Genotyping of Clinical Varicella-Zoster Virus Isolates Collected in China

Abstract

Varicella-zoster virus (VZV) is genetically stable; and various schemes for the genotyping of VZV based on restriction fragment length polymorphisms (RFLPs), PCR, and sequencing have been developed. At least three major genotypes have been recognized among VZV isolates or clinical samples from different locations around the world; however, few data were available for viral isolates from China. In the current study, a collection of 19 VZV isolates from patients with zoster or varicella in the middle eastern region of China was examined for genetic variations. RFLP analysis of DNA fragments of open reading frames (ORFs) 38, 54, and 62 showed that all 19 isolates were PstI and BglI positive and SmaI negative, and this may represent the major restriction pattern of wild-type VZV strains in China. Further analysis of the R5 variable-repeat region in those strains revealed that 9 (47.4%) were type R5A, while the remaining 10 strains (52.6%) were type R5B. On the basis of the sequencing data for ORFs 1, 21, 22, and 54, all 19 Chinese strains could be grouped into genotype J or J1. A novel in-frame 3-nucleotide insertion (CGG) in ORF1 was found in 4 (21%) of the 19 isolates. Additionally three new nucleotide substitutions were detected in two of the isolates. A varicella isolate from the United States, strain MLS, was included in this study as a control for American wild-type VZV, and was found to be type M1, which represents one of the minor genotypes in North America.

Varicella-zoster virus (VZV) is a member of the subfamily Alphaherpesvirinae within the family of Herpesviridae. Primary infection with VZV causes chickenpox (varicella), which usually occurs during childhood. During primary infection, a latent infection is established in the sensory ganglia, such as those of the dorsal roots and trigeminal nerves. The reactivation of VZV results in herpes zoster (shingles). The age at the time of VZV infection is often related to disease severity, with the greatest risk occurring among very young and very old individuals; immunocompromised persons are also at elevated risk for severe disease (4).

The VZV genome is highly conserved and consists of about 125 kb of linear, double-stranded DNA with about 70 open reading frames (ORFs). The interstrain genomic variation in the VZV genome is limited to about 0.1% and consists almost entirely of single-nucleotide changes dispersed evenly across the genome. On the basis of the analysis of single nucleotide polymorphisms (SNPs), at least three or four geographically distinct genotypes have been described (1, 13, 23).

Several VZV genotyping schemes have been developed, and each proposes a different classification for genotyping. By analysis of selected SNPs in ORFs 1, 21, 50, and 54, four genotypes have been identified: A (Africa/Asia), B and C (Europe, North America), and J (Japanese). Genotype B may represent a recombinant between genotypes A and C (1). Loparev et al. reported that by the analysis of a short region (447 bp) in ORF22, three major genotypes can be distinguished: E (European), J (Japanese), and M (mosaic) (13). Genotype J strains were the most common in Japan, and genotype E strains were the most common in temperate latitudes. The genotype M strains were subdivided into M1, M2, M3, and M4 (12, 13, 16, 22). Another proposed typing approach is examination of the allele distribution in the R5 variable region between ORF60 and ORF61 among VZV isolates (6, 7).

Few VZV isolates or samples from China have been genotyped, despite the enormous population of the country. Loparev et al. (13) analyzed three clinical samples from patients with varicella in south China, and all were type M2. The same study also included three strains from north China, and all were found to be type E. We now present the genotyping data for 19 isolates from patients with zoster or varicella in the middle eastern region of China; all strains were identified as genotype J or J1 by previously published genotyping methods (1, 13), with a few isolates having genetic variations and sequences different from those of isolates described previously.

MATERIALS AND METHODS

Patients and clinical samples.

Nineteen VZV isolates were obtained from 17 cases of zoster and 2 cases of varicella (Table 1). All of the patients (12 males and 7 females) were referred to the Dermatology Clinic of the First Affiliated Hospital of Anhui Medical University, Hefei, China, between July 2007 and January 2008. Hefei is the capital of Anhui Province, which is located in the middle eastern part of China at a latitude of 35°N. Vesicle fluid was collected from skin lesions, placed in viral transport medium, and transported to the laboratory within 4 h on wet ice. In all cases, virus was recovered by inoculation of fluid from a single vesicle on a human embryonic lung fibroblast (HELF) monolayer and was confirmed by immunofluorescence with an anti-VZV gE monoclonal antibody (Biodesign International).

TABLE 1.

