- Email: [email protected]
Characterization of rat spinal cord neurons cultured in defined media on microelectrode arrays
Neuroscience Letters 271 (1999) 179±182
Characterization of rat spinal cord neurons cultured in de®ned media on microelectrode arrays P. Manos, J.J. Pancrazio, M.G. Coulombe, W. Ma, D.A. Stenger* Center for Bio/Molecular Science and Engineering, Code 6910, Naval Research Laboratory, Washington, DC 20375, USA Received 12 March 1999; received in revised form 6 July 1999; accepted 6 July 1999
Abstract Previous efforts to utilize mammalian spinal cord neurons as biosensor elements have relied on neuronal: glial cocultures maintained in serum-containing media. We have examined the feasibility of culturing primary spinal cord neurons in serum-free medium, modi®ed for neuronal longevity, on fabricated microelectrode arrays. Embryonic day 15 rat spinal cord cells were plated on trimethoxysilyl-propyldiethylenetriamine coated microelectrode arrays comprised of gold recording sites passivated with silicon nitride. Immunocytochemistry was performed to verify the presence of neurons and quantitatively assess astrocytes using antibodies against glial ®brillary acidic protein on the silicon nitride substrates. Modi®cations to culture media enabled viable neuronal culture to extend from approximately 14 days in vitro (DIV) to 40 DIV on the arrays containing only 1:1 ^ 0:5% (mean ^ SEM) astrocytes. Extracellular recording revealed tetrodotoxin-sensitive spontaneous electrical activity from the enriched neuronal culture. Threshold detection of extracellular potentials showed an increase in spike rate as a function of glutamate concentration with neurotoxicity at elevated levels. This approach suggests that functional measures related to biosensor applications, pharmacological screening, or the evaluation of neurological disease models can be implemented in a de®ned culture system. q 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: De®ned media; Extracellular recording; Immunocytochemistry; Longevity; Microelectrode array; Serum-free media; Spinal cord; Cultured neurons
Biosensors utilize a biological component as a sensing element compatible with electrical, chemical, acoustic, or optical transduction. Assays utilizing cells as sensors offer the advantage of providing insight into the functional, physiological signi®cance of a toxin. Our long-term goal is the development of biosensor systems that can provide a rapid assessment of potential environmental threats. To this end, we are interested in the utility of excitable cells as sensor elements where the perturbation of cellular biopotentials might indicate the presence of a biologically-active agent. Extracellular recording of neuronal action potentials has been proposed as a basis for cell-based functional assay or biosensor operation [5,6]. Past work has relied on a coculture of neurons and glia in serum-containing media to provide neuronal longevity [4]. In this co-culture, neurons tend to grow on astrocytes which form a barrier between neurons and microelectrode contacts leading to attenuation of bioelectrical signals [3]. Pure neuronal cultures offer the * Corresponding author. Tel.: 11-202-404-6035; fax: 11-202767-9598. E-mail address: [email protected] (D.A. Stenger)
advantage of interpreting the pharmacological data as a direct effect on neurons rather than an indirect effect from the co-cultured glial cells. Our previous effort to generate glia-free neuron culture resulted in cultures with a lifetime of 14±17 days in vitro (DIV) [11], consistent with other reports for neuronal longevity [12,16]. Long-term cultures are required for neurons to develop and gain mature phenotypic and functional properties necessary for electrophysiological and pharmacological studies. In the present study, extracellular recording and immunocytochemistry techniques were used to address the feasibility of a puri®ed population of neurons, cultured under serum-free conditions, for cellular assay development. We report a modi®ed media formulation that extends neuronal longevity and function to 40 DIV. Spinal cord neurons were isolated by enzymatic dissociation of ventral and dorsal tissue obtained from embryonic day 15 Sprague±Dawley rat embryos primarily as described for hippocampal neurons using papain [13]. Cells were cultured without exposure to serum or mitotic inhibitors in Eagles minimal essential medium (MEM) modi®ed as
