03 February 2000
|
03 February 2000
|
|
|
| Nature 403, 549 - 553 (2000) © Macmillan Publishers Ltd. |
|
D. E. SHULZ*, R. SOSNIK†, V. EGO*, S. HAIDARLIU† & E. AHISSAR†
State-dependent learning is a phenomenon in which the retrieval of newly
acquired information is possible only if the subject is in the same sensory
context and physiological state as during the encoding phase1.
In spite of extensive behavioural and pharmacological characterization2, no cellular counterpart of this phenomenon has been reported. Here
we describe a neuronal analogue of state-dependent learning in which cortical
neurons show an acetylcholine-dependent expression of an acetylcholine-induced
functional plasticity. This was demonstrated on neurons of rat somatosensory
'barrel' cortex, whose tunings to the temporal frequency of whisker
deflections were modified by cellular conditioning. Pairing whisker stimulation
with acetylcholine applied iontophoretically yielded selective lasting modification
of responses, the expression of which depended on the presence of exogenous
acetylcholine. Administration of acetylcholine during testing revealed frequency-specific
changes in response that were not expressed when tested without acetylcholine
or when the muscarinic antagonist, atropine, was applied concomitantly. Our
results suggest that both acquisition and recall can be controlled by the
cortical release of acetylcholine.
The ascending cholinergic system3 has long been considered
to be a candidate for mediating behavioural control of neuronal plasticity4-9. This hypothesis is supported by behavioural and
neurophysiological studies in the auditory10-13 and
somatosensory systems14-18. Whereas these studies
demonstrated the permissive role of acetylcholine (ACh) during the induction
of cortical plasticity10-14, they did not address
the possibility that ACh is also involved in the expression of the induced
modifications. To examine this potential role of ACh, single- (n =
99) and multi-unit (n = 85) activities were recorded extracellularly
from the barrel field19 of anaesthetized adult rats, using a
multi-electrode array composed of one or two tungsten-in-glass electrodes
and one combined electrode for recording and iontophoresis of ACh. Temporal-frequency
tuning curves (TFTCs) were obtained by mechanically deflecting the principal
vibrissa at four different frequencies between 2 and 11 Hz (in a few cases
14 Hz was also applied), covering the frequency range predominantly used by
the animal while exploring its environment20. Typically, TFTCs
of barrel cortex neurons show decreased spike counts and increased latencies
with increasing frequencies21. The TFTC was determined again
during ACh iontophoresis, and then pairing occurred. Pairing consisted of
repetitive whisker deflection at one fixed frequency (5, 8 or 11 Hz) accompanied
by ACh iontophoresis. After pairing, the TFTC was determined, once without
ACh and once during application of ACh, thereby restoring the physiological
conditions under which the pairing was carried out. Pairing caused frequency-specific modification of the TFTCs that was expressed
exclusively under ACh application. Three examples of significant increments
in response, induced by pairing and revealed with ACh, are depicted in Fig. 1 (Fig. 1a–c, two-tailed
Kolmogorov–Smirnov, P < 0.0005 compared with control). The principal
whisker for each cell was stimulated at 5 (Fig. 1a),
8 (Fig. 1b) or 11 Hz (Fig. 1c)
during pairing. The potentiation of the response was maximal for the conditioned
frequency in each case and affected both the phasic (due to each whisker deflection)
and tonic (due to the entire train of deflections) components of the response.
The summed response (phasic + tonic per stimulus cycle) is referred to herein
as 'responsiveness'. No change was expressed when the responses
were recorded without ACh (Fig. 1a–c, Kolmogorov–Smirnov
test, P > 0.1). The modifications could be reversed by a second pairing
using a different stimulation frequency. A new frequency-specific enhancement
in response after a second pairing is shown in Fig. 1d.
Usually, the second pairing resulted also in a significant reduction of response
to the initially paired frequency (11 Hz, Fig. 1d) (see Supplementary Information). When the extinction of the effect
was analysed by repeatedly testing the TFTC without and with ACh, the response
modification was still statistically significant at least 45 min after the
pairing but only under ACh (four out of four cells, two-tailed Kolmogorov–Smirnov,
P < 0.0005). The consistent lack of expression without ACh, during
periods that were interleaved with successful expression with ACh, excludes
the possibility that the observed effects correspond to a delayed expression
of cholinergic-induced plasticity22 (see Supplementary
Information).
