Embryonic senescent cells re-enter cell cycle and contribute to tissues after birth

Dear Editor, Cellular senescence (or senescence) has been regarded as a stable form of cell cycle arrest by in vitro cell culture experiments. Recent studies indicate that senescence is associated with aging and diseases, including cancers. For instances, it suppresses tumor progression by halting the growth of premalignant cells, and promotes wound healing by preventing excessive tissue fibrosis or induction of cell dedifferentiation. Targeting senescent cells could restore tissue homeostasis in response to aging, chemotoxicity, or injury. In addition to these pathological conditions in adults, cellular senescence also occurs in physiological states such as mammalian mouse and human embryonic development. Embryonic senescent cells have been reported to be non-proliferative and subjected to clearance from tissues after apoptosis at late embryonic stage. However, the interpretation for clearance of senescent cells at late embryonic stage is based on the disappearance of Cdkn1a (P21) expression and senescence-associated beta-galactosidase (SAβ-Gal) activity, two commonly used senescence markers in the field. Currently, there is no genetic fate mapping evidence for senescent cell fate in vivo. By lineage tracing of P21 senescent cells, we found that embryonic senescent cells labeled at mid-embryonic stage gradually lost P21 expression and SAβ-Gal activity at late embryonic stage. Unexpectedly, some of the previously labeled senescent cells re-entered the cell cycle and proliferated in situ. Moreover, these previously labeled senescent cells were not cleared at late embryonic stage and remained in the tissue after birth. This study unravels in vivo senescent cell fates during embryogenesis, indicating their potential plasticity. We first performed SAβ-Gal staining on embryos and found SAβGal signals in the apical ectodermal ridge (AER) at E10.5–E14.5. We hardly detected positive signals in the AER at E15.5 and afterwards (Fig. 1a). SAβ-Gal activity in AER was validated by staining on tissue sections (Supplementary information, Figure S1a). To confirm the specificity of SAβ-Gal staining for senescence (pH 6.0), we stained embryos at pH 6.5 and pH 7.0 for technical controls as previously described. Indeed, we did not detect any positive SAβ-Gal signal at E10.5–E14.5 (Supplementary information, Figure S1b). These results were consistent with previous studies, demonstrating that senescent cells as detected by SAβ-Gal staining were present at E10.5–E14.5, whereas SAβ-Gal activity disappeared after E15.5 (Fig. 1a, b). Therefore, SAβ-Gal activity could be mainly restricted to midbut not late embryonic stage. These experimental data have been interpreted as indicating that SAβ-Gal senescent cells underwent apoptosis and were cleared from tissues at late embryonic stage. However, an alternative explanation could be that a subset of senescent cells gradually lost SAβ-Gal activity but survived in the tissue at late embryonic stage. The in vivo senescent cell fate currently remains unknown and untested, as to date there is no fate mapping study on senescent cells. P21 is a molecular mediator for embryonic senescence and is highly expressed in SAβ-Gal senescent cells of AER. To trace the cell fate of senescent cells during embryogenesis, we generated the P21-CreER mouse line by knocking CreER cDNA into the stop codon of P21 (Fig. 1c). 2A self-cleaving peptide sequence was used to allow simultaneous expression of CreER and P21 in P21 cells (Fig. 1c). Immunohistochemistry for P21 or estrogen receptor (ESR, for detection of CreER) in P21-CreERmouse forelimbs showed their similar expression patterns to SAβ-Gal activity pattern in embryonic limbs (Fig. 1d, compare to Fig. 1a), suggesting that senescent cells of AER expressed high levels of P21 at mid-stage (e.g., E10.5–E13.5). The CreER expression in embryonic forelimbs at E10.5–E13.5 was largely within AER, recapitulating endogenous P21 expression (Fig. 1d, e). However, the expression of both CreER and P21 was reduced at E14.5 and not detected at E15.5 (Fig. 1d, e). These data demonstrated that CreER was successfully knocked in at the P21 gene locus (Fig. 1c). We further validated coexpression of P21 and ESR (CreER) in AER by immunostaining (Fig. 1f). Taken together, the above data showed that SAβ-Gal activity and P21 gene expression were highly restricted to senescent cells at AER at mid-embryonic stage, consistent with the previous study. We next crossed P21-CreER with R26-tdTomato reporter for genetic lineage tracing of P21 senescent cells during embryogenesis. Tamoxifen pulse treatment leads to translocation of CreER into the nucleus of P21 cells, allowing subsequent Cre-loxP recombination to remove the transcriptional stop region for tdTomato expression (Fig. 1g). We administered tamoxifen at E10.5 or E11.5 to label P21 senescent cells at mid-embryonic stage and then collected tissue samples from E12.5 to birth (P0) for analysis (Fig. 1h). Without tamoxifen treatment, embryos exhibited no detectable tdTomato signal (Fig. 1i), indicating no leakiness of P21-CreER. In tamoxifen-treated samples, we could readily detect tdTomato cells in the AER of developing forelimbs at mid-embryonic stage (Fig. 1j). To confirm that the labeled P21 cells after tamoxifen treatment were indeed senescent cells, we first isolated tdTomato and tdTomato cells from E12.5 limbs by fluorescence-activated cell sorting (FACS, Supplementary information, Figure S2a). By co-staining with SAβ-Gal and tdTomato on FACS-isolated cells, we detected SAβ-Gal activity in tdTomato cells but not in tdTomato cells (Supplementary information, Figure S2b), demonstrating that P21/tdTomato cells at E12.5 were senescent cells. The co-expression of SAβ-Gal, P21, and ESR was also confirmed by immunostaining on consecutive sections of P21-CreER limbs (Supplementary information, Figure S2c). To further demonstrate that the P21 cells labeled at early stage were senescent cells, we generated P21-tdTomato knock-in allele by targeting tdTomato cDNA into the stop codon of P21 with addition of 2A self-cleaving peptide sequence (Supplementary information, Figure S2d). Immunostaining and FACS analysis


