The Worldwide Spread of the Tiger Mosquito as Revealed by Mitogenome Haplogroup Diversity

Abstract

In the last 40 years, the Asian tiger mosquito Aedes albopictus, indigenous to East Asia, has colonized every continent except Antarctica. Its spread is a major public health concern, given that this species is a competent vector for numerous arboviruses, including those causing dengue, chikungunya, West Nile, and the recently emerged Zika fever. To acquire more information on the ancestral source(s) of adventive populations and the overall diffusion process from its native range, we analyzed the mitogenome variation of 27 individuals from representative populations of Asia, the Americas, and Europe. Phylogenetic analyses revealed five haplogroups in Asia, but population surveys appear to indicate that only three of these (A1a1, A1a2, and A1b) were involved in the recent worldwide spread. We also found out that a derived lineage (A1a1a1) within A1a1, which is now common in Italy, most likely arose in North America from an ancestral Japanese source. These different genetic sources now coexist in many of the recently colonized areas, thus probably creating novel genomic combinations which might be one of the causes of the apparently growing ability of A. albopictus to expand its geographical range.

Keywords: Aedes albopictus; haplogroups; mitochondrial DNA; mitogenomes; tiger mosquito.

FIGURE 1

Phylogeny of A. albopictus mitogenomes. The Bayesian (left) and MP (right) trees are shown in the top inset. The posterior probability for the major nodes in the Bayesian tree is reported in red, whereas bootstrap values (1000 replications) are reported in red in the MP tree. These trees encompass 25 novel and two previously published sequences (Table 1). A magnified MP tree is also shown reporting all mutations that characterize the 27 mitogenomes except those linking mitogenome #27 to the A1′2 node (see below). The asterisk (^∗) indicates the location of these mutations (347 in the coding region and 16 in control region), which are listed in Supplementary information for Figure 1. For the phylogeny construction, the entire coding region variation of all mitogenomes was included as well as some control region mutations (see “Material and Methods”). The published sequences (#17 and #27, Table 1) are from the Jiangsu Province, China (Zhang et al., 2015) and from Taipei, Taiwan, respectively. The mitogenome from a mosquito of the Italian Rimini strain (#1, marked by the arrow) was employed to number the mutations shown on the branches. Mutations are transitions unless a base is explicitly indicated for transversions (to A, G, C, or T) or a suffix for indels (0.1, d). Heteroplasmic positions are marked by an “h.” Recurrent mutations within the phylogeny are underlined (and in italics if present in mitogenome #27) and back mutations are marked with the suffix @. The numerous mutations shared only by the published mitogenomes #17 and #27 are marked with the suffix §. Taking also into account that, despite their extensive coding region differentiation, mitogenomes #17 and #27 intriguingly harbor virtually identical control region sequences, it is likely that at least some of the mutations marked with the suffix § are mistakes. Colors illustrate geographic origins. Table 1 provides additional information concerning the geographic origin and haplogroup affiliation of each sample. Length variation (insertions/deletions) in a poly-A stretch beginning at np 3808 was not considered. Note that for mitogenomes #9 and #13 (in squared boxes) the sequence variation was assessed only partially for the coding region and not at all for the control region (Supplementary information for Figure 1). The sub-haplogroup A1a1a1 and its derivatives of probable North American origin are encircled.

FIGURE 2

Organization of tandem repeats in the control region of A. albopictus mitogenomes. This schematic representation is based on the complete sequence information of mitogenomes #1 (Rimini strain) and #27 (Figure 1; Table 1). Two types of repeats (I and II) were observed with copy number differing for both repeats in the two mitogenomes. Type I consists of a ∼190 bp repeat unit whereas type II is made up of a shorter unit of ∼42 bp. The exact length (in bp) of each repeat is reported inside the box. The size of boxes representing repeats is not proportional to repeat lengths. Shown nucleotide positions are relative to sequence #1. Variation in size due to variable number of tandem repeats in the two regions, when available, is provided in Table 1. Black diamonds represent in the order from left to right: a poly-T stretch motif, a poly-T stretch motif followed by a GC-rich block, and a short poly-A stretch motif. All of these are conserved in different mosquito species (Dueñas et al., 2006).

FIGURE 3

Frequency (%) of A. albopictus mtDNA haplogroups in worldwide populations and possible diffusion routes. Arrows indicate possible diffusion routes from the native home-range (South-East Asia) and subsequent dispersals. Populations and corresponding frequency values are listed in Table 2. Note that the Chinese laboratory strain was not included. The original world map is from the website (http://www.freeworldmaps.net).