ATM facilitates mouse gammaherpesvirus reactivation from myeloid cells during chronic infection

Abstract

Gammaherpesviruses are cancer-associated pathogens that establish life-long infection in most adults. Insufficiency of Ataxia-Telangiectasia mutated (ATM) kinase leads to a poor control of chronic gammaherpesvirus infection via an unknown mechanism that likely involves a suboptimal antiviral response. In contrast to the phenotype in the intact host, ATM facilitates gammaherpesvirus reactivation and replication in vitro. We hypothesized that ATM mediates both pro- and antiviral activities to regulate chronic gammaherpesvirus infection in an immunocompetent host. To test the proposed proviral activity of ATM in vivo, we generated mice with ATM deficiency limited to myeloid cells. Myeloid-specific ATM deficiency attenuated gammaherpesvirus infection during the establishment of viral latency. The results of our study uncover a proviral role of ATM in the context of gammaherpesvirus infection in vivo and support a model where ATM combines pro- and antiviral functions to facilitate both gammaherpesvirus-specific T cell immune response and viral reactivation in vivo.

Keywords: ATM; Gammaherpesvirus; Myeloid cell; Reactivation; T cell response.

Figure 1. Validation of M-Cre ATM^c/c mouse model

(A) Conditional ATM alleles (ATM^c/c) include LoxP sequences flanking exons 57 and 58 (PI3K domain of ATM). Cre-mediated recombination results in deletion of exons 57 and 58. Arrows indicate the location of primers used in panel B. (B, C) Bone marrow derived macrophages were generated from M-Cre positive and M-Cre negative mice. (B) Following 14 days of culture, relative abundance of intact and recombined atm alleles was measured by qPCR analysis using primers shown in panel A. (C) Whole cell lysates were collected at day 9 and day 14 of culture and ATM and β-actin protein expression assessed by western blot analysis. (D, E) Peritoneal cells were harvested from M-Cre positive or MCre negative mice and CD11c^-CD11b⁺F4/80⁺ cells sorted by FACS. Sorted cells were either analyzed for surface protein expression by flow cytometry (D) or subjected to western blot analysis using indicated antibodies (E).

Figure 2. Myeloid cell-specific ATM depletion attenuates establishment of MHV68 latency

M-Cre positive and negative littermates (3-5 mice/group for each experiment) were intranasally (A-D) or intraperitoneally (E, F) inoculated with 10⁴ PFU of MHV68. Splenocytes and peritoneal cells were harvested at 20 (A-D) or 16 (E, F) days post infection and pooled from mice in each experimental group. Frequencies of reactivation (A, C, E) and MHV68-infected cells (B, D, F) were determined using limiting dilution assays. Data were pooled from 2-5 independent experiments and subjected to statistical analyses.

Figure 3. Myeloid-restricted ATM depletion does not affect MHV68-specific immune response

M-Cre positive or negative littermates were mock-inoculated or intranasally infected with MHV68 as described in Fig. 2. (A, B) Splenocytes and peritoneal cells were harvested at 16-20 days post infection and stained with either anti-CD8 or anti-CD4 in combination with anti-CD3 or with anti-B220 or anti-NK1.1. Data show the absolute number of B220⁺, NK1.1⁺, CD3⁺CD8⁺ double positive, and CD3⁺CD4⁺ double positive cells. (C, D) Splenocytes or peritoneal cells harvested from MHV68- or mock-infected mice of indicated genotypes were stained with anti-CD8 and anti-CD3 in combination with either ORF6₄₈₇/D^b or ORF61₅₂₄/K^b MHC class I tetramers. Data were pooled from at least 2-3 independent experiments. Floating bars represent the mean and each symbol represents an individual mouse. (E) Serum was collected at 16 days post infection and serial dilutions assayed for MHV68-specific immunoglobulin.

Figure 4. Myeloid cell-specific ATM deficiency fails to alter relative abundance of myeloid cell types in the peritoneum

M-Cre positive or negative littermates were intranasally infected with MHV68 as in Fig. 2. Peritoneal cells were harvested at 16-20 days post infection and stained with antibodies against CD11c, GR-1, CD11b, and F4/80 for analysis by flow cytometry. (A) Representative flow cytometry plots showing gating strategy in all experimental groups. (B) Absolute number and frequency of indicated myeloid cells populations defined based on the gating strategy shown in panel A. Data were pooled from 2-3 independent experiments, 3-4 mice/group. Error bars represent SEM.

Figure 5. Induction of the DNA damage response stimulates MHV68 reactivation from latency in an ATM-dependent manner

Peritoneal cells were harvested from BL6 mice infected with MHV68 for 28 days. Following harvest, cells were subjected to indicated doses of γ-irradiation, and allowed to reactivate MHV68 ex vivo in untreated medium (A, B, No Tx group in C), or medium supplemented with 10μM Ku-55933 or DMSO (C). (A, C) Cell-associated DNA was isolated at indicated times post irradiation and viral DNA measured by qPCR with subsequent normalization to cellular GAPDH. To facilitate pooling of data from multiple experiments, in (A), GAPDH-normalized viral DNA levels measured at 24 and 48h post irradiation were further normalized to those observed in mock-irradiated cells at 0h of reactivation (set to 1). Similarly, in (C), where only a single time point was examined (24h post irradiation), viral DNA levels in unirradiated untreated virus-reactivating cultures were set to 1, and the rest of experimental conditions normalized to that value. (B) Frequency of MHV68 reactivation was measured using limiting dilution assay. Data were pooled from 2-3 independent experiments. *p<0.05.

Figure 6. ATM expression attenuates levels of reactive oxygen species in MHV68-infected macrophages

M-Cre positive and negative primary macrophages were derived from bone marrow of littermates. Macrophages were infected at an MOI of 10 or mock-treated. Hydroethidine (HE) was added at 7 or 23 h post infection at 10μM. Macrophages were harvested at 8 (A) or 24 (B) hours post infection and superoxide determined by measuring levels of 2-hydroxyethidium (2-OH-E⁺) using HPLC. 2-OH-E⁺ levels were normalized to protein levels for each sample. Finally, baseline levels observed in mock-infected control macrophages were set at 1 for each examined time point, and the rest of experimental samples within a particular time point normalized to the corresponding mock. Data were pooled from 2-3 independent experiments.

Figure 7. Effect of myeloid cell-specific ATM deficiency on long-term MHV68 latency

Cre positive or negative littermates were intranasally infected with MHV68 as in Fig. 2. Splenocytes and PEC were harvested at 42 days post infection and pooled from 3-5 mice in each experimental group. Frequency of reactivation (A, B) and MHV68 genome positive cells (C, D) was determined using limiting dilution assays. Data were pooled from 2-3 independent experiments. (E) Representative flow diagram of sorted macrophages and non-macrophage peritoneal cells harvested from M-Cre positive MHV68-infected mice. (F) Western analysis of sorted populations presented in panel C. (G) Working model. A combination of proviral and antiviral functions of ATM balances host-gammaherpesvirus interaction in vivo. ATM restricts gammaherpesvirus reactivation by promoting the development of virus-specific adaptive immune response. In parallel, ATM expression within the myeloid cells facilitates gammaherpesvirus reactivation.