Abstract
Background
The innate immune system can recall previous immunologic challenges and thus respond more effectively to subsequent unrelated challenges, a phenomenon called trained immunity. Training the innate immune system before surgery might be a potential option to prevent bone and joint infection.
Questions/purposes
(1) Does the training process cause adverse effects such as fever or organ injury? (2) Does training the innate immune system confer broad-spectrum protection against bone and joint infection in a mouse model? (3) Does trained immunity remain effective for up to 8 weeks in this mouse model?
Methods
After randomization and group information blinding, we trained the innate immune system of C57BL/6 mice (n = 20 for each group) by intravenously injecting them with either 0.1 mg of zymosan (a toll-like receptor 2 agonist), 0.1 mg of lipopolysaccharide (a toll-like receptor 4 agonist), or normal saline (control). For assessing the host response and possible organ injury after training and infection challenge, we monitored rectal temperature, collected blood to determine leukocyte counts, and performed biochemical and proinflammatory cytokine analyses. After 2 weeks, we then assessed whether trained immunity could prevent infections in an intraarticular implant model subjected to a local or systemic challenge with a broad spectrum of bacterial species (Staphylococcus aureus, Escherichia coli, Enterococcus faecalis, Streptococcus pyogenes, or Pseudomonas aeruginosa) in terms of culture-positive rate and colony counts. The proportion of culture-positive joint samples from trained and control groups were compared after 4 weeks. Finally, we increased the interval between training and bacterial challenge up to 8 weeks to assess the durability of training efficacies.
Results
Training with zymosan and lipopolysaccharide caused mild and transient stress in host animals in terms of elevated rectal temperature and higher blood urea nitrogen, creatinine, alanine aminotransferase, and aspartate aminotransferase levels. Trained mice had fewer culture-positive joint samples after local inoculation with S. aureus (control: 100% [20 of 20]; zymosan: 55% [11 of 20], relative risk 0.55 [95% CI 0.37 to 0.82]; p = 0.001; lipopolysaccharide: 60% [12 of 20], RR 0.60 [95% CI 0.42 to 0.86]; p = 0.003) and systemic challenge with S. aureus (control: 70% [14 of 20]; zymosan: 15% [3 of 20], RR 0.21 [95% CI 0.07 to 0.63]; p = 0.001; lipopolysaccharide: 15% [3 of 20], RR 0.21 [95% CI 0.07 to 0.63]; p = 0.001) than controls. We observed similar patterns of enhanced protection against local and systemic challenge of E. coli, E. faecalis, S. pyogenes, and P. aeruginosa. Zymosan-trained mice were more effectively protected against both local (control: 20 of 20 [100%], zymosan: 14 of 20 [70%], RR 0.70 [95% CI 0.53 to 0.93]; p = 0.02) and systemic (control: 70% [14 of 20]; zymosan: 30% [6 of 20], RR 0.43 [95% CI 0.21 to 0.89]; p = 0.03) challenge with S. aureus for up to 8 weeks than controls.
Conclusions
Trained immunity confers mild stress and broad-spectrum protection against bone and joint infection in a mouse model. The protection conferred by immunity training lasted up to 8 weeks in this mouse model. The results of the current research support further study of this presurgical strategy to mitigate bone and joint infection in other large animal models.
Clinical Relevance
If large animal models substantiate the efficacy and safety of presurgical immunity training-based strategies, clinical trials would be then warranted to translate this strategy into clinical practice.
Introduction
The host’s immune system is typically divided into an innate system and an adaptive immune system. The innate immune system responds rapidly and nonspecifically to pathogens [4]. In contrast, the adaptive immune system responds more slowly and specifically to pathogens, and depends more on immunologic memory; that is, specific recognition of an antigen the host had previously encountered [4]. Conventional wisdom about the way the immune system is organized asserts that immunologic memory is limited in the adaptive immune system because it is molded by antigen-specific responses [25].
Recently, however, emerging facts have greatly challenged this concept [17]. Evidence for immunologic memory in the innate immune system was first discovered in invertebrates and plants [13, 21]. In humans, the existence of a form of innate immune memory was demonstrated in patients who acquired nonspecific and long-term protection against certain diseases after receiving Bacillus Calmette-Guérin, smallpox, and measles vaccines [2, 5, 15]. The term trained immunity was proposed to describe this kind of phenomenon, wherein the innate immune system recalls a previous challenge, enabling it to respond more effectively to another similar or dissimilar microbial stimulus in the future [11]. The idea of a trained innate immune system has also been demonstrated in multiple preclinical models using heat-killed Candida albicans, β-glucans, zymosan, and Bacillus Calmette-Guérin vaccination [1, 6, 18]. In these animal models, the innate immune system was successfully trained to improve its response to a second challenge. More recently, a Phase 1 clinical trial demonstrated that Bacillus Calmette-Guérin vaccination might improve the clinical and immunologic response against malaria in humans [23].
Training the innate immune system before elective surgery might be a potential option to prevent bone and joint infection. The main advantage of training the innate immune system is that trained immunity is nonspecific and thus theoretically effective against all types of pathogens. Additionally, to determine an appropriate timing for surgery after training, it is essential for to investigate how long the transient stress and protection will last.
Thus, this prompted us to ask the following research questions: (1) Does the immune training process cause adverse effects such as fever and organ injury? (2) Does training the innate immune system confer broad-spectrum protection against bone and joint infection in a mouse model? (3) Does trained immunity remain effective for up to 8 weeks in this mouse model?
Materials and Methods
All animal experiments and animal welfare were conducted in accordance with protocols approved by the institutional animal care and use committee of our hospital. The study was designed and performed according to the Animal Research: Reporting In Vivo Experiments guidelines [12].
