, a riboflavin-requiring derivative of AP100 (7). A. pleuropneumoniae strains were cultured at 35°C in a waterbath with shaking at 150 rpm in brain heart infusion (BHI; Difco Laboratories, Detroit, Mich.) containing NAD (10 µg/ml; Sigma Chemical Company, St. Louis, Mo.). Riboflavin (Sigma) was added to a final concentration of 200 µg/ml when necessary. E. coli XL1-blue MRF (Stratagene, La Jolla, Calif.), used for plasmid construction and analysis, was cultured in Luria-Bertani (LB) medium. Ampicillin was added to 100 µg/ml for plasmid selection in E. coli and to 50 µg/ml for selection in A. pleuropneumoniae.

Promoter trap vector pTF86. The promoter trap vector pTF86 was developed as an IVET vector (6). The vector contains a T4 terminator, a unique BamHI cloning site, promoterless luciferase genes (luxAB) genes from Vibrio harveyi, and promoterless riboflavin genes (ribBAH) from B. subtilis in shuttle vector pGZRS19 (25). This plasmid is capable of replicating in both A. pleuropneumoniae and E. coli. The copy number of pTF86 was shown to be 8 to 10 copies per cell in A. pleuropneumoniae (6). We have not measured the copy number of pTF86 in E. coli, but qualitative analysis indicates that the copy number is substantially higher than in A. pleuropneumoniae.

Construction of an A. pleuropneumoniae promoter library. Chromosomal DNA of AP100 was prepared and partially digested with Sau3A. Fragments ranging from 0.4 to 1.0 kb in size were purified from an agarose gel and ligated into the alkaline phosphatase-treated BamHI site of pTF86 (6). The ligation mixture was electroporated into AP233 as previously described (6). Single colonies were selected and subcultured onto BHI plates containing riboflavin (200 µg/ml), ampicillin (50 µg/ml), and NAD (10 µg/ml).

Screening of the promoter library. Qualitative screening of transformants for expression of the luxAB genes in vitro was performed on a Hamamatsu C1966 photonic microscope system. Colonies on agar plates were exposed to 50 µl of N-decyl aldehyde (Sigma) distributed evenly on a glass petri dish lid. The plate was then analyzed using the standard settings for high Lux expression colonies. The camera aperture was set at 11, the intensity setting on the keyboard was set at 8, the bit select was set at 0 to 3, and the offset was set at +1. Colonies that exhibited a high degree of Lux expression compared with known standard A. pleuropneumoniae Lux expression controls were selected for further analysis.

RNA isolation. RNA isolation was performed as described by Xiong et al. (27) with modifications. Briefly, A. pleuropneumoniae cells containing the hot promoter plasmids were grown in 5 ml of BHI medium supplemented with NAD, riboflavin, and ampicillin as described above. The cells were grown to an optical density at 520 nm of 0.8 at 37°C. Then 1.5 ml of the culture was transferred to a cold microcentrifuge tube, and 2 µl of chloramphenicol (20 mg/ml) was added. Cells were pelleted by centrifugation for 30 s at 13,000 × g. The cell pellet was resuspended in 200 µl of STET buffer (18). and 200 µl of phenol-chloroform by vortexing for 30 s. Samples were placed at 100°C for 1 min. All of the subsequent steps were performed in a cold room at 4°C. The samples were centrifuged for 3 min at 13,000 × g. The aqueous phase was removed, extracted with an equal volume of chloroform, and centrifuged again for 3 min at 13,000 × g. The aqueous phase was precipitated by adding 100 µl of 7.5 M ammonium acetate and 600 µl of 100% ethanol. After incubation at 80°C for 15 min, RNA was pelleted by centrifugation for 10 min at 13,000 × g. The pellet was washed with absolute ethanol, resuspended in 20 µl of diethyl pyrocarbonate (DEPC)-treated H₂O with RNase inhibitor (Promega, Madison, Wis.), and stored at 80°C until reverse transcription reactions were performed.

