TREATMENT OF AGING OR AGE-RELATED DISORDERS USING XBP1

Abstract

Described is a targeted gene therapy for use in the delay or treatment of a symptomatic stage of aging and/or age-related disease in a subject, in particular to maintain or restore endoplasmic reticulum proteostasis. The gene therapy comprises the administration of a therapeutically effective amount of a pharmaceutically acceptable composition comprising X-box binding protein 1 (XBP1) or an agent that stimulates neuronal expression of XBP1 in the brain of the subject.

Claims

1. A method for the delay or treatment of a symptomatic stage of aging in a subject, the method comprising administering to the subject a therapeutically effective amount of a pharmaceutically acceptable composition comprising X-box binding protein 1 (XBP1) or an agent that stimulates expression of a polypeptide comprising XBP1 in the brain of the subject to maintain or restore endoplasmic reticulum proteostasis in the subject.

2. The method of claim 1, wherein the composition comprises an adeno-associated virus (AAV) vector.

3. The method of claim 2, wherein the AAV vector induces expression of XBP1s.

4. The method of claim 2, wherein the AAV vector induces expression of a fusion protein comprising XBP1 and ATF6, optionally joined by a linker.

5. The method of claim 1, wherein the XBP1 is a mammalian XBP1, preferably human XBP1.

6. The method of claim 4, wherein the ATF6 is a mammalian ATF6, preferably human ATF6.

7. The method of claim 2, wherein the AAV vector is of a serotype selected from the group consisting of AAV2, AAV6, AAV7, AAV8 and AAV9.

8. The method of claim 7, wherein the AAV vector is of the serotype AAV2 or AAV6.

9. The method of claim 1, wherein the symptomatic stage of aging is a decline in basal motor and/or cognitive function associated with aging.

10. The method of claim 1, wherein administration of the composition substantially prevents, reduces, reverses or delays decay in basal cognitive and/or motor capacity.

11. The method of claim 1, wherein the composition is administered to an aged mammalian subject.

12. The method of claim 11, wherein the subject is suffering from age-related cognitive decline.

13. The method of claim 1, wherein the subject is human.

14. The method of claim 1, wherein the pharmaceutically acceptable composition is administered systemically or locally.

15. The method of claim 1, wherein the pharmaceutically acceptable composition is administered by a nasal route or by direct intraventricular or intrathecal injection and the composition passes the haemato-encephalic barrier.

16. The method of claim 1, wherein the agent stimulates expression of a polypeptide comprising XBP1 in the hippocampus.

17. A method for the prevention or treatment of an age-related disorder in a subject, the method comprising administering to the subject a therapeutically effective amount of X-box binding protein 1 (XBP1) or an agent that stimulates or induces expression or over-expression of XBP1 in the brain.

18. The method of claim 17, wherein the disorder is age-related cognitive decline.

19. The method of claim 17, wherein the disorder is age-related motor dysfunction.

20. The method of claim 17, wherein the disorder is progeria or accelerated ageing.

21. The method of claim 19, wherein the agent comprises an adeno-associated virus (AAV) vector.

22. The method of claim 21, wherein the AAV vector induces expression of XBP1s.

23. The method of claim 21, wherein the AAV vector induces expression of a fusion protein comprising XBP1 and ATF6, optionally joined by a linker.

24. The method of claim 17, wherein the XBP1 is a mammalian XBP1, preferably human XBP1.

25. The method of claim 23, wherein the ATF6 is a mammalian ATF6, preferably human ATF6.

26. The method of claim 21, wherein the AAV vector is of a serotype selected from the group consisting of AAV2, AAV6, AAV7, AAV8 and AAV9.

27. The method of claim 26, wherein the AAV vector is of the serotype AAV2 or AAV6.

28. The method of claim 17, wherein the agent is administered to a mammalian subject suffering from an age-related disorder.

29. The method of claim 28, wherein the subject is human.

30. The method of claim 17, wherein a pharmaceutically acceptable composition comprising XBP1 or the agent is administered systemically or locally to the subject.

31. The method of claim 30, wherein the pharmaceutically acceptable composition is administered by a nasal route or by direct intraventricular or intrathecal injection and the composition passes the haemato-encephalic barrier.

