Lateral Turns in the Lamprey. II. Activity of Reticulospinal
Neurons During the Generation of Fictive Turns
PATRIQ FAGERSTEDT, GRIGORI N. ORLOVSKY, TATIANA G. DELIAGINA, STEN GRILLNER, AND
FREDRIK ULLÉN
Nobel Institute for Neurophysiology, Department of Neuroscience, Karolinska Institutet, SE-171 77 Stockholm, Sweden
Received 21 June 2000; accepted in final form 11 June 2001
Fagerstedt, Patriq, Grigori N. Orlovsky, Tatiana G. Deliagina,
Sten Grillner, and Fredrik Ullén. Lateral turns in the lamprey. II.
Activity of reticulospinal neurons during the generation of fictive
turns. J Neurophysiol 86: 2257–2265, 2001. We studied the neural
correlates of turning movements during fictive locomotion in a lamprey in vitro brain–spinal cord preparation. Electrical stimulation of
the skin on one side of the head was used to evoke fictive turns.
Intracellular recordings were performed from reticulospinal cells in
the middle (MRRN) and posterior (PRRN) rhombencephalic reticular
nuclei, and from Mauthner cells, to characterize the pattern of activity
in these cell groups, and their possible functional role for the generation of turns. All recorded reticulospinal neurons modified their
activity during turns. Many cells in both the rostral and the caudal
MRRN, and Mauthner cells, were strongly excited during turning. The
level of activity of cells in rostral PRRN was lower, while the lowest
degree of activation was found in cells in caudal PRRN, suggesting
that MRRN may play a more important role for the generation of
turning behavior. The sign of the response (i.e., excitation or inhibition) to skin stimulation of a neuron during turns toward (ipsilateral),
or away from (contralateral) the side of the cell body was always the
same. The cells could thus be divided into four types: 1) cells that
were excited during ipsilateral turns and inhibited during contralateral
turns; these cells provide an asymmetric excitatory bias to spinal
networks and presumably play an important role for the generation of
turns; these cells were common (n ⫽ 35; 52%) in both MRRN and
PRRN; 2) cells that were excited during turns in either direction; these
cells were common (n ⫽ 19; 28%), in particular in MRRN; they could
be involved in a general activation of the locomotor system after skin
stimulation; some of the cells were also more activated during turns in
one direction and could contribute to an asymmetric turn command; 3)
one cell that was inhibited during ipsilateral turns and excited during
contralateral turns; and 4) cells (n ⫽ 12; 18%) that were inhibited
during turns in either direction. In summary, our results show that, in
the lamprey, the large majority of reticulospinal cells have responses
during lateral turns that are indicative of a causal role for these cells
in turn generation. This also suggests a considerable overlap between
the command system for lateral turns evoked by skin stimulation,
which was studied here, and other reticulospinal command systems,
e.g., for lateral turns evoked by other types of stimuli, initiation of
locomotion, and turns in the vertical planes.
In all classes of vertebrates, the spinal cord contains neuronal networks that can generate the basic motor pattern of
locomotion. Descending supraspinal systems activate these
networks and modulate their activity, to produce alterations in
speed, steering maneuvers, postural corrections, and other adaptations of the locomotor movements to the surrounding and
the goals of the organism (Grillner 1985; Orlovsky et al. 1999).
Here we study the functional organization of the descending
control system for lateral turning movements in the lamprey, a
lower vertebrate belonging to the cyclostomes. Lampreys perform lateral turns spontaneously (McClellan and Hagevik
1997; Ullén et al. 1997), but they can also be evoked by various
types of sensory stimuli, e.g., unilateral eye illumination (negative phototaxis) (Ullén et al. 1995, 1997), illumination of tail
skin photoreceptors (Deliagina et al. 1995; Harden-Jones 1955;
Young 1935), mechanical or electrical skin stimulation (McClellan 1984; McClellan and Hagevik 1997), and olfactory
stimulation (Kleerekoper 1972; Kleerekoper and Mogensen
1963; Teeter 1980). Neural correlates of turning movements
(fictive turns) can be evoked in vitro by electrical stimulation
of the skin of the head (Fagerstedt and Ullén 2001; McClellan
and Hagevik 1997). A detailed analysis of the motorneuron
activity during turns is presented in the accompanying paper
(Fagerstedt and Ullén 2001).
Higher vertebrates have a number of descending systems
that can influence spinal locomotor networks. However, in the
lamprey, supraspinal commands are mediated essentially by
reticulospinal (RS) and vestibulospinal pathways. The RS system contains the largest number of cells (Bussières 1994) and
is also the only system that reaches the middle and caudal parts
of the spinal cord. RS neurons are located in four nuclei: the
mesencephalic reticular nucleus, and the anterior, middle, and
posterior rhombencephalic reticular nuclei (MRN, ARRN,
MRRN, and PRRN, respectively; Fig. 1A). They project
mainly to the ipsilateral spinal cord. Mauthner cells and accessory Mauthner cells are situated lateral to the MRRN and
project to the contralateral spinal cord (Brodin et al. 1988;
Bussières 1994; Bussières et al. 1999; Ronan 1989; Rovainen
1983; Swain et al. 1993).
Recordings of the mass activity in the left and right RS
pathways in intact, freely behaving lampreys have shown that,
during lateral turns, the RS activity on the side toward which
the animal turns is stronger than that on the opposite side
(Deliagina et al. 2000). However, it has remained unclear
Present address and address for reprint requests: F. Ullén, Neuropediatrics,
Dept. of Woman and Child Health, Astrid Lindgren Children’s Hospital,
SE-171 76 Stockholm, Sweden (E-mail: Fredrik.Ullen@neuro.ki.se).
The costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked ‘‘advertisement’’
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
INTRODUCTION
www.jn.org
0022-3077/01 $5.00 Copyright © 2001 The American Physiological Society
2257
2258
FAGERSTEDT, ORLOVSKY, DELIAGINA, GRILLNER, AND ULLÉN
FIG. 1. A: the brain stem of the lamprey (view from
above) with the mesencephalic reticular nucleus, and the
anterior, middle, and posterior rhombencephalic reticular
nuclei (MRN, ARRN, MRRN, and PRRN, respectively)
indicated. B: experimental arrangement. The in vitro preparation, consisting of the brain and spinal cord attached to
the underlying cranium and notochord as well as skin of the
rostral part of the head, was positioned in a chamber divided into two pools by an agar barrier. The rostral pool
contained the brain and was perfused with normal Ringer
solution. The caudal pool contained the spinal cord and was
perfused with a Ringer solution containing D-glutamate to
elicit fictive locomotion. The skin of the head was stimulated by two bipolar silver wire stimulation electrodes, one
on each side. Fictive locomotion and fictive turning movements evoked by the skin stimulation were recorded by two
pairs of left and right extracellular glass pipette electrodes
positioned on one rostral and one more caudal pair of
ventral roots (VR). Responses in reticulospinal (RS) cells in
MRRN and PRRN during fictive turns were recorded intracellularly with a microelectrode. One pair of left and
right caudal extracellular glass pipette electrodes was positioned on the spinal cord surface to allow recordings of
RS action potentials and antidromic stimulation of intracellularly recorded neurons.
which groups of RS neurons are responsible for this asymmetry
in descending influences. In the present study, we used an in
vitro preparation of the isolated lamprey brain and spinal cord
and performed intracellular recordings from RS cells in the
MRRN and the PRRN and from Mauthner cells, to elucidate
the possible roles of different groups of RS neurons for eliciting lateral turns.
Preliminary accounts of parts of this study have appeared in
abstract form (Fagerstedt et al. 1998; Ullén et al. 1998).
METHODS
Experiments were performed on 14 adult (15–30 cm) North American
silver lampreys, Ichthyomyzon unicuspis. The preparation and surgery
were described in detail in the accompanying paper (Fagerstedt and Ullén
2001). In brief, the preparation consisted of the head with exposed brain
and rostral part of the spinal cord (20 –50 segments) and was mounted in
an experimental chamber, divided by a barrier into two pools, which were
filled with Ringer solution (Fig. 1B).
Fictive lateral turns were evoked by electrical stimulation of the
skin on one side of the head with bipolar silver wire electrodes (0.5
mm diam). The activity of individual RS neurons in the two largest
reticular nuclei, MRRN and PRRN, and Mauthner cells were recorded
intracellularly with sharp glass micropipettes filled with 4 M potassium acetate and 0.1 M potassium chloride (Fig. 1, A and B). The
activity of motoneurons was recorded extracellularly with two pairs of
glass pipette electrodes (20 –30 ␮m tip diameter) positioned over the
ventral roots at the level of segments 11–19 and segments 18 –33,
respectively, so that the distance between the electrode pairs was
around 10 segments (Fig. 1B). Data on ventral root activity at different
rostrocaudal levels during turns are presented in the accompanying
paper (Fagerstedt and Ullén 2001); here, only data from the rostral
ventral root electrodes will be included. Two additional glass pipette
electrodes (tip diameter 50 –75 ␮m) were positioned on the left and
right dorsal surface of the spinal cord, close to the caudal end of the
preparation (segment 22–38; Fig. 1B). They were used to determine to
which side of the spinal cord each recorded RS neuron projected and
its conduction velocity. This was done either by stimulating the RS
neuron intracellularly, and recording the arrival of the orthodromically
propagated action potentials in the spinal cord, or by stimulating the
J Neurophysiol • VOL
axon of the RS neuron extracellularly, and recording the antidromically propagated action potentials in the cell body.
Fictive locomotion, characterized by rhythmical bursts of motoneuron
activity alternating between the left and the right side was evoked by
administering D-glutamate (0.5–2 mM) to the spinal cord pool. Fictive
turns were evoked by electrical skin stimulation (250- to 1,500-ms trains
of 2-ms pulses at 10 –50 Hz) and were characterized by a transient
increase of the burst duration and intensity on the turning side, and of the
cycle duration (see RESULTS) (Fagerstedt and Ullén 2001). Turns toward
the side of the recorded cell were classified as ipsilateral, whereas turns
in the opposite direction were classified as contralateral. In 18 cells, the
effect of high-frequency intracellular stimulation (10 –50 Hz for 10 –30 s)
on the locomotor rhythm was also investigated.
Data acquisition and analysis of the ventral root activity was described
in the accompanying paper (Fagerstedt and Ullén 2001). In summary, a
locomotor cycle was defined as starting with the onset of a burst of
ventral root activity on the side of the skin stimulus, and ending with the
onset of the successive burst on the same side; burst duration was defined
as the time between onset and termination of the same locomotor burst;
and burst intensity as the area of the rectified burst (burst amplitude)
divided with its burst duration. The activity of RS cells during ipsilateral
and contralateral fictive turns was analyzed separately. For neurons that
fired action potentials during the turn, the number of action potentials, the
firing frequency, and the total duration of the burst were measured. For
two activated cells, where a larger number of turns with varying amplitude were recorded, a correlation analysis between turn amplitude in the
rostral segments and cellular response was performed. Turn amplitude
was determined either by looking at changes in burst intensity or in cycle
duration, since these two parameters to some extent vary independently
(Fagerstedt and Ullén 2001). The effects of intracellular stimulation on
the locomotor rhythm were evaluated by comparing the mean burst
intensity, burst duration, and cycle duration during the period of stimulation with the mean values of the same parameters during a control
period of approximately equal duration (t-test), immediately preceding
the stimulation.
