Influenza A virus elicits peri-vascular adipose tissue inflammation and vascular dysfunction of the aorta in pregnant mice

Abstract

Influenza A virus (IAV) infection during pregnancy initiates significant aortic endothelial and vascular smooth muscle dysfunction, with inflammation and T cell activation, but the details of the mechanism are yet to be clearly defined. Here we demonstrate that IAV disseminates preferentially into the perivascular adipose tissue (PVAT) of the aorta in mice. IAV mRNA levels in the PVAT increased at 1-3 days post infection (d.p.i) with the levels being ~4-8 fold higher compared with the vessel wall. IAV infection also increased Ly6Clow patrolling monocytes and Ly6Chigh pro-inflammatory monocytes in the vessel wall at 3 d.p.i., which was then followed by a greater homing of these monocytes into the PVAT at 6 d.p.i. The vascular immune phenotype was characteristic of a "vascular storm"- like response, with increases in neutrophils, pro-inflammatory cytokines and oxidative stress markers in the PVAT and arterial wall, which was associated with an impairment in endothelium-dependent relaxation to acetylcholine. IAV also triggered a PVAT compartmentalised elevation in CD4+ and CD8+ activated T cells. In conclusion, the PVAT of the aorta is a niche that supports IAV dissemination and a site for perpetuating a profound innate inflammatory and adaptive T cell response. The manifestation of this inflammatory response in the PVAT following IAV infection may be central to the genesis of cardiovascular complications arising during pregnancy.

Fig 1. Influenza A virus (IAV) primarily disseminates into the PVAT of pregnant mice compared to the arterial wall.

Eight-to-twelve-weeks-old pregnant (E12 gestation) C57BL/6 mice were intranasally inoculated with PBS or Hk-x31 (X-31; 10⁴ PFU) for arterial wall and PVAT tissue assessment at 6 hours post infection (h.p.i), 1, 3 and 6 days post infection (d.p.i). (A) Schematic of infection schedule and experiments (created with BioRender.com), the presence of IAV burden in the arterial wall and PVAT was confirmed through qPCR, using a cycle threshold of <31 cycles as a confirmed infection. (B—C) Vessel wall and PVAT gene expression of viral PA normalized to GAPDH or RPS18. (D) Representative immunofluorescence image of the arterial wall of pregnant PBS and Hk-x31 infected mice at 6 h, or 3 and 6 d.p.i labeled with IAV nucleoprotein antibody (green). (E) Representative immunofluorescence image of the PVAT of pregnant PBS at 6 h.p.i and Hk-x31 infected mice at 6 h, or 3 and 6 d.p.i labeled with IAV nucleoprotein antibody (green). Data are represented as mean ± SEM (pregnant PBS, n = 4–8; pregnant X-31, n = 4–8 of at least two to three independent experiments). All fold change calculations of the X-31 group were measured via qPCR, performed against the PBS group within its respective timepoint and normalised against RPS18 (except otherwise stated). Statistical analysis was performed using unpaired t-test against their respective PBS control. * P<0.05.

Fig 2. IAV infection triggers early onset endothelial dysfunction in pregnant mice.

Vascular function was assessed at 6 h and 1 d.p.i in isolated thoracic aortic rings of pregnant mice inoculated with PBS or Hk-x31 (X-31; 10⁴ PFU). (A) Schematic of infection schedule and experiments (created with BioRender.com) (B) Endothelium-dependent vasodilation to acetylcholine (ACh) at 6 h.p.i. (C) 1 d.p.i vascular function assessment to endothelium-dependent vasodilator—ACh. (D) 1 d.p.i vascular reactivity assessment to endothelium independent vasodilator–SNP. Vascular relaxation is calculated as a % of pre-constriction to U-46619 (thromboxane agonist). Data are represented as mean ± SEM (pregnant PBS, n = 4–8; pregnant X-31, n = 4–8 of at least two independent experiments). Statistical analysis was conducted using a two-way analysis of variance (ANOVA) followed by Holm’s Sidak post-hoc multiple comparison. * P<0.01, #P<0.001.

Fig 3. IAV induces an inflammatory and oxidative stress phenotype in both the arterial wall and PVAT in pregnant mice.

