Epstein-Barr virus infection as potential indicator of the occurrence and clinical presentation of systemic lupus erythematosus

Introduction The relationship between Systemic lupus erythematosus (SLE) and Epstein-Barr virus (EBV) infection has been suggested for decades, but the underlying mechanism of the EBV influence on SLE development remains to be elucidated. Methods The goals of this research, which included 103 SLE patients and 99 controls, were to investigate the association of the parameters of EBV infection and SLE, to explore whether pooled demographic, clinical and EBV markers achieve a more significant effect on SLE development than each of them individually, and to evaluate EBV nuclear antigen 1 (EBNA1) and latent membrane protein 1 (LMP1) gene polymorphisms in isolates from SLE patients. Results Comprehensive results related to serological, molecular and sequence markers of EBV infection in SLE patients demonstrated even 24 times higher possibility of having SLE if there is the presence of anti-EBV-EA(D) (early antigen) IgG antibodies (OR=24.086 95%CI OR=2.86-216.07, p=0.004). There was the same distribution of glucocorticoids (p=0.130), antimalarials (p=0.213), and immunosuppressives (p=0.712) in anti-EBV-EA(D) IgG positive and negative SLE patients. Further, higher anti-EBV-EA(D) IgG antibodies titers were identified as independent factors associated with lymphopenia, hematological SLE manifestation (OR=1.041, 95%CI OR=1.01-1.08, p=0.025, while a higher titer of anti-CA (viral capsid antigen) IgG antibodies (OR=1.015, 95%CI OR=1.01-1.03, p=0.019) and positive RF (rheumatoid factors) (OR=4.871, 95%CI OR=1.52-15.61, p=0.008) were identified as independent factors associated with alopecia within SLE. Finally, novel data on EBV EBNA1 and LMP1 gene polymorphisms in lupus are reported. Conclusion The results support further investigation targeting EBV as a prognostic marker and therapeutic goal for lupus.


Introduction:
The relationship between Systemic lupus erythematosus (SLE) and Epstein-Barr virus (EBV) infection has been suggested for decades, but the underlying mechanism of the EBV influence on SLE development remains to be elucidated.

Methods:
The goals of this research, which included 103 SLE patients and 99 controls, were to investigate the association of the parameters of EBV infection and SLE, to explore whether pooled demographic, clinical and EBV markers achieve a more significant effect on SLE development than each of them individually, and to evaluate EBV nuclear antigen 1 (EBNA1) and latent membrane protein 1 (LMP1) gene polymorphisms in isolates from SLE patients.

