Identification and function of ryanodine receptor subtype 3 in non-pregnant mouse myometrial cells

Abstract

Subtype 3 of the ryanodine receptor (RYR3) is a ubiquitous Ca²⁺ release channel which is predominantly expressed in smooth muscle tissues and certain regions of the brain. We show by reverse transcription-polymerase chain reaction (RT-PCR) that non-pregnant mouse myometrial cells expressed only RYR3 and therefore could be a good model for studying the role of endogenous RYR3. Expression of RYR3 was confirmed by Western blotting and immunostaining. Confocal Ca²⁺ measurements revealed that in 1.7 mm extracellular Ca²⁺, neither caffeine nor photolysis of caged Ca²⁺ were able to trigger any Ca²⁺ responses, whereas in the same cells oxytocin activated propagated Ca²⁺ waves. However, under conditions of increased sarcoplasmic reticulum (SR) Ca²⁺ loading, brought about by superfusing myometrial cells in 10 mm extracellular Ca²⁺, all the myometrial cells responded to caffeine and photolysis of caged Ca²⁺, indicating that it was possible to activate RYR3. The caffeine-induced Ca²⁺ responses were inhibited by intracellular application of an anti-RYR3-specific antibody. Immunodetection of RYR3 with the same antibody revealed a rather homogeneous distribution of fluorescence in confocal cell sections. In agreement with these observations, spontaneous or triggered Ca²⁺ sparks were not detected. In conclusion, our results suggest that under conditions of increased SR Ca²⁺ loading, endogenous RYR3 may contribute to the Ca²⁺ responses of myometrial cells.

Three genes encoding ryanodine receptors (RYR1, RYR2 and RYR3) have been detected in mammalian tissue (Sorrentino et al. 2000). RYR1 was initially identified in skeletal muscle (Zorzato et al. 1990), RYR2 is primarily associated with cardiac and some smooth muscles (Otsu et al. 1990), and RYR3 is the most widely expressed (Giannini et al. 1992; Sorrentino & Volpe, 1993). Although each isoform may be responsible for activating Ca²⁺ release from internal stores, the contribution of the different RYR isoforms in Ca²⁺ signalling is not completely understood. Using RYR3 knockout mice, it has been reported that RYR3 may contribute with RYR1 to induce Ca²⁺ sparks in neonatal skeletal myocytes (Bertocchini et al. 1997). In addition, overexpression of RYR3 in dyspedic myotubes has been reported to produce Ca²⁺ sparks similar to those induced in frog skeletal myocytes (Ward et al. 2000). However, using an antisense strategy, it appears that in vascular myocytes, both RYR1 and RYR2 are required for Ca²⁺ release during Ca²⁺ sparks and Ca²⁺ waves induced by activation of L-type Ca²⁺ current or by application of caffeine, with no participation from RYR3 (Coussin et al. 2000). Moreover, when both RYR1 and RYR2 are inhibited with antisense oligonucleotides and under conditions of increased sarcoplasmic reticulum (SR) Ca²⁺ loading, RYR3 can be activated by caffeine and localized increases in [Ca²⁺]_i (Mironneau et al. 2001).Since all these studies were performed in cell types expressing several subtypes of RYRs or in conditions of overexpression of RYR3, the physiological role of endogenous RYR3 was not clearly assessed. Previous data have reported that in cultured myometrial cells from pregnant rats and intact strips from pregnant and non-pregnant rats, caffeine is unable to induce Ca²⁺ release from the SR (Arnaudeau et al. 1994; Taggart & Wray, 1998), suggesting that RYR1 and/or RYR2 subtypes are not expressed in these cells. However, analysis of RYR subtypes by RT-PCR has led to conflicting results. In non-pregnant human myometrium, RYR3 seems to be expressed in isolation whereas in pregnant human myometrium, RYR2 and RYR3 have been detected (Awad et al. 1997). Moreover, in pregnant human and rat myometrium, all three RYRs have been reported (Martin et al. 1999a,b), but the effects of caffeine remain controversial. When a full-length cDNA encoding the rabbit uterine RYR3 is expressed in HEK293 cells, these cells exhibit a strong caffeine response (Chen et al. 1997), suggesting that alternative splice variants might be involved in the caffeine sensitivity of RYR3s (Miyatake et al. 1996).

In order to study the functional role of endogenous RYR3, we examined the possibility that myometrial cells from non-pregnant mouse may express only the RYR3 subtype using RT-PCR, Western blotting and immunocytochemistry. We investigated the effects of caffeine, oxytocin and [Ca²⁺]_i jumps induced by flash photolysis of caged Ca²⁺ in isolated myometrial cells under conditions of normal and increased SR Ca²⁺ loading. We show that RYR3 is insensitive to both caffeine and increases in [Ca²⁺]_i under conditions of normal SR Ca²⁺ loading but can become activated by the same agents under conditions of increased SR Ca²⁺ loading.

METHODS

Cell preparation

The investigation conforms with the European Community and French guiding principles in the care and use of animals. Authorization to perform animal experiments was obtained from the French Ministère de l'Agriculture et de la Pêche.

