7. B. L. Tempel, Y. N. Jan, L. Y. Jan, Nature 332, 837 (1988). 8. A. Baumann, A. Grupe, A. Ackermann, 0. Pongs, EMBOJ. 7, 2457 (1988). 9. Oocytes (Dumont stage V-VI) were harvested (20) from adult female X. laevis under anesthesia [0.35% MS-222 (3-aminobenzoic acid ethyl ester), Sigma]. Theca and follicular layers were removed by incuba- tion for 3 hours in Ca2'-free and 82.5 mM NaCl ND-96 solution. Denuded oocytes were ijected 1 to 24 hours later with 0.5 to 50 ng of mRNA (in 50 nl). Oocytes were incubated at 18°C for up to 120 hours in ND-96, which is 96 mM NaCl, 2 mM KC1, 1.8 mM CaC12, 1 mM MgC12, 2.5 mM sodium pyruvate, 0.5 mM theophylline, 5 mM Hepes, and gentamycin (50 p.g/ml). 10. Recordings of membrane current were made 24 to 96 hours after injection of mRNA. Oocytes were continuously supafiLsed (3 m/min) with ND-96 (pyruvate, theophylline, and gentamycin omitted) at 230 to 250C and were voltage-damped with two microelectrodes (resistance 0.1 to 1 megohm) by using standard tochniques [A. S. Finkel and P. W. Gage, in Voltage and Patch Clamping with Microelec- trodes, T. G. Smith, H. Lecr, S. J. Redman, P. W. Gage, Eds. (Williams and Wilkins, Baltimore, 1985), pp. 47-94]. Single sweep current traces were low-pass filtered at 3 or 10 kHz before recording and are presented without leak subtraction. Oocytes (89%) displayed large K+ currents 48 to 96 hours after injection of 25 to 100 ng of RBK-1 mRNA (two RNA preparations injected into 45 oocytes from four Xceopus donors). The outward current evoked by stepping from -90 (or -70) to -30 mV was 4.3 ± 0.45 pA (peak) and 3.1 ± 0.33 pA after 1 s (n = 30). Leak currents during these steps did not exceed 0.2 pA. No rapidly developing outward currents were observed in uninjected or water- injected oocytes (n = 44). As reported [N. Dascal, CRC Crit. Rev. Biochem. 22, 317 (1987)], slowly developing outward currents were occasionally ob- served in uninjected oocytes during steps to -20 or 0 mV, but these did not exceed 0.1 gsA and were not affected by tetraethyla4mmonium (1 mM) or 4-AP (3 mM). The peak amplitude of the outward current was strongly related to the amount of mRNA injected. After 48 to 60 hours, the current evoked by stepping from -90 to -30 mV was 0.08 ± 0.06 pA (2 out of 8 positivc) for 500 pg of mRNA, 0.82 ± 0.18 pA (12 out of 14 positive) for 5 ng of mRNA and 1.81 ± 0.37 pA (7 out of 8 positive) for 50 ng of mRNA. 11. For K+ concentrations of 2 to 20 mM, the reversal potential of the tail current was measured by step- ping to 0 mV for 10 ms and then measuring the current amplitude immediately after stepping back to different potentials. For 40, 60, and 80 mM KV, reversal of the current was observed directly. NaCI was reduced when KCI was increased in these eperiments, so the good agreement with the Nernst equation implies that Na+ was not significantly permeable. The current was also unaffected by sub- stituting isethionate for 90% of the C10 ions (n = 3). 12. The time constant for activation was 2.7 ± 0.3 ms at -20 mV, and 6.0 ± 1.3 ms at -40 mV (mean + SEM, n = 5); measurements were inaccurate at 0 and +20 mV because currents may have been contaminated by capacitative transients. At the end of the depolarizing step the current declined with a time constant of 5.0 ± 0.3 ms (2 mM K+, at -70 mV, n = 4). The faster time constants of inactiva- tion during a sustained depolarization were 68.3 ± 8.4 ms at -40 mV, 36.5 ± 1.0 ms at -10 mV, and 43.9 ± 13.0 ms at +10 mV (n = 4); the slower time constants were 11.0 ± 2.3 s at -40 mV, 5.8 ± 1.4 s at -10 mV, and 10.8 ± 3.9 s at +10 mV (n = 4). Recovery from inactivation reached 60% ± 10% (n = 4) within 1 s and was complete betwee 30 and 60 s. 13. J. A. Connor and C. F. Stevens, J. Physiol. (London) 213, 21 (1971); M. A. Rogawski, Trends Neurosci. 8, 214 (1985). 14. J. V. Halliwell, I. B. Odhman, A. Pelchen-Matthews, J. 0. Dolly, Proc. Natl. Acad. Sci. U.S.A. 83, 493 (1986); M. Segal and J. L. Barker, J. Neurophysiol. 51, 1409 (1984); M. Segal, M. A. Rogawski, J. L. Barker, J. Neurosci. 4, 604 (1984). 15. B. Gustaffson, M. Gaivan, P. Grafe, H. A. Wig- strom, Nature 299, 252 (1982). 16. H. Kasai, D. Kamcyama, K. Yamaguchi, J. Fukuda, Biophys. J. 49, 1243 (1986). 17. H. Rehm and M. Lazdunski, Proc. Natl. Acad. Sci. U.S.A. 85, 4919 (1988). 18. M. Noda et al., Nature 312, 121 (1984). 19. L. C. Timpc, Y. N. Jan, L. Y. Jan, Neuron 1, 659 (1988). 