Experimental Overview
Randomization and group information blinding were used to reduce the bias of this study. We trained the innate immune system of C57BL/6 mice (n = 20 for each group) by intravenously injecting them with either 0.1 mg of zymosan (a toll-like receptor 2 agonist), 0.1 mg of lipopolysaccharide (a toll-like receptor 4 agonist), or normal saline (control). We assessed the host response and possible organ injury after training by monitoring rectal temperature, collecting blood to determine leukocyte counts, and performing biochemical and proinflammatory cytokine analyses (Fig. 1A). After 2 weeks, we then assessed whether trained immunity could prevent infections in an intraarticular implant model subjected to a local or systemic challenge with five most common bacterial species in terms of culture-positive rate and colony counts (Fig. 1A). Finally, we increased the interval between training and bacterial challenge up to 8 weeks to assess the durability of training efficacies (Fig. 1B).
Fig. 1.
A-B This schematic illustration shows the study design. (A) The mice were trained with 0.1 mg of zymosan, or 0.1 mg of lipopolysaccharide. After 2 weeks, the mice were challenged with Staphylococcus aureus, Escherichia coli, Enterococcus faecalis, Streptococcus pyogenes, or Pseudomonas aeruginosa. (B) The mice were trained with 0.1 mg zymosan and challenged with S. aureus. Different interval between training and bacterial challenge (2, 4, 6, and 8 weeks) were used to assess the durability of training efficacies; ALT = alanine transaminase; AST = aspartate transaminase; BUN = blood urea nitrogen; Cre = creatinine; IL-1 = interleukin-1; TNF-α = tumor necrosis factor α; IL-6 = interleukin-6; CXCL-1 = C-X-C motif chemokine ligand-1. A color image accompanies the online version of this article.
Immunity Training and Stress Assessment after Training
Female C57BL/6 mice aged 8 weeks (Laboratory Animal Center of Shanghai Jiaotong University) were housed in a temperature- and humidity-controlled environment (20° to 23° C; 50%) on a 12-hour light-dark schedule. Free access to food and tap water was provided. All mice were bred and maintained under specific pathogen-free conditions at an animal facility accredited by the American Association for the Accreditation of Laboratory Animal Care at Johns Hopkins and housed according to procedures described in the Guide for the Care and Use of Laboratory Animals [7]. We trained the immune system of the mice by intravenously injecting either 0.1 mg of zymosan (a toll-like receptor 2 agonist) or 0.1 mg of lipopolysaccharide (a toll-like receptor 4 agonist), 14 days before surgery, unless specified otherwise. These doses were based on the doses used in previous studies [6, 20]. Blood was collected from the infraorbital veins of mice 12 hours after the intravenous injection of experimental training substances. Whole blood was centrifuged at 1500 g for 10 minutes to separate the serum. The serum was used to determine biochemical values such as alanine transaminase, aspartate transaminase, blood urea nitrogen, and creatinine levels via an automatic biochemical analyzer (Beckman Coulter Diagnostics, Brea, CA, USA). Complete blood cell counts were performed using an automatic hematology analyzer (Beckman Coulter Diagnostics). Serum levels of several representative proinflammatory cytokines (interleukin-1-β [IL-1β], interleukin-6 [IL-6], tumor necrosis factor-α [TNF-α], and C-X-C motif chemokine ligand-1 [CXCL-1]) were measured using ELISA and a ProcartaPlex kit (ThermoFisher, Waltham, MA, USA), according the manufacturer’s instructions. Rectal temperature was monitored with a rectal probe (Kent Scientific Corp, Torrington, CT, USA).
Bacterial Inoculum Preparation
We chose five representative species of bacteria that are most frequently isolated in bone and joint infections [24, 26]. The following strains of bacteria were obtained from the American Type Culture Collection (Manassas, VA, USA): Staphylococcus aureus (BAA-1556), Escherichia coli (BAA-196), Enterococcus faecalis (BAA-2573), Streptococcus pyogenes (BAA-362), and Pseudomonas aeruginosa (BAA-2793). These bacteria were maintained on tryptic soy agar containing 5% sheep blood (BD™ Trypticase™ Soy Agar II with 5% sheep blood, Becton Dickinson, Franklin Lakes, NJ, USA) at 37° C and under aerobic conditions. Fresh bacterial cultures were prepared 24 hours before surgery. Stock solutions of the inoculums were prepared by collecting a colony of bacteria from the culture plates using a sterile cotton swab. After washing the collected bacteria three times in sterile normal saline, we adjusted the concentration of bacteria to 107 colony-forming units per mL. A standard optical density curve was used to determine the concentration of bacteria in the inoculum. For the local and systemic challenge, 1 × 104 and 1 × 106 colony-forming units of bacteria were applied, respectively.
Surgical Procedures and Bacterial Inoculation
According to the anesthetic and analgesia guideline from the institutional animal care and use committee, the mice were anesthetized using 2.5% inhalational isoflurane after induction with an intraperitoneal injection of ketamine (80 mg/kg) and xylazine (7.5 mg/kg). We treated mice with buprenorphine (1 mg/kg) for postoperative analgesia in this study. Aseptic techniques were used throughout the surgical procedures. After the adequate depth of anesthesia had been achieved, we disinfected the surgical site with povidone-iodine and ethanol swabs and placed a sterile drape over the mouse. The articular surface of the right distal femur was exposed via a medial parapatellar approach to access the distal femur. The femoral canal was identified and a 25-gauge sterile needle was inserted in a retrograde fashion through the intercondylar notch. The tract was then reamed manually with a 22-gauge needle. A surgery-grade K-wire (0.88 mm in diameter and 12 mm in length) was inserted into the reamed tract, leaving 1 mm of wire protruding from the joint space. For the local-inoculation groups of animals, 1 × 104 colony-forming units of bacteria were applied to the K-wire immediately before insertion. The arthrotomy site was closed with interrupted 4-0 Vicryl® (Ethicon, Somerville, NJ, USA) sutures. For the systemic-injection groups, mice received a single intravenous injection of 1 × 106 colony-forming units of bacteria into the tail vein immediately after wound closure. No postsurgical dressing of the wound was used in this study.
