Adult onset leukoencephalopathy with axonal spheroids and pigmented glia (ALSP) and Nasu–Hakola disease: lesion staging and dynamic changes of axons and microglial subsets

Abstract

The brains of 10 Japanese patients with adult onset leukoencephalopathy with axonal spheroids and pigmented glia (ALSP) encompassing hereditary diffuse leukoencephalopathy with axonal spheroids (HDLS) and pigmentary orthochromatic leukodystrophy (POLD) and eight Japanese patients with Nasu–Hakola disease (N‐HD) and five age‐matched Japanese controls were examined neuropathologically with special reference to lesion staging and dynamic changes of microglial subsets. In both diseases, the pathognomonic neuropathological features included spherically swollen axons (spheroids and globules), axon loss and changes of microglia in the white matter. In ALSP, four lesion stages based on the degree of axon loss were discernible: Stage I, patchy axon loss in the cerebral white matter without atrophy; Stage II, large patchy areas of axon loss with slight atrophy of the cerebral white matter and slight dilatation of the lateral ventricles; Stage III, extensive axon loss in the cerebral white matter and dilatation of the lateral and third ventricles without remarkable axon loss in the brainstem and cerebellum; Stage IV, devastated cerebral white matter with marked dilatation of the ventricles and axon loss in the brainstem and/or cerebellum. Internal capsule and pontine base were relatively well preserved in the N‐HD, even at Stage IV, and the swollen axons were larger with a higher density in the ALSP. Microglial cells immunopositive for CD68, CD163 or CD204 were far more obvious in ALSP, than in N‐HD, and the shape and density of the cells changed in each stage. With progression of the stage, clinical symptoms became worse to apathetic state, and epilepsy was frequently observed in patients at Stages III and IV in both diseases. From these findings, it is concluded that (i) shape, density and subsets of microglia change dynamically along the passage of stages and (ii) increase of IBA‐1‐, CD68‐, CD163‐ and CD204‐immunopositive cells precedes loss of axons in ALSP.

INTRODUCTION

Many diseases or pathological conditions widely involving the cerebral white matter in adult humans have been reported, but among them, only five are characterized by extensive spherically swollen axons (spheroids and globules): hereditary diffuse leukoencephalopathy with axonal spheroids (HDLS) 6, 7, 16, 21, 27, 30, 33, 34, 35, 39, 45, 70, 71, 83, Nasu–Hakola disease (N‐HD) 17, 18, 19, 24, 46, 50, 52, 53, 60, 79, 80, pigmentary orthochromatic leukodytrophy (POLD) 12, 28, 82, sudanophilic leukodystrophy (SLD) 57, 58 and traumatic diffuse brain injury 77. Recently adult onset leukoencephalopathy with axonal spheroids and pigmented glia (ALSP) was proposed as a comprehensive term encompassing HDLS and POLD because of similar characters of clinical and pathological features, and presence of mutations of colony stimulating factor 1 receptor (CSF1R) in the diseases 37, 49, 55, 66, 67, 86.

ALSP shows autosomal dominant inheritance and affected individuals suffer progressive dementia, depression and epilepsy in adulthood, with neuropathologically extensive degeneration of the cerebral white matter with axonal spheroids 21, 39, 70. It has been recently reported that CSF1R mutations account for between 10% and 25% of adult onset leukodystrophies 43. ALSP brains have revealed decreased expression of CSF1R on microglial cells 39, 70.

N‐HD, first established by Hakola et al. 17, 18 and Nasu et al. 52, 53, shows autosomal recessive inheritance with progressive dementia, epilepsy and bone fracture in adulthood. It is characterized by marked degeneration of the cerebral white matter with spheroids, bone fracture and peculiar membranous structures in the bone marrow and adipose tissue 17, 19, 52, 53. The causative genes are TREM2 (triggering receptor expressed on myeloid cells 2) and DNAX activating protein of 12 kDa (DAP12) 10, 40, 56, 61, 63. It has been reported that mutations in TREM2 and DAP12 lead to abnormalities in TREM2 function 48. Investigations of N‐HD brains have revealed that DAP12 immunoreactivity is absent 75, but exclusively positive in cases harboring a specific nucleotide substitution in DAP12 72.

CSF1R and TREM2/DAP12 are reportedly involved in microglial activation in the brain and bone marrow 2, 54, 62, and it has been shown that signal transduction of CSF1R and TREM2/DAP12 meets at the downstreams 47.

The above findings indicate that ALSP and N‐HD may share a common pathomechanism and neuropathological features. Staging system of histologic characters and severities of myelinated fiber loss of the white matter lesion were reported by Alturkustani et al. in the ALSP 3, however, it remains to be clarified whether the two diseases show similar patterns of progression and lesion staging in the brain, and share similar neuropathology, especially in terms of microglial alteration. These issues are very important to understand the pathomechanisms and develop treatments, and when attempting to diagnose the two diseases by brain imaging, and differentiating them neuropathologically in the absence of any genetic information or data on mutations in the CSF1R and TREM2/DAP12 genes.

In the present study, the brains of ten patients with ALSP, eight patients with N‐HD and five age‐matched controls were examined neuropathologically with special reference to lesion staging and microglial alteration, correlation between pathological stages and clinical symptoms and the pathomechanisms of these diseases were discussed.

MATERIALS AND METHODS

Examined cases

We examined 10 autopsied brains of ALSP patients from six families (six males and four females; ages at autopsy: 41–63 years; mean age ± standard deviation, 53.1 ± 7.7 years) 25, 30, 38, 39, 70, 81, 87 and eight autopsied brains of N‐HD patients from seven families (four males and four females; ages at autopsy: 38–55 years, 43.9 ± 5.9 years) 4, 5, 19, 31, 52, 53, 73, 74, 76. All of these patients were Japanese. Mutations in CSF1R were identified in all of the ALSP patients and p.M48WfsX6 mutation in DAP12 was identified in one of the N‐HD patients (N‐HD #5). The mutations are heterozygous in ALSP cases and homozygous in a N‐HD case. Detailed clinical characteristics (causative genes, family history, clinical diagnosis, cognitive decline, depression/anxiety, behavioral changes, frontal releasing signs, pyramidal tract signs, parkinsonism, epilepsy, modified Rankin scale 84 and bone fractures) of the ten ALSP and eight N‐HD patients are shown in Table 1. Most of N‐HD cases had not been examined the mutation of DAP12/TREM2, but the cases showed characteristic features as N‐HD in clinical symptoms including young onset, dementia, epilepsy and bone fracture and pathological findings such as spheroids in the cerebral white matter and membranous cystic lesions in the bone marrow. It is not clear whether N‐HD will be differentiated by causative genes, such as DAP12 and TREM2. Thus N‐HD patients were examined as a group in the present study as having distinctive clinical and pathological findings. All of the clinical information was obtained from the clinical charts recorded by the respective attending physicians.

Table 1.

Clinical characteristics of 14 ALSP and eight N‐HD patients.

Patient	Family line	Sex	Age at onset	Age of death	Duration of disease (years)	Causative gene CSF1R	Family history	Brain weight (g)	Clinical diagnosis	Neuropsychiatric symptoms							mRS at death 65	Bone fracture	Membranous lesion	References
										Cognitive decline	Depression/ Anxiety	Behavioral changes	Frontal releasing	Pyramidal tract signs	Parkinsonism signs	Epilepsy
ALSP 1	SH1	M	40	41	1	p. K793T	−	1430	juvenile AD	+	+	+	+	−	+	−	3	−	−	38
ALSP 2	OK2	F	40	44	4	p. I794T	+ (AD)	1160	AD	+	+	+	+	+	+	+	5	−	nd or ne	81
ALSP 3	MY1	M	40	53	13	p. I794T	+ (AD)	1000	familial LE	+	nd	+	+	nd	+	+	5	−	nd or ne
ALSP 4	MY1	M	40	56	16	p. I794T	+ (AD)	980	familial LE	+	+	nd	+	nd	nd	+	5	−	nd or ne
ALSP 5	YK	F	41	54	13	p. S688EfsX13	−	910	AD	+	−	+	+	+	−	−	5	−	−	25, 39
ALSP 6	AI1	F	48	63	15	p. C653Y	+ (AD)	885	familial LE	+	nd	nd	+	nd	nd	+	5	−	nd or ne	49, 70
ALSP 7	MY1	M	50	56	6	p. I794T	+ (AD)	980	familial LE	+	+	+	+	nd	nd	−	5	−	nd or ne
ALSP 8	MY1	M	52	58	6	p. I794T	+ (AD)	1010	familial LE	+	−	+	+	+	nd	+	5	−	nd or ne
ALSP 9	AI2	M	56	62	6	p. S759F	+ (AD)	1152	nd	+	+	+	+	+	+	−	5	−	nd or ne	30, 87
ALSP 10	AI1	F	No	44	0	p. C653Y	+ (AD)	1330	Pulmonary tuberculosis	−	−	−	−	−	−	−	0	−	nd or ne	49, 70
			symptoms*
						DAP12/TREM2
N‐HD 1	KN	M	14	43	29	ne	−	680	N‐HD	+	−	+	+	+	nd	+	5	+	+	4
N‐HD 2	YK2	F	20	49	29	ne	−	840	N‐HD	+	−	+	+	+	+	+	5	+	+	5
N‐HD 3	TK1	M	23	42	19	ne	+ (AR)	910	N‐HD	+	−	+	+	+	nd	+	5	+	+	5
N‐HD 4	SH3	M	26	38	12	unknown †	−	1050	nd	+	−	+	+	nd	nd	+	5	+	+	19, 31, 52, 53
N‐HD 5	HY	M	28	42	14	p. M48WfsX6 ‡	−	1180	N‐HD	+	−	+	+	nd	nd	+	5	+	+	73, 74, 76
N‐HD 6	TK2	F	29	37	8	ne	+ (AR)	1010	N‐HD	+	−	+	+	+	+	+	4	−	+	5
N‐HD 7	MY2	F	33	45	12	ne	+ (AR)	935	nd	+	−	+	+	+	+	+	5	−	+
N‐HD 8	MY2	F	33	55	22	ne	+ (AR)	580	N‐HD	+	−	+	+	+	+	−	5	−	ne §

*ALSP patient #10 was a younger sister of ALSP patient #6. The neuropathology in patient #10 included leukoencephalopathy with axonal spheroids, although no neuropsychiatric symptoms had been evident during her lifetime.

