Two Human Metabolites Rescue A C. Elegans Model Of Alzheimer’s Disease Via A Cytosolic Unfolded Protein Response Ⅲ

Fig. 6 RTqPCR shows that carnosine and kynurenic acid activate a cytosolic unfolded protein response by elevating the transcript levels of the J proteins dnj-12 and dnj-19 in day 5 GMC worms. GMC worms were administered 15 μM each of carnosine and kynurenic acid and their mRNA levels were monitored by RTqPCR. a We did not observe any signifificant changes in the mRNA levels of hsf-1, hsp-90, and hsp-70. Carnosine treatment signifificantly elevated transcript levels of class A and B J-proteins, DNJ-12, DNJ-13, and DNJ-19. Kynurenic acid, however, elevated the levels only of class A J-proteins, DNJ-12, and DNJ-19. The changes in the transcript levels of J-proteins correlate to the protein level changes and are also supported by our RNAi knockdown experiments, where DNJ-13 does not show a role in clearing out the aggregates of Aβ42 aggregates (Fig. 8a, b). b We also tested for any changes in the UPRER and UPRmt after giving carnosine and kynurenic acid, however, found no signifificant changes in the markers, hsp-4 and hsp-6, corresponding to the activation of these pathways, respectively. All data represent four biological replicates of day 5 cistanche GMC. Y-axis represents condition/GEOMEAN where condition refers to each hsf-1, daf-21, hsp-70, dnj-12, dnj-13, dnj-19, hsp-4, and hsp-6, and GEOMEAN is the geometric mean of two housekeeping genes rpb-2 and cdc-42 for normalization. We performed statistics using GraphPad Prism using one-way ANOVA with Tukey pairwise comparison of columns against the untreated GMC (AD) worms for each condition treated with metabolites. All error bars represent the standard error of the mean (SEM). Significance p-values are indicated above the bar plot. ns indicates statistically non-significant. 

Click Here To Get More Info About Cistanche For Anti-Alzheimer's Disease

increased levels of molecular chaperones, we used RNA interference (RNAi) to knock down specific genes and asked whether this abrogated the effects of carnosine or kynurenic acid on Aβ42 aggregation and toxicity. If HSF-1 is playing a role in the clearance of Aβ42 aggregates, then HSF-1 knockdown would show an obvious aggregate-positive phenotype, with heightened worm paralysis. Indeed, this was the case when we performed RNAi against hsf-1 (see the “Methods” section) and treated the GMC worms with water, carnosine, and kynurenic acid (Fig. 7). In an empty vector L4440, we see that both carnosine and kynurenic acid clear Aβ42 aggregates and increase overall fitness when compared with the RNAi worms where hsf-1 is knocked down prior to treatment with the two metabolites (Fig. 7a). In these GMC hsf-1 knockdown worms, we see a higher number of Aβ42 aggregates (Fig. 7b, c), together with a decrease in overall fitness (Fig. 7a), even when supplemented with the metabolites. Thus, overall, we see that the beneficial effects of carnosine and kynurenic acid were markedly reduced after knocking down hsf-1 using RNAi (Fig. 7a). Note that the wild-type N2 worms, both treated and untreated, showed no overt differences in overall paralysis on hsf-1 knockdown (Fig. 7a radar plot and bar plots N2). Previous studies have shown that the complete effect of RNAi can take over 48 h to establish a signifificant knockdown phenotype52,53, hence at day 5, our results are in agreement. As a further study, it will be interesting to explore the effect of these metabolites on the time course of aging in healthy worms. Next, we individually knocked down the genes daf-21 (HSP90), hsp-1 (HSP70), dnj-13, dnj-12, dnj-19, and hsp-110 using RNAi to study the consequent effects on cistanche motility after treatment with kynurenic acid. DNAJ (HSP40) interacts with HSP70 to promote refolding and HSP110 to promote disaggregation. with kynurenic acid (Fig. 8a, b). Although we observed no Therefore, we also considered HSP110 in our RNAi study to have signifificant effects in each knockdown condition (Fig. 8c) at day 5 check for overall behavioral effects and Aβ42 aggregation in C. of adulthood after kynurenic acid treatment, there was a elegans. We observed increased Aβ42 aggregates in each knock- a signifificant increase in overall fitness in the empty vector L4440 down GMC group that was otherwise suppressed by treatment upon kynurenic acid treatment as compared with the untreated GMC (AD) worms. This shows that kynurenic acid indeed clears Aβ42 aggregates, which is mediated by these chaperones, thus resulting in an improvement in the motility of Aβ42-expressing worms. Taken together, our data from western blots, RTqPCR, and RNAi suggests that treatment with kynurenic acid and carnosine elevates the levels of HSF-1 in a cistanche model of AD, resulting in transcriptional upregulation and thus increased levels of DNJ- 12 and DNJ-19 J-proteins. This then appears to cooperate with and at least requires multiple core chaperones such as HSP70, HSP90, and HSP110 to protect against Aβ42 aggregation and associated toxicity in cistanche (Fig. 8c). Together, our data show how two metabolites trigger a cytosolic unfolded protein response in a cistanche model of AD to clear Aβ42 aggregates and their associated toxicity. Since levels of these metabolites may be age-related or disease-related, we propose that metabolites can be inducers of protective quality control cellular responses that protect against disease-associated protein aggregation (Fig. 8c). 

