Thiazole Functionalization of Thiosemicarbazone for Cu(II) Complexation: Toward Highly Efficient Anticancer Drugs with Promising Oral Bioavailability

1. Introduction

Thiosemicarbazones (TSCs) are a class of Schiff-base derivatives that exhibit diverse biologically beneficial activities, such as antibacterial, antiviral, and enzyme inhibitory activities [1,2]. Most notably, TSCs have long been reported as a highly promising class of drugs for a broad-spectrum of cancers through a wide range of anticancer mechanisms, such as ribonuclease reductase and topoisomerase II inhibition [3,4,5,6]. Other mechanisms include the quick sequestration of cell-proliferative-dependent ions (i.e., Fe³⁺ and Cu²⁺) to elicit the anticancer effect [4]. The anticancer effect of TSCs is primarily linked to, and profoundly affected by, their structures that feature an S, N chelator [4,5]. In this regard, additional introduction of an N coordination site, such as α-pyridyl serves as an advantage as it forms a more stable N, N, S pincer-like chelation that results in more effective ion removal [7]. In this sense, TSCs are particularly effective for treating iron overload diseases such as leukemia and neuroblastoma [8]. Recently findings also suggested that, upon capturing of metal ions by TSCs, a different anticancer mechanism based on reactive oxygen species (ROS) chemistry can be commenced by taking advantage of the unique characteristics of the tumor microenvironment (TME), such as the overexpression of H₂O₂ and GSH [9,10,11,12]. The ROS generation can subsequently strengthen the chemotherapeutic outcome of TSCs.

The promising potential of TSCs as anticancer drugs has been manifested by several drugs that successfully entered clinical trials, such as 3-aminopyridine-2-carboxaldehyde thiosemicarbazone (Triapine), 5-hydroxypicolinaldehyde thiosemicarbazone (5-HP), and 4-(pyridine-2-yl)-N-([(8E)-5,6,7,8-tetrahydroquinolin-8-ylidene]amino)piperazine-1-carbothioamide (COTI-2) (Chart 1). Nevertheless, numerous side effects were also associated with these drugs, i.e., the formation of methemoglobin hypoxia by Triapine[13], severe hematological and gastrointestinal side effects for 5-HP[14], and drug resistance and other adverse effects (i.e., nausea, vomiting, and fatigue) for COTI-2 [15,16].

In this work, we report the synthesis of a novel TSC-based anticancer drug of N′-(di(pyridin-2-yl)methylene)-4-(thiazol-2-yl)piperazine-1-carbothiohydrazide (HL) featuring di-2-pyridylketone for chelating-bridging and a thiazole spectator (Chart 1). Ligand HL can readily associate Cu²⁺ to give the corresponding coordination complexes of [CuCl(L)]₂ (1), and [Cu(NO₃)(L)]₂ (2). The structures of HL, 1, and 2 were confirmed by various spectroscopic techniques and ultimately authenticated by single-crystal X-ray diffraction studies. HL, 1, and 2 are highly toxic against hepatocellular carcinoma cell lines Hep-G2, PLC/PRF/5, and HuH-7. Notably, these species are extremely toxic for PLC/PRF/5, with the corresponding half maximal inhibitory concentration (IC₅₀) values as low as 3.26 nmol/mL (HL), 2.18 nmol/mL (1), and 2.54 × 10⁻⁵ nmol/mL (2). Our preliminary pharmacokinetic studies of HL further revealed that oral administration (per os, PO) exhibits a significantly longer half-life t_1/2 of 21.61 ± 9.4 h as compared to that of the intravenous (i.v.) administration of 11.88 ± 1.66 h. These results highlight that HL and its relevant functional coordination complexes can be promising chemotherapeutic drugs for oral administration.