†In N‐HD patient #4, a conversion at nucleotide position 116 had been reported previously 30, but later this had been proven to be false positive by the authors (Dr. K. Sano and Dr. N. Amano, personal communication).

‡DAP12 c. 141delG: p. M48WfsX6.

§N‐HD patients #7 and #8 were sisters and autopsy showed membranous lesions in the adipose tissue in patient #7, although neither showed any bone fractures during their clinical courses.

M; male, F; female, age; years old, ne; not examined, nd; not described, FTD; frontotemporal dementia, LE; leukoencephalopathy, AD; Alzheimer's disease, VaD; vascular dementia, mRS; modified Rankin Scale 82.

ALSP patient #1 had been diagnosed on the basis of genetic examination when mild cognitive disturbance appeared. He died of sepsis unexpectedly in the early stage of the disease 38. ALSP patients #6 and #10 were sisters, and ALSP patients #3, #4, #7 and #8 were cousins. CSF1R p. I794T was recently detected in this family (personal communication to YM, one of the present authors) (Table 1).

Five age‐matched controls (three males and two females) were used for comparison of the neuropathological findings. This study was performed in accordance with the provisions of the Declaration of Helsinki (1995), and was approved by the Ethics Committees of Shinshu University (No. 2149).

Neuropathological procedures

Formalin‐fixed paraffin‐embedded serial 6‐µm‐thick coronal sections of the frontal, parietal, temporal and occipital lobes, transverse sections of the brainstem and transverse or sagittal sections of the cerebellum were subjected to neuropathological examination. Hematoxylin and eosin (HE), Klüver–Barrera (K‐B), Bodian, Masson trichrome, periodic acid‐Schiff (PAS) and Holzer preparations of the sections were made. Immunohistochemistry was performed using primary antibodies against phosphorylated (p)‐neurofilament (NF), non‐p‐NF, IBA‐1, cluster of differentiation (CD) 68, CD163, CD204, CSF1R, DAP12, TREM2, CD3, CD20, myelin basic protein (MBP), glial fibrillary acidic protein (GFAP), p‐trans activation responsive region‐DNA‐binding protein of 43 kDa (p‐TDP‐43), synaptophysin (SYP), p‐tau, p‐α‐synuclein, human ß amyloid and ubiquitin (Ub) (Table 2). Characteristics of IBA‐1, CD68, CD163, CD204 are summarized in Table 3.

Table 2.

Antibodies used.

Detection	Antibody type	Dilution	Increase of antigenicity	Source
Phosphorylated neurofilament	Mouse/monoclonal (clone SMI31)	1:2000	Boiling	COVANCE, Princeton, USA
Non‐phosphorylated neurofilament	Mouse/monoclonal (clone SMI32)	1:500	Autoclaving	COVANCE, Princeton, USA
Iba1	Rrabbit/polyclonal	1:4000	Autoclaving	Wako Pure Chemical Industries, Osaka, Japan
CD163	Mouse/monoclonal (clone 10D6)	1:200	Autoclaving	Leica Biosystems, Newcastle Upon Tyne, UK
CD204	Mouse/monoclonal (clone SRA‐E5)	1:200	Autoclaving	TransGenic, Kobe, Japan
CD68	Mouse/monoclonal (clone KP1)	1:800	Boiling	DAKO, Glostrup, Denmark
Colony stimulating factor 1 receptor (CSF1R)	Rabbit/polyclonal	1:400	Protease K	LifeSpan BioSciences, Seattle, USA
DNAX activating protein of 12 kDa (DAP12)	Rabbit/polyclonal	1:500	Autoclaving	Santa Cruz Biotechnology, Dallas, USA
Triggering receptor expressed on myeloid cells 2 (TREM2)	Rabbit/polyclonal	1:100	Autoclaving	Sigma, St. Louis, USA
CD3	Rabbit/monoclonal (clone SP7)	1:400	Autoclaving	Nichirei Bioscience, Tokyo, Japan
CD20	Rabbit/monoclonal (clone EP459Y)	1:300	Non	Abcam, Cambridge, UK
Myelin basic protein (MBP)	Rabbit/polyclonal	1:600	Boiling	DAKO, Glostrup, Denmark
Glial fibrillary acidic protein (GFAP)	Mouse/monoclonal (clone 6F2)	1:100	Autoclaving	DAKO, Glostrup, Denmark
Phosphorylated TAR DNA binding protein 43 (p‐TDP43)	Mouse/monoclonal (clone 11‐9)	1:10 000	Autoclaving	COSMO BIO, Tokyo, Japan
Synaptophysin (SYP)	Mouse/monoclonal (clone 171B5)	1:1000	Autoclaving	Medical & Biological Laboratories, Japan
Phosphorylated tau	Mouse monoclonal (clone AT8)	1:3000	Non	Innogenetics, Ghent, Belgium
Phosphorylated synuclein	Mouse monoclonal (clone #64)	1:20 000	Formic acid	Wako Pure Chemical Industries, Osaka, Japan
Human beta amyloid	Mouse monoclonal (clone 6F/3D)	1:800	Formic acid	DAKO, Glostrup, Denmark
Ubiquitin (UB)	Rabbit/polyclonal	1:1000	Autoclaving	DAKO, Glostrup, Denmark

Antigenicity was increased by autoclaving (121°C, 20 min) in 0.01 mol/L citrate buffer solution (pH 6.0), boiling in a microwave oven (500 W, 7 min) in 0.01 mol/L citrate buffer solution (pH 7.6), 99% formic acid for 5 min or 0.005 µg/mL protease K in 10% normal horse serum in 0.01 mol/L phosphate‐buffered saline with Triton X‐100 (pH 7.6) at 37°C for 30 min.

Table 3.

Characteristics of markers for microglia.

	Molecule	Function	References
IBA‐1	calcium‐binding protein	microglial activation	1, 26
CD163	hemoglobin scavenger receptor	macrophage activation	15, 44, 51
CD204	macrophage scavenger receptor	propagation of lesion	32, 51
CD68	lysosome	phagocytic activity	51

Immunohistochemical staining was performed using the avidin–biotin–peroxidase complex (ABC) method (Vectastain ABC Elite kit, Vector, Burlingame, CA, USA). Non‐specific binding of the ABC system reagents was blocked by pretreating the sections with 0.3% hydrogen peroxide in methanol and a normal blocking serum, and then incubating them with the required primary antibody overnight at 4°C. The sections were then rinsed in phosphate‐buffered saline with Triton X‐100 (PBST) and incubated for 1 h with the secondary reagent containing a biotinylated anti‐rabbit or anti‐mouse IgG antibody (diluted 1:200) at 37°C, and finally with the ABC solution for 1 h at room temperature (RT). As the secondary reagent, EnVision^TM Dual link System‐HRP (DAKO) was used for CSF1R for 30 min. The sections were subjected to a peroxidase reaction with 30 μL ImmPACT^TM DAB Chromogen Concentrate (Vector) (diluted 1:2, with 50 mmol/L Tris HCl (pH 7.6)) in 1 mL ImmPACT^TM Diluent (Vector) at RT. After stopping the reaction, the sections were rinsed in tap water, then stained with hematoxylin.

For CD68, p‐NF and MBP, antigenicity for immunohistochemistry was increased by boiling the sections in 0.01 mol/L citrate‐buffered solution (pH 7.6) in a microwave oven (750 W, 25 min); for IBA‐1, CD163, CD204, DAP12, TREM2, CD3, non‐p‐NF, GFAP, p‐TDP‐43, SYP and Ub by autoclaving (121°C, 20 min) in 0.01 mol/L citrate‐buffered solution (pH 6.0); for p‐alpha synuclein, sections were incubated with 99% formic acid for 5 min at RT, and for CSF1R by incubation with proteinase K (0.005 μg/mL, 37°C, 30 min). Primary antibodies used for immunohistochemistry are listed in Table 2.

In the frontal and parietal lobes, double staining was performed using PAS and immunohistochemistry for GFAP or IBA‐1 with SMI‐31, and double immunohistochemistry was performed for p‐NF with MBP, IBA‐1 with CD163, and CD163 with CD204. For double immunohistochemistry for p‐NF with MBP, the sections were boiled in a microwave oven, anti‐p‐NF antibody (1:2000) was applied, and the sections were then subjected to peroxidase reaction with 30 µL ImmPACT^TMDAB Chromogen concentrate (Vector, Burlingame, CA, USA) (diluted 1: 2, by 50 mM TrisHCl (pH 7.6)) in 1 mL of ImmPACT^TM Diluent (Vector) for 5 min at RT, using a single staining procedure. After pretreatment in Denaturing Solution (A:B = 1:2) (Biocare Medical, Concord, CA, USA) for 3 min, the sections were incubated in PBST (pH 7.6). Anti‐MBP antibody (1:400) was added and the sections were then incubated for 17 h at 4°C. They were then washed three times in PBST, and incubated with secondary biotinylated antibody for 1 h at 37°C. They were again washed three times in PBST, followed by use of the avidin–biotin–immunoperoxidase complex method and reaction with Vina Green Chromogen (Biocare Medical, Concord, CA, USA). For double immunohistochemistry for IBA‐1 and CD163, the sections were processed by autoclaving, and then we first applied anti‐CD163 antibody (1:100), followed by staining with Vectastain Blue (Vector). Then, anti‐IBA‐1 antibody (1:2000) was applied, and the sections were subjected to the peroxidase reaction. For double immunohistochemistry for CD163 and CD204, after autoclaving we first applied anti‐CD163 antibody (1:200), and the sections were then subjected to the peroxidase reaction. Anti‐CD204 antibody (1:50) was then applied, and the sections were stained with Vectastain Blue (Vector).

