Enhanced Anticancer Potential of Pd(II)-Thiosemicarbazone Complexes: Selectivity, Mechanisms, and 3D Models

Abstract

Background/Objectives: Cancer remains a major global health challenge, driving the search for novel chemotherapeutic agents. This study aimed to evaluate the structural and biological properties of a series of Pd(II) complexes containing triphenylphosphine and thiosemicarbazone ligands, in order to assess their potential as anticancer agents. Methods: Six Pd(II) complexes with the general formula [PdCl(PPh₃)(TSC)] were synthesized and fully characterized by NMR (¹H, ¹³C, ³¹P), FTIR, mass spectrometry, and X-ray diffraction. Their cytotoxic effects were investigated through in vitro assays using 2D and 3D cancer cell models, including clonogenic, wound healing, cell cycle, and apoptosis assays via flow cytometry. Results: Complexes from the B family demonstrated significantly higher cytotoxicity than those from the C family, particularly against ovarian (IC₅₀ < 1 µM) and breast (IC₅₀~2 µM) cancer cell lines. These compounds exhibited superior potency and selectivity compared to cisplatin, with high selectivity indices toward non-tumor cells. Mechanistic studies revealed both cytotoxic and cytostatic effects depending on structural variations, with apoptosis identified as the primary mechanism of cell death. PdB1, in particular, induced a marked increase in late apoptotic populations and maintained its cytotoxic activity in 3D spheroid models by promoting disintegration, loss of cell adhesion, and nuclear fragmentation. Conclusions: The findings underscore the therapeutic promise of Pd(II) complexes, especially PdB1, as potent and selective antineoplastic agents capable of acting effectively in complex tumor environments and potentially overcoming chemoresistance.

1. Introduction

Cancer is characterized by the uncontrolled proliferation of cells caused by genetic and epigenetic alterations that disrupt normal regulatory mechanisms, including those that control apoptosis. These transformed cells can form primary tumors and, in malignant forms, invade surrounding tissues and metastasize to distant organs, driving disease progression and complicating treatment outcomes [1,2,3]. This complexity underscores the ongoing need for innovative therapeutic strategies to improve clinical outcomes. Chemotherapy remains a central component of treatment protocols, using drugs that interfere with cell division or induce programmed cell death [4,5,6].

Cisplatin, a platinum-based chemotherapeutic agent, is widely used in the treatment of a variety of solid tumors. Its cytotoxicity is due to the aquation of its chloride ligands, which enables covalent binding to DNA and the formation of intra- and interstrand crosslinks [7,8,9,10]. These adducts inhibit DNA replication and transcription, ultimately inducing apoptosis. Despite its efficacy, the long-term success of cisplatin is limited by several factors. In particular, the emergence of drug resistance, whether intrinsic or acquired, significantly reduces its therapeutic potential. Resistance mechanisms include increased drug efflux, enhanced DNA repair, inactivation by thiol containing biomolecules, and impaired apoptosis signaling [11]. In addition, severe side effects such as nephrotoxicity, neurotoxicity, and ototoxicity necessitate the development of alternative metal-based agents with improved safety and efficacy profiles.

In response to these limitations, researchers have explored a variety of metal ions as the basis for novel antitumor agents [12,13,14]. Palladium(II) stands out because of its chemical similarity to platinum(II), particularly in terms of coordination geometry and ligand exchange properties [15,16]. In contrast to cisplatin, Pd(II) complexes often exhibit faster ligand exchange kinetics, which can be modulated by appropriate ligand design. A notable example of a palladium-based drug is Tookad (padeliporfin), a Pd(II)-porphyrin complex approved for the treatment of prostate cancer via photodynamic therapy [17,18].

Thiosemicarbazones (TSCs) are versatile N,S-donor ligands derived from the condensation of thiosemicarbazides with aldehydes or ketones. Their ability to chelate metal ions combined with their intrinsic biological activity has led to extensive investigation of their antitumor, antiviral and antibacterial properties [19,20]. Among them, triapine is a clinically relevant TSC that is currently undergoing phase II and III trials for several types of cancer [21]. Structural modification of TSCs, particularly by substitution on the aromatic or aliphatic moieties, offers a powerful strategy to fine-tune pharmacological properties and metal-binding characteristics.

In this context, the present study investigates the influence of structural variations in thiosemicarbazone ligands on the antitumor activity of Pd(II) complexes. Six [PdCl(PPh₃)(TSC)] compounds were synthesized and fully characterized by NMR spectroscopy (¹H, ¹³C, ³¹P), FTIR, high resolution mass spectrometry, and single crystal X-ray diffraction. The TSC ligands used in this work are derived from the ketones benzylideneacetone (B) and chalcone (C). The substituents on nitrogen 3 of the thiocarbazides are –H (1), –CH₃ (2), and –CH₂CH₃ (3). The combination of these reagents results in the formation of ligands B1, B2, B3, C1, C2, and C3, and the complexation and addition of triphenylphosphine led to the formation of the complexes PdB1, PdB2, PdB3, PdC1, PdC2, and PdC3; see Scheme 1. It is important to mention that the PdB2 complex was reported in a previous work [22], and will have its biological studies extended in this work. Their cytotoxic potential was evaluated in both tumor and non-tumor cell lines, with further mechanistic studies addressing their ability to inhibit tumor cell proliferation, migration, and survival. Furthermore, 3D tumor models were investigated to gain deeper insight into their biological behavior and potential as next-generation antineoplastic agents [23].

2. Materials and Methods

2.1. Materials

The PdB2 complex was synthesized previously by the author of this work, and its characterization is available in the literature [22]. The ligands used were provided by collaborators and the syntheses and characterizations are already reported [24].

The solvents used were from Labsynth, AcrosOrganics and Merck P.A. (Geel, Belgium). The reagents chloride palladium (II) (99%), triphenylphosphine (PPh₃), Cesium iodide (99.99%), human serum albumin (HSA), calf thymus—DNA, DPPH, and MTT were purchased from Sigma-Aldrich (St. Louis, MO, USA).

2.2. Physical Measurements

¹H, ¹³C{¹H} and ³¹P{¹H} NMR spectra were obtained using the BRUKER ARX 9.4T spectrometer (BRUKER, Billerica, MA, USA). Dept135, COSY, HSQC, and HMBC experiments were performed for the complex PdC3. All the data obtained was processed using the MestReNova 12.0 software.

