Human Cytomegalovirus Delays Neutrophil Apoptosis and Stimulates the Release of a Prosurvival Secretome

Abstract

Human cytomegalovirus (HCMV) is a major cause of viral disease in the young and the immune-suppressed. At sites of infection, HCMV recruits the neutrophil, a cell with a key role in orchestrating the initial immune response. Herein, we report a profound survival response in human neutrophils exposed to the clinical HCMV isolate Merlin, but not evident with the attenuated strain AD169, through suppression of apoptosis. The initial survival event, which is independent of viral gene expression and involves activation of the ERK/MAPK and NF-κB pathways, is augmented by HCMV-stimulated release of a secretory cytokine profile that further prolongs neutrophil lifespan. As aberrant neutrophil survival contributes to tissue damage, we predict that this may be relevant to the immune pathology of HCMV, and the presence of this effect in clinical HCMV strains and its absence in attenuated strains implies a beneficial effect to the virus in pathogenesis and/or dissemination. In addition, we show that HCMV-exposed neutrophils release factors that enhance monocyte recruitment and drive monocyte differentiation to a HCMV-permissive phenotype in an IL-6-dependent manner, thus providing an ideal vehicle for viral dissemination. This study increases understanding of HCMV-neutrophil interactions, highlighting the potential role of neutrophil recruitment as a virulence mechanism to promote HCMV pathology in the host and influence the dissemination of HCMV infection. Targeting these mechanisms may lead to new antiviral strategies aimed at limiting host damage and inhibiting viral spread.

Keywords: apoptosis; human cytomegalovirus; monocyte; neutrophil; polymorphonuclear leukocyte.

Figure 1

Human cytomegalovirus (HCMV) inhibits the rate of constitutive neutrophil apoptosis. (A) Morphological analysis of neutrophil apoptosis in uninfected control (Mock) and infected (HCMV strain Merlin) cells at 20 h postinfection, shown with GM-CSF-treated positive control (MOI 3, n = 9), with (B) representative photomicrographs of cells with apoptotic cells indicated by arrows (magnification ×1,000). (C) Flow cytometry analysis of neutrophil apoptosis in cells infected with Merlin, by annexin V binding at 20 h (MOI 3, n = 7), with (D) representative flow cytometry plots shown. (E) Scanning electron microscopy (EM) images of neutrophils infected with Merlin for 20 h versus control, with smooth apoptotic cells indicated by arrow (bar represents 20 μm). (F) Colorimetric assay for caspase-3-activity performed at 0, 6, and 18 h on Mock (−) or HCMV strain Merlin (+) infected cells (MOI 3, n = 3). (G) Neutrophils infected with Merlin at an MOI of 3, 1, or 0.3 were assayed for apoptosis by morphology (n = 5–6) and by (H) flow cytometry (n = 3–5) at 20 h. (I) Neutrophils infected with HCMV strain TB40/e at an MOI of 3, 1, or 0.3 were assayed for apoptosis by morphology at 20 h (n = 3–8). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 2

Human cytomegalovirus (HCMV) promotion of neutrophil survival is independent of viral gene expression. (A) HCMV Merlin inoculum was ultracentrifuged at 180,000 g for 2 h, and the supernatant was removed from the virion pellet and used to infect neutrophils (virion-free inoculum). Apoptosis was assessed at 20 h by morphology (n = 3) compared to mock-infected cells and cells infected with non-centrifuged Merlin inoculum (Merlin). (B) RNA or DNA was isolated from neutrophils and RNA from human foreskin fibroblasts (HFFs) 18 h after infection with HCMV strain Merlin, and RNA was subjected to reverse transcription, and all samples were then amplified in an IE72 PCR or GAPDH PCR control before loading onto the gel. (C) qPCR showing gene expression levels of IE72 RNA isolated at 6 h from neutrophils infected with Merlin versus control, relative to Merlin-infected fibroblasts (MOI 3, n = 3). (D) Neutrophils were mock infected or incubated with HCMV for 24 h and then stained for immediate early (IE) antigen, shown with Hoescht nuclear counter stain and Merlin-infected HFFs as a positive control. (E) Morphological (n = 6) and (F) flow cytometric analysis (n = 5) of neutrophil apoptosis in cells infected with Mock or UV-inactivated Merlin at 20 h (MOI 3). Data represent mean ± SE, **p < 0.01, ***p < 0.001.

