European Journal of Neurology 2006, 13: 528532

doi:10.1111/j.1468-1331.2006.01284.x

Topography of a-internexin-positive neuronal aggregates in 10 patients with

neuronal intermediate filament inclusion disease
R. A. Armstronga and N. J. Cairnsb
a
Vision Sciences, Aston University, Birmingham, UK; and bDepartments of Neurology, and Pathology and Immunology, Washington
University in St Louis, School of Medicine, St Louis, MO, USA

Keywords:

a-internexin, intermediate
laments, neuronal inclusions, neuronal intermediate lament inclusion
disease, topographic
pattern
Received 28 February 2005
Accepted 26 June 2005

Abnormal neuronal intermediate lament (IF) inclusions immunopositive for the type
IV IF a-internexin have been identied as the pathological hallmark of neuronal
intermediate lament inclusion disease (NIFID). We studied the topography of these
inclusions in the frontal and temporal lobe in 68 areas from 10 cases of NIFID. In the
cerebral cortex, CA sectors of the hippocampus, and dentate gyrus granule cell layer,
the inclusions were distributed mainly in regularly distributed clusters, 50800 lm in
diameter. In seven cortical areas, there was a more complex pattern in which the
clusters of inclusions were aggregated into larger superclusters. In 11 cortical areas, the
size of the clusters approximated to those of the cells of origin of the cortico-cortical
pathways but in the majority of the remaining areas, cluster size was smaller than
400 lm. The topography of the lesions suggests that there is degeneration of the
cortico-cortical projections in NIFID with the formation of a-internexin-positive
aggregates within vertical columns of cells. Initially, only a subset of cells within a
vertical column develops inclusions but as the disease progresses, the whole of the
column becomes aected. The corticostriate projection appears to have little eect on
the cortical topography of the inclusions.

Introduction
Four proteins comprise the type IV neuronal intermediate lament (IF) proteins, viz., a-internexin and the
neurolament (NF) triplet proteins composed of light
(NF-L), medium (NF-M), and heavy (NF-H) subunits.
a-Internexin is the rst IF protein expressed in postmitotic neurons of the developing nervous system [1,2]
but in the adult brain, expression of a-internexin is
conned to mature neurons. NF proteins play distinct
roles in lament assembly and in the radial growth of
axons [3] and recent studies suggest a signicant role for
these proteins in nervous system disease. For example,
mutations of the NF-L gene have been linked to subtypes of Charcot-Marie-Tooth disease and of the NF-H
gene are associated with some cases of sporadic
amylotrophic lateral sclerosis (ALS) [4]. In addition,
neuronal death occurs in transgenic mice that express
NF and a-internexin [3] whilst in focal cortical dysplasia a disorder characterized by disorganized cortical
cytoarchitecture there is increased expression of IF
proteins including a-internexin in ballooned neurons [5].
Abnormal neuronal IF aggregates immunopositive
for a-internexin have recently been identied as
Correspondence: Richard A. Armstrong, Vision Sciences, Aston
University, Birmingham, B4 7ET, UK (tel.: 0121 359 3611;
fax: 0121 333 4220; e-mail: r.a.armstrong@aston.ac.uk).

528

the pathological hallmark of neuronal intermediate

lament inclusion disease (NIFID) [6], a disease of early
onset with a variable clinical phenotype including
frontotemporal lobar degeneration, pyramidal, and
extrapyramidal signs [79]. Neuropathologically, there
is degeneration of the cerebral cortex, striatum, and
brain stem with neuronal loss in the frontal, parietal,
and temporal cortices, and in the hippocampus [7,8,10].
Abnormal lamentous inclusions are commonly found
in neurodegenerative disease [11] but the majority are
characterized by abnormal cellular aggregates of tau
(tauopathies) or a-synuclein (a-synucleinopathies).
NIFID extends the molecular classication of neurodegenerative disorders and implicates novel mechanisms of pathogenesis [6].
In the cerebral cortex of the tauopathies and a-synucleinopathies, the pathological inclusions often have
a distinct topography, viz., they occur in distinct
clusters which exhibit a regular periodicity parallel to
the pia mater [12]. This has led to the hypothesis that
lamentous inclusions may develop in relation to clusters of cells associated with specic cortico-cortical and
cortico-hippocampal projections and result from the
degeneration of these cells. In addition, there is extensive neuronal loss in the striatum in NIFID that may
inuence the topography of the inclusions in the cortex
[13]. A previous study [14] reported data on the
topography of NF-H-labeled neuronal inclusions in

