Universität zu Köln

Meyer, David Helmut ORCID: 0000-0002-5667-4720 (2024). Development and Application of Aging Clocks. PhD thesis, Universität zu Köln.

PDF Dissertation_DavidHelmutMeyer_PrintVersion.pdf Download (108MB)

Abstract

Aging clocks have emerged as powerful tools in the field of aging biology. These clocks utilize various biomarkers to estimate the biological age, overall health status, and pace of aging of an organism. Unlike chronological age, which measures linearly the time elapsed since birth, biological age considers factors such as genetics, lifestyle, and environmental exposure that affect the aging process and lead to inter-individual differences. By providing an assessment of an organisms’ health status, aging clocks can aid personalized healthcare, and accelerate aging research by giving surrogate endpoints in clinical trials for the identification and evaluation of geroprotective interventions. Currently, the field primarily focuses on three key areas of research: 1.) the search and validation of accurate biological aging clocks is still ongoing, with various clocks being built for different species and data modalities. 2.) The underlying mechanisms and interpretation of aging clocks is under debate with no clear consensus on what aging clocks are measuring. 3.) And lastly, the use of aging clocks in the identification and evaluation of geroprotective interventions. This is currently largely constrained to their use as surrogate endpoints of clinical trials, limiting their applicability. In this thesis, we first developed an accurate transcriptomic aging clock based on the novel concept of binarization (BitAge). Transcriptomic aging clocks faced limitations due to the inherent variability and age-dependent increase in variation of transcriptomic data, leading to their relative underperformance compared to epigenetic aging clocks. Here, I show that binarizing transcriptomic data enables the usage of transcriptomic data for training aging clocks that rival epigenetic aging clocks. Leveraging existing lifespan data for the nematode Caenorhabditis elegans for temporal rescaling, moreover, allowed highly accurate predictions, not only of the chronological, but especially the biological age. In the second part of this thesis, we investigated the underlying mechanisms of aging clocks and what they ultimately might be measuring. We show that accumulating stochastic variation is sufficient to build aging clocks that accurately predict the chronological and biological age. All tested epigenetic aging clocks, including the most recent pan-mammalian clock, and our own transcriptomic aging clock, correlate with the amount of artificially added stochastic variation to a biological ground state. Surprisingly, we found that an aging clock can be built using just one biological sample and artificially induced stochastic variation accumulation. Even clocks trained with only one biological sample enabled highly correlated predictions with the chronological age and revealed significant differences among samples subjected to lifespan interventions. In the last part of this thesis, we applied our aging clocks to a pseudo-bulk dataset of neuronal cell classes of the nematode Caenorhabditis elegans. We identified almost two-fold aging rate differences 2 between the youngest and oldest predicted neuron classes, and showed that these biological age differences are associated with neurodegeneration in vivo. We then used the predicted age of all neuronal cell classes to identify transcriptomic trajectories over the predicted age (NeuronAge). We show that enriched pathways of genes that are correlated with NeuronAge are conserved in human and mice, thereby bringing the field of cross-species aging transcriptome comparisons to species as far evolutionary apart as Caenorhabditis elegans and humans. Lastly, we performed an in silico drug screen and identified known and novel neuroprotective small molecule compounds that could be validated in vivo, demonstrating that our identified hits do decelerate the age-related neurodegeneration in Caenorhabditis elegans.

Item Type:	Thesis (PhD thesis)
Creators:	Creators Email ORCID ORCID Put Code Meyer, David Helmut david.meyer@uni-koeln.de orcid.org/0000-0002-5667-4720 UNSPECIFIED
URN:	urn:nbn:de:hbz:38-736682
Date:	2024
Language:	English
Faculty:	Faculty of Mathematics and Natural Sciences
Divisions:	CECAD - Cluster of Excellence Cellular Stress Responses in Aging-Associated Diseases
Subjects:	Data processing Computer science Life sciences
Uncontrolled Keywords:	Keywords Language Aging clocks English Bioinformatics English Caenorhabditis elegans UNSPECIFIED Neurodegeneration English
Date of oral exam:	26 July 2024
Referee:	Name Academic Title Schumacher, Björn Prof. Dr. Beyer, Andreas Prof. Dr. Rotblat, Barak Prof. Dr.
Refereed:	Yes
URI:	http://kups.ub.uni-koeln.de/id/eprint/73668