- Email: [email protected]
The effect of autonomous alpha-CaMKII expression on sensory responses and experience-dependent plasticity in mouse barrel cortex
Neuropharmacology 41 (2001) 771–778 www.elsevier.com/locate/neuropharm
The effect of autonomous alpha-CaMKII expression on sensory responses and experience-dependent plasticity in mouse barrel cortex Stanislaw Glazewski a, Rafael Bejar b, Mark Mayford b, Kevin Fox a
a,*
Cardiff School of Biosciences, Cardiff University, Museum Avenue, Cardiff CF10 3US, UK b The Scripps Research Institute, ICDN Room 206, La Jolla, CA USA Received 11 April 2001; received in revised form 20 June 2001; accepted 22 June 2001
Abstract The calcium/calmodulin kinase II (CaMKII) autophosphorylation site is thought to be important for plasticity, learning and memory. If autophosphorylation is prevented by a point mutation (T286A) LTP is blocked in the hippocampus and cortex. Conversely, if the point mutation mimics autophosphorylation (T286D) a range of frequencies that normally produce LTP in wild types cause LTD instead. In order to test whether the αCaMKII-T286D mutation increases levels of depression in vivo, we examined the effect of the αCaMKII-T286D transgene on plasticity induced in the barrel cortex by whisker deprivation. Surprisingly, the mutation did not affect depression or potentiation. However, in animals reared with the transgene turned on from birth, the surround receptive field responses were greater than normal. This effect may be due to the potentiating action of autophosphorylated CaMKII during early development. 2001 Elsevier Science Ltd. All rights reserved. Keywords: Phosphorylation; LTP; LTD; Somatosensory; Neocortex
1. Introduction Calcium/calmodulin kinase II (CaMKII) is a major postsynaptic density protein thought to be involved in synaptic plasticity (Kennedy et al., 1983; Miller and Kennedy, 1985). In general, mutations of the CaMKII gene affect learning, memory, LTP and experiencedependent plasticity. Knockout of the αCaMKII gene or inactivation of the autophosphorylation site by point mutation (T286A), cause abolition of these forms of plasticity (Silva et al., 1992; Glazewski et al. 1996, 2000; Kirkwood et al., 1997; Giese et al., 1998). However, increasing the proportion of calcium-independent αCaMKII with a transgenic form of calcium-independent αCaMKII (T286D) produces a more complex phenotype (Mayford et al., 1995). Substitution of the threonine autophosphorylation site with an aspartate produces a constitutively active CaMKII molecule. This
* Corresponding author. Tel.: +44-2920-874632; fax: +44-2920874744. E-mail address: [email protected] (K. Fox).
mutation might be expected to occlude LTP by saturating it. Instead, the ability to produce LTP is preserved, but the T286D mutation causes a shift in the frequency at which LTP can be induced. A range of frequencies that evoke LTP in a wild type cause LTD, while LTP is only produced at far higher tetanus frequencies than normal. The effect can be summarised as a rightwards shift in the BCM curve in a direction favouring production of LTD (Mayford et al., 1995). The BCM curve describes the relationship between the activity of the cell and the direction of synaptic change (namely, a direction towards potentiation or depression). A crucial aspect of the theory is that the activity threshold for potentiation is itself variable and dependent on the prior history of the neurone’s activity. The BCM theory has been used to explain a number of effects known to occur in experience-dependent plasticity (Bienenstock et al., 1982). For example, it is able to explain the way in which weak inputs potentiate when there are no strong inputs to depolarise the neurone. Such a situation occurs during reverse lid suture for example, where the first eye to be closed drives very few cells in the visual cortex after a few days of deprivation,
0028-3908/01/$ - see front matter 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 8 - 3 9 0 8 ( 0 1 ) 0 0 0 9 7 - 1
772
S. Glazewski et al. / Neuropharmacology 41 (2001) 771–778
and yet, despite the weakness of the synaptic drive from the recently closed eye, it is able to potentiate its inputs on reversing the eye-lid suture (Blakemore and Van Sluyters, 1974; Bienenstock et al., 1982; Clothiaux et al., 1991). The BCM theory is also consistent with a number of other observations in ocular dominance and whisker deprivation plasticity. One of the central features of the theory is the idea that the threshold at which potentiation rather than depression is produced is not fixed but variable (Bienenstock et al., 1982). The observation that the cross over point at which LTP is induced can be altered by a CaMKII mutation is therefore of interest and might be expected to have an effect on experience-dependent plasticity (Mayford et al., 1995; Bear et al., 1987). To date, the relationship between the BCM curve when applied to LTP and the BCM theory applied to experience-dependent plasticity has not been fully investigated. A mutation that does not prevent plasticity but which shifts the BCM curve therefore provides an opportunity to examine the relationship. One might predict that a mutation which sets a higher threshold for LTP and which favours the production of LTD might reduce experience-dependent potentiation and increase the chances of experience-dependent depression. Both potentiation and depression can be measured in mouse barrel cortex by inducing plasticity by whisker deprivation (Glazewski et al., 1996). We therefore induced plasticity in mice expressing the T286D mutated form of αCaMKII and looked at the effect on plasticity of spared and deprived whisker responses. Experiencedependent depression decreases between about 2 and 6 months of age in mice: we looked at depression at an age where depression is waning in order to increase the likelihood of seeing an increase in depression as a result of the mutation should it occur. The results suggest that experience-dependent potentiation and depression are not affected by expression of autonomous αCaMKII. However, we also found that the level of surround receptive field whisker responses are elevated if the transgene is on during early development. 2. Methods 2.1. Subjects 2.1.1. Plasticity experiments Recordings were made from 1123 neurones in layer II/III of the barrel cortex of 22 undeprived (543 neurones) and 24 deprived (580 neurones) mice. The animals were either double (CaM-tet) or single (tet and CaM) transgenics or their wild-type littermates, at the age of 3–4 months. All transgenic animals were kept on doxycycline (40 µg/l) administered via the drinking water, until 1 week before starting deprivation. Recordings were also made from 11 animals at an age
exceeding 6 months (284 neurones recorded). All recordings were conducted blind to genotype. Animals were genotyped by PCR as described before (Mayford et al., 1995). 2.1.2. Time course experiments To check whether introduction of the transgene has an influence on the magnitude of principal and surround responses recorded in the barrel cortex, a time-course experiment was performed. The mice were initially kept on a doxycycline diet until 6 months of age, and were recorded at 5, 7, 12, 15 and 42 days after withdrawal of doxycycline (296 neurones recorded from 11 animals). 2.2. Deprivation To induce plasticity D1 vibrissa was spared for 17– 19 days, while all surrounding vibrissae were deprived every second day. Deprived vibrissae were allowed to regrow for 8–10 days before recording. The deprivation technique used has been found not to affect vibrissae innervation (Li et al., 1995) and is described in detail elsewhere (Glazewski et al., 1998). 2.3. Recording Anaesthesia was induced with metofane and maintained with urethane (1.5 g/kg of body weight). The depth of anaesthesia was monitored during the experiment and supplemented if necessary by small injections of urethane. To record, the skull was thinned 2–4 mm lateral to midline and 0–3 mm caudal from bregma. A small hole was made in the skull before each electrode penetration using a hypodermic needle. The neurones were sampled every 50 µm throughout the depth of the penetration (0–300 µm). All experiments were compliant with the UK 1986 Scientific Procedures Act. 2.4. Stimulation and analysis Vibrissae were stimulated using a computer-controlled piezoelectric stimulator. Fifty 1° stimuli were delivered at 1 Hz to every whisker that might lie in the receptive field for each neurone. Two parameters of the response were measured: the averaged magnitude of response (i.e. the number of spikes counted in the 5- to 50-ms poststimulus interval minus spontaneous activity) and the modal latency (from whisker stimulation to first spike). The average magnitude of response (recorded in the barrel columns immediately adjacent to the D1 barrel column) to stimulation of principal (P), spared (i.e. D1) and surround whiskers were calculated for each animal. The average for each group of animals tested was then calculated from the average from each individual in the group. The relative response to stimulation of the spared vib-
