close

Вход

Забыли?

вход по аккаунту

?

Morphological changes at paranodes in IgM paraproteinaemic neuropathy

код для вставкиСкачать
MICROSCOPY RESEARCH AND TECHNIQUE 34~507-521(1996)
Node Distribution and Packing Density in the Rat CNS-PNS
Transitional Zone
JOHN FRAHER
Department of Anatomy, University College, Cork, Irelund
KEY WORDS
CNS-PNS transitional node spacing, Distribution, Packing density,
Morphometry
ABSTRACT
The density of nodes of Ranvier was examined at CNS, PNS, and transitional zone
(TZ) levels of rat lumbar ventral motoneurone fibres. It was found to be significantly greater in the
TZ than at the other levels: The differencewas sevenfold for the ventral root and at least fourfold
for central fibre levels.
Node distribution and spacing was examined within the two main types of TZ found in rat ventral
rootlets: the first, in which the TZ is short and is approximately on a level with the surface of the
cord; and the second, in which it is much longer and extends into the proximal part of the rootlet.
Node spacing was estimated as nearest neighbour distance, the true distance between adjacent
node centres. This is a better estimate of node spacing than simple density since it measures the
actual linear distance between nodes over which any interaction between them would be likely to
take place. Despite marked differences in the dimensions of the two types of TZ, nearest neighbour
distance distribution was very similar in each, suggesting that similar mechanisms may influence
their spacing during development.
The TZ contains especially large amounts of interstitial tissue, mainly composed of astrocyte
processes, separating the fibres traversing it. The proportion of the TZ composed of interstitium was
over three times that in the ventral root and nearly twice that at the CNS level studied. The large
amounts of astrocytic tissue in the TZ may be related to the high packing density of nodes. It may
function to regulate extracellular ionic concentrations in the TZ and to maintain a stable ionic
environment for the transitional nodes. 8 1996 Wiley-Liss, Inc.
INTRODUCTION
The transitional zone (TZ) forms an irregular interface between the CNS and PNS. It extends over a short
length of each fibre bundle passing between them (Fig.
1). It is characterised by a sharp discontinuity in a
variety of tissue types. Myelin sheaths peripheral to it
are formed by Schwann cells. Those central to it are
formed by oligodendrocytes. At the interface each myelinated fibre possesses a transitional node of Ranvier
bounded by one of each of the two myelinating cell
types, termed the transitional Schwann cell and the
transitional oligodendrocyte (Bristol and Fraher, 1987;
Fraher 1978a; Fraher and Kaar 1984,1986). Much of
the remaining tissue of the TZ is composed of astrocytes, which form a specialised glia limitans in it. Myelinated fibres pass through this in tunnels. Because
the TZ is short and each fibre possesses a node as it
passes through it, the density of transitional nodes is
likely to be high relative to other places in the PNS and
CNS, in which nodes are generally not in phase with
one another but are scattered more or less irregularly
along the fibres. The TZ therefore represents an area of
considerable potential interest for studies on the packing, distribution, metabolism, and even interaction of
nodes. The packing density and distribution of transitional nodes are influenced by the amount, arrangement, and organisation of astrocytic tissue in the TZ.
This separates the myelinated fibres as they traverse
8
1996 WILEY-LISS,INC.
the "Z, not only at their nodes but also over the central
and peripheral paranodes bounding these. It helps to
keep them apart and so,arguably, also helps to prevent
interaction between them.
Transitional node density and distribution were investigated using ventral motoneurone fibres of adult
rat cervical and lumbar ventral roots. Having emerged
fkom their cell bodies in the ventral spinal grey matter
(Fig. 1) these fibres are grouped into bundles. As they
cross the spinal cord white matter these bundles are
termed intramedullary rootlets. They emerge into the
PNS as ventral rootlets which eventually unite distally
to form the ventral roots. Close to the spinal cord surface, central and peripheral nervous tissues overlap
within the rootlet at the irregular CNS-PNS interface.
As a result, a segment of the rootlet contains both types
of tissue. This segment is the transitional zone.
Node density was investigated at three distinct levels along rat ventral motoneurone fibre bundles: in the
intramedullary rootlets, where internodal length is
relatively short; in ventral rootlets, where the internodes are relatively long (Fraher 1978b);and in the TZ.
Node spacing, i.e., centre to centre distance between
Received February 13, 1995; accepted in revised form March 20,1995.
Address reprint requests to ProfessorJohn P. Fraher, Departmentof Anatomy,
University College, Cork,Ireland.
J. FRAHER
508
a
b
Fig. 1. a: Diagrammatic transverse section of upper cervical spinal cord giving rise to two ventral rootlets (arrows).The area outlined
is shown enlarged in c. b: Diagrammatic longitudinal section showing
spinal cord at mid-lumbar level giving rise to ventral rootlets (arrows). The area outlined is shown enlarged in d. c , d Enlargements of
the areas outlined in a and b, a t upper cervical and mid-lumbar levels,
respectively. In each, four myelinated fibres (asterisks) are shown
running in an intramedullary rootlet (double arrowheads) and in the
ventral rootlet (arrowheads)continuous with it. A central tissue projection (CTP)extends into the lumbar rootlet. Each myelinated fibre
possesses a transitional node of Ranvier (arrows) as it traverses the
interface between central (dark) and peripheral (pale) nervous tissue.
This is bounded centrally by the transitional oligodendrocyteand peripherally by the transitional Schwann cell. The cervical CNS-PNS
interface is irregular lying at a level a t and slightly deep to the surrounding spinal cord surface. The length of rootlet containing both
central and peripheral nervous tissue is termed the transitional zone
(TZ).(Reproduced from Bristol et al., 1993, with permission of S.
Karger AG.)
adjacent nodes, was also studied in the TZ, where node
density was found to be highest.
may be identified in the rat. Node density and distribution were examined in the two principal contrasting
types of zone found in rat spinal cord. In the first of
these (classed here as type l), the TZ is centred on the
surface of the cord and extends a little above and below
it (Fig. 2a). The glia limitans is thickened at the TZ.
Around the perimeter of the fibre bundle, a t and deep
to the plane of the CNS surface, the TZ is often surrounded by a thick sleeve of astrocytic tissue which
contains no nerve fibres. This may be broader towards
TRANSITIONAL ZONE MORPHOLOGY
The form of the TZ influences the distribution of
transitional nodes within it. Its general appearance,
position, and extent vary markedly between nerves
and even in some instances between the constituent
rootlets of a given nerve (for a review see Fraher,
1992). On this basis, at least eight different types of TZ
TZ NODE DENSITY AND DISTRIBUTION
509
Fig. 2. Diagrammatic longitudinal sections through (a)cervical and (b) lumbar rootlets and at right
angles to the spinal cord surface (arrows), showing the two types of transitional zone in which node
density was studied. Dark: CNS tissue; pale: PNS tissue.
the cord surface than at deeper levels, so that the TZ is
wedge-shaped on longitudinal section. In such examples, the centre of the mass of astrocytic tissue is traversed by the fibre bundle (Fig. 2a). The superficial
part of the glia limitans is invaginated by the central
ends of the transitional Schwann cells and its deep part
contains the distal ends of the transitional oligodendrocytes (Fig. 1).The transitional nodes between the two
lie within the substance of the TZ, mostly at, slightly
superficial, and slightly deep to the plane of the surrounding cord surface (see below). TZs of this kind are
found in rat cervical ventral rootlets. The second type
of TZ studied is found in most rat thoracic and lumbar
ventral rootlets which have been examined. In this
type, classed here as type 2 (Fig. 2b), central nervous
tissue extends distally into the free rootlet as a central
tissue projection. This tapers distally and has an irregular conical or wedge shape. It is composed of astrocytic
tissue traversed by motoneurone fibres. Its surface is
formed of thickened glia limitans and its core resembles white matter. Peripheral nervous tissue surrounds
the central tissue projection and comprises the complementary portion of the rootlet over the extent of the TZ
(Fig. 1).In most cases the central tissue projection lies
asymmetrically in the rootlet, being closer to its deep
than to its superficial surface. This type of zone is generally confined to the rootlet and does not extend to a
significant degree deep to the cord surface. In some
places, especially where a number of adjacent rootlets
emerge from the CNS close to one another, the glia
limitans intervening between them is thickened and
the rootlets emerge through a thick pad of interwoven
astrocyte processes, common to all of them.
