Multi‐omic rejuvenation of naturally aged tissues by a single cycle of transient reprogramming

A single period of transient OSKM expression in naturally aged mice is sufficient to reverse DNA methylation changes that occur upon aging in the pancreas, liver, spleen, and blood.

“We found that this is sufficient to reverse DNA methylation changes that occur upon aging in the pancreas, liver, spleen, and blood.”

DM enhancer at Chr10:68231400-68232600 showed a trend to be demethylated after OSKM activation.

“For example, DM enhancer at Chr10: 68231400–68232600 showed a trend to be demethylated after OSKM activation (Figure 1d).”

No remethylation was detected in these two positions after transient OSKM activation.

“However, we could not detect remethylation in these two positions after transient OSKM activation (Figure S1n).”

PCA showed that aging-DM enhancers in OSKM mice were distinct from old mice and displaced toward the young group.

“As before, PCA analysis showed that aging‐DM enhancers in OSKM mice were distinct from old mice and displaced toward the young group (Figure 1c).”

Transient OSKM expression in naturally aged mice reverts transcriptional changes, especially regarding biological processes known to change during aging.

“Similarly, we observed reversion of transcriptional changes, especially regarding biological processes known to change during aging.”

Out of nine CpGs expected to gain methylation with aging, eight showed the expected increase in methylation with aging, specifically near genes Tpt1 (2 close CpGs), Stambpl1, Dnmbp (2 consecutive CpGs), Akt1, Gm12339, and Gm17678.

“Out of them, eight showed the expected increase in methylation with aging, specifically, those near genes Tpt1 (2 close CpGs), Stambpl1, Dnmbp (2 consecutive CpGs), Akt1, Gm12339, and Gm17678.”

The methylation profile of old-OSKM mice was placed by PCA between the young and old groups, revealing partial epigenetic rejuvenation of aging-sensitive promoters.

“Interestingly, the methylation profile of old‐OSKM mice was placed by PCA between the young and old groups, revealing partial epigenetic rejuvenation of aging‐sensitive promoters (Figure 1a).”

Pyrosequencing confirmed that two aging-hypomethylated CpGs (1700011F14Rik and Ptprj) were hypomethylated with aging.

“We also tested two aging‐hypomethylated CpGs (1700011F14Rik and Ptprj) and pyrosequencing confirmed that both were hypomethylated with aging.”

Histological changes observed in the pancreas after one week of OSKM induction are reversed two weeks after removal of doxycycline.

“Importantly, the histological changes observed in the pancreas after one week of OSKM induction were reversed two weeks after removal of doxycycline (Figure S1a).”

All eight regions presented reduced methylation in very old-OSKM mice treated with doxycycline, in some cases becoming indistinguishable from young mice.

“Interestingly, all eight regions presented reduced methylation in very old‐OSKM mice treated with doxycycline, in some cases becoming indistinguishable from young mice (Figure 1g, Figure S1l).”

Reversion of methylation was consistently more profound after 4 weeks of recovery than after 2 weeks of recovery.

“Of note, reversion of methylation was consistently more profound after 4 weeks of recovery than after 2 weeks of recovery.”

By comparing old and young pancreata, a list of differentially methylated (DM) promoters was identified.

“By comparing old and young groups, we identified a list of differentially methylated (DM) promoters (Figure S1f).”

Detection of pluripotency markers in pancreas requires two weeks of OSKM expression.

“Of note, the detection of pluripotency markers in pancreas required two weeks of OSKM expression (Figure S1b).”

Upon OSKM induction for 1 week in 55-week-old reprogrammable mice, clear histological changes are observed in the pancreas.

“Upon OSKM induction for 1 week, we observed clear histological changes in the pancreas (Figure S1a).”

Some serum metabolites and biomarkers altered with aging are restored to young levels upon transient reprogramming.

“Finally, some serum metabolites and biomarkers altered with aging were also restored to young levels upon transient reprogramming.”

Using a separate cohort of old mice and a different technique (pyrosequencing), rejuvenation was confirmed in 8 out of 11 tested regions that showed aging-dependent methylation changes.

“Using a separate cohort of old mice and a different technique (pyrosequencing), we confirmed rejuvenation in 8 out of 11 tested regions that showed aging‐dependent methylation changes.”

PCA of aging-DM promoters clearly segregated young from old mice.

“Focusing on these aging‐DM promoters, we performed principal component analyses (PCA) that clearly segregated young from old mice.”

The majority of OSKM-induced demethylation in the pancreas happened after turning off OSKM expression, during the 2 weeks of recovery period.

