Mast Cell Protease 5 Mediates Ischemia-Reperfusion Injury of Mouse Skeletal Muscle1
http://www.100md.com
免疫学杂志 2005年第11期
Abstract
Ischemia with subsequent reperfusion (IR) injury is a significant clinical problem that occurs after physical and surgical trauma, myocardial infarction, and organ transplantation. IR injury of mouse skeletal muscle depends on the presence of both natural IgM and an intact C pathway. Disruption of the skeletal muscle architecture and permeability also requires mast cell (MC) participation, as revealed by the fact that IR injury is markedly reduced in c-kit defective, MC-deficient mouse strains. In this study, we sought to identify the pathobiologic MC products expressed in IR injury using transgenic mouse strains with normal MC development, except for the lack of a particular MC-derived mediator. Histologic analysis of skeletal muscle from BALB/c and C57BL/6 mice revealed a strong positive correlation (R2 = 0.85) between the extent of IR injury and the level of MC degranulation. Linkage between C activation and MC degranulation was demonstrated in mice lacking C4, in which only limited MC degranulation and muscle injury were apparent. No reduction in injury was observed in transgenic mice lacking leukotriene C4 synthase, hemopoietic PGD2 synthase, N-deacetylase/N-sulfotransferase-2 (enzyme involved in heparin biosynthesis), or mouse MC protease (mMCP) 1. In contrast, muscle injury was significantly attenuated in mMCP-5-null mice. The MCs that reside in skeletal muscle contain abundant amounts of mMCP-5 which is the serine protease that is most similar in sequence to human MC chymase. We now report a cytotoxic activity associated with a MC-specific protease and demonstrate that mMCP-5 is critical for irreversible IR injury of skeletal muscle.
Introduction
The classical C pathway and naturally occurring IgM from B-1 lymphocytes are essential components of hind limb ischemia-reperfusion (IR)4 injury in the mouse (1, 2). These data have led to the hypothesis that activation of the classical C pathway with formation of membrane-attack complexes leads to muscle injury. However, the finding that muscle damage is markedly attenuated in C-sufficient but mast cell (MC)-deficient B6D2F1-KitWf/KitWf (Wf/Wf) and WBB6F1-KitW/KitW (W/Wv) mice (3, 4, 5) indicates that IR-mediated injury is a more complex pathologic response than currently understood and that MCs play a prominent role in this cytotoxic injury.
After activation, mature MCs quickly exocytose the contents of their secretory granules, which include histamine, serotonin, heparin- and chondroitin sulfate-serglycin proteoglycans, multiple serine proteases, carboxypeptidase A3, and preformed cytokines such as TNF-. MC activation also results in the rapid generation of the eicosanoids PGD2 and cysteinyl leukotriene C4 (LTC4) (6, 7), followed hours later by increased expression of numerous cytokines and chemokines (8, 9). Although this diverse array of MC-derived products contributes to the vascular leakage and cellular infiltration characteristic of a brisk host inflammatory response, it has not been shown that any of these MC-derived mediators participate in IR-mediated injury of skeletal muscle cells. Classical activation of MCs in adaptive immune responses occurs by cross-linking of IgE bound to FcRI or by the direct interaction of preformed immune complexes with the low-affinity Ig receptor FcRIII (10, 11). In addition, several other signals associated with innate immunity, including the anaphylatoxins C3a and C5a and adenosine, can activate MCs (12, 13, 14, 15, 16).
The apparent requirement for both the classical C pathway and MCs in IR-induced injury suggested that MC degranulation occurs as a consequence of C activation. In this study, we sought to test the hypothesis that the MC is a critical effector cell downstream of C and then to establish the role of the MC independent of the broad impact of c-kit deficiency on the hemopoietic lineage and other cell types (17, 18, 19). We demonstrate reduced levels of both MC degranulation and muscle injury in C4-deficient mice, thereby confirming an essential role for the C pathway in IR injury of muscle. We then delineate the contribution of individual MC-derived mediators by quantitative analysis of the extent of hind limb IR injury and MC degranulation in MC-sufficient strains with targeted disruption of a particular mediator. We demonstrate that of all of the tested MC mediators, only mouse MC protease (mMCP)-5 plays a significant effector role in severe IR-induced rhabdomyolysis of skeletal muscle.
Materials and Methods
Statistical analysis
Student’s t test for paired samples was used for direct comparisons of observations made in control and knockout animals. Pearson’s correlation coefficient was used in estimating the relationship between muscle injury and MC degranulation for wild-type control animals. A p <0.05 was considered to be statistically significant.