2. Results and Discussion

2.1. Synthesis and X-ray Structure Characterization of HL, [CuCl(L)]₂ (1), and [Cu(NO₃)(L)]₂ (2)

The synthesis of ligand HL was achieved by a five-step protocol to give an overall yield of 9.3% (Scheme 1). The purity of HL was confirmed by microanalysis, ¹H, and ¹³C nuclear magnetic resonance (NMR) spectroscopy (Figure S1). The assumed connectivity of HL was also unambiguously determined by single-crystal X-ray diffraction studies (Figure 1a).

Coordination complexes of 1 and 2 were prepared accordingly by reacting HL and the respective Cu(II) sources under appropriate conditions (Scheme 2). Single crystal X-ray structure data for 1 (Figure 1b) and 2 (Figure 1c) indicated that both compounds form dimeric structures in the solid state, enabled by the head-to-tail complementary Cu−N coordination. In 1, each Cu²⁺ shows a square pyramidal coordination geometry, fulfilled by an S, one azomethinic N, one α-pyridyl N atom from an L ligand, and one Cl⁻ that defines the square plane, in addition to one pyridyl N atom from adjacent L ligand that occupies the apical position. A similar connectivity is also found in 2 except that one Cl⁻ is replaced by a chelating NO₃⁻, yielding a distorted octahedral coordination geometry for Cu²⁺.

Notably, the ligands in both 1 and 2 undergo a change of conjugation with the loss of hydrazinic proton and the conversion of C=S into C−S bond. Such structure variation is seen in a variety of transition metal complexes of α-pyridyl thiosemicarbazone chelators, such as Triapine complexes of Fe³⁺/Ga³⁺[17,18], di-2-pyridyl ketone 4,4-dimethyl-3-thiosemicarbazone (HDp44mT) complexes of Fe³⁺[19], as well as several Cu²⁺ complexes [20]. Bormio Nunes et al. also reported that, for Fe³⁺/Cu²⁺ complexes of COTI-2, the N−N single bond of the ligand is inherited in the final chelating complexes [16]. As a result of the C=S to C−S bond variation, the C=S bond distance in HL of 1.681(5) Å increased to 1.739(4) Å in 1 and 1.744(6) Å in 2 (Table S1). Nevertheless, the N−N distance in HL of 1.354(5) Å remained nearly unaltered as compared to the N=N distances in 1 (1.354(4) Å) and 2 (1.351(6) Å) due to the additional coordination of one N to Cu²⁺.

2.2. Spectroscopic Characterizations of HL, 1, and 2

Energy dispersive X-ray spectroscopy (EDS) revealed an atomic ratio of Cu : Cl : S of 2.0 : 2.2 : 4.2 in 1 and Cu : S of 2.0 : 4.6 in 2 (Figure S2), consistent with the derived ratio of Cu : Cl : S of 1.0 : 1.0 : 2.0 (1) and Cu : S of 1.0 : 2.0 (2) from single-crystal data, and the elements Cu, S, N and C are evenly distributed in their elemental mapping diagrams (Figure S2). The Fourier-transform infrared (FT-IR) spectra show that upon coordination with Cu²⁺, the C=N bond vibration peaked at 1571 cm⁻¹ in HL shifts to 1593 cm⁻¹ for both 1 and 2 (Figure S3). Meanwhile, bands at 850 cm⁻¹ (1) and 856 cm⁻¹ (2) which were assignable as C=S bond vibrations, also shifted as compared to that of HL (869 cm⁻¹) upon bonding with Cu²⁺ [21,22]. In the ultraviolet-visible (UV-Vis) spectroscopy (Figure 2), ligand-to-metal charge transfer (LMCT) bands in 1 and 2 are found at 426 nm as compared to HL [23]. Meanwhile, the ligand-centered UV-Vis absorption band of the HL undergoes a significant blue shift due to the conjugation change and charge transfer reactions. The LMCT leads to the partial reduction of Cu²⁺ which is favorable for Cu⁺-induced reactions, such as the ROS generation via the Fenton-like process [24]. The Cu 2p X-ray photoelectron spectroscopy (XPS) of both 1 and 2 revealed two sets of binding energies (Figure 3). Peaks at 933.8/953.5 eV (1) and 933.8/953.8 eV (2) are assignable as the spin-orbit splitting of Cu 2p_3/2 and 2p_1/2 of Cu²⁺ [25,26,27]. Two satellite peaks at 943.3/963.3 eV (1) and 943.4/963.0 eV (2) are also attributed to Cu²⁺ ions due to its d⁹ electronic configuration that is susceptible to photoreduction to give a more stable d¹⁰ configuration.

