THE JOURNAL OF COMPARATIVE NEUROLOGY 412:186–192 (1999)

Effect of Nerve Crush on Perikaryal

Number and Volume of Neurons in Adult
Rat Dorsal Root Ganglion
J. DEGN,1* T. TANDRUP,1,2 AND J. JAKOBSEN1
1Department of Neurology, University Hospital of Aarhus, DK-8000 Aarhus C, Denmark
2Stereological Research Laboratory, University of Aarhus, DK-8000 Aarhus C, Denmark

ABSTRACT
Assumption-free stereological methods were applied to assess the effect of nerve crush on
perikaryal number and mean volume of neuronal subpopulations in adult rat dorsal root
ganglion (DRG). The L5 spinal nerve of 20 Wistar rats was crushed approximately 7 mm distal
to the DRG, and the contralateral spinal nerve and DRG were left intact and used as controls.
After four, 15, 45, and 120 days, the rats were killed, and the tissue was fixed and processed for
subsequent preparation of 30-µm-thick sections. Estimates of neuron number were obtained
with the optical fractionator technique and estimates of the mean perikaryal volume with the
vertical planar rotator principle. Perikaryal loss was progressive during the early study
period but stabilized 45 days after nerve injury. The mean number (n) of all neurons in intact
L5 DRG was 16,400 (S.D. ⫽ 2,000). The loss of perikarya was 16% (P ⬍ 0.05) after four days,
15% (P ⬍ 0.05) after 15 days, 30% (P ⫽ 0.059) after 45 days, and 34% (P ⬍ 0.05) after 120 days.
B cells were lost at an earlier time than were A cells, and the B cell loss was more pronounced
(39% vs. 22%, respectively, after 120 days). For A cells, the mean perikaryal volume was
initially reduced but was normalized at the end of the study. Distributions of perikaryal
volume showed that the curves of both A and B cells were uniformly displaced toward smaller
values 15 and 45 days after injury. Neuronal loss caused by crush seems similar to that seen in
rats exposed to permanent axotomy (Vestergaard et al. [1997] J Comp Neurol 388:307–312) at
the same location, indicating that survival of perikarya is not dependent on possibility for
fiber growth. J. Comp. Neurol. 412:186–192, 1999. r 1999 Wiley-Liss, Inc.

Indexing terms: neuronal degeneration; nerve injury; stereology; cell size

Axotomy to provoke changes in dorsal root ganglion
(DRG) neuron morphology is an important model in stud-
ies of regeneration and survival of injured neurons during
neuroprotective treatment. Although the model of axon
injury has many advantages, a major obstacle in treat-
ment studies has been the lack of assumption-free tech-
niques to estimate cell loss. Most studies of axotomy have
reported a 15–50% neuronal loss in DRG. The most
pronounced loss has been found after proximal injury of
axons of neonatal animals. Reports in the literature con-
cerning the loss of DRG neurons after nerve injury are,
however, variable. One of the reasons for this large varia-
tion is the use of different counting methods (Coggeshall,
1992), differences of age (neonatal vs. adult), gender,
severity, and type of injury. Furthermore, the opportunity
for regeneration, the proximity of the injury to the peri-
karyon (proximal vs. peripheral), and time after injury
could affect the estimates.
By using stereological techniques, it is now possible to
eliminate some of the biases and methodological problems
of previous studies concerning neuron number and size. A
recent study (Vestergaard et al., 1997) using stereological
methods showed that perikaryal volume of L5 DRG neu-
rons was reduced as early as 4 days after permanent
axotomy. Cell bodies were lost after 15 days, predomi-
nantly due to loss of B cells. This loss was progressive
during the 45 days of study and was associated with
shrinkage of neuronal perikarya. In the present study, the
effect of nonpermanent axotomy after a crush lesion
(Bridge et al., 1994) was examined. The aims of the project
were to study the temporal effects of nerve crush on total
perikaryal number and mean perikaryal volume of neuro-
nal subpopulations in adult rat DRG and compare the
effect of crush with data from previous studies of perma-
Grant sponsor: University and County of Aarhus, Denmark.
*Correspondence to: J. Degn, Department of Neurology, University
Hospital of Aarhus, DK-8000 Aarhus C, Denmark. E-mail: degn@mailme.dk
Received 27 October 1998; Revised 5 April 1999; Accepted 29 April 1999

r 1999 WILEY-LISS, INC.