0304-3940/99/$ - see front matter q 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S03 04 - 394 0( 9 9) 00 52 0- 0
180
P. Manos et al. / Neuroscience Letters 271 (1999) 179±182
follows to yield MEM/N31: N3 [12], which includes 10 mg/ml BSA, 200 mg/ml transferrin, 200 mM putrescine, 60 nM sodium selenite, 20 ng/ml triiodothyronine, 10 mg/ ml insulin, and 40 ng/ml corticosterone, and supplemented with 14 mM creatine, 0.5 mM ethanolamine, and 1 mM bhydroxybutyrate and 1 mM fumarate [9]. The cells were plated at a density of 8 £ 10 4/cm 2 in MEM/N31 onto microelectrode arrays coated with trimethoxysilyl-propyldiethylenetriamine (DETA, United Chemicals, Bristol, PA). The arti®cial surfaces composed of DETA self-assembled monolayers (SAMs) have some similarities to polylysine ®lms in their physical characteristics and their ability to support growth and neurite extension [14]. A 50% medium change was performed at days 4, 7, 10, 14, and once a week thereafter. Cells were maintained at 378C in a humidi®ed incubator with 5% CO2/95% air. Cultures were prepared for immuno¯uorescent analysis by paraformaldehyde ®xation, and Triton X-100 permeablization as previously described [10]. The primary and secondary antibodies were diluted in PBS with 5% bovine serum albumin (BSA), pH 7.5, and were incubated sequentially for 60±90 mm at room temperature. The arrays were treated with the primary antibody solution consisting of a 1:500 dilution of monoclonal mouse anti-glial ®brillary acidic protein (GFAP) antibody (Chemicon, Temecula, CA) and a 1:300 dilution of rabbit anti-neuron-speci®c enolase antibody (Chemicon) in PBS with 0.05% saponin and 1% BSA. After three brief rinses with PBS, the arrays were stained for approximately 1 h at room temperature in a secondary antibody solution containing donkey anti-mouse Fab(2) fragment of IgG conjugated with ¯uorescein (Jackson Immunoresearch Laboratories, West Grove, PA), prepared at a 1:75 dilution in PBS with 0.05% saponin and 1% bovine serum albumin. To quantify intact cells, Hoechst 33258 (Bio-Rad, Life Sciences Research, Hercules, CA), an intercolating dye with blue emission speci®c for double stranded DNA, was added during the second antibody incubation at a ®nal concentration of 0.25±0.5 mg/ml as described previously [10]. Control experiments where sera replaced the primary antibody were performed to determine conditions for acceptable background levels. Samples were viewed and photographed on an upright epi¯uorescence microscope (Nikon, model E-800) allowing simultaneous excitation of the ¯uorophores. Planar microelectrode arrays were designed and fabricated at the Center for Integrated Systems at Stanford University. Each array has 32 gold microelectrode sites that are 14 mm in diameter. Details of array fabrication and microelectrode preparation have been described earlier [11]. The arrays were cleaned and coated with DETA for cell culture as described previously [8]. Biopotentials from the microelectrode array were bandpass ®ltered (low pass comer frequency 300 Hz; high pass comer frequency 3 kHz) and ampli®ed by 10 000 using a high input impedance ampli®er (model DAM50, World Precision Instruments, Sarasota, FL). To achieve electrical connection between
the array bond pad and the DAM50, a micromanipulator was used to position gold leads for direct contact. A stainless steel housing was ®tted around the planar array to form a 0.5 ml-chamber and allow single-ended measurements by grounding the bathing solution. Extracellular activity was
Fig. 1. Primary rat spinal cord neurons cultured on microelectrode arrays. Data were representative of the analysis of ®ve cultures. (A) Phase contrast photomicrograph showing neuronal cell bodies and processes cultured on microelectrode arrays fabricated with industry standard thin-®lm photolithography [11]; culture was 7 DIV. (B) Double immuno¯uorescence photomicrograph identifying nuclei in intact cells (blue) and an extremely low level of astrocytes (green) using Hoechst 33258 and antibodies against glial ®brillary acidic protein, respectively; culture was 15 DIV. Previous immunostaining work had shown that the vast majority of the intact cells stained positively for neuron speci®c enolase as expected for the neuron-enriched culture.
