- Email: [email protected]
In vitro reconstruction and functional development of the superior colliculus in the retinotectal pathway
Neuroscience Letters 545 (2013) 96–101
Contents lists available at SciVerse ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
In vitro reconstruction and functional development of the superior colliculus in the retinotectal pathway Shinya Hirota ∗ , Hiroyuki Moriguchi 1 , Atsushi Saito, Kousuke Inoue, Akitoshi Murakami, Kiyoshi Kotani, Yasuhiko Jimbo Department of Human and Engineered Environmental Studies, Graduate School of Frontier Sciences, University of Tokyo, 5-1-5, Kashiwanoha, Kashiwa, Chiba 277-8563, Japan
h i g h l i g h t s
g r a p h i c a l
a b s t r a c t
• Retinotectal pathways were reconstructed on MEA substrates by cocultures. • Spontaneous activities of co-cultured retinas and SCs were recorded for one month. • Functional connections between co-cultured retinas and SCs were formed. • The signal input from retina affects the functional development of SC.
a r t i c l e
i n f o
Article history: Received 6 February 2013 Received in revised form 12 March 2013 Accepted 6 April 2013 Keywords: Superior colliculus (SC) Microelectrode array (MEA) substrates Co-culture Spontaneous activity Development
a b s t r a c t In order to examine the formation of a neural network and the functional development of a visual pathway, we performed in vitro reconstruction of the retinotectal pathway using organotypic explants and co-culture methods. Retinas and superior colliculus (SC) slices obtained from embryonic rats were cocultured on microelectrode array (MEA) substrates for four weeks. We observed retinal ganglion cell neurites innervating SC slices that evoked responses in retinas or SC slices after applying electrical stimulation. Functional connections between retinas and SC slices were formed in the cultures. At the same time, spontaneous electrical activities were recorded from both the retinas and SC slices over the four weeks. In the co-cultured SC slices, sporadic firings were initially observed at 3–4 days in vitro (DIV), and thereafter the frequency of spontaneous firing increased and synchronized activities occurred after two weeks in vitro (WIV). In most of the single-cultured SC slices, however, only sporadic firings were observed over four weeks. In addition, the retinas and SC slices were co-cultured to enable the exchange of soluble factors with each other via culture medium but not via direct neural connections. The activity patterns resembled ones of single-cultured SC slices. These results suggest that signal inputs from retinas through direct neural connections affect the development of SCs in the retinotectal pathway. © 2013 Elsevier Ireland Ltd. All rights reserved.
1. Introduction
∗ Corresponding author at: Department of Neurosurgery, University of Texas MD Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030, USA. Tel.: +1 713 834 6240. E-mail addresses: [email protected], sh [email protected] (S. Hirota). 1 Present address: Quantitative Biology Center (QBiC), Integrated Biodevice Research Unit, RIKEN, 2-2-3, Minatojima-minamimachi, Chuo-ku, Kobe, Hyogo 6500047, Japan. 0304-3940/$ – see front matter © 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.neulet.2013.04.020
The retinotectal pathway, which is formed by the retinal projection to the superior colliculus (SC), plays an important role in the control of orienting eye and head movement in vertebrates. Precise topographic projection from the retina to the SC is established, as is the lateral geniculate nucleus (LGN) and the primary visual cortex [23]. The retinotectal pathway is useful to elucidate the formation and development of neural circuits in visual pathways using in vitro systems [24].
S. Hirota et al. / Neuroscience Letters 545 (2013) 96–101
The vertebrate retinotectal pathway uses a multistep developmental process [3]. In rats, retinal ganglion cell (RGC) axons initially enter the SC on embryonic day 16 (E16) and form synapses with SC neurons on E17 [16]. At the same time, spontaneous firings of RGCs occur on E16 [1] and are transmitted along optic axons on postnatal day 0 (P0) [4]. Previous studies have suggested that signal inputs from the retina regulate the refinement of connections between RGCs and SC neurons by cell death and axon retraction [2]. Electrical activity is also thought to be important after embryonic development for the refinement of connections in brain [22]. Little is known about spontaneous activity patterns in SC during development. The earliest spontaneous activity of SC was recorded at P5-6 [10]. The synaptic transmission between RGC and SC neuron was detectable on P1 using in vitro models [15] and P10 using in vivo models [18]. In general, spontaneous activity is triggered by external input [12], but the activity in the cortex occurs spontaneously in the absence of any signal input from other regions [21]. The former suggests that signal input from retinas evokes SC activity and therefore affects SC development, while the latter seems to indicate that SC activity occurs spontaneously or is affected by retina-derived soluble factors. In this study, we investigated which were more critical in the development of SC: signal input from retinas or retina-derived factors. By development, we mean the construction of functional circuits in SC using retinotectal pathways on MEA substrates. The MEA substrate is a culture dish embedded with multiple electrodes to enable non-invasive recording of electrical activity and stimulation with various spatial-temporal patterns [11]. Using organotypic culture on MEA substrates enables us to monitor spontaneous neural activity of both tissues while preserving organotypic organizations over several weeks. The co-culture model does not display an in vivo-like morphology or functions, but it does enable the investigation of the effect of signal input or soluble factors by separating from other regions [6]. In our previous study, spontaneous activity from co-cultured retinas and SCs were recorded up to 11 days in vitro (DIV) using P0-2 rats [9]. Considering that the major developmental process of retinotectal pathways occurs before the eye-opening around P14 [18], it is important to record spontaneous activities in SCs in retinotectal pathways for at least 2–3 weeks in vitro (WIV). To promote cell survival and neurite growth [7], we attempted co-cultures of retinas and SCs taken from younger rats, E17-18. As a result, retinotectal pathways, in which we recorded spontaneous activity for four weeks and signaling between retinas and SCs, were constructed on MEA substrates. Next, in order to investigate the effect of the retina on SC development, spontaneous activity patterns in co-cultured or single-cultured SCs were analyzed for four weeks. We also examined whether signal input from the retina or retina-derived soluble factors affected the development of SC by recording spontaneous activity in co-cultured SCs when plating the polydimethylsiloxane (PDMS, Dow Corning Toray) compartments between the retinas and SCs.
