The NO Pathway Acts Late during the Fertilization Response in Sea Urchin Eggs*

Both the inositol 1,4,5-trisphosphate (InsP3) and ryanodine receptor pathways contribute to the Ca2+ transient at fertilization in sea urchin eggs. To date, the precise contribution of each pathway has been difficult to ascertain. Evidence has accumulated to suggest that the InsP3 receptor pathway has a primary role in causing Ca2+ release and egg activation. However, this was recently called into question by a report implicating NO as the primary egg activator. In the present study we pursue the hypothesis that NO is a primary egg activator in sea urchin eggs and build on previous findings that an NO/cGMP/cyclic ADP-ribose (cADPR) pathway is active at fertilization in sea urchin eggs to define its role. Using a fluorescence indicator of NO levels, we have measured both NO and Ca2+ at fertilization and establish that NO levels rise after, not before, the Ca2+ wave is initiated and that this rise is Ca2+-dependent. By inhibiting the increase in NO at fertilization, we find not that the Ca2+transient is abolished but that the duration of the transient is significantly reduced. The latency and rise time of the transient are unaffected. This effect is mirrored by the inhibition of cGMP and cADPR signaling in sea urchin eggs at fertilization. We establish that cADPR is generated at fertilization, at a time comparable to the time of the rise in NO levels. We conclude that NO is unlikely to be a primary egg activator but, rather, acts after the initiation of the Ca2+ wave to regulate the duration of the fertilization Ca2+ transient.

In sea urchins the immediate consequence of sperm-egg fusion is the generation of a single large transient elevation in cytosolic free calcium ([Ca 2ϩ ] i ), which is absolutely required for egg activation (1)(2)(3). Sea urchin eggs contain both InsP 3 1 and ryanodine receptor channels. The microinjection of either InsP 3 , ryanodine, or the ryanodine receptor agonist cADPR mobilizes [Ca 2ϩ ] i and activates eggs (4 -6). The presence of these receptors in sea urchin eggs has been demonstrated immunologically (7,8), and a sea urchin ryanodine receptor has now been cloned (9). Both these channel types contribute to the Ca 2ϩ transient at fertilization. Studies have shown that microinjection of the InsP 3 receptor antagonist heparin in conjunction with antagonists of the ryanodine receptor, such as ruthenium red and the cADPR inhibitor 8-amino-cADPR, block the Ca 2ϩ transient at fertilization (10,11). Injection with heparin alone, unless used at relatively high concentrations (12), does not abolish the fertilization Ca 2ϩ transient (10,11,13). cADPR and ryanodine receptor antagonists block the transient only in the presence of InsP 3 antagonists (10,11), if at all (14).
The precise roles of the InsP 3 and ryanodine receptor pathways at fertilization in sea urchins have been more difficult to ascertain. There is evidence to suggest that the InsP 3 receptor pathway has a primary role in causing Ca 2ϩ release and egg activation. InsP 3 is produced by members of the phosphatidylinositol phospholipase C (PLC) enzyme family and has been shown to be generated at fertilization (15)(16)(17)(18). Chemical inhibitors of PLC, pentosan polysulfate and U73122, can abolish fertilization Ca 2ϩ transients in sea urchin eggs (12,17). Molecular inhibition studies, using dominant negative PLC␥SH2 domain constructs, show that sperm-activated Ca 2ϩ release and egg activation are markedly delayed, then blocked, when activation of this enzyme is inhibited (19 -21). In mouse eggs, diffusion of a newly characterized PLC, PLC, from sperm to egg triggers Ca 2ϩ oscillations and embryo development (22). These data strongly suggest an InsP 3 -based initiation mechanism for the fertilization calcium wave.
