HDAC1/2 inhibitor therapy improves multiple organ systems in aged mice

Summary Aging increases the risk of age-related diseases, imposing substantial healthcare and personal costs. Targeting fundamental aging mechanisms pharmacologically can promote healthy aging and reduce this disease susceptibility. In this work, we employed transcriptome-based drug screening to identify compounds emulating transcriptional signatures of long-lived genetic interventions. We discovered compound 60 (Cmpd60), a selective histone deacetylase 1 and 2 (HDAC1/2) inhibitor, mimicking diverse longevity interventions. In extensive molecular, phenotypic, and bioinformatic assessments using various cell and aged mouse models, we found Cmpd60 treatment to improve age-related phenotypes in multiple organs. Cmpd60 reduces renal epithelial-mesenchymal transition and fibrosis in kidney, diminishes dementia-related gene expression in brain, and enhances cardiac contractility and relaxation for the heart. In sum, our two-week HDAC1/2 inhibitor treatment in aged mice establishes a multi-tissue, healthy aging intervention in mammals, holding promise for therapeutic translation to promote healthy aging in humans.

and chemical screening, our team has been pioneering transcriptome-based drug screening for longevity interventions.For example, our approaches have identified (1) HSP90 inhibitors as proteostasis-inducing longevity interventions, 12 (2) longevity compounds with minimized probabilities of side effects in humans, 14 (3) the acetylcholine receptor as a target to activate the pro-longevity transcription factor FOXO3, 15,16 and (4) the antiretroviral zidovudine to activate the pro-longevity transcription factor ATF4. 17 In addition to our own work, in silico drug screening has been used to identify a novel treatment for metabolic disorder, 18 identify mimetics for the calorie restriction longevity intervention, 19 and in general, de-risk early phase drug screening. 20n the current work, we performed multiple in silico drug screens using transcriptional profiles of 2,837 small molecules testing their ability to mimic known genetic longevity interventions.We identified one compound that was most commonly found to mimic the transcriptional profile of the genetic longevity interventions.This benzamide-based small molecule, termed compound 60 (Cmpd60; aka Merck60 or BRD 692), is a selective histone deacetylase 1 and 2 (HDAC1/2) inhibitor.Cmpd60 was previously shown to repress growth in certain hematologic malignancies in vitro 21 and in vivo to cross the blood-brain barrier to reduce anxiety in mice. 22Here, we used a combination of molecular, phenotypic, and bioinformatic analyses in multiple disease cell models and mouse models for age-related disease to establish if Cmpd60 acts as a geroprotector.Indeed, we found that Cmpd60 treatment attenuates age-associated phenotypes across multiple organ systems, including the kidney, brain, and heart.This is in line with our finding that Cmpd60's transcriptional signature mimics diverse longevity interventions and with other individual accounts of certain (pan-or class-specific) HDAC inhibitors benefiting individual diseases. 23Our work establishes for the first time specifically HDAC1/2 inhibition as a healthy aging intervention in mammals, further demonstrates that a single molecule can have pleiotropic beneficial effects for healthy aging on multiple organ systems, and paves the way for the development of more potent geroprotective therapeutics in mammals capable of recapitulating the benefits of diverse known genetic longevity interventions.