Postchallenge Bacterial Cultures and Species Identification
The mice were monitored for 4 weeks and then euthanized with carbon dioxide to harvest joint specimens for culturing. Briefly, the knee was disinfected and draped in a fashion identical to that used for the previous K-wire implantation. The joint capsule, entire femur, and intramedullary K-wire were harvested separately and placed into three sterile tubes, each containing 3 mL of normal saline solution. The femur and joint capsule were homogenized with a tissue grinder (Merck, Darmstadt, Germany). K-wires were sonicated to release any bacteria from the biofilm. Then, we inoculated culture plates (same tryptic soy agar used above; Becton Dickinson) with 30 µL of the sample supernatant and incubated the plates at 37° C for 24 hours. Positive results were defined as formation of at least one bacterial colony after 24 hours of culturing. The culture plates were photographed using a digital camera. Bacterial colonies were quantified using ImageJ analysis (National Institutes of Health, Bethesda, MD, USA). The software with the colony counter plugin could automatically count the colonies in captured photographs. Importantly, we set an a priori threshold for reliable bone and joint infection diagnosis. If the sum of three colony counts (bone, joint capsule, and K-wire) was less than 20 colonies in a given mouse, the results from this mouse would be excluded from further consideration. In this study, we did not encounter this.
Bacterial species were identified with 16S ribosomal DNA sequencing. We used a MicroSeq 500 microbial identification system (ThermoFisher), according to the manufacturer’s instructions. Briefly, to extract genomic DNA, we harvested a bacterial colony and suspended the cells in 100 µL of PrepMan™ Ultra Sample Preparation Reagent (ThermoFisher). The sample was incubated for 10 minutes at 100° C and centrifuged to obtain the supernatant. We amplified the 16S rDNA region with Master Mix (ThermoFisher) and purified extension products using a MicroSEQ™ ID Ultra Sequencing Strips Kit (ThermoFisher). Next, the sample was subjected to electrophoresis and analyzed using Applied Biosystems™3500 Series Genetic Analyzers (ThermoFisher). All bacteria isolated from the joints of mice in this study were identified via 16S ribosomal DNA sequencing to be the identical species of the initial inoculum, indicating that there was no extra-experimental microbial contamination.
Primary and Secondary Study Outcomes
Our primary study outcome was culture-positive rate after bacterial challenge. We assessed this by harvesting and culturing the bone, K-wire and joint capsules 4 weeks after the challenge. Our secondary study outcomes was the bacterial burden of the joint after bacterial challenge. We assessed it by quantitatively count the forming colonies on agar.
Sample Size Calculation, Randomization, and Blinding
The sample size was based on a Type I error set to 0.05, with a power of 80%. The superiority margin was set to 0. According to a pilot study we performed (data not shown; n = 10 for systemic E. coli), to detect a 40% difference between trained (20%) and control groups (60%), 16 mice would be required. With this number of mice, we achieved greater than 99% power to detect a statistically significant difference between locally challenged groups of mice. Considering the possible loss of mice and experimenter variability in applying bacteria, we finally overpowered the sample size to 20 for each group. Simple randomization was performed to assign the mice into the respective groups. Except for the researcher (JL) who prepared the training substances (zymosan, lipopolysaccharide, and placebo), all researchers in this study were blinded to the group assignment information.
Statistical Analysis
Differences in the proportions of outcomes between groups were compared using Fisher’s exact test. Significance was evaluated using the nonparametric Wilcoxon rank sum test for comparing continuous variables between different groups. IBM SPSS Statistics version 26 (Chicago, IL, USA) was used for the statistical analysis.
Results
Host Stress Response After Immunity Training
Training with zymosan or lipopolysaccharide caused mild but transient stress to the host. For both groups of mice, the mean rectal temperature reached a peak of less than 1° C above the normal baseline temperature 12 hours after injection and rapidly declined to the baseline temperature within 30 hours (Fig. 2A). The levels of proinflammatory cytokines (IL-1β, IL-6, tumor necrosis factor-alpha, and CXCL-1) increased in the serum of trained mice (Fig. 2B), likely reflecting the response to the zymosan or lipopolysaccharide injections. Consistently, the serum levels of blood urea nitrogen, creatinine, alanine transaminase, and aspartate transaminase increased transiently after immune training and declined back to baseline after 36 hours (Fig. 2C). Taken together, training with zymosan or lipopolysaccharide caused mild but transient stress to the host in terms of rectal temperature and blood urea nitrogen, creatinine, alanine transaminase, and aspartate transaminase levels.
Fig. 2.