Apart from the axons and their myelin sheaths, the
bulk of the central tissue component of the TZ is formed
by a thick weave of astrocyte processes. In larger TZs
astrocyte perikarya are also found within the zone.
However, the rootlets under consideration were small
and most astrocyte perikarya lay around the periphery
of the zone, most often at the surface of the CNS surrounding the attachment of the rootlet to the cord. It is
noteworthy that over its Schwann cell paranode, its
transitional node, and its course over a short distance
central to that, each myelinated fibre traverses the TZ
in an individual tunnel walled by astrocyte processes.
These processes separate it from the neighbouring myelinated fibres, and also from unmyelinated axon bundles in those TZs which contain these. The central end
of the transitional Schwann cell typically lies in an
invagination in type 1zones and in a groove in type 2
zones (Fig. 3a,b). In type 1 zones the depth of the invaginations averages around 30 pm but may be up to
50 pm or even more (Fraher, 1978a; Fraher and Kaar,
1982). The wall of the invagination is formed by astrocyte processes (Fig. 3a). In many instances these show
a tendency to be circumferentially arranged. Their inner aspect is covered by a layer of basal lamina. Within
the invagination is the central end of the transitional
Schwann cell, comprising its paranode and a variable
length of the adjacent part of the internode, depending
on its depth. The transitional node lies at the deep end
of the invagination (Fraher and Kaar, 1984; Fraher et
al., 1988). Transitional node structure and development have been described in detail (Fraher and Kaar,
1984; Rossiter and Fraher, 1990). Some of its essential
features are shown in Figure 4. At this level, astrocyte
processes project into the node gap and tend to form a
partition separating the paranode of the transitional
Schwann cell from that of the oligodendrocyte. The
CNS and PNS compartments are further separated
from one another by the basal lamina which forms a
disc-like projection into the node gap between the astrocyte processes and the Schwann cell (Fig. 3a,b).
Here the basal lamina is reflected onto the external
surface of the Schwann cell paranode. Some collagen
fibrils lie in the gap between the two layers of basal
lamina. In the deeper parts of the invaginations the
astrocytic walls surrounding the invaginations are
commonly complete. More superficially they tend to be
incomplete. In type 1 zones the invaginations are
closely packed and as a result the astrocytic tissue is
arranged as a honeycomb with the mouths of the invaginations lying close to the plane of the surrounding
cord surface (Bristol and Fraher, 1989; Livesey and
Fraher, 1992).
In type 2 TZs the central end of each transitional
Schwann cell lies in an individual open-sided groove on
the surface of the central tissue projection (Fig. 3b).
a
b
I
Is
e*=
r
-
S
Fig. 3. a:Diagrams showing the relationship of astrocytic tissue to
myelinated fibres in a type 1transitional mne in planes longitudinal
(Is) and transverse (ts) to the rootlet as shown in the key at top left.
The central end of each transitional Schwann cell myelin sheath (S)is
invaginated below the plane of the CNS surface (IS). Each invagination is walled by astrocytic tissue arranged like a honeycomb (ts).The
basal lamina covering the Schwann cell is continuous with that covering the astmcytic tissue lining the invaginations at the level of the
transitional node (N), where it projects into the node gap. Central to
the node the myelin sheath is formed by a transitional oligodendroA fringe of slender glial processes (F) projects distally for a
cyte (0).
short distance beyond the plane of the surrounding CNS surface. Astrocytic tissue: shaded. Myelin: black. Basal lamina: dashed lines.
@&produd from Fraher, 1992, with permission of Pergamon Press
plc.) b Diagrams showing the relationship of astmcytic tissue to myelinated fibres in a type 2 transitional zone in planes longitudinal (Is)
and transverse (ts)to the rootlet, as shown in the key at top left. The
central end of each transitional Schwann cell myelin sheath (S)lies in
a groove on the surface of the central tissue projection, the wall of
which is formed by astrocytic tissue. The basal lamina covering the
Schwann cell is continuous with that of the astrocyte processes forming the groove at the transitional node (N), central to which the myAstrocytic
elin sheath is formed by a transitional oligodendrocyte(0).
tissue: shaded. Myelin: black. Basal lamina: dashed lines. (Reproduced from Fraher, 1992, with permission of Pergamon Press plc.)
Fig. 4. Electron micrographs of CNS-PNStransitional nodes. a:
Longitudinal section through an adult rat transitional node and the
related peripheral (leR) and central (right) paranodes. The peripheral
paranode lies in an invagination (arrowheads)bounded externally by
astrocyte processes. x 9,900. b Longitudinal section through a deep,
narrow node gap. A slender astrocyte process (arrowhead) projects
into the gap. It separates the oligodendrocyteterminal pockets from a
deep invagination of the extracellular space bounded on both sides by
basal lamina (arrow).At the bottom of the gap the two layers of basal
lamina are continuous with one another and project further into the
node BB a single layer (double arrows) which separates loose oligoden-
drocyte pockets (above)from a Schwann cell cytoplasmic process (below). x 35,100. c: Longitudinal section through a transitional node,
showing an astrocyte process extending into the node gap from an
adjacent perikaryon. x 17,500.d,e: Transverse &ions through definitive transitional nodes. d Microvilli (leR) overlying the nodal
axon (right) are irregular and entangled with one another. X 35,600.
e: The axon has a dense subaxolemmal undercoating and is covered
with irregular Schwann cell processes. x 45,500. (Parts b, d, e reproduced from Fraher and Kaar, 1986, with permission of Cambridge
University Press.)
512
J. FRAHER
Fig. 5. Diagrammatic longitudinal sections through ventral rootlets containing (a)type 1 and (b)
type 2 transitional zones of cervical and lumbar ventral rootlets, respectively, a t right angles to the
surface of the spinal cord (arrows), showing typical patterns of transitional node distribution in each
(dots).
This averages around 20 pm in length and often surrounds the myelinated internode progressively more
completely as the transitional node is approached.
Most transitional nodes lie at the central end of such a
groove or else in a short invagination continuous with
its central end.
Deep to the transitional node in both types of zone,
the paranodes of transitional oligodendrocytic myelin
sheaths of adjacent fibres are separated by astrocytic
tissue. With increasing depth, this becomes progressively less in amount and the neighbouring myelin
sheaths gradually approach one another as the thickness of the astrocytic partition between them lessens.
