Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Mar 27.
Published in final edited form as: Mol Microbiol. 2011 Oct 20;82(4):836–850. doi: 10.1111/j.1365-2958.2011.07831.x

RbsB (NTHI_0632) mediates quorum signal uptake in nontypeable Haemophilus influenzae strain 86-028NP

Chelsie E Armbruster 1, Bing Pang 1, Kyle Murrah 1, Richard A Juneau 1, Antonia C Perez 1, Kristin ED Weimer 1, W Edward Swords 1,*
PMCID: PMC3609414  NIHMSID: NIHMS325110  PMID: 21923771

Summary

Nontypeable Haemophilus influenzae (NTHI) is a respiratory commensal and opportunistic pathogen, which persists within biofilms on airway mucosal surfaces. For many species, biofilm formation is impacted by quorum signaling. Our prior work shows that production of autoinducer-2 (AI-2) promotes biofilm development and persistence for NTHI 86-028NP. NTHI 86-028NP encodes an ABC transporter annotated as a ribose transport system that includes a protein (RbsB) with similarity to the Escherichia coli LsrB and Aggregatibacter actinomycetemcomitans RbsB proteins that bind AI-2. In this study, inactivation of rbsB significantly reduced uptake of AI-2 and the AI-2 precursor dihydroxypentanedione (DPD) by NTHI 86-028NP. Moreover, DPD uptake was not competitively inhibited by ribose or other pentose sugars. Transcript levels of rbsB increased in response to DPD and as bacteria approached stationary-phase growth. The NTHI 86-028NP rbsB mutant also formed biofilms with significantly reduced thickness and total biomass and reduced surface phosphorylcholine, similar to a luxS mutant. Infection studies revealed that loss of rbsB impaired bacterial persistence in the chinchilla middle-ear, similar to our previous results with luxS mutants. Based on these data, we conclude that in NTHI 86-028NP, RbsB is a LuxS/AI-2 regulated protein that is required for uptake of and response to AI-2.

Keywords: Nontypeable Haemophilus influenzae, autoinducer-2, biofilm, otitis media

Introduction

Nontypeable Haemophilus influenzae (NTHI) inhabits the nasopharynx and upper airways of children and healthy adults (Erwin & Smith, 2007, Faden et al., 1991), and can cause opportunistic infections of the airway mucosa that include bronchitis, sinusitis, and otitis media (OM) (Erwin & Smith, 2007, Pichichero, 2000). Among these infections, OM is of particular public health importance. For example, OM is an extremely pediatric infection, affecting the majority of all children (Klein, 2000, Mandel et al., 2008), and is a leading cause for pediatric office visits and new antibiotic prescription to children (Finkelstein et al., 2001).

OM infections are often chronic and/or recurrent in nature, and can be highly resistant to immune clearance and antibiotic treatment. Thus, it is generally thought that OM involves persistence of bacteria within biofilm communities (Costerton et al., 1999, Bakaletz, 2007, Post et al., 2007). Bacterial biofilms have been observed directly in patient samples (Hall-Stoodley et al., 2006, Hoa et al., 2009, Nistico et al., 2011, Post, 2001) and in the chinchilla experimental model for OM (Ehrlich et al., 2002, Post, 2001). As biofilms provide protection from the immune response and antibiotic treatment, understanding the mechanisms involved in their formation during chronic infection may provide new targets for disruption and treatment of biofilm-related infections (Armbruster & Swords, 2010).

Biofilm formation for many bacterial species is controlled in part through cell density-dependent quorum signaling networks, wherein changes in bacterial population phenotypes are mediated by accumulation of a soluble signaling mediator (Hardie & Heurlier, 2008, Jayaraman & Wood, 2008, Miller & Bassler, 2001, Waters & Bassler, 2005). Such signaling networks include autoinducer-2 (AI-2), a ribose derivative produced by both Gram-negative and Gram-positive bacteria that is conserved among numerous bacterial species and thus referred to as an inter-species signal (Jayaraman & Wood, 2008, Waters & Bassler, 2005). Mutation of the genetic determinant of AI-2 production (luxS) was previously shown to impact biofilm thickness/maturation and persistence of NTHI strain 86-028NP in an experimental animal model of OM (Armbruster et al., 2009). The focus of this study is to understand how strain NTHI 86-028NP internalizes and responds to AI-2.

In Vibrio harveyi, sensing of AI-2 occurs through a two-component system involving LuxP and LuxQ (Bassler et al., 1994). LuxP is similar to periplasmic ribose binding proteins and binds AI-2, while LuxQ contains sensor kinase and response regulator domains to propagate the signal (Henke & Bassler, 2004). Salmonella typhimurium, Escherichia coli, Sinorhizobium meliloti, and Aggregatibacter actinomycetemcomitans utilize the Lsr (LuxS regulated) ABC transporter, an AI-2 transport system similar to the ribose ABC transporter (Miller et al., 2004). Rather than the two-component signal cascade described in V. harveyi, the Lsr system mediates AI-2 uptake via binding by LsrB and transport through a heterodimeric membrane channel (Pereira et al., 2008, Michiko E. Taga, 2001, Xavier & Bassler, 2005, Shao et al., 2007b). In addition to LsrB, it was determined in A. actinomycetemcomitans that the ribose binding protein RbsB also interacts with AI-2 (Shao et al., 2007a, James et al., 2006).

NTHI 86-028NP does not possess a homolog of LuxPQ transport system. However, the NTHI 86-028NP genome contains an ABC transporter annotated as a ribose transport system that includes a binding protein encoded by NTHI_0632, designated rbsB (Harrison et al., 2005). The NTHI 86-028NP rbsB gene product has 40.3 percent identity to E. coli LsrB, and 87 percent identity to A. actinomycetemcomitans RbsB that was found to bind AI-2. The predicted RbsB proteins from H. influenzae 86-028 NP and A. actinomycetemcomitans are virtually identical in length (292 vs. 293 amino acids). LsrB from A. actinomycetemcomitans is predicted to be somewhat larger (365 amino acids). Based on these observation and the similarity between the Lsr AI-2 transport system and ribose transporters, it was hypothesized that NTHI 86-028NP may utilize RbsB for internalization of and response to AI-2. An rbsB mutant was therefore generated in NTHI 86-028NP to determine the impact of this mutation on production and internalization of AI-2, biofilm formation, and the ability of NTHI to establish a chronic infection in the chinchilla model of OM.

Results

Construction and confirmation of NTHI 86-028NP rbsB::Cm

A null allele of rbsB (NTHI_0632) was constructed by insertion of a chloramphenicol resistance cassette into the coding sequence (see Experimental Procedures). This allele was introduced into NTHI 86-028NP by natural transformation and verified by PCR (data not shown). No differences in growth rate were observed for NTHI 86-028NP rbsB::Cm compared to the parental strain in rich broth culture. NTHI 86-028NP and NTHI 86-028NP rbsB::Cm also grew similarly in minimal media with glucose as a carbon source, and neither strain was able to utilize ribose as a sole carbon source (data not shown). Notably, supernatant samples taken during growth in rich media and assessed for AI-2 using the Vibrio harveyi bioluminescence assay revealed a significant accumulation of AI-2 in NTHI 86-028NP rbsB::Cm supernatants during late-exponential and early-stationary phase (Figure 1). In contrast, the level of AI-2 in supernatant harvested from cultures of the parental strain decreased at these time points, suggesting that strain NTHI 86-028NP begins to internalize AI-2 in an rbsB-dependent manner as it approaches stationary phase. Importantly, genetic complementation of NTHI 86-028NP rbsB::Cm via insertion of the wild type rbsB allele into an intergenic region of the chromosome (IRB, see Experimental Procedures) restored the ability of this strain to decrease AI-2 levels in stationary phase.

Figure 1. AI-2 accumulates in late-exponential phase cultures of NTHI 86-028NP rbsB::Cm.

Figure 1

Open in a new tab

NTHI 86-028NP, NTHI 86-028NP rbsB::Cm, and the complemented strain NTHI 86-028NP rbsB::Cm IRB::Sp were cultured in sBHI to stationary phase and supernatant samples were removed and stored at −20°C. Determination of AI-2 production in thawed supernatants was performed using the Vibrio harveyi bioluminescence assay. Significance was determined by a two-way ANOVA with a post-hoc test. *P<0.05. Error bars indicate mean and standard deviation for three independent experiments and duplicate samples for bioluminescence.

