Gammaherpesvirus-driven plasma cell differentiation regulates virus reactivation from latently infected B lymphocytes

Abstract

Gammaherpesviruses chronically infect their host and are tightly associated with the development of lymphoproliferative diseases and lymphomas, as well as several other types of cancer. Mechanisms involved in maintaining chronic gammaherpesvirus infections are poorly understood and, in particular, little is known about the mechanisms involved in controlling gammaherpesvirus reactivation from latently infected B cells in vivo. Recent evidence has linked plasma cell differentiation with reactivation of the human gammaherpesviruses EBV and KSHV through induction of the immediate-early viral transcriptional activators by the plasma cell-specific transcription factor XBP-1s. We now extend those findings to document a role for a gammaherpesvirus gene product in regulating plasma cell differentiation and thus virus reactivation. We have previously shown that the murine gammaherpesvirus 68 (MHV68) gene product M2 is dispensable for virus replication in permissive cells, but plays a critical role in virus reactivation from latently infected B cells. Here we show that in mice infected with wild type MHV68, virus infected plasma cells (ca. 8% of virus infected splenocytes at the peak of viral latency) account for the majority of reactivation observed upon explant of splenocytes. In contrast, there is an absence of virus infected plasma cells at the peak of latency in mice infected with a M2 null MHV68. Furthermore, we show that the M2 protein can drive plasma cell differentiation in a B lymphoma cell line in the absence of any other MHV68 gene products. Thus, the role of M2 in MHV68 reactivation can be attributed to its ability to manipulate plasma cell differentiation, providing a novel viral strategy to regulate gammaherpesvirus reactivation from latently infected B cells. We postulate that M2 represents a new class of herpesvirus gene products (reactivation conditioners) that do not directly participate in virus replication, but rather facilitate virus reactivation by manipulating the cellular milieu to provide a reactivation competent environment.

Figure 1. Requirement of MHV68 M2 protein for efficient reactivation from B cell latency.

(A) Schematic illustration MHV68 recombinant viruses harboring a hygromycin phosphotransferase gene fused to EGFP (Hygro-EGFP) in the neutral locus of either wild type (WT-HE) or an M2-null mutant virus (M2.Stop-HE). (B) Latently infected WT-HE and M2.Stop-HE M12 cells were either mock treated or stimulated with TPA and expression of viral replication-associated antigen detected with either an anti-MHV68 serum generated in infected rabbits or antibodies raised against a peptide from the replication-associated ORF59 encoded viral antigen. Antibodies against tubulin were used to control for amount of cell lysate loaded. (C) Virus production from WT-HE and M2.stop-HE M12 cells as a function of time post-TPA stimulation. Virus titers were determined by plaque assay on NIH 3T3 fibroblasts as previously described . (D) WT or M2.Stop-HE M12 cells were either mock treated or stimulated with TPA at 24 hr post-transfection with an M2 expression plasmid (pEF/M2-AU1) or the parental empty expression vector (pEF). Cell lysates harvested at 24 hr post-TPA induction were analyzed by immunoblotting using a rabbit anti-MHV68 antiserum generated from MHV68 infected rabbits .

Figure 2. Distribution of M2 null virus in newly formed, marginal zone and follicular B cells in the spleen at the peak of latency is similar to wild type MHV68.

(A and B) Insertion of a YFP expression cassette into the viral genome, or addition of an AU1 epitope tag on the C-terminus of M2, did not alter the phenotype of the M2 null virus. Latency and reactivation phenotypes of several YFP expressing MHV68 recombinant viruses were assessed using standard limiting dilution analyses to compare the eYFP expressing M2.Stop recombinant virus to the well characterized M2.Stop mutant, as well as different wt YFP expressing viruses - one of which contains an AU1 epitope tag on the C-terminus of M2. All data were pooled from two independent virus isolate, five mice infected with 100 pfu of the indicated virus via intranasal inoculation. (A) Limiting dilution PCR analyses of the frequency of splenocytes harboring viral genome following infection with the indicated recombinant viruses. (B) Limiting dilution analyses of MHV68 reactivation following explant of splenocytes harvested at day 16 post-infection. (C) Similar levels of marginal zone (MZ), follicular (FC) and newly formed (NF) B cells in spleens infected with wt MHV68-YFP or M2 null viruses (M2.stop-YFP). (D) Percentage of MZ, FC and NF B cells in virus infected (YFP+ cells) splenocytes with wt MHV68-YFP or M2 null viruses. (E) Percentage of activated B cells present in wt MHV68-YFP or M2 null virus infected spleen cells at day 16 post-infection. The p values indicated in the figure were determined by two-tailed, unpaired t test with a 95% confidence level.

Figure 3. M2 null virus infected B cells form germinal centers, but exhibit a defect in immunoglobulin isotype class switching.