S. Glazewski et al. / Neuropharmacology 41 (2001) 771–778
rissae versus the response to the principal whisker (denoted F) was calculated for each cell (where (F)=D1/(D1+P), and D1 and P are averaged D1 vibrissa response and principal vibrissa response, respectively, for that cell). A vibrissa dominance histogram (VDH) could then be composed from the frequency of occurrence of each class of relative response (F). VDHs were compared using the weighted vibrissae dominance index (WVDI) calculated for each animal and averaged for treatment groups: (WVDI)=(0F0+1F1+…9F9)/9N where, F0 is the number of cells in the 0–0.099 band, F1 is the number of cells in the 0.1–0.199 band etc. and N is the total number of cells in a particular sample. An ANOVA was used to check for effects and interactions between variables and t-tests were used for post-hoc analysis. As described before these parameters (D1, WVDI) are distributed normally (Glazewski et al., 1996). Where distributions were not normal (deprived principal whisker responses, receptive field sizes, and latency of response) a chi-square or Kruskal–Wallis test was performed. Cumulative distribution functions (CDFs) were used to analyse the skewed distribution of principal whisker responses caused by deprivation in adolescents. Differences in CDFs were assessed using a Kolmogorov– Smirnov two-sample test. 2.5. Histological identification At the end of each electrode penetration a small lesion was made in layer IV (1 uA, DC, 10 s, tip negative). This served to mark the location of each penetration. After each experiment the animal was deeply anaesthetised, perfused through the heart with 0.2 M phosphatebuffered saline, followed by 3.7% buffered solution of formaldehyde and 10% sucrose in buffered formaldehyde. The brain was removed, the cortex flattened as described before (Strominger and Woolsey, 1987) and left overnight in 20% sucrose in a buffered solution of formaldehyde. Sections of 40 µm were cut tangential to the surface of the flattened cortex using a freezing microtome and the tissue was reacted for cytochrome oxidase (Wong-Riley, 1979). Sections were later analysed using a camera lucida system. 2.6. Gene expression assay Gene expression levels were estimated by performing RT-PCR on whole forebrain RNA samples (Mayford et al., 1995). To inhibit gene expression, doxycyclin was administered at a concentration of 40 ng/g in the diet. The time-course of gene expression was examined following a return to normal diet. The primers used amplified a 苲350 bp segment of both endogenous and mutant CaMKII mRNA. The resulting cDNA fragments were run on a gel and probed with a radiolabelled oligonucleotide specific for CaMKII-Asp286.
773
3. Results 3.1. The effect of the T286D mutation on plasticity Animals were reared with the transgene off from birth until 2 weeks before induction of plasticity. Plasticity was induced in the barrel cortex by sparing the D1 whisker and depriving the others for a period of 18 days (Glazewski et al., 1996). After allowing one week for the deprived whiskers to regrow, we assayed the responses generated in layer II/III neurones by stimulating the deprived and spared whisker for each cell. The response of the spared whisker relative to that of the principal whisker is measured by the vibrissae dominance index for each cell (see Section 2). The average index is shown in Fig. 1 for normal (undeprived) and deprived wild-type animals. The average increases from 0.16±0.04 in undeprived animals to 0.41±0.04 in deprived wild types following deprivation (where a value of 0 would indicate responses from the principal whisker only and a value of 0.5 equals input from spared and principal whiskers and a value of 1.0 would indicate responses from the spared whisker only). The average WVDI also increases following deprivation for animals expressing the autonomous (T286D) form of αCaMKII, from 0.11±0.04 to 0.29±0.03. An ANOVA showed that there was an effect of deprivation (F(1, 20)=32.0, p⬍10−4) and genotype (F(1, 20)=5.3, p⬍0.05) but no interaction between the two (F(1, 20)=1.05, p=0.32). This implies that while significant plasticity occurs in both cases, the WVDI values are slightly lower in mutants than wild types, independent of deprivation. The WVDI is affected by both the principal whisker response and the spared whisker response. Therefore, we looked at each measure individually. Fig. 1 shows that deprivation causes an increase in the spared whisker response in wild types from an average value (±sem) of 0.34±0.12 spikes per stimulus (sp/st) to 0.84±0.12 sp/st. A similar though slightly lower level of potentiation is seen in the T286D animals, from 0.39±0.12 to 0.61±0.11 sp/st. An ANOVA showed that there was an effect of deprivation (F(1, 20)=9.05, p⬍10−2) but not of genotype (F(1, 20)=0.49, p=0.49) on the spared whisker response, suggesting that the mutant form of αCaMKII does not interfere with experience-dependent potentiation. Finally, we also examined the effect of deprivation on the deprived principal whisker response. The CDFs for wild types and T286D mutants are shown in Fig. 1. It can be seen that the distributions are shifted leftward to lower values for both mutants and wild types. However, neither effect was statistically significant (p⬎0.05, Kolmogorov–Smirnov two-sample test). In summary, potentiation is not significantly decreased nor depression increased by expression of the autonomous form of CaMKII.
774
S. Glazewski et al. / Neuropharmacology 41 (2001) 771–778
Fig. 1. The effect of autonomous alpha-CaMKII on plasticity. (A) The weighted vibrissae dominance index is shown for wild-type and transgenic animals. The weighted vibrissae dominance is low in undeprived animals indicating that the principal whisker response is dominant over the D1 whisker response. However, in both wild types and transgenics, the WVDI increases indicating a shift of response away from the principal whisker toward the spared D1 whisker response. (A value of 0.5 indicates approximately equal contribution from principal and spared whisker input.) (B) Deprivation causes an increase in the spared whisker response in both wild types and transgenic mutants. (C) The CDF for the response of the deprived (regrown) principal whisker (empty squares) is shifted left toward lower values by deprivation compared with undeprived animals (full diamonds). (D) Similarly, in transgenic animals the deprived principal whisker response (empty squares) is shifted to lower values compared with the control (full diamonds). The depression is no greater in transgenics (C) than in wild types (D).
3.2. The timecourse of T286D CaMKII expression To establish that the transgene had turned on effectively in animals receiving doxycycline from birth we studied the effect of withdrawal of doxycycline on transgene mRNA production. By using RT-PCR to amplify the mRNA and a radiolabelled probe for the transgenic form of CaMKII we were able to detect the presence and amount of the transgene being expressed. Fig. 2(A) shows how the T286D form of αCaMKII increases following withdrawal of doxycyclin. The densitometricallymeasured levels are shown relative to levels in wild types and plotted in Fig. 2(B). It can be seen that the message is induced within 1 day.
In animals reared with the gene turned off, the transgene levels increase when the doxycyclin is withdrawn over the first 14 days. Levels are no higher at 90 days, indicating that the expression levels have saturated at about 14 days. In animals reared with the gene turned on from birth, expression shows a similar timecourse to that in older animals with doxycyclin withdrawn (Fig. 2(B)). Similar levels of expression are achieved in animals reared with the gene on and off within a few days of withdrawal of doxycyclin. However, if the animals are reared from birth with the gene on until adulthood, far higher levels are achieved than in animals reared from birth with the gene turned off (Fig. 2(B), right most point).
