One of the major limitations in the use of genetically modified mice for studying cognitive functions is the lack of regional and temporal control of gene function. To overcome these limitations, a forebrain-specific promoter was combined with the tetracycline transactivator system to achieve both regional and temporal control of transgene expression. Expression of an activated calcium-independent form of calcium-calmodulin-dependent kinase II (CaMKII) resulted in a loss of hippocampal long-term potentiation in response to 10-hertz stimulation and a deficit in spatial memory, a form of explicit memory. Suppression of transgene expression reversed both the physiological and the memory deficit. When the transgene was expressed at high levels in the lateral amygdala and the striatum but not other forebrain structures, there was a deficit in fear conditioning, an implicit memory task, that also was reversible. Thus, the CaMKII signaling pathway is critical for both explicit and implicit memory storage, in a manner that is independent of its potential role in development.
Get full access to this article

View all available purchase options and get full access to this article.

Already a Subscriber?


Scoville W. B., Milner B., J. Neurol. Neurosurg. Psychiatry 20, 11 (1957);
Squire L. R., Psychol. Rev. 99, 195 (1992).
Schacter D. L., Tulving E., Memory SystemsMIT PressCambridge, MA1994.
Blanchard D. C., Blanchard R. J., J. Comp. Physiol. Psychol. 81, 281 (1972);
Davis M., in The Amygdala: Neurobiologial Aspects of Emotion, Memory, and Mental Dysfunction,, Aggleton J. P., Ed. (Wiley-Liss, New York, 1992), p. 255;
LeDoux J. E., ibid., p 339.
Grant S. G. N., et al., Science 258, 1903 (1992).
Silva A. J., Paylor R., Wehner J. M., Tonegawa S., ibid. 257, 206 (1992).
Silva A. J., Stevens C. F., Tonegawa S., Wang Y., ibid.: 201.
Mayford M., Wang J., Kandel E. R., O'Dell T. J., Cell 81, 891 (1995).
Bach M. E., Hawkins R. D., Osman M., Kandel E. R., Mayford M., ibid.: 905.
Mayford M., Abel T., Kandel E. R., Curr. Opin. Neurobiol. 5, 141 (1995).
Gossen M., Bujard H., Proc. Natl. Acad. Sci. U.S.A. 89, 5547 (1992).
Furth P. A., et al., ibid. 91, 9302 (1994).
Miller S. G., Kennedy M. B., Cell 44, 861 (1986).
Burgin K. E., et al., J. Neurosci. 10, 1788 (1990).
Hanson P. I., Schulman H., Annu. Rev. Biochem. 61, 559 (1992).
Fong Y.-L., Taylor W. L., Means A. R., Soderling T. R., J. Biol. Chem. 264, 16759 (1989).
Thiel G., Czernik A. J., Gorelick F., Nairn A. C., Greengard P., Proc. Natl. Acad. Sci. U.S.A. 85, 6337 (1988).
Waldmann R., Hanson P. I., Schulman H., Biochemistry 29, 1679 (1990).
Lisman J., Trends Neurosci. 17, 406 (1994).
Bland B. H., Prog. Neurobiol. 26, 1 (1986).
The CaMKIIα promoter consisted of 8.5 kb of genomic DNA upstream of the transcription initiation site of the mouse CaMKIIα gene, as well as 84 base pairs of the 5′ noncoding exon. Genomic DNA was isolated from a C57 B16/J mouse spleen cosmid library with a rat genomic probe consisting of a 0.4-kb Ava I fragment comprising the transcription-initiation region of rat CaMKIIα (40). The tTA gene from plasmid pUHD 15-1 (10) was flanked by an artificial intron and splice sites at the 5′ end (41) and by a polyadenylation signal from SV40 at the 3′ end. The cDNA with intron and polyadenylation signal was placed downstream of the 8.5-kb CaMKII promoter fragment. The cDNAs for Escherichia coli lacZ and mouse CaMKIIα were similarly flanked by the hybrid intron and polyadenylation signal and placed downstream of the tet-O promoter element of plasmid pUHD 10-3 (10). The CaMKIIα gene was a full-length cDNA (4.8 kb) isolated from a C57B16/J mouse brain cDNA library. The lacZ gene carried an SV40 large T antigen nuclear localization signal as well as the 3′ untranslated region (UTR) of CaMKIIα, which targets the mRNA to dendrites (42).
