Targeted inhibition of CD47 expression in Human cancer cell lines on treatment with Cardiac Glycosides

doi:10.21203/rs.3.rs-2361555/v1

Download PDF

Short Report

Targeted inhibition of CD47 expression in Human cancer cell lines on treatment with Cardiac Glycosides

https://doi.org/10.21203/rs.3.rs-2361555/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The interest has grown in therapeutically targeting Hypoxia-inducible factor 1α (HIFα), which directly stimulates the expression of cluster of differentiation 47 (CD47) on the cell surface that suppresses phagocytosis in cancer cells. Increased expressions of CD47 and counter interaction with signal regulatory protein alpha (SIRPα) enable cancer cells to avoid cell-mediated cell destruction. On the other side, decreased expression of SIRPα was reported to promote growth. Thus, colossal concern and curiosity arise in identifying the molecular mechanism behind this suppressive effect of CGs in controlling cancer cells. We first report that cardiac glycosides (CGs) inhibit HIFα and CD47 in human breast, lung, and liver cancer cells. Furthermore, our analysis through TCGA (The Cancer Genome Atlas) data shows that these three potential genes correlate with poor survival in breast, lung, and liver cancers. Our molecular docking and molecular dynamic simulations studies demonstrated the interaction of the mentioned CGs with target proteins and identified the stability. Importantly, unlike any other anti-CD47 antibodies, the antitumor activity of CGs has been evaluated in many cancers with no hematologic toxicities. These findings would help to develop clear strategies to target CD47 and SIRPα interaction through HIF1-α inhibitors to promote phagocytosis.

HIF1-α

CD47

SIRPα

Cardiac glycosides

Cancer therapy

Phagocytosis

Molecular docking

Molecular dynamic simulations

Therapeutic targets

The cancer cells appear to disrupt immune responses by provoking a milieu that helps immunologic forbearance. This immune-privileged status is accomplished by representing existing homeostasis methods that alleviate immune responses with those that avert prolonged inflammation and autoimmunity. Hence, identifying the human immune system's failure against tumor cells remains critical in cancer drug discovery. One significant explanation could be that tumors further appoint the loss of immune checkpoint pathways to avoid immune cells (Chen & Mellman, 2013). Investment and progression of cancer depend on various hallmark features, including constant cell proliferation, resistance to drugs for evading apoptosis, facilitating replicative mortality, and invasion, mainly by creating an immune-suppressive environment (Hanahan & Weinberg, 2011). A century ago, tumor immune surveillance was discovered; however, it has been recognized that only adaptive immunity plays a significant role in the anticancer mechanism by eradicating cancer cells (Swann & Smyth, 2007). The role of innate immunity through macrophages in promoting phagocytosis via tumor pathogenesis was recognized recently (Chao, Majeti, et al., 2012). Macrophages exhibit phagocytic mechanisms by triggering the "eat" signal to maintain cell clearance and pathogen defense mechanisms (Moraes et al., 2017). Recent studies have reported that tumor cells escape from macrophage-mediated phagocytosis by generating anti-phagocytic signals ("do not eat") through CD47 (Chao, Weissman, et al., 2012). CD47 expresses in various hematological and solid tumors, including breast, liver, and lung, in several other tumors, including adverse prognostic indicators for cell survival(Petrova et al., 2017). As a transmembrane glycoprotein, CD47 up-regulates in tumor cells and binds with SIRPα on the macrophages to inhibit phagocytosis (Tsai & Discher, 2008).

Exposure of cancer cells to reduced O₂ could lead to the activation of hypoxic-induced factors, known as heterodimeric transcriptional activators, specifically HIF1-α, HIF-2α, and significantly expressed HIF-1β subunits (Semenza, 2016). Among these subunits, HIF1-α and HIF-2α are known to regulate at high levels in almost all cancers (Semenza, 2010). Increased synthesis and decreased degradation always lead to the overexpression of HIF1-α protein levels in cancers; the overexpression of HIF1-α results in advanced cancer patients (Rankin & Giaccia, 2008). Previous studies have indicated that CGs, such as Digoxin, Ouabain, and Proscillaridin A, can effectively inhibit the function of HIF1-α and HIF2-α in tumor xenografts (Zhang et al., 2008). Cancer cells use hypoxia to show resistance to anticancer drugs and enhance aggressive growth (Triner & Shah, 2016). HIFs use various genes, including p53, Beclin, MDR1, and Bcl-2, to induce therapy resistance for cancer cells. Consequently, inhibition of HIFs could lead to cell death through apoptosis or necrosis by attenuating several signaling pathways(Matsumoto et al., 2010). Besides all these efforts from the past two decades, HIF inhibitors accomplish not approved for treating cancers (Kim et al., 2017)

Stimulatingly, recent studies reported that HIF1-α expression leads to the activation of the CD47 gene in breast cancer cells (Zhang et al., 2015). Small molecules targeting this CD47-SIRPα axis and HIF1-α could lead to the discovery of a novel target for cancer treatment. CD47-SIRPα antagonists could encourage phagocytosis and antigen uptake, triggering acquired responses and connecting innate and adaptive immunity (Veillette& Chen, 2018). As CGs are reported as inhibitors for HIF1-α and are responsible for the activation of CD47, the current study is intended to identify the role of HIF1-α inhibitors in CD47-SIRPα immune checkpoint signaling. Therefore, CD47-SIRPα obstructive agents may help improve the anticancer effects of CGs. Previous studies from our lab have reported the anticancer activities of CGs (PS, ST, and LC) in breast, lung, and liver cancer cells(Reddy, Ghosh, et al., 2020). In the present study, we have hypothesized and initiated an approach to identify the role of CGs in the CD47-SIRPα immune checkpoint pathway along with their inhibitory properties against HIF1-α in vitro and in silico studies.

