**Isothiocyanates (ITCs) 1-(Isothiocyanatomethyl)-4 phenylbenzene and 1-Isothiocyanato-3,5 bis(trifluoromethyl)benzene—Aldehyde Dehydrogenase (ALDH) Inhibitors, Decreases Cisplatin Tolerance and Migratory Ability of NSCLC**

**Jolanta Kryczka 1, \* , Jakub Kryczka 2 , Łukasz Janczewski 3 , Anna Gajda 3 , Andrzej Fr ˛aczyk 4 , Joanna Boncela 2 , Beata Kolesi ´nska <sup>3</sup> and Ewa Brzezia ´nska-Lasota 1**


3

4

**\*** Correspondence: jolanta.kryczka@umed.lodz.pl

**Abstract:** One of the main treatment modalities for non-small-cell lung cancer (NSCLC) is cisplatinbased chemotherapy. However, the acquisition of cisplatin resistance remains a major problem. Existing chemotherapy regimens are often ineffective against cancer cells expressing aldehyde dehydrogenase (ALDH). As such, there is an urgent need for therapies targeting ALDH-positive cancer cells. The present study compares the anticancer properties of 36 structurally diverse isothiocyanates (ITCs) against NSCLC cells with the ALDH inhibitor disulfiram (DSF). Their potential affinity to ALDH isoforms and ABC proteins was assessed using AutoDockTools, allowing for selection of three compounds presenting the strongest affinity to all tested proteins. The selected ITCs had no impact on NSCLC cell viability (at tested concentrations), but significantly decreased the cisplatin tolerance of cisplatin-resistant variant of A549 (A549CisR) and advanced (stage 4) NSCLC cell line H1581. Furthermore, long-term supplementation with ITC 1-(isothiocyanatomethyl)-4-phenylbenzene reverses the EMT phenotype and migratory potential of A549CisR to the level presented by parental A549 cells, increasing E-Cadherin expression, followed by decreased expression of ABCC1 and ALDH3A1. Our data indicates that the ALDH inhibitors DSF and ITCs are potential adjuvants of cisplatin chemotherapy.

**Keywords:** non-small-cell lung cancer; cisplatin resistance; aldehyde dehydrogenase; isothiocyanates; disulfiram; epithelial to mesenchymal transition

#### **1. Introduction**

Since its introduction into clinical trials in 1971 and subsequent Food and Drug Administration approval in 1978, cisplatin represents a major landmark in the history of successful anti-cancer therapeutics. It has changed the management of several solid malignancies, including lung cancer, which remains the second-most-common cancer globally and the leading cause of cancer death [1]. Approximately 2.2 million of new cases of lung cancer are estimated to occur each year worldwide, with a mean 5-year survival rate of 22% [2,3].

**Citation:** Kryczka, J.; Kryczka, J.; Janczewski, Ł.; Gajda, A.; Fr ˛aczyk, A.; Boncela, J.; Kolesi ´nska, B.; Brzezia ´nska-Lasota, E. Isothiocyanates (ITCs) 1-(Isothiocyanatomethyl)-4 phenylbenzene and 1-Isothiocyanato-3,5 bis(trifluoromethyl)benzene— Aldehyde Dehydrogenase (ALDH) Inhibitors, Decreases Cisplatin Tolerance and Migratory Ability of NSCLC. *Int. J. Mol. Sci.* **2022**, *23*, 8644. https://doi.org/10.3390/ ijms23158644

Academic Editor: Angela Stefanachi

Received: 22 July 2022 Accepted: 29 July 2022 Published: 3 August 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

One of the factors that contributes to such a low survival rate is the development of cisplatin resistance [4,5]. Cisplatin resistance is very complicated because of the multifactorial composition of several mechanisms including DNA repair, induction of anti-apoptotic signals, and the active efflux of drugs from the cell cytoplasm. Additionally, cisplatinresistant cells were proven to undergo Epithelial to Mesenchymal Transition (EMT), which plays a substantial role in cancer progression and metastasis, increasing cancer cell motility and invasiveness [5]. EMT renders cancer cells virtually impervious to the majority of anticancer drugs, decreasing cell proliferation and directly influencing the expression of ABC-family transporters, which are involved in multidrug resistance [6,7].

Recently, aldehyde dehydrogenase (ALDH) was confirmed to play a role in drug resistance in lung cancer [8]. A bioinformatic analysis of metabolic enzymes found cisplatinresistant NSCLC to express ALDH [9]. ALDH family members are cytosolic or mitochondrial isoenzymes that are responsible for oxidizing intracellular aldehydes. They play a role in the oxidation of retinol to retinoic acid in early stem cell differentiation [10]. Several of the 19 genes known to encode the ALDH family, such as ALDH1A1, ALDH1A2, ALDH1A3, ALDH1A7, ALDH3A1, ALDH4A1, ALDH5A1, ALDH6A1, and ALDH9A1, are considered the cancer stem cells (CSC) markers involved in drug resistance [11–13]. Notably, currently used chemotherapy regimens were shown to be ineffective against ALDH-positive, cisplatin-resistant cancer cells [8]. Importantly, while ALDH1 activity has been reported in a number of NSCLC cell lines and tumor samples, its role in chemotherapy resistance remains unclear [4]. Therefore, it may be beneficial to investigate therapies targeting drugresistant cancer cells expressing ALDH [8]. One group of small-molecule ALDH inhibitors that may be promising candidates for anti-cancer therapy is the isothiocyanates (ITCs).

Isothiocyanates are commonly known for their anti-cancer properties. They are lowmolecular-weight, natural, organic composites characterized by a pungent odor [14], with the general formula R–NCS. They are found in cruciferous vegetables, such as radish, horseradish, wasabi, broccoli, or Brussels sprouts [15,16] and are produced by the reaction of glucosinolates with myrosinases [17–20]. In addition to natural ITCs such as benzyl isothiocyanate [21,22], phenethyl isothiocyanate [23,24] or best-tested sulforaphane (SFN) [25–30], synthetic counterparts of ITCs have been produced by modification with a fluorine atom [31,32], phosphorous group [33,34], or other functional groups [35]. ITCs are used in organic chemistry as substrates to synthesize inter alia heterocyclic compounds or thioamides [36]. However, these compounds are best-known for their anticancer activity [27,37–41] and antibacterial activity [42–44].

In prostate cancer, ITCs decreased the concentration of the anti-apoptotic proteins Bcl-2 (B-cell lymphoma 2) and Bcl-xl (B-cell lymphoma-extra large) and increased the expression of pro-apoptotic proteins Bax (BCL2 Associated X, Apoptosis Regulator) and activate caspases [45]. Additionally, they decreased the activity of apoptosis inhibitors such as cIAP1 (cellular inhibitor of apoptosis protein-1), cIAP2 (cellular inhibitor of apoptosis 2), and XIAP (X-linked inhibitor of apoptosis protein) and induced Apaf1 (Apoptotic Peptidase Activating Factor 1) protein activity [46]. In addition, in human embryonic kidney cell line HEK293, ITCs were found to suppress transcription of histone deacetylases (HDACs), thus deregulating apoptosis- and differentiation-controlling mechanisms [47]. Furthermore, several studies have confirmed that ITCs suppress both the angiogenesis of human umbilical vein endothelial cells (HUVECs) [48] and metastasis of B16F-10 melanoma [33,49]. Additionally, benzyl isothiocyanates BITCs were proven to inhibit the phosphorylation activity of three major mitogen-activated protein kinases (MAPKs), ERK1/2, p38, and p-JNK1/2, thus presenting direct anti-metastatic activities in SK-Hep1 cells [50].

To improve the current therapeutic efficacy of this cisplatin, alternative strategies are needed to overcome resistance. In the present study, 36 structurally different ITCs (**1**–**36**) were synthesized and tested in vitro using lines of lung cancer cells and their cisplatinresistant variants A549, A549CisR, and NCI-H1581. The protein level and function of two isoforms of ALDH were noted, these being potential markers of cisplatin resistance. Furthermore, the effects of targeting ALDH3A1 and ALDH7A1 by chemical inhibition

were assessed in terms of their ability to re-sensitize resistant lung tumor cells to the cytotoxic effects of cisplatin. Moreover, as multiple ITCs were proven to present many, ostensibly not related, anticancer properties, due to different ITCs cellular targets, ITCs 1-(isothiocyanatomethyl)-4-phenylbenzene (named **19**) was screened in regards to possible antimetastatic abilities. ITCs 19 supplementation significantly decreases proteolytic abilities and in the aftermath the invasiveness of A549CisR, thus highlighting multifactorial anticancer properties.

#### **2. Results**

#### *2.1. Characteristics of the Cisplatin-Resistant A549 Cell Line*

A cisplatin-resistant variant of A549 (named A549CisR) was created by constant culturing in increasing cisplatin concentrations (1–10 µM), resulting in an IC<sup>50</sup> value of 150 µM, compared to 75 µM for A549 (data not shown). A549CisR acquired a mesenchymal-like phenotype manifested by upregulation of mesenchymal marker N-cadherin, with simultaneous repression of epithelial marker E-cadherin (Figure 1A), as noted previously [51,52]. The degree of mesenchymal properties acquired via EMT varied between cells from an epithelial-like status, through a mixed epithelial/mesenchymal (E/M hybrid) form to a strongly mesenchymal phenotype. The hybrid and mesenchymal cells exhibited increased invasive features and circulating tumor cell (CTC) characteristics, suggesting that EMT plays an important role during metastatic dissemination [53]. The A549CisR EMT phenotype was followed by changes in ALDH3A1, ALDH7A1, and ABC protein expression (Figure 1A). A549CisR presented significantly higher ALDH3A1 (stem cell marker [10]), ALDH7A1, ABCC1, and ABCC4 expression and lower expression of ABCG2 compared to A549. Interestingly, ALDH1A1 expression did not differ (data not shown). Furthermore, we noticed changes in cell morphology, A549CisR cells became larger, spindle-shaped, and less densely packed than the parental A549 (Figure 1B). The cisplatin-induced EMT model mimics a natural shift toward higher aggression, and increased the migratory ability and metastasis obtained by chemo-resistant cancer cells during cancer progression [5,54].

Regarding the effect of cisplatin resistance related changes on NSCLC migration, A549CisR presented a higher 2D migration rate (observed in wound healing assay) than the parental sensitive variant (A549) (Figure 1C), even though this variant is considered to be a fairly aggressive NSCLC model [55]. This further demonstrates the increased metastatic potential associated with cisplatin resistance.

#### *2.2. ALDH Inhibitor—Disulfiram Impact on Suppression of Cisplatin Resistance*

The study also analyzed the impact of disulfiram (DSF), a well-known ALDH inhibitor, on A549 and A549CisR. DSF has been used to control alcohol abuse for many decades; however, it has recently been found to have strong anticancer activity both in vitro and in cancer xenografts [56]. DSF treatment (3 µM, 48 h) slightly reduces the mesenchymal phenotype of the cisplatin-resistant variant of A549, restoring the expression of E-cadherin, an epithelial marker (Figure 1D), to the level observed in parental A549. However, the phenotype was not fully restored, as the expression of N-cadherin, a mesenchymal marker, remained unchanged. Importantly, DSF partially re-sensitized A549CisR cells to cisplatin treatment (Figure 1E). Supplementation with 3 µM DSF followed by 75 µM cisplatin (24 h) significantly reduced the tolerance of A549CisR to cisplatin treatment. DSF impact on A549CisR viability is presented in Supplementary Materials (Figure S1).

— — **Figure 1.** Characteristics of the cisplatin-resistant A549 cell line and DSF reversion of cisplatin resistance. (**A**) Western blot analysis of EMT markers and ABC proteins, performed in standard reducing SDS PAGE conditions. (**B**) Cell morphology (phase contrast microscopy). (**C**) Wound healing such as analysis of resistant-cell migration. A549 and A549CisR cells were seeded on a six-well plate and grown to confluence. Next, a wound was created and rinsed twice with PBS. New full medium was added. Wounded area was visualized after 0, 4, 8, 24, and 48 h using an OLYMPUS IX53 microscope and calculated by ImageJ software, *n* = 4, \* *p* < 0.05; \*\* *p* < 0.005; \*\*\* *p* < 0.001, NS—not statistically significant. (**D**) DSF treatment reverses cisplatin resistance. Standard SDS-PAGE Western blot analysis of EMT markers. The A549CisR cells were treated with 3 µM DSF (48 h), and cell lysates were obtained using M-PER Mammalian Protein Extraction Reagent #78501 as described in Materials and Methods. (**E**) To determine the impact of DSF on cell viability, A549CisR cells were seeded on 96-well plates and treated with 75 µM cisplatin, 3 µM DSF, and 3 µM DSF 75 µM cisplatin. After 48 h, cell viability was tested using WST-1 assay (ScienCell, Research Lab., Carlsbad, CA, USA); *n* = 3, \* *p* < 0.05; \*\* *p* < 0.005; \*\*\* *p* < 0.001, NS—not statistically significant.

#### *2.3. Analysis of Cisplatin Resistance in NSCLC Cells Based on Changes in mRNA Expression of ALDH Family Proteins, EMT Marker, and ABC Proteins According to GEO Data*

Delivery of cisplatin-resistant NSCLC cells is considered the standard approach (Materials and Methods section). However, to confirm whether the phenotypical changes occurring in the A549CisR variant are universal, the differences in mRNA expression between A549 and A549CisR were compared using the Gene Expression Omnibus GSE108214 database (Figure 2). A549CisR demonstrated significantly higher ALDH3A1 (stem cell marker) [10] and ALDH7A1 mRNA expression compared to parental A549, similarly to those obtained by our Western blot results. Interestingly, no statistically significant changes were observed in the mRNA expression of the most well-known stem cell marker, ALDH1A1 (Figure 2A) [10].

**Figure 2.** Cisplatin-resistance-related changes in mRNA expression of ALDH family proteins (**A**),

–

–

—

EMT markers (**B**), and ABC proteins (**C**). NSCLC cell line A549 and its cisplatin-resistant variant A549CisR mRNA levels were analyzed using microarray dataset GSE108214, acquired from the public Gene Expression Omnibus (GEO) databases [59]. Normality test (Shapiro–Wilk) was performed, followed by the Mann–Whitney U test; \* *p* < 0.05; \*\* *p* < 0.005; \*\*\* *p* < 0.001, not statistically significant no indicator.

In addition, A549 demonstrated significant upregulation of all major EMT markers, including vimentin (VIM), fibronectin (FN1), and N-Cadherin (CDH-2) (Figure 2B), with no repression of epithelial E-Cadherin (CDH-1). This suggests that acquisition of cisplatin resistance is accompanied by the development of an advanced (not yet fully completed) hybrid EMT phenotype, with strong migratory potential. Recently, EMT was confirmed to be an important regulator of several ABC proteins, as the promoters of ABC transporters contain several binding sites for EMT-inducing transcription factors such as: Twist, Snail, and ZEB. Thus, this leads to increased ABC protein mRNA expression and increased broad, multidrug resistance [7,57,58]. Analysis of the GSE108214 data set confirmed that the mRNA expression of ABCB1, ABCC1, and ABCC4 by cisplatin-resistant variants of A549 (A549CisR) was significantly upregulated, whereas ABCG2 was downregulated (Figure 2C).

#### *2.4. Assessment of Affinity of Isothiocyanates to Chosen ADHD and ABC Family Protein*

In vivo, DSF is rapidly metabolized to diethyldithiocarbamate (DDTC), which is further converted to *S*-methyl-*N*,*N*-diethyldithiocarbamate (Me-DDTC) and *S*-methyl-*N*,*N*diethylthiocarbamate (DETC). Subsequently, P450 catalyzes the oxidation of DETC, and Me-DDTC produces DETC-sulfoxide (DETC-SO) and *S*-methyl-*N*,*N*-diethyldithiocarbamatesulfoxide (Me-DDTC-SO) and -sulfone (Me-DTC-SO2), metabolites that are most likely directly involved in ALDH inhibition. Importantly, when downstream steps of DSF metabolism are blocked by a chemical P450 inhibitor, liver ALDH remains uninhibited, confirming that it is the metabolites of DSF that are the true inhibitors of ALDH in vivo [60].

Therefore, the study examined the effects of isothiocyanate DSF analogues (ITCs) synthesized for the purpose of the study. In silico molecular modeling allowed the selection of the three most promising compounds: (2-isothiocyanatoethane-1,1-diyl)dibenzene named **18**, 1-(isothiocyanatomethyl)-4-phenylbenzene named **19**, and 1-isothiocyanato-3,5 bis(trifluoromethyl)benzene named **36** (Figure 3A); these presented the highest affinity to the several available models (RCSB Protein Data Bank) of ALDH isoforms, including ALDH3A1 and ALDH7A1, which were upregulated in the cisplatin-resistant variant of A549 (Figure 3B), and the ABC proteins involved in drug resistance (ABCB1, ABCC1, and ABCG2) (Figure 3C). Among the tested ITCs, the highest average affinity to all selected proteins was demonstrated by **18**, **19**, and **36** (Figure 3D).

**Figure 3.** Molecular modeling of isothiocyanate (ICT) affinity to the tested ALDH and ABC proteins. (**A**) Chemical structures of three chosen ICTs: **18**, **19**, and **36**. (**B**) ICT affinity to the models of ALDH isoforms and (**C**) ABC proteins obtained from RCSB Protein Data Bank. (**D**) Average affinity of ICTs to all selected proteins, calculated as average of all affinity value to all structural models of each protein.

