Transcriptional analysis of the 5' terminus of the flp fimbrial gene cluster from Actinobacillus actinomycetemcomitans

Elaine M. Haase, Joshua O. Stream and Frank A. Scannapieco

Department of Oral Biology, School of Dental Medicine, 109 Foster Hall, University at Buffalo, State University of New York, Buffalo, NY 14214, USA

Correspondence
Elaine M. Haase
haase{at}acsu.buffalo.edu


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Fresh isolates of the oral bacterial pathogen Actinobacillus actinomycetemcomitans exhibit a fimbriated, rough colony phenotype. Evidence suggests that the fimbrial subunit gene flp is part of a cluster of 14 genes (flp to tadG) thought to encode proteins involved in the synthesis, assembly and export of these fimbriae. To determine the transcriptional organization of the 5' terminus of this gene cluster, total RNA from rough and smooth phenotype variants of A. actinomycetemcomitans strain 283 were analysed by RT-PCR. Primers designed to amplify regions spanning gene junctions or multiple genes yielded amplicons at each individual gene junction from flp to tadD for both the rough and smooth variants. Semi-quantitative RT-PCR of the rcpA to tadZ amplicon revealed that significantly more mRNA was transcribed from the rough than the smooth variant. Longer amplicons encompassing flp to tadZ (3·9 kb) and tadA to tadD (2·1 kb) were also detected, but only from the rough variant. Rapid amplification of cDNA ends (RACE) was used to identify the 5' end of the mRNA containing flp. Antisense primers located within rcpC, orfB and flp-2 enabled amplification of a RACE product that was subsequently isolated and subcloned into pGEM-T. DNA sequencing indicated that the 5' end of the mRNA was located at a G or T nucleotide -102 to -101 nt upstream of flp. Corresponding {sigma}70 consensus sequences were located at -10 (TATAAT) and -35 (TTGCAT) relative to the transcription start site. These data confirm that the flp gene cluster is an operon transcribed as a polycistronic message commencing from a G or T nucleotide located in the intergenic region upstream of flp. Promoter function of the flp upstream region was confirmed using a lacZ reporter gene construct transformed into Escherichia coli. RT-PCR analysis further suggests that although transcription does occur in both the rough and smooth variants, full-length transcripts are rapidly degraded or are significantly downregulated in the smooth variant.

Abbreviations: DEPC, diethyl pyrocarbonate; GSP, gene-specific primer; RACE, rapid amplification of cDNA ends


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The oral bacterium Actinobacillus actinomycetemcomitans is a Gram-negative coccobacillus most notably associated with localized juvenile periodontitis, an aggressive form of early onset periodontitis affecting children 10 to 17 years old, and some cases of refractory adult periodontitis (Haraszthy et al., 2000; Slots & Dahlen, 1985; Tanner et al., 1979; Zambon, 1985). It is a member of the HACEK (Haemophilus aphrophilus, Haemophilus parainfluenzae, Actinobacillus actinomycetemcomitans, Cardiobacterium hominis, Eikenella corrodens and Kingella kingae) group of human pathogens responsible for approximately 3 % of endocarditis (el Khizzi et al., 1997; Serra & Tonato, 1969; Steckelberg et al., 1990; Vandepitte et al., 1977) and other extra-oral infections (Burgher et al., 1973; Martin et al., 1967; Muhle et al., 1979; Overholt, 1966; Page & King, 1966). It has also been associated with atheromatous plaques and may play a role in their formation (Haraszthy et al., 2000). Newer niches are being discovered for A. actinomycetemcomitans, yet the molecular pathogenesis of the organism is not well defined.

A. actinomycetemcomitans, like many other bacteria, exhibits phase transition in fimbriae expression when grown in culture (Gally et al., 1993; Marrs et al., 1988; Mims, 1976; Schwan et al., 1992; Snellings et al., 1997). Differential expression of fimbriae and other adhesins is likely in response to environmental factors such as pH, temperature, oxygen or iron concentration (Roosendaal et al., 1986; Scannapieco et al., 1987). Primary isolates of A. actinomycetemcomitans recovered from both oral and systemic infections are of the rough colony phenotype and produce abundant peritrichous fimbriae that aggregate to form bundles, typical of some type IV fimbriae (Fine et al., 1999a; Inouye et al., 1990; Strom & Lory, 1993). In vitro subculture irreversibly converts this organism to a non-fimbriated, planktonic smooth colony phenotype (Fine et al., 1999a; Haase et al., 1999; Inouye et al., 1990; Rosan et al., 1988). In vitro studies have shown that the rough colony variant adheres better to saliva-coated hydroxyapatite, while the smooth colony variant invades oral epithelial cells significantly better than the rough colony variant (Fine et al., 1999b; Fives-Taylor et al., 1995; Meyer et al., 1991; Rosan et al., 1988). Similarly, in an in vivo study of adhesion, the rough phenotype colonized rat teeth significantly longer than the smooth phenotype variant (Fine et al., 2001). Recently, intracellular A. actinomycetemcomitans has been observed within buccal epithelial tissue taken from periodontally normal individuals (Rudney et al., 2001). The phenotype of the intracellular organisms in vivo is not known. While this rough to smooth phenotype conversion has yet to be demonstrated in vivo, it cannot be ruled out; it may play a role in the response of the organism to different host microenvironments. Some of the genes expressed specifically by the rough colony variant are likely involved in initial adhesion and colonization of host mucosal surfaces. Downregulation of these genes may permit phase transition and tissue invasion.

