Genetic characterization of a novel recombinant echovirus 30 strain causing a regional epidemic of aseptic meningitis in Hokkaido, Japan, 2017.

A regional epidemic of aseptic meningitis caused by echovirus 30 (E30) occurred in Hokkaido, Japan, during the period of August-December 2017. To investigate their phylogenetic relationship to other human enteroviruses, we determined the complete genomic nucleotide sequences of isolates from this outbreak. Phylogenetic analysis of the viral capsid protein 1 gene showed that the strains were most closely related to E30 strains detected in Germany, France, and Russia in 2013. In contrast, the region encoding the viral protease and the RNA-dependent RNA polymerase had a close phylogenetic relationship to non-E30 enteroviruses detected in the United Kingdom and Switzerland in 2015-2017, suggesting that a recombination event had occurred.

Echovirus 30 (E30) is one of the serotypes belonging to the species Enterovirus B [1] and is a major pathogen causing aseptic meningitis (AM) in both children and adults. Large epidemics caused by E30 have been frequently reported globally [2]. We experienced a regional epidemic of AM caused by E30 in Hokkaido, Japan, during the period of August-December 2017. Details of the epidemic and the clinical features were presented previously [3]. In this study, we determined the complete nucleotide sequences of viruses from the outbreak to determine their molecular features and their phylogenetic relationship to other strains of enterovirus B (EV-B). Virus isolation, RNA extraction, and sequencing were performed as described earlier [4]. Genome amplification and the nucleotide sequencing were carried out using specific primers (Supplementary Table S1). We determined the complete nucleotide sequences of four isolates, named E30/Hokkaido.JPN/21208 (Hokkaido21208), 21214, 21246, and 21326, that were collected on August 31, September 2 and 18, and November 24, 2017, and submitted them to the GenBank database under the accession numbers LC416533 to LC416536, respectively.

To investigate the genetic relationships of these isolates to globally circulating strains, phylogenetic analysis based on the viral capsid protein 1 (VP1) gene, the viral protease gene, and the RNA-dependent RNA polymerase (3CD) gene were carried out. Phylogenetic analysis based on the VP1 gene (876 bases) was performed in comparison with 248 E30 strains isolated in 33 countries during 1957-2017, and analysis based on the 3CD gene (1938 bases) was performed with 243 EV-B strains isolated in 33 countries during 1948-2017 (Supplementary Tables S2 and S3). With respect to choosing the strains, we took into account the genetic classifications of the strains that had been evaluated [2, 4–7]. Phylogenetic trees were constructed using Molecular Evolutionary Genetic Analysis software (MEGA v.5) [8], using the maximum-likelihood method based on the general time-reversible model. The genetic lineages were confirmed by bootstrap values using 1000 pseudoreplicates [9]. Furthermore, the nucleotide sequences were aligned with those of 243 strains (Supplementary Table S3) of EV-B using MEGA v.5. The gap-opening and extension penalties were set to 15 and 6.66, respectively, and the alignments were corrected manually to match the open reading frame (ORF). In order to identify potential recombination sites in the genome, a similarity plot analysis was performed using Simplot software version 3.5.1 [10] with a sliding window of 400 nucleotides moving in steps of 50 nucleotides.

The genomes of four E30 strains from Hokkaido consisted of 7428 nucleotides including a 5’ untranslated region (UTR) of 742 nucleotides followed by an ORF encoding a polypeptide of 2194 amino acids and a 3’ UTR of 101 nucleotides. The nucleotide and amino acid sequences of these strains were highly conserved (99.6-9% and 99.8-9% identity), suggesting that a single lineage of E30 had spread. The sequence of the 5’ UTR and the P1 genomic region of the strain Hokkaido21208 showed the highest identity to those of the strain 13-759, which was isolated in Germany in 2013 [11] (Table 1). While a high degree of nucleotide sequence similarity was also found in the 2A and the 2B genes, the sequence identity values from the 2C gene to the 3’ UTR were considerably lower (ranging from 71.2% to 84.0%). However, despite this divergence, the amino acid identity values from the P1 region to the P3 region ranged from 90.9% to 99.6% (Table 1).