RbsB mediates AI-2 uptake in NTHI 86-028NP

To further investigate the role of rbsB in depletion or internalization of AI-2, NTHI 86-028NP luxS::Kn was utilized to study the kinetics of AI-2 depletion during growth, independent of AI-2 production. Broth cultures were inoculated with NTHI 86-028NP luxS::Kn and supplemented with the synthetic AI-2 precursor (S)-4,5-Dihydroxy-2,3-pentanedione (DPD) as indicated, and samples were taken to assess the amount of DPD remaining in the cultures over time (Figure 2A). By the time the cultures reached stationary phase, NTHI 86-028NP luxS::Kn had depleted the majority of DPD present in the culture. In contrast, only a minimal loss of DPD was observed in an uninoculated control. As expected, NTHI 86-028NP luxS::Kn did not produce detectable AI-2 at any point during the course of growth. Notably, mutation of rbsB in an NTHI 86-028NP luxS::Kn background almost completely abrogated depletion of DPD by this strain (Figure 2A). In five replicate experiments, NTHI 86-028NP luxS::Kn depleted 60% of DPD from culture supernatants on average by the time the cultures reached an OD600 of approximately 0.85, while the double mutant only depleted 15% of total DPD signal (Figure 2B). In addition, DPD levels were unaffected by cell-free culture supernatants from NTHI 86-028NP luxS::Kn and NTHI 86-028NP rbsB::Cm luxS::Kn, indicating that DPD is not being degraded by factors present in the culture supernatants (data not shown).

Figure 2. Mutation of rbsB limits AI-2 depletion.

Figure 2

Open in a new tab

NTHI 86-028NP luxS::Kn and NTHI 86-028NP rbsB::Cm luxS::Kn were cultured in sBHI media alone or sBHI supplemented with 0.2 μM of the chemically synthesized AI-2 precursor (S)-4,5-Dihydroxy-2,3-pentanedione (DPD) and samples were taken to assess depletion of DPD by V. harveyi bioluminescence. An uninoculated control sample of sBHI supplemented with DPD was included as a negative control for depletion. (A) A representative graph showing kinetics of AI-2 depletion by NTHI 86-028NP luxS::Kn and NTHI 86-028NP rbsB::Cm luxS::Kn. Error bars indicate mean and standard deviation for duplicate samples. (B) Combined results of five independent experiments with duplicate samples showing the percent of AI-2 depleted by NTHI 86-028NP luxS::Kn versus NTHI 86-028NP rbsB::Cm luxS::Kn for culture supernatants taken at an OD600 of ~0.85, normalized to uninoculated control samples. Error bars indicate mean and standard deviation.

Because rbsB is designated as the binding protein for a ribose transport system, we next tested the impact of ribose on DPD uptake. NTHI 86-028NP luxS::Kn was cultured in media supplemented with 0.2 μM DPD alone or with increasing concentrations of ribose, and samples were taken when the cultures reached an OD600 of approximately 0.85 for assessment of DPD depletion (Table 1). The addition of 0.1 mM ribose had no effect on DPD depletion, while ribose in excess of 1 mM inhibited depletion of DPD. Notably, this inhibition of depletion only occurred when ribose was present in 5000-fold molar excess, suggesting a higher affinity for AI-2 than for ribose. Incubation with millimolar concentrations of another pentose sugar (xylose) also limited DPD depletion, while the addition of sucrose had no impact (Table 1). Based on these results, we conclude that in NTHI strain 86-028NP, rbsB is part of an AI-2 transport system with lower affinity for pentose sugars.

Table 1.

Inhibition of DPD depletion

Treatment Percent DPD Depletiona Standard Deviation P Valueb
No Treatment 58.66 10.91 -
Ribose
 0.1 mM 42.68 7.92 0.07
 1.0 mM 34.12 3.57 0.01
 10 mM 2.28 10.02 < 0.01
Xylose
 0.1 mM 60.46 10.72 0.98
 1.0 mM 66.65 2.33 0.07
 10 mM 0.93 16.83 < 0.01
Sucrose
 0.1 mM 60.50 1.13 0.19
 1.0 mM 61.20 10.89 0.17
 10 mM 58.10 2.26 0.10
Open in a new tab
a

Percent DPD depleted from culture at an OD600 of approximately 0.800 compared to an uninoculated control.

b

P values determined by unpaired t test of percent DPD depletion with the indicated treatment compared to an untreated control.

RbsB transcript levels correlate with AI-2 production

For some bacterial species, expression of the AI-2 transport system is induced by the presence of AI-2 (Taga et al., 2001). It was therefore hypothesized that if the protein encoded by rbsB functions as an AI-2 binding protein, transcript levels may increase during late-exponential phase when bacteria reach peak AI-2 production. NTHI 86-028NP and NTHI 86-028NP luxS::Kn were therefore cultured in rich media to stationary phase, and samples were taken at various time points during growth to assess transcript levels for luxS and rbsB by real time RT-PCR (Figure 3). For NTHI 86-028NP, luxS transcript levels increased by approximately two-fold during late-exponential phase (Figure 3A), and rbsB transcript levels increased by approximately five-fold during late-exponential to early-stationary phase (Figure 3B). Notably, both luxS and rbsB transcript levels remained low in the luxS mutant (Figure 3A and 3B). As NTHI 86-028NP luxS::Kn was shown to deplete AI-2 from culture in Figure 2, low level expression of rbsB must still be sufficient to mediate AI-2 uptake, whereas increased expression of rbsB may be optimal for internalization of the high concentration of AI-2 encountered during early-stationary phase. Additionally, incubation of the luxS mutant with DPD resulted in a transient 8-fold increase in rbsB transcript levels compared to a control culture (Figure 3C), indicating that rbsB transcription increases in response to AI-2 signaling. From these data, we conclude that transcription of rbsB, and luxS to a lesser extent, are regulated by AI-2 signaling.

Figure 3. AI-2 increases transcription of luxS and rbsB.

Figure 3

Open in a new tab

NTHI 86-028NP and NTHI 86-028NP luxS::Kn were cultured in sBHI media to stationary phase. Samples were taken during lag phase (OD600 0.200), exponential phase (OD600 0.500), late-exponential phase (OD600 0.750), and early stationary phase (OD600 0.950) for isolation of RNA and real time RT-PCR analysis of luxS (A) and rbsB (B) transcript levels. Values represent the ratio of luxS or rbsB to gyrA transcript. Error bars indicate mean and standard deviation for duplicate samples. (C) NTHI 86-028NP luxS::Kn was cultured in sBHI to an OD600 of ~0.650 and supplemented with either 0.2 μM DPD or sterile water. Samples were taken prior to the addition of DPD or water and at 10, 30, 40, 50, and 90 minutes afterwards for RNA isolation and real time RT-PCR analysis of rbsB transcript levels. Values represent the fold increase in rbsB transcript (normalized to gyrA) for the DPD treated samples compared to untreated samples. Error bars indicate mean and standard deviation for duplicate sampling from two independent experiments.

Both rbsB and luxS promote biofilm maturation

Previous studies with the NTHI 86-028NP luxS mutant showed that loss of AI-2 quorum signaling altered overall biofilm development and maturation, resulting in biofilms with reduced biomass and average thickness (Armbruster et al., 2009). As rbsB is necessary for depletion of DPD, we hypothesized that mutation of rbsB, and the concomitant loss of AI-2 internalization would diminish biofilm maturation, as we have previously observed for luxS mutants. To test this hypothesis, NTHI 86-028NP, NTHI 86-028NP luxS::Kn, and NTHI 86-028NP rbsB::Cm were cultured in a continuous flow system to establish biofilms. Both mutants formed visible biofilms in the continuous flow system. Biofilms visualized by Live/Dead staining revealed differences in biofilm thickness by confocal laser scanning microscopy (CLSM) for NTHI 86-028NP luxS::Kn and NTHI 86-028NP rbsB::Cm compared to the parental strain (Figure 4). Representative three-dimensional images of biofilms formed by each strain demonstrate the reduced biofilm biomass for the quorum signaling mutants (Figure 4A-C). Similar to our previous work with NTHI 86-028NP luxS::Kn, both the luxS mutant and 86-028NP rbsB::Cm formed biofilms with significantly decreased biomass (Figure 4D), surface to biovolume ratio (Figure 4E), average thickness (Figure 4F), and maximum thickness (Figure 4G) compared to the parental strain.

Figure 4. Mutation of rbsB results in a similar biofilm defect as mutation of luxS.

