Bioprospection of Natural Sources of Polyphenols with Therapeutic Potential for Redox-Related Diseases

Plants are a reservoir of high-value molecules with underexplored biomedical applications. With the aim of identifying novel health-promoting attributes in underexplored natural sources, we scrutinized the diversity of (poly)phenols present within the berries of selected germplasm from cultivated, wild, and underutilized Rubus species. Our strategy combined the application of metabolomics, statistical analysis, and evaluation of (poly)phenols’ bioactivity using a yeast-based discovery platform. We identified species as sources of (poly)phenols interfering with pathological processes associated with redox-related diseases, particularly, amyotrophic lateral sclerosis, cancer, and inflammation. In silico prediction of putative bioactives suggested cyanidin–hexoside as an anti-inflammatory molecule which was validated in yeast and mammalian cells. Moreover, cellular assays revealed that the cyanidin moiety was responsible for the anti-inflammatory properties of cyanidin–hexoside. Our findings unveiled novel (poly)phenolic bioactivities and illustrated the power of our integrative approach for the identification of dietary (poly)phenols with potential biomedical applications.


Yeast plasmids and strains
All plasmids and strains used in this study are listed in Table S1 and S2. To construct p426_GAL1pr-FLAG-HTT103Q-GFP, the sequence GAL1pr-FLAG-HTTp103-GFP from p425GAL1_HTT103Q was amplified by PCR and cloned into the p426 vector using the In-Fusion Cloning kit (TAKARA Clontech). p426_GAL1pr-GFP-A42 was generated by inserting the sequence GFP-A42, obtained by the double digestion of p416_GPDpr-GFP-A42 with BamHI/SmaI, into the p426 vector.
p425_GAL1pr-GFP-A42 was generated by subcloning the sequence GAL1pr-GFP-A42 into the SacI/HindIII restriction sites of p425 vector.

Yeast growth conditions
SC medium containing 1% raffinose was used for growth of PD and ALS integrative yeast models.
Synthetic dropout CSM-URA medium containing 1% raffinose was used for growth of ALS episomal yeast model. Synthetic dropout CSM-LEU, CSM-LEU-URA and CSM-HIS-URA-TRP media supplemented with standard concentrations of the required amino acids and containing 1% raffinose, were used for growth of HD, AD and RAS-RAF-interaction yeast models, respectively. Growth of Crz1-activation yeast model was performed in SC medium containing 2% glucose and Crz1 activation was induced with 1.8 mM MnCl2. Cells cultures were prepared as described in Materials and Methods.

Growth assays.
Growth assays were carried out as described in Materials and Methods section.

Growth curve data analysis methods
Raw data were exported from Excel and read into R software for plots construction, calculation of growth parameters and performance of statistical analysis to compare curves (our unpublished data).
Briefly, values of optical densities at 600 nm (raw OD) were read for 9 replicates, and the corresponding blank values. The procedure of Toussaint and Conconi [6] was then implemented: raw ODs were subtracted by the correspondent blank value to give the corrected OD values; whenever corrected OD values were negative (which implies that the OD was smaller or equal to the correspondent blank value), a corrected OD of 0.001 was used; corrected ODs were divided by the minimum OD and then transformed applying the natural logarithm (giving lnODs). Corrected OD values divided by the minimum OD were represented graphically. To calculate the growth parameters we used Adjustment of a model-free spline (nonparametric) and Model fitting (parametric) approaches.

Flow cytometry
Growth assays were carried out as described in Materials and Methods. To analyze cell viability with PI, cells were incubated with 20 μg/mL of PI for 30 min at 30ºC protected from light. FCM was performed in a FACS BD Calibur, equipped with a blue solid state laser (488 nm), green fluorescence channel 530/30 nm, and orange red fluorescence channel 610/20 nm.

Fluorescence microscopy
Fluorescence microscopy was carried out as described in Materials and Methods.

Protein extraction and immunoblotting
Protein extraction and immunoblotting were carried out as described in Materials and Methods, using the TCA protein extraction protocol.