2.3. Hydroxyl Radical (•OH) Production

The generation of •OH from H₂O₂ by HL, 1, and 2 was assayed by 3,3′,5,5′-tetramethylbenzidine (TMB) [12,28,29]. The colorless TMB can be quickly oxidized by •OH to give a characteristic blue charge-transfer complex oxTMB with intense absorption at 652 nm [30,31]. Different concentrations of HL (5, 10, 20, 30 μg/mL) or 1 and 2 containing equivalent HL were treated with (TMB) solution containing 100 μM H₂O₂, which is assumed as approximately the concentration reported to be present in tumor cells, or five times that in normal cells [32,33]. After 20 min of incubation with HL, there is no oxTMB found as suggested by both the UV-Vis and the direct observation of solution colors (Figure S4). In sharp contrast, both 1 and 2 can induce the generation of •OH under otherwise identical conditions in a concentration-dependent manner with 2 further outperforms 1 (Figure 4, and Figure S4). In addition, the MeOH solutions of HL, 1, and 2 at HL / L concentrations of 30 μg/mL remained stable upon keeping for 72 h (Figure S5), as evidenced by their unaltered UV-Vis absorption patterns throughout the experiments.

2.4. CCK-8 Assay for HL, 1, and 2

The cytotoxicity of HL, 1, and 2 against three hepatocellular carcinoma cell lines Hep-G2, PLC/PRF/5, and HuH-7 were evaluated by cell counting kit-8 (CCK-8) assay. As shown in Figure 5, all three complexes demonstrate superior cytotoxicity originating from the ligand, particularly for the PLC/PRF/5 cell line. Notably, 1 and 2 demonstrated higher cytotoxicity as compared to the HL, presumably due to the additional ROS generation induced by the overexpressed H₂O₂ in the cancer cells. The IC₅₀ values for HL, 1, and 2 against PLC/PRF/5 were as low as 3.26 nmol/mL, 2.18 nmol/mL, and 2.54 × 10⁻⁵ nmol/mL, which is favorably lower than those against Hep-G2 (78.83 nmol/mL for HL, 38.11 nmol/mL for 1, and 16.86 nmol/mL for 2), and HuH-7 (192.20 nmol/mL for HL, 176.60 nmol/mL for 1, and 87.53 nmol/mL for 2) cell lines (Table S2).

2.5. Intracellular ROS Generation by HL, 1, and 2

Given the efficient •OH production in water induced by 1 and 2, we next investigated in vitro ROS generation by 1 and 2 in the representative HuH-7 cell line using 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) as a fluorescence probe, and compared to that of HL [34,35]. The non-emissive DCFH-DA can readily be uptaken by the cell, where it is hydrolyzed into 2,7-dichlorodihydrofluorescein (DCFH) by cellular enzymes. Upon oxidation by ROS, DCFH can convert to 2,7-dichlorofluorescein (DCF) which gives bright green fluorescence [29,36]. As shown in Figure 6, HL alone produced negligible ROS. In sharp contrast, 1 and 2 can induce the formation of strong fluorescence characterizing the effective generation of ROS in the HuH-7 cells.