As antibody controls, the primary antisera were either omitted or were replaced with normal rabbit or mouse serum. Several specimens of neural and non‐neural tissue from the patients served as positive or negative tissue controls. The preparations were examined by light microscopy.

The findings were evaluated neuropathologically and semiquantitatively. Neuropathological lesion staging was based on the degree and area of axon loss in both diseases. The degree of loss of myelinated fibers, the appearance of swollen axons, loss of neurons and changes in the morphological distribution of immunoreactivity for IBA‐1, CD163, CD204, CD68, CFS1R, DAP12 and DAP12 in the centrum semiovale were examined and expressed as a diagram for each stage. The severity of the neuropathological findings was graded semiquantitatively based on the severity of loss of findings or the frequency of findings into four using bars with different thickness and blank space.

T cells were identified as small round cells immunopositive for CD3, and B cells as immunopositive for CD20, and round cells several micrometers in diameter with relatively dark round nuclei evident by hematoxylin–eosin staining were identified as oligodendroglial cells.

Cells immunoreactive for IBA‐1, CD163, CD204, CD68, CFS1R or DAP12 in the centrum semiovale were examined by light microscopy in 10 fields of view at ×400 magnification. When more than 10 immunopositive cells were present in each field, the count was expressed as “many,” 5–9 cells as “moderate” and fewer than 4 as “few.” The severity of neuronal loss in the frontal cortex was examined in 5 fields of view at ×20 magnification. Also the degree of axon loss was examined in various white matter regions in each brain in 5 fields of view at ×400 magnification. For both evaluations, <20% loss was expressed as “slight,” loss of 21–50% as “moderate” and loss of >51% as “severe.” The frequency of swollen axons was examined in various white matter regions in each brain in 5 fields of view at ×20 magnification. The presence of >5 swollen axons per field was expressed as “many,” 2–4 as “moderate” and <1 as “few”. The ratio of cells immunopositive for IBA‐1 and CD163 or CD163 and CD204 was evaluated by dividing the number of cells positive for both by the number of cells positive for either in each brain region in 10 fields of view using light microscopicy at ×400 magnification

RESULTS

Loss of axons in the brain white matter and staging of lesions

No significant loss of axons was identified in control brains.

ALSP

Based on the severity and area of axon loss, lesions were considered to be divisible into four stages: Stage I: Patchy loss of myelinated fibers in the cerebral white matter without remarkable cerebral atrophy. Stage II: Patchy loss of large areas of myelinated fibers in the cerebral white matter with slight cerebral atrophy. Good preservation of U‐fibers in the cerebrum and white matter in the brainstem and cerebellum. Slight dilation of the lateral and third ventricles. Stage III: Extensive degeneration of the cerebral white matter including the corpus callosum, internal capsule and some parts of U‐fibers. Atrophy of the thalamus, and moderate dilatation of the lateral and third ventricles. Good preservation of the optic tract, lateral geniculate body, optic radiation, white matter in the cerebellum and transverse fibers and tegmentum of the brainstem. Stage IV: Cerebral white matter devastation with marked atrophy in the frontal white matter, centrum semiovale, temporal white matter, corpus callosum and thalamus and severe dilatation of the lateral ventricles. Thinning of the cerebral cortex. Relatively good preservation of the hippocampus. Severe degeneration of the cerebellar white matter and myelinated fibers in the pontine base (Figures 1A–O and 3A–D).

Figure 1.

Lesion staging in ALSP. Stage I: Small patchy loss of myelinated fibers in the frontal white matter and a large patch in the parietal white matter (A and B, arrows). The cerebral white matter shows no atrophy. Stage II: Large patchy loss of myelinated fibers in the frontal white matter and centrum semiovale (C and D, arrows) with slight atrophy of the cerebral white matter. Well preserved U‐fibers (C and D, asterisks) and white matter in the brainstem and cerebellum. Slightly dilated lateral and third ventricles (D and E). Stage III: Extensive degeneration of the cerebral white matter (F, arrows) including the corpus callosum (G, arrows), internal capsule (H, arrow) and some parts of U‐fibers (F, arrowheads). The thalamus appears atrophic (H, arrowhead), and moderate dilatation of the lateral and third ventricles is evident (H). However, the optic tract, lateral geniculate body (H, asterisk), optic radiation, white matter in the cerebellum and transverse fibers and tegmentum of the brainstem are well preserved (H–J). The superior cerebellar peduncles are well preserved (I, asterisk). Stage IV: Devastation of the cerebral white matter with marked atrophy in the frontal white matter (K, arrows), centrum semiovale, temporal white matter (M, arrow), corpus callosum and thalamus (L, arrowhead), and severe dilatation of the lateral ventricles (K). Thickening of the cerebral cortex is slightly to moderately decreased (K and M). Relative preservation of the hippocampus (M, asterisk). Severe degeneration of the cerebellar white matter and the pontine base (N and O, arrows; lesions). The superior cerebellar peduncles show moderate involvement (N, arrowheads). A–O: Klüver–Barrera preparation. Magnifications among cerebrum, cerebellum and brainstem are different.

Devastation of the temporal white matter was observed in ALSP patients at Stages III and IV, and marked thinning of the temporal cortex was evident in patients at Stage IV, although the hippocampus was relatively well preserved (Figure 1M).

Regarding the corticospinal tracts in ALSP, the brain at Stage I showed suspected loss of axons in the internal capsule, cerebral peduncle, corticospinal tract in the pontine base and medullary pyramids. Frontobulbar tracts were unremarkable. A few swollen axons were evident in the internal capsule. In the brain at Stage II, slight loss of axons was observed in the internal capsule, cerebral peduncle, corticospinal tract in the pontine base (Figure 1E), pyramids and lateral corticospinal tracts in the spinal cord. Scattered swollen axons were revealed in the internal capsule and cerebral peduncle including frontobulbar tracts. The brain at Stage III demonstrated marked loss of axons in the internal capsule, cerebral peduncle including frontobulbar tracts, corticospinal tract in the pontine base, pyramids and lateral and anterior corticospinal tracts of the spinal cord. Many swollen axons were evident in the internal capsule, cerebral peduncle, corticospinal tract in the pontine base (Figure 1I) and pyramids. Some patients showed many swollen axons in the corticospinal tracts of the spinal cord, whereas others showed a few. In the brain at Stage IV, the corticospinal tracts in the internal capsule, cerebral peduncle, pontine base (Figure 1N) and pyramids were devastated, and no swollen axons were seen. Severe loss of axons and a few swollen axons were observed in the corticospinal tracts of the spinal cord at this stage.

Specific brain findings in individual patients with ALSP were as follows. Patient #10 showed loss of myelinated fibers in small patches in the frontal white matter and a large patch in the parietal white matter, as reported by Riku et al. 70. The cerebral white matter showed no atrophy (Stage I, Table 4, Figure 1A,B arrows; patches, Figure 3A). Patients #1 showed patchy spotted loss of myelinated fibers in the frontal white matter and centrum semiovale with slight atrophy of the cerebral white matter (Figure 1C,D, arrows; patches). Although the cerebral white matter showed slight but diffuse degeneration, U‐fibers (Figure 1C, asterisk) and white matter in the brainstem and cerebellum were well preserved (Figure 1E). The lateral and third ventricles showed slight dilatation (Stage II, Table 4, Figure 1C‐E and 3B). Patients #2, #4, #5 and #9 showed extensive degeneration of the cerebral white matter (Figure 1F, arrows) including the corpus callosum (Figure 1G, arrows), internal capsule (Figure 1H, arrow) and some parts of U‐fibers (Figure 1F, arrowheads). The thalamus was atrophic (Figure 1H, arrowhead), and moderate dilatation of the lateral and third ventricles was evident (Figure 1H). However, the optic tract, lateral geniculate body (Figure 1H, asterisk), optic radiation, white matter in the cerebellum and transverse fibers and tegmentum of the brainstem were well preserved (Stage III, Table 4, Figures 1I,J and 3C). Patients #3, #6 and #7 showed devastation of the cerebral white matter with marked atrophy of the frontal white matter (Figure 1K, arrows), centrum semiovale, internal capsule (Figure 1L, arrow), temporal white matter (Figure 1M, arrow), corpus callosum and thalamus (Figure 1L, arrowhead), and severe dilatation of the lateral ventricles. The thickness of the cerebral cortex was slightly to moderately decreased (Figure 1K,M). The hippocampus was relatively well preserved (Figure 1M, asterisk). The cerebellar white matter and the myelinated fibers in the pontine base were severely degenerated (Stage IV, Table 4, Figure 1N,O, arrows; lesions, and Figure 3D). The superior cerebellar peduncles were moderately affected (Figure 1N, arrowheads).

Table 4.

Neuropathological findings and staging of lesions in the white matter.