β-Galactosidase assays
β-Galactosidase assays were carried out as described in Materials and Methods. Monitoring of Crz1 activation in solid medium was performed similarly with the exception that cells were patched onto solid glucose medium supplemented or not with MnCl2 for 90 min before the overlay procedure.

Statistical analysis
Statistical analyses were carried out as described in Materials and Methods. Figure S1. PCA of the correlation matrix of the entire positive mode dataset, including blanks, QC samples (S. lycopersicum Crimson, Indigo [7] and Purple [8]) and Rubus samples. Principal components 1, 2, 3, 4 and 5 explain 25.36, 9.85, 6.88, 6.29 and 5.39% of the variation, respectively. The PCA plots indicate a clear distinction between the samples and QCs and blanks. The QCs consist of tomato juice from tomatoes expressing different classes of (poly)phenols thereby confirming that the analytical method utilized in this study successfully distinguishes Rubus (green) from non-Rubus material and different tomato juices based on their (poly)phenolic profiles.

Supplementary information: SMART discovery platform
A yeast-based screening platform was used for the identification of Rubus bioactivities. We used various yeast models of redox-related neurodegenerative diseases (NDs) based on the expression of human genes associated with different NDs, (SNCA [2], FUS/TLS [5], HTTpQ103 [9] or Aβ42 [10]) fused to GFP (Figures S3a-S6a). Upon induction of expression of each protein with galactose, the growth of yeast cells was impaired (Figures S3b-S6b, S3c-S6c). Growth data was modeled using nonlinear parametric regression to estimate the growth parameters (final biomass, maximum growth rate, lag time, doubling time and area under curve -AUC), as well as the percentage of protection (with 95% confidence intervals) (Figures S3c-S6c). The AUC parameter was used to calculate the protection factor of Rubus extracts towards each disease pathological process.
Protein expression, inferred by the increase in GFP fluorescence signal (Figures S3d-S6d) was accompanied by an increase of propidium iodide-permeable cells in the PD model, indicating also cytotoxicity that was associated with the accumulation of protein inclusions (Figures S3e-S6e). We used the various yeast models for the identification of bioactive extracts/compounds that interfered with specific cellular pathologies associated with NDs.
In addition, we also included models of cancer-associated cellular pathologies in the SMART genes (Figure S7a), through the measurement of β-galactosidade activity (Figures S7b,c) and cellular growth in media devoid of lysine (Figures S7d,e). This simple system greatly facilitates the identification of bioactive molecules potentially inhibiting RAS/RAF pathological interactions in large collections.
As a common link between neurodegeneration and cancer, inflammation models were also included in the discovery platform. Crz1 is the yeast homologue of NFAT, a transcription factor controlling inflammatory responses in humans. Similarly to NFAT, Crz1 regulation is modulated by the calcium (Ca 2+ )-signaling pathway, which culminates in calcineurin (CaN) activation by calmodulin, Crz1 dephosphorylation and nuclear translocation [11,12]. The yeast Ca 2+ /CaN/Crz1 reporter strain encodes lacZ under the control of a promoter bearing Crz1 binding sites (Crz1-lacZ model) [4] (Figure S8a), representing an easy tool to assess Crz1 activation through the measurement of β-galactosidade activity [12] (Figure S8c). An additional strain, encoding the fusion CRZ1-GFP driven by the native CRZ1 promoter (Figure S8b) allowed assessment of Crz1 nuclear accumulation in cells with disturbed Ca 2+ cytosolic levels [4] (Figure S8d). Given the evolutionary conservation of NFAT and Crz1 activation mechanisms, reinforced by the conserved activity of FK506 and cyclosporin immunosuppressants in both yeast and humans [13], the yeast Ca 2+ /CaN/Crz1 reporter assay represents an easy and reliable tool to identify small molecules with potential to attenuate NFAT-mediated inflammatory responses.         can be a special case of pattern B, in which different features may exert bioactive properties, thus when the feature is present in low amounts, the protective effect could be conferred by another compound.