Epstein–Barr virus in autoimmune diseases
E´ric Toussirot*MD, PhD
Department of Rheumatology, University Hospital Jean Minjoz, Besanc ¸on; and EA 3186 Agents Pathoge `nes et
Inflammation, University of Franche Comte ´, Besanc ¸on, France
Jean Roudier MD, PhD
INSERM U639, Universite ´ de la Me ´diterrane ´e, Marseille; and Rheumatology Department, La Conception
Teaching Hospital, APHM Marseille, France
Autoimmune diseases such as systemic lupus erythematosus (SLE), rheumatoid arthritis (RA),
and primary Sjo ¨gren’s syndrome (pSS) are complex disorders with a genetic background and
the involvement of environmental factors, including viruses. The Epstein–Barr virus (EBV) is
a plausible candidate for playing a role in the pathophysiology of these diseases. Both SLE and
RA are characterized by high titers of anti-EBV antibodies and impaired T-cell responses to
EBV antigens. Compared with normal subjects, elevated EBV load in peripheral blood has
been observed in SLE and RA. EBV DNA or RNA has been evidenced in target organs of RA
(synovium) or pSS (salivary glands). Finally, molecular mimicry has been demonstrated between
EBV proteins and self antigens in these three conditions. In addition, SLE, RA, and pSS are asso-
ciated with an increased risk of lymphoma with a potential role for EBV. The influence of new
and emergent treatments of these autoimmune diseases (biological therapies) on EBV load and
the course of latent EBV infection requires further studies.
Key words: autoimmune diseases; Epstein–Barr virus; molecular mimicry.
Autoimmune diseases such as systemic lupus erythematosus (SLE), rheumatoid
arthritis (RA), and primary Sjo ¨gren’s syndrome (pSS) are complex disorders with var-
ied clinical manifestations. The pathogenesis of these diseases is not fully understood
and involves genetic factors, as suggested by familial cases and concordance rates in
monozygotic twins. Beside the genetic component, environmental factors contribute
to the development or the occurrence of disease flares. Various environmental factors
have been suspected as potential etiologic factors in autoimmune diseases, including
* Corresponding author: Tel.: þ33 3 81 66 82 41; Fax: þ33 3 81 66 86 86.
E-mail address: email@example.com (E´. Toussirot).
1521-6942/$ - see front matter ª 2008 Elsevier Ltd. All rights reserved.
Best Practice & Research Clinical Rheumatology
Vol. 22, No. 5, pp. 883–896, 2008
available online at http://www.sciencedirect.com
physical agents, tobacco (for RA and SLE), drugs (in drug-induced SLE), and infectious
agents. Among the infectious triggering agents, viruses have long been suspected to
promote the development of autoimmune diseases. Many viruses have been proposed
as potential initiating factors for SLE, RA, and pSS. This chapter focuses on the evi-
dence associating a candidate virus, the Epstein–Barr virus, to SLE, RA, and pSS,
with special emphasis on the pathophysiological links with these diseases.1–3
EPSTEIN–BARR VIRUS BIOLOGY
The EBV (or human herpes virus 4) is a g herpes virus whose genome is a 172-kb dou-
ble strand of DNA. It infects around 98% of the world’s population. It initially repli-
cates in nasopharyngeal epithelial cells and B lymphocytes. To penetrate the cells,
the EBV envelope glycoprotein gp350 binds to the type 2 complement receptor or
CD21. After primary infection, the virus persists in latent form within memory B cells.
At this stage, latent cycle genes produce a limited set of EBV proteins: Epstein–Barr
nuclear antigen (EBNA), latent membrane antigen (LMA), and terminal proteins
(TPs). The virus reactivates occasionally, switching from latent to lytic cycle, with
the production and expression of transactivating proteins, structural viral proteins,
and envelope glycoproteins. The antigens typical for the lytic cycle are the viral capsid
antigen (VCA), the early antigen (EA), and the membrane antigen (MA) (Table 1). The
lytic and the latent cycles have distinctive serological profiles: antibodies to VCA, EA,
and envelope antigens are produced early during the infection, whereas antibodies to
EBNA develop later and persist throughout life. EBV exerts a wide range of immuno-
modulating effects, including inhibition of apoptosis, inhibition of the anti-EBV effects
on interferon g within B cells, and changes in the production of pro-inflammatory cy-
tokines such as tumor necrosis factor alpha (TNF-a), interleukin (IL)-1, and IL-6.4In
addition, EBV produces a viral cytokine that shares those properties of IL-10 that allow
the virus to limit the host immune response.5
To be a good candidate for a causal role in autoimmune diseases, a suspected virus
must have the following properties3,6:
? It must be ubiquitous for explaining the worldwide prevalence of the autoimmune
? Autoimmune diseases have a chronic evolution with periods of flares; thus, the virus
must persist within the body throughout life.
? It must be capable to alter or modify the host immune response such as the pro-
duction of antibodies which are commonly observed in SLE, pSS, and also RA.
? Other properties include the inhibition of apoptosis, the limitation of anergy or
? The virus infection must precede the development of clinical symptoms.
Table 1. Main antigens expressed by the Epstein–Barr virus during the latent cycle and the lytic cycle.
