Coxiella burnetii Virulent Phase I and Avirulent Phase II Variants Differentially Manipulate Autophagy Pathway in Neutrophils

ABSTRACT

Coxiella burnetii is an obligate intracellular Gram-negative bacterium that causes Q fever in humans. The virulent C. burnetii Nine Mile phase I (NMI) strain causes disease in animal models, while the avirulent NM phase II (NMII) strain does not. In this study, we found that NMI infection induces severe splenomegaly and bacterial burden in the spleen in BALB/c mice, while NMII infection does not. A significantly higher number of CD11b⁺ Ly6G⁺ neutrophils accumulated in the liver, lung, and spleen of NMI-infected mice than in NMII-infected mice. Thus, neutrophil accumulation correlates with NMI and NMII infection-induced inflammatory responses. In vitro studies also demonstrated that although NMII exhibited a higher infection rate than NMI in mouse bone marrow neutrophils (BMNs), NMI-infected BMNs survived longer than NMII-infected BMNs. These results suggest that the differential interactions of NMI and NMII with neutrophils may be related to their ability to cause disease in animals. To understand the molecular mechanism underlying the differential interactions of NMI and NMII with neutrophils, global transcriptomic gene expressions were compared between NMI- and NMII-infected BMNs by RNA sequencing (RNA-seq) analysis. Interestingly, several genes involved in autophagy-related pathways, particularly membrane trafficking and lipid metabolism, are upregulated in NMII-infected BMNs but downregulated in NMI-infected BMNs. Immunofluorescence and immunoblot analyses indicate that compared to NMI-infected BMNs, vacuoles in NMII-infected-BMNs exhibit increased autophagic flux along with phosphatidylserine translocation in the cell membrane. Similar to neutrophils, NMII activated LC3-mediated autophagy in human macrophages. These findings suggest that the differential manipulation of autophagy of NMI and NMII may relate to their pathogenesis.

INTRODUCTION

Coxiella burnetii is an obligate intracellular Gram-negative bacterium that causes the zoonotic disease Q fever (1). The Nine Mile strain of C. burnetii exists as two phase variants based on the lipopolysaccharide (LPS) structure expressed on the cell surface: the Nine Mile phase I variant (NMI) exhibits full-length LPS, and the phase II (NMII) variant has truncated LPS. A genomic comparison showed that relative to the genome of NMI strain, homologs of genes CBU_0679 to CBU_0697 were completely deleted in the genome of NMII strain, which eliminates several LPS biosynthesis genes and is associated with the production of a severely truncated LPS (2). Several in vivo studies demonstrated that NMI strain causes fever, weight loss, or splenomegaly in mice and guinea pigs, while NMII strain does not (3). However, the mechanism underlying how NMI and NMII strains differentially cause disease in animal models remains unclear.

Neutrophils are abundant cells in the circulation that potentially favor early killing of intracellular bacteria without involving significant acquired immunity (4, 5). Neutrophils have a short half-life and are inherently preprogrammed to die by constitutive apoptosis without causing damage to tissues. However, interaction with bacteria results in neutrophil turnover. Neutrophil turnover is microbe specific, where the pathogen decides the fate of neutrophils, ranging from prolonging the neutrophil life span to causing rapid neutrophil lysis after phagocytosis. During nonpathogenic bacterial invasion, neutrophils kill the bacteria through phagocytosis and degranulation of lytic enzymes, reactive oxygen intermediates, and cytokines (6). However, during infectious conditions, uncontrolled recruitment of neutrophils eventually increases the severity of infection, leading to subsequent inflammatory responses (7). Neutrophils are not only associated with various acute responses during infections and inflammations but also appear to be involved in chronic, indolent human inflammatory diseases (8). Thus, neutrophils exhibit a multifaceted role during intracellular bacterial infections, and the heterogeneity of neutrophils is usually microbe specific. The heterogeneity in response includes differences in the production of effector molecules, resulting in prolonged survival or immediate cell lysis (9).

Our previous studies demonstrated that neutrophils play an important role in host defense against C. burnetii infection (10, 11). Notably, the phagocytosis of NMI and NMII by neutrophils differs, in that NMII is more susceptible (12), which might relate to the nonpathogenic nature of NMII. Thus, the difference in the effector responses of neutrophils against NMI and NMII may be responsible for their differential ability to induce inflammatory responses in animals. Therefore, understanding the molecular mechanism underlying the neutrophil heterogeneity against C. burnetii virulent NMI and avirulent NMII variants would open a new avenue of neutrophil immunity in response to intracellular bacterial pathogens.

Developments in modern genomics and proteomics technologies provide a great opportunity to utilize RNA-sequencing (RNA-seq) technology to understand the changes in transcriptional regulation and the protein synthesis associated with neutrophil function during infection and inflammation (13, 14). Recent reports specified that hypervariable expression of immunologically relevant genes is associated with environmental exposure, which implicates the genetic and epigenetic manipulations in neutrophils by pathogens and other external influences (15). However, no study has been performed utilizing RNA-seq to compare the neutrophil responses against virulent and avirulent strains of intracellular bacterial pathogens. By use of this technology, a deep understanding of the global expression variations between NMI- and NMII-infected neutrophils could provide novel insights into the molecular pathways and associated effector functions of neutrophils. In addition, information about NMI-specific activation of genes and pathways may lead to the development of novel key strategies to prevent neutrophil-mediated inflammatory responses during C. burnetii infection.

In this study, we used a systems biology approach to examine whether the differential interaction of virulent NMI and avirulent NMII with neutrophils relates to their ability to cause disease in a mouse model. The results suggest that NMI’s and NMII’s differential manipulation of autophagy in neutrophils may be responsible for their different abilities to cause disease in mice. This study provides novel insights into the neutrophil-mediated immunity against an intracellular bacterial pathogen.

RESULTS

Neutrophils are more highly accumulated in NMI-infected mouse organs than in NMII-infected mouse organs.

Earlier investigations demonstrated the involvement of neutrophils in invoking the innate immune response during intracellular bacterial infection and the inflammation response. To understand the role of neutrophils during NMI and NMII infections, BALB/c mice were infected with 1 × 10⁷ of NMI or NMII bacteria. The numbers of neutrophils (CD11b⁺ Ly6G⁺) in liver, lung, and spleen were quantified using flow cytometry at 7 and 14 days postinfection (dpi). The numbers of neutrophils in the liver, lung, and spleen were significantly higher in NMI-infected mice than in uninfected and NMII-infected mice (Fig. 1A and B). Splenomegaly and bacterial burden in the spleen were also measured at 7 and 14 dpi and used as parameters to assess the severity of infection. The results indicate that NMI-infected mice developed significantly greater splenomegaly (Fig. 1C) and a higher bacterial burden in the spleen (Fig. 1D) than uninfected and NMII-infected mice. Collectively, these results demonstrate that neutrophil accumulation correlates with the inflammatory response during NMI infection, suggesting that accumulation of neutrophils might be important for the NMI-induced inflammatory response.

FIG 1.

Accumulation of neutrophils is associated with inflammation. BALB/c mice were challenged with 1 × 10⁷ cells of C. burnetii NMI or NMII. (A) Flow cytometry analysis of Ly6G⁺ neutrophils in the liver, lungs, and spleen. Dot plots show the increase in CD11b⁺ Ly6G⁺ cells after 14 days of infection in lungs, liver, and spleen during NMI infection but not in NMII and control groups. Gated cells are highlighted in blue, and the total population in highlighted in black. The percentages shown in the graphs represent the total CD11b⁺ Ly6G⁺ neutrophils. (B) Bar graph showing increases in neutrophil population in lungs, liver, and spleen after NMI infection at 7 and 14 dpi but not in the case of NMII. (C) Bar graph showing the splenomegaly index after NMI infection at 7 and 14 dpi but less significant results or no inflammation post-NMII infection. (D) Bar graph showing the log₁₀ com1 genomic copy number determined by RT-PCR in NMI- and NMII-infected spleen. An asterisk at the top of a bar indicates that the value is significantly different from that for the uninfected control at a P of <0.005.

NMI-infected BMNs survive longer than NMII-infected BMNs.