Figure 1 Plasticity of cortical responses expressed during ACh application in
four different units.
Full legend
High resolution image and legend (59k)
Overall, 33% (39 out of 119) of the single- and multi-units recorded with
the combined electrode showed a statistically significant TFTC modification
when tested with ACh after pairing. In contrast, when measured without ACh,
TFTC changes were observed in fewer cases (21%, 25 out of 119; 
2 test, P < 0.05). This difference was also valid for single
units: 30% showed a modified TFTC (17 out of 57) when tested with ACh, whereas
only 14% (8 out of 57) were modified when measured without ACh (
2 test, P < 0.05). Most of the changes expressed with ACh
were response potentiations specific to the paired frequency (76%). Consistent
with studies in the auditory cortex10, most of the effects observed
during testing without ACh were decreases in responsiveness (87.5%). The overlap
between the populations of cells displaying significant TFTC changes after
pairing with and without ACh was small: only 3 out of the 22 modified single
units showed effects in both conditions.
We tested whether ACh application is required during pairing in order to
induce TFTC changes by repeating the pairing protocol without ACh. In most
cases (17 out of 19), whisker stimulation without ACh induced no changes in
the TFTC. Additional evidence supporting the permissive role of ACh in the
induced plasticity was obtained from cells recorded simultaneously with a
tungsten-in-glass electrode other than the combined electrode used for iontophoresis.
Only 7% of the units (3 out of 45) recorded with tungsten-in-glass electrodes
(that is, units that were probably beyond the maximal distance from the ejection
site where ACh is still effective in modifying neuronal firing activity (S.
H. et al., unpublished data)) were modified after pairing when tested
with ACh, even though they were activated by the stimulus. In contrast, 33%
(39 out of 119) of the cells simultaneously recorded by the combined electrode
were significantly modified (2 test, P < 0-002),
confirming that the modifications are observed preferentially within the region
of ACh application.
Part of the response modifications described here may have resulted from the presence of ACh during the second and fourth TFTCs of the protocol and not from the pairing. To isolate the effect of the fixed-frequency pairing, the population of cells submitted to pairing was compared with a control population for which the four TFTCs were applied with no conditioning period in between (the time interval between the second and the third TFTCs was kept the same as for the original protocol). Responses were quantified by a weighted ratio between the response to stimulation at a given frequency and the averaged response to all other frequencies (see Methods). Figure 2 shows the cumulative distributions of changes in weighted ratio observed in each condition. Consistent with other studies23, 24, and independently of the pairing, the response variability was larger (two-tailed F-test, P < 0-001) when tested with ACh (Fig. 2a, dashed line) than without ACh (Fig. 2a, solid line), whereas the mean was unchanged (two-tailed Mann-Whitney U-test, P > 0.46). However, the introduction of the fixed-frequency pairing induced an additional effect: the relative strength of the response to the paired frequency was significantly potentiated (Fig. 2b, compare solid and dashed curves, one-tailed Mann–Whitney U-test, P < 0-001), whereas the variability was unchanged (two-tailed F -test, P > 0.37). These potentiations were specific for the paired frequency, as the distribution of changes for unpaired frequencies (Fig. 2b, dotted line) was indistinguishable from the control distribution (Fig. 2b, dashed line, two-tailed Mann–Whitney U-test, P > 0.1), and significantly different from the distribution of changes at the paired frequency (Fig. 2b, solid line, two-tailed Mann–Whitney U-test, P < 0.0001).
 |
Figure 2 Statistical analysis of ACh-induced response modifications for all
tested units.
Full legend High resolution image and legend (20k) |
Statistically, the entire population was potentiated by fixed-frequency pairings. However, when each unit was analysed separately, only a subpopulation exhibited significant modifications. We examined the dependency of these modifications on the frequency of the paired stimulus (Fig. 3). The significant potentiations observed during testing with ACh were maximal for the paired frequency (Fig. 3a–c) and differed from changes at other frequencies (Fig. 3d–f; analysis of variance (ANOVA), F(1,43) = 5.45, 6.68 and 6.43 for 5, 8 and 11 Hz, respectively, P < 0.05). On average, the TFTCs' reorganization after pairing was such that paired and unpaired frequencies showed, respectively, relative gains and losses in response (Fig. 3g, right box; ANOVA, F(1,43) = 12.07, P < 0-005). No significant reorganization of the TFTCs was observed when testing without ACh ( Fig. 3g, left box; ANOVA, F(1,43) = 0.45, P > 0.5).
 |
Figure 3 Reorganization of TFTCs expressed with ACh after pairing.
