Learning-induced remodelling of inhibitory synapses in the motor cortex

Abstract

Robust structural and functional plasticity occurs at excitatory synapses in the motor cortex in response to learning. It is well established that local spinogenesis and the subsequent maintenance of newly formed spines are crucial for motor learning. However, despite local synaptic inhibition being essential for shaping excitatory synaptic input, less is known about the structural rearrangement of inhibitory synapses following learning. In this study, we co-expressed the structural marker tdTomato and a mEmeraldtagged intrabody against gephyrin to visualize inhibitory synapses in layer 2/3 cortical neurons of wild-type CD1 mice. We found that a 1-day accelerated rotarod paradigm induced robust motor learning in male and female adult CD1 mice. Histological analyses revealed a significant increase in the surface area of gephyrin puncta in neurons within the motor cortex but not in the somatosensory cortex upon motor learning. Furthermore, this learning-induced reorganization of inhibitory synapses only occurred in dendritic shafts and not in the spines. These data suggest that learning induces experience-dependent remodelling of existing inhibitory synapses to fine-tune intrinsic plasticity and input-specific modulation of excitatory connections in the motor cortex.

Headnote

Robust structural and functional plasticity occurs at excitatory synapses in the motor cortex in response to learning. It is well established that local spinogenesis and the subsequent maintenance of newly formed spines are crucial for motor learning. However, despite local synaptic inhibition being essential for shaping excitatory synaptic input, less is known about the structural rearrangement of inhibitory synapses following learning. In this study, we co-expressed the structural marker tdTomato and a mEmeraldtagged intrabody against gephyrin to visualize inhibitory synapses in layer 2/3 cortical neurons of wild-type CD1 mice. We found that a 1-day accelerated rotarod paradigm induced robust motor learning in male and female adult CD1 mice. Histological analyses revealed a significant increase in the surface area of gephyrin puncta in neurons within the motor cortex but not in the somatosensory cortex upon motor learning. Furthermore, this learning-induced reorganization of inhibitory synapses only occurred in dendritic shafts and not in the spines. These data suggest that learning induces experience-dependent remodelling of existing inhibitory synapses to fine-tune intrinsic plasticity and input-specific modulation of excitatory connections in the motor cortex.

Keywords: learning, plasticity, inhibition, postsynaptic density, gephyrin

1. Introduction

The precise arrangement of neurotransmitter receptors at the postsynaptic membrane is crucial for efficient synaptic transmission and plasticity. Modular scaffolding molecules such as postsynaptic density protein 95 (PSD-95) and gephyrin are critical in anchoring and clustering neurotransmitter receptors, concentrating these receptors at excitatory and inhibitory synapses, respectively [1-3]. Activity-dependent remodelling of postsynaptic scaffolding molecules contributes to the plasticity of functional synapses during development and in response to changes in patterns of neural activity or sensory experiences [4,5]. This is particularly evident in excitatory synapses, where dendritic spines undergo extensive structural remodelling during synaptic plasticity and learning [6-13]. It is well established that the spine volume positively correlates with the area of PSD and a-amino-3hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptor content and, hence, synaptic strength [8,10,14,15]. Although stimulated spines undergo rapid growth, the addition of PSD-95 molecules occurs at a slower time scale [16-19]. The gradual increase of PSD-95 in size and complexity promotes spine stability and precise alignment with presynaptic structures [19-23].

Accumulating evidence has demonstrated that the plasticity of inhibitory synapses plays essential roles in fine-tuning synaptic efficacy, refining neural circuitry and maintaining network synchronization and stability [24,25]. Major inhibitory neurotransmitter receptors, including y-aminobutyric acid type-A (GABAA) and glycine receptors, are clustered at the synapse and anchored to the cytoskeleton by the postsynaptic scaffolding molecule, gephyrin [3,26]. Similar to PSD-95 and AMPA receptors, gephyrin and GABA receptors are organized into subsynaptic nanodomains in neurons [27-29]. Activity-dependent redistribution of gephyrin molecules into and out of inhibitory PSDs is a major regulatory process that determines the dynamic changes in the strength of inhibitory neurotransmission during bidirectional synaptic plasticity [30-36]. Importantly, long-term in vivo imaging of inhibitory synapses in the layer 2/3 neurons of the mouse visual cortex has also revealed the dynamic nature of gephyrin clusters that continuously assemble and dissemble on the dendritic shaft and dually innervated dendritic spines [37-39]. However, little is known about the structural reorganization of inhibitory synapses in the context of learning.

Here, we expressed a genetically encoded fibronectin intrabody generated by mRNA display (FingR) against gephyrin [40], fused with the fluorescent protein mEmerald [41], to label inhibitory synapses along the dendritic regions of layer 2/3 cortical neurons in CD1 mice. We found that a 1-day rotarod motor learning task induced a robust increase in the size of gephyrin puncta in the dendritic shaft but not in the dendritic spine of neurons in the motor cortex. The overall gephyrin puncta density was not altered. Importantly, these effects were only observed in the motor cortex and not in the somatosensory cortex of the trained mice. Taken together, our results demonstrate that the PSD of inhibitory synapses undergoes robust structural remodelling following motor skill acquisition, which has important implications in neuronal plasticity, and the formation and refinement of the neural circuitry underlying learned movements.