#### *2.5. Impact of ITCs on the Reversion of Cisplatin Resistance, EMT Phenotype and Cell Migration*

μM for ITC μM for ITC – The chosen isothiocyanates are well-tolerated by the A549 cell line across a broad concentration range for 24 h: the IC<sup>50</sup> values were approximately 80 µM for ITC **18** but are not present for ITC **19** and are 360 µM for ITC **36**. The ITCs demonstrated no cytotoxic effect against A549 or A549CisR at concentrations of 0.1–5 µM (Figure 4A). However, **19** and **36** significantly repressed the cisplatin resistance of A549CisR cells at 10-fold lower concentrations (0.3 µM), similar to DSF itself (3 µM) (Figure 4B). No statistically significant changes in cisplatin resistance were observed for combined 0.3 µM ITC **18** and 75 µM cisplatin (Figure 4B). Additionally, ITCs 19 impact on cisplatin resistance repression was

tested in 3 µM concentration, presenting nearly 50% higher sensitivity compared to the untreated, resistant variant A549 (Figure 4B).

— **Figure 4.** Induction of cisplatin-resistance reversion by isothiocyanates. (**A**) The impact of isothiocyanate concentration on A549 cell viability—WST-1 assay after 24 h supplementation with compounds **18**, **19**, or **36** (ScienCell, Research Lab., Carlsbad, CA, USA). (**B**) Reversion of cisplatin resistance (WST-1 assay). (**C**) Downregulation of ALDH3A1 by ITCs. Long-term (3 µM, 15 days) impact of compound **19** on EMT marker and ABC protein levels. (**D**) Standard SDS-PAGE Western blot. (**E**) Wound-healing-like analysis of the impact of compound **19** on cisplatin-resistant NSCLC cell migration. A549CisR and A549CisR supplemented with compound **19** (3 µM, 15 days) cells were seeded on a six-well plate, grown to confluence. A wound was made and rinsed twice with PBS. New full medium was added. The wounded area was visualized after 0, 4, 8, 24, and 48 h using an OLYMPUS IX53 microscope and calculated by ImageJ software, *n* = 3. (**F**) ITC **19** increases cisplatin sensitization of advanced NSCLC cells NCI-H1581. \* *p* < 0.05; \*\* *p* < 0.005; \*\*\* *p* < 0.001, not statistically significant—no marker.

As compound **19** presented the strongest cisplatin-resistance reversion abilities and the highest tolerability in both tested cell lines, it was chosen for further study of its long-term

impact on cisplatin-resistant NSCLC cells when applied at the same concentration as DSF itself (3 µM). Fifteen-day supplementation with 3 µM ITC **19** resulted in significant repression of ALDH3A1 expression (similarly to **18** and **36**) (Figure 4C) suggesting the reversion of a stem-cell-like phenotype. Furthermore, compound **19** was found to partially reverse EMT phenotype, presenting increased expression of the epithelial marker E-cadherin, with no significant changes in N-cadherin expression (Figure 4D) (similar to DSF). Phenotype reversion was also accompanied by the acquisition of an ABC protein-expression pattern, characteristic of the non-resistant A549 parental variant (Figure 4D), i.e., a decrease in ABCC1 and an increase in ABCG2 expression.

Furthermore, acquisition of cisplatin resistance increased the migratory abilities of A549CisR cells, rendering them highly metastatic. However, treatment with ITC **19** (15 days, 3 µM) significantly lowered 2D migration rate, as indicated by wound healing assay (Figure 4E).

#### *2.6. Isothiocyanate #19 Increases Cisplatin Sensitization of Advanced NSCLC Cells*

As ITC **19** significantly reduces cisplatin resistance in A549CisR, it was also tested against NCI-H1581, a stage 4 NSCLC cell line (CRL-5878) (Figure 4F). NSCLC predominantly consists of adenocarcinoma (AC) and squamous cell carcinoma (SCC). H1581 represents the smallest subfraction (10%) of NSCLC: a large cell carcinoma (LCC) that tends to grow rapidly and spread more aggressively than some other forms of lung cancer. H1581 is characterized by high focal amplification of FGFR1 [61] overexpression, which is related to increased aggressiveness, metastasis, and poor prognosis in various cancer types (especially in NSCLC) [62]. Activation of FGFR1 was reported to initiate EMT in several cancer types, including primary or secondary drug resistant lung cancer and lung cancer cell lines such as H1581 [63]. Importantly, H1581 cell line presents an EMT-derived, highly drug-resistant, cancer stem-cell-like phenotype, with increased ALDH activity. Inhibition of FGFR1 and ALDH activity suppress the growth, viability, and stem-cell-like phenotype of H1581 [64]. Thus, H1581 may be considered a well indicator, partially proving ALDH importance in drug resistance. Supplementation of H1581 with 20 µM cisplatin (24 h) had no significant impact on cell viability; however, the combination of 3 µM **19** and 20 µM cisplatin significantly decreased cell viability, as observed using a standard WST-1 assay (Figure 4F). ITCs 19′ s impact on H1581 viability is presented in Supplementary Materials (Figure S2).

#### *2.7. Isothiocyanate #19 Decreases Proteolytical Abilities and Invasive Properties of Cisplatin-Resistant NSCLC Cells*

The obtained cisplatin-resistant variant of A549 presents a significantly higher migration rate as observed in the wound healing assay. However, migration has two main types: "path finding" (amoeboid migration type) and "path generating" (mesenchymal type of migration). Amoeboid migration is characterized by rounded cell morphology, low adhesion, high migration velocity, extensive cell body deformations caused by actin protrusions or hydrostatic membrane blebs, and its independence of extracellular matrix (ECM) degradation. Thus, amoeboid migration is based on cells' abilities to find and fit into existing "paths". On the other hand mesenchymal type of migration is acquired in non-mesenchymal cells via EMT and strongly depends on proteolytic degradation of ECM components (mainly via matrix metalloproteinases—MMPs), which enables crossing of the anatomical boundaries and in-aftermath metastasis [65,66]. Cisplatin resistance is often accompanied by increased metastatic potential. A549CisR presents higher proteolytic abilities than parental A549, as visualized by confocal microscopy imaging (zymography in situ assay—white arrows point increased gelatinolytic effect) (Figure 5A) and calculated using fluorescent dequenching (DQ) gelatin assay (Figure 5B). A549CisR supplemented with 3 µM **19** decreases gelatin degradation to the level observed for parental A549 and A549CisR treated with MMP2 inhibitor ARP101 (24 h, 10 µM). Interestingly, gelatin degradation presented by A549CisR supplemented with both 10 µM ARP101 and 3 µM **19** is slightly, yet statistically significant,

lower than the one observed for either of the compounds alone, whereas the MMP2 protein level remains unchanged by **19** (Figure 5C). An increased ability to cleave ECM components such as collagen (or its degraded form—gelatin) is required by cancer cells during invasion and metastasis, as it provides physical disintegration of anatomical boundaries allowing for invasion of surrounding as well as distant metastasis [66]. Thus, invasive properties of A549 and A549CisR were tested (Figure 5D,E). Cells were treated with or without 3 µM **19** for 24 h, and next were transferred to gelatin coated 8 µm pore size upper chamber of Nunc Cell Culture Inserts in starving medium (with or without 3 µM **19**). A full medium was used in lower chamber as chemoattractant. Cells were allowed to enter the membrane pores thru gelatin layer for 3 h. Next, the medium and the gelatin from the top surface of the membrane were removed, the invaded cells on the bottom surface of the membrane were washed 2× with PBS, and then fixed, stained with Hoechst 33342, and counted in five random spots. The cisplatin-resistant variant presents a significantly higher invasion rate than the parental A549 cells. Furthermore, supplementation with **19** significantly decreases invasion of A549CisR, presenting no statistically significant effect on A549.

**Figure 5.** Repression of proteolytical and invasive abilities of A549CisR by isothiocyanates. (**A**) Confocal

—

—

—

microscopy image of A549 and A549CisR proteolytic degradation of FITC-conjugated DQ gelatin (Life Technologies, Waltham, MA, USA) (zymography in situ). Cleaved DQ gelatin becomes fluorescently active as highlighted by white arrows. F-actine stained with Texas Red-X Phalloidin, nuclei stained with Hoechst 33342. (**B**) Gelatinolytic activity of A549 and A549CisR calculated by measuring fluorescence intensity of cleaved DQ gelatin. MMP2 inhibitor ARP101 used as an additional control to verify ITCs impact on DG gelatin degradation. Calculated fluorescence intensity in arbitrary units of fluorescence were et as control—100% for A549. \* *p* < 0.05, not statistically significant— NS. (**C**) Standard SDS-PAGE Western blot analysis of MMP2 expression in A549, A549CisR, and A549CisR supplementation with compound **19** (24 h 3 µM). (**D**) Gelatin transwell invasion assay based on [57,67]. A549 and A49CisR were treated with or without compound **19** (24 h 3 µM) and transferred to upper chamber in starving medium supplemented with or without **19** 3 µM. Full medium was used as chemoattractant in lower chamber. Invasion was visualized and calculated after 3 h in 5 random spots. \* *p* < 0.05. (**E**) Gelatin transwell invasion assay visualization by fluorescence microscopy—random representative image.

#### **3. Discussion**

Currently, the best treatment for most types of carcinomas is surgical excision of the early primary tumor with proper histopathologic margins [68]. However, often due to late diagnosis (advanced stage of tumor), cancer cells are able to increase their mass and invade surrounding/distant tissue, thus preventing surgical removal. In such cases, chemotherapy remains one of the basic treatment modality [5,54]. This fact is extremely important in case of lung cancer: due to the lack of effective early-detection methods, it has one of the highest mortality rates among cancers [1].

Cisplatin is commonly used in many lung cancer types, including squamous cell carcinoma (SCC), large cell carcinoma (LCC), and adenocarcinoma (AC) [69]. Although cisplatin shows remarkable effects during initial treatment, a large majority of patients develop resistance as treatment proceeds, presenting a higher number of secondary tumors after period of remission [5,54]; for example, 30–55% of non-small-cell lung cancer (NSCLC) patients (both adenocarcinoma and squamous cell carcinoma) [70] suffer from cisplatinresistant cancer reoccurrence within one year of surgery and associated chemotherapy. There are many mechanisms responsible for cisplatin therapy failure such as DNA-damage repair, cell-death inhibition, drugs efflux and inactivation, drug-target alteration, and metabolic shift. However, one of the most prominent mechanisms is EMT induction, which allows an epithelial cell to obtain a mesenchymal phenotype, resulting in an increased ability to migrate [5]. Importantly, cisplatin-related EMT leads to the acquisition of a migratory phenotype, which allows passage across anatomical boundaries and the invasion of both local and distant tissue [71]. Interestingly, inhibition of Ataxia Telangiectasia Mutated (ATM) results in reversion of the EMT phenotype in cisplatin-resistant NSCLC cells, inhibiting cell invasion and tumor metastasis [72]. Although cisplatin resistance is driven by multiple, often ostensibly unrelated mechanisms, the Western blot analysis of our present cisplatin-resistant A549CisR found it to correspond on the protein level with the mRNA profile of a previously described A549CisR variant [73], which is publicly available in the GEO database as the GSE108214 dataset.

Furthermore, in our A549CisR cell line, cisplatin resistance manifests as ongoing EMT, indicated by upregulation of N-cadherin and repression of E-cadherin, which substantially increased migratory potential. Previous studies have also noted a similar co-existence between cisplatin resistance and EMT induction [72,74]. Importantly, our derived A549CisR cell line, representing an EMT phenotype, demonstrated ABCC1 and ABCC4 upregulation and ABCG2 downregulation; this relationship was observed for GSE108214, while ABCG2 repression has also been noted in advanced colorectal cancer cells undergoing EMT and patient samples [57]. Several EMT-inducing transcription factors, such as Twist, Snail, and ZEB, were recently confirmed to be important regulators of certain ABC proteins, directly interacting with E-Box sites in their corresponding promoter regions [7,57]. This mechanism may be utilized by cisplatin-resistant cells to increase overall multidrug resistance, as ABC proteins have broad spectrum of transported chemotherapeutic agents such as 5-fluorouracil, irinotecan, doxorubicin, mitoxantrone, or vinblastine, to name a few [75,76].

Currently, potential supplements to cisplatin-based chemotherapy are being sought [77– 79]. One such candidate is Disulfiram (DSF), which is an aldehyde dehydrogenase (ADH) inhibitor that has been used as a first-line anti-alcoholism Drug The drug has been reported to cause cell-cycle arrest in the G2/M phase and enhance cisplatin sensitivity in NSCLC lines [80]. Recently, disulfiram has been called "a novel repurposed drug for cancer therapy", with an anti-cancer effect noted in several cancer types, including liver, breast, prostate, pancreatic, and NSCLC [81]. Therefore, the present study examines the potential of ADH inhibitors for improving the treatment of cisplatin-resistant NSCLC. Moreover, interestingly, aldehyde dehydrogenase (ALDH) is considered to be marker of NSCLC circulating tumor cells (CTC), which indicate an advanced EMT phenotype [82].

Our data confirm that DSF triggers re-sensitization of a cisplatin-resistant NSCLC cell line (A549CisR). Although DSF is known to inhibit NF-kB signaling, proteasome activity, and aldehyde dehydrogenase (ALDH) activity and to induce endoplasmic reticulum (ER) stress and autophagy, the exact mechanisms of its anti-cancer properties remain unclear [81].

Our findings indicate that DSF reverses an acquired mesenchymal phenotype to a certain extent, repressing the expression of mesenchymal marker N-cadherin. Interestingly, DSF is rapidly metabolized to diethyldithiocarbamate (DDTC), which is further converted to *S*-methyl-*N*,*N*-diethyldithiocarbamate (DETC) and *S*-methyl-*N*,*N*-diethyldithiocarbamate (Me-DDTC), and it is these metabolites of DSF that are the true inhibitors of ALDH in vivo [60].

The present study evaluates the potential of de novo synthesized isothiocyanates (ICTs) that resemble DSF. Out of 36 tested ITCs, the three most promising compounds were chosen, viz. **18**, **19**, and **36**: these were found to manifest strong affinity to ALDH isoforms and ABC proteins based on in silico analysis. Interestingly, **18** and **19** have also presented strong anti-tumoral properties in a xenograft zebrafish model [83]. The chosen ITCs appear to be strong anti-cisplatin-resistance agents: they were found to significantly repress cisplatin tolerance in both A549CisR and in the stage 4 NSCLC cell line NCI-H1581 at concentrations 10-fold lower than DSF. ALDH family and/or ABC proteins most probably are not the only important target for ITCs, that are beneficial during anticancer therapy. Benzyl isothiocyanates BITCs are one of the most extensively studied ITCs with regard to cancer chemoprevention, which was proven to inhibit the phosphorylation activities of three major mitogen-activated protein kinases (MAPKs): ERK1/2, p38, and p-JNK1/2 [50].

In a previous study, ITCs were found to demonstrate anti-proliferative activity in vitro in the human colon, uterus, mammary gland, and lung carcinoma cell lines. ITC treatment led to cell-cycle arrest and cell death. ITCs have also been found to be effective against a murine mammary gland carcinoma 4T1 model in vivo, with administration resulting in reduced tumor mass [33]. Furthermore, 3 µM ITC analogs inhibited the motility of three highly malignant cell lines derived from cervical (HeLa), glioblastoma (U87), and breast (MDA-MB-231) carcinomas [83]. This finding is consistent with our present observations.

DSF has been used to treat alcohol abuse for about 70 years. Typically, patients receiving DSF-based therapy are exposed to high doses for a long period of time, i.e., six or more months, during which time the drug appears to possess non-lethal properties, if not mixed with alcohol [84].

In the present study, 15-day supplementation with **19** was found to partially reverse the EMT phenotype (restoration of E-cadherin expression) stem-cell-like phenotype (downregulation of ALDH3A1) of the tested on A549CisR cells, and repress their migratory potential to the level observed in parental A549 cells.

