And Postsynaptic Activity in Thalamocortical Axon Branching

Role of pre- and postsynaptic activity in thalamocortical axon branching

Akito Yamadaa,1,2, Naofumi Uesakaa,b,1, Yasufumi Hayanoa, Toshihide Tabatab,c, Masanobu Kanob, and Nobuhiko Yamamotoa,3

aNeuroscience Laboratories, Graduate School of Frontier Biosciences, Osaka University, 1-3 Yamadaoka, Suita, Osaka 565-0871, Japan; bDepartment of Neurophysiology, Graduate School of Medicine, University of Tokyo, 7-3-1, Hongo, Bunkyo-ku, Tokyo 113-0033, Japan; and cLaboratory for Neural Information Technology, Graduate School of Science and Engineering, University of Toyama, 3190 Gofuku, Toyama, Toyama 930-8555, Japan

Edited* by Carla J. Shatz, Stanford University, Stanford, CA, and approved March 9, 2010 (received for review January 21, 2009) Axonal branching is thought to be regulated not only by genetically axon branching is inhibited by reducing the activity at either of these defined programs but also by neural activity in the developing nervous locations, suggesting that both pre- and postsynaptic activity is system. Here we investigated the role of pre- and postsynaptic activity required for the development of TC axon branching. in axon branching in the thalamocortical (TC) projection using organotypic coculture preparations of the thalamus and cortex. Individual TC Results axons were labeled with enhanced yellow fluorescent protein by TC axon branching was studied in cocultures of the thalamus and transfection into thalamic neurons. To manipulate firing activity, a cortex by introducing a plasmid encoding enhanced yellow fluo- vector encoding an inward rectifying potassium channel (Kir2.1) was rescent protein (EYFP) into thalamic cells. Spontaneous firing introduced into either thalamic or cortical cells. Firing activity was activity was monitored with multielectrode dishes (MED) during monitored with multielectrode dishes during culturing. We found that culturing (Fig. 1A). As previously shown (18), spontaneous firing axon branching was markedly suppressed in Kir2.1-overexpressing (mostly field potential-like activity) developed during the second thalamic cells, in which neural activity was silenced. Similar suppres- week in vitro in both thalamic and cortical cells (1.34 ± 0.45 Hz in the sion of TC axon branching was also found when cortical cell activity cortical explant, n =4;0.96± 0.32 Hz in the thalamic explant, n =4; was reduced by expressing Kir2.1. These results indicate that both pre- Fig. 1 C and H and Table 1). Furthermore, cortical activity was highly and postsynaptic activity is required for TC axon branching during correlated with thalamic activity (Fig. 1C and Fig. S1). Most of the development. negative potentials measured in the cortex were synchronized with thalamic potentials (85.7 ± 8.5%, n = 4). In the presence of such development | neocortex | thalamus | neural activity | organotypic culture spontaneous activity, most EYFP-labeled axons formed elaborate branches (the number of branch points, 13.2 ± 2.5, n = 17) in cortical uring development, axons navigate to their target regions and explants after 2 weeks in vitro (Table 1; see also Figs. 2A and 3A). In ± n form elaborate branches when they make synaptic connections contrast, TC axons formed few branches (2.2 0.6, = 17) in D fi with multiple target cells. It has been demonstrated that axon guid- cocultures where all ring activity was blocked by tetrodotoxin (TTX) ance to the target region is regulated by attractive and repulsive application during the second week in vitro (Table 1 and Fig. S2) (18). fl molecular cues that are expressed in particular spatiotemporal pat- We then studied the in uence of pre- and postsynaptic activity on fi terns (1). Similar molecular mechanisms are thought to influence TC axon branching by speci cally reducing neural activity in either axon branching (2). In addition, neural activity such as firing and thalamic or cortical cells. First, axon branching was examined in synapticactivitycanalsoaffectbranchformation(3–5). An intrigu- Kir2.1-overexpressing thalamic cells. To observe axonal extension, the eyfp vector was cotransfected with a kir2.1 expression vector into ing and unanswered problem is how neural activity regulates axonal B branching. thalamic cells (Fig. 1 ). Immunohistochemistry with an antibody against a peptide tag indicated that more than 95% of EYFP- The thalamocortical (TC) projection in the mammalian cortex is a D–F well-characterized system in which to investigate activity-dependent labeled cells expressed Kir2.1 protein (Fig. 1 ). In most cases, up to 20 labeled cells (less than 0.2% of the total number of thalamic axon branching. In the developing sensory cortices, TC axons form B elaborate terminal arbors, whose size and complexity are altered by cells) were sparsely distributed in the thalamic explant (Fig. 1 ). Whole-cell perforated-patch recordings showed that spontaneous neural activity. In the primary visual cortex of higher mammals such action potential firing was dramatically diminished in Kir2.1-over- as cats, ferrets, and monkeys, TC axons serving left and right eyes are expressing thalamic cells (0.017 ± 0.01 Hz, n =3;Fig.1G and H and segregated into eye-specific stripes (6). This segregation is estab- Table 1), with hyperpolarized membrane potentials (Fig. S3A). In lished during development (7), but is disrupted by blockade of ret- addition, membrane resistance was substantially decreased (Fig. inal activity (8, 9). Regarding individual axon arbors, it is known that S3A), which may be attributable to an increase in conductance after monocular deprivation, TC axons serving the deprived and caused by overexpression of the potassium channel. However, nondeprived eyes shrink and expand their arbors, respectively (10, excitatory postsynaptic responses were evoked readily in Kir2.1- 11). A recent study in which monocular deprivation was combined overexpressing thalamic cells when recurrent corticothalamic fibers with silencing cortical activity has further suggested that correlations between pre- and postsynaptic activity play a dominant role in segregation of axon arbors (12). In accordance with this view, molecular Author contributions: A.Y., N.U., and N.Y. designed research; A.Y., N.U., Y.H., and N.Y. machinery in pre- and postsynaptic sites has also been shown to performed research; T.T. and M.K. contributed new reagents/analytic tools; A.Y., N.U., Y.H., affect arbor formation of TC axons in the somatosensory cortex (13– T.T., M.K., and N.Y. analyzed data; and A.Y., N.U., and N.Y. wrote the paper. 15). However, the relative role of pre- and postsynaptic activity in TC The authors declare no conflict of interest. axon branching remains unclear. *This Direct Submission article had a prearranged editor. To address this issue, we investigated TC axon branching in 1A.Y. and N.U. contributed equally to this work. fi cocultures of the thalamus and cortex by manipulating the ring 2Present address: Pharmacology Research Laboratories, Pharmaceutical Research Division, activity of thalamic (presynaptic) and cortical (postsynaptic) cells. Takeda Pharmaceutical Co. Ltd., Osaka 532-8686, Japan. To do this, neural activity of either thalamic or cortical cells was 3To whom correspondence should be addressed. E-mail: [email protected]. selectively silenced by means of overexpression of Kir2.1, an inward This article contains supporting information online at www.pnas.org/cgi/content/full/ rectifying potassium channel (16, 17). Our findings suggest that TC 0900613107/DCSupplemental.

