Brain Research, 538 (1991) 347-350 347 Elsevier BRES 24491 A test of the spine resistance hypothesis for LTP expression John Larson and Gary Lynch Center for the Neurobiology of Learning and Memory, University of California, Irvine, CA 92717 (U.S.A.) (Accepted 2 October 1990) Key words: Long-term potentiation; Dendritic spine; N-methyl-D-aspartatereceptor; Quisqualate receptor; Temperature; CA1; Hippocampus Long-term potentiation (LTP) consists of an enhanced response to released transmitter by the quisqualate/AMPA subclass of glutamate receptors with little change in the slower currents generated by the NMDA receptor subclass. Recent computer simulations suggest that a decrease in the resistance of dendritic spines would selectivelyaugment fast synaptic currents and this could produce the pattern of results found with LTP. The present experiments tested this hypothesis by asking whether non-NMDA responses slowed by low temperature to resemble NMDA responses could express LTP.Slow non-NMDA responses recorded at 25 °C did express LTP, indicating that the time courses of NMDA responses cannot explain why they do not express LTE The results, therefore, do not support the hypothesis that spine resistance changes are responsible for the enhanced transmission. Long-term potentiation (LTP) of synaptic transmission is a candidate mechanism for memory storage in the mammalian brain. It is well established that induction of LTP in field CA1 of hippocampus involves activation of NMDA (N-methyl-D-aspartate) receptors 6A° and a rise in calcium concentrations in the postsynaptic region2°'21; however the nature of the enduring substrate responsible for expression of the potentiation remains controversial. Electron microscopic studies have shown that the poten- tiation effect is associated with an increase in certain types of synapses but it is unclear if this reflects synaptogenesis or a transformation of existing con- nections 3'sA7As. LTP is expressed by postsynaptic re- sponses to released transmitter that are mediated by receptors of the quisqualate/AMPA (Q/A) type but not of the NMDA type 12'22'23. This could indicate that LTP is due to a change in the properties of Q/A receptors or their associated ionophores. Receptor binding studies have thus far not provided evidence for changes in either number or affinity of Q/A receptors during LTP expres- sion, but a number of technical factors could prevent detection of such a change 19. An alternative explanation for LTP expression is that a morphological alteration of dendritic spines increases current flow into the dendrite by decreasing the electrical resistance of the spine neck 2'7'13'24'26. The conditions under which spine shape could modify synaptic strength require both that the synaptic conductance be large enough and the spine neck conductance small enough that the synaptic current is attenuated by a large local depolarization in the spine head 2'26. The degree of synaptic current attenuation could then be reduced by decreases in neck resistance as might occur with an increase in neck diameter or a decrease in neck length. Computer simulations of spines also show that slow synaptic currents are much less affected by neck resis- tance changes than are fast currents 26 and this could account for the relative lack of LTP expression by NMDA receptors, since NMDA-mediated responses are slower than Q/A-mediated responses 5'9't2'23. If this were the case, it would be expected that LTP could not be expressed by non-NMDA (Q/A) receptor-mediated re- sponses under conditions in which they have time-courses similar to those generated by NMDA receptors that do not express LTE We tested this hypothesis in the present experiments by slowing Q/A responses with low temper- atures until they resembled NMDA responses at physi- ological temperature. The results indicate that slow Q/A responses can express LTP and thus argue against the spine resistance hypothesis for LTP expression. Experiments were conducted on hippocampal slices maintained at the interface between a perfusion bath and a humidified, oxygen-rich (95% 02/5% CO2) atmo- sphere. Immediately after preparation, slices were placed in the recording chamber at a temperature of either 25 or 35 °C and maintained at the same temperature through- out the experiment. Slices were perfused (1 ml/min) with medium containing (in mM): 124 NaCl, 3 KCI, 1.25 KHEPO4, 2.5 MgSO4, 3.4 CaCI2, 10 D-glucose, 26 NaHCO3, and 3 L-ascorbate. Bipolar stimulation elec- Correspondence: J. Larson, Center for the Neurobiology of Learning and Memory, University of California, Irvine, CA 92717, U.S.A. 0006-8993/91/$03.50 (~) 1991 Elsevier Science Publishers B.V. (Biomedical Division)