2. Results

2.1. Visualization of endogenous gephyrin puncta in cortical neurons of (D1 mice

Fluorescent protein tagging is the most conventional method for investigating protein localization and dynamics in cells. However, overexpression of scaffolding proteins in the PSD often results in the alteration of synaptic function. For example, overexpression of PSD-95 causes enhanced maturation of excitatory synapses by increasing the size and numbers of dendritic spines and synaptic AMPA receptors [42]. Likewise, overexpression of gephyrin has also been reported to affect the density and size of gephyrin puncta and glutamatergic synapses [43]. To avoid these undesirable side effects, we visualized inhibitory synapses by expressing a recombinant FingR intrabody that recognizes endogenous gephyrin (Gphn.FingR) and fused it with a fluorescent protein [40], as it is well established that overexpression of Gphn.FingR does not affect the size of the gephyrin puncta and the electrophysiological properties of neurons [40]. To enhance the brightness of the existing green fluorescent protein (GFP)-tagged Gphn.FingR, we generated a construct that encodes Gphn.FingR-mEmerald fusion protein on a vector that contains the CCR5 zinc finger and KRAB(A) repressor transcriptional control system [40]. When transfected into primary hippocampal neurons in culture, we observed perfect colocalization of Gphn.FingR-mEmerald-CCR5TC with endogenous gephyrin but not with PSD-95 puncta, which confirmed the specificity of this construct (figure 1a).

To visualize gephyrin-positive inhibitory synapses in vivo, we co-expressed two plasmids that encode Gphn.FingR-mEmerald-CCR5TC and the structural marker tdTomato via in utero electroporation in CD1 mice at embryonic day 15 (E15). Histological analyses of the adult electroporated mice validated the expression of these constructs in layer 2/3 neurons in the motor and somatosensory cortices (figure 1b). Moreover, the visualization of these histological sections on a spinning disc confocal microscope with higher magnification revealed that most of the Gphn.FingR-mEmerald-CCR5TC puncta reside along the dendritic shaft (figure 1c). Unbound intrabodies, which were in excess, localized in the nucleus, where they turn off the transcription of new transcripts encoding Gphn.FingR (figure 1c). These data demonstrate the utility of the Gphn.FingR- mEmerald-CCR5TC construct to mark inhibitory synapses both in vitro and in vivo, which is consistent with results reported in a recent study that used Gphn.FingR-msScarlett [45].

2.2. Motor learning induces structural remodelling of inhibitory synapses

To study the effect of acute learning on the structure of inhibitory synapses, we quantified the endogenous gephyrin puncta in layer 2/3 cortical neurons of in utero electroporated CD1 mice co-expressing Gphn.FingR-mEmerald-CCR5TC and tdTomato (figure 2a). We chose motor learning as an experimental model due to the presence of excitatory synapse remodelling and dendritic spine dynamics in the motor cortex and their essential roles for improved motor performance after learning [11-13,46]. We adopted the accelerated rotarod task in which mice were trained to maintain their balance on a rotating rod with increasing acceleration over 20 sessions in a single day. At the end of the training sessions, the mice showed a significant increase in the latency to fall off the accelerated rotarod compared to untrained mice (figure 2b). This improved motor performance was observed in both genders (figure 2c), demonstrating that the single-day accelerated rotarod task is a robust motor skill learning paradigm in adult CD1 mice.

Next, we performed histological analyses on the brains of adult CD1 mice and quantified the number and size of Gphn.FingR-mEmerald puncta in the principal neurons within layer 2/3 of the motor cortex of the trained and untrained animals (figure 2d). We observed a significant increase in the surface area of individual gephyrin puncta (figure 2e) across multiple neurons (figure 2f) in the motor cortex of rotarod-trained mice, compared to naive animals. We also saw a significant increase in the volume of individual gephyrin puncta (figure 2g). However, it did not reach statistical significance when we compared the mean volume across neurons (р = 0.06, figure 2/1). Despite the increase in the size of inhibitory synapses following motor learning, the number of gephyrin puncta did not appear to change (figure 2i). These effects were specific to neurons in the motor cortex as they were not observed in somatosensory cortical neurons (figure 2j-0).

