Natural Phaeosphaeride A Derivatives Overcome Drug Resistance of Tumor Cells and Modulate Signaling Pathways

In the present study, natural phaeosphaeride A (PPA) derivatives are synthesized. Anti-tumor studies are carried out on the PC3, K562, HCT-116, THP-1, MCF-7, A549, NCI-H929, Jurkat, and RPMI8226 tumor cell lines, and on the human embryonic kidney (HEK293) cell line. All the compounds synthesized turned out to have better efficacy than PPA towards the tumor cell lines listed. Among them, three compounds exhibited an ability to overcome the drug resistance of tumor cells associated with the overexpression of the P-glycoprotein by modulating the work of this transporter. Luminex xMAP technology was used to assess the effect of five synthesized compounds on the activation of intracellular kinase cascades in A431 cells. MILLIPLEX MAP Multi-Pathway Magnetic Bead 9-Plex was used, which allowed for the simultaneous detection of the following nine phosphorylated protein markers of the main intracellular signaling pathways: a universal transcription factor that controls the expression of immune-response genes, apoptosis and cell cycle NFκB (pS536); cAMP-dependent transcription factor (CREB (pS133); mitogen-activated kinase p38 (pT180/pY182); stress-activated protein kinase JNK (pT183/pY185); ribosomal SK; transcription factors STAT3 (pS727) and STAT5A/B (pY694/699); protein kinase B (Akt) (pS473); and kinase regulated by extracellular signals ERK1/2 (pT185/pY187). The effect of various concentrations of PPA derivatives on the cell culture was studied using xCelligence RTCA equipment. The compounds were found to modulate JNK, ERK1/2, and p38 signaling pathways. The set of activated kinase cascades suggests that oxidative stress is the main probable mechanism of the toxic action of PPA derivatives.


Introduction
Multiple drug resistance (MDR) of tumor cells is a major obstacle for the chemotherapy of malignant neoplasms. MDR is a system for protecting cells simultaneously from many drugs that differ in their chemical structure and mechanism of action on the cells [1]. Among the known molecular causes of drug resistance in human tumors is the currently with the survival and resistance of tumor cells, is an important area. The key signaling pathways associated with proliferation, apoptosis, differentiation, and also underlying the emergence of drug resistance, include MAPK signaling [22], PI3K/Akt/mTOR signaling [23,24], and NF-κB signaling [25]. For a detailed assessment of the effect of PPA derivatives on signaling pathways, LuminexxMAP technology is used in this work [26,27].
This article includes information on the synthesis of novel PPA derivatives, a biological evaluation of novel and known derivatives, including the measurements of the cytotoxic effects, and the ability to overcome drug resistance and modulate signaling pathways.
Secondary and cyclic amines were used to treat the mesylate. The reactions proceeded in the presence of TEA in acetonitrile at room temperature and led to amino derivatives 1-9 in a 16-43% yield. The ROESY spectrum of 5 showed a correlation between the methyl protons at δ 1.05 (H-7) and the proton at δ 3.14 (H-6), confirming the inversion of configuration of the C-6 atom (see supplementary materials). Products 1 and 10 were synthesized by our research group in 2017 and 2021 [19,21], respectively, and herein we show their new biological activities.

Assessment of the Cytotoxicity Level on Cultures of Tumor Cells
At the first stage, the level of cytotoxicity of the obtained compounds was assessed on a panel of tumor cultures of various histogenesis (Tables 1 and 2). The obtained compounds differ in the level of cytotoxicity, which, in most cases, is higher than the control substance etoposide. The synthesized compounds are especially active against cell cultures of tumors of the hematopoietic system, in particular, against multiple myeloma cells (NCI-H929) and acute T-cell leukemia (Jurkat). As the number of carbon atoms Scheme 1. Synthesis of target compounds. The atom numbering scheme is that of Clardy et al. [14] and differs from the IUPAC one.
Secondary and cyclic amines were used to treat the mesylate. The reactions proceeded in the presence of TEA in acetonitrile at room temperature and led to amino derivatives 1-9 in a 16-43% yield. The ROESY spectrum of 5 showed a correlation between the methyl protons at δ 1.05 (H-7) and the proton at δ 3.14 (H-6), confirming the inversion of configuration of the C-6 atom (see Supplementary Materials). Products 1 and 10 were synthesized by our research group in 2017 and 2021 [19,21], respectively, and herein we show their new biological activities.  Tables 1 and 2). The obtained compounds differ in the level of cytotoxicity, which, in most cases, is higher than the control substance etoposide. The synthesized compounds are especially active against cell cultures of tumors of the hematopoietic system, in particular, against multiple myeloma cells (NCI-H929) and acute T-cell leukemia (Jurkat). As the number of carbon atoms in the alkyl groups in the dialkylamine substituents increased, the activity of the substances decreased, so the IC50 was not checked for products 4,6,7.