Non-pregnant C57BL/6 mice (20–25 g) were killed by cervical dislocation. The longitudinal muscle layer of myometrium was cut into several pieces and incubated for 10 min in low-Ca²⁺ (40 μm) physiological solution (HBSS), and then 0.8 mg ml⁻¹ collagenase (EC: 3.4.24.3), 0.20 mg ml⁻¹ pronase E (EC: 3.4.24.31) and 1 mg ml⁻¹ bovine serum albumin were added at 37 °C for 20 min. After this time, the solution was removed and pieces of myometrium were incubated again in fresh enzyme solution at 37 °C for 20 min. Tissues were placed in a enzyme-free solution and triturated using a fire-polished Pasteur pipette to release the cells. Cells were seeded at a density of 10³ cells mm⁻² on glass slides. A similar protocol was used to isolate mouse duodenal myocytes, as previously reported (Morel et al. 1997). Myometrial cells were used either freshly isolated or after a short-term primary culture in medium M199 containing 2 % fetal calf serum, 2 mm glutamine, 1 mm pyruvate, 20 units ml⁻¹ penicillin and 20 μg ml⁻¹ streptomycin; they were kept in an incubator gassed with 95 % air and 5 % CO₂ at 37 °C and used within 30 h. Normal physiological solution contained 130 mm NaCl, 5.6 mm KCl, 1 mm MgCl₂, 1.7 mm CaCl₂, 11 mm glucose, and 10 mm Hepes (adjusted to pH 7.4 with NaOH). For experiments that used antibodies, the myometrial cells were held at −50 mV with a standard patch clamp technique using a List EPC7 patch clamp amplifier (Darmstadt, Eberstadt, Germany). The whole-cell recording mode was performed with patch clamp pipettes of 2–5 MΩ resistance. The basic pipette solution contained 130 mm CsCl and 10 mm Hepes (pH 7.3 with CsOH).

Reverse transcription-polymerase chain reaction

Total RNA was extracted from mouse brain and from either freshly isolated smooth muscle cells or cultured cells within 30 h (about 100 cells) using the RNeasy mini kit (Qiagen, Hilden, Germany) following the instructions of the supplier. The reverse transcription (RT) reaction was performed using the Sensiscript RT kit (Qiagen). Total RNA was incubated with random primers (Promega, Lyon, France) at 65 °C for 5 min. After a cooling time of 15 min at 25 °C, RT mix was added and the total mixture was incubated 60 min at 37 °C. The resulting cDNA was stored at −20 °C. PCR was performed with 2 μl cDNA (0.9 μg cDNA for myometrium and duodenum or 0.6 μg cDNA for brain), 1.25 units HotStartTaq DNA polymerase (Qiagen), 1 μm of each primer and 200 μm of each deoxynucleotide triphosphate, in a final volume of 50 μl. The polymerase chain reaction (PCR) conditions were 95 °C for 15 min, then 35 cycles at 94 °C for 1 min, 60 °C (RYR1 and RYR2) or 56 °C (RYR3) for 1 min and 72 °C for 1 min and at the end of PCR, samples were kept at 72 °C for 10 min for final extension and then stored at 4 °C. Reverse transcription and PCR were performed with a thermal cycler (Eppendorf, Le Pecq, France). Amplification products were separated by electrophoresis (2 % agarose gel) and visualized by ethidium bromide staining. The minimum detection of RYR amplification products was obtained with 15 ng cDNA. Gels were photographed with EDAS 120 and analysed with KDS1D 2.0 software (Kodak Digital Science, Paris, France). Sense (s) and antisense (as) primer pairs specific for RYR1, RYR2 and RYR3 were designed on the known cloned receptor sequences as stated in the GenBank sequence database (accession numbers: X83932, X83933, X83934, respectively) with Lasergene software (DNASTAR, Madison, WI, USA). The nucleotide sequence and the length of the expected PCR products (in parentheses) for each primer pair were, respectively:

RYR1(s)	GAAGGTTCTGGACAAACACGGG;
RYR1(as)	TCGCTCTTGTTGTAGAA TTTGCGG (435 bp);
RYR2(s)	GAATCAGTGAGTTACTGGGCATGG;
RYR2(as)	CTGGTCTGAGTTCTCCAAAAGC (635 bp) (Coussin et al. 2000);
RYR3(s)	AGGTGATCAACAAGTATGGA;
RYR3(as)	CAACAGATGAGCAGCAAAGA (RYR3-I: 273 bp and RYR3-II: 614 bp) (Miyatake et al. 1996).

After electrophoresis, the amplified DNA fragments were cleaned and purified with the Qiaquick gel extraction kit (Qiagen). PCR fragments were sequenced by the Qiagen sequencing service. The deduced DNA sequences of RYR1, RYR2, RYR3-I and -II fragments were 98–99 % identical to the published sequences.