7562–7567 | PNAS | April 20, 2010 | vol. 107 | no. 16 www.pnas.org/cgi/doi/10.1073/pnas.0900613107 Downloaded by guest on September 28, 2021 Fig. 1. Suppression of spontaneous firing in Kir2.1-overexpressing thalamic Fig. 2. Control and Kir2.1-overexpressing thalamic axons in coculture prep- cells in vitro. (A) Cocultures of thalamic and cortical slices on the MED after 2 arations. Thalamic cells were electroporated with eyfp (A) or with eyfp + kir2.1 weeks in culture. (B) Kir2.1 was expressed together with EYFP in thalamic (B) after 1-2 days in culture. Labeled TC axons were observed in the cocultured neurons in a coculture. (C) The upper and lower traces show extracellular cortical explants after 14 days in culture. (C) TC axons transfected with eyfp recording from cortical and thalamic explants, respectively, in a normal and kir2.1were observed after 9 days in culture. Contrast is reversed. (Scale bar: coculture preparation. In C, 85% of cortical cell firing is synchronized with A, 100 μm.) (A–C) Interrupted lines indicate the pial surface of cortical explant. thalamic cell firing with a delay of ≈4 ms (estimated by cross-correlation Arrows indicate the presumed layer 4 boundaries. analysis; see Materials and Methods). (D–F) Immunohistochemistry of Kir2.1- overexpressing thalamic cells. EYFP-labeled cells (D) were immunostained with anti-FLAG antibody (E). (F) A merged image. (G) Perforated-patch recording from intact (Upper) and transfected (Lower) thalamic cells. (H) Conversely, branch tip length was significantly increased by Kir2.1 Quantitative analysis of spontaneous firing activities of cortical, thalamic, overexpression (Table 1). In contrast, these parameters including and Kir2.1-overexpressing thalamic cells. The numbers in parentheses rep- branch points (12.2 ± 3.4, n = 9) were unchanged in cells that resent the number of cultures tested. CTX, cortex; Th, thalamus; KirTh, μ – μ expressed a mutant Kir2.1 lacking rectification (Table 1 and Fig. Kir2.1-overexpressing thalamus. (Scale bar: A and B, 500 m; D F,20 m.) NEUROSCIENCE S5). Thus, TC axon branching was markedly reduced by silencing thalamic cell activity. were stimulated electrically (Fig. S3B). The morphological features We also observed Kir2.1-expressing TC axons at earlier devel- of these transfected cells were also similar to those of intact thalamic opmental stages. They grew extensively in the cortical explant, and cells (Fig. S4 A, B, E,andF). Thus, it is unlikely that the basic aspects some of them traveled with growth cones (Fig. 2C). This aspect was of cellular morphology and the appearance of corticothalamic similar to that for normal TC axons in vitro (18, 19), which indicates synaptic transmission are altered by Kir2.1 overexpression. Fur- that axonal growth is not influenced by Kir2.1 overexpression. It is thermore, overall firing activity in cortical and thalamic explants was kir2.1 also unlikely that overproduced branches are later eliminated. not affected by introducing into a small number of thalamic Branch formation was further studied in cocultures where cells (Fig. S3C and Table 1). cortical cell activity was suppressed. As TC axons form con- Under these conditions, TC axon branching was investigated after 2 weeks in culture. Kir2.1-overexpressing TC axons extended nections with multiple target cells in vitro as well as in vivo, we into the cortical explant with little branching (Fig. 2 A and B). The attempted to overexpress Kir2.1 in numerous cortical cells. To kir2.1 dsred number of branch points was roughly one-fourth in Kir2.1-over- achieve this, and were cotransfected into the cells that expressing cells (2.9 ± 0.7, n = 14), compared with control (Fig. 3 A are destined to become layer 4, using in utero electroporation. and B and Table 1). The branch density (the number of branch tips After birth, cortical slices containing DsRed-positive cells were per the total branch length) was also much decreased (Table 1). dissected and cocultured with a thalamic slice (Fig. 4A). To label

