Filamentous white matter prion protein deposition is a distinctive feature of multiple inherited prion diseases

Research governance, post mortem examination and tissue preparation

Ethical approval for these studies was obtained from the Local Research Ethics Committee of UCL Institute of Neurology/National Hospital for Neurology and Neurosurgery. Informed consent to use the tissue for research was obtained. Clinical data, disease history, PRNP mutation and PRNP codon 129 status were available for all cases. Autopsies were carried out in a post mortem room designated for high risk autopsies. Brains were removed and stored in buffered formalin. The frontal, temporal, parietal and occipital cortex and the cerebellum were dissected during the post mortem procedure and immersed in 10% buffered formalin for up to one week, immersed into 98% formic acid for one hour and postfixed for 24 h in 10% buffered formalin. Tissues were then processed through graded alcohols and embedded in paraffin wax. In a small subset of cases, a small tissue block was prepared from the frontal cortex and white matter and immersed into glutaraldehyde for preparation of semithin resin section and electron microscopy. Frozen tissue of frontal brain and cerebellum was collected routinely and banked. The remaining brains were stored permanently in 10% formalin and additional brain regions were dissected and processed to paraffin blocks at later time points.

Antibodies and immunohistochemistry

In this study we used anti-PrP ICSM35 (D-Gen Ltd, London, UK, raised in Prnp^o/o mice against recombinant human PrP as previously described [12];) and anti PrP KG9 TSE Resource Centre, Roslin Institute University of Edinburgh [13]. ICSM35 recognises residues 93–102 and KG9 binds to residues 140–180 of human PrP. To detect abnormal PrP deposition, mounted sections were treated with 98% formic acid for 5 min, placed on an automated Ventana Discovery staining machine, heated to 95°C in 2.1 mM Tris–HCl, 1.3 mM EDTA, 1.1 mM sodium citrate, pH 7.8, for 30 min, digested for 16 min with a low concentration of protease (Protease 3, 0.02 U/ml alkaline protease, order no 760–2020 Ventana Medical Systems), incubated in Superblock for 10 min, then exposed to the primary antibody followed by biotinylated anti mouse IgG secondary antibody and visualised using the iView detection kit Haematoxylin was used as counterstain.

Immunofluorescence

Immunofluorescent dual labelling was carried out on selected, representative frontal lobe sections of cases with 4OPRI, 6OPRI, P102L, A117V, D178N, E200K mutations and sCJD 129MM, MV, and VV. For dual labelling and co-localisation of abnormal PrP in white matter structures we applied the same antigen retrieval conditions as above. Then we dispensed the antibody ICSM35 (1:1500) together with either anti-Neurofilament (NF200, SIGMA N5389; 1:100) or anti-Myelin Basic Protein (SMI94, Covance, 1:250) followed by the secondary antibodies labelled with fluorochromes Alexa 488 (goat anti Mouse IgG1) or Alexa 546 (goat anti mouse IgG2b). Images were taken on a ZEISS LSM510Meta Laser Scanning confocal microscope, using a Plan-Apochromat 63×/1.4 Oil objective. All images were taken at a scanning resolution that matched the optical resolution of the lens (pixel size corresponding to 99.6 nm on the section).

Transmission electron microscopy

Tissue blocks of no more than 2 mm³ were immersed in 2.5% glutaraldehyde, treated with formic acid for 24 h and following postfixation on glutaraldehyde, processed for resin (Epon) embedding. Resin sections were stained with Toluidine blue and suitable regions were selected for electron microscopy. Ultra-thin sections were stained with lead citrate and examined in a Jeol 100-CXII electron microscope. Images were recorded on a 4Megapixel SIS Megaview digital camera.