“We observed that the majority of OSKM‐induced demethylation in the pancreas happened after turning off OSKM expression, that is, during the 2 weeks of recovery period.”

The large majority of genes upregulated by OSKM in old mice corresponded to genes with reduced expression upon aging.

“Interestingly, the large majority of genes upregulated by OSKM in old mice (pink dots) corresponded to genes with reduced expression upon aging (Figure 2b, Figure S2a).”

A total of 23 metabolites were significantly changed between young and very old mice when combining two independent serum metabolomics experiments.

“When combining the two independent experiments, a total of 23 metabolites were identified as significantly changed between young and very old mice (Figure S5a,b, Table S10).”

In spleen, PCA analysis of aging-DM promoters and enhancers did not allow inferring epigenetic rejuvenation of the tissue.

“In the spleen, PCA analysis of aging‐DM promoters and enhancers did not allow to infer an epigenetic rejuvenation of this tissue (Figure S4a,e)”

Old-OSKM mice increased their levels of collagen-related genes.

“and, remarkably, old‐OSKM mice increased their levels of collagen‐related genes (Figure 2h, Figure S2j).”

When the same analysis was performed after randomizing the samples, only 44 gene-sets were found.

“When the same analysis was performed after randomizing the samples (i.e., samples were randomly assigned to the three experimental groups: young, old, and old‐OSKM), only 44 gene‐sets were found (see pattern 1 in Table S7).”

Pattern analysis identified a total of 179 gene-sets rejuvenated by OSKM.

“This analysis identified a total of 179 gene‐sets rejuvenated by OSKM (see pattern 1 in Table S6).”

Four aging-associated serum metabolites-4-hydroxyproline, thymine, trimethyl-lysine, and indole-3-propionic acid-were reversed after reprogramming in two cohorts of very old mice.

“Out of these aging‐associated metabolites, 4 were reversed after reprogramming (Figure 5a).”

• Also stated in: § Serum metabolomic profiling reveals systemic benefits · plain

  “These metabolites are 4‐hydroxyproline, thymine, trimethyl‐lysine, and indole‐3‐propionic acid”

• Also stated in: § DISCUSSION · plain

  “We have detected four serum metabolites that changed with aging and were reversed by OSKM in the two cohorts of old mice: 4‐hydroxyproline, thymine, trimethyl‐lysine, and indole‐3‐propionic acid.”

• AssumptionRests on an assumption: The serum metabolome is a valid readout of systemic rejuvenation.

The total difference in Hsf4 methylation between reprogrammable and non-reprogrammable very old mice was 8%.

“resulting in a total difference of 8% methylation between the reprogrammable and the non‐reprogrammable mice (Figure 4b).”

In spleen, OSKM transient expression rescued the age-associated decline of seven genes from the Mouse Aging Cell Atlas.

“Interestingly, OSKM transient expression rescued the age‐associated decline of seven of these genes while the senescent marker p16INK4a (Cdkn2a) was unchanged (Figure 4c, Figure S4i).”

In spleen, the senescent marker p16INK4a (Cdkn2a) was unchanged after OSKM.

“Interestingly, OSKM transient expression rescued the age‐associated decline of seven of these genes while the senescent marker p16INK4a (Cdkn2a) was unchanged (Figure 4c, Figure S4i).”

All these gene-set alterations were ameliorated in old-OSKM mice.

“All these gene‐set alterations were ameliorated in old‐OSKM mice (Figure 2e, Figure S2b‐f).”

In two cohorts of very old OSKM mice, liver levels of p16INK4a, Mcp1, and Cxcl2 did not decline after one cycle of OSKM.

“As expected, all of these were increased with aging; however, their levels did not decline in any of the two cohorts of very old‐OSKM mice (Figure 3c, Figure S3h,i).”

In reprogrammable very old mice, Hsf4 methylation decreased by ~4% over the same 5-week period.

“In contrast, during the same period of time, reprogrammable mice reduced their average methylation levels by ~4%”

• Also stated in: § DISCUSSION · plain

  “This was in contrast to mice exposed to one cycle of reprogramming (1 week of OSKM expression followed by 4 weeks of recovery) in which Hsf4 methylation was decreased or maintained.”

Old-OSKM samples behaved like young samples when compared to untreated old control mice.

“Notably, old‐OSKM samples behaved like young samples when compared to untreated old control mice (Figure 2c‐d, Table S5).”

By comparing transcriptome of old versus young pancreata, a list of significantly differentially expressed genes (aging-DEGs, n=217) was identified.

“By comparing the transcriptome of old versus young samples, we identified a list of significantly differentially expressed genes (aging‐DEGs, n = 217) (Table S4).”

