The Effect of 1,2,4-Triazole-3-thiol Derivatives Bearing Hydrazone Moiety on Cancer Cell Migration and Growth of Melanoma, Breast, and Pancreatic Cancer Spheroids

4-Phenyl-3-[2-(phenylamino)ethyl]-1H-1,2,4-triazole-5(4H)-thione was used as a starting compound for the synthesis of the corresponding 1,2,4-triazol-3-ylthioacetohydrazide, which reacts with isatins and various aldehydes bearing aromatic and heterocyclic moieties provided target hydrazones. Their cytotoxicity was tested by the MTT assay against human melanoma IGR39, human triple-negative breast cancer (MDA-MB-231), and pancreatic carcinoma (Panc-1) cell lines. The selectivity of compounds towards cancer cells was also studied. In general, the synthesized compounds were more cytotoxic against the melanoma cell line. N′-(2-oxoindolin-3-ylidene)-2-((4-phenyl-5-(2-(phenylamino)ethyl)-4H-1,2,4-triazol-3-yl)thio)acetohydrazide, N′-((1H-pyrrol-2-yl)methylene)-2-((4-phenyl-5-(2-(phenylamino)ethyl)-4H-1,2,4-triazol-3-yl)thio)acetohydrazide and N′-(2-hydroxy-5-nitrobenzylidene)-2-((4-phenyl-5-(2-(phenylamino)ethyl)-4H-1,2,4-triazol-3-yl)thio)acetohydrazide were identified as the most active among all synthesized compounds in 3D cell cultures. N′-(4-(dimethylamino)benzylidene)-2-((4-phenyl-5-(2-(phenylamino)ethyl)-4H-1,2,4-triazol-3-yl)thio)acetohydrazide inhibited all cancer cell migration, was characterized as relatively more selective towards cancer cells, and could be further tested as an antimetastatic candidate.


Introduction
Cancer is a major global public health problem and is growing as the leading cause of mortality, accounting for vast morbidity worldwide. It is a malignant disease characterized by rapid and uncontrolled cell proliferation [1]. Pancreatic carcinoma is associated with one of the worst clinical outcomes due to its aggressive, metastatic, and drug-resistant nature. Pancreatic cancer accounts for 7% of all cancer deaths [2]. Breast cancer is the leading cause of cancer morbidity, disability, and mortality in women worldwide. It was the most diagnosed malignancy in 2020 [3]. Among skin cancers, melanoma is the most lethal form, accounting for 75% of deaths due to skin cancer, although it is diagnosed only in 4% of skin cancer cases [4].
Cytotoxic drugs are among the most important treatments used for cancer; however, their permeability and efficiency are low. Furthermore, a variety of anticancer cytotoxic drugs show different adverse side effects due to the low selectivity of the antiproliferative action and allow tumors to develop resistance to multiple chemotherapeutic drugs [5]. The search for new effective anticancer agents with superior selectivity towards cancer cells is of crucial importance [6]. Small-molecule kinase inhibitors have been widely investigated as novel anticancer therapeutics. Kinases are enzymes that transfer a phosphate group to a protein and regulate various cellular functions such as proliferation, differentiation, migration, metabolism, and angiogenesis by activating several signalling pathways [7]. Kinases have also been frequently found to be deregulated and overexpressed in cancerous tissues. Therefore, modification of the kinase activity by employing small molecules has emerged as a strategic approach for cancer treatment [8].
1,2,4-Triazole is one of the most significant nitrogen-containing scaffolds in the field of medicinal chemistry due to its diverse biological properties, including anticancer activity [9][10][11][12]. Among the azoles, triazoles are the most stable compounds and are difficult to cleave. 1,2,4-Triazoles act as important pharmacophores by interacting with biological receptors with high affinity due to their dipole character, hydrogen bonding capacity, rigidity, and solubility [13]. The 1,2,4-triazole nucleus is stable to metabolism and acts as an important pharmacophore by interacting at the active site of a receptor as a hydrogen bond acceptor and as a donor. Due to its polar nature, the triazole nucleus can increase the solubility of the ligand and significantly improve the pharmacological profile of the drug [14]. Many anticancer agents such as fluconazole, tebuconazole, triadimefon, and ribavirin bear a 1,2,4-triazole moiety, revealing their potential in the development of novel anticancer agents [15]. Triazole heterocycles that incorporate sulfur in the form of mercaptoand thione-substitution show more potency compared to their parent derivatives [16]. A variety of biological activities, such as anticancer, antimicrobial, antitubercular, hypoglycemic, etc., properties have been reported for a large number of 1,2,4-triazole-3-thione and 1,2,4-triazole- 3-thiol derivatives [17][18][19].