97
10 m × 10 m, and arranged with a 50 m distance. The former was used for the co-cultures of retinas and SC slices and the latter was used for the single-cultured SC slices. 2.2. Organotypic cultures All animal experiments were carried out in accordance with the guidelines for animal experiments at University of Tokyo. A total of seven pregnant Wistar rats at E17-18 (Charles River) were deeply anesthetized with diethyl ether and the embryos were surgically removed. The retinas and SC slices were taken from these embryos using ice-cold artificial cerebrospinal fluid (ACSF) saturated with 95% O2 and 5% CO2 . The ACSF contained 149 mM NaCl, 2.8 mM KCl, 2 mM CaCl2 , 1 mM MgCl2 , 10 mM HEPES, and 10 mM d-glucose, pH 7.4–7.5. Retinas were dissected free from the surrounding tissues of eyes, and were cut into 4–6 pieces. SCs were cut into 300 m-slices using a vibratome (Pro 7 Linear Slicer, Dosaka EM). Both tissues were placed on polyethyleneimine (PEI, Sigma) and laminin (Gibco) double-coated MEA substrates and separated at a distance of 500 m (Fig. 1). The culture medium was a 1:1 mixture of Dulbecco’s modified Eagle’s medium and Ham’s F-12 (Invitrogen) supplemented with 5% fetal bovine serum (FBS, Invitrogen), 1% Penicillin/Streptomycin (Invitrogen), 5 g/ml Insulin (Sigma), 20 nM Progesterone (Sigma), 20 nM Hydrocortisone (Sigma), 100 g/ml Transferrin (Sigma), 100 M Putrescine (Sigma), and 30 nM sodium selenite (Wako), pH 7.3. The co-cultures were maintained in a humidified 5% CO2 incubator at 37 ◦ C. The medium was exchanged every 2–3 days. 2.3. Live cell imaging with DiI The neuronal projections from retinas to SCs in vitro were confirmed. Retinas were incubated for 50 min at 37 ◦ C in the DiI solution consisting of a Vybrant DiI cell-labeling solution (Invitrogen) diluted 200 times with ACSF and then washed with ACSF three times. Retinas labeled with DiI were cultured with SC slices on culture dishes at a distance of more than 500 m. 2.4. Extracellular recordings and electrical stimulations Recordings and stimulations were carried out in a box with an atmosphere kept at 37 ◦ C and 95% air/5% CO2 to minimize damage to cells. Electrical signals were amplified with a 64 channel amplifier (NF Corp., Gain: ×10,000) and filtered with a band pass filter (100 Hz–5 kHz). The amplified signals were then sampled at intervals of 40 s using an AD converter (National Instrument) and recorded on the hard disk of a PC. Evoked responses in one tissue were observed after applying electrical stimulation to the other tissue to confirm the formation of functional connections between retina and SC. Stimulation waveforms consisted of 10 biphasic pulses (0.1 ms at +1 V followed by 0.1 ms at −1 V) at a frequency of 0.1 Hz.
2. Materials and methods
2.5. Fabrication of PDMS compartment
2.1. Fabrication of MEA substrates
We fabricated a PDMS compartment that enabled the exchange of soluble factors via culture medium but prevented the formation of neural connections. The compartment had a box shape of 200–300 m wide and 250 m high.
We fabricated MEA substrates with two types of electrode patterns (Fig. 1). The electrode circuit was made of indium tin oxide slightly coated on the glass substrate and covered with Si-based photoresist as the insulation layer. Recording and stimulation sites were coated with platinum black to reduce the interface impedance. The first type of MEA substrates was split into 4 × 8 electrodes separated by 500 m. Each electrode within a group was 30 m × 30 m, and was separated from their neighbors by 180 m. The other type was split into 8 × 8 electrodes, each
3. Results 3.1. Co-culture of retina and SC on MEA substrates Retinas and superior colliculus (SC) slices were co-cultured on MEA substrates for four weeks (n = 10, Fig. 2). Neurites extended
98
S. Hirota et al. / Neuroscience Letters 545 (2013) 96–101
Fig. 1. Retina-SC co-culture and SC culture on microelectrode array (MEA) substrates. Two different kinds of MEA substrates were used for the co-culture and the SC culture. The co-cultured retina and SC slice were arranged on each block of MEA substrates such that the tip of the retina was oriented toward the surface of the SC in view of the anatomical structure of the retinotectal pathway as oriented in vivo. In addition, retinas were mounted such that the RGCs generating action potentials contact with the electrodes on the MEA substrates.
from the retinas and SC slices until 24 h in vitro. The neurite outgrowth and cell migration from the retina extended toward the SC slice but did not expand to the opposite sides (Fig. 2A). SC axons grew radially. The time course of spontaneous activity in co-cultured retina and SC over four weeks is shown in Fig. 2B as raster plots. The thresholds of signal detection were set as the five times standard deviations of amplitudes in the noise region. In the retina, spontaneous activity was observed after 2–3 days in vitro (DIV). Sporadic firings generated from 1 to 3 WIV and continuous firings were recorded from several electrodes at 4 WIV. In the SC, spontaneous activity was recorded after 3 DIV and the shift of activity pattern was observed until 4 WIV. Sporadic firings occurred until 1 WIV. Thereafter, firing rates increased and
synchronized bursts generated locally at 3 WIV, and synchronized bursts around the SC slice generated at 4 WIV. 3.2. Functional connection formed between co-cultured retina and SC Morphological connections between retinas and SC slices were established after 1 WIV (Fig. 3A, left). To examine whether neuronal projections were formed anterogradely in the retina-SC co-cultures, we placed DiI crystals into retinas and cultured with SC slices on culture dishes at a distance of more than 500 m (Fig. 3A, right). Consequently, the retinal cells and their axons that were labeled with DiI could be identified. The connection and the upper regions of the co-cultured SC slices were robustly stained. This
Fig. 2. Spontaneous activity recorded from the co-cultured retina and SC slice over four weeks. (A) A phase-contrast image of a retina-SC co-culture on the MEA substrates at 4 weeks in vitro (WIV). (B) The spontaneous activity recorded from co-cultured retina and SC at 1–4 WIV.