Table 1

Nucleotide and amino acid sequence identity between strain Hokkaido21208 and strain 13-759

The result of similarity plot analysis correlated with the above results (Fig. 1). From the 5’ UTR to the 2B gene, the strain Hokkaido21208 was most closely related to the strain 13-759. However, the similarity of the 2C gene dropped abruptly to the same level as with other serotypes of EV-B. By contrast, the genomic region between the 2C gene and the 3’ UTR of the strain Hokkaido21208 did not show significant similarity to any of the strains compared, suggesting that a recombination event had occurred at the border between the 2B and 2C genes. This breakpoint has been previously identified as a recombination hot spot [12].

Fig. 1

Similarity plot analysis comparing strain Hokkaido21208 with EV-B strains. The nucleotide sequence of strain Hokkaido21208 was compared with those of 223 strains of EV-B (Supplementary Table S3), using Simplot software version 3.5.1 with a sliding window of 400 nucleotides moving in steps of 50 nucleotides. The strains described in this article are indicated by colored lines, and other strains are indicated by gray lines

A previous study indicated that the VP1 genes of E30 detected in Japan before 1998 were divided into six genogroups [5]. These genogroups correspond to the lineages E30_d, e, and g based on an advanced report by Bailly et al. [6]. In this study, our analysis indicated that the Hokkaido strains belong to the lineage E30_f (Fig. 2A), forming a novel cluster with Japanese strains detected in Fukushima prefecture, and with German, French, and Russian strains detected during 2008-2014. In contrast, several Japanese strains detected during 2008-2014 in Toyama, Osaka, Oita, and Okinawa prefectures, located mainly in western Japan, were classified in lineage E30_b (Fig. 2A). These data suggest that the Hokkaido strains might have originated overseas, in particular, Europe. E30 strains have been detected almost every year in Japan [13]. However, due to insufficient genetic information of these isolates, the detailed phylogenetic and topological relationships among Japanese strains and the time point of importation of lineage E30_f could not be clarified.

Fig. 2

Phylogenetic trees based on the VP1 genes of E30 strains and the 3CD genes of EV-B strains. (A) Phylogenetic tree based on the VP1 genes of E30 strains. (B) Phylogenetic tree based on the 3CD genes of EV-B strains. Bootstrap values above 50% are shown at the nodes. The scale bar represents nucleotide substitutions per site. Black circles indicate the Hokkaido strains. Black triangles indicate the strains highlighted in Fig. 1

Amino acid substitutions in VP1 of the Hokkaido strains were found in at least 20 positions in comparison with the strain 25-0284/osaka.AB908259/JPN/2013, which belongs to lineage E30_b (Table 2). Among these sites, 13 substitutions (56Y/F, 87E/D, 120I/V, 145V/I, 156R/K, 157S/G, 234N/S, 247K/R, 269V/I, 274E/D, 284S/T, 285T/N, and 289L/M) showed divergence from these lineages. Codons 87 and 156, which comprise a portion of the BC loop and the EF loop, respectively, have been predicted not only to be exposed on the surface but also to represent putative positive selection sites [6, 14]. Besides, several substitutions are in the C-terminal region of VP1 (amino acids 260-292), which has been recognized as one of the main antigenic regions of enteroviruses [15-17]. The emergence of an epidemic would implicitly indicate the lack of herd immunity in the community. Furthermore, the novel strains, which have different antigenic properties, even within the same serotype, could have escaped the host’s immune system. Earlier studies have demonstrated that E30 is antigenically heterogeneous and divided into three major groups [18, 19]. The VP1 nucleotide sequences of these groups are classified into the lineage E30_a or c, while the differences in antigenic properties do not correlate with the comprehensive classification [6]. Under these circumstances, it is important to explore whether the strains are pathogenic and able to spread beyond a regional scale in the near future. However, the strain-specific features that might be involved in disease are still not well understood [11, 20], and more-detailed study is needed to evaluate the characteristics of each lineage and to understand the potential endemic risk and the dynamics of virus transmission.