Cell migration is a complicated process consisting of cell-body polarization, followed by the formation and extension of cell protrusions; these protrusions adhere to the substratum, and cell contraction moves the cell body forward toward the leading edge. The migration cycle is completed by deadhesion of the attachments at the rear of the cell [85]. This cascade of events requires substantial energy expense in the form of ATP [86]. Thus,

since ALDH mediates the production of NADH, which is used as an energy source for ATP production during oxidative phosphorylation (OXPHOS) [87], long-term ALDH inhibition may slow the migration rate. Importantly, while OXPHOS uses NADH supplied from ALDH in cancer cells, OXPHOS obtains NADH via the tricarboxylic acid cycle (TCA cycle) in normal cells. Thus, targeting cancer cell OXPHOS by inhibiting ALDH could selectively reduce their ATP levels, significantly repressing many of the mechanisms of cancer cells [88]. High ALDH3A1 expression and activity correlates with cell proliferation and resistance against drug toxicity, such as cyclophosphamide, ifosfamide and trofosfamide. Repression of cells proliferation and drug resistance can be observed upon ALDH3A1 directly inhibition by the administration of specific synthetic inhibitors, antisense oligonucleotides, or siRNA [89,90]. Thus, even though ALDH3A1 expression was downregulated in a discrete manner by the tested ITCs in A549CisR cells, inhibition of its activity is the most important factor influencing cell response. Furthermore, mitochondria are able to interact with the nucleus through the retrograde signaling mechanism; this results in the activation of diverse nuclear responses that regulate survival rate, metastasis, and drug resistance [91]. Additionally, mesenchymal type of migration that strongly relies on proteolytical degradation of ECM components is acquired in non-mesenchymal cells via EMT [65,66]. Thus, often cisplatin-resistant cells, presenting EMT phenotype, are characterized by increased proteolytical abilities and a path-generating type of migration [71]. In this study, tested A549CisR presented enhanced gelatinolytic and invasive properties in comparison to the parental A549. Both invasion and gelationlysis were significantly suppressed by 24 h supplementation with 3 µM ITCs **19**. Gelatin is mainly degraded by matrix metalloproteinases 2 and 9 (MMP2 and MMP9), thus, inhibition of their activity results in altered invasion and metastasis. Importantly, **19** in the tested concentration decreases gelatin proteolysis to the level presented by the well-known MMP2 inhibitor ARP101 (10 µM, 24 h), with no changes observed in the MMP2 protein level. Importantly, MMP2 is produced as an inactive proenzyme that requires activation, canonically performed by MMP14 (non-canonically performed by other enzymes, such as MMP2 itself or Cathepsins); thus, the unchanged protein level upon **19** supplementation is less relevant in terms of MMP2 activity and involvement in cell migration [67]. This effect was presented also by benzyl isothiocyanates (BITCs), which was proven to downregulate both MMP2 and MMP9 but, more importantly, to increase the mRNA level of tissue inhibitor of matrix metalloproteinases-2 (TIMP-2) [50]. Thus, we can strongly assume that compound **19** acts as indirect suppressor of MMP2 activity, however, the exact mechanism is yet to be discovered.

Furthermore, in the A549CisR cell line, the ABC protein levels were restored to the parental non-resistant variant: ABCC1 was downregulated whereas ABCG2 was upregulated. Interestingly, neither ABCC1 nor ABCG2 are cisplatin exporters [75,76]; hence, it was unclear why cisplatin resistance regulates ABC protein levels, while not being a substrate for particular transporters itself, and why compound **19** reverses their level to the one observed in parental A549 cells [92]. However, cisplatin-resistant cells exhibit an EMT phenotype, which has been proposed to be the main cause of the primary and acquired drug resistance in several cancer types [7,93]. The observed downregulation of ABCG2 in the more mesenchymal A549CisR cell line may be difficult to explain, but our data correspond well with previous findings, indicating that the downregulation of ABCG2, at both the mRNA and protein level, reverses the correlation with mesenchymal markers and acquisition of advance EMT in CRC cells (in vivo and in vitro) [57]. Furthermore, long-term supplementation with compound **19** partially represses and reverses EMT, i.e., restores E-Cadherin expression, and EMT triggers the metastatic and drug-resistance properties of cisplatin-resistant NSCLC cells.

Importantly, compound **19** presents high affinity to ABC proteins, including ABCB1, ABCC1, and ABCG2, which are present in various pharmaceuticals, such as Doxorubicin, Paclitaxel, Vinblastine, Methotreaxate (MTX), Irinotecan, and Topotecan [75,76]. This can potentially increase their cytosolic accumulation, leading to the repression of cisplatininduced multidrug resistance and the resensitization of cells; however, this needs further

investigation. Compound **19** appears to possess strong anti-cisplatin/anti-multidrug resistance and anti-metastatic properties, and its supplementation during chemotherapy may be highly beneficial for NSCLC patients.

#### **4. Materials and Methods**

#### *4.1. Cell Culturing and Induction of Drug Resistance*

Lung cancer cell lines A549 and NCI-H1581 were obtained from the ATCC (Manassas, VA, USA). The A549 cisplatin-resistant sub-line A549CisR was established by growing A549 cells in the presence of increasing concentrations of cisplatin to a final concentration of 10µM over approximately six months. Both cell lines (A549/A549CisR) were cultured in Ham's F12-K medium (Corning, Manassas, VA, USA), H1581 were cultured in DNEM/HAM F-12 (Corning, Manassas, VA, USA) in a 90–95% humidified atmosphere of 5% CO2; the media were supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Biowest, Nuaillé, France) and the antibiotics streptomycin, penicillin (Biowest), and primocin (Invivogen, San Diego, CA, USA). The cells were plated in 25 cm<sup>2</sup> cell culture flasks and sub-cultured before reaching confluency using Accutase (Biowest). The culture medium was changed every two days. The cells were split 1:10 during each passage.

#### *4.2. Reagents*

Cisplatin [cis-diammineplatinum(II) dichloride], was obtained from Sigma-Aldrich (St. Louis, MO, USA) and dissolved in DMSO. Disulfiram was obtained from Cayman Chemical and dissolved in DMSO. Aliquots were stored at −20 ◦C for up to a maximum of three months and thawed immediately before use. WST-1 [2-(4-Iodophenyl)-3- (4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium] was purchased from ScienCell (Research Lab., Carlsbad, CA, USA) freshly made, #8038; RIPA lysis buffer was purchased from VWR Chemicals, #N653-100 mL; protease inhibitor was purchased from Thermo Fisher Scientific (Waltham, MA, USA), #PIER87785; BCA purchased from Thermo Fisher Scientific, #PIER23225. Primary antibodies: E cadherin #3195P cell signaling, N cadherin #13116P cell signaling, ABCC1 #VMA00330 BioRad (Hercules, CA, USA), ABCC4 #PA5-18315 Thermo Fisher Scientific, ABCG2 #BRB155559 Biorbyt (Cambridge, UK), GAPDH #sc-32233 Santa Cruz Biotechnology (Dallas, TX, USA), α-Tubulin #NB100-690H Novus Biologicals (Littleton, CO, USA), ALDH3A1 #PA5-80332 Thermo Fisher Scientific, ALDH7A1 #MA5-29028 Thermo Fisher Scientific. Secondary HRP-conjugated antibodies were purchased from Santa Cruz Biotechnology.

#### *4.3. Western Immunoblotting*

Total protein was extracted from cells using ice-cold M-PER Mammalian Protein Extraction Reagent #78501 supplemented with the Halt protease inhibitor cocktail (Thermo Scientific, Waltham, MA, USA), and the soluble protein fraction was collected through centrifugation. The protein concentrations in the cell lysates were measured with the BCA method (Pierce/Thermo Scientific, Waltham, MA, USA) and equalized between samples. Protein (40 µg) from whole cell lysates was fractionated on SDS-PAGE gels and transferred to a PVDF or nitrocellulose membrane (BioRad, Hercules, CA, USA). Transfer efficiency and loading were confirmed by reversible staining of the membrane with Ponseau S solution (Sigma-Aldrich, UK) following protein transfer. Membranes were blocked at room temperature with blocking buffer (BioRad, Hercules, CA, USA). Primary antibodies were added in 1:1000–1:5000 dilution and incubated for one hour at RT (Materials and Methods Section 4.2). Membranes were washed 3 × 15′ with TBST and incubated with a secondary horseradish peroxidase (HRP)—labelled antibody for 1 h RT (1:2000). Membranes were washed in 3 × 15′ with TBST following incubation with secondary antibodies. Bound antibody complexes were detected and visualized using Clarity Western ECL Substrate (BioRad, Hercules, CA, USA). Densitometric analysis was carried out using ImageJ software, and percentage expression was represented relative to controls (100%) [94].

#### *4.4. Cell Viability Assay*

The in vitro cell viability effects of DSF, analogs of disulfiram, cisplatin, DSF/cisplatin, and analogs of disulfiram/cisplatin were determined by WST-1 assay. In brief, the cells (2 × 10 5 cells/mL) were seeded on 96-well culture plates and left for 24 h. Next, parental and resistant tumor cells were incubated in 100 µL of fresh medium containing different drug concentrations. Next, after 48 h of incubation, 10 µL of WST-1 reagent (ScienCell, Research Lab., Carlsbad, CA, USA), freshly made, #8038, was added for 2 h. Calculation of cell viability was done by OD450 nm–OD630 nm using the BioTek ELx800 multimode microplate reader. – – – – – –

#### *4.5. Wound Healing (Scratch) Assay*

The 2D migration was tested using wound healing assay. Treated or untreated with ITC **19** (15 days, 3 µM), A549 and A549CisR cells were seeded on a six-well plate and grown to confluence. Wounds were created by scraping monolayer cells using a 20 µL pipet tip, and non-adherent cells were rinsed off twice with PBS. Fresh medium was added with or without 3 µM ITC **19**. The scratch and surrounding cells were imaged at 0 h (immediately after scratching). The area of the wound was visualized (OLYMPUS IX53 microscope, magnification, 100×) and measured 4, 8, 10, and 24 h after scratching. The wound area was calculated by ImageJ software. Cell motility was estimated by quantification of percentage recovery using the equation: R (%) = [1 − (wound area at Tt/wound area at T0)] × 100, where T0 is the wounded area at 0 h, and Tt is the wounded area after th. The assays were replicated three times. grown to confluence. Wounds were created by scraping monolayer cells using a 20 μL − grown to confluence. Wounds were created by scraping monolayer cells using a 20 μL − grown to confluence. Wounds were created by scraping monolayer cells using a 20 μL − grown to confluence. Wounds were created by scraping monolayer cells using a 20 μL − grown to confluence. Wounds were created by scraping monolayer cells using a 20 μL − grown to confluence. Wounds were created by scraping monolayer cells using a 20 μL −

#### *4.6. GEO Database Analysis*

Microarray profiles and the GSE108214 dataset of the NSCLC cell line A549 and its cisplatin-resistant variant (A549CisR) were acquired from the public Gene Expression Omnibus (GEO) databases—National Center for Biotechnology Information (NCBI), USA National Library of Medicine 8600 Rockville Pike, Bethesda, MD 20894, USA (https://www. ncbi.nlm.nih.gov/geo/, accessed on 14 April 2021) [59]; these were described previously [5]. The data were analyzed and presented as box charts, with the median depicted, as described previously [57]. Statistical analysis was performed using Jasp software (https://jasp-stats. org/, accessed on 6 May 2022) [95]. — — — — — —

#### *4.7. Synthesis of ITCs—Similar Compounds to DSF — — — — — —*

The tested isothiocyanates **1**–**36** (Table 1) had previously been synthesized using four known methods [96–98] (Schemes 1–4, Methods A–D). – – – – – – – – – – – – – – – – – – – – – – – –

.


**Table 1.** Structure of tested isothiocyanates **1**–**36** a – – – – – –

**Table 1.** *Cont.*


–

–

–

–

—

—

—

—

– – – –

– – – –

– – – –

– – – –

–

–

–

–

*—*

*—*

*—*

*—*

grown to confluence. Wounds were created by scraping monolayer cells using a 20 μL

grown to confluence. Wounds were created by scraping monolayer cells using a 20 μL

grown to confluence. Wounds were created by scraping monolayer cells using a 20 μL

grown to confluence. Wounds were created by scraping monolayer cells using a 20 μL

−

−

−

−




**Table 1.** *Cont.*

<sup>a</sup> Method A [96]; method B and method C [97]; method D [98].

– – – – – – – – – **Scheme 1.** Method A: microwave-assisted synthesis of isothiocyanates **1**–**6**, **10**, **13**–**28**, and **31**–**36**. – – –

–

–

–

–

–

–

–

–

–

–

– – –

– – –

– – –

– – –

−

−

−

−

−

−

−

−

−

−

−

−

−

−

−

– − **Scheme 3.** Method C: synthesis of isothiocyanate derivatives of methyl ester amino acids **7**–**9** using DMT/NMM/TsO<sup>−</sup> (**74**) as desulfurating reagent.

– – –

−

**Scheme 4.** Method D: synthesis of isothiocyanates **11** and **12** using propane phosphonic acid anhydride (T3P) (**75**).

– – – – – – – – – – – – – – – – – – – The isothiocyanates **1**–**6**, **10**, **13**–**28**, and **31**–**36** were obtained by one-pot, two-step, microwave-assisted (MW) synthesis (Method A). Briefly, a mixture of aliphatic primary amines **37**–**42**, **46**, and **49**–**55** or aromatic amines **56**–**64** and **67**–**72**, carbon disulfide (CS2) and triethylamine (Et3N) (for amines **37**–**42**, **46**, and **49**–**55**), or DBU (for amines **56**–**64** and **67**–**72**) were transformed under normal conditions at room temperature into the intermediate dithiocarbamates **73**. Next, dithiocarbamates **73** were converted without any desulfurating reagent under microwave-assisted reaction (MW: 20 min, 90 ◦C for **37**–**42**, **46**, and **49**–**55** or 100 ◦C for **56**–**64** and **67**–**72**) to ITCs **1**–**6**, **10**, **13**–**28**, and **31**–**36** in good yields (25–98%) (Scheme 1, Table 1) [96].

– − – Isothiocyanates **29**–**30** were also synthesized by one-pot, two-step microwave-assisted synthesis using CS2, DBU, and aromatic amines **65** and **66** as substrates; however, a different MW method was used (Method B). The intermediate dithiocarbamates were synthesized, as shown in Method A. The second step however, performed under microwave-assisted conditions, was accomplished in a shorter time (MW: 3 min and 90 ◦C) and in the presence of 4-(4,6-dimethoxy-1,3,5-triazin-2-yl)-4-methylmorpholinium toluene-4-sulfonate (DMT/NMM/TsO−, **74**) as a desulfurating reagent. Isothiocyanates **29** and **30** were isolated at high yields (67–87%) (Scheme 2) (Table 1) [97].

– − – – – Isothiocyanate derivatives of methyl ester amino acids **7**–**9** were obtained under normal conditions in a one-pot, two-step procedure in the presence of DMT/NMM/TsO<sup>−</sup> (**74**) as desulfurating agent; however, the reaction was performed in normal conditions (Method C). The first step was performed in the presence of CS2, NMM at room temperature in 10 min, using hydrochloride **43**–**45** as substrates; the second step was performed with desulfurating reagent **74** at room temperature and in 30 min. ITCs **7**–**9** were isolated with satisfactory yields (30–51%) after flash chromatography (Scheme 3) (Table 1) [97].

– The last two isothiocyanates **11** and **12** were synthesized in a two-step process with propane phosphonic acid anhydride (T3P) used as a desulfurating reagent (**75**) (Method D). Diamine **47** or hydrobromide **48** was reacted with CS<sup>2</sup> in the presence of Et3N at room temperature for one hour. Next, the reactions were cooled to 4 ◦C, T3P (**75**) was added in two portions, and the mixture was stirred for another two hours at room temperature. Products **11** and **12** were isolated in good yields (55–60%) using flash chromatography (Scheme 4) (Table 1) [98].

–

#### *4.8. Molecular Modeling*

Affinity calculations were performed for 26 ABC proteins and 27 ligands using software dedicated to molecular docking: AutoDockTools v.1.5.6 (La Jolla, California, USA), included in the MGLTools 1.5.6 (La Jolla, California, USA) package, and AutoDock Vina 1.1.2 (La Jolla, California, USA). Files describing proteins were downloaded in pdb format from the RCSB Protein Data Bank [99] and then converted using AutoDockTools to the pdbqt format required by AutoDock Vina.

The structures of ITCs **1**–**37** were drawn in ACD/ChemSketch (Toronto, Ontario, Canada) (Freeware) and then 3D Structure Optimization was performed. The structures were saved in .mol (MDL MOL) format. Next, the structures in .mol format were converted to .pdb (Protein Data Bank) format using OpenBabelGUI. The .pdb format files were used to model docking in the Vina package.

AutoDockTools was also used to determine the size of the search area and its center. For protein molecules with size that exceeded the maximum size of the search area, search sub-areas totaling the entire protein molecule were defined. For example, two search sub-areas covering the entire molecule were created for the 6qex protein, and these were treated as separate cases in the calculations: 6qex(1) and 6qex(2).

As a non-deterministic search algorithm was implemented in AutoDock Vina, each docking variant was calculated 10-fold. As a result, it was necessary to perform 9620 calculations (37 ligands × 26 proteins × 10 searches). As it would be difficult to run AutoDock Vina manually so many times, the calculation process was automated using a script written in Matlab 2021a. This script, in addition to running individual calculation cases, aggregated the results generated by AutoDock Vina. The calculation process is given in the flow diagram presented in Figure 6.

— — — — — — **Figure 6.** Flow diagram of the algorithm controlling the calculations, where i—protein index; j—ligand

—

index; k—search variant index; m—number of proteins; n—number of ligands; A—affinity for the current protein and ligand; Aset—set of affinity values.

The calculations were carried out on a PC equipped with a 12-core/24-thread AMD Ryzen 9 3900X processor. In order to fully utilize all processor cores, three instances of AutoDock Vina were run simultaneously. In order to make a quantitative assessment of the affinity of the studied ligands to selected proteins, the average affinity values of the analyzed ligands for selected proteins, expressed by the relationship (1), were calculated:

$$AA(j) = \frac{\sum\_{i=1}^{i=m} A(i, j)}{m} \tag{1}$$

where *i*—protein index; *j*—ligand index; k—search variant index; *m*—number of proteins; n—number of ligands; *A*—affinity; *AA*—average affinity.

