Life Cycle

Primary infection with EBV results from exposure to the oral secretions of seropositive individuals through kissing, sharing of food, or other intimate contact. The long-accepted concept that EBV infection spreads to B lymphocytes after initial productive (lytic) infection of oral epithelial cells^21,22 has been challenged. Tonsillar biopsies from patients with primary EBV infection did not reveal any infected epithelial cells, but infected lymphocytes were readily seen.^23,24 EBV undoubtedly has clinically significant tropism for epithelial cells, as is seen in nasopharyngeal carcinoma and oral hairy leukoplakia. It remains possible that significant infection of oral epithelial cells occurs in nontonsillar sites or that an initial round of lytic replication precedes spread to the B-cell compartment and the onset of symptoms.^25,26 Infected B lymphocytes incite an intense cytotoxic T-cell response, and these T cells constitute the atypical lymphocytosis characteristic of primary EBV infection.^27,28 In healthy individuals, most infected B lymphocytes are cleared through immune surveillance, but between one and 50 B cells per million remain quiescently infected and serve as the reservoir for lifelong infection of the individual.^29,30 Thus, EBV shares the properties of lifelong latency and persistence with other members of the herpesvirus family. In contrast to that of alpha herpes viruses (herpes simplex virus and varicella-zoster virus), shedding of infectious EBV particles into the saliva from periodic reactivation of latently infected cells is entirely asymptomatic. This shedding occurs in otherwise healthy persons but is more frequent in immunosuppressed hosts (Table 141-1).

The host range of the virus is limited. In vitro cultivation of the virus has been described primarily in B lymphocytes and also in nasopharyngeal epithelial cells of humans and certain nonhuman primates.³¹ EBV binds to its receptor, the CD21 molecule, through an interaction with its major envelope glycoprotein, gp350. CD21 or complement receptor 2 (CR2) transduces signals important for B lymphocyte proliferation and can also be expressed by follicular dendritic cells and nasopharyngeal epithelial cells.^32–38 Another EBV glycoprotein, gp42, binds major histocompatibility complex (MHC) class II molecules, which serve as co-receptors for infection of B cells.^39–43 Gp42 also promotes B-cell infection by forming a heterotrimeric complex with the gH/gL EBV glycoproteins, masking a motif on gH that is important for infection of epithelial cells.⁴⁴ EBV virions released from infected B lymphocytes contain lower levels of gp42 and infect epithelial cells more efficiently than virions derived from epithelial cells. This reciprocal tropism is proposed to enhance EBV shuttling between B lymphocytes and the oral epithelium.⁴⁵

Latent Infection and Growth Transformation

After infection with EBV, B lymphocytes enter the cell cycle and proliferate continuously in a process termed transformation or immortalization; these cells can be propagated in vitro indefinitely.⁴⁶ This ability of EBV to convert peripheral blood B cells into immortalized lymphoblastoid cell lines (LCLs) is widely used in genomic studies as a means of preserving DNA samples from volunteer donors for future use.⁴⁷ In vivo EBV-driven B-cell proliferation is observed during infectious mononucleosis, in which it probably serves to expand rapidly the pool of infected B lymphocytes. These B lymphocytes are usually rapidly cleared from the circulation.^48–51 However, in the absence of an intact immune response, EBV infection can result in life-threatening lymphoproliferative disease (LPD).^52,53 The growth-transforming properties of EBV can act in concert with genetic and environmental cofactors to cause malignant diseases in immunocompetent hosts as well.^54,55