32. The method of claim 17, wherein the agent stimulates expression of a polypeptide comprising XBP1 in the hippocampus.

33. The method of claim 1, wherein the polypeptide comprises an amino acid sequence having at least 85%, at least 90%, at least 99% or at least 99% sequence identity to SEQ ID NO:1 or SEQ ID NO:2.

34. The method of claim 17, wherein the polypeptide comprises an amino acid sequence having at least 85%, at least 90%, at least 99% or at least 99% sequence identity to SEQ ID NO:1 or SEQ ID NO:2.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0056] FIG. 1 demonstrates that AAV2-XBP1s gene delivery to aged mouse hippocampus reverts cognitive, functional and morphological changes associated to normal aging.

[0057] FIG. 1A) Representative photomicrography of an aged wild type mice injected with AA2-XBP1s in CA1 region of the hippocampus (magnification 40×). FIG. 1B) Novel Object Location test, 1C) Novel Object Recognition test, and 1D) Barnes Maze test, indicate that aged mice injected with AAV2-XBP1s in the hippocampus (n=15) present normal cognition when compared to aged control animals (n=14). (p=0.0130; p=0.0218; p=0.065 in Day 2 of Barnes Maze).

[0058] FIG. 2 demonstrates that AAV2-XBP1s gene delivery to aged mouse hippocampus reverts cognitive, functional and morphological changes associated to normal aging.

[0059] FIGS. 2A, 2B, 2C, and 2D) Dendritic spikes quantification shows increased proportion of total and mushroom-like dendritic spikes in the CA1 region of aged mice injected with AAV-XBP1s when compared to controls (n=5, 6; n=8 dendrites per animal; p=0.0092).

[0060] FIG. 3 demonstrates that AAV2-XBP1s gene delivery to aged mouse hippocampus reverts cognitive, functional and morphological changes associated to normal aging.

[0061] FIGS. 3A and 3B) Field excitatory post-synaptic potentials increase following stimuli for 60 minutes in slices from aged mice injected with AAV-XBP1s (n=9, 28) or AAV-mock (n=6, 17) slices (p=0.0245, Unpaired t-test followed by Dunnett's post test).

[0062] FIG. 4 demonstrates that AAV2-XBP1s gene delivery to aged mouse hippocampus reverts cognitive, functional and morphological changes associated to normal aging.

[0063] FIGS. 4A, 4B, 4C, 4D, and 4E) Firing rate of spontaneous hippocam pal activity (SA) in slices from aged AAV-XBP1s or AAV-mock animals; (n=9, 6; p<0.0001) treated with picrotoxin (PTX) or untreated (SA). Frequency distribution in function of time (down) of firing rate and spikes bursts (%).

[0064] FIG. 5 shows that IRE1 ablation in the brain accelerates and exacerbates age-associated cognitive and motor decline in mammals without influencing behavior during youth. FIG. 5A) New Object Recognition test indicates that aged wild type (n=17) mice fail to recognize new presented objects when compared to middle aged (n=24) or young animals (n=14; p<0.0001, One-way ANOVA followed by Tukey's post-test). IRE1.sup.cKO mice, on the other hand, when evaluated by the same test present decreased capacity to recognize novel objects during middle age but not during youth (n=23, 16; two-way ANOVA followed by Sidak's post test, p=0.0004).

[0065] FIG. 5B) The Barnes Maze test indicates that on test day 5, aged IRE1.sup.cKO mice make more mistakes until they can find the target hole when compared to aged wild types, which does not happen during middle age or youth (n=9, 9, 9, 9, 10, 8; Unpaired t Student's test, p=0.0092). FIG. 5C) Contextual fear conditioning test indicates that young IRE1.sup.cKO mice present normal freezing responses compared to wild types (n=7, 10; unpaired t Student's Test, p=0.2305). Middle aged knock out, however, present decreased freezing responses compared to middle aged wild type (n=9, 14; unpaired t Student's test, p=0.0012). FIG. 5D) Hang test was performed in wild type and IRE1.sup.cKO mice at different ages to evaluate motor performance. Old IRE1.sup.cKO mice (n=18) were compared with aged WT animals (n=15) and presented decreased motor function at this age (unpaired t Student's test, p=0.0002) which was not observed during middle age (n=11, 15; p=0.6121) or youth (n=10, 12; p=0.6434). FIG. 5E) The rotarod test was used to evaluate motor performance in wild type (n=16, 8, 18) and IRE1cKO mice (n=12, 8, 6) during youth, middle age or advanced age. Latencies to fall were plotted and compared within each age and showed that only aged animals presented decreased function following IRE1 ablation in the brain (n=17, 8; unpaired Student's t test, p=0.0351). FIG. 5F) Young IRE1.sup.cKO mice present normal behavior as shown by the New Object Location test (n=17, 22; p=0.1741). During middle age, animals lose the capacity to discriminate between objects (n=22, 13; p=0.104).

