Clinical and Pathological Features of FTDP‐17 with MAPT p.K298_H299insQ Mutation

Abstract

Background

MAPT is a causative gene in frontotemporal dementia with parkinsonism linked to chromosome 17 (FTDP‐17), a hereditary degenerative disease with various clinical manifestations, including progressive supranuclear palsy, corticobasal syndrome, Parkinson's disease, and frontotemporal dementia.

Objectives

To analyze genetically, biochemically, and pathologically multiple members of two families who exhibited various phenotypes of the disease.

Methods

Genetic analysis included linkage analysis, homozygosity haplotyping, and exome sequencing. We conducted tau protein microtubule polymerization assay, heparin‐induced tau aggregation, and western blotting with brain lysate from an autopsy case. We also evaluated abnormal tau aggregation by using anti‐tau antibody and PM‐PBB3.

Results

We identified a variant, c.896_897insACA, p.K298_H299insQ, in the MAPT gene of affected patients. Similar to previous reports, most patients presented with atypical parkinsonism. Biochemical analysis revealed that the mutant tau protein had a reduced ability to polymerize microtubules and formed abnormal fibrous aggregates. Pathological study revealed frontotemporal lobe atrophy, midbrain atrophy, depigmentation of the substantia nigra, and four‐repeat tau‐positive inclusions in the hippocampus, brainstem, and spinal cord neurons. The inclusion bodies also stained positively with PM‐PBB3.

Conclusions

This study confirmed that the insACA mutation caused FTDP‐17. The affected patients showed symptoms resembling Parkinson's disease initially and symptoms of progressive supranuclear palsy later. Despite the initial clinical diagnosis of frontotemporal dementia in the autopsy case, the spread of lesions could explain the process of progressive supranuclear palsy. The study of more cases in the future will help clarify the common pathogenesis of MAPT mutations or specific pathogeneses of each mutation.

Microtubule‐associated protein tau (MAPT), encoding tau protein, is recognized as the causative gene in frontotemporal dementia with parkinsonism linked to chromosome 17 (FTDP‐17). Numerous MAPT mutations have been reported to be responsible for FTDP‐17. Abnormal tau deposition occurs in neurodegenerative disorders such as Alzheimer's disease, progressive supranuclear palsy (PSP), corticobasal syndrome, and frontotemporal dementia (FTD). Tau proteins contain three or four microtubule‐binding domains, called three‐repeat tau or four‐repeat tau, respectively. The expression levels of three‐repeat and four‐repeat tau are almost equivalent in healthy individuals. Abnormally deposited tau varies by the disease: both three and four‐repeat tau accumulate in Alzheimer's disease, only three‐repeat tau in Pick disease, and only four‐repeat tau in PSP and corticobasal syndrome. Whereas, frontotemporal lobar degeneration is caused not only by FTDP‐17, but also by various other causes including PSP, corticobasal degeneration, and mutations in TARDBP and FUS. The neurological manifestations caused by the MAPT mutations are extremely diverse, with symptoms of PSP, corticobasal syndrome, Parkinson's disease, FTD, and globular glial tauopathy. ¹ Thus, MAPT mutations are known to cause various clinical symptoms and pathological characteristics, and the study of more cases is necessary for understanding the pathogenesis of FTDP‐17.

Case Series

We studied two families with members diagnosed with Parkinson's disease, PSP, or FTD (Table S1). In patients in whom we could perform neurological examination, the diagnoses were based on diagnostic criteria used internationally by neurologists. ² , ³ , ⁴ , ⁵ We obtained written informed consent from all participants. This study was approved by the Ethical Committee for Human Genome Research of Hiroshima University.

Family 1 had six affected members over two generations: three were diagnosed with Parkinson's disease, two with PSP, and one with FTD (Fig. 1A). The proband, 1‐III‐6, was a woman whose mother had received a diagnosis of Parkinson's disease. The proband experienced tremors in both upper limbs at age 58 and received 600 mg of levodopa per day. Although the tremors were reduced, gait disturbance emerged, and by age 64, she was falling frequently. She also exhibited writing difficulties, neck flexion, and dysarthria. Neurological findings included up‐ and downgaze palsy, positive oculocephalic reflex, marked head drop, trunk and limb rigidity, righting reflex disorder, short‐stepped gait, and retropulsion. She was taking 600 mg of levodopa, 375 μg of pergolide, and 150 mg of amantadine per day. The dose of levodopa was reduced to 300 mg due to worsening of head drop. Imaging showed preserved cerebrum with mild atrophy of the hippocampus and tegmentum (Fig. S1A,B). She received a clinical diagnosis of PSP.

