Regulation of Pulmonary Bacterial Immunity by Follistatin-Like Protein 1

Klebsiella pneumoniae is a common cause of antibiotic-resistant pneumonia. Follistatin-like protein 1 (FSTL-1) is highly expressed in the lung and is critical for lung homeostasis. The role of FSTL-1 in immunity to bacterial pneumonia is unknown. Wild-type (WT) and FSTL-1 hypomorphic (Hypo) mice were infected with Klebsiella pneumoniae to determine infectious burden, immune cell abundance, and cytokine production. FSTL-1 Hypo/TCRδ^−/− and FSTL-1 Hypo/IL17ra^−/− were also generated to assess the role of γδT17 cells in this model.

ABSTRACT

Klebsiella pneumoniae is a common cause of antibiotic-resistant pneumonia. Follistatin-like protein 1 (FSTL-1) is highly expressed in the lung and is critical for lung homeostasis. The role of FSTL-1 in immunity to bacterial pneumonia is unknown. Wild-type (WT) and FSTL-1 hypomorphic (Hypo) mice were infected with Klebsiella pneumoniae to determine infectious burden, immune cell abundance, and cytokine production. FSTL-1 Hypo/TCRδ^−/− and FSTL-1 Hypo/IL17ra^−/− were also generated to assess the role of γδT17 cells in this model. FSTL-1 Hypo mice had reduced K. pneumoniae lung burden compared with that of WT controls. FSTL-1 Hypo mice had increased Il17a/interleukin-17A (IL-17A) and IL-17-dependent cytokine expression. FSTL-1 Hypo lungs also had increased IL-17A⁺ and TCRγδ⁺ cells. FSTL-1 Hypo/TCRδ^−/− displayed a lung burden similar to that of FSTL-1 Hypo and reduced lung burden compared with the TCRδ^−/− controls. However, FSTL-1 Hypo/TCRδ^−/− mice had greater bacterial dissemination than FSTL-1 Hypo mice, suggesting that gamma delta T (γδT) cells are dispensable for FSTL-1 Hypo control of pulmonary infection but are required for dissemination control. Confusing these observations, FSTL-1 Hypo/TCRδ^−/− lungs had an increased percentage of IL-17A-producing cells compared with that of TCRδ^−/− mice. Removal of IL-17A signaling in the FSTL-1 Hypo mouse resulted in an increased lung burden. These findings identify a novel role for FSTL-1 in innate lung immunity to bacterial infection, suggesting that FSTL-1 influences type-17 pulmonary bacterial immunity.

INTRODUCTION

Klebsiella pneumoniae is a Gram-negative bacillus of the family Enterobacteriaceae. While K. pneumoniae commensally inhabits mucosal surfaces of healthy individuals, it is a common cause of pneumonia and bacteremia during nosocomial infection (1). Increasingly, K. pneumoniae infections are resistant to antibiotics, including carbapenems (carbapenem-resistant Enterobacteriaceae [CRE]) and extended-spectrum β-lactams (ESBLs), and thus constitute a major public health threat (2).

Interleukin-17A (IL-17A) and gamma delta T (γδT) cells have been shown to be an integral component of innate host immunity to K. pneumoniae pulmonary infection (3, 4). While numerous IL-17A-associated inflammatory mediators (IL-23, G-CSF, CXCL1, CXCL2, CXCL5, and Lipocalin-2) have been well characterized in the clearance of K. pneumoniae from the lung, γδT cells have been identified as an early, and critical, IL-17A-producing cell type (5–10). Moreover, γδT cells are necessary for the prevention of bacterial systemic dissemination (3).

Follistatin-like protein 1 (FSTL-1) is a secreted glycoprotein with multiple functions in diverse physiological processes, including cardiovascular, respiratory, skeletal, muscular, nervous, and urinary development; signaling; homeostasis; and tissue remodeling; as well as mediating respiratory disease, cardiovascular disease, and cancer (11–23). FSTL-1 has also been shown to have important functions in immune-mediated diseases and immune signaling (24–30). Initially described as a novel mediator of inflammation in arthritis, FSTL-1 has been shown to enhance proinflammatory cytokine and chemokine production in a collagen-induced arthritis murine model (31–34), as well as promote metalloprotease gene transcription in human rheumatoid synoviocytes (35). We have previously shown that FSTL-1 is critical in lung homeostasis, as a decrease in FSTL-1 results in a spontaneous emphysema phenotype and is associated with altered Nur77/Nr4a1-dependent macrophage inflammatory signaling (25). However, the role Nr4a1 in bacterial lung infection is poorly understood, and the impact of FSTL-1 on pulmonary immunity to infection has not been previously identified.