Immunoblotting

For Western blotting analysis, microsomal proteins were separated on 5 % SDS-PAGE minigels and transferred to PVDF (polyvinylidene difluoride) membranes for 70 min at 100 V in a transfer buffer containing 192 mm glycine and 25 mm Tris-HCl (pH 8.3). Microsomes were blocked for 3 h in blocking phosphate buffer solution containing 0.1 % Tween-20 and 3 % bovine serum albumin (adjusted to pH 7.4) and then incubated overnight with the primary antibody at 1 : 2500 dilution (rabbit anti-RYR3 antibody or rabbit anti-InsP₃R antibody). After extensive washing, microsomes were incubated for 2 h with the secondary (anti-rabbit) antibody coupled to the peroxydase (1 : 5000 dilution). Specific antigen detection was performed using OPTI-4CN kit (Bio-Rad) to detect peroxydase activity on PVDF membranes and the Kodak EDAS 120.

Cytosolic Ca²⁺ measurements

For experiments that used intracellular application of antibodies or heparin, fluo 3 (60 μm) was dialysed into the cell through the patch clamp pipette. In the other experiments, cells were loaded by incubation in external physiological solution containing 4 μm fluo 3 acetoxymethylester (fluo 3 AM) for 1 h in an incubator at 37 °C. These cells were washed and allowed to cleave the dye to the active fluo 3 compound for at least 30 min. Images were acquired using the line-scan mode of a confocal Bio-Rad MRC1000 (Bio-Rad, Paris, France) connected to a Nikon Diaphot microscope. Excitation light was delivered by a 25 mW argon ion laser (Ion Laser Technology, Salt Lake City, UT, USA) through a Nikon Plan Apo ×60, 1.4 NA objective lens. Fluo 3 was excited at 488 nm and emitted fluorescence was filtered and measured at 540 ± 30 nm. At the setting used to detect fluo 3 fluorescence, the resolution of the microscope was near 0.4 μm × 0.4 μm × 1.5 μm (x-, y- and z-axis). Images were acquired in the line-scan mode at a rate of 6 ms per scan. Scanned lines were plotted vertically and each line was added to the right of the preceding line to form the line-scan image. In these images, time increased from the left to the right, and position along the scanned line was given by vertical displacement. Fluorescence signals are expressed as pixel per pixel fluorescence ratios (F/F₀), where F is the fluorescence during a response and F₀ is the rest-level fluorescence of the same pixel. Image processing and analysis were performed using COMOS, TCSM and MPL software (Bio-Rad).

Caffeine, oxytocin and other stimulating substances were applied by pressure ejection from a glass pipette for the period indicated on the records. All experiments were carried out at 26 ± 1 °C.

Flash photolysis

Caged Ca²⁺, 1-(4,5-dimethoxy-2-nitrophenyl) EDTA, tetra(acetoxymethylester), (DMNP-EDTA AM) at 15 μm was added to the bathing solution and maintained in the presence of cells for 1 h in an incubator at 37 °C. In some experiments, DMNP-EDTA (1 mm, in the presence of 0.25 mm CaCl₂) was introduced into the cell via the patch clamp pipette, with 3–4 min allowed for equilibration. Photolysis was produced by a 1 ms pulse from a xenon flash lamp (Hi-Tech Scientific, Salisbury, UK) focused to a ∼2 mm diameter spot around the cell. Light was band-pass filtered using a UG11 glass between 300 and 350 mm. Flash intensity could be adjusted by varying the capacitor-charging voltage between 0 and 380 V, which corresponded to a change in the energy input into the flash lamp from 0 to 240 J. On flash photolysis, Ca²⁺ was released within 2–4 ms and the small percentage of conversion of the caged compound (∼20 %) allows application of a second pulse without the Ca²⁺ response being altered (Escobar et al. 1995; Arnaudeau et al. 1997).

RYR labelling

Myometrial cells were immunostained as previously described (Boittin et al. 1999). Briefly, myocytes were incubated in the presence of the anti-RYR3-specific antibody (Giannini et al. 1995) (at 1 : 300 dilution) for 20 h at 4 °C, and with the secondary antibody (donkey anti-rabbit IgG conjugated to fluorescein isothiocyanate, diluted at 1 : 500) for 3 h at 20 °C. Thereafter, cells were mounted in Vectashield. Images of the stained cells were obtained with the Bio-Rad confocal microscope. For immunostaining, the resolution of the microscope was near 0.2 μm × 0.2 μm × 0.5 μm (x-, y- and z-axis).

Chemicals and drugs

Collagenase was obtained from Worthington (Freehold, NJ, USA). Fluo 3, fluo 3 AM, DMNP-EDTA and DMNP-EDTA AM were from Molecular Probes (Leiden, The Netherlands). Caffeine was from Merck (Nogent sur Marne, France). Bay K 8644 was from Bayer (Puteaux, France). Medium M199 was from ICN (Costa Mesa, CA, USA). Ryanodine and cyclopiazonic acid were from Calbiochem (Meudon, France). Fetal calf serum was from Bio Media (Boussens, France). Streptomycin, penicillin, glutamine and pyruvate were from Gibco (Cergy Pontoise, France). All primers were synthesized and purchased from Eurogentec (Seraing, Belgium). Heparin (from porcine intestinal mucosa; relative molecular mass 6000) and all other chemicals were from Sigma (St Louis, MO, USA). The polyclonal rabbit anti-RYR3-specific antibody was developed against purified glutathione-S-transferase fusion protein corresponding to the region of low homology between the transmembrane domains 4 and 5 of RYR3 (Giannini et al. 1995). The polyclonal rabbit anti-InsP₃R antibody was directed against the COOH-terminal amino acids (GGVGDVLRKPS) of the rabbit InsP₃R (407143-S, Calbiochem).

