Aryl hydrocarbon receptor activation alters immune cell populations in the lung and bone marrow during coronavirus infection

Abstract

Respiratory viral infections are one of the major causes of illness and death worldwide. Symptoms associated with respiratory infections can range from mild to severe, and there is limited understanding of why there is large variation in severity. Environmental exposures are a potential causative factor. The aryl hydrocarbon receptor (AHR) is an environment-sensing molecule expressed in all immune cells. Although there is considerable evidence that AHR signaling influences immune responses to other immune challenges, including respiratory pathogens, less is known about the impact of AHR signaling on immune responses during coronavirus (CoV) infection. In this study, we report that AHR activation significantly altered immune cells in the lungs and bone marrow of mice infected with a mouse CoV. AHR activation transiently reduced the frequency of multiple cells in the mononuclear phagocyte system, including monocytes, interstitial macrophages, and dendritic cells in the lung. In the bone marrow, AHR activation altered myelopoiesis, as evidenced by a reduction in granulocyte-monocyte progenitor cells and an increased frequency of myeloid-biased progenitor cells. Moreover, AHR activation significantly affected multiple stages of the megakaryocyte lineage. Overall, these findings indicate that AHR activation modulates multiple aspects of the immune response to a CoV infection. Given the significant burden of respiratory viruses on human health, understanding how environmental exposures shape immune responses to infection advances our knowledge of factors that contribute to variability in disease severity and provides insight into novel approaches to prevent or treat disease.NEW & NOTEWORTHY Our study reveals a multifaceted role for aryl hydrocarbon receptor (AHR) signaling in the immune response to coronavirus (CoV) infection. Sustained AHR activation during in vivo mouse CoV infection altered the frequency of mature immune cells in the lung and modulated emergency hematopoiesis, specifically myelopoiesis and megakaryopoiesis, in bone marrow. This provides new insight into immunoregulation by the AHR and extends our understanding of how environmental exposures can impact host responses to respiratory viral infections.

Keywords: emergency hematopoiesis; hematopoietic stem and progenitor cells; megakaryopoiesis; myeloid cells; respiratory virus infection.

Graphical abstract

Figure 1.

Aryl hydrocarbon receptor (AHR) activation increased morbidity and dampened the adaptive immune response to mouse hepatitis virus 1 (MHV-1) infection. BALB/c mice (6 wk of age, female) were administered either peanut oil vehicle control (Vehicle) or 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD; 10 µg/kg body wt) by gavage 1 day before infection. Infected mice were inoculated intranasally with 5 × 10⁴ plaque-forming units (PFU) of MHV-1. Immunologically naive mice (denoted by N on graphs) were not infected and were euthanized 24 h after vehicle or TCDD administration. A: Cyp1a1 expression in the liver and whole blood of mice exposed to vehicle or TCDD. Infected mice were harvested 6 days after infection. Cyp1a1 expression was quantified by RT-qPCR. Fold change was calculated by the $2^{\mathrm{-}\Delta \Delta {\mathrm{C}}_{\mathrm{T}}}$ method (where C_T is threshold cycle). *P ≤ 0.05, comparing the vehicle and TCDD groups by Student’s t test. B: % body weight change relative to the day of infection. The 2-way ANOVA P value is displayed. C: viral burden (PFU/mL) in livers on the indicated days relative to infection. D: single-cell suspensions of lung-derived immune cells were prepared on the indicated days relative to infection. The number of CD8⁺ cytotoxic T lymphocytes (CTLs; CD3ε⁺CD45R⁻CD8a⁺CD44^hiCD62L^lo) was determined by flow cytometry. E: anti-MHV-1 antibody levels in serially diluted serum were determined by ELISA. Serum was obtained from uninfected [immunologically naive (N)] mice and from mice 10 days after infection. The graph depicts the mean relative level of MHV-1-specific IgG at all dilutions. OD₄₀₅, optical density at 405 nm. Numerical data and P values from post hoc tests for CTL (D) and antibody (E) measurements are in Supplemental Tables S2 and S3. In D, *P ≤ 0.05, comparing the treatment groups on the indicated day after infection [2-way ANOVA, Tukey honestly significant difference (HSD)]; #P ≤ 0.05, comparing the uninfected mice (N) and infected mice (day 2 relative to infection) within either the vehicle- or TCDD-treated group (2-way ANOVA, Dunnett’s). In E, *P ≤ 0.05, comparing the treatment groups at the indicated dilution (2-way ANOVA, Tukey HSD). In B, D, and E, the 2-way ANOVA P value is displayed on each graph. Data are from a time course experiment with 34 mice in each treatment group. Cellular assessments at each point in time relative to infection had 4–6 mice in each treatment group (4 immunologically naive and 6 infected). All data are shown as means ± SE.