[0066] FIG. 6 shows that IRE1 ablation in the brain accelerates and exacerbates age-associated cognitive and motor decline in mammals without influencing behavior during youth. FIGS. 6A, 6B, and 6C) Western Blots for PSD95, synaptophysin and SNAP-25 indicates a decreased Synaptophysin and SNAP-25 content in the hippocampus and cerebral cortex of middle aged and aged IRE1 Nest KO mice when compared to controls (n=5, 6, 6, 6, p values indicated).

[0067] FIG. 7 shows that IRE1 ablation in the brain accelerates and exacerbates age-associated cognitive and motor decline in mammals without influencing behavior during youth. FIGS. 7A, 7B, and 7C) Compound Muscle Action Potentials (CMAPs) indicate there is no difference in voltage amplitudes comparing IRE1.sup.cKO mice with wild types during the course of aging (n=11, 11, 10, 10, 11, 7; p>0.05).

[0068] FIG. 8 shows that IRE1 ablation in the brain accelerates and exacerbates age-associated cognitive and motor decline in mammals without influencing behavior during youth. FIGS. 8A and 8B) Neuromuscular junction evaluation indicate there is no difference in morphology when comparing IRE1.sup.cKO with wild type mice (n=4, 4).

[0069] FIG. 9 shows that XBP1s overexpression in the brain using transgenic mice prevents the emergence of age-associated motor and cognitive dysfunction in mammals. FIG. 9A) New Object Location test indicates that middle aged (n=12, 13; p=0.0007) and aged (n=15, 9; p=0.0417) XBP1s transgenic mice can discriminate between new located objects when compared to age-matched animals. FIG. 9B) New Object Recognition test indicates that aged XBP1s transgenic mice present normal behavior as expected for young wild type mice (n=17, 12; p=0.0130). FIG. 9C) Barnes Maze test shows that aged XBP1s transgenic mice (n=8) spend less time to get to the target hole when compared to aged wild type mice (n=10) in day 1 (p=0.0039), day 2 (p=0.0028), day 3(p=0.018) and day 4 (p=0.003). FIG. 9D) Middle aged (n=8, 8; p=0.040) and aged (n=8, 8; p=0.0056) XBP1s transgenic mice exhibit better scores in the hang test when compared to age-matched wild types. No significant difference is observed during youth (n=8, 8; p=0.107). FIG. 9E) Rotarod test indicates that aged XBP1s transgenic mice (n=10) show increased latency to fall from the rod when compared to aged wild types (n=17) as there is no significant difference during youth (n=16, 12; p=0.566). FIG. 9F) No significant difference was encountered in the morphology of aged XBP1s transgenic mice when compared to wild types (n=7, 7).

[0070] FIG. 10 shows that XBP1s overexpression in the brain using transgenic mice prevents the emergence of age-associated motor and cognitive dysfunction in mammals. FIGS. 10A, 10B, and 10C) CMAP readings indicate there is a decreased amplitude in the muscles of young XBP1s transgenic mice (n=6) when compared to young wild types mice (n=10) that is not maintained during aging when evaluating middle aged (n=11, 8) or aged (n=10, 8) mice.

EXAMPLES

[0071] As the aging process is the most important risk factor to the emergence of dementia and neurodegenerative diseases, a technology has been developed and evaluated to revert the decay in the buffering capacity of the brain during aging, aiming to revert the normal decay in brain function that may even translate into the development of dementia or Alzheimer. In this scenario, a viral vector (adeno-associated virus) has been developed to drive the ectopic overexpression of an active from of the transcription factor XBP1 in mammals. XBP1 is a master regulator of the unfolded protein response (UPR) that establishes pro-survival and repair gene expression programs to restore proteostasis, in addition to controlling synaptic function by directly controlling the expression of synaptic proteins and neurotrophins such as brain-derived neurotrophic factor (BDNF).