PCA of aging-associated DEGs clearly separated old and young samples.

“PCA analysis of these aging‐associated DEGs clearly separated old and young samples as expected.”

PCA placed old-OSKM profiles between old and young ones, suggesting that transient OSKM activation partially restores aging-related transcriptomic changes.

“Remarkably, PCA placed old‐OSKM profiles between old and young ones, suggesting that transient OSKM activation partially restores aging‐related transcriptomic changes (Figure 2a).”

Five of 15 mRNAs from the Mouse Aging Cell Atlas signature were confirmed to be downregulated in the liver of old mice.

“Five of these genes were confirmed to be downregulated in the liver of our old mice (Figure 3c).”

The DNA replication gene-set was downregulated in old versus young control mice.

“whereas the DNA replication gene‐set was downregulated (Figure 2d).”

All five genes recovered young levels of expression in the livers of old-OSKM mice.

“Interestingly, all these five genes recovered young levels of expression in the livers of old‐OSKM mice (Figure 3c).”

From the total combined number of aging-DM promoters and enhancers in liver (n=108), a substantial fraction (61%) underwent rejuvenation.

“From the total combined number of aging‐DM promoters and enhancers (n = 108), a substantial fraction (61%) underwent rejuvenation (Figure 3b, Figure S3a‐f, Table S8).”

Protein homeostasis was improved in old-OSKM mice.

“Again, this process was improved in old‐OSKM mice (Figure 2g, Figure S2i).”

Genes downregulated by OSKM in old mice were genes with upregulated expression upon aging.

“Conversely, those genes downregulated by OSKM in old mice (blue dots) were genes with upregulated expression upon aging (Figure 2b, Figure S2a).”

Nrf2 expression was reset to young levels in very old-OSKM spleens.

“Finally, Nrf2 also declined in very old control spleens and its expression was reset to young levels in very old‐OSKM spleens (Figure 4d).”

In control non-reprogrammable very old mice, Hsf4 methylation increased by 4% over 5 weeks.

“Interestingly, it was possible to detect a 4% increase in Hsf4 methylation in control non‐reprogrammable mice (Figure 4b).”

• Also stated in: § DISCUSSION · plain

  “In the case of control mice, we observed an increase in methylation during this period of 5 weeks.”

These findings were further validated in an independent cohort of very old mice (100 weeks).

“These findings were further validated in an independent cohort of very old mice (100 weeks) (Figure S3h).”

The number of γH2AX-positive cells in liver was similar in very old mice with or without one cycle of OSKM.

“We also measured the number of γH2AX‐positive cells, a marker of in vivo senescence, and again the levels in the liver were similar in very old mice with or without 1 cycle of OSKM (Figure S3j).”

• AssumptionRests on an assumption: γH2AX-positive cells are a marker of in vivo senescence.

Nrf2 declined in very old control spleens.

“Finally, Nrf2 also declined in very old control spleens and its expression was reset to young levels in very old‐OSKM spleens (Figure 4d).”

Hsf4 methylation in blood increases linearly with age from ~20% in young mice (~10 weeks) to ~55% in very old mice (~100 weeks).

“First, we measured by pyrosequencing the methylation levels of the Hsf4 region in the blood of mice of different ages, confirming a linear change of large magnitude, from ~20% methylation in young mice (~10 weeks) to ~55% methylation in very old mice (~100 weeks) (Figure 4a).”

• Also stated in: § DISCUSSION · boosted

  “Indeed, we found that the degree of methylation of these sites in peripheral blood is tightly linked to the age of the mice.”

The mTOR signaling gene-set was upregulated in old versus young control mice.

“As expected, the mTOR signaling gene‐set was upregulated in old versus young control mice (Figure 2c), whereas the DNA replication gene‐set was downregulated (Figure 2d).”

Serum levels of transaminases (AST/GOT and ALT/GPT) of OSKM mice were significantly lower after 1 week of OSKM activation and 2 weeks of recovery.

“Interestingly, the serum levels of transaminases of OSKM mice were significantly lower after 1 week of OSKM activation and 2 weeks of recovery reflecting an improved liver function (Figure 3e).”

Up to 163 promoter and enhancer regions in spleen showed evidence of rejuvenation according to their average methylation level.

“although up to 163 promoter and enhancer regions showed evidence of rejuvenation according to their average methylation level (Figure S4b‐d,f‐h, Table S9).”

Lower levels of serum transaminases (ALT/GPT and AST/GOT) were detected after one cycle of reprogramming.