EFFECT OF NERVE CRUSH ON RAT DRG 187

nent axotomy to estimate whether neuronal survival is

dependent on fiber growth.

MATERIALS AND METHODS

Animals
Twenty 10-week-old inbred male Wistar rats (290–330 g)
were allocated into four groups of five animals each. All
rats were cared for and surgically handled in accordance
with the guidelines specified in the Federation of Euro-
pean Laboratory Animal Science Associations recommen-
dations. The animals were anesthetized with a mixture of
diazepam (5 mg/kg) and pentobarbital (25 mg/kg) injected
intraperitoneally. Laminectomy was performed, and the
L5 spinal nerve was freed from surrounding tissues with a
J-shaped glass tube. A double crush (2 ⫻ 30 seconds) was
applied approximately 7 mm distal to the corresponding
L5 DRG by using a fine smooth-tipped 0.5-mm S.W. Inox
Castroviejo forceps (model Tübingen, Amann Medizintech-
nic, Germany) with a stopping pin to standardize the
lesion. The side for the lesion was chosen at random, and
the contralateral DRG served as the control. Buprenorphi-
num (4 ⫻ 0.1 mg) was admistered during the postoperative
phase.
Fixation and tissue preparation
Four, 15, 45, and 120 days after crush, the rats were
anesthetized and tissue fixed by vascular perfusion with a
4% (w/v) phosphate buffered glutaraldehyde (pH 7.40, 0.08
M) solution through the ascending aorta after a short
prerinse with Tyrode’s buffer (pH 7.76). Following identifi-
cation of the L4 and L5 spinal nerves, the L5 DRGs on both
sides were removed, including a segment of their spinal
nerves a few millimeters distal to the site of the crush.
The ganglia were placed in agar (7%) and rotated at
random around the long, vertical axis (Baddeley et al.,
1986). Tissues were dehydrated through ascending concen-
trations of ethanol solutions (70%, 96%, 96%, 99%, and
99% for 1 hour each) and embedded in glycolmethacrylate
(Technovit 7100, Heraeus Kulzer GmbH, Wehrheim, Ger-
many). Thirty-micrometer-thick sections were cut parellel
to the vertical axis of the ganglia with a Reichert-Jung
2050 Supercut microtome by using Ralph glass knives.
Using a random starting point, every third section in the
series was sampled for staining with cresyl violet acetate
(0.25 g%). Spinal nerves were postfixed with osmium
tetroxide (1%) for 2 hours and embedded in Epon (TAAB
812, Medium, Taab Laboratories Equipment Ltd., United
Fig. 1. Intact (A) and chromatolytic (B) A and B cells 15 days after
Kingdom). One-micrometer cross sections were cut with an spinal nerve injury. In A, the nuclei of one large A neuron (bottom) and
ultramicrotome (LKB Ultrotome III) and stained with a four small B neurons (top) are in focus. Notice the well-defined Nissl
solution of toluidinblue/sodiumborate (1% w/v). granules and the light nucleus with one large, dark, central nucleolus
in the A neurons. The nuclei of B neurons contain multiple nucleoli
Cell type identification located peripherally in the nucleus. In B, the central chromatolysis,
the peripheral mantle of Nissl substance, and the caps around the
The ganglion cells were visualized in three dimensions crenated eccentric displaced nuclei are clear. Scale bar ⫽ 20 µm.
by optical sectioning, which improved the identification
markedly. Two types of neurons could be identified in the
DRG (Andres, 1961; Lieberman, 1976; Rambourg et al., nucleus was light, very often with multiple nucleoli located
1983; Tandrup, 1993). The large A cells (Fig. 1) were peripherally. Only 2% of neurons could not be placed into
dominated by well-defined Nissl granules, with an un- either A or B cell categories.
stained Nissl-free axon hillock region. The nucleus was
lightly stained and usually contained one large, dark, Counting procedures
central nucleolus. In contrast to these characteristic fea- A microscope was connected to a color camera linked to a
tures, the small B cells were more heterogeneous, some personal computer by using morphometric software
with homogeneously stained dark cytoplasm, others with a (C.A.S.T.-Gridt, Olympus DK A/S, Albertslund, Denmark).
light halo in the center of the cell. In all the B cells, the The image of the sections was projected to a computer
188 J. DEGN ET AL.