RESULTS
Response patterns in individual RS cells
Altogether, 72 RS neurons were recorded intracellularly in
MRRN (n ⫽ 42) and PRRN (n ⫽ 30) in 10 preparations. Their
86 • NOVEMBER 2001 •
www.jn.org
RETICULOSPINAL ACTIVITY DURING FICTIVE TURNS
resting membrane potentials ranged between ⫺58 and ⫺80
mV. In around 50% of the cells, a spinal projection of the
neuron could be verified (see METHODS). The main reason that
a spinal projection could not be directly demonstrated in some
cells is most likely that the axon terminated rostral to the spinal
surface electrodes; earlier studies have shown that practically
all larger reticular cells do project to the spinal cord (Wannier
1994). For simplicity, the term “RS neuron” will therefore be
used for all cells recorded in the reticular nuclei. The recorded
RS neurons practically never fired action potentials spontaneously. Rhythmic modulation of the membrane potential in
phase with ipsilateral fictive locomotor activity was sometimes
seen, but the amplitude was low. All neurons were tested
during 2–10 ipsilateral turns and 2–10 contralateral turns (see
METHODS), with the exception of two cells in PRRN and six
cells in MRRN, which could only be tested for turns in one
direction.
As illustrated in Figs. 2 and 3, individual RS cells showed
different characteristic patterns of responses during ipsilateral
and contralateral turns. Excitatory responses consisted either of
a subthreshold wave of depolarization or a depolarization with
action potentials. During ipsilateral turns, most RS neurons
(n ⫽ 53) received an excitatory input. Of these, 26 cells were
depolarized and fired a burst of action potentials (Fig. 2A),
FIG. 2. A and B: typical responses of a cell from MRRN that was activated
during both ipsilateral and contralateral turns (TT3 cell; see text). Ventral root
recordings are from segment 15. A: response during ipsilateral turn. B: response during contralateral turn. This particular cell was more excited during
ipsilateral turns and could contribute to the asymmetric descending turn
command, but other TT3 cells showed the opposite tendency (see text). C and
D: typical responses of a cell from PRRN that was depolarized during ipsilateral turns and hyperpolarized during contralateral turns (TT1 cell; see text).
Rostral ventral root recordings are from segment 15, caudal ones from segment
30. C: response during ipsilateral turn. D: response during contralateral turn.
This particular cell received subthreshold excitatory postsynaptic potentials
(EPSPs) during the recorded ipsilateral turns. TT1 cells that are activated
during ipsilateral turns are presumably important for the generation of the
descending turn command (see text).
J Neurophysiol • VOL
2259
while 25 neurons responded only with subthreshold depolarizations during the turn (Fig. 2C). Two cells responded with
subthreshold depolarizations during some turns but were recruited and fired action potentials during other turns (see Fig.
6, A and B). The remaining cells (n ⫽ 13) were inhibited (not
illustrated). Most RS neurons (n ⫽ 50) received an inhibitory
input (Fig. 2D) during contralateral turns. The remaining cells
(n ⫽ 20) received an excitatory input. Of the latter cells, six
responded with subthreshold waves of depolarization (not illustrated), 13 responded with action potentials (Fig. 2B), while
one cell responded with action potentials during some turns
and with subthreshold depolarizations during the others (not
illustrated).
These results are summarized in Fig. 3, for the neurons
recorded from MRRN and PRRN and also separately for the
neurons from rostral and caudal subdivisions of these nuclei.
The large majority of cells in MRRN were excited during
ipsilateral turns (n ⫽ 33), and most of these (n ⫽ 23) also fired
action potentials. Only five cells were inhibited (Fig. 3A).
During contralateral turns, most MRRN cells (n ⫽ 24) were
inhibited while the remaining cells (n ⫽ 17) were depolarized;
13 cells fired action potentials (Fig. 3A). No differences in the
response characteristics were found between cells in rostral
(n ⫽ 26; Fig. 3B) and caudal MRRN (n ⫽ 13; Fig. 3C). For
three cells the location in MRRN was not noted.
Most RS neurons in PRRN were also excited during ipsilateral turns (n ⫽ 22), but only five of these fired action potentials. The remaining eight cells were inhibited (Fig. 2D). During contralateral turns, almost all (n ⫽ 26) PRRN cells were
inhibited. Only four cells were depolarized, one of which fired
action potentials (Fig. 3D). Cells in rostral PRRN (Fig. 3E)
were more activated than cells in caudal PRRN (Fig. 3F)
during turns in either direction. In the rostral PRRN, five cells
fired action potentials during ipsilateral turns and one cell
during contralateral turns (Fig. 3E), while none of the cells in
caudal PRRN were activated during turning responses evoked
by stimulation from either side (Fig. 3F).
Since several cells received subthreshold excitation during
some turns and fired action potentials during others, these two
responses will not be considered as distinct. In this way, four
main types of cell responses during turns evoked by trigeminal
skin stimulation could be distinguished. Cells with corresponding reponse patterns will be labeled TT1, TT2, TT3,
and TT4 cells, respectively: 1) TT1 cells were excited during
ipsilateral turns and inhibited during contralateral turns
(Fig. 2, C and D); 2) TT2 cells were inhibited during
ipsilateral turns and excited during contralateral turns; 3)
TT3 cells were excited during turns in either direction (Figs.
2, A and B, and 6); and 4) TT4 cells, finally, were inhibited
during turns in either direction.
Tables 1 and 2 show the response patterns of all recorded
cells in MRRN and PRRN. In MRRN, TT3 cells and TT1
cells dominated. Ten of the total 15 TT3 cells in this nucleus
fired action potentials during turns in either direction (Fig.
