Plants synthesize a staggering variety of secondary metabolites that provide a chemically diverse pool of high-value small molecules with potential application for human health that cannot be matched by any synthetic libraries [1
]. The use of plants in traditional medicine dates back to antiquity and is still, despite huge investments into combinatorial chemistry and high-throughput screens, an important source of novel drugs and metabolites with a myriad of underexplored pharmacological and biotechnological applications [2
]. From conventional folk medicine to the scientific validation of protective properties, (poly)phenolic compounds have been implicated and/or identified as underpinning the beneficial health properties of several plants.
is a large and diverse genus of the Rosaceae family comprising more than 250 species with the most commonly known being red and black raspberries and blackberries. These fruits are characterized by their high polyphenolic content and diversity, which make them a major source of (poly)phenols with potential importance for human health. This includes redox-related diseases, such as neurodegeneration and cancer, which is consistent with the well-described role of (poly)phenols in targeting signaling pathways regulating redox homeostasis. Besides sharing oxidative stress and chronic inflammation as common pathological processes, neurodegenerative diseases (NDs) are also known as conformational disorders, as they are associated with protein misfolding and aggregation [3
] in a process thought to lead to neuronal death. Alzheimer’s disease (AD) pathology is associated to the accumulation of Aβ42 amyloid plaques [5
] and hyperphosphorylated tau neurofibrillary tangles [6
]. The accumulation of concentric hyaline cytoplasmic inclusions of α-synuclein (αSyn), known as Lewy bodies (LBs), is the major pathological hallmark of Parkinson’s disease (PD) and other LB diseases [7
]. Proteotoxic aggregates in neuronal cells of Huntington’s disease (HD) patients are formed by N-terminal polyglutamine (polyQ)-expanded huntingtin (HTT) [9
]. In amyotrophic lateral sclerosis (ALS), fused in sarcoma or translocated in liposarcoma (FUS/TLS) protein has been implicated in the formation of toxic aggregates and neuronal demise [10
With the goal of harnessing the diversity of Rubus
(poly)phenols for the discovery of new phenolic compounds of value, we developed an integrative approach that combined the power of metabolomics, for polyphenolic content characterization, and a Simple Molecular Architecture Research Tool (SMART) discovery platform for filtering potential bioactivities to be further explored in advanced pre-clinical models [12
]. The platform is composed of yeast strains expressing human disease genes associated with the most-studied NDs as cited before (Aβ42
]), cancer (RAS
]), and inflammation (CRZ1
], the yeast orthologue of human Nuclear Factor of Activated T-cell—NFAT). This is possible due to the high degree of evolutionary conservation of fundamental biological processes among eukaryotes, which has established the budding yeast as a powerful model for the identification of molecular targets amenable for therapeutic intervention and lead molecules with health-promoting potential [21
]. Benefiting from the easy and low-cost handling, facile genetic manipulation and the possibility to search against specific molecular targets, yeast-based screening technologies have proved to be very useful for the identification of promising drug candidates [12
] including the flavonoids quercetin and epigallocatechin gallate [26
2. Materials and Methods
2.1. Plant Material and Extraction Procedure
A range of different cultivars and species from the Rubus
genus cultivated in Portugal (Odemira) and UK (Dundee) (Table S1
) were manually harvested in the field at full ripeness as assessed by picker. Samples were kept in a cool box until they were transferred to −20 °C storage. Samples were extracted as described by Dudnik et al. [12
]. In summary, approximately 50 g of frozen fruit from each species/cultivar was weighted and transferred into a solvent-proof blender containing 150 mL of pre-cooled 50 ng/mL Morin (Sigma–Aldrich, Gillingham, UK) solution prepared with 0.2% formic acid methanolic solution. Samples were then blended with three pulses of 10 s duration and subsequently filtered using Whatman filter paper grade 1. The filtrate was aliquoted and solvent-dried using a speed-vac (VWR, Lutterworth, UK) and subsequently lyophilized. Dried extracts were flushed with N2
and stored at −20 °C until analysis by Liquid Chromatography coupled to a Time-of-Flight Mass Spectrometer (LC-ToF-MS). The filtrates to be used in the cell assays were solvent-dried using speed-vac, resuspended in CH3
O (50/50), and subjected to solid-phase extraction [27
2.2. Total Phenolic Quantification
Total phenolic content of the eluates were determined using the Folin–Ciocalteu method adapted to a microplate reader [27
]. The eluates were aliquoted, freeze-dried, and frozen at −20 °C.