The present study examines the role of FSTL-1 in lung immunity in a model of K. pneumoniae pulmonary infection. The FSTL-1 hypomorphic (Hypo) mouse, a mouse strain with reduced levels of FSTL-1, demonstrated less bacterial lung burden than the wild type, suggesting that the FSTL-1 Hypo lung environment provides host protection against K. pneumoniae. These mice have an increased population of resident IL-17A-producing γδT cells compared with the wild type; however, crossing FSTL-1 Hypo with the TCRδ^−/− mouse and generating the FSTL-1 Hypo/TCRδ^−/− double knockout did not reverse the phenotype, likely due to compensatory IL-17A⁺ T cell populations. Attenuation or elimination of IL-17A signaling increased pulmonary burden in the FSTL-1 Hypo mouse. Cumulatively, these observations suggest that FSTL-1 mediates control of bacterial lung infection, in part, by influencing the development and function of IL-17A-producing cells.

RESULTS

FSTL-1 Hypo mice have reduced Klebsiella pneumoniae burden in the lung.

To assess the innate immune impact of FSTL-1 in K. pneumoniae pulmonary infection, intratracheal (i.t.) K. pneumoniae-infected WT and FSTL-1 Hypo mice had lung burden determined by CFU at 24 and 48 hours postinfection. A difference in bacterial burden was not found at 24 hours postinfection (Fig. 1A). At 48 hours postinfection, FSTL-1 Hypo mice displayed a significant reduction in K. pneumoniae lung burden by CFU, compared with wild-type mice (Fig. 1B). To assess bacterial dissemination, spleens were harvested as well. At 24 and 48 hours postinfection, FSTL-1 Hypo mice displayed no significant difference in bacterial dissemination compared with wild-type mice by CFU (Fig. 1A and B). Surprisingly, despite the underlying spontaneous emphysema physiology (25), FSTL-1 Hypo mice were able to control K. pneumoniae from the lung, by the 48-hour time point, to a greater extent than wild-type mice, thereby suggesting that germ line Fstl1 mutations resulting in reduced FSTL-1 expression conveyed a protective advantage against K. pneumoniae pulmonary infection.

FIG 1.

FSTL-1 Hypo mice have reduced Klebsiella pneumoniae burden in the lung. (A) Klebsiella pneumoniae infectious burden at 24 hours postinfection in the lung and spleen in WT and FSTL-1 Hypo mice. (B) Klebsiella pneumoniae infectious burden in the lung and spleen at 48 hours postinfection in WT and FSTL-1 Hypo mice. Data are representative of three combined experiments. **, P = 0.0068.

FSTL-1 Hypo mice have increased IL-17A production following Klebsiella pneumoniae pulmonary infection.

We next aimed to identify potential mechanisms underlying the reduced K. pneumoniae burden of the FSTL-1 Hypo lung, where we confirmed reduced Fstl1 transcript levels and FSTL-1 protein abundance (Fig. 2A). At the 24-h time point, when burden was similar, we next identified that FSTL-1 Hypo mice had significantly increased Il17a gene expression and IL-17A protein (Fig. 2B), an integral host defense cytokine against Klebsiella lung infection. IL-17A-associated gene (Csf3, Cxcl1, Cxcl2, Cxcl5, Il23a, Il23r, and Lcn2) transcript levels were significantly elevated in the FSTL-1 Hypo mouse lung at 24 h postinfection (Fig. 2C), and there was an increased trend in the Il22 transcript. Absolute numbers of CD3⁺TCRγδ⁺ and CD11b⁺Ly6G⁺ neutrophils present in the lung were also significantly higher in the FSTL-1 Hypo mouse than those in the wild type (Fig. 2D). These data suggest that a reduced burden in the lung may be mediated by an increase in IL-17A-mediated recruitment of neutrophils.