2, A and B). Of the 17 TT1 cells, nine were activated during
ipsilateral turns (Table 1). The remaining cells consisted of
four TT4 cells and one single cell with a TT2 response
(Table 1). TT1 cells were the most common cell type also in
PRRN (n ⫽ 18; Table 2). Only four of these cells were
activated, however; the remaining 14 cells received subthreshold depolarization during ipsilateral turns (Fig. 2, C
86 • NOVEMBER 2001 •
www.jn.org
2260
FAGERSTEDT, ORLOVSKY, DELIAGINA, GRILLNER, AND ULLÉN
FIG. 3. Summary of responses in rostral and caudal halves of MRRN and PRRN during fictive turns. Cells were either
hyperpolarized or depolarized during turns in a certain direction. Depolarized cells were divided into cells that only received
subthreshold EPSPs (■) and cells that fired action potentials (s). A–C: responses in all MRRN cells (A), and in cells from the rostral
(B) and caudal (C) halves of MRRN, displayed separately. The response patterns in rostral and caudal parts of MRRN were highly
similar. D–F: responses in all (D) PRRN cells, and in cells from the rostral (E) and caudal (F) halves of the nucleus. Cells in caudal
PRRN were less activated than cells in rostral PRRN.
and D; Table 2). TT4 cells constituted the second most
common cell type in PRRN (n ⫽ 8; Table 2). Only four TT3
cells and no TT2 cell were found in this nucleus. Two
Mauthner cells were recorded. Both of these were TT3 cells
and were activated during turns in either direction with
relatively high firing frequency (see Fig. 4).
Response characteristics of cells that were activated
during fictive turns
Response characteristics of individual cells that fired action
potentials during turns are summarized in Fig. 4. Of all 23
MRRN cells that were activated during ipsilateral turns (TT3
and TT1 cells; see Table 1 and above), 17 were selected for
quantitative analysis, while the remaining cells were not included because of relatively unstable recordings or few re1. Responses of individual MRRN neurons during turns
toward (ipsilateral) or away from (contralateral) the side
of the cell body
corded turns. The mean number of spikes per turn for these
MRRN cells showed a wide distribution with a total mean of
6.7 (Fig. 4A). The mean spike train durations and firing frequencies likewise varied within a wide range with total means
of 0.85 s (Fig. 4B) and 10.4 Hz (Fig. 4C), respectively.
During contralateral turns 13 MRRN cells were activated
(TT3 and TT2 cells; see Table 1 and above), of which nine were
selected for quantitative analysis. The magnitude of these
responses did not differ significantly from those seen during
ipsilateral turns, however (see DISCUSSION). The total mean
number of spikes per turn was 6.5 (Fig. 4D). The spike train
durations had a total mean of 0.85 s (Fig. 4E), and the firing
frequencies had a total mean of 13.0 Hz (Fig. 4F). The latter
two parameters could only be calculated for cells that fired
more than one spike per turn (n ⫽ 7).
2. Responses of individual PRRN neurons during turns
toward (ipsilateral), or away from (contralateral) the side
of the cell body
TABLE
TABLE
Ipsilateral Turns
Ipsilateral Turns
Excitation
Contralateral
Turns
Inhibition
Excitation
EPSPs
Spikes
Excitation
Inhibition
EPSPs
Spikes
4*
8†
9†
1†
0†
2‡
1‡
2‡
10‡
The number of recorded cells with one particular response pattern is indicated in each cell of the table. * Cells that were inhibited with turns in either
direction. † Cells that were excited by turns in one direction, and inhibited by
turns in the other direction. ‡ Cells that were excited by turns in any direction.
J Neurophysiol • VOL
Contralateral
Turns
Inhibition
Excitation
EPSPs
Spikes
Inhibition
EPSPs
Spikes
8*
14†
4†
0†
0†
3‡
1‡
0‡
0‡
The number of recorded cells with one particular response pattern is indicated in each cell of the table. * Cells that were inhibited with turns in either
direction. † Cells that were excited by turns in one direction, and inhibited by
turns in the other direction. ‡ Cells that were excited by turns in any direction.
86 • NOVEMBER 2001 •
www.jn.org
RETICULOSPINAL ACTIVITY DURING FICTIVE TURNS
2261
FIG. 4. Response characteristics of all individual activated cells. For mean values, see text. A–C: responses during ipsilateral
turns. Total number of spikes (A), spike train duration (B), and mean spike frequency (C) for PRRN, MRRN and Mauthner (Mth)
cells. The stronger activation of MRRN and Mth cells than PRRN cells was reflected in all studied response parameters (see text).
D–F: responses during contralateral turns. Only one PRRN cell was activated during turns in the contralateral direction. Total
number of spikes (D), spike train duration (E), and mean spike frequency (F) were higher for MRRN and Mauthner cells.
Only four PRRN cells were activated during ipsilateral
turns (TT1 cells; see Table 2). For all parameters, the
responses of these cells were considerably weaker than for
the MRRN cells. The total mean number of spikes per turn
was 2.0 (Fig. 4A), while the total mean spike train duration
was 0.27 s (Fig. 4B), and the total mean firing frequency was
5.2 Hz (Fig. 4C). A single PRRN cell (TT3 cell; see Table
2) fired action potentials during contralateral turns. The
mean number of spikes fired by this cell was 1.8, with a
mean train duration of 0.30 s and a mean firing frequency of
3.0 Hz.
The two recorded Mauthner neurons were strongly activated
during turns in both directions. During ipsilateral turns, the
total mean number of spikes was 16.2, with a mean train
duration of 0.91 s and a mean firing frequency of 50.3 Hz (Fig.
