Supernatant of platelet-Klebsiella pneumoniae coculture induces apoptosis-like death in Klebsiella pneumoniae

ABSTRACT

Multidrug-resistant Klebsiella pneumoniae strains, especially carbapenem-resistant K. pneumoniae, have become a rapidly emerging crisis worldwide, greatly limiting current therapeutic options and posing new challenges to infection management. Therefore, it is imperative to develop novel and effective biological agents for the treatment of multidrug-resistant K. pneumoniae infections. Platelets play an important role in the development of inflammation and immune responses. The main component responsible for platelet antibacterial activity lies in the supernatant stimulated by gram-positive bacteria. However, little research has been conducted on the interaction of gram-negative bacteria with platelets. Therefore, we aimed to explore the bacteriostatic effect of the supernatant derived from platelet-K. pneumoniae coculture and the mechanism underlying this effect to further assess the potential of platelet-bacterial coculture supernatant. We conducted this study on the gram-negative bacteria K. pneumoniae and CRKP and detected turbidity changes in K. pneumoniae and CRKP cultures when grown with platelet-K. pneumoniae coculture supernatant added to the culture medium. We found that platelet-K. pneumoniae coculture supernatant significantly inhibited the growth of K. pneumoniae and CRKP in vitro. Furthermore, transfusion of platelet-K. pneumoniae coculture supernatant alleviated the symptoms of K. pneumoniae and CRKP infection in a murine model. Additionally, we observed apoptosis-like changes, such as phosphatidylserine exposure, chromosome condensation, DNA fragmentation, and overproduction of reactive oxygen species in K. pneumoniae following treatment with the supernatant. Our study demonstrates that the platelet-K. pneumoniae coculture supernatant can inhibit K. pneumoniae growth by inducing an apoptosis-like death, which is important for the antibacterial strategies development in the future.

IMPORTANCE

With the widespread use of antibiotics, bacterial resistance is increasing, and a variety of multi-drug resistant Gram-negative bacteria have emerged, which brings great challenges to the treatment of infections caused by Gram-negative bacteria. Therefore, finding new strategies to inhibit Gram-negative bacteria and even multi-drug- resistant Gram-negative bacteria is crucial for treating infections caused by Gram-negative bacteria, improving the abuse of antibiotics, and maintaining the balance between bacteria and antibiotics. K. pneumoniae is a common clinical pathogen, and drug-resistant CRKP is increasingly difficult to cure, which brings great clinical challenges. In this study, we found that the platelet-K. pneumoniae coculture supernatant can inhibit K. pneumoniae growth by inducing an apoptosis-like death. This finding has inspired the development of future antimicrobial strategies, which are expected to improve the clinical treatment of Gram-negative bacteria and control the development of multidrug-resistant strains.

INTRODUCTION

Klebsiella pneumoniae is one of the common pathogens responsible for drug-resistant opportunistic infections in hospitalized patients (1), and drug-resistant strains of this bacterium are often resistant to multiple antimicrobials, a phenomenon termed “multidrug resistance” (2, 3). K. pneumoniae, a member of the Enterobacteriaceae family, is an opportunistic gram-negative pathogen that poses significant clinical and public health threats (4). In recent years, K. pneumoniae has emerged as one of the most common factors responsible for hospital- and community-acquired infections and a major cause of neonatal sepsis (5). Due to multidrug-resistant K. pneumoniae strains producing extended-spectrum β-lactamases or carbapenemases, K. pneumoniae spreads rapidly worldwide, making the treatment of infections caused by these strains difficult (6–8). It is reported that in Europe alone, such strains cause more than 90,000 infections and more than 7,000 deaths annually. The pathogenicity and multidrug resistance of K. pneumoniae make its treatment even more challenging (9). Therefore, to deal with the threat of bacterial resistance and improve the treatment of K. pneumoniae infections, innovative antibacterial agents and infection control strategies are urgently needed (10, 11).

A large number of studies have shown that platelets (PLTs）play an important role in host defense against pathogens (12). Platelets are anucleated cells derived from megakaryocytes in the bone marrow. Platelets not only play a role in the classical coagulation pathway but also actively participate in the immune response against pathogenic microorganisms (13). Studies have shown that platelets participate in the regulation of immune cells by releasing platelet microvesicles (PMVs), lipid mediators, nucleoside, mitochondrial DNA, growth factors, cytokines, and chemokines (13). Moreover, platelets are also directly involved in immune defense. Platelets participate directly in a bidirectional interaction with microbial pathogens, as they can sense and respond to bacterial infection signals, but at the same time, some pathogenic microorganisms can alter and modulate the structure and function of platelets (14, 15). Similar to innate immune cells, platelets contain a variety of pattern recognition receptors, such as Toll-like receptors (TLRs), C-type lectin receptors, and nucleotide-binding oligomerization domain-like receptors, which recognize different components of bacteria during infection (16, 17). Furthermore, platelets act as sentinel cells that can sense the invasion of pathogens and neutralize the invading bacteria (13). Specifically, activated platelets release antimicrobial peptides via α-granules, which perform important immune functions (12, 18). When pathogens such as bacteria or viruses induce platelet activation (19), it is mainly the α-granules that release antimicrobial peptides containing β-defensin 2, thymosin β4, and other derived antimicrobial peptides (such as fibrinopeptide A and B) to fight against the pathogens directly (20, 21). Humans have more than 30 genes coding for β-defensins, mainly expressed in epithelial cells, but they are also expressed in platelets and are secreted after thrombin stimulation (22). Most of the antibacterial effects of platelets have been shown to be associated with their secretory ability induced by bacteria, with previous reports showing platelets engulfing S. aureus and methicillin-resistant S. aureus (23). Platelet-derived transforming growth factor beta 1 plays an important role in inhibiting the growth of S. aureus (24). Furthermore, various antimicrobial peptides released from platelets, such as thrombocidin-1 (TC-1) and thrombocidin-2 (TC-2), have been found to exert different antibacterial effects against different bacteria. Both the components (TC-1 and TC-2) exhibit a potent bactericidal activity against Bacillus subtilis, to a lesser extent against Escherichia coli and S. aureus, and a fungicidal activity against Cryptococcus neoformans (12). Platelets contain stable messenger RNA transcripts and the translation machinery for protein synthesis despite being anucleated cells, as they inherit them from their megakaryocytic precursors (25). This special feature precisely facilitates the release of a variety of complex types of inhibitory protein molecules by the platelets.