2.1 Cell culture

The following human cancer cells were used for the study, Human breast cell lines (MCF-7), non-small lung cancer cell lines (A549), and Hepatocellular carcinoma cells (HepG2) were cultured as previously described [18]. The cells were maintained at 37°C in a 5% CO₂ environment.

2.2 Reagents

Alexa flour 488, ProLong Gold Antifade Mountant, and TRIzol were obtained from Invitrogen (Massachusetts, USA). PS was purchased from Toronto Research Chemicals (North York, Canada), and ST; LC were obtained from Sigma-Aldrich (St. Louis, U.S.A.). The primary antibodies used for the present study were anti-CD47, anti-SIRPα (Elabsciences), and anti-HIF1-α procured from Elabsciences (Texas, USA). GAPDH obtained from Cell signaling technology (Massachusetts, USA).

2.3 RNA extraction and qRT-PCR

Total cellular RNA was extracted using a TRIzol reagent following the manufacturer's instructions. Briefly, MCF-7, HepG2, and A549 cells were treated with lethal doses (we aim to identify the expressions below the threshold for lethality; we have used the highest concentrations) of PS, ST, and LC for 24 hrs. Cells were then harvested, and the total RNA. It was isolated by following the manufacturer's instructions. An equal concentration of total RNA (2µg) was used for cDNA synthesis. The cDNA synthesis was done using a verso cDNA kit (Thermo Fisher Scientific, India). Real-time PCR performed with SYBR green master mix (Origin chemicals, India). The following primers were used in the present study, HIF1-α, forward: CCTGCAACTGAATCAAGAGGTGC, reverse: CCATCAGAAGGACTTGCTGGCT CD47 forward: AGAAGGTGAAACGATCATCGAGC, reverse: CTCATCCATACCACCGGATCT, and for GAPDH forward: AACGGGAAGCTTGTCATCAATGGAAA, reverse GCATCAGCAGAGGGGGCAGAG. GAPDH is the internal control (housekeeping gene) to normalize RNA in each reaction. The results were expressed as relative abundance and analyzed using the 2–∆∆Ct method. All the experiments are repeated thrice to check the consistency (n = 3).

2.4 Immunoblotting

Following specific treatment incubations, the total cell lysate was obtained and lysed in RIPA buffer (Thermo Fisher Scientific) with a protease inhibitor cocktail (Roche). Estimating total protein concentration was performed using the Bradford protein estimation assay, and an equal concentration of 30 µg loaded in each well of 12% SDS-PAGE. The gel was then transferred to the PVDF membrane with the Trans-Blot® Turbo™ Blotting System (Bio-Rad). The membrane was then blocked with 5% BSA with primary antibody at 4°C overnight. The membrane was washed with a secondary antibody for 1 hr at room temperature. GAPDH was used as the loading control to identify equal loading. C-Digit (LI-COR Lincoln, NE, USA) Chemiluminescent scanner was used to detect. All experiments were repeated three times (n = 3) to ensure consistency.

2.5 Immunofluorescence studies

Around 0.3x10⁶ cells were seeded per well on top of coverslips in a six-well plate and incubated overnight. The cells were then treated with lethal test compounds and incubated for 24 hrs. Then the cells were fixed with 4% paraformaldehyde and 0.1% Triton X for 20 min. The cells were then blocked with 5% BSA and incubated with the primary antibody overnight at 4°C. The coverslip was washed with 1X TBS and set in a secondary antibody (Alexa flour 488) for 2 hrs at room temperature. Then the coverslip was cleaned and stained with 0.1 µg/ml concentration of DAPI, placed on top of the glass slide containing ProLong Gold Antifade Mountant, and sealed with wax. The slides were observed under a fluorescent microscope with 40X Magnification (Leica DMI3000, Wetzlar, Germany).

2.6 Molecular docking

Structures of HIF1-α (3KCX), CD47 (2JJS), and SIRP-alpha (2WNG) were downloaded from RCSB PDB. Missing amino residues and incomplete side chains have been modeled using a modeler (v9.22) for the structure of HIF1-α (3KCX). Protein structures have been prepared, optimized, and minimized using the protein preparation wizard of Schrödinger maestro at pH 7 (+/- 2). Extra water molecules and heteroatoms have removed. All ligands have been prepared, and possible stereoisomers have been generated at pH 7(+/-2) using the LigPrep of the maestro. The grid has been developed around the receptor site to define the docking site. For 3KCX, an already present inhibitor (Clioquinol) has been selected to represent the docking site. The receptor grid for CD47 and SIRPαhave been defined by reported essential interactive residue. All stereoisomers of three ligands were docked with extra precision to HIF1-α, CD47, and SIRP-alpha, one by one using Glide (Schrödinger). After docking, the generated glide and docking scores were analyzed along with the docked structure. Each protein's best-docked structure with each ligand has been exported for further molecular dynamics analysis among all complexes.

2.7 Molecular dynamic simulations

A molecular dynamics simulation study of nine protein-ligand complexes with three different ligands was performed using Gromacs 5.1. Topologies of all proteins were generated using the latest force field, gromos96 54a7. While ligands topologies were generated using the PRODRG server. Further, ligands topologies were included in protein topology before salvation. All protein-ligand complexes were solvated with an spc216 water model in the triclinic box. Hereafter, adding corresponding Na+/Cl- counter ions neutralized solvated systems. Subsequently, neutralized systems were minimized using the steepest descent minimization method, and position restrain was applied for ligand molecules in all complexes. After minimization, equilibration was performed using NVT and NPT ensembles for the 100ps time steps. Unbound structures of all three proteins were also prepared for MD simulation using a similar force field, a salvation system used for bound proteins. Finally, MD was performed for all nine equilibrated protein-ligand complexes and all three unbound proteins for 30 ns. After the MD simulation, Root Mean Square Deviation analysis (RMSD), Root Mean Square Fluctuation analysis (RMSF), and Hydrogen bond occupancy analysis were performed for further research. RMSD and RMSF analysis for 2WNG protein were carried out for a ligand domain of 112 amino acids.