EBV infection of B lymphocytes is characterized by a state of viral latency, in which the genome circularizes in the nucleus and is replicated as an episome in concert with host chromosomes by cell enzymes. The infection is latent in the sense that viral particles are not being produced, but it is anything but quiescent. Limited viral gene expression persists, and these genes exert effects on the infected cell. In vitro latent infection of B lymphocytes with EBV is characterized by the expression of latent infection membrane proteins 1 and 2 (LMP1 and LMP2), six EBV nuclear antigens (EBNAs), and two small, nuclear, noncoding RNAs (EBV-encoded RNAs [EBERs]) that are transcribed by RNA polymerase III (Table 141-2).⁵⁶ Additional EBV transcripts have been detected in latent infection and are termed complementary strand transcripts (CSTs) or BamH1 A rightward transcripts (BARTs). Translation of these transcripts into proteins has not been shown, but they appear to serve as precursors for two of three clusters of EBV-encoded microRNAs (miRNAs) whose role in EBV biology remains unclear.^57,58 Recombinant reverse genetic analysis has determined that only LMP1, EBNA1, EBNA2, EBNA3A, EBNA3C, and EBNALP are critical for B-cell growth transformation.⁵⁶ The mechanisms by which these EBV gene products promote B-lymphocyte growth have been the subject of intense investigation.

After the virus gains entry to susceptible B lymphocytes, EBNA2 and EBNALP are the first proteins expressed. EBNA2 is an acidic transactivator that acts as the major switch to turn on latent virus gene expression and several B-cell gene products (including c-myc, c-fgr, CD21, and CD23). It has no intrinsic sequence-specific DNA binding capacity but rather is targeted to promoters by binding to a host DNA binding protein RBP-Jκ (also called CBF1 or CSL), a downstream component of the Notch signaling pathway.^59,60 By an incompletely understood mechanism, EBNALP cooperates with EBNA2 to activate expression of the remaining nuclear proteins and LMP1 and LMP2.⁶¹ LMP1 is the major EBV-encoded oncogene, and its expression in transgenic mice results in B-cell lymphomas.^62,63 It constitutively activates signaling pathways that mimic the growth and survival signals given to B cells by CD4⁺ T lymphocytes through the CD40 surface glycoprotein. LMP1 sends this signal through its cytoplasmic tail, which binds a set of second messenger proteins similar but not identical to those used by CD40.^64,65 Unlike CD40, LMP1 does not require the presence of ligand to form patches in the cell membrane but self-associates constitutively, approximating its cytoplasmic tails to activate signaling.⁶⁶ This results in the activation of nuclear factor κB (NF-κB); c-jun; upregulation of adhesion molecules (intercellular adhesion molecule 1, LFA-1, and LFA-3); cytokine production; B-cell proliferation; and induction of an antiapoptotic state.^48,55 A second EBV latent membrane protein, LMP2, mimics another signal necessary for B-cell survival.⁶⁷ By interacting with signaling molecules of the B-cell receptor (BCR), LMP2 mimics BCR engagement by constitutive patching in the membrane in a manner analogous to LMP1. LMP2 probably also interferes with normal signaling through the BCR by antigenic stimulation to inhibit activation of lytic viral replication (discussed subsequently). Interestingly, LMP2 is not necessary for EBV-mediated outgrowth of B cells in vitro but is probably a critical component of the viral strategy in vivo. The nuclear protein EBNA1 acts to promote the replication of the viral genome by the host machinery when the virus is in the latent, episomal state and to ensure proper segregation of the EBV genome to both daughter cells. The EBNA3 proteins also interact with the RBP-Jκ DNA binding protein and promote B lymphocyte growth and survival by silencing expression of the p16^INK4 and p14^ARF tumor suppressor gene products.^68,69 The function of the highly expressed, noncoding EBV RNAs (EBERs) is incompletely understood.

EBV-associated malignant diseases are exclusively associated with latent infection and latent gene expression. Three general patterns of expression of EBV-encoded proteins have been observed in association with latency (see Table 141-2).^48,49 Expression of all latent genes is seen in LPD in immunosuppressed hosts, in primary CNS lymphoma of patients with AIDS, and during primary EBV infection (infectious mononucleosis), and this program of gene expression is often referred to as latency III.⁵⁴ EBV-associated nasopharyngeal carcinomas, Hodgkin’s lymphoma, and T-cell lymphomas exhibit a more restricted pattern of EBV gene expression (latency II) that includes LMP1, LMP2, EBNA1, and the EBERs and EBNA1.^70,71,72 In Burkitt’s lymphoma (latency I) only EBERs and EBNA1 are expressed.⁵⁴ The more restricted patterns of latent gene expression in some tumors are probably in part the result of the intense immune response against viral proteins.