In silico transcriptome screening for pharmaceuticals mimicking genetic longevity interventions
In order to identify small molecules that could recapitulate the benefits of multiple genetic longevity interventions, we consulted the GeneAge database, where we found 75 genetic interventions (i.e., either knockdown/outs or overexpressions), which have been documented to extend lifespan. 24We next turned to the library of integrated-network-based cellular signatures (LINCS), an online database and software suite containing mRNA signatures of both drug-treated and genetically perturbed human cell lines. 25,26Cross-referencing our list of 75 genetic longevity interventions with genetic perturbation cell lines, we found transcriptional signatures were available in the LINCS database for 25 of these (Figure 1A).These 25 interventions, along with FOXO3 overexpression as recently described, 15 were used to query the LINCS database consisting of high-certain transcriptomes of 2,837 small molecules present in 8 core cell lines (PC3, VCAP, A375, HA1E, HCC515, HT29, MCF7, and HEPG2), and identify those whose transcriptomic signatures were most similar to at least one of the genetic longevity intervention's transcriptomes (LINCS score >90).To ensure the highest likelihood that our drug list would indeed benefit the aging process, we imposed a filter on the query, requiring that a drug's known target must also be included as a genetic perturbation hit.This resulted in 498 compounds mimicking at least one genetic longevity intervention (Figure 1A).Finally, drugs were ranked according to the number of genetic longevity interventions they transcriptionally mimicked, to form a prioritization ranking (Figure 1B).
When exploring the ranked list of drugs mimicking the most longevity interventions, we noted many well-studied drugs in the context of aging (Figure S1).For example, the 3 rd ranking drug, mimicking 10 out of the 25 genetic interventions, was sirolimus, well known to extend lifespan in diverse model organisms. 27Furthermore, ranked 4 th and 5 th included other molecules that extend lifespan in Caenorhabditis elegans, including digoxin, 28 taxifolin, 29 genistein, 30 and catechin. 31Indeed, many top ranked small molecules from our screen either extend lifespan in model organisms or have other direct links to age-related pathways (Figure S1).However, the top ranked compound, which is the only one to bear transcriptional similarity to 12 out of the 25 genetic longevity interventions, was termed ''compound 60'' (Cmpd60, or ''Merck60'') and had not yet been explored in the context of healthy aging (Figure 1B).
Cmpd60 is a benzamide-based small molecule that selectively inhibits histone deacetylase 1 and 2 (HDAC1/2).Interestingly, Cmpd60 mimicked the effects of the metabolic-related genetic longevity interventions including knockdown of AKT (AKT KD ) and knockdown of multiple components of the insulin signaling pathway including INSR KD and IRS1 KD (Figure 1B), in line with reports that pan-HDAC inhibition can prevent insulin resistance and obesity in mice fed a high-fat diet. 32Furthermore, because HDAC inhibitors as a drug class may harbor some of the most promising geroprotective compounds, 23 we believed Cmpd60 was an intriguing molecule to further explore in the context of geroprotection.

Cmpd60 in aged mice restores youthful molecular and physiological renal parameters
In order to investigate Cmpd60's potential protective effects during aging, we first turned to an in vitro model of renal fibrosis, a hallmark of age-related kidney disease.During aging, senescent tubular epithelial cells (TECs) accumulate in the kidney, 33 which produce a wide range of profibrotic mediators, such as transforming growth factor b (TGF-b). 34This profibrotic cytokine in turn affects TECs' phenotype, promoting a partial epithelial-mesenchymal transition (EMT), ultimately leading to renal fibrosis. 35Partial EMT in TECs is marked by an increased expression of the mesenchymal gene alpha-smooth muscle actin (aSMA) and a decrease in zonula occludens-1 (ZO-1) and E-cadherin. 36,37Indeed, treating TECs with recombinant TGF-b was sufficient to significantly increase aSMA and reduce ZO-1 protein expression (Figure 2A).We then tested if Cmpd60 could prevent EMT in TECs.We used a dose of 1 mM Cmpd60, which is a non-toxic dose (Figure S2A) that effectively increases histone acetylation levels at histone H3K18 and H4K8 (Figures S2B and S2C).Strikingly, Cmpd60 partially prevented EMT upon TGF-b stimulation, reducing aSMA and increasing ZO-1 and E-cadherin protein expression (Figures 2A and S2D).
To determine Cmpd60's geroprotective effects on the kidney at the molecular and physiological levels, we proceeded to treat aged male mice (20 months old) via intraperitoneal injection for 14 days with either Cmpd60 (22.5 mg/kg) or control (Figure 2B).This dosing regimen was based in part on previous studies with Cmpd60. 22EchoMRI measurements showed that fat mass, lean mass, and total body weight did not change between treated and untreated mice, suggesting that Cmpd60 was tolerated at the dose used (Figure S2E), which matched the observation that blood biochemistry markers for renal and liver toxicity did not differ between the two groups (Figure S2F).Assessing acetylation levels revealed an increase of H4K8 acetylation in Cmpd60-treated kidneys, demonstrating efficacy of the intervention (Figures 2C and S2G).
To assess the molecular effects of Cmpd60 on the kidney, we performed RNAseq transcriptomics on kidneys from the treated and untreated aged mice (Table S1).Samples could be readily differentiated using partial least squares discriminant analysis (PLS-DA) (Figure 2D).Exploring the data further, we calculated differential expression between the groups, where we noted that inhibition of HDAC1/2 with Cmpd60 imparted clear differences on the transcriptional landscape (p < 0.01, Table S1).To better understand what these changes were, we performed gene ontology (GO) term and KEGG pathway analyses on the up-and downregulated genes (Figure 2E and Table S2).Here, we found the top upregulated GO term was one often associated to longevity, healthy aging, and oxidative stress protection, namely, glutathione metabolic processes 38 (Figure 2E).These genes included glutathione S-transferase genes (Gstm1, Gsta3, and Gsta4) (Figure S2H), an important family of detoxifying and cytoprotective enzymes crucial for longevity 39 and a protective mechanism against the development of renal fibrosis healthy aging and oxidative stress protection.Given the decrease in partial EMT observed in the in vitro model, we sought to assess how these molecular changes manifest themselves at the physiological level.We performed histological analysis of renal fibrosis by analyzing collagen content detected with picrosirius red.Markedly, we found that the aged Cmpd60-treated mice showed less age-related renal fibrosis than their untreated counterparts (Figures 2F and 2G).Taken together, these findings suggest Cmpd60 alters the transcriptional landscape in aged kidney cells, shifting it toward a profile protective from oxidative stress and conducive to a reduction of renal EMT and agerelated kidney fibrosis.