2.3. Phenolic Profile Determination by LC-ToF-MS
Analysis of sample extracts was performed as described by Dudnik et al. [12
]. Briefly, dried extracts from each species/cultivar were resolubilized in triplicate using 2 mL of a 75% methanol solution with 0.1% formic acid. Five hundred microliters of the extract were decanted into filter vials, sealed with 0.45 mm Polytetrafluoroethylene-lined screwcap (Thomson Instrument Company, London, UK) and transferred into the autosampler. The analysis was achieved using an Agilent LC-ToF-MS system consisting of a quaternary pump (Agilent 1260, Cheadle, UK), a diode-array-detector (DAD) (Agilent 1260), a temperature control device (Agilent 1260), and a thermostat (Agilent 1290) coupled to an Agilent 6224 time-of-flight (ToF) instrument. Five microliters of the sample were injected onto a 2 × 150 mm (4 µm) C18 column (Phenomenex, Torrance, CA, USA) fitted with a C18 4 × 2 mm Security Guard™ cartridge (Phenomenex, Torrance, CA, USA). Sample and column temperature were maintained at 4 °C and 30 °C, respectively. The samples were eluted at a flow rate of 0.3 mL/min using two mobile phases (A: 0.1% formic acid in ultrapure water; B: 0.1% formic acid in 50:50 ultrapure water:acetonitrile) with the following gradient: 0 min 5% B; 4 min 5% B; 32.00 min 100% B; 34.00 min 100% B; 36.00 min 5% B; 40.00 min 5% B.
For optimal electrospray ionization conditions, the nebulizer pressure, drying gas temperature, and drying gas were set to 45 psi, 350 °C, and 3 L/min, respectively. In addition, the DAD was performed at 254, 280, and 520 nm. Morin levels (internal standard) were integrated in Agilent Mass Hunter Quan software (v. B.06.00, Cheadle, UK), and all samples with deviations larger than 10% relative to the dataset mean were reinjected. For all samples, three aliquots were analyzed across three different analytical batches.
2.4. Component Detection, Peak Alignment, and Integration
All chromatograms were evaluated in the same manner using the Agilent Software Profinder v. B.06.00, as previously described by Dudnik et al. [12
], which integrates peak findings in an automated and unbiased way with a peak integration user interface that allows user-driven curation of individual peaks. For positive mode data, the batch recursive molecular feature was used with peak extraction restricted to 2.1–38.00 min of the chromatography and peaks with levels higher than 15,000 counts with potential adducts of +H, +Na+
, and +NH4+
(−H and +Cl−
in negative mode) and a maximum of one charge state. The compound ion count threshold was set at two or more ions, and for alignment purposes the RT window was set at 0.70% + 0.60 min and the mass window was set at 25 ppm + 2 mDa. A post-processing filter to restrict analysis to compounds with more than 15,000 counts and present in at least 3 of the files in at least one sample group (species/line). The find-by-ion options were set to limit the extracted ion chromatogram (EIC) to the expected retention time +/− 0.40 min. The “Agile” algorithm was used for the integration of EIC, with a Gaussian smoothing of 9 points applied before the integration and a Gaussian width of 3 points. Additionally, peak filters were set at over 15,000 counts and the chromatogram formats were set to centroid when available and otherwise profile. The spectrum was extracted at 10% of peak height and excluded if the spectra within the m
range used was above 20% of the saturation. Finally, a post-processing filter was applied and compounds with less than 15,000 counts or present in less than 3 files in at least one sample group (species/line) were excluded. The output of automated peak finding and integration resulted in 542 and 210 molecular features found in the positive and negative modes, respectively. Manual curation resulted in the narrowing down of the molecular features found to 366 and 169 in the positive and negative modes, respectively. These were subsequently used in the statistical analysis.
2.5. Multivariate Analysis
GenStat for Windows, 16th Edition (VSN international Ltd., Hemel Hempstead, UK) was used for all the multivariate analysis performed. A principal component analysis (PCA), based on the correlation matrix, was applied to all the QC samples to ensure that the blank, reference samples, and berry samples were well separated (data not shown). The positive and negative metabolite datasets were analyzed separately and PCA plots were generated for the first 4 principal components. These were subsequently used for selecting the species with the greatest phytochemical differences.