A-C (A) This figure shows the mean change in rectal temperature in reference to the mean baseline temperature (Δ temperature) 6 hours, 12 hours, 18 hours, 24 hours, and 30 hours after immune training with an intravenous injection of saline (control), 0.1 mg of zymosan, or 0.1 mg of lipopolysaccharide. (B) This figure shows the mean serum levels of proinflammatory cytokines (IL-1β, IL-6, TNF-α , and CXCL-1) that were assessed 12 hours after training. (C) This figure shows the mean serum levels of blood urea nitrogen, creatinine, alanine transaminase, and aspartate aminotransferase, which were assessed 12 hours and 36 hours after training. The error bars indicate the SD. The transverse bars indicate the mean. ap < 0.05, control versus lipopolysaccharide; bp < 0.05, control versus zymosan; cp < 0.05, zymosan versus lipopolysaccharide; IL-1β = interleukin-1β; TNF-α = tumor necrosis factor α; IL-6 = interleukin-6; CXCL-1 = C-X-C motif chemokine ligand-1.
Trained Immunity Confers Broad-spectrum Protection Against Bone and Joint Infection
The trained immune system responded more rapidly and robustly to bacterial challenge and thus reduced the infection rate and bacterial burden after bacterial challenge. The number of white blood cells and neutrophils in whole blood from trained mice was higher than that from controls before bacterial challenge, and at early stages after postsurgical bacterial challenge (Fig. 3A). In trained mice, proinflammatory cytokines (IL-1β, IL-6, tumor necrosis factor-alpha, and CXCL-1) increased more rapidly and to a higher level in blood collected after bacterial challenge (Fig. 3B), suggesting that the trained immunity produced a measurable, rapid, and robust immune response to a bacterial challenge in our mouse joint replacement model.
Fig. 3.
A-B These figures show (A) white blood cell, neutrophil, and lymphocyte counts, and (B) serum levels of proinflammatory cytokines (IL-1β, IL-6, tumor necrosis factor alpha, and CXCL-1) after local inoculation (left column) or systemic injection (right column) of various bacterial species at different timepoints, as indicated. The error bars indicate the SD. The transverse bars indicate the mean. ap < 0.05 for control versus LPS; bp < 0.05 for control versus zymosan; cp < 0.05 for zymosan versus lipopolysaccharide; IL-1β = interleukin-1β; TNF-α = tumor necrosis factor α; IL-6 = interleukin-6; CXCL-1 = C-X-C motif chemokine ligand-1.
Presurgical immune system-trained mice had a reduced risk of culture-positive joint samples compared with those in the control groups (Table 1) after local inoculation with S. aureus (control: 100% [20 of 20]; zymosan: 55% [11 of 20], relative risk 0.55 [95% CI 0.37 to 0.82]; p = 0.001; lipopolysaccharide: 60% [12 of 20], RR 0.60 [95% CI 0.42 to 0.86]; p = 0.003), E. coli (control: 90% [18 of 20]; zymosan: 55% [11 of 20], RR 0.61 [95% CI 0.40 to 0.93]; p = 0.03; lipopolysaccharide: 55% [11 of 20], RR 0.61 [95% CI 0.40 to 0.93]; p = 0.03), E. faecalis (control: 85% [17 of 20]; zymosan: 40% [8 of 20], RR 0.47 [95% CI 0.27 to 0.83]; p = 0.008; lipopolysaccharide: 50% [10 of 20], RR 0.59 [95% CI 0.37 to 0.95]; p = 0.04), S. pyogenes (control: 95% [19 of 20]; zymosan: 40% [8 of 20], RR 0.42 [95% CI 0.24 to 0.73]; p < 0.001; lipopolysaccharide: 60% [12 of 20], RR 0.63 [95% CI 0.44 to 0.92]; p = 0.02), and P. aeruginosa (control: 100% [20 of 20]; zymosan: 40% [8 of 20], RR 0.40 [95% CI 0.23 to 0.68]; p < 0.001; lipopolysaccharide: 60% [12 of 20], RR 0.60 [95% CI 0.42 to 0.86]; p = 0.003). Likewise, trained mice had enhanced protection against systemic challenge (Table 2) with S. aureus (control: 70% [14 of 20]; zymosan: 15% [3 of 20], RR 0.21 [95% CI 0.07 to 0.63]; p = 0.001; lipopolysaccharide: 15% [3 of 20], RR 0.21 [95% CI 0.07 to 0.63]; p = 0.001), E. coli (control: 50% [10 of 20]; zymosan: 10% [2 of 20], RR 0.20 [95% CI 0.05 to 0.80]; p = 0.01; lipopolysaccharide: 15% [3 of 20], RR 0.30 [95% CI 0.10 to 0.93]; p = 0.04), E. faecalis (control: 45% [9 of 20]; zymosan: 0% [0 of 20], RR N/A; p = 0.001; lipopolysaccharide: 0% [0 of 20], RR N/A; p = 0.001), S. pyogenes (control: 65% [13 of 20]; zymosan: 10% [2 of 20], RR 0.15 [95% CI 0.04 to 0.60]; p < 0.001; lipopolysaccharide: 20% [4 of 20], RR 0.31 [95% CI 0.12 to 0.78]; p = 0.01), and P. aeruginosa (control: 40% [8 of 20]; zymosan: 0% [0 of 20], RR N/A; p = 0.003; lipopolysaccharide: 5% [1 of 20], RR 0.13 [95% CI 0.02 to 0.91]; p = 0.02). Similarly, trained immunity also resulted in lower mean colony counts in cultured samples from trained mice than in samples from control mice (Fig. 4) after local inoculation and systemic injection with S. aureus, E. coli, E. faecalis, S. pyogenes, or P. aeruginosa.
Table 1.