Eight-to-twelve-week-old pregnant (E12 gestation) C57BL/6 mice were intranasally inoculated with PBS or Hk-x31 (X-31; 10⁴ PFU) for vessel wall and PVAT tissue inflammatory and oxidative stress mRNA transcripts assessment at 6 h, or 1, 3 and 6 d.p.i. (A) Schematic of infection schedule and experiments (created with BioRender.com) (B—E) Vessel wall and PVAT gene expression of pro-inflammatory cytokines TNF-α and IL-6. (F—G) Antiviral mediator IFN-γ mRNA transcript gene expressions as determined via qPCR in the vessel wall and PVAT. (H—I) Vessel wall and PVAT gene expression of oxidative stress marker NOX2. (J—K) Early immune activation marker CD69 mRNA transcript gene expressions as determined via qPCR in the vessel wall and PVAT. Data are represented as mean ± SEM (pregnant PBS, n = 4–8; pregnant X-31, n = 4–8 of at least two independent experiments). All fold change calculations of the X-31 group were measured via qPCR, performed against the PBS group within its respective timepoint and normalised against RPS18. Statistical analysis was performed using unpaired t-test against their respective PBS control. * P<0.05.

Fig 4. IAV promotes innate inflammation via substantial infiltration of monocytes and neutrophils in PVAT compared to arterial wall in pregnant mice.

Separate single cell suspensions were prepared from vessel wall and PVAT digests from pregnant mice that were inoculated with either PBS or Hk-x31 virus (X-31; 10⁴ PFU) at 3 and 6 d.p.i and quantified via flow cytometry for the following cell subsets. (A) Representative dot plots and quantification showing patrolling monocytes (CD11b⁺Ly6C^low) and pro-inflammatory monocytes (CD11b⁺Ly6C^high) in the vessel wall. (B) Ly6G⁺ neutrophils quantification. Quantification results in the vessel wall are also shown. (C) Representative dot plots and quantification showing patrolling monocytes (CD11b⁺Ly6C^low) and pro-inflammatory monocytes (CD11b⁺Ly6C^high) in the PVAT. (D) Ly6G⁺ neutrophils quantification. Quantification results in the PVAT are also shown. All cell populations were measured as absolute number of CD45⁺ population per 25,000 counting beads. Data are represented as mean ± SEM (pregnant PBS, n = 4–6; pregnant X-31, n = 4–6; of at least two independent experiments). Statistical analysis was performed using unpaired t-test against their respective PBS control. * P<0.05.

Fig 5. IAV drives an increase in global CD3⁺ immune T cell infiltration and activation predominantly in the PVAT and periadventitial space of pregnant mice when compared to the arterial wall.

Representative Immunofluorescence image of pregnant PBS and X-31 mice arterial wall and PVAT labelled with CD3 antibody (red) and counterstained with DAPI (blue). Data are representative of pregnant PBS, n = 5–6; pregnant X-31, n = 4–6; of at least two independent experiments.

Fig 6. IAV drives immune T cell infiltration and activation predominantly in the PVAT of pregnant mice when compared to the arterial wall.

Separate single cell suspensions were prepared from vessel wall and PVAT digests from pregnant mice that were inoculated with either PBS or Hk-x31 virus (X-31; 10⁴ PFU) at 3 and 6 d.p.i and quantified via flow cytometry for the following cell subsets. (A) Vessel wall analysis showing CD4⁺ and CD8⁺ T cells at 3 and 6 d.p.i. (B—C) Representative X-31 dot plots and quantification showing PBS and X-31 CD44⁺ and CD69⁺ activated CD4⁺ and CD8⁺ T cells in the vessel wall. (D) PVAT analysis showing CD4⁺ and CD8⁺ T cells at 3 and 6 d.p.i. (E—F) Representative X-31 dot plots and quantification showing PBS and X-31 CD44⁺ and CD69⁺ activated CD4⁺ and CD8⁺ T cells in the PVAT. All cell populations are measured as absolute number of CD45⁺ population per 25,000 counting beads. Data are represented as mean ± SEM (pregnant PBS, n = 3–6; pregnant X-31, n = 4–6; of at least two independent experiments). Statistical analysis was performed using unpaired t-test against their respective PBS control. * P<0.05.