Figure 1.

Pedigree chart and genetic analysis. (A) Pedigree chart of family 1. Solid circles and squares signify affected women and men, respectively, and open circles and square signify unaffected women and men, respectively. Arrows indicate probands; numbers in parentheses indicate ages at onset of symptoms. Six family members over two generations were affected: three with PD, two with PSP, and one with FTD. (B) Pedigree chart of family 2. Nine members were affected over two generations: five with PD, two with PDD, and two with PSP. (C) Results of linkage analysis. The maximum logarithm of odds was 1.2039. The box indicates results of linkage analysis of chromosome 17. (D) Magnification of linkage analysis of chromosome 17 (Chr17). MAPT (red arrow) was present in a relatively long region with a high logarithm of odds. (E) Homozygosity haplotyping results. Haplotype regions shared by affected patients were found on chromosomes 13 and 17. MAPT (red arrow) was present in the candidate region of chromosome 17. (F) Results of Sanger sequencing. A MAPT variant (c.896_897insACA, p.K298_H299insQ) detected by exome sequencing was confirmed. (G) The mutation in MAPT was present in the second microtubule‐binding domain; the second microtubule‐binding domain is indicated by a red box and the others by yellow. Since the second microtubule‐binding domain is only present in four‐repeat tau, the inserted glutamine affected only four‐repeat tau, not three‐repeat tau. BC, breast cancer; CHF, chronic heart failure; FTD, frontotemporal dementia; LC, lung cancer; PD, Parkinson's disease; PDD, Parkinson's disease with dementia; PSP, progressive supranuclear palsy; SC, stomach cancer.

An older brother, 1‐III‐4, began exhibiting upper limb tremor at age 56. Intentional tremor and gait disturbance appeared at age 58. Subsequently, short‐stepped gait and stooped posture became evident, and cognitive dysfunction, compulsive behavior, a mask‐like expression, and muscle rigidity developed. The patient had severe cognitive decline. Neurological abnormalities included up‐ and downgaze palsy, dysarthria, dysphagia, loss of pharyngeal reflex, mild muscle weakness, advanced cervical muscle rigidity, cogwheel rigidity of limbs, resting and intention tremor, hyperreflexia, positive pathological reflexes in both legs, and frontal lobe signs. At age 61, he was taking 600 mg of levodopa and his symptoms had mildly improved. Imaging revealed atrophy of the frontotemporal lobe and tegmentum and marked atrophy of the hippocampus (Fig. S1C–E). He received a clinical diagnosis of PSP.

A younger sister, 1‐III‐9, suffered from gait disturbance at age 56. Later, she exhibited tremor and muscle rigidity. She received a diagnosis of Parkinson's disease at age 57 and began taking levodopa. At age 58, she suddenly lost the abilities to drive a car and move her legs properly. Memory impairment occurred at age 60 and foot dystonia at age 61. From the early stage of the disease, apathy, loss of sympathy, and deficits in executive tasks were observed, but there were no inappropriate behaviors observed. With 1000 mg of levodopa, symptoms improved mildly. After that, her activities of daily living deteriorated rapidly, and she became fully dependent on other people for care. Neurological examination at age 62 showed facial grimacing, hypertonia, grasping reflex, and positive Babinski sign. Head computed tomography and magnetic resonance imaging (MRI) showed marked atrophy of the frontal and temporal lobes. In the same year, akinetic mutism developed, and FTD was clinically diagnosed. At 64, she died of acute intestinal peristalsis and respiratory failure, and an autopsy was performed.