RT-PCR was performed essentially as described (7). Total forebrain RNA (100 ng) was used in each reaction with oligonucleotide primers to amplify a region of the transcript that includes the Thr286→Asp mutation. Equal amounts of amplified cDNA (both wild-type and mutant sequences) were separated on a 3% agarose gel, transferred to nylon membranes, and hybridized with a 32P-labeled oligonucleotide probe specific for the Asp286 mutation (oligonucleotide sequence: 5′-CTTCAGGCAGTCGACGTCCTCCTGTCTGTG-3′). Blots were washed under conditions in which only the Asp286 mutant cDNA was detected (2′ 15 min, 60°C, 0.2′ standard saline citrate). A Northern (RNA) blot of total forebrain mRNA revealed expression of a shorter-than-expected CaMKII-Asp286 transcript (∼3.4 kb). As shown in Fig. 3, this shorter CaMKII-Asp286 transcript did not localize to dendrites, presumably as a result of the loss of a sequence element in the 3′ UTR that is necessary for mRNA targeting to dendrites (42).
Transverse slices (400 μm thick) of mouse hippocampus were prepared and placed in an interface slice chamber perfused with artificial cerebrospinal fluid as described (7). Field excitatory postsynaptic potentials (EPSPs) were elicited once per minute with fine tungsten bipolar stimulation electrodes (0.05-ms pulse duration). Stainless steel recording electrodes were placed in striatum radiatum. The stimulation strength was set to produce 50% of the maximum obtainable EPSP in each slice. Baseline synaptic response was collected for 20 min before the tetanus. The 10-Hz tetanus was delivered for 1.5 min at the same intensity as used in the baseline recording. For doxycycline treatment, animals were administered doxycycline (1 mg/ml) plus 5% sucrose in the drinking water for 2 to 3 weeks, and the slices were the exposed to doxycycline (1 ng/ml) in the perfusate. All animals were 2.5 to 6 months of age at the time of recording.
It would also be useful to suppress transgene expression during development and then activate the gene only in the adult animal. However, we found that treatment of wild-type mice with doxycycline (1 mg/ml) during development impaired adult spatial memory and memory for fear conditioning. This result suggests that doxycycline itself produces a defect in neuronal development. We therefore used transgene suppression only in the adult animal in which the doxycycline treatment did not affect memory. Given the activation of the transgene throughout development, it is possible that the LTP and memory phenotypes observed with the transgene active in the adult animal result from a synergistic interaction between developmental and adult expression rather than a direct acute effect of transgene expression in the adult animal.
On the Barnes circular maze (8), the mice (2.5 to 6 months of age) were tested once a day until they met the criterion (five out of six sessions with three or fewer errors) or until 40 days had elapsed. The order of holes searched was recorded by an observer who was blind to genotype and doxycycline condition, and from these data the number of errors was determined. Errors were defined as searches of any hole that did not have the tunnel beneath it. Searches included nose pokes and head deflections over the hole. At the end of each session the search strategy used was recorded by the observer. The spatial search strategy was operationally defined as reaching the escape tunnel with both error and distance scores ≤3. Distance was calculated by counting the number of holes between the first hole searched within a session and the escape tunnel. A one-factor analysis of variance (ANOVA) (gender) revealed no significant effect of gender for either transgenic or wild-type mice, so the data were collapsed across this variable. For the error data, a three-factor ANOVA (genotype, doxycycline, and session block) with one repeated measure was used. For the spatial search strategy data, the two groups of B22 transgenic mice were compared with a two-way ANOVA (doxycycline and session block) with one repeated measure.