2.6. In Vivo Pharmacokinetic Study of HL.

For pharmacokinetic studies, CD^®(SD) IGS Rats (5−6 weeks old) were randomly divided into two groups (3 for each group) for oral administration (per os, PO) or intravenous administration (i.v.) of HL with dosages of 30 mg kg⁻¹ and 0.5 mg kg⁻¹, respectively (Tables S4 and S5). The blood was then collected from the orbital sinus with a heparinized syringe at different time intervals (0.5, 1.0, 2.0, 4.0, 8.0, and 24 h for PO, and 0.03, 0.5, 1.0, 2.0, 4.0, 8.0, and 24 h for i.v.). High-performance liquid chromatography (HPLC) gives the mean area under the curve (AUC_0-inf) values of 10409.01 ± 5313.4 ng h mL⁻¹ for PO, and 5701.36 ± 2647.15 ng h mL⁻¹ for i.v. HL achieved the largest C_max value of 425.93 ± 176.1 ng mL⁻¹ at a T_max of 1.0 h for PO, while for i.v., the C_max value of 806.43 ± 46.66 ng mL⁻¹ was immediately observed at T_max of 0.03 h, which is expected for i.v. administration. Notably, the PO administration exhibits a significantly longer half-life t_1/2 of 21.61 ± 9.4 h, nearly doubled as compared to that of the i.v. administration of 11.88 ± 1.66 h (Figure 7), certifying that HL has a favorable pharmacokinetic via PO administration.

3. Materials and Methods

3.1. General

tert-Butyl piperazine-1-carboxylate (BOC-PIP; ≥99%, Aladdin), 2-bromothiazole (98%, Crgent Biotech), di-2-pyridyl ketone (98%, Crgent Biotech), NH₂NH₂ (≥95%, adamas), Cu(NO₃)₂ (≥99%, Aladdin), CuO (≥97%, Aladdin), DMF (≥99%, Aladdin), CH₂Cl₂ (≥99%, Aladdin), trifluoroacetic acid (TFA; ≥99%, Aladdin), i-PrOH (≥99%, Aladdin), methyl tert-butyl ether (MTBE, ≥99%, Aladdin), NaHCO₃ (>99%, Crgent Biotech), EtOH (≥95%), and Cs₂CO₃ (≥99%, Crgent Biotech) were commercially available and used without further purifications.

Hep-G2, and PLC/PRF/5, and HuH-7 cell lines were purchased from the Shanghai Institute of Cell Biology, Chinese Academy of Sciences. Phosphate buffer solution (PBS) was purchased from Shanghai Basal Media Technologies Co., Ltd. Cell culturing medium MEM (with NEAA + 10% FBS + 1% P/S), MEM (with NEAA + 1% P/S), DMEM (10% FBS + 1% P/S), McCoy’s 5A (10% FBS + 1% P/S), DMEM (1% P/S), and 0.25% trypsin solution (containing EDTA, dissolved in PBS) were purchased from Procell Life Science & Technology Co., Ltd. The cell counting kit-8 (CCK-8) and the reactive oxygen species detection kit were from APEXBIO and Shanghai Beyotime Biotechnology Co., Ltd., respectively.

¹H nuclear magnetic resonance (NMR) spectra were obtained on a Varian UNITY plus-400/plus-600 NMR spectrometer (Varian, Inc., Palo Alto, CA, USA). Fourier transform infrared (FT-IR) spectra were measured on a Bruker VERTEX 70+HYPERION 2000 FT-IR spectrometer (Bruker AXS GmbH, Germany) using the attenuated total refraction (ATR) technique. Elemental analyses for C, H, and N were measured on a Carlo-Erba CHNO-S microanalyzer (Carlo Erba, Waltham, MA, USA). Ultraviolet-visible (UV-Vis) spectroscopy was obtained on a Varian Cary-50 UV-visible spectrophotometer (Varian, Inc., Palo Alto, CA, USA). X-ray photoelectron spectroscopy (XPS) was performed on an EXCALAB 250 XI X-ray photoelectron spectrometer (Thermo Scientific, Waltham, MA, USA). The high performance liquid chromatography (HPLC) were carried out on an Angilient 1260 Infinity II Bio-SEC system (Agilent Technologies, Inc., CA, USA). The CCK-8 cytotoxicity assay was conducted on a multifunction microplate detector by recording the absorption at 450 nm using a TECAN M1000PRO microplate reader (Tecan, Zürich, Switzerland).

