A preprint reports age-related changes in just 0,05% of DNA methylation sites examined in king penguin blood, compared with 3,7% in humans, with changed sites enriched near genes associated with hematopoietic stem cells
A preprint reports age-related changes in just 0,05% of DNA methylation sites examined in king penguin blood, compared with 3,7% in humans, with changed sites enriched near genes associated with hematopoietic stem cells
On 13 October, Robin Cristofari’s group at the University of Helsinki posted a preprint on bioRxiv. Using 126 blood samples from king penguins of known age, the authors looked for genes with age-related changes in DNA methylation, the chemical marks on DNA.
Epigenetic clocks estimate age from DNA methylation in blood, but what they measure remains debated. Almost all DNA in human blood comes from leukocytes, because red blood cells lack nuclei. Each leukocyte type has its own methylation pattern, and their relative proportions change with age. One explanation is that the clocks detect these changes in cell composition. In an analysis of more than 10 000 samples, the difference between estimated and chronological age was associated with immune cell proportions for all six clocks tested. In an August preprint covered by Ukhvat, changes in cell composition explained most of the change over a decade at 48 DNA sites. Another explanation is that blood cells inherit methylation marks from their precursors, hematopoietic stem cells. Support for this explanation has come mainly from computational models.
In adult birds, erythrocytes retain their nuclei and account for 99,5–99,9% of blood cells, while leukocytes account for 0,1–0,5%. Changes in leukocyte proportions therefore have little effect on blood methylation. The authors predicted that if cell composition explained the age signal, that signal would largely disappear. If the stem cell explanation held, a signal associated with those cells would remain.
An age signal remained, although it was much smaller: methylation changed significantly with age at just 0,05% of the DNA sites examined. In human data subsampled to a comparable sample size, the average was 3,7%, or more than 75 times as many. The authors interpret changes in cell composition as one of the main sources of the age signal in mammals. The two explanations are complementary: changes in cell composition may themselves result from stem cell aging.
In birds older than 2,5 years, after growth has stopped, methylation changes consistently with age at 409 genes. Genes used to identify hematopoietic stem cells are 3,3 times as common in this set as expected. The researchers examined the activity of these genes in mature blood cells from four zoo birds. According to the authors, many of the strongest age signals occur at genes that are silent in mature erythrocytes, including MECOM and PRDM16, which help maintain hematopoietic stem cells in mice. Because erythrocytes do not use these genes, the authors hypothesize that the methylation marks were inherited from precursor cells. They call this the palimpsest model. A palimpsest is a manuscript on which new text has been written over erased text that remains legible: the mature cell’s program is written over inherited marks, but those marks remain readable at silent genes.
To test whether the same genes show age-related changes in humans, the researchers used data from 1 501 participants in the Finnish Twin Study. After adjustment for blood cell composition, 60 of the 357 human genes corresponding to the penguin genes remained associated with age. They used methylation marks in these genes to build a clock called PenguinAge: if the marks come from stem cells, a clock based on them should depend less on leukocyte composition. In 229 participants whose data were not used to train the clock, the mean error was 2,72 years. The difference between the clock’s estimate and chronological age was less strongly associated with leukocyte proportions than for 96,7% of clocks built from random sets of 409 genes. The authors suggest that such clocks could be used to monitor stem cell exhaustion and immune aging.