Latent cycleLytic cycle
EpsteineBarr nuclear antigen (EBNA)
Latent membrane antigen (LMA)
Terminal protein (TP)
Viral capsid antigen (VCA)
Early antigen (EA)
Membrane antigen (MA)
884E´. Toussirot and J. Roudier
EBV fulfills these different properties and is thus a relevant etiologic agent for SLE,
RA, and pSS.3,6
EPSTEIN–BARR VIRUS AND SYSTEMIC LUPUS ERYTHEMATOSUS
SLE is the prototype of autoimmune disease characterized by the presence of autoan-
tibodies. These antibodies are directed against various autoantigenic targets, such as
DNA, histones, Ro/La complex, etc. Numerous links have been documented between
EBV and SLE, and include serologic evidences, abnormal EBV load in peripheral blood
mononuclear cells (PBMC) of SLE patients, impaired control of EBV infection, and
molecular mimicry between EBV proteins and SLE autoantibodies.6–8
Prevalence of Epstein–Barr virus infection
A first line of evidence for the immune relationship between SLE and EBV is the prev-
alence of EBV infection, documented by the anti-EBV humoral response, in SLE pa-
tients compared with controls. However, this is difficult to investigate because EBV
has a ubiquitous repartition and, thus, the prevalence of the serologic response to
EBV is high in the normal population. To specifically evaluate the serologic response
against EBV antigens in SLE, a large population of patients and normal controls is
thus required.6,7One most incisive experiment was the assessment of EBV exposure
in pediatric patients under the age of 19. This study, by James et al, evaluated the se-
rologic response to the EBV VCA antigen (and also the presence of EBV DNA in
PBMC) and the results supported a strong association of EBV infection and SLE
[odds ratio (OR) 49.9, 95% confidence interval (CI) 9.3–1025, p < 10?12].9Similar re-
sults were obtained in a large collection of adult SLE patients: 192 patients with SLE
were compared with 392 age- and sex-matched controls for EBV VCA serology and
common herpes viruses.10This study showed that 99.5% of adults with SLE had sero-
converted to EBV, compared with 95% of normal subjects, giving an OR of 9.35
(p ¼ 0.014). This difference was not found for the other herpes virus.10
Another study performed in Taiwan evaluated the presence of IgA anti-EBNA an-
tibodies (which are a more reliable marker for detecting disease related antibodies)
and also EBV DNase antibodies: both were detected more frequently in SLE patients
than normal subjects.11
Another study examined the influence of age, race, and CTLA-4 polymorphism on
the anti-EBV humoral response in SLE patients.12IgA VCA antibodies were deter-
mined in a cohort of 230 SLE patients and compared with 276 healthy controls. In
African–American lupus, EBV IgA antibodies were found more frequently than
in race-matched controls (OR 5.6, 95% CI: 3.0–10.6); an association of EBV IgA anti-
bodies with lupus was also found in older white patients. A polymorphism of CTLA-4
promoter (?1661 A/G) was associated with anti-EBV IgA response in SLE patients.
These data suggest the influence of age, race, and CTLA-4 genotype on EBV-seropre-
valence in SLE patients, particularly in African–American or older white patients.12
The evidence for higher titers of anti-EBV antibodies in lupus patients has been de-
scribed in several studies (reviewed in ref.7). It was also demonstrated that patients
with SLE and RA possess antibodies that bind to various peptides of the EBNA-1 pro-
tein.13It was of particular interest to note that lupus patients had higher levels of an-
tibodies against three peptides of EBNA-1 than normal subjects. Two of these peptides
Epstein–Barr virus in autoimmune diseases 885
contain a glycine–arginine repeat motif, an important sequence in potential mecha-
nisms of autoimmunity.
Epstein–Barr virus load in patients with systemic lupus erythematosus
Several groups have shown that EBV load is 10- to 100-fold higher in the peripheral
blood of SLE patients than in controls.6This was first demonstrated in 32 pediatric
patients, all of whom had positive EBV DNA in their PBMC compared with 23 of
32 matched controls (OR > 10, p <0.002).9In another study, 82% of adult SLE pa-
tients were positive for EBV DNA in PBMC compared with 49% of controls.14In
SLE patients from Taiwan, EBV DNA was amplifiable in the sera of 41.9% SLE patients
and 3.25% controls (p < 0.05).11Another study examined the frequency of infected
peripheral B cells in SLE.15This study found an around 10-fold increase in the frequen-
cies of EBV-infected cells in 35 SLE patients compared with 44 controls. This high fre-
quency of infected cells was associated with the occurrence of disease flares. In
addition, abnormal expression of viral lytic (BZLF1, 39% of SLE patients) and latent
(LMP-1, 29% of SLE patients; LMP-2a, 18% of SLE patients) genes was detected in
the blood of SLE patients.15In another study, EBV load in PBMC was found to be
more than 15-fold in patients with SLE than in healthy controls. These results were
not influenced by immunosuppressive drugs or SLE disease activity.16Finally, Kang
et al found that SLE patients had a 40-fold increased in EBV load compared with
T -cell response to the Epstein–Barr virus in systemic lupus erythematosus
It has been demonstrated that patients with SLE have an impaired EBV-specific T-cell
response.18A more recent study evaluated specific T-cell responses to EBV antigens
in SLE.17By means of intracellular staining, it was found that the frequencies of
EBV-specific CD69 þ CD4 þT cells producing interferon (IFN)-g were significantly
higher in SLE than in healthy controls; there was a tendency for similar results for
CD69 þ CD4 þ T
CD69 þ CD8 þ T cells producing IFN-g were lower in patients with SLE than in con-
trols. A negative correlation was found between EBV load and CD69 þ CD4 þ T cells
producing IFN-g and a positive correlation with the frequency of CD69 þ CD8þ T
cells producing IFN-g. Another study also found a functional impairment of
CD8þ EBV- specific T cells in SLE: their functional capacity to express IFN-g was de-
creased.19Together, these studies suggest that patients with SLE have defective latent
EBV control resulting from impaired T-cell responses.