Although most inhibitory synapses are found on the soma and dendritic shafts, inhibitory nerve terminals also innervate the dendritic spines of neocortical and hippocampal neurons [37,45,47,48]. Indeed, we were able to observe Gphn.FingR-mEmerald labelling within dendritic spines (figure 3a), which accounted for approximately 10% of the total gephyrin puncta in cortical neurons (figure 3b). To determine the effects of learning on inhibitory synapse remodelling in the dually innervated spines, we performed analyses of Gphn.FingR-mEmerald puncta on the dendritic shaft and spine separately. As expected, we found a significant increase in the surface area (figure 3c) and a trend towards larger volume (figure 3d, p = 0.08) of gephyrin puncta on the dendritic shafts of neurons in the motor cortex in response to acute motor learning. Although previous studies have shown that inhibitory synapses on spines are slightly more dynamic (assembled and removed) than shaft inhibitory synapses or dendritic spines, and that this process is responsive to sensory experience [37,38,49], we did not see any significant changes in the surface area or volume of gephyrin puncta on dendritic spines (figure 3 ,f). However, we cannot exclude the possibility that transient changes of gephyrin dynamics in these dually innervated spines occur outside our observation point (24 hours) before or after motor learning, as this might allow flexible, input-specific modulation of stable excitatory synapses.

3. Discussion

It is well established that the structural reorganization of synapses and active remodelling of the PSD contribute to changes in neuronal plasticity, functional connectivity and circuit refinement during learning [9,50,51]. In the context of motor learning, the formation of spatially clustered dendritic spines provides a means for the local reorganization of the excitatory circuit in layer 2/3 of the motor cortex [13,46,52]. Widespread modulation of inhibitory neurons in the cortex during motor learning has also been reported [52,53]. In this study, we asked whether learning could induce structural remodelling of inhibitory PSDs in the motor cortex following an accelerated rotarod paradigm. By labelling endogenous gephyrin puncta with the genetically encoded intrabody Gphn.FingR-mEmerald, we found that motor learning increased the surface area of gephyrin puncta without affecting their overall density in the dendritic shaft of layer 2/3 neurons in the motor cortex. The learning-related changes in the size of inhibitory PSDs were system-specific as they did not occur in layer 2/3 neurons of the somatosensory cortex.

Previous studies have shown that cortical disinhibition contributes to the expression of structural plasticity in excitatory neurons [52-55], which is associated with a significant reduction in GABA release in the human primary motor cortex [56,57]. In rodents, cortical disinhibition during the initial learning phase is mediated by reduced inhibition by somatostatin- and parvalbumin-positive interneurons onto the apical dendrites and perisomatic regions of layer 2/3 motor neurons, respectively [52,54,55]. Our results might initially seem counterintuitive because the increase in gephyrin puncta size has previously been shown to correlate well with the strength of GABA receptor-mediated currents [30,31,34]. However, a more recent study has shown that an acute augmentation of gephyrin surface area disperses the localization of synaptic GABAA receptors and disrupts the precise nanocolumn alignment at the inhibitory synapses, thereby reducing inhibitory synaptic neurotransmission [58]. The next logical step in testing this hypothesis is to visualize the nanoscale organization of GABA, receptors within these enlarged gephyrin puncta associated with motor learning.

Cortical inhibition during learning is highly dynamic. After the initial learning phase, the excitatory drive onto parvalbuminexpressing interneurons increases [55]. This post-learning hyperexcitation may function to limit excessive structural plasticity and preserve learned information [59-61]. It could also contribute to restoring the balance between excitation and inhibition of the neural network [62]. These two functions are complementary and are not mutually exclusive. It is conceivable that the increase in gephyrin surface area following a single day of rotarod learning may contribute to enhanced cortical inhibition to subserve those functions. Therefore, it is critical that future experiments are focused on measuring inhibitory neurotransmission in the layer 2/3 cortical neurons in the motor cortex 24 hours post-rotarod learning.

In summary, our results provide empirical evidence for learning-induced structural remodelling of inhibitory synapses in the mouse motor cortex following a day of accelerated motor learning. These structural changes after short-term motor learning might reflect the learning-induced modulation of neuronal inhibition as an adaptive state of a dynamic cortical network. Our method should be applicable for future experiments to monitor the structural rearrangement of inhibitory PSDs during the initial learning phase and for tracking their dynamics over multiple days when combined with in vivo two-photon imaging in awake mice. As Gphn.FingR-mEmerald is a genetically encoded intrabody, it can be expressed in a distinct population of neurons with cell-specific promoters to gain deeper insights into the spatial and temporal regulation of structural plasticity of inhibitory PSDs in the brain during learning and memory processes.

4. Material and methods

4.1. Animals

Wild-type CD1 mice were purchased from the University of Queensland Biological Resources. Timed-mating was set up by placing CD1 male and female mice in the same cage overnight. When a vaginal plug was observed, it was considered the first embryonic day (Е0). All animals were housed in a 20°C room on a 12 hour day-and-night cycle. Food and water were supplied ad libitum. All research procedures involving animal use were conducted according to the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes and were approved by the University of Queensland Animal Ethics Committee (2021/AE000100).

4.2. DNA constructs

The pCAG-Gphn.FingR-mEmerald-CCR5TC construct was generated from pCAG-PSD-95.FingR-eGFP-CCR5TC (Addgene 46295) and pCAG-Gphn.FingR-mKate2-IL2RGTC (Addgene 46295), both of which were gifts from Don Arnold [40]. We first replaced the eGFP in pCAG-PSD-95.FingR-eGFP-CCR5TC to generate pCAG-PSD-95.FingR-mEmerald-CCR5TC. We then removed the PSD-95.FingR and subsequently subcloned Gphn.FingR into the parent vector to create pCAG-Gphn.FingRmEmerald-CCR5TC. The pCAG-tdTomato construct was a gift from Patricio Opazo.

