Treatment of osteomyelitis is challenging. Serial débridements and intravenous antibiotics are the standard treatment for an orthopaedic infection. Unfortunately, intravenous antibiotic therapy with several antibiotics requires careful monitoring of the serum concentrations to maintain nontoxic systemic levels. In addition, the blood supply to the infected tissue may be compromised, which would prevent sufficient local antibiotic levels.14,22,28,39 For these reasons, interest in local administration of antibiotics has been increasing.
These locally administered antibiotics may achieve greater than a 20-fold6 antibiotic concentration in local tissue over intravenous administration and seem clinically promising. However, the potential toxic effects of high local antibiotic concentrations on bone and its healing process have not been fully determined. Studies have shown some antibiotics are toxic to osteoblastlike cells, especially when they are present in high concentrations.5,9,10,19,20
During the bone healing process, not only are osteo- blasts present, but there also are undifferentiated mesenchymal stem cells (MSCs), which migrate into the bone- healing site, proliferate, and differentiate.26,38 Any failure in mobilization, proliferation, and differentiation of these progenitor cells will lead to failure in new bone formation and to eventual nonunion.38 Therefore, the potential toxic effect of antibiotics on these progenitor cells may compromise the bone-healing process.
Gentamicin is a potent bactericidal antibiotic against a wide range of organisms. Gentamicin-containing cement implantation was one of the earliest and is one of the most recognized treatments for skeletal infections.8,32 An in vitro experiment measuring elution of the drug around the beads showed levels of 400 to 600 μg/mL the first day.14 With such high concentrations of gentamicin, a potential adverse effect on the bone-healing process may occur. However, information has not been available regarding the possible direct effects of gentamicin on human MSCs (hMSCs), one of the major cells contributing to bone healing.
We determined whether the high concentration of gentamicin achieved after local administration would have toxic effects on the cell viability, proliferation, or osteochondrogenic capacity of hMSCs.
MATERIALS AND METHODS
Human mesenchymal stem cells were isolated from bone marrow under standardized conditions from three healthy donors and used for investigation of toxic effects of gentamicin on hMSCs. After 7 days of gentamicin exposure, we evaluated cell viability, proliferation, and in vitro osteochondrogenic capacity of hMSCs. Some cells from each gentamicin condition were loaded into porous ceramic vehicles and implanted in immunocompromised mice for an in vivo osteochondrogenic assay.
To avoid potential variations caused by gender, age, or race, bone-marrow aspirates were obtained from three healthy Caucasian adult donors of similar age (male, 23-27 years of age) after informed consent was obtained under the terms of an Institutional Review Board-approved protocol.
We established hMSC cultures using previously published methods.7,17
Twenty milliliters of the bone marrow aspirates were enriched for hMSCs by density gradient centrifugation over a Percoll® (Sigma-Aldrich, St Louis, MO) cushion. We determined selective cell adhesion to culture surfaces in the presence of a selected batch of 10% fetal bovine serum.17 The first passage cells were seeded on 60-mm culture dishes or six-well culture dishes at a density of 5 × 103 cells/cm2 and cultured for 7 days with or without gentamicin (0 μg/mL, 50 μg/mL, 100 μg/mL, or 200 μg/mL). After 7 days of gentamicin exposure, we evaluated cell viability, proliferation, and in vitro and in vivo osteochondrogenic capacity. Triplicate cultures were set up for each concentration of antibiotic studied for each cell preparation.
Seven days after culture, we determined suspended trypsin- released cell viability (Passage 2) with a hemacytometer by a trypan blue exclusion test and estimated cell proliferation by total DNA content.
For the DNA assay, 1 mL of 0.1 N NaOH was added per well of the six-well culture dish and then neutralized with 0.1 N HCl in 5 mol/L NaCl and 100 mmol/L NaH2PO4. This mixture was combined with 1 mL of 0.7 μg/mL Hoechst 33258 (Sigma) in water. Fluorescence was read at an excitation of 360 nm and emission of 460 nm and compared with a certified Calf Thymus DNA standard (Sigma). Differences in DNA content are assumed to reflect differences in cell number.