3.2. Synthetic Steps for HL

Ligand HL can be synthesized via a five-step process as indicated below.

Step 1. Synthesis of S1. t-Butyloxy carbonyl (BOC) protected piperazine (BOC-PIP; 400.0 mg, 2.148 mmol) and 2-bromothiazole (529.0 mg, 3.221 mmol) were dissolved in DMF (4 mL) and stirred for 5 min. Cs₂CO₃ (1400 mg, 4.295 mmol) was added as a solid and the mixture heated overnight at 120°C. The formed product was extracted from H₂O using ethyl acetate to obtain S1 as a white solid. Yield (290 mg, 50% based on BOC-PIP). ¹H NMR (400 MHz, CDCl₃) δ 7.209 (s, 1H), 6.605 (s, 1H), 3.566 (s, 4H), 3.466 (s,4H), 1.483 (s, 9H). ¹³C NMR (101 MHz, CDCl₃) δ 172.096, 154.601, 139.567, 107.814, 80.273, 48.456, 28.396.

Step 2. Synthesis of S2. Compound S1 (290.0 mg, 1.078 mmol) was dissolved in CH₂Cl₂ (4 mL) to obtain a light yellow-brown solution and trifluoroacetic acid (TFA, 12 mL) was subsequently added, and the solution quickly turned yellow upon stirring. The mixture was stirred for 2 h and the solvent evaporated, DCM / MeOH = 2 : 1 and NaHCO₃ were consequently introduced to obtain S2 as a light yellow powder. Yield (180 mg, 99% based on S1). ¹H NMR (400 MHz, DMSO-d₆) δ 9.845 (s, 1H), 7.360 (s, 1H), 7.088 (s, 1H), 5.039 (s, 7H), 3.843 (s, 4H), 3.257 (s, 4H). ¹³C NMR (101 MHz, DMSO-d₆) δ 170.534, 133.871, 110.024, 46.268, 41.908.

Step 3. Synthesis of S3. Complex S2 (180.0 mg, 1.065 mmol) was added to CH₂Cl₂ (5 mL) to form a suspension, and Cs₂CO₃ (342.0 mg, 1.917 mmol) was introduced to give a yellow solution. The mixture was stirred at r.t. for 15 h, and the product was washed with H₂O and extracted with CH₂Cl₂. The organic layers were combined and dried. The pure yellow compound of S3 can be isolated by chromatography using CH₂Cl₂/i-PrOH (25 : 1, v/v) as the eluent. Yield (270 mg, 91% based on S2). ¹H NMR (400 MHz, CDCl₃) δ 7.904 (s, 1H), 7.234 (s, 2H), 7.111 (s, 1H), 6.683 (s, 1H), 4.040 (s, 4H), 3.656 (s, 4H). ¹³C NMR (101 MHz, CDCl₃) δ 179.348, 170.996, 139.720, 137.427, 130.303, 119.252, 108.941, 50.694, 48.032.

Step 4. Synthesis of S4. Compound S3 (270.0 mg, 0.968 mmol) was added to EtOH (4 mL) to form a suspension, and NH₂NH₂ (0.0063 mL, excess) was subsequently added. The mixture was heated to reflux at 80°C to yield a white suspension of S4 which was purified by chromatography using CH₂Cl₂/MeOH (25 : 1, v/v) as the eluent. Yield (175 mg, 74% based on S3). ¹H NMR (400 MHz, DMSO-d₆) δ 9.225 (s, 1H), 7.189 (s, 1H), 6.881 (s, 1H), 4.802 (s, 2H), 3.882 (s, 4H), 3.417 (s, 4H). ¹³C NMR (101 MHz, DMSO-d₆) δ 183.080, 171.525, 139.917, 108.901, 48.052, 46.913.

