Feng Gao

Background: Current blood-based biomarkers for the neurodegeneration component of the AT(N) framework, such as neurofilament light chain (NfL), reflect general neuroaxonal injury but lack specificity for Alzheimer's disease (AD). We investigated whether plasma neuronal pentraxin receptor (NPTXR), a synaptic protein involved in excitatory circuit maintenance, could serve as an AD-enriched blood biomarker to track downstream neurodegeneration, disease stage, and clinical prognosis.Methods: In this multicenter cohort study, we analyzed four prospective cohorts from China: a discovery cohort (CANDI; n=699), an independent blinded validation cohort (CLEAR-AD; n=401), and two disease-control cohorts enriched for vascular and non-AD neurodegeneration (cerebral amyloid angiopathy [CAA]/cerebral small vessel disease [CSVD]; n=238, and amyotrophic lateral sclerosis [ALS]; n=160). Participants included cognitively unimpaired individuals and patients with mild cognitive impairment (MCI), AD dementia, non-AD dementia, CAA-related intracerebral hemorrhage, CAA-related inflammation, CSVD-related intracerebral hemorrhage, and ALS. Amyloid status was defined by amyloid PET or cerebrospinal fluid (CSF) Aβ42/Aβ40 ratio. Plasma NPTXR was measured with an in-house Meso Scale Discovery immunoassay and benchmarked against plasma tTau and serum NfL. Primary outcomes were between-group differences in plasma NPTXR and its disease specificity relative to tTau and NfL. Secondary outcomes included its association with amyloid and tau pathology, structural MRI measures, cognitive performance, longitudinal cortical thinning, cognitive decline, and progression from MCI to dementia. Analyses were adjusted for age, sex, education, and APOE ε4, with correction for multiple testing.Findings: Among 1498 participants (874 [57·2%] women; mean [SD] age, 64·2 [9·5] years), plasma NPTXR levels were lower in AD than in cognitively unimpaired controls in both CANDI (mean difference, −103·55 pg/mL; 95% CI, −147·03 to −60·07; p<0·001) and CLEAR-AD (mean difference, −250·45 pg/mL; 95% CI, −369·94 to −130·96; p<0·001). Plasma NPTXR was also reduced in amyloid-positive MCI and showed the greatest reduction in AD dementia. Plasma NPTXR did not differ from controls in CSVD or ALS but was lower in CAA; in contrast, serum NfL was increased in CSVD and ALS. Lower plasma NPTXR was associated with greater Aβ-PET burden, higher CSF pTau181, thinner cortex, and worse cognitive performance. Higher baseline plasma NPTXR was also associated with slower cortical thinning, slower cognitive decline, and lower risk of progression from MCI to dementia (hazard ratio 0·33, 95% CI 0·14–0·80; p=0·014). Interpretation: Plasma NPTXR was consistently reduced across the biologically defined AD continuum, closely tracked AD-related neurodegeneration, and provided robust prognostic information for clinical progression. The relative stability of plasma NPTXR in ALS and CSVD—contrasting sharply with the non-specific elevations of NfL—supports its role as an AD-enriched biomarker of downstream synaptic and neurodegenerative change. Plasma NPTXR could complement amyloid and tau biomarkers for assessing downstream neurodegeneration and improving disease staging, prognosis, and therapeutic monitoring in AD.

Identifying biomarkers that precisely track the neurodegenerative component of Alzheimer's disease (AD) is essential for effective clinical management. Here we show that cerebrospinal fluid (CSF) levels of the synaptic proteins NPTX1 and NPTXR are robust indicators of disease severity and future clinical progression. In two independent, multi-ethnic cohorts spanning the AD continuum (n = 635), lower CSF NPTX levels correlate strongly with cognitive impairment and cortical thinning in AD-vulnerable regions. Longitudinally, baseline NPTX levels predict accelerated brain atrophy and the clinical transition from mild cognitive impairment to dementia, frequently outperforming or complementing established markers such as pTau181 and neurofilament light chain. These findings establish NPTX1 and NPTXR as sensitive, stage-specific markers of synaptic integrity and neurodegeneration. By accurately forecasting disease progression, these biomarkers offer significant potential to enhance patient stratification and provide a crucial tool for monitoring the efficacy of disease-modifying therapies in clinical trials.

