Novel PSEN1 (Q223 L) mutation causes early-onset Alzheimer’s disease: A case report

Abstract

BACKGROUND

Mutations in the PSEN1, PSEN2, and APP genes are known to cause Alzheimer’s disease (AD). Among these, PSEN1 mutations are the most frequent causes of autosomal dominant early-onset AD (EOAD). Patients harboring pathogenic mutations often exhibit considerable clinical heterogeneity. Identifying novel mutations and analyzing their associations with clinical cases is crucial for advancing our understanding of the pathogenesis of AD.

CASE SUMMARY

This report describes the clinical presentation of a family with EOAD. The proband was a 43-year-old Chinese female who presented with a three-year history of cognitive decline for 3 years. Magnetic resonance imaging demonstrated diffuse cerebral cortical atrophy. Next-generation sequencing identified a novel heterozygous c.668A>T mutation in PSEN1, which resulted in a p.Gln223 Leu. cerebrospinal fluid biomarker analysis revealed abnormal levels of amyloid and tau, indicative of underlying Alzheimer’s pathology. Furthermore, ¹⁸F-flortaucipir (AV-1451) positron emission topography and ¹⁸F-florbetapir (AV-45) positron emission topography imaging demonstrated extensive cerebral amyloid beta and tau deposition.

CONCLUSION

We report a novel pathogenic PSEN1 mutation, Q223 L, identified for the first time in a Chinese family with EOAD.

Key Words: Alzheimer’s disease; Mutation; PSEN1; Gene; Pathogenic; Case report

Core Tip: We identified a novel p.Gln223 Leu mutation in PSEN1 in a 43-year-old Chinese female with autosomal dominant early-onset Alzheimer’s disease (EOAD). Comprehensive clinical, neuroimaging, cerebrospinal fluid biomarkers, and ¹⁸F-florbetapir/¹⁸F-flortaucipir positron emission topography findings confirmed typical Alzheimer’s pathology. This mutation occurs at a mutational hotspot of PSEN1, is predicted to be pathogenic by in silico analysis, and has not been previously reported in public databases. Our report expands the mutation spectrum of PSEN1, enriches genotype - phenotype correlations of familial EOAD, and provides new evidence for the genetic diagnosis, pathogenesis research, and genetic counseling of EOAD.

INTRODUCTION

The three genes - PSEN1, PSEN2, and APP - which encode presenilin (PS) 1, PS2, and amyloid precursor protein (APP), respectively, have been identified as causative agents of Alzheimer’s disease (AD)[1-3]. AD with mutations is rare and investigations into its prevalence are limited. In individuals with early-onset AD (EOAD) and a family history of dementia, the prevalence of causative mutations is estimated to be 11%-68%.

Pathogenic mutations in AD predominantly follow an autosomal dominant inheritance and lead to an earlier average age of onset (46.2 years). Additionally, patients carrying pathogenic mutations are more likely to develop noncognitive neurological symptoms, including myoclonus, seizures, spastic paraplegia, extrapyramidal signs, and ataxia[4,5].

Mutations in PSEN1, PSEN2, and APP may contribute to AD by disrupting the proteolytic processing of APP. Specifically, amyloid beta (Aβ) is generated from APP through sequential cleavage by β-secretase and the γ-secretase complex, wherein PS1 and PS2 serve as the essential catalytic components[6-9]. Previous studies using transgenic animal models or transfected cells have demonstrated that these mutations can elevate the production of Aβ (particularly Aβ42) or increase the Aβ42/Aβ40 ratio[3,7,10].

The pathogenesis of AD remains unclear; thus, identifying novel mutations and analyzing their associations with clinical cases are crucial to advancing our understanding of AD pathogenesis. Here, we describe a novel PSEN1 mutation identified in a patient with EOAD and characterize its associated clinical and biomarker profiles.

CASE PRESENTATION

Chief complaints

The proband, a 43-year-old Chinese female, was admitted to the Department of Neurology at the Peking University Shenzhen Hospital. She presented with a history of cognitive decline for 3 years. She was entirely reliant on her husband for assistance with basic activities of daily living and had lost most of her self-care abilities.

History of present illness

The proband presented with a history of cognitive decline over 3 years.

History of past illness

The patient had no significant medical history.

Personal and family history

The proband’s brother (Figure 1, II-2) developed memory impairment at around the age of 20, followed by a slowly progressive course until his death at age 58. However, the exact cause of death and the clinical details remain unknown. The proband’s mother (Figure 1, I-1) developed cognitive impairment in old age; however, the exact timing and specific details were unclear, and she did not seek medical attention at that time.

Figure 1  Sequencing chromatograms showing the presence of the Q223 L mutation in the proband (II-2) and his non-demented 28-year-old son (III-1).