The spectra in the infrared region were obtained on a SHIMADZU IRTracer-100 spectrophotometer (SHIMADZU, Kyoto, Japan), using CsI pellets as a support, and the data obtained was processed using Shimadzu IR Solution, 1.60 and Origin 9.0 software.

UHPLC-QTof-MS/MS was performed by direct infusion, using water (solvent A, 0.1% formic acid) and CH₃CN (solvent C with 0.1% formic acid) as the mobile phase, in isocratic mode, in a ratio of 05:95 v/v, respectively. The mobile phase flow rate was 0.300 mL·min⁻¹ and the injection volume was 3 µL. Quantitative analysis of the samples was performed using an Agilent 6545 qTOF MS system (Agilent Technologies, Santa Clara, CA, USA), equipped with a Jet electrospray interface (ESI) in positive mode. The error (in ppm) was calculated from the theoretical mass and the respective MS spectrum acquired. The calculation was carried out for the most intense isotope of the molecular ion [M]⁺. A range between 100 and 1000 Da was monitored, with a scan rate of 3 spectra s⁻¹ and processed using Mass Hunter Workstation Software version B.08.00.

X-ray diffraction was obtained on a Rigaku XtaLAB mini II diffractometer, MoKα radiation (λ = 0.71073 Å). The SHELXT (2019) software was used to solve the structure using direct methods and successive Fourier difference maps, which allowed the non-hydrogen atoms to be located. Structural refinement was performed using the same software, using the least squares method. The unit cell parameters and absorption corrections were calculated using the CrysAlisPro171 software. Graphical representations of the molecules and crystallographic data tables were generated using the OLEX2 1.5 and MERCURY 4.2.0 software [25,26].

2.3. Synthesis of the Complexes

The palladium precursor complex, [PdCl₂(CH₃CN)₂], was obtained by slowly adding 500 mg of chloride palladium (II) to 50 mL of previously heated acetonitrile. After 4 h of reaction, the solution was concentrated by evaporating the solvent and the solid obtained was filtered. Yield = 80%.

To obtain the palladium complexes, 0.19 mmol of the precursor complex was solubilized in acetonitrile. Then, thiosemicarbazone ligand was added in a 1:1 ratio. After 24 h of reaction, at ambient temperature, 0.19 mmol of triphenylphosphine (PPh₃) was added and the reaction medium was kept under agitation for a further 24 h. After this period, the volume of the solution was reduced and the solid obtained was filtered under vacuum.

[PdCl(L-B1)(PPh₃)]Cl (PdB1) − Yield = 90%. C₂₉H₂₈Cl₂N₃PPdS. FTIR (cm⁻¹): 3451-3237 νNH; 3050 νCH_sp²; 1617 νC=CArH; 1542 δN-H; 1097 τ(P-CPh); 748 νC=S; 688 δC-HAr. ¹H NMR (CDCl₃, 400 MHz) δ(ppm): 13.51 (s, 1H, N(2)-H), 10.03 (s, 1H, N(3)-H), 8.55 (d, J = 16.1 Hz, 1H, H-8), 7.76 (m, 6H, o-HAr PPh₃), 7.70 (m, 2H, H-1 + H-5), 7.57 (m, 3H, p-HAr PPh₃), 7.48 (m, 6H, m-HAr PPh₃), 7.37 (m, 3H, H-2 + H-3 + H-4), 7.24 (m, 1H, H-7), 6.10 (s, 1H, N(3)-H), 2.67 (s, 3H, R1(CH₃)). ³¹P{¹H} NMR (CDCl₃, 162 MHz) δ(ppm) = 30.85. HRMS (QTof, positive mode) m/z: calculed: 622.0459; found: 622.0445 [M]⁺.

[PdCl(L-B3)(PPh₃)]Cl (PdB3) − Yield = 85%. C₃₁H₃₂Cl₂N₃PPdS. FTIR (cm⁻¹): 3408-3237 νNH; 3068-2916 νCH_sp²; 1555 νC=CArH; 1514 δN-H; 1095 τ(P-CPh); 748 νC=S; 689 δC-HAr. ¹H NMR (CDCl₃, 400 MHz) δ(ppm): 12.92 (s, 1H, N(2)-H), 10.08 (s, 1H, N(3)-H), 8.50 (d, J = 16,1 Hz, 1H, H-8), 7.78 (m, 6H, o-HAr PPh3), 7.67 (m, 2H, H-1 + H-5), 7.57 (m, 3H, p-HAr PPh3), 7.49 (m, 6H, m-HAr PPh3), 7.36 (m, 3H, H-2 + H-3 + H-4), 7.18 (d, J = 16.1 Hz, 1H, H-7), 3.24 (q, dd J = 7.2; J = 5.0 Hz, H2, R2(CH₂), 2.64 (s, 3H, R1(CH₃), 1.22 (t, J = 7.1 Hz, 3H, R2(CH₃)). ³¹P{¹H} NMR (CDCl₃, 162 MHz) δ(ppm) = 30.89. HRMS (QTof, positive mode) m/z: calculed: 650.0772; found: 650.0777 [M]⁺.

[PdCl(L-C1)(PPh₃)]Cl (PdC1) − Yield = 65%. C₃₄H₃₀Cl₂N₃PPdS. FTIR (cm⁻¹): 3379-3255 νNH; 3094-3002 νCH_sp²; 1598 νC=CArH; 1533 δN-H; 1095 τ(P-CPh); 750 νC=S; 691 δC-HAr. ¹H NMR (CDCl₃, 400 MHz) δ(ppm): 8.07 (d, J = 16.1 Hz, 1H, H-8), 7.84 (m, 3H, R1(Ph)), 7.59 (m, 9H, H-1 + H-3 + H-5 + o-HAr PPh₃), 7.52 (m, 2H, R1(Ph)), 7.46 (m, 3H, p-HAr PPh₃), 7.38 (m, 6H, m-HAr PPh₃), 7.36 (m, 2H, H-2 +H-4), 6.75 (d, J = 16.1 Hz, 1H, H-7). ³¹P{¹H} NMR (CDCl₃, 162 MHz) δ(ppm) = 29.83. HRMS (QTof, positive mode) m/z: calculed: 684.0616; found: 684.0612 [M]⁺.

