Neuroscience Letters, 99 (1989) 137 141 137

Elsevier Scientific Publishers Ireland Ltd.

NSL 06001

Repetitive burst-firing neurons in the deep layers of

mouse somatosensory cortex

A. A g m o n 1 a n d B.W. C o n n o r s 2
ZDepartment of Neurology, StanJbrd University School of Medicine, Stanford, CA 94305 (U.S.A.) and
:Section of Neurobiology, Division of Biology and Medicine, Brown University, Providence, R1 02912
(u.s.A.)
(Received 20 July 1988; Revised version received 14 December 1988; Accepted 14 December 1988)

Key words: Mouse; Somatosensory cortex; Bursting neuron; Action potential; Pyramidal neuron

Intracellular recordings were made from neurons of the mouse somatosensory cortex isolated in vitro.
Two physiologically distinct classes of pyramidal cells were observed: regular-spiking cells were the major-
ity, and generated accommodating trains of single spikes; bursting cells generated clusters of 2 5 action
potentials, and clusters were separated by prolonged afterhyperpolarizations. The bursting cells of the
mouse neocortex were unusual in producing repetitive bursting during sustained current pulses, and in
being localized to a laminar zone straddling layers V and VI.

The intrinsic firing properties of different neurons in the neocortex are not
uniform. For example, pyramidal cells can be distinguished from aspiny and sparsely
spiny cells by their action potential shapes and repetitive firing patterns [14, 19].
There are also distinctive physiological subclasses of pyramidal cells. Stow and fast
pyramidal tract neurons in the cat differ in both passive membrane properties and
action potentials [6, 16, 23]. Another striking example is the dichotomy between
'regular-spiking' neurons and intrinsically 'bursting' neurons of the guinea pig [9, 19].
In response to a step current stimulus just above threshold, regular-spiking cells
generate single action potentials. In contrast, at threshold bursting cells generate a
cluster of 2-5 spikes riding upon a slow depolarizing envelope. It is important to
stress that the intrinsic behavior of a neuron derives from its complement of voltage-
dependent membrane channels, and is independent of synaptic input.
Intrinsically bursting neurons in guinea pig neocortex have a specific laminar dis-
tribution. Whereas regular-spiking cells are found in all layers from II to VI, bursting
cells have so far been observed only in layers IV and the upper aspect of layer V [9,
19]. Very little is known about the incidence of bursting cells among various other
species and areas of neocortex. In this study we present observations on intrinsically

Correspondence: B.W. Connors, Section of Neurobiology, Box G, Division of Biology and Medicine,
Brown University, Providence, RI 02912, U.S.A.

0304-3940/89/$ 03.50 ~(-';1989 Elsevier Scientific Publishers Ireland Ltd.

138

bursting neurons in the primary somatosensory cortex o f the mouse. Unlike pre-
viously described bursting cells of guinea pig cortex, those of the mouse had a deeper
laminar position, and m a n y could generate repetitive bursts during a prolonged sti-
mulus.
The general methods for preparing, maintaining and recording from cortical slices
have been described [8, 19], however the present study employed several significant
modifications [1 3]. Mice (C57 BL/6) 2 5 weeks old were anesthetized with halo-
thane, decapitated and the brains removed into ice-cold physiological solution.
Slices, 400/tin thick, were cut on a vibratome, either in a coronal plane or in an ob-
lique plane o f section designed to preserve thalamocortical connectivity [1]. Under
transillumination layers IV (the barrels) and the region straddling layers V and VI
were visualized as dark bands [2]: the slice and its laminar borders were traced on
a camera lucida, and recordings were subsequently made in an interface-type
c h a m b e r kept at 32' C. Intracellular recordings were made with microelectrodes filled
with 4 M potassium acetate and bevelled to about 200 MQ. Cells were activated by
step pulses o f current. D a t a were acquired using a computer-based data acquisition
system. The laminar position of each cell was determined by measuring its fractional
position between the pia and the border o f the gray and white matter, and comparing
this measurement with the camera lucida drawing.
An example o f an intracellular recording from a cell that fits the criteria for regu-
lar-spiking neurons [19] is illustrated in Fig. IA. This cell fired trains of individual
action potentials whose frequency a c c o m m o d a t e d during prolonged current pulses.
Spike frequency increased smoothly with current intensity (not shown).

