Epstein–Barr virus (EBV) is an enveloped virus with a double-stranded DNA genome that infects humans and persists for life, principally in B lymphocytes. It belongs to the gammaherpesviruses, a group distinguished by their ability to establish persistent infections in lymphoid cells. More than 90% of adults worldwide have evidence of previous EBV infection. Most infections cause no recognizable illness, particularly when acquired in early childhood, but primary infection during adolescence or adulthood can produce infectious mononucleosis.
EBV occupies an important position in virology, immunology, and cancer biology. It was the first human virus discovered in association with a cancer, and it can induce sustained proliferation of human B cells in laboratory culture. In infected people, a regulated balance between viral persistence and immune surveillance usually prevents disease. Disruption of that balance, together with particular genetic, environmental, or immunological conditions, contributes to several lymphomas, nasopharyngeal carcinoma, and a subset of gastric cancers. EBV infection also has a particularly strong epidemiological relationship with multiple sclerosis.
Its biology depends on switching between productive replication, which generates new virus particles, and latency, in which viral DNA remains within a cell while only selected viral genes are expressed. These alternative states help explain its efficient transmission, lifelong persistence, varied disease associations, and the difficulty of developing treatments that eliminate infection.
Discovery and scientific classification
The discovery of EBV grew out of investigations of an unusual childhood tumor in equatorial Africa. In 1958, the surgeon Denis Burkitt described a distinctive malignancy, often involving the jaw, that later became known as Burkitt lymphoma. Its geographical distribution encouraged investigation of infectious and environmental contributions. Following Burkitt’s presentation in London in 1961, the pathologist Anthony Epstein began studying tumor material with the aim of identifying a possible viral agent.
In 1964, Michael Anthony Epstein, Bert Achong, and Yvonne Barr reported herpesvirus-like particles in cultured cells derived from a Burkitt lymphoma. The discovery depended on maintaining tumor cells in culture and examining them by electron microscopy. It established an association between a previously unknown human virus and a malignancy, but identifying viral particles did not by itself explain whether the virus caused the tumor. Subsequent evidence from serology, molecular studies, experimental infection, and tumor biology established that EBV contributes directly to several cancers.
A second major development connected the virus with infectious mononucleosis. In work published in 1968, Werner and Gertrude Henle and their collaborators documented the appearance of EBV-specific antibodies in relation to mononucleosis. Prospective observations in students subsequently showed that primary EBV infection could precede the characteristic clinical syndrome. These findings brought together a common acute illness and a virus originally discovered through cancer research.
EBV belongs to the family Orthoherpesviridae, historically widely called Herpesviridae, the subfamily Gammaherpesvirinae, and the genus Lymphocryptovirus. Related lymphocryptoviruses infect other primates. EBV is also known as human herpesvirus 4, although the virus name, formal species designation, and clinical abbreviation belong to different naming conventions. Its persistent lymphoid infection contrasts with the predominantly neuronal latency of herpes simplex and varicella-zoster viruses.
Two broad EBV types, conventionally designated type 1 and type 2, differ most conspicuously in sequences encoding EBV nuclear antigen 2 and parts of the EBNA3 family. Type 1 generally transforms B cells more efficiently in standard culture systems. Both types circulate in human populations, their geographical frequencies differ, and mixed infections occur. Whole-genome DNA sequencing has revealed substantially more diversity than this two-type classification captures, including recombination and regional viral lineages. The laboratory reference strain B95-8 has been especially influential, but it contains a deletion relative to many circulating viruses and therefore does not represent every aspect of natural EBV biology.
Virion structure and genome organization
An EBV particle has the layered architecture characteristic of herpesviruses. Its DNA is enclosed in an icosahedral capsid, surrounded by a protein-rich tegument and an outer lipid envelope bearing viral glycoproteins. Complete particles are commonly around 150–200 nanometres in diameter, although their apparent dimensions vary with preparation and imaging method. The envelope is derived from host cellular membranes during viral assembly, but its embedded viral proteins determine important aspects of attachment, membrane fusion, and host-cell recognition.
