Particulate Endocytosis Mediates Biological Responses of Human Mesenchymal Stem Cells to Titanium Wear Debris Chukwuka C. Okafor,1* Hana Haleem-Smith,1 Patrice Laqueriere,2 Paul A. Manner,1,3 Rocky S. Tuan1 1
Cartilage Biology and Orthopaedics Branch, National Institutes of Arthritis, and Musculoskeletal and Skin Diseases, National Institutes of Health, Building 50, Room 1503, MSC8022, Department of Health and Human Services, Bethesda, Martland 20892-8022 2
Division of Bioengineering and Physical Sciences, Office of Research Services, National Institutes of Health, Department of Health and Human Services, Bethesda, Maryland 20892-5755 3
Department of Orthopaedic Surgery, George Washington University, Washington, DC 20037
Received 8 May 2005; accepted 15 August 2005 Published online 31 January 2006 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jor.20075
ABSTRACT: Continual loading and articulation cycles undergone by metallic (e.g., titanium) alloy arthroplasty prostheses lead to liberation of a large number of metallic debris particulates, which have long been implicated as a primary cause of periprosthetic osteolysis and postarthroplasty aseptic implant loosening. Long-term stability of total joint replacement prostheses relies on proper integration between implant biomaterial and osseous tissue, and factors that interfere with this integration are likely to cause osteolysis. Because multipotent mesenchymal stem cells (MSCs) located adjacent to the implant have an osteoprogenitor function and are critical contributors to osseous tissue integrity, when their functions or activities are compromised, osteolysis will most likely occur. To date, it is not certain or sufficiently confirmed whether MSCs endocytose titanium particles, and if so, whether particulate endocytosis has any effect on cellular responses to wear debris. This study seeks to clarify the phenomenon of titanium endocytosis by human MSCs (hMSCs), and investigates the influence of endocytosis on their activities. hMSCs incubated with commercially pure titanium particles exhibited internalized particles, as observed by scanning electron microscopy and confocal laser scanning microscopy, with time-dependent reduction in the number of extracellular particles. Particulate endocytosis was associated with reduced rates of cellular proliferation and cell–substrate adhesion, suppressed osteogenic differentiation, and increased rate of apoptosis. These cellular effects of exposure to titanium particles were reduced when endocytosis was inhibited by treatment with cytochalasin D, and no significant effect was seen when hMSCs were treated only with conditioned medium obtained from particulate-treated cells. These findings strongly suggest that the biological responses of hMSCs to wear debris are triggered primarily by the direct endocytosis of titanium particulates, and not mediated by secreted soluble factors. In this manner, therapeutical approaches that suppress particle endocytosis could reduce the bioreactivity of hMSCs to particulates, and enhance long-term orthopedic implant prognosis by minimizing wear-debris periprosthethic osteolysis. ß 2006 Orthopaedic Research Society.* Published by Wiley Periodicals, Inc. J Orthop Res 24:461–473, 2006
Keywords: mesenchymal stem cells; wear debris; endocytosis; titanium; osteolysis; aseptic loosening; orthopedic prosthesis
INTRODUCTION Correspondence to: Rocky S. Tuan (Telephone: 301-4516854; Fax: 301-435-8017; E-mail:
[email protected]) *Department of Orthopaedic Surgery, Thomas Jefferson University, Philadelphia, PA 19107. ß 2006 Orthopaedic Research Society. *This article is a US Government work and, as such, is in the public domain in the United States of America.
In 2002, about 170,000 Americans and over 1.5 million worldwide underwent total hip arthroplasty,1 a procedure that involves replacement of the acetabulum with an artificial liner, as well as replacement of the proximal femur with a prosthetic stem. The same year, over 40,000 JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
461
462
OKAFOR ET AL.
Americans underwent surgery to revise loosened orthopedic implants. Periprosthetic osetolysis, which leads to pain, loss of function, and surgery revision, is the major complication following total joint arthroplasty. Although total joint arthroplasty is undoubtedly one of the most successful surgical procedures available, it is clear that its longevity is limited by osteolysis, a problem that can be expected to be more pronounced and to increase dramatically in the next few years as America’s baby boomers age. Osteolysis usually progresses to implant failure that requires implant revision, a procedure that not only involves a repeat surgery, but also has a poorer clinical prognosis and a shorter implant survival and stability duration than primary joint arthroplasty.2,3 The risk of implant failure has had a negative impact on younger patients who are excellent candidates for arthroplasty, who must endure prolonged morbidity and joint pain simply because if they undergo the procedure when needed, they might need two or three revisions in their lifetime. Periprosthetic osteolysis is thought to be a tissue response to liberated wear debris generated from continual loading and articulation cycles at the implant interface.4 It is now also apparent that loosening of implants is highly correlated with wear debris accumulation.5–7 Today, as efforts are being made to prolong the life of implants, it is imperative that the mechanisms by which wear debris triggers osetolysis be elucidated. A number of retrieval studies have shown that tissues adjacent to a failed total joint implant contain billions of submicron size particles per gram (dry weight), including titanium particulates.8,9 Another study on aseptic loosened hip implants documented submicron-size titanium particulates in bone marrow not only at the prosthesis site, but also at remote sites such as the iliac crest bone marrow.10 Within the bone marrow stroma are multipotential mesenchymal stem cells (MSCs) that have the ability to differentiate into cells of a variety of mesenchymal lineages upon appropriate induction.11 The potential of these MSCs to differentiate into osteoblastic cells is critical for proper and continuous osseointegration of prosthetic implants. We have recently shown that exposure of human MSCs (hMSCs) to titanium particulates disrupts their osteogenic differentiation12 and enhances apoptosis in vitro.13 These effects on the hMSCs may thus reduce the population of functional osteoblasts at periprosthetic sites, thereby osseointegration at the bone/implant interface.14 JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
In this study, we tested the hypothesis that endocytosis of titanium particles by hMSCs mediates the biological effects observed after exposure to titanium particulates. Our results clearly showed that hMSCs endocytose titanium particulates, and that the biological effects (decreased proliferation, increased apoptosis, disruption of osteogenesis, and decreased substrate adhesion) associated with particulate exposure occur only after particulate endocytosis.
