Calorie restriction and calorie-restriction mimetics activate chaperone-mediated autophagy

Significance

Calorie restriction (CR) is well known to extend healthspan and preserve the response to stress in old organisms. Loss of proteostasis, one of the hallmarks of aging, is associated with a decline in a selective form of autophagy, chaperone-mediated autophagy (CMA). In this study, we demonstrate that CR activates CMA and that a similar activation can be attained in multiple tissues of old mice even after short treatment with CR mimetic (CRMs) compounds. This activation of CMA is required for the improved response to stress conferred by CRMs to old mice. Our findings highlight CMA as a potential gerotherapeutic target.

Abstract

Chaperone-mediated autophagy (CMA) is part of the mammalian cellular proteostasis network that ensures protein quality control, maintenance of proteome homeostasis, and proteome changes required for the adaptation to stress. Loss of proteostasis is one of the hallmarks of aging. CMA decreases with age in multiple rodent tissues and human cell types. A decrease in lysosomal levels of the lysosome-associated membrane protein type 2A (LAMP2A), the CMA receptor, has been identified as a main reason for declined CMA in aging. Here, we report constitutive activation of CMA with calorie restriction (CR), an intervention that extends healthspan, in old rodent livers and in an in vitro model of CR with cultured fibroblasts. We found that CR-mediated upregulation of CMA is due to improved stability of LAMP2A at the lysosome membrane. We also explore the translational value of our observations using calorie-restriction mimetics (CRMs), pharmacologically active substances that reproduce the biochemical and functional effects of CR. We show that acute treatment of old mice with CRMs also robustly activates CMA in several tissues and that this activation is required for the higher resistance to lipid dietary challenges conferred by treatment with CRMs. We conclude that part of the beneficial effects associated with CR/CRMs could be a consequence of the constitutive activation of CMA mediated by these interventions.

A well-maintained proteome is key for supporting proper cellular function and response to stress. The proteostasis network is composed of protein synthesis and folding machineries as well as protein degradation pathways that serve as surveillance mechanisms for all intracellular proteins. Continuous synthesis and degradation of the proteome assures protein quality control and adaptation to changes in the extracellular environment. Intracellular proteins are degraded by two major systems, the ubiquitin proteosome system and the lysosomal/autophagy system. Loss of proteostasis is one of the hallmarks of aging, and its failure has been shown to underlie many of age-associated diseases and contribute to the aggravating effect of aging in those diseases. Chaperone-mediated autophagy (CMA) is a selective form of lysosomal degradation for soluble intracellular proteins with a KFERQ-like motif in their sequence (1, 2). Once this motif is recognized and bound by heat shock cognate 71 kDa protein (HSC70), the substrate protein is delivered to lysosome-associated membrane protein type 2A (LAMP-2A), that serves as the CMA lysosomal receptor and multimerizes into a substrate translocation complex (3–5). Substrate proteins reach the lysosomal lumen through this complex assisted by the lysosomal-resident HSC70 (6–8). CMA degrades oxidized and damaged proteins and hence plays a critical role in intracellular protein quality control (9). Furthermore, CMA also participates in selective proteome remodeling; by degrading fully functional proteins to terminate their function, CMA modulates diverse cellular processes including lipid and glucose metabolism, cell signaling, cell cycle, and transcriptional regulation, among others (9). In fact, in vivo blockage of CMA leads to metabolic deregulation (10–13) comparable to the metabolic syndrome of aging and to degeneration in multiple tissues (14, 15). Although CMA activity is present in most mammalian cells under basal conditions, different stressors are robust activators of this pathway (16). These stressors include oxidative (17), lipotoxic (18), genotoxic (19), proteotoxic (20), and hypoxic (21) stress, all conditions that exacerbate cellular aging (22). In fact, CMA activity is known to decrease with age (23, 24), primarily due to lower stability of LAMP2A at the lysosomal membrane (18, 25, 26). CMA malfunction accelerates onset of different age-related diseases such as metabolic disorders, atherosclerosis, and neurodegeneration (12–14, 27). The tight relation of CMA malfunctioning with disease provides a rationale for the recent interest in finding interventions that could prevent the decline in CMA activity with age.

Over the years, numerous studies have shown the beneficial effects of calorie restriction (CR) as an intervention to slow down aging and expand healthy lifespan (28). CR involves reduction of daily calorie intake by 30 to 50% without causing malnutrition, and it has proven to be advantageous in all organisms tested to date (29). CR is able to delay the onset of many age-related diseases including cancer, diabetes, and atherosclerosis (30, 31). The molecular mechanisms behind the beneficial effects of CR on aging have not been fully elucidated, although there is general agreement that the beneficial effect of this intervention originates from its ability to modulate many hallmarks of aging such as genomic instability, deregulation of nutrient sensing, altered intracellular communication, mitochondria energetics, as well as proteostasis (32). Since this intervention involves nutritional changes and reducing levels of nutrients is a well-known trigger for autophagy, the contribution of autophagy to CR effects has been investigated. CR improves the regulation of macroautophagy, one of the most studied types of mammalian autophagy (33, 34). Restoration of macroautophagy by CR has shown to be behind the lifespan extension observed in organisms such as Caenorhabditis elegans (35). A direct effect of CR on expression of macroautophagy-related genes has also been demonstrated recently in different tissues from old rats (36). Other macroautophagy regulators shown to be impacted by CR include the PI3K/AKT/MTOR and ERK1/2 MAPK signaling pathways (37), protein kinase CK2 (38), and the transcription factor TFEB (39). The regulatory effect of CR on macroautophagy involves, in part, the reversion of age-dependent changes in acetylation of important macroautophagy proteins (40). Modification of the acetylome is one of the most accepted changes behind the beneficial effect of CR and, in fact, constitutes the basis for the development of pharmacological CR mimetics (CRMs) (41). Despite the many beneficial effects of CR, maintaining such lifestyle has proven to be difficult for many individuals, leading to attempts to pharmacologically replicate the beneficial effects of CR while maintaining a regular diet, which set the basis for development of CRMs. CRMs have been defined as compounds designed to mimic the biochemical effects of CR and thus promote change in the cellular acetylome either by 1) reducing the levels of acetyl-CoA, 2) inhibiting certain acetyltransferases, or 3) stimulating deacetylases (42). CRMs such as spermidine reproduce the regulatory effect of CR on macroautophagy. However, despite CMA also playing an important role in protein quality control and proteome maintenance, and being regulated by nutrient availability, until date, the impact of CR or CRMs on CMA and the possible contribution of this type of autophagy to their beneficial effects in old organisms have not been addressed. The absence of CMA in invertebrate species, preferred experimental model for lifespan extension studies, and the difficulty in tracking CMA activity in multiple organs and tissues at the same time are probably responsible for the lack of studies on the effect of CR and CRMs on CMA.

