вход по аккаунту


Distal neuropathy in experimental diabetes mellitus.

код для вставкиСкачать
Distal Neuropathy
in Experimental Diabetes Mehtus
Mark J. Brown, MD, Austin J. Sumner, MD, Douglas A. Greene, MD, Susan M. Diamond, BA,
and Arthur K. Asbury, MD
Although nerve conduction slowing is a well-accepted abnormality in rats with acute experimental diabetes, reports of neuropathological changes in diabetic rat nerves have been inconsistent. To examine this further, we
studied electrophysiological and morphological features of posterior tibial nerves and their distal branches from
four-week streptozotocin-induced diabetic rats and matched controls.
Diabetic rat posterior tibial motor conduction was slowed (mean k 1 SD, 34.8 2 3.1 d s e c ; controls, 41.2 2 2.5
dsec), and evoked muscle response amplitudes were only half of control values. Using quantitative techniques, we
documented a diminution in number of the largest myelinated fibers in otherwise normal mid posterior tibial
nerves, with an increase in the smaller sizes, indicating either a degree of axonal atrophy or impaired fiber growth
during development. The principal pathological finding was active breakdown of myelinated fibers in the most
distal motor twigs of hindfoot muscles supplied by posterior tibial branches, with preservation of fibers in more
proximal segments of these nerves. This anatomical lesion in diabetic nerves could account for both observed
conduction slowing and lowered muscle response amplitudes. A consistent feature in both diabetic and control mid
lateral plantar nerves was a zone of demyelination that apparently occurs at this natural site of nerve entrapment in
rats. Taken together, the pathological abnormalities of peripheral nerve in acute experimental diabetes are best
explained as resulting from a distal axonopathy.
Brown MJ, Sumner AJ, Greene DA, et al. Distal neuropathy in experimental
diabetes mellitus. Ann Neurol 8: 168-178, 1980
The pathological alterations of peripheral nerves in
chronic human diabetic polyneuropathy have been
established by biopsy and autopsy studies (reviewed
in [l, 5,481). The predominant feature is widespread,
seemingly nonspecific distal axonal degeneration,
although segmental demyelination may be an important finding in some cases. These advanced morphological changes appear to account in large part for the
slowed nerve conduction and reduced sensory potential amplitudes that are regularly observed in
symptomatic human diabetic neuropathy [ 18, 21, 29,
321. In contrast, the early structural changes in
human diabetic neuropathy have not yet been
identified with certainty, nor is there a satisfactory
explanation for the subtle clinical and electrophysiological abnormalities that can be found in
newly diagnosed diabetic patients [8, 15, 21, 511. An
extensive nerve biopsy study of asymptomatic newly
diabetic individuals may be unfeasible for ethical reasons. Animal models are therefore of great interest in
searching for the earliest morphological lesion in diabetic neuropathy.
Acute experimental diabetes in rats is accompanied
by electrophysiological abnormalities of peripheral
nerve that suggest an analogy with human diabetic
neuropathy [13, 19,20,41]. Motor nerve conduction
is slowed in rats with diabetes, whether produced by
pancreatectomy o r injection of alloxan or streptozotocin. Despite the unequivocal slowing, down to
70 to 75% of normal in these reports, pathological
findings in nerves of rats with acute experimental
diabetes have been disappointingly few and inconsistent [46].
The investigations described here were designed
to examine the appearance of the posterior tibial
nerve and its distalmost branches from well-characterited acutely diabetic rats with slowed motor nerve
conduction. We report the occurrence of excessive
myelin breakdown in distal intramuscular nerve twigs
of four-week diabetic rats and describe quantitative
alterations of more proximal myelinated fibers. The
possible relationship between these morphological
changes and the observed electrophysiological abnormalities in experimental diabetes is discussed.
From the Department of Neurology and the George S. Cox Research Institute, University of Pennsylvania School of Medicine,
Philadelphia, PA.
Address reprint requests to Mark J. Brown, MD, Department of
Neurology, Hospital of the University of Pennsylvania, 3400
Spruce St, Philadelphia, PA 19104.
Received Oct 11, 1979, and in revised form Nov 27. Accepted for
publication Nov 28, 1979.
168 0364-5134/80/080168-11$01.25 @ 1979 by Mark J. Brown
Table 1 . Plasma Glucose, Weight, and Site of N e r v e s Examined Morphologically for Three Groups of Diabetic and Control Ratsa
No. of Rats with the Following Nerves Examined:
Lateral Plantar
Group 1
Group 2
Group 3
“Values are mean
45 (7) . . .
(7) . . .
646 2 109 (7) 177 f. 39 (7)
174 c 33 (7) 261 2 33 (7)
815 c 166 (6) 190 & 30 (6) 6
318 f. 16(5) 5
728 f. 147 (7) 200
203 I 32 (7) 305
Twigs To
f. 22
SD. Number in parentheses is initial size of each subgroup.
... = not examined.