From the above-described animal experiment, it is evident that there is heterogeneity in neutrophil responses to NMI and NMII infections. To determine if NMI and NMII strains differentially interact with neutrophils in vitro, mouse bone marrow neutrophils (BMNs) were used to examine whether the neutrophil survival rate differs between NMI- and NMII-infected neutrophils by flow cytometry. To achieve maximum purity and avoid neutrophil activation, magnetic separation was used to isolate BMNs. The flow cytometry analysis indicated that the purity of neutrophils (CD11b⁺ Ly6G⁺) isolated by immunomagnetic separation was 98% (Fig. 2A and B). To compare the survival rates between NMI- and NMII-infected neutrophils, 1 × 10⁶ purified BMNs were infected with NMI or NMII at a multiplicity of infection (MOI) of 10 and 100 for 48 h and the viable cells were quantified by flow cytometry. The numbers of CD11b⁺ Ly6G⁺ events were recorded for 1 min and are represented as percent survival relative to that at 0 hpi. As shown in Fig. 2C, the survival rate of NMI-infected BMNs was 43% at an MOI of 10 and 52% at an MOI of 100, which was significantly higher than the survival rate of NMII-infected BMNs, which was 21% at an MOI of 10 and 22% at an MOI of 100. To determine the infection rate, BMNs were infected with NMI or NMII at an MOI of 100 and the infection rate was calculated using immunofluorescence staining by determining the 4′,6′-diamidino-2-phenylindole (DAPI)/C. burnetii ratio in 200 events. As shown in Fig. 2D, the infection rates of NMI- and NMII-infected BMNs at 48 hpi were 32% and 44%, respectively. In addition, the growth curves of NMI and NMII organisms in BMNs at 24, 48, 72, and 96 hpi were also determined by quantifying com1 genomic copy numbers using real-time PCR (RT-PCR). As shown in Fig. 2E, although com1 genomic copy numbers in NMII-infected BMNs are significantly higher than in NMI-infected BMNs at 48 and 72 hpi, the com1 genomic copy numbers decreased over time in both NMI- and NMII-infected BMNs, indicating that both NMI and NMII organisms were unable to replicate in neutrophils. In contrast, as shown in Fig. 2F, mouse tfrc genomic copy numbers are significantly higher in NMI-infected BMNs than in NMII-infected BMNs at 24 and 48 hpi, indicating that live cell numbers in NMI-infected BMNs are significantly higher than in NMII-infected BMNs. This result correlates with the BMN survival rate as determined by fluorescence-activated cell sorting (FACS) and supports that NMI-infected BMNs survive longer than NMII-infected BMNs. Furthermore, live-cell imaging and flow cytometry analysis indicated that NMII-infected BMNs exhibited larger cellular structures with more granularity than NMI-infected BMNs (Fig. 2G and H). Collectively, these results demonstrate that NMI and NMII differentially activate BMNs, which leads to NMI-infected BMNs surviving longer than NMII-infected BMNs.

FIG 2.

NMI-infected purified BMNs exhibit the maximum survival rate. Neutrophils were purified from the bone marrow of BALB/c mice by using magnetic beads, and the purity was confirmed by flow cytometry. (A) Dot plot showing the total cell population from the bone marrow. (B) Dot plot showing the percentage of neutrophils after purification. R7 represents the CD11b⁺ Ly6G⁺ neutrophils. (C) Survival rate of neutrophils after NMI and NMII infection at 48 hpi. The percent neutrophil survival was determined by counting the number of live Ly6G⁺ neutrophils captured for 1 min at 48 hpi. Bars represent the mean result for three individual experiments, and the asterisk at the top of a bar indicates that the value is significantly different from that of the control at a P of <0.005. (D) Infection rate was determined by measuring the ratio of C. burnetii-positive cells to DAPI⁺ cells using Celleste software. The rates of infection of NMI and NMII are displayed with the merged image. (E) The com1 genomic copy numbers were determined in NMI- and NMII-infected BMNs at 24, 48, 72, and 96 hpi and are represented as survival of C. burnetii (%) relative to that at 24 hpi. (F) The genomic copy numbers of tfrc, a mouse endogenous gene, were determined in NMI- and NMII-infected BMNs at 0, 24, 48, 72, and 96 hpi and are represented as survival of C. burnetii (%) relative to that at 0 hpi. (G) Neutrophils infected with NMI or NMII in a 96-well plate were imaged in a Lionheart live-cell imaging system for 48 h. Images captured at 0, 4, 24, and 48 h postinfection indicate that NMII-infected BMNs but not NMI-infected BMNs undergo structural modification over the time of infection. The magnified image shows clear changes in the cell size and structure of NMII-infected BMNs. (H) Flow cytometry analysis shows a shift in forward scatter (FSC) in NMII-infected BMNs compared to the uninfected and NMI-infected cells, confirming the increased cell size of neutrophils.

NMI- and NMII-infected BMNs exhibit differential gene expression patterns.

Advances in next-generation sequencing technology made RNA sequencing (RNA-seq) a reliable technique for analyzing global dynamic changes in gene expression between biological samples. To understand the molecular mechanism behind the heterogeneity in neutrophil responses to NMI and NMII organisms, we compared the RNA-seq pattern of NMI- and NMII-infected BMNs with that of the uninfected BMN controls at 1 hpi (early) and 20 hpi (late). As shown in Fig. 3A to D, the transcriptional profiles of NMI- or NMII-infected BMNs were significantly different from that of the uninfected BMNs. In addition, compared to the response at 1 hpi, a strong transcriptional response was exhibited at 20 hpi. As shown in Fig. 3E to H, the deep data analysis indicated that 8,687 genes were differentially expressed between the C. burnetii-infected and uninfected BMNs, and several unique and common differentially expressed genes (DEGs) were identified between NMI- and NMII-infected BMNs. Compared to uninfected BMNs, 19 genes were upregulated and 1 gene was downregulated in NMII-infected BMNs at 1 hpi, but 137 genes were upregulated and 5 genes were downregulated in NMI-infected BMNs (Fig. 3G). At the later stage (20 hpi) of infection, both NMI- and NMII-infected BMNs exhibited strong and distinct transcriptional responses compared to uninfected BMNs (Fig. 3H). Interestingly, 154 genes were differentially (>3-fold) expressed between NMI- and NMII-infected BMNs (Fig. 3I). These results indicate that NMI and NMII organisms induce a distinct transcriptional response in neutrophils.

FIG 3.

Comparative RNA-seq depicts differentially expressed genes (DEGs) between NMI- and NMII-infected BMNs. A comparative DEG analysis shows the changes between NMI- and NMII-infected neutrophils (1 and 20 hpi) compared to the uninfected neutrophils. (A to D) MA plots showing the M (log ratio)and A (mean average) of differential expression magnitude between PBS control and infected neutrophils. DEGs are represented as red dots. During the early phase of NMI infection (1 h) in the neutrophils, there are notable DEGs, whereas NMII infection did not induce differential expression in the early stage. But in the later phase (20 h), both NMI- and NMII-infected neutrophils expressed a significant differential expression pattern. (E to H) Venn diagrams showing the unique and common DEGs of NMI- and NMII-infected neutrophils at 1 and 20 hpi. (I) Heat map representing the differential expression pattern of genes coding for various cytokines, chemokines, and other signaling molecules between NMI- and NMII-infected neutrophils at 1 and 20 hpi. Blue indicates downregulated genes, and red indicates upregulated genes.

Validation of gene expression pattern by RT-PCR.

To confirm and validate the DEG data, quantitative real-time PCR analysis was performed using a set of 45 selected genes at three different time points (4, 24, and 48 hpi). These genes were distributed across the range of expression changes between NMI- and NMII-infected BMNs predicted to be involved in various immunologically relevant pathways. The results confirmed our DEG findings, and all of the selected 45 genes were differentially expressed between uninfected, NMI-infected, and NMII-infected BMNs (Fig. 4A). Moreover, the pattern of differential expression was also concordant with the DEG results, thus validating our comparative RNA-seq findings. See Table S1 in the supplemental material for details about the list of 45 genes and the primers. Among the 45 genes, the expression patterns of 16 genes are represented in Fig. 4B. Notably, autophagy-related genes PI3Kc2g and Lpl are upregulated in NMII-infected BMNs but downregulated in NMI-infected BMNs. In addition, in comparison to NMI-infected BMNs, several genes, including Plcd1, Rab7, Ifnγ, IL-7, IL17A, Epha4, Colec12, Prdx1, and Gsta3, are highly expressed in NMII-infected BMNs. In contrast, genes such as Ifnβ, IL-5, Sult4a1, and Cxcl2 are more highly upregulated in NMI-infected BMNs than in NMII-infected BMNs. DEG and RT-PCR analyses showed that Sgk1, an antiapoptotic factor, was downregulated in NMII-infected BMNs but slightly upregulated in NMI-infected BMNs (Fig. 4A). Ifnγ was upregulated in NMII-infected BMNs but downregulated in NMI-infected BMNs. Inversely, Ifnß1 was significantly upregulated in NMI-infected BMNs but has no change in expression in NMII-infected BMNs. Ifnα2 gene expression was highly downregulated in NMI-infected BMNs but not in NMII-infected BMNs (Fig. 4B). IL-17 and IL-7 interleukin gene expression was highly upregulated in NMII-infected BMNs but did not change in NMI-infected BMNs. The proinflammatory cytokine IL-5 was significantly upregulated in NMI-infected BMNs but did not change in NMII-infected BMNs (Fig. 4B). These results suggest that the 45 genes reported in this study might play an important role in the heterogeneity of the neutrophil responses against NMI and NMII infection.

