Alzheimer’s disease (AD), the most common neurodegenerative disorder among the elderly, affects tens of millions worldwide. Recent research has brought significant breakthroughs, offering new hope for early diagnosis and treatment.
A team led by Professor Luorongcan from the School of Life Sciences at Lanzhou University and Professor Yujintai from Huashan Hospital affiliated to Fudan University recently published a News & Views article in Nature Aging. The article systematically reviews a study by Professor Thomas K. Karikari’s team from the University of Gothenburg, Sweden, published in Nature Medicine. This study has identified novel biomarkers for the early stages of AD – phosphorylated tau at Serine-262 and Serine-356.
These phosphorylated tau proteins, p-tau Ser-262 and p-tau Ser-356, are key indicators of early tau aggregates. Located in the core sequence of soluble tau aggregates (STAs), they are closely associated with granular, pre-fibrillar tau aggregates before the formation of neurofibrillary tangles (NFTs). Unlike traditional diagnostic methods, such as tau positron emission tomography (PET) and post-mortem histopathology, which often detect pathological changes only after significant cognitive decline, these biomarkers can be detected in cerebrospinal fluid (CSF). This enables early non-invasive assessment of tau pathology, helping clinicians identify high-risk individuals before irreversible damage occurs. Moreover, these biomarkers can distinguish AD from other tauopathies and are independently related to the severity of NFT burden and cognitive decline, regardless of amyloid-β (Aβ) deposition. This discovery also suggests that targeting STAs may be more effective in treatment, as early aggregates are considered more neurotoxic.
Another significant development comes from a study published in Nature Medicine by Sandrine Andrieu and other researchers. They emphasize the transformative potential of artificial intelligence (AI) in overcoming the key bottlenecks in Alzheimer’s research, such as data sharing barriers, long cycles of biomedical discovery and clinical trials, and the disconnect between laboratory insights and clinical practice.
AI can analyze diverse data, including non-targeted and targeted liquid biomarkers, cognitive tests, speech patterns, neuroimaging, and wearable sensor outputs, to identify new digital biomarkers. These biomarkers, which may even be generated by consumer-grade devices, have the potential to revolutionize early detection, enhance patient stratification, and significantly improve clinical trial efficiency. Additionally, an AI-driven research platform that integrates multimodal datasets and promotes privacy-compliant international data sharing is advocated. Such a platform would serve as the center of a research ecosystem, providing equal access and accelerating interdisciplinary and global cooperation. AI is also expected to transform clinical trials by optimizing patient recruitment, stratification, and retention. Real-time data analysis driven by AI can enhance trial efficiency, reduce costs, and lower dropout rates. The use of “digital twins” to simulate individualized disease progression and treatment responses can personalize treatment strategies before clinical implementation, streamlining the drug development process.
In addition, a long – term study by Professor Jia Jianping’s team from Xuanwu Hospital, Capital Medical University, published in The New England Journal of Medicine, is the largest and longest – followed longitudinal cohort study reflecting biomarker changes before AD diagnosis. Conducted in the Chinese population over 20 years, it for the first time revealed the dynamic changes of cerebrospinal fluid and imaging biomarkers from the asymptomatic to symptomatic stages of AD, clarifying the key earliest nodes of pathophysiological changes in AD. This provides a time window for anti-AD new drug development targeting pathological proteins like Aβ and strong evidence for ultra-early diagnosis and precise intervention. The study showed the order and time points of biomarker differences between the AD group and the cognitively normal group: Aβ (18 years before diagnosis), Aβ 42/40 (14 years before diagnosis), p-tau 181 (11 years before diagnosis), t-tau (10 years before diagnosis), NfL (9 years before diagnosis), hippocampal atrophy (8 years before diagnosis), and cognitive decline (6 years before diagnosis).
A milestone study in Nature Medicine also explored the world of proteins in cerebrospinal fluid on an unprecedented scale. It found that synapse proteins, crucial messengers for information transfer between neurons, are closely related to cognitive decline, even surpassing traditional Aβ and tau proteins. The ratio YWHAG:NPTX2 was discovered to predict cognitive outcomes, classify risk levels, and even issue warnings several years before symptoms appear.
Furthermore, a large-scale plasma proteomics analysis published in Nature Aging examined 6,905 aptamers for 6,106 unique proteins in the plasma of more than 3,300 well-characterized individuals. The study identified 416 proteins associated with clinical AD status, 294 of which are new. Seven proteins were identified using machine learning models with high predictive power for both clinical AD (area under the curve (AUC) >0.72) and biomarker-defined AD status (AUC >0.88), and these findings were replicated in multiple external cohorts and orthogonal platforms, highlighting the potential of plasma proteins as biomarkers for early AD detection, monitoring, and guiding treatment decisions.
These latest research results have opened up new directions for the early diagnosis, treatment, and research of Alzheimer’s disease. However, challenges remain. For example, the clinical application of new biomarkers such as p-tau Ser-262 and p-tau Ser-356 needs to be validated in larger and more diverse cohorts. Developing therapeutic drugs targeting these phosphorylated sites requires strict preclinical and clinical testing. Nevertheless, with continuous exploration, these discoveries will surely guide future scientific research and clinical strategies, bringing us closer to the goal of early intervention and effective treatment of Alzheimer’s disease.
Related topics:
