Specific pathogen-free, 4–6-week-old female Balb/c mice and influenza A virus strain A/California/07/2009 (H1N1v) were provided by the Institute of Laboratory Animal Science, Peking Union Medical College, China. The experiments were performed in biosafety level 3 facilities in compliance with governmental and institutional guidelines. This study was carried out in accordance with the recommendations of the Chinese National Guidelines for the Care of Laboratory Animals and the Institutional Animal Care and Use Committee of the Institute of Laboratory Animal Science, Peking Union Medical College. The protocol was approved by the Institutional Animal Care and Use Committee (ILAS-PC-2015-016).

Mice were anesthetized and inoculated intranasally either with virus (10² 50% tissue culture infective dose [TCID₅₀] in 50 µl solution per mice) or, in the control group, an equal quantity of phosphate-buffered saline (PBS). The symptom, body weight, and survival rate of the mice were observed daily.

For each mouse, the whole right lung was fixed in 10% formalin for 24 h and then embedded in paraffin for histological examination. The lung tissue sections (4 µm) were deparaffinized and hydrated using xylene and an alcohol gradient and, then, stained with H&E. The histopathology of the lung tissue was observed by light microscopy.

For each mouse, the whole left lung homogenates were used for virus titration tests using endpoint titration in Madin–Darby canine kidney (MDCK) cells, as described previously (36).

The lungs were cut up and subsequently digested in Dulbecco’s Modified Eagle Medium (DMEM) (Gibco, Life Technologies, New York) containing 0.1% collagenase I (Gibco, Life Technologies, New York) at 37°C for 60 min. The tissue suspension after digestion was filtered through a 75-µm strainer and then washed with DMEM. The total lung lymphocytes were centrifuged by density gradient centrifugation using 40% Percoll and 70% Percoll (GE Healthcare, Amersham, UK). The lung lymphocytes were collected from the interface between the 40% and 70% Percoll. The lymphocytes were washed two times with PBS and then suspended in PBS. The spleens were disrupted using tissue grinders with a 75-µm nylon strainer, and the cell suspensions were lysed using lysing buffer (BD Pharm Lyse™, New Jersey). The number of cells in each suspension was adjusted to 2.0 × 10⁶ per tube.

After stripping the fat and connective tissue around the trachea, the exposed trachea was slowly injected into 2 ml PBS. The PBS was then recovered after 1 min and centrifuged at 1,500 rpm for 10 min at 4°C. The supernatant was collected and stored at −80°C.

Lung homogenates were prepared by homogenizing perfused whole lung tissue using an electric homogenizer for 2 min 30 s in 1 ml PBS. The homogenates were centrifuged at 3,000 rpm for 10 min at 4°C. The supernatant was collected and stored at −80°C.

The lymphocytes were stimulated with leukocyte activation cocktail with BD GolgiPlug™ (BD Pharmingen San Diego, CA, USA) for 4 h at 37°C and 5% CO₂ in an incubator. The cells were pre-incubated with purified rat anti-mouse CD16/CD32 (Mouse BD Fc Block™, 2.4G2, BD Pharmingen™) on ice for 10 min.

For the extracellular cell marker analysis, the cells were incubated with the following fluorescein-conjugated antibodies for 30 min: BV510-anti-CD3ε (145-2C11; BD Biosciences), APC-Cy7-anti-CD3ε (145-2C11; BD Biosciences), PE-Cy7-T-cell receptor (TCR)β (H57-597; BD Biosciences), PE-anti-CD4 (RM4-5; BD Biosciences) PerCP-Cy5.5-anti-CD8a (53-6.7; BD Biosciences), APC-anti-TCR γ/δ (GL3; Biolegend, CA, USA), FITC-anti-TCR Vγ1 (2.11; Biolegend), PE-anti-TCR Vγ4 (UC3, Biolegend), PerCP-efluor710-anti-FasL (MFL3; eBioscience), and PE-Cy7-anti-NKG2D (CX5; eBioscience).

