Sequence analysis of the Epstein-Barr virus (EBV) BRLF1 gene in nasopharyngeal and gastric carcinomas.

BACKGROUND: Epstein-Barr virus (EBV) has a biphasic infection cycle consisting of a latent and a lytic replicative phase. The product of immediate-early gene BRLF1, Rta, is able to disrupt the latency phase in epithelial cells and certain B-cell lines. The protein Rta is a frequent target of the EBV-induced cytotoxic T cell response. In spite of our good understanding of this protein, little is known for the gene polymorphism of BRLF1.
RESULTS: BRLF1 gene was successfully amplified in 34 EBV-associated gastric carcinomas (EBVaGCs), 57 nasopharyngeal carcinomas (NPCs) and 28 throat washings (TWs) samples from healthy donors followed by PCR-direct sequencing. Fourteen loci were found to be affected by amino acid changes, 17 loci by silent nucleotide changes. According to the phylogenetic tree, 5 distinct subtypes of BRLF1 were identified, and 2 subtypes BR1-A and BR1-C were detected in 42.9% (51/119), 42.0% (50/119) of samples, respectively. The distribution of these 2 subtypes among 3 types of specimens was significantly different. The subtype BR1-A preferentially existed in healthy donors, while BR1-C was seen more in biopsies of NPC. A silent mutation A/G was detected in all the isolates. Among 3 functional domains, the dimerization domain of Rta showed a stably conserved sequence, while DNA binding and transactivation domains were detected to have multiple mutations. Three of 16 CTL epitopes, NAA, QKE and ERP, were affected by amino acid changes. Epitope ERP was relatively conserved; epitopes NAA and QKE harbored more mutations.
CONCLUSIONS: This first detailed investigation of sequence variations in BRLF1 gene has identified 5 distinct subtypes. Two subtypes BR1-A and BR1-C are the dominant genotypes of BRLF1. The subtype BR1-C is more frequent in NPCs, while BR1-A preferentially presents in healthy donors. BR1-C may be associated with the tumorigenesis of NPC.

Background

Epstein-Barr virus (EBV) is a ubiquitous human herpesvirus that infects over 90% of the world population. As the causal agent of infectious mononucleosis, EBV is also tightly associated with various malignancies, including Hodgkin's disease, Burkitt's lymphoma(BL), nasopharyngeal carcinoma(NPC), and B and T cell lymphomas in immunocompromised individuals such as AIDS patients and organ transplant recipients[1,2]. It is also responsible for some gastric carcinomas (GC). The EBV infection is found in 80-100% of gastric lymphoepithelioma-like carcinoma cases and 2-16% of common types of gastric adenocarcinoma [3-6]. In Northern China this rate is about 7.0% according to our previous study [7].

After primary infection, EBV establishes a lifelong, asymptomatic state in B cells. However, EBV can periodically reactivate and replicate in a lytic manner [8]. Understanding how viral latency is disrupted is a central focus in herpesvirus biology. Induction of the switch from latency to lytic cycle is associated with expression of immediate-early (IE) protein Rta (R transactivator), the product of the BRLF1 gene [9]. Rta is a 605-amino acid (AA) protein with unknown cellular homologues. The N-terminus of Rta contains an overlapping DNA binding (AA 1 to 320) and dimerization (AA 1 to 232) domain that does not correspond to any described DNA binding motif previously [10]. The transcriptional activation domain is found in the C-terminal region of the protein. An obligatory acidic activation domain (AA 520 to 605) contains highly conserved hydrophobic residues that are predicted to form alpha helices [10]. A weaker accessory activating domain contains two proline-rich subregions (AA 352 to 410 and 450 to 500).

ZEBRA, the product of EBV BZLF1 gene, had been thought to be the only viral protein capable of initiating the lytic cycle [11-14]. In recent years Rta has been found to be able to disrupt latency through activating Zp, the promoter of BZLF1, leading expression of ZEBRA, and thereby stimulation of early lytic genes, DNA replication, and late gene expression [9,15,16]. There are a number of interesting differences between the induction of lytic EBV infection by BZLF1 and that by BRLF1. Zalani S, et al. [17] reported that Rta can disrupt viral latency in an epithelial cell-specific manner (in contrast to the ability of ZEBRA to disrupt latency in B cells), and the mechanisms leading to disruption of EBV latency appear to be cell-type specific. Also, it has been demonstrated that BRLF1, not BZLF1, requires activation of the p38 and c-Jun stress MAP kinase pathways for induction of lytic EBV infection [18], and also requires PI3 kinase activation [19].