Cmpd60 treatment protects against detrimental brain aging processes
Having seen clear benefits of Cmpd60 treatment to the aged renal system and noting prior work of others that demonstrated Cmpd60's ability to cross the blood-brain barrier, 22 we inquired the effects of Cmpd60 on the aged brain.Assessing histone modification in the brain revealed increased acetylation levels (Figures 3A and S3A).Establishing this, we proceeded to perform RNAseq transcriptomics on brains of treated and untreated aged mice (Table S3).Here, PLS-DA readily separated the two groups (Figure 3B), and we applied the same cutoff as for the kidney to assess differential expression (p < 0.01, Figure S3B).Interestingly, assessing enriched GO terms and KEGG pathways revealed an alteration in oxidative phosphorylation processes, downregulated upon treatment (Figure 3C; Table S4).Remarkably, the KEGG pathway of Alzheimer was also downregulated upon Cmpd60 treatment (Figure 3C).This included genes also involved in oxidative phosphorylation such as the NADH:ubiquinone oxidoreductase subunits (NDUFs) (Figure 3D), in line with the finding that decreasing mitochondrial capacity can reduce amyloid-b toxicity. 40bserving this potential beneficial effect, we next asked if Cmpd60 treatment could help prevent neurological decline in a dementia model.To address this, we turned to the APPSWE-1349 mouse model, a transgenic mouse overexpressing an isoform of human Alzheimer beta-amyloid (bA), which shows clear signs of impaired spatial referencing at 9-10 months of age. 41We proceeded to treat APPSWE-1349 mice and control littermates for 14 days with either Cmpd60 (22.5 mg/kg) or control (Figure 3E).We used mice younger than those that show full physiological symptoms, aged 6-7 months, to ensure the greatest chance of intervening in the early, molecular-based processes that occur and contribute to bA accumulation and neurodegeneration.Likewise, we focused on molecular readouts to assess efficacy.Performing RNAseq transcriptomics on brain of these mice (Table S5) and PLS-DA revealed a strong separation of the non-treated transgenic mice but less separation of the Cmpd60-treated transgenic mice from the control littermate mice (Figure 3F).This suggested Cmpd60 treatment was shifting the transgenic mouse profile away from a disease profile toward a non-disease profile.3][44] Taking into account all four groups, namely the transgenic and control mice, both untreated and treated, allowed for an analysis of gene expression changes that Cmpd60 induced, unique to the transgenic disease model.Here, relevant for Cmpd60's potential effects in dementia specifically, we found an upregulation of memory-related GO terms (Figure 3G; Table S6).Some of the differentially expressed genes in this category included Pla22g6, 45 Cx3cr1, 46 Ncam1, 47,48 and Cyfip1 49 (Figure 3H), genes whose expression have been shown to benefit cognitive processes.
Finally, to determine how these transcriptional changes may manifest at the physiological level, we performed histological analysis of brains from the transgenic mice, either Cmpd60 treated or untreated.Although the mice we studied were younger than the age at which aggregates are clearly visible, we found suggestive evidence that pre-aggregates were less present in Cmpd60 treated mice.Specifically, 4 out of 7 untreated mice showed aggregates (57%), whereas only 2 out of 6 Cmpd60-treated mice showed aggregates (33%) (Figure S3F).Taken together, our findings suggest Cmpd60 modifies the brain transcriptional landscape in a manner protective against the brain aging changes and counter to dementia-related processes.