S. Hirota et al. / Neuroscience Letters 545 (2013) 96–101
suggests that RGC axons formed connections and terminated in the upper regions of the SC slices. Next, we tested whether any functional connections formed between the retinas and SCs (n = 10). Evoked responses of cocultured retinas and SCs were observed immediately after 10 electrical stimulations (n = 3). Fig. 3B shows an example of samples with responses that were detected at several electrodes in both regions. Fig. 3C shows electrical activity recorded from the electrodes (retina: RE1-2, SC: SC1-2) before and after the application of
99
electrical stimulation. First, electrical stimulations were applied to the retinal region. In the retinal region, firing rates increased after application at electrode RE1, and in the SC region, there was a marked increase in firing rates after stimulation at electrodes SC1 and SC2. Next, electrical stimulations were applied to the SC region. In the SC region, brief increases in the number of spikes were observed at electrodes SC1 and SC2 after the application of stimulation, and in the retinal region, spike firings occurred after stimulation at electrode RE1. Average firing rates for 10 stimulations at electrodes RE1 and SC2 are shown in Fig. 3D. A significant
Fig. 3. (A) The retinotectal pathway formed by the retinal projection to the SC in culture. An image of a retina-SC co-culture at 8 days in vitro (DIV) (left) and a DiI-stained image (right). When the DiI was placed into the retina, labeled RGC axons anterogradely followed into the SC and formed neural connections between the retina and the SC (arrowhead). RGC axon terminals were observed in the surface of SC (arrow). (B–D) Evoked responses of the SC and retina after electrical stimulations. (B) A phase-contrast image of a retina-SC co-culture on the MEA substrates at 3 WIV (electrode RE1, 2: retinal region, electrode SC1, 2: SC region). (C) Electrical activity at electrodes 1–4 before and after electrical stimulation to the retinal or SC region. (D) Spike rate averages for 10 trials compared before and after stimulation (n = 10, ±SE, * p < 0.05, ** p < 0.01, Student’s t-test).
100
S. Hirota et al. / Neuroscience Letters 545 (2013) 96–101
Fig. 4. (A and B) Spontaneous activity patterns recorded from co-cultured or single-cultured SC slices from 0.5 to 4 WIV. (A) In most of the co-cultured SC slices, spike firings were observed at early stages, synchronized bursts occurred after 2 WIV. (B) In most of the single-cultured SC slices, only spike firings were observed for four weeks. (C) Spontaneous activity patterns recorded from co-cultured retinas and SCs at distances of 500 m without neural connections at 0.5–4 WIV. In all SC slices, no synchronized burst was observed for four weeks.
increase in firing rate was observed at both electrodes after applying stimulations to each region.
generated during four weeks (n = 1). These spontaneous activities in the SCs were significantly different from those in the co-cultured SCs.
3.3. Spontaneous activity in co-cultured and single-cultured SC 4. Discussion To examine the effect of retina on the functional development of SC, retina-derived soluble factors or signal inputs from the retina, we compared activity patterns in co-cultured (n = 10) and singlecultured (n = 10) SC slices. We classified the activity patterns of SCs into three types: synchronized burst around SC slice, local synchronized burst, and sporadic firings. The activity patterns of SCs recorded from co-cultures and from single-cultures are shown in Fig. 4A and B. In the co-cultures, spontaneous SC activity was detectable at 0.5 WIV, and the probability of detecting synchronized burst increased with culture time to a peak at 3 WIV. The sporadic spikes gradually changed into a synchronized burst in most of the co-cultured SCs (Fig. 4A). In the single-cultures, sporadic firings were recorded at 0.5 WIV from SC slices in which synchronized bursts were detectable within four weeks, but the percentage of such samples was just 20% of the total number (Fig. 4B). The shift of spontaneous activity patterns was different between co-cultured and single-cultured SC slices. 3.4. Spontaneous activity in co-cultured SC without neural connection with retina To examine the effect of retina-derived soluble factors on the development of SC, we observed the spontaneous activity in cocultured SCs when plating PDMS compartments between retinas and SCs for four weeks (Fig. 1). The distance between the retina and the SC was approximately 500 m. Activity patterns recorded from both retinas and SCs are shown in Fig. 4C. In the retina region, local synchronized bursts generated until 2 WIV, but then no activity occurred after 3 WIV. In the SC region, only sporadic firings
4.1. In vitro reconstruction of retinotectal pathways We established the reconstruction of the retinotectal pathway in culture and developed a simultaneous recording method of spontaneous activity from retinas and SCs during the formation of functional connections. The shifts of spontaneous activity were recorded from retinas and SCs during the developmental stage of co-cultured neural networks. In the co-cultured retinas, sporadic firings were detected from several electrodes and continuous firing was recorded after 3 WIV. In the co-cultured SCs, sporadic firings began in the first week. Thereafter the frequency of firings increased and synchronized bursts occurred around the SCs after 2–3 weeks. Spontaneous neural activity plays an important role in neural development and in the formation of neural circuits [13]. In cultured hippocampal neural networks, spontaneous uncorrelated firings transform into synchronized activity during development [14]. In vivo, the earliest spontaneous activity in rat SC is detectable at P5–P6. Thereafter, more numerous neurons generate spontaneous firings [18]. Synchronized bursts occur around acute SC slices after P16 [19]. Our results seem to indicate that spontaneous activity occurs during SC development, changing from sporadic firing to synchronized burst. The morphological connection between the retina and SC was observed by DiI. Evoked responses were detected from SCs and retinas after applying electrical stimulations. Our data indicate that the functional connections between retinas and SCs were formed in vitro.
S. Hirota et al. / Neuroscience Letters 545 (2013) 96–101
101
4.2. Signal input from retina triggers spontaneous activity in SC
References
We suggest that the different spontaneous activity patterns in SCs between co-cultures and single-cultures, especially the occurrence of synchronized burst around the SC, depends on retinaderived soluble factors or signal inputs from retinas. Spontaneous synchronized activity occurs commonly in a developing brain and plays an essential role in the formation of functional circuits, i.e., synaptic connections and network plasticity [14]. Although the mechanism of the onset of spontaneous synchronized activity remains unclear [17], its activity provides an indication of neural development in SC. In this experiment, SC slices were cultured with retinas using PDMS compartments that enabled the exchange of soluble factors but prevented the formation of a neural connection. The PDMS compartment enables us to focus on the effect of retina-derived soluble factors on the spontaneous activity pattern in the SC. However, none of the SC slices exhibited synchronized bursts within 4 WIV. This suggests that soluble factors from retinas do not trigger spontaneous firings of SC neurons. In our previous work, spontaneous activity in single-cultured SC slices taken from P0-2 rats was recorded for four weeks [8]. Spontaneous firings were detectable from all SC slices until 1 WIV, and then, synchronized bursts were observed from 70% of the SC slices within the four weeks. This shift of activity pattern was different from one of single-cultured SC slices taken from E17-18 rats. Here, we suggest that the difference of spontaneous activity pattern arises from the signal input to SC via synaptic connections. The synaptic transmission between RGC and SC neuron begins at P0 [4] and may initiate the firing of SC neuron because spontaneous bursts of RGCs strengthen connections to the target and are sufficient to depolarize LGN neurons [20]. Additionally, the cell death occurred in the SC when the signal input from the retina was blocked by Tetrodotoxin, a sodium channel blocker [5]. Therefore, the signal input from the retina might trigger the spontaneous activity in the SC as well as the LGN and the activity itself also promotes SC development.