ALSP		Control	Stage I	Stage II	Stage III	Stage IV
	Patient no.	Control, 5 subjects	ALSP #10	ALSP #1	ALSP #2, 4, 5, 8, 9	ALSP #3, 6, 7
Axonal loss in brain white matter		no remarkable axon loss	patchy axon loss in cerebral white matter without brain atrophy	large patchy axon loss with slight atrophy of cerebral white matter and slight dilatation of ventricles	extensive axon loss in cerebral white matter and dilated ventricles without remarkable axon loss in brain stem / cerebellum	devastated cerebral white matter with marked dilatation of ventricles and axon loss in brain stem / cerebellum
Swollen axon	Front. Cortex	absent	swollen axons (<10 µm); a few, pNF‐ballooned neurons; a few	swollen axons (<15 µm); a few, pNF‐ ballooned neurons; a few	swollen axons (<10 µm); a few / scattered, pNF‐ballooned neurons; a few	swollen axons (<8 µm); many, pNF‐ ballooned neurons; many
	Cereb. W. M.	absent	swollen axons (<20µm); scattered	swollen axons (<40µm); many	swollen axons (<30µm); scattered / many	swollen axons (<15µm); a few
SYP	Front. Cortex	diffuse in neuropile	diffuse in neuropile	diffuse in neuropile	diffuse in neuropile	diffuse in neuropile, swollen axons; a few
	Cereb. W. M.	absent	positive in swollen axons; scattered	positive in swollen axons; many	absent	absent
Iba1	Front. Cortex	ramified cells; a few / scattered	ramified & star‐shaped cells; scattered	slightly swolllen ramified & star‐shaped cells; scattered	swollen ramified cells; many	ramified cells; many
	Cereb. W. M.	slightly swollen ramified cells; a few / scattered	swollen ramified microglia; scattered	ramified & ameboid cells; many	ramified & ameboid cells; a few	ameboid, ramified & star‐shaped cells; a few
CD163	Front. Cortex	ramified & capillary cells; a few	ramified & capillary cells; a few	ramified cells & capillary cells, a few / scattered	ramified cells & capillary cells; a few / many	swollen ramified & capillary cells; a few / many
	Cereb. W. M.	ramified & capillary cells; a few	swollen ramified & capillary cells; scattered	swollen ramifiied & ameboid cells, many / a few	ameboid & capillary cells; many / scattered	ameboid & capillary cells; a few
CD204	Front. Cortex	absent	ramified & capillary cells; a few	ramified cells & capillary cells; scattered / a few	ramified & capillary cells; scattered / a few	swollen ramified cells; many / absent
	Cereb. W. M.	ramified & ameboid cells; a few	ameboid cells; many, ramified & capillary cells; a few	swollen ramified & ameboid cells; many	swollen ramified & ameboid cells; many	ameboid cells; scattered
CD68	Front. Cortex	small glial & capillary cells; scattered	small ramified and capillary cells; scattered	swollen ramified cells; scattered	capillary cells; a few	small ramifed & capillary cells; a few
	Cereb. W. M.	ramified cells with short processes; scattered	large ameboid cells; scattered	large ameboid cells; many	ameboid cells; scattered	small ameboid cells; a few
CSF1R	Front. cortex	absent	star‐shaped & ramified cells; a few	star‐shaped & ramified cells; a few	star‐shaped cells; a few	star‐shaped cells; scattered
	Cereb. W. M.	ramified & star‐shaped cells; many / scattered	star‐shaped & plump ramified cells; many	star‐shaped & plump ramified cells; many	star‐shaped cells; many	star‐shaped & plump ramified cells; many
DAP12	Front. cortex	blood & capillary cells, ramified & star‐shaped cells; many	blood & capillary cells, ramified & star‐shaped cells; many	blood & capillary cells, ramified & star‐shaped cells; many	blood & capillary cells, ramified & star‐shaped cells; many	blood & capillary cells, ramified & star‐shaped cells; many
	Cereb. W. M.	ramified cells; scattered	swollen ramified cells; scattered	swollen ramified cells; many	swollen ramified cells; many	ameboid & ramified cells; a few
TREM2	Front. cortex	neurons; a few / many	neurons; a few	neurons; a few	neurons, blood & capillary cells; a few	blood & capillary cells; a few
	Cereb. W. M.	blood & capillary cells; a few	blood & capillary cells; a few	blood & capillary cells; a few	blood & capillary cells; a few	blood & capillary cells; a few
GFA	Front. cortex	protoplasmic & perivascular cells; scattered	protoplasmic & perivascular cells; scattered	protoplasmic cells & perivascular cells; scattered	protoplasmic & perivascular cells; scattered	protoplasmic cells; scattered
	Cereb. W. M.	fibrillary & perivascular cells; scattered	fibrillary & perivascular cells; scattered	fibrillary & perivascular cells; scattered	large reactive protoplasmic & perivascular cells; scattered	fibrillary & perivascular cells; many

N‐HD		Control	Stage I	Stage II	Stage III	Stage IV
	Patient no.	Control, 5 subjects		N‐HD #6	N‐HD #2, 3, 4, 5, 7	N‐HD #1, 8
Axonal loss in brain white matter		no remarkable axon loss		large patchy axon loss with slight atrophy of cerebral white matter and slight dilatation of ventricles	extensive axon loss in cerebral white matter and dilated ventricles without remarkable axon loss in brain stem / cerebellum	devastated cerebral white matter with marked dilatation of ventricles and axon loss in cerebellum, but relatively preserved brain stem
Swollen axon	Front. cortex	absent	n.e.	swollen axons; a few, pNF‐neurons; a few	swollen axons; a few, pNF‐neurons; a few	swollen axons; a few, pNF‐neurons; a few
	Cereb. W. M.	absent	n.e.	swollen axons (<25 µm); scattered	swollen axons (<18 µm); scattered / a few	swollen axons (<15 µm); scattered
SYP	Front. cortex	diffuse in neuropile	n.e.	diffuse in neuropile	diffuse in neuropile	diffuse in neuropile
	Cereb. W. M.	absent	n.e.	absent	absent	absent
Iba1	Front. cortex	ramfied cells; a few / scattered	n.e.	ramified cells; scattered	ramified cells; scattered / many	swollen ramified cells; many / scattered
	Cereb. W. M.	slightly swollen ramified cells; a few / scattered	n.e.	ramified cells; a few	ameboid cells; scattered / many	ramified & ameboid cells; many / scattered
CD163	Front. cortex	ramified & capillary cells; a few	n.e.	ramified & capillary cells; a few	ramified & capillary cells; a few	ramified cells; a few
	Cereb. W. M.	ramified & capillary cells; a few	n.e.	ramified & capillary cells; a few	ameboid & capillary cells; scattered	ameboid, ramified & capillary cells; a few
CD204	Front. cortex	absent	n.e.	n.e.	capillary cells; a few	capillary cells; a few
	Cereb. W. M.	ramified & ameboid cells; a few	n.e.	n.e.	capillary cells; scattered	ameboid cells; many, capillary cells; a few
CD68	Front. cortex	small glial & capillary cells; a few	n.e.	capillary cells; a few	capillary cells; a few	ameboid, ramified & capillary cells; scattered
	Cereb. W. M.	ramified cells with short processes; scattered	n.e.	small ameboid cells; scatterd	small ameboid & capillary cells; many in N‐HD #3, scatterred in other N‐HD	small ameboid & capillary cells; scattered / many
CSF1R	Front. cortex	absent	n.e.	absent	absent	star‐shaped cells; a few
	Cereb. W. M.	ramified & star‐shaped cells; many / scattered	n.e.	ramified & star‐shaped cells; many / scattered	star‐shaped & ramified cells; scattered	star‐shaped & ramified cells; many
DAP12	Front. cortex	blood & capillary, ramified & star‐shaped cells; many	n.e.	absent	neurons; scattered in N‐HD #3, a few in other N‐HD	neurons; a few
	Cereb. W. M.	ramified cells; scattered	n.e.	absent	small glial cells & swollen axons; many in N‐HD #3, absent in other N‐HD	small glial cells; a few
TREM2	Front. cortex	neurons; a few / many	n.e.	absent	neurons; scattered in N‐HD #3, a few in other N‐HD, blood & capillary cells; a few in other N‐HD than #3	blood & capillary cells; a few
	Cereb. W. M.	blood & capillary cells; a few	n.e.	absent	capillary & blood cells; a few, swollen axons; a few in N‐HD #3,	blood & capillary cells; a few
GFA	Front. cortex	protoplasmic & perivascular cells; scattered	n.e.	protoplasmic & perivascular cells; scattered	protoplasmic & perivascular cells; scattered	protoplasmic cells; scattered
	Cereb. W. M.	fibrillary & perivascular cells; scattered	n.e.	fibrillary cells; many	fibrillary cells; many	fibrillary cells; many

Front; frontal, Cereb. W. M.; cerebral white matter, Numbers in parentheses represent the maximum diameter of swollen axons in patients at each stage.

Figure 3.

