Oosthoek Marlies, Vijverberg Everard G B, Blujdea Elena R, Veld Sjors G J G In't, Avilés Martín Pucheu, Zsadanyi Sára E, Hok-A-Hin Yanaika S, Visser Allerdien, van der Flier Wiesje M, Barkhof Frederik, Del Campo Marta, Schut Martijn C, Bejanin Alexandre, Alcolea Daniel, Teunissen Charlotte E, Vermunt Lisa
Neurochemistry Laboratory, Department of Laboratory Medicine, Amsterdam Neuroscience, Vrije Universiteit Amsterdam, Amsterdam UMC, De Boelelaan 1117, Amsterdam, 1081 HV, The Netherlands.
Alzheimer Center, Department of Neurology, Vrije Universiteit Amsterdam, Amsterdam UMC, De Boelelaan 1117, Amsterdam, 1081 HV, The Netherlands.
Alzheimers Res Ther. 2025 Jul 22;17(1):169. doi: 10.1186/s13195-025-01799-3.
Amyloid-related imaging abnormalities (ARIA) are a common and potentially dangerous side effect in anti-amyloid therapies, creating a need for tools to assess ARIA risk. Several patient factors have been linked to ARIA; namely the presence of microbleeds (MBL), APOE E4 carriership (APOE4), and extremely low CSF Aβ42 concentrations (A). We hypothesize that studying the CSF proteome of Alzheimer's disease (AD) dementia patients from a high-risk group (MBLAPOE4A) can inform on the biological underpinnings of ARIA risk and aid the progress of ARIA risk biomarkers.
We utilized CSF proteomic data of AD (n = 156) and cognitively unimpaired individuals (CU n = 100) of the Amsterdam Dementia Cohort. The proteome of the defined high-risk (n = 13) was compared to low-risk AD group (n = 23), using age and sex corrected linear regressions followed by gene ontology analysis. For biomarker prioritization, we selected proteins that were abnormal in the high-risk group versus low-risk and CU patients. The biomarkers were validated in an independent cohort (high risk n = 14, low risk n = 9) analyzed using customized multiplex panels. Lastly, we assessed biomarker lead co-expression.
Ninety-four proteins differentiated in the high-risk group compared to low-risk (p < 0.05), none surviving FDR correction. These proteins were enriched for synapse-related proteins and axonogenesis. CHIT1 (vs. low-risk AD: FC = 1.0, p = 0.014, vs. CU: FC = 2.4, p < 0.001) and DDAH1 (vs. low-risk AD: FC=-0.31, p = 0.046, vs. CU: FC = 0.5, p < 0.001) were prioritized as biomarker. DDAH1 protein changes replicated in an independent cohort (FC=-0.37, p = 0.010), and CHIT1 replicated on a trend level (FC = 0.70, p = 0.104). DDAH1 levels had the highest co-expression with synaptic process, energy utilization and RNA-binding cell signaling related proteins (R > 0.8).
The findings suggest that in the high risk group, there is a lack of upregulation in synapse and axonogenesis related proteins. High CSF CHIT1 and less increased CSF DDAH1 levels within AD relate to ARIA risk. From the literature, the link to ARIA risk for CHIT1 could be its contribution to innate immunity or vascular amyloid deposition, and for DDAH1 to blood-brain barrier integrity. Biomarker assays are available to assess the potential of CHIT1 and DDAH1 in trials and treatment studies in the clinical setting.
淀粉样蛋白相关成像异常（ARIA）是抗淀粉样蛋白疗法中常见且可能危险的副作用，因此需要评估ARIA风险的工具。一些患者因素与ARIA有关，即微出血（MBL）的存在、APOE E4携带者状态（APOE4）以及脑脊液Aβ42浓度极低（A）。我们假设，研究高危组（MBLAPOE4A）阿尔茨海默病（AD）痴呆患者的脑脊液蛋白质组，可以揭示ARIA风险的生物学基础，并有助于ARIA风险生物标志物的研究进展。
我们利用了阿姆斯特丹痴呆队列中AD患者（n = 156）和认知未受损个体（CU，n = 100）的脑脊液蛋白质组数据。使用年龄和性别校正的线性回归，随后进行基因本体分析，将定义的高危组（n = 13）的蛋白质组与低危AD组（n = 23）进行比较。为了对生物标志物进行优先排序，我们选择了高危组与低危组及CU患者相比异常的蛋白质。使用定制的多重检测板对独立队列（高危n = 14，低危n = 9）中的生物标志物进行验证。最后，我们评估了生物标志物主要共表达情况。
与低危组相比，高危组中有94种蛋白质存在差异（p < 0.05），但无一通过FDR校正。这些蛋白质富含与突触相关的蛋白质和轴突形成相关蛋白。几丁质酶1（CHIT1）（与低危AD组相比：FC = 1.0，p = 0.014，与CU组相比：FC = 2.4，p < 0.001）和二甲基精氨酸二甲胺水解酶1（DDAH1）（与低危AD组相比：FC = -0.31，p = 0.046，与CU组相比：FC = 0.5，p < 0.001）被优先作为生物标志物。DDAH1蛋白变化在独立队列中得到重复（FC = -0.37，p = 0.010），CHIT1在趋势水平上得到重复（FC = 0.70，p = 0.104）。DDAH1水平与突触过程、能量利用和RNA结合细胞信号相关蛋白的共表达最高（R > 0.8）。
研究结果表明，在高危组中，与突触和轴突形成相关的蛋白质缺乏上调。AD患者脑脊液中CHIT1水平较高且DDAH1水平升高较少与ARIA风险相关。从文献来看，CHIT1与ARIA风险的联系可能在于其对先天免疫或血管淀粉样蛋白沉积的作用，而DDAH1与血脑屏障完整性有关。在临床环境中的试验和治疗研究中，有生物标志物检测方法可用于评估CHIT1和DDAH1的潜力。
