Lung microbiome and antibiotic resistome of cystic fibrosis patients attending clinics in the Gauteng Province, South Africa

Abstract

Cystic fibrosis (CF) is a life-limiting hereditary disease characterised by recurrent polymicrobial lung infections and Pseudomonas aeruginosa lung colonisation is associated with increased morbidity and mortality. The life-long administration of antibiotics to treat lung infections leads to the accumulation of antibiotic resistance genes (ARGs) in CF pathogens.

Culture-dependent techniques used in diagnostics resulted in limited pathogens being associated with CF lung disease and antibiotic susceptibility testing does not typically characterise resistance mechanisms employed by the pathogens. This study aimed to use the culture-independent techniques:16S rRNA sequencing and conventional multiplex-polymerase chain reactions (M-PCR) assays for the characterisation of the lung microbiomes and antibiotic resistomes respectively, of the CF patients in Gauteng, South Africa with a special focus on P. aeruginosa. A comparison of the results found with 16S rRNA sequencing and the M-PCR assays was made with shotgun metagenomics, a more comprehensive but costly sequencing technique on selected specimens. Additionally, the genomes of selected P. aeruginosa isolates were analysed using whole genome sequencing (WGS). In total, 22 CF patients (7 to 46 years of age), comprised of 11 P. aeruginosa (PA) colonised and 11 non-P. aeruginosa (NPA) colonised patients provided consent for the collection of sputum specimens at two tertiary academic hospitals in Gauteng, South Africa.

More than 77 bacterial genera were found in the total (22) CF lung microbiomes which were dominated by traditional genera: Staphylococcus (32%), Pseudomonas (21.5%), Streptococcus (12%) and Haemophilus (5.1%). Notable genera such as Pseudomonas (43%) and Achromobacter (1%) were unique to PA microbiomes, while Burkholderia-Caballeronia-Paraburkholderia was unique to NPA microbiomes.

Clinically relevant mobilizable ARGs were found in the 22 CF resistomes where macrolide [ermB 95% (21/22), ermA 91% (20/22) and ermF 91% (20/22)], sulfonamide [sul2 64% (14/22)], aminoglycoside [ant(2”)-Ia 59% (13/22)], beta lactamase [blaTEM 59% (13/22)] and tetracycline [59% (13/22)] ARGs were prevalent. The ARGs: aph(3’)-IIIa-3 [27% (3/11)] and tetL [27% (3/11)] were exclusive to PA patients, while blaZ [91% (10/11)] was exclusive to NPA patients.

A comparison of 16S rRNA and shotgun sequencing on two randomly selected CF sputum specimens found that more bacterial genera were detected with 16S rRNA sequencing [85% (41/48)], than with shotgun sequencing [69% (33/48)]. However, limited agreement was found in the ARGs detected with M-PCR assays and shotgun sequencing.

Thirteen P. aeruginosa isolates analysed with WGS showed that unique strains mostly colonised the lungs of CF patients. Efflux pumps (MexAB-OprM and MexCD-OprJ) and acquired mutations in the crpP and gryA genes were important antibiotic resistance mechanisms detected in the isolates, while adaptive mutations detected in genes such as: lasR, mucA, pilA, pilB and pilC enhanced P. aeruginosa virulence and persistence in CF lungs.

The study CF lung microbiomes were diverse and harboured mobilisable, clinically relevant ARGs. The use of 16S rRNA sequencing proved superior to shotgun sequencing, however, M-PCR assays and shotgun sequencing gave contrasting results. Unique P. aeruginosa strains colonised patient lungs that carried clinically relevant ARGs, virulence genes and mutations. Future longitudinal studies analysing CF microbiomes and antibiotic resistomes and the expression of functional genes are required from the study settings.