Nonetheless an NO-mediated pathway has been proposed to be the primary mechanism for egg activation (23). In sea urchins the external application of NO to intact eggs and egg homogenates generates Ca 2ϩ transients, although this doesn't lead to egg activation (24,25). Eggs can be activated when microinjected with an NO donor (23). Ca 2ϩ elevation in eggs by NO has been demonstrated to require the operation of a cGMPmediated synthesis of cADPR (24,25), and levels of cGMP (26,18) and cADPR (18) are known to rise at fertilization in sea urchin eggs. Kuo and coworkers (23) found that microinjection of an NO scavenger into eggs abolished the fertilization Ca 2ϩ transient and prevented egg activation. Microinjection of recombinant nitric-oxide synthase (NOS) and calmodulin was sufficient to fully activate eggs, indicating that NOS substrates were present. The measurement of an increase in endogenous NO in fertilized eggs very soon after sperm-egg fusion using a fluorescent indicator dye supported the proposition that an NO rise might trigger the Ca 2ϩ wave and egg activation; NO and Ca 2ϩ were not measured simultaneously, so the precise tem-* This work was supported by a grant from the Wellcome Trust. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  poral relationship between the NO rise and the Ca 2ϩ transient could only be inferred (23). It was pointed out that initiation of the Ca 2ϩ transient by NO was unlikely to be due to the activation of a pathway mediated by cADPR, because antagonists of this pathway alone cannot block fertilization. A more recent study of two separate chordate species, where NO and Ca 2ϩ were measured simultaneously, could detect no NO increases either prior to, or after, the generation of the Ca 2ϩ transient, or obtain abolition of the transient with NO-signaling inhibitors (27). We, and others, were also concerned that the NO indicator dye used by Kuo and co-workers (23), diaminofluorescein (DAF-2), might also report the post-fertilization pH increase that occurs in sea urchins (2). The fluorescent product of the reaction of DAF-2 with NO is said to exhibit pH-dependent fluorescence intensity (28).
This report contributes to the debate on the relative roles of the InsP 3 and ryanodine receptor pathways at fertilization in sea urchin eggs. We have focused on the specific role of NO and the cADPR-regulated ryanodine receptor pathway. We have examined the temporal relationship between NO and Ca 2ϩ in sea urchin eggs at fertilization by simultaneously measuring these signaling components. By using specific inhibitors we establish that an NO/cGMP/cADPR pathway plays a major role in regulating the later phase of the fertilization Ca 2ϩ transient, supporting the view that the general role of the NO pathway is to enhance and sustain Ca 2ϩ signals independently of other signaling pathways (25). We have been unable to confirm an early role for NO in initiating the events that lead to egg activation in the sea urchin at fertilization.

EXPERIMENTAL PROCEDURES
Collection of Sea Urchin Eggs-Eggs were obtained by ovulating female Psammechinus miliaris (Marine Biological Station, Millport, Isle of Cumbrae, Scotland) or Lytechinus pictus (Marinus, Long Beach, CA) with an intracoelomic injection of 0.5 M KCl solution. The jelly was removed by several passages through Nitex mesh, and the eggs were suspended in artificial sea water (ASW: 410 mM NaCl, 39 mM MgCl 2 , 15 mM MgSO 4 , 2.5 mM NaHCO 3 , 10 mM CaCl 2 , 10 mM KCl, 1 mM EDTA, at pH 8.0). Unless otherwise stated, treatments and their corresponding controls were carried out on the same batches of eggs.
Imaging of Intracellular [Ca 2ϩ ] i and NO-To visualize NO, we examined the specificity of two available fluorescent indicators, DAF-2, and the more recently available dye DAF-FM. We compared changes in fluorescence of the two dyes in eggs at fertilization, with and without the addition of 100 M 5-(N,N-dimethyl)amiloride hydrochloride (DMA) to block the Na ϩ /H ϩ antiporter and thus prevent pH increases (29,30). DAF-2 responses were largely abolished in eggs fertilized in DMA (DAF-2 fluorescence levels 200 s post-fertilization were on average 9 Ϯ 0.7% higher than resting levels (n ϭ 6), whereas in the presence of DMA the increase was just 2 Ϯ 0.6% (n ϭ 5)). DAF-FM responses were not abolished by DMA treatment. We confirmed that DAF-FM loaded into unfertilized sea urchin eggs reported NO increases by the external application of 5.7 mM of the NO donor sodium nitroprusside (which caused a 6 Ϯ 2% fluorescence increase above resting levels after 5 min in four eggs) and that the fluorescence did not increase with the addition of 10 mM NH 4 Cl (fluorescence levels fell 2 Ϯ 0.5% below resting after 5 min in four eggs), which is known to cause an elevation in cytosolic pH in sea urchin eggs (31). Due to its specificity, DAF-FM was our indicator of choice for measuring NO changes.