Physical examination

On admission, the patient’s vital signs were stable. Her Mini-Mental State Examination and Montreal Cognitive Assessment scores on admission were 8 and 3, respectively. Other neurological examination results were unremarkable. Focal neurological signs were not observed.

Laboratory examinations

DNA sequence analysis revealed a novel heterozygous transition from A to T (c.668A>T) in the proband (Figure 1), which introduced a non-conserved amino acid substitution, replacing the polar hydrophilic glutamine (Q) with a hydrophobic leucine (L) at codon 223 (NM_000021.4: C.668A>T, p.Gln223 Leu). We also screened the patient’s two sons who did not present with dementia. Segregation analysis showed a p.Gln223 Leu mutation in her 28-year-old son (Figure 2, III-1) but not in her 16-year-old son (Figure 2, III-2).

Figure 2 Family pedigree of the patient with the Q223 L mutation. The arrow (p) indicates the proband. Black filled symbols denote affected patients, and white filled symbols denote unaffected family members. M: Q223 L mutation; N: Wild type.

A cerebrospinal fluid (CSF) amyloid test showed decreased Aβ42 levels and Aβ42/Aβ40 ratio, and increased levels of pathological phospho-tau (Table 1). The CSF biomarker profiles were consistent with AD pathology.

Table 1 Levels of amyloid beta 40, amyloid beta 42, and p-Tau181 in cerebrospinal fluid.

CSF levels of biomarkers	Patient’s result	Indication	Normal control
Aβ40	7434.98 pg/mL	-	-
Aβ42	291.64 pg/mL	↓	> 691 pg/mL
Aβ42/Aβ40	0.04	↓	> 0.06
p-Tau181	55.66 pg/mL	↑	< 51 pg/mL

Aβ: Amyloid beta; CSF: Cerebrospinal fluid.

Imaging examinations

Brain magnetic resonance imaging (MRI) and positron emission topography (PET) computed tomography (CT) were performed when the proband was 43 years old.

MRI demonstrated diffuse cerebral cortical atrophy, most prominent in the bilateral medial temporal lobes, with concomitant white matter lesions in the centrum semiovale and corona radiata, and a few cortical microbleeds (Figure 3).

Figure 3 Cranial magnetic resonance imaging scans of the proband. A: Cranial magnetic resonance imaging (MRI) showed diffuse cerebral cortical atrophy, most prominently in the bilateral medial temporal lobes; B: Cranial MRI showed hyperintense lesion in the bilateral white matter in T2 weighted imaging; C: Cranial MRI showed hyperintense lesion in the bilateral white matter in fluid-attenuated inversion recovery images; D: Susceptibility-weighted imaging revealed several cortical microbleeds.

Fluorodeoxyglucose (FDG)-PET revealed diffuse hypometabolism in the right lateral temporal lobe, bilateral posterior cingulate gyri, bilateral precuneus, right superior parietal lobule, and bilateral inferior parietal lobules (the right lobes/Lobules were more affected). FDG-PET revealed diffuse and mildly reduced uptake in the left lateral temporal lobe, left frontal lobe, and left superior parietal lobule (Figure 4A). ¹⁸F-florbetapir (AV-45) PET imaging revealed extensive Aβ deposition throughout the cerebral cortex and cerebellum (Figure 4B), while ¹⁸F-flortaucipir (AV-1451) PET showed abnormal diffuse tau deposition throughout the bilateral frontal, temporal, parietal, and occipital lobes and in the cingulate gyrus (Figure 4C). The results of PET-CT demonstrated extensive cerebral Aβ and tau deposition.

Figure 4 Positron emission topography scans of the proband. A: ¹⁸F-florbetapir (AV-45) positron emission topography imaging showed extensive cerebral cortex and cerebellar amyloid beta deposition; B: ¹⁸F-flortaucipir (AV-1451) PET showed abnormal diffuse tau deposition throughout the bilateral frontal, temporal, parietal, and occipital lobes, and the cingulate gyrus; C: Fluorodeoxyglucose positron emission tomography-positron emission topography revealed diffuse hypometabolism in the right lateral temporal lobe, bilateral posterior cingulate gyri, bilateral precunei, right superior parietal lobule, and bilateral inferior parietal lobules (right lobes/Lobules more affected).

FINAL DIAGNOSIS

The patient was diagnosed with EOAD, which was confirmed by CSF analysis and PET-CT.

TREATMENT

The patient was treated with donepezil (10 mg/day). Concurrently, family members were advised to enhance patient supervision and consider cognitive behavioral therapy.

OUTCOME AND FOLLOW-UP

The patient remains under follow-up, for a total of 6 months so far. The patient’s cognitive decline remained unchanged.

DISCUSSION

To our knowledge, this is the first report of PSEN1 p.Q223 L mutation in a Chinese patient diagnosed with EOAD. The p.Gln223 Leu variant was located within a highly conserved residue in the N-terminal region of the sixth transmembrane domain of PS1. The predominant disease mechanism is a gain of function that alters γ-secretase specificity, shifting the sequential cleavage of APP to favor the production of longer Aβ peptides[3,11-16].