P. Manos et al. / Neuroscience Letters 271 (1999) 179±182
recorded using a perfusion system with an extracellular solution composed of (in mM): NaCl, 140; KCl, 3.5; CaCl2, 1.2; MgSO4, 1.2; NaHCO3, 4.17; NaH2 P04, 1.3; dglucose, 10: pH to 7.2. A custom automatic temperature controller permitted the cell bathing solution to be maintained at 36±378C. Comparisons of the proportions of active electrode sites versus days in culture were accomplished by testing the difference between proportions [15] with Pvalues less than 0.05 considered signi®cant. As shown in Fig. 1A, DETA SAMs presented a suitable surface for the culture of neuronal cells, consistent with earlier work [8,14]. As expected, nearly all the cells on the microelectrode arrays stained positively for neuron speci®c enolase (data not shown), indicating the predominance of neurons in the culture. To assess the presence of astrocytes among the neurons cultured on the microelectrode arrays, immunocytochemistry was utilized after 7 days in culture. For such quantitative assessment, the Hoechst stain offered a clear representation of cell number, the vast majority of which were neuronal. As shown in Fig. 1B, the percentage of astrocytes on the microelectrode arrays was exceedingly low reaching only 1:1 ^ 0:5% (mean ^ STD, n 8 arrays). After 5 weeks in serum-free culture conditions, neuronal cells exhibited poor viability characterized by retraction of processes, phase opaque somas, and cellular detachment from the surface. To determine whether or not spontaneous bioelectrical activity could be detected in these minimal, astrocyte-free cultures, we examined extracellular potentials using the microelectrode arrays. Data were collected from 24 arrays cultured at least 10 up to 33 DIV. Consistent with previous work [11], spontaneous extracellular potentials ranged from 100±800 mV peak-to-peak and varied in form and time course at an individual microelectrode (Fig. 2A), presumably due to variable coupling of cell bodies and/or processes to recording sites. An additional complicating factor my be the presence of complex spikes, or multispike bursts, which have been identi®ed previously in neuronal extracellular recordings [17]. As shown on the expanded time scale of Fig. 2A, individual brief duration spikes consistent with action potentials could be resolved within the recordings. Reversible inhibition of the spike activity by the Na 1 channel blocker tetrodotoxin veri®ed the physiologic basis for the observed extracellular potentials (Fig. 2A); a similar elimination in spike activity was observed in two other experiments. For quanti®cation of spike activity, a threshold for spike crossing was used to detect and record events. Fig. 2B shows changes in spike frequency with serial additions of the excitatory neurotransmitter glutamate resulting in a concentration range of 10± 100 mM. Increased concentrations of glutamate elevated spike frequency until reaching a threshold for neurotoxicity at 100 mM. A similar elevation in spike activity and subsequent toxicity was observed in two other experiments. There was a slight increase in the number of active microelectrode sites with longer DIV. While the percentage of microelectrode sites exhibiting activity was statistically unchanged
181
Fig. 2. Extracellular recording from neuron-enriched culture under serum-free conditions. (A) Reversible inhibition of spontaneous electrical activity by the Na 1 channel blocker tetrodotoxin (30 mM) indicating the physiologic basis of the neuronal spikes; culture was 34 DW. (B) Dose-dependent stimulatory effect of glutamate on extracellular neuronal potentials. An increased frequency of spikes was observed with serial elevations in the glutamate concentration. Sustained exposure to 100 mM glutamate resulted in functional neurotoxicity; culture was 16 DIV.
for cultures 10±20 DIV at 32% (activity detected on 105 of 328 total sites), the percentage signi®cantly (P , 0:05) rose to 48% (25 of 52 sites) for cultures 28±33 DIV. Nevertheless, longer duration DIV led to culture degradation. This brief study demonstrates that extracellular recording can be achieved from neurons cultured in serum-free, de®ned media conditions. The absence of serum and the use of MEM/N31 media greatly diminished astrocytes detectable by immunocytochemistry, indicating that spontaneous extracellular recording can be achieved from an enriched population of mammalian neurons. This approach suggests that pharmacological screening, biosensor applications, or evaluation of neurological disease models can be implemented in a reduced culture system such that any astrocyte contributions to the neuronal response can be systematically evaluated. The conditions used to culture the neurons in the present study, in particular the embryonic status of the donor animal and media conditions, made it extremely unlikely that oligodendrocytes were present in the cultures in a signi®cant proportion [7]. Morphological examination of the cultures indicated that cells with typical oligodendrocyte morphology, i.e. relatively small, dark cell bodies with ®ne processes, were not observed in long-term culture (.2
182
P. Manos et al. / Neuroscience Letters 271 (1999) 179±182
