Abstract
Orientia tsutsugamushi infection can cause acute lung injury and high mortality in humans; however, the underlying mechanisms are unclear. Here, we tested a hypothesis that dysregulated pulmonary inflammation and Tie2-mediated endothelial malfunction contribute to lung damage. Using a murine model of lethal O. tsutsugamushi infection, we demonstrated pathological characteristics of vascular activation and tissue damage: 1) a significant increase of ICAM-1, VEGFR2, and angiopoietin-2 (Ang2) proteins in inflamed tissues and lung-derived endothelial cells (EC), 2) a progressive loss of endothelial quiescent and junction proteins (Ang1, VE-cadherin/CD144, occuludin), and 3) a profound impairment of Tie2 receptor at the transcriptional and functional levels. In vitro infection of primary human EC cultures and serum Ang2 proteins in scrub typhus patients support our animal studies, implying endothelial dysfunction in severe scrub typhus. Flow cytometric analyses of lung-recovered cells further revealed that pulmonary macrophages (MΦ) were polarized toward an M1-like phenotype (CD80+CD64+CD11b+Ly6G-) during the onset of disease and prior to host death, which correlated with the significant loss of CD31+CD45- ECs and M2-like (CD206+CD64+CD11b+Ly6G-) cells. In vitro studies indicated extensive bacterial replication in M2-type, but not M1-type, MΦs, implying the protective and pathogenic roles of M1-skewed responses. This is the first detailed investigation of lung cellular immune responses during acute O. tsutsugamushi infection. It uncovers specific biomarkers for vascular dysfunction and M1-skewed inflammatory responses, highlighting future therapeutic research for the control of this neglected tropical disease.
Author Summary Scrub typhus is a life-threatening disease, infecting an estimated one million people yearly. Acute lung injury is the most common clinical observation; however, its pathogenic biomarkers and mechanisms of progression remain unknown. Here, we used a lethal infection mouse model that parallels certain aspects of severe scrub typhus, primary human endothelial cell cultures, and patient sera to define pathogenic biomarkers following Orientia tsutsugamushi infections. We found a significant increase in the levels of endothelial activation/stress markers (angiopoietins and ICAM-1) in infected mouse lungs and in patient sera, but a progressive loss of endothelium-specific Tie2 receptor and junction proteins (VE-cadherin), at severe stages of disease. These signs of vasculature disruption positively correlated with the timing and magnitude of recruitment/activation of proinflammatory MΦ subsets in infected lungs. Bacterial growth in vitro was favored in M2-like, but not in M1-like, MΦs. This study, for the first time, reveals endothelial malfunction and dysregulated inflammatory responses, suggesting potential therapeutic targets to ameliorate tissue damage and pathogenesis.
Introduction
Scrub typhus is a febrile and potentially lethal illness that infects an estimated one million individuals per year [1]. The disease is caused by infection with the bacterium, Orientia tsutsugamushi. Nearly a third of the human population lives in endemic areas, known as the “tsutsugamushi triangle”, although recent reports have identified scrub typhus in South America, which was previously believed to be free of scrub typhus [1, 2]. Within endemic areas, scrub typhus is reported to cause a substantial proportion (approximately 15-23%) of reported febrile illness [3, 4]. If left untreated, scrub typhus can manifest as interstitial pneumonia, myocardial and hepatic inflammation, and meningoencephalitis. [5]. Mild interstitial pneumonitis is typically the extent of pulmonary involvement during self-resolving or promptly treated scrub typhus. However, life-threatening pathologies can arise in severe cases, including lung hemorrhage, edema buildup, diffuse alveolar damage, and interstitial cellular infiltration [6]. Acute respiratory distress syndrome and lung damage are associated with high mortality and present in 6.75-25% of scrub typhus patients [5, 7]; however, there is no detailed investigation of the underlying mechanisms responsible for pulmonary endothelial dysfunction and inflammation.
Being an obligate intracellular bacterium, O. tsutsugamushi can infect a variety of host cells but primarily replicate in macrophages (MΦs)[8], dendritic cells [9], and endothelial cells (ECs) [6, 8]. The bacteria enter host cells via the phagosome [10] or endosome [11], which they subsequently escape to begin replication in the cytoplasm. Infection-triggered cellular responses, including the activation of activator protein-1 (AP-1) and NF-κB pathways, the production of proinflammatory cytokines (IL-1β, TNF-α, IL-8/CXCL8), and the expression of distinct gene profiles, have been examined in vitro by using primary human umbilical vein endothelial cells (HUVEC) [12], human epithelial/EC-like ECV304 cell line [13], human monocytes or MΦs [14]. Prolonged infection can result in EC death via apoptosis [15], but there is limited information on endothelial responses during the course of O. tsutsugamushi infection [16]. Sublethal O. tsutsugamushi infection studies in outbred Swiss CD-1 mice [17], as well as clinical studies of human patients [18], have shown significant elevation of endothelial activation markers (ICAM-1, VCAM-1, E-Selectin, etc.) in the serum of infected individuals; however, endothelium-focused analyses during in vivo infection remain largely unexplored. Several studies have characterized the response and activation of MΦs during O. tsutsugamushi infection [19, 20]. In vitro and ex vivo experiments have shown that human monocytes/MΦs in the circulation adopt an inflammatory “M1” type transcriptional profile after O. tsutsugamushi infection, although little is known regarding tissue specific macrophages or the presence of alternatively activated “M2”-type macrophages [21].