Characteristics of patients in the study recruited in the Dermatology Clinic, Hefei, China^a

Isolate no.	Sex	Age (yr)	Varicella or zoster	Patient type	History of varicella
1	M	50	Z	A	Childhood in China
2	M	45	Z	A	Not known
3	M	70	Z	H	Childhood in China
4	M	64	Z	A	Childhood in China
5	M	34	Z	A	Not known
6	M	67	Z	A	Not known
7	M	35	Z	A	Childhood in China
8	M	30	Z	A	Childhood in China
9	F	67	Z	A	Childhood in China
10	M	71	Z	A	Not known
11	M	74	Z	H	Childhood in China
12	F	80	Z	H	Not known
13	F	40	Z	A	Childhood in China
14	M	18	V	A	Not known
15	F	76	Z	A	Not known
16	M	47	Z	A	Childhood in China
17	M	20	Z	A	Childhood in China
18	M	17	V	A	Not known
19	F	32	Z	A	Childhood in China

^aAbbreviations: M, male; F, female; V, varicella; Z, zoster; A, ambulatory; H, hospitalized.

VZV DNA isolation and PCR assays.

DNA was purified from 200 μl of virus-infected cells within three serial passages by using a MiniBEST viral RNA/DNA extraction kit (version 3.0; Takara Biotechnology Co. Ltd., Dalian, China), according to the manufacturer's instructions. The same method was used to extract DNA from HELFs infected with the Oka vaccine strain (v-Oka), the parental Oka strain (p-Oka), and an American wild-type strain (strain MLS) (gifts from Anne A. Gershon, Columbia University). DNA from uninfected HELFs was used as a negative control in all experiments. The concentration of DNA was determined by spectrometry, and the DNA was stored at −20°C.

PCR amplifications were carried out in 50-μl volumes that included 2 μl of template DNA, 2.5 U of Ex Taq DNA polymerase (Takara Biotechnology), 0.4 μM of each forward and reverse primer, and 100 μM of each deoxynucleoside triphosphate (dATP, dTTP, dGTP, dCTP) (Advanced Takara Biotechnology, Dalian, China). The final reaction volume was adjusted to 50 μl with DNase- and RNase-free water. Thermal cycling comprised an initial hot start at 94°C for 90 s, followed by 30 to 35 cycles of amplification (94°C for 25 s, 52 to 56°C for 45 s, and 72°C for 25 s) and a final extension at 72°C for 10 min. For each primer set, a PCR-negative control, in which DNA extracted from uninfected HELFs was used as the template, was included. For detection of the PCR products, 5 μl of the PCR products was loaded onto 1.5% agarose gels and the gels were electrophoresed at 100 V for 30 to 60 min. For separation of the DNA fragments after amplification by PCR based on the R5 region of the VZV genome, 5 μl of the PCR products was loaded onto a 2% agarose gel and the gel was electrophoresed at 80 V for 1.5 h. The gels were stained with ethidium bromide and visualized under UV light. Two types of molecular weight markers, MspI- digested pUC18 DNA and a 100-bp ladder (both from TianGen Biotech, Beijing, China), were used to identify the sizes of the PCR products. All of the oligonucleotide primers used for PCR were published previously (Table 2) and were synthesized by Shanghai Sangon Biological Engineering Technology & Services Co., Ltd., China.

TABLE 2.

Primers used for genotyping of VZV isolates

ORF (restriction site)	Primer sensea	Primer sequence (5′→3′)	Genome positionb	Reference
ORF 1	F	TCA GCT GGC TTT TCT AAG AAT TCG	427-450	2
	R	TAT TTT TGG GAT CCG CAA TGT C	932-911
ORF 21	F	TGG CGC GGT TTA AAT GAA TTG A	33497-33518	2
	R	CAC GTG TAG CTC CAA AAA CCT AGG	33999-33976
ORF 54 (BglI)	F	CGT AAT GCA TAA CAG GCC AAC AC	95005-95027	2
	R	AAA CCT GGC GTC AAA CAT TAC A	95501-95497
ORF 22	F	GGG TTT TGT ATG AGC GTT GG	37837-37856	13
	R	CCC CCG AGG TTC GTA ATA TC	38283-38264
ORF 38 (PstI)	F	AAG TTT CAG CCA ACG TGC CAA TAA A	69060-69084	7
	R	AGA CGC GCT TAA CGG AAG TAA CG	69706-69684
ORF 62 (SmaI)	F	TTC CCA CCG CGG CAC AAA CA	106036-106055	14
	R	GGT TGC TGG TGT TGG ACG CG	106303-106284
R5	F	GGC AAA TAC TTA GAC CGT TTT	101992-102012	6
	R	TAA TGG ACT TTT AAT GGA TTG	102350-102329

^aF, forward; R, reverse.