Family 2 had nine affected individuals over two generations: five were diagnosed with Parkinson's disease, two with Parkinson's disease with dementia (PDD), and two with PSP (Fig. 1B). The proband, 2‐III‐8, was a man whose mother were diagnosed Parkinson's disease. Gait disturbance was his initial symptom, at age 58. He later exhibited a mask‐like expression, dysarthria, cogwheel rigidity of the left upper limb, tremor of the left upper and lower limbs, stooped posture, and righting reflex disorder. Head MRI revealed general brain atrophy and diffuse T2 hyperintensity in the deep white matter. Iodine‐123 iodoamphetamine single‐photon emission computed tomography showed decreased blood flow in the frontal and temporal lobes. At age 63, up‐ and downgaze palsy, short‐stepped and freezing of gait, and festination developed. Although 300 mg of levodopa alleviated symptoms somewhat, the wearing‐off effect was prominent. Imaging at age 65 showed mild midbrain atrophy. At 69, his symptoms worsened, with severe rigidity, increased cervical muscle tone, forced grasping, hyperreflexia, and positive pathological reflexes, making speech and body movement remarkably difficult. He was taking 300 mg of levodopa, 2.5 mg of pramipexole, 7.5 mg of selegiline, and 300 mg of entacapone with little improvement. He received a clinical diagnosis of PSP.

In a younger sister, 2‐III‐10, clumsiness was the initial symptom, at age 56. At age 57, gait disturbance developed, and she started taking levodopa. At age 58, she received 300 mg of levodopa, 1.5 mg of pramipexole, and 5 mg of selegiline per day with mild improvement of symptoms. Neurologically, the patient exhibited upward gaze palsy, nonfluent speech, and hyperreflexia, alongside bradykinesia and rigidity. Subsequently, her symptoms progressed gradually; cognitive dysfunction, behavioral disinhibition, apathy, writing disorder, dysarthria, dysphagia, and severe tremor appeared at age 59. Based on her clinical symptoms, her diagnosis was considered to be PSP. Dysphagia worsened at age 60; dyskinesia, dysuria, and trismus were present at age 61. Because of frequent pneumonia, she underwent gastrostomy, and tube feeding was started. She died of pneumonia at age 65.

We performed genetic analyses on these patients. The candidate region obtained from linkage analysis and homozygous haplotyping included chromosome 17, which contains the MAPT gene (Fig. 1C–E). From exome sequencing and segregation analysis of family 1, we identified a causative variant, MAPT (NM_005910):c.896_897insACA,p.K298_H299insQ (Table S2). Exome sequencing of a patient from another family identified the same MAPT variant, cosegregating with the disease. Variant confirmation via Sanger sequencing (Fig. 1F) showed that it belonged to the second microtubule‐binding domain (Fig. 1G). Details of the genetic and biochemical analyses are provided in the Supplementary Materials (Supplemental Methods, Supplemental Data, Fig. S2–S5).

Pathological Findings

Postmortem examination was performed 6 hours after the death of patient 1‐III‐9. The brain weighed 800 g. Frontal and temporal lobe atrophy, midbrain atrophy (Fig. 2A–C), and depigmentation of the substantia nigra were observed macroscopically (Fig. 2C). Histological examination revealed severe cortical neuronal loss in the frontal and temporal lobes. In other lesions, including those in the anterior horn of the spinal cord and in the brainstem, the remaining neurons had numerous basophilic inclusions in the cytoplasm (Fig. 2D). In the internal segment of the globus pallidus, basophilic cytoplasmic inclusions were scattered (Fig. 2E). These inclusions stained positively with Gallyas–Braak silver (Fig. 2F). However, few neurofibrillary tangles and senile plaques were observed. Immunostaining also revealed that these inclusions were positive for AT8 (Fig. 2G) and RD4 (Fig. 2H) but not for RD3 (Fig. 2I). Cytoplasmic vacuoles around the inclusions were visualized with AT8 (Fig. 2G). RD4‐positive neurons were also found in the anterior horn of the spinal cord (Fig. 2J). Electron microscopy revealed many cytoplasmic vacuoles (Fig. 2K). In between cytoplasmic vacuoles, neurofibrillary tangles were scattered (Fig. 2L). Immunostaining with AT8 and fluorescence staining with PM‐PBB3 (Fig. 3) revealed AT8‐positive neuronal inclusion bodies in the midbrain and medulla oblongata, and the inclusions stained positively with PM‐PBB3. The inclusion bodies had a heterogeneous inside structure and were arranged in a honeycomb‐like pattern. Tufted astrocyte‐like structures, also AT8‐positive, were found in the thalamus. In the subthalamic nucleus, AT8‐positive tau inclusions were found in neurons, and coiled body–like structures thought to be AT8‐positive tau inclusions in oligodendroglia were observed.