Data analysis

Data are expressed as means ± s.e.m.; n represents the number of tested cells. Significance was tested by means of Student's t test. P values < 0.05 were considered significant.

RESULTS

RYR subtypes expressed in non-pregnant mouse myometrial cells

Expression of RYR subtypes was detected in mouse brain and in smooth muscle cells from myometrium and duodenum by RT-PCR. All three RYR subtypes were present in mouse brain (Fig. 1A) and two RYR3 mRNA isoforms were detected: the upper band is compatible with RYR3-II (614 bp) and the lower band is compatible with RYR3-I (273 bp), as proposed by Miyatake et al. (1996). In contrast, only the two RYR3 mRNAs were detected in mouse myometrial cells (Fig. 1B) whereas both RYR2 and RYR3 mRNAs were detected in duodenal myocytes (Fig. 1C). Similar results were obtained from freshly isolated and cultured myometrial cells (n = 6). Figure 1D shows a Western blot analysis on samples of mouse myometrium and brain. The anti-RYR3 antibody recognized a band of ∼550 kDa molecular mass corresponding to the value previously reported for RYR3 (Bertocchini et al. 1997; Sonnleitner et al. 1998), whereas the anti-InsP₃ antibody recognized a band of ∼250 kDa in brain and myometrium (n = 3). No cross-reactivity of the antibodies with the two proteins was detectable. Immunodetection of the RYR3s in cell confocal sections with the specific anti-RYR3 antibody (Giannini et al. 1995) revealed that these receptors were distributed in the whole sections in both freshly isolated(Fig. 2A) and cultured (Fig. 2B) myometrial cells. Compiled data in the absence and presence of the specific anti-RYR3 antibody clearly revealed the specific fluorescence in myometrial cells (Fig. 2C). These results indicate that non-pregnant mouse myometrial cells express only RYR3 and therefore, are a good model for studying the function of endogenous RYR3.

Figure 1. Detection of ryanodine and inositol 1,4,5-trisphosphate receptors (RYRs and InsP₃Rs, respectively) in freshly dissociated mouse non-pregnant myometrial cells.

A-C, cDNA fragments of RYR1 (lane 1), RYR2 (lane 2) and RYR3 (lane 3) were amplified from mouse brain, myometrial and duodenal cells, respectively. The amplified DNA fragments were separated on a 2 % agarose gel and visualized by staining with ethidium bromide. Molecular size standards are indicated in base pairs (bp). Two alternatively spliced RYR3 variants, RYR3-I (273 bp) and RYR3-II (614 bp), were present in brain and smooth muscle cells. Similar results were obtained from six different mice. D, Western blot analysis on microsomes from mouse brain (150 μg protein) and myometrium (Myo; 70 μg protein), separated on 5 % SDS-PAGE with the anti-RYR3 and anti-InsP₃ antibodies. Molecular mass is indicated in kilodaltons. Similar results were obtained from three different mice.

Figure 2. Immunostaining of RYR3 in confocal cell sections of non-pregnant myometrial cells.

A, freshly isolated myocytes stained in the presence or absence of the specific anti-RYR3 antibody. B, staining of a cultured myometrial cell for 30 h. C, compiled data illustrating the mean fluorescence in the absence or presence of the anti-RYR3 antibody with the number of cells tested indicated in parentheses. Cells were obtained from six different mice.

Confocal Ca²⁺ signals evoked by caffeine, oxytocin and flash photolysis of caged Ca²⁺

In myometrial cells, oxytocin acts through production of inositol 1,4,5-trisphosphate (InsP₃) (Marc et al. 1986) and subsequent stimulation of InsP₃Rs to induce Ca²⁺ release from the SR (Arnaudeau et al. 1994). In 1.7 mm Ca²⁺-containing solution (1.7 mm [Ca²⁺]_o), both freshly isolated and cultured myometrial cells responded to application of 0.1–0.25 μm oxytocin by producing propagated Ca²⁺ waves (Fig. 3A). As illustrated in Fig. 4, the peak amplitudes of the oxytocin-induced Ca²⁺ responses, measured from a 2 μm region of the line-scan image, were similar in freshly isolated and cultured myometrial cells. For example, in cultured myometrial cells the peak amplitude of these responses was 2.24 ± 0.06 (Δ(F/F₀), n = 65). In contrast, application of 10–50 mm caffeine was unable to induce any Ca²⁺ responses in the same cells which responded to oxytocin (Fig. 3A and Fig. 4). RYRs can be directly activated by an increase in [Ca²⁺]_i in the vicinity of the receptors. In vascular myocytes, flash photolysis of Ca²⁺-loaded DMNP-EDTA has been shown to instantaneously elevate the [Ca²⁺]_i within 2–4 ms and to activate Ca²⁺ responses (Arnaudeau et al. 1997; Boittin et al. 1998). In 1.7 mm [Ca²⁺]_o-containing solution, flash pulses evoked Ca²⁺ transients in cultured myometrial cells that appeared to be uniform and rapidly declined within 500 ms (Fig. 5A). Application of 10 μm ryanodine for 5 min had no effect on these Ca²⁺ transients suggesting that they corresponded to Ca²⁺ jumps due to release of Ca²⁺ ions from the caged molecule (Fig. 5C).

