Ca2+ Signaling Occurs via Second Messenger Release from Intraorganelle Synthesis Sites

Summary Cyclic ADP-ribose is an important Ca2+-mobilizing cytosolic messenger synthesized from β-NAD+ by ADP-ribosyl cyclases (ARCs). However, the focus upon ectocellular mammalian ARCs (CD38 and CD157) has led to confusion as to how extracellular enzymes generate intracellular messengers in response to stimuli. We have cloned and characterized three ARCs in the sea urchin egg and found that endogenous ARCβ and ARCγ are intracellular and located within the lumen of acidic, exocytotic vesicles, where they are optimally active. Intraorganelle ARCs are shielded from cytosolic substrate and targets by the organelle membrane, but this barrier is circumvented by nucleotide transport. We show that a β-NAD+ transporter provides ARC substrate that is converted luminally to cADPR, which, in turn, is shuttled out to the cytosol via a separate cADPR transporter. Moreover, nucleotide transport is integral to ARC activity physiologically because three transport inhibitors all inhibited the fertilization-induced Ca2+ wave that is dependent upon cADPR. This represents a novel signaling mechanism whereby an extracellular stimulus increases the concentration of a second messenger by promoting messenger transport from intraorganelle synthesis sites to the cytosol.


Production of recombinant ARC proteins
Rib-hydrolase domains (amino acid residues in brackets) of ARCα (19 to 300), ARC® (23 to 295) and ARCγ (54 to 285) were cloned into pGEX-2TKP and pPICZαA for expression in either bacteria (as GST-fusion proteins) or in yeast (as secreted proteins), respectively.

b) Stratified eggs
For stratification, eggs were centrifuged on a cushion of 1 M sucrose at 17,500 g for 30 min. To dislodge cortical granules [3], eggs were suspended in 400 mM urethane immediately before and during centrifugation. In addition to the distended morphology, stratification was verified by appropriate migration of the nucleus (stained with 10 μg/ml Hoechst 33342), yolk platelets (stained with 1 μM Lysotracker Red DND-99) mitochondria (stained with 300 nM tetramethylrhodamine ethyl ester), and the appearance of the oil droplet (LC Davis and AJ Morgan, unpublished data, 2008).

c) Cortical lawns
Live eggs were washed 3× in Ca 2+ -free ASW and settled onto poly-L-lysine coated glass coverslips. The eggs were then ruptured with a jet of cold Ca 2+ -free buffer (0.5 M NaCl, 10 mM KCl, 2.5 mM NaHCO 3 , 62.5 mM NaOH, 20 mM EGTA, pH 8) which sheared away the cytoplasm leaving the cortices of the eggs bound to the solid support [4].

d) Egg homogenates
Following a previous protocol [5], de-jellied eggs were washed twice in Ca 2+ -free ASW supplemented with 1 mM EGTA and then twice with Ca 2+ -free ASW. Finally, eggs were washed with an intracellular medium without EGTA (GluIM: 250 mM potassium gluconate, 250 mM N-methylglucamine, 1 mM MgCl 2 , 20 mM HEPES, pH 7.2) and homogenized on ice in GluIM (50%) supplemented with an ATP regenerating system (2 mM ATP, 20 U/ml creatine phosphokinase, 20 mM phosphocreatine) and EDTA-free protease inhibitor cocktail (Complete TM , Roche) using a pre-chilled Dounce glass tissue homogenizer, size 'A' pestle. Homogenates were centrifuged for 10 s at 13, 000 g to remove cortical granules, and the supernatants aliquoted and stored at -80 o C.
To examine the effect of EGTA upon vesicle integrity during homogenate preparation ( Figure S3), homogenates (50%) were prepared as above with the following modifications: ASW washes and GluIM at all steps was prepared with or without 10 mM EGTA and the removal of cortical granules by centrifugation was omitted.