In this study, we investigated the effect of chronic long-term CR and acute short-term administration of CRMs to rodents on their CMA activity and asked whether these interventions could prevent/reverse the decline in CMA in old animals. We show that CR and CRMs lead to constitutive upregulation of CMA. Both interventions increase lysosomal levels of LAMP2A, the limiting component of CMA, in old mice, in part, by favoring stabilization of this protein at the lysosomal membrane. Furthermore, we demonstrate that CMA is necessary to achieve the protective effect of CRMs against nutritional stress induced in mice by a high carbohydrate diet. These results support that activation of CMA contributes to part of the beneficial effects mediated by CR and CRMs and highlight the gerotherapeutic potential of modulating this type of autophagy.

Results

CMA Is Constitutively Activated during CR.

To test the possible effects of CR on CMA, we utilized 12- and 22-mo-old rats fed normally (ad libitum; AL) or CR (40% reduced caloric intake from 16 wk of age; CR) and compared to 4-mo AL fed rats (Fig. 1A). We focused our study on the liver, because CMA has been extensively characterized in this organ and decline in hepatocyte CMA has been shown to start by 12 mo of age both in rats and mice. The most direct way to measure CMA, independently of the activity of other proteolytic pathways, is by using a well-established in vitro system with isolated lysosomes (7, 43) [SI Appendix, Fig. S1A shows that purity (enrichment) and recovery of lysosomal enzymes in the isolated fractions were comparable in all conditions]. We isolated CMA-active lysosomes (a population of lysosomes competent for CMA owing to the presence of HSC70 inside the lysosomal lumen; CMA+ lysosomes) (7) and incubated them with radiolabeled GAPDH or RNase A, two well-characterized CMA substrate proteins (44). Degradation of these proteins by CMA, measured as their conversion into radiolabeled amino acids, is attained after their binding to the lysosomal membrane and translocation into the lumen. Thus, this assay recapitulates all steps of CMA. We found that, as reported earlier (23, 24, 26), the ability of lysosomes to degrade the radiolabeled substrates decreases as the animals age (Fig. 1B and SI Appendix, Fig. S1B). Strikingly, this decrease was not observed in lysosomes isolated from age-matched CR animals, which instead showed comparable (or even higher) substrate degradation to young AL animals (Fig. 1B and SI Appendix, Fig. S1B). We next explored the ability of lysosomes to up-regulate CMA in response to 20 h starvation [a known CMA activator) (45)] and observed similar significantly higher CMA-dependent protein degradation in the two CR groups (Fig. 1C). To assess whether higher degradation in the CR group was a result of higher activity of the lysosomal proteases or to actual changes in the lysosomal binding and internalization of the CMA substrates, we compared substrate degradation by the same lysosomes but after disrupting the integrity of their membrane with a hypotonic shock to allow free access of proteases to substrates. Proteolytic activity of the two substrates was comparable in lysosomes from AL and CR 12-mo rats and was slightly decreased in 22-mo CR compared to AL, which we attributed to the higher presence of endogenous substrates internalized in the CR group (Fig. 1D and SI Appendix, Fig. S1C). These findings suggest that the observed increase in CMA substrate degradation by lysosomes isolated from CR rats was indeed a result of enhanced substrate binding/uptake.

Fig. 1.

CMA is constitutively activated by CR. (A) Experimental scheme. Lysosomes were isolated from livers of AL or CR male rats, 4, 12, and 22 mo of age. (B and C) In vitro uptake of [¹⁴C]GAPDH (a known CMA substrate) by intact lysosomes isolated from livers of fed (B) or starved (C) 4, 12, and 22 mo AL or CR male rats. Degradation of [¹⁴C]GAPDH was measured as percentage of proteolysis. n = 4 to 6 independent experiments (i.e.) (D) Proteolysis of [¹⁴C]GAPDH in lysosomes broken in parallel with the same groups as in (B and C). n = 5, i.e., (E) In vitro binding (Left) and uptake (Right) of GAPDH by intact liver lysosomes isolated from 4, 12, and 22 mo AL or CR male rats. n = 7 to 10, i.e., (F) Experimental scheme. Mouse fibroblasts (NIH3T3 cells) were maintained for 14 d in serum from AL or CR rats. (G and H) Long-lived intracellular protein degradation in cells maintained in AL or CR serum labeled with [³H]Leucine for 48 h (G). The contribution of lysosomes to total protein degradation was analyzed by treating cells with lysosomal protease inhibitors (NH₄Cl and leupeptin, NL) (H) n = 2, i.e., with two technical replicates. (I) Representative images of NIH3T3 cells expressing KFERQ-PS-Dendra2 reporter maintained in AL or CR serum for 2 wk and switched to serum-supplemented or serum-free media for 18 h. Bottom: Boxed areas at higher magnification where contrast has been uniformly modified to better visualize puncta against the cytosolic signal of the reporter. Nuclei are stained with DAPI. Right shows quantification of the number of fluorescent reporter puncta per cell. n = 3, i.e., with at least 1,000 cells per condition using high-content microscopy. Values are mean ± SEM. The two-way ANOVA test was used. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001, ns: not significant.

We separately analyzed these two CMA steps using another in vitro assay in which purified single CMA substrate proteins are incubated with lysosomes untreated or treated with inhibitors of lysosomal proteases to prevent the degradation of the internalized substrate (7). In untreated lysosomes, the internalized protein is rapidly degraded and only protein bound to the membrane can be detected (binding), whereas uptake can be calculated as the difference of protein associated with treated and untreated lysosomes. As shown in Fig. 1E, the amount of GAPDH bound to the lysosomal membrane (Left) and translocated inside the lysosomes (Right) was significantly higher in the lysosomes from CR animals at the two ages analyzed. These results indicate that CR constitutively up-regulates CMA in hepatocytes through increased lysosomal binding and internalization of its substrates.