The delicate balance of EC quiescence and activation is crucial during systemic infection. While EC activation promotes adherence and recruitment of innate and adaptive immune cells for pathogen clearance, prolonged activation can lead to EC cytotoxicity, impaired barrier function, and host mortality [22]. One of the critical mechanisms to control EC activation status and cellular function is through competitive interactions between angiopoietin-1 (Ang1) and Ang2 ligands with their receptor, Tie2, a protein tyrosine kinase that is predominately expressed on ECs in humans and mice. Tie2 activation and phosphorylation via binding with Ang1 (produced by pericytes and platelets [23]) promote EC quiescence, which limits leukocyte adhesion and maintains EC survival and vascular barrier integrity [24]. Infection- or inflammation-triggered release of Ang2 (normally stored within the Weibel-Palade bodies in ECs [25, 26]) can compete with Ang1 binding Tie2 to antagonize its function [27]. Inhibition of Tie2 signaling via Ang2 binding stimulates leukocyte adhesion, vascular barrier destabilization, and inflammation [28, 29]. Thus, Ang2/Ang1 expression ratios and Tie2 activation status are important biomarkers for the pathogenesis of systemic infection, such as severe sepsis and malaria [30, 31].
To investigate endothelial alterations during severe O. tsutsugamushi infection, we have recently developed a lethal intravenous O. tsutsugamushi infection model in C57BL/6 mice [32-34], which parallels aspects of severe scrub typhus in humans. In our lethal models, bacterial loads in both the spleen and liver reached peak levels around or shortly after the onset of disease (days 6-8 post-infection), and are reduced significantly by days 10-12. In contrast, lung bacterial loads remain elevated throughout infection [35]. All mice expire by days 12-13, suggesting unknown mechanisms of pathogenesis are at work during late infection [32, 35]. Given that the lungs are a major organ for O. tsutsugamushi infection in humans and in different animal models, and that elevated ratios of ANG2/ANG1 transcripts are pathological hallmarks in lethal infection models [32, 35], we hypothesize that dysregulated pulmonary inflammation and Tie2/Ang2-mediated endothelial dysfunction contribute to disease pathogenesis at late stages of O. tsutsugamushi infection.
In this study, we utilized a lethal infection model in C57BL/6 mice, in vitro infection systems (human EC cultures, and bone marrow-derived MΦ subsets), as well as sera from scrub typhus patients to reveal positive correlations between vascular dysfunction, activation of innate immune cells, and disease progression. We focused on two cellular subsets known to be sites of O. tsutsugamushi replication, ECs and MΦs, to characterize their activation and polarization in vivo. To the best of our knowledge, this is the first report to delineate MΦ subsets in inflamed lungs and their positive correlation with Tie2 malfunction during late stages of severe infection.
Results
Pulmonary EC activation and tight junction disruption during infection in mice
Given that the lung is a major site of O. tsutsugamushi infection in humans and animal models [6, 32] and that EC activation and disruption of vascular barrier integrity are principal steps for acute lung injury in sepsis and pneumonia models [36], we sought to investigate pulmonary EC activation in C57BL/6 mice following i.v. inoculation with a lethal dose of O. tsutsugamushi Karp (∼1.325×106 viable bacteria in 200 µl of PBS). Inoculation via this route establishes bacterial replication in the lungs accompanied by interstitial pneumonitis and alveolar thickening (S1A Fig, [35]). Immunofluorescent staining of frozen lung sections revealed increased ICAM-1-positive (green) vascular staining on days 2 (D2) to 9, as well as a close association of bacterial (red) with activated endothelium at D9 (Fig 1A, boxed area, Fig 1C). To examine endothelial structure and adherens junctions, we co-stained lung sections with GSL I-B4 lectin (specific for α-galactose residues known to be enriched on the surface of EC) and anti-VE-cadherin/CD144 (an adherens junction protein, red), as described in our previous report for neuroinflammation [34]. The VE-cadherin staining was intense and homogenous in the control tissues, but markedly reduced in D2 samples; VE-cadherin staining was nearly absent in some foci of D6 and D9 samples (Fig 1B-C), implying the reduction of junction proteins. In conjunction with the endothelial junction proteins, we co-stained the epithelial junction protein, occludin, and GSL I-B4 lectin. Consistently, we found a progressive loss of occludin staining on the bronchial epithelium during the infection and a near absence of staining in D9 samples (S1B Fig). These data suggest a progressive and severe loss of vascular barrier integrity in the infected lungs, especially at late stages of acute infection prior to host death (D9) [32].