II/111 --'C--'--;_
j .... IV

V
o • • • e'''
B I Vl
i, II I
wm

S L ....] 20
0.3 nA
40 msec

Fig. 1. Different firing patterns of mouse neocortical neurons during injected current pulses. A: recording
from a representative regular-spiking neuron of layer V. B: recording from a repetitively bursting neuron
located near the border of layers V and VI. Calibration bars apply to both A and B. C: laminar location
of 33 neurons from 6 experiments, superimposed on a map traced from a photograph of a living slice. Site
of regular-spiking cells identified by dots, bursting cells by asterisks. The horizontal position of each sym-
bol is arbitrary. Boundaries of layer IV and the layer V/VI border marked with dashed lines.
 139

In contrast, many neurons exhibited repetitive burst-firing behavior. Fig. 1B shows

a recording from one such cell. During a positive current pulse 330 ms long, the cell
generated 3 bursts of 3-5 action potentials each. The mean burst frequency during
the pulse was about 9 Hz. Action potential bursts always rode upon a slow depolariz-
ing envelope of about 20 mV, and each burst was followed by a trough of relative
hyperpolarization. The mean resting potentials of regular-spiking ( - 7 2 mV) and
bursting cells ( - 74 mV) were not significantly different.
The laminar location of bursting neurons was not random. Fig. IC plots the
recording site of all presumed pyramidal neurons encountered during 6 experiments.
Dots represent regular-spiking cells, asterisks represent bursting cells. Of the 33 neu-
rons, 4 were classified as bursting cells; each of the 4 was located in the vicinity of
the border of layers V and VI. Out of a total sample of 92 presumed pyramidal cells
recorded in slices from 33 animals, 13 of 37 (35%) in layers V/VI were bursting cells
while none of the 45 cells in layers II/III and IV exhibited bursting properties as
defined here.
The pattern of bursting was stimulus-dependent and varied from cell-to-cell. Fig.
2A illustrates repetitive bursting at 2 different current intensities in another cell.
Mean burst frequency was 9 Hz at the low, and 16 Hz at the high intensity. Burst
size was 2 or 3 spikes, with smaller bursts predominating at higher currents. Some
neurons could not sustain repetitive bursting beyond 2 or 3 bursts, and reverted to
single-spike firing either following a bursting episode or during it (Fig. 2B). The indi-
vidual spikes in these cases were followed by prominent depolarizing afterpotentials
(Fig. 2B, arrows).
Two regular-spiking neurons of layers V/VI and one repetitive bursting neuron

40 msec

Fig. 2. Firing patterns of 2 different bursting neurons from layers V/VI. A: repetitivebursts generated
by step current stimuli of 0.44 nA (bottom) and 0.66 nA (top). B: bursts and single-spikefiringin response
to 0.50 nA (bottom) and 0.80 nA (top). Arrowheads identifyexamples of depolarizing afterpotentialsfol-
lowing single spikes. Calibrations apply to all traces.
140

