NTHI strain 86-028NP is a minimally passaged clinical isolate obtained from Nationwide Children's Hospital which has been extensively characterized in the chinchilla model of OM and has been sequenced and annotated (Harrison et al., 2005, 2007). The parental NTHI strain 86-028NP::rpsL_A128G is a streptomycin resistant strain constructed as previously described (Carruthers et al., 2012). Construction of an unmarked, non-polar deletion mutant of the sapF gene was performed by the recombineering strategy as previously described (Tracy et al., 2008; Shelton et al., 2011). Briefly, primers 5′ TTCACTAAAAATCCACCAACCA 3′ and 5′ TGGACGTATAAAAATTGTAGATGG 3′ were used to amplify sapF and 1 kb of the flanking DNA both 5′ and 3′ to sapF. The subsequent amplicon was ligated into the pGEM-T Easy vector (pAV001) and transformed into E. coli strain DY380. In parallel, primers, 5′ AAAACAACCGCCACCCCTTTTATACTTAATTGCAAAGGAAATGAATAATGATTCCGGGGATCCGTCGACC 3′ and 5′ AAAGGGACGAGTGTTCATTTTTCCTTTCCTTAGTGAGTGTTTTTGTCTTTTGTAGGCTGGAGCTGCTTCG 3′, each containing 50 base pairs of DNA homologous to the 5′ and 3′ ends of the sapF gene were used to amplify the spec-rpsL cassette from pRSM2832. This amplicon was then electroporated into strain E. coli DY380/pAV001 to form strain DY380/pAV002, in which the sapF gene in pAV001, has been replaced by the cassette. The plasmid pAV002 was then used to transform NTHI 86-028NP::rpsL_A128G and transformants were selected by growth on spectinomycin-containing chocolate II agar plates. To generate a nonpolar deletion mutant, the sapF mutant was transformed with plasmid pRSM2947, grown at 32°C, and FLP expression was induced using anhydrotetracycline. The cells were cured of the plasmid by growth at 37°C.

Complementation of the sapF mutant was constructed as previously described (Mason et al., 2006). Briefly sapF was amplified from NTHI 86-028NP and ligated downstream the sap promoter into pSpec1 and transformed into Haemophilus strain Rd. A spectinomycin resistant clone was isolated and the plasmid stored as pSWS001. This plasmid was then transformed into NTHI 86-028NP::rpsL_A128GΔsapF and a spectinomycin resistant complemented strain was stored as NTHI 86-028NP::rpsL_A128GΔsapF/pSWS001.

Green fluorescent protein (GFP) reporter strains were engineered to monitor morphological plasticity phenotypes in the sapF deletion mutant strain and the parental strain NTHI 86-028NP::rpsL_A128G. An EcoRI/SphI restriction fragment was removed from pRSM2211 (Mason et al., 2003). The fragment contains the NTHI OMP P2 promoter cloned upstream the gfpmut3 gene. The P2:gfp fusion was cloned into the EcoR1/SphI site of pSpec1, constructed as previously described (Mason et al., 2006) and the plasmid saved as pGM1.1. This plasmid was then transformed into NTHI 86-028NP generating strain RSM2713. The plasmid (pGM1.1) was isolated from RSM2713 and used to transform parental strain NTHI 86-028NP::rpsL_A128G and the ΔsapF strain, generating strain NTHI 86-028NP::rpsL_A128G/pGM1.1 and NTHI 86-028NP::rpsL_A128GΔsapF/pGM1.1, respectively.

NTHI was grown on chocolate II agar (Becton Dickinson, Sparks, MD), in brain heart infusion broth supplemented with 2 μg heme/mL and 1 μg NAD/mL (supplemented brain-heart infusion; sBHI) or in defined iron-source (DIS) medium supplemented with 0, 2, or 20 μg heme/mL at 37°C in 5% CO₂, prepared as previously described (Mason et al., 2006). To monitor the requirement of the SapF ATPase on heme import and growth after heme starvation, the bacterial strains were grown overnight on chocolate agar at 37°C and 5% CO₂, sub-cultured into nitric acid-washed 15 mL round bottom glass tubes containing pre-warmed DIS medium with either 0 or 2 μg heme/mL and grown for 16 h to deplete internal stores of heme for use in subsequent experiments, as previously described (Mason et al., 2011). Heme-iron utilization was monitored in the iron-starved or non-starved cultures by measuring bacterial growth every hour for 16 h using the Synergy HT kinetic plate reader (BioTek, Bath, UK) as previously described (Mason et al., 2011). Number of viable bacteria was confirmed by serial dilution and plating to determine the colony forming units (CFU) per mL of culture.

