Influenza induces lung lymphangiogenesis independent of YAP/TAZ activity in lymphatic endothelial cells

Abstract

The lymphatic system consists of a vessel network lined by specialized lymphatic endothelial cells (LECs) that are responsible for tissue fluid homeostasis and immune cell trafficking. The mechanisms for organ-specific LEC responses to environmental cues are not well understood. We found robust lymphangiogenesis during influenza A virus infection in the adult mouse lung. We show that the number of LECs increases twofold at 7 days post-influenza infection (dpi) and threefold at 21 dpi, and that lymphangiogenesis is preceded by lymphatic dilation. We also show that the expanded lymphatic network enhances fluid drainage to mediastinal lymph nodes. Using EdU labeling, we found that a significantly higher number of pulmonary LECs are proliferating at 7 dpi compared to LECs in homeostatic conditions. Lineage tracing during influenza indicates that new pulmonary LECs are derived from preexisting LECs rather than non-LEC progenitors. Lastly, using a conditional LEC-specific YAP/TAZ knockout model, we established that lymphangiogenesis, fluid transport and the immune response to influenza are independent of YAP/TAZ activity in LECs. These findings were unexpected, as they indicate that YAP/TAZ signaling is not crucial for these processes.

Fig. 1

Lymphatic vessels dilate during influenza infection. (A) VEGFR3-positive lymphatic vessels (brown; also denoted with red stars) visualized via IHC are more dilated in influenza-infected lungs than in controls. (B) For each mouse, 20 lymphatic vessel diameters were measured using 3 separate histologic sections; mean and SD are shown for n = 3 mice per condition per time point (3-, 7-, and 21-dpi). Statistical significance by unpaired two tailed t-test with Welch’s correction.

Fig. 2

Influenza infection induces pulmonary lymphangiogenesis. (A) PROX1-positive LEC nuclei (brown; also denoted with black arrowheads) visualized via IHC are more numerous in influenza-infected lungs than in controls at 7 and 21 dpi. (B) For each mouse, LEC nuclei were quantified using 3–5 tissue sections from influenza-infected and control mice; mean and SD are shown for n = 3–5 mice per condition per time point. Statistical significance by unpaired one tailed t-test with Welch’s correction.

Fig. 3

LECs proliferate in response to influenza infection. (A) C57Bl/6 mice were administered saline or PR8 followed by EdU injections, and lungs were harvested at 7 dpi followed by IF staining for PROX1 and EdU. White arrowheads indicate PROX1+/EdU+ cells. (B) The number of PROX1 + LECs that costained for EdU was quantified; mean and SD are shown for n = 9 mice per condition. Statistical significance by unpaired two-tailed Mann Whitney test. (C) Prox1-CreER^T2/TdTomato mice were administered tamoxifen prior to saline or PR8 administration, and lungs were harvested at 7 dpi followed by IF staining for PROX1 and tdTomato. White arrowheads indicate PROX1+/TdTomato+ cells. (D) The number of PROX1 + LECs that costained for tdTomato was quantified; mean and SD are shown for n = 5 mice per condition. Statistical significance by unpaired two-tailed Mann Whitney test.

Fig. 4

Lymphatic fluid drainage from the lung is enhanced during influenza infection. (A) Lung wet to dry weight ratios were obtained at 7 dpi from C57Bl/6 mice administered saline or PR8; mean and SD are shown for n = 9 mice per condition. Statistical significance by unpaired two-tailed Mann Whitney test. (B) Example histologic section of a mediastinal LN stained for VEGFR3 (red) and Dextran-488 (green). (C) Fluorometric quantification of Dextran-488 in mediastinal LN 15 min after instillation into left lung lobe. Mean and SD are shown for n = 8–10 mice per condition. Statistical significance by unpaired two-tailed Mann Whitney test.