Primer extension. Primer extension analysis was performed using RNA isolated as above that was not more than 1 day old. The primers were 5'-end labeled using [-³²P]ATP and T4 polynucleotide kinase (Gibco-BRL, Rockville, Md.) (20) and purified using Centri-step spin columns (Boehringer-Mannheim). Following end labeling and purification, 1 to 2 µl of labeled primer was incubated with 10 µl of RNA at 85°C for denaturation. The mixture was centrifuged for 10 s at 13,000 × g and allowed to cool to room temperature for 1 h to allow the primer to anneal to the mRNA. Then 4 µl of avian myeloblastosis virus (AMV) reverse transcriptase buffer, 2 µl of DEPC-treated H₂O, 1 µl of 1.25 mM deoxynucleoside triphosphate mix, and 1 µl of AMV reverse transcriptase (Boehringer Mannheim Biochemicals, Indianapolis, Ind.) were added, and the mixture was incubated for 1 h at 42°C. The reaction was stopped by adding 4 µl of STOP solution, provided in Sequenase 2.0 kits (U.S. Biochemical, Cleveland, Ohio).

DNA sequencing. For sequencing the promoter fragments, 5 ml of E. coli XL1-Blue MRF containing the appropriate promoter fragment plasmid was grown for 16 h in LB medium supplemented with ampicillin (100 µg/ml). Plasmid DNA was prepared using Qiagen spin columns (Qiagen Inc., Chatsworth, Calif.) and eluted in 100 µl of water. Plasmids were sequenced with the same primer that was used for reverse transcription by the dideoxy chain termination method using the Sequenase 2.0 kit (U.S. Biochemical) and [³⁵S]dATP (Amersham Corp, Arlington Heights, Ill.). The products of the sequencing reactions and the reverse transcription reaction were separated on an 8.0% polyacrylamide gel containing 8 M urea to analyze the mRNA 5' ends of each promoter fragment clone. The gels were exposed to film for 24 h, and the mRNA 5' ends were determined.

DNA and protein sequence analysis. Nucleotide and amino acid sequences were compared with sequences in the GenBank and EMBL databases using the FASTA and BLAST programs (1). Alignments of the promoter regions were done with MultiAlin version 5.4.1 (3) and the Genetics Computer Group (GCG) suite of programs (8).

Construction of promoter deletion mutant. A deletion mutant (SD2P) was constructed from clone SD2 using a PCR strategy that utilized long-range PCR with primers that contained internal NotI sites. Two primers were designed to prime away from the promoter region. A long-range PCR kit (Boehringer Mannheim) was used to amplify the entire plasmid from the specific primers. The products were then digested with NotI (Boehringer-Mannheim) and ligated overnight at 4°C using T4 DNA ligase (Boehringer-Mannheim). The ligation mixture was electroporated into AP233, and individual transformants were screened for Lux expression as described above. SD2P was sequenced as described above.

Construction of a functional promoter in the deletion mutant SD2P. Two primers were designed to reconstruct the promoter region in SD2P. The primers were used to amplify the promoter region from SD2 by PCR. The primers contained internal NotI sites, and following digestion with NotI for 90 min at 37°C, the promoter fragment was ligated into NotI-digested SD2P. The ligation mixture was electroporated into AP233, and the transformants (SD2PR) were screened for Lux expression. SD2PR was sequenced as described above.

Construction of promoter clones from two characterized E. coli promoters. Two pairs of oligonucleotide primers were designed to clone the bla promoter from pUC18 and the promoter from the str operon from the E. coli genome. The str promoter has a high degree of similarity with the E. coli consensus promoter sequence (16). These two promoters were amplified by PCR and cloned into the BamHI site in pTF86.