Several components of the immune system contribute to the highly efficient control of virus replication and proliferation of immortalized, EBV-infected cells in healthy individuals, and probably the most important components are HLA-restricted specific cytotoxic T lymphocytes (CTLs). As EBV can switch directly from the latent state into the lytic cycle without any expression of further latent proteins [20], CTL directed against latent proteins might not be able to prevent the ongoing viral replication. Therefore, CTL directed against immediate-early (IE) proteins is a pivotal step to control the virus lytic activation. Rta has been demonstrated to have multiple epitopes recognized by EBV-specific CTL [21,22]. Delineation of sequence variations of CTL epitopes may help the development of an effective control of EBV replication and cell proliferation.

A notable feature of EBV-associated malignancies is variation in incidence and the proportion of EBV-positive tumors in different geographic regions [23,24]. The disparity is poorly understood. To explore the potential association of the EBV-associated malignancies with integrated EBV sequence variations, as well as the possibility of a CTL-based control of EBV replication and cell proliferation, we analyzed the sequence variation of EBV BRLF1 gene in EBVaGCs, NPCs and healthy donors.

Results

Sequence variation of BRLF1 gene

The sequence of BRLF1 gene coding 605 AAs was successfully amplified in 34 EBVaGCs, 57 NPCs, and 28 TWs, respectively. All the sequences were compared with the prototype B95-8 sequence. Nucleotide changes were detected in 31 loci, 14 of which resulted in AA changes. Among the 17 loci with silent nucleotide changes, one (at 103654) was detected with an A/G interchange in all the specimens tested. The translated AA mutations from the sequence variations were summarized in Figure 1. According to the phylogenetic tree (Figure 2), 5 distinct subtypes of BRLF1 were identified among the observed 119 specimens, namely subtype BR1-A, BR1-B, BR1-C, BR1-D, and BR1-E. Two subtypes, BR1-A and BR1-C, were found to be dominant in the total specimens.

Figure 1

Observed BRLF1 sequence variations in EBVaGC, NPC biopsies and TWs of healthy donors in Northern China. The numbers in the first row correspond to the amino acid positions and the numbers in the second correspond to the nucleotide positions, under which the B95-8 prototype amino acid and nucleotide sequences are listed. Different patterns are noted to the far left column, while the specimens showing identical sequences to each other are listed by a representative isolate in the second column. The following numbers separated by "/" denote the number of the identical sequences from EBVaGC, NPC and TW, respectively. Only sequences different from B95-8 are indicated. The small letters denote the nucleotide, and the amino acids are denoted by capital letters. The GD1 sequence was taken from EBV genomes AY961628 [25]. C666-1 and SNU-719 were two EBV-positive cell lines [26,27], whose sequences were obtained by using PCR-direct sequencing method.

Figure 2

Phylogenetic tree drawn from the BRLF1 amino acid sequences of 21 representative isolates using neighbor-joining method. BRLF1 subtypes BR1-A, BR1-B, BR1-C and BR1-D were shown on the right. In this figure the conserved subtype BR1-E (represented by GC4) was not listed.

The subtype BR1-A, which was represented by NPC87, was detected in 42.9% (51/119) of samples. Forty specimens in this subtype had 2 common coding changes: 377(Ala→Glu), 542(Ser→ Asn), while 6 specimens only had residue 377, 5 specimens only had residue 542 changes. Interestingly, residue 489 caused different AA changes among different specimens, namely, Gln→Arg in 28 specimens, Gln→Lys in 21 specimens. Besides these 3 residues, 3 isolates (TW165, TW121, and NPC6) showed Val→Ile interchange at residue 479. Silent changes in this pattern were detected in 3 residues: 486(CCG→CCA), 510(GAA→GAG), 572(CCC→CCA) (data not shown). The prevalent AA mutations at residues 377, 489 and 542 of this subtype were identical to the GD1 strain, which is a representative EBV strain isolated from NPC patients in Guangdong, China [25]. Also, the BRLF1 gene in C666-1 cell line, which was established from an undifferentiated NPC biopsy in Southern China [26], harbored only these three mutations.

