Polymorphonuclear leukocyte degranulation with exposure to polymethylmethacrylate nanoparticles Frank J. Papatheofanis* and Riad Barmada Bone Metabolism Laboratory, Department of Orthopaedics, The University of Illinois, College of Medicine, Chicago, Illinois 60680 Polymethylmethacrylate (PMMA) is clinically employed in a wide range of orthopaedic procedures. The etiology of the inflammatory reaction of recipient tissues to PMMA remains unresolved. Classically, polymorphonuclear leukocytes (PMNs) release cytoplasmic lysosomal granules when exposed to a variety of proinf lammatory stimuli. Such degranulation contributes, and partially defines, the local tissue reaction to this foreign material. In the present investigation, PMMA particles (50-60 nm) were mixed with human PMNs, and the amount of lactate dehydrogenase, lysozyme, and
P-glucuronidase released from the cells was quantitated. In all cases, a dosedependent increase in degranulation followed the addition of increasing amounts of PMMA to the PMNs. In addition, the migration of PMNs was diminished in a dose-dependent manner with exposure to increasing amounts of the cement. These results suggested that PMMA stimulates the release of leukocyte lysosomal contents and alters the migration characteristics of these cells in a manner that is consistent with the local inflammatory reaction to this cement.
A zone of radiolucency appears at the bone-cement interface within weeks following prosthetic repair of hip joints with polymethylmethacrylate (PMMA)-stabilized components.' This radiolucent layer is comprised of a synovial-like membrane that contains fibrous tissue with sheets of histiocytes and foreign body giant cells admixed with particulate PMMA debris.' Additionally, this membrane produces large amounts of osteolytic prostaglandin E2 (PGE2) and ~ollagenase.~ Aseptic loosening of prosthetic components, and, consequently, prosthetic joint failure, has been ascribed to cellular changes at the bone-cement interfa~e.~ An understanding of these changes may eventually lead to the design of biomaterials that are less likely to generate such an intense chronic inflammatory response. The particulate PMMA that is sloughed during the implantation of joint prostheses, and during the period of time such prostheses remain fixed in situ, stimulates the recruitment and proliferation of foamy histiocytes and multinucleated giant cells. Furthermore, polymorphonuclear leukocytes *To whom correspondence should be addressed. Journal of Biomedical Materials Research, Vol. 25, 761-771 (1991) CCC 0021-9304/91/060761-11$4.00 0 1991 John Wiley & Sons, Inc.
PAPAT H EOFANIS AND BARM ADA
(PMNs), lymphocytes, plasma cells, fibrocytes, and other marrow cells are also well-represented in the synovial-like membrane that encapsulates these prostheses at the bone-cement interface.2The continued presence of PMNs in this membrane suggests the possibility of a vital role for this cell type in the creation and maintenance of the membrane. Interestingly, the early participation of PMNs in the reaction of recipient tissues to PMMA remains poorly understood. PMNs serve a critical role in host defense and represent the initial circulating phagocytic capacity of an organism. PMN phagocytosis and chemotaxis are linked to respiratory burst activity. Phagocytosis is stimulated by activation of membranal NADPH oxidases which catalyze the reduction of oxygen to superoxide anion (02).5 D e g r a d a t i o n of lysosomal contents also occurs in conjunction with the activation of PMN phagocytic activity.‘ Furthermore, other inflammatory mediators are elaborated on activation of phagocytosis that include vasodilatory prostaglandins, chemotactic leukotrienes and complement components, and fever-producing interle~kins.~ The lysosomal granules that are released during phagocytosis contain an abundance of inflammatory mediators in the form of enzymes. Specific granules contain lysozyme, alkaline phosphatase, and collagenase, whereas azurophil granules contain acid hydrolases, cationic proteins, and some neutral proteases. Lactate dehydrogenase is an enzymatic marker of general cellular damage that may not be due to degranulation but is often released by cells during phagocytosis. Previous reports indicated that particulate PMMA (10-100 pm) stimulated a significant histologic response in recipient bone that was consistent with the characteristics of the synovial-like membrane that was described in loosened prosthetic joint surfaces.2 Particles in the range of nm to p m diameters are sloughed during the lifetime of a PMMAstabilized implant.’I8 The present investigation sought to determine whether PMNs degranulate in response to nanoparticulate PMMA, thereby indicating activation of these phagocytic cells. MATERIALS AND METHODS
Neutrophil preparation Blood was drawn by venipuncture from three volunteers who did not have implants of any kind and had not experienced infection or febrile illness in the 6 months prior to donation. The blood was drawn into 60-mL syringes and apportioned into tubes containing 1500 units sodium heparin dissolved in 5 mL HBSS. The blood and heparin were gently mixed and transferred to 16 x 125 mm plastic centrifuge tubes that already contained 5 mL of FicollHypaque solution. The Ficoll-Hypaque solution was prepared by mixing 40 g sodium diatrizoate and 25.4 g Ficoll400 (Sigma Chemical Co., St. Louis, MO) that was brought to a total volume of 267 mL with triple-distilled, deionized water. The gradient solution was subsequently passed through a 0.22-pm filter (Millipore Corp., Bedford, MA). The centrifuge tubes were spun at 350 g for 45 min. The neutrophil-rich band was removed, washed, and resuspended
to a final dilution of 1 X lo6 PMNs/mL HBSS and was stored in ice until used. The final cell suspension contained >98% viable cells by trypan blue exclusion, and >98% of these cells were PMNs by differential count.