[PdCl(L-C2)(PPh₃)]Cl (PdC2) − Yield = 68%. C₃₅H₃₂Cl₂N₃PPdS. FTIR (cm⁻¹): 3460-3250 νNH; 3068-2907 νCH_sp²; 1581 νC=CArH; 1489 δN-H; 1095 τ(P-CPh); 745 νC=S; 692 δC-HAr. ¹H NMR de (CDCl₃, 400 MHz) δ(ppm): 13.67 (s, 1H, N(2)-H), 10.18 (s, 1H, N(3)-H), 8.15 (d, J = 16.1 Hz, 1H, H-8), 7.84 (m, 2H, R1(Ph)), 7.77 (m, 2H, R1(Ph)), 7.64 (m, 3H, H-1 + H-5 + R1(Ph)), 7.57 (m, 6H, o-HAr PPh3), 7.51 (m, 3H, p-HAr PPh₃), 7.42 (m, 9H, H-2 + H-3 + H-4 + m-HAr PPh₃), 6.91 (d, J = 16.1 Hz, 1H, H-7), 2.91 (s, 3H, R2(CH₃)). ³¹P{¹H} NMR (CDCl₃, 162 MHz) δ(ppm) = 31.70. HRMS (QTof, positive mode) m/z: calculed: 698.0772; found: 698.0773 [M]⁺.

[PdCl(L-C3)(PPh₃)]Cl (PdC3) − Yield = 69%. C₃₆H₃₄Cl₂N₃PPdS. FTIR (cm⁻¹): 3434-3236 νNH; 3054-2929 νCH_sp²; 1587 νC=CArH; 1520 δN-H; 1097 τ(P-CPh); 745 νC=S; 685 δC-HAr. ¹H NMR (CDCl₃, 400 MHz) δ(ppm): 13.67 (s, 1H, N(2)-H), 10.27 (s, 1H, N(3)-H), 8.18 (d, J = 16.1 Hz, 1H, H-8), 7.86 (m, 2H, R1(Ph₃)), 7.79 (m, 2H, R1(Ph₃)), 7.62 (m, 3H, H-1 + H-5 + R1(Ph₃)), 7.58 (m, 6H, o-HAr PPh₃), 7.51 (m, 3H, p-HAr PPh₃), 7.42 (m, 9H, H-2 + H-3 + H-4 + m-HAr PPh₃), 6.90 (d, J=16.1 Hz, 1H, H-7) 3.29 (dd J = 7.2; J = 5.0 Hz, 2H, R2(-CH₂)), 1.23 (t, J = 7.2 Hz, 3H, R2(-CH₃)). ³¹P{¹H} NMR (CDCl₃, 162 MHz) δ(ppm) = 31.71. HRMS (QTof, positive mode) m/z: calculed: 712.0929; found: 712.0931 [M]⁺.

2.4. Cell Culture Tests (In Vitro)

The cell lines used in the tests were A2780 (ovarian cancer), A2780cis (ovarian cisplatin resistant cancer), A549 (lung cancer), MRC5 (non-tumoral lung cancer), MDA-MB-231 (triple-negative breast cancer), MCF-7 (breast cancer), SK-BR-3 (breast cancer), MCF-10A (non-tumoral breast cancer), and A375 (skin cancer—melanoma).

The MCF-10A cell line was grown in DMEM/F-12 culture medium, supplemented with 5% horse serum, human epidermal growth factor (EGF) (20 ng/mL), hydrocortisone (0.05 mg/mL), insulin (0.01 mg/mL), penicillin (1%), streptomycin (1%), and L-glutamine (2 mM). The A2780, A2780cis, and MCF-7 cell lines were cultured in RPMI-1640 medium, and the other lines in DMEM medium; both culture media were supplemented with 10% (v/v) FBS (fetal bovine serum), antibiotics, and antimycotics. Cultivation was carried out in culture flasks stored in a humid incubator at 37 °C, with an atmosphere of 5% CO₂.

2.5. Cell Viability (IC₅₀)

The cell viability test to obtain the minimum inhibitory concentration of 50% (IC₅₀) was made using a 96-well microplate, in which a 150 μL solution containing 1.5 × 10⁴ cells suspended in culture medium was added to each well. The plates were kept for 24 h in a humidified incubator (37 °C, 5% CO₂), after which 0.75 μL of a DMSO solution (0.5% v/v) containing the compounds of interest was added to each well. The plates were then incubated for a further 48 h. The ratio of viability of cells was obtained using the MTT method. Through its reduction via the mitochondrial process, purple formazan is formed, which is proportional to cell viability. After incubating the microplates, 50 μL of a solution of MTT (1 mg∙mL⁻¹) in PBS was added to each well and the cells were incubated again in the oven for 4 h. Finally, the solution was removed from each well and 100 μL of isopropyl alcohol with 10% DMSO was added to solubilize the Formazan crystals. Absorbance measurements (540 nm) were then taken in each well using a BioTek EPOCH spectrophotometer (BioTek, Winooski, WT, USA). The data obtained was analyzed using Excel 360 and GraphPadPrism 8.0.2.

2.6. Clonogenic Assay

For this test, 600 cells of the A2780cis strain were added to each well of a six-well microplate. After incubating for 24 h in a 5% CO₂ atmosphere, different concentrations of the compounds were added to the wells and incubated again for five days. After this period, the culture medium was changed and the plates were incubated for a further five days. Finally, the culture medium was removed and washed with PBS. Subsequently, a solution of methanol and acetic acid (3:1 v/v) was added for five minutes, and then the cells were stained with methyl violet solution. The plates were washed with distilled water. After drying, they were photographed using the iBright™ FL1500 (ThermoFisher, Waltham, MA, USA) and analyzed using ImageJ 1.53 software.

2.7. Cell Morphology Assay

1.2 × 10⁵ A2780cis cells were added to each well of a twelve-well microplate. After incubating for 24 h in a 5% CO₂ atmosphere, different concentrations of the compounds were added to the wells and the plates were incubated again for 48 h. Using a Nikon Eclipse TS 100 inverted microscope (Nikon, Minato, Tokyo) coupled to a Moticam digital camera (1.3 Megapixels; Moticam, Xiamen, China), the plate wells were photographed at 0, 24, and 48 h, and the morphological changes observed were analyzed.

