In recent years, the biological effects of extracellular purine nucleotides acting through P2 receptors have been studied in many cell and tissue types and ATP is now recognized as an important messenger molecule in cell–cell communication (Burnstock, 1997). Thus far, attention has been largely focused on the role of purinergic signaling in mediating changes in short-term cellular activity. However, there is growing evidence that purinergic signaling may also play a role in long-term cellular communication, including cell proliferation, differentiation, and apoptosis (Neary et al., 1996; Abbracchio and Burnstock, 1998). These processes are central to embryonic development, and the specific functions of purine receptor subtypes during skeletal muscle development are the focus of interest in this study.
P2 receptors are classified into two main families, P2X and P2Y, based on molecular structure, transduction mechanisms, and pharmacological properties (Ralevic and Burnstock, 1998). The P2Y receptors are G protein–coupled receptors, which act principally by activating phospholipase C, leading to the formation of inositol 1,4,5-trisphophate and mobilization of intracellular Ca2+. In contrast, P2X receptors are a ligand- gated ion channel family, and activation of these receptors by extracellular ATP elicits a flow of cations (Na+, K+, and Ca2+) across the plasma membrane. To date, seven P2X receptor subunits (P2X1-7) capable of assembling homo- or heteromultimeric receptors (Torres et al., 1999) have been cloned from mammalian species.
The functional importance of these receptor subtypes is yet to be fully understood, particularly with regard to their potential trophic actions. The P2X5 receptor subunit has been shown to be expressed in the proliferating and differentiating cell layers of stratified squamous epithelial tissues (Gröschel-Stewart et al., 1999) suggesting that ATP signaling by means of the P2X5 receptor may play a role in these processes. Conversely, the P2X7 receptor subunit has been strongly linked to apoptosis (Surprenant et al., 1996; Collo et al., 1997; Gröschel-Stewart et al., 1999). There is increasing evidence to suggest that P2X7 activation can induce apoptosis in several cell types and that this process is dependent on the caspase signaling cascade (Coutinho Silva et al., 1999; Ferrari et al., 1999).
Responses to ATP have been demonstrated in embryonic tissues, notably skeletal muscle. By using the patch clamp recording technique, a transmitter-like action of ATP on the cell membranes of myoblasts and myotubes cultured from 12-day chick embryos was first demonstrated by Kolb and Wakelam (1983). These results were confirmed by Hume and Thomas (Hume and Thomas, 1988; Thomas and Hume, 1990a,b, 1993; Thomas et al., 1991), who showed that ATP elicits a potent, P2-dependent depolarizing action on cultured chick myotubes. Similar responses, characteristic of P2 receptors, have been demonstrated in the mouse myoblast cell line, C2C12 (Henning et al., 1993; Henning, 1997). In chick muscle, ATP-responses were also shown to be developmentally regulated (Wells et al., 1995). ATP-elicited contractions were most apparent in early development (embryonic day 6 [E6]) and were absent by embryonic day 17. More recently, immunohistochemistry has been used to describe the developmentally regulated expression pattern of two members of the P2X family, P2X5 and P2X6, in the skeletal muscle of the chick embryo (Meyer et al., 1999).
However, the role of ATP and the identity of the P2X receptor subtypes involved in the development of mammalian skeletal muscle are still largely undefined. In this study, we use immunohistochemistry to investigate the expression of P2X receptors and their relationship to acetylcholine receptors (AChRs) in developing rat skeletal muscle. By using polyclonal antibodies raised against the seven different rat P2X receptor peptides (Oglesby et al., 1999), we demonstrate the timing and localization of receptor expression. In summary, we show the sequential and developmental expression of three receptor subtypes, P2X2, P2X5, and P2X6, perhaps indicating that ATP acting by means of P2X receptors may play a key role in skeletal muscle formation.
P2X2, P2X5, and P2X6 Immunoreactivity in Developing Rat Skeletal Muscle
Staining procedures were carried out for all seven of the P2X receptors at all embryonic stages; however, we found immunoreactivity in developing skeletal muscle only for P2X2, P2X5, and P2X6. Control experiments, performed by preabsorbing the antibodies with the corresponding peptides, showed no immunostaining, confirming the specificity of our findings. The expression of these receptors was followed from E12 to postnatal day (P) 21 in various muscles, including the intercostal, paravertebral, and limb muscles. Results from the muscle groups examined showed that the timing of receptor expression was subtype-specific.
