explormg the then current dilemmas and difficulties m PrP immunostammg A
five-center study was set up in which duplicate slides were circulated for PrP
lmmunostainmg to all the centers, each center using Its own protocol and anti-
bodies. Subsequently, all the stained slides have been circulated for assess-
ment of stammg results to each of the five centers m turn and a consensus
report 1sbeing drawn up. It has emerged that despite some slight varlatlon m
the intensity of staining, similar patterns of PrP deposition were noted m each
case regardless of the antibody used or of slight variations m protocol (each
center used at least one of the pretreatment steps described, usually exposure to
formic acid). Control cases,such as age matched, nondemented subjects, and
casesof Alzheimer disease, did not show PrP stammg
Pretreatments that have been advocated m the literature include those listed
in Table 1. The protocol that 1scurrently m use m our laboratory is shown m
Table 2. In summary, the pretreatments that we use before exposure to the
primary antibody include hydrated autoclaving, formic acid for 5 mm, and
guamdine thlocyanate for 2 h (Table 2). Although in our experience guamdine
thiocyanate has proved useful, the results of the five-center workshop study
suggest that this pretreatment could be excluded with no significant detriment
to PrPCJDvisuahzatlon. In contrast, the formic acid and hydrated autoclavmg
pretreatments are extremely useful, leading to consistent and unequivocal pat-
terns of PrPCJDmununolocallzatlon. In particular, sections that are slightly
positive with other pretreatments always show more widespread and specific
stammg followmg the autoclave pretreatment. We have experimented with
other pretreatments (Table l), including protease K and pepsin (but have found
that they are extremely detrimental to tissue preservation); hydrolytic auto-
claving (m 2.5 mA4HCl as compared with distilled water; this appears to con-
fer no advantage over hydrated autoclaving); and microwave enhancement
(1 mm in formic acid). Our experience with over 150 cases of prion disease
suggeststhat the current immunocytochemical protocol and pretreatments pro-
vide the most reliable results. In most cases a panel of three or four anti-PrP
antibodies 1sused for diagnostic confirmation (Table 3) (2,13).
4. PrP lmmunostaining in Sporadic CJD Cases
Given the range of pathological changes seen with routine staining methods
m casesof sporadic CJD it 1snot, perhaps, surprising that PrPCJDimmunolocali-
zation also varies from case to case. The patterns of PrPCJDdeposltlon include:
confluent sponglform change m the cortical
1. Cortex In those cases with severe
gray matter, irregular diffuse deposits of PrPCJD are seen m relation to the
sponglform vacuoles (Figs. 1 and 2). However, when the sponglform change 1s
represented only by small discrete parenchymal vacuoles, PrPCJD stammg may
PrP Antibodies Currently in Use in Edinburgh
Clonality Antigen Source
lA8 Polyclonal Scrapie fibrils 1 in 5000 o/n0 J. Hope,
KG9 Monoclonal Recombinant bovine PrP 1 in 200 o/n C. Birkett,
3F4 Monoclonal Synthetic peptide (residues 1 in 2000 o/n R. Kascsak,
108-l 11 of human PrPC) New York
SP40 1 in 1500 o/n B. Anderton,
Polyclonal Synthetic peptide (partial
sequence sheepPrP) London
â€œo/n = overnight.
Fig. 1. Immunolocalization of PrPCJD in the cortical gray matter from a patient with
sporadic CJD. Darkly stained PrP CJDdeposits are seen in association with confluent
spongiform change through the full thickness of the cortical ribbon from cortical sur-
face above to the white matter below.
be apparently absent in gray matter unless specifically enhanced with autoclav-
ing pretreatment, which reveals diffuse granular staining of the gray matter in
such cases (13). In some cases, distinct perineuronal staining is seen particularly
in layers 3 and 5 of the cortex (Fig. 3). Astrocytes may display granular cytoplas-
mic staining either as an isolated feature or in company with vacuolar deposits of
PrPCJD (Fig. 4). Plaques are not usually visible in the cortex, although PA4 plaques
Diagnosis of Human Prion Disease 65
Fig. 2. PrPCJD immunolocalization in association with severe spongiform change in
the cortical gray matter from a case of sporadic CJD.
