doi:10.1093/brain/awp160
Brain 2009: 132; 2180–2195
| 2180
BRAIN
A JOURNAL OF NEUROLOGY
Ablation of adhesion molecule L1 in mice favours
Schwann cell proliferation and functional recovery
after peripheral nerve injury
Daria Guseva,1 Doychin N. Angelov,2 Andrey Irintchev1,3 and Melitta Schachner1,4,5
1
2
3
4
5
Zentrum für Molekulare Neurobiologie, Universität Hamburg, Hamburg, Germany
Anatomical Institute I, University of Cologne, Cologne, Germany
Department of Otorhinolaryngology, Friedrich-Schiller-University Jena, Jena, Germany
W. M. Keck Center for Collaborative Neuroscience and Department of Cell Biology and Neuroscience, Rutgers University, NJ, USA
Center for Neuroscience, Shantou University Medical College, Shantou, China
Correspondence to: Melitta Schachner,
Zentrum für Molekulare Neurobiologie,
Universität Hamburg,
Martinistrasse 52,
D-20246 Hamburg,
Germany
E-mail: [email protected]
Correspondence may also be addressed to: Andrey Irintchev,
Department of Otorhinolaryngology,
Friedrich-Schiller-University Jena,
Lessingstrasse 2, D-07740 Jena,
Germany
E-mail: [email protected]
The adhesion molecule L1 is one of the few adhesion molecules known to be beneficial for repair processes in the adult central
nervous system of vertebrates by promoting axonal growth and neuronal survival. In the peripheral nervous system, L1 is
up-regulated by myelination-competent Schwann cells and regenerating axons after nerve damage but its functional role has
remained unknown. Here we tested the hypothesis that L1 is, as in the central nervous system, beneficial for nerve regeneration
in the peripheral nervous system by performing combined functional and histological analyses of adult L1-deficient mice
(L1y/) and wild-type (L1y/+) littermates. Contrary to our hypothesis, quantitative video-based motion analysis revealed
better locomotor recovery in L1y/ than in L1y/+ mice at 4–12 weeks after transection and surgical repair of the femoral
nerve. Motoneuron regeneration in L1y/ mice was also enhanced as indicated by attenuated post-traumatic loss of motoneurons, enhanced precision of motor reinnervation, larger cell bodies of regenerated motoneurons and diminished loss of inhibitory
synaptic input to motoneurons. In search of mechanisms underlying the observed effects, we analysed peripheral nerves at short
time-periods (3–14 days) after transection and found that Schwann cell proliferation is strongly augmented in L1y/ versus
L1y/+ mice. L1-deficient Schwann cells showed increased proliferation than wild-type Schwann cells, both in vivo and in vitro.
These findings suggest a novel role for L1 in nerve regeneration. We propose that L1 negatively regulates Schwann cell
proliferation after nerve damage, which in turn restricts functional recovery by limiting the trophic support for regenerating
motoneurons.
Keywords: adhesion molecule L1; femoral nerve; Schwann cells; peripheral nerve regeneration; motoneuron
Received December 14, 2008. Revised April 25, 2009. Accepted May 12, 2009. Advance Access publication June 18, 2009
ß The Author (2009). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved.
For Permissions, please email: [email protected]
L1 and nerve regeneration
Introduction
The neural cell adhesion molecule L1 is a glycoprotein of the
immunoglobulin superfamily expressed in most, if not all, neurons
in the central nervous system (CNS). During CNS development,
L1 is targeted to the surface of developing axons and growth
cones and mediates outgrowth, adhesion, fasciculation and
guidance of axons as well as neuronal migration and survival
(Hortsch, 2003; Maness and Schachner, 2007). Perturbations
of these processes by mutations in the L1 gene lead to brain
malformations and dysfunctions in humans and mice (Maness
and Schachner, 2007). L1 is also implicated in CNS regeneration
in adult vertebrates. After spinal cord injury in zebrafish, the
expression of L1.1, a homolog of the mammalian L1, is increased
in successfully regenerating descending axons but not in ascending
projections that fail to regenerate (Becker et al., 2005). Also,
morpholino-mediated reduction of L1.1 expression leads to
deficient regeneration (Becker et al., 2004). Similar to poorly
regenerating neurons in zebrafish, mammalian neurons fail to
up-regulate L1 expression after CNS trauma (Styren et al., 1995;
Mason et al., 2003) but L1 expression is up-regulated in CNS
neurons when a peripheral nervous system (PNS) nerve bridge is
introduced (Anderson et al., 1998; Woolhead et al., 1998) or L1 is
overexpressed in Purkinje cells (Zhang et al., 2005). When L1
expression in neurons and glia is induced by viral transduction
(Chen et al., 2007), when L1 overexpressing embryonic stem
cells are transplanted (Chen et al., 2005), or when the axonal
growth-inhibiting environment is enriched in exogenous L1
(Roonprapunt et al., 2003), recovery from spinal cord injury in
rodents is enhanced. These findings clearly indicate that L1 is a
molecule promoting CNS regeneration.
In the PNS, L1 is expressed in axons and Schwann cells during
embryonic and early postnatal development and remains
expressed by non-myelinating Schwann cells in the adult
(Faissner et al., 1984; Nieke and Schachner, 1985; Martini and
Schachner, 1986, 1988). Constitutive ablation of L1 in mice
does not lead to apparent PNS disorders. Neural crest cell
migration is reduced in the developing gut of L1y/ mice but in
the adult the entire gastrointestinal tract is colonized by enteric
neurons (Anderson et al., 2006). It has, thus, been suggested that
L1 mutations contribute to Hirschsprung’s disease, a human condition in which enteric neurons are absent from the distal bowel.
L1y/ mice have mild impairments in locomotion and severely
reduced pain sensitivity, but these deficits have been attributed
to abnormal function of the CNS (Thelin et al., 2003; Jakeman
et al., 2006). Sprouting of calcitonin gene-related peptide positive
sensory axons into the lesion site, which spontaneously occurs
after spinal cord injury and is associated with development of
pain, is reduced in L1y/ mice (Deumens et al., 2007). In peripheral nerves of adult L1y/ mice, myelination is normal but
the ensheathment of non-myelinated axons is disturbed
(Dahme et al., 1997; Haney et al., 1999). Peripheral nerve
injury in adult rodents leads to up-regulation of L1 expression in
myelination-competent Schwann cells distal to the injury site and
enhanced expression in regenerating axons (Daniloff et al., 1986;
Martini and Schachner, 1988; Martini, 1994). However, the
functional relevance of this up-regulation has not been elucidated.
Brain 2009: 132; 2180–2195
| 2181
We speculated, considering the regeneration-promoting properties
of L1 in the CNS, that L1 is beneficial also for PNS regeneration.
To test this hypothesis, we analysed peripheral nerve regeneration
in L1y/ mice and wild-type (L1y/+) littermates using a wellestablished paradigm, femoral nerve transection and repair in
adult mice combined with quantitative video-based analysis of
motor recovery and cellular correlates of the functional outcome
(Irintchev et al., 2005a; Eberhardt et al., 2006; Simova et al.,
2006; Ahlborn et al., 2007). Contrary to our expectation, however, we found superior outcome of nerve repair in L1y/ mice
suggesting that ablation of L1 leading to enhanced proliferation of
Schwann cells is beneficial for peripheral nerve regeneration.
Materials and Methods
Animals
We used 12-week-old L1-deficient (L1y/) mice from a colony
generated by insertion of a tetracycline-controlled transactivator into
the second exon of the L1 gene (Rolf et al., 2001; Saghatelyan et al.,
2004). The L1y/ mice were males carrying the mutant allele of the
X-chromosome-linked L1 gene. The wild-type (L1y/+) mice used as
controls were male littermates of the L1y/ mice carrying the
wild-type L1 allele. Unlike the L1y/ mice generated initially in
our laboratory, which express residual amounts of L1 (Dahme
et al., 1997), these L1y/ mice were L1 null mice. Importantly,
the L1y/ mice used here were maintained on a 129SvJ/NMRI
genetic background which attenuates some abnormal features
seen in L1y/ mice bred on a C57BL/6J background such as overt
hydrocephalus, low body mass (60–80% of L1y/+ mice), poor breeding and high mortality of the L1y/ offspring within the first 2 months
after birth. Before and after the experiments, the mice were kept on
standard laboratory food and tap water ad libitum, with an artificial
12 h light/dark cycle. All treatments of the animals were conducted
in accordance with the German and European Community laws on
protection of experimental animals.
Experimental design
In Experiment 1 (see ‘Nerve regeneration’ below), we analysed functional recovery in L1y/ mice and L1y/+ littermates during a 12-week
recovery period following femoral nerve injury (Fig. 1A). Subsequently,
these animals were subjected to retrograde labelling of regenerated
motoneurons (Fig. 1B) and, following a survival period of 1 week,
spinal cords and both injured and intact contralateral femoral nerves
were sampled for histological analyses (Fig. 1A). In addition to these
‘injured’ mice, i.e. subjected to femoral nerve transection and surgical
repair, groups of control ‘non-injured’ L1y/ and L1y/+ mice
were included in this experiment. These mice were subjected only to
retrograde tracing (Fig. 1C) and used for morphological analysis of
quadriceps motoneurons. In Experiment 2 (‘Cell proliferation during
Wallerian degeneration’, Fig. 1D), the femoral nerve of L1y/ and
L1y/+ mice was transected and 3, 5, 7 and 14 days later the distal
nerve segments were harvested for analyses of cell proliferation.
Finally, in Experiment 3 an in vitro analysis was performed to
investigate the proliferation of Schwann cells derived from 7-day-old
L1y/ and L1y/+ mice.
2182
| Brain 2009: 132; 2180–2195
D. Guseva et al.
Figure 1 Schematic representation of the experimental design. In Experiment 1, L1y/ and L1y/+ mice were subjected to unilateral
transection and repair of the femoral nerve (at 0 weeks in A) proximal to the bifurcation of the sensory saphenous branch and the
mixed quadriceps branch (‘Nerve transection’ in B). Recovery of function was followed by repeated video recordings over a 12-week
period (A, arrowheads). At 12 weeks after injury, retrograde labelling of regenerated motoneurons was performed by transecting the
two nerve branches distal to the bifurcation and application of two different fluorescence dyes, fluoro-ruby and Fast Blue to the
saphenous and the quadriceps branch, respectively (B, ‘Retrograde labelling’). One week later, femoral nerves and spinal cords were
sampled for morphological analyses (A, elbow arrow). Additional control mice were subjected to retrograde labelling only (C) and
served as controls in the morphological analyses. The circles on the left hand side in C represent motoneurons with axons targeting the
quadriceps branch only and thus retrogradely labelled by Fast Blue but not with fluoro-ruby. After nerve regeneration, motor axons
regenerate correctly to the quadriceps branch or aberrantly to the saphenous branch, or both nerve branches. Application of the two
tracers allows the detection of both aberrantly and correctly projecting motoneurons (circles on the left hand side in B). The arrow at
‘Analysis of axons’ points to the approximate level at which axons were analysed in intact and regenerated nerves. In Experiment 2,
mice of both genotypes were subjected to femoral nerve transection (0 day, D) and femoral nerves were sampled for analysis of
Schwann cell proliferation at 3, 5, 7 and 14 days after injury (elbow arrows). Quantitative assessment of Schwann cell proliferation was
performed at the level indicated in C. For further details see text.
Nerve regeneration
Transection and surgical repair of the femoral nerve
Surgery was performed as described previously (Simova et al., 2006).
The mice were anesthetized by intraperitoneal injections of 0.4 mg/kg
fentanyl (Fentanyl-Janssen, Janssen, Neuss, Germany), 10 mg/kg
droperidol (Dehydrobenzperidol, Janssen) and 5 mg/kg diazepam
(Valium 10 Roche, Hoffman – La Roche, Grenzach-Wyhlen,
Germany). The right femoral nerve was exposed and transected
approximately 3 mm proximal to the bifurcation of the saphenous
and quadriceps branches (Fig. 1B). The cut ends of the nerve were
inserted into a polyethylene tube (3 mm length, 0.58 mm inner diameter, Becton Dickinson, Heidelberg, Germany) and fixed with a single
epineural 11-0 nylon stitch (Ethicon, Norderstedt, Germany) so that a
2 mm gap was present between the proximal and distal nerve stumps.
L1 and nerve regeneration
Brain 2009: 132; 2180–2195
| 2183
The tube was filled with phosphate buffered saline (PBS, pH 7.4) and
the skin was closed with 6-0 sutures (Ethicon). To prevent hypothermia, the mice were then kept in a warm room (35 C) until full
recovery from anaesthesia.
Analysis of motor function
Functional analysis was performed over a time-period of 12 weeks
(Fig. 1A) using a quantitative video-based approach developed in
our laboratory (single-frame motion analysis, Irintchev et al., 2005a).
To evaluate locomotor function, mice were accustomed, in 3–4 trials,
to beam-walking prior to operation. In this test, the animal walks
voluntarily from one end of a horizontal beam (length 1000 mm,
width 40 mm) towards its home cage located at the other end of
the beam. For all mice, a rear view of one walking trial was captured
once prior to the operation and at different time-points after surgery
(Fig. 1A) with a high-speed camera (A602fc, Basler, Ahrensburg,
Germany) at 100 frames per sec and stored on a personal computer
in Audio Video Interleaved (AVI) format. The video sequences were
examined using SIMI-Motion 7.0 software (SIMI Reality Motion
Systems, Unterschleissheim, Germany). Selected frames in which the
animals were seen in defined phases of the step cycle were used for
measurements of two parameters: the heels-tail angle (HTA, Fig. 2A
and C) and the foot-base angle (FBA, Fig. 2B and D) as described
previously (Irintchev et al., 2005a). Both parameters are directly
related to the ability of the quadriceps muscle to keep the knee joint
extended during contralateral swing phases. As a relative measure
of functional recovery at different time-points after nerve injury, we
calculated the stance recovery index (RI), which is a mean of the RI for
the heels-tail angle and the foot-base angle. The index for each angle
is calculated in percent as:
RI ¼ ½ðXreinn Xden Þ=ðXpre Xden Þ 100,
Figure 2 Analysis of motor function. (A–D) Single video
frames from recordings of beam walking of an L1y/ and an
L1y/+ mouse at prior to femoral nerve repair (0 day). In A and
C the animals are seen at mid-stance of the right hind limb and
maximum altitude of the contralateral swing. Such video
frames are used to measure the heels-tail angle as shown by
the lines drawn in both panels. Note the lower position of right
heel in A compared with C. This difference results in a larger
heels-tail angle (measure from the dorsal aspect) in the L1y/
mouse compared with the L1y/+ mouse. B and D show video
frames in which the right paw of the same mice at take-off
position. Such frames are used for measuring the foot-base
angle shown by lines drawn in the panels. A more pronounced
external rotation of the paw in B compared with D results in a
smaller foot-base angle (measured from the medial aspect) in
the L1y/ mouse compared with the L1y/+ mouse. (E–H)
Time-course of motor recovery after femoral nerve lesion.
Mean values SEM of heels-tail angles (E), foot-base angles
(F) and limb PLRs (G) at different time-points after femoral
nerve transection in L1y/ and L1y/+ mice. Preoperative
values are plotted at 0 day. Number of animals studied per
group are indicated in E. (H) Shows individual animal values of
the stance RI calculated for the heels-tail and the foot-base
where Xpre, Xden and Xreinn are values prior to operation, during the
state of denervation (7 days after injury), and at any given time-point
during reinnervation, respectively.
A third parameter, the limb protraction length ratio (PLR), was
evaluated from video-recordings of voluntary pursuit movements of
the mice (Irintchev et al., 2005a). The mouse, when held by its tail
and allowed to grasp a pencil with its forepaws, tries to catch the
object with its hindpaws and extends both hindlimbs simultaneously.
In non-injured animals, the relative length of the two extremities, as
estimated by lines connecting the most distal mid-point of the extremity with the anus, is approximately equal and the PLR (ratio of
the right to left limb length) is close to 1.0. After denervation,
the limb cannot be extended maximally and the PLR becomes
significantly 41.0.
Retrograde labelling of motoneurons
At 12 weeks after nerve transection the animals were anaesthetized
with fentanyl, droperidol and diazepam for retrograde labelling
of regenerated motoneurons (Fig. 1A, Simova et al., 2006). After
exposure of the right femoral nerve, a piece of Parafilm (Pechiney
Plastic Packaging, Chicago, IL, USA) was inserted underneath
the nerve and the two nerve branches were transected approximately 5 mm distal to the bifurcation (Fig. 1B). Fluorescence retro-
angles (see Materials and methods section for details) at 12
weeks after injury. A RI of 100% indicates complete recovery.
Asterisk indicates significant difference (P50.05, one-way
ANOVA with Tukey’s post hoc test) between the genotypes at
given time-point.
2184
| Brain 2009: 132; 2180–2195
D. Guseva et al.
grade tracers were applied to the cut nerve ends in powder form:
fluoro-ruby (tetramethylrhodamine dextran, Molecular Probes,
Leiden, The Netherlands) to the saphenous (cutaneous) branch (exclusively sensory branch), Fast Blue (EMS-Chemie, Großumstadt,
Germany) to the quadriceps (muscle) branch (mixed nerve containing
sensory and motor axons). Thirty minutes after dye application, the
nerve stumps were rinsed with PBS and the wound was closed. This
labelling procedure allows visualization of all femoralis motoneurons
that have successfully regenerated their axons beyond the lesion
site. In addition, it is possible to differentiate between motoneurons
with axons correctly projecting into the quadriceps branch (blue circle
in Fig. 1B) and aberrantly projecting motoneurons sending axons to
the saphenous branch only or to both the saphenous and the quadriceps branch (red and purple circle in Fig. 1B, respectively). The same
labelling procedure applied to ‘non-injured’ L1y/ and L1y/+ mice, i.e.
mice which had not been subjected to nerve transection and repair
(Fig. 1C), allows to estimate the normal number of quadriceps
motoneurons all of which project to the quadriceps branch only
(blue circles in Fig. 1C).
Preparation of tissue for morphological analyses
One week after retrograde labelling (Fig. 1A) the mice were
anesthetized by intraperitoneal injection of 16% sodium pentobarbital
solution (NarcorenÕ , Merial, Hallbergmoos, Germany, 5 ml/g body
weight) and were transcardially perfused with 4% formaldehyde in
0.1 M sodium cacodylate buffer, pH 7.3. The lumbar spinal cord was
removed, post-fixed overnight at 4 C and then immersed in 15%
sucrose solution in 0.1 M cacodylate buffer, pH 7.3, for 1 day at
4 C. Afterwards the tissue was frozen for 2 min in 2-methyl-butane
(isopentane, Carl Roth, Karlsruhe, Germany) pre-cooled to 30 C.
For sectioning, the spinal cord segment was attached to a cryostat
specimen holder using TissueTekÕ (Sakura Finetek Europe,
Zoeterwoude, The Netherlands). Serial transverse sections of 25 mm
thickness were cut on a cryostat (Leica CM3050, Leica Instruments,
Nußloch, Germany) and picked up on SuperFrostÕ Plus glass slides
(Roth, Karlsruhe, Germany). The sections were air-dried for at least
1 h at room temperature (RT) and mounted in anti-quenching medium
(Fluoromount G, Southern Biotechnology Associates, Biozol, Eching,
Germany). These sections were used for counting retrogradely labelled
motoneurons.
Both injured and intact contralateral femoral nerves were dissected
from perfusion fixed animals. The tissue samples were post-fixed in
1% osmium tetroxide (Polysciences Europe, Eppelheim, Germany) in
0.1 M sodium cacodylate buffer, pH 7.3, for 1 h at RT, dehydrated and
embedded in resin according to standard protocols. One-mm thick
cross-sections from the motor and sensory nerve branches were cut
at a distance of 3 mm distal to the bifurcation (Fig. 1B) and stained
with 1% toluidine blue/1% borax in distilled water for analysis of axon
numbers, diameters, degree of myelination (g-ratios) in regenerated
and intact nerve branches.
Counting of retrogradely labelled motoneurons
Serial transverse sections of spinal cord were examined under a
fluorescence microscope (Axiophot 2, Zeiss, Oberkochen, Germany)
with appropriate filter sets. Each section was examined using a
40 objective by focusing through the section thickness starting
from the top surface. All profiles, except those visible at the top
surfaces of sections, were counted (Simova et al., 2006). The application of this simple stereological principle prevented double counting of
labelled cells and allowed precise evaluation of cell numbers.
Immunofluorescence staining of perisomatic axon
terminals
Immunofluorescence staining was performed as described (Irintchev
et al., 2005b) using commercial antibodies at optimal dilutions
(Table 1). Sections of spinal cord containing retrogradely labelled
motoneurons were freed from coverslips and mounting medium by
soaking in PBS. Antigen retrieval was performed by immersion into
0.01 M sodium citrate solution, pH 9.0, heated to 70 C in a water
bath for 30 min. Blocking of non-specific binding sites was then
Table 1 Primary antibodies used in this study
Antibody
Cellular phenotypes/
structures recognized
Host
Code/clone
Dilution
Source
S-100
S-100
p75 NGF receptor
Glial fibrillary acidic
protein (GFAP)
Laminin B2/1
Class III b-tubulin
Macrophage antibody
F4/80
L1
Mature
Mature
Mature
Mature
Rabbit
Mouse
Rat
Mouse
Z0311
MAB079-1
ab27007
MAB3402
1:15 000
1:1000
1:1000
1:500
DakoCytomation, Hamburg, Germany
Chemicon, Hofheim, Germany
Abcam, Cambridge, UK
Chemicon
Basal laminae
Axons
Macrophages
Rat
Rabbit
Rat
RT-795-P1ABX YYY
PRB-435P-100
MCA497R
1:100
1:2000
1:300
NeoMarkers, Fremont, CA, USA
Covance, Berkeley, CA, USA
AbD Serotec, Raleigh, NC, USA
Regenerating axons,
non-myelinating
Schwann cells
Apoptotic cells
Proliferating cells
Proliferating cells
Rat
555
1:500
Rabbit
Rabbit
Mouse
AF835
AB15580
G3G5
1:2000
1:500
1:100
Mouse
131 011
1:1000
Antibody produced in our laboratory
and purified by protein G affinity
chromatography (Appel et al., 1995)
R&D Systems, Minneapolis, MN, USA
Abcam
Developmental Hybridoma Bank, Iowa City,
IA, USA
Synaptic Systems, Göttingen, Germany
Goat
AB144P
1:100
Chemicon
Caspase-3 activated
Ki67
Bromodeoxyuridine
(BrdU)
Vesicular GABA
transporter (VGAT)
Choline acetyltransferase
(ChAT)
Schwann
Schwann
Schwann
Schwann
cells
cells
cells
cells
Inhibitory (GABAergic
and glycinergic)
synaptic terminals
Cholinergic synaptic
terminals
L1 and nerve regeneration
performed using PBS containing 0.2% Triton X-100 (Sigma), 0.02%
sodium azide (Merck, Darmstadt, Germany) and 5% normal goat or
normal donkey serum for 1h at RT. Incubation with primary antibodies
against VGAT or ChAT (Table 1), diluted in PBS containing 0.5%
lambda-carrageenan (Sigma) and 0.02% sodium azide, was carried
out at 4 C for 3 days. For a given antigen, all sections were stained
in the same solution kept in screw-capped staining plastic jars (capacity
35 ml, 10 slides, Roth). After washing in PBS, appropriate
Cy3-conjugated secondary antibodies (Jackson ImmunoResearch
Europe, Suffolk, UK) diluted 1:200 in PBS-carrageenan solution were
applied for 2 h at RT. After a subsequent washing in PBS, cell nuclei
were stained for 10 min at RT with bis-benzimide solution (Hoechst
33258 dye, 5 mg/ml in PBS, Sigma) and the sections were mounted in
anti-quenching medium. Specificity of staining was controlled by omitting the primary antibody or replacing it by an equivalent amount
of non-immune IgG or serum derived from the same species as the
specific antibody. These controls were negative.
Analyses of motoneuron size and perisomatic synaptic
terminals
Sections containing retrogradely labelled motoneurons and additionally
stained for ChAT and VGAT were used to estimate soma areas and
motoneuron perisomatic coverage (Apostolova et al., 2006; Simova
et al., 2006). Linear density (number per unit length) of perisomatic
terminals was estimated for motoneurons that correctly projected to
the motor nerve branch of the femoral nerve (identified by Fast Blue
back-labelling). Stacks of images of 1 mm thickness were obtained on
a confocal microscope (Leica DM IRB, Leica, Wetzlar, Germany) using
a 63 oil immersion objective and digital resolution of 1024 1024
pixels. One image per cell at the level of the largest cell body crosssectional area was used to measure soma area, perimeter and number
of perisomatic puncta. These measurements were performed using the
Image Tool 2.0 software program (University of Texas, San Antonia,
TX, USA, http://ddsdx.uthscsa.edu/dig/).
Analyses of axons in the regenerated nerve branches
Total numbers of myelinated axons per nerve cross-section were estimated on an Axioskop microscope (Zeiss) equipped with a motorized
stage and Neurolucida software-controlled computer system
(MicroBrightField Europe, Magdeburg, Germany) using a 100 oil
objective (Simova et al., 2006). Axonal and nerve fibre diameters
were measured in a random sample from each section. For sampling,
a grid with line spacing of 60 mm was projected into the microscope
visual field using the Neurolucida software. Selection of the reference
point (0 coordinates) of the grid was random. For all myelinated axons
crossed by or attaching the vertical grid lines through the sections,
mean orthogonal diameters of the axon (inside the myelin sheath)
and of the nerve fibre (including the myelin sheath) were measured
(Fig. 6A). The mean orthogonal diameter is calculated as a mean of
the line connecting the two most distal points of the profile (longest
axis) and the line passing through the middle of the longest axis at
right angle (Irintchev et al., 1990). The degree of axonal myelination
was estimated by the ratio axon to fibre diameter (g-ratio).
Analysis of cell proliferation during
Wallerian degeneration
Femoral nerve surgery was performed as described above (Fig. 1B)
avoiding the use of polyethylene tubes. The mice (n = 2–3 per
genotype and time-point) were perfused 3, 5, 7 or 14 days after
injury (Fig. 1D) and the femoral nerve segments of 15 mm length,
Brain 2009: 132; 2180–2195
| 2185
including the proximal and distal stumps, were post-fixed and
cryoprotected by sucrose infiltration as described above. For indirect
immunofluorescence, the tissue samples were immersed in TissueTekÕ
medium and frozen in liquid nitrogen. Longitudinal (5 mm thick)
and cross (10 mm thick) cryostat sections were used in single and
double-labelling immunofluorescence experiments using antibodies
listed in Table 1. Indirect immunofluorescence was performed as
described above for perisomatic terminals using the primary antibodies
alone for single labelling or mixed at optimal dilutions for double
labelling. Appropriate Cy3- and Cy2-conjugated antibodies preabsorbed with normal sera from diverse species to prevent crossreactions (Multiple Labeling antibodies, Jackson ImmunoResearch)
were used in double labelling experiments. Specificity of staining was
controlled by omitting the primary antibody or replacing it by an
equivalent amount of non-immune IgG or serum. These controls
were negative.
Proliferating cells were quantified using nerves sampled 5 days
after femoral nerve transection (Fig. 1D), a time-point of maximal
cell proliferation as indicated by Ki67 staining (Fig. 8D–F0 ). Cell
counts were performed on an Axioskop microscope equipped with
a motorized stage and Neurolucida software-controlled computer
system using a 100 oil objective. Ten spaced-serial cross-sections
(125 mm apart) cut at approximately 3 mm distal to the site of nerve
transection were used for the analysis per animal. These sections
were stained with anti-p75 NGF receptor and anti-Ki67 antibodies to
visualize Schwann cells and proliferating cells, respectively, and with
bis-benzimide solution to reveal cell nuclei. Four types of nuclei were
distinguished (Fig. 10A): Ki67+ nuclei in p75+ cells (type Ia, proliferating Schwann cells), Ki67 nuclei in p75+ cells (Ib, non-proliferating
Schwann cells), Ki67+ nuclei in p75 cells (IIa, proliferating nonSchwann cells) and Ki67 nuclei in p75 cells (IIb, non-proliferating
non-Schwann cells). The total number of nuclei per section and the
numbers of Ia, Ib, IIa, IIb nuclei were quantified. The extent of
Schwann cell proliferation was estimated by the ratio of type Ia to
type IIa nuclei (proliferating Schwann cells to proliferating nonSchwann cells) and numbers of Ia nuclei normalized to all nuclei in
a tissue section and nerve cross-sectional area.
Analysis of Schwann cell proliferation in vitro
Sciatic and femoral nerves were dissected from 7-day-old L1y/+ and
L1y/ mice, freed of connective tissue and transferred into Petri dishes
with ice-cold Ham’s F-12 medium (PAA, Cölbe, Germany). The nerves
were treated with 0.25% trypsin (Sigma, Deisenhofen, Germany) and
0.03% collagenase (Sigma) in Dulbecco’s modified Eagle medium
(DMEM, PAA) for 30 min at 37 C. Enzymatic digestion was stopped
by rinsing 2 with F-12 medium. The tissue samples were then
transferred into complete (DMEM/Ham’s F-12) medium containing
0.01% DNase I (Sigma). The complete medium consisted of equal
parts DMEM and F-12 media and was supplemented with 60 ng/ml
progesterone, 16 mg/ml putrescine, 5 mg/ml insulin, 400 ng/ml
L-thyroxine, 160 ng/ml sodium selenite, 10.1 ng/ml triiodo-thyronine,
38 ng/ml dexamethasone (all from Sigma), 7.9 mg/ml glucose (VWR,
Darmstadt, Germany), 0.01% bovine serum albumin, 100 IU/ml
penicillin/streptomycin, 2 mM L-glutamine, 100 mM sodium pyruvate
(all from PAA) and 20% horse serum (Invitrogen, Karlsruhe,
Germany). Mechanical dissociation was performed by gentle trituration
with a fire-polished Pasteur pipette. After adding of DMEM, the cell
suspension was centrifuged at 500g for 10 min at 4 C in a centrifuge
tube on top of a 4% BSA cushion. The supernatant was discarded
and the cells were resuspended in 1 ml fresh pre-warmed (37 C)
complete medium, counted and plated at a density of 2 106
cells/ml on 0.01% poly-L-lysine-coated coverslips. After 4 h, 20 mM
2186
| Brain 2009: 132; 2180–2195
5-bromo-20 -deoxyuridine (BrdU, Sigma) was applied to the cultures
and the cells were grown for additional 48 h at 37 C in a
CO2-enriched (5%) humidified atmosphere. Finally, the culture
medium was removed and, after rinsing with PBS, the cells were
fixed with 4% formaldehyde in PBS for 30 min at RT, rinsed 3
with PBS and incubated with 2N HCl for 30 min at 37 C to denature
DNA. After rinsing 2 with PBS, non-specific binding was blocked by
incubation for 30 min with 5% normal goat serum (see below) in PBS
at RT. The fixed cells were incubated with a mouse monoclonal antiBrdU antibody (see below) overnight at 4 C, washed 3 5 min with
PBS, incubated with a Cy3-conjugated goat anti-mouse secondary
antibody (see below) for 1 h at RT, washed once with PBS and
mounted in anti-quenching medium.
Analysis of Schwann cell proliferation was performed using confocal
images of randomly selected fields obtained on a confocal microscope
(Olympus FV10-ASW, Hamburg, Germany). The percentage of
proliferating Schwann cells (ratio of BrdU+ to all Schwann cells) was
estimated in 20 fields, each containing 100–200 cells, per cell culture
preparation. At least three cultures were analysed per genotype.
Schwann cells were identified by phase contrast microscopy as elongated bi- or tripolar cells with an oval soma (Fig. 11A–C). Fibroblasts
appeared as flattened, polymorphic cells with large, round nuclei
(Fig. 11A–C).
Results
Gait abnormalities in non-injured
L1y/ mice
Frame-by-frame analyses of video sequences recorded prior
to nerve transection (Fig. 1A) revealed differences between
L1y/ and L1y/+ littermates. First, during the mid-stance phase,
at maximum altitude of the contralateral swing, almost the entire
planar paw surface of L1y/ mice contacted the ground (Fig. 2A).
In contrast, only the digits of L1y/+ mice were on the ground
during this gait phase (Fig. 2C). Second, the external rotation of
the paw at take-off position was more pronounced in L1y/ than
in L1y/+ mice (Fig. 2B and D). Consequently, L1y/ mice had
different heels-tail angles and the foot-base angles (Fig. 2A–F),
two measures used to estimate quadriceps muscle function after
femoral nerve lesion (Irintchev et al., 2005a). However, the
observed deviations of the angle values from normal in uninjured
L1y/ mice were opposite to the changes induced by femoral
nerve injury (Fig. 2E and F) and, thus, could not be attributed
to quadriceps muscle weakness.
Enhanced functional recovery after
nerve repair in L1y/ mice
Despite genotype-related differences in gait prior to operation,
damage of the femoral nerve-induced changes in the heels-tail
and the foot-base angles similar in magnitude in L1y/ and
L1y/+ mice (compare values at 0 and 1 week after injury in
Fig. 2E and F). These alterations were caused by impaired extensor
function of the quadriceps muscle leading to abnormally high
heel position and external rotation of the ankle at gait cycle positions similar to those seen in Fig. 2A and C and Fig. 2B and D,
D. Guseva et al.
respectively. Importantly, at 1 week after injury, no significant
differences were found between L1y/ and L1y/+ mice (Fig. 2E
and F, ANOVA with Tukey’s post hoc test). After the first week,
the angles in L1y/+ mice gradually returned to normal but
recovery was deficient, as compared to normal values, even
after 12 weeks (Fig. 2E and F, P50.05, ANOVA with Tukey’s
post hoc test). Locomotor improvement in L1y/ mice was
superior compared with L1y/+ mice at 4–12 weeks after injury
(Fig. 2E and F). Also, as estimated by the stance RI, a measure
of the individual degree of recovery calculated for both angles, the
final outcome of femoral nerve repair was significantly better in
L1y/ than in L1y/+ mice (87 14% versus 60 8.5% recovery,
P50.05, t-test, Fig. 2H).
In addition to ground locomotion, we also evaluated the ability
of the animals to extend the knee joint during movements without
body weight support using the ‘pencil’ test (Irintchev et al.,
2005a). This ability, estimated by the limb PLR (length of the
non-injured hindlimb/length of injured limb during maximum
extension), was similar in L1y/ and L1y/+ mice before operation
(Fig. 2G). Also, the degree of disability at 1 week after operation
and the recovery thereafter were similar in the two genotypes
(Fig. 2G). In conclusion, the combined results show that L1
ablation leads to a superior functional outcome.
Reduced numbers of quadriceps
motoneurons in non-injured
L1y/ mice
The number of retrogradely labelled quadriceps motoneurons was
significantly lower, by 29%, in non-injured L1y/ mice compared
non-injured L1y/+ littermates (Fig. 3). In both phenotypes, only
Fast Blue labelled motoneurons were observed indicating that all
motoneurons project to the quadriceps nerve branch only
(Fig. 1C). No small (microglial) cells loaded with Fast Blue, as a
result of phagocytosis of retrogradely labelled motoneurons
(Simova et al., 2006), were detected in spinal cord sections from
L1y/ and L1y/+ mice indicating that motoneuron death in
L1y/ mice during the 1 week labelling period was not the
reason for the reduced number of labelled motoneurons. The finding of reduced number of motoneurons in L1y/ mice, in conjunction with a lower than normal number of myelinated axons in
non-injured nerves (see following section) and normal body mass
in L1y/ mice (35 1.7 g versus 36 2.8 g in L1y/ and L1y/+
mice, respectively, P 40.05, t-test), indicates that L1 ablation leads
to a deficient development and/or maintenance in the adult of the
quadriceps motoneuron pool.
Better motoneuron survival and
regeneration in injured L1y/ mice
At 12 weeks after nerve repair, the total number of motoneurons
retrogradely labelled through the motor (quadriceps) or the
sensory (saphenous) branch only, or through both branches of
the femoral nerve, was significantly reduced (32%) in injured
compared with non-injured L1y/+ mice (Fig. 3). In contrast to
L1y/+ mice, the total number of regenerated motoneurons in
L1 and nerve regeneration
Figure 3 Analysis of motoneuron regeneration. Mean
numbers of motoneurons (SEM) labelled through the motor
or sensory branch only (grey and dark grey bars, ‘Motor’ and
‘Sensory’, respectively), through both branches (light grey bars,
‘Both’), and the sum of neurons in the first three categories
(black bars, ‘Total’) 12 weeks after femoral nerve repair in
L1y/ and L1y/+ mice (right-hand half of the graph, ‘Injured’,
6/6 animals of L1y/ and L1y/+ mice, respectively) and
non-injured mice (‘Non-injured’ on the left-hand side, 10/13
animals of L1y/ and L1y/+ mice, respectively). In non-injured
mice retrograde tracing but no previous nerve injury was
performed. Asterisks and hash indicate differences between
non-injured and injured groups of the same genotype and
between L1y/+ and L1y/ animals under the same
experimental conditions, respectively (P50.05, one-way
ANOVA with Tukey’s post hoc test).
injured L1y/ mice was similar to the number in non-injured L1y/
animals (Fig. 3). In both genotypes, small fractions of all labelled
motoneurons projected aberrantly to the sensory branch only, or
to both the motor and the sensory branches (Fig. 3). Also in both
genotypes, the majority of regenerated motoneurons had axons
which had re-grown only into the appropriate motor nerve
branch—a feature of preferential motor reinnervation (PMR,
Fig. 3) (Brushart, 1988). However, the degree of PMR, estimated
as index of PMR (motoneurons projecting to the motor branch/
motoneurons projecting to the sensory branch), was significantly
higher in L1y/ mice compared with L1y/+ mice (4.4 1.1 versus
2.0 0.52, respectively, P50.05, t-test). In conclusion, the results
show better motoneuron regeneration in L1y/ compared with
L1y/+ mice: motoneuron loss and/or failure of regeneration are
reduced and precision of motor reinnervation is improved.
L1 deficiency does not influence
axonal regeneration
Next we analysed non-injured and regenerated nerves morphologically to assess axonal numbers, axonal diameters and degree
of myelination. In both the motor and sensory branches of the
femoral nerve, we found a significant reduction in the total
number of myelinated axons in non-injured L1y/ mice (by 17
and 26%, respectively) compared with L1y/+ littermates (Fig. 4A
and B). At 12 weeks after injury, the numbers of regenerated
axons in the motor and sensory branches of the femoral
Brain 2009: 132; 2180–2195
| 2187
Figure 4 Analysis of myelinated axons in non-injured and
regenerated motor (A) and sensory (B) nerve branches at
12 weeks after injury in L1y/ and L1y/+ mice (‘Injured’,
n = 11 and 10, respectively) and in non-injured contralateral
nerves (‘Non-injured’, n = 13 and 6, respectively). Asterisks
and hash indicate differences between non-injured and injured
groups of the same genotype and between L1y/+ and L1y/
animals under the same experimental conditions, respectively
(P50.05, one-way ANOVA with Tukey’s post hoc test).
nerve were reduced compared with non-injured nerves in both
L1y/ (by 27 and 34%, respectively) and L1y/+ mice (by
26 and 35%, respectively, Fig. 4A and B). The axonal numbers
in the regenerated nerve branches were not significantly different
between L1y/ and L1y/+ mice. Also, no differences between the
genotypes were found for axonal diameters and relative degree of
myelination (g-ratio) in non-injured or regenerated nerves (Figs 5
and 6) with one exception. The frequency distribution of axonal
diameters in the non-injured motor branch of L1y/ mice was
partially shifted towards larger values (P50.05, Kolmogorov–
Smirnov test) indicating that a fraction of the axons in L1y/
mice had larger than normal diameters (Fig. 5). With regard to
nerve repair, we can conclude that L1 ablation in both axons
and Schwann cells does not influence axonal regrowth and
remyelination as estimated at 12 weeks after injury.
Larger soma size and attenuated loss of
inhibitory synaptic input to regenerated
L1y/ motoneurons
The soma size of motoneurons in L1y/ and L1y/+ mice with
or without preceding injury of the femoral nerve (‘Injured’ and
‘Non-injured’ groups in Fig. 7D, respectively) was evaluated only
for motoneurons projecting to the quadriceps muscle as detected
by retrograde labelling (blue cells in Figs 1B and 7A). Both
non-injured and injured L1y/ mice had larger motoneurons
compared with the respective groups of L1y/+ mice (by 20 and
16%, respectively, Fig. 7D).
Analysis of inhibitory (GABAergic and glycinergic, VGAT+)
terminals and modulatory cholinergic (ChAT+) terminals around
motoneuron perikarya revealed differences between the genotypes in mice without injury (‘Non-injured’ groups in Fig. 7E and
F). Density (number per mm) of VGAT+ terminals was lower and
that of ChAT+ puncta higher in L1y/ compared with L1y/+ mice
(20 and 22%, respectively). After nerve injury (‘Injured’ groups in
Fig. 7E and F), densities of ChAT+ terminals in both genotypes
were not changed compared with the respective non-injured
2188
| Brain 2009: 132; 2180–2195
Figure 5 Analysis of axon diameters in regenerated and
non-injured nerves. Shown are normalized frequency
distributions of axon diameters in non-injured (A and B) and
regenerated (C and D) motor (A and C) and sensory (B and D)
nerve branches of the femoral nerve in L1y/ and L1y/+ mice.
Regenerated nerves were studied 12 weeks after injury.
Number of axons and nerves studied per group are shown in
the panels. A significant difference between the genotypes was
detected only for the non-injured motor nerve branches (A,
P50.05, Kolmogorov–Smirnov test). This difference indicates
increased proportion of large-diameter axons in the motor
branch of L1y/ mice. Abnormally thick axons (413 mm) were
not present in L1y/ animals as illustrated by images of
semithin sections from the motor nerve branches of an L1y/
and an L1y/+ mouse (E and F).
group as also were VGAT+ boutons in L1y/ mice. In contrast to
L1y/ mice, VGAT+ terminals were reduced in injured compared
with non-injured L1y/+ animals (22%). We can conclude that
synaptic coverage related to nerve injury differs in L1y/+ and L1y/
mice. Particularly interesting from a functional point of view is
the finding that nerve repair-related loss of inhibitory terminals in
L1/+ mice was not found in L1y/ mice.
Enhanced Schwann cell proliferation
in L1y/ mice after nerve damage
We analysed sections from femoral nerves 3, 5, 7 and 14 days
after injury in L1y/+ and L1y/ mice (Fig. 1D) using the L1 antibody 555 and antibodies recognizing Schwann cell and axonal
antigens, S-100 and ß-tubulin, respectively (data not shown).
D. Guseva et al.
Figure 6 Analysis of myelination in regenerated and
non-injured nerves. Mean orthogonal diameters of the axon
(A, black arrows) and of the nerve fibre (A, white arrows) were
measured and the degree of myelination was estimated by the
ratio axon to fibre diameter (A, g-ratio). Shown are normalized
frequency distributions of g-ratios in non-injured (B and C) and
regenerated (D and E) motor (B and D) and sensory (C and E)
nerve branches of the femoral nerve in L1y/ and L1y/+ mice.
Regenerated nerves were studied 12 weeks after injury.
Number of axons and nerves studied per group are identical
with those shown in Fig. 4. No differences between the
genotypes were detected for any pair of samples (P40.05,
Kolmogorov–Smirnov test).
While L1 could not be detected in L1y/ samples, L1 was
up-regulated after nerve transection. The number of S-100positive cells was higher, at any time-point studied (2–3 animals
studied per time-point and genotype), in L1y/ versus L1y/+ mice
(Fig. 8A–C). A similar difference between the genotypes was
found for another Schwann cell marker, GFAP (not shown).
These observations suggested that L1 deficiency influences
Schwann cell proliferation after nerve injury. We used an antibody
against the Ki67 antigen to assess cell proliferation. At any timepoint after injury studied, the number of Ki67-positive cells in
longitudinal sections of L1y/ mice exceeded by at least two-fold
the numbers in L1y/+ mice (Fig. 8D–F0 ). To verify the possibility
that most of these proliferating cells are Schwann cells, we
performed co-localization experiments in sections of the femoral
nerve 5 days after injury using antibodies against Ki67
and Schwann cell markers, S-100 and GFAP (Fig. 1D). No
co-localization of Ki67+ nuclei and S-100+ (Fig. 9A–A0 ) or GFAP+
(not shown) cell profiles was found. However, most Ki67+ nuclei
in both genotypes appeared to co-localize with laminin, a basal
L1 and nerve regeneration
Brain 2009: 132; 2180–2195
| 2189
Figure 7 Analysis of motoneuron soma size and perisomatic nerve terminals. The estimation was performed only for correctly
projecting motoneurons (retrogradely labelled with Fast Blue, but not with fluoro-ruby, applied to the motor and sensory nerve branch,
respectively, A). Confocal images (1-mm thick optical slices) show VGAT+ (B) and ChAT+ (C) terminals around motoneuron cell bodies
(asterisks) in sections from an L1y/+ mouse studied 12 weeks after nerve repair. Scale bar indicates 25 mm for A–C. (D–F) Soma area of
ChAT+ and VGAT+ motoneurons (D) and linear densities of VGAT+ (E) and ChAT+ (F) puncta in non-injured mice (subjected to
retrograde tracing only) and mice studied 12 weeks after nerve repair (‘Injured’). Shown are group mean values + SEM values calculated
from individual mean values. Between 200 and 350 cells from 5 to 12 animals were analysed per group and parameter. Asterisks and
hash indicate differences between non-injured and injured groups of the same genotype and between L1y/+ and L1y/ animals under
the same experimental conditions, respectively (P50.05, one-way ANOVA with Tukey’s post hoc test).
lamina component produced by Schwann cells (Fig. 9B–B0 ).
No co-localization of Ki67+ nuclei and the macrophage marker
F4/80 was observed (Fig. 9C–C0 ). Finally, analysis of apoptotic
cells using an antibody against activated caspase-3 did not reveal
differences between the genotypes at any time-point studied
(not shown).
Since the differentiation-associated proteins S-100 and GFAP
did not co-localize with Ki67, Schwann cell proliferation was quantified using an antibody against the p75 NGF receptor which is
expressed in proliferating Schwann cells. At 5 days after femoral
nerve transection (Fig. 1D), when cell proliferation was estimated
to be maximal (Fig. 8D–F0 ), the number of proliferating Schwann
cells (both p75+ and Ki67+, designated as type Ia in Fig. 10A) was
higher in L1y/ than in L1y/+ mice irrespective of whether this
number was normalized to proliferating non-Schwann cells (Ki67+,
p75, type IIa, Fig. 10B)—by 26%, to all nuclei in a nerve crosssection (by 36%, Fig. 10C) or nerve cross-sectional area (by 20%,
Fig. 10D). Thus, L1 deficiency leads to enhancement of Schwann
cell proliferation after nerve injury.
Enhanced proliferation of Schwann cells
from L1y/ mice in vitro
Immunofluorescence detection of BrdU and phase contrast
analysis revealed BrdU labelling in nuclei of two types of
cell in the cultures: elongated bi- or tripolar cells with an
oval soma and small nuclei, considered to be Schwann cells
(Brockes et al., 1979; Scarpini et al., 1986; Honkanen et al.,
2007), and fibroblast-like cells with flattened, polymorphic appearance and large round nuclei (Fig. 11A–C). Quantitative analysis
revealed that a higher proportion of the cells with Schwann
cell morphology had incorporated BrdU in cultures derived from
L1y/ mice compared with L1y/+ cultures (57 and 39%, respectively, Fig. 11D), a finding supporting the observations made
in vivo.
Discussion
The results of this study show that L1 deficiency in mice has a
favourable impact on functional recovery after peripheral nerve
damage. Improved functional outcome was paralleled by positive
effects on regeneration including attenuated post-traumatic loss of
motoneurons, enhanced precision of reinnervation of the femoral
motor, and thus appropriate, nerve branch by motoneuron axons,
larger cell bodies of regenerated motoneurons and diminished loss
of inhibitory synaptic input to motoneurons. These phenomena are
associated with enhanced proliferation, and thus higher numbers,
of the growth-supportive denervated Schwann cells in injured
nerves of L1y/ mice.
2190
| Brain 2009: 132; 2180–2195
D. Guseva et al.
Figure 8 Analysis of Schwann cells proliferation in vivo. Longitudinal cryostat sections of 5 mm thickness from distal stumps of
transected femoral nerves of L1y/ mice (A–F) and L1y/+ littermates (A0 –F0 ) dissected 3–7 days after injury (Fig. 1D). Staining for the
Schwann cell marker S-100 (A–C0 ) reveals higher numbers of S-100+ cells in L1y/ mice (A–C) than in L1y/+ littermates (A0 –C0 ).
Staining of sections serial to those shown in panels A–C0 with an antibody against the proliferation antigen Ki67 (D–F0 ) reveals higher
numbers of proliferating cells in L1y/ mice than in L1y/+ mice at all time-points studied. Scale bars in A and D indicate 50 and
100 mm for A–C0 and D–F0 , respectively.
Functional and structural deficits in
uninjured L1y/ mice
Analyses of non-injured L1y/ mice revealed deficits, as compared
with L1y/+ littermates, in numbers of quadriceps motoneurons
and numbers of myelinated axons in the motor and sensory
branches of the femoral nerve. These differences could be
caused by target factors such as, for example, reduced muscle
mass despite normal body weight. Since L1 has been found to
support neuronal survival in vitro (Hulley et al., 1998; Chen
et al., 1999; Loers et al., 2005), we favour the view that
L1 deficiency in motoneurons leads to loss of motoneurons and
sensory neurons which in turn results in lower numbers of axons in
the peripheral nerves. Deficient motoneuron numbers may also
contribute to the mild gait deficits observed in our L1y/ mice
on 129/SvJ/NMRI background and previously in L1y/ mice on
a C57BL/6J/129S7 genetic background (strain B6; 129S7L1camtm1Sor/J, Jackson Laboratory, Bar Harbor, ME, USA,
Jakeman et al., 2006). If numbers of muscle fibres are not reduced
by ablation of L1, motoneuron deficit would lead to an increase
in motor unit size (number of muscle fibres innervated by one
motoneuron) in L1y/ mice. In association with abnormal sensory
inputs, this could contribute to reduced precision of motor control.
In line with the idea of increased motor unit size are the observations of larger somata of retrogradely labelled quadriceps
motoneurons and increased axonal diameters in the motor
branch of the femoral nerve of L1y/ mice. In contrast to other
deficits, non-injured L1y/ mice had the same myelination as
L1y/+ animals in both branches of the quadriceps nerve estimated
by g-ratio.
Enhanced outcome of nerve injury in
L1y/ mice
Previous studies have shown that homophilic L1 interactions
promote neurite outgrowth on cultured Schwann cells (Bixby
et al., 1988; Kleitman et al., 1988; Seilheimer and Schachner,
1988). Therefore, our initial expectation was to find at least
delayed recovery, if not inferior final outcome in L1y/ mice.
We surprisingly observed better locomotor recovery in L1y/
mice than in L1y/+ littermates. L1y/ mice showed subtle
abnormalities in gait prior to operation. However, the degree of
L1 and nerve regeneration
Brain 2009: 132; 2180–2195
| 2191
L1 and synaptic connectivity of
motoneurons
Figure 9 Analysis of proliferating cells in vivo. Cross-sections
from the quadriceps (motor) branch (on the left hand side in
each pair of sections) and the saphenous (sensory) branch
of the femoral nerve of L1y/ mouse (A–C) and an L1y/+
littermate (A0 –C0 ) cut from nerve segments distal from the
site of injury at 5 days after injury (Fig. 1D). Ki67+ nuclei
co-localize with laminin (B and B0 ) but not with S-100
(A and A0 ) or with the macrophage marker F4/80 (C and C0 ).
Scale bar indicates 25 mm.
impairment after injury was similar in L1y/ and L1y/+ mice as
indicated by the parallel and equal in extent changes occurring
between 0 and 7 days in both genotypes (Fig. 2E and F).
Therefore, L1y/ and L1y/+ mice suffer from the injury to the
same extent. Better functional recovery in L1y/ mice became
apparent first at 4 weeks after injury indicating lack of advantages
during the early phase of regeneration. At 12 weeks after injury,
motoneuron survival/axonal regrowth, as assessed by retrograde
labelling, was also better in L1y/ mice. Compared with noninjured mice of the same genotype, number of regenerated
retrogradely labelled motoneurons was reduced in L1y/+ but not
in L1y/ mice. Also, PMR after nerve injury was more
pronounced and somata of regenerated motoneurons were
larger in L1y/ than in L1y/+ mice. Similar effects on motoneuron
survival and regrowth and a similar time-course of functional
improvement were found after femoral nerve injury and application of a peptide mimic of the HNK-1 (human natural killer cell)
glycan in mice (Simova et al., 2006) which is expressed in
myelinated axons of the motor, but not sensory branches of the
mouse femoral nerve. We interpret such findings as indications of
trophic effects on motoneurons after axotomy, an interpretation in
line with the observation of enhanced Schwann cell proliferation in
L1y/ mice observed here. It is, however, also conceivable, that
the enhanced functional improvement in L1y/ mice is additionally related to attenuated susceptibility to injury of motoneurons
that have survived as the most resistant ones to developmental
elimination.
This study provides also, for the first time, data on perisomatic
coverage of motoneurons by VGAT+ (glycinergic and
GABAergic) and ChAT+ (cholinergic) axon terminals after nerve
injury. Considering the results for L1y/+ mice, two observations
appear important. First, the number of cholinergic boutons on
motoneurons was similar before and after injury. This is in contrast
to the dramatic reduction of this synaptic input to lumbar motoneurons after thoracic spinal cord injury in mice which correlates
with reduced locomotor recovery (Apostolova et al., 2006;
Jakovcevski et al., 2007). These two observations suggest that
injury-induced alterations in the modulatory cholinergic innervation of motoneurons (Miles et al., 2007) do not contribute to
motor deficits after nerve regeneration. Second, regenerated
motoneurons were deficiently innervated by inhibitory terminals
at 12 weeks after injury compared with non-injured motoneurons
in L1y/+ mice. This deficit is likely related to functional
impairments after nerve regeneration since perisomatic inhibition
is crucial for normal motoneuron function. Considering genotype
specific effects, it is interesting that non-injured L1y/ motoneurons were surrounded by more cholinergic and less inhibitory
boutons than L1y/+ motoneurons. At present it cannot be
judged whether these differences in synaptic coverage represent
compensations for or contribute to functional impairments in noninjured L1y/ mice. And finally, no loss of inhibitory terminals
in regenerated L1y/ motoneurons compared with non-injured
L1y/ motoneurons was found—in contrast to L1y/+ mice—a
feature that might favour enhanced locomotor recovery. The
combined observations indicate that L1 deficiency causes alterations, as compared with L1y/+ mice, in synaptic connectivity of
motoneurons in the non-injured spinal cord and after peripheral
nerve regeneration.
Enhanced Schwann cell proliferation
as a mechanism underlying enhanced
functional recovery in L1y/ mice
After nerve injury, the Schwann cells in the distal nerve stump
switch their mode from myelination of electrically active axons
to growth support for regenerating axons (Mirsky and Jessen,
1999; Scherer and Salzer, 2001). Denervated Schwann cells
proliferate, align in chains, bands of Büngner, inside endoneurial
tubes (basal laminae deposited by Schwann cells) and act as
cellular and molecular guides for the regenerating axons.
Moreover, the switch in Schwann cell phenotype after injury
is associated with downregulation of myelin-associated genes,
up-regulation of several growth-associated genes, and increased
production of neurotrophic factors and extracellular matrix
proteins supporting neuronal survival and axonal regrowth
(Fu and Gordon, 1997; Scherer and Salzer, 2001; Moran and
Graeber, 2004). If this support is reduced by, for instance, longterm denervation during adulthood or persistence of Schwann cell
immaturity in early postnatal animals, regeneration is compromised
(Fenrich and Gordon, 2004; Moran and Graeber, 2004). In light of
2192
| Brain 2009: 132; 2180–2195
D. Guseva et al.
Figure 10 Analysis of Schwann cell proliferation in vivo. Cross-section from the quadriceps (motor) branch of the femoral nerve of an
L1y/+ mouse (A) cut 3 mm distal from the site of injury at 5 days after injury (Fig. 1D) stained for the Schwann cell antigen p75 and
the proliferation marker Ki67. Four types of cell were distinguished and quantified in 10 space-serial sections per animal: p75+ cells with
Ki67+ nuclei (type Ia, proliferating Schwann cells), p75+ cells with Ki67 nuclei (Ib, non-proliferating Schwann cells), p75 cells with
Ki67+ nuclei (IIa, proliferating non-Schwann cells) and p75 cells with Ki67 nuclei (IIb, non-proliferating non-Schwann cells). The
numbers of proliferating Schwann cells (type Ia) were significantly higher in L1y/ than in L1y/+ mice (*P50.05, t-test, n = 3 animals
per genotype) as indicated by their ratio to proliferating non-Schwann cells (type IIa, B), and numbers per section normalized to the
number of all nuclei (C) or nerve cross-sectional area (D). The values shown in B–D are group mean values (+SEM) calculated
from individual mean values (10 sections analysed per animal). Scale bar in A indicates 50 mm.
this knowledge, better regeneration and functional recovery in
L1y/ mice can be explained by enhanced Schwann cell proliferation during the period of Wallerian degeneration. Higher number
of Schwann cells may provide enhanced trophic support by neurotrophins, cytokines and other factors thus positively influencing,
not only motoneurons, but also sensory neurons and improve their
survival and promote the restoration of afferent connections.
A relationship between L1 expression and Schwann cell proliferation has not been reported or considered previously. The only
exception is the finding that reduction of Schwann cell proliferation by serum deprivation in vitro augments L1 expression
(Takeda et al., 2001). One possible explanation for the effect of
L1 ablation on cell proliferation in vivo is delayed differentiation,
and thus prolonged proliferation, of Schwann cells after contact
with regrowing axons as a result of L1 deficiency in both
interacting cell types. Disruption of homophilic L1 interactions
(Bixby et al., 1988; Kleitman et al., 1988; Seilheimer and
Schachner, 1988) may underlie such an effect. While this hypothesis has to be tested, our experiments with primary Schwann cell
cultures, which are devoid of nerve cells but contain L1-negative
fibroblast-like cells, indicate that L1-mediated control of cell
proliferation is exerted via Schwann cell—extracellular matrix
interactions or cis-interactions between L1 and other molecules
on the Schwann cell membrane. One possibility is that L1 suppresses cell proliferation by binding to integrins on the surface of
Schwann cells. For example, interaction of L1 with a7b1-integrin,
a receptor for laminin-2 on Schwann cells (Chernousov et al.,
2007), through RGD motifs in the sixth immunoglobulin domain
of L1, can modulate integrin-mediated signalling and thus cell
proliferation (Itoh et al., 2004, 2005; Armstrong et al., 2007;
Maness and Schachner, 2007). Another possibility is that loss of
the interaction of L1 with the fibroblast growth factor receptor 1
L1 and nerve regeneration
Brain 2009: 132; 2180–2195
| 2193
(FGFR1) in L1y/ Schwann cells promotes proliferation (Dong
et al., 1997). Upon fibronectin type III domain-mediated binding
of L1 to FGFR1, the receptor is phosphorylated and this favours
differentiation and axonal growth in neurons rather than proliferation (Loers et al., 2005; Kulahin et al., 2008). The importance of
FGF-2 signalling in the control of Schwann cell proliferation has
previously been demonstrated using mice overexpressing FGF-2
(Jungnickel et al., 2006). These transgenic mice show enhanced
Schwann cell proliferation but, in contrast to our L1y/ mice, no
superior functional outcome of peripheral nerve injury compared
with wild-type mice was detected. These observations do not
contradict our hypothesis that increased numbers of Schwann
cells during Wallerian degeneration favour the final outcome of
nerve injury since FGF-2 overexpression stimulates signalling not
only via the regeneration-promoting FGFR1, but also via FGFR3,
which has a negative impact on nerve regeneration (Jungnickel
et al., 2004).
L1 and myelination
Figure 11 Analysis of Schwann cell proliferation in vitro.
Primary culture from peripheral nerves of an L1y/+ mouse
treated with BrdU for 48h to label proliferating cells (the same
field on panels A–C). Schwann cells were identified by phase
contrast microscopy as elongated bi- or tripolar cells with an
oval luminescent soma (A and C). The number of Schwann
cells containing BrdU-labelled nuclei (two cells indicated by
arrows, B and C) was estimated in cultures of both genotypes
and expressed as percentage of all Schwann cells. Large cells
with BrdU-positive nuclei (B) and barely visible cell bodies (A)
were classified as fibroblast-like cells and were not analysed.
The proportion of BrdU-labelled Schwann cells was significantly
higher in cultures derived from L1y/ mice than in L1y/+
cultures (D, *P50.05, t-test). Values shown in D are
means + SEM.
Although L1 is involved in myelination in vitro (Seilheimer et al.,
1989; Wood et al., 1990; Coman et al., 2005), peripheral nerve
myelination in L1y/ mice is normal (Dahme et al., 1997; Haney
et al., 1999). Here, we found similar degrees of myelination in
L1y/ and L1y/+ mice for both non-injured and regenerated
nerves. These findings are in agreement with the conclusion of
Martini and Schachner (1986, 1988) based on analyses of
L1 expression during nerve development and regeneration, that
L1 is not essential for the maintenance of compact myelin.
However, the expression pattern observed in these studies suggested that L1 is involved in initial axon-Schwann cell interactions
and at the onset of myelination. Our findings, therefore, do not
preclude a role of L1 in myelination. During normal development
until adulthood and within a 12-week recovery period after nerve
injury, compensatory mechanisms may overcome an L1-related
delay in myelination.
In conclusion, the results of this study indicate that L1, in
contrast to its role in the injured CNS, has a negative impact on
peripheral nerve regeneration, implicating L1’s suppression of
Schwann cell proliferation and thus reduced trophic support that
proliferating, denervated Schwann cells convey to neurons.
Ensuing L1-related enhancement of differentiation may then be
counteractive to regeneration after injury in the adult PNS.
These findings warrant experiments designed to test the potential
of a functional, but transient block of L1 expression by Schwann
cells in the regenerating nerve, using, for example, neutralizing
antibodies, to improve of the outcome of nerve injury. An
alternative strategy would be to transiently stimulate Schwann
cell proliferation, and thereby improve the functional outcome
after injury, by applying Schwann cell mitogens during the
period of Wallerian degeneration.
Acknowledgements
We are grateful to Gabriele Loers for providing L1 antibody and
help with Schwann cells culture, Peggy Puthoff for supplying
2194
| Brain 2009: 132; 2180–2195
transgenic mice, Emanuela Szpotowicz for technical assistance and
Achim Dahlmann for genotyping. M.S. is New Jersey Professor of
Spinal Cord Research.
Funding
Hertie Stiftung (1.01.1/06/005).
References
Ahlborn P, Schachner M, Irintchev A. One hour electrical stimulation
accelerates functional recovery after femoral nerve repair. Exp
Neurol 2007; 208: 137–44.
Anderson PN, Campbell G, Zhang Y, Lieberman AR. Cellular and
molecular correlates of the regeneration of adult mammalian CNS
axons into peripheral nerve grafts. Prog Brain Res 1998; 117: 211–32.
Anderson RB, Newgreen DF, Young HM. Neural crest and the development of the enteric nervous system. Adv Exp Med Biol 2006; 589:
181–96.
Apostolova I, Irintchev A, Schachner M. Tenascin-R restricts
posttraumatic remodeling of motoneuron innervation and functional
recovery after spinal cord injury in adult mice. J Neurosci 2006; 26:
7849–59.
Appel F, Holm J, Conscience JF, von Bohlen and Halbach F, Faissner A,
James P, Schachner M. Identification of the border between fibronectin type III homologous repeats 2 and 3 of the neuronal cell adhesion
molecule L1 as neurite outgrowth promoting and signal transducing
domain. J Neurobiol 1995; 28: 297–312.
Armstrong SJ, Wiberg M, Terenghi G, Kingham PJ. ECM molecules mediate both Schwann cell proliferation and activation to enhance neurite
outgrowth. Tissue Eng 2007; 13: 2863–70.
Becker CG, Lieberoth BC, Morellini F, Feldner J, Becker T, Schachner M.
L1.1 is involved in spinal cord regeneration in adult zebrafish.
J Neurosci 2004; 24: 7837–42.
Becker T, Lieberoth BC, Becker CG, Schachner M. Differences in the
regenerative response of neuronal cell populations and indications
for plasticity in intraspinal neurons after spinal cord transection in
adult zebrafish. Mol Cell Neurosci 2005; 30: 265–78.
Bixby JL, Lilien J, Reichardt LF. Identification of the major proteins
that promote neuronal process outgrowth on Schwann cells in vitro.
J Cell Biol 1988; 107: 353–61.
Brockes JP, Fields KL, Raff MC. Studies on cultured rat Schwann cells. I.
Establishment of purified populations from cultures of peripheral nerve.
Brain Res 1979; 165: 105–18.
Brushart TM. Preferential reinnervation of motor nerves by regenerating
motor axons. J Neurosci 1988; 8: 1026–31.
Chen J, Bernreuther C, Dihne M, Schachner M. Cell adhesion molecule
L1-transfected embryonic stem cells with enhanced survival support
regrowth of corticospinal tract axons in mice after spinal cord injury.
J Neurotrauma 2005; 22: 896–906.
Chen J, Wu J, Apostolova I, Skup M, Irintchev A, Kügler S, et al. Adenoassociated virus-mediated L1 expression promotes functional recovery
after spinal cord injury. Brain Res 2007; 130: 954–69.
Chen S, Mantei N, Dong L, Schachner M. Prevention of neuronal cell
death by neural adhesion molecules L1 and CHL1. J Neurobiol 1999;
38: 428–39.
Chernousov MA, Kaufman SJ, Stahl RC, Rothblum K, Carey DJ.
Alpha7beta1 integrin is a receptor for laminin-2 on Schwann cells.
Glia 2007; 55: 1134–44.
Coman I, Barbina G, Charlesa P, Zalc B, Lubetzki C. Axonal signals in
central nervous system myelination, demyelination and remyelination.
J Neurol Sci 2005; 233: 67–71.
D. Guseva et al.
Dahme M, Bartsch U, Martini R, Anliker B, Schachner M, Mantei N.
Disruption of the mouse L1 gene leads to malformations of the
nervous system. Nat Genet 1997; 17: 346–9.
Daniloff JK, Levi G, Grumet M, Rieger F, Edelman GM. Altered expression of neuronal cell adhesion molecules induced by nerve injury and
repair. J Cell Biol 1986; 103: 929–45.
Deumens R, Lübbers M, Jaken RJ, Meijs MF, Thurlings RM, Honig WM,
et al. Mice lacking L1 have reduced CGRP fibre in-growth into spinal
transection lesions. Neurosci Lett 2007; 420: 277–81.
Dong Z, Dean C, Walters JE, Mirsky R, Jessen KR. Response of Schwann
cells to mitogens in vitro is determined by pre-exposure to serum, time
in vitro, and developmental age. Glia 1997; 20: 219–30.
Eberhardt KA, Irintchev A, Al-Majed AA, Simova O, Brushart TM,
Gordon T, et al. BDNF/TrkB signaling regulates HNK-1 carbohydrate
expression in regenerating motor nerves and promotes functional
recovery after peripheral nerve repair. Exp Neurol 2006; 198: 500–10.
Faissner A, Kruse J, Nieke J, Schachner M. Expression of neural cell
adhesion molecule L1 during development, in neurological mutants
and in the peripheral nervous system. Brain Res 1984; 317: 69–82.
Fenrich K, Gordon T. Canadian Association of Neuroscience
review: axonal regeneration in the peripheral and central nervous
systems – current issues and advances. Can J Neurol Sci 2004; 31:
142–56.
Fu SY, Gordon T. The cellular and molecular basis of peripheral nerve
regeneration. Mol Neurobiol 1997; 14: 67–116.
Haney CA, Sahenk Z, Li C, Lemmon VP, Roder J, Trapp BD. Heterophilic
binding of L1 on unmyelinated sensory axons mediates Schwann cell
adhesion and is required for axonal survival. J Cell Biol 1999; 146:
1173–83.
Honkanen H, Lahti O, Nissinen M, Myllylä RM, Kangas S, Päiväläinen S,
et al. Isolation, purification and expansion of myelination-competent,
neonatal mouse Schwann cells. Eur J Neurosci 2007; 26: 953–64.
Hortsch M. Neural cell adhesion molecules – brain glue and much more!
Front Biosci 2003; 8: d357–9.
Hulley P, Schachner M, Lübbert H. L1 neural cell adhesion molecule is
a survival factor for fetal dopaminergic neurons. J Neurosci Res 1998;
53: 129–34.
Irintchev A, Draguhn A, Wernig A. Reinnervation and recovery of mouse
soleus muscle after long-term denervation. Neuroscience 1990; 39:
231–43.
Irintchev A, Rollenhagen A, Troncoso E, Kiss JZ, Schachner M. Structural
and functional aberrations in the cerebral cortex of tenascin-Cdeficient mice. Cereb Cortex 2005b; 15: 950–62.
Irintchev A, Simova O, Eberhardt KA, Morellini F, Schachner M. Impacts
of lesion severity and TrkB deficiency on functional outcome of
femoral nerve injury assessed by a novel single-frame motion analysis
in mice. Eur J Neurosci 2005a; 22: 802–8.
Itoh K, Cheng L, Kamei Y, Fushiki S, Kamiguchi H, Gutwein P, et al.
Brain development in mice lacking L1-L1 homophilic adhesion. J Cell
Biol 2004; 165: 145–54.
Itoh K, Fushiki S, Kamiguchi H, Arnold B, Altevogt P, Lemmon V.
Disrupted Schwann cell-axon interactions in peripheral nerves of
mice with altered L1-integrin interactions. Mol Cell Neurosci 2005;
30: 624–9.
Jakeman LB, Chen Y, Lucin KM, McTigue DM. Mice lacking L1 cell
adhesion molecule have deficits in locomotion and exhibit enhanced
corticospinal tract sprouting following mild contusion injury to the
spinal cord. Eur J Neurosci 2006; 23: 1997–2011.
Jakovcevski I, Wu J, Karl N, Leshchyns’ka I, Sytnyk V, Chen J, et al. Glial
scar expression of CHL1, the close homolog of the adhesion molecule
L1, limits recovery after spinal cord injury. J Neurosci 2007; 27:
7222–33.
Jungnickel J, Gransalke K, Timmer M, Grothe C. Fibroblast growth factor
receptor 3 signaling regulates injury-related effects in the peripheral
nervous system. Mol Cell Neurosci 2004; 25: 21–9.
Jungnickel J, Haase K, Konitzer J, Timmer M, Grothe C. Faster nerve
regeneration after sciatic nerve injury in mice over-expressing basic
fibroblast growth factor. J Neurobiol 2006; 66: 940–8.
L1 and nerve regeneration
Kleitman N, Simon DK, Schachner M, Bunge RP. Growth of embryonic
retinal neurites elicited by contact with Schwann cell surfaces is
blocked by antibodies to L1. Exp Neurol 1988; 102: 298–306.
Kulahin N, Li S, Hinsby A, Kiselyov V, Berezin V, Bock E. Fibronectin type
III (FN3) modules of the neuronal cell adhesion molecule L1 interact
directly with the fibroblast growth factor (FGF) receptor. Mol Cell
Neurosci 2008; 37: 528–36.
Loers G, Chen S, Grumet M, Schachner M. Signal transduction pathways
implicated in neural recognition molecule L1 triggered neuroprotection
and neuritogenesis. J Neurochem 2005; 92: 1463–76.
Maness PF, Schachner M. Neural recognition molecules of the
immunoglobulin superfamily: signaling transducers of axon guidance
and neuronal migration. Nat Neurosci 2007; 10: 19–26.
Martini R. Expression and functional roles of neural cell surface molecules
and extracellular matrix components during development and
regeneration of peripheral nerves. J Neurocytol 1994; 23: 1–28.
Martini R, Schachner M. Immunoelectron microscopic localization of
neural cell adhesion molecules (L1, N-CAM, and MAG) and their
shared carbohydrate epitope and myelin basic protein in developing
sciatic nerve. J Cell Biol 1986; 103: 2439–48.
Martini R, Schachner M. Immunoelectron microscopic localization of
neural cell adhesion molecules (L1, N-CAM, and myelin-associated
glycoprotein) in regenerating adult mouse sciatic nerve. J Cell Biol
1988; 106: 1735–46.
Mason MR, Lieberman AR, Anderson PN. Corticospinal neurons
up-regulate a range of growth-associated genes following intracortical,
but not spinal, axotomy. Eur J Neurosci 2003; 18: 789–802.
Miles GB, Hartley R, Todd AJ, Brownstone RM. Spinal cholinergic
interneurons regulate the excitability of motoneurons during
locomotion. Proc Natl Acad Sci USA 2007; 104: 2448–53.
Mirsky R, Jessen KR. The neurobiology of Schwann cells. Brain Pathol
1999; 9: 293–311.
Moran LB, Graeber MB. The facial nerve axotomy model. Brain Res Rev
2004; 44: 154–78.
Nieke J, Schachner M. Expression of the neural cell adhesion molecules
L1 and N-CAM and their common carbohydrate epitope L2/HNK-1
during development and after transection of the mouse sciatic nerve.
Differentiation 1985; 30: 141–51.
Rolf B, Kutsche M, Bartsch U. Severe hydrocephalus in L1-deficient mice.
Brain Res 2001; 891: 247–52.
Roonprapunt C, Huang W, Grill R, Friedlander D, Grumet M, Chen S,
et al. Soluble cell adhesion molecule L1-Fc promotes locomotor
recovery in rats after spinal cord injury. J Neurotrauma 2003; 20:
871–82.
Brain 2009: 132; 2180–2195
| 2195
Saghatelyan AK, Nikonenko AG, Sun M, Rolf B, Putthoff P, Kutsche M,
et al. Reduced GABAergic transmission and number of hippocampal
perisomatic inhibitory synapses in juvenile mice deficient in the neural
cell adhesion molecule L1. Mol Cell Neurosci 2004; 26: 191–203.
Scarpini E, Meola G, Baron P, Beretta S, Velicogna M, Scarlato G. S-100
protein and laminin: immunocytochemical markers for human
Schwann cells in vitro. Exp Neurol 1986; 93: 77–83.
Scherer SS, Salzer JL. Axon-Schwann cell interactions during
peripheral nerve degeneration and regeneration. In: Jessen KL,
Richardson WD, editors. Glial cell development: basic principles and
clinical relevance. 2nd edn., Oxford, UK: Oxford University Press;
2001. p. 299–330.
Seilheimer B, Persohn E, Schachner M. Antibodies to the L1 adhesion
molecule inhibit Schwann cell ensheathment of neurons in vitro. J Cell
Biol 1989; 109: 3095–103.
Seilheimer B, Schachner M. Studies of adhesion molecules mediating
interactions between cells of peripheral nervous system indicate a
major role for L1 in mediating sensory neuron growth on Schwann
cells in culture. J Cell Biol 1988; 107: 341–51.
Simova O, Irintchev A, Mehanna A, Liu J, Dihne M, Bachle D, et al.
Carbohydrate mimics promote functional recovery after peripheral
nerve repair. Ann Neurol 2006; 60: 430–37.
Styren SD, Miller PD, Lagenaur CF, DeKosky ST. Alternate strategies
in lesion-induced reactive synaptogenesis: differential expression
of L1 in two populations of sprouting axons. Exp Neurol 1995; 131:
165–73.
Takeda Y, Murakami Y, Asou H, Uyemura K. The roles of cell adhesion
molecules on the formation of peripheral myelin. Keio J Med 2001; 50:
240–8.
Thelin J, Waldenström A, Bartsch U, Schachner M, Schouenborg J. Heat
nociception is severely reduced in a mutant mouse deficient for the L1
adhesion molecule. Brain Res 2003; 965: 75–82.
Wood PM, Schachner M, Bunge RP. Inhibition of Schwann cell myelination in vitro by antibody to the L1 adhesion molecule. J Neurosci 1990;
10: 3635–45.
Woolhead CL, Zhang Y, Lieberman AR, Schachner M, Emson PC,
Anderson PN. Differential effects of autologous peripheral nerve
grafts to the corpus striatum of adult rats on the regeneration of
axons of striatal and nigral neurons and on the expression of
GAP-43 and the cell adhesion molecules N-CAM and L1. J Comp
Neurol 1998; 391: 259–73.
Zhang Y, Bo X, Schoepfer R, Holtmaat AJ, Verhaagen J, Emson PC, et al.
Growth-associated protein GAP-43 and L1 act synergistically to
promote regenerative growth of Purkinje cell axons in vivo. Proc
Natl Acad Sci USA 2005; 102: 14883–8.