[1] A. Bansal, J.H. Singer, B.J. Hwang, W. Xu, A. Beaudet, M.B. Feller, Mice lacking specific nicotinic acetylcholine receptor subunits exhibit dramatically altered spontaneous activity patterns and reveal a limited role for retinal waves in forming ON and OFF circuits in the inner retina, J. Neurosci. 20 (2000) 7672–7681. [2] A.R. Chandrasekaran, D.T. Plas, E. Gonzalez, M.C. Crair, Evidence for an instructive role of retinal activity in retinotopic map refinement in the superior colliculus of the mouse, J. Neurosci. 25 (2005) 6929–6938. [3] D.A. Feldheim, D.D. O’Leary, Visual map development: bidirectional signaling, bifunctional guidance molecules, and competition, Cold Spring Harb. Perspect. Biol. 2 (2010). [4] R.E. Foster, B.W. Connors, S.G. Waxman, Rat optic nerve: electrophysiological, pharmacological and anatomical studies during development, Brain Res. 255 (1982) 371–386. [5] L. Galli-Resta, M. Ensini, E. Fusco, A. Gravina, B. Margheritti, Afferent spontaneous electrical activity promotes the survival of target cells in the developing retinotectal system of the rat, J. Neurosci. 13 (1993) 243–250. [6] W. Guido, F.S. Lo, R.S. Erzurumlu, An in vitro model of the kitten retinogeniculate pathway, J. Neurophysiol. 77 (1997) 511–516. [7] A. Hafidi, G. Lanjun, D.H. Sanes, Age-dependent failure of axon regeneration in organotypic culture of gerbil auditory midbrain, J. Neurobiol. 41 (1999) 267–280. [8] S. Hirota, K. Inoue, H. Moriguchi, Y. Takayama, Y. Jimbo, Recording of spontaneous activity in cultured superior colliculus slices using microelectrode array, IEEJ Trans. EIS. 131 (2011) 1355–1360. [9] S. Hirota, H. Moriguchi, Y. Takayama, Y. Jimbo, In vitro reconstruction of visual information-processing system using co-cultured retina and superior colliculus, IEEJ Trans. EIS. 129 (2009) 1923–1928. [10] S.K. Itaya, S. Fortin, S. Molotchnikoff, Evolution of spontaneous activity in the developing rat superior colliculus, Can. J. Physiol. Pharmacol. 73 (1995) 1372–1377. [11] Y. Jimbo, A. Kawana, Electrical stimulation and recording from cultured neurons using a planar electrode array, Bioelectrochem. Bioenerg. 29 (1992) 193–204. [12] H.A. Johnson, D.V. Buonomano, Development and plasticity of spontaneous activity and up states in cortical organotypic slices, J. Neurosci. 27 (2007) 5915–5925. [13] L.C. Katz, C.J. Shatz, Synaptic activity and the construction of cortical circuits, Science 274 (1996) 1133–1138. [14] X. Li, W. Zhou, M. Liu, Q. Luo, Synchronized spontaneous spikes on multielectrode array show development of cultured neuronal network, Conf. Proc. IEEE Eng. Med. Biol. Soc. 2 (2005) 2134–2137. [15] F.S. Lo, R.R. Mize, Retinal input induces three firing patterns in neurons of the superficial superior colliculus of neonatal rats, J. Neurophysiol. 81 (1999) 954–958. [16] R.D. Lund, A.H. Bunt, Prenatal development of central optic pathways in albino rats, J. Comp. Neurol. 165 (1976) 247–264. [17] A.K. McCabe, S.L. Chisholm, H.L. Picken-Bahrey, W.J. Moody, The self-regulating nature of spontaneous synchronized activity in developing mouse cortical neurons, J. Physiol. 577 (2006) 155–167. [18] S. Molotchnikoff, S.K. Itaya, Functional development of the neonatal rat retinotectal pathway, Dev. Brain Res. 72 (1993) 300–304. [19] P. Phongphanphanee, K. Kaneda, T. Isa, Spatiotemporal profiles of field potentials in mouse superior colliculus analyzed by multichannel recording, J. Neurosci. 28 (2008) 9309–9318. [20] W. Rockhill, J.L. Kirkman, M.M. Bosma, Spontaneous activity in the developing mouse midbrain driven by an external pacemaker, Dev. Neurobiol. 69 (2009) 689–704. [21] W.L. Shew, H. Yang, T. Petermann, R. Roy, D. Plenz, Neuronal avalanches imply maximum dynamic range in cortical networks at criticality, J. Neurosci. 29 (2009) 15595–15600. [22] N.C. Spitzer, Electrical activity in early neuronal development, Nature 444 (2006) 707–712. [23] J.W. Triplett, M.T. Owens, J. Yamada, G. Lemke, J. Cang, M.P. Stryker, D.A. Feldheim, Retinal input instructs alignment of visual topographic maps, Cell 139 (2009) 175–185. [24] A. Wizenmann, M. Bahr, Growth characteristics of ganglion cell axons in the developing and regenerating retino-tectal projection of the rat, Cell Tissue Res. 290 (1997) 395–403.
5. Conclusion We reconstructed retinotectal pathways on MEA substrates to observe spontaneous activity in both retinas and SCs and the functional connections between tissues for one month. The difference of activity patterns in SCs between co-cultures and single-cultures was observed. Activity patterns of co-cultured SCs in a no-neural connection with retinas were similar to single-cultured SCs. These results suggest that not retina-derived soluble factors but rather direct signal inputs from retinas that mainly trigger spontaneous firings of SC neurons and affect the development of SCs. Acknowledgements We thank Dr. Joseph H. McCarty at Department of Neurosurgery, University of Texas MD Anderson Cancer Center for constructive comments on the paper.