Table 1. Branching aspects of TC axons and ﬁring activity of thalamic and cortical cells Th→CTX KirTh→CTX mKirTh→CTX Th→KirCTX Th→KirCTX (tri) Th→CTX (tri) Th→CTX (TTX)

(na = 17, nc =7) (na =14,nc =7) (na =9,nc =6) (na =12,nc =5) (na =12,nc =4) (na =7,nc =3) (na =17,nc =9)

† † Branch points 13.2 ± 2.5 2.9 ± 0.7* 12.2 ± 3.4 5.3 ± 1.4 5.8 ± 1.6 11.9 ± 3.9 2.2 ± 0.6* Total length, mm 2.14 ± 0.22 1.59 ± 0.21 2.84 ± 0.57 1.77 ± 0.24 1.55 ± 0.16 2.48 ± 0.65 1.22 ± 0.11† Density 5.4 ± 0.7 2.1 ± 0.3* 4.0 ± 0.5 3.1 ± 0.4† 3.7 ± 0.7 4.7 ± 0.8 2.4 ± 0.3* Tip length, μm 61.8 ± 7.0 268 ± 54* 121 ± 19 141 ± 39 109 ± 27 119 ± 21 244 ± 52*

(nc =4) (nc =2) (nc =4) (nc =5) (nc =5) (nc =2)

Th activity, Hz 0.96 ± 0.32 1.03 ± 0.20 N.D. 0.27 ± 0.09 0.81 ± 0.11 0.81 ± 0.11 0.0 ± 0.0 ‡ 0.017 ± 0.01 CTX activity, Hz 1.34 ± 0.45 1.21 ± 0.10 N.D. 0.17 ± 0.04 0.61 ± 0.14 1.40 ± 0.30 0.0 ± 0.0

Each value represents the mean and SEM. For branching, cross-comparison analysis shows no signiﬁcant difference between KirTh→CTX, Th→KirCTX, Th→KirCTX (tri), and Th→CTX (TTX). Likewise, there is no signiﬁcant difference between Th→CTX, mKirTh→CTX and Th→CTX (tri). CTX, cortex; Th, thalamus; KirTh, Kir2.1-overexpressing thalamus; mKirTh, mutant Kir2.1-overexpressing thalamus; KirCTX, Kir2.1-overexpressing cortex; (tri), the triple culture; (TTX),

TTX treatment; na, the number of axons sampled; nc, the number of cultures. *P < 0.01. † P < 0.05 (Dunnett’s test). ‡ This value was obtained from Kir2.1-expressing thalamic cells (n = 3) by whole-cell patch recording.

Yamada et al. PNAS | April 20, 2010 | vol. 107 | no. 16 | 7563 Downloaded by guest on September 28, 2021 Fig. 5. Postsynaptic responses in kir2.1-transfected cortical explants. Post- synaptic currents were recorded in control (A) and Kir2.1-expressing (B) cells Fig. 3. Decreased axonal branching in Kir2.1-overexpressing thalamic neu- in the same kir2.1-transfected cortical explants by stimulating the thalamic rons. (A) TC axon branching in normal cocultures of the thalamus and cortex explant. The current was also recorded in cortical cells (naïve control) in after 2 weeks in culture. (B) Axon branching of Kir2.1-overexpressing tha- normal cocultures (C). The peak amplitudes of EPSCs were plotted against lamic neurons in the cortex. Arrows indicate the presumed layer 4 boun- stimulus intensity (D). All electrophysiological experiments were performed daries. CTX, cortex; Th, thalamus; KirTh, Kir2.1-overexpressing thalamus. after 13–15 days in culture. *P < 0.05 (Dunnett’s test).

TC axons, the eyfp vector was electroporated to thalamic neurons points, 5.3 ± 1.4, n = 12; Table 1). The frequency of spontaneous at 1 day in vitro, as described. activity was decreased considerably in kir2.1-transfected cortical After 2 weeks in vitro, more than a thousand DsRed-labeled explants (0.17 ± 0.04 Hz, n = 4; Fig. 4 C and D and Table 1). The cells (1,000–2,000 cells, roughly 4–8% of the total number of silenced cells, which were densely distributed in the upper layers cortical cells) were primarily found in the upper cortical layers (10–20% of upper-layer cells; Fig. S4 H and I), are thought to (Fig. 4A and Fig. S4H). Kir2.1-overexpressing cortical cells had impair the emergence or spreading of field potential-like firing slightly hyperpolarized membrane potentials and smaller mem- activity by suppressing synchronous firing, which may be generated brane resistances, as was the case for thalamic cells (Fig. S3A; see by local circuits consisting of subsets of cortical neurons (21–23). also Fig. S5). Postsynaptic responses evoked by thalamic stim- As a consequence, firing activity was also suppressed in thalamic ulation were also examined in Kir2.1-expressing cortical cells. The explants (0.27 ± 0.09 Hz, n = 4; Fig. 4 C and D and Table 1). This amplitude of excitatory postsynaptic currents was 3× smaller than may be attributable to a significant reduction in excitatory trans- in naïve cortical explants (Fig. 5; see also Fig. S3D), but was not mission from the cortex to the thalamus, as spontaneous firing different from that in nonexpressing cells in kir2.1-transfected originated from the cortical explant more frequently (Fig. S1) (24). cortical explants (Fig. 5). This suggests that basic synaptic prop- To make thalamic cells active, another cortical slice was added to erties are affected by the silencing effect of kir2.1 rather than by the coculture (Fig. 4B). Consequently, firing activity in the thalamus transfection itself. The cellular morphology was not obviously was restored substantially in the triple cultures (0.81 ± 0.11 Hz in the different from that of normal cortical cells (Fig. S4 C, D, E, and G), thalamus of the triple culture, n = 5; 0.96 ± 0.32 Hz in the thalamus although the possibility cannot be ruled out that Kir2.1 expression of the normal coculture, n =4;Fig.4C and D and Table 1). In affects fine morphology (20). contrast, spontaneous firing remained low in Kir2.1-transfected We found that TC axon branching was markedly decreased in cortical explants (0.61 ± 0.14 Hz, n =5),comparedwithfiring cocultures of kir2.1-transfected cortex with thalamus (branch activity in normal cocultured cortex (1.34 ± 0.45 Hz, n = 4) and the

