Academic Background

1970, B.A., Occidental College
1979, Ph.D., Wesleyan University

Stimulus-secretion Coupling at Nerve Terminals

The small size and inaccessibility of most nerve terminals has not, until recently, allowed direct measurement of their individual electrophysiological properties. Thus, the molecular details of how ionic currents control the release of neuroactive substances remain undetermined. We can isolate nerve terminals from the rat posterior pituitary (PP) which respond to depolarization by releasing, in a calcium (Ca²⁺) dependent manner, peptide neurohormones via exocytosis. A combination of patch-clamp, biochemical, imaging, and molecular biological techniques are being used to try to understand how electrical patterns of activity, drugs (such as alcohol), endogenous transmitters (such as ATP), and toxins regulate Ca²⁺ entry and subsequent transmitter release at these nerve terminals.

Ca²⁺ channels
We have characterized "L"- and "N"-like Ca²⁺ channels in these PP terminals, which are quite different from their counterparts on the cell body. Dihydropyridine-Ca²⁺ channel agonists and antagonists, which modulate the L-type activity, have no effect on the "Nt"-type channel which is most susceptible to block by omega (w)-conotoxin GVIA. Recently we have been able to prevent dialysis of the cytoplasm of the terminal by utilizing amphotericin B to perforate nerve terminals through the patch in the pipette. This technical improvement has enabled us to perform hitherto impossible studies on the Ca²⁺ channels in the "intact" nerve terminals, including the testing of synthetic and mutated toxins. This has led to the positive identification of a third or "Q" type of Ca²⁺ channel in AVP-releasing terminals. Funnel-Web spider toxins and w-conotoxin MVIIC block this Ca²⁺-current component. Most recently, we have identified a fourth or "R" type of Ca²⁺ channel in OT-releasing terminals.

Localization
Fluorescently-tagged specific blockers of the L and N type Ca²⁺ channels allowed us to determine their distribution in these terminals (see Fig. 1) using a laser confocal microscope. These preliminary results indicate that it should be feasible not only to localize different types of Ca²⁺ channels in individual PP terminals, but to even do so in relation to possible release sites.

Figure 1. Localization of Ca²⁺ channels. The DM-bodipy labelled dihydropyridine (specific for L-type channel) appeared to be uniformly distributed (A, right panel) in all types of terminals (A, left panel: using transmitted light), and even in endocrine cells. In contrast, the fluorescein-labeled w-Cgtx GVIA probe (specific for N-type) was mainly found in discrete hot spots (B: 4X zoom of two terminals) and only on neurohypophysial terminals. It was also possible, using reflected light scattering (i.e., the "Tyndal effect") to observe the NSG within individual NH terminals (C: 4X zoom of two terminals). These disappear rapidly in response to puffs of high (100 mM) K⁺, perhaps indicating exocytotic release at particular sites.

Electrical activity
We have studied several other currents in order to understand how bursting electrical activity is generated and regulated. We have characterized a novel Ca²⁺ activated K⁺ channel, specifically blocked by the anti-hypertension drug, tetrandrine, in the PP terminals that could play a role in terminating bursts. This Ca²⁺-activated K⁺ channel is activated by intracellular applications of Mg-ATP, apparently via an endogenous kinase. We are currently attempting to identify this kinase and the physiological effector for this "up-regulation". We are also attempting to determine why different bursting patterns of activity are necessary to release vasopressin vs. oxytocin from these nerve terminals.

ATP feedback
Since substances, such as ATP, are co-released from neurosecretory granules (NSG), we investigated whether ATP could actually affect peptide secretion. ATP exhibited a bi-phasic effect, initially potentiating via a P2x2 receptor and then inhibiting via an A₁, receptor vasopressin release. Furthermore, both the Ca²⁺-activated K⁺ and the N-type Ca²⁺ channels are modulated by ATP co-released with the peptides. The endogenous effects of both ATP and its metabolite, Adenosine, could help explain the differential burst effects on release of the neuropeptides.

Ethanol effects
Ingestion of ethanol (EtOH) is known to result in a reduction of plasma arginine-vasopressin (AVP) levels in mammals. Release of AVP from nerve terminals isolated from the rat neurohypophysis was very sensitive to EtOH. Patch clamping of these terminals indicated that both inactivating and long-lasting calcium currents were reduced in EtOH, but that the long-lasting single channel currents were most sensitive. EtOH-induced decreases in plasma AVP levels can be explained by EtOH's inhibition of Ca²⁺- and potentiation ofCa²⁺-activated K⁺-currents in the nerve terminals.

We (in collaboration with S. Treistman) have shown for the first time that EtOH can directly affect specific ion channels involved in a physiological response.

Opioid effects
Opioids interact with receptors on neurons, leading to a variety of effects, e.g., analgesia, euphoria, and diuresis. It is not known, however, whether these effects are at somata and/or synapses in the central nervous system (CNS). The electrical and secretory activities of the hypothalamo-neurohypophysial system (HNS) are affected by both exogenous and endogenous opioids. Furthermore, the HNS develops tolerance and dependence to morphine during chronic administration suggesting that this CNS system is a good model for studying the physiological mechanisms underlying these phenomena. For example, µ-opioid specific effects on oxytocin release can be explained by its targeting of R-type Ca-channels found only on oxytocin terminals.

Calcium sparks in nerve terminals
As a model for exocytosis we have been studying, in collaboration with John Walsh, the role of intracellular Calcium in nerve terminals. We have now shown that Ca-sparks or "syntillas" exist in neurohypophysial terminals of the mouse. Most recently, we have been studying the activation of these ryanodine Ca²⁺ channels by voltage and the identity of the Ca-stores in nerve terminals.

Mechanism of exocytosis

We have been attempting to reconstitute channel-forming proteins from the NSG of rat and bovine PP terminals in order to study their properties. We have observed both an anion and a Ca²⁺ activated cation channel in these membranes. We have hypothesized that the NSG channels may play a direct role in the mechanism underlying exocytosis (see Figure 2). Blockers of this channel also block release of peptides from the permeabilized PP terminals. Most importantly, both the Ca²⁺-activated NSG channel and Ca²⁺-dependent release are inhibited by an antibody directed against the putative Ca²⁺-binding site of synaptophysin, an integral NSG membrane protein. We are now in the process of utilizing a molecular genetic approach to try to determine if and how this vesicular channel is involved in exocytosis. Most recently, we have discovered that the NSG channel is actually the ryanodine receptor and thus could mediate “syntillas” in nerve terminals.

Figure 2. Model of depolarization-secretion coupling: (1) A complex of proteins serve to "dock" the NSG. This complex possibly involves VAMP (synaptobrevin) and synaptotagmin (p65) interacting via SNAP-25 with syntaxin. (2) Synaptotagmin is liberated by the binding ofa-SNAP (a) and then by NSF. According to our data, Calcium (Ca) enters through, at least, three types of Ca channels and elevates its intracellular levels [Ca]. (3) We hypothesize that Ca could then bind to synaptotagmin and relieve its inhibition (-) of the NSG channel that we have shown to likely be synaptophysin. (4) Synaptophysin can interact with certain plasma membrane proteins such as physophilin, and the opening of the apposed channels could then form a gap junction-like "fusion pore" across the two membranes. (5) When the fusion pore is open, extracellular cations would move into the NSG down their concentration and/or electrical gradients. (6) The subsequent osmotic increase would force water to enter the NSG and cause them to swell. (8) Entry of ions, would also disrupt the matrix inside the vesicle and lead to subsequent expulsion of the contents of the NSG, perhaps even through the fusion pore itself.