Impact of prophylactic CpG Oligodeoxynucleotide application on ...

0 downloads 0 Views 2MB Size Report
May 8, 2015 - TLR-9 ligand CpG oligodeoxynucleotide type B (CpG ODN) induces a ... In this rat model, prophylactic administration of a single dose of CpG ...
Bone 78 (2015) 194–202

Contents lists available at ScienceDirect

Bone journal homepage: www.elsevier.com/locate/bone

Original Full Length Article

Impact of prophylactic CpG Oligodeoxynucleotide application on implant-associated Staphylococcus aureus bone infection Shneh Sethi a,⁎, Ulrich Thormann b, Ursula Sommer c, Sabine Stötzel c, Walid Mohamed c, Reinhard Schnettler b, Eugen Domann a, Trinad Chakraborty a, Volker Alt b a b c

Institute of Medical Microbiology, Justus-Liebig-University Giessen, 35392 Giessen, Germany Department of Trauma Surgery Giessen, Justus-Liebig-University Giessen, 35392 Giessen, Germany Laboratory of Experimental Trauma Surgery, Justus-Liebig-University Giessen, 35392 Giessen, Germany

a r t i c l e

i n f o

Article history: Received 11 December 2014 Revised 17 April 2015 Accepted 21 April 2015 Available online 8 May 2015 Edited by: Michael Amling Keywords: Chronic bone infection Staphylococcus aureus CpG-ODN Implants Cytokines

a b s t r a c t TLR-9 ligand CpG oligodeoxynucleotide type B (CpG ODN) induces a proinflammatory environment. We evaluated the effects of a preoperative CpG ODN application in an implant-associated Staphylococcus aureus bone infection model by monitoring bacterial loads and cytokine and chemokine levels. A total of 95 rats were used in four different groups: CpG ODN group (group 1; n = 25), non-CpG-ODN group (group 2; n = 25); saline pretreatment (group 3; n = 25), and one uninfected group (group 4; n = 20). A single dose of CpG-ODN was administered to the left tibialis anterior muscle 3 days prior to surgery and the tibia midshaft was osteotomized, stabilized by an intramedullary implant and subsequently contaminated with 103 colony forming units (CFUs) of S. aureus in groups 1–3. The osteotomy gap in animals of group 4 was not contaminated with S. aureus and those animals did not receive any pretreatment. CpG ODN administration resulted in significant reduction of the bacterial load in tibia tissue homogenate and on the implant surface on day 1 post-infection compared to non-CpG-ODN pretreatment (p b 0.05; p b 0.05). Reductions in bacterial CFUs, compared to non-treated (saline) controls, were approximately 67% and 77% for bone tissue homogenates and implants. No bacteria were detected in uninfected rats. Early reduction of bacterial CFUs in the tibia was accompanied by increased levels of proinflammatory mediators MIP-2, IL-1β and RANTES in bone tissue milieu of the CpG ODN treated group compared to controls. At day 42 post-infection, bone marrow tissue of rats pretreated with CpG ODN had comparable high bacterial CFU numbers as the non-CpG ODN or saline treated groups. Microbiological analysis of implants removed from CpG ODN treated rats showed high bacterial growth densities on their surfaces which were not different from those observed in controls. In histology, all animals of groups 1–3 showed established infected non-unions. Additionally, inflammatory mediator profiles in bone marrow homogenates of CpG ODN treated rats resembled those seen in infected controls. In this rat model, prophylactic administration of a single dose of CpG ODN, resulted in marked reduction of S. aureus load in the infected tibia during the initial stage of infection but failed to prevent development of chronic infection over time. © 2015 Elsevier Inc. All rights reserved.

Introduction Staphylococcus aureus is an important gram-positive bacterium causing a variety of infections in humans which include many superficial, systemic nosocomial or community acquired infections involving various organs such as the skin and soft tissue, joints, lungs, brain and bone [1]. During recent years, increasing incidence of infections involving methicillin-resistant strains of S. aureus (MRSA) has become a serious clinical problem [2]. Osteomyelitis and implant-associated bone ⁎ Corresponding author at: Institute for Clinical Chemistry and Laboratory Medicine, Klinikum Stuttgart, Kriegsbergstr. 62, 70174 Stuttgart, Germany. E-mail address: [email protected] (S. Sethi).

http://dx.doi.org/10.1016/j.bone.2015.04.030 8756-3282/© 2015 Elsevier Inc. All rights reserved.

infections caused by S. aureus are a major clinical challenge because of the high resistance to traditional antibiotic therapies and host immune defenses and therefore surgical interventions are necessary to eradicate the bacteria [3,4]. Regardless of the route of infection, the bone environment permits bacterial multiplication and typically allows bacteria to develop into surface attached three dimensional structures called biofilms [4,5]. A key feature of the biofilm-associated S. aureus is the switching from the single-celled planktonic (free swimming) stage of growth style into slowly replicating sessile mode of growth composed of surface attached aggregates of microcolonies embedded in mass of selfproduced extracellular polymeric substance also matrix [5–7]. The switching of the planktonic bacterial population to the sessile mode of