FIG 2.

FSTL-1 Hypo mice have increased IL-17A production, and IL-17-associated pathways, following Klebsiella pneumoniae infection. WT and FSTL-1 Hypo mouse lung 24 hours postinfection. (A) Fstl1/FSTL-1 is reduced in FSTL-1 Hypo mice, as determined by gene expression and protein. (B) Il17a/IL-17A is elevated in FSTL-1 Hypo mice, as determined by gene expression and protein. (C) IL-17A-associated gene expression is elevated in FSTL-1 Hypo mice. All mRNA transcript levels were determined relative to reference gene Hprt1. (D) CD11b⁺Ly6G⁺ and CD3⁺ TCRγδ⁺ cell numbers are increased in FSTL-1 Hypo mice. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

γδT cells are the primary source of IL-17A during K. pneumoniae lung infection.

We next sought to identify the cellular source of this increased IL-17A production. Using flow cytometry, we evaluated the abundance of cell types known to produce IL-17A (CD3⁻, CD3⁺CD4⁺, CD3⁺CD8⁺, and CD3⁺TCRγδ⁺) in the lung. There were no significant differences in the percentage of CD3⁺CD4⁺ or CD3⁺CD8⁺ cells 24 hours postinfection (see Fig. S1B and C in the supplemental material); however, CD3⁺ TCRγδ⁺ percentages were increased in the FSTL-1 Hypo lung compared to WT (Fig. 3B). Next, we assessed to what extent these cell types produced IL-17A during K. pneumoniae infection. Compared to the CD3⁺ cells, CD3⁻ cells produced negligible percentages of IL-17A regardless of genotype (Fig. 3A). FSTL-1 Hypo lung CD3⁺TCRγδ⁺ cells produced significantly higher percentages of IL-17A than those of wild-type lungs (Fig. 3C). Unlike CD3⁺TCRγδ⁺ cells, CD3⁺CD4⁺ and, to a lesser extent CD3⁺CD8⁺, cells produced similar amounts of IL-17A when comparing WT to FSTL-1 Hypo mice (Fig. 3C and D; see Fig. S2D in the supplemental material), suggesting that FSTL-1 Hypo mice have increased IL-17A-producing γδT cells during K. pneumoniae lung infection compared with those of wild-type mice, potentially conferring protection against pulmonary infection.

FIG 3.

FSTL-1 Hypo mice have increased γδ⁺IL-17A⁺ cells following Klebsiella pneumoniae infection. WT and FSTL-1 Hypo mouse lung 24 hours postinfection. CD3⁺IL-17A⁺ (A), CD3⁺TCRγδ⁺ (B), and TCRγδ⁺IL-17A⁺ (C) cells are elevated in FSTL-1 Hypo mice by flow cytometric analysis. (D) CD4⁺IL-17A⁺ cells are not significantly different between FSTL-1 Hypo and WT mice by flow cytometric analysis. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

FSTL-1 Hypo mice have greater lung-resident TCRγδ⁺IL-17A⁺ cell abundance.

To determine whether the increase in lung IL-17A-producing γδT cells in FSTL-1 Hypo mice was elicited by K. pneumoniae infection or was a preexisting developmental immune phenotype, we examined lung-resident cell percentages and IL-17A production by flow cytometry. The uninfected lungs of a FSTL-1 Hypo mouse displayed a similar dominant CD3⁺TCRγδ⁺IL-17A⁺ lung phenotype compared with that of wild-type lungs (Fig. 4A to C). FSTL-1 Hypo lungs display a significantly greater percentage of CD3⁺IL-17A⁺ cells and CD3⁺TCRγδ⁺ cells than wild-type lungs, with no significant contribution of IL-17A from CD3⁻, CD3⁺CD4⁺, or CD3⁺CD8⁺ cells (Fig. 4; see Fig. S3 in the supplemental material). These data suggest reduced FSTL-1 production in the FSTL-1 Hypo mouse may lead to the development of a lung environment with a greater type-17 immune potential, as manifested by increased lung-resident IL-17A-producing γδT cells.