weeks). Functional neurotoxicity at elevated glutamate concentrations was observed, although at a somewhat higher concentration than reported previously (50 mM) [5]. With regard to culture longevity, previous efforts to culture spinal cord neurons in serum-free conditions using MEM/N3 led to viable neurons for only 14 DIV [11]. In contrast, we have extended the longevity of the culture to approach 40 DIV through the use of MEM/N31 containing three key components: fumarate, a metabolite of the tricarboxcylic acid cycle; b -hydroxybutyrate, a preferred energy substrate in developing nervous system [2]; and creatine at relevant in vivo concentrations. Future efforts to approach the long-term neuronal viability (.150 DIV) achieved with neuronal: glia co-culture systems [5] may involve the re®nement of media conditions and/or the array surface. Perhaps, similar supplementation of media such as B27/neurobasal, which can support neuronal culture for over 4 weeks [1], will further extend the longevity of glia-free, de®ned neuronal culture. This work was supported by the OSD Counterproliferation Support Program through DARPA. The opinions and assertions contained herein are the private ones of the authors and are not to be construed as of®cial or re¯ecting the views of the Department of the Navy. The authors thank Dr. Michael Geusz, Bowling Green State University, for consultation in extracellular electrophysiology, and Dr. Joanne D. Andreadis, Naval Research Laboratory, for her insightful comments. [1] Brewer, G.J., Torricelli, J.R., Evege, E.K. and Price, P.J., Optimized survival of hippocampal-neurons in B27-supplemented Neurobasal TM, a new serum-free medium combination. J. Neurosci. Res., 35 (1993) 567±576. [2] Edmond, J., Robbins, R.A., Bergstrom, J.D., Cole, R.A. and deVellis, J., Capacity for substrate utilization in oxidative metabolism by neurons, astrocytes and oligodendrocytes from developing brain in culture. J. Neurosci. Res., 18 (1987) 551±561. [3] Gross, G.W., Internal dynamics of randomized mammalian neuronal networks in culture. In D.A. Stenger and T.M. McKenna (Eds.), Enabling Technologies for Cultured Neuronal Networks, Academic Press, San Diego, CA, 1994, pp. 277±317. [4] Gross, G.W. and Kowalski, J.M., Experimental and theoretical analysis of random nerve cell network dynamics. In P. Antognetti and V. Milutonovic (Eds.), Neural Networks: Concepts, Applications and Implementations, Vol. 4, Prentice Hall, Englewood, NJ, 1991, pp. 47±110. [5] Gross, G.W., Rhoades, B.K., Azzazy, H.M.E. and Wu, M.C.,
[6] [7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15] [16]
[17]
The use of neuronal networks on multielectrode arrays as biosensors. Biosen. Bioelectron., 10 (1995) 553±567. Gross, G.W., Rhoades, B. and Jordan, R., Neuronal networks for biochemical sensing. Sensors Actuators B, 6 (1992) 1±8. Hansson, E., Ronnback, L., Persson, L.I., Lowenthal, A., Noppe, M., Alling, C. and Karlsson, B., Cellular composition of primary cultures from cerebral cortex, striatum, hippocampus, brain-stem and cerebellum. Brain Res., 300 (1984) 9±18. Jung, D.R., Cuttino, D.S., Pancrazio, J.J., Manos, P., Custer, T., Sathanoori, R.S., AIi, L.E., Coulombe, M.G., Bey, P.P., Czamaski, M.A., Borkholder, D.A., Kovacs, G.T.A., Stenger, D.A. and Hickman, J.J., Cell-based sensor microelectrode array characterized by imaging XPS, SEM, impedance measurements and extracellular recordings. J. Vac. Sci. Technol. A, 16 (1998) 1183±1188. Manos, P., Bryan, G.K. and Edmond, J., Creatine kinase activity in cultured postnatal rat brain development and in cultured neurons, astrocytes and oligodendrocytes. J. Neurochem., 56 (1991) 2101±2107. Manos, P. and Edmond, J., Immuno¯uorescent analysis of creatine kinase in cultured astrocytes by conventional and confocal microscopy: a nuclear localization. J. Comp. Neurol., 326 (1992) 273±282. Pancrazio, J.J., Bey, P.P., Lobee, A., SubbaRao, M., HuiChuan, C., Howard, L.L., Gosney, W.M., Borkholder, D.A., Kovacs, G.T.A., Manos, P., Cuttino, D.S. and Stenger, D.A., Description and demonstration of a CMOS ampli®er-based system with measurement and stimulation capability for bioelectrical signal transduction. Biosen. Bioelectron., 13 (1998) 971±979. Romijin, H.J., van Huizen, F. and Wolters, P.S., Towards an improved serum-free, chemically de®ned medium for longterm culturing of cerebral cortex tissue. Neurosci. Biobehav. Rev., 8 (1984) 301±334. Schaf¯ier, A.E., Barker, J.L., Stenger, D.A. and Hickman, J.J., Investigation of the factors necessary for growth of hippocampal neurons in a de®ned system. J. Neurosci. Methods, 62 (1995) 111±119. Stenger, D.A., Pike, C.J., Hickman, J.J. and Cotman, C.W., Surface determinants of neuronal survival and growth on self-assembled monolayers in culture. Brain Res., 630 (1993) 136±147. Walpole, R.E. and Myers, R.H., Probability and Statistics for Engineers and Scientists, Macmillan, New York, 1978. Weiss, S., Pin, J.P., Sebben, M., Kemp, D.E., Sladeczek, F., Gabrion, J. and Bockaert, J., Synaptogenesis of cultured striatal neurons in serum-free medium: a morphological and biochemical study. Proc. Natl. Acad. Sci. USA, 83 (1986) 2238±2242. Welsh, J.P. and Schwarz, C., Multielectrode recording from the cerebellum. In M.A.L. Nicolelis (Ed.), Methods for Neural Ensemble Recordings, CRC Press, Boca Raton, Fl, 1999, pp. 79±100.