^bBased on the sequence with GenBank accession number NC_001348.

Restriction enzyme reactions.

Restriction endonuclease digestion of the PCR products was performed with 7 μl of the PCR product, 10 U of endonuclease (BglI, PstI, or SmaI; Fermentas Inc.), and 2 μl of the accompanying 10× endonuclease buffer; the final reaction volume was adjusted to 20 μl with DNase- and RNase-free water. The reaction mixtures were incubated overnight at 37°C for BglI and PstI or 25°C for SmaI. The BglI- and PstI-digested products were separated by gel electrophoresis on 3% agarose (SBS Genetech, Beijing, China) at 80 V for 150 min and 100 min, respectively. The SmaI-digested products were electrophoresed on 4% agarose gel at 80 V for 3.5 h.

Sequencing.

The PCR products of the VZV genes (ORFs 1, 21, 22, and 54) were purified with an agarose gel DNA purification kit (Takara Biotechnology) and were sequenced by using a BigDye Terminator (version 3.0) cycle sequencing kit (Applied Biosystems, United Kingdom), according to the manufacturer's instructions. Automated DNA sequencing was carried out on an ABI Prism 3730XL DNA analyzer (Perkin-Elmer Applied Biosystems, Warrington, United Kingdom). Sequencing was performed with the forward primers listed in Table 2; however, reverse primers were employed to sequence the other strand whenever a different or an ambiguous nucleotide was found when the sequence was compared with the published sequence of strain Dumas (GenBank accession number NC_001348) in the homologous position.

RESULTS

Restriction fragment length polymorphism (RFLP) analysis of ORF38 and ORF54.

SNPs in ORF38 (PstI) and ORF54 (BglI) have commonly been used as genetic markers in VZV epidemiologic and vaccine studies (10, 13, 15). After PCR amplification with the primers for ORF38 listed in Table 2, a 647-bp fragment was obtained, as expected. Subsequent PstI digestion of the PCR products yielded two fragments of 357 bp and 290 bp from all 19 VZV isolates as well as strain MLS (data not shown). In contrast, the 647-bp fragments obtained from strains v-Oka and p-Oka were not cleaved by PstI. These results indicate that all 19 Chinese VZV isolates and strain MLS evaluated in this study contain the PstI site in ORF38, whereas both v-Oka and p-Oka were negative for the PstI site, as reported previously (10, 18).

PCR amplification of DNA with primers for ORF54 produced amplicons of 497 bp. The amplicons were digested with BglI, which resulted in two fragments of 256 bp and 241 bp. All of the viruses analyzed in this study, including the 19 Chinese isolates, strain MLS, and strains v-Oka and p-Oka, were positive for the BglI site (data not shown). The results of the analysis for both v-Oka and p-Oka are consistent with previous observations (10, 18).

RFLP analysis of ORF62.

PCR amplification of DNA fragments in ORF62 with the primers in Table 2 produced amplicons of 268 bp. Cleavage of the products from the 19 isolates with SmaI resulted in a set of 153-, 79-, and 36-bp DNA fragments (data not shown). The same result was achieved with the product amplified with the DNA of strains MLS and p-Oka. In contrast, the PCR product of strain v-Oka was cleaved into a set of 112-, 79-, and 41/36-bp fragments (data not shown). These results indicate that only the PCR product from v-Oka has the SmaI site, as was documented previously (14), while all 19 Chinese VZV isolates and strains MLS and p-Oka were negative for the SmaI site. Thus, the 19 Chinese strains were identified as wild-type VZV.

In summary (Table 3), the RFLP analysis showed that all 19 Chinese isolates as well as strain MLS were PstI positive (PstI⁺), BglI⁺, and SmaI negative (SmaI⁻) (typical of wild-type VZV); strain p-Oka was PstI⁻, BglI⁺, and SmaI⁻; and strain v-Oka was PstI⁻, BglI⁺, and SmaI⁺.

TABLE 3.

Summary of genotyping of VZV strains analyzed in this study

Strain	ORF38 PstI site	ORF54 BglI site	ORF62 SmaI site	R5 type(s)	SNP in ORF22a	SNP in ORFs 1, 21, and 54b
MLS	+	+	−	R5A	M1	A1
v-Oka	−	+	+	R5B	J	J2
p-Oka	−	+	−	R5B	J	J1
Chinese isolates	+	+	−	R5A (47.4%), R5B (52.6%)	J	J1

^aFrom Loparev et al. (13).