To support our immunofluorescent results, we prepared single-cell suspensions from mouse lung tissues and used flow cytometry to examine the frequency of recovered CD31+CD45- ECs and their surface ICAM-1 expression. We found that compared with the mock controls, infected lung tissues contained a significant increase in the frequencies of ICAM-1+CD31+CD45- ECs at D6 and D9, respectively (p < 0.001, p < 0.01, Fig 1D), while there were approximately 5-folds reduction in the frequencies of total ECs at D6 and D9 (p < 0.001, S1C Fig). We also examined EC expression of vascular endothelial growth factor receptor 2 (VEGFR2), a critical factor controlling vascular permeability and barrier function [37]. Compared with the mock controls, there was an approximately 6-fold increase in the frequencies of VEGFR2+CD31+CD45- ECs in the lung tissues at D6 and D9 (Fig 1D). These flow cytometry data reinforced the immunofluorescent results, indicating marked endothelial activation and damage at D6 (the onset of disease) and D9 (prior to host demise). To validate these findings in mice, we infected primary HUVEC cultures with different doses of O. tsutsugamushi (3 and 10 MOI) and found a dose-dependent increase in ICAM1 and IL8/CXCL8 transcripts at 24 h post-infection (S1D Fig). Collectively, these data indicate infection-triggered endothelial stress and activation, accompanied by progressive vascular damage and tight junction disruption during the course of infection.
Alterations in the angiopoietin-Tie2 system during Orientia infection
Currently, there are no detailed in vivo studies to define molecular mechanisms underlying O. tsutsugamushi infection-associated vascular damage. For other severe and systemic infections caused by bacteria or viruses, alterations in angiopoietin proteins or their functional Tie2 receptor is considered as one of the key mechanisms for vascular dysfunction [38, 39]. Given our previous findings of elevated ANG2/ANG1 mRNA ratios in mice with severe scrub typhus and in O. tsutsugamushi-infected HUVECs [33], we speculated that impairment in Tie2 function occurs in severe scrub typhus [16]. To test this hypothesis, we examined Ang1, Ang2, and Tie2 protein levels in the lung tissues via immunofluorescent staining (Figs 2-3). While the mock controls contained relatively high levels of Ang1 (green), with relatively low levels of Ang2 (red), we found a modest decrease in Ang1-positive staining but a steady increase in Ang2 staining, during the course of infection (Fig 2A). These IFA results were quantified, demonstrating the significant decrease of Ang1 and a significant increase of Ang2 expression during O. tsutsugamushi infection that correlated with disease progression (Fig 2B). To validate our findings mouse models, we measured scrub typhus patient sera via specific ELISA assays and found a statistically significant increase in circulating Ang2 levels, which correlated with their O. tsutsugamushi-specific antibody titers (p < 0.05 and p < 0.01, comparing IFA titers of 1:640 and 1:1280 with the control subjects, Fig 2C). These human data support our findings obtained from mouse tissues, indicating the potential utility of serum Ang2 levels as a molecular biomarker of scrub typhus severity.
IFA staining of the Ang1/2 receptor, Tie2, revealed clear Tie2 staining in mock infected controls, however, positive Tie2 staining was nearly absent in some foci of D6 and D9 samples (Fig 3A). To validate these findings, we used Western blot analyses of lung tissues and confirmed a striking reduction of phosphorylated Tie2 (pTie2) and total Tie2 levels at both D6 and D9, as compared with either the mock and D2 samples (Fig 3B), implying impairments at the translational and functional levels. The qRT-PCR analyses further validated a statistically significant decrease in TIE2 mRNA levels in the lungs at D6 and D10 (p < 0.01, compared with the mock controls, Fig 3C), implying impairment at the transcriptional level. These data, together with our previous studies [32, 33], indicate that marked Ang2 production, accompanied with severe impairment in the Tie2 functions, are pathogenic mechanisms of severe vascular damage in O. tsutsugamushi infection.
M1-like responses in the lungs of infected mice
Having documented progressive endothelial damage and alterations in endothelium-specific biomarkers following O. tsutsugamushi infection (Figs 1 and 2), we then examined the timing and magnitude of leukocyte recruitment and activation. Although some reports described leukocyte involvement in O. tsutsugamushi-infected mouse spleen and brain [34, 40], there are no detailed studies of innate immune responses in infected lungs. Using immunofluorescent staining, we found that CD45+ leukocytes and CD3+ T cells were accumulated around Ang2-positive foci in the lungs at D6 and D10, and that CD45-Ang2 or CD3-Ang2 co-stained foci were readily detectable at D10 (yellow, S2A Fig). Flow cytometric analyses revealed a 20-fold increase in total numbers of CD4+ T cells at D10, but a statistically significant decrease in percentages of these cells at D6 and D10, respectively (p < 0.0001, compared with mock controls, S2B Fig). In contrast, there was a 50-fold increase in total numbers and 2.3-fold increase in percentages of CD8+ T cells at D10 (S2C Fig). These findings were consistent with the known importance of CD8+ T cells during O. tsutsugamushi infection in mice [40, 41].