Figure 4

Open in a new tab

Biofilms formed by NTHI 86-028NP, NTHI 86-028NP luxS::Kn, and NTHI 86-028NP rbsB::Cm under continuous flow conditions were visualized with a live/dead stain by CLSM. Z-series images were used to create representative volume views of biofilms formed by NTHI 86-028NP (A), NTHI 86-028NP luxS::Kn (B), and NTHI 86-028NP rbsB::Cm (C). Z-series images were also exported to COMSTAT to obtain biofilm measurements, including total biomass (D), surface to biovolume ratio (E), average biofilm thickness (F), and maximum biofilm thickness (G). Statistical significance determined by unpaired t test, error bars indicate standard error of the mean (n=8).

Stationary biofilms were also established by NTHI 86-028NP, NTHI 86-028NP luxS::Kn, NTHI 86-028NP rbsB::Cm, and the complemented strain NTHI 86-028NP rbsB::Cm IRB::Sp to verify the results obtained under continuous media flow conditions and to determine the impact of genetic complementation on biofilm formation by the rbsB mutant (Figure 5). Live/Dead staining and CLSM of biofilms formed by each strain demonstrated that NTHI 86-028NP luxS::Kn (Figure 5D) and NTHI 86-028NP rbsB::Cm (Figure 5C) also formed thinner biofilms with reduced total biomass (Figure 5E) and average thickness (Figure 5F) as compared to the parental strain NTHI 86-028NP (Figure 5A) when cultured under stationary conditions. Notably, genetic complementation of rbsB fully restored biofilm formation by this strain as the biofilms formed by NTHI 86-028NP rbsB::Cm IRB::Sp were similar to those formed by the parental strain (Figure 5B, 5E, 5F). From these studies, we conclude that mutation of rbsB and the resulting loss of AI-2 internalization impacts biofilm formation in a similar manner as loss of AI-2 production.

Figure 5. Genetic complementation fully restores biofilm formation by NTHI 86-028NP rbsB::Cm.

Figure 5

Open in a new tab

NTHI 86-028NP, NTHI 86-028NP luxS::Kn, NTHI 86-028NP rbsB::Cm, and the complemented strain NTHI 86-028NP rbsB::Cm IRB::rbsB were cultured in sBHI media and allowed to establish stationary biofilms for 12 h. Biofilms were visualized by live/dead staining and CLSM. Vertical z-series images were used to create representative volume views of biofilms formed by NTHI 86-028NP (A), NTHI 86-028NP rbsB::Cm IRB::rbsB (B), NTHI 86-028NP rbsB::Cm (C), and NTHI 86-028NP luxS::Kn (D). Z-series images were also exported to COMSTAT to obtain measurements of total biofilm biomass (E) and average thickness (F). Statistical significance determined by unpaired t test. Error bars indicate mean and standard deviation (n=6).

To further investigate the impact of AI-2 quorum signaling on biofilm formation, NTHI 86-028NP, NTHI 86-028NP luxS::Kn, and NTHI 86-028NP rbsB::Cm stationary biofilms were formed in the presence or absence of DPD, and biofilms were analyzed for changes in total biomass (Figure 6). In media lacking DPD, NTHI 86-028NP formed thick biofilms while NTHI 86-028NP luxS::Kn and NTHI 86-028NP rbsB::Cm biofilms had significantly decreased total biomass (Figure 6A) and average thickness (Figure 6B), as previously observed. The addition of 0.2 μM DPD to NTHI 86-028NP did not significantly alter biofilm formation. For NTHI 86-028NP luxS::Kn, the addition of DPD fully restored biofilm biomass and thickness to the level observed for the parental strain. Notably, NTHI 86-028NP rbsB::Cm did not respond to DPD, as indicated by low biofilm biomass and average thickness under both treatment conditions. Taken together, these results provide concrete evidence that the biofilm defects observed for NTHI 86-028NP luxS::Kn are the direct result of loss of AI-2 signaling rather than a metabolic defect due to mutation of luxS, and that rbsB is required by NTHI strain 86-028NP to respond to AI-2.

Figure 6. NTHI 86-028NP requires rbsB to respond to exogenous AI-2.

Figure 6

Open in a new tab

NTHI 86-028NP, NTHI 86-028NP luxS::Kn, and NTHI 86-028NP rbsB::Cm were cultured in sBHI media alone (white bars) or sBHI supplemented with 0.2 μM DPD (gray bars) and allowed to establish stationary biofilms for 12 h. Biofilms were visualized by live/dead staining and CLSM. Vertical z-series images were exported to COMSTAT for analysis of total biofilm biomass (A) and average thickness (B). Statistical significance determined by unpaired t test. **P<0.01 compared to NTHI 86-028NP. *P<0.01 compared to NTHI 86-028NP luxS::Kn. Error bars indicate mean and standard deviation (n=6).

The decreased biofilm thickness and total biomass observed for NTHI 86-028NP luxS::Kn were previously attributed in part to alterations in lipooligosaccharide (LOS) moieties (Armbruster et al., 2009). Specifically, NTHI 86-028NP luxS::Kn had decreased phosphorylcholine (PCho), an LOS modification that correlates with biofilm maturation and persistence in vivo (Hong et al., 2007a, Hong et al., 2007b). The level of surface-accessible PCho on NTHI 86-028NP rbsB::Cm was assessed by whole-bacterium ELISA for comparison to NTHI 86-028NP, NTHI 86-028NP luxS::Kn, and NTHI 86-028NP licD::Kn, a mutant lacking the PCho transferase that is thus unable to decorate LOS with PCho (Figure 7). As previously reported, NTHI 86-028NP luxS::Kn had significantly reduced surface-accessible PCho compared to the parental strain. NTHI 86-028NP rbsB::Cm also had reduced PCho as compared to the parental strain. Notably, the amount of surface-accessible PCho detected on NTHI 86-028NP rbsB::Cm was almost identical to the level detected on NTHI 86-028NP luxS::Kn. NTHI 86-028NP licD::Kn had negligible surface-accessible PCho, as expected.

Figure 7. Mutation of rbsB reduces surface-accessible PCho.

Figure 7

Open in a new tab

A modified whole-bacterium ELISA was used to detect surface phosphorylcholine using an anti-PCho monoclonal antibody. Significance determined by unpaired t test, error bars indicate standard error of the mean (n=12).

The biosynthetic genes involved in PCho decoration of LOS, as well as other LOS moieties, are subject to phase variation via slip-strand mispairing within repeat regions (Weiser, 2000). Therefore, the phase status of seven LOS biosynthetic genes was assessed for NTHI 86-028NP rbsB::Cm for comparison to the parental strain (see Experimental Procedures). No differences were observed for the phase status of genes involved in PCho decoration of LOS (Table 2). Additionally, NTHI 86-028NP rbsB::Cm phase status only differed from that of NTHI 86-028NP for oafA, which was generally in the “ON” phase for NTHI 86-028NP rbsB::Cm colonies and in the “OFF” phase within most NTHI 86-028NP colonies. Interestingly, altered phase status of oafA was also observed for NTHI 86-028NP luxS::Kn (Armbruster et al., 2009). The results of the phase status assessment clearly show that the reduction in surface-accessible PCho observed for NTHI 86-028NP rbsB::Cm was not due to phase variation of the LOS biosynthetic genes necessary for the addition of PCho to LOS.

Table 2.

Analysis of tetrameric repeat regions

NTHI 86-028NP NTHI 86-028NP rbsB::Cm
Number of colonies Number of colonies
oafA
 ON 8 7
 OFF 24 1
lex2A
 ON 2 0
 OFF 32 6
lic1A
 ON 31 8
 OFF 0 0
lic2A
 ON 24 8
 OFF 0 0
lic3A
 ON 24 7
 OFF 0 0
lic3A2
 ON 8 8
 OFF 0 0
lgtC
 ON 38 8
 OFF 0 0
a1769
 ON 8 8
 OFF 0 0
lav
 ON 0 0
 OFF 8 7
Open in a new tab
a

1769 refers to NTHI_1769, a hypothetical glycosyltransferase

Both rbsB and luxS promote the establishment of chronic NTHI infection

Our previous study of NTHI 86-028NP luxS::Kn showed that production of AI-2 was a critical factor in the ability of nontypeable H. influenzae to establish a persistent infection in the chinchilla model of otitis media (Armbruster et al., 2009). Based on the similarities between NTHI 86-028NP luxS::Kn and NTHI 86-028NP rbsB::Cm biofilm formation and PCho expression, it was hypothesized that mutation of rbsB and the resulting defect in AI-2 internalization would result in a similar persistence defect in the chinchilla model of otitis media. To test this hypothesis, chinchillas were inoculated via transbullar injection of approximately 103 cfu/ear of NTHI 86-028NP, NTHI 86-028NP luxS::Kn, or NTHI 86-028NP rbsB::Cm, and bacterial load was determined at 7, 14, 21, and 28 days post-infection. Comparable numbers of each strain were recovered at 7 and 14 days post-infection (Figure 8). By 21 days post-infection, a significant persistence defect was observed for the quorum signaling mutants with 13% of ears (2 of 16) infected with NTHI 86-028NP luxS::Kn and 31% of ears (5 of 16) infected with NTHI 86-028NP rbsB::Cm clear of visible signs of infection and with bacterial loads below the limit of detection. A similar trend was observed at 28 days post-infection, with bacterial loads below the limit of detection observed in 33% ears (2 of 6) infected with NTHI 86-028NP luxS::Kn and 50% of ears (4 of 8) infected with NTHI 86-028NP rbsB::Cm. Notably, for animals infected with the parental strain NTHI 86-028NP a bacterial load greater than the initial inocula of 103 cfu/ear was detected at all times post-infection. We thus conclude that rbsB encodes a protein critical for AI-2 quorum signaling in NTHI strain 86-028NP, that the AI-2 signaling system promotes the establishment of biofilms in vitro, and that AI-2 signaling also promotes the establishment of chronic infection by NTHI.