Next, we confirmed these findings in intact cells, taking advantage of a previously established cell culture model of CR based on supplementation of the cultured media for 2 wk with serum from AL or CR rats (46) (Fig. 1F). Analysis of long-lived protein degradation rates (preferential substrates of the lysosomal system) after labeling newly synthesized proteins in the cultured cells with ³H-leucine for 48 h (25) revealed significantly higher rates of intracellular protein degradation in cells maintained in CR serum as compared to those cultured in AL serum (Fig. 1G). Differences in degradation between both groups were abolished in the presence of lysosomal protease inhibitors (ammonium chloride (NH₄Cl) and leupeptin, NL) (Fig. 1H) supporting that the upregulation in protein degradation with CR was occurring in lysosomes. As in the studies with isolated lysosomes, not only basal lysosomal protein degradation but the one induced by switching cells to serum-free medium after radiolabeling was also significantly higher in cells previously exposed to CR serum (SI Appendix, Fig. S1 D and E). To directly analyze CMA in the cultured cells, we next utilized a previously described fluorescent CMA reporter KFERQ-PS-Dendra2 that when delivered to lysosomes via CMA highlights these compartments as fluorescent puncta (47). Quantification of fluorescent puncta per cell provides a reliable assessment of changes in CMA activity. We found that cells cultured in CR serum had a threefold increase in CMA reporter fluorescent puncta as compared to those maintained in AL serum (Fig. 1I) and upon culture in serum-free media, the cells previously maintained in CR serum showed an even higher increase in the number of puncta than the AL serum-free cells (Fig. 1I). The CMA reporter mainly tracks binding of the reporter protein to LAMP2A, because once translocated into lysosomes, it loses its fluorescence due to unfolding (47). Although in most instances substrate binding directly correlates with the level of internalization, to confirm that this was also the case in the CR-induced increase in reporter binding, we inhibited lysosomal degradation by treating cells with NL and then used an antibody to detect the Dendra protein accumulated in the lysosomal lumen (48). As shown in SI Appendix, Fig. S1F, CR serum increased the amount of KFERQ-Dendra internalized and degraded in lysosomes, confirming thus higher CMA activity.

In addition to the higher CMA activity per lysosome detected in the in vitro system with isolated lysosomes, the studies in cultured cells revealed an expansion of the pool of lysosomes active for CMA in cells maintained in CR serum. We have previously reported that changes in the lysosomal pool active for CMA can be monitored by analyzing colocalization of LAMP2A (present in all types of lysosomes) with HSC70 (only present in CMA+ lysosomes) (7). Costaining for both proteins revealed a significant increase in their colocalization in cells maintained in CR serum when compared to AL, that was comparable to the one observed upon serum removal in cells previously maintained in AL serum (SI Appendix, Fig. S1G). Interestingly, we did not observe differences in LAMP2A/HSC70 colocalization upon serum removal between CR and AL serum pretreated cells, suggesting that while constitutive upregulation of basal CMA activity by CR was attained through combination of an increase in both activity per lysosome and number of lysosomes capable to perform CMA, the increase in inducible CMA by CR was mostly a result of higher activity per lysosome. As described before, activation of CMA by CR under basal conditions associated with mobilization of CMA-active lysosomes (LAMP2A+ HSC70+) toward the perinuclear area (17, 23, 49) resembling the distribution observed only in AL cells upon serum removal (SI Appendix, Fig. S1 G, H).

Overall, these results support a constitutive activation of CMA by CR in both the in vivo and in vitro models tested.

CR Stabilizes Lysosomal LAMP2A Levels.

To elucidate the mechanism(s) behind the observed CR-mediated constitutive activation of CMA, we first analyzed possible transcriptional changes in the components of the CMA network, the subset of genes that contribute to this pathway either as effectors or as positive or negative regulators (2). Expression level of LAMP2A, the limiting CMA component, was reduced with age but old CR mice still showed reduced expression when compared to young mice (Fig. 2A and SI Appendix, Fig. S2A). Interestingly, CR up-regulated expression of several of the chaperones known to participate in CMA and restored the up-regulated expression of several CMA negative regulators (such as BMAL1, AKT1) that displayed increased expression in 22-mo-old mice (Fig. 2A and SI Appendix, Fig. S2A). Since all these genes also participate in other cellular processes, to place their changes in the context of CMA, we utilized the recently developed CMA score that predicts the overall impact on CMA taking into consideration the weight and direction (activation or inhibition) of each of these genes on CMA (14). To calculate the CMA score, we used the expression values of each of the components of the CMA network and multiplied the expression of LAMP2A by 2 since it is the limiting component of this pathway. Expression levels of effectors and positive CMA regulators were added up and the additive expression of the negative regulators was subtracted from this value, which was corrected by the sum of assigned weights. We observed that the CMA score, predicted through transcriptional changes, was indeed significantly decreased in 22-mo-old AL mice, whereas in old CR mice, the CMA score increased to levels comparable to those in young mice (Fig. 2B) coinciding with the observed constitutive upregulation of CMA in livers from CR rodents.

Fig. 2.

CR stabilizes lysosomal LAMP2A levels. (A) Heat map of the expression of components of the CMA network in livers of 4-mo, 22-mo AL, CR male mice. n = 5 mice. (B) CMA index calculated from the expression of genes shown in A. (C) Immunoblot for LAMP2A and other abundant lysosomal membrane proteins in homogenate (HOM) and CMA-active lysosomes (CMA+) from livers of 3, 12, and 22 mo AL or CR male rats. Quantification for 3 and 22 mo is shown on the Right. Quantification for 3 and 12 mo is shown in SI Appendix, Fig. S2B. n = 2 (homogenates) and 4 (lysosomes) rats. (D) Immunoblot for key CMA chaperones in CMA+ and CMA− lysosomes from animals as in C. Cathepsin B (CATH B) and LAMP1 are shown as a lysosomal enrichment marker and levels of LAMP2A in isolated membranes from the same lysosomes are shown at the Bottom for comparison of CR induced changes. Equal amount of total protein was loaded per lane in the lysosome blots (Top), whereas the total amount of membrane isolated from equal protein amount of lysosomes was loaded in the bottom gel to account for differences in cargo content among the fractions. Ponceau staining of the lysosomal membrane fractions shows that only lysosomal membranes from 12-mo-old rats display changes in cargo/membrane ratio. Densitometric values for L2A in the membranes upon correction for protein loaded and relative to the 3 mo fraction are shown. (E) Immunoblot for LAMP2A (Left) and percentage of protein remaining at the indicated times (Right) in CMA+ lysosomes from 4, 12, and 22 mo CR rat livers incubated at 37 °C in the absence or presence (+PI) of protease inhibitors. n = 4 rats. (F) Immunoblot for LAMP2A (L2A) of lysosomal membranes from 12 and 22 mo AL or CR, incubated at 37 °C for 20 min. Black arrow: full length (FL) L2A. White arrowheads: L2A fragments (f). Right: Densitometric scanning of L2A from a lower exposure image to better resolve bands. Areas of FL and the different L2A fragments (f1 to f4) of each condition were calculated and the percentage of total L2A present in each fraction is shown in the graph. This experiment was repeated twice with similar results. (G) Immunoblot for LAMP2A in detergent resistant (DR, lipid microdomains), intermediate (Int), and solubilized (Sol) fractions of CMA+ lysosomal membranes from 3 and 22 mo AL or CR male rats. Quantification of percentage in DR fractions is shown on the Right. n = 3 rats. Molecular weight markers of the regions shown in the immunoblots and Ponceau staining are shown on the Right. One-way ANOVA (B and C) and two-way ANOVA (E and G) tests were used. Values are mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant.

