Sarcolemmal Complement Membrane Attack Complex Deposits During Acute Rejection of Myofibers in Nonhuman Primates

Abstract

We have previously studied in nonhuman primates several aspects of the acute rejection of myofibers, including the histological characteristics, the mechanisms of myofiber elimination by the T cells, and the development of anti-donor antibodies. Here, we report the participation of the complement membrane attack complex (MAC) in this context. We used muscle sections of macaques from experiments of allogeneic muscle precursor cell transplantation with confirmed rejection of the graft-derived myofibers. Sections were stained with hematoxylin and eosin, alizarin red and for immunodetection of MAC, CD8, CD4, C3, C4d, and immunoglobulins. The prominent finding was the presence of sarcolemmal MAC (sMAC) deposits in biopsies with ongoing acute rejection or with recent acute rejection. The numbers of sMAC-positive myofibers were variable, being higher when there was an intense lymphocyte infiltration. Few sMAC-positive myofibers were necrotic or had evidence of sarcolemma permeation. The immunodetection of C3, C4d, and immunoglobulins did not provide significant elements. In conclusion, sMAC deposits were related to myofiber rejection. The fact that the vast majority of sMAC-positive myofibers had no signs of necrosis or sarcolemmal permeation suggests that MAC would not be harmful to myofibers by itself.

INTRODUCTION

Transplantation of muscle precursor cells (MPCs) has potential application for the treatment of skeletal muscle diseases (1–3). On one hand, the transplanted cells can be integrated into the recipient’s myofibers, after which these myofibers will contain both native and graft-derived myonuclei (4). Graft-derived myonuclei can thus restore a genetically missing protein whose absence causes a myopathy, such as dystrophin in Duchenne muscular dystrophy (5–8). On the other hand, grafted cells could generate new myofibers, a property useful for regenerative medicine in muscles that have lost myofibers due to different pathological causes (4). Excluding autologous transplantation of non-genetically modified cells (9, 10), the long-term survival of the graft-derived myofibers in both cases depends on an appropriate control of acute rejection (11).

In the course of clinical trials of MPC transplantation in myopathic patients (5–8), we were confronted with a lack of well-defined elements to diagnose myofiber rejection (12). To fill this void, we conducted studies in nonhuman primates (NHPs) (12–14). NHPs have much closer phylogenetic relationship with humans than other mammals (15), and for this reason they are crucial in transplant immunology research (16–18). Unlike mice (19), NHPs share with humans important transplant immune parameters (20–23), and acute rejection is driven by the same immune elements with comparable histological features and cadence (18). In fact, rejection in composite tissue transplantation in humans was characterized based on NHP studies, while results in mice and other non-primate mammals could not be extrapolated to humans (24).

Until now, we have defined the histological features of myofiber rejection in NHPs (12), identified possible mechanisms of myofiber elimination by T lymphocytes (13), and confirmed that there is production of circulating anti-donor antibodies during this rejection (14). Here, we report the participation of the complement’s membrane attack complex (MAC) in this context.

MATERIALS AND METHODS

We analyzed samples of our bank of frozen muscle sections corresponding to macaques in which myofiber rejection after MPC allotransplantation was induced by 2 protocols:

(1) Protocol IW (immunosuppression withdrawal). Monkeys were immunosuppressed with high levels of tacrolimus for 4 weeks and then tacrolimus administration was stopped to trigger rejection of the graft-derived myofibers (which have completed regeneration at that time). To monitor the graft by histology, 4 monkeys received cells labeled with β-galactosidase (β-Gal). To confirm that the findings were due to the allogeneic context and not only to β-Gal expression, 3 monkeys received cells with no genetic manipulation. To monitor the graft in this case, we grafted male-derived cells in females and we detected the Y chromosome by PCR in the cell-grafted muscles. (2) Protocol LI (low immunosuppression, n = 3). Monkeys were immunosuppressed at suboptimal tacrolimus levels from the beginning, and the graft was monitored by histology after transplantation of β-Gal-positive cells.

Animals

We included samples of 10 male and female cynomolgus monkeys (Macaca fascicularis), age range 44–57 months. They are identified here by the acronym of the protocol followed by a number. For surgeries, we used general anesthesia with 1.5%–2% isofluorane in oxygen after induction with ketamine (10 mg/kg) and glycopyrrolate (0.05 mg/kg). Buprenorphine (0.01 mg/kg twice a day) was given 3 days for postoperative analgesia. The procedures were authorized by the Laval University Animal Care Committee.