FIG 4.

Validation of subset of DEGs by real-time PCR. (A) A heat map represents the differential expression pattern of 45 selected genes in NMI- and NMII-infected neutrophils at 4, 24, and 48 hpi. Data represent ΔΔC_T values of the genes compared to that of uninfected neutrophils with β-actin as the housekeeping gene. Scale, blue represents downregulated genes, while red represents upregulated genes. (B) Bar graphs indicating the relative expression of the 16 selected genes that are discussed in the article at 4, 24, and 48 hpi, with error bars indicating the standard deviation. The data represent the mean of results of three biological replicates and three technical replicates.

Upregulated genes in NMII-infected BMNs involved in autophagy-related pathways.

Given the profound changes in the expression pattern of the subset of genes identified by RNA-seq analysis and validated by RT-PCR analysis, these genes might be associated with few or many biological pathways involving the immunological response of neutrophils. To investigate the difference in pathway activation between NMI and NMII, upregulated genes from NMII-infected BMNs and downregulated genes from NMI-infected BMNs were subjected to protein network analysis using the STRING database. The protein-protein interaction network indicated that genes upregulated in NMII-infected BMNs (Fig. 5) but downregulated in NMI-infected BMNs (Fig. 6) are involved in the Rap1 signaling pathway and the neutrophil degranulation pathway (Fig. 5B and Fig. 6B). A few genes, such as Plau, Plaur, and Lpl, are upregulated in NMII-infected BMNs but downregulated in NMI-infected BMNs. However, a different set of genes involved in those pathways was differentially expressed between NMI- and NMII-infected BMNs, which indicates that NMI and NMII might manipulate BMNs via different mechanisms. Upregulation of phagosome-related genes and Rab regulation genes in NMII-infected BMNs suggests that NMII might induce autophagy in neutrophils. Notably, three genes involved in PI3P regulation, PI3Kc2g, Plcd1, and Lpl, were significantly upregulated in NMII-infected BMNs but downregulated in NMI-infected BMNs (Fig. 4B). Protein network analysis indicated that the three corresponding proteins are functionally interconnected with proteins involved in autophagy-related pathways. Functional annotation indicated that these three proteins are involved in lipid metabolism. Reactome pathway analysis indicated the interaction of PI3Kc2g, Plcd1, and Lpl genes in the membrane trafficking of multiple cellular organelles, including autophagosomes (Fig. 7).

FIG 5.

Protein network analysis of upregulated genes in NMII-infected BMNs. Prediction of the protein network was done using the STRING database with the upregulated genes (>3-fold) from NMII-infected BMNs. (A) Network analysis showing that the genes are interconnected in multiple protein network pathways, including the PI3K/AKT pathway (purple), RAP1 signaling pathway (green), cellular response to stimulus (red), and focal adhesion (yellow). (B) Graph indicating the number of genes involved in each pathway.

FIG 6.

Protein network analysis of downregulated genes in NMI-infected BMNs. NMI-infected neutrophils undergo immediate differential gene expression during the early stage and also survive for a longer time. To understand the molecular mechanism, the downregulated genes (>3-fold) were used to predict the protein network using the STRING database. (A) Mapping of downregulated genes in NMI-infected neutrophils revealed that pathways such as vesicle-mediated transport (yellow), focal adhesion (purple), angiogenesis (green), and RAP1 signaling (red) are critically downregulated. (B) Graph indicating the number of genes involved in each pathway.

FIG 7.

Reactome pathway analysis depicts NMII influences on membrane trafficking. Reactome pathway analysis was performed with selected DEGs that are upregulated in NMII-infected BMNs. The pathway indicates that conversion of glutathione (GSH) to GSSG and membrane trafficking in autophagosomes are activated in NMII-infected BMNs. COLEC12, ITGB2, SLC23A1, SULT2B1, GSTA3, PI3Kc2g, PLCD1, and LPL genes are represented by pink boxes, the targets such as PAPS, PAP, DAG, Fatty acid are represented by green ovals, and the by-products such as ADP, ATP are represented by orange ovals.

NMII strain induces LC3-mediated autophagy in neutrophils.

Protein-protein interaction and reactome pathway analyses indicated that the major DEGs were associated with autophagy and survival pathways, suggesting that NMI and NMII organisms can differentially activate autophagy in neutrophils. To confirm this phenomenon, NMI- or NMII-infected BMNs were stained with CYTO-ID autophagy detection dye and analyzed by flow cytometry. The CYTO-ID dye selectively labels accumulated autophagic vacuoles, which will stain autophagosomes for monitoring autophagic flux in live cells. As shown in Fig. 8A and B, the number of CYTO-ID-positive (CYTO-ID⁺) cells in NMII-infected BMNs was significantly higher than that in uninfected and NMI-infected BMNs. In addition, the number of CYTO-ID⁺ cells in NMII-infected BMNs at an MOI of 100 was significantly higher than in NMII-infected BMNs at an MOI of 10. These results indicate that NMII infection induced a robust autophagy response in neutrophils compared to NMI. An immunofluorescence assay and immunoblotting were also used to confirm if NMI and NMII differentially activate key autophagy-related proteins in neutrophils. As shown in Fig. 8C, larger amounts of LC3I and cleaved LC3II, which are involved in the maturation of autophagosomes, were detected in NMII-infected BMNs than in uninfected and NMI-infected BMNs at both 24 and 48 hpi. However, although the amounts of proteins involved in early phagosome formation, such as Beclin, ATG5, and Rab7, were similar between NMI- and NMII-infected BMNs at 24 hpi, larger amounts of those proteins were detected in NMII-infected BMNs than in NMI-infected BMNs at 48 hpi. The intensity units of the Beclin, ATG5, Rab7, and LC3II protein bands in NMII-infected BMNs were significantly higher than those in uninfected and NMI-infected BMNs at 48 hpi (Fig. 8D). Furthermore, colocalization of C. burnetii with LC3 (Fig. 8E) and Beclin (Fig. 8F) proteins was observed in both NMI- and NMII-infected BMNs, suggesting that C. burnetii is localized inside the phagosome that expresses LC3 and Beclin. Collectively, these results suggest that NMII infection may activate autophagosome maturation via LC3 lipidation while NMI infection does not.

FIG 8.

NMI infection inhibits, but NMII infection activates, autophagy in neutrophils. Neutrophils were infected with NMI or NMII at an MOI of 10 and 100, and the difference in autophagy and apoptosis was measured using flow cytometry and microscopy. (A) Neutrophils were stained with CYTO-ID autophagy detection dye after 48 hpi, and flow cytometry was performed. A histogram indicates that more NMII-infected BMNs than control and NMI-infected BMNs are CYTO-ID positive. (B) Graph showing the number of autophagy-positive neutrophils after infection with NMI or NMII at an MOI of 10 and 100 at 48 hpi. Error bars represent the standard deviation of results of three individual samples, and an asterisk indicates that the value is significantly different from that of the uninfected control at a P of <0.005 and # symbol indicates that the value is significantly different from NMI at a P of <0.005. (C) Western Blot analysis of autophagy-related proteins Beclin, ATG5, Rab7, and LC3 and the internal control, β-actin, in uninfected BMNs and NMI- or NMII-infected BMNs at an MOI of 100 at 24 and 48 hpi. (D) Relative intensity of Beclin, ATG5, Rab7, and LC3II markedly increased in NMII-infected BMNs compared to uninfected and NMI-infected BMNs. Data represent the mean value of results of three individual experiments. An asterisk at the top of a bar indicates that the value is significantly different from that of the control at a P of <0.005, whereas a number sign indicates that the value is significantly different from that of the NMI-infected BMNs at a P of <0.005. (E) Confocal microscopy images of BMNs infected with NMI or NMII at an MOI of 100 after 48 hpi, stained with DAPI (blue), LC3 (green), and C. burnetii (red). Merged images represent the colocalization of LC3 and C. burnetii. (F) Confocal microscopy images of BMNs infected with NMI or NMII at an MOI of 100 after 48 hpi, stained with DAPI (blue), Beclin (green), and C. burnetii (red). Merged images represent the colocalization of Beclin and C. burnetii.

