Pathogen surveillance in the informal settlement, Kibera, Kenya, using a metagenomics approach

Abstract

Background: Worldwide, the number of emerging and re-emerging infectious diseases is increasing, highlighting the importance of global disease pathogen surveillance. Traditional population-based methods may fail to capture important events, particularly in settings with limited access to health care, such as urban informal settlements. In such environments, a mixture of surface water runoff and human feces containing pathogenic microorganisms could be used as a surveillance surrogate.

Method: We conducted a temporal metagenomic analysis of urban sewage from Kibera, an urban informal settlement in Nairobi, Kenya, to detect and quantify bacterial and associated antimicrobial resistance (AMR) determinants, viral and parasitic pathogens. Data were examined in conjunction with data from ongoing clinical infectious disease surveillance.

Results: A large variation of read abundances related to bacteria, viruses, and parasites of medical importance, as well as bacterial associated antimicrobial resistance genes over time were detected. Significant increased abundances were observed for a number of bacterial pathogens coinciding with higher abundances of AMR genes. Vibrio cholerae as well as rotavirus A, among other virus peaked in several weeks during the study period whereas Cryptosporidium spp. and Giardia spp, varied more over time.

Conclusion: The metagenomic surveillance approach for monitoring circulating pathogens in sewage was able to detect putative pathogen and resistance loads in an urban informal settlement. Thus, valuable if generated in real time to serve as a comprehensive infectious disease agent surveillance system with the potential to guide disease prevention and treatment. The approach may lead to a paradigm shift in conducting real-time global genomics-based surveillance in settings with limited access to health care.

Fig 1. Geographical overview and location and description of the residence clusters 9 and 10 of the urban slum city of Kibera, Nairobi, Kenya.

Sampling points marked with a red circle, brown lines indicate hill contours, dark green mark the PBIDS, black lines separate the residence clusters, light green mark the urban slum city. Photograph provided by the author Eric Ngeno.

Fig 2. Data from the PBIDS from the Kibera clusters 9 and 10 in the period from June 18, 2014 (week 25) to August 20, 2014 (week 34).

All cluster: black; cluster 9: red; cluster 10: green; X-axis is cases.

Fig 3. Relative read abundance (in RPM: Reads per million) of 27 human pathogens, (10 viral, 5 parasites, 12 bacterial) in sewage from Kibera.

Red: cluster 9; blue: cluster 10. The dotted horizontal lines show upper limits for each cluster. Note that viral data are shown on the logarithmic scale (log10). Note that scale is individual for each pathogen.

Fig 4. Heatmap showing changes in AMR abundance over time in clusters 9 and 10.

Relative abundance (FPKM) was calculated for AMR at drug class level. AMR classes (rows) are clustered according to co-abundance using complete linkage clustering of Euclidean distances. Data were mean-standardized (Z-scores) within each drug class, enabling within-class, cross-sample interpretation. Colors represent log (ln) transformed relative abundances (FPKM).