Platinum(IV) Prodrugs Incorporating an Indole-Based Derivative, 5-Benzyloxyindole-3-Acetic Acid in the Axial Position Exhibit Prominent Anticancer Activity

Kinetically inert platinum(IV) complexes are a chemical strategy to overcome the impediments of standard platinum(II) antineoplastic drugs like cisplatin, oxaliplatin and carboplatin. In this study, we reported the syntheses and structural characterisation of three platinum(IV) complexes that incorporate 5-benzyloxyindole-3-acetic acid, a bioactive ligand that integrates an indole pharmacophore. The purity and chemical structures of the resultant complexes, P-5B3A, 5-5B3A and 56-5B3A were confirmed via spectroscopic means. The complexes were evaluated for anticancer activity against multiple human cell lines. All complexes proved to be considerably more active than cisplatin, oxaliplatin and carboplatin in most cell lines tested. Remarkably, 56-5B3A demonstrated the greatest anticancer activity, displaying GI50 values between 1.2 and 150 nM. Enhanced production of reactive oxygen species paired with the decline in mitochondrial activity as well as inhibition of histone deacetylase were also demonstrated by the complexes in HT29 colon cells.


Introduction
The classical cytotoxic platinum(II) drug, cis-diammineplatinum(II) dichloride (cisplatin) and its direct analogues, trans-L(1R,2R-diaminocyclohexane) oxalatoplatinum(II) (oxaliplatin) and cis-diammine(1,1-cyclobutanecarboxylato)platinum(II) (carboplatin) (Figure 1), are a class of square planar platinum(II) anticancer drugs, which have irrevocably evolved as working pillars of chemotherapy [1][2][3][4][5].The potency of platinum(II) drugs is demonstrated through their ability to induce deoxyribonucleic acid (DNA) lesions by covalently binding to DNA [6].Consequently, DNA replication and transcription are effectively ceased, which induces apoptosis [7].It is also noteworthy that while DNA is the therapeutic target of these drugs, inevitably, they also exhibit non-specific binding with phospholipids, thiols, ribonucleic acid, and other biomolecules that contain nitrogen, sulfur, and oxygen atoms [8][9][10].Unfortunately, this singularity reflects the non-selectivity of the drugs resulting in decreased therapeutic efficacy and increased incidence of pernicious side effects, which patients commonly experience when receiving platinum-based regimens [11][12][13][14][15]. Furthermore, drug resistance is another clinical limitation associated with the use of platinum(II) drugs, which often develops when treatment is prolonged [16,17].Because of this, combinatorial treatment is employed, wherein therapeutic agents are administered together with platinum(II) drugs in an attempt to enhance treatment effectivity [18].However, this type of regimen can complicate patient care and interfere with therapeutic outcomes since each co-administered agent has a unique pharmacokinetic index [19,20].
Int. J. Mol.Sci.2024, 25, 2181 2 of 33 [16,17].Because of this, combinatorial treatment is employed, wherein therapeutic agents are administered together with platinum(II) drugs in an attempt to enhance treatment effectivity [18].However, this type of regimen can complicate patient care and interfere with therapeutic outcomes since each co-administered agent has a unique pharmacokinetic index [19,20].Cisplatin, oxaliplatin and carboplatin are effective anticancer agents.Regrettably, their disadvantageous traits proved to be burdensome to patients, and thus it is only ethical to develop safer chemotherapeutic alternatives that can improve cancer treatment experience.Many platinum-based complexes have been previously synthesised and investigated in the hope of finding another chemotherapeutic agent that is as potent as cisplatin, but without the toxic side effects, and has the potential to circumvent chemoresistance [21][22][23][24].Some examples include zeniplatin [25], spiroplatin [26], enloplatin [27], aroplatin [28] and miboplatin [29] (Figure 2).These third-generation platinum(II) complexes were in clinical trials, but were eventually withdrawn either because of demonstrated severe toxicity or limited antitumour activity.A collection of structurally related platinum(II) complexes have also been investigated and numerous studies have confirmed their impressive biological activity against multiple human cell lines [30][31][32][33][34][35].The chemical structures of these complexes are represented as [Pt II (HL)(AL)] 2+ , where HL is a bulky heterocyclic ligand (i.e., 1,10-phenanthroline (phen), 5-methyl-1,10-phenanthroline (5-Mephen) or 5,6-dimethyl-1,10-phenanthroline (5,6-Me2phen)) and AL is a chiral ancillary ligand (i.e., 1S,2S-diaminocyclohexane (SS-DACH or DACH)).