#### *4.9. Fluorescent Dequenching (DQ) Gelatin Assay*

The surface of 96-well plates was coated with 75 µL 0.1 mg/mL DQ gelatin (Life Technologies, Waltham, MA, USA) overnight at 4 ◦C and then washed 3× with PBS. Then, 25 × 10<sup>5</sup> cells/mL were added for 24 h to earlier prepared DQ gelatin-coated dishes in full medium. Additionally, as a control medium supplemented with 10 µM MMP-2 inhibitor ARP101 was used. FITC fluorescence generated by the cleavage of DQ gelatin was measured using a Thermo Labsystem Fluoroscan Ascent reader (ThermoFisher Scientific, Waltham, MA, USA), fit with FITC excitation and emission filters. Data are presented as the percent of increase above background fluorescence (100%) observed in the control A549 cell line [67].

Visualization of DQ gelatin assay (zymography in situ assay) was performed as described by us in [67]. Briefly, cells were grown on FITC gelatine-coated chamber slides (Life Technologies, Waltham, MA, USA) until 60–70% confluency and were subsequently incubated with Hoechst 33342 (Molecular Probes/Life Technologies, Waltham, MA, USA) for 15 min in the incubator. The cells were washed with PBS (3×), fixed for 10 min in CellFIXTM (1% formaldehyde, 0.35% methanol, 0.09% sodium azide) from BD Biosciences (cat no. 340181) for 10 min, washed with PBS (3 × 5 min), and blocked with 3% BSA/PBS at RT for 1 h. After washing with PBS, the slides were incubated with F-actin probe—Texas Red-X Phalloidin (ThermoFisher Scientific # T7471) at RT for 20 min in the dark. The slides were washed with PBS and mounted with Mowiol (Sigma-Aldrich, St. Louis, MO, USA), and the cells were visualized under a confocal microscope (Nikon D-Eclipse C1; Nikon, Tokyo, Japan) with a 40× objective and were analyzed with EZ-C1 version 3.6 software (Tokyo, Japan).

#### *4.10. Transwell Invasion Assay*

Transwell invasion assay was performed using our modified protocol [57,67]. Briefly: Nunc Cell Culture Inserts (transwell) with 8.0µm pore diameter (#141006) were covered with 50 µL 0.2% gelatin 1 h 37 ◦C. Next, gelatin was carefully removed. A549 or A549CisR cells were treated with or without 3 µM #19 for 24 h. Then, they were trypsinized, washed 2× with medium, and transferred (2 × 10<sup>5</sup> cells/chamber) to upper chamber of Nunc Cell Culture Inserts in 0.1% BSA medium—supplemented with or without 3 µM #19. Full medium in lower chamber was used as chemoattractant. Next, medium and the gelatin from the top surface of the membrane were removed, invaded cells on the bottom surface of the membrane were washed 2× with PBS and then fixed for 5 min with CellFIXTM. Cells were dyed at RT 15 min with Hoechst 33,342 (Molecular Probes/Life Technologies, Waltham, MA, USA). Finally, membranes were cut out from chambers and placed on microscope glass, and number of cells that migrate into the membrane was counted in 5 random spots.

#### **5. Conclusions**

The presence of significant drug resistance, as well as the combination of cisplatin resistance and increased metastatic potential in the case of NSCLC, remains an obstacle during anticancer therapy. Chemotherapy can result in the selection of highly resistant and metastatic cancer cells, and this may be one of the reasons why patients present a high number of secondary tumors after apparent post-chemotherapeutic reemission [54]. Thus, long-term therapeutic strategies should focus on adjuvant "adaptive therapies" before resistance emerges [100]. Our research indicates that DSF and the tested ITCs have no impact on NSCLC cell viability at the tested concentrations, but significantly repress the cisplatin resistance of both cisplatin-resistant and metastatic NSCLC cells. Furthermore, supplementation with the tested ITCs significantly decreased the metastatic potential of all tested NSCLC models, reversing EMT toward an epithelial phenotype and decreasing the migratory potential as well as the proteolytic and invasive abilities. Therefore, the tested ITCs, especially ITC **19**, possess anti-drug-resistant and anti-metastatic potential, which may be of value during cisplatin-based chemotherapy.

**Supplementary Materials:** The supporting information can be downloaded at: https://www.mdpi. com/article/10.3390/ijms23158644/s1.

**Author Contributions:** J.K. (Jolanta Kryczka) generated the A549CisR cell line, designed the research, carried out all experiments, prepared the first draft of the manuscript; J.K. (Jakub Kryczka) participated in A549CisR cell-line generation, performed GEO database analysis, participated in data evaluation and interpretation of migration assay; J.B. critically revived the manuscript; Ł.J. synthesized ITCs; A.G. synthesized ITCs; B.K. participated in the substantive evaluation and discussion of the results; A.F. performed molecular modeling; E.B.-L. designed the research, mentored and developed the protocols, provided data evaluation and interpretation. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the National Center for Research and Development (Warsaw, Poland) within an InterChemMed grant (POWR.03.02.00–00-I029/16). This research was funded by the Medical University of Lodz (Statute No. 503/1-013-02/503-11-001-19-00).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Gene Expression Omnibus dataset #GSE108214 is available at https: //www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE108214, accessed on 21 July 2022.

**Acknowledgments:** The authors Ł.J. and A.G. express their gratitude to Tadeusz Gajda for their helpful discussions.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


### *Article* **The Multidirectional Effect of Azelastine Hydrochloride on Cervical Cancer Cells**

**Ewa Trybus \* , Teodora Król \* and Wojciech Trybus**

Department of Medical Biology, The Jan Kochanowski University, Uniwersytecka 7, 25-406 Kielce, Poland; wojciech.trybus@ujk.edu.pl

**\*** Correspondence: ewa.trybus@ujk.edu.pl (E.T.); teodora.krol@ujk.edu.pl (T.K.)

**Abstract:** A major cause of cancer cell resistance to chemotherapeutics is the blocking of apoptosis and induction of autophagy in the context of cell adaptation and survival. Therefore, new compounds are being sought, also among drugs that are commonly used in other therapies. Due to the involvement of histamine in the regulation of processes occurring during the development of many types of cancer, antihistamines are now receiving special attention. Our study concerned the identification of new mechanisms of action of azelastine hydrochloride, used in antiallergic treatment. The study was performed on HeLa cells treated with different concentrations of azelastine (15–90 µM). Cell cycle, level of autophagy (LC3 protein activity) and apoptosis (annexin V assay), activity of caspase 3/7, anti-apoptotic protein of Bcl-2 family, ROS concentration, measurement of mitochondrial membrane potential (∆ψm), and level of phosphorylated H2A.X in response to DSB were evaluated by cytometric method. Cellular changes were also demonstrated at the level of transmission electron microscopy and optical and fluorescence microscopy. Lysosomal enzyme activities-cathepsin D and L and cell viability (MTT assay) were assessed spectrophotometrically. Results: Azelastine in concentrations of 15–25 µM induced degradation processes, vacuolization, increase in cathepsin D and L activity, and LC3 protein activation. By increasing ROS, it also caused DNA damage and blocked cells in the S phase of the cell cycle. At the concentrations of 45–90 µM, azelastine clearly promoted apoptosis by activation of caspase 3/7 and inactivation of Bcl-2 protein. Fragmentation of cell nucleus was confirmed by DAPI staining. Changes were also found in the endoplasmic reticulum and mitochondria, whose damage was confirmed by staining with rhodamine 123 and in the MTT test. Azelastine decreased the mitotic index and induced mitotic catastrophe. Studies demonstrated the multidirectional effects of azelastine on HeLa cells, including anti-proliferative, cytotoxic, autophagic, and apoptotic properties, which were the predominant mechanism of death. The revealed novel properties of azelastine may be practically used in anti-cancer therapy in the future.

**Keywords:** azelastine; oxidative stress; autophagy; apoptosis; mitotic catastrophe

#### **1. Introduction**

Drug resistance is a very big problem in most advanced cancers [1,2]. The biggest obstacle in cancer chemotherapy, including the treatment of cervical cancer, is resistance to cisplatin, among others, resulting from the induction of autophagy and inhibition of tumor cell apoptosis [3,4]. The process of programmed cell death can also be inhibited during oncogenesis. Cancer cells with multiple genetic and epigenetic alterations avoid apoptosis, which is initially triggered by the transformation process itself and then by the unfavorable tumor environment and the implemented therapy [5,6]. The resulting limitations in cancer therapy contribute to increased mortality. Therefore, recently, a new trend in worldwide research has become the search for alternative treatments, also inducing other types of cell death, especially among compounds already used in other therapies [5,7–9], an example of which may be antihistamines (AHs).

**Citation:** Trybus, E.; Król, T.; Trybus, W. The Multidirectional Effect of Azelastine Hydrochloride on Cervical Cancer Cells. *Int. J. Mol. Sci.* **2022**, *23*, 5890. https://doi.org/ 10.3390/ijms23115890

Academic Editors: Angela Stefanachi and Marcin Majka

Received: 9 April 2022 Accepted: 23 May 2022 Published: 24 May 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Antihistamines (AHs), due to their proven strong anti-inflammatory and anti-allergenic properties, are widely used worldwide as first-line drugs in the treatment of numerous allergic diseases [10]. Their mechanism of action involves stabilization of the inactive form of histamine H1 receptor, thereby blocking the action of histamine [11–13], which, as a major mediator of inflammatory response, not only underlies many allergic diseases [14], but is also directly involved in the regulation of biological processes during the development of various types of cancer, including cervical cancer [1,15,16]. Hence, in recent years, attention has been focused on the potential antitumor properties of antihistamines, both among the long-used and new second-generation representatives. Compounds have been identified, that alone or in combination with other drugs show significant activity against various types of cancer cells, confirmed both in vitro and in clinical trials. An example is astemizole, a second-generation drug, that has been described as an inhibitor of hepatocellular carcinoma proliferation [17] as well as an inducer of apoptotic death in various human melanoma cell lines [18,19]. In the case of terfenadine, the ability to induce apoptosis in prostate cancer cells has been demonstrated [20]. In turn, new representatives of AHs, more often used in practice, i.e., loratadine and its active metabolite desloratadine, improve survival in breast cancer [21–23] and skin melanoma [24]. Additionally, desloratadine has properties to induce apoptosis of T-cell lymphoma cells [25], and loratadine interferes with cell cycle progression of human colon cancer cells by increasing their sensitivity to radiation [26], and improves survival for ovarian cancer [7].

Azelastine hydrochloride (a phthalazinone derivative) is commonly used especially in the topical treatment of respiratory diseases, i.e., in allergic rhinitis (also in the course of asthma and COPD), vasomotor rhinitis, and as part of the prophylaxis and therapy of allergic conjunctivitis [27–30]. Furthermore, recent in vitro studies have demonstrated the ability of this compound to prevent and inhibit SARS-CoV-2 infection in nasal tissue [31]. Azelastine was also included in the list of compounds that exhibit lysosomotropic properties and have the ability to accumulate in the lungs when administered systemically, which creates the potential to achieve an effective drug concentration and was therefore recommended for use in patients with SARS-CoV-2 [32]. It should be emphasized that azelastine is a new representative of the second generation of H1 receptor antagonists, characterized by a different chemical structure than other preparations from this group [33] and high selectivity to the receptor, and thus low risk of side effects and very good tolerance both in adults and children [34–38]. It was also found that this group has an equivalent or faster onset of action compared to the first generation AHs [39]. Numerous scientific studies confirm that the biological properties of H1 receptor antagonists, including azelastine, also result from the possibility of non-receptor activity [13,40–42], which offers a broad perspective for the discovery of new therapeutic properties of these compounds.

In recent years, azelastine has also been tested for anti-inflammatory [43], antibacterial [44], and antiparasitic properties [45]. In turn, little attention has been paid to research into the potential anticancer mechanisms of this compound. So far, the property of azelastine to induce apoptosis in human colorectal adenocarcinoma cells (HT-29 line) has been described, where the tested compound at concentrations of 10 µM-20 µM, independently of the receptor, decreased the expression of Bcl-2 protein and caused significant changes in mitochondria [46]. In another study [47], azelastine at a concentration of 5 µM sensitized KBV20C cells to the effects of vincristine (in combination administration), leading to decreased cell viability, arrest in G2 phase, and increased apoptosis. The results of the cited studies inspired the present study.

Therefore, due to the well-known resistance of HeLa cells to chemotherapy, which manifests itself by induction of autophagy and blockade of apoptosis, we decided to study the changes occurring in these cells under the influence of azelastine hydrochloride in the context of induction of apoptosis as well as other types of cell death as potential anticancer mechanisms of action of this compound.

#### **2. Results**

#### *2.1. Azelastine Induces Apoptosis in HeLa Cells*

Exposure of cells to azelastine resulted in an increase in the frequency of both cells in early (Annexin V-PE+/7-AAD−) and late apoptosis (Annexin V-PE+/7-AAD+).

At a concentration of 15 µM, apoptotic cells were over 26% (*p* ≤ 0.0001) and at 25 µM over 34% (*p* ≤ 0.0001) (Figure 1A,C). Azelastine at a concentration of 45 µM further increased the number of apoptotic cells to 60.13% (*p* ≤ 0.0001). Subsequent concentrations (60 µM and 90 µM) significantly increased the number of apoptotic cells, which were more than 93% and 98%, respectively (*p* ≤ 0.0001), with a clear predominance of cells with a late apoptotic phenotype.

**Figure 1.** Proapoptotic effects of azelastine hydrochloride. Control cells and treated for 48 h with azelastine at concentrations of 15 µM, 25 µM, 45 µM, 60 µM, and 90 µM. (**A**) Level of apoptosis determined by annexin V-PE/ 7-AAD staining. Live cells (annexin V-PE−/7-AAD−), cells in early (annexin V-PE+/7-AAD−) and late-stage apoptosis (annexin V-PE+/7-AAD+), and necrotic cells (V-PE−/7-AAD+). (**B**) Changes in 3/7 caspase activity. Live cells (caspase 3/7-/7-AAD−), cells in early (caspase 3/7+/7-AAD−) and late apoptosis (caspase 3/7+/7-AAD+), dead cells (caspase 3/7-/7-AAD+). (**C**) Percentage of apoptotic cells dependent on azelastine concentration. (**D**) Cell viability as determined by the MTT assay. (**E**) Changes in nuclei of cells labeled with 4′ ,6-diamidino-2-phenylindole (DAPI). Control cells showing normal cell nuclei morphology. Cells treated with azelastine showing changes typical of apoptosis, i.e., marked condensation of chromatin and fragmentation of cell nucleus. Images were taken at 4000× magnification. Data representative of three parallel experiments correspond to mean values ± standard error (SE). Differences were statistically confirmed at: \*\*\* *p* < 0.001.

Moreover, microscopic analysis (DAPI staining) showed that azelastine induces typical nuclear changes for apoptosis i.e., chromatin condensation and nuclear fragmentation, especially at concentrations of 60 µM and 90 µM (Figure 1E). The obtained results were dependent on the concentration of the test compound.

#### *2.2. Azelastine Induces Caspase 3/7-Dependent Apoptosis*

As shown in Figure 1B,C, caspase 3/7 activity increased significantly (*p* ≤ 0.0001) under the azelastine concentrations used. At 15 µM and 25 µM, cells with active caspase were more than 24% and 26%, respectively, and more than 55% at 45 µM concentration. The highest caspase 3/7 activity was shown for concentrations of 60 µM (96.4%) and 90 µM (98.7%). These results indicate a caspase 3/7 activation-dependent proapoptotic effect of azelastine.

#### *2.3. Azelastine Inhibits the Viability of HeLa Cells*

The MTT assay showed a highly statistically significant (*p* ≤ 0.0001) reduction in the ability of the cells to reduce the dye compared to the control, which was taken as 100% (Figure 1D). Already at the lowest concentration of 15 µM, the cell viability was 86% and at subsequent concentrations of 25 µM and 45 µM, it decreased significantly to 68.33% and 51.33%. However, the lowest percentage of living cells was obtained as a consequence of the highest concentrations of the test compound, i.e., 60 µM (8%) and 90 µM (about 4%). Azelastine inhibited mitochondrial metabolic activity to a concentration-dependent degree, which was also indicative of mitochondrial membrane damage.