Fig. 5

There is no significant difference in lymphatic histologic scoring or lymphatic fluid drainage from the lung between Cre(−) and Cre(+) YAP/TAZ^△LEC littermates at baseline or during influenza. (A) Experimental design; image created in BioRender. (B,C) VEGFR3-positive lymphatic vessels (brown, also denoted with red stars; airway, AW; blood vessel, BV) visualized via IHC in Cre(−) or Cre(+) littermates had similar diameters at baseline and during influenza at 7 or 16 dpi. For each mouse, 20 lymphatic vessel diameters were measured using 3 separate histologic sections; mean and SD are shown for n = 3 mice per condition per time point. Statistical significance by unpaired two-tailed Mann Whitney test. (D–F) PROX1-positive LEC nuclei (brown) were visualized via IHC at baseline or during influenza at 7 or 16 dpi for Cre(−) and Cre(+) littermates. For each mouse, LEC nuclei were quantified using 3 tissue sections from influenza-infected and control mice; mean and SD are shown for n = 3 mice per condition per time point. Statistical significance by unpaired two tailed t-test with Welch’s correction. (G) Prox1 and Flt4 mRNA was quantified from left lobes collected from Cre(−) or Cre(+) littermates at 7 dpi after influenza infection. Median and IQR are shown for n = 7–17 mice per group. Statistical significance by unpaired two-tailed Mann Whitney test. (H) Lung wet to dry weight ratios were obtained at 7 dpi from Cre(−) or Cre(+) littermates administered saline or PR8 IT instillations; mean and SD are shown for n = 5–16 mice per condition per timepoint. Statistical significance by unpaired two-tailed Mann Whitney test. (I–J) Fluorometric quantification of Dextran-488 in mediastinal LN 50 min (saline) or 15 min (influenza) after IT instillation into left lung lobe of Cre(−) or Cre(+) littermates. Mean and SD are shown for n = 3–10 mice per condition per timepoint. Statistical significance by unpaired two-tailed Mann Whitney test.

Fig. 6

There is no significant difference in infection severity between Cre(−) and Cre(+) YAP/TAZ^△LEC littermates after influenza challenge. (A) Weight change for C57Bl/6 mice and Cre(−) and Cre(+) YAP/TAZ^△LEC littermates after saline or influenza challenge. Mean and SEM are shown. (B) Influenza PFU recovered from left lobes of Cre(−) or Cre(+) littermates at 7 dpi after influenza challenge. Median and IQR are shown for n = 12 mice per group. Statistical significance by unpaired two-tailed Mann Whitney test. (C) Nucleoprotein (NP) mRNA was quantified from left lobes collected from Cre(−) or Cre(+) littermates at 7 dpi after influenza infection. Median and IQR are shown for n = 16–17 mice per group. Statistical significance by unpaired two-tailed Mann Whitney test. (D,E) Histopathological scoring by a blinded veterinary pathologist was performed using H&E-stained lung sections from Cre(−) or Cre(+) littermates at 7 dpi after influenza infection; representative sections are shown. Mean and SD are shown for n = 3 mice per group. Statistical significance by unpaired two-tailed Mann Whitney test.

Fig. 7

There is no significant difference in mLN immunophenotype between Cre(−) and Cre(+) YAP/TAZ^△LEC littermates after influenza challenge. An 11-color flow cytometry panel was used to determine the immunophenotype of the mLN of Cre(−) or Cre(+) littermates at 2 dpi (A) and 7 dpi (B). After the exclusion of doublets and debris, dead cells were excluded using live/dead staining. Immune cells were identified using the pan-hematopoietic marker CD45 (including separate stains for intravascular vs tissue resident CD45(+) cells). The panel included markers for T-cells (CD4, CD8), B-cells (CD19) as well as dendritic cell subtypes (Ly6C, CD11b and CD103 positive and negative subsets of CD11c(+) cells). Monocytes, macrophages and granulocytes were classified generally using Ly6C, CD64, Ly6G and Siglec F. Mean and SD are shown for n = 3–5 mice per group per timepoint.