For morphological analysis by fluorescence with Green Plasma and DAPI, 1 × 10⁴ cells/well of A2780cis were added to 96-well plates and incubated for 24 h. After this, the cells were treated with different concentrations of the compound and incubated for 48 h. Then the medium was removed and the cells were fixed with methanol for 5 min, and 50 μL of CellMasK Green Plasma membrane was added. After 5 min, 50 µL of DAPI was added and incubated at room temperature for a further 10 min. Images were taken using a fluorescence microscope (CELENA^® S Logos Biosystems, Anyang-si, Republic of Korea) with the appropriate filters.

2.8. Cell Viability (Kit: LIVE/DEAD^®)

1.0 × 10⁴ A2780cis cells were added to each well in a 96-well plate. After 24 h, the cells were treated with different concentrations of the compound and incubated again for 48 h. Then 100 µL of the LIVE/DEAD^® solution prepared as described by the manufacturer was added. The images were acquired using a fluorescence microscope (CELENA^® S logos biosystems) with a 4× objective, carried out in triplicate. Live and dead cells were counted using ImageJ 1.53 software and the IC₅₀ value was calculated using GraphPad Prism 8.0.2.

2.9. Cell Migration Assay—“Wound Healing”

4.0 × 10⁵ A2780cis cells were added to each well of a 12-well plate. The plate was incubated for 24 h and the formation of a monolayer of cells was observed. Using a micropipette tip, a scratch (wound) was made in the center of the well and culture medium and different compound solutions were added.

Using the digital camera attached to the microscope, the wells were photographed at 0, 24, and 30 h. The images obtained were analyzed using ImageJ software. It was then possible to calculate the rate of wound reduction in relation to time and the values obtained were analyzed using GraphPadPrism 8.0.2 software.

2.10. Cell Cycle Analysis

A2780cis cells (1 × 10⁵ cells/well) were seeded in 12-well plates. After incubation in CO₂ atmosphere at 37 °C for 24 h, the cells were treated with different concentrations of complex PdB1 for 24 h. Then, the cells were collected, washed with ice-cold PBS, and fixed in 70% ethanol at −20 °C for 24 h. After fixation, the cells were centrifuged at 2000 rpm and 4 °C for 5 min, resuspended in 150 μL of PBS buffer containing RNase A (0.2 mg/mL) and hypotonic fluorochrome solution (PI 5 μg/ mL), and incubated for 30 min before being analyzed on an Accuri C6 (BD Biosciences, Franklin Lakes, NJ, USA) flow cytometer that recorded 10,000 events. The cell cycle phase distribution was analyzed in triplicate by using the FlowJo 10.8 Software. Untreated cells were the negative control.

2.11. Apoptosis Assay

Apoptosis in A2780cis cells induced by the complexes was analyzed by flow cytometry; a PE-Annexin-V Apoptosis Detection kit (BD Biosciences) was employed. The cells (1 × 10⁵ cells/well) were seeded in a 24-well plate. After incubation for 24 h, the cells were exposed to increasing concentrations of complex PdB1 for 24 h. After treatment, the cells were collected, centrifuged at 2000 rpm and 4 °C for 5 min, washed with ice-cold PBS, and resuspended in 150 µL of binding buffer. PE-Annexin V (2.5 µL) and 7ADD (2.5 µL) were added, which was followed by incubation in the dark at room temperature for 20 min. Then, the cells were centrifuged at 2000 rpm and 4 °C for 5 min, resuspended in 400 µL of binding buffer, and analyzed on an Accuri C6 (BD Biosciences) flow cytometer; 10,000 events were recorded. Fluorescence was quantified in triplicate with the Cell Quest software (BD Biosciences). Untreated cells were the negative control.

2.12. Three-Dimensional Assay

The 3D culture assay was performed with the A2780cis cell line using the Greiner Bio-One 96-well magnetic bioprinting kit. In total, 150 µL of magnetic nanoparticles were added to the cells grown in a 25 cm² bottle and incubated for 24 h. Then, 3750 cells/well were added to repellent 96-well plates, taken to the magnetic drive, and incubated for 4 days for the formation of steroids. Once the spheroids had been obtained, they were treated with the compound PdB1 at different concentrations. Finally, images of the spheroids were recorded at 0 and 48 h, and images with fluorescence markers (DAPI and PI) were also obtained for the 48 h period using a logos biosystems CELENA^® S fluorescence microscope.

3. Results and Discussion

3.1. Synthesis and Characterization

The Pd(II) complexes were synthesized using the substitution reaction of the ligands present in the precursor [Pd(Cl)₂(CH₃CN)₂], as shown in Scheme 1. Substitution of the acetonitriles coordinated to the Pd(II) ion occurs by an activated associative mechanism, in which the TSC ligand approaches the metal center via the axial position (d_z²), and the penta-coordinated intermediate was formed, followed by the exit of CH₃CN. After the first substitution, the ligand attacks the metal center again and substitutes the second acetonitrile, forming a five-membered chelated ring [22].

Once the acetonitrile ligands are replaced and the chelate ring is formed, the palladium (II) complexes form a bimetallic intermediate in the solution, which is constituted by the chelate rings formed and two clorides in bridge [27]. The intermediate is broken up with the addition of triphenylphosphine after 24 h of reaction.

According to the ¹H NMR spectra of the complexes, the signals from the CH₃ and CH₂CH₃ groups are found in the more protected region of the spectrum. The signals found in the region higher than 10 ppm refer to H-N(2) and H-N(3). The H-N(2) signal was the most acidic of the amines, and there was a shift from 8.37 ppm of the free ligand (L-C3) to 13.66 ppm of the complex (PdC3). Thus, coordination to the metal center represented a reduction in the electronic density around this bond.

Due to the great structural similarity between the Pd(II) compounds synthesized, a detailed evaluation was made of PdC3, which is the coordination compound with the greatest structural complexity in the series (all the other ¹H NMR spectra are available in Figures S2–S7). The H signals of the aromatic rings were found between the signals of the vinyl hydrogens. In the L-C3 ligand, the H(7) and H(8) signals were observed at 6.33 and 7.02 ppm, respectively; however, with coordination to Pd(II), there was a shift in the electronic density of the ligand to supply the positive charge of the metal, generating a change in the chemical environment of the vinyl signals, which were shifted to 6.90 [H(7)] and 8.18 [H(8)] ppm. The ¹H-¹H COSY experiment was performed for the PdC3 complex, allowing the observation of H–H correlations from both the TSC ligand and the triphenylphosphine coordinated to the metal center (Figure S8).