2006 EFNS

a-internexin-positive neuronal aggregates

a restricted number of brain areas in four cases of

NIFID. The objective of this study was to determine
whether the inclusions characterized by a-internexin
immunoreactivity in a larger group of NIFID cases
exhibited a similar pattern of topographic distribution
to those of the tauopathies and a-synucleinopathies and
whether the topography is more likely to be attributable
to cortico-cortical than corticostriate degeneration.

Materials and methods

Cases

Ten cases of NIFID (see Table 1) obtained from Canada, Norway, Spain, Japan (one case from each), and
from France, the UK and the USA (two cases from
each) were studied [13]. Patients had no family history
of psychiatric or neurological disorders. All patients
displayed the histological hallmarks of NIFID and are
described in detail by Cairns et al. [13]. Three of the
cases (A, F, I) have been studied previously [14].
Histological methods

After death, the consent of the next of kin was obtained

for brain removal, following local Ethical Committee
procedure and the 1995 Declaration of Helsinki (as
modied Edinburgh 2000). Brain tissue was preserved
in buered 10% formalin or 4% paraformaldehyde.
Tissue blocks were taken from the medial frontal cortex
at the level of the genu of the corpus callosum, motor
cortex, and temporal lobe at the level of the lateral
geniculate body. Tissue was xed in 10% phosphatebuered formal saline and embedded in paran wax.
Immunohistochemistry (IHC) was performed on 6- to
8-lm sections with an antibody that recognizes ainternexin (Zymed Laboratories Inc., San Francisco,
CA, USA). Sections were counterstained with hematoxylin.
Table 1 Demographic features and gross brain weight of the neuronal
intermediate lament inclusion disease (NIFID) cases studied

Case

Sex

Age at onset
(years)

Age at death
(years)

Duration
(years)

Brain
weight (g)

A
B
C
D
E
F
G
H
I
J

F
F
M
F
M
M
M
M
F
M

23
25
32
38
39
47
48
48
52
56

28
29
35
41
42.5
50
52
61
54.7
60

5
4
3
3
3.5
3
4
13
2.7
4

860
710
NA
904
950
1200
1310
850
813
1250

NA, data not available.

2006 EFNS European Journal of Neurology 13, 528532

529

Morphometric methods

In areas of the cerebral cortex, a-internexin-positive

inclusions were counted along a strip of tissue (3200
6400 lm in length) parallel to the pia mater, using
250 50 lm sample elds arranged contiguously. The
sample elds were located both in the upper (laminae
II/III) and lower (laminae V/VI) cortical laminae, the
short edge of the sample eld orientated parallel with
the pia mater and aligned with guidelines marked on the
slide. In the hippocampus, the inclusions were counted
from area CA1 to CA4, the short dimension of the
contiguous sample eld being aligned with the alveus.
Inclusions were also observed in the dentate gyrus
granule cells [10] and the sample eld was aligned with
the upper edge of the granule cell layer.
Data analysis