#### *2.4. Azelastine Generates ROS Inducing Changes in Mitochondrial Structure and Induces Endoplasmic Reticulum Stress*

Compared to the control image (Figure 2(A1)), mitochondria with clear matrix and an irregular arrangement of mitochondrial cristae were observed already at the lowest azelastine concentration (15 µM) (Figure 2(A2)). As a result of 25 µM concentration, mitochondria showed a significant enlargement, a highly clear matrix, and a reduction in the mitochondrial cristae. Swollen channels of the rough endoplasmic reticulum were also visible in their close proximity (Figure 2A(3,3a)). In turn, the cytoplasm of cells exposed to 45 µM azelastine (Figure 2A(4,4a)) was dominated by swollen mitochondria with a strongly clear matrix, with disorganization of the inner mitochondrial membrane and pronounced damage to the cristae. Mitochondria were also characterized by disruption of the mitochondrial membrane, resulting in leakage of matrix contents into the cytoplasm (Figure 2(A4)). In contrast, visible in the microphotographs the rough endoplasmic reticulum appeared as dilated channels (Figure 2A(4,4a)). At subsequent azelastine concentrations of 60 µM and 90 µM (Figure 2A(5–6a)), the mitochondria were characterized by increased structure disorganization indicating significant damage, and in the vicinity of these structures remained the rough endoplasmic reticulum in the form of strongly widened and swollen cisterns. Compared with the control group, in which ROS (+) cells constituted 3.29%, treatment of HeLa cell with azelastine resulted in concentration-dependent intracellular ROS production (Figure 2C,D). The concentrations of 15 µM and 25 µM showed a small but statistically significant increase in ROS (+) cells to 12.4% and 24.8%, respectively (*p* ≤ 0.0001). Increasing the azelastine concentration to 45 µM resulted in increased generation of reactive oxygen species, ROS (+) cells accounted for more than 45% (*p* ≤ 0.0001). Significant levels of cellular ROS (+) were observed following azelastine treatment at concentrations of 60 µM (48.93%) and 90 µM (49.99%) (*p* ≤ 0.0001). The induction of reactive oxygen species generation correlated with a progressive decrease in mitochondrial membrane potential (Figure 2E,F).The lowest percentage of cells with mitochondrial membrane depolarization was shown at a concentration of 15 µM (9.76%) (*p* ≤ 0.0001). At 25 µM and 45 µM, cells with reduced mitochondrial membrane potential were 14.05% and 15.89%, respectively (*p* ≤ 0.0001). The highest percentage of cells with mitochondrial membrane depolarization (being more than 50%; *p* ≤ 0.0001) was found for concentrations of 60 µM and 90 µM. These results were confirmed in the imaging of rhodamine 123-labeled mitochondria, as it was shown that with increasing azelastine concentration, there is a gradual extinction of green fluorescence emission, significant in the range 45–90 µM (Figure 2B). The azelastine-induced increase in the level of reactive oxygen species, contributed to increased oxidative stress and stress of the rough endoplasmic reticulum and to the induction of apoptotic changes.

**Figure 2.** Changes in mitochondria, induction of oxidative stress, and endoplasmic reticulum stress in HeLa cells caused by the action of azelastine hydrochloride. Control cells (**A1**). Azelastine concentration-dependent ultrastructural changes indicative of apoptosis (**A2**–**6a**); brightening of the matrix and irregular arrangement of cristae in the mitochondria of cells subjected to the 15 µM concentration (**2**); at 25 µM, visible enlarged mitochondria with reduction of mitochondrial cristae remaining in close proximity to the dilated channels of the rough endoplasmic reticulum (**3**,**3a**); at a concentration of 45 µM, mitochondria with enhanced damage characteristics are present, i.e., strongly enlarged with disruption of the mitochondrial membrane (**4**) and damaged mitochondrial cristae (**4a**) and altered rough endoplasmic reticulum in the form of dilated channels (**4**,**4a**); at concentrations of 60 µM (**5**,**5a**) and 90 µM (**6**,**6a**), visible mitochondria with severe disorganization of the structure indicating damage, and rough endoplasmic reticulum located in their vicinity with strongly enlarged and swollen cisterns. Explanation of abbreviations: N—nucleus, M—mitochondria, AG—Golgi apparatus, RER—rough endoplasmic reticulum, AV—autophagic vacuoles, Lp—primary lysosomes, Ls—secondary lysosomes. Images were taken at 11,500× magnification. (**B**) Gradual and azelastine concentration-dependent loss of green fluorescence derived from rhodamine 123-labeled mitochondria. (**C**) Generation of reactive oxygen species and (**D**) percentage of ROS (+) cells as a result of azelastine. (**E**) Changes of mitochondrial membrane potential (∆ψm) and the percentage of cells with mitochondrial membrane depolarization (**F**) at different azelastine concentrations. Each sample was analyzed in triplicate. The differences were statistically confirmed at: \*\*\* *p* < 0.001.

#### *2.5. Azelastine Induces DNA Damage*

The 48 h effect of azelastine induced a concentration-dependent increase in phosphorylated H2A.X in response to DNA double strand breaks (DSBs) (Figure 3B,C). The increase in DNA damage was as follows for the subsequent concentrations: 15 µM (12.9%, *p* ≤ 0.0001), 25 µM (18.38%, *p* ≤ 0.0001), 45 µM (24.44%, *p* ≤ 0.0001), and 60 µM (28.92%, *p* ≤ 0.0001). At the highest concentration of 90 µM, cells with DBS accounted for more than 30% (*p* ≤ 0.0001) of all analyzed cells compared to the control (2.79%). DNA damage in azelastine-treated cells could have led to apoptosis.

**Figure 3.** Bcl-2 protein inactivation and DNA damage demonstrated in cells exposed to 48 h action of azelastine hydrochloride. (**A**) Cells expressing Bcl-2 are clustered in the top two quadrants of the dot plot (inactivated and activated). Over 60% are dephosphorylated after treatment with azelastine at 60 µM and 90 µM, confirming inactivation of the Bcl-2 signaling pathway. (**B**) Azelastine at concentrations of 15–90 µM generates DNA damage that induces DNA repair mechanisms such as *γ*H2AX. (**C**) Azelastine concentration-dependent percentage of cells with double-stranded DNA damage and (**D**) percentage of cells with Bcl-2 protein inactivation. Data representative of three parallel experiments correspond to mean values ± standard error (SE). Differences were statistically confirmed at: \*\*\* *p* < 0.001.

#### *2.6. Azelastine Inactivates the Bcl-2 Protein*

Azelastine induced inactivation of the anti-apoptotic protein Bcl-2 in HeLa cells (Figure 3A,D). At concentrations of 15 µM and 25 µM, cells with inactivated Bcl-2 protein represented 13.17% and 22.47% (*p* ≤ 0.0001), respectively, relative to the control (3.1%). Progressive changes in the expression of the test protein were shown at higher concentrations, i.e., 45 µM (40.56%), 60 µM (62.82%), and 90 µM (65%) (*p* ≤ 0.0001), which indicates a mechanism of proapoptotic action of azelastine involving the mitochondrial pathway.

#### *2.7. Azelastine Enhances Vacuolization and Apoptotic Changes in HeLa Cells—Morphological Evaluation*

In cells exposed to 48 h of azelastine, a significant concentration-dependent increase in the number of cells with vacuolization changes in the cytoplasm was observed (Figure 4D). Compared to control values (16 cells), the highest number of cells with enhanced vacuolization was observed at 15 µM (2119 cells) and 25 µM (2010 cells) (*p* ≤ 0.0001). Within the vacuole, a strong eosinophilic material destined for degradation was visible (Figure 4A(2,2a,3)). In contrast, at higher concentrations of the test compound, vacuolization changes showed a decreasing trend (Figure 4D). A lower but equally highly statistically significant result (1282 cells) was shown at a concentration of 45 µM (Figure 4A(4,4a)). On the other hand, at concentrations of 60 µM and 90 µM (Figure 4A(5–6a)), the presence of the lowest number of vacuolized cells was confirmed.

**Figure 4.** Morphological changes indicating the induction of vacuolating and apoptotic changes and a decrease in the dividing capacity of Hela cells as a consequence of 48 h treatment with azelastine hydrochloride. (**A1**) Control cells with normal morphology, including interphase cells and cells with

multiple mitotic figures. (**A2**–**6a**) Cells treated with azelastine at concentrations of 15–90 µM; (**A2**–**3a**) cells with numerous vacuolization changes in the cytoplasm, strongly eosinophilic material is visible within the vacuoles, which is destined for degradation, indicating the process of autophagy; (**A4**,**4a**) cells with intensive cytoplasmic vacuolization and a pyknotic cell nucleus with visible partial chromatin condensation; (**5**–**6a**) strong pro-apoptotic effect of azelastine at concentrations of 60 µM and 90 µM, expressed by the presence of numerous apoptotic cells and cells with efferocytosis. Explanation of markings: 1—interphase, 2—prophase, 3—prometaphase, 4—metaphase, 5—telophase, 6—cytokinesis, 7—vacuolization of cytoplasm, 8—vacuoles with visible digestion material, 9—apoptotic cells, 10—binucleated cells in apoptosis, 1—-binucleated cells with vacuolization, 12—giant cells, 13—abnormal segregation of chromosomes, 14—multinucleated cells in apoptosis, 15—cells with features of vacuolization and apoptosis, 16—multinucleated cells with vacuolization, 17—efferocytosis, 18—cells with phagocytosed material (by efferocytosis), which were directed toward the apoptotic pathway. Hematoxylin and eosin staining. Images were taken at 4000× magnification. (**B**) Cell cycle changes of HeLa line cells treated for 48 h with azelastine at concentrations of 15–90 µM analyzed by flow cytometry. (**C**) Percentage of cells in different phases of the cell cycle indicating blocking of cells in S phase. (**D**) Azelastine concentration-dependent number of vacuolated and apoptotic cells; at concentrations of 15–25 µM, azelastine induced vacuolization changes; at a concentration of 45 µM, there was a reduction in vacuolization changes in favor of apoptotic cells, while at concentrations of 60–90 µM, azelastine promoted apoptosis. (**E**) Changes in the mitotic index indicating an inhibition of the dividing capacity of azelastine. Average values from three independent experiments. The differences were statistically confirmed at: \*\*\* *p* < 0.001.

The action of azelastine on tested cells resulted in the simultaneous appearance of cells with apoptotic changes such as reduced size, increased cytoplasm staining, pyknotic nucleus with chromatin condensation, and the presence of apoptotic bodies. At concentrations of 15 µM and 25 µM, apoptotic cells accounted for 236 and 963 (*p* ≤ 0.0001), respectively (Figure 4A(2–3a),D), compared to the control (11 cells). However, at 45 µM, the number increased significantly to 1708 (*p* ≤ 0.0001) (Figure 4A(4,4a),D). The highest value was observed at 60 µM (2988 cells) and 90 µM (2992 cells) (*p* ≤ 0.0001) (Figure 4A(5–6a),D). It should be noted that at a concentration of 45 µM (Figure 4A(4,4a)), there were cells with simultaneously observed features such as enhanced vacuolization of the cytoplasm and a pyknotic cell nucleus with partial chromatin condensation, indicating a gradual switch from vacuolization to apoptotic changes. The presence of phagocytosed apoptotic cells was observed within the cytoplasm of living cells (Figure 4A(4a,5a)) and those that were directed into the apoptosis pathway (Figure 4A(6,6a)), indicating induction of the efferocytosis process.

#### *2.8. Azelastine Blocks Cells in S Phase and Reduces Mitotic Index*

Cytometric analysis (Figure 4B,C) showed a statistically significant (*p* ≤ 0.0001) increase in the number of cells arrested in the S phase of the cell cycle, progressing with azelastine concentration. At the concentration of 15 µM, these cells accounted for 34.23%. Slightly higher results were obtained at a concentration of 25 µM (40.64%) and at 45 µM (44.47%). However, at 60 µM and 90 µM, there was a 2-fold increase in the number of cells in the above-mentioned phase as compared to the control (28.77%). At the same time, in the concentration range of 25–90 µM, there was a significant reduction in the number of cells in the G0/G1 phase (*p* ≤ 0.0001).

Comparison of cells incubated with azelastine at all concentrations used (15–90 µM) with cells from the control group (considered as 100%) showed statistically significant (*p* ≤ 0.0001) decrease in mitotic index (Figure 4E). Already at a concentration of 15 µM azelastine, the dividing capacity of the cells decreased significantly to 22% and this was also the highest recorded result, because at the other concentrations of the test compound, i.e., 25 µM, 45 µM, 60 µM and 90 µM, the mitotic index decreased to, respectively, 7%, 3%, 2%, and 1%. These changes demonstrate the antiproliferative properties of azelastine.

#### *2.9. Azelastine Induces Mitotic Catastrophe*

Morphological analysis showed that azelastine at 15 µM resulted in changes considered morphological markers of mitotic catastrophe (Figure 5). These included multiple abnormalities occurring during mitotic division, such as the presence of anaphase bridges (Figure 5(A2)), tripolar metaphase (Figure 5(A1)), and pentapolar anaphase (Figure 5(A3)). Azelastine also induced the formation of micronuclei (micronucleation) (Figure 5B), which were present in the highest and also statistically significant numbers at a concentration of 15 µM (Figure 5A(3–6)). Furthermore, the data obtained indicated clear multinucleation due to the action of the test compound (Figure 5B). The highest results were recorded at a concentration of 15 µM with 372 binucleated cells, 132 multinucleated cells, and 23 giant cells (at *p* ≤ 0.0001). At the next concentration of 25 µM, the results remained high in the range of statistically significant values, there were 267 binucleated cells, 90 multinucleated cells, and 12 giant cells found. However, at 45 µM azelastine, the number of binucleated, multinucleated, and giant cells significantly decreased to 47, 20, and 9, respectively, while at high concentrations of 60 µM and 90 µM, it was further reduced to levels below control values (Figure 5B).

Of note are the vacuolization (Figure 5C) and apoptotic (Figure 5D) changes observed simultaneously in cells with multinucleation. At low concentrations of azelastine (15 µM and 25 µM), vacuolization changes predominated over apoptotic ones, whereas at 45–90 µM, bi- and multinucleated cells were directed towards the apoptotic pathway. The results indicate that azelastine induces mitotic catastrophe, which precedes the onset of apoptosis.

#### *2.10. Azelastine Enhances Degradation Processes*

Analysis of changes at the ultrastructural level revealed numerous autophagic vacuoles in the cytoplasm of cells with azelastine at a concentration of 15 µM (Figure 6A(1–1b)); vacuoles were differentiated in size and content indicating different stages of degradation. In the studied cells, expanded Golgi apparatuses and dilated channels of the rough endoplasmic reticulum were present; these changes indicated the intensification of the process of synthesis of proteins crucial for subsequent stages of intracellular digestion. The presence of numerous mitochondria (Figure 6(A1a)) in the tested cells may result from the increased demand for ATP necessary for the macroautophagy process. Also at 25 µM concentration (Figure 6A(2–2b)), numerous and highly enlarged autophagic vacuoles containing material at different stages of degradation were shown, and vacuoles at the formation stage were also present (Figure 6(A2b)). In the lumen of these structures, large fragments of the cytosol with organelles were visible (Figure 6A(2,2b)), and some vacuoles took the form of emptiness and clearly demarcated from the cytoplasm spaces (Figure 6(A2a)). Swollen mitochondria (Figure 6(A2a)), dilated channels of rough endoplasmic reticulum, and reduced Golgi apparatus (Figure 6(A2)), whose membranes could be used for vacuole formation, were also observed in the cells. In contrast, in cells exposed to azelastine at 45 µM concentration (Figure 6A(3–3b)), the number of autophagic vacuoles was reduced; however, they had different shapes and covered a large area of the cytosol. In addition, altered mitochondria and single, slightly dilated channels of rough endoplasmic reticulum were present within the cytoplasm of these cells. When cells were treated with high concentrations of azelastine 60 µM and 90 µM (Figure 6A(4–5b)), the presence of secondary lysosomes was clearly marked alongside altered cell nuclei with local chromatin condensation (Figure 6(A4a), an expanded nuclear envelope (Figure 6A(4a,5,5a)), often with features of fragmentation (Figure 6(A5b)). There were also single, damaged mitochondria (Figure 6A(5a,5b)) and dilated channels of rough endoplasmic reticulum (Figure 6A(4–5a)). The demonstrated changes were dependent on the concentration of azelastine and indicated the intensification of the degradation processes. The progressive degradation observed at high concentrations may indicate a switch of cellular metabolism with the possibility of triggering programmed cell death.

**Figure 5.** Morphological markers of mitotic catastrophe induced in HeLa cells exposed to 48 h treatment of azelastine hydrochloride. (**A**) Intensified changes (multipolar mitosis, micronucleation, and multinucleation) in cells treated with 15 µM azelastine. Explanation of markings: 1—cells with vacuolization, 2—binucleated cells with vacuolization, 3—apoptotic giant cell, 4—multinucleated cells with vacuolization, 5—tripolar metaphase, 6—apoptotic cells, 7—cells with micronuclei, 8—anaphase bridges, 9—multinucleated cells with micronuclei, 10—efferocytosis, 11—pentapolar anaphase, 12—giant cells, 13—binucleated cells in apoptosis, 14—multinucleated cells in apoptosis, 15—giant cells with vacuolization and micronuclei. Hematoxylin and eosin staining. Images were taken at a magnification of 4000×. (**B**) Distribution of cells with micronuclei, bi-, multinucleated cells, and giant cells at different concentrations of azelastine 15–90 µM. Azelastine concentration-dependent induction of vacuolization (**C**) and apoptotic (**D**) changes in bi-, multinucleated, and giant cells. Average values from three independent experiments. The differences were statistically confirmed at: \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001.

**Figure 6.** The intensification of degradation processes in HeLa cells with azelastine hydrochloride. (**A**) Ultrastructural changes; (**1**–**1b**) in the cytoplasm of cells very numerous autophagic vacuoles filled with material at various stages of degradation can be seen, and an extensive Golgi apparatus and dilated channels of the rough endoplasmic reticulum; (**2**–**2b**) clear changes in cells indicating a progressive degradation process, i.e., multiple vacuoles at different stages of digestion; (**3**–**3b**), the image shows vacuoles differentiated in terms of shape, covering large areas of the cytosol, indicating intensive degradation; (**4**–**5b**) dominant changes in the form of numerous primary and secondary lysosomes and changes at the level of the cell nucleus (local chromatin condensation, enlarged nuclear envelope and fragmentation). The changes obtained in the concentrations of 60 µM and 90 µM of azelastine indicate a progressive degradation as well as the initiation of cell death by apoptosis. (**B**) Histograms of cells indicating azelastine-induced autophagy as manifested by an increase in fluorochrome fluorescence intensity (red area) indicating LC3 protein activation. In the concentration range of 45–90 µM, the fluorescence intensity decreased, which argued for a switch of autophagy to apoptosis. Cells were stained with anti-LC3/Alexa Fluor® 555 conjugated antibody and the fluorescence intensity was measured cytometrically. (**C**) Changes in cathepsin D and L activities (mean ± SE) in the lysosomal fraction of HeLa cells after 48 h of exposure to different concentrations of azelastine. Results are the average of three independent experiments. Differences were statistically confirmed at: \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001.