Full legend High resolution image and legend (23k) |
The ACh-dependent expression of the enhancement in response to the paired frequency was blocked by the muscarinic antagonist, atropine. Figure 4 shows an example of a significant frequency-specific potentiation (Fig. 4a, 8 Hz, red line; Kolmogorov–Smirnov, P < 10-7) that is absent when atropine and ACh are iontophoresed together during testing (Fig. 4a, green line; Kolmogorov–Smirnov, P > 0.89). Two minutes after the end of atropine application, a significantly enhanced response to the paired frequency was recovered with ACh ( Fig. 4a, orange line; Kolmogorov–Smirnov, P < 10 -7). Overall, the muscarinic nature of the effect has been confirmed in all the cells showing an ACh-dependent expression of plasticity and tested with atropine (five out of five cells; Fig. 4b).
 |
Figure 4 Atropine blocks the ACh-dependent expression of plasticity.
Full legend High resolution image and legend (25k) |
The temporal response properties of populations of auditory cortical cells can be modified after extensive periods of tone presentations at a given repetition rate paired with stimulation of the nucleus basalis13. We have shown that single units of the somatosensory barrel cortex can show a rapidly induced ACh-dependent plasticity of temporal response properties. Furthermore, we have shown that the expression of ACh-induced modifications is also regulated by increased cortical ACh. The altered responsiveness to a specific stimulus frequency, which was associated with increased ACh levels, was expressed only in the presence of ACh. The requisite for a similarity between the acquisition and the recall conditions is analogous to a "state-dependent learning"1, 2—a phenomenon in which newly acquired information may become available for retrieval only if the endogenous state of the brain and the sensory context present at the time of the original encoding episode are reinstated at the time of testing. In our anaesthetized animals, the increased cholinergic levels were induced by exogenous applications; however, in the awake animal, endogenous activation of the cholinergic system probably provides the required levels of cortical ACh for both memory formation5, 8 and recall25.
Methods
Animal preparation and electrophysiologyExperiments
were carried out on adult Wistar albino rats (300
25 g) obtained from
the Animal Breeding Unit of The Weizmann Institute of Science. Maintenance,
manipulations and surgery were according to institutional animal welfare guidelines.
Experimental procedures were similar to those used previously26, 27.
Briefly, anaesthetized rats (urethane, 1.5 g kg-1) were mounted
in a modified stereotaxic device28 which allows free access
to the somatosensory cortex and to vibrissae. The right postero-medial barrel
subfield was exposed, the dura removed and neural activity recorded with a
multi-electrode array composed of two tungsten-in-glass electrodes and a combined
electrode29 composed of a tungsten core surrounded by six micropipettes.
The pipettes were filled with acetylcholine chloride (1 M, pH 4.5),
atropine sulphate (0.1 M, pH 4.5) and NaCl (3 M) for current balance.
In most cases, the tungsten-in-glass and combined electrodes were lowered
independently into different barrels. Data from units recorded by the combined
electrodes (n = 132) and the tungsten-in-glass electrodes (n
= 52) were analysed separately.
Vibrissae stimulation and protocolWhiskers were stimulated
by a linear electromagnetic vibrator (pulses of 10 ms, 5-ms rise time and
5-ms fall time, 160 µm at 
5 mm from the snout). Temporal
frequency tuning curves (TFTCs) were obtained by deflecting the principal
vibrissa at different frequencies in the following order: 2, 5, 8, 11 and
in a few cases 14 Hz; 45 s interval; (14), 11, 8, 5, 2 Hz, with interblock
intervals of 10 s. Stimuli were applied at each frequency in blocks of 12
consecutive trains of 4 s + 1 s intertrain interval each. Before pairing,
the TFTC was determined first without and then during ACh iontophoresis. Pairing
consisted of 24 trains of stimulation (each of 4 s + 1 s intertrain interval)
of the vibrissa at one fixed temporal frequency (5, 8 or 11 Hz) accompanied
with ACh iontophoresis (20–80 nA). After pairing, the TFTC was determined
without ACh and once again with ACh. In some experiments (n = 16 cells),
two additional TFTCs were determined, one during combined iontophoresis of
ACh and atropine (60 nA) and another during ACh application alone. For 53
cells out of 119 only one pairing was applied. For the other recordings, the
pairing was repeated several times at the same or different frequencies.