4.3. Antibodies

The following antibodies were obtained from commercial sources: mouse anti-gephyrin (Synaptic Systems Cat. no. 147 011, RRID:AB_887717, used at 1:50 dilution), chicken anti-MAP2 (Abcam Cat. no. ab92434, RRID:AB_2138147, used at 1:2000 dilution), and mouse anti-PSD-95 (BioLegend Cat. no. 810401, RRID:AB_2564750, used at 1:500 dilution). Alexa-conjugated secondary antibodies were purchased from Invitrogen and used at 1 : 500 dilution.

4.4. In utero electroporation

In order to label layer 2/3 neocortical neurons, time-mated CD1 pregnant dams were electroporated at embryonic stage (E) 15 for all experiments [63]. Briefly, the pregnant dams were anaesthetized with an intraperitoneal injection of ketamine (120 mg kg) and xylazine (10 mg kg"). The embryos were then exposed from the abdominal cavity by laparotomy under sterile conditions. Each embryo was positioned so that the lateral telencephalic ventricles were visible through the uterine wall. One microlitre of plasmid solution (containing 0.5 ug of pCAG-Gphn.FingR-mEmerald-CCR5TC, 0.5 ug pCAG-tdTomato and 0.0025% Fast Green dye) was injected into the lateral ventricles with glass-pulled micropipettes and a Picospritzer. Five 100 ms square pulses of 30-35 V were delivered at 1 Hz with 1 to 3 mm paddles (ECM 803, BTX). Once this procedure was completed, the uterine horns were replaced inside the abdominal cavity, and the incision was sutured twice. Animals were then subcutaneously injected with 1 ml of sterile saline solution and recovered in a humidified chamber at approximately 28°C. For pain relief, an edible buprenorphine gel pack was prepared by injecting MediGel with 0.2 ml of buprenorphine solution (0.026 mg ml") and positioned next to the recovering dam in the cage. Dams were then monitored daily until they gave birth to live pups. The fluorescent patches in the cortex of pups at postnatal day 2 were checked using fluorescence goggles. Only pups with fluorescent patches were kept until they reached the desired experimental age.

4.5. Accelerated rotarod task

Adult CD1 mice (2-6 months old) were trained on an accelerated rotarod system (Ugo Basile" SRL) to induce motor learning according to a protocol adapted from [11] and performed during daytime (light phase). Before training began, the mice were habituated to their cage and behaviour room for 30 min and then on a stationary rotarod apparatus for 2 min. During the training sessions, mice were first exposed to the fixed-speed protocol (8 r.p.m.), allowing them to acclimatize to the rotating movement as they rarely fall off the rod at this relatively low speed. Following this, the speed of the rotarod was increased stepwise from 4 to 60 r.p.m. over 2 min. During the training phase, the mice were gently tapped on the hind limbs with a plastic Pasteur pipette to teach them to face forward on the rotating rod, helping them to stay on the rotarod at higher speeds. After falling, the mice were immediately placed back on the rotarod. This 1-day protocol involved 20 training sessions, split into two training blocks with a 5 min interval. To assess motor learning, each mouse underwent three trials of the acceleration protocol (5 min inter-trial intervals), and the latency to fall was recorded.

4.6. Histology

Following deep anaesthesia, mice were transcardially perfused with 0.9% saline followed by a 4% paraformaldehyde (PFA) solution prepared in 0.1 M phosphate-buffered saline (PBS). The brains were post-fixed in 4% PFA solution for 48 hours and stored in 0.1 M PBS containing 1% sodium azide at 4°C. They were then embedded in 3.5% agarose and sectioned into 50 pm coronal slices using a vibratome (Leica VT 1000S). Coronal brain sections were collected and washed three times in 0.1 M PBS and stored in 0.1 M PBS containing 1% sodium azide at 4°C. They were then mounted on positively charged slides (UberFrost®, InstrumeC), dried and post-fixed with 4% PFA for 10 min. After washing with PBS, the mounted sections were stained with 0.1% 4',6-diamidino-2"-phenylindole (DAPI) solution for 10 min, washed and coverslipped (Menzel-Glaser) with anti-fade mounting medium (Dako).

4.7. Microscopy and image analyses

Wide-field fluorescence imaging was performed using a Zeiss upright Axio-Imager Z1 fluorescence microscope on a 10x Plan-Apochromat objective. High-resolution fluorescence images were acquired with a Diskovery spinning disc confocal microscope at various magnifications using 40x water immersion and 60x oil immersion objectives. Individual dendrites from the 60x images were carefully traced for analysis using Neurolucida 360 software (MBF Bioscience, RRID:SCR_016788). The entire dendritic trees of individual neurons were reconstructed in three dimensions from the series of 50 um coronal sections. Individual dendritic spines were semi-automatically detected and classified across all image traces using identical detection parameters (table 1). Similarly, individual gephyrin puncta were detected automatically in all traces using consistent parameters.