In the ¹³C{¹H} NMR spectrum (100 MHz) obtained for the PdC3 complex (Figure S9), it was possible to observe the signals for the aliphatic carbons at 13.69 and 40.27 ppm, which corresponded to the N-ethyl substituent of the TSC ligand. To identify the C17 signal, the DEPT-135 experiment was carried out (Figure S10). The signals referring to the carbons of the vinyl group (C7 and C8) were found at 119.70 and 150.18 ppm, and then the most unprotected signals were assigned to C9 (166.04 ppm) and C16 (207.12 ppm), which correspond to the carbons bonded directly to the N and S atoms, respectively, which participate in the coordination to the Pd(II) metal in the formation of the complex. The aromatic carbons are found in the 127 to 135 ppm region, where it was possible to identify the signals relating to the phenyl groups. Additionally, the ¹H-¹³C-HSQC spectrum was obtained for the PdC3 complex (Figure S11). The experiment corroborates the proposed structure.

The ³¹P{¹H} NMR experiment was also carried out and the chemical shift values of the different complexes synthesized were similar to each other. There was a shift in the free PPh₃ signal from −6.81 ppm to 30 ppm (Figure S12), indicating a reduction in the electronic density at the P of the triphenylphosphine due to coordination with Pd(II). Also, the PdB3 complex was tested for stability in DMSO. Using ³¹P{¹H} NMR, the complex analyzed remained stable after 48 h and showed no change in chemical shift (Figure S13).

According to the infrared spectra (Figure S14), in the region between 3500 and 3200 cm⁻¹, bands referring to the νN-H stretching are found. In the region of 3100 cm⁻¹, the νC-H stretching bands of sp² hybridization were assigned. Bands in the region below 3000 cm⁻¹ were observed for all the complexes, with the exception of the PdC1 complex, which has no carbons with sp³ hybridization. In the 1600–1400 cm⁻¹ region, strong intensity bands were detected referring to the νC=C and νC=N stretches. The band at 1090 cm⁻¹, referring to the asymmetric deformation τ(P-CPh), indicates the coordination of the PPh₃ ligand to the metal center. The band referring to the ν_SC=S stretch was found in the free TSC ligands at 750 cm⁻¹ and in the spectra of the complexes around 710 cm⁻¹, indicating the weakening of the C=S bond [28]. In addition, the band referring to the νC=N stretch was strengthened, showing that the coordination of the TSC ligand to Pd(II) resulted in changes around the thioamide bonds.

The mass spectrometry technique verified the presence of the complexes in their cationic form, as well as detecting the cation related to the labilization of Cl⁻ and H⁺ (Figures S15–S20). Both cations detected showed the Pd isotopic pattern, with the two most intense peaks referring to the ¹⁰⁶Pd and ¹⁰⁸Pd isotopes, which is in line with the isotopic abundance described in the literature [29,30]. The errors (in ppm) calculated from the theoretical masses and the experimental results did not exceed 3 ppm, which is a clear indication that the proposed structures were obtained. The calculation was carried out for the most intense isotope of the molecular ion [M]⁺.

According to the XRD data, the complexes could be divided into those that crystallized in a cationic and neutral form, and all the structural representations are available in Figures S21–S26 and the crystallographic data in Tables S1–S3. In all the structures obtained, the distorted square-planar geometry around the metal center was observed, and the complexes that crystallized in the neutral form showed a shorter N(2)-C(10) bond distance and a longer S-C(10) bond distance, which was an expected change due to the deprotonation of N(2), which generates a change in the electronic density of the complexes and, consequently, an increase in the double bond character between N and C(10). The effects of deprotonation could be seen in the Pd-S bond, wherein the neutral species this value is less than 2.261 Å, and in the cationic species it has more elongated bonds (2.267–2.279 Å). In the PdC3 system, the Pd-S bond difference between the neutral and protonated species was 0.036 Å.

Single crystals of the PdC3 complex were obtained in two different structures. The structure named PdC3-Cl was obtained from the reaction solution and had the TSC ligand as protonated and Cl⁻ as the counter-ion. In turn, the PdC3-DMSO structure had its single crystal obtained in a complex-DMSO solution and showed the TSC ligand deprotonated with the presence of the DMSO solvent in its three-dimensional structure and with the chloride ligand remaining bound to the Pd(II) complex. The crystallographic structures obtained for this same complex showed different conformations (Figure 1). This behavior is related to the characteristics of each solvent. The chloroform/acetonitrile solution is less basic than DMSO. Therefore, in DMSO, the complex deprotonates, while in the chloroform/acetonitrile solution it does not deprotonate. It is important to mention that this behavior does not influence the cytotoxicity of the compound, since the tests are carried out in buffered solutions.

3.2. Cell Tests

3.2.1. Cell Viability (IC₅₀)

The cytotoxic activity of the complexes was assessed against a range of tumor and non-tumor cell lines. For clarity, the cell viability results are grouped according to breast (

Table 1. IC₅₀ values (µmol·L⁻¹) of the complexes in 48 h of incubation for the human breast cell lines MDA-MB-231, MCF-7, SK-BR-3, and MFC-10A.

R1	R2	Complexes	MDA-MB-231	MCF-7	SK-BR-3	MFC-10A
	H	PdB1	1.45 ± 0.1	3.79 ± 0.4	1.42 ± 0.07	1.58 ± 0.07
Me	Me	PdB2	0.92 ± 0.08	1.47 ± 0.2	1.48 ± 0.06	―
	Et	PdB3	1.04 ± 0.06	0.91 ± 0.008	<2	―
	H	PdC1	5.97 ± 0.1	17.68 ± 1	2.25 ± 0.3	2.64 ± 0.06
Ph	Me	PdC2	9.66 ± 0.4	>50	>50	―
	Et	PdC3	>50	>50	>50	―
		Cisplatin	33.14 ± 0.3	8.64 ± 0.8	5.86 ± 0.2	23.90 ± 0.7

Abbreviations: Me (methyl), Et (ethyl), and Ph (phenyl).