Within the intramedullary rootlet oligodendrocytic
sheaths are commonly directly opposed to one another
in the manner typical of many sheaths throughout the
CNS (Peters et al.. 1991).
NODE DISTRIBUTION
The astrocyte processes of the TZ are arranged and
distributed so that they tend to isolate the transitional
nodes and the paranodes proximal and distal to them
from neighbouring fibres. Node separation is also increased by having nodes of adjacent fibres longitudinally offset relative to one another (Fig. 5a,b). In type
1 TZs, most nodes of a given bundle are irregularly
distributed within a short segment of rootlet close to
the plane of the cord surface. While nodes of neighbouring fibres may lie at markedly different levels, a tendency is sometimes evident for the more laterally
placed nodes to lie deeper than those in the medial part
of the rootlet (Fig. 5a). The disc of rootlet in which the
nodes lie is therefore inclined to face ventrolaterally. In
a study of node distribution in a large number of TZs,
the most superficial node was found on average to lie
19 pm above the plane of the cord surface and the deep
est 13 pm below it (Bristol et al., 1993).
Nodes of type 2 TZs are also longitudinally offset
relative to one another, by the very fact of being distributed over the surface of the central tissue projection. Nevertheless, they are usually not uniformly distributed. Three-dimensional reconstructions (Fig. 6 )
show that in some instances they are clustered on the
surface of the central tissue projection, while other areas of the projection surface are devoid of them. Clustering is prominent in some rootlets, but is absent in
others. Fewer nodes tend to be found on the dorsal than
on the ventral surface of the central tissue projection.
This accords with the eccentric position of the projection in the rootlet. Because of this, the ventral surface
of the projection is more extensive than the dorsal and
so more fibres will penetrate the former than the latter.
While nodes of large and small ventral motoneurone
fibres (see below) are frequently intermingled, a tendency is sometimes seen for each type to be aggregated
into clusters on the central tissue projection surface
(Figs. 5b,6).
MAmRIALS AND METHODS
Node Density in CNS, PNS, TZ
Studies of node density on central, transitional, and
peripheral segments of ventral motoneurone axons
were carried out on alternating sequential series of
thin and semithin sections extending along the TZs
and also along the intramedullary bundles and the
ventral root fascicles central and peripheral to them,
respectively.
Specimens were futed, exposed, removed, and processed for routine transmission electron microscopy,
details of which have been given previously (Fraher &
Bristol, 1990; Fraher and Kaar, 1984; Kaar et al.,
1983). In brief, rats were anaesthetized with a 1:3 ::
ch1oroform:ether mixture and killed by intravascular
perfusion of fixative (2% glutaraldehyde and 2.5%
paraformaldehyde in an orthophosphate buffer at a pH
of 7.2-7.4). Supplementary fixation was carried out by
irrigating the spinal subarachnoid space with the same
fixative (Kaar et al., 1983). Twenty-five specimen
blocks from mid-lumbar levels containing the distal
ventral root, transitional zone, and intramedullary
rootlet (Fig. 1)were prepared from each animal. Alternating sequential series of thick (0.5 pm) and thin (100
nm) transverse sections at each of these three levels
were cut on a previously calibrated Reichert OMU4
Ultracut-E ultramicrotome. Serial light and electron
micrographs were taken of the whole transverse sec-
Fig. 6. Views of computerized three-dimensional representations
of central tissue projections (CTPs) into rat lumbar ventral rootlets,
showing the distribution of transitional nodes of large (unmarked
pale spots)and small (pale spots containing solid circles) fibres. a 4
CTP 1 from lateral, medial, dorsal, and ventral viewpoints, respectively; e-g: CTP 2 from medial, dorsal, and lateral viewpoints, respectively; h-k CTP 3 from ventral, medial, dorsal, and lateral viewpoints, respectively.
514
J. FRAHER
TABLE 1 . Mean node density in lumbar intramedullary rootlets
IIMR). tmnsitional zones (TZ).and ventml roots NR)'
Mean (* s.e.m.)
IMR
6.6 (? 0.8)
Tz
VR
13.8 ( 5 1.5)
2.0 ( 2 0.2)
Ratios
Tz/IMR
TZNR
2.3
7.8
'Ratios between densities also shown. (Reproduced from Fraher and Bristol,
1990, with permission of Cambridge University Press.)
tion of the TZ at 6 pm intervals over its entire length.
These were used to trace individual fibres and locate
nodes throughout each series.
Ten upper cervical spinal segments and attached
roots were obtained, each from a different adult animal, and were processed for transmission electron microscopy as described above. Fifty-four TZs were examined by making alternating sequential series of thin
and semithin sections. The series of sections were extended both centrally and peripherally beyond the limits of the TZ into the intramedullary and ventral rootlets, respectively. The positions of nodes and the
outlines of the TZ were determined on photomicrographs as before.
Node density was compared at CNS, PNS, and TZ
levels of lumbar rootlets, which contain type 2 TZs
(Figs. 1,2). To do this, individual rat lumbar ventral
motoneurone axon bundles were examined in intramedullary rootlets, ventral roots, and TZs. Node
density was estimated simply as number of nodes per
unit volume in each of 25 ventral rootlets, (1)in the TZ
as a whole and (2) in the central tissue projection. To
calculate TZ volume the outline of the TZ was first
traced on the photomicrograph a t each TZ level, i.e., a t
6 pm intervals. TZ cross-sectional area was measured
from each photomicrograph, using a Kontron Mini-IPS
automatic image analyser. In most cases the cross sectional area of the TZ varied between levels. Accordingly, the volume of the segment between each pair of
serially adjacent levels of section measured was estimated as that of a solid conical segment (Documenta
Geigy, 19621,the height of which was the sum of the
thicknesses of the intervening thick and thin sections.
By adding their volumes, the total TZ volume was
readily calculated. A computerised reconstruction of
each such conical segment was carried out using the
Kontron mini-IPS or IBAS system. Summation of these
segments enabled three-dimensional reconstructions of
the entire TZ to be produced. Nodes were identified on
the photomicrographs and on the intervening sections.
The total node number within each TZ was determined
and node density was calculated as the number of
nodes per unit volume of TZ (Table 1).The central tissue projection was also readily identifiable on transversely sectioned lumbar rootlets, since it was pale and
contrasted with the surrounding PNS tissue. Its outline was traced on the photomicrograph at each level.
Accordingly, three-dimensional reconstructions of central tissue projections (Fig. 6)were produced and their
dimensions and volumes measured, as described above
for the TZ. Node density was also calculated at different levels along the TZ and in rootlets of different size
(Table 2).
TABLE 2. Mean node density in lumbar TZ (DJ and central tissue
pmjection (CTP)(DJ'
(a) In small (15 fibres or fewer), medium (16-30 fibres), and large
(more than 30 fibres) rootlets
Rootlet
trpe
Small
Medium
Large
Mean no.
of fibres
12
21
42
DCt nodes/
D, nodes/
lo5 pmS
22.4
41.0
10.2
16 km3
36.4
17.6
18.6
(b) In moximal. middle and distal thirds of the TZ.
Rootlet level
Proximal third
Middle third
Distal third
D, nodes/
(? s.d.1
11.2(? 8.1)
14.0 (2 7.3)
15.8 (213.1)
los urn'
D,, nodes/
10' um' (? 8.d.)
13.2 ( ? 10.1)
26.3 (r15.9)
72.5 (r36.7)
'(Repmduced from Fraher, 1992, with permission of Pergamon
plc.)