In vitro site-directed mutagenesis of asd promoter clone. Four pairs of oligonucleotide primers were designed that encode four specific mutations in the spacer region of the asd promoter clone. Each primer pair primes in opposite directions away from the promoter region on opposite strands of the plasmid. The primer pairs allow amplification of the entire asd plasmid and incorporation of the desired mutations in the spacer region. The primers were used in a system modified from the QuikChange site-directed mutagenesis method (Stratagene). The asd promoter plasmid was purified from E. coli XL1-Blue MRF and used as the template in four PCRs. The PCR cycling conditions were an initial denaturation at 95°C for 30 s, followed by 15 cycles of 95°C for 30 s, 55°C for 1 min, and 68°C for 18 min. The reactions were then cooled to 37°C, and 10 U of DpnI was added to each reaction to digest the methylated template asd plasmid. Following digestion with DpnI, a portion of each reaction was used to transform competent E. coli XL1-Blue MRF. The transformants were screened and sequenced to confirm the presence of the desired mutation in the spacer region.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Construction and screening of an A. pleuropneumoniae promoter library. The promoter trap vector used in this study was developed as an IVET vector (6). This vector, designated pTF86, contains a T4 terminator, a unique BamHI cloning site, promoterless V. harveyi luxAB genes, and promoterless ribBAH genes from B. subtilis in a shuttle vector, pGZRS19 (25), capable of replicating in both A. pleuropneumoniae and E. coli. When a DNA fragment with an active A. pleuropneumoniae promoter is inserted in the appropriate orientation into the BamHI cloning site, the luxAB genes and the ribBAH genes are expressed. This restores to AP233, a riboflavin-requiring derivative of virulent A. pleuropneumoniae serotype 1 (ATCC 27088), the ability to grow in the absence of riboflavin, but more importantly for this study, these clones have Lux activity. A library of potential promoter clones was constructed in pTF86 and transformed into AP233, and clones were screened qualitatively for Lux activity by photonic camera (Fig. 1). Clones with a high degree of Lux expression, such as the asd and TF7 clones shown in Fig. 1, were chosen for further study. Inserts from each of these hot promoter clones were sequenced, using a primer complementary to the 5' end of the luxA coding region as well as additional internal primers as needed. The nucleotide and predicted amino acid sequences of these inserts were used to search the GenBank and EMBL databases, and putative identification of the genes encoded was made based on homology data (Fig. 2).

View larger version (52K): [in this window] [in a new window]   FIG. 1.   Qualitative screen of an A. pleuropneumoniae promoter library. (A) Photograph of duplicate colonies on an agarose plate. (B) Photograph of luminescence using a photonic camera. The asd clone and the TF7 clone have a high degree of lux expression. The iviA clone, which was recovered from an IVET experiment, is an example of a clone with low lux expression (6). The vector pTF86, containing no insert, is also shown.

View larger version (28K): [in this window] [in a new window]   FIG. 2.   Alignment of 12 promoter regions from A. pleuropneumoniae. The promoter regions marked with an asterisk were identified in previous studies. The 10 and 35 regions are shown in bold, and the consensus sequence is shown at the bottom. The experimentally found transcriptional start sites are underlined and capitalized. Dashes indicate gaps to align the promoter regions. The distance between each transcriptional start site and the start codon and the sequence of the RBS for each of the clones is shown. The relative strength of each promoter clone is given as Lux activity in RLU/OD520. These measurements were performed on three different days and are averages of 15 measurements. The standard deviation was calculated, and the error for each measurement is given. The database homology is given for the promoter clones. The homologies are all based on comparison with H. influenzae. Clone SD7 is homologous to HI0017, which is a hypothetical open reading frame in H. influenzae. Clone SD2 is homologous to lcfA, a long-chain fatty acid coenzyme A ligase; TF8 is homologous to kdtA, a 2-keto-3-deoxyoctolosonic acid transferase involved in lipopolysaccharide biosynthesis; SD20 is homologous to top3, a DNA topoisomerase; SD9 is homologous to xerB, an aminopeptidase; and TF7 is homologous to dnaB, a DNA helicase. The percent identity and the amino acid (aa) overlap are also given. Clone SD10 has a short open reading frame (18 codons), and no homologous protein was found. N/A, data not available; ND, not done. GenBank accession numbers: SD2, AF275726; SD7, AF275727; SD9, AF275728; SD10, AF275729; SD20, AF275730; TF7, AF275731; and TF8, AF275732.

