INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

TWO OF THE FIRST EVENTS of fertilization are the generation of an ion current across the oocyte plasma membrane (20) and a transient increase in intracellular Ca²⁺ (9, 29). In many species including echinoderms, amphibians, and mammals, the fertilization current is triggered by the release of intracellular Ca²⁺ (7, 16, 21). A notable exception, however, is the ascidian, where the fertilization current is Ca²⁺ independent (4, 5, 26). The ion channel responsible for the ascidian fertilization current is a large, nonspecific ion channel with a reversal potential of about +20 mV (5). Although the physical properties of this channel are well characterized, the physiological trigger of the channel is not known (4, 26).

One of the early events stimulated at fertilization is the metabolism of nicotinamide nucleotides (10, 28). One such metabolite, cyclic ADP-ribose (cADPR), has been shown to behave as a potent Ca²⁺-mobilizing enzyme in some systems (12, 18). A second metabolite, ADP-ribose, is much less well known as a second messenger and is thought to be involved uniquely in signaling through nonenzymatic ADP ribosylation (25). In this manuscript, we show that the plasma membrane of Ciona intestinalis oocytes contains an ion channel that is gated by ADP-ribose. Furthermore, our data suggest that this channel is the previously characterized "fertilization channel" (5). An inhibitor of nicotinamide nucleotide breakdown blocks both currents induced by nicotinamide nucleotide injection and the fertilization current, suggesting that ascidian sperm induce the fertilization current by stimulating the breakdown of nicotinamide nucleotides to ADP-ribose.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Collection and preparation of oocytes. Oocytes were dissected from the ascidian C. intestinalis collected from the Bay of Naples and kept in tanks with running seawater until use. Oocytes were manually dechorionated using steel needles and placed in an injection chamber containing 2 ml of natural filtered seawater from the Bay of Naples. Fragments were prepared by cutting ascidian oocytes with steel needles on extrusion through the chorion. The sizes of fragments were measured, and fragments with a diameter of 20 µm were used for the experiments. Ca²⁺-free seawater (0 Ca²⁺) contained (in mM) 500 NaCl, 10 KCl, 50 MgSO₄, 2.5 NaHCO₃, and 10 EGTA, pH 8.0. Constituents of low-Na⁺ seawater (0 Na⁺) were (in mM) 5 NaCl, 495 choline chloride, 10 KCl, 10 CaCl₂, 25 MgSO₄, 25 MgCl₂, 2.5 NaHCO₃, and 10 HEPES, pH 8.0. High-Ca²⁺ seawater contained (in mM) 370 NaCl, 27 MgCl₂, 28 MgSO₄, 2.5 NaHCO₃, 100 CaCl₂, 10 KCl, and 1 EDTA, pH 8.0.

Microinjection and electrophysiological techniques. Standard patch pipettes of 2 µm diameter and 10 M resistance were used for both microinjection and electrophysiology. Pipettes were backfilled with reagents dissolved in an intracellular solution (ICS) containing (in mM) 200 K₂SO₄, 20 NaCl, and 10 HEPES, pH 7.5, unless otherwise stated. After formation of a gigaohm seal, the patch was ruptured and reagents injected by pressure using an Eppendorf Transjector 5246 (the pressure injection system is required to introduce reagents into large cells such as ascidian oocytes). Injection volumes were estimated by estimating the size of the pulse in the oocyte, measured by the displacement of cytoplasm after an injection at a controlled pressure. Control injections of up to 10% of oocyte volume of ICS did not affect oocytes. Membrane potentials were held at 80 mV for all experiments except where stated. Currents were recorded with a List L/M-EPC7 patch-clamp amplifier in the whole cell voltage-clamp configuration and stored on a microcomputer with a Bio-Rad CRS-400 electrophysiology/ion measurement system. cADPR, ADP-ribose, cyclic aristeromycin diphosphate ribose, and 8-NH₂-cADPR analogs were supplied by Dr. Anthony Galione. Heat-inactivated cADPR was produced by boiling cADPR for 45 min. This compound had no activity as a Ca²⁺-mobilizing agent in sea urchin homogenates, confirming the inactivation of cADPR to ADP-ribose (data not shown). All other reagents were obtained from Sigma except where stated.

For single-channel recording, cADPR (5 µM pipette concentration) and ADP-ribose (10 nM pipette concentration) were introduced in a continuous flow into an oocyte through an electrode in the whole cell configuration. The same electrode was used to clamp the membrane potential. A second electrode containing ICS was used in the cell-attached patch configuration. The single-channel electrode was held at 0 mV. Excised patch recordings in the outside-out configuration were prepared by clamping an unfertilized oocyte in whole cell configuration with a pipette containing 10 nM ADP-ribose and then gently pulling the pipette off the oocyte membrane. Currents were recorded with a List L/M-EPC7 patch-clamp amplifier and stored on videotape for subsequent analysis.

