Tympanometric configurations and middle ear findings in experimental otitis media

Haemophilus responses to nutritional immunity: epigenetic and morphological contribution to biofilm architecture, invasion, persistence and disease severity

In an effort to suppress microbial outgrowth, the host sequesters essential nutrients in a process termed nutritional immunity. However, inflammatory responses to bacterial insult can restore nutritional resources. Given that nutrient availability modulates virulence factor production and biofilm formation by other bacterial species, we hypothesized that fluctuations in heme-iron availability, particularly at privileged sites, would similarly influence Haemophilus biofilm formation and pathogenesis. Thus, we cultured Haemophilus through sequential heme-iron deplete and heme-iron replete media to determine the effect of transient depletion of internal stores of heme-iron on multiple pathogenic phenotypes. We observed that prior heme-iron restriction potentiates biofilm changes for at least 72 hours that include increased peak height and architectural complexity as compared to biofilms initiated from heme-iron replete bacteria, suggesting a mechanism for epigenetic responses that participate in the changes observed. Additionally, in a co-infection model for human otitis media, heme-iron restricted Haemophilus, although accounting for only 10% of the inoculum (90% heme-iron replete), represented up to 99% of the organisms recovered at 4 days. These data indicate that fluctuations in heme-iron availability promote a survival advantage during disease. Filamentation mediated by a SulA-related ortholog was required for optimal biofilm peak height and persistence during experimental otitis media. Moreover, severity of disease in response to heme-iron restricted Haemophilus was reduced as evidenced by lack of mucosal destruction, decreased erythema, hemorrhagic foci and vasodilatation. Transient restriction of heme-iron also promoted productive invasion events leading to the development of intracellular bacterial communities. Taken together, these data suggest that nutritional immunity, may, in fact, foster long-term phenotypic changes that better equip bacteria for survival at infectious sites.

Clinical management of upper and lower respiratory tract diseases caused by nontypeable Haemophilus influenzae (NTHI) is a significant socioeconomic burden. Therapies targeting the pathogenic lifestyle of NTHI remain non-existent due to a lack of understanding of host microenvironmental cues and bacterial responses that dictate NTHI persistence. Iron availability influences bacterial virulence traits and biofilm formation; yet, host sequestration of iron serves to restrict bacterial growth. We predicted that fluctuations in availability of iron-containing compounds, typically associated with infection, would impact NTHI pathogenesis. We demonstrated that transient restriction of heme-iron triggered an epigenetic developmental program that enhanced NTHI biofilm architecture, directly influenced by induced morphological changes in bacterial length. Heme-iron restricted bacteria were primed for survival in the mammalian middle ear, due in part to an observed reduction in host inflammation coinciding with a striking reduction in host mucosal epithelial damage, compared to that observed in response to heme-iron replete NTHI. Moreover, transiently restricted NTHI were more invasive of epithelial cells resulting in formation of intracellular bacterial communities. Our findings significantly advance our understanding of how host immune pressure and nutrient availability influence pathogenic behaviors that impact disease severity.

Epitope mapping immunodominant regions of the PilA protein of nontypeable Haemophilus influenzae (NTHI) to facilitate the design of two novel chimeric vaccine candidates

We designed and tested three PilA-derived vaccine candidates in a chinchilla model of ascending nontypeable Haemophilus influenzae (NTHI)-induced otitis media (OM). Delivery of antiserum directed against each immunogen conferred varying degrees of protection. Presentation of a B-cell epitope derived from the OMP P5 adhesin at the N-terminus of recombinant soluble PilA protein (as opposed to the C-terminus), resulted in a protective chimeric immunogen that combined epitopes from two distinct NTHI adhesins (type IV pili and OMP P5). Incorporating protective epitopes derived from two NTHI adhesins/virulence determinants into a single pediatric vaccine candidate to prevent OM has multiple potential inherent advantages.

The fourth surface-exposed region of the outer membrane protein P5-homologous adhesin of nontypable Haemophilus influenzae is an immunodominant but nonprotective decoying epitope

Nontypable Haemophilus influenzae is a major cause of otitis media and other mucosal infections. After natural disease in children and experimental disease in chinchillas, we found a hierarchical pattern of immunodominance among the four surface-exposed regions of the P5-homologous adhesin, with the greatest response directed to region 4. However, Ab to region 4 is not protective. When this natural but biased response was refocused to region 3 by immunization, augmented bacterial clearance and protection from ascending otitis media was observed. Collectively, the data indicate that region 4 contains a highly immunodominant but nonprotective decoying epitope, the presence of which dampens the immune response to a subdominant but protective epitope in region 3.

