The turnover of brain proteins is critical for organism survival, and its perturbations are linked to pathology. Nevertheless, protein lifetimes have been difficult to obtain in vivo. They are readily measured in vitro by feeding cells with isotopically labeled amino acids, followed by mass spectrometry analyses. In vivo proteins are generated from at least two sources: labeled amino acids from the diet, and non-labeled amino acids from the degradation of pre-existing proteins. This renders measurements difficult. Here we solved this problem rigorously with a workflow that combines mouse in vivo isotopic labeling, mass spectrometry, and mathematical modeling. We also established several independent approaches to test and validate the results. This enabled us to measure the accurate lifetimes of ~3500 brain proteins. The high precision of our data provided a large set of biologically significant observations, including pathway-, organelle-, organ-, or cell-specific effects, along with a comprehensive catalog of extremely long-lived proteins (ELLPs).

Scaden enables robust cell type deconvolution of complex samples across data types, using a deep learning–based model.

A fundamental problem in biomedical research is the low number of observations available, mostly due to a lack of available biosamples, prohibitive costs, or ethical reasons. Augmenting few real observations with generated in silico samples could lead to more robust analysis results and a higher reproducibility rate. Here, we propose the use of conditional single-cell generative adversarial neural networks (cscGAN) for the realistic generation of single-cell RNA-seq data. cscGAN learns non-linear gene–gene dependencies from complex, multiple cell type samples and uses this information to generate realistic cells of defined types. Augmenting sparse cell populations with cscGAN generated cells improves downstream analyses such as the detection of marker genes, the robustness and reliability of classifiers, the assessment of novel analysis algorithms, and might reduce the number of animal experiments and costs in consequence. cscGAN outperforms existing methods for single-cell RNA-seq data generation in quality and hold great promise for the realistic generation and augmentation of other biomedical data types.

Aging and increased amyloid burden are major risk factors for cognitive diseases such as Alzheimer’s disease (AD). Effective therapies for these diseases are lacking. Here, we evaluated mouse models of age-associated memory impairment and amyloid deposition to study transcriptome and cell type–specific epigenome plasticity in the brain and peripheral organs. We determined that aging and amyloid pathology are associated with inflammation and impaired synaptic function in the hippocampal CA1 region as the result of epigenetic-dependent alterations in gene expression. In both amyloid and aging models, inflammation was associated with increased gene expression linked to a subset of transcription factors, while plasticity gene deregulation was differentially mediated. Amyloid pathology impaired histone acetylation and decreased expression of plasticity genes, while aging altered H4K12 acetylation–linked differential splicing at the intron-exon junction in neurons, but not nonneuronal cells. Furthermore, oral administration of the clinically approved histone deacetylase inhibitor vorinostat not only restored spatial memory, but also exerted antiinflammatory action and reinstated epigenetic balance and transcriptional homeostasis at the level of gene expression and exon usage. This study provides a systems-level investigation of transcriptome plasticity in the hippocampal CA1 region in aging and AD models and suggests that histone deacetylase inhibitors should be further explored as a cost-effective therapeutic strategy against age-associated cognitive decline.