A Moraxella catarrhalis vaccine to protect against otitis media and exacerbations of COPD: An update on current progress and challenges

ABSTRACT

Moraxella catarrhalis is a major cause of morbidity and mortality worldwide, especially causing otitis media in young children and exacerbations of chronic obstructive pulmonary disease in adults. This pathogen uses several virulence mechanisms to colonize and survive in its host, including adherence and invasion of host cells, formation of polymicrobial biofilms with other bacterial pathogens, and production of β-lactamase. Given the global impact of otitis media and COPD, an effective vaccine to prevent M. catarrhalis infection would have a huge impact on the quality of life in both patient populations by preventing disease, thus reducing morbidity and health care costs. A number of promising vaccine antigens have been identified for M. catarrhalis. The development of improved animal models of M. catarrhalis disease and identification of a correlate of protection are needed to accelerate vaccine development. This review will discuss the current state of M. catarrhalis vaccine development, and the challenges that must be addressed to succeed.

Introduction

Moraxella catarrhalis (Mcat) is a human-restricted, respiratory tract pathogen that is a major cause of middle ear infections, or otitis media (OM), in young children and exacerbations in adults with chronic obstructive pulmonary disease (COPD). OM and COPD exacerbations cause substantial morbidity in both patient populations, and can lead to fatality in COPD patients.¹ There is also significant financial burden associated with infections in these 2 clinical settings, with doctor visits, emergency room visits, lost wages due to missed days at work, and cost of treatment.^2-6 Altogether, in the US, the estimated annual cost of OM is between $3-5 billion dollars, while the societal cost of COPD is estimated around $31,000 per patient annually.^7,8 Despite the immense morbidity and financial burden, there is currently not a vaccine available to prevent Mcat-associated diseases. Substantial recent progress has been made in identifying candidate vaccine antigens, making a review of the area particularly timely. Therefore, the purpose of this review is to discuss the importance of Mcat as a pathogen in OM and COPD, and discuss the current state of vaccine development to prevent Mcat diseases in vulnerable patient populations.

Mcat: The pathogen

First discovered in the late 19^th century, for a long time Mcat was thought to be a harmless commensal species related to the nonpathogenic Neisseria spp that are also present in the normal flora of the upper respiratory tract.^9,10 However, further investigation with more rigorous methods has revealed an important role for Mcat as a human respiratory tract pathogen. To this end, Mcat uses several virulence mechanisms, which facilitate colonization of the respiratory tract and opportunistic disease. Some of these key mechanisms are discussed as follows:

(1) Surface-associated molecules for adhesion and immune evasion: Adhesion molecules expressed on the bacterial cell surface, including UspA proteins, Hag/MID, and others, mediate Mcat binding to host surface receptors expressed on epithelial cells lining the respiratory tract.^11-14 This organism can also invade and survive within epithelial cells, thus evading immune detection and innate immune defenses. Many of these adhesion molecules are multifunctional. For example, UspA2 functions as an adhesin, and also binds terminal components of complement to evade complement-mediated killing via the classical pathway.^15-20

Similar to many other species of Gram-negative bacteria, Mcat produces outer membrane vesicles (OMVs) that act as a “decoy” to misdirect innate immune cells and other host immune components. Secretion of OMVs is a deliberate process that allows selectivity of OMV contents based on environmental cues and other factors involved in biofilm formation and nutrient acquisition.²¹ OMVs are derived from the bacterial cell envelope and are composed of molecules expressed on the outer membrane of the bacterium. Virulence factors, periplasmic proteins and other molecules targeted for export are also present in OMVs.^21,22

While also acting as an adhesin, the Hag/MID molecule has unique immunomodulatory properties that contribute to survival of Mcat in the respiratory tract.²³ Briefly, Hag/MID binds to mIgD expressed on naïve B cells in the nasopharyngeal associated lymphoid tissues (NALT). This interaction in concert with other signals stimulate the B cells to produce nonspecific, nonreactive IgM molecules.²⁴ However, this Mcat-induced B cell response is reversible through Th2 cell interaction, and Mcat-specific antibodies are produced.²⁵

