Combinational therapy with antibiotics and antibiotic-loaded adipose-derived stem cells reduce abscess formation in implant-related infection in rats

Abstract

Implant-related infection is difficult to treat without extended antibiotic courses. However, the long-term use of antibiotics has led to the development of multidrug- and methicillin-resistant Staphylococcus aureus. Thus, alternatives to conventional antibiotic therapy are needed. Recently, mesenchymal stem cells have been shown to have antimicrobial properties. This study aimed to evaluate the antimicrobial activity and therapeutic effect of local treatment with antibiotic-loaded adipose-derived stem cells (ADSCs) plus an antibiotic in a rat implant-associated infection model. Liquid chromatography/tandem mass spectrometry revealed that ADSCs cultured in the presence of ciprofloxacin for 24 h showed time-dependent antibiotic loading. Next, we studied the therapeutic effects of ADSCs and ciprofloxacin alone or in combination in an implant-related infection rat model. The therapeutic effects of ADSCs plus antibiotics, antibiotics, and ADSCs were compared with no treatment as a control. Rats treated with ADSCs plus ciprofloxacin had the lowest modified osteomyelitis scores, abscess formation, and bacterial burden on the implant among all groups (P < 0.05). Thus, local treatment with ADSCs plus an antibiotic has an antimicrobial effect in implant-related infection and decrease abscess formation. Thus, our findings indicate that local administration of ADSCs with antibiotics represents a novel treatment strategy for implant-associated osteomyelitis.

Figure 1

Cell viability based on the measurement of mitochondrial oxidative activity after exposure CPFX for 24 h (A) or 7 days (B). Bars represent the median ± interquartile range of the percentages of the control (100%). (C) Alizarin red and ALP staining of ADSCs and ADSCs loaded with 100 mg/L CPFX (ADSCs-ant) at 2 weeks. (D–F) RT-qPCR results. ALP, alkaline leukocyte phosphatase (D), osteocalcin (E), rCRAMP, rat cathelicidin-related antimicrobial peptide (F). There was a significant difference in osteocalcin, whereas there were no significant differences in ALP and rCRAMP between ADSCs and ADSCs-ant.

Figure 2

Evaluating the antibiotic loading and releasing ability of ADSCs. (A–C) CPFX concentrations in cells or CM determined by LC–MS/MS at the indicated time points after CPFX loading or release. (A) CPFX concentrations after loading onto ADSCs and BMSCs. *P < 0.05; ns, no significant difference. The green bar represents a control without cells to confirm that CPFX did not adsorb to the plate. (B) CPFX concentrations released by cells in CM. (C) CPFX concentrations in cells after release. The broth dilution method was used to evaluate the antimicrobial activity of ADSCs-ant and CM on S. aureus. The protocol is shown in (D), the results in (E). Row B: serial 1:2 dilution of ADSCs-ant, Row C: serial 1:2 dilution of CM of ADSCs-ant, Row D: serial 1:2 dilution of ADSCs, Row E: serial 1:2 dilution of CM of ADSCs, Row F: serial 1:2 dilution of CPFX (starting stock solution 128 mg/l; MIC at 1:2 dilution = 0.125 mg/l), Row G: control bacterial growth in medium without CPFX. (F) ADSCs-ant induced visible growth inhibition at a dilution of 1:8 (row B), and the CM of ADSCs-ant induced growth inhibition at a dilution of 1:2 (row C). Growth inhibition was not induced by ADSCs and their CM (rows D and E).

Figure 3

Evaluation of the modified osteomyelitis score. (A) No treatment, (B) ADSCs-ant, (C) antibiotic, (D) ADSCs. (A) Soft tissue swelling. (B) Abscess formation. The yellow arrow indicates abscess formation. (C) The ADSCs-ant group showed the lowest modified osteomyelitis score among all groups. *P < 0.05 by Kruskal–Wallis test followed by Dunn’s post-hoc test.

Figure 4

μCT imaging analysis of the femur of rats (LaTheta). (A) μCT image of a representative femur of each group; (1) no treatment, (2) ADSCs-ant, (3) antibiotic, (4) ADSCs. (B) Evaluation of the Hounsfield Unit value using the DICOM viewer software (Synapse Vincent). (C) Healthy bone ratio at the proximal screw in all groups. *P < 0.05 versus no treatment and ADSCs; **P < 0.001 versus no treatment and ADSCs; ns, no significant difference by Kruskal–Wallis test followed by Dunn’s post-hoc test. No treatment: mean 0.57, SD 0.07, range 0.47–0.72; ADSCs-ant: mean 0.64, SD 0.05, range 0.52–0.73; antibiotic: mean 0.62, SD 0.06, range 0.53–0.74; ADSCs: mean 0.56, SD 0.06, range 0.46–0.66. (D) Healthy bone ratio at the distal screw in all groups. There were no significant differences between groups by Kruskal–Wallis test followed by Dunn’s post-hoc test.

Figure 5

Histological analysis. (A–C) Microscopic images of formalin-fixed, H&E-stained paraffin sections. Ab indicates abscess formation. Cb indicates cortical bone. The white arrow indicates partial disappearance of the cortical bone. The black arrow indicates necrosis of the cortical bone. Ca indicates cancellous bone. (D) The abscessed area in total area was evaluated in three regions, including at the distal screw hall, the proximal screw hall, and the region between both screw halls. (E) Abscessed area in total area in all groups. The abscessed area was significantly lower in the ADSCs-ant group than in the no-treatment group (P < 0.05 by ordinary one-way ANOVA followed by Sidak’s post-hoc test).

Figure 6

Effects of ADSCs and antibiotic on bacterial burden on the implant and in soft tissue as determined by CFU analysis. (a) No treatment, (b) ADSCs-ant, (c) antibiotic, (d) ADSCs. #P < 0.05, by Kruskal–Wallis test followed by Dunn’s post-hoc test; *P < 0.05, by ordinary one-way ANOVA followed by Sidak’s post-hoc test; **P < 0.05 by Brown-Forsythe and Welch ANOVA followed by Dunnett’s T3 post-hoc test.

Figure 7

Protocol for implant-related infection model establishment in rats.