MATERIALS AND METHODS Reagents Unless stated otherwise, all reagents and tissue culture materials were purchased from Sigma (St. Louis, MO). Penicillin–streptomycin, fungizone, and fetal bovine serum (FBS) were obtained from GIBCO-BRL (Gaithersburg, MD), and DMEM/F12 mediuim with L-glutamine and 15 mM HEPES from BioWhittaker (Walkersville, MD). Titanium Particle Preparation and Characterization Commercially pure titanium (cpTi) particles from Sigma-Aldrich were treated to remove >99.94% of adherent endotoxins according to Ragab et al.15 Particles were first passivated with a 25% nitric acid wash at 708C for 1 h followed by three washes with sterile phosphatebuffered saline (PBS), and further sterilized by incubation in 70% ethanol for 30 min at room temperature. Particles were then incubated in five alternating cycles of 0.1 N NaOH/95% ethanol and 25% nitric acid; in each cycle, incubation in the alkali–ethanol mixture was done at 308C for 20 h, while the incubation in 25% nitric acid was done at room temperature for 20 h. Limulus amoebocyte lysate assay (Associates of Cape Cod, Inc., Falmouth, MA) of the treated particles showed the absence of endotoxin levels exceeding 0.005 EU/mL. Particles were washed three times with sterile PBS, resuspended in DMEM/F12 medium containing 10% FBS, 100 U/mL penicillin, 100 mg/mL streptomycin, and 0.25 mg/mL amphotericin B, and stored in stock suspension at 48C. For experiments, the cpTi stock suspension was diluted with culture medium to provide the needed concentration of particles per hMSC. Particle size distribution and particle concentrations were determined using a Multisizer 3 Coulter Counter (Beckman Coulter, Miami, FL) as described previously,12 and particles were observed by means of scanning electron microscopy (SEM) (Hitachi S-4500 Scanning Electron Microscope, Hitachinaka, Japan). Isolation and Culture of hMSCs hMSCs were isolated from bone marrow aspirated from the femoral heads of patients undergoing total hip DOI 10.1002/jor
PARTICULATE ENDOCYTOSIS BY HUMAN MESENCHYMAL STEM CELLS
arthroplasty for primary osteoarthritis.12 Patients with congenital malformation, neoplasia, avascular necrosis, or viral infection were excluded. This study was approved by the Institutional Review Boards of George Washington University (Washington, DC) and Thomas Jefferson University (Philadelphia, PA), and was carried out in compliance to all requirements for patient consent. Approximately 4 mL volume of bone marrow was gently vortexed and suspended in 10 mL of DMEM/F12 medium, drawn through a 10-cc syringe with a 20G11/2 needle (Becton-Dickinson, Franklin Lakes, NJ), and placed in a 50-mL centrifuge tube. This process was repeated until the marrow was sufficiently separated from bone tissue and debris; the suspension was then subjected to centrifugation at 1000 rpm for 5 min, the supernatant was removed, and the cells were resuspended in 20 mL of DMEM/F12 medium and counted using a hemocytometer. After resuspension to a concentration of 3 106 cells/mL in DMEM/F12 medium supplemented with 10% FBS, 100 U/mL penicillin, and 100 mg/mL streptomycin, and 0.25 mg/mL amphotericin B, cells were seeded in T150 tissue culture flasks, then incubated in 5% CO2 at 378C and allowed to adhere for 24 h. Serum lot was selected on the basis of hMSC proliferation in an undifferentiated state and support of osteogenic differentiation determined by alkaline phosphatase staining.7 Nonadherent cells were aspirated 24 h after seeding, and adherent cells were supplied fresh medium every 72 h and expanded until 75% confluency (expansion time of 2–3 weeks) prior to use. In this study, bone marrow was aspirated from three patients.
Treatment of hMSCs with Particulates Culture expanded hMSCs were seeded at 2.5 105 cells/ mL per well in six-well plates and allowed to adhere for 24 h at 378C. Initially, appropriate particle concentration was selected based on the effects of varying concentrations of titanium particles (0 to 10,000 particles per cell) on apoptosis induction detected from flow cytometry-based propidium iodide cell cycle analysis (see below); the results revealed that particle concentration as low as 1000 per cell had significant
463
effect on hMSC apoptosis. Based on this finding (see Results), cells were incubated with DMEM/F12/10% FBS containing 1000 cpT; particles/cell, unless stated otherwise. The treatment groups are shown in Table 1, with control cells incubated in medium alone. The treatments (24 h) are: (1) Ti, cultures treated with 1000 titanium particles/cell; (2) CM, cultures treated with 25% conditioned medium derived from Group 1; (3) CD, cultures pretreated with 2 mM cytochalasin for 30 min prior to the 24-h incubation period; (4) CD þ Ti, cytochalasin D-treated cells as in Group 3 exposed to 1000 titanium particles/cell; and (5) Latex, cells treated with 1000 particles/cell of 0.5-mm latex beads (Polysciences, Warrington, PA).