To determine the topographic patterns of the inclusions, the data were analysed by spatial pattern analysis
[15,16]. This method uses the variance-mean ratio
(V/M) of the data to determine whether the inclusions
were distributed randomly (V/M 1), regularly
(V/M < 1), or were clustered (V/M > 1) along a strip
of tissue. Counts of inclusions in adjacent sample elds
were added together successively to provide data for
increasing eld sizes, e.g. 50 250, 100 250,
200 250 lm, etc., up to a size limited by the length of
the strip sampled. V/M was then plotted against eld
size to determine whether the clusters of inclusions were
regularly or randomly distributed and to estimate the
mean cluster size parallel to the tissue boundary. A
V/M peak indicates the presence of regularly spaced
clusters whilst an increase in V/M to an asymptotic
level suggests the presence of randomly distributed
clusters. The statistical signicance of a peak was tested
using a t-test [16]. The spatial patterns of the histological features in upper and lower laminae of the
cerebral cortex and in the cortex and hippocampus
were compared using chi-square (v2) contingency table
tests.

Results
Typical examples of the topographic patterns shown by
the a-internexin aggregates are shown in Fig. 1. In case
D (inferior temporal gyrus), a V/M peak at eld size
50 lm indicates the presence of regularly distributed
clusters of lesions of average diameter 50 lm. A V/M
peak is also present at eld size 200 lm suggesting that
the inclusions are aggregated into larger superclusters
at least 200 lm in diameter. In case C (motor cortex),
there was a single V/M peak at eld 800 lm suggesting

530

R. A. Armstrong and N. J. Cairns

6.0
5.5
5.0

Variance/mean

4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5

200

400

600

800

1000

1200

1400

1600

1800

Field size ( m)

Figure 1 Examples of the topographic pattern exhibited by ainternexin-positive neuronal cytoplasmic inclusions in neuronal
intermediate lament inclusion disease (NIFID) (closed symbols:
case D, inferior temporal gyrus; open symbols: case C, motor
cortex).

the presence of clusters of inclusions 800 lm in diameter distributed parallel to the pia mater.
The topographic patterns exhibited by the inclusions
in each brain region of each case are shown in Table 2.
In the cerebral cortex, 39 of 54 (72%) gyri showed a
regular distribution of clusters of lesions, a random
distribution was present in four of 54 (7%) gyri, a
regular distribution in one of 54 (2%) gyri and large
non-regular clusters (>800 lm) in two of 54 (4%) gyri.
In the CA sectors of the hippocampus and in the dentate gyrus, regularly distributed clusters of inclusions

were present in seven of 14 (50%) areas studied, a

random distribution in two of 14 (14%) areas, a regular
distribution in one of 14 (7%) areas, and large clusters
were present in four of 14 (29%) areas. In seven areas,
there was evidence of clustering at two scales, i.e.
smaller sized clusters of inclusions were aggregated into
larger clusters. In laminae II/III of the cortical areas
exhibiting clustering, clusters were in the size range
400800 lm in six of 23 (26%) gyri, smaller than
400 lm in 15 of 23 (65%) and larger then 800 lm in
two of 23 (9%) gyri. In laminae V/VI, cluster sizes of
inclusions were between 400 and 800 lm in ve of 19
(26%) gyri, smaller than 400 lm in 12 of 19 (63%) gyri,
and larger than 400 lm in two of 19 (11%) gyri. The
distribution of topographic patterns of the inclusions
was similar in laminae II/III compared with V/VI
(v2 1.54, P > 0.05) and in the cerebral cortex compared with the hippocampus and dentate gyrus (v2
5.38, P > 0.05). A comparison of the data from the
seven new cases with cases A, F, and I studied previously suggests similar topographic patterns are present
(v2 1.85, P > 0.05).