The linear viral genome is approximately 170 kilobase pairs long, with differences among strains arising partly from repetitive sequences and deletions. It encodes dozens of proteins, including structural components, replication enzymes, transcriptional regulators, and immune-modulating products. It also produces abundant noncoding RNAs and multiple microRNAs. The two small RNAs EBER1 and EBER2 are especially useful as laboratory markers because they accumulate to high levels in many latently infected cells.
Terminal repeat sequences occur at the ends of virion DNA. After the genome reaches the cell nucleus, the ends ordinarily join to form a circular extrachromosomal molecule, or episome. During latency this DNA associates with histones and other cellular proteins, acquiring a chromatin organization that helps regulate viral gene expression. Unlike several other tumor viruses, EBV does not normally require integration into a host chromosome for persistence or oncogenic activity.
Different regions of the genome support distinct biological programs. The latent origin of replication, oriP, and the viral protein EBNA1 allow episomes to be maintained in dividing cells. Separate lytic origins support the extensive DNA amplification needed to make new virions. Viral genes expressed during productive replication include a DNA polymerase, its accessory factors, capsid proteins, and proteins involved in packaging DNA and assembling infectious particles. This separation between latent maintenance and productive replication is central to the virus’s response to cellular conditions.
Envelope glycoproteins have complementary functions rather than serving as interchangeable attachment factors. The abundant glycoprotein gp350/220 promotes attachment to B cells, principally through complement receptor 2, also called CD21. The gH/gL complex, gp42, and gB participate in receptor engagement and fusion. Their relative abundance affects which cells the virus enters efficiently. Nonstructural viral products are equally important to disease: latent membrane proteins alter cellular signaling, nuclear antigens reshape transcription, and viral microRNAs influence both infected-cell survival and interactions with immunity. The genome thus encodes not only the machinery for particle production but also a sophisticated system for controlling the infected cell between episodes of replication.
Transmission, population distribution, and cellular entry
EBV circulates worldwide, and humans are its principal natural reservoir. Transmission occurs mainly through saliva, including intimate contact and other forms of salivary exchange. Infectious virus may be shed by people without symptoms, both after primary infection and intermittently during long-term carriage. This persistent, often unnoticed shedding helps explain why infection is so widespread. Transmission through transplanted organs or blood products also occurs, although these are not the main routes sustaining infection in the general population.
The age of first infection varies with social and environmental circumstances. In many populations, infection occurs predominantly in early childhood and is often clinically silent or resembles other minor childhood illnesses. Where acquisition is delayed until adolescence or young adulthood, infectious mononucleosis becomes more conspicuous. These differences reflect patterns of exposure rather than the existence of a childhood-specific and an adult-specific virus. A person’s EBV antibody status generally remains positive after infection, making population serosurveys useful for reconstructing age-related exposure.
The mouth and oropharynx provide the main setting for entry and shedding. EBV can infect both B cells and epithelial cells, but the precise sequence of cellular events during natural acquisition is not completely resolved. Lymphoid tissues of the tonsils provide close contact between these cell populations. Experimental systems support several routes of transfer, including direct interaction between infected B cells and epithelial cells, while cell-free infection varies substantially with cell type and experimental conditions.
For B-cell entry, gp350 binding to CD21 concentrates virus at the cell surface. The viral glycoprotein gp42 interacts with class II major histocompatibility complex molecules, helping trigger the fusion machinery. EBV infection therefore depends on more than simple attachment: receptor engagement must be coordinated with gH/gL and gB to deliver the capsid into the cell. Following intracellular transport, viral DNA reaches the nucleus and begins the transcriptional program appropriate to the newly infected cell.