Schematic representation of lesion progression and staging in ALSP and N‐HD. At Stage I of ALSP, axonal loss occurs as irregular patches in the centrum semiovale and frontal white matter (A). At Stage II, loss of axons is evident in the cerebral white matter, with relative predominance in the frontal lobe (B). At Stage III, cerebral involvement is aggravated, and the pontine base is degenerated (C). At Stage IV, in addition to aggravated cerebral and pontine base lesions, cerebellar involvement is also evident (D). In N‐HD, involvement of the temporal lobe is more evident at Stage II than at the same stage in ALSP, and the internal capsule is clearly spared in N‐HD (E). At N‐HD Stage III, cerebral involvement is aggravated, but the internal capsule and pontine base are spared (F). At Stage IV, in addition to aggravation of the cerebral lesions, cerebellar involvement is evident, but the internal capsule and pontine base are relatively spared. Temporal lesions are more marked at every Stage in N‐HD than is the case in ALSP (G). Red marking; regions of axon loss

N‐HD

Based on the severity and area of axon loss, lesions were also considered to be divisible into four stages, although none of the patients had Stage I lesions. Stage II: Slight atrophy and irregular patches of myelinated fiber loss in the frontal, temporal and parietal white matter, centrum semiovale and corpus callosum with moderate dilatation of the lateral ventricles. Good preservation of U‐fibers and white matter in the pons and cerebellum (Figures 2A–C and 3E). Stage III: Moderate atrophy and loss of myelinated fibers of the frontal, temporal, parietal and occipital white matter, centrum semiovale and corpus callosum. Atrophic temporal cortex and hippocampus. Severe deterioration of the thalamus, hippocampus and parahippocampal gyrus with loss of neurons. Good preservation of the internal capsule, optic tract, Ammon's horn and white matter in the brainstem and cerebellum (Figures 2D–G and 3F). Stage IV: Marked atrophy and loss of axons in the frontal and temporal white matter, centrum semiovale and corpus callosum with severe dilatation of the lateral ventricles. Deterioration of the thalamus, hippocampus and parahippocampal gyrus. Degeneration of the cerebellar white matter, especially posteriorly, but relative preservation of the internal capsule, optic tract, pontine base and superior cerebellar peduncles (Figures 2H–K and 3G).

Figure 2.

Lesion staging in N‐HD. Stage II: Slight atrophy and irregular patches of myelinated fiber loss (A and B, arrows) in the frontal white matter, centrum semiovale and corpus callosum with moderate dilatation of the lateral ventricles (B) slight atrophy of the anterior nucleus of thalamus (B, arrowhead). Good preservation of U‐fibers (A and B, asterisks) and the white matter of the pons (C) and cerebellum. Stage III: Moderate atrophy of the frontal white matter, the centrum semiovale (D and E, arrows) and the corpus callosum (E, arrowhead; i). Evident involvement of U‐fibers (E, arrowheads). Severe deterioration of the thalamus (E, arrowhead; ii), but sparing of the internal capsule (E, asterisk; iii), optic tract (E, asterisk; iv), dentate gyrus (E, asterisk; v) and white matter of the brainstem and cerebellum (F and G). Stage IV: Marked atrophy and loss of axons in the frontal and temporal white matter, centrum semiovale (H and I, arrows) and corpus callosum (I, arrowhead; i) with severe dilatation of the lateral ventricles (H and I). Deterioration of the thalamus (I, arrowhead; ii), hippocampus and parahippocampal gyrus (I, arrowhead; iii), but relative preservation of the internal capsule (I, asterisk; iv) and optic tract (I, asterisk; v). Degeneration of the cerebellar white matter, especially the posterior part (K, arrow), but relative preservation of the pontine base (J, asterisk; vi) and superior cerebellar peduncles (J, asterisks; vii). A–K: K‐B. Magnifications among cerebrum, cerebellum and brainstem are different.

In the corticospinal tracts of the brain in N‐HD at Stage II, a few swollen axons were evident in the internal capsule including frontobulbar tracts, corticospinal tract in the pontine base (Figure 2C) and medullary pyramids, without remarkable loss of axons. At Stage III there was suspected loss of axons in the internal capsule, cerebral peduncle including frontbulbar tracts, corticospinal tract in the pontine base (Figure 2F), pyramids and corticospinal tracts of the spinal cord. A few swollen axons were seen in the internal capsule, cerebral peduncle, corticospinal tracts in the pontine base and pyramids. At Stage IV there was moderate loss of axons in the internal capsule, cerebral peduncle including frontobulbar tracts, corticospinal tract in the pontine base (Figure 2J) and pyramids. A few swollen axons were also evident there. Slight loss of axons was notable in the lateral corticospinal tracts of the spinal cord with a few swollen axons.

Specific brain findings in individual patients with N‐HD were as follows. Patient N‐HD #6 showed slight atrophy and irregular patches of myelinated fiber loss (Figure 2A,B, arrows) in the frontal white matter, centrum semiovale and corpus callosum with moderate dilatation of the lateral ventricles. U‐fibers (Figure 2A,B, asterisks) and the white matter of the pons and cerebellum appeared well preserved (Stage II, Table 4, Figures 2A–C and 3E). Patients #2, #3, #4, #5 and #7 showed moderate atrophy of the frontal white matter, and the centrum semiovale (Figure 2D,E, arrows), corpus callosum (Figure 2E, arrowhead; i) and thalamus (Figure 2E, arrowhead; ii) showed severe deterioration; however, the internal capsule (Figure 2E, asterisk; iii), optic tract (Figure 2E, asterisk; iv), dentate gyrus (Figure 2E, asterisk; v) and white matter of the brainstem and cerebellum were spared (Stage III, Table 4, Figures 2F,G and 3F). Patients #1 and #8 showed marked atrophy and loss of axons in the frontal and temporal white matter, centrum semiovale (Figure 2H,I, arrows) and corpus callosum (Figure 2I, arrowhead; i) with severe dilatation of the lateral ventricles (Figure 2H,I). The thalamus (Figure 2I, arrowhead; ii), hippocampus and parahippocampal gyrus (Figure 2I, arrowhead; iii) also showed deterioration, but the internal capsule (Figure 2I, asterisk; iv) and optic tract (Figure 2I, asterisk; v) were relatively preserved. The cerebellar white matter, especially the posterior part, was degenerated (Figure 2K, arrow), but the pontine base (Figure 2J, asterisks; vi) and superior cerebellar peduncles (Figure 2J, asterisks; vii) were relatively well preserved (Stage IV, Table 4, Figure 3G).

Spherically swollen axons (spheroids and globules) in the cerebral white matter

Few swollen axons were observed in the control brains (Table 4, Figures 4Ai and 6).

Figure 4.

Immunohistologic examination on axons and microglia. A. Axonal loss and swollen axons in frontal white matter in control brains and each stage of ALSP and N‐HD. In control brains there were a few swollen axons (Ai). In ALSP there were relatively many swollen axons, 10 µm in diameter, in patches of axon loss in the cerebral white matter at Stage I (Aii, arrows). At Stage II, many swollen axons, 40 µm in diameter, were distributed diffusely in the cerebral white matter (Aiii, arrows). At Stages III and IV, the frontal white matter and centrum semiovale were degenerated. At Stage III there were a few swollen axons with a maximum diameter of 30 µm, whereas at Stage IV their maximum diameter was 15 µm (Aiv, arrows and Av). At Stages I and II, swollen axons in the centrum semiovale were immunopositive for synaptophysin (Insets in Aii and Aiii). In N‐HD, swollen axons with a maximum diameter of 25 µm were scattered in the frontal white matter and centrum semiovale at Stage II (Avi, arrows). At Stage III, a few swollen axons, with a maximum diameter of 18 µm, were present in the frontal white matter and centrum semiovale (Avii, arrow). At Stage IV, there were a few swollen axons 15 µm in maximal diameter (Aviii). However, the swollen axons were negative for synaptophysin. Ai–viii; Immunohistochemistry for p‐NF with periodic acid‐Schiff staining. Insets in Aii–iii; Immunohistochemistry for synaptophysin. Bars; 50 µm. B. IBA‐1‐ and/or CD163‐immunopositive cells in the frontal white matter in control brains and each stage of ALSP and N‐HD. In control brains, few or scattered IBA‐1‐immunopositive ramified cells and capillary‐associated cells were seen in the frontal cortex and white matter. About 5% of the IBA‐1‐immunopositive cells were also immunopositive for CD163 (Bi–ii). In ALSP, scattered IBA‐1‐immunopositive ramified cells and capillary‐associated cells were seen in the frontal cortex at Stage I. CD163 was immunopositive in about 5% of the cells (Biii). In the cerebral white matter at Stage I, many plump ramified and ameboid cells were evident, and CD163 was immunopositive in about 50% of them (Biv). At Stage II, many large ramified IBA‐1‐immunopositive cells were present in the frontal cortex, and CD163 was immunopositive in about 60% of them (Bv). In the cerebral white matter at the same stage, many ameboid IBA‐1‐immunopositive cells were observed, and almost all of them were CD163‐immunopositive (Bvi). At Stage III, scattered plump ramified cells with IBA‐1 immunopositivity were present in the frontal cortex, and about 50% of them were CD163‐immunopositive (Bvii). In the cerebral white matter at Stage III, many ameboid and plump ramified cells with IBA‐1 immunopositivity were evident, and about 80% of them were immunopositive for CD163 (Bviii). At Stage IV, many plump ramified cells with IBA‐1 immunopositivity were seen in the frontal cortex, and CD163 was immunopositive in about 5% of them (Bix). In the cerebral white matter at this stage, scattered IBA‐1‐immunopositive ameboid cells were observed, and about 80% of them were immunopositive for CD163 (Bx). CD163 was immunopositive in the cytoplam and proximal portion of the cell processes of ramified or ameboid cells (Biv–x). In N‐HD, many ramified IBA‐1‐immunopositive cells were seen in the frontal cortex at Stage II, and about 5% of them were immunopositive for CD163 (Bxi). In the cerebral white matter at the same stage, a few IBA‐1‐immunopositive ramified cells were present, and about 20% of them were immunopositive for CD163 (Bxii). At Stage III, many IBA‐1‐immunopositive plump ramified cells and capillary‐associated cells were evident in the frontal cortex, and about 5% of them were imunopositive for CD163 (Bxiii). In the cerebral white matter at Stage III, scattered IBA‐1‐immunopositive plump ramified and capillary‐associated cells were seen, and about 50% of them were immunopositive for CD163 (Bxiv). At Stage IV, many plump ramified cells and capillary‐associated cells with IBA‐1 immunopositivity were seen in the frontal cortex, and about 5% of the capillary‐associated cells were immunopositive for CD163 (Bxv). In the cerebral white matter at Stage IV, many IBA‐1‐immunopositive plump ramified cells were seen, and about 5% of them were immunopositive for CD163 (Bxvi). Double immunohistochemistry for Iba1 (brown) and CD163 (green). Bars; 20 µm. C. CD68‐immunopositive cells in the frontal white matter of control brains and each stage of ALSP and N‐HD. In control brains, CD68 was immunopositive in scattered small ramified cells with short processes and capillary wall‐forming cells (Ci). In ALSP, CD68 was immunopositive in large ameboid cells, and capillary wall‐forming cells. The density of CD‐immunopositive cells increased in the frontal cortex and white matter, being maximal at Stage II (Cii–v). In N‐HD, scattered CD‐68‐immunopositive small ameboid cells were present in the cerebral white matter and cortex (Cvi and Cviii–ix), but only patient #3 showed many medium‐sized ameboid cells in the cerebral white matter (Cvii). Immunohistochemistry for CD68. Bars; 50 µm.