Bactericidal assays were performed to evaluate AMP resistance as previously described (Mason et al., 2005). Briefly, bacteria were grown overnight on chocolate II agar plates, inoculated into 5 mL of pre-warmed sBHI medium, grown for 3 h at 37°C in 5% CO₂, and then diluted with sBHI to 10⁸ CFU/mL. The bacteria were further diluted 1:100 in 10 mM sodium phosphate (pH 7.4) containing 1% sBHI to 10⁶ CFU/mL. In a 96-well polypropylene plate (Falcon), 90 μL of 10 mM sodium phosphate containing 1% sBHI was added to each well. Human β-defensin (hBD-3, PeproTech) or human cathelicidin LL-37 (Phoenix Pharmaceuticals) was serially diluted in the wells so that each well retained 90 μL of the appropriate concentration of peptide. 10 μL bacteria (10⁴ CFU) were added to each well, and the plate was incubated for 1 hour at 37°C in 5% CO₂. Following incubation, bacterial concentrations were determined by serial dilution and growth on chocolate II agar, and data were expressed as CFU/mL of viable bacteria.

Starved and non-starved cultures were used to determine the influence of heme starvation and SapF deficiency on biofilm structure and organization. Starved and non-starved cultures were matched for CFU/mL in DIS medium and 2.5 × 10⁵ bacteria were added to each well of a Nunc Lab-Tek8-well chamber slide containing DIS medium supplemented with or without 2 μg heme/mL. The chamber slide was then incubated at 37°C with 5% CO₂ and biofilms were grown for 48 hours with a media change at 24 hours. The biofilms were stained with BacLite™ LIVE/DEAD stain (Molecular Probes, Grand Island, NY) for the strains not containing the pGM1.1 GFP reporter plasmid. Biofilm structure and organization were imaged with an Axiovert 200M inverted epifluorescence microscope equipped with the Apotome attachment for improved fluorescence resolution and an Axiocam MRM CCD camera (Carl Zeiss Inc., Thornwood, NY). Experiments were repeated in triplicate and representative samples are shown.

Healthy adult chinchillas (Chinchilla lanigera) purchased from Rauscher's Chinchilla ranch (LaRue, OH) with no evidence of middle ear infection were used to assess the biological impact of SapF function. Chinchillas were anesthetized with 2 mg xylazine/kg (Fort Dodge Animal Health, Fort Dodge, IA) and 10 mg ketamine/kg (Phoenix Scientific Inc., St. Joseph, MO) and 86-028NP::rpsL_A128G or 86-028NP::rpsL_A128GΔsapF were inoculated transbullarly into the middle ear (86-028NP::rpsL_A128G= 4.19 × 10³; 86-028NP::rpsL_A128G ΔsapF = 8.34 × 10³) of a chinchilla or via passive inhalation for nasopharyngeal colonization (86-028NP::rpsL_A128G = 1.34 × 10⁸; 86-028NP::rpsL_A128GΔsapF = 2.05 × 10⁸). At the indicated days post-inoculation, middle ear effusions were collected by epitympanic taps through the superior bullae, and directly obtained from the inferior bullae. Nasopharyngeal lavage fluids were obtained by passive inhalation of 500 μL of pyrogen free sterile saline into one naris and collection of the lavage through the other naris as the liquid was exhaled. Middle ear infection was monitored daily by video otoscopy for signs of inflammation and middle ear fluid accumulation. All animal experiments were performed using accredited conditions for animal welfare approved by the Institutional Animal Care and Use Committee (Welfare Assurance Number A3544-01) at The Research Institute at Nationwide Children's Hospital, AR08-00027.

We have previously shown that the SapA and SapD Sap transporter subunits are required for resistance to the cationic peptide (r)cBD-1 (the chinchilla homologue of human β-defensin-3, hBD-3; Mason et al., 2005, 2006). We further demonstrated that AMP transport to the bacterial cytoplasm, as a mechanism to decrease periplasmic accumulation of these toxic peptides, was dependent upon SapBC permease production (Shelton et al., 2011). We hypothesized that the two ATPase subunits, SapD and SapF, would energize transport of AMP molecules through the SapBC permease. To this end, we exposed the unmarked, nonpolar sapF mutant or the marked, non-polar sapD mutant (Mason et al., 2006) to increasing concentrations of the AMPs hBD-3 or LL-37 and monitored bacterial survival compared to that of the parental (NTHI 86-028NP::rpsL_A128G) or wild type (NTHI 86-028NP) strains, respectively. In marked contrast to the SapD ATPase, whereby the sapD mutant was susceptible to both hBD-3 and LL-37 (Figures 1B,D, respectively), we observed that the SapF ATPase was dispensable for conferring the resistant phenotype at all concentrations of LL-37 and at hBD-3 concentrations less than 2 μg/ml (microgram/ml) (Figures 1C,A, respectively). These data suggest that each ATPase may mediate unique functions with the Sap ABC transporter.

