Multi-peaked adaptive landscape for chikungunya virus evolution predicts continued fitness optimization in Aedes albopictus mosquitoes

Abstract

Host species-specific fitness landscapes largely determine the outcome of host switching during pathogen emergence. Using chikungunya virus (CHIKV) to study adaptation to a mosquito vector, we evaluated mutations associated with recently evolved sub-lineages. Multiple Aedes albopictus-adaptive fitness peaks became available after CHIKV acquired an initial adaptive (E1-A226V) substitution, permitting rapid lineage diversification observed in nature. All second-step mutations involved replacements by glutamine or glutamic acid of E2 glycoprotein amino acids in the acid-sensitive region, providing a framework to anticipate additional A. albopictus-adaptive mutations. The combination of second-step adaptive mutations into a single, 'super-adaptive' fitness peak also predicted the future emergence of CHIKV strains with even greater transmission efficiency in some current regions of endemic circulation, followed by their likely global spread.

Figure 1. Phylogenetic relationships and spread of IOL CHIKV strains in the Indian Ocean region since 2004.

(a) Origin and reconstructed spread of sl1-3 and Kerala 2009 (sl4) strains. (b) Phylogenetic tree generated using complete genomic sequences of 91 representative IOL CHIKV strains constructed using the maximum likelihood method. Bootstrap values, sl1-4-specific and E1-A226V mutations are indicated. The SL07 strain used as a genetic background for all in vivo experiments is highlighted with an asterisk. The purple arrow highlights strain LK(EH)CH18608 with the E2-K233E substitution. The bar indicates 0.2% nucleotide sequence divergence.

Figure 2. Effect of sl1-3-specific mutations on CHIKV fitness following competition in A. albopictus.

(a) Schematic representation of the experimental design. A. albopictus mosquitoes from a Thai colony were presented with blood meals (BM) containing 10⁶p.f.u. ml⁻¹ of a 1:1 mixture of viruses (based on p.f.u.) expressing either sl1 (b,d), sl2 (c,e), sl3 (f,h) or sl3-B (g,i) signature mutations and wt (SL07-226) virus, and processed at 10 dpi. Graphs (b,c,f,g) show numbers and percentages of mosquitoes containing mutant (mut), wt or both viruses in individual mosquito heads (representing disseminated infections). For each virus pair, the difference in number of mosquitoes with disseminated infection of mutant versus wt was tested for significance with a one-tailed McNemar test. Gel images (d,e,h,i) show change in relative RNA levels for each virus competition pair. RNA was extracted from BM at 0 dpi or 4 pools of 10 mosquitoes per pool at 10 dpi followed by RT-PCR analysis and ApaI digestion to distinguish marked from unmarked competitors. The relative fitness (RF) was determined as a geometric mean of the ratio between mut:wt bands in the sample amplicon, divided by the starting ratio of mut:wt bands in the BM amplicon.

Figure 3. Effect of sl1, sl2 and sl3-B-specific mutations on CHIKV fitness in Vero and HFF cells.

(a) Schematic representation of the experimental design. Cells were infected at a multiplicity of 0.1 p.f.u. per cell in triplicate with a 1:1 mixture of viruses (based on p.f.u.) expressing either sl1- (b), sl2- (c) or sl3-B-specific (d) mutations (mut) and SL07-226V (wt) virus. At 1 and 2 dpi, supernatants were collected for RNA extraction and RT-PCR analysis. The relative fitness (RF) within a given competition was determined as the geometric mean of the ratio between mut and wt bands in the sample amplicon, divided by the starting ratio of mut to wt bands in the amplicon derived from the inoculum used for infection of cells.

Figure 4. Effect of sl1-, sl2- and sl3-B-specific mutations on CHIKV fitness in infant mice.

(a) Schematic representation of the experimental design. 2 to 3-day-old CD-1 mice were infected with 10² p.f.u. of a 1:1 mixture of viruses (based on p.f.u.) expressing sl1- (b), sl2- (c) or sl3-B-specific (d) mutations (mut) and SL07-226V (wt) virus. At 1 and 2 dpi, CHIKV RNA was extracted from the blood of individual mice and used for RT-PCR analysis. The relative fitness (RF) within a given competition was determined as the geometric mean of the ratio of mut to wt bands in amplicons derived from the mouse blood, divided by the starting ratio of mut and wt bands in the amplicon from the inoculum used for infection of mice.

Figure 5. Effect of individual or combined expression of sl1–specific mutations on CHIKV initial infection and replication in midguts of A. albopictus.

(a) Schematic representation of the experimental design. A 1:1 ratio (based on p.f.u.) of viruses that differ by the set of indicated substitutions (see tables on the left of each figure) was presented to a Thailand colony of A. albopictus at a final blood meals (BM) titre of 1 × 10⁶ p.f.u. ml⁻¹. At 1, 2 or 3 dpi, mosquito midguts were collected in pools of 10. RNA was extracted from each pool followed by RT-PCR analysis and ApaI digestion. (b–e) Tables show genetic differences between the competing viruses. The relative fitness (RF) for each time point was determined as a geometric mean of ratio of mut to wt bands in the amplicon derived from the sample, divided by the starting ratio of mut and wt bands in the BM amplicon.