For intracellular cytokine staining, the extracellularly stained cells were fixed and permeabilized using BD Cytofix/Cytoperm™ and BD Perm/Wash™ (BD Biosciences) and, then, incubated with the following fluorescein-conjugated antibodies: PE-Cy7-anti-IFN-γ (XMG1.2; eBioscience) and PE-Cy7-anti-IL-17A (17B7; eBioscience). Finally, samples were acquired using the fluorescence-activated cell sorting (FACS) Canto II system (BD Biosciences). The data were analyzed using a Kaluza analysis with Flowjo 6.1 software.

For γδT cell depletion, two mAbs with specificity for different TCR-γ/δ epitopes were used to maximize γδT cell depletion as previously described (37); 125 µg anti-mouse TCR γ/δ mAbs (GL3; Biolegend) and 125 µg anti-mouse TCR γ/δ mAbs (UC7, Biolegend) were injected intraperitoneally (i.p.) 1 day before infection. For the Vγ1⁺γδT and Vγ4⁺γδT subsets depletion, 250 µg anti-mouse TCR Vγ1 mAbs (2.11; Sungene, Tianjin, China) and anti-mouse TCR Vγ4 mAbs (UC3-10A6; Sungene), respectively, were injected i.p. both 3 and 1 days before infection, respectively. Hamster IgG (HTK888, Biolegend) was used as isotype control. Control mice were injected i.p. with 250 µl PBS or hamster IgG 1 day before infection. Flow cytometry was used to evaluate the depletion of γδT cells and the Vγ1⁺γδT and Vγ4⁺γδT subsets, as previously described in detail (for γδT cell depletion) (38).

A multiplex biometric immunoassay was used to measure the cytokines in the serum and BALF according to the manufacturer’s instructions [ProcartaPlex Mouse Cytokine & Chemokine Panel 1 (13 plex); eBioscience]. The levels of cytokines were determined using a multiplex array reader from the Luminex™ Instrumentation System (Bio-Plex Workstation; Bio-Rad Laboratories, San Diego, CA, USA). The concentration was calculated using Bio-Plex Manager Software provided by the manufacturer (39).

Bicinchoninic acid protein assay kit (Beyotime Biotechnology, Jiangsu, China) was used to determine the total protein concentration in the BALF. Briefly, 20 µl BALF and 200 µl BCA were mixed in the wells of a 96-well plate. The microplate was incubated at 37°C for 30 min. The optical density at 562 nm (OD₅₆₂) was measured using a microplate reader (Thermo Lab systems Multiskan MK3, Finland). An automatic biochemical analyzer was used to determine LDH activity (InTec PRODUCTS, Inc., Xiamen, China) in the BALF following the manufacturer’s protocol.

γδT cells, CD4⁺T cells, and CD8⁺T cells were purified from the lungs of wild-type mice or infected mice using mouse TCRγ/δ⁺T cell isolation kit, mouse CD4 (L3T4) MicroBeads, and CD8a (Ly-2) MicroBeads (Miltenyi Biotechnology, Germany) according to the manufacturer’s instructions, respectively. The cell purity reached to 95% as determined by flow cytometry. Purified γδT cells, CD4⁺T cells, and CD8⁺T cells were stimulated by virus and cultured with IL-1β (10 ng/ml) and/or IL-23 (10 ng/ml) for 36 h. The concentrations of IL-17A in cultured supernatants were determined using an enzyme-linked immunosorbent assay (ELISA) (R&D, MN, USA).

The data are presented as mean ± SEMs. Analysis of variance (ANOVA) was used for analysis of the differences between three or more groups and t tests were used for analysis of the differences between two groups. Statistical graphs were obtained using GraphPad Prism 5 software. The p value of the difference in survival was determined using the Kaplan–Meier log-rank test. Differences were considered statistically significant when p < 0.05 and highly statistically significant when p < 0.01 and p < 0.001.