Table 2

Amino acid substitutions in VP1 of E30 strains detected in Japan after 2010

Strain	Lineage	Amino acid number
					1	1	1	1	1	2	2	2	2	2	2	2	2	2	2	2	2	2
		5	8	8	1	2	4	5	5	0	3	4	6	6	7	7	8	8	8	8	8	9
		6	4	7	0	0	5	6	7	4	4	7	3	9	4	7	4	5	7	8	9	0
Hokkaido strains/2017	f	Y	V	E	M	I	V	R	S	S	N	K	R	V	E	G	S	T	N	P	L	S
Se47/Fukushima/2014	f	.	T	.	.	.	.	.	.	.	.	.	.	.	.	.	.	.	.	.	.	.
Se91/Fukushima/2014	f	.	T	.	.	.	.	.	.	.	.	.	.	.	.	.	.	.	.	L	.	.
Se98/Fukushima/2014	f	.	T	.	.	.	.	.	.	.	.	.	.	.	.	.	.	.	.	.	.	L
25-0284/Osaka/2013	b	F	A	D	.	V	I	K	G	N	S	R	K	I	D	S	T	N	G	V	M	K
22-0338/Osaka/2010	b	F	A	D	.	V	I	K	G	N	S	R	K	I	D	S	T	N	S	A	M	M
489-NS/Toyama/2010	b	F	A	D	.	V	I	K	G	.	S	R	.	I	D	F	T	N	*	*	*	*
F-Aug-1/Toyama/2010	b	F	A	D	I	V	I	K	G	N	S	R	.	I	D	S	T	N	*	*	*	*

Dots (.) indicate amino acid residues identical to those in the Hokkaido strains

Asterisks (*) indicate that there was no sequencing information

The 3CD genes of EV-B are divided phylogenetically into three groups (1, 2, and 3) (Fig. 2B) [4, 7]. Our findings indicate that the Hokkaido strains belong to group 3 and form a cluster with coxsackievirus type B2 (CVB2), CVB3, CVB4, CVB5, E3, E6, E11, E13, enteroviruses 85, and B106 strains detected in European and Asian countries during the last 20 years. Among these strains, CVB4, CVB5, and E3 strains detected in the United Kingdom and Switzerland during 2015-2017 occupied the nearest branch (Fig. 2B). No other E30 strain except for the Hokkaido strains was found in this cluster. In contrast, the E30 strain 13-759, whose VP1 gene showed a high degree of similarity to those of the Hokkaido strains, was classified into group 2. Group 2 included several EV-B strains detected intermittently in Europe, suggesting that this genome has been maintained there since the previous century (Fig. 2B) [6]. In addition to the data for the VP1 genes described above, chronological and phylogenetic information about the 3CD genes strongly suggest that the Hokkaido E30 strains originated in Europe.

The emergence of E30 has been found to exhibit a cyclic pattern with dominance and disappearance of individual recombinants over a period of years [21-27]. The turnover period for such recombinants has been calculated to be from 1.8 to 5 years [28, 29]. Therefore, long-term surveillance based on detection of virus in clinical specimens and environmental materials would contribute to our understanding trends and epidemiological patterns [30-34]. Furthermore, investigation of the antigenic properties of circulating viruses is fundamental to understand their pathogenic features and, ultimately, to promote measures that are important for preventive public health.

Below is the link to the electronic supplementary material.

Supplementary material 1 (PDF 204 kb)

Supplementary material 2 (PDF 322 kb)

Supplementary material 3 (PDF 339 kb)