cells producingTNF-a. Bycontrast,thefrequenciesof
Molecular mimicry between lupus autoantigens and the
Beside these serologic and functional aspects of the association between SLE and EBV,
other evidence came from the findings of molecular investigations of SLE autoantigens
and certain EBV proteins.6,8Cross-reactive epitopes from EBV were identified in pa-
tients with infectious mononucleosis and then in patients with pSS and SLE. These pep-
tides contain glycine–alanine-rich sequences and cross-react with nuclear antigens.20
Indeed, SLE is characterized by the production of autoantibodies targeting different
specific epitopes of nuclear proteins, such as Sm, Ro, or La complex. The Sm proteins
886E´. Toussirot and J. Roudier
are polypeptide components of the spliceosome involved in the splicing of nuclear
RNAs to messenger RNAs. The major Sm proteins, which are antigenic in SLE, include
Sm B/B0, D1, D2, and D3.8One major humoral immune target of the Sm B/B0protein
is a proline-rich repeated motif PPPGMRPP. EBNA-1 contains a similar sequence
PPPGRRP. This EBV sequence was commonly recognized by antibodies from lupus
patients, but not by anti-EBV-positive normal subjects. Immunization with the
PPPGMRPP motif can lead to lupus autoimmunity and clinical features of SLE.21Immu-
nization with the EBNA-1 PPPGRRP peptide can also lead to lupus autoimmunity. An-
other Sm antigenic target, Sm D1, located between amino acids 95 and 119, has a very
similar cross-reactive sequence with amino acids 35–58 of EBNA-1. Lupus-patient an-
tibodies directed against the Sm D1 are capable of binding the EBNA-1 35–58 peptide
and, conversely, immunization of animals with the 35–58 EBNA-1 peptide induces the
production of antibodies that can react with EBNA-1 and Sm D1 95–119.8,22
Molecular mimicry from a second autoantigen system has been described in SLE. It
concerns the anti-Ro antibodies, which are observed in up to 40% of lupus patients. A
maintarget ofRoprotein wasidentified andlocatedataminoacidposition169–180.Im-
munization of animals with the 169–180 sequence of Ro leads to antibodies against the
peptide of immunization, but also to clinical features such renal insufficiency and leuco-
penia. Purified 160–180 anti-Ro antibodies also bind to a cross-reactive sequence of
EBNA-1. Immunization of animals with this cross-reactive sequence of EBNA-1 leads
to autoimmunity and clinical features of lupus.8,23However, this EBNA-1 sequence
tif, whereas the cross-reactive epitope of EBNA-1 is GGSGSGPRHRDGVRR, located in
the 58–72 region of EBNA-1. Animals immunized with Ro169–180 or EBNA-1 58–72
developed both anti-EBNA-1 and anti-Ro antibodies.8
Together, these data support clear molecular links between anti-viral EBNA-1 anti-
bodies and lupus anti-Sm and anti-Ro autoantibodies (Table 2).
Finally, there is clinical evidence to associate EBV with SLE based on clinical descrip-
tions of the development or exacerbation of lupus after EBV infection.24,25
EPSTEIN–BARR VIRUS AND RHEUMATOID ARTHRITIS
RA is a chronic inflammatory joint disease responsible for pain and the development of
bone erosions and cartilage destruction. The pathophysiology of RA is complex and
includes genetic factors, hormonal status, psychological factors, and immune deregu-
lation, as well as environmental factors. The EBV fulfills the different characteristics
to be a plausible candidate for a causal role in RA (see above).
Table 2. Molecular mimicry between lupus associated autoantigens and Epstein–Barr virus proteins.
(amino acid position)
Amino acid sequence EBV antigen
(amino acid position)
Amino acid sequence
Sm D1 (95e119)
GGSGSGPRHRDGVRR Ro (169e180)EBNA-1 (58e72)
Epstein–Barr virus in autoimmune diseases 887
The evidence linking EBV to RA includes serologic data, cell-mediated control of
EBV, cross-reactivity between EBV proteins and human self proteins, the identification
of the EBV genome in human synovial membrane, and the cell-mediated response to
EBV proteins within the joint.26
Humoral response to Epstein–Barr virus antigens in rheumatoid arthritis
Anti-EBV titers are higher in patients with RA than in controls. Historically, serum re-
activity to the nuclei of EBV-infected B cells was described in patients with RA. This
target B-cell antigen was termed RANA (rheumatoid arthritis nuclear antigen); reac-
tivity to RANA was observed in up to 67% of patients with RA compared with 8% of
controls.3In fact, RANA was subsequently identified as EBNA-1, which contains a re-
peated glycine–alanine sequence, a dominant epitope during infectious mononucleosis
and also in RA.
Cellular immune response to Epstein–Barr virus in rheumatoid arthritis
As was observed in SLE, patients with RA have a poor control of EBV infection.4In-
deed, cultured B cells infected with EBV induce immortal lymphoblastoid cell lines.
The rate of cultured B cells that transform into lymphoblastoid lines is higher in
RA. A study using the human leukocyte antigen (HLA)-peptide tetramer technique
showed that EBV-specific CD8þ T cells were directed predominantly to lytic cycle
peptides and that IFN-g production by these T cells was significantly reduced in RA
compared with controls.27
Molecular mimicry between Epstein–Barr virus antigens and self proteins
There are several examples of molecular resemblance between EBV and self antigens
in RA. Several sequences are found in EBNA-6 and HLA-DQ*030228: a six-amino-acid
sequence is shared by EBV reading frame BPLF1 and Mycobacterium tuberculosis 65 kD
heat shock protein, a protein that induced arthritis in the animal model of adjuvant
arthritis29; glycine–alanine repeat sequences found in EBNA-1 are also expressed in
cytoskeleton proteins, such as cytokeratin and type 2 collagen30; the RA genetic sus-
ceptibility sequence QKRAA located in HLA-DRB1*0401, a susceptibility factor for
RA, shares a similar sequence in the gp110 EBV glycoprotein, a lytic cycle antigen ex-
pressed in the endoplasmic reticulum and on the cell membrane and the viral capsid.31
B and T cells cross-react to the QKRAA in gp110 and HLA-DRB1*0401.31In RA, we
found decreased T-cell response to gp110; this response was correlated with inflam-
mation parameters and systemic involvement (Table 3).32
Table 3. Molecular mimicry between human self antigens and Epstein–Barr virus proteins.