Biological Activities
Assessment of the Cytotoxicity Level in Primary Cultures of Soft Tissue Sarcomas Furthermore, the cytotoxicity level of the several most active compounds in the primary cultures of soft tissue sarcomas (SMTs) was assessed [28], i.e., cultures obtained from samples of SMT tumors, and passed several times outside the body. It is believed that such cultures have a molecular profile that corresponds more with cells from the human body, in comparison with that of stable cultures of tumor cells [29] (Table 3). 1.6 ± 0.7 1.9 ± 0.5 7.8 ± 0.6 8 3.9 ± 0.8 2.3 ± 0.8 4.8 ± 0.5 9 2.9 ± 1.1 4.1 ± 1.1 13 ± 0.9 The synthesized compounds have cytotoxic activity against primary cultures of SMT of various histological forms.

Overcoming Drug Resistance of Tumor Cells
At the next stage, the cytotoxic activity of compounds 1, 2, and 9 (Scheme 2) was investigated on several pairs of cell lines that differ in the expression of P-gp and the method by which they were obtained. K562/i-S9 cells are the result of the ABCB1 transfection into K562 cells, which increases P-gp expression, but should not seriously affect other signaling pathways that could lead to drug resistance. K562/i-S9_Dox cells were obtained from K562/i-S9 by selection with low concentrations of doxorubicin to increase P-gp expression. Subline HBL-100/Dox is a derivative of HBL-100 cells, obtained by long-term selection with increasing concentrations of doxorubicin, which led not only to an increase in P-gp expression, but also to a change in the signaling pathways in this subline. Cells of the K562/i-S9_Dox subline express the highest level of P-gp: 94.7% ± 2.3 (p < 0.0001 relative to K562, and p = 0.04 relative to K562/i-S9). In cells of the K562/i-S9 line, the expression of P-gp is 80% ± 8.2 (p < 0.0001 relative to K562). At 0.7% ± 0.5, P-gp is practically not expressed in K562 cells (Figure 1). an increase in P-gp expression, but also to a change in the signaling pathways in this subline. Cells of the K562/i-S9_Dox subline express the highest level of P-gp: 94.7% ± 2.3 (p < 0.0001 relative to K562, and p = 0.04 relative to K562/i-S9). In cells of the K562/i-S9 line, the expression of P-gp is 80% ± 8.2 (p < 0.0001 relative to K562). At 0.7% ± 0.5, P-gp is practically not expressed in K562 cells ( Figure 1).
Scheme 2. The structure of compounds 1, 2, and 9. The described models were used to study the ability of the obtained compounds to overcome drug resistance associated with the overexpression of P-gp. As can be observed from Table 4, these cultures differ in resistance to dox, so K562/i-S9 and K562/i-S9_Dox cells are 11 (p = 0.0002) and 28 (p < 0.0001) times more resistant than K562, respectively. At the same time, K562/i-S9_Dox is 2.6 times more resistant than K562/i-S9 (p = 0.0002), which corresponds to a higher expression of P-gp in the K562/i-S9_Dox subline. The obtained compounds 1, 2 and 9 have a cytotoxic effect on cells, with the same efficiency, regardless of the overexpression of P-gp. However, we noted a slightly greater sensitivity of K562/i-S9 cells to 1 (p = 0.015) relative to K562, and a tendency towards increased sensitivity to 9 (p = 0.08) relative to K562. an increase in P-gp expression, but also to a change in the signaling pathways in this subline. Cells of the K562/i-S9_Dox subline express the highest level of P-gp: 94.7% ± 2.3 (p < 0.0001 relative to K562, and p = 0.04 relative to K562/i-S9). In cells of the K562/i-S9 line, the expression of P-gp is 80% ± 8.2 (p < 0.0001 relative to K562). At 0.7% ± 0.5, P-gp is practically not expressed in K562 cells ( Figure 1).
Scheme 2. The structure of compounds 1, 2, and 9. The described models were used to study the ability of the obtained compounds to overcome drug resistance associated with the overexpression of P-gp. As can be observed from Table 4, these cultures differ in resistance to dox, so K562/i-S9 and K562/i-S9_Dox cells are 11 (p = 0.0002) and 28 (p < 0.0001) times more resistant than K562, respectively. At the same time, K562/i-S9_Dox is 2.6 times more resistant than K562/i-S9 (p = 0.0002), which corresponds to a higher expression of P-gp in the K562/i-S9_Dox subline. The obtained compounds 1, 2 and 9 have a cytotoxic effect on cells, with the same efficiency, regardless of the overexpression of P-gp. However, we noted a slightly greater sensitivity of K562/i-S9 cells to 1 (p = 0.015) relative to K562, and a tendency towards increased sensitivity to 9 (p = 0.08) relative to K562.  The described models were used to study the ability of the obtained compounds to overcome drug resistance associated with the overexpression of P-gp. As can be observed from Table 4, these cultures differ in resistance to dox, so K562/i-S9 and K562/i-S9_Dox cells are 11 (p = 0.0002) and 28 (p < 0.0001) times more resistant than K562, respectively. At the same time, K562/i-S9_Dox is 2.6 times more resistant than K562/i-S9 (p = 0.0002), which corresponds to a higher expression of P-gp in the K562/i-S9_Dox subline. The obtained compounds 1, 2 and 9 have a cytotoxic effect on cells, with the same efficiency, regardless of the overexpression of P-gp. However, we noted a slightly greater sensitivity of K562/i-S9 cells to 1 (p = 0.015) relative to K562, and a tendency towards increased sensitivity to 9 (p = 0.08) relative to K562. Similar results were obtained for the HBL-100 and HBL-100/Dox line pairs (Table 5). HBL-100/Dox cells were 556 times and 23 times more resistant to doxorubicin and etoposide, respectively (p < 0.0001), but there was no difference in resistance to 1 and 2 between cells. Only for 9 is there a tendency towards a slightly higher resistance of HBL-100/Dox cells compared to HBL-100. Thus, the compounds studied in this work overcome drug resistance associated with P-gp overexpression, which has been shown in several pairs of sensitive and resistant tumor cell lines.