To further evaluate the roles of γδT cells during the early stage of influenza A (H1N1) pdm09 infection, a mouse model of severe infection was established based on our previous study (40). After infection with the virus, the mice exhibited piloerection, listlessness and loss of appetite. The body weight gradually decreased by 32.48% (Figure 1A), and the mortality was 83.33% at 9 dpi (Figure 1B). Compared to uninfected control mice, in the lungs of infected mice, alveolar space collapse and gradually aggravated histopathological injury were observed and there was a high degree of inflammatory cell infiltration in the lung tissues (Figure 1C). Meanwhile, the concentration of total protein and activity of LDH in the BALF, which represented the pathological damage to the lungs, significantly increased from 1 dpi and peaked on 5 and 3 dpi, respectively (Figure 1D; Figure S1 in Supplementary Material).

The changes in the levels of γδT cells in the lungs of infected or uninfected control mice at multiple time points were analyzed using flow cytometry. As shown in Figure 1E, the percentage of γδT cells among the lung CD3⁺ T lymphocytes in the control mice was 0.59%, but it was significantly raised to 1.04, 1.54, and 1.06% at 1, 3, and 7 dpi, respectively, in the infected mice. Correspondingly, the absolute number of γδT cells also increased significantly from 4.43 × 10⁴ per lung in the control mice to 9.0 × 10⁴ per lung at 3 dpi in the infected mice. Unlike αβT cells, γδT cells leave the thymus as mature cells and the differentiation of functional subtypes of γδT cells is largely preprogrammed in the thymus during the fetal stage (41).

We next investigated whether the accumulated γδT cells in the lungs were recruited from the spleen or peripheral blood. As shown in Figure S2A in Supplementary Material, the percentage of γδT lymphocytes in the peripheral blood was slightly raised at 1 dpi and then reduced to half of the normal level at 3 and 7 dpi. The percentage of γδT lymphocytes in the spleen gradually decreased from 0.62% in the control mice to 0.47, 0.41, and 0.27% at 1, 3, and 7 dpi in the infected mice, respectively. The absolute number of γδT cells also reduced by 44.4 and 88.5% at 3 and 7 dpi, respectively (Figure S2B in Supplementary Material). These data suggest that the spleen might serve as the original source of γδT cells upon pulmonary infection. In addition to γδT lymphocytes, the CD3⁺αβTCR⁺T lymphocytes, and the CD3⁺αβTCR⁺CD4⁺T and CD3⁺αβTCR⁺CD8⁺T subsets, were analyzed. As shown in Figure 1F, there were no clear differences in the percentage and number of CD4⁺T and CD8⁺T cells among the CD3⁺αβTCR⁺T lymphocytes in the infected lungs compared with the control lungs.

Individual γδT-cell subsets are often associated with tissue-specific homing and functions (34, 42). The Vγ1⁺γδT and Vγ4⁺γδT subsets in the lungs and spleens after infection were further investigated. As shown in Figure 1G, the percentage of Vγ4⁺γδT cells gradually increased after infection, and the proportion of Vγ4⁺γδT cells among the γδT cells in the lungs significantly increased from 28.97% in the control mice to 41.48% in the infected mice at 3 dpi. However, the percentage of Vγ1⁺γδT cells slightly declined, with the proportion of Vγ1⁺γδT cells among the γδT cells in the lungs slightly declining from 29.15% in the control mice to 23.87% in the infected mice at 3 dpi. However, the percentage of Vγ4⁺γδT cells in the spleens decreased slightly after infection and the percentage of Vγ1⁺γδT cells showed no obvious change (Figure S2C in Supplementary Material). These results indicate that γδT lymphocytes, especially the Vγ4⁺γδT subset, quickly infiltrated into the lungs at the early stage of the severe influenza A infection, indicating that they have immune protective or immunopathological injury roles.