20. B. Rudy et al., ibid., p. 649. 21. Supported by Department of Health and Human Services grants DA03160, DA03161, DA04154, DK32979, and MH40416. We thank H. Lcster for help with establishing the oocyte injection methods, C. Miller for providing charybdotoxin, and Y. Wu for technical help. 5 December 1988; accepted 2 February 1989 Inescapable Versus Escapable Shock Modulates Long-Term Potentiation in the Rat Hippocampus TRACEY J. SHORS, THOMAS B. SEIB, SEYMOUR LEVINE, RicHARD F. THOMPSON A group of rats was trained to escape low-intensity shock in a shuttle-box test, while another group of yoked controls could not escape but was exposed to the same amount and regime of shock. After 1 week of training, long-term potentiation (LTP) was measured in vitro in hippocampal slices. Exposure to uncontrollable shock massively impaired LTP relative to exposure to the same amount and regime of controllable shock. These results provide evidence that controllability modulates plasticity at the cellular-neuronal level. E XPOSURE TO INESCAPABLE SHOCK in laboratory animals has been linked to marked changes in endocrine ac- tivity and central nervous system neuro- chemistry (1), suppressed immunological function (2), increased gastric ulceration (3), reduced activity (4), weight loss (5), de- creased aggression and lowered dominance status (6), and analgesia (7). Most of these effects can be ameliorated when the animal can control the aversive event; control is defined as the capacity to make an instru- mental response to an aversive stimulus. Of particular interest are the learning deficits observed after exposure to inescap- able shock (8). These deficits cover a wide range of tasks (9) and are often transferred from one task to another (10). Similar re- gimes have a detrimental effect on LTP in the rat hippocampus (11). LTP is a form of neuronal plasticity characterized by an in- crease in synaptic response to a constant volley after brief tetanic stimulation of affer- cnt fibers (12). Because of its relatively long time course, localization in the hippocam- pus (although not exclusively), and correla- tion with behavioral learning (13-17), LTP has been suggested as a component of asso- ciative memory formation (18). Prior exposure to uncontrollable shock eliminated LTP in the in vitro hippocampal slice preparation (11). To determine wheth- er this effect, like those described above, T. J. Shors, T. B. Seib, R. F. Thompson, Department of Pchology, University of Southern Californa, Los An- eLie, CA 90089. S.Lvin, Deartentof Psychiatry, Stanford Universi- ty, tnor,C 930. could be ameliorated by permitting the ani- mal to exert control, we placed Long-Evans male rats (n = 12), weighing 200 to 250 g, and a second group of yoked controls (n = 12) in identical soundproof shuttle boxes. Boxes were linked to a scrambled- shock generator, and the rats were subjected to low-intensity shock (60 Hz, 1 mA) every minute for 30 min. Yoked controls could not escape, but experimental animals were able to escape by running through an arch- way (8 cm by 8 cm) and tripping a balance switch that shut off the current to the boxes of both groups simultaneously. After seven daily sessions of 30 shock presentations with an intertrial interval (m) of 60 s, the experimental group had mastered the behav- ior to the extent that the duration of each shock had dropped on average from 3.8 to 1.5 s and over 75% of the responses were less than 1.5 s (Fig. 1). Immediately after the seventh morning of training, animals from both groups were killed and hippocampal slices (400 ,um) were prepared (19). Twelve additional Long-Evans males were taken directly from their home cages and killed, and hippocam- pal slices were prepared. Trunk blood was collected from all rats for corticosterone radioimmunoassay (20). Recordings were performed "blind" by the experimenter. Extracellular field potentials were record- ed from the cell body layer of CA1 after pulsed stimulation of the Schaffer collateral branches of CA3 pyramidal cell axons. After a 10-min stability period, input-output functions were obtained. The potential be- fore tetanus was set at one-half the maxi- SCIENCE, VOL. 244 224- on November 22, 2014 www.sciencemag.org Downloaded from on November 22, 2014 www.sciencemag.org Downloaded from on November 22, 2014 www.sciencemag.org Downloaded from