Eggs of P. miliaris were transferred to poly-D-lysine (10 mg/ml)coated glass coverslips and maintained at 16°C in ASW. Fura-2 dextran, pentapotassium salt (10,000 M r , 10 mM in the pipette, Molecular Probes) in a buffer consisting of 0.5 M KCl, 20 mM PIPES, 0.1 mM EGTA at pH 6.8, was pulsed into eggs with drawn borosilicate glass micropipettes (GC150F-10, Clark Electromedical Instruments) using gas pressure, to a final concentration of ϳ50 -100 M. Injection volumes did not exceed 1.5% of the cell volume. Following injection, eggs were incubated in ASW containing 50 M DAF-FM DA for 15 min, followed by a 30 -40 min post-incubation step in ASW, or a 15-min post-incubation step in ASW followed by 15 min in low sodium ASW (50 mM NaCl, 360 mM choline Cl, 39 mM MgCl 2 , 15 mM MgSO 4 , 2.5 mM KHCO 3 , 10 mM CaCl 2 , 10 mM KCl, 1 mM EDTA at pH 8.0) supplemented with 100 M DMA with fertilization being carried out in this medium. For DAF-2 experi-ments, 50 M DAF-2 DA was added externally to eggs, and the procedure followed was then the same as that for DAF-FM loading in the absence of DMA. Oxyhemoglobin (Sigma, H-0267, lot 20K7618; 46% purity, assayed spectroscopically) was co-loaded with fura-2 dextran, whereas dibromo-BAPTA was loaded into cells previously injected with fura-2 dextran, prior to incubation with DAF-FM followed by 100 M DMA. Because inhibiting the fertilization pH changes was found to result in enhanced NO production and Ca 2ϩ transients of longer duration, experiments using oxyhemoglobin and dibromo-BAPTA were performed in the presence of DMA, so that differences in levels of NO produced and the Ca 2ϩ transient generated due to the antagonists were more readily apparent when compared with controls. All work was carried out at 16°C, and all post-injection steps were performed under red light. Eggs were illuminated on a stage of a Nikon Diaphot 300 microscope using a 100-watt mercury lamp light source and a UV-F 20ϫ numerical aperture 0.8 objective (Nikon). Simultaneous Ca 2ϩ and NO levels were measured using narrow band-pass filters (350, 380, and 490 nm, Chroma) housed in a Lambda 10-2 filter wheel (Sutter Instrument Co.), and a polychroic beamsplitter (61000v2bs, Chroma) with emission filter (61000v2m, Chroma). Images were collected using a charge-coupled device camera (Photometrix Coolsnap fx, Roper Scientific). Both the filter wheel and acquisition by the charge-coupled device camera were controlled by Metafluor software version 4.0 (Universal Imaging Corp., West Chester, PA). Free cytosolic Ca 2ϩ concentration was determined by ratioing average fluorescence intensity values calculated from the images excited at 350 and 380 nm, and standard CaCl 2 solutions were used to calibrate the system with corrections for viscosity (32). Images were processed by IDL software (Research Systems Inc.) on an Indigo 2 workstation (Silicon Graphics).
Treatment of Eggs with cGMP, 8-Bromo-cADPR, Nicotinamide, and (R p )-cAMP-S-All compounds were introduced into eggs of L. pictus using the same injection buffer and procedure as above. The amount injected never exceeded 1% of cell volume. For treatments involving cGMP, eggs were double-injected, first with fura-2 dextran (final concentration of ϳ20 M) or fura-2 dextran and ␤-NAD ϩ or inhibitors, and then with cGMP or injection buffer (control). For inhibitor treatments, control responses to cGMP were performed after each experiment on eggs from the same batch to exclude the possibility of a reduction in responsiveness to cGMP with time. Ca 2ϩ levels were measured using ratio photometry and calibrated as described previously (3). (R p )-cAMP-S (final concentration, 2-4 mM) was co-loaded into eggs of L. pictus with fura-2 dextran (final concentration, ϳ100 M) and Ca 2ϩ levels measured and calibrated as for the simultaneous Ca 2 ϩ and NO experiments, replacing the polychroic beamsplitter and emission filter with a 400-nm dichroic and 510-nm band-pass emission filter (Nikon).
Treatment of Eggs with 8-Amino-cADPR-Eggs of L. pictus were prepared as above and microinjected with a mixture of fura-2 pentapotassium salt and 8-NH 2 -cADPR (final intracellular concentration of ϳ10 M for both compounds) in injection buffer (0.5 M KCl, 20 mM PIPES, at pH 6.7). Experiments were performed at 20 -22°C. Free cytosolic Ca 2ϩ concentration was determined by ratioing fluorescence intensities at 340 and 380 nm using an emission wavelength of 510 nm. Ratio images were obtained using a fluorometric imaging system (monochromator, TILL Photonics, Germany) and Ionvision software (Improvision Ltd., UK) as described previously (33).