Peptide and recombinant antigens for protection against bacterial middle ear infection

Passive immunization of chinchillas with serum specific for either LB1 or for LPD-LB1 (f)(2,1,3) prior to challenge with heterologous NTHI isolates (relative to diversity in region three of P5-fimbrin), significantly inhibited the signs and incidence of otitis media (P < or = 0.01) induced by any of the challenge isolates. The ability of these antisera to induce total eradication of NTHI from the nasopharynx was not however equivalent among challenged cohorts. The data thus suggested that while early, complete eradication of NTHI from the nasopharynx was highly protective, reduction of the bacterial load to below a critical threshold level appeared to be similarly effective. Both immunogens thus remain strong vaccine candidates.

Development of the middle ear in neonatal chinchillas. I. Birth to 14 days

Measures of middle-ear function in humans show large differences among neonates, infants, and adults. In contrast, hearing sensitivity is essentially mature at birth. Hypotheses that have been proposed to explain the developmental changes in middle-ear function include: (i) contaminating effects of the immature neonatal ear-canal wall and (ii) persistent fetal tissue in the ear canal, tympanic membrane (TM), and middle-ear space. To better understand the relationships between middle-ear function, hearing sensitivity and the structure of the middle ear, 30 chinchillas, aged 1-14 days, were studied. Middle-ear function was assessed by multifrequency tympanometry with probe tones ranging from 226 to 2,000 Hz. Hearing sensitivity was measured by auditory brainstem response using clicks and 1, 2, 4, and 8 kHz tone bursts. Structural characteristics were analyzed from temporal bone histologic preparations. At all frequencies, the acoustic admittance of the neonatal car is very low and tympanometric patterns are complex and irregular, compared to adult animals. The admittance is essentially constant from 1 to 14 days, indicating that developmental changes occur over a much wider age span than that investigated here. Hearing sensitivity of the chinchilla appears to be mature at birth. Histologic analysis indicated that there were no age-related changes in TM thickness, TM diameter, distance from TM to promontory, and stapes footplate diameter. There were small increases in bone thickness, middle-ear area, mastoid bulla area, and in the perimeters of the middle ear and mastoid bulla. There were no significant amounts of loose mesenchyme or other fetal tissue in the middle-ear space.

Variability of outer membrane protein P1 and its evaluation as a vaccine candidate against experimental otitis media due to nontypeable Haemophilus influenzae: an unambiguous, multifaceted approach

Candidate vaccine antigens for preventing otitis media caused by nontypeable Haemophilus influenzae (NTHI) should possess one or more conserved epitopes. We sought to evaluate the candidacy of P1, a surface-expressed outer membrane protein knowing that this antigen is subject to diversifying selection. Therefore, we selected NTHI strains from among >500 phylogenically variant isolates representative of the diversity found in natural populations of H. influenzae. Twenty-three variants of P1 (</=95% similarity) were identified among 42 strains. When chinchillas were immunized with recombinant P1 (rP1) obtained from one of these isolates (BCH-3), all animals developed antibodies specific for rP1. Immunized animals were protected against disease when challenged with BCH-3, but not with an ompP1 mutant of BCH-3 or a strain (BCH-2) possessing a heterologous P1 (91% identity). We conclude that (i) while P1 induces protection against NTHI-mediated otitis media, development of a polyvalent vaccine reflecting the variability of P1 would be necessary to construct an efficacious vaccine and (ii) use of a phylogenically characterized collection of representative isolates in concert with gene sequencing, cloning, gene inactivation, and animal testing offers an efficient, rational, and rigorous strategy for evaluating the potential problems associated with variability of vaccine targets and specificity of related immune responses.