Interaction of the Obtained Compounds with the P-Glycoprotein
We investigated the potential interaction of some synthesized compounds with P-gp (Table 6, Figure 2) in classical experiments on the release of rhodamine Rd123 (fluorescent agent, P-gp substrate). After 20 min of incubation with Rd123, almost 100% of the cells of the K562/i-S9_Dox line are stained, but after removing Rd123 from the mixture, after about 30 min, the number of luminous cells drops to 17%, because P-gp removes Rd123 from the cells. However, the addition of verapamil inhibits this process, and 90% of the cells remain luminous. All the investigated compounds act in the same direction, while substances 2 and 9 act with an efficiency similar to that of verapamil: the number of luminous cells is 82% and 80%, respectively. Compound 1 has a weaker interaction with P-gp than verapamil; the number of luminous cells is 73%. The experiments show that the synthesized compounds modulate the activity of P-gp, significantly reducing it, with approximately the same efficiency as the classical inhibitor verapamil.

Influence of Phaeosphaeride A Derivatives on the Key Signaling Pathways
LuminexxMAP technology was used to assess the effects of compounds 1, 2, 5, 9, 10 and PPA (the most active and the most inactive among all PPA derivatives) on the activation of intracellular kinase cascades in A431 cells. A431 cells are characterized by a high inducibility of transcription factors STAT3 and STAT5. In turn, phaeosphaeride A and its derivatives are probably inhibitors of the STAT3 gene [14]. Therefore, the A431 cell line was chosen as a model for searching for possible targets for the action of phaeosphaeride A derivatives. Table 7 presents IC50 data for phaeosphaeride A derivatives on A431 cells.  The experiments show that the synthesized compounds modulate the activity of P-gp, significantly reducing it, with approximately the same efficiency as the classical inhibitor verapamil.