S. Glazewski et al. / Neuropharmacology 41 (2001) 771–778
775
Fig. 3. The near neighbour whisker response as a function of time after withdrawal of doxycycline. As shown, the transgene is activated 2 days after withdrawal of doxycycline, but the near neighbouring whisker responses are not affected by this event. Average responses from individual animals are shown as empty squares while full circles indicate the averages for particular time points.
3.4. The effect of the 7286D mutation on development
Fig. 2. The timecourse of CaMKIIT286A expression following withdrawal of doxycylin. (A) Expression levels of transgenic CaMKIIT286A increase over a period of 90 days following withdrawal of doxycylin in animals carrying the Tet transactivator gene and the gene for the T286D form of CaMKII. cDNA fragments from forebrain are shown radiolabelled for an oligonucleotide specific for CaMKIIT286A (see Section 2) (note only a weak signal in the wild-type sample). WT, wildtype mouse, Asp286, mutant transgene only (no tTA transgene), all others are from doubly transgenic mice. (B) Quantification of the assay shows that gene expression rises quickly over the first 2 days. Each unit of induction was calculated as the expression difference between double transgenic mouse on doxycyclin minus basal expression in the T286D mouse without tTA transgene, i.e., induction unit=‘day 0’ signal⫺‘T286D’ signal. Empty circles represent animals raised with the gene off, full circles animals raised with the gene on. The number of animals for each averaged point is shown adjacent to that time point.
In the experiments described in the preceding sections the transgene had been turned off during development until 2 weeks before the start of the deprivation experiment. In order to test the effect of the transgene on development we also looked at a number of animals reared with the gene turned on from birth and continuing into the deprivation period. Fig. 4 shows that, as with the animals in which the gene had been activated in adult life, potentiation was essentially unaffected by deprivation. In both wild types and T286D mutants the spared whisker response was potentiated equally in wild types (D1=1.37±0.38 sp/st) and T286D mutants (D1=1.39±0.38 sp/st) and these values were not different (F(1, 5)=0.0006, p=0.98). However, there was a clear difference in the level of surround receptive field response, which was greater in the T286D animals
3.3. Effect of the T286D mutation on surround receptive field properties We measured the surround receptive field response of neurones in layers II/III of the barrel field at different times after having discontinued administration of doxycycline. Fig. 2 shows that the transgene expression increases to approximately stable levels after 2 days. If anything there is a minor decrease in receptive field strength during this period (Fig. 3) but not a significant change. Expression levels are higher at 14 and 90 days than at 1 or 2 days after turning the gene on, but this is not reflected by any change in sensory responses at 42 days (Fig. 3). The results therefore suggest that there is no effect of the transgene on sensory responses in the cortex when the transgene is turned on after development of the cortex.
Fig. 4. The effect of autonomous αCaMKII on development. The response magnitude of the deprived near neighbour response (black bars) and the spared D1 whisker response (hatched bars) are shown for animals reared with the transgene expressed from birth. Left: plasticity is normal in wild-type animals with the D1 whisker response showing significant potentiation compared with the deprived near neighbouring whisker responses. Right: in transgenic animals, the spared whisker response is potentiated to the same degree as in wild types. However, the deprived surround receptive field whisker response is significantly greater than that found in wild-type animals.
776
S. Glazewski et al. / Neuropharmacology 41 (2001) 771–778
(NN=1.03±0.15) than in wild types (NN=0.34±0.15), and this difference was significant (F(1, 5)=10.3, p⬍ 0.05).
4. Discussion This study shows that increasing the proportion of the autonomous form of αCaMKII has very little effect on the level of potentiation or depression in the barrel cortex induced by whisker deprivation when expressed in adult animals. Conversely, if the gene is active during development the effect of the autonomous form of αCaMKII on receptive field response levels is profound. If the gene is active during development, the surround receptive field responses are far greater than normal, suggesting that either a larger number of synapses form between the barrel columns or that a similar number form but they are relatively potentiated. This result is analogous to the finding that hippocampal place fields were less focused in animals reared with the transgene turned on (Rotenberg et al., 1996). There are two possible reasons why the surround receptive field responses are greater in the animals reared with the transgene on from birth. One possibility is the existence of a sensitive period during development when the autonomous form of αCaMKII is able to affect receptive field development. The other possibility is that the surround receptive fields are only potentiated if a higher level of transgene expression occurs. The levels of transgene mRNA expression were far higher in animals reared from birth with the transgene on than in animals reared from birth with the transgene off, which would then explain why only the animals reared with the gene on from birth were affected (Fig. 2(B)). This argument assumes that the protein levels are related to the mRNA levels in a simple manner. This may be a reasonable assumption as in heterozygote αCaMKII knockouts approximately half the protein is expressed compared to wild types (Chapman et al., 1995). In favour of the other argument, that development proceeds through a period sensitive to the αCaMKII autophosphorylation state, it has previously been shown that autonomous forms of CaMKII can have a profound effect on early neuronal development. In a series of experiments conducted on development of synaptic function in hippocampal cultures and frog tectum it has been found that a virally introduced form of calcium independent CaMKII can alter the development of dendritic arbors (Zou and Cline, 1996; Wu and Cline, 1998) as well as cause insertion of AMPA receptor subunits at synapses (Wu et al., 1996; Shi et al., 1999). The effect on AMPA receptor insertion would be expected to cause a potentiated surround receptive field response if it developed without compensating mechanisms. This may be what has happened in the case of the animals reared
from birth with the gene turned on. The surround receptive field whiskers drive the cell at higher levels than normal because of higher numbers of synapses receiving AMPA receptor insertion during development. The reason the principal whisker responses are not elevated in these cases may be due to saturation of the response at a maximum. These studies may point to a role for phosphorylation of CaMKII in the early development of cortical synapses, but raise the question of why developmental deficits are not found in animals carrying only an unphosphorylatable form of CaMKll (Giese et al., 1998; Glazewski et al., 2000). One possibility is that the alpha isoform of CaMKII has a relatively minor role during this very early period of development when natural expression levels of the alpha isoform are low (Burgin et al., 1990). In this way the absence of the autophosphorylated form would have a minor effect while greater than normal expression of the phosphorylated form would have an effect. As mentioned above, the other possibility is that there is a threshold level of CaMKII T286D expression below which the transgene has no effect on sensory response levels. This would explain why the animals reared with the gene off showed no effect on surround receptive field responses. The reason why the animals reared with the gene on do show an effect would be due to the high levels of transgene expression in the adult rather than a direct effect on development. Unfortunately, our studies do not distinguish between these two possibilities. Our results are not predicted by the shift in the frequency that causes LTD rather than LTP in the T286D animals (Mayford et al., 1995). In wild-type animals a 5 Hz stimulus produces LTP in the hippocampus while in T286D animals it causes LTD. There are at least two reasons why such a change in the frequency of LTP induction does not affect experience-dependent potentiation. First, it is not clear how closely LTD corresponds to the experience-dependent depression induced in barrel cortex by whisker deprivation. Although depression is known to be dependent on cortical activity (Wallace et al., 2001), the synaptic mechanism is not known. Second, the frequencies involved in barrel cortex are probably higher than those produced by LTD experiments even following deprivation. The spared whisker response of layer IV cells increases in peak frequency following deprivation from approximately 50 to 75 Hz (Kelly et al., 1999). Both frequencies would cause LTP if applied as an electrical tetanus to the hippocampal circuitry. Conversely, deprivation reduces a trimmed whisker response from a peak of about 50 Hz to about 10 Hz and again both would, in wild types, cause LTP in the hippocampus (Kelly et al., 1999). In addition, depriving a whisker still leaves a relatively high level of spontaneous activity (Kelly et al., 1999). This suggests that the timing of the pre- and post-synaptic responses may
S. Glazewski et al. / Neuropharmacology 41 (2001) 771–778
be more important than the absolute frequency for determining whether experience-dependent depression or potentiation occurs. The spontaneous activity may produce depression by occurring out of phase with depolarisation of the post-synaptic neurone or by failing more often than not to depolarise the post-synaptic neurone (Feldman, 2000; Fox, 2000). It would be useful to know whether out of phase pre- and post-synaptic activity leads to depression via the same molecular mechanisms as induced by low frequency stimulation. In earlier experiments it has been found that the autophosphorylation state of CaMKII is important for experience-dependent plasticity and LTP in the barrel cortex (Glazewski et al., 2000). It is conceivable that the T286D mutation does not affect cortical plasticity because it is expressed in parallel with the endogenous form of the molecule in this transgenic animal. Finally, while potentiation seems to be dependent on αCaMKII in the cortex we have not found any evidence that depression is dependent on αCaMKII in three αCaMKII mutants to date (Glazewski et al. 1996, 2000; present study).