Previously, we have identified two outer-membrane proteins, RcpA and RcpB, expressed exclusively by the rough colony variant (Haase et al., 1999). RcpA is a homologue of the secretin component of the type II secretion system and is possibly involved in pore or channel formation (Hardie et al., 1996; Pugsley, 1993). The function of RcpB remains unknown. These proteins are part of a cluster of at least 14 genes encoding proteins postulated to be responsible for the synthesis, assembly and export of Flp fimbriae (Haase et al., 1999; Inoue et al., 1998; Kachlany et al., 2000, 2001). The 5' terminus of flp shares some homology with type IV fimbriae (Inoue et al., 1998; Skerker & Shapiro, 2000). In addition, the arrangement of structural and secretory genes in a single locus resembles the type IVb pilin genes represented by the bundle-forming pilin (Bfp) of enteropathogenic Escherichia coli and the toxin co-regulated pilin (TCP) of Vibrio cholerae (Ogierman et al., 1993; Stone et al., 1996). However, the flp fimbrial gene cluster of A. actinomycetemcomitans better resembles the recently described gene cluster encoding a novel pilus in Caulobacter crescentus (Skerker & Shapiro, 2000). Transposon mutagenesis of 11 of 14 genes within the flp fimbrial gene cluster in A. actinomycetemcomitans converted a rough colony variant to the smooth colony phenotype (Kachlany et al., 2000, 2001). Based on these data, we are investigating the rough to smooth phenotype conversion at the molecular level by focusing on the transcription of genes within this fimbrial gene cluster. The purpose of this study is to determine if the genes within the flp fimbrial gene cluster are transcribed as a polycistronic message, if the gene cluster is differentially upregulated in the rough phenotype and the location of the putative operon promoter.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains and culture conditions.
Four isogenic rough/smooth pairs of A. actinomycetemcomitans (strains 29, 283, 361 and A26) used in this study were previously described (Haase et al., 1999). A. actinomycetemcomitans strain HK1651, obtained from the American Type Culture Collection, was received as a smooth phenotype. A. actinomycetemcomitans strains were cultivated as previously described (Haase et al., 1999). Briefly, to ensure maintenance of culture phenotype and maximal expression of fimbriae, each strain of A. actinomycetemcomitans was first cultured from frozen stock to Trypticase soy blood agar supplemented with 0·5 % (w/v) yeast extract (Difco), haemin (5 µg ml-1) and menadione (0·2 µg ml-1) (TSBY) for 48–72 h at 37 °C anaerobically (5 % CO2, 10 % H2, 85 % N2) (Haase et al., 1999; Scannapieco et al., 1987). One to five well-isolated colonies with the desired colony morphology were subcultured into a starter culture of Trypticase soy broth (BBL) supplemented with 0·6 % (w/v) yeast extract and 0·04 % (w/v) NaHCO3 and incubated anaerobically at 37 °C for 18–22 h. Cultures were diluted 1 : 10 into fresh TSBY and incubated under the same conditions. All broth cultures were grown statically.

RNA isolation and analysis.
Total RNA was prepared from A. actinomycetemcomitans by a modification of the SDS lysis/CsCl cushion procedure (Deretic & Konyecsni, 1989; Kolodrubetz et al., 1996). A 200 ml culture of rough phenotype cells or a 50 ml culture of smooth phenotype cells was harvested in mid- to late-exponential phase of growth by centrifugation at 8000 g for 15 min. The growth phase was determined by comparing the OD495 with previous growth curves of each strain and phenotype. The cells were washed with cold lysis buffer (50 mM Tris/HCl, pH 7·5) and resuspended in 5 ml lysis buffer. Each sample was passed through a 20-gauge needle five times. One millilitre of 20 % (w/v) SDS was added and vortexed briefly. Samples were incubated for 1–5 min at 37 °C and vortexed once every minute until cells were completely lysed. Then 4 g solid CsCl was added and mixed slowly by gentle inversion for 2 min. An additional 8 ml cold lysis buffer was added and mixed, and the precipitate removed by centrifugation at 15 000 g for 10 min. The supernatants (13 ml) were carefully layered onto 4 ml 5·7 M CsCl cushions. Total RNA was pelleted by ultracentrifugation (Beckman L8-M with a SW28 rotor) at 102 000 g at 20 °C for 28 h. Supernatants were removed by inversion and the pellets dissolved in 100 µl diethyl pyrocarbonate (DEPC)-treated water, 3 vols 95 % (v/v) ethanol were added and RNA was pelleted by centrifugation and stored in ethanol at -70 °C. Pellets were reconstituted in 100 µl DEPC-treated water, aliquoted and stored at -70 °C. RNA was evaluated for quantity and quality at A260 and A280, and on agarose formaldehyde gels.

Northern blot analysis.
A 30 µg sample of total RNA from each isogenic rough/smooth pair of A. actinomycetemcomitans strains 29 and A26 was applied to a 0·7 % (w/v) agarose/formaldehyde gel containing ethidium bromide and run as per standard protocol (ECL system, Amersham Pharmacia). The RNA was transferred to Hybond-N neutral nylon membrane by capillary transfer and hybridized with a 292 bp internal fragment corresponding to nt 406–697 of rcpA. The DNA probe was amplified by PCR from genomic DNA strain 283R using the forward primer (5'-GCTCGTTCACAAGAAGAAAGCC-3') and the reverse primer (5'-ATCCACCTCCGAAACCGAAG-3'), and then randomly labelled with biotin (Invitrogen). The blot was washed under very low stringency conditions. A first wash series at 65 °C in 0·6 M NaCl, 60 mM sodium citrate and 0·5 % (w/v) SDS was followed by a second wash at ambient temperature in 0·3 M NaCl, 30 mM sodium citrate and 0·1 % (w/v) SDS. The blot was developed according the Photogene protocol (Invitrogen) and exposed to BioMax MS film (Kodak).

RT-PCR of adjacent genes.
Prior to first strand synthesis, total RNA (5 µg) was pretreated with 1 U RQ1 RNase-free DNase (Promega) according to the manufacturer's protocol to remove any contaminating DNA. The entire 10 µl sample of treated RNA was used as template to synthesize first strand cDNA as per the manufacturer's protocol (Invitrogen). An antisense gene-specific primer (GSP; 2 pmol) was added to the total RNA (5 µg) and denatured at 70 °C for 10 min. After a quick chill on ice, 4·0 µl 5x times; first strand buffer (0·25 M Tris/HCl, pH 8·3, 0·375 M KCl, 15 mM MgCl2), 2·0 µl 0·1 M DTT and 1·0 µl 10 mM dNTP mix were added and incubated for 2 min at 42 °C. Superscript II reverse transcriptase (200 U; Invitrogen) was added to the reaction tubes only and DEPC-treated water was added to the negative control tubes to bring the total volume to 20 µl. The mixture was incubated for 50 min at 42 °C followed by heat inactivation of the reaction at 70 °C for 15 min. A reverse transcription negative control was included for each RNA template. To analyse each contiguous gene junction, the antisense primer used for first strand synthesis (cDNA) was located approximately 300–420 nt downstream of the translation start codon for each ORF; this primer was designated as the RT primer (Table 1; RT). Prior to PCR, to ensure that the cDNA template was free of RNA complementary to the cDNA, 1 µl (2 U) RNase H (Invitrogen) was added to 10 µl cDNA and incubated at 37 °C for 20 min.


View this table:
[in this window]
[in a new window]
 
Table 1. Oligonucleotide primers used in RT-PCR in this study

 
For PCR amplification of each contiguous gene junction, a forward primer was designed within 250–500 nt upstream of the translation start codon for each gene junction flp-1/flp-2, flp-2/orfAB, orfB/rcpC, rcpC/rcpA, rcpA/rcpB, rcpB/tadZ, tadZ/tadA, tadA/tadB, tadB/tadC and tadC/tadD (see Figs 2a and 3 for the gene cluster organization). Similarly, a nested reverse primer was designed upstream of the appropriate RT primer but still 3' of the start codon (Table 1; F, R). In a 50 µl PCR reaction, 2 µl cDNA was used as template along with 15 pmol each primer. Typically the annealing temperature was five degrees below the Tm for each primer. The PCR program used was as follows: preincubate at 94 °C for 3 min, denature at 94 °C for 1 min, anneal for 1 min, extend at 72 °C for 1·5 min for 45 cycles, followed by a 10 min incubation at 72 °C. A negative control containing no template and a positive control containing 500 ng genomic DNA from A. actinomycetemcomitans strains 283 or HK1651 were included in each assay. Amplified PCR products were resolved on a 1 % (w/v) agarose gel and stained with ethidium bromide.