“We also detected lower levels of serum transaminases (ALT/GPT and AST/GOT) after one cycle of reprogramming.”

Spleen and liver showed evidence of rejuvenation of a transcriptional signature of aging derived from the Mouse Aging Cell Atlas.

“Nevertheless, spleen and liver did show evidence of rejuvenation of a transcriptional signature of aging derived from the Mouse Aging Cell Atlas (Zhang et al., 2021).”

• cites Zhang et al., 2021Not checked yet

  Not checked in the free run.

No evidence of reduction in senescence markers was observed.

“We did not observe however evidence of reduction in senescence markers.”

The 2-4 week recovery period was sufficient for full restoration of normal histology in the pancreas.

“This recovery period was sufficient for a full restoration of normal histology in the pancreas.”

About half of the methylation changes observed at the time of switching off OSKM disappeared after 2 weeks.

“Notably, about half of the changes observed at the time of switching off OSKM disappeared after 2 weeks.”

No associations were established between specific aging-DM regions and mRNA levels of the associated genes.

“We have not been able to establish associations between specific aging‐DM regions and the mRNA levels of the associated genes.”

One cycle of OSKM (1 week induction and 2 weeks recovery) was sufficient to reset the levels of aging-DEGs to a younger state in pancreas.

“Similar to the methylation data, one cycle of OSKM (1 week of induction and 2 weeks of recovery) was sufficient to reset the levels of aging‐DEGs to a younger state.”

In pancreas, the collagen synthesis gene set was downregulated with aging and rescued by OSKM.

“Notably, our transcriptomic analyses identified collagen synthesis as a gene‐set downregulated with aging and rescued by OSKM in the pancreas (Figure 2h).”

• Also stated in: § DISCUSSION · plain

  “Of note, the transcriptional network for collagen synthesis was downregulated with aging and rescued by OSKM in the pancreas.”

In spleen and liver, methylation changes in aging-DM regions were not as pronounced as in pancreas, as judged by principal component analyses.

“In the case of the spleen and liver, the methylation changes observed in aging‐DM regions were not as pronounced as in the case of the pancreas, as judged by principal component analyses.”

The majority of gain-of-methylation events observed after recovery were already present at the time of extinguishing OSKM expression.

“The majority of gain‐of‐methylation events observed after recovery were already present at the time of extinguishing OSKM expression.”

Pyrosequencing measurement of individual aging-methylated CpGs consistently showed more demethylation 4 weeks post-OSKM expression compared to 2 weeks.

“In line with these observations, pyrosequencing measurement of individual aging‐methylated CpGs consistently showed more demethylation 4 weeks post‐OSKM expression compared to 2 weeks.”

A single period of OSKM expression can drive epigenetic, transcriptomic, and metabolomic changes toward a younger configuration in multiple tissues and in the serum.

“These observations indicate that a single period of OSKM expression can drive epigenetic, transcriptomic, and metabolomic changes toward a younger configuration in multiple tissues and in the serum.”

In mice, cycles of OSKM in adult individuals improve their capacity to respond to tissue injury.

“In mice, cycles of OSKM in adult individuals improve their capacity to respond to tissue injury (Chen et al., 2021; Ocampo et al., 2016).”

• cites Chen et al., 2021Not checked yet

  Not checked in the free run.

• cites Ocampo et al., 2016Not checked yet

  Not checked in the free run.

The recovery period is important for the demethylation of aging-DMRs.

“This, together with the RRBS data above (Figure 1e,f), is another indication that the recovery period is important for the demethylation of aging‐DMRs.”

Reprogrammed cells can subsequently differentiate into somatic cells that are rejuvenated relative to their parental cells.

“Moreover, reprogrammed cells can subsequently differentiate into somatic cells that are now rejuvenated relative to their parental cells (Lapasset et al., 2011; Liu et al., 2011).”

• cites Lapasset et al., 2011Not checked yet

  Not checked in the free run.

• cites Liu et al., 2011Not checked yet

  Not checked in the free run.

Reprogramming erases molecular and cellular traits of aging acquired by somatic cells throughout their lifespan.

“This process, generally known as reprogramming, erases molecular and cellular traits of aging acquired by somatic cells throughout their lifespan (Lapasset et al., 2011; Liu et al., 2011; Mahmoudi & Brunet, 2012; Petkovich et al., 2017; Rando & Chang, 2012).”

• cites Lapasset et al., 2011Not checked yet

  Not checked in the free run.

• cites Liu et al., 2011Not checked yet

  Not checked in the free run.

• cites Mahmoudi & Brunet, 2012Not checked yet

  Not checked in the free run.