Particle Endocytosis Assay Cultured hMSCs were seeded at 5 104 cells per well in 24-well plates and allowed to adhere for 24 h at 378C. After cells adhered, 1 107 titanium particles (2000 particles per cell) were added to each well in duplicates. At 0-, 2-, 6-, 8-, 12-, and 24-h time points, supernatant of respective wells were collected, then cells were trypsinized with 0.05% trypsin for 5 min to remove and collect all titanium particle attached to the cell membranes but not internalized. Controls consisted of 1 107 particles added to control well without cells and processed similarly. The same assay was also carried out on cytochalasin D-treated hMSCs. Particles were pelleted from the suspension solution collected per time point by centrifugation at 1000 rpm for 5 min, resuspended in 1 mL of DMEM/F12 medium, and counted by means of a Multisizer 3 Coulter Counter. Particle Internalization Examined by SEM hMSCs were seeded at 3 104 cells per 13 mm diameter Thermanox plastic coverslip (EM Sciences, Fort Washington, PA), placed in wells of 24-well plates, and allowed to adhere for 24 h at 378C, and were then subjected to particle treatment for 24 h. Control cells were not treated with any particles. Other controls included: (1) cells first prefixed with buffered 2.5%
Table 1. Treatment Groups to Analyze the Effect of Titanium Particles on Mesenchymal Stem Cells Groups Control Ti CM CD Ti þ CD Latex
DOI 10.1002/jor
Treatment Cells maintained in regular DMEM-F12 medium, supplemented with 10% FBS Cells maintained as in the Control Group, with exposure to titanium particulates (1000/cell unless stated otherwise) for 24 h Maintained as in the Control Group, with supplementation of 25% conditioned medium derived from cells in the Ti Group Maintained as in the Control Group, in the presence of 2 mM cytochalasin D for 30 min Treated as Ti and CD combined (CD first, medium removed, followed by Ti at 1000 particles/cell, unless stated otherwise) Treated as in Ti, except with latex particles (0.5 mm; 1000 particles/cell)
JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
464
OKAFOR ET AL.
glutaraldehyde, (2) cells pretreated with 2 mM cytochalasin D, and (3) cells incubated at 48C before and during 24-h treatment with titanium particles. After 24-h treatment, all cultures were washed twice with PBS and fixed in buffered 2.5% glutaraldehyde, dehydrated through a graded series of ethanol, and critical point dried with hexamethyldisilazane (EM Sciences, Fort Washington, PA). The coverslips containing fixed and dehydrated cells were mounted onto aluminum stubs, and sputter-coated (MED 010 minideposition system, Balzers Union, Liechtenstein) with 15 nm of gold and examined using a Hitachi S-4500 SEM, equipped with a cold cathode field emission gun at an accelerating voltage of 30 kV. Images were acquired using the secondary electron and backscattered electron detectors and recorded digitally. Particle Internalization Assay by Confocal Laser Scanning Microscopy hMSCs from passage 2 to 4, seeded at 3 104 cells per 13-mm diameter glass coverslip (EM Sciences) placed in wells of 24-well plates, were treated for 24 h with varying concentrations of titanium particles or latex beads, or with no particles. Cells were then gently rinsed three times with PBS, fixed with 4% paraformaldehyde for 15 min at room temperature, and stained with 2 mg/mL acridine orange (Molecular Probes, Eugene, OR) in PBS for 20 min at room temperature. Viewed using a scanning confocal microscope (Olympus Fluo View FV500, Melville, NY), endocytosed particles were visualized in optical sections obtained parallel to the substrate at 0.5 mm intervals along the z-axis. Internalized particles appeared as black holes within the cytoplasm of the hMSCs.10 About 50 cells were examined per coverslip. Cell Proliferation Assay
Table 1. After treatment, cells were washed with PBS, collected after dissociation with 0.25% trypsin, and pelleted by centrifugation. Cells were then labeled in the dark at 48C for 1 h with propidium iodide solution (20 mg/mL propidium iodide and 1 mg/mL RNase in PBS). Before flow cytometry, any remaining, free titanium particulates were carefully removed by washing. Control consisted of cells first fixed with ice-cold methanol, then exposed to particles for 24 h, and stained with propidium iodide as described above. All samples were then analyzed using fluorescenceactivated cell sorting (FACS) with a DIVA Flow Cytometer (Becton Dickinson, Heidelberg, Germany). Each treatment set was done in duplicates. Apoptosis was assayed similarly for hMSC cultures seeded at 2.5 105 cells/well in six-well plates and treated for 72 h at 378C with 10 ng/mL IL-1a, 10 ng/mL IL-6, 20 ng/mL TNF-a, 20 ng/mL TRAIL, or a mixture of all four cytokines at the respective concentrations. Osteogenesis Assay Cells treated with particles as described in Table 1 were cultured with fresh DMEM/F12/10% FBS containing osteogenic supplements (OS) (50 mg/mL L-ascorbate2-phosphate, 0.1 mM dexamethasone, 10 mM bglycerolphosphate, and 10 nM 1a,25-(OH)2D3).11,16 The culture medium was changed every 3 days, and the cultures were allowed to differentiate for 10 days, and then evaluated for osteogenic differentiation on the basis of alkaline phosphatase activity. As a negative control, nonparticle loaded cells were cultured with DMEM/F12/10% FBS without osteogenic supplements. For alkaline phosphatase histochemistry, cultures were fixed with 4% paraformaldehyde in PBS, stained with Sigma ALP Kit according to the manufacturer’s protocol, and examined with bright field light microscopy. The percentage of alkaline phosphatase positive cells per well was quantified by counting cells in three randomly selected microscopic fields under a 20 objective.