2.8 Gene expression analysis

The pre-processed, FPKM-normalized gene expression data from TCGA cohorts (breast cancer, BRCA; liver cancer, LIHC; and lung cancer, LUAD, LUSC) were downloaded from Broad GDAC Firehose (https://gdac.broadinstitute.org/; last accessed Dec 2019; data version 2016_01_28). Kaplan-Meier analyses were performed using GraphPad Prism 7.0 (GraphPad Software, La Jolla, CA, USA). LogA log-rank test was used to assess significance.

2.9 In-silico ADMET prediction

The ADMET properties of PS, ST, and LC were predicted using the SWISSADME database. The root of prediction software is the associations of physicochemical properties such as log P and solubility with the pharmacokinetics of the CGs. The software calculated the physicochemical, pharmacological properties, and toxic parameters of CGs.

2.10 Statistical analysis

All the values in the present study were expressed as mean ± SD. Statistical analyses were either performed using GraphPad Prism 7.0 or Origin pro9.0. The student t-test was used to determine the significant differences between the two independent groups. P values were considered, and p < 0.5 is considered statistically significant. All the experiments were repeated three times (n = 3).

3.1. CGs inhibit the expression of Hypoxia Inducible Factor 1α

In this study, we investigated whether the selected CGs inhibit the expression of HIF1-αand was assessed through real-time qPCR and western blotting analysis. For the experiment, an approximate number of 1*10⁵ cells (MCF-7, A549, and HepG2) were plated and incubated with lethal doses of PS (MCF-7- 100nm, A549-100nm, and HepG2-100nm), ST (MCF-7-2µM, A549-1µM, and HepG2-2.5µM), and LC (MCF-7-1.2µM, A549-0.16µM, and HepG2-0.7µM). As the lethal doses should inhibit ~ 90% of cancer cells' growth, all the subsequent experiments were conducted with the indicated concentrations. Destructive dose treatment of these compounds resulted in decreased expressions of HIF1-α in MCF-7, A549, and HepG2 cells, as shown in Fig. 1A. Furthermore, to validate this effect, we isolated the total cell lysate and performed a western blot using a primary rabbit antibody of HIF1-α. PS, ST, and LC were shown to decrease the protein levels of HIF1-α. A dramatic reduction in the gene and protein expressions of HIF1-α was identified in the treatment conditions compared to controls (untreated), as shown in Fig. 1B. The decreased expression of HIF1-α suggested that CGs can block the aggressive growth of cancer cells. The present study demonstrates that CGs are effective inhibitors for HIF1-α and possesses new therapeutic possibilities to treat cancers.

3.2. CG.s inhibits the expression of CD47 in cancer cells

We performed real-time qPCR and western blotting analysis to study the cytotoxic effect of CGs through inhibiting CD47 directly. It has been previously reported that blocking CD47 activity with anti-CD47 antibodies increased the phagocytosis of cancer cells (Tseng et al., 2013). With this basis, the current study was designed to identify whether CGs could inhibit the activity of CD47 in cancer cells from promoting phagocytosis. Subsequently, HIF1-α is a known activator for CD47; we intended to check whether inhibition of HIF1-α could have any role in expression variations of CD47. MCF-7, A549, and HepG2 cells were pre-treated with PS, ST, and LC for 24 hrs and isolated the total RNA and cell lysate for further experiments.

The mRNA levels of CD47 showed consistent downregulation. CGs-induced MCF-7, A549, and HepG2 cells showed downregulation of CD47 consistently in Fig. 1C. To confirm that CGs inhibit CD47 expression, we confirmed protein expressions using western blot analysis. The obtained results demonstrated that the expression of CD47 was reliably inhibited in CGs treatment. These results were also stable with our RT-PCR analysis, which revealed that the results were consistently the same. Through scrutinized analysis, we have shown that cancer cells with decreased CD47 expression are more likely to be phagocytosed.

3.3. Immunofluorescence analysis

We then sought to explore different backgrounds of the cell under treatment and control conditions through immunofluorescence studies. In this immunofluorescence analysis, treatment with PS, ST, and LC resulted in a significant decrease in the expressions of HIF1-α and CD47. The results suggested that HIF1-α, CD47, and SIRPα were widely distributed around the nucleus and changed the specific localization compared to control (Untreated) cells. Here we have shown the localization of HIF1-α in MCF-7 cells (Fig. 2), A549 (Supplementary Fig. 1), and HepG2 cells (Supplementary Fig. 2), as well as intracellular distribution of HIF1-α, and CD47, in cancer cells was assessed by immunofluorescence using specific antibodies with appropriate controls. The target proteins have been identified within the membrane, cytoplasm, and nucleus of these cells of the breast, lung, and liver cancer cells. These data suggest that HIF1-α and CD47 may perform a nuclear function in a range of cancer cells and can be targeted for drug discovery.

3.4. Molecular docking

The PS, ST, and LC ligand structures were docked with the target proteins of HIF1-α, CD47, and SIRPα. The HIF1-α–PS complex was stabilized by two hydrogen bond interactions with THR182 and GLN189 residues with a -7.219 glide score. In contrast, HIF1-α–ST complex alleviated one hydrogen bond with GLN189 residue with a -5.296 glide score. The HIF1-α–five hydrogen bonds alleviate the LC complex, and residues, namely, SER77, TYR79, LYS92, GLN133, and ASP187, were actively interacting with the ligand molecule and formed hydrogen bonds with a glide score of -8.786. The Docking score and hydrogen bond interactions for three compounds with three ligands were presented in Table 1. The protein-ligand complex is shown in Fig. 3.