We previously demonstrated that transport of heme across the inner membrane requires the Sap transporter permease, SapBC, and the SapA periplasmic binding protein (Mason et al., 2011). For technical reasons associated with the requirement of SapD for potassium homeostasis, we have not been able to determine the contribution of SapD to heme-iron transport in NTHI. However, as SapF is not required for potassium homeostasis (data not shown), we determined whether SapF was required for the energy-dependent transport of heme. Thus, we predicted that loss of SapF would impair heme-iron utilization, measured by changes in bacterial growth, particularly under conditions of heme-iron starvation. We first monitored whether loss of SapF altered growth kinetics of NTHI compared to that of the parental strain when grown in DIS medium. We observed a delay in entrance into logarithmic growth in the sapF mutant compared to that of the parent strain (Figure 2A). However, the SapF-deficient strain maintained a similar growth rate compared to that of the parent strain throughout the remainder of the culture period, suggesting a compensatory mechanism for heme acquisition in the absence of SapF protein or a requirement of SapF for heme acquisition during the transition from stationary phase to logarithmic growth.

In order to monitor whether environmental heme-iron starvation would provoke more severe growth defects, the parent and sapF mutant strains were depleted of internal stores of heme-iron by incubation of each strain in DIS medium without heme for 16 h. Bacteria were matched for optical density and viable bacteria were enumerated by plating and overnight growth on agar (See “Material and Methods”; Figure 2B). We observed a statistically significant decrease (∼2-fold) in the number of colonies observed in the absence of SapF when compared to that of the parental strain (Figure 2B). The addition of sapF in trans restored the number of colonies produced by the sapF mutant (Figure 2B).

We next investigated whether SapF was required for recovery from heme-iron depletion of internal stores during planktonic growth (Figure 3). To this end, the parent strain, a SapF-deficient strain, or a SapF-complemented mutant strain was heme-iron starved for 16 h in DIS medium depleted of iron sources. The bacteria were then cultured in the presence of one of two concentrations of heme for 16 h (Figure 3). The heme-iron-starved parent strain failed to grow when cultured in heme-iron-depleted medium, which indicated that the microorganism was sufficiently starved of all internal iron stores (Figure 3A). Growth was restored to the parental strain when cultured in the presence of either concentration of heme-iron (Figures 3B,C). Although no growth defects were observed with the SapF-deficient strain during heme-iron starvation (data not shown), the bacteria demonstrated some modest growth but failed to regain logarithmic growth in the presence of heme-iron (Figures 3B,C). Complementation of the sapF mutation fully restored the growth of the heme-iron-starved microorganism compared to that of the parent strain (Figures 3B,C). These data indicated that SapF is critical for heme acquisition and utilization in NTHI.

A biofilm is a highly organized, multicellular network of bacteria encased in a matrix which results in a robust architecture. Iron availability has been shown to influence biofilm formation (Johnson et al., 2005, 2008; Hindre et al., 2008), and iron availability is limited in the human host. Given the observation that SapF contributes to the uptake and utilization of heme-iron, we investigated whether the loss of SapF ATPase likewise resulted in changes to NTHI biofilm architecture. Biofilm development was monitored for the sapF mutant strain or the parental strain when cultured in a heme-iron-replete medium (Figure 4). The parental strain forms a biofilm that primarily consists of a mat-like architecture with sporadic spiked short towers (Figures 4A,C). In marked contrast, loss of SapF resulted in more robust, thicker tower formation (Figures 4B,D). The differences in biofilm architecture were more readily observed when comparing the three-dimensional surface plot produced from a series of fluorescent optical sections (Figures 4C,D). The altered architecture in the absence of SapF suggests that environmental stresses (i.e., heme-iron depletion) may enhance the overall complexity of NTHI biofilms.