Mice were infected with a 100 pfu of either MHV68-YFP or M2.Stop-YFP viruses via intranasal inoculation. (A) Total germinal center B cells (CD3⁻/B220⁺/GL7^hi/CD95^hi) present in the spleen are reduced in M2.stop-YFP infected mice compared to MHV68-YFP infected mice at day 16. Representative flow cytometry plots are shown for 3–4 individual mice. (B) Scatter plot depicting the percentage of total B220+ splenocytes that are germinal center B cells (GL7+/CD95+) in naïve mice, MHV68-YFP or M2stop-YFP infected spleens at day 16 post-infection. (C) The distribution of wild type and M2 null virus infected B cells (YFP⁺ cells) in the spleens of infected mice. (D) Percentage of wild type and M2 null virus infected (YFP+) B cells vs uninfected B cells (YFP-) in the spleen that exhibited a germinal center phenotype. (E) Percentage of splenic B cells that were IgG2a+ in wild type vs M2 null virus infected mice at day 16 post-infection. (F) M2.stop-YFP⁺ spleen cells showed an impaired ability to class-switch to IgG2a.

Figure 4. Access of MHV68 to the plasma cell reservoir during MHV68 chronic infection requires a functional M2 gene.

Mice were infected with a 100 pfu of either MHV68-YFP or M2.Stop-YFP viruses via intranasal inoculation. (A) The absence of MHV68 infected plasma cells in mice infected with the M2.Stop-YFP virus. Flow cytometry analyses, gated on YFP+ cells, to determine the presence of virus infected plasma cells (CD138⁺/B220⁻) in the spleen at day 16 post-infection. Data shown were compiled from representative wild type (MHV68-YFP) and M2 null virus (M2.Stop-YFP) infected mice (4 mice per virus). (B) Similar levels of total plasma cells in mice infected with wild type vs. the M2 null MHV68. Compiled data showing the percentage of plasma cells present in wt MHV68 (MHV68-YFP) infected or M2 null virus (M2.stop-YFP) infected spleen cells at day 16 post-infection. (C) Percentage of plasma cells in virus infected (YFP+) and uninfected (YFP-) splenocytes harvested at day 16 post-infection. (D) ELISPOT analyses of antibody secreting cells (ASC) present in wt MHV68 (MHV68-YFP) infected splenocytes (YFP+ cells) or bulk splenocytes harvested at day 16 post-infection.

Figure 5. Plasma cells contribute significantly to spontaneous MHV68 reactivation following explant.

(A and B) Purification of plasma cells and non-plasma cells by flow cytometry. Shown are representative flow plots unsorted (A) and post-sorting (B), as well as the gates used to isolate the plasma cells and non-plasma cell populations. Cells were sorted from MHV68-YFP infected spleens at day 16 post-infection. (C) Limiting dilution determinations of the frequency of splenocyte populations harboring the MHV68 genome (left panel) or spontaneously reactivating virus upon explant (right panel). Splenocytes were harvested from C57Bl/6 mice at day 16 post-infection. For the reactivation analyses, both intact cells (filled symbols) and mechanically disrupted cells (open symbols) were plated to distinguish the presence of pre-formed infectious virus from reactivating virus. (D) ELISPOT analyses of antibody secreting cells in the unsorted and purified plasma cell and non-plasma cell populations analyzed in (A).

Figure 6. M2 can drive plasma cell differentiation in a B lymphoma cell line.

(A) Appearance of plasmacyte morphology following transfection of the murine BCL-1 lymphoma cells with a M2 expression vector. GFP fluorescence in BCL-1 lymphoma cells transfected with either vector control (MSCV-M2.Stop-IRES-GFP) or an M2 expression vector (MSCV-M2-IRES-GFP). GFP expression, driven from an internal ribosome entry site introduced downstream of the M2 open reading frame, was examined at 24 hr post-transfection. (B) M2 expressing BCL-1 cells exhibited increased size and granularity compared to M2.Stop transfected cells, as determined by changes in forward and side light scatter. Filled gray histograms were gated on the GFP expressing cell population in BCL-1 cells transfected with the M2.Stop vector, while the black open histograms were gated on the GFP expressing cell population in BCL-1 cells transfected with the M2 expression vector. (C) RT-PCR analysis of transcripts associated with the plasma cell differentiation program. -, untreated BCL-1 cells; LPS, BCL-1 cells treated with 20 ug/ml LPS and RNA prepared at 24 hr post-treatment; V, BCL-1 cells transfected with a control vector (MSCV-M2.Stop-IRES-GFP); M2, BCL-1 cells transfected with an M2 expression vector (MSCV-M2-IRES-GFP). BCL-1 cells were transfected using a Nucleofector (Amaxa) and RNA prepared at 24 hours post-transfection. (D) Analysis of IgM levels in the tissue culture supernatant of mock treated vs LPS stimulated, or vector control (MSCV-M2.Stop-IRES-GFP) vs M2 expression vector (MSCV-M2-IRES-GFP) transfected BCL-1 cells. As indicated, IgM levels were determined at 48 hours post-treatment. (E) Model of M2 induction of plasma cell differentiation leading to virus reactivation from latency.