FIG 4.

FSTL-1 Hypo mice have greater lung-resident γδ⁺IL-17A⁺ cell abundance. CD3⁺IL-17A⁺ (A), CD3⁺TCRγδ⁺ (B), and TCRγδ⁺IL-17A+ (C) cells are elevated in uninfected FSTL-1 Hypo lungs by flow cytometric analysis. (D) CD4⁺IL-17A⁺ cells are not significantly different between FSTL-1 Hypo and WT uninfected lungs by flow cytometric analysis. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

FSTL-1 Hypo/TCRδ^−/− mice have reduced Klebsiella pneumoniae lung burden but are not protected against bacterial dissemination.

Given the increase in lung-resident IL-17A-producing γδT cells of a FSTL-1 Hypo mouse, we next sought to examine the necessity of γδT cells in conferring FSTL-1 Hypo protection against K. pneumoniae by generation of a FSTL-1 Hypo/TCRδ^−/− double-knockout mouse (Fig. 5A and B). We hypothesized that if FSTL-1 Hypo mice were protected via lung-resident IL-17A-producing γδT cells, the absence of γδT cells in the FSTL-1 Hypo/TCRδ^−/− mouse would phenocopy the TCRδ^−/−, as determined by K. pneumoniae infectious burden. Thus, lung and spleen K. pneumoniae burdens were determined 48 hours postinfection by CFU in FSTL-1 Hypo, TCRδ^−/−, and FSTL-1 Hypo/TCRδ^−/−. FSTL-1 Hypo mice had a significantly lower lung burden than that of TCRδ^−/− mice, with a similar burden compared with FSTL-1 Hypo/TCRδ^−/− mice (Fig. 5C). These findings suggest that FSTL-1 Hypo-mediated lung protection is independent of γδT cells.

FIG 5.

FSTL-1 Hypo-TCRδ^−/− mice have reduced Klebsiella pneumoniae lung burden, but are not protected against bacterial dissemination. (A) TCRδ⁺ cells are absent in naive lung of Tcrδ^−/− and FSTL-1 Hypo/Tcrδ^−/− mice, as determined with a fluorescence-activated cell sorter (FACS). (B) Fstl1 and Trdc gene expression in naive lung of FSTL-1 Hypo, Tcrδ^−/−, and FSTL-1 Hypo/Tcrδ^−/− mice. All mRNA transcript levels were determined relative to reference gene Hprt1. (C) Klebsiella pneumoniae infectious burden in lung and spleen 48 hours postinfection in FSTL-1 Hypo, Tcrδ^−/−, and FSTL-1 Hypo/Tcrδ^−/− mice. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

When we examined bacterial dissemination, TCRδ^−/− demonstrated an increase in splenic bacterial burden relative to FSTL-1 Hypo mice (Fig. 5C). However, FSTL-1 Hypo/TCRδ^−/− mice also showed increased K. pneumoniae burden in spleen CFU compared with that of FSTL-1 Hypo spleen (Fig. 5C). This finding supports the necessity of γδT cells in protection from bacterial dissemination in the context of FSTL-1 Hypo, which is consistent with prior observations that γδT cells are required for the control of bacterial dissemination during pulmonary K. pneumoniae infection (3).

FSTL-1 Hypo/TCRδ^−/− mice have exaggerated, compensatory lung-resident IL-17A⁺-producing cells.