Fig. 4. Spontaneous activity of thalamic and cortical cells in cocultures or triple cultures containing Kir2.1-overexpressing cortex. (A) Coculture of the thalamus with the cortical slice where Kir2.1 plus DsRed were overexpressed before culturing by in utero electroporation. Thalamic and cortical explants were placed on an MED (Top). The same cultures were observed by fluorescence microscopy (Bottom) after 2 weeks in culture. (B) Similar to A, with the exception that a thalamic explant was sandwiched with between a Kir2.1-overexpressing and an untransfected cortical explant on the MED. (C) The upper two traces show the extracellular recording from cocultures of the thalamic explant and Kir2.1-overexpressing cortical explant. The lower three traces correspond to the extracellular potentials from the triple culture. (D) Relative firing activity (normalized by the average firing frequency of normal thalamic or cortical activity) in thalamic and Kir2.1- overexpressing explants in the cocultures (Left). Similarly, the relative activities of thalamic explants, Kir2.1-overexpressing cortical explants and additional cortical explants are shown for the triple culture (Right). The numbers in brackets represent the numbers of samples tested. (Scale bar: A and B, 1 mm.)

7564 | www.pnas.org/cgi/doi/10.1073/pnas.0900613107 Yamada et al. Downloaded by guest on September 28, 2021 (Fig. S2) (18), it is likely that both pre- and postsynaptic activity is required for normal TC axon branching (Fig. 7; see also Table 1). The finding that branch elaboration is promoted only by coactivation of pre- and postsynaptic cells may not fit the case of axon arbor remodeling in eye-specific projections of the visual cortex. Indeed, elaborate axon arbors are observed in the cat visual cortex when both pre- and postsynaptic cells are silent (12, 25). This is also true in the remodeling of axon arbors in the retinotectal projection (26, 27). In contrast, TC axons did not form elaborate branches in TTX-treated cocultures and in cocultures of the thalamus and Kir2.1-expressing cortex, in which both pre- and postsynaptic activity was suppressed. This is consistent with the features of geniculocortical axons observed at the early developmental stages in TTX- infused visual cortex (28). Therefore, the mechanism that underlies initial arbor formation seems to be different from those regulating plastic changes or maintenance in later developmental stages. Previous studies using neuronal silencing with Kir2.1 have suggested that presynaptic activity plays a dominant role in axon branching (27, 29, 30). In particular, competitive interactions between a retinal ganglion cell and adjacent cells have been sug- Fig. 6. TC axon branching in Kir2.1-overexpressing and intact cortical gested to be crucial in retinotectal axon branching (27). However, explants in the triple cultures. TC axon branching was reduced in Kir2.1- it is unlikely that the same mechanism acts on branch formation of overexpressing cortical explants (A) but not in intact cortical explants (B). Th, TC axons. When kir2.1 was introduced into a larger number of thalamus; CTX (tri), intact cortex in the triple culture; KirCTX (tri), Kir2.1- thalamic cells, the labeled axons still formed few branches in the overexpressing cortex in the triple culture. All TC axons were observed after 2 weeks in culture. Arrows indicate the presumed layer 4 boundaries. cortical explant (Fig. S6). In accordance with this, recent evidence suggests that competition among presynaptic axons may be less important than homosynaptic mechanisms for the elaboration of additional cortex in the triple cultures (1.40 ± 0.30 Hz, n =5;Fig.4C eye-specific geniculocortical axon arbors (12). NEUROSCIENCE and D and Table 1). In the triple cultures, TC axons did not form A characteristic aspect of our cultures was that thalamic and fi C extensive branches in the Kir2.1-transfected cortex (Fig. 6A). cortical cell ring was highly synchronized (Fig. 1 and Fig. S1). Quantitative analysis confirmed that the number of branch points Our previous study using the same coculture preparations showed that synchronized firing activity between thalamic and cortical cells was significantly smaller (5.8 ± 1.6, n = 12) for TC axons invading develops during the second week in vitro, when axonal branches the Kir2.1-trasnsfected cortex than control, although the total are added in great numbers (18). Thus, precise synchrony rather branch length and the density were not statistically different (Table than the overall amounts of pre- and postsynaptic activity for a 1). In contrast, TC axons invading the intact cortical explant in the given period may play an important role in the promotion of axon triple culture formed branches in a fashion (11.9 ± 3.9, n =7)similar branching (18, 31–34), although our investigations did not dis- to that in control (Fig. 6B and Table 1). Thus, axonal branching was tinguish between these two possibilities. The NMDA receptor, a decreased in the more silent cortex, which indicates that cortical cell coincidence detector of pre- and postsynaptic activity, is likely to activity is also required for TC axon branching. be involved in this process (13, 35, 36). Indeed, a large NMDA receptor-mediated component was generated in cortical cells by Discussion thalamic stimulation (Fig. S7), as found in the developing sensory The present findings show that TC axon branching is impaired in cortices (37–39). In addition, the fact that the AMPA/NMDA ratio coculture preparations, where either thalamic or cortical cell was lower in the silenced cortical explants than in naïve ones activity is silenced by Kir2.1 overexpression. Together with the implies that the normal development of the AMPA component finding that axon branching is reduced in the presence of TTX might also be related to branch formation (Fig. S7). There are several possible mechanisms to account for how coactivation of pre- and postsynaptic elements might underlie the axonal morphological changes. A plausible mechanism is that a retrograde signal may induce branches in growing axons. The retrograde substance could be released from postsynaptic cells when both pre- and postsynaptic elements are coactivated. Neurotrophins are candidate molecules for this process (40), as they promote thalamic axon growth in vitro (41). Similarly, the expression of branch-promoting and inhibiting molecules (2, 42) synthesized by postsynaptic cortical cells may be regulated by electrical activity. Activity-dependent expression of receptor molecules for these ligands may also be crucial. For example, Eph expression in motor axons is altered by their firing activity (43). It has also been shown that activity blockade inhibits axon responsiveness to ephrin-As and leads to the disruption of topographic mapping in the retinotectal projection (44). Finally, regulation of cytoplasmic signaling such as that driven by NMDA receptor activation may also be specifically Fig. 7. Relation between the number of branch points and firing activity. Firing activities of thalamic and cortical explants are shown in X and Y axes, involved in activity-dependent processes (45, 46). respectively. The number of branch points in each culture type is shown on the In summary, TC axon branching was promoted only when both z axis. CTX, cortex; Th, thalamus; KirTh, Kir2.1-overexpressing thalamus; thalamic and cortical cells were coactivated in vitro. This sug- KirCTX, Kir2.1-overexpressing cortex; (tri), triple culture; (TTX), TTX treatment. gests that coactivation of pre- and postsynaptic elements not only