Characterization of rat spinal cord neurons cultured in de®ned media on microelectrode arrays P. Manos, J.J. Pancrazio, M.G. Coulombe, W. Ma, D.A. Stenger* Center for Bio/Molecular Science and Engineering, Code 6910, Naval Research Laboratory, Washington, DC 20375, USA Received 12 March 1999; received in revised form 6 July 1999; accepted 6 July 1999
Abstract Previous efforts to utilize mammalian spinal cord neurons as biosensor elements have relied on neuronal: glial cocultures maintained in serum-containing media. We have examined the feasibility of culturing primary spinal cord neurons in serum-free medium, modi®ed for neuronal longevity, on fabricated microelectrode arrays. Embryonic day 15 rat spinal cord cells were plated on trimethoxysilyl-propyldiethylenetriamine coated microelectrode arrays comprised of gold recording sites passivated with silicon nitride. Immunocytochemistry was performed to verify the presence of neurons and quantitatively assess astrocytes using antibodies against glial ®brillary acidic protein on the silicon nitride substrates. Modi®cations to culture media enabled viable neuronal culture to extend from approximately 14 days in vitro (DIV) to 40 DIV on the arrays containing only 1:1 ^ 0:5% (mean ^ SEM) astrocytes. Extracellular recording revealed tetrodotoxin-sensitive spontaneous electrical activity from the enriched neuronal culture. Threshold detection of extracellular potentials showed an increase in spike rate as a function of glutamate concentration with neurotoxicity at elevated levels. This approach suggests that functional measures related to biosensor applications, pharmacological screening, or the evaluation of neurological disease models can be implemented in a de®ned culture system. q 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: De®ned media; Extracellular recording; Immunocytochemistry; Longevity; Microelectrode array; Serum-free media; Spinal cord; Cultured neurons
Biosensors utilize a biological component as a sensing element compatible with electrical, chemical, acoustic, or optical transduction. Assays utilizing cells as sensors offer the advantage of providing insight into the functional, physiological signi®cance of a toxin. Our long-term goal is the development of biosensor systems that can provide a rapid assessment of potential environmental threats. To this end, we are interested in the utility of excitable cells as sensor elements where the perturbation of cellular biopotentials might indicate the presence of a biologically-active agent. Extracellular recording of neuronal action potentials has been proposed as a basis for cell-based functional assay or biosensor operation [5,6]. Past work has relied on a coculture of neurons and glia in serum-containing media to provide neuronal longevity [4]. In this co-culture, neurons tend to grow on astrocytes which form a barrier between neurons and microelectrode contacts leading to attenuation of bioelectrical signals [3]. Pure neuronal cultures offer the * Corresponding author. Tel.: 11-202-404-6035; fax: 11-202767-9598. E-mail address: [email protected] (D.A. Stenger)
advantage of interpreting the pharmacological data as a direct effect on neurons rather than an indirect effect from the co-cultured glial cells. Our previous effort to generate glia-free neuron culture resulted in cultures with a lifetime of 14±17 days in vitro (DIV) [11], consistent with other reports for neuronal longevity [12,16]. Long-term cultures are required for neurons to develop and gain mature phenotypic and functional properties necessary for electrophysiological and pharmacological studies. In the present study, extracellular recording and immunocytochemistry techniques were used to address the feasibility of a puri®ed population of neurons, cultured under serum-free conditions, for cellular assay development. We report a modi®ed media formulation that extends neuronal longevity and function to 40 DIV. Spinal cord neurons were isolated by enzymatic dissociation of ventral and dorsal tissue obtained from embryonic day 15 Sprague±Dawley rat embryos primarily as described for hippocampal neurons using papain [13]. Cells were cultured without exposure to serum or mitotic inhibitors in Eagles minimal essential medium (MEM) modi®ed as