Cultured hMSCs from passage 2 to 4 were seeded at 2.5 105 cells/mL per well in six-well plates and treated as described in Table 1. After treatment, cells were trypsinized, replated in 96-well plates at 10,000 cells per well, incubated at 378C for the indicated periods of time, and analyzed for cell proliferation using the Dojindo Cell Counting Kit-8 (CCK-8, Dojindo Molecular Technologies, Inc., Gaithersburg, MD). Calibration was done using cultures containing known numbers of viable cells to generate the standard curve. After 24, 72, and 120 h of incubation, 10 mL of CCK-8 solution was added to each well of a 96-well plate, the plates were then incubated at 378C for 2 h, and A450 was measured using a VICTOR17 V microplate reader (Perkin-Elmer Life Sciences, Boston, MA).
For adhesion studies, cells were treated as described above (see Table 1) for 12 h, collected after dissociation using 0.05% trypsin, and then plated at 5 104 cells per well in 24-well plates. At 15, 30, 45, 60, and 120 min after plating, hMSCs adhesion kinetics was measured fluorimetrically as described previously by using the fluorescent vital dye, 20 70 -bis (2-carboxyethyl)5carboxyfluorescein acetyoxymethylester (BCECF-AM) (Molecular Probes, Junction City, OR).18 Cell numbers were determined by comparison to a standard curve of dye intensities released from known cell numbers.
Apoptosis Assay Using Flow Cytometry
Statistical Analysis
Cultured hMSCs seeded at 2.5 105 cells per well in six-well plates were treated for 24 h as described in
All values are expressed as the mean of multiple separate experiments SEM. Data were analyzed,
JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
Cell Adhesion Assay
DOI 10.1002/jor
PARTICULATE ENDOCYTOSIS BY HUMAN MESENCHYMAL STEM CELLS
465
and statistical significance was determined by analysis of variance (Smith’s Statistical Package, Claremont, CA), and where appropriate, post hoc testing was performed using the Student’s t-test. p 0.05 was required for statistical significance.
RESULTS Endocytosis of Titanium Particulates by hMSCs The size distribution of the titanium particles used in this was determined using a Coulter Counter, and the results are presented in Table 2. SEM based on secondary electrons and backscatter electrons revealed that the particles existed also as both individual entities as well as larger aggregates (Fig. 1). Following 24-h treatment with titanium particles, SEM revealed particles internalized within hMSCs (Fig. 2B and C). When endocytosis was inhibited by treatment with cytochalasin D, culturing at 48C, or cell fixation prior to particulate treatment, no particles were observed within cells (Fig. 2D–F). To confirm that hMSCs actively endocytosed titanium particles in their vicinity, hMSCs treated with titanium particles were observed as a function of time, and progressively lower abundance of extracellular particles was apparent (Fig. 3). Cells that were first treated with cytochalasin D prior to incubation with particulates did not exhibit particle endocytosis until 12 h later. Final confirmation of particulate endocytosis by hMSCs was carried out with confocal laser scanning microscopy, illustrated by the z-scan image stacks in Figure 4. The images of sections parallel to the substrate showed multiple, punctate dark particles (Fig. 4A and B), and clusters of dark particles were fully surrounded by the acridine orange-stained cytoplasm of the hMSCs (Figs. 4A, iii–ix), not visible above and below the cell body (Figs. 4A, i and ii, and xi and xii). hMSCs that were not exposed to titanium particles did not exhibit similar, internalized dark objects (results
Table 2. Size Distribution of Titanium Particulatesa Parameter Mean (SD) Median Mode
Size (mm) 0.519 0.125 0.488b 0.426
a Determined as diameter using Multisizer 3 Coulter Counter (see Materials and Methods). b 78.89% less than 0.772 mm; 99.3% less than 2.650 mm.
DOI 10.1002/jor
Figure 1. Scanning electron microscopy image of cpTi particles. (A) Secondary electron detector image, (B) backscattered electron image. Titanium particles are seen forming aggregates of various sizes. Bar ¼ 30 mm.
not shown). hMSCs were capable of internalizing a large number of titanium particles, which appeared to be localized principally in extranuclear space of the cell, that is, the cytoplasm (Fig. 4B). Taken together, these observations strongly suggest that hMSCs carried out active endocytosis of titanium particulates, subsequently localizing them to the cytoplasm. Effect of Titanium Particulate Endocytosis on Cell Proliferation of hMSCs As shown in Figure 5, exposure of hMSCs to 1000 titanium particles per cell dramatically decreased the rate of cellular proliferation, compared to hMSCs not treated with any particles or hMSCs treated with 0.5-mm latex beads. Interestingly, hMSCs first treated with cytochalasin D prior to incubation with titanium particulates were less severely affected, suggesting that blocking endocytosis of particulates overcame their effect on cell proliferation, even in the presence of cytochalasinmediated perturbation of the cytoskeleton.