Contents lists available at SciVerse ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
In vitro reconstruction and functional development of the superior colliculus in the retinotectal pathway Shinya Hirota ∗ , Hiroyuki Moriguchi 1 , Atsushi Saito, Kousuke Inoue, Akitoshi Murakami, Kiyoshi Kotani, Yasuhiko Jimbo Department of Human and Engineered Environmental Studies, Graduate School of Frontier Sciences, University of Tokyo, 5-1-5, Kashiwanoha, Kashiwa, Chiba 277-8563, Japan
h i g h l i g h t s
g r a p h i c a l
a b s t r a c t
• Retinotectal pathways were reconstructed on MEA substrates by cocultures. • Spontaneous activities of co-cultured retinas and SCs were recorded for one month. • Functional connections between co-cultured retinas and SCs were formed. • The signal input from retina affects the functional development of SC.
a r t i c l e
i n f o
Article history: Received 6 February 2013 Received in revised form 12 March 2013 Accepted 6 April 2013 Keywords: Superior colliculus (SC) Microelectrode array (MEA) substrates Co-culture Spontaneous activity Development
a b s t r a c t In order to examine the formation of a neural network and the functional development of a visual pathway, we performed in vitro reconstruction of the retinotectal pathway using organotypic explants and co-culture methods. Retinas and superior colliculus (SC) slices obtained from embryonic rats were cocultured on microelectrode array (MEA) substrates for four weeks. We observed retinal ganglion cell neurites innervating SC slices that evoked responses in retinas or SC slices after applying electrical stimulation. Functional connections between retinas and SC slices were formed in the cultures. At the same time, spontaneous electrical activities were recorded from both the retinas and SC slices over the four weeks. In the co-cultured SC slices, sporadic firings were initially observed at 3–4 days in vitro (DIV), and thereafter the frequency of spontaneous firing increased and synchronized activities occurred after two weeks in vitro (WIV). In most of the single-cultured SC slices, however, only sporadic firings were observed over four weeks. In addition, the retinas and SC slices were co-cultured to enable the exchange of soluble factors with each other via culture medium but not via direct neural connections. The activity patterns resembled ones of single-cultured SC slices. These results suggest that signal inputs from retinas through direct neural connections affect the development of SCs in the retinotectal pathway. © 2013 Elsevier Ireland Ltd. All rights reserved.
1. Introduction
∗ Corresponding author at: Department of Neurosurgery, University of Texas MD Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030, USA. Tel.: +1 713 834 6240. E-mail addresses: [email protected], sh [email protected] (S. Hirota). 1 Present address: Quantitative Biology Center (QBiC), Integrated Biodevice Research Unit, RIKEN, 2-2-3, Minatojima-minamimachi, Chuo-ku, Kobe, Hyogo 6500047, Japan. 0304-3940/$ – see front matter © 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.neulet.2013.04.020
The retinotectal pathway, which is formed by the retinal projection to the superior colliculus (SC), plays an important role in the control of orienting eye and head movement in vertebrates. Precise topographic projection from the retina to the SC is established, as is the lateral geniculate nucleus (LGN) and the primary visual cortex [23]. The retinotectal pathway is useful to elucidate the formation and development of neural circuits in visual pathways using in vitro systems [24].
S. Hirota et al. / Neuroscience Letters 545 (2013) 96–101
The vertebrate retinotectal pathway uses a multistep developmental process [3]. In rats, retinal ganglion cell (RGC) axons initially enter the SC on embryonic day 16 (E16) and form synapses with SC neurons on E17 [16]. At the same time, spontaneous firings of RGCs occur on E16 [1] and are transmitted along optic axons on postnatal day 0 (P0) [4]. Previous studies have suggested that signal inputs from the retina regulate the refinement of connections between RGCs and SC neurons by cell death and axon retraction [2]. Electrical activity is also thought to be important after embryonic development for the refinement of connections in brain [22]. Little is known about spontaneous activity patterns in SC during development. The earliest spontaneous activity of SC was recorded at P5-6 [10]. The synaptic transmission between RGC and SC neuron was detectable on P1 using in vitro models [15] and P10 using in vivo models [18]. In general, spontaneous activity is triggered by external input [12], but the activity in the cortex occurs spontaneously in the absence of any signal input from other regions [21]. The former suggests that signal input from retinas evokes SC activity and therefore affects SC development, while the latter seems to indicate that SC activity occurs spontaneously or is affected by retina-derived soluble factors. In this study, we investigated which were more critical in the development of SC: signal input from retinas or retina-derived factors. By development, we mean the construction of functional circuits in SC using retinotectal pathways on MEA substrates. The MEA substrate is a culture dish embedded with multiple electrodes to enable non-invasive recording of electrical activity and stimulation with various spatial-temporal patterns [11]. Using organotypic culture on MEA substrates enables us to monitor spontaneous neural activity of both tissues while preserving organotypic organizations over several weeks. The co-culture model does not display an in vivo-like morphology or functions, but it does enable the investigation of the effect of signal input or soluble factors by separating from other regions [6]. In our previous study, spontaneous activity from co-cultured retinas and SCs were recorded up to 11 days in vitro (DIV) using P0-2 rats [9]. Considering that the major developmental process of retinotectal pathways occurs before the eye-opening around P14 [18], it is important to record spontaneous activities in SCs in retinotectal pathways for at least 2–3 weeks in vitro (WIV). To promote cell survival and neurite growth [7], we attempted co-cultures of retinas and SCs taken from younger rats, E17-18. As a result, retinotectal pathways, in which we recorded spontaneous activity for four weeks and signaling between retinas and SCs, were constructed on MEA substrates. Next, in order to investigate the effect of the retina on SC development, spontaneous activity patterns in co-cultured or single-cultured SCs were analyzed for four weeks. We also examined whether signal input from the retina or retina-derived soluble factors affected the development of SC by recording spontaneous activity in co-cultured SCs when plating the polydimethylsiloxane (PDMS, Dow Corning Toray) compartments between the retinas and SCs.