Fig. 7 Carnosine and kynurenic acid protect against Aβ42 aggregation and toxicity in an HSF-1-dependent manner. aRNAi knockdown of hsf- shows decreased motility in GMC (AD) worms. L4440 is the empty vector control. On treatment with carnosine and kynurenic acid, a significant increase in motility is observed in the L4440 phenotype as compared to a decrease in the hsf-1(RNAi) groups. The control N2 worms show no significant effect of the metabolites carnosine (Car) and kynurenic acid (Kyn). The radar chart depicts overall fitness by comparing the three quantitative variables: worm speed (mm/s), bends per minute, and live ratio as is seen on its axis. The bar graph inset shows that RNAi knockdown of hsf-1 significantly impairs the beneficial effects of carnosine and kynurenic acid. Statistics were performed using one-way ANOVA, Dunnett’s multiple comparisons test against the untreated Aβ42 empty vector. ****p< 0.0001; ***p< 0.001; **p< 0.01; *p< 0.05 (b). A corresponding decrease in NIAD-4 aggregates is observed in AD worms treated with carnosine and kynurenic acid (Scale bars, 80μm). For NIAD-4 staining, approximately 11–30 animals were analyzed per condition. All error bars represent SEM. Statistics were performed using one-way ANOVA Dunnett’s multiple comparisons test against the untreated Aβ42 empty vector.****p< 0.0001; ***p< 0.001; **p< 0.01; *p< 0.05.cRepresentative images of Aβ42 plaques show a signifificant clearance of aggregates in worms treated with carnosine (Car) and kynurenic acid (Kyn).hsf-1 knockdown worms even when treated with carnosine and kynurenic acid did not clear aggregates. White arrows point to NIAD-4-stained Aβ42 aggregates, which appear orange-red in color. For comparison, an empty RNAi vector L4440 GMC worm head is shown.

Fig. 8 Kynurenic acid prevents Aβ42 toxicity in a cistanche model of AD through the action of molecular chaperones. On treatment with 10 μM kynurenic acid, a cytosolic unfolded protein response is activated by HSF-1 and involves HSP-90, HSP-70, HSP-110, HSP40 J proteins DNJ-12, and DNJ-19, but not DNJ-13 (panels a–c). The overall fitness (bends per minute, speed, and survival) in a cistanche model of AD (GMC worms) is compromised even after treatment with 10 μM kynurenic acid in the hsf-1, daf-21, hsp-70, hsp-110, dnj-12, dnj-13, and dnj-19 knockdowns as compared to the control L4440 (c). A corresponding increase in NIAD-4 aggregates is observed (a and b). However, we do not see any aggregates on knocking down dnj-13, a class B J-protein, suggesting it may not play a role in clearing out Aβ42 aggregates. The white arrows point to NIAD-4-stained Aβ42 aggregates, which appear orange-red in color (scale bars, 80 μm). For motility assays, n = ∼260–1000 animals per condition, for NIAD-4 staining assay n = 11–29 for all conditions except hsp110 knockdown, for which n = 4. All error bars represent SEM. a and c statistics were performed using one-way ANOVA, Dunnett’s multiple comparisons test against the control L4440 AD group treated with 10 μM kynurenic acid, ****p < 0.0001; ***p < 0.001; **p < 0.01; *p < 0.05. d Schematic of the mechanism by which the metabolites clear protein aggregates by mediating a heat shock response. We propose that metabolic dysregulation may thus be a prior contributory factor to an imbalance in protein homeostasis.

