Nov 10, 1997 - The authors thank Dr Angelo Azzi and Dr Polly Matzinger for their helpfull suggestions and comments on the manuscript, and Dr Maria. Marone ...
International Immunology, Vol. 10, No. 3, pp. 267–273
© 1998 Oxford University Press
Myoblasts produce IL-6 in response to inflammatory stimuli Stefania Gallucci, Carlo Provenzano, Paola Mazzarelli, Flavia Scuderi and Emanuela Bartoccioni Institute of General Pathology, Catholic University, Largo Francesco Vito 1, 00168 Rome, Italy Keywords: inflammatory myopathy, muscle cells, polymyositis, protein kinase
Abstract Muscle fibers are the target of T cell-mediated cytotoxic reactions in polymyositis and inclusion body myositis, while the success of myoblast transplantation depends on the absence of an immune rejection against the myofibers. In order to study the behaviour of muscle cells in an inflammatory milieu, we investigated the production of IL-6 and its modulation, including the second messenger pathways controlling it, in in vitro highly purified human myoblast cultures. We found that IL-1β, tumor necrosis factor (TNF)-α and lipopolysaccharide (LPS) stimulated myoblast IL-6 secretion in a dose- and time-dependent manner, whereas forskolin and cholera toxin did not. HA1004 at 10 µM did not significantly affect the IL-1β- and TNF-α-induced IL-6 secretion, suggesting that cAMP and protein kinase A are not sufficient to stimulate this process. To investigate the role of protein kinase C (PKC) in this signal transduction, we employed the inhibitor calphostin C, and the activators phorbol-12-myristate-13-acetate (PMA) and calcium ionophore A23187. Calphostin C blocked IL-6 secretion, PMA had a small stimulatory effect and A23187 had no effect; moreover, PKC down-regulation by PMA did not inhibit IL-1β stimulation, while it reduced TNF-α stimulation. These data indicate that different PKC isoforms may be involved in TNF-α and IL-1β signal transduction. Such a difference can distinguish the action of two traditionally ‘overlapping’ inflammatory cytokines. Our data suggest that muscle cells, like myoblasts, satellite cells and in vivo regenerating myofibers, may discriminate between different stimuli and produce IL-6 when activated in response to muscle injury. Introduction Immune reactions against muscle cells play a central role in the pathogenesis of many muscular diseases such as polymyositis (PM) and inclusion body myositis (1), and can also compromise the success of many strategies for in vivo gene transfer based upon myoblast transplantation (2,3). Pathological specimens obtained from patients affected by these diseases clearly reveal muscle fiber invasion and killing by immune mononucleated cells, and immunohistochemical studies have demonstrated cytokines such as IL-1β and IFN in the inflamed muscle (4,5). In spite of this, very little is known about the sequence of pathogenic events leading to myofiber attack by lymphocytes and the relative contribution given by different cell types to the genesis of the lesions observed in autoimmune myositis. Skeletal muscle fibers are large syncitial cells, in close contact with mononucleated myogenic cells, the so-called satellite cells. Following an injury, the latter transform into activated myoblasts which proliferate and differentiate to restore tissue integrity. In vitro
studies have demonstrated that myoblasts can be stimulated by inflammatory cytokines such as IFN-γ and tumor necrosis factor (TNF)-α to express MHC class II and adhesion molecules, and to function as ‘facultative’ antigen-presenting cells (6). Moreover, we have previously reported MHC-II and intercellular adhesion molecule (ICAM)-1 expression on muscle fibers from PM biopsies but not from normal tissue (7). It is conceivable that, in the presence of proinflammatory stimuli, muscle fibers and myoblasts may actively participate in the triggering and/or amplification of the destructive reactions of autoimmune myopathies. We established an in vitro model to dissect the role of muscle cells in the proinflammatory circuit of PM lesions, in which highly purified human myoblasts were stimulated with proinflammatory agents such as IL-1β, TNFα or lipopolysaccharide (LPS). Our data show that myoblasts can actively respond to an inflammatory environment by selectively expressing and secreting IL-6, a cytokine which influences myofibers by activating cachectic proteolysis (8)
Correspondence to: E. Bartoccioni Transmitting editor: L . Steinman
Received 30 June 1997, accepted 10 November 1997