Node density within the intramedullary rootlet and
in individual ventral root fascicles was calculated in a
similar way. Intramedullary rootlets are generally
compact on cross section but are in some cases deeply
indented by the surrounding central nervous tissue of
the spinal cord. However, they are usually not delineated sharply from the surrounding cord tissue by any
clearly defined barrier (Fig. 7c). This necessitates unequivocal identification of the intramedullary bundle
fibres. This was achieved by following all ventral motoneurone fibres centrally from the TZ. To do this, all
constituent fibres were numbered a t the TZ and traced
centrally on serial photomicrographs as far as the surface of the ventral horn grey matter. The cross-sectional area of the intramedullary rootlet at any given
level was defined as the sum of the areas of all such
fibres, including their myelin sheaths. Some astrocytic
and oligodendrocytic tissue which contributed to the
intramedullary bundle was excluded in measuring
cross sectional area for the following reasons. Cell nuclei and perikarya apposed to the bundle were not included in cross-sectional area calculation because both
astrocytes and oligodendrocytes contribute in varying
degrees to both the intramedullary rootlet and the surrounding central nervous tissue (Fig. 74. Blood vessels
were also excluded in area calculation. On the basis of
these criteria, and using the Kontron-IPS system, the
outline of the intramedullary rootlet a t each level was
traced on the photomicrograph over the digitizer tablet
and its cross-sectional area measured. Using these values, the volume of the length of intramedullary rootlet
studied was calculated as for that of the TZ, described
above. The total number of nodes in each length of
rootlet studied was determined by examining the photomicrographs and serial transverse sections in the
photomicroscope. Node density was then calculated as
before (Table 1).
PNS node density was measured in the distal ventral
root immediately proximal to its junction with the dorsal root ganglion. At this level nerve fibres form clearly
defined fascicles, each bounded by a perineurial
sheath. The node density in 54-60 pm lengths of 36
different fascicles (7 from each of 4 animals and 8 from
TZ NODE DENSITY AND DISTRIBUTION
515
Fig. 7. Transverse sections through (a) a ventral rootlet, (b)a TZ, (c) an intramedullary bundle, at
the upper cervical level. In b the TZ is in the upper half of the illustration, transitional fibres being
surrounded by pale astrocytic tissue. In c the intramedullary bundle fibres are sectioned mostly transversely, while those of the surrounding cord are sectionedlongitudinally.Bars:10 pm. (Basedon Bristol
et al., 1993. Reproduced with permission of S. Karger AG.)
the fifth) was calculated from the volume and the total
number of nodes in each, in a manner similar to that
described above, using photomicrographsof every sixth
serial transverse section (Table 1).
Node Spacing in TZ
Estimation of node density simply as number per
unit volume is relatively crude. It fails to take account
of a number of features of transitional node distribution which are evident from three-dimensional reconstructions (Figs. 5,6). These include variations in longitudinal and transverse nodal separation, clustering
of nodes, and also the frequently noted absence of nodes
from large parts of the TZ. Node spacing can be more
accurately estimated by measuring the distance between a node and its nearest neighbour (the nearest
neighbour distance).
Nearest neighbour distance was studied in type 1
and type 2 TZs. The alternating sequential series of
thick and thin sections of upper cervical and mid-lum-
bar TZs used were prepared as described in Node Density in CNS, PNS, TZ, the material used for the lumbar
study being the same as for the node density study. In
both types of zone, photomicrographs of serial sections
extending over the entire length of the TZ were used.
Using the Kontron-IPS system the outline of the TZ
was determined from each photomicrograph. The position of each transitional node within it was identified
and its coordinates recorded, using the digitizer tablet.
Computer-assisted three-dimensional reconstructions
of the distribution of all nodes in each TZ studied were
thereby produced. The programme used allowed stacking of the serial profiles of the TZ (generated from the
digitizer tablet). It also provided the three-dimensional
coordinates of the centre of each node within the TZ.
Using these, the distance between each node centre
and the centre of its nearest neighbour was calculated:
its nearest neighbour distance. This was determined
for each TZ examined.
In the mid-lumbar rootlets, nearest neighbour dis-
516
J. FRAHER
TABLE 3. Mean (? s.e.m.) nearest neighbour distances for nodes in
proximal, middle, and distal thirds of lumbar TZs, and for TZ
as a whole
h x i m a l third
Nearest neighbour distance (pm)
Middle third
Distal third
a
Whole TZ
(Repmduced from Fraher, 1992, with permission of Pergamon Press plc.)
tances were calculated for large and small fibres separately. Fibres were separated into the two classes on
the basis of their peripheral internodal diameters.
Each fibre in the TZ was traced distally from its node
using the serial photomicrographs. Its peripheral internodal diameter was measured from electronmicrographs of thin sections immediately distal to the TZ.
Axon calibre was bimodally distributed. It was therefore possible to allocate each fibre within the distribution to the large or the small class on the basis of probability plots (Cox, 1966) using an iterative method
based on the EM algorithm (Dempster et al., 1977;
Kaar and Fraher, 1985). The dividing diameter between the classes was 7.0 pm. Each node was therefore
identified as belonging to the large or small class, and
distributions of nearest neighbour distances were produced separately for each fibre class. Nearest neighbour
distance was calculated for 439 large and 300 small
ventral rootlet fibres. The two fibre classes were not
distinguished in the cervical TZs. Mean values and
overall distributions (Table 3) were determined (Fig. 8).
To permit comparison of node density and node spacing estimates at cervical and lumbar levels, node density was calculated for cervical TZs in a manner similar
to that described in Node Density in CNS, PNS, TZ for
lumbar TZs. Using the data derived from this, various
type 1 and type 2 TZ parameters were also compared
(Table 4).
Fibre Spacing in TZ
As they pass through the TZ, myelinated fibres seem
to be relatively widely separated from one another
(Fig. 7b). This is the case even though many are nodal
or paranodal at any particular level of section and so
have a smaller cross sectional area than elsewhere.
The amount of interstitial tissue seems to be significantly increased compared with both the intramedullary rootlet and the ventral rootlet and serves to separate the fibres from one another. To study this, fibre
spacing in the TZ was compared with that in the ventral and intramedullary rootlets by measuring nearest
neighbour distances and determining the amounts of
interstitial material separating the myelinated fibres
at each of the three levels.
This study was carried out on upper cervical rootlets.
These were used to examine separation in a population
consisting predominantly of ventral motoneurone fibres, since they contain little if any autonomic outflow.
Electron microscopic studies confirmed that few if any
unmyelinated or thinly myelinated fibres were present
in them. This permitted examination of rootlet composition at light microscopic levels. Specimens were prepared by standard methods for electron microscopy
a
w 25
P
b
20
15
10
5
0
12
2l
30 -I
>33
N. N. DIST.
(crW
Fig. 8. Percentage frequency histograms showing the overall distribution of nearest neighbour distance (N.N. DIST.) (black) in (a)
upper cervical and (b)mid-lumbar rootlets. In b the histogram also
shows the contribution of values for gamma fibres (white)to the total.
(Reproduced from Bristol et al., 1993,with permission of S. Karger
AG.)