Figure 2.

Neuropathological autopsy findings of the patient with frontotemporal dementia (1‐III‐9). (A) Frontal and temporal lobe atrophy was observed. (B) The bilateral internal segments of the globus pallidus exhibited atrophy and pigmentation. (C) Midbrain atrophy and dyspigmentation were observed. (D) The cytoplasm of the remaining neurons had numerous basophilic inclusions in the nuclei of the brainstem. (E) The cytoplasmic basophilic inclusions were also scattered in the globus pallidus. (F) Gallyas–Braak silver staining revealed numerous neuronal cytoplasmic inclusions and oligodendrocytic coiled bodies. However, few neurofibrillary tangles and senile plaques were observed. (G) The cytoplasmic inclusions stained positively with phosphorylated tau (AT8). Immunostaining also revealed AT8‐positive cytoplasmic vacuoles around inclusions. (H) The inclusions were positive for four‐repeat tau antibody. (I) The inclusions did not have three‐repeat tau immunoreactivity. (J) The inclusions of four‐repeat tau were also found in the anterior horn of the spinal cord. (K) Electron microscopy showed that neurons had many cytoplasmic vacuoles. (L) Neurofibrillary tangles were scattered in between cytoplasmic vacuoles.

Figure 3.

Immunostaining with AT8 and fluorescent staining with PM‐PBB3. AT8‐positive tau inclusion bodies were found in neurons of the midbrain and medulla. The inclusion bodies had an affinity for PM‐PBB3 and were arranged in a distinct honeycomb‐like pattern. Tufted astrocyte‐like structures were found in the thalamus. In the subthalamic nucleus, intraneuronal AT8‐positive tau inclusion bodies and coiled body–like structures were observed, and the latter were considered AT8‐positive tau inclusion bodies in oligodendrocytes. The size of the magnified images in the right two columns is 30 × 30 μm.

Discussion

In this study, we presumed that the causative gene for familial parkinsonism could be detected in these families with multiple patients presenting with various types of parkinsonism. We then identified insACA mutations in MAPT causing FTDP‐17 by genetic and biochemical analysis. Recently, the same mutation has been reported in Japanese patients, and biochemical abnormalities of the mutation have shown similar results. ⁶ In the early stages of the disease, severe dementia was rare, but parkinsonism was observed in most of our cases. However, many cases reported so far are thought to manifest with atypical parkinsonism. ⁷ Common clinical characteristics of our patients included tremor, supranuclear oculomotor dysfunction, gait disturbance, and severe rigidity. Early symptoms are similar to those of Parkinson's disease, but as the disease progresses, the symptoms resemble those of PSP and dementia was present.

Most patients with the MAPT mutations present with an FTD phenotype in which the main clinical manifestations of the behavioral variant frontotemporal dementia are combined with varying degrees of parkinsonism. In addition, various phenotypes such as those of Parkinson's disease, PDD, PSP, corticobasal syndrome, Pick disease, and globular glial tauopathy are observed in some cases. In particular, patients with C‐terminal mutations such as p.K369I, p.E372G, and p.G389R may exhibit tauopathy similar to that in Pick disease. ⁸ , ⁹ Mutations in MAPT around the second microtubule‐binding domain containing the insACA mutation, including p.N296del, p.N296N, p.P301L, p.P301T, and IVS10 + 16C > T, cause clinically atypical parkinsonism whose symptoms are similar to those of PSP, as well as those of FTD. ¹⁰ , ¹¹ The p.P301L mutation manifests as symptoms not only of FTD but also of globular glial tauopathy and corticobasal degeneration.

The pathological findings of FTDP‐17 caused by a MAPT mutation are diverse. ¹² Many cases are characterized by frontotemporal lobar atrophy and various inclusions of tau proteins. Neuronal loss is widespread, affecting the frontal and temporal lobes, globus pallidus, substantia nigra, striatum, dentate nucleus, and spinal cord. In our autopsy case, the clinical diagnosis was FTD, but pathological examination revealed neuronal loss not only in the cerebral cortex and anterior horn cells but also in the substantia nigra. The ratio of three‐ and four‐repeat tau varies according to the location of the mutation, but in our patient, the tau protein that abnormally accumulated in neurons was mainly four‐repeat tau, which reflects the location of the insACA mutation.