Radioreceptor Assay of cADPR Levels-Eggs of L. pictus (0.8 -1.5 ml) were dejellied, washed twice in ASW, and made to a final volume of 20 ml with ASW. 350 l of sperm was added, and the solution continuously mixed. At given time points, 2-ml aliquots were removed and centrifuged for 6 s at 6000 rpm. Excess ASW was removed, and eggs were treated with 3 M perchloric acid (1:1 w/v). The precipitated proteins were removed by centrifugation, and the supernatants were neutralized with 2 M KHCO 3 . To remove contaminating nucleotides that weakly interfere with [ 32 P]cADPR binding, tissue samples were treated with NADglycohydrolase (0.25 unit/ml), nucleotide pyrophosphatase (1.75 units/ml), alkaline phosphatase (50 units/ml), and apyrase (5 units/ml) for 4 h, as described previously (34). Acid extracts were stored at Ϫ20°C prior to use in the radioreceptor assay. The binding assay was performed as previously described (35). Briefly, tracer [ 32 P]cADPR was synthesized from a high specific activity (1000 Ci/nmol) radiolabeled form of its precursor [ 32 P]NAD ϩ . The cyclization reaction was carried out for 2 h at room temperature using 250 mCi of [ 32 P]NAD ϩ , 100 ng ml Ϫ1 of ADP-ribosyl cyclase, and 5 mM Tris-HCl, pH 7.4. cADPR was purified using high-performance liquid chromatography, and fractions containing cADPR were neutralized with Tris base. The binding assay consisted of sea urchin egg homogenate made up to a final concentration of 0.5 mg/ml protein in IM buffer (333 mM N-methylglucamine, 333 mM potassium acetate, 27 mM HEPES, 1.3 mM MgCl 2 ; pH titrated to 7.2 with acetic acid). Approximately 22.5 fmol of [ 32 P]cADPR was incubated with homogenate in a total volume of 250 ml for 10 min at room temperature. The binding reaction was terminated by rapid filtration with a Brandel cell harvester using GF/B filters washed immediately before filtration with ice-cold IM and washed twice with 2-4 ml of ice-cold IM immediately after filtration. Radioactivity retained on the filters was determined using standard scintillation counting techniques. Inhibitory effects of acid extracts on [ 32 P]cADPR binding were compared with standard curves constructed using authentic cADPR. As a control, each sample was heat treated at 85°C for 45 min to hydrolyze the cADPR to ADPR. The inhibitory effect on [ 32 P]cADPR binding of all samples was abolished by heat treatment.
Image and Statistical Analyses-Average pixel intensity values for 350-, 380-, and 490-nm excitation wavelengths were determined from a region of interest bordering the outside of labeled eggs using Metafluor image analysis software and background subtracted using values determined from a specified background region. DAF-2 and DAF-FM intensity values were baseline-corrected using the measured resting values recorded prior to insemination. Statistical analyses used a Student t test (one-tail, unpaired). Data were expressed as the means Ϯ S.E. for n eggs tested.

Levels of NO, as Measured with DAF-FM, Rise after the Initiation of the Ca 2ϩ Wave in Fertilized Sea Urchin Eggs-
DAF-FM is said to be a more faithful indicator of NO above pH 5.8 (28) and has been used to image NO production in several cell types, including rat neuronal cells and smooth muscle cells (37,38). We confirmed the specificity of DAF-FM in sea urchin eggs at fertilization (see "Experimental Procedures"). Using DAF-FM, in combination with the ratiometric Ca 2ϩ indicator dye fura-2, we were able to simultaneously measure NO and Ca 2ϩ levels in intact eggs at fertilization. Simultaneous measurements of Ca 2ϩ and DAF-FM fluorescence demonstrated that DAF-FM was not directly sensing Ca 2ϩ under our experimental conditions. In P. miliaris eggs loaded with DAF-FM and subsequently fertilized, an increase in fluorescence was found to occur after the onset of the Ca 2ϩ wave (Fig. 1, A and  B). On average, DAF-FM fluorescence started to increase 68.4 Ϯ 5.4 s post-fertilization (PF) or 5.3 Ϯ 4.2 s after peak Ca 2ϩ levels were reached (n ϭ 16). Increases in DAF-FM fluorescence were not attenuated when eggs were fertilized in low sodium sea water in the presence of 100 M 5-(N,N-dimethyl)amiloride hydrochloride (DMA, Fig. 1B) to block the Na ϩ /H ϩ antiporter and thus prevent an elevation of the pH (29,30). We observed differences in the shape of the fertilization Ca 2ϩ transient when DMA-treated, DAF-FM-loaded eggs were compared with those loaded with DAF-FM only (Fig. 1B, panel i). The duration of the Ca 2ϩ transient was significantly prolonged. The time taken for [Ca 2ϩ ] i to fall to 25% of the peak level was 189.8 Ϯ 5.5 s (n ϭ 16) in control eggs compared with 263.6 Ϯ 6 s in DMA-treated (n ϭ 14, p ϭ Ͻ 0.001, t test). The response of DAF-FM was also affected (Fig. 1B, panel ii). The magnitude of the response was significantly greater (the DAF-FM fluorescence levels, 200 s PF, were an average of 6.9 Ϯ 0.9% greater than resting in DMA-treated eggs and 3.9 Ϯ 0.4% greater than resting in untreated eggs; p ϭ Ͻ0.01, t test). These results confirm that NO levels increase in sea urchin eggs at fertilization and that this rise occurs during the period of sustained Ca 2ϩ elevation after the Ca 2ϩ wave has been generated. Interestingly, the enhanced levels of NO production in fertilized eggs in which the pH change is inhibited correlate with a Ca 2ϩ transient of increased duration.
