THE JOURNAL OF COMPARATIVE NEUROLOGY 353:129-142 (1995)
Morphology of Developing Rat
Genioglossal Motoneurons Studied In
Vitro: Relative Changes in Diameter and
Surface Area of Somata and Dendrites
P.A. NUNEZ-ABADES AND W.E. CAMERON
Departments of Neuroscience and Psychiatry,
University of Pittsburgh, Pittsburgh, Pennsylvania 15260
ABSTRACT
This study describes the postnatal change in size of motoneurons in the hypoglossal
nucleus that innervate the genioglossus muscle. Such anatomical information is essential for
determining the cellular mechanisms responsible for the changes observed in the electrical
properties of these motoneurons during postnatal development. The cells analyzed here are
part of an earlier study (Nufiez-Abades et al. [1994] J. Comp. Neurol. 339:401-420) where 40
genioglossal (GG) motoneurons from four age groups (1-2, 5-6, 13-15, and 19-30 postnatal
days) were labeled by intracellular injection of neurobiotin in an in vitro slice preparation of the
rat brainstem and their cellular morphology was reconstructed in three-dimensional space.
The sequence of postnatal dendritic growth can be described in two phases. The first phase,
between birth (1-2 days) and 13-15 days, was characterized by no change in either dendritic
diameter (any branch order) or dendritic surface area of GG motoneurons. However,
maturation of the dendritic tree produced more surface area at greater distances from the soma
by redistributing existing membrane (retracting some terminal branches). During the second
phase, between 13-15 days and 19-30 days, the dendritic surface area doubled as a result of an
increase in the dendritic diameter across all branch orders and a generation of new terminal
branches. In contrast to the growth exhibited by the dendrites, there was little discernible
postnatal growth of somata.
At all ages, dendrites of GG motoneurons show the largest amount of tapering in the firstand second-order dendrites. The calculated dendritic trunk parameter deviated from a value
1.0, indicating that the dendritic tree of developing GG motoneurons cannot be modeled
accurately as an equivalent cylinder. However, the value of this parameter increased with age.
Strong correlations were found between the diameter of the first-order dendrite and the
dendritic surface area, dendritic volume, combined dendritic length, and, to a lesser extent, the
number of terminal dendrites in GG motoneurons. Correlations were also found between somal
and dendritic geometry but only when data were pooled across all age groups.
These data support earlier studies on kitten phrenic motoneurons, which concluded that
postnatal growth of motoneurons was not a continuous process. Based on the fact that there
was no growth in the first 2 weeks, the changes in the membrane properties described during
this phase of postnatal development (e.g., decrease in input resistance) cannot be attributed to
increases in the total membrane surface area of these motoneurons. It is proposed that these changes
result from changes in the specific membrane resistance of these neurons. D 1995 Wiley-Liss, Inc.
Indexing terms: brainstem, hypoglossal nucleus, upper airways, respiration,three-dimensional
reconstruction
Motoneuronal morphology and total membrane surface
area, in particular, have been cited as a critical factor in
determining the input-output
Of
neurons (Burke, 1981 1987).The geometry ofthe dendritic
arbors influences how synaptic inputs generated at various
9
o 1995 WILEY-LISS, INC.
AcceptedAugust23, 1994
Address reprint requests to William E, Cameron, PhD, Department of
Neuroscience, 446 Crawford Hall, University of Pittsburgh, Pittsburgh, PA
15260.
P.A. NUNEZ-ABADES AND W.E. CAMERON
130
sites on the tree will impact the integration at the axon
hillock (Barrett, 1975). This paper is part of a larger study
using in vitro brainstem slices to follow the postnatal
changes in the morphological and electrophysiological properties of developing genioglossal (GG) motoneurons (NufiezAbades et al., 1993,1994).In a previous paper, we identified
times during postnatal development when major changes
occur in the active and passive membrane properties of GG
motoneurons (Nufiez-Abades et al., 1993). Some of these
observed changes in the electrical properties of the motoneuron could be due solely to changes in cellular morphology.
For example, there was a large decrease (50%) in mean
input resistance of GG motoneurons between the first and
the second weeks of postnatal life. One simple explanation
for this decrease in resistance could be a doubling of
membrane surface area. The present data demonstrate that
this is not the case.
In the first paper examining the morphology of these
developing motoneurons (Nufiez-Abades et al., 1994), we
described the postnatal changes in the dendritic branching
of developing GG motoneurons. During the first 2 postnatal
weeks, we found that the dendritic trees undergo a decrease
in branching complexity (loss of terminal branches) that is
simultaneous with an increase in the length of other
branches. By the third postnatal week, the dendritic tree
reelaborates by regenerating new branches, probably as a
result of collateral outgrowths from terminal branches.
Three-dimensional analysis revealed that this spatial redistribution of GG dendrites during postnatal life was nonrandom. In the present study, we have reexamined the temporal pattern of growth exhibited by GG motoneurons by
focusing on the changes in the distribution of membrane
area as a function of the distance from the soma. These data
describing the changes in the dendritic geometry are important for generating an electrical model of the developing
neuron (Rall et al., 1992).
On the basis of morphometric analyses of intracellularly
labeled mammalian neurons, investigations have shown
several common principles that govern the growth and
maintenance of dendrites. A strong positive correlation has
been found between the diameter of the first-order dendrites and the surface area, volume, and combined length of
the dendrite in several populations of adult cat spinal
motoneurons (Ulfhake and Kellerth, 1981, 1983, 1984;
Cameron et al., 1985; Rose et al., 1985; Cullheim et al.,
1987) and rat neocortical pyramidal cells (Larkman,
1991a,b). This relationship has also been described by
workers from two laboratories in studies of developing
spinal motoneurons of the kitten (Ulfhake and Cullheim,
1988; Ramirez and Ulfhake, 1991; Cameron et al., 1991a).
The present study extends these investigations to the
development of brainstem motoneurons and to a different
species, the rat. Preliminary reports of these data have been
presented elsewhere (He et al., 1992; Nufiez-Abades et al.,
1992).
MATERIALS AND METHODS
The morphological data were obtained in an earlier study
of intracellularly labeled GG motoneurons with neurobiotin. The details of the surgery, recording, intracellular
injection, and histochemical processing have been described
previously (Nufiez-Abades et al., 1994). In brief, rats were
deeply anesthetized with halothane, tracheotomized, ventilated, transcardially perfused with a cold ( 1 ~ 4 ° Csucrose)
artificial cerebral spinal fluid (ACSF), and then quickly
decapitated. The brainstem was removed and sectioned at
300 km in the transverse plane using a Vibraslice and then
incubated at room temperature for at least 1 hour before
transfer to the recording chamber. GG motoneurons had
been retrogradely labeled by intramuscular injection of
dextran rhodamine made several days prior to surgery.
Intracellular impalements of GG motoneurons were made
in the ventromedial region of the hypoglossal nucleus with
the help of an upright microscope equipped with epifluorescence. A total of 40 GG motoneurons were selected for
reconstruction and they represented each of the following
groups: 1-2 days (n = lo), 5-6 days (n = lo), 13-15 days
(n = lo), and 19-30 days (n = 10).Seven cells in each group
had all their dendritic processes confined to the slice, and
these processes could be traced to terminations measuring
less than 1 bm in diameter. The remaining motoneurons
could only be partially reconstructed because one or two
dendritic branches left the plane of the section; these
branches exiting the slice always had diameters less than 1
km. The partially reconstructed cells were equally distributed among the age groups. The completely reconstructed
dendrites from these partially reconstructed cells were
included in the analysis involving individual dendrites, but
data from these cells were excluded in analyses dealing with
the morphometrics of the entire neuron.
