tion, and regeneration: Role of macrophages. Proceedings of the National Academy of Sciences of the United States of America 84:8735-8739, 1987b.

Hosang, M., and Shooter, E.M. The internalization of nerve growth factor by high affinity receptors on pheochromocytoma PC12 cells. EMBO Journal 6:1197-1202, 1987.

Johnson, D.; Lanahan, A.; Buck, C.R.; Sehgal, A.; Morgan, C.; Mercer, E.; Bothwell, M.; and Chao, M. Expression and structure of the human NGF receptor. Cell 47:545-554, 1986.

Korsching, S.; Auburger, G.; Heumann, R.; Scott, J.; and Thoenen, H. Levels of nerve

growth factor and its mRNA in the central nervous systems of the rat correlate with cholinergic innervation. EMBO Journal 4:1389-1393, 1985.

Layer, P., and Shooter, E.M. Binding and degradation of nerve growth factor by PC12 pheochromocytoma cells. Journal of Biological Chemistry 258:3012-3018, 1983. Lindholm, D.; Heumann, R.; Meyer, M.; and Thoenen, H. Interleukin-1 regulates synthesis of nerve growth factor in nonneuronal cells of rat sciatic nerve. Nature 330:658-659, 1987.

Masiakowski, P., and Shooter, E. M. Nerve growth factor induces the genes for two

proteins related to a family of calcium binding proteins in PC12 cells. Proceedings of the National Academy of Sciences of the United States of America 85:1277-1281, 1988.

Millbrandt, J. Nerve growth factor rapidly induces a gene which encodes a possible transcriptional regulatory factor. Science 238:797-799, 1987.

Millbrandt, J. Nerve growth factor induces a gene homologous to the glucocorticoid receptor. Neuron 1:183-188, 1988.

Mobley, W.C.; Rutkowski, J.L.; Tennenkoon, G.I.; Gemski, J.; Buchanan, K.; and Johnston, M.V. Nerve growth factor increases choline acetyltransferase activity in developing basal forebrain neurons. Molecular Brain Research 1:53-62, 1986. Radeke, M.J.; Misko, T.P.; Hsu, C.; Herzenberg, L.A.; and Shooter, E.M. Gene transfer and molecular cloning of the rat nerve growth factor receptor. Nature 325:593-597, 1987.

Rohrer, H.; Heumann, R.; and Theonen, H. The synthesis of nerve growth factor (NGF) in developing skin is independent of innervation. Developmental Biology 128:240-244, 1988.

Schwab, M.E.; Otten, U.; Agid, Y.; and Thoenen, H. Nerve growth factor in the rat CNS: Absence of specific retrograde axonal transport and tyrosine hydroxylase induction in locus coeruleus and substantia nigra. Brain Research 168:473-483, 1979.

Shelton, D.L., and Reichardt, L.F. Studies on the expression of the beta nerve growth factor (NGF) gene in the central nervous system: Level and regional distribution of NGF mRNA suggest that NGF functions as a trophic factor for several distinct populations of neurons. Proceedings of the National Academy of Sciences of the United States of America 83:2714-2718, 1986.

Taniuchi, M.; Clark, H.B.; and Johnson, E.M., Jr. Induction of nerve growth factor receptor in Schwann cells after axotomy. Proceedings of the National Academy of Sciences of the United States of America 83:4094-4098, 1986.

Thoenen, H., and Barde, Y.A. Physiology of nerve growth factor. Physiological Review 60:1284-1334, 1980.

Thoenen, H.; Bandtlow, C.; and Heumann, R. The physiological function of nerve

growth factor in the central nervous system: Comparison with the periphery. Review of Physiological and Biochemical Pharmacology 109:145-178, 1987.

Vale, R.D. Intracellular transport using microtubule-based motors. Annual Review of Cell Biology 3:347-378, 1987.

Will, B., and Hefti, F. Behavioral and neurochemical effects of chronic intraventricular injections of nerve growth factor in adult rats with fimbria lesions. Behavioural Brain Research (Amsterdam) 17:17-24, 1985.

Williams, L.R.; Varon, S.; Peterson, G.M.; Wictorin, K.; Fischer, W.; Bjorklund, A.; and Gage, F.H. Continuous infusion of nerve growth factor prevents basal forebrain neuronal death after fimbria fornix transection. Proceedings of the National Academy of Sciences of the United States of America 83:9231-9235, 1986.

