HU protein affects transcription of surface polysaccharide synthesis genes in Porphyromonas gingivalis

Abstract

K-antigen capsule synthesis is an important virulence determinant of the oral anaerobe Porphyromonas gingivalis. We previously reported that the locus required for synthesis of this surface polysaccharide in strain W83 (TIGR identification PG0106 to PG0120) is transcribed as a large (∼16.7-kb) polycistronic message. Through sequence analysis, we have now identified a 77-bp inverted repeat located upstream (206 bp) of the start codon of PG0106 that is capable of forming a large hairpin structure. Further sequence analysis just upstream and downstream of the capsule synthesis genes revealed the presence of two genes oriented in the same direction as the operon that are predicted to encode DNA binding proteins: PG0104, which is highly similar (57%) to DNA topoisomerase III, and PG0121, which has high similarity (72%) to DNA binding protein HU (β-subunit). In this report, we show that these two genes, as well as the 77-bp inverted repeat region, are cotranscribed with the capsule synthesis genes, resulting in a large transcript that is ∼19.4 kb (based on annotation). We also show that a PG0121 recombinant protein is a nonspecific DNA binding protein with strong affinity to the hairpin structure, in vitro, and that transcript levels of the capsule synthesis genes are downregulated in a PG0121 deletion mutant. Furthermore, we show that this decrease in transcript levels corresponds to a decrease in the amount of polysaccharide produced. Interestingly, expression analysis of another polysaccharide synthesis locus (PG1136 to PG1143) encoding genes involved in synthesis of a surface-associated phosphorylated branched mannan (APS) indicated that this locus is also downregulated in the PG0121 mutant. Altogether our data indicate that HU protein modulates expression of surface polysaccharides in P. gingivalis strain W83.

FIG. 1.

Schematic representation of the capsule locus. (A) The position of the predicted hairpin structure created by base pairing of the 77-bp inverted repeat is indicated as a segmented line, and predicted promoter regions are shown with black arrows. The number of base pairs on either side of the hairpin and those contained within the loop are indicated. The ORF numbers correspond to TIGR CMR identifications. (B) Sequence and alignment of the 77-bp inverted repeat.

FIG. 2.

Northern blot analysis of PG0106 (A), PG0104 (B), and PG0121 (C) expression in the PG0106 (capsule) mutant. Total RNA was isolated from WT strain W83 and DEL0106::Erm mutant (Δ106) cells in the mid-exponential growth phase, as described in Materials and Methods. Expression of the three genes (PG0104, PG0106, and PG0121) in the WT and the capsule operon null mutant (Δ106) are shown. The 10 size marker bands correspond to 9.0, 6.0, 5.0, 4.0, 3.0, 2.5, 2.0, 1.5, 1.0, and 0.5 kb. Arrows indicate hybridization to large transcripts (>9 kb) and transcripts in the 5- to 6-kb size range.

FIG. 3.

RT-PCR analysis to evaluate transcriptional linkage of PG0121 to the capsule operon. Total RNA was isolated from the WT strain W83 and the DEL0121::Erm mutant (ΔHU), and cDNA was generated as described in Materials and Methods and using the MonsterScript cDNA synthesis kit. In the subsequent PCRs, the following primer combinations were used. (A) RT106_F/RT108_R (lane 1), RT108_F/RT110_R (lane 2), RT113_F/RT116_R (lane 3), RT116_F/RT118_R (lane 4), and RT118_F/RT120_R (lane 5); the gel shows linkage throughout the operon and the lower transcript levels of the K-antigen synthesis genes (except for PG0106 to PG0108) in the ΔHU mutant. (B) TIGR106_F/TIGR106_R (lane 1; positive control) and RT119_F/RT121_R (lane 2); the gel shows linkage of PG0121 to the capsule operon. M, DNA size markers.

FIG. 4.

RT-PCR analysis to evaluate transcriptional linkage of PG0104 to the 5′ end of the capsule operon and transcription through the hairpin region. Total RNA was isolated from the wild-type strain, and cDNA was generated as described in Materials and Methods with the QuantiTect reverse transcription kit. In the subsequent PCRs, the following primer combinations were used: RT_end104 (lane 1), RT_link104 (lane 2), RT_hairpin (lane 3), RT_link106 (lane 4), RT_start106 (lane 5). The figure demonstrates linkage of PG0104 to the hairpin, transcription of the hairpin, and linkage of the hairpin to PG0106. M, DNA size markers.

FIG. 5.

