Background
Epstein-Barr virus (EBV) is a member of the
Lymphocryptovirus genus,
Gammaherpesvirinae subfamily of the
Herpesviridae family of viruses. This virus is associated with a wide variety of diseases, both benign and malignant, which ubiquitously infect humans and persist for the lifetime of the individual. During its life cycle, EBV has latent and productive (lytic) phases. The latent phase maintains the virus long-term in its host and can lead to the productive phase where virus is reactivated and produced allowing it to be transmitted. During the two phases, EBV expresses a set of viral gene products in its life cycle and some of these genes were proved to possess the potential to cause changes in the interactions between the virus and the host's immune system [
1,
2].
The biology and pathogenesis of EBV has been the focus of many studies but the clinical management of the disease is poorly understood. Whether certain EBV genotypes are involved in the pathogenesis of specific EBV-related diseases has been the subject of investigation in recent years. Several viral variants can be distinguished according to polymorphisms in EBV genes, such as EBV nuclear antigen (EBNA) and BZLF1, a potent regulator of the switch from latency to lytic phases encoded by the EBV
Bam HI fragment Z. EBV genotypes can be categorized as type 1 or type 2 on the basis of marked allelic polymorphisms within the EBNA2, 3A, 3B, and 3C genes [
3,
4]. Both EBV types have been detected in immunocompromised and immunocompetent hosts but type 1 EBV is predominant in Asian nasopharyngeal carcinoma and has a greater potential to transform B lymphocytes than EBV type 2. Type 2 EBV, on the other hand, enters the lytic cycle more readily than type 1 EBV [
5‐
7]. Sequence diversity of the BZLF1 gene promoter zone (Zp) (from -221 to +12, with respect to the transcription start site of BZLF1) have also been identified and variants are differentially distributed among malignant and non-malignant cells [
8,
9].
Childhood EBV infection is typically asymptomatic but can also induce three types of non-malignant disorders, including infectious mononucleosis (IM), EBV-associated hemophagocytic lymphohistiocytosis (EBV-HLH) and chronic active EBV infection (CAEBV). Certain linkages exist between these diseases where IM, usually a benign self-limiting disease, can develop to EBV-HLH and CAEBV in some patients. Likewise, EBV-HLH progresses very rapidly and becomes a life-threatening disease without immunosuppressive therapy, which occurs during the process of CAEBV sometimes or in association with fulminant IM[
10‐
12]. CAEBV is characterized by chronic or recurrent IM-like symptoms persisting over a long period of time and has a high likelihood of developing into EBV related malignant diseases, such as T/NK cell lymphomas, with a high fatality rate [
13‐
15]. Thus, this study aimed to investigate the association of BZLF1 Zp variants and type 1 and type 2 EBV and to explore the relationship between these EBV genotypes and clinical phenotypes of EBV-associated diseases in Chinese children.
In this study, EBV DNA from blood samples of 206 patients with IM, EBV-HLH, CAEBV, and healthy controls was examined by PCR for EBNA2 and EBNA3C subtypes (EBV type 1 and type 2) and Zp variants. This case-control study is the first investigation to explore the association between EBV subtypes and BZLF1-Zp variants and EBV infection in the China children population.
Discussion
This case-control study is the first investigation to explore the association between EBV subtypes and BZLF1-Zp variants and EBV infection in the China children population. In this study, statistical analysis determined that differences in the distribution of Zp variants were significant in the four patient categories. The frequency of the Zp-V3 variant in the CAEBV group was statistically higher than for other categories (
P ≤ 0.01), while Zp-P was predominant in all categories except CAEBV. This suggests that the Zp-P variant EBV was the most common variant found in China and that infection by Zp-V3 is strongly correlated to CAEBV. The Zp-V3 variant is significantly associated with malignancy in both immunocompetent and immunocompromised patients [
8,
9] and the higher frequency of the Zp-V3 variant in CAEBV patients observed in the current study suggests that CAEBV is more likely an entity of pre-malignancy. Similarly, the Zp-V4 variant was also identified in this study and was most associated with IM and healthy control cases. Zp-V1 was identified as a novel variant and was detected in 10 IM and health control cases but not in CAEBV and EBV-HLH. The absence or low level of ZP-V1 and Zp-V4 in CAEBV and HLH reflects a less severe pathogenesis than for the Zp-V3 variant which may enhance the tumorigenicity of EBV.
