A Pipeline for Natural Small Molecule Inhibitors of Endoplasmic Reticulum Stress

Abstract

The homeostasis of eukaryotic cells is inseverable of that of the endoplasmic reticulum (ER). The main function of this organelle is the synthesis and folding of a significant portion of cellular proteins, while it is also the major calcium reservoir of the cell. Upon unresolved ER stress, a set of stress response signaling pathways that are collectively labeled as the unfolded protein response (UPR) is activated. Prolonged or intense activation of this molecular machinery may be deleterious. It is known that compromised ER homeostasis, and consequent UPR activation, characterizes the pathogenesis of neurodegenerative diseases. In an effort to discover new small molecules capable of countering ER stress, we subjected a panel of over 100 natural molecules to a battery of assays designed to evaluate several hallmarks of ER stress. The protective potential of these compounds against ER stress was evaluated at the levels of calcium homeostasis, key gene and protein expression, and levels of protein aggregation in fibroblasts. The most promising compounds were subsequently tested in neuronal cells. This framework resulted in the identification of several bioactive molecules capable of countering ER stress and deleterious events associated to it. Delphinidin stands out as the most promising candidate against neurodegeneration. This compound significantly inhibited the expression of UPR biomarkers, and displayed a strong potential to inhibit protein aggregation in the two aforementioned cell models. Our results indicate that natural products may be a valuable resource in the development of an effective therapeutic strategy against ER stress-related diseases.

Keywords: drug discovery; endoplasmic reticulum stress; natural products; neurodegeneration; unfolded protein response.

FIGURE 1

Physico-chemical properties of the library used (molecular weight, NP-likeness score, AlogP, Number of Lipisnki violations), and distribution of the chemical (super)classes of the molecules present in the library.

FIGURE 2

Impact of a library of natural products on MRC-5 fibroblasts viability, as determined by MTT reduction assays. Results represent the percentage of the control and correspond to, at least, three independent experiments, each performed in triplicate.

FIGURE 3

Cytosolic calcium levels in MRC-5 fibroblasts in response to co-incubation of calcium ionophore A23187 at 5 µM and nontoxic molecules (50 µM), as determined by the fluorescence of the probe Fura-2/AM. (A) Heatmap representing all molecules tested; (B) Molecules that caused a statistically significant decrease in calcium levels when compared to the positive control. Results correspond to the mean of, at least, three independent experiments, each performed in triplicate.

FIGURE 4

Effect of 5-deoxykaempferol, delphinidin, fisetin, myristic acid, naringenin, naringin, sennoside B, spermine and quercetin-3-O-β-D-glucoside on Tg-induced UPR signaling, as determined by RT-qPCR. Results correspond to the mean ± standard error of the mean of, at least, four independent experiments, each performed in duplicate. GAPDH was used as a reference gene for normalization of expression.

FIGURE 5

(A) Impact of delphinidin, fisetin, myristic acid, sennoside B and quercetin-3-O-β-D-glucoside on Aβ_25-35-induced protein aggregation in MRC-5 cells, determined by the fluorescence of thioflavin T. Results correspond to the mean ± standard error of the mean of, at least, three independent experiments, individually performed in triplicate. (B) The protective effect of bioactive molecules was captured under a fluorescence microscope. The fluorescence signal of thioflavin T is represented in the histograms. (C) Effect of delphinidin on Tg-induced BiP expression in MRC-5 cells, as evaluated by Western blotting. Results correspond to the mean ± standard error of the mean of, at least, three independent experiments. (D) Images obtained through Western blotting analysis.

FIGURE 6

(A) Effect of delphinidin and fisetin on Aβ_25-35-induced protein aggregation in SH-SY5Y cells, as evaluated by determination of the fluorescence of thioflavin T. Results correspond to the mean ± standard error of the mean of, at least, three independent experiments, each performed in triplicate. (B) The effect of bioactive molecules was observed under a fluorescence microscope. The fluorescence signal of thioflavin T is represented in the histograms. (C) Effect of delphinidin on Tg-induced BiP expression in SH-SY5Y cells, as evaluated by Western blotting. Results correspond to the mean ± standard error of the mean of, at least, three independent experiments. (D) Example of the obtained Western blotting results.

FIGURE 7

Correlation matrix for the features used for the chemical library used in this work.

FIGURE 8

Different dimensionality reduction methods applied to chemical library under study. The active molecule delphinidin is presented in orange and inactive molecules in blue. (A) variation explained: 0.9. (B) number of components: 2, metric MDS, max_iterations=300. (C: left) number of neighbors: 45; minimum distance: 0.01. (C: right) number of neighbors: 45; minimum distance: 0.1. (D) perplexity: 50; number of iterations: 300.