S. Sethi et al. / Bone 78 (2015) 194–202

growth is considered to be under genetic control and an intramicrobial cell-to-cell communication system termed “Quorum sensing” (QS) contributes to a variety of functional and phenotypic alterations seen in bacteria residing in the biofilm [8]. It is considered that bacterial biofilms are the cause of most chronic and recurrent infections whereas their planktonic counterparts account for the acute phase of the infections [4,5]. In contrast to the planktonic S. aureus organisms, the biofilm-associated bacterial phenotype avoids killing by host immune defenses and conventional antibiotics [9–11]. It is assumed that this enables the bacteria to cause chronic and recurrent infections which are difficult to treat with conventional antibiotics. Various prophylactic strategies have been evaluated for preventing such infections before biofilms are developed. These include coating or impregnation of implants with antibiotics or antimicrobial peptides [12–16] and the use of antibiotics contained in active release biomaterials [17]. Reportedly, interleukin-12 nanoscale coatings at the implant–tissue interface can be used to treat open-fracture infections [18]. An alternative approach is the stimulation of host immune defenses through vaccination which is currently receiving increasing attention, but as yet, have met with limited success [19]. Previous studies in animal models have shown that S. aureus induces an immediate pro-inflammatory innate immune response at the infection sites that enables the host to protect against planktonic staphylococcal populations (reviewed in [20–22]). Essentially, S. aureus induced innate response is mediated by the activation of toll-like receptors (TLRs) in which multiple secreted chemokines/cytokines and their receptors play a crucial role in the recruitment of different inflammatory cells (predominantly neutrophils and macrophages) at the site of infection. Evidence from both in vitro and in vivo studies have shown that host neutrophils represent the major effector cell type involved in the killing of planktonic S. aureus populations [20–23]. Intriguingly, however, under certain conditions excessive accumulation of neutrophils may also lead to disease progression [24–27]. CpG (cytosine-phosphate-guanosine) oligodeoxynucleotide (CpG ODN) sequences are motifs present in bacterial DNA. Unmethylated CpG ODNs are known to stimulate a variety of host innate and adaptive immune responses including the production of a variety of proinflammatory cytokines and chemokines [28,29]. CpG ODNs are well known for their immunomodulating effects via binding to TLR-9 receptor positive cells and are being explored in various possible applications including protection against a variety of pathogens and as an adjuvant with vaccines which are in clinical trials [28–30]. In the present study, we investigated the impact of TLR 9 ligand CpG ODN treatment in a rat model of implant-associated S. aureus osteomyelitis which closely resembles a contaminated open fracture state and assessed the bacterial load and the production of selected cytokines and chemokines at the site of infection. Because S. aureus organisms undergo a planktonic phase before they colonize the implant and form biofilms [9,10], it was reasoned that CpG ODN induced innate proinflammatory immune functions may clear or at least limit the initial inoculum of planktonic organisms to a level which would not allow for subsequent establishment of infected non-unions.

195

with 0.9% saline (group 3, n = 25); uninfected control (group 4, n = 20). For groups 1–3, 15 and 10 animals were planned for euthanasia on post-op day 1 and post-op day 42, respectively. For the uninfected control group, 10 animals were used for observations of post-operative day 1 and day 42, respectively.

Animals Male Sprague Dawley rats (12 weeks old) weighing around 300 g were obtained from Charles River Laboratories (Sulzfeld, Germany). Rats were housed in a restricted-access room. Food and water were provided ad libitum.

S. aureus strain A clinical strain of S. aureus (EDCC 5055, available under DSM No. 28763) was used in this study. The strain was isolated from a patient with a wound infection and was shown to exhibit strong hemolytic activity and to be an inducer of biofilm [31]. The strain was cultivated in brain heart infusion broth (BHI) at 37 °C under vigorous shaking overnight. A 4 hour subculture of the 1:50 diluted overnight culture was performed. Following these steps, a 1:10 dilution in phosphate-buffered saline (PBS) was made and the number of colony forming units (CFUs) was determined by serial dilutions and standardized at a concentration of 103 CFU/20 μl. The viability of the bacteria and the correct CFU/20 μl of each suspension were verified after inoculation of the bacteria using standard agar plating tests (see below). Synthetic CpG ODNs Phosphorothioate oligodeoxynucleotide (ODN) sequences containing unmethylated CpG motifs and control ODN sequences were as follows (CpG motifs are underlined): CpG ODN (ODN1826, sequence: 5′-tcc atg acg ttc ctg acg tt-3′) and as a negative control the ODN1826control (GpC-DNA, sequence: 5′-tcc atg agc ttc ctg agc tt-3′). The ODNs were provided by Invivogen (San Diego, California, USA) who synthesized and purified the products. The lyophilized ODNs were stored at − 20 °C on arrival and before use were suspended in pyrogen-free saline also supplied by the firm. The ODNs were tested before lyophilisation for bacterial contamination and found negative. Briefly, the testing was done by assaying for the activation of NF-kB (nuclear factor-kappa Beta)/AP-1 (adaptor protein-1) via stimulation of TLR2 and TLR4 in HEK-Blue TLR2 and HEL-Blue TLR4 cell lines respectively. TLR2 recognizes peptidoglycan, lipoteichoic acid and lipoprotein from gram positive bacteria and TLR4 is the receptor for bacterial lipopolysaccharide and lipid A. The cell lines HEL-Blue TLR2 and HEL-Blue TLR4 are designed for monitoring the presence of proteins and endotoxins respectively.

Materials and methods Study design Animal experiments followed the animal welfare act of the national institute of health and guide for care and use of laboratory animals and were carried out after approval by the local regional government and accord with German animal protection laws of the district government of Giessen (Ref. no: V54-19c2015(1)GI20/28Nr.97/2011). A total of 95 rats were used for the study and the animals were randomly assigned into four different treatment groups and into two observation periods (post-operative day 1 or post-operative day 42 (Table 1): CpG-ODN (group 1, n = 25); non-CpG-ODN (group 2, n = 25); vehicle

Table 1 Overview of study design. Group

Treatment

n

Number of animals euthanized at day 1 post-infection

Number of animals euthanized at day 42 post-infection

1 2 3 4

CpG-ODN Non-CpG-ODN Vehicle Uninfected controls

25a 25b 25c 20a

14 13 13 9

10 10 10 10

a

1 rat was lost directly after anesthesia. 1 rat was lost directly after anesthesia and 1 rat was lost due to complex fracture during osteotomy. c 2 rats were lost directly after anesthesia. b