Proportions of positive culture results by specimen type and treatment group after local inoculation of bacteria
| Specimen | Control (n = 20) | Zymosan (n = 20) | Relative risk (95% CI) | p value | Lipopolysaccharide (n = 20) | Relative risk (95% CI) | p value |
| Staphylococcus aureus | |||||||
| Bone | 20 | 9 | 0.45 (0.28 to 0.73) | <0.001 | 12 | 0.60 (0.42 to 0.86) | 0.003 |
| K-wire | 20 | 11 | 0.55 (0.37 to 0.82) | 0.001 | 12 | 0.60 (0.42 to 0.86) | 0.003 |
| Joint capsule | 20 | 10 | 0.50 (0.32 to 0.78) | <0.001 | 11 | 0.55 (0.37 to 0.82) | 0.001 |
| Whole jointa | 20 | 11 | 0.55 (0.37 to 0.82) | 0.001 | 12 | 0.60 (0.42 to 0.86) | 0.003 |
| Escherichia coli | |||||||
| Bone | 18 | 9 | 0.50 (0.30 to 0.83) | 0.006 | 10 | 0.56 (0.35 to 0.88) | 0.01 |
| K-wire | 17 | 11 | 0.65 (0.42 to 1.00) | 0.08 | 11 | 0.65 (0.42 to 1.00) | 0.08 |
| Joint capsule | 18 | 10 | 0.56 (0.35 to 0.88) | 0.01 | 11 | 0.61 (0.40 to 0.93) | 0.03 |
| Whole jointa | 18 | 11 | 0.61 (0.40 to 0.93) | 0.03 | 11 | 0.61 (0.40 to 0.93) | 0.03 |
| Enterococcus faecalis | |||||||
| Bone | 15 | 7 | 0.47 (0.24 to 0.89) | 0.03 | 9 | 0.60 (0.35 to 1.04) | 0.11 |
| K-wire | 17 | 8 | 0.47 (0.27 to 0.83) | 0.008 | 9 | 0.53 (0.32 to 0.89) | 0.02 |
| Joint capsule | 14 | 6 | 0.43 (0.21 to 0.89) | 0.03 | 8 | 0.57 (0.31 to 1.05) | 0.11 |
| Whole jointa | 17 | 8 | 0.47 (0.27 to 0.83) | 0.008 | 10 | 0.59 (0.37 to 0.95) | 0.04 |
| Streptococcus pyogenes | |||||||
| Bone | 18 | 7 | 0.39 (0.21 to 0.72) | 0.001 | 12 | 0.67 (0.45 to 0.98) | 0.07 |
| K-wire | 19 | 8 | 0.42 (0.24 to 0.73) | <0.001 | 11 | 0.58 (0.39 to 0.87) | 0.008 |
| Joint capsule | 17 | 6 | 0.35 (0.18 to 0.71) | 0.001 | 10 | 0.59 (0.37 to 0.95) | 0.04 |
| Whole jointa | 19 | 8 | 0.42 (0.24 to 0.73) | <0.001 | 12 | 0.63 (0.44 to 0.92) | 0.02 |
| Pseudomonas aeruginosa | |||||||
| Bone | 19 | 7 | 0.37 (0.20 to 0.68) | <0.001 | 12 | 0.63 (0.44 to 0.92) | 0.02 |
| K-wire | 20 | 8 | 0.40 (0.23 to 0.68) | <0.001 | 12 | 0.60 (0.42 to 0.86) | 0.003 |
| Joint capsule | 19 | 7 | 0.37 (0.20 to 0.68) | <0.001 | 12 | 0.63 (0.44 to 0.92) | 0.02 |
| Whole jointa | 20 | 8 | 0.40 (0.23 to 0.68) | <0.001 | 12 | 0.60 (0.42 to 0.86) | 0.003 |
Refers to any positive culture results of bone, K-wire, or joint capsule. Values are presented as the number of positive culture results.
Table 2.
Proportions of positive culture results by specimen type of and treatment group after systemic injection of bacteria
| Specimen | Control (n = 20) | Zymosan (n = 20) | Relative risk 95% CI) | p value | Lipopolysaccharide (n = 20) | Relative risk (95% CI) | p value |
| Staphylococcus aureus | |||||||
| Bone | 14 | 3 | 0.21 (0.07 to 0.63) | 0.001 | 3 | 0.21 (0.07 to 0.63) | 0.001 |
| K-wire | 12 | 3 | 0.25 (0.83 to 0.75) | 0.008 | 3 | 0.25 (0.83 to 0.75) | 0.008 |
| Joint capsule | 13 | 3 | 0.23 (0.08 to 0.69) | 0.003 | 3 | 0.23 (0.08 to 0.69) | 0.003 |
| Whole jointa | 14 | 3 | 0.21 (0.07 to 0.63) | 0.001 | 3 | 0.21 (0.07 to 0.63) | 0.001 |
| Escherichia coli | |||||||
| Bone | 10 | 2 | 0.20 (0.05 to 0.80) | 0.01 | 2 | 0.20 (0.05 to 0.80) | 0.01 |
| K-wire | 9 | 1 | 0.11 (0.02 to 0.80) | 0.008 | 1 | 0.11 (0.02 to 0.80) | 0.008 |
| Joint capsule | 9 | 2 | 0.22 (0.06 to 0.90) | 0.03 | 3 | 0.33 (0.11 to 1.05) | 0.08 |
| Whole jointa | 10 | 2 | 0.20 (0.05 to 0.80) | 0.01 | 3 | 0.30 (0.10 to 0.93) | 0.04 |
| Enterococcus faecalis | |||||||
| Bone | 8 | 0 | N/A | 0.003 | 0 | N/A | 0.003 |
| K-wire | 8 | 0 | N/A | 0.003 | 0 | N/A | 0.003 |
| Joint capsule | 9 | 0 | N/A | 0.001 | 0 | N/A | 0.001 |
| Whole jointa | 9 | 0 | N/A | 0.001 | 0 | N/A | 0.001 |
| Streptococcus pyogenes | |||||||
| Bone | 10 | 2 | 0.20 (0.05 to 0.80) | 0.01 | 3 | 0.30 (0.10 to 0.93) | 0.04 |
| K-wire | 11 | 1 | 0.09 (0.01 to 0.64) | 0.001 | 2 | 0.18 (0.05 to 0.72) | 0.006 |
| Joint capsule | 13 | 2 | 0.15 (0.04 to 0.60) | 0.001 | 4 | 0.31 (0.12 to 0.78) | 0.01 |
| Whole jointa | 13 | 2 | 0.15 (0.04 to 0.60) | 0.001 | 4 | 0.31 (0.12 to 0.78) | 0.01 |
| Pseudomonas aeruginosa | |||||||
| Bone | 7 | 0 | N/A | 0.008 | 1 | 0.14 (0.02 to 1.06) | 0.04 |
| K-wire | 7 | 0 | N/A | 0.008 | 1 | 0.14 (0.02 to 1.06) | 0.04 |
| Joint capsule | 8 | 0 | N/A | 0.003 | 1 | 0.13 (0.02 to 0.91) | 0.02 |
| Whole jointa | 8 | 0 | N/A | 0.003 | 1 | 0.13 (0.02 to 0.91) | 0.02 |
Refers to any positive culture results of bone, K-wire, or joint capsule. Values are compared with control by using Fisher’s exact test.