268 Production of IL-6 by human myoblasts and myoblast proliferation (9), and which is also involved in the promotion of immune responses mainly mediated by B and cytolytic T cells (CTL) (10,11). To date, no information is available on the molecular mechanisms by which IL-1β and TNF-α control the myoblast IL-6 production. No second messengers mediating the effect of IL-1β and TNF-α on IL-6 secretion have been conclusively identified in any cell type. Following receptor binding, IL-1β and TNF-α activate multiple protein kinases (12), increasing protein phosphorylation. Both protein kinase A (PKA), following the activation of adenylate cyclase (13), and protein kinase C (PKC) are supposed to play a role in IL-1β signal transduction. These kinases, through different pathways, could take part in the induction of IL-6 expression, as its promoter contains both cAMP responsive element (CRE) and NF-κB binding sequences (14). We used pharmacological manipulation of IL-6 secretion, and found an intriguing difference in the cellular response to two largely overlapping and pleiotropic stimuli, i.e. IL-1β and TNF-α. In our model these two cytokines use different signaling pathways: IL-1β action appears to be mediated by an atypical PKC isoform, insensitive to Ca21 and phorbol esters, while a phorbol-12-myristate13-acetate (PMA)-sensitive, PKC-dependent signal transduction pathway seems to be involved in TNF-α activation. These observations suggest that muscle cells can finely distinguish and integrate multiple environmental signals too. Methods
Myoblast treatment Myoblasts were seeded at 63103/well in 96-well plates. The following day the myoblast monolayer was incubated at 37°C either in fresh control medium alone, or in the same medium supplemented with cytokines or LPS for 4, 8, 24 or 48 h, alone or in various combinations, to obtain time– and dose– response curves. CTX (0.01–1 µg/ml) and FSK (1–100 µM) were added to the cell cultures for 8 and 24 h; PDBu or PMA with or without A23187 were added for 1 (stimulation) or 24 h (down-regulation) at the concentrations described in Results and were then removed before cytokine stimulation. Myoblasts were pre-incubated with HA1004 (10–100 µM) for 1 h, then the cells were incubated for a further 4–8 h with IL-1β or TNFα. Calphostin C was used for 30 min, then IL-1β or TNF-α were added and cells were incubated for a further 4 h. At the end of each incubation the supernatants were collected, centrifuged and frozen at –20°C until IL-6 determination. The viability of myoblasts was always .95% as determined by the Trypan blue exclusion method. No significant differences of growth rate were found under these different experimental conditions as measured by the MTT staining technique and by [3H]thymidine incorporation (data not shown). IL-6 determination The IL-6 concentration in myoblast culture supernatants was determined by an ELISA kit (Duo-set; Genzyme). The sensitivity of this assay is 31.2 pg/ml. Some of the results were confirmed by a biological assay using the IL-6-dependent B9 hybridoma cell line (kindly provided by L. Aarden).
Cell cultures In vitro human myoblast cultures were established from muscle specimens of subjects undergoing major orthopedic surgery, as previously described (15,16). The resulting myoblasts could be sub-cultured for .30 population doublings. Myoblast preparations were purified by magnetic cell separation using an anti-Leu 19 mAb and goat anti-mouse Ig-coated magnetic beads (Dynal, Oslo, Norway) (15). The purity of each preparation was .98% as assessed by cytofluorimetric analysis (Profile II; Coulter, Miami, FL) with an anti-Leu-19 mAb (Becton Dickinson, San Jose, CA), which cross-reacts with the neural cell adhesion molecule N-CAM (CD56) (17). Myoblast identity was also assayed by the capacity of myoblasts to fuse into multinucleated myotubes and then by immunofluorescence staining with mAb MF20 (kindly provided by G. Cossu), which recognizes all sarcomeric myosins (18). Cytokines and reagents We used human recombinant TNF-α (sp. act. 13108 U/mg; Boehringer Mannheim, Mannheim, Germany), human rIL-1β (sp. act. 13108 U/mg; Genzyme, Cambridge, MA) and LPS (Sigma, St Louis, MS). We used the adenylate cyclase activators forskolin (FSK) and cholera toxin (CTX), and the PKC stimulators PMA, phorbol-12,13-dibutyrate (PDBu) and the calcium ionophore A23187 (Sigma), as well as the kinase inhibitors HA1004 [N-(2-guanidinoethyl)-5-isoquinolinesulfonamide dihydrochloride] (Calbiochem, San Diego, CA) and calphostin C (LC Laboratories, La¨ufelfingen, Switzerland). MTT and DMSO were purchased from Sigma. [3H]Thymidine was from Amersham (Amersham, UK).