(2) Antibiotic resistance: Antimicrobial resistance is a major worldwide crisis that is impacting the effectiveness of treatment and management of many infections caused by bacterial pathogens.²⁶ Misuse of antibiotics, including physicians over-prescribing and incomplete consumption of antimicrobial courses by patients, contribute to the acquisition and spread of antibiotic resistance genes in many pathogens, including Mcat. The first report describing β-lactam resistance in Mcat was published in 1976.^27,28 During this time, β-lactamase detection rates were relatively low, with ∼40 percent of clinical isolates producing β-lactamase. Based on structural studies and sequence analyses, it was hypothesized that Mcat acquired the bro β-lactamase gene from an unknown Gram-positive organism. Over the next decade, this antibiotic resistance gene spread at an alarming rate to virtually 100% of strains within 30 y of its discovery. Even more concerning is the growing body of evidence that indicates this β-lactamase, which is tethered to the membrane, is expressed on OMVs, and can act distally to provide passive protection to coinfecting species of bacteria, including Streptococcus pneumoniae and nontypeable Haemophilus influenzae (NTHi) that would otherwise be susceptible to treatment.^27,29-33

(3) Biofilm formation: Like many other species of pathogenic bacteria, Mcat forms biofilms within the host. Biofilm formation is an important virulence factor for bacterial survival, impacting the duration and severity of disease.^30,31,34,35 Within the biofilm, the bacteria are encased in a protective polymeric matrix, which conveys resistance to host immune defenses. Additionally, the bacteria in biofilms are in a reduced metabolic state and experience alternative gene expression profiles, contributing to enhanced resistance to antimicrobial killing.^36-38

Mcat typically colonizes the nasopharynx of a large proportion of infants and children; the rate of colonization decreases with age. Mcat shares an ecological niche in the human respiratory tract with pneumococcus and NTHi, which also cause OM and exacerbations of COPD. Studies of patient samples and experimental models of disease demonstrate that Mcat forms polymicrobial biofilms with both NTHi and pneumococcus within the host.^30,31,34 In fact, polymicrobial biofilm formation with Mcat enhances survival in vivo of both species in comparison to single species biofilms. As previously discussed, formation of multispecies biofilms has also been implicated as a contributing factor to antimicrobial treatment failure, especially in conjunction with β-lactamase producing bacteria like Mcat.

Rationale for a Mcat vaccine

In view of the role of Mcat as a prevalent and important pathogen in OM and COPD, these susceptible patient populations would benefit tremendously from a vaccine against this organism. An Mcat vaccine could prevent disease by reducing nasopharyngeal colonization, blocking migration to the middle ear and lower airways, and/or by inhibiting growth at these sites, without changing nasopharyngeal colonization. While reduction of nasopharyngeal colonization may facilitate prevention of disease, regulatory agencies will likely focus on prevention of disease for approval rather than prevention of nasopharyngeal colonization. Beyond preventing Mcat infections, vaccination may facilitate treatment of pneumococcus and NTHi infections by eliminating a potential copathogen that can provide passive protection from β-lactam antibiotic therapy and other mechanisms.

Clinical manifestations of Mcat infection

Otitis media

OM is a common childhood disease affecting millions young children worldwide. Approximately 50-80% of children in the US have at least one episode of OM by 1 y of age, with a peak in incidence of disease in children 3 y of age (80–90%).^39-42 Consequently, OM is associated with substantial morbidity, causing pain and discomfort from inflammation and pressure from fluid in the middle ear, developmental delays due to hearing impairment at a time that is critical for speech and language development, and a great deal of stress for parents caring for their afflicted children.^43,44 In persistent and recurrent cases, insertion of tubes under general anesthesia becomes necessary to relieve pressure, drain fluid, and prevent rupture of the tympanic membrane.^45,46

The 3 major bacterial pathogens that cause OM include pneumococcus, NTHi, and Mcat. Mcat causes 5 to 30% of cases of OM.^47-50 Estimates of the proportion of OM caused by Mcat vary widely between studies, due to differences in geographical location, sample collection, and bacterial detection methods used. However, Mcat tends to be severely underestimated in OM, the reasons for which will be discussed in more detail later. Mcat is typically 1) the first otopathogen to colonize the nasopharynx, 2) the first otopathogen to cause an episode of OM, and 3) has a rate of isolation in middle ear fluids from children with OM that remains relatively stable throughout early childhood.^50-52