Discussion
The data suggest that the a-internexin-positive neuronal
inclusions occur in clusters with, in many areas, the
clusters exhibiting a regular periodicity parallel to the
tissue boundary. This type of topography is similar to
that reported for lamentous inclusions characterized
by tau and a-synuclein reactivity [17 20] and for
NF-H-labeled inclusions in the three original cases of

Table 2 Topographic pattern of the a-internexin-positive neuronal cytoplasmic inclusions in areas of the frontal and temporal lobe in 10 cases of
neuronal intermediate lament inclusion disease (NIFID)
Brain region and lamina
FC

MC

ITG

PHG

Case

II/III

V/VI

II/III

V/VI

II/III

V/VI

II/III

V/VI

HC

DG

A
B
C
D
E
F
G
H
I
J

200

400

50
R, >800
400
R
R

50, 400

Rg
50, 400
100
100
50

50, 800
50
>800
100
400

800
R
50
>800
50
50, 400

200
100

Rg
400
100, 400, 400
100
100

50

200
100

50
100
100
Rg
R

50

50

100
50
R
200
50

50, 200

>800
R
100, 400
200
50

50

50

400
>800
>800
R
>800

50

50

100
Rg
50
200
>800

FC, frontal cortex; MC, motor cortex; ITG, inferior temporal gyrus; PHG, parahippocampal gyrus; HC, hippocampus; DG, dentate gyrus.
Data represent the dimensions ( lm) measured parallel to the pia mater, alveus or dentate gyrus granule cell layer of regularly distributed clusters of
IF aggregates with the following exceptions: R random distribution, Rg regular distribution of individual aggregates. Data preceded by >
indicate large-scale clustering without regular spacing. Where two gures are present, clustering occurs at two scales, i.e. smaller clusters are
aggregated into larger superclusters. Indicates either tissue section unavailable or density of inclusions was too low to determine topography.

2006 EFNS European Journal of Neurology 13, 528532

a-internexin-positive neuronal aggregates

NIFID [14]. Hence, the seven new cases of NIFID that

originate from several centers reveal essentially the
same types of topographic pattern as reported previously [14]. The data also suggest that a-internexin and
NF-H IHC reveal the same topographic patterns and
are therefore likely to be staining the same populations
of neuronal inclusions.
The topography of the clusters of aggregates within
the cerebral cortex and hippocampus suggests that the
inclusions could be related to the cells of origin of
specic cortico-cortical and cortico-hippocampal projections [2123]. The cells of origin of these projections
are clustered and occur in bands that are more or less
regularly distributed along the cortical strip. Individual
bands of cells, approximately 500800 lm in width,
traverse the cortical laminae in columns [21]. In
approximately 70% of cortical regions studied, the
clusters of aggregates were regularly distributed parallel
to the pia mater consistent with their development in
relation to these cell clusters. In eleven cortical areas,
the estimated width of the aggregate clusters approximated to the size of the cells of origin of the corticocortical projections. In many of the remaining brain
regions, the aggregates developed in much smaller
clusters, usually between 50 and 100 lm in diameter.
Hence, the inclusions may aect only a specic subset
of cells within a column characterized by a particular
neurotransmitter or antigenic determinant. In seven
cortical areas, however, the smaller regularly distributed clusters of inclusions appeared to be aggregated
into larger superclusters and in a further six areas,
clusters larger than 800 lm in diameter were present.
This observation suggests that as the disease progresses
the smaller clusters of inclusions develop into larger
clusters [17] and more of the cell column becomes
aected.
There is extensive neuronal loss in the striatum in
NIFID [13] and this may also contribute to the cortical topography of the inclusions. Virtually all the
major cortical areas project to the caudate nucleus and
putamen. These projections are topographically
organized, e.g. the frontal cortex projects to the
anterior part of the head of the caudate nucleus and
the precommissural part of the putamen [24]. Projections to the striatum have their origin in lamina V and
are grouped in columns distinct from those of other
subcortical projections [25,26]. Hence, degeneration of
this pathway in NIFID could inuence the topography
of the inclusions especially in the lower laminae.
However, there was no signicant dierence in the
topography of the inclusions in the upper and lower
laminae suggesting that the patterns in both regions
are more likely to reect degeneration of the corticocortical pathways.

