Abstract
The Caenorhabditis Intervention Testing Program (CITP) is an NIH-funded research consortium of investigators who conduct analyses at three independent sites to identify chemical interventions that reproducibly promote health and lifespan in a robust manner. The founding principle of the CITP is that compounds with positive effects across a genetically diverse panel of Caenorhabditis species and strains are likely engaging conserved biochemical pathways to exert their effects. As such, interventions that are broadly efficacious might be considered prominent compounds for translation for pre-clinical research and human clinical applications. Here, we report results generated using a recently streamlined pipeline approach for the evaluation of the effects of chemical compounds on lifespan and health. We studied five compounds previously shown to extend C. elegans lifespan or thought to promote mammalian health: 17α-estradiol, acarbose, green tea extract, nordihydroguaiaretic acid, and rapamycin. We found that green tea extract and nordihydroguaiaretic acid extend Caenorhabditis lifespan in a species-specific manner. Additionally, these two antioxidants conferred assay-specific effects in some studies—for example, decreasing survival for certain genetic backgrounds in manual survival assays in contrast with extended lifespan as assayed using automated C. elegans Lifespan Machines. We also observed that GTE and NDGA impact on older adult mobility capacity is dependent on genetic background, and that GTE reduces oxidative stress resistance in some Caenorhabditis strains. Overall, our analysis of the five compounds supports the general idea that genetic background and assay type can influence lifespan and health effects of compounds, and underscores that lifespan and health can be uncoupled by chemical interventions.
Introduction
Elaboration of clinical strategies that delay health declines associated with aging is anticipated to markedly improve life quality for the elderly and their families. One research area focused on such a goal seeks to identify compound interventions that can prolong life and/or promote healthy aging. Toward this end, model organisms are invaluable in initial compound screening as these models offer low expense, ease of culture, short lifespans and often simple health assays. One of the most widely studied aging models, Caenorhabditis elegans, has proven a useful system for identifying and characterizing compounds that robustly extend lifespan and healthspan1–3.
The Caenorhabditis Intervention Testing Program (CITP), an NIH-funded research consortium consisting of investigators at three independent sites (Rutgers University, the University of Oregon, and the Buck Institute for Research on Aging), is tasked to identify pharmacological interventions with the potential to extend Caenorhabditis lifespan and healthspan in a robust manner. The founding principle for the CITP is that compounds with positive effects across a genetically diverse population engage conserved biochemical pathways that promote healthy aging. The CITP is distinctive in testing compounds across a genetically diverse panel of Caenorhabditis strains and species to identify interventions that promote lifespan extension independent of genetic background. Another distinctive feature of the CITP effort is that studies are replicated as closely as possible at the three geographically distinct sites. To date, the CITP has reported on the lifespan and healthspan effects of more than 27 compounds in more than 250,000 individuals across nearly 280 trials4–9.
The initial CITP studies focused on longevity as the sole endpoint for anti-aging intervention evaluation4–7. For the current suite of test compounds, we expanded and modified our workflow4 to evaluate the potential of compounds to extend lifespan and/or healthspan across Caenorhabditis species and strains (Supplemental Figure 1).
Here, we report results from five compounds in the CITP testing pipeline: 17α-estradiol, acarbose, green tea extract (GTE), nordihydroguaiaretic acid (NDGA), and rapamycin. 17α-estradiol, a weak endogenous steroidal estrogen, has been reported to alleviate age-related metabolic dysfunction and inflammation in male mice10, to protect against neurodegeneration in cell and animal models of Parkinson’s disease11, and to extend lifespan in genetically heterogenous male mice12,13. Acarbose, an anti-diabetic drug that inhibits alpha-glucosidase, has been shown to prevent age-related glucose intolerance14 and to limit postprandial hyperglycemia in mice15,16. Likewise, acarbose has been shown to extend median lifespan in genetically heterogenous mice12,13. GTE is rich in antioxidant polyphenols, and has been reported to reduce the risk of coronary heart disease and certain forms of cancer17, as well as to provide neuroprotection against diseases such as Alzheimer’s18. GTE increases lifespan in both flies19,20 and mice21, and its primary constituent flavonoid, epigallocatechin gallate (EGCG), has been shown to extend mean lifespan in C. elegans22, although that result may be context dependent23,24. NDGA, a lignin found in the creosote bush, possesses both antioxidant and anti-inflammatory properties25,26, and has been shown to increase lifespan in mice12,27. Finally, rapamycin, an mTOR kinase inhibitor, has been shown to increase lifespan in a variety of model organisms, including mice and flies28–30.