screen with superimposed counting frames. Systematic,

uniformly random sampling of counting fields was per-
formed by using a stepping motor and a microcator (Dr.
Johannes Heidenhouin, Germany) attached to the stage of
the microscope.
Estimates of neuron number were obtained by using the
optical fractionator (West et al., 1991) and optical disectors
(Sterio, 1984; Gundersen, 1986; Gundersen et al., 1988;
Braendgaard et al., 1990). The number of perikarya (n)
was estimated by multiplying counted nuclei (Q⫺) within
the disectors with the inverse sampling fractions:

A H
N⫽3· 兺Q ⫺
· ·
a h

where a/A is the area fraction within the disectors and h/H
is the fraction of the section thickness used. Area of
counting frames was 3,500 µm2 for A cells and 1,680 µm2
for B cells. The step length was 冑A ⫽ 185 µm, and disector
height measured with the microcator was h ⫽ 15 µm. The
mean section thickness was 23 µm (H) as measured with
the microcator in every fifth disector sample, at a total
magnification of 2,918⫻. All measurements were per-
formed blindly.
A counting frame with a top plane a few micrometers
below the surface and a bottom plane 15 µm further below
defined the limits of the optical disectors. The nucleus was
chosen as the counting unit because all DRG neurons have
exactly one nucleus. All neuronal nuclei coming into focus
and belonging to the counting frames (Gundersen, 1977)
were counted (Q⫺) because the plane of focus was lowered
15 µm through the transparent section by using a short
focus depth objective (PlanApo 60⫻, numerical aperture ⫽
1.40 oil). Nuclei in focus at the top plane were not included,
as opposed to nuclei at the bottom plane.
For volume estimation, cells should have equal sam-
pling probability, and measurements in the sampled cells
have to be performed in either isotropic or vertical planes.
This was ensured by the use of vertical sections (Baddeley
et al., 1986).
The mean volume of the perikarya was estimated by
using the vertical planar rotator principle (VPR; Jensen
and Gundersen, 1993) incorporated in the semiautomatic
software C.A.S.T.-Gridt. The VPR is an unbiased local
volume estimator based on two-dimensional information
obtained from a single optical plane through the neuron.
The VPR provides an estimate of perikarya size from
measurements of intersections between the cell boundary
and three lines perpendicular to the vertical axis. All
measurements were performed from the unique point (the
nucleolus) to the cell borders in the optical plane, where
the nucleolus is delineated most clearly. If several nucleoli
were present in a cell, the measurements were made from
the largest of the nucleoli. The magnification used was
1,756⫻.
Statistics
Because of markedly right-skewed distributions, loga-
rithmic transformation was applied to all data. For num-
ber and volume, the statistical significance of group mean
differences between crushed and control DRG neurons was
evaluated by using Student’s paired t-test with a 5% limit
of significance.
Fig. 2. Cross sections of the L5 spinal nerve a few millimeters
distal to the nerve crush after 4 (B) and 120 (C) days. A: Control. Scale
bar ⫽ 50 µm.
RESULTS
The rats suffered little from the operative procedures
and had an initial minor mean weight deficit of 50 g
(30–70 g) compared with the control rats.
An average of 84 (54–108) A cells and 133 (92–193)
B cells were counted per intact DRG. Seventy-three (51–
98) A and 97 (57–137) B cells were counted per crushed
DRG. The mean thickness of the sections (H) was 23
(S.D. ⫽ 2) µm.
The DRG neurons changed morphologically after the
nerve crush (Fig. 1; Nathaniel and Nathaniel, 1973). The
most conspicuous signs were chromatolysis with periph-
eral displacement of a crenated nucleus and proliferation
of satellite cells especially around large A cells. Frequently,
the nuclei showed small indentations of the nuclear mem-
brane. After 120 days, accumulation of perineuronal satel-
lite cells (Nageotte bodies) was observed (Nageotte, 1922).
In the spinal nerve distal to the crush, the predominant
features were Wallerian degeneration after 4, 15, and 45
days (Fig. 2). Regenerating clusters of small axons were
observed after 120 days.
Distributions of perikaryal volume were bimodal (A and
B cells). There seemed to have been a relatively large cell
loss of B cells 45 and 120 days after crush, with left
displacement of both A and B cell curves 15 and 45 days,
respectively, after injury (Fig. 3). The mean total number
of neurons in intact L5 DRG was 16,400 (S.D. ⫽ 2,000),
23% of the cell bodies being A cells. There was a progres-
sive and significant loss of DRG neurons during the
experimental period. The individual loss of DRG neurons
EFFECT OF NERVE CRUSH ON RAT DRG 189

