Chitinases and chitinase-like proteins are an evolutionary conserved group of proteins. In the absence of chitin synthesis in mammals, the conserved presence of chitinases suggests their roles in physiology and immunity, but experimental evidence to prove these roles is scarce. Chitotriosidase (chit1) is one of the two true chitinases present in mammals and the most prevalent chitinase in humans. In this study, we investigated the regulation and the role of chit1 in a mouse model of Klebsiella pneumoniae lung infection. We show that chitinase activity in bronchoalveolar lavage fluid is significantly reduced during K. pneumoniae lung infection. This reduced activity is inversely correlated with the number of neutrophils. Further, instilling neutrophil lysates in lungs decreased chitinase activity. We observed degradation of chit1 by neutrophil proteases. In a mouse model, chit1 deficiency provided a significant advantage to the host during K. pneumoniae lung infection by limiting bacterial dissemination. This phenotype was independent of inflammatory changes in chit1−/− mice as they exerted a similar inflammatory response. The decreased dissemination resulted in improved survival in chit1−/− mice infected with K. pneumoniae in the presence or absence of antibiotic therapy. The beneficial effects of chit1 deficiency were associated with altered Akt activation in the lungs. Chit1−/− mice induced a more robust Akt activation postinfection. The role of the Akt pathway in K. pneumoniae lung infection was confirmed by using an Akt inhibitor, which impaired health and survival. These data suggest a detrimental role of chit1 in K. pneumoniae lung infections.
Lung infections are the eighth leading cause of death in the United States. Pneumonia due to bacterial pathogens is a major clinical challenge. Because of anatomical and physiological reasons, lungs are constantly exposed to microbial agents. To deal with invading microbes, lungs are well equipped with various host defense mechanisms (1, 2). However, opportunistic pathogens often overcome these host defense mechanisms, especially in persons with an impaired host defense, such as those in hospitals with underlying diseases or those on ventilators (3). Klebsiella pneumoniae is one such opportunistic Gram-negative bacterium (4). Pulmonary bacterial infections due to K. pneumoniae result in substantial mortality and therapeutic costs, and K. pneumoniae is the third leading cause of hospital-acquired bacterial pneumonia (4, 5). Significant increase in mortality is observed when infections spread to the peripheral organs, which can lead to sepsis and septic shock. Mortality due to K. pneumoniae bacteremia can be as high as 50–70% (6, 7). Although antibiotics are the main therapeutic intervention used, therapeutic failure is common, and significant mortalities still persist (8). The emergence of multidrug-resistant strains of this pathogen, especially strains that produce carbapenamase, are putting further limitations on the currently available therapies (9). Although the discovery rate of new antibiotics has been dismal (10), exploring host mechanisms that contribute to the regulation of infection could help to develop new therapies. Further, there is always a risk of emerging resistant strains against newly developed antibiotics. In contrast, host-targeted therapies are not prone to develop resistance by pathogens, at least theoretically. However, very little is known about the mechanisms involved in bacterial dissemination from the lung to systemic circulation and the host factors that are responsible for this phenomenon.
Chitinase and chitinase-like proteins (CLPs) are a conserved group of proteins that belong to the 18–glycosyl hydrolase family (11). In the absence of chitin biosynthesis in mammals, chitinases were hypothesized to have important roles in physiology. Their well-documented regulation in various diseases and pathological conditions further supports their important biological roles (12–14). Chitotriosidase (chit1) is one of the two true chitinases present in mammals, the other one being acidic mammalian chitinase (AMCase) (15, 16). Whereas chit1 is the most prominent chitinase present in humans, mice express both chit1 and AMCase (17). The role of chit1 in fungal and parasitic infection is indicated by both epidemiological and experimental studies (18–21). A significant proportion of various populations, including Asian, white, and Indian populations, are deficient (5–20%) in enzymatic activity, mainly due to a 24-bp duplication mutation (22, 23). Interestingly, these individuals without chitinase activity do not express any obvious related phenotypes or reported abnormalities. However, how these individuals respond to bacterial infections is not known.
The role of chit1 has been proposed in host immunity mainly because of its high expressions at anatomical locations where host pathogen interactions take place, such as the lungs and the gut (17, 24). Further, supporting this belief, chit1 is stored in and released from macrophages, one of the main effector cells against invading pathogens (25). However, the regulation and role of chit1 in bacterial infection has not been explored in vivo.
In the current study, we explored the regulation and the role of chit1 in K. pneumoniae lung infection. In this article, we report that chitinase activity is significantly downregulated during K. pneumoniae infection, which is mediated by the degradation of chit1 by neutrophil elastases. In this infection model, we show that chit1 deficiency provides a significant advantage to the mice by limiting bacterial dissemination. This resulted in improved survival in mice in the presence or absence of antibiotic therapy. Altered Akt activation in chit1-deficient mice was associated with improved outcomes in K. pneumoniae infection.