Analysis of the protein levels of the CMA receptor LAMP2A revealed that, as previously reported, levels of LAMP2A in CMA+ lysosomes decreased in 22-mo-old rats (23, 24), while total cellular levels remained unchanged (Fig. 2C). Consistent with the observed CMA upregulation (Fig. 1 B and E and SI Appendix, Fig. S1B), 22-mo CR rats showed lysosomal levels of LAMP2A comparable to those in young animals (Fig. 2C). These changes in lysosomal LAMP2A seem specific for this protein as levels of LAMP1, another lysosomal membrane protein, remained unchanged, and although total cellular levels of LAMP2C, a spliced variant of the lamp2 gene, increased with age and with CR, lysosomal levels of this variant were comparable in CR and AL rats (Fig. 2C). Despite the upregulation of CMA observed in mid-age CR animals compared to AL (Fig. 1 B and E and SI Appendix, Fig. S1B), we did not find a consistent increase in levels of LAMP2A between lysosomes from AL and CR rats at 12 mo of age (SI Appendix, Fig. S2B); however, the increase became evident at both ages when equal amounts of lysosomal membranes, instead of total lysosomal content, were analyzed (Fig. 2D). It is likely that the higher affluency of endogenous substrates to lysosomes from CR animals, as a result of their increased CMA activity, makes the contribution of cargo higher than the contribution of lysosomal resident proteins when loading equal amounts of total lysosomes.

We have previously described that by 12 mo of age, when lysosomal levels of LAMP2A start to decrease in rat and mouse livers, the number of lysosomes containing HSC70, and hence able to perform CMA, significantly increases. We attributed this change to an attempt to compensate for the reduced levels of the LAMP2A receptor per lysosome by recruiting additional lysosomes to perform CMA, since all types of lysosomes already contain LAMP2A and the inactive ones are only missing HSC70 in their lumen (23). We can separately isolate the subgroup of lysosomes usually competent for CMA (CMA+) and those normally engaged in other types of lysosomal degradation (CMA−) (7). Although overall levels of HSC70 remained unchanged with age or CR (SI Appendix, Fig. S2C), comparative analysis of both types of lysosomes confirmed the presence of higher levels of HSC70 in CMA- lysosomes by 12 mo of age, coinciding with the beginning of decrease in LAMP2A levels at the lysosomal membrane (Fig. 2D). Interestingly, we did not detect this increase in HSC70 in CMA- lysosomes from CR mice (Fig. 2D). Since the presence of HSC70 in the group of lysosomes usually inactive for CMA observed by 12 mo in the control group was no longer evident in the CR mice, we propose that preservation of lysosomal LAMP2A levels by CR made the additional compensatory recruitment of lysosomes to perform CMA observed in old animals unnecessary. Total cellular levels (SI Appendix, Fig. S2C) and distribution of other CMA-related chaperones and of LAMP1 between both groups of lysosomes (Fig. 2D) were comparable in CR and AL rats.

We next investigated the mechanism(s) whereby CR preserves high levels of LAMP2A at the lysosomal membrane. Since we confirmed that as in mice, levels of LAMP2A mRNA also remained unchanged upon CR in rats (SI Appendix, Fig. S2D), we focused on the posttranscriptional regulation of LAMP2A at the lysosome, where most of the CR-associated increase in LAMP2A levels seems to occur (Fig. 2 C and D). Changes in LAMP2A degradation at the lysosomal membrane have been shown to modulate CMA activity in conditions of nutritional stress such as starvation (25), and this tightly regulated degradation has proven disrupted with age (26). In fact, under normal conditions, turnover of LAMP2A in lysosomes requires its lateral mobilization to specific lipid microdomains at the lysosomal membrane where it is cleaved at the transition between the transmembrane and luminal regions (50). However, with age, changes in the lipid composition of the lysosomal membrane reduce entry of LAMP2A into these microdomains (18) and result instead in up-regulated aberrant degradation of this protein by lysosomal proteases (26). We can reproduce LAMP2A-regulated degradation in isolated lysosomes by incubating them in the absence of substrate and presence of calcium, which promotes binding of cathepsin A (the protease responsible for the cleavage of LAMP2A) to the microdomains (25, 51). As reported before, we observed a marked decrease in this regulated degradation of LAMP2A in lysosomes isolated from old animals, but interestingly, old mice subjected to CR showed similar degradation profile as young animals (Fig. 2E). To visualize the previously described abnormal fragmentation of LAMP2A in old liver lysosomes, we incubated only the lysosomal membranes, to prevent rapid degradation of LAMP2A fragments in the lumen, and found that they were no longer observed in lysosomes isolated from age-matched CR rats (Fig. 2F). As further confirmation of the restoration of regulated degradation of LAMP2A by CR, we analyzed the association of this protein to lysosomal lipid microdomains using previously established isolation procedures (50). We found increased levels of LAMP2A in the detergent-resistant microdomains isolated from old CR animals compared to old AL animals (Fig. 2G).

These findings support the idea that CR preserves CMA activity until late in life by maintaining proper turnover of the receptor protein for this pathway.

CR Mimetics Activate CMA.