Effect of Titanium Particulate Endocytosis on Apoptosis of hMSCs Exposure of hMSCs to increasing concentrations of titanium particles resulted in a dose-dependent increase in apoptotic cells. Apoptosis induction was observed even at particle concentrations as low as 500 particles per cell, whereas a concentration of 5000 particles per cell resulted in almost complete apoptosis. As shown in Figure 6, FACS analysis showed that cells that had endocytosed titanium particles were more apoptotic than cells that had not endocytosed particulates (treated with cytochalasin D). Interestingly, exposure to 25% conditioned medium derived from particulate-treated cells did not significantly increase the proportion of apoptotic hMSCs. JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
466
OKAFOR ET AL.
Figure 2. Scanning electron microscopy (SEM) images of hMSCs exposed to titanium particles under various treatment conditions. hMSCs were treated with 1000 titanium particles/cell under conditions described in Table 1. Cultures were viewed by SEM using secondary electron (SE) detectors for cell surface topography (A1, B1, C1, D1, E1, F1, and G1), as well as backscattered electron (BSE) detectors to image difference in chemical compositions in the specimen at incidence angles deeper than the cell surface, to highlight endocytosed particles (A2, B2 and 3, C2 and 3, D2, E2, F2, and G2; B2 and 3, and C2 and 3 are identical positive/negative images). (A) Control hMSC culture, not exposed to titanium particulates—no particles seen under BSE. (B, C) Ti Group—clearly showing the presence of endocytosed particles under BSE. (B) Low magnification; (C) high magnification. (D, E) Ti þ CD Group—Ti particles seen adherent to cell surface only, and not internalized. (F) Similar to Ti Group, except cells were prefixed with paraformaldehye prior to exposure to titanium particles—particles seen distributed on cell surface and not internalized. (G) hMSCs þ Ti incubated at 48C—the low temperature suppressed endocytosis, as Ti particles were seen only outside cells. Bar ¼ 4 mm (C), 10 mm (A, E), or 20 mm (B, D, F, and G).
Effect of Titanium Particulate Endocytosis on Osteogenic Differentiation of hMSCs hMSCs subjected to different treatments as described in Table 1 were treated with osteogenic supplements for 10 days to induce osteogenesis. As shown in Figures 7 and 8, hMSCs treated with titanium particles showed a significant decrease in the percentage of alkaline phosphatasepositive, osteoblastic cells, when compared to untreated cells. Particulate endocytosis was again an important factor, because the suppression of JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
osteogenesis was reduced with cytochalasin D pretreatment of the hMSCs. In addition, cells treated with 25% conditioned medium, or cells exposed to 0.5-mm latex beads were also significantly less affected than cells that were treated with and endocytosed titanium particles. Effect of Titanium Particulate Endocytosis on hMSC Substrate Adhesion Efficient substrate adhesion is one of the hallmarks of MSCs, and has been widely utilized for DOI 10.1002/jor
PARTICULATE ENDOCYTOSIS BY HUMAN MESENCHYMAL STEM CELLS
Figure 3. Rate of hMSCs endocytosis of titanium particulates. hMSCs were treated with 1000 titanium particles/cell as described in Table 1. hMSCs rapidly endocytosed surrounding titanium particulates. Treatment with cytochalasin D, a reversible endocytosis inhibitor prevents endocytosis for about 8–12 h after treatment.
their isolation.11,19 We therefore examined whether exposure to and endocytosis of titanium particulates would affect hMSC substrate adhesion. As shown in Figure 9, analysis of adhesion kinetics revealed that hMSCs treated with 1000 titanium particles per cell were significantly less adherent, when compared to hMSCs not treated with any particles. hMSCs pretreated with cytochalasin D before exposure to titanium particulates showed improved adhesion kinetics, similar to hMSCs treated with cytochalasin D only. Treatment of hMSCs with 25% conditioned medium derived from particulate-treated hMSCs only resulted in mild suppression of cell adhesion. Exposure to latex beads did not affect hMSC adhesion (results not shown).
DISCUSSION We have previously shown that hMSCs exposed to titanium particulates wear debris exhibit a host of biological responses that include reduced cell proliferation, decreased cell viability, increased apoptosis, and suppressed osteogenic differentiation.12–14 The aim of this study is therefore to examine the mechanisms by which exposure to titanium particulates influences the biology of hMSCs to result in these effects. Our hypothesis is that endocytosis of titanium particulates precedes and is responsible for these biological effects. DOI 10.1002/jor
467
Our results reported here for the first time that hMSCs actively endocytose submicron-size titanium particulates. We have also corroborated the biological responses of hMSCs upon treatment with titanium particulates to particle endocytosis. SEM and confocal microscopy showed that the titanium particulates are internalized by hMSCs from the immediate vicinity of the cells, that is, no accumulation of particles is seen adjacent to the cells. The particles appear to be lodged within the cytoplasm, and from the confocal images, most likely exist as either aggregates or in close proximity to one another to form the observable, dark structures. In comparison, SEM observations showed that titanium particulates are relatively abundant in the vicinity (excluding particles that appear adherent to the cell surface) of hMSCs that have been prefixed, or incubated at 48C, or pretreated with cytochalasin D, an inhibitor of vesicular endocytosis. In addition to confirming our previous findings that exposure to titanium particulates reduces cell viability and enhances apoptosis,13 as well as suppresses osteogenic differentiation,12 we observe that cell–substrate adhesion is also significantly inhibited. All of these effects are specific to titanium particles, because incubation with 0.5-mm latex beads had no effect. These effects are also dependent on active particle endocytosis, because the effects are reduced by suppressing endocytotic activity with cytochalasin D or incubation at 48C. That conditioned medium alone is ineffective again supports the importance of direct cellular interaction with and endocytosis of the titanium particles in mediating their effects. It should be noted that cytochalasin D is a reversible inhibitor of endocytosis,20 and to avoid prolonged effects on the cytoskeleton, only 30- to 60-min incubation with cytochalasin D is done in this study. As shown in Figure 3, cytochalasin D-treated cells are able to recover their ability to endocytose titanium particulates 6 to 8 h past the initial treatment; thus, the biological effects of exposure to titanium particulates are only reduced, but not eliminated, by cytochalasin D pretreatment. Presumably, if endocytosis could be continuously inhibited (perhaps by other means), the effects of particulate exposure could be more or completely blunted. It is noteworthy that conditioned medium from particulate-treated MSCs has little effect on the parameters measured here, supporting the notion that the effects of the particles require their direct interaction with the cells, specifically via JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
468
OKAFOR ET AL.