Self antigens EpsteineBarr virus antigen
Cytokeratin (type II collagen)
QKRAA sequence in HLA-DRB1*04101
Glycineealanine repeat sequences of EBNA-1
QKRAA sequence of gp110
888E´. Toussirot and J. Roudier
Cell-mediated response to Epstein–Barr virus antigens within the joint
CD8 þ T cell clones that recognized BZLF1 and BMLF1 proteins of EBV (BZLF1 and
BMLF1 reading frames encode for transactivators produced during the lytic cycle)
were isolated from the synovial joint fluid in two patients with RA. This response
was stronger in the joint than in peripheral blood.33We also found a T-cell response
to gp110 protein in the synovial fluid from patients with RA that was higher than sy-
novial fluid T-cell reactivity to the gp110 of patients with other inflammatory rheu-
matic disease.34These data support the existence of a cell-mediated response to
EBV lytic cycle antigens within the joints.
Presence of the Epstein–Barr virus in the synovial membrane
Different techniques have been used to demonstrate the presence of EBV in sy-
novial fluid or tissue from RA patients, with limited success (reviewed in ref.26). Im-
munohistochemical studies using antibodies to EBV labeled a small proportion of
synovial lymphocytes. The same problem was obtained with in situ hybridization, giv-
ing a positive response in between 8 and 62% of RA patients. In polymerase chain
reaction (PCR) studies, patients with RA were significantly more likely than controls
to have EBV DNA or RNA in PBMC, saliva, synovial fluid or synovial membrane;
positive results ranged from 6 to 67%.26A recent study analyzed the relationship
between the presence of EBV genome in the synovium and the genetic predisposi-
tion to RA: Individuals who had both an HLA-DR4 susceptibility allele and a positive
test for EBV DNA in synovial tissue had a 41-fold increase risk for RA compared
with subjects who had a negative test for EBV DNA and none of the RA suscepti-
Epstein–Barr virus load in patients with rheumatoid arthritis
The percentage of EBV-infected B cells is higher in RA than in healthy subjects. In keep-
ing with these findings, the EBV burden in patients with RA has been investigated in
two studies35,36, both of which showed that EBV load was about 10-fold higher in
RA patients than in controls.
Epstein–Barr virus and anti-citrullinated antibodies
in rheumatoid arthritis
Antibodies directed against cyclic citrullinated peptides (CCP) are highly specific for
RA. EBV antigens can undergo post-translational citrullination, thus becoming targets
for anti-CCP. Indeed, patients with RA have antibodies to citrullinated EBV peptides in
their serum. These antibodies were found in 45% of RA patients compared to 5% of
healthy controls or non-RA inflammatory diseases.37The viral anti-citrullinated
antibody correlated with anti-CCP titers. It was also demonstrated that these EBV-
citrullinated antibodies reacted with citrullinated peptides that are antigen targets in
RA such as fibrin.38Thus, EBV may induce immune response against citrullinated
Epstein–Barr virus in autoimmune diseases 889
EPSTEIN–BARR VIRUS AND PRIMARY SJO¨GREN’S SYNDROME
pSS is a common autoimmune disease characterized by autoimmunity against exocrine
glands. Lymphocyte cells infiltrate salivary and lacrimal glands, leading to altered func-
tion of the glands and a sicca syndrome. pSS is also characterized by a systemic pro-
duction of autoantibodies. The pathogenesis of the disease remains unclear, but there
is some evidence for the participation of viruses, including EBV. Indeed, EBV fulfills in
part the different conditions to be a likely candidate in the pathogenesis of pSS.1,28:
? One evident reason is that EBV replicates in the oropharynx during the primary
infection; it has normal site of latency in salivary and lacrimal glands.
? EBV is known to induce strong T-cell responses; thus, immune recognition of the
virus at its site of latency or activation might contribute to exocrine gland damage.
? EBV is a B-cell polyclonal activator that can induce the production of various autoan-
tibodies similar to those found in patients with pSS. pSS predisposes to the develop-
ment of polyclonal and monoclonal lymphoproliferate disorders (LPD). This is also
observed in allograft recipients, who are at increased risk for EBV-associated LPD.28
? Antibodies directed against SS-B proteins are found in the serum of patients with
pSS. They precipitate a cytoplasmic protein, which is complexed to EBV-encoded
small RNAs (EBER 1 and EBER 2).28
? There are some descriptions of patients who developed pSS after EBV acute
One main argument for the association between pSS and EBV is the presence of
EBV DNA in saliva and/or salivary gland biopsies more frequently in patients with
pSS than in healthy controls. However, there are contradictory results in the litera-
ture; these can be explained by the heterogeneity of the patient population and also
by technical reasons. In fact, different techniques were used to evidence EBV DNA
in salivary tissue (immunohistochemistry method, in situ hybridization, and PCR).
One group found a high prevalence of EBV in patients with pSS compared with healthy
controls. Using the immunohistochemistry method, Fox et al found cytoplasmic stain-
ing of epithelial cells with monoclonal antibody against EBV-EA antigen in 57% of pa-
tients with pSS compared with 0% of healthy controls; they confirmed their results
using the PCR method.40Using in situ hybridization, Mariette et al found EBV DNA
in 50% of patients with pSS and 8% of healthy controls. EBV DNA staining was ob-
served in the epithelial cells of salivary glands. Using PCR, EBV DNA was detected
in salivary-gland specimens from 86% of patients with pSS and 29% of controls sub-
jects.41Another study, performed by Wen et al, found similar results with positive sal-
ivary gland biopsies for EBV DNA by in situ hybridization in 57% of patients with pSS
compared with 0% in normal subjects.42Conversely, there are studies with negative
results. One study by Venables et al found no differences between patients with pSS
and normal controls for the presence of EBV in salivary glands, either by in situ hybrid-
ization or immunohistochemical method.43Collectively, EBV genes or proteins are de-
tected in the salivary (and the lacrimal) glands of around 50% of patients with pSS, but
these results are not confirmed by all authors.