Tau abnormally accumulated in the brain can be visualized by imaging with PBB3 and PM‐PBB3. ¹³ [Carbon‐11]PBB3 positron emission tomography performed in patients with the p.N279K mutation of MAPT showed tau deposition in the midbrain and medial temporal areas. ¹⁴ Autopsy revealed the affinity of inclusion bodies for PM‐PBB3 as a result of abnormal aggregation of tau. We speculated that tau imaging could be used to observe the spread of abnormally accumulated tau lesions. Pathological analysis showed that the tau lesions had spread to the striatum, midbrain, and medulla oblongata, and we observed many neuronal inclusion bodies that had a unique structure not found in patients with sporadic PSP and other patients with FTDP‐17. Phosphorylated tau‐positive inclusion bodies were found not only in neurons but also in astrocytes and oligodendrocytes.

In conclusion, we studied two families affected by FTDP‐17 with insACA mutations in MAPT. As in previous reports of MAPT mutations, the patients had diverse clinical symptoms. Most patients initially exhibited clinical symptoms similar to those of PSP. Pathological findings were obtained in the autopsy of the patient with FTD, but lesions involving the substantia nigra could explain the disease process in other cases. Further understanding of the pathogenesis of FTDP‐17 requires the study of more cases.

Author Roles

(1) Research project: A. Conception, B. Organization, C. Execution; (2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique; (3) Manuscript: A. Writing of the first draft, B. Review and Critique.

H.M.: 1A, 1B, 1C, 2A, 2C, 3A

T.K.: 1C, 3A

Y.M.: 1C

M.O.: 1C

N.S.: 1C, 3A, 3B

T.M.: 1C, 3A, 3B

Y.S.: 1C

H.S.: 1C

Y.Y.: 1C

T.T.: 1C, 3B

Y.I.: 1C

H.I.: 1C

H.M.: 1C, 3B

M.H.: 1B, 1C

K.A.: 1C

T.S.: 1B

A.T.: 1B

H.K.: 1A, 1B

Disclosures

Ethical Compliance Statement: This study was approved by the Ethical Committee for Human Genome Research of Hiroshima University under the number Hi‐43‐25 and written informed consent was obtained from all participants before inclusion. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this work is consistent with those guidelines.

Funding Sources and Conflict of Interest: This work was supported by Takeda Science Foundation (HK) and Grants‐in‐Aid for Scientific Research from the Japan Society for the Promotion of Science (HMorino, 15 K15083; HK, KAKENHI 26242085). The authors declare that there are no conflicts of interest relevant to this work.

Financial Disclosures for the Previous 12 Months: The authors declare that there are no additional disclosures to report.

Supporting information

Data S1 Linkage analysis and homozygosity haplotyping. Genome‐Wide Human SNP Array 6.0 was used to genotype peripheral lymphocyte‐derived genomic DNA. Merlin software was used for linkage analysis in five patients from family 1. Homozygosity haplotyping was performed to identify common haplotype regions among patients.

Exome sequencing. Exome sequencing was conducted in samples from three patients in family 1 and one patient in family 2 using SeqCap EZ Human Exome Library v2.0. Variant calling and functional prediction were performed using standard tools.

Purification of recombinant tau. Mutant tau isoforms were expressed in Escherichia coli, purified, and quantified using SDS‐PAGE.

Microtubule polymerization assay. Effects of wild‐type and mutant tau on microtubule polymerization were assessed using porcine tubulin and fluorescence monitoring.

Heparin‐induced tau filament formation. Tau filament formation was quantified using thioflavin T assay after incubation with heparin.

Atomic force microscopy. Atomic force microscopy was used to observe tau filaments in solution.

Tissue extraction. Sarkosyl‐insoluble fractions were extracted from various tissues using a standardized protocol.

Western blotting. Proteins from tissue extracts were analyzed using western blotting with specific antibodies against tau isoforms.

Histopathological analysis. Brain and spinal cord sections were analyzed for neuronal loss, gliosis, and tau pathology using various staining methods and electron microscopy.