5. Conclusion Our studies have shown that expression of the autonomously active form of αCaMKII (T286D) does not affect plasticity in animals reared with the gene on from birth nor in animals reared with the gene ‘off’ from birth but ‘on’ during the deprivation period. However, autonomous αCaMKII does have an effect on surround receptive field magnitude in animals reared with the gene on from birth. However, we do not know at present whether this is due to an effect of αCaMKII during development or to the higher levels of transgene expressed in adulthood in animals reared with the gene on from birth.
Acknowledgements We gratefully acknowledge the assistance of Mervyn McKenna for all histology. Grant from the National Institutes of Health NS27759 (K.F.) and the MRC (K.F.) supported this work.
References Bear, M.F., Cooper, L.N., Ebner, F.F., 1987. A physiological basis for a theory of synapse modification. Science 237, 42–48. Bienenstock, E., Cooper, L.N., Munroe, P., 1982. Theory for the development of neuron selectivity: orientation specificity and binocular interaction in the visual cortex. Journal of Neuroscience 2, 32–48. Blakemore, C., Van Sluyters, R.C., 1974. Reversal of the physiological effect of monocular deprivation in kittens: further evidence for a sensitive period. Journal of Physiology 237, 195–216. Burgin, K.E., Waxham, M.N., Rickling, S., Westgate, S.A., Mobley,
777
W.C., Kelly, P.T., 1990. In situ hybridization histochemistry of Ca2+/calmodulin-dependent protein kinase in developing rat brain. Journal of Neuroscience 10, 1788–1798. Chapman, P.F., Frenguelli, B.G., Smith, A., Chen, C.M., Silva, A.J., 1995. The alpha-Ca2+/calmodulin kinase II: a bidirectional modulator of presynaptic plasticity. Neuron 4, 591–597. Clothiaux, E.E., Bear, M.F., Cooper, L.N., 1991. Synpatic plasticity in visual cortex: comparison of theory with experiment. Journal of Neurophysiology 66, 1785–1804. Feldman, D.E., 2000. Timing based LTP and LTD at vertical input layer II/III pyramidal cells in rat barrel cortex. Neuron 27, 45–56. Fox, K., 2000. Timing is everything. Neuron 27, 1–3. Giese, K.P., Fedorov, N.B., Filipkowski, R.K., Silva, A.J., 1998. Autophosphorylation at Thr286 of the alpha calcium-calmodulin kinase II in LTP and learning. Science 279, 870–873. Glazewski, S., Chen, C-H., Silva, A., Fox, K., 1996. The requirement for αCAMKII in experience-dependent plasticity of the barrel cortex. Science 272, 421–423. Glazewski, S., McKenna, M., Jacquin, M., Fox, K., 1998. The nature and origins of experience-dependent depression of vibrissae responses in rat barrel cortex. European Journal of Neuroscience 10, 2107–2116. Glazewski, S., Giese, K.P., Silva, A., Fox, K., 2000. The role of the alpha-CaMKII autophosphorylation in neocortical experiencedependent plasticity. Nature Neurosci. 3 (9), 911–918. Kelly, M.K., Carvel, G.E., Kodger, J.M., Simons, D.J., 1999. Sensory loss by selective whisker removal produces immediate disinhibition in the somatosensory cortex of behaving rats. Journal of Neuroscience 19, 9117–9125. Kennedy, M.B., Bennett, M.K., Erondu, N.E., 1983. Biochemical and immunochemical evidence that the ‘major postsynaptic density protein’ is a subunit of a calmodulin-dependent protein kinase. Proceedings of the National Academy of Science USA 80, 7357–7361. Kirkwood, A., Silva, A., Bear, M.F., 1997. Age dependent decreases of synaptic plasticity in the neocortex of alphaCaMKII mutant mice. Proceedings of the National Academy of Science USA 94, 3380– 3383. Li, X., Glazewski, S., Lin, X., Elde, R., Fox, K., 1995. The effect of vibrissae deprivation on follicle innervation, neuropeptide synthesis in the trigeminal ganglion and S1 barrel field cortical plasticity. Journal of Comparative Neurology 357, 465–481. Mayford, M., Wang, E.R., O’Dell, T.J., 1995. CaMKII regulates the frequency response function of hippocampal synapses for the production of both LTP and LTD. Cell 81, 891–904. Miller, S.G., Kennedy, M.B., 1985. Distinct forebrain and cerebellar isozymes of type II Ca2+/calmodulin-dependent protein kinase associate differently with the postsynaptic density fraction. Journal of Biological Chemistry 260, 9039–9046. Rotenberg, A., Mayford, M., Hawkins, R.D., Kandel, E.R., Muller, R.U., 1996. Mice expressing activated CaMKII lack low frequency LTP and do not form stable place cells in the CA1 region of the hippocampus. Cell 87, 1351–1361. Shi, S.H., Hayashi, Y., Petralia, R.S., Zaman, S.H., Wenthold, R.J., Svoboda, K., Malinow, R., 1999. Rapid spine delivery and redistribution of AMPA receptors after synaptic NMDA receptor activation. Science 284, 1811–1816. Silva, A.J., Stevens, C.F., Tonegawa, S., Wang, Y., 1992. Deficient hippocampal long-term potentiation in alpha-calcium-calmodulin kinase II mutant mice. Science 257, 201–206. Strominger, R.N., Woolsey, T.A., 1987. Templates for locating the whisker area in a fresh flattened mouse and rat cortex. Journal of Neuroscience Methods 22, 113–118. Wallace, H., Liming, K., Glazewski, S., Fox, K., 2001. The role of cortical activity in experience-dependent potentiation and depression of sensory responses in barrel field. Journal of Neuroscience 21 (11), 3881–3894.
778
S. Glazewski et al. / Neuropharmacology 41 (2001) 771–778
Wong-Riley, M., 1979. Changes in visual system of monocularly sutured or enucleated cats demonstrable with cytochrome oxidase histochemistry. Brain Research 171, 11–28. Wu, G.Y., Cline, H.T., 1998. Stabilisation of dendritic arbour structure in vivo by CaMKII. Science 279, 222–226.
Wu, G., Malinow, R., Cline, H.T., 1996. Maturation of a central glutamatergic synapse. Science 274, 972–976. Zou, D.J., Cline, H.T., 1996. Expression of constitutively active CaMKII in target tissue modifies presynaptic axon arbour growth. Neuron 16, 529–539.
