Presynaptic factors in the regulation of DSI expression in hippocampus

Abstract

We studied the mechanisms by which GABA release is reduced in the retrograde signaling process called depolarization-induced suppression of inhibition (DSI). DSI is mediated by endocannabinoids in acute and cultured organotypic hippocampal slices. We examined a variety of K⁺ channel antagonists to determine the nature of the K⁺ channel that, when blocked, reduces DSI. Among 4-AP, TEA, dendrotoxin, Cs, margatoxin, and charybdotoxin, only 4-AP was highly effective in blocking DSI, suggesting that a K⁺ channel composed in part of K_V1.4, K_V1.5 or K_V1.7 subunits can readily regulate DSI. The inhibition of DSI by 4-AP is largely overcome by reducing [Ca²⁺]_o, however, suggesting that DSI expression can be prevented by saturation of the release process when a K_V1.X channel is inhibited. DSI of agatoxin- and TTX-insensitive mIPSCs was unaffected by 4-AP, but was largely occluded by ω-conotoxin GVIA, indicating that block of presynaptic N-type Ca²⁺ channels is probably a major mechanism of DSI expression. Significant DSI of mIPSCs remained in ω-conotoxin, hence we infer that block of N-channels does not fully explain hippocampal DSI expression.

Introduction

Regulation of GABAergic inhibition is a critical factor in the control of neuronal excitability in the central nervous system. An intriguing form of GABA inhibitory postsynaptic potential (IPSP) regulation is the retrograde signaling process called “depolarization-induced suppression of inhibition” DSI, (Alger and Pitler, 1995), which is present in hippocampus (Pitler and Alger, 1992, Alger et al., 1996, Morishita and Alger, 1997), cerebellum (Vincent et al., 1992, Vincent and Marty, 1993, Llano et al., 1991) and neocortex (Zilberter, 2000). DSI is mediated by the release of an endogenous cannabinoid (endocannabinoid) from postsynaptic target cells in the hippocampus (Ohno-Shosaku et al., 2001, Wilson et al., 2001) and cerebellum (Kreitzer and Regehr, 2001b, Diana et al., 2002). The type I cannabinoid receptor, CB1R, is the major cannabinoid receptor in the brain (Pertwee, 1997; Ameri, 1999). Once released, the endocannabinoid activates CB1Rs on presynaptic GABAergic interneurons near their terminal regions (Katona et al., 1999, Hajos et al., 2000, Tsou et al., 1999), and thereby transiently suppresses GABA release, but the mechanism by which release is suppressed by CB1R activation is not yet fully understood. In this paper, we address some unresolved issues regarding the expression mechanism of DSI in the hippocampal CA1 region. DSI represents the first well-defined cellular process mediated by endocannabinoids, and DSI and related processes (e.g., DSE, Kreitzer and Regehr, 2001a, Maejima et al., 2001) no doubt significantly influence neuronal network activity.

Vincent and Marty (1993) recorded simultaneously from two neighboring cerebellar Purkinje cells, and reported that cerebellar DSI can be manifested in two distinct ways. The induction of DSI in one cell, the primary cell, simultaneously induced DSI in a nearby, secondary cell, even though the secondary cell was not given the DSI induction protocol. Bath application of TTX prevented expression of DSI in the secondary cell, but not in the primary cell. Thus, one expression mechanism of cerebellar DSI involves a TTX-sensitive process. Even in TTX, however, DSI continued to be manifested in the primary cell, demonstrating that there is also a TTX-insensitive DSI expression mechanism, and suggesting that this process affects the interneuron terminal. It is unclear whether or not there is more than one DSI mechanism in hippocampus.

Wilson et al. (2001) reported that DSI and the actions of a CB1R agonist are mimicked and occluded by blocking N-type Ca²⁺ channels. It is well established that activation of CB1R decreases Ca²⁺ currents through N-type channels (Caulfield and Brown, 1992, Mackie and Hille, 1992), and therefore Wilson et al. (2001) argued that DSI is mediated by a decrease in Ca²⁺ influx into the GABAergic nerve terminals. Voltage-dependent Ca²⁺ influx into cerebellar climbing fibers (Kreitzer and Regehr, 2001a) and basket cell nerve terminals (Diana et al., 2002) is decreased by the activation of CB1Rs. Hence reduction of nerve terminal Ca²⁺ influx appears to be a mechanism of cerebellar DSI expression. Nevertheless, occlusion by ω-conotoxin is not unequivocal evidence for inhibition of N-channels as the expression mechanism of DSI (see Discussion), and we investigated the hypothesis of N-channel involvement in DSI in more detail.

The potassium (K) channel blocker 4-AP, at a low concentration, abolishes DSI (Alger et al., 1996, Morishita et al., 1998) but the identity of the channel responsible, and how 4-AP blocks DSI remain unclear. It could be that 4-AP-sensitive channels are activated during DSI, as activation of CB1Rs also enhances certain K⁺ conductances (Kirby et al., 2000). Because blocking these channels represents a potentially powerful way of regulating DSI, it will be very important to learn their identity. An increase in K conductance could either reduce release by shunting the presynaptic action potential or cause a presynaptic block in action-potential conduction. On the other hand, 4-AP can overcome certain forms of neurotransmitter-induced presynaptic inhibition by increasing Ca²⁺ influx into the terminals to such an extent that the release process becomes “saturated”, and thereby insensitive to the relatively small decreases in Ca²⁺ caused by the presynaptic inhibitory neurotransmitter (Klapstein and Colmers, 1992). Lowering [Ca²⁺]_o reduces the ability of 4-AP to counteract the effects of cannabinoids on hippocampal IPSCs (Hoffman and Lupica, 2000).

Paired-pulse stimulation is often used as an indirect probe of the mechanisms that determine the probability of transmitter release from the nerve terminal (Zucker, 1989). The paired-pulse-evoked inhibitory postsynaptic current (eIPSC) ratio (PPR) reportedly increases during DSI (Wilson and Nicoll, 2001, Ohno-Shosaku et al., 2001). This is a conventional prediction of the use-dependent model of transmitter release, which says that PPR reflects the state of the transmitter release process at the nerve terminal. An increase in PPR appeared to support the conclusion that DSI is expressed mainly at the terminal. Yet a number of studies (Waldeck et al., 2000, Kraushaar and Jonas, 2000, Kim and Alger, 2001) have reported that experimentally decreasing the probability of GABA release, e.g., by activating presynaptic inhibitory receptors with baclofen, or reducing release with Cd²⁺ , does not invariably change PPR, contrary to expectations of the use-dependent model. Thus, paired-pulse plasticity of eIPSCs may not reflect a use-dependent process. Moreover, the method often used to calculate PPR can yield artifactual results (Kim and Alger, 2001). We have re-examined this issue in order to try to resolve the discrepancies in the reports on the behavior of the PPR in DSI because of its potential for illuminating the expression mechanism of DSI.

We investigated the regulation of DSI expression in the rat hippocampal CA1 region to explore these issues.