Figure 3. Typical Ca²⁺ responses evoked by oxytocin and caffeine in mouse non-pregnant myometrial cells under conditions of normal and increased SR Ca²⁺ loading.

A, in 1.7 mm [Ca²⁺]_o, application of 0.25 μm oxytocin induced a propagated Ca²⁺ response whereas in the same cell, application of 10 mm caffeine was unable to induce any Ca²⁺ response. B, in 10 mm [Ca²⁺]_o for 1 h, both oxytocin (0.25 μm) and caffeine (10 mm) induced propagated Ca²⁺ responses in all the cells tested. Traces show (from top to bottom) substance application, line-scan fluorescence image and averaged fluorescence ratio (F/F₀) from a 2 μm region of the line-scan image (indicated by vertical bar). Myocytes were loaded with fluo 3 AM and used within 30 h of primary culture.

Figure 4. Compiled data showing the effects of oxytocin and caffeine in freshly isolated and cultured mouse non-pregnant myometrial cells.

Peak amplitude (Δ(F/F₀)) of the Ca²⁺ responses induced by 0.25 μm oxytocin (open bars) and 10 mm caffeine (filled bars) with the number of cells tested indicated in parentheses. Myocytes were used within 30 h of primary culture. Cells were obtained from six different mice.

Figure 5. Effects of normal and increased SR Ca²⁺ loading on the Ca²⁺ responses evoked by flash photolysis of caged Ca²⁺.

A, in 1.7 mm [Ca²⁺]_o, typical time course of averaged fluorescence ratio (F/F₀) from the entire line-scan image to application of a UV flash pulse of 37 J. B, in 10 mm [Ca²⁺]_o, typical time course of the entire line-scan fluorescence ratio to application of a 37 J flash pulse. C, compiled data showing the effects of ryanodine (10 μm) and heparin (1 mg ml⁻¹) applied for 5 min on the Ca²⁺ responses evoked by flash photolysis of caged Ca²⁺ (37 J flash pulses) in 1.7 mm (open bars) and 10 mm [Ca²⁺]_o (hatched bars). Ratio of peak values in 10 and 1.7 mm [Ca²⁺]_o in control conditions and in the presence of 1 mg ml⁻¹ heparin or 10 μm ryanodine for 5 min (filled bars). The number of cells tested is indicated in parentheses. Asterisks indicate values significantly different from those obtained in control conditions. Except for the experiments with heparin, myocytes were loaded with fluo 3 AM and DMNP-EDTA AM and the external solution contained either 1.7 or 10 mm Ca²⁺. Intracellular application of heparin was performed with the patch-clamp pipette and fluo 3 and DMNP-EDTA were added to the pipette solution. Cells were obtained from three different mice and used within 30 h of primary culture.

When cultured myometrial cells were incubated in 10 mm [Ca²⁺]_o for 1 h, both the amplitude and upstroke velocity of the oxytocin-induced Ca²⁺ responses measured in the same cells increased significantly from 2.27 ± 0.12 (Δ(F/F₀)) to 2.78 ± 0.16 (n = 15) and from 6.69 ± 0.70 (Δ(F/F₀) s⁻¹) to 9.31 ± 0.65 (n = 15), respectively (Fig. 3B and Fig. 6). Longer exposure to 10 mm [Ca²⁺]_o did not significantly change the parameters of oxytocin-induced Ca²⁺ responses (n = 26). These results suggest that the SR Ca²⁺ content of myometrial cells is increased by the sustained elevation in extracellular [Ca²⁺] and may modify both amplitude and upstroke velocity of Ca²⁺ responses, as previously shown in vascular myocytes (Mironneau et al. 2001). To assess the role of the SR Ca²⁺ loading in the generation of large and fast Ca²⁺ responses to oxytocin, the effects of 10 μm cyclopiazonic acid were first investigated on the oxytocin-induced Ca²⁺ responses. Inhibition of the Ca²⁺ uptake capacity of the intracellular store by cyclopiazonic acid resulted in a small elevation of the basal [Ca²⁺]_i and the suppression of the oxytocin-induced Ca²⁺ response in the continuous presence of cyclopiazonic acid for 5 min (n = 6; not shown). In a second set of experiments, oxytocin (0.25 μm) was applied in Ca²⁺-free, 0.5 mm EGTA-containing solution for 20 s (a time sufficient to remove voltage-dependent Ca²⁺ current) in cultured myocytes superfused either in 1.7 mm [Ca²⁺]_o or 10 mm [Ca²⁺]_o. After 20 s in Ca²⁺-free solution, the amplitude of the oxytocin-induced Ca²⁺ responses (Δ(F/F₀)) was 2.05 ± 0.13 (n = 11) in myometrial cells pretreated with 1.7 mm [Ca²⁺]_o and 2.51 ± 0.17 (n = 11) in cells pretreated with 10 mm [Ca²⁺]_o, indicating an increase in Ca²⁺ response amplitude similar to that obtained in 1.7 mm [Ca²⁺]_o. Taken together, these results suggest that the increased accumulation of Ca²⁺ in the SR is responsible for the large and fast oxytocin-induced Ca²⁺ responses under conditions of increased [Ca²⁺]_o.