The effects of gentamicin on in vitro osteogenic capacity of hMSCs were evaluated by the expression of alkaline phosphatase (ALP), matrix calcium deposition, and total DNA content after inducing osteogenic differentiation in monolayer cultures. In brief, the hMSCs (Passage 1), after exposure to different concentrations of gentamicin for 7 days, were trypsin-released, counted, and seeded in six-well culture dishes at a density of 3 × 103 cells/cm2. In vitro osteogenesis was induced with osteogenic induction medium (DMEM-LG with 10% fetal bovine serum, 0.1 μmol/L dexamethasone (Sigma), and 0.05 mmol/L ascorbate-2-phosphate (WAKO, Richmond, VA), referred to as OS medium.11 On Day 10, 2 mmol/L β-glycerolphosphate (Sigma) was added to the OS medium. Alkaline phosphatase activity was assayed at Days 0, 4, 8, 10, 12, and 16; total DNA content was assayed at Days 0, 4, 8, 12, 16, 20, 24, and 28; and matrix calcium deposition/mineralization was assayed at Day 28 after culture in control or OS medium. The results for ALP activity and calcium content are presented as per well and also normalized to DNA content.
For ALP assay, 1 mL of a 1 mg/mL solution of ALP substrate (p-nitrophenyl phosphate; Sigma) in a 50 mmol/L glycine buffer and 1 mmol/L MgCl2·6H2O was added per well of a six-well dish. After 3 minutes, the solution was removed and transferred to a tube containing an equal volume of 1 mol/L NaOH. The absorbance of the resulting solution was read at 405 nm and was compared with a series of dilutions of p-nitrophenol (Sigma). After the ALP assay solution was removed, the cultures were stored at −70°C until DNA quantification.
For the calcium assay, the calcium in the osteogenic hMSC culture was extracted with 0.6 mol/L HCl on Day 28 of culture. Aliquots of the extract were mixed with reagent from a commercial calcium assay kit (Biotron, Hemet, CA) and the absorbance was read at 575 nm. Calcium concentration was determined with a standard curve generated from a series of dilutions of CaCl2.17
The chondrogenic capacity of cells from each study group was evaluated in an aggregate culture system.29,37 Cells (Passage 1) first were expanded in monolayer cultures in the presence or absence of gentamicin and then passaged for aggregate culture. In brief, cells were trypsinized, counted, and resuspended in chondrogenic medium (DMEM-HG supplemented with 1% ITS + Premix™ [BD, Franklin Lakes, NJ], 100 umol/L ascorbate-2- phosphate, 10−7 mol/L dexamethasone, 0.1 mmol/L nonessential amino acids [Gibco, Carlsbad, CA], 1 mmol/L sodium pyruvate [Gibco], and 10 ng/mL transforming growth factor-β1 [R&D Systems, Minneapolis, MN]) at a density of 1.25 × 106 cells/mL. Aliquots containing 2.5 × 105 cells (Passage 2) were placed in polypropylene 96-well plates.29 On Day 21, triplicate aggregates from each group were processed for histologic and immunohistologic evaluations or for glycosaminoglycan and DNA quantifications.
The aggregate glycosaminoglycan content was quantified by a previously described method.2 Aggregates were digested with papain.23 Twenty-five-μL aliquots of the papain-digested extracts were mixed with 250 μL of 0.02% safranin O (Sigma). The mixtures were filtered through a nitrocellulose membrane with a dot blot apparatus (Bio-Rad, Hercules, CA). The individual dots were cut from the filter, and incubated in 10% cetylpyridinium chloride. The absorbance of these extracts was read at 536 nm and was compared with chondroitin sulfate standards (Seikagaku America, Falmouth, MA). We also determined the aggregate DNA content, measured as described above. To additionally evaluate chondrogenesis in the aggregates, samples were formalin-fixed and paraffin- embedded. Adjacent 7-μm sections were stained for collagen Types II and X.
To evaluate the in vivo osteochondrogenic potential of cells from each group, hMSC-loaded ceramic cubes were implanted in immunocompromised mice as described previously.3,16 Briefly, hMSCs (Passage 1), after culture for 7 days in control medium with or without gentamicin (0 μg/mL, 50 μg/mL, 100 μg/mL, or 200 μg/mL), were trypsin-released, and loaded into ceramic cubes. Cell-loaded cubes were implanted subcutaneously on the dorsal surface of the immunocompromised mice.