97
10 m × 10 m, and arranged with a 50 m distance. The former was used for the co-cultures of retinas and SC slices and the latter was used for the single-cultured SC slices. 2.2. Organotypic cultures All animal experiments were carried out in accordance with the guidelines for animal experiments at University of Tokyo. A total of seven pregnant Wistar rats at E17-18 (Charles River) were deeply anesthetized with diethyl ether and the embryos were surgically removed. The retinas and SC slices were taken from these embryos using ice-cold artificial cerebrospinal fluid (ACSF) saturated with 95% O2 and 5% CO2 . The ACSF contained 149 mM NaCl, 2.8 mM KCl, 2 mM CaCl2 , 1 mM MgCl2 , 10 mM HEPES, and 10 mM d-glucose, pH 7.4–7.5. Retinas were dissected free from the surrounding tissues of eyes, and were cut into 4–6 pieces. SCs were cut into 300 m-slices using a vibratome (Pro 7 Linear Slicer, Dosaka EM). Both tissues were placed on polyethyleneimine (PEI, Sigma) and laminin (Gibco) double-coated MEA substrates and separated at a distance of 500 m (Fig. 1). The culture medium was a 1:1 mixture of Dulbecco’s modified Eagle’s medium and Ham’s F-12 (Invitrogen) supplemented with 5% fetal bovine serum (FBS, Invitrogen), 1% Penicillin/Streptomycin (Invitrogen), 5 g/ml Insulin (Sigma), 20 nM Progesterone (Sigma), 20 nM Hydrocortisone (Sigma), 100 g/ml Transferrin (Sigma), 100 M Putrescine (Sigma), and 30 nM sodium selenite (Wako), pH 7.3. The co-cultures were maintained in a humidified 5% CO2 incubator at 37 ◦ C. The medium was exchanged every 2–3 days. 2.3. Live cell imaging with DiI The neuronal projections from retinas to SCs in vitro were confirmed. Retinas were incubated for 50 min at 37 ◦ C in the DiI solution consisting of a Vybrant DiI cell-labeling solution (Invitrogen) diluted 200 times with ACSF and then washed with ACSF three times. Retinas labeled with DiI were cultured with SC slices on culture dishes at a distance of more than 500 m. 2.4. Extracellular recordings and electrical stimulations Recordings and stimulations were carried out in a box with an atmosphere kept at 37 ◦ C and 95% air/5% CO2 to minimize damage to cells. Electrical signals were amplified with a 64 channel amplifier (NF Corp., Gain: ×10,000) and filtered with a band pass filter (100 Hz–5 kHz). The amplified signals were then sampled at intervals of 40 s using an AD converter (National Instrument) and recorded on the hard disk of a PC. Evoked responses in one tissue were observed after applying electrical stimulation to the other tissue to confirm the formation of functional connections between retina and SC. Stimulation waveforms consisted of 10 biphasic pulses (0.1 ms at +1 V followed by 0.1 ms at −1 V) at a frequency of 0.1 Hz.
2. Materials and methods
2.5. Fabrication of PDMS compartment
2.1. Fabrication of MEA substrates
We fabricated a PDMS compartment that enabled the exchange of soluble factors via culture medium but prevented the formation of neural connections. The compartment had a box shape of 200–300 m wide and 250 m high.
We fabricated MEA substrates with two types of electrode patterns (Fig. 1). The electrode circuit was made of indium tin oxide slightly coated on the glass substrate and covered with Si-based photoresist as the insulation layer. Recording and stimulation sites were coated with platinum black to reduce the interface impedance. The first type of MEA substrates was split into 4 × 8 electrodes separated by 500 m. Each electrode within a group was 30 m × 30 m, and was separated from their neighbors by 180 m. The other type was split into 8 × 8 electrodes, each
3. Results 3.1. Co-culture of retina and SC on MEA substrates Retinas and superior colliculus (SC) slices were co-cultured on MEA substrates for four weeks (n = 10, Fig. 2). Neurites extended
98
S. Hirota et al. / Neuroscience Letters 545 (2013) 96–101
Fig. 1. Retina-SC co-culture and SC culture on microelectrode array (MEA) substrates. Two different kinds of MEA substrates were used for the co-culture and the SC culture. The co-cultured retina and SC slice were arranged on each block of MEA substrates such that the tip of the retina was oriented toward the surface of the SC in view of the anatomical structure of the retinotectal pathway as oriented in vivo. In addition, retinas were mounted such that the RGCs generating action potentials contact with the electrodes on the MEA substrates.
from the retinas and SC slices until 24 h in vitro. The neurite outgrowth and cell migration from the retina extended toward the SC slice but did not expand to the opposite sides (Fig. 2A). SC axons grew radially. The time course of spontaneous activity in co-cultured retina and SC over four weeks is shown in Fig. 2B as raster plots. The thresholds of signal detection were set as the five times standard deviations of amplitudes in the noise region. In the retina, spontaneous activity was observed after 2–3 days in vitro (DIV). Sporadic firings generated from 1 to 3 WIV and continuous firings were recorded from several electrodes at 4 WIV. In the SC, spontaneous activity was recorded after 3 DIV and the shift of activity pattern was observed until 4 WIV. Sporadic firings occurred until 1 WIV. Thereafter, firing rates increased and
synchronized bursts generated locally at 3 WIV, and synchronized bursts around the SC slice generated at 4 WIV. 3.2. Functional connection formed between co-cultured retina and SC Morphological connections between retinas and SC slices were established after 1 WIV (Fig. 3A, left). To examine whether neuronal projections were formed anterogradely in the retina-SC co-cultures, we placed DiI crystals into retinas and cultured with SC slices on culture dishes at a distance of more than 500 m (Fig. 3A, right). Consequently, the retinal cells and their axons that were labeled with DiI could be identified. The connection and the upper regions of the co-cultured SC slices were robustly stained. This
Fig. 2. Spontaneous activity recorded from the co-cultured retina and SC slice over four weeks. (A) A phase-contrast image of a retina-SC co-culture on the MEA substrates at 4 weeks in vitro (WIV). (B) The spontaneous activity recorded from co-cultured retina and SC at 1–4 WIV.