Since we demonstrated that SapF activity was essential for heme acquisition and utilization by NTHI, we monitored NTHI biofilm formation and architecture following heme-iron starvation in the presence or absence of heme for 24 h. Biofilms that develop following restoration of heme-iron to depleted cultures revealed phenotypic changes in the sapF mutant strain (Figure 5B) compared to the parent strain (Figure 5A). The biofilms were more punctate in appearance, reflective of bacterial aggregates that appeared to be primed for bacterial growth in the presence of heme. Complementation of the sapF mutation restored a smoother biofilm phenotype similar to that of the parental strain (compare Figures 5A,C). These alterations in biofilm architecture and bacterial morphology dependent upon SapF function were exaggerated under heme-iron starvation conditions (Figure 5E). Changes in bacterial morphology were also evident in the parental and complemented mutant strains when subjected to heme-iron starvation conditions. These data suggest that heme-iron starvation influences changes in NTHI biofilm architecture.

We became interested in investigating the punctate structures (Figure 5E) and overall architecture of the more robust tower structures (Figure 4D) formed in the absence of SapF. In contrast to the mat-like architecture of the parental strain (Figure 6A), microscopic evaluation at higher magnification revealed that the towers were formed from individuals that appear in an open lacework architecture (Figure 6B). In many cases the live dead stain revealed a chain of nucleoids that appear to be floating in the culture medium but securely attached to the biofilm proper. The apparently suspended nucleoid chain suggested that the NTHI may be forming bacterial chains or filaments. To better characterize the interbacterial interactions of the NTHI individuals within the biofilms, we transitioned to the use of an NTHI strain that constitutively produces GFP to visualize the intact cytoplasm for better differentiation of overall cell length. We thus sought to examine alterations in biofilm growth and architecture when both the parent and mutant strain were depleted of internal heme-iron stores. The environmentally depleted bacteria were placed into a biofilm chamber slide for 24 h of culture in the absence of environmental heme-iron (Figures 6C,D). Depletion of heme-iron resulted in the inhibition of cell division leading to the formation of elongated NTHI with a continuous cytoplasm as indicated by the uniform distribution of GFP (Figures 6C,D). These filamentous NTHI formed patches or clusters that were readily observed even at lower magnification (Figure 5E). This phenotypic change in morphology was accentuated in the sapF mutant strain as a majority of viable bacteria were filamentous compared to the parent strain (compare Figures 6C,D). These data demonstrate morphological plasticity occurs under conditions of heme-iron limitation. The morphological plasticity observed here can account for the decreases in CFU observed above (Figure 2B). In addition to the filamentous form being enriched at early stages in biofilm development in response to heme-iron limitation, filamentous NTHI were also observed throughout the towers formed in the absence of SapF (Figure 6B). Bacterial filamentation in response to environmental stressors has been shown to provide selective survival advantages to the bacterium (Justice et al., 2008; Horvath et al., 2011). Since the host microenvironment is initially iron deplete, the ability to filament may provide NTHI a survival advantage during colonization and biofilm formation.

Recently, multiple functions of the Sap transporter have been revealed through the use of strains deficient in the periplasmic binding protein, SapA (Mason et al., 2006, 2011; Shelton et al., 2011). We have demonstrated here that SapF was required for heme utilization, and coordination of NTHI biofilm architecture, traits that are important to establish acute OM. We have also previously shown that resistance to AMPs is essential for persistence of NTHI in vivo (Shelton et al., 2011). Therefore, we took advantage of our observation that SapF is dispensable for antimicrobial resistance to elucidate the contribution of heme-iron utilization toward colonization of the chinchilla nasopharynx and middle ear (Figure 7). We hypothesized that heme-iron utilization is a key trait required for survival in a host environment. In order to test this hypothesis we infected four chinchillas each with the parental NTHI strain or the sapF mutant strain and monitored bacterial survival in the nasopharynx and middle ear over a 14-day period. The absence of SapF dramatically attenuates persistence in the nasopharynx (Figures 7A,C) and middle ear (Figures 7B,D) as evidenced by the more rapid clearance of NTHI by day 14 (Chi square p = 0.04 and p = 0.05, respectively). The sapF mutant was cleared from the nasopharynx in three of the four animals on day 14 (Figures 7A,C). Moreover, the sapF mutant was cleared from the middle ear in five of eight ears on day 14 (Figures 7B,D). In contrast, the majority of the ears infected with the parental isolate had ongoing infections two weeks following bacterial inoculation. It is clear that a loss of SapF alters NTHI persistence and that SapF is required for normal long-term infection of the middle ear space, suggesting that heme-iron utilization is required for persistence in the host.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.