NMII infection also induced an LC3-mediated autophagy in human macrophages.

THP-1 cell-derived macrophages were used to investigate if NMI and NMII organisms can differentially activate autophagy in human cells. Following phorbol 12-myristate 13-acetate (PMA) differentiation of monocytes into macrophages, THP-1 cells were infected with NMI or NMII at an MOI of 100 for 24 h and then extracellular bacteria were removed and fresh culture medium was added. The expression of autophagy-related proteins was compared between NMI- and NMII-infected human macrophages by an immunofluorescence assay and immunoblotting at 48 hpi. As shown in Fig. 9A, larger amounts of Beclin, ATG5, Rab7, LC3I, and cleaved LC3II proteins were detected in NMII-infected macrophages than in uninfected and NMI-infected macrophages. The intensity units of the Beclin, ATG5, Rab7, LC3I, and LC3II protein bands in NMII-infected macrophages were significantly higher than those in uninfected and NMII-infected macrophages (Fig. 9B). In addition, although colocalization of C. burnetii with LC3 (Fig. 9C) and Beclin (Fig. 9D) proteins was observed in both NMI- and NMII-infected macrophages, the fluorescence intensities of LC3 and Beclin expression in NMII-infected macrophages are stronger than those in uninfected and NMI-infected macrophages. These results demonstrate that NMI and NMII organisms differentially activate autophagy in both murine and human innate immune cells.

FIG 9.

NMI infection inhibits, but NMII infection activates, autophagy in human macrophages. THP1-derived macrophages were infected with NMI or NMII at an MOI of 100, and the difference in autophagy was measured using immunoblotting and microscopy. (A) Western Blot analysis of autophagy-related proteins Beclin, ATG5, Rab7, and LC3 and an internal control, β-actin, in uninfected macrophages or macrophages infected with NMI or NMII at an MOI of 100 at 48 hpi. (B) The relative intensities of Beclin, ATG5, Rab7, and LC3II markedly increased in NMII-infected macrophages compared to uninfected and NMI-infected macrophages. Data represent the mean value of results of three individual experiments. An asterisk at the top of a bar indicates that the value is significantly different from that of the control at a P of <0.005, while a # sign indicates that the value is significantly different from that of the NMI-infected macrophages at a P of <0.005. (C) Confocal microscopy images of macrophages infected with or without NMI or NMII at an MOI of 100 after 48 hpi, stained with DAPI (blue), LC3 (green), and C. burnetii (red). Merged images represent the colocalization of LC3 and C. burnetii. (D) Confocal microscopy images of BMNs infected with or without NMI or NMII at an MOI of 100 after 48 hpi, stained with DAPI (blue), Beclin (green), and C. burnetii (red). Merged images represent the colocalization of Beclin and C. burnetii.

NMII infection induced phosphatidylserine translocation in the membrane of neutrophils.

RNA-seq analysis indicated that NMII infection in BMNs upregulated the expression of several genes involved in lipid metabolism. Autophagy is intricately related to the metabolism of lipids, namely phospholipids, among which phosphatidylserine (PS) is a well-studied lipid that is normally restricted to the cytoplasmic face of the plasma membrane but externalized early during cell death. Therefore, to determine the changes in lipid metabolism in neutrophil membranes during C. burnetii infection, PS translocation was assessed using an annexin-V-FLUOS kit and analyzed by flow cytometry. As shown in Fig. 10A, higher levels of annexin-V fluorescence-expressing cells were detected in NMII- or dotA:Tn mutant-infected BMNs than in uninfected and NMI-infected BMNs at 24 hpi. However, the propidium iodide (PI)-stained cells were indistinguishable between uninfected and C. burnetii-infected BMNs at 24 hpi. These results suggest that NMII- and dotA:Tn-infected BMNs undergo PS translocation without inducing cell death at 24 hpi. To further confirm this hypothesis, live-cell imaging was performed using a pSIVA (polarity-sensitive indicator of viability and apoptosis) probe and PI to visualize PS translocation in NMI-, NMII-, or dotA:Tn-infected BMNs at 24, 48, and 72 hpi. As shown in Fig. 10B, the percentages of pSIVA-positive cells in NMII- and dotA:Tn-infected BMNs were similar but significantly higher than those in uninfected and NMI-infected BMNs at 24 hpi. In contrast, the percentages of PI-positive cells in NMI, NMII, or dotA:Tn infected BMNs had no significant difference from those in uninfected BMNs at 24 hpi (Fig. 10C). In addition, as shown in Fig. 10D, although more pSIVA-stained cells were observed in NMII- and dotA:Tn-infected BMNs than in uninfected and NMI-infected BMNs, the levels of PI-stained cells were similar in uninfected and C. burnetii-infected BMNs at 24 hpi. These observations correlate with the flow cytometry data and support the hypothesis that NMII- and dotA:Tn-infected BMNs undergo PS translocation without inducing cell death at 24 hpi. In addition, although the percentages of pSIVA-positive cells in NMII- and dotA:Tn-infected BMNs were similar, they were significantly higher than those in uninfected and NMI-infected BMNs at 48 and 72 hpi (Fig. 10B). These results indicate that NMII- and dotA:Tn-infected BMNs undergo PS translocation but that NMI-infected BMNs inhibit PS translocation. In contrast, compared to uninfected and dotA:Tn-infected BMNs, the percentage of PI positive cells was significantly lower in NMI- or NMII-infected BMNs at 48 and 72 hpi (Fig. 10C). It is notable that the percentage of PI-positive cells in NMI-infected BMNs was significantly lower than that in NMII-infected BMNs at 48 and 72 hpi (Fig. 10C). These results indicate that both NMI and NMII strains can inhibit or delay neutrophil programming death, although the NMI strain showed a stronger ability than the NMII strain, and that the dotA:Tn mutant strain was unable to inhibit or delay neutrophil death.

FIG 10.

NMII induces phosphatidylserine translocation in the membrane of neutrophils. (A) Neutrophils were infected with NMI, NMII, or the dotA:Tn mutant at an MOI of 100 and stained with annexin-V and propidium iodide dyes after 24 hpi, and flow cytometry analysis was performed. A histogram shows the shift in fluorescence of annexin-V (top) and propidium iodide (bottom). The percentage of annexin-V- or propidium iodide-positive cells is displayed in each histogram. BMNs were infected with NMI, NMII, or the dotA:Tn mutant at an MOI of 100, and the cells were stained with pSIVA probe (green) and propidium iodide (red); cell nuclei were stained with DAPI (blue). Images were acquired after 24, 48, and 72 hpi using an EVOS M7000 fluorescent imaging system. (B to D) The numbers of live cells stained with pSIVA and propidium iodide were calculated using Celleste software and are represented as the percentage of pSIVA (B)- and propidium iodide (C)-stained cells; images are representative of three individual experiments (D). Data represent the mean value derived from microscopic images recorded in five different locations. An asterisk at the top of a bar indicates that the value is significantly different from that of the control at a P of <0.005, whereas a number sign indicates that the value is significantly different from that of the NMI-infected BMNs at a P of <0.005.

NMII induced glutathione-mediated oxidative stress in neutrophils.

Glutathione (GSH) deficiency in cells triggers autophagy, where the GSH/GSSG ratio is crucial in this process (16). Our DEG analysis (Fig. 4) showed that the gene encoding Sult2B1, which converts dehydroepiandrosterone (DHEA) to dehydroepiandrosterone-sulfate (DHEAS), is upregulated in NMII-infected BMNs but not in NMI-infected BMNs. Moreover, reactome pathway analysis indicated that the GstA3 gene might be involved in the oxidation of GSH to GSSG, leading to the Keap1-Nrf2 autophagy lysosomal pathway (Fig. 7). In addition, DEG analysis indicated that the Slc23A1 gene is specifically upregulated in NMII-infected BMNs. The ascorbic acid transporter gene Slc23a1 is essential for vitamin C transport into cells, which is required for the oxidation of GSH to GSSG (17). These observations suggest that the genes Sult2B1, GstA3, and Slc23a1 might be involved in the conversion of GSH to GSSG, which might be associated with the activation of autophagy and apoptosis in NMII-infected BMNs.