Cell Transplantation

We used MPCs of our bank of frozen cells obtained from muscle biopsies made in 2 cynomolgus monkeys. One of the cell lines was infected in vitro with a replication-defective retroviral vector LNPOZC7 (gift of Dr Constance Cepko, Harvard University, Boston, MA) encoding a LacZ reporter gene (25). MPCs were cultured in MCDB-120 (26) with 15% FBS (Hyclone, Logan, UT), 10 ng/mL basic fibroblast growth factor (Feldan, St-Laurent, Canada), 0.5 mg/mL bovine serum albumin (Sigma, St-Louis, MO), 1.0 µM dexamethasone (Sigma), and 5 µg/mL human insulin (Sigma). MPCs were transplanted in the biceps brachii, quadriceps and gastrocnemius, using matrices of percutaneous injections (100/cm²), with 100–250 µL precision syringes (Hamilton, Reno, NV) and 27-gauge 0.5-inch disposable needles. The number of cells injected per cm³ of muscle (µ = 24.3 × 10⁶ ± SD 4.5 × 10⁶) was the same in all sites for a given animal. Transplants were performed on sites of ∼1 cm³ in the IW group, and in the whole biceps brachii in the LI group. To identify these sites during biopsies, 2 stitches of polypropylene 4.0 suture (Ethicon, Somerville, NJ) were placed ∼5 mm from both sides of each site. For more details see references (12, 14, 25).

Immunosuppression Protocols

An intramuscular formulation of tacrolimus (a generous gift from Astellas Pharma, Inc., Osaka, Japan) was administered once a day, beginning 5–7 days before transplantation. Blood samples were taken at different intervals to quantify tacrolimus blood levels using an IMx tacrolimus II kit for micro-particle enzyme immunoassay (Abbott, Wiesbaden, Germany). In the IW protocol, we maintained tacrolimus blood levels during 4 weeks over those needed for MPC allotransplantation using this tacrolimus formulation (>50 µg/L). Intramuscular tacrolimus was a suspension of low solubility and worked as a deposit formulation, thus tacrolimus blood levels decreased slowly over several weeks after stopping administration and rejection occurred when tacrolimus levels fell below 10 µg/L (12, 14). In the LI protocol, tacrolimus blood levels ranged between 38 and 18 µg/L during the follow-up (12).

Sampling

IW group: cell-grafted sites were biopsied at 4 weeks post-transplantation and then every 2 weeks until the graft was absent in at least 2 consecutive biopsies. Successive biopsies were taken in different muscles, restarting by the first muscle once the others were completed. LI group: biopsies were taken in the cell-grafted sites at 4, 12, and 20 weeks post-transplantation. Biopsies were mounted in embedding medium and snap frozen in liquid nitrogen. Serial sections of 10–15 µm were made in a cryostat.

Histological Analysis

Sections were stained with hematoxylin and eosin (H&E) for standard histological analysis. For β-Gal detection, sections were fixed in 0.25% glutaraldehyde and incubated 24 hours at room temperature in a solution containing 0.4 mM X-gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside) (Boehringer Mannheim, Vienna, Austria) from a 20-mg/mL stock in dimethylformamide, 1 mM magnesium chloride, 3 mM potassium ferrocyanide, and 3 mM potassium ferricyanide in PBS. Sections were also stained for 5 minutes in a 2% alizarin red (Sigma) solution at pH 5.4, followed by a brief rinse in PBS at pH 5.4.

The following elements were detected by immunohistochemistry. MAC, using a mouse anti-human/monkey C5b-9 monoclonal antibody (mAb) (1:200, clone aE11, Abcam, Cambridge, MA). CD8-positive lymphocytes, using either a mouse anti-human CD8 mAb (1:300, clone RPA-T8, BD Biosciences, Mississauga, Canada) or a rabbit anti-human/sheep CD8 polyclonal antibody (pAb) (1:300, Abcam). CD4-positive lymphocytes, using a mouse anti-human CD4 mAb (1:300, clone L200, BD Biosciences). Other immunohistochemical reactions were done with a biotinylated goat pAb to monkey IgG, IgA, and IgM [H&L]—(1:1200, OriGene Technologies, Rockville, MD), a rabbit pAb to human/monkey C3 (1:800, Abcam) and a mouse mAb to human C4d (1:800, Abcam). For immunohistochemistry, nonspecific binding was blocked by 30-minute incubation with 10% FBS in PBS. Sections were incubated 1 hour with the primary antibody. Unconjugated antibodies were followed by a 30-minute incubation with a biotinylated anti-mouse antibody (1:150, Dako, Copenhagen, Denmark) or a biotinylated anti-rabbit antibody (1:150, Dako). Biotinylated antibodies were followed by a 30-minute incubation with streptavidin-Cy3 (1:700, Sigma). For double immunostaining, an anti-mouse IgG antibody conjugated to Alexa Fluor 488 (1:300; Molecular Probes, Eugene, OR) was also used as second antibody. Antibodies and streptavidin were diluted in PBS, pH 7.4, containing 1% FBS. Incubations were performed at room temperature.