Isatin derivatives have been widely recognized in cancer therapeutics as protein kinase inhibitors. Sunitinib is an oxindole-containing clinically used drug for the treatment of renal cell carcinoma [26]. Another derivative of oxindole, nintedanib, was approved in 2020 in the United States for the treatment of interstitial lung diseases such as idiopathic pulmonary fibrosis and chronic fibrosis with a progressive phenotype. It has an effective antiproliferative feature, inhibiting angiokinase and restricting growth factor, consequently being amongst the most potent indolinone compounds [8]. Many researchers have exploited the isatin moiety taking advantage of NH in the first position, the carbonyl position C2, and C3 for the synthesis of various derivatives that possess different biological activities including anticancer effects in different types of cancer [27][28][29]. Diverse substitution at the C5 position has been associated with the promising activity of synthesized oxindole derivatives such as progesterone antagonist, vasopressin antagonist, anti-Alzheimer, phosphate inhibitor, kinase inhibitor, neuroprotection, NMDA blocking, and anticancer, antioxidant, antibacterial, and anti-HIV activity [30].
Based on the pharmacological implications of the moieties discussed above and as a continuation of our search for biologically active hybrid compounds that carry heterocyclic fragments [32][33][34][35], we report herein the synthesis of a series of 1,2,4-triazole-3-thiol derivatives bearing 5-substituted 2-oxindole-hydrazone moieties. Another series of hybrid compounds includes hydrazones bearing a 1,2,4-triazole-3-thiolyl moiety along with various aromatic and heterocyclic fragments. Hybrid compounds that contain pyrimidine, pyrrole, and pyrazole moieties have been extensively shown to possess antiproliferative properties [36][37][38][39][40][41]. Three human cancer cell lines, namely melanoma (IGR39), triple-negative breast cancer (MDA- , and pancreatic carcinoma (Panc-1) were chosen for this study. These tumour types are characterized as very aggressive and invasive and are usually resistant to available chemotherapeutics [42][43][44]. In order to reduce the development of cancer resistance, existing anti-cancer drugs are often combined with novel therapeutics that could improve their penetration into cancer cells, reduce the efflux from the cells, or improve the effectiveness of cancer therapy by affecting different metabolic pathways [45]. Kinase inhibitors are one of the drug groups that are used in combinatorial studies [46]. We decided to explore novel compounds in 3D cell cultures (tumor spheroids) and migration assays, and to identify the most promising ones for further development.