[Pt II (phen)(SS-DACH)] 2+ (PHENSS), [Pt II (5-Mephen)(SS-DACH)] 2+ A collection of structurally related platinum(II) complexes have also been investigated and numerous studies have confirmed their impressive biological activity against multiple human cell lines [30][31][32][33][34][35].The chemical structures of these complexes are represented as [Pt II (H L )(A L )] 2+ , where H L is a bulky heterocyclic ligand (i.e., 1,10-phenanthroline (phen), 5-methyl-1,10-phenanthroline (5-Mephen) or 5,6-dimethyl-1,10-phenanthroline (5,6-Me 2 phen)) and A L is a chiral ancillary ligand (i.e., 1S,2S-diaminocyclohexane (SS-DACH or DACH)).[Pt II (phen)(SS-DACH)] 2+ (PHENSS), [Pt II (5-Mephen)(SS-DACH)] 2+ (5MESS) and [Pt II (5,6-Me 2 phen)(SS-DACH)] 2+ (56MESS) are three notable examples from this class (Figure 3).56MESS has proven to be the most potent and has been reported to promote epigenomic changes by interfering with the cytoskeleton, cell-cycle proteins, mitochondria, and nuclear DNA [36,37].Despite the remarkable in vitro activity of 56MESS, Researchers envisioned that designs based on kinetically inert octahedral platinum(IV) complexes would deliver safer, well-tolerated and effective anticancer treatment [41][42][43][44][45].It is generally accepted that platinum(IV) complexes are the likely chemotherapeutic replacements for their equivalent platinum(II) parent complexes because of their kinetic inertness and potential for bioactive functionalisation.Commonly, platinum(IV) complexes are prepared by oxidation of the platinum(II) precursors with hydrogen peroxide [46,47].This synthetic step generates the platinum(IV) scaffold, incorporating two axial positions with two hydroxides coordinated and readily available for substitution by bioactive ligands.While cisplatin, carboplatin and oxaliplatin are being exploited as precursor cores for new platinum(IV) complexes [48][49][50][51], we contribute to this approach by utilising the potent platinum(II) complexes, PHENSS, 5MESS and 56MESS.With the vast number of bioactive ligands available to coordinate to platinum(IV), the indole-based derivative, 5-benzyloxyindole-3-acetic acid (5B3A) caught our interest.We have selected 5B3A as axial ligand because it incorporates an indole pharmacophore, which has been used in drug design and discovery [52,53].The bicyclic nature of the compound exhibits multiple therapeutic properties [54,55]; and for this reason, it is found in many drugs, specifically those that target cancer [56].Recent studies have also shown the effective use Researchers envisioned that designs based on kinetically inert octahedral platinum(IV) complexes would deliver safer, well-tolerated and effective anticancer treatment [41][42][43][44][45].It is generally accepted that platinum(IV) complexes are the likely chemotherapeutic replacements for their equivalent platinum(II) parent complexes because of their kinetic inertness and potential for bioactive functionalisation.Commonly, platinum(IV) complexes are prepared by oxidation of the platinum(II) precursors with hydrogen peroxide [46,47].This synthetic step generates the platinum(IV) scaffold, incorporating two axial positions with two hydroxides coordinated and readily available for substitution by bioactive ligands.While cisplatin, carboplatin and oxaliplatin are being exploited as precursor cores for new platinum(IV) complexes [48][49][50][51], we contribute to this approach by utilising the potent platinum(II) complexes, PHENSS, 5MESS and 56MESS.With the vast number of bioactive ligands available to coordinate to platinum(IV), the indole-based derivative, 5-benzyloxyindole-3-acetic acid (5B3A) caught our interest.We have selected 5B3A as axial ligand because it incorporates an indole pharmacophore, which has been used in drug design and discovery [52,53].The bicyclic nature of the compound exhibits multiple therapeutic properties [54,55]; and for this reason, it is found in many drugs, specifically those that target cancer [56].Recent studies have also shown the effective use of the indole scaffold in creating new metal-based anticancer drugs with improved activity [57,58].We expect that by coordinating 5B3A to the platinum(IV) form of PHENSS, 5MESS and 56MESS, we will enhance the anticancer properties by providing an additional mode of action against cancer cells.