Bacterial OM often follows a viral upper respiratory tract infection.^54-57 Inflammation of the Eustachian tube caused by the viral infection creates negative pressure behind the tympanic membrane. Eventually, this negative pressure induces insufflation of nasopharyngeal secretions containing OM pathogens, introducing them into the middle ear cavity. Therefore, nasopharyngeal colonization is a necessary prerequisite for development of OM. Preventing nasopharyngeal colonization through vaccination will eliminate the main risk factor for development of disease. Furthermore, vaccination against this organism could provide some level of protection from disease, similar to trends seen in pneumococcus following introduction of the conjugate vaccines.^58-61 Thus, development of an OM vaccine, targeting all 3 pathogens would be most advantageous for preventing disease in this vulnerable patient population. However, more research is needed to identify and assess a vaccine antigen(s) that would induce protection against Mcat in humans; this point will be discussed further in following sections.

Mcat is often underestimated as a cause of otitis media for 2 reasons:

1) The reduced metabolic state in biofilm bacteria renders culture-based methods of detection less reliable because these methods are designed to detect growth of planktonic bacteria. An important mechanism of pathogenesis in OM, especially in chronic OM, is the formation of biofilms in the middle ears. Often, biofilms have been implicated as the reason many cases of chronic or recurrent OM occur, contributing to the morbidity associated with this disease. Use of PCR-based identification methods has greatly improved detection of biofilm-associated bacteria, and further demonstrates the importance of Mcat as a pathogen in OM. Recent studies that use PCR-based detection methods reveal that Mcat is present in a far greater proportion of middle ear fluids that detected by culture alone.^49,50,62 Future studies of OM should include both culture and PCR-based methods to assess the etiology of OM.

2) Mcat is also overlooked because it appears to cause a clinically milder form of otitis media compared with the pneumococcus, and is more frequently isolated in polymicrobial infections with NTHi and pneumococcus.^52,63 Synergy between Mcat and other otopathogens enhances growth and virulence of NTHi and pneumococcus during these infections.^30,31,52,53 For example, Mcat confers β-lactam resistance to other otopathogens, which may be a contributing factor in treatment failures of OM.^30,31,64-67 Therefore, vaccination against Mcat may enhance the success rate of antimicrobial treatment in OM by eliminating or reducing the presence of an organism capable of conferring passive resistance to otherwise susceptible pathogens during coinfection.

COPD

COPD is a progressive inflammatory lung disease that causes enormous morbidity and mortality in adults worldwide. The course of COPD is marked by intermittent periods of worsening symptoms, called exacerbations, including increased sputum production, increased sputum purulence, and increased shortness of breath compared with baseline symptoms. Over time, these infections cause significant damage to the airways and alveolar tissues, resulting in progression of disease. In advanced stages, pulmonary function is severely impaired, leading to respiratory failure and death.

Mcat is the second leading bacterial cause of exacerbations in COPD patients, after NTHi.⁵ This organism causes approximately 2 to 4 million exacerbations of COPD in the US annually, accounting for approximately 10% of all exacerbations.^5,68

Current state of Mcat vaccine development

Although the potential impact of an Mcat vaccine for preventing disease in young children and adults with COPD is substantial, there are several key issues that hinder further progress; namely, lack of a known correlate(s) of protection and lack of an animal model that simulates Mcat infection in humans. Despite these obstacles, great strides have been made in identifying potential vaccine antigens, effective adjuvant formulations, and immunization routes that could be used in humans. In this section, the progress, challenges, and potential paths to success in Mcat vaccine development will be discussed.

Identifying protective antigens

Animal model systems

The most widely used model for testing potential Mcat vaccine antigens is the mouse pulmonary clearance model (Table 1).^69-72 Immunized and control mice are challenged with Mcat by introducing a high inoculum of bacteria into the lungs, through intratracheal inoculation, intranasal inoculation, or inhalation of aerosolized bacteria. Mice are allowed to clear some of the bacteria for several hours before the remaining bacterial load is determined in homogenized lungs. Accelerated clearance of bacteria from the lungs of antigen-immunized mice compared with vehicle controls is indicative of a protective immune response. A limitation of this model is that Mcat does not survive for long in the airways of challenged mice; a naïve mouse will typically clear most strains of Mcat within 24 hours. The inability to establish colonization in mice limits the window that immune responses against Mcat can be tested to within just a few hours. In spite of the limitations, this model is reliable, consistent, reproducible among research groups, and, therefore, generally accepted as the best model for testing Mcat vaccine antigens.

