Antiviral prevention of sepsis induced cytomegalovirus reactivation in immunocompetent mice

Abstract

Introduction: Immunocompetent patients can reactivate latent cytomegalovirus (CMV) during critical illness and reactivation is associated with significantly worse outcomes. Prior to clinical trials in humans to prove causality, we sought to determine an optimal antiviral treatment strategy.

Methods: Mice latently infected with murine CMV (MCMV) received a septic reactivation trigger and were randomized to receive one of four ganciclovir regimens or saline. Lungs were evaluated for viral transcriptional reactivation and fibrosis after each regimen. Influences of ganciclovir on early sepsis-induced pulmonary inflammation and T-cell activation were studied after sepsis induction.

Results: All ganciclovir regimens reduced measurable MCMV transcriptional reactivation, and 10mg/day for 7 or 21 days was most effective. Lower dose (5mg/kg/day) or delayed therapy was associated with significant breakthrough reactivation. Higher doses of ganciclovir given early were associated with the lowest incidence of pulmonary fibrosis, and delay of therapy for 1 week was associated with significantly worse pulmonary fibrosis. Although bacterial sepsis induced activation of MCMV-specific pulmonary T-cells, this activation was not influenced by ganciclovir.

Conclusion: These results suggest that antiviral treatment trials in humans should use 10mg/kg/day ganciclovir administered as early as possible in at-risk patients to minimize reactivation events and associated pulmonary injury.

Figure 1. Cytomegalovirus transcriptional reactivation during bacterial sepsis

Mice latently infected with murine cytomegalovirus (MCMV) had bacterial sepsis induced by cecal ligation and puncture (CLP). After CLP, mice received ganciclovir or saline treatments as indicated. Three weeks after CLP and subsequent antiviral treatment, mouse lung homogenates were evaluated for viral transcriptional reactivation. Representative gels from nested reverse transcription polymerase chain reactions (RT-PCR) from each group are shown. Each lane represents lung tissue results from individual mice. RT results for GAPDH confirm mRNA extraction, and murine cytomegalovirus (MCMV) mRNA represents MCMV gene glycoprotein B transcription. Presence of DNA contamination was evaluated by NO-RT controls for each specimen. Negative and positive refer to technique controls. All RT-PCR evaluations were performed in triplicate.

Figure 2. Ganciclovir treatment influences reactivation and pulmonary injury

Mice latently infected with murine cytomegalovirus (MCMV) had bacterial sepsis induced and received one of the listed ganciclovir regimens or saline (control). A. Three weeks later lungs from surviving mice were evaluated for MCMV transcriptional reactivation. Each bar represents the percentage reactivation (# reactivation/n) for each treatment group. * indicates significant difference from saline controls (p<0.05), and @ indicates significant difference from 10mg/kg 21 days group (p<0.05). All evaluations were performed in triplicate. B. Surviving mice were also evaluated for pulmonary fibrosis. Lungs were fixed, embedded, sectioned, and stained with Gomori’s trichrome. Images were acquired, color segmented, and fibrosis quantitated as percent of pixels. Each bar represents the mean ± standard error for 5–7 mice.

Figure 2. Ganciclovir treatment influences reactivation and pulmonary injury

Mice latently infected with murine cytomegalovirus (MCMV) had bacterial sepsis induced and received one of the listed ganciclovir regimens or saline (control). A. Three weeks later lungs from surviving mice were evaluated for MCMV transcriptional reactivation. Each bar represents the percentage reactivation (# reactivation/n) for each treatment group. * indicates significant difference from saline controls (p<0.05), and @ indicates significant difference from 10mg/kg 21 days group (p<0.05). All evaluations were performed in triplicate. B. Surviving mice were also evaluated for pulmonary fibrosis. Lungs were fixed, embedded, sectioned, and stained with Gomori’s trichrome. Images were acquired, color segmented, and fibrosis quantitated as percent of pixels. Each bar represents the mean ± standard error for 5–7 mice.

Figure 3. Bacterial sepsis induces pulmonary T-cell activation

Flow cytometry for CD8+ T-cells was performed on pulmonary lymphocytes isolated from mice. A. Representative lymphocyte gating and CD8 histograms for murine cytomegalovirus (MCMV) naïve or latently infected mice (non-septic). B. Graphical summary of CD8 percentages in lungs of non septic mice. C. Pulmonary lymphocytes from MCMV latent mice isolated 1 day after bacterial sepsis (CLP) or no sepsis (No CLP) were evaluated for CD8+ activation by colocalizing with activation marker CD43 using flow cytometry. For graphs each data point represents result from one mouse, and bars represent mean ± standard error.

Figure 3. Bacterial sepsis induces pulmonary T-cell activation

Flow cytometry for CD8+ T-cells was performed on pulmonary lymphocytes isolated from mice. A. Representative lymphocyte gating and CD8 histograms for murine cytomegalovirus (MCMV) naïve or latently infected mice (non-septic). B. Graphical summary of CD8 percentages in lungs of non septic mice. C. Pulmonary lymphocytes from MCMV latent mice isolated 1 day after bacterial sepsis (CLP) or no sepsis (No CLP) were evaluated for CD8+ activation by colocalizing with activation marker CD43 using flow cytometry. For graphs each data point represents result from one mouse, and bars represent mean ± standard error.

Figure 4. Ganciclovir does not prevent sepsis induced T-cell activation in lungs

A. Flow cytometry was performed on pulmonary lymphocytes isolated from mice previously infected with murine cytomegalovirus (MCMV) 1 day after induction of sepsis (CLP) or sepsis after ganciclovir pretreatment (CLP/GCV). Cells were incubated with fluorophore labeled anti-CD8 and CD43 antibodies as well as MCMV specific peptide loaded MHC-I tetramers specific for MCMV proteins pp89 and m164. Representative flow scatter plot shows gating for CD8+Tetramer-positive (CD8+TET+) and CD8+Tetramer-negative (CD8+TET−) cells that were subsequently evaluated for activation (CD43^hi). Activation results are summarized graphically for tetramer-positive cells (above) and tetramer negative cells (below). Ganciclovir had no significant impact upon MCMV-specific T-cell activation. Significantly more MCMV specific T-cells were activated (CD43^hi) than non-MCMV-specific T-cells after sepsis (analysis not shown). For graphs each data point represents result from one mouse, and bars represent mean ± standard error. Students T-test was used for comparisons and p-values are shown.

Figure 4. Ganciclovir does not prevent sepsis induced T-cell activation in lungs

A. Flow cytometry was performed on pulmonary lymphocytes isolated from mice previously infected with murine cytomegalovirus (MCMV) 1 day after induction of sepsis (CLP) or sepsis after ganciclovir pretreatment (CLP/GCV). Cells were incubated with fluorophore labeled anti-CD8 and CD43 antibodies as well as MCMV specific peptide loaded MHC-I tetramers specific for MCMV proteins pp89 and m164. Representative flow scatter plot shows gating for CD8+Tetramer-positive (CD8+TET+) and CD8+Tetramer-negative (CD8+TET−) cells that were subsequently evaluated for activation (CD43^hi). Activation results are summarized graphically for tetramer-positive cells (above) and tetramer negative cells (below). Ganciclovir had no significant impact upon MCMV-specific T-cell activation. Significantly more MCMV specific T-cells were activated (CD43^hi) than non-MCMV-specific T-cells after sepsis (analysis not shown). For graphs each data point represents result from one mouse, and bars represent mean ± standard error. Students T-test was used for comparisons and p-values are shown.