Yamada et al. PNAS | April 20, 2010 | vol. 107 | no. 16 | 7565 Downloaded by guest on September 28, 2021 strengthens synaptic plasticity but also induces structural changes total branch length), were performed. Statistical evaluation was performed by in cortical circuits during development. means of one-way ANOVA followed by Dunnett’s post hoc test.

Materials and Methods Analysis of Cellular Morphology and Cell Counts. To study cellular morphology × × A schematic timeline of the present experiments, including culture prepara- and the number of labeled cells, images were taken with a 10 or 20 objective tions, gene transfer, electrophysiological recording, and anatomical obser- lens. To determine the total number of cells in thalamic and cortical slices, μ vation, is provided in Fig. S8. cultures were cut into 20- m sections, followed by Nissl staining. Then, the ratio of the number of labeled cells to the total number of cells was estimated Organotypic Slice Culture. All experiments were performed according to the for each cortical or thalamic explant. Indeed, the percentage was estimated to – guidelines established by the animal welfare committees of Osaka University. be 3 6%. The ratio of Kir2.1-expressing cortical cells was also evaluated by Cocultures of the cortex with the thalamus were prepared as described pre- dissociating cortical slices, which had been transfected with dsred + kir2.1 by in viously (47). In brief, the dorsal thalamic region was dissected from E15 rat utero electroporation. The cortical slices were dissected at P1, and were 2+ 2+ ’ embryos (Sprague-Dawley), and cortical slices were dissected from sensory incubated in Ca Mg -free Hanks solution containing trypsin (0.25%). After cortices of postnatal day (P)1 or P2 rats. The thalamic and cortical slices were dissociation, the number of labeled cells and the total number of dissociated – plated on a membrane ﬁlter (Millicell-CM PICMORG50; Millipore), which was cells were counted. The percentage was estimated to be 5 10% by this coated with rat tail collagen. To monitor neural activity, cortical and thalamic method. Finally, the ratio of Kir2.1-expressing cells was evaluated as the explants were plated on MEDs (Alpha MED Sciences) (18). The culture medium average of the values obtained by the both methods. consisted of a 1:1 mixture of DMEM and Ham’s F-12 (Invitrogen) with several supplements (47). These cultures were maintained at 37 °C in an environment Immunohistochemistry. Cultured slices were ﬁxed with 4% paraformaldehyde

