A novel mouse model for septic arthritis induced by Pseudomonas aeruginosa

Septic arthritis is one of the most aggressive joint diseases. Although caused predominantly by S. aureus, Gram-negative bacteria, Pseudomonas aeruginosa among them, account for a significant percentage of the causal agents of septic arthritis. However, septic arthritis caused by P. aeruginosa has not been studied thus far, due to lack of an animal model. NMRI mice were inoculated with different doses of P. aeruginosa. The clinical course of septic arthritis and radiological changes of joints were examined. Furthermore, the host molecular and cellular mechanisms involved in P. aeruginosa-induced septic arthritis were investigated. Inoculation of mice with P. aeruginosa caused septic arthritis in a dose-dependent manner. Neutrophil depletion led to higher mortality and more severe joint destruction (p < 0.01). In contrast, monocyte depletion resulted in higher mortality (p < 0.05) but similar arthritis severity compared to controls. Mice depleted of CD4+ T-cells inoculated with P. aeruginosa displayed less severe bone damage (p < 0.05). For the first time, a mouse model for P. aeruginosa septic arthritis is presented. Our data demonstrate that neutrophils play a protective role in P. aeruginosa septic arthritis. Monocytes/macrophages, on the other hand, are only essential in preventing P. aeruginosa-induced mortality. Finally, CD4+ T-cells are pathogenic in P. aeruginosa septic arthritis.

. P. aeruginosa induces septic arthritis in mice. Naval Medical Research Institute (NMRI) mice were inoculated with different doses of Pseudomonas aeruginosa (P. aeruginosa) (2.2 × 10 6 -2.8 × 10 8 colony-forming units [CFU]/mouse) and followed for up to 10 days. The severity (A) of arthritis in the mice were observed for 10 days post-infection. Y values represent measurements from surviving mice only for the respective days. Changes in body weight expressed as percentages of the initial body weight (B) and cumulative survival (C) of the mice. Cumulative bone destruction scores (D) and frequency of bone destruction (E) of the joints from all 4 limbs of NMRI mice as assessed by micro-computed tomography scan. A representative microcomputed tomography images (F) of an intact knee joint from healthy NMRI mouse and destroyed knee joints from NMRI mice with septic arthritis on day 3, 7 and 10 post-infection. The arrows indicate bone destruction. Representative photomicrographs (G) of histologically intact knee joint from a healthy NMRI mouse (upper panel) and of a heavily inflamed knee joint with severe bone and cartilage destruction from NMRI mouse with septic arthritis inoculated with P. aeruginosa (lower panel), stained with hematoxylin and eosin. Original magnification, ×10. The asterisk indicates heavily inflamed synovium. Abscess scores of the kidneys from the mice sacrificed 10 days post-infection (H) and, bacterial load of P. aeruginosa in kidneys of the mice (I). Levels of the pro-inflammatory cytokine Interleukin 6 (IL-6) (J) and chemokine monocyte chemoattractant protein 1 (MCP-1) (K) in serum were determined after termination of the experiment on day 10 post-infection. The data from 2 independent experiments were pooled, (n = 5-11/group). Statistical evaluations were performed using the Mann-Whitney U test (A,B,D,H-K), Log-rank Mantel cox (C) and Fisher's exact test (E). Data are expressed as mean values ± SEM. *p < 0.05; **p < 0.01, ***p < 0.001. www.nature.com/scientificreports www.nature.com/scientificreports/ aeruginosa compared to mice receiving lower doses ( Fig. 1D-F). The subgroup analyses of bone destruction are shown in Supplementary File (see Supplementary Tables S1 and S2).
A dose-dependent pattern regarding weight loss among mice receiving different doses of P. aeruginosa was also observed (Fig. 1B). Mice inoculated with higher doses of P. aeruginosa lost significantly more weight (p < 0.05) during the course of the experiment compared to mice receiving lower doses of bacteria.
Significantly more mice succumbed to P. aeruginosa in the group receiving the highest dose of bacteria compared to all other groups (p < 0.05) (Fig. 1C). In fact, 60% of the mice in the highest bacterial dose group had died by day 9.