Muscle sections were analyzed by standard microscopy with epifluorescence and bright-field optics. Selected sections with double fluorescent immunostaining were analyzed with an epifluorescent Olympus IX70 microscope (Olympus, Tokyo, Japan) with a FluoView 300 confocal microscopy scanning unit (Olympus) using lasers of 488 nm and 543 nm.

Polymerase Chain Reaction

Total DNA was extracted from sections of muscle biopsies performed in (i) the 3 females grafted with male MPCs non-β-Gal labeled; (ii) a male macaque as positive control; and (iii) a female (non-grafted site) as negative control. Primers that amplify a 1610 bp region of the cynomolgus Y chromosome were used (27): J130, 5′-CGTGTCTTTCCTCATGGCTTC-3′ (forward) and CDY-2r, 5′-CTTTACCATGGATTCGACCC-3′ (reverse). For PCR conditions see references (12, 14). PCR products were loaded on 1% agarose gel, stained with ethidium bromide and scanned with an AlphaImager digital imaging system, avoiding saturation. RAG gene DNA was detected to control the quality of the extracted DNA.

Detection of Anti-Donor Cell Antibodies

In the IW group, blood samples were taken before immunosuppression and concomitantly with the biopsies to extract the serum (14). MPCs from the same batch as transplanted were incubated 1 hour with the sera of the transplanted monkey. Cells were rinsed in PBS and incubated for 15 minutes with 1/40 goat anti-monkey antibody conjugated to FITC (Immunoconjugate, Tilburg, The Netherlands) to be analyzed in a flow cytometer at 488 nm. For more details see reference (14).

Quantitative Analysis

We quantified the sectional area of the biopsy that was β-Gal-positive in monkeys that received β-Gal-labeled cells, as well as the sectional area of the biopsy occupied by lymphocyte accumulations in H&E-stained sections of all monkeys (12). These data were reviewed for this study. To quantify myofibers with sarcolemmal MAC (sMAC) deposits, we counted the number of sMAC-positive myofibers in one of the sections in which they were more abundant, regardless of the intensity of labeling and if sMAC covered partially or completely the myofiber contour. We then measured the surface of grafted muscle to express the result as sMAC-positive myofibers per mm² of muscle section. These quantifications were done using ImageJ software (NIH ImageJ 1.49v, Bethesda, MD) and corroborated in several cases by direct counting at the microscope. ImageJ was also used to quantify the density of the PCR bands corresponding to the Y chromosome in monkeys that received unlabeled cells.

RESULTS

MAC Detection

The prominent finding in this study was the presence of sMAC associated with myofiber rejection. Since C5b-9 is a neoantigen formed only in the quaternary configuration of MAC, the absence of native complement labeling made MAC detection very clear (Fig. 1). The numbers of sMAC-positive myofibers were variable and in some cases strikingly high, as described below.

FIGURE 1.