Experimental mice
All mice used in our experiments were C57BL/6J background. All mouse operations were done according to the guidelines of the Institutional Animal Care and Use Committee (IACUS) at the Institute for Nutritional Sciences and Institutes for Biochemistry and Cell Biology, Shanghai institutes for Biological Sciences, Chinese Academy of Sciences. Detection of vaginal plug in the morning was designated as E0.5. Tamoxifen (Sigma: T5648-5G) treated by oral gavage at the indicated times (0.05-0.10 mg/g). R26-tdTomato mouse line was reported previously 1  generate new knockin mouse lines in this study.

Genomic PCR
Genomic DNA was extracted from embryonic yolk sac or mouse tail. Tissues were incubated in proteinase K overnight at 55 °C, followed by centrifugation at 13000rpm for 8min to obtain supernatant with genomic DNA. The supernatant was added equal volume isopropanol to precipitate DNA and washed in 70% ethanol. These steps need centrifugation at 13000rpm for 3min to obtain precipitate with genomic DNA. All DNA was genotyped with specific primers that distinguished the knock-in allele from the wild-type allele. For R26-tdTomato line, primers

SAβ-Gal staining
SAβ-Gal staining was performed according to the Senescence β-Galactosidase Staining Kit (Cell Signaling Technology, #9860). Embryos and tissue sections were incubated in PBS and then fixed in 1X fixative solution (need to dilute 10X fixative solution to 1X with distilled water) at room temperature for 10-15min. After two times washing in PBS, embryos and tissue sections were treated with X-Gal staining cocktail (adjusted pH to 6.0 using NaOH or HCl) at 37°C for 5-6 hours. X-Gal staining cocktail needs to be prepared by mixing with distilled water, 10X staining solution, 100X solution A and 100X solution B. The staining circumstances need to be in low CO 2 condition. Embryos and tissues sections were washed in PBS to stop the reaction and then fixed in 4% fresh PFA for 1 hour. Pictures were taken using a Zeiss stereo microscope (AxioZoom V16).

Whole-mount P21 and ESR staining
Whole-mount immunohistochemistry was performed as previously described 2 . Embryos were collected in PBS and then fixed in 4% paraformaldehyde (PFA) overnight at 4°C. After three times washing with PBT (PBS with 0.1% Tween20) for 10 min at 4°C, embryos were sequentially dehydrated through 25%, 50%, 75%, 100% methanol for 15 min at room temperature.
Embryos were then left in PBT overnight with constant rotation. Embryos were incubated with secondary antibodies (Immpress anti-rabbit, Vector lab, MP-7401-50; Immpress anti-rat, Vector lab, MP-744-15) for 2 hours at room temperature, followed by 4 times of washing in PBT for 1 hour at 4°C with constant rotation. Embryos were then left in PBT overnight with constant gentle rotation. In the next day, the ImmPACT DAB kit was used to for whole-mount immunohistochemistry. After signal development, the embryos were fixed in the 4% PFA for 1 hour at room temperature. Images were taken using a Zeiss stereo microscope (AxioZoom V16).