The current research on the antibacterial mechanism of platelets mainly focuses on the direct and indirect antibacterial effects of platelets. The direct antibacterial effect is manifested as the interaction between platelet glycoproteins, particles, bacterial surface proteins, and antimicrobial peptides, such as CXCL4, CXCL7, TGF-β, and β-defensin. Indirect antibacterial effect is mediated by plasma proteins such as fibrinogen, von Willebrand factor, immunoglobulin G (IgG), platelet glycoproteins GPIIbIIIa and GPIbα, FcγRIIα, complement receptors, and TLRs (26–28).

The role of platelets in the innate immune system has been well established in previous studies. Platelets directly induce DNA damage and the inhibition of S. aureus division (23, 24). Platelets have also been shown to inhibit the growth of methicillin-resistant S. aureus by inducing hydroxyl radical-mediated apoptosis-like cell death (29). Meanwhile, the supernatant of platelets incubated with S. aureus has been reported to exert the same inhibitory effect as platelets do on the bacteria (30). A wealth of research on gram-positive bacteria has demonstrated that it is crucial for platelets to release bactericidal agents when they bind bacteria and get activated (31, 32). However, there is a paucity of studies on the interaction between gram-negative bacteria and platelets, and little is known about the mechanisms underlying the action of gram-negative bacteria. Moriarty et al. (31). have shown that E. coli can induce platelet aggregation and antibacterial effect in an IgG-dependent and FcγIIa-dependent manner, but the secretion of antimicrobial substances remains not clarified (33). In addition, due to the diversity and complexity of pathogens, the mechanism underlying platelet-mediated antibacterial defense and the clinical potential of platelets remain poorly understood.

To investigate the inhibitory effect of platelets on K. pneumoniae bacteria and the underlying mechanism, we assessed whether the supernatant obtained from platelet-K. pneumoniae coculture possesses an antibacterial activity that can inhibit K. pneumoniae growth. The research on the function and mechanism of antibacterial effect of platelets may pave the way for the development of novel antibacterial drugs and help in mitigating the abuse of antibiotics and controlling the spread of multidrug-resistant bacteria.

RESULTS

Supernatant of platelet-K. pneumoniae coculture inhibits K. pneumoniae growth in vitro

Using flow cytometry, we determined that the purity of washed platelets (CD41a⁺) was more than 90% (Fig. S1A through C). Platelets (100–300 × 10⁹/L) and K. pneumoniae (10⁸ CFU/mL) were cocultured, and the supernatant of the platelet-K. pneumoniae coculture was obtained (Fig. S2A through C). Then, the platelet-K. pneumoniae coculture supernatant was extracted and added to the K. pneumoniae cultures, with the optimal ratio of the supernatant to K. pneumoniae being set to 10⁵ CFU/mL. We observed that the addition of the platelet-K. pneumoniae coculture supernatant inhibited the growth of K. pneumoniae cultures. According to the optical density at 600 nm (OD_600nm) value and bacterial colony counts of K. pneumoniae, we found that with the extension of coculture time, the bacteria density of the supernatant-treated K. pneumoniae group was significantly lower than that of the K. pneumoniae group and S_PLT + KP control group at 8 h, and the inhibition of K. pneumoniae growth mediated by the platelet-K. pneumoniae coculture supernatant lasted for 24 h (Fig. 1A and B). As shown in Fig. 1C and D, turbidity of the supernatant-treated K. pneumoniae group was much lower than that of the K. pneumoniae and S_PLT + KP control group. In addition, the counts of bacteria in Luria-Bertani (LB) plates were markedly lower than those in K. pneumoniae and S_PLT + KP control group at 8 h (Fig. 1E and F).

Fig 1.

Platelet-K. pneumoniae coculture supernatant inhibits K. pneumoniae growth in vitro. K. pneumoniae was added to the platelet-K. Pneumoniae coculture supernatant. Sn +KP: K. pneumoniae cocultured with platelet-K. pneumoniae coculture supernatant. S_PLT + KP: K. pneumoniae cocultured with supernatant of platelet lysate, as a control group. KP: untreated K. pneumoniae, as a negative control. IPM: K. pneumoniae treated with Imipenem, as a positive control. (A) The growth curve of K. pneumoniae cocultured with or without platelet-K. pneumoniae coculture supernatant for 24 h, according to OD_600nm. (B) The counts of K. pneumoniae in each group at 2, 4, 6, 8, and 10 h. (C) The photos of K. pneumoniae in each group at 8 h. (D) Analysis of OD_600nm in each group at 8 h. (E) The counts of K. pneumoniae in each group at 8 h. (F) Statistical results of the counts in (E). All results have been tested at least three times. Statistical analysis of the data was performed by one-way analysis of variance (ANOVA), Tukey multiple comparisons.

To explore whether the antibacterial effect of the supernatant was mainly due to platelet activation after co-culture of platelets with KP, we measured the platelet activation rate after co-culture of platelets with K. pneumoniae. The results showed that platelet activation was significantly increased after coculture of platelets with K. pneumoniae (Fig. S3A and B). In addition, when platelet activation was inhibited by the antiplatelet drugs Aspirin and Ticagrelor (Fig. S3C through F), the antibacterial effect of the supernatant of platelet-K. pneumoniae co-culture was reduced (Fig. S3G). These results demonstrate that the platelet-K. pneumoniae coculture supernatant can directly inhibit K. pneumoniae growth in vitro and platelet activation is the main reason for the inhibitory effect of the platelet-K. pneumoniae coculture supernatant.

Platelet-K. pneumoniae coculture supernatant inhibits K. pneumoniae growth in a murine infection model

To confirm our findings in vivo, we aimed to replicate the inhibition of K. pneumoniae growth by the platelet-K. pneumoniae coculture supernatant in a murine infection model. First, C57BL/6 mice were infected with K. pneumoniae via transtracheal injection to establish a K. pneumoniae infection murine model (Fig. S4A). The body weight, number of white blood cells (WBCs), PLTs, and level of hemoglobin (Hb) of mice decreased after injecting K. pneumoniae (Fig. S4B through E). Moreover, the number of K. pneumoniae in the lungs and bronchoalveolar lavage fluid (BALF) of K. pneumoniae-infected mouse group was markedly higher than that in the PBS-treated mouse group (Fig. S4F through H). The mice were then euthanized 24 h after infection, and the mouse lungs were analyzed using hematoxylin and eosin staining. The results showed that there were more leukocytes in the lungs of the K. pneumoniae-infected group, and the lung histological injury score (34, 35) was higher than that of the control group (Fig. S4I and J). These data suggested that the K. pneumoniae infection murine model was successfully established in this study.