Epithelial entry uses a different receptor environment and does not follow the same gp42-dependent mechanism. Ephrin receptor A2 is one identified epithelial entry factor, and integrins and additional cellular proteins can contribute in particular models. The producer cell also affects subsequent tropism. Virions produced in B cells tend to contain less gp42 because gp42 interacts with class II molecules during production; these particles can be relatively efficient at infecting epithelial cells. Epithelial-cell-derived particles can carry more gp42 and efficiently infect B cells. This reciprocal relationship provides a mechanistic explanation for how movement between epithelial and lymphoid compartments can support infection, although it does not by itself establish every step of transmission in humans.
Latency, reactivation, and lifelong persistence
After entering a B cell, EBV can activate a growth program that drives the cell into proliferation. In culture, this process produces continuously growing lymphoblastoid cell lines, an experimental system central to understanding viral transformation. In healthy people, however, unrestricted growth is usually prevented by immunity and by progressive restriction of viral gene expression. Long-term persistence is associated chiefly with resting memory B cells, which can retain viral episomes while expressing very few viral proteins.
Several latency programs are distinguished by their patterns of viral gene expression. Latency III includes the EBV nuclear antigens EBNA1, EBNA2, EBNA-LP, and the EBNA3 proteins, together with latent membrane proteins and noncoding RNAs. It supports B-cell growth but exposes numerous targets to immune recognition. Latency II generally includes EBNA1 and latent membrane proteins, while latency I is characterized chiefly by EBNA1 among the standard latent proteins. Latency 0 describes resting infected cells without detectable expression of these latent proteins. These categories are useful frameworks rather than a complete inventory of every transcriptional state in every tissue.
Latent proteins manipulate normal B-cell signals. EBNA2 works through cellular DNA-binding factors to activate viral and host transcription. LMP1 supplies signaling resembling that of the activated CD40 receptor and can stimulate NF-κB and other pathways associated with survival and proliferation. LMP2A can provide signals analogous to those from the B-cell receptor, permitting survival under conditions in which normal receptor signaling is absent or altered. EBNA1 binds oriP, supports episomal maintenance, and helps viral genomes segregate as infected cells divide.
The differentiation state of the infected cell is closely connected to the choice between latency and replication. Memory-cell formation provides a relatively quiet reservoir, whereas differentiation toward an antibody-secreting plasma cell can favor reactivation. The immediate-early viral proteins BZLF1 and BRLF1 initiate transcriptional changes that activate the lytic program. BZLF1 is particularly notable for recognizing selected methylated viral DNA sequences, allowing the epigenetic state of the latent genome to participate in the transition toward productive infection.
During lytic replication, early proteins support viral DNA synthesis, and later proteins build and package new particles. Capsids form in the nucleus; subsequent steps involve nuclear egress, acquisition of tegument, and final envelopment through cellular membrane compartments. Productive replication can damage or destroy the infected cell, unlike silent episomal persistence. Not every episode of viral gene activation completes this sequence: abortive or incomplete lytic programs occur in some settings. Lifelong carriage therefore consists of heterogeneous infected-cell states rather than a uniform, continuously productive infection, and asymptomatic reactivation is not equivalent to a new primary infection.
Immune control and infectious mononucleosis
EBV infection elicits coordinated innate and adaptive responses. Natural killer cells contribute to early control, particularly against cells undergoing lytic infection. Virus-specific T cells subsequently recognize infected cells displaying viral peptides, while antibodies can neutralize extracellular particles and provide durable evidence of exposure. This response usually limits the expanding infected-cell population without eliminating every latent genome. The resulting equilibrium allows lifelong persistence in an otherwise healthy host.
Infectious mononucleosis is most characteristic of primary infection acquired during adolescence or adulthood. After an incubation period commonly around four to six weeks, the typical syndrome includes fever, pharyngitis, enlarged lymph nodes, and fatigue. Enlargement of the spleen and liver can occur, and liver-enzyme elevations are common. The severity and duration vary: prominent throat symptoms may settle before fatigue and reduced physical capacity disappear. Primary EBV infection in young children is less likely to produce the classic syndrome.