Estimates of channel density were made by dividing the saturating current obtained after injection of ADP-ribose into ascidian oocyte fragments by the peak single-channel currents. The data were obtained using a clamped membrane potential of 80 mV.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

An inward current is triggered by nicotinamide nucleotides in ascidian oocytes. Pressure injection of nicotinamide nucleotides and nicotinamide nucleotide metabolites into C. intestinalis oocytes through a micropipette in the whole cell voltage-clamp configuration generated inward currents that varied in amplitude in a dose-dependent manner (Fig. 1A). The most potent compound was ADP-ribose, which induced an inward current at intracellular concentrations as low as 10 nM. Microinjection of ADP did not induce plasma membrane currents, even at a concentration of 100 µM in the oocyte (Table 1), suggesting that the inward currents observed after ADP-ribose injection are due to an effect of this molecule and not due to its metabolism to ADP. cADPR is hydrolyzed to ADP-ribose by cADPR hydrolase (reviewed in Refs. 12 and 18). Cyclic aristeromycin disphosphate ribose, a poorly hydrolyzable analog of cADPR (1), did not trigger inward currents (Table 1). In contrast, heat inactivation of cADPR (which produces ADP-ribose) caused a dramatic increase in peak current after injection of an equivalent concentration of cADPR (see Fig. 1A and Table 1). These data suggest that cADPR is hydrolyzed to ADP-ribose before the current is induced and therefore imply that ADP-ribose is the unique trigger that gates the current.

View larger version (20K): [in this window] [in a new window]   View larger version (10K): [in this window] [in a new window]   View larger version (15K): [in this window] [in a new window]   View larger version (18K): [in this window] [in a new window]   Fig. 1.   A: plasma membrane currents gated by nicotinamide nucleotide and their metabolites. Top: raw data from a typical experiment. Arrow, time of injection. Peak current is recorded as maximum inward current triggered by microinjection of reagent. Bottom: dose-response curve for microinjection of nicotinamide nucleotide indicated. Symbols represent mean with SE as bars. L, ligand concentration; I, current; ADPR, ADP-ribose; HI-cADPR, heat-inactivated cyclic ADP-ribose. See Table 1 for statistics. B: timing of development of plasma membrane ion currents after injection of labeled reagents. Arrow, time of injection. See Table 1 for statistics. C: nicotinamide nucleotides and ADP-ribose gate currents through same plasma membrane mechanism. ADP-ribose was injected into 20-µm-diameter fragments of ascidian oocytes to give a saturating current. Current does not increase when ADP-ribose is coinjected with any of the 4 nicotinamide nucleotides. ADP-ribose was injected to 1 µM (n = 13). ADP-ribose (1 µM) was then coinjected with 500 µM NAD+ (n = 6), 50 µM NADH (n = 7), 100 µM NADP+ (n = 5), or 200 µM NADPH (n = 5). Data are shown as mean with error bars representing SE. D: nicotinamide inhibits membrane current induced by microinjection of nicotinamide nucleotides. Solid bars, controls (nicotinamide absent). Open bars, oocytes bathed in 20 mM nicotinamide. Bars represent mean, with error bars representing SE. Levels of significance according to Student's t-test: * 95%, ** 99%.

Microinjection of nicotinamide nucleotides triggered inward currents (Fig. 1A and Table 1), but with a 0.2- to 0.5-s latency (Fig. 1B and Table 1). The latency for the currents triggered by nicotinamide nucleotide microinjection suggests that enzymatic modification of these nucleotides takes place before the current is gated. We tested whether nicotinamide nucleotide precursors gated inward currents through the same mechanism as ADP-ribose by coinjecting a saturating concentration of ADP-ribose and a nicotinamide nucleotide into 20-µm diameter ascidian oocyte fragments (fragments are required to determine a saturating concentration of ADP-ribose without saturating the patch-clamp amplifier). The peak current did not increase in size (Fig. 1C), suggesting that nicotinamide nucleotide breakdown forms ADP-ribose. The breakdown of NAD⁺ to cADPR by ADP-ribosyl cyclase and to ADP-ribose by NADase can be blocked through metabolic end-product inhibition by the addition of nicotinamide (2, 24). In our system, 20 mM nicotinamide significantly inhibited the currents triggered by microinjection of nicotinamide nucleotides, without significantly affecting the currents triggered by ADP-ribose or cADPR microinjection (Fig. 1D and Table 1).