4.8. Primary neuronal cultures and transfection

Sprague-Dawley rat embryos (E18) were used to prepare primary hippocampal neurons as described previously [64]. Briefly, hippocampi were isolated and dissociated with 30 U of papain suspension (Worthington) for 20 min in a 37°C water bath. A single-cell suspension was obtained by triturating tissues with fire-polished glass Pasteur pipettes and then plated at a density of 8 x 10 cells on a poly-L-lysine-coated coverslip in Neurobasal growth medium supplemented with 2 mM Glutamax, 1% penicillin/streptomycin and 2% B-27. Neurons were then maintained in a humidified 5% CO) tissue culture incubator at 37°C and fed twice a week with the Neurobasal growth medium. Hippocampal neurons were transfected at days in vitro (DIV) 13-14 with Lipofectamine 2000 (Invitrogen) and processed at DIV 15-16.

4.9. Immunostaining

Coverslips containing transfected hippocampal neurons were washed in PBS and fixed with a 4% parafix solution (4% PFA and 10% sucrose in 1x PBS) for 10 min at room temperature. After washing, they were permeabilized in 0.25% Triton-X (in PBS) for 10 min, washed with PBS and blocked for 1 hour in 10% normal goat serum (NGS in PBS). The neurons were then incubated with primary antibodies in 3% NGS solution at 4°C overnight, followed by secondary antibodies in 3% NGS solution for 1 hour at room temperature. After washing, the coverslips were mounted onto glass slides with an anti-fade solution. Images were collected with a 63x oil immersion objective on a Zeiss LSM510 confocal microscope.

4.10. Statistics

The sample size in the figure legends represents individual puncta or neurons generated from 4 electroporated CD1 mice. Statistical analysis was performed in GraphPad Prism 10.0 using a two-tailed unpaired t-test or the non-parametric Mann-- Whitney test. For all statistical tests, р <0.05 was considered significant. Unless stated otherwise, all data are reported as mean + standard error of the mean (s.e.m.).

Ethics. All research procedures involving animal use were conducted according to the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes and were approved by the University of Queensland Animal Ethics Committee (2021/AE000100). Data accessibility. The data that support the findings of this study are available from the corresponding author upon reasonable request. Declaration of Al use. We have not used Al-assisted technologies in creating this article. Authors' contributions. N.B.: conceptualization, formal analysis, investigation, methodology, visualization, writing- original draft, writing-review and editing; A.P.: conceptualization, formal analysis, investigation, methodology, writing-review and editing; S.S.D.: methodology, resources, validation; S.G.: investigation; L.R.F.: methodology; V.A.: conceptualization, funding acquisition, supervision, visualization, writing-review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein. Conflict of interest declaration. We declare we have no competing interests. Funding. This work was supported by the Australian Medical Research Future Fund (Clem Jones Centre for Ageing Dementia Research Flagship Project Grant to V.A.). М.А. holds an Australian Research Council (ARC) Future Fellowship (ЕТ220100485). S.S.D. was supported by a University of Queensland Research Training Scholarship. Acknowledgements. We thank staff members from the Queensland Brain Institute (QBI) Animal House and Behavioural Facility and the QBI Advanced Microscopy Facility for their professional support, Rowan Tweedale for editing the manuscript and Nela Durisic for providing the gephyrin antibodies. Image acquisition and analysis were performed at the QBI Advanced Microscopy Facility using the Neurolucida 360 software generously supported by the Australian Government through an Australian Research Council (ARC) ШЕЕ grant (LE100100074).

Sidebar

Cite this article: Bhembre N, Paolino A, Das SS, Guntupalli 5, Fenlon LR, Anggono V. 2024 Learning-induced remodelling of inhibitory synapses in the motor cortex. Open Biol. 14: 240109.

Received: 29 April 2024

Accepted: 26 September 2024

Subject Areas: neuroscience, cellular biology

References

References

1. Sheng M, Kim E. 2011 The postsynaptic organization of synapses. Cold Spring Harb. Perspect. Biol. 3, 005678. (doi:10.1101/cshperspect.a005678)

2. Droogers WJ, MacGillavry HD. 2023 Plasticity of postsynaptic nanostructure. Mol. Cell. Neurosci. 124, 103819. (doi:10.1016/j.mcn.2023.103819)

3. Pizzarelli В, Griguoli M, Zacchi P, Petrini EM, Barberis A, Cattaneo A, Cherubini Е. 2020 Tuning GABAergic inhibition: gephyrin molecular organization and functions. Neuroscience 439, 125-136. (doi:10.1016/j.neuroscience.2019.07.036)

4. Choquet D, Triller A. 2013 The dynamic synapse. Neuron 80, 691-703. (doi:10.1016/j.neuron.2013.10.013)

5. Chen H, Tang AH, Blanpied TA. 2018 Subsynaptic spatial organization as a regulator of synaptic strength and plasticity. Curr. Opin. Neurobiol. 51, 147-153. (doi:10.1016/j.conb.2018. 05.004)