Table 1

Residues interacting from CD47, SIRPα, and HIF1α with Peruvoside, Strophanthidin, and Lanatoside C with LibDock scores and hydrogen bond forming residues.
Compound name	Target protein	Residues forming H-bond	Residues within 4⁰ distance
Peruvoside	CD47	THR7, SER9, GLY92	PHE4, LYS6, LYS8, VAL10, PHE12, VAL19, ILE21, PRO22, CYS23, PHE24, TRP40, HIS90, THR91, ASN93, TYR94, THR95, CYS96, GLU97
Peruvoside	SIRPα	LEU30, GLY34, PRO35, VAL33, GLN52, LYS53, ARG69, ASN71, LYS93	VAL27, ILE31, PRO32, ILE36, TYR50, ASN51, GLU54, SER66, LYS68, GLU70, PHE74
Peruvoside	HIF1α	GLU225, GLY337	PHE111, PHE224, TYR 228, PRO229, PRO231, GLN241, PRO333, GLN334, VAL336, PRO338, LEU340, ASN341
Strophanthidin	CD47	THR7, LYS8, SER9, THR95	PHE4, LYS6, VAL10, ILE21, PRO22, CYS23, PHE24, TRP40, GLY92, ASN93, TYR94, CYS96
Strophanthidin	SIRPα	GLN52, SER66, ASN71	LEU30, VAL33, GLY34, PRO35, ILE36, THR67, LYS68, ARG69, PHE74, LYS93
Strophanthidin	HIF1α	TYR228, PRO229	PHE224, TYR230, PRO231, GLN241, MET319, ILE322, GLU323, LEU340, ILE344
Lanatoside C	CD47	THR7, LYS8, VAL10, HIS90, THR91, GLY92, TYR94, CYS96	PHE4, LYS6, SER9, PHE12, ILE21, PRO22, CYS23, TRP40, PHE42, LYS43, ASP62, SER89, ASN93, THR95
Lanatoside C	SIRPα	LEU30, GLY34, GLN52, PHE57, THR67, ARG69, ASN71.	ILE31, PRO32, VAL33, PRO35, ILE36, TYR50, GLY55, HIS56, PRO58, VAL60, THR62, GLU65, SER66, LYS68, MET72, PHE74
Lanatoside C	HIF1α	GLU57, GLU59, THR302	ASN58, MET275, ALA300, PRO301, PRO303, LYS311, GLN314

Furthermore, we extended our study to identify the interactions of these compounds with CD47. The CD47–PS complex is stabilized by three hydrogen bonds with a -4.605 glide score, with LYS43, ASN93, and THR95 actively involved in hydrogen bond formation. The CD47–ST complex is alleviated by two hydrogen bonds, shown with a -2.714 glide score. Residues such as SER89 and HIS90 were actively involved in hydrogen bond formation. The CD47–LC complex is highly stable compared to all the other ligands interacting with CD47. This complex has shown the six strongest hydrogen bonds with the target molecule. The glide score of this interaction was identified as -7.393. Residues such as ASP62, SER85, ASP86, SER89, THR91, and THR95 were actively involved in hydrogen bond formation (Fig. 3). Many other amino acids have interacted with ligands, tabulated the residues in Table 1. Then we extended our study to identify the interactions of these compounds with SIRPα. The results were as follows: SIRPα has shown an excellent binding affinity with PS by forming three hydrogen bonds with GLY34, GLN52, and ARG59 residues. The glide score of this interaction was identified as -5.24 for this interaction.

In contrast, the SIRPα–ST complex is alleviated by one hydrogen bond with a glide score of -3.648. LYS93 was actively involved in hydrogen bond formation. The SIRPα–LC complex is highly stable compared to all other ligands interacting with SIRPα. The glide score of this interaction was identified as -8.786. Residues such as LEU30, THR62, GLU65, SER66, and ARG69 were actively involved in hydrogen bond formation (Fig. 3). Other than hydrogen bond-forming residues, many other amino acids interacted with these ligands and model scores. All those amino acids are shown in Table 1.

3.5. Molecular dynamic simulations

Molecular dynamics simulation studies were performed to analyze the conformational stability of HIF1-α CD47 and SIRPα with PS, ST, and LC. Comparative root means square deviation (RMSD) analysis of bound protein compared to unbound protein reveals the complex's stability. Root Mean Square Fluctuation (RMSF) analysis helps find more or less amino acid fluctuation around the binding domain. The dynamics of hydrogen bond formation help determine the significant interaction between ligand and protein required to stabilize the complex. For LC, the average number of hydrogen bonds per frame is 3.013. Key residues forming hydrogen bonds as a donor are GLN189 (28%), SER170 (13.9%), ASN137 (22.5%), TYR79 (21.5%), and as acceptor are GLN133 (10.9%), GLU91 (11.1%). For PS, the average number of hydrogen bonds per frame is 1.123. Key residues forming hydrogen bonds as a donor are GLN189 (22.1%) and GLN133 (12.5%), and acceptors are none. For Strophanthidin, the average number of hydrogen bonds per frame is 0.201. Key residues forming hydrogen bonds as the donor are TYR79 (18.5%) and ASP187 (43%, 24.3%) as the acceptor. Native HIF1 exhibits instability in RMSD throughout the simulation. Nevertheless, HIF1-α with ligand complexes is stable. ST has deference instability due to more deviation in the first ten ns. Even though it is stable later. However, no significant difference in RMSF was observed among any complex other than exceptional loop regions. Radiuses of gyration represent the compactness of protein throughout the simulation. Native protein is less compact than complex. Among all complexes, HIF1-α _ PS has uniformity in compactness than the other two complexes. Protein complexes with LC and PS ligands do not show consistent RMSD graph behavior than unbound protein and protein ST complexes. Root mean square fluctuation (RMSF) of protein-bound with ST ligand fluctuates less than another complex. CYS96 and PHE12 of CD47 occupied hydrogen bonds with ST ligand for > 57% and > 17% simulation time. Hydrogen bond occupancy between LC and protein residue TYR94 and CYC96 is > 42% and > 23%, respectively. PS Ligand formed a hydrogen bond with TYR94 and PHE4 residues with > 34% and > 21% occupancy, respectively.