of humidified 95% air and 5% CO2. followed by extensive washes with PBS. The slices were incubated overnight at 4 °C with anti-FLAG mAb (1:1,000, ANTI-FLAG M2; Sigma). The fluo- Preparation of Plasmids and Gene Delivery into Cultured Cells. The coding rescence signal was visualized by donkey anti-mouse IgG-Cy3 (1:200; Sigma). region of eyfp was cloned into a pCAGGS vector (48). Rat kir2.1 cDNA (GenBank accession no. NM017296, Ensemble ENSRNOG00000004720) was obtained by Recording of Spontaneous Activity. To examine spontaneous activities of RT-PCR with primers (forward; 5′-TTCTAAAGCAGAAACACTGG-3′,reverse;5′- cortical and thalamic cells, extracellular recording was performed on MED CATCAGACTGTGTAGCGA -3′) and P2 rat brain cDNA. The product was further (interpolardistance,0.3mm;inputimpedancewiththepreamplifier>1MΩ), as processed by PCR (forward primer; 5′-GCTCGAGGAAGCATGGGCAGTGTGCGT-3′, described previously (18). In brief, extracellular voltages were recorded at reverse primer; 5′-GAATTCCTACTTGTCGTCATCGTCTTTGTAGTCTACTATCTCC- several locations for more than 5 min every day. Negative potentials with GATTCTCGCCT-3′) to add FLAG tag (underlined) to the carboxyl terminus. The amplitudes above a set threshold (1.5× the maximal amplitude of the baseline PCR product was inserted into the pCAGGS vector. To generate nonconducting noise) were counted as spikes with Axograph software (Axon Instruments). kir2.1 mutant, the pore-region amino acid motif GYG was mutated to AAA (16). Firing activity of each explant was represented as an average of the firing rates Cotransfection of pCAGGS-eyfp and pCAGGS-kir2.1 (or pCAGGS-kir2.1 mutant) during 11–14 days in culture. The synchronism between spontaneous thalamic was performed for organotypic coculture preparations to examine the detail of and cortical activities was defined as the percentage of thalamic firing that axon morphology. time-overlapped with cortical firing during the observation time (5 min). Electroporation with glass microelectrodes was performed after 1–2 days Cross-correlation analysis was performed to examine the time difference in culture to introduce these vectors into thalamic cells (18, 49). In brief, the between thalamic and cortical potentials, using Axograph software. – μ μ plasmid solution of pCAGGS-eyfp (1 2 g/ L) or a mixture of pCAGGS-eyfp Tomeasure spontaneous activity from individual Kir2.1-overexpressing cells – μ μ – μ μ (1 2 g/ L) + pCAGGS-kir2.1 (4 5 g/ L) was pressure ejected onto the sur- in the cocultures, whole-cell perforated-patch recordings were performed face of the explants with a glass micropipette, and electrical pulses (10 trains under current-clamp mode using an upright microscope (BX51WI; Olympus). μ of 200 square pulses of 1 ms duration, 200 Hz, 500 A) were applied with The pipette solution consisted of (in mM) 140 D-gluconate potassium salt, 10 another glass microelectrode. NaOH, 10 Hepes, 8 MgCl2; pH was adjusted to 7.35 with HCl; 0.005 volume dimethyl sulfoxide solution of amphotericin B (0.2 mg/mL) was added before In Utero Electroporation. In utero electroporation was performed to over- recordings. The recording chamber (culture dish) was perfused at a rate of 1–2 express Kir2.1 in the cortical cells that are destined to become layer 4 cells (50, mL/min with a saline whose composition was (mM) 123 NaCl, 4.2 KCl, 10 Hepes, 51, 52). Pregnant rats at E16 were deeply anesthetized with Nembutal. The 10 D-glucose, 2 CaCl2, and 1 MgCl2 (pH 7.35). abdomen was surgically opened without opening the uterus itself. A mixture of pCAGGS-dsred (2 μg/μL) and pCAGGS-kir2.1 (2 μg/μL) was injected Recording of Synaptic Currents. To obtain synaptic currents from thalamic and into one cerebral vesicle. Platinum electrodes were positioned beside the cortical cells, whole-cell recordings were performed. All experiments were uterus, and square pulses (30 V; 50 ms) were delivered five times with carried out at 32 °C. Resistances of patch pipettes were 2–3MΩ when filled electroporator (CUY20; NepaGene). After electroporation, embryos were with an intracellular solution composed of (in mM) 60 CsCl, 10 Cs D-gluconate, allowed to develop until birth. 20 TEA-Cl, 20 BAPTA, 4 MgCl2, 4 ATP, and 30 Hepes (pH 7.3, adjusted with CsOH). The pipette access resistance was compensated by 70%. The compo- Confocal Imaging and Quantitative Analysis. EYFP-labeled TC axons were sition of the standard bathing solution was (in mM) 125 NaCl, 2.5 KCl, 2 CaCl ,1 observed by confocal microscopy (MRC-600; Bio-Rad) (18), with an argon laser 2 MgSO , 1.25 NaH PO , 26 NaHCO , and 20 glucose, bubbled with 95% O and and a filter set (excitation, 488 nm; emission long-pass filter, 515 nm). DsRed- 4 2 4 3 2 5% CO . Picrotoxin (100 μM) was always added to block inhibitory synaptic positive cells were observed with another filter set (excitation, 514 nm; emis- 2 transmission. Electrical stimulation (duration, 0.1 ms; amplitude, 0–50 μA) to sion long-pass filter, 565 nm). EYFP-labeled axons were easily distinguished the cortical and thalamic explants was applied with bipolar tungsten elec- with this filter set even when DsRed-labeled cells were present in cortical trodes (interpolar distance, 0.5 mm). explants. Images were collected with 10× and 40× objective lenses (768 × 512 pixels for 1,097 × 731 μm or 274 × 183 μm) at 1- to 5-μm steps to obtain the entire axon arbors (2–20 optical sections). Individually distinguishable axons ACKNOWLEDGMENTS. We thank Drs. Edward S. Ruthazer, Björn Granseth, Noriyuki Sugo, and Ryuichi Shirasaki for critical reading of this manuscript. We were drawn using National Institutes of Health image custom-made macros (a also thank Drs. Yukio Komatsu and Yoshio Hata for helpful suggestions. This generous gift from Edward Ruthazer, Montreal Neurological Institute, McGill work was supported by Grants-in-Aid for Scientific Research Projects 15300107, < μ University, Montréal, Canada). Small processes ( 5 m) were excluded from 18021021, and 18300105 (to N.Y.). 17023001 and 21220006 (to M.K.) from the analysis. Quantitative analysis, including the number of branch points, axonal Japanese Ministry of Education, Culture, Sports, Science and Technology, and tip length, total branch length, and branch density (the number of branches/ the Novartis Foundation (Japan) for the Promotion of Science.