Before E15, there was no immunoreactivity for any of the P2X receptors in the developing muscle. Immunoreactivity for P2X5, but not P2X2 or P2X6, began at E15. Staining with P2X5 was restricted to a small number of cells at the ends of the muscles, close to the myotendinous junctions (Fig. 1A). Immunohistochemistry with an antibody against skeletal myosin on sequential sections from the same embryo confirmed that the staining was confined to the areas of developing muscle (Fig. 1B) described above. This skeletal marker was used throughout this study to identify areas of future muscle. Immunoreactivity for P2X5 strengthened during development; by E16 staining was more widespread within muscles already shown to have positive immunoreactivity for P2X5 and began to appear in muscles that had previously been negative (Fig. 1C,D). This stage also marked the appearance of P2X6 immunoreactivity within skeletal muscle. P2X6 staining was strong and extensive, immunopositive cells could be seen throughout muscle blocks and in all muscle groups tested (Fig. 2A). Immunoreactivity for P2X5 and P2X6 was still apparent at E18 (Figs. 2B,C,D, 3), although in the case of P2X5, it had become restricted to the ends of developing muscle fibers (Fig. 6A). By E20, immunoreactivity for both P2X5 and P2X6 within the skeletal muscle had disappeared. However, it was at these final stages of prenatal development, E18 onward, that P2X2 expression began to appear. Initially P2X2 was expressed most strongly in the intercostal (Fig. 4C,D) and paravertebral muscles (Fig. 4B) and was almost entirely absent in the limb musculature (Fig. 4A). P2X2 expression was most marked in the muscles of the lower limb at E20 (Fig. 5A). Immunoreactivity for P2X2 continued for at least 1 week postnatally (Fig. 5B) but had reduced in strength to near adult expression by P14 (Fig. 5C,D).
In summary, we found three P2X receptors to be expressed in developing skeletal muscle, P2X2, P2X5, and P2X6. These receptors were expressed sequentially, P2X5 being the first receptor to be expressed (E15) followed by P2X6 (E16) and finally P2X2 (E18). This sequence of receptor expression was maintained for all muscles tested. Expression of P2X5 and P2X6 was restricted to in utero development, whereas staining for P2X2 continued after birth. Immunoreactivity for the P2X5 and P2X6 receptor subtypes was most transient, E15-E18 and E16-E18, respectively, whereas P2X2 expression was more sustained, being reduced to near adult expression in the second postnatal week (E18-P14).
Double-Labeling for P2X2 and Acetylcholine Receptors
Double-labeling experiments for P2X2 and AChRs, localized with α-bungarotoxin-Texas Red, were carried out at E18, E20, P7, P14, P21, and adult. As described above, P2X2 was expressed from E18 to P14 (Fig. 5). AChR staining was restricted to large membrane clusters at the center of the muscle fibers, i.e., prospective endplate sites (Fig. 6B–D). Consequently, only a subset of clusters is seen in any one section. At no stage did we observe colocalization of P2X2 and AchR clusters (Fig. 6B–D). This finding included adult muscle where endplates (positive for α-bungarotoxin) as well as extrasynaptic regions were P2X2 negative.
P2X5 Immunoreactivity in Cells Located Between Muscle Fibers
Although P2X5 expression in developing skeletal muscle was restricted to the prenatal period, P2X5-immunopositive cells were apparent within the muscle mass at P7. These cells appeared to be in between developing muscle fibers. This was confirmed by using double labeling with anti-skeletal muscle myosin (anti-PM) and anti-P2X5, which showed a complete absence of overlapping expression (Fig. 7A). By P21, P2X5 staining both within and outside muscle fibers had largely disappeared. To identify the P2X5-positive cells present at P7, further double-labeling experiments were performed by using anti-neurofilament 200 or anti-smooth muscle actin. In both cases, there was limited coexpression with many cells remaining positive only for P2X5 (Fig. 7B-D). This finding would suggest that P2X5 stained a mixed population of cells at P7 including nerves, smooth muscle cells, and probably endothelial cells.
ATP, coreleased with acetylcholine, has been shown to modulate the developing neuromuscular synapses of Xenopus embryos (Fu and Poo, 1991; Fu, 1995) and to have direct transmitter-like actions on developing chick skeletal muscle (Kolb and Wakelam, 1983). In chick, ATP has potent P2-dependent, depolarizing actions on myoblasts in vitro (Hume and Honig, 1986) and causes muscular contraction in vivo (Wells et al., 1995). Furthermore, Wells et al. (1995) demonstrate that sensitivity to ATP is developmentally regulated and returns in denervated skeletal muscle. Experiments on the mouse myoblast cell line, C2C12, showing P2-dependent responses, suggest that ATP may also be an important signaling molecule in mammalian skeletal muscle development (Henning et al., 1993).