Table 1.

Potential Mcat vaccine antigens that have been tested in an animal model of infection.

Antigen	Function	Route	Adjuvant	Animal model	Functional antibodies	References
CopB	Iron acquisition	IV	None	MPCM	Bactericidal	85-89
CysP	SBP of ABC transporter, binds sulfate	SC	IFA	MPCM	—	90
dLOS	Adhesin, endotoxin	SC	Conjugated TT or HMP + Ribi-700	MPCM		89,91-101
		IP	Passive immunization w/immunized rabbit serum	MPCM
		IN	Conjugated CRM9 + CT	MPCM	Bactericidal
		SC	Conjugated CRM9 + Ribi-700	MPCM
		IP	Passive immunization with mAb	MPCM
Hag/MID	Adhesin, binds IgD	IP	CFA/IFA	MPCM	—	13,87,102-109
LbpB	Binds lactoferrin	—	—	—	—	110
M35	Porin	IPP/IT	None	MPCM	Opsonization	111-113
McaP	Autotransporter, adhesin	—	—	—	—	114,115
McmA	Autoagglutination, adhesin	—	—	—	—	116
MhaB1, MhaB2	Adhesin	SC	CFA/IFA	CNCM	—	73
Msp22	Surface lipoprotein, binds heme	IN	IC31	MPCM	—	69,70,108,117
		SC	IFA	MPCM
Msp75	Homology to succinic dehydrogenase	IN	CT	MPCM	—	69,117
OlpA	Complement resistance, binds factor H	—	—	—	—	118,119
OMP CD	Adhesin, binds mucin	SC	Alum, MPL + Alum, QS-21, conjugated to dLOS + Ribi-700	MPCM
		IPP/IT	IFA	MPCM	Neutralization	79,108,109,120-124
		IN	AdDP	MPCM
OMP E	Fatty acid transport	—	—	—	—	125,126
OMP G1a, OMP G1b	Lipoprotein, putative copper transport protein	—	—	—	—	127-129
OppA	SBP of ABC transporter, binds peptides	IN	CT	MPCM	—	108,130-132
PilA	Surface molecule	—	—	—	—	14,108,133
SBP2	SPB of ABC transporter, binds arginine	SC	IFA	MPCM	—	134,135
TbpB	Transferrin binding protein	SC	AlPO4 + MPL	MPCM	Bactericidal	87-89,109,136,137
UspAs	Adhesins, binds complement and ECM components	SC	QS-21, Conjugated to dLOS + Ribi-700	MPCM	Bactericidal	12,87-89,107,109,138-141

Abbreviations: SBP - substrate-binding protein, ABC - ATP-binding cassette, ECM - extracellular matrix, SC - subcutaneous, IP - intraperitoneal, IN - intranasal, IPP - intra-Peyer's patch, IT - intratracheal, IV - intraveneous, CFA - complete Freund's adjuvant, IFA - incomplete Freund's adjuvant, TT - tetanus toxoid, HMP - high molecular weight proteins, CRM - Cross-reactive mutant of diphtheria toxoid, CT - cholera toxoid, mAb - monoclonal antibodies, AdDP - adamantylamide dipeptide, MPL - monophosphoryl lipid A, MPCM - mouse pulmonary clearance model, CNCM - Chinchilla nasopharyngeal colonization model. Hyphen (-) indicates an area with no published results. Note: none of the proposed antigens have been assessed in clinical trials.