In the conditioned fear task (8), freezing was defined as a total lack of movement with the exception of respiration and was measured by an experimenter who was blind to genotype and doxycycline condition. The percentage of time spent freezing to context and cue was calculated. No significant effect of gender was observed in the B22 or B21 transgenic mice or the wild-type mice, so the data were collapsed across this variable. Freezing to context and cue on testing day was analyzed by two three-factor ANOVAs (genotype, line, and doxycycline) that were used to compare the B22 and B21 transgenic and wild-type mice. Two one-way ANOVAs were used to compare the amount of freezing 6 weeks later to cue and context in B21 transgenics on doxycycline, B21 transgenics switched to water, and wild-type mice.
Kim J. J., Fanselow M. S., Science 256, 675 (1992).
LeDoux J. E., Cicchetti P., Xagoravis A., Romanski L. M., J. Neurosci. 10, 1062 (1990).
Kita H., Kitai S. T., J. Comp. Neurol. 298, 40 (1990);
Canteras N. S., Somerly R. B., Swanson L. W., ibid. 360, 213 (1995).
Pain sensitivity was measured in B21 transgenic (n = 4) and wild-type mice (n = 6). The mice were placed individually in a mouse operant chamber with a metal grid floor and given 1-s foot shocks of increasing intensity (for example, 1, 2, 3, 4 mA …). An experimenter who was blind to the genotype of the mice recorded the intensity of foot shock required to elicit each of the following three responses: startles, vocalizations, and jumps. A t test for each response revealed no significant effect of genotype.
Unconditioned freezing in the presence of an intruder was measured in B21 transgenic (n = 8) and wild-type mice (n = 10) in a Nalgene plastic metabolism cage. The mice and intruder were placed in the upper and lower chambers, respectively. The chambers were separated by a metal grid floor. A seven-week-old male Sprague-Dawley rat served as the intruder and was placed in the lower chamber 10 min before introduction of the mouse. The amount of unconditioned freezing occurring during the first 120 s after the mouse was introduced was measured by an experimenter who was blind to genotype. A t test revealed no significant effect of genotype.
Cummings J. A., Mulkey R. M., Nicoll R. A., Malenka R. C., Neuron 16, 825 (1996).
Neveu D., Zucker R. S., ibid.: 619.
Meyer T., Hanson P. I., Stryer L., Schulman H., Science 256, 1199 (1992).
Pettit D. L., Perlman S., Malinow R., ibid. 266, 1881 (1994);
Lledo P.-M., et al., Proc. Natl. Acad. Sci. U.S.A. 92, 11175 (1995).
The expression of the CaMKII-Asp286 transgene in the CA1 region of the B22 mice was patchy; that is, some neurons expressed the transgene well, whereas in other neurons expression was absent. This patchy expression precluded an assessment of LTP in this line of mice by means of field recordings, which sample many synapses from different neurons in a region. However, it is assumed that in those neurons where the transgene was strongly expressed in these mice, a shift in the LTP/LTD frequency threshold would occur.
O'Keefe J., Nadel L., The Hippocampus as a Cognitive MapOxford Univ. PressNew York1978
Quirk G. J., Repa C., LeDoux J. E., Neuron 15, 1029 (1995).
Rogan M. T., LeDoux J. E., ibid.: 127.
Churchland P. S., Sejnowski T. J., The Computational BrainMIT PressCambridge, MA1992.
Sunyer T., Sahyoun N., Proc. Natl. Acad. Sci. U.S.A. 87, 278 (1990).
Choi T., Huang M., Gorman C., Jaenisch R., Mol. Cell. Biol. 11, 3070 (1991).
Mayford M., Baranes D., Podsypanina K., Kandel E. R., Proc. Natl. Acad. Sci. U.S.A. 93, 13250 (1996).
We thank R. Axel and T. Jessell for critically reading this manuscript; J. Finkelstein for maintaining and genotyping the mice; R. Shih, V. Winder, and L. Varshavsky for help with behavioral experiments; C. Lam for help with figures; H. Ayers and I. Trumpet for typing the manuscript; and M. Osman for animal care. This research was supported by the Howard Hughes Medical Institute and the National Institute of Mental Health.