To determine whether the administration of platelet-K. pneumoniae coculture supernatant inhibits K. pneumoniae growth and relieves K. pneumoniae infection in mice, the supernatant obtained from the platelet-K. pneumoniae coculture was transfused into the K. pneumoniae-infected mice after 24 h infection (Fig. 2A). We observed the changes in body weight, WBCs, PLTs, and Hb in mice within 48 h (Fig. S5A through D) and found that the infection symptoms in mice were alleviated following treatment with the platelet-K. pneumoniae coculture supernatant. The body weight of mice in the K. pneumoniae-infected and supernatant-treated groups was lower than that in the PBS group (Fig. S5E). The counts of WBCs in the K. pneumoniae-infected and supernatant-treated groups significantly decreased compared to that in the PBS group (Fig. S5F). Meanwhile, PLTs count in the supernatant-treated group was lower than that in the PBS and K. pneumoniae group (Fig. S5G). Finally, there was no difference in the levels of Hb across the three groups (Fig. S5H). The mice were euthanized 48 h after K. pneumoniae infection, and we observed that the entire lungs of all mice from the K. pneumoniae-infected mouse group exhibited red appearance, indicating a significant bleeding in the lungs. However, lung bleeding was alleviated after treatment with the platelet-K. pneumoniae coculture supernatant (Fig. 2B). Notably, the number of K. pneumoniae in the lungs and BALF was significantly reduced after 48 h infection (Fig. 2C and D; Fig. S5I). Additionally, damage to the lungs was also reduced in mice from the supernatant-treated group, as assessed from pathological examination (Fig. 2E and F). Moreover, the analysis of BALF cells using Swiss Giemsa staining showed significant changes in the WBC and red blood cell (RBC) counts, indicating that the inflammatory cell infiltration and hemorrhage in the mouse lungs were alleviated (Fig. 2G through I). Furthermore, the cytokine levels in BALF indicated that the release of proinflammatory cytokines in the supernatant-treated mouse group was reduced compared with that in the K. pneumoniae group (Fig. 2J through M). These results suggest that the platelet-K. pneumoniae coculture supernatant can inhibit K. pneumoniae growth in vivo.

Fig 2.

Platelet-K. pneumoniae coculture supernatant transfusion alleviates the pulmonary symptoms of K. pneumoniae-infected mice. C57BL/6 mice were infected with K. pneumoniae by intratracheal injection and then treated with or without platelet-K. pneumoniae coculture supernatant (n = 5). PBS: mice were injected intratracheal with PBS as a negative control group. KP: mice were injected intratracheal with K. pneumoniae. Sn +KP: mice were injected with platelet-K. pneumoniae coculture supernatant after K. pneumoniae infection. (A) Schematic diagram of platelet-K. pneumoniae coculture supernatant transfusion in K. pneumoniae infected mice. (B) Lungs of mice in each group. (C) Statistical histogram of bacteria count in lung. (D) Statistical histogram of bacteria count in BALF. (E) Representative images of H&E staining of lung. (F) Statistical histogram of histological score analysis of lung injury in the mice in (E). (G) Representative images of Swiss staining analysis of cell precipitates in BALF. (H) Statistical histogram of the RBC count of BALF in each group. (I) Statistical histogram of the WBC count of BALF in each group. (J) Statistical histogram of IL-6 level in the serum of the mice in each group. (K) Statistical histogram of IL-6 level in the BALF of the mice in each group. (L) Statistical histogram of TNF-α level in the BALF of the mice in each group. (M) Statistical histogram of IL-1β level in the BALF of the mice in each group. All results have been tested at least three times. Statistical analysis of the data was performed by one-way ANOVA, Tukey multiple comparisons.

Platelet-K. pneumoniae coculture supernatant induces oxidative stress damage and apoptosis-like death in K. pneumoniae

As the platelet-K. pneumoniae coculture supernatant was found to exert inhibitory effects on K. pneumoniae in vivo and in vitro, we sought to reveal the potential mechanism underlying this process. To this end, we first used scanning electron microscopy to observe the ultrastructure of the bacterium. When K. pneumoniae was cultured with the platelet-K. pneumoniae coculture supernatant in the culture medium, its morphology became significantly distorted and shrunken (Fig. 3A). Moreover, as the formation of biofilm is part of the dynamic process of bacterial growth and biofilm can enhance the resistance of bacteria to host defense and antibiotics (36–38), we used scanning electron microscopy to observe biofilm formation in the K. pneumoniae cultures. We found that biofilm formation was significantly reduced in the supernatant-treated group compared with that in the K. pneumoniae control group (Fig. 3B).

Fig 3.

Platelet-K. pneumoniae coculture supernatant induces apoptosis-like death of K. pneumoniae. K. pneumoniae were cocultured with or without platelet-K. pneumoniae coculture supernatant for 6 h. (A and B) The ultrastructure of K. pneumoniae was observed by scanning electron microscopy. (A) Scale bars: 5 µm (upper panel); 3 µm (lower panel). The original magnification was 10,000 (upper panel), 20,000 (lower panel) in each group. (B) Scale bars: 10 µm (upper panel); 5 µm (lower panel). The original magnification was 5,000 (upper panel), 10,000 (lower panel) in each group. (C) The ultrastructure of K. pneumoniae was observed by transmission electron microscopy. Scale bar: 500 nm (upper panel); 200 nm (lower panel). The original magnification was 20,000 (upper panel), 40,000 (lower panel) in each group. (D) PS exposure was detected by flow cytometry. (E) Statistical results of FITC-Annexin V fluorescence intensity in (D). (F) DiBAC₄(3)-labeled bacterial membrane potential was detected by flow cytometry. (G) Statistical results of FITC-DiBAC₄(3) fluorescence intensity in (F). (H) V-A-D-FMK labeled intracellular caspase of bacteria was detected by FACS analysis. (I) Statistical results of V-A-D-FMK fluorescence intensity in (H). (J) DNA damage of K. pneumoniae was detected by FACS analysis with TUNEL. (K) Statistical results of fluorescence intensity in (J). (L) FITC-DCFH-DA labeled intracellular ROS of K. pneumoniae was detected by flow cytometry. (M) Statistical results of fluorescence intensity in (L). (N) Validation of RNA-sequence results by qRT-PCR. Naked: unstained K. pneumoniae. KP: untreated K. pneumoniae, as a negative control. Sn +KP: K. pneumoniae treated with platelet-K. pneumoniae coculture supernatant. PC: K. pneumoniae treated with MMC (5 µg/mL) for 6 h, as a positive control. All results have been tested at least three times. Statistical analysis of the data was performed by one-way ANOVA, Tukey multiple comparisons.