Cmpd60 treatment improves cardiac function
Having noted Cmpd60's beneficial effects on the aged kidney and brain, with relevance for two serious and undertreated age-related dysfunctions of renal failure and dementia, we next inquired as to the effects of Cmpd60 on one of the organs most contributing to age-related death: the heart.Our initial analysis did not reveal significant acetylation changes in histone H3 or H4 (Figures S4A-S4C).Nonetheless, to further explore Cmpd60's cardiac-related effects more deeply, we performed RNA-seq transcriptomics on hearts from aged treated and untreated mice (Table S7).Here, we again observed samples to be readily distinguishable upon PLS-DA (Figure 4A).
Upon evaluating differential expression in the Cmpd60-treated versus untreated heart samples, we noted far greater transcriptional changes following Cmpd60 treatment in the heart compared with either the kidney or the brain.Accordingly, we applied a stricter cutoff to assess differential expression (adjusted p value <0.05) (Figure 4B).Although we found fewer GO enrichments and KEGG pathways related to altered oxidative phosphorylation processes, strikingly, we found the top enriched GO terms were related to heart valve development, suggesting profound changes influencing heart function may be occurring upon Cmpd60 treatment (Figure 4C; Table S8).This upregulation included genes such as SMAD family member 6 (Smad6), ADAM metallopeptidase with thrombospondin type 1 motif 9 (Adamts9), and elastin microfibril interface 1 (Emilin1), members of gene families who have all been linked to cardiovascular outcomes, with either deficiency proving detrimental or abundance proving beneficial [50][51][52] (Figure 4D).Spurred by these promising findings, we turned to an in vitro assay of cardiac functioning.Here, we assessed contraction (percentage of sarcomere shortening) and relaxation (return velocity) in adult rat ventricular cardiomyocytes.Remarkably, and in line with our in vivo findings at the transcriptional level, we found Cmpd60-treated ventricular cardiomyocytes showed both an improved contraction and relaxation parameters (Figures 4E and 4F).Taken together, this suggests Cmpd60 treatment modifies the cardiac transcriptional level and manifests itself at the functional level to improve age-related cardiac outcomes.

A consensus model of Cmpd60's effects
Having identified tissue-specific benefits of Cmpd60, we next inquired whether a conserved expression profile existed among the different tissues of the treated mice.To accomplish this, we assessed the overlap of differentially expressed genes in the kidney (p value<0.05),brain (p value<0.05),or heart (adjusted p value<0.05).We identified 41 genes upregulated (Figure 5A) and 30 genes downregulated (Figure 5B) in common between the three tissues following Cmpd60 treatment.Among these 71 genes, for example, were genes including upregulated Mapk3, Tgm2, and Spns2 and downregulated Mrps28 and Fzd8 (Figure 5C).Transcription factor analysis querying diverse motif databases revealed six motifs (transfac-pro-M00797, cisbp-M6275, swissregulon-hs-HIF1A.p2,transfac-pro-M00466, transfac-pro-M07043, homer-TACGTGCV-HIF-1a) associated with Hif1a target genes (Figure 5D), suggesting Cmpd60 treatment increases oxidative stress resistance, an observation in line with the main transcriptional changes observed in the kidney and brain.Taken together, our work suggests both tissue-specific effects of Cmp60 treatment, such as Gsta2/3/4 and Gstp1/3 in the kidney, Wnt5a in the brain, and Scx and Emilin1 in the heart, as well as common transcriptional changes shared between tissues, oriented around Hif1a target gene expression.Together, the cumulated effects of these molecular changes may result in the age-reversing qualities we observed following Cmpd60 treatment in old mice.