highly expressed in tear and saliva and a-fodrin is a cytoskeleton protein. Serum from
patients with pSS, but not from those with other autoimmune diseases or from healthy
controls, reacted with both lipocalin and a-fodrin, and also with an EBV protein.44
890 E´. Toussirot and J. Roudier
It was also recently demonstrated that the saliva of patients with pSS can activate
EBV. In this study, the influence of saliva was examined on an EBV-negative cell line
transfected with BZLF1, a transactivating EBV gene promoter. The results showed
that saliva of pSS patients exerts a significant effect on EBV gene promoter.45
Antibodies against EBV antigens are elevated in the serum of patients with pSS.28
Primary EBV infection is generally asymptomatic or subclinical in immunocompetent
subjects, but can cause a self-limiting disease in young adult, i.e. infectious mononucle-
osis. However, EBV is associated with the development of malignant diseases (Burkitt
lymphoma, nasopharyngeal carcinoma). In addition, in immunocompromised individ-
uals (hereditary immunodeficiency, transplant recipients or patients with acquired
immunodeficiency syndrome [AIDS]), EBV might result in the development of severe
or lethal LPD. The described EBV-associated LPDs range from infectious-mononucle-
osis-like disease to malignant lymphoma. To date, there is no specific treatment for
latent EBV infection, but the understanding of the main pathogenic mechanisms
involved during the viral infection provided the development of potential treatments
of EBV-related disorders.46
The therapeutic approach against EBV-related disorders includes antiviral agents,
immune modulators, and chemotherapeutic drugs (Table 4). Antiviral agents such as
aciclovir, ganciclovir, and vidarabine can be used in cases of EBV replication in infected
cells, or for AIDS patients who have developed oral hairy leukoplakia. Anti-EBV cyto-
toxic T cells derived from donors have been used in cases of transplant recipients who
Table 4. Specific therapeutic procedure for Epstein–Barr-virus-associated conditions.
EBV-associated conditionsMain pathogenic mechanism Therapeutic strategies
Proliferation of latently
Surgery, chemotherapy, irradiation
Primary or hereditary
Proliferation of latently
EBV-infected cells related to
hematopoietic stem cell
transplantation, EBV-specific CTL
syndrome: transplant recipients,
Proliferation of latent
EBV-infected cells related
Cessation or decrease of
immunosuppressive drugs, antiviral
agents Monoclonal antibodies (anti-
Oral hairy leukoplakia (in AIDS)Replication of EBV in
Aciclovir, ganciclovir, vidarabine
Severe infectious mononucleosis
Immune response against
plasmapheresis, bone marrow
antibodies (anti-CD20, anti-TNF-a)
AIDS, acquired immunodeficiency syndrome; CTL, cytotoxic T cells; EBV, Epstein–Barr virus.
Epstein–Barr virus in autoimmune diseases891
developed EBV-related LPD or increased EBV DNA in peripheral blood following
transplantation. Alternative and recent treatments include monoclonal antibodies to
CD21 (the EBV receptor), to CD24 (pan-B cell antigen), and to CD20 (a B-cell anti-
gen); anti-IL-6 monoclonal antibodies have also been used. Severe infectious mononu-
cleosis with hemophagocytic lymphohistiocytosis can occur in patients with hereditary
immunodeficiency and may be fatal. In this situation, treatment with etoposide has
been shown to be very effective. In hereditary syndromes with immunodeficiency,
anti-CD20 monoclonal antibody has also been used – in association with antiviral
agents or immunoglobulins – to eliminate EBV-infected B cells. Finally, anti TNF-
a agents have been used in one case of hemophagocytic lymphohistiocytosis that
occurred in a patient with possible X-linked lymproliferative syndrome.47
Some cases of LPD have been reported in patients with autoimmune diseases. One
case of infectious-mononucleosis-like disease was diagnosed in a patient with RA
receiving low-dose methotrexate. In this case, a favorable outcome was obtained
with symptomatic treatment.48However, more severe disorders, such as malignant
lymphoma, were observed with methotrexate therapy in RA.49In addition, some cases
of hemophagocytic syndrome associated to EBV have been described in SLE. These
clinical features, with an often fatal course, have been associated to reactivation of
aggressive EBV infection.50
RISK OF LYMPHOMA IN AUTOIMMUNE DISEASES AND
RELATIONSHIPS WITH THE EPSTEIN–BARR VIRUS
There is an increased risk of lymphoma in RA, particularly in male patients or those
with highly active disease. The EBV has been associated with these lymphomas in
15–30% of cases. The effect of RA treatments on lymphoma risk (methotrexate or
anti-TNF-a agents) is still under debate. The exact role of EBV in the development
of these lymphomas is unclear. However, it is conceivable that impairment in cell-
mediated immunity might lead to the uncontrolled proliferation of EBV-infected
B cells. RA is associated with a high EBV burden in peripheral blood. Combined
with the immunosuppression related to the disease itself, or with its treatments,
this can promote the development of B-cell malignancies.49
pSS is associated with an increased risk of non-Hodgkin lymphoma.50Lymphomas
occurring in patients with pSS are mainly low-grade marginal-zone lymphomas.51The
association of these lymphomas with EBV is not clear. LMP protein and EBER RNAs
of EBV were not detected in a series of 16 non-Hodgkin lymphomas from patients
with pSS.51In situ hybridization and PCR were used to detect EBV RNA in the tumors
had positive staining for EBV sequences in the tumor specimens but no association was
found between EBV-positive salivary glands and EBV-positive lymphoma.52
Systemic lupus erythematosus is associated with an increased risk of B-cell malig-
nancies, mainly non-Hodgkin lymphoma.53Aggressive, diffuse, large B-cell lymphomas
predominates in patients with SLE. Associated risk factors for lymphoma in SLE were
female sex, sicca symptoms, and lung involvement, but not immunosuppressive
drugs.54As was described in RA, more intense disease activity and/or long-lasting dis-
ease might be indications of a higher risk of lymphoma development in SLE.55In an
international study of cancer in SLE, the standardized incidence ratio for non-Hodgkin
lymphoma was 3.64 (95% CI 2.63–4.93).56EBV RNAs have occasionally been detected
in lymphoma biopsy specimens by in situ hybrization.57
892 E´. Toussirot and J. Roudier
IMPACT OF THERAPIES FOR AUTOIMMUNE DISEASES ON
EBV VIRAL LOAD
EBV load is higher in patients with RA, SLE, and pSS than in healthy controls. One rel-
evant question is the influence of specific drugs of these autoimmune diseases on the
natural course of EBV. Conventional treatments such as methotrexate do not influ-
ence EBV load in RA as well as infliximab, an anti-TNF-a monoclonal antibody.36
The evolution of the EBV viral load in RA patients under anti-TNF-a (infliximab or eta-
nercept) has been examined in a 5-year follow-up study on 140 patients. The general
findings of this study were that EBV load was stable over time under methotrexate
and/or TNF-a suppressors.58However, this question has not been specifically exam-
ined with the use of other biological therapies of RA.