RNA extraction and RT-PCR Because of the low number of cells that could be cultured, which did not yield enough material for Northern blotting, we decided to measure specific mRNA levels by semiquantitative RT-PCR. Myoblasts were seeded at 63104/well in 12-well plates. At the end of each incubation cells were lysed in the culture dish by adding the appropriate amount (1 ml/10 cm2) of Trizol Reagent (Gibco/BRL, Grand Island, NY) and passing the cell lysate several times through a pipette. The lysates were then processed according to the manufacturer instructions to obtain total RNA, which was then used for the RT-PCR. In previous experiments we set up the method for multiplex IL-6 and aldolase RT-PCR: total RNA, equivalent to that contained in either 1000 or 2000 cells (the same quantity was used throughout each amplification experiment), is reverse transcribed to cDNA and then amplified in a single tube in a 50 µl volume using the Perkin-Elmer GeneAmp EZ rTth RNA PCR kit, in a Perkin-Elmer 9600 thermal cycler. Reverse transcription was carried out for 45 min at 60°C, while the two-temperature PCR was carried out for 30 sec at 95°C and 45 sec at 62°C. Primers for IL-6 have been chosen so that they span the splicing junction of two exons, to ensure that no amplification of any contaminant genomic DNA can occur, while aldolase mRNA was amplified using the A25 and H20 primers reported in Chelly et al. (19). To obtain a semiquantitative measurement of the RNA, amplification was carried on for only 28–30 cycles, to maintain the product in the linear phase of growth; 10 µl aliquots of the reaction mixture were run on a 3% Metaphor (FMC Products, Rockland,
Production of IL-6 by human myoblasts 269 MA) agarose gel in TAE buffer, containing 1 µg/ml of ethidium bromide. Aldolase was the internal standard to account for differences in the reaction efficiency and gel loading. Band intensity was assessed by scanning and quantification using the Phoretix 1D program (Phoretix International, Newcastle Upon Tyne, UK); results were then normalized according to aldolase intensity and are expressed as arbitrary units (AU). Statistical analysis Statistical evaluation of data was performed using Student’s t-test. Results Constitutive and cytokine-induced production of IL-6 by human myoblasts Myoblasts were purified using an anti-Leu-19 mAb coupled to magnetic beads. More than 98% of the cultured cells expressed markers for muscle cell lineage and maintained the ability to fuse (not shown). Low levels of constitutive IL-6 production were detected in supernatants of myoblasts after as little as 4 h of incubation in fresh culture medium. Exposure to IL-1β (10–2000 pg/ml), TNF-α (10–200 U/ml) or LPS (2,5– 100 ng/ml) increased the amount of IL-6 secreted in the medium in a dose-dependent manner (Fig. 1). As in other cell types (20), IL-1β was the most potent inducer of IL-6 secretion with a 30- to 40-fold increase over unstimulated IL6 production, as compared to a 5- to 7-fold increase induced by LPS and a 3- to 4-fold by TNF-α (note the scale difference in Fig. 1). The biological assay using the IL-6-dependent B9 hybridoma cell line confirmed the ELISA results, demonstrating the production of a functionally active cytokine (data not shown). Kinetic studies showed an increase of IL-6 secretion already after a 4 h exposure to the stimuli and their effect continued over the entire experimental period (48 h) (Fig. 1). IL-1β and TNF-α were able to potentiate each other (Table 1): using a sub-optimal dose of IL-1β, the contemporaneous addition of TNF-α induced an increase of IL-6 secretion to levels higher than the sum of their individual effects (P , 0.001 for IL-1β 10 pg/ml and TNF-α 10 U/ml). No detectable levels of TNF-α or IL-1β proteins have been found in supernatants of myoblasts under basal conditions or after stimulation (data not shown), suggesting that TNF-α and IL-1β do not induce each other’s production. By RT-PCR, we observed an increase in IL-6-specific mRNA in treated myoblasts as compared to untreated cells, while aldolase messengers remained unmodified. Time course experiments showed that IL-1β or TNF-α stimulation on IL-6 mRNA was already detectable after 2 h of exposure and IL-1β-induced IL-6 mRNA continued to accumulate in the following 2 h (Fig. 2). IL-6 mRNA levels were lower after 4 h of TNF-α stimulation than after 2 h, a phenomenon that we observed in every experiment we did, possibly because TNFα induces a smaller and more restricted in time stimulation of IL-6 mRNA. These data suggest that IL-1β and TNF-α affect IL-6 production through an increase of IL-6-specific mRNA: the minor stimulation of mRNA induced by TNF-α can be the basis of the lower levels of IL-6 protein produced by myoblasts.
Fig. 1. Human myoblasts secrete IL-6 in response to IL-1β (a), TNFα (b) and LPS (c). Myoblasts (63103/well) were treated with the indicated stimulators for 4, 24 or 48 h, at the indicated concentrations. IL-6 concentration was determined by ELISA in the supernatants. The data are the media of three independent experiments; bars represent SE.