Different aspects linked to sarcolemmal membrane attack complex (sMAC) detection in cell-grafted muscles during acute rejection of myofibers. (A–C) Serial cross-sections (monkey IW-3, 8 weeks after tacrolimus withdrawal) processed for fluorescent immunodetection of C5b-9 (A) and CD8 (C), and stained with hematoxylin and eosin (H&E) (B). They are shown at low magnification to illustrate the abundance of sMAC-positive myofibers (A), focal lymphocyte accumulations (B, dark basophilic spots), and CD8-positive cells (C, the inset is a magnification of the indicated region). The regions in the rectangles in A and B are respectively enlarged in D and E to better see the sMAC-positive myofibers and the topographic correlation with the lymphocyte accumulations. The intensity of sMAC labeling is variable, and myofibers with more intense sMAC labeling are essentially close or within the lymphocyte accumulations. (F) At higher magnification, the myofibers indicated by arrowheads in D and E illustrate that they are sMAC-positive (D) but have a preserved intermyofibrillar network (F). The “a” arrow in D and E indicates a sMAC-positive myofiber that exhibits the classical aspect of lymphocyte invasion of non-necrotic myofibers (the intermyofibrillar network is preserved, as can be seen in the inset in E) typical of the T lymphocyte attack on myofibers. The “b” arrow in D and E points to a myofiber with lymphocyte invasion that is not clearly sMAC-positive in the serial section. (G, H) Serial cross-sections (monkey IW-3, 10 weeks after tacrolimus withdrawal) stained with alizarin red ([AR] G) and processed for fluorescent immunodetection of C5b-9 (H). The arrows point to 2 AR-positive myofibers that are sMAC-positive in the serial section, while the arrowheads point to AR-positive myofibers that have no sMAC. Some sMAC-positive myofibers are not stained with AR (which illustrate the AR reaction in the majority of the sMAC-positive myofibers [asterisks]). Scale bars: A–C = 1 mm; D, E, G, H = 100 µm; F = 25 µm.

IW Group

The labels used to identify the graft (β-Gal and Y chromosome) were detected in the biopsies during 6–10 weeks after tacrolimus withdrawal, after which they disappeared (Fig. 2, blue and green graphs). In most cases (6/7), there was a drastic reduction of the label (a decrease in µ = 82% ± 9% [SD] with respect to the average of the previous values) prior to its complete disappearance. The disappearance of these labels was associated with (a) focal lymphocyte accumulations (Figs. 1A–E and 2, mauve graphs), essentially composed by CD8-positive (Fig. 1C) and CD4-positive cells (not shown); and (b) the development of de novo circulating anti-donor cell antibodies (Fig. 2, orange bars). This configures a picture of acute rejection of the graft-derived myofibers (12–14).

FIGURE 2.

Correlation of sarcolemmal membrane attack complex (sMAC) with other transplantation parameters in the monkeys of the IW (immunosuppression withdrawal) protocol. Graphs for each monkey (in columns) display the number of sMAC-positive myofibers (red graphs: number of sMAC-positive myofibers per mm² of grafted muscle in the biopsy), graft-specific labels (blue graphs: percentage of the sectional area of the muscle that was β-Gal-positive; green graphs: density of the Y chromosome PCR band in ImageG units), and lymphocyte accumulations (violet graphs: percentage of the grafted area in the biopsy occupied by lymphocyte accumulations). The arrows designated as “sampling” at the top indicate the moment of both muscle biopsies and blood samples for detection of anti-donor cell antibodies (ADCA). X-axes indicate weeks after tacrolimus withdrawal. Orange bars in the top show the period in which de novo circulating ADCA were detected. Both the peak of sMAC-positive myofibers and the peak of lymphocyte infiltration concur with the abrupt decrease and/or the disappearance of the grafted-cell labels. In the red graphs, the periods in which the MAC labeling was slight and more frequently partial are lighter than the periods in which MAC labeling was intense and more frequently complete around the myofiber contour.

Few or no sMAC-positive myofibers (range = 0–2.2 per mm²) were observed in the biopsies that preceded the drastic reduction of the donor label or its disappearance (Fig. 2, red graphs). During this period, MAC labeling was slight and frequently partial around the myofibers. In the subsequent period, the amount of sMAC-positive myofibers increased significantly and then decline with different patterns (Fig. 2, red graphs). Some macaques exhibited an intense peak of sMAC-positive myofibers in one of the biopsies, which in some was preceded by a lesser increase and followed by smaller values (Fig. 2, red graphs). In this period, MAC labeling was generally intense and complete around the myofiber contours (Fig. 1A, D, H). The exception was the final biopsy of monkeys IW-1 and IW-6, in which MAC labeling was slight and partial as in the pre-rejection period. There was no significant difference (p = 0.99 using a t-test) in the maximal value of sMAC-positive myofibers per mm² between monkeys grafted with β-Gal-positive cells (µ = 40.2 ± 14.6 SD) and unlabeled cells (µ = 40.7 ± 46.8 SD), nor qualitative difference in MAC labeling.