Polymethylmethacrylate Polymethylmethacrylate (PMMA; Simplex) was purchased from Howmedica, Rutherford, NJ. PMMA was polymerized according to the manufacturer's instructions. The polymerized PMMA was milled and particles measuring 50-60 nm in diameter were isolated by employing sieves that permitted this cutoff range. The polymerized PMMA was washed with copious amounts of sterile saline before use. Polystyrene beads (50 nm in diameter) were employed as physical controls. Degranulation assays PMNs (1 x lo6 cells) were mixed with the indicated amounts of PMMA or polystyrene and incubated at 37°C in a gently shaking water bath for 15 min. On completion of the incubation period, the PMNs were centrifuged and the supernatant was assayed for lysozyme (EC 22.214.171.124), lactate dehydrogenase (LDH; EC 126.96.36.199), and P-glucuronidase (EC 188.8.131.52). Briefly, LDH release, a marker of general cellular damage, was determined by monitoring the rate of NADH oxidation at 340 nm. For this assay, purified beef heart (Worthington, Freehold, NJ) and sodium pyruvate served as standards.' Similarly, lysozyme degranulation, a marker for specific and azurophil granules, was determined by monitoring the rate of lysis of Micvococcus Zysodeikticus where purified chicken egg white lysozyme served as the enzyme standard. The rate of substrate lysis was monitored at 450 nm." Finally, P-glucuronidase, an azurophil granule marker, was determined following incubation with the substrate phenolphthalein-glucuronic acid. The extent of reaction was measured at 550 nm." The percent total LDH, lysozyme, and Pglucuronidase release were expressed relative to the percent of total (100%) enzyme activity released on exposure of PMNs to 0.2% Triton X-100 detergent. In this regard, 100% release of LDH, lysozyme, and P-glucuronidase corresponded to 390 -+ 62 AU/106 PMNs, 4.1 2 0.5 pg/106PMNs, and 2.0 +- 0.7 p g / lo6 PMNs, respectively. Where indicated, 0.1 pM f-met-leu-phe (Sigma), a secretagogue that triggers release of both specific and azurophil granules, was added to the incubation suspensions as a positive control of degranulation. Migration assays Briefly, PMNs (1 x lo6 cells), that had been previously incubated as indicated, were loaded into the upper compartment of Boyden chambers (Ahlco Corp., Southington, CT) according to method 2 of Maderazo and Woronick.12
PAPATHEOFANIS A N D BARMADA
After 60 min of air incubation, cell migration was determined by measuring the distance attained by the leading front cells in the chamber. Buffer served as the positive control whereas addition of 1 mM sodium cyanide to the PMNs served as the negative control for this assay system. Statistics
Experimental values represent the results of quadruplicate determinations (n = 4) and are expressed as the mean k standard deviation of the mean. The Student’s t distribution test was employed for analysis and p < 0.05 was accepted as significant. RESULTS
Neutrophil degranulation in response to f-met-leu-phe stimulation (i.e., positive degranulation control) resulted in the release of 51 ? 5.1% of lysozyme and 33 & 4.7% of P-glucuronidase contained in these cells. These amounts are comparable to literature values derived from stimulation with identical amounts of the same secretogogue: 45.8 5 4.4% (lysozyme) and If PMNs were boiled (5 min at 100OC) or 29.5 k 2.7% (p-gluc~ronidase).’