4, A–C). During contralateral turns the values of the same
parameters were 11.7 spikes, 0.80 s and 15.5 Hz, respectively
(Fig. 4, D–F).
Responses during ipsilateral and contralateral turns were
compared cell by cell in eight of the 10 MRRN TT3 cells
that were activated during turns in both directions. Notably,
these cells did not show similar changes of response parameters with change in turn direction (see DISCUSSION). Four
cells fired less spikes per turn during contralateral than
ipsilateral turns, while three cells fired more spikes during
contralateral turns and one cell had the same number of
spikes (Fig. 5A). Both the mean spike train duration and the
mean firing frequency were decreased during contralateral
turns in about one-half of the cells and increased in the
remaining cells (Fig. 5, B and C).
J Neurophysiol • VOL
Correlation between turn amplitude and activity
of RS neurons
We did not systematically try to evoke turns with large
differences in amplitude, to study the correlation between the
turn amplitude and the activity of RS neurons. Some observations could be made, however, which indicate that the generation of large turns is accompanied by both a recruitment of
more RS cells and an increased response in already active cells.
Some neurons (two in PRRN, one in MRRN) were depolarized below threshold for action potentials in weaker turns but
were activated during stronger turns, as illustrated in Fig. 6, A
and B. In other neurons, an increase of turn amplitude was
accompanied by an increase of the burst duration and frequency. Figure 6, C and D, shows an example of a MRRN TT3
cell that was activated during a weaker turn (Fig. 6C) and
increased both number of spikes, train duration, and spiking
frequency during a stronger turn (Fig. 6D).
An analysis of the correlation between the cell response and
the turn amplitude was done for the two recorded Mauthner
cells, where a larger number of turns (n ⫽ 21) could be
examined. This revealed strong positive correlations between
parameters of the cell response and turn amplitude on the
contralateral side. The number of spikes showed the strongest
correlation with increases in cycle duration during contralateral
turns (k ⫽ 0.90), as did spike train duration (k ⫽ 0.92). Spike
frequency showed the strongest correlation with increase in
burst intensity during contralateral turns (k ⫽ 0.55). Correlations with ipsilateral turn amplitude were low (⫺0.3 ⬍ k ⬍
0.3). One can note that the Mauthner cell is a contralaterally
projecting neuron (see DISCUSSION).
86 • NOVEMBER 2001 •
www.jn.org
2262
FAGERSTEDT, ORLOVSKY, DELIAGINA, GRILLNER, AND ULLÉN
FIG. 5. Responses during ipsilateral and contralateral turns in all that were activated during turns in either direction (TT3 cells;
see text). Mean values of the different responses are shown for individual cells, normalized to the mean response during ipsilateral
turns. A–C: the total number of spikes fired by each cell during ipsilateral and contralateral turns (A) as well as the spike train
durations (B) and the firing frequencies (C) are shown. Some cells were more excited during ipsilateral turns than during
contralateral turns, whereas other cells showed the opposite pattern or had similar responses during turns in any direction (see text).
The cell with a spike train duration and firing frequency of 0 during contralateral turns fired only one action potential.
Intracellular stimulation of RS cells
Intracellular stimulation (10 –50 Hz; see METHODS) of the
recorded RS cells were performed in 18 cells (10 MRRN cells,
seven PRRN cells, and one Mauthner cell). Significant effects
on the locomotor rhythm were seen in 12 cases; both increases
and decreases of the locomotor frequency were observed.
Effects on burst intensity were predominantly excitatory on the
ipsilateral side and inhibitory on the contralateral side of the
spinal cord, but different cells influenced rostral and caudal
ventral roots differently. Figure 7 shows an example of a TT3
FIG. 6. Recruitment of new RS cells, and increased intensity of firing of
already active cells with increased turn amplitude. Ventral root recordings are
from segment 15. A and B: example of a TT3 cell from PRRN that received
subthreshold EPSPs during weak ipsilateral turns (A), but was activated during
strong ipsilateral turns (B). C and D: example of a TT3 cell from MRRN that
was weakly activated during small ipsilateral turns (C), but responded more
strongly during large turns in the same direction (D).
J Neurophysiol • VOL
cell from MRRN that evoked a strong response, similar to a
fictive turn, when stimulated at 20 Hz. The locomotor frequency was decreased, and the burst intensity was increased on
the ipsilateral side and decreased on the contralateral side of
the spinal cord.
DISCUSSION
General properties of the descending command
for lateral turns
During lateral turns in the swimming lamprey, a mechanical
wave with larger amplitude than the normal locomotor waves
is propagated rostrocaudally along the body, so that the orientation of the body is shifted toward the new axis of swimming,
initially in the rostral part, then at progressively more caudal
levels. The pattern of muscle activity during turning is characterized by an increase in EMG burst amplitude, duration, and
proportion (burst duration divided by cycle duration) on the
side toward which the animal is turning, as well as by an
increase in cycle duration. An increase in burst amplitude on
the contralateral side is commonly seen after the initial turn
cycle (Fagerstedt and Ullén 2001; McClellan 1984; McClellan
and Hagevik 1997; Ullén et al. 1993). All these changes in the
locomotor rhythm can be seen in the in vitro preparation of the
isolated brain and spinal cord, in which fictive lateral turns are
evoked by skin stimulation (Fagerstedt and Ullén 2001; McClellan and Hagevik 1997). The basic components of the turn
motor pattern are thus centrally generated.