6. Trachtenberg JT, Chen BE, Knott GW, Feng G, Sanes JR, Welker E, Svoboda K. 2002 Long-term in vivo imaging of experience-dependent synaptic plasticity in adult cortex. Nature 420, 788-794. (doi:10.1038/nature01273)

7. Holtmaat A, Wilbrecht L, Knott GW, Welker E, Svoboda К. 2006 Experience-dependent and cell-type-specific spine growth in the neocortex. Nature 441, 979-983. (doi:10.1038/ nature04783)

8. Matsuzaki M, Honkura N, Ellis-Davies GCR, Kasai H. 2004 Structural basis of long-term potentiation in single dendritic spines. Nature 429, 761-766. (doi:10.1038/nature02617)

9. Mas, Zuo Y. 2022 Synaptic modifications in learning and memory: a dendritic spine story. Semin. Cell Dev. Biol. 125, 84-90. (doi:10.1016/j.semcdb.2021.05.015)

10. Holtmaat A, Svoboda К. 2009 Experience-dependent structural synaptic plasticity in the mammalian brain. Nat. Rev. Neurosci. 10, 647-658. (doi:10.1038/nrn2699)

11. Hayashi-Takagi A, Yagishita S, Nakamura M, Shirai Е, Wu YI, Loshbaugh AL, Kuhlman В, Hahn KM, Kasai H. 2015 Labelling and optical erasure of synaptic memory traces in the motor cortex. Nature 525, 333-338. (doi:10.1038/nature 15257)

12. Yang G, Pan Е, Gan WB. 2009 Stably maintained dendritic spines are associated with lifelong memories. Nature 462, 920-924. (doi:10.1038/nature08577)

13. XuT, Yu X, Perlik AJ, Tobin WF, Zweig JA, Tennant К, Jones T, Zuo Y. 2009 Rapid formation and selective stabilization of synapses for enduring motor memories. Nature 462, 915- 919. (doi:10.1038/nature08389)

14. Holler S, Kóstinger G, Martin KAC, Schuhknecht GFP, Stratford KJ. 2021 Structure and function of a neocortical synapse. Nature 591, 111-116. (doi:10.1038/s41586-020-03134-2)

15. Arellano JI, Benavides-Piccione В, Defelipe J, Yuste В. 2007 Ultrastructure of dendritic spines: correlation between synaptic and spine morphologies. Front. Neurosci. 1, 131-143. (doi:10.3389/neuro.01.1.1.010.2007)

16. Meyer D, Bonhoeffer T, Scheuss V. 2014 Balance and stability of synaptic structures during synaptic plasticity. Neuron 82, 430-443. (doi:10.1016/j.neuron.2014.02.031)

17. Bosch M, Castro J, Saneyoshi T, Matsuno H, Sur M, Hayashi Y. 2014 Structural and molecular remodeling of dendritic spine substructures during long-term potentiation. Neuron 82, 444-459. (doi:10.1016/j.neuron.2014.03.021)

18. Sun Y, Smirnov M, Kamasawa N, Yasuda R. 2021 Rapid ultrastructural changes in the PSD and surrounding membrane after induction of structural LTP in single dendritic spines. J. Neurosci. 41, 7003-7014. (doi:10.1523/JNEUROSCI.1964-20.2021)

19. Wegner W, Steffens H, Gregor C, Wolf F, Willig KI. 2022 Environmental enrichment enhances patterning and remodeling of synaptic nanoarchitecture as revealed by STED nanoscopy. elife 11, e73603. (doi:10.7554/eLife.73603)

20. Hruska M, Henderson ? Le Marchand SJ, Jafri H, Dalva MB. 2018 Synaptic nanomodules underlie the organization and plasticity of spine synapses. Nat. Neurosci. 21, 671-682. (doi: 10.1038/541593-018-0138-9)

21. Cane M, Maco B, Knott G, Holtmaat A. 2014 The relationship between PSD-95 clustering and spine stability in vivo. J. Neurosci. 34, 2075-2086. (doi:10.1523/JNEUROSCI.3353-13. 2014)

22. Clavet-Fournier V, Lee C, Wegner W, Brose N, Rhee J, Willig KI. 2024 Pre- and postsynaptic nanostructures increase in size and complexity after induction of long-term potentiation. ¡Science 27, 108679. (doi:10.1016/j.isci.2023.108679)

23. Tang AH, Chen H, Li TP, Metzbower SR, MacGillavry HD, Blanpied TA. 2016 A trans-synaptic nanocolumn aligns neurotransmitter release to receptors. Nature 536, 210-214. (doi:10. 1038/nature19058)

24. Kullmann DM, Moreau AW, Bakiri Y, Nicholson E. 2012 Plasticity of inhibition. Neuron 75, 951-962. (doi:10.1016/j.neuron.2012.07.030)