Materials and Methods
All animal studies were done according to Institutional Animal Care and Use Committee–approved protocols at Yale University.
Bacterial infection model
Mouse-adapted laboratory strain of K. pneumoniae (ATCC 43186) was grown on Luria broth plates from the glycerol stocks stored in −80°C. The next day, a single colony was transferred to liquid broth to culture overnight, and the next day it was subcultured for 1 h to bring the bacteria into log phase of growth. The numbers of CFUs were estimated by measuring OD at 600 nm and were confirmed by plating the inoculum. Mice were inoculated intratracheally by injecting 50 μl of PBS solution containing 5000 CFUs of K. pneumoniae.
For intratracheal inoculation, mice were anesthetized using a mixture of ketamine and xylazine (100 and 10 mg/kg, respectively). A small incision was made on the neck to expose the trachea, and bacterial suspension was instilled directly into the trachea. The wound was sealed with Vetbond surgical glue, and the mice were observed until they recovered from anesthesia.
For peritoneal infection, 104 CFUs of K. pneumoniae, suspended in 200 μl of PBS, were injected into the peritoneal cavity of mice. Mice were euthanized at 24 h postinfection to harvest peritoneal lavage fluid, spleen, and bronchoalveolar lavage fluid (BAL). Peritoneum was lavaged twice with 5 ml of sterile PBS each time.
For Pseudomonas aeruginosa infections, PAO1 strain was grown in the similar manner as described for K. pneumoniae. Mice were inoculated intratracheally with 1 × 107 CFUs per mouse, and organs were harvested at 12 h postinfection.
For LPS administration in the lungs, LPS from P. aeruginosa or from K. pneumoniae (Sigma-Aldrich, St. Louis, MO) was used at 5 μg/mouse by intratracheal route, and mice were sacrificed at 12 h postinfection.
Purification of neutrophils and preparation of neutrophil lysates
Mouse bone marrows were isolated from donor mice and flushed to isolate bone marrow cells. Purified neutrophils were isolated using a neutrophil isolation kit from Stemcell Technologies according to the manufacturer’s protocol. This kit uses a negative selection method using magnetic beads.
Measurement of chitinase activity
Chitinase activity in the BAL samples was measured using a fluorescence-based assay. BAL samples (5 μl) were incubated with 4-methylumbelliferyl β-D-N,N′-diacetylchitobioside hydrate solution for 15 min at 37°C, and the fluorescence intensity was measured at 450 nm. BAL samples from IL-13 transgenic mice were used as positive controls, as previous studies have indicated significant elevation of chitinase activity in these mice (26).
Western blot studies
Western blots of BAL samples were performed by loading 30 μl of BAL samples from each mouse in 4–20% gradient SDS-PAGE gel. For lung tissue and cell lysates, 30 and 10 μg of protein, respectively, were loaded. After electrophoresis, the gels were transferred to a polyvinylidene difluoride membrane using the Trans-Blot system from Bio-Rad. Membranes were then blocked with 5% milk for 1 h and incubated with primary Abs overnight. HRP-labeled secondary Abs were used. Bands were detected by using HRP substrate. Band intensities were measured using the Bio-Rad ChemiDoc MP Imaging System from Bio-Rad. Abs to Phospho-Akt (T308), total Akt, and anti-rabbit secondary Abs were purchased from Cell Signaling Technology (Danvers, MA). Abs to chit1 and AMCase were purchased from LSBio (Seattle, WA). β-Actin was purchased from Santa Cruz Biotechnologies.
Total RNA from the lung tissue was extracted using the Qiagen RNAeasy Kit as per the manufacturer’s instructions. cDNA was synthesized using the iScript reverse transcriptase kit using instructions provided with the kit. Quantitative PCR assays were performed using the SYBR Green Master Mix. The following primers were used in this study: AMCase forward: 5′-ACA AGC ATC TCT TCA CTG TCC TGG T-3′, reverse: 5′-TGG ATG TTG GAA ATC CCA CCA GCT-3′; chit1 forward: 5′-CGG CAG GAA CTA AAT CTT CCA T-3′, reverse: 5′-TGG GCG TGG CTC AGG TAT-3′; 18S forward: 5′-GCA ATT ATT CCC CAT GAA CG-3′, reverse: 5′-AGG GCC TCA CTA AAC CAT CC-3′.
Harvesting BAL samples
Mice were euthanized at a given point of time postinfection or at baseline as indicated. The trachea was exposed by making a small cut on the neck and then inserted with a 22-gauge catheter as described before (27). Lungs were lavaged by two aliquots of 750 μl of ice-cold sterile PBS. BAL was kept on ice or 4°C until further processing. These BAL samples were centrifuged to pellet the cells, and the cell-free supernatant was collected in separate tubes and stored at −80°C for further analysis.