ALSP

Patient #10, who was at Stage I, showed scattered swollen axons a few µm in diameter in the cerebral white matter and relatively many swollen axons, with a maximum diameter of 10 µm, were seen in patches of axon loss (Table 4, Figure 4Aii, arrows). Many swollen axons, with a maximum diameter of 40 µm, were diffusely evident in the centrum semiovale at Stage II, and their density was highest among the ALSP patients examined (Table 4, Figure 4Aiii, arrows). The frontal white matter and centrum semiovale at Stages III and IV showed devastation and a few swollen axons, their maximum diameter being 30 µm at Stage III and 15 µm at Stage IV (Table 4, Figures 4Aiv, arrows, 4Av and 6). At Stage II there were few swollen axons in the U‐fibers, but many were evident at Stages III and IV.

Swollen axons in the centrum semiovale at Stages I and II were immunopositive for synaptophysin (Table 4, Insets in Figure 4Aii,iii). Double immunohistochemistry for p‐NF and MBP revealed that many swollen axons were covered by myelin (Figure 5Ci).

Figure 5.

Immunohistochemical examination of CSF1R, DAP12, myelinated fibers and lymphocytic infiltration. a. CSF1R‐immunopositive cells in the frontal cortex and white matter in control brains and each stage of ALSP and N‐HD. In control brains, CSF1R‐immunopositive plump ramified or star‐shaped cells were numerous or scattered in the frontal white matter, but sparse in the frontal cortex (ai–ii). In ALSP, many CSF1R‐immunopositive star‐shaped or plump ramified cells were seen in the cerebral white matter, but were sparse in the frontal cortex at Stages I–IV (aiii–x). In N‐HD, CSF1R‐immunopositive ameboid or plump ramified cells were numerous or scattered in the cerebral white matter at Stages II–IV, but were absent or minimal in the frontal cortex at all stages (axi–xvi). a: Immunohistochemistry for CSF1R. Bars; 20 µm. b. DAP12‐immunopositive cells in the frontal cortex and white matter in control brains and each stage of ALSP and N‐HD. In control brains, DAP12 was immunopositive in many ramified and star‐shaped cells, blood cells and capillary‐associated cells in the frontal cortex, and in scattered ramified cells in the cerebral white matter (bi–ii). At Stages I, II and III of ALSP, many swollen ramified cells in the centrum semiovale were immunopositive for DAP12, and in the frontal cortex at Stage II, many DAP12‐immunopositive ramified and star‐shaped cells, blood cells and capillary wall‐forming cells were seen (biii–x). DAP12 was negative in N‐HD brains (bxi–ii and bxv–xviii) except in patient #3, who was at Stage III, and showed many DAP12‐immunopositive small glial cells in the cerebral white matter and blood cells in the frontal cortex (bxiii–xiv). b: Immunohistochemistry for DAP12. Bars; 20 µm. c. Myelinated swollen axons, perivascular lymphocytic infiltration in the lesions. In the cerebral white matter in ALSP, many swollen axons were covered by myelin (ci), and scattered perivascular T‐cell infiltration was evident at Stages II and III (cii). ci: Double immunohistochemistry for p‐NF (brown) and MBP (green). Bar; 20 µm, cii: Immunohistochemistry for CD3. Bar; 50 µm.

N‐HD

Swollen axons with a maximum diameter of 25 µm were scattered in the frontal white matter and centrum semiovale at Stage II, and the density was highest at the Stage (Table 4, Figure 4Avi, arrows). At Stage III, when their maximum diameter was 18 µm, were a few in the frontal white matter and centrum semiovale (Table 4, Figure 4Avii, arrow). At Stage IV, a few swollen axons with a maximal diameter of 15 µm were evident (Table 4, Figures 4Aviii and 6). Swollen axons in patients with N‐HD were not immunopositive for synaptophysin (Table 4), and their density was relatively higher in the U‐fibers than in the centrum semiovale at Stages II–IV.

Frontal cortex

No remarkable swollen axons or loss of neurons, axons were evident in the control brains (Table 4, Figure 6).

Figure 6.

Dynamic changes of axons and microglial subsets along the passage of stages in ALSP and N‐HD. The severity of the neuropathological findings in Figures 4 and 5 and Table 4 was graded semiquantitatively based on the severity of loss of findings or the frequency of findings into four using bars with different thickness and blank space. “Slight or a few” was expressed as thin bars, “moderate or scattered” by bars of moderate thickness, and “severe or many” by thick bars. Blank spaces indicate absence of the cells examined. Dotted lines with question marks indicate non‐examined items due to absence of available autopsy cases. Patient at Stage I showed scattered swollen axons in the cerebral white matter without remarkable loss of axons and relatively many swollen axons were seen in patches of axon loss. The finding might indicate axon swelling precedes axon loss. The number and size of swollen axons in the cerebral white matter were much greater in ALSP than in N‐HD, and were maximal at Stage II of ALSP. This feature roughly corresponds with the alteration of microglial cells in ALSP, whereas this tendency is not evident in N‐HD. Alteration of microglia, especially an increased number of CD68‐immunopositive cells in the cerebral white matter, may precede axonal alteration in ALSP. Increase of IBA‐1‐, CD163‐ and CD204‐immunopositive cells precedes loss of axons in ALSP. Number of CSF1R‐immunopositive cells does not decrease in ALSP cerebral white matter. The number and size of swollen axons are small in N‐HD, though severe loss of axons occurs in both N‐HD and ALSP. Microglial changes are far less evident in N‐HD than in ALSP. N‐HD case #3 showing relatively many CD68‐ and DAP12‐immunopositive cells compared with other patients at Stage III, perhaps indicating that the patient suffered from mutation of TREM2 or single‐base substitution of DAP12 as reported by Sasaki et al. 72.

ALSP

Swollen axons measuring up to 15 µm in diameter were few in the cortex at Stages I and II, scattered at Stage III and numerous at Stage IV. The cerebral cortex showed no remarkable loss of neurons, although the cortex showed marked thinning at Stage III, and loss of neurons in the frontal cortex was evident only at Stage IV. Synaptophysin immunoreactivity showed no marked alteration in the frontal cortex at Stages I to III, but was decreased in the cortex at Stage IV. Scattered p‐NF‐immunopositive ballooned neurons were evident in the frontal cortex at Stages I–III, but were numerous at Stage IV (Table 4, Figure 6).

N‐HD

No marked neuronal loss was observed in the frontal cortex of patients with N‐HD, but a few swollen axons and p‐NF‐immunopositive ballooned neurons were evident (Table 4, Figure 6).

Double immunohistochemistry for IBA1 and CD163

In control brains, a few or scattered IBA‐1‐immunopositive ramified cells and capillary‐associated cells were seen in the frontal cortex and cerebral white matter, and about 5% of the IBA‐1‐immunopositive cells were immunopositive for CD163 (Table 4, Figures 4Bi,ii and 6).

ALSP

At Stage I, scattered ramified cells and capillary‐associated cells were immunopositive for IBA‐1 in the frontal cortex. CD163 was immunopositive in about 5% of those cells (Figure 4Biii). In the cerebral white matter, many IBA‐1‐immunopositive plump ramified and ameboid cells were evident, and CD163 was immunopositive in about 50% of them (Table 4, Figure 4Biv). At Stage II, many large ramified cells were immunopositive for IBA‐1 in the frontal cortex, and CD163 was immunopositive in about 60% of them (Figure 4Bv). In the cerebral white matter, many ameboid IBA‐1‐immunopositive cells were present, and almost all of them were CD163‐immunopositive (Table 4, Figure 4Bvi). At Stage III, scattered plump ramified cells with IBA‐1 immunopositivity were evident in the frontal cortex, and about 50% of them were also CD163‐immunopositive (Figure 4Bvii). In the cerebral white matter, many ameboid and plump ramified cells with IBA‐1 immunopositivity were evident, and about 80% of them were also immunopositive for CD163 (Figure 4Bviii). At Stage IV, many plump ramified cells were immunopositive for IBA‐1 in the frontal cortex, and about 5% of them were also immunopositive for CD163 (Figure 4Bix). The cerebral white matter contained scattered IBA‐1‐immunopositive ameboid cells, of which about 80% were also immunopositive for CD163 (Figure 4Bx). CD163 was immunopositive in the cell cytoplasm and proximal portion of the cell processes of ramified or ameboid cells (Table 4, Figure 4Biv–x).