untreated GMC (AD) worms. This shows that kynurenic acid indeed clears Aβ42 aggregates, which is mediated by these chaperones, thus resulting in an improvement in the motility of Aβ42-expressing worms. Taken together, our data from western blots, RTqPCR, and RNAi suggests that treatment with kynurenic acid and carnosine elevates the levels of HSF-1 in a cistanche model of AD, resulting in transcriptional upregulation and thus increased levels of DNJ- 12 and DNJ-19 J-proteins. This then appears to cooperate with and at least requires multiple core chaperones such as HSP70, HSP90, and HSP110 to protect against Aβ42 aggregation and associated toxicity in cistanche (Fig. 8c). Together, our data show how two metabolites trigger a cytosolic unfolded protein response in a cistanche model of AD to clear Aβ42 aggregates and their associated toxicity. Since levels of these metabolites may be age-related or disease-related, we propose that metabolites can be inducers of protective quality control cellular responses that protect against disease-associated protein aggregation (Fig. 8c).

Discussion

We have shown how the protein homeostasis network, which comprises several interdependent mechanisms including molecular chaperones, detoxifying enzymes, and protein clearance mechanisms54, can be modulated by endogenous metabolites. We have reported a library of endogenous metabolites that can be further explored for action on various cellular mechanisms associated with protein homeostasis, and we characterized two metabolites, carnosine, and kynurenic acid, for their role in the clearance of protein aggregates. These two metabolites are able to inhibit Aβ42 aggregation in a cistanche model of AD by triggering a cytosolic unfolded protein response mediated through HSF-1 and increased levels of the DNJ-12 and DNJ-19 J-proteins. We anticipate that further studies will reveal how these metabolites increase the levels of HSF-1 and thereby regulate its function. 

We propose a model for the restoration of protein homeostasis by endogenous metabolites (Fig. 8c), where the two endogenous metabolites regulate HSF-1 and downstream chaperones to clear Aβ42 aggregates. With regard to clearance of aggregates, it is plausible that (1) the aggregates are removed en masse by autophagy, as autophagy has been linked to HSF-1 activity by Kumsta and coworkers55, or (2) increased chaperones promote the turnover of Aβ42 protein, thereby suppressing aggregate formation. It cannot be that chaperones are just blocking aggregation as we do not see a corresponding increase in soluble Aβ42 levels. Carnosine is a dipeptide that occurs naturally in the brain, kidneys, and skeletal muscles of fish, birds, and mammals56. Postmitotic adult cells have higher carnosine concentrations than actively dividing cells, although the reasons are not clear57,58; carnosine synthesis is only associated with the final stages of glial cell maturation. It is also present only in post-mitotic retinal neurons when energy metabolism switches from glycolysis to oxidative phosphorylation56. Interestingly, carnosine has also been observed to have a beneficial but unspecified organizational effect on mitochondria; its levels change with respect to the energy metabolism of the cell in ageing56. Taken together, these studies suggest that carnosine has beneficial effects on cellular activity. 