1. Tessier-Lavigne M, Goodman CS (1996) The molecular biology of axon guidance. 4. Cline H (2003) Sperry and Hebb: Oil and vinegar? Trends Neurosci 26:655–661. Science 274:1123–1133. 5. Dantzker JL, Callaway EM (1998) The development of local, layer-speciﬁc visual cortical 2. Dent EW, Barnes AM, Tang F, Kalil K (2004) Netrin-1 and semaphorin 3A promote or axons in the absence of extrinsic inﬂuences and intrinsic activity. J Neurosci 18:4145–4154. inhibit cortical axon branching, respectively, by reorganization of the cytoskeleton. J 6. Hubel DH, Wiesel TN, LeVay S (1977) Plasticity of ocular dominance columns in Neurosci 24:3002–3012. monkey striate cortex. Philos Trans R Soc Lond B Biol Sci 278:377–409. 3. Katz LC, Shatz CJ (1996) Synaptic activity and the construction of cortical circuits. 7. LeVay S, Stryker MP, Shatz CJ (1978) Ocular dominance columns and their develop- Science 274:1133–1138. mentinlayerIVofthecat’s visual cortex: A quantitative study. JCompNeurol179:223–244.

7566 | www.pnas.org/cgi/doi/10.1073/pnas.0900613107 Yamada et al. Downloaded by guest on September 28, 2021 8. Huberman AD, Speer CM, Chapman B (2006) Spontaneous retinal activity mediates 31. Bi G, Poo M (2001) Synaptic modification by correlated activity: Hebb’s postulate development of ocular dominance columns and binocular receptive fields in v1. revisited. Annu Rev Neurosci 24:139–166. Neuron 52:247–254. 32. Chandrasekaran AR, Plas DT, Gonzalez E, Crair MC (2005) Evidence for an instructive 9. Stryker MP, Harris WA (1986) Binocular impulse blockade prevents the formation of role of retinal activity in retinotopic map refinement in the superior colliculus of the ocular dominance columns in cat visual cortex. J Neurosci 6:2117–2133. mouse. J Neurosci 25:6929–6938. 10. Antonini A, Stryker MP (1996) Plasticity of geniculocortical afferents following brief 33. Butts DA, Kanold PO, Shatz CJ (2007) A burst-based “Hebbian” learning rule at or prolonged monocular occlusion in the cat. J Comp Neurol 369:64–82. retinogeniculate synapses links retinal waves to activity-dependent refinement. PLoS 11. Antonini A, Stryker MP (1993) Rapid remodeling of axonal arbors in the visual cortex. Biol 5:e61. Science 260:1819–1821. 34. Torborg CL, Hansen KA, Feller MB (2005) High frequency, synchronized bursting 12. Haruta M, Hata Y (2007) Experience-driven axon retraction without binocular drives eye-specific segregation of retinogeniculate projections. Nat Neurosci 8:72–78. imbalance in developing visual cortex. Curr Biol 17:37–42. 35. Iwasato T, et al. (2000) Cortex-restricted disruption of NMDAR1 impairs neuronal 13. Lee LJ, Iwasato T, Itohara S, Erzurumlu RS (2005) Exuberant thalamocortical axon patterns in the barrel cortex. Nature 406:726–731. fi – arborization in cortex-speci c NMDAR1 knockout mice. J Comp Neurol 485:280 292. 36. Schlaggar BL, Fox K, O’Leary DD (1993) Postsynaptic control of plasticity in developing fi 14. Rebsam A, Seif I, Gaspar P (2002) Re nement of thalamocortical arbors and somatosensory cortex. Nature 364:623–626. emergence of barrel domains in the primary somatosensory cortex: A study of normal 37. Iwakiri M, Komatsu Y (1993) Postnatal development of NMDA receptor-mediated – and monoamine oxidase a knock-out mice. J Neurosci 22:8541 8552. synaptic transmission in cat visual cortex. Brain Res Dev Brain Res 74:89–97. fi 15. Lu HC, et al. (2006) Role of ef cient neurotransmitter release in barrel map 38. Yamada K, Yamamoto N, Toyama K (2000) Development of NMDA and non-NMDA – development. J Neurosci 26:2692 2703. receptor-mediated excitatory synaptic transmission in geniculocortical and cortico- ’ 16. Burrone J, O Byrne M, Murthy VN (2002) Multiple forms of synaptic plasticity cortical connections in organotypic coculture preparations. Eur J Neurosci 12: triggered by selective suppression of activity in individual neurons. Nature 420: 3854–3862. 414–418. 39. Crair MC, Malenka RC (1995) A critical period for long-term potentiation at 17. Johns DC, Marx R, Mains RE, O’Rourke B, Marbán E (1999) Inducible genetic thalamocortical synapses. Nature 375:325–328. suppression of neuronal excitability. J Neurosci 19:1691–1697. 