To confirm the above hypothesis, we analyzed the intracellular levels of GSH and GSSG at 4, 24, and 48 hpi in BMNs infected with NMI and NMII at an MOI of 100. As shown in Fig. 11A and B, the GSH level decreased but the GSSG level increased in a time-dependent manner in NMII-infected BMNs, but the levels of GSH and GSSG did not change over time in NMI-infected BMNs. These results support the pathway analysis findings that the reduction in the GSH/GSSG ratio by the Slc23A1, Sult2B1, and GstA3 genes during NMII infection may lead to activation of autophagy and apoptosis in neutrophils.

FIG 11.

Increased intracellular IL-17 and GSSG levels in NMII-infected BMNs. (A and B) Intracellular GSH and GSSG levels were determined with a glutathione colorimetric kit, and the results are represented as fold change compared with the control. Data showed that GSH is decreased whereas GSSG is increased in NMII-infected neutrophils. No significant change was observed in NMI-infected and uninfected neutrophils. (C) Flow cytometry was used to determine the changes in intracellular IL-17 in NMI- and NMII-infected neutrophils 48 hpi. A histogram shows that NMII-infected neutrophils were stained with higher IL-17 levels than NMI-infected and uninfected neutrophils. (D) Graph showing the percentage of IL-17-positive cells determined by flow cytometry. Data represent the mean value of results of three individual experiments, and an asterisk at the top of a bar indicates that the value is significantly different from that of the control at a P of <0.005.

NMI and NMII infection differentially activate intracellular IL-17 in BMNs.

Apart from the membrane trafficking genes, the expression of a few interleukin genes was also significantly different between NMI- and NMII-infected BMNs. DEG patterns and RT-PCR analysis indicated that IL-17 was more highly expressed in NMII-infected BMNs than in NMI-infected neutrophils. Flow cytometry analysis further confirmed that the level of intracellular IL-17 was higher in NMII-infected BMNs than in NMI-infected BMNs (Fig. 11C and D). IL-17 allows neutrophils to function as antigen-presenting cells and to attract macrophages to kill the bacteria. This suggests that IL-17 might also be involved in the heterogeneity of the neutrophil response against NMI and NMII infections.

DISCUSSION

Previous studies demonstrated that the virulent C. burnetii NMI strain causes fever and an inflammatory response in immunocompetent mice and guinea pigs while the avirulent NMII strain does not (18–20). However, the mechanism underlying how NMI and NMII strains differentially cause disease in animal models remains unclear. In this study, we found that NMI-infected mice developed significantly greater splenomegaly and a higher bacterial burden in the spleen than NMII-infected mice. Flow cytometry analysis indicated that higher numbers of CD11b⁺ Ly6G⁺neutrophils accumulated in the lung, liver, and spleen of NMI-infected mice but the numbers of neutrophils were similar between uninfected and NMII-infected mice. In addition, an in vitro survival study indicated that NMI-infected BMNs survive longer than NMII-infected BMNs. These results suggest that neutrophil accumulation may correlate with the inflammatory response during NMI infection and that accumulation of neutrophils might be important for NMI-induced splenomegaly in mice. Several reports showed that uncontrolled recruitment of neutrophils during infectious conditions eventually increases the severity of inflammatory responses (7), and our findings indeed confirmed the increased recruitment of neutrophils and prolonged neutrophil survival during NMI infection. In addition, RNA-seq revealed significant DEGs in NMI-infected BMNs at 1 hpi, indicating that NMI activates a unique molecular response in BMNs at the early stage of invasion. Similarly, DEGs at 20 hpi suggest that NMI and NMII induce differential responses in neutrophils at the late stage of infection. In addition, to understand the involvement of the selected DEGs during the course of infection, RT-PCR quantification was performed at 4, 24, and 48 hpi to validate the early and late expression patterns of the selected genes, which revealed time-dependent changes in the expression of 45 selected genes between NMI- and NMII-infected neutrophils. These data suggest that the differential modulation of neutrophil responses of the NMI and NMII strains may contribute to their distinct ability to cause diseases in animal models.

Comparative transcriptomics, RT-PCR quantification, and pathway analysis demonstrated that the majority of genes upregulated in NMII-infected BMNs are functionally interconnected and involved in autophagy-related pathways, including Rab regulation, neutrophil degranulation, membrane trafficking, and EPH signaling pathways. Macroautophagy plays critical roles in the majority of immune cells, including neutrophils, in the clearance of intracellular bacterial pathogens, where the bacterial cells enclosed in vesicles are released to lysosomes for degradation. Previous reports demonstrated that the LC3-mediated maturation of autophagosomes is a critical event in NMII-infected macrophages, several primary cells, and cell lines (21, 22, 23). To determine if NMI and NMII strains differentially manipulate autophagy pathways in innate immune cells, we analyzed LC3-mediated autophagy in NMI- and NMII-infected BMNs. Our results showed that (i) LC3⁺ phagosomes in NMII-infected BMNs were significantly higher than in NMI-infected BMNs, (ii) autophagy-related LC3 and Beclin proteins were strongly expressed and colocalized with C. burnetii in NMII-infected BMNs but fewer LC3⁺/Beclin⁺ cells were detected in NMI-infected BMNs, (iii) larger amounts of LC3I and cleaved LC3II proteins were detected in NMII-infected BMNs than in NMI-infected BMNs, and (iv) more large cellular structures with high granularity were observed in NMII-infected BMNs than in NMI-infected BMNs, demonstrating that NMII infection induces maturation of autophagy in neutrophils while NMI infection does not. In addition, the observations that larger amounts of Beclin, LC3I, and cleaved LC3II proteins were detected in NMII-infected THP-1 cell-derived human macrophages than in NMI-infected macrophages and that the fluorescence intensities of LC3 and Beclin expression in NMII-infected macrophages are stronger than those in NMI-infected macrophages demonstrate that NMII infection induces maturation of autophagy in human macrophages but NMI infection does not. Mizushima and Yoshimori (24) suggested that comparison of the amounts of LC3II irrespective of LC3I between samples can be an indicator for autophagic flux in cells. Therefore, based on significant increases of LC3II protein in NMII-infected neutrophils, it is possible to conclude that NMII-infected neutrophils undergo maturation of autophagosome. Raoult et al. (25) also demonstrated that NMII induces Rab7-mediated phagolysosome formation in macrophages but that NMI does not activate phagolysosome maturation in macrophages. This is consistent with our finding that NMII infection induces LC3-mediated phagolysosome maturation in neutrophils. The maturation of autophagosomes and involvement of autophagy-related proteins during NMII infection have also been demonstrated in primary human alveolar macrophages and Chinese hamster ovary cells (23, 26, 27). Additionally, a recent investigation revealed a functional link between CLTC (clathrin heavy chain) and autophagy during NMII infection in HeLa cells (28). Furthermore, Newton et al. (29) reported that a functional Cig2 protein of NMII is important for interactions between the Coxiella containing vacuole (CCV) and host autophagosomes in HeLa cells. Together, these data suggest that NMII infection-induced maturation of autophagy is cell type independent, which may be responsible for triggering the host defense mechanisms leading to clearance of bacteria in NMII-infected animals.

The observation that, other than LC3, the proteins involved in the early phase of autophagy, Beclin, ATG5, and Rab7, are similar between NMI- and NMII-infected BMNs at 24 hpi indicates that the vacuole formation ability may be similar between NMI- and NMII-infected BMNs. This hypothesis was supported by the previous findings that NMI and NMII strains exhibit similar CCV-forming abilities and replication patterns (30). It is notable that although ATG5 in NMI-infected BMNs was significantly higher than in NMII-infected BMNs at 24 hpi, NMII-infected BMNs exhibited more ATG5 than NMI-infected BMNs at 48 hpi. In addition, Rab7 was overexpressed in both NMI- and NMII-infected BMNs at 24 hpi but was significantly increased in only NMII-infected BMNs at 48 hpi. ATG5 was known to be critical for phagosome formation, maturation, and fusion of autophagosome with the lysosome (31). Thus, these observations suggest that phagosome formation might be similar between NMI- and NMII-infected BMNs during the early stage of infection but that the maturation and fusion of autophagosome are activated only in NMII-infected BMNs at the late stage of infection, which might lead to the clearance of NMII by neutrophils.