270 Production of IL-6 by human myoblasts Table 1. IL-1β and TNF-α synergize in inducing IL-6 secretion [human myoblasts (63103/well) were treated with the indicated stimulators for 24 h at the concentrations shown] Interleukins
IL-6 (ng/ml 6 SD)
Control IL-1β 10 pg/ml IL-1β 100 pg/ml TNF-α 10 U/ml TNF-α 100 U/ml IL-1β 10 pg/ml 1 TNF-α 10 U/ml IL-1β 10 pg/ml 1 TNF-α 100 U/ml
0.430 6 0.086 3.011 6 0.468 19.412 6 4.766 1.598 6 0.433 2.315 6 0.190 6.408 6 0.161a 14.160 6 1.086b
aP , 0.001; IL-1β 10 pg/ml and TNF-α 10 U/ml together versus separate stimulation. bP 5 0.036; IL-1β 10 pg/ml and TNF-α 100 U/ml together versus separate stimulation.
Fig. 2. IL-6 mRNA levels as measured by semiquantitative RT-PCR. (a) Stimulation by IL-1β for 2 (C) and 4 (C) h, TNF-α for 2 (C) and 4 (C) h, IL-1β for 2 (C) and 4 (C) h with Calphostin inhibition; unstimulated control is in lane (7). (b) Control is in lane (1); stimulation by TNF-α for 4 h (C), forskolin for 4 h (C), PMA for 1 h followed by a 4 h incubation in basal medium (C); PKC down-regulation by 24 h PMA treatment followed by stimulation by IL-1β for 4 h (C) and TNF-α for 2 h (C); IL-6 and aldolase (ALD) amplification products are indicated.
Activation of adenylate cyclase does not stimulate IL-6 production To study whether a cAMP-dependent pathway is involved in IL-6 production, myoblasts were treated with the adenylate cyclase activators CTX (0.01–1 µg/ml) and FSK (1–100 µM) for 4 (Fig. 2) or 24 (Fig. 3a) h. Neither compound significantly enhanced the basal levels of IL-6 secretion; further, RT-PCR studies show that FSK did not affect IL-6 mRNA levels (Fig. 2), as measured at 4 h after initial stimulus. These data both suggest that activation of the cAMP-dependent signal transduction pathway is not sufficient to induce IL-6 production in myoblasts.
Effect of phorbol esters and calcium ionophore A23187 on IL-6 production To evaluate whether a phorbol ester-dependent pathway could play a role in myoblast IL-6 production, cells were preincubated with the PKC activators PMA (10–1000 nM) and calcium ionophore A23187 (0.01–10 µM), either alone or together for 1 h; the compounds were then washed away and the cells incubated for a further 4 or 8 h in medium alone. After PMA induction, IL-6 secretion was only 4-fold higher than the basal level (Fig. 3b) and did not change even when higher concentrations (1 µM) were used (data not shown). Similar results were obtained with PDBu, a more hydrophilic phorbol ester with similar biological effects, suggesting that the low stimulation was not due to PKC down-regulation because of an incomplete removal of PMA (data not shown). The calcium ionophore A23187 had no effect, either alone or when associated with PMA or PDBu. By RT-PCR, we could not detect any significant variation of IL-6 mRNA levels by stimulation with PMA (Fig. 2), confirming the protein findings. Effects of protein kinase inhibitors on IL-1β- and TNF-αinduced IL-6 production To further evaluate PKA or PKC involvement in IL-6 induction by IL-1β and TNF-α, we inhibited these enzymes with the isoquinoline derivative HA1004, which inhibits PKA at a lower concentration than it does for PKC (Ki 2.3 µM for PKA and 40 µM for PKC). Myoblasts were pre-incubated with HA1004 for 30 min, then IL-1β (50 or 100 pg/ml) or TNF-α (100 U/ml) were added for 4 h: at 10 µM, a concentration which inhibits PKA but not PKC, HA1004 had no significant inhibitory effect on IL-6 secretion, while at 100 µM (a concentration which blocks both PKA and PKC) it caused a 40% inhibition for IL-1β and a 90% inhibition for TNF-α (Fig. 4). These results suggested that PKC, but not PKA, is involved in IL-1β and TNF-α signal transduction leading to IL-6 production. To confirm PKC involvement, we used calphostin C, the most specific PKC inhibitor presently available with a Ki of 50 nM for PKC and ..50 µM for PKA (21). We used a 30 min pre-incubation followed by the addition of IL-1β or TNF-α for 4 h. As shown in Fig. 4, 100 nM calphostin C, a concentration which is only twice its Ki for PKC, decreases the stimulation of IL-6 production by IL-1β and TNF-α by 50%, while almost a 100% inhibition was reached at 400 nM, a concentration which is at least 10 times lower than the Ki for PKA or other protein kinases (.5 µM) (21), further supporting the hypothesis that PKC mediates IL-6 production. By RT-PCR (Fig. 2), we found that no IL-6 mRNA was present after a 2 h IL-1β (100 pg/ml) stimulation of IL-6 production in calphostin C-inhibited cells (400 nM). All protein kinase inhibition experiments were done with short time stimulations (2–4 h) in order to minimize any possible cellular damage by protein kinase inhibitors. No toxic effect of any compound was detected by either Trypan blue exclusion or the MTT staining techniques at the end of each incubation period. To further study the role of phorbol esters- and Ca21insensitive isoforms of PKC in IL-6 production, we depleted the myoblasts of phorbol ester-sensitive PKC isoforms by prolonged treatment with PMA (1 µM) or PDBu (2 µM) for 24 h. After three washes, myoblasts were incubated for 8 h with control medium, with or without IL-1β or TNF-α; to evaluate
Production of IL-6 by human myoblasts 271
Fig. 3. Effect of PKA and PKC stimulators on IL-6 secretion. (a) CTX and FSK were added to myoblasts (63103/well) for 24 h. (b) PMA and/ or calcium ionophore A23187, dissolved in DMSO, were added to the culture medium for 1 h and the supernatants were collected after 8 h. The data are representative of four independent experiments.