TABLE 4.Mean (t s.d3 ovemll values at cervical and lumbar levels
for: TZ length, TZ volume, fibre number, and transitional
node density*
Length
Volume
Fibre
Nodedensity
Level
(s.d.) pm (8.d.) lo3 pm3 no. (s.d.) (s.d.) n0./105pm3
Cervical
29.4 (5.6)
69.7 (37) 13.6 (3.8)
27.8(8.7)
Lumbar 163.4(49.4) 322.0(120) 31.1(6.4)
13.8(4.3)
Ratio U C
5.5
4.6
2.3
0.5
*All correspondingmean cervical and lumbar values dif€ered significantly from
one another Ip 5 0.01).
(Bristol et al., 1993; Fraher and Kaar, 1985; Kaar et
al., 1983). Alternating sequential series of thin and
semithin transverse sections of adult rat upper cervical
rootlets were prepared as described above (Node Density in CNS, PNS, TZ).The series extended from the
intramedullary rootlets through the TZs and into the
ventral rootlets. In each of over 50 rootlets, computerised morphometric analysis was carried out on photomicrographs a t a single level of the entire cross section of the intramedullary rootlet, of the ventral rootlet
and also at the mid-level of the TZ. Packing density
was measured in two ways. Firstly, the proportion of
bundle cross-sectional area made up by fibres and by
interstitium was measured a t each level for each rootlet (Table 5a). For this purpose, the fibre was defined as
the axon together with all components of the oligodendrocyte or Schwann cell surrounding it. The remaining
tissues comprised the interstitium. These consisted of
the cellular and extracellular components of the endoneurium in the ventral rootlet, and of astrocytic tissue
together with extracellular space both in the intramedullary rootlet and a t the mid-level of the TZ. The sec-
517
TZ NODE DENSITY AND DISTRIBUTION
TABLE 5. Mean packing density of cervical ventml motoneurone fibres at intmmedullary bundle (IMB),
transitwnul ZOM (TZ),and ventml rootlet (VR)levels'
(a) Percentage of bundle
composed of fibres
IMB
TZ
VR
(b) Nearest neighbour
distance (Fm)
(c) Mean fibre crosssectional area (pm2)
Animalno.
IMB
TZ
VR
IMB
TZ
VR
Overall
mean (s.d.1 62.5 (2.9) 33.4 (4.9) 80.0 (0.2) 5.1 (0.48) 7.1 (0.71) 6.7 (0.13) 28.6 (4.6) 47.9 (9.9) 62.0 (8.4)
n
21
16
15
265
283
240
119
109
110
'The table shows (a) the proportionof the bundle made up by fibres (interstitiumcomprises the remainder), (b) nearest neighbour
distance values, and (c) fibre cross sectional area. For each of these parameters the mean value is given for each animal studied,
as is the overall mean value together with the standard deviation. IMB:Fibres composed of axon8 and ensheathing oligodenh y t i c myelin and cytoplasm. TZ Fibres composed of axona, ensheathing Schwann cell, or oligcdendrocytic myelin and c y b
plasm, 88 well as transitional nodes. VR: Fibres composed of axona and ensheathing Schwann cell myelin and cytoplasm. n:
Number of rootlets examined in calculating percentage composition. of fibre8 examined in calculating nearest neighbour distance, and of fibres examined in calculating cross-sectional area. (Reproducedfrom Bristol et al., 1993, with permission of S.
Karger AG.)
ond estimate of packing density used nearest neighbour distance. This was also calculated for all fibres in
the single transverse plane selected as for the area
measurements at each of the three levels. It was determined from the coordinates of the centre of gravity of
each fibre at each level (Table 5b). Measurements
made in the TZ used the centres of all fibres irrespective of whether they had been sectioned at nodal, paranodal, juxtanodal, or internodal levels. In addition to
the above, using the same sections, mean fibre cross
sectional area was determined a t ventral rootlet, TZ,
and intramedullary rootlet levels (Table 5c). This included the axon, the myelin sheath and any cytoplasm
of the myelinating cell external to this. To determine
cross-sectional area, all transversely sectioned fibres a t
each of these three levels were numbered and over 100
fibres were selected randomly for measurement. The
results are given in Table 5.
RESULTS
Node Density
Node density, measured in lumbar motoneurone
axon bundles, was greatest in the TZ (Table 1).It was
statistically significantly greater (P < 0.001) than in
either the intramedullary rootlet or the ventral root. It
averaged more than twice the density measured in the
former and almost eight times that in the latter. Since
all the nodes in the TZ are located on or in the central
tissue projection, a more accurate estimate of transitional node density is given by their density within
this. This may be regarded as the better of the two
estimates, since there are few if any nodes in the PNS
tissue compartment of the TZ, virtually all nodes lying
on the surface of or within the central tissue projection.
The number of nodes per unit volume of the central
tissue projection was readily determined from the computerised reconstructions and was found to be 26.3
nodes/106 pm3, approximately twice that for the TZ as
a whole. As regards intramedullary node density, the
method used to determine this overestimated the packing density because some astrocytic and oligodendrocytic material which contributed in part at least t o the
intramedullary rootlet was excluded in calculating intramedullary bundle volume (Fig. 7) (see Materials
and Methods.) For the above reasons, therefore, it is
likely that true transitional node density is probably
substantially more than four times greater than that in
the intramedullary rootlet and is around fifteen times
greater than in the ventral root. The markedly greater
density of transitional nodes is clearly related to the
fact that all fibres possess a node as they traverse a
short length of the bundle. Consequently, the nodes are
much more nearly in register with one another than at
either CNS or PNS levels, where no such restrictions
on their distribution apply.
If node densities along the intramedullary rootlets
and in the ventral root reflect values generally within
the CNS and PNS, respectively, then the TZ may be a
site of uniquely high node density. Transitional node
density seems to have some relationship to rootlet size.
Density in the central tissue projection was significantly greater in small roots containing 15 or fewer
myelinated fibres than in the remainder (Table 2a).
Node density also varied between proximodistal levels
along the TZ. Values were determined a t proximal,
middle, and distal thirds. While they showed some tendency to increase proximodistally within the TZ as a
whole, this was much more marked within the central
tissue projection (Table 2b) where the density was over
five times greater in the distal than in the proximal
third.
Node Spacing
Node spacing, estimated by nearest neighbour distance, was measured for type 1 and type 2 TZs a t upper
cervical and mid-lumbar levels, respectively. Mean
values for centre to centre spacing were similar a t both
levels, being 14.1 (s.d. 7.4)and 12.7 (s.d. 6.7)pm, respectively. The similarity of values for the two zone
types is also evident in the distributions of nearest
neighbour distance a t cervical and lumbar levels (Fig.
8). While the distributions of neu-est neighbour distances for large and small lumbar fibres had a similar
form, the nodes of small fibres tended to have significantly smaller nearest neighbour distances than those
of large fibres (Fig. a), the mean values being 9.7 and
12.2pm, respectively. Around 40% of small nodes, and
a similar proportion of large nodes had nearest neighbour distances of 10 pm or less. Nearest neighbour distance varied considerably both within and between the
groups of rootlets which went to form individual roots.
Despite this, there was a tendency for nodes to be furthest apart in the proximal third of the lumbar TZ and
closest together in the middle and distal thirds, in
518
J. FRAHER
which they were similarly spaced (Table 3). Nodes with
particularly large nearest neighbour distances tended
to be located at the proximal or distal end of the TZ.
The data obtained for cervical and lumbar TZs allowed comparison of the two TZ types in terms of a
number of parameters other than nearest neighbour
distance (Table 4). In terms of their overall length and
volume, cervical zones were on average only one-fifth
the size of lumbar zones. Cervical rootlets contained
somewhat less than half (44%) of the numbers of fibres
found in lumbar rootlets. Mean node density, estimated
simply as number per unit volume, in cervical TZs was
twice that in lumbar zones.