Interleukin-17 is a solid link between the innate and adaptive immune responses and can exert both beneficial and deleterious effects. Our previous study reported on the pivotal roles of the proinflammatory cytokine IL-17A in immunopathological lung damage during severe influenza virus infection (12). The concentrations of IL-17A in the BALF and serum of the control and infected mice were detected. As shown in Figure 2A, the concentrations of IL-17A in the BALF significantly increased in the infected mice by 3.26- and 3.61-fold at 3 and 7 dpi, respectively. Meanwhile, the concentration of IL-17A in the serum was also significantly increased (1.84-fold) in the infected mice at 3 dpi (Figure 2B).

The proportion and number of IL-17A-secreting cells, including IL-17-producing γδT cells (γδT17), Th17 cells, and Tc17 cells, in the lungs were assayed using flow cytometry. As shown in Figures 2C–E, γδT cells, Th cells, and Tc cells could secret IL-17A in the uninfected and infected lungs, but the vast majority of the IL-17A⁺ cells were γδT cells. The percentage of γδT17 cells among the γδT cells was significantly increased from 8.97% in the uninfected mice to 21.73% in the infected mice at 1 dpi and returned to the normal level at 3 and 7 dpi. The percentages of Th17 and Tc17 cells were also obviously increased at 1, 3, and 7 dpi. However, the proportion of Th17 and Tc17 among the CD4⁺ or CD8⁺ cells was significantly lower than the proportion of γδT17 cells among the γδT cells (Figures 2C,D).

Consistent with this, the absolute number of γδT17 cells was also significantly increased from 2.74 × 10³ in the uninfected lungs to 6.15 × 10³ in the infected lungs at 1 dpi. The number of γδT17 cells (6.15 × 10³ cells/lung) was also far greater than the number of Th17 cells (1.35 × 10³ cells/lung) and Tc17 cells (0.75 × 10³ cells/lung) at 1 dpi and, then, gradually decreased at 3 and 7 dpi. However, the number of Th17 and Tc17 gradually increased as the infection progressed and reached a similar level to the number of γδT17 cells at 7 dpi (Figure 2E).

To accurately support our conclusion that γδT cells were the principal producers of IL-17A, IL-17A⁺ cells among the lung lymphocytes were first gated and, then, surface markers for T-cell subtypes were analyzed. As shown in Figures 2F,G, at 1 dpi, about 56, 36, and 2% of the IL-17A⁺ cells were γδT, CD4⁺Th, and CD8⁺Tc cells, respectively. These results further proved that γδT cells were the major source of IL-17A in the lungs at the early stage of severe infection with influenza A (H1N1) pdm09, which indicated that γδT cells might be involved in lung immunopathological injury.

IL-17A⁺γδT cells were significantly increased after infection, but the major subsets of IL-17A-secreting γδT cells in severe influenza virus infection were unclear. As shown in Figures 3A–C, the expression of IL-17A in Vγ1⁺γδT and Vγ4⁺γδT cells was detected using flow cytometry. The percentage of IL-17A-positive cells among the Vγ1⁺γδT and Vγ4⁺γδT subsets was significantly increased at 1 dpi and then attenuated at 3 and 7 dpi. Moreover, the proportion of IL-17A-positive cells among the Vγ4⁺γδT cells was far higher than that among the Vγ1⁺γδT cells. The proportion of IL-17A-positive cells among the Vγ1⁺γδT cells was 1.8, 4.4, 0.59, and 0.35%, but the proportion among the Vγ4⁺γδT cells was 20.2, 46.6, 12.4, and 5.9% in the uninfected control, 1, 3, and 7 dpi lungs, respectively (Figure 3B). The absolute number of IL-17A⁺Vγ1⁺γδT and IL-17A⁺Vγ4⁺γδT cells was significantly increased at 1 dpi and then declined at 3 and 7 dpi. Furthermore, the absolute number of IL-17A⁺Vγ4⁺γδT cells was far greater than the absolute number of IL-17A⁺Vγ1⁺γδT cells. The number of IL-17A⁺Vγ4⁺γδT cells (1.07 × 10⁴) was tenfold that of IL-17A⁺Vγ1⁺γδT cells (0.99 × 10³) at 1 dpi (Figure 3C). In contrast, in the spleen, only a slight increase in IL-17A-positive cells was observed among the γδT cells and the Vγ4⁺γδT subset, but not the Vγ1⁺γδT subset, at 1 dpi, and the percentage of IL-17A⁺γδT lymphocytes in the spleen was then gradually reduced to half the normal level at 3 and 7 dpi. This indicates the recruitment of γδT cells from the spleen to the lung after infection (Figure S3 in Supplementary Material; IL-17A⁺Vγ1⁺γδT subset flow chart and histogram not shown). These results demonstrated that Vγ4⁺γδT cells were the major source of IL-17A in the lungs at the early stage of severe influenza A (H1N1) pdm09 infection.