The effect of autonomous alpha-CaMKII expression on sensory responses and experience-dependent plasticity in mouse barrel cortex Stanislaw Glazewski a, Rafael Bejar b, Mark Mayford b, Kevin Fox a
a,*
Cardiff School of Biosciences, Cardiff University, Museum Avenue, Cardiff CF10 3US, UK b The Scripps Research Institute, ICDN Room 206, La Jolla, CA USA Received 11 April 2001; received in revised form 20 June 2001; accepted 22 June 2001
Abstract The calcium/calmodulin kinase II (CaMKII) autophosphorylation site is thought to be important for plasticity, learning and memory. If autophosphorylation is prevented by a point mutation (T286A) LTP is blocked in the hippocampus and cortex. Conversely, if the point mutation mimics autophosphorylation (T286D) a range of frequencies that normally produce LTP in wild types cause LTD instead. In order to test whether the αCaMKII-T286D mutation increases levels of depression in vivo, we examined the effect of the αCaMKII-T286D transgene on plasticity induced in the barrel cortex by whisker deprivation. Surprisingly, the mutation did not affect depression or potentiation. However, in animals reared with the transgene turned on from birth, the surround receptive field responses were greater than normal. This effect may be due to the potentiating action of autophosphorylated CaMKII during early development. 2001 Elsevier Science Ltd. All rights reserved. Keywords: Phosphorylation; LTP; LTD; Somatosensory; Neocortex
1. Introduction Calcium/calmodulin kinase II (CaMKII) is a major postsynaptic density protein thought to be involved in synaptic plasticity (Kennedy et al., 1983; Miller and Kennedy, 1985). In general, mutations of the CaMKII gene affect learning, memory, LTP and experiencedependent plasticity. Knockout of the αCaMKII gene or inactivation of the autophosphorylation site by point mutation (T286A), cause abolition of these forms of plasticity (Silva et al., 1992; Glazewski et al. 1996, 2000; Kirkwood et al., 1997; Giese et al., 1998). However, increasing the proportion of calcium-independent αCaMKII with a transgenic form of calcium-independent αCaMKII (T286D) produces a more complex phenotype (Mayford et al., 1995). Substitution of the threonine autophosphorylation site with an aspartate produces a constitutively active CaMKII molecule. This
* Corresponding author. Tel.: +44-2920-874632; fax: +44-2920874744. E-mail address: [email protected] (K. Fox).
mutation might be expected to occlude LTP by saturating it. Instead, the ability to produce LTP is preserved, but the T286D mutation causes a shift in the frequency at which LTP can be induced. A range of frequencies that evoke LTP in a wild type cause LTD, while LTP is only produced at far higher tetanus frequencies than normal. The effect can be summarised as a rightwards shift in the BCM curve in a direction favouring production of LTD (Mayford et al., 1995). The BCM curve describes the relationship between the activity of the cell and the direction of synaptic change (namely, a direction towards potentiation or depression). A crucial aspect of the theory is that the activity threshold for potentiation is itself variable and dependent on the prior history of the neurone’s activity. The BCM theory has been used to explain a number of effects known to occur in experience-dependent plasticity (Bienenstock et al., 1982). For example, it is able to explain the way in which weak inputs potentiate when there are no strong inputs to depolarise the neurone. Such a situation occurs during reverse lid suture for example, where the first eye to be closed drives very few cells in the visual cortex after a few days of deprivation,
0028-3908/01/$ - see front matter 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 8 - 3 9 0 8 ( 0 1 ) 0 0 0 9 7 - 1
772
S. Glazewski et al. / Neuropharmacology 41 (2001) 771–778
and yet, despite the weakness of the synaptic drive from the recently closed eye, it is able to potentiate its inputs on reversing the eye-lid suture (Blakemore and Van Sluyters, 1974; Bienenstock et al., 1982; Clothiaux et al., 1991). The BCM theory is also consistent with a number of other observations in ocular dominance and whisker deprivation plasticity. One of the central features of the theory is the idea that the threshold at which potentiation rather than depression is produced is not fixed but variable (Bienenstock et al., 1982). The observation that the cross over point at which LTP is induced can be altered by a CaMKII mutation is therefore of interest and might be expected to have an effect on experience-dependent plasticity (Mayford et al., 1995; Bear et al., 1987). To date, the relationship between the BCM curve when applied to LTP and the BCM theory applied to experience-dependent plasticity has not been fully investigated. A mutation that does not prevent plasticity but which shifts the BCM curve therefore provides an opportunity to examine the relationship. One might predict that a mutation which sets a higher threshold for LTP and which favours the production of LTD might reduce experience-dependent potentiation and increase the chances of experience-dependent depression. Both potentiation and depression can be measured in mouse barrel cortex by inducing plasticity by whisker deprivation (Glazewski et al., 1996). We therefore induced plasticity in mice expressing the T286D mutated form of αCaMKII and looked at the effect on plasticity of spared and deprived whisker responses. Experiencedependent depression decreases between about 2 and 6 months of age in mice: we looked at depression at an age where depression is waning in order to increase the likelihood of seeing an increase in depression as a result of the mutation should it occur. The results suggest that experience-dependent potentiation and depression are not affected by expression of autonomous αCaMKII. However, we also found that the level of surround receptive field whisker responses are elevated if the transgene is on during early development. 2. Methods 2.1. Subjects 2.1.1. Plasticity experiments Recordings were made from 1123 neurones in layer II/III of the barrel cortex of 22 undeprived (543 neurones) and 24 deprived (580 neurones) mice. The animals were either double (CaM-tet) or single (tet and CaM) transgenics or their wild-type littermates, at the age of 3–4 months. All transgenic animals were kept on doxycycline (40 µg/l) administered via the drinking water, until 1 week before starting deprivation. Recordings were also made from 11 animals at an age
exceeding 6 months (284 neurones recorded). All recordings were conducted blind to genotype. Animals were genotyped by PCR as described before (Mayford et al., 1995). 2.1.2. Time course experiments To check whether introduction of the transgene has an influence on the magnitude of principal and surround responses recorded in the barrel cortex, a time-course experiment was performed. The mice were initially kept on a doxycycline diet until 6 months of age, and were recorded at 5, 7, 12, 15 and 42 days after withdrawal of doxycycline (296 neurones recorded from 11 animals). 2.2. Deprivation To induce plasticity D1 vibrissa was spared for 17– 19 days, while all surrounding vibrissae were deprived every second day. Deprived vibrissae were allowed to regrow for 8–10 days before recording. The deprivation technique used has been found not to affect vibrissae innervation (Li et al., 1995) and is described in detail elsewhere (Glazewski et al., 1998). 2.3. Recording Anaesthesia was induced with metofane and maintained with urethane (1.5 g/kg of body weight). The depth of anaesthesia was monitored during the experiment and supplemented if necessary by small injections of urethane. To record, the skull was thinned 2–4 mm lateral to midline and 0–3 mm caudal from bregma. A small hole was made in the skull before each electrode penetration using a hypodermic needle. The neurones were sampled every 50 µm throughout the depth of the penetration (0–300 µm). All experiments were compliant with the UK 1986 Scientific Procedures Act. 2.4. Stimulation and analysis Vibrissae were stimulated using a computer-controlled piezoelectric stimulator. Fifty 1° stimuli were delivered at 1 Hz to every whisker that might lie in the receptive field for each neurone. Two parameters of the response were measured: the averaged magnitude of response (i.e. the number of spikes counted in the 5- to 50-ms poststimulus interval minus spontaneous activity) and the modal latency (from whisker stimulation to first spike). The average magnitude of response (recorded in the barrel columns immediately adjacent to the D1 barrel column) to stimulation of principal (P), spared (i.e. D1) and surround whiskers were calculated for each animal. The average for each group of animals tested was then calculated from the average from each individual in the group. The relative response to stimulation of the spared vib-