Western blotting and glycotyping

Brain tissue homogenates (10% w/v) from frozen samples of frontal cortex were prepared in Dulbecco’s phosphate buffered saline lacking Ca²⁺ or Mg²⁺ ions using tissue grinders or by serial passage through needles of decreasing diameter. Aliquots (typically 20 μl) were removed and proteinase K added from a 1 mg/ml stock solution (prepared in water) to give a final concentration in the sample of 50 μg/ml. Following incubation at 37°C for 60 min, samples were centrifuged at 16100 g for 1 min before termination of the digestion by the addition of an equal volume of 2 × SDS sample buffer (125 mM Tris–HCl, pH 6.8, 20% v/v glycerol, 4% w/v sodium dodecyl sulphate, 4% v/v 2-mercaptoethanol, 0.02% w/v bromophenol blue) containing 8 mM 4-(2-aminoethyl)-benzene sulfonyl fluoride and immediate transfer to a 100°C heating block for 10 min. Samples were centrifuged at 16100 g for 1 min prior to electrophoresis in 16% tris-glycine gels (Invitrogen). Gels were electroblotted onto PVDF membrane (Immobilon-P; Millipore) and subsequently blocked in PBS containing 0.05% v/v Tween-20 (PBST) and 5% non-fat milk powder for 60 min. After washing in PBST, the membranes were incubated with anti-PrP monoclonal antibody 3F4 (0.2 μg/ml in PBST) before washing in PBST (60 min) and incubation with alkaline-phosphatase-conjugated goat anti-mouse IgG secondary antibody (Sigma-Aldrich Prod. No. A2179) diluted 1:10, 000 in PBST. After washing (1 h with PBST and 2 × 5 min with 20 mM Tris pH 9.8 containing 1 mM MgCl₂) blots were incubated for 5 min in chemiluminescent substrate (CDP-Star; Tropix Inc) and visualized on Biomax MR film (Kodak). Control samples (of known PrP^Sc type) were analysed in parallel and run on the same blots to enable assignment of molecular strain type according to the London classification [14, 15].

Examination and quantification

We analysed 35 autopsy cases of inherited prion disease seen at the National Prion Clinic, NHNN and these were compared with 26 cases with sCJD (9 cases with codon 129 MM, 10 with MV and 7 with VV polymorphism). The inherited cases included the following PRNP gene mutations (numbers of cases in brackets): 4 octapeptide/96 base pair repeat insertion [4OPRI] (3), 5OPRI/120 bp (1), 6OPRI/144 bp (4), P102L (10), A117V (4), D178N (3), E200K (9) and V210I (1).

Brains were examined in the following areas: Frontal, temporal, parietal, occipital and cerebellar cortex. More regions were available for some of the brains, sometimes with very extensive sampling, but we restricted our analysis to the above-mentioned representative areas of subcortical white matter. Using a semi-quantitative scoring scale (Table 1), all features were recorded for the cortical areas and the cerebellum. Intensity and the deposition pattern of PrP were recorded for the cortical grey matter and the subcortical white matter, as well as the cerebellar molecular and granular layers and the cerebellar white matter. The following features were observed and recorded: in the cortical grey matter, there was synaptic labelling and deposition of plaques, in addition filamentous PrP was present in the white matter fibres extending into the cortex. In the white matter, we recorded small granules as well as thin threads of variable length, which depended on the orientation of the white matter tracts. The scores 0-3 reflect the abundance of PrP deposits (threads or granules), which is listed in Table 1. The quantity and the pattern of cortical prion pathology was recorded for all cases as shown in Table 2.

Score (granules or threads)	Number of inclusions (High power field (HPF) with a 40× objective)
0	Less than 1 inclusion per 4 HPF
1	More than 1 inclusion per 4 HPF, up to 5 inclusions per 1 HPF
2	5-20 inclusions per 1 HPF
3	More than 20 inclusions per 1 HPF

Scoring scheme used to semiquantitatively assess the frequency and density of granular or filamentous inclusions in DAB stained paraffin sections on a LEICA DM2500 with a 40× HCX Pan Apochromat objective.