A dose-dependent pattern regarding the kidney abscess score was also observed: mice inoculated with the highest dose of P. aeruginosa developed more macroscopic kidney abscesses compared to those inoculated with lower doses of the bacteria (p < 0.05) (Fig. 1H). Although not statistically significant, dose-dependent tendencies for higher CFU counts were observed in the kidneys of mice receiving the higher doses of P. aeruginosa (5.6 × 10 7 CFU/mouse, Fig. 1I).
Interestingly, a dose-dependent pattern regarding the serum IL-6 and MCP-1 levels were observed in the mice. Mice receiving higher doses of P. aeruginosa (5.6 × 10 7 CFU/mouse) had significantly higher levels of both IL-6 (p < 0.05) and MCP-1 (p < 0.001) serum compared to mice receiving lower doses (Fig. 1J,K).
The definitive diagnosis of septic arthritis is the recovery of bacteria in the joints. To this end, joints from P. aeruginosa infected mice were collected, homogenized and plated. Five out of six mice had at least one joint that was positive for CFU counts.
To study which cytokines are responsible for the onset of P. aeruginosa-induced septic arthritis, supernatants collected from joint homogenates of P. aeruginosa infected mice (7 × 10 7 CFU/mouse) were compared to homogenates from healthy mice. Significantly elevated levels of TNF-α and IL-6 were observed among the P. aeruginosa infected mice compared to healthy mice (see Supplementary Fig. S1). However, no differences with regard to IL-10 and MCP-1 were observed between the groups (see Supplementary Fig. S1).
Neutrophils are protective against P. aeruginosa-induced septic arthritis. To understand the role of neutrophils in P. aeruginosa-induced septic arthritis, mice were either depleted of neutrophils by anti-Ly6G antibodies or received isotype control antibodies. Mice lacking neutrophils exhibited significantly higher severity (p < 0.01) as well as higher frequency of bone erosions (p < 0.05) when infected by P. aeruginosa compared to isotype controls ( Fig. 2A-C). The subgroup analyses of bone destruction are shown in Supplementary File (see Supplementary Table S3).
Neutrophils are absolutely crucial for the defence against P. aeruginosa-induced mortality, as mice depleted of neutrophils had significantly higher mortality (p < 0.01), compared to their isotype controls (Fig. 2D). Only 40% of the mice depleted of neutrophils survived until termination of the experiment while the survival rate for the controls was 100%.
Monocytes/macrophages are essential in protection against P. aeruginosa-induced mortality. We proceeded further and investigated the role of monocytes/macrophages in P. aeruginosa-induced septic arthritis by depleting blood monocytes using clodronate liposomes. In contrast to mice depleted of neutrophils, the lack of monocytes/macrophages did not have any impact on the severity or on the frequency of bone  Supplementary Table S4). However, monocytes/macrophages, just like neutrophils, are essential for the protection of the host against P. aeruginosa-induced mortality. Mice lacking macrophages had significantly higher mortality (p < 0.05) compared to their controls (Fig. 3D), with approximately 45% of the macrophage-depleted mice succumbing to the disease compared to none from the control mice.

Septic arthritis induced by P. aeruginosa is mediated through CD4 but not CD8 T-cells. Next,
we investigated the role of T-cells in P. aeruginosa-induced septic arthritis by selectively depleting either CD4 or CD8 T-cells in the mice. Interestingly, the severity of bone destruction was significantly lower in mice depleted of CD4 T-cells (p < 0.05) compared to mice treated with isotype control antibody (Fig. 4A,C). This is in line with our   Supplementary Table S5). Intriguingly, mice depleted of CD4 T-cells displayed significantly higher levels of serum MCP-1 (p < 0.01) compared to both isotype controls as well as mice lacking CD8 T-cells during P. aeruginosa i.v. infection (Fig. 4D).