Enumeration of bacterial burden in BAL, lung, and spleen
Small aliquots of BAL samples were serially diluted in sterile broth and plated on agar plates. Aseptically isolated left lung or spleens were homogenized in 1 ml of sterile PBS and then serially diluted and plated on agar plates. Numbers of CFUs were estimated by counting the number of colonies on the agar plates after overnight incubation at 37°C.
Cell count and differential counts
Total numbers of WBCs in the BAL samples were collected by resuspending the cell pellet obtained from BAL samples from mice into PBS. The cells were counted using the Beckman Coulter cell counter. Approximately 1 × 105 cells were used to prepare cytospin slides, which were stained with HEMA-3 stain, and the numbers of macrophages and neutrophils were estimated by counting at least 200 cells per slide.
Lung pathology scoring
Lungs were inflated with 0.5% low-melting agar and fixed in formalin. Tissue sections were stained with H&E. Lung scoring was done based on scores from 0 to 4, in which 0 is no pathologic condition and 4 is severe. Scores for peribronchial and perivascular inflammation were added for each mouse.
ELISA and Bio-Plex assays
Cytokines in the BAL samples were measured using conventional sandwich ELISA duoset kits from R&D as per the manufacturer’s instructions. Briefly, 96-well plates were coated overnight with capture Ab. Unbound Ab was washed off the next day, and the test samples were loaded and incubated for 2 h. Detection Ab was added after washing off the samples. Detection Ab was washed off, and streptavidin HRP was added to incubate for 20 min. After unbound streptavidin was washed off, TMB substrate was added to react with bound HRP for 20 min. Stop solution was added to stop the reaction, and absorbance was measured at 450 nm. The Bio-Plex cytokine assay was performed using Bio-Plex Pro Mouse Cytokine kit from Bio-Rad according to the instruction provided with the kit.
For survival studies, mice were infected with 5 × 103 CFUs of K. pneumoniae and were observed every day for mortality. Mice were euthanized once they were considered a humane concern and considered as dead. Blood samples were collected from the orbital sinus of some of these mice while mice were under anesthesia by ketamine/xylazine at given time points to estimate bacterial burden in the blood. For survival with antibiotics, two doses of antibiotics were administered at 48 and 60 h postinfection.
Primary macrophage and cell line culture
Mouse bone marrow–derived macrophages (BMDMs) from wild type and chit1−/− mice were developed by culturing the bone marrow in RPMI 1640 containing 10% FBS and 20% of L929 cell-conditioned media for 7 d. Cells were treated with K. pneumoniae LPS at 500 ng/ml for the indicated time points. RAW 264.7 cells from the American Type Culture Collection were grown in DMEM with 10% FBS and 1% penicillin/streptomycin solution.
Data were analyzed using GraphPad prism software version 7. Two groups were compared using Student t test. Experiments comprising more than two groups were analyzed using one-way ANOVA, and Sidak multiple comparison test was used. For survival studies, Kaplan–Meier curves were prepared, and data were analyzed using Wilcoxon test.
Airway chitinase activity is downregulated during infection and inflammation
Chitinase activity is dynamically regulated during various inflammatory conditions and is believed to be an important mediator of inflammation and disease progression (12, 16, 28). To understand the regulation of chitinase activity during bacterial lung infection, chitinase activity was measured in the BAL samples of mice infected with K. pneumoniae for different durations. At baseline, mice have a marked chitinase activity which decreases in a time-dependent fashion during K. pneumoniae infection, and by 48 h postinfection, a significant drop in chitinase activity was observed (596 ± 43 versus 390 ± 33 arbitrary units [AU], p < 0.05, Fig. 1A). A similar drop in the chitinase activity was observed during Pseudomonas infection or sterile inflammation with the bacterial product LPS (394 ± 32 versus 223 ± 26 AU, p < 0.01, Fig. 1B). Unlike BAL samples, no detectable chitinase activity was observed in the peritoneal lavage fluids obtained from mice (Supplemental Fig. 2C). To determine whether infection with K. pneumoniae leads to the downregulation of the expression of chitinases, the mRNA levels of chit1 and AMCase were measured. Expression levels of chit1 (1.00 ± 0.42 versus 0.94 ± 0.43, p = ns, Fig. 1C) and AMCase (1.00 ± 0.40 versus 1.03 ± 0.46, p = ns, Fig. 1D) in the lungs of infected mice were similar to those of uninfected mice, suggesting that the chitinase expression is not responsible for the decreased activity during infection. Together, our results suggest that chitinase activity is downregulated during bacterial infection and LPS-mediated inflammation without altering the lung expression of true chitinases.