Our previous study has shown that platelets can induce DNA damage in S. aureus and methicillin-resistant S. aureus. Therefore, we hypothesized that the supernatant of platelet-K. pneumoniae coculture may also inflict DNA damage in K. pneumoniae. Indeed, we observed changes in the morphology and internal structure of K. pneumoniae using transmission electron microscopy after the treatment of K. pneumoniae cultures with the platelet-K. pneumoniae coculture supernatant. As shown in Fig. 3C, we observed an evident agglutination of K. pneumoniae chromatin as well as irregular eversion and blurring of the cell wall and membrane after treatment with the platelet-K. pneumoniae coculture supernatant. These morphological outcomes showed that the cell wall, cell membrane, and cell nucleus of K. pneumoniae were damaged following supernatant treatment.

Antibiotic treatment can induce cell apoptosis and cause the targeted cells to exhibit characteristic markers of apoptosis, including phosphatidylserine (PS) exposure, chromosome condensation, and DNA fragmentation (29, 39). We found that the platelet-K. pneumoniae coculture supernatant had an antibiotic-like effect on K. pneumoniae in our previous study. Therefore, PS exposure in K. pneumoniae was assessed using Annexin V staining and flow cytometry, which confirmed the occurrence of apoptosis-like death in K. pneumoniae following treatment with the platelet-K. pneumoniae coculture supernatant. The percentages of Annexin V-positive K. pneumoniae in the supernatant-treated and control groups were 6.91% and 1.14%, respectively (Fig. 3D and E), suggesting that the platelet-K. pneumoniae coculture supernatant treatment can induce PS exposure in K. pneumoniae.

Next, the membrane potential of supernatant-treated K. pneumoniae was measured using the lipophilic anionic fluorescent dye DiBAC₄(3), which is sensitive to cellular membrane potential. Compared to that in the untreated K. pneumoniae group, the fluorescence intensity of DiBAC₄(3) in supernatant treatment group increased significantly (Fig. 3F and G), indicating that treatment with the supernatant induced depolarization of the K. pneumoniae membrane potential. We also assessed the activity of intracellular caspase in K. pneumoniae using V-A-D-FMK. Intracellular fluorescence indicated a stable binding of V-A-D-FMK-FITC to bacterial proteins that possess binding affinity for a general caspase substrate. The increase in fluorescence intensity reflects an increase in the concentration of these proteins. Results showed that the fluorescence intensity of V-A-D-FMK was enhanced in the supernatant-treated group, implying that the activity of intracellular caspase increased in K. pneumoniae after supernatant treatment (Fig. 3H and I).

DNA fragmentation is one of the most important indicators of apoptosis (39). In the previous morphological observations, we could see that the platelet-K. pneumoniae coculture supernatant caused chromatin condensation in K. pneumoniae; hence, we sought to evaluate DNA damage in K. pneumoniae using the terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay and flow cytometry. The fluorescence intensity of the labeled DNA fragments increased significantly after treatment with the platelet-K. pneumoniae coculture supernatant, further validating that DNA damage, indeed, occurred in K. pneumoniae after treatment with the supernatant (Fig. 3J and K).

Reactive oxygen species (ROS) are a product of biological respiration metabolism. When cells are exposed to external environmental pressures, oxidative stress occurs and excessive ROS are produced, adversely damaging the DNA and proteins in cells, eventually leading to cell death (40). Therefore, we assessed the levels of ROS using DCFH-DA-labeled probe and found that the fluorescence intensity of DCFH-DA in K. pneumoniae increased after treatment with the platelet-K. pneumoniae coculture supernatant (Fig. 3L and M). We detected the expression of ROS-related stress genes in K. pneumoniae and found that the expression of ndhO, ndhR, qoxC, and qoxD genes was upregulated and the azr gene was downregulated after platelet-K. pneumoniae coculture supernatant treatment (Fig. 3N). Increased expression of qoxC and qoxD suggests reduced electron supply to the oxidative respiratory chain, and the decreased expression of AZR indicates compromised electron supply to the oxidative respiratory chain (29, 41–43) .These results provide evidence that K. pneumoniae suffered oxidative stress following platelet-K. pneumoniae coculture supernatant treatment.

Additionally, we examined the levels of pH, lactate dehydrogenase (LDH), glucose, and other ions in K. pneumoniae cultures treated with the platelet-K. pneumoniae coculture supernatant to explore whether the changes in physical or chemical environments affect the growth of K. pneumoniae (Fig. S6A through G). Results showed that the levels of LDH, Na⁺, Cl⁻, CO₂, and pH increased, while those of K⁺ and Ca²⁺ decreased (Fig. S6H through N). These data indicated that the osmotic pressure in K. pneumoniae was altered following treatment with the platelet-K. pneumoniae coculture supernatant, which might partially explain the damage to the cell membrane and cell wall. Meanwhile, the increase in LDH levels further indicated that the metabolism of K. pneumoniae was gravely affected. Collectively, these data suggested that the platelet-K. pneumoniae coculture supernatant induced oxidative stress damage and apoptosis-like changes in K. pneumoniae. The imbalance of Na⁺, Cl⁻, K⁺, and Ca²⁺ concentrations and the increase in LDH levels also indicated that the homeostasis and metabolic process in K. pneumoniae were adversely impaired, potentially causing injury and death of the bacterial population.

N-acetylcysteine inhibits the oxidative stress injury and apoptosis-like death induced upon treatment with the platelet-K. pneumoniae coculture supernatant

We have demonstrated that the production of intracellular ROS increased and the apoptosis-like changes appeared following treatment of K. pneumoniae with the platelet-K. pneumoniae coculture supernatant. We hypothesized that apoptosis was triggered by the overproduction of ROS induced by antimicrobial peptides in coculture supernatant.