Step 5. Synthesis of HL. Compound S4 (40.0 mg, 0.165 mmol) was dissolved in EtOH (10 mL) and stirred for 10 min. Di-2-pyridyl ketone (34.0 mg, 0.181 mmol) was subsequently introduced and the mixture was heated to reflux at 80°C overnight. Upon cooling to r.t., the crude powder of HL was recrystallized with MeOH / MTBE (5 : 1, v/v) to give a pure product of yellow-green solid of HL. Yield (18.8 mg, 28% based on S4). Single crystals of HL were grown by slow evaporation of a MeOH/DMSO (v/v = 1 : 1) solution of HL. Anal. Calcd for C₁₉H₁₉N₇S₂: C 55.72, H 4.68, N 23.94; found: C 55.26; H 4.93; N, 23.49. IR (ATR, cm⁻¹): 3053(w), 2842(w), 1572(m), 1519(vs), 1487(s), 1464(s), 1382(s), 1339(m), 1305(vs), 1279(vs), 1220(vs), 1204(vs), 1130(vs), 1104(s), 1048(s), 1016(vs), 995(vs), 962(s), 916(s), 898(w), 870(m), 858(w), 802(vs), 785(s), 762(m), 743(vs), 723(vs), 693(vs), 669(m), 656(m), 642(s), 618(s). ¹H NMR (400 MHz, DMSO-d₆) δ 14.64 (s, 1H), 8.89 (s, 1H), 8.61 (s, 1H), 7.98 (s, 3H), 7.60 (s, 3H), 7.20 (s, 1H), 6.89 (s, 1H), 4.16 (s, 4H), 3.32 (s, 4H). ¹³C NMR (101 MHz, DMSO-d₆) δ 170.77, 148.40, 147.95, 139.47, 137.79, 137.30, 126.84, 124.64, 123.78, 108.38, 48.33, 47.45.

3.3. Synthesis and Characterization of [CuCl(L)]₂ (1)

Method 1: HL (100 mg, 0.244 mmol) was introduced in 20 mL MeOH to give a yellow suspension, and CuCl₂•2H₂O (42 mg, 0.246 mmol) in 10 mL MeOH was subsequently introduced dropwise. The formed mixture was stirred at 40°C for 8 h. The solvent was removed, and the formed brown powder was recrystallized in MeOH and Et₂O to obtain the crystal of 1.

Method 2: CuO (21.4 mg, 0.269 mmol) and HCl (19.6 mg, 0.537 mmol) were mixed in 3 mL of MeOH and the mixture stirred at 40°C. With the gradual dissolution of CuO, the solution turned pale green. After 10 min, HL (100 mg, 0.244 mmol) was introduced and the solution became dark brown. The mixture was stirred for an additional 2 h and then filtered, washed with a mixture of MeOH/MTBE (1 : 1, v/v; MTBE = methyl tert-butyl ether) to give a brown powder of [CuCl(L)]₂ which was dried under vacuo. Yield (220 mg, 89% based on HL). Anal. Calcd for C₃₈H₄₀Cl₂Cu₂N₁₄S₄: C 44.79, H 3.96, N 19.24; found: C 44.18, H 3.37, N 18.96. IR (ATR, cm⁻¹): 3077(w), 2993(w), 2645(w), 1592(s), 1563(m), 1476(w), 1458(w), 1440(vs), 1418(s), 1383(s), 1364(vs), 1318(w), 1289(m), 1275(s), 1247(vs), 1193(m), 1169(w), 1130(s), 1089(w), 1058(w), 1012(s), 982(w), 966(w), 956(w), 923(s), 850(m), 812(m), 795(s), 772(w), 738(m), 709(m), 686(w), 664(s), 644(s), 612(m).

