Chemical reprogramming ameliorates cellular hallmarks of aging and extends lifespan

Abstract

The dedifferentiation of somatic cells into a pluripotent state by cellular reprogramming coincides with a reversal of age-associated molecular hallmarks. Although transcription factor induced cellular reprogramming has been shown to ameliorate these aging phenotypes in human cells and extend health and lifespan in mice, translational applications of this approach are still limited. More recently, chemical reprogramming via small molecule cocktails have demonstrated a similar ability to induce pluripotency in vitro, however, its potential impact on aging is unknown. Here, we demonstrated that chemical-induced partial reprogramming can improve key drivers of aging including genomic instability and epigenetic alterations in aged human cells. Moreover, we identified an optimized combination of two reprogramming molecules sufficient to induce the amelioration of additional aging phenotypes including cellular senescence and oxidative stress. Importantly, in vivo application of this two-chemical combination significantly extended C. elegans lifespan and healthspan. Together, these data demonstrate that improvement of key drivers of aging and lifespan extension is possible via chemical-induced partial reprogramming, opening a path towards future translational applications.

Keywords: Aging; Cellular Reprogramming; Chemical Reprogramming; Epigenetics; Lifespan.

Figure 1. Chemical-induced partial reprogramming significantly improves aged hallmarks in aged human fibroblasts.

(A) Schematic representation of chemical-induced partial reprogramming via 6 days of treatment with the seven chemicals previously shown to induce mouse chemical iPSC (Hou et al, 2013). (B, C) Immunofluorescence and quantification of γH2AX following 7c treatment (B) or Doxorubicin (100 nM) and 7c treatment (C). (D, E) Immunofluorescence and quantification of H3K9me3 (D) and H3K27me3 (E) following 7c treatment. (F) mRNA levels of senescence-associated and age-related stress response genes in the p53 tumor suppressor pathway following 7c treatment (6 days). (G) Principal component analysis (PCA) of control (blue) and 7c-treated (orange) fibroblasts. (H) Gene ontology (GO) enrichment analysis following 7c treatment with developmental (in pink) and cell cycle (in blue) pathways highlighted. (I) Heatmap showing the expression pattern of differentially expressed genes associated with developmental pathways following 7c treatment. Data were median ± IQR (B–E), mean ± SEM (F). (B, C) n = 2, (D–I) n ≥ 3. Statistical significance was assessed by comparison to untreated control using a paired two-tailed t-test (B–F). NES normalized enrichment scores. Source data are available online for this figure.

Figure 2. Chemical-induced partial reprogramming via 7 chemicals does not fully induce multiparameter rejuvenation.

(A) MTS quantification of cell density following 7c treatment until confluence. (B) Immunofluorescence and quantification of Ki67 following 7c treatment (6 days, “6D”). (C) Immunofluorescence and quantification of γH2AX following 7c treatment (6D) in proliferative (10% FBS) and non-proliferative (1% FBS) conditions. (D) Fluorescence detection and quantification of reactive oxygen species (ROS) following 7c treatment (6D). Data were mean ± SEM (A, D), median ± IQR (C). (A, B) n = 2, (C) n = 1, (D) n ≥ 3. Statistical significance was assessed by comparison to untreated control using a paired two-tailed t-test (A, D). OD optical density. Source data are available online for this figure.

Figure 3. Optimized cocktail (2c) efficiently improves multiple molecular hallmarks of aging.

(A) Immunofluorescence and quantification of γH2AX following TCP + Repsox (2c, 5 µM each) treatment (6 days, “6D”). (B, C) Immunofluorescence and quantification of H3K9me3 (B) and H3K27me3 (C) following 2c treatment (6D). (D) Senescence-associated beta-galactosidase (SA-beta-gal) staining and quantification following Doxorubicin (100 nM) in 2c pretreated fibroblasts (6D). (E) mRNA levels of senescence-associated p21 expression following doxorubicin (100 nM) in 2c pretreated fibroblasts (29 days, “29D”). (F) SA-beta-gal staining and quantification of replicative-induced senescence (RIS) following long-term (29D) 2c treatment. (G) mRNA levels of senescence-associated and age-related stress response genes in the p53 tumor suppressor pathway following 2c treatment. (H) MTS quantification of cell density following 2c treatment until confluence. (I) Fluorescence detection and quantification of ROS following 2c treatment (6D). Data were median ± IQR (A–C), mean ± SEM (D–I). (A–C, E–G, I) n ≥ 3, (D, H) n = 2. Statistical significance was assessed by comparison to untreated control using paired two-tailed t-test (A–C, F, H, I), one-way ANOVA and Dunnett correction (D, E, G). OD optical density. Source data are available online for this figure.

Figure 4. Characterization of 2c transcriptomic profile.

(A) Principal component analysis (PCA) of control (blue) and 2c-treated (green) fibroblasts. (B) Gene ontology (GO) enrichment analysis following 2c treatment with developmental and cell migration (in pink) and RNA splicing and post-translational (in blue) pathways highlighted. (C) Heatmap showing the expression pattern of differentially expressed genes associated with developmental pathways following 2c treatment. (D) Venn diagram showing the overlapping differentially expressed genes between 2c and 7c treatments. (E) Gene expression changes of key fibroblast markers following 2c and 7c treatment. Data are median ± IQR with whiskers extending to the furthest values within 1.5·IQR (E). (A–E) n ≥ 3. Statistical significance was assessed by comparison to the untreated control using exact hypergeometric probability (total reference gene count of 18000) (D). NES normalized enrichment scores, CPM counts per million.