To confirm whether the increased ROS levels are related to the apoptosis-like death of K. pneumoniae after treatment with the platelet-K. pneumoniae coculture supernatant, the antioxidant N-acetylcysteine (NAC) was added into the coculture system, and the changes in PS exposure, membrane potential, and intracellular caspase activity of K. pneumoniae were assessed. After the production of ROS in the supernatant-treated K. pneumoniae cultures was inhibited by NAC (Fig. 4G and H), we discovered that the decline in ROS levels led to an increase in K. pneumoniae population in the cultures (Fig. S7A through C) as well as reduction in the PS exposure (Fig. 4A and B), cell membrane depolarization (Fig. 4C and D), and intracellular caspase activity (Fig. 4E and F) of K. pneumoniae. These results signified that the apoptosis-like changes in K. pneumoniae were attenuated because ROS production was inhibited. Taken together, the platelet-K. pneumoniae coculture supernatant appeared to induce apoptosis-like death in K. pneumoniae by promoting the overproduction of ROS (Fig. S7D).

Fig 4.

NAC inhibits the apoptosis of K. pneumoniae induced by platelet-K. pneumoniae coculture supernatant. K. pneumoniae were cocultured with or without platelet-K. pneumoniae coculture supernatant for 8 h, and 6 mM NAC was added to the coculture system. Naked: unstained K. pneumoniae. Sn +KP: K. pneumoniae treated with platelet-K. pneumoniae coculture supernatant. Sn +KP + NAC: NAC was added to the platelet-K. pneumoniae coculture supernatant cocultured with K. pneumoniae. KP + NAC: NAC was added to 1640 RPMI medium with K. pneumoniae. KP: untreated K. pneumoniae, as a negative control. PC: K. pneumoniae treated with MMC (5 µg/mL) for 6 h, as a positive control. (A) PS exposure was detected by flow cytometry. (B) Statistical results of FITC-Annexin V fluorescence intensity in (A). (C) Bacterial membrane potential of K. pneumoniae was detected by flow cytometry with DiBAC₄(3). (D) Statistical results of DiBAC₄(3) fluorescence intensity in (C). (E) Intracellular caspase of K. pneumoniae was detected by flow cytometry with V-A-D-FMK. (F) Statistical results of V-A-D-FMK fluorescence intensity in (E). (G) The ROS of K. pneumoniae was detected by flow cytometry with DCFH-DA.(H) Statistical results of DCFH-DA fluorescence intensity in (G). All results have been tested at least three times. Statistical analysis of the data was performed by one-way ANOVA, Tukey multiple comparisons.

Platelet-K. pneumoniae coculture supernatant exerts antimicrobial effects on CRKP in vitro and in vivo

To further explore whether the platelet-K. pneumoniae coculture supernatant exerts the same direct antibacterial effects on CRKP, we monitored the growth of CRKP after platelet-K. pneumoniae coculture supernatant treatment. First, clinical isolate strains were obtained, and the drug resistance and virulence genes of CRKP were determined. The results showed that the string test of CRKP was negative, the capsular serotype was K20, and the virulence genes were magA, fimH, and Aero (Fig. S9A through F). Then, the platelet-K. pneumoniae coculture supernatant was extracted and added to the CRKP cultures. The turbidity of the supernatant-treated CRKP group was found to be lower than that of the CRKP control group (Fig. S8A). In addition, the OD_600nm decreased following treatment with the platelet-K. pneumoniae coculture supernatant (Fig. S8B). The counts of CRKP populations in LB plates were also lower than those in the control group (Fig. S8C). These observations demonstrated that the platelet-K. pneumoniae coculture supernatant can directly inhibit CRKP growth in vitro.

To verify whether platelet-K. pneumoniae coculture supernatant transfusion can inhibit CRKP growth and relieve CRKP infection in mice, a CRKP-infected murine model was first established via the transtracheal injection of CRKP into mice. The platelet-K. pneumoniae coculture supernatant was then injected into the CRKP-infected mice via the tail veins. We monitored the changes in body weight, WBC counts, PLT counts, and Hb level in mice within 48 h (Fig. S10A through G) and found that the infection symptoms in the mice were alleviated after transfusion of the supernatant. We also observed pulmonary congestion and edema in mice following euthanasia 48 h after infection. The results showed that pulmonary congestion and edema in the mice were relieved (Fig. S10H), and the number of bacterial cells in the lungs and BALF was significantly reduced after supernatant treatment (Fig. S10I). We collected and centrifuged the BALF to evaluate the cell deposits at the bottom of the tubes and found that the deposits in the CRKP-infected group contained red blood cells, with the number of cells in the supernatant-treated group being slightly lower than that in the infected group (Fig. 5A). Moreover, the analysis of BALF cells in the WBC and RBC counts, indicating that the levels of inflammatory cell infiltration and hemorrhage in the mouse lungs were alleviated after supernatant transfusion (Fig. 5B and C). The proportions of granulocytes, T lymphocytes, and B lymphocytes in the cell deposits were then analyzed via flow cytometry; results showed that the proportion of CD11b⁺ cells in the CRKP group increased significantly, while the proportions of CD41⁺, CD3⁺, and B220⁺ cells in the supernatant-treated group were much higher than those in the CRKP group (Fig. 5D through G). These results indicate that the platelet-K. pneumoniae coculture supernatant might play an important role in mobilizing murine immune cells against CRKP infection. In addition, hematoxylin and eosin staining showed that inflammatory cell infiltration was reduced and that the pathological damage of alveolar tissue was improved after supernatant treatment (Fig. 5H and I). In the supernatant-treated group, the levels of proinflammatory cytokines (IL-6, TNF-α, and IL-1β) were lower than those in the CRKP-infected group (Fig. 5J). Taken together, these results demonstrate that platelet-K. pneumoniae coculture supernatant transfusion can inhibit CRKP growth and defend against CRKP infection in vivo.

Fig 5.

