articles NATURE CELL BIOLOGY VOL 5 FEBRUARY 2003 www.nature.com/naturecellbiology 109 The signal flow and motor response controling chemotaxis of sea urchin sperm U. Benjamin Kaupp*†‡, Johannes Solzin*†, Eilo Hildebrand*, Joel E. Brown†§, Annika Helbig*†, Volker Hagen¶, Michael Beyermann¶, Francesco Pampaloni* and Ingo Weyand*† *Institut für Biologische Informationsverarbeitung, Forschungszentrum Jülich, 52425 Jülich, Germany †Marine Biological Laboratory, Woods Hole, MA 02543, USA §Department of Ophthalmology, Albert Einstein College of Medicine, New York, NY 10461 USA ¶Forschungsinstitut für Molekulare Pharmakologie, 13125 Berlin, Germany ‡e-mail: a.eckert@fz-juelich.de Published online: 31 January 2003, DOI: 10.1038/915 The signalling pathway and the behavioural strategy underlying chemotaxis of sperm are poorly understood. We have studied the cellular events and motor responses that mediate chemotaxis of sperm from the sea urchin Arbacia punctulata. Here we show that resact, a chemoattractant peptide, initiates a rapid and transient rise in the concentration of cyclic GMP, followed by a transient influx of Ca 2+ . The binding of a single resact molecule elicits a Ca 2+ response, and 50–100 bound molecules saturate the response. The ability to register single molecules is remi- niscent of the single-photon sensitivity of rod photoreceptors. Both resact and cyclic nucleotides cause a turn or brief tumbling in the swimming path of sperm. We conclude that a cGMP-mediated increase in the Ca 2+ concentra- tion induces the primary motor response of sperm to the chemoattractant. C hemotaxis of sperm is well established in marine inverte- brates, most notably in sea urchins, but there is also evidence for chemotaxis in mammals 1–4 . Eggs attract the sperm by dif- fusible factors. Resact, the chemoattractant of the sea urchin A. punctulata, is a peptide that binds to a receptor guanylyl cyclase 5 . Resact belongs to a family of sperm-activating peptides 6 . Stimulation of sperm with peptides evokes changes in intracellular pH, Ca 2+ concentration, cAMP and cGMP concentrations, and membrane potential (for reviews see refs 5, 7, 8). But the sequence of cellular events and their causal relations remain unresolved. In addition, the macroscopic and microscopic behaviour of sperm (their swimming trajectories and flagellar waveforms) in a gradient of the attractant are largely unknown. In previous studies, intracel- lular events and behavioural responses might have been either missed or misinterpreted owing to insufficient time resolution. In addition, peptide concentrations covering almost six orders of magnitude have been used previously, and some recorded respons- es may have reflected adaptational mechanisms rather than the pri- mary reaction. To characterize the resact-induced signals in sperm of A. punc- tulata, we have resolved the kinetics of changes in the intracellular concentrations of cAMP, cGMP and Ca 2+ using rapid mixing tech- niques. We have also developed caged compounds that generate ‘instantaneous’ step increases in the concentrations of resact and cyclic nucleotides. We show that resact elicits a fast rise in cGMP concentration, followed by a transient increase in Ca 2+ concentra- tion. This Ca 2+ response can be triggered by a single molecule of resact. Releasing cGMP from a caged compound elicits a similar Ca 2+ signal. Changes in cAMP concentration are not involved in the initial cellular events. When stimulated by resact or cyclic nucleotides, sperm undergo turns or tumbling episodes, which result from a transient increase in the asymmetry of flagellar beat that requires external Ca 2+ . Our results have implications for the framework of models 8–12 of chemotactic signalling in sperm. Results Resact rapidly increases the cGMP concentration. Unstimulated sperm contained 1.1 ± 0.7 pmol cGMP per 10 8 cells (mean ± s.d.; n = 9). Resact rapidly stimulated a large increase in cGMP concen- tration in a dose-dependent fashion (Fig. 1a, b). The mean peak increase in cGMP concentration for 250 nM resact was about 30- fold. For high resact concentrations, the cGMP concentration rose to a peak within 200 ms and then decayed to lower values that were graded with the stimulus. The resting level of cAMP was 8.3 ± 7.3 pmol cAMP per 10 8 cells (n = 10). The resact-induced cAMP responses were smaller and slower than were the cGMP responses (Fig. 1c). Stimulation with 250 nM resact increased cAMP concentration by about four- fold. After 200 ms, the cAMP concentration had not changed appreciably, whereas the cGMP concentration had risen to a peak (Fig. 1d). At lower concentrations of resact (250 pM to 2.5 nM), the changes in cAMP concentration were variable: the increases were not graded with resact concentrations and varied during the meas- urement (10 s). Only two of nine experiments (one with and one without the phosphodiesterase (PDE) inhibitor isobutylmethylx- anthine; IBMX) showed a statistically significant and steady increase in cAMP concentration. By contrast, the continuous increases in cGMP concentration were graded with the resact con- centration down to 250 pM (Fig. 1b). Thus, cAMP concentration responds less sensitively to resact than does cGMP concentration. Preincubation with IBMX did not significantly alter the resting cGMP concentration or the initial rate of the rise in cGMP con- centration, but either decelerated or completely abolished the decay from the initial peak (Fig. 1e); thereby indicating that the decay is controlled by PDE activity. By contrast, the resting cAMP concen- tration rose by about threefold (23.8 ± 3.5 pmol cAMP per 10 8 cells; n = 3) in the presence of IBMX. Resact (250 nM) further increased cAMP concentration by about sevenfold (data not shown). At all resact concentrations, the increase in cyclic ©2003 Nature Publishing Group