Figure 4. Internalization status of titanium particulates in hMSCs visualized by confocal microscopy. hMSCs were treated with titanium particles (1000/cell), stained with 2 mg/mL acridine orange in PBS, and observed by confocal laser scanning microscopy. Images were obtained as optical sections parallel to the substrate at 0.5 mm intervals, and about 50 cells on each coverslip were examined. Internalized titanium particulates appeared in the sequential confocal sections as dark holes within the acridine orange-stained cytoplasm of the hMSCs. (A) z-Scan confocal sections parallel to the substrate traversing the entire cell body of a single hMSC, showing multiple dark holes and clusters of dark holes fully surrounded by the cytoplasm. Bar ¼ 20 mm. (B) A representative hMSC showing a large number of internalized titanium particulates, with a nucleus-sparing pattern (also see SEM images in Fig. 2). Bar ¼ 20 mm.
endocytosis. In preliminary studies, we have treated hMSCs with one or more combinations of the following inflammatory cytokines, IL-1b, IL-6, TRAIL, and TNF-a, at concentrations up to 20 mg/ JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
mL for 72 h, and found that direct treatment with these cytokines, similar to the conditioned medium, elicited negligible apoptosis (data not shown). This observation suggests that hMSCs are DOI 10.1002/jor
PARTICULATE ENDOCYTOSIS BY HUMAN MESENCHYMAL STEM CELLS
Figure 5. Effect of titanium particulate endocytosis on hMSC proliferation. hMSCs were treated with 1000 titanium particles/cell as described in Table 1. The Tionly group showed significantly reduced cell proliferation compared to untreated control or the latex group. When titanium particle treatment was done in the presence of cytochalasin D (CD þ Ti) to inhibit endocytosis, cell proliferation was only slightly affected.
naturally insensitive to these cytokines that are known to be liberated by macrophages upon phagocytosis of wear debris particles,21–23 at least as far as apoptosis is concerned, and thus release of
Figure 6. Effect of titanium endocytosis on hMSC apoptosis. hMSCs subjected to treatment with titanium particles (1000 and 1000 particles/cell) under conditions described in Table 1 showed significantly increased rate of apoptosis, compared to untreated control and latex treated cultures. Pretreatment with cytochalasin D (30 and 60 min) at 1000 titanium particles/cell significantly reduced the level of apoptosis. Interestingly, conditioned medium (CM; 25%) from titanium particle-treated cells did not affect apoptosis. *p 0.05. DOI 10.1002/jor
469
these cytokines is unlikely to the key factor responsible for the effects of titanium particulates as observed here. On the other hand, whether hMSCs are capable of responding to titanium particles by secreting stress responsive chemokines, as in the case of osteoblasts,24 remains to be investigated. Both mechanical and biological theories have been put forward to explain aseptic loosening of orthopedic implants. The mechanical theory attributes factors, such as repeated daily peak and localized stresses imposed on the prosthesis– osseous interface during normal activities, to the generation of necrotic bone tissues and their replacement by soft tissues.4,17,25–27 The biological theory of orthopedic implant loosening states that submicroscopic wear debris particles, produced from the gliding surfaces of the prosthetic components and found in periprosthetic sites, are phagocytosed by tissue macrophages, and provoke inflammation, leading to the release of cytokines, such as IL-1, TNF-a, and prostaglandins, that activate osteoclasts and result in osteolysis and aseptic implant loosening.21,28 –33 Our results suggest an additional pathway leading to aseptic osteolysis and implant loosening.14 As illustrated in Figure 10, wear debris particles are generated from continual articulation and gliding of implant surfaces, particularly over time. Although some of these particulates are phagocytosed by macrophages, the multipotential hMSCs that line the bone marrow cavity also actively endocytose these particles. As a consequence, the proliferation, and adhesive and osteogenic activities of the hMSCs are significantly diminished, while their apoptosis rate increases, in a manner dependent on the concentration of and the time of exposure to the wear debris particles. The hMSCs in the closest proximity to the highest concentration of liberated wear debris particles will be affected the most, and as the effects on their biological activities are exacerbated, these cells will detach from the prosthetic device, creating a gradually expanding periprosthetic cavity. In addition, that the osteoprogenitor cells are less adherent and less capable of osteogenic differentiation during remodeling of the periprosthetic bone will result in an imbalance between bone growth and bone resorption (already enhanced due to activation of macrophages via particle phagocytosis mediated release of proinflammatory cytokines). This imbalance will also contribute to the development of a gradually expanding periprosthetic cavity. The expanding cavity allows for easier migration of wear debris JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
470
OKAFOR ET AL.
