Identification of new fisetin analogs as kinase inhibitors: Data on synthesis and anti-skin cancer activities evaluation

This article contains supplemental datasets of the recently published related research article “Synthesis, Inverse Docking-Assisted Identification and in vitro Biological Characterization of Flavonol-based Analogs of Fisetin as c-Kit, CDK2 and mTOR Inhibitors against Melanoma and Non-melanoma Skin Cancers” by Roy et al., [1]. It provides in-depth data not included in the original co-submission on the biophysical, molecular docking, and biological characterization of newly synthesized flavonol-based analogs of fisetin, a natural dietary small molecule with anticancer and anti-inflammatory properties. These synthetic small molecules were investigated as new, potential single and/or multi-kinase inhibitors of the cyclin-dependent kinase-2 (CDK2), receptor tyrosine kinases (c-KITs), and mammalian targets of rapamycin (mTOR) targets, potentially active against melanoma or non-melanoma skin cancers. Furthermore, this data-in-brief article comprises additional sets of results on several aspects of the properties of the dual and multiple kinase inhibitor compounds’ effects that were not presented in the associated article, including the activated targets that are dysregulated in skin cancers; the effects on markers of apoptosis; on colony formation; and in scratch wound healing assays. The study has identified a panel of novel fisetin analogs that are either single- or multi-kinase inhibitors, which may be further developed as active for the treatment of melanoma and non-melanoma skin cancers. The dataset presented herein will be utilized for additional studies aiming to establish a biological platform to steer for predictive and experimental screening of novel flavonoids and analogs in relevant organoids, humanized animal models and in vivo disease models. The present results should also serve as a key stepping-stone towards enabling target-structure-based design, synthesis and initial testing of novel analogs or derivatives of fisetin. The current study may eventually lead to the development of safe, promising and preclinical candidate entities for treatment of skin and other forms of cancers as well as various other human diseases, which can possibly add to the general armamentarium of promising and safe drugs for health promotion.


Keywords:
Anticancer evaluation skin cancers fisetin analogs target prediction computational docking kinase activities melanoma and epidermoid carcinoma Skin Cancers" by Roy et al., [1] . It provides in-depth data not included in the original co-submission on the biophysical, molecular docking, and biological characterization of newly synthesized flavonol-based analogs of fisetin, a natural dietary small molecule with anticancer and anti-inflammatory properties. These synthetic small molecules were investigated as new, potential single and/or multi-kinase inhibitors of the cyclin-dependent kinase-2 (CDK2), receptor tyrosine kinases (c-KITs), and mammalian targets of rapamycin (mTOR) targets, potentially active against melanoma or nonmelanoma skin cancers. Furthermore, this data-in-brief article comprises additional sets of results on several aspects of the properties of the dual and multiple kinase inhibitor compounds' effects that were not presented in the associated article, including the activated targets that are dysregulated in skin cancers; the effects on markers of apoptosis; on colony formation; and in scratch wound healing assays. The study has identified a panel of novel fisetin analogs that are either single-or multi-kinase inhibitors, which may be further developed as active for the treatment of melanoma and nonmelanoma skin cancers. The dataset presented herein will be utilized for additional studies aiming to establish a biological platform to steer for predictive and experimental screening of novel flavonoids and analogs in relevant organoids, humanized animal models and in vivo disease models. The present results should also serve as a key stepping-stone towards enabling target-structure-based design, synthesis and initial testing of novel analogs or derivatives of fisetin. The current study may eventually lead to the development of safe, promising and preclinical candidate entities for treatment of skin and other forms of cancers as well as various other human diseases, which can possibly add to the general armamentarium of promising and safe drugs for health promotion.