IL-1 does not play a crucial role in P. aeruginosa-induced septic arthritis. We have previously
shown that treatment with IL-1 receptor antagonist (Ra) aggravates S. aureus-induced septic arthritis and sepsis in mice 8 . In the present study, we investigated the role of IL-1 in mice infected with P. aeruginosa. IL-1 Ra treatment tended to aggravate P. aeruginosa arthritis, but no significant difference regarding either the severity or the frequency of bone erosion was observed ( Fig. 5A-C).

Antibiotic treatment reverses bone erosion in mice.
We investigated if the development of bone erosion in P. aeruginosa-induced septic arthritis could be prevented with appropriate antibiotic treatment. Indeed, antibiotic treated mice had less severe arthritis (p = 0.05) as well as tendency toward lower frequency (p = 0.06) of bone erosion compared to PBS control mice ( Fig. 6A-C).
Weight loss is an important parameter in our animal model that provides indication of the systemic effect of the disease. Here, we observed that PBS-treated control mice lost more weight (p = 0.05) compared to antibiotic-treated mice on day 3 post-infection (Fig. 6D). Fig. 1, mice receiving higher dose of P. aeruginosa developed sepsis that led to death during the course of the experiment. We investigated if treatment with antibiotics could reverse this effect. Indeed, PBS-treated mice succumbed to P. aeruginosa infection within few days after infection whereas antibiotic-treated mice had significantly lower mortality (p = 0.003) (see Supplementary Fig. S2).

Discussion
S. aureus remains the most common cause of septic arthritis. In comparison to S. aureus, P. aeruginosa is somewhat under-reported, but still one of the most common Gram-negative bacteria causing septic arthritis, especially in intravenous drug abusers 3 . Interestingly, despite similar surgical interventions and antibiotic treatment, patients with septic arthritis caused by P. aeruginosa had a lower remission rate than those infected with S. aureus, suggesting that P. aeruginosa is at least as clinically virulent as S. aureus 14 . The use of a unique animal model for S. aureus septic arthritis developed by our laboratory has clarified the involvement of several bacterial virulence factors, as well as the roles of various host immune cell types and cytokines involved in the pathogenesis of the disease [8][9][10][11][12] . However, there is no available animal model for P. aeruginosa septic arthritis. In the current study, we have established a novel mouse model for P. aeruginosa septic arthritis and demonstrated that neutrophils play a role in protection against disease. This model will also be useful to identify the virulence factors of P. aeruginosa in future studies. www.nature.com/scientificreports www.nature.com/scientificreports/ Optimal arthritogenic dose is relatively narrow for the clinical P. aeruginosa strain used in this study, as the dose of 2.8 × 10 8 CFU/mouse induced more than 50% mortality and 2.2 × 10 6 CFU/mouse caused almost no signs of septic arthritis. Thus, the dose for different laboratory and clinical strains should be titrated and tested in the model. Our next step is to find the arthritis dose for the most used P. aeruginosa strains such as PA01, whose complete genome sequence has been reported 15 and a library with large numbers of mutants are available to study the roles of different virulence factors in P. aeruginosa septic arthritis. The doses chosen should also be adjusted according to the purpose of the study or hypothesis in the different experiments. For example, to understand the role of neutrophils in P. aeruginosa arthritis, the bacterial dose should be kept at a lower level, since marked down-regulation of innate immunity by neutrophil depletion leads to uncontrolled systemic infection and death, which was the case in our study.
Pro-inflammatory cytokines are an essential part of the immune response required to eliminate invading microorganisms. Nevertheless, high levels of TNF-α and IL-6 have been shown to aggravate joint destruction in septic arthritis 16,17 and anti-TNF therapy is known to ameliorate S. aureus septic arthritis 18 . In the current study, significantly higher levels of TNF-α and IL-6 were observed in supernatants of joint homogenates collected from P. aeruginosa infected mice compared to supernatants from healthy mice.