Influence of Phaeosphaeride A Derivatives on the Key Signaling Pathways
LuminexxMAP technology was used to assess the effects of compounds 1, 2, 5, 9, 10 and PPA (the most active and the most inactive among all PPA derivatives) on the activation of intracellular kinase cascades in A431 cells. A431 cells are characterized by a high inducibility of transcription factors STAT3 and STAT5. In turn, phaeosphaeride A and its derivatives are probably inhibitors of the STAT3 gene [14]. Therefore, the A431  We used the MILLIPLEX MAP Multi-Pathway Magnetic Bead 9-Plex-Cell Signaling Multiplex Assay, which allows the simultaneous detection of nine phosphorylated proteins that are markers of the main intracellular signaling pathways (phosphorylation site is indicated in parentheses), a universal transcription factor that controls the expression of immune response genes; apoptosis and cell cycle NFκB (pS536); cAMP-dependent transcription factor (CREB (pS133); mitogen-activated kinase p38 (pT180/pY182); stressactivated protein kinase JNK (pT183/pY185); ribosomal SK; transcription factors STAT3 (pS727) and STAT5A/B (pY694/699); protein kinase B (Akt) (pS473); and kinase regulated by extracellular signals ERK1/2 (pT185/pY187).
The effect of various concentrations of PPA derivatives on the cell culture was studied using xCelligence RTCA equipment (AceaBioscience, San Diego, CA, USA). The concentration closest to the IC50 (10 µM) was chosen as the experimental concentration. The selected concentration of phaeosphaeride A derivatives allows for the assessment of the early effect of compounds on cells and compares their effectiveness. In this experiment, the signaling pathways Akt (pSer473), p70S6K (pThr412), NFkB (p65-pSer536), and CREB (pSer133) were not activated ( Table 8). The stress-activated signaling pathways ERK/MAP kinase 1/2 (pThr185/pTyr187), JNK (pThr183/pTyr185), and p38 (pThr180/pTyr182) were activated for these compounds. The fluorescence intensity of the active forms ERK1, JNK, and p38 increased by 1.5-3 times under the action of 1, 2, and 9, i.e., substances with the greatest cytotoxic effect (Figure 3). Compound 2 exhibited the maximum degree of activation. These pathways were only slightly activated under the action of 5, 10, and PPA, which showed a significantly lower cytotoxic effect on tumor cultures. All compounds, with the exception of 2, did not significantly affect the activation of the transcription factors STAT5A/B (pTyr694/699) and STAT3 (pSer727). Compound 2 slightly activated these proteins, probably through the strong activation of the ERK1, JNK, and p38 proteins [30].
The set of activated kinase cascades suggests that oxidative stress is the main mechanism of the toxic action of PPA derivatives, leading to cell death [31][32][33]. In further work, it would be expedient to evaluate the markers of oxidative stress under the action of PPA derivatives and the kinetics of their activation on various cell lines.

Chemistry
The 1H NMR spectra were acquired on a Bruker AVANCE III 400 MHz NMR spectrometer in CDCl 3 . Optical rotations were acquired on an Optical Activity Polaar 3005 Polarimeter using a 2.5 cm cell with an Na 589 nm filter and the concentration of the samples was denoted as c. The mass spectra data were acquired on a Thermo Scientific TSQ Quantum Access Max Mass spectrometer. High-resolution mass spectra (HRMS) were acquired on a LTQ OrbitrapVelos spectrometer and on a Bruker MicrOTOF. FTIR spectra were acquired on a Shimadzu IR Affinity-1 spectrometer. Organic solvents used were dried by standard methods when necessary. Commercially available reagents were used without further purification. All reactions were monitored by TLC with EMD/Merck KGaA silica gel-coated plates, with visualization by UV light and by charring with 0.1% ninhydrin in EtOH. Column chromatography was performed using Merck 60 Å 70-230 mesh silica gel. The optical density was determined using a Multiskan FC spectrophotometer (Thermo Scientific, Waltham, MA, USA) at a wavelength of 540 nm when using the MTT assay.

MTT Test
Cells MCF-7, PC-3, A549, and HCT-116, and primary cultures of STSwere seeded in 96-well plates, 6 × 10 3 cells per well in 90 µL of culture medium. Suspension cultures were seeded at a concentration of 25 × 10 3 cells per well in 135 µL of medium. The test substances were dissolved in DMSO (Applichem, Council Bluffs, IA, USA) until a concentration of 1 × 10 −2 M was obtained; for the subsequent dilution of the substances, a serum-free culture medium was used. The next day, for adhesive cultures, and on the same day, for suspension cultures, substances were added to the plates in a volume of 10 µL (adhesive cultures) and 15 µL (suspension cultures) per well at various concentrations; each concentration was added in 3-4 repetitions. The cells were cultured in the presence of substances for 48 h. Then, MTT reagent (PanEko, Moscow, Russia) was added to the wells at a concentration of 5 mg/mL in a volume of 20 µL per well. After 1.5-2 h, the medium with the reagent was removed, and the precipitate of the crystals of formazan was dissolved in 60 µL of DMSO. The optical density was determined using a Multiskan FC spectrophotometer (Thermo Scientific, Waltham, MA, USA) at a wavelength of 540 nm.
Based on the results of the experiments, a schedule was plotted of the dependence of the percentage of surviving cells (from control) on the concentration of the compound, and the sought concentration IC50 was calculated. The experiments were carried out in 3-4 times.