Our studies of 17α-estradiol, acarbose, GTE, NDGA, and rapamycin underscore the complexities of assessing biological outcomes of candidate lifespan extending treatments. Using our standardized protocols, we find that the antioxidants GTE and NDGA extend Caenorhabditis lifespan in a species-specific manner. GTE and NDGA tests also revealed some assay-specific outcomes—in certain genetic backgrounds we found decreased survival in manual longevity assays, whereas we measured extended lifespan when we determined outcomes using the automated C. elegans Lifespan Machines (ALM). GTE and NDGA affected swimming ability in a strain-specific manner, and GTE lowered oxidative stress resistance in some Caenorhabditis strains. Lifespan and healthspan could thus be uncoupled as evaluated by our approach, with outcomes influenced by genetic background. Overall, our findings on this test set of interventions underscore how impactful genetic background, selected health assay, and protocol details are in the assessment of intervention effects. Interventions that meet the high bar of efficacy across a broad range of genetic backgrounds and across multiple experimental approaches may prove the exception but would establish definitive priority for testing in mammalian models.
Results
The CITP pipeline for the evaluation of compounds on lifespan and healthspan
We created a workflow plan to evaluate the ability of a chemical intervention to extend lifespan and/or healthspan across Caenorhabditis species and strains (Supplemental Figure 1). The CITP pipeline involves preliminary testing in a single lab to identify a potentially effective dosage range (phase one study; Supplemental Figure 2) and to target a bioactive dose across strains (phase two; Supplemental Figure 3) before we undertake survival assays in all three CITP labs (phase three; Figure 1). In phase four, we evaluate the effect of bioactive compounds that passed the preliminary tests for robust impact on longevity (phases 1-3) on select health measures (Figures 2 and 3). We publish our findings on compounds without positive lifespan that exit the pipeline6–8 (Supplemental Figure 1).
The antioxidants green tea extract and NDGA extend Caenorhabditis lifespan in a species-specific manner
Using the streamlined pipeline we describe above, we evaluated a test set of compounds for lifespan effects. Among this test set, we found that rapamycin, acarbose, and 17α-estradiol were broadly ineffectual. Rapamycin only altered lifespan for two (C. briggsae HK104 and JU1373) of the nine tested strains, while 17α-estradiol only had an effect on JU1630 (Figure 1). We also found that acarbose did not statistically change the lifespan of any of the tested strains.
The two antioxidants we tested, GTE and NDGA, exerted the largest effect on lifespan in manual survival assays (Figure 1). GTE extended lifespan in two species, C. elegans (two of three strains) and C. tropicalis (all three strains) (Figure 1a). The effect on lifespan was more pronounced in C. elegans, with mean lifespan extended >15% in both MY16 and JU775. In the laboratory C. elegans strain N2, we observed lifespan extension via GTE in only one lab, and the pooled results across labs was insignificant. We also observed lifespan extension with GTE in C. tropicalis, with all three strains tested showing small but significant increases in mean lifespan; the largest effect was in JU1630 (~11% increase in mean lifespan). In contrast, we found that all C. briggsae strains exhibited a significant decrease in survival when exposed to GTE, with mean lifespan decreasing by 10-19%. Overall, we observe a strong species-specific effect on longevity with GTE exposure.
We found that NDGA also impacted Caenorhabditis longevity in a species-specific manner (Figure 1b). In C. elegans, much like with GTE, we measured an 8-16% increase in the mean lifespan of strains MY16 and JU775, while N2 had a similar, albeit insignificant, lifespan increase. The effect of NDGA on lifespan was more variable than we recorded for GTE as we observed a significant decrease in survival in five of six C. briggsae and C. tropicalis strains.