196

S. Sethi et al. / Bone 78 (2015) 194–202

Pretreatment with CpG ODN The effect of CpG ODN pretreatment was assessed in a rat model of S. aureus implant-associated osteomyelitis. The rats were divided into 4 groups (n = 8 per group). Three days prior to surgery and bacterial infection, the rats were treated as follows: each rat in group 1 was injected intramuscularly into the left tibialis anterior muscle with a single dose (500 μg) of CpG ODN; group 2 animals were administered a single dose (500 μg) of non-CpG ODN; group 3 rats received only vehicle (0.9% saline) injection and group 4 rats were included as uninfected controls (no bacterial contamination of the osteotomy and no treatment with CpG ODN). We selected CpG ODN delivery, 3 days prior to subsequent infection with S. aureus because of the reports in the literature which suggest that this period is needed to pre-activate innate immune effectors capable of exerting antimicrobial effects [28,49,50], further detailed in the discussion. Surgical implantation and S. aureus infection After 72 h of pretreatment, osteotomy of the midshaft tibia and simultaneous S. aureus infection was performed as described previously in an infected non-union model mimicking an open fracture situation of a long bone in humans resulting in an infected non-union [31]. Briefly, surgery was carried out under general anesthesia (ketamine (60 mg/kg body weight), xylazine (6 mg/kg bodyweight) and atropine (0.1 mg/kg bodyweight)) under sterile conditions. Following disinfection and draping of the left lower leg the midshaft tibia was osteotomized via an oscillating saw and then stabilized 22 gauge (0.7 mm diameter) cannula (B. Braun, Melsungen, Germany). The osteotomy site of each rat of groups 1–3 was then contaminated with 20 μl of a bacterial suspension containing 103 CFUs of S. aureus strain DSM No. 28763 whereas the uninfected control group 4 group received only saline. The fascia was then sutured with 3/0 vicryl and the lesion was closed with 4/0 prolene and additionally metal clamps were applied. The rats were allowed to full weight bearing by providing stable intraoperative conditions after intramedullary fixation and were held on a free diet and water. The total observation time was 6 weeks and general conditions such as wound healing, weight bearing and temperature were checked by a veterinarian. The total observation time was postoperative day 1 to access the early effect of CpG ODN administration on bacterial load and day 42 so as to determine whether chronic bone infection had been established. Post-operative period and euthanasia General health conditions of the animals, wound healing, weight bearing and temperature were daily monitored by a veterinarian. On post-operative day 1 or day 42 the animals received general anesthesia for blood sample collection before euthanization by CO2. Bone tissue processing for bacterial counts and cytokine analysis The rats assigned in different groups were euthanized at days 1 and 42 post-infection and after stripping of most muscle and connective tissue, bone marrow tissue from each rat was removed and weighed from one tibial half and then homogenized in 1.5 ml PBS. A part of the homogenate from each rat was used immediately for obtaining bacterial CFU counts. The remaining homogenates were stored at −80 °C for cytokine assays. Quantification of bacterial loads Bone marrow tissue homogenates from each rat were serially diluted (10-fold) in sterile PBS and 100 μl of each dilution plated onto BHI

agar plates to quantify the number of CFUs. The plates were incubated at 37 °C for 24 h, CFUs grown on agar plates were counted and the number of CFUs per gram bone tissue for each dilution determined. Results are expressed as CFU per gram bone marrow tissue. Furthermore, at indicated time points, implants were removed aseptically from the infected site, and rolled over BHI agar plates for bacterial analysis. The plates were incubated overnight at 37 °C and bacterial counts performed by using serial 10-fold dilutions. It was observed that the implants removed from the tibias of infected rats on postinfection day 1 developed individual colonies when rolled onto agar surfaces which were easy to count. However, rolling of wires removed from infected tibias on day 42 post-infection produced bacterial lawns all along the wire surfaces and were difficult to quantify. In order to quantify the bacterial growth on implant surfaces, a semiquantitative grading scheme was used. After rolling of implants on BHI agar plates, the overnight bacterial growth at 37 °C was classified as follows: (1 +) represents growth in the first zone where the implants were rolled, (2+) growth in the second zone and (3+) growth in the third zone. Analysis of cytokine levels A custom designed Bio-Plex rat cytokine assay (Bio-Rad, Hercules, CA, USA) was used, according to the manufacturer's instructions, for the simultaneous quantification of a panel of selected cytokines and chemokines at the site of infection (surrounding tissue of the implant). These included, interleukin-1alpha (IL-1alpha), interleukin-1beta (IL-1beta), interleukin-4 (IL-4), interleukin-6 (IL-6), interleukin-10 (IL-10), interleukin-17 (IL-17), interferon gamma (IFN-gamma), macrophage inflammatory protein 2 (MIP-2), regulated on activation, normal T-cell expressed and secreted (RANTES), tumor necrosis factor alpha (TNF-alpha), interleukin-12p40 (IL-12p40), monocyted chemoattractant protein-1 (MCP-1). Briefly, bone tissue samples (see above) were homogenized on ice in sterile PBS containing an EDTA-free protease inhibitor cocktail (Roche Diagnostics, Mannheim, Germany). The homogenates were centrifuged for 15 min at 14,000 ×g at 4 °C. Supernatants were collected and the concentration of total protein in test samples determined by using a standard protein assay (Coomassie Bradford) protein assay kit (Thermo Scientific) and adjusted to 500 μg/ml of total protein. The test samples were incubated on 96-well plate with magnetic beads coated with cytokine/chemokine specific antibodies followed by incubations with detection antibodies and finally with streptavidin–pycoerythrin conjugate. Samples were tested in duplicate along with the provided standards and the values expressed as pg/ml; mean ± SD using the Bioplex 200 system (Bio-Rad). Statistical analysis The infection results were analyzed using ANOVA and post-hoc analyses, with a p value of b0.05 denoting statistical significance. Cytokine data are presented as means ± standard deviation using Students t test, with p values of b 0.05 denoting statistical significance. Histology Tibias from two animals of each group were used for histological analysis. The harvested tibias were fixed in phosphate-buffered 4% paraformaldehyde (Merck, Darmstadt, Germany) at 4 °C for a maximum of three days, washed six times with 0.1 M phosphate buffer (pH 7.2– 7.4) and embedded in Technovit 9100 according to the manufacturers' protocol (Heraeus Kulzer, Hanau, Germany). After embedding, the tibia block was dissected into 4 parts to provide a proximal epiphyseal and a proximal metaphyseal part for transverse grindings and a diaphyseal part or longitudinal grindings. For histological analysis these