Figure 7. Effect of titanium endocytosis on osteogenic differentiation of hMSCs. Cultures of hMSCs were treated as described in Table 1 and induced to undergo osteogenesis as described in Materials and Methods. Osteogenesis was assessed on the basis of alkaline phosphatase activity detected histochemically. Osteogenesis was lower in cultures treated with 1000 (E) or 2,000 (G) titanium particles per cell (internalized titanium particulates were readily visible), compared to untreated osteogenic cultures (B). Treatment with conditioned medium (C) or latex particles (I) did not affect osteogenesis. Pretreating cultures with cytochalasin D (F, H) appeared to protect the effect of the titanium particles in terms of osteogenesis; treatment with cytochalasin alone did not affect osteogenesis (D). Bar ¼ 20 mm. OS, osteogenic medium.
further down the prosthetic stem, where they will again be actively endocytosed by hMSCs. In this manner, the process repeats itself and continues until stability of the prosthesis is significantly compromised, resulting in implant loosening. Thus, we propose that both bone growth and bone resorption are affected by the release
Figure 8. Quantitation of the effect of titanium particle endocytosis on osteogenic differentiation of hMSCs. Similar to the histochemical results in Figure 7, the results presented here quantitatively show that hMSC differentiation was significantly suppressed upon exposure to titanium particles (1000 and 2000 per cell). Pretreatment with cytochalasin D protected the cells from the effect of the titanium particles. *p < 0.05, compared to the positive control (þOS). JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
Figure 9. Effect of titanium particle endocytosis on hMSC substrate adhesion. hMSCs were treated with titanium particles (1000/cell) as described in Table 1. Titanium particulate endocytosis by hMSCs substantially suppressed their adhesive ability to less than 50% of untreated control at 120 min (values at all time points were significantly different; *p 0.04). Pretreatment with cytochalasin D (Ti þ CD) partially protected the hMSCs from the titanium particle effect (adhesion at 75% of control at 120 min). On the other hand, treatment with conditioned medium (25%) from titanium particle treated hMSCs (CM) or cytochalasin D (CD) initially suppressed cell adhesion, although the level at 120 min was similar to that of untreated control. DOI 10.1002/jor
PARTICULATE ENDOCYTOSIS BY HUMAN MESENCHYMAL STEM CELLS
471
Figure 10. Schematic representation of wear debris mediated periprosthetic osteolysis involving MSCs. (Left) Particulates generated from prosthetic wear localized to periprosthetic space are presented to bone cells, monocytes/macrophages, and MSCs. (Right) The effects on MSCs include suppression of cell proliferation, substrate adhesion, and osteogenic differentiation, and increased apoptosis. In combination with other known effects of wear debris on monocytes/macrophages and osteoblasts (see text), the rate of bone resorption exceeds rate of bone formation, and the periprosthetic space is enlarged, eventually resulting in aseptic loosening of the prosthesis.
of wear debris in periprosthetic sites to contribute to clinically observed aseptic implant loosening. Although our data clearly demonstrate that hMSCs actively endocytose titanium particulates and that this activity precedes and is required for the particulate-mediated effects on proliferation, substrate adhesion, apoptosis, and osteogenic activities of the hMSCs, the exact mechanistic signals responsible are currently unknown. The combined, and perhaps synergistic, effects on bone growth and bone resorption contribute to the pathogenesis of aseptic osteolysis. Alternatively, the cellular responses observed here may have resulted from the endocytosed particles occupying space in and gradually filling up the cytoplasm, thereby interfering with regular cellular functions. Unlike macrophages, whose primary function is the clearing of foreign substances via phagocytosis to trigger inflammation, MSCs are unlikely to actively endocytosis under DOI 10.1002/jor
normal conditions. Thus, when MSCs respond by endocytosing titanium particles, the internalized particles are likely to pose significant physical interference within the cytoplasm, such as interrupting cytoskeletal organization, and impact cellular functions. Ongoing studies aim to determine whether hMSCs that have endocytosed titanium particles release pro-inflammatory cytokines as macrophages do,31 or whether they produce factors that stimulate osteoclastogenesis as osteoblasts do.24 In this manner, we will gain further insights into the possible paracrine consequences of the action of titanium particles on hMSCs; such information may provide the basis for the development of pharmacologic therapeutics for periprosthetic osteolysis.14 Finally, we are also interested in determining whether changes in integrin profile may be responsible for the reduced substrate adhesivitiy of hMSCs that have endocytosed titanium particles. JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
472
OKAFOR ET AL.
ACKNOWLEDGMENTS We thank Dr. Kristien Zaal, NIAMS, for her technical assistance in confocal microscopy, and Dr. Mark Wang for sharing insights from his previous work on the effects of particulates on mesenchymal stem cells. This work was supported by NIH ZO1 AR41113. C.C.O. was an Undergraduate Scholar of the National Institutes of Health.
13.
14.
15.