Published by Elsevier Inc. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ) Table   Subject Chemistry and Biological sciences Specific subject area Medicinal Chemistry, Molecular Biology and Skin Cancer Research Type of data Table  Figure How data were acquired Dataset was collected as part of a synthesis and biophysical and biological characterization of novel flavonol derivatives of fisetin targeting specific kinases as anticancer agents [1] . The synthesized flavonol derivatives produced from microwave-assisted methods were purified and characterized using different spectroscopic and spectrometric techniques, such as 1 H-NMR, 13 C-NMR, IR, GC-MS and HR-ESIMS [1] . Cell cultures were prepared, analyzed, and protein content was quantified using the BCA reagent assay (Thermo Fisher Scientific, Waltham, MA, USA). Protein electrophoresis: SDS-PAGE method (Protean 3 system, Bio-Rad, Hercules, CA, USA), followed by protein expression via the western blotting method, using the Bio-Rad analysis and imaging system as earlier described [1 , 2] . Analysis and graphs in the figures, including in vitro enzymatic kinase activities, western blot, scratch ( continued on next page ) wound assays, and colony formation were plotted using GraphPad PRISM program suite, version 8 (GraphPad Software, Inc., La Jolla, CA, USA), and molecular docking analysis [ 1 , 2] . In silico prediction of skin permeation, lipophilicity, absorption, distribution, metabolism, excretion (ADMET) properties of the analogs compared to the reference drug fisetin were performed and analyzed using the online SwissADME platform of the Swiss Institute of Bioinformatics [3] . Data format Raw Analyzed Parameters for data collection Data were collected at the end of each experiment; each experiment was repeated at least three times. The post experiment analysis parameters for consideration were: effect or no effect, with statistical consideration for any significant differences. Description of data collection For western blot data, when exposed, faint and auto-exposed bands were obtained; the images were captured and analyzed using the Biorad image analysis software system. Densitometric data were collected, tallied and graphs were plotted to compare changes in the different treatment groups. The variations in protein and enzymes expressions in lysates and the differences in the proportion of colony sizes and numbers and wound closure areas were analyzed as in the associated research article [1] and as reported earlier [2] . Differences between samples treated with and without test compounds were measured and analyzed including protein and enzyme expression kinetics.

Value of the Data
• This additional dataset provides further insights into the characteristics of the different flavonol compounds synthesized as single or multi-kinase inhibitors. These compounds exhibited varying degrees of biophysical properties and biological effects on human skin cancer cell lines compared to minimal effects on untreated control groups. The data provide a clear picture of the range diversity of the effects of the different synthetic small molecular structures examined. • This data-in-brief article contains results that can be used by scientists and researchers as a reference and guide towards future investigations that may explore the potential therapeutic diversity as well as repurposing of these compounds for treating other human diseases. • This dataset may help to formulate new additional research hypotheses to explore the preclinical effects of these compounds toward ultimate clinical use. Also, scientists can utilize the data to improve drug discovery and development in anticancer and adjuvant therapies, alone or in combination with other known drugs. • The work presented informs further investigations on the toxicological and safety profiles in lieu of their formulations alone or in combination with existing FDA approved drugs.

Data Description
This dataset article provides supplementary data to the associated research article by Roy et al., [1] . Fig. 1 A-I reports the 1 H-NMR, 13 C-NMR, and HR-MS data for the three most active  Table 1 Synthesis of Fisetin analogs via microwave-assisted Claisen-Schmidt condensation and modified Algar-Flynn-Oyamada (AFO) reactions. analogs F9, F17 and F20 [1] . The general approach for the synthesis of Fisetin analogs and their structures provided in Table 1 . The full/detailed properties of all 24 flavonol analogs synthesized are provided in Supplementary dataset 1 and their raw NMR data files provided in ASCII file format and the raw HR-MS data files in .txt format linked in the (HR-MS Raw Files) in the source repository. Table 2 reports the results from the in silico prediction of absorption, distribution, metabolism, excretion (ADME) of the three potent fisetin analogs in the associated article (1). All 24 analogs were predicted to possess improved cell penetration properties, with the three most active being in the order as F17 > F9 > F20 , compared to fisetin, the parent natural compound. The table also shows the lipophilicity profile for which all compounds have good oral bio-absorption and intestinal absorption, but with less effective sub-lingual absorption. The Table 2 dataset presents on the three most active analogs, based on the log Kp value, is favored as follows F17 > F9 > F20 respectively, as compared to fisetin ( F0 ). This reveals an important property for the skin pharmaceutical applications, an indication that each derivative has a better skin permeation (cm/s) ability than the parent compound, fisetin. Detailed raw ADME prediction data for all 24 flavonol compounds synthesized are displayed in supplementary dataset 2. Tables 3 -14 report results of RMSD values obtained after Cross-docking/re-docking for the selection of PDB for EGFR ( Table 2 ), c-Kit ( Table 4 ), Akt ( Table 5 ), MET ( Table 6 ), MEK1 ( Table 7 ), VEGFR1 ( Table 8 ), Table 2 In silico prediction of absorption, distribution, metabolism and excretion (ADME) for the most potent fisetin analogs, namely F9, F17 and F20 . MAPK ( Table 9 ), mTOR ( Table 10 ), PI3K ( Table 11 ), PIP5K1 α ( Table 12 ), FGFR ( Table 13 ), and CDK2 ( Table 14 ) related to the parent article(1). Fig. 2 reports potency data analyzed from examination of the inhibitory activities of the selected, most active analogs compared to fisetin on the following kinases: A) c-Kit, B) CDK2/Cyclin A, C) CDK2/Cyclin E and D) mTOR. The attributes of IC 50 values ( μM) of calculated ranked inhibitor potency values are expressed as means ± SD of three independent assays and are listed and described in Table 3 of the related research article          [1] . Fig. 3 reports the mean western blot bands (A-B) and their corresponding densitometric data analysis in bar graph format (C-D) of the different melanoma (A375) and non-melanoma (A431) cells treated in the presence or absence of test analogs F9 and F17 , showing the differential expression of apoptosis induction via activation of the extrinsic and intrinsic apoptotic pathways. The data for F20 are described in the related research article [1] . Fig. 4 summarizes the Western blot band results (A-B) and the Bar graph plots data (C-D) from the analysis of the effects of different concentrations of flavonol analogs F9 and F17 on the protein expression levels of cyclin A2, cyclin E2, CDK2, phosphorylated c-Kit and Stat3 in 48h treated A375 and A431 skin cancer cells. In Fig. 5 , the Western blot bands (Upper panel) and the Bar graph plots (lower panels) exhibit analyzed data from the effects of different concentrations of analogs F9 and F17 on the protein expression levels of phosphorylated p90RSK/Akt/mTOR/MAPK (ERK1/2) and ribosomal  , components of the intrinsic apoptosis pathway in 48h treated A375 and A431 cells. A375 and A431 cells were incubated in the absence or presence of flavonol analogs (F9 and F17; 0, IC 50 , 2xIC 50 ; μM, 48h), and the whole-cell lysates of cells treated with/without F9 and F17 were subjected to SDS-polyacrylamide gel electrophoresis. Equal protein loading was confirmed by reprobing for β-actin or vinculin as loading control, and the actual protein levels were normalized to the loading control and expressed as percentage. The Western blot data shown are representative of immunoblot of more than two independent experiments with similar results. Fig. 3 (C-D). Potent flavonol analogs F9 and F17 induce apoptosis through activation of the extrinsic and intrinsic apoptotic pathways in melanoma and non-melanoma cells as shown graphically. The data expressed as the percentage of which analogs F9 and F17 significantly suppressed the protein expression levels of pro-and-cleaved caspases ( 3 ,7 and -8), PARP (116 kDa) and cleaved PARP (85 kDa), as well as Bax and Bcl-2 in (C) A375 and (D) A431 as compared with untreated control cells. Bar graphs represent mean ± SD of results of relative quantitative density values for the blots normalized with an internal loading control from three independent experiments. The statistical significance was determined using one-way ANOVA and Dunn's multiple comparison test, and p < 0.05 ( * ) was considered significant.  Fig. 7 describes the data from the determination of the long-term dose-dependent effect of the analogs F9 and F17 on the percentage of colonies when compared to the respective untreated cutaneous carcinoma control cell lines. All the data for compound F20 are displayed in the associated/parent research article [1] . The supplementary raw data files are available in the link.