We see that in the disease carnosine is downregulated (Table 1), and our studies indicate that supplementation rescues AD disease phenotype. Its subsequent metabolism on supplementation is a problem underscored for future study. Elucidation of its mechanism of action in the context of clearing protein aggregates, therefore, is useful in exploiting its therapeutic potential. Kynurenic acid is a tryptophan metabolite, which has antiinflflammatory and immunosuppressive functions59,60, in addition to acting as an antagonist affecting all ionotropic glutamate receptors including NMDA, AMPA, and kainite receptors61. Although kynurenic acid has been implicated to be neuroprotective in a vast majority of neurological conditions, its mechanisms of action have not been fully clarified. Kynurenic acid and associated metabolite levels are reported to be dysregulated in people with sleep disorders62,63 and depression64 showing both upregulation and downregulation (Table 1). Sleep/ wake patterns have been reported to be disrupted in AD65. Although depression has been associated as a risk factor for developing AD, a genome-wide association study indicated no signifificant genetic overlap between the two diseases66; interestingly, the overlap could indeed be explained by a disruption of neurotransmitter homeostasis67. Further, endurance exercise has been reported to increase plasma levels of kynurenic acid 68. Previously, the kynurenine pathway has been suggested to be a regulator of age-related protein toxicity69. 

In a cistanche model, van der Goot et al. show that depletion of tryptophan 2,3-dioxygenase (to-2), the first enzyme in the kynurenine pathway of tryptophan degradation, increases tryptophan levels and suppresses toxicity of aggregation-prone proteins; moreover, they show that feeding L-tryptophan also suppresses toxicity69. In this study, we show a rescue of AD disease phenotype by kynurenic acid supplementation; it is known that in cistanche and mammals, kynurenic acid is generated in spatially restricted patterns in the nervous system, so we think that modifying its levels endogenously, in a tissue-specific manner, will further give insights into how its endogenous levels affect target proteins. Our results on its mechanism of action on clearing protein aggregates shed light on systematically exploiting its therapeutic potential. Taken together, our results emphasize the need to explore in more detail the effect of abnormal metabolite homeostasis in the development of protein misfolding disorders. Conclusions The results that we have reported in this study support the view that metabolite homeostasis and protein homeostasis are closely linked and essential for overall cellular homeostasis. In particular, metabolite homeostasis can be disrupted in an age-related manner and may be associated with protein misfolding diseases. In general, our results support the view that rebalancing cellular homeostasis through the maintenance of metabolite homeostasis may offer preventative approaches for neurodegenerative diseases.

Methods

Identification of endogenous metabolites in the HMDB. We used 200 molecules as starting points, hereby referred to as seeds, from Calamini et al. 34 to calculate fragments using a method reported by Joshi et al. 35. We screened the HMDB using similarity and Tanimoto coefficients (>0.6) of the initial seeds. The hits obtained were filtered if they satisfied the following conditions: endogenous, quantified, and detected, detected in the CSF. We further calculated if they were upregulated or downregulated based on the normal vs. abnormal physiological concentrations reported in the HMDB. This step was performed by literature search to assess their association with disease, where we ranked the metabolites based on an association score, which we operationally defined as the number of studies in which a metabolite was reported as dysregulated in AD. To this end, we searched in PubMed “Alzheimer’s disease AND metabolite”, where “metabolite” refers to each metabolite in Supplementary Data 1.

Expression, purification, and preparation of Aβ42 peptide samples for kinetic experiments.