3.4. Synthesis and Characterization of [Cu(NO₃)(L)]₂ (2)

Cu(NO₃)₂ (50.5 mg, 0.269 mmol) and HL (100 mg, 0.244 mmol) were mixed in 3 mL MeOH and the mixture was stirred at 40°C to obtain a dark brown solution, and upon further stirring for 2 h, the solution turned dark green. The mixture was filtered and washed with a mixture of MeOH/MTBE (1 : 1, v/v; MTBE = methyl tert-butyl ether) to give a brown powder of 2 which was dried under vacuo. Single crystals were obtained by slow diffusion of Et₂O into a MeOH solution of 2. Yield (235 mg, 90% based on HL). Anal. Calcd for C₃₈H₃₆Cu₂N₁₆O₆S₄: C 42.73, H 3.40, N 20.98; found: C 42.38, H 3.33, N 20.81. IR (ATR, cm⁻¹): 3080(w), 2846(w), 1592(m), 1520(s), 1483(s), 1461(vs), 1424(vs), 1380(s), 1338(w), 1278(vs), 1239(vs), 1203(vs), 1140(vs), 1102(m), 1052(w), 1012(vs), 920(s), 897(w), 855(m), 819(m), 792(s), 744(m), 724(w), 700(m), 665(m), 615(m).

3.5. Single Crystal X-ray Crystallography

Diffraction data for HL, 1, and 2 were acquired either on a Bruker APEX II CCD X-ray diffractometer (Bruker AXS GmbH, Germany) using Mo-Kα (λ = 0.71073 Å) (HL) or Ga-Kα (λ = 1.34138 Å) irradiation (1 and 2). Refinement and reduction of the collected data were achieved using the program Bruker SAINT and absorption corrections were performed using a multi-scan method [37]. All the crystal structures were solved by direct methods and refined on F² by full-matrix least-squares techniques with SHELXTL-2016 [38]. Crystallographic data for HL, 1, and 2 have been deposited in the Cambridge Crystallographic Data Center (CCDC) as supplementary publication numbers 2341904 (HL), 2341905 (1), and 2341906 (2). These data can be obtained free of charge either from the CCDC via www.ccdc.cam.ac.uk/data_request/cif or from the Supporting Information. A summary of the key crystallographic data for HL, 1, and 2 is listed in Table 1. Selected bond distances and angles were listed in Table S1.

3.6. In Vitro Cytotoxicity Evaluation by CCK-8 Assay

The HuH-7 cell line was cultured in DMEM + 10% FBS + 1% P/S and DMEM + 1% P/S. Cells grew as a monolayer and were detached upon confluence using trypsin (0.5% w/v in PBS). The cells were harvested from the cell culture medium by incubating in trypsin solution for 3 min, and then centrifuged with the supernatant subsequently discarded. 3 mL portion of serum-supplemented cell culture medium was added to neutralize any residual trypsin. The cells were re-suspended in serum-supplemented DMEM at a concentration of 5 × 10⁴ cells/mL. Cells were cultured at 37°C and 5% CO₂ for the CCK-8 studies. HuH-7 cells were seeded at a density of 2 × 10⁴ cells per well in 90 µL of culture medium (DMEM + 10% FBS + 1% P/S), and cultured for 24 h at 37°C and 5% CO₂ for attachment. The culture medium was then replaced by a serum-free medium (DMEM + 1% P/S) containing various concentrations of HL, 1, and 2 (Pharmaceuticals solubilized with 2 parts per thousand DMSO). All experiments were carried out with three replicates (n = 3), and the untreated cells served as the 100% cell viability control, while the cell-free medium (DMEM + 10% FBS + 1% P/S + CCK-8) served as the blank. HuH-7 cells were directly incubated for a period of 72 h. After incubation, 100 μL of culture medium (DMEM + 10% FBS + 1% P/S) and 10 μL CCK-8 was introduced and cultured for an additional 2 h before spectrophotometric measurement at 450 nm on a microplate reader. The relative cell viability (%) related to control cells was calculated using the equation:

$$V\%={\displaystyle \frac{{\left[A\right]}_{experimental}-{\left[A\right]}_{blank}}{{\left[A\right]}_{control}-{\left[A\right]}_{blank}}}\times 100\%$$

in which V% is the percentage of cell viability, [A]_experimental is the absorbance of the wells culturing the treated cells, [A]_blank is the absorbance of the blank, and [A]_control is the absorbance of the wells culturing untreated cells. The cytotoxicity assessment for Hep-G2 and PLC/PRF/5 resembled that of HuH-7 except that the culturing media was different, viz MEM (NEAA) + 10% FBS + 1% P/S and MEM (NEAA) + 1% P/S for both Hep-G2 and PLC/PRF/5.

3.7. Detection of Intracellular Reactive Oxygen Species

HuH-7 cells were introduced into a 12-well plate (4 × 10⁵ per well) in 1 mL of growing media (DMEM + 10% FBS + 1% P/S) and the cells were incubated at 37°C under 5% CO₂ for 24 h for attachment. The growing media was removed and the wells were replenished with HL, 1, and 2 in 1 mL growing media (with 2 ppm of DMSO) with equivalent HL concentrations of 2 μg/mL. The cells were incubated for an additional 2 h and then the culturing media was replaced with 10 μmol/mL DCFH-DA in DMEM + 1% P/S (1 mL), prepared by diluting (1 : 1000) a concentrated DCFH-DA following the supplied protocols. The cells were incubated for a further 20 min and then the culturing media was removed. The cells were further rinsed with culturing media (DMEM + 1% P/S) three times to remove the residual DCFH-DA, and observed under a BD5000 inverted microscope to estimate and compare the ROS generation abilities of these materials.

3.8. Pharmacokinetics of HL

CD^®(SD) IGS Rats (5−6 weeks old) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. All animal experiments were used in accordance with the requirements of the Experimental Animal Welfare Ethics Committee of Transcenta Diagnostic Technology (Suzhou) Co., LTD (PZ-20211210001). For pharmacokinetic experiments, SD rats were randomly divided into two groups, with 3 for PO administration (30 mg kg⁻¹) and i.v. administration (0.5 mg kg⁻¹), respectively. Plasma was collected from the jugular vein with the presence of a heparinized syringe at different time intervals, viz. 0.5, 1.0, 2.0, 4.0, 8.0, and 24 h for PO, and 0.03, 0.5, 1.0, 2.0, 4.0, 8.0, and 24 h for i.v. Plasma was first centrifuged at 3500 rpm for 10 min, then 0.05mL of the plasma was extracted, then trifluoroacetic acid was used to acid precipitate protein, and NaOH to neutralize the solution. The mixture was diluted using MeCN: H₂O (5 : 5 v/v). After 10 min of precipitating, supernatant fluids were collected by centrifugation at 10000 rpm for 5 min and filtered with a syringe through a 0.22 μm hydrophilic membrane filter and measured by HPLC method. For the HPLC assay, the analytical column was an Agilent ZORBAX SB C₁₈ column (4.6 mm × 150 mm, 5 μm. The mobile phase was MeCN: H₂O (5:5 v/v), the flow rate of the mobile phase was 0.8 mL min⁻¹ and the UV detector was at 245 nm.

4. Conclusions

In this work, we report that the thiosemicarbazone derivative of HL and its Cu(II) complexes 1 and 2 demonstrated superior anticancer performances against hepatocellular carcinoma cell lines Hep-G2, PLC/PRF/5, and HuH-7. For 1 and 2, the involvement of ROS as catalyzed by the Fenton-like process is also obvious. Intriguingly, the pharmacokinetic studies revealed that HL can be successfully absorbed via PO administration, with a favorable half-life of 21.61 ± 9.4 h, almost double that of via i.v. administration, rendering HL promising clinical potential. We believe that the association of such type of ligand with diverse metal ions may further equip the drug with additional beneficial features, such as the modulation of acute toxicity, toxicity via ROS, and on-demand metal supplements during the curation process.