To understand the impact of IL-17 in the FSTL-1 Hypo/TCRδ^−/− mouse, in the setting of K. pneumoniae lung infection, we sought to determine the immune phenotype. Knowing that FSTL-1 Hypo mice have an increase in lung-resident IL-17A-producing γδT cells, we next sought to phenotype the type-17 immune response in the FSTL-1 Hypo/TCRδ^−/− mouse. Uninfected FSTL-1 Hypo/TCRδ^−/− mice exhibited a significant increase in the percentage of lung-resident CD3⁺IL-17A⁺ cells compared with that of both FSTL-1 Hypo and TCRδ^−/− mice, with little to no contribution of IL-17A from the CD3⁻ population (Fig. 6A). In the FSTL-1 Hypo lung, the dominant IL-17A-producing cells are the TCRγδ⁺ cells (Fig. 6B to D). However, TCRδ^−/− and FSTL-1 Hypo/TCRδ^−/− lung-resident IL-17A-producing cells are significantly dominated by CD3⁺CD4⁺ and CD3⁺CD4⁻CD8⁻ cell populations (Fig. 6D). CD8⁺ cells were not a significant contributor of IL-17A (see Fig. S4D in the supplemental material). These data suggest that, within the milieu of a FSTL-1 Hypo mouse, elimination of γδT cells promotes a compensatory development of IL-17A production via CD3⁺CD4⁺ cells and CD3⁺CD4⁻CD8⁻ cells.

FIG 6.

FSTL-1 Hypo-TCRδ^−/− mice have exaggerated, compensatory lung-resident IL-17A⁺-producing cells. Percentage of CD3⁺IL-17A⁺ (A), TCRγδ⁺IL-17A⁺ (B), and CD4⁺IL-17A⁺ (C) cells in uninfected FSTL-1 Hypo, Tcrδ^−/−, and FSTL-1 Hypo/Tcrδ^−/− lungs by flow cytometric analysis. (D) Percentage of CD4⁺TCRγδ⁺ cells of CD3⁺IL-17A⁺ population by flow cytometry in FSTL-1 Hypo, Tcrδ^−/−, and FSTL-1 Hypo/Tcrδ^−/− uninfected lungs. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Attenuation of IL-17A signaling in FSTL-1 Hypo mice increased Klebsiella pneumoniae lung burden.

We next aimed to determine the importance of increased IL-17A production in FSTL-1 Hypo for protection against K. pneumoniae pulmonary infection using two independent techniques. First, we employed the anti-IL-17A neutralization antibody in the FSTL-1 Hypo mouse and assessed bacterial lung and spleen burden by CFU at 48 hours postinfection. Compared with the IgG control antibody group, the anti-IL-17A-treated group had significantly increased lung CFU and no difference in spleen CFU (Fig. 7A and B). Next, a FSTL-1 Hypo/Il17ra^−/− double-transgenic mouse was assessed for infectious burden 48 hours following K. pneumoniae pulmonary infection. Compared with the FSTL-1 Hypo mouse, the FSTL-1 Hypo/Il17ra^−/− mouse had significantly increased lung burden and yet no difference in bacterial dissemination (Fig. 7C and D), consistent with the impact of IL-17A neutralization. Together, these findings suggest that the FSTL-1 Hypo mouse lung environment protects against pulmonary K. pneumoniae infection in an IL-17A-dependent manner.

FIG 7.

Attenuation of IL-17A signaling in FSTL-1 Hypo mice increased Klebsiella pneumoniae lung burden. Klebsiella pneumoniae infectious burden at 48 h in the lung (A) and spleen (B) of FSTL-1 Hypo mice treated with anti-IL-17A neutralizing antibody or IgG control. Klebsiella pneumoniae infectious burden in the lung (C) and spleen (D) at 48 hours postinfection in FSTL-1 Hypo and FSTL-1 Hypo/Il17ra^−/− mice. **, P < 0.01. Data for A and B are representative of 3 individual experiments; data from C and D are combined from two individual experiments.

DISCUSSION

Antibiotic-resistant infections are a worldwide public health crisis, specifically carbapenem-resistant Enterobacteriaceae (CRE), including Klebsiella pneumoniae (2). K. pneumoniae is also a common cause of hospital-acquired infection and pneumonia. Development of new approaches to treat and prevent infection and colonization with antibiotic-resistant organisms, such as K. pneumoniae, is increasingly urgently needed. Modulating host immunity to aid in the clearance of infection is an understudied approach, and a deeper understanding of host-pathogen interactions that control K. pneumoniae lung infection is required. Several aspects of this pulmonary host defense pathway have been characterized (36); T cell immune responses are essential for orchestrating effective bacterial clearance, including IL-17-mediated signaling (5, 10, 37–39). Moore et al. established that interferon gamma (IFN-γ) is required for effective K. pneumoniae responses in the lung (40). Alpha beta T (αβT) cell responses are dispensable for controlling acute infection, although Th17 cells confer antibody-independent protection against subsequent heterologous infections, including K. pneumoniae (37). During acute infection, γδT cell responses are required to control infection, including K. pneumoniae dissemination, cumulatively suggesting that innate type-17 immunity is critical for controlling K. pneumoniae lung infection and dissemination (41).