To a lesser extent, chinchillas have also been used to assess potential Mcat vaccine antigens.⁷³ In this model, Mcat is given intranasally by allowing anesthetized chinchillas to inhale the inoculum drop-wise. Mcat colonizes the nasopharynx of naïve chinchillas for about one week. To assess potential vaccine antigens, the nasopharyngeal tissues and or lavages are harvested and the remaining bacterial counts are determined 3 d post-challenge. Chinchillas have also been used as a model for OM. This model has been used successfully with NTHi and pneumococcus, but further development is needed for Mcat. In one version of this model, otopathogens are introduced into the middle ears through direct injection into the superior bullae through the thin layer of bone at the top of the skull.^74,75 In this model, Mcat survives for only about 24 hours in a single species infection. However, a more natural route of inoculation has been used that more accurately mimics middle ear infection in humans. A multispecies infection is established with other otopathogens or a respiratory virus through intranasal inoculation, similar to the nasopharyngeal colonization model.^30,76-78 Inflammation in the upper respiratory tract induces Eustachian tube dysfunction and the otopathogens then ascend into the chinchilla middle ears. Nonetheless, this model has several major disadvantages: 1) chinchillas are a limited resource, with only a few sources in the US that have these animals available for research laboratories, 2) OM only develops in a fraction of the infected animals following intranasal inoculation and 3) addition of other pathogens further complicates this model, which adds another source of variability that may obscure results.

In conclusion, development of animal models of disease has been difficult for Mcat because it is a human-restricted pathogen. The mouse pulmonary clearance model and the chinchilla model both have limitations for use in vaccine development for Mcat. Thus, the development of animal models that more appropriately simulate Mcat infection in humans would advance the field of vaccine development to prevent Mcat infections.

In vitro studies

While testing in an animal model of disease is important for identifying potential vaccine antigens, there are several potential Mcat antigens that have demonstrated promising characteristics through in vitro studies as well. The candidate vaccine antigens in Table 1 are conserved and expressed on the bacterial cell surface during infection, and are therefore accessible for antibody binding. Antibody-mediated clearance is hypothesized to be important in protection against Mcat. Hence, the majority of these studies involve assessing functional antibodies generated to the antigen of interest in bactericidal and opsonophagocytosis assays. Additionally, binding of neutralizing antibodies to antigens could block key binding sites and inhibit functions within the host, so antibody neutralization assays are also important. For example, antibodies to the potential vaccine antigen, OMP CD, inhibit binding to mucin.⁷⁹ Thus, potential vaccine antigens can be identified in vitro by demonstrating the generation of functional antibodies in immunized animals or in patient samples. To this end, in vitro functional assays have been used in place of an animal model of disease for vaccine development in Neisseria meningitidis, which highlights the potential usefulness of in vitro functional assays for Mcat vaccine development. However, very few of the candidate antigens in Table 1 have published results demonstrating antibody functional capabilities. These antigens were not all tested using the same assay parameters, which may account for the variability in results among these antigens. Thus, there is a need for standardized in vitro methods that can be applied to potential antigens under evaluation as vaccine candidates.

Antigens

Assessment of potential vaccine antigens has led to the identification of myriad potential molecules that could be used in a vaccine against Mcat (Table 1). These antigens share many characteristics that effective vaccine antigens possess: the antigen 1) is conserved among strains, 2) is expressed on the cell surface during infection, 3) is highly immunogenic, and 4) induces potentially protective responses in immunized animals.

The currently licensed pneumococcal conjugate vaccine that includes protein D of NTHi (Synflorix™) has demonstrated ∼30% efficacy in prevention of first episodes of OM caused by NTHi, establishing a proof of principle that a surface protein antigen of a non encapsulated Gram negative otopathogen induces protection against OM in humans. Including additional NTHi antigens to increase efficacy is a rational approach that is being pursued by several research groups. Based on this observation, it is likely that an effective vaccine to prevent OM caused by Mcat, also an unencapsulated Gram negative otopathogen, will require a multivalent vaccine as well. Based on a search of clinicaltrial.gov, no Mcat vaccine antigens are currently in clinical trials; thus, there is little information to guide the choice of optimal Mcat vaccine antigen at this time. We propose testing several candidate vaccine antigens that have different functions for Mcat in parallel. For example, a multivalent vaccine that contains a substrate binding protein, an adhesin and a human matrix binding protein (Table 1) would have the potential to induce protection by blocking multiple key steps in pathogenesis in addition to facilitating immune clearance through multiple targets. Given the substantial failure rate of vaccine antigens due to a variety of factors (formulation considerations, stability, toxicity, efficacy, etc) it will be important to assess multiple antigens in parallel.