However, this work was performed before knowledge of the existence, let alone the properties of the P2X receptor subtypes (Ralevic and Burnstock, 1998). Our findings suggest that expression of the P2X receptors could account for the ATP sensitivity of developing skeletal muscle. By using immunohistochemistry, we show developmentally regulated expression of three P2X receptor subtypes, P2X2, P2X5, and P2X6, in rat skeletal muscle. These receptors are expressed in a sequential manner, which was consistent for all muscles tested: P2X5 was expressed first (E15-E18), followed by P2X6 (E16-E18), and finally P2X2 (E18-P14).
The temporal and spatial overlap in the expression of P2X2, P2X5, and P2X6 receptor subunits, which we observed, raises the possibility that both homomeric and heteromeric receptors may be present in developing skeletal muscle. In sensory neurones of the nodose ganglion, P2X2 and P2X3 subunits are coexpressed (Vulchanova et al., 1997). The characteristics of the P2X receptors present on these cells can only be accounted for by the presence of a P2X2/3 heteromeric receptor (Lewis et al., 1995). Biochemical studies have shown that 11 heteromeric receptors can be formed by pairwise combination of P2X subunits (Torres et al., 1999). However, of these only P2X2/3, P2X1/5, P2X4/6, and P2X2/6 have been demonstrated in functional studies (Lewis et al., 1995; Torres et al., 1998; Lê et al., 1998; King et al., 2000). Further double-labeling studies will be required to determine more closely the extent of P2X receptor coexpression in developing skeletal muscle. This issue is highlighted by our findings at P7, demonstrating the expression of both P2X2 and P2X5, but whereas P2X2 stains skeletal muscle, P2X5 is expressed on a range of other cell types, including smooth muscle and nerve fibers. However, it is worth noting that the study of Torres et al. (1999) suggests that P2X6 does not form a homomeric receptor, suggesting that any functional role will be in coassembly with either P2X2 or P2X5 subunits.
The developmentally regulated expression of the P2X receptors suggests that the channels formed are of functional significance. The timing of receptor expression seems to be closely related to key events in skeletal muscle development (Fig. 8), notably secondary myotube formation and the establishment of mature neuromuscular junctions. We suggest that ATP signaling by means of P2X5 and P2X6 could be involved in the former event, whereas P2X2 could be related to the latter.
Skeletal muscle cells are formed in two stages: primary myotubes develop first, and are followed after a delay by secondary myotubes, which will form the majority of muscle fibers in the adult tissue (Kelly and Zacks, 1969; Harris, 1981; Ontell and Kozeka, 1984; Ross et al., 1987). Secondary myotubes develop from the middle of the muscle in close association with the neuromuscular junctions of guiding primary myotubes (Duxson et al., 1989). They extend by the asynchronous fusion of myoblasts at their ends (Zhang and McLennan, 1995) to finally attach to the muscle tendons (Duxson and Usson, 1989). Immunohistochemical data from our study suggest the involvement of P2X receptors in secondary myotube formation. We detected no immunoreactivity for P2X5 (the first receptor to be expressed) during at least the early stages of primary myotube formation (Ross et al., 1987). However, P2X receptors could be involved in the formation of secondary myotubes, which begins at E16 in the intercostal muscles (Kelly and Zacks, 1969), precisely the same time at which there is strongest immunoreactivity for P2X5 and P2X6.