2.6. Yeast Strains, Plasmids, and Transformation
Strains and plasmids used in this study are listed in Tables S3 and S4
, respectively. The W303-1A_FUS and W303-1A_T strains were obtained by transformation of the W303-1A strain with plasmids pAG303_GAL1pr-FUS and pAG303_ GAL1pr-ccdB previously linearized with BstZ
17I. Yeast transformation procedures were carried out as indicated using the lithium acetate standard method [28
2.7. Yeast Growth Conditions
Synthetic complete (SC) medium (0.67% yeast nitrogen base without amino acids (YNB) (DifcoTM
Thermo Scientific Inc., Waltham, MA, USA) and 0.79 g/L complete supplement mixture (CSM) (MP Biomedicals, Inc.—Fisher Scientific, Irvine, CA, USA)), containing 1% raffinose was used for growth of PD and ALS integrative yeast models. Synthetic dropout CSM-URA
medium (0.67% YNB and 0.77 g/L single amino acid dropout CSM−URA
(MP Biomedicals, Inc.—Fisher Scientific, Irvine, CA, USA)) containing 1% raffinose was used for growth of AD and ALS episomal yeast models. For growth of the HD model, a synthetic dropout SC-LEU medium was used (0.67 % YNB and 0.54 g/L 6-amino acid dropout CSM–ADE–HIS–LEU–LYS–TRP–URA
(MP Biomedicals, Inc.—Fisher Scientific, Irvine, CA, USA), supplemented with standard concentrations of the required amino acids and containing 1% (w/v
) raffinose. For growth of the RAS–RAF interaction yeast model, CSM–HIS–URA–TRP
media was used (0.67% YNB and 0.54 g/L 6-amino acid dropout CSM–ADE–HIS–LEU–LYS–TRP–URA
(MP Biomedicals, Inc.—Fisher Scientific, Irvine, CA, USA), supplemented with standard concentrations of the required amino acids and containing 1% raffinose. In all conditions, medium containing glucose (control, disease-protein OFF) and galactose (disease-protein ON), at a final concentration of 2%, were used for the repression or induction of disease protein expression, respectively. Growth of Crz1 activation yeast model was performed in SC medium containing 2% (w/v
) glucose, and Crz1 activation was induced with 1.8 mM MnCl2
]. Radicicol (Sigma, Gillingham, UK) and FK506 (Cayman Chemicals, Ann Arbor, MI, USA) were used as positive controls for the yeast models of RAS–RAF interaction and Crz1 activation, respectively.
A pre-inoculum was prepared in raffinose or glucose (for Crz1 activation model) medium. Cultures were incubated overnight at 30 °C under orbital shaking, diluted in fresh medium, and incubated under the same conditions until the optical density at 600 nm (OD600) reached 0.5 ± 0.05 (log growth phase). Cultures were then diluted according to the equation: ODi × Vi = (ODf/(2(t/gt)) × Vf, where ODi = initial optical density of the culture, Vi = initial volume of culture, ODf = final optical density of the culture, t = time (usually 16 h), gt = generation time of the strain, and Vf = final volume of culture. Readings were performed in a 96 well microtiter plate using a Biotek Power Wave XS plate spectrophotometer (Biotek® Instruments, Winooski, VT, USA). Dried extracts of Rubus species/cultivar obtained after total phenolic compounds determination were re-solubilized in adequate growth medium for the cellular assays.
2.8. Growth Assays
For the phenotypic growth assays, the strains were grown as described to OD600 0.1 ± 0.01 and were inoculated (OD600 0.2 ± 0.02) in medium supplemented or not with the indicated concentrations of Rubus extracts. After 6 h, OD600 nm was adjusted to 0.05 ± 0.005, serial dilutions were performed with a ratio of 1:3, and 5 μL of each dilution was spotted onto solid medium containing glucose or galactose as the sole carbon sources. Growth was recorded after 48 h incubation at 30 °C. Images were acquired using ChemidocTM XRS and Image-Lab® 6.0.1 software (Biorad, Hercules, CA, USA). For the growth curves, yeast cultures were diluted to OD600 0.12 ± 0.012 in fresh medium supplemented or not with the indicated concentrations of the extracts in a 96 well microtiter plate. After 2 h incubation at 30 °C, cultures were further diluted to OD600 0.03 ± 0.003 in repressing (glucose) or inducing (galactose) media supplemented or not with the extracts. The cultures were incubated at 30 °C with shaking for 24 h or 48 h (for the AD model) and cellular growth was monitored hourly by measuring OD600 using a Biotek Power Wave XS Microplate Spectrophotometer (Biotek® Instruments, Winooski, VT, USA). The areas under the curve (AUC) were integrated using the Origin 6 software (OriginLab, Northampton, MA, USA). For the RAS/RAF interaction model, final biomass was calculated by normalizing OD600 of cultures after 48 h incubation at 30 °C to the initial OD600.