Notably, among several DEGs, genes involved in lipid metabolism, including Pi3kc2g, Plcd1, and Lpl, were more highly expressed in NMII-infected BMNs than in NMI-infected BMNs. Reactome pathway analysis also indicated that these genes are mainly involved in the positive regulation of PtdIns(4,5)P2 and PtdIns(4)P, specifically in lipid metabolism and inositol phosphate metabolism. PtdIns(4,5)P2 is well established as an essential lipid messenger at the plasma membrane and other intracellular components, including autolysosomes (32, 33). The composition of lipids such as phosphatidylinositol and phosphatidylserine on the inner leaflets of the plasma membrane is necessary for maintaining the integrity of the phospholipid bilayer (34). Thus, the observation that genes involved in lipid metabolism were more highly expressed in NMII-infected BMNs than in NMI-infected BMNs suggests that NMII-infected BMNs might undergo PS translocation in the membrane. The results that higher levels of annexin-V fluorescence-expressing cells were detected in NMII- or dotA:Tn-infected BMNs than in NMI-infected BMNs at 24, 48, and 72 hpi support this hypothesis and demonstrate that NMII- or dotA:Tn-infected BMNs undergo lipid metabolism-related changes in the membrane but NMI-infected BMNs do not. Since PS externalization on the cell surface of neutrophils serves as a common recognition signal for macrophages and is associated with the cell death program in neutrophils (35), it is possible to conclude that NMII-infected neutrophils might be cleared by the macrophages but NMI-infected neutrophils may escape this host defense mechanism. This might explain why higher numbers of neutrophils accumulated in the lung, liver, and spleen of NMI-infected mice but there was no neutrophil accumulation in NMII-infected mice. In contrast, the observation that PI-positive cells in uninfected and dotA:Tn-infected BMNs were significantly higher than in NMI- or NMII-infected BMNs at 48 and 72 hpi demonstrates that both NMI and NMII strains can inhibit or delay neutrophil programming death while the dotA:Tn mutant strain was unable to inhibit or delay neutrophil death. In addition, the result that PI-positive cells in NMI-infected BMNs were significantly lower than those in NMII-infected BMNs at 48 and 72 hpi indicates that the NMI strain is stronger than the NMII strain in inhibiting or delaying neutrophil programmed death. The ability of the NMII strain to inhibit neutrophil death has been demonstrated in our previous study (10), which indicates that NMII inhibits cell death in neutrophils by exploiting the Mcl1 protein. However, the present study also demonstrated that NMII infection induces PS translocation in neutrophils, which may lead to cell death, while NMI infection inhibits PS translocation in neutrophils. The difference between NMI and NMII in inducing PS translocation in neutrophil membranes may be responsible for their distinct abilities to inhibit or delay neutrophil programming death. Collectively, these results demonstrate that (i) NMII- and dotA:Tn-infected BMNs undergo significant PS translocation compared to NMI-infected BMNs, (ii) uninfected neutrophils undergo PS translocation-independent cell death, (iii) although both NMI and NMII strains can inhibit or delay neutrophil death, the NMI strain showed a stronger ability to inhibit or delay neutrophil death, and (iv) the dotA:Tn mutant strain lost the ability to inhibit or delay neutrophil death.

DEG analysis indicated that three GSH-related genes, Sult2B1, Slc23A1, and GstA3, were significantly upregulated in NMII-infected BMNs, but these genes were not upregulated in NMI-infected neutrophils. Colorimetric analysis of GSH and GSSG levels in C. burnetii-infected neutrophils also showed that the level of GSH was reduced, but the level of GSSG was increased, in NMII-infected BMNs. Decreasing levels of GSH have been linked to reactive oxygen species (ROS) generation, which accelerates mitochondrial damage and induces apoptosis (36). It has been shown that reduction of the levels of GSH/GSSG is also associated with increased reactive oxygen species levels, which further activates the Keap1-Nrf2 autophagy lysosomal pathway (37). Moreover, regulation of glutathione is important to maintain the oxidative stress during bacterial infection (38). Thus, alteration in the GSH/GSSG levels might be associated with the autophagy in NMII-infected BMNs. These findings suggest that reduction of GSH/GSSG levels might be associated with autophagy in NMII-infected BMNs.

Interleukins produced by neutrophils shape innate and adaptive immune responses. This study showed that gene expression and intracellular protein expression of IL-17 was upregulated in NMII-infected BMNs. IL-17 is important for the recruitment of macrophages for phagocytosis of neutrophils at the site of infection, which prevents the inflammatory response (39). This observation suggests that NMII bacteria activate IL-17 expression in neutrophils, which might further induce macrophage-mediated clearance of neutrophils that are undergoing autophagy and apoptosis. Another interleukin gene, IL-7, is also upregulated in NMII-infected BMNs. IL-7 is a type I glycoprotein that is essential for T cell development and homeostatic proliferation, mouse B cell development, and the generation of CD4⁺ and CD8⁺ memory T cells (40). Moreover, NMII infection-upregulated EPH signaling in neutrophils might also activate T cells. Thus, it is possible that NMII-infected BMNs could attract more T cells to the site of infection, as CD4⁺ and CD8⁺ T cells are important for the clearance of C. burnetii infection (41). However, the gene encoding proinflammatory cytokine IL-5 is upregulated in NMI-infected BMNs but not in NMII-infected BMNs. IL-5 induces a pulmonary inflammatory response via eosinophil-mediated inflammation in lungs, which results in airway inflammation, airflow obstruction, and airway hyperresponsiveness (42). This finding suggests that NMI-infected BMNs activate the IL-5-mediated inflammatory response at the site of infection, which may explain why significant inflammation was observed in the lung, liver, and spleen of NMI-infected mice. The Ifnγ gene was upregulated in NMII-infected BMNs but was downregulated in NMI-infected BMNs. It has been shown that Ifnγ released by neutrophils is critical for the activation of macrophages, leukocytes, and cell killing and links innate and adaptive immunity (43). Thus, the release of Ifnγ from NMII-infected neutrophils may help in the clearance of NMII bacteria from the localized infectious site while NMI bacteria may inhibit this host defense mechanism by downregulation of Ifnγ in neutrophils.

In summary, this study demonstrates that NMII infection induces neutrophil membrane remodeling via PS translocation associated with LC3 lipidation-mediated autophagosome maturation in neutrophils, which might be responsible for triggering host-protective immune responses. In contrast, NMI infection inhibits autophagosome maturation in neutrophils, resulting in prolonged survival, secretion of proinflammatory interleukins, and recruitment of more neutrophils to the site of infection. Thus, targeted activation of autophagy in neutrophils might resolve inflammation in organs of infected animals, thereby preventing severe tissue damage and mortality during intracellular bacterial infection.

In addition, it is important to mention that most of the genes activated by NMI in neutrophils are potentially involved in survival pathways, dysregulated autophagy, and apoptosis mechanisms, which are common features in tumor formation. Considering this fact, NMI infection might have a strong ability to induce cancer. This hypothesis was supported by a previous report that demonstrated an association between C. burnetii infection and B-cell non-Hodgkin lymphoma in humans (44). These observations highlight the potential of this intracellular bacterial infection to cause human cancers. However, further studies are required to determine whether activation of the survival pathway and manipulation of autophagy by NMI relate to cancer development in both in vitro and in vivo models.

Conclusions.

Accumulation of neutrophils is positively correlated with the inflammatory response during NMI infection in mouse lungs, liver, and spleen. NMI-infected BMNs survived longer than NMII-infected BMNs, suggesting that the longer survival of neutrophils might increase the accumulation of neutrophils in various organs. Comparative RNA-seq analysis revealed that the transcriptional profile of neutrophils infected with NMI and NMII exhibit significant DEGs. The pathway analysis demonstrated that NMII-infected BMNs undergo autophagy, whereas NMI infection inhibits autophagy in neutrophils. In addition, LC3 staining and PS externalization assays confirmed that NMII infection induces autophagosome maturation and apoptosis in neutrophils, while NMI infection does not induce autophagy and apoptosis. Collectively, these findings suggest that NMI and NMII differentially manipulate neutrophil autophagy and that apoptosis may be responsible for their different abilities to cause diseases in animal models.

MATERIALS AND METHODS

Ethics statement.

All research involving animals was conducted in accordance with the animal care and use guidelines, and all animal use protocols were approved by the Animal Care and Use Committee at the University of Texas at San Antonio and the University of Missouri–Columbia.

C. burnetii culture.

C. burnetii Nine Mile phase I (NMI) clone 7 (RSA 493) and Nine Mile phase II (NMII) clone 4 (RSA 439) were used for both in vitro and in vivo experiments. The bacterial cells were cultured in acidified citrate cysteine medium 2 (ACCM-2) as previously described (45). The bacterial cells were pelleted by centrifugation at 15,000 × g for 30 min, followed by 2 washes with sterile 1× phosphate-buffered saline (PBS). The pathogenic strain NMI was cultured and maintained and all the experiments that involved NMI were handled in the biosafety level 3 (BSL3) laboratory at the University of Texas at San Antonio and the University of Missouri Laboratory for Infectious Disease Research (MU-LIDR).