Herewith, we report the syntheses and structural characterisation of three platinum(IV) complexes of type [Pt IV (H L )(A L )(X)(OH)] 2+ , where X is the indole-based axial ligand, 5B3A: 4).The purity and chemical structures of the complexes were confirmed through high-performance liquid chromatography (HPLC), nuclear magnetic resonance ( 1 H-NMR; two-dimensional correlation spectroscopy (2D-COSY); heteronuclear multiple quantum correlation ( 1 H-195 Pt-HMQC)), ultraviolet-visible (UV), circular dichroism (CD) and elemental microanalysis experiments.All complexes were purified via a flash chromatography system.To determine the anticancer activity of the resultant complexes, they were screened against twelve human cell lines including HT29 colon, U87 glioblastoma, MCF-7 breast, A2780 ovarian, H460 lung, A431 skin, Du145 prostate, BE2-C neuroblastoma, SJ-G2 glioblastoma, MIA pancreas, the cisplatin-resistant ADDP ovarian variant, and the non-tumour-derived MCF10A breast line.In addition to solubility, stability and lipophilicity studies, the reduction behaviour of P-5B3A, 5-5B3A and 56-5B3A upon treatment with ascorbic acid (AsA) was also determined with the aid of 1 H-NMR and one-dimensional 195 Pt-NMR (1D-195 Pt-NMR).The capacity of the complexes to generate reactive oxygen species (ROS), alter mitochondrial activity via causing mitochondrial membrane potential (MtMP) changes, and inhibit histone deacetylase (HDAC) were also investigated to determine how these parameters may affect their overall biological activity.
of the indole scaffold in creating new metal-based anticancer drugs with improved activity [57,58].We expect that by coordinating 5B3A to the platinum(IV) form of PHENSS, 5MESS and 56MESS, we will enhance the anticancer properties by providing an additional mode of action against cancer cells.
Herewith, we report the syntheses and structural characterisation of three platinum(IV) complexes of type [Pt IV (HL)(AL)(X)(OH)] 2+ , where X is the indole-based axial ligand, 5B3A: 4).The purity and chemical structures of the complexes were confirmed through high-performance liquid chromatography (HPLC), nuclear magnetic resonance ( 1 H-NMR; two-dimensional correlation spectroscopy (2D-COSY); heteronuclear multiple quantum correlation ( 1 H-195 Pt-HMQC)), ultraviolet-visible (UV), circular dichroism (CD) and elemental microanalysis experiments.All complexes were purified via a flash chromatography system.To determine the anticancer activity of the resultant complexes, they were screened against twelve human cell lines including HT29 colon, U87 glioblastoma, MCF-7 breast, A2780 ovarian, H460 lung, A431 skin, Du145 prostate, BE2-C neuroblastoma, SJ-G2 glioblastoma, MIA pancreas, the cisplatin-resistant ADDP ovarian variant, and the non-tumour-derived MCF10A breast line.In addition to solubility, stability and lipophilicity studies, the reduction behaviour of P-5B3A, 5-5B3A and 56-5B3A upon treatment with ascorbic acid (AsA) was also determined with the aid of 1 H-NMR and onedimensional 195 Pt-NMR (1D-195 Pt-NMR).The capacity of the complexes to generate reactive oxygen species (ROS), alter mitochondrial activity via causing mitochondrial membrane potential (MtMP) changes, and inhibit histone deacetylase (HDAC) were also investigated to determine how these parameters may affect their overall biological activity.