Figure 3

Human cytomegalovirus (HCMV)-stimulated neutrophil survival does not require internalization and does not signal through known receptors for HCMV. (A) Detection of viral particles (arrows) by transmission electron microscopy both outside and (inset) within HCMV strain Merlin-infected neutrophils 30 min postinfection (bar represents 500 nm). (B) Cytochalasin D or latrunculin A (10 μM) pretreatment did not influence neutrophil apoptosis following infection, compared with vehicle control (MOI 3, n = 3). (C) Morphological analysis of neutrophil apoptosis at 18 h in cells preincubated with the receptor tyrosine kinase inhibitor 324841 prior to Mock or Merlin infection (MOI 3, n = 3). (D) Morphological analysis of neutrophil apoptosis at 18 h in neutrophils pretreated with the CD13 inhibitors actinonin or anti-CD13 clone WM15 before Mock or Merlin infection (MOI 3, n = 3). Neutrophils were incubated with inhibitors of (E) TLR2 (n = 3) or (F) TLR4 (n = 3) prior to infection with Merlin (MOI 3), and apoptosis was assessed by morphology at 20 h. Dose–response curves are shown measuring neutrophil apoptosis at 20 h by cell morphology and flow cytometry using annexin V for (G) the TLR2 agonist Pam3CSK4 (n = 1) and (H) the TLR4 agonist LPS (n = 1). (I) Morphological analysis of neutrophil apoptosis in mock and Merlin-infected ultrapure neutrophils at 20 h (MOI 1, n = 3). Data represent mean ± SE, *p < 0.05. NS, not significant.

Figure 4

Human cytomegalovirus (HCMV)-induced neutrophil survival signals through multiple pathways. (A) Neutrophils were preincubated with the PI3K inhibitors LY294002 or wortmannin, the p38-MAPK inhibitor SB203580 or the JNK inhibitor SP600125 prior to infection with Merlin, and apoptosis was assessed by morphology at 20 h (MOI 3, n = 3–5). (B) Western blot analysis for total and phosphorylated AKT in neutrophils incubated with mock or HCMV at 5, 10, 15, and 20 min (MOI 3). (C) Neutrophils were preincubated with the NF-κB inhibitors IKK inhibitor VII or BAY-11-7085 (3 or 10 μM), with or without addition of the ERK1/2 inhibitor U0126, prior to HCMV infection, and apoptosis was assessed by flow cytometry using annexin V at 20 h (MOI 3, n = 2–4). (D) Neutrophils were preincubated with the ERK1/2 inhibitors PD98059 or U0126 (1, 10, or 20 μM), with or without the addition of BAY-11-7085, prior to infection with HCMV, and apoptosis was assessed by flow cytometry using annexin V at 20 h (MOI 3, n = 2–4). (E) Western blot analysis for ERK1/2 phosphorylation or IκB degradation was performed on mock or Merlin-infected neutrophils at 0, 5, 10, 15, 20, 30, and 40 min postinfection (MOI 3). (F) Western blot analysis for MCL-1 and Bax expression performed at 0 or 6 h after infection with mock or HCMV (MOI 3). Data represent mean ± SE, *p < 0.05, **p < 0.01, and ***p < 0.001.

Figure 5

Human cytomegalovirus (HCMV)-induced neutrophil survival is viral strain specific. (A) The ULb′ region of the HCMV genome containing viral genes UL133–UL151 is lost in AD169 and replaced with inverted repeat RL14-1. (B) Neutrophils were incubated with the laboratory-adapted HCMV strain AD169 for 20 h, and apoptosis was assessed by cell morphology and (C) flow cytometry, compared to mock-infected neutrophils or neutrophils incubated with the clinical HCMV strain Merlin (MOI 3, n = 3–4). Neutrophils were incubated with wild-type Merlin or mutant Merlin strains lacking one of three genes from the ULb′ region of the HCMV genome known to code for virion components, and apoptosis was assessed at 20 h by (D) morphology and (E) flow cytometry (MOI 3, n = 3–4). Neutrophils were incubated with mutant strains lacking one of five ULb′ genes potentially coding for virion components, and apoptosis was assessed at 20 h by (F) morphology and (G) flow cytometry (MOI 3, n = 3). Neutrophils were incubated with wild-type Merlin versus Merlin lacking the UL131A gene, and apoptosis was assessed at 20 h by (H) morphology and (I) flow cytometry (MOI 3, n = 3). Data represent mean ± SE. **p < 0.01, ***p < 0.001, and ****p < 0.0001. NS, not significant.