Fig. 3. Frequency distributions of mean perikaryal volume of A (the two right curves in each panel)
and B (the two left curves in each panel) dorsal root ganglia bodies four, 15, 45, and 120 days after injury
(solid lines) of the L5 spinal nerves. Compare with intact (dotted lines) nerves. Values are mean numbers
and error bars indicate S.E.M.

after crush of the L5 spinal nerve in adult rats was TABLE 1. Perikaryal Number and Loss of Neurons of the L5 DRG
After Crush1
115–8,900, with an average reduction of 2,700 (S.E.M. ⫽
950) four days after the injury, 2,500 (S.E.M. ⫽ 800) after Days Intact (S.D.) Crush (S.D.) Difference (S.E.M.)
15 days, 4,500 (S.E.M. ⫽ 1,700) after 45 days, and 5,700 A neurons
(S.E.M. ⫽ 850) after 120 days. 4 4,134 (548) 3,328 (803) 806 (508)
The relative mean loss of all perikarya was 16% (P ⬍ 15 3,869 (567) 3,612 (374) 257 (135)
0.05) four days after the crush, 15% (P ⬍ 0.05) after 15 45
120
3,336 (752)
3,848 (482)
3,204 (875)
3,003 (565)
132 (171)
845 (198)*
days, 30% (P ⫽ 0.059) after 45 days, and 34% (P ⬍ 0.05)
B neurons
after 120 days (Table 1, Fig. 4). The loss of A cells (22%) 4 12,699 (947) 10,820 (2,257) 1,879 (963)
was only significant 120 days after the injury, whereas 15 12,767 (2,661) 10,598 (1,482) 2,169 (764)*
45 11,588 (2,181) 7,193 (1,566) 4,395 (1,620)
there was an early and statistically significant loss of B 120 12,879 (1,866) 7,881 (2,113) 4,998 (915)*
cells (17%) after 15 days. Also, the neuronal loss was larger
Total
for B cells (39%) than for A cells (22%) after 120 days (Ta-
4 17,011 (1,168) 14,355 (2,168) 2,656 (954)*
ble 1). 15 16,785 (2,337) 14,333 (1,366) 2,451 (791)*
Geometric mean perikaryal volume of intact A cells was 45 15,070 (2,641) 10,567 (1,955) 4,504 (1,720)
120 16,782 (1,757) 11,059 (2,087) 5,722 (860)*
63,200 µm3 (S.D. ⫽ 11,600) and 10,800 µm3 (S.D. ⫽ 1,100)
1Perikaryal number and loss of neurons of the L5 dorsal root ganglion (DRG) 4, 15, 45,
of intact B cells. Perikaryal volume of A cells was un- and 120 days after unilateral crush of its spinal nerve, values being means (S.D.s and
changed four days after the crush, diminished by 28% (P ⬍ S.E.M.s). Asterisk indicates a statistically significant difference of cell numbers at the
0.05) after 15 days, by 31% (P ⬍ 0.05) after 45 days, and 0.05 level between the intact and the crushed side in the same rats.