View larger version (38K):
[in this window]
[in a new window]
 
Fig. 2. RT-PCR from flp to tadA. (a) Schematic of the 5' terminus of the flp–rcp–tad gene cluster. The white arrow indicates no amplicon recovered; black arrows indicate amplicons recovered. (b) Amplicons (3·9 kb) obtained from strains 283, 361 and A26 in long RT-PCR of flp-1 to tadZ (c) Amplicons (1044 bp) obtained from strains 283 and A26 at different cycles in semi-quantitative RT-PCR of rcpA to tadZ.

 


View larger version (10K):
[in this window]
[in a new window]
 
Fig. 3. RT-PCR from tadA to tadD. (Left) Schematic of the 3' terminus of the flp–rcp–tad gene cluster. The black arrow indicates amplicon obtained from tadA to tadD. (Right) Amplicon (2·1 kb) obtained from strain 283. (R, rough; S, smooth phenotypes.)

 
Semi-quantitative PCR.
Reverse transcription and PCR amplification were performed as described, except a 15 µl aliquot was removed from each sample every five cycles starting at cycle 15 and ending at cycle 40. Amplicons at each time point were compared by agarose gel electrophoresis with ethidium bromide staining. PCR primers RcpA-F5a and TadZ-R1 (Table 1) were used to amplify a 1044 bp region spanning three genes from rcpA to tadZ from strains 283 and A26.

Long RT-PCR.
To amplify longer transcripts, the same protocols used for gene junction analysis were performed with the following modifications: DNase treatment was reduced to 15 min or eliminated, the first strand synthesis incubation temperature was increased to 50 °C and the PCR amplification program was changed. The long PCR program was preincubation at 94 °C for 1 min, 45 cycles of 94 °C for 30 s, 53 °C for 30 s, 68 °C for 6 min, with a final extension of 68 °C for 10 min.

RACE.
Amplification of the flp cDNA 5' end was performed using the 5' RACE System Version 2.0 (Invitrogen). Total RNA was isolated from A. actinomycetemcomitans 283 rough colony variant using a CsCl cushion as described above. Total RNA (5 µg) was pretreated with RQ1 RNase-free DNase (1 U; Promega) according to a modification of the manufacturer's protocol by incubating at 37 °C for 15 min in a total volume of 10 µl. First strand cDNA was synthesized in a modification of the 5' RACE System Version 2.0 protocol. The entire volume of DNase-digested total RNA (5 µg), 2·5 pmol GSP1 (RcpC-R2) in l µl, and 1 µl 10 mM dNTP mix were combined, denatured for 5 min at 65 °C and chilled on ice for 2 min. To the mixture, 4 µl 5x times; first strand buffer and 2 µl 0·1 M DTT were added and incubated at 50 °C for 2 min. Superscript II (200 U) reverse transcriptase or DEPC-treated water (1 µl) was added and incubated at 50 °C for 50 min, followed by incubation at 70 °C for 15 min to deactivate the enzyme. One microlitre of RNase H was then added and incubated at 37 °C for 20 min. Prior to purification, 5 µl DEPC-treated water was added to a final volume of 26 µl. The GlassMAX DNA spin cartridge purification of cDNA was performed according to the manufacturer's protocol. Tailing of 10 µl purified cDNA using terminal deoxynucleotidyl transferase and dCTP was done according to instructions, except that the sample was incubated on ice for 1 h. Confirmation of cDNA was performed after each step by PCR using two GSPs upFlp-F3 and RcpC-R1. All PCR products were analysed using agarose gel electrophoresis, ethidium bromide staining and 1 kb molecular size standards (Invitrogen) or 100 bp standards (MBI Fermentas).

Tailed cDNA was amplified by primary PCR using 10 µM each primer, the 5' RACE abridged anchor primer (AAP) and GSP2 (OrfB-R3). PCR was performed according to protocol except that 3 µl template was used in the reaction. The thermocycling program was as follows: preincubation at 94 °C for 1 min, followed by 45 cycles of 94 °C for 30 s, 46 °C for 30 s and 68 °C for 4 min, followed by an extension at 68 °C for 4 min. Nested PCR was performed using primary PCR product diluted 1 : 100 as template and 10 µM each of the nested primers abridged universal amplification primer (AUAP) and GSP3 (Flp2-R1). The PCR program was 94 °C for 2 min followed by 35 cycles of 94 °C for 1 min, 53 °C for 1 min and 72 °C for 2 min, followed by 72 °C for 10 min. To verify the specificity of the band of interest, a gel plug containing the band was used as template in a nested PCR using two GSPs, upFlp-F3 and Flp2-RS5. For DNA sequencing, the band of interest was excised from the gel, purified using the Wizard PCR prep DNA purification kit (Promega), quantified by gel electrophoresis by comparison to DNA low mass standards (Invitrogen) and ligated into plasmid pGEM-T (Promega). The plasmid was transformed into E. coli DH5{alpha}. Clones containing the PCR product were selected by blue/white screening, colony PCR and purified by the Wizard Plus SV miniprep DNA purification kit (Promega). The PCR product was sequenced from the plasmid in both directions using the T7 and SP6 primers at the Biopolymer Facility at Roswell Park Cancer Institute.

Construction of a reporter gene plasmid in E. coli.
The region upstream of flp (-48 to -485 bp) was amplified by PCR (35 cycles) from genomic DNA of A. actinomycetemcomitans strain 283 rough phenotype using EcoRI-tagged GSP FlpF1.eco (5'-GCGCGAATTCGCAGACAATATAGCACAAT-3') and BamHI-tagged GSP FlpR1.bam (5'-GCGCGGATCCCCTTGAGTTTGGATTTAATG-3'). The PCR product was subcloned into pGEM-T (Promega) and transformed into E. coli DH5{alpha} (Invitrogen). Colony PCR of white colonies with insert-specific primers was used to identify transformants. Transformant pEHG437 was digested with EcoRI/BamHI and ligated into EcoRI/BamHI digested promoter-probe plasmid pRS415 (Simons et al., 1987) such that the BamHI restriction endonuclease site was at the 3' terminus of the flp promoter region–lacZ junction. The flp promoter–lacZ fusion construct was introduced into chemically competent E. coli DH5{alpha} cells. Colonies that were blue on Luria–Bertani (LB) agar containing 64 µg X-Gal ml-1 (Fisher), 38 µg IPTG ml-1 (Invitrogen) and 100 µg ampicillin ml-1 (Sigma), and red on MacConkey agar (Difco) were selected for colony PCR. A transformant with a 437 bp insert, pEHR437, was confirmed by PCR using vector-specific primers (pRS415-F1: 5'-CGCCACATAGCAGAACTT-3' and pRS415-R1: 5'-TCTTCGCTATTACGCCAG-3') and GSPs FlpF1.eco and FlpR1.bam. The insert size was confirmed by restriction digestion. Orientation of the insert and junctions were confirmed by DNA sequence analysis.