Figure 6. Effect of individual or combined expression of sl3-B-specific mutations on CHIKV initial infection and replication in midguts of A. albopictus.

(a) Schematic representation of the experimental design. A 1:1 ratio (based on p.f.u.) of viruses that differ by the sets of indicated substitutions (see tables on the left of each figure) was presented to a Thailand colony of A. albopictus at final blood meals (BM) titre of 1 × 10⁶ p.f.u. ml⁻¹. At 1, 2 and 3 dpi, mosquito midguts were collected in pools of 10. RNA was extracted from each pool followed by RT-PCR analysis and ApaI digestion. (b–d) Tables show genetic differences between the competing viruses. The relative fitness (RF) for each time point was determined as the geometric mean of the ratio of mut to wt bands in amplicons derived from each sample, divided by the starting ratio of mut and wt bands in the BM amplicon.

Figure 7. Effect of mutations located on E2-210-252 axis on CHIKV fitness in A. albopictus.

(a) Atomic structure of the CHIKV E3 (grey)/E2 (blue)/E1 (yellow) spike complex constructed based on (PDB ID:3N44), demonstrating locations of introduced substitutions (pink spheres). Yellow and green spheres highlight positions of first- and second-step adaptive substitutions. (b) Schematic representation of the experimental design; A. albopictus mosquitoes from a Thai colony were presented with blood meals (BM) containing 10⁶ p.f.u. ml⁻¹ of a 1:1 mixture of wt (SL07-226V_Apa) and virus containing a single mutation located on E2-210-252 axis (E2-T213Q (c,g), E2-H232Q (d,h), E2-L248Q (e,i), E2-K254Q (f,j), E2-K233Q (k,o), E2-K233E (l,p), E2-K234E (m,q) and E2-K252E (n,r)), and processed at 7 dpi. Graphs (c–f,k–n) show numbers and percent of mosquitoes containing mutant, wt or containing both viruses in individual mosquito heads (representing disseminated infections) as determined from RT-PCR amplicons. For each virus pair the difference in number of mosquitoes with disseminated infection of mutant versus wt was tested for significance with a one-tailed McNemar test. Gel images (g–j,o–r) show change in relative RNA levels for each virus pair. RNA was extracted from BM at 0 dpi or 4 pools of 9–10 mosquitoes per pool at 7 dpi followed by RT-PCR analysis and ApaI digestion to distinguish marked (wt) from unmarked (mutant) competitors. The relative fitness (RF) was determined as a geometric mean of the ratio between mut:wt bands in an amplicon derived from the sample, divided by the starting ratio of mut:wt bands in the BM. ND, data not determined.

Figure 8. Effect of combined expression of E2-L210Q and E2-K252Q on CHIKV fitness in A. albopictus.

(a,b) above figures is a schematic representation of viruses used in the competition assay. Blood meals (BMs) containing 3 × 10⁵ p.f.u. ml⁻¹ of 1:1 mixture of viruses (DM and SL07-226V-210Q) (a,c) and (DM and SL07-226V-252Q) (b,d) were presented to A. albopictus. Disseminated CHIKV infection (a,b) and relative CHIKV RNA levels in mosquito bodies (c,d) were assayed at 10 dpi as described above. Graphs (a,b) show numbers and percent of mosquitoes containing DM, single mutant (SM; E2-L210Q or E2-K252Q), or containing both viruses in individual mosquito heads (representing disseminated infections). For each virus pair the difference in number of mosquitoes with disseminated infection of DM versus SM was tested for significance with a one-tailed McNemar test. Gel images (c,d) show change in relative RNA levels for each virus pair. RNA was extracted from BM at 0 dpi or 4 pools of 8 mosquitoes per pool at 10 dpi followed by RT-PCR analysis and ApaI digestion to distinguish marked from unmarked competitors in amplicon gels. The relative fitness (RF) was determined as a geometric mean of the ratio between DM: SM amplicon bands in the sample, divided by the starting ratio of DM: SM bands in the BM.

Figure 9. Hypothetical and empirical fitness landscapes for a multi-step process of host switching by CHIKV.

(a) The traditional hypothetical model^, postulates an initial decline in fitness upon a host change due to the non-overlapping fitness landscapes, followed by gradual adaptive evolution to a single peak in a recipient host, which also results in a fitness decline for infection of the donor host. (b) Empirical model of CHIKV adaptation to A. albopictus shows the lack of an initial fitness decline upon shifting from A. aegypti, and the existence of multiple, independent second-step adaptive peaks available after the initial adaptive E1-A226V substitution. Green indicates ancestral amino acids and red indicates derived. The Asian plateau represents the inability of Asian strains to reach the first-step adaptive peak owing to its dependence on the epistatic E1-A226V/E1-98T interaction (blue). Fitness for A. aegypti infection is not affected by any of the A. albopictus-adaptive substitutions.