Next, we performed simulations for SIRPα. The results suggested that RMSD analysis of the ligand-binding domain of the SIRPα protein-ligand complex does not show a significant difference with unbound protein. Nevertheless, a comparative RMSF examination of the graph reveals that a specific domain of nearly ten amino acids between 65 to 74 considerably drops the RMSF value for LC and PS bounded proteins. For LC specified protein domain, almost ten amino acid sequences from 26 to 36 fluctuate less. PHE57 amino acid participates in hydrogen bond formation with LC and PS ligands with > 22% and > 49% occupancy, respectively. GLU70 and ASN71 build hydrogen bonds with > 32% and > 18% occupancy with ST ligand (Fig. 4).

3.6. Analysis of TCGA dataset mRNA expressions of HIF1-α, CD47, and SIRPα in human breast, lung, and liver cancers

Over expression of HIF1-α and CD47 is associated with decreased survival in several cancers, including breast, lung, and liver. Subsequently, down-regulation of SIRPα resulted in increased cancer proliferation in breast, liver, and pancreatic cancers. Based on the data obtained from the TCGA-RNA-Seq dataset, we analyzed the predictive value of HIF1-α, CD47, and SIRPα in breast, lung, and liver cancer patients with overall survival (the length of time from either the diagnosis) and progression progression-free survival rate length of time during and after treatment of cancer) (Supplementary Fig. 6). The present study used the quantile version of the graphs to obtain consistency. The low (blue) is quantile 1 (lowest 25%), and the high (red color) is quantile 4 (highest 25%). Here the high and low are based on the quantile of the FPKM expression values. TCGA data show that the three genes correlate with poor survival in breast, lung, and liver cancers, suggesting the potential benefit of inhibiting the HIF1-α, CD47, and SIRPα in these cancers

3.7. In-silico ADMET toxicity screening of CGs

The Swiss ADME was used for the in silico study of ADMET (absorption, distribution, metabolism, elimination, and toxicity) and drug-likeness of PS, ST, and LC. Here are the predicted Pharmacokinetics (Table 2) and drug-like properties of CGs and their degradation products represented in Table 3 and Supplementary Fig. 7. The drug-likeness parameters of CGs suggest that the pink area represents the optimal range for each property (lipophilicity: XLOGP3 between 0.7 and + 5.0, size: MW between 150 and 500 g/mol, polarity, TPSA between 20 and 130 Å, 2 solubility, log S not higher than 6, saturation: the fraction of carbons in sp3 hybridization not less than 0.25, and flexibility: no more than 9 rotatable bonds. LC has the most miniature likeness for oral absorption because of its flexibility. At the same time, ST falls within the excellent bioavailability region. The GI absorption for PS and LC is high, which might cause adverse effects by entering the bloodstream through the gastrointestinal tract. On the other side, the GI absorption for LC was less, suggesting non-toxicity to the gastrointestinal tract. CGs show high log KP values (permeability coefficient), which might result in toxicity.

Table 2

Pharmacokinetics of CGs and its degradation products
Pharmacokinetics
Properties	Peruvoside (PS)	Strophanthidin (ST)	Lanatoside C (LC)
GI absorption	High	High	Low
BBB permeant	No	No	No
P-gp substrate	Yes	Yes	Yes
CYP1A2 inhibitor	No	No	No
CYP2C19 inhibitor	No	No	No
CYP2C9 inhibitor	No	No	No
CYP2D6 inhibitor	No	No	No
CYP3A4 inhibitor	No	No	No
Log Kp (Skin permeation)	-8.90 cm/s	-8.31 cm/s	-12.26 cm/s

Table 3

Drug-likeness of CGs according to several protocols.
Drug likeliness
Parameters	Peruvoside (PS)	Strophanthidin (ST)	Lanatoside C (LC)
Lipinski	Yes; 1 violation: MW > 500	Yes; 0 violation	Yes; 3 violation: MW > 500, Nor0 > 10, NHorOH > 5
Ghose	NO; 3 violations: MW > 480, MR > 130, #atom > 70	Yes	No; 3 violations: MW > 480, MR > 130, #atom > 70
Veber	Yes	Yes	No; 2 violations: Rotors > 10, TPSA > 140
Egan	No; 1 violation: TPSA > 131.6	Yes	No; 1 violation: TPSA > 131.6
Muegge	YES	Yes	No; 5 violations: MW > 600, TPSA > 150, #rings > 7, H-acc > 7, H-do > 5
Bioavailability score	0.55	0.55	0.17

Overcoming immune recognition is the fundamental mechanism by which the cancer cells escape phagocytosis by macrophages. Evidence has suggested that CGs possess a broad range of anticancer and antiviral activities against various cancer and viral ailments(Reddy et al., 2019; Reddy et al., 2020; Reddy et al., 2020). Pertinent studies in identifying the anticancer activity of CGs have speculated that interrupting different circuits of signal transduction in the cancer system induces their anti-proliferative response. Contemporary research further validated these compounds' substantial role as immune system modulators causing immunogenic cell death (Škubník et al., 2021). Moreover, clinical studies by Olona et al. assessing the cytotoxicity of Ouabain in human macrophages showed that the compound induced cell death through apoptosis and pyroptosis (Olona et al., 2022).

Furthermore, Shandell et al. recently reported an interdepended association between the activation of IDO1, an immune checkpoint protein, and Na⁺/K⁺-ATPase (NKA) inhibition in cancer cells. The study revealed that the CGs, namely Ouabain and Digitoxin downregulated IDO1 by inhibiting the NKA pump(Shandell et al., 2022). These studies specify the potential anticancer property of cardiac glycosides via impeding various signaling complexes and immunomodulating the tumor microenvironment.