Inflammatory cells mediate downregulation of chitinase activity during lung infection by cleaving chit1
Upregulation of chitinase activity is often associated with the type 2 inflammatory responses, especially those mediated by eosinophils (14, 26). However, the role of inflammatory cells such as neutrophils in regulating chitinase activity is not known. We found that neutrophil infiltration in the lung during K. pneumoniae infection (Fig. 2A) inversely correlated with chitinase activity in the BAL of mice (r = −0.93, p < 0.05, Fig. 2B). No significant correlation was observed with macrophage infiltration or chitinase activity (r = −0.84, p = ns, Fig. 2D). To establish a causal relationship between neutrophils and downregulation of chitinase activity, we instilled neutrophil lysates directly into the lungs of mice and measured chitinase activity in the BAL samples of mice. Similar to infection and inflammation, instillation of neutrophil lysates resulted in a significant decrease in chitinase activity in the BAL samples (800 ± 45 versus 601 ± 39 AU, p < 0.05, Fig. 2E). By Western blot analysis, we observed a significant reduction in the levels of chit1 protein during infection as well as upon instillation of neutrophil lysates (Fig. 2F, 2G). The levels of AMCase did not change in the BAL samples postinfection or after instillation of neutrophil lysate. In agreement with these results, we also observed that the treatment of BAL fluid with neutrophil elastase significantly degrades chit1, but no susceptibility of AMCase to elastase was observed (Fig. 2H). Together, these results suggest that neutrophil infiltration during lung infection or sterile inflammation contributes toward decreased chitinase activity mediated by proteolytic cleavage of chit1.
Chit1 deficiency in mice limits bacterial dissemination and weight loss during Klebsiella infection
Chit1 is the most prominent chitinase present in humans and also contributes significantly to chitinase activity in mice (17) (Supplemental Fig. 2). To understand the role of chitinases during lung infection, we infected chit1-deficient (chit1−/−) mice to study their weight loss, lung bacterial burden, spleen bacterial burden (as a marker of dissemination), and total protein content in BAL (as a marker of lung injury). These data show that at 48 h postinfection, chit1−/− mice maintain their weight closer to baseline compared with wild type mice (weight loss in wild type 14.64 ± 0.84% versus 12.37 ± 0.74% in chit1−/− mice, p = 0.05, Fig. 3A). The decreased weight loss corresponded to decreased bacterial burden in spleen of chit1−/− mice by a log change (6.06 ± 0.26 in wild type versus 5.05 ± 0.25 log CFUs/spleen in chit1−/− mice, p < 0.01 Fig. 3B), whereas similar bacterial burden in the BAL was observed (6.20 ± 0.18 in wild type versus 5.97 ± 0.17 log CFUs/ml BAL in chit1−/− mice, p = ns, Fig. 3C). Similar to that in BAL, the bacterial burden in the lung tissue was not different among the two groups (5.52 ± 0.45 versus 6.39 ± 0.41, p = ns). No difference in the lung injury was observed, as indicated by the total protein content in the BAL (169.50 ± 13.25 in wild type versus 186.80 ± 12.95 μg/ml BAL in chit1−/− mice, p = ns, Fig. 3D). In contrast, upon peritoneal infection, similar bacterial burdens were observed in the peritoneal lavage fluid, BAL, and spleen in wild type and chit1−/− mice (Supplemental Fig. 3B–D). Similar weight loss between wild type and chit1−/− mice was observed during peritoneal infection (Supplemental Fig. 3A), along with similar levels of inflammation (Supplemental Fig. 3E, 3F), suggesting lung-specific protective effects of chit1 deficiency during K. pneumoniae infection. Interestingly, unlike BAL fluid, which has significant chitinase activity at baseline, the peritoneal lavage fluid had no detectable chitinase activity (Supplemental Fig. 2C).
To test whether these protective effects are specific to K. pneumoniae or extend to other Gram-negative infections, we use a Pseudomonas infection mouse model. Using this mouse model, we did not see any protection in chit1−/− mice, as evident by increased bacterial load in lung tissues as well as in the spleen (Supplemental Fig. 3G–I). Together, these results suggest that chit1 specifically regulates bacterial dissemination during K. pneumoniae lung infection without significantly altering pulmonary bacterial clearance and lung injury.