Assay of ß-galactosidase specific activity.
An overnight culture of E. coli DH5{alpha} containing pEHR437 grown in LB medium supplemented with 100 µg ampicillin ml-1 was diluted 1 : 200 in fresh medium and incubated at 37 °C to an OD600 0·4. A 1·5 ml aliquot of culture was added to 1·5 ml 2x times; Buffer Z (120 mM Na2HPO4, 80 mM NaH2PO4, 20 mM KCl, 2 mM MgSO4) containing freshly added 50 µl chloroform, 1 µl 10 % SDS and 13·5 µl ß-mercaptoethanol, vortexed for 5 s and stored overnight at 4 °C for maximum cell permeabilization. Tubes were vortexed for 10 s and the chloroform allowed to settle prior to removing a 1 ml aliquot for analysis. Samples (in triplicate) were preincubated at 28 °C and the reaction initiated by the addition of prewarmed 0·2 ml ONPG (4 mg ml-1) in 100 mM phosphate buffer, pH 7·0. Reactions were incubated for 20 min at 28 °C and stopped with 0·5 ml 1 M Na2CO3. When necessary, samples were diluted in Buffer Z prior to the addition of ONPG to maintain A420 readings between 0·1 and 0·8. Prior to absorbance readings, a 1 ml aliquot was removed and centrifuged for 5 min at 12 000 g to remove cells. ß-Galactosidase activity in Miller units was calculated as (A420/OD600x1000)÷(assay time in minxvolume in ml) (Hernandez & Bremer, 1990; Miller, 1972).


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
RNA isolation and analysis
It was considerably more difficult to obtain sufficient total RNA from the rough colony variant than from the smooth colony phenotype. A starting volume of cell culture for the rough colony variant that was four times greater than the smooth colony variant was required to obtain an equivalent amount of total RNA by the CsCl/SDS method. Other methods incorporating guanidine isothiocyanate, i.e. Trizol (Invitrogen) alone, with or without sonication, Fast-prep tubes (Bio101) or freeze/thaw conditions were insufficient to lyse the tenaciously autoaggregating microcolonies of the rough strain. The addition of lysozyme and proteinase K helped to lyse the cells and obtain RNA. However, when the quality of total RNA was evaluated by gel electrophoresis, rRNA from the smooth strain was consistently intact, while that from the rough strain was often degraded. Good quality RNA was more frequently obtained from the rough strains using the classical, but time consuming, SDS lysis/CsCl cushion method. When equal concentrations of total RNA obtained from the isogenic rough/smooth pairs of strains 29 and A26 using this method were resolved on a 0·7 % (w/v) agarose/formaldehyde gel containing ethidium bromide, three rRNA bands characteristic of A. actinomycetemcomitans: 23S{alpha} (1·8 kb), 23Sß (1·2 kb) and 16S (1·6 kb) were present in each preparation with no evidence of RNA degradation (Fig. 1, top) (Haraszthy et al., 1992; Preus et al., 1992; Zambon & Sunday, 1989).



View larger version (83K):
[in this window]
[in a new window]
 
Fig. 1. Northern blot analysis. (Top panel) Total RNA (30 µg) per lane from A. actinomycetemcomitans strains 29 and A26 (R, rough; S, smooth phenotypes), resolved on a 0·7 % (w/v) agarose/formaldehyde gel. Three rRNA bands are evident at 1·8 kb (23S{alpha}), 1·2 kb (23Sß) and 1·5 kb (16S). (Bottom panel) Northern blot of the above gel probed with a 292 bp biotinylated internal fragment of rcpA.

 
Northern blot analysis
To obtain a signal by Northern blot, the gel had to be heavily loaded (30 µg) with total RNA and the wash stringency kept very low. As shown in Fig. 1 (bottom), the biotinylated probe from an internal fragment of rcpA hybridized only to total RNA from the rough colony, but not the isogenic smooth colony variants. Binding of the probe to 1·2–1·8 kb bands suggested binding to rRNA, or more likely, co-migrating mRNA, since there was no binding in the corresponding region of the RNA from the smooth phenotype variants. The probe also bound as a smear to the area between 2·9 and 4·4 kb. The smearing, likely caused by degradation of mRNA with homology to the probe, made it impossible to assess the exact transcript size(s). Therefore, alternative methods of transcript analysis by RT-PCR were required.

RT-PCR
To gain preliminary evidence of a polycistronic message containing flp and other contiguous genes within the cluster, and considering the putative instability of the mRNA (Fig. 1), transcript analysis was accessed at each individual gene junction from flp to tadD (Figs 2a and 3). Total RNA from strain 283 rough variant was compared to the smooth variant by RT-PCR. Amplicons were obtained from consecutive gene junctions flp to tadD for both the rough and smooth phenotypes suggesting a polycistronic message (data not shown). Gene junctions further downstream (tadDEFG) were not analysed. In each assay where a cDNA amplicon was obtained, the corresponding control template obtained without reverse transcriptase was negative, confirming the lack of DNA contamination of the RNA samples.

Semi-quantitative PCR
To demonstrate differential gene expression between the rough and smooth variants, the relative amount of target sequence was compared by sampling the PCR products at multiple time points during the amplification before reaching the plateau at 40 cycles. As shown in Fig. 2(c), amplicons from cDNA templates first appeared in cycle 20 in both the rough variants of strains 283 and A26. At this time point product was present in both rough and smooth variants for strain 283, although the amount from the rough strain was considerably greater through cycle 35. This quantitative difference persisted until cycle 40, the plateau phase of PCR, when no discernible difference in the amount of product obtained from the rough and smooth variants was observed. Thus, short transcripts (<2 kb) encompassing contiguous gene junctions or multiple genes from flp to tadD are present in both the rough and smooth variants, but the quantity from the rough variants is significantly greater than from the smooth variants.