Having observed that the FSTL-1 Hypo mouse controls K. pneumoniae lung burden, we then determined cellular and molecular constituents of this phenotype, specifically FSTL-1 regulation of γδ-17 T cells. FSTL-1 Hypo mice expressed elevated proinflammatory cytokines, including IL-17A, following K. pneumoniae infection, which was associated with improved lung burden, consistent with the protective role of type-17 immune responses against bacterial lung infection, although the impact of FSTL-1 on IL-17A-producing cells in the lung has not previously been known.

Determining that IL-17A production in the FSTL-1 Hypo lung is dominated by γδT cells suggested a novel function of FSTL-1, namely, controlling innate type-17 cell immunity in the lung. Interestingly, the observation that FSTL-1 Hypo mice have increased γδT-17 cell abundance prior to infection argues that FSTL-1 likely regulates natural, rather than inducible, γδT-17 cell development. Furthermore, the mechanism(s) by which FSTL-1 confers this phenotype (i.e., regulation of γδT cell thymic differentiation, proliferation, apoptosis, or cell extrinsic factors) remain to be defined. However, the findings that FSTL-1 Hypo mice have enhanced control of K. pneumoniae infection are the first identification of a potential therapeutic role for FSTL-1 in pulmonary host defense.

The observations of FSTL-1 Hypo/TCRδ^−/− mice highlight several insights into the cellular requirements and tissue-specific nature of the FSTL-1 Hypo phenotype. Because the FSTL-1 Hypo/TCRδ^−/− and FSTL-1 Hypo mice have similar K. pneumoniae burdens in the lung, this suggests that γδT cells are not required for the lung-specific effects of FSTL-1. Conversely, FSTL-1 Hypo/TCRδ^−/− have a bacterial dissemination similar to that of TCRδ^−/− mice, suggesting that FSTL-1-mediated control of dissemination requires γδT cells, or a subset thereof. Of note, Moore et al. have shown that TCRδ^−/− mice exhibit compensatory host responses to K. pneumoniae pulmonary infection, as we also observed; they also noted an unaltered ability of the TCRδ^−/− mice to recruit neutrophils, which we also observed in the TCRδ^−/− and FSTL-1 Hypo/TCRδ^−/− mice (data not shown), as well as, to produce compensatory IL-17A levels. Nonetheless, the pulmonary immune response of FSTL-1 Hypo suggests a role for FSTL-1 in the development of resident IL-17A-producing cells, including γδT-17 cells. Furthermore, by showing that IL-17A is critical for FSTL-1 Hypo protection, this work implicated FSTL-1 in regulating the development, activation, and function of IL-17-producing cells, although the mechanism remains unknown. Further examination of the link between FSTL-1 and IL-17A production in the lung (and other tissues) is needed to determine the cytokines and cells involved in FSTL-1 dependent type-17 immunity.

Within the lung, FSTL-1 has critical developmental, homeostatic, and repair functions (12, 13, 21, 25). Although we recently identified that the FSTL-1 Hypo mouse develops spontaneous pulmonary emphysema, a contribution of type-17 immune responses was not identified in that model. However, FSTL-1 Hypo mice did have reduced Nr4a1 expression, and in vitro exogenous FSTL-1 maintained Nr4a1 suppression of NF-κB in macrophages. Other work has shown Nr4a1/Nur77 inhibit Th17 immune responses, although the role of Nr4a1 in pulmonary host defense is poorly understood (42, 43). The experimental approaches to study FSTL-1 in vivo are complicated by the perinatal lethality of germ line Fstl1 knockout (12). Further study of the interplay of FSTL-1, Nr4a1, and type-17 immunity in the lung is warranted.