Introduction
Systemic lupus erythematosus (SLE) is a female predominant systemic autoimmune disease.It is characterized by the involvement of multiple organs and significant morbidity through pathogenic autoantibody production caused by the autoactivation of T and B cells.SLE includes repeated episodes of exacerbation, known as flares, and remission.Similar to other autoimmune disorders, the etiology of SLE still needs to be clarified entirely.The interplay between various genetic and environmental risk factors contributes to the onset and perpetuation of SLE.From external factors, infection, particularly Epstein-Barr virus (EBV) infection, has multiple potential roles in driving autoimmunity (1).
EBV is one of the most common human viruses.As a member of the Human Herpesviruses family, it latently infects up to 99% of the world's population with the ability to survive in cells for life (2).With such unique capacities, it represents a constant challenge to the host.After the primary lytic infection of epithelial cells, EBV maintains latency in B lymphocytes with the expression of a limited number of genes defined with different types of latency (3).However, it occasionally reactivates the lytic cycle resulting in shifts between different sets of expressed genes (4).The humoral response includes the synthesis of antiviral antibodies against antigens of both the lytic and latent phases.Firstly, anti-EA (early antigen) and anti-CA (viral capsid antigen), during the first 3-4 weeks and at the time of the onset of the clinical symptoms, respectively (5).Although anti-CA IgG levels persist for life, the levels are higher during lytic infection (primary or reactivation) (6).On the other hand, EA IgG could be found in 85% of the acute infection, with usual persistence of 3 months, but with detectable levels years after primary infection in some cases (5).It is estimated that up to 30% of healthy subjects with a history of EBV have EA (D) IgG (5).Moreover, these antibodies could be detectable during reactivations or in immunocompromised patients (5,6).Anti-EBNA-1 IgG is usually undetectable during the first month after the onset of clinical symptoms and is therefore indicative of past infection (5).
The relationship between SLE and EBV infection has been suggested for decades, but the underlying mechanism of the EBV influence on SLE progression remains to be elucidated.Several models have been proposed, among which antibody production through typical molecular mimicry and epitope spreading is the most important.The EBV genome encodes human homolog proteins such as latent protein, EBV nuclear antigen 1 (EBNA1), and the lytic protein, viral IL-10, that in SLE patients stimulate humoral and inflammatory immune responses (7).In particular, EBNA1 epitopes induce the production of autoantibodies against C1q and cross-react with SLE autoantigens, including Sm and Ro (7,8).Based on the previously described knowledge, it was suggested that impaired control of EBV infection and cell-mediated immunity, with altered T cell function and the presence of multiple antibodies, leads to polyclonal expansion of EBV-infected B cells (9).Finally, autoreactive B cell activation during flares leads to frequent relapsing-remitting episodes of EBV infection, contributing to a vicious circle of repeated immune stimulations (10).Recent examples of more frequent EBV reactivations in SLE patients include the presence and elevation of serological markers CA IgG and EA IgG, higher frequencies of EBVinfected cells, higher viral loads and blood mononuclear cells, and increased expression of lytic genes (11).Moreover, in SLE patients, EBV-specific cytotoxic T and dendritic cells have reduced functionality, decreasing interferon levels after EBV infection (12,13).The inverse relationship between decreased EBV lytic cycle responsive T cells and levels of anti-EBV antibodies is also followed by a negative correlation with the SLE disease activity index (SLEDAI) (14).
The balance that EBV establishes with the immune system and the host's capacity to control cell proliferation in some EBVassociated diseases could be braked by the consequences influenced by specific variability of EBV genes (15,16).In this regard, EBNA1 and latent membrane protein 1 (LMP1) genes are particularly interesting, with EBNA1 as the most prominently expressed EBV gene and LMP1 as the crucial viral oncogene.Their nucleotide variants and sequence specificities in previous studies showed potential contributions to the malignancy development (15,17,18).However, investigation of EBV gene variability and whether any mutation in viral genes contributes to SLE pathology has never been published before.Moreover, understanding this diversity in EBVassociated diseases revealed the possibility of predicting potential viral epitope targets to determine a treatment strategy (19).
Guided by the knowledge gaps and numerous unsolved hypotheses about SLE evolution, the goals of this research were set: a) to investigate the association between the respondents' characteristics, including the parameters of EBV infection and SLE; b) to explore whether pooled markers achieve a more significant effect on SLE development than each of them individually, c) to evaluate EBV EBNA1 and LMP1 gene polymorphisms in isolates from SLE patients.

Study design and participants
This prospective cohort study included 103 SLE patients diagnosed and treated at the Clinic for Allergology and Immunology University Clinical Center of Serbia and the Institute of Rheumatology, Belgrade, between June 2020 and November 2021.All SLE patients met at least 4 ACR SLE 1997 classification criteria (20).Patients with significant comorbidities (severe cardiac, pulmonary, and psychiatric diseases or dementia) and those unable to cooperate were excluded from the study.After detailed clinical interviews and physical examination of patients, all relevant demographic, clinical, and laboratory data were collected from their medical records.Disease activity was assessed using SLEDAI 2K scale (21) and Physician Global Assessment (PGA) (21), while fatigue was evaluated by the FACIT Fatigue Scale (version 4) (22).Patients with SLEDAI 2K≥6 were considered to have active SLE.Lupus Low Disease Activity State (LLDAS) was defined as SLEDAI-2K score ≤4, no activity in major organs, and PGA <1 (23,24), while SLEDAI-2K score = 0 defined remission.The control group consisted of 99 participants.After a thorough interview and physical examination, we excluded those with positive personal or family histories of systemic autoimmune diseases and suggestive signs and symptoms of systemic autoimmune or active malignant disease.The exclusion criteria for both study cohorts were age under 18, unfeasibility to cooperate, and malignancies.All participants provided written informed consent.The study was performed in accordance with the Declaration of Helsinki and was approved by the Ethical Board of the Faculty of Medicine, University of Belgrade (No 1550/IX-14).