Chit1−/− mice have similar inflammatory responses during K. pneumoniae lung infection
To understand the possible mechanisms behind improved health and limited bacterial dissemination in the absence of chit1, we sought to determine the inflammatory response during lung infection with K. pneumoniae. K. pneumoniae infection in mice leads to the elevation of many of the inflammatory cytokines, resulting in inflammatory cell recruitment (data not shown). Compared with wild type mice, chit1−/− mice had similar inflammatory cell infiltration at 48 h after K. pneumoniae infection (0.55 ± 0.07 in wild type versus 0.57 ± 0.09 × 106 per mouse in chit1−/− mice, p = ns, Fig. 4A). The numbers of neutrophils (0.32 ± 0.07 in wild type versus 0.28 ± 0.06 × 106 per mouse in chit1−/− mice, p = ns, Fig. 4B) and macrophages (0.19 ± 0.02 in wild type versus 0.27 ± 0.04 × 106 per mouse in chit1−/− mice, p = ns, Fig. 4C) were similar in these mice BALs after lung infection. Histological analysis revealed cell infiltration and consolidation in the lung tissue of infected mice, but there was no apparent difference between lungs from wild type and chit1−/− mice as indicated by lung pathology scores (4.73 ± 0.59 in wild type versus 4.9 ± 0.57 in chit1−/− mice, p = ns, Fig. 4K, 4L). The levels of cytokines were also measured in the BAL samples obtained from the infected mice. The levels of various inflammatory cytokines such as TNF-α, IL-β, IL-6, IL-12, IFN-γ, and IL-17 were not different between the two groups upon infection (Fig. 4D–I). Similarly, the anti-inflammatory cytokine IL-10 was similar between infected wild type and chit1−/− mice (Fig. 4J). The levels of other tested cytokines such as IL-2, IL-4, and IL-5 were below the detection limit (1 pg/ml) in BAL samples of both wild type and chit1−/− mice. Corresponding to these results, BMDMs obtained from wild type and chit1−/− mice had similar cytokine responses when stimulated with LPS (Supplemental Fig. 4D, 4E). Also, cells of the macrophage cell line RAW 264.7 produced similar cytokines in response to LPS in the presence or absence of recombinant chit1 (Supplemental Fig. 4F, 4G). Interestingly, BMDMs obtained from chit1−/− mice had better control on bacterial growth in vitro compared with the BMDMs obtained from wild type mice (Supplemental Fig. 4H).
Overall, these results indicate that chit1 deficiency does not play a significant role in regulating inflammatory response during lung infection but still contributes to the protection against K. pneumoniae infection in macrophages.
Chit1−/− mice regulate early dissemination of bacteria and have a survival advantage with K. pneumoniae lung infection with or without antibiotic therapy
To better understand the mechanisms of lower bacterial burden in the spleen of chit1−/− mice, we sought to determine the time course of bacterial dissemination in these mice. After K. pneumoniae infection, bacterial burden was measured in the blood samples of these mice. Chit1−/− mice have lower bacterial burden in their blood compared with wild type mice, which is evident as early as 24 h postinfection and remains higher up to 48 h (196.18 versus 8.82 × 103 CFU/ml at 24 h, 185.40 versus 28.75 × 103 CFU/ml at 36 h, and 544.67 versus 23.64 × 104 CFU/ml at 48 h, p = 0.011), suggesting better control of bacterial dissemination in the chit1−/− mice (Fig. 5A). Next, we investigated if limiting the bacterial dissemination results in better survival in chit1−/− mice. As expected, we observed a significant survival advantage in the chit1−/− mice during lung infection. On day 3, whereas 70% of mice succumbed to infection in the wild type group, only 44% died of infection in the chit1−/− group (Fig. 5B; p < 0.05). To further mimic a clinical situation, we administered two doses of antibiotics in the mice at 48 and 60 h postinfection to understand the role of chit1 deficiency in this model. Whereas even after two doses of antibiotics, wild type mice maintained detectable bacterial burden in their circulation, the chit1−/− mice effectively cleared circulating pathogens after similar doses of antibiotics (Fig. 5C). The lower bacterial burden in chit1−/− mice after antibiotic treatment resulted in a rather dramatic increase in the survival during K. pneumoniae lung infection. Whereas ∼80% of mice survived in the chit1−/− group, only 10% of mice survived in the wild type group (Fig. 5D; p < 0.05). Taken together, these data strongly suggest that chit1 deficiency provides mice with a survival advantage, with or without antibiotics, because of better control of early dissemination.
Altered Akt activation in chit1−/− mice during lung infection
Akt signaling has been shown to play important roles during host–pathogen interactions (29). CLP have been shown to regulate the Akt signaling pathway. AMCase has been shown to protect epithelial cell apoptosis by regulating the PI3K–Akt pathway, whereas chitinase 3–like protein 1 (Chil1) directly activates Akt signaling (30, 31). We sought to determine whether chit1 regulates the Akt pathway during K. pneumoniae infection. Our data indicate that chit1−/− mice maintain a lower Akt activation compared with wild type mice at baseline (1.00 ± 0.14 in wild type mice versus 0.45 ± 0.03 in chit1−/− mice, p < 0.01, Fig. 6A). Interestingly, upon K. pneumoniae lung infection, chit1−/− mice have a significantly higher activation of Akt signaling compared with wild type mice (1.00 ± 0.07 in wild type mice versus 1.27 ± 0.10 in chit1−/− mice, p < 0.05 Fig. 6B). Similar to the lungs, BMDMs from chit1−/− mice had significantly lower activation of Akt at baseline (1.0 in wild type cells versus 0.79 ± 0.04 in chit1−/− cells, p < 0.05); however, upon stimulation with LPS, a robust activation of Akt was observed in chit1−/− macrophages (0.79 ± 0.04 at baseline versus 1.14 ± 0.12 with LPS stimulation, p < 0.005) but not in wild type macrophages (1.0 at baseline versus 0.96 ± 0.04 with LPS stimulation, p = ns) (Fig. 6C). In contrast, there were no differences in the activation of the MAPK pathway as measured by the activation of P42/44 or P38 at either baseline or upon infection (Fig. 6D, 6E, 6G, 6H). Similar results were obtained in BMDM stimulation with LPS (Fig. 6F, 6I).