returned to normal after 120 days. Perikaryal volume of B

cells was diminished by 9% (n.s.) after four days, by 27% Even though the perikarya of the large A cells went
(P ⬍ 0.05) after 15 days, by 25% (P ⬍ 0.05) after 45 days, through the most pronounced degree of shrinkage, these
and by 19% (P ⬍ 0.05) after 120 days (Table 2). cells appeared to restore their original mean volume
190 J. DEGN ET AL.

from those in the present study with regard to the type of

injury. Technically, ganglia were bisected before embed-
ding in that study, whereas whole ganglia were used in the
present study. However, in both studies, stereological
techniques were applied. In that study, the combined
perikaryal loss of A and B cells was 6%, 19%, 22%, and 35%
after four, eight, 15, and 45 days, respectively, and the
reduction of perikaryal volume (both A and B cells) was
33%, 34%, 29%, and 25% respectively. In both studies
(Vestergaard et al., 1997; present study), relatively more B
cells were lost, whereas A cell perikarya were more se-
verely diminished. Taken together, these findings indicate
that a proximal lesion leads to more cell loss than a more
peripheral lesion, whereas neuronal survival seems inde-
pendent of the possibility for fiber growth.
A recent study by Dreetz et al. (personal communication)
showed that p75-knockout mice lacked 39% A cells and
57% B cells. Volume of A cells was 30% less than that of
controls, and B cells were reduced by 18%. Small B cells
have been shown to respond early and more intensely with
Fig. 4. Number of cell bodies of A cells (circles) and B cells chromatolysis than do larger A cells (Lieberman, 1971). B
(triangles) in the intact (In) and contralateral dorsal root ganglion cells have high-affinity receptors for nerve growth factor
four, 15, 45, and 120 days after spinal nerve crush (Cr). Bars are group (NGF), whereas A cells bind neurotrophin 3 (Richardson et
mean values. al., 1986; Sebert and Shooter, 1993). In accordance with
these observations, the present study emphasizes the
TABLE 2. Perikaryal Volume and Shrinkage of Neurons of the L5 DRG different vulnerability of the various neuronal subpopula-
After Crush1 tions in rat DRG and points to a possible role of neuro-
Days Intact (S.D.) Crush (S.D.) Difference (S.E.M.) trophic factor for the differential vulnerability.
Several investigators (Ygge et al., 1981) have pointed
A neurons
4 58,962 (11,226) 59,760 (7,744) ⫹798 (4,134)
out that the total pool of thoracic DRG neurons is distrib-
15 70,382 (14,070) 50,841 (12,486) 19,541 (3,326)* uted in a heterogeneous fashion between adjacent DRGs
45 68,886 (8,608) 47,860 (6,619) 21,026 (3,035)* (e.g., between the T2 and T3 DRGs). For this reason,
120 54,531 (4,347) 53,038 (7,734) 1,493 (2,219)
counting of all neurons in the L4, L5, and L6 DRGs (the
B neurons
main contributors to the sciatic nerve) has been proposed,
4 10,581 (1,226) 9,634 (1,755) 947 (741)
15 10,863 (1,004) 7,947 (2,307) 2,916 (845)* and the use of the combined number has been proposed to
45 10,924 (1,182) 8,176 (1,058) 2,748 (296)* avoid irregular spatial distribution of perikarya to bias the
120 10,822 (1,470) 8,784 (1,495) 2,038 (499)*
comparison (Serbert and Shooter, 1993). In the present
1Perikaryal volume (µm3 ) and shrinkage of neurons of the L5 dorsal root ganglia (DRG)
study, the mean total number of neurons in intact L5 DRG
4, 15, 45, and 120 days after unilateral crush of its spinal nerve, values being geometric
means (S.D.s and S.E.M.s). Asterisk indicates a statistically significant difference of cell was found to be rather constant (16,400, S.D. ⫽ 2,000). In