The marked disparity between the patterns of findings for node density and nearest neighbour distance
on comparison of the cervical and lumbar TZs (Table4)
can be explained in terms of node distribution within
the zones. Density appears lower in lumbar than in
cervical TZs, because in the former the nodes tend to be
distributed on the surface of the central tissue projection (Figs. 5,6). As a result, few are present within its
substance. Accordingly, substantial volumes of lumbar
TZs lack nodes. These voids, however, contribute to
central tissue projection volume, which leads to a low
density value. This highlights the lack of precision of
simple density measurement. It also reflects the common occurrence of node clustering in type 2 TZs. There
is a greater tendency for small than for large nodes to
be clustered with other nodes of the same class (Fraher,
1992). This is shown by comparison of nearest neighbour distances. About 30% of small nodes and 20% of
large nodes have nearest neighbours of the same class
less than 10 pm distant. This is counter to the relative
proportions of numbers of nodes of the two size classes
within the rootlet, large nodes being 50%more numerous than small nodes. This underlines the preferential
tendency for small fibres to be closer together than
large fibres.
In the cervical TZs, in contrast to the lumbar, there
are few large areas which entirely lack nodes and
which consist principally of astrocytic tissue. Here the
main factor which might tend to distort values for node
density is the occurrence of transitional nodes at unusually distal levels, within the rootlet, or unusually
proximal levels, deep to the cord surface. Both of these
factors would result in high values for TZ length and
volume and therefore would tend to lower values for
node density. As with the lumbar TZs, estimation of
node packing density using nearest neighbour distances avoids any distortion of values following from
the lack of refinement inherent in calculating TZ volume from relatively simple measurements.
The frequency distribution of nearest neighbour distances reflects the range of variation in the degree to
which nodes are offset relative to one another. It is
noteworthy that the distributions are similar at cervical and lumbar levels. This indicates that nodal offsetting is very similar in both locations, despite the wide
differences in basic TZ morphology and dimensions and
in node density and distribution. Nodal offsetting occurs in lumbar zones through the nodes being distributed over the surface of the irregularly conical central
tissue projection and in cervical zones through the
nodes on adjacent fibres being placed at different levels. These findings suggest that similar mechanisms
may operate during development to determine the
spacing of nodes and their locations along neighbouring fibres in the two different types of TZ.
At both cervical and lumbar levels very similar proportions of nodes were less than 10 pm from the nearest neighbour, from node centre to node centre. The
true distance between adjacent nodal axolemmae is
considerably less than this, especially in the case of
large diameter fibres. Assuming a nodal diameter of
half that of the internodal axon (Kaar, 1984), an approximate nodal axonal diameter of 5 pm may be estimated for large fibres. If the centres of such a pair of
nodes, located in the same transverse plane, were separated from each other by 10 km, then their axolemmae would be separated by 5 pm, a distance corresponding to as little as one axonal diameter. If nodes
with the same centre to centre spacing are longitudinally offset from one another by a distance less than
the length of the nodal axolemma, then the actual separation between the axolemmae would be even less
than that. Such close proximity raises the possibility of
interaction between the fibres involved.
Fibre Spacing
Morphometry strongly confiied the impression
(Fig. 7) that the proportion of interstitium in the TZ
was considerably greater than in either the intramedullary or the ventral rootlets, being over three times
that in the former and almost twice that in the latter
(Table 5a). It more than compensated for the relatively
small mean cross-sectional area of the fibres as they
traversed the TZ.As a result, fibres in the TZ had a
greater mean nearest neighbour distance than those in
the ventral rootlet. However, the difference was not
statistically significant. Fibres constituted only one
third of the bundle cross sectional area in the TZ,
whereas at the other two levels the proportion which
they comprised was around twice this or even more. In
the rat cervical TZ, stereological analysis shows that
the interstitium consists in large part of astrocyte tissue, though small proportions of basal lamina, extracellular space, and collagen fibres are present, especially in its superficial third (Fraher and Kaar, 1982).
The presence of large amounts of interstitium is related to relatively wide separation of the fibres in the
TZ. This may lessen the likelihood of functional interaction between them. Astrocytic lamellae are in some
cases concentrically arranged around the nodal, paranodal, or internodal segments of individual fibres in
the TZ.This is frequently the case during TZ development, especially in relation to bare axon segments at
and central to the level of the presumptive transitional
node (Fraher, 1978a). More commonly, the spaces between TZ fibres consist of fine astrocyte processes of no
preferred orientation, separated by extracellular space,
and in some areas surrounded by basal lamina. The
resulting series of cell membranes and extracellular
material could present significant barriers to the passage of ions or the flow of electrical currents between
the fibres so separated from one another and would also
TZ NODE DENSITY AND DISTRIBUTION
provide extensive areas of surface for uptake of ions
from the interstitial spaces.
FUNCTIONAL CORRELATES
If the node densities in the intramedullary and ventral rootlets resemble those generally found in the CNS
and PNS, respectively, then node density is greater in
the TZ than elsewhere. Accordingly, the possibility of
interaction between myelinated fibres because of proximity of their nodes and the need to maintain a stable
ionic environment may be greater in the TZ than in
any other location. Since the greatest node densities
were found in the smallest rootlets, any interaction is
likely to be greatest in these.
It is known that the electrical fields around active
myelinated fibres in the PNS induce changes in the
excitability of neighbouring inactive fibres in vitro
(Blair and Erlanger, 1932;Marrazzi and Lorente de N6,
1944; Rosenblueth, 1941, 1944). Stampfli (1954) suggested that adjacent myelinated fibres could influence
one another's activity if their nodes were close enough.
Since action potential magnitude, electrical fields, and
ionic fluxes are maximal at nodes (Tasaki and Tasaki,
19501, it is here that adjacent fibres are most likely to
interact with one another. Nearest neighbour distances
show that the axolemmae of transitional nodes are
about 5 Fm distant from one another in around one
third of both large and small L4 fibres (Bristol et al.,
1993). "his, however, is a relatively long distance,
given the potential pathways for shunting in the fibre
bundle. The probability of interaction between transitional nodes is also lessened by the fact that the internodes bounding them centrally and peripherally are
shorter than average (Carlstedt, 1981; Fraher, 1978b).
This follows from the theoretical studies of Waxman
and Brill (1978), which showed that short internodes
facilitate transmission past focally demyelinated
zones. They could have a similar effect along lengths of
fibre with short internodes, thereby increasing the
safety factor (Hildebrand et al., 1985). Certain of the
morphological features of the TZ would therefore seem
not to favour impulse spread between the transitional
nodes of Ranvier, even though these are more closely
packed than elsewhere.
The tendency for astrocyte processes to project into
the central portion of the transitional node gap shows
some similarities with their arrangement at CNS
nodes. Astrocyte processes also project into the node
gap at CNS nodes, where they become closely associated with the nodal axolemma (Black and Waxman,
1988;Gilmore, 1963,1971;Hildebrand, 1971a,b).It has
been suggested that the astrocyte processes in conjunction with the oligodendrocyte together perform functions at CNS nodes corresponding to those carried out
by the Schwann cell in relation to PNS nodes (Black
and Waxman, 1988). As at CNS nodes, the arrangement and size of the astrocyte processes extending into
the transitional node gap lack a constant pattern.