Results
Time course of muscle injury and MC degranulation after reperfusion of ischemic skeletal muscle
In our initial study of hind limb IR injury in W/Wv mice and their +/+ littermates, MC degranulation and muscle injury were assessed after 3 h of reperfusion (4). Others have assessed muscle damage 24 h after reperfusion (3, 5). To determine the time course of MC degranulation relative to skeletal muscle injury, we assessed each at 1, 3, and 6 h after the initiation of reperfusion in C57BL/6 mice. Skeletal muscle damage was expressed as the number of injured fibers per 50 fc. MC degranulation was quantitated as the number of degranulated MC in 20 low-power (x10 objective) fields (LPF). Both irreversible muscle damage and MC degranulation exhibited time-dependent changes only after the initiation of reperfusion. Although skeletal muscle injury was maximal at the 3-h time point, substantial injury was seen consistently 1 h after the initiation of reperfusion (Fig. 1A). MC degranulation showed a similar time course, with detectable activation and exocytosis of granules occurring within 1 h and reaching maximal levels by 3 h (Fig. 1B). Comparable muscle damage and MC degranulation also were noted in the IR-treated BALB/c mouse (Fig. 2). For both wild-type mouse strains, the extent of muscle damage was strongly correlated with the percentage of MCs exhibiting degranulation (R2 = 0.85, Fig. 2).
The C-derived peptides C3a and C5a are capable of eliciting activation and secretion of mediators from MCs (12, 13, 14), and both anaphylatoxins are generated during the formation of the membrane-attack complex C5b-9. C5a has been implicated in IR-induced injury of the kidneys (30). However, when we evaluated strains that lacked one of the anaphylatoxin receptors, either C3aR or C5aR, the extent of MC degranulation and muscle injury was comparable to that of their wild-type littermates (Fig. 2 and Table I). These unexpected findings suggest either that C3aR and C5aR have redundant activities or that the MCs are activated via another C-dependent mechanism.
Role of specific MC mediators in IR-induced muscle injury
Mouse MCs elaborate numerous cytokines/chemokines, most of which are expressed 2–6 h after cellular activation in vitro. In our time course study (Fig. 1), substantial muscle damage was apparent after 1 h of reperfusion, suggesting little opportunity for transiently induced cytokines and chemokines to play a significant role. However, because some populations of mouse MCs store small amounts of preformed TNF- in their secretory granules (31), we evaluated IR injury in a mouse strain unable to express this cytokine and determined that muscle injury was not attenuated (5 mice, n = 1, data not shown).
The cysteinyl leukotrienes are permeability-enhancing mediators, and mice lacking LTC4S do not exhibit increases in MC- and macrophage-dependent microvascular permeability (23, 32). PGD2 is the other prominent eicosanoid released from activated MCs, and this arachidonic acid metabolite also increases microvascular permeability (33). IR-induced muscle-cell damage in LTC4S- and hemopoietic PGD2S-null mice was found to be comparable to that in wild-type mice (Fig. 2 and Table II). Thus, the two most prominent eicosanoids generated from MCs do not contribute significantly to MC-dependent IR-mediated skeletal muscle damage.
Based on the rapid time frame associated with IR injury, the MC-derived mediators most likely involved are the preformed secretory granule proteases. The MCs that reside in the muscles of BALB/c mice express mMCP-5 and mMCP-7, but not mouse transmembrane tryptase (mTMT)/tryptase /Prss31 (38). The MCs in the C57BL/6 mouse are unable to express mMCP-7 because of a point mutation in the exon 2/intron 2 splice site of this tryptase gene (39). In contrast, the levels of mTMT are markedly increased in the MCs of this mouse strain relative to those in BALB/c mice (38). The similarity in the extent of hind limb IR injury in the BALB/c and C57BL/6 mouse strains (Fig. 2) implies that neither mMCP-7 nor mTMT is the primary MC-derived factor that mediates muscle injury. Because we previously observed that some intra-alveolar macrophages are coated with immunoreactive mMCP-1 in our hind limb IR model (4), we considered the possibility that mMCP-1 released from mucosal MCs could exert its effects at a distant site such as the muscle. The extent of muscle injury and MC activation-secretion in mMCP-1-null mice was comparable to that in wild-type mice (Fig. 2 and Table II). Thus, mMCP-1 is not the relevant factor.
Of the remaining mMCP candidates, mMCP-5 is the most abundant serine protease in every safranin-positive mouse MC examined to date (40). After IR treatment of wild-type and mMCP-5/BL6–/– mice, mean muscle injury was 48 ± 4.7% and 15.6 ± 2.1% (mean ± SEM, n = 4, p < 0.01), respectively, for a reduction in injury of 68% (Figs. 2 and 4), even though MC degranulation was comparable (63 ± 7.5% and 73 ± 10%, respectively). To exclude other defects that may have occurred during development of this mMCP-5-null mouse strain, we also examined a second mMCP-5-null strain produced from an independently derived embryonic stem cell clone and backcrossed onto the BALB/c genetic background. These mice also showed dramatic protection against muscle injury; mean injury in the BALB/c controls was 60 ± 3.3% (mean ± SEM, 12 mice, 2 experiments) while the mean injury in the 5 mMCP-5/BALB–/– mice was 19.4 ± 3.1% (p < 0.001) for a reduction in injury of 68% (Fig. 2). MC degranulation was comparable, 74.3 and 75.7%, respectively. The extent of muscle injury relative to MC degranulation in the two mMCP-5-null strains was well below the trend line for all other strains and fell outside of the 95% confidence interval for this analysis (Fig. 2). Excluding mMCP-5, the correlation coefficient for the trend line including all of the normal mice and transgenic strains with targeted deletions tested was R2 = 0.83.