Fig. 4.
This figure shows the number of colony-forming units in cultures derived from surgical-site samples after local inoculation (left column) or systemic injection (right column) with the indicated pathogens. Cultures were obtained from the bone, K-wire, and joint capsule. The transverse bars indicate the median; CFU = colony forming units; ap < 0.05 for control versus lipopolysaccharide; bp < 0.05 for control versus zymosan; cp < 0.05 for zymosan versus lipopolysaccharide.
Protection Against Bone and Joint Infection Caused by S. Aureus Persists for at Least 8 Weeks After Trained Immunity
Compared with control mice, zymosan-trained mice were more effectively protected against both local (control: 20 of 20 [100%], zymosan: 14 of 20 [70%], relative risk 0.70 [95% CI 0.53 to 0.93], p = 0.02) and systemic (control: 70% [14 of 20] zymosan: 30% [6 of 20], relative risk 0.43 [95% CI 0.21 to 0.89]; p = 0.03) challenge with S. aureus for up to 8 weeks than controls (Table 3).
Table 3.
Proportions of positive culture results (S. aureus) by specimen type and timepoint after immune-system training
| Specimen | Control (n = 20) | 2 weeks (n = 20) | 4 weeks (n = 20) | 6 weeks (n = 20) | 8 weeks (n = 20) | Relative risk (95% CI) a | p valuea |
| Local inoculation, n | |||||||
| Bone | 20 | 9 | 12 | 11 | 14 | 0.70 (0.53 to 0.93) | 0.02 |
| K-wire | 20 | 11 | 11 | 13 | 13 | 0.65 (0.47 to 0.90) | 0.008 |
| Joint capsule | 20 | 10 | 12 | 12 | 12 | 0.60 (0.42 to 0.86) | 0.003 |
| Whole jointb | 20 | 11 | 12 | 13 | 14 | 0.70 (0.53 to 0.93) | 0.02 |
| Systemic injection, n | |||||||
| Bone | 14 | 3 | 4 | 4 | 6 | 0.43 (0.21 to 0.89) | 0.03 |
| K-wire | 12 | 3 | 3 | 3 | 4 | 0.33 (0.13 to 0.86) | 0.02 |
| Joint capsule | 13 | 3 | 4 | 3 | 6 | 0.46 (0.22 to 0.97) | 0.06 |
| Whole jointb | 14 | 3 | 4 | 4 | 6 | 0.43 (0.21 to 0.89) | 0.03 |
Control mouse versus 8 weeks trained mouse by using Fisher’s exact test.
Refers to any positive culture results of bone, K-wire, or joint capsule.
Discussion
Preclinical studies have demonstrated that the innate immune system could be trained with immunogens to improve its response to subsequent challenge. A Phase 1 clinical trial demonstrated that Bacillus Calmette-Guérin vaccination could successfully improve the response against malaria in humans [23]. Training the innate immune system before elective surgery might be a potential option to prevent bone and joint infection. The most attractive feature of this strategy is that trained immunity is nonspecific and thus theoretically broad-spectrum. In this study, we demonstrated that training the mice with zymosan and lipopolysaccharide caused mild and transient stress in terms of body temperature and blood biochemical parameters. Trained immunity confers protection against all five strains of bacteria in a mouse model of bone and joint infection. The protection conferred by immunity training lasted up to 8 weeks in this mouse model.
Limitations
The current work had several limitations. First, low-grade but sustained inflammation after training, termed as overtraining, has been implicated in the pathophysiology of chronic and autoinflammatory disorders of the immune system. We did not assess the risks of overtraining in this study. Our protocol was unlikely to cause overtraining because overtraining commonly develops after repeated training [16]. Nevertheless, investigators should pay close attention to this issue in future translational studies. Second, our experimental immunogens, zymosan and lipopolysaccharide, are classic agents used for training in experimental research but clearly would not be the most appropriate ones for the clinic, because these substances are natural components of bacteria. The range between effective and toxic doses might be relatively narrow for these substances. Instead, many approved vaccines (for example, the Bacillus Calmette-Guérin and measles vaccines) might be more suitable training immunogens for clinical studies because of their well-established safety profiles. Third, this study lacked longitudinal monitoring of the bacterial burden and host immune response, which clearly carried valuable information regarding immunity training. As a proof-of-concept study, the main conclusion was not affected by this limitation. However, this issue might be very valuable for future investigations, especially those focusing on the mechanistic aspect.