Hydrazone derivatives [9][10][11][12][13][14][15][16][17][18][19] were synthesized from hydrazide 3 and the corresponding aldehydes in methanol under reflux in the yield range of 29-98% (Scheme 2) [35,49]. The structures of the synthesized compounds have been confirmed by 1 H and 13 C NMR, IR, and MS data. In the 1 H NMR spectrum for 15 ( Figure S43 in the Supplementary Materials), multiplets at 7.88-8.25 integrated for two protons are attributed to the H13 proton and the one in the CH group of the pyrazole ring. In the 1 H NMR spectrum for 16 ( Figure  S46 in the Supplementary Materials), two protons in the pyrazole ring resonated as multiplets at 6.52-6.65 ppm and 7. 35-8.18 ppm (this signal overlaps with the resonances of aromatic protons and the one of the H13 proton). The presence of the methyl group attached to the N atom in pyrazole moiety was confirmed by the singlet at 3.40 ppm. The resonance attributed to the C11 carbon atom in the 13 C NMR spectra of hydrazones [9][10][11][12][13][14][15][16][17][18][19] was shifted downfield by ~1 ppm compared to the respective resonance in the 13   2-Oxindole derivatives 4-8 were synthesized from hydrazide 3 and the corresponding isatins in methanol under reflux in good yields (63-75%) [35]. In the 1 H NMR spectra of 4-8, the proton of the secondary amine group adjacent to the C 1 carbon resonated as a singlet at 5.65 ppm (Figures S10, S13, S16, S19, and S22 in the Supplementary Materials). The proton in the amide group (NHC 12 ) resonated in the range of 11. 5-13.4 ppm, while the proton singlets at the higher field in the range of 10. 8-11.9 ppm are attributed to the secondary amine group in the 2-oxindole moiety. The 1 H NMR spectra of hydrazones 4-8 display double sets of resonances of the NHC 12 protons, and a 2-oxindole NH proton due to restricted rotation around the amide bond (Figures S10, S13, S16, S19, and S22 in the Supplementary Materials). This splitting of the proton resonances indicates that in DMSO-d 6 , hydrazones exist as a mixture of Z/E isomers with respect to the hindered rotation around the amide bond. Usually, the Z isomer predominates [32,33]. In the 1 H NMR spectra, the NH protons of Z isomers resonate in a lower field with respect to the resonances attributed to E isomers [48]. Two carbonyl group carbon resonances (C 14 and C 12 , respectively) in the range of 162.  ppm confirm the presence of the 2-oxindole moiety along with the amide group in molecules 4-8 (Figures S11, S14, S17, S20, and S23 in the Supplementary Materials).
Hydrazone derivatives 9-19 were synthesized from hydrazide 3 and the corresponding aldehydes in methanol under reflux in the yield range of 29-98% (Scheme 2) [35,49]. The structures of the synthesized compounds have been confirmed by 1 H and 13 C NMR, IR, and MS data. In the 1 H NMR spectrum for 15 ( Figure S43 in the Supplementary Materials), multiplets at 7.88-8.25 integrated for two protons are attributed to the H 13 proton and the one in the CH group of the pyrazole ring. In the 1 H NMR spectrum for 16 ( Figure S46 in the Supplementary Materials), two protons in the pyrazole ring resonated as multiplets at 6.52-6.65 ppm and 7. 35-8.18 ppm (this signal overlaps with the resonances of aromatic protons and the one of the H 13 proton). The presence of the methyl group attached to the N atom in pyrazole moiety was confirmed by the singlet at 3.40 ppm. The resonance attributed to the C 11 carbon atom in the 13 C NMR spectra of hydrazones 9-19 was shifted downfield by~1 ppm compared to the respective resonance in the 13

Cytotoxicity
The synthesized compounds showed different activity against human cancer cell lines at 50 µM concentration. This concentration was chosen based on the solubility of the compounds and our previous experience. It has been shown that the higher concentrations of 50-100 µM allow distinguishing the most active compounds [33,50] and is often used for a primary screening of anticancer agents [51]. Our calculated EC 50 values correspond to the 50% lethal concentration (LC 50 ), which is calculated by the US National Cancer Institute (NCI) during screening of compounds [52]. However, NCI performs screening on cancer cell panels, and does not evaluate the compound effects on normal cells (e.g., fibroblasts); thus, the obtained results usually lack information about compound selectivity towards tumours. In addition to compound screening in the cell monolayer, NCI recently has introduced the hollow fibre assay, which includes a short-term in vitro culture (up to 48 h) of a panel of 12 human tumour cell lines in fibres, and the results obtained well represent compound activity in tumour xenografts [53]. Thus, we also decided to first select the most promising candidates by the MTT assay and then test their activity in tumour spheroids.
Surprisingly, the synthesized hydrazones showed relatively high activity against the IGR39 cell line used in the screening experiments ( Figure 1), except compounds 1, 2, 3, 5, 12, 13, 14, 16, and 19. This type of cancer (malignant melanoma) is usually considered as a relatively resistant one due to many different resistance mechanisms [54]. Triple-negative breast cancer cells MDA-MB-231 and pancreatic cancer cells Panc-1 were less sensitive to a majority of the tested compounds. In general, triple-negative breast cancer cells are characterized by their high resistance to many chemotherapeutic drugs due to high expression of P-glycoprotein, the stemness properties, and other mechanisms [55]. Pancreatic cancer is one of the most difficult cancer types to treat and, therefore, is characterized by high resistance to many available anticancer agents [56].