The normal IR injury in the NDST-2 null mice that still transcribe the gene but are deficient in granule-stored mMCP-5 implies that the NDST-2 null MC could still be providing the protease needed for the injury. Frozen sections of skeletal muscle, immunohistochemically evaluated for the presence of mMCP-5, showed a diffuse staining pattern associated with the MCs of the NDST-2–/– mice (Fig. 5A). No mMCP-5 was detectable in the MCs of mMCP-5/BL6–/– mice (Fig. 5B) and a distinct granule staining pattern was associated with the skeletal muscle MC in wild-type C57BL/6 mice (Fig. 5C). The chloroacetate esterase reactivity demonstrates the smaller size of the MCs found in the skeletal muscle of these two null strains (Fig. 5, D and 5, E) as compared with the MCs in the wild-type strain (Fig. 5F). The chloroacetate esterase activity in MCs of the two null strains is attributed to the presence of mMCP-4 that has been detected by activity in another NDST-2–/– strain (36) and by immunoblot analysis in peritoneal MCs from the mMCP-5/BL6–/– mice (data not shown).
Discussion
We have identified a cytotoxic function for one MC granule protease in IR injury of mouse skeletal muscle. We made this observation while establishing that MCs participate quantitatively with natural IgM and C in producing IR injury of hind limb skeletal muscle of the mouse. The results demonstrate that the prominent MC secretory granule protease mMCP-5 is essential for the occurrence of irreversible muscle damage and that its release depends on classical C activation. The extensive skeletal muscle injury with local MC activation during IR contrasts with the absence of severe local cellular injury after conventional MC activation through receptors such as FcRI, FcRIII, or c-kit.
These studies extend earlier results that implicated the MCs in hind limb IR by demonstrating protection against injury in MC-deficient Wf/Wf and W/Wv mouse strains (3, 4, 5). The most recent assessment by Bortolotto et al. (5) demonstrated that reconstitution of the skeletal muscle with <10% of the normal number of MCs was able to restore 85% of the injury observed in the wild-type controls. Yet, the genetic abnormality of the receptor tyrosine kinase c-kit in these strains not only prevents survival of peripheral tissue MCs but also creates deficiencies in other hemopoietic lineages, such as erythrocytes and neutrophils, as well as affects melanogenesis and gametogenesis (17, 18, 19, 41, 42, 43). Thus, our results confirm the role of the MC in a mouse strain with normal hemopoietic lineage development and demonstrate the critical role of the MC proteases by identifying a requirement for a specific MC function through targeted disruption of a MC-specific gene.
The role of the local MCs in producing irreversible skeletal muscle injury was suggested by time-dependent studies in which a quantitative progression of cell injury was associated with an incremental reduction of intact MCs at the lesion (Fig. 1) and not in adjacent unaffected skeletal muscle (data not shown). This interpretation is supported by an analysis of the percentage of injured muscle fibers and the percentage of degranulated MCs in the same muscle (Fig. 2). The percentage of injured skeletal muscle fibers is directly correlated with the percentage of degranulated MCs in the injury site with a correlation coefficient of 0.85, indicating that 85% of the muscle injury is directly correlated with MC activation. Analysis of the various null mutant mice shows that with the exception of the mMCP-5-null mice, values from the other transgenic animals evaluated in our study lie near the linear trend line for the comparison of skeletal muscle injury and loss of intact MCs obtained from injured +/+ control mice. The skeletal muscle MCs of both the mMCP-5- and NDST-2-null strains have small granules due to the absence of core components of the secretory granule complex and a concomitant small cell size (Fig. 5). Trichinella spiralis infection in each of the strains elicits mucosal MCs in normal numbers and size, indicating that lineage development is normal and that the small size of the skeletal muscle MC only reflects the decrease in granule constituents (Ref.20 and R. L. Stevens, unpublished data). Nonetheless, they can be enumerated and assessed for exocytosis with chloroacetate esterase staining in both resting and injured skeletal muscle (Figs. 4 and 5). Importantly, both possessed virtually the same number of MCs per 20 LPF in their hind limb skeletal muscle and responded to IR with comparable levels of degranulation. Yet the mMCP-5 null mice experienced minimal skeletal muscle injury, whereas the NDST-2 null mice were severely injured. The surprising finding that the degree of IR injury in the NDST-2-null mouse strain is similar to that in wild-type mice suggests that the small amount of mMCP-5 detected by immunohistochemistry (Fig. 5) is sufficient for extensive muscle damage in this animal model. By analogy to the finding that heparin restricts the enzymatic activity of recombinant mMCP-6 (44) and purified rat MC protease-1 (45), it is possible that the absence of heparin in the NDST-2-null mouse results in a marked increase in the relative lytic activity of the small amount of mMCP-5. Despite their granule storage deficiency, the mMCP-5 gene is transcribed at a normal rate in the MCs of NDST-2-null mice (20). Thus, another possibility is that extensive muscle damage occurs in NDST-2-null mice because the constitutive release of mMCP-5 augments the response of the tissue to other factors active in hind limb IR injury. Because the NDST-2-null mouse also has less MC carboxypeptidase A (20, 36, 37, 46), any role for this exopeptidase acting in concert with mMCP-5 would again depend on the lack of heparin binding.