Some elements of innate immunity, including neutrophils 19 , complement factors 11 and natural killer cells 20 , are protective in S. aureus septic arthritis. On the other hand, monocytes/macrophages are pathogenic in arthritic lesions, but protective in septic lethality 21 . In agreement with the data from S. aureus-induced septic arthritis, our current study compellingly demonstrates that neutrophils are the most crucial immune cells for better survival rates and for less severe septic arthritis, in our model of P. aeruginosa-induced septic arthritis. Indeed, both neutropenic mice and MyD88 deficient mice that cannot recruit neutrophils to lungs are highly susceptible to fatal P. aeruginosa lung infection 22 . CXC chemokines are critical mediators of neutrophil-mediated host defence in P. aeruginosa pneumonia 23,24 . Interestingly, P. aeruginosa itself possesses immune evasion capacity by producing ExoS and ExoT, two type III secreted effectors, blocking reactive oxygen species production by neutrophils 25 .
Remarkably, in mice with chemotherapy-induced neutropenia, vaccine-induced lung macrophage expansion protects against lethal P. aeruginosa pneumonia 26 , suggesting the potent role of monocytes/macrophages in Pseudomonas infections. In mice with septic arthritis caused by S. aureus, depletion of mononuclear phagocytes by etoposide led to deteriorated bacterial clearance and higher mortality 21 . The severity of arthritic lesions was, however, less pronounced in macrophage-depleted mice, suggesting a pathogenic role of macrophages in the development of septic arthritis. In the current study, monocyte/macrophage depletion by administering clodronate liposomes significantly increased mortality, which is in line with previous findings in S. aureus septic arthritis 21 . However, the protective effect of monocytes/macrophages depletion was totally absent in our animal model, as monocyte-depleted mice had actually slightly more severe bone erosions, demonstrating that macrophages are not pathogenic in P. aeruginosa-induced septic arthritis. CD4+ T lymphocytes constitute various T helper cell subsets including Th1, Th2, Th17, Tfh and Tregs. Within CD4+ subsets, there is a large proportion of inflammatory cells that contribute to autoimmune and infectious diseases 27 . CD4+ T cells are known to be directly involved in the development of autoimmune arthritis 27 , S. aureus septic arthritis 13 and Lyme arthritis 28 . Similarly, depletion of CD4+ but not CD8+ cells significantly reduced the severity and frequency of bone erosions in P. aeruginosa-induced septic arthritis, strongly indicating that CD4+ T cells are pathogenic in this disease. Which subsets of CD4+ T cells are responsible for this? Interferon-gamma www.nature.com/scientificreports www.nature.com/scientificreports/ (IFN-γ) producing Th1 cells are considered to be a major player in development of rheumatoid arthritis 29 . IFN-γ released by Th1 cells activates macrophages to produce pro-inflammatory cytokines such as TNF 30 . Th17 cells, another subset of CD4+ T cells, appear to play a potent role in the onset of autoimmune arthritis in several experimental models 31,32 . IL-17 blocking agents have been successfully used in psoriatic arthritis and ankylosing spondylitis 33,34 . Therefore, Th1 and Th17 are most likely the culprits, although further studies are needed to elucidate more detailed mechanisms.
MCP-1 is crucial for recruiting monocytes, neutrophils, as well as T-cells, to the site of an infection 35,36 . However, MCP-1 has been implicated in the pathogenesis of several diseases, e.g., rheumatoid arthritis 37 . Furthermore, MCP-1 has been shown to be upregulated in P. aeruginosa corneal infection in mice 38 and treatment with anti MCP-1 antibodies resulted in significant reductions in severity of corneal damage and neutrophil infiltration 39 . Indeed, MCP-1 serum levels were elevated in mice with P. aeruginosa septic arthritis compared to healthy controls, suggesting that MCP-1 is implicated in P. aeruginosa septic arthritis. However, the role of MCP-1 in this disease might be much more complicated, since mice depleted of CD4+ T-cells displayed less severe bone destruction but higher MCP-1 serum levels compared to control animals.
IL-1β is produced predominantly by macrophages and plays potent roles in the early recruitment of neutrophils and subsequent bacterial killing in P. aeruginosa pulmonary infection 40,41 . IL-1Ra treatment has recently been shown to aggravate S. aureus septic arthritis 8 . Similarly, in the current study, IL-1Ra treatment showed a tendency towards enhanced severity and frequency of bone erosions caused by P. aeruginosa infection although statistically significant differences in this study could not be reached, probably due to small sample size.