The ADP-ribose channel is a large, nonspecific ion channel. The peak inward current generated by injection of ADP-ribose was attenuated but not totally abolished when oocytes were bathed in Ca²⁺-free seawater (Fig. 2). In contrast, the peak current increased in amplitude in high-Ca²⁺ seawater (Fig. 2). Replacement of external Na⁺ by choline also reduced the peak ADP-ribose current (Fig. 2). These experiments suggest that the ADP-ribose channel is permeable to both Ca²⁺ and Na⁺. Both whole cell and single-channel current-voltage curves for currents induced by ADP-ribose or cADPR show a reversal potential of between +15 and +20 mV (Fig. 3, A and B; Table 2), again suggesting that the channel is nonspecific. The single-channel data give an estimate of unitary conductance of 140 pS for both cADPR and ADP-ribose (Fig. 3C; Table 2). From these data, and the data in Fig. 1C, we estimate the channel density to be 46 µm² (see MATERIALS AND METHODS). Furthermore, apart from the range 20 to +20 mV, where single-channel currents were immeasurable, the data indicate that single-channel open probability is voltage independent (Fig. 3D), suggesting that opening of the ADP-ribose channel is not affected by membrane potential.

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Properties of plasma membrane current induced by ADP-ribose microinjection. ADP-ribose was injected to 10 or 200 nM. Bars represent mean ± SE. Control, injection of intracellular solution (ICS) to 1% in natural filtered seawater (NSW); 0 Ca2+ and 0 Na+ are 0 Ca2+ and low-Na+ seawaters, respectively. Hi Ca2+, seawater containing 100 mM Ca2+. ** Significance at 99% level, Student's t-test.

View larger version (14K): [in this window] [in a new window]   View larger version (18K): [in this window] [in a new window]   View larger version (10K): [in this window] [in a new window]   View larger version (11K): [in this window] [in a new window]   Fig. 3.   Electrophysiological properties of the cADPR/ADP-ribose-sensitive channel. A: current-voltage (I-V) graph of peak currents recorded at different clamped membrane potentials. cADPR () and ADP-ribose () were used at 5 µM and 10 nM intracellular concentration, respectively. Data represent mean with bars representing SE (n = 6 for both reagents). B: single-channel currents recorded with ADP-ribose and cADPR. cADPR and ADP-ribose intracellular concentrations were as in A. Control, single-channel trace before addition of reagent. ADP-ribose excised patch represents a recording in outside-out configuration after excision of a patch of membrane from the oocyte. ADP-ribose was 100 nM in the pipette (n = 3 for all experiments). C: single-channel reversal potential (Em) for currents triggered by nicotinamide nucleotide metabolites. Data represent mean and SE (n = 3 for both reagents); , cADPR; , ADP-ribose. D: analysis of single-channel currents. Top: open/close times for ADP-ribose channel. Data from a single-channel recording at 80 mV were analyzed. Solid bars, open time; open bars, close time. Bins are of 10-ms intervals for open time and 50-ms intervals for close time. Bottom: probability of ADP-ribose channel opening. Data in top were used to calculate open probability of the ADP-ribose channel at diverse membrane potentials. Apart from range 20 to +20 mV, open probability remained equal. Dip in probability between 20 and +20 mV could be due to the inability to measure currents at these potentials.

We ruled out the contribution of Ca²⁺ in gating the current, because the peak inward current gated by either cADPR or ADP-ribose was augmented, not diminished, after injection of Ca²⁺ chelators to buffer cytoplasmic Ca²⁺ (Fig. 4). Furthermore, neither eight-substituted analogs of cADPR, which competitively inhibit cADPR-induced Ca²⁺ release (27), nor the ryanodine receptor agonist ryanodine or antagonist ruthenium red (11) had any detectable effect on channel activity gated by cADPR (Fig. 4). These data suggest that the channel is not gated by local Ca²⁺ increases triggered by cADPR or through ryanodine receptors.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   ADP-ribose channel is not gated by Ca2+. Data represent mean with bars as SE. Currents triggered by 5 µM cADPR and 10 nM ADP-ribose are taken as controls. In presence of BAPTA, 10 mM intracellular concentration, there is a significant rise in the peak current triggered by 5 µM cADPR or 10 nM ADP-ribose. 8-NH2-cADPR represents an experiment where oocytes were preloaded with the cADPR inhibitor 8-NH2-cADPR to 100 µM, and the currents triggered by 5 µM cADPR were measured. 8-Br-cADPR represents an experiment using a similar cADPR inhibitor. Intracellular concentration of 8-Br-cADPR was 200 µM. Ryanodine was injected alone to an intracellular concentration of up to 10 µM. In final experiment, oocytes were preinjected with Ca2+-induced Ca2+ release antagonist ruthenium red (Ruth. Red) to 500 µM. ** Significance at 99% level, Student's t-test.