Cytotoxicity
The synthesized compounds showed different activity against human cancer cell lines at 50 µM concentration. This concentration was chosen based on the solubility of the compounds and our previous experience. It has been shown that the higher concentrations of 50-100 µM allow distinguishing the most active compounds [33,50] and is often used for a primary screening of anticancer agents [51]. Our calculated EC50 values correspond to the 50% lethal concentration (LC50), which is calculated by the US National Cancer Institute (NCI) during screening of compounds [52]. However, NCI performs screening on cancer cell panels, and does not evaluate the compound effects on normal cells (e. g., fibroblasts); thus, the obtained results usually lack information about compound selectivity towards tumours. In addition to compound screening in the cell monolayer, NCI recently has introduced the hollow fibre assay, which includes a short-term in vitro culture (up to 48 h) of a panel of 12 human tumour cell lines in fibres, and the results obtained well represent compound activity in tumour xenografts [53]. Thus, we also decided to first select the most promising candidates by the MTT assay and then test their activity in tumour spheroids.
Surprisingly, the synthesized hydrazones showed relatively high activity against the IGR39 cell line used in the screening experiments ( Figure 1), except compounds 1, 2, 3, 5, 12, 13, 14, 16, and 19. This type of cancer (malignant melanoma) is usually considered as a relatively resistant one due to many different resistance mechanisms [54]. Triple-negative breast cancer cells MDA-MB-231 and pancreatic cancer cells Panc-1 were less sensitive to a majority of the tested compounds. In general, triple-negative breast cancer cells are characterized by their high resistance to many chemotherapeutic drugs due to high expression of P-glycoprotein, the stemness properties, and other mechanisms [55]. Pancreatic cancer is one of the most difficult cancer types to treat and, therefore, is characterized by high resistance to many available anticancer agents [56]. In general, among the synthesized 1,2,4-triazole-3-thiol derivatives, hydrazones 17 and 18 bearing 2-hydroxybenzene or 2-hydroxy-5-nitrobenzene moiety, respectively, were identified as the most active ones against all tested cancer cell lines. Meanwhile, hydrazones bearing 5-fluoro-or 5-trifluoromethoxy-2-oxindole moiety 7 and 8, respectively, were the most active among the synthesized compounds against the IGR39 cell line. Interestingly, the least active derivatives against IGR39 (compounds 14 and 19) showed relatively higher activity towards the triple-negative breast cancer cell line. These com- In general, among the synthesized 1,2,4-triazole-3-thiol derivatives, hydrazones 17 and 18 bearing 2-hydroxybenzene or 2-hydroxy-5-nitrobenzene moiety, respectively, were identified as the most active ones against all tested cancer cell lines. Meanwhile, hydrazones bearing 5-fluoro-or 5-trifluoromethoxy-2-oxindole moiety 7 and 8, respectively, were the most active among the synthesized compounds against the IGR39 cell line. Interestingly, the least active derivatives against IGR39 (compounds 14 and 19) showed relatively higher activity towards the triple-negative breast cancer cell line. These compounds contain pyrrole and 4-(methylthio)benzene fragments, respectively, in their structure, which could be a key element to obtain a selectivity against the MDA-MB-231 cell line. Two more compounds, 4 and 10, showed a moderate selectivity against the melanoma IGR39 cell line; therefore, they were also included in further studies. In total, seven of the most active compounds (4, 7, 8, 10, 14, 17, and 18) were selected for further studies, and their effective concentrations that reduce cell viability by 50% (EC 50 values) were determined ( Figure 2). ture, which could be a key element to obtain a selectivity against the MDA-MB-231 cell line. Two more compounds, 4 and 10, showed a moderate selectivity against the melanoma IGR39 cell line; therefore, they were also included in further studies. In total, seven of the most active compounds (4, 7, 8, 10, 14, 17, and 18) were selected for further studies, and their effective concentrations that reduce cell viability by 50% (EC50 values) were determined ( Figure 2). Hydrazone 10 bearing p-(dimethylamino)benzene moiety has been shown to be the least active out of seven selected compounds, but it was most selective towards cancer cell lines (EC50 = 22.3 ± 2.5 µM against IGR39, EC50 = 9.7 ± 1.6 µM against MDA- , and EC50 = 26.2 ± 1.0 µM against the Panc-1 cell line) compared to fibroblasts (EC50 = 61.4 ± 2.8 µM). The cytotoxic effects of hydrazone bearing the 5-trifluoromethoxy-2-oxindole moiety 8 against Panc-1 were the lowest among the seven tested compounds (EC50 = 38.5 ± 1.6 µM). However, its activity against melanoma and breast cancer cells was comparable to that of the other most active compounds. Furthermore, compound 8 showed a lower cytotoxicity against fibroblasts, which makes it more promising.