Discussion
Fig. 4. Effect of PKA and PKC inhibitory treatments on IL-1β- and TNF-α-induced IL-6 secretion. Cells were incubated with Calphostin (Cph) or HA1004 (HA) for 30 min, then IL-1β or TNF-α were added to the cultures for 4 h. PKC down-regulation (DR) was obtained by PMA treatment (1 µM) for 24 h, then the cells were washed and further incubated for 4 h with IL-1β or TNF-α. Results are expressed as percent of IL-6 secretion induced by IL-1β or TNF-α used at the concentrations shown. The data are representative of five independent experiments.
IL-6 mRNA levels, myoblasts were incubated with IL-1β or TNF-α for 4 and 2 h respectively. PKC down-regulation did not significantly affect IL-1β-stimulated IL-6 production, while it reduced TNF-α-stimulation both at the protein (Fig. 4) and mRNA level (Fig. 2). These results suggest that IL-1β and TNF-α control IL-6 secretion using different signal transduction pathways: a Ca21-independent and phorbol ester-insensitive PKC isoform, perhaps an atypical PKC, is mostly involved in the IL-1β induction, while a PMA-sensitive, PKC-dependent signal transduction pathway seems to play a key role in TNF-α activation of IL-6 secretion in human myoblasts. The IL-1β stimulation of IL-6 secretion after PKC down-regulation by PMA was completely inhibited by 400 nM calphostin C (data not shown), further supporting the hypothesis of the involvement of an atypical PKC in this process.
We have established fibroblast-free pure human myoblast cultures from satellite cells and analyzed them for secretion of the proinflammatory cytokine IL-6. Myoblast cultures were classified as pure on the basis of cytofluorimetric and immunohistochemical analysis: .98% of the cells expressed CD56 and sarcomeric myosin. Moreover, after serum deprivation, they fused to form typical multinucleated myotubes. In this paper we present evidence that human myoblasts secrete IL-6 under basal conditions and in response to several pro-inflammatory agents, including IL-1β, TNF-α and LPS. In our experiments, IL-1β was the most potent inducer of IL-6 secretion, followed in decreasing order by LPS and TNF-α. IL-1β and TNF-α are structurally unrelated cytokines which bind to distinct membrane receptors, yet their spectra of biological effects overlap so that these two cytokines are often considered to be interchangeable. So far, it has been proposed that different pathways transduce their signal from the membrane to the nucleus (22). The IL-6 promoter has response elements to the three major signal transduction pathways, i.e. PKC, cAMP/PKA and calcium (23): there is general agreement that the involvement of the cAMP/PKA or PKC signal transduction pathways vary according to the cell type. We have not been able to stimulate the production of IL-6 in myoblasts using the adenylate cyclase activators CTX and FSK. The stimulatory action of IL-1β and TNF-α was not blocked by concentrations of HA1004 which inhibit PKA without affecting PKC. These data led us to investigate whether PKC controls IL-6 production in our cells. At present, 12 different PKC isoforms have been identified and cloned in mammals (24). They are grouped in three subfamilies: the first group comprises the four classical PKC (cPKC) isoforms (α, βI, βII and γ) which are activated by Ca21 and diacylglycerol (DAG). The second includes five new PKC (nPKC) (δ, ε, η, θ and µ) which are insensitive to Ca21. The third and most recently identified group contains three atypical PKC (aPKC) (ζ, λ and ι) (24) which do not respond to Ca21, DAG or phorbol esters (25). Calphostin C, which interacts with the regulatory domain of PKC, inhibits aPKCζ with an IC50 similar to that for cPKC and nPKC (21,26). Phorbol esters like PMA can activate both
272 Production of IL-6 by human myoblasts cPKC and nPKC, which are down-regulated by a prolonged PMA treatment. The activity of the aPKC is not affected by the same manipulations (27,28). In our model, calphostin C, the most specific PKC inhibitor, can completely block the stimulatory effect of both IL-1β and TNF-α. On the other hand, PMA alone induced not more than a 3-fold increase of IL-6 production and its effect was not influenced by the presence of calcium ionophore A23187, suggesting that the PMAsensitive PKC are poorly involved in this process. PKC downregulation did not affect the IL-1β-induced IL-6 production, while it substantially reduced the TNF-α-induced one. All together, these data suggest that PKC transduces the signal for both IL-1β and TNF-α. The different response of the two cytokines to down-regulation can be explained by their use of distinct PKC isoforms: IL-1β is likely to use an aPKC for signal transduction, while a PMA-sensitive cPKC or nPKC probably transduces the TNF-α signal. PMA and TNF-α give the same level of stimulation, further supporting the hypothesis that TNF-α signals its effect through some PMA-sensitive PKC isoform. Moreover, in our model TNF-α is a poor IL-6 inducer by itself but seems to cooperate with IL-1β in IL-6 induction (Table 1), an effect which warrants further investigation. Occupation of different receptors can induce the activation of distinct PKC isoforms, which at some point activate a common cascade of mediators, explaining the effect we observed. Such a difference in the use of kinase isoforms can distinguish the action of two traditionally ‘overlapping’ inflammatory cytokines. Further investigations could clarify whether a differential PKC involvement in IL-1β and TNF-α signaling is peculiar to myoblasts or represents a more general difference in the signaling properties of these two cytokines . The in vivo counterparts of cultured myoblasts are regenerating cells which, following a myofiber injury, develop from satellite cells, proliferate, fuse and differentiate into new mature myofibers (29). Presently, attempts are made to use allogeneic or autologous myoblasts as carriers for gene therapy because of their property of fusing with mature muscle fibers in vivo, becoming an integral part of the muscle tissue (3). These efforts are hindered by the early rejection of transplanted myoblasts and hybrid myofibers by CTL or NK cells before the fusion process is completed (30). An inflammatory or immune response against myofibers occurs in autoimmune muscle diseases too: in PM CTL and macrophages surround and focally invade non-necrotic muscle fibers (31). It has been demonstrated that muscle-infiltrating lymphocytes are restricted in their TCR Vα/β repertoire in PM patients, while autoinvasive T cells are clonally expanded (32,33); moreover, it has been shown that in inflammatory myopathies there is an intense in situ proliferation and differentiation of infiltrating CD31 lymphocytes (34). All these data support the hypothesis that a local immune reaction underlies the pathological process observed both in myoblast rejection and inflammatory myopathies. In an inflammatory environment, macrophage-produced cytokines such as IL-1β and TNF-α could induce activated satellite cells to secrete other cytokines like IL-6. IL-6, in turn, synergizes with IL-1β and TNF-α in co-stimulating immune responses: as an example, the combination of IL-1β and IL-6 induces naive CD81 cells to proliferate and acquire lytic
capabilities, in the absence of accessory cells and CD41 help (35). Studies performed on IL-6-knockout mice have shown that they are unable to mount a normal inflammatory response to localized tissue damage while they show a normal systemic inflammation to the injection of bacterial LPS (36) and that this phenomenon can be ascribed to a reduced in situ production of chemokines (37), supporting the idea that the IL-6 system plays a pivotal role in local inflammatory reactions by amplifying leukocyte recruitment. The clinicopathological relevance of this mechanism is proved by the effect of systemically administered anti-IL-6 antibodies in experimental autoimmune encephalomyelitis (EAE): treated animals show reduced development of EAE, associated with the absence of inflammatory infiltrates in the central nervous system, further supporting that IL-6 is an essential mediator of the local inflammatory response (38). When dealing with the muscle tissue, however, we have to take into account the direct effects of IL-6 on muscle cells: it can stimulate proteolysis in C2C12 myotubes in vitro (8) and induce myoblast proliferation (9), possibly depending on specific metabolic or culture conditions. Our data, along with literature data, corroborate our hypothesis that in vivo the local secretion of IL-6 by muscle cells may be a central element in the initiation and progression of transplanted myoblast rejection and autoimmune muscle diseases: it can be a CTL proliferation and differentiation agent, an autocrine proteolysis-inducing factor of damaged myotubes, and a proliferation-stimulating agent for satellite cells to replace the destroyed myofibers. Acknowledgements The authors thank Dr Angelo Azzi and Dr Polly Matzinger for their helpfull suggestions and comments on the manuscript, and Dr Maria Marone for her help in band scanning and quantitation. This work was in part supported by Telethon Italy grants 509 to E. B. and 532 to C. P., MURST 40% to E. B. and CNR 96.03732.CT14 to E. B.