Example 1: Delivery of AAV2-XBP1s into the Hippocampus of Aged Mice Reverts Cognitive Decline

[0072] In this example, proteostasis was restored to the brain of aged mice. A local gain-of-function approach was used to assess whether the direct delivery of XBP1s using AAV-based gene therapy to the brain could block the dysfunction in cognition observed during the aging process.

[0073] Middle aged and aged wild type animals were stereotaxically injected with AAV2-XBP1s (prepared e.g. as described in WO 2017/059554) in both hippocampus (FIG. 1A), as this brain region is directly associated to the process of learning and memory, crucial to a normal phenotype in all cognitive tests implemented here. Extraordinarily, middle aged and old mice overexpressing XBP1s presented cognitive function comparable to young animals, which is not observed in aged AAV-mock injected mice (FIGS. 1B, 1C, and 1D). This finding indicates that gene delivery of AAV2-XBP1s to the brain is sufficient to revert age-associated cognitive dysfunction in wild type animals.

[0074] The distribution of mushroom like dendritic spikes was then analyzed, as its counting is associated with increased cognitive capacity in brain circuits and represents a mechanism of synaptic plasticity. The results indicate an increased number of dendritic spikes/rea in mice that received AAV2-XBP1s treatment in the hippocampus (FIGS. 2A to 2D).

[0075] Finally, electrophysiological properties of hippocampal circuitry were evaluated in aged animals treated with AAV2-XBP1s. After electrical stimulation of Schaffer Collateral Fibers, an increased long-term potentiation (LTP) was observed in postsynaptic terminals which is sustained one hour after stimuli (FIGS. 3A and 3B). Interestingly, the slope increase is higher in AAV2-XBP1s injected animals when compared to aged AAV2-mock injected. When evaluating the basal firing rate in hippocampal slices, a decreased response was observed in AAV2-XBP1s injected animals when compared to controls, suggesting that inhibitory circuits are preserved in aged mice following XBP1s overexpression (FIGS. 4A to 4E).

[0076] AAV2-XBP1s was delivered into the hippocampus of aged mice that already presented a cognitive decline. The results indicated the aged animals showed a full reversion of the age-associated cognitive decline when compared to controls animals, recovering synaptic capacity. Thus, this data indicates that XBP1s gene delivery to the brain mediated by AAV infection reverts age-associated dysfunction of the brain at behavioral, morphological and functional levels.

Example 2: Bursting UPR Proteostasis by Expressing XBP1s Improved the Quality of Brain Aging at the Level of Cognitive Function

[0077] In this example, the function of the IRE1-XBP1s axis in the aging process of the mammalian brain was evaluated at the motor and cognitive level using different behavioral approaches. The results show that genetically disrupting ER stress sensor IRE1 function in the brain accelerates and exacerbates age-associated phenotypes. Importantly, increasing XBP1s content in the brain, specifically in the hippocampus, is sufficient to slow down or even block cognitive and motor dysfunction in elderly mice. These results correlate with altered electrophysiological responses, synaptic proteins content and brain-derived neuronal factor (BDNF) levels in the brain of aged mice.

[0078] To evaluate if genetic disruption of the IRE1-XBP1s pathway in the brain would influence age associated cognitive and motor dysfunction, conditional transgenic mice with a mutated form of IRE1 lacking the RNase domain (IRE1.sup.cKO) in the brain were used as previously reported (12). These animals present no mRNA splicing of XBP1s in brain tissue indicating that the pathway was fully inactivated. When comparing wild types with IRE1.sup.cKO mice during youth, no differences were detected in any of the behavioral tests implemented (FIGS. 5A to 5F, 6A to 6C, 7A to 7C, 8A and 8B.). Age matched wild type animals were then compared with different conditional transgenic mice with gain or loss-of-function experiments to evaluate whether genetic manipulation of IRE1-XBP1s axis in the brain may alter age-associated cognitive disruption.