The highest values of sMAC-positive myofibers coincided in some cases (4/7) with the highest values of lymphocyte infiltration, while in the rest they were observed in the biopsies performed 2 weeks later (Fig. 2). Myofibers within or close to lymphocyte accumulations had usually the most intense sMAC labeling (Fig. 1D, E). Regarding the de novo circulating anti-donor antibodies, the first significant increase in sMAC-positive myofibers coincided in 2/7 cases with the first detection of antibodies, in 2/7 cases preceded it in 2 weeks, and in the other cases was observed 2 or 4 weeks later (Fig. 2).

LI Group

There were significant numbers of β-Gal-positive myofibers at 4 weeks, followed by a drastic loss of β-Gal in subsequent biopsies (Fig. 3A). Lymphocyte accumulations were observed throughout the follow-up (Fig. 3A). This also configured a pattern of cellular rejection of the graft-derived myofibers (12), although slower than in the IW protocol. Myofibers with sMAC were observed in all biopsies (µ = 9.2 ± 4 SD per mm², range 3.6–17.2) with the same histological characteristics as during rejection in the IW protocol.

FIGURE 3.

Sarcolemmal membrane attack complex (sMAC) in the LI (low immunosuppression) protocol. (A) Panel illustrates the correlation of sMAC with other parameters, similarly and at similar scale as in Figure 2: sMAC-positive myofibers per mm² of grafted muscle (red graphs), as well as the percentages of the sectional area of the cell-grafted muscle that were β-Gal-positive (blue graphs) or occupied by lymphocyte accumulations (violet graphs). (B–D) Serial cross-sections of a cell-grafted muscle region in monkey LI-2, 4 weeks post-transplantation, stained for fluorescent immunodetection of C5b-9 (B), for β-Gal demonstration (C), and with hematoxylin and eosin (H&E) (D). This biopsy shows the correlation between β-Gal-positive myofibers and sMAC. All sMAC-positive myofibers are β-Gal-positive (some marked with asterisks). However, most β-Gal-positive myofibers have no sMAC, as seen in the regions indicated by arrowheads. An accumulation of lymphocytes (arrows in B–D) surrounds β-Gal-positive myofibers, most of which are sMAC-positive. Myofibers close to this lymphocyte accumulation, as in other cases, are those with the most intense sMAC labeling. Scale bars = 100 µm.

Given that biopsies at 4 weeks had significant numbers of β-Gal-positive myofibers and sMAC (Fig. 3A), it was possible to correlate both. Essentially all sMAC-positive myofibers were β-Gal-positive in these biopsies (Fig. 3B, C). The majority of β-Gal-positive myofibers, however, had no sMAC (Fig. 3B, C). As in the IW group, myofibers with more intense sMAC labeling were close to or within lymphocyte accumulations (Fig. 3B, D). In subsequent biopsies there were significant numbers of sMAC-positive myofibers whereas β-Gal-positive myofibers were substantially reduced or disappeared (Fig. 3A).

Searching for Evidence of Myofiber Damage in sMAC-Positive Myofibers

Some myofibers had sarcoplasmic MAC during the lymphocyte infiltration (Fig. 4). This indicates loss of membrane integrity allowing the complement to penetrate the myofiber and form MAC deposits in the sarcoplasm, which is diagnostic of myofiber necrosis (28). Some myofibers with sarcoplasmic MAC had sMAC (Fig. 4A, C) but in others it could not be observed (Fig. 4D, F). The vast majority of sMAC-positive myofibers, however, had no sarcoplasmic MAC or H&E evidence of any phase of necrosis (i.e. hypercontracted dark myofibers, hyaline myofibers, or phagocytosis [29]). On the contrary, they exhibited a preserved intermyofibrillar network (Fig. 1F), an indication that myofibers are non-necrotic (30). We observed sMAC-positive myofibers with lymphocyte invasion but preserved intermyofibrillar network (Fig. 1E), that is, the typical image of lymphocyte invasion of non-necrotic myofibers characteristic of the T-cell attack in muscles (12, 31–33).

FIGURE 4.