Identification of mRNA 5' ends or putative transcriptional start sites. To identify the approximate distance of each mRNA 5' end from the luxAB genes, a primer was designed that is complementary to the 5' end of the luxA gene. A reverse transcription reaction was performed on all of the promoter clones using this primer. The products were run next to a [-³²P]ATP-labeled 100-bp ladder on an 8.0% polyacrylamide gel. This step allowed us to design primers that were close enough to each transcriptional start site to obtain results with single-base-pair accuracy. Following sequence analysis of each clone, primers were designed complementary to the coding strand for each clone and within 100 bp of the mRNA 5' end. Primer extension experiments were completed, and the cDNA product from each of the reverse transcription reactions was run next to a sequencing ladder that used the same primer as the reverse transcription reaction. Figure 2 shows the DNA sequences of the regions upstream from the putative transcriptional start sites found in primer extension experiments. Examples of the autoradiographs of the primer extension gels are shown in Fig. 4 and 5.

Identification of a consensus promoter sequence. In Fig. 2, the promoters from the nine promoter clones analyzed in this study are aligned with three promoters identified in previous research (5, 9, 15). The alignment was performed by using MultiAlin version 5.4.1 and the GCG suite of programs (3, 8). The consensus sequence determined from this alignment consists of nucleotides that occur in more than 50% of the clones at any position. The two regions identified in Fig. 2 are the 10 and 35 regions. The 10 region is positioned 5 to 12 bp upstream of the transcription start site. The consensus sequence of this region is TATAAT. The spacing between the 10 and 35 region ranged between 13 and 16 bp. The consensus sequence of the 35 region was TTRAA, where R can be either G or A. The A. pleuropneumoniae consensus promoter sequence was compared with consensus housekeeping promoters for E. coli (11), B. subtilis (12), Streptomyces sp. (21), Campylobacter jejuni (26), Corynebacterium glutamicum (20), Caulobacter crescentus (17), and Mycobacterium paratuberculosis (2) (Table 1).

Analysis of RBS and transcriptional start site of A. pleuropneumoniae promoters. The initiating nucleotide in 11 of the 12 promoter clones was predicted to be A or T. The ribosome-binding site (RBS) for each of the promoter clones is identified in Fig. 2. Using 16S rRNA sequences the predicted RBS for A. pleuropneumoniae is AGGAGG (4). The promoter clones identified in this study have a consensus RBS of AGGAGG.

Construction of promoter deletion clone to verify loss of promoter function. A promoter mutant was constructed to verify that the deletion of a putative promoter would cause loss of Lux expression and the lack of a cDNA product in a reverse transcription experiment. Clone SD2 was chosen because it has relatively high Lux expression and there were convenient sites to design primers inside the promoter fragment. The primers that were designed are shown schematically in Fig. 3. The primers were designed to delete a 114-bp fragment that included the putative transcriptional start site, the 10 region, the 35 region, and some upstream and downstream sequence. The deletion does not remove the binding site for the primer used in the primer extension mapping experiments. Long-range PCR was performed, and 20 transformants (SD2P1 to SD2P20) were chosen for further analysis. Of the 20 transformants that were chosen for Lux screening, all had no Lux expression when visualized with the photonic camera. A photonic camera image of AP233 clones containing the deleted promoter and the wild-type promoter is shown in Fig. 3 as well as the quantitative Lux expression data. One of the SD2P clones was sequenced to confirm that the promoter region was indeed deleted and the remainder of the genomic DNA fragment was still intact. Primer extension analysis of SD2 (lane 1) as well as SD2P (lane 2) is shown in Fig. 4. The sequencing ladder was prepared from the same primer that was used for the reverse transcription of both SD2 and SD2P. The loss of the cDNA product in lane 2 is further evidence that the region deleted contained the sequences necessary for the initiation of transcription in A. pleuropneumoniae.

View larger version (12K): [in this window] [in a new window]   FIG. 3.   Construction of a promoter deletion mutant from clone SD2. Map of the insert structure of the wild-type SD2 clone and SD2P. The 35 and 10 regions are noted as well as the transcriptional start site (Tx). Sequence analysis of the mutant showed that a total of 114 bp had been deleted, ranging from 82 bp upstream of the transcriptional start site to 32 bp downstream of the start site. The deletion did not delete the primer-binding site for the reverse transcription primer (MM116). (A) Photograph of the colony growth of each clone. (B) Photonic camera image.

View larger version (63K): [in this window] [in a new window]   FIG. 4.   Primer extension analysis of clones SD2 (lane 1) and SD2P (lane 2). The sequencing reactions and the reverse transcription reactions were performed with an identical primer. Lane 1 contains the reverse transcription reaction from the SD2 clone, which contains a wild-type promoter. Lane 2 is the reverse transcription reaction from the SD2P clone that contains the promoter deletion. The start site is highlighted with an arrow, and the 10 and 35 regions are bracketed.