All motoneurons were reconstructed using the Eutectic
Neuron Tracing System running on an IBM-compatible
computer (Capowski, 1989).With this system, the neurons
were reconstructed by making multiple measurements
(approximately every 5-10 km) of each dendritic branch
using a ~ 6 dry
0 objective (NA = 0.80). Each point was
designated as either a branch point, a continuation point, or
a natural (terminal) ending and was assigned X, Y, and Z
and diameter values. This form of analysis provides a more
accurate estimation of dendritic length, surface area, and
volume than measurements derived from planar reconstructions (cells drawn in two dimensions with the aid of a
drawing tube attachment; see Cameron et al., 1991a, for
details). With the Eutectic system, the combined dendritic
length was calculated as the summed lengths of all branches
of a dendrite, and the total extent of the dendritic territory
(area of influence) was quantified by calculating the minimum surface area required to enclose all the dendritic
terminals of the neuron. Because of the difficulty of determining the somatodendritic boundary, the cell bodies and
proximal dendrites were manually drawn using a camera
lucida attachment. The geometry of the cell bodies was
approximated by the largest ellipse that could be placed
within the boundaries of the cell (Schade et al., 1961). The
major and minor diameters (axes) of these ellipses and the
diameters of the first-order dendrites were measured from
the manual reconstruction using a digitizing tablet. Soma1
surface area and volume were calculated using the equation
for a prolate spheroid (Webber and Pleschka, 1976).
The terminology used to describe the dendritic tree is the
same as described previously in Nufiez-Abades et al. (1994).
In all reconstructed cells, each dendritic branch was assigned a branch order following a centrifugal ordering
method (cell body to distal terminal), and each branch was
identified as a preterminal or a terminal branch. The term
preterminal branch was used to identify parts of a dendrite
that connected the soma with the first branching point or
two successive branching points. A terminal branch was
defined as a branch that connected a branching point with a
CHANGES IN SIZE OF DEVELOPING GENIOGLOSSAL MOTONEURONS
131
TABLE 1. Quantitative Data for Developing Genioglossal Motoneurons'
Postnatal age (days)
Major diameter (bm)
Minor diameter (bm)
Mean diameter (bm)
Minorlmajor diameter
Soma1 surface area iprnz)
Total dendritic surface area (&m2J
Dendriticisomal surface area
1-2
5-6
13-15
19-30
29.3 ? 3.2 (10)
19.9 i 2.4 (10)
24.6 i 2.3 (10)
0.68 2 0.08 (10)
1,664 ? 319 (10)
14,033 t 1,589 (7)
9.1 2 1.9 ( 7 )
27.2 2 2.6 (10)
18.0 i 3.1 (10)
22.6 i 2.3 (10)
0.67 2 0.12 (10)
1,395 i 331 (10)
12,951 2,103 (7)
10.0 2 1.9 (7)
30.2 i 5.4 (10)
19.1 i 2.1 (10)
24.7 i 2.6 (10)
0.65 2 0.14 (10)
1,607 i 252 (10)
13,158 i 3,149 (7)
8.7 i 2.6 ( 7 )
32.1 i 5.9 (10)
21.0 i 3.5 (10)
26.5 i 4.1 (10)
0.66 i 0.10 (10)
1,921 t 571 (10)
24,406 i 2,366 (7)***
12.7 i 1.8 (7)*
*
'Values are means i SD. The total number in the samples is indicated in parentheses. Asterisks show level of significance between sequential age groups.
'P < 0.05.
**P< 0.01.
***P< 0.001
termination. Also, each dendritic branching point was
assigned a vertex index (Berry and Flinn, 1984; Verwer and
Van Pelt, 1985) indicating whether the parent branch gave
rise to two terminal branches (Va-node), one terminal
branch and to another parent branch (Vh-nOde),or two new
parent branches (no terminal branches; V,-node). The
diameter of dendritic branches was calculated as the average of multiple measurements along the entire branch
(every 5-10 Fm), and this value was analyzed separately for
terminal and preterminal branches. Further analysis of the
preterminal branches was undertaken with respect to the
branch order or type of branching node.
When modeling the neuron as an equivalent cylinder
(Rall, 1959, 19771, one assumes that the dendritic tree can
generate a constant surface area at varying distances from
the soma. This assumption was tested by calculating the
combined dendritic trunk parameter (Cameron et al., 1985,
1991a; Ulfhake and Cullheim, 1988). The dendritic trunk
parameter was defined as the sum of the daughter branch
diameters, each raised to 312 power divided by the diameter
of the first-order branch raised to the 312 power. If the
dendritic tree was a perfect equivalent cylinder, then the
value of this trunk parameter at various distances from the
cell body would be 1.0. The data of all dendrites from each
age group were pooled, and the dendritic trunk parameter
was calculated at intervals of 100 pm.
To quantitate the symmetry of dendritic branching, the
daughter branch ratio (DBR) was calculated for each
branch point as the ratio of the diameters of the larger
daughter segment to that of the smaller daughter segment
(Larkman, 1991a).A symmetrical branching would have a
ratio of 1.0 with increasing values of the DBR indicating
greater asymmetry. Tapering of dendrites was quantified
by measuring the amount of taper per 100 km of branch
length:
- ddistal)x 1001length of branch,
Taper = (dproximal
where dproximal
and ddistalare diameters of the proximal and
distal ends of the dendritic branch, respectively, and the
value is expressed in micrometers of taper per 100 pm
dendritic length. A positive value was derived when there
was tapering and a negative value when the branch expanded.
For each geometric parameter, the statistical differences
between two age groups' data were tested using the nonparametric Mann-Whitney U test, while comparisons among
more than two groups were accomplished with the KruskalWallis analysis of variance. In the text, statistical significance has been indicated as follows: *P < 0.05, **P < 0.01,
and ***P < 0.001. The functions describing the relation-
ships between the various quantitative measures of dendritic morphology were determined by the method of least
squares fit. The data points were fitted by linear, power,
exponential, or second-order polynomial functions, depending upon which yielded the largest correlation coefficient (r).
RESULTS
Original data consisting of light photomicrographs of
intracellularly labeled neurons and their respective threedimensional, computer-assisted reconstructions have been
presented in an earlier paper (Nuhez-Abades et al., 1994).
In addition, representative reconstructions of cells from
each of the four age groups were also presented in this
earlier work.
Growth of soma relative to dendrites
In all cases, the shape of the cell body was best fit by an
ellipse. Analysis of major, minor, and mean [(major +
minor)/2] diameters of the cell body revealed no changes in
any of these dimensions during postnatal development and,
as a result, the ratio of minor to major diameters was
unchanged with age (Table 1).At all four age groups, the
ratio of 0.7 indicated a slight eccentricity of GG cell body;
however, no preferential orientation of this eccentricity was
noted. Motoneurons maintained their multipolar shape
during development. Furthermore, the orientation of the
major diameter was totally unrelated to the nonrandom
spatial orientation of the dendrites (Nufiez-Abades et al.,
1994). Without significant change in major or minor diameters, it follows that there was no significant increase in
calculated somal surface area with age. Therefore, the adult
size and shape of the GG somata were apparently established at birth.
