Supranuclear Palsy, Progressive, 1


A number sign (#) is used with this entry because of evidence that progressive supranuclear palsy-1 (PSNP1) is caused by heterozygous mutation in the gene encoding microtubule-associated protein tau (MAPT; 157140) on chromosome 17q21.


Progressive supranuclear palsy (PSP) is the second most frequent cause of degenerative parkinsonism. In addition to parkinsonism, the clinical symptoms include early postural instability, supranuclear gaze palsy, and cognitive decline. Neuropathologically, the disorder is characterized by abundant neurofibrillary tangles, which differ in both distribution and composition from those associated with Alzheimer disease. In progressive supranuclear palsy, the tangles are primarily localized to subcortical regions and are found in both neurons and glia, whereas in Alzheimer disease they are more widespread, largely cortical, and limited to neurons. They also have different characteristics at the ultrastructural level (Baker et al., 1999).

Kertesz (2003) suggested the term 'Pick complex' to represent the overlapping syndromes of frontotemporal dementia (FTD; 600274), primary progressive aphasia (PPA), corticobasal degeneration (CBD), progressive supranuclear palsy, and FTD with motor neuron disease. He noted that frontotemporal dementia may also be referred to as 'clinical Pick disease,' and that the term 'Pick disease' (172700) should be restricted to the pathologic finding of Pick bodies.

Genetic Heterogeneity of Progressive Supranuclear Palsy

Other loci for PSP have been mapped to chromosome 1q31 (PSNP2; 609454) and 11p12-p11 (PSNP3; 610898).

See also Parkinson-dementia syndrome and atypical progressive supranuclear palsy (260540).

Clinical Features

Rojo et al. (1999) reported 12 pedigrees of familial progressive supranuclear palsy, confirmed by pathology in 4 probands. Pathologic diagnosis was confirmed by internationally agreed criteria. The spectrum of the clinical phenotypes in these families was variable, including 34 typical cases of PSP (12 probands plus 22 secondary cases), 3 patients with postural tremor, 3 with dementia, 1 with parkinsonism, 2 with tremor, dystonia, gaze palsy, and tics, and 1 with gait disturbance. Ros et al. (2005) provided an update on 1 of the families reported by Rojo et al. (1999). Three family members were affected. At 37 years of age, the proband developed an akinetic-rigid syndrome, gait disturbance, frequent falls, micrographia, dysarthria, eyelid apraxia, abolition of upgaze, and hyperreflexia with unilateral extensor plantar response. PET scan showed reduced striatal fluorodopa uptake. The patient showed disease progression, with axial dystonia, mutism, complete vertical gaze palsy, and dysphagia, leading to death at age 45 years. Neuropathologic examination showed atrophy of the mesencephalon, pons, striatum, and subthalamic nuclei, and depigmentation of the substantia nigra. There was tau protein accumulation in neurons and glia, primarily in the mesencephalon. Protein analysis of brain tissue detected hyperphosphorylated tau protein and overexpression of tau isoforms with 4 microtubule-binding repeats.

Nath et al. (2003) provided a detailed review of the clinical features of 187 patients with PSP in the United Kingdom. The most common symptoms included bradykinesia, falls, cognitive impairment, bulbar or language problems, and visual disturbances, such as diplopia. Other variable features included tremor, retrocollis, limb dystonia, and favorable response to levodopa. Average age at onset was 66 years, and the mean survival was 5 to 6 years.

Tuite et al. (2005) reported a family in which several members had clinical features consistent with PSP and others had features consistent with corticobasal degeneration. Neuropathologic examination of 2 affected sibs showed features of PSP in 1 sib and features of CBD in the other. Specifically, PSP features included tufted astrocytes, whereas CBD features included ballooned neurons and astrocytic plaques. Both patients had severe neuronal loss in the substantia nigra. Genetic studies of 4 affected family members demonstrated the H1/H1 tau haplotype, but no mutations in the MAPT gene were identified. There was a history of remote consanguinity, and Tuite et al. (2005) suggested a unifying genetic etiology in this family.