S. Hirota et al. / Neuroscience Letters 545 (2013) 96–101
suggests that RGC axons formed connections and terminated in the upper regions of the SC slices. Next, we tested whether any functional connections formed between the retinas and SCs (n = 10). Evoked responses of cocultured retinas and SCs were observed immediately after 10 electrical stimulations (n = 3). Fig. 3B shows an example of samples with responses that were detected at several electrodes in both regions. Fig. 3C shows electrical activity recorded from the electrodes (retina: RE1-2, SC: SC1-2) before and after the application of
99
electrical stimulation. First, electrical stimulations were applied to the retinal region. In the retinal region, firing rates increased after application at electrode RE1, and in the SC region, there was a marked increase in firing rates after stimulation at electrodes SC1 and SC2. Next, electrical stimulations were applied to the SC region. In the SC region, brief increases in the number of spikes were observed at electrodes SC1 and SC2 after the application of stimulation, and in the retinal region, spike firings occurred after stimulation at electrode RE1. Average firing rates for 10 stimulations at electrodes RE1 and SC2 are shown in Fig. 3D. A significant
Fig. 3. (A) The retinotectal pathway formed by the retinal projection to the SC in culture. An image of a retina-SC co-culture at 8 days in vitro (DIV) (left) and a DiI-stained image (right). When the DiI was placed into the retina, labeled RGC axons anterogradely followed into the SC and formed neural connections between the retina and the SC (arrowhead). RGC axon terminals were observed in the surface of SC (arrow). (B–D) Evoked responses of the SC and retina after electrical stimulations. (B) A phase-contrast image of a retina-SC co-culture on the MEA substrates at 3 WIV (electrode RE1, 2: retinal region, electrode SC1, 2: SC region). (C) Electrical activity at electrodes 1–4 before and after electrical stimulation to the retinal or SC region. (D) Spike rate averages for 10 trials compared before and after stimulation (n = 10, ±SE, * p < 0.05, ** p < 0.01, Student’s t-test).
100
S. Hirota et al. / Neuroscience Letters 545 (2013) 96–101
Fig. 4. (A and B) Spontaneous activity patterns recorded from co-cultured or single-cultured SC slices from 0.5 to 4 WIV. (A) In most of the co-cultured SC slices, spike firings were observed at early stages, synchronized bursts occurred after 2 WIV. (B) In most of the single-cultured SC slices, only spike firings were observed for four weeks. (C) Spontaneous activity patterns recorded from co-cultured retinas and SCs at distances of 500 m without neural connections at 0.5–4 WIV. In all SC slices, no synchronized burst was observed for four weeks.
increase in firing rate was observed at both electrodes after applying stimulations to each region.
generated during four weeks (n = 1). These spontaneous activities in the SCs were significantly different from those in the co-cultured SCs.
3.3. Spontaneous activity in co-cultured and single-cultured SC 4. Discussion To examine the effect of retina on the functional development of SC, retina-derived soluble factors or signal inputs from the retina, we compared activity patterns in co-cultured (n = 10) and singlecultured (n = 10) SC slices. We classified the activity patterns of SCs into three types: synchronized burst around SC slice, local synchronized burst, and sporadic firings. The activity patterns of SCs recorded from co-cultures and from single-cultures are shown in Fig. 4A and B. In the co-cultures, spontaneous SC activity was detectable at 0.5 WIV, and the probability of detecting synchronized burst increased with culture time to a peak at 3 WIV. The sporadic spikes gradually changed into a synchronized burst in most of the co-cultured SCs (Fig. 4A). In the single-cultures, sporadic firings were recorded at 0.5 WIV from SC slices in which synchronized bursts were detectable within four weeks, but the percentage of such samples was just 20% of the total number (Fig. 4B). The shift of spontaneous activity patterns was different between co-cultured and single-cultured SC slices. 3.4. Spontaneous activity in co-cultured SC without neural connection with retina To examine the effect of retina-derived soluble factors on the development of SC, we observed the spontaneous activity in cocultured SCs when plating PDMS compartments between retinas and SCs for four weeks (Fig. 1). The distance between the retina and the SC was approximately 500 m. Activity patterns recorded from both retinas and SCs are shown in Fig. 4C. In the retina region, local synchronized bursts generated until 2 WIV, but then no activity occurred after 3 WIV. In the SC region, only sporadic firings
4.1. In vitro reconstruction of retinotectal pathways We established the reconstruction of the retinotectal pathway in culture and developed a simultaneous recording method of spontaneous activity from retinas and SCs during the formation of functional connections. The shifts of spontaneous activity were recorded from retinas and SCs during the developmental stage of co-cultured neural networks. In the co-cultured retinas, sporadic firings were detected from several electrodes and continuous firing was recorded after 3 WIV. In the co-cultured SCs, sporadic firings began in the first week. Thereafter the frequency of firings increased and synchronized bursts occurred around the SCs after 2–3 weeks. Spontaneous neural activity plays an important role in neural development and in the formation of neural circuits [13]. In cultured hippocampal neural networks, spontaneous uncorrelated firings transform into synchronized activity during development [14]. In vivo, the earliest spontaneous activity in rat SC is detectable at P5–P6. Thereafter, more numerous neurons generate spontaneous firings [18]. Synchronized bursts occur around acute SC slices after P16 [19]. Our results seem to indicate that spontaneous activity occurs during SC development, changing from sporadic firing to synchronized burst. The morphological connection between the retina and SC was observed by DiI. Evoked responses were detected from SCs and retinas after applying electrical stimulations. Our data indicate that the functional connections between retinas and SCs were formed in vitro.
S. Hirota et al. / Neuroscience Letters 545 (2013) 96–101
101
4.2. Signal input from retina triggers spontaneous activity in SC
References
We suggest that the different spontaneous activity patterns in SCs between co-cultures and single-cultures, especially the occurrence of synchronized burst around the SC, depends on retinaderived soluble factors or signal inputs from retinas. Spontaneous synchronized activity occurs commonly in a developing brain and plays an essential role in the formation of functional circuits, i.e., synaptic connections and network plasticity [14]. Although the mechanism of the onset of spontaneous synchronized activity remains unclear [17], its activity provides an indication of neural development in SC. In this experiment, SC slices were cultured with retinas using PDMS compartments that enabled the exchange of soluble factors but prevented the formation of a neural connection. The PDMS compartment enables us to focus on the effect of retina-derived soluble factors on the spontaneous activity pattern in the SC. However, none of the SC slices exhibited synchronized bursts within 4 WIV. This suggests that soluble factors from retinas do not trigger spontaneous firings of SC neurons. In our previous work, spontaneous activity in single-cultured SC slices taken from P0-2 rats was recorded for four weeks [8]. Spontaneous firings were detectable from all SC slices until 1 WIV, and then, synchronized bursts were observed from 70% of the SC slices within the four weeks. This shift of activity pattern was different from one of single-cultured SC slices taken from E17-18 rats. Here, we suggest that the difference of spontaneous activity pattern arises from the signal input to SC via synaptic connections. The synaptic transmission between RGC and SC neuron begins at P0 [4] and may initiate the firing of SC neuron because spontaneous bursts of RGCs strengthen connections to the target and are sufficient to depolarize LGN neurons [20]. Additionally, the cell death occurred in the SC when the signal input from the retina was blocked by Tetrodotoxin, a sodium channel blocker [5]. Therefore, the signal input from the retina might trigger the spontaneous activity in the SC as well as the LGN and the activity itself also promotes SC development.