Given the ability of CR to constitutively activate CMA in vitro and in vivo and the reported beneficial effects of preserving the activity of this autophagic pathway through genetic manipulation in old mice (13, 52), we next set to investigate the effect of CR mimetics (CRMs) on CMA. We chose CRMs that modulate protein acetylation through two different mechanisms: 1) spermidine and a new-generation compound, C646, that inhibit acetyltransferases proven to act as a sink of acetyl-CoA in aging (i.e., EP300 acetyltransferase), and 2) hydroxycitrate and the new-generation compound SB204990 that modulate acetyl-CoA levels by inhibiting ATP citrate lyase (42, 53–58). All of these CRMs have shown promising results in promoting healthspan in multiple species (42, 59). As expected, these CRMs reduced cytoplasmic protein acetylation in mouse fibroblasts (NIH3T3) (SI Appendix, Fig. S3 A and B). To analyze the effect of these CRMs on CMA, we first treated the mouse fibroblasts stably expressing the CMA reporter (KFERQ-PS-Dendra2), with increasing concentrations of the CRMs and prioritized comparisons by their mechanism of action. All of them induced a dose-dependent increase in CMA (measured as increased number of reporter fluorescent puncta), albeit of different magnitude and dose range (Fig. 3 A and B). We confirmed that CRMs activate CMA as early as 24 h and that, as anticipated, the effect of the new-generation CRMs was long lasting (Fig. 3 C and D and SI Appendix, Fig. S3 C and D). As shown in Fig. 3 E and F and SI Appendix, Fig. S3 E and F, CRMs also induced an increase in the amount of KFERQ-Dendra internalized and degraded in lysosomes, thus confirming higher CMA activity.

Fig. 3.

CRMs activate CMA in vitro. (A and B) Scheme of CMA reporter and representative images (A) and quantification (B) of CMA activity in mouse fibroblasts (NIH3T3 cells) stably expressing KFERQ-PS-Dendra2 CMA reporter, treated with C646, SB204990 (SB), Spermidine (SP), or Hydroxycitrate (HC) at the indicated concentrations. Nuclei are highlighted with DAPI. Quantification of CMA reporter is shown as the number of fluorescent puncta per cell. n = 5 to 7 independent experiments (i.e.), with >500 cells per condition using high-content microscopy. (C and D) Representative images (C) and quantification (D) of CMA activity in cells as in A, treated with C646 (10 μM) or SB204990 (50 μM) for the indicated times. Quantification of CMA reporter shown as the number of fluorescent puncta per cell. Insets show higher magnification regions. Nuclei are highlighted with DAPI. n = 3, i.e., (E and F) Representative images of immunofluorescence for Dendra (E) and quantification (F) of KFERQ-Dendra puncta per cell (Left) and degradation (Right) in cells as in A, treated with C646 (10 μM) or SB204990 (SB, 50 μM), untreated (None) or treated with NH₄Cl and leupeptin (+NL) to prevent lysosomal proteolysis. Insets on the Right show boxed areas at higher magnification. Nuclei are shown with DAPI. Quantification shows the number of fluorescent puncta per cell quantifying >50 cells per condition. n = 3, i.e., One-way ANOVA test (Left) and multiple comparisons relative to None (Right) are shown. Values are mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001, ns: not significant.

The stimulatory effect of CRMs on CMA in vitro motivated us to test whether this effect could also be reproduced in vivo in animals already displaying an age-related decline in CMA. To this effect, we used 18-m-old KFERQ-Dendra transgenic mice (KFERQ-PS-Dendra2 mice) in which the CMA reporter is expressed ubiquitously (48). Out of the four CRMs tested in cells in culture, we chose hydroxycitrate (HC) and C646 for in vivo testing based on their pharmacokinetic properties and to cover two different mechanisms of action and compound generation. CRMs were administered intraperitoneally for two consecutive days, and CMA activity was measured as the number of Dendra⁺ puncta colocalized to the endolysosomal marker LAMP1 in the liver, kidney, and pancreas, three organs where we have seen a marked reduction in CMA activity at this age. Labeling of the endolysosomes with LAMP1 also allowed us to analyze possible changes in this compartment and in the percentage of lysosomes competent for CMA. This brief 2-d treatment with either of the CRMs was sufficient to elicit a significant increase in the number of Dendra⁺/LAMP1⁺ puncta in the liver (Fig. 4 A and B). We also observed a significant increase in the percentage of lysosomes competent for CMA (Fig. 4C) that only in the case of C646 was associated with an increase in the overall number of endolysosomes (LAMP1⁺ compartments) (SI Appendix, Fig. S4A). Interestingly, the effect of CRMs on CMA seems to show some cell and tissue-specific differences both in the magnitude and in the underlying mechanism. Thus, while both CRMs were efficient in significantly up-regulating CMA in the acinar cells of the exocrine pancreas (Fig. 4 D and E), this was attained without significant changes in the percentage of CMA-active lysosomes (although an increasing trend was noticeable) or in the total number of endolysosomes in the pancreas of these animals (Fig. 4F and SI Appendix, Fig. S4B). In the kidney, only treatment with HC resulted in a significantly increased number of Dendra⁺ LAMP1⁺ puncta (Fig. 4 G and H) although an increase in the percentage of lysosomes active for CMA was noticeable for both C646 and HC (Fig. 4I). HC had no effect on the total number of endolysosomes in the kidney, whereas C646 decreased the LAMP1⁺ puncta (SI Appendix, Fig. S4C), explaining why despite having a higher percentage of CMA-active lysosomes the overall number of these lysosomes, and consequently, CMA activity was lower than in the HC treated mice. Overall, these results support that CRMs are able to activate CMA in old animals in the different tissues tested, by increasing the number of CMA-competent lysosomes in the liver and kidney but not in the pancreas, wherein CMA activity increased mostly because of higher activity per lysosome.

Fig. 4.

