NEURONAL TL'rllOR CELLS WITH EXCITABLE MEMBRANES GROWhT IN VITRO BY PHILLIP NELSON, %'INFRED RUFFSER, AND I\~ARSHALL NIRENBEHG N.~TIO?ihL INSTITUTES OF Hb:.iLTH, BETHESD.\, M.\RYL.\ND Communicated July 14, 1969 Abstract.--l\Iouse neuroblastoma cells grown ill z4ro are capable of genernting action potentials in response to electrical stimulation. A wide spectrum of re- sponses was observed with different cells ranging from passive responses to action potentials. A hypothesis is proposed concerning the acquisition of electrical ex- citabilit,y as neuroblasts mature into neurons. Tissue culture methodology affords many opportunities to study molecular aspects of information processin g by cells from the nervous system.`, 2 During the last few years we have investigated the properties of tumor cells and normal cells derived from the nervous system and grown in L&O. During the course of these studies, Augusti-Tocco and Sate told us that. they bad established clonal lines of cells from a transplantable mouse neuroblastomn3 that contain choline acetylase, acetyl cholinesterase, and tyrosine hydroxylase.4 Human neuro- blastomas also have been cultured in ailroj and have been shown to contain en- zymes related to norepinephrine synthesis and catabolism.`+8 We have investigated electrophysiologic properties of the mouse neuroblnstoma of Augusti-Tocco and Sato and wish to report that neuroblastoma cells, grown in vitro, are capable of generating a&ion potentials that are characteristic of neurons. lI~at,erials and dle&&.--Seuroblastoma C-1300 is a spontaneous tumor nlaintained since 1940 by serial transplantation in strain .1/J mice. Mice bearing this tumor were obtained from the Jackson Memorial Laboratory, Bar Harbor, Maine. Keuroblastoma cells were dissociated and grown in vitro as described by Augusti-Tocco and Satoh except that Dulbecco's modification of Eagle `S medium9 was used in place of F-IO, and cells were grown in an atmosphere of 10% CO, and 90% air. Cells have not been cloned. The average doubling time of cell.5 is 18 hr; hence, cells have been re- peatedly trypsinized and subcultured. The medium contains inorganic ions in the following concentrations (mM) : Sa+, 154; I<+, 5.4; Ca++, 1.8; Mg++, 0.8; Cl-, 118; and HCOS-, 44. Plastic petri dishes, 35 or 60 mm in diameter and containing neuroblastoma cells in the medium described above, lvere placed in a plastic chamber on the stage of a Zeiss inverted microscope. The temperature of the medium, measured with a thermistor probe, was maintained at 35oC with a heating element below the microscope stage. A humidified mixture of 10% CO* and 90% air flowed over the surface of the culture, Glass micropipette electrodes with tip diameter <0.5 P and filled with 3 91 KC1 were used to record transmembrane potentials from the cultured cells. The resistance of the electrodes ranged from 10 to 200 megohms; most were from 20 to 80 megohms when measured in the culture medium. The-microelectrode was arranged in a bridge circuit and connected to a Bak unity gain amplifier. This circuit allows transmembrane poten- tials to be recorded while currents are passed across the cell membranes through the im- paling microelectrode.1o Re~.Q.s.-The morphologic characteristics of mouse neuroblastoma cells cultured in &w resemble those described by Augusti-Tocco and Sato.4 How- 1004 VOL. 64. 1969 BIOCHEXISTRY: XELSOX ET AL. 100.5 ever, the cells used in the present study are not clonal cell lines, so a heterogeneous ceil population is expected. The following procedure was used to study the electrophysiologic properties of neuroblastoma cells in vitro. The cell was photographed and the tip of a micro- electrode was insert,ed in the interior of the cell; another electrode was immersed in the extracellular medium; within several minutes, when stable records could be obtained, the difference in voltage across the cell membrane recorded by the intracellular electrode was determined. Pulses of current 50-100 msec in dura- tion were then passed through the electrodes