25. Capogna M, Castillo PE, Maffei A. 2021 The ins and outs of inhibitory synaptic plasticity: neuron types, molecular mechanisms and functional roles. Eur. J. Neurosci. 54, 6882-6901. (doi:10.1111/ejn.14907)

26. Tyagarajan SK, Fritschy JM. 2014 Gephyrin: a master regulator of neuronal function? Nat. Rev. Neurosci. 15, 141-156. (doi:10.1038/nrn3670)

27. Crosby KC, Gookin SE, Garcia JD, Hahm KM, Dell'Acqua ML, Smith KR. 2019 Nanoscale subsynaptic domains underlie the organization of the inhibitory synapse. Cell Rep. 26, 3284- 3297.(doi:10.1016/j.celrep.2019.02.070)

28. Nair D, Hosy E, Petersen JD, Constals A, Giannone G, Choquet D, Sibarita JB. 2013 Super-resolution imaging reveals that AMPA receptors inside synapses are dynamically organized in nanodomains regulated by PSD95. J. Neurosci. 33, 13204-13224. (doi:10.1523/JNEUROSCI.2381-12.2013)

29. MacGillavry HD, Song Y, Raghavachari S, Blanpied TA. 2013 Nanoscale scaffolding domains within the postsynaptic density concentrate synaptic AMPA receptors. Neuron 78, 615- 622. (doi:10.1016/j.neuron.2013.03.009)

30. Petrini EM et al. 2014 Synaptic recruitment of gephyrin regulates surface GABAA receptor dynamics for the expression of inhibitory LTP. Nat. Commun. 5, 3921. (doi:10.1038/ ncomms4921)

31. Flores CE, Nikonenko I, Mendez P, Fritschy JM, Tyagarajan SK, Muller D. 2015 Activity-dependent inhibitory synapse remodeling through gephyrin phosphorylation. Proc. Nat! Acad. Sci. USA 112, E65-E72. (doi:10.1073/pnas.1411170112)

32. Pennacchietti F et al. 2017 Nanoscale molecular reorganization of the inhibitory postsynaptic density is a determinant of gabaergic synaptic potentiation. J. Neurosci. 37, 1747- 1756. (doi:10.1523/JNEUROSCI.0514-16.2016)

33. Hanus C, Ehrensperger MV, Triller A. 2006 Activity-dependent movements of postsynaptic scaffolds at inhibitory synapses. J. Neurosci. 26, 4586-4595. (doi:10.1523/JNEUROSCI. 5123-05.2006)

34. Muir J, Arancibia-Carcamo IL, MacAskill AF, Smith KR, Griffin LD, Kittler JT. 2010 NMDA receptors regulate GABAA receptor lateral mobility and clustering at inhibitory synapses through serine 327 on the y2 subunit. Proc. Nat! Acad. Sci. USA 107, 16679-16684. (doi:10.1073/pnas. 1000589107)

35. Bannai H et al. 2015 Bidirectional control of synaptic GABAAR clustering by glutamate and calcium. Cell Rep. 13, 2768-2780. (doi:10.1016/j.celrep.2015.12.002)

36. Battaglia S, Renner M, Russeau M, Côme E, Tyagarajan SK, Lévi S. 2018 Activity-dependent inhibitory synapse scaling is determined by gephyrin phosphorylation and subsequent regulation of GABA(A) receptor diffusion. eNeuro 5, e0203-0217.2017. (doi:10.1523/ENEURO.0203-17.2017)

37. Chen JL, Villa KL, Cha JW, So PTC, Kubota Y, Nedivi E. 2012 Clustered dynamics of inhibitory synapses and dendritic spines in the adult neocortex. Neuron 74, 361-373. (doi: 10. 1016/j.neuron.2012.02.030)

38. Villa KL, Berry KP, Subramanian J, Cha JW, Oh WC, Kwon HB, Kubota Y, So PTC, Nedivi E. 2016 Inhibitory synapses are repeatedly assembled and removed at persistent sites in vivo. Neuron 89, 756-769. (doi:10.1016/j.neuron.2016.01.010)

39. van Versendaal D et al. 2012 Elimination of inhibitory synapses is a major component of adult ocular dominance plasticity. Neuron 74, 374-383. (doi:10.1016/j.neuron.2012.03. 015)

40. Gross GG et al. 2013 Recombinant probes for visualizing endogenous synaptic proteins in living neurons. Neuron 78, 971-985. (doi:10.1016/j.neuron.2013.04.017)

41. Balleza E, Kim JM, Cluzel P. 2018 Systematic characterization of maturation time of fluorescent proteins in living cells. Nat. Methods 15, 47-51. (doi:10.1038/nmeth.4509)

42. El-Husseini AE, Schnell E, Chetkovich DM, Nicoll RA, Bredt DS. 2000 PSD-95 involvement in maturation of excitatory synapses. Science 290, 1364-1368.