S. Glazewski et al. / Neuropharmacology 41 (2001) 771–778
rissae versus the response to the principal whisker (denoted F) was calculated for each cell (where (F)=D1/(D1+P), and D1 and P are averaged D1 vibrissa response and principal vibrissa response, respectively, for that cell). A vibrissa dominance histogram (VDH) could then be composed from the frequency of occurrence of each class of relative response (F). VDHs were compared using the weighted vibrissae dominance index (WVDI) calculated for each animal and averaged for treatment groups: (WVDI)=(0F0+1F1+…9F9)/9N where, F0 is the number of cells in the 0–0.099 band, F1 is the number of cells in the 0.1–0.199 band etc. and N is the total number of cells in a particular sample. An ANOVA was used to check for effects and interactions between variables and t-tests were used for post-hoc analysis. As described before these parameters (D1, WVDI) are distributed normally (Glazewski et al., 1996). Where distributions were not normal (deprived principal whisker responses, receptive field sizes, and latency of response) a chi-square or Kruskal–Wallis test was performed. Cumulative distribution functions (CDFs) were used to analyse the skewed distribution of principal whisker responses caused by deprivation in adolescents. Differences in CDFs were assessed using a Kolmogorov– Smirnov two-sample test. 2.5. Histological identification At the end of each electrode penetration a small lesion was made in layer IV (1 uA, DC, 10 s, tip negative). This served to mark the location of each penetration. After each experiment the animal was deeply anaesthetised, perfused through the heart with 0.2 M phosphatebuffered saline, followed by 3.7% buffered solution of formaldehyde and 10% sucrose in buffered formaldehyde. The brain was removed, the cortex flattened as described before (Strominger and Woolsey, 1987) and left overnight in 20% sucrose in a buffered solution of formaldehyde. Sections of 40 µm were cut tangential to the surface of the flattened cortex using a freezing microtome and the tissue was reacted for cytochrome oxidase (Wong-Riley, 1979). Sections were later analysed using a camera lucida system. 2.6. Gene expression assay Gene expression levels were estimated by performing RT-PCR on whole forebrain RNA samples (Mayford et al., 1995). To inhibit gene expression, doxycyclin was administered at a concentration of 40 ng/g in the diet. The time-course of gene expression was examined following a return to normal diet. The primers used amplified a 苲350 bp segment of both endogenous and mutant CaMKII mRNA. The resulting cDNA fragments were run on a gel and probed with a radiolabelled oligonucleotide specific for CaMKII-Asp286.
773
3. Results 3.1. The effect of the T286D mutation on plasticity Animals were reared with the transgene off from birth until 2 weeks before induction of plasticity. Plasticity was induced in the barrel cortex by sparing the D1 whisker and depriving the others for a period of 18 days (Glazewski et al., 1996). After allowing one week for the deprived whiskers to regrow, we assayed the responses generated in layer II/III neurones by stimulating the deprived and spared whisker for each cell. The response of the spared whisker relative to that of the principal whisker is measured by the vibrissae dominance index for each cell (see Section 2). The average index is shown in Fig. 1 for normal (undeprived) and deprived wild-type animals. The average increases from 0.16±0.04 in undeprived animals to 0.41±0.04 in deprived wild types following deprivation (where a value of 0 would indicate responses from the principal whisker only and a value of 0.5 equals input from spared and principal whiskers and a value of 1.0 would indicate responses from the spared whisker only). The average WVDI also increases following deprivation for animals expressing the autonomous (T286D) form of αCaMKII, from 0.11±0.04 to 0.29±0.03. An ANOVA showed that there was an effect of deprivation (F(1, 20)=32.0, p⬍10−4) and genotype (F(1, 20)=5.3, p⬍0.05) but no interaction between the two (F(1, 20)=1.05, p=0.32). This implies that while significant plasticity occurs in both cases, the WVDI values are slightly lower in mutants than wild types, independent of deprivation. The WVDI is affected by both the principal whisker response and the spared whisker response. Therefore, we looked at each measure individually. Fig. 1 shows that deprivation causes an increase in the spared whisker response in wild types from an average value (±sem) of 0.34±0.12 spikes per stimulus (sp/st) to 0.84±0.12 sp/st. A similar though slightly lower level of potentiation is seen in the T286D animals, from 0.39±0.12 to 0.61±0.11 sp/st. An ANOVA showed that there was an effect of deprivation (F(1, 20)=9.05, p⬍10−2) but not of genotype (F(1, 20)=0.49, p=0.49) on the spared whisker response, suggesting that the mutant form of αCaMKII does not interfere with experience-dependent potentiation. Finally, we also examined the effect of deprivation on the deprived principal whisker response. The CDFs for wild types and T286D mutants are shown in Fig. 1. It can be seen that the distributions are shifted leftward to lower values for both mutants and wild types. However, neither effect was statistically significant (p⬎0.05, Kolmogorov–Smirnov two-sample test). In summary, potentiation is not significantly decreased nor depression increased by expression of the autonomous form of CaMKII.
774
S. Glazewski et al. / Neuropharmacology 41 (2001) 771–778
Fig. 1. The effect of autonomous alpha-CaMKII on plasticity. (A) The weighted vibrissae dominance index is shown for wild-type and transgenic animals. The weighted vibrissae dominance is low in undeprived animals indicating that the principal whisker response is dominant over the D1 whisker response. However, in both wild types and transgenics, the WVDI increases indicating a shift of response away from the principal whisker toward the spared D1 whisker response. (A value of 0.5 indicates approximately equal contribution from principal and spared whisker input.) (B) Deprivation causes an increase in the spared whisker response in both wild types and transgenic mutants. (C) The CDF for the response of the deprived (regrown) principal whisker (empty squares) is shifted left toward lower values by deprivation compared with undeprived animals (full diamonds). (D) Similarly, in transgenic animals the deprived principal whisker response (empty squares) is shifted to lower values compared with the control (full diamonds). The depression is no greater in transgenics (C) than in wild types (D).
3.2. The timecourse of T286D CaMKII expression To establish that the transgene had turned on effectively in animals receiving doxycycline from birth we studied the effect of withdrawal of doxycycline on transgene mRNA production. By using RT-PCR to amplify the mRNA and a radiolabelled probe for the transgenic form of CaMKII we were able to detect the presence and amount of the transgene being expressed. Fig. 2(A) shows how the T286D form of αCaMKII increases following withdrawal of doxycyclin. The densitometricallymeasured levels are shown relative to levels in wild types and plotted in Fig. 2(B). It can be seen that the message is induced within 1 day.
In animals reared with the gene turned off, the transgene levels increase when the doxycyclin is withdrawn over the first 14 days. Levels are no higher at 90 days, indicating that the expression levels have saturated at about 14 days. In animals reared with the gene turned on from birth, expression shows a similar timecourse to that in older animals with doxycyclin withdrawn (Fig. 2(B)). Similar levels of expression are achieved in animals reared with the gene on and off within a few days of withdrawal of doxycyclin. However, if the animals are reared from birth with the gene on until adulthood, far higher levels are achieved than in animals reared from birth with the gene turned off (Fig. 2(B), right most point).