In summary, we demonstrate that we have successfully established a mouse model for P. aeruginosa induced septic arthritis. Our results strongly suggest that neutrophils are protective for both septic arthritis as well as P. aeruginosa induced mortality. However, monocytes/macrophages are protective against P. aeruginosa induced death but exhibited no role against septic arthritis. CD4+ T cells play a pathogenic role in septic arthritis. Our model system is useful, not only to understand the pathogenesis of P. aeruginosa septic arthritis, but also to study virulence factors of P. aeruginosa. At the termination of the experiments, the mice were anaesthetized with medetomidine (Orion Pharma, Finland) and ketamine hydrochloride (Pfizer AB, Sweden), blood from the axillary artery was collected and the mice were immediately sacrificed as previously described 8 .

Methods
The experiments were performed as follows: (1) mice (n = 5/group) were inoculated with 4 different doses of P. aeruginosa ranging from 2.2 × 10 6 -2.8 × 10 8 CFU/mouse to study the dose-dependent kinetics of arthritis; 2) mice (n = 9-10/group) were depleted of either neutrophils or macrophages and infected with P. aeruginosa (2.0-5.6 × 10 6 CFU/mouse) to study the roles of neutrophils and macrophages in P. aeruginosa-induced septic arthritis; (3) mice (n = 5/group), depleted of either CD4 or CD8 T-cells, were infected with P. aeruginosa (1.2 × 10 7 CFU/ mouse), in order to investigate the impact of T-cells in P. aeruginosa septic arthritis; (4) mice (n = 5/group) treated with IL-1 Ra were inoculated with P. aeruginosa (2 × 10 6 CFU/mouse) to study the role of IL-1 in septic arthritis; (5) mice (n = 4/group) were infected with P. aeruginosa (5.6 × 10 7 CFU/mouse) and treated with PBS or ciprofloxacin to investigate if treatment could prevent the development of bone destruction caused by P. aeruginosa in mice; (6) mice (n = 7-8/group) were infected with P. aeruginosa (1.7 × 10 8 CFU/mouse) and treated with PBS or ciprofloxacin to investigate if treatment could protect septic lethality in mice.
Two observers blinded to the treatment groups regularly weighed and clinically examined the mice for incidences and severity of arthritis. The mice were sacrificed on days 9-10 and sera were collected to assess the cytokine levels. In addition, the kidneys and the paws were obtained for assessment of bacterial counts and radiological examination of bone erosions, respectively.
To study whether the bacteria invaded and exist in the joints, mice (n = 10) were inoculated with P. aeruginosa (7 × 10 7 CFU/mouse), and different joint groups (forepaws and wrists, elbows, shoulders, back paws and ankles, knees, and hips) from each animal were collected separately and homogenized for CFU counts on day 7 when the clinical arthritis became evident. www.nature.com/scientificreports www.nature.com/scientificreports/ Neutrophil depletion. For selective depletion of murine blood neutrophils 44 , 400 μg of specific monoclonal anti-Ly6G antibody (clone 1A8; BioXCell) in 200 μl of PBS was intraperitoneally (i.p.) injected on day −1, before infection with P. aeruginosa, and on day 1, post-infection. The control mice were treated with isotype control antibody (clone 2A3; BioXCell).

In vivo
T cell depletion. The T-cell depletion procedures were carried out as described before 45 . For selective depletion of CD4 T-cells, a specific rat anti-mouse CD4 monoclonal antibody (clone GK1.5; BioXCell) was used, whereas to selectively deplete CD-8 T-cells in mice, a rat anti-mouse CD8α (clone 2.43; BioXCell) monoclonal antibody was used. The control mice for both groups were treated with a rat IgG2b isotype control (clone LTF-2; BioXCell) monoclonal antibody. Mice received i.p. a dose of 400 μg/mouse/antibody in 200 μl of PBS on day −1, before infection with P. aeruginosa, and on days 3 and 7, post-infection.