and the changes in voltage across the cell membrane were determined as a function of time. Steady current was then passed through the electrodes to adjust the voltage across the cell membrane to about -60 mv. Pulses of current again were passed through the electrodes and perturbations in voltage across the cell membrane were determined. This procedure allowed us to test, the excitable properties of the cell membrane at the voltage level obtained after penetration of the cell by the electrode and at a standard level of -GO mv. Our objective was to determine whether neuroblastoma cells are electrically excitable in vitro and whether they generate neuron-like action potentials in response to stimuli. Of 259 cells that were examined, 149 were studied in detail. Ilost of the cells were found to have "active" membranes; however, the degree of activity varied widely. Excitable cells were found that were capable of generat- ing action potentials, as well as cells with less active membranes. An example of an excitable cell generating action potential is shown in Figure ln and B. A resting potential of -40 mv was obtained upon penetration of the cell by the electrode. A steady current of about 2 namp was passed through the electrode to adjust the voltage across the cells membrane to -tG mv. Then a series of five pulses of stimulating current approximately 70 msec in duration, but, differing in intensity, were passed across the membrane (the upper traces of Fig. lB), evoking the changes in membrane voltage corresponding to the lower traces of Figure 1B. The weakest stimulus, labeled a in the Figure, evoked a small, smoothly increasing perturbation in voltage. The most intense stimulating pulses elicited action potentials. Partial responses were elicited by stimuli of intermediate intensity. Kine per cent of the cells examined were capable of generating action potentials. Stimulation of such cells for five msec, a pulse of current briefer than the duration of the action potential, elicited an action potential. Action potentials were elicited two days after cells were dissociated with trypsin and then subcultured. A cell exhibiting only a partial response is shown in Figure 1C and D. A resting potential of -20 mv was found on penetration of this cell, but even when the voltage across the membrane was adjusted to -GO mv with steady current, a pulse of stimulating current elicited only a partial response. A fully developed action potential could not be evoked in this cell. The observed response to a pulse of current is dependent upon the level of transmembrane voltage at the time of stimulation (Fig. 3). When the voltage across the membrane of the cell shown in Figure 24 was adjusted to -60 mv, a current pulse of one nnmp evoked a partial response (Fig. 2s). A current pulse FIG. I.--(A, n). The large cell IIC:\~ 111~ renter of the photograph was studied eleca- t rophysiologioally. Cells had been sllb- carlltured 7 days earlier and were past the logarithmic phase of growth. Total time in v&o was 103 days. The 100 p bar applies to both (A ) and (C). Intraaellrdar record- ings from the cell are shown in (8). Five oscilloscope traces are superimposed. The lines shown in the upper part, of this and subsequent, figllres correspond to the (:\II'- rents that, were passed t.hrorlgh the elee- trades across the cell membral,e (rj~lhl o,rlind~); the lower curves irtdic::~lc the &allges itI voltage ac'ross the (~~11 Mom- brane (IP~( o,dina[e) evoked by these ~II'- reuts. After the resting potential was de- termilled, 2 namp of steady crrrrellt were passed t,hrorlgh the elect,rodes to at1jtM the voll,age across the cell membrane 1.0 -65 mv. Then t.he cell membrane ww stimulated with pldses of current as it)- dicated by ON and OFF. The weakest' stimrllating pldse, labeled a, elicited a small passive voltage response, also labeled a. The most, intense stimlllating pulses (e) elichited action potentials. (C, U). The large ~11 near 1 he WIIICI was examilted elec:trol~h~siologic~:~lly. Tha ill is from 1 he same u11l~~lu: tlc3Ghocl if1 (.