S. Sethi et al. / Bone 78 (2015) 194–202

grindings were stained with Toluidine-blue (Merck, Darmstadt, Germany). Enzymehistochemical analysis (TRAP staining): After deplastification with 2-methoxyethyl acetate (Merck, Darmstadt, Germany) three times for 5 min, samples were rehydrated and pretreated with 0.1 M sodium acetate buffer (pH 5.2) and then subsequently incubated in NaphtholAS-TR phosphate in N-N-Dimethyl-formamide (both: Sigma-Aldrich, Taufkirchen, Germany) and sodium tartrate (Merck, Darmstadt, Germany) with Fast Red TR salt (Aldrich Chemical Company, Milwaukee, USA) at 37 °C for 1 h. Hematoxylin (Shandon-Instant-Hematoxylin; Thermo Scientific 1 + 3 in bidest. Water) was used for counterstaining and Kaiser's glycerol gelatine, phenolfree (Roth, Karlsruhe, Germany) for covering. Images were taken using an Axioplan 2 Imaging System (Carl Zeiss, Germany) in combination with a Leica DC500 camera, acquired with the Leica IM 1000 software and scale bars were inserted using Adobe Photoshop CS6. Results Clinical observations 89 of the 95 animals survived the planned observation period and could be evaluated as determined in the Materials and methods section (Table 1). One rat of group 1 died directly after surgery and anesthesia, one rat of group 2 sustained a complex tibia fracture during osteotomy and had to be euthanized and another animal died directly after surgery and anesthesia. Two animals of group 3 and one animal of group 4 also died shortly after the surgical procedure. None of the remaining 89 animals showed signs of sepsis during the entire observation period and the wounds healed uneventfully. Effect of CpG ODN B pretreatment on bacterial loads associated with the implant and surrounding bone tissue Only S. aureus organisms were recovered and enumerated. The data (Fig. 1) indicate that the numbers of bacterial CFUs in bone tissues of infected CpG ODN pretreated rats were significantly reduced at day 1 post-infection when compared to those for non-CpG ODN treated control rats (CpG ODN vs non-CpG ODN; mean ± SD, 3351 ± 1804 CFUs vs 15227 ± 12375) (p b 0.05). Notably, the numbers of CFUs in bone tissues from infected saline treated rats were not reduced and were

197

comparable to the numbers observed in the non-CpG ODN treated control rats (saline vs non-CpG ODN 10,039 ± 7627 vs 15,227 ± 12,375). Additionally, when the implants were examined for bacterial growth on day 1 post-infection, low numbers of CFU counts were detectable on their surfaces when removed from CpG ODN treated rats (119 ± 200 CFUs) compared to the numbers observed on implants originating from non-CpG ODN treated rats (588 ± 473 CFUs) or from saline treated rats (516 ± 410 CFUs) (Fig. 2). At day 42 post-infection substantially higher numbers of CFUs were found to be associated with the bone marrow tissue from CpG ODN pretreated rats (26,170 ± 20,176 CFU) which were comparable to those observed in the bone marrow tissue of infected rats which were either pretreated with non-CpG ODN (positive control) (11,132 ± 10,144 CFU) or saline (negative control) (70,775 ± 87,999 CFU). Microbiological analysis of the implants removed from rats administered CpG ODN, non-CpG ODN or saline revealed no significant differences in bacterial growth densities on their surfaces. In all the three infected groups, the implants displayed bacterial growth densities which ranged between grades 2+ and 3+. No detectable bacterial growth was observed on implant surfaces removed from control uninfected rats. Thus it can be concluded that CpG ODN pretreatment is not effective in curtailing bacterial bone tissue load over longer time periods. Bone tissue cytokine/chemokine profiles at the infection site The quantification of a selected panel of cytokines and chemokines showed that at day 1 post-infection, infected CpG ODN treated rats developed significantly elevated bone tissue levels of the chemokine MIP-2 (p b 0.05) and the cytokine IL-1β (p b 0.05), which is known to activate many cell types, compared to the control group of uninfected rats. Notably, the level of another inflammatory chemokine, RANTES was also elevated in CpG ODN treated rats (Fig. 3). The animals of control groups (uninfected, infected non-CpG ODN treated or saline treated) showed relatively low levels of MIP-2, IL-1β and RANTES. Several other inflammatory mediators tested such as MCP-1, TNF-alpha, IL-12, IL-4, IL-6, IL-17, IL-1alpha, and IFN-gamma were either barely detectable or not present in the bone tissue homogenates of CpG ODN pretreated rats and this was also the case in the control groups (data not shown). Cytokine and chemokine analysis of bone tissue homogenates at day 42 post-infection showed that compared to non-infected controls, the infected implanted rats, irrespective of prior treatment, had increased concentrations of TNF-alpha (p b 0.05), IL-1alpha (p b 0.05), IL-1β

Fig. 1. Pretreatment with CpG ODN reduces early bone tissue associated S. aureus CFU. Rats were infected with 103 CFU EDCC 5055 (DSM 28763) at the osteotomy site after osteotomy and implant stabilization and treatment with CpG ODN, non-CpG ODN and vehicle was administered 3 days prior to infection as detailed under Materials and methods. Bone tissue was recovered at day 1 (A) and day 42 (B) post-infection to quantify bacterial burden. Bacterial load is shown as CFU/g bone marrow tissue and CFU/implant. Bars represents SDs. (*p b 0.05).

198

S. Sethi et al. / Bone 78 (2015) 194–202

Histology

Fig. 2. Implant-associated S. aureus CFUs after prior administration of CpG ODN, non-CpG ODN or only saline. Following infection and respective treatment implants were recovered at day 1 post-infection to quantify bacterial burden. Bacterial load is shown as CFU/ implant. Bars represent SDs. (*p b 0.05).

(p b 0.05), RANTES, MIP-2, MCP-1 (p b 0.05) and IL-17 (Fig. 3). Notably, the analysis also revealed the presence of significant amounts of IL-10, a prototype of a Th2 anti-inflammatory cytokine in animals of all infected groups compared to non-infected animals (p b 0.05).