Inhibition of Akt pathway impairs bacterial clearance, dissemination, and survival in Klebsiella lung infection
To further understand the role of Akt during K. pneumoniae lung infection, we used Akt inhibitor wortmannin. A single dose of wortmannin impaired health (weight loss 4.78% ± 0.51 in controls versus 9.14% ± 0.65 in treated, p < 0.005, Fig. 7A) and decreased survival in mice (median survival of 4 d in controls versus 2 d in treated, p < 0.01, Fig. 7B). As expected, the wortmannin-treated mice had significantly elevated bacterial load in the spleen (3.21 ± 0.556 versus 7.09 ± 0.30 log CFUs/spleen) and BAL (5.46 ± 0.14 versus 6.85 ± 0.04 log CFUs/ml BAL) (Fig. 7C, 7D, respectively), suggesting the importance of the Akt pathway during K. pneumoniae lung infection.
Lung infections result in significant morbidity and mortality in humans. In the fight against infectious bacteria, antibiotics provided us with a significant edge over pathogens, but this advantage diminished with the development of antibiotic resistance in pathogens (32). Currently, multidrug-resistant strains are threatening to revert all the advantages achieved in the last eight decades. The rapidly evolving antibiotic resistance and slow antibiotic discovery prompted us to explore host mechanisms that might boost host immunity to help survive infections.
Chitinases are an evolutionally conserved group of proteins in organisms from yeasts, arthropods, and fruit flies to mammals, including humans (Supplemental Fig. 1). Shared and conserved areas of the chit1 gene can be seen across the different species from the depicted gene tree. The presence of chitinases in humans and other mammals has been puzzling in the absence of chitin synthesis. The obvious role of chitinases in physiology or host defense has not been proved experimentally. Also, a significant human population (5–20% of healthy human population) is devoid of chitinase activity because of a 24-bp mutation (22, 23, 33). However, a well-documented regulation of chitinases during many diseases and pathological conditions (16, 34, 35) suggests specific regulatory mechanisms of the activity and functions associated with chitinases.
In the current study, we explored the regulation of chitinases during lung infection with K. pneumoniae, a leading pathogen in hospital-acquired pneumonia and in patients with complex lung diseases (5, 36, 37). K. pneumoniae infection resulted in a time-dependent decrease in chitinase activity in the BAL fluid (Fig. 1A), which was independent of gene expression of AMCase and chit1 (Fig. 1C, 1D). The role of live pathogens in decreasing chitinase activity in the infected mice was excluded owing to similar downregulation of chitinase activity in LPS-injected mice (Fig. 1B). These data suggest that live pathogens are not essential to decrease chitinase activity.
To understand the mechanisms underlying the downregulation of chitinase activity, we characterized the inflammatory response during K. pneumoniae infection. We observed not only a strong negative correlation between chitinase activity and the number of neutrophils in BAL but also that direct instillation of neutrophil lysates in the lung effectively decreased chitinase activity, establishing a causal role of neutrophils. In our BAL samples from mice that were infected with K. pneumoniae or instilled with neutrophil lysates, we observed lower levels of chit1 protein, suggesting a mechanism for decreased chitinase activity (Fig. 2F, 2G). A similar breakdown of chit1 was observed when BAL samples were incubated with neutrophil elastase (Fig. 2H). Supporting our observations, a recent report studying the interaction of neutrophil enzymes with chit1 in fungus-infected cystic fibrosis patients reported that neutrophil enzymes can directly cleave chit1 present in the BAL samples of patients with cystic fibrosis (38). Interestingly, we did not see a reduction in the levels of AMCase, the second chitinase present in mammals (Fig. 2F, 2G). Supporting this observation, treatment of BAL with neutrophil elastase had no effect on AMCase levels. Overall, we propose in this article that neutrophils recruited during pulmonary infections directly decrease chitinase activity by cleaving chit1 with minimal effect on AMCase. It is important to note that only airway lavage fluids possessed significant chitinase activity at baseline, which was absent in the peritoneal lavage fluids (Supplemental Fig. 2C).
Many pathological conditions and diseases such as Gaucher disease, interstitial lung disease, chronic obstructive pulmonary disease, and diabetes have been shown to be associated with increased chitinase activity (16, 34, 35, 39). However, these studies did not explore the type of inflammation or cell types involved in these pathological conditions or the role of chit1 as a contributor to the underlying disease or pathophysiological conditions. Also, whether increased chitinase activity contributes to the increased susceptibility to infections in many of these diseases is not known.