Figure 8. Mutation of rbsB limits bacterial persistence in vivo.

Figure 8

Open in a new tab

Chinchillas were infected via transbullar inoculation with ~103 cfu of NTHI 86-028NP (filled circles), NTHI 86-028NP luxS::Kn (open triangles), or NTHI 86-028NP rbsB::Cm (open diamonds). Total bacterial load per ear was determined by serial dilution and plating of effusion fluid, middle ear lavage, and bullar homogenate at 7, 14, 21, and 28 days post-infection. Data points represent total bacterial load for individual ears, and data shown are the combined results for three independent studies (two 21-day studies and one 28-day study). Day 7: NTHI 86-028NP n=15 ears, NTHI 86-028NP luxS::Kn n=17 ears, NTHI 86-028NP rbsB::Cm n=15 ears. Day 14: NTHI 86-028NP n=15 ears, NTHI 86-028NP luxS::Kn n=15 ears, NTHI 86-028NP rbsB::Cm n=15 ears. Day 21: NTHI 86-028NP n=14 ears, NTHI 86-028NP luxS::Kn n=16 ears, NTHI 86-028NP rbsB::Cm n=16 ears. Day 28: NTHI 86-028NP n=8 ears, NTHI 86-028NP luxS::Kn n=6 ears, NTHI 86-028NP rbsB::Cm n=8 ears. The dashed line indicates limit of detection. Error bars represent the geometric mean and 95% confidence intervals. Significance determined by log transformation of the data and unpaired t test.

Discussion

It has long been recognized that bacteria can persist within biofilm communities during chronic disease, and establishment of these bacterial communities affords protection from both immune clearance and antibiotic treatment (Anderson & O’Toole, 2008, Stewart, 2002, Ghannoum & O’Toole, 2004, Costerton et al., 2003). Due to the high rate of antibiotic resistance for chronic biofilm diseases such as OM, research focused on identifying factors which contribute to the establishment of bacterial biofilms and persistence will further our understanding of biofilm-associated disease and may aid in the development of new therapeutics. One factor which contributes to biofilm formation, maturation, and dispersal for many species is bacterial cell-to-cell communication via quorum signals (Davies et al., 1998, Hardie & Heurlier, 2008). Our previous work established that NTHI 86-028NP produces AI-2 and requires luxS to form mature biofilms and establish a chronic infection (Armbruster et al., 2009), but it was unclear as to how strain 86-028NP was responding to this signaling mediator. In this study, we used a genetic approach to identify rbsB as being required for uptake of and response to AI-2 by NTHI 86-028NP, and further addressed the role of AI-2 quorum signaling in biofilm development and persistence. The results of this study clearly show that sensing of AI-2 promotes the establishment of mature NTHI biofilm communities in vitro and during experimental OM.

While previous work from our laboratory reported a prominent role for luxS in biofilm maturation and persistence, the product of luxS is a component of the activated methyl cycle and therefore plays a role in bacterial metabolism. Thus, for some species, the effects of a luxS mutation on biofilm development and virulence can be attributed to a metabolic defect rather than interference with production and sensing of AI-2 (Rezzonico & Duffy, 2008, Sztajer et al., 2008, Winzer et al., 2002, Winzer et al., 2003). However, through the use of the AI-2 precursor DPD in the present study, we have clearly shown that the biofilm defects of a luxS mutant can be fully complemented by the addition of quorum signal. The results of this study also prove that mutation of a gene involved in the response to AI-2 in a strain with luxS intact still results in defects in biofilm maturation, indicating that this phenotype in NTHI 86-028NP is due to loss of AI-2 quorum signaling rather than any potential metabolic impact of the luxS mutation. Both the luxS mutant and the rbsB mutant exhibited decreased phosphorylcholine levels, indicating that AI-2 signaling regulates the formation of mature NTHI biofilms at least in part through modulation of LOS composition (Armbruster et al., 2009). Additionally, the results of the phase status analyses in concert with our previous study of the luxS mutant clearly exclude the possibility of an unlinked phase variation event resulting in the observed changes in biofilm maturation.

In E. coli, the LuxS-regulated (Lsr) operon for internalization of AI-2 includes lsrA, lsrC, lsrD, lsrB, the genes lsrF, lsrG, and lsrE that appear to be involved in degradation of AI-2, and a divergently transcribed gene cluster containing lsrK and the repressor lsrR (Li et al., 2007, Taga et al., 2003, Wang et al., 2005, Xavier & Bassler, 2005). In contrast, the ribose transport operon of NTHI 86-028NP which includes rbsB is composed of rbsD, rbsA, rbsC, rbsB, rbsK, and rbsR (Harrison et al., 2005). This configuration is more similar to the σ-ribose transport operon of Aggregatibacter actinomycetemcomitans HK1651 (www.oralgen.lanl.gov) than to the Lsr transport operon. Additionally, rbsK and rbsR do not appear to be divergently transcribed in NTHI, and the genes involved in degradation of AI-2 are lacking. However, as NTHI 86-028NP lacks a true Lsr transporter, this strain appears to be utilizing the putative ribose transporter as an AI-2/LuxS-regulated transporter for uptake of and response to AI-2. In support of this hypothesis, transcription of the lsr operon in S. typhimurium and E. coli is regulated by AI-2 and thus decreased in a luxS mutant (Taga et al., 2001, Xavier & Bassler, 2005). Therefore, the data reported herein showing low rbsB transcript levels in NTHI 86-028NP luxS::Kn and upregulation of transcript levels following the addition of DPD prove that rbsB is regulated by LuxS and AI-2 in NTHI 86-028NP.

Based on these results, rbsB in NTHI 86-028NP clearly represents an AI-2-regulated and luxS-regulated (Lsr) gene even though the operon containing rbsB is more similar to the ribose transport operon of A. actinomycetemcomitans than the Lsr operon of E. coli. In addition, the minimal impact of pentose sugars on depletion of DPD (unless in excess of 5000 times the DPD concentration) supports the conclusion that, for NTHI strain 86-028NP, rbsB is utilized as an AI-2 transporter at high cell density. It is therefore intriguing to speculate that the rest of the ribose transport operon of NTHI 86-028NP may be similarly regulated by AI-2 and function as part of a high-affinity AI-2 transporter, in addition to its possible function as a pentose sugar transporter. However, further research will be necessary to determine how the operon is regulated without apparent divergent transcription of the operon repressor. The downstream phosphorylation and detection of AI-2 and the signaling induced in response to detection of AI-2 also remain to be investigated in NTHI.

In the context of experimental OM, both the luxS mutant and the rbsB mutant were below the limit of detection in 13% to 50% of samples taken at 21 days post-infection or later while NTHI 86-028NP continued to persist at high levels in all animals infected with this strain. Based on these data quorum signaling does not appear to be required for acute infection, but the ability to both produce and respond to the AI-2 signal clearly contribute to the establishment of a chronic infection. This defect in persistence during chronic infection may be related to the biofilm defects observed for the quorum signaling mutants, in that minimal biofilm formation by the mutant strains may be sufficient to promote persistence during acute infection, but the formation of mature phosphorylcholine-positive biofilm communities may be necessary for the establishment and maintenance of a chronic infection and evasion of the host immune response.