Up to 13-15 days, there was no increase in total membrane surface area (see Table 1).Therefore, dendritic to
somal surface area ratio was unchanged during the first 2
postnatal weeks. Between 13-15 and 19-30 days, the
dendrites exhibited a significant increase in surface area at
a time when somal dimensions increased by only 16%.The
dendritic surface area nearly doubled between 13-15 days
and 19-30 days and thereby increased the dendritic to
somal surface area ratio by 46% (Table 1). Based on
anatomical and physiological criteria (see Nuhez-Abades et
al., 1994), rat GG motoneurons appear to be mature by 19
days of postnatal life. If the 19-30-day-old group represents
the adult state, then the cell bodies exhibit less growth
during development than the dendrites exhibit.
Table 2 presents the number and diameter of the firstorder dendrites in each age group. The increase in dendritic
-
P.A. NUNEZ-ABADES AND W.E. CAMERON
132
TABLE 2. Quantitative Data of Developing Genioglossal Dendrites'
Postnatal age (days)
1-2
No. dendriteslneuron
Diameter first-orderdendrite
Comb. diameter first-order dendrite ibm)
Diam. all dendritic branches (bm)
Diam terminal branches ( p m )
Total surface arealneuron (+m2)
Dendritic volumelneuron !rm3)
Comb. dendritic lengthlneuron ipm)
No. terminalslneuron
Area of influence (urn2)
5.8 i 1.1(10)
2.5 2 1.5 (65)
14.6 ? 2.2 (10)
0.97 i 0.91 (340)
0.57 i 0 19 (191)
15,624 ? 1,669 (7)
5,682 ? 1,050 (7)
4,360 ? 1,429 (7)
27.0 2 8.3 (7)
128,088 2 39,146 (7)
5-6
13-15
19-30
6.1 i 11(10)
2.2 i 1.5 (58)
12.7 ? 1.9 (10)
1.00 5 0 . 9 1 (259)
0.62 5 0.25 (150i*
14,282 2 2,240 (7)
4,356 t 1,101 (7)
4,596 t- 788 (7)
21.3 i- 4.1 (7)
181,022 t- 36,090 (7)
5.7 i 1.1 (10)
2.1 2 1.4 (54)
12.2 i 2.4 (10)
1.04 ? 1.05 (191)
0.63 ? 0.20 (116)
14,717 ? 3,056 (71
4,970 ? 1,816 (7)
4,241 ? 1,059 (7)
16.0 5 4.3 (7)
176,145 ? 55,728 (7)
6.5 5 0.7 (10)
3.2 i 1.5 (62)***
20.7 i 3.0 (lo)***
1.36 ? 1.13 1297)**
0.78 5 0.30 i172)***
26,365 ? 2,696 (7)"*
11,757 ? 3,115 (7i***
6,244 t- 704 (7)'*
24.6 i- 1.4 (7)**
312,616 i 110,862 (7)'
~~
'Values and statistical representation as in Table 1.
Fig. 1. Dendritic branch diameter of developing genioglossal (GG) motoneurons as a function of
dendritic branch order and age. In the oldest age group, there was an increase in mean diameter at all
dendritic branch orders. Dendritic diameter was calculated as the average of all dendrites of a given branch
order. The cellular dimensions presented in this and the next six figures represent the means from seven
totally reconstructed cells in each age group.
surface area observed was not due to an increase in the
number of dendrites per cell; their number remained
unchanged postnatally. However, there was an increase in
the diameter of first-order dendrites between 13-15 days
and 19-30 days that also produced an increase in the
combined diameter of first-order dendrites.
Morphometry of individual dendrites
To understand how growth of surface area occurs in
different parts of the dendritic tree, we analyzed the
changes occurring in different generations of branching
(terminal and preterminal) and nodal types. The mean
diameter of all dendritic branches showed no significant
growth up to 13-15 days (Table 21, then increased significantly ( P < 0.01) from 1.04 f 1.05 km at 13-15 days to
1.36 ? 1.13 Frn at 19-30 days. The mean branch diameter
in the different branch orders is shown in Figure 1.There
was a rapid decline in diameter between the first and the
second branch orders in all age groups. For all four age
groups, the diameter of the second branch order was
approximately one-half that of the first branch order. In
higher order branches of the dendritic tree, the decline in
mean branch diameter was less pronounced. Between birth
and 13-15 days, there was no significant increase in
diameter of any branch order. However, the growth between 13-15 days and 19-30 days was characterized by a
significant increase (P < 0.05) in the diameter of all
dendritic branch orders. This increase ( P < 0.01) was most
evident in the first and second branch orders. A similar
pattern emerged when the caliber of terminal and pretermi-
CHANGES IN SIZE OF DEVELOPING GENIOGLOSSAL MOTONEURONS
nal branches was analyzed separately (Fig. 2A,B), where
there was an increase in mean diameter in all dendritic
branch orders (either preterminal or terminal branches) a t
19-30 days. In relative terms, the decrease in diameter
between subsequent branch orders was larger for preterminal than for terminal branches. There were also growth
patterns in preterminal branches that were specific to nodal
vertex type (Fig. 2 0 . In the four age groups, the diameter
of preterminal branches at V,-nodes was statistically smaller
(P < 0.001) than that at v h - or V,-nodes. At 1-2 days and
19-30 days, the diameter of the different nodal types was
V, < Vb < V, while at 5-6 days and 13-15 days, the pattern
was V, < Vb = V,.
Figure 3A shows the relationship between dendritic
branch diameter and dendritic branch length for preterminal and terminal branches (inset) at 19-30 days. Although
only data from the 19-30 day group is presented in Figure
3, the average diameter of the terminal branch was positively correlated (P < 0.001) with the branch length for all
four age groups. In contrast, this relationship of diameter
to length for the preterminal branches was best fit by a
second-order polynomial function ( P < 0.05). When this
analysis of preterminal branches was extended to specify
the nodal type (Fig. 3B), a negative linear correlation (P <
0.001) was found for only Vh-nodes while no correlations
were found for either V,- and V,-nodes.
-0- 1-2days
--(I 5-6days
h
5
v
+13-15days
1.0-
19-30 days
-A-
t
4-
E
i
.-
k
+
o.8-
0.60.4-
'
0.2
,
1
2
3
4
5
6
7
8
7
8
Branch Order
B
4
h
5
4
v
t
3
4-
a,
5
i
i
-
2
i
1
.-2
Geometry of dendritic trees
In order to predict how synaptic current generated at
various locations on the dendritic tree will passively spread
to the cell body and initial segment, it is important to
describe the geometry of the motoneuron membrane. In
Figure 4, the distribution of dendritic surface area is plotted
as a function of distance along the dendrite for each of the
four age groups. There was a rapid decline in dendritic
surface area within the first 300 km from the soma at all
ages. In general, maturation of the GG dendritic tree
produces more surface area at greater dendritic distances
from the soma, as evidenced by the fact that there were no
dendrites (and, therefore, no surface area) beyond 700 km
from the soma found prior to 13-15 days. Total neuronal
surface area, volume, and combined dendritic length remained unchanged in the first 2 weeks of postnatal life
(Table 2). At 13-15 days, growth beyond 700 Fm was
counterbalanced by a small decrease in dendritic surface
area within the first 200 km. In a previous paper (NubezAbades et al., 1994), we described that the elongation in
terminal branches at 13-15 days was accomplished by
redirecting membrane of resorbed branches. Later (beyond
13-15 days), the dendritic surface area doubled as a result
of increases in both diameter and branch length. This
growth was achieved by increasing the dendritic diameter
along the whole dendritic tree (see Fig. 1) and increasing
the combined dendritic length of the neuron, in part by
reelaborating new terminal branches (see Table 2 and
Niifiez-Abades et al., 1994). Coincident with the increase in
the dendritic surface area between 13-15 days and 19-30
days, there was also roughly a doubling in the dendritic
volume and the area of influence.