As depicted in the 1 H-NMR spectrum of P-5B3A (Figure S9), the protons of the heterocyclic ligand, phen, resonated between 8 and 10 ppm.The most deshielded or downfield phen protons, H2 and H9, were assigned to the doublets at 9.40 and 9.38 ppm, which exhibit an equal J-coupling constant of 5.4 Hz.The doublet of doublets at 9.06 ppm was assigned to H4 and H7 protons, and both have J-coupling constants of 4.8 and 8.1 Hz, respectively.Subsequently, a multiplet was recorded at 8.32 ppm which was induced by  S11).The acquired NMR data are comparable to published data [40,[61][62][63].
As depicted in the 1 H-NMR spectrum of P-5B3A (Figure S9), the protons of the heterocyclic ligand, phen, resonated between 8 and 10 ppm.The most deshielded or downfield phen protons, H2 and H9, were assigned to the doublets at 9.40 and 9.38 ppm, which exhibit an equal J-coupling constant of 5.4 Hz.The doublet of doublets at 9.06 ppm was assigned to H4 and H7 protons, and both have J-coupling constants of 4.8 and 8.1 Hz, respectively.Subsequently, a multiplet was recorded at 8.32 ppm which was induced by H3 and H8 protons.The remaining protons in the phen ligand, H5 and H6, were assigned to the sharp singlet at 8.26 ppm.It is also noteworthy that the broad signals adjacent to the signals of H4, H7, H3, H8, H2 and H9 protons were caused by the amines of the chiral ancillary ligand, SS-DACH (Figure S9).
When comparing the multiplicities demonstrated for P-5B3A in the aromatic region with the multiplicities for its corresponding platinum(II) precursor and platinum(IV) scaffold, PHENSS and PHENSS(IV)(OH) 2 , slight variations are evident as shown in Figure 6.Upon coordinating 5B3A to the platinum(IV) scaffold, PHENSS(IV)(OH) 2 , an upfield shift was detected for the following phen protons: H4, H7, H3, H8, H5 and H6 (Figure 6).This phenomenon may have been caused by the possible π-π stacking interactions between the aromatic rings of 5B3A and the phen, and this upfield movement was also observed by structurally related complexes in the literature [59,61].Overall, the variations in multiplicity reported for P-5B3A is primarily influenced by the coordination of the axial ligand, 5B3A.
When comparing the multiplicities demonstrated for P-5B3A in the aromatic region with the multiplicities for its corresponding platinum(II) precursor and platinum(IV) scaffold, PHENSS and PHENSS(IV)(OH)2, slight variations are evident as shown in Figure 6.Upon coordinating 5B3A to the platinum(IV) scaffold, PHENSS(IV)(OH)2, an upfield shift was detected for the following phen protons: H4, H7, H3, H8, H5 and H6 (Figure 6).This phenomenon may have been caused by the possible π-π stacking interactions between the aromatic rings of 5B3A and the phen, and this upfield movement was also observed by structurally related complexes in the literature [59,61].Overall, the variations in multiplicity reported for P-5B3A is primarily influenced by the coordination of the axial ligand, 5B3A.Notable differences in chemical multiplicity were also demonstrated by the protons of 5B3A originating from the benzene and indole bicyclic rings (benzene fused with pyrrole) (Figures 6 and S9).The pyrrole proton (Hpyrrole) proved to be the most deshielded as it resonated furthest downfield as a singlet at 10.3 ppm, as shown in Figures 6 and S9.The distorted multiplicities between 7.37-7.53ppm were assigned to the benzene protons, a, b, c, d and e (Figures 6 and S9).The irregular doublet and doublet of doublets at 7.02 and 6.69 ppm were assigned to the indole protons, i and h, respectively.As for the remaining indole protons, f and g, a singlet at 6.46 ppm was recorded, which indicates overlapping.Of further note, the 2D-COSY of P-5B3A was also analysed to elucidate the assignment of the indole protons of 5B3A (Figure S12).As shown in Figure S12, f and h proved to be the only protons coupled to Hpyrrole.Proton h was found to be coupled to protons g and i (Figure S12).Furthermore, the methylene protons of 5B3A, β and α, was assigned to the singlet and the multiplet at 4.91 and 3.23 ppm, respectively (Figure S9).Within the aliphatic Notable differences in chemical multiplicity were also