It is known that γδT cells can induce very fast, innate-like responses through TCR-dependent or -independent signaling, involving proinflammatory cytokines, the NKG2D receptor, or pattern-recognition receptors. It has been reported that the production of IL-17A by γδT cells appeared to be largely independent of TCR activation and IL-1β and IL-23 were sufficient to trigger abundant secretion of IL-17A by γδT cells (14). As shown in Figures 4A,B, the concentration of IL-1β and IL-23 in the BALF was significantly improved at the early stage of infection, which might promote the secretion of IL-17A by γδT cells.

We next explored whether the secretion of IL-17A by γδT cells was induced by the virus or cytokines. γδT cells, CD4⁺T cells, and CD8⁺T cells were isolated from lung using magnetic-activated cell sorting (MACS) and, then, stimulated with influenza A (H1N1) pdm09 virus, and IL-1β and/or IL-23. As shown in Figure 4C, γδT cells could only secret a small quantity of IL-17A when cultured with medium and the virus could not induce the expression IL-17A directly. However, IL-23 could induce a significant increase in IL-17A, IL-1β could induce a slight increase, and IL-23 and IL-1β had no synergistic effect. Thus, interestingly, the virus was not direct cause for inducing IL-17A production by γδT cells. Moreover, the virus, IL-23, or IL-1β could not induce the expression IL-17A by CD4⁺T cells and CD8⁺T cells (Figures S4A,B in Supplementary Material). To further confirm the dominantly producing of IL-17A by γδT cells, γδT cells, CD4⁺T cells, and CD8⁺T cells from infected lung at 1 dpi were isolated using MACS and, then, cultured to detect IL-17A in supernatants. As shown in Figure 4D, the concentration of IL-17A from the supernatants of cultured γδT cells was significantly higher than that of CD4⁺T cells or CD8⁺T cells. Moreover, the ability to secrete IL-17A of γδT cells was not further amplified by IL-23 or IL-1β (Figures S4C–E in Supplementary Material). These findings suggested that influenza A (H1N1) pdm09 virus might infect epithelial cells, macrophages, or dendritic cells and cause them to secret IL-1β and IL-23, which then promote IL-17A production by γδT cells in an antigen-independent manner.

Interferon-γ is an important early-phase antiviral effector cytokine secreted by activated γδT cells, which can act together with other immune effectors (particularly NK cells) for rapid and efficient virus control (25–27, 43). In the peripheral tissues, the functional orientation of γδT cells depends on the microorganism encountered, with IFN-γ production dominating antiviral responses (44). On the other hand, stimulation with IL-1β and IL-23 can also induce the co-production of IFN-γ by γδT17 cells (45). Flow cytometry was used to further analyze the expression of IFN-γ among γδT cells and their dominant subsets. As shown in Figure 5, the percentage of IFN-γ-positive cells among the γδT cells significantly increased from 0.14% in the uninfected lungs to 2.42, 1.31, and 1.25% in the infected lungs at 1, 3, and 7 dpi, respectively. The percentages of IFN-γ-positive cells among the Vγ1⁺γδT and Vγ4⁺γδT subsets all increased significantly after infection, with similar rates of increase; this was different from the IL-17A production, which was mainly derived from Vγ4⁺γδT cells. However, IFN-γ⁺γδT cells, as well as the IFN-γ⁺Vγ1⁺γδT and IFN-γ⁺Vγ4⁺γδT subsets, in the infected spleens gradually decreased compared with the levels in the uninfected mice (Figure S5 in Supplementary Material).