Experimental Design, Materials and Methods
This study's major purpose was to validate the newly synthesized and fully characterized flavonol analogs [4] , as anticancer agents, with different effects on basal and squamous carcinoma of the skin cells in vitro . Inclusion criteria included in the supplementary datasets were not extensively presented in the associated article. These data complement and further enhance the results presented in the associated article [1] . These datasets and their analyses aimed at improving our understanding of the physicochemical and other pharmaceutical properties of the compounds in vitro were explored through the use of the online SWISS-ADME platform of the Swiss Institute of Bioinformatics ( http://www.swissadme.ch ) [3] , and compared with the same  (F9 and F17;0,IC 50 ,2xIC 50 ;μM,48h), and western blotting performed as described in the method section. Equal protein loading was confirmed by reprobing for β-actin as loading control, and protein levels were normalized to the loading control and expressed as percentage.
The Western blot data shown are representative of immunoblots of three independent experiments with similar results. (C and D) The data expressed in the Bar graphs represent mean ± SD of relative quantitative normalized density values in percentage with an internal loading control from three independent experiments. The analogs F9 and F17 significantly suppressed the protein expression levels of these in (C) A375 and (D) A431 as compared with untreated control cells. For bar graphs, the statistical significance was determined using one-way ANOVA and Dunn's multiple comparison test, and p < 0.05 ( * ) was considered significant. descriptors obtained for fisetin, the reference compound. A detailed description of the procedures of study-specific investigation and the molecular and cellular assay protocols are essentially as presented in the related research article [1] . Equal protein loading was confirmed by reprobing for β-actin or vinculin as a loading control, and the actual protein levels were normalized to the loading control and expressed as a percentage. The Western blot data shown are representative of immunoblot of more than two independent experiments with similar results. Statistical analysis was conducted, and graphics were designed using GraphPad PRISM program suite, version 8 (GraphPad Software, Inc., La Jolla, CA, USA). The data expressed in the Bar graphs represent mean ± SD of values in percentage control from three independent experiments; the statistical significance was determined using one-way ANOVA and Bonferoni's multiple compar- Fig. 5. Potent flavonol analogs F9 and F17 inhibit the protein expression levels of phosphorylated p90RSK/Akt/mTOR/MAPK (ERK1/2) and ribosomal protein S6 in A375 melanoma and A431 non-melanoma cells. Role of the flavonol analogs F9 and F17 on the protein expression levels of cyclin A2, cyclin E2, CDK2, phosphorylated Akt, p90RSK, mTOR, MAPK (ERK1/2) and ribosomal protein in 48h treated (A) A375 and (B) A431 cells. A375/A431 cells were incubated with/without analogs (F9 and F17; 0, IC 50 , 2xIC 50 ; μM, 48h), and western blotting performed as described in method section. Equal protein loading was confirmed by reprobing for β-actin as loading control, and protein levels were normalized to the loading control and expressed as percentage. The Western blot data shown are representative of immunoblot of three independent experiments with similar results (C and D). The data expressed in the Bar graphs represent mean ± SD of relative quantitative normalized density values in percentage with an internal loading control from three independent experiments. The analogs F9 and F17 significantly suppressed the protein expression levels of these in (C) A375 and (D) A431 as compared with untreated control cells. For bar graphs, the statistical significance was determined using one-way ANOVA and Dunn's multiple comparison test, and p < 0.05 ( * ) was considered significant. (C and D) significant dose-dependent decrease in cultured cells' scratch wound healing areas was observed in the presence of selected potent flavonol hits, compared to untreated control cells, after 48 h of incubation. The data expressed in the Bar graphs represent mean ± SD of scratched wound area values in percentage compared to 0h control and 24h controls from three independent experiments. The analogs F9 and F17 significantly suppressed wound healing area (C) A375 and (D) A431 as compared with untreated control cells. The statistical significance was determined using one-way ANOVA and Dunn's multiple comparison test, and p < 0.05 ( * ) was considered significant. (C and D) The data expressed in the Bar graphs represent mean ± SD of values in percentage control from threeindependent experiments, the statistical significance was determined using one-way ANOVA and Bonferoni's multiple comparison test, and p < 0.05 ( * ) was considered significant.
ison test or Dunn's multiple comparison test. P-values ≤ 0.05 were considered statistically significant.

Ethics Statement
The primary normal human epidermal keratinocytes used were isolated from new born foreskin biopsies or adult skin biopsies and established in culture using published protocols, [5] , and were obtained under a University of Wisconsin-Madison-approved institutional review board protocol. The experiments were conducted following the Declaration of Helsinki principles.