numbers at the 0.05 level between the intact and the crushed side in the same rats. addition, four other studies (Tandrup, 1993, 1995; Tan-
drup and Braendgaard, 1994; Vestergaard et al., 1997)
using the same unbiased stereological techniques have
earlier than B cells. In contrast, the volume of B cells still estimated the number to be in the range of 15,700–17,900.
was reduced by 19% (P ⬍ 0.05) after 120 days. Obviously, the number of L5 DRG neurons is rather
constant. The present study did not demonstrate any
side-to-side asymmetry.
DISCUSSION Not all L4-L6 DRG neurons project into the sciatic
nerve. Devor et al. (1985) estimated that fewer than 50% of
Most studies of axotomy (mainly of the peripheral type)
have reported a 15–50% neuronal loss of DRG cell bodies all neurons of the ganglia L4–L6 do. Therefore, peripheral
(Ranson, 1906; Cavanaugh, 1951; Aldskogius and Risling, axotomy at the level of the sciatic nerve does not affect all
1981; Risling et al., 1983a,b; Ygge and Aldskogius, 1984; DRG neurons, and for this reason, a direct comparison
Arvidsson et al., 1986; Rich et al., 1987; Schmalbruch, between peripheral and proximal axotomy is difficult.
1987a; Himes and Tessler, 1989; Baranowski et al., 1994). The principal findings of the present study confirm that
The most pronounced loss has been found after proximal DRG neurons are lost when the adult spinal nerve is
injury of axons of neonatal animals (Himes and Tessler, crushed. The estimates are free of assumptions about
1989; Ygge, 1989). neuron size, shape, and orientation. Furthermore, the
Recent studies based on assumption-free stereological techniques are not dependent on shrinkage or swelling of
methods have indicated that the loss of DRG cell bodies the neurons or the surrounding tissues during prepara-
after peripheral axotomy at the level of the sciatic nerve is tion. The principles of volume estimation also are assump-
14–30% (Eriksson, 1997; Tandrup et al., personal commu- tion free, but absolute estimates are dependent on prepara-
nication), whereas more proximal axotomy leads to a 35% tion artifacts. Because the changes observed represent a
loss (Vestergaard et al., 1997). The results obtained by combination of loss of cell bodies and reduction of peri-
Vestergaard et al. (1997) on permanent axotomy only differ karyal volume, interpretation of the displacements of the
EFFECT OF NERVE CRUSH ON RAT DRG
distributions of perikaryal volume is difficult. However, in
Figure 3, the left displacement of the B cell curves 15 and
45 days after injury indicates shrinkage of cells, whereas
the two other panels attest loss of cells only. To show the
difficulties of interpretation, the left–downward displace-
ment of the B cell curve four days after crush is mentioned.
One explanation could be a selective loss of large B cells.
Also, primary volume reduction could occur in combination
with cell loss of medium-sized B cells. Eventually, small B
cells could increase their size while large B cells reduce
their volume dramatically. We suggest that the displace-
ments occur in the vertical and horizontal directions
rather than by means of pivot shift. This interpretation is
supported by the horizontal curve migration initially (15
days) toward smaller and later (120 days) toward origi-
nally volumes.
The more pronounced neuronal loss found after proxi-
mal injury of axons suggests that reduction in volume is
dependent on the amount of axoplasm lost during axotomy.
Because neuronal volume consists of both perikaryal and
axonal contributions, estimation of total cell volume is
complex. The importance of the relatively drastic reduc-
tion in Schwann-cell-derived support should be taken into