Adjuvant

Selection of adjuvant is an important factor in vaccine design. Adjuvants enhance immune responses, and help direct specific types of immune responses. Since a correlate of protection for Mcat has not been identified, it is not known what specific responses are important for protection against Mcat. However, based on knowledge from similar extracellular bacteria, including NTHi and pneumococcus, Th2 and Th17 responses are thought to be important in mediating protection against Mcat because these responses enhance production of functional antibodies.^80-82 Thus, selecting adjuvants that skew responses toward Th2 and Th17, like aluminum salts (e.g. AlPO₄), may be more beneficial than adjuvants that skew heavily toward Th1 type responses, like AS01.⁸³

The adjuvants used to test potential Mcat vaccines in animal models are listed in Table 1. Interestingly, both Th1 and Th2 directed adjuvants have been used successfully with Mcat antigens in mice, suggesting that cellular immunity may play a large role in protective immunity against Mcat. Further enhancements to animal models of disease, in vitro studies, and more in depth studies into the mechanisms behind the immune responses to these potential vaccines will aid in identifying the correlate(s) of protection against this organism.

Routes of administration

Immunization route also plays a critical role in directing immune responses to a vaccine. Most of the vaccines approved for use in humans today are given via intramuscular injection. However, a large body of work suggests mucosal immunization shifts responses to vaccines to responses that favor protection at mucosal sites; an important observation that may inform future vaccination strategies, especially to mucosal pathogens like Mcat. To date, the only approved mucosal vaccine for human use is FluMist®, a quadravalent influenza vaccine that delivers live-attenuated virus in an intranasal mist. Based on studies performed since its introduction in the early 2000s, this vaccine induces protective mucosal immune responses, but does not induce strong systemic immune responses.⁸⁴ Based on immunization studies in animals, both systemic and mucosal immunization strategies have potential to induce protective immune responses against Mcat (Table 1). Therefore, development of a mucosal vaccine given intranasally has great potential for providing protective immunity to Mcat without needing to use a more painful route of administration (i.e. intramuscular injection). A major caveat of this strategy is that there are very few mucosal vaccines approved for use in humans. Without enough data to determine duration of immunity following mucosal immunization, developing a mucosal vaccine for Mcat may not be the best option for now.

Clinical and regulatory strategies for an Mcat vaccine

Altogether, vaccination against Mcat has the potential to have a profound impact on OM and COPD. An Mcat vaccine could prevent up to one-third of exacerbations of COPD. Given the underestimation of the role of Mcat in OM based on culture, and the role of Mcat as a co-pathogen, an effective Mcat vaccine may have a broader impact than is apparent based on the current literature. Given the importance of the pneumococcus, NTHi and Mcat in OM, a vaccine testing strategy must account for the reality that preventing OM by any one of the 3 pathogens alone will not have an optimal impact on preventing OM. Indeed, pneumococcal conjugate vaccines induce serotype specific protection against OM, but have not had a major impact on the overall burden of OM. Therefore, instead of targeting one or 2 individual pathogens in clinical trials, the ideal approach would be to design trials to target all 3 bacterial causes of OM. This will require innovative approaches to testing multivalent vaccines. With early exposure as a major risk factor for later development of OM, an all-inclusive OM vaccine may be most effective if given in infancy following the same schedule as the current childhood vaccines.

Concluding remarks and future directions

Mcat is an important pathogen that is severely underestimated in OM and COPD, but is gaining more appreciation as a pathogen as a result of more rigorously designed studies and the use of PCR-based detection methods. Several promising vaccine candidates are ready for further development. However, these antigens must first undergo formulation optimization, studies to validate safety and efficacy, and product made in accordance with current good manufacturing practice (cGMP) regulations before testing in clinical trials. Enhancing partnerships between academia, industry, and regulatory bodies would aid in moving promising antigens along the vaccine development pipeline. Ultimately, a vaccine that will prevent OM and exacerbations of COPD will need to induce protection from infection by the 3 major pathogens in these clinical settings: NTHi, S. pneumoniae, and Mcat.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.