Monocytes and MΦs are particularly noteworthy leukocytes during O. tsutsugamushi infection, as they can act as a target for bacterial replication and a propagator of the inflammatory response [8, 14], possibly playing a role in O. tsutsugamushi dissemination from skin lesions [9, 42]. While in vitro infection predominantly drives human monocytes/MΦs to M1-like transcription programs [9, 42], our current knowledge on O. tsutsugamushi-MΦ interactions in the lungs is still limited. Using IFA staining, we observed co-localization of bacteria (green) with IBA-1+ MΦs (red) in mouse lungs (S3A Fig). To define monocyte/MΦ responses, we applied reported protocols and gating strategies for flow cytometric analysis of mouse lung monocyte/MΦ subsets [43] (Fig 4A). Compared with the mock controls, D6 and D9 samples had 4-to 5-fold increases in the frequency of CD64+CD11b+Ly6G- alveolar/interstitial monocytes/MΦs, as well as 6- and 14-fold increases in total cell numbers, respectively (p < 0.01 and p < 0.001, Fig 4B). Of note, nearly all (∼97%) pulmonary MΦs displayed an M1-like phenotype (CD80+CD64+CD11b+Ly6G-, Fig 4C) at D9. In contrast, while the mock and D2 lung samples contained ∼3.2% of M2-like (CD206+CD64+CD11b+Ly6G-) cells, these cells were barely detectable at D6 or D9 (Fig 4D). These data suggest extensive recruitment and/or activation of M1-like cells, but marked loss and/or suppression of M2-like cells, during the progression of disease. Likewise, lung qRT-PCR assays confirmed a significant up-regulation of M1 markers (IFNγ, FPR2, CD38, and NOS2), but not M2 markers (CD206, EGR2), at D9 (Fig 5A and B). While the IL10 up-expression was previously reported by our lab and other groups [44], we also detected an increased expression of ARGINASE1 (Fig 5B), a marker known for M2 polarization and the growth of other intracellular pathogens [45]. Together with data shown in Fig 4, we concluded that at the onset of disease and beyond, O. tsutsugamushi infection preferentially stimulated pro-inflammatory innate responses in M1-like monocytes/MΦs, which correlate with the onset of vascular damage (Figs 1-3).
MΦ polarization in favor of Orientia replication in vitro
Because we had demonstrated differential monocyte/MΦ responses in vivo, we asked how MΦ polarization might influence intracellular growth of the bacteria. We generated bone marrow-derived MΦs from naïve C57BL/6 mice, polarized cells via pretreatment with LPS (100 ng/ml) or recombinant IL-4 (rIL-4, 10 ng/ml) for 24 h, infected cells with O. tsutsugamushi (MOI 5), and measured bacterial loads at different time points. Flow cytometry and gene profile analyses of primed but uninfected cells confirmed their corresponding polarization to either classically activated M1 or alternatively activated M2 phenotypes (S3B and C Fig), as documented by others [46]. At 48 h post-infection, IL-4-primed M2 cells contained significantly increased loads of bacteria (judged by the copy number of Orientia 47-kDa gene) than LPS-primed M1 cells (p < 0.01, Fig 6A). At 72 h post-infection, M2 cells contained 10-fold more bacteria than M1 cells (Fig 6A), with extensive accumulation of bacteria (green) within IBA-1-positive MΦs (red, Fig 6B).
Discussion
Despite being an important emerging infectious disease, detailed immunological studies of scrub typhus patient samples or O. tsutsugamushi-infected animal tissues during the course of disease are scarce. In this study, we focused on ECs and monocytes/MΦs in the lungs of lethally infected mice to examine the activation of these cellular subsets known to be important cellular targets of O. tsutsugamushi in vivo. Our findings revealed important parameters and cell-specific alterations associated with acute lung injury and pathogenesis. The endothelium in infected lungs presented progressive Tie2 malfunction, increased Ang2 and ICAM-1 expression and pro-inflammatory MΦs at the onset of disease and severe stages of infection. Since lung damage and vascular malfunction are hallmarks of scrub typhus severity in patients [6], a better understanding of pathogenesis associated with acute lung injury is important for disease control and management.
The molecular characteristics of endothelial alterations during O. tsutsugamushi infection in vivo have not been explored. The present study provided evidence for the mechanisms underlying pulmonary injury and vascular dysfunction during O. tsutsugamushi infection. First, the timing of ICAM-1 and VEGFR2 expression on the surface of lung-derived CD31+CD45- ECs was concurrent with the appearance of signs of vascular injury and decrease in cell junction proteins (Fig 1 and S1 Fig). Since ICAM-1 promotes circulating immune cells to bind to the endothelium and extravasate into inflamed tissues [47], increased ICAM-1 surface expression likely contributed to immune cell influx into the infected mouse lungs (Fig 5, S2 Fig). Our observed ICAM1 and IL-8/CXLC8 up-regulation in infected human EC cultures (S1C Fig) were consistent with other reported studies of scrub typhus patients [48]. VEGFR2 is known to be the primary receptor for vascular endothelial growth factor (VEGF) in the endothelium, and the VEGF/VEGFR2 axis regulates microvascular permeability via interacting with VE-cadherin and tight junction proteins [49, 50]. In our hands, we consistently detected a marked increase in VEGFR2 on the surface of infected ECs, but a significant reduction of VE-cadherin (adhesion junctions) in the lungs at D6 and D9. While it is unclear whether increased VEGFR2 expression is directly linked to diminished VE-cadherin expression in infected lungs, our data suggest compromised endothelial barrier integrity during severe O. tsutsugamushi infection in the lungs. Second, a notable reduction of Tie2 proteins was concurrent with significant Ang2 production and/or release at the severe stages of infection (Figs 2-3). To date, there are no reports for Tie2 expression levels in scrub typhus patients or animal models, although our group previously showed increased ANG2/ANG1 transcript ratios in O. tsutsugamushi-infected human EC cultures and mouse tissues [32]. Our findings of significant and progressive reduction in Tie2 protein and its mRNA levels, as well as functional pTie2 level, in the lungs of lethally infection mice are important from basic research and clinical points of view. It is known that the Ang1/Tie2 axis is essential for vascular remodeling and endothelial cell stabilization, as either knockout Ang1 or Tie2 in mice is embryonic lethal [51, 52]. Given the critical function of Tie2 receptor in vascular physiology and integrity, it will be important to further examine whether our observed reduction in Ang1 and Tie2 is due to direct endothelial damage or signaling from nearby pericytes and recruited immune cells. Research in these areas would be of great value, as angiopoietin- or Tie2-targeted therapies have been evaluated as alternative treatment strategies for severe sepsis [53, 54], severe dengue [55], and in cerebral malaria [38] infection models to restore endothelial quiescence during infection. Our clinical observation that increased serum Ang2 in human scrub typhus patients correlates with O. tsutsugamushi-specific antibody titers demonstrates the utility of Ang2 as a pathogenic biomarker, and highlights the potential of use Ang2- or Tie2-targeted therapies for severe scrub typhus, as in patients with severe sepsis [29, 30] and malaria [56].