It has been reported that the fluorescence levels of the NO indicator DAF-2 increase very early during the fertilization response and the consequent suggestion is that NO is a very early player in the fertilization signaling cascade (23). In this study we analyzed whole egg average fluorescence intensity immediately after sperm-egg fusion, in DAF-2-loaded eggs of P. miliaris (50 M, applied externally) fertilized in normal sea water (n ϭ 11), and DAF-FM loaded eggs with and without the addition of DMA (n ϭ 30) and failed to find the small increases in fluorescence previously observed. Careful inspection of both DAF-2 (Fig. 1C, panel i) and DAF-FM fluorescence images (Fig.  1C, panel ii) at the point of initiation of the fertilization Ca 2ϩ transient again offered no confirmation of these earlier observations. Analysis of a 30-m diameter region (average egg diameter, 120 m) at the point of sperm egg fusion (as defined by the initial Ca 2ϩ increase) in 17 eggs did not reveal any local increase in fluorescence. We estimate that, given the noise in our measurements, if a local NO increase occurs, the fluorescence increase locally must be Ͻ0.3% above mean resting values for DAF-FM-labeled eggs and Ͻ0.8% for those labeled with DAF-2.
Scavenging NO Shortens the Duration of the Fertilization Ca 2ϩ Transient-Oxyhemoglobin scavenges NO (39 -41). To determine whether prolongation of the Ca 2ϩ transient that we observed in DMA-treated eggs was due to NO production, we microinjected oxyhemoglobin (12-28 M) into DMA-treated P. miliaris eggs. Oxyhemoglobin both delayed and very markedly attenuated the post-fertilization increase in DAF-FM fluorescence (Fig. 1D). In control eggs, the start of the DAF-FM fluorescence increase was observed 60.6 Ϯ 8 s after peak [Ca 2ϩ ] i (108.2 Ϯ 8.1 s PF, n ϭ 10), whereas in oxyhemoglobin loaded eggs, where an increase occurred, it was significantly later, at 128.6 Ϯ 12.6 s after the [Ca 2ϩ ] i peak (171.7 Ϯ 13 s PF, n ϭ 8, p Ͻ 0.001, t test). In oxyhemoglobin-loaded eggs DAF-FM increased on average only 1.0 Ϯ 0.4% above resting levels 200 s PF (n ϭ 10), whereas control eggs showed a 7.3 Ϯ 0.8% increase (n ϭ 10, p Ͻ 0.001, t test). For two out of ten oxyhemoglobin-loaded eggs, no increase in DAF-FM fluorescence was detected during fertilization, but Ca 2ϩ transients leading to egg activation did occur. In the presence of oxyhemoglobin, the latency, magnitude, and rise time of the fertilization Ca 2ϩ transient was unaffected. The average control latency, rise time, and amplitude were 27.3 Ϯ 1.5 s, 21.6 Ϯ 0.8 s, and 1.04 Ϯ 0.1 M, respectively, in control eggs (n ϭ 10), and the corresponding values for 10 oxyhemoglobin-loaded eggs were 24.1 Ϯ 1.3 s, 22.5 Ϯ 2.7 s, and 1.1 Ϯ 0.06 M. However, the rate at which calcium decreases after the [Ca 2ϩ ] i peak was significantly increased. The time taken for a fall to 25% of peak [Ca 2ϩ ] i was 243.1 Ϯ 14.6 s (n ϭ 10) in control eggs, compared with 192.9 Ϯ 9.9 s in oxyhemoglobin-loaded eggs (n ϭ 10, p Ͻ 0.01, t test). These observations show that inhibiting the NO rise in sea urchin eggs does not abolish the fertilization Ca 2ϩ transient and has no effect on the initiation or rate of rise of the transient. However, inhibiting the NO rise shortens the transient. Because enhanced NO levels lengthen the transient (Fig. 1B), these data imply a causal relationship between NO production and Ca 2ϩ transient duration.