The ADP-ribose channel is gated by sperm at fertilization. The reversal potential, conductance, and Ca²⁺-independent properties of the ADP-ribose channel closely resemble the properties of a previously reported ion channel gated by ascidian sperm at fertilization (4, 5). This suggests that sperm trigger the hydrolysis of nicotinamide nucleotides at fertilization to ADP-ribose and that this triggers the fertilization current. We tested this hypothesis by measuring the current induced by ascidian sperm in the presence of nicotinamide. When 50 mM nicotinamide was added to the bath, the peak current induced at fertilization was strongly inhibited (Fig. 5A and Table 3). The ascidian fertilization current includes both Ca²⁺-dependent and Ca²⁺-independent components (unpublished observations). The fertilization current observed when cytoplasmic Ca²⁺ is buffered by Ca²⁺ chelators (26) is also blocked by nicotinamide (Fig. 5 and Table 3), strongly suggesting that this current is triggered by a mechanism involving nicotinamide nucleotide metabolism and is not Ca²⁺ dependent. Single-channel data for the sperm-induced current demonstrated a unitary conductance of 140 pS and reversal potential of +15 mV (Fig. 5B and Table 2), strongly suggesting that the ADP-ribose channel and the fertilization channel are equivalent.

View larger version (12K): [in this window] [in a new window]   Fig. 5.   A: effect of nicotinamide on membrane current at fertilization. Top trace, control fertilization current [natural filtered seawater (NSW)]. Second trace (nicotinamide), currents observed at fertilization when oocytes were bathed in 50 mM nicotinamide dissolved in NSW. Third trace (BAPTA), an experiment where BAPTA was microinjected to 10 mM intracellular concentration, and oocyte fertilized in NSW. Fourth trace (BAPTA + nicotinamide), an oocyte microinjected with BAPTA to 10 mM intracellular concentration, incubated in 50 mM nicotinamide in NSW and fertilized. Scale bars represent time and current. B: single-channel trace of ion channel opened at fertilization. Scale bar represents time against peak single-channel current. Current has an equivalent unitary conductance to ADP-ribose channel. See Table 2 for statistics.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

In this paper, we have shown that microinjection of nicotinamide nucleotides and nicotinamide nucleotide metabolites triggers an inward current in ascidian oocytes by opening a specific ion channel. Our data point strongly to ADP-ribose as the physiological trigger for the ion channel. The channel has a unitary conductance of 140 pS and furthermore appears to be nonselective. The presence of large, nonselective ion channels in the oocyte plasma membrane appears very unusual. However, such ion channels do exist in other species; for example, Ca²⁺-gated nonspecific ion channels can be found in the plasma membrane of the sea urchin (8). The unique property of the ion channel characterized in the present paper is the fact that a second messenger, ADP-ribose, gates the channel. We believe that similar channels have not been previously reported in any system. In fact, although ADP-ribose is a well-characterized mediator in several cell processes (25), it is not generally thought to be a second messenger.

ADP-ribose and related compounds are produced through the metabolism of nicotinamide nucleotides (14, 15, 18). Generally, NAD⁺ is the precursor nucleotide, at least in systems where cADPR is the second messenger (17, 23). In the present data, NAD⁺ was the least sensitive nucleotide precursor in terms of peak current produced. It has previously been noted that all forms of adenine dinucleotide can be metabolized into active forms, in terms of Ca²⁺ release (3). These data suggest either that NAD⁺ is not exclusively metabolized by the nicotinamide nucleotide metabolic pathway or that enzymatic conversion between different forms of nicotinamide nucleotide occurs before metabolism of the active form occurs. The enzymatic constituents of the C. intestinalis nicotinamide nucleotide metabolic pathway are currently being examined by our laboratory and our collaborators.

Nicotinamide nucleotides are known to be metabolized at fertilization (10, 27). This suggests that nicotinamide nucleotide metabolites are important second messengers at fertilization. Although doubts remain as to the function of cADPR in the sea urchin oocyte at fertilization (19, 22), our data strongly suggest that nicotinamide nucleotide metabolites play an active role at fertilization. The fact that nicotinamide blocks both the fertilization current and the current induced by the injection of nicotinamide nucleotides suggests that the pathway of nicotinamide nucleotide metabolism to ADP-ribose is present in ascidians and that sperm sensitize this pathway at fertilization.

The mechanism of activation of the nicotinamide nucleotide metabolic pathway in other systems appears to involve the production of nitric oxide (13, 24, 32). We have recently shown that nitric oxide induces an inward current in ascidian oocytes through nicotinamide nucleotide metabolism (13), suggesting that nitric oxide may be a mediator in the generation of the ascidian fertilization current. Interestingly, soluble sperm extracts do not contain molecules capable of gating the fertilization current in ascidians (30, 31). Possibly, direct injection of nitric oxide by the sperm contributes to this process.