Groups of outbred male Wistar rats weighing 145 to 1 5 5
gm were rendered diabetic by intravenous injection of
streptozotocin (Upjohn Company, Kalamazoo, MI), 70 mg
per kilogram of body weight, in 0.01 M citrate buffer, p H
5.5, after an overnight fast. Their age-matched cage mates
served as controls. The diagnosis of diabetes was established in surviving rats if nonfasting plasma glucose concentration exceeded 300 mg/dl both 48 hours and four
weeks after injection. Diabetic rats and controls were
housed in metabolic cages and fed a standard diet including
all known nutritional requirements as well as a free myoinositol level of 0.011% (Nutritional Biochemicals Corporation, Cleveland, OH). This highly reproducible model of
experimental diabetes is described in detail elsewhere [20].
Diabetic and control animals were processed alternately to
avoid systematic variations in handling. At the time of
sacrifice each animal was weighed, and plasma was obtained
for glucose determination.
To measure motor conduction velocity of the posterior
tibial nerve, we anesthetized diabetic and control rats with
sodium pentobarbital, 30 to 40 m d k g intraperitoneally.
Core temperature was maintained at 37°C with an infrared
heating lamp. The left leg was extended and secured to
facilitate distance measurements. Stimuli of 0.1 msec duration were applied to the left posterior tibial nerve at the
popliteal fossa and at the ankle. To reduce local volume
conduction, stimulus intensity was limited to the smallest
voltage that consistently produced the maximum muscle
response. Evoked muscle action potential amplitudes were
recorded with unipolar subcutaneous pin electrodes, inserted with standardized placement in the plantar surface of
the left foot. Distal and proximal latencies and potential
amplitudes were measured from photographs of oscilloscope recordings, and velocity was calculared. These and all
other measurements were expressed as mean SD. Statis-
tical significance between group means was examined using
the two-tailed t test.
For morphological studies, diabetic and control rats were
given sufficient sodium pentobarbital to ensure deep
anesthesia. We cannulated the abdominal aorta with a fine
plastic catheter, incised the inferior vena cava, and perfused
the aorta and lower limbs with 50 ml of 0.9% saline solution, followed by 200 ml of 3.6% glutaraldehyde in 0.1 M
phosphate buffer. Portions of the posterior tibial nerve and
its branches were removed on the right side using the dissecting microscope (Table 1). Specimens were fixed for an
additional 24 hours in 3.6% glutaraldehyde in 0.1 M phosphate buffer. Adjacent cross and longitudinal segments of
each nerve were postfixed in 2% osmium tetroxide in 0.1
M phosphate buffer, dehydrated in graded alcohol and
propylene oxide, and embedded in epoxy. Cross and longitudinal sections 2 fi thick were cut and mounted on glass
slides and examined with the light microscope.
During the course of this study we prepared, using identical methods, three groups of diabetic and control rats (see
Table 1).We removed 1 cm segments of the mid posterior
tibial nerves from 6 diabetic and 5 control cage mates in
Group 1 for qualitative and quantitative study. Posterior
cibial nerve conduction in these rats had been measured on
the day of sacrifice (Table 2). Cross and longtudinal sections were prepared. To look for early nodal and paranodal
myelin changes that could account for conduction slowing,
we teased 50 osmicated single fibers in 100% glycerin from
posterior tibial nerves of 3 diabetic rats in Group 1. These
were evaluated with differential interference contrast optics. Segments of posterior tibial nerves from 3 diabetic and
3 control rats in Group 1 were stained with copper ferrocyanide [lo] and with ferric ion in phosphate buffer [36]
to examine cation binding activity of the nodal axolemma
and gap substance. Single fibers from these stained
specimens were teased in liquid epoxy resin, and 2 p
Brown et al: Distal Neuropathy in Diabetes
Table 2. Posterior Tibial Nerve ElectrophysaologicalStudies of Four-Week Diabetic and Control Rats (Group 1 ) a
Rats (N = 5 )
Motor conduction velocity (m/sec)
Muscle potential amplitude (mv)
Distal latencv (msec)
Rats (N = 5)
41.2 5 2.5
2.8 & 0.9
1.0 0.1
p < 0.01
= 0.01
"All values are mean +- 1 SD
NS = not significant.
epoxy-embedded, hardened longitudinal sections of the
same material were mounted on slides. Nodes of Ranvier
were examined with the light microscope using phase optics.
To quantitate the myelinated fiber composition of the
posterior tibial nerve in the Group 1 rats, we prepared
photomicrographs of whole cross sections enlarged and
printed to known final magnification. Total endoneurial
area was measured directly with the aid of a calibrated
Elographics digitizer-Hewlett-Packard 9815A calculator
system programmed to compute area using Simpson's rule.