NO Production at Fertilization Is Ca 2ϩ -dependent-It has been demonstrated that the treatment of unfertilized eggs with the calcium ionophore ionomycin causes NO production as measured by the accumulation of nitrite (23). The Ca 2ϩ chelator dibromo-BAPTA is very effective at suppressing Ca 2ϩ transients and gradients (42,43). Microinjection of 2.5-7 mM dibromo-BAPTA into P. miliaris eggs before insemination prevents both the fertilization Ca 2ϩ wave (Fig. 2B) and the DAF-FM increase (Fig. 2D). To control for the possibility that BAPTA itself might scavenge NO, we added sodium nitroprusside to BAPTA-injected eggs labeled with DAF-FM and noted NO increases equivalent to control eggs (not shown). This finding demonstrates that the post-fertilization NO increase is dependent on an increase in [Ca 2ϩ ] i .
cGMP Effects Ca 2ϩ Release via the Ryanodine Receptor Pathway and Regulates the Duration of the Ca 2ϩ Transient at Fertilization-NO has been shown to trigger Ca 2ϩ release in intact eggs and egg homogenates and to promote cGMP synthesis (25). The inhibition of NO-and cGMP-mediated Ca 2ϩ release by nicotinamide, an ADP-ribosyl cyclase inhibitor, implicates cADPR as the downstream effector of this pathway (24,25). We confirm and extend these findings. Fig. 3 shows that the  (Table I). Moreover, the activation response to lower doses of cGMP was potentiated by pre-injection of ␤-NAD ϩ (100 M), the cADPR precursor, and enabled concentrations of cGMP as low as 1 M to activate eggs (Table I). The role of cGMP at fertilization was examined by injecting eggs of L. pictus with 2-4 mM (R p )-cAMP-S, which inhibits cGMP-dependent protein kinase (44) and is effective in sea urchin egg homogenates (45). The presence of (R p )-cAMP-S reduced the duration of the fertilization Ca 2ϩ transient (Fig. 3C). The mean time of a fall to 50% of peak [Ca 2ϩ ] i was 146.5 Ϯ 13 s in control eggs (n ϭ 8) compared with 111 Ϯ 9.2 s in the presence of (R p )-cAMP-S (n ϭ 7, p ϭ Ͻ0.05, t test).  Ͻ0.01, t test). These findings suggest cGMP regulates Ca 2ϩ release during the later phase of the fertilization Ca 2ϩ transient via the activation of cGMP-dependent protein kinase, which leads to cADPR production.
cADPR Levels Increase after the Rise of the Fertilization Ca 2ϩ Transient-No imaging method yet exists to measure cADPR in single eggs, so cADPR levels in a population of fertilized L. pictus eggs were measured to establish the temporal relationship between NO production, cADPR synthesis, and the fertilization Ca 2ϩ transient. cADPR levels have been previously shown to increase by 30 s after insemination in egg populations of the sea urchin species Anthocidaris crassispina and Hemicentrotus pulcherrimus (18). We have previously shown (16) that at appropriate sperm densities a peak of around 80% of eggs in a population are undergoing a Ca 2ϩ transient at 25 s after insemination. Fig. 4A shows that levels of endogenous cADPR increase by 25 s after fertilization and reach a peak 50 s after insemination of eggs of L. pictus. Levels of cADPR then [Ca 2ϩ ] i was measured using fura-2 dextran and ratio photometry. C, inhibition of cGMP-dependent protein kinases using (R p )-cAMP-S (2-4 mM) significantly reduced the duration of the fertilization Ca 2ϩ transient. Plots are of whole cell [Ca 2ϩ ] i calculated from images of a representative example of a fura-2 dextran-loaded egg treated with (R p )-cAMP-S (2.7 mM, boldface line) and a control (thin line). All quoted concentrations are intracellular. Note that the small, immediate increase in [Ca 2ϩ ] i after microinjection is an artifact of microinjection (13). decline to resting levels by ϳ150 s. cADPR production appears to lag the fertilization Ca 2ϩ transient. This is in contrast to previously measured InsP 3 production, which precedes, and coincides with, the fertilization Ca 2ϩ transient (16; reproduced in this report as Fig. 4B).
cADPR Regulates the Later Phase of the Fertilization Ca 2ϩ Transient-The timing of cADPR production suggests it may have a role during the later phase of the Ca 2ϩ transient. We tested this idea by using 8-amino-cADPR, a competitive inhibitor (Fig. 5). After microinjection of 10 M 8-amino-cADPR into eggs of L. pictus, the duration of the fertilization Ca 2ϩ transient (the time taken for [Ca 2ϩ ] i levels to fall to 500 nM postmaximum, or ϳ25% of peak values) was decreased significantly to 119 Ϯ 19 s from 279 Ϯ 55 s in the control (p Ͻ 0.05, t test, n ϭ 12). This result demonstrates that cADPR plays a major part in sustaining the Ca 2ϩ transient.