Northern blot analysis of PG0121 (HU) and PG0106 (capsule operon) expression in the PG0121 (HU) mutant. Total RNA was isolated from WT strain W83 and the DEL0121::Erm mutant (ΔHU) as described in Materials and Methods. (A) Expression of PG0121 during exponential growth in strain W83 and the corresponding DEL0121::Erm mutant. (B) Expression of PG0106 during exponential growth in the parent strain W83 compared to expression during exponential (Exp) and stationary-phase growth (SP) in the corresponding DEL0121::Erm mutant, showing that the K-antigen capsule synthesis genes are downregulated in the ΔHU mutant.

FIG. 6.

qPCR analysis of K-antigen capsule synthesis genes in the PG0121 mutant and the complemented strain (pTgroES-HU). Total RNA was isolated from the parent strain W83, the DEL0121::Erm mutant (PG0121 mutant), and the complemented strain harboring plasmid pTgroES-HU, as described in Materials and Methods. To assess expression from all potential transcripts, cDNA was generated with random primers and the MonsterScript cDNA synthesis kit. The relative fold changes in expression shown are the averages of three separate experiments.

FIG. 7.

Comparison of polysaccharide profiles of different strains based on PAGE. (A) Crude polysaccharide extracted from 5-ml cultures of the parent strain W83, the PG0104 mutant (Δ0104), the PG0106 mutant (Δ0106), and the PG0121 mutant (ΔHU) was analyzed by SDS-PAGE. Both the PG0106 and PG0121 mutants showed less capsule production than the parent W83 strain. The PG0104 mutant showed no change in capsule production and displayed a pattern similar to W83. (B) Polysaccharide profile analysis of the parent strain (W83), the PG0106 mutant (Δ0106), and the PG0121 mutant (ΔHU). For each strain, the left lane contained 50 μg of extracted polysaccharide and the right lane contained 100 μg. As expected, the PG0106 mutant exhibited less capsular polysaccharide and was lacking in a number of high-molecular-weight bands. However, the profile of the HU mutant was similar to the parent W83 strain, indicating that although the HUβ mutant is deficient in capsule production, the strain is not severely defective in biosynthesis of polysaccharides. (C) Double immunodiffusion assay results with antibodies to K-antigen capsule. All strains cross-reacted with the antiserum except for ΔPG0106 (the K-antigen null mutant) and strain 381, a K-antigen minus strain, indicating that neither a PG0104 nor a PG0121 mutation completely abolishes capsule production.

FIG. 8.

Comparison of capsular structures by electron microscopy of the parent strain W83 with the K-antigen capsule minus mutant (Δ0106) and the ΔPG0121 (HU) mutant. Direct magnification levels of images were variable: W83 (×60,000); W83Δ0121 (×80,000), K-antigen capsule null strain W83Δ0106 (×100,000). The HU mutant strain produced less capsular polysaccharide than the parent strain W83.

FIG. 9.

Northern analysis of the complemented strain (pTgroES-HU) shows restoration of expression of large capsule operon transcripts. The blot was hybridized with a mixture of two probes: a 233-bp probe for PG0106 (capsule operon) and a 137-bp probe for PG0121 (HU). Shown is the expression of a large transcript(s) that was restored when HU was supplied in trans under the control of the groES (weak) promoter.

FIG. 10.

qPCR analysis of the A-LPS synthesis locus (PG1136 to PG1145) in the PG0121 mutant. Total RNA was isolated in triplicate from the parent strain W83 and the DEL0121::Erm mutant (PG0121 mutant). To assess expression from all potential coding transcripts, cDNA was generated with random primers. The relative fold change of expression, showing downregulation of this locus in the HU mutant, is the average of three separate experiments.

FIG. 11.

Purification of recombinant PG0121 His-tagged fusion protein (HU) and gel mobility shift analysis of HU binding to the hairpin region. (A) SDS-PAGE analysis results, showing purification of the PG0121 His-tagged fusion protein. (B) Western blot results for the PG0121 His-tagged fusion protein, detected with anti-His tag antibodies. (C) Recombinant His-tagged PG0121 at the concentrations indicated was incubated with 100 ng of the inverted repeat target probe (see Materials and Methods for details) in a buffer containing high levels of salt (200 mM NaCl). (D) The same assay as shown in panel C was performed with a target probe of the promoter region minus the hairpin structure. (E) Competition assays using a DIG-labeled probe indicate high affinity for the hairpin (see Materials and Methods for details).