The second common subtype BR1-C (represented by NPC57) was detected in 42.0% (50/119) of samples. This subtype contained 3 common signature residues: residue 273 (Arg→Met), 316(Lys→Glu), 542(Ser→Asn). Additionally, some isolates showed one or more additional sequence variations at other positions. Residues 284(Gly→Ser), 288(Thr→Ser), 371(Pro→Gln) coexisted in 22 specimens, while 23 specimens contained residue 371 only; one specimen (GC95) contained residue 288 only. At residue 489, 22 specimens had Gln/Arg interchange; 9 specimens had Gln/Lys interchange. An interchange Tyr/His at residue 292 was detected in 22 specimens; interchange Val/Ile at residue 479 in 19 specimens. This subtype involved 16 silent mutations in different isolates (data not shown).

The rest 3 subtypes were only detected in small numbers of specimens. The subtype BR1-B shared 2 common AA substitutions: residues 273(Arg→Met), 316(Lys→Glu). Additional Gln/Arg interchange at residue 489 was detected in 6 specimens. The subtype BR1-D was detected in 5 specimens, which had 5 common residues: 273 (Arg→Met), 316(Lys→Glu), 403(Pro→Ser), 406(Leu→Phe), 414(Gly→Val). The last subtype BR1-E, which is represented by the strain GC4, had the same AA sequence with the prototype B95-8, except for a silent mutation at 103654. This subtype was only found in 3 EBVaGCs. The BRLF1 gene in EBV-positive GC cell line SNU-719, which was established from a Korea GC patient [27], also showed conserved sequence.

Distribution of BRLF1 subtypes in EBVaGCs, NPCs, and TWs

The frequency of BRLF1 subtypes in EBVaGCs, NPCs, and TWs of healthy donors was summarized in Table 1. Fisher's exact test was used to determine the difference of the BRLF1 subtypes among the EBVaGCs, NPCs, and the TWs. Two subtypes, BR1-A and BR1-C, were dominant in the tested specimens. BR1-A was detected in 42.9% (51/119) of total specimens, that were 13/34(38.2%) EBVaGCs, 19/57 (33.3%) NPCs, 19/28 (67.8%) TWs. BR1-C was found in 50 specimens (42.0%), including 12/34 (35.3%) EBVaGCs, 30/57(52.7%) NPCs, and 8/28 (28.6%) TWs. The present rate of BR1-A in TWs (67.8%,19/28) was significantly higher than in EBVaGCs (38.2%,13/34) or NPCs (33.3%,19/57); while, BR1-C was seen more in NPCs (52.7%,30/57) than in EBVaGCs (35.3%,12/34), or in TWs (28.6%,8/28) (P = 0.0019).

Table 1

Distribution of BRLF1 subtypes in EBVaGCs, NPCs, and TWs

BRLF1 subtypes	EBVaGC(n = 34)	NPC(n = 57)	TWs(= 28)
BR1-A	13(38.2%)	19(33.3%)	19(67.8%)
BR1-B	6(17.6%)	4(7.0%)	0
BR1-C	12(35.3%)	30(52.7%)	8(28.6%)
BR1-D	0	4(7.0%)	1(3.6%)
BR1-E	3(8.9%)	0	0

Variation analysis in BRLF1 functional domains

As an important transactivator, BRLF1 gene harbors 3 vital functional domains: dimerization, DNA binding, and transactivation domains [10]. The AA mutations in BRLF1 functional domains were summarized in Table 2. In this study, the domain of dimerization (AA 1 to 232) was found to be stably conserved, where no AA mutations were detected (data not shown). Five residues (273, 284, 288, 292 and 316) were detected to have AA mutations in the DNA binding domain. The prevalent mutations in this domain were found in residue 273 (R→M) and 316 (K→E). The mutation R→M at residue 273 affected 50% (17/34) of EBVaGCs, 66.7% (38/57) of NPCs, 32.1% (9/28) of TW isolates; the mutation K→E at residue 316 affected 52.9%(18/34) of EBVaGCs, 66.7% (38/57) of NPCs, 32.1% (9/28) of TW isolates. The distribution of mutation R→M at residue 273 was significantly different among 3 types of specimens (χ2 = 9.28, P < 0.01; EBVaGC vs NPC: χ2 = 2.47, P > 0.05; EBVaGC vs TW: χ2 = 2.01, P > 0.05; NPC vs TW: χ2 = 9.05, P < 0.01); likewise, mutation K→E at residue 316 was distributed differently among 3 types of specimens(χ2 = 9.08, 0.01