Figure 6. Compiled data showing the effects of normal and increased SR Ca²⁺ loading on the Ca²⁺ responses evoked by oxytocin and caffeine.

A, peak amplitude (Δ(F/F₀)) of the Ca²⁺ responses induced by 0.25 μm oxytocin (open bars) and 10 mm caffeine (filled bars) in 1.7 and 10 mm [Ca²⁺]_o for 1 h. B, upstroke velocity (Δ(F/F₀) s⁻¹) of the Ca²⁺ responses induced by 0.25 μm oxytocin (open bars) and 10 mm caffeine (filled bars) in 1.7 and 10 mm [Ca²⁺]_o for 1 h with the number of cells tested indicated in parentheses. Asterisks indicate values in 10 mm [Ca²⁺]_o significantly different from those obtained in 1.7 mm [Ca²⁺]_o. Myocytes were loaded with fluo 3 AM and the external solution contained either 1.7 or 10 mm Ca²⁺. Cells were obtained from four different mice and used within 30 h of primary culture.

Under conditions of increased SR Ca²⁺ loading, applications of 10 mm caffeine as well as flash photolysis of caged Ca²⁺ induced Ca²⁺ responses in all the cells tested (Fig. 3B and Fig. 5). The peak amplitude and upstroke velocity of the caffeine-induced Ca²⁺ responses, measured from a 2 μm region of the line-scan image were 1.84 ± 0.25 (Δ(F/F₀), n = 15) and 1.43 ± 0.45 (Δ(F/F₀) s⁻¹, n = 15), respectively (Fig. 6). In Ca²⁺-overloaded myocytes, flash photolysis of caged Ca²⁺ induced a large Ca²⁺ transient followed by a sustained plateau (Fig. 5B). The peak amplitude of the Ca²⁺ responses evoked by 37 J flash pulses was 2.32 ± 0.15 (n = 10; Fig. 5C). Application of 10 μm ryanodine for 5 min on Ca²⁺-overloaded myocytes inhibited the Ca²⁺ response evoked by 37 J flash pulses (Fig. 5C). The remaining Ca²⁺ response in Ca²⁺-overloaded cells pretreated with ryanodine was similar to the Ca²⁺ response obtained in control cells superfused in 1.7 mm [Ca²⁺]_o (Fig. 5C). The ratio of the Ca²⁺ transients (peak value) obtained in 10 and 1.7 mm [Ca²⁺]_o was estimated in control conditions and in cells pretreated with heparin (1 mg ml⁻¹) or ryanodine (10 μm) for 5 min (Fig. 5C). The results showed that Ca²⁺ release induced by flash photolysis of caged Ca²⁺ was unaffected by heparin but blocked by ryanodine in agreement with the exclusive participation of RYRs in the Ca²⁺-induced Ca²⁺ release mechanism.

Spontaneous Ca²⁺ sparks were not detected in Ca²⁺-overloaded myometrial cells (n = 57). Various experimental conditions, such as applications of low concentrations of caffeine, ryanodine or Bay K 8644 (an L-type Ca²⁺ channel agonist), have been reported to trigger and increase the frequency of Ca²⁺ sparks in vascular myocytes (Arnaudeau et al. 1996; Coussin et al. 2000). Applications of 5 nm Bay K 8644 (n = 35), 1 μm ryanodine (n = 39; Fig. 7) or 5 mm caffeine for 1–3 min (data not shown) were ineffective in inducing the generation of Ca²⁺ sparks in non-pregnant mouse myometrial cells isolated from five different mice.

Figure 7. Typical Ca²⁺ responses evoked by low concentrations of Bay K 8644 and ryanodine in mouse non-pregnant myometrial cells under conditions of increased SR Ca²⁺ loading.

In 10 mm [Ca²⁺]_o for 1 h, both 5 nm Bay K 8644 (A) and 1 μm ryanodine (B) induced homogeneous increases in [Ca²⁺]_i without detection of localized Ca²⁺ events. Traces show (from top to bottom) substance application, line-scan fluorescence image and averaged fluorescence ratio from the entire line-scan image. Myocytes were loaded with fluo 3 AM and used within 30 h of primary culture. Similar results were obtained in 35–39 cells from five different mice.