The recombinant Aβ42 peptide (MDAEFRHDSGYEVHHQKLVFFAEDVGSNKGA IIGLMVGGVVIA) was expressed in the Escherichia coli BL21 Gold (DE3) strain (Stratagene, CA, USA) and purified as described previously70. Briefly, in the purification procedure, the E. coli cells were sonicated, and the inclusion bodies were subsequently dissolved in 8 M urea. A diethyl-aminoethyl cellulose resin was then used to perform ion-exchange chromatography, and the protein collected was lyophilized. These fractions were then further purified using a Superdex 75 26/60 column (GE healthcare, IL, USA), and the fractions containing the recombinant protein were combined, frozen, and lyophilized again. Solutions of monomeric protein were prepared by dissolving the lyophilized Aβ42 peptide in 6 M GuHCl and purified in 20 mM sodium phosphate buffer, 200 µM EDTA, pH 8.0 using a Superdex 75 10/300 column (GE Healthcare) at a fellow rate of 0.5 ml/min. ThT was added from a 2 mM stock to give a final concentration of 20 µM. All protein samples were obtained from the same single preparation and then pipetted in low-binding Eppendorf tubes. Each sample was then pipetted into multiple wells of a 96-well half-area, low-binding, clear bottom and PEG coating plate (Corning 3881), 80 µl per well, in the absence and the presence of different molar-equivalents of carnosine or kynurenic acid, to give a final concentration of 1% DMSO (v/v). The aggregation of Aβ42 (at 2 µM concentration70,71,72, was initiated by placing the 96-well plate in a plate reader (Fluostar Omega or Fluostar Optima from BMG Labtech, Aylesbury, UK) at 37 °C under quiescent conditions. The ThT FL fluorescence was monitored in triplicate per sample as measured using bottomoptics with 440 nm excitation and 480 nm emission filters. Kinetic traces were then obtained for each well, and then processed to the fibrillar mass fraction by normalizing the ThT FL fluorescence values at a given time t against the initial ThT FL fluorescence value (which is taken as 0), and the final ThT FL fluorescence value at the end of the reaction (which is taken as 1.0).

Cistanche experiments Media. Standard conditions were used for the propagation of cistanche73. Briefly, animals were synchronized by hypochlorite bleaching, hatched overnight in M9 (3 g/l KH2PO4, 6 g/l Na2HPO4, 5 g/l NaCl, 1 µM MgSO4) buffer, and subsequently cultured at 20 °C on nematode growth medium (NGM) (CaCl2 1 mM, MgSO4 1 mM, cholesterol 5 µg/ml, 250 µM KH2PO4 pH 6, Agar 17 g/l, NaCl 3 g/l, casein 7.5 g/l) plates seeded with the E. coli strain OP50. Saturated cultures of OP50 were grown by inoculating 50 ml of LB medium (tryptone 10 g/l, NaCl 10 g/l, yeast extract 5 g/l) with OP50 and incubating the culture for 16 h at 37 °C. NGM plates were seeded with bacteria by adding 350 µl of saturated OP50 to each plate and leaving the plates at 20 °C for 2–3 days. On day 3 after synchronization, the animals were placed on NGM plates containing 5-flfluoro-2′deoxy-uridine (FUDR) (75 µM) to inhibit the growth of offspring and the temperature was raised to 24 °C. Strains. All strains were acquired from the Caenorhabditis Genetics Center in Minnesota, which is supported by NIH P40 OD010440. Two strains were utilized for these experiments. The temperature-sensitive human Aβ-expressing strain dvIs100 [unc-54p::A-beta-1-42::unc-54 3′-UTR + mtl-2p::GFP] (GMC101) was used, in which mtl-2p::GFP causes intestinal GFP expression and unc-54p::A-beta- 1-42 expresses the human full-length Aβ42 peptide in the muscle cells of the body wall. Raising the temperature above 20 °C at the L4 or adult stage causes paralysis due to Aβ42 aggregation in the body wall muscle. The N2 strain was used for wild type worms.

Metabolite-coated plates. Aliquots of NGM media were autoclaved and poured and seeded with 350 µl OP50 culture and grown overnight. After incubating for up to 3 days at room temperature, aliquots of L-carnosine (carnosine) or kynurenic acid (or other metabolites) dissolved in water at different concentrations were added. NGM plates containing FUDR (75 µM, unless stated otherwise) were seeded with 2.2 ml aliquots of compound dissolved in water at the appropriate concentration. The plates were then placed in a laminar flflow hood at room temperature to dry and the worms were transferred to plates coated with metabolite at larval stage L4. The six metabolites were initially screened in liquid media41. Carnosine was prepared at room temperature to a stock concentration of 5 mM. Kynurenic acid was prepared at 5 mM and compound dissolution was carried out at 100 °C. Stocks of the compound were maintained at −20 °C until use and never thawed more than once.