Fluorescence staining. Fluorescence staining of human brain sections was performed using PM‐PBB3 to detect tau aggregates.

Statistical analysis. Statistical analysis was performed using R, with significance set at P < 0.05 or <0.01.

Supplemental Figure S1. MRI features of the patient 1‐III‐6 and 1‐III‐4. (A, B) MRI of patient 1‐III‐6 was performed at age 64 years, 6 years after onset of disease. There was atrophy of the temporal lobe predominantly on the right side, and the midbrain was also atrophic. (C–E) MRI of patient 1‐III‐4 was performed at the age of 61 years, 5 years after onset. The patient presented with severe cognitive dysfunction. In addition to atrophy of the midbrain, there was severe temporal lobe atrophy predominantly on the left. MRI, magnetic resonance imaging; T1WI, T1‐weighted image; T2WI, T2‐weighted image; FLAIR, fluid attenuated inversion recovery.

Supplemental Figure S2. The insACA mutation reduces the ability of tau to polymerize microtubules. (A) Coomassie Brilliant Blue staining of purified recombinant wild‐type, insACA, P301L, and P301S mutant tau 2N4R isoforms separated by sodium dodecyl‐sulfate polyacrylamide gel electrophoresis (SDS‐PAGE). (B) Time course of tubulin polymerization absence (solid triangles) or presence of the wild‐type (solid circles), insACA (open circles), P301L (open squares), and P301S tau isoforms (open diamonds). (C) Graph of fluorescence units indicates that polymerized microtubules were quantified by 15 min (black bars) and 30 min (white bars) of incubation. Data represent means ± standard errors (n = 4). The insACA mutation significantly reduced the ability of tau to polymerize microtubules. (D) Upper panel show tubulin (arrowhead) and various tau proteins (arrow) in microtubule fractions. Lower panel show quantification of polymerized tubulin by 30 min of incubation. Data represent means ± standard errors (n = 3). Statistical significance was evaluated with analysis of variance, followed by Dunnett's test (*P < 0.05 and **P < 0.01, mutant tau isoforms vs. wild‐type tau).

Supplemental Figure S3. Heparin‐induced aggregation of insACA mutant tau. (A) Time course of thioflavin T fluorescence in heparin‐induced aggregation of the wild‐type (solid circles), insACA (open circles), P301L (open squares), and P301S tau isoforms (open triangles). The shaded area represents tau aggregation between 0 and 8 hours of incubation. (B) Enlargement of the shaded area in part A, showing thioflavin T fluorescence in the initial phase of tau aggregation. Data represent means ± standard errors (n = 3). InsACA mutant tau aggregated more rapidly than did wild‐type tau. (C) Thioflavin T fluorescence of wild‐type (black bars), insACA (white bars), P301L (hatched bars), and P301S tau isoforms (gray bars) after 4 and 168 h of incubation. (D) Sarkosyl‐insoluble tau was quantified after 168 h of incubation of wild‐type, insACA, P301L, and P301S tau isoforms. Although the thioflavin T fluorescence value of aggregated insACA tau was lower than that of wild‐type tau, the amount of aggregated insoluble insACA tau was comparable with that of P301S mutant tau. Data represent means ± standard errors (n = 3). Statistical significance was evaluated with analysis of variance followed by Dunnett's test (*P < 0.05 and **P < 0.01, mutant tau isoforms vs. wild‐type tau).

Supplemental Figure S4. Atomic force microscopy. Images show filamentous or granular formation of recombinant tau. Mutant tau had fewer granular aggregates and more elongated filamentous aggregates than did wild‐type tau.

Supplemental Figure S5. Western blotting of sarkosyl‐soluble and insoluble tau fractions. Samples were prepared from the spinal cord of a patient with frontotemporal dementia with parkinsonism linked to chromosome 17 (FTDP‐17), the spinal cord of a patient with amyotrophic lateral sclerosis (ALS) without tauopathy, the brain of a patient with Alzheimer's disease (AD), and the brain of an rTg4510 mouse. Arrowheads indicate various tau isoforms. The left two panels were immunoblotted with Tau12 to detect total tau, and the right two panels were immunoblotted with PHF1 to detect phosphorylated tau.

TABLE S1.

TABLE S2.