Antioxidants that decrease survival in certain genetic backgrounds extend lifespan in automated C. elegans Lifespan Machine assays
To increase throughput, we have implemented ALMs31 (Supplemental Figure 1, Phase Two) at all three CITP sites. The ALM is a lifespan analysis platform built on flatbed scanner technology that enables life-long imaging of animals at one-hour intervals, increasing both throughput and temporal resolution for data sampling. Our previous testing of compounds NP1, resveratrol, propyl gallate, thioflavin t, and α-ketoglutarate revealed that most ALM trials recapitulated outcomes from manual plate-based assays, although we did identify light sensitivity of some compounds as a factor that could change outcome 5. During preliminary experiments on ALMs we found that both GTE and NDGA increased lifespan in strains that also reported decreased lifespans in manual assays. More specifically, for GTE assays on the ALM, all C. briggsae strains exhibited increased survival compared to the control. For NDGA treatment, four of the six C. briggsae and C. tropicalis strains exhibited positive lifespan results on the ALM. Difference in outcome might be the consequence of increased light exposure on the ALMs32, as light is known to induce photooxidative stress33. ALM protocols also involve fewer potentially damaging animal transfers, and feature no exposure to freshly treated compound plates in mid-to late-life, in contrast to manual assays34,35. Our highly replicated data, however, emphasize that the methodological approach to lifespan determination is a factor in experimental outcome.
GTE and NDGA impact on swimming ability is strain-specific
With the goal of maintaining health over the lifetime, compounds that slow the age-related decline in swimming ability are of particular interest to the CITP. We previously showed that swimming ability as a measure of locomotion with compound treatment in Caenorhabditis does not always correlate with lifespan, and thus swim locomotion is an assay with potential to identify compounds that may improve healthspan independent of lifespan (CITP, in preparation). Because both GTE and NDGA induced strong species-specific effects on lifespan, we addressed whether these compounds could improve locomotion.
We found that with GTE treatment, all C. elegans strains showed an improvement in swimming ability (Figure 2a). The improved locomotion effect was age-dependent in MY16 and JU775 (8-9% increase in mean swim score at old age as compared to the control, respectively), while N2 showed an overall robust improvement in locomotion (25% increase at young age, 146% increase at old age). The C. tropicalis strains (in which GTE improved lifespan), exhibited strain-specific responses in swimming ability: strain JU1630 showed no difference in locomotion with GTE treatment, while strain QG834 exhibited a small but significant decrease in swimming ability (2-4% decrease with age). JU1373 exhibited a strong age-related response in which GTE robustly decreased swimming ability at young age (14% decrease) but improved swimming ability during old age (16% increase). The effect of GTE on swimming ability in C. briggsae was minimal but strain-specific as we recorded decreases in swimming ability at young age in strains AF16 (23% decrease) and ED3092 (13% decrease), but a small increase at old age in HK104 (6% increase).
We also tested the effect of NDGA on locomotion in our panel of Caenorhabditis strains. With NDGA treatment, swimming ability improved in five of six C. elegans and C. briggsae strains, but none of the C. tropicalis strains (Figure 2b). The effect in C. elegans was observed in two strains, with N2 gaining a general increase (11-13% increase in mean swimming with age) and with JU775 exhibiting a reduction in the age-related decline of swimming ability (6% increase at old age). Interesting, all three C. briggsae strains showed an age-dependent improvement of locomotion at either young age (ED3092, 6% increase), or old age (HK104 and ED3092, 19-21% increase, respectively), irrespective of their reduced lifespan with NDGA treatment. C. tropicalis, which showed either no effect or a negative lifespan effect in response to NDGA, had no difference in swimming ability with NDGA treatment.
Overall, our tests underscore that: 1) longevity and locomotory health are not well correlated; 2) interventions can elicit marked differences in locomotory healthspan, even in the absence of longevity changes.