), ovarian (

Table 2. IC₅₀ values (µmol·L⁻¹) of the complexes in 48 h of incubation for the ovarian tumor cell lines A2780 and A2780cis, with the resistance factor (RF) value.

R1	R2	Complexes	A2780	A2780cis	RF
	H	PdB1	0.12 ± 0.01	0.20 ± 0.01	1.66
Me	Me	PdB2	0.13 ± 0.01	0.21 ± 0.01	1.62
	Et	PdB3	0.23 ± 0.01	0.26 ± 0.01	1.13
	H	PdC1	0.40 ± 0.01	0.62 ± 0.01	1.55
Ph	Me	PdC2	3.93 ± 0.2	1.66 ± 0.01	0.42
	Et	PdC3	22.78 ± 1	11.1 ± 0.6	0.48
		Cisplatin	14.24 ± 0.05	25.09 ± 0.9	1.76

), and lung and skin (

Table 3. IC₅₀ values (µmol·L⁻¹) of the complexes in 48 h of incubation for A549, MRC5, and A375 cell lines.

R1	R2	Complexes	A549	MRC5	A375
	H	PdB1	44.8 ± 2	1.70 ± 0.2	1.34 ± 0.08
Me	Me	PdB2	30.4 ± 0.4	6.85 ± 0.3	―
	Et	PdB3	12.99 ± 1	9.28 ± 0.3	―
	H	PdC1	>50	46.77 ± 2	>50
Ph	Me	PdC2	>50	>50	―
	Et	PdC3	>50	>50	―
		Cisplatin	13.01 ± 0.4	21.61 ± 0.8	12.21 ± 0.62

) cell lines. The thiosemicarbazone ligands and [PdCl₂(MeCN)₂] were tested against the MDA-MB-231, A549, MRC5, and A2780cis cell lines. For these cell lines, the TSC ligands and Pd(II) precursor showed no cytotoxic effect at the tested concentrations, generally presenting IC₅₀ values higher than 50 µmol·L⁻¹.

For breast cancer, the tumor cell lines MDA-MB-231, MCF-7, and SK-BR-3 were tested, along with the non-tumorigenic line MCF-10A, which was included only for the PdB1 and PdC1 complexes. Overall, the compounds demonstrated high cytotoxicity, with low IC₅₀ values in most of the tested lines, except for those in the C series. Specifically, PdC2 and PdC3 were largely inactive in breast cancer lines, while PdC1 exhibited significantly lower IC₅₀ values compared to the standard drug cisplatin, except for the MCF-7 cell line (Figure 2).

Among all tested compounds, those in the B series were the most potent, consistently showing the lowest IC₅₀ values towards both tumor and non-tumor breast cell lines. The compounds also overcome cisplatin, which was used as a reference drug. Notably, their activity against MDA-MB-231 and SK-BR-3 was constant despite structural variations in the R₂ substituent, suggesting that these changes had little effect on the efficacy of these lines. In contrast, for MCF-7 cells, increasing the size of the R₂ group increased activity, with PdB3 yielding the lowest IC₅₀ value in this line.

To further evaluate the potential of these complexes, their activity was also tested in ovarian cancer cell lines A2780 and A2780cis, the latter being resistant to cisplatin. This comparison aimed to determine whether resistance mechanisms affecting cisplatin would also impact the new compounds.

Interestingly, IC₅₀ values in ovarian cell lines were generally lower than in other tumor types, indicating strong cytotoxic effects, as well as other recently reported palladium complexes [31,32,33]. The resistance factor (RF), calculated as IC₅₀(A2780cis)/IC₅₀(A2780), revealed that B series complexes were more effective against the non-resistant A2780 line (RF > 1), suggesting reduced efficacy in resistant cells. In contrast, the C series showed mixed behavior, with some compounds maintaining activity in both lines.

Within the C series, an inverse correlation emerged between the size (or molecular weight) of the R₂ substituent and cytotoxic activity: larger substituents led to higher IC₅₀ values. On the other hand, the B complexes remained largely unaffected by changes to R₂, exhibiting a more consistent cytotoxic profile.

Contrary to the strong activity observed in breast and ovarian cell lines, the complexes exhibited lower cytotoxicity against the A549 lung cancer cell line. Of the series tested, only the B family showed measurable IC₅₀ values within the concentration range evaluated. Interestingly, an inverse trend was observed for A549 compared to the MCF-7 cell line: the complex with the largest R₂ substituent (PdB3) showed the highest activity and yielded the lowest IC₅₀ within this series. Nevertheless, the overall IC₅₀ values for A549 remained higher than those observed for other tumor models, indicating a generally reduced sensitivity.

For the non-tumor lung fibroblast line MRC-5, the cytotoxic profile followed previous trends, with the B series again showing superior activity. In this case, however, the lowest IC₅₀ was observed for the complex with the smallest R₂ group (PdB1), suggesting that bulkier substituents may reduce efficacy in non-tumor lung cells.

In addition, two representative complexes, PdB1 and PdC1, were evaluated against the A375 melanoma cell line. Of the two, only PdB1 had a significant effect on cell viability, further confirming the higher potency and broader spectrum of activity of the B compounds across multiple tumor types.

In summary, the evaluation of cytotoxic activity in various tumor and non-tumor cell lines revealed consistent trends in structure–activity relationships, particularly highlighting the superior performance of the B-Series. A key observation in multiple cell lines was the increased cytotoxicity associated with the presence of a terminal methyl group (-CH₃) in the R₂ substituent. This is following the literature reports suggesting that methylation of bioactive molecules can enhance antiproliferative effects and inhibit cell migration, possibly by altering membrane permeability or interaction with intracellular targets [34,35,36].

To evaluate the therapeutic potential of the complexes, the selectivity index (SI) was calculated (

Table 4. Selectivity index (SI *) in relation to the non-tumor cell line MRC5.