In addition, while a clear trend towards clearance was observed for the quorum signaling mutants it is notable that the animals which still had detectable levels of the mutant strains at later time points carried a similar bacterial load as that observed for animals infected with the parental strain. As these infection studies are conducted using an outbred population, this difference between apparent clearance and persistence may reflect heterogeneity in the immune responses of individual chinchillas and their ability to combat the infection. However, as all 52 ears infected with the parental strain exhibited a bacterial load higher than the initial inocula regardless of the number of days post-infection, these results suggest that the quorum signaling mutants are most likely more susceptible to factors of the host immune response than the parental strain. The results of the infection studies thus suggest that either only some of the animals were able to mount a sufficient immune response to promote clearance of the mutant strains, or that the quorum signaling mutants managed to overcome the defect in AI-2 signaling and/or biofilm maturation to establish a persistent population in some of the infected ears. Further research will be necessary to distinguish between these possibilities.

Similar to the results of the in vitro biofilm studies, the results of experimental infection with the luxS mutant and the rbsB mutant prove that the persistence defect previously observed for NTHI 86-028NP luxS::Kn was due to loss of AI-2 quorum signaling rather than any potential metabolic defects or to alterations in other quorum signaling networks, such as AI-3, which may be impacted by a luxS mutation (Kendall et al., 2007). However, it still remains a possiability that host factors produced during the course of infection could impact NTHI signaling networks and contribute to biofilm formation and persistence. The data presented herein thus support a prominent role for AI-2 quorum signaling in concert with other as yet undetermined factors that promote the establishment of a chronic, biofilm-associated infection. It is also possible that a mutant deficient in both the ability to produce and respond to AI-2 may have a more severe persistence defect beyond what was observed for either quorum signaling mutant strain.

NTHI remains a leading cause of OM and other upper respiratory tract infections, despite promising vaccine candidates. Our data indicate that AI-2 quorum signaling is utilized by strain 86-028NP to promote biofilm formation and establishment of chronic, persistent infections. Furthermore, our previous work indicates that AI-2 produced by NTHI impacts Moraxella catarrhalis biofilm formation, antibiotic resistance, and persistence in the chinchilla model of OM (Armbruster et al., 2010, Verhaegh et al.). Therefore, therapeutics aimed at interfering with AI-2 binding and signaling may prevent biofilm development or disrupt already established biofilms to augment current OM treatments. For NTHI 86-028NP, rbsB appears to be a promising target for interference with AI-2 signaling as mutation of this gene clearly has a significant biological impact on AI-2 uptake, biofilm development, and persistence during experimental OM. However, it should be noted that there is considerable genomic heterogeneity among NTHI strains and that some NTHI strains possess both the ribose transport operon and an Lsr transport operon, similar to A. actinomycetemcomitans. Thus, it is possible that other NTHI strains may utilize different means for uptake of quorum signal, or that other strains may respond to AI-2 signaling in a different manner. Further investigation into transport and sensing of AI-2 in NTHI will allow for identification of virulence factors controlled by AI-2 signaling, and may reveal new target candidates for disruption of quorum signaling as a therapy for treatment of NTHI infections.

Experimental Procedures

Bacterial strains and culture conditions

All strains of nontypeable H. influenzae were cultivated in brain heart infusion (BHI) medium (Difco) supplemented with 10 μg/mL hemin (ICN Biochemicals) and 10 μg/mL NAD (Sigma); this medium is referred to below as supplemented BHI (sBHI). Nontypeable H. influenzae (NTHI) strain 86-028NP is a nasopharyngeal isolate from a child with chronic otitis media (Bakaletz et al., 1988) for which the genomic sequence has been determined (Harrison et al., 2005). NTHI 86-028NP luxS::Kn contains a kanamycin resistance cassette disrupting the autoinducer-2 synthase (NTHI_0621) (Armbruster et al., 2009). NTHI 86-028NP licD::Kn contains a kanamycin resistance cassette disrupting the phosphorylcholine transferase (NTHI_1594) (West-Barnette et al., 2006).

Generation of NTHI 86-028NP rbsB::Cm

A 2313 bp DNA fragment containing rbsB (NTHI_0632) was amplified from NTHI 86-028NP genomic DNA using primers GCA ATC GCC GCT TCA ATG G and CTA CCG CTA CCC CCG TCA GG. The amplicon was ligated into pCR2.1 following InVitrogen protocol to generate pCR-rbsB. pCR-rbsB was digested with SspI, a unique site within rbsB, for ligation with a chloramphenicol resistance cassette from pCMR (Whitby et al., 1998). The resulting construct was transformed into chemically-competent E. coli to generate pCR-rbsB::Cm. The expected sequence was confirmed by PCR with primers GGT TTG GCT GTT TCT GGC TCT GC and CGA TAG TTG CTG CCA TTT TGC CAC, and by sequence analysis. The plasmid DNA was linearized with BamHI and introduced into NTHI 86-028NP via natural transformation as described previously (Hong et al., 2007a, Hong et al., 2007b). Transformants were plated onto sBHI agar containing 1.5 μg/ml chloramphenicol (Sigma) and incubated for 48 h at 37 °C. Transformants carrying the expected construct were verified by PCR.

Generation of NTHI 86-028NP rbsB::Cm IRB::rbsB

A 1363 bp intergenic region (herein designated as intergenic region B, or IRB) is contained within the NTHI 86-028NP genome between NTHI_1846 and NTHI_1849. A 2248 bp segment containing IRB was amplified from NTHI 86-028NP genomic DNA using primers AGC TTC CGC CGC CAG TAA AAT and CCA AAG CGT CAG CGG ATG C. The amplicon was ligated in pCR2.1 using the TA Cloning Kit (InVitrogen) to create pCR-IRB. White colonies were verified to contain IRB using primers CCG CGT CCG AAA TAT GTG CTA C and ACT CCG CGT TAT CCC CGT CTT A. pCR-IRB was digested with SspI and dephosphorylated. A spectinomycin resistance cassette was excised from pSpec (Whitby et al., 1998) with EcoRV and ligated into the prepared pCR-IRB vector. The resulting construct was transformed into chemically-competent E. coli to generate pCR-IRB::Sp. pCR-IRB::Sp was digested with StuI and dephosphorylated. pCR-rbsB (described above) was digested with BtgI and BsmFI, and the 1134 bp band containing rbsB was gel extracted and ligated into pCR-IRB::Sp. The resulting construct was transformed into chemically-competent E. coli to generate pCR-IRB::rbsB. Transformants were selected by growth on LB plates with 100 μg/ml spectinomycin (Sigma), screened using primers M13F (a universal primer in pCR2.1) and CAC GGT ACC ATT TCT TGC TG (within the spectinomycin resistance cassette), and verified by restriction digest with NdeI and EcoRV to generate three bands (4118 bp, 3435 bp, and 938 bp). pCR-IRB::rbsB was linearized using BamHI and introduced into NTHI 86-028NP rbsB::Cm via natural transformation. Transformants were plated onto sBHI agar containing 30 μg/ml spectinomycin, and verified by PCR to contain both rbsB and rbsB::Cm using primers GGT TTG GCT GTT TCT GGC TCT GC and CGA TAG TTG CTG CCA TTT TGC CAC.

Generation of NTHI 86-028NP rbsB::Cm luxS::Kn

A double mutant strain was generated by disruption of luxS (NTHI _0621) in NTHI 86-028NP rbsB::Cm. The previously published plasmid pUCluxS::Kn (Armbruster et al., 2009) was linearized with SfoI and introduced into NTHI 86-028NP rbsB::Cm via natural transformation. Transformants were plated onto sBHI agar containing 30 μg/ml ribostamycin (Sigma), and verified by PCR to contain both luxS::Kn and rbsB::Cm.

Autoinducer-2 production and depletion

NTHI 86-028NP and NTHI 86-028NP rbsB::Cm from overnight plate cultures were diluted in 20 ml of sBHI to an OD600 of approximately 0.05 and grown to stationary phase in sidearm flasks at 37°C with shaking at 150 rpm. Samples were taken at hourly intervals, centrifuged for 3 minutes at 13000 rpm, and supernatants were stored at −20°C until bioluminescence assays were performed. Luminescence produced by V. harveyi BB170 (Bassler et al., 1994) following 3 h incubation with supernatant samples was determined in a Turner Designs TD-20/20 luminometer for 10 s. Data are reported as relative light units [counts per 10 s].