Passive transfer of antiserum specific for immunogens derived from a nontypeable Haemophilus influenzae adhesin and lipoprotein D prevents otitis media after heterologous challenge

We recently determined that passive transfer of serum directed against a synthetic peptide called LB1 or a recombinant fusion protein immunogen [LPD-LB1(f)(2,1,3)] could prevent otitis media after challenge with a homologous nontypeable Haemophilus influenzae (NTHI) isolate. NTHI residing in the nasopharynx was rapidly cleared from this site, thus preventing it from ascending the eustachian tube and inducing otitis media in chinchillas compromised by an ongoing viral upper respiratory tract infection. While LB1 is based solely on one NTHI adhesin, the latter immunogen, LPD-LB1(f)(2,1,3), was designed to incorporate two NTHI antigens shown to play a role in the pathogenesis of otitis media; lipoprotein D (LPD) and the P5-homologous fimbrin adhesin. The design of LPD-LB1(f)(2,1,3) also accommodated for the recently demonstrated existence of three major groupings, based on amino acid sequence diversity, in the third surface-exposed region of P5-fimbrin. LPD-LB1(f)(2,1,3) was thus designed to potentially confer broader protection against challenge by diverse strains of NTHI. Chinchillas were passively immunized here with serum specific for either LB1 or for LPD-LB1(f)(2,1,3) prior to challenge with a member of all three groups of NTHI relative to diversity in region 3. The transferred serum pools were also analyzed for titer, specificity, and several functional activities. We found that both serum pools had equivalent ability to mediate C'-dependent killing and to inhibit adherence of NTHI strains to human oropharyngeal cells. When passively transferred, both serum pools significantly inhibited the signs and incidence of otitis media (P </= 0.01) induced by any of the three challenge isolates. Despite providing protection against disease, the ability of these antisera to induce total eradication of NTHI from the nasopharynx was not equivalent among NTHI groups. These data thus suggested that while early, complete eradication of NTHI from the nasopharynx was highly protective, reduction of the bacterial load to below a critical threshold level appeared to be similarly effective.

Protection against development of otitis media induced by nontypeable Haemophilus influenzae by both active and passive immunization in a chinchilla model of virus-bacterium superinfection

Three separate studies, two involving active-immunization regimens and one involving a passive-transfer protocol, were conducted to initially screen and ultimately more fully assess several nontypeable Haemophilus influenzae outer membrane proteins or their derivatives for their relative protective efficacy in chinchilla models of otitis media. Initial screening of these antigens (P5-fimbrin, lipoprotein D, and P6), delivered singly or in combination with either Freund's adjuvant or alum, indicated that augmented bacterial clearance from the nasopharynx, the middle ears, or both anatomical sites could be induced by parenteral immunization with P5-fimbrin combined with lipoprotein D, lipoprotein D alone, or the synthetic chimeric peptide LB1 (derived from P5-fimbrin), respectively. Data from a second study, wherein chinchillas were immunized with LB1 or lipoprotein D, each delivered with alum, again indicated that clearance of nontypeable H. influenzae could be augmented by immunization with either of these immunogens; however, when this adjuvant was used, both antibody titers in serum and efficacy were reduced. A third study was performed to investigate passive delivery of antisera directed against either LB1, lipoprotein D, nonacylated lipoprotein D, or a unique recombinant peptide designated LPD-LB1(f)2,1,3. The last three antiserum pools were generated by using the combined adjuvant of alum plus monophosphoryl lipid A. Passive transfer of sera specific for LB1 or LPD-LB1(f)2,1,3 to adenovirus-compromised chinchillas, prior to intranasal challenge with nontypeable H. influenzae, significantly reduced the severity of signs and incidence of otitis media which developed (P </= 0.001). Collectively, these data indicate the continued merit of further developing LB1 and LPD-LB1(f)2,1,3 as components of vaccines for otitis media.

Multifrequency tympanometry and histopathology in chinchillas with experimentally produced middle ear pathologies

Middle ear lesions were produced in chinchillas by introducing fascia, Gelfoam, silastic or a combination of these materials into the labyrinthine and mastoid bullae. After 1-2 months, conventional 226 Hz and multifrequency tympanometry (MFT) were performed and the animals were sacrificed for histopathological analysis of the middle ear. Specific middle ear lesions were created including abnormal tympanic membrane thickness, tympanic membrane mass, tympanic membrane adhesion, mastoid bulla obstruction, ossicular adhesions and reduced middle ear volume. Several potentially strong associations emerged between tympanometric measures and pathological conditions. The strongest were the combinations of thick tympanic membrane and irregular multifrequency tympanograms, tympanic membrane mass and notched 226 Hz admittance tympanograms, tympanic membrane adhesions and low 226 Hz admittance, and tympanic membrane adhesions and low resonant frequency. 226 Hz static admittance had a sensitivity of 0.73 and a specificity of 0.75 for detecting significant middle ear pathology. Tympanometric width was not an effective diagnostic test, separately or in combination with static admittance. The most effective test based on 226 Hz measures was the combination of low admittance or admittance notch. MFT was superior to 226 Hz tympanometry. The combination of low resonance or double resonance or irregular patterns had a sensitivity and specificity of 0.91 and 1.0, respectively. Combining 226 Hz tympanometry with MFT improved the performance relative to 226 Hz tympanometry alone but not relative to MFT alone. The results support clinical reports that multifrequency tympanometry detects some middle ear pathologies that are not detected by conventional 226 Hz tympanometry.