These lie central to the disc of basal lamina, which also
projects into the node gap and which represents the
continuity between the Schwann cell basal lamina of
the PNS and the astrocyte basal lamina of the glia
limitans. Peripheral to the disc of basal lamina the
519
structure of the node closely resembles that found in
PNS nodes generally, since it contains microvillous
processes arising from the central end of the transitional Schwann cell. If, as suggested by Black and
Waxman (19881, the differences in structure of CNS
and PNS nodes reflect different physiological characteristics at them, then transitional nodes are likely to
be complex with the basal lamina forming a sharply
defined structural boundary between the two environments. However, its effectiveness as a functional
boundary must be doubtful.
The exceptionally high proportions of astrocytic tissue in the TZ as a whole, combined with the fact that
astrocyte processes extend into the transitional node
gap, could influence node function in a number of
ways. Perinodal astrocyte processes may be involved in
the production of gap substance, the regulation of extracellular ion levels, or the regulation of node metabolism (Hildebrand, 1971a,b), functions corresponding
to those proposed for perinodal Schwann cell processes
peripherally (Black and Waxman, 1988; Landon and
Hall, 1976; Landon and Langley, 1971; Rosenbluth,
1981; Waxman and Black, 1987). They could also be
involved in the synthesis and turnover of ion channels
and perhaps of other axolemmal components which are
subsequently transferred to the nodal axolemma
(Black and Waxman, 1988; Gray and Ritchie, 1985).If
there is a demand for such functions in the TZ, the high
density of nodes is well matched by the large amount of
astrocytictissue projecting into and surrounding them.
These functions could also be performed by the
Schwann cell microvilloid processes related to the peripheral parts of the node gap.
Astrocyte plasma membranes in tissue culture include numerous ion channels (reviewed by Barres et
al., 1990). If these are present in vivo they could be
involved in ion fluxes. In the TZ there are large numbers of small astrocyte processes, often with complex
irregular profiles, and therefore having high surface to
volume ratios (Rossiter and Fraher, 1990). These processes could therefore carry large numbers of ion channels with the capacity for extensive ion movements
through them. Good evidence exists (Barres et al.,
1990) that glial cells can regulate extracellular potassium by passive flux,which is in some cases accompanied by chloride movements. These ions are stored in
astrocytes and later returned to neurons. These processes may help to regulate extracellular ion concentrations. Such a function could be especially important
in the TZ where node concentration is exceptionally
high. The dense mass of astrocyte processes in the TZ,
surrounding the paranodes of the fibres traversing it
and also extending into the node gap, could function in
this way. They could prevent ion diffusion away from
the nodes and the consequent development of high concentrations in the extracellular spaces throughout the
TZ, which could in turn interfere with nodal function.
The large amounts of densely packed astrocyte processes in the TZ, by being intimately related to the
nodal axon and by compartmentalizingthe nodes and
their adjacent paranodes, could control the environment of the transitional nodes and help to maintain the
conditions in which they function efficiently. Ion chan-
520
J. FRAHER
nerve fibers by extrinsic action potentials. Am. J. Physiol., 101:
559-564.
Bristol, D.C., and Fraher, J.P. (1987) A morphometric study of the
CNS-PNS transitional zone in rat lumbar ventral roots. J. Anat.,
152:236.
Bristol, D.C., and Fraher, J.P. (1989) Experimental traction injuries
of ventral spinal nerve roots. A scanning electron microscopic
study. Neuropathol. Appl. Neurobiol., 15:649-561.
Bristol, D.C., Cremin, P.M., and Fraher, J.P. (1993) Spacing of central-peripheral transitional nodes of rat motoneurones. A d a Anat.,
148206-212.
Carlstedt, T. (1981) An electron-microscopicalstudy of the developing
transitional region in feline S1dorsal rootlets. J . Neurol. Sci., 50:
357-372.
Coggeshall, R.E. (1986) Non-classical features of dorsal root ganglion
cell organisation. In: Spinal Merent Processing. T.L. Yaksh, ed.
Plenum Press, New York, pp. 83-96.
Coggeshall, R.E., Emery D.G., Ito, H.,and Maynard, C.W. (1977) Unmyelinated and small myelinated axons in rat ventral roots. J.
Comp. Neurol., 173:175-184.
Cox, D.R. (1966) Notes on the analysis of mixed frequency distributions. Br. J. Math. Statist. Psychol., 1939-47.
Dempster, A.P., Laird, N.M., and Rubin D.B. (1977) Maximum likelihood fmm incomplete data via the EM algorithm. J.R. Statist.
SOC.,B391-38.
Documenta, Geigy (1962) 6th ed. Geigy, Basle, p. 143.
Fraher, J.P. (1978a) The maturation of the ventral mt-spinal cord
transitional zone. J. Neurol. Sci., 36427-449.
Fraher, J.P. (1978b)Quantitative studies on the maturation of central
and peripheral parts of individual ventral motoneuron axons. 11.
Internodal length. J. Anat., 127:l-15.
Fraher, J.P. (1992) The CNS-PNS transitional zone of the rat. Morphometric studies at cranial and spinal levels. Prog. Neurobiol.,
38261-316.
Fraher, J.P., and Bristol, D.C. (1990) High density of nodes of Ranvier
in the CNS-PNS transitional zone. J. Anat., 170:131-137.
Fraher, J.P., and Kaar, G.F. (1982) The maturation of the ventral
root-spinal cord transitional zone. Part 2. A quantitative ultrastructural study of the dynamics of its early development. J. Neurol. Sci.,
53:63-75.
Fraher, J.P., and Kaar, G.F.(1984) The transitional node of Ranvier
at the junction of the central and peripheral nervous systems: An
ultrastructural study of its development and mature form. J. Anat.,
139215-238.
Fraher, J.P., and Kaar, G.F. (1985) The development of alpha and
gamma motoneuron fibres in the rat. 11. A comparative ultrastructural study of their central and peripheral myelination. J. Anat.,
141:89-103.
Fraher, J.P., and Kaar, G.F. (1986) The lumbar ventral root-spipal
cord transitional zone in the rat. A morphological study during
development and at maturity. J. Anat., 145:109-122.
Fraher, J.P., Kaar, G.F., Bristol, D.C., and Rossiter, J.P. (1988) Development of ventral spinal motoneurone fibres: A correlative study
of the growth and maturation of central and peripheral segments of
large and small fibre classes. Pmg. Neurobiol., 31:199-239.
Gilmore, S.A. (1963) The effects of x-irradiation on the spinal cords of
neonatal rats. II. Histological observations. J. Neumpathol. Exp.
Neurol., 22:294-301.
Gilmore, S.A. (1971) Neuroglial population in the spinal white matter
of neonatal and early postnatal rats: An autoradiographic study on
numbers of neuroglia and changes in their proliferative activity.
Anat. Rec., 171283-292.
ACKNOWLEDGMENTS
Gray, P.T.A., and Ritchie, J.M.(1985) Ion channels in Schwann and
glial cells. Trends Neurosci., 8:411-415.
The work reported was supported by a number of Hildebrand, C. (19718) Ultrastructural and light-microscopicstudies
grants from the Health Research Board of Ireland and
of the nodal region in large myelinated fibres of the adult feline
spinal cord white matter. Acta Physiol. Scand., 36443-71.
the Wellcome Trust.