Tiago et al. [57] have shown that the presence of stroma may enhance the drug resistance of melanoma in vitro, due to the interaction between tumour and stroma, which means that selectivity in cell monolayers does not represent the in vivo situation quite well. Cancer-associated fibroblasts are known to also change the resistance of cancer cells in monolayers to chemotherapy [58]. Furthermore, many studies have indicated not very large differences between cytotoxicity against normal and cancer cells, and there are still some ongoing debates on how to interpret the obtained results [59] and what selectivity index could be considered as "good" enough. Usually, the higher toxicity towards cancer cells at a value twice or lower than towards normal cells, is considered to be already showing selectivity [60]. In our scenario, compounds 8 and 10 were even more selective against several cancer cell lines.
All seven compounds showed higher cytotoxicity towards malignant melanoma IGR39 cell lines, especially pyrrole derivative 17, and they could be further developed as anti-melanoma drugs. Dacarbazine, used to treat melanoma, is much less active than these compounds and inhibits melanoma A375 cell proliferation at concentrations of 25-100 µM [61]. There are limited studies on compounds against the IGR39 cell line, which may show the importance of our study. In our previous experiments on new sunitinib derivatives, Hydrazone 10 bearing p-(dimethylamino)benzene moiety has been shown to be the least active out of seven selected compounds, but it was most selective towards cancer cell lines (EC 50 = 22.3 ± 2.5 µM against IGR39, EC 50 = 9.7 ± 1.6 µM against MDA- , and EC 50 = 26.2 ± 1.0 µM against the Panc-1 cell line) compared to fibroblasts (EC 50 = 61.4 ± 2.8 µM). The cytotoxic effects of hydrazone bearing the 5-trifluoromethoxy-2-oxindole moiety 8 against Panc-1 were the lowest among the seven tested compounds (EC 50 = 38.5 ± 1.6 µM). However, its activity against melanoma and breast cancer cells was comparable to that of the other most active compounds. Furthermore, compound 8 showed a lower cytotoxicity against fibroblasts, which makes it more promising.
Tiago et al. [57] have shown that the presence of stroma may enhance the drug resistance of melanoma in vitro, due to the interaction between tumour and stroma, which means that selectivity in cell monolayers does not represent the in vivo situation quite well. Cancer-associated fibroblasts are known to also change the resistance of cancer cells in monolayers to chemotherapy [58]. Furthermore, many studies have indicated not very large differences between cytotoxicity against normal and cancer cells, and there are still some ongoing debates on how to interpret the obtained results [59] and what selectivity index could be considered as "good" enough. Usually, the higher toxicity towards cancer cells at a value twice or lower than towards normal cells, is considered to be already showing selectivity [60]. In our scenario, compounds 8 and 10 were even more selective against several cancer cell lines.
All seven compounds showed higher cytotoxicity towards malignant melanoma IGR39 cell lines, especially pyrrole derivative 17, and they could be further developed as anti-melanoma drugs. Dacarbazine, used to treat melanoma, is much less active than these compounds and inhibits melanoma A375 cell proliferation at concentrations of  µM [61]. There are limited studies on compounds against the IGR39 cell line, which may show the importance of our study. In our previous experiments on new sunitinib derivatives, we identified several highly active derivatives 4001, 4007, and 4008 that possessed an antiproliferative effect at 140-500 nM concentrations [34]. However, the BRAF inhibitor dabrafenib, which is used as a chemotherapeutic drug for melanoma treatment, reduces melanoma cell viability at a nanomolar concentration [62].