Background: Amyloid and cerebrovascular pathologies play critical role in the pathogenesis of Alzheimer's disease (AD). Matrix metalloproteinases (MMPs) is known to degrade Aβ peptides, while they also contribute to cerebrovascular pathology. Presently, the precise roles of MMPs in AD remain unclear. Therefore, we investigated relationships between CSF levels of MMPs and amyloid and cerebrovascular pathologies. Method: A total of 299 participants were included from the CANDI cohort, including 50 cognitively unimpaired (CU) individuals, 69 patients with MCI, 137 AD patients, and 43 non‐ADD patients. MMP2, MMP9, Aβ38, Aβ40 and Aβ42 were measured in the CSF. Participants underwent structural MRI and AV45 PET scan. And fluid biomarkers (Qalb and CSF PDGFRβ) were used to assess cerebrovascular pathologies. Result: CSF MMP2 levels in MCI and AD patients were significantly lower compared to CU individuals and non‐ADD patients. Decreased CSF MMP2 levels, but not MMP9 were associated with lower levels of CSF Aβ peptides (Aβ38, Aβ40, Aβ42) and higher Aβ PET SUVR. Notably, CSF MMP2 was positively associated with multiple cortical regions within the AD continuum, but not in the non‐ADD continuum. Meanwhile, we observed positive correlations between CSF MMP2 and MMP9 with cerebrovascular pathologies, including WMH volume, CMB, lacunes, PVS, and Qalb. Additionally, lower CSF MMP2 was associated with longitudinal brain atrophy. Conclusion: This study suggested the involvement of MMPs in the cerebrovascular damage and the protective role of MMP2 in amyloid pathology, which highlight that the clinical application of MMP inhibitors should consider the protective effect of MMP2 on amyloid pathology.

Magnetic resonance imaging (MRI), combined with artificial intelligence techniques, has improved our understanding of brain structural change and enabled the estimation of brain age. Neurodegenerative disorders, such as Alzheimer’s disease (AD), have been linked to accelerated brain aging. In this study, we aimed to develop a deep-learning framework that processes and integrates MRI images to more accurately predict brain age, cognitive function, and amyloid pathology. In this study, we aimed to develop a deep-learning framework that processes and integrates MRI images to more accurately predict brain age, cognitive function, and amyloid pathology.We collected over 10,000 T1-weighted MRI scans from more than 7,000 individuals across six cohorts. We designed a multi-modal deep-learning framework that employs 3D convolutional neural networks to analyze MRI and additional neural networks to evaluate demographic data. Our initial model focused on predicting brain age, serving as a foundational model from which we developed separate models for cognition function and amyloid plaque prediction through transfer learning. The brain age prediction model achieved the mean absolute error (MAE) for cognitive normal population in the ADNI (test) datasets of 3.302 years. The gap between predicted brain age and chronological age significantly increases while cognition declines. The cognition prediction model exhibited a root mean square error (RMSE) of 0.334 for the Clinical Dementia Rating (CDR) regression task, achieving an area under the curve (AUC) of approximately 0.95 in identifying ing dementia patients. Dementia related brain regions, such as the medial temporal lobe, were identified by our model. Finally, amyloid plaque prediction model was trained to predict amyloid plaque, and achieved an AUC about 0.8 for dementia patients. These findings indicate that the present predictive models can identify subtle changes in brain structure, enabling precise estimates of brain age, cognitive status, and amyloid pathology. Such models could facilitate the use of MRI as a non-invasive diagnostic tool for neurodegenerative diseases, including AD.

β-secretase 1 (BACE1), known for its role in amyloid-β production associated with Alzheimer’s disease (AD), has also been suggested to be elevated in patients with Type 2 diabetes mellitus (T2DM). Notably, BACE1 could cleave the insulin receptor (InsR), leading to reduced InsR levels, which may impair insulin signaling and contribute to insulin resistance. Presently, we observed decreased InsR levels and impaired glucose disposal in the livers of mice with systemic overexpression of BACE1 (HUBC mice). This suggests that elevated BACE1 could contribute to insulin resistance by shedding membrane InsR. Additionally, mice fed a high-fat diet (HFD), a well-established model of T2DM, displayed increased BACE1 levels and decreased InsR. To further investigate whether inhibiting BACE1 could enhance insulin sensitivity and alleviate symptoms of diabetes, we treated HFD mice with the BACE1 inhibitor Elenbecestat. Remarkably, the administration of Elenbecestat restored InsR levels and improved their downstream signaling pathways, leading to increased insulin sensitivity and enhanced glucose tolerance. In summary, our findings suggest that inhibiting BACE1 can restore InsR expression and improve insulin-signaling sensitivity, ultimately resulting in enhanced diabetic phenotypes.