REFERENCES 16. 1. American Academy of Orthopaedic Surgeons Publications. Projections of total hip and total knee arthroplasty in the United States: projections to 2030. http://www.aaos.org/wordhtml/pdfs_r/tjr.pdf and http://www.aaos.org/wordhtml/press/arthropl. htm. 2. Callaghan JJ, Forest EE, Olejniczak JP. 1985. Results of revision for mechanical failure after cemented total hip replacement, 1979 to 1982. A two to five-year follow-up. J Bone Joint Surg Am 67:1074–1085. 3. Hanssen AD, Rand JA. 1988. A comparison of primary and revision total knee arthroplasty using the kinematics stabilizer prosthesis. J Bone Joint Surg Am 70:491–499. 4. Mjoberg B. 1997. The theory of early loosening of hip prostheses. Orthopedics 20:1169–1175. 5. Jacobs JJ, Shanbhag A, Glant TT, et al. 1994. Wear debris in total joint replacements. J Am Acad Orthop Surg 4:212–250. 6. Jazrawi LM, Jaffe WL. 1999. Ceramic bearings in total hip arthroplasty: early clinical results and review of long term experiences. Arthroplasty Arthroscopic Surg 10:1–7. 7. Wang JC, Yu WD, Sandhu HS, et al. 1999. Metal debris from titanium spinal implants. Spine 24: 899–903. 8. Hirakawa K, Bauer TW, Stulberg BN, et al. 1996. Characterization and comparison of wear debris from failed total hip implants of different types. J Bone Joint Surg Am 78:1235–1243. 9. Margevicius KJ, Bauer TW, McMahon JT, et al. 1994. Isolation and characterization of debris in membranes around total joint prostheses. J Bone Joint Surg Am 76:1664–1675. 10. Engh CA Jr, Moore KD, Winh TN, et al. 1997. Titanium prosthetic wear debris in remote bone marrow. A report of two cases. J Bone Joint Surg Am 79:1721–1725. 11. Pittenger MF, Mackay AM, Beck SC, et al. 1999. Multilineage potential of adult human mesenchymal stem cells. Science 284:143–147. 12. Wang ML, Nesti LJ, Tuli R, et al. 2002. Titanium particles suppress expression of osteoblastic phe-
JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
notype in human mesenchymal stem cells. J Orthop Res 20:1175–1184. Wang ML, Tuli R, Manner PA, et al. 2003. Direct and indirect induction of apoptosis in human mesenchymal stem cells in response to titanium particles. J Orthop Res 21:697–707. Wang ML, Sharkey PF, Tuan RS. 2004. Particle bioreactivity and wear-mediated osteolysis. J Arthroplasty 19:1028–1038. Ragab AA, Motter RVD, Lavish SA, et al. 1999. Measurement and removal of adherent endotoxin from titanium particles and implant surfaces. J Orthop Res 17:803–809. Caterson EJ, Nesti LJ, Danielson KG, et al. 2002. Human marrow-derived mesenchymal progenitor cells: isolation, culture expansion, and analysis of differentiation. Mol Biotechnol 20:245–256. Brown TD, Stewart KJ, Nieman JC, et al. 2002. Local head roughening as a factor contributing to variability of total hip wear: a finite element analysis. J Biomech Eng 124:691–698. Sinha RK, Tuan RS. 1994. Surface composition of orthopaedic implant metals regulates cell attachment, spreading and cytoskeletal organization of primary human osteoblasts in vitro. Clin Orthop 305:258–272. Bruder SP, Jaiswal N, Haynesworth SE. 1997. Growth kinetics, self-renewal, and the osteogenic potential of purified human mesenchymal stem cells during extensive subcultivation and following cryopreservation. J Cell Biochem 64:278–294. Schafer DA. 2002. Coupling actin dynamics and membrane dynamics during endocytosis. Curr Opin Cell Biol 14:76–81. Blaine TA, Rosier AN, Puzas JE, et al. 1996. Increased levels of tumor necrosis factor-alpha and interleukin-6 protein and messenger RNA in human peripheral blood monocytes due to titanium particles. J Bone Joint Surg Am 78:1181–1192. Fritz EA, Glant TT, Vermes C, et al. 2002. Titanium particles induce the immediate early stress responsive chemokines IL-8 and MCP-1 in osteoblasts. J Orthop Res 20:490–498. Maloney WJ, James RE, Smith RL. 1996. Human macrophage response to retrieved titanium alloy particles in vitro. Clin Orthop 322: 268–278. Granchi D, Amato I, Battistelli L, et al. 2005. Molecular basis of osteoclastogenesis induced by osteoblasts exposed to wear particles. Biomaterials 26:2371–2379. Callaghan JJ, Brown TD, Pedersen DR, et al. 2002. Choices and compromises in the use of small head sizes in total hip arthroplasty. Clin Orthop 405:144–149. Fisher J. 1998. Wear of polyethylene in artificial hip joints: superolateral wear of the acetabulum. J Bone Joint Surg Br 80:190–191.
DOI 10.1002/jor
PARTICULATE ENDOCYTOSIS BY HUMAN MESENCHYMAL STEM CELLS
27. Grottkau BE, Noordin S, Shortkroff S, et al. 2002. Effect of mechanical perturbation on the release of PGE(2) by macrophages in vitro. J Biomed Mater Res 59:288–293. 28. Friedman RJ, Black J, Galante JO, et al. 1994. Current concepts in orthopaedic biomaterials and implant fixation. Inst Course Lect 43:233–255. 29. Glant TT, Jacobs JJ. 1994. Response of three murine macrophage populations to particulate debris: bone resorption in organ cultures. J Orthop Res 12:720–731.
DOI 10.1002/jor
473
30. Goodman S. 2005. Wear particulate and osteolysis. Orthop Clin North Am 36:41–48. 31. Jacobs JJ, Roebuck KA, Archibeck M, et al. 2001. Osteolysis: basic science. Clin Orthop Rel Res 393:71–77. 32. Konttinen YT, Zhao D, Beklen A, et al. 2005. The microenvironment around total hip replacement prostheses. Clin Orthop 430:28–38. 33. Wooley PH, Schwarz EM. 2004. Aseptic loosening. Gene Ther 11:402–407.
JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006