{ ). The prdse of stim\dating crlrrellt is indicated by the Ilpper lirles, the restdtallt, change iu membratke voltage ,I)? the lowet wlrve. The inflec+tion WI the rlsltlg phase of ihe voltage CIIrve c~orresponds to :I partial respolrse of the rell membral~r. I I 1 I I ON OFF 2 -30 G -1 i-50 I -70 / f t I 1 I I I I J 0 20 40 60 80 100 MILLISECONDS 0 20 40 60 80 100 MILLISECONDS VOL. 64, 1969 BIOCHEMISTRY: NELSON ET .4L. 1OOT -60.7 1 I I I I 1 I f 1 1 0 20 40 60 80 100 0 20 40 60 80 100 MILL I SECONDS MI LLI SECONDS FIG. Z.-(A). The cell is from a culture that had been treated with trypsin to dissociate t,he cells 5 days earlier. The cell was exnmiued electrophysiologicatly soon after the culture had passed I he logarithmic phase of growth. Total time in vitro was 108 days. (B, C, D). ;\n active response is shown in (B); examples of rectification in (C, D). The lines labeled D uud II indicate pulses of current that evoke changes in voltage across the cell membrane labeled rl and h, respectively. S designates t,he steady current used to adjust the v11lt age across I hex cell memhrsne, s. of the same intensit#y but of opposite direction elicited a different wave form, a smoothly changing, uninflected, voltage transient. In Figure 2C, the steady voltage across the cell membrane was adjusted to - 25 mv. Then, current pulses that increased or decreased trnnsmembrane voltage did not evoke partial re- sponses; however, the decrease in trnnsmembrane voltage was smaller than the increase. When the steady voltage was adjusteli to -35 mv (Fig. 20), asym- metric responses (rectification) were obtained, but the direction of asymmetry was reversed. However, at a steady membrane voltage of -15 mv, symmetric responses to pulses of current were obtained (not shown). The cell illustrated in Figure 3d generated repetitive action potentials in the absence of a stimulating current (Fig. 3B) and also in response to a pulse of stimulating current (Fig. 3C). The action potential shown in Figure 30 was eJicited by turning qff a pulse of current that increased the voltage across the cell membrane. The response appears to be identical to the "off-excitation" phenomenon exhibited by many neurons. -Action potentials, partial responses, and rectificat,ion are characteristic of 1008 BIOCHEMISTRY: .YEI.SO:\- ET A-1 I,. I'ROC. pr'. A. s. 0 40 60 120 160 200 0 40 80 120 160 200 MILLISECONDS MILLISECONDS FIG. 3.-(A). The large round cell shown in the upper left part of the photograph and in- dicated by the microelectrode was studied. The culture was in the stationary phase of growth: total time in M-0 was 66 days. (B). Repetitive action potentials recorded soon after the cell was penetrated by the micro- electrode. The action potentials occurred in the absence of stimulating current. The dashes in this and the following figure represent portions of the records that, were filled in dnring re- production of the records. (C). Two action potentials elicited bp a pulse of current. (D). An example of "off-excitation " is shown. An action potential was evoked by turlGng 08 the pulse of current. The top of the action potential is not shown. "active" cell membranes and are due to voltage-dependent changes in the perme- ability of cell membranes." "Active" membranes could not be demonstrated in 36 per cent of the cells studied. Such membranes are termed "passive" be- cause &en intense pulses of current produce changes in membrane voltage that have the simple form of voltage trace a in Figure 1C or trace h in Figure 2C and are symmetrical. The relative frequencies of membrane responses are summarized in Table 1. "Active" responses were observed with 65 per cent of the cells examined. Seu- ron-like action potentials were observed with 9 per cent of the ceils, partial TABLF. 1. Summary of cell membrane responses. Membrane properties Active membranes Action potential Partial response Rectification Passive membranes Total Number of cells 13 25 58 53 149 Total (%I Resting potential (mvolts) 9 40 17 24 39 20 35 17 100 22 VOL. 64, 1969 BIOCHEMISTRY: NELSON ET AL. 1009 responses with 17 per cent, and recttication with 39 per cent of the cells. Thirty- five per cent of the cells examined had `Lpassive" membranes. Discussion.