To understand the role of chitinases in lung infections, we used mice that are deficient in chit1. Chit1 is the most prominent chitinase in humans, and almost all chitinase activity can be attributed to it (16). However, AMCase has been reported to be the major chitinase in mice (15, 26). Neither the role of chit1 in mouse lung nor its contribution to the overall chitinase activity has been well appreciated (17). In this report, we present data to show that chit1 is a significant contributor to the chitinase activity in mice, as chit1 knockout mice have significantly lower chitinase activity in BAL and serum (Supplemental Fig. 2). Emphasizing the importance of chit1, K. pneumoniae–infected chit1−/− mice exhibited a significant protected phenotype in which they lost significantly less weight compared with the wild type mice, suggesting better health in these mice (Fig. 3A). In agreement to the overall health, the chit1−/− mice have significantly lower bacterial dissemination to the spleen (Fig. 3B), a marker of bacterial dissemination, while maintaining similar bacterial load in the lung as well as similar protein levels in the BAL (a marker of lung leakage) (Fig. 3C, 3D). In agreement with similar bacterial burden and airway protein content, no difference in pathologic condition was observed between the two groups postinfection (Fig. 4K, 4L). However, the dissemination phenotype was visible at early time points, as indicated by decreased bacterial counts in the blood of chit1−/− mice (Fig. 5A). Also, the trend toward decreased bacterial number in the blood of chit1−/− mice suggests that the significant decreased bacterial load in the spleens of chit1−/− mice is due to limited dissemination from the lung rather than better peripheral control of infection. This theory is further supported by the fact that similar dissemination and health was observed in chit1−/− mice when infected by i.p. route (Supplemental Fig. 3). Chit1 is believed to be an important mediator of host response against pathogens based on its presence in macrophages, where it performs important functions during pulmonary infections. It is interesting to note that it was hypothesized that chit1 might possess antibacterial properties by itself or that it can promote antibacterial activity of lysozymes (40). Experiments showed that chit1 does not possess any bactericidal activity of its own or boost the bactericidal activity of lysozymes (40). Our study shows a better control of bacterial infection in chit1−/− mice, which further refutes any direct antibacterial role of chit1 against K. pneumoniae in vivo. In this regard, our data suggest that not only is chit1 dispensable for immunity against K. pneumoniae, it also plays a detrimental role in K. pneumoniae lung infection. Further, these effects are specific to K. pneumoniae lung infection. In a mouse model of Pseudomonas lung infection, we did not observe any protection in chit1−/− mice; indeed, our data show that they had higher bacterial load in the lung and spleen. Although Pseudomonas and Klebsiella are both Gram-negative pathogens associated with nosocomial lung infections, they have many differences in their virulence factors and pathomechanisms. Flagella and type-3 secretion system constitute important virulence factor in P. aeruginosa infection, whereas capsule is an important component of K. pneumoniae virulence (41, 42). Their virulence also varies to great extent in our mouse models, in which inoculation of only 5 × 103 is sufficient to cause a severe lung infection which effectively disseminates to other organs for Klebsiella. In contrast, 1 × 107 CFUs of P. aeruginosa are needed to establish severe infection in the lungs, with minimal dissemination to the periphery. These obvious differences might account for differences observed in chit1−/− mice between these two pathogens in this study.
The controlled dissemination in chit1−/− mice during K. pneumoniae infection resulted in increased survival, further emphasizing the advantage provided by chit1 deficiency during K. pneumoniae infection. However, the survival advantage provided by chit1 deficiency was limited to just a 1-d extension of survival in our infection model (median survival of 3 d versus 4 d, Fig. 5B). This might be attributed to the high lethality of K. pneumoniae and its ability to keep multiplying in the mouse despite an intact immune system. We also used antibiotics in our model to mimic a clinical scenario; patients with pneumonia are often treated with antibiotics (43). Two doses of the antibiotic treatment increased survival in both the groups compared with the mice that did not receive antibiotics (Fig. 5B versus Fig. 5D). However, chit1−/− mice had significantly increased survival upon antibiotic treatment compared with the wild type mice that received similar antibiotic treatment (Fig. 5D).
To understand the mechanisms that improved bacterial control in these mice, we sought to determine the inflammatory response in these mice, as inflammation plays an important role in limiting bacterial growth and spread during K. pneumoniae lung infection (44, 45). Lung infection produced a similar inflammatory response in wild type and chit1−/− mice at 48 h, as indicated by the number of recruited cells including neutrophils and macrophages, which are the major cell types in the airways of K. pneumoniae–infected mice (Fig. 4A–C). We also characterized inflammation at early time points postinfection and found a similar inflammatory response, even at early phase of infection (data not shown), which suggests that chit1 has a limited impact on the inflammatory response during K. pneumoniae bacterial infection. In agreement with these results, chit1−/− mice invoked similar levels of inflammatory responses upon administration of LPS in the lung (Supplemental Fig. 4A–C).