DISCUSSION
In this work, we used an in silico drug screening platform and identified a single molecule, the HDAC1 and HDAC2 inhibitor Cmpd60, which possessed transcriptional signatures mimicking diverse genetic longevity interventions.In line with this, Cmpd60 demonstrated distinct effects across multiple organ systems where it was able to attenuate age-related phenotypes.In the kidney, Cmpd60 treatment increased protective gene expression related to oxidative stress regulation and reduced fibrosis possibly via reduced partial EMT detected in in vitro studies.This is in line with several studies supporting the link between decreased oxidative stress and amelioration of renal fibrosis. 53,54In the brain, Cmpd60 treatment showed transcriptional changes conducive to improved cognitive functioning and molecular indications of neuroprotection in both naturally aged brain and a dementia mouse model brain.In the heart, Cmpd60 resulted in cardiac-remodeling-related transcriptional changes and benefitted cardiomyocyte functioning.
With Cmpd60 demonstrating such diverse age-related benefits across multiple organs, a question remains as to how these effects are mediated.HDAC inhibition has previously been suggested to benefit health through a plethora of mechanisms, including FOXO3 activation, 55 Klotho upregulation, 56 or reversing age-related acetylation changes, among others. 23Notably, these have all been explored in diverse models and organs.The likeliest answer therefore is that HDAC inhibition modifies a tissue-specific epigenetic landscape, creating beneficial tissue-specific responses (Figure 5E).Because aging is accompanied by alterations in histone acetylation patterns and global loss of transcriptional control, 57 one tantalizing possibility is that Cmpd60 reverses these acetylation changes and attenuates the aging phenotype in a tissue-specific manner.Our findings at the transcriptional level, including an upregulation of oxidative stress protection and alterations in metabolic gene expression, catered to each organ, support this idea.It remains to be seen how each organ achieved such benefits and how these findings can further translate to benefit human health.
One remarkable occurrence we noted is the beneficial effects of Cmpd60 treatment in cardiac tissue, despite the fact that no changes in histone acetylation levels were observed in this system.This suggests that Cmpd60's HDAC-targeting effects may not be responsible for the changes observed in the heart and rather that the effects of Cmpd60 in the heart may be either (1) indirect, e.g., systemic effects from another body system that crosstalk with the heart, or (2) affecting acetylation levels of proteins other than HDACs.This seems to be different compared with the changes we have observed in the kidney and brain where histone acetylation changes were clearly observed.Indeed, HDAC inhibitors in general have been hypothesized to benefit the aging process by targeting non-histone-related proteins (as well as histones), 23 and Cmpd60 may also act through multiple mechanisms, again in a tissue-specific manner.The heart may be an example of this in our study.
Further exploring the link between Cmpd60 and Hif1a would be of great interest.This is especially the case considering past studies that have demonstrated HDACs to activate Hif1a, which therefore implies that a general suppression of Hif1a results upon treatment with HDAC inhibitors. 58Different HDACs and HDAC inhibitors may have different regulatory effects on Hif1a, depending on dose and cell line used. 59ere, it should be noted that although our RNA-seq and bioinformatic analyses have revealed a clear link between Hif1a and Cmpd60, further in vitro studies cotreating cells with Cmpd60 and an Hif1a inhibitor would be required to formalize this relationship.In our kidney cell model, we observed a rescue of markers associated with pEMT following treatment with both TGF-b and Cmpd60.This restoration is likely facilitated by the modulation of Hif-1a.Indeed, previous research has illustrated a time-dependent increase in Hif-1a levels in proximal tubular epithelial cells exposed to TGF-b.Moreover, inhibiting Hif-1a effectively inhibits TGF-b-induced EMT and attenuates kidney fibrosis, which aligns with our findings. 53,60Therefore, the connection between Hif1a and Cmp60 should be seen as a candidate mechanism, requiring formal validation, and it is likely that Cmpd60 may work through other means as well.Indeed, although our Hif1a analyses showed suppression of Hif1a-related genes, it also demonstrated activation of other Hif1a-regulated genes.

Limitations of the study
Several limitations should be considered with our study.For example, our study design involved treating aged animals and assessing a final time point after the treatment period for molecular and physiological changes.With this design, we did not assess aged mice before treatment, and we cannot discern whether or not Cmpd60 acted to (1) rejuvenate the aged animals or (2) attenuate age-related changes that developed during the treatment period.Because the treatment period was relatively short, it can be expected that most changes observed after the treatment were the result of a reversal of aging phenotypes.However, a follow-up study where histology and RNA-seq of aged animals prior to treatment are collected would be required to address this fully, as well as young control animals for comparisons.Furthermore, another limitation of our work may be that our initial drug screen using datasets from the Broad Institute included many cancerous cell lines. 26Although this approach has been used before-by ourselves and others-identifying compounds benefiting health through diverse mechanisms not related to cancer, 15,18,61 it could theoretically produce a confounding factor.It would be interesting to see what other compounds may emerge from similar screens when cancerous cell lines are excluded.Nonetheless, our current screen performed in this study has functioned to identify Cmpd60 as a candidate compound capable of addressing multiple aging phenotypes, meriting further investigation in-of-itself.