We have shown a novel role for FSTL-1 as an impactful component of host immunity during Klebsiella pneumoniae pulmonary infection, by impacting lung-resident immune cell IL-17 production. The present findings reinforce the necessity of innate γδT cell and IL-17 responses in effective pulmonary immunity to Klebsiella infection and highlight a potential strategy of immunomodulation for enhancing bacterial clearance in the mammalian lung.

MATERIALS AND METHODS

Animal experiments.

Wild-type C57BL/6 (WT; Jackson Laboratory, Bar Harbor, ME), FSTL-1 Hypo, TCRδ^−/− (B6-TCRdelta⁻, B6.129P2-Tcrd^tm1Mom/J; Jackson Laboratory, Bar Harbor, ME), IL17ra^−/− (Il17ra^tm1Koll, Il17ra KO; Jackson Laboratory, Bar Harbor, ME), FSTL-1 Hypo/Tcrδ^−/−, and FSTL-1 Hypo/IL17ra^−/− mice were bred, housed, and subjected to experimental protocols in accordance with the Institutional Animal Care and Use Committee (IACUC) of the University of Pittsburgh School of Medicine. FSTL-1 Hypo/Tcrδ^−/− mice were generated in our institution by crossing FSTL-1 Hypo mice with Tcrδ^−/− mice. FSTL-1 Hypo/IL17ra^−/− were generated by crossing FSTL-1 Hypo mice with IL17ra^−/− mice. Novel double-transgenic strains were verified by PCR using allele-specific primers and resolution with agarose gel electrophoresis as described (10, 44) (Fig. S1A and B). In vivo studies were performed with 8- to 15-week-old female and male mice. For in vivo neutralization of IL-17A, anti-IL-17A (17F3, BioXcell) or an isotype control (MOPC-21, BioXcell) antibody was administered at 200 μg of intratracheal (i.t.) antibody per mouse both and 400 μg of intraperontaneal (i.p.) antibody on the day prior to infection (day −1), the day of infection (day 0) and on day +1. On day 2, mice were administered only 400 μg i.p. antibody per mouse.

Klebsiella pneumoniae infection.

Klebsiella pneumoniae strain 396, a hypervirulent mucoid K1 serotype clinical isolate (37), was grown overnight at 37°C and 250 rpm in tryptic soy broth (TSB) and then the following day was subcultured in TSB at a dilution of 1:1,000 for 2 hours at 37°C and 250 rpm to achieve log-phase growth. Bacteria were pelleted and washed with sterile phosphate-buffered saline (PBS) at a centrifugation speed of 5,000 × g for 5 minutes. The cell pellet was diluted in sterile PBS and given at 1 × 10⁴ CFU per mouse via intratracheal (i.t.) administration. At 24 or 48 hours following K. pneumoniae administration, mice were sacrificed and lung and spleen were harvested. To assess lung bacterial burden and bacterial dissemination by CFU, the left lung lobe and spleen, respectively, were collected in sterile PBS and subjected to homogenization. The right lung lobe was collected in TRIzol reagent (Invitrogen) for RNA isolation and subsequent gene expression analysis or collected in sterile PBS for collagenase digestion and flow cytometry.

Real-time PCR.

RNA was isolated from lung lobe in TRIzol reagent (Invitrogen) according to the manufacturer’s protocol. RNA was normalized and transcribed into cDNA using iScript reverse transcription supermix (Bio-Rad) according to manufacturer’s protocol. The following primer sequences/assays were used: Hprt (Mm03024075_m1), Fstl1 (Mm00433371_m1), Il-17a (Mm00439618_m1), Csf3 (Mm00438334_m1), Cxcl1 (Mm04207460_m1), Cxcl2 (Mm00436450_m1), Cxcl5 (Mm00436451_g1), Il22 (Mm00444241_m1), Il23a (Mm00444241_m1), Il23r (Mm00444241_m1), Lcn2 (Mm00444241_m1), Gapdh (forward, 5′-AAT GTG TCC GTC GTG GAT CTG A-3′; reverse, 5′-GAT GCC TGC TTC ACC ACC TTC T-3′), and Trdc (forward, 5′-TGTTGCTTGTCTGGTGAAAGAT-3′; reverse 5′-GCACTGTACTTCCCGCTGG-3′). Depending on the primer set, real-time PCRs were run on a Bio-Rad CFX96 real-time system using iQ SYBR green supermix (Bio-Rad) or TaqMan universal PCR mastermix (Applied Biosystems).