#### *2.11. Azelastine Activates Cathepsin D and L*

As shown in the study, the effect of azelastine compared to control (which was taken as 100%) resulted in concentration-dependent changes in cathepsin D and L activity (*p* ≤ 0.0001) (Figure 6C). The highest increase in the activity of enzymes to 179.96% and 177.54% occurred at the concentrations of 15 µM and 25 µM, respectively. At 45 µM, the enzymes activity was found to be 173.89%. Further increase in the concentration of the test compound to 60 µM and 90 µM resulted in reduction of cathepsin D and L activities to 144.33% and 120.53%, respectively. The behavior of lysosomal enzymes is a reflection of the degradation processes activated by azelastine.

#### *2.12. Azelastine Induces Autophagy by Increasing LC3 Protein Levels*

According to the principle of the assay used, LC3 is a cytoplasmic protein involved in autophagosome formation during autophagy, which is translocated from the cytoplasm to the interior of autophagosomes, and its fluorescence is monitored cytometrically. According to the studies performed, azelastine induced autophagy depending on the concentration (Figure 6B). The highest intensity of fluorescence in cells was observed at concentrations of 15 µM and 25 µM, it was 139.3% and 143.36%, respectively, compared to the cells of the control group (gray area) (48.9%). With increasing azelastine concentrations, a gradual reduction in dye emission in labeled cells was observed to 95.3% at 45 µM and to 75.2% at 60 µM. At the highest concentration used (90 µM), a further reduction of the fluorescence intensity to 46.9% was demonstrated.

#### **3. Discussion**

Despite continuous advances in anticancer therapy, low treatment efficacy with concomitant high side effects is still a major problem [16]. Therefore, in the search for potential chemotherapeutic agents, particular attention is paid to the safety of the drug and its good tolerability [18]. Such features may be possessed by the well-studied new-generation H1 antihistamines, which have almost completely displaced the old-generation drugs used in anti-allergic treatment [48]. Another important aspect in the search for new oncological treatment options is the complexity of the oncological disease. The success of cancer therapy is also influenced by the possibility of modulating molecular and cellular factors found in the tumor and its microenvironment [16]. Thus, the identification of compounds with multidirectional mechanisms of action is crucial for the further development of anticancer therapies [6], and azelastine, used in anti-allergy treatment, may be such a drug.

The results obtained from our study allow us to conclude that the studied compound induced in HeLa cells two important processes for anticancer therapy, namely autophagy and apoptosis (Figure 4D).

At low concentrations (15 µM and 25 µM), azelastine clearly promoted autophagy while apoptosis remained low. The induction of autophagy is indicated by an increased number of cells with intensified vacuolization of the cytoplasm (Figure 4A(2–3a),D). An important role in this process is played by the maintenance of an acidic pH inside the vacuole, which was documented by the presence of a strongly eosinophilic content within the large vacuoles of the studied cells (Figure 4A(2,3)). Adequate pH is necessary for the activity of lysosomal enzymes required to digest cellular material [49,50]. In our study, we showed that azelastine treatment caused a marked increase in the activity of lysosomal enzymes, i.e., cathepsin D and L (Figure 6C). The revealed concentration-dependent increase in lysosomal hydrolases activity was correlated with ultrastructural changes of studied cells, indicating an increase in degradative processes. The numerous autophagic vacuoles seen in the microphotographs (Figure 6A(1–2b)), which are very large and contain fragments of cytosol with organelles, indicate the presence of macroautophagy. This was also confirmed by examining the autophagy-specific marker, LC3 protein, where the highest fluorescence intensity (i.e., 139.3% and 143.36%) was found at the lowest concentrations of azelastine (15 µM and 25 µM, respectively) (Figure 6B).

As shown in Figure 4D, cells with morphological features typical of apoptosis clearly gained advantage at 45 µM azelastine, and with increasing concentrations of 60 µM and 90 µM, they constituted more than 90% of all analyzed cells. The switch of autophagy to apoptosis is documented in Figure 4A(4,4a), where cells with characteristics of both types of cell death are seen. Such a condition can be associated with progressive degradation of organelles, confirmed by the presence of giant autophagic vacuoles in cells loaded with 45 µM of azelastine (Figure 6A(3,3a)), as well as the presence of increased numbers of primary and secondary lysosomes at high concentrations (60 µM and 90 µM) of the test compound (Figure 6A(4–5b)). In the studied cells, enlarged mitochondria were visible next to the vacuoles (Figure 6A(5a,5b)), which according to the literature could be related to the increasing demand for ATP, necessary for enhanced autophagy as well as for triggering programmed cell death [51]. The nuclei of the cells also showed altered morphology, including chromatin condensation and fragmentation (Figure 6A(4a,5b)), which was confirmed by DAPI staining (Figure 1E). The pro-apoptotic effect was additionally confirmed by the cytometric method; it was shown that azelastine, depending on the concentration, significantly induced the number of apoptotic cells with the dominance of the late-apoptotic phenotype. These values increased as follows: up to 60% at a concentration of 45 µM, 93% at 60 µM, and 98% at 90 µM (Figure 1A,C).

Autophagy and apoptosis are interconnected and can occur in the same cell in response to a given stimulus simultaneously or separately [52,53]. According to Fimia and Piacentini [54], induction of apoptosis is often associated with increased autophagy. In the presence of apoptotic stimuli, autophagy may be an adaptive response or a distinct type of cell death [55]. The tendency to change the regulation of both processes demonstrated in our studies was dependent on the concentration of azelastine. The targeting of cells to the apoptotic pathway was likely the result of a failed attempt to restore cellular homeostasis as a consequence of increased cellular stress [56]. During excessive autophagy, mitochondria tend to show accelerated production of reactive oxygen species due to increased oxidative phosphorylation. A slight, but statistically significant increase in the ROS level (Figure 2C,D), was noted already at the lowest concentrations of the tested compound (15 µM and 25 µM). The ROS values increased significantly at the concentration of 45 µM, which could have triggered the apoptosis process in the tested cells. Similar results were obtained by the team of Nicolau-Galmés [55] in a study on terfenadine, an old generation antihistamine, which enhanced autophagy and consequently led to the induction of apoptosis.

The oxidative stress activated by azelastine in HeLa cells was correlated with the simultaneous increase in the level of phosphorylated H2A.X (Figure 3B,C). The results obtained in this study indicate the participation of ROS in inducing DNA damage, which could have been a signal to trigger apoptosis. The significantly reduced division capacity of HeLa cells (Figure 4E) and their arrest in the S phase of the cell cycle (Figure 4B,C) may be associated with the DNA damage response. As shown in the literature, cell proliferation may be crucial for tumor development and progression, and histamine may be the main mediator of this process in various types of cancer [16]. On the other hand, DNA damage and inhibition of cell proliferation are among the important mechanisms of action of anticancer drugs [57]. The various properties of azelastine demonstrated in this study can also be used in anticancer therapy.

The antiproliferative properties of azelastine are also confirmed by the mitotic catastrophe induction capacity demonstrated in studies, documented in Figure 5A showing the abnormal course of mitosis. This process was most likely induced by DNA damage, and resulted in demonstrated changes such as multinucleation and micronucleation (Figure 5A(1–6),B). Of particular note is the presence of giant, multinucleated cells (Figure 5(A4)). At high concentrations of the tested compound, cells with significant nuclear changes eventually underwent apoptosis (Figure 5D), which is considered one of the necessary final steps in the course of mitotic catastrophe. The mitotic catastrophe shows a strong mechanistic relationship with the cellular and molecular changes accompanying

carcinogenesis and therefore seems to be a preferentially stimulated process in cancer cells [58–60]. Compounds promoting mitotic death, such as azelastine, may be a promising therapeutic alternative for apoptosis-resistant cancer cells.

In cells, the functions of "stress sensors" are performed by mitochondria and they are the central executors of apoptosis [61] as well as the course of mitotic catastrophe [59]. As our results showed, the induction of apoptosis by azelastine was also associated with the mitochondrial pathway. At the level of submicroscopic studies, already at low concentrations of azelastine, mitochondria were enlarged with a clear matrix (Figure 2A(2,3,3a)). However, under the influence of high concentrations, enhanced changes were demonstrated, with significant mitochondrial damage (Figure 2A(4–6a)) and disorganization of the inner membrane. At the same time, the cytometric analysis determined the highest percentage of cells with depolarization of the mitochondrial membrane (over 50%) for the concentrations of 60 µM and 90 µM (Figure 2E,F). The violation of the mitochondrial membrane integrity was confirmed by the concentration-dependent gradual quenching of green fluorescence emissions from labeled mitochondria (Figure 2B), which was also associated with the demonstrated inactivation of the Bcl-2 protein (Figure 3A,D) and activation of the executive caspases (Figure 1B,C). We also demonstrated the cytotoxic effect of azelastine related to the reduction of metabolic activity of mitochondria using the MTT test. Depending on the concentration, this compound reduced the viability of HeLa cells (Figure 1D), which at the highest concentration (90 µM) was only 4%. According to the studies by Cornet-Masana [62] conducted on leukemia lines, mitochondria in cancer cells are characterized by numerous changes, which, according to Pathania's team [63], makes them more susceptible to therapies aimed at the metabolism of cancer cells.

The analysis performed at the level of submicroscopic changes revealed that mitochondrial disorganization is also accompanied by significant changes in the profile of the rough endoplasmic reticulum. It has been reported that even at low concentrations of azelastine, there is significant dilatation of reticulum channels (Figure 2A(3,3a)). These changes intensified as a consequence of the action of increasing concentrations of the tested compound, and at its high concentrations, they became significantly swollen (Figure 2A(5–6a)), which can be explained by the stress of the reticulum. The revealed changes in the endoplasmic reticulum homeostasis may be induced by an increased level of ROS [64,65], which is also confirmed by the results of our research.

Cao and Kaufman [66] emphasize in their works the importance of spatial and functional distribution in cells of organelles such as mitochondria and endoplasmic reticulum. Also in the analyzed electronograms, close proximity of altered mitochondria and expanded channels of rough endoplasmic reticulum was demonstrated (Figure 2A(5a–6a)), indicating a functional relationship between these organelles and which may be relevant to the processes regulating apoptotic cell death. Pro-apoptotic factors derived from mitochondria induce signals from the rough endoplasmic reticulum, which in turn cause changes in the mitochondria. On the other hand, reticulum stress can lead to mitochondrial dysfunction and consequent oxidative stress, followed by impaired homeostasis and apoptosis [67–69]. Apoptosis involving endoplasmic reticulum stress has attracted a lot of attention in recent years [64]. Mild stress of cancer cells can lead to the activation of adaptive mechanisms, however, therapeutic benefits of compounds that induce endoplasmic reticulum stress and put cells on the apoptosis pathway have been confirmed for certain types of cancer cells [70,71]. In our studies, azelastine induced in HeLa cells oxidative stress, stress of rough endoplasmic reticulum, and mitochondrial dysfunction, which by reinforcing each other, disrupted cellular functions and activated proapoptotic signals [65,66,68]. A similar mechanism of action has been reported for terfenadine, an old generation antihistamine in relation to A375, HT144, and Hs294T cell lines [55]. However, in studies on the action of rupatadine, ebastine, and loratadine in relation to acute myeloblastic leukemia cells, the cytotoxicity of these compounds consisted of bidirectional, mitochondrial-lysosomal action, ROS generation, and reduction of mitochondrial metabolic activity, which led to the activation of caspase 3 and 7 and induction of the apoptosis pathway [62].

Efferocytosis-phagocytosis of dead cells was also observed under the action of azelastine (Figure 4A(4a,5,6,6a)). According to the literature, this process under certain conditions can be performed by "non-professional phagocytes" [72–74]. In the context of carcinogenesis, efferocytosis suppresses the body's natural immune response, then facilitates the immune escape of tumor cells while promoting the tumor microenvironment [50]. This process not only affects the proliferation, invasion, metastasis, and angiogenesis of cancer cells, but also regulates adaptive responses and decreases the positive response to radiotherapy and to many commonly used anticancer antibodies and chemotherapeutic agents [75]. The data obtained in our study indicate, that initially azelastine at low concentrations (Figure 4(A4a)) induced efferocytosis in the context of an adaptive response of HeLa cells, and then cells with phagocytosed apoptotic cells were directed to the apoptotic pathway (Figure 4A(6,6a)). Such action is in contrast to that of traditional therapies, which induce apoptosis of tumor cells and increase subsequent efferocytosis, which suppresses the inflammatory response [76]. Thus, the demonstrated property of azelastine indicates an additional possibility of the interference of the compound in the tumorigenesis, and at the same time, fits in the current view of combining traditional therapies with therapies targeting the efferocytosis process in order to improve their effectiveness [50].

#### **4. Material and Methods**

#### *4.1. In Vitro Culture Conditions*

Human cervical adenocarcinoma cells (HeLa cells) were cultured in a Direct Heat incubator (Thermo Scientific, Waltham, MA, USA), under standard culture conditions, i.e., 37 ◦C and 5% CO2, on a modified DMEM medium (GIBCO, New York, NY, USA), containing 10% fetal calf serum (Biowest, Nuaillé, France) and a mixture of antibiotics (penicillin G, streptomycin, amphotericin B) (Corning, Manassas, VA, USA). The HeLa cells were purchased from the American Type Tissue Culture Collection (Rockville, MD, USA). Cells were treated for 48 h with azelastine hydrochloride (≥98% HPLC), (4-[(4 chlorophenyl)methyl]-2-(1-methylazepan-4-yl) phthalazin-1-one hydrochloride), which was purchased from Sigma Aldrich (St. Louis, MO, USA). The following concentrations of the test compound were used in the experiment: 15 µM, 25 µM, 45 µM, 60 µM, and 90 µM.

According to the literature data, the tested concentrations are used in research on antihistamine drugs conducted on cancer cell lines. Control cells were cultured in complete maintenance medium without the addition of the test compound.

#### *4.2. Assessment of Cell Viability—MTT Test*

The level of cytotoxicity of azelastine against HeLa cells was determined using MTT (3-(4,5-dimethyl-2-yl)-2,5-diphenyltetrazolium bromide) reduction assay. Cells seeded in Falcon 96-well plates (Fisher Scientific, Waltham, MA, USA) after azelastine treatment were stained with MTT solution (1 mg/mL) (Sigma Aldrich, St. Louis, MO, USA). After 2 h of incubation of the cells with the dye, dimethylsulfoxide (DMSO) was applied to solubilize the formed formazan crystals. Optical density was measured at 570 nm using a Synergy 2 multi-detector microplate reader (BioTek, Winooski, VT, USA). Cell viability was calculated in comparison with the control group using Gen5 software.

#### *4.3. Visualization of Apoptotic Cells under A Fluorescence Microscope*

Morphological evaluation of nuclei of control and tested cells was performed using 4 ′ ,6-diamidino-2-phenylindole (DAPI) staining. Cells cultured in dishes (Falcon, Fisher Scientific, Waltham, MA, USA) were stained with 2.5 µg/mL DAPI solution (Sigma Aldrich, St. Louis, MO, USA) for 15 min and then washed with PBS. The preparations were analyzed using a Nikon Eclipse Ti epi-fluorescence inverted microscope (Nikon Instruments Inc., Melville, NY, USA).

#### *4.4. Detection of Apoptosis*

Phosphatidylserine externalization in azelastine-exposed cells was assessed using Annexin V and Dead Cell test kit (Merck Millipore, Burlington, MA, USA). Control and azelastine treated cells were detached using 0.25% trypsin-EDTA (Corning, New York, NY, USA), centrifuged and washed with PBS. Then, cells were stained with annexin V (100 µL) for 20 min at room temperature in the dark. The fluorescence intensity was analyzed using a Muse analyzer (Merc Millipore, Burlington, MA, USA).

#### *4.5. Activity of Caspase-3/7*

The level of caspase-3/7 activation was measured using a caspase-3/7 assay kit (Merck-Millipore, Burlington, MA, USA). After 48 h of incubation with azelastine, cells were harvested by trypsinization and incubated at 37 ◦C with 5 µL of Caspase-3/7 working solution (as per protocol). Then, 150 µL of Caspase 7-AAD working solution was added to the cells. Detection of caspase-positive cells was performed using a Muse analyzer (Merck-Millipore, Burlington, MA, USA).