This variant was determined to be damaging in silico using a Rare Exome Variant Ensemble Learner (REVEL), with a score of 0.978. REVEL is an ensemble method that integrates predictions from multiple computational tools to improve the accuracy of the pathogenicity classification of missense variants. REVEL scores range from 0 to 1, with higher scores indicating a greater probability of the variant being pathogenic[11]. The pathogenicity of the p.Q223 L mutation has not been documented in the AD and Frontotemporal Dementia mutation databases (http://www.molgen.ua.ac.be/admutations/) and The Human Genetics Mutation Database (HGMD, http://www.hgmd.org). To our knowledge, this is the first report of a novel p.Q223 L mutation and it’s associated with AD.

Missense mutations in the same codons have also been reported. Codon 223 of the PSEN1 gene represents a mutational hotspot associated with autosomal dominant EOAD. Two distinct missense mutations in this codon, p.Gln223Arg (Q223R)[17] and p.Gln223 Lys (Q223K)[18,19], are linked to clinical presentations featuring earlier onset, faster progression, and a spectrum of atypical neurological symptoms. Q223R mutation has been reported in a patient with EOAD who presented with cognitive decline accompanied by dysarthria and spastic gait disorder. The diagnosis of the patient with the Q223R mutation was supported by CSF biomarkers and FDG-PET findings consistent with AD[17]. This Q223R variant was classified as pathogenic in the Alzforum database (https://www.alzforum.org/). Functional evidence from cell-based assays strongly supports the pathogenicity of the PSEN1 Q223K variant. When expressed in PSEN1/PSEN2 double-knockout neuroblastoma cells, this variant significantly increased the Aβ42/Aβ40 ratio, a well-established biochemical hallmark of pathogenic mutations in AD. This Q223K variant was classified as “Pathogenic” in previous study[18-20]. The convergence of these two different pathogenic substitutions at the same residue underscores the critical functional importance of this region within the PS1 protein, likely disrupting γ-secretase activity and promoting an amyloidogenic pathway while also leading to diverse and striking non-amnestic clinical features.

Notably, the patient exhibited distinct radiological features. The patient showed white matter hyperintensities (WMHs) with scattered cortical and subcortical microbleeds on the MRI. Despite being more frequently associated with vascular dementia, these findings have also been documented in patients with AD. Mao et al[21] reported that WMHs were a universal finding (7/7) in a cohort of genetically confirmed EOAD patients. WMHs were predominantly located in the periventricular regions, especially around the occipital and frontal horns, and presented as mildly punctate foci rather than large confluent areas. Susceptibility weighted imaging revealed no microbleeding in any patient in this cohort. The authors discussed two potential pathways for WMHs in EOAD: Ischemic injury from small vessel disease and Wallerian degeneration secondary to cortical neuron loss. Given the absence of vascular risk factors and microbleeds in their patients, they suggested that axonal degeneration following neuronal loss is a more plausible primary mechanism for the observed WMHs in these young genetically determined EOAD cases. However, the underlying mechanisms that link AD to WMHs and microbleeds remain unclear. Furthermore, 18F-florbetapir (AV-45) PET imaging revealed extensive Aβ deposition in this case - a finding that aligns with the established understanding that PSEN1 mutations often lead to widespread amyloid accumulation early in life.

In summary, several lines of evidence support the possible pathogenicity of the PSEN1 p.Gln223 Leu variant: (1) The proband was diagnosed with EOAD, supported by biomarker evidence indicating Alzheimer’s pathology, and had a clear family history consistent with autosomal dominant inheritance; (2) The mutation was absent in several large population databases; (3) In silico prediction analysis suggested that it is pathogenic and might have an effect on PS1 protein function; (4) It was located in a highly conserved residue among species; and (5) Two different mutations were previously reported in the same codon associated with EOAD. Based on the above evidence, this variation is likely pathogenic.

However, this study had some limitations. Complete segregation analysis could not be performed because of the death of other affected family members. In addition, we were unable to perform functional studies to examine the possible effects of this mutation on the pathogenesis of AD. Another limitation was that we did not perform biomarker testing on the 28-year-old asymptomatic son of the proband carrying the same mutation. Given the son’s asymptomatic status, biomarker testing was performed to avoid potential adverse psychological consequences. The patient will be enrolled in a rigorous clinical follow-up program with future biomarker testing contingent on the emergence of symptoms.

CONCLUSION

Here, we report a novel pathogenic PSEN1 p.Gln223 Leu mutation underlying EOAD, as evidenced by clinical, imaging, and biomarker data. This finding contributes to the genetic architecture of AD and has direct implications for genetic counseling and clinical surveillance of patients’ families.