[0079] Interestingly, when comparing those strains during middle age, an early disruption of cognitive function in knock-outs was detected that was expected to happen only in advanced age, as evidenced in aged wild type animals in the new object recognition test (FIG. 5A). When analyzing old IRE1.sup.cKO animals in the Barnes Maze test, we showed that they make more mistakes until they can find the target hole when compared to aged wild type mice (FIG. 5B). Such phenotype is absent during middle age or youth (FIG. 5B). Finally, we performed the Contextual Fear Conditioning (CFC) test in those strains to assess learning and memory with an alternative assay. Results indicate that although young IRE1.sup.cKO present normal phenotype, during middle age those mice display significantly fewer freezing responses when compared to middle aged wild types (FIG. 5C). In conjunction, those findings indicate that ablating a functional IRE1-XBP1s pathway in the brain accelerates or even exacerbates age-associated cognitive disruption although it drives no further cognitive disability during youth. We also measured by western blot the levels of pre- and post-synaptic proteins in IRE1.sup.cKO animals in different ages (FIGS. 6A, 6B, and 6C). Results indicate there is a significant decrease in the content of pre-synaptic proteins synaptophysin and SNAP-25 in the cerebral cortex and hippocampus of middle aged and old IRE1.sup.cKO mice, which correlated with behavioral dysfunction in those time windows.

[0080] To assess the contribution of XBP1s to this process age matched conditional mutants overexpressing XBP1s in the brain, e.g. as described in WO2016/106458 and WO2016/106458, were evaluated using the same cognitive tests. Middle aged XBP1s transgenic mice present normal cognition when evaluated in the New Object Location (NOL) and New Object Recognition (NOR) tests, as expected for young wild type mice (FIGS. 9A and 9B). Remarkably, old XBP1s transgenic mice showed phenotype comparable to young wild types in the Barnes Maze as detected by the latency to find the target hole (FIG. 9C). These results indicate that XBP1s overexpression in the brain attenuates or even blocks the emergence of age-associated brain dysfunction.

Example 3: IRE1-XBP1s Axis Modulates Age-Associated Motor Dysfunction

[0081] Experiments were also carried out using genetically modified mice for XBP1s (transgenic overexpressing active XBP1s in neurons) or IRE1 alpha (conditional knockout mice in the brain) and the normal cognitive and motor decline of mice during aging was evaluated. Remarkably, a functional role of brain UPR during aging is demonstrated here using classical gain- and loss-of-function experiments taking advantage of transgenic animals overexpressing XBP1s in the entire neuronal tissue and knock outs for the upstream UPR sensor IRE1 that could not express XBP1s in the brain. Following the aging course in those different mutant lines, it was observed that the IRE1-XBP1s pathway could alter age-associated cognitive and motor decline as shown by different tasks. Overall, the results indicate that therapeutic strategies to improve ER proteostasis are a possible candidate for interventions to reduce the risk of developing dementia and neurodegeneration and can also be used to ameliorate the normal motor and cognitive decay associated to the aging process.

[0082] Comparing IRE1.sup.cKO with wild type mice during youth, no significant differences were observed in motor function. However, old IRE1.sup.cKO mice presented poorer motor function when compared to old wild types, as evidenced both in the rotarod test and the hanging test. These results suggest an age-dependent phenotype of IRE1 deficiency in motor dysfunction (FIGS. 5D and 5E). When evaluating the amplitudes of compound muscle action potentials (CMAPs) in three distinct muscle groups, aged wild type mice presented decreased amplitudes in all muscles analyzed (FIGS. 7A, 7B, and 7C). Of interest, there is no significant difference between wild types and IRE1.sup.cKO animals in any age analyzed (FIGS. 7A, 7B, and 7C), indicating that age-dependent disruption observed in the knockout animals is likely dependent of higher motor circuits in the brain. Consistent with those findings, no significant alterations were observed in the morphology of neuromuscular junctions of mutant mice when compared to wild types (FIGS. 8A and 8B).

[0083] Age matched XBP1s transgenic mice were compared with wild-type animals to evaluate changes in motor disability during the course of aging. Remarkably, middle aged and aged transgenic animals presented better motor function when compared to wild types as shown by rotarod and hanging tests (FIGS. 9D and 9E) both in the hanging and rotarod tests although no changes could be measured during youth, thus suggesting that XBP1s overexpression in the brain could attenuate age associated motor dysfunction without any further alteration during youth. Again, no significant alterations were observed in the morphology of neuromuscular junctions in those mice (FIG. 9F). Young XBP1s transgenic mice presented slight decreased amplitudes in CMAP readings in the gastrocnemius and tibialis anterior muscles when compared to young wild types although no alterations were observed in middle aged or old animals (FIGS. 10A, 10B, and 10C).