#### *4.6. Analysis of Ultrastructural Changes*

Cells for electron microscopy were fixed in 3% glutaraldehyde (Serva Electrophoresis GmbH, Heidelberg, Germany) followed by 2% OsO4 (Spi, West Chester, PA, USA) in cacodyl buffer. The material was then dehydrated in an ascending series of ethanol solutions (10–99.8%) and embedded in Epon 812 epoxy resin (Serva Electrophoresis GmbH, Heidelberg, Germany), followed by polymerization at 40 ◦C and 60 ◦C. The epoxy blocks were cut into ultra-thin sections on a Leica EM UC7 ultramicrotome (Leica Biosystems, Wetzlar, Germany), and the obtained preparations were further contrasted with uranyl acetate and lead citrate. Analysis was performed using a Tecnai G2 Spirit transmission electron microscope (FEI Company, Hillsboro, OR, USA) equipped with a Morada camera (Olympus, Soft Imagine Solutions, Münster, Germany). The interpretation of the changes in azelastine-exposed cells was based on the image of control cells.

#### *4.7. Measurement of the Mitochondrial Membrane Potential (*∆*ψm)*

The decrease in ∆ψm was analyzed using the Muse MitoPotential Assay kit (Merck Millipore). Cells after incubation with azelastine were resuspended in 95 µL of Muse MitoPotential working solution and incubated at 37 ◦C for 20 min. The cells were then stained with 7-AAD dead cell marker (5 µL) at room temperature for 5 min, and the cell suspension was analyzed by flow cytometry.

#### *4.8. Microscopic Evaluation of Changes in the Potential of Mitochondrial Membrane*

After 48 h incubation with azelastine, cells were fixed in 4% paraformaldehyde and then incubated for 30 min with rhodamine 123 (Sigma Aldrich, St. Louis, MO, USA) at a concentration of 5 µg/mL ethanol. The used fluorochrome binds to metabolically active mitochondria, so the fading of fluorescence is proportional to the decrease in mitochondrial membrane potential. The cells were then washed with PBS and analyzed under a Nikon A1 confocal microscope based on a Nikon Eclipse Ti inverted microscope (Nikon Instruments Inc., Melville, NY, USA) and equipped with Nikon Nis Elements AR software (Nikon Instruments Inc., Melville, NY, USA).

#### *4.9. Oxidative Stress Analysis*

The Muse Oxidative Stress Assay kit (Merck Millipore, Burlington, MA, USA) based on intracellular detection of superoxide radicals was used to investigate the level of reactive oxygen species. As according to the manufacturer's instructions, cells were treated with Muse Oxidative Stress Reagent working solution (190 µL) after 48 h incubation with azelastine. Samples were then incubated at 37 ◦C for 30 min and the percentage of gated ROS (−) and ROS (+) cells with ROS activity were analyzed.

#### *4.10. Assessment of Bcl-2 Protein Phosphorylation*

Changes in Bcl-2 phosphorylation in HeLa cells were assessed using the Muse™ Bcl-2 Activation Dual Detection Assay kit (Merck-Millipore, Guyancourt, France) according to the manufacturer's instructions. Two direct conjugated antibodies were used in the kit, i.e., phospho-specific anti-phospho-Bcl-2 (Ser70)-Alexa Fluor® 555 and a conjugated anti-Bcl-2-PECy5 antibody to measure total Bcl-2 expression levels. The degree of activation of the Bcl-2 pathway was assessed by measuring Bcl-2 phosphorylation relative to total Bcl-2 expression in the tested cells.

#### *4.11. DNA Damage Assessment*

To determine whether azelastine causes DNA damage, cells were fixed and permeabilized with Muse Fixation Buffer and Permeabilization Buffer reagents, followed by staining with anti-phospho-Histone H2A.X (Ser139) and anti-phospho-ATM (Ser1981) antibodies according to the instructions for the Muse H2A.X Activation Dual Detection kit (Millipore, Darmstadt, Germany).

#### *4.12. Cell Cycle Analysis*

Cells were analyzed using the Muse Cell Cycle Assay Kit (Merck Millipore, Burlington, MA, USA). Cells were trypsinized and centrifuged, and the obtained cell pellet was fixed in 70% ice-cold ethanol. Cells were then treated with Muse Cell Cycle Reagent (Merck Millipore, Burlington, MA, USA) for 30 min and then analyzed with a Muse analyzer (Merck Millipore, Burlington, MA, USA).

#### *4.13. Visualization of Morphological Changes and Assessment of the Dividing Capacity of HeLa Cells*

Cells were cultured on sterile coverslips in Falcon dishes (Fisher Scientific, Waltham, MA, USA) in DMEM medium supplemented with azelastine (test cells) or without test compound (control cells). Methanol-fixed cells were stained with Harris hematoxylin and eosin, then dehydrated in an ascending series of ethanol solutions and immersed in xylene. Each preparation was analyzed based on a control image, taking into account changes mainly concerning cell nucleus (presence of bi- and multinucleated cells, giant cells, cells with micronuclei, with chromatin condensation, with pyknotic nucleus), cytoplasm (increased or decreased pigmentation, vacuolization changes, presence of apoptotic bodies), and mitotic division (presence of cells in particular phases of division and abnormal mitotic figures). Quantitative and qualitative analysis of morphological changes in the studied cells and photographic documentation were performed using a Nikon Eclipse 80i microscope with Nikon NIS Elements D 3.10 software (Nikon Instruments Inc., Melville, NY, USA). The mitotic index was evaluated by determining the number of cells in each phase of mitotic division, and the result was expressed as a percentage. In preparations, 3000 cells each were analyzed in three independent experiments (9000 cells/concentration), and the final score for a given trait was the mean value.

#### *4.14. Evaluation of Cathepsin D and L Activity Levels*

After 48 h of incubation with azelastine, cells were trypsinized, resuspended in 0.25 M sucrose solution and homogenized using a Potter S homogenizer (Sartorius, Gottingen, Germany). The homogenate was initially centrifuged at an overload of 700× *g*, for 10 min. The extranuclear supernatant was then centrifuged at 20,000× *g* overload for 20 min, and the obtained lysosomal pellet was resuspended in Triton X-100 (Sigma-Aldrich, St. Louis, MO, USA). The activities of degradative enzymes, cathepsin D and L, were determined in the lysosomal fraction according to the modified Langner's method. According to the procedure, 2% azocasein (Sigma-Aldrich, St. Louis, MO, USA), 0.2 M acetate buffer pH = 5.0, and 10% TCA (+4 ◦C) were used. After incubation at 37 ◦C, samples were centrifuged, and enzyme activity was measured by colorimetric method at 366 nm using a Spekol 1500 spectrophotometer (Analytik Jena GmbH, Jena, Germany). Simultaneously,

the total protein content (at 680 nm) was determined using the Lowry's method modified by Kirschke and Wiederanders. Enzyme activity was expressed as µmol/mg protein/hour.

#### *4.15. LC3-Antibody Detection*

The level of azelastine-induced autophagy was assessed by cytometric assay using Autophagy LC3 antibody (Merck Millipore, Burlington, MA, USA). The kit includes reagent to selective membrane permeabilization (Autophagy Reagent A) that allows to distinguish between cytosolic and autophagic LC3. This is accomplished by extracting the cytosolic protein while protecting the LC3 that is translocated to and remains intact in autophagosomes. Addition of Anti-LC3 Alexa Fluor® 555 and Autophagy Reagent B to the cells allows quantification of LC3 by measuring fluorescence using flow cytometry. According to the protocol, cells were seeded in Falcon 96-well plates. Autophagy A reagent in EBSS medium (Corning, Corning, NY, USA) was then added to the cells and incubated for 4 h under CO<sup>2</sup> atmosphere, followed by washing with HBSS (Corning, Corning, NY, USA), trypsinization, and centrifugation. The supernatant was removed and anti-LC3 Alexa Fluor® 555 and Autophagy Reagent B were added to the cells and incubated on ice for 30 min in the dark. The samples were then analyzed using flow cytometry technique. Cells that were treated with serum-free medium for 4 h were used as a positive control.

#### *4.16. Statistical Analysis*

Statistical analysis of the study results was performed using one-way analysis of variance (ANOVA) with multiple post-hoc comparisons using Tukey's test. Differences were considered statistically significant at *p* < 0.05. Statistica 10.0 software (StatSoft, Krakow, Poland) was used for data analysis.

#### **5. Conclusions**

In our study we demonstrated potential anticancer properties of azelastine based on autophagic, proapoptotic, cytotoxic, or antiproliferative activity, which, taking into account safety of its application and potent anti-inflammatory properties, can be regarded as features of a compound that is part of the current canon of fight against cancer. Azelastine may be therefore an alternative method of oncological treatment, which requires further research.

**Author Contributions:** Conceptualization, E.T., W.T. and T.K.; methodology, E.T. and W.T.; software, E.T. and W.T.; validation, E.T.; formal analysis, E.T., W.T. and T.K.; investigation, E.T., W.T. and T.K.; resources, E.T., W.T. and T.K.; data curation, E.T., W.T. and T.K.; writing—original draft preparation, E.T.; writing—review and editing, E.T., W.T. and T.K.; visualization, E.T. and W.T.; supervision, E.T.; project administration, E.T., W.T. and T.K.; funding acquisition, E.T., W.T. and T.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the statutory activity of the Jan Kochanowski University in Kielce, grant No. SUPB.RN.21.248.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data that support the findings of this study are available from the corresponding author upon reasonable request.

**Acknowledgments:** We would like to thank Anna Lankoff from the Center of Radiobiology and Biological Dosimetry of the Institute of Nuclear Chemistry and Technology, Warsaw, Poland for making the cell line available for research.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


### *Article* **Arylquin 1 (Potent Par-4 Secretagogue) Inhibits Tumor Progression and Induces Apoptosis in Colon Cancer Cells**

**Yi-Ting Chen 1,2 , Tzu-Ting Tseng 1 , Hung-Pei Tsai <sup>3</sup> and Ming-Yii Huang 4,5,6, \***


**Abstract:** Colorectal cancer (CRC) is one of the most common gastrointestinal cancers worldwide. Current therapeutic strategies mainly involve surgery and chemoradiotherapy; however, novel antitumor compounds are needed to avoid drug resistance in CRC, as well as the severe side effects of current treatments. In this study, we investigated the anticancer effects and underlying mechanisms of Arylquin 1 in CRC. The MTT assay was used to detect the viability of SW620 and HCT116 cancer cells treated with Arylquin 1 in a dose-dependent manner in vitro. Further, wound-healing and transwell migration assays were used to evaluate the migration and invasion abilities of cultured cells, and Annexin V was used to detect apoptotic cells. Additionally, Western blot was used to identify the expression levels of N-cadherin, caspase-3, cyclin D1, p-extracellular signal-regulated kinase (ERK), p-c-JUN N-terminal kinase (JNK), and phospho-p38, related to key signaling proteins, after administration of Arylquin 1. Xenograft experiments further confirmed the effects of Arylquin 1 on CRC cells in vivo. Arylquin 1 exhibited a dose-dependent reduction in cell viability in cultured CRC cells. It also inhibited cell proliferation, migration, and invasion, and induced apoptosis. Mechanistic analysis demonstrated that Arylquin 1 increased phosphorylation levels of ERK, JNK, and p38. In a mouse xenograft model, Arylquin 1 treatment diminished the growth of colon tumors after injection of cultured cancer cells. Arylquin 1 may have potential anticancer effects and translational significance in the treatment of CRC.

**Keywords:** Arylquin 1; colon cancer; tumor progression; apoptosis

#### **1. Introduction**

Colorectal cancer (CRC) is the third most common malignancy in men and the second in women globally. In 2020, there were almost two million newly diagnosed cases of CRC, and it caused nearly one million deaths [1]. Wide excision for resectable cancer combined with postoperative chemotherapy remains the first choice of therapy in cases of CRC. Combined radiotherapy is now regarded as the standard treatment for advanced rectal cancer. However, not only chemotherapy but also radiation may cause unbearable side effects and toxicities to lead treatment failure due to early drug withdrawal. The development of innovative drugs that induce tumor apoptosis is the main focus of research into antineoplastic therapeutic agents. To find selective cancer-targeted therapeutics remains one of the greatest challenges.

Arylquin 1, a new molecule first identified by scientists at the University of Kentucky, is regarded as a secretagogue of prostate apoptosis response-4 (Par-4) in healthy cells. Par-4,

**Citation:** Chen, Y.-T.; Tseng, T.-T.; Tsai, H.-P.; Huang, M.-Y. Arylquin 1 (Potent Par-4 Secretagogue) Inhibits Tumor Progression and Induces Apoptosis in Colon Cancer Cells. *Int. J. Mol. Sci.* **2022**, *23*, 5645. https:// doi.org/10.3390/ijms23105645

Academic Editor: Angela Stefanachi

Received: 26 April 2022 Accepted: 17 May 2022 Published: 18 May 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

regarded a tumor specific suppressor protein, is identified in normal cells, but it is often decreased or absent in neoplastic cells to evade the destiny of apoptosis [2]. Extracellular Par-4 can induce apoptosis through the binding of the 78 kDa glucose-regulated protein (GRP78) on tumor cell membranes [3]. Intracellularly, as seen in prostate cancers, Par-4 mainly activates the signaling pathway of the Fas death receptor and inhibits cellular pro-survival mechanisms [4]. Downregulation of Par-4 protein during tumorigenesis has been proposed associated with poor prognosis and treatment resistance in many solid tumors to highlight the importance as a critical event of therapeutic target agent. For example, Par-4 level was decreased in human renal cell carcinoma compared with normal renal tubular cells [5]. Low level of Par-4 was found correlated with low survival period in glioblastoma patients, and Tamoxifen-induced cell death was alleviated by Par-4 specific siRNA in vitro [6]. In breast cancer, knockdown of Par-4 increased the proliferation and reduced the chemosensitization [7]. However, gene therapy such as transfection with plasmids or adenoviral vectors is very complex and hundreds of challenges remain to be overcome. Therefore, replenishment of Par-4 by Arylquin 1 via sensitizing tumor cells to apoptosis is promising. Previously, Arylquin 1 has been reported to promote morphology changes and decrease viability in various cancer cells [8]; however, any anticancer effects of Arylquin 1 in CRC have yet to be determined.

The present study aimed to investigate the cytotoxic effects and mechanisms of Arylquin-1-induced cell death in CRC. To our knowledge, this is the first study to explore the therapeutic value of Arylquin 1 specifically in CRC.

#### **2. Results**

#### *2.1. Arylquin-1-Attenuated Cell Viability in CRC Cell Lines*

To explore the effects of Arylquin 1 on the proliferation of cultured CRC cells, cells were incubated with Arylquin 1 at different concentrations for 72 h, after which cell viability was assessed using the MTT assay. SW620 and HCT116 cells demonstrated reduced viability at doses of 0.25, 0.5, 1, 1.5, 2, 2.5, and 3 µM Arylquin 1 relative to the 0 µM control (Figure 1). These data showed a dose-dependent reduction in proliferation after Arylquin 1 treatment. The IC<sup>50</sup> concentrations of Arylquin 1 were 1.8 and 2.3 µM for SW620 and HCT116 cells, respectively. μ μ μ

**Figure 1.** Viability of SW620 and HCT116 cells treated with different doses of Arylquin 1.

*2.2. Arylquin-1-Inhibited Cell Migration, Invasion, and Epithelial–Mesenchymal Transition (EMT) in CRC Cells*

μ To investigate the effects of cell migration after treatment with Arylquin 1, we used a wound-healing assay and compared the results between a control group and the Arylquin-1-treated groups. In both SW620 and HCT116 cells, 1, 1.5, and 2 µM Arylquin 1 markedly inhibited cell migration at 16, 24, and 48 h after administration (Figure 2). These data also indicated the dose-dependent effects of Arylquin 1 on cell migration in CRC cells.

**Figure 2.** Wound-healing assay and percentage of cell migration in cultured SW620 and HCT116 cells 0, 16, 24, and 48 h after treatment with Arylquin 1; quantitative data are expressed as mean ± SEM. \*\* *p* < 0.01 and \*\*\* *p* < 0.001 compared to the control group.

μ To evaluate cell invasion, we used a Matrigel invasion assay. In both SW620 and HCT116 cells, 1, 1.5, and 2 µM Arylquin 1 significantly inhibited cell invasion 24 h after administration (Figure 3). As in the wound-healing and proliferation assays, Arylquin 1 inhibited the cell migration in CRC cells in a dose-dependent manner. μ

**Figure 3.** Wound-healing assay and percentage of cell migration in cultured SW620 and HCT116 cells 24 h after treatment with Arylquin 1; quantitative data are expressed as mean ± SEM. \*\*\* *p* < 0.001 compared to the control group.

Cadherins are markers of metastasis used to evaluate the ability of cells to undergo an epithelial–mesenchymal transition (EMT). SW620 and HCT116 cells treated with Arylquin 1 were subjected to Western blot (Figure 4), showing that N-cadherin levels were significantly lower in Arylquin-1-treated CRC cells in a dose-dependent manner, indicating that Arylquin 1 attenuates EMT in CRC.

97

β 0 **Figure 4.** Western blot for N-cadherin, with β-actin-loading control, in SW620 and HCT116 cells treated with Arylquin 1; quantitative data are expressed as mean ± SEM; \*\* *p* < 0.01 and \*\*\* *p* < 0.001 compared to the control group.

#### *2.3. Arylquin 1 Promotes Apoptosis in Cultured CRC Cells*

μ As shown in Figure 5, the highest level of cleaved caspase-3 and the lowest level of Cyclin-D1 were detected 72 h after treatment with a dose of 2 µM Arylquin 1, indicating that apoptotic activity is positively associated with Arylquin 1 dose strength in both SW620 and HCT116 cells. Moreover, BCl2 levels showed a tendency to significantly decrease after treatment with Arylquin 1 in HCT116 cells.

