Genomic Epidemiology of Vibrio cholerae O1 Associated with Floods, Pakistan, 2010

Muhammad Ali Shah1, Ankur Mutreja1, Nicholas Thomson1, Stephen Baker, Julian Parkhill, Gordon Dougan, Habib Bokhari, and Brendan W. Wren

Author affiliations: COMSATS Institute of Information Technology, Islamabad, Pakistan (M.A. Shah, H. Bokhari); Wellcome Trust Sanger Institute, Cambridge, UK (A. Mutreja, N. Thomson, J. Parkhill, G. Dougan); Oxford University Clinical Research Unit, Ho Chi Mihn City, Vietnam (S. Baker); London School of Hygiene and Tropical Medicine, London, UK (M.A. Shah, S. Baker, B.W. Wren)

Abstract

In August 2010, Pakistan experienced major floods and a subsequent cholera epidemic. To clarify the population dynamics and transmission of Vibrio cholerae in Pakistan, we sequenced the genomes of all V. cholerae O1 El Tor isolates and compared the sequences to a global collection of 146 V. cholerae strains. Within the global phylogeny, all isolates from Pakistan formed 2 new subclades (PSC-1 and PSC-2), lying in the third transmission wave of the seventh-pandemic lineage that could be distinguished by signature deletions and their antimicrobial susceptibilities. Geographically, PSC-1 isolates originated from the coast, whereas PSC-2 isolates originated from inland areas flooded by the Indus River. Single-nucleotide polymorphism accumulation analysis correlated river flow direction with the spread of PSC-2. We found at least 2 sources of cholera in Pakistan during the 2010 epidemic and illustrate the value of a global genomic data bank in contextualizing cholera outbreaks.

In 2010, a surge in cholera cases seriously threatened public health across Pakistan, where previously sporadic cases of cholera had been reported (1–3). In late July and August 2010, record monsoon rainfall and the simultaneous glacier melt resulted in the worst flooding in the recorded history of Pakistan, affecting an area of 61,776 square miles and displacing >20 million persons (4). A cholera outbreak ensued, and the World Health Organization (WHO) reported 164 laboratory-confirmed cases with the help of National Institute of Health and other allied departments in Pakistan (5).

Despite the paucity of data on the impact of cholera in Pakistan before 2010, seasonal epidemics are known to have occurred every year since then. Cholera is endemic to South Asia (6) and the Bay of Bengal (7,8) and is spread through contaminated food and water, often after civil unrest or natural disasters. Pakistan is particularly at risk for waterborne disease because it is an agricultural economy with one of the most expansive water distribution systems in the world. This vast irrigation system depends largely on the Indus River, which originates on the northern slopes of the Kailash mountain range in India and runs north to south through the entire length of Pakistan with many tributaries, including the Zaskar, the Shyok, the Nubra, and the Hunza, converging in the northern region and flowing through the provinces of Ladakh, Baltistan, and Gilgit.

Not all Vibrio cholerae strains cause major disease outbreaks. Although V. cholerae has >200 serogroups, only serogroups O1 and O139 are associated with epidemics. Serogroup O1 isolates can be assigned to 2 biotypes, classical and El Tor; the latter is responsible for the current seventh pandemic that has spread in global radiations, or waves, originating in the Bay of Bengal (8). The clinical severity of cholera is associated with the production of cholera enterotoxin (CT), which is encoded by a gene on the 6.9-kb CTX prophage integrated within chromosome 1 of all pandemic V. cholerae O1 cholera strains (9). Historically, CTX prophages have been categorized as CTXclassical or CTXEl Tor on the basis of DNA sequence of the rstR and the sequence variation in ctxB gene. During the last 2 decades, new variants of El Tor biotypes have emerged and have been used to differentiate V. cholerae isolates (10). However, such approaches do not have the resolution required to stratify the highly clonal V. cholerae O1 isolates of the seventh pandemic sufficiently to understand their precise phylogeny and relate that to geographic distribution and spread.

The application of whole-genome sequence analyses has revolutionized our ability to resolve the V. cholerae O1 El Tor populations and more precisely determine the patterns of spread of cholera within the worst affected countries of the world. Clarifying the routes of spread of cholera in Pakistan provides the unprecedented opportunity to inform public health provision. This study showed that the 2010 cholera outbreak was, in fact, an epidemic within an epidemic explained by 2 independent introductions of cholera in the country, 1 from the south and 1 following the flood water as it moved from north to south along the Indus River.