Effects of the anti-RYR3 antibody

In order to confirm that the Ca²⁺ responses induced by caffeine and flash photolysis of caged Ca²⁺ under conditions of increased SR Ca²⁺ loading were due to activation of RYR3, we tested the effects of anti-InsP₃R and anti-RYR3 antibodies which have proved useful in blocking Ca²⁺ responses in vascular myocytes. Concentration-dependent inhibitory effects and specificity of these antibodies has been previously reported in vascular myocytes (Boittin et al. 1999; Mironneau et al. 2001). Under conditions of increased SR Ca²⁺ loading (in 10 mm [Ca²⁺]_o for 1 h), intracellular applications of 10 μg ml⁻¹ anti-InsP₃R antibody for 7 min had no significant effect on the caffeine-induced Ca²⁺ responses (Fig. 8A) but inhibited in a concentration-dependent manner the oxytocin-induced Ca²⁺ responses (Fig. 8B). The specificity of this antibody was also illustrated by the absence of effect of the boiled (95 °C for 30 min) anti-InsP₃R antibody on the oxytocin-induced Ca²⁺ response (Fig. 8B). Intracellular applications of 10 μg ml⁻¹ anti-RYR3 antibody for 7 min slightly reduced the oxytocin-induced Ca²⁺ response but this effect was not significant (Fig. 8B). In contrast, the anti-RYR3 antibody inhibited in a concentration-dependent manner the caffeine-induced Ca²⁺ responses (Fig. 8A). Similarly, the Ca²⁺ responses evoked by flash photolysis of caged Ca²⁺ (estimated as in Fig. 5C, right panel) were unaffected by the intracellular presence of the anti-InsP₃R antibody but blocked by 3 μg ml⁻¹ anti-RYR3 antibody (data not shown). The specificity of the anti-RYR3 antibody was also illustrated by the absence of effect of the boiled (95 °C for 30 min) antibody on the caffeine-induced Ca²⁺ response (Fig. 8A). Taken together, these results show that in Ca²⁺-overloaded myometrial cells the RYR3 is responsible for the Ca²⁺ responses evoked by caffeine and flash photolysis of caged Ca²⁺.

Figure 8. Effects of anti-RYR3 and anti-InsP₃R antibodies on caffeine- and oxytocin-induced Ca²⁺ responses under conditions of increased SR Ca²⁺ loading.

A, effects of 10 μg ml⁻¹ anti-InsP₃R antibody and increasing concentrations of anti-RYR3 antibody (applied intracellularly for 7 min) on the Ca²⁺ responses (Δ(F/F₀)) induced by 10 mm caffeine in cells superfused in 10 mm [Ca²⁺]_o for 1 h. B, effects of 10 μg ml⁻¹ anti-RYR3 antibody and increasing concentrations of anti-InsP₃ antibody (applied intracellularly for 7 min) on the Ca²⁺ responses (Δ(F/F₀)) induced by 0.25 μm oxytocin in cells superfused in 10 mm [Ca²⁺]_o for 1 h. Control: open bars; in the presence of anti-RYR3 antibody: filled bars; in the presence of anti-InsP₃ antibody: hatched bars. The number of cells tested is indicated in parentheses. Asterisks values significantly different from those obtained in control conditions. Cells were obtained from three different mice and used within 30 h of primary culture; they were loaded with fluo 3 and the different antibodies through the patch pipette.

DISCUSSION

In this paper, we report that mouse non-pregnant myometrial cells express only the RYR3 isoform and that activation of this isoform by caffeine and localized increases in [Ca²⁺]_i becomes possible under conditions of increased SR Ca²⁺ loading. The homogeneous distribution of RYR3s in confocal cell sections and the absence of detection of spontaneous and triggered Ca²⁺ sparks suggest that RYR3s do not constitute Ca²⁺ release units in myometrial cells.

Under conditions of normal SR Ca²⁺ loading (when the cells were superfused in 1.7 mm extracellular Ca²⁺), propagated Ca²⁺ waves were triggered by oxytocin whereas caffeine was ineffective in inducing any Ca²⁺ response in mouse non-pregnant myometrial cells. These results are in good agreement with previous data obtained in pregnant rat myometrial cells (Arnaudeau et al. 1994) or in intact myometrial strips from pregnant and non-pregnant rats (Taggart & Wray, 1998) as well as in human myometrial cells in culture (Lynn et al. 1995; Morgan & Gillespie, 1995). Similarly, in vascular myocytes where both RYR1 and RYR2 have been inhibited by an antisense oligonucleotide strategy, the remaining RYR3s were unable to produce Ca²⁺ sparks and waves in myocytes superfused in 1.7 mm [Ca²⁺]_o (Coussin et al. 2000). Our results obtained in non-pregnant mouse myometrial cells revealed that (i) RYR3 was the sole subtype identified by RT-PCR and its expression was confirmed by Western blot analysis and immunolabelling; (ii) endogenous RYR3s expressed in isolation could not be activated by caffeine and [Ca²⁺]_i jumps induced by flash photolysis of caged Ca²⁺ under conditions of normal SR Ca²⁺ loading. Similar results were obtained in freshly isolated myometrial cells as well in cultured cells within 30 h. These results are at variance with data obtained recently from pregnant rat myometrial cells where ∼30 % of the cells respond to caffeine (Martin et al. 1999b). A first possibility may be that a proportion of freshly isolated pregnant rat myometrial cells shows increased SR Ca²⁺ loading, so that RYR3 can be activated by caffeine (as demonstrated in the present study). A second possibility may be that RYR1 or RYR2 which are known to be sensitive to caffeine, could be expressed occasionally in pregnant rat myometrial cells. For example, in pregnant human myometrial cells where both RYR2 and RYR3 are expressed, caffeine is effective in inducing Ca²⁺ responses (Awad et al. 1997). Although the three RYR subtypes have been detected in the whole pregnant rat myometrium (Martin et al. 1999b), it has not been demonstrated that all the RYR subtypes are expressed in myometrial cells. Moreover, identification of the RYR subtypes at the protein level is needed before concluding that these subtypes are involved in Ca²⁺ signalling of rat myometrial cells. Thus, our results indicate that in non-pregnant mouse myometrial cells, Ca²⁺ release from the SR appears to depend exclusively on the opening of InsP₃-gated channels.