GTE reduces oxidative stress resistance in some Caenorhabditis strains
We next assayed oxidative stress resistance with GTE treatment across our panel of strains because GTE exerted the most widespread effect on longevity and locomotion of the two antioxidants tested. In seven of the nine strains, GTE treatment reduced the animal’s ability to resist oxidative stress for at least one age tested (Figure 3). This effect was most prominent in the C. briggsae strains, which exhibited significant decreases in survival at all ages tested (ranging from a 17% decrease in mean survival in old AF16 to a 65% decrease in young ED3092). Previous exposure to a mild stressor can result in a more robust response to future stressors, a process known as hormesis36. Continuous exposure to an antioxidant may subsequently reduce reactive oxygen species that promote an organism’s ability to mount an enhanced oxidative stress response when removed from the compound. While GTE does have antioxidant properties37, the inability to resist oxidative stress may reflect an anti-hormetic effect. The only exceptions to the failure of GTE to increase oxidative stress resistance were in C. tropicalis JU1373, which exhibited an age-dependent increase in oxidative stress survival (22% increase at old age), and JU1630, with no effect at either age. C. elegans JU775 also had an increase in oxidative stress survival as compared to the control during old age (35% increase), though this increase was paired with a decrease in oxidative stress survival at young age (31% decrease).
We have found ALM-based thermotolerance assays to be challenging to reproduce across labs (CITP, in preparation). Still, as we continue to assess how broadly across compounds health measures are impacted, we tested thermotolerance consequent GTE treatment. As we previously observed, there was a high amount of variability within our dataset that made it difficult to identify significant effects on thermotolerance, but our results suggest that GTE may protect against thermosensitivity as most of the strains exhibited lifespan extension with GTE treatment (Supplemental Figure 4).
Discussion
Aging is a complex trait with a broad range of observed outcomes. The inter-individual variability in outcome necessitates studies of significant size and replicate number to reliably identify anti-aging interventions. Because of the necessary study sizes, and the need to follow individuals over their lifetime, initial characterization of compound effects on aging is frequently performed in short-lived animal models. The oft-unspoken assumption is that the near universality of phenotypic aging in multi-cellular animals reflects shared underlying causes of aging, and insights gained in model systems will be generalizable across taxa. However, genetically identical individuals in the same environment can exhibit strikingly different aging trajectories. Whether that variability in outcome simply reflects differences in aging symptoms due to unidentified environmental differences or stochastic biological events or, importantly, if those differences reflect different causes of aging experienced by individuals is not known. The CITP attempts to account for issues of inter-individual differences and stochastic events by testing compounds at a large experimental scale using standardized protocols34,35, replicating experiments at three independent sites, and statistically partitioning the variance to determine the sources of variability (Supplemental Table 1 for this study). Additionally, the CITP tests compound interventions across a genetically diverse panel of Caenorhabditis strains and species to identify interventions that are effective independent of genetic background because they act through a widely conserved aging mechanism. The large number of animals studied, and the numbers of independent replicates in CITP assays, increase confidence that reported outcomes observed reflect true biology. Outcomes also underscore that few compounds may be expected to exert overarching impact on lifespan and mobility health. When such compounds are identified, a concerted effort to move them into to mammalian translation pipeline should be mounted.
The CITP workflow
The CITP workflow needs to account for the multiplicative effects on labor needs due to replication of lifespan analyses at the three CITP sites, and the coverage across a genetic diversity panel. As such, CITP lifespan analyses are relatively laborintensive, even when using short lived nematode models. The addition of health measures to the workflow necessitated further streamlining of CITP workflow (see Supplemental Figure 1). In the restructured CITP workflow we evaluate a compound’s ability to extend lifespan and/or healthspan across Caenorhabditis species and strains (Supplemental Figure 1). The pipeline involves preliminary testing in a single lab to narrow down the dosage range (phase one) and target the optimal dose across strains (phase two) before replicating lifespan results across labs (phase three) and evaluating the effect of robust positive compounds on healthspan (phase four). We found that compounds that did not look promising in the preliminary testing would not benefit from further investigation and have since introduced early exit points in the workflow whereby compounds without positive lifespan results are published and dropped from the pipeline 6–8.