Monocytes/MΦs play important roles in infection with O. tsutsugamushi and other closely related Rickettsia species [19, 57]. Previous findings [19] are consistent with our observation of an increased accumulation of CD11b+Ly6G- MΦs/monocytes and the close association of IBA-l+ phagocytes with Orientia in the lungs at D6 and D9. Yet, our findings of selective recruitment/expansion of M1-skewed MΦs/monocytes, and the suppression of M2 cell activation, in the infected mouse lungs were particularly novel and important. This is the first in vivo evidence for M1-polarized MΦ responses in O. tsutsugamushi-infected mouse lungs, which was consistent with previous in vitro and ex vivo studies for M1-skewed gene transcription programs in O. tsutsugamushi-infected human monocytes and MΦs [21, 42]. At present, it is still unclear whether the predominately M1 MΦ population we observed contributes to the killing of O. tsutsugamushi and the damage of vascular tissues, as seen for protective versus pathogenic roles in other lung infection models [58]. Given the reports that anti-Ang2 antibody treatment during pulmonary bacterial infection can decrease MΦ recruitment and inflammation [26], and that signaling via Tie2 on monocytes/MΦs can promote a proinflammatory profile [59], it will be interesting to determine how Ang2/Ang1 signaling on both endothelial cells and MΦs promote cellular recruitment and activation during O. tsutsugamushi infection.
Conditions that promote the killing or growth of O. tsutsugamushi remain unclear, in part due to difficulties in bacterial cultivation, genetic modification, or visualization for studying the host-bacterium interactions [60]. Currently, there are no available reports or data for the phenotype of Orientia-carrying MΦ subsets or the roles of arginase-1 in vivo. Our in vitro comparative studies revealed limited bacterial replication in LPS-primed M1-like MΦs (Fig 6), supporting a previous report for the role of NOS2-mediated mechanisms in the control of O. tsutsugamushi Karp strain [19]. Yet, our in vitro findings were contradictory to another reported study, in which NO-enhanced the growth of O. tsutsugamushi Ikeda bacteria was observed in LPS-activated RAW 264.7 murine macrophages at days 6 to 8 post-infection [20]. While the use of different bacterial strains, MΦ sources, and examination times may account for these discrepancies, our study calls for careful examination of the intracellular niche for the replication of these obligate intracellular bacteria under different or physiological conditions. For bone marrow-derived MΦs, we found comparable bacterial loads in LPS- and IL-4-primed cells at 3 h, implying similar attachment and invasion of the bacteria under these two treatments. However, IL-4-primed M2 cells contained 10-fold more O. tsutsugamushi than LPS-primed M1 cells at 72 h. At present, our data did not exclude the possibility of bacterial growth in LPS-primed M1 cells after prolonged time in vitro, or in M1-vs. M2-like MΦs in mouse lungs or other organs. Given the recent report that specific MΦ responses, such as miR-155 and IL-10 production, correlate with prevention of cytokine storm in severe O. tsutsugamushi infection [61], it will be important to further examine whether the strong type-1 inflammatory responses in vivo [32] and M1-skewed responses (Fig 5) are responsible for marked decrease in M2 MΦs in the lungs. The use of transgenic mouse strains for tracking MΦ subsets would also help reveal whether O. tsutsugamushi bacteria preferentially replicate within M2 MΦs in vitro or contribute to the impairment of type 2 immune responses.