N‐HD

At Stage II, many ramified cells were immunopositive for IBA‐1 in the frontal cortex, and CD163 was immunopositive in about 5% of them (Figure 4Bxi). In the cerebral white matter, a few IBA‐1‐immunopositive ramified cells were seen, and about 20% of them were immunopositive for CD163 (Figure 4Bxii). At Stage III, many IBA‐1‐immunopositive plump ramified cells and capillary‐associated cells were evident in the frontal cortex, and about 5% of them were also imunopositive for CD163 (Figure 4Bxiii). In the cerebral white matter, scattered plump ramified and capillary‐associated cells were immunopositive for IBA‐1, and about 50% of them were also immunopositive for CD163 (Figure 4Bxiv). At Stage IV, many plump ramified cells and capillary‐associated cells with IBA‐1 immunopositivity were evident in the frontal cortex, and about 5% of them, which were capillary‐associated cells, were immunopositive for CD163 (Figure 4Bxv). In the cerebral white matter, many IBA‐1‐immunopositive plump ramified cells were present, of which about 5% were also immunopositive for CD163 (Table 4, Figure 4Bxvi).

Double immunohistochemistry for CD163 and CD204

In control subjects, a few CD204‐immunopositive ramified cells were observed in the cerebral white matter, but such cells were absent in the frontal cortex. Many of these cells were also immunopositive for CD163 (Table 4).

ALSP

At Stage I, CD204 was immunopositive in many ameboid cells, and in a few ramified and capillary‐associated cells in the cerebral white matter and frontal cortex (Table 4). At Stages II and III, many swollen ameboid and ramified cells in the cerebral white matter and a few/scattered ramified or capillary‐associated cells in the frontal cortex were immunopositive for CD204 (Table 4). At Stage IV, CD204 immunopositivity was evident in scattered ameboid cells in the cerebral white matter, and in many/scattered swollen ramified cells in the frontal cortex (Table 4). Most of these cells were also immunopositive for CD163.

N‐HD

A few CD204‐immunopositive ramified, ameboid or capillary‐associated cells were present in the cerebral white matter and frontal cortex (Table 4), and most of them were immunopositive for CD163.

Immunohistochemistry for CD68

CD68 was immunopositive in scattered small ramified cells with short processes and capillary‐associated cells in the control brains (Table 4, Figures 4Ci and 6).

ALSP

CD68‐immunopositive ramified or capillary‐associated cells were scattered or sparsely present in the frontal cortex at every stage. CD68 immunopositivity was evident in large ameboid or ramified cells in the cerebral white matter, being scattered at Stages I and III, numerous at Stage II and few at Stage IV (Table 4, Figures 4Cii–v and 6).

N‐HD

Scattered CD68‐immunopositive small ameboid and ramified cells were present in the cerebral white matter and cortex at Stages II–IV (Table 4, Figures 4Cvi, viii–ix and 6), but many medium‐sized ameboid cells in the cerebral white matter were evident only in patient #3 at Stage III (Table 4, Figures 4Cvii and 6).

Immunohistochemistry for CSF1R

Many/scattered CSF1R‐immunopositive ramified or star‐shaped cells were observed in the cerebral white matter of the control brains, but were absent in the frontal cortex (Table 4, Figures 5Ai–ii and 6).

ALSP

At Stages I–IV, many CSF1R‐immunopositive star‐shaped or ramified cells were evident in the cerebral white matter, and a few were seen in the frontal cortex (Table 4, Figures 5Aiii–x and 6).

N‐HD

At Stages II–IV, many/scattered CSF1R‐immunopositive ameboid or ramified cells were evident in the cerebral white matter, but were absent or only sparse in the frontal cortex (Table 4, Figures 5Axi–xvi and 6).

Immunohistochemistry for DAP12

In control brains, DAP12 was immunopositive in many ramified and star‐shaped cells, blood cells and capillary‐associated cells in the frontal cortex, and in scattered ramified cells in the cerebral white matter (Table 4, Figures 5Bi–ii and 6).

ALSP

Many ramified and star‐shaped cells, blood cells and capillary‐associated cells were DAP12‐immunopositive in the frontal cortex at every stage. Many DAP12‐immunopositive swollen ramified cells were evident in the centrum semiovale at Stages I, II and III, but were few at Stage IV (Table 4, Figures 5Biii–x and 6).

N‐HD

DAP12 was negative in N‐HD brains at Stages II and III (Figure 5Bxi,xii,xv–xviii) except for patient #3, who was at Stage III and showed DAP12 immunopositivity in many small glial cells in the cerebral white matter and blood cells in the frontal cortex (Table 4, Figures 5Bxiii–xiv and 6). A few capillary‐associated cells and glia were positive at Stage IV of N‐HD (Figures 5Bxvii–xviii and 6).

Immunohistochemistry for TREM2

In control brains, TREM2 was immunopositive in a few to many neurons in the cerebral cortex, and in blood cells and capillary wall‐forming cells in the cerebral white matter (Table 4).

ALSP

A few neurons, blood cells or capillary‐forming cells were TREM2‐immunopositive in the cerebral white matter and frontal cortex at each stage (Table 4).

N‐HD

TREM2 was negative in the brain at Stage II, but immunopositive in scattered or a few neurons, blood cells or capillary‐forming cells in the cerebral white matter and frontal cortex at Stages III and IV (Table 4).

Oligodendroglial cells

In both ALSP and N‐HD, the density of oligodendroglial cells was decreased slightly in the frontal white matter at Stage III, but marked there at Stage IV.

Immunohistochemistry for glial fibrillary acidic protein

In control brains, scattered glial fibrillary acidic protein (GFAP)‐immunopositive protoplasmic, fibrillary and perivascular cells were observed in the cerebral white matter and frontal cortex (Table 4).

ALSP

Many GFAP‐immunopositive large reactive protoplasmic cells were evident in the cerebral white matter at Stage III, as demonstrated using immunohistochemistry for alpha‐B‐crystallin by Baba et al. 7. Scattered GFAP‐immunopositive protoplasmic or perivascular cells were present in the frontal cortex (Table 4).

N‐HD

Many GFAP‐positive fibrillary cells were evident in the cerebral white matter at Stages II–IV. Scattered GFAP‐immunopositive protoplasmic and perivascular cells were seen in the frontal cortex (Table 4).

Lymphocytic infiltration

Scattered T‐cells had infiltrated the perivascular space in the cerebral white matter at Stages II and III of ALSP (Figure 5Cii), but no B‐cell infiltration was evident. No remarkable lymphocytic infiltration was seen in the control and N‐HD brains.

Other neuropathological findings including aging‐related features

No significant neuropathological findings such as tau‐immunopositive structures, senile plaques, p‐alpha synuclein‐positive structures, p‐TDP‐43‐inclusions or infarcts were observed in the brains of patients with ALSP and N‐HD.

Correlation between lesion stages and clinical symptoms

By modified Rankin scale (mRS) 84 in ALSP, no neurological symptoms (mRS 0) at Stage I, mild to severe cognitive impairment and/or disability (mRS 3–5) at Stage II, but severe disability or apathetic state (mRS 5) at Stage III were seen. In N‐HD, moderate severe disability (mRS 4) at Stage II and severe disability or apathetic state (mRS 5) at Stages III and IV were observed.

In the present series examined, epilepsy occurred in the patients at Stages III and IV, but not in the Stages I or II of ALSP. In N‐HD, epilepsy examined only one case at Stage II, but seven at Stages III and IV.

DISCUSSION

In ALSP and N‐HD, the staging and progression pattern of lesions based on the severity and area of axon loss are considered to be divisible into four, though in the present series the total number of examined cases was limited and no autopsy cases of Stage I N‐HD were available. The severity of loss of axons, frequency and size of swollen axons and changes in shape and subsets of microglia are diverse among patients with ALSP and N‐HD. Staging of the lesions makes these diversity integrated and classified clearly. The analysis of characters of lesions in each mutation of causative genes is a problem left for future.

Briefly, the stages of ALSP are as follows: Stage I, myelinated fiber loss occurs in a scattered patchy manner in the centrum semiovale without atrophy of cerebral white matter; Stage II, large patches in the cerebral white matter with preservation of U‐fibers and slight atrophy of the cerebral white matter are evident; Stage III, marked degeneration of the cerebral white matter with cerebral atrophy and dilation of the lateral and third ventricles, but preservation of the cerebellum and transverse fibers in the pontine base; Stage IV, devastation of the cerebral white matter, degeneration of cerebellar white matter and myelinated fibers in the pontine base is evident (Table 4, Figure 1).

In N‐HD, initiation of myelinated fiber loss was not clarified because no brain specimens at Stage I without neurological symptoms were available. The other stages of N‐HD are as follows: Stage II, slight atrophy of the cerebrum and irregular patches of myelinated fiber loss in the cerebral white matter with moderate dilatation of the lateral ventricles; Stage III, moderately atrophic cerebrum with degeneration of the cerebral white matter and dilated lateral and third ventricles. Good preservation of the cerebellar white matter and pontine base; Stage IV, marked atrophy of the brain with devastation of the cerebral white matter, severe dilatation of the lateral ventricles and deterioration of the cerebellar white matter, but relative preservation of the internal capsule and pontine base (Table 4, Figure 2).

Chronological lesion progression pattern in ALSP was elucidated by observation of two sisters (ALSP patients #6 and #10) 70 and in other patients by brain MRI 35, 78. Thus, differences in the severity and area of brain white matter degeneration in each patient were not due to differences in mutated genes, but to lesion progression. Attending clinicians should pay attention to small patchy lesions in the cerebral white matter of healthy or demented middle‐aged adults, since ALSP caused by CSF1R mutation may occur occasionally in solitary (de novo) patients 36, 39.