2006 EFNS European Journal of Neurology 13, 528532

531

The a-internexin neuronal aggregates may be the

cause or the result of neural degeneration in NIFID
[27]. A number of studies have linked neuronal degeneration to the accumulation of neuronal IF-positive
aggregates. Transgenic mice overexpressing a-internexin, for example, develop cerebellar torpedoes with
high levels of neuronal IFs in large pyramidal cells of
the cerebral cortex and areas of the thalamus [2].
Furthermore, transgenic mice expressing altered NF
proteins develop motor neuron degeneration and
NF-positive aggregate formation similar to cases of
sporadic ALS [28]. Alternatively, abnormal protein
aggregates may accumulate in cell bodies in vulnerable
systems as a consequence of synaptic disconnection or
the breakdown of neural pathways [12]. Deciency of
NF-L in transgenic mice results in IF-positive inclusions
and death of motor neurons [29]. However, the protein
aggregates do not provoke the death of neurons unlike
the toxic neuronal IF aggregates induced by peripherin
overexpression [29]. In addition, a-internexin activity
increases in crushed, transected, and resected facial
nerves after 7 days suggesting that it is a reaction to
axonal damage and suggests a role for a-internexin in
neuronal regeneration [30]. These results suggest that
the accumulation of a-internexin-positive aggregates in
NIFID could be a response to the neurodegeneration
aecting the cortico-cortical pathways rather than
being its cause.

Acknowledgements
We thank Drs R.H. Perry, C. Duyckaerts, F. CruzSanchez, K. Skullerud, E. Bigio, M. Gearing, I.R.A.
Mackenzie, and H. Yokoo for providing tissue sections
for this study, the sta of the Center for Neurodegenerative Disease Research of the University of Pennsylvania School of Medicine for technical support.
Support for this work was provided by a grant from the
Welcome Trust (GR066166A1A) to N.J.C.

References
1. Ching GY, Liein RKH. Roles of head and tail domains in
alpha-internexins self-assembly and coassembly with the
neurolament triplet proteins. Journal of Cell Science
1998; 111: 321333.
2. Ching GY, Chien CL, Flores R, Liein RKH. Overexpression of alpha-internexin causes abnormal neurolamentous accumulations and motor coordination
decits in transgenic mice. Journal of Neuroscience 1999;
19: 29742986.
3. Julien JP. Neurolament functions in health and disease.
Current Opinion in Neurobiology 1999; 9: 554560.
4. Lariviere RC, Julien JP. Functions of intermediate laments in neuronal development and disease. Journal of
Neurobiology 2004; 58: 131148.

532

R. A. Armstrong and N. J. Cairns

5. Taylor JP, Slater R, French J, Balluch G, Crino PB.

Transcription of intermediate lament genes is enhanced
in focal cortical dysplasia. Acta Neuropathologica 2001;
102: 141148.
6. Cairns NJ, Zhukareva V, Uryu K, et al. a-Internexin is
present in the pathological inclusions of neuronal intermediate lament inclusion disease. American Journal of
Pathology 2004; 164: 21532161.
7. Bigio EH, Lipton AM, White CL, Dickson DW, Hirano
A. Frontotemporal dementia and motor neurone degeneration with neurolament inclusion bodies: additional
evidence for overlap between FTD and ALS. Neuropathology and Applied Neurobiology 2003; 29: 239253.
8. Cairns NJ, Perry RH, Jaros E, et al. Patients with a novel
neurolamentopathy: dementia with neurolament inclusions. Neuroscience Letters 2003; 341: 177180.
9. Josephs KA, Holton JL, Rossor MN, et al. Neurolament inclusion body disease: a new proteinopathy? Brain
2003; 126: 22912303.
10. Cairns NJ, Jaros E, Perry RH, Armstrong RA. Temporal
lobe pathology of human patients with neurolament
inclusion disease. Neuroscience Letters 2004; 354: 245247.
11. Goedert M, Spillantini MG, Serpell LC, et al. From
genetics to pathology: tau and alpha-synuclein assemblies
in neurodegenerative diseases. Philosophical transactions
of the Royal Society of London. Series B, Biological
sciences 2001; 356: 213227.
12. Armstrong RA, Cairns NJ, Lantos PL. What does the
study of spatial patterns tell us about the pathogenesis of
neurodegenerative disorders? Neuropathology 2001; 21:
112.
13. Cairns NJ, Grossman M, Arnold SE, et al. Clinical and
neuropathologic variation in neuronal intermediate lament inclusion disease (NIFID). Neurology 2004; 63:
13761384.
14. Cairns NJ, Armstrong RA. Clustering of neuronal inclusions in dementia with neurolament inclusions. Acta
Neuropathologica 2003; 106: 125128.
15. Armstrong RA. The usefulness of spatial pattern analysis
in understanding the pathogenesis of neurodegenerative
disorders, with particular reference to plaque formation in
Alzheimers disease. Neurodegeneration 1993; 2: 7380.
16. Armstrong RA. Analysis of spatial patterns in histological
sections of brain tissue. Journal of Neuroscience Methods
1997; 73: 141147.
17. Armstrong RA. Is the clustering of neurobrillary tangles
in Alzheimers patients related to the cells of origin of
specic cortico-cortical projections? Neuroscience Letters
1993; 160: 5760.