~ treated with cyanide (i.e., negative degranulation controls) before addition of this secretogogue, no degranulation of lysosomes occurred following this stimulation (data not shown). Addition of polystyrene nanoparticles (physical controls) resulted in the release of all three enzymes in a statistically significant manner ( p < 0.05 at >10 mg polystyrene/mL) that was dosedependent (Fig. 1). The amounts of lysosomal enzymes released into the supernatant were significantly greater ( p < 0.05) where the PMNs were exposed to PMMA versus polystyrene nanoparticles. For example, the amount of LDH released (Fig. 2) was significantly greater on exposure to PMMA versus polystyrene nanoparticles. Addition of 50 and 100 mg/mL of PMMA resulted in 13.4% and 19.8% more LDH release, respectively, compared to the addition of polystyrene. Similar results were noted for the release of lysozyme (Fig. 3) and P-glucuronidase (Fig. 4). In all cases, the amount of enzyme released into the supernatant was greater following exposure to PMMA nanoparticles as compared to polystyrene nanoparticles. Also, the release of these three enzymes demonstrated a dose-dependent relationship relative to the amount of nanoparticles added. The amounts of nanoparticles added are representative of literature protocols and based on estimates of PMMA wear debris formation in model system^.'^ Figure 5 summarizes the decreased migration of PMNs exposed to increasing amounts of PMMA and polystyrene nanoparticles. At 50 mg/mL the distance travelled by the PMNs was 12% and 11.5% less on exposure to PMMA versus control and polystyrene nanoparticle exposure, respectively. In all cases, the decreased migration distance observed with PMMA exposure was significantly greater than that observed for control (94 2 4.6 mm) or polystyrene-exposed PMNs.
PMN DEGRAN UL ATION
8 7 6
3 2 1
Polystyrene, mg / ml Figure 1. Enzyme release following stimulation with increasing amounts of polystyrene nanoparticles. P-GlucU, P-glucuronidase; Lyso, lysozyme; LDH, lactate dehydrogenase; n = 4.
Loosening of PMMA-stabilized prosthetic components is an important and frequent complication of joint replacement. Indeed, osteolytic damage around PMMA-fixed prostheses has even been described in clinically functional, well-fixed femoral pro~theses.'~ PMMA debris is sloughed at the implant site 30 fl
Nanoparticles, mg / ml Figure 2. Release of lactate dehydrogenase following exposure to increasing amounts of PMMA and polystyrene nanoparticles. PMMA, polymethylmethacrylate; polyS, polystyrene; n = 4; 100% release corresponds to 390 f 62 AU/106 PMNs.
PAPATHEOFANIS A N D BARMADA
Nanoparticles, mg / ml Figure 3. Lysozyme release after exposure to increasing amounts of PMMA and polystyrene. 100% release corresponds to 4.1 2 0.5 &lo6 PMNs and n = 4.
as a result of mechanical wear and breakdown of rhe implant, and particulate debris is also generated during implantation and from residual unpolymerized cement. Little is known about the initial reaction of the reticuloendothelial system and the response of phagocytic mechanisms to the presence of PMMA in situ. Particulate PMMA debris at the bone-cement interface has been reported to stimulate a foreign body histiocytic and giant cell response after only weeks of implantation.’ Likewise, particulate PMMA alters
Nanoparticles, mg / ml Figure 4. Lysosomal release of P-glucuronidase following exposure to increasing amounts of PMMA and polystyrene nanoparticles. 100% release represents 2.0 0.7 pg/106 PMNs and n = 4.
Nanoparticles, mg / ml Figure 5. Migration distance of PMNs exposed to increasing amounts of particulate polystyrene (solid columns) and PMMA (stippled columns). Statistical values indicated above each matched set of data represent the result of experiments performed in quadruplicate.