Long PCR
Using conventional RT-PCR conditions the longest multigene transcript detected was about 1·6 kb. Solely increasing the incubation temperature of reverse transcription to reduce possible secondary structure of mRNA and altering the PCR conditions to amplify longer templates was not consistently successful. It was suspected that DNase treatment might be degrading the mRNA. When total RNA in DEPC-treated water heated as per the manufacturer's instructions to 37 °C for 30 min followed by 65 °C for 10 min was compared to unheated samples by RT-PCR, the heated samples yielded no product while the unheated samples did (data not shown). The corresponding RT template controls omitting reverse transcriptase were negative for both heated and unheated samples. This indicated that DNase treatment conditions were degrading the mRNA. Therefore, long RT-PCR was performed using total RNA not pretreated with DNase. Alternatively, a reduced 15 min incubation time with DNase was performed. As shown in Fig. 2(b), an amplicon of approximately 3·9 kb was obtained using PCR primers specific for flp (Flp1-F2) and tadZ (TadZ-R1) only from the rough variants of strains 283, A26 and 361; these products were obtained only when DNase pretreatment was omitted. The detection of multiple overlapping transcripts obtained by combining various primers and DNase-treated total RNA as template supported the presence of a polycistronic transcript. The following shorter transcripts were detected using conventional RT-PCR: 465 nt transcript from upstream flp-1 (-22) to flp-2 (+117) using primers upFlp1-F3, Flp2-R2 and Flp2-R1; 846 nt transcript from upstream flp (-22) to orfB (+253) using primers upFlp1-F3, OrfB-R2 and OrfB-R3; and 1133 nt transcript from upstream flp-1 (-22) to rcpC (+62) using primers upFlp1-F3, RcpC-R2 and RcpC-R1 (Fig. 2a and data not shown). Thus, a polycistronic message appears to be transcribed from the flp gene cluster at least from flp to tadZ. Similarly, using both DNase-treated and untreated total RNA in long PCR, a 2·1 kb transcript between tadA and tadD was detected only from the rough variant of strain 283 (Fig. 3). Using the same starting concentration of total RNA and RT-PCR conditions, the quantity of amplicon appeared to be considerably less than that of flp to tadZ even though amplification of genomic DNA was comparable. This suggests that less transcript encompassing the tadA to tadD region is present as compared to the upstream regions. Alternatively, this may be the result of technical difficulties in obtaining stable mRNA.

RT-PCR analysis of the potential flp promoter
Upstream of flp is an intergenic region of approximately 559 bp between orfX and flp (GenBank, AB005741.2). orfX is likely not part of the flp gene cluster as it is transcribed in the opposite orientation to flp. It was assumed that the promoter for flp fimbrial subunit or the polycistronic message containing flp would be located within the intergenic sequence, as it is in many fimbrial operons. Primers were designed to determine the approximate location of the 5' end of the flp transcript within this intergenic region. RT-PCR using primers upFlp-F2 (-485) and Flp2-R1 (+117) detected no transcript. However, a 139 bp amplicon was obtained using primers upFlp-F3 (-22) and Flp2-R1 (+117), suggesting that the transcription start site was located between -22 and -485 nt upstream of the flp translation start site (Fig. 2a and data not shown).

RACE
RACE was used to determine the exact location of the flp promoter (Fig. 4a). The GSP1 primer for first strand cDNA synthesis (RcpC-R2) was chosen because the annealing temperature of this primer and its nested primer RcpC-R1 were close to that of AAP (abridged anchor primer). However, primary PCR of tailed cDNA using primers AAP and RcpC-R1 failed to give the expected approximately 1·4 kb product, or even any bands at the calculated annealing temperature or several degrees lower. The manufacturer reports that it is difficult to get amplicons greater than 1·0 kb using Taq DNA polymerase in RACE. To verify that cDNA of the approximate expected length was present after each step of the RACE procedure, PCR done with a GSP located near the ribosome-binding site of flp (upFlp-F3) and reverse GSP within rcpC (RcpC-R1) amplified the expected 1·2 kb product (data not shown). When GSP2 (OrfB-R3) located closer to the 5' end of the gene cluster was used for primary PCR and the thermocycling program was adjusted for long PCR, several bands ranging from 350 to 976 bp were obtained (Fig. 4b). One band (976 bp) was of the expected size to contain the promoter upstream of flp. Using the diluted PCR reaction and nested primers [AUAP (abridged universal amplification primer) and GSP3 OrfA-R1] several bands ranging from 300 to 576 bp were obtained after 35 cycles (Fig. 4b). When this reaction was run for 45 cycles, all bands below 500 bp were also faintly visible in the negative (no template) PCR control, indicating that they were most likely nonspecific products (data not shown). Only the 576 bp band was unique. This band could be amplified with two GSPs (upFlp-F3 and Flp2-RS5) indicating that this was indeed a gene-specific band (data not shown). DNA sequencing of the 576 bp amplicon indicated that it was tailed with G residues and that the flp transcription start site was located -101 nt upstream from the flp translational start site (Fig. 4c). However, due to the C tailing, it is not possible to rule out that the start site is located at the G nucleotide -102 nt upstream of flp. Potential E. coli {sigma}70 consensus sequences are located -10 (TATAAT) and -35 (TTGCAT) upstream of the transcription start site with 16 nt between them (Fig. 5).



View larger version (34K):
[in this window]
[in a new window]
 
Fig. 4. RACE. (a) Schematic of RACE strategy, primers and products. (b) Left panel: amplicons obtained in primary PCR using primers AAP (abridged anchor primer) and GSP2. Lane 1, 1 kb standards; lane 2, PCR negative control; lane 3, dCTP-tailed cDNA template; lane 4, 100 bp standards. Right panel: amplicons obtained in nested PCR using primers AUAP (abridged universal amplification primer) and GSP3. Lane 1, 1 kb standards; lane 2, PCR negative control; lane 3, primary PCR template (1 : 100). (c) 5' end sequence of 576 bp amplicon with anchor nucleotides in parentheses followed by tailed (G) region. Transcriptional start site (GT) is indicated in bold.

 


View larger version (31K):
[in this window]
[in a new window]
 
Fig. 5. Promoter region of the flp operon. The flp translation start site and ribosome-binding site (RBS) are indicated. Transcription start site G(T) (black arrowhead) is located -102 (-101) bp upstream of the translation start site. The RNA polymerase consensus sequences at -10 and -35 relative to the transcription start site and the AT-rich region upstream of the promoter are underlined.

 
Functional analysis of the promoter region
ß-Galactosidase activity of the promoter-probe construct pEHR434, containing the promoter site in the region upstream of flp but lacking the ribosome-binding site, was compared with that of the host strain with and without vector alone. As shown in Table 2, the ß-galactosidase activity of pEHR434 was at least 1000-fold greater than the background level of the pRS415 vector alone. This confirms the presence of a DNA sequence in the upstream region of flp that is able to function as a promoter in E. coli as suggested by the {sigma}70 sequences identified in the promoter region.