Hildebrand, C. (1971b) Ultrastructural and light-microscopicstudies
of the developing feline spinal cord white matter. I. The nodes of
REFERENCES
Ranvier. Acta Physiol. Scand., 364:81-101.
Barres, B.A., Chun, L.L.Y., and Corey, D.P. (1988) Ion channel ex- Hildebrand, C., Remahl, S., and Waxman, S.G. (1985) Axo-glial relations in the retina-optic nerve junction of the adult rat: Electronpression by white matter glia: I. Type 2 astrocytes and oligodendrocytes. Glia, 1:lO-30.
microscopic observations. J. Neurocytol., 14:597-617.
Barres, B.A., Chun, L.L.Y., and Corey, D.P. (1990) Ion channels in Kaar, G.F. (1984) Development of the transitional region between
vertebrate glia. Ann. Rev. Neurwci., 13441-474.
central and peripheral nervous systems. A morphologicaland quanBlack, J.A., and Waxman, S.G. (1988) The perinodal astrocyte. Glia,
titative study of the L5 ventral root. Ph.D. Thesis, National Uni1:169-183.
versity of Ireland. Dublin. Ireland.
Blair. E.A.. and Erlamer. J. (1932) On the effects of uolarization on Kaar, G.F., and Fraher, J.P. (1985) The development of alpha and
nels in oligodendrocytes may also play a part in ion
fluxes related to axonal activity, whereby they also
accumulate potassium and chloride from the extracellular spaces (Barres et al., 1988). The tapering oligodendrocyte paranodes in the TZ are surrounded by progressively thicker sleeves of astrocyte processes as the
transitional node is approached. It could be that ions
taken up by these astrocyte processes are transferred to
the paranodal axon via the oligodendrocyte cytoplasm.
Gap junctions between the astrocyte processes and the
oligodendrocyte paranodal loops (Black and Waxman,
1988) would facilitate this. Alternatively, the astrocytic sleeve could take up ions which have diffused outwards through the oligodendrocytic paranodes. In either case, the astrocytic sleeve would function to
isolate the transitional paranode from its neighbours.
Transitional zones represent areas of exceptionally
high concentrations of nodes. Accordingly, even though
their structure is specialised relative to CNS and PNS
nodes, they are well suited for study of node functions
in vivo, especially of possible electrical interactions between them, as well as ionic fluxes in the nodal environment. Nodes in type 2 TZs are particularly suitable
because the zones lie in small discrete rootlets superficial to the cord surface (Bristol et al., 1993;Kaar and
F'raher, 1986, 1987) which can therefore be localised
relatively easily. These TZs possess a number of advantages for node study over those on central fibre bundles. In the latter, nodes are not in phase with one
another and their component fibres are likely to be
intermingled with those of other fibre bundles which
have different origins, courses, distributions, and functions. Type 2 TZs of dorsal rootlets are relatively more
easily accessed for experimental study than those in
ventral roots. However, those ventral rootlets which
lack a significant autonomic outflow, such as rat L4,
offer the possibility of studying a relatively pure population of myelinated axons mostly belonging to a single class, namely, those of ventral motoneurones (Coggeshall et al., 1977).These rootlets may contain a small
number of sensory axons, but only a few of these are
myelinated (Coggeshall, 1986;Coggeshall et al., 1977;
Kim et al., 1987;Loeb, 1976). Furthermore, the muscles innervated through individual ventral rootlets
may be identified electrophysiologically or by tracer
methods. This offers the possibility of correlating functional studies of nodes and their surrounding TZ astrocyte processes with those of the muscles innervated by
them.
I
TZ NODE DENSITY AND DISTRIBUTION
gamma motoneuron fibres in the rat. I. A comparative ultrastructural study of their central and peripheral axon growth. J. Anat.,
141~77-88.
Kaar, G.F., and Fraher, J.P. (1986) The sheaths surrounding the attachments of rat lumbar ventral roots to the spinal cord: A light and
electron microscopical study. J. Anat., 148:137-146.
Kaar, G.F., and Fraher, J.P. (1987) The vascularization of the centralperipheral transitional zone of rat lumbar ventral rootlets: A morphological and morphometric study. J. Anat., 150145-154.
Kaar, G.F., O'Sullivan, V.R., and Fraher, J.P. (1983) Combined Vaacular perfusion and in situ fixation of rat spinal nerve roots. Ir. J.
Med. Sci., 152:222.
Kim, J.9 Shin, H.K.9 and Chung, J.M. (1987) Many Ventral root Of
in the cat are third branches Of dorsalroot ganglion
cells. Brain Res., 417:304-314.
Landon, D.N., and Hall, S. (1976) The myelinated fiber. In: The Peripheral N
~ D.N.~Landon,
~ 4
..
chapman and ~ ~ 1London,
1 ,
pp.
,I-LVU.
rnc
Landon, D.N., and Langley, O.K. (1971) The local chemical environment of the node of Ranvier. A study of cation binding. J. Anat.,
108419-432.
Livesey, F.J., and Fraher, J.P. (1992) Experimental traction injuries
of cervical spinal nerve roots: A scanning EM study of rupture patterns in fresh tissue. Neuropathol. Appl. Neurobiol., 18376-386.
Loeb, G.E. (1976) Ventral root projections of myelinated dorsal root
ganglion cells in the cat. Brain Res., 106:159-165.
Marrazzi, A.S., and Lorente de N6 (1944) Interaction of neighbouring
fibres in myelinated nerve. J . Neurophysiol., 7:83-101.
521
Peters, Palay, and Webster (1991) The Fine Structure of the Nervous
System. 3rd Ed. Oxford University Press, Oxford.
k n b l u e t h , A. (1941) The stimulation of myelinated axons by nerve
impulses in adjacent myelinated =om. Am.J. Physiol., 132119128.
Rosenblueth, A. (1944) The interaction of myelinated nerve fibres in
A ~J.,physiol,, 1m656-670.
mammalian nerve
to ionophom localk n b l u t h , J. (1981) Fr-fraCtum
ization in normal and myelin-deficient nerves. In: Advances in Neurology, vol. 31. Demyelinating Diseases: Basic and Clinical Elecbven
press,
trophysiology~S.G. waxmanand J.M. Rit,.hie, d,
New York, pp. 391-418.
bssiter, J.P., and l+aher, J.P. (1990) ~ ~ t e ~
of central
i ~ and
~ l
peripheral nervous tissues in rat dorsolateral vagal rootlet transitional zones. J. Neurocytol., 19385-407.
S t h p f l i , R. (1954) Saltatov conduction in nerve. Physiol. Rev., 3 4
101-112.
Tasaki, I., and Tasaki, T. (1950) The electrical field which a transmitting nerve fiber produces in the fluid medium. Biochim. Biophys.
A&, 5~335-342.
Waxman, S.G., and Black, J.A. (1987) Macromolecular structure of
the Schwann cell membrane. Perinodal microvilli. J. Neurol. Sci.,
7723-34.
Waxman, S.G., and Brill, M.H. (1978) Conduction through demyelinated plaques in multiple sclerosis: Computer simulations of facilitation by short internodes. J. Neurol. Neurosurg. Psychiatry, 41:
408-416.
i
~
~
Документ
Категория
Без категории
Просмотров
3
Размер файла
1 780 Кб
Теги
paraproteinaemic, change, igm, neuropathy, paranodes, morphological
1/--страниц
Пожаловаться на содержимое документа