The fact that this MC-dependent cytotoxicity occurs in a setting of IR with C activation preceding degranulation argues for a critical combination of events, including binding of natural IgM, which leads to activation of C and then MCs. The critical role of mMCP-5 in this setting may be operative via several different mechanisms. The protease might directly activate a pronecrotic pathway or inactivate an antinecrotic pathway outside the muscle cell or inside the cell if it is able to gain entry via the C5b-9 membrane pore. Alternately, mMCP-5 could activate a protease cascade that ultimately results in proteolytic disruption of the basement membrane, sarcolemma, the subsarcolemma cytoskeleton of skeletal muscle, or a factor that regulates the viability of the muscle cell. Whatever its mechanism, our findings demonstrate that irreversible skeletal muscle injury in IR-treated mice is exquisitely dependent on mMCP-5 and independent of several other MC-derived mediators.
]
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by National Institutes of Health Grants GM-52585, HL-36110, and HL-63284.
2 Current address: Cincinnati Children’s Hospital Medical Center, 3333 Burnet Avenue, ML7028, Cincinnati, OH 45229-3039.
3 Address correspondence and reprint requests to Dr. Michael F. Gurish, Brigham and Women’s Hospital, Smith Building, Room 624, 1 Jimmy Fund Way, Boston, MA 02115. E-mail address: mgurish{at}rics.bwh.harvard.edu
4 Abbreviations used in this paper: IR, ischemia-reperfusion; fc, fibers counted; LPF, low-power field; LTC4, leukotriene C4; LTC4S, LTC4 synthase; PGD2S, PGD2 synthase; MC, mast cell; mMCP, mouse MC protease; NDST, N-deacetylase/N-sulfotransferase; mTMT, mouse transmembrane tryptase.
Received for publication May 14, 2004. Accepted for publication March 17, 2005.
References
Weiser, M. R., J. P. Williams, F. D. Moore, Jr, L. Kobzik, M. Ma, H. B. Hechtman, M. C. Carroll. 1996. Reperfusion injury of ischemic skeletal muscle is mediated by natural antibody and complement. J. Exp. Med. 183: 2343-2348.
Kyriakides, C., W. Austen, Jr, Y. Wang, J. Favuzza, L. Kobzik, F. D. Moore, Jr, H. B. Hechtman. 1999. Skeletal muscle reperfusion injury is mediated by neutrophils and the complement membrane attack complex. Am. J. Physiol. 277: C1263-C1268.
Lazarus, B., A. Messina, J. E. Barker, J. V. Hurley, R. Romeo, W. A. Morrison, K. R. Knight. 2000. The role of mast cells in ischaemia-reperfusion injury in murine skeletal muscle. J. Pathol. 191: 443-448.
Mukundan, C., M. F. Gurish, K. F. Austen, H. B. Hechtman, D. S. Friend. 2001. Mast cell mediation of muscle and pulmonary injury following hindlimb ischemia-reperfusion. J. Histochem. Cytochem. 49: 1055-1056.
Bortolotto, S. K., W. A. Morrison, X. Han, A. Messina. 2004. Mast cells play a pivotal role in ischaemia reperfusion injury to skeletal muscles. Lab. Invest. 84: 1103-1111.
Katz, H. R., M. B. Raizman, C. S. Gartner, H. C. Scott, A. C. Benson, K. F. Austen. 1992. Secretory granule mediator release and generation of oxidative metabolites of arachidonic acid via Fc-IgG receptor bridging in mouse mast cells. J. Immunol. 148: 868-871.