Examples of interventions that exert varied effects across species
As a group dedicated to identifying highly reproducible pharmacological interventions that extend lifespan, promote fundamental functionality such as locomotory ability, or both, with efficacy that applies over a genetically diverse Caenorhabditis test set, the CITP has first focused on testing compounds published to be effective in extending lifespan in nematodes or other model systems4–7. The five additional compounds we studied here, 17α-estradiol, acarbose, green tea extract, nordihydroguaiaretic acid, and rapamycin, are representative of those interest class interventions. We evaluated the five compounds for longevity modulation across a genetically diverse test set and pursued two of the most potent promoters of longevity, GTE and NDGA, for impact on locomotory health and stress resistance. We observed that the antioxidants GTE and NDGA modestly extend Caenorhabditis lifespan in a speciesspecific manner (Figure 1). Additionally, the antioxidants exhibited differences according to survival assay protocol, with observed decreased survival for certain genetic backgrounds in manual survival assays contrasting with extended lifespan as determined on the automated C. elegans Lifespan Machines (Figure 1 and Supplemental Figures 2–3). GTE and NDGA confer strain specific impact on swimming ability (Figure 2), but GTE reduces oxidative stress resistance in some Caenorhabditis strains (Figure 3), suggesting that GTE mechanistic targets and the underlying cause(s) of diminished health for these assessments may reflect different processes. While lifespan and health can certainly be uncoupled, and both are plausible targets for intervention, this study combined with previous observations underscores the complex challenge to finding universal lifespan and healthspan extending interventions.
Implications for identifying compounds that slow aging using C. elegans
Our current study holds implications for future study design with chosen methodology potentially complicating identification of anti-aging compounds. For example, the differences observed in compound lifespan effects for manual and automated longevity studies for GTE and NDGA potentially reflect environmental differences between assay types that can profoundly alter compound longevity effect. The differences in measured effects may be easy to explain, for example GTE has been shown to be protective under photo-damage38, and the frequent illumination during automated lifespan analysis could induce photo-damage. Additionally, previous studies have concluded that the most abundant catechin in green tea, epigallocatechin gallate, didn’t extend lifespan under normal laboratory conditions, but did under stress conditions23, consistent with our observations if ALM assays are mildly stressful. Consistent with that interpretation, we previously found that under CITP protocols, measured lifespans for untreated animals are typically shorter in automated analyses relative to manual assays5. But the question remains; what environment is the most informative for translation to subjects outside of a controlled experimental environment? Additionally, the disconnect between lifespan and health measurements, and the genetic background dependency of those effects suggests additional complications that need to be addressed in a screening experimental design. In total, against a small but expanding test set, we observe that compound genetic background and assay type can give rise to differences in lifespan evaluation and health assessment4,5.
Stepping back to offer some perspective, it is likely unrealistic to expect many blockbuster aging interventions using broad based outcome evaluation. Nonetheless, the pursuit of the particular compounds that can positively move outcomes across genetic backgrounds and via multiple measures of heath might be the critical strategy that spotlights priority interventions for mammalian testing.
Conclusion
Our study of five interventions (NDGA, GTE, 17α-estradiol, acarbose, and rapamycin) underscore the complexities of assessing biological outcomes of candidate aging interventions. Using our standardized protocols, we find that the antioxidants GTE and NDGA extend Caenorhabditis lifespan in a speciesspecific manner. GTE and NDGA tests also revealed some assayspecific outcomes—in certain genetic backgrounds we found decreased survival in manual longevity assays, whereas we measured extended lifespan when we determined outcomes using the automated C. elegans Lifespan Machines. GTE and NDGA affected swimming ability in a strain-specific manner, and GTE lowered oxidative stress resistance in some Caenorhabditis strains. Lifespan and healthspan appear uncoupled. Overall, our findings on this test set of interventions underscore how impactful genetic background, selected health assay, and protocol details are in the assessment of intervention effects. Interventions that meet the high bar of efficacy across a broad range of genetic backgrounds and across multiple experimental approaches may prove the exception, but such capacity would establish definitive priority for testing in mammalian models.
Materials and methods
Strains
All natural isolates used were obtained from the Caenorhabditis Genetics Center (CGC) at the University of Minnesota: C. elegans N2, MY16, and JU775; C. briggsae AF16, ED3092, and HK104; C. tropicalis JU1630, JU1383, and QG834. Worms were maintained at 20°C on 60 mm NGM plates seeded with Escherichia coli OP50-1.
