The microbiome of otitis media and development of a probiotic to prevent otitis media in Indigenous Australian children

Abstract

Background: Indigenous Australian children have endemic rates of otitis media (OM), impacting negatively on development, schooling and employment. Current attempts to prevent and treat OM are largely ineffective. Beneficial microbes are used successfully in a range of diseases and show promise in OM in non-Indigenous children. We aim to explore the role of beneficial microbes in OM in Indigenous Australian children.Aims: 1) Explore the knowledge gaps pertaining to upper respiratory tract (URT)/ middle ear microbiota (pathogens and commensals) in relation to OM in indigenous populations globally by systematic review of the literature.2) To explore the URT microbiota in Indigenous Australian children in relation to ear/ URT health and infection.3) To explore the ability of commensal bacteria found in the URT of Indigenous children to inhibit the growth of the main otopathogens.Methods: The systematic review of the PubMed database was performed according to PRISMA guidelines, including screening of articles meeting inclusion criteria by two independent reviewers.To explore the URT microbiota, we cross-sectionally recruited Indigenous Australian children from two diverse communities. Demographic and clinical data were obtained from parent/carer interview and the child’s medical record. Swabs were obtained from the nasal cavity, buccal mucosa and palatine tonsils and the ears, nose and throat were examined. Samples were analysed using a culturomics approach with MALDI-TOF isolate identification. Real-time PCR was used to qualify otopathogen loads and detect respiratory viruses. Culture-independent analysis of the nasal microbiota was examined using16S rRNA amplicon next generation sequencing (NGS).The bacterial interference of lactobacilli and alpha haemolytic streptococci (AHS) were investigated using agar overlay and cell-free supernatant. Promising isolates underwent whole genome sequencing (WGS) to investigate genetic markers of URT tropism, antibiotic resistance and virulence genes. In vitro antibiotic susceptibility was examined for ampicillin, amoxicillin + clavulanic acid, and azithromycin.Results: The systematic review included 25 papers encompassing Indigenous Australian, Alaskan, and Greenlandic children. It identified high rates of nasopharyngeal colonisation with the main otopathogens in indigenous children with OM. There was significant heterogeneity between studies, particularly in microbiological methods, which were largely limited to culture-based detection of the main otopathogens with an absence of data regarding commensal bacterial flora in the upper airway.We recruited 103 Indigenous Australian children aged 2-7 years (mean 4.7 years). Seventeen (16.5%) children were ‘healthy’ (normal examination and no history of OM). Investigation of nasal microbiota showed that children with a history of OM/ current OM/URT infection (URTI) had higher otopathogen detection and loads, and rhinovirus detection compared to healthy children (all p < 0.04). Investigation of network relationships revealed a strong correlation between high otopathogens loads in children with a history of OM/ current OM/URTI. Healthy children demonstrated a more complex network of correlated genera and a strong correlation between Corynebacterium pseudodiphtheriticum and Dolosigranulum pigrum. 16S NGS showed that Dolosigranulum was ubiquitous across all otitis groups but correlated with different genera in each group. Ornithobacterium was only detected with 16S NGS and was identified in children with current/ historical OM. It was absent/ at low relative abundance in the healthy children. Ornithobacterium was strongly correlated with Helcococcus, Dichelobacter and clustered around Streptococcus and Haemophilus.In relation to nose health, children with purulent rhinorrhoea had higher nasal otopathogen detection and loads, and rhinovirus detection compared to those with healthy noses (all p < 0.04). Children with healthy noses had a strong correlation between C. pseudodiphtheriticum and D. pigrum.Twenty-six lactobacilli isolates and 66 AHS isolates from 17 remote children were tested against otopathogens. Lactobacilli could readily inhibit the growth of otopathogens; three Lactobacillus rhamnosus isolates were more effective than commercially available strains, L. rhamnosus GG and L. rhamnosus LB21. AHS were less effective inhibitors, although some isolates were able to inhibit Streptococcus pneumoniae. Three lactobacilli progressed to WGS. One, L. rhamnosus (3160), had SpaCBA genes coding for pili to adhere to epithelial cells. We detected minor antibiotic resistance genes coding for antibiotic efflux pump and a ribosomal protection protein, neither associated with typical URT antimicrobials. The lactobacilli were susceptible to typical URT antimicrobials in vitro. Screening for virulence genes detected genes for two putative capsule proteins that have been described in bacteria from other genera.Conclusion: We have demonstrated the importance of bacterial relationships in the expression of URT health or disease. Poor ear/ URT health is related to strong correlation between high otopathogens loads, suggesting otopathogen synergism. Healthy children demonstrate a strong relationship between C. pseudodiphtheriticum and D. pigrum, which is not seen in other phenotypes, suggesting that C. pseudodiphtheriticum-D. pigrum synergism supports URT health. We detected Ornithobacterium, likely Candidatus Ornithobacterium hominis, and in this population was correlated with a novel bacterium which appears to be related to poor upper respiratory tract/ear health. We found lactobacilli that readily inhibited otopathogens with in silico and in vitro support a positive safety profile.