Levi-Schaffer, F., E. T. Dayton, K. F. Austen, A. Hein, J. P. Caulfield, P. M. Gravallese, F. T. Liu, R. L. Stevens. 1987. Mouse bone marrow-derived mast cells cocultured with fibroblasts: morphology and stimulation-induced release of histamine, leukotriene B4, leukotriene C4, and prostaglandin D2. J. Immunol. 139: 3431-3441.(J. Pablo Abonia2,*, Danie)
Ischemia with subsequent reperfusion (IR) injury is a significant clinical problem that occurs after physical and surgical trauma, myocardial infarction, and organ transplantation. IR injury of mouse skeletal muscle depends on the presence of both natural IgM and an intact C pathway. Disruption of the skeletal muscle architecture and permeability also requires mast cell (MC) participation, as revealed by the fact that IR injury is markedly reduced in c-kit defective, MC-deficient mouse strains. In this study, we sought to identify the pathobiologic MC products expressed in IR injury using transgenic mouse strains with normal MC development, except for the lack of a particular MC-derived mediator. Histologic analysis of skeletal muscle from BALB/c and C57BL/6 mice revealed a strong positive correlation (R2 = 0.85) between the extent of IR injury and the level of MC degranulation. Linkage between C activation and MC degranulation was demonstrated in mice lacking C4, in which only limited MC degranulation and muscle injury were apparent. No reduction in injury was observed in transgenic mice lacking leukotriene C4 synthase, hemopoietic PGD2 synthase, N-deacetylase/N-sulfotransferase-2 (enzyme involved in heparin biosynthesis), or mouse MC protease (mMCP) 1. In contrast, muscle injury was significantly attenuated in mMCP-5-null mice. The MCs that reside in skeletal muscle contain abundant amounts of mMCP-5 which is the serine protease that is most similar in sequence to human MC chymase. We now report a cytotoxic activity associated with a MC-specific protease and demonstrate that mMCP-5 is critical for irreversible IR injury of skeletal muscle.
Introduction
The classical C pathway and naturally occurring IgM from B-1 lymphocytes are essential components of hind limb ischemia-reperfusion (IR)4 injury in the mouse (1, 2). These data have led to the hypothesis that activation of the classical C pathway with formation of membrane-attack complexes leads to muscle injury. However, the finding that muscle damage is markedly attenuated in C-sufficient but mast cell (MC)-deficient B6D2F1-KitWf/KitWf (Wf/Wf) and WBB6F1-KitW/KitW (W/Wv) mice (3, 4, 5) indicates that IR-mediated injury is a more complex pathologic response than currently understood and that MCs play a prominent role in this cytotoxic injury.
After activation, mature MCs quickly exocytose the contents of their secretory granules, which include histamine, serotonin, heparin- and chondroitin sulfate-serglycin proteoglycans, multiple serine proteases, carboxypeptidase A3, and preformed cytokines such as TNF-. MC activation also results in the rapid generation of the eicosanoids PGD2 and cysteinyl leukotriene C4 (LTC4) (6, 7), followed hours later by increased expression of numerous cytokines and chemokines (8, 9). Although this diverse array of MC-derived products contributes to the vascular leakage and cellular infiltration characteristic of a brisk host inflammatory response, it has not been shown that any of these MC-derived mediators participate in IR-mediated injury of skeletal muscle cells. Classical activation of MCs in adaptive immune responses occurs by cross-linking of IgE bound to FcRI or by the direct interaction of preformed immune complexes with the low-affinity Ig receptor FcRIII (10, 11). In addition, several other signals associated with innate immunity, including the anaphylatoxins C3a and C5a and adenosine, can activate MCs (12, 13, 14, 15, 16).
The apparent requirement for both the classical C pathway and MCs in IR-induced injury suggested that MC degranulation occurs as a consequence of C activation. In this study, we sought to test the hypothesis that the MC is a critical effector cell downstream of C and then to establish the role of the MC independent of the broad impact of c-kit deficiency on the hemopoietic lineage and other cell types (17, 18, 19). We demonstrate reduced levels of both MC degranulation and muscle injury in C4-deficient mice, thereby confirming an essential role for the C pathway in IR injury of muscle. We then delineate the contribution of individual MC-derived mediators by quantitative analysis of the extent of hind limb IR injury and MC degranulation in MC-sufficient strains with targeted disruption of a particular mediator. We demonstrate that of all of the tested MC mediators, only mouse MC protease (mMCP)-5 plays a significant effector role in severe IR-induced rhabdomyolysis of skeletal muscle.
Materials and Methods
Statistical analysis
Student’s t test for paired samples was used for direct comparisons of observations made in control and knockout animals. Pearson’s correlation coefficient was used in estimating the relationship between muscle injury and MC degranulation for wild-type control animals. A p <0.05 was considered to be statistically significant.