High-magnification phase photomicrographs of each fascicle were taken using an alternating-field pattern to avoid
selection bias. At the time of data analysis these fields were
found to cover a mean of 37% of diabetic nerve endoneurial areas and 33% of control endoneurial areas. Myelinated fiber area was obtained by digitizing the longer or
major (d,) external axis and shorter o r minor (d,) external
axis of each myelinated fiber and by computing area as if
the fibers were ovals:
Table 3. Semiquantitative Scales for Evaluating Myelhated
Fibers in Peripheral Nerve Cross Sections'
This method is considerably faster than and almost as reliable as measuring the true area from multiple perimeter
points, and is preferable to estimating area by measuring
major or minor axes alone [27].Size-frequency distribution
and number of myelinated fibers per square millimeter of
endoneurial area as well as fibers per nerve were calculated
for each posterior tibial nerve [4]. Mean frequency of each
fiber size category and mean number of fibers per nerve
were then determined for diabetic and control nerves and
were compared using the two-tailed t test.
Lateral plantar nerve trunks were examined at three sites
(see Table 1). The proximal segment was obtained 2 to 3
mm distal to the calcuneus, the middle segment was the
next 1 cm portion, and the distal segment was a 6 mm
length ending just proximal to the origin of the branch to
small plantar foot muscles. Three diabetic and 3 control
middle lateral plantar segments were prepared for single
fiber teasing. Distal sensory branches of medial plantar
nerves in the foot were obtained from 6 diabetic and 5
control rats (Group 1).
Initially, distal lateral plantar nerve branches and superficial and intramuscular twigs to plantar interosseus muscles
were obtained from a small number of diabetic and control
rats in which the posterior tibial motor conduction velocities had been measured (Group 2). Because of the remote
possibility of traumatic damage to these nerves during
electrophysiological testing, we prepared an additional 7
diabetic and 7 control rats not subjected to electrophysiological studies (Group 3 ) . To evaluate the myelinated fiber density and extent of fiber breakdown in these
nerves, we coded the slides of cross sections to obscure
their origin and rated them according to a semiquantitative
scale (Table 3 ) . Differences in group "neuropathy scores"
were evaluated for statistical significance using the MannWhitney test.
170 Annals of Neurology Vol 8 N o 2
August 1980
Maximum complement found in controls
Mild reduction in MF
Moderate reduction in MF
<5% of maximum or no MF
None present, MF are present
<1 myelin ovoid/25 fibers
1-3 myelin ovoidd25 fibers
> 3 myelin ovoids/25 fibers
None present, but no MF are seen in nerve
a"Neuropathy score" = A + B, an estimate of acute and chronic
MF changes, and can range from 1 (most abnormal) to 8.
MF = myelinated fibers.
Sixteen diabetic rats survived four weeks after injection of streptozotocin, the time when all studies were
carried out; there were 16 morphological controls.
All streptozotocin-injected rats were diabetic, with
plasma glucose levels well above 300 mg/dl (see
Table 1). Diabetic rats appeared chronically ill and
weighed less than age-matched controls. There were
no clinical signs of sensory or motor impairment.
Mean posterior tibial nerve conduction was 34.8 +- 3.1
m/sec in the diabetic rats, significantly less ( p < 0.01)
than that in age-matched controls (see Table 2).
Plantar muscle evoked response amplitudes from
900 *
< .05
P < .05
P =.01
Control l n = 5 )
2820 * 160 MF/Ncrvc
= Diobelic ( n = 6I
2850t190 MFINerve
Myelinated Fiber Size
Fig 1 . Myelinatedfiber (MF) size-frequency distribution for
control and diabetic rat posterior tibial nerves. The overall
number offibers per diabetic nerve is normal, but there is a
35 % decrease in the population of large fibers (area >40 p2,
diameter >7.2 G)with a similar increase in the smallfiber
population. Statistically significant p values are indicated.
Bars represent I SD.
diabetic animals were reduced to less than half of
control values (Table 2). These abnormalities served
to confirm the expected presence of neuropathy
using electrophysiological criteria. Distal motor
latencies did not differ in diabetics and controls.
By light microscopy we found no abnormalities in
cross or longitudinal sections of posterior tibial
nerves. Nodal and paranodal myelin of teased single
fibers from 3 diabetic posterior tibial nerves was
normal by light microscopy. There was no evidence
of active or chronic demyelination. Copper ferrocyanide and ferric ion stained the nodal gap substance and adjacent axolemma of diabetic nerves to
the same extent as controls. Myelinated fiber density
of posterior tibial nerves was normal in diabetic rats
(2,850 f 190 M F/nerve, N = 6; control, 2,820 f
160 M F/nerve, N = 5 ) . Myelinated fiber sizefrequency distribution was altered, however: there
was a mild reduction in the number of largestdiameter axons in diabetic nerves with an apparent
compensatory increase in smaller-diameter axons
(Fig 1). This reduction was largely of myelinated
fibers with an external area of 40 p2 or more.
Whereas control nerves contained 780 If; 240 fibers
of this size, diabetic nerves had only 510 f 140 ( p <
0.05). The endoneurial area of 6 diabetic posterior
tibial nerves measured from the innermost lamella of
the perineurium was 0.15 -+ 0.02 p2 and was similar
to that of 5 controls (0.16 ? 0.01 p2).