The unusual circulating lymphocytes that gave mononucleosis its name are largely activated cytotoxic T cells rather than the infected B cells themselves. A substantial expansion of EBV-specific CD8 T cells can occur, especially against lytic viral antigens. The clinical illness therefore reflects both infection and the vigorous immune response to it. Tissue inflammation, lymphoid enlargement, and circulating immune mediators help explain why the systemic symptoms can be marked even though the long-term infected reservoir is comparatively small.
EBV also limits immune recognition through several mechanisms. Some lytic proteins interfere with antigen processing or presentation, while other products alter cytokine signaling. BCRF1 encodes a viral relative of interleukin-10 with immunomodulatory effects. EBNA1 contains a glycine–alanine repeat region that reduces the efficiency of particular pathways of antigen production and presentation, although EBNA1 is not immunologically invisible. Restricting the number of viral proteins expressed during latency is itself a powerful means of reducing exposure to immune surveillance.
Most uncomplicated mononucleosis resolves without a lasting organ disorder, but recognized complications include upper-airway obstruction from lymphoid swelling, autoimmune hemolytic anemia, thrombocytopenia, and neurological manifestations. Splenic rupture is uncommon but potentially severe. EBV is not the only cause of a mononucleosis-like illness: cytomegalovirus, acute HIV infection, toxoplasmosis, and several other conditions can produce overlapping findings. Conversely, persistent fatigue alone does not establish ongoing EBV replication, and a strong antibody response can remain detectable long after acute illness has ended.
Contributions to human cancers
EBV is classified by the International Agency for Research on Cancer as carcinogenic to humans. Its causal contribution differs among tumor types, and infection alone is not sufficient to explain cancer development. The great majority of infected people never develop an EBV-associated malignancy. Tumor risk instead reflects interactions among viral persistence, host genetics, somatic mutations, immune control, and environmental exposures. Different tumors also select different viral latency programs, showing that EBV does not promote every malignancy through one universal mechanism.
In Burkitt lymphoma, a defining feature is deregulation of the cellular MYC gene, usually through a chromosomal translocation involving an immunoglobulin locus. EBV is present in most endemic African cases but in a smaller proportion of sporadic cases elsewhere. Intense exposure to Plasmodium falciparum malaria contributes to the epidemiological setting of endemic disease through effects that include B-cell stimulation and altered immune control. The restricted latency pattern in many Burkitt tumors differs markedly from the broad growth program seen when EBV transforms normal B cells in culture.
EBV is also present in a subset of classical Hodgkin lymphomas. In these tumors, viral products are localized to the malignant Hodgkin and Reed–Sternberg cells rather than merely to unrelated infected lymphocytes in the tissue. LMP1 and LMP2A can support signaling and survival in cells with abnormal or ineffective B-cell receptor function. The proportion of EBV-positive cases varies with age, geographical setting, histological subtype, and immune status. EBV additionally contributes to several other B-cell and T- or natural-killer-cell lymphoid malignancies.
Among epithelial cancers, nasopharyngeal carcinoma has a particularly strong association with EBV, especially the nonkeratinizing forms common in southern China and parts of Southeast Asia. Viral involvement occurs alongside inherited susceptibility and environmental factors; it is not explained simply by how common EBV infection is in the population. EBV-associated gastric carcinoma forms another distinct molecular group, accounting for roughly one in ten gastric cancers in many series, with considerable variation among studies. These tumors show characteristic patterns of DNA methylation and other molecular changes.
Several lines of evidence distinguish meaningful viral involvement from incidental infection. Demonstrating EBV within malignant cells, rather than in neighboring immune cells, is essential. Shared viral terminal-repeat patterns can support infection preceding expansion of a tumor clone, while functional studies show how latent proteins, viral RNAs, and altered host regulation contribute to survival and growth. Some tumor cells remain dependent on viral functions, but the relevant dependencies vary by cancer. The presence of EBV therefore provides biological and sometimes diagnostic information without implying that all EBV-positive tumors behave identically or require the same treatment.