Fig. 3. Laminar neuronal staining of membranes and processes with antibody KG9.
Microcystic spongiform change is seen in the adjacent unstained cortex.
may be present in small number as an aging phenomenon or, rarely, as a sign of
concurrent presence of Alzheimer disease. The comu ammonis of the temporal
cortex is frequently spared both in terms of spongiform change and of PrPCJD.
2. Basal ganglia: PrPCJD immunostaining is sometimes visible in the basal ganglia,
particularly in the form of small plaques in the white matter when the gray matter
shows spongiform change (Fig. 5). Fine granular staining may be visible in the
areas of gray matter.
Fig. 4. Focal deposits of PrPCJD associated with spongifotm vacuoles in the cortex
of a case of sporadic CJD. Neighboring neurons (arrow) are unstained. Occasional
astrocytes show granular cytoplasmic staining for PrPCJD (closed arrow).
3. Cerebellum: In a minority of cases of sporadic CJD small PrPcJD positive plaques
may be present in any of the layers of the cortex and in the white matter but are
particularly prominent in the Purkinje cell layer and in the granular layer (Fig. 6).
The plaques are circular in profile and may have a pale center, a dark circumfer-
ential layer, and radiating spicules from the outer surfacethe so-called kuru
plaques (26) (Fig. 7). The majority of cases do not have well-defined plaques and
show more diffuse positivity in the granular layer and focally in the molecular
layer (Fig. 8). The white matter is unstained in such cases. The presence of
plaques in the cerebellum bears no obvious relationship to the presence of
spongiform change in the molecular layer but does correspond with the presence
of cerebellar atrophy. The Purkinje cells are always PrPcJD negative (2,13) (Fig. 8).
Perineuronal staining may be visible in the dentate nucleus.
4. Brain stem: A minority of cases show scattered small plaques in the brainstem,
particularly in the peripheral white matter tracts (Fig. 9). More typically, in one-
third of cases of sporadic CJD granular positivity is seen in the gray matter of the
periaqueductal region, tectum, substantia nigra, and pontine nuclei. Staining asso-
ciated with neurons may be very obvious in some of these nuclei, particularly in
the olive (Figs. 9 and 10). The corticospinal tracts are always negative.
5. Spinal cord: In these cases that show positivity in the brainstem, PrPCJD
immunopositivity is seen in the substantia gelatinosa and dorsal horns of the spi-
nal cord gray matter (27) (Fig. 11). These cases also show small plaques within
the peripheral white matter tracts and linear PrP CJDdeposits extending into the
white matter (Fig. 12).
Diagnosis of Human Prion Disease
Fig. 5. Section of basal ganglia from a case of sporadic CJD showing spongiform
change in areas of gray matter at the top and bottom of the figure. PrPCJD-positive
plaques are present in the intervening white matter as well as in the gray matter. Most
of the spongiform gray matter is apparently unstained for PrPCJD.
It is interesting to note that, whereas in some instances, PrPCJD accumulates
in those areas that show the most profound pathological change (such as the
cerebral cortex in the presence of confluent spongiform change as shown in
Figs. 1 and 2) or that are associated with clinical signs and symptoms (cerebel-
lar ataxia with cerebellar atrophy, displaying plaques in the granular layer as
shown in Fig. 6), there are also considerable discrepancies between the local-
ization of classical pathology (as revealed by routine staining) and PrPCJD
immunolocalization (2). Thus, quite strong positivity for PrPCJDmay be evi-
dent in parts of the CNS, such as spinal cord and brain stem (Figs. 9-12),
which rarely show recognizable cellular changes or spongiform degeneration
in human cases,and conversely, the cerebellar molecular layer may show con-
spicuous microcystic spongiform change without colocalizing evidence of
PrPCJD(13). Similarly, we have not thus far been successful in demonstrating
PrPCJDby immunocytochemical techniques in dura mater or in the pituitary
gland, both of which have been implicated in the transmission of prion dis-
easesfrom one human case to another. Further work needs to be undertaken in
this field since there are several possible explanations for this finding. Our
initial studies with spinal cord tissue proved negative immunocytochemically,
but with improving techniques and a wider range of material, we have now
Fig. 6. Section of cerebellum from a case of sporadic CJD displaying PrPCJD-posi-
tive plaques. The plaques are most numerous in the granular layer but also show focal
concentrations in the Purkinje cell layer at the interface between the largely unstained
molecular layer and the granular layer.