Multifrequency tympanometry in chinchillas

Multifrequency tympanometry (MFT), using probe frequencies ranging from 226-2,000 Hz, was performed on normal chinchillas to obtain normative data against which to compare results from animals with middle ear pathology. A series of validating experiments was conducted to determine the effects of anatomical alterations of the middle ear on MFT. These included artificially extending the ear canal, opening the bulla, injecting saline into the middle ear, and disrupting the ossicular chain. The results indicate that MFT characteristics of chinchilla ears are qualitatively similar to those observed in normal humans and patients with middle ear disease, and MFT provides information that is not available from the 226-Hz tympanogram.

Synergistic effect of adenovirus type 1 and nontypeable Haemophilus influenzae in a chinchilla model of experimental otitis media

We recently reported the development of a chinchilla model of experimental otitis media (OM) that uses a pediatric clinical isolate of adenovirus type 1 (4) and in which an active infection with the wild-type strain was demonstrated. To expand upon these findings, this study was designed to determine whether we could demonstrate adenovirus infection-induced predisposition to bacterial OM in the chinchilla, as has been shown in human epidemiological studies (D. A. Clements, F. W. Henderson, and E. C. Neebe, p. 27-29, in D. J. Lim, C. D. Bluestone, J. O. Klein, D. J. Nelson, and P. L. Ogra, ed., Proceedings of the Fifth International Symposium on Recent Advances in Otitis Media, 1993; F. W. Henderson, A. M. Collier, M. A. Sanyai, et al., N. Engl. J. Med. 306:1377-1383, 1982). In addition, we were interested in determining whether altering the order of pathogen acquisition would further affect the outcome of disease incidence and severity. Toward this end, cohorts of chinchillas were inoculated intranasally with a strain of nontypeable Haemophilus influenzae (NTHi) (86-028NP) which colonizes the chinchilla nasopharynx but does not consistently induce culture-positive OM when inoculated intranasally (L. O. Bakaletz, T. M. Hoepf, D. J. Lim, and B. Tallan, Abstr. 90th Annu. Meet. Am. Soc. Microbiol. 1990, abstr. B-66, p. 37, 1990), adenovirus type 1 and then inoculated 7 days later with NTHi, NTHi and then inoculated 7 days later with adenovirus type 1, or both pathogens concurrently. All cohorts were observed over a 35-day period and assessed for incidence and severity of OM by several methodologies. The data collectively indicated that all animals receiving both pathogens developed OM of greater severity than those receiving only a single agent. Adenovirus inoculation followed 7 days later by NTHi inoculation was the order of pathogen acquisition which induced the most prolonged presence of NTHi in both the nasopharynx and the middle ear, the most severe tympanic membrane inflammation overall, and the most significant damage to and altered function of both middle ear and eustachian tube mucosae.

Influenza A virus-induced otitis media and mucociliary dysfunction in the guinea pig