Effect on Cell Migration
The 'Wound healing' assay was used to evaluate compound activity on cell migration. The most active compounds 4, 7, 8, 10, 14, 17, and 18 were tested for their effect on human melanoma IGR39, human triple-negative breast cancer MDA- , and pancreatic carcinoma Panc-1 cell migration at 10 µM concentration. The effect of this concentration on  S58-60) using the MTT assay (the procedure is described in Supplementary Materials).
Several compounds have been shown to possess an effect on tested cancer cell migration (Figure 3). In general, the effect on IGR39 cell migration was more strongly expressed after a shorter duration (24 h For decades, 3D cell cultures have been widely used as a model to test the anticancer activity of novel substances. This model is considered to represent the real tumour microenvironment much better compared to conventional cell monolayers (2D models). Tu-  The compounds did not reduce viability of cells by more than up to 80% after 48 h of incubation, except the compounds 14 and 17 (IGR39 cell viability after 48 h of incubation with these compounds was (67.9 ± 1.8)% and (76.0 ± 7.2)% ( Figures S58-60 in Supplementary Materials). The viability of Panc-1 cells after 72 h was reduced up to (74.3 ± 3.4)% by compound 4, and up to (79.7 ± 3.8)% by compound 17. In general, there was no correlation between cell viability and cell migration at 10 µM concentration, so it could not be concluded that the migration inhibition effect was the result of reduced cell proliferation. Of course, compound effects on cell proliferation cannot be excluded, but we assume that there could be many still unknown effects on cell metabolic pathways more specifically related to the migration inhibition, and the reduction in cell viability could be a final consequence of these effects.
Malignant melanoma and triple-negative breast cancer cell lines are characterized by high aggressiveness and invasiveness. As a result, in order to combat the resistance, an approach of inhibiting cancer cell migration has been widely studied [63,64]. There is a need for more effective novel compounds with a potential to reduce cell migration. One of the possible solutions could be a combination strategy where the new compound or drug is added in combination with already clinically used highly effective anticancer agents [65]. Mishra et al. [66] determined that the pan-PI3K inhibitor combined with the widely used anticancer drug doxorubicin suppresses cancer cell growth, proliferation, and migration. Inhibition of protein kinase B (AKT) and SRC in the presence of gemcitabine significantly reduced the metastatic potential of pancreatic cancer cells by suppressing the phosphorylation of mTOR and ERK in cells [67]. The synthesized compounds contain fragments characteristic of kinase inhibitors, and they are expected to possess a similar mechanism of action. However, more detailed studies are needed to prove their mechanism of action.

Activity in 3D Cell Cultures (Spheroids)
For decades, 3D cell cultures have been widely used as a model to test the anticancer activity of novel substances. This model is considered to represent the real tumour microenvironment much better compared to conventional cell monolayers (2D models). Tumour spheroids are one of the most simplified 3D cell models, and are characterised by hypoxia formation in their core as well as the gradient of tested substances [68]. The effects of 10 µM solutions of the seven selected hydrazones 4, 7, 8, 10, 14, 17, and 18 on melanoma IGR39, triple-negative breast cancer MDA-MB-231 and pancreatic carcinoma Panc-1 cell spheroid growth were evaluated ( Figure 4). As shown in Figure 4A, the most active compounds against all three types of spheroids were 14, 17, and 18. Melanoma spheroid growth was mostly inhibited by compounds 4, 14, and 18; IGR39 spheroids began to disintegrate after 8 days of incubation with compound 17. Meanwhile, Panc-1 spheroids were more resistant to compound treatment, but their growth delay was also observed after incubation with almost all compounds tested ( Figure 4B).