While the mouse model used here allows us to examine host-bacterium interactions and immune alterations, this model has some intrinsic limitations, as it bypasses bacterial dissemination from the local/skin sites. Nevertheless, this model has several advantages over self-limited infection models following the subcutaneous or intradermal inoculation of the bacteria via needles into inbred strains of mice [19, 62], for the examination of innate immune responses in visceral organs. Compared to a lethal infection initiated by feeding O. tsutsugamushi-infected mites on outbred mice, a technically challenging model with high variability [63], our model provides more consistent results and lethality, permitting the analysis of a given host molecule using gene-targeted knockouts on the C57BL/6 background. More importantly, our lethal model mimics certain pathological aspects of severe scrub typhus observed in humans, uncovering tissue-specific immune alterations that have never been described previously. For example, our findings of elevated Ang2 proteins in O. tsutsugamushi-infected lungs and increased ANG2 expression in multiple organs [32] are consistent with clinical studies of scrub typhus patients (Fig 2), which supports the potential for monitoring serum Ang2 levels as an indicator of disease severity and treatment outcome.
In summary, this study has revealed new insights into immune dysregulation and pathogenesis of severe scrub typhus. Through comprehensive analyses of O. tsutsugamushi-infected mouse lung tissues, we have provided the first evidence for the production of endothelial destabilizing factors and in vivo polarization of lung recruited MΦs. Our findings of polarized M1-like responses in the lungs at late stages of disease argue for immune-based restriction of bacterial replication as well as immunopathogenesis. While the molecular mechanisms underlying host-bacterium interaction and immune dysregulation remains unclear at this stage, it is conceivable that serum and tissue Ang2 levels would be a molecular biomarker for severe scrub typhus and a potential therapeutic target for treatment. A better understanding of infection-versus immune-mediated dysregulation will help design treatment strategies for severe scrub typhus cases.
Materials and Methods
Mouse infection and ethics statement
Female C57BL/6 mice were purchased from Envigo (Huntingdon, United Kingdom), maintained under specific pathogen-free conditions and used at 6-9 weeks of age, following protocols approved by the Institutional Animal Care and Use Committee (protocols # 9007082B and 1302003) at the University of Texas Medical Branch (UTMB) in Galveston, TX. All mouse infection studies were performed in the ABSL3 facility in the Galveston National Laboratory located at UTMB; all tissue processing and analysis procedures were performed in the BSL3 or BSL2 facilities. All procedures were approved by the Institutional Biosafety Committee, in accordance with Guidelines for Biosafety in Microbiological and Biomedical Laboratories. UTMB operates to comply with the USDA Animal Welfare Act (Public Law 89-544), the Health Research Extension Act of 1985 (Public Law 99-158), the Public Health Service Policy on Humane Care and Use of Laboratory Animals, and the NAS Guide for the Care and Use of Laboratory Animals (ISBN-13). UTMB is a registered Research Facility under the Animal Welfare Act, and has a current assurance on file with the Office of Laboratory Animal Welfare, in compliance with NIH Policy.
O. tsutsugamushi Karp strain was used herein; all infection studies were performed with the same bacterial stocks prepared from Vero cell infection, for which infectious organisms were quantified via a qPCR viability assay [35, 64]. Mice were inoculated intravenously (i.v.) with ∼1.325×106 viable bacteria (a lethal dose, 200 µl) or PBS and monitored daily for weight loss and signs of disease. In most cases, tissue samples (4-5 mice/group) were collected at 2, 6, and 9 (or 10) days post-infection and inactivated for immediate or subsequent analyses.
Ethical approval for human samples used in this work was granted by the Institutional Review Board of both Seoul National University Hospital (IRB NO 1603-136-751) and Chungnam National University Hospital (IRB NO 2014-12-006). All patients and healthy volunteers provided written informed consent prior to sample collection.
Immunofluorescence microscopy and quantification
Mouse lung tissues were processed for immunofluorescent staining, as in our previous report [34]. Briefly, 6-μm frozen sections were blocked and incubated with the following rat or rabbit anti-mouse antibodies (1:200, purchased from Abcam, Cambridge, MA, USA, unless specified): anti-ICAM1, anti-Ang1, anti-Ang2 (R&D Systems/Biotechne, McKinley Place NE, Minnesota), anti-Tie2, anti-VE-cadherin (adherence junctions), anti-occludin (epithelial tight junctions), anti-IBA-1 (ionised calcium binding adapter molecule-1, a MΦ marker), anti-CD45 (BD Bioscience, San Jose, CA, USA), or anti-Ang2 (R&D Systems). Staining endothelium in sections was done with FITC-conjugated Griffonia Simplicifolia lectin I (1:100, GSL I-B4 lectin, Vector Lab, Burlingame, CA, USA). Bacteria was stained in various sections using rabbit anti-O. tsutsugamushi Karp serum (1:500) [35]. Bound antibodies were visualized by using Alexa Fluor 488- or 555-conjugated, goat anti-rat or anti-rabbit IgG (H+L, 1:1,000-1:2,000, Life Technologies, Grand Island, NY, USA). All sections were stained with DAPI (1:5,000, Sigma-Aldrich, St. Louis, MO, USA). Infected sections stained with secondary Abs and DAPI only served as negative controls to optimize staining conditions. For each section, at least 6 low- and 6 high-magnification fields of the lung sections were imaged on a Carl Zeiss Axio Observer fluorescence microscope (Carl Zeiss Microscopy LLC, Thornwood, NY, USA) equipped with ApoTome and Zen imaging software. Acquisition settings were identical among samples of different experimental groups. Representative images at each time point are presented. To measure fluorescent intensity of given markers, gray levels across the entire tissue image (low-magnification, 4 independent images per group) were measured via ImageJ under the same parameter setting. Data are presented at mean + SEM of the group.