In this study, the focus of interest is on the descending brain
stem commands causing lateral turns. Fibers reaching the spinal cord in lamprey originate in the reticular nuclei (Brodin et
al. 1988; Ronan 1989; Rovainen 1983; Swain et al. 1993), the
vestibular nuclei (Bussières 1994; Bussières et al. 1999;
Rovainen 1979), tectum, pretectum, and scattered cell groups
in diencephalon (Bussières 1994). Of these, the reticulospinal
and vestibulospinal systems are by far the largest, containing at
least around 2,400 and 400 cells, respectively, while only a few
fibers from the other systems reach the spinal cord (Bussières
1994; Swain et al. 1993). Furthermore, fictive turns can be
evoked in rhombencephalic preparations (Fagerstedt and Ullén
2001), i.e., reticulospinal cells in the MRRN and the PRRN
86 • NOVEMBER 2001 •
www.jn.org
RETICULOSPINAL ACTIVITY DURING FICTIVE TURNS
2263
FIG. 7. Example of turning-like changes of the locomotor
pattern evoked by intracellular stimulation of a single neuron in
MRRN, at 20 Hz. During the stimulation, the cycle duration
increased and the burst intensity, duration, and proportion increased on the stimulated side.
together with the vestibulospinal system are sufficient to evoke
lateral turns. Vestibular fibers reach only the most rostral part
of the spinal cord (Rovainen 1979). Recent experiments have
also shown that vestibulospinal influences on spinal motorneurons are 3–5 times weaker than reticulospinal influences on the
same segment, and that the strength of the vestibulospinal
influences tapers off rapidly in more caudal segments, to become negligible beyond segment 10 –12 (P. V. Zelenin, personal communication). For these reasons we focused our attention on neurons in the two large rhombencephalic reticular
nuclei, the MRRN and the PRRN.
Skin stimulation in quiescent animals can evoke an avoidance behavior that includes arousal, lateral turning, and initiation of locomotion (McClellan 1984; McClellan and Grillner
1983). Therefore the responses in RS neurons to the skin
stimulation employed in the present study most likely reflect
commands both for a general activation of the locomotor
system and for a lateral turn. Activation of the spinal locomotor
network is presumably caused by bilaterally symmetric commands transmitted via RS pathways, whereas lateral turns are
caused by asymmetric commands: recordings of mass activity
in RS pathways in intact lampreys have shown that initiation of
locomotion is preceded by a bilateral activation of the RS
system, while turning movements are accompanied by an increase of RS activity on the side toward which the animal turns
(Deliagina et al. 2000). Additional evidence for the importance
of ipsilateral commands for turning was obtained in lesion
experiments: a rostral hemisection of the spinal cord prevents
turning toward the lesioned side (Fagerstedt and Ullén 2001;
Ullén et al. 1997). Finally, mathematical modeling has shown
that, in many models of the lamprey spinal locomotor networks, turn-like alterations of ongoing locomotor activity can
be evoked by commands directed to one side of the spinal cord
(Kozlov et al. 2001; McClellan and Hagevik 1997; Ullén et al.
1998). RS cells that respond differentially to ipsilateral and
contralateral turns were therefore considered as candidates for
generating the turning command in the present study.
Role of different nuclear regions for the generation of turns
The largest proportion of cells that fired action potentials
during fictive turns was found in MRRN, followed by rostral
PRRN, and finally caudal PRRN, where few activated cells
were found. No difference in the level of activity was found
between the rostral and caudal halves of MRRN. The Mauthner
cells were strongly activated during turns in either direction
(see Fig. 4). These findings suggest that MRRN is an important
nucleus for the generation of lateral turns as a result of skin
stimuli. In this preparation, however, locomotion was not
evoked from the brain stem. The level of excitation of at least
J Neurophysiol • VOL
some RS cells was therefore somewhat lower than in vivo,
when swimming is initiated from the brain stem and accompanied by depolarizing plateaus and spiking activity in reticular
cells (Kasicki et al. 1989; Viana Di Prisco et al. 1997). Since
PRRN contains a large number of cells with an asymmetric
response during lateral turns, it is therefore possible that the
relative importance of PRRN for the generation of turn commands increases at higher levels of tonic RS activity.
Putative roles of different RS cell types for the generation
of turns
All RS cells responded with either excitation or inhibition
during turns in either direction; the sign of these responses was
always the same. RS neurons therefore naturally fall into four
groups with regard to their responses during lateral turns: TT1,
TT2, TT3, and TT4 cells (see RESULTS). The distribution of these
cell types in MRRN and PRRN reflected the overall pattern of
excitability discussed above. More excited cells (e.g., spiking
TT3 cells) occurred in MRRN, whereas TT4 cells were most
common in caudal PRRN (see Tables 1 and 2).
TT1 cells are likely to play an important role for the generation of lateral turns. These cells could provide an excitatory
bias to spinal networks involved in the ipsilateral turn generation, while they are inhibited during turns in the other direction. TT1 cells were common in both MRRN and PRRN, but
TT1 cells that actually fired action potentials during fictive
turns, and thus could contribute to the observed modulations of
the locomotor pattern, were found mainly in MRRN (see Table
1). Increasing the excitability in the brain stem could presumably recruit more TT1 cells in PRRN (see above). Only one
TT2 cell was recorded, in MRRN (Table 1), and whether it
projected to the ipsilateral or contralateral spinal cord was
unfortunately not determined.
TT3 cells were common in MRRN (Table 1). The activation
of these cells could represent a general arousal and activation
of the locomotor system. Some TT3 cells, however, were more
strongly activated during ipsilateral turns than during contralateral turns, and may in addition contribute to the asymmetric
turn command. In other TT3 cells, the opposite pattern was
found; such cells could also contribute to the asymmetric turn
command if they project contralaterally. Some TT3 cells, finally, showed no clear differences in their responses depending
on turn laterality. These cells are probably not directly involved in the specification of turn direction. If they play a role
for turn generation, they could possibly be involved in triggering the turning event, or influence the turn amplitude by
exciting spinal networks involved in turn generation.