Figure 2.

Monocytes were reduced in 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)-treated animals during mouse hepatitis virus 1 (MHV-1 infection). BALB/c mice were dosed with vehicle or TCDD (p.o.) 1 day before infection (intranasal) with MHV-1. Single-cell suspensions of lung-derived immune cells were obtained and used for flow cytometry. A: graph shows the mean number of leukocytes (CD45⁺ cells) in the lung before and after infection. B: gating strategy used to identify leukocytes (CD45⁺ cells), neutrophils (CD45⁺Ly6G⁺CD11b⁺ cells), macrophages (CD45⁺Ly6G⁻F4/80⁺ cells), and monocytes (CD45⁺Ly6G⁻CD11b⁺Ly6C⁺ cells) (58, 59). C–H: line graphs depict the number (C) and percentage (F) of neutrophils, the number (D) and percentage (G) of macrophages, and the number (E) and percentage (H) of monocytes. In A, C, and E, the symbols for the mean values in uninfected vehicle- and TCDD-treated mice overlap. Data are shown as means ± SE with 6 mice per group per day on each day after infection and 4 uninfected mice per treatment group (denoted by N). *P ≤ 0.05, comparing treatment groups on the specified day after infection [2-way ANOVA, Tukey honestly significant difference HSD)]. #P ≤ 0.05, comparing uninfected mice (N) to infected mice (day 2 relative to infection) within either the vehicle- or TCDD-treated group (2-way ANOVA, Dunnett’s). Supplemental Tables S2 and S4 contain all numerical data and P values from 2-way ANOVA and post hoc tests.

Figure 3.

Aryl hydrocarbon receptor (AHR) activation shifted the frequency of Ly6C^lo and Ly6C^hi monocytes in the infected lung. BALB/c mice were given either vehicle or 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) 1 day before infection with mouse hepatitis virus 1 (MHV-1). There were 6 mice per treatment group per day relative to infection and 4 uninfected mice per treatment group (denoted by N). Single-cell suspensions of lung-derived immune cells were prepared on the indicated days relative to infection. Graphs depict the mean percentage (±SE) of monocytes that were Ly6C^lo (A) and Ly6C^hi (B). The 2-way ANOVA P value (treatment × time) is indicated on each graph. *P ≤ 0.05, comparing across treatment groups on the indicated day after infection [Tukey honestly significant difference (HSD)]. #P ≤ 0.05, comparing between uninfected (N) group and infected (day 2 relative to MHV-1 infection) group within either the vehicle-treated animals or TCDD-treated animals (Dunnett’s). All numerical data and P values from the 2-way ANOVA and post hoc tests are in Supplemental Tables S2 and S4.

Figure 4.

Aryl hydrocarbon receptor (AHR) activation altered specific cell populations within the mononuclear phagocyte system (MPS) during mouse hepatitis virus 1 (MHV-1) infection. BALB/c mice (6 mice per treatment group per day) were given vehicle or 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) 1 day before infection. Uninfected, or immunologically naive (N), mice were also included, with 4 mice per treatment group. Single-cell suspensions of lung-derived immune cells were prepared on the indicated days relative to infection. After gating on CD45⁺Ly6G⁻CD45R⁻, cells were subdivided based on expression of Ly6C, F4/80, CD11c, and CD11b with Boolean gating. A depicts 15 subpopulations (Pop) that expressed at least 1 of these 4 markers and indicates the number assigned to each combination of markers. B–M: graphs depict the mean percentage (±SE) of all 15 subpopulations, broadly organized as monocytes (B–E), macrophages (F–I), and dendritic cells (DCs; J–M). N–P: graphs depict the mean percentage (± SE) of 3 subpopulations that express a combination of markers not phenotypically categorized as monocyte, macrophage, or DC. Q: graph depicts the mean percentage (±SE) of each subpopulation in immunologically naive (N) vehicle- or TCDD-treated mice. R–V: graphs depict the mean percentage (±SE) of each subpopulation separated by day relative to infection (2, 6, 8, 10, and 14). Percentage values represent the percentage of each population out of total leukocytes in the lung. In E, I, and M–P the symbols representing the mean values in immunologically naive vehicle- and TCDD-treated mice (white and blue circles) overlap. The number of all 15 subpopulations over time is in Supplemental Fig. S2. The 2-way ANOVA (treatment × time) P value is represented in D, F, H, and K–N. *P ≤ 0.05, comparing the treatment groups at single points in time [2-way ANOVA, Tukey honestly significant difference (HSD)].