^bFrom Barrett-Muir et al. (1) and Quinlivan et al. (19).

Analysis of R5 alleles.

In order to study the allele distribution in the R5 variable region among the VZV isolates, PCR was performed with the primer pairs used by Hawrami and Breuer (6) (Table 2). Allele B yielded a band of 471 bp among the PCR products and was found in 10 (52.6%) of the 19 VZV isolates as well as in strains v-Oka and p-Oka (Fig. 1; Table 3). The PCR product for allele A at the R5 locus is 359 bp, which is 112 bp shorter than the PCR product of allele B, and was observed in nine (47.4%) Chinese isolates as well as strain MLS.

FIG. 1.

DNA fragments after amplification of the R5 region by PCR of. Lane M, 100-bp molecular size marker; lanes 1 to 19, DNA fragments from 19 clinic isolates, respectively, amplified by PCR; lanes 20 to 22, DNA fragments from strains MLS, v-Oka, and p-Oka, respectively, amplified by PCR; lane 23, PCR-negative control.

DNA sequencing analysis.

In order to compare the genotypes of the Chinese isolates in this study to those of isolates from other parts of the world, the amplicons of ORFs 1, 21, 22, and 54 were sequenced; and the informative polymorphic markers are listed in Fig. 2 and are summarized in Table 3.

FIG. 2.

Genomic variations of 19 Chinese isolates and 1 U.S. isolate, strain MLS. The sequence positions are based on the published genomic sequence of strain Dumas (GenBank accession number NC_001348). Green cells, genotype C markers; yellow cells, genotype J markers; purple cells, markers unique to various single-nucleotide mutations.

According to the nomenclature scheme proposed by Barrett-Muir et al. (1), all 19 Chinese isolates and strain p-Oka are genotype J1 on the basis of the SNP profiles in VZV ORFs 1, 21, and 54. The isolates are genotyped as J, according to the genotyping scheme of Loparev et al. (13). The genotype of strain v-Oka is J2. Strain MLS is genotype A1, according to the scheme of Barrett-Muir et al. (1), or genotype M1, according to the scheme of Loparev et al. (13) (Table 3).

DISCUSSION

VZV is highly infectious, causes large outbreaks of varicella in naive populations, and establishes lifelong latency; the subclinical reactivation of VZV, however, is believed to occur less frequently than the reactivation of other alphaherpesviruses (5, 13). The low number of reproductive cycles in the infected host may restrict opportunities for the introduction of mutations. The VZV genome is one of the most highly conserved among the genomes of human viral pathogens, with only about 0.1% sequence variation occurring between any two wild-type strains. Its highly conserved genome complicates genotyping and epidemiological surveillance efforts; nonetheless, several robust methods have led to a wealth of information on the genotypic variation among VZV isolates (2, 6, 10, 13, 15, 16, 17, 23).

This is the first study to analyze the genotypes of VZV clinical isolates from the middle eastern region of China. On the basis of an analysis of 38 polymorphic markers in four ORFs (ORFs 1, 21, 22, and 54), considerable homology was found among the 19 clinical strains. They can be placed in the same genotype by either one of the two commonly used typing schemes: genotype J1, on the basis of the SNP profiles of VZV ORFs 1, 21, and 54 (1), or genotype J, on the basis of the SNP profile in ORF22 (13). In contrast to a previous report that indicated that all three isolates from south China tested were genotype M and all the three isolates from north China tested were genotype E (13), all 19 strains analyzed in this study were found to be genotype J by the same genotyping method. These data suggest that VZV strains of at least three different genotypes are circulating in China and that each of them may be dominant in a certain geographic area.

Interestingly, three novel consecutive nucleotide (CGG) in-frame insertions (CCG in the coding strand) were present in ORF1 between nucleotides 780 and 781 of the Dumas strain (GenBank accession number NC_001348; Fig. 2) in 4 (21%) of the 19 isolates. The insertion results in the addition of a proline in the cytoplasmic portion of the ORF1 protein, a presumed type 2 membrane protein of VZV.

In addition, three new nucleotide substitutions were found in 2 of the 19 Chinese VZV isolates, strains 8 and 19. In strain 8, a synonymous C → T mutation in ORF21 (nucleotide 33608) and a nonsynonymous T → G mutation in ORF54 (nucleotide 95150) were found. The latter introduced an arginine to replace glycine at the nucleotide of 279 of this 769-amino-acid virus capsid portal protein. In strain 19, a nonsynonymous C → A substitution in ORF22 (nucleotide 38055) was observed and resulted in a threonine replacement of asparagine in the middle of the 2,673-amino-acid viral tegument protein.