CRMs activate CMA in vivo in old mice. (A) Representative immunofluorescence staining for Dendra and LAMP1 of the liver of 18-mo-old KFERQ-Dendra male mice treated with saline, C646, or Hydroxycitrate (HC) for two consecutive days. The Inset shows higher magnification of the boxed area. Arrows: Dendra+ puncta. (B and C) Quantification of Dendra+ LAMP1+ puncta per cell as total (B) and as percent of LAMP1+ endolysosomes (C). An average of 45 cells, over four fields per animal were quantified. Nuclei are stained with DAPI. n = 5 mice per group. (D) Representative immunofluorescence staining of the pancreas of mice as in (A) with Dendra and LAMP1. The Inset shows higher magnification. Arrows: Dendra+ puncta. (E and F) Quantification of Dendra+ LAMP1+ puncta per cell as total (E) and as percent of LAMP1+ endolysosomes (F). An average of 200 cells, over four fields per animal were quantified. Nuclei are stained with DAPI. n = 5 mice per group. (G) Representative immunofluorescence staining of the kidney of mice as in (A) with Dendra and LAMP1. The Inset shows higher magnification. Arrows: Dendra+ puncta. (H and I) Quantification of Dendra+ LAMP1+ puncta per cell as total (H) and as percent of LAMP1+ endolysosomes (I). An average of 40 cells, over four fields per animal were quantified. Nuclei are stained with DAPI. n = 5 mice per group. (J and K) Representative immunoblots (J) and quantification (K) for the indicated proteins in liver fractions of total homogenate (Homg.), cytosol (Cyto.), CMA-active lysosomes (CMA+), and CMA-inactive lysosomes (CMA−) from liver of saline (S), C646 (C), and hydroxycitrate (H) treated 18-mo-old mice as in A. n = 5 mice per group. (L) Quantification of HSC70 in CMA+ and CMA− lysosomes from blots as in J. n = 3 mice per group. (M and N) Heat map of the expression in the liver of the indicated CMA network genes (M) and CMA score calculated from M (N) for the three groups of mice. n = 5 mice per group. Molecular weight markers of the regions shown in the immunoblots and Ponceau staining are shown on the Right. Values are mean ± SEM. The one-way ANOVA test was used. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001, ns: not significant.

As in the case of CR, acute treatment with CRMs did not affect the proteolytic activity of the lysosomal luminal enzymes (SI Appendix, Fig. S4D) or the stability of the lysosomal membrane detected as the release of the lysosomal enzyme hexosaminidase (SI Appendix, Fig. S4E). While total levels of LAMP2A in liver homogenate and CMA-inactive lysosomes remained unchanged, we found a significant increase in the levels of this protein in CMA+ lysosomes isolated from livers of C646 treated old mice and an upward trend with HC treatment (Fig. 4 J and K). In contrast, lysosomal levels of HSC70 and the distribution of this chaperone between CMA+ and CMA− lysosomes remained unchanged after the acute treatments with CRMs (Fig. 4 J and L). To determine whether the increase observed in LAMP2A was transcriptionally regulated, we analyzed the expression of LAMP2A and the other two spliced variants of the lamp2 gene. We found that expression of LAMP2A, LAMP2B, and LAMP2C in CRM-treated mouse livers was comparable to saline-treated mice (SI Appendix, Fig. S4F). Transcriptional analysis of the other components of the CMA network did not reveal significant changes in the expression of any of the CMA effectors or positive and negative regulators (Fig. 4M and SI Appendix, Fig. S4G). In agreement with the absence of transcriptional changes, the CMA score calculated from the expression of each of the CMA components remained unchanged in the livers of CRM-treated animals further confirming that the overall increase in CMA activity with CRM was not due to transcriptional upregulations (Fig. 4N).

Overall, our results support an activating effect of CRMs on CMA in multiple organs of aged mice, even when administered for a short period of time.

Beneficial Effects of CRMs Are Dependent on CMA.

A common feature of aging organisms, their reduced capacity to respond to stress, has proven to be corrected through use of CRMs in different experimental settings. To investigate whether these beneficial effects of CRMs in the aging response to stress could be dependent on the observed upregulation of CMA, we challenged a group of 18-m-old wild-type and CMA-deficient mice (^L2AKO) (27) with high carbohydrate diet (HCD) for 4 wk. We choose this challenge because of the well-characterized protective effect of CRMs against dietary challenges (60) and the described role of CMA in regulation of hepatic lipid and glucose metabolism. CMA regulates hepatic glycolysis by degradation of glycolytic enzymes and mobilization of lipid stores by degrading lipid droplet proteins, thus explaining why CMA-deficient mice display marked hepatosteatosis upon dietary challenges (10, 27). Since C646 showed consistent results in both our in vitro and in vivo studies, we decided to treat the HCD-challenged mice with C646 for the last 3 wk of the dietary challenge (Fig. 5A). Hematoxylin/eosin staining of these mouse livers revealed features of both micro- and macrovesicular steatosis in wild-type mice that were markedly more discrete in the C646 treated mice (Fig. 5 B and C). We confirmed the lipid nature of these vacuoles using staining with BODIPY493/503 which demonstrated reduction in both the number of lipid droplets per hepatic area and per cell upon treatment with C646 (Fig. 5 D and E). As expected, we observed more pronounced steatotic features (Fig. 5 B and C) and higher number of lipid droplets (Fig. 5 D and E) in ^L2AKO mice, but in this case, treatment with C646 failed to show any improvement in steatosis or lipid droplets compared to the untreated mice (Fig. 5 B–E). These results suggest that the protection against hepatosteatosis provided by the CRM was dependent on having functional CMA activity.

Fig. 5.

Beneficial effects of CRM are CMA dependent. (A) Scheme of the interventions in 18-mo-old wild-type (WT) and LAMP2AKO (^L2AKO) male mice. Briefly, equal animal groups were fed AL or a high carbohydrate diet (HCD) for 4 wk with or without CRM (C646) treatment in the last 3 wk. (B and C) Representative hematoxylin and eosin (H&E) staining images of livers of mice in A (B) and quantification of fat vacuoles (C) as percent area of the tissue. n = 7 to 8 mice per group. (D and E) Representative images of BODIPY 493/503 staining [to stain lipid droplets (LD)] (D) and quantification of LD (E) as the number of LD per area (Top) and per cell (Bottom) for the three dietary groups in livers from WT and ^L2AKO mice. n = 7 to 8 mice per group. (F and G) H&E staining (F) and quantification of mean cell area (G) and size distribution of eWAT cells (H) in WT (Left) and ^L2AKO (Right) mice. n = 7 to 8 mice per group. The two-way ANOVA test was used. Values are means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001, ns: not significant.