The osteotomy gap was clear visible in all animals on day 1 which was also still visible on day 42 with the absence of cortical bridging confirming the induction of an infected non-union site in all animals of groups 1–3 (Fig. 4). Only in negative control group (group 4) in which the implant was not contaminated during surgery, bone healing activity with cortical bridging was visible on day 42 (Fig. 5). Accumulations of immunocompetent cells or granulocytes could be found neither in the negative control on post-operative day 1 or day 45 nor in the positive control (group 3) on post-operative day 1. Little amounts of granulocytes and bacteria can be seen in the samples from CpG ODN (group 1) and non-CpG ODN (group 2) treated animals on postoperative day 1. At day 42, immunocompetent cells and bacteria can also be detected in the positive control and in considerable numbers in the CpG ODN and non-CpG ODN treated groups (Fig. 4). Clusters of granulocytes can be found at the osteotomy site, at small bone fragments and bone sequesters and/or in the medullary space on day 42 (Figs. 5A–C). These accumulations are mostly in direct vicinity of bacteria. Bacteria can be either in the intramedullary space, but more often inside of cavities or caverns, where there is only limited or even no contact with granulocytes. Accumulations of immunocompetent cells consist almost entirely of granulocytes and appear homogenously in all groups. TRAP-staining was performed to demonstrate the presence of activated macrophages and osteoclasts. TRAP-positive cells were rarely seen on postoperative day 1 but more frequently on day 42. No TRAP-positive cells could be found within clusters of granulocytes. TRAP-positive cells were predominantly located in areas with bone impairment and bone resorption processes (Fig. 5D). In the samples of post-operative day 42, small amounts of TRAP-positive cells were found in the bone marrow, but never in direct vicinity to accumulations of granulocytes.

Fig. 3. Profiles of selected cytokines/chemokines in the bone marrow tissue of infected rats pretreated with CpG ODN, non-CpG ODN or saline and non-infected rats. Bone marrow tissue from infected CpG ODN, non-CpG ODN and vehicle treated rats as well as from non-infected rats were collected and analyzed for cytokines/chemokines at day 1 and day 42 as detailed under Materials and methods. Bars represent SDs. (*p b 0.05).

S. Sethi et al. / Bone 78 (2015) 194–202

199

Fig. 4. Histology of the rat tibias at osteotomy site stained with toluidine-blue performed for the negative control, CpG ODN and non-CpG ODN treated animals at both time points (day 1 and day 42; longitudinal sections). A–C at day 1/D–F at day 42 post-infection euthanized animals; A + D: positive control; B + E: CpG-ODN treated animals; C + F: non-CpG ODN treated animals. Insert pictures show a higher magnification of the framed area; G: accumulations of granulocytes; double arrow: osteotomy area; bs: bone sequester; c: cortical bone; vb: vital bone; arrows: bacteria. Osteotomy site without sings of infection (I = implant). A) osteotomy site without sings of infection. B) osteotomy site without sings of infection. C) area of former osteotomy with massive signs of bone resorption (x) and infection. D) area of former osteotomy with without bridging of the gap with accumulation of granulocytes bone sequesters as signs for infected non-union. E) accumulation of granulocytes around a bone sequester in the non-healed osteotomy gap as sign of infected non-union.

Discussion Synthetic CpG ODNs are known for their adjuvant role in the induction of antigen-specific immune responses (adaptive immunity) to codelivered antigens and in addition have the potential to induce innate (non-specific) protective effects against certain infections [25–27]. The aim of the present study was to examine whether the prophylactic administration of TLR9 ligand CpG ODN type B would modulate the course of infection in a rat model of S. aureus implant-related chronic disease. The data presented herein reveal that rats prophylactically treated with 500 μg CpG ODN by injecting into the tibialis anterior muscle, 3 days prior to surgical intervention and infection, markedly curtailed the inoculated bacterial numbers early after infection. However, this treatment failed to eliminate and clear the infection over time. Thus, when euthanized on day 42 post-infection the CpG ODN pretreated rats showed no significant differences in the numbers of bacterial CFUs associated with the bone marrow when compared to those of rats which were treated with non-CpG ODN or PBS. The data further suggest that the early reduction of bone tissue and implant-associated bacteria was specific and TLR9 dependent because ODN containing non-CpG motifs was ineffective in reducing the

bacterial CFUs. It may be mentioned that because of the limited number of rats which were used for this preliminary proof of principal study, it was not possible to monitor bacterial CFUs at time points between days 1 and 42 and need to be addressed in future work. It is possible that in our animal model, CpG initiated innate proinflammatory environment which presumably accounts for the observed anti-S. aureus effect at an early time after infection was compromised to some level by the subsequent surgery and infection induced immune actions. Conceivably, a diverse immune environment with various effector cell types and a bone milieu rich in multiple pro- and anti-inflammatory mediators accounts for the incomplete bacterial clearance and this subsequently results in the development of biofilm and chronic infection. Bio-Plex cytokine array analysis results have shown that the early reduction of bacterial load in response to CpG ODN treatment was accompanied by a significant increase (compared to control groups) in the concentrations of the chemokines MIP-2 and RANTES as well as the cytokine IL-1β in the bone marrow homogenate. MIP-2 is a critical neutrophil attracting chemokine and IL-1β plays an important role in triggering neutrophil dependent innate protective immune responses [20,32]. Indeed, in a murine model of post-arthroplasty S. aureus joint infection, Il-1β was found to play a protective role [33]. Others have

200

S. Sethi et al. / Bone 78 (2015) 194–202

Fig. 5. Histology of distribution of bacteria, granulocytes and TRAP positive cells on day 42. A–C: Toluidine blue staining; D: TRAP staining; A: non-CpG ODN treated animal; B: positive control; C: negative control; D: CpG ODN treated animal. Insert pictures show a higher magnification of the framed area; G: accumulations of granulocytes; double arrow: osteotomy area; c: cortical bone; vb: vital bone; gt: granulation tissue. A) Cross section with implant (I = implant) and the surrounding tissue and some bacteria (arrow) inside the hollow space of the implant. B) Longitudinal section of the osteotomy site with persisting gap between the bone fragments and bone sequester inside the medullary space with some bacteria (arrow) and granulocytes surrounding the bone sequester as sign for infected non-union. C) Longitudinal section of the osteotomy site with bone neoformation (nb) in the gap between the bone fragments. D) TRAP positive cells (arrow) located in the bone marrow with direct contact to vital bone.