Elevation of extracellular [Ca²⁺] to 10 mm created the conditions for an increased SR Ca²⁺ loading, as previously reported in cardiac and vascular myocytes (Gyorke & Gyorke, 1998; Lukyanenko et al. 1999; Mironneau et al. 2001). Inhibition of Ca²⁺ accumulation in the SR by cyclopiazonic acid completely suppressed the oxytocin-induced Ca²⁺ responses, whereas removal of extracellular Ca²⁺ for 20 s had no effect on the enhanced oxytocin-induced Ca²⁺ responses in 10 mm [Ca²⁺]_o, indicating that they were dependent on Ca²⁺ release from the SR. Increase in SR Ca²⁺ loading was illustrated by a significant enhancement in amplitude and upstroke velocity of the oxytocin-induced Ca²⁺ responses. Under these conditions, caffeine and photolysis of caged Ca²⁺ induced Ca²⁺ responses in all cells tested, indicating that RYR3 became activatable. Interestingly, the upstroke velocity of the caffeine-induced Ca²⁺ response in non-pregnant mouse myometrial cells is similar to that measured in vascular myocytes where both RYR1 and RYR2 are inactivated by injection of antisense oligonucleotides (Coussin et al. 2000), suggesting that the properties of RYR3 expressed in vascular and visceral smooth muscle cells may be similar. Immunodetection of RYR3s in mouse myometrial cells revealed a rather homogeneous fluorescence distribution in confocal cell sections suggesting that RYR3s did not form clustered units. This observation is in agreement with the absence of spontaneous and triggered Ca²⁺ sparks in all the cells tested in 10 mm [Ca²⁺]_o, whatever the experimental protocols used (applications of low concentrations of Bay K 8644, ryanodine or caffeine).

The existence of several regulatory sites on the RYR protein complex arises essentially from single channel experiments on RYRs incorporated in lipid bilayers. Several reports have shown that RYR3 can be activated by low cytosolic Ca²⁺ concentration in the presence of 1 mm ATP (Chen et al. 1997; Murayama et al. 1999). These results support the idea that cytosolic Ca²⁺ and ATP may act synergistically to modify the open probability of RYR3 (Manunta et al. 2000). However, in native cells expressing only RYR3, nanomolar increases in [Ca²⁺]_i are unable to activate RYR3. Another important modulator of RYR3 activity is luminal [Ca²⁺]. Increase in luminal [Ca²⁺] has been reported to cause changes in RYR gating (Gyorke & Gyorke, 1998; Lukyanenko et al. 1999). Recently, it has been shown that the open probability of cardiac RYR is strongly increased by raising the luminal [Ca²⁺] from 10 μm to 1 mm (Ching et al. 2000). This effect is removed in the presence of luminal trypsin, suggesting the existence of luminal Ca²⁺ binding sites located either on the RYR or on a closely associated protein which may regulate RYR gating (Ching et al. 2000). Our results confirm that luminal [Ca²⁺] may exert a modulation of endogenous RYR3, leading to an increased channel activity. In addition, the slow gating kinetics of recombinant RYR3 observed at nanomolar [Ca²⁺]_i and in the presence of a physiological concentration of ATP (Manunta et al. 2000) may explain the slow upstroke velocity of caffeine-induced Ca²⁺ release in cells expressing only RYR3s (Coussin et al. 2000 and present study).

Expression of a full-length cDNA encoding the rabbit uterine RYR3 in HEK293 cells leads to a Ca²⁺ release channel sensitive to ryanodine and caffeine (Chen et al. 1997). The spliced variants described by Miyatake et al. (1996) could offer an alternative explanation for various caffeine sensitivity. Since the two RYR3 spliced variants were identified in mouse myometrial cells, which did not respond to caffeine in 1.7 mm [Ca²⁺]_o, and in duodenal myocytes, which responded to caffeine in 1.7 mm [Ca²⁺]_o (Morel et al. 1997), this ruled out the possibility that these spliced variants could be responsible for sensitivity to caffeine. In duodenal myocytes, it is likely that sensitivity to caffeine is related to expression of RYR2.

In conclusion, our data show the expression of RYR3 in non-pregnant mouse myometrial cells and its functional activity only under conditions of increased SR Ca²⁺ loading.