Trichrome staining of the livers also revealed a significant increase in fibrosis due to HCD, which was not as severe when wild-type mice were treated with CRM; however, ^L2AKO mice did not show any signs of improvements when treated with CRM (SI Appendix, Fig. S5A). Liver weights were comparable among all the cohorts of mice (SI Appendix, Fig. S5B). Despite the fact that both groups of wild-type and ^L2AKO mice on HCD had similar daily food intake (SI Appendix, Fig. S5C), wild-type mice that received CRM treatment maintained their body weight similar to AL diet mice, while untreated mice displayed significant gain in weight (SI Appendix, Fig. S5D). ^L2AKO mice displayed only a trend toward weight gain upon HCD that was not corrected by CRM treatment (SI Appendix, Fig. S5D). The HCD gain in body weight was, as expected, mostly due to an increase in fat mass (SI Appendix, Fig. S5E) also confirmed by the higher weight of both epididymal (eWAT) and subcutaneous (sWAT) white adipose tissue depots (SI Appendix, Fig. S5F). A decrease in fat mass and weight of the fat depots was observed in C646-treated wild-type mice, which was not observed in the C646-treated ^L2AKO cohort (SI Appendix, Fig. S5 E and F). Histologically, we observed that the increase in the size of the adipose cells in wild-type mice subjected to HCD was clearly prevented with CRM treatment; however, ^L2AKO mice did not show any improvement in adipocyte size with CRM treatment (Fig. 5 F and G). Furthermore, we also noticed significantly higher proportion of smaller cells in eWAT of wild-type HCD mice that were treated with CRM, which was not observed in ^L2AKO mice (Fig. 5H). The results of the experiments with wild-type mice are consistent with the previously described protective effects of CRM, whereas the failure to improve with CRM treatment observed in the ^L2AKO mice supports that the beneficial effects of CRM depend on CMA.

Discussion

Our work emphasizes the activating effect of CR and CRMs on CMA as well as the importance of CMA in achieving the beneficial effects of these interventions in aged mouse livers.

Taking advantage of in vitro systems with isolated lysosomes and fluorescent reporters to track CMA activity, in this work, we show that CR leads to constitutive upregulation of CMA in vivo and in vitro, that sustained CR since early in life prevents the age-dependent decline in CMA activity and that acute interventions with CRMs in old mice are effective in restoring normal levels of CMA in multiple tissues (Fig. 6). The findings with the CRMs are interesting because of their potential value as gerotherapeutic interventions. Adaptation to strict dietary limitations is difficult for many individuals, and use of CRMs is an attractive alternative strategy (41). We show that all the CRMs analyzed in this study have a stimulatory effect on basal CMA and are also capable to restore CMA activity in old organisms albeit with different efficiency. The basis of these differences in efficiency may be a consequence of their different mechanism of action and/or bioavailability properties in vivo. Future studies are required to investigate whether the stabilizing effect of CR on LAMP2A at the lysosomal membrane identified in this work in the liver is also behind the restoration of CMA observed with the CRMs in other tissues and whether longer periods of treatment may be needed to observe significant increases in tissues in which LAMP2A stability could be more severely compromised. Last, it is also possible that CRMs could have tissue-specific effects on different components of the CMA machinery.

Fig. 6.

Effect of CR and CRMs on CMA. CMA decreases with age due to a reduction in levels of LAMP2A at the lysosomal membrane. In this work, we show that: 1) CR in rats from 4 mo of age prevents the age-dependent decline in CMA activity in the liver (Top); 2) CRMs administered acutely to 18-mo-old mice induce upregulation of CMA in multiple organs (Bottom); 3) CMA is required for the protective effect of CRMs against a dietary challenge (diet rich in carbohydrates).

In our studies, CR showed an additive effect to the upregulation of CMA in response to nutrient removal, in support of the idea that CR and starvation may regulate CMA through different signaling mechanisms. In serum-deprived cultured cells, CMA upregulation may occur due to the elimination of potentially CMA-inhibitory factors present in the serum. Although serum from CR animals may also have reduced levels of some of those factors, recent studies have identified complex changes in serum metabolites associated with CR (61). It is thus possible that some of the increasing metabolites, whose changes will not be recapitulated by mere serum removal, are responsible for the observed CMA upregulation. In fact, increased LAMP2A levels have been described in the presence of ketone bodies (62, 63), which should be elevated in CR interventions and consequently could contribute to the increase in CMA observed in this study.

As a step forward to understand the possible contribution of the observed constitutive activation of CMA by CR and CRMs on the beneficial effects of these interventions, we show here the need for fully functional CMA in the protective effects of CRMs against a high carbohydrate challenge. Treatment with CRM (C646) shows multiple evidence of protection against this challenge in wild-type mice but not in CMA-incompetent mice. These findings highlight the previously reported importance of CMA in hepatic lipid metabolism by modulating both lipophagy and cytosolic lipolysis of LD (10). Hepatocyte-specific L2AKO mice are unable to mobilize liver lipids efficiently, and additionally they have reduced degradation of lipogenic enzymes which makes these animals more prone to develop steatosis (10, 12). C646 is a histone acetyltransferase inhibitor with more selectivity for CBP/EP300 (53), and in our experiment, it was shown to be the most effective in increasing CMA activity in aged mice. CBP/EP300 can interact with and acetylate many transcription factors including SREBP1, which is important for fatty acid and cholesterol biosynthesis, and increase their activity (53, 64). It is possible that C646 indirectly inhibits proteins such as SREBP1 by enzymatically inhibiting CBP/EP300 and that this effect is further enhanced by the increase in CMA activity, which aids in removal of perilipins and subsequently lipolysis. However, it is also attractive to think that upregulation of CMA may contribute directly to reduce CBP/EP300 levels as a recent study has demonstrated that EP300 is amenable to degradation by CMA at least in macrophages (65). Further research is necessary for better understanding and identification of targets of C646 and other CRMs and how they could potentially regulate CMA and other pathways. In the case of macroautophagy, acetylation has emerged as an important regulatory mechanism, thus making possible a direct effect of CRMs on acetylases that modulate essential macroautophagy components (40). The role of acetylation in the regulation of CMA has been less explored. To date, only histone deacetylase 6 (HDAC6) has proven to be stimulatory on CMA through deacetylation of HSP90, a chaperone required to stabilize binding of CMA substrates to the LAMP2A receptor (66) or by preventing acetylation of CMA substrates and facilitate their recognition by HSC70 (67). Enhanced acetylation of CMA substrates such as tau has also shown to exert an inhibitory effect on CMA through direct “clogging” of the CMA translocation complex (68). Interestingly, substrate acetylation can instead promote degradation of some proteins through CMA by completing a CMA targeting motif missing the mandatory glutamine residue [since acetylation of a lysine makes it behave as a glutamine and to be recognized by HSC70 (2)]. Future studies are needed to discriminate whether changes in acetylation are an additional mechanism behind the observed upregulation of CMA by CRMs or whether they indeed contribute to the higher stability of LAMP2A at the lysosomal membrane under these conditions.