Donker Kaat et al. (2009) reported 176 patients who fulfilled the diagnostic criteria for PSP, including 65 possible, 91 probable, and 20 definite. The mean age at PSP onset was 66.5 years. Presenting symptoms included gait disorder with falls (65%), behavioral changes (21%), bradykinesia (20%), cognitive decline (16%), stiffness (14%), speech problems (9%), visual complaints (6%), and tremor (4%). Tremor at disease onset occurred significantly more frequently in patients with PSP who had a positive family history compared to patients who did not (9% vs 1%, p = 0.02). Neuropathologic examination of 20 patients were consistent with the diagnosis. Detailed results of 5 patients showed mild frontal atrophy and depigmentation of the substantia nigra and locus ceruleus. There was severe neuronal loss and gliosis in the pallidum, subthalamic nucleus, and substantia nigra. Immunohistochemistry with the MAPT AT8 antibody (which stains 4-repeat tau) showed neurofibrillary tangles, tufted astrocytes, and coiled bodies with threads in the striatum and caudate; a variable number of these were found in the subthalamic nucleus and thalamus. The neocortex showed a few pre-tangles and tufted astrocytes in the frontal, temporal, and parietal cortex.

Neuroradiologic Studies

Piccini et al. (2001) studied regional cerebral dopaminergic function and glucose metabolism in members of 2 large kindreds with familial PSP in an effort to identify subclinical cases. They studied 3 clinically affected PSP patients who showed significant reduction in caudate and putamen uptake of (18)F-dopa in positron emission tomography (PET) studies, along with a significant reduction in striatal, lateral, and medial premotor area and dorsal prefrontal cortex glucose metabolism. In 4 of 15 asymptomatic relatives, caudate and putamen (18)F-dopa uptake was 2.5 SDs lower than the normal mean. These 4 subjects and a fifth asymptomatic relative with normal (18)F-dopa uptake showed a significant reduction of cortical and striatal glucose metabolism in a pattern similar to that of their affected relatives.

To calculate the rate of brain atrophy in 6 patients with PSP, Josephs et al. (2006) used serial MRI scans and a technique called the 'boundary shift integral,' which calculates the integral change in the brain/CSF fluid boundary over the interior and exterior surfaces of the brain. The mean interval between baseline and final MRI scan was 3.1 years. PSP Patients had a rate of cerebral atrophy and ventricular expansion of 1.3% and 7% per year, respectively, compared to 0.4% and 1.8%, respectively, in control individuals. Josephs et al. (2006) suggested that these benchmark rates and the boundary shift integral technique could be used to monitor disease progression and response to therapy in PSP.

Clinical Variability

Nicholl et al. (2003) described a form of PSP characterized by fatal respiratory hypoventilation in 2 sibs from a consanguineous marriage. The authors called the disorder 'tauopathy and respiratory failure.' The 29-year-old pregnant sister developed dyspnea with stridor, and later had a generalized seizure that left her unconscious. Despite therapy, she died after 9 days. Her 30-year-old brother developed cough syncope, dyspnea, and central apnea, which progressed over 40 months, leading to death at age 34 years. During the illness, he showed slow smooth pursuit and impaired saccades, mild rigidity and bradykinesia, and myoclonic jerks. Neuropathologic examination showed widespread neuronal eosinophilia and pyknosis with gliosis in multiple brain regions, consistent with hypoxic brain damage. There was pervasive tau pathology in neuronal perikarya, neurites, and threads in the gray matter of the hippocampus, thalamus, and pons, but not in the cerebral cortex. Functional studies indicated that the mutated protein showed reduced binding to microtubules as well as increased fibrillization and aggregation. Both sibs carried the H1/H1 haplotype associated with PSP. Nicholl et al. (2003) commented on the unusual apparent autosomal recessive inheritance of this tauopathy. In a review of the role of tau in neurodegenerative diseases, Quadros et al. (2007) classified the disorder in the sibs reported by Nicholl et al. (2003) as PSP.