The analysis presented in this report provides strong evidence that VZV isolates with the same genotype from the same area may include variants that can be distinguished. For example, even though the 19 isolates in this study all belonged to genotype J1, they may vary by three nucleotide insertions and a few single-nucleotide substitutions and may also have different R5 alleles. Those differences may be useful in epidemiological surveys, including investigations of nosocomial outbreaks.

It has been shown that the majority of wild-type VZV strains in temperate climates (such as North America, Europe, and eastern Australia) are PstI⁺ and BglI⁻ in ORF38 and ORF54, respectively, while the dominant pattern in countries with a tropical or a subtropical climate is PstI⁺ and BglI⁺ (such as Asia, Africa, and the Caribbean) (10, 11, 13, 20, 21). The 19 isolates in this study were collected in a city in China with a latitude of 31.85°N and a typical temperate climate, but all were found to be PstI⁺ and BglI⁺. This observation is consistent with the findings presented in a previous report in which all three isolates from south China were PstI⁺ and BglI⁺ (13). Although the 19 Chinese isolates in this study were genotyped as J or J1, analysis of the PstI site in ORF38 showed differences between VZV strains from China and those from Japan. In contrast to the dominance of PstI⁺ in Chinese strains, 25 to 47% of Japanese strains, including p-Oka and v-Oka, were PstI⁻ in various studies (8, 10, 13, 18, 21).

In addition, strain MLS, which was also evaluated in this study, was from New York City and was found to be genotype A1 according to the SNPs in ORFs 1, 21, and 54 or genotype M1 according to the SNPs in ORF22. Both genotypes A1 and M1 were found to prevail in Africa, Asia, and the Far East. Because the dominant VZV genotype in the United States is B or C by the method of Barrett-Muir et al. (1) or E by the method of Loparev et al. (13), the genotype of strain MLS represents a minor genotype in the United States. The RFLP pattern of strain MLS was PstI⁺, BglI⁺, and SmaI⁻, which is also more typical of Asian than North American VZV isolates.

The typing of a repeat region in the VZV genome, R5, was performed because the distribution of R5 alleles was previously found to be geographically related (6, 20, 21). For example, the R5A allele is dominant in European and North American strains, whereas R5B is the major allele in Japanese strains, including strain v-Oka. In this study, PCR analysis of the R5 variable repeat region showed that 9/19 (47.4%) of the Chinese strains and strain MLS had the R5A allele. The remaining 10 (52.6%) strains, along with strains v-Oka and p-Oka, had the R5B allele; the latter finding is consistent with previous observations (21). No strain with the C allele was found among the strains analyzed in this study. Our data suggest that neither R5A nor R5B is a dominant R5 allele in the Chinese strains, a result that seems different from those of analyses performed in other countries (6, 20, 21).

Several VZV genotyping schemes have been developed, but as yet, there is no commonly agreed upon genotype nomenclature (1, 13, 23). On the basis of the results of the heteroduplex mobility assay and phylogenetic analysis, selected SNPs in ORFs 1, 21, 50, and 54 were found to be useful for the study of genetic variation among VZV isolates and resulted in the identification of four genotypes, genotypes A, B, C, and J (1, 2, 19). This genotyping scheme was used to identify variants of VZV circulating in the United Kingdom and seven other countries on four continents (1, 2, 3). Another genotyping method (9, 12, 13, 15), based on the analysis of a short region (447 bp) in ORF22 and an additional region in either ORF21 or ORF50, was developed and could distinguish five major genotypes (genotypes E1, E2, J, M1, and M2) and two minor genotypes (genotypes M3 and M4). Genotype J strains were the most common in Japan, and genotype E strains were the most common in temperate latitudes. Strains belonging to genotypes M1 and M2 were the most common in tropical and subtropical regions. In different studies, an M3 genotype strain was isolated in the United States (22), and M4 genotype strains were isolated in Spain and France (12).

Another approach based on phylogenetic analysis of four glycoprotein genes (gH, gL, gB, and gE) and the IE62 major transactivator gene resulted in the division of four clades, clades A and D (Europe/North America) and clades B and C (Japan/Singapore) (5, 23). Two isolates from Thailand, however, were found to be segregated within clade D and therefore they seemed more Western than Asian in nature. Future analysis of the SNP profiles of those four VZV genes from Chinese isolates may provide further insights into the variants of the Asian strains in the genotyping scheme.