Recent studies have shown that although intermittent fasting and CR can stimulate adaptative macroautophagy and increase longevity, in certain instances, prolonged CR can stimulate macroautophagic cell death (69). Although the molecular mechanism behind this switch from adaptative to detrimental autophagy remains unclear, there seems to be a relation with the underlying energetic status of the cell. Contrary to macroautophagy where sustained activation can lead to depletion of organelles such as mitochondria and the subsequent negative impact on the cellular energetic balance, in the case of CMA, only proteins primed for degradation upon identification by HSC70 can be targeted to this pathway. It is possible that this selective recognition of cargo prevents unwanted degradation of other proteins thus making CMA upregulation safe. We propose that interventions such as CR and CRMs when implemented in conditions with reduced lysosomal capacity for binding and internalization of substrates by CMA such as aging are beneficial because by restoring lysosomal functionality they allow for degradation of the backlog of proteins already primed for degradation by the chaperone. In this respect, we do not anticipate negative effects of sustained upregulation of CMA.

Limitations of the current work include the exclusive focus on male rodents. Recent work has shown a sex- and strain-specific beneficial effect of CR on mice, wherein the positive effect of CR on healthspan requires adjustment of the extent of CR (70). Although we have only used one mouse strain, the fact that the same induction in CMA activity could be observed in rats points toward this effect being likely applicable to multiple mammalian species. However, future studies are needed to compare changes in CMA in response to CR in different mouse strains and in females. Intratissue heterogeneity in CMA response to CRM, as the one observed here in the pancreas, needs to be explored to determine whether it originates from cell-type specific differences or differences in CRM biodistribution inside the organ. The requirement for CMA for the protective effect of CRM is only shown in an acute context in this work. The clear dependence on CMA for that effect provides now rationale for future long-term studies addressing whether extension of life- and healthspan mediated by CR and CRMs is also dependent on functional CMA. This is important because, despite the almost complete ablation of response to CRM in the CMA-deficient mice observed in this work, we anticipate that when performing similar epistasis experiments in the context of longevity, multiple hallmarks of aging may come into play. It will be interesting then to determine what fraction of the beneficial effect in longevity is due to CMA upregulation. One limitation to currently perform these studies is that animals with systemic blockage of CMA show shorter lifespan and degenerative and metabolic derangements making it difficult to extrapolate conclusions from this model when CRM interventions are imposed at advanced ages. We are currently generating a mouse model with inducible knock-out of LAMP2A that will help overcome this limitation.

Overall, our findings demonstrate the efficacy of CR in preventing the age-dependent decline in CMA and of CRMs in restoring CMA activity in old organisms. Considering the previously demonstrated beneficial effect of genetically preventing the decline in LAMP2A with age on liver proteostasis and function (24), the protective effect against atherosclerosis of the same genetic intervention at the systemic level (13) and the recently demonstrated protective effect of small molecule activators of CMA in mouse models of neuronal and retinal degeneration (14, 15), we propose that the increase in CMA activity upon CR and CRMs could also be behind some of the similar beneficial effect on those systems and disease models shown for these interventions.

Materials and Methods

Animals, Cells, and Reagents.

Male Fisher-344 rats of ages 4, 12, and 22 mo were obtained from the NIA aged rodent colonies maintained on a sterilized Purina Mills 5L79 diet (similar to the NIH31M) and subjected to a 40% reduction in calories from 4 mo of age. CRMs studies were performed in wild-type C57BL/6J male mice (JAX 000664), KFERQ-Dendra2 reporter mice (48), and LAMP2A knock-out (^L2AKO) mice (14). The HCD contained 70% carbohydrates, 20% protein, 10% fat (D12450J, Research Diets). NIH3T3 cells control or stably expressing the CMA reporter KFERQ-Dendra2 (47) were used for the in vitro CR system with serum from AL or CR rats and for the CRM treatments. All reagents used in this study are detailed in SI Appendix, Materials and Methods and Tables S1 and S2. Primers used for quantitative PCR analysis are listed in SI Appendix, Table S3.

CMA Assays.

CMA was analyzed using the following assays: 1) CMA reporter (KFERQ-PS-Dendra2) in cultured cells; 2) in vitro incubation of CMA substrates with lysosomes isolated from liver through differential centrifugation and flotation in discontinuous density gradients (7) and quantification of the amount of substrate degraded (43, 49), bound to the membrane and internalized into the lysosomal lumen (49); and 3) analysis of number of fluorescence Dendra puncta colocalizing with LAMP1 in tissues from KFERQ-Dendra mice (48). Degradation of LAMP2A was analyzed by incubating intact lysosomes or isolated lysosomal membranes and blotting for LAMP2A at different incubation times. Lysosomal membrane microdomains were isolated as described previously (50).

Cells and Tissue Imaging.

Direct fluorescence images were acquired with a high-content microscope; immunofluorescence images in antibody-stained cells were acquired with an Axiovert 200 fluorescence microscope (Carl Zeiss Microscopy). Tissues were cut in 40 μm sections using Leica VT 1000S vibratome, and after stained free floating, they were imaged using the Leica TCS SP8 (LeicaMicrosystem) microscope with a 40× objective for antibody-stained tissues or a Leica light microscope DMD108 for H&E stained sections. Images were quantified after thresholding using ImageJ.

Sample Size and Statistical Analysis.

All data presented are mean ± SEM. Prior to statistical testing, normality was assessed using the Shapiro–Wilk test. We used the two-tailed t test for comparison between two groups, one-way ANOVA for comparisons with more than two groups, and two-way ANOVA for comparison of different conditions in several groups. We used power analysis to calculate the number of animals per experiment and condition based on our previous results with dietary challenges, the KFERQ-Dendra mouse reporter, and with chemical activators of CMA. Statistical analyses were performed in GraphPad Prism 10.0. All quantifications were done blind to the experimentalist and only unblinded for grouping and statistics. No animals were discarded from the analysis.

Supplementary Material

Appendix 01 (PDF)

Appendix 02 (PDF)

Dataset S01 (XLSX)

Data, Materials, and Software Availability

All reagents generated in this study are available from the lead contact with a completed material transfer agreement. There are no restrictions on data availability in this manuscript. All raw data shown in main and supplementary figures is included in the manuscript and/or provided as an excel worksheet under supporting information organized by figures, and it includes statistics along with exact P values. This manuscript does not report original code.