Onset of ages was generally younger in N‐HD than in ALSP. As the Stage progressed, the age at death tended to be high and disease course to be long. However, there was difference among individuals, as ALSP #4 had 16 years duration and #5, 13 years at Stage III, but ALSP #7 at Stage IV had only 6 years duration. Further examination will be needed to elucidate the factor regulating the speed of lesion progression. With progression of the Stage, clinical symptoms became worse to apathetic state, and epilepsy may be induced at the Stages of severe destruction of the cerebral white matter in both diseases. Microglial phenotype changes in ALSP and N‐HD may be linked to depression and/or anxiety observed in patients with these diseases, as reported recently 11, 69.

As for neuropathological differentiation between ALSP and N‐HD, the internal capsule and pontine base are relatively well preserved in the latter, even at Stage IV, whereas the hippocampus is relatively well preserved in ALSP in comparison to N‐HD. Well preservation of the convolutional white matter of the motor cortex, internal capsule and pyramidal tracts is a remarkable character of the N‐HD patients. The further examination seems to be necessary to elucidate the mechanism. ALSP shows more evident microglial alteration, especially increased numbers of cells immunopositive for CD68, CD163, CD204 and DAP12, than N‐HD. Regarding the neuropathological differentiation of patients with DAP12 or TREM2 in N‐HD, N‐HD #3 represented increased number of DAP12‐ and CD68‐immunopositive cells in the lesions (Table 4, Figures 4, 5, 6). The finding may show the patient suffered from mutation of TREM2 or single‐base substitution of DAP12 as reported by Sasaki et al. 72.

IBA‐1 has been shown to be a calcium‐binding protein specific for microglia 1, 26, and its function has been considered to be essential for microglial activation 23. CD163 is a hemoglobin scavenger receptor and a macrophage‐specific protein, and upregulated expression of the receptor is one of the major changes occurring in macrophage activation 15, 51. CD204 has been reported to be a myeloid cell marker and to be a macrophage scavenger receptor (scavenger receptor A: SR‐A), and plays roles in propagation of lesions in various diseases such as inflammation, sepsis and ischemic injury 32, 51. CD68 is a macrophage/monocyte/neutrophil/basophil marker 51. Macrophage colony stimulating factor (M‐CSF) has been reported to be secreted by T cells and to maintain the macrophage population through combination with its protein tyrosine kinase receptor, CSF1R 48.

On the microglial activation, the present study revealed that microglia were swollen, and considered to be activated in the cerebral white matter in ALSP, especially at Stages II and III. Actually in those Stages, number of immunopositive cells for IBA‐1, CD163, CD68 and DAP12 increased. From the findings, CSF1R was also considered to be increased in expression, even if the molecular structure was transformed by mutation of CSF1R, and immunohistochemistry showed that the number of CSF1R‐immunopositive cells in the cerebral white matter was preserved in ALSP.

It had been reported that CSF1R gene mutations in cultured cells cause loss of signal transduction function 20, 66. However, immunohistochemistry has demonstrated mutant CSF1R protein in cultured cells, and its internal localization from the cell surface has been reported 20. Accordingly it is considered that mutation of the CSF1R gene does not inevitably lead to depletion of CSF1R protein in the brains of patients with ALSP. From this viewpoint, there is a possibility that immunohistochemistry for CSF1R would yield different results, even in the same autopsy case, according to the method employed.

Immunoblotting and immunohistochemistry of the ALSP brain using a polyclonal antibody against CSF1R (Santa Cruz Biotechnology) 70 or a polyclonal antibody against the C‐terminal of CSF1R (C20, Santa Cruz Biotechnology) 39 have yielded negative results, but in the present study, immunohistochemistry employing a rabbit polyclonal antibody against CFS1R from LifeSpan BioSciences, USA, was used and boosting of immunoreactivity using protease K, the EnVision^TM Dual link System and ImmPACT^TM DAB Chromogen Concentrate yielded positive results. In general, a negative result of immunohistochemistry using a particular antibody does not necessarily indicate absence of material bearing the epitope. The authors think that the present data would be useful for other researchers using their own antibodies or methods.

On the correlation between axonal alteration and microglial response, axonal swellings occurring at Stage I and being maximal at Stage II may be a degenerative reaction of axons due to inflammatory processes. Axonal damages may occur simultaneously to the axonal swelling after the Stage I and maximal at Stage II. After the Stages, damaged axons eliminated by the phagocytic cells, and to be devastated state of the Stage IV in ALSP. Inflammatory process in the cerebral cortex occurred one Stage later than in the white matter. Activation and inflammatory reaction demonstrated by immunohistochemistry for markers of the microglia were far weaker in N‐HD than ALSP, and axonal swelling was smaller in number and size than ALSP. Severe loss of axons in N‐HD similar to that in ALSP may be induced by different mechanism from N‐HD.

Patient at Stage I showed scattered swollen axons in the cerebral white matter without remarkable loss of axons and relatively many swollen axons were seen in patches of axon loss. The finding might indicate axon swelling precedes axon loss. The number and size of swollen axons in the cerebral white matter were much greater in ALSP than in N‐HD, and were maximal at Stage II of ALSP. This feature roughly corresponds with the alteration of microglial cells in ALSP, whereas this tendency is not evident in N‐HD. Alteration of microglia, especially an increased number of CD68‐immunopositive cells in the cerebral white matter, may precede axonal alteration in ALSP. Increase of Iba1‐, CD163‐ and CD204‐immunopositive cells precedes loss of axons in ALSP (Table 3, Figures 4 and 6). The existence of many swollen axons covered by myelin may indicate that axon swelling occurs earlier than loss of myelin or oligodendroglia as reported by Lin et al. 42, Yazawa et al. 87 and Kinoshita et al. 35. Presence of synaptophysin‐immunopositive swollen axons observed in ALSP may indicate a stasis of axonal flow, as reported by Baba et al. 7 observing amyloid precursor protein in the swollen axons in ALSP. U‐fibers are considered to be involved later than centrum semiovale both in ALSP and N‐HD.

It has been reported that CSF1R and CD163 locate on the anti‐inflammatory M2 cells but not on pro inflammatory M1 cells 44, 48. Alzheimer's disease (AD) brains show a robust microglial inflammatory response. It has been reported that CD163‐immunopositive microglia are located in the centers of senile plaques, but are not associated with neurofibrillary tangles 65. Many CD68‐immunopositive cells appear in AD brains from the early phase 8. In an AD mouse model, the APP + PS1 mouse, a distinctive shift from M2 to M1 activation in the brain has been observed 29. Parkinson's disease is characterized by loss of dopaminergic neurons and the presence of alpha‐synuclein‐containing aggregates in the substantia nigra, as well as chronic inflammation 85. In the substantia nigra it had been reported that alpha‐synuclein deposition is correlated with major histocompatibility complex (MHC) Class II‐immunoreactive M1 cells, and that the disease duration is short in patients with a large number of CD68‐immunoreactive cells in the substantia nigra 14.

In the white matter lesions of multiple sclerosis (MS), M1 microglia produce pro‐inflammatory cytokines leading to an increase of demyelination, whereas M2 microglia produce anti‐inflammatory cytokines and participate in synapse repair and remodeling 13. MHCII‐immunoreactive M1 cells are located at the edges of the lesions, which constitute the front line of the disease 68, or MHCII microgliaparticipate in remyelination 59. Microglia can switch between M1 and M2 phenotypic profiles in the MS lesions 64. M1‐related genes increased from 7 days postinfection (dpi) up to 98 dpi, then decreased, but M2‐related genes increased continuously from 14 dpi up to 196 dpi in Theiler's murine encephalomyelitis 22.

In the spinal cords of ALS mice, which have a C57Bl/6 genetic background and overexpress the G93A mutation in the Cu²⁺/Zn²⁺ superoxide dismutase gene (mSOD1), and are a transgenic animal model of inherited human ALS, CD68‐immunopositive cells gradually increase from the middle to the end‐stage 9, but CD163 mRNA expression is decreased in the end‐stage 41. Thus, as the disease accelerates, a shift occurs from beneficial immune responses (involving M2 microglia) to deleterious immune responses (involving M1 microglia) 88.

In the white matter lesions of ALSP observed in the present study, CD68‐immunopositive cells appear in the early phase, as is the case in AD 8, whereas CD163‐immunopositive cells increase in lesions at the middle stage and then decrease once destruction is extensive, unlike the features in AD 65. This pattern of progression of ALSP also differs from that in PD, ALS mice and Theiler' murine encephalomyelitis.

As to the mechanism of lesion progression in ALSP, CD68‐immunopositive cells appear before axonal swelling or loss, suggesting that the cells are a type of lesion initiator as well as being responsible for phagocytosis. In lesions at the middle stage, CD163‐immunopositive M2 cells are activated, possibly playing a role in immunosuppression. Overall, the microglial alteration observed in the white matter lesions of ALSP is a unique feature in comparison with the other degenerative or inflammatory diseases described above.

Conclusively, the present study provide that (i) the lesion progression pattern is similar in both ALSP and N‐HD, and the lesion severity can be divided into four stages, and that (ii) ALSP shows larger and more numerous swollen axons and more severe microglial changes than is the case in N‐HD. (iii) Microglia dynamically changed in shape, density and subsets along the passage of the stages in both diseases. (iv) Increase of IBA‐1‐, CD68‐, CD163‐ and CD204‐immunopositive cells precedes loss of axons in ALSP. (v) With progression of the Stage, clinical symptoms became worse to apathetic state, and epilepsy induced at the stages of severe destruction of the cerebral white matter in both diseases.