Results
Time course of muscle injury and MC degranulation after reperfusion of ischemic skeletal muscle
In our initial study of hind limb IR injury in W/Wv mice and their +/+ littermates, MC degranulation and muscle injury were assessed after 3 h of reperfusion (4). Others have assessed muscle damage 24 h after reperfusion (3, 5). To determine the time course of MC degranulation relative to skeletal muscle injury, we assessed each at 1, 3, and 6 h after the initiation of reperfusion in C57BL/6 mice. Skeletal muscle damage was expressed as the number of injured fibers per 50 fc. MC degranulation was quantitated as the number of degranulated MC in 20 low-power (x10 objective) fields (LPF). Both irreversible muscle damage and MC degranulation exhibited time-dependent changes only after the initiation of reperfusion. Although skeletal muscle injury was maximal at the 3-h time point, substantial injury was seen consistently 1 h after the initiation of reperfusion (Fig. 1A). MC degranulation showed a similar time course, with detectable activation and exocytosis of granules occurring within 1 h and reaching maximal levels by 3 h (Fig. 1B). Comparable muscle damage and MC degranulation also were noted in the IR-treated BALB/c mouse (Fig. 2). For both wild-type mouse strains, the extent of muscle damage was strongly correlated with the percentage of MCs exhibiting degranulation (R2 = 0.85, Fig. 2).
The C-derived peptides C3a and C5a are capable of eliciting activation and secretion of mediators from MCs (12, 13, 14), and both anaphylatoxins are generated during the formation of the membrane-attack complex C5b-9. C5a has been implicated in IR-induced injury of the kidneys (30). However, when we evaluated strains that lacked one of the anaphylatoxin receptors, either C3aR or C5aR, the extent of MC degranulation and muscle injury was comparable to that of their wild-type littermates (Fig. 2 and Table I). These unexpected findings suggest either that C3aR and C5aR have redundant activities or that the MCs are activated via another C-dependent mechanism.
Role of specific MC mediators in IR-induced muscle injury
Mouse MCs elaborate numerous cytokines/chemokines, most of which are expressed 2–6 h after cellular activation in vitro. In our time course study (Fig. 1), substantial muscle damage was apparent after 1 h of reperfusion, suggesting little opportunity for transiently induced cytokines and chemokines to play a significant role. However, because some populations of mouse MCs store small amounts of preformed TNF- in their secretory granules (31), we evaluated IR injury in a mouse strain unable to express this cytokine and determined that muscle injury was not attenuated (5 mice, n = 1, data not shown).
The cysteinyl leukotrienes are permeability-enhancing mediators, and mice lacking LTC4S do not exhibit increases in MC- and macrophage-dependent microvascular permeability (23, 32). PGD2 is the other prominent eicosanoid released from activated MCs, and this arachidonic acid metabolite also increases microvascular permeability (33). IR-induced muscle-cell damage in LTC4S- and hemopoietic PGD2S-null mice was found to be comparable to that in wild-type mice (Fig. 2 and Table II). Thus, the two most prominent eicosanoids generated from MCs do not contribute significantly to MC-dependent IR-mediated skeletal muscle damage.
Based on the rapid time frame associated with IR injury, the MC-derived mediators most likely involved are the preformed secretory granule proteases. The MCs that reside in the muscles of BALB/c mice express mMCP-5 and mMCP-7, but not mouse transmembrane tryptase (mTMT)/tryptase /Prss31 (38). The MCs in the C57BL/6 mouse are unable to express mMCP-7 because of a point mutation in the exon 2/intron 2 splice site of this tryptase gene (39). In contrast, the levels of mTMT are markedly increased in the MCs of this mouse strain relative to those in BALB/c mice (38). The similarity in the extent of hind limb IR injury in the BALB/c and C57BL/6 mouse strains (Fig. 2) implies that neither mMCP-7 nor mTMT is the primary MC-derived factor that mediates muscle injury. Because we previously observed that some intra-alveolar macrophages are coated with immunoreactive mMCP-1 in our hind limb IR model (4), we considered the possibility that mMCP-1 released from mucosal MCs could exert its effects at a distant site such as the muscle. The extent of muscle injury and MC activation-secretion in mMCP-1-null mice was comparable to that in wild-type mice (Fig. 2 and Table II). Thus, mMCP-1 is not the relevant factor.
Of the remaining mMCP candidates, mMCP-5 is the most abundant serine protease in every safranin-positive mouse MC examined to date (40). After IR treatment of wild-type and mMCP-5/BL6–/– mice, mean muscle injury was 48 ± 4.7% and 15.6 ± 2.1% (mean ± SEM, n = 4, p < 0.01), respectively, for a reduction in injury of 68% (Figs. 2 and 4), even though MC degranulation was comparable (63 ± 7.5% and 73 ± 10%, respectively). To exclude other defects that may have occurred during development of this mMCP-5-null mouse strain, we also examined a second mMCP-5-null strain produced from an independently derived embryonic stem cell clone and backcrossed onto the BALB/c genetic background. These mice also showed dramatic protection against muscle injury; mean injury in the BALB/c controls was 60 ± 3.3% (mean ± SEM, 12 mice, 2 experiments) while the mean injury in the 5 mMCP-5/BALB–/– mice was 19.4 ± 3.1% (p < 0.001) for a reduction in injury of 68% (Fig. 2). MC degranulation was comparable, 74.3 and 75.7%, respectively. The extent of muscle injury relative to MC degranulation in the two mMCP-5-null strains was well below the trend line for all other strains and fell outside of the 95% confidence interval for this analysis (Fig. 2). Excluding mMCP-5, the correlation coefficient for the trend line including all of the normal mice and transgenic strains with targeted deletions tested was R2 = 0.83.