For in vitro studies, cells were seeded onto coverslips in 24-well plates (Falcon Corning, Corning, NY, USA). At indicated times of infection, slides were washed, fixed with 4% PFA for 20 min, and permeabalized with Triton X-100 for 15 min. After blocking with 10% BSA/3% goat serum for 1 h, cells were incubated with serum collected from Orientia-infected mice (1:1,500) or rabbit anti-IBA-1 (1:250, Abcam) (1:50) at 4°C overnight and then with a secondary Ab: goat anti-mouse Alexa Fluor 488 (Thermo Fisher Scientific, Waltham, MA, USA) or donkey anti-rabbit Alexa Fluor 594 (Invitrogen/Thermo Fisher Scientific) and DAPI (1:1000, Thermo Fisher Scientific). The cover slips were mounted on slides by using an Antifade Mountant solution (Invitrogen/Thermo Fisher Scientific). Images were taken using an Olympus IX51 microscope (Olympus Corporation, Tokyo, Japan).
Flow cytometry
Equivalent portions of lung tissues were harvested from infected and control mice, minced, and digested with 0.05% collagenase type IV (Gibco/Thermo Fisher Scientific) in Dulbecco’s Modified Eagle’s Medium (DMEM, Sigma-Aldrich, St. Louis, MO) for 30 mins at 37°C. Minced tissues were loaded into Medicons and homogenized using a BD Mediamachine System (BD Biosciences, Franklin Lakes, NJ). Lung single-cell suspensions were made by passing lung homogenates through 70-µm cell strainers. Spleen homogenates were made by passing tissue through a 70-µm strainer. Lymphocytes were enriched by using Lympholyte-M Cell Separation Media (Burlington, NC). Red blood cells were removed by using Red Cell Lysis Buffer (Sigma-Aldrich). Leukocytes were stained with the Fixable Viability Dye (eFluor 506) (eBioscience/Thermo Fisher Scientific, Walthalm, MA) for live/dead cell staining, blocked with FcγR blocker, and stained with fluorochrome-labeled antibodies (Abs). The following Abs purchased from Thermo Fisher Scientific and BioLegend (San Diego CA): PE-Cy7-anti-CD3ε (145-2C11), Pacific Blue-anti-CD4 (GK1.5), APC-Cy7-anti-CD8a (53-6.7), APC-anti-Ly6G (1A8-Ly6G), APC-anti-CD31 (390), PE-anti-VEGFR2 (Avas12a1), FITC-anti-ICAM-1 (YN1/1.7.4), Pacific Blue-anti-CD45 (30-F11), PE-anti-CD80 (16-10A1), BV421-anti-CD206 (CO68C2), FITC-anti-CD64 (X54-5/7.1), PerCP-Cy5.5-anti-CD11b (M1/70). Cells were fixed in 2% paraformaldehyde overnight at 4°C before cell analysis. Data were collected by a BD LSRFortessa (Becton Dickinson, San Jose, CA) and analyzed using FlowJo software version 8.86 (Tree Star, Ashland, OR). As previously reported for mouse lung tissues [65] CD45+CD31- and CD45-CD31+ cells were considered hematopoietic cells and endothelial cells by flow cytometry, respectively.
Western blot
Protein from lung tissues was extracted with a RIPA lysis buffer (Cell Signaling Technology, Danvers, MA) and quantified with BCA Protein Assay kit (Thermo Fisher Scientific). Protein samples (40 μg/lane) were loaded onto 4-20% SDS-PAGE gels (Bio-Rad Laboritories, Hercules, CA, USA) and transferred onto nitrocellulose membranes (Bio-Rad Laboratories). After blocking non-specific binding sites, membranes were respectively incubated with rabbit Abs specific to mouse Tie2 (1:500, Abcam), phospho-Tie2 (1:400, R&D System, USA), and β-actin (1:15000, Novus Biologicals, USA), and an anti-rabbit secondary antibody (SouthernBiotech, Birmingham, AL, USA). After treatment with the Maximum Sensitivity Substrate (Thermo Fisher Scientific) for 1 min, the light signals were captured by Luminescent Image Analyzer (ImageQuant LAS 4000, GE Healthcare Bio-Sciences AB, Sweden). Protein bands were quantified by using image analysis software (ImageJ). Three independent experiments were performed.
Infection of mouse bone marrow-derived MΦs
Bone marrow cells were collected from mouse femur and tibia and treated with a red cell lysis buffer (Sigma). For MΦ generation, bone marrow cells were grown in DMEM (Gibco) with 10% FBS, penicillin/streptomycin antibiotics, and 40 ng/ml M-CSF (BioLegend) and incubated at 37°C. Media was changed at day 3, and cells were collected at day 7 and seeded onto 6- or 24-well plates for overnight. MΦs were treated with either 100 ng/ml LPS (for M1 polarization) or 10 ng/ml mouse rIL-4 (for M2 polarization, Peprotech, Rocky Hill, NJ) for 24 h. Cells were then infected with O. tsutsugamushi (5 MOI) and centrifuged at 2,000 RPM for 5 min to synchronize infection of the cells.