Abbreviations CRE CTL CTX DAG EAE FSK ICAM LPS PDBu PKA PKC PM PMA TNF
cAMP responsive element cytotoxic T lymphocyte cholera toxin diacylglycerol experimental autoimmune encephalomyelitis forskolin intercellular adhesion molecule lipopolysaccharide phorbol-12,13-dibutyrate protein kinase A protein kinase C polymyositis phorbol-12-myristate-13-acetate tumor necrosis factor
References 1 Dalakas, M. C. 1991. Polymyositis, dermatomyositis, and inclusion-body myositis. N. Engl. J. Med. 325:1487. 2 Huard, J., Roy, R., Guerette, B., Verreault, S., Tremblay, G. and Tremblay, J. P. 1994. Human myoblast transplantation in immunodeficient and immunosuppressed mice: evidence of rejection. Muscle & Nerve 17:224. 3 Rando, T. and Blau, H. 1994. Primary mouse myoblast purification,
Production of IL-6 by human myoblasts 273
4
5
6
7
8
9
10
11
12
13
14
15
16 17
18
19
characterization, and transplantation for cell-mediated gene therapy. J. Cell. Biol. 125:1275. Isenberg, D. A., Rowe, D., Shearer, M., Novick, D. and Beverly, P. C. L. 1986. Localization of interferons and interleukin 2 in polymyositis and muscular dystrophy. Clin. Exp. Immunol. 63:450. Schubert, W., Kontozis, L., Sticker, G., Schwan, H., Haraldsen, G. and Jerusalem, F. 1988. Immunofluorescent evidence for presence of interleukin-1 in normal and diseased human skeletal muscle. Muscle & Nerve. 8:890. Goebels, N., Michaelis, D., Wekerle, H. and Hohlfeld, R. 1992. Human myoblasts as antigen-presenting cells. J. Immunol. 149:661. Bartoccioni, E., Gallucci, S., Scuderi, F., Ricci, E., Servidei, S., Broccolini, A. and Tonali, P. 1994. MHC class I, MHC class II and intercellular adhesion molecule-1 (ICAM-1) expression in inflammatory myopathies. Clin. Exp. Immunol. 95:166. Ebisui, C., Tsujinaka, T., Morimoto, T., Kan, K., Iijima, S., Yano, M., Kominami, E., Tanaka, K. and Monden, M. 1995. Interleukin6 induces proteolysis by activating intracellular proteases (cathepsins B and L, proteasome) in C2C12 myotubes. Clin. Sci. Colch. 89:431. Austin, L. and Burgess, A. W. 1991. Stimulation of myoblast proliferation in culture by leukaemia inhibitory factor and other cytokines. J. Neurol. Sci. 101:193. Muraguchi, A., Hirano, T., Tang, T., Matsuda, Y., Horii, Y., Nakajima, K. and Kishimoto, T. 1988. The essential role of B cell stimulatory factor 2 (BSF-2/IL-6) for the terminal differentiation of B cells. J. Exp. Med. 167:332. Okada, M., Kitahara, M., Kishimoto, S., Matsuda, T., Hirano, T. and Kishimoto, T. 1988. IL-6/BSF-2 functions as a killer helper factor in the in vitro induction of cytotoxic T cells. J. Immunol. 141:1543. Guy, G. R., Chua, S. P., Wong, N. S., Ng, S. B. and Tan, Y. H. 1991. Interleukin-1 and tumor necrosis factor activate common multiple protein kinases in human fibroblasts. J. Biol. Chem. 266:14343. Zhang Y., Lin, J., Yip, Y. K. and Vilcek, J. 1988. Enhancement of cAMP levels and of protein kinase activity by TNF and IL-1 in human fibroblasts: role in the induction of IL-6. Proc. Natl Acad. Sci. USA 85:6802. Tanabe, O., Akira, S., Kamiya, T., Wong, G. G., Hirano, T. and Kishimoto, T. 1988. Genomic structure of the murine IL-6 gene: high degree of conservation of potential regulatory sequences between mouse and human. J. Immunol. 141:3875. Bartoccioni, E., Michaelis, D. and Hohlfeld, R. 1994. Constitutive and cytokine-induced production of interleukin-6 by human myoblasts. Immunol. Lett . 42:135. Blau, H. M. and Webster, C. 1981. Isolation and characterization of human muscle cells. Proc. Natl Acad. Sci. USA 78:5623. Illa, I., Leon-Monzon, M. and Dalakas, M. C. 1992. Regenerating and denervated human muscle fibers and satellite cells express neural cell adhesion molecule recognized by monoclonal antibodies to natural killer cells. Ann. Neurol. 31:46. Bader, D., Masaki, T. and Fischman, D. A. 1982. Immunochemical analysis of myosin heavy chain during avian myogenesis in vivo and in vitro. J .Cell .Biol. 95:763. Chelly, J., Kaplan, J.-C., Maire, P., Gautron, S. and Kahn, A. 1988. Transcription of dystrophin gene in human muscle and nonmuscle tissues. Nature 333:858.