Platelet-K. pneumoniae coculture supernatant transfusion alleviates the the pulmonary symptoms of CRKP infected mice. C57BL/6 mice were infected with CRKP by intratracheal injection and then treated with or without platelet-K. pneumoniae coculture supernatant(n = 5). PBS: mice were injected with PBS into the trachea, as negative control group. CRKP: mice were injected with CRKP into the trachea. Sn+CRKP: mice were injected with platelet-K. pneumoniae coculture supernatant after CRKP infection. (A) Photos of cells in BALF in each group. (B) Statistical histogram of the WBC count in BALF of each group. (C) Statistical histogram of the RBC count in BALF of each group. (D) The detection of the percentage of CD41 positive cells by flow cytometry. (E) The detection of the percentage of CD11b positive cells by flow cytometry. (F) The detection of the percentage of B220 positive cells by flow cytometry. (G) The detection of the percentage of CD3 positive cells by flow cytometry. (H) Representative images of H&E staining of lung. (I) Statistical histogram of histological score analysis of lung injury in the mice in (H). (J) Statistical histogram of IL-6, TNF-α, and IL-1β levels in the BALF of the mice in each group. All results have been tested at least three times. Statistical analysis of the data was performed by one-way ANOVA, Tukey multiple comparisons.

DISCUSSION

In the present study, we successfully demonstrated that the platelet-K. pneumoniae coculture supernatant can inhibit the growth of K. pneumoniae both in vivo and in vitro. These findings reveal a novel role of platelets in combating the infections caused by K. pneumoniae and its drug-resistant strain, CRKP. In addition, this study lays a foundation for future studies on the transfusion of platelets or their derivatives and the development and application of new antibacterial agents.

Platelets are complex cells containing three different types of granules: the α, dense or δ, and lysosomal granules. The α-granules carry diverse proteins, cytokines, chemokines, and growth factors, while the δ-granules contain small molecules such as adenosine diphosphate, serotonin, glutamate, histamine, and calcium, which are necessary for hemostasis (44). In the present study, we found that the changes in bacterial turbidity in the cultures of K. pneumoniae treated with platelet-K. pneumoniae coculture supernatant were significantly different from the changes caused by platelet-S. aureus coculture supernatant in the corresponding S. aureus cultures. Therefore, we hypothesized that antimicrobial proteins may have formed by the platelets against K. pneumoniae in the platelet-K. pneumoniae coculture supernatant and the production of these antimicrobial peptides depended on platelets activation and it may be closely related to the special translation mode of platelets (45, 46). This study clearly demonstrated that treatment with the platelet-K. pneumoniae coculture supernatant had a significant inhibitory effect on K. pneumoniae in vitro and may play an important role in ameliorating K. pneumoniae-induced bacterial pneumonia.

In this study, treatment with the platelet-K. pneumoniae coculture supernatant significantly increased ROS production and induced apoptosis-like death in K. pneumoniae. The inhibition induced by the platelet-K. pneumoniae coculture supernatant in K. pneumoniae cultures observed in the present study might be the combined outcome of bacterial cell membrane damage and DNA damage. In the process of mediating apoptosis-like changes in bacteria, ROS production serves not only as one of the apoptotic indicators but also as a key factor in the occurrence of apoptosis (47). A number of studies have confirmed that bacteria produce excessive ROS to interfere with the bacterial oxidative phosphorylation and tricarboxylic acid cycle (48), especially important regulators of highly harmful apoptosis, which can directly damage DNA, lipids, and proteins (33). In addition, excessive ROS can cause cell membrane depolarization, eventually leading to apoptosis-like death in bacteria (39, 47). As mentioned previously, platelet-derived antimicrobial peptides present in the platelet-K. pneumoniae coculture supernatant may play an important role in inhibiting K. pneumoniae growth, changing the ionic environment of bacterial survival and causing oxidative stress damage in bacteria (Fig. S6). Previous reports have also revealed that antimicrobial peptides, such as cationic host defense peptides, contain positive charges and can bind to anionic bacterial membranes, causing lipid clustering and increasing membrane permeabilization (45, 46), thereby inhibiting bacterial proliferation. In addition, combined treatment with NAC inhibits the antibacterial activities of daunomycin, moxifloxacin, and oxacillin against S. aureus (29, 48). Our study supports the hypothesis that ROS play a key role in the supernatant-mediated apoptosis-like changes in K. pneumoniae. The inhibitory effect of the platelet-K. pneumoniae coculture supernatant on K. pneumoniae is particularly similar to the effect of antibiotics (39, 49). More importantly, the platelet-K. pneumoniae coculture supernatant also exerts the same inhibitory effect on CRKP in vivo and in vitro. In addition, transfusion of platelet-K. pneumoniae coculture supernatant alleviated the symptoms of pneumonia in K. pneumoniae- and CRKP-infected mice. These findings are of great significance for the research and development of antimicrobial agents in the future and would provide novel insights into potential clinical treatment of patients with refractory or drug-resistant infections.

Although our study provides evidence that the supernatant of platelet-K. pneumoniae coculture mediates the inhibition of K. pneumoniae growth, it has several limitations. Although we noted that the main molecules that are released from platelets upon their activation (including β-defensin 2, thymosin β4, CXCL4, CXCL7, CXCL12, and TGF-β, all of which have potent antibacterial effects) have already been reported (50–54), the exact composition of the platelet-K. pneumoniae coculture supernatant and the components that were associated with antibacterial activity in our study are yet to be elucidated. Additionally, although the transfusion of platelet-K. pneumoniae coculture supernatant alleviated infection symptoms in the K. pneumoniae-infected mice model, our observations do not preclude the participation of neutrophils and other immune cells in producing the observed outcome in vivo. Despite the key role that the supernatant plays in the anti-K. pneumoniae effect and the induction of ROS overproduction, the detailed mechanism underlying the platelet-K. pneumoniae coculture supernatant-induced apoptosis-like changes in K. pneumoniae remains unclear, and this is a research direction we would like to undertake in the future.

In conclusion, Our results demonstrated that the platelet-K. pneumoniae coculture supernatant directly inhibited the growth of K. pneumoniae and induced apoptosis-like death in K. pneumoniae. Additionally, we showed the overproduction of ROS-mediated oxidative stress damage in K. pneumoniae. The inhibitory effect of the platelet-K. pneumoniae coculture supernatant on drug-resistant strain CRKP was also shown in our study. As platelets are related to infection and inflammation, the antimicrobial function that platelets may perform cannot be ignored, but more research is needed to determine whether the platelet-bacteria coculture supernatant exhibits similar antimicrobial effects even in the presence of other pathogenic microorganisms in such a coculture setup. It is imperative that our understanding of the functions of platelets and platelet-related derivatives be improved, as this may provide alternative approaches for improving antimicrobial therapy, developing novel antimicrobial agents, controlling bacterial diseases, and solving the emerging crisis of antibiotic resistance.