Lytic Infection

Latent infection can be activated to lytic infection by stimulation of host B cells by certain chemicals, calcium ionophores, or antibodies to surface immunoglobulin.⁷³ The physiologic signals that reactivate EBV lytic replication are unknown, but signaling through the B-cell receptor after antigenic stimulation is a possible scenario. After this inciting event, two EBV-encoded transcriptional activators are expressed: BZLF1 and BRLF1. Expression of these immediate early genes leads to a cascade of events that culminate in the production of early EBV gene (early antigen [EA]) products responsible for viral replication (e.g., thymidine kinase, DNA polymerase) and late (structural) genes of the virus including viral capsid antigens (VCAs).⁷⁴ Lytic infection produces EBV virions and can cause host cell death.

Incidence of Infection

Clinically apparent infectious mononucleosis is more common in populations in which primary EBV exposure is delayed until after the first decade of life. The disease is diagnosed most frequently among adolescents of higher socioeconomic groups in industrialized countries.⁸⁰ The incidence of infectious mononucleosis in a large epidemiologic study in the United States was 45.2 cases per 100,000 per year and was highest in the 15- to 24-year-old age group.⁸¹ The incidence was the same for women as for men, but the peak age-specific incidence occurred 2 years earlier in women. Infectious mononucleosis is 30 times more frequent in whites than in blacks. The infrequency of infectious mononucleosis among blacks, noted as early as 1940, is probably a reflection of earlier primary EBV infection and the higher frequency of subclinical infections in children.^82–84 No clear seasonal incidence has been noted.

Methods of Spread

The virus persists in the B-cell compartment for the life of the infected host and can be cultured from throat washings from 10% to 20% of healthy adults, from 50% of kidney transplant recipients, and from greater proportions of those critically ill with leukemia or lymphoma (see Table 141-1).^85–87 Approximately 50% of men with human immunodeficiency virus type-1 (HIV-1) infection who have sex with men shed EBV in oropharyngeal secretions.⁸⁸ Low titers of EBV are present in throat washings of persons with infectious mononucleosis.^89–91 Susceptible roommates of students with infectious mononucleosis or with inapparent EBV infection have EBV seroconversion no more frequently than the general susceptible college population.^17,79 Only 6% of those with infectious mononucleosis cite previous contact with another case of infectious mononucleosis.⁸¹ EBV DNA or protein, or both, have also been identified in parotid duct and uterine cervical epithelia, although the implications of this distribution are unclear with respect to viral transmission.^92,93

EBV, like other herpesviruses, is relatively labile in the laboratory, and the virus has not been recovered from environmental sources, including fomites. These data suggest that EBV is a widespread agent that is not particularly contagious and that most cases of infectious mononucleosis are probably contracted by intimate contact between susceptible individuals and asymptomatic shedders of EBV. Among young adults, spread of the virus may be facilitated by the transfer of saliva with kissing.^94,95 EBV infection has been linked to sexual intercourse; however, other studies have found that kissing with coitus conferred no additional risk of EBV infection compared with kissing without coitus.^96,97 Serologic evidence suggests that the virus may also be spread among susceptible individuals within families.^98,99 EBV has also been spread via blood transfusion and after open heart surgery as the postpump perfusion syndrome.¹⁰⁰ Most postpump perfusion infectious mononucleosis is, however, heterophile negative and attributable to CMV.

Although several apparent epidemics of infectious mononucleosis have been described, these reports have not been substantiated with EBV serologic data and have lacked rigorous epidemiologic, clinical, or laboratory support. Some of these have resulted from errors in the performance of Monospot tests.¹⁰¹ On the basis of the previously discussed information, true epidemics of infectious mononucleosis are unlikely to occur.