Animal infection.

Eight-week-old female BALB/c mice were anesthetized using isoflurane gas. A total of 1 × 10⁷ NMI or NMII bacterial cells in 400 µL of PBS were used to challenge each mouse via intraperitoneal injection. Animals were housed in sterile microisolator cages, and all the infections were carried out in an animal biosafety level 3 (ABSL3) facility at the MU-LIDR. Animals were provided food and water ad libitum. Splenomegaly was measured after 7 or 14 days of infection by comparing the percentage of splenomegaly [(spleen weight/body weight) × 100] with that of uninfected controls. Each experiment contained four animals per group.

Cell suspension preparation and flow cytometry.

To identify the accumulation of neutrophils in different tissues, mouse organs such as lungs, liver, and spleen were dissected under sterile conditions at 7 or 14 dpi. Dissected tissues were immediately homogenized, and the cell suspension of each organ was filtered through 100-µm nylon mesh using complete RPMI medium and pelleted by centrifugation at 500 × g for 5 min. Then, erythrocytes (RBCs) in the cell suspension were lysed by adding 1× ACK lysis buffer and incubated at room temperature for 5 min, and the reaction was stopped by the addition of excess RPMI medium and pelleted by centrifugation at 500 × g for 5 min. Finally, the cells were resuspended in MACS buffer (Miltenyi Biotec, Inc.) and counted using a hemocytometer. The total number of neutrophils in each organ with or without infection was counted using flow cytometry. Briefly, 1 × 10⁶ cells were stained with antibody cocktail containing anti-mouse Ly6G (APC-cy7) (BioLegend) and anti-mouse CD11b (fluorescein isothiocyanate [FITC]) (BioLegend) for 30 min at 4°C without light. Cells were pelleted by centrifugation at 500 × g for 5 min, followed by two washes with MACS buffer, and fixed with 4% paraformaldehyde before analysis in a MoFlo XDP cell sorter (Beckman Coulter). A total of 100,000 cellular events were recorded for each sample, and the data were analyzed using Summit software.

Bacterial quantification.

The bacterial load in culture stock and the bacterial burden in spleen cells were determined by quantitative real-time PCR (qRT-PCR) analysis of the com1 copy number from the genomic DNA. Briefly, genomic DNA was isolated either from the culture or from spleen cells using lysis buffer (1 M Tris, 0.5 M EDTA, 7 mg/mL glucose, 28 mg/mL lysozyme), and 10 µl proteinase K (20 mg/mL) was added to each sample prior to incubation at 60°C for 18 h. In addition, 21 µl 10% SDS was added, samples were incubated at room temperature for 1 h, and DNA was extracted using the High Pure PCR template preparation kit (Roche, Indianapolis, IN) according to the manufacturer’s instructions. SYBR green (Applied Biosystems, Foster City, CA)-based qPCR analysis was performed using com1-specific primers, and the absolute bacterial count was determined using the standard curve generated using recombinant plasmid DNA (com1 gene ligated into pBluescript vector).

Neutrophil isolation.

Neutrophils were isolated from bone marrow of 8-week-old female BALB/c mice, which are housed in pathogen-free microisolator cages fed with regular diet at the University of Missouri laboratory animal facility. Neutrophils were isolated from the bone marrow of the mice by using a Miltenyi magnetic beads kit. Briefly, mice were euthanized using CO₂ exposure, and cervical dislocation was performed to kill the animal. Using sterile techniques, the femur and tibia were collected and the bone marrow cells were flushed with 3 mL of Hanks balanced salt solution (HBSS) (Mg-, Ca-) using a 25-gauge needle and centrifuged at 400 × g for 10 min. The pellet was resuspended with ACK lysis buffer for 5 min until the RBCs lysed completely. Immediately after RBC lysis, the reaction was neutralized by adding an equal volume of HBSS, and the contents were passed through the 70-µm cell strainer and centrifuged at 400 × g for 10 min. The cell pellet was washed with 10 mL of HBSS and centrifuged at 400 × g for 10 min. The pellet obtained was used for magnetic separation using a neutrophil isolation kit (Miltenyi Biotech GmBH, Germany) in accordance with the manufacturer’s instructions. Trypan blue was used to determine the viability, which was greater than 99%, the cells were counted using a hemocytometer, and the purity of the neutrophils was assessed by flow cytometry using anti-mouse Ly6G (phycoerythrin [PE]) antibody (BioLegend), which was greater than 98% for each assay.

THP1 macrophage culture and C. burnetii infection.

Human monocyte-like (THP-1) cells (TIB-202; ATCC) were maintained in 25-cm² tissue culture flasks containing RPMI 1640 medium (Gibco) supplemented with 10% fetal bovine serum (Gibco), 100 U/mL of penicillin, and 100 µg/mL of streptomycin (Gibco) at 37°C in 5% CO₂. Before infection, THP-1 monocytic cells were cultured in 6-well tissue culture plates (for protein extraction) or in a Lab-Tek II Chamber Slide (for microscopy) in the presence of 200 nM phorbol 12-myristate 13-acetate (PMA). After 24 h of induction, monocytes differentiated into adherent, macrophage-like cells, the medium containing PMA was removed, and cells were replenished with fresh RPMI medium lacking PMA. Twenty-four hours after addition of the fresh medium, the cells were washed with PBS before use in infection studies.

Adherent macrophage cells maintained in a 6-well plate or chamber slide containing RPMI lacking antibiotic were infected with NMI, NMII, or the dotA:Tn NMII mutant at an MOI of 100 and incubated at 37°C in 5% CO₂. After 24 hpi, medium containing C. burnetii was removed, washed with PBS to remove the extracellular bacteria, and replenished with fresh RPMI medium. At 48 hpi, the cells were subjected to an immunofluorescence assay or protein extraction.

Live-cell imaging.

Freshly isolated neutrophils were infected with either NMI or NMII at an MOI of 100 in a 96-well plate and incubated in the onstage incubator with 5% CO₂ at 37°C for 48 h, and the morphological changes in the neutrophils were captured every 5 min using a Lionheart microscope with a 60× objective (BioTek). Finally, the images and videos were prepared by stitching the montages using Gen5 software.

Bacterial challenge, RNA isolation, and Illumina sequencing.

Freshly isolated neutrophils were plated in 6-well plates at 5 × 10⁶/well, infected with NMI or NMII at an MOI of 100, and incubated at 37°C in 5% CO₂ for 4, 24, and 48 h. Total RNA was isolated from 5 × 10⁶ neutrophils challenged with or without strain NMI or NMII of C. burnetii by using an RNeasy minikit (Qiagen) after 1 hpi and 20 hpi and stored at −80° C until further use. Five groups (3 biological replicates) of total RNA samples (uninfected BMNs, NMI-infected BMNs 1 hpi [NMI1], NMI-infected BMNs 20 hpi [NMI20], NMII-infected BMNs 1 hpi [NMII1], NMII-infected BMNs 20 hpi [NMII20]) were used for the analysis. The integrity of the isolated total RNA was determined by capillary electrophoresis fragment analysis (University of Missouri DNA Core Laboratory), and the 28S/18S rRNA ratio as well as RNA integrity number (RIN) was calculated. All the samples exhibited a RIN score over 9.5, and the samples were subjected to library preparation. Strand-specific RNA-seq libraries were prepared using the Illumina TruSeq HT stranded total RNA library prep kit (Illumina, San Diego, CA), in accordance with the manufacturer’s instructions. All the libraries were normalized, pooled, and sequenced with the Illumina HiSeq 2000 sequencer using a 100-nucleotide paired-end protocol.

RNA-seq data analysis.

The RNA-seq data were aligned against the mouse genome (https://ftp.ensembl.org/pub/release-87/fasta/mus_musculus/dna/Mus_musculus.GRCm38.dna_sm.primary_assembly.fa.gz) using STAR, and the reads were sorted by SAMtools. Further, the strandness information was obtained using the RSeQC-2.6.4 tool by comparing the mouse RseqC reference file (https://sites.google.com/site/liguowangspublicsite/home/GRCm38_mm10_Ensembl.bed.gz?attredirects=0&d=1). Raw counts were converted to log₂ values, and the raw data were normalized using HTSeq by 0.75 quartile normalization. Significance and fold changes in differential expression between the samples were estimated using Gfold. All the genes with a 3-fold difference were considered significant.