Staining for P2X2 begins in the intercostal muscles at E18, which is the peak of polyneuronal innervation (Dennis et al., 1981). It is well established that nerve-induced activity plays a crucial role in regulating the expression and distribution of receptors and channels on the muscle membrane, notably the acetylcholine receptor. Whereas developing myotubes express clusters of AChRs throughout the membrane, adult skeletal muscle is characterized by the striking concentration of receptors in the postsynaptic membrane and their virtual absence from the extrasynaptic membrane. This reorganization is in part activity-dependent (Vrbova et al., 1995; Sanes and Lichtman, 1999). In the case of rat intercostal muscles, synaptic inputs are first detected at E15 and produce clustering of AChRs by E16 (Dennis et al., 1981). This finding is confirmed in our results which show acetylcholine clusters in prospective endplate regions. However, on the basis of intracellular recordings in response to acetylcholine application (generally considered to be more sensitive than detection by α-bungarotoxin) the extrajunctional AChR density, particularly in regions close to the myotendinous junction, remains high after birth and only declines during the first postnatal week (Diamond and Miledi, 1962). Changes in the expression of the P2X receptor subtypes, particularly P2X2, mirrors and may even contribute to this process because P2X2 receptor expression is also maintained during the first postnatal week and declines subsequently. In our dual-labeling experiments, extrajunctional AChRs were probably not visible due to the low detection sensitivity of α-bungarotoxin compared with intracellular recording. Interestingly, no expression of P2X2 has been reported in chick skeletal muscle (Meyer et al., 1999). This may reflect differences in the development of mammalian as opposed to avian neuromuscular junctions.
The surprising abundance of these receptors raises the issue of what the potential sources of ATP in this developing system might be. It is well established that ATP is stored within the synaptic vesicles of presynaptic nerve terminals and is coreleased with acetylcholine (Silinsky and Hubbard, 1973). Muscles are also known to secrete substantial amounts of ATP in response to electrical activity (Landmesser and Morris, 1975). However, there are other important nerve-independent sources of extracellular ATP. Osteoblasts and chondrocytes, which are obviously closely associated with developing skeletal muscle, have been shown to be capable of releasing ATP (Bowler et al., 1998; Lloyd et al., 1999) and could be associated with the marked expression of P2X5 at myotendinous junctions.
In conclusion, we show for the first time that three members of the P2X family, P2X2, P2X5, and P2X6, are expressed in developing rat skeletal muscle. These receptors are expressed sequentially (P2X5, followed by P2X6, and finally P2X2) and appeared to be developmentally regulated. The timing of receptor expression is closely related to key events in skeletal muscle development, notably the generation of secondary myotubes and the redistribution of acetylcholine receptors, suggesting that ATP may have trophic actions on developing mammalian skeletal muscle. However, the precise roles of the P2X receptor subtypes in skeletal muscle development remain to be investigated.
Mating of Rats and Embryo Preparation
For dated pregnancies, female Sprague-Dawley rats in estrus were placed overnight with primed male rats and examined in the morning for the presence of a vaginal plug. The day of finding the plug was designated embryonic day 0 (E0) and the day of birth, postnatal day 0 (P0). Pregnant rats, killed by a rising concentration of CO2 and confirmed by cervical dislocation, were used to obtain embryos at E12, E14, E15, E16, E18, and E20. Once removed, some embryos were placed decapitated and intact onto a cork block, covered with OCT compound, and frozen in liquid nitrogen-cooled isopentane. Limb buds were dissected from the remaining embryos, mounted on cork blocks, and frozen separately in liquid nitrogen-cooled isopentane. Skeletal muscle samples were also taken from rat pups killed at P7, P14, and P21. Segments of rib cage and lower limb were removed and prepared as previously described. Cryostat sections were cut at 12 μm and collected on gelatinized slides.
Longitudinal sections of tibialis anterior muscle were prepared from animals at P7, P14, P21, and adult. Muscles were stretched and pinned on Sylgard, fixed in 4% paraformaldehyde for 2 hr at room temperature, and transferred to 20% sucrose in phosphate-buffered saline (PBS) and left overnight. Tissue was placed longitudinally and embedded in OCT on cork blocks. Cryostat sections were cut at 50 μm and collected in PBS.
The immunogens used for the production of polyclonal antibodies were synthetic peptides corresponding to the carboxy termini of the cloned rat P2X receptors, covalently linked to keyhole limpet hemocyanin. The peptide sequences are as follows. P2X1: amino acids 385–399, ATSSTLGLQENMRTS; P2X2: amino acids 458–472, QQDSTSTDPKGLAQL; P2X3: amino acids 383–397, VEKQSTDSGAYSIGH; P2X4: amino acids 374–388, YVEDYEQGLSGEMNQ; P2X5: amino acids 437–451, RENAIVNVKQSQILH; P2X6: amino acids 357–371, EAGFYWRTKYEEARA; P2X7: amino acids 555–569, TWRFVSQDMADFAIL.