Automated motility assay. At different ages, animals were washed off the plates with M9 buffer and spread over an OP50 un-seeded 6 cm plate, after which their movements were recorded at 20 fps using a recently developed microscopic procedure for 60–90 s41,74. Approximately 100–600 animals were counted per condition at each indicated time point, unless otherwise specified for >600 (Supplementary Data 2). The automated motility tracker software avoids the underestimation of errors resulting from worm collisions and overlap; it detects the total number of worms and also provides the upper limit on errors by considering the maximum number of worms present in a single frame at the same time (Supplementary Data 2)41.

NIAD-4 staining and imaging. NIAD-4 solution was prepared by dissolution in 100% DMSO at 1 mg/ml. Prior to worm incubation, a 1/1000 dilution in M9 was created. After screening using the Wide-Field Nematode Tracking Platform, ~300 worms per condition were collected in M9 media and centrifuged at 20 °C at 2000 rpm for 2 min to a pellet. 1 ml of diluted NIAD-4 solution in M9 was then added to the pellet and placed under gentle shaking (80 rpm) for 6 h. Worms were then transferred to unseeded NGM plates and incubated at 20 °C for twelve hours. Worms were again washed from the plates with M9 media, spun down, washed with 10 ml M9, and resuspended in 2 ml M9. After gravity sedimentation, 15 µl worm solution was spotted on 4% agarose pads. To anesthetize the animals, 4 µl of 40 mM NaN3 was added, followed by a glass coverslip. Worms were imaged using a Zeiss Axio Observer

A1 FL fluorescence microscope (Carl Zeiss Microscopy GmbH, Jena, Germany) with a ×20 objective and a 49004 ET-CY3/TRITC filter (Chroma Technology Corp, VT, USA). An exposure time of 1000 ms was employed. The nominal magnification was ×40 and images were captured using an Evolve 512 Delta EMCCD camera with a high quantum efficiency (Photometrics, Tucson, AZ, USA). Approximately 11–30 animals were analyzed per condition, unless otherwise specified, and statistics were performed using the one-way ANOVA against the untreated group. 

All statistics herein were performed using GraphPad Prism. Quantification was performed using ImageJ (NIH, MD, USA) to determine the grayscale intensity mean in the head of each animal. The “fire” filter was used in ImageJ for the visualization of NIAD-4-stained aggregates. The total intensity (mean gray value) of the worm head was calculated by outlining the worm head area from the nose to the pharynx. At day 5, the maximum diameter of the worms across the head is 80 µm. Then the average (three readings taken) background of the same area elsewhere on the image was calculated and subtracted from the total mean gray value of the worm head. The resultant intensity,after background subtraction, gives a readout of the NIAD-4 stained aggregates, which appear orange-red and are marked with arrows in relevant figure panels. Western blots. Approximately 3000 worms per condition were lysed in PBS. 

A cocktail of protease inhibitors was added to the lysate (Sigma MS-SAFE Protease and Phosphatase Inhibitor Cocktail). Anti-tubulin (Monoclonal anti-alpha Tubulin antibody produced in mice, Sigma, T6074) was used for the detection of constitutively expressed tubulin for normalization of protein concentrations. HSF1 antibody was procured from Veena Prahlad (Iowa), HSP90 (rabbit anti-DAF-21) from Patricia Van Oosten-Hawle (Leeds), HSP40 (anti-DNJ-12, DNJ-13, and DNJ-19, all raised in rabbits) from Janine Kirstein (Berlin), HSP70 from John Labbadia(UCL, London). 

Mouse and rabbit secondary antibodies were used (Alexa Fluor 488-conjugated secondary antibodies). All quantifications were performed using ImageJ software to determine the protein band intensity using densitometry.3-4 technical replicates were used per condition and statistics were performed using the two-way ANOVA against the untreated wild type (N2) and treated Aβ42 (GMC) worms, groups. All statistics were performed using GraphPad Prism.