macrophage, fibroblast, and osteoblast function: and the cytotoxicity of PMMA to these cells has been postulated to precipitate the granulomatous response observed at the bone-cement i n t e r f a ~ eParticulate .~ PMMA stimulates an inflammatory response at the cellular level that includes release of acid phosphataseI6 and gluco~aminidase~~ from macrophages and increased release of PGEz and IL-1 from the synovial-like membrane present at the bone-cement interfa~e.~”’ Preliminary results also indicate that PMMA stimulates the production of superoxide anion radical (02’) in PMNs exposed to particulate cement.” Figure 6 summarizes a possible mechanism that may account for PMMAmediated osteolysis at sites of prosthetic implantation. Particulate PMMA may activate membranal NADPH oxidases in phagocytic cells such as the PMN. Such activation generates superoxide anion radicals, and other partially reduced oxygen species, that react with membranal polyunsaturated fatty acids to produce lipid hydroperoxides.” These hydroperoxides are activators of prostaglandin and leukotriene biosynthesis. During such an oxidative burst, prostaglandins and leukotrienes are released by these activated cells. Increased amounts of Ca2+are also taken up by these activated cells. This Caz+activates proteases and lipases that attack the cell membrane. Also, oxidative phosphorylation is impaired in mitochondria as a result of the shift in intracellular calcium pools. Consequently, ATP depletion leads to destabilization of lysosomal membranes, and in conjunction with concomitant cell membrane destruction by Ca2+-drivenproteases and lipases, the contents of these lysosomes is expressed extracellularly. Alterations in ATP and Ca2+ homeostasis also lead to outright degranulation of lysosomal enzymes and the release of other inflammatory mediators. In combination, the release of these agents and enzymes contributes to osteolysis at the bone-cement
PAPATHEOFANIS AND BARMADA
LYSOZYME 4 1 DH ALKALINE PHOSPHATASC COLLAGENASE NEUTRAL PROTEASES GLUCOsAMl N IDAsE
DEGRANULAT I ON
Figure 6. A possible mechanism for PMMA-stimulated osteolysis at the site of prosthetic implantation. PUFA, polyunsaturated fatty acid; ROOH, lipid hydroperoxide; 20 :4, arachidonic acid; PGH synthase, prostaglandin G synthase; LTs, leukotrienes; PAF, platelet activating factor.
interface. Moreover, these mediators activate and recruit more cells to the site of the inflammatory reaction and therein potentiate the inflammatory response. The investigation of aseptic loosening of PMMA-stabilized components is complicated by a number of factors. Debate persists regarding the response of cells to bulk versus particulate PMMA in situ. Recent results, however, distinguish the florid inflammatory reaction and subsequent foreign body re-
PMN DEGRAN UL ATION
sponse to particulate PMMA as separate and distinct from any response to bulk cement that may occur over years of implantation.’ Moreover, most involved cells must phagocytize or respond to digestable debris before any reaction to this biomaterial may be initiated.7 However, phagocytic cells not only respond to particulate PMMA but also to inert materials such as polyethylene” and metals such as titanium.16 PGE2, collagenase, gelatinase, and IL-1 are produced by synovial-like membranes from cementless endoprostheses to the same or greater extent than that produced by cemented endoprostheses.” To some extent, the bone resorption observed at the interface between recipient tissues and PMMA may result from a very subtle alteration of focal phagocytic cell activity that may be due to local damage to the endosteum or periosteum. The present results indicated that initial phagocytic activity against particulate PMMA may be initiated by otherwise unstimulated, nascent PMNs, and that PMMA is cytotoxic to PMNs (i.e., degranulation of azurophilic and specific granules and decreased migration). PMNs are involved in phagocytic killing of ingested microorganisms that is related to oxygen consumption, free radical formation, alterations in cellular oxidoreductive capacity, and enzyme release secondary to degranulation and phagolysosome formation. Consequently, the persistence of PMNs in synovial-like membranes, albeit a very minor component of such membranes, at the bone-cement interface’ suggests that the role of PMNs might not only involve the initial phagocytic response to this biomaterial. Since PMNs elaborate a wide variety of chemotactic agents and other mediators of inflammation, they may serve a critical role in the chronic maintenance of an acute inflammatory reaction to PMMA wear debris. Ordinarily, after the initial neutrophilic attack, macrophages, fibroblasts, and vascular endothelial cells begin proliferating to synthesize local granulation tissue. This tissue is indicative of repair and characteristic of wound healing. This stage of repair may never be fully achieved in PMMA-recipient bone because of the persistence of PMNs in the involved tissues. The presence of PMNs may result in the chronic restimulation or reactivation of other phagocytic cells and in the persistent release of inflammatory mediators thereby creating a local cellular environment of cyclical acute inflammatory events. In effect, such a condition may account for the osteolysis at the PMMA interface since such lysis occurs at an accelerated rate during acute inflammationz3and is also evident within only a few weeks after prosthetic implantati~n.’~
References 1. H.-G. Willert, J. Ludwig, and M. Semlitsch, “Reaction of bone to methylmethacrylate after hip arthroplasty. A long-term gross, light microscopic, and scanning electron microscopic study,” J. Bone It. Surg., 56-A, 1368-1382 (1974). 2. S. B. Goodman, V. L. Fornasier, and J. Kei, “The effects of bulk versus particulate polymethylmethacrylate on bone,” CIin. Orthop., 232, 256262 (1988).
PAPATHEOFANIS AND BARMADA
5. 6. 7. 8.