Immune deficiency, chronic active disease, and multiple sclerosis
Failure of immune surveillance can change EBV from a largely controlled infection into a cause of severe lymphoid proliferation. After solid-organ or hematopoietic stem-cell transplantation, immunosuppression may permit infected B cells to expand, producing post-transplant lymphoproliferative disorder (PTLD). This category spans several pathological entities rather than a single uniform disease. Many early cases are EBV-associated, although EBV-negative forms also occur. An EBV-seronegative recipient exposed to an EBV-positive donor represents an important risk setting for primary infection under immunosuppression.
In advanced HIV-associated immune deficiency, EBV contributes to particular lymphomas and to oral hairy leukoplakia, a usually painless white lesion of the tongue associated with productive infection in epithelium. Inherited immune disorders can reveal especially precise requirements for EBV control. X-linked lymphoproliferative disease caused by pathogenic variants in SH2D1A, which encodes the signaling protein SAP, can produce severe disease after EBV infection. Other defects affecting T-cell and natural-killer-cell activation, killing, or coordination also predispose to serious EBV-associated illness.
Chronic active EBV disease is a rare systemic disorder, not a synonym for prolonged tiredness after mononucleosis. It involves persistent or recurrent systemic symptoms, elevated EBV burden, and demonstrable EBV-infected cellular populations, often T or natural killer cells in the systemic entity recognized in current classifications. Organ involvement can include hepatitis, splenomegaly, cytopenias, and inflammatory complications. Establishing the identity of infected cells is important because a positive blood PCR result alone does not define the disorder.
EBV is also an important infectious trigger of hemophagocytic lymphohistiocytosis, a syndrome of severe immune activation. Excessive activation of lymphocytes and macrophages produces high cytokine levels and can damage multiple organs. The relationship among infection, impaired cytotoxic function, and inflammatory amplification is central to its pathogenesis. EBV-associated disease in these settings is therefore not always explained by the quantity of virus alone; the infected cell lineage and the character of the immune response also matter.
The relationship with multiple sclerosis is supported by unusually strong longitudinal evidence. A 2022 study of United States military personnel found that MS risk increased approximately 32-fold following EBV seroconversion among participants initially negative for EBV; this was a relative-risk estimate in a specific cohort, not an individual probability of developing MS. Proposed mechanisms include molecular mimicry between EBV proteins and nervous-system antigens, altered B-cell biology, and persistent immune dysregulation. Experimental work has identified cross-reactive responses involving EBNA1 and the human protein GlialCAM. EBV infection remains vastly more common than MS, and these findings do not mean that infection alone is sufficient or that the complete disease mechanism has been established.
Laboratory diagnosis and interpretation
Laboratory investigation of EBV serves several different purposes: identifying recent primary infection, documenting past exposure, measuring viral DNA in defined high-risk settings, and demonstrating virus in diseased tissue. These questions require different tests. A result that is informative for one purpose can be misleading for another. In particular, detecting antibodies does not necessarily indicate current viral replication, and detecting EBV DNA does not by itself identify the cause of a patient’s symptoms.
Serology commonly measures antibodies to viral capsid antigen, abbreviated VCA, and to EBV nuclear antigen, usually EBNA1. VCA-specific IgM generally appears early in primary infection and usually disappears over subsequent weeks or months. VCA-specific IgG appears during acute infection and usually persists for life. Antibodies to EBNA1 generally develop later, often two to four months after symptom onset. Thus VCA IgM with absent EBNA antibodies supports recent primary infection, whereas VCA IgG together with EBNA IgG usually indicates past infection. Individual patterns can be atypical, particularly in immunocompromised people.
Heterophile antibody tests, often associated with the Monospot name, detect antibodies that react with antigens unrelated to EBV itself. They can support recognition of infectious mononucleosis in an appropriate clinical setting but are not EBV-specific. Sensitivity is lower early in illness and in young children, and false-positive results occur. Antibodies to early antigen are also imperfect markers of active disease: some previously infected healthy people retain them. High or persistent antibody titres therefore cannot be interpreted as a direct measurement of ongoing viral activity.