Data analysis To keep the initial state comparable among cells, only the first paired frequency was considered for statistical tests. The effect was assessed systematically on the test period immediately after the last pairing at that frequency. The relative strength of the response to a given frequency was quantified by the weighted ratio (WR) = (R f – AvgR) / (Rf + AvgR), where Rf is the response to stimulation at a given frequency (spike count over 60 ms from the stimulus onset) and AvgR is the averaged response to stimulation at all other frequencies. This ratio, which takes values from –1 to +1, was calculated for each of the 24 trains of stimuli, with Rf and AvgR values computed from corresponding trains across frequencies presented during the same TFTC. To assess the effect of conditioning, the 24 values obtained from TFTCs before and after pairing were statistically compared (two-tailed Kolmogorov-Smirnov, significance level P < 0.01). The comparison was performed independently for the TFTCs obtained without and with ACh, for each frequency. To assess the frequency specificity of the effect, cells were grouped as a function of the paired frequency (5, 8 or 11 Hz), and the differences in weighted ratios were averaged across cells (see Fig. 3d–f). This analysis was done on all cells showing a statistically significant change in weighted ratio values for any of the tested frequencies (paired and non-paired), thus avoiding any bias towards the paired frequency. The weighted values were statistically compared using multi-factor ANOVA with repeated measures.
Supplementary information is available on Nature's World-Wide Web site (http://www.nature.com) or as paper copy from the London editorial office of Nature.
Received 8 October 1999;accepted 15 November 1999
| 1. | Izquierdo, I. in Neurobiology of Learning and Memory (eds Lynch, G., McGaugh, J. L. & Weinberger, N. M.) 333-358 (Guilford, New York, 1984). |
| 2. | Gordon, W. C. & Klein, R. L. in Animal Learning and Cognition (ed. Mackintosh, N. J.) 255-279 (Academic, San Diego, 1994). |
| 3. | Mesulam, M. M., Mufson, E. J., Wainer, B. H. & Levey, A. I. Central cholinergic pathways in the rat: an overview based on an alternative nomenclature (Ch1-Ch6). Neuroscience 10, 1185-1201 (1983). Links |
| 4. | Richardson, R. T. & DeLong, M. R. A reappraisal of the functions of the nucleus basalis of Meynert. Trends Neurosci. 11, 264-267 (1988). Links |
| 5. | Singer, W. in Brain Organization and Memory: Cells, Systems, and Circuits (eds McGaugh, J. L., Weinberger, N. M. & Lynch, G.) 211-233 (Oxford Univ. Press, New York, 1990). |
| 6. | Ahissar, E. & Ahissar, M. Plasticity in auditory cortical circuitry. Curr. Opin. Neurobiol. 4, 580-587 (1994). Links |
| 7. | Weinberger, N. M. Dynamic regulation of receptive fields and maps in the adult sensory cortex. Annu. Rev. Neurosci. 18, 129-158 (1995). Links |
| 8. | Dykes, R. Mechanisms controlling neuronal plasticity in somatosensory cortex. Can. J. Physiol. Pharmacol. 75, 535-545 (1997). Links |
| 9. | Edeline, J. M. Learning-induced physiological plasticity in the thalamo-cortical sensory systems: a critical evaluation of receptive field plasticity, map changes and their potential mechanisms. Prog. Neurobiol. 57, 165-224 (1999). Links |
| 10. | Metherate, R. & Weinberger, N. M. Acetylcholine produces stimulus-specific receptive field alterations in cat auditory cortex. Brain Res. 480, 372-377 (1989). Links |
| 11. | Edeline, J. -M., Hars, B., Maho, C. & Hennevin, E. Transient and prolonged facilitation of tone-evoked responses induced by basal forebrain stimulations in the rat auditory cortex. Exp. Brain Res. 97, 373-386 (1994). Links |
| 12. | Kilgard, M. P. & Merzenich, M. M. Cortical map reorganization enabled by nucleus basalis activity. Science 279, 1714-1718 (1998). Links |