There are several assumptions regarding the geometry of
the motoneuron that permit us to approximate the dendritic tree as an equivalent cylinder (Rall, 1959, 1977).One
such assumption is that the dendritic trunk parameter (see
Materials and Methods) should be approximately equal to
1.0 at varying distances from the soma. In Figure 5 , the
133
-
?!?
a
1
1
2
3
4
5
6
Branch Order
C
4h
5
-t
v
3-
a,
5
B
L:
0
C
2 -
I
m!
1
-
1
VC
vb
Va
Branch Type
Fig. 2. Dendritic diameter of terminal (A) and preterminal (B)
branches as a function of branch order for the four age groups. C:
Average branch diameter for each preterminal vertex type for each age
group. All changes in diameter between nodal vertices were significant
( P < 0.001)except those between V,- and Vb-nodes for 5-6 and 13-15
days. Symbols are mean values with standard error bars.
value of the dendritic trunk parameter has been plotted for
four age groups as a function of distance from the soma
along the dendrite. Similar to the distributions found for
surface area, the values of the dendritic trunk parameter
P.A. NUNEZ-ABADES AND W.E. CAMERON
134
A
700
600
500
400
300
1
v
1
v
n
7
L
E
a
v
100
c
S
B
e,
.-c4
o-
n
3
5
4
6
7
700
0
L
U
c
a,
2
1
o a
600
a
500
o
V,-nodes
A
Vb-lIOdeS
A
400
Vc-nodes
A
0
300
200
100
1
2
3
4
5
6
7
Dendritic Diameter (pm)
Fig. 3. Relationship between branch diameter and branch length in
19-30 days old GG motoneurons. The diameter represents the mean of
all measurements (5-10 km increments) along the segment. A Preterminal branches. Inset: Terminal branches. Preterminal branches in A
are best fit by a second-order polynomial function (r = -0.44) and
terminal branches by a linear function (r = 0.66). B: Preterminal
branches vs. diameter replotted specifying the nodal vertex type. A
significant linear correlation was found only for preterminal branches
ofVb-nodaltype (r = -0.52, P < 0.001).
were characterized by a rapid decline in this parameter over
the first 300 pm. This plot demonstrates that the dendritic
trunk parameter decreases with distance at all ages studied.
The large deviations of this value from 1.0 indicate that
developing GG dendrites are not accurately modeled as an
equivalent cylinder.
The observed deviations of the dendritic tree from an
equivalent cylinder could result from asymmetrical branching and/or tapering of branches (Barrett and Crill, 1974).
Branch symmetry was evaluated by calculating the daughter branch diameter ratio (DBR). When DBRs were calcu-
lated for the GG motoneurons in the four age groups, all
ratios exceeded 1.0 to varying degrees. Initially, similar
values were obtained for 1-2 days and 5-6 days (1.31 ? 0.43
and 1.29 2 0.43, respectively) while significantly larger
( P < 0.05) values were found at 13-15 days (1.54 & 0.96)
and 19-30 days (1.52 0.51). Therefore, the asymmetry at
branching points in GG motoneurons increased with age.
Further analyses were done taking into account the nodal
vertex type (Table 3). V,-nodes had a value closer to 1.0
than those found in V,- and Vb-nodes at four age groups.
While the values of DBR remained unchanged in V,-nodes,
*
CHANGES IN SIZE OF DEVELOPING GENIOGLOSSAL MOTONEURONS
135
a,
0
(d
‘t:
3
a
L
a,
n
Fig. 4. Distribution of dendritic surface area per cell at 100 pm
intervals from soma along the dendrite as function of age. In this figure
and the next two figures plotting distance along the dendrite, the X axis
starts at 100 pm, not at zero, which would be equivalent to the
boundary with the soma. Note that the data are calculated at regular
increments of distance from soma (e.g., 100 pm, 200 pm, etc.). In order
to extend greater distances, the dendrite GG motoneurons at 13-15
days generated less surface area within the first 600 pm than that
found for the dendrites at 1-2 days.
DBR in V,- and Vb-nodes increased significantly (P < 0.001
and P < 0.05, respectively) postnatally.
Another parameter indicates that dendritic asymmetry is
an unequal length from the soma to dendritic terminations
among dendritic branches of the same dendrite. Figure 6
shows the mean number of terminal branches per cell as a
function of distance along the dendrite at 100 pm intervals
from the soma. The distribution of terminal endings was
similar between 1-2 and 5-6 days, with a peak at 300 pm
from the soma. However, dendrites at 13-15 and 19-30
days produced fewer numbers of terminal endings within
the first 200 pm from the soma and a larger number of
them at a distance at 700 pm and beyond. The lower
number of terminal branches found in the first 400 pm at
13-15 days was due to the less complex branching of the
dendrites at this age (Nufiez-Abades et al., 1994). In
general, terminal branches end at nonuniform lengths from
the soma at all four age groups. Although our analysis has
been restricted to anatomical distance, a close relationship
between electrotonic and anatomical distance has been
described in cortical neurons (Larkman et al., 1992). This
would imply that dendritic branches of GG motoneurons
may have differences in electrotonic length that, in turn,
would produce a decrease in dendritic trunk parameter.
The contribution that unequal lengths to dendritic terminals make to decreasing the dendritic trunk parameter has
been demonstrated for cat cervical motoneurons (Rose et
al., 1985).
A second factor contributing to the decrease in dendritic
trunk parameter is dendritic tapering (Barrett and Crill,
1974; Rose et al., 1985). Figure 7 shows the taper (measured in pmi100 pm length) as a function of dendritic
branch order. The largest taper was found in the first
branch order in the older animals and, to a lesser extent, in
the first- and second-order dendrites of the newborn.
Therefore, tapering may account for the rapid decline in
dendritic trunk parameter found within the first 200-300
pm of the dendritic tree at birth and in the first 200 pm of
the dendritic tree of older GG motoneurons. Further
analysis of tapering was conducted by subdividing branches
into terminal and preterminal branches. Tapering in terminal or preterminal branches of adult GG motoneurons (1.24
& 0.95 and 0.40 ? 0.29, respectively) had values similar to
those at birth (0.92 0.79 and 0.35 5 0.29, respectively).
For all four age groups, preterminal branches exhibit
greater tapering (all significant, at least P < 0.05) than that
found in terminal branches. The quantitation of tapering in
the finer terminal dendrites may be restricted by our ability
to accurately resolve diameters less than 0.5 pm. Given this
limited resolution, there might be real differences between
ages in tapering in higher order branches that cannot be
resolved at the level of light microscopy. Because of the
importance of the surface area measurement to the conclusions of this paper, we calculated that surface area would
decrease by less than 5%if all terminal branches tapered to
endings of 0.1-0.2 pm instead of 0.4-0.5 pm.
*
P.A. NUNEZ-ABADES AND W.E. CAMERON
136
Fig. 5. Distribution of the dendritic trunk parameter as a function of distance from soma and age. The
dendritic trunk parameter was calculated as the sum of all dendritic branch diameters raised to 312 power
measured at 100 p,m intervals, then divided by the sum of the first-order dendritic diameters raised to 312
power. Values less than 1.0 indicate that the developing dendritic tree deviates from an ideal equivalent
cylinder.