demonstrated by the protons of 5B3A originating from the benzene and indole bicyclic rings (benzene fused with pyrrole) (Figures 6 and S9).The pyrrole proton (H pyrrole ) proved to be the most deshielded as it resonated furthest downfield as a singlet at 10.3 ppm, as shown in Figures 6 and S9.The distorted multiplicities between 7.37-7.53ppm were assigned to the benzene protons, a, b, c, d and e (Figures 6 and S9).The irregular doublet and doublet of doublets at 7.02 and 6.69 ppm were assigned to the indole protons, i and h, respectively.As for the remaining indole protons, f and g, a singlet at 6.46 ppm was recorded, which indicates overlapping.Of further note, the 2D-COSY of P-5B3A was also analysed to elucidate the assignment of the indole protons of 5B3A (Figure S12).As shown in Figure S12, f and h proved to be the only protons coupled to H pyrrole .Proton h was found to be coupled to protons g and i (Figure S12).Furthermore, the methylene protons of 5B3A, β and α, was assigned to the singlet and the multiplet at 4.91 and 3.23 ppm, respectively (Figure S9).Within the aliphatic region (1-3 ppm), the recorded resonances exhibited by the protons of the SS-DACH (Figure S9) are also consistent with literature data [40,63,64].[40,59,63].The 1 H-195 Pt-HMQC spectrum of P-5B3A highlights the correlation of protons H2 and H9 (9.12 ppm) and of H3 and H8 (8.10 ppm) with the platinum peaks at 498 ppm (Figure S15).No peak was found at −2800 ppm (Figure S16), which confirmed the absence of platinum(II) species.This further supports that not only the sample used for the NMR experiments are free of platinum(II) species denoting purity, the 5B3A ligand was also successfully coordinated to PHENSS(IV)(OH) 2 and occupied one axial position.
The 1 H-NMR assignment of multiplicities for the spectra of 5-5B3A and 56-5B3A (Figures S10 and S11) followed the same rationale to that discussed for P-5B3A.With respect to the 1 H-NMR spectra of 5-5B3A and 56-5B3A (Figures S10 and S11), some distinctions in multiplicity were evident when compared to those observed for P-5B3A, most specifically in the aromatic region.While these distinctions are mostly attributed to the coordination 5B3A ligand, another factor that contributes to these is the methylation of the heterocyclic rings of 5-5B3A and 56-5B3A.This phenomenon has been observed previously with similar complexes in the literature [59,61,65].Finally, the 1 H-195 Pt-HMQC spectra of 5-5B3A and 56-5B3A (Figures S17-S20) also confirmed the successful coordination of 5B3A to the respective platinum(IV) scaffolds, indicating the axial ligand only occupying one axial position.
In addition to NMR, elemental microanalysis was performed on an Elementar Vario MICRO (Elementar Analysensysteme GmbH) to further confirm the chemical composition and purity of P-5B3A, 5-5B3A and 56-5B3A.The resulting elemental microanalysis is consistent with the calculated values for the prodrugs as nitrates with co-crystallised H 2 O molecules, as summarised in the Materials and Methods section for P-5B3A.2H 2 O, 5-5B3A.3H 2 O and 56-5B3A.2H 2 O.

UV and CD Spectral Analysis
The molecular electronic transitions exhibited by P-5B3A, 5-5B3A and 56-5B3A in the UV measurements are comparable to published data (Figures S21-S23) [40,61,62,66].The acquisition of UV spectra was achieved as previously described [59,61,65,67].Experiments were obtained in triplicates; three separate stock solutions were prepared for each complex.All spectra were baseline corrected.Mean extinction coefficients (ε) were determined with standard deviation and errors based on the constructed plot curves.CD measurements were also undertaken to confirm that the chirality of the ancillary ligand, SS-DACH, was retained by P-5B3A, 5-5B3A and 56-5B3A during syntheses.Furthermore, the absorption bands exhibited in the CD spectra of P-5B3A, 5-5B3A and 56-5B3A (Figures S24-S26) are also comparable to published data [40,61,62,66].Due to the structural similarities of the complexes, no significant differences were observed in their UV or CD spectra.All characteristic peaks exhibited by P-5B3A, 5-5B3A and 56-5B3A in their UV and CD spectra are summarised in Table 1.