Complexes	MDA-MB-231	MCF-7	SK-BR-3	A2780	A2780cis	A549
PdB1	1.2	0.4	1.2	14.2	8.5	0.0
PdB2	7.4	4.7	4.6	52.7	32.6	0.2
PdB3	8.9	10.2	―	40.3	35.7	0.7
PdC1	7.8	2.6	20.8	116.9	75.4	―
Cisplatin	0.7	2.5	3.7	1.5	0.9	1.7

* SI = [IC₅₀ of non-tumor cell line]/[IC50 of tumor cell lines].

) as the ratio between the IC₅₀ values for the non-tumor MRC-5 line and those for the tumor cell lines. For synthetic organic and organometallic compounds, SI values greater than 10 are considered indicative of promising selectivity [37,38,39]. Notably, the PdC1 complex exhibited exceptionally high SI values of 75.4 for the resistant A2780cis and 116.9 for the non-resistant A2780 line, suggesting strong selectivity for ovarian tumor cells.

Based on these findings, the PdC1 complex was selected for further in-depth studies to elucidate its mechanism of action in the in vitro tumor microenvironment. The PdB1 complex was also included in the subsequent experiments due to its close structural similarity to PdC1 and its consistent cytotoxic profile, providing a complementary perspective to understand the influence of structural variations on biological activity.

3.2.2. Clonogenic Assay

This assay allows evaluation of the effects of compounds on the long-term proliferative potential of cancer cells, specifically their ability to survive, divide, and form colonies after treatment. Colony formation assays reveal whether cells can recover and proliferate after exposure to compounds, thus providing insight into both cytotoxic and cytostatic effects.

In this context, both the PdB1 and PdC1 complexes exhibited a predominantly cytotoxic profile (Figure 3 and Figure 4). In particular, the PdB1 complex significantly reduced colony formation even at sub-IC₅₀ concentrations, suggesting that it interferes with essential cellular functions required for survival and regrowth. This suggests a potent effect not only on rapidly dividing cells, but potentially on the clonogenic potential of tumor cells, which is closely linked to tumor progression and recurrence.

The ability of PdB1 to exert such inhibitory effects at lower concentrations may be due to its structural characteristics, including the presence of a terminal methyl group in the R₂ substituent, which has been associated with enhanced cellular uptake and interaction with biomolecular targets. PdC1 also reduced colony formation, although to a lower extent, which is consistent with its lower activity in certain cell lines but still supports its cytotoxic potential.

These results reinforce the relevance of colony formation assays as a complement to IC₅₀-based assessments and highlight the potential of PdB1, especially at lower, sublethal concentrations, as a candidate for further studies aimed at targeting tumor-cell persistence and resistance mechanisms.

3.2.3. Cell Morphology Assay

Evaluation of cell morphology provides valuable insight into the effects of compounds on cell structure and viability. In this study, we used the A2780cis tumor cell line, which is characterized by an adherent growth pattern. Changes such as decreased cell density and the presence of cellular debris serve as morphological indicators of cytotoxicity and cell death.

Microscopic analysis of treated A2780cis cells revealed a marked cytotoxic response at concentrations above the IC₅₀ of the compounds tested (Figure 5 and Figure S27). Specifically, we observed a significant reduction in the number of adherent cells, the appearance of rounded and detached cells often indicative of apoptosis or loss of membrane integrity, and an accumulation of cell fragments. These morphological changes strongly suggest that the compounds exert a potent cytotoxic effect at higher concentrations.

The correlation between increased compound concentration and the observed morphological changes reinforces the potential of these agents to disrupt cellular homeostasis and induce cell death.

In addition to the initial morphological evaluation, further analysis was performed using fluorescent staining to better characterize the cellular changes induced by the compound PdB1. Two markers were used: DAPI, which binds to DNA and highlights nuclear morphology, and Green Plasma, which stains the plasma membrane. This dual-staining approach allowed for more detailed observation of potential morphological changes associated with cell death mechanisms.

Following treatment of A2780cis cells with PdB1, starting at concentrations as low as ½ × IC₅₀, a significant shift in cell morphology was observed (Figure 6). Green Plasma staining revealed a transition to a more spherical shape and a reduction in total cell area, suggesting membrane retraction and early apoptotic features. At the same time, DAPI staining showed pronounced nuclear changes, including chromatin condensation and nuclear fragmentation hallmarks of apoptosis. These morphological changes became increasingly evident with increasing concentrations of PdB1, supporting a dose-dependent induction of apoptosis.

3.2.4. Cell Viability (Kit: LIVE/DEAD^®)

To further investigate the cytotoxic effects of PdB1 on A2780cis cells, viability was assessed using the LIVE/DEAD^® viability/cytotoxicity kit. This assay distinguishes between live and dead cells based on membrane integrity and esterase activity. Viable cells fluoresce green due to intracellular conversion of non-fluorescent calcein-AM to fluorescent calcein, while non-viable cells fluoresce red due to uptake of ethidium homodimer-1 (EthD-1), which binds to nucleic acids in cells with compromised membranes [40,41].

Fluorescence microscopy revealed a marked increase in red-fluorescing (non-viable) cells beginning at ½ × IC₅₀, indicating early signs of cell death (Figure 7). This effect intensified with increasing concentrations of PdB1, suggesting a clear dose-dependent response.

Notably, the viability profile observed with the LIVE/DEAD^® assay closely matched the MTT assay results, reinforcing the reproducibility and reliability of the findings. These consistent results confirm the cytotoxic efficacy of PdB1 against A2780cis cells.

Taken together with the morphological changes observed by both brightfield and fluorescence staining (DAPI and Green Plasma), the data support that PdB1 induces apoptotic cell death, likely through mechanisms involving membrane disruption and nuclear condensation. In summary, these findings highlight the therapeutic potential of PdB1 as a candidate for targeting cisplatin-resistant ovarian cancer cells.

3.2.5. Cell Migration Assay (Wound Healing)

This experiment was designed to evaluate the ability of the Pd(II) complexes to inhibit cell migration using a wound healing assay. Cell migration was assessed by comparing the initial wound area with the area remaining after treatment. As expected, the control group treated with 0.5% (v/v) DMSO showed the highest degree of wound closure, indicating robust migratory activity in the absence of cytotoxic compounds.

In contrast, cells treated with Pd(II) complexes showed a marked reduction in wound healing, suggesting an impaired migratory capacity (Figure 8). The inhibitory effect on migration was most pronounced at concentrations near IC₅₀, where approximately 50% of the wound area remained unhealed. This observation suggests that the compounds not only exert cytotoxic effects, but also interfere with cellular mechanisms involved in migration, a key factor in cancer invasion and metastasis.