Protein network analysis.

To understand the key pathways that are differentially activated between NMI- and NMII-infected neutrophils and to analyze the interactions between the selected DEGs, network analysis was performed using the STRING database (46). Briefly, the top 500 genes which were significantly differentially expressed between NMI- and NMII-infected neutrophils were separated from the master list, the data were submitted to the STRING database via the Multiple Proteins option, and the data were searched against the mouse database. For analyzing protein-protein interactions between the selected genes, we employed the mouse protein-protein interaction database from STRING online database version 11.0, and the interaction prediction was determined with high confidence (0.700). Also, functional ontology analysis was performed, and the key molecular functions that are found in common within the selected genes were determined. Further, the interaction between the selected genes involved in specific pathways was also predicted.

RT-PCR analysis.

To confirm the RNA-seq data, RT-PCR analysis was performed. As mentioned earlier, RNA was isolated, the first strand was synthesized from the total RNA using a QuantiTect reverse transcription kit (Qiagen), and gene-specific PrimeTime primers (Integrated DNA Technologies, Inc.) were used to amplify and quantify the selected differentially expressed genes. qRT-PCR analysis was performed using standard 20-µL reaction mixtures with the SYBR green master mix (Invitrogen) in a StepOnePlus RT-PCR instrument with the standard two-step amplification method. The relative expression of target genes was quantified using the ΔΔC_T relative quantification method by comparison with ß-actin as an internal control.

Fluorescence assay for autophagy.

Autophagy in BMNs infected with either NMI or NMII was determined by performing a fluorescence-based autophagy assay. The Enzo CYTO-ID autophagy detection kit was used in accordance with the instructions provided by the supplier. The kit works on the basis of a proprietary fluorescent probe that has been reported to label autophagosomes. Briefly, 1 × 10⁶ purified BMNs were cultured in a sterile 6-well plate with NMI or NMII at an MOI of 10 or 100 for 24 h in a humidified chamber at 37°C under 5% CO₂. Uninfected BMNs were maintained as a control. After staining, the cells were fixed in 4% paraformaldehyde (PFA) prior to flow acquisition. The percentage of autophagy-positive cells was determined using an FL1 detector in a MoFlo XDP cell sorter (Beckman Coulter). Also, the green fluorescent cells were visualized using a Lionheart FX automated microscope (BioTek Instruments, Inc.).

Immunofluorescence staining.

A total of 1 × 10⁶ freshly isolated and purified BMNs were placed into each well of a Lab-Tek II Chamber Slide (Thermo Fisher), infected with NMII at an MOI of 100, and incubated in a humidified chamber at 37°C under 5% CO₂. After 48 hpi, the cells were washed with sterile 1× phosphate-buffered saline (PBS), fixed with 2% paraformaldehyde (PFA), and permeabilized with ice-cold (−20°C) methanol. The cells were stained intracellularly with rabbit anti-C. burnetii polyclonal antibodies for 1 h at room temperature and then stained with mouse anti-LC3 or anti-Beclin (autophagy sampler kit; BioLegend) and counterstained with goat anti-rabbit IgG (H+L) cross-adsorbed secondary antibody, PE, and goat anti-mouse IgG (H+L) cross-adsorbed secondary antibody, Alexa Fluor 488, for 1 h at 37°C. The nuclei were stained with 4′,6′-diamidino-2-phenylindole (DAPI) for 5 min at 4°C. Finally, the slide was mounted in ProLong Gold antifade reagent (Invitrogen, Carlsbad, CA) and allowed to cure overnight at room temperature. Microscopic analysis was performed using a Zeiss 710 NLO 2P system, and images were analyzed using Zen software.

Immunoblotting.

Western blotting was performed to quantify the autophagy-related proteins in NMI- and NMII-infected BMNs. After 24 and 48 h of infection, BMNs were washed and protein was isolated using radioimmunoprecipitation assay (RIPA) buffer, and total protein was quantified using a Pierce Coomassie Plus (Bradford) assay kit. Equal amounts of protein (50 µg/sample) were separated using 4 to 20% polyacrylamide gels and then transferred to a polyvinylidene difluoride (PVDF) membrane using an iBlot 2 instrument (Thermofisher). The membranes were blocked with 5% nonfat dry milk in Tris buffer (pH 7.5) containing 200 mM Tris, 1.38 M NaCl, and 0.1% Tween 20 for 2 h and then incubated with a primary antibody specific for Beclin, ATG5, Rab7, LC3, or anti-β-actin (autophagy sampler kit; BioLegend) at 4°C overnight in 5% bovine serum albumin and treated with horseradish peroxidase (HRP)–goat anti-mouse IgG secondary antibody (BioLegend) for 1 h at room temperature. The bands were visualized using an enhanced chemiluminescence Western blot detection kit (Thermo Fisher Scientific), and chemiluminescent protein bands were captured and the images were processed on an iBright 1500 instrument. The protein intensity was measured by ImageJ software, and the ratios of β-actin to the target proteins are represented as relative intensities. Each experiment was repeated at least three times.

Staining of phosphatidylserine-positive neutrophils.

An annexin-V-FLUOS staining kit was purchased from Roche Diagnostics GmbH (Penzberg, Germany). Briefly, 1 × 10⁶ purified BMNs were cultured in a sterile 6-well plate with NMI or NMII at an MOI of 10 or 100 for 24 h in a humidified chamber at 37°C under 5% CO₂. Uninfected BMNs were maintained as a control. Cells were centrifuged and washed three times with PBS. Cell pellet was then resuspended in 100 µl of annexin-V-FLUOS labeling solution. After incubation at room temperature for 30 min, samples were analyzed on a MoFlo XDP cell sorter (Beckman Coulter). Further, to confirm the phosphatidylserine (PS) exposure in BMNs infected with NMI, NMII, and the dotA:Tn mutant at 24, 48, and 72 hpi, live-cell imaging was performed using a pSIVA kit (Bio-Rad), which specifically binds with the exposed phosphatidylserine. Propidium iodide was used to visualize the membrane-compromised BMNs, and the nuclear stain DAPI was used to visualize the nucleus in BMNs. All the reagents were added according to the manufacturer’s instructions. Images were acquired using an EVOS M7000 microscope and analyzed and quantified using Celleste software. Every analysis included at least 200 cells from three individual wells, and the percentages of pSIVA- and PI-positive cells are represented as the pSIVA or PI/DAPI ratio.

Measurement of intracellular GSH/GSSG.

Changes in GSH and GSSG concentrations in NMI- or NMII-infected neutrophils were measured using the colorimetric method. Briefly, 1 × 10⁶ purified BMNs infected with NMI or NMII at an MOI of 100 were cultured in a sterile 6-well plate. Uninfected BMNs were maintained as a control. The cells were pelleted and washed twice with ice-cold PBS, cell lysates were prepared in ice-cold 5% 5-sulfo-salicylic acid dehydrate (SSA), and the quantity of GSH and GSSG was measured using a glutathione colorimetric detection kit (Invitrogen) according to the manufacturer’s protocol. The absorbance was measured at 405 nm using a Molecular Devices SpectraMax Plus plate reader and SoftMax software or an Infinite F50 (Tecan, Switzerland) microplate reader, and the fold change in GSH and GSSG concentration was determined by comparing the NMI or NMII results with the control. All the experiments were performed in triplicate.

Determination of intracellular IL17AF by flow cytometry.

Intracellular IL-17 expression was determined by intracellular staining of BMNs infected with either NMI or NMII using IL17AF antibody targeting mouse IL-17A/IL-17F heterodimer. For this assay, 1 × 10⁶ purified BMNs infected with NMI or NMII at an MOI of 100 were cultured in a sterile 6-well plate. Uninfected BMNs were maintained as a control. The cells were pelleted and washed three times with PBS, and 100 μL of Fix/Perm medium A was added to each sample, pulse vortexed, and incubated at room temperature for 15 min in the dark. Cells were washed twice with 2 mL of PBS. Then, anti-mouse IL17AF-PE (Invitrogen) antibody at a concentration of 0.2 mg/mL was added to cells in 100 μL of permeabilization medium B and incubated for 20 min at room temperature with protection from light. Cells were washed twice with 2 mL of PBS and fixed in 4% PFA prior to flow acquisition.

Statistical analysis.

Statistical analysis was performed by Welch’s unpaired t test using GraphPad Prism 9.00 software. For all analysis, a P value of <0.05 was deemed significant.