S. Glazewski et al. / Neuropharmacology 41 (2001) 771–778
775
Fig. 3. The near neighbour whisker response as a function of time after withdrawal of doxycycline. As shown, the transgene is activated 2 days after withdrawal of doxycycline, but the near neighbouring whisker responses are not affected by this event. Average responses from individual animals are shown as empty squares while full circles indicate the averages for particular time points.
3.4. The effect of the 7286D mutation on development
Fig. 2. The timecourse of CaMKIIT286A expression following withdrawal of doxycylin. (A) Expression levels of transgenic CaMKIIT286A increase over a period of 90 days following withdrawal of doxycylin in animals carrying the Tet transactivator gene and the gene for the T286D form of CaMKII. cDNA fragments from forebrain are shown radiolabelled for an oligonucleotide specific for CaMKIIT286A (see Section 2) (note only a weak signal in the wild-type sample). WT, wildtype mouse, Asp286, mutant transgene only (no tTA transgene), all others are from doubly transgenic mice. (B) Quantification of the assay shows that gene expression rises quickly over the first 2 days. Each unit of induction was calculated as the expression difference between double transgenic mouse on doxycyclin minus basal expression in the T286D mouse without tTA transgene, i.e., induction unit=‘day 0’ signal⫺‘T286D’ signal. Empty circles represent animals raised with the gene off, full circles animals raised with the gene on. The number of animals for each averaged point is shown adjacent to that time point.
In the experiments described in the preceding sections the transgene had been turned off during development until 2 weeks before the start of the deprivation experiment. In order to test the effect of the transgene on development we also looked at a number of animals reared with the gene turned on from birth and continuing into the deprivation period. Fig. 4 shows that, as with the animals in which the gene had been activated in adult life, potentiation was essentially unaffected by deprivation. In both wild types and T286D mutants the spared whisker response was potentiated equally in wild types (D1=1.37±0.38 sp/st) and T286D mutants (D1=1.39±0.38 sp/st) and these values were not different (F(1, 5)=0.0006, p=0.98). However, there was a clear difference in the level of surround receptive field response, which was greater in the T286D animals
3.3. Effect of the T286D mutation on surround receptive field properties We measured the surround receptive field response of neurones in layers II/III of the barrel field at different times after having discontinued administration of doxycycline. Fig. 2 shows that the transgene expression increases to approximately stable levels after 2 days. If anything there is a minor decrease in receptive field strength during this period (Fig. 3) but not a significant change. Expression levels are higher at 14 and 90 days than at 1 or 2 days after turning the gene on, but this is not reflected by any change in sensory responses at 42 days (Fig. 3). The results therefore suggest that there is no effect of the transgene on sensory responses in the cortex when the transgene is turned on after development of the cortex.
Fig. 4. The effect of autonomous αCaMKII on development. The response magnitude of the deprived near neighbour response (black bars) and the spared D1 whisker response (hatched bars) are shown for animals reared with the transgene expressed from birth. Left: plasticity is normal in wild-type animals with the D1 whisker response showing significant potentiation compared with the deprived near neighbouring whisker responses. Right: in transgenic animals, the spared whisker response is potentiated to the same degree as in wild types. However, the deprived surround receptive field whisker response is significantly greater than that found in wild-type animals.
776
S. Glazewski et al. / Neuropharmacology 41 (2001) 771–778
(NN=1.03±0.15) than in wild types (NN=0.34±0.15), and this difference was significant (F(1, 5)=10.3, p⬍ 0.05).
4. Discussion This study shows that increasing the proportion of the autonomous form of αCaMKII has very little effect on the level of potentiation or depression in the barrel cortex induced by whisker deprivation when expressed in adult animals. Conversely, if the gene is active during development the effect of the autonomous form of αCaMKII on receptive field response levels is profound. If the gene is active during development, the surround receptive field responses are far greater than normal, suggesting that either a larger number of synapses form between the barrel columns or that a similar number form but they are relatively potentiated. This result is analogous to the finding that hippocampal place fields were less focused in animals reared with the transgene turned on (Rotenberg et al., 1996). There are two possible reasons why the surround receptive field responses are greater in the animals reared with the transgene on from birth. One possibility is the existence of a sensitive period during development when the autonomous form of αCaMKII is able to affect receptive field development. The other possibility is that the surround receptive fields are only potentiated if a higher level of transgene expression occurs. The levels of transgene mRNA expression were far higher in animals reared from birth with the transgene on than in animals reared from birth with the transgene off, which would then explain why only the animals reared with the gene on from birth were affected (Fig. 2(B)). This argument assumes that the protein levels are related to the mRNA levels in a simple manner. This may be a reasonable assumption as in heterozygote αCaMKII knockouts approximately half the protein is expressed compared to wild types (Chapman et al., 1995). In favour of the other argument, that development proceeds through a period sensitive to the αCaMKII autophosphorylation state, it has previously been shown that autonomous forms of CaMKII can have a profound effect on early neuronal development. In a series of experiments conducted on development of synaptic function in hippocampal cultures and frog tectum it has been found that a virally introduced form of calcium independent CaMKII can alter the development of dendritic arbors (Zou and Cline, 1996; Wu and Cline, 1998) as well as cause insertion of AMPA receptor subunits at synapses (Wu et al., 1996; Shi et al., 1999). The effect on AMPA receptor insertion would be expected to cause a potentiated surround receptive field response if it developed without compensating mechanisms. This may be what has happened in the case of the animals reared
from birth with the gene turned on. The surround receptive field whiskers drive the cell at higher levels than normal because of higher numbers of synapses receiving AMPA receptor insertion during development. The reason the principal whisker responses are not elevated in these cases may be due to saturation of the response at a maximum. These studies may point to a role for phosphorylation of CaMKII in the early development of cortical synapses, but raise the question of why developmental deficits are not found in animals carrying only an unphosphorylatable form of CaMKll (Giese et al., 1998; Glazewski et al., 2000). One possibility is that the alpha isoform of CaMKII has a relatively minor role during this very early period of development when natural expression levels of the alpha isoform are low (Burgin et al., 1990). In this way the absence of the autophosphorylated form would have a minor effect while greater than normal expression of the phosphorylated form would have an effect. As mentioned above, the other possibility is that there is a threshold level of CaMKII T286D expression below which the transgene has no effect on sensory response levels. This would explain why the animals reared with the gene off showed no effect on surround receptive field responses. The reason why the animals reared with the gene on do show an effect would be due to the high levels of transgene expression in the adult rather than a direct effect on development. Unfortunately, our studies do not distinguish between these two possibilities. Our results are not predicted by the shift in the frequency that causes LTD rather than LTP in the T286D animals (Mayford et al., 1995). In wild-type animals a 5 Hz stimulus produces LTP in the hippocampus while in T286D animals it causes LTD. There are at least two reasons why such a change in the frequency of LTP induction does not affect experience-dependent potentiation. First, it is not clear how closely LTD corresponds to the experience-dependent depression induced in barrel cortex by whisker deprivation. Although depression is known to be dependent on cortical activity (Wallace et al., 2001), the synaptic mechanism is not known. Second, the frequencies involved in barrel cortex are probably higher than those produced by LTD experiments even following deprivation. The spared whisker response of layer IV cells increases in peak frequency following deprivation from approximately 50 to 75 Hz (Kelly et al., 1999). Both frequencies would cause LTP if applied as an electrical tetanus to the hippocampal circuitry. Conversely, deprivation reduces a trimmed whisker response from a peak of about 50 Hz to about 10 Hz and again both would, in wild types, cause LTP in the hippocampus (Kelly et al., 1999). In addition, depriving a whisker still leaves a relatively high level of spontaneous activity (Kelly et al., 1999). This suggests that the timing of the pre- and post-synaptic responses may
S. Glazewski et al. / Neuropharmacology 41 (2001) 771–778
be more important than the absolute frequency for determining whether experience-dependent depression or potentiation occurs. The spontaneous activity may produce depression by occurring out of phase with depolarisation of the post-synaptic neurone or by failing more often than not to depolarise the post-synaptic neurone (Feldman, 2000; Fox, 2000). It would be useful to know whether out of phase pre- and post-synaptic activity leads to depression via the same molecular mechanisms as induced by low frequency stimulation. In earlier experiments it has been found that the autophosphorylation state of CaMKII is important for experience-dependent plasticity and LTP in the barrel cortex (Glazewski et al., 2000). It is conceivable that the T286D mutation does not affect cortical plasticity because it is expressed in parallel with the endogenous form of the molecule in this transgenic animal. Finally, while potentiation seems to be dependent on αCaMKII in the cortex we have not found any evidence that depression is dependent on αCaMKII in three αCaMKII mutants to date (Glazewski et al. 1996, 2000; present study).