Host animals were sacrificed after 6 weeks. Ceramic cubes were removed, fixed, decalcified, embedded, cut into 5-μm- thick sections, and stained with toluidine blue. Stained sections were examined for the presence of bone or cartilage. All sections were viewed by two individuals who were blinded to the sample identity. The total number of pores in each ceramic implant and the number of pores containing bone or cartilage were determined for each histologic section, and the percentage of bone- positive or cartilage-positive pores was calculated.16
Each continual variable resulting from individual experiments was expressed as mean ± SEM of triplicates and compared with the results obtained for hMSCs cultured in the absence of gentamicin. Data were entered and analyzed with an SPSS statistical program (SPSS Inc, Chicago, IL). We used Student's t test to assess the difference between the groups for each continuous variable. For the t test, each variable for each group, which was treated with different concentrations of gentamicin, was compared with the group treated without gentamicin. Before the analysis, the p value was set at 0.05 for each test.
Gentamicin did not elicit an adverse effect on the viability of hMSCs at the concentrations tested (Fig 1A). In contrast, high concentrations of gentamicin inhibited cell proliferation. The DNA contents of the cultures were decreased (p < 0.05) to 76% ± 22% and 50% ± 25% for cells cultured with 100 μg/mL and 200 μg/mL gentamicin, respectively, compared with the hMSCs cultured in medium without gentamicin (Fig 1B).
Gentamicin inhibited (p < 0.05) cell proliferation and ALP activity in vitro during the early culture phase (Days 4 and 8). Total DNA content was lower (p < 0.05) in cells preexposed to gentamicin at 100 μg/mL and 200 μg/mL compared with cells preexposed in medium without gentamicin (Fig 2A). However, this difference was not observed for cells cultured in control medium (Fig 2B). Decreases (p < 0.05) in ALP activity were detected during the early stage of osteogenesis (Days 4 and 8) in a dose- dependent manner in hMSCs preexposed to gentamicin at 50 μg/mL and greater in OS medium (Fig 2C). However, at all assay times, ALP activity was similar between groups for data normalized to DNA content (Fig 2D). Analysis of the culture matrix calcium deposition showed no differences between each group in the amounts of calcium content per well or normalized to DNA content (data not shown).
For in vitro chondrogenesis, the high concentrations of gentamicin had an adverse effect on the chondrogenic differentiation of hMSCs. The pellet DNA content was similar between the groups in the amount of DNA per pellet at the times analyzed (Fig 3A). In contrast, analysis of the glycosaminoglycan content showed pellets made with cells expanded in gentamicin at 100 μg/mL and 200 μg/mL contained less (p < 0.05 ) glycosaminoglycan per pellet (Fig 3B) and normalized to DNA content (Fig 3C). Type II and Type X collagen were detected throughout the entire extracellular matrix of pellets made with cells expanded in 0 or 50 μg/mL gentamicin, except for the peripheral layers (Fig 4). In contrast, these molecules were detected in only a portion of the extracellular matrix of pellets made with cells expanded in 100 or 200 μg/mL gentamicin (Fig 4H, I, K, L).
High concentrations of gentamicin also showed adverse effects on in vivo osteochondrogenic differentiation. A greater (p < 0.05) amount of bone was present in ceramic cubes loaded with hMSCs expanded with the lower concentrations of gentamicin (0 μg/mL and 50 μg/mL) than in cubes loaded with cells cultured in high concentrations of gentamicin (100 μg/mL and 200 μg/mL) (Table 1). Most of the osseous material was deposited along the pores located in the peripheral regions of the ceramic cubes (Fig 5A). Some empty space could be found in the ceramic cube. This is because of dissolution of ceramic material during the decalcification step (Fig 5B). Only a small amount of cartilage, which is identified by large round cells surrounded by matrix (Fig 5C), is produced by hMSCs in this model, as indicated by cartilage-containing pores representing a small proportion of pores identified as bone-positive or cartilage-positive. Table 1 provides the percent distribution for cartilage and bone in these specimens.