View this table:
[in this window]
[in a new window]
 
Table 2. ß-Galactosidase activity

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In this paper, we describe the analysis of the flp (flp–rcp–tad) fimbrial gene cluster in A. actinomycetemcomitans from flp to tadD. Previously, in complementation studies of a spontaneous smooth strain, the rough adherent phenotype could only be fully restored when the 12 kb flptadG region, not just when tadA–tadG was present on a plasmid (Kachlany et al., 2000). This suggests that the flp–tadZ loci of the fimbrial gene cluster are essential to the presentation of the adherent rough colony phenotype. Thus, our transcriptional analysis focused on establishing whether this region forms a single operon or transcriptional regulatory unit. Initially, Northern blot analysis comparing total RNA from the rough and smooth colony variants of A. actinomycetemcomitans and probed with DNA from an internal fragment of rcpA revealed that rcpA was transcribed in the rough variants and not the isogenic smooth variants. However, smearing of the bands precluded any definitive estimation of unitary transcript size. Similarly, Ramer et al. (1996) were unable to detect mRNA of the rcpA homologue bfpB by Northern blot analysis under any conditions, but could detect it by a more sensitive RNase-protection assay. We chose to use RT-PCR to analyse transcripts because of the small amount of RNA required, ease of analysis, sensitivity and availability of GSPs.

The flp fimbrial gene cluster was initially examined from flp to tadD by assaying each gene junction by RT-PCR. Transcripts containing multiple contiguous genes were detected from both the rough and smooth variants. The largest amplicon detected was 2·1 kb containing tadAtadD from the rough variant only. Another 3·9 kb transcript from flp to tadZ was detected only when total RNA was not pretreated with DNase prior to RT-PCR. Together, these RT-PCR analyses provide evidence that at least all the genes between flp and tadZ, as well as between tadA and tadD are transcribed as polycistronic messages.

In contrast, previous studies had shown by Northern blot analysis that a flp transcript of approximately 550 nt was present only in the rough colony variant of A. actinomycetemcomitans (Ruparelia et al., 2000). Furthermore, it has been proposed previously that the fimbrial subunit was transcribed separately from the rest of the gene cluster (Ruparelia et al., 2000). The presence of a shorter transcript only in the rough variant by Northern blot analysis may reflect both quantity of transcript produced and/or the sensitivity of the assay. Sequence analysis of the 98 bp flp–orfA intergenic region revealed a potential stem–loop structure that could act as a transcriptional attenuator (Inoue et al., 1998). The presence of an attenuator distal to the fimbrial subunit is seen in type IV pilus gene expression (Brown & Taylor, 1995; Klemm, 1994; Ramer et al., 1996; Villar et al., 1999). An attenuator after the subunit gene enables transcription of many more copies of the subunit, while offering a limited read-through of the complete transcript containing downstream genes. Though fimbrial operon transcripts often encode the fimbrial subunit in combination with downstream genes, frequently only the fimbrial subunit mRNA is detectable by Northern blot (Brown & Taylor, 1995; Collison et al., 1996). In A. actinomycetemcomitans, the Northern blot analysis indicated that significant fimbrial subunit transcript was produced only from the rough colony variant, while high sensitivity RT-PCR indicated that a polycistronic transcript was generated from at least seven genes. It is likely that the entire gene cluster from flp to tadG is transcribed as a polycistronic message. However, the full-length transcript has yet to be detected, probably due to mRNA instability.

Gene junction transcription analysis indicated that transcription occurred in both the rough and smooth variants when PCR was carried out for 45 cycles. However, semi-quantitative analysis revealed that significantly more transcript was produced by the rough variants. This supports the less sensitive Northern blot data in this paper and in previous studies showing transcript present only in the rough variant. Therefore, we propose that the flp operon is transcribed by both the rough and smooth variants, but it is more rapidly degraded in the smooth variant so that few, if any, full-length transcripts are translated. It may also be that the transcript is severely downregulated in the smooth variant. Furthermore, it is probable that a smaller transcript containing only flp-1 is transcribed in stoichiometrically greater amounts.

The arrangement of the promoter proximal to a major fimbrial subunit gene is typical of the type IV fimbriae biogenesis. The 559 bp intergenic sequence upstream of flp was the probable location of the operon promoter. A potential ribosome-binding site (AGGAG) is located -11 nt upstream of the translation start codon (ATG) of flp (Inoue et al., 1998). Preliminary RT-PCR analysis indicated that the promoter was likely located between -485 and -22 bp upstream of flp. Previously, Inoue et al. (1998) predicted the promoter region to be located in the region -126 to -154 nt upstream of flp (GenBank AB005741.2). Here, the transcriptional start site was identified as a T or G nucleotide located at a position -101 or -102 bp upstream from the flp translational start site using 5' RACE. Given that a majority of transcriptional start sites are purines, it is likely that the G nucleotide is the true start site. Analysis of the immediate upstream DNA sequence revealed a potential E. coli {sigma}70 promoter sequence at -10 (TATAAT) and -35 (TTGCAT) separated by 16 nt. Immediately upstream of the -35 site is an AT-rich tract suggesting a possible regulatory sequence which would enhance fimbriae expression (Owen-Huphes et al., 1992; Puente et al., 1996; Ross et al., 1993). Significant ß-galactosidase activity of the transcriptional reporter gene fusion containing the region upstream of flp indicated the presence of DNA sequences that could function as a promoter in E. coli. Previously, the well-studied leukotoxin promoter of A. actinomycetemcomitans was also found to contain E. coli {sigma}70 promoter sequences functional in E. coli as well as A. actinomycetemcomitans (Brogan et al., 1994; Kolodrubetz et al., 1996).

In summary, the flp fimbrial subunit gene along with its putative prepilin peptidase (orfB) and the type II secretion homologues (rcpA and tadA) is transcribed as a polycistronic message. Transcription of this operon is significantly greater in the rough than the smooth variant of A. actinomycetemcomitans. The putative operon promoter is located at nucleotide -101 or -102 in the intergenic region just upstream of flp. The rough to smooth phenotype conversion is essentially irreversible in vitro. Further analysis of the promoter region should provide insight into the regulation of fimbrial expression and phenotype transition.