5. Conclusion Our studies have shown that expression of the autonomously active form of αCaMKII (T286D) does not affect plasticity in animals reared with the gene on from birth nor in animals reared with the gene ‘off’ from birth but ‘on’ during the deprivation period. However, autonomous αCaMKII does have an effect on surround receptive field magnitude in animals reared with the gene on from birth. However, we do not know at present whether this is due to an effect of αCaMKII during development or to the higher levels of transgene expressed in adulthood in animals reared with the gene on from birth.
Acknowledgements We gratefully acknowledge the assistance of Mervyn McKenna for all histology. Grant from the National Institutes of Health NS27759 (K.F.) and the MRC (K.F.) supported this work.
References Bear, M.F., Cooper, L.N., Ebner, F.F., 1987. A physiological basis for a theory of synapse modification. Science 237, 42–48. Bienenstock, E., Cooper, L.N., Munroe, P., 1982. Theory for the development of neuron selectivity: orientation specificity and binocular interaction in the visual cortex. Journal of Neuroscience 2, 32–48. Blakemore, C., Van Sluyters, R.C., 1974. Reversal of the physiological effect of monocular deprivation in kittens: further evidence for a sensitive period. Journal of Physiology 237, 195–216. Burgin, K.E., Waxham, M.N., Rickling, S., Westgate, S.A., Mobley,
777
W.C., Kelly, P.T., 1990. In situ hybridization histochemistry of Ca2+/calmodulin-dependent protein kinase in developing rat brain. Journal of Neuroscience 10, 1788–1798. Chapman, P.F., Frenguelli, B.G., Smith, A., Chen, C.M., Silva, A.J., 1995. The alpha-Ca2+/calmodulin kinase II: a bidirectional modulator of presynaptic plasticity. Neuron 4, 591–597. Clothiaux, E.E., Bear, M.F., Cooper, L.N., 1991. Synpatic plasticity in visual cortex: comparison of theory with experiment. Journal of Neurophysiology 66, 1785–1804. Feldman, D.E., 2000. Timing based LTP and LTD at vertical input layer II/III pyramidal cells in rat barrel cortex. Neuron 27, 45–56. Fox, K., 2000. Timing is everything. Neuron 27, 1–3. Giese, K.P., Fedorov, N.B., Filipkowski, R.K., Silva, A.J., 1998. Autophosphorylation at Thr286 of the alpha calcium-calmodulin kinase II in LTP and learning. Science 279, 870–873. Glazewski, S., Chen, C-H., Silva, A., Fox, K., 1996. The requirement for αCAMKII in experience-dependent plasticity of the barrel cortex. Science 272, 421–423. Glazewski, S., McKenna, M., Jacquin, M., Fox, K., 1998. The nature and origins of experience-dependent depression of vibrissae responses in rat barrel cortex. European Journal of Neuroscience 10, 2107–2116. Glazewski, S., Giese, K.P., Silva, A., Fox, K., 2000. The role of the alpha-CaMKII autophosphorylation in neocortical experiencedependent plasticity. Nature Neurosci. 3 (9), 911–918. Kelly, M.K., Carvel, G.E., Kodger, J.M., Simons, D.J., 1999. Sensory loss by selective whisker removal produces immediate disinhibition in the somatosensory cortex of behaving rats. Journal of Neuroscience 19, 9117–9125. Kennedy, M.B., Bennett, M.K., Erondu, N.E., 1983. Biochemical and immunochemical evidence that the ‘major postsynaptic density protein’ is a subunit of a calmodulin-dependent protein kinase. Proceedings of the National Academy of Science USA 80, 7357–7361. Kirkwood, A., Silva, A., Bear, M.F., 1997. Age dependent decreases of synaptic plasticity in the neocortex of alphaCaMKII mutant mice. Proceedings of the National Academy of Science USA 94, 3380– 3383. Li, X., Glazewski, S., Lin, X., Elde, R., Fox, K., 1995. The effect of vibrissae deprivation on follicle innervation, neuropeptide synthesis in the trigeminal ganglion and S1 barrel field cortical plasticity. Journal of Comparative Neurology 357, 465–481. Mayford, M., Wang, E.R., O’Dell, T.J., 1995. CaMKII regulates the frequency response function of hippocampal synapses for the production of both LTP and LTD. Cell 81, 891–904. Miller, S.G., Kennedy, M.B., 1985. Distinct forebrain and cerebellar isozymes of type II Ca2+/calmodulin-dependent protein kinase associate differently with the postsynaptic density fraction. Journal of Biological Chemistry 260, 9039–9046. Rotenberg, A., Mayford, M., Hawkins, R.D., Kandel, E.R., Muller, R.U., 1996. Mice expressing activated CaMKII lack low frequency LTP and do not form stable place cells in the CA1 region of the hippocampus. Cell 87, 1351–1361. Shi, S.H., Hayashi, Y., Petralia, R.S., Zaman, S.H., Wenthold, R.J., Svoboda, K., Malinow, R., 1999. Rapid spine delivery and redistribution of AMPA receptors after synaptic NMDA receptor activation. Science 284, 1811–1816. Silva, A.J., Stevens, C.F., Tonegawa, S., Wang, Y., 1992. Deficient hippocampal long-term potentiation in alpha-calcium-calmodulin kinase II mutant mice. Science 257, 201–206. Strominger, R.N., Woolsey, T.A., 1987. Templates for locating the whisker area in a fresh flattened mouse and rat cortex. Journal of Neuroscience Methods 22, 113–118. Wallace, H., Liming, K., Glazewski, S., Fox, K., 2001. The role of cortical activity in experience-dependent potentiation and depression of sensory responses in barrel field. Journal of Neuroscience 21 (11), 3881–3894.
778
S. Glazewski et al. / Neuropharmacology 41 (2001) 771–778
Wong-Riley, M., 1979. Changes in visual system of monocularly sutured or enucleated cats demonstrable with cytochrome oxidase histochemistry. Brain Research 171, 11–28. Wu, G.Y., Cline, H.T., 1998. Stabilisation of dendritic arbour structure in vivo by CaMKII. Science 279, 222–226.
Wu, G., Malinow, R., Cline, H.T., 1996. Maturation of a central glutamatergic synapse. Science 274, 972–976. Zou, D.J., Cline, H.T., 1996. Expression of constitutively active CaMKII in target tissue modifies presynaptic axon arbour growth. Neuron 16, 529–539.