Co-immunodetection of C5b-9 (green fluorescence) and CD8 (red fluorescence) in cross sections of a biopsy taken in a cell-grafted site at the peak of lymphocyte infiltration using confocal microscopy. Some myofibers (asterisks) present linear sarcolemmal membrane attack complex (sMAC) deposits covering completely, or partially, the myofiber contour. Two necrotic myofibers (A, D, arrows) are identified by the sarcoplasmic MAC labeling, which indicates damage of the sarcolemma allowing penetration of complement that was activated and formed intracellular MAC deposits. (A) The necrotic myofiber has also sMAC (arrowhead). (D) No sMAC is discernible in the necrotic myofiber. Most of the sMAC-positive myofibers in the images are in contact with CD8-positive lymphocytes. (D–F) The myofiber with sarcoplasmic MAC deposition is not surrounded by CD8-positive lymphocytes, which may be due to the fact that it is already in an advanced stage of necrosis (as the intense intracellular MAC labeling suggests) and was possibly left aside by the CD8-positive lymphocytes. Scale bars = 50 µm.

To observe if non-necrotic sMAC-positive myofibers had at least the sarcolemma permeable to small ions, we performed AR stain, which is used to detect penetration of extracellular calcium into myofibers in human (30) and animal pathology (34). Although AR-positive myofibers were observed in some biopsies (Fig. 1G), few were sMAC-positive in serial sections (Fig. 1H). The vast majority of sMAC-positive myofibers (even those with intense sMAC labeling) showed no AR staining (Fig. 1G, H).

We tried to correlate sMAC with C3, C4d, and immunoglobulins, but their immunodetection failed to provide clear details to draw conclusions. Immunodetection of monkey immunoglobulins produced an intense labeling in the perimysium and endomysium, making it impossible to differentiate an eventual sarcolemmal deposition (even testing different antibody concentrations). C3 and C4d immunolabeling was weak (also even when testing different antibody concentrations) and we were unable to identify deposits. Because obtaining positive controls in cynomolgus tissues for C3 and C4d is not possible, we cannot draw conclusions on their absence.

DISCUSSION

We have previously reported that acute rejection of myofibers in NHPs is characterized by focal accumulations of CD8-positive and CD4-positive lymphocytes in the endomysium, partially or completely surrounding myofibers and frequently invading some of them (12, 13). This study adds a new element to this picture, that is, the presence of MAC deposits in the sarcolemma. As with the cellular (12) and humoral (14) responses, the transplantation of cells without transgenes confirmed that this was due to the allogenic context and not to β-Gal expression. It should be noted that the mAb used to detect MAC is specific to a neoepitope exposed on C9 when incorporated into MAC (35) and, in addition to humans, identifies MAC in non-human primates (36), dogs (34), and pigs (37). Otherwise, the labeling observed in the necrotic myofibers coincides with that reported for C5b-9 in human myopathies (28), and the sarcolemmal labeling coincides with the fact that MAC is typically formed on the surface of targets.

Given that sMAC deposits were so clear, constant and in some cases profuse, we wonder whether sMAC plays a role in myofiber rejection. Complement is a major component of graft failure in organ transplantation primarily by the interaction with endothelial cells, thus producing vascular injury (38, 39). MAC is formed on the surface of targets as a result of any of the 3 activation pathways of complement, that is, classical, alternative, or lectin pathways (40). The cylindrical macromolecular structure of MAC forms trans-membrane channels that allow extracellular water, ions, and molecules to freely penetrate the target, causing its lysis (41, 42). Consequently, our first question was whether sMAC rendered the sarcolemma leaky and predisposed to myofiber necrosis.

In fact, the vast majority of sMAC-positive myofibers did not exhibit evidence of necrosis or sarcolemmal permeation. There were some necrotic myofibers associated with the lymphocyte infiltration (detected by sarcoplasmic MAC deposition), but it is not possible to determine whether this was produced by T lymphocytes (13) or by sMAC. In any case, the numbers of necrotic myofibers were very low compared to the large numbers of non-necrotic sMAC-positive myofibers. This is evident, even though myofiber necrosis is often segmental and some degree of necrosis may be outside the plane of the sections. We then used AR to observe whether sMAC-positive myofibers had intracellular calcium deposits. Since this dye forms a red complex with deposits of calcium salts, AR-positive myofibers imply that the sarcolemma was permeable to small ions such as calcium (30). AR-positive myofibers precede the histological manifestations of necrosis in Duchenne muscular dystrophy (30), and evidence myofiber damage earlier than sarcoplasmic MAC in NHPs (43). Therefore, the fact that very few sMAC-positive myofibers were AR-positive (and that AR-positive myofibers were not always related to sMAC) seems to indicate that sMAC did not produce sarcolemma permeation.