Quantitative PCR and reverse transcriptase PCR (qPCR and qRT-PCR)
To determine bacterial loads, bone marrow-derived MΦs were collected at 3, 24, 48, and 72 hpi by using a DNeasy kit (Qiagen) and used for qPCR assays, as previously described [32]. Bacterial loads were normalized to total nanogram (ng) of DNA per µL for the same sample, and data are expressed as the gene copy number of 47-kDa protein per picogram (pg) of DNA. The copy number for the 47-kDa gene was determined by known concentrations of a control plasmid containing single-copy insert of the gene. Gene copy numbers were determined via serial dilution (10-fold) of the control plasmid.
To measure host gene expression, mouse tissues or in vitro-infected cells were respectively collected in RNALater (Ambion, Austin, TX) or Trizol solution at 4°C overnight to inactivate infectious bacteria and stored at -80°C for subsequent analyses. Total RNA was extracted by using RNeasy mini kits (Qiagen) and digested with RNase-free DNase (Qiagen); cDNA was synthesized with the iScript cDNA synthesis kit (Bio-Rad Laboratories, Hercules, CA). The quantitative RT-PCR (qRT-PCR) assays were performed with iTaq SYBR Green Supermix and a CFX96 Touch Real-Time PCR Detection System (Bio-Rad). PCR assays were denatured for 3 min at 95°C, followed by 40 cycles of 10s at 95°C and 30s at 60°C. Melt-curve analysis was also used to check the specificity of the amplification reaction. Relative abundance of transcripts was calculated by using the 2-ΔΔCT method and compared to housekeeping genes glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or β-actin. Primers used in these analyses are listed in Table S1.
Human umbilical vein endothelial cell (HUVEC) infection
HUVECs (Cell Application, San Diego, CA) were maintained in complete Prigrow I medium supplemented with 3% heat-inactivated FBS (Applied Biological Materials, Vancouver, Canada) in 5% CO2 at 37°C. All in vitro experiments were performed between cell passages 5 and 7, as described previously [33, 66]. For infection, HUVECs were cultivated in Prigrow I medium with 10% FBS and seeded onto 6-well plates (Corning Inc., Corning, NY). Confluent monolayers were infected with Orientia (3 and 10 MOI) for 24 h and compared with uninfected controls.
Human serum collection and measurement of Ang2 by ELISA
Human serum samples were collected from healthy volunteers (n = 8) and scrub typhus patients (n = 32) after obtaining informed consent at the Chungnam National University Hospital in Daejeon, South Korea. Scrub typhus diagnosis was confirmed based on clinical symptoms and a positive serology: a 4-fold or greater rise in the titer of paired plasma or single cut-off titer of an IgM antibody ≥ 1:160 by an indirect immunofluorescence antibody assay (IFA) against O. tsutsugamushi antigens or passive hemagglutination assay (PHA) during hospital admission. Healthy volunteers had never been previously diagnosed with scrub typhus, and their sera were negative when examined by IFA. Patient plasma samples were classified into four groups based on their IFA titers. Ang2 concentration was determined by using a commercial ELISA kit (Abcam), according to manufacturer’s instructions.
Human antibody titer measured by IFA
L929 cells infected with three strains of O. tsutsugamushi (Boryong, Karp, and Gilliam strains) were harvested, mixed in equal amounts, and used as antigens to measure total IgG titers against O. tsutsugamushi via IFA. Briefly, infected L929 cells were harvested, washed with PBS, seeded onto Teflon-coated spot slides, and fixed with cold acetone for 10 min. The slides were stored at -70°C until use. Two-fold serially diluted (1:40 to 1:1280 in PBS) patient sera was added to the antigen-coated spot on the slide and incubated for 30 min in a wet chamber at room temperature. An Alexa Fluor 488-conjugated goat anti-human IgG (diluted 1:1000 in PBS, Molecular Probes, Waltham, MA, USA) was used as the secondary antibody. The stained slides were examined under an Olympus FV1000 laser scanning confocal microscope (Olympus, Tokyo, Japan). The endpoint titer of IFA was defined as the highest titer showing a fluorescence signal above the background.
Statistical analysis
Data were presented as mean ± standard errors of the mean (SEM). Differences between individual treatment and control groups were determined by using Student’s t test, utilizing Welch’s correction when appropriate. One-way ANOVA was used for multiple group comparisons with a Tukey’s Post Hoc for comparisons between groups. Statistically significant values are referred to as *, p<0.05; **, p<0.01; ***, p<0.001; and ****, p<0.0001.
Supporting information
S1 Table. Real-time PCR primers of human, murine, and bacterial genes. The primer sequences used in this study (listed in the 5’ to 3’ direction).
Acknowledgements
The authors thank Dr. David Walker for generously providing BSL-3 laboratory equipment and space for conducting experiments, Nicole Mendell for continued training and technical help in many aspects of this study, Dr. Juan Olano and the UTMB Immunology Joint Lab Group for providing helpful suggestions, and Dr. Sherry Haller for manuscript editing assistance.