Activity-Dependent Potentiation of Synaptic Transmission From L30 Inhibitory Interneurons of Aplysia Depends on Residual Presynaptic Ca But Not on Postsynaptic Ca

Fischer, Thomas M., Robert S. Zucker, and Thomas J. Carew. with a time course of milliseconds; augmentation , which Activity-dependent potentiation of synaptic transmission from L30 lasts for seconds to tens of seconds; and posttetanic potentiainhibitory interneurons of Aplysia depends on residual presynaptic tion (PTP), with a time course on the scale of minutes Ca but not on postsynaptic Ca. J. Neurophysiol. 78: 2061– (reviewed in Fisher et al. 1997; Magleby 1987; Zucker 1989, 2071, 1997. Activity-induced short-term synaptic enhancement 1996). A large body of evidence, beginning with the work (STE) is a common property of neurons, one that can endow of Katz and Miledi (1968), suggests that STE depends on neural circuits with the capacity for rapid and flexible information the action of Ca, which enters the nerve terminal during processing. Evidence from a variety of systems indicates that the activation ([Ca]i ) . Support for this hypothesis, generally expression of STE depends largely on the action of residual Ca , known as the ‘‘residual calcium hypothesis’’, derives largely which enters the presynaptic terminal during activity. We have shown previously that a Ca-dependent STE in the inhibitory from three lines of evidence: extracellular calcium is necessynapse between interneurons L30 and L29 in the abdominal gansary for the induction of synaptic enhancement (i.e., Katz glion of Aplysia californica has a functional role in regulating the and Miledi 1968); manipulations that directly increase gain of the siphon withdrawal circuit through facilitated recurrent [Ca]i increase synaptic strength (e.g., Kamiya and Zucker inhibition onto the L29s. In the present paper, we further explore 1994; Zengel et al. 1994); and calcium imaging experiments the role of Ca in L30 STE by examining two basic issues: 1) demonstrate a strong correlation between the accumulation What is the role of residual presynaptic Ca in the maintenance of [Ca]i during activation and synaptic enhancement (e.g., of L30 STE? We examine this question by first inducing STE in Atluri and Regehr 1996; Delaney and Tank 1994; Delaney the L30s then rapidly buffering presynaptic free calcium through et al. 1989; Regehr et al. 1994). the use of the photoactivated Ca chelator diazo-4, which was preloaded into the L30 neurons. Three forms of STE in the L30s However, until the development of photolabile Ca buffwere examined: frequency facilitation (FF), augmentation (AUG), ers capable of producing rapid and controlled manipulations and posttetanic potentiation (PTP). In each case, the activationof [Ca]i (Adams and Tsien 1993; Adams et al. 1989), it induced enhancement of the L30 to L29 synapse was reduced to was difficult to demonstrate directly that the maintenance of preactivation levels at the first test pulse following photolysis of synaptic enhancement depends on the continued action of diazo-4. 2) What is the role of postsynaptic Ca in the induction residual [Ca]i in the synaptic terminal. Using the diazo of L30 STE? We examine whether there is a postsynaptic requireseries of Ca buffers, which undergo a rapid rise in Ca ment of elevated Ca for the induction of L30 STE by first inaffinity with binding kinetics similar to bis-(o-aminojecting the calcium chelator bis-(o-aminophenoxy)-N,N,N *,N *tetraacetic acid (BAPTA) into the postsynaptic cell L29 (at levels phenoxy)-N,N,N *N *-tetraacetic acid (BAPTA) on photolysufficient to block transmitter release from the L29s), to prevent sis by ultraviolet (UV) light, Kamiya and Zucker (1994) any increase in postsynaptic intracellular Ca that may occur durdemonstrated that facilitation, augmentation, and PTP at the ing L30 (presynaptic) activation. We found that BAPTA injection crayfish neuromuscular junction are all eliminated on reducdid not effect either the induction or the time course of FF, AUG, tion of [Ca]i by diazo photolysis. In the present work, we or PTP in the L30s. Taken collectively, our data indicate that all have employed a similar methodology to examine the calforms of STE in the L30s depend on presynaptic free cytosolic cium dependence of STE at a central synapse formed from Ca for their maintenance but do not require the elevation of the L30 inhibitory interneurons onto the L29 excitatory interpostsynaptic Ca for their induction. neurons in the abdominal ganglion of Aplysia californica (Fig. 1) . The L30 to L29 inhibitory synapse exhibits forms