TABLE 3. Daughter Branch Ratios'
+1-2 days
Postnatal age (days)
1-2
5-6
13-15
V,-nodes 1 10 ? 0.15 162) 1.12 ? 0.17 (55) 1.10 i- 0 13 137)
Vb-nodes 1.41 i 0.40 (62) 1.48 i 0.63 (33) 2.12 i 1.29 (281*
V,-nodes 1.58 i- 0.68 ( 2 5 ) 1.44 ? 0.42 (20) 1.56 ? 0.86 (9)
19-30
1.32 2 0 29 152)***
1.71 i 0.62 (591*
1.45 -t 0.28 (13)
6-
43- 5-6 days
A
--t 13-15 days
*19-30 days
'Values and statistical representation as in Table 1
Relationship between somal and
dendritic geometry
Several dimensions of the soma and dendrites that are
relatively easy to measure have been shown to predict the
extent and complexity of the dendritic tree in cat spinal
motoneurons (Zwaagstra and Kernell, 1980; Ulfhake and
Kellerth, 1981, 1983). These relationships are also found
throughout postnatal development of cat spinal motoneurons (Ulfhake and Cullheim, 1988; Cameron et al., 1991a).
The plots of dendritic dimensions as a function of somal size
are presented in Figure 8 for developing brainstem (GG)
motoneurons of the rat. Correlations were found between
the somal diameter and the combined dendritic diameter of
first-order dendrites (r = 0.42, P < 0.01; Fig. 8A), and the
dendritic surface area (r = 0.50, P < 0.001; Fig. 8B).
Similarly, correlations were found between the somal surface area and the combined dendritic diameter (r = 0.49,
P < 0.01; Fig. 8C), and the dendritic surface area (r = 0.60,
P < 0.001; Fig. 8D) for data pooled across age. However, no
\
200
400
600
800
1000
Distance Along Dendrite ( c m )
Fig. 6. Relationship between the number of terminal branches per
cell as a function of distance along dendrites. On the abscissa, the values
represent the mean number of dendrites having terminals with a 100
p,m increment. Note that dendritic branches terminate at varying
lengths from the soma in all four age groups.
correlations between these parameters were found when
analyzed within any individual age group. Furthermore, no
correlations were found between either the somal dimensions and the number of first-order dendrites, or the
CHANGES IN SIZE OF DEVELOPING GENIOGLOSSAL MOTONEURONS
137
Fig. 7. Absolute tapering (&mi100 pm) of developing GG dendrites within various branch orders. Note
that the largest amount of tapering is concentrated in the first and second branch orders for all four age
groups.
number of terminal branches for the pooled data (not
illustrated).
It is important to know if there are any dimensions of the
dendrites that could serve to predict the size and complexity
of a dendritic tree without performing laborious, detailed
reconstructions. Strong correlations were found within the
individual age groups for geometry intrinsic to the dendrites. Figure 9 illustrates the relationship between the
diameter of the first-order dendrite and the dendritic
surface area for each age group. Linear correlation coefficients (r) values ranged between 0.94 and 0.97 (all significant, P < 0.001). Similar results were found in Figure 1OA
for the relationship between the diameter of the first-order
dendrite and the dendritic volume at all ages (r = 0.920.95, P < 0.001). In general, the slope of the regression
lines varied during development with the oldest age group
tending to have a steeper slope than that of the newborn
group.
The size of the dendritic tree is also determined by the
diameter and the combined dendritic length of the dendrite
in GG motoneurons. Linear correlations were also found
between combined dendritic length and dendritic surface
area (r = 0.93-0.96; not shown) and dendritic volume (r =
0.76-0.84; not shown). Because both dendritic surface area
and volume are correlated positively to the diameter of the
first-order dendrites as well as the dendritic combined
length, it follows that the two latter parameters should be
interrelated. The correlation coefficients for the combined
dendritic length as a function of the diameter of the
first-order dendrite ranged from r = 0.85-0.93 across the
four age groups (Fig. 10B). During postnatal development,
there was a transient increase in the slope of the regression
line between 1-2 and 5-6 days that subsided at the older
ages. This transient implies a different temporal pattern for
the elongation in GG motoneurons with age, as previously
described by Nufiez-Abades et al. (1994).
Weaker correlations were found between the number of
terminal branches and the diameter of the first-order
dendrite (r = 0.75-0.84), dendritic surface area (r =
0.75-0.871, and dendritic volume (r = 0.63-0.76) than
those correlations mentioned above. The strongest correlation for the number of terminal branches (r = 0.82-0.91)
was found with dendritic combined length (Fig. 11).There
was a trend for the regression to decrease its slope with age.
This change in slope reflects the elimination of terminal
branches observed between 5-6 days and 13-15 days, and
the reelaboration of the original dendritic complexity at
19-30 days (Nufiez-Abades et al., 1994).
DISCUSSION
The goal of this study was to describe the growth of rat
GG motoneurons. During postnatal development, the sequence of changes in size, as previously found for changes
in dendritic branch structure (Nufiez-Abades et al., 19941,
can be described in two phases for GG motoneurons. The
first phase (from birth to 13-15 days) was characterized by
no growth in either dendritic diameter or dendritic surface
area. However, the cell morphology is not static during this
period; rather, maturation produces more surface area at
greater distances from the soma by redistributing preexistent membrane. During the second phase (beyond 13-15
days), the dendritic surface area doubled due to both an
increase in dendritic diameter at all branch orders and the
P.A. NUNEZ-ABADES AND W.E. CAMERON
138
0 1-2 days
A,
0 13-15 days
C
5 28
28
v
A
A
L
$ 24
24
A
A
E 20
A
1
Y
AL
n
L
16
16
6. 12
12
-
cv
E
Am
A
&
6
m
* Rm
0 0
0
I
18
22
26
30
25-
34
30
A
A
A
6
12
18
24
30 -
A
A
0
5
'0"
A
c9
v
b
Ao
$
B
Q
m
a
A
42
0
A
20
.(d
2
A
A
25 -
30
A
A
A
A
A A
A
A
20 -
20-
Soma1 Diameter (pm)
Fig. 8. Relationship between mean somal diameter and combined
dendritic diameter of first-order dendrites (A) and total dendritic
surface area (B). Relationship between somal surface area (SA) and
combined dendritic diameter (C) and total dendritic surface area (D).
Data from seven totally reconstructed cells and three partially recon-
structed cells in each age group were used for A and C, while only data
from seven totally reconstructed cells in each age group were used for B
and D (see Materials and Methods). Significant correlations ( P < 0.01)
were found only for data pooled across ages but not for any individual
age group.
generation of new terminal branches. In contrast to the
growth observed in dendrites, there was no statistically
significant growth in any somal dimensions measured
postnatally, suggesting that adult size may be achieved by
birth in GG motoneurons. Finally, we found strong correlations between various aspects of dendritic geometry in
developing GG motoneurons similar to those found in other
studies of adult spinal motoneurons.
motoneurons had essentially achieved their adult size and
shape at birth. This discrepancy between the species could
be due to differences in methodology (intracellular vs.
retrograde labelling), or a more accurate estimate of mean
diameter from the cat study based on a larger sample
(100-150 cells per age group), or simply that rat GG
motoneurons mature earlier in postnatal life than those in
the cat.