Figure 5. Treatment with 2c increases C. elegans lifespan and healthspan.

(A) Survival of N2 C. elegans upon treatment with TCP (50 µM), Repsox (50 µM), and 2c (TCP + Repsox, 50 µM each). (B–D) Survival of N2 C. elegans upon treatment with 2c (B), Repsox (C), and TCP (D) at 50, 100, or 200 µM. (E) Progeny generation and reproductive span (indicated with #) of the unmated hermaphrodite C. elegans upon treatment with 2c. (F) Quantification of the germline tumoral mass development at young (day 5), middle (day 13), and old age (day 22) upon 2c treatment. (G, H) Mean movement speed (G), swimming speed (H), and swimming activity (I) of N2 C. elegans upon treatment with 2c. (J) Resistance to oxidative stress was measured at 3, 7, 14, and 19 days of adulthood upon 2c treatment. (K) Survival of C. elegans to paraquat-induced oxidative stress upon 2c treatment at 14 days of adulthood. (L) Thermotolerance of C. elegans to 2 h HS at day 19 of adulthood upon 2c treatment. Data were mean ± SEM (E–I). (E–G, J) n ≥ 10, (H–I) n ≥ 4. The number of animals and measures are detailed in the Methods. Statistical significance was assessed by comparison to vehicle control using log-rank (Mantel-Cox) test for each lifespan (A–D, J–L), one-way ANOVA and Dunnett correction (F), two-way ANOVA mixed-effects analysis and Geisser-greenhouse correction (G), paired two-tailed t-test for day 21 of adulthood (H, I). HS Heat shock. Source data are available online for this figure.

Figure EV1. Optimized cocktail (2c) multiparameter rejuvenation of aging hallmarks are recapitulated in human keratinocytes.

(A, B) MTS quantification of cell density following treatment of human epidermal keratinocyte with TCP (A) and Repsox (B). Red arrows indicate selected concentrations. (C) Immunofluorescence and quantification of γH2AX following 2c treatment in keratinocytes (6 days, “6D”). (D, E) Immunofluorescence and quantification of H3K9me3 (D) and H3K27me3 (E) following 2c treatment (6D) in keratinocytes. (F) SA-beta-gal staining and quantification of senescence following 2c treatment (6D) in keratinocytes. (G, H) Immunofluorescence and quantification of γH2AX (G) senescence-associated beta-galactosidase (SA-beta-gal) (H) following doxorubicin (100 nM) treatment in 2c pretreated keratinocytes (6D). Data were mean ± SEM (A, B, F, H), median ± IQR (C, E, G). (A–H) n ≥ 3. Statistical significance was assessed by comparison to untreated control using paired two-tailed t-test (C–F), one-way ANOVA and Dunnett correction (G, H).

Figure EV2. Reduced 2c cocktail shows long-lasting effects upon treatment and no dedifferentiation.

(A, B) Immunofluorescence and quantification of H3K9me3 (A) and H3K27me3 (B) upon 2c treatment for 6 days (6 days, “6D”) followed by 30 days chemical withdrawal. (C) Scratch assay analysis upon 2c treatment in adult dermal fibroblasts (DF, 6D) alongside a neonatal control (BJ). (D) Gene ontology (GO) terms overlap between 2c and 7c treatments, with developmental (in pink) and cell cycle (in blue) pathways highlighted. (E) Heatmap showing the expression pattern of differentially expressed genes associated with pluripotency following 2c and 7c treatment, and induction of chemical iPSC (ciPSC). (F) Heatmap of Spearman’s correlation of gene expression associated with aging, 2c and 7c. (G) Fluorescence detection and quantification of fibroblast identity marker COL1A1 following 2c and 7c treatments. Data were median ±  IQR (A, B), mean ± SEM (C, E). (A–G) n ≥ 3. Statistical significance was assessed by comparison to untreated control using paired two-tailed t-test (A, B), two-way ANOVA mixed-effects analysis and Geisser-Greenhouse correction (C), Spearman’s correlation test (F), one-way ANOVA and Dunnett correction (G). NES normalized enrichment scores, FC fold change.

Figure EV3. Treatment with 2c improves multiple healthspan parameters in C. elegans.

(A) Survival of N2 C. elegans upon treatment with Metformin (50 mM). (B) Progeny production and reproductive span (denoted by #) of unmated hermaphrodite C. elegans upon treatment with 2c and Metformin. (C) Germline tumoral mass (denoted by t) development at young (day 5), middle (day 13), and old age (day 22) upon 2c treatment in linearized nematodes. (D) Mean body length of N2 C. elegans upon treatment with 2c at 50 µM. (E) Maximum movement speed of N2 C. elegans upon treatment with 2c at 50 µM. (F) 30-second movement map of N2 C. elegans at day 9 and day 18 upon treatment with 2c. (G) Mean swimming asymmetry, amplitude, and max speed of N2 C. elegans upon treatment with 2c at 50 µM. (H) Thermotolerance of N2 C. elegans to 4 h HS at 7 days of adulthood upon 2c treatment. Data were mean ± SEM (B, D, E, G). (B, D, E) n ≥ 10, (G) n ≥ 4. Number of animals and measures are detailed in Methods. Statistical significance was assessed by comparison to vehicle control using log-rank (Mantel-Cox) test (A), by comparison to vehicle control using two-way ANOVA mixed-effects analysis and Geisser-Greenhouse correction (D, E), paired two-tailed t-test for day 21 of adulthood (G), log-rank (Mantel-Cox) test (H). HS heat shock.