Diabetic nerves to the interosseus muscles were
morphologically abnormal because of reduced numbers of myelinated fibers and abundance of myelin
debris (Figs 2-4). Numerous myelin ovoids appeared
in cross and longitudinal sections of most fascicles in
all of 10 diabetic plantar interosseus nerves (Figs 3,
4). Occasional fibers undergoing degeneration also
were found in 5 of 10 control nerves. Neuropathy
scores for coded sections of diabetic interosseus
nerves were significantly lower than for control
nerves taken at the same sites (Fig 5 ) . The difference
from controls was greater in the distal intramuscular
branches (p < 0.001) than in the more proximal
superficial branches ( p < 0.01). Among diabetics, the
distal nerves were more severely affected than the
proximal nerves (p < 0.05); there was no difference
between control nerves taken at the proximal or the
more distal sites. We did not observe large-diameter
demyelinated fibers, neurofilamentous accumulations, or focal enlargements of diabetic axons.
At the most proximal portion of the lateral plantar
nerves, myelinated fibers were normal although rare
Brown et al: Distal Neuropathy in Diabetes
myelin ovoids were observed in diabetic but not
control specimens. Moving distally, the mid lateral
plantar nerves of both diabetic and control animals
had a myxomatous, loosely packed endoneurium
with abundant small Renaut bodies (Fig 6). Myelin
density appeared less than that usually found in
distal nerve trunks. This was largely the result of
thinned or absent myelin sheaths around single larger-diameter axons. Teased single fibers through this
region confirmed the presence of segmental demyelination and remyelination. Schwann cell numbers
appeared increased, and occasionally there were thin
onion bulb formations. These changes of demyelination and remyelination were present in all rat
nerves examined but were more prominent in the
172 Annals of Neurology Vol 8 No 2
August 1980
Fig 2. Control rat. (A)Longitudinal section of superficial
portion of nerve t o plantar interosseus muscles. Myelinatedfiber
structures appear normal. (9)Cross section from the same site
demonstrates normal appearance offibers. A Renaut body is
present (R). (2 p epoxy sections; phase optics. Bar = 10 p.)
diabetic nerves (see Fig 6). Myelinated fibers in
the lateral plantar segment distal to this, just before
the site of branching of the nerve to the interosseus
muscles, were more densely concentrated within the
endoneurial space than in the adjacent demyelinated
region, but fiber density did not return to that found
proximal to the focal area of demyelination. However, individual fibers were normal in appearance and
Fig 3 . Diabetic rat. (A)Longitudinal section of superjicial
portion of nerve i o plantar interosseus muscles. Myelin ovoids
have replaced the myelinated fibers. ( B ) Cross section from the
same site shows abundant myelin debris and few remaining
normalfbers. A Renaut body is present (R).(2 p epoxy section; phase optics. Bar = 10 p.)
did not differ between diabetics and controls. Sensory branches of the medial plantar nerves also were
normal aside from the rare occurrence of myelin
ovoids in both diabetic and control nerves.
Several groups of investigators [ 3 4 , 35, 38, 521 have
described segmental demyelination and axonal de-
generation in nerves of rats with chronic (four
months to two years) experimental diabetes. The
changes are not unique to diabetes and can be features of normal aging in rats [22, 341 as well as humans [30, 331. Even so, they provide morphological
support for the position that peripheral neuropathy
occurs in alloxan- and streptozotocin-induced diabetes in rats and simulates features of human diabetic
neuropathy. These myelinated fiber abnormalities
represent a late stage of neuropathy and do not indicate the nature of the early, potentially reversible
morphological changes in experimental diabetes.
The principal evidence for the occurrence of
neuropathy in acute (less than three months) experimental diabetes in rats has been electrophysiological.
Brown et al: Distal Neuropathy in Diabetes
Slowed motor conduction has been measured at a
variety of sites, including sciatic [20], posterior tibia1
[41], and caudal nemeS [ 191. Since conduction slowing in streptozotocin diabetes can be
rigorous insulin therapy [20], it is unlikely that the
electrophysiological abnormalities result from a direct toxic effect of streptozotocin o n nerves. Structural confirmation of myelinated fiber abnormalities
174 Annals of Neurology
Vol 8 No 2
August 1980
Fig 4. (A) Cross section of intramuscularportion of nerve t o
plantar interosseas muscles i n control animal. Myelinated
fibers appear normal. ( B ) Diabetic intramuscular nerve obtained from the same site. There is severe loss of myelinated
fibers and widespread myelin debris. (2 p epoxy sections;phase
optics. Bar = 10 p.)