TUNEL assay
Tissue sections were fixed in 4% PFA for 10 min at room temperature, followed by washing in PBS for 15 min. The slides were then incubated in PBST (PBS with 0.1% triton X-100) for 2 min at 4°C. Tissue sections were incubated in TUNEL fluorescein (50 µl enzyme solution dilute in 450 µl label solution, Roche, 11684795910) for 1 hour at room temperature. After sections were washed in PBS, immunofluorescent staining was performed at following steps. Images were acquired on Olympus confocal microscope (FV1200).

EdU staining
Pregnant mice were treated with 10 µg/g EdU (ThermoFisher, A10044) at 2 hours before collected through intraperitoneal injection (for quantification of total proliferated ratio, 10 µg/g EdU was treated everyday from E12.5). Tissue sections were blocked in PBSST (5% normal donkey serum in PBST) for 1 hour at room temperature. Click-iT reaction cocktail (Invitrogen, C10337) need to be prepared in 15 minutes before used. Slides were incubated in Click-iT reaction cocktail for 30 min at room temperature in dark, followed by three times washing in PBS for 10 min at room temperature. Immunofluorescent staining was performed at following steps.
Images were acquired on Olympus confocal microscope (FV1200).

Immunofluorescent staining
Immunofluorescent staining was performed as previously described 3 . Embryos or tissues were collected in PBS, and then fixed in 4% PFA for 30-50 minutes. After three washes in PBS, embryos or tissues were dehydrated in 30% sucrose (dissolved in PBS) overnight and embedded in OCT (Sakura). Next, cryosections of 9-10µm in thickness were collected. After air dry for 1-1.5 hour at room temperature, sections were incubated with 30% H 2 O 2 for 10 minutes. After two times of washing in PBS for 10 min at room temperature, slides were blocked in blocking buffer (5% donkey serum, 0.1% Triton X-100 in PBS) for 30 min at room temperature, followed by tissue sections were mounted with mounting medium containing the nuclear stain DAPI (Vector Lab). Images were acquired on Olympus confocal microscope (FV1200). The obtained images were analyzed by ImageJ software (NIH).

FACS and SAβ-Gal assay
The forelimbs were harvested from E12.5 embryos, and then finely minced and transferred to the digestion mix (RPMI1640, 5% FBS, 1% PSG and 300U Collagenase3). Tissues were incubated with the mix for 20-30 minutes at 37°C with gentle shaking. Tissue mix was centrifuged (1000rpm, 5min) and the precipitate was re-suspended in 2ml pre-warmed 0.05% trypsin/EDTA, The FACS-isolated tdTomato + and tdTomatocells were stained with SAβ-Gal using SPiDER-βGal kit (SG03, Dojindo). Cells were cultured in the dish at 37°C overnight in a 5% CO 2 incubator. The cultured cells were washed and then 1ml Bafilomycin A1 working solution was added for incubation at 37°C for 1h in a 5% CO 2 incubator to inhibit endogenous β-galactosidase activity. Bafilomycin A1 working solution was discarded and cells were incubated with 1ml SPiDER-βGal working solution for 2 hours at 37°C in a 5% CO 2 incubator. After supernatant removal, cells were washed with culture medium twice for 1 hour. The stained cells were fixed at 4% PFA for 5-10 minutes at room temperature, and then stained with DAPI (1:2000) for 15min at room temperature. Images were acquired on Olympus confocal microscope (FV1200).

Statistical analysis
All data were representative of 5-6 individuals, as indicated in each figure legend, and presented as mean values ± S.D.. Statistical comparisons between different sets were made with analysis of normalization and variance, followed by two-sided unpaired Student's t-test for analyzing differences between two data groups. P value < 0.05 was considered to be statistically significant difference. All mice were randomly assigned to different experimental groups. No statistical method was used to predetermine sample size.