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

Age-related changes in cellular metabolism can affect brain homeostasis, creating conditions that are permissive to the onset and progression of neurodegenerative disorders such as Alzheimer’s and Parkinson’s diseases. Although the roles of metabolites have been extensively studied with regard to cellular signaling pathways, their effects on protein aggregation remain relatively unexplored. By computationally analyzing the Human Metabolome Database, we identified two endogenous metabolites, carnosine, and kynurenic acid, that inhibit the aggregation of the amyloid beta peptide (Aβ) and rescue a C. elegans model of Alzheimer’s disease. We found that these metabolites act by triggering a cytosolic unfolded protein response through the transcription factor HSF-1 and downstream chaperones HSP40/Jproteins DNJ-12 and DNJ-19. These results help rationalize previous observations regarding the possible anti-aging benefits of these metabolites by providing a mechanism for their action. Taken together, our findings provide a link between metabolite homeostasis and protein homeostasis, which could inspire preventative interventions against neurodegenerative disorders.

Click Here To Get More Info About How Cistanche Prevents Alzheimer's and Parkinson's diseases

Ask for more:

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

Alzheimer's disease (AD) is a complex disorder characterized by the presence of aberrant protein deposits in brain tissues1–3. Although the molecular origins of this disease are still to be firmly established, it is generally recognized that the presence of protein deposits is associated with the dysregulation of the protein homeostasis network4,5. In turn, this system is part of a wider cellular homeostasis system, which includes a variety of other components, including the metabolite homeostasis system6,7. Metabolites affect key steps in cellular pathways where they act as substrates for enzymes or as signaling molecules for the activation of biochemical pathways. For example, caloric restriction and intermittent fasting retard structural and functional decline during aging in laboratory rodents and monkeys8–10. 

Moreover, signaling pathways have evolved to respond to cellular stresses in aging, with the AMPK and mTOR pathways being widely studied in cellular development and aging11–14. It is therefore important to understand how neurodegenerative disorders associated with aging, such as AD, are linked with aberrations in cellular homeostasis. The specific hypothesis that we investigate here is whether the dysregulation of metabolite levels is associated with the aberrant aggregation of proteins in AD. We approach this by testing if supplementing the dysregulated metabolites may prevent or clear the aggregation of misfolded proteins. The impairment of various biological processes contributes to the onset and progression of AD, including DNA repair, inflflammation, mitochondrial function, oxidative stress, neuronal network excitability, metal ion homeostasis, autophagy, hypothalamic regulation, adult neurogenesis, and protein homeostasis (Fig. 1)1,15–28. Small molecule metabolites play key roles in these biochemical processes, but it is still not known in detail which among these roles are first dysregulated and are associated with the pathophysiological onset and progression of the disease. In this context, whether the phenomenon of protein misfolding and aggregation is a cause or a consequence of the impairment in these biological processes is yet to be firmly established. A wide range of studies has indicated that protein aggregation can directly result in the activation of the downstream biochemical cascades that signifificantly advance the pathology of AD29. The toxic aggregate species are recognized to be Aβ and Tau oligomers, whose abnormal interaction with cellular components can impair neuronal functions29,30. Given our present understanding of AD, slowing or inhibiting protein aggregation may contribute to preventing the advancement of the disease. Thus, understanding which pathways contribute towards this objective can offer insightful windows of opportunity for preventative and therapeutic interventions, as well as provide guidelines for sustainable ways of living (e.g. lifestyle factors, exercise, and diet) that suppress AD and promote healthy aging.