DISCUSSION
Both InsP 3 -and ryanodine-receptor channels participate in the sea urchin fertilization Ca 2ϩ response (10,11). Previous work has uncovered a signaling pathway that could regulate ryanodine receptors at fertilization. Data are consistent with the notion that NO leads to cGMP production via guanylate cyclase activation, and elevated cGMP promotes the production of cADPR by ADP-ribosyl cyclase (24,25), leading to ryanodine receptor activation and Ca 2ϩ release. Because fertilizationassociated increases in both cGMP (18,26), cADPR (18), and NO levels (23) have been measured, it is clear that the NO/ cGMP/cADPR pathway is present in eggs and activated at fertilization. What is less clear is the contribution that the pathway makes to the physiology of Ca 2ϩ release at fertilization. On the one hand, it is suggested that an NO-mediated pathway is the primary mechanism for egg activation (23), while on the other there are abundant data that suggest that the InsP 3 signaling pathway plays the primary role (12, 14, 17, 19 -22). In this study we have used a physiological approach to dissect the contribution that the NO/cGMP/cADPR pathway makes in generating the fertilization calcium signal and show that the pathway acts late in the response.
Nitric Oxide Is Generated after the Fertilization Ca 2ϩ Transient Is Initiated-We have used a recently available, and pHstable, NO indicator dye, DAF-FM, to show that NO levels do indeed rise at fertilization. For the first time in the sea urchin egg we have measured NO and Ca 2ϩ levels simultaneously. We TABLE I Release of the fertilization envelope by cGMP and its inhibition The percentage of activated eggs, and the latency between injection and the start of envelope elevation, are shown for the indicated treatments. cGMP intracellular concentration ranged from 1 M (null effect) to 100 M (approximately 90% of eggs activated). Nicotinamide, ␤-NAD ϩ , and 8-bromo-cADPR, at the indicated intracellular concentrations, were injected 2-5 min before cGMP injection. Previous injection with KCl did not affect the cGMP response. have been unable to detect any increase in NO prior to the initiation of the Ca 2ϩ wave after sperm-egg fusion using DAF-FM, either locally at the point of sperm-egg fusion, or globally throughout the egg. When we performed the same analysis using the indicator DAF-2 used in a previous study (23), we again failed to detect an increase in NO prior to the initiation of the wave.
Nitric Oxide Production Depends on the Fertilization Ca 2ϩ Transient and Is Sensitive to pH-The two major ionic signals at fertilization in sea urchins are the Ca 2ϩ transient and the pH change (2). As we have seen, preventing the pH increase significantly enhances NO production as measured by DAF-FM, and microinjection of the Ca 2ϩ chelator BAPTA blocks both the fertilization Ca 2ϩ transient and the NO increase. This demonstrates that NO production is dependent on the Ca 2ϩ increase at fertilization and may be regulated by changes in pH.
Nitric Oxide Regulates the Later Phase of the Ca 2ϩ Transient-The temporal characteristics of the NO rise may give us clues to its physiological role. A lack of a detectable early increase using either DAF-2 or DAF-FM is not readily reconciled with a role in initiation of the Ca 2ϩ transient. This is supported by the absence of any alteration in rise time or latency (Table II) of the fertilization Ca 2ϩ transient by concentrations of an NO scavenger, oxyhemoglobin, sufficient to suppress significantly the NO increase we observed later during fertilization. These results contrast markedly with those of Kuo et al. (23). It could be argued that, due to the impurity of commercially available oxyhemoglobin, we were unable to inject concentrations of oxyhemoglobin previously reported to abolish egg activation (23). Although this is a possibility, we have found that abolition of the NO rise does not coincide with an abolition of the fertilization Ca 2ϩ transient. Because NO levels rise during the period of sustained Ca 2ϩ elevation after the peak, a role for regulating this phase of the Ca 2ϩ signal is possible. There are two pieces of evidence for this: inhibiting the pH change at fertilization results in enhanced NO production linked to Ca 2ϩ transients of longer duration, and shorter duration Ca 2ϩ transients are seen in cells where NO increases are inhibited.
The NO/cGMP/cADPR/RyR Pathway Acts Late at Fertilization-We have extended previous findings using agonists and antagonists of cGMP and cADPR signaling to show that a cGMP/cADPR/RyR pathway, previously demonstrated in homogenates, operates in a very similar manner in intact, unfertilized eggs. However, the use of such inhibitors at fertilization has been limited to demonstrating that the ryanodine receptor pathway played a role at fertilization (10,11). Here we show that the effect of an inhibitor of cGMP-dependent protein kinase or a competitive cADPR antagonist is to curtail the fertilization Ca 2ϩ transient. It is also clear from this work that the same inhibitors have no effect on the latency of the fertilization Ca 2ϩ transient (Table II). These observations, together with those using the NO scavenger oxyhemoglobin, indicate that a major function of the NO/cGMP/cADPR/RyR pathway is to sustain and prolong the sea urchin fertilization Ca 2ϩ signal.
Measurements of cADPR Levels at Fertilization Suggest cADPR Is a Regulator, Not an Initiator, of the Ca 2ϩ Transient-Measurements of cGMP production at fertilization indicate that levels of this messenger rise quickly after fertilization to a peak at around 30 s (18,26). Here we show a similar time course for cADPR production in L. pictus, which is in agreement with the time course for cADPR production at fertilization in two other sea urchin species, A. crassispina and H. pulcherrimus (18). It is very difficult to draw conclusions about the temporal sequence of cGMP and cADPR production, because, with the resolution available to us, both seem to peak at about the same time, roughly coincident with the peak of the population Ca 2ϩ transient. What we can say is that the timing of cADPR production at fertilization appears to be later than that of InsP 3 measured in a previous study (16) in L. pictus and suggests that cADPR regulates the later phase of the transient. Our data indicate cADPR levels peak at 200 nM at fertilization; cADPR levels of 40 -150 nM are reported to release Ca 2ϩ in intact sea urchin eggs on microinjection (5,10,11), so the increases we observe are physiologically relevant. One paradox is that the previously measured levels of cGMP in A. crassispina and H. pulcherrimus (18) are orders of magnitude lower than the concentration required to activate intact eggs on microinjection (46, and the present study). However, this is not unexpected if the role of the NO/cGMP/cADPR/RyR pathway is to enhance an existing Ca 2ϩ transient, rather than initiate it. On this interpretation, the activation of eggs by NO, cGMP, and cADPR is not a physiological response, although it reveals the presence of the signaling pathway.
The Stimulation of the NO/cGMP/cADPR/RyR Pathway by Ca 2ϩ -In direct contrast to our observation that the NO increase is Ca 2ϩ -dependent, it has been found that artificial Ca 2ϩ increases do not stimulate the production of cGMP (18,26). This is surprising given that sea urchin eggs are known to contain the Ca 2ϩ /calmodulin-dependent neuronal NOS isoform and can generate NO when Ca 2ϩ is artificially increased (23). The resolution of this paradox will depend upon a robust and readily available method of measuring cGMP with good time resolution.
Conclusions-The prevalent hypothesis of egg activation is that InsP 3 is generated by a protein that diffuses from sperm to egg when sperm and egg fuse at fertilization (47,48). In some species this protein may be a phospholipase C (22) and in others a component of tyrosine kinase signaling pathways (49 -51). It is generally accepted that the cGMP/cADPR pathway does not initiate activation, because blockers of the pathway cannot by themselves block fertilization, in marked contrast to antagonists of the phosphoinositide signaling pathways (12, 14, 17, 19 -21). Our data are consistent with this majority view and assign a late, sustaining role to the NO/cGMP/cADPR/RyR pathway at fertilization in sea urchin eggs. Finally, it may be that, although the NO pathway itself is widely distributed (52)(53)(54), its involvement at fertilization is confined to echinoderms or perhaps echinoids. Neither frog (55), ascidian (56), nor mammalian oocytes (57, 58) rely on the RyR receptor at fertilization, and neither mouse nor ascidian oocytes showed any evidence of increases in NO at fertilization (27). One obvi-

TABLE II
The latency of the fertilization Ca 2ϩ transient is unaffected by modulators of the NO, cGMP, and cADPR signalling pathways The mean latent period (ϮS.E.), from the start of the ''cortical flash'' to the start of the spike rise, was measured. The NO inhibitor oxyhemoglobin (12-28 M), the cGMP-kinase inhibitor (R p )-cAMP-S (2-4 mM), and the cADPR inhibitors nicotinamide (10 mM, external) and 8-bromo-cADPR (10 -30 M) had no significant effect on the length of the latent period. This was also true in eggs treated with the cADPR precursor ␤-NAD ϩ (200 M). ous difference between echinoids and most other animals is that their eggs are arrested at interphase of the cell cycle rather than in the meiotic metaphase. Whether this correlation between the stage of meiotic arrest and the presence of an NO signaling cascade is physiologically significant remains to be determined.