Host Stress Response After Immunity Training
High levels of zymosan or lipopolysaccharide could induce the death of host cells by eliciting stress [19]. As a consequence, intravenous injection of high-dose zymosan or lipopolysaccharide could lead to acute injuries on multiple organs such as kidney and liver [8, 14]. In this proof-of-concept study, we used two classic agents (zymosan and lipopolysaccharide) for immunity training and demonstrated that intravenous injection of low-dose zymosan or lipopolysaccharide could successfully train the innate immunity without causing severe stress and organ injuries. Interestingly, lipopolysaccharide triggered a slightly stronger stress than zymosan. However, training with lipopolysaccharide, in contrast, had a relatively weaker protection compared with zymosan. As a typical phenomenon after innate immunity training [9], the number of neutrophils instead of lymphocytes in circulation from trained mice was higher than that from controls, indicating the existence of mild stress lasting for at least 2 weeks. This phenomenon might explain why the host could respond to the challenge more rapidly and robustly.
Trained Immunity Confers Broad-spectrum Protection Against Bone and Joint Infection
For certain pathogens, the most economical and reliable strategy for infection prevention is enhancing patients’ immune defense by using vaccines [22]. However, unlike many other infectious diseases, bone and joint infections are caused by a variety of microorganisms [24, 26], highlighting the need for an effective strategy against a broad spectrum of bacteria. In the current study, mice whose immune systems were trained with zymosan or lipopolysaccharide outperformed untrained ones in the face of infectious postsurgical challenges. To determine whether the protection conferred by the training was effective against a broad spectrum of pathogens, we challenged the mice with five of the most common species of bacteria, accounting for more than 90% of the overall bone and joint infections [24, 26]. Our findings have laid the foundation for a larger-animal experiment, and if this strategy is comparably effective and safe, future clinical studies in humans thereafter might be warranted to translate these findings into clinical practice.
Protection Against Bone and Joint Infection Caused by S. Aureus Persists for at Least 8 Weeks After Trained Immunity
We also demonstrated that training the innate immune system with zymosan or lipopolysaccharide conferred long-lasting protection in this mouse model. The persistence of protection conferred by presurgical immunity training is an important outcome, one that is relevant to future clinical translation. The molecular basis for trained immunity involves transcriptional and epigenetic reprogramming of immune cells and their progenitors [17]. Thus, for vaccines against Bacillus Calmette-Guérin or measles, for example, the nonspecific prevention conferred might last for months, at the very least [3]. In the present study, we demonstrated that 2 months after the initial training, the protective effects were still sufficiently robust to protect the mice from both systemic and local challenge with S. aureus. This immunological memory of trained immunity typically lasts from 1 week to several months [10]. Although this protection might not be as persistent as that observed for classic immunologic memory, 8 weeks is clearly adequate for orthopaedic device implantation to be completed. Please note this time window for surgery after training might be broadened or narrowed in large animals and humans. Future investigation should pay close attention to this important issue.
Conclusions
Trained immunity confers mild stress and broad-spectrum protection against bone and joint infection in a mouse model. The protection conferred by immunity training lasted up to 8 weeks in this mouse model. Further study evaluating this presurgical strategy should be performed in larger-animal models and then, if the outcomes are favorable, in humans.
Acknowledgments
We thank Changqing Zhang MD, PhD, for his constructive suggestion on the study design.
Footnotes
The institution of one or more of the authors (HZ) has received, during the study period, funding from the Clinical Research Project of Shanghai Municipal Health Commission (Grant 20194Y0254). The institution of one or more of the authors (XZ) has received, during the study period, funding from the National Natural Science Foundation of China (Grant 81672144, 81974331) and a Shanghai Municipal Education Commission-Gaofeng Clinical Medicine Grant (Grant 20161429).
Clinical Orthopaedics and Related Research® neither advocates nor endorses the use of any treatment, drug, or device. Readers are encouraged to always seek additional information, including FDA approval status, of any drug or device before clinical use.
Each author certifies that his or her institution approved the animal protocol for this investigation and that all investigations were conducted in conformity with ethical principles of research.
The first two authors contributed equally to this manuscript.
All ICMJE Conflict of Interest Forms for authors and Clinical Orthopaedics and Related Research® editors and board members are on file with the publication and can be viewed on request.