TT4 cells were predominantly found in PRRN, and especially in the caudal half of this nucleus. Since these cells were
86 • NOVEMBER 2001 •
www.jn.org
2264
FAGERSTEDT, ORLOVSKY, DELIAGINA, GRILLNER, AND ULLÉN
not active during normal fictive locomotion, nor during fictive
turns, they appear unlikely to play a role for the control of the
fictive turns studied in these experiments. One possibility
would be that TT4 cells are involved in other types of behavior,
and that they get inhibited during lateral turns to avoid behavioral conflict.
The present data clearly indicate that the generation of large
turns involves both recruitment of additional RS neurons and
larger responses in cells that are active already during turns
with smaller amplitude. At the RS level, turn amplitude is thus
most likely represented in the activity of the whole population
of turn-related neurons, where the contribution of each individual cell will depend on its level of activity, as well as on the
strength and specificity of its connections to the segmental
networks in the spinal cord. One can note that several cells that
were activated during turns in the present study produced
turn-like changes of one or more parameters of the locomotor
activity when stimulated intracellularly at a high frequency,
supporting the notion that they are involved in the generation
of turns.
The increase in burst intensity during turns can presumably
in part be explained by direct excitation of motoneurons, while
effects on cycle duration require that interneurons of the locomotor pattern generator are affected. Cycle duration and burst
intensity can, at least to some extent, vary independently during turns (Fagerstedt and Ullén 2001). Preliminary data in the
present study indicate that, at least in some cells, changes in
burst intensity are mainly reflected in the firing frequency of
active RS command neurons, while changes in cycle duration
correlate best with the total duration of the train of action
potentials. In this way, an independent regulation of these two
parameters could be obtained by modulating different aspects
of the response in the same command neurons.
RS neurons participate in the generation of different types of
commands; one extreme would be that all RS neurons are
active to some degree during all types of commands; the other
extreme would be that different commands involve separate,
nonoverlapping subpopulations of RS cells. In this regard, it is
striking that all RS neurons recorded in the present study
responded during lateral turns, albeit in some cases below the
threshold for the generation of action potentials. This suggests
a considerable overlap between the command system for lateral
turns and other RS command systems. Lateral turns could, e.g.,
involve an asymmetric activation of the same ipsilateral and
contralateral RS neurons that are involved in initiation and
maintenance of locomotion. If this were the case, unilateral
activation of this pool of neurons should give an increase in
cycle duration, while a bilateral activation of the same cells
decreases the cycle duration. Observations on rostrally hemisected animals in vivo suggest that at least some neurons
involved in lateral turns may be separate from the neurons
initiating locomotion, however; these animals swim along
straight lines, although the descending activity is completely
unilateral, and perform normal spontaneous lateral turns toward the intact side, with the same frequency as intact animals
(Ullén et al. 1997).
Relation of the lateral turn command system to other RS
command systems
REFERENCES
In this paper we have discussed the functional organization
of a descending command system for lateral turns evoked by
skin stimulation. The system for lateral turning is an integral
part of the steering system. Lateral turns can be evoked not
only by skin stimuli, but also by stimuli of a number of other
modalities (visual, olfactory, lateral line; see INTRODUCTION) and
may also be performed spontaneously. To what extent the
control systems for lateral turns, performed in different contexts, share common mechanisms at the RS level has not been
investigated. Visually evoked turns (negative phototaxis) have
been shown to be accompanied by the same types of modulations of the locomotor pattern as turns evoked by skin stimulation (Ullén et al. 1993; Wallén et al. 1994). Having one
common system for the generation of turns at the immediate
supraspinal level would avoid unnecessary redundancy.
A second, related issue concerns the relation of the command system for lateral turns to the descending control systems
for other forms of behavior, such as initiation of forward and
backward locomotion, braking, and turns in the vertical planes.
Since the RS system is the main system for the transmission of
all descending commands in the lamprey, one can assume that,
in general, different types of command are encoded as different
spatiotemporal patterns of activity in the whole population of
RS neurons. This raises the question of to what extent the same
J Neurophysiol • VOL
We thank D. Parker for valuable comments on the manuscript.
Support by grants from the National Institute of Neurological Disorders and
Stroke (NS-38022), the Howard Hughes Medical Institute (75195-544801), the
Swedish Medical Research Council (MFR 3026 and 11554), and the Royal
Academy of Science is gratefully acknowledged. F. Ullén was supported by a
postdoctoral grant from the Swedish Research Council for Engineering Sciences (TFR 282-96-905), the Swedish Brain Foundation, and by grants from
Åke Wibergs Stiftelse and Magnus Bergwalls Stiftelse.
BRODIN L, GRILLNER S, DUBUC R, OHTA Y, KASICKI S, AND HÖKFELT T.
Reticulospinal neurons in lamprey: transmitters, synaptic interactions and
their role during locomotion. Arch Ital Biol 126: 317–345, 1988.
BUSSIÈRES N. Les Systèmes Descendants chez la Lamproie. Étude Anatomique
et Fonctionelle (PhD thesis). Montreal, Canada: Univ. of Montréal, 1994.
BUSSIÈRES N, PFLIEGER JF, AND DUBUC R. Anatomical study of vestibulospinal
neurons in lampreys. J Comp Neurol 407: 512–526, 1999.
DELIAGINA TG, ULLÉN F, GONZALEZ MJ, EHRSSON H, ORLOVSKY GN, AND
GRILLNER S. Initiation of locomotion by lateral line photoreceptors in
lamprey: behavioural and neurophysiological studies. J Exp Biol 198: 2581–
2591, 1995.