Thus, the present study concerns the role of CGs as immune checkpoint inhibitors to enhance the phagocytic potentiality of macrophages against cancer cells. Approved immune checkpoint inhibitors have increased cancer patients' survival rates and renovated modern oncology (Petrova et al., 2017). Intra-tumoral hypoxic condition is a common phenomenon in cancer cells, leading to the expression of a group of genes (CREB, NF-kB, STAT3, and CD47) that promotes cancer progression and tumor vascularization (Semenza, 2016). Previous reports have suggested that hypoxia plays a significant role in stimulating the activity of CD47 in breast tumors and thus acts as a transcriptional activator of CD47. Increased expression of CD47 in cancer cells leads to decreased phagocytosis and cancer progression(Hayat et al., 2020; Jiang et al., 2021).

However, the innate immune system plays a crucial role in phagocytosis, cancer progression, and obstruction by mediating CD47-SIRPα immune checkpoint signaling. Therapeutic approaches/inhibitors inhibiting this pathway's mechanism could discover a novel approach to eliminating the tumor cells through phagocytosis (Chao, Weissman, et al., 2012). Recent studies on this mechanism have demonstrated that SIRPαFc (TTI-621) enhances the macrophage-mediated phagocytosis in many solid tumors (~ 97% of the patient sample were shown sensitivity towards TTI-621), which suggests that targeting the CD47-SIRPα axis is a more beneficial approach to treat cancers(Johnson et al., 2019; Liu et al., 2022). The interaction of CD47 with SIRPα results in the "do not eat me" signal release, which essentially blocks tumor phagocytosis. To our knowledge, very few reports on inhibitors block CD47-SIRPα interaction. For example, RRx-001 is known as an anticancer agent (Phase III) whose function is to inhibit the expression of CD47 in cancer cells and inhibits the expression of SIRPα on monocytes/macrophages (Cabrales, 2019).

The SIRP family contains a cluster of proteins, which includes a receptor (SIRPα), an activating receptor (SIRPβ), and one non-signaling receptor (SIRPγ). The expression patterns of these three proteins are distinct. For instance, SIRPα expresses in monocytes, cancer-derived macrophages, and hematopoietic stem cells in the tumor microenvironment. In contrast, SIRPβ expresses in macrophages and neutrophils, and SIRPγ expresses in NK. Cells and lymphocytes(Takahashi, 2018). Among these, SIRPα contains sub-domains (SIRPα1 and SIRPα2) whose function is the same. SIRPα overexpression is associated with reduced receptiveness towards tyrosinekinase ligands (EGF) and insulin. Down regulation of SIRPα is connected with several biochemical pathways that promote cancer progression. For example, in prostate cancer cells, the down regulation of SIRPα is identified, and the increased expression results in a reduced number of live cells (Yao et al., 2017). Similar results were observed in breast cancer cells. SIRPα was down-regulated, and this protein's up regulation leads to the inhibition of cancer progression in nude mice (Yamasaki et al., 2007). In liver cancer cells, the expression of SIRPα is less than that of normal cells. Simultaneously, the up regulation decreased cancer cells through the SIRPα1/SHP-2 dependent mechanism (Qin et al., 2006).

Inhibitors targeting CD47/SIRPα could block communication between cancer cells and macrophages, which could be a practical approach to treating cancers (Kershaw & Smyth, 2013; Vonderheide, 2015). The present study shows that CGs are crucial in inhibiting HIF1-α and CD47. Two essential findings in our study suggested that PS, ST, and LC inhibited the expression of HIF1-α in breast, lung, and liver cancer cells. Second, the present study elucidated the dual inhibitory property of CGs against HIF1-α and CD47, thus demonstrating the importance of HIF1-α inhibitors that may also work as inhibitors for CD47. However, further studies on phagocytosis assay, could give better hope for developing CGs as therapeutic candidates to treat cancer.

It has been demonstrated that CGs possesses a broad range of anticancer activity. The evidence has suggested that HIF1-α, CD47, and SIRPα play a crucial role in cancer progression, immunity, and phagocytosis. TCGA data showed that HIF1-α, CD47, and SIRPα are associated with poor survival in breast, lung, and liver cancers and suggests the potential benefit of targeting the HIF1-α, CD47, and SIRPα in these cancers. To dissect the mechanism by which CGs induce cell death in cancer cells, we examined the effect of PS, ST, and LC on HIF1-α and CD47 to validate the impact of compounds on these genes and thus regulate cancer growth. Our study demonstrated that the PS, ST, and LC could significantly decrease the expression of HIF1-α and CD47 in cancer cells. Moreover, the blockade of HIF1-α with CGs can lead to the inhibition of CD47, which further inhibits tumor growth and metastasis. Nevertheless, studies are needed to confirm the role of CGs in macrophage-mediated phagocytosis targeting the CD47-SIRPα immune checkpoint mechanism.

Funding: The Science and Engineering Research Board (SERB) Govt. of India funds through the EMR (2016/003715/BBM) scheme supported this work.

Acknowledgments: The authors acknowledge the Science and Engineering Research Board (SERB) Govt. of India for financial assistance. We Thank to Dhanasekhar Reddy for his help for executing the experimental work. HP thanks ICMR-SRF, Govt. of India, and the research facilities supported by the Central University of Kerala.

