RESUMO
Background: Lower respiratory tract infection (LRTI) is a leading cause of infant morbidity and mortality globally. LRTI may be caused by viral or bacterial infections, individually or in combination. We investigated associations between LRTI and infant nasopharyngeal (NP) viruses and bacteria in a South African birth cohort. Methods: In a case-control study of infants enrolled in the Drakenstein Child Health Study (DCHS), LRTI cases were identified prospectively and age-matched with controls from the cohort. NP swabs were tested using quantitative real-time polymerase chain reaction (qPCR) and 16S rRNA gene amplicon sequencing. We calculated adjusted Conditional Odds Ratios (aORs) for qPCR targets and used mixed effects models to identify differentially abundant taxa between LRTI cases and controls and explore viral-bacterial interactions. Results: Respiratory Syncytial Virus (RSV) [aOR: 5.69, 95% CI: 3.03-10.69], human rhinovirus (HRV) [1.47, 1.03-2.09], parainfluenza virus [3.46, 1.64-7.26], adenovirus [1.99, 1.08-3.68], enterovirus [2.32, 1.20-4.46], Haemophilus influenzae [1.72, 1.25-2.37], Klebsiella pneumoniae [2.66, 1.59-4.46], or high-density (> 6.9 log10 copies/mL) Streptococcus pneumoniae [1.53, 1.01-2.32] were associated with LRTI. Using 16S sequencing, LRTI was associated with increased relative abundance of Haemophilus (q = 0.0003) and decreased relative abundance of Dolosigranulum (q = 0.001), Corynebacterium (q = 0.091) and Neisseria (q = 0.004). In samples positive for RSV, Staphylococcus and Alloprevotella were present at lower relative abundance in cases than controls. In samples positive for parainfluenza virus or HRV, Haemophilus was present at higher relative abundance in cases. Conclusions: The associations between bacterial taxa and LRTI are strikingly similar to those identified in high-income countries, suggesting a conserved phenotype. RSV was the major virus associated with LRTI. H. influenzae appears to be the major bacterial driver of LRTI, acting synergistically with viruses. The Gram-positive bacteria Dolosigranulum and Corynebacteria may protect against LRTI, while Staphylococcus was associated with reduced risk of RSV-related LRTI. Funding: National Institutes of Health of the USA, Bill and Melinda Gates Foundation, National Research Foundation South Africa, South African Medical Research Council, L'Oréal-UNESCO For Women in Science South Africa, Australian National Health and Medical Research Council.
RESUMO
Introduction: Short-read amplicon sequencing studies have typically focused on 1-2 variable regions of the 16S rRNA gene. Species-level resolution is limited in these studies, as each variable region enables the characterisation of a different subsection of the microbiome. Although long-read sequencing techniques take advantage of all 9 variable regions by sequencing the entire 16S rRNA gene, they are substantially more expensive. This work assessed the feasibility of accurate species-level resolution and reproducibility using a relatively new sequencing kit and bioinformatics pipeline developed for short-read sequencing of multiple variable regions of the 16S rRNA gene. In addition, we evaluated the potential impact of different sample collection methods on our outcomes. Methods: Using xGen™ 16S Amplicon Panel v2 kits, sequencing of all 9 variable regions of the 16S rRNA gene was carried out on an Illumina MiSeq platform. Mock cells and mock DNA for 8 bacterial species were included as extraction and sequencing controls respectively. Within-run and between-run replicate samples, and pairs of stool and rectal swabs collected at 0-5 weeks from the same participants, were incorporated. Observed relative abundances of each species were compared to theoretical abundances provided by ZymoBIOMICS. Paired Wilcoxon rank sum tests and distance-based intraclass correlation coefficients were used to statistically compare alpha and beta diversity measures, respectively, for pairs of replicates and stool/rectal swab sample pairs. Results: Using multiple variable regions of the 16S ribosomal Ribonucleic Acid (rRNA) gene, we found that we could accurately identify taxa to a species level and obtain highly reproducible results at a species level. Yet, the microbial profiles of stool and rectal swab sample pairs differed substantially despite being collected concurrently from the same infants. Conclusion: This protocol provides an effective means for studying infant gut microbial samples at a species level. However, sample collection approaches need to be accounted for in any downstream analysis.
RESUMO
Background: The molecular epidemiology of carbapenem-resistant Enterobacterales in Cape Town remains largely unknown. Objectives: This study aimed to describe the molecular epidemiology, resistome, virulome and mobilome of carbapenem-resistant Klebsiella pneumoniae (CRKP) within Cape Town to guide therapy, antimicrobial stewardship and infection prevention and control practices. Methods: Eighty-five CRKP isolates from hospitalized patients underwent WGS as part of a prospective, multicentre, cross-sectional study, conducted between 1 November 2020 and 30 November 2022, across public-sector and private-sector hospitals in Cape Town, South Africa. Results: MLST revealed three novel types, ST6785, ST6786 and ST6787, while the most common were ST219, ST307, ST17, ST13 and ST2497. Different predominant clones were noted in each hospital. The most common carbapenemase gene was blaOXA-48-like, detected in 71% of isolates, with blaNDM detected in 5%. Notably, co-detection of two carbapenemase genes (blaOXA-48-like and blaNDM) occurred in 13% of isolates. The yersiniabactin siderophore was detected in 73% of isolates, and was most commonly associated with the ICEKp5 mobile element. All carbapenemases were located on plasmids. The genes blaOXA-181 and blaOXA-232 colocalized with a ColKP3 replicon type on assembled contigs in 83% and 100% of cases, respectively. Conclusions: CRKP epidemiology in Cape Town reflects institutionally dominant, rather than regional, clones. The most prevalent carbapenemase gene was blaOXA-48-like, in keeping with CRKP epidemiology in South Africa in general. Emerging clones harbouring both blaOXA-48-like and blaNDM, such as ST17, ST2497 and the novel ST6787, are a concern due to the limited availability of appropriate antimicrobial agents in South Africa.