0304-3940/99/$ - see front matter q 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S03 04 - 394 0( 9 9) 00 52 0- 0
180
P. Manos et al. / Neuroscience Letters 271 (1999) 179±182
follows to yield MEM/N31: N3 [12], which includes 10 mg/ml BSA, 200 mg/ml transferrin, 200 mM putrescine, 60 nM sodium selenite, 20 ng/ml triiodothyronine, 10 mg/ ml insulin, and 40 ng/ml corticosterone, and supplemented with 14 mM creatine, 0.5 mM ethanolamine, and 1 mM bhydroxybutyrate and 1 mM fumarate [9]. The cells were plated at a density of 8 £ 10 4/cm 2 in MEM/N31 onto microelectrode arrays coated with trimethoxysilyl-propyldiethylenetriamine (DETA, United Chemicals, Bristol, PA). The arti®cial surfaces composed of DETA self-assembled monolayers (SAMs) have some similarities to polylysine ®lms in their physical characteristics and their ability to support growth and neurite extension [14]. A 50% medium change was performed at days 4, 7, 10, 14, and once a week thereafter. Cells were maintained at 378C in a humidi®ed incubator with 5% CO2/95% air. Cultures were prepared for immuno¯uorescent analysis by paraformaldehyde ®xation, and Triton X-100 permeablization as previously described [10]. The primary and secondary antibodies were diluted in PBS with 5% bovine serum albumin (BSA), pH 7.5, and were incubated sequentially for 60±90 mm at room temperature. The arrays were treated with the primary antibody solution consisting of a 1:500 dilution of monoclonal mouse anti-glial ®brillary acidic protein (GFAP) antibody (Chemicon, Temecula, CA) and a 1:300 dilution of rabbit anti-neuron-speci®c enolase antibody (Chemicon) in PBS with 0.05% saponin and 1% BSA. After three brief rinses with PBS, the arrays were stained for approximately 1 h at room temperature in a secondary antibody solution containing donkey anti-mouse Fab(2) fragment of IgG conjugated with ¯uorescein (Jackson Immunoresearch Laboratories, West Grove, PA), prepared at a 1:75 dilution in PBS with 0.05% saponin and 1% bovine serum albumin. To quantify intact cells, Hoechst 33258 (Bio-Rad, Life Sciences Research, Hercules, CA), an intercolating dye with blue emission speci®c for double stranded DNA, was added during the second antibody incubation at a ®nal concentration of 0.25±0.5 mg/ml as described previously [10]. Control experiments where sera replaced the primary antibody were performed to determine conditions for acceptable background levels. Samples were viewed and photographed on an upright epi¯uorescence microscope (Nikon, model E-800) allowing simultaneous excitation of the ¯uorophores. Planar microelectrode arrays were designed and fabricated at the Center for Integrated Systems at Stanford University. Each array has 32 gold microelectrode sites that are 14 mm in diameter. Details of array fabrication and microelectrode preparation have been described earlier [11]. The arrays were cleaned and coated with DETA for cell culture as described previously [8]. Biopotentials from the microelectrode array were bandpass ®ltered (low pass comer frequency 300 Hz; high pass comer frequency 3 kHz) and ampli®ed by 10 000 using a high input impedance ampli®er (model DAM50, World Precision Instruments, Sarasota, FL). To achieve electrical connection between
the array bond pad and the DAM50, a micromanipulator was used to position gold leads for direct contact. A stainless steel housing was ®tted around the planar array to form a 0.5 ml-chamber and allow single-ended measurements by grounding the bathing solution. Extracellular activity was
Fig. 1. Primary rat spinal cord neurons cultured on microelectrode arrays. Data were representative of the analysis of ®ve cultures. (A) Phase contrast photomicrograph showing neuronal cell bodies and processes cultured on microelectrode arrays fabricated with industry standard thin-®lm photolithography [11]; culture was 7 DIV. (B) Double immuno¯uorescence photomicrograph identifying nuclei in intact cells (blue) and an extremely low level of astrocytes (green) using Hoechst 33258 and antibodies against glial ®brillary acidic protein, respectively; culture was 15 DIV. Previous immunostaining work had shown that the vast majority of the intact cells stained positively for neuron speci®c enolase as expected for the neuron-enriched culture.