Loss of collagens occurs with age in pancreas.

“Finally, loss of collagens occurs with age in pancreas (Riopel & Wang, 2014) and, remarkably, old‐OSKM mice increased their levels of collagen‐related genes (Figure 2h, Figure S2j).”

• cites Riopel & Wang, 2014Not checked yet

  Not checked in the free run.

Lower serum transaminase levels after OSKM activation and recovery reflect improved liver function.

“Interestingly, the serum levels of transaminases of OSKM mice were significantly lower after 1 week of OSKM activation and 2 weeks of recovery reflecting an improved liver function (Figure 3e).”

• AssumptionRests on an assumption: Serum transaminase levels are a valid marker of liver function, such that lower levels reflect improved liver function.

A single cycle of OSKM expression has a detectable rejuvenating effect on tissues such as liver, spleen, and blood that do not express high levels of OSKM and do not manifest obvious signs of histological alterations.

“We conclude that a single cycle of OSKM expression has a detectable rejuvenating effect on tissues such as liver, spleen, and blood that do not express high levels of OSKM and do not manifest obvious signs of histological alterations.”

• AssumptionRests on an assumption: The observed molecular rejuvenation features are not explained by changes in cell composition.
• AssumptionRests on an assumption: OSKM transgene expression is mild in liver, spleen, and blood.

A single period of OSKM expression is able to rejuvenate a fraction of the methylation changes that occur with aging in promoters and enhancers.

“We conclude that a single period of OSKM expression is able to rejuvenate a fraction of the methylation changes that occur with aging in promoters and enhancers.”

Transient OSKM activation appears to orchestrate a positive reconfiguration of the transcriptome that opposes key hallmarks of aging.

“Overall, transient OSKM activation appears to orchestrate a positive reconfiguration of the transcriptome that opposes key hallmarks of aging.”

A single period of transient OSKM in vivo may not be sufficient to rejuvenate the transcriptome of senescent cells present in aged liver.

“These observations suggest that a single period of transient OSKM in vivo may not be sufficient to rejuvenate the transcriptome of the senescent cells present in the aged liver.”

• AssumptionRests on an assumption: γH2AX-positive cells are a marker of in vivo senescence.

Cellular senescence is a potent barrier for reprogramming.

“Cellular senescence is a potent barrier for reprogramming (Haridhasapavalan et al., 2020)”

• cites Haridhasapavalan et al., 2020Not checked yet

  Not checked in the free run.

The Hsf4 intragenic region is differentially methylated during aging in various tissues.

“This particular region has been identified by independent studies as differentially methylated during aging in various tissues (Beerman et al., 2013; Han et al., 2018; Taiwo et al., 2013)”

• cites Beerman et al., 2013Not checked yet

  Not checked in the free run.

• cites Han et al., 2018Not checked yet

  Not checked in the free run.

• cites Taiwo et al., 2013Not checked yet

  Not checked in the free run.

Serum 4-hydroxyproline is considered to reflect total collagen levels.

“its presence in the serum is considered to reflect total collagen levels (Gabr et al., 2017).”

• cites Gabr et al., 2017Not checked yet

  Not checked in the free run.

The Hsf4 region has a large magnitude of change in percentage of methylation from young to old individuals.

“it is extraordinary due to the large magnitude of change in percentage of methylation from young to old individuals (Han et al., 2018).”

• Also stated in: § DISCUSSION · boosted

  “A remarkable property of this methylated region is the large magnitude of change in percentage of methylation from young to old individuals (Han et al., 2018).”

• cites Han et al., 2018Not checked yet

  Not checked in the free run.

Transcriptional rejuvenation may be easier to achieve compared to epigenetic rejuvenation.

“It seems, therefore, that transcriptional rejuvenation may be easier to achieve compared to epigenetic rejuvenation.”

• AssumptionRests on an assumption: OSKM transgene expression is mild in liver, spleen, and blood.

Senescent cells are generally resistant to being reprogrammed.

“Senescent cells are generally resistant to be reprogrammed (Haridhasapavalan et al., 2020).”

• cites Haridhasapavalan et al., 2020Not checked yet

  Not checked in the free run.

A single period of transient OSKM may not be sufficient to rejuvenate the transcriptome of senescent cells present in aged tissues.

“Therefore, a single period of transient OSKM may not be sufficient to rejuvenate the transcriptome of the senescent cells present in aged tissues.”

The disappearance of about half of methylation changes after 2 weeks illustrates the existence of unstable changes induced by OSKM.

“This illustrates the existence of unstable changes induced by OSKM and the importance of allowing a recovery period to identify durable changes.”