[0084] These results indicate that the IRE1-XBP1s axis exerts a major role in sustaining brain health span during aging in mammals. Thus, targeting this pathway may prove a powerful tool to modulate undesired outcomes of the aging process thus increasing health and life quality of the elderly population.

REFERENCES

[0085] 1. Houtkooper R H, Argmann C, Houten S M, Cant6 C, Jeninga E H, Andreux P A, et al. The metabolic footprint of aging in mice. Sci Rep (2011) 1. [0086] 2. Petersen R C. Mild cognitive impairment as a diagnostic entity. Journal of Internal Medicine (2004) 256(3):183-94. [0087] 3. Brayne C. The elephant in the room—healthy brains in later life, epidemiology and public health. Nat Rev Neurosci (2007) Mar. 8(3):233-9. [0088] 4. Sanford, Sarah, et al. Independence, loss, and social identity: Perspectives on driving cessation and dementia. Dementia (2018):1471301218762838. [0089] 5. Kaushik S, Cuervo AM. Proteostasis and aging. Nature Medicine (2015) 21(12):1406-15. [0090] 6. Martinez, Gabriela et al. Endoplasmic reticulum proteostasis impairment in aging. Aging Cell (2017) Aug.; 16(4): 615-623. [0091] 7. Hetz C. The unfolded protein response: Controlling cell fate decisions under ER stress and beyond. Nat Rev Mol Cell Biol (2012) 13(2):89-102. [0092] 8. Hetz, C., & Saxena, S. ER stress and the unfolded protein response in neurodegeneration. Nature reviews. Neurology (2017) Aug. 13(8):477-491. [0093] 9. Valdes P, Mercado G, Vidal R L, Molina C, Parsons G, Court F A, et al. Control of dopaminergic neuron survival by the unfolded protein response transcription factor XBP1. Proc Natl Acad Sci USA (2014) May 6;111(18):6804-9. [0094] 10. Hetz C, Thielen P, Matus S, Nassif M, Court F, Kiffin R, et al. XBP-1 deficiency in the nervous system protects against amyotrophic lateral sclerosis by increasing autophagy. Genes Dev (2009) Oct. 1;23(19):2294-306. [0095] 11. Vidal R L, Figueroa A, Court F A, Thielen P, Molina C, Wirth C, et al. Targeting the UPR transcription factor XBP1 protects against huntington's disease through the regulation of foxol and autophagy. Hum Mol Genet (2012) May 15;21(10):2245-62. [0096] 12. Duran-Aniotz, C., Cornejo, V. H., Espinoza, S., Ardiles, A. O., Medinas, D. B., Salazar, C., & Scheper, W. IRE1 signaling exacerbates Alzheimer's disease pathogenesis. Acta Neuropathologica (2017) Sep.; 134(3):489-506. [0097] 13. LIU, Sheng-Yuan, et al. Polymorphism-116 C/G of Human X-box-Binding Protein 1 Promoter is Associated with Risk of Alzheimer's Disease. CNS neuroscience & therapeutics (2013) 19(4):229-234. [0098] 14. Martinez, Gabriela, et al. “Regulation of memory formation by the transcription factor XBP1.” Cell reports (2016) 14(6):1382-1394. [0099] 15. Rampon, C., et al. Effects of environmental enrichment on gene expression in the brain. Proc. Natl. Acad. Sci. USA (2000) 97:12880-12884. [0100] 16. Hayashi, A., et al. The role of brain-derived neurotrophic factor (BDNF)-induced XBP1 splicing during brain development. J. Biol. Chem. (2007) 282:34525-34534. [0101] 17. Hayashi, A., Kasahara, T., Kametani, M., and Kato, T. Attenuated BDNF-induced upregulation of GABAergic markers in neurons lacking Xbp1. Biochem. Biophys. Res. Commun. (2008) 376:758-763. [0102] 18. Taylor R C, Dillin A. XBP-1 is a cell-nonautonomous regulator of stress resistance and longevity. Cell (2013) 153(7):1435-47. [0103] 19. LUIS, Nuno Miguel, et al. Intestinal IRE1 is required for increased triglyceride metabolism and longer lifespan under dietary restriction. Cell reports (2016)