40. Cabelli RJ, Hohn A, Shatz CJ (1995) Inhibition of ocular dominance column formation 18. Uesaka N, Hayano Y, Yamada A, Yamamoto N (2007) Interplay between laminar by infusion of NT-4/5 or BDNF. Science 267:1662–1666. specificity and activity-dependent mechanisms of thalamocortical axon branching. J 41. Hanamura K, Harada A, Katoh-Semba R, Murakami F, Yamamoto N (2004) BDNF and Neurosci 27:5215–5223. NT-3 promote thalamocortical axon growth with distinct substrate and temporal 19. Yamamoto N, Higashi S, Toyama K (1997) Stop and branch behaviors of dependency. Eur J Neurosci 19:1485–1493. geniculocortical axons: A time-lapse study in organotypic cocultures. J Neurosci 17: 42. Yamamoto N, et al. (2000) Inhibitory mechanism by polysialic acid for lamina-specific 3653–3663. branch formation of thalamocortical axons. J Neurosci 20:9145–9151. 20. Cancedda L, Fiumelli H, Chen K, Poo MM (2007) Excitatory GABA action is essential for 43. Hanson MG, Landmesser LT (2004) Normal patterns of spontaneous activity are morphological maturation of cortical neurons in vivo. J Neurosci 27:5224–5235. required for correct motor axon guidance and the expression of specific guidance 21. Mao BQ, Hamzei-Sichani F, Aronov D, Froemke RC, Yuste R (2001) Dynamics of – spontaneous activity in neocortical slices. Neuron 32:883–898. molecules. Neuron 43:687 701. 22. Yoshimura Y, Dantzker JL, Callaway EM (2005) Excitatory cortical neurons form fine- 44. Nicol X, et al. (2007) cAMP oscillations and retinal activity are permissive for ephrin signaling during the establishment of the retinotopic map. Nat Neurosci 10:340–347. scale functional networks. Nature 433:868–873. NEUROSCIENCE 23. Rumpel S, LeDoux J, Zador A, Malinow R (2005) Postsynaptic receptor trafficking 45. Ohnami S, et al. (2008) Role of RhoA in activity-dependent cortical axon branching. J – underlying a form of associative learning. Science 308:83–88. Neurosci 28:9117 9121. 24. Yamamoto N, Kurotani T, Toyama K (1989) Neural connections between the lateral 46. Nakazawa T, et al. (2008) Regulation of dendritic spine morphology by an NMDA geniculate nucleus and visual cortex in vitro. Science 245:192–194. receptor-associated Rho GTPase-activating protein, p250GAP. J Neurochem 105: – 25. Hata Y, Tsumoto T, Stryker MP (1999) Selective pruning of more active afferents when 1384 1393. fi cat visual cortex is pharmacologically inhibited. Neuron 22:375–381. 47. Yamamoto N, Yamada K, Kurotani T, Toyama K (1992) Laminar speci city of extrinsic 26. Ruthazer ES, Akerman CJ, Cline HT (2003) Control of axon branch dynamics by cortical connections studied in coculture preparations. Neuron 9:217–228. correlated activity in vivo. Science 301:66–70. 48. Niwa H, Yamamura K, Miyazaki J (1991) Efficient selection for high-expression 27. Hua JY, Smear MC, Baier H, Smith SJ (2005) Regulation of axon growth in vivo by transfectants with a novel eukaryotic vector. Gene 108:193–199. activity-based competition. Nature 434:1022–1026. 49. Uesaka N, Nishiwaki M, Yamamoto N (2008) Single cell electroporation method for 28. Herrmann K, Shatz CJ (1995) Blockade of action potential activity alters initial axon tracing in cultured slices. Dev Growth Differ 50:475–477. arborization of thalamic axons within cortical layer 4. Proc Natl Acad Sci USA 92: 50. Saito T, Nakatsuji N (2001) Efficient gene transfer into the embryonic mouse brain 11244–11248. using in vivo electroporation. Dev Biol 240:237–246. 29. Mizuno H, Hirano T, Tagawa Y (2007) Evidence for activity-dependent cortical wiring: 51. Tabata H, Nakajima K (2001) Efficient in utero gene transfer system to the developing Formation of interhemispheric connections in neonatal mouse visual cortex requires mouse brain using electroporation: Visualization of neuronal migration in the projection neuron activity. J Neurosci 27:6760–6770. developing cortex. Neuroscience 103:865–872. 30. Wang CL, et al. (2007) Activity-dependent development of callosal projections in the 52. Fukuchi-Shimogori T, Grove EA (2001) Neocortex patterning by the secreted signaling somatosensory cortex. J Neurosci 27:11334–11342. molecule FGF8. Science 294:1071–1074.

Yamada et al. PNAS | April 20, 2010 | vol. 107 | no. 16 | 7567 Downloaded by guest on September 28, 2021