It has already been proven by different research groups that measuring only the size of spheroids is not enough to judge about the effectiveness of the compound. The size of the spheroid does not necessarily correlate with the cell viability in 3D [69]; therefore, in our study we also tested cell viability in all spheroids at the end of the experiment ( Figure 4C). Interestingly, cell viability was from 1.5-fold to 2.8-fold lower compared to the control in all types of spheroids, despite the size of spheroids being not so different from that of the control group. Such a phenomenon could be explained by the different spheroid morphology due to distinct hypoxic properties and variant gradient of the compounds tested. Several studies have already shown that kinase inhibitors might reduce spheroid size by strengthening cell-cell interaction [70]. On the contrary, Aihara et al. [71] discovered that LATS kinase inhibitors block the Hippo signalling pathway and could promote cell proliferation under 3D culture conditions. The study only shows that kinase inhibitors may possess different effects depending on their specific metabolic pathways, and more extensive studies are needed. growth was mostly inhibited by compounds 4, 14, and 18; IGR39 spheroids began to disintegrate after 8 days of incubation with compound 17. Meanwhile, Panc-1 spheroids were more resistant to compound treatment, but their growth delay was also observed after incubation with almost all compounds tested ( Figure 4B). It has already been proven by different research groups that measuring only the size of spheroids is not enough to judge about the effectiveness of the compound. The size of the spheroid does not necessarily correlate with the cell viability in 3D [69]; therefore, in our study we also tested cell viability in all spheroids at the end of the experiment ( Figure  4C). Interestingly, cell viability was from 1.5-fold to 2.8-fold lower compared to the control in all types of spheroids, despite the size of spheroids being not so different from that of the control group. Such a phenomenon could be explained by the different spheroid morphology due to distinct hypoxic properties and variant gradient of the compounds tested. Several studies have already shown that kinase inhibitors might reduce spheroid size by strengthening cell-cell interaction [70]. On the contrary, Aihara et al. [71] discovered that LATS kinase inhibitors block the Hippo signalling pathway and could promote cell proliferation under 3D culture conditions. The study only shows that kinase inhibitors may possess different effects depending on their specific metabolic pathways, and more extensive studies are needed. In summary, hydrazones 4, 14, and 18 were identified as the most promising anticancer agents out of a series of 1,2,4-triazole-3-thiol derivatives. They were shown to possess a moderate cytotoxicity against the tested cancer cell lines (EC 50 values were in the range of 2-17 µM). Although these derivatives are not very selective against all cancer cells compared to fibroblasts, yet they were several times more cytotoxic against the triple-negative breast cancer cell line, which is characterized as very invasive and lacking specific targets for chemotherapeutics. Furthermore, selected compounds showed higher or comparable activity to dacarbazine and erlotinib, drugs already approved to treat malignant melanoma and pancreatic cancer. Furthermore, N -(4-(dimethylamino)benzylidene)-2-((4-phenyl-5-(2-(phenylamino)ethyl)-4H-1,2,4-triazol-3-yl)thio)acetohydrazide (10) was identified as relatively selective towards cancer cells and showed promising results in migration assays on MDA-MB-231 and Panc-1 cells. It is worth exploring its effects on cell migration and invasion in more sophisticated methods such as 3D invadopodia formation or cell migrations in spheroid models and identifying specific pathways that contribute to these effects. The development of antimetastatic agents is facing major challenges [72], and metastasis often leads to death from more aggressive types of cancer, which only proves the need to search for novel effective migrastatic agents.

Chemical Reagents and Instruments
Reagents were purchased from Sigma-Aldrich (St. Louis, MO, USA) and TCI Europe N.V. (Zwijndrecht, Belgium). The reaction course and purity of the synthesized compounds were monitored by TLC using aluminium plates precoated with silica gel 60 F254 (Mer-ckKGaA, Darmstadt, Germany). The melting points were determined on a MEL-TEMP (Electrothermal, A Bibby Scientific Company, Burlington, NJ, USA) melting point apparatus and are uncorrected. FT-IR spectra (ν, cm −1 ) were recorded on a Perkin-Elmer Spectrum BX FT-IR spectrometer using KBr pellets. The 1 H and 13 C-NMR spectra were recorded in DMSO-d 6 on a Bruker Avance III (400 MHz, 101 MHz) spectrometer operating in the Fourier transform mode. Chemical shifts (δ) are reported in parts per million (ppm) calibrated from TMS (0 ppm) as an internal standard for 1 H NMR, and DMSO-d 6 (39.43 ppm) for 13 C NMR. Mass spectra were obtained on a Bruker maXis UHR-TOF mass spectrometer (Bruker Daltonics, Bremen, Germany) with ESI ionization.
