Phylogenetic Analysis of Transmission Dynamics of Dengue in Large and Small Population Centers, Northern Ecuador

Abstract

Although dengue is typically considered an urban disease, rural communities are also at high risk. To clarify dynamics of dengue virus (DENV) transmission in settings with characteristics generally considered rural (e.g., lower population density, remoteness), we conducted a phylogenetic analysis in 6 communities in northwestern Ecuador. DENV RNA was detected by PCR in 121/488 serum samples collected from febrile case-patients during 2019-2021. Phylogenetic analysis of 27 samples from Ecuador and other countries in South America confirmed that DENV-1 circulated during May 2019-March 2020 and DENV-2 circulated during December 2020-July 2021. Combining locality and isolation dates, we found strong evidence that DENV entered Ecuador through the northern province of Esmeraldas. Phylogenetic patterns suggest that, within this province, communities with larger populations and commercial centers were more often the source of DENV but that smaller, remote communities also play a role in regional transmission dynamics.

Keywords: Dengue; Ecuador; dengue virus; phylogenetics; remoteness; transmission; urban; vector-borne infections; viruses; whole-genome sequencing.

Figure 1

Locations of the 6 rural communities in Esmeraldas Province and the city of Esmeraldas for study of transmission dynamics of dengue in large and small population centers, northern Ecuador.

Figure 2

Number of febrile cases per month for DENV-1 (blue) and DENV-2 (orange) in study of transmission dynamics of dengue in large and small population centers, northern Ecuador, May 2019–December 2021. During April 2020–November 2020 (green), sampling was not conducted because of COVID-19 lockdown in Ecuador. DENV, dengue virus.

Figure 3

Maximum-clade credibility tree for dengue virus serotype 1 in study of transmission dynamics of dengue in large and small population centers, northern Ecuador. Tree was constructed using whole-genome sequences from Ecuador (blue), Colombia (green), Venezuela (purple), and Argentina (red) generated in BEAST software (https://beast.community). Within Ecuador, Esmeraldas Province samples are within the yellow rectangles and El Oro Province samples are within the orange rectangle, combined with E gene sequences from Norte de Santander department, Colombia. Posterior probabilities are shown in internal nodes. E, envelope.

Figure 4

Location of El Oro Provinces in Ecuador, Nariño and Putumayo departments in Colombia, and Venezuela in relationship to Esmeraldas Province in study of transmission dynamics of dengue in large and small population centers, northern Ecuador.

Figure 5

Maximum-clade credibility tree for dengue virus serotype 2 in study of transmission dynamics of dengue in large and small population centers, northern Ecuador. Tree was constructed using whole-genome sequences from Ecuador (blue), Colombia (green), and Venezuela (purple), combined with envelope gene sequences from Colombia departments (Nariño, Valle del Cauca, Putumayo, and Norte de Santander), and generated in BEAST software (https://beast.community). Within Ecuador, samples from Esmeraldas Province are within the yellow shaded area, and El Oro Province samples are within the orange shaded area. Posterior probabilities are shown in internal nodes.

Figure 6

Subclade tree of dengue virus serotype 1 from rural communities of Esmeraldas Province in study of transmission dynamics of dengue in large and small population centers, northern Ecuador. Gradient of remoteness is classified as remote communities with no road access (green), communities with road access (pink), commercial center (yellow), and urban (blue). Subclade nodes are labeled with posterior probabilities generated in BEAST software (https://beast.community).

Figure 7

Subclade tree of dengue virus serotype 2 from rural communities of Esmeraldas Province in study of transmission dynamics of dengue in large and small population centers, northern Ecuador. Gradient of remoteness is classified as remote communities with no road access (green), communities with road access (pink), commercial center (yellow), and urban (blue). Subclade nodes are labeled with posterior probabilities generated in BEAST software (https://beast.community).