The normal IR injury in the NDST-2 null mice that still transcribe the gene but are deficient in granule-stored mMCP-5 implies that the NDST-2 null MC could still be providing the protease needed for the injury. Frozen sections of skeletal muscle, immunohistochemically evaluated for the presence of mMCP-5, showed a diffuse staining pattern associated with the MCs of the NDST-2–/– mice (Fig. 5A). No mMCP-5 was detectable in the MCs of mMCP-5/BL6–/– mice (Fig. 5B) and a distinct granule staining pattern was associated with the skeletal muscle MC in wild-type C57BL/6 mice (Fig. 5C). The chloroacetate esterase reactivity demonstrates the smaller size of the MCs found in the skeletal muscle of these two null strains (Fig. 5, D and 5, E) as compared with the MCs in the wild-type strain (Fig. 5F). The chloroacetate esterase activity in MCs of the two null strains is attributed to the presence of mMCP-4 that has been detected by activity in another NDST-2–/– strain (36) and by immunoblot analysis in peritoneal MCs from the mMCP-5/BL6–/– mice (data not shown).
Discussion
We have identified a cytotoxic function for one MC granule protease in IR injury of mouse skeletal muscle. We made this observation while establishing that MCs participate quantitatively with natural IgM and C in producing IR injury of hind limb skeletal muscle of the mouse. The results demonstrate that the prominent MC secretory granule protease mMCP-5 is essential for the occurrence of irreversible muscle damage and that its release depends on classical C activation. The extensive skeletal muscle injury with local MC activation during IR contrasts with the absence of severe local cellular injury after conventional MC activation through receptors such as FcRI, FcRIII, or c-kit.
These studies extend earlier results that implicated the MCs in hind limb IR by demonstrating protection against injury in MC-deficient Wf/Wf and W/Wv mouse strains (3, 4, 5). The most recent assessment by Bortolotto et al. (5) demonstrated that reconstitution of the skeletal muscle with <10% of the normal number of MCs was able to restore 85% of the injury observed in the wild-type controls. Yet, the genetic abnormality of the receptor tyrosine kinase c-kit in these strains not only prevents survival of peripheral tissue MCs but also creates deficiencies in other hemopoietic lineages, such as erythrocytes and neutrophils, as well as affects melanogenesis and gametogenesis (17, 18, 19, 41, 42, 43). Thus, our results confirm the role of the MC in a mouse strain with normal hemopoietic lineage development and demonstrate the critical role of the MC proteases by identifying a requirement for a specific MC function through targeted disruption of a MC-specific gene.
The role of the local MCs in producing irreversible skeletal muscle injury was suggested by time-dependent studies in which a quantitative progression of cell injury was associated with an incremental reduction of intact MCs at the lesion (Fig. 1) and not in adjacent unaffected skeletal muscle (data not shown). This interpretation is supported by an analysis of the percentage of injured muscle fibers and the percentage of degranulated MCs in the same muscle (Fig. 2). The percentage of injured skeletal muscle fibers is directly correlated with the percentage of degranulated MCs in the injury site with a correlation coefficient of 0.85, indicating that 85% of the muscle injury is directly correlated with MC activation. Analysis of the various null mutant mice shows that with the exception of the mMCP-5-null mice, values from the other transgenic animals evaluated in our study lie near the linear trend line for the comparison of skeletal muscle injury and loss of intact MCs obtained from injured +/+ control mice. The skeletal muscle MCs of both the mMCP-5- and NDST-2-null strains have small granules due to the absence of core components of the secretory granule complex and a concomitant small cell size (Fig. 5). Trichinella spiralis infection in each of the strains elicits mucosal MCs in normal numbers and size, indicating that lineage development is normal and that the small size of the skeletal muscle MC only reflects the decrease in granule constituents (Ref.20 and R. L. Stevens, unpublished data). Nonetheless, they can be enumerated and assessed for exocytosis with chloroacetate esterase staining in both resting and injured skeletal muscle (Figs. 4 and 5). Importantly, both possessed virtually the same number of MCs per 20 LPF in their hind limb skeletal muscle and responded to IR with comparable levels of degranulation. Yet the mMCP-5 null mice experienced minimal skeletal muscle injury, whereas the NDST-2 null mice were severely injured. The surprising finding that the degree of IR injury in the NDST-2-null mouse strain is similar to that in wild-type mice suggests that the small amount of mMCP-5 detected by immunohistochemistry (Fig. 5) is sufficient for extensive muscle damage in this animal model. By analogy to the finding that heparin restricts the enzymatic activity of recombinant mMCP-6 (44) and purified rat MC protease-1 (45), it is possible that the absence of heparin in the NDST-2-null mouse results in a marked increase in the relative lytic activity of the small amount of mMCP-5. Despite their granule storage deficiency, the mMCP-5 gene is transcribed at a normal rate in the MCs of NDST-2-null mice (20). Thus, another possibility is that extensive muscle damage occurs in NDST-2-null mice because the constitutive release of mMCP-5 augments the response of the tissue to other factors active in hind limb IR injury. Because the NDST-2-null mouse also has less MC carboxypeptidase A (20, 36, 37, 46), any role for this exopeptidase acting in concert with mMCP-5 would again depend on the lack of heparin binding.