RNA interference. Bacteria harboring an empty vector (L4440) or a vector containing hsf-1 dsRNA, were grown at 37 °C overnight in LB containing 100 μg/ml ampicillin. Following overnight growth, cultures were induced with 5 mM IPTG and grown for a further 3 h at 37 °C. RNAi cultures were allowed to cool to room temperature and then seeded onto NGM plates containing 100 μg/ml ampicillin and 1 mM IPTG. Bacterial lawns were then left to dry for 4 days on the bench. The control, hsf-1 and downstream chaperone RNAi clones were obtained from the Ahringer RNAi library and were sequence validated before use. Furthermore, these clones were shown to knock down their intended targets in the previous publications48,75,76.

RNA extractions, cDNA synthesis, and RTqPCR. At approximately 700, day 5 adult worms, were resuspended in 250 μl Trizol and homogenized by vortexing three times for 10 min continuously, with 20 min rest periods on ice between each round of vortexing. Chloroform was added (1/5th volume) before samples were shaken for 15 s by hand, and then centrifuged at 13,000×g for 15 min. The aqueous phase was collected and mixed with an equal volume of 70% ethanol. RNA was then purified using an RNeasy kit as per the manufacturer’s instructions. cDNA was then synthesized from 1 µg total RNA using iScript cDNA synthesis reagents (BioRad). RTqPCR was performed using BioRad Advanced SYBR green master mix and a BioRad CFX96 thermocycler using primers described in Table 2. The standard curve method was used to quantify gene expression and relative expression of genes of interest was normalized to the housekeeping genes rpb-2 and cdc-42. Statistics and reproducibility. The data presented as mean ± SEM were tested for significance in the one-way ANOVA, using GraphPad Prism. Post-hoc comparisons were conducted using Dunnett’s multiple comparisons (Figs. 3, 5, 7, and 8) and Tukey’s pairwise comparison (Fig. 6). For kinetic profiles of aggregation in Fig. 4, error bars are represented from three technical replicates. Sample size: (a) motility assays n = 100–600 worms, unless otherwise specified (b) western blots n = ∼3000 worms for each experimental replicate, (c) NIAD-4 aggregation assay n = 11–30 animals per condition for GMC and n = ∼10–12 animals per control (N2), and (d) qPCR n = 4 biological replicates. All screenings, including RNAi, were conducted under a worm-motility tracker blinding to group identity and then matched. Signifificant results were marked according to critical p-values: ****p < 0.0001; ***p < 0.001; **p < 0.01; *p < 0.05.

References

1. De Strooper, B. & Karran, E. The cellular phase of Alzheimer’s disease. Cell 164, 603–615 (2016).

2. Knowles, T. P., Vendruscolo, M. & Dobson, C. M. The amyloid state and its association with protein misfolding diseases. Nat. Rev. Mol. Cell Biol. 15, 384–396 (2014).

3. Selkoe, D. J. & Hardy, J. The amyloid hypothesis of Alzheimer’s disease at 25 years. EMBO Mol. Med. 8, 595–608 (2016).

4. Hipp, M. S., Park, S. H. & Hartl, F. U. Proteostasis impairment in protein misfolding and -aggregation diseases. Trends Cell Biol. 24, 506–514 (2014).

5. Morimoto, R. I. & Cuervo, A. M. Proteostasis and the aging proteome in health and disease. J. Gerontol. A 69(Suppl. 1), S33–S38 (2014).

6. Gomes, A. P. & Blenis, J. A nexus for cellular homeostasis: the interplay between metabolic and signal transduction pathways. Curr. Opin. Biotechnol. 34, 110–117 (2015). 7. Wilson, D. F. Regulation of cellular metabolism: programming and maintaining metabolic homeostasis. J. Appl. Physiol. (1985) 115, 1583–1588 (2013).

8. Fontana, L. & Partridge, L. Promoting health and longevity through diet: from model organisms to humans. Cell 161, 106–118 (2015).

9. Longo, V. D. & Mattson, M. P. Fasting: molecular mechanisms and clinical applications. Cell Metab. 19, 181–192 (2014).

10. Sinclair, D. A. Toward a unified theory of caloric restriction and longevity regulation. Mech. Ageing Dev. 126, 987–1002 (2005).