-The results show that mouse neuroblastoma cells grown in vitro are electrically excitable and are capable of generating action potentials. Neu- rons with action potentials have been found in explant12* l3 and dissociated cellI cultures. However, normal mature neurons usually do not divide, whereas at least some neuroblastoma tumor cells retain the ability to divide in vitro and also exhibit properties expected of neurons. Augusti-Tocco and Sate' have cloned these cells and have shown that they contain choline acetylase, acetyl cholin- esterase, and tyrosine hydroxylase. We have confirmed these observations and have also shown that the cells synthesize catechols and catechol derivatives in vitro. l6 Sixty-five per cent of the cells examined electrophysiologically had "active" membranes. However, a wide spectrum of responses was observed with different cells, even when membrane properties were studied at the same transmembrane voltage and under relatively stable recording conditions. Although cell injury may have contributed to the variability of results, it is unlikely that such a wide spectrum of responses would be observed with mature neurons. We propose as a working hypothesis that the range of membrane properties ob- served is expressed in the following sequence a-s neuroblasts mature into mUTOm: passive TespoIzses, delayed rectification, partial respor1ses, and action potentials. The molecular events responsible for these phenomenon are not fully understood. It is possible that some neuroblastoma cells differentiate in vitro and one or more genetic programs corresponding to functional action potentials are expressed, or that events related to cell division affect the capacity of cells to generate action potentials. Alternatively, cultures may contain many neuroblastoma cell lines that differ genetically and are fixed at different stages of differentiation. Experi- ments are in progress to resolve these questions and also to determine whether neuroblastoma cells are capable of forming synapses in vitro. The mouse neuro- blastoma cell system of Augusti-Tocco and `Sato and other cell lines derived from the nervous system provide many opportunities to explore various aspects of neurobiology. Note added in proof: Similar results have now been obtained with clonal neuroblastoma lines. We would like to express our appreciation to Gabriella &u&i-Tocco and Gordon Sato for telling us about their experiments with neuroblastoma cells prior to publication. 1 Murray, M. R., in Cells and Tissues in Culture, vol. 2, ed. E. N.Wiier (New York: Academic Press, 1965), p. 373. z Geiger, R. S., in Intemahnal Review of Neurobiology, vol. 5, ed. C. C. Pfeiffer and J. R. Smythies (New York: Academic Press, 1963), p. 1. a Augusti-Tocco, C., and G. Sate, Seminar at the Neurosciences Research Program meeting, Feb. 5, 1969, Brookline, Mass. 4 Augusti-Tocco, G., and G. Sato, these PROCEEDINGS, in press. 6 Murray, XI. R., and A. P. Stout, Amer. J. Path., 23, 429 (1947). a Bohuon, C., E. II. LaBrosse, M. Assicot, and A. Amar-Costesec, in Recent Results in Cancer Research, vol. 2 Neuroblastomu, Biochemical Studies, ed. C. Bohuon (New York: Springer-Verlag, 1966), p. 16. 7 Goldstein, M., B. Anagnoste, and M. N. Goldstein, Science, 160, 767 (1968). 1010 BIOCHEMISTRY: NELSON ET AL. Pnoc. N. A. S. 8 von Stud&z, W., Phmm. Rev., 18,645 (1966). * Dulbecco, R., and G. Freeman, ~kckqy, 8,396 (1959). 10 Araki, T., and T. Otani, J. Neurophysiol., 18,472 (1955). 11 Grundfest, H., Federdion Proc., 26, 1613 (1967). I* Chin, S. M., J. Con@. Neural., 104, 285 (1956). 1) Hild, W., and I. Tasaki, J. Newophysiol., 25, 277 (1962). 14 Scott, B. S., V. E. Engelbert, and K. C. Fisher, Ezp. Newol., 23,230 (1969). 16 Wilson, S., J. Farber, R. Rosenberg, T. Amano, P. Nirenberg, N. Seeds, and M. Nirenberg, unpubliihed data.