The dendritic tree of GG motoneurons represents 90% of
the total surface area at birth. This percentage is smaller
than values reported in adult spinal motoneuronal pools. In
cat spinal motoneurons, the dendritic tree accounts for
greater than 95% of the total membrane surface area in
adults (Ulfhake and Kellerth, 1981, 1983; Cameron et al.,
1985; Rose et al.,1985; Cullheim et al., 1987). Therefore,
the somata of these rat brainstem motoneurons represents
a larger relative area available for potential synaptic contacts than the somata of cat spinal motoneurons. In
addition, we found that the dendritic to somal surface area
ratio increased postnatally in GG motoneurons, as previously reported for developing lumbosacral and phrenic
motoneurons (Ulfhake and Cullheim, 1988; Cameron et al.,
1991a).This change in the ratio represents a different time
Postnatal changes in size of the receptive
domains
To date, a variety of studies have described the rate of
growth of the soma in a number of motor nuclei (Mellstrom
and Skoglund, 1969; Sat0 et al., 1978; Rose et al., 1984;
Arvidsson et al., 1987; Tatton and Theriault, 1988; Cameron and He, 1989). One of these studies examined the
growth in cat GG motoneurons (Brozanski et al., 1989).
Using a similar method to measure somal dimensions,
these authors reported that the shape and size of the
somata changed during postnatal development. Specifically, the mean somal diameter value at birth was only
two-thirds of the adult value and the soma became more
elongated with age. In contrast, we found that rat GG
1
CHANGES IN SIZE OF DEVELOPING GENIOGLOSSAL MOTONEURONS
10000
8ooo
1
1-2 days (n=55)
6000
$
.-o
+
I
I
I
I
I
I
1
13-15 days (n=54)
8000
0
1
10000
3
cn
-
5-6 days (n=58)
I
n/
I
139
19-30 days (n=62)
o
0
6000
4000
2000
I
1
2
3
4
5
6
7
8
1
I
I
I
I
I
I
I
2
3
4
5
6
7
8
Diameter of First Order Dendrites (pm)
Fig. 9. Relationship between the diameter of the first-order dendrite and the dendritic surface area for the four age groups. The points
were fit by a least squares analysis yielding the following equations: y =
999x - 78, r = 0.94 (1-2 days); y = 1212x - 396, r = 0.97 (5-6 days); y
= 1139x - 288, r = 0.95 (13-15 days); y = 1371x - 776, r = 0.95 (19-30
days). All correlation coefficients significant (P < 0.001). Data of totally
reconstructed dendrites (number in parentheses) from ten cells (seven
total and three partial) in each age group were used in this and the next
figure.
course of growth for the soma as compared to the dendrites.
A similar progression was found in developing triceps surae
motoneurons of the cat where the soma reached its adult
size at 44-46 postnatal days while dendrites were only
one-half of their adult surface area (Ulfhake and Cullheim,
1988). In general, it would appear that the somal surface
area constitutes a larger target for afferent synapses in
developing motoneurons than it does in the adult motoneurons.
Only a few studies have provided information to date on
the postnatal elaboration of the dendritic tree. One study
found that the diameter of the dendrites of cat lumbar
motoneurons increase during postnatal development (Tatton et al., 1983). Growth in diameter has been reported to
occur between two months and the adult in phrenic motoneurons (Cameron et al., 1991a),while 50% of the diameter
increase in lumbosacral motoneurons takes place after
three weeks of age (Ramirez and Ulfhake, 1991). Growth in
membrane surface area in developingtriceps surae motoneurons of the cat has been reported as a continuous process
(Ulfhake and Cullheim, 1988). However, investigations of
phrenic motoneurons in the cat (Cameron et al., 1991a),
and in the rat (Spielmann et al., 1992), as well as the
present study of GG motoneurons in the rat show that the
process of motoneuronal growth is not continuous. Instead,
there are periods when there is no increase in total surface
area, and these periods coincide with a remodeling of the
dendritic tree that produces a substantial change in branching complexity and dendritic length. These dramatic changes
in morphology are accompanied by important changes in
the electrophysiological properties of these cells (Cameron
et al., 1991a; Mazza et al., 1992; Ndfiez-Abades et al., 1993).
Functional significance
One goal of the present study was to determine whether
the growth of total membrane surface area could explain
the rapid decrease in mean input resistance of GG motoneurons measured between 5-6 days (85.1 2 25.9 MR) and
13-15 days (45.5 ? 18.6MR; Nufiez-Abades et al., 1993). In
the present study, we conclude that there was no increase
in surface area between 5-6 days and 13-15 days and,
therefore, the increase in membrane surface area cannot
account for the decrease found in input resistance. A
similar conclusion was derived from work in cat phrenic
motoneurons where the mean input resistance of those
neurons decreased from 4.1 r 0.7 MR at two weeks to 1.9 &
1.0 MR at one month (Cameron et al., 1991b). During that
period, there was a similar lack of increase in total membrane surface area (Cameron et al., 1991a). Therefore, we
propose that the mechanisms to explain the decrease in
input resistance involves a change at the level of the unit
membrane. One possibility to explain the decrease in input
resistance is a decrease in specific membrane resistance of
P.A. NUNEZ-ABADES AND W.E. CAMERON
140
A
5000
1
0
-
E
4000
-z
3
2500
/'g-30
B
5
2000
a
_I
0
._
c
.-
1
i
1500 -
U
C
t3
u
a
n
1000
-
C
.-
0
500-
I
I
I
I
I
I
I
I
1
1
2
3
4
5
6
7
8
Diameter of First Order Dendrites (pm)
Fig. 10. A: Relationship between the diameter of the first-order
dendrite and the dendritic volume for the four postnatal age groups.
The equations for the linear regression line are: y = 506x - 245, r =
0.93 (1-2 days); y = 473x - 232, r = 0.95 ( 5 4 days); y = 533x - 318,
r = 0.92 (13-15 days); y = 783x - 659, r = 0.93 (19-30 days). B:
Relationship between the diameter of the first-order dendrite and the
combined length dendritic for the four postnatal age groups. The
equations of the linear regression are: y = 294x + 47, r = 0.85 (1-2
days); y = 414x - 102, r = 0.93 (5-6 days); y = 355x - 21, r = 0.87
(13-15 days); y = 318x - 96, r = 0.85 (19-30 days). All significance
levels exceededP < 0.001.
the motoneuron. This hypothesis is supported, in part, by
the observed decrease in membrane time constant (from
10.0
4.2 to 7.3 f: 3.3 ms) found during this period
(between 5-6 and 13-15 days) in GG motoneurons (NubezAbades et al., 1993). Even stronger support comes from
direct measurements of specific resistance that we have
made in developing GG motoneuron using combined electrophysiology and morphometry. These data indicate a transient decrease in the mean specific membrane resistance of
GG motoneurons at 13-15 days (Nunez-Abades et al.,
1992).
Systematic correlations have been described between
various aspects of dendritic geometry in adult cat spinal
motoneurons (Ulfhake and Kellerth, 1981, 1983, 1984;
Cameron et al., 1985; Rose et al., 1985; Cullheim et al.,
1987). In the present study, we found a strong correlation
between the diameter of the first-order dendrite and the
*
dendritic surface area, dendritic volume, combined dendritic length, and, to lesser extent, with the number of
terminal dendrites in GG motoneuron. This relationship
provides an important tool to estimate dendritic parameters, such as total membrane surface area, without the
need to make detailed measurements of the entire tree. In
addition, dendritic surface area can be estimated for neurons that have been only partially labeled or have processes
that leave the plane of section. Furthermore, the existence
of these relationships across motor pools and species suggests that there are general principles governing the growth
and maintenance of mammalian motoneuron dendrites.