| 13. | Kilgard, M. P. & Merzenich, M. M. Plasticity of temporal information processing in the primary auditory cortex. Nature Neurosci. 8, 727-731 (1998). |
| 14. | Rasmusson, D. D. & Dykes, R. W. Long-term enhancement of evoked potentials in cat somatosensory cortex produced by co-activation of the basal forebrain and cutaneous receptors. Exp. Brain Res. 70, 276-286 (1988). Links |
| 15. | Jacobs, S. E. & Juliano, S. L. The impact of basal forebrain lesions on the ability of rats to perform a sensory discrimination task involving barrel cortex. J. Neurosci. 15, 1099-1109 (1995). Links |
| 16. | Baskerville, K. A., Schweitzer, J. B. & Herron, P. Effects of cholinergic depletion on experience-dependent plasticity in the cortex of the rat. Neurosci. 80, 1159-1169 (1997). |
| 17. | Maalouf, M., Miasnikov, A. A. & Dykes, R. W. Blockade of cholinergic receptors in rat barrel cortex prevents long-term changes in the evoked potential during sensory preconditioning. J. Neurophysiol. 80, 529-545 (1998). Links |
| 18. | Sachdev, R. N., Lu, S. M., Wiley, R. G. & Ebner, F. F. Role of the basal forebrain cholinergic projection in somatosensory cortical plasticity. J. Neurophysiol. 79, 3216-3228 (1998). Links |
| 19. | Woolsey, T. A. & Van der Loos, H. The structural organization of layer IV in the somatosensory region (SI) of mouse cerebral cortex. The description of a cortical field composed of discrete cytoarchitectonic units. Brain Res. 17, 205-242 (1970). Links |
| 20. | Carvel, G. E. & Simons, D. J. Biometric analyses of vibrissal tactile discrimination in the rat. J. Neurosci. 10, 2638-2648 (1990). Links |
| 21. | Ahissar, E., Haidarliu, S. & Zacksenhouse, M. Decoding temporally encoded sensory input by cortical oscillations and thalamic phase comparators. Proc. Natl Acad. Sci. USA 94, 11633-11638 (1997). Links |
| 22. | Webster, H. H. et al. Long-term enhancement of evoked potentials in racoon somatosensory cortex following co-activation of the nucleus basalis of Meynert complex and cutaneous receptors. Brain Res. 545, 292-296 (1991). Links |
| 23. | Howard, M. A. & Simons, D. J. Physiologic effects of nucleus basalis magnocellularis stimulation on rat barrel cortex neurons. Exp. Brain Res. 102, 21-33 (1994). Links |
| 24. | Ahissar, E., Haidarliu, S. & Shulz, D. E. Possible involvement of neuromodulatory systems in cortical Hebbian-like plasticity. J. Physiol. (Paris) 90, 353-360 (1996). Links |
| 25. | Delacour, J., Houcine, O. & Costa, J. C. Evidence for a cholinergic mechanism of learned changes in the responses of barrel field neurons of the awake and undrugged rat. Neuroscience 34, 1-8 (1990). Links |
| 26. | Shulz, D. E., Cohen, S., Haidarliu, S. & Ahissar, E. Differential effects of acetylcholine on neuronal activity and interactions in the auditory cortex of the guinea-pig. Eur. J. Neurosci. 9, 396-409 (1997). Links |
| 27. | Haidarliu, S. & Ahissar, E. Spatial organization of facial vibrissae and cortical barrels in the guinea pig and golden hamster. J. Comp. Neurol. 385, 515-527 (1997). Links |
| 28. | Haidarliu, S. An anatomically adapted, injury-free headholder for guinea pigs. Physiol. Behav. 60, 111-114 (1996). Links |
| 29. | Haidarliu, S., Shulz, D. E. & Ahissar, E. A multi-electrode array for combined microiontophoresis and multiple single-unit recordings. J. Neurosci. Meth. 56, 125-131 (1995). |
Acknowledgements. We thank Y. Dudai, Y. Frégnac, H. Markram, P. Salin and M. Segal for helpful comments, and A. Bagady for help with statistical analysis. V.E. and D.E.S. were supported by the French Embassy in Israel during their visits to the laboratory of E.A. where the experiments were done. This work was supported by PICS CNRS, Ministère des Affaires Etrangères Français, AFIRST, HFSP, US-Israel Binational Science Foundation, Israel and the MINERVA Foundation, Germany.
Nature © Macmillan Publishers Ltd 2000 Registered No. 785998 England.