Preprint: a new DNA methylation index of cell specialization changed little during partial reprogramming, while epigenetic clocks indicated rejuvenation
Preprint: a new DNA methylation index of cell specialization changed little during partial reprogramming, while epigenetic clocks indicated rejuvenation
On October 11, Andrew Teschendorff’s group at the Chinese Academy of Sciences posted a manuscript describing the EpiDiff model on bioRxiv before peer review. The model uses DNA methylation, chemical marks on DNA, to calculate a cell specialization index ranging from 0 (a pluripotent cell capable of becoming any cell type) to 1 (a fully specialized cell).
Epigenetic clocks estimate cell age from the same marks. During partial reprogramming, which involves several days of exposure to factors that convert cells to a pluripotent state, these clocks indicate rejuvenation. Full reprogramming erases cell specialization, so partial reprogramming is stopped earlier. The index is intended to detect this loss of specialization: the authors of the preprint developed it for use alongside epigenetic clocks.
The index was trained on 146 samples: 25 pluripotent cell samples, 86 fetal tissue and cord blood samples (representing an intermediate state), and 35 adult tissue samples. It uses 1298 DNA sites where methylation differs from that in pluripotent cells by at least 30 percentage points in fetal tissues and 50 in adult tissues. The authors chose these thresholds to exclude changes caused by cell division. The index responds to loss of specialization: cells fully reprogrammed to pluripotency have very low values.
The authors applied the index to data from two previous studies. In the 2020 study, fibroblasts (skin cells) and cells lining blood vessels received six reprogramming factors for four days. The clocks estimated that treated cells were 3,4 years younger, while the index remained unchanged. In the 2022 study, fibroblasts from donors aged 38–53 years received the factors for 10–17 days, compared with approximately 50 days for full reprogramming. Cells selected for signs of successful reprogramming were then cultured without the factors. Their index remained close to, though below, the value of the original fibroblasts. The preprint authors conclude that partial reprogramming largely preserves cell specialization.
The index increases with age even within individual cell types isolated before measurement: liver cells, bronchial epithelial cells, astrocytes (supporting cells in the brain), and naive T cells that have not encountered a pathogen. The index combines signs of specialization across all cell lineages: in a simulated mixture of two mature cell types, it is higher than in either type alone, so the presence of other mature cells raises it. The authors attribute the increase in naive T cells to this admixture: mature cells also carry the marker used to select naive T cells, and their proportion increases with age.
The authors explain the increase in liver cells, bronchial epithelial cells, and astrocytes as drift: with age, DNA methylation in these cells shifts toward patterns associated with other cell lineages. To test this explanation, they selected sites among the 1298 that distinguish the embryonic germ layers: 247 are characteristic of the mesoderm (which gives rise to muscle, bone, and blood), 139 of the ectoderm, and 70 of the endoderm. In most datasets, age-related methylation changes at these sites follow the same direction as changes during embryonic cell specialization more often than changes at randomly selected sites do. This pattern is strongest at mesoderm-associated sites; in astrocytes, it occurs only at those sites.
The authors call this epigenetic drift toward mesenchyme, the embryonic connective tissue. Biologist Juan Carlos Izpisua Belmonte’s group reported similar drift in gene activity across more than 40 human tissues in the journal Cell: partial reprogramming reduced this drift before cells lost their specialization.