It has been reported that human Vγ9Vδ2 T cells can be directly activated by NKG2D (46). Activated γδT cells can lyse and eliminate virus-infected cells by enhancing their cytotoxicity effector functions (44). It has also been reported that human Vγ9Vδ2 T cells can recognize and efficiently kill human or avian influenza virus-infected monocyte-derived macrophages. The cytotoxicity of Vγ9Vδ2 T cells is dependent on NKG2D activation and is mediated by the Fas–FasL and perforin–granzyme B pathways (27–30). To determine the phenotypes of the γδT cells and their subsets in the mice lungs, we examined the cell surface markers NKG2D and FasL in γδT cells from both control and infected mice lungs using flow cytometry. The percentage of NKG2D-positive cells among lung γδT cells, and Vγ1⁺γδT and Vγ4⁺γδT subsets, gradually increased, and there were significant differences at 3 and 7 dpi compared to the levels in the uninfected lungs (Figure 6A). Similar results were found in the infected γδT cells, and Vγ1⁺γδT and Vγ4⁺γδT subsets, in the spleens, but the positive rates were far lower than in the lungs (Figure S6A in Supplementary Material).

As the infection process, the percentage of FasL-positive cells among lung γδT cells and Vγ1⁺γδT subsets gradually increased but not among the Vγ4⁺γδT subset (Figure 6B). The percentage of FasL-positive cells among γδT cells was significantly raised from 15.8% in the uninfected lungs to 28.7, 31.6, and 37.6% in the infected lungs at 1, 3, and 7 dpi, respectively (Figure 6B). The FasL-positive cells in the infected γδT cells, Vγ1⁺γδT and Vγ4⁺γδT subsets, in the spleens were only significant increased at 1 dpi, and the positive rates were far lower than that in the lungs (Figure S6B in Supplementary Material).

After we confirmed that γδT cells and the Vγ4⁺γδT subset were the major source of the elevated IL-17A that mediated lung immunopathological injury in severe infection, we further analyzed whether depletion of γδT cells or the Vγ4⁺γδT subset could alleviate lung injury. The neutralizing antibodies against TCRγ/δ, TCR Vγ4, TCR Vγ1 or isotype control were i.p. injected, as shown in Figure 7A. After anti-TCRγ/δ mAb treatment at 1 day before infection, 92.9% of γδT cells in the lungs were depleted on the day of infection (the percentage of γδT cells in the lungs was decreased from 0.57% in the PBS-treated control group to 0.04%), and the depletion lasted until 7 dpi (Figure 7B). For Vγ1⁺γδT and Vγ4⁺γδT subset depletion, anti-TCR Vγ1 or anti-TCR Vγ4 mAbs, respectively, were injected twice on 3 and 1 days before infection. The results showed that almost all the lung Vγ1⁺γδT subset was depleted on the day of infection, and this could also last until 7 dpi (Figure 7C). In addition, 51.88 and 62.42% of the lung Vγ4⁺γδT subset were depleted on the day of infection and 7 dpi, respectively (Figure 7C). Meanwhile, almost all the Vγ1⁺γδT or Vγ4⁺γδT subsets were depleted in the blood and spleen on the day of infection and 7 dpi (Figure S7 in Supplementary Material).