S.R. Goldring, A.L. Schiller, M. Roelke, C.M. Rourke, D.A. O’Neill, and W. H. Harris, “The synovial-like membrane at the bone-cement interface in loose total hip replacements and its proposed role in bone lysis,” J. Bone It. Surg., 65-A, 575-584 (1983). S. R. Goldring, M. Jasty, M. Roelke, C.M. Rourke, F.R. Bringhurst, and W. H. Harris, “Formation of a synovial-like membrane at the bonecement interface,” Arthritis Rheum., 29, 836-845 (1986). P. Bellavite, E. Papini, L. Zeni, V. Della Bianca, and F. Rossi, ”Studies on the nature and activation of 02--forming NADPH oxidase of leukocytes,“ Free Rad. Res. Cornrn., 1, 11-29 (1985). B. M. Babior, “The respiratory burst of leukocytes,” I. Clin. Invest., 73, 599-604 (1984). R. Sandborg and J. Smolen, ”Early biochemical events in leukocyte activation,” Lab. Invest., 59, 300-305 (1988). M. Perry, C. Frondoza, L. Jones, and D. S. Hungerford, “The response of macrophages, fibroblasts and osteoblasts to PMMA and metal particles in tissue culture,” Trans. Orthop. Res. SOC.,15, 486 (1990). W. E.C. Wacker, D. D. Ulmer, and B. L. Vallee, “Metalloenzymes and myocardial infarction. 11. Malic and lactic dehydrogenase activities and zinc concentrations in serum,“ N.Engl. J. Med., 255,449-456 (1956). D. Shugar, ”Measurement of lysozyme activity and the ultraviolet inactivation of lysozyme,” Biockim. Biopkys. Acta., 8, 302-308 (1952). G. Brittinger, R. Hirschhorn, S. D. Douglas, and G. Weissman, ”Studies on lysosomes,” J Cell. Biol., 37, 394-411 (1968). E.G. Maderazo and C. L. Woronick, “Micropore filter assay of human granulocyte locomotion: Problems and solutions,” Clin. Immunol. lmmunoputkol., 11, 196-211 (1978). L. Kilpatrick, B.-Z. Garty, K.F. Lundquist, K. Hunter, C.A. Stanley, L. Baker, S. D. Douglas, and H. M. Korchak, “Impaired metabolic function and signaling defects in phagocytic cells in glycogen storage disease type lb,” J. Clin. Invest., 86, 196-202 (1990). E Betts, T. Wright, E. Salvati, and A. Boskey, ”Barium content of tissues from revision total hip arthroplasties,” Trans. Orthop. Res. Soc., 15, 457 (1990). W. J. Maloney, M. Jasty, C. R. Bragdon, and W. H. Harris, ”Focal osteolysis in association with a well-fixed cemented femoral component,” Trans. Orthop. Res. Soc., 15, 450 (1990). J. Mitchell, B. Evans, P. Ducheyne, and J. Cuckler, ”Comparison of the inflammatory response of PMMAkitanium particles in the rat intramuscular implant model,” Trans. Orthop. Res. Soc., 15, 485 (1990). R.G. Davis, R. L. Smith, S.B. Goodman, and J. L. Lerman, ”Bone cement stimulates lysosomal enzyme activity in adherent mononuclear cells,” Trans. Orthop. Res. SOC.,15, 234 (1990). T.S. Thornhill, S. Shortkroff, H.-P. Hsu, K. Keller, M. Spector, C. Mintzer, R. Nowak, and N. Lane, ‘A canine model for the study of the interface ’pseudo-membrane’ in loose cemented femoral stems,” Trans. Orthop. Res. Soc., 15, 232 (1990). F. J. Papatheofanis and R. Barmada, “Superoxide anion radical production by neutrophils on exposure to polymethylmethacrylate,” Trans. Orthop. Res. Soc., 15, 484 (1990). F. J. Papatheofanis and W. E. M. Lands, ”Lipoxygenase Mechanisms,” in Biochemistry of Arackidonic Acid Metabolism, Martinus Nijhoff Publ., Boston, 1985, pp. 9-39. P. Campbell, S. Nasser, D. Millett, and H.C. Amstutz, “A study of the effects of polyethylene wear debris in cemented and uncemented implants,” Trans. Orthop. Res. Soc., 15, 441 (1990).
PM N DEGRAN U L AT ION
K. J. Kim, S.C. Wilson, and H. E. Rubash, “Comparison study of interface tissues in cementless and cemented prostheses,” Trans. Orthop. Xes. Soc., 15, 236 (1990). 23. J.T. Chambers, “The pathophysiology of the osteoclast,” J. Clin. Puthol., 38, 241-247 (1985). 22.
Received May 18, 1990 Accepted February 6, 1991