Fig. 1 Metabolite homeostasis and protein homeostasis in Alzheimer’s disease (AD). a AD is associated with a wide range of dysfunctions in cellular homeostasis (colored boxes). A variety of endogenous metabolites are involved in regulating these cellular processes (gray box insets). b Small molecule endogenous metabolites can modulate the protein aggregation processes associated with the disease. By designing systematic studies we can uncover the exact molecular mechanisms by which they balance protein homeostasis.

and misfolding, and the unfolded protein responses (UPR) in the endoplasmic reticulum and mitochondria31,32. Many heat shock proteins from the HSR are molecular chaperones that guide the conformation of proteins during biogenesis and prevent the misfolding and aggregation that interfere with cellular function21. Since the cytoplasm is also a pool of various metabolites produced and shuttled across biochemical networks, it is expected that their levels and threshold concentrations in the cell may have an effect as small molecule regulators on the protein homeostasis network, thus keeping aggregation-prone proteins in their soluble state. Patel and colleagues showed that ATP enhances protein solubility at physiological concentrations (mM range)33. One can thus envisage that other endogenous metabolites may also contribute to the modulation of protein aggregation. Enhanced expression of molecular chaperones, which is primarily regulated by the transcription factor heat shock factor 1 (HSF-1), has been shown to restore protein homeostasis in a variety of protein misfolding disease models, suggesting that this mechanism may represent a promising target for preventative and therapeutic approaches34. Calamini et al. identified new classes of small-molecule protein homeostasis regulators that induce HSF1-dependent chaperone expression and restore protein folding in multiple protein conformational disease models. These beneficial effects on proteome stability are mediated by HSF1, FOXO, NRF2, and the chaperone machinery34. In fact, endogenous metabolites are also small molecules that regulate key biological processes. Since the levels of these metabolites can change in response to environmental cues, it is important to understand their direct effects on protein homeostasis. In this work, we asked whether or not endogenous metabolites can influence the aggregation of proteins, and if they do, what is their mechanism of action. Is it a direct influence by binding to the monomeric, oligomeric, and fibrillar species or aggregates, or is it indirect by acting on various other cellular homeostatic mechanisms? Through our investigations, we identified two human endogenous metabolites that trigger a cytoplasmic unfolded protein response by increasing the levels of HSF-1, and downstream HSP40 co-chaperones J-proteins, ultimately resulting in the clearance of Aβ aggregates in a C. elegans model of AD.

Results

Identification of endogenous metabolites as small molecule protein homeostasis regulators. 

In order to identify metabolites associated with AD, we started by calculating fragments and similarity scores35 (see the “Methods“ section) of the small molecules reported as protein homeostasis regulators by Calamini et al.34 (Fig. 2a). Using these fragments, we screened the Human Metabolome Database (HMDB)36 for metabolites detected in the cerebrospinal fluid (CSF). As it is in direct contact with the extracellular space of the brain, the CSF provides a window on brain metabolism by reflecting the biochemistry of the brain37 in the form of its associated metabolites. In this work, we only considered endogenous metabolites, as we wanted to look into molecules active in primary cellular mechanisms, not involving exogenous metabolic products. Our reasoning is that focusing on endogenous metabolites can help identify fundamental cellular processes that are compromised in neurodegeneration as a response to the environment and/or cellular insults. Since the detection and quantification of metabolites in body fluids is challenging, we filtered only those metabolites that were extensively detected and quantified. We then used the KEGG database38 to identify biochemical pathways associated with these metabolites, and further, the metabolites that are particularly dysregulated in neurodegeneration (Supplementary Data 1). In the case of dysregulation, the levels of upregulation and downregulation of these metabolites are available from the HMDB and associated literature available therein. We then performed a literature search to corroborate the dysregulated status of the metabolites in neurodegeneration. We ranked this status as an association score, by performing a PubMed search (see the “Methods” section) (Supplementary Data 1). In principle, the association score describes the association of AD with the identified endogenous metabolite across studies reported in PubMed. Classifying the identified metabolites into upregulated and downregulated gives an insight into the effects of their normal concentrations on maintaining cellular homeostasis and their dysregulation in disease. We found 186 of the 380 endogenous CSF metabolites to be dysregulated in neurodegeneration (Supplementary Data 1). We then focused on six of these metabolites: 1-methylhistidine, homovanillic acid, melatonin, L-carnosine, vanylglycol, and kynurenic acid (Table 1), which were selected because they shared overlapping fragments with known HSF-1 protein homeostasis regulators (Fig. 2b). We then investigated experimentally their possible roles in modulating the aggregation of Aβ42, the highly cytotoxic 42-residue form of Aβ.