were successfully filled with the fluorescent dye Lucifer yellow CH. All 3 had the clas-
sical morphological characteristics of neocortical pyramidal cells [10].
The repetitive bursting behavior of neurons in mouse neocortex is reminiscent of
the endogenously generated bursting activity of neurons in many invertebrate ganglia
(e.g. ref. 4) and some nuclei of the mammalian central nervous system (e.g. refs. 5,
15, 18, 24). The ionic mechanisms of repetitive bursting vary greatly among different
cell types, although many of the mammalian systems seem to depend in part upon
a low threshold calcium current to sustain the burst. Such a current apparently exists
in neocortical pyramidal cells [11], however the mechanism(s) of intrinsic bursting
in neocortex remain to be explored.
This study describes two functional classes of neurons in the mouse primary soma-
tosensory cortex: regular-spiking and bursting cells. A similar distinction was pre-
viously made for pyramidal cells in the neocortex of guinea pigs [9, 19]. However,
in the present work, bursting cells were different in 2 respects: a) they often fired
repetitive bursts during a sustained current stimulus, and b) they were found in the
zone on both sides of the border of layers V and VI. In contrast, the bursting cells
so far described in the guinea pig generate only one burst per step-current stimulus,
followed by single-spike firing, and their soma positions are in layer IV and upper
layer V. Whether these are true species differences or reflect different methodologies
remains to be explored.
While neocortical neurons from a variety of intracellular studies in vivo and in
vitro meet the criteria for the regular-spiking designation [6, 9, 16, 20, 22, 23], there
have been very few unambiguous reports of neocortical neurons with the ability to
generate intrinsic bursts. Following the initial descriptions in parietal and cingulate
neocortex of guinea pigs in vitro [9, 12, 19], bursting cells were observed in rat motor
cortex in vivo [17] and epileptic temporal neocortex of humans maintained in vitro
(D.A. McCormick, personal communication). These limited data, plus the present
results, suggest that intrinsically bursting neurons may be a common feature of dif-
ferent areas of neocortex in a variety of species from mouse to man.
The functions of intrinsically bursting cells in cerebral cortex are unknown. Several
lines of evidence indicate that they receive markedly less inhibitory input than regu-
lar-spiking cells, whether l¥om intracortical stimulation [7, 21] or from thalamus-
evoked inputs [I]. Agmon [1], using a novel slice preparation, has also shown that
intrinsically bursting cells receive much less monosynaptic input from the ventrobasal
nucleus of the thalamus than do regular-spiking cells. Bursting neurons have also
been implicated in the initiation of synchronized epileptiform discharges [7, 8, 12, 13],
and their behavior suggests that they are highly interconnected with each other.
Finally, their restricted laminar location in the infragranular layers suggests that all
bursting cells may project to the same subcortical target(s). Thus, the intrinsic mem-
brane properties of pyramidal neurons may correlate with both their excitatory and
inhibitory synaptic connections. Further elucidation of bursting cell morphology and
connections, both intrinsic and extrinsic, should help to illuminate their functions.
 141

We thank M.J. Gutnick for helpful comments on the manuscript. These studies
were supported by MH17047, NS 01271, NS 25983 and the Klingenstein Fund.

