A DNA methylation atlas of human and mouse kidney reveals that epigenetic aging in chronic kidney disease is concentrated almost entirely in renal tubular cells, driven by physical collapse of their 3D chromatin architecture
A DNA methylation atlas of human and mouse kidney reveals that epigenetic aging in chronic kidney disease is concentrated almost entirely in renal tubular cells, driven by physical collapse of their 3D chromatin architecture
On September 24, a team from Altos Labs, Washington University in St. Louis, and Indiana University published a single-cell epigenomic atlas of the kidney: 64,000 nuclei from human and mouse cells, covering DNA methylation, chromatin accessibility, spatial tissue mapping, and three-dimensional genome architecture, all measured in the same samples. In people with chronic kidney disease, renal tubular cells are epigenetically "older" than their calendar age by nearly five years on average. Immune, vascular, and connective-tissue cells in the kidney show no such acceleration.
Renal tubules are part of the nephron, the functional unit of the kidney: the nephron is what reclaims glucose, amino acids, salts, and water from the filtered fluid that would otherwise be lost. Using the sciMET method, the researchers profiled the methylomes (maps of chemical marks on DNA that do not alter the genetic code itself but control which genes are switched on and which are switched off) of 12 human donors (7 with disease, 5 healthy) and 6 mice (3 young, 3 old). The pattern of these marks allows highly accurate estimation of the biological age of a tissue through what are known as epigenetic clocks.
The clocks accelerated in disease in only one cell type: tubular epithelium. This finding was confirmed independently in published sequencing data from 30 kidney tissue samples. Bulk-tissue analysis typically blurs such differences between cell types, a blind spot that a 17-tissue human atlas demonstrated this past summer. In healthy donors aged 20 to 90, the genes that change activity in tubular cells with age track in the same direction as genes that change in the same cells during disease: the tubular tissue of a patient with chronic kidney disease looks as though its normal aging has been sharply accelerated.
The authors then combined methylation data with a physical map of how DNA is folded inside the nucleus, using the scMethyl-Hi-C method. DNA in the nucleus folds into ordered three-dimensional compartments: some regions are physically brought close together and remain accessible for transcription, while others are tightly packed and locked away. In healthy tubular cells, 107 such locked compartments are specific to the cell's identity. In disease, these compartments lose their packing in synchrony with the loss of methyl marks at the same loci, and the unlocking releases genes for senescence-associated inflammatory signals, tissue scarring, and dedifferentiation of tubular cells.
The authors tested this mechanism beyond the laboratory. They compared the list of unlocked genes against blood proteomics from 54,219 UK Biobank participants and found that three proteins encoded by these genes independently predict declining kidney function across the entire population. Separately, tubular epithelium diverged between human and mouse more than any other kidney cell type, specifically in chromatin-packaging and stress-response genes. Basal metabolism is conserved, but the stress-response programs are evolutionarily younger and less refined: the same vulnerable category of genes.
The paper was published in Nature Aging. The lead authors are from the same group of Kun Zhang at Altos Labs and Sanjay Jain who showed in Nature in 2023 that after injury a tubular cell either recovers or becomes trapped in a damaged state. Chronic kidney disease affects roughly 674 million people worldwide, yet until now it was unclear what physically holds a cell in that trapped state. There is now a concrete answer: collapse of the three-dimensional chromatin architecture in the very regions that make the cell what it is.