e) Cell surface complexes (CSCs)
CSCs were prepared as before [6]. Dejellied eggs were washed three times in ASW and suspended in intracellular medium (IM: 220 mM potassium glutamate, 500 mM glycine, 10 mM NaCl, 5 mM MgCl 2 , 10 mM EGTA, 2.5 mM MgATP, 5 mM dithiothreitol, EDTA-free protease inhibitor cocktail (Complete TM , Roche), pH 6.8). The eggs were washed three times with IM and homogenised in a pre-chilled Dounce glass homogenizer with 3-5 strokes of a tight-fitting pestle. CSCs were pelleted at 1,000 g for 1 min and suspended in fresh IM. Centrifugation was repeated until the wash buffer was clear and when only large sheets of egg cortex were visualised under the light microscope. Pellets of CSC were either used immediately or frozen in liquid nitrogen and stored at -80 o C.

f) Cortical granules
Cortical granules (CGs) were detached from the CSCs by sucrose displacement [7]: icecold 1 M sucrose containing, 1 mM EGTA and EDTA-free protease inhibitor cocktail (Complete TM , Roche) was added to a pellet of CSCs by allowing the solution to flow gently down the side of the tube and to wash over the top of the pellet without disturbing it. The sucrose solution was removed and the CSC pellet was resuspended in 10× volume of the sucrose solution. The tube was swirled by hand on ice and incubated on ice for 60 min. CG detachment was checked with phase constrast microscopy. When detachment of CGs was complete, the plasma membrane-vitelline layer (PMVL) components of the CSC were pelleted by centrifugation at 1,000 g for 20 min. The supernatant, which contains the CGs was removed and pelleted at 20,000 g for 30 min.

g) CSC or CG Lysis
CSCs or CGs were lysed by resuspension in 10 mM HEPES, pH 7 or 10 mM acetate buffer, pH 5 followed by incubation on ice for 10 min. Membranes (particulate fraction) were separated from the soluble fraction by ultracentrifugation at 100,000 g for 1 h. The pellets were resuspended in the same volume of lysis buffers. Hypotonic lysis was confirmed using confocal microscopy and the release of Lysotracker Red staining (AJ Morgan and LC Davis, unpublished data, 2008).

Immunoblotting analysis
Protein samples were mixed with Laemmli sample buffer (under reducing conditions) and proteins resolved in 10% acrylamide gels by SDS-PAGE and blotted onto nitrocellulose membranes using 20 mM sodium phosphate buffer, pH 6.7. Membranes were blocked in PBS (10 mM phosphate buffer, 2.7 mM KCl, 137 mM NaCl, pH 7.4) containing 0.5% Tween and 5% dry skimmed milk and incubated with affinity-purified antibodies. HRPconjugated anti-rabbit IgG (Sigma) was used as a secondary antibody and specific bands visualized by chemiluminescence using ECL reagents (GE healthcare).
Eggs and cortical lawns were fixed with 4% paraformaldehyde in modified PBS (10 mM phosphate buffer, 10 mM KCl, 450 mM NaCl, pH 7.4). Eight-μm sections of ovary were fixed with 10% formalin. Fixed samples were permeabilized with 0.2% Triton X-100 where indicated. After blocking non-specific sites with 2% goat serum, samples were incubated with affinity-purified antibodies (or control IgGs) and labelling detected with fluorescently-labelled secondary antibodies. Immunofluorescence was viewed on a confocal laser scanning microscope and images collected in the Multitrack mode which alternates laser lines to reduce bleed through. Excitation/emission (nm): green (488/505-530), red (543/>560).

Immunogold electron microscopy
Eggs were fixed with modified Karnovsky's solution, post-stained with 0.01% OsO 4 , and embedded in Spurr's resin. Silver-gold sections (approximately 90 nm) were placed on nickel grids and non-specific sites blocked. Samples were incubated with affinity-purified antibodies (or control IgGs) and labelling detected with secondary antibodies conjugated to 15 nm colloidal gold particles. Sections were post-fixed with 2% glutaraldehyde, and stained with uranyl acetate and lead citrate. Sections were visualized at 100 KeV with a Philips (New York, NY, USA) 410 electron microscope with an Advantage HR CCD camera from Advanced Microscopy Techniques (AMT) run by proprietary software.