Together these observations suggest that sMAC was not directly harmful to myofibers, that is, did not produce the entry of extracellular medium leading to necrosis (at least in the vast majority of myofibers). This does not rule out the possibility that MAC could fulfill another role in myofiber rejection, since there was recent evidence that the complement has other functions than direct lysis of targets in transplantation, for example, influencing T cell responses (44).

Another intriguing observation is that there was no constant concurrence between sMAC and graft-derived labels. The only condition in which we found sMAC in confirmed graft-derived myofibers was in the biopsies of the LI group at 4 weeks. In subsequent biopsies of the LI group and in the IW group there were significant numbers of sMAC-positive myofibers (including intense peaks) when the histology and PCR indicated that most or all allogeneic cells were already eliminated. Indeed, the same occurred between lymphocyte infiltration and graft-derived labels, as if the immune attack in this context exceeded the times of alloantigen rejection and persisted for some time after the alloantigen elimination.

If we consider 3 factors, sMAC seems to be essentially related to the lymphocyte infiltration: (i) There was a close temporal nexus between sMAC and lymphocyte infiltration in the IW group, with the particularity that sMAC peaks never preceded lymphocyte peaks: either both were simultaneous or sMAC peaks were observed immediately after lymphocyte peaks. The same applies to the first increase in sMAC and the first increase in lymphocyte infiltration. (ii) There was a topographical correlation between lymphocyte infiltrates and sMAC-positive myofibers, and muscle bundles with profuse lymphocyte infiltration were those with profuse sMAC-positive myofibers. (iii) The sMAC labeling was more intense close to and within lymphocyte accumulations. Considering these observations, we postulate that the immune cell reaction induced sMAC deposition in myofibers present in the inflammatory zone regardless of whether they were allogeneic or not. In this sense, experiments with mice suggested that the cellular rejection not only destroys graft-derived myofibers but also autologous myofibers of the recipient due to a “bystander” effect attributed to release of cytokines by immune cells (45).

The steps of the complement cascade that preceded sMAC deposition remain to be determined. The intense immunoglobulin staining in the interstitium made it impossible to determine if there were sarcolemmal IgG or IgM deposits, which are required to initiate the classic complement pathway. However, sMAC appearance did not seem to be related to de novo circulating anti-donor antibodies if we consider that in 2 cases these antibodies were detected later than the appearance of significant numbers and even at the peak of sMAC. The techniques used here did not allow us to determine if there were sarcolemmal C3 or C4d deposits, which would have given some indication of the activation pathways. As noted in human myopathies (46), this could be due to the fact that C3, for example, is rapidly removed by proteolysis whereas MAC is highly insoluble and resistant to proteolysis (42). However, given that we were unable to obtain positive controls for C3 and C4d in cynomolgus tissues, we cannot conclude in their absence.

We previously reported histological resemblances between myofiber rejection in NHPs and human polymyositis (12), a disease also caused by the attack of CD8-positive lymphocytes against myofibers (47). Polymyositis is similarly characterized by accumulations of T lymphocytes surrounding myofibers and T cell invasion of non-necrotic myofibers (31–33). Our study introduces a difference between these conditions because sMAC was not observed in adult polymyositis (46). However, sMAC-positive myofibers were reported in human myopathies such as fascioscapulohumeral muscular dystrophy (46), limb girdle muscular dystrophy (46), laminin-positive congenital muscular dystrophy (46), dysferlinopathies (48, 49), and X-linked vacuolated myopathy (50, 51). This is intriguing because these myopathies have neither an immune origin nor a relevant immune component, but confirms that MAC can be deposed on human myofibers. Interestingly, sMAC-positive myofibers in these myopathies were non-necrotic (46, 48, 49) and there was no evidence of sarcolemmal permeation with AR (46) as in most sMAC-positive myofibers in our study.

In conclusion, the present observations show the participation of MAC during myofiber rejection in macaques, an animal phylogenetically close to humans and with similar transplant immunology, opening the door to studies to elucidate its role in this process. It would also be interesting to see whether sMAC-positive myofibers occur in composite tissue transplants as well as in gene therapy of the skeletal muscle, to the extent that there could be immune reactions against myofibers in which genetic modifications were made.

This work was supported by a grant of the Jesse’s Journey Foundation for Gene and Cell Therapy of Canada to Daniel Skuk and a grant of the Canadian Institutes of Health Research to Jacques P. Tremblay.

The authors have no duality or conflicts of interest to declare.