Depletion of γδT cells or the Vγ4⁺γδT subset but not the Vγ1⁺γδT subset significantly reduced the severe influenza A (H1N1) pdm09 infection-induced weight loss (Figure 8A) and also significantly improved the survival rate (Figure 8B). The survival rate was only 16.7% at 14 dpi [with 10² TCID₅₀ influenza A (H1N1) pdm09 virus], but it was significantly improved (to 83.3%) at 14 dpi after γδT cell depletion. Surprisingly, depletion of the Vγ4⁺γδT subset led to a similar survival rate to that after depletion of the γδT cells, but depletion of the Vγ1⁺γδT subset did not change the survival rate (Figure 8B).

Further analysis showed that the improved survival rate was not related to the viral load but to reduced ALI. Compared to the PBS or isotype control group, the viral titers at 3 and 7 dpi were similar in the γδT, Vγ1⁺γδT, and Vγ4⁺γδT cell depletion groups (Figure 8C). Histopathological analysis of the lungs of the PBS or isotype antibody-treated control mice showed gradually increased histopathological injury and a large amount of infiltrating inflammatory cells. Depletion of γδT cells significantly ameliorated the lung immunopathological injury and attenuated the inflammatory cell infiltration (Figure 8D). Consistent with the survival rate findings, depletion of the Vγ4⁺γδT subset but not the Vγ1⁺γδT subset also ameliorated lung injury and attenuated inflammatory cell infiltration (Figure 8D). In addition, the concentration of total protein in the BALF also decreased from 2.61 mg/ml in the control group to 0.71 and 0.89 mg/ml in the γδT and Vγ4⁺γδT cell depletion groups at 3 dpi, respectively (Figure 8E). Moreover, the activity of LDH in the BALF was decreased from 104.56 U/l in the control group to 17.18 and 7.06 U/l in the γδT and Vγ4⁺γδT cell depletion groups, respectively (Figure 8F).

We next detected whether the alleviated lung injury in the γδT or Vγ4⁺γδT cell depletion groups was related to changes in cytokines. As shown in Figures 9A,B, depletion of γδT cells or the Vγ4⁺γδT subset significantly decreased the amount of IL-17A in lung and the concentration of IL-17A in BALF compared to PBS or isotype antibody control group. The amount of IL-17A in lung of γδT cells depletion group and Vγ4⁺γδT subset depletion group were decreased to 27.91 and 31.21% of that in isotype antibody control group, respectively. Meanwhile, the concentrations of IL-17A in BALF of γδT cells depletion group and Vγ4⁺γδT subset depletion group were also reduced to 30.5 and 36%, respectively, when compared to isotype antibody control group. However, depletion of Vγ1⁺γδT subset showed similar level IL-17A in lung and BALF with PBS or isotype antibody control group. Since pulmonary γδT cells could also secret a small amount of IFN-γ, the levels of IFN-γ in lung and BALF of different depletion groups were also detected. As shown in Figures 9C,D, depletion of γδT cells, as well as Vγ1⁺γδT and Vγ4⁺γδT subsets, had no obvious changes when compared with PBS and isotype control groups. As intraperitoneal injection of depletion antibodies depletes systemic γδT lymphocytes, the IL-17A and IFN-γ in serum were also detected. As shown in Figure S8A in Supplementary Material, depletion of γδT cells or the Vγ4⁺γδT subset significantly decreased the serum concentration of IL-17A to 14.3 and 23.5% of the concentrations in the PBS-treated control group at 3 dpi. Puzzlingly, depletion of the Vγ1⁺γδT subset also decreased the concentration of IL-17A, which might be due to systemic depletion. Meanwhile, the concentration of IFN-γ also showed a slightly lower after depletion of γδT, Vγ1⁺γδT, and Vγ4⁺γδT cells at 2 dpi compared to isotype control group, but there was a sharp rise at 7 dpi, which might have been due to the differentiation of Th1 cells (Figure S8B in Supplementary Material).