MATERIALS AND METHODS

Preparation of human washed platelets

Human apheresis platelets were collected from 8 donors, centrifuged at 213g for 5 min, and then washed gently with washing solution (PBS containing 10% ACD solution) twice. Residual plasma and red blood cells were removed after washing, and the platelets were resuspended to achieve a count of 200 × 10⁹/L in 1640 RPMI medium (Gibco). The purity of the washed platelets was detected by blood autoanalyzer (XP-100, Sysmex, Kobe, Japan) and flow cytometry. All relevant procedures were performed in accordance with the requirements of the Xijing Hospital Ethics Committee (KY20212120-C-1).

Preparation of mice washed platelets

C57BL/6 mice, 6- to 8-week-old, were from the Animal Center of the Fourth Military Medical University. All animal experiments were conducted in accordance with the regulations on animal protection and management of the Fourth Military Medical University (IACUC-20130014). In this study, blood was collected from the hearts of mice and stored in PBS solution (contained 10% ACD solution). The blood was centrifuged twice (150g each time) for 5 min in order to remove the white and red blood cells. The upper layer of the fluid contained platelets, which are then transferred to a new sterile tube and washed twice. Finally, platelet precipitates were collected and suspended in 1640 RPMI medium with a count of 200 × 10⁹ /L.

Preparation, culture, and counting of bacteria

K. pneumoniae (ATCC number: 700603, Manassas, VA, USA) was inoculated on LB agar plates at 37°C for 24 h in order to obtain the single colony. Then, the single colony of K. pneumoniae was inoculated in LB and cultured at 37°C for 6–8 h to logarithmic growth phase.

In vitro experiments, K. pneumoniae was washed with sterile PBS for three times, and the concentration of bacteria was determined by a spectrophotometer (UV-2550; Shimadzu, Kyoto, Japan) at 600 nm wavelength (OD_600nm). Finally, K. pneumoniae was diluted to 1 × 10⁵ CFU/mL for testing. The numbers of bacteria were determined through serial dilution and plating assay. Five tubes were prepared, and 900 µL sterile PBS was added to each tube. Then, added 100 µL of the bacterial solution (1 × 10⁹ CFU/mL) to tube 1, mixed it and absorbed 100 µL into tube 2, diluted it to tube 5 in turn. One hundred microliters of bacterial solution was for plating, and the bacteria were incubated at 37℃ for 20 h. In vivo experiments, K. pneumoniae was washed with sterile PBS three times to prevent aggregation and then suspended with PBS to the count of 1 × 10⁸ CFU/mL.

CRKP were isolated from the clinical samples in laboratory department of Xijing Hospital of Fourth Military Medical University. CRKP was cultured and counted in the same way as K. pneumoniae in vitro studies, and it was diluted to 1 × 10⁵ CFU/mL for testing. In vivo studies, CRKP was washed with sterile PBS three times to prevent aggregation and then suspended with PBS to the count of 1 × 10⁸ CFU/mL. The drug resistance of clinically isolated CRKP was detected by disc diffusion test, and the presence of drug resistance genes and virulence factors was determined by PCR. In addition, tigecycline and imipenem were used to confirm drug resistance of the CRKP strains. The MICs of tigecycline for CRKP clinical isolated strains were detected by broth microdilution method in Mueller-Hinton (MH) medium. First, 180 µL of diluted bacterial solution (1 × 10⁵ CFU/mL) was added to the tube 1, 100 µL of diluted bacterial solution was added from tube 2 to 11, and sterile MH medium was added to the tube 12 as a control. Then, added 20 µL of the drug solution to tube 1, mixed it and absorbed 100 µL into tube 2, diluted it to tube 11 in turn, respectively. Finally, the drug concentration was 1,280, 640, 320, 160, 80, 40, 20, 10, 5, 2.5, and 1.25 µg/mL. The bacteria were incubated at 37°C for 20 h, and MIC was the lowest drug concentration to completely inhibit the growth of bacteria.

Preparation of platelet-K. pneumoniae coculture supernatant

In vitro and vivo study, K. pneumoniae (1 × 10⁸ CFU/mL) and human (or mice) platelets (200 × 10⁹ /L) were mixed and cocultured in 1,640 RPMI medium at 37°C for 6–8 h under shaker. The coculture of K. pneumoniae and platelets was centrifuged at 10,000g for 10 min after 6–8 h of incubation. The supernatant of the coculture was then transferred to a sterile tube, and the supernatant containing secreted platelets compounds was filtered using a sterile filter (0.2 µm pore size) to eliminate all cells. The supernatant was incubated again with the same bacteria for 6 h at 37°C under shaker.

Mice model of K. pneumoniae and CRKP infection

K. pneumoniae or CRKP (100 µL, 1 × 10⁸ CFU/mL) was injected into the trachea of mice, and the blood was collected through the orbital veins at 0 and 24 h in each group. The volume of the blood collected each time was 30 µL, and it was stored in a PBS solution (containing 20% ACD solution) for analysis of WBC, and platelet counts using the blood autoanalyzer and the number of bacteria in BALF and lungs was counted using LB plates. The mice were euthanized 24 h after infection, and the lungs were collected for the analysis by HE staining.

Transfusion of platelet-K. pneumoniae coculture supernatant into mice

C57BL/6 mice (6- to 8-week-old) were divided into PBS, KP/CRKP, and supernatant transfusion (Sn + KP/Sn + CRKP) groups (5 mice each group, 3 female, and 2 male). In the supernatant transfusion group, platelet-K. pneumoniae coculture supernatants (300 µL) were transfused via tail vein injection after K. pneumoniae/CRKP infection for 24 h, and the blood was collected through the vein at 0, 24, and 48 h. WBCs and PLTs in the peripheral blood of mice were collected at different times and detected by the blood autoanalyzer.

The mice were euthanized 48 h after K. pneumoniae/CRKP infection, and the needle of a syringe was inserted into the trachea of each mouse; we instilled PBS into the mouse lung and gently drew back fluid after all lobes of lung are inflated. Re-instilled the returned fluid into mouse lung and collected the first 1 mL return and put on ice, repeated at least for three times. The lungs were collected for HE staining. In each tissue sample, five random areas were scored, and the mean value was calculated by the modified scoring system described by Wang et al. and Hasan et al. (34, 35). The histology score was the sum of the following four parameters: size of alveolar spaces, thickness of alveolar septa, alveolar fibrin deposition, and neutrophil infiltration (0: absent and appears normal; 1: light; 2: moderate; 3: strong; and 4: intense; total score is 16). The bacteria count in the lungs and BALF were analyzed by using LB plates. WBCs and PLTs in mice BALF were detected by the blood autoanalyzer and flow cytometry, cell precipitates were visualized by Swiss Giemsa staining, while cytokines in BALF and serum were measured by ELISA.