INTRODUCTION: We investigated the specific factors driving abnormal angiogenesis in Alzheimer's disease (AD) and its role in cerebrovascular lesions and neurodegeneration. METHODS: We assessed cerebrovascular pathologies, amyloid‐beta (Aβ), and tau pathologies in post mortem human brains and detected 12 angiogenic factors in cerebrospinal fluid (CSF) from the China Aging and Neurodegenerative Disease Initiative (CANDI) cohort. RESULTS: We observed severe blood‐brain barrier damage and elevated levels of the vascular marker CD31 in human AD brains, which had a stronger correlation with tau pathology than Aβ pathology. Consistently, only CSF pTau181 showed positive associations with CSF angiogenesis factors (soluble vascular endothelial growth factor receptor 2 [sVEGFR2], VEGF‐C, VEGF‐D, placental growth factor [PLGF], Angiopoietin2, and Serpin E1), but not CSF Aβ42/40. Additionally, higher levels of CSF sVEGFR1, soluble Tyrosine‐protein kinase receptor 2 [sTIE2], PLGF, and interleukin 8 [IL8], as well as lower levels of CSF urokinase‐type plasminogen activator [uPA], were associated with worsen cerebrovascular pathologies and neurodegeneration. DISCUSSION: Our findings indicate that tau pathology may play a critical role in pathological angiogenesis, contributing to cerebrovascular lesions and neurodegeneration in AD. HIGHLIGHTS: BBB damage and elevated vascular marker CD31 in human AD brains had a stronger correlation with tau pathology than Aβ pathology. CSF pTau181 mediated the effect of CSF Aβ42/40 on CSF sVEGFR1 and sTIE2. Only CSF pTau181 showed positive associations with sVEGFR2, VEGF‐C, VEGF‐D, PLGF, Angiopoietin2, and Serpin E1. Higher sVEGFR1, sTIE2, PLGF, and IL8, and lower uPA in CSF, were associated with cerebrovascular pathologies and neurodegeneration.

Background: β ‐secretase 1 (BACE1) plays a key role in amyloidogenic pathway and is considered a new mechanism for cerebral small vessel disease (CSVD). We explore the potential role of plasma BACE1 in CSVD and the pathological process it may be involved in. Methods and Results: We enrolled 163 participants with CSVD (114 cerebral amyloid angiopathy and 49 hypertensive hemorrhage), and 96 cognitively unimpaired elders and 40 participants with Alzheimer's disease as controls. We measured BACE1 activity using a synthetic fluorescence substrate enzyme‐linked immunosorbent assay. We used regression models to investigate associations between BACE1 and imaging and blood markers as well as clinical outcomes in CSVD. Plasma BACE1 activity was significantly higher in CSVD (median 862.0 relative fluorescence units [RFU], interquartile range 700.6–1032.9) compared with elder controls (522.5 RFU, 438.3–662.1, P P =0.032). For CSVD, there was an association between plasma and cerebrospinal fluid BACE1 activity (β=0.50, P =0.030). Plasma BACE1 activity was negatively associated with hippocampal volume (coefficient −2.40, P P P =0.006) but not with specific cerebrovascular imaging markers. During 2‐year follow‐up, BACE1 activity was independently related to dementia conversion (odds ratio, 17.72; P =0.001) but not stroke in CSVD. Conclusions: Plasma BACE1 activity is moderately increased and associated with neurodegeneration and cognitive impairment risk in CSVD. It indicates that BACE1 is a promising biomarker especially for CSVD‐related neurodegeneration.

Background: The exact impacts of β‐site APP cleaving enzyme (BACE1) on brain atrophy and cognitive decline in the Alzheimer’s disease (AD) remain not fully understood. This study aimed to investigate the relationship between BACE1 in cerebrospinal fluid (CSF) and amyloid‐β (Aβ) pathology, neurodegeneration and cognitive function.Method: This study involved 359 participants from original individuals of the China Aging and Neurodegenerative Disorder Initiative (CANDI) cohort, who underwent measurements of AD biomarkers. Among them, 259 individuals underwent baseline structural magnetic resonance imaging (MRI) scans, 101 individuals underwent ¹⁸F‐labeled florbetapir (AV45) positron emission tomography (PET) scans and 117 individuals received ¹⁸F‐labeled fluorodeoxyglucose (FDG) PET scans. Additionally, 95 participants underwent follow‐up cognitive tests, and 81 participants received follow‐up structural MRI scans.Results: Out of the 359 participants, 216 (60.17%) were women. The mean [SD] age was 62.18 [8.48] years, and the mean [SD] follow‐up times for MRI and cognitive tests were 1.52 (0.55) years and 1.79 (0.87) years, respectively. CSF BACE1 levels were elevated in patients with mild cognitive impairment caused by AD or AD. CSF BACE1 levels were correlated with soluble Aβ40 (β = 0.469, p < .001) and Aβ42 (β = 0.174, p = .001), but not AV45‐PET signal. Furthermore, CSF BACE1 levels were positively associated with cortical thickness in specific brain regions, including the insula and supramarginal areas. BACE1 was also linked to cortical glucose metabolism in frontal and supramarginal areas after adjusting regional Aβ burden. Additionally, the longitudinal study revealed that elevated BACE1 levels were associated with reduced cortical thinning in specific brain regions and less cognitive decline (β = ‐0.219, p = .05) over time.Conclusions: While increased CSF BACE1 is involved in promoting Aβ production, higher levels of BACE1 are associated with better cortical thickness and cognitive function. Maintaining the normal physiological function of BACE1 in the brain is important for cognitive function.