43. Yu W, Blas ALD. 2008 Gephyrin expression and clustering affects the size of glutamatergic synaptic contacts. J. Neurochem. 104, 830-845. (doi:10.1111/j.1471-4159.2007.05014.x)

44. Franklin KBJ, Paxinos G. 2007 The mouse brain in stereotaxic coordinates, 3rd edn. London, UK: Academic Press.

45. Kleinjan MS et al. 2023 Dually innervated dendritic spines develop in the absence of excitatory activity and resist plasticity through tonic inhibitory crosstalk. Neuron 111, 362-371. (doi:10.1016/j.neuron.2022.11.002)

46. Peters AJ, Chen SX, Komiyama T. 2014 Emergence of reproducible spatiotemporal activity during motor learning. Nature 510, 263-267. (doi:10.1038/nature13235)

47. Kubota Y, Hatada S, Kondo S, Karube F, Kawaguchi Y. 2007 Neocortical inhibitory terminals innervate dendritic spines targeted by thalamocortical afferents. J. Neurosci. 27, 1139- 1150. (doi:10.1523/JNEUROSCI.3846-06.2007)

48. Chiu CQ, Lur G, Morse TM, Carnevale NT, Ellis-Davies GCR, Higley MJ. 2013 Compartmentalization of GABAergic inhibition by dendritic spines. Science 340, 759-762. (doi:10.1126/ science.1234274)

49. Knott GW, Quairiaux C, Genoud C, Welker E. 2002 Formation of dendritic spines with GABAergic synapses induced by whisker stimulation in adult mice. Neuron 34, 265-273. (doi: 10.1016/50896-6273(02)00663-3)

50. Caroni P, Donato F, Muller D. 2012 Structural plasticity upon learning: regulation and functions. Nat. Rev. Neurosci. 13, 478-490. (doi:10.1038/nrn3258)

51. Peters AJ, Liu H, Komiyama T. 2017 Learning in the rodent motor cortex. Annu. Rev. Neurosci. 40, 77-97. (doi:10.1146/annurev-neuro-072116-031407)

52. Chen SX, Kim AN, Peters AJ, Komiyama T. 2015 Subtype-specific plasticity of inhibitory circuits in motor cortex during motor learning. Nat. Neurosci. 18, 1109-1115. (doi:10.1038/ nn.4049)

53. Ren C, Peng K, Yang R, Liu W, Liu C, Komiyama T. 2022 Global and subtype-specific modulation of cortical inhibitory neurons regulated by acetylcholine during motor learning. Neuron 110, 2334-2350.(doi:10.1016/j.neuron.2022.04.031)

54. Hattori R, Kuchibhotla KV, Froemke RC, Komiyama T. 2017 Functions and dysfunctions of neocortical inhibitory neuron subtypes. Nat. Neurosci. 20, 1199-1208. (doi:10.1038/nn. 4619)

55. Donato F, Rompani SB, Caroni P. 2013 Parvalbumin-expressing basket-cell network plasticity induced by experience regulates adult learning. Nature 504, 272-276. (doi:10.1038/ nature12866)

56. Kolasinski J, Hinson EL, Divanbeighi Zand AP, Rizov A, Emir UE, Stagg CJ. 2019 The dynamics of cortical GABA in human motor learning. J. Physiol. 597, 271-282. (doi:10.1113/ JP276626)

57. Stagg CJ, Bachtiar V, Johansen-Berg H. 2011 The role of GABA in human motor learning. Curr. Biol. 21, 480-484. (doi:10.1016/j.cub.2011.01.069)

58. Olah SS et al. 2023 Acute reorganization of postsynaptic GABA(A) receptors reveals the functional impact of molecular nanoarchitecture at inhibitory synapses. Cell Rep. 42, 113331. (doi:10.1016/j.celrep.2023.113331)

59. Cichon J, Gan WB. 2015 Branch-specific dendritic Ca" spikes cause persistent synaptic plasticity. Nature 520, 180-185. (doi:10.1038/nature14251)

60. Shibata K, Sasaki Y, Bang JW, Walsh EG, Machizawa MG, Tamaki M, Chang LH, Watanabe T. 2017 Overlearning hyperstabilizes a skill by rapidly making neurochemical processing inhibitory-dominant. Nat. Neurosci. 20, 470-475. (doi:10.1038/nn.4490)

61. Isaacson JS, Scanziani M. 2011 How inhibition shapes cortical activity. Neuron 72, 231-243. (doi:10.1016/j.neuron.2011.09.027)

62. Turrigiano G. 2011 Too many cooks? Intrinsic and synaptic homeostatic mechanisms in cortical circuit refinement. Annu. Rev. Neurosci. 34, 89-103. (doi:10.1146/annurev-neuro060909-153238)

63. Paolino A, Fenlon LR, Kozulin P, Haines E, Lim JWG, Richards LJ, Suárez R. 2020 Differential timing of a conserved transcriptional network underlies divergent cortical projection routes across mammalian brain evolution. Proc. Nat! Acad. Sci. USA 117, 10554-10564. (doi:10.1073/pnas. 1922422117)

64. Tan JZA et al. 2023 Copine-6 is a Ca·· sensor for activity-induced AMPA receptor exocytosis. Cell Rep. 42, 113460. (doi:10.1016/j.celrep.2023.113460)