Conclusion
As most studies on HDAC inhibitors focus on one specific tissue, our study is unique in that it looks at the effects of HDAC inhibition in three different organs: kidney, brain, and heart.This enabled us to recognize an overlapping gene expression profile in all three tissues, associated with Hif1a target genes.Although we identified tissue-specific benefits of Cmpd60, it should be noted that HDAC inhibitors are also known for their undesirable side effects. 62Despite, or thanks to, their many diverse on-and off-target effects, HDAC inhibitors nonetheless benefit a range of preclinical age-related disease models. 23We therefore recommend future research to assess dose-dependent effects of HDAC1/HDAC2 inhibitors in multiple organs.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following: DNase (QIAGEN).RNA was quantified with a NanoDrop 2000 spectrophotometer (Thermo Scientific; Breda, The Netherlands) and stored at-80 C until use.
RNA libraries were prepared and sequenced with the Illumina platform by Genome Scan (Leiden, The Netherlands).The NEBNext Ultra II Directional RNA Library Prep Kit for Illumina was used to process the sample(s).The sample preparation was performed according to the protocol "NEBNext Ultra II Directional RNA Library Prep Kit for Illumina" (NEB #E7760S/L).Briefly, mRNA was isolated from total RNA using the oligo-dT magnetic beads.After fragmentation of the mRNA, cDNA synthesis was performed.This was used for ligation with the sequencing adapters and PCR amplification of the resulting product.The quality and yield after sample preparation was measured with the Fragment Analyzer.The size of the resulting products was consistent with the expected size distribution (a broad peak between 300-500 bp).Clustering and DNA sequencing using the NovaSeq6000 was performed according to manufacturer's protocols.A concentration of 1.1 nM of DNA was used.NovaSeq control software NCS v1.6 was used.

Transcriptome analysis and visualization
Data processing was performed using R version 4.1.1.Genes were reannotated using the Ensembl genome database and the biomaRt package. 68Resulting p-values were corrected for multiple testing using the Benjamini-Hochberg false discovery rate where applicable.Biological process (BP) overrepresentation analysis was performed using Clusterprofiler 4.0.5 69 and org.Mm.eg.db (version 3.13.0).Gene selection (for Figures 2E, 3C, 3G, 4C, S3D, and S3E) was done based on p-value less than 0.01, and log Fold change larger than 0 for up-regulated genes, smaller than 0 for down-regulated genes.Partial least squares discriminant analysis (PLS-DA) was performed on normalized cpm value (genes with zero expression were filtered out) using MixOmics version 6.16.3. 70Upset plots were generated using UpsetR version 1.4.0. 71For transcription factor (TF) binding motif over-represention, analysis was performed using RcisTarget 1.14.0. 72Shared up-regulated genes (pvalue < 0.05, log Fold change larger than 0 ) between brain, kidney and heart were used as input gene list.The same was performed for shared down-regulated genes.The following file (mm9-500bp-upstream-7species.mc9nr.feather)was used to specify the gene-motif rankings.''motifAnnotations_mgi_v9 '' was used for motif annotation to transcription factors.[75]

Histology and immunostaining
Paraffin-embedded kidney and brain tissues were processed for (immuno)histological analysis.To quantify the percentage of interstitial fibrosis, Picro Sirius red histological staining was performed to detect collagen content.Kidney tissue slides were incubated with 0.2% Picro Sirius Red (PSR) solution (pH 2.0) for 1h followed by incubation with 0.01M HCl.The amount of PSR-positive staining per high power field (20x magnification) was quantified by Image J software.Beta amyloid plaques in brain slides were identified with beta Amyloid (1-42) antibody (Genetex: GTX134510).Quantification of the percentage of amyloid plaques was performed by the neuropathologist in a blinded manner.

Tissue lysates
Freeze dried tissues (kidney, brain and heart) were homogenized in lysisbuffer (120mM Tris pH 6.8, 4% SDS, 20% glycerol supplemented with protease inhibitors) and stored overnight at -20C.The next day the homogenates were passed through a 21G needle and protein was measured using a BCA kit (Thermo Scientific).
Twenty mg of protein was loaded onto a 4-12% Bis-Tris gradient gel (Invitrogen) and separated proteins were transferred on PVDF membrane (Millipore).After blocking aspecific signal, membranes were incubated overnight at 4 C with primary antibodies listed in KRT.The following day membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies for 1 hour at RT. Detection was done by ECL western blotting substrate (Thermo Scientific) and images were obtained on a LAS 4000 (ImageQuant).Band intensity was quantified through ImageJ.