Local administration of antibiotics is well established for treatment of osteomylitis.14,15,22 After an infection is resolved, local antibiotics usually are removed and the bone- healing process resumes to restore the normal physiologic function.15,22 In this study, we investigated potential adverse effects of antibiotic preexposure on the sequence of bone healing after treatment with local antibiotics. We used human mesenchymal stem cells, the major cells that contribute to the bone-healing process,26,38 from healthy adults to assess the effects of gentamicin on osteochondrogenesis. The results indicate gentamicin does not have cytotoxicity on hMSCs at the concentrations tested, but exerted adverse effects on the proliferation and differentiation of hMSCs at concentrations of 100 μg/mL and 200 μg/mL. Gentamicin at these concentrations inhibits in vitro differentiation, as indicated by glycosaminoglycan content, Type II and Type X collagen deposition, cell proliferation, and ALP activity expression. The inhibitory effects were exerted on osteogenesis and chondrogenesis, although cells were consistently more sensitive to the inhibitory effects on in vitro chondrogenesis and early stages of osteogenesis. Although the marrow samples were derived from independent human donors, similar responses were detected with respect to the in vitro functional characteristics analyzed in this study. Comparable effects were observed when cells were loaded into porous ceramic vehicles and implanted into an in vivo osteogenic assay.
Our study has several limitations. The marrow donors' average age was 25 years (range, 23-27 years). The osteogenic potential of hMSCs from this age group might be greater compared with cells derived from older adults,4,21 but it has been reported the osteogenic capacity is maintained during aging.12,30 The effect of aging on the osteogenic potential of hMSCs is an open question. In addition, hMSCs do not produce a copious amount of cartilage in this in vivo model. We were unable to estimate the direct effects of gentamicin on in vivo chondrogenesis of hMSCs. Also, to maintain an exponential proliferation rate and prevent spontaneous differentiation potential17 and to mimic the clinical practice for which local antibiotics usually are removed within weeks after the implantation,14,15,22 hMSCs were cultured with or without gentamicin for 7 days, before they reached confluence. In a clinical setting, the elution of local antibiotic usually follows first-order kinetics.27,33,34 Therefore, based on our results, we do not know whether the gentamicin produces any different effects on hMSCs with a longer or shorter exposure period or with kinetically changing concentrations.
Osteomyelitis is a serious postoperative complication in orthopaedic surgery. Serial débridements combined with intravenous and local antibiotics are the standard treatment for properly selected patients.15,39 Implantable gentamicin-containing polymethylmethacrylate (PMMA) beads have been one of the most widely used local antibiotics in clinical practice.33,34 The reliability and efficiency of local gentamicin-containing PMMA bead therapy for osteomyelitis have been documented.32,33 This technique delivers high local concentrations of gentamicin while avoiding systemic toxicity. The popularity of gentamicin beads is related to the high and sustained release of the antibiotic and the commercial availability of the beads (Septopal, Biomet-Merck, Bridgend, UK). In one canine animal model study, gentamicin-impregnated PMMA bead implantation resulted in a wound hematoma concentration as high as 212 μg/mL after 7 days.34 In a clinical study with gentamicin beads, wound-effluent concentration of the antibiotic reached 345.6 μg/mL.34 These findings imply, during local antibiotic administration in the clinical setting, intrinsic hMSCs at these sites are exposed to high concentrations of gentamicin, which exceeds the critical osteochondrogenesis-affecting level of 100 μg/mL.
Most of the safety concerns of gentamicin therapy are focused on the serum level because of the potential nephrotoxicity and ototoxicity.13 Results of our studies, reported here, imply the high concentrations of antimicrobial agents in local tissues achieved by local administration may raise another potential safety concern, even if the systemic absorption of the locally delivered gentamicin is limited and results in extremely low serum levels.27,34 Some investigators have examined the effects of antibiotics on local tissue, such as fibroblasts35 and osteoblasts (Table 2).5,9,10,19,20 Tobramycin and gentamicin have been shown to substantially decrease the growth of osteoblast- like cells at concentrations of 400 μg/mL and 100 μg/mL, respectively.10,19 Similarly, rifampin, ciprofloxacin, cefazolin, and vancomycin also have been shown to substantially decrease cell growth at concentrations of 3 μg/mL, 20 μg/mL, 200 μg/mL, and 1000 μg/mL, respectively.5,9,20 These experiments were performed with a human osteosarcoma cell line (MG63)5,19,20 or with osteoblastlike cells isolated from human cancellous bone.9,10 However, these data were not sufficient to provide a full understanding of the effects of antibiotics on the bone- healing process.