shown that IL-1β is upregulated in the tissue which surrounds failed joint replacement implants and is involved in periprosthetic osteolysis [34,35]. IL-1β and IL-1α represent 2 of the isoforms of interleukin-1, a proinflammatory cytokine which is produced by a variety of cells in the bone environment including: monocytes, macrophages, osteoblasts and osteoclasts and can induce the proliferation of osteoclast precursors and the activation of mature osteoblasts [36]. The chemokine RANTES which is produced by cell types such as epithelial cells, endothelial cells, fibroblasts and T-lymphocytes has been reported to chemoattract multiple cells types including T-cells, monocytes, NK-cells and dendritic cells [37] and is protective in mucosal infections [38]. Analysis of bone tissue homogenates at post-infection day 42 showed that compared to non-infected controls, infected rats with implants had increased concentrations of TNF-alpha, IL-1 alpha, IL-1β, RANTES, MIP-2, MCP-1 and IL-17. However, there were no significant differences in the levels among rats which were pretreated with CpG ODN, non-CpG ODN or saline. Notably, the levels of the antiinflammatory cytokine IL-10 were also found to be increased in the tissue homogenates of rats in these 3 groups. It has been suggested that IL-10 suppresses infection-stimulated bone resorption in-vivo [39]. The results of the present study are in line with previous investigations reporting protective effects associated with CpG ODN administration against a variety of infections [28–30] including S. aureus mastitis [40]. The results are also reminiscent of a previous study, which used EP67 peptide for treating in a mouse model of catheter-associated S. aureus biofilm infection [41]. The authors reported that administration of EP67 at the time of infection and then at 24 and 48 h post-infection, resulted in significantly reduced local bacterial growth at day 14 post-

infection and attenuated biofilm development but had no effect on established biofilms. EP67 is a C5a receptor agonist, which like CpG ODN is capable of inducing a proinflammatory environment [41]. The types of immune responses induced by CpG ODNs depend on their structural features which determines differences in their uptake and intracellular internalization [42–44]. We selected the type B CpG ODN because of the reports that this ODN type not only can activate humoral immunity (B-cells) but also is a strong inducer of nuclear factor (NF-kB) signaling (a critical transcription factor) and production of cytokines involved in the activation of multiple cell populations [42,43] including bone cells such as osteoblasts [45]. Previous reports examining CpG ODN induced protective effects against different pathogens used both systemic and local routes of administration. Available evidence suggests that although the systemic route of CpG ODN delivery confers effective protection against certain pathogens [28–30], it can also lead to increased disease progression [46,47]. Compartmentalized application of CpG ODN has been reported to be more effective against pathogens causing localized infections such as pneumonia caused by Klebsiella pneumoniae [48]. Importantly, prior reports on CpG ODN induced protection against pathogens have shown kinetic differences. Thus, at least 2–3 days prior treatment with CpG ODN are required for inducing effective protection against bacterial pathogens such as K. pneumoniae [48], Burkholderia pseudomallei [49], Escherichia coli [50], Helicobacter pylori [38] and S. aureus [40]. It has been argued that this time period is needed for the upregulation of the innate immune functions involved in anti-microbial effects [28,49, 50]. Notably, post-infection treatment with CpG ODN against acute

S. Sethi et al. / Bone 78 (2015) 194–202

infections when compared to certain chronic infections has generally proved ineffective [49,50]. We chose to deliver CpG ODN by injecting into the tibialis anterior muscle, because this route has been reported to elicit strong protection against S. aureus in rat model of mastitis [40]. The underlying mechanism by which CpG ODN exerts its immunoprotective effect in the bone marrow following administration in the tibialis muscles is unclear. It has been reported that following subcutaneous or intramuscular application, substances with low molecular weights are primarily absorbed via the capillaries and those with high molecular weights primarily via lymph vessels [51]. It is possible that because of its small size (approximately 6 kDa), CpG ODN directly finds its way in the tibia via capillaries and is then taken up by TLR-9 positive cells. Alternatively, the high dose used may lead to its delivery via lymph vessels. It is important to note that TLR 9 has been demonstrated in multiple immune cell subsets [28,30,42,43] as well as in typical bone cells such as osteoblasts and osteoclast precursors [45,52,53]. The precise identification of the effector cell types and inflammatory mediators, which account for the early reduction of bone tissue bacterial load in response to CpG ODN and the subsequent development of chronic infection remain unclear. Bone is known to consist of multiple cell types including the typical “bone cells” (osteoblasts, osteoclasts, and osteocytes), hematopoietic cells (stem cells, progenitor cells, and megakaryocytes), stromal cells (mesenchymal cells) as well as immune cells such as lymphocytes, macrophages and dendritic cells. Multiple cell-types can produce a variety of cytokines and chemokines in the bone environment, which play a regulatory role in the control of bone repair, remodeling and pathogen clearance [36,54]. We now therefore require detailed analysis, such as immunohistochemical studies to focus on the immune cell types and specific inflammatory cytokines/chemokines important in CpG ODN mediated effects observed in the present model system. Finally, it is important to mention that safety studies on CpG ODNs have reported that injection of high doses by a systemic route (i.v., i.p.) in mice result in toxic effects [55], presumably because of excessive stimulation of the innate immune system [56]. In a rat safety evaluation model, daily intramuscular injections of CpG ODN type B in high doses (150 μg per animal) for 28 days, revealed no macroscopic signs of any abnormalities at the injection site. However, the authors observed local inflammatory cell infiltration, hyperplasia of fibrous tissue at the site of injection and increased inflammatory response in the inguinal lymph nodes [57]. The effective CpG ODN B dose/duration (for instance repeated CpG ODN administrations), route and mode of delivery remain to be established for our rat model of S. aureus implant-related chronic infection. In conclusion, the results of this study have indicated that a significant fraction of the inoculated planktonic bacteria is killed at the infection site before they can adhere/colonize the implant surfaces. This is of clinical significance because it raises the possibility that the prophylactic use of CpG ODN may facilitate the early effective clearance of bacteria during the planktonic stage of infection with conventional antibiotics which otherwise have no beneficial effect.