Figure	Case no	Mutation	Codon 129 status	PrP Scmolecular strain type*	Gender	Age at death	Disease duration (mo)	Amount of PrP Scin WM		PrP pattern WM T = Threads D = Dots		Other pathology
								FB	Cer	FB	Cer	Cortical PrP pattern	Spongiosis
	1	V210L	MV	2	M	69	5	+	—	T, D	—	S1, PN1, D2	+
	2	E200K	MM	2	M	65	4	+	—	D	—	S2, GR1, PN1	+
	3	E200K	MV	ND	M	67	24	+	—	D	—	S1	+
	4	E200K	MM	ND	F	57	6	++	+	D	D	S1, PN1	+
Figure 2D-F, Figure 4A, B	5	E200K	MV	2/3	F	45	38	+++	+	T, D	D	S3, GR2	+++
	6	E200K	MM	ND	M	61	6	++	+	D	D	S2, GR2, PV2	+
	7	E200K	MM	2	M	65	5	—	—	—	—	S1, PN1	++
	8	E200K	MM	2	M	65	9	+	—	D		S1	+
	9	E200K	MM	2	F	69	4	+	+	D	—	S2, PN1	++
	10	E200K	MM	ND	M	47	19	++	++	T, D	T, D	S2, D1, PV2	+
	11	D178N	MM	Negative	M	60	7	+++	+	D	D	P3, D2	o
Figure 2A-C, Figure 3I, J	12	D178N	MV	ND	M	60	24	++	—	D	—	S2, GR1, P1	+++
	13	D178N	MV	ND	F	51	20	—	—	—	—	S1, GR1	++
	14	A117V	MV	Negative	M	43	6	+++	++	T	D	S2, P3, D3	o
	15	A117V	VV	Negative	M	40	46	+++	++	T	D	S2, P3	++
Figure 1M-P, Figure 3G, H	16	A117V	MV	Negative	M	47	47	+++	—	T	—	S1, P3	o
Figure 5C, D	17	A117V	MV	ND	F	40	98	+++	+	T	T, D	S2, P3	++
	18	P102L	MV	Negative	M	53	64	++	++	T	T, D	P3, D2	o
	19	P102L	MM	1	M	58	53	+	+	T	D	P3, D2	++
	20	P102L	MM	ND	M	64	24	—	+	—	T, D	S1, P1	o
Figure 1I-L	21	P102L	MM	ND	M	59	49	+++	+++	T, D	T, D	P3, D2	++
	22	P102L	MV	2	F	64	33	+++	++	D	D	S2, P3, D2	++
	23	P102L	MV	ND	M	61	50	+++	+++	T, D	T, D	S2, P3, D1	+++
	24	P102L	MM	2	F	65	64	+	++	T, D	T, D	S1, P3, D2	++
Figure 3E, F	25	P102L	MM	Negative	F	62	48	++	+	D	D	S2, P3, D1, PV1	+
	26	P102L	MM	Negative;8 kDa	M	52	155	+	+	D	D	P2	o
	27	P102L	MM	ND	F	66	81	+	+++	D	T, D	S1, P3	++
	28	6OPRI	MV	Negative	F	49	8	+++	+	T	D	S1	o
Figure 1E-H	29	6OPRI	MV	Negative	F	46	0	+++	+++	T	T, D	S2, P3, D3	o
	30	6OPRI	MM	ND	M	44	118	+++	+++	T, D	D	S3, GR1, P1, D1	+
Figure 3C, D, Figure 5A, B	31	6OPRI	MM	ND	F	45	72	+++	++	T	T, D	S1, PN1	+
	32	5OPRI	MM	ND	M	57	323	+	+	T, D	T, D	S1	+
	33	4OPRI	MM	ND	F	56	7	+++	+	T	T, D	S3, GR1	+
Figure 1A-D, Figure 3A, B	34	4OPRI	MM	ND	M	73	14	++	++	T, D	T, D	S1	+++
	35	4OPRI	MM	2	F	62	5	++	++	T	T, D	S1, GR1, PN1	++

*Proteinase K-digested PrP fragment size according to the London classification [14, 15].

ND, not done.

All case numbers (first column) are cross-referenced to Figures 1, 2, 3, 4, 5 and 6. The columns on the right describe the pattern of PrP deposition in the white matter. The semi quantitative density scores of white matter PrP are represented in column 8. mo: months, M: male/methionine, F: female, V: valine, WM: white matter, FB = Forebrain, Cer = Cerebellum, T = Threads, D = Dots. The two columns on the right describe the pattern and intensity of cortical PrP deposition as described in [11] and the degree of spongiform changes. S = synaptic, PN = perineuronal net, GR = Granular; P = well demarcated plaques, D = diffuse plaques; and PV = perivascular deposits. The number indicates the semi-quantitative intensity or density of the respective feature.