2.9. Flow Cytometry
Cell cultures at OD600 0.2 ± 0.02 were exposed or not to the indicated concentrations of Rubus extracts for 6 h. Cultures were further diluted to OD600 0.2 ± 0.02 in glucose and galactose supplemented or not with the indicated concentrations of Rubus extracts for 12 h. Flow cytometry was performed in a FACS BD Calibur equipped with a blue solid-state laser (488 nm) and green fluorescence channel 530/30 nm. Data analysis was performed using FlowJo software (BD, San Jose, CA, USA), and the cell doublets exclusion was performed based on Forward-A and -W scatter parameters. A minimum of 30,000 events were analyzed for each experiment. Results are expressed as the percentage of GFP positive cells as compared to the control.
2.10. Fluorescence Microscopy
Yeast cells subjected to the same treatment as above were monitored for the formation of disease-protein intracellular inclusions or nuclear translocation of Crz1 by fluorescence microscopy using a Leica DMRA2 fluorescence microscope (Leica, Wetzlar, Germany) equipped with a CoolSNAP HQ CCD camera (1.3MPx monochrome). Images were analyzed using ImageJ 1.8.0 software (NIH, Bethesda, MD, USA).
2.11. Protein Extraction
Tris-based buffer (TBS) [30
] or trichloroacetic acid (TCA)/MURB (50 mM sodium phosphate, 25 mM MES pH 7.0, 1% SDS, 3 M urea, 0.5% 2-mercaptoethanol, 1 mM sodium azide) [31
] were used for total protein extraction. Aliquots corresponding to OD600
1–2 of cultured cells were harvested by centrifugation at 5000× g
for 3 min. For TBS extraction, cells were resuspended in TBS supplemented with protease and phosphatase inhibitors, disrupted with glass beads (3 cycles of 30 s in the vortex and 5 min on ice), and cell debris were removed by centrifugation at 700× g
for 3 min. Total protein was quantified using the MicroBCA kit (Thermo Fisher Scientific, Waltham, MA, USA) according the manufacturer’s instructions. Samples were incubated at 95 °C for 10 min before SDS-PAGE.
For the TCA/MURB protocol, cells were first resuspended in TCA to a 10% final concentration, and the samples were incubated for 20 min at −20 °C. The cells were harvested by centrifugation at 15,000× g for 3 min, washed twice with acetone, and the air-dried cell pellet was resuspended in MURB supplemented with protease and phosphatase inhibitor cocktails. Cells suspensions were disrupted with glass beads (3 cycles of 30 s in the vortex and 5 min on ice), the samples were incubated at 70 °C for 10 min, and unlysed cells were removed by centrifugation at 10,000× g for 1 min.
Equal volumes of protein extract, normalized to the OD600 of cell cultures (for TCA/MURB protocol), or equal concentrations of total proteins (for TBS protocol) were loaded in a 15% SDS-PAGE. The Trans-Blot Turbo transfer system (BioRad, Hercules, CA, USA) was used to transfer proteins to a 0.22 µm nitrocellulose membrane according to the manufacturer’s specifications. Membranes were washed with TBS, blocked with 5% skim milk in TBS-Tween for 1 h at room temperature, and incubated overnight at 4 °C with antibodies against GFP (1:5000, Neuromab, Davis, CA, USA), FUS/TLS (1:1000, Millipore, Burlington, MA, USA), and PgK1 (1:5000, Life Technologies Corporation, Carlsbad, CA, USA). Membranes were washed three times with TBS-Tween and incubated with horseradish peroxidase-conjugated secondary antibodies (1:10,000, Pierce, Waltham, MA, USA) for 2 h at room temperature. Protein signals were detected using Amersham ECL Prime Detection Reagent (GE Healthcare, Chicago, IL, USA) and signal intensity was estimated using the ImageJ 1.8.0 software (NIH, Bethesda, MD, USA).