Comparison of transcriptional start site identified from plasmid mRNA and from AP233 containing the endogenous copy of asd gene. We compared the mRNA 5' ends that were identified from AP233 containing no plasmids and AP233 containing the asd plasmid. This was done to verify that the transcriptional start sites mapped from the plasmid clones were identical to the start sites mapped from A. pleuropneumoniae containing only the endogenous copy of the gene. The results of the primer extension experiments are shown in Fig. 5. Lane 1 shows the putative transcriptional start site mapped from total cellular RNA isolated from AP233 containing no plasmids, and lane 2 shows the putative transcriptional start site identified from the asd promoter clone. The same primer was used in each reaction, and each putative transcriptional start site maps to the same position.

View larger version (78K): [in this window] [in a new window]   FIG. 5.   Comparison of transcriptional start sites mapped from mRNA isolated from AP233 (lane 1) and the asd clone (lane 2). The same primer was used for reverse transcription reactions and sequencing reactions. The reaction in lane 2 was done with 15 µg of total RNA isolated from AP233, and the reaction in lane 1 was completed with 1.5 µg of total RNA isolated from the asd clone. The start site is highlighted with an arrow, and the 10 and 35 regions are bracketed.

Quantitative expression of E. coli promoters in E. coli and A. pleuropneumoniae. Two E. coli promoters were cloned in pTF86, and their expression was analyzed by quantitative Lux assays (Table 2). Clone pSDbla contains the -lactamase promoter from pUC18, and clone pSDstr contains an E. coli promoter from the str operon that is very similar to the E. coli consensus promoter. Both promoters have a spacing of 17 bp, and the sequences of their 10 and 35 elements are shown in Table 2. Both promoters were strongly expressed in E. coli, as predicted, but were only weakly expressed in A. pleuropneumoniae.

In vitro site-directed mutagenesis of asd promoter clone to assess expression differences based on spacer length. Site-directed mutagenesis was performed on the asd promoter clone to alter the spacer length of the clone. Four mutants were constructed by inserting or deleting adenosine nucleotides with spacing between the 10 and 35 elements ranging from 14 to 18 bp. The sequences of each of the promoter mutants as well as the wild-type asd promoter clone and Lux expression for each of the clones in E. coli and A. pleuropneumoniae are shown in Table 2. Note that a spacing of 18 bp in the A. pleuropneumoniae asd promoter should be equivalent to a spacing of 17 bp in an E. coli promoter, due to the different lengths of the 35 element in these species. In E. coli, expression increased with increased spacing, and the strongest expression was found in the asd18 promoter. In contrast, in A. pleuropneumoniae expression was strongest at a spacing of 16 bp, with high expression at shorter spacings but no significant expression in the asd18 promoter.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In this study we identified nine strongly expressed or hot promoters from A. pleuropneumoniae using a promoter trap vector, pTF86. This vector, which encodes the luxAB genes from V. harveyi, allows quantification of Lux expression and therefore the relative strength of the promoter driving the expression. The nine promoter clones were analyzed by primer extension to determine their transcriptional start sites with single-base-pair accuracy. The sequences upstream of the transcriptional start sites were aligned along with three previously identified promoters to determine a consensus promoter sequence for A. pleuropneumoniae.

The consensus sequence for the 10 region is TATAAT, which is identical to that of E. coli and B. subtilis. The 35 region was aligned, and the consensus sequence is TTRAA. The 35 consensus is identical in the first, second, and fourth positions to the sequences of E. coli and B. subtilis. The spacing between the 10 region and the transcriptional start site ranges from 5 to 10 bp, with an average of 7 bp of separation, which is similar to promoters in other eubacteria. The spacing between the 10 and 35 sequences in A. pleuropneumoniae ranges between 13 and 16 bp. This spacing is shorter than that of the promoters found in most other eubacteria.