For AI-2 uptake studies, sBHI was supplemented with 0.2 μM (S)-4,5-Dihydroxy-2,3-pentanedione (DPD, Omm Scientific) when indicated, inoculated with ~108 cfu of NTHI (OD600 ~0.15), and incubated at 37°C and 150 rpm. Samples were taken at hourly intervals, centrifuged, and stored at −20°C as above. This concentration of DPD was chosen as it elicits approximately equivalent luminescence from V. harveyi as NTHI 86-028NP late exponential phase culture supernatant (OD600 of approximately 0.750). For inhibition studies, NTHI 86-028NP luxS::Kn cultures were supplemented with ribose (Sigma), xylose (Sigma), or sucrose (Sigma) to a final concentration of 0.1, 1.0, or 10 mM as indicated and samples were taken at mid-logarithmic phase of growth (OD600 of ~0.85).

Real time RT-PCR

NTHI 86-028NP and NTHI 86-028NP luxS::Kn were grown in sBHI to stationary phase, and 0.5 ml samples were taken at the indicated optical densities for isolation of RNA as per RNeasy Mini kit instructions (Qiagen). On-column DNase treatment of samples was performed using the RNase-Free DNase set (Qiagen). RNA concentration and purity were determined using a NanoDrop ND-1000 spectrophotometer (version 3.1.2). RT-PCR master mixes were prepared using the TaqMan One-Step RT-PCR Master Mix Reagents kit (Applied Biosystems). A complete list of primers and probes is provided in Table 3 (all primers were purchased from Eurofins, and Fam/Tamra probes were purchased from Sigma). Each RT-PCR reaction was conducted in duplicate using 100 ng total RNA in a final volume of 25 μl per well. Thermocycling conditions in an ABI Prism 7000 Sequence Detection System (Applied Biosystems) were as follows: 1 cycle of 48°C for 30 minutes, 1 cycle of 95°C for 10 minutes, and 40 cycles of 95°C for 15 seconds and 60°C for 1 minute. Transcript levels were determined for luxS, rbsB, and gyrA using the 7000 System SDS Software (version 1.2.3f2). For each analysis, control experiments lacking reverse transcriptase treatment were performed and showed no detectable product.s

Table 3.

Real time PCR primers

Primer Sequence
gyrA Forward 5′-GCATTACCTGACGTTCGAGATG-3′
gyrA Reverse 5′-CCTTCGCGATCCATTGAGAA-3′
gyrA Probe 5′-6TTTAAAACCAGTTCACCGCCGCGTAC0-3′
luxS Forward 5′-TGCGGAAGCTATACGGAACA-3′
luxS Reverse 5′-CCTATACCGCGTGCGATAACA-3′
luxS Probe 5′-6TCCTTAGAAGATGCACACGAAATTGCCAA0-3′
rbsB Forward 5′-TGATGCTCATAAATTCAATGTGCTT-3′
rbsB Reverse 5′-TTTCCGTTACATTCAAACCTTTTG-3′
rbsB Probe 5′-6CCAGTCAGCCAGCAGATTTTGATCGA0-3′
Open in a new tab

Continuous flow system

NTHI biofilms were established in a continuous-flow system as previously described (Hong et al., 2007b). NTHI 86-028NP, NTHI 86-028NP luxS::Kn, and NTHI 86-028NP rbsB::Cm were diluted to ~108 cfu/ml in sBHI broth, injected into the port of a commercial microscopy flow-cell (Stovall), and incubated for 3 h at 37°C without medium flow to permit bacterial surface attachment. Continuous flow of sBHI media was initiated at a rate of ~60 ml per hour and maintained for 24 h.

Stationary biofilm system

NTHI stationary biofilms were established in Lab-TekII two-chamber no. 1.5 German coverglass slides (Nunc). NTHI 86-028NP, NTHI 86-028NP luxS::Kn, NTHI 86-028NP rbsB::Cm, and NTHI 86-028NP rbsB::Cm IRB::rbsB were diluted to ~108 cfu/ml in sBHI broth (or sBHI supplemented with 0.2 μM DPD as indicated). Chambers of the coverglass slides were inoculated with 2 ml of the bacterial suspension, and biofilms were established at 37°C and 5% CO2 for 12 hours.

Confocal Laser Scanning Microscopy (CLSM)

Biofilm samples were washed once with PBS and stained with a LIVE/DEAD BacLight viability kit (InVitrogen). Z-series images of biofilms were collected using either a Zeiss LSM510 CLSM microscope or a Nikon Eclipse C1 CLSM microscope. Between five and eight image stacks, each representing a different field of view, were compiled for each strain. The Z-series images were visualized using the Nikon Elements software and exported into MATLAB (version 5.1) for COMSTAT analysis as previously described (Heydorn et al., 2002).

Whole-bacteria ELISA

ELISAs were performed on intact bacterial cells as described previously (Jones et al., 2002, Armbruster et al., 2009). Bacteria were diluted in distilled H2O to an OD600 of 0.150 (~108 cfu/ml), and 0.1 ml was dispensed into each well of a MaxiSorp-coated 96-well microtiter plate (Nunc) and dried overnight uncovered at 37°C for 24 h. Plates were washed once and blocked for 30 min in TSBB buffer (10 mM Tris, 0.5 M NaCl, 0.5% Tween, pH = 8.0). Wells were incubated with 0.1 ml anti-phosphorylcholine monoclonal antibody HAS (Statens Serum Institut) overnight at room temperature, washed once, and incubated for 1 h in the dark with goat-anti-mouse IgM-horseradish peroxidase (HRP) conjugate. Wells were then washed once and incubated in the dark for approximately 20 min with 0.1 ml of TMB substrate reagent (OptEIA, Becton-Dickinson). The reaction was stopped with 0.05 ml of 2 N H2SO4, and OD450 measured using a plate reader (Labsystems Multiskan Plus).

Assessment of phase-variable genes

Determination of ON/OFF state of phase-variable genes was performed as previously described (Erwin et al., 2006, Armbruster et al., 2009). Individual colonies of each NTHI strain were harvested from plate cultures suspended in 0.1 ml of 10% suspension of Chelex (BioRad). Genomic DNA was prepared by vortexing suspensions for 15 sec, boiling for 10 min, chilling on ice for 2 min, and centrifugation at 8000 x g for 2 min. Supernatants (0.05 ml) were transferred to fresh tubes for PCR amplification and sizing by means of GeneMapper (Applied Biosystems, Inc., Foster City, CA) analysis at the DNA Sequencing Facility of the WFUHS Biomolecular Resource Laboratory as previously described (Armbruster et al., 2009). Numbers of repeats for each region were estimated based on fragment size, and used in accordance with the previously defined genomic sequence of each region to define “on” or “off” status of each phase-variable region.

Chinchilla infection studies

Bacterial persistence and biofilm formation in the middle ear cavity were assessed as described previously (Hong et al., 2007a, Hong et al., 2007b, Armbruster et al., 2009). Chinchillas were purchased from Rauscher’s chinchilla ranch (Larue, OH) and allowed to acclimate to the vivarium for >7 d prior to infection. No animals showed visible signs of illness prior to infection. The animals were anesthetized with isofluorane and infected via transbullar injection with ~103 CFU of NTHI 86-028NP, NTHI 86-028NP luxS::Kn, or NTHI 86-028NP rbsB::Cm. All inocula were confirmed by plate-counts. The degree of inflammation was assessed at 48 h intervals by digital otoscopic examination using a qualitative scoring system of 1 to 4, with 4 being the most severe (Hong et al., 2007a). Animals were euthanized at 7, 14, 21, and 28 days post-infection and bullae were aseptically opened. Fluids within the middle ear were recovered and middle ear lavage was performed using 1.0 ml sterile PBS. Both bullae were then excised and homogenized in 10 ml sterile PBS (Hong et al., 2007a). Total cfu/ear were determined by combining bacterial counts obtained for homogenized bullae and middle-ear effusion fluid with lavage.

Statistics

Significance was determined by nonparametric t test, unpaired t test, or two-way ANOVA with post-hoc tests of significance as indicated. All P values are two-tailed at a 95% confidence interval. Analyses were performed using GraphPad Prism, version 5 (GraphPad Software, San Diego, CA).

Acknowledgments

The authors acknowledge excellent assistance by Gayle Foster, Jim Turner, the WFU sequencing facility, and helpful comments and critiques from Dr. Steven Mizel and Dr. Marlena Westcott. This work was supported by grants from NIH/NIDCD (DC007444 and DC10051) awarded to W. E. Swords. Chelsie E. Armbruster and Kristin E.D. Weimer were supported by an NIH training grant (T32 AI07401).

