mTert induction in p21-positive cells counteracts capillary rarefaction and pulmonary emphysema

Lung diseases develop when telomeres are shortened beyond a critical point. We have constructed a mouse model in which the catalytic subunit of telomerase (mTert), or its catalytically inactive form (mTertCI), is expressed from the p21Cdkn1a promoter. We found that this particular expression of mTert reduces senescence of endothelial cells (EC) in lungs of aged mice, as well as emphysema and pulmonary perivascular fibrosis. We also show that mTert counteracts the decline in capillary density in aged mice and promotes the maintenance of high numbers of Cd34+ cells, identified as a subclass of endothelial cells with proliferative capacity. In line with these results, young p21+/Tert mice treated with a VEGF receptor inhibitor combined with hypoxia are also protected against senescence and emphysema induced by this treatment. The catalytic activity of mTert is required for all the effects observed. However, and unexpectedly, we found that both mTert and mTertCI expression significantly reduced p21 levels in the lungs of aged mice. mTert thus protects against age-related and induced loss of capillary vessels and subsequent lung emphysema. Droplet-based single-cell RNA-sequencing on 3 samples of total cells from mouse lung tissue. Three genotypes were simultaneously analyzed. Single-cell 3'-RNA-Seq samples were prepared using single cell V2 reagent kit and loaded in the Chromium controller according to standard manufacturer protocol (10x Genomics, PN-120237) to capture 6.000 cells. Briefly, dissociated lung cells are encapsulated using microfluidic device. RNAs are captured on beads coated of oligos containing an oligo-dTTT, UMIs and a specific barcode. After reverse transcription, cDNAs are washed, PCR-amplified and washed again before analysis on a Bioanalyzer (Agilent) for quality control. Finally, libraries are prepared following standard Illumina protocol and sequenced on a NovaSeq sequencer (Illumina). Raw sequences are demultiplexed and reads are mapped onto the mm10 reference genome using the v2.3 Cell Ranger pipeline (10X Genomics) to generate a count matrix for each sample. The digital matrices were filtered by cell type (on clusters composed of the same cell type), to remove low-quality cells with low UMI counts and cells with relatively high mitochondrial DNA content. Outlier analysis was performed with perCellQCMetrics from the scatter package. An upper cutoff was manually determined for each sample based on a plot of gene count versus UMI count or % of mito-chondrial genes, to have at least 1000 UMIs, number of transcripts ranging between 1000 and 30000 and at most 14% mitochondrial transcripts. The quality was consistent across samples, and differences in RNA and gene content could be ascribed to cell-type-specific effects. DGE matrices from all sam-ples, sequenced at different time, were then merged and subsequently normalized using the deconvo-lution normalization method in the scran R package in order to correct for differences in read depth and library size inside and between samples.