Conflicts of Interest: The authors declare no conflict of interest

Cabrales, P. (2019). RRx-001 Acts as a Dual Small Molecule Checkpoint Inhibitor by Downregulating CD47 on Cancer Cells and SIRP-α on Monocytes/Macrophages. Translational Oncology, 12(4), 626–632. https://doi.org/10.1016/j.tranon.2018.12.001
Chao, M. P., Majeti, R., & Weissman, I. L. (2012). Programmed cell removal: A new obstacle in the road to developing cancer. Nature Reviews Cancer, 12(1), Article 1. https://doi.org/10.1038/nrc3171
Chao, M. P., Weissman, I. L., & Majeti, R. (2012). The CD47–SIRPα pathway in cancer immune evasion and potential therapeutic implications. Current Opinion in Immunology, 24(2), 225–232. https://doi.org/10.1016/j.coi.2012.01.010
Chen, D. S., & Mellman, I. (2013). Oncology Meets Immunology: The Cancer-Immunity Cycle. Immunity, 39(1), 1–10. https://doi.org/10.1016/j.immuni.2013.07.012
Hanahan, D., & Weinberg, R. A. (2011). Hallmarks of Cancer: The Next Generation. Cell, 144(5), 646–674. https://doi.org/10.1016/j.cell.2011.02.013
Hayat, S. M. G., Bianconi, V., Pirro, M., Jaafari, M. R., Hatamipour, M., & Sahebkar, A. (2020). CD47: Role in the immune system and application to cancer therapy. Cellular Oncology, 43(1), 19–30. https://doi.org/10.1007/s13402-019-00469-5
Jiang, Z., Sun, H., Yu, J., Tian, W., & Song, Y. (2021). Targeting CD47 for cancer immunotherapy. Journal of Hematology & Oncology, 14(1), 180. https://doi.org/10.1186/s13045-021-01197-w
Johnson, L. D. S., Banerjee, S., Kruglov, O., Viller, N. N., Horwitz, S. M., Lesokhin, A., Zain, J., Querfeld, C., Chen, R., Okada, C., Sawas, A., O’Connor, O. A., Sievers, E. L., Shou, Y., Uger, R. A., Wong, M., & Akilov, O. E. (2019). Targeting CD47 in Sézary syndrome with SIRPαFc. Blood Advances, 3(7), 1145–1153. https://doi.org/10.1182/bloodadvances.2018030577
Kershaw, M. H., & Smyth, M. J. (2013). Making Macrophages Eat Cancer. Science, 341(6141), 41–42. https://doi.org/10.1126/science.1241716
Liu, J., Meng, Z., Xu, T., Kuerban, K., Wang, S., Zhang, X., Fan, J., Ju, D., Tian, W., Huang, X., Huang, X., Pan, D., Chen, H., Zhao, W., & Ye, L. (2022). A SIRPαFc Fusion Protein Conjugated With the Collagen-Binding Domain for Targeted Immunotherapy of Non-Small Cell Lung Cancer. Frontiers in Immunology, 13, 845217. https://doi.org/10.3389/fimmu.2022.845217
Matsumoto, S., Yasui, H., Mitchell, J. B., & Krishna, M. C. (2010). Imaging Cycling Tumor Hypoxia. Cancer Research, 70(24), 10019–10023. https://doi.org/10.1158/0008-5472.CAN-10-2821
Moraes, L. A., Kar, S., Foo, S. L., Gu, T., Toh, Y. Q., Ampomah, P. B., Sachaphibulkij, K., Yap, G., Zharkova, O., Lukman, H. M., Fairhurst, A.-M., Kumar, A. P., & Lim, L. H. K. (2017). Annexin-A1 enhances breast cancer growth and migration by promoting alternative macrophage polarization in the tumour microenvironment. Scientific Reports, 7(1), Article 1. https://doi.org/10.1038/s41598-017-17622-5
Olona, A., Hateley, C., Guerrero, A., Ko, J.-H., Johnson, M. R., Anand, P. K., Thomas, D., Gil, J., & Behmoaras, J. (2022). Cardiac glycosides cause cytotoxicity in human macrophages and ameliorate white adipose tissue homeostasis. British Journal of Pharmacology, 179(9), 1874–1886. https://doi.org/10.1111/bph.15423
Petrova, P. S., Viller, N. N., Wong, M., Pang, X., Lin, G. H. Y., Dodge, K., Chai, V., Chen, H., Lee, V., House, V., Vigo, N. T., Jin, D., Mutukura, T., Charbonneau, M., Truong, T., Viau, S., Johnson, L. D., Linderoth, E., Sievers, E. L., … Uger, R. A. (2017). TTI-621 (SIRPαFc): A CD47-Blocking Innate Immune Checkpoint Inhibitor with Broad Antitumor Activity and Minimal Erythrocyte Binding. Clinical Cancer Research, 23(4), 1068–1079. https://doi.org/10.1158/1078-0432.CCR-16-1700
Qin, J.-M., Yan, H.-X., Liu, S.-Q., Wan, X.-W., Zeng, J.-Z., Cao, H.-F., Qiu, X.-H., Wu, M.-C., & Wang, H.-Y. (2006). Negatively regulating mechanism of Sirpalpha1 in hepatocellular carcinoma: An experimental study. Hepatobiliary & Pancreatic Diseases International, 5(2), 246–251.
Rankin, E. B., & Giaccia, A. J. (2008). The role of hypoxia-inducible factors in tumorigenesis. Cell Death & Differentiation, 15(4), Article 4. https://doi.org/10.1038/cdd.2008.21
Reddy, D., Ghosh, P., & Kumavath, R. (2020). Strophanthidin Attenuates MAPK, PI3K/AKT/mTOR, and Wnt/β-Catenin Signaling Pathways in Human Cancers. Frontiers in Oncology, 9. https://www.frontiersin.org/articles/10.3389/fonc.2019.01469
Reddy, D., Kumavath, R., Barh, D., Azevedo, V., & Ghosh, P. (2020). Anticancer and Antiviral Properties of Cardiac Glycosides: A Review to Explore the Mechanism of Actions. Molecules, 25(16), Article 16. https://doi.org/10.3390/molecules25163596
Reddy, D., Kumavath, R., Ghosh, P., & Barh, D. (2019). Lanatoside C Induces G2/M Cell Cycle Arrest and Suppresses Cancer Cell Growth by Attenuating MAPK, Wnt, JAK-STAT, and PI3K/AKT/mTOR Signaling Pathways. Biomolecules, 9(12), Article 12. https://doi.org/10.3390/biom9120792