DELIAGINA TG, ZELENIN PV, FAGERSTEDT P, GRILLNER S, AND ORLOVSKY GN.
Activity of reticulospinal neurons during locomotion in the freely behaving
lamprey. J Neurophysiol 83: 853– 863, 2000.
FAGERSTEDT P AND ULLÉN F. Lateral turns in the lamprey. I. Patterns of
motoneuron activity. J Neurophysiol 86: 2246 –2256, 2001.
FAGERSTEDT P, ULLÉN F, DELIAGINA T, WALLÉN P, AND GRILLNER S. Lateral
turns in the lamprey. I. Activity of spinal interneurons during fictive turns.
Soc Neurosci Abstr 24: 1156, 1998.
GRILLNER S. Neurobiological bases of rhythmic motor acts in vertebrates.
Science 228: 143–149, 1985.
HARDEN-JONES FR. Photokinesis in the ammocoete larva of the brook lamprey.
J Exp Biol 32: 492–503, 1955.
KASICKI S, GRILLNER S, OHTA Y, DUBUC R, AND BRODIN L. Phasic modulation
of reticulospinal neurones during fictive locomotion and other types of
spinal motor activity in lamprey. Brain Res 484: 203–216, 1989.
KLEEREKOPER H. The sense organs. In: The Biology of Lampreys, edited by
Hardisty IC and Potter MW. New York: Academic Press, 1972, vol. 2, p.
373– 404.
86 • NOVEMBER 2001 •
www.jn.org
RETICULOSPINAL ACTIVITY DURING FICTIVE TURNS
KLEEREKOPER H AND MOGENSEN J. Role of olfaction in the orientation of
Petromyzon marinus. I. Response to a single amine in prey’s body odor.
Physiol Zool 36: 347–360, 1963.
KOZLOV A, ULLÉN F, FAGERSTEDT P, AURELL E, LANSNER A, AND GRILLNER S.
Mechanisms for lateral turns in lamprey in response to descending unilateral
commands: a modeling study. Biol Cybern. In press.
MCCLELLAN AD. Descending control and sensory gating of ‘fictive’ swimming
and turning responses elicited in an in vitro preparation of the lamprey
brainstem/spinal cord. Brain Res 302: 151–162, 1984.
MCCLELLAN AD AND GRILLNER S. Initiation and sensory gating of ‘fictive’
swimming and withdrawal responses in an in vitro preparation of the
lamprey spinal cord. Brain Res 269: 237–250, 1983.
MCCLELLAN AD AND HAGEVIK A. Descending control of turning locomotor
activity in larval lamprey: neurophysiology and computer modelling. J Neurophysiol 78: 214 –228, 1997.
ORLOVSKY GN, DELIAGINA TG, AND GRILLNER S. Neuronal Control of Locomotion—From Mollusc to Man. New York: Oxford Univ. Press, 1999.
RONAN M. Origins of the descending spinal projections in petromyzontid and
myxinoid agnathans. J Comp Neurol 281: 54 – 68, 1989.
ROVAINEN CM. Electrophysiology of vestibulospinal and vestibulo-reticulospinal systems in lampreys. J Neurophysiol 42: 745–766, 1979.
ROVAINEN CM. Neurophysiology. In: The Biology of Lampreys, edited by
Hardisty IC and Potter MW. New York: Academic Press, 1983, vol. 4A, p.
1–136.
SWAIN GP, SNEDEKER JA, AYERS J, AND SELZER ME. Cytoarchitecture of
spinal-projecting neurons in the brain of the larval sea lamprey. J Comp
Neurol 336: 194 –210, 1993.
J Neurophysiol • VOL
2265
TEETER J. Pheromone communication in sea lampreys (Petromyzon marinus):
implications for population management. Can J Fish Aquat Sci 37: 2123–
2132, 1980.
ULLÉN F, DELIAGINA TG, ORLOVSKY GN, AND GRILLNER S. Spatial orientation
in lamprey. II. Visual influences on orientation during locomotion and in the
attached state. J Exp Biol 198: 675– 681, 1995.
ULLÉN F, DELIAGINA TG, ORLOVSKY GN, AND GRILLNER S. Visual pathways
for postural control and negative phototaxis in lamprey. J Neurophysiol 78:
960 –976, 1997.
ULLÉN F, FAGERSTEDT P, EKEBERG Ö, LANSNER A, ORLOVSKY GN, AND
GRILLNER S. Lateral turns in the lamprey. II. The brainstem descending
command system— computer modelling and lesion studies. Soc Neurosci
Abstr 24: 1156, 1998.
ULLÉN F, ORLOVSKY GN, DELIAGINA TG, AND GRILLNER S. Role of dermal
photoreceptors and lateral eyes in initiation and orientation of locomotion in
lamprey. Behav Brain Res 54: 107–110, 1993.
VIANA DI PRISCO G, PEARLSTEIN E, ROBITAILLE R, AND DUBUC R. Role of
sensory-evoked NMDA plateau potentials in the initiation of locomotion.
Science 278: 1122–1125, 1997.
WALLÉN P, ULLÉN F, DELIAGINA TG, ORLOVSKY GN, AND GRILLNER S. Modulations of the normal locomotor pattern during visually evoked yaw turns
in the lamprey. Soc Neurosci Abstr 20: 1408, 1994.
WANNIER T. Rostro-caudal distribution of reticulospinal projections from different brainstem nuclei in the lamprey. Brain Res 666: 275–278, 1994.
YOUNG JZ. The photoreceptors of lampreys. I. Light-sensitive fibres in the
lateral line nerves. J Exp Biol 12: 229 –238, 1935.
86 • NOVEMBER 2001 •
www.jn.org