   ACKNOWLEDGEMENTS
 
We would like to thank V. James Hernandez for his technical advice and critical review of the manuscript. We also thank Bruce Roe, Fares Z. Najar, Sandy Clifton, Tom Ducey, Lisa Lewis and David Dyer for the use of unpublished nucleotide sequence data provided by the A. actinomycetemcomitans Genome Sequencing Project at the University of Oklahoma. This work was supported in part by grants DE 09838 (F. A. S) and DE 14466-01 (E. M. H.) from the National Institute Dental and Craniofacial Research.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Brogan, J. M., Lally, E. T., Poulsen, K., Kilian, M. & Demuth, D. R. (1994). Regulation of Actinobacillus actinomycetemcomitans leukotoxin expression: analysis of the promoter regions of leukotoxic and minimally leukotoxic strains. Infect Immun 62, 501–508.[Abstract]

Brown, R. C. & Taylor, R. K. (1995). Organization of tcp, acf, and toxT genes within a ToxT-dependent operon. Mol Microbiol 16, 425–439.[Medline]

Burgher, L. W., Loomis, G. W. & Ware, F. (1973). Systemic infection due to Actinobacillus actinomycetemcomitans. Am J Clin Pathol 60, 412–415.[Medline]

Collison, S. K., Clouthier, S. C., Doran, J. L., Banser, P. A. & Kay, W. W. (1996). Salmonella enteritidis agfBAC operon encoding thin, aggregative fimbriae. J Bacteriol 178, 662–667.[Abstract]

Deretic, V. & Konyecsni, W. M. (1989). Control of mucoidy in Pseudomonas aeruginosa: transcriptional regulation of algR and identification of the second regulatory gene, algQ. J Bacteriol 171, 3680–3688.[Medline]

el Khizzi, N., Kasab, S. A. & Osoba, A. O. (1997). HACEK group endocarditis at the Riyadh Armed Forces Hospital. J Infect 34, 69–74.[Medline]

Fine, D. H., Furgang, D., Schreiner, H. C., Gonocharof, P., Charlesworth, J., Ghazwan, G., Fitzgerald-Bocarsly, P. & Figurski, D. H. (1999a). Phenotypic variation in Actinobacillus actinomycetemcomitans during laboratory growth: implications for virulence. Microbiology 145, 1335–1347.[Abstract]

Fine, D. H., Furgang, D., Kaplan, J., Charlesworth, J. & Figurski, D. H. (1999b). Tenacious adhesion of Actinobacillus actinomycetemcomitans strain CU1000 to salivary-coated hydroxyapatite. Arch Oral Biol 44, 1063–1076.[CrossRef][Medline]

Fine, D. H., Goncharoff, P., Schreiner, H., Chang, K. M., Furgang, D. & Figurski, D. (2001). Colonization and persistence of rough and smooth colony variants of Actinobacillus actinomycetemcomitans in the mouths of rats. Arch Oral Biol 46, 1065–1078.[CrossRef][Medline]

Fives-Taylor, P., Meyer, D. & Mintz, P. (1995). Characteristics of Actinobacillus actinomycetemcomitans invasion of and adhesion to cultured epithelial cells. Adv Dent Res 9, 55–62.[Abstract]

Gally, D. L., Bogan, J. A., Eisenstein, B. I. & Blomfield, I. C. (1993). Environmental regulation of the fim switch controlling type 1 fimbrial phase variation in Escherichia coli K-12; effects of temperature and media. J Bacteriol 175, 6186–6193.[Abstract]

Haase, E. M., Zmuda, J. L. & Scannapieco, F. A. (1999). Identification and molecular analysis of rough-colony-specific outer membrane proteins of Actinobacillus actinomycetemcomitans. Infect Immun 67, 2901–2908.[Abstract/Free Full Text]

Haraszthy, V. I., Sunday, G. J., Bobek, L. A., Motley, T. S., Preus, H. & Zambon, J. J. (1992). Identification and analysis of the gap region in the 23S ribosomal RNA from Actinobacillus actinomycetemcomitans. J Dent Res 71, 1561–1568.[Abstract]

Haraszthy, V. I., Zambon, J. J., Trevisan, M., Zeid, M. & Genco, R. J. (2000). Identification of periodontal pathogens in atheromatous plaques. J Periodontol 71, 1554–1560.[Medline]

Hardie, K. R., Seydel, A., Guilvout, I. & Pugsley, A. P. (1996). The secretin-specific, chaperone-like protein of the general secretory pathway: separation of proteolytic protection and piloting functions. Mol Microbiol 22, 967–976.[Medline]

Hernandez, V. J. & Bremer, H. (1990). Guanosine tetraphosphate (ppGpp) dependence of the growth rate control of rrnB P1 promoter activity in Escherichia coli. J Biol Chem 265, 11605–11614.[Abstract/Free Full Text]

Inoue, T., Tanimoto, I., Ohta, H., Kato, K., Murayama, Y. & Kazuhiro, F. (1998). Molecular characterization of low-molecular-weight component protein, Flp, in Actinobacillus actinomycetemcomitans fimbriae. Microbiol Immunol 42, 253–258.[Medline]

Inouye, T., Ohta, H., Kokeguchi, S., Fukui, K. & Kato, K. (1990). Colonial variation and fimbriation of Actinobacillus actinomycetemcomitans. FEMS Microbiol Lett 69, 13–18.[CrossRef]

Kachlany, S. C., Planet, P. J., Bhattacharjee, M. K., Kollia, E., DeSalle, R., Fine, D. H. & Figurski, D. H. (2000). Nonspecific adherence by Actinobacillus actinomycetemcomitans requires genes widespread in bacteria and archaea. J Bacteriol 182, 6169–6176.[Abstract/Free Full Text]

Kachlany, S. C., Planet, P. J., Desalle, R., Fine, D. H., Figurski, D. H. & Kaplan, J. B. (2001). flp-1, the first representative of a new pilin gene subfamily, is required for non-specific adherence of Actinobacillus actinomycetemcomitans. Mol Microbiol 40, 542–554.[CrossRef][Medline]

Klemm, P. (1994). Fimbriae: Adhesion, Genetics, Biogenesis, and Vaccines. Boca Raton, FL: CRC Press.

Kolodrubetz, D., Spitznagel, J., Wang, B., Phillips, L. H., Jacobs, C. & Kraig, E. (1996). Cis elements and trans factors are both important in strain-specific regulation of the leukotoxin gene in Actinobacillus actinomycetemcomitans. Infect Immun 64, 3451–3460.[Abstract]

Marrs, E. F., Ruehl, W. W., Schoolnik, G. K. & Falkow, S. (1988). Pilin-gene phase variation of Moraxella bovis is caused by an inversion of the pilin genes. J Bacteriol 170, 3032–3039.[Medline]

Martin, B. F., Derby, B. M., Budzilovich, G. N. & Ransohoff, J. (1967). Brain abscesses due to Actinobacillus actinomycetemcomitans. Neurology 17, 833–837.[Medline]

Meyer, D. H., Sreenivasan, P. K. & Fives-Taylor, P. M. (1991). Evidence for invasion of a human oral cell line by Actinobacillus actinomycetemcomitans. Infect Immun 59, 2719–2726.[Medline]

Miller, J. H. (1972). Experiments in Molecular Genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Mims, C. A. (1976). Host and microbial factors influencing susceptibility. In The Pathogenesis of Infectious Disease, pp. 196–199. Edited by C. A. Mims. New York: Academic Press.