Information & Authors


Published In

Volume 274Issue 52936 December 1996
Pages: 1678 - 1683


Received: 8 August 1996
Accepted: 23 October 1996
Published online: 6 December 1996


Request permissions for this article.



Mark Mayford
The authors are at the Center for Neurobiology and Behavior, College of Physicians and Surgeons of Columbia University, and Howard Hughes Medical Institute, 722 West 168 Street, New York, NY 10032, USA.
Mary Elizabeth Bach
The authors are at the Center for Neurobiology and Behavior, College of Physicians and Surgeons of Columbia University, and Howard Hughes Medical Institute, 722 West 168 Street, New York, NY 10032, USA.
Yan-You Huang
The authors are at the Center for Neurobiology and Behavior, College of Physicians and Surgeons of Columbia University, and Howard Hughes Medical Institute, 722 West 168 Street, New York, NY 10032, USA.
Lei Wang
The authors are at the Center for Neurobiology and Behavior, College of Physicians and Surgeons of Columbia University, and Howard Hughes Medical Institute, 722 West 168 Street, New York, NY 10032, USA.
Robert D. Hawkins
The authors are at the Center for Neurobiology and Behavior, College of Physicians and Surgeons of Columbia University, and Howard Hughes Medical Institute, 722 West 168 Street, New York, NY 10032, USA.
Eric R. Kandel*
The authors are at the Center for Neurobiology and Behavior, College of Physicians and Surgeons of Columbia University, and Howard Hughes Medical Institute, 722 West 168 Street, New York, NY 10032, USA.


* To whom correspondence should be addressed.

Metrics & Citations



View Options






Information & Authors
Published In
issue cover image
Volume 274|Issue 5293
6 December 1996
Submission history
Received:8 August 1996
Accepted:23 October 1996
Published in print:6 December 1996
Metrics & Citations
Article Usage
Export citation

Select the format you want to export the citation of this publication.

Cited by
  1. PARIS farnesylation prevents neurodegeneration in models of Parkinson’s disease, Science Translational Medicine, 13, 604, (2021)./doi/10.1126/scitranslmed.aax8891
  2. Molecular Mechanism for Age-Related Memory Loss: The Histone-Binding Protein RbAp48, Science Translational Medicine, 5, 200, (200ra115-200ra115), (2021)./doi/10.1126/scitranslmed.3006373
  3. Transgenic Inhibition of Synaptic Transmission Reveals Role of CA3 Output in Hippocampal Learning, Science, 319, 5867, (1260-1264), (2021)./doi/10.1126/science.1151120
  4. Spine-Type-Specific Recruitment of Newly Synthesized AMPA Receptors with Learning, Science, 319, 5866, (1104-1107), (2021)./doi/10.1126/science.1149967
  5. Tau Suppression in a Neurodegenerative Mouse Model Improves Memory Function, Science, 309, 5733, (476-481), (2021)./doi/10.1126/science.1113694
  6. Cognitive Memory: Cellular and Network Machineries and Their Top-Down Control, Science, 306, 5695, (435-440), (2021)./doi/10.1126/science.1101864
  7. The Message and the Messenger: Delivering RNA in Neurons, Science's STKE, 2002, 126, (pe16-pe16), (2021)./doi/10.1126/stke.2002.126.pe16
  8. Conditional Restoration of Hippocampal Synaptic Potentiation in GluR-A-Deficient Mice, Science, 292, 5526, (2501-2504), (2021)./doi/10.1126/science.1059365
  9. NMDA Receptor-Dependent Synaptic Reinforcement as a Crucial Process for Memory Consolidation, Science, 290, 5494, (1170-1174), (2021)./doi/10.1126/science.290.5494.1170
  10. Neuroscience: Breaking Down Scientific Barriers to the Study of Brain and Mind, Science, 290, 5494, (1113-1120), (2021)./doi/10.1126/science.290.5494.1113
Share article link

Share on social media
Get Access
Log in to view the full text

AAAS Log in

AAAS login provides access to Science for AAAS members, and access to other journals in the Science family to users who have purchased individual subscriptions, as well as limited access for those who register for access.

Log in via OpenAthens.
Log in via Shibboleth.
More options

Purchase digital access to this article

Download and print this article for your personal scholarly, research, and educational use.

Purchase this issue in print

Buy a single issue of Science for just $15 USD.

View Options


No eLetters have been published for this article yet.

eLetters is an online forum for ongoing peer review. Submission of eLetters are open to all. eLetters are not edited, proofread, or indexed. Please read our Terms of Service before submitting your own eLetter.