7 0
zz 4o- o
Nerves l o Plantar lnlerosseous Muscles
Fig 5 . Comparison of control and diabetic rat nerves t o plantar
interosseus muscles at two sites, using a semiquantitative morphological “neuropathyscore” (see Table 3). Diabetic nerves
scored lower than controls at both sites, with the distal intramuscular portion significantly more affected than the more
proximal superficial portion.
in acute experimental diabetes has been found only
recently [6, 231.
In the present study we did not detect clinical signs
of motor or sensory nerve impairment in any of our
four-week diabetic rats; to our knowledge none have
been reported by others. Whenever possible our
electrophysiological measurements were undertaken
in the same animals in which the nerves subsequently
were studied morphologically (see Table 1). This was
not always technically feasible. However, the diabetic
response to streptozotocin injection among 150 gm
rats was highly consistent, as determined by the observed degree of hyperglycemia and reduction of
body weight during and at the termination of an experiment (see Table 1; [201). We therefore assumed
that a n y peripheral nerve changes occurring in one
group of rats would be present to a similar degree in
other groups. We observed that mean posterior tibial
motor nerve conduction in diabetic rats slows, to
about 16% less than in age-matched controls, after
four weeks of streptozotocin-induced diabetes. In
addition to reduced motor conduction, we found that
plantar interosseus muscle response amplitudes obtained after stimulating the posterior tibial nerve
Fig 6. Whole cross section of control (A)and diabetic ( B ) rat
mid lateral plantar nerves at same magnification. Compared
with other sites along this nerve, the endoneurial matrix i s
increased and has a myxomatous appearance. Myelin sheaths
in both nerves are ohen thinner than would be expectedfor
axon diameter, suggesting demyelination and remyelination.
By inspection, myelinatedfibers are less abundant in the diabetic nerve. (2 p epon sections; phase optics. Bar = SO p.)
Brown et al: Distal Neuropathy in Diabetes
were about half those of controls. Terminal motor
latencies were not prolonged.
A unique finding in this study was the presence of
myelinated fiber debris in distal twigs of diabetic
posterior tibial nerve branches when more proximal
portions did not show these changes. Distal fiber degeneration could be the consequence of a proximal
focal compression neuropathy present in the feet of
these cachectic young diabetic rats, and to a lesser
degree in controls. Compression neuropathies occur
in forepaws of normal guinea pigs, but generally after
18 months of age or more [17, 311. Human diabetics
with neuropathy appear to have a predisposition to
pressure palsies [321, as do patients with other
neuropathies [47].Diabetic rat nerves could have the
same susceptibility. However, distal fiber breakdown
in lateral plantar twigs of our diabetic rats became
more extensive as the distance from the presumed
compression site increased, and occasional myelin
ovoids were found in the proximal diabetic lateral
plantar nerves, above the area of apparent compression neuropathy. These observations are more compatible with a graded distal dying-back neuropathy
171 or a distal axonopathy [45].However, we cannot
yet eliminate compression as a pathogenic factor affecting diabetic rat nerves.
We and others [41]could not distinguish qualitatively the myelinated fibers of four-week diabetic
mid posterior tibial nerves from controls. In an
abstract, Preston [35] reported an increased incidence of segmental demyelination in teased nerve
fibers from chronically diabetic rats. Since demyelination is a feature in some human diabetic nerves [so]
and would account for conduction slowing, this abnormality in nerves of acutely diabetic rats has been
diligently sought by others [24, 411. None has been
successful. Like them, we could not find evidence of
meaningful nodal or internodal myelin changes in
either longitudinal sections or teased single fibers
from diabetic rats. We were unable to confirm the
observation that cation binding at nodes is decreased
in diabetic rat nerves [ d o ] ;Sharma and Thomas [41]
also found normal staining in this region. Despite
these discouraging results, recent support for an early
morphological lesion of myelin has come from
Fukuma and co-workers [ 161, whose freeze-fracture
preparations of two-week diabetic rat sciatic nerve
myelin contained fewer intramembranous particles
than controls. While the importance of these particle
alterations is far from clear, even minor changes in
diabetic myelin could alter its resistivity and slow
saltatory conduction [ 141.
The myelinated fiber spectrum of well-fixed posterior tibial nerves from our four-week diabetic rats
with slowed motor conduction shifted away from the
larger diameter fibers toward the smaller ones, with
176 Annals of Neurology
Vol 8 No 2
August 1980
no change in total fiber numbers. This concurs with
the observations of Jakobsen [231, who reported
“axonal dwindling” or axonal atrophy in the peroneal
nerves of perfusion-fixed diabetic rats when compared with similarly handled controls. Malnutrition
in diabetes may explain this apparent axonal atrophy.
While food restriction for four weeks sufficient to
duplicate the weight loss of diabetes did not produce
axonal atrophy in Jakobsen’s adult rats [23],the normal enlargement of largest diameter myelinated
fibers was impaired in nerves of chronically starved
young rats [42].Axonal atrophy has been described in
experimental neuropathies [3, 91 and in genetic and
acquired human neuropathies [ l l , 121. The basis of
axonal atrophy or failure of maturation in experimental diabetic neuropathy is unknown. Diminished
axoplasmic transport in diabetic rats [39]could be responsible for reduction in axoplasm and distal axonal
caliber, although this explanation is speculative.