consideration when dealing with proximal injury. In future
studies, application of gene-modified Schwann cells produc-
ing neurotrophic factors (e.g., NGF) is a promising pros-
pect to examine what influences the amount of cell loss.
It appears that the perikaryal loss was progressive
during the early study period but stabilized 45 days after
nerve injury (B cells) and that predominantly B cells were
lost. Also, B cells were lost earlier than were A cells.
Because cell bodies were lost and individual perikaryal
volume was reduced simultaneously, interpretation of the
displacements of the distributions of perikaryal volume is
difficult. However, the mean perikaryal volume was ini-
tially reduced but regained dimensions toward the end of
the study. The neuronal loss caused by a crush seems
similar to the loss in studies of permanent axotomy,
indicating that survival of perikarya is not dependent on
the possibility for fiber growth after axotomy.
ACKNOWLEDGMENTS
This study was made possible by the financial and
technical support from the Department of Neurology,
Aarhus University Hospital and a scholarship from the
County of Aarhus. We express gratitude for unfailing
expert assistance by laboratory technician Kirsten Kand-
borg. We are grateful to Jens Nyengaard, M.D., and Hans
Jørgen Gundersen, M.D., for rewarding discussions about
stereological issues.
LITERATURE CITED
Aldskogius H, Risling M. 1981. Effect of sciatic neurectomy on neuronal
number and size distribution in the L7 ganglion of kittens. Exp Neurol
74:597–604.
Andres KH. 1961. Untersuchungen über den feinbau von spinalganglien.
Zeitschr Zellforsch 55:1–48.
Arvidsson J, Ygge J, Grant G. 1986. Cell loss in lumbar dorsal root ganglia
and transganglionic degeneration after sciatic nerve resection in the
rat. Brain Res 373:15–21.
191
Baddeley AJ, Gundersen HJG, Cruz-Orive LM. 1986. Estimation of surface
area from vertical sections. J Microsc 142:259–276.
Baranowski AP, Priestley JV, McMahon SB. 1994. The consequence of
delayed versus immediate nerve repair on the properties of regenerat-
ing sensory nerve fibres in the adult rat. Neurosci Lett 168:197–200.
Braendgaard H, Evans SM, Howard CV, Gundersen HJG. 1990. The total
number of neurons in the human neocortex unbiasedly estimated using
optical disectors. J Microsc 157:285–304.
Bridge PM, Ball DJ, Mackinnon SE, Nakao Y, Brandt K, Hunter DA, Hertl
C. 1994. Nerve crush injuries—a model for axonotmesis. Exp Neurol
127:284–290.
Cavanaugh MV. 1951. Quantitative effects of the peripheral innervation
area on nerve and spinal ganglion cells. J Comp Neurol 94:181–219.
Coggeshall RE. 1992. A consideration of neural counting methods. Trends
Neurosci 15:9–13.
Devor M, Govrin Lippmann R, Frank I, Raber P. 1985. Proliferation of
primary sensory neurons in adult rat dorsal root ganglion and the
kinetics of retrograde cell loss after sciatic nerve section. Somatosens
Res 3:139–167.
Eriksson P. 1997. The effects of neurotrophic substances on primary
sensory neurons following peripheral nerve injury [thesis]. Stockholm:
Department of Neuroscience, Karolinska Instituttet.
Gundersen HJG. 1977. Notes on the estimation of numerical density of
arbitrary profiles: the edge effect. J Microsc 111:219–223.
Gundersen HJG. 1986. Stereology of arbitrary particles. J Microsc 143:
3–45.
Gundersen HJG, Bagger P, Bendtsen TF, Evans SM, Korbo L, Marcussen
N, Møller A, Nielsen K, Nyengaard JR, Pakkenberg B, Sørensen FB,
Vesterby A, West MJ. 1988. The new stereological tools: Disector,
fractionator, nucleator and point sampled intercepts and their use in
pathological research and diagnosis. Acta Pathol Microbiol Immunol
Scand 96:857–881.