Flow Cytometry
To assess the expression of P-gp in K562, K562/i-S9, and K562/i-S9_Dox cells, 5 × 10 5 cells were fixed in 4% formaldehyde solution, then washed twice in PBS, after 30 min they were incubated with monoclonal antibody FITC Mouse Anti-Human P-glycoprotein, clone 17F9 (BD Biosciences, San Jose, CA, USA) was washed twice to remove antibodies in PBS. The number of luminous cells was analyzed using a Becton Dickinson FACScan flow cytometer. The results were analyzed using Diva and CyFlodgic software.
Evaluation of the Release of Rhodamine 123 (Rd123) from Cells with an Overexpression of P-Glycoprotein To assess the interaction of the investigated compounds with P-gp, we used the test for the release of rhodamine 123 (Rd123) (Sigma-Aldrich, St. Louis, MO, USA) from cells with P-gp overexpression K562/i-S9_Dox. The principle of the test is as follows: Rd123 is a fluorophore and a known P-gp substrate. During incubation, Rd123 enters the cells by simple diffusion, and then Rd123 begins to interact with P-gp, which releases it from the cell. Cells in which Rd123 is present are stained and can be detected by flow cytometry. The more efficiently P-gp works, the less stained cells in the population remain over time. If inhibitors or modulators of P-gp activity are added to the cells, then the cells remain stained; if substances do not interact with P-gp, then the cell population loses its luminosity.
The experiment used K562/i-S9_Dox cells with P-gp overexpression. Cells (5 × 10 5 per point) were incubated with Rd123 at a concentration of 3.5 µg/mL for 20 min in serum-free RPMI1640 medium. Furthermore, the efficiency of Rd123 entry into cells was assessed. Subsequently, the cells were washed once from Rd123 in a PBS solution, and the precipitate was divided into several parts; the cells were placed in a clean serum-free medium to assess the efficiency of Rd123 release, as well as in a serum-free medium with the addition of the following compounds at a concentration of 20 µM: verapamil (a classic inhibitor of P-gp) and test compounds 1, 2, and 9. Incubation with substances took place for 35 min in 1 mL of serum-free medium. Then, the cells were precipitated, washed with PBS solution, and the % of stained cells in the population were assessed on a flow cytometer.

Statistical Analysis
The results were presented as the mean ± SD. GraphPad Prism 6.0 was used for statistical analysis. The statistical significance was evaluated by unpaired t-test.

Evaluation of the Effect of Phaeosphaeride A Derivatives on Key Signaling Pathways
To determine the integral cellular response using the xCelligence RTCA equipment (AceaBioscience, Santa Clara, CA, USA), 20 thousand cells were inoculated into a well of a specialized plate allowing the determination of the cell index in real time and cultured in complete growth medium. The next day after the seeding, the medium was replaced with an experimental one (compounds at concentrations of 50, 20, 10, 5, 2, 1, and 0.5 µm). The cell index was monitored for 3 days after the addition of the xenobiotic. The concentration closest to the IC50 (10 µM) was chosen as the experimental concentration.
We used a MILLIPLEX MAP Multi-Pathway Magnetic Bead 9-Plex-Cell Signaling Multiplex Assay kit (Merck/Millipore), according to the manufacturer's protocol. A431 cells were cultured in DMEM with 10% FBS in 24-well plates. Compounds were added to the subconfluent cells at a concentration of 10 µM in DMEM medium for 1 h. Each compound was added in 5 experimental repeats. Cell lysates were prepared with lysis buffer (Millipore) with protease inhibitors (Complete, Roche); the protein concentration in the lysates was determined by the Lowry method. In samples equalized in protein concentration, the active forms of kinase cascades were determined using a Bio-Plex 200 fluorescent analyzer (Bio-Rad).

Conclusions
Thus, the obtained PPA derivatives can overcome the drug resistance of tumor cells associated with the overexpression of P-glycoprotein by modulating the work of this transporter, which makes them promising in terms of creating new anticancer drugs aimed at treating resistant malignant neoplasms. A potential mechanism of the toxic effect of PPA derivatives associated with oxidative stress has also been identified. In addition, the effective two-stage synthesis of the presented compounds based on natural substances makes the development of this project even more attractive.