There is some controversy as to whether these principles
extend to vertebrates in general. Weak correlations ( P <
0.01) between first-order dendrite diameter and combined
dendritic length (r = 0.404),and between dendritic diameter and total membrane surface area (r = 0.461), have
been reported for brachial and lumbar spinal motoneurons
in the frog (Birinyi et al., 1992). The greater variability
found for these correlations in frog rnotoneurons may
reflect differences in the methods of measuring cellular
geometry. First, Birinyi and colleagues (1992) calculated
their stem dendrite diameter as the average of three
measurements (one at the boundary of the cell body, a
second at the first branch point, and the third midway
between these two sites). In the mammalian studies, the
dilated areas of the first-order dendrite occurring at the
boundary with the cell body and at first branch point have
always avoided because they inaccurately reflect the mean
segment diameter. Thus, the calculated diameter in the
frog study would introduce greater variability when compared with the diameters measured in the mammalian
literature. In addition, the sample was small (5 cells per age
group). Furthermore, the motoneurons analyzed in the frog
study represent a mix of cells from the brachial and lumbar
spinal cord without any reference to the target muscle that
they innervate. It has been demonstrated that variability in
these dendritic correlations increases when measurements
from motoneurons supplying different hindlimb muscles in
the cat are pooled together (Ulfhake and Kellerth, 1983).
This effect would be compounded by pooling motoneurons
from two extremes of the spinal cord. Given the differences
between the amphibian and mammalian analyses, the
relationships between various dimensions of the dendritic
tree may be a principle in all vertebrates neurons.
In addition to changes in these relationships among
different motor pools, there are also changes associated
with development. As described previously for lumbosacral
and cervical motoneurons (Ulfhake and Cullheim, 1988;
Cameron et al., 1991a), we found that the nature of the
relationship between dendritic parameters is adjusted postnatally in GG motoneurons. For example, a given diameter
of the first-order dendrite in an adult animal will generate a
larger dendritic surface area and volume than a dendrite of
similar caliber in a newborn animal. These types of changes
may be necessary to maintain synaptic efficacy. The generation of more dendritic surface area provides more sites for
synaptic inputs onto the motoneuron, and a proliferation of
synaptic inputs (increase in synaptic current) may be
required to compensate for the decrease in input resistance
found in GG motoneurons and to maintain the motor
output.
The growth of the dendritic membrane area will also alter
the dendritic cable properties of the cell and, thus, the
postsynaptic effects of its synapses (Rall, 1977; Horwitz,
CHANGES IN SIZE OF DEVELOPING GENIOGLOSSAL MOTONEURONS
12 1
0
141
-
0
1-2 days (n=55)
-
5-6 days (n=58)
0
86-
El
.-C
-0
c
w
4-
2-
.l2
-I
a
10
._
1
7
13-15 days (n=54)
-
19-30 days (n=62)
0
-
8
0
0
6
-
4
-
2
I
500
1000 1500 2000 2500
500
I
,
I
I
1000 1500 2000 2500
Combined Dendritic Length (pm)
Fig. 11. Relationship between combined dendritic length and number of terminal branches for the four
age groups. The equations ofthe linear regression are: y = 0 . 0 0 4 7 ~+ 0.62, r = 0.91; y = 0 . 0 0 3 2 ~ 0.92, r =
0.90; y = 0 . 0 0 2 6 ~+ 0.64, r = 0.82; y = 0 . 0 0 3 3 ~ 0.66, r = 0.87 for 1-2, 5-6, 13-15, and 19-30 days,
respectively. All correlations significant ( P < 0.001).
+
1981; Wolf et al., 1992).The detailed morphology generated
in the present study has been used to test some anatomical
assumptions underlying the equivalent cylinder model of
the motoneuron (Rall, 1959, 1977). We found that the
dendritic trunk parameter deviated from a value of 1.0,
indicating that developing GG dendrites are not well suited
to be modeled as an equivalent cylinder at any age. We
concluded that the rapid decline in dendritic trunk parameter in the first 300 Fm of the dendritic tree was due to the
large amount of tapering found in the first- and secondorder dendrites at four age groups. Because tapering is
small in terminal branches and in the higher branch order
dendrites, nonuniform lengths of branches may be more
responsible for the observed decrease in dendritic trunk
parameter beyond 400 Fm. However, tapering in the terminal branches could be underestimated because of the
accuracy to which we could resolve diameters less than 0.5
pm. Furthermore,tapering could be influenced by the way
of estimation itself. The absolute amount of taper, as used
here, is influenced by the decrease in average branch
diameter that occurs with increasing distance (branch
order); therefore, tapering may be underestimated (Rose et
al., 1985). To avoid this problem, different formulas have
been used to normalize taper values (i.e., by the branch
start diameter) that in some cases overestimate the degree
of expansion, or overestimate the degree of taper as well
(see Rose et al., 1985).
In the present study, using measurements of absolute
taper, we found that dendritic tapering was most marked
+
on first-order dendritic branches, as previously demonstrated for cervical motoneurons of the cat using normalized tapering (Rose et al., 1985) or absolute tapering
(Cameron et al., 1991a). In addition, values of absolute
tapering allow data to be compared between different motor
pools. Further investigations are now required to explore
the electrotonic properties of developing mammalian motoneurons, and to determine whether dendritic growth is
regulated to maintain a constant electrotonic length, or
whether other mechanisms are employed to maintain synaptic efficacy of even the most remote synapses.
ACKNOWLEDGMENTS
The authors thank Mr. Fang He for his technical assistance in reconstructing the motoneurons with the Eutectic
Neuron Tracing System and Mr. Darrell Henze for his
comments on the manuscript. This work was supported by
a grant from the SIDS program of the NICHD (HD 22703).
P.A.N.-A. was supported by a postdoctoral fellowship from
the Ministerio de Educacion y Ciencia, Spain.
LITERATURE CITED
Arvidsson, U.,J. Svedlund, P.-A. Lagerback, and S. Cullheim (1987) An
ultrastructural study of the synaptology of y-motoneurons during the
postnatal development in the cat. Dev. Brain Res. 37:303-312.
Barrett, J.N. (1975) Motoneuron dendrites: Role in synaptic integration.
Fed. Proc. 34:1398-1407.
142
Barrett, J.N., and W.E. Crill (1974) Specific membrane properties of cat
motoneurons. J. Physiol. (London)239:301-324.
Berry, M., and R. Flinn (1984)Vertex analysis of Purkinje cell dendritic trees
in the cerebellum of the rat. Proc. R. Soc. London [Biol.] 221:321-348.
Birinyi, A,, M. Antal, E. Wolf, and G. Szekely (1992) The extent of the
dendritic tree and the number of synapses in the frog motoneuron. Eur.
J. Neurosci. 4:1003-1012.
Brozanski, B.S., R.D. Guthrie, E.A. Volk, and W.E. Cameron (1989) Postnatal growth of genioglossal motoneurons. Pediatr. Pulmonol. 7:133-139.
Burke, R.E. (1981) Motor units: Anatomy, physiology and functional
organization.In V.B. Brooks (ed): Handbook of Physiology. Section I.