[1] A. Bansal, J.H. Singer, B.J. Hwang, W. Xu, A. Beaudet, M.B. Feller, Mice lacking specific nicotinic acetylcholine receptor subunits exhibit dramatically altered spontaneous activity patterns and reveal a limited role for retinal waves in forming ON and OFF circuits in the inner retina, J. Neurosci. 20 (2000) 7672–7681. [2] A.R. Chandrasekaran, D.T. Plas, E. Gonzalez, M.C. Crair, Evidence for an instructive role of retinal activity in retinotopic map refinement in the superior colliculus of the mouse, J. Neurosci. 25 (2005) 6929–6938. [3] D.A. Feldheim, D.D. O’Leary, Visual map development: bidirectional signaling, bifunctional guidance molecules, and competition, Cold Spring Harb. Perspect. Biol. 2 (2010). [4] R.E. Foster, B.W. Connors, S.G. Waxman, Rat optic nerve: electrophysiological, pharmacological and anatomical studies during development, Brain Res. 255 (1982) 371–386. [5] L. Galli-Resta, M. Ensini, E. Fusco, A. Gravina, B. Margheritti, Afferent spontaneous electrical activity promotes the survival of target cells in the developing retinotectal system of the rat, J. Neurosci. 13 (1993) 243–250. [6] W. Guido, F.S. Lo, R.S. Erzurumlu, An in vitro model of the kitten retinogeniculate pathway, J. Neurophysiol. 77 (1997) 511–516. [7] A. Hafidi, G. Lanjun, D.H. Sanes, Age-dependent failure of axon regeneration in organotypic culture of gerbil auditory midbrain, J. Neurobiol. 41 (1999) 267–280. [8] S. Hirota, K. Inoue, H. Moriguchi, Y. Takayama, Y. Jimbo, Recording of spontaneous activity in cultured superior colliculus slices using microelectrode array, IEEJ Trans. EIS. 131 (2011) 1355–1360. [9] S. Hirota, H. Moriguchi, Y. Takayama, Y. Jimbo, In vitro reconstruction of visual information-processing system using co-cultured retina and superior colliculus, IEEJ Trans. EIS. 129 (2009) 1923–1928. [10] S.K. Itaya, S. Fortin, S. Molotchnikoff, Evolution of spontaneous activity in the developing rat superior colliculus, Can. J. Physiol. Pharmacol. 73 (1995) 1372–1377. [11] Y. Jimbo, A. Kawana, Electrical stimulation and recording from cultured neurons using a planar electrode array, Bioelectrochem. Bioenerg. 29 (1992) 193–204. [12] H.A. Johnson, D.V. Buonomano, Development and plasticity of spontaneous activity and up states in cortical organotypic slices, J. Neurosci. 27 (2007) 5915–5925. [13] L.C. Katz, C.J. Shatz, Synaptic activity and the construction of cortical circuits, Science 274 (1996) 1133–1138. [14] X. Li, W. Zhou, M. Liu, Q. Luo, Synchronized spontaneous spikes on multielectrode array show development of cultured neuronal network, Conf. Proc. IEEE Eng. Med. Biol. Soc. 2 (2005) 2134–2137. [15] F.S. Lo, R.R. Mize, Retinal input induces three firing patterns in neurons of the superficial superior colliculus of neonatal rats, J. Neurophysiol. 81 (1999) 954–958. [16] R.D. Lund, A.H. Bunt, Prenatal development of central optic pathways in albino rats, J. Comp. Neurol. 165 (1976) 247–264. [17] A.K. McCabe, S.L. Chisholm, H.L. Picken-Bahrey, W.J. Moody, The self-regulating nature of spontaneous synchronized activity in developing mouse cortical neurons, J. Physiol. 577 (2006) 155–167. [18] S. Molotchnikoff, S.K. Itaya, Functional development of the neonatal rat retinotectal pathway, Dev. Brain Res. 72 (1993) 300–304. [19] P. Phongphanphanee, K. Kaneda, T. Isa, Spatiotemporal profiles of field potentials in mouse superior colliculus analyzed by multichannel recording, J. Neurosci. 28 (2008) 9309–9318. [20] W. Rockhill, J.L. Kirkman, M.M. Bosma, Spontaneous activity in the developing mouse midbrain driven by an external pacemaker, Dev. Neurobiol. 69 (2009) 689–704. [21] W.L. Shew, H. Yang, T. Petermann, R. Roy, D. Plenz, Neuronal avalanches imply maximum dynamic range in cortical networks at criticality, J. Neurosci. 29 (2009) 15595–15600. [22] N.C. Spitzer, Electrical activity in early neuronal development, Nature 444 (2006) 707–712. [23] J.W. Triplett, M.T. Owens, J. Yamada, G. Lemke, J. Cang, M.P. Stryker, D.A. Feldheim, Retinal input instructs alignment of visual topographic maps, Cell 139 (2009) 175–185. [24] A. Wizenmann, M. Bahr, Growth characteristics of ganglion cell axons in the developing and regenerating retino-tectal projection of the rat, Cell Tissue Res. 290 (1997) 395–403.
5. Conclusion We reconstructed retinotectal pathways on MEA substrates to observe spontaneous activity in both retinas and SCs and the functional connections between tissues for one month. The difference of activity patterns in SCs between co-cultures and single-cultures was observed. Activity patterns of co-cultured SCs in a no-neural connection with retinas were similar to single-cultured SCs. These results suggest that not retina-derived soluble factors but rather direct signal inputs from retinas that mainly trigger spontaneous firings of SC neurons and affect the development of SCs. Acknowledgements We thank Dr. Joseph H. McCarty at Department of Neurosurgery, University of Texas MD Anderson Cancer Center for constructive comments on the paper.