Samples
SLE patients' serum and whole blood samples were collected at the Clinic for Allergology and Immunology University Clinical Center of Serbia and at the Institute of Rheumatology, Belgrade.Control group serum and whole blood samples were obtained from volunteers at the Institute of Rheumatology, Belgrade and the Institute of Microbiology and Immunology, Faculty of Medicine, University of Belgrade.After collecting 5 ml of blood in plain vacutainers, sera were separated by centrifugation.Plasma was also separated by centrifugation from 5 ml of ethylenediaminetetraacetic acid (EDTA) blood collected in vacutainer tubes.Two tubes (sera and plasma) from each patient were immediately tested or stored at -70°C until further analysis.

EBV serological testing
Anti-EBV antibodies against CA (IgG), CA (IgM), EA (IgG), EA (IgM), and EBNA1 (IgG) were identified and measured using commercial ELISAs according to manufacturer's instructions in collected sera (Euroimmun, Lubeck, Germany).The assays used recombinant [EBNA1 and EA(D)] or purified (CA) antigens.Standard calibrators were used in each assay to calculate index values/ optical density (OD) ratios, which serve as a quantitative measure of IgG antibody levels or a semi-quantitative measurement of IgM antibody levels.All assays met pre-determined quality control measures based on positive, negative, and blank controls.The positivity of IgG antibody presence was defined by a cut-off value of 20 relative units (RU/ml).For the sample values that were out of linearity defined by the used assays, re-measurement was done as following: samples were remeasured in a new test run at a dilution of e.g.1:400 and results in RU/ml read from the calibration curve were multiplied by a factor 4. The positivity of IgM antibody presence was defined as OD ratio ≥1.1.Absorbances were recorded on a Multiscan FC microplate reader (Thermo Scientific, Massachusetts, USA) using a wavelength of 405 nm with background subtraction at 650 nm.

EBV DNA detection
Viral DNA was isolated from 200 mL plasma using a PureLink Genomic DNA Mini Kit (Invitrogen by Thermo Fisher Scientific, Massachusetts, USA) according to the manufacturer's instructions.Two hundred and two DNA isolates were further used in a nested-PCR method to amplify two EBV genes: EBNA1 and LMP1.Amplifications of the C terminus of EBNA1 and C terminus of LMP1 were performed by nested PCRs as described previously (25), using primers that were reported by Lorenzetti et al. and Li et al. (26,27).

EBV EBNA1 and LMP1 carboxy-terminal regions sequencing with sequence analysis
For purification, performing cycle sequencing reactions and sequencing of EBNA1 and LMP1, both sense and antisense strands, principles described before were used (25).
LMP1 and EBNA1 nucleotide sequences 506-bp and 329-bp long were separately aligned and compared with a reference sequence in Bioedit 7.0.5.3 software (28).EBNA1 subtypes and sub-variants were defined after inspection of signature amino acid changes at the following positions: 471, 475, 476, 479, 487, 492, 499, 500, 502, 517, 520, 524, 525, 528, and 533.Characterization of LMP1 variants was performed using the same software by examination of amino acid changes described by Edwards et al. (29).The sequence used in both alignments as the reference was downloaded from the GenBank/ EMBL7DDBJ database under the accession number V01555.

Data analysis
Numerical data were described with the arithmetic mean and standard deviation or median and interquartile range (IQR), depending on the data distribution.Normal distribution was evaluated by mathematical (Kolmogorov-Smirnov and Shapiro-Wilk test, skewness and kurtosis, and coefficient of variation) and graphical methods (histogram, box-plot, Q-Q diagram).Categorical variables were presented as absolute and relative numbers in n (%).Student t-test for two independent samples or Mann-Whitney U test was used to compare numerical data between study cohorts, depending on the data distribution.The chi-square test was used to test the difference in the distribution of categories of two independent samples.
To evaluate all possible factors influencing a person's probability of having SLE, univariate logistic regression analysis was performed first, with multivariate analysis after that with significant variables from the previous analysis in the model.Variance Inflation Factor (VIF) multicollinearity testing method and correlation coefficients were used.All variables with VIF>5 were eliminated from the multivariate model.Also, according to the co-variance matrix and correlation coefficients, one of two correlated variables with a lower p-value in univariate logistic regression was eliminated.Forward Wald regression method was applied, and only the last step was presented in the results.The odds ratio (OR), 95% confidence interval of odds ratio (95% CI OR), and p-value were reported.Complete statistical analysis was performed in SPSS Statistics for Windows Version 26.0 (IBM Corp. Armonk, NY).
The sample size was calculated by G Power 3.1.9.2 for the major outcomethe presence of anti-EBV-EA(D) IgG antibodies in SLE patients and control according to previously known data (7) and for the effect size of 0.3, error of the I type (a) of 0.05, and statistical power of 0.8.It was obtained that 88 observation units per study group would be enough to determine the differences between them.