β **Figure 5.** Western blot for Cyclin D1, cleaved caspase-3, and BCl2, with a β-actin-loading control, in SW620 and HCT116 cells 72 h after treated with Arylquin 1; quantitative data are expressed as mean ± SEM. \* *p* < 0.05, \*\* *p* < 0.01, and \*\*\* *p* < 0.001 compared to the control group.

μ

### 98

The Annexin V/FITC assay was performed to identify apoptotic cells. A high proportion of viable cells was observed in the untreated control group in both SW620 (93.52 ± 3.1%) and HCT116 (83.65 ± 5.37%) cells (Figure 6). In both cell lines, a pattern of the cell population shifting from viable cells to early apoptotic stages to late apoptotic stages was observed, and this was associated with the Arylquin 1 dosage. However, a significant increase in early and late apoptotic/necrotic cells was observed after treatment with 2 µM Arylquin 1 in SW620 cells (19.68 ± 5%). These results revealed that Arylquin 1 can induce apoptosis and necrosis in a dose-dependent manner.

**Figure 6.** Annexin V/FITC Assay: (**A**) dot plot of Annexin V/FITC in SW620 and HCT116 cells after treatment with Arylquin 1; (**B**) quantification of Annexin V/FITC in SW620 and HCT116 cells. EA—early apoptosis; LA/NEC—late apoptosis/necrosis. All data are expressed as mean ± SEM. \* *p* < 0.05 compared with corresponding controls.

#### *2.4. Arylquin 1 Regulated Apoptosis via the MAPK Pathway in CRC Cells*

To further clarify the signaling pathways involved and whether CRC cell viability decreased as a consequence of Arylquin-1-induced apoptosis, SW620 and HCT116 cells treated with Arylquin 1 were subjected to Western blot for proteins downstream of the mitogen-activated protein kinase (MAPK), Akt, and signal transducer and activator of transcription 3 (STAT3) pathways. The MAPK family consists of three major subfamilies of related proteins (extracellular-signal-regulated kinases [ERKs], c-Jun N-terminal kinase [JNK], and p38). In both SW620 and HCT116 cells, Arylquin 1 administration led to the upregulation of ERK, p38, and JNK expression (Figure 7).

β **Figure 7.** Western blot for p-ERK, ERK, p-p38, p38, p-JNK, JNK, p-Akt, Akt, and Stat3, with a β-actin-loading control, in (**A**) SW620 and (**B**) HCT116 cells treated with Arylquin 1; quantitative data are expressed as mean ± SEM; \* *p* < 0.05, \*\* *p* < 0.01, and \*\*\* *p* < 0.001 compared to the control group. β

#### *2.5. Arylquin 1 Suppressed the Growth of CRC Cells in Mice*

Having validated our cell model in vitro, we next evaluated the in vivo effects of Arylquin 1 on HCT116 tumor xenograft growth. Tumor or control (empty) cell suspensions were injected subcutaneously into the flanks of 6-week-old mice, and tumor growth was evaluated and registered periodically (Figure 8). Arylquin 1 was injected intraperitoneally on day 7 after cell implantation. Our results demonstrate a significant and marked reduction in tumor volume in the Arylquin 1 treatment group, from day 14 after cell injection until the end of the experiment at day 42 (all *p* < 0.01). The Arylquin 1 treatment group showed significantly slower tumor growth compared with the untreated control group, displaying final tumor volumes of 127 ± 69 and 1042 ± 157 mm<sup>3</sup> , respectively.

**Figure 8.** (**A**) Gross examination and (**B**) tumor growth curves of mouse xenografts of human colorectal cancer with and without Arylquin 1 treatment; quantitative data are expressed as mean ± SEM; \*\*\* *p* < 0.001 compared to the control group.

μ

μ

#### **3. Discussion**

CRC is one of the most commonly diagnosed cancers and is a leading cause of cancerrelated death. Currently, the chemotherapeutic strategy for the treatment of advanced CRC is based on a variety of reagents, sometimes accompanied by lethal and dose-dependent adverse side effects. Thus, it is worth exploring potential drug substitutions to increase treatment effectiveness and cause fewer severe side effects in patients with CRC.

In the present study, various assays were used to investigate the antitumor effects of Arylquin 1 in human CRC. We first identified that treatment with Arylquin 1 inhibited growth in cultured SW620 and HCT116 cells with an IC<sup>50</sup> of 1.8 and 2.3 µM, respectively. In mice injected with cultured human CRC cells, treatment with Arylquin 1 decreased tumor size. Arylquin 1 was also found to inhibit cell migration and invasion. Western blot showed that Arylquin 1 treatment led to downregulated N-cadherin expression.

Arylquin 1, a molecular compound regarded as a Par-4 secretagogue in healthy cells, was identified in 2014 by Rangnekar et al. [9]. This structure–activity study has been confirmed, with results showing that Arylquin 1 targets vimentin and causes the release of Par-4 [10]. Par-4, a 38 kDa tumor suppressor protein, is expressed in healthy cells, but it is usually downregulated or silenced in cancer cells to elude the induction of apoptosis [2]. Recently, the intracellular apoptotic effects of Par-4 and the ability of cancer cells to inhibit Par-4 release were carefully investigated [4]. Apoptosis induced by interactions between Par-4 and the tumor cell surface receptor GRP78 leads to programmed cell death [3,10]. Moreover, Min et al. reported that Arylquin 1 can induce lysosomal membrane permeabilization, with useful antitumor effects [8]. Chloroquine, another secretagogue of Par-4, has also been identified to induce apoptosis and inhibit tumor metastasis. Arylquin 1 and Chloroquine share a common pharmacophore [11]. Therefore, the potential cytotoxic activity of Arylquin 1 to bolster the release of Par-4 is a promising therapeutic advance.

In our study, the ability of inhibiting cell migration and invasion by Arylquin 1 treatment was noted. We found that Arylquin 1 attenuated N-cadherin expression in CRC. EMT has been identified to play an important role in tumor progression and metastasis. Upregulation of vimentin and N-cadherin regulated by a complex signaling pathway is the hallmark of EMT [12]. In previous studies, vimentin, which is highly involved in EMT and metastasis, was reported as the primary target of Arylquin 1 to inhibit the spread of lung cancer [9,13]. Silencing endogenous Par-4 using siRNA has also been shown to promote tumor growth and metastasis resulted in increased N-cadherin expression in pancreatic cancers [14]. Upregulation of Vimentin and N-cadherin were also detected associated with the interaction of Par-4 in cervical cancer during transforming growth factor (TGF)-β-induced EMT [15]. Furthermore, overexpression of Par-4 was proved to reduce EMT which was later promote apoptosis in pancreatic cancer [16]. It seems that our results may provide evidence that Aryqulin-1 may enhance the anti-EMT effects via Par-4 in tumor cells.

From the Annexin V/FITC assay in the present study, it was observed that both SW620 and HCT116 cells underwent apoptosis after Arylquin 1 treatment. The proportion of cells observed undergoing apoptosis increased as the dose increased. This shows that Arylquin 1 can induce apoptosis in a dose-dependent manner in both SW620 and HCT116 cells. It seems that CRC cell viability decreased is a consequence of Arylquin-1-induced apoptosis. There are two main pathways involved in apoptosis, the mitochondria-related intrinsic pathway and the death-receptor-related extrinsic pathway [17,18]. In the present study, downregulation of BCl2 and upregulation of caspase-3, which is involved in both intrinsic and extrinsic apoptosis pathways, were identified in HCT116 cells treated by Arylquin 1. Rangnekar et al. have previously shown that Par-4 upregulation induces apoptotic death in prostate cancer [19]. Wang et al. further reported a direct correlation between Par-4 levels and apoptotic activity induced by 5-fluorouracil in CRC [20]. Moreover, CRC tumor growth—namely the tumor growth rate as reflected by size after 42 days—was also suppressed by Arylquin 1 in mice in the present study, confirming the apoptosis-inducing effects of Arylquin 1.

Higher apoptosis percentage with a significant decrease in BCl2 expression was identified in HCT116 cells than SW620 cells in our data. As we know, the programmed cell death leaded by the interactions between Par-4 and the tumor cell surface receptor GRP78 has been investigated [2]. GRP78 is regarded as the central regulator of endoplasmic reticulum stress in apoptosis. Upregulation of GRP78 in tumor cells has been identified associated with reduce tumorigenicity and increase sensitivity to DNA crosslinking agents due to promote its localization in cell surface [21,22]. The oncogenic role of GRP78 draws attention to its value as a prognostic and drug response marker to mediate the therapeutic efficiency [23]. Therefore, the variable surface and cellular location of GRP78 expressions may alter and result in the different behaviors of SW620 and HCT116 cells.

In the present study, we found that both promoted apoptosis and increased expression of phosphorylated ERK in CRC treated with Arylquin 1 were observed, in addition to JNK and p38 upregulation. The MAPK signaling pathway, including ERK1/2, JNK/SAPK, and p38, is involved in cell proliferation, apoptosis, and metastasis of cancer cells, depending on the cell type and stimulus [24]. Therefore, modulating this pathway can provide a potential approach to treating CRC. ERK has been reported to promote cell growth and control migration [25,26]. On the other hand, tumor-necrosis-factor-associated apoptosis-inducing ligand and TGF-β were identified to induce apoptosis via ERK-mediated upregulation of death receptors in colon cancer cells [27,28]. Activation of JNK has also been shown to be involved in cell survival, proliferation, migration, invasion, and cell death [29,30]. Previously, ERK and JNK were reported to participate in the regulation of apoptosis (intrinsic and extrinsic; [30,31]). JNK can activate apoptotic pathway by regulating the activities of mitochondria directly or upregulating proapoptotic genes through transcriptional factors [32]. Activation of JNK pathway of apoptosis induced by Vernodalin or TRAIL was found in colon cancer cells [33,34]. p38 is also involved in the signal integration of migration, metastasis, and apoptosis in cancer [35]. In HeLa cells, Lee et al. reported that TRAIL induced apoptosis via p38 activation stimulated by reactive oxygen species [36]. Our results support the crucial role of MAPK in the regulation of proliferation, migration, metastasis, and apoptotic processes in CRC cells. Since these proteins act on tumor behavior, their dysregulation by Arylquin 1 highlights the novel mechanisms that may produce its antiproliferative, anti-invasive, and apoptotic functions.

#### **4. Materials and Methods**

#### *4.1. Cell Culture*

SW620 and HCT116 cells were purchased from the Bioresource Collection and Research Center (Taiwan) and cultured in Dulbecco's Modified Eagle Medium (DMEM) (Gibco; 12800-017, Waltham, MA; USA) with 10% fetal bovine serum (FBS) at 37 ◦C in an atmosphere of 5% CO2.

#### *4.2. Cell Viability*

Cell lines were seeded on a 24-well plate at a density of about 3 × 10<sup>4</sup> cells in 500 µL of DMEM with 10% FBS in each well. The viable cell counts were performed following the MTT assay after culturing with different doses of Arylquin 1 (0, 0.25, 0.5, 1, 1.5, 2, 2.5, and 3 µM) for 72 h to identify the IC<sup>50</sup> dose.

#### *4.3. Migration and Invasion Assays In Vitro*

Cell migration was investigated using the wound-healing assay (Cat. Nr. 80209; Ibidi GmbH, Graefelfing, Germany) using two-sided wound gaps. The wound-healing assay was performed in 24-well plates seeded with 3 × 10<sup>5</sup> cells per insert and cultured at 37 ◦C for 24 h, after which 0, 1, 1.5, or 2 µM Arylquin 1 was added to the wells. After 24 h, these cells were washed twice with phosphate-buffered saline (PBS) and photographed at 16, 24, and 48 h after washing. The transwell migration assay (COR3452; Corning Inc., Corning, NY, USA) was used to measure cell invasion in vitro. Incubated cells treated with 0, 1, 1.5, or 2 µM Arylquin 1 were seeded at 5 × 10<sup>3</sup> cells per insert, and the lower chamber of the

transwell assay was filled with 0.5 mL DMEM with 10% FBS. After 24 h, cells that remained on the upper surface of the transwell membrane were removed by a cotton swab. Cells that had passed through the transwell membrane to the bottom of the insert were fixed using formalin, stained using 0.5% crystal violet, and quantified via manual counts from photographs of six randomly selected fields.

#### *4.4. Protein Extraction and Western Blot*

All samples were prepared in 100 µL of RIPA lysis buffer, and 30 µg protein from each sample was loaded in the wells of a sodium dodecyl sulfate–polyacrylamide gel electrophoresis gel; gels were run at 80 V for 2 h. The separated proteins were then transferred from the gel to a polyvinyl difluoride membrane with a 400 mA current for 2 h. After 1 h in blocking buffer, the membranes were incubated with primary antibodies overnight, followed by incubation with secondary antibodies for 1 h. The primary antibodies used were anti-β-actin (MAB1501R; Sigma-Aldrich; Burlington, MA, USA), anti-p-p38 (9211S; Cell Signaling; Danvers, MA, USA), anti-p38 (9212S; Cell Signaling; Danvers, MA, USA), anti-p-JNK (4668S; Cell Signaling; Danvers, MA, USA), anti-JNK (9252S; Cell Signaling; Danvers, MA, USA), anti-p-ERK (9101S; Cell Signaling; Danvers, MA, USA), anti-ERK (9102S; Cell Signaling; Danvers, MA, USA), anti-p-Akt (9271S; Cell Signaling; Danvers, MA, USA), anti-Akt (9272S; Cell Signaling; Danvers, MA, USA), anti-Stat3 (sc-8019; Santa Cruz, CA, USA), anti-N-cadherin (22018-1-AP; Proteintech; Rosemont, IL, USA), anti-Cyclin D1 (60186-1-Ig; Proteintech; Rosemont, IL, USA), anti-cleaved caspase-3 (9661S; Cell Signaling; Danvers, MA, USA), and anti-BCl2 (4223S; Cell Signaling; Danvers, MA, USA). Horseradish peroxidase (HRP)-conjugated secondary antibodies were goat anti-rabbit IgG (H + L)-HRP (C04003; Croyez, Taipei, Taiwan) and goat anti-mouse IgG (H + L)-HRP (C04001; Croyez, Taipei, Taiwan). Signals were detected using an enhanced chemiluminescent solution (Western Lightning Plus; PerkinElmer, Waltham, MA, USA) with a chemiluminescence imager (Minichemi; Thermo Fisher Scientific, Waltham, MA, USA).

#### *4.5. Flow Cytometry*

Cells were transfected to 6-well plates and treated with 0, 1, 1.5, or 2 µM Arylquin 1 for 72 h. Both detached and attached cells were centrifuged at 1000 rpm for 5 min, washed once with 1× PBS, then treated according to the manufacturer's instructions using the Muse® Annexin V and Dead Cell Kit (Cat. No. MCH100105; MilliporeSigma, Burlington, MA, USA).

#### *4.6. Animal Model*

Six-week-old male NU/NU nude mice were obtained from BioLasco Taiwan (Taipei, Taiwan). All animal experiments followed the protocols of the Institutional Animal Care and Use Committee of Kaohsiung Medical University (IACUC Approval No: 110266) and were performed according to the Guiding Principles for the Care and Use of Laboratory Animals. Mice were acclimatized to a 12:12 h light/dark cycle at 24 ± 1 ◦C with ad libitum access to food and water. Human HCT116 cells were chosen for tumor xenograft to evaluate the effects of Arylquin 1 on tumor growth, invasion, and metastasis of colon cancer according to previous studies [37–39]. HCT116 cells were subcutaneously injected at a density of 1 × 10<sup>7</sup> in NU/NU mice. One week after injection, 150 uM/kg Arylquin 1 was injected intraperitoneally. Tumor volume (mm<sup>3</sup> ) was measured three times a week, calculated as (length × width<sup>2</sup> )/2 on days 7, 10, 14, 21, 28, 35, and 42. Mice were killed 42 days after the injection of tumor cells.

#### *4.7. Statistical Analysis*

All statistical analyses were performed with SPSS 19.0 (IBM Corp., Armonk, NY, USA). Intensities of Western blot bands were digitally analyzed using ImageJ software. All tests were two-sided, and a *p*-value less than 0.05 was considered statistically significant.

#### **5. Conclusions**

In summary, the present study demonstrates for the first time that Arylquin 1, as a secretagogue of Par-4, can inhibit the proliferation of CRC cells in vitro and can decrease migration, invasion, and metastasis. Moreover, promoting apoptosis in vitro and suppressing tumor size in vivo were also observed. In conclusion, Arylquin 1 may be used as a potential anticancer drug with strong translational significance in the treatment of CRC.

**Author Contributions:** Y.-T.C. conducted data analysis and manuscript drafting; the cell and animal studies were finished with the assistance of T.-T.T. and H.-P.T.; M.-Y.H. was involved in conceptualization and manuscript review editing and led the project. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by grants from the Ministry of Science and Technology (MOST 107-2320-B-037-018, MOST 110-2314-B-037-075-MY2), Kaohsiung Medical University Chung-Ho Memorial Hospital (KMUH107-7R87, KMUH108-8M66, KMUH109-9M78, KMUH110-0R72, KMUH-DK(B)110005-2), and Kaohsiung Medical University (KMU-Q108003).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

**Conflicts of Interest:** All authors declare no conflict of interest.

#### **References**