Mesenchymal stem cells play an important role in studies of the bone-healing process, not only because of the importance of transforming MSCs into chondrocytes and osteoblasts in the repair process,26,38 but also because MSCs provide a useful model for evaluating the multiple factors responsible for the progression of cells from un- differentiated precursors to osteoblasts and eventually terminally differentiated osteocytes.11 Additionally, the majority of fracture repair is effected by secondary bone healing,18 which is mediated by endochondral ossification following the sequence of fracture-site early cartilage formation, chondrocyte hypertrophy, cartilage mineralization, and eventual secondary ossification.24 The similarity between the in vitro hMSC chondrogenic pellet culture system and fracture callus could be interpreted to mean this model could be useful in studying the factors involved in the fracture repair process.1,25,31,36
Finally, gentamicin applied topically is effective in treating bone infection as it is delivered at high concentrations that avoid systemic toxicity. However, before using this approach, the high local levels achieved must be determined and shown to have no adverse effect on the surrounding tissue. Based on our results, gentamicin at 100 μg/mL concentration and greater has a substantial inhibitory effect on hMSCs in vitro and in vivo. If reparative hMSCs appear at sites of local administration of gentamicin, hMSC contribution to bone healing might be compromised.
We thank Dr. In- Hwan Song for providing expertise in histo- logic data analysis.
1. Ashhurst DE. Collagens synthesized by healing fractures. Clin Orthop Relat Res
2. Carrino DA, Arias JL, Caplan AI. A spectrophotometric modification of a sensitive densitometric Safranin O assay for glycosaminoglycans. Biochem Int
3. Dennis JE, Haynesworth SE, Young RG, Caplan AI. Osteogenesis in marrow-derived mesenchymal cell porous ceramic composites transplanted subcutaneously: effect of fibronectin and laminin on cell retention and rate of osteogenic expression. Cell Transplant
4. D'Ippolito G, Schiller PC, Ricordi C, Roos BA, Howard GA. Age- related osteogenic potential of mesenchymal stromal stem cells from human vertebral bone marrow. J Bone Miner Res
. 1999;14: 1115-1122.
5. Edin ML, Miclau T, Lester GE, Lindsey RW, Dahners LE. Effect of cefazolin and vancomycin on osteoblasts in vitro. Clin Orthop Relat Res
6. Gerhart TN, Roux RD, Horowitz G, Miller RL, Hanff P, Hayes WC. Antibiotic release from an experimental biodegradable bone cement. J Orthop Res
7. Haynesworth SE, Goshima J, Goldberg VM, Caplan AI. Characterization of cells with osteogenic potential from human marrow. Bone
8. Hedstrom SA, Lidgren L, Torholm C, Onnerfalt R. Antibiotic containing bone cement beads in the treatment of deep muscle and skeletal infections. Acta Orthop Scand
9. Isefuku S, Joyner CJ, Simpson AH. Toxic effect of rifampicin on human osteoblast-like cells. J Orthop Res
10. Isefuku S, Joyner CJ, Simpson AH. Gentamicin may have an adverse effect on osteogenesis. J Orthop Trauma
11. Jaiswal N, Haynesworth SE, Caplan AI, Bruder SP. Osteogenic differentiation of purified, culture-expanded human mesenchymal stem cells in vitro. J Cell Biochem
12. Justesen J, Stenderup K, Eriksen EF, Kassem M. Maintenance of osteoblastic and adipocytic differentiation potential with age and osteoporosis in human marrow stromal cell cultures. Calcif Tissue Int
13. Kahlmeter G, Dahlager JI. Aminoglycoside toxicity: a review of clinical studies published between 1975 and 1982. J Antimicrob Chemother
14. Klemm KW. Antibiotic bead chains. Clin Orthop Relat Res
. 1993; 295:63-76.
15. Lazzarini L, Mader JT, Calhoun JH. Osteomyelitis in long bones. J Bone Joint Surg Am
16. Lennon DP, Edmison JM, Caplan AI. Cultivation of rat marrow- derived mesenchymal stem cells in reduced oxygen tension: effects on in vitro and in vivo osteochondrogenesis. J Cell Physiol
. 2001; 187:345-355.