[3]

[4] [5]

[6] [7] [8] [9] [10] [11] [12] [13]

[14] [15] [16]

[17]

[18]

[19]

[20]

[21] [22] [23] [24] [25] [26] [27] [28] [29]

[30] [31]

Acknowledgments This study was funded by grants of the Deutsche Forschungsgemeinschaft (DFG) awarded to VA (Grant: AL 1368/ 1-1) and SS (Grant: SE 2291/1-1). We thank the German Center for Infection Research (DZIF). We thank Katrin Koslowski for her technical support. References [1] Lowy FD. Staphylococcus aureus infections. N Engl J Med Aug. 20 1998;339(8): 520–32. [2] Carleton HA, Diep BA, Charlebois ED, Sensabaugh GF, Perdreau-Remington F. Community-adapted methicillin-resistant Staphylococcus aureus (MRSA): popula-

[32]

[33]

[34]

[35]

[36] [37]

201

tion dynamics of an expanding community reservoir of MRSA. J Infect Dis Nov. 15 2004;190(10):1730–8. Montanaro L, Speziale P, Campoccia D, Ravaioli S, Cangini I, Pietrocola G, et al. Scenery of Staphylococcus implant infections in orthopedics. Future Microbiol Nov. 2011; 6(11):1329–49. Fux CA, Stoodley P, Hall-Stoodley L, Costerton JW. Bacterial biofilms: a diagnostic and therapeutic challenge. Expert Rev Anti Infect Ther Dec. 2003;1(4):667–83. Archer NK, Mazaitis MJ, Costerton JW, Leid JG, Powers ME, Shirtliff ME. Staphylococcus aureus biofilms: properties, regulation, and roles in human disease. Virulence Sep. 2011;2(5):445–59. Flemming HC, Wingender J. The biofilm matrix. Nat Rev Microbiol Sep. 2010;8(9): 623–33. Otto M. Staphylococcal biofilms. Curr Top Microbiol Immunol 2008;322:207–28. Parsek MR, Greenberg EP. Sociomicrobiology: the connections between quorum sensing and biofilms. Trends Microbiol Jan. 2005;13(1):27–33. Mah TF, O'Toole GA. Mechanisms of biofilm resistance to antimicrobial agents. Trends Microbiol Jan. 2001;9(1):34–9. Stewart PS. Mechanisms of antibiotic resistance in bacterial biofilms. Int J Med Microbiol Jul. 2002;292(2):107–13. Leid JG. Bacterial biofilms resist key host defences. Microbe 2009;4(2):66–70. De SK, Contreras R. Human antimicrobial peptides: defensins, cathelicidins and histatins. Biotechnol Lett Sep. 2005;27(18):1337–47. Schmidmaier G, Lucke M, Wildemann B, Haas NP, Raschke M. Prophylaxis and treatment of implant-related infections by antibiotic-coated implants: a review. Injury May 2006;37(Suppl. 2):S105–12. Giuliani A, Pirri G, Nicoletto SF. Antimicrobial peptides: an overview of a promising class of therapeutics. Cent Eur J Biol 2007;2(1):1–33. Hickok NJ, Shapiro IM. Immobilized antibiotics to prevent orthopaedic implant infections. Adv Drug Deliv Rev Sep. 2012;64(12):1165–76. Alt V, Kirchhof K, Seim F, Hrubesch I, Lips KS, Mannel H, et al. Rifampicin-– fosfomycin coating for cementless endoprostheses: antimicrobial effects against methicillin-sensitive Staphylococcus aureus (MSSA) and methicillin-resistant Staphylococcus aureus (MRSA). Acta Biomater Oct. 2014;10(10):4518–24. McMillan DJ, Lutton C, Rosenzweig N, Sriprakash KS, Goss B, Stemberger M, et al. Prevention of Staphylococcus aureus biofilm formation on metallic surgical implants via controlled release of gentamicin. J Biomed Sci Eng 2011;4:535–42. Li B, Jiang B, Boyce BM, Lindsey BA. Multilayer polypeptide nanoscale coatings incorporating IL-12 for the prevention of biomedical device-associated infections. Biomaterials May 2009;30(13):2552–8. Lattar SM, Noto LM, Denoel P, Germain S, Buzzola FR, Lee JC, et al. Protein antigens increase the protective efficacy of a capsule-based vaccine against Staphylococcus aureus in a rat model of osteomyelitis. Infect Immun Jan. 2014; 82(1):83–91. Sethi S, Chakraborty T. Role of TLR-/NLR-signaling and the associated cytokines involved in recruitment of neutrophils in murine models of Staphylococcus aureus infection. Virulence Jul. 2011;2(4):316–28. Miller LS, Cho JS. Immunity against Staphylococcus aureus cutaneous infections. Nat Rev Immunol Aug. 2011;11(8):505–18. Hanke ML, Kielian T. Deciphering mechanisms of staphylococcal biofilm evasion of host immunity. Front Cell Infect Microbiol 2012;2:62. Rigby KM, DeLeo FR. Neutrophils in innate host defense against Staphylococcus aureus infections. Semin Immunopathol Mar. 2012;34(2):237–59. Nathan C. Neutrophils and immunity: challenges and opportunities. Nat Rev Immunol Mar. 2006;6(3):173–82. Cheng OZ, Palaniyar N. NET balancing: a problem in inflammatory lung diseases. Front Immunol 2013;4:1. Craig A, Mai J, Cai S, Jeyaseelan S. Neutrophil recruitment to the lungs during bacterial pneumonia. Infect Immun Feb. 2009;77(2):568–75. Thwaites GE, Gant V. Are bloodstream leukocytes Trojan Horses for the metastasis of Staphylococcus aureus? Nat Rev Microbiol Mar. 2011;9(3):215–22. Krieg AM. Therapeutic potential of Toll-like receptor 9 activation. Nat Rev Drug Discov Jun. 2006;5(6):471–84. Kanzler H, Barrat FJ, Hessel EM, Coffman RL. Therapeutic targeting of innate immunity with Toll-like receptor agonists and antagonists. Nat Med May 2007; 13(5):552–9. Klinman DM. Immunotherapeutic uses of CpG oligodeoxynucleotides. Nat Rev Immunol Apr. 2004;4(4):249–58. Alt V, Lips KS, Henkenbehrens C, Muhrer D, Oliveira Cavalcanti MC, Sommer U, et al. A new animal model for implant-related infected non-unions after intramedullary fixation of the tibia in rats with fluorescent in situ hybridization of bacteria in bone infection. Bone May 1 2011;48(5):1146–53. Miller LS, O'Connell RM, Gutierrez MA, Pietras EM, Shahangian A, Gross CE, et al. MyD88 mediates neutrophil recruitment initiated by IL-1R but not TLR2 activation in immunity against Staphylococcus aureus. Immunity Jan 2006;24(1):79–91. Bernthal NM, Pribaz JR, Stavrakis AI, Billi F, Cho JS, Ramos RI, et al. Protective role of IL-1beta against post-arthroplasty Staphylococcus aureus infection. J Orthop Res Oct. 2011;29(10):1621–6. Epstein NJ, Warme BA, Spanogle J, Ma T, Bragg B, Smith RL, et al. Interleukin-1 modulates periprosthetic tissue formation in an intramedullary model of particleinduced inflammation. J Orthop Res May 2005;23(3):501–10. St Pierre CA, Chan M, Iwakura Y, Ayers DC, Kurt-Jones EA, Finberg RW. Periprosthetic osteolysis: characterizing the innate immune response to titanium wear-particles. J Orthop Res Nov. 2010;28(11):1418–24. Lorenzo J, Horowitz M, Choi Y. Osteoimmunology: interactions of the bone and immune system. Endocr Rev Jun. 2008;29(4):403–40. Rollins BJ. Chemokines. Blood Aug. 1 1997;90(3):909–28.