enced by immunosuppressive drugs (methotrexate, azathioprine, ciclosporin,
mycofenilate mofitile, or cyclophosphamide).15As stated above, the specific impact of
emergent new drugs for SLE (anti-CD20, anti-CD40) on EBV load is unknown.
Data on EBV load in patients with pSS are not available. Thus, the influence of new
biological treatments (anti-CD20) on EBV viral burden is unknown.
The relationships between EBVand autoimmune diseases are complex and involve sev-
eral mechanisms. In SLE, RA, and pSS, the molecular resemblance between EBV pro-
teins and self antigens provides a plausible and fascinating mechanism of disease
induction. A proposed scenario was the interrelation between genetic predisposition
and EBV infection, a cross-reaction between self antigens and EBV viral proteins, the
development of autoantibodies (anti-Sm or anti-Ro antibodies in SLE, anti-CCP anti-
bodies in RA), an epitope-spreading process that can lead to the development of
additional non-cross-reactive autoepitopes, and finally the clinical manifestations.8
EBV has multiple impacts on host immunity and can adapt to its environment. For
instance, EBV produces an IL-10-like cytokine, it can modulate IL-6 release, it induces
the production of TNF-a by infected B cells, and it can inhibit apoptosis of infected
B cells by the production of a Bcl-2 homolog. All these biological properties of EBV
are relevant to, and can contribute to, the pathophysiology of autoimmune disorders.
? EBV is a common etiologic factor for SLE, RA, and pSS. A common mechanism
for the intervention of EBV in the pathophysiology of these diseases is a molec-
ular mimicry between EBV antigens and self proteins.
? These autoimmune diseases are both characterized by high titers of anti-EBV
antibodies, a high proportion of EBV-infected cells with elevated viral load in
peripheral blood compared to controls and an increased risk of lymphoma.
? EBV has been detected in the synovium of patients with RA and in the salivary
glands of patients with pSS.
? Severe EBV-related illness or the development of LPD can occur in patients
with autoimmune diseases as a consequence of EBV infection or reactivation.
Epstein–Barr virus in autoimmune diseases893
1. James JA, Harley JB & Scofield RH. Role of viruses in systemic lupus erythematosus and Sjo ¨gren
syndrome. Current Opinion in Rheumatology 2001; 13: 370–376.
2. James JA, Harley JB & Scofield RH. Epstein-Barr virus and systemic lupus erythematosus. Current Opinion
in Rheumatology 2006; 18: 462–467.
3. Vaughan JH. The Epstein-Barr virus in autoimmunity. Springer Seminars in Immunopathology 1995; 17:
4. Lotz M & Roudier J. Epstein-Barr virus and rheumatoid arthritis. Cellular and molecular aspects. Rheu-
matology International 1989; 9: 147–152.
5. Kanegane H, Wakiguchi H, Kanegane C et al. Viral interleukin-10 in chronic active Epstein-Barr virus
infection. The Journal of Infectious Diseases 1997; 176: 254–257.
6. Harley JB, Harley ITW, Guthridge JM et al. The curiously suspicious: a role for Epstein-Barr virus in
lupus. Lupus 2006; 15: 768–777.
7. McClain MT, Harley B & James JA. The role of Epstein-Barr virus in systemic lupus erythematosus. Fron-
tiers in Bioscience 2001; 6: 137–147.
*8. Poole BD, Scofield RH, Harley JB et al. Epstein-Barr virus and molecular mimicry in systemic lupus
erythematosus. Autoimmunity 2006; 39: 63–70.
9. James JA, Kaufman KM, Farris AD et al. An increased prevalence of Epstein-Barr virus infection in young
patients suggests a possible etiology for systemic lupus erythematosus. The Journal of Clinical Investigation
1997; 100: 3019–3026.
*10. James JA, Neas BN, Moser KL et al. Systemic lupus erythematosus in adults is associated with previous
Epstein-Barr virus exposure. Arthritis and Rheumatism 2001; 44: 1122–1146.
11. Lu JJY, Chen DY, Hsich CWet al. Association of Epstein-Barr virus infection with systemic lupus eythe-
matosus in Taiwan. Lupus 2007; 16: 168–175.
12. Parks CG, Cooper GS, Hudson LL et al. Association of Epstein-Barr virus with systemic lupus eryth-
ematosus. Effect modification by race, age and cytotoxic lymphocyte - associated antigen 4 genotype.
Arthritis and Rheumatism 2005; 52: 1148–1159.
*13. Petersen J, Rhodes G, Roudier J et al. Altered immune response to glycine-rich sequences of Epstein-
Barr nuclear antigen-1 in patients with rheumatoid arthritis and systemic lupus erythematosus. Arthritis
and Rheumatism 1990; 33: 993–1000.