1 Agrnon, A., Intrinsic Properties and Synaptic Connectivity of Mouse Barrel Cortex Neurons: Correla-
tion between Firing Patterns and Thalamocortical Inputs, Ph.D. dissertation, Stanford University,
1988, 173 pp.
2 Agmon, A. and Connors, B.W., Visualization of the barrel field in living rat neocortical slices, Soc.
Neurosci. Abstr., 10 (1984) 493.
3 Agmon, A. and Connors, B.W., The thalamocortical and transcallosal barrel field slice preparations:
preserving extrinsic connections in vitro, Soc. Neurosci. Abstr., 11 (1985) 750.
4 Alving, B.O., Spontaneous activity in isolated somata of Aplysia pacemaker neurons, J. Gen. Physiol.,
51 (1968) 29-45.
5 Andrew, R.D. and Dudek, F.E., Burst discharge in mammalian neuroendocrine cells involves an
intrinsic regenerative mechanism, Science, 221 (1983) 1050-1052.
6 Calvin, W.H. and Sypert, G.W., Fast and slow pyramidal tract neurons: an intracellular analysis of
their contrasting repetitive firing properties in the cat, J. Neurophysiol., 39 (1976) 420~34.
7 Chagnac-Amitai, Y. and Connors, B.W., Intrinsic excitability and synaptic connectivity of neocortical
pyramidal neurons are correlated: implications for epileptogenesis, Epilepsia, 29 (1988) 7 I0.
8 Connors, B.W., Initiation of synchronized neuronal bursting in neocortex, Nature (Lond.), 310 (1984)
685 687.
9 Connors, B.W., Gutnick, M.J. and Prince, D.A., Electrophysiological properties of neocortical neu-
rons in vitro, J. Neurophysiol., 48 (1982) 1302 1320.
10 Feldman, M.L., Morphology of the neocortical pyramidal neuron. In A. Peters and E.G. Jones (Eds.),
Cerebral Cortex, Vol. 1, Cellular Components of the Cerebral Cortex, Plenum, New York, 1984, pp.
123-200.
l 1 Friedman, A. and Gutnick, M.J., Low-threshold calcium electrogenesis in neocortical neurons, Neu-
rosci. Lett., 81 (1987) 117 122.
12 Gutnick, M.J. and Friedman, A., Synaptic and intrinsic mechanisms of synchronization and epilepto-
genesis in the neocortex. In U. Heinemann, M. Klee, E. Neher and W. Singer (Eds.), Calcium Electro-
genesis and Neuronal Functioning, Exp. Brain Res. Ser., Vol. 14, 1986, pp. 327-335.
13 Gutnick, M.J., Connors, B.W. and Prince, D.A., Mechanisms of neocortical epileptogenesis in vitro,
J. Neurophysiol., 48 (1982) 1321-1335.
14 Huettner, J.E. and Baughman, R.W., The pharmacology of synapses formed by identified corticocolli-
cular neurons in primary cultures of rat visual cortex, J. Neurosci., 8 (1988) 160-175.
15 Jahnsen, H. and Llinas, R., Electrophysiological properties of guinea-pig thalamic neurons: an in vitro
study. J. Physiol. (Lond.), 349 (1984) 205-226.
16 Koike, H., Okada, Y., Oshima, T. and Takahashi, K., Accommodative behavior of cat pyramidal tract
cells investigated with intracellular injection of currents, Exp. Brain Res., 5 (1968) 173-188.
17 Landry, P., Wilson, C.J. and Kitai, S.T., Morphological and electrophysiological characteristics of py-
ramidal tract neurons in the rat, Exp. Brain Res., 57 (1984) 177-190.
18 Llinas, R. and Yarom, Y., Electrophysiology of mammalian inferior olivary neurones in vitro. Differ-
ent types of voltage-dependent ionic conductances. J. Physiol. (Lond.), 315 (1981) 549-567.
19 McCormick, D.A., Connors, B.W., Lighthall, J.W. and Prince, D.A., Comparative physiology of pyra-
midal and sparsely spiny neurons of the neocortex, J. Neurophysiol., 54 (1985) 782-806.
20 Ogawa, T., Ito, S. and Kato, H., Membrane characteristics of visual cortical neurons in in vitro slices,
Brain Res., 226 (1981) 315-319,
21 Silva, L.R., Chagnac-Amitai, Y. and Connors, B.W., Inhibitory synaptic input correlates with intrinsic
membrane properties in neocortical pyramidal cells, Soc. Neurosci. Abstr., 14 (1988) 883.
22 Stafstrom, C.E., Schwindt, P.C. and Crill, W.E., Repetitive firing in layer V neurons from the cat neo-
cortex in vitro, J. Neurophysiol., 52 (1984) 264-289.
23 Takahashi, K., Slow and fast groups of pyramidal tract cells and their respective membrane properties,
J. Neurophysiol., 28 (1965) 908-924.
24 Wilcox, K.S., Gutnick, M.J. and Christoph, G.R., Electrophysiological properties of neurons in the
lateral habenula nucleus: an in vitro study, J. Neurophysiol., 59 (1988) 212 225.