Table 1.A summary of the peak maxima in the UV and CD spectra of P-5B3A, 5-5B3A and 56-5B3A.

Platinum(IV) Complexes
Prominent peaks were detected at the wavelengths ~200 and ~279-289 nm for all complexes, as shown in Figure 7.
The anticipated metal-to-ligand charge transfer interactions and π-π* transitions were exhibited in the UV spectra of the complexes (Figures S21-S23), and are consistent with similar complexes in the literature [40,61,62,66].These interactions are the effect of the heterocyclic ligands of the complexes, phen, 5-Mephen and 5,6-Me2phen which exhibit ligand-centred π-π* transitions.The patterns observed in the UV spectra of P-5B3A, 5-5B3A and 56-5B3A (Figures 7 and S21-S23), especially the bathochromic shift, are heavily influenced by their heterocyclic ligands.This observation corresponds to published data [40,61,62,66].Prominent peaks were detected at the wavelengths ~200 and ~279-289 nm for all complexes, as shown in Figure 7. Based on the UV spectra of P-5B3A (Figures 7 and S21), a prominent absorption band was detected at 204 nm, which was followed by a dip between ~240 and 250 nm, denoting a hypochromic shift.Subsequently, a hyperchromic shift was observed when a less Based on the UV spectra of P-5B3A (Figures 7 and S21), a prominent absorption band was detected at 204 nm, which was followed by a dip between ~240 and 250 nm, denoting a hypochromic shift.Subsequently, a hyperchromic shift was observed when a less prominent peak appeared at 279 nm, which was followed by a weak band between 300-310 nm, as shown in Figure 7.The absorption bands detected for P-5B3A were also evident in the UV spectra of 5-5B3A and 56-5B3A (Figures 7, S22 and S23).Moreover, the absorptions bands shifted towards a longer wavelength, which indicates bathochromic shift.Interestingly, a characteristic band was observed for 56-5B3A between ~240 and 250 nm which appears like a weak shoulder.Finally, it is also noteworthy that the characteristic absorption bands for uncoordinated 5B3A are evident within the absorption bands for the complexes (Figure 7).This provides further evidence of the contributions of 5B3A when coordinated to platinum.Overall, the UV spectra demonstrated by P-5B3A, 5-5B3A, 56-5B3A also show evidence of the presence or lack of methyl groups on the heterocyclic ligands.
CD measurements were also performed to confirm that the chirality of the chiral ancillary ligand, SS-DACH was retained during syntheses.Chirality is an essential feature that directly contributes to the potency of the studied complexes, and this has been the subject of previous studies [33][34][35]68].The CD spectra of P-5B3A, 5-5B3A and 56-5B3A (Figures S24-S26), demonstrated no significant differences.Strong peaks were evident at lower wavelengths, which is consistent with previously reported structurally similar platinum(IV) complexes [40,61,62,66].The CD spectra of P-5B3A (Figure S24), exhibited prominent negative absorption bands (200-240 nm), followed by weaker positive absorption bands between 250 and 280 nm.This pattern was also observed in the CD spectra of 5-5B3A (Figure S25) and 56-5B3A (Figure S26), although 56-5B3A exhibited slightly more prominent positive absorption bands between 250 and 280 nm.