EBV infection status in SLE patients and controls
The prevalence of EBV DNA (EBNA1 and/or LMP1 gene) was 11% in the plasma of SLE patients, without statistically significant difference in comparison with controls (p=0.893).Anti-EBV-CA IgM, anti-EBV-EA(D) IgG, and anti-EBV-EA(D) IgM antibodies were more prevalent in SLE patients than in controls (p<0.001,p<0.001, and p=0.008, respectively).The status of EBV infection was determined based on the results from both EBV DNA and anti-EBV antibody testing.The active EBV infection was declared if EBV DNA and/or anti-EBV IgM antibodies were present.SLE patients more often had active EBV infection than controls (42% vs 18%, p<0.001) (Table 2).SLE patients with latent EBV infection were significantly more often treated with immunosuppressive drugs than those with active infection (37% vs. 19%, p=0.047), while the distribution of other medications (corticosteroids, antimalarials, pulse therapy) was the same between SLE patients with active and latent EBV infection (p=0.797,p=0.306, and p=0.773, respectively) (Available in Supplementary Material Table 1).

Evaluation of factors associated with SLE
All applicable EBV and clinicodemographic parameters were evaluated for defined outcomes (presence of SLE, the active form of SLE disease, and different clinical manifestations of SLE: rash, arthritis, ulcerations, alopecia, nephritis, leucopenia, lymphopenia, thrombocytopenia).Univariate, then multivariate regression analysis was performed.All variables with p<0.05 in univariate regression were included in the multivariate model taking into consideration collinearity.The complete analysis is available in the Supplementary Material Tables 2-12.

Variations of EBV EBNA1 and LMP1 carboxy-terminal regions sequences
Eight sequences of the EBNA1 gene fragment (coordinates 109261-109590) from 8 SLE patients were analyzed and compared with the B95-8 prototype strain.Inspection included signature amino acid changes at the following positions: 471, 475, 476, 479, 487, 492, 499, 500, 502, 517, 520, 524, 525, 528, and 533.Based on amino acid substitutions, isolates were grouped into two prototype subtypes, P-Ala and P-Thr (Table 4).Since none of the obtained sequences fully corresponded to the prototype sequences, further analysis of additional nucleotide variability and amino acid substitutions was performed.Three subvariants (sv) were identified: P-Ala-sv-2 and -3; P-Thr-sv-7.All three subvariants were previously found in Serbian isolates, however, P-Ala-sv-3 and P-Thr-sv-7 were described as newly identified (18).P-Ala-sv-3 differed from P-Ala-sv-2 in position 499, where the new subvariant retained the same nucleotide combination as the prototype sequence.P-Thr-sv-7 was characterized by four changes compared to the prototype sequence (positions 476, 487, 492, and 524) or two changes compared to the P-Thr sequence (499 and 520).
Eight sequences of the LMP1 gene fragment (coordinates 168719-168213) from 8 SLE patients were aligned and compared with the B95-8 prototype strain.Characterization of LMP1 variants was performed using by well-known examination of amino acid changes described by Edwards et al. (29).Due to divergence from the reference sequence, presence of deletions, number of 11-amino acid repeats and total amino acid changes, isolates were divided into four variants: B95-8, North Carolina, Mediterranean and China1 (Table 5).

