On the dynamics of acute EBV infection and the pathogenesis of infectious mononucleosis

Abstract

Memory B cells latently infected with Epstein-Barr virus (mB(Lats)) in the blood disappear rapidly on presentation with acute symptomatic primary infection (acute infectious mononucleosis [AIM]). They undergo a simple exponential decay (average half-life: 7.5 +/- 3.7 days) similar to that of normal memory B cells. The cytotoxic T lymphocyte (CTL) response to immediate early (IE) lytic antigens (CTL(IEs)) also decays over this time period, but no such correlation was observed for the CTL response to lytic or latent antigens or to the levels of virions shed into saliva. We have estimated the average half-life of CTL(IEs) to be 73 (+/- 23) days. We propose that cycles of infection and reactivation occur in the initial stages of infection that produce high levels of mB(Lats) in the circulation. Eventually the immune response arises and minimizes these cycles leaving the high levels of mB(Lats) in the blood to decay through simple memory B-cell homeostasis mechanisms. This triggers the cells to reactivate the virus whereupon most are killed by CTL(IEs) before they can release virus and infect new cells. The release of antigens caused by this large-scale destruction of infected cells may trigger the symptoms of AIM and be a cofactor in other AIM-associated diseases.

Figure 1

Schematic drawing of our model of EBV persistence. During acute infection, EBV transmitted via saliva infects naive B cells turning them into latently infected B-cell blasts (bB_Lats) and driving their proliferation and differentiation through a germinal center (gcB_Lat) into latently infected memory B cells (mB_Lats) that then migrate to peripheral blood where EBV persists for the lifetime of its host. When mB_Lats migrate back to the tonsil, they occasionally receive signals to undergo terminal differentiation into plasma B cells, which triggers reactivation of the virus. This results in the subsequent expression of lytic genes (eg, IE and EA) and ultimately production of free virions that can then infect new B cells or shed into saliva to infect new hosts. This process in turn stimulates an aggressive antiviral CTL response. The infected cells expressing targets recognized by CTLs are indicated. This response effectively reduces the number of cells that complete viral replication so that during persistent infection little or no new infection of B cells occurs. Whether the stable levels of mB_Lats during persistence are maintained through memory B-cell homeostasis, low-level infection of new B cells, or a combination of the 2 is unresolved.

Figure 2

Hypothesized mechanisms for the rapid decline of mB_Lats. The mB_Lats may simply die, perhaps via apoptosis induced by homeostatic culling. This would not stimulate a virus-specific CTL response. Alternatively, the cells may turn on viral latent proteins or initiate viral replication and be killed by CTLs.

Figure 3

The level of infected mB_Lats over time in the blood of AIM patients. (A) From the time of initial visit and diagnosis to approximately 1 year for 3 patients. The dotted and unbroken horizontal lines are, respectively, the upper limit and the median values for the blood of healthy carriers (n = 30). (B,C) Detailed time course during the first 50 days for 6 patients. The frequency of infected cells (FOIs) is expressed as the percentage of infected memory B cells and is plotted on a log scale to demonstrate the simple exponential decay in the weeks following diagnosis.

Figure 4

Correlation between the levels of mB_Lats and the CTL responses against lytic (IE and EA) and latent (LAT) antigens in the blood of an AIM patient (IM4) immediately following diagnosis. (A) Level over time of mB_Lats expressed as the log of the percentage of infected memory B cells. (B-D) Levels over time of CTL responses against IE, EA, and LAT antigens, respectively, expressed as the log of the percentages of tetramer-binding cells in the total CD8⁺ T-cell population. (E-G) Correlation between the frequency of mB_Lats and of CTL against IE, EA, and LAT antigens, respectively.

Figure 5

Correlation between the levels of mB_Lats, the CTL responses against lytic antigens (IE and EA) in the blood, and shed virion in the saliva of an AIM patient (IM10) immediately following diagnosis. (A) Level over time of mB_Lats expressed as the log of the percentage of infected memory B cells. (B,C) Levels over time of CTL responses against IE and EA antigens, respectively, expressed as the log of the percentages of tetramer-binding cells in the total CD8⁺ T-cell population. Note the values for CTL_Lats are not shown because none was detected at any time point. (D) Level of virions shed into saliva expressed as the logs of the number of virions per milliliter of saliva sample. (E) As in panel D, but from 40 days after diagnosis until approximately one year later. (F) As in panel D, but for patient IM9. (G) As in panel E, but for patient IM9.

Figure 6

Comparison of virus shedding levels in the saliva of 2 groups of AIM patients: at 1 to 2 weeks (n = 8 patients) and 6 to 12 months (n = 8 patients) after diagnosis.

Figure 7

Schematic drawing of the proposed model for the course of acute EBV infection leading to AIM. Acute EBV infection leads to ever-increasing rounds of production of mB_Lat reactivation and infection of new B cells resulting in high levels of mB_Lats and virus shedding. This stimulates strong CTL responses that drastically reduce this one-way circuit, leaving high levels of mB_Lats that simply decay through apoptosis according to the normal half-life of a memory B cells. Apoptosis triggers viral reactivation leading to the production of cells expressing viral IE proteins that are rapidly destroyed by CTL_IEs. Hence the levels of CTL_IEs fall in parallel with those of the mB_Lats. We hypothesize that this massive cell death causes the symptoms of AIM, leading patients to enter the clinic at a time when the levels of mB_Lats and CTL_IEs are rapidly falling but levels of CTLs against other viral target (EA and LAT) and saliva shedding are already leveling off toward the levels found in persistent infection (dashed lines indicating the first 6 weekly visits to the clinic in our study).