In the present study, the endoneurial area of perfused diabetic posterior tibial nerves was normal at a
time when motor conduction in the same nerves was
slowed. In contrast, Jakobsen [ 2 5 ] found that the
volume of endoneurial structures was decreased and
“structureless” area increased in osmium-fixed, frozen cross sections of sciatic nerves from four-week
diabetic rats. He thought that artifacts of fixation did
not play a role in his results and speculated that an
abnormality of the blood-nerve barrier could result
in endoneurial swelling and an alteration of the endoneurial milieu, thus affecting nerve conduction.
Physiological studies using tracers with molecular
weight as small as 12,000 [26, 441 have failed to
demonstrate the postulated blood-nerve barrier
permeability defect. Since abnormal diffusion of
molecules smaller than those available for tracer
studies could affect the function of diabetic nerves,
physiologically important blood-nerve or bloodperineurial barrier abnormalities cannot yet be
excluded despite the lack of convincing morphological or physiological evidence.
Could either axonal atrophy or distal fiber degeneration account for the electrophysiological abnormalities that we and others have found in acute experimental diabetes? Considering the first possibility,
we think it unlikely that, by itself, a uniform 35%
reduction in the number of motor and sensory fibers
with an external area greater than 40 p2 would result
in decreased motor conduction. Although evoked response amplitudes would be lowered in this situation, 65% of the largest fibers would remain unaffected, and maximum conduction velocity should be
normal [2]. However, selective motor fiber atrophy
(or failure of maturation) could slow motor conduction and yet be obscured morphologically by the
large sensory fiber complement remaining in the
mixed myelinated fiber histograms. The effect of abnormally small “largest” fibers on conduction has
been demonstrated experimentally in malnourished
rats; fastest sensory conduction is reduced in these
animals, in which myelinated axons have not reached
a normal diameter for age [43].
As for the second possibility, a distal axonopathy
with breakdown of the plantar interosseus nerves
could account for both the reduced action potential
amplitudes and slowed nerve conduction velocities
that we observed in diabetic rats. The plantar interosseus nerves supply a major portion of those small
foot muscles for which the action potentials form the
bulk of our measured evoked responses. Thus, loss
of terminal motor axons and subsequent failure of
their motor units to fire explain the reduced muscle
potential amplitudes that we recorded. But preservation of even a few of the fast-conducting motor axons
would result in normal maximum velocity. However,
if the largest-diameter and most distal fibers were
selectively damaged first, with only the smallerdiameter motor fibers remaining in functional contact
with motor end-plates, then maximum conduction
would be measured only through the slowerconducting fiber population. This mechanism has
been proposed to explain median nerve forearm
slowing in patients with the carpal tunnel syndrome
[49]. Selective early involvement of the largest as
well as longest fibers has been recognized in other
experimental peripheral neuropathies [45]but not in
experimental or human diabetic neuropathy.
Another consideration is that the electrophysiological and coexisting morphological abnormalities in acute experimental rat diabetes could be
unrelated phenomena occurring independently as a
result of the pervasive metabolic disorder in diabetes.
Eliasson 1131 observed that in vitro rat mixed sciatic
nerve conduction is slowed in diabetes. The sciatic
nerve is thought to be morphologically normal in
early diabetes [41], and Eliasson’s direct nerve recording system did not include distal motor axons. If
his findings are substantiated by others using current
methods and the isolated nerve segments are found
to be morphologically normal by present-day criteria,
then the independence of the two processes will be
established. In human diabetic neuropathy the magnitude of nerve conduction slowing cannot be accounted for completely by the observed morphological changes, indicating at least a degree of
independence in the two processes [ 2 ] .
Our current hypothesis is that experimental diabetic neuropathy in the rat is a distal axonopathy and
that the earliest fiber breakdown appears in terminal
portions of susceptible nerves. This position is supported by our finding that active axonal degeneration
is present in plantar interosseus nerves when more
proximal portions of the same nerve are normal, and
could explain the accepted electrophysiological
findings in experimental diabetes. We speculate that
a similar mechanism operates in human diabetic
polyneuropathy, in which clinical [l,5, 481 and electrophysiological [28] evidence suggests that the
longest nerve fibers are affected first. Morphological
support for a distal axonopathy in human diabetes
comes from the work of Reske-Nielsen et a1 [37],
who found changes of distal axonal degeneration in
muscle biopsies from all of 7 juvenile diabetics only
four weeks or less after the clinical onset of their
symptoms. However, further studies will be required
to determine conclusively whether the acute changes
in distal nerve fibers of diabetic rats are a reflection of
distal axonopathy or are a unique experimental
analog of the clinical liability of neuropathic patients
to pressure palsies.
Supported by US Public Health Service Grants NS08075,
IP3OAM-19525, and R01 AM16186, and by the Muscular Dystrophy Association, Inc.