Carnosine and kynurenic acid rescue a C. elegans model of AD. C. elegans is a well-characterized model of aging sharing fundamental biochemical similarities with humans. 

Moreover, with an average lifespan of 2–3 weeks, it makes for a powerful in vivo tool for studying protein aggregation. Since we are investigating endogenous metabolites and wanted to investigate if and by which mechanism metabolites can clear aggregates, we considered selecting the C. elegans in vivo system for screening instead of an in vitro screening approach. We first carried out a dose-dependent screening (0, 1, 5, and 10 μM) (extended data in Supplementary Figs. 1 and 2) of the six metabolites that we identified as proteostasis regulators to test the overall fitness of a well-established C. elegans model, GMC101 (GMC), in which Aβ42 is overexpressed in the large muscle cells39. This model shows age-dependent inclusion formation and related toxicity, which can be measured by a decrease in the number of body bends per minute (BPM), an increase in paralysis rate, and a decrease in the speed of movement39 (Fig. 3). We report any signifificant benefits upon metabolite treatment as changes in overall fitness, where we consider fitness as the total behavioral response of worms as a function of BPM, speed (mm/s), and live ratio. We used Thioflflavin-T (ThT) as a positive control which has been previously shown to profoundly extend lifespan and slow aging in adult C. elegans40; these beneficial effects of ThT depend on HSF-1, the stress resistance and longevity transcription factor SKN-1, molecular chaperones, autophagy, and proteasomal functions40 (Supplementary Figs. 1 and 2). GMC and wild-type N2 worms were shifted onto the metabolites at the L4 stage, in order to avoid developmental effects, and subjected to motility assays on day 5 of adulthood when the phenotypic manifestations of Aβ42 toxicity are prominent (Fig. 3). We found that three metabolites, carnosine, any glycol, and kynurenic acid, improved the worm fitness as compared with the untreated GMC worms (Supplementary Figs. 1, 2, and for detailed characterization refer Fig. 3). We used ThT as a positive control for worm fifitness40. We found that the results obtained with carnosine and kynurenic acid were highly reproducible in terms of enhanced motility and markedly decreased NIAD-4-stained aggregates as compared to the untreated worms, and thus selected these two metabolites for further characterization.

Fig. 2 Identification of endogenous metabolites as protein homeostasis regulators that are dysregulated in AD. a Schematic of the computational fragment-based strategy that we used to identify endogenous metabolites from the Human Metabolome Database (HMDB) as candidate protein homeostasis regulators that are also dysregulated in AD. b Using this approach, we identified six candidate metabolites that serve as protein homeostasis regulators (Table 1, Supplementary Fig. 1 and Supplementary Data 1).

At different doses of metabolites, both carnosine and kynurenic acid show a clear response in the improvement of motility at day 5 (Fig. 3b–e) as compared with the initial (younger) days of adulthood. The GMC motility (in BPM) was highest at a concentration of 15 μM carnosine and 10 μM kynurenic acid, with a corresponding lower count of aggregates (Fig. 3f, g). We saw that carnosine is equally effective in the concentration range 10–25 μM, and. kynurenic acid was most effective in lowering aggregate number at a 15 μM concentration.

In both cases, further increase in concentrations (≥50 μM) did not show any improvement in motility, possibly because of unrelated toxic effects of the metabolites in C. elegans (Supplementary Fig. 1h, i). We did not observe any overt effects of carnosine and kynurenic acid on changes in worm motility and total fitness score in wild-type N2 worms (Supplementary Fig. 4). Total fitness score is calculated using metrics that include body bends, speed, paralysis rates, area per animal, and mean errors. The resulting fingerprint is generated in the form of a radar chart using open-source software as described in Perni et al.41.