Physicochemical Investigations 2.2.1. Lipophilicity
The order of complexes by increasing lipophilicity is P-5B3A < 5-5B3A < 56-5B3A (Table 2 and Figure S27).It can be deduced that increasing methylation on the heterocyclic ligand increased lipophilicity.This trend agrees with published data [59,61,65,66,69].The determined log k w values are consistent and within the range of the log k w values determined for similar complexes reported in the literature [59,61].Furthermore, cisplatin has a reported log k w value of −2.19 [70,71], which is significantly lower than the acquired log k w values of P-5B3A, 5-5B3A and 56-5B3A (Table 2).This suggests that P-5B3A, 5-5B3A and 56-5B3A are expected to have better absorption properties than cisplatin.  -1 Pt-NMR spectroscopy were utilized to investigate the reduction activity of the platinum(IV) complexes, P-5B3A, 5-5B3A and 56-5B3A, using previously reported methods [59,61,65].The biological reductant selected for this study is AsA.Briefly, sequential 1 H-NMR experiments were performed for 1 h at 37 • C, proceeded by 1D- 195 Pt-NMR at 400 and −2800 ppm (30 min per region).It is noteworthy that the purity of each complex was examined prior the reduction experiments ensuring that no traces of platinum(II) species were evident in the samples used.All complexes were dissolved in PBS and DMSO-d 6 , without the reductant.Preliminary 1D-195 Pt-NMR measurements were conducted at 37 • C at 400 and −2800 ppm.This was completed to ensure that the PBS and DMSO-d 6 do not interfere with the complexes by causing unwanted reduction.From the preliminary 1D-195 Pt-NMR spectra, it was confirmed that the PBS and DMSO-d 6 did not cause the complexes to reduce (Figures S28-S30).Upon treatment of the platinum(IV) complexes with AsA, there was considerable reduction observed in the 1 H-NMR and 1D-195 Pt-NMR spectra acquired (Figures S31-S36).Notable changes in chemical multiplicity were recorded in the 1 H-NMR reduction spectra of P-5B3A, 5-5B3A and 56-5B3A (Figures S31, S33 and S35), especially along the aromatic region.The estimated reduction time points for P-5B3A, 5-5B3A and 56-5B3A are also outlined in Table 3.The expanded 1 H-NMR reduction spectra for P-5B3A at different time intervals showing distinct changes in chemical multiplicity are highlighted in Figure 8.The separate doublets exhibited by H2 and H9 protons and the multiplet exhibited by H4 and H7 protons (in the platinum(IV) form) resolved into a doublet (Figure 8).Subsequently, an upfield shift was observed for H3, H8, H5 and H6 protons.Notably, the H pyrrole of the axial ligand, 5B3A appeared further downfield at 10.8 ppm upon dissociating from platinum.Another notable change in multiplicity was demonstrated by the indole protons, f and g.As shown in Figure 8, these protons resonated as a singlet at 6.46 ppm (in the platinum(IV) form), but eventually resolved into two separate singlets as it dissociated from the platinum.These changes in multiplicity show that the platinum(IV) complex, P-5B3A successfully reduced to its core platinum(II) precursor, PHENSS.To further confirm this, 1D-195 Pt-NMR measurements were performed at 37 • C at 400 and −2800 ppm immediately after the final 1 H-NMR measurement.At approximately 1.5 h, the 1D-195 Pt-NMR reduction spectra of P-5B3A was acquired and no platinum(IV) peak was found at 400 ppm, as shown in Figure S32.However, at −2800 ppm, 30 min after the first 1D-195 Pt-NMR reduction spectra were acquired, a prominent peak was recorded, which indicates the presence of the platinum(II) precursor, PHENSS (Figure S32).The changes in multiplicity observed in the 1 H-NMR reduction spectra of P-5B3A (Figures 8 and S31) were comparable to those observed for 5-5B3A and 56-5B3A (Figures S33 and S35).5-5B3A and 56-5B3A were also confirmed to have successfully reduced to their core platinum(II) precursors, 5MESS and 56MESS, respectively, based on the acquired 1D-195 Pt-NMR reduction spectra (Figures S34 and S36).Because P-5B3A, 5-5B3A and 56-5B3A are reduced to their core platinum(II) precursors, this further confirms that they are prodrugs.T is time in min.The expanded 1 H-NMR spectrum of the untreated complex (control) was also included for comparison.

Stability Measurements
The stability of P-5B3A, 5-5B3A and 56-5B3A was monitored via analytical HPLC.Surprisingly, P-5B3A and 5-5B3A did not show any signs of significant hydrolysis for up to 1 week according to the HPLC chromatograms obtained (Figures S37 and S38), only minimal peak traces at ~5 min denoting the presence of the equivalent platinum(II) precursors.While 56-5B3A did not completely hydrolyse in solution, a trace platinum(II) peak was also recorded at ~6 min which gradually became more prominent over the week (Figures S37 and S38).Overall, the platinum(IV) complexes, P-5B3A, 5-5B3A and 56-5B3A proved to be adequately stable in aqueous solution for up to 1 week at 37 °C without any significant hydrolysis involved.