We thank Sandra Jo Radich and Sally Lattimer for their technical
assistance and Barbara J. Whiting for preparing the manuscript.
1. Asbury AK, Johnson PC: Pathology of Peripheral Nerve.
Philadelphia, Saunders, 1978
2. Behse F, Buchthal F, Carlsen F: Nerve biopsy and conduction
studies in diabetic neuropathy. J Neurol Neurosurg Psychiatry 40:1072-1082, 1977
3. Brown MJ, DeJesus CPV, Pleasure DE, Asbury AK: Nerve
conduction slowing precedes demyelination in experimental
methyl n-butyl ketone (MBK) neuropathy. Presented at the
IVth International Congress on Neuromuscular Diseases,
September 1978
4. Brown MJ, Iwamori M, Kishimoto Y , et al: Endoneurial lipid
composition of normal human sural nerve. Ann Neurol
5:239-244, 1979
5. Brown MJ, Martin JR, Asbury A K Painful diabetic
neuropathy. A morphometric study. Arch Neurol 33:164171, 1976
6. Brown MJ, Sumner AJ, Greene DA, et al: Distal neuropathy
in acute experimental diabetes. Neurology 29590, 1979
7. Cavanagh JB: The significance of the “dying back’ process in
experimental and human neurological disease. Int Rev Exp
Pathol 3:219-267, 1964
8. Chochinov RH, Ullyot LE,Moorhouse JA: Sensory perception thresholds in patients with juvenile diabetes and their
close relatives. N Engl J Med 286:1233-1237, 1972
9. Clark AW, Griffin JW, Price DL: The axonal pathology in
chronic IDPN intoxication, J Neuropathol Exp Neurol
39~42-55, 1980
10. DeBaecque C, Raine CS, Spencer PS: Copper binding at PNS
nodes of Ranvier during demyelination and remyelination in
the perineurial window. Neuropathol App Neurobiol
2~459-470, 1976
11. Dyck PJ, Johnson WJ, Lambert EH, O’Brien PC: Segmental
demyelination secondary to axonal degeneration in uremic
neuropathy. Mayo Clin Proc 46:400-431, 1971
12. Dyck PJ, Lais AC: Evidence for segmental demyelination secondary to axonal degeneration in Friedreichs ataxia. In
Brown et al: Distal Neuropathy in Diabetes
Kakulas BK (ed): Clinical Studies in Myology. Amsterdam,
Excerpta Medica, 1973
13. Eliasson SG: Nerve conduction changes in experimental diabetes. J Clin Invest 43:2353-2358, 1964
14. Eliasson SG: Properties of isolated nerve fibers from alloxanized rats. J Neurol Neurosurg Psychiatry 32:525-529,
15. Fraser DM, Campbell IW, Ewing DJ, et al: Peripheral and
autonomic nerve function in newly diagnosed diabetes mellitus. Diabetes 26:546-550, 1977
16. Fukuma M, Carpentier J-L, Orchi L, et al: An alteration in
internodal myelin membrane structure in large sciatic nerve
fibers in rats with acute strepcozotocin diabetes and impaired
nerve conduction velocity. Diabetologia 15:65-72, 1978
17. Fullerton PM, Gilliatt R W Median and ulnar neuropathy in
the guinea-pig. J Neurol Neurosurg Psychiatry 30:393-402,
18. Gilliatt RW, Willison RG: Peripheral nerve conduction in
diabetic neuropathy. J Neurol Neurosurg Psychiatry
25~11-18, 1962
19. Goto I, Peters HA: Serial in vivo determination of motor
conduction velocity in tails of alloxanized non-diabetic and
diabetic rats. J Neurol Sci 22:177-182, 1974
20. Greene DA, DeJesus PV, Winegrad Al: Effects of insulin and
dietary myoinsotol on impaired peripheral motor nerve conduction velocity in acute streptozotocin diabetes. J Clin Invest
21. Gregersen G: Diabetic neuropathy: Influence of age, sex,
metabolic control, and duration of diabetes on motor coFduction velocity. Neurology (Minneap) 17:972-980, 1967