This difference in spacer length was analyzed by constructing mutants of the asd promoter clone to assess the effects of both inserting and deleting nucleotides in the spacer region. The optimum spacing for the asd promoter in A. pleuropneumoniae was found to be 16 bp, but a high degree of expression was also seen with spacings of 14 and 15 bp (Table 2). However, a significant loss of Lux expression was seen in the clones with spacer lengths of 17 and 18 bp. In comparison, the E. coli clones had their highest Lux expression at 17 and 18 bp of spacing between the 10 and 35 elements. In addition, both the E. coli str promoter and the bla promoter from pUC18 were strongly expressed in E. coli but only weakly expressed in A. pleuropneumoniae. It is clear that there is a difference in the consensus promoter structure between E. coli and A. pleuropneumoniae and that A. pleuropneumoniae promoters have a shorter spacing in the region between the 10 and 35 elements than do those of E. coli.

No sigma factors have been characterized to date in A. pleuropneumoniae. However, it is likely that multiple sigma factors control A. pleuropneumoniae gene expression under different environmental conditions, as in other eubacteria. The promoter clones in this study were chosen for three reasons. First, each has a high degree of lux expression when grown on standard laboratory media. Second, the clones identified in the IVET vector (pTF86) are also expressed in vivo. Third, the seven promoter clones (asd, SD2, ribG, TF8, SD20, SD9, and TF7) that were putatively identified have homologous genes in E. coli that are transcribed by sigma-70 promoters (10). While these facts do not rule out an alternative sigma factor, we expect that these genes are transcribed by a similar housekeeping sigma factor in A. pleuropneumoniae.

To confirm our promoter identification strategy, we compared the mRNA 5' ends that were identified from AP233 containing no plasmids and AP233 containing the asd plasmid. The cDNA products from the reverse transcriptase reactions were identical. This result verified that the putative transcriptional start site that we identified using the promoter trap vector was indeed the same as the promoter used in the A. pleuropneumoniae genome to drive gene expression.

To further confirm our promoter identification, we constructed a deletion mutant from the SD2 clone. In this mutant, SD2P, 114 bp were deleted, including the 35 and 10 regions and the putative transcriptional start site, but not the primer-binding site for the reverse transcriptase reaction or the ATG. The mutant was sequenced, and the region that was deleted was confirmed. There was no Lux expression in this mutant, and no primer extension product was detected. The promoter region from SD2 was cloned back into the SD2P mutant, and the resulting clone, SD2PR, showed Lux expression levels that were identical to that of the wild-type SD2 and an identical primer extension product. Therefore, we conclude that the 114-bp segment contains a true promoter, and we infer that the mRNA 5' end that we mapped represents a transcriptional start site.

The nine promoters that were identified in this study demonstrated various degrees of promoter strength. However, the strength of each could not be directly related to how similar the 10 and 35 sequences were to the proposed consensus sequence. This variability in expression suggests that there are other factors that modulate the strength of individual promoters. In other eubacteria, there are many factors that can affect the degree of promoter activity. The UP element and the 16 and 45 sequences modulate the activity of individual promoters and can do so with expression differences of up to 100-fold, depending on the changes that are made (21, 24). Additionally, the spacer region between the 10 and 35 sequences plays a pivotal role in recognition of the promoter and also the promoter strength. In addition to the number of nucleotides in the spacer region, it has been shown that the sequence of the spacer has an effect on strength (14). We have shown with the asd promoter clone mutant data that by inserting or deleting a single base pair, we can alter the expression by up to sevenfold in A. pleuropneumoniae. We theorize that additional factors contribute to modulation of A. pleuropneumoniae gene expression.

This study has identified a consensus promoter structure for general housekeeping genes in A. pleuropneumoniae. This can be used to identify additional A. pleuropneumoniae promoters and provides a baseline for future transcriptional studies in A. pleuropneumoniae. In the process of characterizing the promoter structure, we have shown that A. pleuropneumoniae promoters are different from those of E. coli and B. subtilis. The promoters of A. pleuropneumoniae have a shorter spacing between the 10 and 35 elements. A longer spacing of 17 or 18 bp drastically reduced promoter activity. This finding explains why some E. coli promoters do not function well in A. pleuropneumoniae and demonstrates that not all eubacterial consensus promoters are identical to those identified for E. coli and B. subtilis.