Morphological analysis

To analyze the ultrastructure of K. pneumoniae, the samples were fixed with 2.5% and 3% glutaraldehyde respectively at 4°C for at least 12 h for slicing. The samples were sent to the center of electron microscopy in Fourth Military Medical University. Sections were examined with transmission electron microscopy (JEM-1230; Jeol USA, Peabody, MA, USA) and scanning electron microscopy (Hitachi S-3400N, Japan), and digital images were captured with a CCD camera (Olympus, Tokyo, Japan).

Biofilm formation

The platelet-K. pneumoniae coculture supernatant incubated with K. pneumoniae in 24-well plates, and sterile cell slides were placed in each well. The cocultured liquid was incubated at 37°C for 7 days. 1640 RPMI medium was gently discarded, and an equal volume of fresh 1640 RPMI medium was slowly added along the side wall of the well every 24 h. Finally, the medium was discarded and washed once with sterile PBS; then, 1 mL of 3% glutaraldehyde was added and fixed samples at 4°C for 12 h. The all samples were sent to the center of electron microscopy in Fourth Military Medical University.

Flow cytometry analysis

Analysis of platelet activation by flow cytometry

Briefly, platelet was cultured with K. pneumoniae as described above. Then, (1 µL each) anti-Human CD62P (BD Biosciences, USA) were added into 19 µL platelets for staining and incubated in darkness at 22°C for 15 min. After incubation, washed the platelets with buffer and suspended in 300 µL of buffer. Finally, the sample was analyzed by flow cytometry.

Analysis of phosphatidylserine exposure about K. pneumoniae

K. pneumoniae was cultured and washed as described above. The samples were diluted to 1:100 with 1 × binding buffer. Then, Annexin V and PI (1 µL each) (FITC Annexin V Apoptosis Detection Kit I, BD Biosciences, USA) were added into 100 µL diluted cells for staining and incubated in darkness at 22°C for 15 min. After incubation, washed the cells with 1 × binding buffer and suspended in 500 µL of 1 × binding buffer. Finally, the sample was analyzed by flow cytometry.

Detection of membrane depolarization

DiBAC₄(3) (Sigma-Aldrich, St. Louis, MO, USA) was used to detect the membrane potential of K. pneumoniae. K. pneumoniae were cultured and washed as described above. DiBAC₄(3) (0.1 µg/µL, 1 µL) was added to 1 mL diluted cells for staining. The cells were then incubated in darkness at room temperature for 30 min. After staining, the FITC-DiBAC₄(3) fluorescence intensity of the samples was analyzed by flow cytometry.

Analysis of bacterial proteins with caspase-like substrate affinity. FITC-conjugated pan-caspase inhibitor peptide, Z-V-d-FMK (Intracellular Caspase Detection ApoStat, USA), was used to detect whether platelet-K. pneumoniae coculture supernatant could induce the expression of bacterial proteins that can bind to caspase substrate peptides. The cells of K. pneumoniae were cultured and washed as described above. Then, 10 µL Z-V-d-FMK-FITC was added into 1 mL diluted cells and incubated at 37°C in the dark for 30 min. Next, the cells were washed with PBS and suspended with 500 µL PBS. Finally, the fluorescence intensity of Z-V-d-FMK-FITC was analyzed by flow cytometry.

Detection of K. pneumoniae DNA fragmentation

K. pneumoniae were added to platelet-K. pneumoniae coculture supernatant for 8 h. The samples were collected and DNA fragmentation was detected by TUNEL assay (In Situ Cell Death Detection Kit, Fluorescein, Roche, Mannheim, Germany). All steps were performed according to the manufacturer’s instructions and analyzed by flow cytometry.

Analysis of reactive oxygen species (ROS) by flow cytometry

K. pneumoniae were added to platelet-K. pneumoniae coculture supernatant for 8 h. The samples were collected, and the ROS level was detected by DCFH-DA (5 µM, KeyGen Biotech, Nanjing, China). All steps were performed according to the manufacturer’s instructions and analyzed by flow cytometry.

RNA analysis

After 8 h of coculture with or without platelet-K. pneumoniae coculture supernatant, RNA was extracted from K. pneumoniae using the RNeasy RNA isolation kit (Qiagen, Dusseldorf, Germany). The expression of five genes related to oxidative phosphorylation, viz. ndhO, ndhR, qoxC, qoxD, azr, was analyzed using qRT-PCR; 16S rRNA was used as the housekeeping gene. The sequences of primers used for analyzing other genes were as follows:

16S rRNA, F (5′-3′): ACGTGGATAACCTACCTATAAGACTGGGAT

R (5′-3′): TAACCTTACCAACTAGCTAATGCAGCG

ndhO, F (5′-3′): TGCGACCATGTTTCTTTGCG

R (5′-3′): GCCACGCTTACCATTTGACC

ndhR, F (5′-3′): CGTCGTTAATGCAGGCTGATG

R (5′-3′): CACGAAACGCATTGACTGGA

qoxC, F (5′-3′): CATCCGCTATACACCATCCCT

R (5′-3′):TCGCTAGGTATCGTTTGGGC

qoxD, F (5′-3′): TAAGTGTGAAGAGTGACCGCC

R (5′-3′): TTGGCTTTGCATTCGTCCAAG

azr, F (5′-3′): TTCAGTTTTATGCGGTTCGGC

R (5′-3′): TGCTGAAGGACCACAAGGTTT

Statistical analysis

Statistical analyses were performed using GraphPad software (Prism v.5.01 La Jolla, CA, USA), and the results of flow cytometry were analyzed using FlowJoV7.6 software. All data are presented as mean ± standard error of the mean. The differences between the mean variables of the two groups were determined using Student’s t test. One-way analysis of variance (ANOVA) with Tukey’s post hoc test was used for multiple group analyses. Statistical significance was set at P＜0.05.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/spectrum.01279-23.

Supplemental figures. spectrum.01279-23-s0001.pdf.

Fig. S1 to S10.

OPEN PEER REVIEW. reviewer-comments.pdf.

An accounting of the reviewer comments and feedback.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.