Quantitative polymerase chain reaction measures EBV DNA and has important roles in transplantation and selected malignancies. Results depend on the sample. Whole blood includes DNA within infected circulating cells, whereas plasma more strongly reflects cell-free material, which may arise from tissue injury, tumor-cell turnover, or other sources. Results from plasma, whole blood, and isolated cells are not interchangeable. Differences among assays, reporting units, and clinical settings also limit the use of a single universal threshold for disease prediction.
For tissue diagnosis, in situ hybridization for EBERs is a principal method of locating infected cells. It is particularly useful in lymphomas and epithelial tumors because it shows which cells contain the virus. Immunohistochemistry for selected viral proteins can provide complementary information but may be negative when the relevant protein is absent from that latency program. In nasopharyngeal carcinoma, circulating plasma EBV DNA can function as a biomarker of tumor burden and contribute to risk assessment or monitoring in defined settings. None of these applications makes an isolated positive EBV result equivalent to a diagnosis of cancer, and tissue morphology and the wider clinical context remain fundamental.
Treatment, prevention, and experimental approaches
The distinction between latent and lytic infection strongly shapes treatment. Drugs that inhibit viral DNA polymerase act mainly during productive replication, when the viral replication machinery is active. They do not reliably remove the latent episomes carried by resting B cells. This explains why antiviral activity in laboratory systems or reduced salivary shedding does not necessarily translate into elimination of infection or substantial improvement in uncomplicated infectious mononucleosis.
Care for ordinary mononucleosis is primarily supportive rather than directed at eradicating EBV. Trials of acyclovir and related agents have not established a clear routine clinical benefit for uncomplicated disease, despite effects on viral shedding. Corticosteroids are not a general treatment for EBV infection; their clinical use is limited to selected serious complications. A different situation exists for EBV-associated malignancies and severe lymphoproliferative disorders, where therapy is determined by the pathological diagnosis, immune setting, and affected organs rather than simply by the presence of the virus.
In transplant-associated disease, reducing immunosuppression can help restore antiviral immune control, although its feasibility is constrained by the risk to the transplant. Rituximab, an antibody directed against CD20, is important in many CD20-positive B-cell PTLD settings, sometimes alongside chemotherapy or other interventions. Its target is the B cell, not a viral protein, and it is not a treatment for every EBV-infected cellular lineage. Likewise, therapies for Hodgkin lymphoma, Burkitt lymphoma, nasopharyngeal carcinoma, and gastric carcinoma differ substantially despite their possible shared viral association.
Adoptive cellular therapy aims to restore the missing antiviral response more directly. EBV-specific T cells can recognize viral antigens displayed by infected cells and have been used particularly for post-transplant disease. Tabelecleucel, an allogeneic EBV-specific T-cell product, received European Union authorization in 2022 for defined relapsed or refractory EBV-positive PTLD settings. Other approaches under study include engineered cellular therapies, strategies targeting EBNA1 or viral dependencies, and attempts to induce lytic gene expression so that tumor cells become more susceptible to antiviral drugs or immune attack. These strategies differ in maturity and are not interchangeable.
Vaccine development has focused on preventing acquisition, reducing infectious mononucleosis, or strengthening immune control after exposure. Candidate antigens include gp350 and combinations of entry glycoproteins intended to elicit neutralizing antibodies against infection of both B cells and epithelial cells. An earlier randomized gp350 vaccine trial reduced mononucleosis without clearly preventing infection, demonstrating that protection against disease and protection against acquisition are distinct outcomes. Newer platforms include multivalent protein and messenger-RNA approaches. Whether vaccination can also reduce EBV-associated cancers or MS requires much longer follow-up and appropriately designed studies, because these outcomes are uncommon relative to infection and may arise many years after initial exposure.