Statistical analysis

4OPRI [96 base pair insert] (1 male, 2 female); Figure 1A-D, Figure 3A, B

The mean age for this group is 63.7 years (56–73) (Cases #33-35). The cerebral grey matter of these cases shows both axonal threads and granular synaptic stain (Figure 1A). The white matter also shows threads and dots of abnormal PrP (Figure 1B,C). The cerebellar cortex displays the classical “striped” or “tigroid” pattern of PrP depositions, perpendicular to the cerebellar surface and the Purkinje cell layer as reported previously [11, 16] and the white matter comprises rare to moderately frequent dots and thread-like staining (Figure 1D). Filamentous deposits were seen with both antibodies (Figure 1B, ICSM35 and 1C, KG9). All three 4OPRI cases show consistent axonal labelling in cerebral grey and white matter as well as cerebellar white matter, in a pattern and intensity similar to that observed in the 6OPRI group [11, 17]. Co-localisation studies (Figure 3A,B) with PrP (red) and neurofilaments (green) show occasional parallel course of PrP threads and axons (Figure 3A,A’), while both longitudinal and transverse sections of myelin sheaths confirm the presence of the threads within the myelin structure Figure 3B,B).

5 OPRI [120 bp insert] (1 male)

The age of this patient was 57 years, with disease duration of several years (Case #32). The onset of his illness was difficult to determine as he gradually developed neurological symptoms in his forties which superimposed a pre-existing personality disorder with aggressiveness and outbursts. The brain showed PrP deposits that are similar to sporadic sCJD and the characteristic striping pattern, which is generally seen in the cerebellar cortex with 4OPRI and 6OPRI mutations, was absent. The cerebral white matter showed subtle white matter threads and there was no white matter PrP in the cerebellum.

6OPRI [144 bp insert] (1 male, 3 female); Figure 1E-H, Figure 3C,D, Figure 5A,B

The mean age for this group is 46 years (44–49) (cases #28-31). In all cases the cerebral grey matter contains abundant axonal threads (Figure 1E), as well as a granular synaptic stain and scattered microplaques. These threads are present in the myelinated fibres that radiate from the white matter into the cortex. The subcortical white matter contains PrP positive threads (Figure 1F, ICSM35 and 1G, KG9) and to a lesser extent, fine granules. The cerebellar molecular layer shows the classical, distinct “stripy” or “tigroid” pattern with an orientation perpendicularly to the cerebellar surface as described before [11, 18–20]. This pattern is a hallmark of octapeptide insert mutations and unlikely to be related to the filamentous deposits. Threads are also seen in the cerebellar granule layer (Figure 1H) and the cerebellar white matter shows variable amounts of dot- and thread-like staining. Case #30 shows a staining pattern that varies from the other cases with a 6OPRI mutation, in that the grey matter displays granular synaptic positivity without thread labelling and the white matter shows a predominance of granular labelling and fewer threads. In conclusion, this group shows consistent, strong filamentous positivity in both, grey and white matter (Figure 3), similar to the group with 4OPRI mutation (see above). The cerebellar white matter shows thread pathology, a feature also seen in cases with P102L and 4OPRI mutations. Co-localisation studies show a striking overlap of PrP with myelin sheaths, whilst colocalisation with axons was below detection limits, confirming that PrP threads are located extra-axonal, within the sheath (Figure 3C,D). However, ultrastructural studies of case #31 show granular protein deposits also in the axoplasm but, in keeping with immunofluorescence studies, more substantial deposits in the myelin sheath, which appears displaced and split by the amyloid aggregates (Figure 5A,B).

P102L (6 male, 4 female); Figure 1I-L, Figure 3E, F

The mean age for this group is 60.8 years (53–66) (cases #18-27). The cerebral grey matter of these cases contains frequent core- and multicentric plaques with diffuse granular synaptic pattern as described before [21, 22] and occasional threads composed of abnormal PrP (Figure 1I). The subcortical white matter contains scattered plaques of small size with variable fine granular and thread-like positivity (Figure 1J, ICLM35 and 1K, KG9). The cerebellar cortex shows plaques and large granular aggregates. The white matter contains variable plaques, granules and short threads (Figure 1L). In the majority of cases with P102L mutation, a small, and at most moderate amount of filamentous labelling can be observed in the white matter and less dense axonal positivity in the grey matter. In more than two thirds of the cases filamentous labelling is also present in the cerebellar white matter. In immunofluorescence studies however, nearly no filamentous material was detected (Figure 3E,F).