3.2.6. Cell Cycle Analysis

Cell cycle analysis was performed to determine whether treatment with the Pd(II) compounds leads to alterations in cell cycle distribution or induction of cell death in A2780cis cells. Experiments were performed at concentrations corresponding to IC₅₀.

The results (Figure 9) showed a clear accumulation of cells in the sub-G1 phase after treatment. This phase is not part of the normal cell cycle, but reflects a population of cells with fragmented DNA, which is often associated with apoptosis. Thus, the increase in sub-G1 does not indicate classical cell cycle arrest, but instead the presence of dead or dying cells [42].

These results suggest that the primary effect of the Pd(II) compounds at these concentrations is the induction of cell death via apoptosis rather than the interruption of specific cell cycle checkpoints. This is consistent with other assays performed in this study that indicate apoptotic features, further supporting the cytotoxic potential of the compounds.

3.2.7. Apoptosis Assay by Anexin V

To further investigate the mechanism of PdB1-induced cell death, an apoptosis assay was performed using the Annexin V-FITC/PI kit followed by flow cytometry analysis. This approach allows differentiation between live, early apoptotic, late apoptotic, and necrotic cells based on phosphatidylserine exposure and membrane integrity.

Analyzing Figure 10, it is observed that in the control group (treated with 0.5% DMSO), the majority of the cell population (86.4%) was located in Q3, indicating a predominantly viable population with intact membranes and no phosphatidylserine externalization. However, after treatment with PdB1 at ½ × IC₅₀, a significant reduction in viable cells was observed, with only 37.1% remaining in Q3. At the same time, there was a significant increase in the Q2 quadrant, indicating cells undergoing late apoptosis.

This dose-dependent shift in the apoptotic profile suggests that PdB1 promotes programmed cell death rather than necrosis. The relatively modest presence of cells in Q1 (necrosis) further supports this conclusion. The increase in the Q2 population at sub-lethal concentrations also suggests that PdB1 initiates a time-dependent or progressive apoptotic process that culminates in late-stage apoptosis as exposure continues.

These results are consistent with previous findings from sub-G1 cell cycle analysis and DAPI staining, which demonstrated nuclear condensation and DNA fragmentation, both typical features of apoptosis. Taken together, these data confirm that apoptosis is the primary mode of cell death induced by PdB1 and support its potential as a pro-apoptotic agent against cisplatin-resistant ovarian cancer cells.

3.2.8. Three-Dimensional Assay

The morphological assay in 3D cell culture was performed using the A2780cis tumor cell line to evaluate structural changes induced by the PdB1 complex in a more physiologically relevant and drug-resistant tumor model. Unlike traditional 2D monolayer cultures, 3D spheroid models better mimic the in vivo tumor microenvironment by promoting cell–cell and cell–matrix interactions, as well as gradients of oxygen, nutrients, and drug penetration. This model is particularly useful for evaluating compounds targeting chemoresistant cancers [43,44,45].

Brightfield microscopy images taken after 48 h of treatment showed continued spheroid growth at lower concentrations of PdB1, indicating limited cytotoxicity under these conditions (Figure 11). However, significant morphological changes were observed at a concentration of 25 µM. The spheroids showed a marked reduction in compaction and cohesion, with disrupted borders and visible disintegration of the spheroid structure. These features are indicative of loss of cell–cell adhesion and structural integrity, often associated with cell death within the spheroid core and periphery.

Such changes suggest that PdB1 is able to penetrate the dense 3D architecture and exert cytotoxic effects even under conditions that typically confer resistance to conventional treatments such as cisplatin. The morphological disruption observed at 25 µM highlights the potential of PdB1 to overcome microenvironment-mediated resistance mechanisms.

Fluorescence staining with DAPI and propidium iodide (PI) was used to further evaluate the extent and spatial distribution of cell death within A2780cis spheroids following treatment with the PdB1 complex (Figure 12). In this assay, DAPI selectively binds to DNA and labels the nuclei of all cells with blue fluorescence, while PI is membrane-impermeable and selectively labels the nuclei of dead cells with compromised membrane integrity with red fluorescence. The overlap of the two signals results in a purple hue, indicating nuclei of cells that have undergone cell death.

Fluorescence microscopy images revealed a clear dose-dependent response to PdB1 treatment. At lower concentrations, DAPI staining was predominant, indicating a higher proportion of intact, viable cells. In contrast, at 12.5 µM, the presence of strong PI fluorescence indicated a significant level of cell death within the spheroids, particularly around the outer layers where drug penetration is initially more effective. At the highest concentration tested (25 µM), the spheroids showed complete disintegration, accompanied by intense violet staining resulting from the co-localization of DAPI and PI signals. This strongly suggests that PdB1 at this concentration induces widespread and possibly complete cell death within the 3D structure.

In particular, the diffuse PI staining and internal labeling of the spheroids demonstrate that PdB1 has effective permeability, allowing it to reach deeper cell layers within the tumor-like structure. However, when compared to 2D monolayer assays such as MTT, it is evident that higher concentrations are required to induce similar cytotoxic effects in 3D culture, underscoring the increased resistance of spheroids due to factors such as limited drug diffusion, altered cell metabolism, and enhanced survival pathways.

4. Conclusions

In this study, new Pd(II) complexes were successfully proposed, synthesized, and characterized. The structures of most compounds were elucidated, and single crystals were obtained. Cellular assays revealed that the B-family complexes were more toxic than those of the C family, with the R1 substituent having a greater influence on cell viability, while R2 modulated cytotoxicity depending on the cell lineage. The R1 group played a crucial role in determining the cytotoxic and cytostatic profiles, with aldehydes exhibiting a cytostatic effect near the IC₅₀ and ketones displaying a cytotoxic profile. The carbonyl group modulated cytotoxic activity, with smaller aldehydes and ketones increasing cytotoxicity, while bulkier substituents reduced this effect. The complexes also demonstrated antimigratory potential, the ability to alter cell morphology, and apoptosis induction. The 3D assay indicated that the PdB1 complex effectively induces cell death in physiologically relevant tumor environments, highlighting its potential as a promising candidate in the search for new metallodrugs.