202

S. Sethi et al. / Bone 78 (2015) 194–202

[38] Raghavan S, Nystrom J, Fredriksson M, Holmgren J, Harandi AM. Orally administered CpG oligodeoxynucleotide induces production of CXC and CC chemokines in the gastric mucosa and suppresses bacterial colonization in a mouse model of Helicobacter pylori infection. Infect Immun Dec. 2003;71(12):7014–22. [39] Sasaki H, Hou L, Belani A, Wang CY, Uchiyama T, Muller R, et al. IL-10, but not IL-4, suppresses infection-stimulated bone resorption in vivo. J Immunol Oct. 1 2000; 165(7):3626–30. [40] Zhu YM, Miao JF, Fan HJ, Zou SX, Chen WH. Protective effect of CpG-DNA against mastitis induced by Staphylococcus aureus infection in a rat model. Int Immunopharmacol Apr. 2007;7(4):435–43. [41] Hanke ML, Heim CE, Angle A, Sanderson SD, Kielian T. Targeting macrophage activation for the prevention and treatment of Staphylococcus aureus biofilm infections. J Immunol Mar. 1 2013;190(5):2159–68. [42] van Ojik HH, Bevaart L, Dahle CE, Bakker A, Jansen MJ, van Vugt MJ, et al. CpG-A and B oligodeoxynucleotides enhance the efficacy of antibody therapy by activating different effector cell populations. Cancer Res Sep. 1 2003;63(17):5595–600. [43] Gursel M, Verthelyi D, Gursel I, Ishii KJ, Klinman DM. Differential and competitive activation of human immune cells by distinct classes of CpG oligodeoxynucleotide. J Leukoc Biol May 2002;71(5):813–20. [44] Sasai M, Linehan MM, Iwasaki A. Bifurcation of Toll-like receptor 9 signaling by adaptor protein 3. Science Sep. 17 2010;329(5998):1530–4. [45] Zou W, Amcheslavsky A, Bar-Shavit Z. CpG oligodeoxynucleotides modulate the osteoclastogenic activity of osteoblasts via Toll-like receptor 9. J Biol Chem May 9 2003;278(19):16732–40. [46] Xin L, Shelite TR, Gong B, Mendell NL, Soong L, Fang R, et al. Systemic treatment with CpG-B after sublethal rickettsial infection induces mouse death through indoleamine 2,3-dioxygenase (IDO). PLoS One 2012;7(3):e34062. [47] Ito S, Pedras-Vasconcelos J, Klinman DM. CpG oligodeoxynucleotides increase the susceptibility of normal mice to infection by Candida albicans. Infect Immun Sep. 2005;73(9):6154–6.

[48] Deng JC, Moore TA, Newstead MW, Zeng X, Krieg AM, Standiford TJ. CpG oligodeoxynucleotides stimulate protective innate immunity against pulmonary Klebsiella infection. J Immunol Oct. 15 2004;173(8):5148–55. [49] Wongratanacheewin S, Kespichayawattana W, Intachote P, Pichyangkul S, Sermswan RW, Krieg AM, et al. Immunostimulatory CpG oligodeoxynucleotide confers protection in a murine model of infection with Burkholderia pseudomallei. Infect Immun Aug. 2004;72(8):4494–502. [50] Gomis S, Babiuk L, Godson DL, Allan B, Thrush T, Townsend H, et al. Protection of chickens against Escherichia coli infections by DNA containing CpG motifs. Infect Immun Feb. 2003;71(2):857–63. [51] Ballard BE. Biopharmaceutical considerations in subcutaneous and intramuscular drug administration. J Pharm Sci Mar. 1968;57(3):357–78. [52] Zou W, Schwartz H, Endres S, Hartmann G, Bar-Shavit Z. CpG oligonucleotides: novel regulators of osteoclast differentiation. FASEB J Mar. 2002;16(3):274–82. [53] Hayashi S, Yamada T, Tsuneto M, Yamane T, Takahashi M, Shultz LD, et al. Distinct osteoclast precursors in the bone marrow and extramedullary organs characterized by responsiveness to Toll-like receptor ligands and TNF-alpha. J Immunol Nov. 15 2003;171(10):5130–9. [54] Cho SW, Soki FN, Koh AJ, Eber MR, Entezami P, Park SI, et al. Osteal macrophages support physiologic skeletal remodeling and anabolic actions of parathyroid hormone in bone. Proc Natl Acad Sci U S A Jan. 28 2014;111(4):1545–50. [55] Yamamoto G, Kobayashi H, Hikichi T, Irisawa A, Obara K, Ohira H. Synthetic oligodeoxynucleotides induce gastritis in mice. Fukushima J Med Sci Jun. 2009; 55(1):23–31. [56] Heikenwalder M, Polymenidou M, Junt T, Sigurdson C, Wagner H, Akira S, et al. Lymphoid follicle destruction and immunosuppression after repeated CpG oligodeoxynucleotide administration. Nat Med Feb. 2004;10(2):187–92. [57] Liu L, Shen L, Liu X, Yu Y, Li Y, Wang L, et al. A safety study of a B-class CpG ODN in Sprague-Dawley rats. J Appl Toxicol Jan. 2012;32(1):60–71.