Van Leeuwen et al. (2006) detected aberrant frameshifted proteins, APP+1 (APP; 104760) and UBB+1 (UBB; 191339), within the neuropathologic hallmarks of Alzheimer disease (AD; 104300) and other MAPT-related dementias, including Pick disease, progressive supranuclear palsy, and less commonly frontotemporal dementia. Van Leeuwen et al. (2006) postulated that accumulation of APP+1 and UBB+1, which represents defective proteasome function, contributes to various forms of dementia.

Botella-Lopez et al. (2006) found increased levels of a 180-kD reelin (RELN; 600514) fragment in CSF from 19 patients with AD compared to 11 nondemented controls. Western blot and PCR analysis confirmed increased levels of reelin protein and mRNA in tissue samples from the frontal cortex of AD patients. Reelin was not increased in plasma samples, suggesting distinct cellular origins. The reelin 180-kD fragment was also increased in CSF samples of other neurodegenerative disorders, including frontotemporal dementia, PSP, and Parkinson disease (PD; 168600).


De Yebenes et al. (1995) described 7 families with multiple affected individuals. These included 3 suggesting autosomal dominant inheritance (see 609454) and 1 family in which there was parental consanguinity, suggesting recessive inheritance. Although the majority of cases of progressive supranuclear palsy appear to be sporadic, there may be rare genetically determined forms.

Golbe et al. (1996) failed to demonstrate substantial contribution of heredity to the cause of progressive supranuclear palsy in a study of 91 personally examined patients (75 were age-matched with controls). They pointed out that the size of their study may not have been sufficient to demonstrate a subtle familial aggregation in progressive supranuclear palsy.

The presence of affected members in at least 2 generations in 8 of the 12 families reported by Rojo et al. (1999) and the absence of consanguinity suggested autosomal dominant transmission with incomplete penetrance. Rojo et al. (1999) concluded that hereditary PSP is more frequent than previously thought and that the scarcity of familial cases may be related to a lack of recognition of the variable phenotypic expression of the disease.

Donker Kaat et al. (2009) found that 57 (33%) of 172 probands with PSP had at least 1 first-degree relative who had dementia or parkinsonism, compared to 131 (25%) of the control subjects (odds ratio of 1.5). In addition, patients with PSP had more first-degree relatives with parkinsonism compared to controls, (OR of 3.9). Twelve patients with PSP (7%) fulfilled the criteria for autosomal dominant transmission of PSP, dementia, tremor, or parkinsonism. One patient was found to have a P301L mutation in the MAPT gene (157140.0001). The findings supported familial aggregation of parkinsonism in progressive supranuclear palsy.

Molecular Genetics

Conrad et al. (1997) demonstrated an association between progressive supranuclear palsy and a dinucleotide (TG) polymorphic repeat between exons 9 and 10 of the microtubule-associated protein tau gene (MAPT; 157140). This association was subsequently confirmed by several other studies (Oliva et al., 1998; Higgins et al., 1998; Bennett et al., 1998). In each case, an overrepresentation of the most common allele (a0) and genotype (a0a0) was reported for the progressive supranuclear palsy group. However, due to the nature of the dinucleotide polymorphism, it was considered unlikely that this variation was biologically significant in the disease process, but was in fact in disequilibrium with other polymorphisms. Baker et al. (1999) described 2 extended haplotypes, H1 and H2, that cover the human tau gene, and showed that the most common haplotype, H1, was significantly overrepresented in patients with PSP, extending earlier reports of the association between the intronic dinucleotide polymorphism a0 and the disorder.

Using an autosomal recessive model, Higgins et al. (1998) found that progressive supranuclear palsy was in linkage disequilibrium with the tau gene, but not with the alpha-synuclein (163890) gene.