14. Yu SF, Wu HC, Tsai WC et al. Detecting Epstein-Barr virus DNA from peripheral blood mononuclear
cells in adult patients with systemic lupus erythematosus in Taiwan. Medical Microbiology and Immunology
2005; 194: 15–120.
? A B- and T-cell cross-reaction between EBV antigen (gp110) and self protein
(HLA-DRB1*0401) has been demonstrated in RA, whereas in other autoim-
mune diseases only cross-reactive autoantibodies have been described. Thus,
in SLE and pSS, the demonstration of a T-cell reactivity against similar se-
quences on EBV and self antigens is required.
? EBV load is elevated in autoimmune diseases but data are lacking on the
changes of this viral burden under biologics.
? The risk of lymphoma is higher in autoimmune diseases than in the control
population. In immunosuppressed transplant recipients, the elevation of EBV
load can predict the emergence of lymphoma. Perhaps monitoring EBV load
in patients receiving biologics for autoimmune disease will be useful for evalu-
ating the risk of LPD.59
894 E´. Toussirot and J. Roudier
15. Gross AJ, Hochberg D, Rand Wet al. EBVand systemic lupus erythematosus: a new perspective. Journal
of Immunology 2005; 174: 6599–6607.
*16. Moon UY, Park SJ, Oh STet al. Patients with systemic lupus erythematosus have abnormally elevated
Epstein-Barr virus load in blood. Arthritis Research & Therapy 2004; 6: R295–R302.
17. Kang I, Quan T, Nolasco H et al. Defective control of latent Epstein-Barr virus infection in systemic
lupus erythematosus. Journal of Immunology 2004; 172: 1287–1294.
18. Tsokos G, Magrath IT & Balow JE. Epstein-Barr virus infection induces normal B cell responses but de-
fective suppressor T cell responses in patients with systemic lupus erythematosus. Journal of Immunology
1983; 131: 1797–1804.
19. Berner BR, Tary-Lehman M, Yonkers NL et al. Phenotypic and functional analysis of EBV specific mem-
ory CD8 cells in SLE. Cellular Immunology 2005; 235: 29–38.
20. Vaughan JH, Nguyen MD, Valbrecht JR et al. Epstein-Barr virus induced autoimmune responses. The
Journal of Clinical Investigation 1995; 95: 1316–1327.
21. James JA & Harley JB. Lupus humoral autoimmunity after short peptide immunization. Annals of the
New York Academy of Sciences 1997; 815: 124–127.
22. Riemekasten G, Marell J, Trebelijahr G et al. A novel epitope on the C terminus of SmD1 is recognized
by the majority of sera from patients with systemic lupus erythematosus. The Journal of Clinical Investi-
gation 1998; 102: 754–763.
23. Mc Clain MT, Heinlen LD, Dennis GJ et al. Early events in lupus humoral autoimmunity suggest initiation
through molecular mimicry. Nature Medicine 2005; 11: 85–89.
24. Verdolini R, Bugatti R, Giangacomi M et al. Systemic lupus eythematosus induced by Epstein-Barr virus
infection. The British Journal of Dermatology 2002; 146: 877–881.
25. Kasapcopur O, Ergul Y, Kutlug S et al. Systemic lupus erythematosus due to Epstein-Barr virus or
Epstein-Barr virus infection provocating acute exacerbation of systemic lupus erythematosus? Clinical
Rheumatology 2006; 26: 765–767.
26. Toussirot E & Roudier J. Pathophysiological links between rheumatoid arthritis and the Epstein-Barr
virus: an update. Joint Bone Spine 2007 Oct; 74(5): 418–426.
27. Klatt T, Ouyang Q, Flad Tet al. Expansion of peripheral CD8 þ CD28 - T cells in response to Epstein-
Barr virus in patients with rheumatoid arthritis. The Journal of Rheumatology 2005; 32: 239–251.
28. Fox RI, Luppi M, Pisa P et al. Potential role of Epstein-Barr virus in Sjo ¨gren’s syndrome and rheumatoid
arthritis. The Journal of Rheumatology 1992; 19(Suppl. 32): 18–24.
29. Van Eden W, Hogervorst J, Hensen E et al. A cartilage mimicking epitope on a 65 kD mycobacterial heat
shock protein: adjuvant arthritis as amodel for human rheumatoid arthritis. Current Topics in Microbiology
and Immunology 1989; 145: 27–43.
30. Baboonian C, Venables PJW, Willaim DG et al. Cross reaction of antibodies to a glycine alanine repeat
sequence of Epstein-Barr virus nuclear antigen 1 with collagen, cytokeratin and actin. Annals of the Rheu-
matic Diseases 1991; 50: 772–775.
*31. Roudier J, Petersen J, Rhodes G et al. Susceptibility to rheumatoid arthritis maps to a T cell epitope
shared by HLA-Dw4 DR beta 1 chain and the Epstein-Barr virus glycoprotein gp110. Proceedings of
the National Academy of Sciences of the United States of America 1989; 86: 5104–5108.
32. Toussirot E, Wendling D, Tiberghien P et al. Decreased T cell precursor frequencies to Epstein-Barr
virus glycoprotein gp110 in peripheral blood correlate with disease activity and severity in patients
with rheumatoid arthritis. Annals of the Rheumatic Diseases 2000; 59: 533–538.
*33. Scotet E, David-Ameline J, Peyrat MA et al. T cell response to Epstein-Barr virus transactivators in
chronic rheumatoid arthritis. The Journal of Experimental Medicine 1996; 184: 1791–1800.
34. Toussirot E, Wendling D, Luka J et al. Increased frequency of T cells specific for Epstein-Barr virus gly-
coprotein gp110 in the synovial fluid of patients with rheumatoid arthritis. Arthritis Research & Therapy
2003; 5: S25.
*35. Saal JG, Krimmel M, Steidle M et al. Synovial Epstein-Barr virus infection increases the riskof rheumatoid
arthritis in individuals with the shared HLA-DR4 epitope. Arthritis and Rheumatism 1999; 42: 1485–1496.