Enzyme-linked immunosorbent assay.

Lung homogenates were assayed for mouse FSTL-1 and the IL-17A protein. Lung homogenates were generated by homogenization of left lung lobe in 1 ml sterile PBS. The abundance of mouse FSTL-1 protein in lung homogenates was measured by an inhouse sandwich enzyme-linked immunosorbent assay (ELISA) protocol. A capture antibody (MAB1738; R&D Systems) was adhered to the bottom of a flat-bottomed 96-well plate overnight at 4°C. The plate was washed five times with wash buffer (PBS containing 0.05% Tween 20). The plate was then blocked with blocking buffer (PBS containing 1% bovine serum albumin and 5% sucrose) for 1 hour at room temperature. Standards and experimental samples were incubated for 1 hour at room temperature and then washed as previously stated. A biotinylated detection antibody (BAF1738; R&D Systems) was added for 1 hour at room temperature and then washed. Streptavidin-horseradish peroxidase (HRP) (18-4100-51; eBioscience/ThermoFisher) was added, incubated for 20 minutes at room temperature, and then washed. An OptEIA 3,3',5,5' tetramethylbenzidine (TMB) substrate kit (BD Biosciences) was used to visualize protein concentration. A total of 1 M sulfuric acid was used to stop the substrate reaction, and the plate was read at a wavelength of 450 nm on a Synergy H1 plate reader (BioTek) using Gen5 v2.01 software. The commercially available Mouse IL-17A (homodimer) ELISA Ready-SET-Go! Kit (Invitrogen) was used according to the manufacture protocol’s for assessing IL-17A protein levels.

Flow cytometry.

Single-cell suspensions were obtained by digestion of lung lobe in Dulbecco’s modified Eagle medium (DMEM) containing 4 mg/ml collagenase (Sigma-Aldrich) and 0.2 mg/ml DNase (Sigma-Aldrich) at 37°C with agitation for 1 hour, and then the digest was strained through a 70-μm filter. ACK lysis buffer (Gibco) was used to lyse red blood cells according to the manufacturer’s protocol. Cell labeling was performed using the Foxp3/transcription factor staining buffer set (Invitrogen eBioscience) according to the manufacturer’s protocol. Cells were labeled with the following antibodies: anti-IL-17A (eBiol7B7), anti-CD3 (17A2), anti-TCRγδ (eBioGL3), anti-CD4 (RM4-5), and anti-CD8 (53-6.7). Data acquisition was performed on an LSR Fortessa analyzer (SORP) using FACSDiva software v8.0.1. Data were analyzed using FlowJo software v10.1 (TreeStar).

Absolute cell counts.

Lung was digested and subjected to red blood cell lysis as prepared for flow cytometry. Total lung cells were enumerated with 0.2% trypan blue solution and an Invitrogen Countess automated cell counter. Percentages of lung CD11b⁺Ly6G⁺ neutrophils were obtained by flow cytometry using primary antibodies against mouse CD11b (M1/70) and Ly6G (1A8), LSR Fortessa, and FlowJo software v10.1. The absolute neutrophil number per mouse lung was enumerated using the total lung cell digest count and the flow cytometry percentage of CD11b⁺Ly6G⁺ cells. Lung CD3⁺TCRγδ⁺ cells were enumerated for absolute cell count in the same manner using flow cytometry antibodies, anti-CD3 (17A2), and anti-TCRγδ (eBioGL3).

Statistical analysis.

Data preparation and statistical analysis were performed with GraphPad Prism software v8. For analyses comparing two groups of data, an unpaired t test with assumed Gaussian distribution was utilized. For a concurrent analysis of three or more data sets, a nonmatching or nonpairing ordinary one-way analysis of variance (ANOVA) with assumed Gaussian distribution, where the mean of each data set was compared to the mean of every other data set, was utilized. Statistical significance was assigned to data analyses yielding probability values less than five percent (P < 0.05).