References

  1. Anderson GG, O’Toole GA. Innate and induced resistance mechanisms of bacterial biofilms. Curr Top Microbiol Immunol. 2008;322:85–105. doi: 10.1007/978-3-540-75418-3_5. [DOI] [PubMed] [Google Scholar]
  2. Armbruster CE, Hong W, Pang B, Dew KE, Juneau RA, Byrd MS, Love CF, Kock ND, Swords WE. LuxS promotes biofilmmaturation and persistence of nontypeable Haemophilus influenzae in vivo via modulation of lipooligosaccharides on the bacterial surface. Infect Immun. 2009;77:4081–4091. doi: 10.1128/IAI.00320-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Armbruster CE, Hong W, Pang B, Weimer KED, Juneau RA, Turner J, Swords WE. Indirect Pathogenicity of Haemophilus influenzae and Moraxella catarrhalis in Polymicrobial Otitis Media Occurs via Interspecies Quorum Signaling. mBio. 2010;1(3):e00102-00110–e00102-00119. doi: 10.1128/mBio.00102-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Armbruster CE, Swords WE. Interspecies bacterial communication as a target for therapy in otitis media. Expert Rev Anti Infect Ther. 2010;8:1067–1070. doi: 10.1586/eri.10.109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bakaletz LO. Bacterial biofilms in otitis media: evidence and relevance. Pediatr Infect Dis J. 2007;26:S17–19. doi: 10.1097/INF.0b013e318154b273. [DOI] [PubMed] [Google Scholar]
  6. Bakaletz LO, Tallan BM, Hoepf T, DeMaria TF, Birck HG, Lim DJ. Frequency of fimbriation of nontypable Haemophilus influenzae and its ability to adhere to chinchilla and human respiratory epithelium. Infect Immun. 1988;56:331–335. doi: 10.1128/iai.56.2.331-335.1988. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bassler BL, Wright M, Silverman MR. Multiple signalling systems controlling expression of luminescence in Vibrio harveyi: sequence and function of genes encoding a second sensory pathway. Mol Microbiol. 1994;13:273–286. doi: 10.1111/j.1365-2958.1994.tb00422.x. [DOI] [PubMed] [Google Scholar]
  8. Costerton JW, Stewart PS, Greenberg EP. Bacterial biofilms: a common cause of persistent infections. Science. 1999;284:1318–1322. doi: 10.1126/science.284.5418.1318. [DOI] [PubMed] [Google Scholar]
  9. Costerton W, Veeh R, Shirtliff M, Pasmore M, Post C, Ehrlich G. The application of biofilm science to the study and control of chronic bacterial infections. J Clin Invest. 2003;112:1466–1477. doi: 10.1172/JCI20365. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Davies DG, Parsek MR, Pearson JP, Iglewski BH, Costerton JW, Greenberg EP. The Involvement of Cell-to-Cell Signals in the Development of a Bacterial Biofilm. Science. 1998;280:295–298. doi: 10.1126/science.280.5361.295. [DOI] [PubMed] [Google Scholar]
  11. Ehrlich GD, Veeh R, Wang X, Costerton JW, Hayes JD, Hu FZ, Daigle BJ, Ehrlich MD, Post JC. Mucosal biofilm formation on middle-ear mucosa in the chinchilla model of otitis media. JAMA. 2002;287:1710–1715. doi: 10.1001/jama.287.13.1710. [DOI] [PubMed] [Google Scholar]
  12. Erwin AL, Allen S, Ho DK, Bonthuis PJ, Jarisch J, Nelson KL, Tsao DL, Unrath WC, Watson ME, Jr, Gibson BW, Apicella MA, Smith AL. Role of lgtC in resistance of nontypeable Haemophilus influenzae strain R2866 to human serum. Infect Immun. 2006;74:6226–6235. doi: 10.1128/IAI.00722-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Erwin AL, Smith AL. Nontypeable Haemophilus influenzae: understanding virulence and commensal behavior. Trends Microbiol. 2007;15:355–362. doi: 10.1016/j.tim.2007.06.004. [DOI] [PubMed] [Google Scholar]
  14. Faden H, Waz MJ, Bernstein JM, Brodsky L, Stanievich J, Ogra PL. Nasopharyngeal flora in the first three years of life in normal and otitis-prone children. Ann Otol Rhinol Laryngol. 1991;100:612–615. doi: 10.1177/000348949110000802. [DOI] [PubMed] [Google Scholar]
  15. Finkelstein JA, Davis RL, Dowell SF, Metlay JP, Soumerai SB, Rifas-Shiman SL, Higham M, Miller Z, Miroshnik I, Pedan A, Platt R. Reducing Antibiotic Use in Children: A Randomized Trial in 12 Practices. Pediatrics. 2001;108:1–7. doi: 10.1542/peds.108.1.1. [DOI] [PubMed] [Google Scholar]
  16. Ghannoum MA, O’Toole GA. Microbial biofilms. ASM Press; Washington, D.C.: 2004. p. xiii.p. 426. [410] p. of plates. [Google Scholar]
  17. Hall-Stoodley L, Hu FZ, Gieseke A, Nistico L, Nguyen D, Hayes JD, Forbes M, Greenberg DP, Dice B, Burrows A, Wackym P, Stoodley P, Post JC, Ehrlich GD, Kerschner JE. Direct detection of bacterial biofilms on the middle ear mucosa of children with chronic otitis media. JAMA. 2006;296:202–211. doi: 10.1001/jama.296.2.202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hardie KR, Heurlier K. Establishing bacterial communities by ‘word of mouth’: LuxS and autoinducer 2 in biofilm development. Nat Rev Microbiol. 2008;6:635–643. doi: 10.1038/nrmicro1916. [DOI] [PubMed] [Google Scholar]
  19. Harrison A, Dyer DW, Gillaspy A, Ray WC, Mungur R, Carson MB, Zhong H, Gipson J, Gipson M, Johnson LS, Lewis L, Bakaletz LO, Munson RS., Jr Genomic sequence of an otitis media isolate of nontypeable Haemophilus influenzae: comparative study with H. influenzae serotype d, strainKW20. J Bacteriol. 2005;187:4627–4636. doi: 10.1128/JB.187.13.4627-4636.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Henke JM, Bassler BL. Three parallel quorum-sensing systems regulate gene expression in Vibrio harveyi. J Bacteriol. 2004;186:6902–6914. doi: 10.1128/JB.186.20.6902-6914.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Heydorn A, Ersboll B, Kato J, Hentzer M, Parsek MR, Tolker-Nielsen T, Givskov M, Molin S. Statistical analysis of Pseudomonas aeruginosa biofilm development: impact of mutations in genes involved in twitching motility, cell-to-cell signaling, and stationary-phase sigma factor expression. Appl Environ Microbiol. 2002;68:2008–2017. doi: 10.1128/AEM.68.4.2008-2017.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hoa M, Tomovic S, Nistico L, Hall-Stoodley L, Stoodley P, Sachdeva L, Berk R, Coticchia JM. Identification of adenoid biofilms with middle ear pathogens in otitis-prone children utilizing SEM and FISH. Int J Pediatr Otorhinolaryngol. 2009;73:1242–1248. doi: 10.1016/j.ijporl.2009.05.016. [DOI] [PubMed] [Google Scholar]
  23. Hong W, Mason K, Jurcisek JA, Novotny LA, Bakaletz LO, Swords WE. Phosphorylcholine decreases early inflammation and promotes the establishment of stable biofilm communities of nontypeable Haemophilus influenzae strain 86–028NP in a chinchilla model of otitis media. Infect Immun. 2007a;75:958–965. doi: 10.1128/IAI.01691-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hong W, Pang B, West-Barnette S, Swords WE. Phosphorylcholine expression by nontypeable Haemophilus influenzae correlates with maturation of biofilm communities in vitro and in vivo. J Bacteriol. 2007b;189:8300–8307. doi: 10.1128/JB.00532-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. James D, Shao H, Lamont RJ, Demuth DR. The Actinobacillus actinomycetemcomitans ribose binding protein RbsB interacts with cognate and heterologous autoinducer 2 signals. Infect Immun. 2006;74:4021–4029. doi: 10.1128/IAI.01741-05. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Jayaraman A, Wood TK. Bacterial quorum sensing: signals, circuits, and implications for biofilms and disease. Annu Rev Biomed Eng. 2008;10:145–167. doi: 10.1146/annurev.bioeng.10.061807.160536. [DOI] [PubMed] [Google Scholar]