Discussion
After numerous attempts to prove a cause-and-effect relationship between EBV infection and SLE, more data on EBVdependent features still need to be included for this evidence (30).The only fact is that the association between SLE and EBV increased reactivation due to insufficient latency regulation or enhanced transition from latent to lytic phase (30, 31).This confirmation was established by a few dozen studies based on detecting EBV infection markers in lupus patients (9).However, most studies measured a limited number of viral markers, quite often without a control group to compare findings.Thus, the latest publications did not include IgM antibodies to CA and EA, free viral DNA or gene sequence variability (7).Moreover, European research was limited primarily to Denmark and Italy, with rare exceptions from France, Netherlands and Hungary (9).To our knowledge, comprehensive analysis of EBV markers in lupus patients and their potential relationship with characteristics, different clinical presentations of SLE or therapeutic modalities is rare among literature data.Therefore, identifying independent EBV factors for SLE and our study's first molecular analysis of EBV genes represent a valuable contribution to this still under-investigated scientific field.
It is known that occasional EBV reactivations can be diagnosed serologically and by molecular detection of viral DNA.Although the specific serological interpretation of EBV reactivation in SLE patients may be missed because of a possible reflection of a non-specific antibody increase due to an autoimmune background, the available results say otherwise.The antibody responses towards other herpesviruses, including cytomegalovirus (CMV) and herpes simplex virus (HSV), are similar between SLE patients and controls (30).Interestingly, however, the dominant EBV latency type maintained during lupus could not be clearly defined using commonly known criteria.In particular, the combination of BZLF1 and LMP1 in SLE patients probably represents dysregulated latent gene expression in the direction of intermediate latency form between latency II and III (30).On the other hand, EBV reactivation in these patients does not have to be a consequence of immune dysregulation but perhaps immunosuppression (11).
This research emphasized the hypothesis that the immune response to EBV activity represents an important element in the sequence of events that follows the clinical course of SLE.According to the results, there is undoubtedly a highly significant EBV activity compared to the control group.Firstly, anti-EBV-CA IgM, anti-EBV-EA(D) IgG, and anti-EBV-EA(D) IgM antibodies were more prevalent in SLE patients than controls.Moreover, all of the mentioned antibodies together with anti-EBV-EBNA1 IgG had also significantly higher titers in SLE patients than controls.Previous research showed similar results, including prospective investigation of SLE disease transition (7,11).Based on all  it might be essential to interpret it together with the results of Wood et al. (7).In one of their latest paper, anti-EBV-EA IgG antibodies were higher when expression of interferon (IFN) and inflammatory or lymphoid and monocyte responses were higher.In addition, those patients with higher anti-EBV-EA(D) IgG responses were more likely to have renal involvement.Another publication that described this potential association with lupus nephritis detected the expression of EBV genes in renal tissue (49).Finally, in our study, a higher titer of anti-CA IgG antibodies and positive RF revealed independent factors related to alopecia within SLE.This finding should be analyzed in further investigations also.
The results, including the interplay between smoking and EBV in autoimmunity development, are scarce and conflicting.In those reported in multiple sclerosis (MS) patients, there are indications of competing antagonism, where mentioned exposures compete to affect the outcome (50).This study did not reveal any similar findings.However, shorter smoking duration was shown as an independent factor associated with lymphopenia in lupus patients.
According to the last systemic review with meta-analysis, only seven studies tested and compared EBV DNA presence between SLE and the control group (9).In addition to high heterogeneity, the positive rate for EBV DNA in SLE and control group was 55.1% and 20.7%, respectively.This difference was not described as significant due to the small number of studies, participants and heterogeneity.Results from our study did not reveal any difference between patients and healthy participants, as it was also published lately by Han et al. (46).Many authors proved a higher viral load in lupus of up to 40fold greater increase regardless of immunosuppressive therapy (38).On the other hand, some authors did not find any viral genome at all (51).Thus, the large discrepancy in the stratification of SLE patients and the selection of samples for DNA testing must be addressed.Unfortunately, previous reports refer to peripheral B cells, PBMC and even sera (7,38,46,48).Unlike them, our study tested viral DNA from free cell blood compartments depicting the amount of virus released during active replication and tendentially eliminating the amount that rests in latency.It was estimated that healthy carriers have approximately 1 to 50 infected cells per million leukocytes or 1 to 30 copies of EBV DNA per million leukocytes in whole blood.Thus, besides B cells or PBMC samples, any biopsy tissue may contain B lymphocytes and harbor amplifiable EBV DNA, too (52).
In our small group of positive EBV DNA isolates from SLE patients, we reported characteristics of EBNA1 and LMP1 gene sequences as an additional novelty.In need of more data on EBV gene sequences in lupus, reports on EBV gene sequences in other systemic autoimmune diseases are also extremely rare.They could be only found in RA or multiple sclerosis (MS) (19,53,54).In addition, those reports included only EBNA1 and EBNA2 variants.Similar to our study, previous authors did not prove the correlation of EBV variability with the risk of autoimmune disease (19,54).However, some assumptions exist that specific virus genetic variants contribute to autoimmune development (53).EBNA1 was described as an unusual immunogen and antigen (31).Gene sequences were assessed to identify potentially specific EBNA1 gene variability associated with those irregularities.Together with the previous reports from Serbian EBNA1 isolates, the dominance of P-Ala and P-Thr subtypes again confirmed one of the earliest hypotheses of European distribution in non-malignant diseases and geographically specific distribution (18) (55).On the other hand, EBNA1 variability patterns in this study did not at all correlate with previous findings in RA isolates (54).Interesting is that only P-Ala-sv-3 isolates lacked mucocutaneous presentation.This finding could be further investigated on a more significant number of isolates.
Although the LMP1 gene is significantly heterogeneous with greater variability than most other EBV genes, there is no literature on LMP1 polymorphisms in systemic autoimmune diseases.Even though this EBV oncogene is primarily significant in EBV-related cancers, its immunogenic potential, dependent on its variability, should be addressed.Thus, LMP1 variants with 30-bp deletion correlate with higher tumorigenic activity and lower immunogenic potential of EBV (15).One of those polymorphisms characterizes China1, identified in two of our isolates.According to some suggestions, China1-specific changes in Human leukocyte antigen (HLA) virus epitopes (within LMP1) could enable LMP1 expression, which leads to the inability to be recognized by LMP1-specific CD8+ cytotoxic T lymphocytes (56).It would be essential to follow whether patients with these isolates would develop some EBV-related cancers.With the identification of 4 different LMP1 variants, sequences from this study showed expected diversity for this geographic area.Due to the small sample and the diversity shown, specific LMP1 variants did not suggest any association with the clinical presentation of lupus.However, the largest group consisted of NC isolates, which should be investigated further because of the earlier hypothesis that the NC variant cannot inhibit T-cell proliferation and natural killer cytotoxicity due to amino acid substitutions responsible for immunosuppressive functions (56).
When it comes to study limitations, it would be more potent if the design was prospective with longitudinal follow-up of exposed and unexposed populations from the period without disease until the moment of SLE diagnosis.A follow-up period after the diagnosis of SLE would yield valuable results too.The same could be added to monitoring the influence of individual EBV gene polymorphisms on the potential development of the underlying disease and EBVassociated tumors.However, this kind of study would require large cohorts and would last long with an unpredictable outcome.There are more aspects related to other autoimmune diseases or other viruses and novel diagnostic tools that could be addressed also: lack of EBV sequences from other autoimmune diseases, serology data for other viruses, and droplet digital PCR for detection of low concentrations of EBV.Moreover, as immunological parameters were part of routine laboratory diagnosis and monitoring of SLE patients, these data for the study were collected from the patients' medical records.Without additional immunological testing of the healthy participants, the analysis included immunological parameters only in the SLE population.

Conclusion
This study obtained a comprehensive analysis of serological, molecular and sequence markers of EBV infection in SLE patients, providing additional evidence and new hypotheses about the role of EBV infection in SLE development, particularly anti-EBV immune response.The results revealed even 24 times higher possibility of having SLE if there is the presence of anti-EBV-EA(D) IgG antibodies.Additionally, higher titers of anti-EBNA1 IgG antibodies and lower titers of anti-EBV-CA IgG antibodies were also found to be independently associated with the presence of lupus.Identification of independent viral factors associated with specific lupus manifestations for the first time revealed anti-EBV-EA(D) IgG antibodies as an independent factor associated with lymphopenia and a higher titer of anti-CA IgG antibodies as an independent factor associated with alopecia.Finally, the novel data on EBV EBNA1 and LMP1 gene polymorphisms in lupus have been reported.

FIGURE 5
FIGURE 5 ± 13.62 years.More detailed demographic, clinical and laboratory characteristics of study cohorts are shown in Table 1.SLE patients

TABLE 1
Baseline characteristics of the study cohorts.

TABLE 2
Prevalence of EBV DNA, anti-EBV antibodies and active EBV infection.
*for the level of significance of 0.05, according to Chi-square test § or Fisher's exact test€.If the p value is in bold, it means that statistical significance is achieved.

TABLE 3
Factors associated with Systemic lupus erythematosus and its manifestations.

TABLE 4 EBNA1
C-terminal nucleotide and amino acid changes found in EBV isolates in SLE patients.

TABLE 5
Distribution of three LMP1 characteristics in EBV isolates in SLE patients.11-amino acid/33-bp repeats located between amino acids 250 and 308.The prototype B95-8 sequence has four perfect repeats with a disruption of 5 amino acids between the second and the third repeat.