There is much epidemiologic, clinical and laboratory evidence that viral infection is involved in otitis media with effusion (OME). However, few studies have demonstrated any direct influence of viruses on the tubotympanum. The purpose of this study was to establish the effect of influenza A virus having invaded the tubotympanum and so elucidate the possible mechanism by which this virus contributes to the pathogenesis of OME. Eighty guinea pigs with normal otoscopic findings were inoculated with 0.2 ml suspension of influenza A (3.3 x 10(8) PFU/ml) into the tympanic cavity through the tympanic membrane. To serve as controls, the same number of guinea pigs were injected with 0.2 ml of physiologic saline solution into the tympanic cavity. At 3, 7, 14, and 28 days postinoculation, they were used for examination of the mucociliary function. Middle ear effusions as well as mucociliary dysfunction were observed only in the animals inoculated with the virus. The ciliary activity in the bulla was declined at any time examined. On the other hand, the ciliary activity in the Eustachian tube and the tympanic orifice was slightly lowered between 7 and 14 days, but the level was not different from that of the controls. However, the number of active ciliated cells (showing more than 500 beats/min) was significantly smaller than that of the controls. The mucociliary clearance time of the tubotympanum was more prolonged than that of the controls at 3, 7, and 14 days, and returned to the control level at 28 days. A variety of morphological changes were observed in the tubotympanum treated with the virus. Major pathologies observed included general inflammatory cell infiltration, vacuolation and other degeneration of ciliated cells, and vascular damage and increased vascular permeability. Regeneration of cilia or ciliated cells followed the degeneration, which included an increased number of basal cells and newly formed centrioles. However, the viral infection also affected the epithelial cells with new centrioles. Our study demonstrates that viral infection can evoke mucociliary dysfunction of the tubotympanum and create increased susceptibility to bacteria. Therefore, viral infection may enhance bacterial infectious processes in the tubotympanu thereby contributing to the occurrence of OME.

Experimental animal models for studying antimicrobial pharmacokinetics in otitis media

Antimicrobial treatment of otitis media, especially drug dosing considerations, is largely empiric, with few reported pharmacologic studies of drug distribution into the middle ear. A chinchilla animal model of serous and purulent otitis media has been used for some time to investigate mechanisms of disease pathogenesis. This model was adapted to investigate the penetration of amoxicillin, trimethoprim, and sulfamethoxazole into middle ear effusion. Purulent otitis media was produced by direct middle ear inoculation with type 7F Streptococcus pneumoniae. Serous otitis media was produced by eustachian tube obstruction using silastic sponge or Coeflex cement, but the Coeflex caused an undesirable local inflammatory response. The three antibiotics were administered to chinchillas with serous and purulent middle ear effusion. Plasma and ear fluid drug concentrations were measured by liquid chromatography and demonstrated the value of this model in assessing antibiotic penetration.

Eustachian tube histopathology during experimental influenza A virus infection in the chinchilla

The eustachian tubes of 29 influenza A virus-infected chinchillas were examined for histopathologic signs at intervals up to 21 days after inoculation to elucidate the pathologic basis of negative middle ear pressure, which occurs during viral respiratory tract infection in humans. In the animal model, eardrum inflammation and negative middle ear pressure mirrored epithelial damage in the eustachian tube and the accumulation of cellular and mucous debris in the tubal lumen. Epithelial damage was greatest in the proximal two thirds of the tube near the nasopharynx, whereas goblet cell metaplasia and increased secretory activity was greatest in the distal, tympanic one third of the tube. These results provide a morphologic correlate to the development of negative middle ear pressure, and perhaps explain the pathologic basis for purulent otitis media during viral respiratory tract infection.

Different virulence of influenza A virus strains and susceptibility to pneumococcal otitis media in chinchillas

We have previously shown that chinchillas infected with a multiply passaged laboratory strain of influenza A/NWS/33 (H1N1) develop negative middle-ear pressure; polymorphonuclear leukocyte oxidative, bactericidal, and chemotactic dysfunction; and increased susceptibility to pneumococcal otitis media. Because influenza A virus strains show different virulence in humans, three such strains were compared in the chinchilla model. Negative middle-ear pressure and tympanic membrane inflammation developed significantly more often in chinchillas infected with wild-type H3N2 virus than with either wild-type H1N1 virus or an attenuated, cold-adapted H3N2 vaccine strain, CR29. Marked depression in polymorphonuclear leukocyte chemiluminescent activity also developed significantly more often in H3N2 infected animals than in H1N1- or CR29-infected animals. Intranasal challenge of influenza virus-infected animals with type 7 Streptococcus pneumoniae resulted in a significantly greater occurrence of pneumococcal otitis media in H3N2-infected animals than in H1N1-, CR29-, or non-influenza-infected control animals. Clearance of pneumococci from nasal washings of animals infected with wild-type H3N2 was significantly delayed in comparison with the other groups. Thus, the previously demonstrated increased susceptibility to otitis media among children infected with H3N2 influenza virus may relate to the capacity of this strain to induce negative middle-ear pressure, polymorphonuclear leukocyte dysfunction, and alteration in the mucosal clearance of pneumococci.