Adult rat ventricular cardiomyocyte isolation and contractility measurement
The animal experiments were performed in accordance with the guidelines from the Directive 2010/63/EU of the European Parliament on the protection of animals used for scientific purposes and approved by the ethics committees of Amsterdam University Medical Centers, VUMC location, Amsterdam, the Netherlands.Adult rat left ventricular cardiomyocytes (CMs) were isolated as described previously. 76,77Briefly, adult wild-type Wistar rats were terminated under anesthesia, followed by chest opening and heart extraction.The heart was cannulated through the aorta and perfused on a Langendorf perfusion set-up with liberase enzyme solution until the tissue was sufficiently digested.The atria and right ventricle were removed and the left ventricle was minced into small pieces and triturated.Subsequently, the cell suspension was filtered and re-suspended in CaCl 2 buffers of increasing Ca 2+ concentrations to reach a final concentration of 1mM.The isolated adult CMs were finally re-suspended in plating medium containing Medium 199 (Lonza, BE12-117F), 1% penicillin/streptomycin (Lonza, DE17-602DE) and 5% fetal bovine serum (PAA, A15-101), and seeded on 1% laminin (L2020-1MG, Sigma)-coated plates (24-well format Costar culture plate, Corning, 3524).One hour after plating, the medium was refreshed with maintenance medium containing Medium 199, 1% penicillin/streptomycin and Insulin-Transferrin-Sodium Selenite Supplement (Sigma-Aldrich; insulin, 10 mg lÀ1; transferrin, 5.5 mg lÀ1; and selenium 5mglÀ1).Subsequently, the cells were stimulated with 5mM Cmpd60 (or corresponding vehicle, DMSO) for 2 hours at 37 C in humidified air with 5% CO 2 .After the stimulation, the contraction and relaxation of the CMs were measured with the MultiCell microscope system (CytoCypher, Amsterdam, the Netherlands) coupled to the Ionoptix high-speed sarcomere length measuring software (Ionoptix LLC, Westwood, Massachusetts).Unloaded intact rat CMs were monitored following field stimulation, and sarcomere shortening was measured and analyzed with the automated, batch analysis software transient analysis tools (Cytosolver, CytoCypher) to determine the contraction and relaxation profiles of the cells.

RNA seq
Statistical analyses were performed using the edgeR v3.26.8 (Robinson et al., 2010) and limma/voom v 3.40.6 78 R packages.All genes with more than 2 counts in at least 3 of the samples were kept.Count data were transformed to log2-counts per million (logCPM), normalized by applying the trimmed mean of M-values method 79 and precision weighted using voom (Law et al., 2014).Differential expression was assessed using an empirical Bayes moderated t test within limma's linear model framework including the precision weights estimated by voom. 78,80Resulting p values were corrected for multiple testing using the Benjamini-Hochberg false discovery rate.Data processing was performed using R v3.6.1 and Bioconductor v3.9.Partial least-squares discriminant analysis (PLS-DA) was performed using mixomics (Rohart  et al., 2017) setting a variable of importance (VIP) score of greater than 1 as significant.Resulting p values (where applicable) were corrected for multiple testing using the Benjamini-Hochberg false discovery rate.Genes were re-annotated using biomaRt using the Ensembl genome databases (v91).

In vivo and in vitro assays
Statistical analyses were performed using PRISM (9.5.1) and specific tests and corrections for multiple hypothesis testing are listed in either each experiment's figure legend or corresponding methods section.

Figure 1 .
Figure 1.Compound screen strategy and results to identify geroprotectors mimicking genetic longevity interventions (A) General outline of screening strategy.GeneAge database was consulted for listing of genetic interventions, which were cross-referenced against the LINCS transcriptome database of cellular perturbations.Compounds best matching a genetic longevity intervention at the transcriptional level were selected for further evaluation, and only those whose drug targets were also present in knockdowns in the screen were included.This resulted in 498 compounds that were ranked based on how many different genetic longevity interventions their transcriptional profiles could recapitulate.(B) The top drugs ranked first (top, Cmpd60 as top-ranked small molecule) to last (bottom) according to how many genetic interventions they mimic (blue indicates a positive hit).