17. Lennon DP, Haynesworth SE, Bruder SP, Jaiswal N, Caplan AI. Human and animal mesenchymal progenitor cells from bone marrow: identification of serum for optimal selection and proliferation. In Vitro Cell Dev Biol
18. McKibbin B. The biology of fracture healing in long bones. J Bone Joint Surg Br
19. Miclau T, Edin ML, Lester GE, Lindsey RW, Dahners LE. Bone toxicity of locally applied aminoglycosides. J Orthop Trauma
20. Miclau T, Edin ML, Lester GE, Lindsey RW, Dahners LE. Effect of ciprofloxacin on the proliferation of osteoblast-like MG-63 human osteosarcoma cells in vitro. J Orthop Res
21. Mueller SM, Glowacki J. Age-related decline in the osteogenic potential of human bone marrow cells cultured in three-dimensional collagen sponges. J Cell Biochem
22. Murray WR. Use of antibiotic-containing bone cement. Clin Orthop Relat Res
23. Ponticiello MS, Schinagl RM, Kadiyala S, Barry FP. Gelatin-based resorbable sponge as a carrier matrix for human mesenchymal stem cells in cartilage regeneration therapy. J Biomed Mater Res
. 2000; 52:246-255.
24. Reddi AH. Bone and cartilage differentiation. Curr Opin Genet Dev
25. Reddi AH. Cell biology and biochemistry of endochondral bone development. Coll Relat Res
26. Rockwood CA Jr, Green DP, Bucholz RW, Heckman JD. Rockwood and Green's Fractures in Adults
ed. Philadelphia, PA: Lippincott-Raven; 1996.
27. Salvati EA, Callaghan JJ, Brause BD, Klein RF, Small RD. Reim- plantation in infection: elution of gentamicin from cement and beads. Clin Orthop Relat Res
28. Smilack JD, Flittie WH, Williams TW Jr. Bone concentrations of antimicrobial agents after parenteral administration. Antimicrob Agents Chemother
29. Solchaga LA, Penick K, Porter JD, Goldberg VM, Caplan AI, Welter JF. FGF-2 enhances the mitotic and chondrogenic potentials of human adult bone marrow-derived mesenchymal stem cells. J Cell Physiol
30. Stenderup K, Justesen J, Eriksen EF, Rattan SI, Kassem M. Number and proliferative capacity of osteogenic stem cells are maintained during aging and in patients with osteoporosis. J Bone Miner Res
31. Topping RE, Bolander ME, Galian G. Type X collagen in fracture callus and the effects of experimental diabetes. Clin Orthop Relat Res
32. Vecsei V, Barquet A. Treatment of chronic osteomyelitis by necrectomy and gentamicin-PMMA beads. Clin Orthop Relat Res
33. Wahlig H, Dingeldein E. Antibiotics and bone cements: experimental and clinical long-term observations. Acta Orthop Scand
. 1980; 51:49-56.
34. Wahlig H, Dingeldein E, Bergmann R, Reuss K. The release of gentamicin from polymethylmethacrylate beads: an experimental and pharmacokinetic study. J Bone Joint Surg Br
. 1978;60: 270-275.
35. Williams RJ III, Attia E, Wickiewicz TL, Hannafin JA. The effect of ciprofloxacin on tendon, paratendon, and capsular fibroblast metabolism. Am J Sports Med
36. Yao KL, Todescan R Jr, Sodek J. Temporal changes in matrix protein synthesis and mRNA expression during mineralized tissue formation by adult rat bone marrow cells in culture. J Bone Miner Res
37. Yoo JU, Barthel TS, Nishimura K, Solchaga L, Caplan AI, Goldberg VM, Johnstone B. The chondrogenic potential of human bone- marrow-derived mesenchymal progenitor cells. J Bone Joint Surg Am
38. Yoo JU, Johnstone B. The role of osteochondral progenitor cells in fracture repair. Clin Orthop Relat Res
39. Zalavras CG, Patzakis MJ, Holtom P. Local antibiotic therapy in the treatment of open fractures and osteomyelitis. Clin Orthop Relat Res