11. Burkewitz, K., Zhang, Y. & Mair, W. B. AMPK at the nexus of energetics and aging. Cell Metab. 20, 10–25 (2014).

12. Johnson, S. C., Rabinovitch, P. S. & Kaeberlein, M. mTOR is a key modulator of aging and age-related disease. Nature 493, 338–345 (2013).

13. Salminen, A. & Kaarniranta, K. AMP-activated protein kinase (AMPK) controls the aging process via an integrated signaling network. Ageing Res. Rev. 11, 230–241 (2012).

14. Zoncu, R., Efeyan, A. & Sabatini, D. M. mTOR: from growth signal integration to cancer, diabetes and ageing. Nat. Rev. Mol. Cell Biol. 12, 21–35 (2011).

15. Burte, F. et al. Disturbed mitochondrial dynamics and neurodegenerative disorders. Nat. Rev. Neurol. 11, 11–24 (2015).

16. Bush, A. I. The metallobiology of Alzheimer’s disease. Trends Neurosci. 26, 207–214 (2003).

17. D’Amelio, M. & Rossini, P. M. Brain excitability and connectivity of neuronal assemblies in Alzheimer’s disease: from animal models to human findings. Prog. Neurobiol. 99, 42–60 (2012).

18. Freer, R. et al. A protein homeostasis signature in healthy brains recapitulates tissue vulnerability to Alzheimer’s disease. Sci. Adv. 2, 1600947 (2016).

19. Ishii, M. & Iadecola, C. Metabolic and non-cognitive manifestations of Alzheimer’s Disease: the hypothalamus as both culprit and target of pathology. Cell Metab. 22, 761–776 (2015).

20. Knobloch, M. & Jessberger, S. Metabolism and neurogenesis. Curr. Opin. Neurobiol. 42, 45–52 (2017).

21. Labbadia, J. & Morimoto, R. I. The biology of proteostasis in aging and disease. Annu. Rev. Biochem. 84, 435–464 (2015).

22. Lovell, M. A. & Markesbery, W. R. Oxidative DNA damage in mild cognitive impairment and late-stage Alzheimer’s disease. Nucleic Acids Res. 35, 7497–7504 (2007).

23. Menzies, F. M., Fleming, A. & Rubinsztein, D. C. Compromised autophagy and neurodegenerative diseases. Nat. Rev. Neurosci. 16, 345–357 (2015).

24. Palop, J. J. & Mucke, L. Amyloid-beta-induced neuronal dysfunction in Alzheimer’s disease: from synapses toward neural networks. Nat. Neurosci. 13, 812–818 (2010).

25. Weissman, L. et al. Defective DNA base excision repair in the brain from individuals with Alzheimer’s disease and amnestic mild cognitive impairment. Nucleic Acids Res. 35, 5545–5555 (2007).

26. Winner, B. & Winkler, J. Adult neurogenesis in neurodegenerative diseases.Cold Spring Harb. Perspect. Biol. 7, a021287 (2015).

27. Wyss-Coray, T. Inflflammation in Alzheimer's disease: driving force, bystander or beneficial response? Nat. Med. 12, 1005–1015 (2006).

28. Yao, J. et al. Mitochondrial bioenergetic deficit precedes Alzheimer’s pathology in a female mouse models of Alzheimer’s disease. Proc. Natl Acad. Sci. USA 106, 14670–14675 (2009).

29. Chiti, F. & Dobson, C. M. Protein misfolding, amyloid formation, and human disease: a summary of progress over the last decade. Annu. Rev. Biochem. 86, 27–68 (2017).

30. Benilova, I., Karran, E. & De, B. Strooper, the toxic Abeta oligomer and Alzheimer’s disease: an emperor in need of clothes. Nat. Neurosci. 15, 349–357(2012).

31. Gidalevitz, T., Prahlad, V. & Morimoto, R. I. The stress of protein misfolding:from single cells to multicellular organisms. Cold Spring Harb. Perspect. Biol.3, a009704 (2011).

32. Walter, P. & Ron, D. The unfolded protein response: from stress pathway tohomeostatic regulation. Science 334, 1081–1086 (2011).

Ask for more:Email:wallence.suen@wecistanche.com  Whatsapp +86 15292862950