36. Balandraud N, Meynard JB, Auger I et al. Epstein-Barr virus load in the peripheral blood of patients with
rheumatoid arthritis. Accurate quantification using real-time polymerase chain reaction. Arthritis and
Rheumatism 2003; 48: 1223–1228.
37. Anzilotti C, Merlini G, Pratesi F et al. Antibodies to viral citrullinated peptide in rheumatoid arthritis.
The Journal of Rheumatology 2006; 33: 647–651.
Epstein–Barr virus in autoimmune diseases 895
*38. Pratesi F, Tommasi C, Anzilotti C et al. Deiminated Epstein-Barr virus nuclear antigen 1 is a target of
anti-citrullinated protein antibodies in rheumatoid arthritis. Arthritis and Rheumatism 2006; 54: 733–741.
39. Wittingham S, McNeilage J & Mackay IR. Primary Sjo ¨gren’s syndrome after infectious mononucleosis.
Annals of Internal Medicine 1985; 102: 490–493.
40. Fox RI, Pearson G & Vaughan JH. Detection of Epstein-Barr virus associated antigens and DNA in sal-
ivary gland biopsies from patients with Sjo ¨gren’s syndrome. Journal of Immunology 1886; 137: 3162–3168.
41. Mariette X, Gozlan J, Clerc D et al. Detection of Epstein-Barr virus DNA by in situ hybridization and
polymerase chain reaction in salivary gland biopsy specimens from patients with Sjo ¨gren’s syndrome.
The American Journal of Medicine 1991; 90: 286–294.
42. Wen S, Shimizu N, Yoshimaya H et al. Association of Epstein-Barr virus with Sjo ¨gren’s syndrome: dif-
ferential EBVexpression between epithelial cells and lymphocytes in salivay glands. The American Journal
of Pathology 1996; 149: 1511–1517.
43. Venables PJW, Teo CG, Baboonian C et al. Persistence of Epstein-Barr virus in salivary gland biopsies
from healthy individuals and patients with Sjo ¨gren’s syndrome. Clinical and Experimental Immunology
1989; 75: 359–364.
44. Navone R, Lunardi C, Gerli R et al. Identification of tear lipocalin as a novel autoantigen target in
Sjo ¨gren’s syndrome. Journal of Autoimmunity 2005; 25: 229–234.
45. Nagata Y, Inoue H, Yamada K et al. Activation of Epstein-Barr virus by saliva from Sjo ¨gren’s syndrome
patients. Immunology 2004; 111: 223–229.
46. Okano M. The evolving therapeutic approaches for Epstein-Barr virus infection in immunocompetent
and immunocompromized individuals. Current Drug Targets. Immune, Endocrine and Metabolic Diseases
2003; 3: 137–142.
*47. Okano M & Gross TG. Advanced therapeutic and prophylactic strategies for Epstein-Barr virus infec-
tion in immunocompromized patients. Expert Review of Anti-Infective Therapy 2007; 5: 403–413.
48. Shoda H, Nakazaki K, Izutsu Ket al. Epstein-Barr virus associated mononucleosis caused by weekly low
dose methotrexate therapy in a rheumatoid arthritis patient. Scandinavian Journal of Rheumatology 2006;
49. Dawson TM, Starkebaum G, Wood BL et al. Epstein-Barr virus, methotrexate and lymphoma in patients
with rheumatoid arthritis and primary Sjo ¨gren’s syndrome: case series. The Journal of Rheumatology
2001; 28: 47–53.
50. Mariette X. Lymphomas in patients with Sjo ¨gren’s syndrome: review of the literature and physiopath-
ologic hypothesis. Leukemia & Lymphoma 1999; 33: 93–99.
51. Royen B, Cazals-Hatem D, Sibilia J et al. Lymphomas in patients with Sjo ¨gren’s syndrome are marginal
zone B - cell neoplasms, arise in diverse extranodal and nodal sites, and are not associated with viruses.
Blood 1997; 90: 766–775.
52. Hirose Y, Sugai S, Masaki Yet al. Epstein-Barr virus study in malignant lymphoma in Sjo ¨gren’s syndrome.
International Journal of Hematology 1999; 69: 174–179.
53. Xu Y & Wiermick PH. Systemic lupus erythematosus and B-cell hematologic neoplasm. Lupus 2001; 10:
54. Lo ¨fstro ¨m B, Baklin C, Sundstro ¨m C et al. A closer look at non-Hodgkin’s lymphoma cases in a national
Swedish systemic lupus erythematosus cohort: a nested case-control study. Annals of the Rheumatic
Diseases 2007; 66: 1627–1632.
55. Bernatsky S, Ramsey-Goldman R, Rajan R et al. Non -Hodgkin lymphoma in systemic lupus erythema-
tosus. Annals of the Rheumatic Diseases 2005; 64: 1507–1509.
56. Bernatsky S, Boivin JF, Joseph L et al. An international cohort study of cancer in systemic lupus eryth-
ematosus. Arthritis and Rheumatism 2005; 52: 1481–1490.
57. Kojima M, Nakamura S, Futamura N et al. Malignant lymphoma in patients with rheumatic diseases
other than Sjo ¨grens’ syndrome: a review of the Japanese literature. Japanese Journal of Clinical Oncology
1997; 27: 84–90.
*58. Balandraud N, Guis S, Meynard JB et al. Long term treatment with methotrexate or TNF inhibitors
does not increase Epstein Barr virus load in patients with rheumatoid arthritis. Arthritis and Rheumatism
2007; 57: 762–767.
59. Balandraud N, Roudier J & Roudier C. What are the links between Epstein-Barr virus, lymphoma, and
tumor necrosis factor antagonism in rheumatoid arthritis? Seminars in Arthritis and Rheumatism 2005;
34(Suppl. 1): 31–33.
896E´. Toussirot and J. Roudier