Figure 6

Human cytomegalovirus (HCMV)-exposed neutrophils produce a pro-survival secretome. Freshly isolated neutrophils were cultured with control media or supernatants from mock or Merlin-infected neutrophils (Merlin SN) collected at 20 h, and apoptosis was determined by (A) morphology and (B) flow cytometry at 20 h (MOI 3, n = 3). (C) Supernatant from mock or Merlin-infected monocytes was added to freshly isolated neutrophils and apoptosis determined by morphology at 18 h (MOI 3, n = 3). (D) Supernatant from mock or AD169-infected neutrophils was added to freshly isolated neutrophils and apoptosis determined by morphology at 18 h (MOI 3, n = 3). (E) Supernatant harvested from mock and HCMV-infected neutrophils at 20 h was centrifuged at 130,000 g for 2 h, and residual virus in the ultracentrifuged pellet and supernatant (UC HCMV SN) was assessed by transferring to fibroblasts for 24 h prior to staining for HCMV immediate early (IE) antigen, compared to mock SN and non-ultracentrifuged HCMV supernatant (HCMV SN) (representative image shown). (F) The ultracentrifuged supernatant was transferred to fresh neutrophils and survival assessed by morphology at 20 h compared to non-ultracentrifuged supernatant (n = 4–5, MOI 3). Data represent mean ± SE. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 7

The human cytomegalovirus (HCMV)-exposed neutrophil secretome is high in IL-8. (A) MesoScale assay plate with three representative images, with each spot from three 10-plex wells representing one cytokine. The surrounding rings are imaging artifacts. (B) Results from MesoScale assay of mock and HCMV-infected (strain Merlin and AD169) neutrophil supernatants identified significant upregulation of IL-8, IL-6, TNFα, MIP-1α, IL-13, and IL-10 and downregulation of (C) MDC and IL-16 (MOI 3, n = 5). (D) IL-8 levels were measured in supernatants from mock- or HCMV-infected neutrophils at 20 h by ELISA (MOI 3, n = 5). (E) Following incubation with neutralizing antibodies against IL-8, supernatants were added to neutrophils and (F) apoptosis by morphology determined at 20 h, shown relative to control cells (MOI 3, n = 2–4). Data represent mean ± SE. *p < 0.05, **p < 0.01.

Figure 8

Supernatants from human cytomegalovirus (HCMV)-exposed neutrophils promote monocyte migration and activation to a permissive phenotype in an IL-6-dependent manner. (A) Monocytes were cultured in transwell plates in the presence of control media, or media from mock- or HCMV-infected neutrophils, or from LPS-activated monocytes as a positive control, and migration assessed by microscopy after 2 h (n = 3). (B) Autologous donor monocytes were cultured in control media, mock-infected neutrophil media, or Merlin-infected neutrophil media, and freshly infected with Merlin, alongside IL-4/GM-CSF-induced dendritic cells (DCs) as a positive control. Cells were stained for immediate early (IE) gene expression at 24 h and counterstained with Hoechst nuclear dye, and (C) relative levels of infection calculated from 10 fields of view (MOI 3, n = 3). (D) Neutralization of IL-6 in supernatant from mock or Merlin-infected neutrophils significantly inhibited the number of monocytes positive for IE gene expression (n = 3). (E) IL-6 levels by ELISA in viral inoculum and in mock-infected and Merlin-infected ultrapure neutrophils (purity >99%) (n = 2). (F) Monocytes were incubated with media (monocyte) or supernatants from mock- or Merlin-infected neutrophils for 3 days. Media was then removed, and allogeneic CD4+ T cells were added to each well along with phytohemagglutinin (PHA) to all conditions except control. T cell proliferation after 6 days of co-culture was assessed by counting and shown relative to T cells cultured alone (control) or in the presence of PHA alone (PHA) (n = 2). (G) Representative flow cytometry plots of autologous monocytes cultured in X-vivo-15 media (control), mock-infected neutrophil media, or HCMV-infected neutrophil media (MOI 3) for 3 days before staining for MHC class I and II. Data represent mean ± SE. *p < 0.05.