Interventions
Compound intervention treatments were performed as previously described4. The compounds used include green tea extract (LKT Laboratories, Inc. G6817, lot #2595901), nordihydroguaiaretic acid (Sigma-Aldrich 74540, lot #BCBQ4489V), α-estradiol (Sigma-Aldrich E8750, lot #016M4175V), rapamycin (LC laboratories R-5000, lot #ASW-135, and acarbose (Sigma-Aldrich A8980, lot #MKBS1059V0). Compounds were obtained as solids and dissolved in either water or DMSO (dimethyl sulfoxide) to obtain stock solutions, with either water or DMSO used to treat control agar plates. DMSO stock solutions (both compound and control) were then further diluted with water to create working solutions to allow for even distribution across the agar plate while maintaining a final concentration of 0.25% DMSO. Agar plates were treated with compound stock solutions such that the final volume was assumed equal to the volume of the agar.
Manual lifespan assay
Per the previously published CITP standard operating procedure4, synchronized populations were generated via timed egg lays on 60 mm NGM plates. At day one of adulthood, 50 worms were transferred to 35 mm NGM plates containing 51 μM FUdR and compound intervention (or the solvent control). Worms were then transferred to fresh plates and scored as alive or dead on day two and five of adulthood for C. elegans and C. briggsae, or day two and four of adulthood for C. tropicalis. Thereafter, worms were transferred once weekly and scored every Monday, Wednesday, and Friday until dead. Death was defined as a lack of response when stimulated with a platinum wire.
Automated lifespan assay
ALM assays were performed as previously described5,32,34, based on modification of the protocols published for the Lifespan Machine31. Briefly, worms were age synchronized and transferred to intervention plates as described above. One week post egg lay, animals were transferred to 50 mm tight-lidded, intervention treated, modified NGM plates containing 51 μM FUdR and 100 μM nystatin and loaded onto the ALMs. Scanner data was collected and analyzed using the Lifespan Machine software (https://github.com/nstroustrup/lifespan; 31 and strainspecific posture files5).
Swimming ability assay
Swimming ability was measured per the standard CITP protocol (CITP, in prep;39), using the C. elegans Swim Test system (CeleST)40,41. Worms were age-synchronized and exposed to compound intervention during adulthood as described above, until swimming measurements were collected at ages 6 and 12 of adulthood (C. elegans and C. tropicalis), or ages 8 and 16 of adulthood (C. briggsae). Videos were processed using the CeleST software (https://github.com/DCS-LCSR/CeleST).
Thermotolerance assay
The ability for animals to withstand heat stress was measured as previously published (CITP, in prep), utilizing a modification of the ALM protocol. Worms were synchronized and aged as adults on compound intervention plates, as stated above, until the desired testing age (adult day 6 and 12 for C. elegans and C. tropicalis, days 8 and 16 for C. briggsae). At this time, animals were placed onto 50 mm plates with modified NGM containing 100 μM nystatin, without FUdR or compound intervention, at a density of 70 worms per plate. Plates without lids were then transferred to Automated Lifespan Machines in an incubator set to 32°C and 50% humidity. Scanner data were collected using an increased scan speed and reduced resolution to provide proper temporal resolution, needed as a result of the shortened survival under these conditions. Images analyzed using the Lifespan Machine software (https://github.com/nstroustrup/lifespan)31 and strain-specific posture files 5 and deaths were validated by hand.
Oxidative stress resistance assay
Resistance to oxidative stress was measured per the standard CITP procedure (CITP, in prep). Worms were prepared and aged as adults on interventions as mentioned above, with the same ages tested as in the swim test and thermotolerance assays. At the desired age, animals were transferred at a density of 70 worms per plate to 50 mm tight-lidded plates with modified NGM containing 40 mM paraquat (or methyl viologen dichloride, from Sigma-Aldrich), 51 μM FUdR, 100 μM nystatin. Plates were then transferred to ALMs at 20°C and scanner data collection, processing, and analyzing was done with the same methodology mentioned for the thermotolerance assay.
Tables
Supplemental Figures
Acknowledgements
We acknowledge the members of the Lithgow, Driscoll and Phillips labs for helpful discussions. We thank the CITP Advisory Committee and Ronald Kohanski (National Institute on Aging) for extensive discussion. We thank Asher Cutter, Marie-Anne Félix, and Christian Braendle for providing strains that they had directly collected. Additional strains were provided by the CGC, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440). This work was supported by funding from National Institutes of Health grants (U01 AG045844, U01 AG045864, U01 AG045829, U24 AG056052) and the Glenn Foundation for Medical Research and the Larry L. Hillblom Foundation.