Reddy, D., Kumavath, R., Tan, T. Z., Ampasala, D. R., & Kumar, A. P. (2020). Peruvoside targets apoptosis and autophagy through MAPK Wnt/β-catenin and PI3K/AKT/mTOR signaling pathways in human cancers. Life Sciences, 241, 117147. https://doi.org/10.1016/j.lfs.2019.117147
Semenza, G. L. (2010). Defining the role of hypoxia-inducible factor 1 in cancer biology and therapeutics. Oncogene, 29(5), Article 5. https://doi.org/10.1038/onc.2009.441
Semenza, G. L. (2016). The hypoxic tumor microenvironment: A driving force for breast cancer progression. Biochimica et Biophysica Acta (BBA) - Molecular Cell Research, 1863(3), 382–391. https://doi.org/10.1016/j.bbamcr.2015.05.036
Shandell, M. A., Capatina, A. L., Lawrence, S. M., Brackenbury, W. J., & Lagos, D. (2022). Inhibition of the Na+/K+-ATPase by cardiac glycosides suppresses expression of the IDO1 immune checkpoint in cancer cells by reducing STAT1 activation. Journal of Biological Chemistry, 298(3), 101707. https://doi.org/10.1016/j.jbc.2022.101707
Škubník, J., Pavlíčková, V., & Rimpelová, S. (2021). Cardiac Glycosides as Immune System Modulators. Biomolecules, 11(5), 659. https://doi.org/10.3390/biom11050659
Swann, J. B., & Smyth, M. J. (2007). Immune surveillance of tumors. The Journal of Clinical Investigation, 117(5), 1137–1146. https://doi.org/10.1172/JCI31405
Takahashi, S. (2018). Molecular functions of SIRPα and its role in cancer (Review). Biomedical Reports, 9(1), 3–7. https://doi.org/10.3892/br.2018.1102
Triner, D., & Shah, Y. M. (2016). Hypoxia-inducible factors: A central link between inflammation and cancer. The Journal of Clinical Investigation, 126(10), 3689–3698. https://doi.org/10.1172/JCI84430
Tsai, R. K., & Discher, D. E. (2008). Inhibition of “self” engulfment through deactivation of myosin-II at the phagocytic synapse between human cells. Journal of Cell Biology, 180(5), 989–1003. https://doi.org/10.1083/jcb.200708043
Tseng, D., Volkmer, J.-P., Willingham, S. B., Contreras-Trujillo, H., Fathman, J. W., Fernhoff, N. B., Seita, J., Inlay, M. A., Weiskopf, K., Miyanishi, M., &Weissman, I. L. (2013). Anti-CD47 antibody-mediated phagocytosis of cancer by macrophages primes an effective antitumor T-cell response. Proceedings of the National Academy of Sciences, 110(27), 11103–11108. https://doi.org/10.1073/pnas.1305569110
Veillette, A., & Chen, J. (2018). SIRPα–CD47 Immune Checkpoint Blockade in Anticancer Therapy. Trends in Immunology, 39(3), 173–184. https://doi.org/10.1016/j.it.2017.12.005
Vonderheide, R. H. (2015). CD47 blockade as another immune checkpoint therapy for cancer. Nature Medicine, 21(10), Article 10. https://doi.org/10.1038/nm.3965
Yamasaki, Y., Ito, S., Tsunoda, N., Kokuryo, T., Hara, K., Senga, T., Kannagi, R., Yamamoto, T., Oda, K., Nagino, M., Nimura, Y., & Hamaguchi, M. (2007). SIRPα1 and SIRPα2: Their role as tumor suppressors in breast carcinoma cells. Biochemical and Biophysical Research Communications, 361(1), 7–13. https://doi.org/10.1016/j.bbrc.2007.06.159
Yao, C., Li, G., Cai, M., Qian, Y., Wang, L., Xiao, L., Thaiss, F., & Shi, B. (2017). Prostate cancer downregulated SIRP-α modulates apoptosis and proliferation through p38-MAPK/NF-κB/COX-2 signaling. Oncology Letters, 13(6), 4995–5001. https://doi.org/10.3892/ol.2017.6070
Zhang, H., Lu, H., Xiang, L., Bullen, J. W., Zhang, C., Samanta, D., Gilkes, D. M., He, J., & Semenza, G. L. (2015). HIF-1 regulates CD47 expression in breast cancer cells to promote the evasion of phagocytosis and the maintenance of cancer stem cells. Proceedings of the National Academy of Sciences, 112(45), E6215–E6223. https://doi.org/10.1073/pnas.1520032112
Zhang, H., Qian, D. Z., Tan, Y. S., Lee, K., Gao, P., Ren, Y. R., Rey, S., Hammers, H., Chang, D., Pili, R., Dang, C. V., Liu, J. O., & Semenza, G. L. (2008). Digoxin and other cardiac glycosides inhibit HIF-1α synthesis and block tumor growth. Proceedings of the National Academy of Sciences, 105(50), 19579–19586. https://doi.org/10.1073/pnas.0809763105

SupplemetaryFiguresed.1.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

Targeted inhibition of CD47 expression in Human cancer cell lines on treatment with Cardiac Glycosides

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials And Methods

2.1 Cell culture

2.2 Reagents

2.3 RNA extraction and qRT-PCR

2.4 Immunoblotting

2.5 Immunofluorescence studies

2.6 Molecular docking

2.7 Molecular dynamic simulations

2.8 Gene expression analysis

2.9 In-silico ADMET prediction

2.10 Statistical analysis

3. Results

3.1. CGs inhibit the expression of Hypoxia Inducible Factor 1α

3.2. CG.s inhibits the expression of CD47 in cancer cells

3.3. Immunofluorescence analysis

3.4. Molecular docking

3.5. Molecular dynamic simulations

3.7. In-silico ADMET toxicity screening of CGs

4. Discussion

5. Conclusions

Declarations

References

Supplementary Files

Status:

Version 1