Measurement of depleted cells by flow cytometry.
The experimental protocols investigating the success rate of the cellular depletions mentioned above are described elsewhere 45 . The depletion efficacies for the targeted cells as analysed by flow cytometry were as follows: neutrophils (99.6%), monocytes (85.6%), CD4 T-cells (97.3%) and CD8 T-cells (90%) 45 .
Treatment with IL-1 receptor antagonist. In order to study the effect of IL-1 in P. aeruginosa-induced septic arthritis, Kineret ® (Anakinra; Amgen), IL-1 Ra, was used, that has previously been shown to block biological function of murine IL-1 8,46 . Anakinra (400 μg/mouse in 100 μl of PBS) was administered subcutaneously daily for one week prior to infection of the mice with P. aeruginosa. The mice were treated daily until day 10 post-infection. PBS served as control.
Antibiotics treatment. To test if antibiotic treatment could reverse the negative consequences of P. aeruginosa on our septic arthritis model, mice were treated with ciprofloxacin, which has previously been shown to be effective against P. aeruginosa infection in murine model 47  Clinical arthritis evaluation. Observers blinded to the treatment groups visually examined the presence of arthritis in all four limbs of each mouse. Arthritis was defined as erythema and/or swelling of the joints. To evaluate the severity of arthritis, a clinical scoring system ranging from 0-3 was used, as previously described 8,48 . Bacteriologic examination. The kidneys from the mice were aseptically collected and observers blinded to the treatment groups assessed abscess formation; a scoring system ranging from 0-3 was used, as previously described 8 . Afterwards, the kidneys were homogenized, plated on blood agar plates, and quantified as CFUs.
Radiological evaluation of arthritis by micro-CT. The mice were sacrificed, the joints removed, fixed in 4% paraformaldehyde for 3 days and transferred to PBS for 24 hours. Thereafter, all limb joints were scanned with Skyscan1176 micro-CT (Bruker, Antwerp, Belgium) as previously described 8,9,49 . The projection images were reconstructed into three-dimensional images using NRECON software (version 1.6.9.8; Bruker) and analyzed with CT-Analyzer (version 2.7.0; Bruker). After reconstruction, experienced observers (A.A and T.J) evaluated, in a blinded manner, the extent of bone and cartilage destruction on a grading scale from 0-3, as previously described 8,9,45,49 . Histopathological evaluation of arthritis. After the scanning, representative joints were decalcified, embedded in paraffin and sectioned with microtome. Tissue sections were thereafter stained with haematoxylin and eosin, as previously described 45,49 . Homogenate preparation and bacteriologic examination. The different joint groups were aseptically removed, homogenized with an Ultra Turrax T25 homogenizer (IKA, Staufen, Germany), diluted in PBS, spread on horse blood agar plates, and incubated for 24 hours at 37 °C, as previously described 8,45 . Viable counts of bacteria were performed and quantified as CFUs. A cutoff-point of 10 CFUs was applied, joints with more than 10 CFUs were considered positive. The homogenates were centrifuged at 13,000 rpm for 10 minutes and the supernatants were collected for cytokine analysis.
Measurement of cytokine and chemokine levels. DuoSet ELISA Kits (R&D Systems, Abingdon, UK) were used to quantify the levels of chemokine monocyte chemoattractant protein (MCP-1) and cytokines tumor necrosis factor alpha (TNF-α), interleukin 6 (IL-6) and interleukin 10 (IL-10) in supernatants from knee joint homogenates, as well as in serum.
Statistical analysis. For statistical analysis, GraphPad Prism version 7.0b software for Mac (GraphPad software, La Jolla, CA, USA) was used. To assess statistical significances, the Mann-Whitney U test, Fischer's exact test, and Mantel Cox log-rank test, as appropriate, were used. Results are reported as the mean values ± standard error of the mean (SEM), unless indicated otherwise. A p value < 0.05 was considered statistically significant.

Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.