Cytokines contribute to the orchestration of the inflammatory response (45–47). Levels of many cytokines, both inflammatory and anti-inflammatory, were measured to find out if chit1 plays a role in regulating cytokine response during lung infection. Postinfection, both wild type and chit1−/− mice produced similar levels of cytokines in their lungs (Fig. 4D–J), suggesting a limited role of chit1 in regulating cytokine response during K. pneumoniae lung infection. These data explain the similar numbers of cell recruitment in the lung during infection in chit1−/− mice. In agreement with these in vivo data, BMDMs from wild type and chit1−/− mice, or macrophage cell line RAW 264.7 upon treatment with chit1, produced similar cytokine levels (Supplemental Fig. 4). These data exclude the role of chit1 in regulating the inflammatory response to K. pneumoniae bacteria or bacterial product LPS.
To elucidate the mechanisms that might contribute to the improved outcome in lung infection, we explored the regulation of the Akt pathway during lung infection. True chitinase AMCase and CLP BRP 39 have been shown to regulate Akt activity (30, 48). The Akt pathway plays an important role in both immune and structural cells. It has been reported that Akt activation can increase the phagocytic activity in macrophages to engulf more of the bacteria while also increasing extracellular trap formation in neutrophils (49, 50). Similarly, in epithelial cells, Akt has been shown to control bacterial transmigration across the gut epithelial monolayer (51). In this study, we observed that while at baseline, chit1−/− mice maintained a lower level of Akt activation, but upon infection with K. pneumoniae, chit1−/− had a robust increase in Akt activation (Fig. 6A, 6B). Similar to the whole-lung Akt level, BMDMs from chit1−/− mice had a lower Akt phosphorylation at baseline. However, upon stimulation with LPS, chit1−/− BMDMs had robust Akt activation, which was absent in wild type macrophages (Fig. 6C). This altered activation of Akt in chit1−/− macrophages was associated with increased ability to control growth of K. pneumoniae in vitro. We believe that a contribution of both immune cells and structural cells might have contributed to the altered dissemination phenotype we observed in chit1−/− during K. pneumoniae infection. We also determined the role of chit1 in regulating the MAPK pathway by measuring the activation of P42/44 and P38 MAPK proteins. There was no significant difference in the MAPK protein levels either at baseline or upon infection (Fig. 6D–I).
To understand the role of Akt, we used a pharmacological inhibitor of Akt in our K. pneumoniae infection mouse model. Akt inhibition led to a significant impairment of the host’s ability to maintain health and survival by impairing bacterial clearance and increasing bacterial dissemination (Fig. 7), suggesting the importance of the Akt pathway in K. pneumoniae lung infection.
Various theories have been proposed to explain the presence of chitinase in mammals. Earlier evidence suggested increased chitinase deficiency among white populations, whereas a conserved presence was observed in African populations (21). This study indicated that improved living conditions in white populations have led to enzyme deficiency, whereas the continuous threat of malaria and other parasitic infections provided selection pressure to retain intact chit1 gene in African populations (21). The same group demonstrated elevated levels of chit1 in the colostrum of African women compared with white women (52). However, later observations did not support these initial reports. Studies demonstrated no significant correlation between parasitic information and chit1 genotype (33, 53, 54). Furthermore, high prevalence of chitinase deficiency was observed in Peruvians with high prevalence of enteroparasites and high consumption of chitin-containing food, refuting previous beliefs that chitinases play a role in the digestion of chitin-containing food or protection against parasitic infection (33). The obvious role of chit1 in fungal infection was proposed because of its in vitro fungicidal activity (15, 25), but this role was refuted by a recent study using an in vivo fungal infection model (55). In this study, chit1 deficiency was shown to provide a survival advantage in a cryptococcal infection model by limiting pathological inflammation mediated by chit1-mediated chitin recognition (55). This study refutes that chit1 has an important role in the host defense against fungal infections.
It is interesting to compare the effects of chit1 in lung infection with other members of chitinase and CLP. Our laboratory has reported that Chil1 plays an essential role in lung infection with Streptococcus. Chil1 protects macrophages against Streptococcus-induced pyroptosis, which leads to improved bacterial clearance and limited pathologic condition (56). Similarly, in Pseudomonas lung infection, absence of chil1 results in an exaggerated inflammatory response and a diminished survival (57), suggesting its essential role in both Gram-positive and Gram-negative bacterial infections. These studies show distinct roles between different chitinases and CLP.
Overall, our study suggests that chitinase activity is actively regulated during bacterial lung infection, mainly by proteolytic cleavage of chit1 mediated by infiltrated neutrophils. Deficiency of chit1 provides a survival advantage to the host during lung infection by limiting bacterial dissemination. This might provide a new therapeutic target to increase host immunity during bacterial infections. These data also suggest that the loss of chitinase activity in humans provides an edge during bacterial lung infection by putting a selection pressure on the mutant gene.
This work was supported by National Heart, Lung, and Blood Institute Grants R01HL126094 (to C.S.D.C.) and ALA-513385 (to L.S.).
The online version of this article contains supplemental material.
The authors have no financial conflicts of interest.