P. Manos et al. / Neuroscience Letters 271 (1999) 179±182
recorded using a perfusion system with an extracellular solution composed of (in mM): NaCl, 140; KCl, 3.5; CaCl2, 1.2; MgSO4, 1.2; NaHCO3, 4.17; NaH2 P04, 1.3; dglucose, 10: pH to 7.2. A custom automatic temperature controller permitted the cell bathing solution to be maintained at 36±378C. Comparisons of the proportions of active electrode sites versus days in culture were accomplished by testing the difference between proportions [15] with Pvalues less than 0.05 considered signi®cant. As shown in Fig. 1A, DETA SAMs presented a suitable surface for the culture of neuronal cells, consistent with earlier work [8,14]. As expected, nearly all the cells on the microelectrode arrays stained positively for neuron speci®c enolase (data not shown), indicating the predominance of neurons in the culture. To assess the presence of astrocytes among the neurons cultured on the microelectrode arrays, immunocytochemistry was utilized after 7 days in culture. For such quantitative assessment, the Hoechst stain offered a clear representation of cell number, the vast majority of which were neuronal. As shown in Fig. 1B, the percentage of astrocytes on the microelectrode arrays was exceedingly low reaching only 1:1 ^ 0:5% (mean ^ STD, n 8 arrays). After 5 weeks in serum-free culture conditions, neuronal cells exhibited poor viability characterized by retraction of processes, phase opaque somas, and cellular detachment from the surface. To determine whether or not spontaneous bioelectrical activity could be detected in these minimal, astrocyte-free cultures, we examined extracellular potentials using the microelectrode arrays. Data were collected from 24 arrays cultured at least 10 up to 33 DIV. Consistent with previous work [11], spontaneous extracellular potentials ranged from 100±800 mV peak-to-peak and varied in form and time course at an individual microelectrode (Fig. 2A), presumably due to variable coupling of cell bodies and/or processes to recording sites. An additional complicating factor my be the presence of complex spikes, or multispike bursts, which have been identi®ed previously in neuronal extracellular recordings [17]. As shown on the expanded time scale of Fig. 2A, individual brief duration spikes consistent with action potentials could be resolved within the recordings. Reversible inhibition of the spike activity by the Na 1 channel blocker tetrodotoxin veri®ed the physiologic basis for the observed extracellular potentials (Fig. 2A); a similar elimination in spike activity was observed in two other experiments. For quanti®cation of spike activity, a threshold for spike crossing was used to detect and record events. Fig. 2B shows changes in spike frequency with serial additions of the excitatory neurotransmitter glutamate resulting in a concentration range of 10± 100 mM. Increased concentrations of glutamate elevated spike frequency until reaching a threshold for neurotoxicity at 100 mM. A similar elevation in spike activity and subsequent toxicity was observed in two other experiments. There was a slight increase in the number of active microelectrode sites with longer DIV. While the percentage of microelectrode sites exhibiting activity was statistically unchanged
181
Fig. 2. Extracellular recording from neuron-enriched culture under serum-free conditions. (A) Reversible inhibition of spontaneous electrical activity by the Na 1 channel blocker tetrodotoxin (30 mM) indicating the physiologic basis of the neuronal spikes; culture was 34 DW. (B) Dose-dependent stimulatory effect of glutamate on extracellular neuronal potentials. An increased frequency of spikes was observed with serial elevations in the glutamate concentration. Sustained exposure to 100 mM glutamate resulted in functional neurotoxicity; culture was 16 DIV.
for cultures 10±20 DIV at 32% (activity detected on 105 of 328 total sites), the percentage signi®cantly (P , 0:05) rose to 48% (25 of 52 sites) for cultures 28±33 DIV. Nevertheless, longer duration DIV led to culture degradation. This brief study demonstrates that extracellular recording can be achieved from neurons cultured in serum-free, de®ned media conditions. The absence of serum and the use of MEM/N31 media greatly diminished astrocytes detectable by immunocytochemistry, indicating that spontaneous extracellular recording can be achieved from an enriched population of mammalian neurons. This approach suggests that pharmacological screening, biosensor applications, or evaluation of neurological disease models can be implemented in a reduced culture system such that any astrocyte contributions to the neuronal response can be systematically evaluated. The conditions used to culture the neurons in the present study, in particular the embryonic status of the donor animal and media conditions, made it extremely unlikely that oligodendrocytes were present in the cultures in a signi®cant proportion [7]. Morphological examination of the cultures indicated that cells with typical oligodendrocyte morphology, i.e. relatively small, dark cell bodies with ®ne processes, were not observed in long-term culture (.2
182
P. Manos et al. / Neuroscience Letters 271 (1999) 179±182