Dual Modes of Excitatory Synaptic Transmission in the Brain

J.M. Bekkers and C.F. Stevens

Section of Molecular Neurobiology, Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, CT 06510

During the years since several distinct types of excitatory amino acid receptor were identified by their pharmacological properties (Watkins and Evans 1981), further studies have revealed that the associated channels fall into distinct functional classes (Mayer and Westbrook 1987). Although the situation will doubtless become more complicated when genetically defined gene families of synaptic receptors have been elucidated, many workers now identify two main kinds of excitatory amino acid receptors: the NMDA (N-methyl-D-aspartate) type and the non-NMDA type, here denoted the QUIS (quisqualate) class of receptor. Our goal is to define functional differences between the channels associated with these receptor types and explore some consequences for the operation of neuronal circuits. We conclude that the two receptor types subserve two quite distinct information processing modes in the brain.

Jahr and Stevens (1987) described two main classes of single-channel currents in cultured hippocampal pyramidal cells that they identified with the NMDA and QUIS pharmacological types: a large (50 pS conductance) channel class corresponds to the NMDA type, whereas a small (20 pS conductance) channel class corresponds to the QUIS pharmacological category. NMDA is the most effective at causing the large channel class to open, and quisqualate (or kainate) is best for opening the small channels, but all glutamate-related agonists can produce at least some openings of both channel types. Although the pharmacological properties of these two types of channels overlap somewhat, their functional properties are nonetheless distinct. Both channel types are normally activated at excitatory synapses by glutamate, and both are permeable to sodium and potassium ions. In three important regards, however, the properties of the two channel classes differ: (1) the QUIS type opens only briefly, for several milliseconds, whereas the NMDA type opens in bursts that are sometimes many hundreds of milliseconds long (Nowak et al. 1984; Cull-Candy and Usowicz 1987; Jahr and

Stevens 1987); (2) the QUIS type is not measurably permeant to calcium ions, whereas the NMDA type is highly permeant to this ionic species (MacDermott et al. 1986; Jahr and Stevens 1987); and (3) the conductance change produced by the QUIS-type channel is independent of membrane potential, whereas the NMDA-type channel has, by virtue of a voltage-dependent magnesium ion blocking mechanism, a conductance strongly influenced by voltage (MacDonald et al. 1982; Nowak et al. 1984; Mayer et al. 1984). These three differences, together with the distribution of the channel types in synaptic membranes, have profound implications for brain information processing.

Methods

Neonatal rat hippocampal neurons from the CA1 field or the CA1 and CA3 fields combined were maintained for several weeks in culture (see methods described in Jahr and Stevens 1987) and then studied with paired whole cell recordings. Typically, one cell (presynaptic) was stimulated to evoke an action potential in its synaptic terminals, and the resultant postsynaptic currents were recorded, under voltage clamp, from the second (postsynaptic) neuron. Electrophysiological studies were done at room temperature.

Results

Excitatory synaptic conductances have two functionally distinct components. Synaptic currents with two distinct phases were found (figure 1), as would have been anticipated from earlier studies (Dale and Roberts 1985; Thomson 1986; Dale and Grillner 1986; Forsythe and Westbrook 1988). The first phase increased rapidly to a peak and decayed with a time constant of around 4 ms, and the second phase increased more slowly (with about the same time course as the first component's decay) and then declined about two orders of magnitude more slowly than the first component; the decay time constant averaged 400 ms (figure 2).

The first component arises from QUIS channels, and the second from NMDA channels. This conclusion is based on a detailed comparison of single channel properties with synaptic conductance properties and involves pharmacological manipulations (the first component is unaffected by APV whereas the second component is completely blocked) and investigations of ion permeability (both components have a reversal potential of nearly zero millivolts in normal solutions, and the slow component reversal potential is shifted appropriately by changes in extracellular calcium concentration) and the effects of voltage on gating for various extracellular magnesium con

2+

Vh (mV)

Figure 1 Family of excitatory postsynaptic currents (epscs) recorded under voltage clamp, showing fast and slow components. A. Currents at V1=± 80 mV, ± 60 mV, +40 mV, each trace an average of 2-3 sweeps. The external solution was mammalian Ringers with 0.1 mM Mg, 1 μM glycine, 1 μM strychnine, and 50 μM picrotoxin; the patch pipette contained CsCl. B. Normalized conductances of the slow components in panel A, obtained from fits of a single exponential to the slow part of each trace. The smooth curve is fitted by eye. Note the reduction in conductance at hyperpolarised potentials due to Mg block.

2+