20 Van Snick, J. 1990. Interleukin-6: an overview. Annu. Rev. Immunol. 8:253. 21 Kobayashi, E., Nakano, H., Morimoto, M. and Tamaoki, T. 1989. Calphostin C (UCN-1028C), a novel microbial compound, is a highly potent and specific inhibitor of protein kinase C. Biochem. Biophys. Res. Commun. 159:548. 22 Dinarello, C. A. 1994. The interleukin-1 family: 10 years of discovery. FASEB J. 8:1314. 23 Sehgal, P. B., Wang, L., Rayanade, R., Pan, H. and Margulies, L. 1995. Interleukin-6-type cytokines. Ann. NY Acad. Sci. 762:1. 24 Dekker, L. V. and Parker, P. J. 1994. Protein kinase C—a question of specificity. Trends Biochem. Sci. 19:73. 25 Hug, H. and Sarre, T. F. 1993. Protein kinase C isoenzymes: divergence in signal transduction? Biochem. J. 291:329. 26 Mizuno, K., Noda, K., Ueda, Y., Hanaki, H., Saido, T., Ikuta, T., Kuroki, T., Tamaoki, T., Hirai, S., Osada, S. and Ohno, S. 1995. UCN-01, an anti-tumor drug, is a selective inhibitor of the conventional PKC subfamily. FEBS Lett. 359:259. 27 Keenan, C., Kelleher, D. and Long, A. 1995. Regulation of nonclassical protein kinase C isoenzymes in a human T cell line. Eur. J. Immunol. 25:13. 28 Gschwendt, M., Leibersperger, H., Kittstein, W. and Marks, F. 1992. Protein kinase Cζ and η in murine epidermis: TPA induces down-regulation of PKCη but not of PKCζ. FEBS Lett. 307:151. 29 Hohlfeld, R. and Engel, A. G. 1994. The immunobiology of muscle. Immunol. Today 15:269. 30 Hohlfeld, R. and Engel, A. G. 1990. Lysis of myotubes by alloreactive cytotoxic T cells and Natural Killer cells. Relevance to myoblast transplantation. J. Clin. Invest. 86:370. 31 Hohlfeld, R. and Engel, A. G. 1991. Coculture with autologous myotubes of cytotoxic T cells isolated from muscle in inflammatory myopathies. Ann. Neurol. 29:498. 32 Mantegazza, R., Andreetta, F., Bernasconi, P., Baggi, F., Oksenberg, J. R., Simoncini, O., Mora, M., Cornelio, F. and Steinman, L. 1993. Analysis of T cell receptor repertoire of muscleinfiltrating T lymphocytes in polymyositis. Restricted V alpha/beta rearrangements may indicate antigen-driven selection. J. Clin. Invest. 91:2880. 33 Bender, A., Ernst, N., Iglesias, A., Dornmair, K., Wekerle, H. and Hohlfeld, R. 1995. T cell receptor repertoire in polymyositis: clonal expansion of autoaggressive CD81 T cells. J. Exp. Med. 181:1863. 34 Lindberg, C., Oldfors, A. and Tarkowski, A. 1995. Local Tcell proliferation and differentiation in inflammatory myopathies. Scand. J. Immunol. 41:421. 35 Renauld, J. C., Vink, A. and Van Snick, J. 1989. Accessory signals in murine cytolityc T cell response: dual requirement for IL-1 and IL-6. J. Immunol. 143:1894. 36 Fattori, E., Cappelletti, M., Costa, P., Sellitto, C., Cantoni, L., Carelli, M., Faggioni, R., Fantuzzi, G., Ghezzi, P. and Poli, V. 1994. Defective inflammatory response in IL-6-deficient mice. J. Exp. Med. 180:1243. 37 Romano, M., Sironi, M., Toniatti, C., Polentarutti, N., Fruscella, P., Ghezzi, P., Faggioni, R., Luini, W., van Hinsbergh, V., Sozzani, S., Bussolino, F., Poli, V., Ciliberto, G. and Mantovani, A. 1997. Role of IL-6 and its soluble receptor in induction of chemochines and leukocyte recruitment. Immunity 6:315. 38 Gijbels, K., Brocke, S., Abrams, J. S. and Steinman, L. 1995. Administration of neutralizing antibodies to interleukin-6 (IL-6) reduces experimental autoimmune encephalomyelitis and is associated with elevated levels of IL-6 bioactivity in central nervous system and circulation. Mol. Med. 1:795.