Himes BT, Tessler A. 1989. Death of some dorsal root ganglion neurons and
plasticity of others following sciatic nerve section in adult and neonatal
rats. J Comp Neurol 284:215–230.
Jensen EBV, Gundersen HJG. 1993. The rotator. J Microsc 170:35–44.
Lieberman AR. 1971. The axon reaction. Int Rev Neurobiol 14:49–124.
Lieberman AR. 1976. Sensory ganglia. In: London DN, editor. The periph-
eral nerve. London: Chapman and Hall. p 188–278.
Nageotte J. 1922. L’organisation de la matière dans ses rapports avec la
vie—études d’anatomie générale et de morphologie expérimentale sur
le tissu conjonctif et le nerf. Paris: Librairie Félix Alcan.
Nathaniel EJH, Nathaniel DR. 1973. Electron microscopic studies of spinal
ganglion cells following crushing of dorsal roots in adult rat. J Ultra-
struct Res 45:168–182.
Rambourg AY, Clermont T, Beaudet A. 1983. Ultrastructural features of six
types of neurons in rat dorsal root ganglia. J Neurocytol 12:47–66.
Ranson SW. 1906. Retrograde degeneration in the spinal nerves. J Comp
Neurol Psychol 16:125–153.
Rich KM, Luszczynski JR, Osborne PA, Johnson EM. 1987. Nerve growth
factor protects adult sensory neurons from cell death and atrophy
caused by nerve injury. J Neurocytol 16:261–268.
Richardson PM, Issa VM, Riopelle RJ. 1986. Distribution of neuronal
receptors for nerve growth factor in the rat. J Neurosci 6:2312–2321.
Risling M, Aldskogius H, Hildebrand C. 1983a. Effekts of sciatic nerve
crush on L7 spinal root and dorsal root ganglia in kittens. Exp Neurol
79:176–187.
Risling M, Aldskogius H, Hildebrand C, Remahl S. 1983b. Effekts of sciatic
nerve resection on L7 spinal roots and dorsal root ganglia in adult cats.
Exp Neurol 82:568–580.
Schmalbruch H. 1987a. Loss of sensory neurons after sciatic nerve section
in the rat. Anat Rec 219:323–329.
Schmalbruch H. 1987b. The number of neurons in dorsal root ganglia
L4–L6 of the rat. Anat Rec 219:315–322.
Sebert ME, Shooter EM. 1993. Expression of mRNA for neurotrophic
factors and their receptors in the rat dorsal root ganglion and sciatic
nerve following nerve injury. J Neurosci Res 36:357–367.
Sterio DC. 1984. The unbiased estimation of number and sizes of arbitrary
particles using the disector. J Microsc 134:127–136.
Tandrup T. 1993. A method for unbiased and efficient estimation of number
and mean volume of specified neuron subtypes in rat dorsal root
ganglion. J Comp Neurol 329:269–276.
192
Tandrup T. 1995. Are the neurons in the dorsal root ganglion pseudounipo-
lar? A comparison of the number of neurons and number of myelinated
and unmyelinated fibres in the dorsal root. J Comp Neurol 357:341–
347.
Tandrup T, Braendgaard H. 1994. Number and volume of rat dorsal root
ganglion cells in acrylamide intoxication. J Neurocytol 23:242–248.
Vestergaard S, Tandrup T, Jakobsen J. 1997. Effect of permanent axotomy
on number and volume of dorsal root ganglion cell bodies. J Comp
Neurol 388:307–312.
West MJ, Slomianka L, Gundersen HJG. 1991. Unbiased stereological
J. DEGN ET AL.
estimation of the total number of neurons in the subdivisions of the rat
hippocampus using the optical fractionator. Anat Rec 231:482–497.
Ygge J. 1989. Neuronal loss in lumbar dorsal root ganglia after proximal
compared to distal sciatic nerve resection: a quantitative study in the
rat. Brain Res 478:193–195.
Ygge J, Aldskogius H. 1984. Intercostal nerve transection and its effect on
the dorsal root ganglion. A quantitative study on thoracic ganglion cell
numbers and sizes in the rat. Exp Brain Res 55:402–408.
Ygge J, Aldskogius H, Grant G. 1981. Asymmetries and symmetries in the
number of thoracic dorsal root ganglion cells. J Comp Neurol 202:365–
372.