The compound effect on cell viability was calculated using the formula: where A-mean of absorbance of the tested compound, A 0 -mean of absorbance of blank (no cells, positive control), and A NC -mean of absorbance of negative control (only cells, no treatment).
The EC 50 values of the most active hydrazones 4, 7, 8, 10, 14, 17, and 18 were established using the same MTT procedure. The compound serial dilutions from 50 µM to 1.56 µM were made in a medium and added to the cells in triplicates. The EC 50 value representing the concentration of a compound causing 50% reduction of cancer cell metabolic activity was calculated using the Hill equation.

Wound Healing' Assay
The 'Wound healing' assay was used to evaluate the inhibitory effect of the most active hydrazones 4, 7, 8, 10, 14, 17, and 18 on cell migration, as described elsewhere [74]. Briefly, melanoma IGR39, human triple-negative breast cancer MDA- , and pancreatic carcinoma Panc-1 cells were seeded in 24-well plates at a density of 6 × 10 4 cells/well and incubated for 48 h in previously described cell culture medium at 37 • C in a humidified atmosphere containing 5% CO 2 . Then the scratch was made using a 100 µL pipette tip. Cells were washed once with PBS, and the fresh medium containing 10 µM of compounds 4, 7, 8, 10, 14, 17, and 18 was added. Medium containing 0.1% DMSO was used as a negative control. Cells were incubated at 37 • C in a humidified atmosphere containing 5% CO 2 .
The 'Wounds' were captured at the intervals of 0 h, 24 h, 48 h and 72 h (only for the Panc-1 cell line) from scratch under the phase contrast microscope at a 4× magnification. The wound area was analysed using the ImageJ program (National Institute of Health, Bethesda, MD, USA).

Compound Activity in Cell 3D Cultures (Spheroids)
Cancer cell spheroids were formed using the magnetic 3D Bioprinting method, as described elsewhere [75]. Briefly, melanoma IGR39, triple-negative breast cancer MDA-MB-231 cells, pancreatic cancer Panc-1 cells, and human fibroblasts at 70% confluency in a 6-well plate were incubated with Nanoshuttle (n3D Biosciences, Inc., Houston, TX, USA) for 8 h at 37 • C in a humidified atmosphere containing 5% CO 2 . Then cells were trypsinized, centrifuged, and seeded into an ultra-low attachment 96-well plate in a volume of 100 µL (1.5 × 10 3 cancer cells and 1.5 × 10 3 human fibroblasts/well). The plate was placed on a magnetic drive and incubated for 2 days at 37 • C in a humidified atmosphere containing 5% CO 2 . Then the fresh medium containing 10 µM of the tested compound was added to the wells. Spheroids were captured every two days using the Olympus IX73 inverted microscope (OLYMPUS CORPORATION, Tokyo, Japan). Quantitative analysis of compound anticancer activity in spheroids was performed using ImageJ, version 1.53o (National Institutes of Health, USA) and Microsoft Office Excel 2016 software (Microsoft Corporation, Redmond, WA, USA).
On the last day of incubation, 10 µL of WST-1 reagent (Sigma-Aldrich Co, St. Louis, MO, USA) was added to each well with spheroids. After 10 h of incubation, 50 µL of liquid from each well were transferred to the new 96-well plate and the absorbance was measured at 460 and 530 nm using a multi-detection microplate reader. Spheroid cell viability was calculated using a formula provided in Section 3.2.2.

Statistical Analysis
All biological experiments were repeated at least three times, calculating the mean and standard deviation. The data was processed using the Microsoft Office Excel 2016 software (Microsoft Corporation, Redmond, WA, USA) and the IBM SPSS Statistics version 27.0 package. Statistical analysis was performed by using the Student's t-test. The level of significance was set as p < 0.05. In order to determine significant differences between values, analysis of variance (ANOVA) followed by a Tukey post-hoc test was performed.