22. Grover-Johnson N, Spencer PS: Peripheral nerve abnormalities in aging rats. J Neuropathol 38:316, 1979
23. Jakobsen J: Axonal dwindling in early experimental diabetes.
I. A study of cross sectioned nerves. Diabetologia 12:539546, 1976
24. Jakobsen J: Axonal dwindling in early experimental diabetes.
11. A study of isolated nerve fibers. Diabetologia 12:547-553,
25. Jakobsen J: Peripheral nerves in early experimental diabetes.
Expansion of the endoneurial space as a cause of increased
water content. Diabetologia 14:113-119, 1978
26. Jakobsen J, Malmgren L, Olsson Y: Permeability of the
blood-nerve barrier in the streptozotocin-diabetic rat. Exp
Neurol60:277-285, 1978
27. Karnes J, Robb R, O'Brien PC, et al: Computerized image
recognition for morphornetry of nerve attribute of shape of
sampled transverse sections of myelinated fibers which best
estimates their average diameter. J Neurol Sci 34:43-5 1,
28. Kimura J, Yamada T, Stevland NP: Distal slowing of motor
nerve conduction velocity in diabetic polyneuropathy. J
Neurol Sci 42:291-302, 1979
29. Lamontagne A, Buchthal F: Electrophysiological studies in
diabetic neuropathy. J Neurol Neurosurg Psychiatry
331442-452, 1970
30. Lascelles RG, Thomas PK: Changes due to age in internodal
length in the sural nerve in man. J Neurol Neurosurg Psychiatry 29:40-44, 1966
31. Marotte LR: An electron microscope study of chronic median
nerve compression in the guinea-pig. Acta Neuropathol
(Berl) 27:69-82, 1974
32. Mulder DW, Lambert EH, Bastron JA, Sprague RG: The
178 Annals of Neurology Vol 8 No 2 August 1980
neuropathies associated with diabetes. A clinical and electromyographic study of 103 unselected diabetic patients.
Neurology (Minneap) 11:275-284, 1961
33. Ochoa J, Mair WGP: The normal sural nerve in man. 11.
Changes in the axons and Schwann cells due to aging. Acta
Neuropathol (Berl) 13:217-239, 1969
34. Powell H, Knox D, Lee S , et al: Alloxan diabetic neuropathy:
electron microscopic studies. Neurology (Minneap)
27:60-66, 1977
35. Preston GM: Peripheral neuropathy in the alloxan-diabetic rat.
J Physiol (Lond) 189:49P-50P, 1967
36. Quick DC, Waxman SD: Ferric ion, ferrocyanide, and inorganic phosphate as cytochemical reactants at peripheral nodes
of Ranvier. J Neurocytol 6:555-570, 1977
37. Reske-Nielsen E. Harmsen A, Vorre P: Ultrastructure of
muscle biopsies in recent, short-term and long-term juvenile
diabetes. Acta Neurol Scand 55:345-362, 1977
38. Sahgal VK, Ahuja MMS, Roy S, Singh N : Diabetic
neuropathy. An experimental study in alloxanized rats, with
special reference to insulin therapy. Acta Diabetol Lat
9:983-1006, 1972
39. Schmidt RE, Matschinsky FM, Godfrey DA, et al: Fast and
slow axoplasmic flow in sciatic nerve of diabetic rats. Diabetes
24: 1081-1085, 1975
40. Seneviratne KN, Weerasuriya A: Nodal gap substance in diabetic nerve. J Neurol Neurosurg Psychiatry 37:502-5 13,
41. Sharma AK, Thomas PK: Peripheral nerve structure and
function in experimental diabetes. J Neurol Sci 23:l-15,
42. Sima A: Studies of fiber size in developing sciatic nerve and
spinal roots in normal, undernour+hed, and rehabilitated rats.
Acta Physiol Scand 91:suppl 406, 1974
43. Sima A, Jankowska E: Sensory nerve conduction velocity as
correlated to fiber size in experimental undernutrition in the
rat. Neuropathol Appl Neurobiol 1:31-37, 1975
44. Sima AAF, Robertson DM: The perineurial and blood-nerve
barriers in experimental diabetes. Acta Neuropathol (Berl)
441189-195, 1978
45. Spencer PS, Schaumburg HH: Pathobiology of neurotoxin
axonal degenerations. In Waxman SG (ed): Physiology and
Pathobiology of Axons. New York, Raven, 1978, pp 265-282
46. Thomas PK: Human and experimental diabetic neuropathy.
In Canal N , Pozza G (eds): Peripheral Neuropathies. New
Yoik, Elsevier North Holland, 1978
47. Thomas P K Screening for peripheral neuropathy in patients
treated by chronic hemodialysis. Muscle Nerve 1:396-399,
48. Thomas PK, Eliasson SG: Diabetic neuropathy. In Dyck PJ,
Thomas PK, Lambert EL (eds): Peripheral Neuropathy.
Philadelphia, Saunders, 1975
49. Thorns PK, Fullerton PM: Nerve fiber size in the carpal tunnel syndrome. J Neurol Neurosurg Psychiatry 26:520-527,
50. Thomas PK, Lascelles RG: The pathology of diabetic
neuropathy. Q J Med 35:489-509, 1966
5 1. Ward JD, Barnes CG, Fisher DJ: Improvement in nerve conduction following treatment in newly diagnosed diabetics.
Lancet 1:428-430, 1971
52. Yagihashi S, Kudo K, Nishihira M: Peripheral nerve structures of experimental diabetes in rats and the effect of insulin
treatment. Tohoku J Exp Med 127:35-44, 1979
Без категории
Размер файла
1 248 Кб
experimentov, diabetes, neuropathy, mellitus, distal
Пожаловаться на содержимое документа