2.13. β–Galactosidase Assays
For monitoring of RAS/RAF interaction, cell cultures at OD600
1 ± 0.1 were exposed or not to the indicated concentrations of Rubus
extracts for 90 min. The extracts were removed, cells were patched at a density of 4.5 × 107
onto solid medium containing glucose or galactose, and incubated 3 h at 30 °C. The assay was revealed by overlaying the cells with 5-Bromo-4-Chloro-3-Indolyl β-d
-Galactopyranoside (X-Gal) solution (0.5% agarose, 50% LacZ buffer, 0.2% SDS, 2 mg X-Gal/mL and at 70 °C). Plates were maintained at 30 °C and monitored until the development of the blue color [32
For quantitative measurements of β–galactosidase activity, cell cultures at OD600
0.5 ± 0.05 were diluted to OD600
0.1 ± 0.01 and challenged or not with extracts and the pure compounds for 90 min. Just before cell lysis, OD600
of cultures were recorded. Cells were incubated with Y-PER Yeast Protein Extraction Reagent (ThermoFisher Scientific) in 96 well microtiter plates for 20 min at 37 °C, LacZ buffer containing 4 mg/L 2-Nitrophenyl β-d
-galactopyranoside (ONPG) was added, and plates were incubated at 30 °C [32
]. The OD420
were monitored using a Biotek Power Wave XS Microplate Spectrophotometer (Biotek®
Instruments, Winooski, VT, USA). Miller units were calculated as described previously [33
2.14. Quantitative Real-Time PCR
The qRT-PCR analyses were performed according to the MIQE guidelines (Minimum Information for Publication of Quantitative Real-Time PCR Experiments) [34
]. Total RNA was extracted using the ENZA yeast RNA extraction kit (OMEGA, Norcross, GA, USA). After cleaning, 200–300 ng of total RNA was used for reverse-transcription with qScript™ cDNA superMix kit (Quanta Biosciences Inc., Gaithersburg, MD, USA). The qRT-PCR was performed in a LightCycler 480 Instrument (Roche, Basel, Switzerland), using LightCycler 480 SYBR Green I Master (Roche, Basel, Switzerland) to evaluate expression of the FUS–GFP
(5′-ACGGACACTTCAGGCTATGG-3′; 5′-CCCGTAAGACGATTGGGAGC-3′) (GeneID: 2521), PMR1
(5′-CACCTTGGTTCCTGGTGATT-3′; 5′-CCGGTTCATTTTCACCAGTT-3′) (GeneID: 852709), PMC1
(5′-GTGGCGCACCATTTTCTATT-3′; 5′-TACTTCATCGGGGCAGATTC-3′) (GeneID: 852878), and GSC2
(5′-CCCGTACTTTGGCACAGATT-3′; 5′-GACCCTTTTGTGCTTTGGAA-3′) (GeneID: 852920) genes. Standard curves were constructed for each gene and expression was calculated by the relative quantification method with efficiency correction using LightCycler 480 Software version 184.108.40.206 (Roche, Basel, Switzerland). Both ACT1
(5′-GATCATTGCTCCTCCAGAA-3′; 5′-ACTTGTGGTGAACGATAGAT-3′) and PDA1
(5′-TGACGAACAAGTTGAATTAGC-3′; 5′-TCTTAGGGTTGGAGTTTCTG-3′) were used as reference genes. The results were expressed as fold-change mRNA levels relative to the control (mRNA fold change) of at least three independent biological replicates.
2.15. Microglia-Induced Inflammation Model
Microglial N9 cells were cultured in EMEM (Eagle Minimum Essential Media, Sigma–Aldrich, Gillingham, UK) media, supplemented with 1% (v/v) L-glutamine (Biochrom AG, Berlin, Germany), 1% (v/v) penicillin/streptomycin, and 10% FBS (Fetal Bovine Serum, Gibco, Waltham, MA, USA). Cell cultures were maintained at 37 °C in 5% (v/v) CO2, and split at sub-confluent cultures (about 60–80%). Cells were then detached by agitation before suspension of the culture media with a pipette (no cellular detaching agent was used). For immunostaining, cells were grown at 5 × 104 cell/well in 24 well plates containing coated coverslips and cultured overnight. Cells were pre-incubated or not with 5 mM of the indicated compounds for 6 h in culture media with reduced FBS to 0.5% (v/v). The medium was discarded, and cells were washed with PBS. Fresh culture media containing 300 ng/mL of LPS (Lipopolysaccharide) or 3 mM ATP (Sigma-Aldrich–Poole, Gillingham, UK) was added and cell cultures were incubated for 1 h to induce transcription factor nuclear localization. For nitric oxide (NO) and tumor necrosis factor alpha (TNF-α) quantifications, cells were seeded at 5 × 105 cells/well in 6 well plates, cultured overnight, and pre-incubated or not with 5 mM of the indicated compounds for 6 h in culture media with reduced FBS to 0.5% (v/v). The medium was discarded, cells were washed with PBS, and incubated in fresh culture media containing 300 ng/mL of LPS for 24 h.