  27. Jones PA, Samuels NA, Phillips NJ, Munson RS, Bozue JA, Arseneau JA, Nichols WA, Zaleski A, Gibson BW, Apicella MA. Haemophilus influenzae type B strain A2 has multiple sialyltransferases involved in lipooligosaccharide sialylation. J Biol Chem. 2002;277:14598–14611. doi: 10.1074/jbc.M110986200. [DOI] [PubMed] [Google Scholar]
  28. Kendall MM, Rasko DA, Sperandio V. Global effects of the cell-to-cell signaling molecules autoinducer-2, autoinducer-3, and epinephrine in a luxS mutant of enterohemorrhagic Escherichia coli. Infect Immun. 2007;75:4875–4884. doi: 10.1128/IAI.00550-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Klein JO. The burden of otitis media. Vaccine. 2000;19:S2–8. doi: 10.1016/s0264-410x(00)00271-1. [DOI] [PubMed] [Google Scholar]
  30. Li J, Attila C, Wang L, Wood TK, Valdes JJ, Bentley WE. Quorum sensing in Escherichia coli is signaled by AI-2/LsrR: effects on small RNA and biofilm architecture. J Bacteriol. 2007;189:6011–6020. doi: 10.1128/JB.00014-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Mandel EM, Doyle WJ, Winther B, Alper CM. The incidence, prevalence and burden of OM in unselected children aged 1–8 years followed by weekly otoscopy through the “common cold” season. Int J Pediatr Otorhinolaryngol. 2008;72:491–499. doi: 10.1016/j.ijporl.2007.12.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Michiko E, Taga JLSBLB. The LuxS-dependent autoinducer AI-2 controls the expression of an ABC transporter that functions in AI-2 uptake in Salmonella typhimurium. Mol Microbiol. 2001;42:777–793. doi: 10.1046/j.1365-2958.2001.02669.x. [DOI] [PubMed] [Google Scholar]
  33. Miller MB, Bassler BL. Quorum sensing in bacteria. Annu Rev Microbiol. 2001;55:165–199. doi: 10.1146/annurev.micro.55.1.165. [DOI] [PubMed] [Google Scholar]
  34. Miller ST, Xavier KB, Campagna SR, Taga ME, Semmelhack MF, Bassler BL, Hughson FM. Salmonella typhimurium recognizes a chemically distinct form of the bacterial quorum-sensing signal AI-2. Mol Cell. 2004;15:677–687. doi: 10.1016/j.molcel.2004.07.020. [DOI] [PubMed] [Google Scholar]
  35. Nistico L, Kreft R, Gieseke A, Coticchia JM, Burrows A, Khampang P, Liu Y, Kerschner JE, Post JC, Lonergan S, Sampath R, Hu FZ, Ehrlich GD, Stoodley P, Hall-Stoodley L. An Adenoid Reservoir for Pathogenic Biofilm Bacteria. J Clin Microbiol. 2011 doi: 10.1128/JCM.00756-10. JCM.00756–00710. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Pereira CS, McAuley JR, Taga ME, Xavier KB, Miller ST. Sinorhizobium meliloti, a bacterium lacking the autoinducer-2 (AI-2) synthase, responds to AI-2 supplied by other bacteria. Mol Microbiol. 2008;70:1223–1235. doi: 10.1111/j.1365-2958.2008.06477.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Pichichero ME. Recurrent and persistent otitis media. Pediatr Infect Dis J. 2000;19:911–916. doi: 10.1097/00006454-200009000-00034. [DOI] [PubMed] [Google Scholar]
  38. Post JC. Direct evidence of bacterial biofilms in otitis media. Laryngoscope. 2001;111:2083–2094. doi: 10.1097/00005537-200112000-00001. [DOI] [PubMed] [Google Scholar]
  39. Post JC, Hiller NL, Nistico L, Stoodley P, Ehrlich GD. The role of biofilms in otolaryngologic infections: update 2007. Curr Opin Otolaryngol Head Neck Surg. 2007;15:347–351. doi: 10.1097/MOO.0b013e3282b97327. [DOI] [PubMed] [Google Scholar]
  40. Rezzonico F, Duffy B. Lack of genomic evidence of AI-2 receptors suggests a non-quorum sensing role for luxS in most bacteria. BMC Microbiol. 2008;8:154. doi: 10.1186/1471-2180-8-154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Shao H, James D, Lamont RJ, Demuth DR. Differential interaction of Aggregatibacter (Actinobacillus) actinomycetemcomitans LsrB and RbsB proteins with autoinducer 2. J Bacteriol. 2007a;189:5559–5565. doi: 10.1128/JB.00387-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Shao H, Lamont RJ, Demuth DR. Autoinducer 2 Is Required for Biofilm Growth of Aggregatibacter (Actinobacillus) actinomycetemcomitans. Infect Immun. 2007b;75:4211–4218. doi: 10.1128/IAI.00402-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Stewart PS. Mechanisms of antibiotic resistance in bacterial biofilms. Int J Med Microbiol. 2002;292:107–113. doi: 10.1078/1438-4221-00196. [DOI] [PubMed] [Google Scholar]
  44. Sztajer H, Lemme A, Vilchez R, Schulz S, Geffers R, Yip CY, Levesque CM, Cvitkovitch DG, Wagner-Döbler I. Autoinducer-2-regulated genes in Streptococcus mutans UA159 and global metabolic effect of the luxS mutation. J Bacteriol. 2008 doi: 10.1128/JB.01086-07. Epub ahead of print. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Taga ME, Miller ST, Bassler BL. Lsr-mediated transport and processing of AI-2 in Salmonella typhimurium. Mol Microbiol. 2003;50:1411–1427. doi: 10.1046/j.1365-2958.2003.03781.x. [DOI] [PubMed] [Google Scholar]
  46. Taga ME, Semmelhack JL, Bassler BL. The LuxS-dependent autoinducer AI-2 controls the expression of an ABC transporter that functions in AI-2 uptake in Salmonella typhimurium. Molecular Microbiology. 2001;42:777–793. doi: 10.1046/j.1365-2958.2001.02669.x. [DOI] [PubMed] [Google Scholar]
  47. Verhaegh SJ, Snippe ML, Levy F, Verbrugh HA, Jaddoe VW, Hofman A, Moll HA, van Belkum A, Hays JP. Colonization of healthy children by Moraxella catarrhalis is characterized by genotype heterogeneity, virulence gene diversity and co-colonization with Haemophilus influenzae. Microbiology. 157:169–178. doi: 10.1099/mic.0.042929-0. [DOI] [PubMed] [Google Scholar]
  48. Wang L, Li J, March JC, Valdes JJ, Bentley WE. luxS-dependent gene regulation in Escherichia coli K-12 revealed by genomic expression profiling. J Bacteriol. 2005;187:8350–8360. doi: 10.1128/JB.187.24.8350-8360.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Waters CM, Bassler BL. Quorum sensing: cell-to-cell communication in bacteria. Annu Rev Cell Dev Biol. 2005;21:319–346. doi: 10.1146/annurev.cellbio.21.012704.131001. [DOI] [PubMed] [Google Scholar]
  50. Weiser JN. The generation of diversity by Haemophilus influenzae. Trends Microbiol. 2000;8:433–435. doi: 10.1016/s0966-842x(00)01839-4. [DOI] [PubMed] [Google Scholar]
  51. West-Barnette S, Rockel A, Swords WE. Biofilm growth increases phosphorylcholine content and decreases potency of nontypeable Haemophilus influenzae endotoxins. Infect Immun. 2006;74:1828–1836. doi: 10.1128/IAI.74.3.1828-1836.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Whitby PW, Morton DJ, Stull TL. Construction of antibiotic resistance cassettes with multiple paired restriction sites for insertional mutagenesis of Haemophilus influenzae. FEMS Microbiol Lett. 1998;158:57–60. doi: 10.1111/j.1574-6968.1998.tb12800.x. [DOI] [PubMed] [Google Scholar]
  53. Winzer K, Hardie KR, Burgess N, Doherty N, Kirke D, Holden MT, Linforth R, Cornell KA, Taylor AJ, Hill PJ, Williams P. LuxS: its role in central metabolism and the in vitro synthesis of 4-hydroxy-5-methyl-3(2H)-furanone. Microbiology. 2002;148:909–922. doi: 10.1099/00221287-148-4-909. [DOI] [PubMed] [Google Scholar]
  54. Winzer K, Hardie KR, Williams P. LuxS and autoinducer-2: Their contribution to quorum sensing and metabolism in bacteria. Adv Appl Microbiol. 2003;53:291–396. doi: 10.1016/s0065-2164(03)53009-x. [DOI] [PubMed] [Google Scholar]
  55. Xavier KB, Bassler BL. Regulation of uptake and processing of the quorum-sensing autoinducer AI-2 in Escherichia coli. J Bacteriol. 2005;187:238–248. doi: 10.1128/JB.187.1.238-248.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES