| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Center for Paralysis Research, Department of Basic Medical Sciences, School of Veterinary Medicine, Purdue University, West Lafayette, Indiana 47907, USA
1Correspondence: Center for Paralysis Research, Department of Basic Medical Sciences, Purdue University, West Lafayette, Indiana 47907, USA. E-mail: cpr{at}vet.purdue.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: nerve fusion • nerve repair • neurotrauma
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
. This loss of sensory and
motor communication across the injury site can produce a permanent
paralysis and loss of sensation in regions below the level of the
spinal injury. Furthermore, it is clear the most damaging aspect of
progressive ‘secondary injury’ (3)
of spinal cord
parenchyma relative to the loss of behavioral functioning is its effect
on white matter. Localized mechanical, biochemical, and
anoxic/ischemic injury to white matter may be sufficient to cause the
failure of axolemmas to function as a barrier or fence to the
unregulated exchange of ions (1)
. This in turn compromises
both the structural integrity of this region of the nerve fiber and its
ability to conduct impulses along the cable. For example, elevated
intracellular Ca2+ induces depolymerization of
microtubules and microfilaments, producing a focal destruction of the
cytoskeleton (2
, 4
, 5)
. When K+
rushes down its electrochemical gradient out of the cell, the resultant
elevated extracellular concentration contributes to localized
conduction block (1
, 6)
. It would be of particular importance to interrupt this progression of events after acute injury to the spinal cord. This could theoretically be accomplished by immediately repairing or sealing regions of compromised membrane with hydrophilic polymers or surfactants, which could retard or reverse the permeabilization of nerve fiber membranes. This in turn might both rescue these nerve fibers from further degeneration and restore variable levels of physiological and behavioral function.
In recent years the administration of several types of polymers and
surfactants to injured cells has been shown to seal or repair their
membranes, reversing the progressive permeabilization produced by the
insult. For example, application of the molecular surfactant Poloxamine
1107 markedly reduced the leakage of hemoglobin from erythrocytes
damaged by radiation (7)
, whereas Poloxamer 188
facilitated the functional recovery of fibroblasts exposed to
potentially fatal heat shock (8)
. Membrane fusogens such
as polyethylene glycol (PEG) are well known to fuse the membranes of
numerous cells in culture, producing giant multinucleated ones
(9
10
11
12)
. This class of large hydrophilic molecules can
likewise reverse the permeabilization of ruptured cell membranes as
well as anatomically reconnect their severed processes.
Using a new in vitro isolation and recording chamber
(6
, 13
, 14)
, we have demonstrated that a local application
of an aqueous solution of PEG can functionally and anatomically
reconnect completely severed strips of guinea pig spinal cord ventral
white matter. This treatment produced axonal fusion associated with the
immediate recovery of compound action potential (CAP) propagation
through the original plane of transection (14)
. In
addition, PEG can rapidly restore physiological functioning in 100% of
severely crushed spinal cords similarly monitored in
isolation(15)
. Since most clinical spinal cord injuries
are compressive/contusive injuries (1)
, we wanted to know
whether the striking in vitro response to PEG could be
duplicated in an in vivo spinal cord injury model.
Here we show that in adult guinea pigs possessing severe and standardized spinal cord compression, CAP propagation through the lesion can be restored within minutes after a brief application of PEG. This physiological recovery was also associated with a behavioral recovery. To further explore the possible clinical significance of these data, we also delayed the application of PEG for ~8 h after spinal injury and again observed similar recoveries in the PEG-treated animals, but not in a sham-treated control group.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 17)
. Subsequently, a constant-displacement 15 s
compression of the spinal cord was performed using a modified forceps
possessing a detente (18)
. In this experiment, the
lesioning procedure had previously been calibrated to produce an
immediate and total loss of CAP conduction through the injury and
behavioral functioning of the cutaneous trunci muscle reflex (CTM; see
below). For some somatosensory-evoked potential (SSEP) measurements or
to sedate animals for behavioral testing and videotaping, guinea pigs
were injected with 0.1 cc Na+ pentobarbital, 50
mg/ml. Surgery and functional testing were carried out under protocols
approved by the Purdue University Animal Care and Use Committee in
accordance with federal, state, and university guidelines governing
animal use in research.
PEG application
An aqueous solution of PEG (either 400 or 1800 daltons, 50% by
weight in distilled water) was applied with a pipette to the exposed
injury for 2 min in experimental animals and then removed by
aspiration. As in prior in vitro experiments (14
, 15)
, we did not detect a difference in the response to these two
solutions, so these data are pooled in this report. The site of PEG
application was immediately lavaged with isotonic Krebs’ solution
(NaCl 124 mM, KCl 2 mM,
KH2PO4 1.24 mM,
MgSO4 1.3 mM, CaCl2 1.2 mM,
dextrose 10 mM, NaHCO3 26 mM, sodium ascorbate 10
mM) and any excess PEG and/or Krebs’ solution was removed by
aspiration. PEG was not applied to the injury in sham-treated animals;
however, the site was lavaged with Krebs’ solution, which was
subsequently removed by aspiration. The wounds were closed and animals
were kept warm until awaking with heat lamps. Guinea pigs were housed
individually and fed ad libidum.
In the first experiment, we attempted to repeat the remarkable complete
reversal of functional loss within minutes of severe spinal injury as
observed in in vitro trials (14
, 15)
. Thus, PEG
was applied within ~15 min of spinal cord compression (experimental
n=14, control n=11). In the second experiment,
PEG application was delayed for ~8 h (experimental n=11,
control n=11). The former groups were evaluated for ~4
days and the latter for ~1 month after PEG application. In both
experiments, documentation of CTM behavior was combined with
physiological recording.
An additional four PEG-treated animals were monitored for 1 day after injury, at which time their spinal cord was again exposed at the site of the original injury and crushed again at this location using the same technique as reported above.
Behavioral analysis of the CTM reflex
CTM behavior is observed as a corrugated rippling of back skin
in response to light tactile stimulation (Fig. 1B
). The behavior is dependent on afferent sensory projections
organized as a long tract of axons in each ventral funiculus of the
spinal cord, just lateral to the spinothalamic tract (19
, 20)
(Fig. 1A
). The reflex is bilaterally organized as
segmental receptive fields, displays little supraspinal control, and
usually is permanently lost after severe spinal injury, producing a
bilateral region of areflexia below the level of the lesion (17
, 19
, 20)
(Fig. 1A, C
). In such cases, recovery of the
CTM reflex in response to tactile or electrical stimulation within the
region of areflexia usually is not observed for the life of the animal.
The anatomy, physiology, and character of the CTM behavior, both normal
and in response to lesioning, have been reported in both rat and guinea
pigs (19
, 20)
.
|
To visualize and quantify the CTM behavior, the shaved back skin of
sedated guinea pigs was touched with a monofilament probe, producing
contraction of the skin in uninjured or intact receptive fields (Fig. 1B
). The boundary between responsive and unresponsive back
skin was marked onto the back skin with a marker while the entire study
period was videotaped from a platform-mounted camera above. Animals
were arranged on a background grid to facilitate the registration of
successive video images. Video images were acquired to an Intel Dual
Pentium Pro computer. Superimposing of images, the coloring of
receptive field boundaries made on the back skin of the animals during
CTM testing, and the general management of video images was performed
using Adobe Photoshop software. Final plates were constructed with
Microsoft PowerPoint software and printed on an Epson Stylus Color 800
printer. Quantitative planimetry of the unit area of receptive fields,
or regions of behavioral loss and recovery, was carried out using IP
Lab Spectrum software.
Physiological recording of SSEPs
A pair of subdermal electrodes stimulated nerve impulses from
the tibial nerve of the hind leg (stimuli trains in sets of 200 at 3
Hz; stimulus amplitude 
3 mA square wave, 200 µs duration).
Evoked volleys of CAPs were conducted into the spinal cord, projected
to, and recorded from the sensory cortex of the brain. Recording of the
nerve impulses at the brain used a pair of subdermal electrodes located
above the level of the contralateral cortex, with reference electrodes
located in the ipsilateral pinna of the ear. Stimulation, recording,
signal averaging, and the computer management of this physiological
data used a Nihon Kohden Neuropak 4 stimulator/recorder and PowerMac G3
computer.
Evoked CAPs are called SSEPs; such measurements were carried out in
every animal prior to spinal cord injury (Fig. 2
). In all animals (at any test period), the failure to record an SSEP
after stimulation of the tibial nerve was further confirmed to be due
to a lack of conduction through the spinal cord lesion by a control
test carried out on the same animal. In this procedure, the medial
nerve of the forelimb was stimulated, initiating evoked potentials in a
neural circuit unaffected by the crush injury (Fig. 2A-C
).
To perform this test, recording electrodes were left in place while
stimulating electrodes were relocated to stimulate the median nerve,
using identical parameters of stimulation.
|
Statistics
The Mann Whitney two-tailed test was used to compare the means
of the data derived from experimental and sham-treated groups. To
compare the proportions between groups, Fishers exact test was used.
All tests were performed using INSTAT software.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) was not
statistically different between either of the two experiments or
between sham-treated and PEG-treated guinea pigs in either experiment
(P>0.4, Student’s t test, two-tailed).
Only one animal died during the course of this study.
Behavioral loss and recovery of the CTM reflex
In both experiments, 19 of the 22 sham-treated animals did not
recover CTM functioning (Table 1
and Table 2
). During the first experiment, CTM functioning actually worsened by day
4 in two control animals (the region of CTM loss increased by 2% and
15% respectively; Table 1
). In contrast, CTM functioning recovered in
10 of 14 PEG-treated animals in the first experiment (~80%; Fig. 1D
, Table 1
) and in > 90% of experimental animals in
the second experiment. In all PEG-treated animals, the restored region
of CTM competent back skin was observed within the first day after
treatment and continued to increase in size with time (Tables 1
and 2)
.
For example, the average unit area of back skin recovering CTM
sensitivity nearly doubled from ~12% (day 1) to ~20% by 1 month
after application in the second experiment (Table 2)
. The increased
proportion of animals recovering CTM function and the average increase
in the areas of recovered CTM competent back skin in response to PEG
were both statistically significant (Tables 1
and 2)
.
|
|
Physiological measurements of conduction through the spinal cord
injury
Physiological measurements of SSEP conduction were performed in
every animal prior to spinal cord injury and within 5–15 min after
surgery (Fig. 2B, C
; Fig. 3
) to provide a basis for later comparison. In the uninjured animal,
SSEPs were typically observed to segregate into two peaks: early
arriving (latency ~ 20–30 ms) and a later arriving SSEP
(~35–45 ms; Fig. 2B, C
; Fig. 3
). In the first experiment,
subsequent records were taken at ~30 min, 1 h, 24 h, and 4
days after PEG treatment. In the second experiment, subsequent
measurements were made ~6–8 h, 18–24 h, 3 days, 2 wk, and 1 month
after the delayed application of PEG. In all animals, the failure to
record an SSEP after stimulation of the tibial nerve was further
confirmed to be due to a lack of conduction through the injury by a
control procedure carried out on the same animal, where the medial
nerve of the forelimb was stimulated. In all cases, this produced a
characteristic SSEP for this spinal circuit unaffected by the injury
(Fig. 2A-C
; Fig. 3
).
|
In this investigation, sham-treated animals never regained the ability to conduct SSEPs through the injury site
In the first experiment, a detectable SSEP was recorded within a
few minutes after PEG application. Quantitative evaluation of 10 of
these animal’s electrical records showed that SSEP amplitudes
continued to improve, averaging ~40% of their preinjury level and
displaying more typical latencies with time (Fig. 2C, D
).
Remarkably, within minutes of the spinal injury, the total loss of
physiological functioning was reversed in 23 of 25 PEG-treated animals.
In the two animals that did not immediately respond to PEG application,
SSEP recovery was later observed at the 2 wk time point (Fig. 3
, Table 2
). In the four animals whose recovered SSEPs were tested by reinjury,
the second compression of the spinal cord at the original injury site
completely eliminated recovered SSEPs, confirming these were conducted
through the lesion.
In summary, all 34 PEG-treated animals recovered SSEP conduction in
contrast to the complete failure of all control guinea pigs to conduct
evoked potentials through the lesion. Only 3 of 22 sham-treated animals
recovered CTM function in both experiments, whereas 20 of 25
PEG-treated animals recovered variable amounts of CTM functioning,
which continued to improve with time (Tables 1
, 2)
.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 19
, 21)
.
An 8 h delay in this application still resulted in a similar
recovery of these functions. In sharp contrast, sham-treated animals
never recovered the ability to conduct nerve impulses, and the minor
occurrence of spontaneous recovery of CTM function was rare compared to
the PEG-treated group. We chose to compress the spinal cord since this
is more typical of clinical injuries (1
, 18)
.
It is likely that this severe constant displacement injury produced
axotomy in an unknown proportion of nerve fibers. However, it is
unlikely that the ability of PEG to fuse the proximal and distal
segments of axons (14)
was responsible for the recovery of
physiological and behavioral function reported here. We have shown that
reconnection of transected axons by PEG requires precise alignment and
precise pressure applied to the carefully abutted proximal and distal
segments (14)
. This was not possible or attempted in this
study. Rather, we think the interaction of PEG with damaged axolemmas
of crushed spinal cords led to an immediate sealing of breaches in
these membranes, producing a reversal of permeabilization.
Molecular repair of cell membranes
Historically, PEG has been used to fuse many individual
cells in vitro into one large multinucleated cell
(9
10
11
12
, 22)
. This technique was an early advancement
allowing the transfer of genetic material between cells and the
production of giant cells, such as neurons, to facilitate
electrophysiological study of their membranes. Hence, the name
‘fusogen’ is sometimes applied to this class of water-soluble
polymers that are able to produce membrane union and the intermingling
of cytoplasms. PEG-mediated membrane fusion is still an important model
for endogenous biomembrane and vesicular fusion though its mode of
action is unclear (22
, 23)
. It is believed that PEG
induces a dehydration of closely approximate membranes, allowing their
structural components to resolve into each other (22)
.
After rehydration, complex molecular morphologies of the lamellae
apparently spontaneously reassemble based in part on the polar forces
associated with the aqueous phase of the membrane. Some have argued
that the PEG-mediated fusion process must involve first rupture and
then resealing of the adjacent bilayers (24
, 25)
whereas
others have opined that membrane rupture is not required for this
process (26)
.
Our preliminary in vitro evaluation of the anatomy of axonal
repair after mechanical compression (13)
has revealed that
a 2 min application of PEG produced sealing of membrane lesions at the
site of a standardized compression. Sealing was indicated by the
exclusion of horseradish peroxidase uptake by injured fibers in the
PEG-treated group compared to sham-treated spinal cords (to be reported
elsewhere). Such immediate repair of membrane breaches sufficient to
inhibit the uptake of large molecular weight dyes should also arrest or
reduce permeabilization allowing the nonspecific flux of ions across
it. We think it is this ‘sealing ’ behavior of PEG that both restores
excitability and reverses anatomical dissolution of the nerve fiber. It
may be that PEG functions as a detergent or sealant in a manner
hypothesized for surfactants such as the poloxamines, polaxomers, and
dextrans (27
, 28)
. These agents may cover or be absorbed
into regions of membrane damage where their hydrophobic cores may
interact with the lipid domain of the bilayer while their hydrophilic
tails extend into the adjacent aqueous domains of the membrane, somehow
binding, closing, or covering discontinuities in the bilayer
(29)
. Surfactant-mediated membrane sealing has been shown
to prevent myonecrosis and calcein leakage from damaged skeletal muscle
cells after electrical trauma (8
, 27
, 28)
, facilitate
recovery from lethal heat shock in fibroblasts (29)
, and
reduce tissue damage in testicular ischemia-reperfusion injury
(30)
. Even though the interaction of surfactants and large
hydrophilic molecules like PEG with cell membranes may or may not share
certain biophysical mechanisms of action, it is likely that their
biological/medical importance is this shared ability to seal
compromised cell membranes. Exploitation of this ability by using both
types of agents offers a new way of thinking about the treatment of
cell and tissue trauma.
Trauma to nervous tissue as a function of trauma to the single cell
It is well known that the anatomical consequences of
compressive/contusive injuries are more severe days rather than hours
after injury. This fact has led to the concept of ‘secondary injury’
and a proliferation of theories of how such delayed and progressive
destruction of spinal cord tissue may occur (1
, 3
, 31)
.
This pathophysiology is complex and incorporates biochemical
alterations (such as free radical-mediated damage, endogenous
neurotoxicity, and derangement of the ionic extracellular environments
inside and outside cells), progressive ischemic injury, and
cell-mediated damage—for example, ‘bystander’damage caused by
infiltration of healthy parenchyma by enormous numbers of macrophages
(32
, 33)
. Although these distinctions are important in
understanding spinal cord injury, it may be more convenient and
instructive to consider their consequences as end points in a final
common pathway that begins with permeabilization of the membranes
(2)
of neurons principally, but not exclusively. For
example, disruption of the axolemma may be so severe as to lead to
axotomy, but less severe or ‘nondisruptive ’ injury to the nerve
fiber may also do so (4
, 5)
. In both scenarios,
unregulated entry of Ca2+ into the cytosol
appears to be the seminal event in the progressive dissolution of the
axon. Some axons may survive intact but fail to conduct action
potentials across the region of damage (31)
. Conduction
block may result from focal demyelination, accompanied by a breakdown
in the ion exchange/ion exclusion properties of normal membrane.
Axotomy or conduction block both lead to a break in the conduction
pathway and functional deficits. It is at this level where the action
of fusogens and membrane sealants swiftly restrict or reduce ion
leakage into or out of injured cells helping to restore ionic
equilibrium, membrane excitability, and ultimately preserving crucial
anatomies.