Fig. 7. High power view of a plaque in the molecular layer of the cerebellar cortex
in a patient with sporadic CJD. The larger plaque has a pale center, a dark edge, and
radiating outer tibrils, displaying the classic kuru plaque appearance.
Diagnosis of Human Prion Disease
Fig. 8. Cerebellar cortex in a case of sporadic CJD without plaques. There is diffuse
granular staining, which is focally accentuated, in the granular and molecular layers.
The Purkinje cell layer is unstained.
Fig. 9. Section of the medulla in sporadic CJD showing immunolocalization of PrPcJD
in the olivary nucleus and in the subpial white matter where small plaques are present.
described immunopositivity for PrPCJDin the spinal cord in 30% of cases of
In many respects, the PrPCJDimmunostaining patterns in human cases are
similar to those described in mice with transmitted scrapie agent (18) although
the localization of PrPCJDis slightly different in the two species.
Fig. 10. High power view of the olivary nucleus shown in Fig. 9. Perineuronal and
diffuse granular staining is seen in the gray matter.
5. PrPCJD Staining Patterns in Other Prion Dementias
In other prion dementias, the patterns of PrPCJD immunopositivity some-
times resemble, and are sometimes quite widely divergent from, those seen in
the CNS of patients with sporadic CJD as described. Although these differing
patterns sometimes segregate with the particular genotype of the individual or
with the mode of acquisition of the transmissible agent, there is also a diversity
of phenotype within individual genetic abnormalities as well as among cases
of sporadic disease (2,27-30).
5.1. Inherited Prion Diseases Including Familial CJD and GSS
These cases have a different clinical presentation from sporadic CJD and
may display a longer duration of illness with a predominantly cerebellar symp-
tomatic presentation. Quite a number of different PrP gene mutations and
inserts have now been described in these families and to a certain extent, the
findings are linked to particular genetic abnormalities. As might be expected
from the clinical presentation, the major pathological changes are found in the
cerebellum, which shows predominantly plaques in familial disorders. In cases
of GSS, large multicentric plaques, more complex than the compact kuru-type
plaque, are seen scattered in the molecular layer and in the granular layer of the
cerebellum. These plaques are readily seen with routine stains (Fig. 13) and
those that bind to amyloid, but are very strongly positive for PrPCJD (Fig. 14).
The cerebellum in GSS has a heavier burden of plaques (Fig. 14) than in those
cases of CJD (Fig. 6) that also display cerebellar plaques. The plaques may
have associated microglial and neuritic abnormalities and in this way resemble
those seen in Alzheimer disease (31). Antibodies for PA4 amyloid may
colocalize in these large complex plaques (32). Spongiform change may be
Diagnosis of Human Prion Disease 71
Fig. 11. Computer image of PrPCJD immunolocalization in the spinal cord of a
growth hormone recipient who developed CJD. The strongest positivity is noted in the
substantia gelatinosa and dorsal horns.
Fig. 12. Section of thoracic spinal cord showing the junctional zone between gray
and white matter. Immunopositivity for PrPcJD extends along cell processes from gray
matter into the white matter.
absent from the cerebellar molecular layer and may be very focal in the cere-
brum, which generally shows less conspicuous changes than in cases of spo-
radic CJD. In one particular GSS kindred, very numerous neurofibrillary
tangles have been described (29) and when combined with the presence of
amyloid plaques, represent a possible source of diagnostic confusion and may
be mistakenly interpreted as the changes of familial Alzheimer disease.
Further study of these interesting families presents a useful opportunity for
linking particular neurodegenerative manifestations with identified genetic
Fig. 13. Routinely stained section of the cerebellar cortex in a case of GSS. Eosino-
philic plaques are conspicuous in the molecular later (above) and are also present
within the granular layer (below).
5.2. Acquired Prion Diseases
This virtually extinct condition, which was associatedwith cannibalistic ritu-
als in the natives of the Fore tribe in Papua New Guinea, presented initially
with cerebellar signs and symptoms and progressed to dementia only at a late