A117V (3 male, 1 female); Figure 1M-P, Figure 3G,H, Figure 5C,D

The mean age for this group is 43 years (40–47) (cases #14-17). The grey matter contains frequent granular plaques with diffuse, moderate-to-strong granular synaptic labelling as well as strongly labelled short (Figure 2M inset) and long filamentous white matter deposits (Figure 2M). In all three cases the white matter shows frequent and intense thread-like positivity with sparse fine granules and plaques (Figure 1N, ICLM35 and 2O, KG9). The cerebellar cortex mainly contains patches of plaques and the white matter shows occasional plaques with none or occasional dot-like PrP deposits as described before [23, 24]. In conclusion, this group displays a strong, dense filamentous PrP labelling in the cerebral white matter, similar to the 6OPRI and 4OPRI cases (see above), with significant filamentous PrP extending to the myelinated axons in the cortical grey matter and no cerebellar threads. A cross section of the pons (Figure 1Q) show occasional circular PrP deposits, strongly suggestive of a localisation in the myelin sheath. Immunofluorescent labelling confirms extensive co-localisation of PrP threads in the sheaths of myelinated fibres, and no axonal PrP. The dual labelling of PrP with neurofilaments demonstrates no colocalisation of the signal (Figure 3G, G’), whilst there is frequent and robust dual labelling of myelin sheaths and filamentous PrP (Figure 3H, H’). All A117V cases strongly accumulate filamentous PrP and it is generally very easy to detect filaments in the white matter of A117V cases. Ultrastructural studies of case #17 show substantial deposits of amyloid in the myelin sheath (Figure 4C,D).

D178N (2 male, 1 female); Figure 2A-C, Figure 3I,J

The mean age for this group is 57 years (51–60); cases #11-13. Both patients showed clinical CJD phenotype. Here, the cortical grey matter shows diffuse, weak to moderate fine granular synaptic stain with plaques of variable size. The subcortical white matter shows a wide variability (none to very frequent) of fine granular positivity but no thread like labelling (Figure 2A, ICSM35 and 3B, KG9). The cerebellar cortex shows patchy synaptic stain and granular aggregates with variable amounts of condensed granular PrP, occasionally forming microplaques, and the subcortical white matter has very rare or PrP dot-like positivity (Figure 2C). Overall these cases show little PrP labelling in the white matter and no thread-like PrP deposits. No threads were detected by immunofluorescence, confirming the observations by light microscopy (Figure 3I,J).

E200K (6 male, 3 female); Figure 2D-F, Figure 4A, B

The mean age for this group is 60 years (45–69) (cases #2-10). The cortical grey matter shows diffuse, granular or fine dot-like synaptic staining that ranges from weak to strong intensity, similar to that in sCJD. In addition, there was cortical perineuronal PrP deposition with variable amounts of microplaques, and there are rare filamentous PrP deposits. The subcortical white matter contains variable amounts (sparse to very frequent) of fine granular positivity (Figure 2D) and occasional white matter threads (Figure 2D, inset, ICSM35, and 2E, KG9). The cerebellar cortex shows patchy synaptic staining and granular aggregates with or without microplaques. The cerebellar white matter shows rare granular positivity and threads in some cases (Figure 2F). In one case (case 5 in Table 2) the cerebellar white matter showed frequent filamentous staining as well as dot-like positivity (Figure 2F, Figure 4A,B). This patient displayed an unusual clinical phenotype (younger age and significantly longer disease duration) compared to the other patients with the same mutation, which may explain the difference to the patients with older age and shorter disease duration. Overall, the PrP staining patterns in brains with the E200K mutation are similar to those observed in sCJD [25]. Among the cases with PRNP gene mutations, this is the only group that shows perineuronal PrP labelling. Double labelling immunofluorescence of one case (#5) shows filamentous PrP deposits. There is co-localisation of PrP in MBP positive myelin sheaths whilst PrP in axonal structures is below the detection limit (Figure 4A,B).

V210I (1 male)

In this single case (#1) with this mutation, the findings are very similar to those in E200K mutations. We found only cerebral PrP^Sc, some of which formed fine threads but no cerebellar PrP^Sc. The cortex shows synaptic labelling with a homogenous distribution across all cortical layers, similar to that of sCJD.

Sporadic CJD (26 cases), Figure 4C-H