The Nervous System, Vol. 11, Motor System. Washington, DC: American
Physiological Society, pp. 345422.
Burke, R.E. (1987) Synaptic efficacy and the control of neuronal inputoutput relations. TINS 10:4245.
Cameron, W.E., and F. He (1989) Morphology of developing motoneurons
innervating the medial gastrocnemius of the cat. Dev. Brain Res.
49.253-263.
Cameron, W.E., D.B. Averill, andA.J. Berger (1985) Quantitative analysis of
the dendrites of cat phrenic motoneurons stained intracellularly with
horseradish peroxidase. J. Comp. Neurol. 231:91-101.
Cameron, W.E., F. He, P. Kalipatnapu, J.S. Jodkowski, and R.D. Guthrie
(1991a) Morphometric analysis of phrenic motoneurons in the cat during
postnatal development. J. Comp. Neurol. 314:l-14.
Cameron, W.E., J.S. Jodkowski, F. He, and R.D. Guthrie (1991b) Electrophysiological properties of developing phrenic motoneurons in the cat. J.
Neurophysiol. 3t671-679.
Capowski, J.J. (19891 Computer Techniques in Neuroanatomy. New York:
Plenum Press.
Cullheim, S., J.W. Fleshman, L.L. Glenn, and R.E. Burke (1987) Membrane
area and dendritic structure in type-identified triceps surae alpha
motoneurons. J. Comp. Neurol. 25568-81.
He, F., P.A. Nubez-Abades, and W.E. Cameron (1992) Morphology of
developing hypoglossal motoneurons studied in slice preparations of the
rat brainstem. SOC.
Neurosci. Abstr. 18:1049.
Honvitz, B. (1981) Neuronal plasticity: How changes in dendritic architecture can affect the spread of postsynaptic potentials. Brain Res. 224.412418.
Larkman, A.U. (1991a) Dendritic morphology of pyramidal neurones of the
visual cortex of the rat: I. Branching Patterns. J. Comp. Neurol.
306: 30 7-3 19.
Larkman, A.U. (1991b) Dendritic morphology of pyramidal neurones of the
visual cortex of the rat: 11. Parameter correlations. J. Comp. Neurol.
306t320-331.
Larkman, A.U., G. Major, K.J. Stratford, and J.J.B. Jack (1992) Dendritic
morphology of pyramidal neurones of the visual cortex of the rat. I V
Electrical geometry. J. Comp. Neurol. 323:137-152.
Mazza, E., P.A. Nufiez-Ahades, J.M. Spielmann, and W.E. Cameron (19921
Anatomical and eledrotonic coupling in developing genioglossal motoneurons of the rat. Brain Res. 598t127-137.
Mellstrom, A., and S. Skoglund (1969) Postnatal changes of neuron volume
and nuclear volume in lamina IX of C7 and L7. Acta Physiol. Scand.
(Suppl.) 3 3 1 3 - 4 9 .
Nubez-Abades, P.A., J.M. Spielmann, and W.E. Cameron (1992) Postnatal
changes in specific membrane resistance of rat brainstem motoneurons.
Soc. Neurosci. Abstr. 18:1049.
Nubez-Abades, P.A., J.M. Spielmann, G. Barrionuevo, and W.E. Cameron
(1993) In vitro electrophysiology of developinggenioglossal motoneurons
in the rat. J. Neurophysiol. 70:1401-1411.
P.A. NUNEZ-ABADES AND W.E. CAMERON
Nunez-Abades, P.A., F. He, G. Barrionuevo, and W.E. Cameron (1994)
Morphology of developing rat genioglossal motoneurons studied in vitro:
Changes in the dendritic length, branching pattern, and spatial distribution. J. Comp. Neurol. 339:401420.
Rall, W. (1959) Branching dendritic trees and motoneuron membrane
resistivity. Exp. Neurol. 1:491-527.
Rall, W. (1977) Core conductor theory and cable properties of neurons. In
E.R. Kandel (ed): Handbook of Physiology. Section I. The Nervous
System, Part 1, Vol. I, Cell Biology of Neurons. Washington, DC:
American Physiological Society, pp. 39-97.
Rall, W., R.E. Burke, W.R. Holmes, J.J.B. Jack, S.J. Redman, and I. Segev
(1992) Matching dendritic neuron models to experimental data. Physiol.
Rev. 72:S159-S186.
Ramirez, V., and B. Ulfhake (1991) Postnatal development of cat hind limb
motoneurons supplying the intrinsic muscles of the foot sole. Dev. Brain
Res. 62189-202.
Rose, D., N. Larnicol, and B. Duron (19841 An HRP study of the cat’s spinal
respiratory motoneurones during postnatal development. Exp. Brain
Res. 56:458467.
Rose, P.K., S.A. Kierstead, and S.J.Vanner (1985) A quantitative analysis of
the geometry of cat motoneurons innervating neck and shoulder muscles.
J. Comp. Neurol. 239939-107.
Sato, M., N. Mizuno, and A. Konishi (1978) Postnatal differentiation of cell
body volumes of spinal motoneurons innervating slow-twitch and fasttwitch muscles. J. Comp. Neurol. 17537-36.
Schade, J.P., and A. Van Harreveld (1961) Volume distribution of moto- and
interneurons in the peroneus-tibialis neuron pool in the cat, J. Comp.
Neurol. 11 7:387-398.
Spielmann, J.M., F. He, and W.E. Cameron (1992) Changes in membrane
properties of developing rat phrenic motoneurons studied in vitro. Soc.
Neurosci. Abstr. 18:440.4.
Tatton, W.G., and E. Theriault (19881Postnatal growth of medial gastrocnemius motoneurons in the kitten. Dev. Brain Res. 43.191-206.
Tatton, W.G., M. Hay, D. McCrea, and I.C. Bruce (1983) Postnatal dendritic
development in motoneurons: Evaluation by a Monte Carlo technique.
Exp. Brain Res. 52:461465.
Ulfhake, B., and J.-0.Kellerth (1981) A quantitative light microscopic study
of the dendrites of cat spinal a-motoneurons after intracellular staining
with horseradish peroxidase. J. Comp. Neurol. 202571-583.
Ulfhake, B., and J.-0.Kellerth (1983) A quantitative morphological study of
HRP-labeled cat a-motoneurons supplying different hindlimh muscles.
Brain Res. 264:l-19.
Ulfbake, B., and J.-0.Kellerth (1984) Electrophysiological and morphological measurements in cat gastrocnemius and soleus a-motoneurons.
Brain Res. 307:167-179.
Ulfhake, B., and S. Cullheim (1988) Postnatal development of cat hind limb
motoneurons. 111: Changes in size of motoneurons supplying the triceps
surae muscle. J. Comp. Neurol. 2781103-120.
Verwer, R.W.H., and J. Van Pelt (1985) Topological analysis of binary tree
structures when occasional multifurcations occur. Bull. Math. Biol.
47t305-316.
Webber, C.L., and K. Pleschka (1976) Structural and functional characteristics of individual phrenic motoneurons. Pflugers Arch. 364t113-121.
Wolf, E., A. Birinyi, and G. Szekely (1992) Simulation of the effect of
synapses: The significance of the dendritic diameter in impulse propagation. Eur. J. Neurosci. 4t1013-1021.
Zwaagstra, B. and D. Kernell (1980) Sizes of soma and stem dendrites in
intracellularly labeled alpha-motoneurons of the cat. Brain Res. 204.295309.