weeks). Functional neurotoxicity at elevated glutamate concentrations was observed, although at a somewhat higher concentration than reported previously (50 mM) [5]. With regard to culture longevity, previous efforts to culture spinal cord neurons in serum-free conditions using MEM/N3 led to viable neurons for only 14 DIV [11]. In contrast, we have extended the longevity of the culture to approach 40 DIV through the use of MEM/N31 containing three key components: fumarate, a metabolite of the tricarboxcylic acid cycle; b -hydroxybutyrate, a preferred energy substrate in developing nervous system [2]; and creatine at relevant in vivo concentrations. Future efforts to approach the long-term neuronal viability (.150 DIV) achieved with neuronal: glia co-culture systems [5] may involve the re®nement of media conditions and/or the array surface. Perhaps, similar supplementation of media such as B27/neurobasal, which can support neuronal culture for over 4 weeks [1], will further extend the longevity of glia-free, de®ned neuronal culture. This work was supported by the OSD Counterproliferation Support Program through DARPA. The opinions and assertions contained herein are the private ones of the authors and are not to be construed as of®cial or re¯ecting the views of the Department of the Navy. The authors thank Dr. Michael Geusz, Bowling Green State University, for consultation in extracellular electrophysiology, and Dr. Joanne D. Andreadis, Naval Research Laboratory, for her insightful comments. [1] Brewer, G.J., Torricelli, J.R., Evege, E.K. and Price, P.J., Optimized survival of hippocampal-neurons in B27-supplemented Neurobasal TM, a new serum-free medium combination. J. Neurosci. Res., 35 (1993) 567±576. [2] Edmond, J., Robbins, R.A., Bergstrom, J.D., Cole, R.A. and deVellis, J., Capacity for substrate utilization in oxidative metabolism by neurons, astrocytes and oligodendrocytes from developing brain in culture. J. Neurosci. Res., 18 (1987) 551±561. [3] Gross, G.W., Internal dynamics of randomized mammalian neuronal networks in culture. In D.A. Stenger and T.M. McKenna (Eds.), Enabling Technologies for Cultured Neuronal Networks, Academic Press, San Diego, CA, 1994, pp. 277±317. [4] Gross, G.W. and Kowalski, J.M., Experimental and theoretical analysis of random nerve cell network dynamics. In P. Antognetti and V. Milutonovic (Eds.), Neural Networks: Concepts, Applications and Implementations, Vol. 4, Prentice Hall, Englewood, NJ, 1991, pp. 47±110. [5] Gross, G.W., Rhoades, B.K., Azzazy, H.M.E. and Wu, M.C.,
[6] [7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15] [16]
[17]
The use of neuronal networks on multielectrode arrays as biosensors. Biosen. Bioelectron., 10 (1995) 553±567. Gross, G.W., Rhoades, B. and Jordan, R., Neuronal networks for biochemical sensing. Sensors Actuators B, 6 (1992) 1±8. Hansson, E., Ronnback, L., Persson, L.I., Lowenthal, A., Noppe, M., Alling, C. and Karlsson, B., Cellular composition of primary cultures from cerebral cortex, striatum, hippocampus, brain-stem and cerebellum. Brain Res., 300 (1984) 9±18. Jung, D.R., Cuttino, D.S., Pancrazio, J.J., Manos, P., Custer, T., Sathanoori, R.S., AIi, L.E., Coulombe, M.G., Bey, P.P., Czamaski, M.A., Borkholder, D.A., Kovacs, G.T.A., Stenger, D.A. and Hickman, J.J., Cell-based sensor microelectrode array characterized by imaging XPS, SEM, impedance measurements and extracellular recordings. J. Vac. Sci. Technol. A, 16 (1998) 1183±1188. Manos, P., Bryan, G.K. and Edmond, J., Creatine kinase activity in cultured postnatal rat brain development and in cultured neurons, astrocytes and oligodendrocytes. J. Neurochem., 56 (1991) 2101±2107. Manos, P. and Edmond, J., Immuno¯uorescent analysis of creatine kinase in cultured astrocytes by conventional and confocal microscopy: a nuclear localization. J. Comp. Neurol., 326 (1992) 273±282. Pancrazio, J.J., Bey, P.P., Lobee, A., SubbaRao, M., HuiChuan, C., Howard, L.L., Gosney, W.M., Borkholder, D.A., Kovacs, G.T.A., Manos, P., Cuttino, D.S. and Stenger, D.A., Description and demonstration of a CMOS ampli®er-based system with measurement and stimulation capability for bioelectrical signal transduction. Biosen. Bioelectron., 13 (1998) 971±979. Romijin, H.J., van Huizen, F. and Wolters, P.S., Towards an improved serum-free, chemically de®ned medium for longterm culturing of cerebral cortex tissue. Neurosci. Biobehav. Rev., 8 (1984) 301±334. Schaf¯ier, A.E., Barker, J.L., Stenger, D.A. and Hickman, J.J., Investigation of the factors necessary for growth of hippocampal neurons in a de®ned system. J. Neurosci. Methods, 62 (1995) 111±119. Stenger, D.A., Pike, C.J., Hickman, J.J. and Cotman, C.W., Surface determinants of neuronal survival and growth on self-assembled monolayers in culture. Brain Res., 630 (1993) 136±147. Walpole, R.E. and Myers, R.H., Probability and Statistics for Engineers and Scientists, Macmillan, New York, 1978. Weiss, S., Pin, J.P., Sebben, M., Kemp, D.E., Sladeczek, F., Gabrion, J. and Bockaert, J., Synaptogenesis of cultured striatal neurons in serum-free medium: a morphological and biochemical study. Proc. Natl. Acad. Sci. USA, 83 (1986) 2238±2242. Welsh, J.P. and Schwarz, C., Multielectrode recording from the cerebellum. In M.A.L. Nicolelis (Ed.), Methods for Neural Ensemble Recordings, CRC Press, Boca Raton, Fl, 1999, pp. 79±100.