Figure 2 .
Figure 2. Influence of Cmpd60 on aging kidney (A) Representative western blot of tubular epithelial cells (TECs) treated with 20 ng/mL of recombinant TGF-b and with and without Cmpd60 (1 mM) for 72 h.Protein lysates of TECs blotted for anti-ZO-1, anti-E-cadherin, anti-aSMA, and b-actin.Cmpd60 suppresses markers for partial EMT, a hallmark of age-related renal fibrosis (n = 3)/group.(B) Schematic of aged mouse treatment regimen with Cmpd60 and analyses.(C) Relative histone H4 acetylation levels (H4K8Ac) assessed by western blot in renal tissue of aged mice with and without Cmpd60 treatment.Protein expression was normalized against H4 total and expressed as mean G SEM. Mann Whitney test was used to determine statistical differences.**p < 0.01.(n = 5-6)/group.(D) PLS-DA analysis of aged mice treated with and without Cmpd60 (n = 5-6)/group.(E) Top GO terms of upregulated processes in aged mice treated with Cmpd60 (see also Table S2).(F) Representative histological images of picrosirius red staining in kidney of aged mice with and without Cmp60.(n = 5-6)/group.(G) Quantification of interstitial fibrosis determined by the percentage of positive picrosirius red staining/high power field, in mice treated with and without Cmpd60.Percentage of positive staining was assessed with ImageJ software.Data are expressed as mean G SEM and the Mann Whitney test was used to determine statistical significance.*p < 0.05, (n = 5-6)/group.

Figure 3 .
Figure 3. Cmpd60 treatment supports healthy brain aging (A) Relative expression of histone H4 acetylation levels (H4k8Ac) assessed by western blot in brain tissue of aged mice treated with control and Cmpd60.Protein expression was normalized against H4 total and expressed as mean G SEM. Mann Whitney t test was used to determine statistical differences.**p < 0.01.n = 5-6/ group.

Figure 3 .
Figure 3. Continued (B) PLS-DA of RNA-seq transcriptome comparing Cmpd60 treated and untreated brain, n = 6 per group.(C) Downregulated KEGG terms resulting from Cmpd60 treatment.(D) Boxplot of counts per million (CPM) expression values of genes in Cmpd60-treated mouse brain from the GO term enrichment of Alzheimer disease.Fill represents condition; gray for control and blue for Cmpd60.(E) Schematic of dementia mouse model and treatment.(F) PLS-DA of RNA-seq transcriptome comparing dementia mice, controls, treated, and untreated.(G)Top Go terms of interaction between the 4 groups, revealing altered cognitive processes.(H) Boxplot of CPM expression values of genes from the GO term enrichment of cognition in (E).Fill represents condition; gray for control and blue for Cmpd60.

Figure 4 .
Figure 4. Cmpd60 treatment benefits cardiac tissues (A) PLS-DA of RNA-seq, aged heart, treated vs. untreated (n = 5-6/group).(B) Volcano plot of RNA-seq differential expression, aged heart, treated vs. untreated (n = 5-6/group).Genes with p value <0.01 were colored (red: upregulated, blue: downregulated).(C) Top GO terms from upregulated genes.(D) Boxplot of CPM expression values of genes in Cmpd60-treated mouse heart from the GO term enrichment of heart valve development.(E) Treatment of cardiomyocytes with Cmpd60 increased contraction, as shown by increased % sarcomere shortening.(F) Treatment of cardiomyocytes with Cmpd60 improved relaxation, as assessed by higher return velocity (n = 4, corresponding to 4 independent experiments; 30-40 CMs were measured per condition per experiment; data are represented as mean G SD, p < 0.05, unpaired t test).

Figure 5 .
Figure 5.A consensus model of Cmpd60's effects (A) Comparison of the unique and shared upregulated genes in the three tissues: kidney, heart, and brain.Forty-one genes are commonly upregulated in the three tissues (highlighted in blue).(B) Comparison of the unique and shared downregulated genes in the three tissues.Thirty genes are commonly downregulated in the three tissues (highlighted in blue) (C) Heatmap of the log fold change of genes with shared regulation in three tissues (for visualization purposes, log fold changes exceeding 2 were capped at 2, while values below À2 were capped at À2). (D) Network for transcription factor Hif1a, one of the top predicted TFs based on motif overrepresentation of the commonly changed genes among the three tissues.Squares represent different motifs annotated to Hif1a.Edges connect each motif to the genes contributing to its enrichment.(E) Model of Cmpd60's geroprotective effects, which are due to both tissue-specific and conserved transcriptional changes, producing net aging-protective effects.