Since the destruction of spinal cord white matter is dynamic, delayed,
progressive, and produced by different mechanisms of action, it is not
surprising that a delay in PEG treatment still produced recoveries in
physiological and behavioral function. The contribution to any
functional and physiological deficit arising from mechanical
compression of nervous tissue should be viewed as a balance between the
irreversible loss of critical anatomies and endogenous mechanisms of
repair such as spontaneous axonal sealing (34)
. It is
likely that PEG may facilitate several of these repair processes and at
different times after injury.
The immediate recovery of excitability implicates a sealing function of
PEG, but it is not a perfect seal. We have reported that the amplitudes
of CAPs restored by PEG treatment were nearly doubled by subsequent
application of the fast potassium channel blocker 4-aminopyridine
(15)
. Thus PEG repaired membranes probably remain somewhat
leaky to potassium, which pushes their resting potential toward that of
K+ and limits CAP amplitudes. Recently we have
developed a method to apply PEG to clinical cases of spinal injury in
dogs (see below). We have noted that during the 2 min application of
PEG to the exposed spinal cord, the minor seepage of blood from
capillaries too small to be cauterized during the laminectomy procedure
was immediately stopped (J. P. Toombs and R. B.
Borgens, unpublished observations). We believe it possible that PEG may
seal or fuse very small breaches in endothelium and that if this occurs
within the contused cord, it may also contribute to a reduction in the
progressive vascular insult accompanying acute severe compression.
The CTM behavioral model and central nervous system (CNS) injury
We do not use walking or other indirect measures of posture and
limb movement in spinal rodents (35
, 36)
as a means to
study the loss and recovery of behavior (37)
. This is due
in part to the prevalence of locally controlled and generated pattered
stepping in small animal models of spinal cord injury, the largely
unknown character of the anatomical basis subserving these complicated
behaviors, and the large numbers of animals required to credibly
demonstrate differences between control and experimental treatments
(37)
. We used a defined sensorimotor behavior, the CTM
reflex, as a behavioral index of spinal cord white matter integrity
(17
, 19
, 21
, 37)
. In transected spinal cords, the CTM
never spontaneously recovers for the life of the animal, whereas in
severe compression injured cords spontaneous recovery is infrequent, as
shown here. Moreover, propagation of nerve impulses through identified
CTM spinal tracts within the ventral funiculus of the cord must be
reestablished to the identified pools of motor neurons in order to
restore CTM functioning. Finally, regions of CTM loss do not recover by
collateral cutaneous sprouting since peripheral cutaneous innervation
is undisturbed by the spinal injury (17
, 19
, 37)
.
This report provides clear evidence of a behavioral recovery
dependent on an identified neural circuit within the damaged mammalian
CNS in response to this experimental treatment. Together with our
previous reports (13
, 14)
, it proposes molecular repair
and fusion of nerve membranes as a novel treatment of severe trauma to
both peripheral nervous system and CNS tissue. Our continuing
experiments will determine the critical window of time for PEG
application, test PEG derivatives [such as PEG/4-aminopyridine
(14)
and PEG-complexed free radical scavengers] designed
to improve PEG-mediated repair, and move clinical tests of PEG
application to naturally produced cases of paraplegia in dogs
(38
, 39)
.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Saadoun, B. A. Bell, A. S. Verkman, and M. C. Papadopoulos Greatly improved neurological outcome after spinal cord compression injury in AQP4-deficient mice Brain, April 1, 2008; 131(4): 1087 - 1098. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Liu-Snyder, M. P. Logan, R. Shi, D. T. Smith, and R. B. Borgens Neuroprotection from secondary injury by polyethylene glycol requires its internalization J. Exp. Biol., April 15, 2007; 210(8): 1455 - 1462. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Wang, Y. Fu, P. Zickmund, R. Shi, and J.-X. Cheng Coherent Anti-Stokes Raman Scattering Imaging of Axonal Myelin in Live Spinal Tissues Biophys. J., July 1, 2005; 89(1): 581 - 591. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. B. Borgens, R. Shi, and D. Bohnert Behavioral recovery from spinal cord injury following delayed application of polyethylene glycol J. Exp. Biol., January 1, 2002; 205(1): 1 - 12. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. S. Duerstock and R. B. Borgens Three-dimensional morphometry of spinal cord injury following polyethylene glycol treatment J. Exp. Biol., January 1, 2002; 205(1): 13 - 24. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