References
- 1.Angulo M, Reyes-Becerril M, Cepeda-Palacios R, Angulo C. Oral administration of debaryomyces hansenii CBS8339-beta-glucan induces trained immunity in newborn goats. Dev Comp Immunol. 2020;105:103597. [DOI] [PubMed] [Google Scholar]
- 2.Bartlett JG. Smallpox vaccination and patients with human immunodeficiency virus infection or acquired immunodeficiency syndrome. Clin Infect Dis. 2003;36:468-471. [DOI] [PubMed] [Google Scholar]
- 3.Benn CS, Netea MG, Selin LK, Aaby P. A small jab - a big effect: nonspecific immunomodulation by vaccines. Trends Immunol. 2013;34:431-439. [DOI] [PubMed] [Google Scholar]
- 4.Berthelot JM, Sibilia J. Trained immunity and autoimmune disease: Did Eve sin before Adam? Joint Bone Spine. 2019;86:293-295. [DOI] [PubMed] [Google Scholar]
- 5.Butkeviciute E, Jones CE, Smith SG. Heterologous effects of infant BCG vaccination: potential mechanisms of immunity. Future Microbiol. 2018;13:1193-1208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ciarlo E, Heinonen T, Theroude C, Asgari F, Le Roy D, Netea MG, Roger T. Trained immunity confers broad-spectrum protection against bacterial infections. J Infect Dis. [Published online ahead of print December 31, 2019]. DOI: 10.1093/infdis/jiz692. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Council NR. Guide for the Care and Use of Laboratory Animals . Washington, DC: National Academies Press; 2010. [Google Scholar]
- 8.Gu L, Liu J, Xu D, Lu Y. Polydatin prevents LPS-induced acute kidney injury through inhibiting inflammatory and oxidative responses. Microb Pathog. 2019;137:103688. [DOI] [PubMed] [Google Scholar]
- 9.Gyssens IC, Netea MG. Heterologous effects of vaccination and trained immunity. Clin Microbiol Infect. 2019;25:1457-1458. [DOI] [PubMed] [Google Scholar]
- 10.Hamada A, Torre C, Drancourt M, Ghigo E. Trained immunity carried by non-immune cells. Front Microbiol. 2018;9:3225. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Hu Z, Lowrie D, Fan X. Statins as adjunctive therapy against tuberculosisthe balance between statin-induced anti-TB effect and trained immunity suppression. J Infect Dis. 2020;222:335-336. [DOI] [PubMed] [Google Scholar]
- 12.Kilkenny C, Browne WJ, Cuthill IC, Emerson M, Altman DG. Improving bioscience research reporting: The ARRIVE guidelines for reporting animal research. PLOS Biology. 2010;8:e1000412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kurtz J, Franz K. Innate defence: evidence for memory in invertebrate immunity. Nature. 2003;425:37-38. [DOI] [PubMed] [Google Scholar]
- 14.Lebeaupin C, Proics E, de Bieville CH, Rousseau D, Bonnafous S, Patouraux S, Adam G, Lavallard VJ, Rovere C, Le Thuc O, Saint-Paul MC, Anty R, Schneck AS, Iannelli A, Gugenheim J, Tran A, Gual P, Bailly-Maitre B. ER stress induces NLRP3 inflammasome activation and hepatocyte death. Cell Death Dis. 2015;6:e1879. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mina MJ, Metcalf CJ, de Swart RL, Osterhaus AD, Grenfell BT. Long-term measles-induced immunomodulation increases overall childhood infectious disease mortality. Science. 2015;348:694-699. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Mulder WJM, Ochando J, Joosten LAB, Fayad ZA, Netea MG. Therapeutic targeting of trained immunity. Nat Rev Drug Discov. 2019;18:553-566. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Netea MG, Joosten LA, Latz E, Mills KH, Natoli G, Stunnenberg HG, O'Neill LA, Xavier RJ. Trained immunity: A program of innate immune memory in health and disease. Science. 2016;352:aaf1098. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Quintin J, Saeed S, Martens JHA, Giamarellos-Bourboulis EJ, Ifrim DC, Logie C, Jacobs L, Jansen T, Kullberg BJ, Wijmenga C, Joosten LAB, Xavier RJ, van der Meer JWM, Stunnenberg HG, Netea MG. Candida albicans infection affords protection against reinfection via functional reprogramming of monocytes. Cell Host Microbe. 2012;12:223-232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Rodriguez M, Domingo E, Alonso S, Frade JG, Eiros J, Crespo MS, Fernandez N. The unfolded protein response and the phosphorylations of activating transcription factor 2 in the trans-activation of il23a promoter produced by beta-glucans. J Biol Chem. 2014;289:22942-22957. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Rusek P, Wala M, Druszczynska M, Fol M. Infectious agents as stimuli of trained innate immunity. Int J Mol Sci. 2018;19:456. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Sharrock J, Sun JC. Innate immunological memory: from plants to animals. Curr Opin Immunol. 2020;62:69-78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Verani JR, Baqui AH, Broome CV, Cherian T, Cohen C, Farrar JL, Feikin DR, Groome MJ, Hajjeh RA, Johnson HL, Madhi SA, Mulholland K, O'Brien KL, Parashar UD, Patel MM, Rodrigues LC, Santosham M, Scott JA, Smith PG, Sommerfelt H, Tate JE, Victor JC, Whitney CG, Zaidi AK, Zell ER. Case-control vaccine effectiveness studies: Preparation, design, and enrollment of cases and controls. Vaccine. 2017;35:3295-3302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Walk J, de Bree LCJ, Graumans W, Stoter R, van Gemert GJ, van de Vegte-Bolmer M, Teelen K, Hermsen CC, Arts RJW, Behet MC, Keramati F, Moorlag S, Yang ASP, van Crevel R, Aaby P, de Mast Q, van der Ven A, Stabell Benn C, Netea MG, Sauerwein RW. Outcomes of controlled human malaria infection after BCG vaccination. Nat Comm. 2019;10:874. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Wallander K, Jorup-Ronstrom C, Ullberg M, Tornblom I, Ottosson C, Giske CG. Etiology of bone and joint infections: a case series of 363 consecutive patients from an orthopaedic infection unit. Infect Dis. 2016;48:618-625. [DOI] [PubMed] [Google Scholar]
- 25.Walsh MC, Takegahara N, Kim H, Choi Y. Updating osteoimmunology: regulation of bone cells by innate and adaptive immunity. Nat Rev Rheumatol. 2018;14:146-156. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Zhu H, Li X, Zheng X. A descriptive study of open fractures contaminated by seawater: infection, pathogens, and antibiotic resistance. Bioed Res Int. 2017;2017:2796054. [DOI] [PMC free article] [PubMed] [Google Scholar]