Molecular Epidemiology of B3 and D8 Measles Viruses through Hemagglutinin Phylogenetic History

Abstract

Of the 24 known measles genotypes, only D8 and B3 are responsible for outbreaks in the last years in Europe, Asia, and America. In this study the H gene of 92 strains circulating between 2015 and 2019 in Lombardy, Northern Italy, and 1273 H sequences available in GenBank were analyzed in order to evaluate the genetic variability and to assess the conservation of the immunodominant sites. Overall, in Lombardy we observed the presence of four different B3 and three different D8 clusters, each one of them including sequences derived from viruses found in both vaccinated and unvaccinated subjects. Worldwide, the residue 400 within the H protein, a position located within the main immune epitope, is mutated in all circulating strains that belong to the two globally endemic genotypes, B3 and D8. Our data demonstrate the usefulness of measles virus (MV) H gene sequencing. Indeed, the monitoring the H protein epitopes of circulating strains could be included in the measles laboratory surveillance activities in order to improve and optimize strategies for measles control, as countries go towards elimination phase.

Keywords: H gene; genotype B3; genotype D8; hemagglutinin; immunodominant epitopes; measles surveillance strategies; measles virus.

Figure 1

Phylogenetic analysis of genotype B3 full H gene. The tree was obtained with the maximum-likelihood method [27], based on the Kimura 2 parameters model [28], identified as the best-fitting model after the model test analysis, using MEGA X [29]. A discrete Gamma distribution was used to model evolutionary rate differences among sites (+G = 0.2035) and branch support (1000 bootstrap iterations [30]) is provided next to nodes. Sequences from genotypes B1 (MVi/Younde.CAE/12.83) and B2 (MVi/Libreville.GAB/8.84) were used as outgroup. Branches corresponding to genotype (B3), sub-genotypes (B3.1 and B3.2), clade (400V), and clusters considered in this study (COMO, Unnamed, D/L/SD, NIGER) are indicated. Accession numbers of sequences used to generate this tree are available in Supplementary Figure S2.

Figure 2

Phylogenetic analysis of genotype B3 H nucleotide sequences belonging to clade B3-400V. The tree was obtained with the maximum-likelihood method [27], based on the Kimura 2 parameters model [28], identified as the best-fitting model after the model test analysis, using MEGA X [29]. A discrete Gamma distribution was used to model evolutionary rate differences among sites (+G = 0.1) and branch support (1000 bootstrap iterations [30]) is provided next to nodes. Other B3 sequences were used as outgroup. Clusters considered in this study are indicated on the right and sequences from Lombardy are indicated by a black circle for non-vaccinated individuals or red shape for vaccinees (triangle: one dose, circle: two doses). Empty circles correspond to partial sequences whose placement was deduced based on a different analysis (Figure S3). Amino acid mutations associated to main branches are indicated in green boxes, those associated to the investigated strains are indicated by black and red boxes for unvaccinated and vaccinated individuals, respectively, while substitutions causing a reversion to the original genotype are in blue.

Figure 3

Phylogenetic analysis of genotype D8 H gene. The tree was obtained with the maximum-likelihood method [27], based on the Tamura 3-parameter model [32], identified as the best-fitting model after the model test analysis, using MEGA X [29]. A discrete Gamma distribution was used to model evolutionary rate differences among sites (+G = 0.2684) and branch support (1000 bootstrap iterations [30]) is provided next to nodes. Sequences from genotypes D5 (Bangkok.THA/93/1) and D9 (Victoria.AUS/12.99) were used as outgroup. Branches corresponding to genotype (D8), clade (400T), sub-clades (A, B, C), and clusters considered in this study (Unnamed, GIR-SOMNATH, LONDON/OSAKA) are indicated. Accession numbers of sequences used to generate this tree are available in Figure S4.

Figure 4

Phylogenetic analysis of genotype D8 H nucleotide sequences belonging to clade D8-400T. The tree was obtained with the maximum-likelihood method [27], based on the Tamura 3-parameters model [32], identified as the best-fitting model after the model test analysis, using MEGA X [29]. A discrete Gamma distribution was used to model evolutionary rate differences among sites (+G = 0.2068) and branch support (1000 bootstrap iterations [30]) is provided next to nodes. Other D8 sequences were used as outgroup. Clusters considered in this study are indicated on the right and sequences from Lombardy are indicated by a black circle for non-vaccinated individuals or red shape for vaccinees (triangle: one dose, circle: two doses). Empty shapes correspond to partial sequences whose placement was deduced based on a different analysis (Figure S5). Amino acid mutations associated to main branches are indicated in green boxes while those associated to the investigated strains are indicated by black and red boxes for unvaccinated and vaccinated individuals, respectively.

Figure 5

Sequence conservation at the level of the H immune epitopes. The figure shows alignments of the sequence logos obtained for each immune epitope and for each genotype. Immune epitopes (SSE: sugar-shielded epitope; RBE: receptor-binding epitope; NE: neutralizing epitope; LE: loop epitope; HNE: hemagglutinating and noose epitope) as well as amino acid positions are indicated at the top, while genotypes with the relative number of available H sequences are indicated on the left.