In the last two decades, 16S rRNA amplicon sequencing has brought substantial insights into the composition of the human microbiome in different body districts, including the lung, and its role in health and disease [7]. Multiple studies in respiratory diseases have described that perturbation in microbiome balance can influence disease progression and severity, increasing inflammation and acute lung injury accompanied by exacerbations of symptoms [26].
In this study, we analysed the composition of the upper airway microbiome in a cohort of CVID patients. We studied the oropharyngeal niche as it has been shown to reflect colonisation of the lower airways [17–19], which are hardly accessible. Moreover, we also investigated the impact of CVID-related clinical variables on the microbiome composition. Our data showed a strong association between the oropharyngeal microbiome composition and CVID status. Adults with CVID showed a severely reduced diversity in the respiratory microbiome which resulted as enriched with potentially pathogenic bacteria. This observation was also confirmed by a machine-learned-based comparative metagenomics tool, a classification model recently validated in microbiome analysis [25]. Principal coordinates analysis of 16S rRNA sequencing data showed high dispersion in the CVID group without any apparent clustering, reflecting a higher intra-group variation in microbiome composition.
At taxonomic levels, in CVID we observed the relative expansion of the genus Streptococcus, with a decreasing abundance for the families Prevotellaceae, Fusobacteriaceae, Veillonellaceae, Campylobacteraceae, and Flavobacteriaceae. High levels of these latter families are generally correlated with a healthy status [6] and similarly, reduced abundance is associated with respiratory diseases such as pneumonia and COPD, and smoking [27, 28]. Moreover, depletion of Prevotella has been found to be linked to increased COPD severity and downregulation of genes promoting host defence [29].
In our cohort, the alpha diversity was severely reduced in CVID patients with undetectable serum IgA levels, who showed the expansion of the Haemophilus and Streptococcus genera. Previously, Berbers et al. found a significantly increased bacterial load of the oropharyngeal microbiome in CVID patients compared to healthy controls, with a relative enrichment of Prevotellaceae in those with serum IgA deficiency. However, they did not observe a dominance of the Haemophilus and Streptococcus genera in patients with low IgA levels [15]. The high relative abundance of Haemophilus and Streptococcus in our cohort is consistent with our previous identification by conventional culture methods of a high H. influenzae and S. pneumoniae mucosal carriage in CVID patients with low IgA serum levels. In these patients, carriage represents a bacterial reservoir, increasing the risk for respiratory infection exacerbations [16]. Here, we have observed that the presence of the pathobionts Streptococcus and Haemophilus in the airways of CVID patients is associated with the perturbation of microbiome with reduced diversity and differences in the relative abundance of the taxa compared to healthy individuals.
Bacterial dysbiosis and the consequent overgrowth of a few species have been associated with respiratory diseases since chronic inflammation facilitates the growth of selected species in the microbial community, which becomes resident rather than transient [30]. Low biomass of species belonging to the Streptococcus and Haemophilus genera are normally pathobionts colonising the oropharynx and lungs of healthy adults [31]. However, due to environmental factors or immunological perturbations, they can overgrow, disseminate, and ultimately cause infections [6, 32]. Oral microbes can also represent a source of respiratory infections in subjects with an impaired immune response, establishing a vicious cycle between oral dysbiosis and respiratory diseases [2, 33].
Under healthy conditions, secretory IgA protects the host by shaping the microbiome and neutralising toxins and viruses without causing inflammation by the inability to activate the complement cascade. Moreover, IgA blocks colonisation and penetration of pathogenic bacteria by binding receptors on the fimbriae, clearing unwanted particles, and promoting the sampling of antigens [8–10]. In this framework, it should be considered the impossibility of replacing IgA and IgM at the mucosal level in CVID patients [34, 35]. This emphasises the need for additional therapeutic interventions, such as using aerosolized IgA to prevent bacterial dysbiosis in the respiratory tract.
Due to their immune defect, CVID patients are at high risk for developing recurrent infections, particularly by encapsulated bacteria [36, 37], evolving towards lung damage [38], with a negative impact on quality of life and survival [36, 39]. Our data demonstrate that severe/moderate COPD status and age are associated with a less rich diversity in microbiome in CVID. The dominance of the genera Haemophilus and Streptococcus, with the addition of the genus Rothia was also observed in CVID with severe/moderate COPD compared to patients with mild or no COPD, in analogy with data from non-CVID cohorts with severe COPD [30]. Streptococcus expansion and its associated metabolites have been previously related to worsening lung function and exacerbations in COPD [40]. Moreover, Haemophilus dominance has been associated with airway neutrophil inflammation and disease severity in COPD and asthma [41, 42]. These data suggest that perturbation in the respiratory microbiome in CVID patients with COPD might contribute to lung disease progression and morbidity.
The main limitation of the study is the choice of the oropharynx as a sampling site. Oropharyngeal samples have been proven to sufficiently represent the structure of the lower airway microbial community [17–19]. However, due to its anatomical localization, this niche is enriched with the bacterial community present in the oral cavity. A further limitation is the lack of longitudinal sampling. Thus, the data provided warrants repeated investigation since the microbiome might change over time, making a single–time point evaluation interesting but not sufficient [43]. Lastly, given the heterogeneity of CVID, the limited number of individuals analysed could have limited the identification of a microbiome signature associated with CVID-related clinical variables.
In conclusion, these results demonstrate that the oropharyngeal niche of CVID patients has a distinct microbiome structure with a reduced diversity. The subset of CVID patients with COPD or undetectable IgA showed a microbiome ecosystem enriched with Streptococcus and Haemophilus, representing a possible source of infection and inflammation. These results pave the way to rethink the management of CVID, as we need to prevent lung damage and new approaches to replace IgA mucosal defect. Given the immunological interactions in the gut-lung axis [44], treatment with immunobiotics might also gain attention considering their potential to confer protection against infections by modulating innate and adaptive antimicrobial immunity [45, 46]. In this same line, as intestinal probiotics have been shown to reduce the number and duration of upper respiratory tract infections [47], their use has potential to contrast respiratory infections. As an alternative approach, the possibility of restoring microbial diversity, correcting dysbiosis, and limiting the abundance of pathobionts at the oro-pharyngeal level with the use of probiotics designed to target that specific niche deserves evaluations and future investigations.