Higgins et al. (1999) found that 22 unrelated patients with PSP had a common extended haplotype of the tau gene, designated HapA, characterized by a homozygous polymorphism in the 5-prime splice site untranslated region of exon 1, 2 missense mutations in exon 4A, and a nonsense mutation in the 5-prime splice site of exon 8. Higgins et al. (2000) found that 51 of 52 patients with PSP had the HapA haplotype.

In 25 unrelated individuals with PSP and 6 individuals with corticobasal degeneration, Higgins et al. (1999) failed to demonstrate the R406W tau mutation (157140.0003) that had been reported by Hutton et al. (1998) in a family segregating an atypical form of progressive supranuclear palsy in an autosomal dominant manner.

Morris et al. (2000) analyzed alpha-synuclein, tau, synphilin (603779), and APOE (107741) genotypes in 50 patients with PSP. They confirmed the predisposing effect of the tau H1 haplotype described by Baker et al. (1999) and others, and found no association between PSP and alpha-synuclein, synphilin, or APOE.

Litvan et al. (2001) examined 63 patients with PSP and found that the presence of the tau H1/H1 genotype was significantly greater in patients compared to controls, as previously reported. There was no difference between PSP cases with one H1 or two H1 alleles in the age of onset, severity, or survival of patients, thus showing that tau genotyping does not predict the prognosis of PSP. However, Litvan et al. (2001) noted that the majority of the PSP patients carried the H1/H1 genotype (88.9%) and none of the patients carried the H2/H2 genotype, thus limiting the conclusions of the study.

Among 96 patients with PSP, Poorkaj et al. (2002) identified 1 patient with a mutation in a highly conserved position in exon 1 of the MAPT gene (157140.0019).

In 2 sibs from a consanguineous marriage who presented with a form of progressive supranuclear palsy designated 'tauopathy and respiratory failure,' Nicholl et al. (2003) identified a homozygous 1291C-T transition in exon 12 of the MAPT gene, resulting in a nonconserved ser352-to-leu (S352L) substitution in the N-terminal repeat of the tau protein. Both sibs carried the H1/H1 haplotype associated with PSP.

De Silva et al. (2003) presented evidence suggesting that the Q7R polymorphism of the saitohin gene (607067), which is located within intron 9 of the MAPT gene, is associated with PSP: 47 of 49 (95.9%) affected patients had the QQ genotype, whereas 37 of 62 (59.7%) controls had the QQ genotype. However, the authors noted that the Q allele, which represents the tau H1 haplotype, was also the most common haplotype in normal controls.

Using single-nucleotide polymorphisms, Pittman et al. (2004) mapped linkage disequilibrium (LD) in the regions flanking MAPT and established the maximum extent of the haplotype block on chromosome 17q21.31 as a region covering approximately 2 Mb. The gene-rich region extended centromerically beyond the corticotropin-releasing hormone receptor-1 gene (CRHR1; 122561) to a region of approximately 400 kb, where there was a complete loss of LD. The telomeric end was defined by an approximately 150-kb region just beyond the WNT3 (165330) gene. The authors showed that the entire, fully extended H1 haplotype was associated with PSP, which implicates several other genes in addition to MAPT as candidate pathogenic loci.

Rademakers et al. (2005) and Pittman et al. (2005) used a large collection of pathologically confirmed PSNP samples to fine map PSNP risk on H1 chromosomes in PSNP cases and controls. PSNP risk was associated with an extended subhaplotype (H1c), and the risk for PSNP was narrowed to a 22-kb region in intron 0 of MAPT by examining younger patients with, presumably, a larger genetic component to their disease. The most likely explanation of the association of the MAPT H1 haplotype and PSNP is that variants in the H1 (and H2) haplotypes confer risk of (protect against) disease by altering expression at the locus, with the risky H1 haplotype expressing higher levels of MAPT.

In affected members of the family reported by Rojo et al. (1999), Ros et al. (2005) identified a heterozygous mutation in the MAPT gene (157140.0025).

Associations Pending Confirmation

For discussion of a possible association of PSP with repeat expansion in the ATXN2 gene, see 601517.0002.