The fact that this MC-dependent cytotoxicity occurs in a setting of IR with C activation preceding degranulation argues for a critical combination of events, including binding of natural IgM, which leads to activation of C and then MCs. The critical role of mMCP-5 in this setting may be operative via several different mechanisms. The protease might directly activate a pronecrotic pathway or inactivate an antinecrotic pathway outside the muscle cell or inside the cell if it is able to gain entry via the C5b-9 membrane pore. Alternately, mMCP-5 could activate a protease cascade that ultimately results in proteolytic disruption of the basement membrane, sarcolemma, the subsarcolemma cytoskeleton of skeletal muscle, or a factor that regulates the viability of the muscle cell. Whatever its mechanism, our findings demonstrate that irreversible skeletal muscle injury in IR-treated mice is exquisitely dependent on mMCP-5 and independent of several other MC-derived mediators.
]
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by National Institutes of Health Grants GM-52585, HL-36110, and HL-63284.
2 Current address: Cincinnati Children’s Hospital Medical Center, 3333 Burnet Avenue, ML7028, Cincinnati, OH 45229-3039.
3 Address correspondence and reprint requests to Dr. Michael F. Gurish, Brigham and Women’s Hospital, Smith Building, Room 624, 1 Jimmy Fund Way, Boston, MA 02115. E-mail address: mgurish{at}rics.bwh.harvard.edu
4 Abbreviations used in this paper: IR, ischemia-reperfusion; fc, fibers counted; LPF, low-power field; LTC4, leukotriene C4; LTC4S, LTC4 synthase; PGD2S, PGD2 synthase; MC, mast cell; mMCP, mouse MC protease; NDST, N-deacetylase/N-sulfotransferase; mTMT, mouse transmembrane tryptase.
Received for publication May 14, 2004. Accepted for publication March 17, 2005.
References
Weiser, M. R., J. P. Williams, F. D. Moore, Jr, L. Kobzik, M. Ma, H. B. Hechtman, M. C. Carroll. 1996. Reperfusion injury of ischemic skeletal muscle is mediated by natural antibody and complement. J. Exp. Med. 183: 2343-2348.
Kyriakides, C., W. Austen, Jr, Y. Wang, J. Favuzza, L. Kobzik, F. D. Moore, Jr, H. B. Hechtman. 1999. Skeletal muscle reperfusion injury is mediated by neutrophils and the complement membrane attack complex. Am. J. Physiol. 277: C1263-C1268.
Lazarus, B., A. Messina, J. E. Barker, J. V. Hurley, R. Romeo, W. A. Morrison, K. R. Knight. 2000. The role of mast cells in ischaemia-reperfusion injury in murine skeletal muscle. J. Pathol. 191: 443-448.
Mukundan, C., M. F. Gurish, K. F. Austen, H. B. Hechtman, D. S. Friend. 2001. Mast cell mediation of muscle and pulmonary injury following hindlimb ischemia-reperfusion. J. Histochem. Cytochem. 49: 1055-1056.
Bortolotto, S. K., W. A. Morrison, X. Han, A. Messina. 2004. Mast cells play a pivotal role in ischaemia reperfusion injury to skeletal muscles. Lab. Invest. 84: 1103-1111.
Katz, H. R., M. B. Raizman, C. S. Gartner, H. C. Scott, A. C. Benson, K. F. Austen. 1992. Secretory granule mediator release and generation of oxidative metabolites of arachidonic acid via Fc-IgG receptor bridging in mouse mast cells. J. Immunol. 148: 868-871.
Levi-Schaffer, F., E. T. Dayton, K. F. Austen, A. Hein, J. P. Caulfield, P. M. Gravallese, F. T. Liu, R. L. Stevens. 1987. Mouse bone marrow-derived mast cells cocultured with fibroblasts: morphology and stimulation-induced release of histamine, leukotriene B4, leukotriene C4, and prostaglandin D2. J. Immunol. 139: 3431-3441.(J. Pablo Abonia2,*, Danie)