Muhle, I., Rau, J. & Ruskin, J. (1979). Vertebral osteomyelitis due to Actinobacillus actinomycetemcomitans. J Am Med Assoc 241, 1824–1825.[CrossRef][Medline]

Ogierman, M. A., Zabihi, S., Mourtzios, L. & Manning, P. A. (1993). Genetic organization and sequence of the promoter-distal region of the tcp gene cluster of Vibrio cholerae. Gene 126, 51–60.[Medline]

Overholt, B. F. (1966). Actinobacillus actinomycetemcomitans endocarditis. Arch Intern Med 117, 99–102.[CrossRef][Medline]

Owen-Huphes, T. A., Pavitt, G. D., Santos, D. S., Sidebotham, J. M., Hulton, C. S., Hinton, J. C. & Higgins, C. F. (1992). The chromatin-associated protein H-NS interacts with curved DNA topology and gene expression. Cell 71, 255–265.[Medline]

Page, M. I. & King, E. O. (1966). Infection due to Actinobacillus actinomycetemcomitans and Haemophilus aphrophilus. N Engl J Med 275, 181–188.[Medline]

Preus, H. R., Sunday, G. J., Haraszthy, V. I. & Zambon, J. J. (1992). Rapid identification of Actinobacillus actinomycetemcomitans based on analysis of 23S ribosomal RNA. Oral Microbiol Immunol 7, 372–375.[Medline]

Puente, J. L., Bieber, D., Ramer, S. W., Murray, W. & Schoolnik, G. K. (1996). The bundle-forming pili of enteropathogenic Escherichia coli: transcriptional regulation by environmental signals. Mol Microbiol 20, 87–100.[Medline]

Pugsley, A. P. (1993). The complete general secretory pathway in gram-negative bacteria. Microbiol Rev 57, 50–108.[Abstract]

Ramer, S. W., Bieber, D. & Schoolnik, G. K. (1996). BfpB, an outer membrane lipoprotein required for the biogenesis of bundle-forming pili in enteropathogenic Escherichia coli. J Bacteriol 178, 6555–6563.[Abstract]

Roosendaal, B., van Bergen en Henegouwen, P. M. P., Mooi, F. R. & de Graaf, F. K. (1986). Regulatory aspects of K99 fimbriae synthesis. In Protein–Carbohydrate Interactions in Biological Systems: the Molecular Biology of Microbial Pathogenicity, pp. 57–59. Edited by S. Normark, B. E. Uhlin & H. Wolf-Watz. New York: Academic Press.

Rosan, B., Slots, J., Lamont, R. J., Listgarten, M. A. & Nelson, G. M. (1988). Actinobacillus actinomycetemcomitans fimbriae. Oral Microbiol Immunol 3, 58–63.[Medline]

Ross, W., Gosink, K. K., Salomon, J., Igarashi, K., Zou, C., Ishihama, A., Severinov, K. & Gourse, R. L. (1993). A third recognition element in bacterial promoters: DNA binding by the alpha subunit of RNA polymerase. Science 262, 1407–1413.[Medline]

Rudney, J. D., Chen, R. & Sedgewick, G. J. (2001). Intracellular Actinobacillus actinomycetemcomitans and Porphyromonas gingivalis in buccal epithelial cells collected from human subjects. Infect Immun 69, 2700–2707.[Abstract/Free Full Text]

Ruparelia, S., McGill, E. S. & Spitznagel, J. K. (2000). Pilin gene expression by Actinobacillus actinomycetemcomitans (abstract). J Dent Res 79, 393.

Scannapieco, F. A., Millar, S. J., Reynolds, H. S., Zambon, J. J. & Levine, M. J. (1987). Effect of anaerobiosis on the surface ultrastructure and surface proteins of Actinobacillus actinomycetemcomitans (Haemophilus actinomycetemcomitans). Infect Immun 55, 2320–2323.[Medline]

Schwan, W. R., Seifert, H. S. & Duncan, J. L. (1992). Growth conditions mediate differential transcription of fim genes involved in phase variation of type I pili. J Bacteriol 174, 2367–2375.[Abstract]

Serra, P. & Tonato, M. (1969). Subacute bacterial endocarditis due to Actinobacillus actinomycetemcomitans. Amer J Med 47, 809–812.[Medline]

Simons, R. W., Houman, F. & Kleckner, N. (1987). Improved single and multicopy lac-based cloning vectors for protein and operon fusions. Gene 53, 85–96.[CrossRef][Medline]

Skerker, J. M. & Shapiro, L. (2000). Identification and cell cycle control of a novel pilus system in Caulobacter crescentus. EMBO J 19, 3223–3234.[Abstract/Free Full Text]

Slots, J. & Dahlen, G. (1985). Subgingival microorganisms and bacterial virulence factors in periodontitis. Scand J Dent Res 93, 119–127.[Medline]

Snellings, N. J., Tall, B. D. & Venkatesan, M. M. (1997). Characterization of Shigella type 1 fimbriae; expression, FimA sequence, and phase variation. Infect Immun 65, 2462–2467.[Abstract]

Steckelberg, J. M., Melton, L. J., Ilstrup, D. M., Rouse, M. S. & Wilson, W. R. (1990). Influence of referral bias on the apparent clinical spectrum of infective endocarditis. Am J Med 88, 582–588.[Medline]

Stone, K. D., Zhang, H., Carlson, L. K. & Donnenberg, M. S. (1996). A cluster of fourteen genes from enteropathogenic Escherichia coli is sufficient for the biogenesis of a type IV pilus. Mol Microbiol 20, 325–337.[Medline]

Strom, M. S. & Lory, S. (1993). Structure-function and biogenesis of the type IV pili. Annu Rev Microbiol 47, 565–596.[CrossRef][Medline]

Tanner, A. C. R., Haffer, C., Brathall, G. T., Visconti, R. A. & Socransky, S. S. (1979). A study of the bacteria associated with advancing periodontitis in man. J Clin Periodontol 6, 278–307.[Medline]

Vandepitte, J., DeGeest, H. & Jousten, P. (1977). Subacute bacterial endocarditis due to Actinobacillus actinomycetemcomitans. J Clin Pathol 30, 842–846.[Abstract]

Villar, M. T., Helber, J. T., Hood, B., Schaefer, M. R. & Hirschberg, R. L. (1999). Eikenella corrodens phase variation involves a posttranslational event in pilus formation. J Bacteriol 181, 4154–4160.[Abstract/Free Full Text]

Zambon, J. J. (1985). Actinobacillus actinomycetemcomitans in human periodontal disease. J Clin Periodontol 12, 1–20.[Medline]

Zambon, J. J. & Sunday, G. J. (1989). Absence of the 23S ribosomal RNA subunit in Actinobacillus actinomycetemcomitans. J Dent Res 68, 218.

Received 3 June 2002; revised 30 September 2002; accepted 8 October 2002.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Haase, E. M.
Articles by Scannapieco, F. A.
Articles citing this Article
PubMed
PubMed Citation
Articles by Haase, E. M.
Articles by Scannapieco, F. A.
Agricola
Articles by Haase, E. M.
Articles by Scannapieco, F. A.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2003 Society for General Microbiology.