Biology subjects

Melzer, T.

Publications and source records attributed to Melzer, T..

4 recordsLinked to original sources

bioRxiv2026

Thalamic nuclei insights into Alzheimer's disease

INTRODUCTIONThalamic nuclei support multiple cognitive processes, yet their integrity in biologically-defined Alzheimers disease (AD) remains unknown. METHODAmyloid status was determined using PET Centiloids >24 in 1,327 participants from ADNI. Combined with clinical diagnosis, this yielded six groups: amyloid-negative or positive CN-MCI-dementia/AD. Thalamic nuclei volumes were extracted from T1-weighted MRI using the HIPS-THOMAS algorithm. RESULTSLarge volume reductions in the anteroventral, mediodorsal, and pulvinar nuclei were observed in amyloid-positive MCI and AD. Reduced volumes were also evident in amyloid-positive CN, supporting preclinical AD. Adding the anteroventral nucleus improved cognitive status classification in Random Forest analyses. A phenotypic model integrating thalamic nuclei clearly distinguished amyloid-positive groups from amyloid-negative CN and reclassified non-AD patients with 68% of amyloid-negative MCI subjects as CN-like, and 27% of amyloid-positive CN as MCI-like. DISCUSSIONThalamic volumetry from conventional T1-weighted MRI enhances clinical insight into AD and provides a practical biomarker for disease intervention.

neuroscience↗

bioRxiv2026

Source-space EEG functional connectivity and prediction of cognition in Parkinsons disease: No added benefit of individualized head models over standard templates

IntroductionCognitive decline is a major non-motor feature of Parkinsons disease (PD), but reliable and accessible biomarkers remain limited. Resting-state electroencephalography (EEG) is a promising candidate because it is low-cost, portable, and well suited to repeated assessment. Recent work has increasingly focused on source-space functional connectivity (FC) for the prediction of cognition. However, the influence of source modelling based on an individualized MRI-based head model relative to that based on standard template model is unknown. MethodsTo compare these two source-space EEG FC methods, we analysed EEG data from the New Zealand Parkinsons Progression Programme, including 136 people with PD and 51 age-similar controls. Source space resting-state EEG, parcellated with the HCP-MMP1 atlas, was used to derive amplitude envelope correlation (AEC) and debiased weighted phase lag index (dwPLI) across six canonical frequency bands. The resulting twenty-four FC modalities were evaluated using six machine-learning regression algorithms within a nested cross-validation framework. ResultsTheta-, alpha-, and beta-band FC showed the most consistent prediction of global cognition. The strongest performance was observed for theta- and alpha-band AEC and dwPLI features (max R{superscript 2} = 0.170, 95% CI = 0.067-0.262; max r = 0.439, 95% CI = 0.328-0.537). Standard and individualized head models showed comparable predictive performance across nearly all modalities. The feature-importance neuroanatomical patterns for Cole-Anticevic networks were also similar between the two head-model options. ConclusionsWe found that source-space resting-state EEG FC can predict cognitive performance in PD. The comparability of the two head models suggests that the more user-friendly and less resource-intensive standard template head model is sufficient for this purpose. This supports feasible, scalable, and clinically accessible EEG-based FC biomarkers of cognition in PD.

neuroscience↗

bioRxiv2024

Improving Predictability, Test-Retest Reliability and Generalisability of Brain-Wide Associations for Cognitive Abilities via Multimodal Stacking

Brain-wide association studies (BWASs) have attempted to relate cognitive abilities with brain phenotypes, but have been challenged by issues such as predictability, test-retest reliability, and cross-cohort generalisability. To tackle these challenges, we proposed a machine-learning "stacking" approach that draws information from whole-brain magnetic resonance imaging (MRI) across different modalities, from task-fMRI contrasts and functional connectivity during tasks and rest to structural measures, into one prediction model. We benchmarked the benefits of stacking, using the Human Connectome Projects: Young Adults (n=873, 22-35 years old) and Human Connectome Projects-Aging (n=504, 35-100 years old) and the Dunedin Multidisciplinary Health and Development Study (Dunedin Study, n=754, 45 years old). For predictability, stacked models led to out-of-sample r[~].5-.6 when predicting cognitive abilities at the time of scanning, primarily driven by task-fMRI contrasts. Notably, using the Dunedin Study, we were able to predict participants cognitive abilities at ages 7, 9, and 11 using their multimodal MRI at age 45, with an out-of-sample r of 0.52. For test-retest reliability, stacked models reached an excellent level of reliability (ICC>.75), even when we stacked only task-fMRI contrasts together. For generalisability, a stacked model with non-task MRI built from one dataset significantly predicted cognitive abilities in other datasets. Altogether, stacking is a viable approach to undertake the three challenges of BWAS for cognitive abilities. Significance statementScientists have had limited success in predicting cognitive abilities from brain MRI. We proposed a machine learning method, called stacking, to draw information across different types of brain MRI. Using three large databases (n=2,131, 22-100 years old), we found stacking to make the prediction of cognitive abilities 1) closer to actual cognitive scores when applied to a new individual, not part of the modelling process, 2) reliable over times and 3) applicable to the data collected from different age groups and MRI scanners. Indeed, stacking, especially with fMRI task contrasts, allowed us to use MRI of people aged 45 to predict their childhood cognitive abilities reasonably well. Accordingly, stacking may help MRI realise its potential to predict cognitive abilities.

neuroscience↗

bioRxiv2023

The Listeria monocytogenes persistence factor ClpL is a potent stand-alone disaggregase

Heat stress can cause cell death by triggering the aggregation of essential proteins. In bacteria, aggregated proteins are rescued by the canonical Hsp70/AAA+ (ClpB) bi-chaperone disaggregase. Man-made, severe stress conditions applied during e.g. food-processing represent a novel threat for bacteria by exceeding the capacity of the Hsp70/ClpB system. Here, we report on the potent autonomous AAA+ disaggregase ClpL from Listeria monocytogenes that provides enhanced heat resistance to the food-borne pathogen enabling persistence in adverse environments. ClpL shows increased thermal stability and enhanced disaggregation power compared to Hsp70/ClpB, enabling it to withstand severe heat stress and to solubilize tight aggregates. ClpL binds to protein aggregates via aromatic residues present in its N-terminal domain (NTD) that adopts a partially folded and dynamic conformation. Target specificity is achieved by simultaneous interactions of multiple NTDs with the aggregate surface. ClpL shows remarkable structural plasticity by forming diverse higher assembly states through interacting ClpL rings. NTDs become largely sequestered upon ClpL ring interactions. Stabilizing ring assemblies by engineered disulfide bonds strongly reduces disaggregation activity, suggesting that they represent storage states.

biochemistry↗