Figure 5.

Aryl hydrocarbon receptor (AHR) activation altered primitive hematopoietic stem and progenitor cell populations during mouse hepatitis virus 1 (MHV-1) infection. One day before infection, BALB/c mice were given vehicle or 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) by gavage. Bone marrow cells from 2 hind legs were harvested from 6 infected mice per group per day and 4 uninfected mice (denoted by N) from each treatment group. A: graph shows the mean number of bone marrow cells before and after infection. B: gating strategy used to identify hematopoietic stem cell (HSC) and multipotent progenitor (MPP) subsets by flow cytometry. C and D: the percentage (C) and number (D) of hematopoietic stem and progenitor cells (HSPCs; Lineage⁻cKit⁺Sca1⁺ cells). E and F: the percentage (E) and number (F) of HSCs (Lineage⁻cKit⁺Sca1⁺CD34⁻CD135⁻CD48⁻CD150⁺). G–J: the percentage of multipotent progenitor cells (MPP^M; Lineage⁻cKit⁺Sca1⁺CD34⁺CD135⁻CD48⁻CD150⁺) (G), lymphoid-biased progenitor cells (MPP^Ly; Lineage⁻cKit⁺Sca1⁺CD34⁺CD135⁺CD48⁺CD150⁻) (H), megakaryocyte-erythroid-biased progenitor cells (MPP^Meg/E; Lineage⁻cKit⁺Sca1⁺CD34⁺CD135⁻CD48⁺CD150⁺) (I), and myeloid-biased progenitor cells (MPP^GM; Lineage⁻cKit⁺Sca1⁺CD34⁺CD135⁻CD48⁺CD150⁻) (J). K–N: the number of MPP^M (K), MPP^Ly (L), MPP^Meg/E (M), and MPP^GM (N). Note that in A and N the white and blue circles representing the mean (±SE) values in uninfected vehicle- and TCDD-treated mice overlap. All data are shown as means ± SE, and the P values in E, F, and H–J are derived from the 2-way ANOVA. *P ≤ 0.05, comparing vehicle and TCDD treatment at single point in time [2-way ANOVA, Tukey honestly significant difference (HSD)]. #P ≤ 0.05, comparing uninfected (N) to means on day 2 after infection within treatment group (2-way ANOVA, Dunnett’s). Numerical data and statistical output from ANOVA and post hoc tests are in Supplemental Tables S5 and S6).

Figure 6.

Aryl hydrocarbon receptor (AHR) activation altered lineage-committed progenitor cells during mouse hepatitis virus 1 (MHV-1) infection. BALB/c mice were given either vehicle or 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) 1 day before infection with MHV-1. Bone marrow cells were collected on the indicated days relative to infection. There were 6 infected mice per group per day and 4 uninfected mice per group (denoted as N). A: the gating strategy used to identify common lymphoid progenitor cells (CLPs), granulocyte-monocyte progenitor cells (GMPs), and megakaryocyte progenitor cells (MkPs) by flow cytometry. B–G: the number (B) and percentage (C) of CLPs (Lineage⁻cKit^lowSca1^lowCD135⁺CD127⁺), the number (D) and percentage (E) of MkPs (Lineage⁻cKit⁺Sca1⁻CD150⁺CD41⁺), and the number (F) and percentage (G) of GMPs (Lineage⁻cKit⁺Sca1⁻CD150⁻CD41⁻CD16/32⁺). In B–G, the symbols (white and blue circles) representing the mean (±SE) for vehicle- and TCDD-treated immunologically naive mice overlap. The P values represented in C–G are from the 2-way ANOVA. #P ≤ 0.05, comparing immunologically naive mice (N) to infected mice (day 2 after infection) within treatment group (2-way ANOVA, Dunnett’s). *P ≤ 0.05, comparing vehicle and TCDD treatment at single point in time [2-way ANOVA, Tukey honestly significant difference (HSD)]. All numerical data and additional statistical comparisons can be found in Supplemental Tables S5 and S6.