Immunostaining was performed as described by Figueira et al. [35
], using rabbit polyclonal anti-NF-κB p65 (C-20) (1:200, Santa Cruz Biotechnology, Santa Cruz, CA, USA) or rabbit polyclonal anti-NFAT1c (1:200, Cell Signalling, Danvers, MA, USA) as primary antibodies and Alexa 594 anti-rabbit IgG (1:500) (Invitrogen, Carlsbad, CA, USA) as the secondary antibody. Nuclei were counterstained with DAPI. Cells were washed three times with PBS between each incubation. Widefield images were acquired on a Leica DMRA2 upright microscope, equipped with a CoolSNAP HQ CCD camera, using a 63× 1.4NA oil immersion objective, DAPI + TRITC fluorescence filter sets. Post-acquiring treatment was performed using ImageJ 1.8.0 software (NIH, Bethesda, MD, USA).
2.17. Nitric Oxide (NO) Quantification
The NO release to media was quantified as described by Ii et al. [36
] using the Griess Reagent (Sigma–Aldrich, Gillingham, UK), according to manufacturer’s instructions. After incubation with LPS, cell media were removed and immediately analyzed for nitrite quantification. Standard curves of sodium nitrite (0–25 μmol/L) were prepared and absorbance was acquired in a Synergy HT microplate reader (Biotek®
Instruments, Winooski, VT, USA).
2.18. TNF-α Quantification
The TNF-α release was assayed by ELISA according to the manufacturer’s instructions (PeproTech®; Princeton Business Park, Rocky Hill, NJ, USA). All reagents and plates used were provided in the kit. For the standard curve, recombinant murine TNF-α (PeproTech®) was diluted from 0–2 μg/L. The plate was incubated at room temperature in a Synergy HT microplate reader (Biotek® Instruments, Winooski, VT, USA) for 35 min, with 5 min intervals Abs405 readings.
2.19. Statistical Analysis
The results reported in this study are the average of at least independent biological triplicates and are represented as the mean ± SEM. Analysis of variance with Tukey’s HSD (Honest Significant Difference) multiple comparison test (α = 0.05) using SigmaStat 3.10 (Systat, Chicago, IL, US) was used to assess the differences among treatments.
This study described the use of an integrative approach, combining the power of metabolomics, cellular assays and potent statistical analysis, to identify novel health-promoting attributes in underexplored (poly)phenol sources. The rationale involved the selection of the most chemically diverse samples of an extensive Rubus collection followed by the determination of health-promoting activities using a SMART discovery platform.
Overall, the study allowed the identification of (poly)phenol-enriched extracts and single compounds from Rubus modulating pathological processes of redox-related diseases responsible for major societal and economic impacts as well as provided some clues regarding the possible molecular mechanisms underlying their protective activity. Our objective was to deliver novel plant (poly)phenolic bioactives with the potential to be exploited either in food engineering and in the pharmaceutical and biotechnological sector as nutraceutical/therapeutic alternatives for redox-related chronic diseases. Of course, in therapeutic applications, development of formulations for controlled delivery to target tissues should further be developed to ensure that physiologically relevant concentrations of bioactive compounds reach their target sites.
Although the number of novel plant (poly)phenolics with potential bioactivity is limited, we tentatively identified several phytochemicals in Rubus
, such as triterpenoids, benzoyl di-hexoside, hydroxysphingosine, and a leucine isomer, which have not been extensively studied such as Rubus
-derived bioactive compounds. Interestingly, (−)-epicatechin has previously been described as possessing synergistic effects [58
], and while our model does not tackle synergistic and antagonistic effects, it is possible that significant results from the runs test could be associated with synergistic effects rather than intrinsically high bioactivity.
While more work is necessary in compound annotation, development of the statistical model in order to cope with synergistic and antagonistic effects and validation of bioactivities in advanced models, this report highlights the feasibility of this strategy for the replication and identification of novel bioactive lead molecules from crude extracts from berry fruits.