The polyclonal antibodies were raised by multiple monthly injections of New Zealand rabbits with the peptides (performed by Research Genetics, Huntsville, AL). The specificity of the antisera was verified by immunoblotting with membrane preparations from CHO-K1 cells expressing the cloned P2X1–7 receptors. The antibodies recognized only one protein of the expected size in the heterologous expression systems and were shown to be receptor subtype-specific (Oglesby et al., 1999). For immunostaining of cryostat sections, the avidin-biotin (ABC) technique was used according to the protocol developed by Llewellyn-Smith et al. (1992, 1993). Air-dried serial sections of the tissues were fixed in 4% formaldehyde and 0.2% of a saturated solution of picric acid in 0.1 M phosphate buffer (pH 7.4) for 2 min. After washing in PBS for 15 min, endogenous peroxidase activity was blocked by treating the sections with 0.4% H2O2 and 50% methanol for 10 min. Nonspecific binding sites were blocked by a 20-min preincubation in 10% normal horse serum (NHS) in PBS. Sections were incubated overnight at room temperature in 5 μg/ml of P2X1–7 antibody or 2 μg/ml anti-skeletal myosin (Sigma, Poole, UK), in 10% NHS in PBS + 2.5% NaCl at room temperature. Subsequently, sections were incubated with biotinylated donkey anti-rabbit IgG (Jackson ImmunoResearch Lab, West Grove, PA) diluted 1:500 in 1% NHS in PBS for 1 hr, and then with ExtrAvidin peroxidase (Sigma) diluted 1:1500 in PBS for 1 hr at room temperature. For color reaction, a solution containing 0.05% 3,3′-diaminobenzidine 0.04% nickel ammonium sulphate, 0.2% β-D-glucose, 0.004% ammonium nitrate, and 1.2U/ml glucose oxidase in 0.1 M PBS was applied for 8 min. Sections were washed three times with PBS after each of the above steps except after the preincubation. Control experiments were carried out with the primary antibody omitted from the staining procedure and the primary antibody preabsorbed with the peptides used to immunize the rabbits, according to the protocol described by Meyer et al. (1999).
For immunofluorescent staining, air-dried serial sections of tissue were fixed in 4% formaldehyde in 0.1 M phosphate buffer (pH 7.4) for 2 min. After washing in PBS for 15 min, sections were treated as before with 10% NHS in PBS to block nonspecific binding. Sections were then incubated overnight at room temperature in 5 μg/ml of P2X5 or P2X2 antibody in 10% NHS in PBS + 2.5% NaCl. P2X expression was visualized by using Oregon-green–labeled goat anti-rabbit IgG (Jackson ImmunoResearch Lab) applied for 1 hr at 1:100. Double staining for acetylcholine receptors was performed by further incubation of sections with α-bungarotoxin-Texas Red at 1:1,000 for 1 hr at room temperature. For neurofilament 200 or smooth muscle actin, double staining was performed by incubation of sections with either mouse anti-neurofilament 200 IgG (Sigma) or mouse anti-smooth muscle actin (Sigma) at 1:100 and 1:1,000, respectively, overnight. Staining was visualized by incubating with TRITC-labeled goat anti-mouse antibody (Jackson ImmunoResearch Lab) at 1:100 for 1 hr.
In the case of double staining for P2X5 and skeletal myosin, sections were incubated overnight at room temperature in 1 μg/ml of P2X5 antibody in 10% NHS in PBS + 2.5% NaCl. This step was followed by incubation with biotinylated donkey anti-rabbit IgG (Jackson ImmunoResearch Lab) diluted 1:500 in 1% NHS in PBS for 1 hr, ExtrAvidin peroxidase diluted 1:1500 in PBS for 1 hr, tyramide amplification for 8 min (Tyramide Amplification Kit, NEN Life Science Products, Boston, MA), and finally streptavidin-fluorescein (Amersham, UK) at 1:200 for 30 min. Sections were washed three times with PBS + Tween (0.05%) after each of the above steps. Staining for skeletal muscle myosin was performed by further incubation of sections overnight in 2 μg/ml rabbit anti-PM in 10% NHS in PBS + 2.5% NaCl, followed by visualization with goat anti-rabbit cy3 (Jackson ImmunoResearch Lab), 1:200 in 1% NHS in PBS for 1 hr. For the longitudinal 50-μm sections, photographs were generated by using confocal microscopy.
The authors thank Tim Arnett, Michelle Bardini, and Ute Gröschel-Stewart for help and technical assistance and Roche Bioscience, Palo Alto, California, for their support. Mina Ryten is funded by University College Medical School and Astrid Hoebertz by the Arthritis Research Campaign.