18. Armstrong RA, Cairns NJ, Lantos PL. Dementia with

Lewy bodies: clustering of Lewy bodies in human
patients. Neuroscience Letters 1997; 224: 4144.
19. Armstrong RA, Cairns NJ, Lantos PL. Clustering of Pick
bodies in Picks disease. Neuroscience Letters 1998; 242:
8184.
20. Armstrong RA, Cairns NJ, Lantos PL. Clustering of
cerebral cortical lesions in patients with corticobasal
degeneration. Neuroscience Letters 1999; 268: 58.
21. Hiorns RW, Neal JW, Pearson RCA, Powell TPS. Clustering of ipsilateral cortico-cortical projection neurons to
area 7 in the rhesus monkey. Proceedings of the Royal
Society of London 1991; 246: 19.
22. De Lacoste M, White CL, III. The role of cortical connectivity in Alzheimers disease pathogenesis: a review and
model system. Neurobiology of Aging 1993; 14: 116.
23. Delatour B, Blanchard V, Pradier L, Duyckaerts C.
Alzheimer pathology disorganizes cortico-cortical circuitry: direct evidence from a transgenic animal model.
Neurobiology of Diseases 2004; 16: 4147.
24. Brodal A. Neurological Anatomy in Relation to Clinical
Medicine. Oxford University Press, New York and
Oxford, 1981.
25. Jones EG, Coulter JD, Burton H, Porter R. Cells of origin
and terminal distribution of corticostriatal bers arising in
the sensory-motor cortex of monkeys. Journal of Comparative Neurology 1977; 181: 291348.
26. Brown LL, Smith DM, Goldblum LM. Organising principles of cortical integration in the rat neostriatum: corticostriate map of the body surface is an ordered lattice of
curved laminae and radial points. Journal of Comparative
Neurology 1998; 392: 468488.
27. Armstrong RA, Cairns NJ, Lantos PL. Are pathological
lesions in neurodegenerative disorders the cause or the
effect of the degeneration? Neuropathology 2002; 22: 114
127.
28. Wong NKY, He BP, Strong MJ. Characterisation of
neuronal intermediate lament protein expression in
cervical spinal motor neurons in sporadic amylotrophic
lateral sclerosis (ALS). Journal of Neuropathology and
Experimental Neurology 2000; 59: 972982.
29. Beaulieu JM, Jacomy H, Julien JP. Formation of intermediate lament protein aggregates with disparate effects
in two transgenic mouse models lacking the NF light unit.
Journal of Neuroscience 2000; 20: 53215328.
30. McGraw TS, Mickle JP, Shaw G, Streit WJ. Axonally
transported peripheral signals regulate alpha-internexin
expression in regenerating motoneurons. Journal of
Neuroscience 2002; 22: 49554963.

2006 EFNS European Journal of Neurology 13, 528532