Assays for ARC activity a) TLC assay
The synthesis of [ 32 P]cADPR was assessed by incubation of either yeast-expressed recombinant ARCs or CSC and CG samples with 16.2 nM [ 32 P]-β-NAD + (GE Healthcare) for 1 h at 20 o C and reaction products separated on silica gel thin layer chromatography (TLC) sheets (Merck) in a mixture of 30% water / 70% ethanol / 0.2 M ammonium bicarbonate [9]. Autoradiography was carried out after exposure on a storage phosphor screen (Typhoon, GE Healthcare). Spots were verified by the co-migration position of authentic β-NAD + , cADPR and ADPR. Synthesis of cADPR was also independently confirmed by a cycling assay which measures cADPR mass [10]. Specific ARC activity of cortical granules was calculated by dividing the absolute cADPR production by the density of the ARCβ band on an immunoblot. b) Ca 2+ release assay L. pictus egg homogenates (with cortical granules removed) were sequentially diluted from 50% stocks to 2.5% over a period of 3 h at 17 o C using GluIM, containing an ATP regenerating system and 3 μM Fluo-3 (Invitrogen) [5]. Fluorescence was monitored at 17 o C in a PerkinElmer LS-50B fluorimeter using excitation 506 ± 4 nm, and emission 526 ± 4 nm. To assess whether transport inhibitors affected either intracellular Ca 2+ release channels or ARC activity, messengers or β-NAD + respectively were added to the homogenate following a 3 minute pre-incubation with either 0.1% DMSO (control) or the indicated concentration of inhibitor. Agents were added from 100× stocks. Fluorescence was calibrated using the standard equation [Ca 2+ ] = K d (F -F min )/(F max -F), using a K d of 0.4 μM; F min (where F is fluorescence) and F max were determined by the addition of 500 μM EGTA and 10 mM Ca 2+ respectively at the end of each run.

Nucleotide uptake
CSCs were incubated for various times with ~16 nM [ 32 P]β-NAD + in incubation buffer (GluIM supplemented with 10 mM EGTA and 100 μM ATP-Na + salt) at 20 o C and separated from free nucleotide using vacuum filtration (Whatman GF/B filters) and three washes of ice-cold GluIM supplemented with 10 mM EGTA and 160 μM unlabelled β-NAD + . For zero time points CSCs were added to excess ice-cold wash buffer containing [ 32 P]β-NAD + and filtered immediately. Non-specific binding was determined by preincubation with 100 μM unlabelled 'cold' β-NAD + . Luminal 32 P was distinguished from 32 P bound to the surface by including 300 μM digitonin in the wash buffer once the CSCs were filtered. Digitonin immediately lyses cortical granules without grossly affecting their morphology (AJ Morgan unpublished data, 2008). Where indicated, dipyridamole (DPM), NBTI, or Indoprofen were preincubated for 2 min prior to addition of [ 32 P]β-NAD + . GF/B filters were exposed to a phosphor-storage screen and scanned on a Typhoon imager (GE Healthcare).
In order to identify the 32 P-labelled nucleotides that accumulate in the cortical granule lumen, CSCs were preincubated for 2 min with DMSO or DPM prior to incubation for 18 min with 50 nM [ 32 P]β-NAD + in incubation buffer (GluIM supplemented with 10 mM EGTA and 100 μM ATP-Na + salt) at 20 o C. Following rapid vacuum filtration (Whatman GF/B filters) and three washes of ice-cold GluIM supplemented with 10 mM EGTA and 500 μM unlabelled β-NAD + , the cortical granules were lysed in 4.8% TCA to release 32 P -labelled luminal nucleotides, precipitate protein and terminate ARC activity. Precipitated protein was pelleted and the supernatants were concentrated 10-fold under vacuum before separation and identification of 32 P-labelled nucleotides by TLC analysis (see TLC assay above). Time zero intensities were subtracted from each nucleotide spot and data normalized to the respective DMSO controls.