A novel Zp variant that differed from Zp-P by one substitution at position -100 (T→G) was detect in this study in only one patient with EBV-HLH. Due to the infrequent isolation of this variant, we did not include this data in correlations with disease. Previously described Zp variants, Zp-PV, Zp-V1-104, Zp-V1-105 or Zp-V1-119 [
9,
16] were not detected in any patient samples. Although it may be chance that these isolates were not detected, specific ethnic groups and geographical restrictions are likely to contribute to the narrow distribution of variants observed in the current study. The detection of different new variants suggests that the accumulation of viral mutations may contribute to the variations observed within the host during virus persistence.
Similar to other studies that reported that type 1 EBV was predominant in Asian nasopharyngeal carcinoma (86.5-96%) [
17,
18], the current study also found that type 1 EBV was predominant in all four categories (81.3-95%). Also in agreement to these studies, type 2 EBV infection was rarely detected (4-13.5%). These findings suggest that the diagnosis of EBV types 1 and 2 in patients is not likely to be useful for predicting susceptibility to EBV-related diseases in Chinese children. Although patients with Zp-V4 or Zp-V1 variants were always type 1 EBV carriers, this study did not confirm that Zp variants segregated by EBV type due to the extremely lower frequency of type 2 EBV in the Chinese study population. Gutierrez et al. had previously shown that the Zp-V3 variant was exclusively associated with type 2 EBV infection; however, the current study found that variant Zp-V3 co-existed with both EBV types. Geographic regions, sample sizes or various diseases are like to result in these differences.
The prevalence of co-existence EBV Zp variants within the four categories studied ranged from 14.6 to 37.5%. The majority of these co-existence viruses occurred in patients with CAEBV and EBV-HLH and always was found associated with Zp-P and not other variants. It is likely that the majority of people are first infected with a more prevalent variant like Zp-P, the predominant variant found in this study, but does not rule out the possibility that new point mutations are likely to be arised during EBV replication of in its hosts from pre-existing variant. In this way, the balance of one pre-existing virus variant which could be controlled by its host, may be disturbed by a specific new variant. Thus, virus replication, tropism, or immune evasion in its hosts could be greatly enhanced after acquiring this new variant. As the Zp-V3 variant was associated with severe diseases in this study, the Zp-V3 type point mutations derived from Zp-P are likely to be associated with a more invasive capacity than Zp-V1 or Zp-V4 variants. Taken together, superinfection by multiple strains of EBV, especially the presence of the Zp-V3 variant, may be a contributing factor in the development of severe EBV infections in children. Thus, these findings may give some prospect to explore the differential distribution of Zp variants in susceptible populations and their association with severe or even fatal EBV diseases. A close dynamic follow-up on patients carrying EBV from an early stage of infection may help us understand how the host immune response allows such mutations to occur.
Just how an individual acquires such mixtures of Zp variants is unknown. This could occur by simultaneous acquisition or by the serial accumulation from exposure to different variant carriers. It seems implausible that such co-infections can be co-acquired from a carrier who was shedding multiple variants in saliva, because it is unclear how the source can accumulate multiple infections before transmitting those orally shedding multiple EBV variants to the next. As infection by EBV with the Zp-P variant was a prerequisite for co-existence in this study, it is possible that an individual is more likely to acquire a prevalent variant, such as Zp-P, at first exposure to the virus, and then the host immunity to this variant is developed. However, part of hosts may fall short in resisting another different variant the next time. It is more likely that some instances of co-existence are the result of serial acquisition over time from independent sources. The host could gain multiple variants dynamically in the light of this hypothesis, but this cannot explain why co-infections harbor no more than two variants. It is plausible that the co-existence variants observed in this study may be new variants that were generated from point mutations of a pre-existing variant. However, the complex relationship between EBV variants and the host requires further investigation of other EBV-related diseases, various ethnic groups, different tissues or a study including a larger sample size.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
YKJ carried out most of the studies and drafted the manuscript. GL and SY participated parts of the studies and writing. ZDX and KLS provided consultation and preparation of the final report. All authors read and approved the final manuscript.