Monitoring intracellular Ca 2+ concentration and vital staining in intact eggs
L. pictus eggs on poly-lysine-coated glass coverslips were maintained in ASW at room temperature and microinjected with a Ca 2+ -sensitive fluorescent dye, fluo-4 dextran To estimate the intracellular dipyridamole concentration, we exploited the intrinsic, weak UV fluorescence of dipyridamole recorded on the confocal microscope (excitation 364 nm, emission >385 nm). Acquisition settings were such that unlabelled eggs showed little autofluorescence, but the signal increased upon addition of 30 μM dipyridamole (extracellular concentration) to the bath.
The endoplasmic reticulum was stained in live eggs according to previous methods [12].
In summary, a saturated solution of DiI (DiIC 18 (3), Invitrogen) was prepared in soyabean oil (Sigma-Aldrich) and microinjected into live eggs (injection pressure ~5500 hPa for 6-8 s). The DiI diffused from the central oil droplet into the contiguous membrane system of the ER in 15-30 min and was imaged on a confocal laser scanning microscope using a 40 × objective (1-µm slice) and a red filter set (excitation/emission: 543/>560 nm). For colocalization of ER and acidic vesicles, eggs were incubated with 1 µM Lysotracker Green DND-26 (Invitrogen) immediately after DiI injection. Following dye equilibration, red and green channels were collected in the confocal Multitrack mode (to reduce potential bleed-through) and using a 63 × objective (0.7-µm slice).

Data Analysis
Data are presented as the mean ± standard error of the mean of n preparations. Statistical analysis was conducted using Student's t test (for two means) and an analysis of variance  [13,14]. Tabulated are the amino acid alternatives found in multiple independent clones. Next to each amino acid position, the observed alternative amino acids are indicated in one-letter code. Polymorphisms at amino acids 29 and 308 (ARCβ) and 36 and 77 (ARCγ) explain some of the differences between our sequences and those found by [15]. (F) Schematic representation of predicted topology of mature S. purpuratus ARC proteins. The domains indicated are ribosyl-hydrolase (Ribhydrolase) predicted by Pfam (accession number PF02267), GPI-anchor (yellow lines) predicted by GPI-SOM and transmembrane segments (cylinders) predicted by TMHMM2.
Signal peptides and signal anchors present in the precursor proteins were predicted by SignalP. ARCα has a predicted cleavable signal peptide attached via a GPI-anchor like CD157 [16] and the Schistosoma ARC [17]; ARCβ also has a predicted cleavable signal peptide, bound through a transmembrane segment close to the C-terminal tail that extends into the cytoplasm; ARCγ has a predicted signal anchor, transmembrane sequence at the N terminus with an exoplasmic C terminus, in a similar manner to CD38 [16] Species abbreviations: Ac (Aplysia californica), Ak (Aplysia kurodai), Bt (Bos taurus), Cf

Figure S4 Model of receptor coupling mechanisms to ARC
In the unstimulated (basal) cell, ARC is sequestered in the lumen of an acidic organelle and is effectively inactive (e.g. limited by substrate availability). Extracellular stimuli trigger nucleotide transport across the vesicle membrane and cADPR synthesized within the lumen is translocated to the cytoplasmic ryanodine receptors to evoke Ca 2+ release (the lower ARC symbol is 'greyed out' for clarity). ARC activity may be terminated by multiple mechanisms: fusion of the exocytotic vesicle with the plasma membrane (PM) elevates the luminal pH to inhibitory levels; exocytosis dilutes substrate/product, extrudes some ARCβ from the cell (perhaps in concert with proteolytic clipping); nucleotide transport may also be inhibited. It is unlikely that β-NAD + transporters and ARC are simultaneously, fully active under basal conditions since this would deleteriously consume cytosolic pools of β-NAD + . Our model does not differentiate between regulation at the level of the β-NAD + or cADPR transport.

Figure S5 Dipyridamole loading of intact eggs
The weak UV fluorescence of dipyridamole was exploited as an index of its time course of loading into intact eggs. Unlabeled eggs were mounted on a confocal microscope equipped with a UV laser and 30 μM dipyridamole (DPM) added to the extracellular medium at the time indicated and the intracellular fluorescence recorded until a plateau was attained. At equilibrium, the intracellular concentration of DPM is, at most, the same as the extracellular concentration. A 3-min preincubation (cf. Fig. 4) gives an upper estimate for the intracellular DPM concentration as 11.3 ± 1.8% of the bath concentration.

Figure S8
Comparison of the effect of inhibitors of phosphodiesterase, cyclooxygenase and nucleotide transport upon fertilization-induced Ca 2+ responses.