In Vitro Anticancer Activity
The in vitro anticancer activity of the synthesised platinum(IV) complexes, P-5B3A, With respect to the estimated reduction time points in Table 3, there was a notable difference in the rate of reduction exhibited by the complexes (Table 3).P-5B3A was the only complex to have reduced more slowly, with 50% reduction occurring between 10 and 15 min and maximal reduction at 60 min (Figures 8 and S31).In comparison, 5-5B3A reduced by 50% between 5 and 10 min, while maximal reduction occurred at 30 min (Figure S33).Interestingly, 56-5B3A was confirmed to have the fastest rate of reduction with maximal reduction at 10 min (Figure S36).This indicates that the differences in reduction rates are primarily regulated by the heterocyclic ligands rather than the coordinated 5B3A ligand.This trend has also been observed previously with similar complexes in the literature [59,61].These preliminary results provided insight into how these complexes reduce in the presence of a reducing agent like AsA.We expect the complexes to reduce differently under physiological conditions (i.e., in the presence of human serum or plasma).

Stability Measurements
The stability of P-5B3A, 5-5B3A and 56-5B3A was monitored via analytical HPLC.Surprisingly, P-5B3A and 5-5B3A did not show any signs of significant hydrolysis for up to 1 week according to the HPLC chromatograms obtained (Figures S37 and S38), only minimal peak traces at ~5 min denoting the presence of the equivalent platinum(II) precursors.While 56-5B3A did not completely hydrolyse in solution, a trace platinum(II) peak was also recorded at ~6 min which gradually became more prominent over the week (Figures S37 and S38).Overall, the platinum(IV) complexes, P-5B3A, 5-5B3A and 56-5B3A proved to be adequately stable in aqueous solution for up to 1 week at 37 • C without any significant hydrolysis involved.
Table 4.The in vitro cytotoxicity values of 5B3A, P-5B3A, 5-5B3A and 56-5B3A in multiple human cell lines, expressed as GI 50 (nM).For comparison, the GI 50 values of cisplatin, oxaliplatin and carboplatin were also included [59,61,65].Significant increase in potency was identified by a paired T-test with a two-tailed distribution (n = 3).Group differences were considered statistically significant if * p < 0.05, ** p < 0.01, *** p < 0.001 and **** p < 0.0001 compared with cisplatin.P-5B3A, 5-5B3A and 56-5B3A proved to be potent in most cell lines (Table 4), although P-5B3A was least active in the BE2-C neuroblastoma cell line, with a GI 50 value of 13,000 ± 5000 nM.5-5B3A and 56-5B3A were found to be more active than cisplatin, oxaliplatin and carboplatin in all the cell lines, as shown in Table 4.The average GI 50 values of the complexes were calculated from all cell lines and with the order of increasing cytotoxicity: P-5B3A < 5-5B3A < 56-5B3A.The complexes incorporating the methylated heterocyclic ligands, 5-5B3A and 56-5B3A demonstrated greater anticancer potential compared to P-5B3A.This trend in cytotoxicity was anticipated, as this has been previously observed with the equivalent platinum(II) precursors (PHENSS, 5MESS and 56MESS) and platinum(IV) scaffolds (PHENSS(IV)(OH) 2 , 5MESS(IV)(OH) 2 and 56MESS(IV)(OH) 2 ) (Table 5), as well as other similar complexes in the literature [31,38,59,61,62,65,66,69,73].These results corroborate the notion that increasing methylation of the heterocyclic ligand enhances the biological activity of the complexes.
The selectivity cytotoxicity index (SCI) of the synthesised platinum(IV) complexes, P-5B3A, 5-5B3A and 56-5B3A, and cisplatin was also calculated by dividing their GI 50 in the normal breast cancer cell line (MCF10A) to their GI 50 in all cancer cell lines (Table 6).Generally, a higher SCI corresponds to improved selectivity towards cancer cells [74,75].The complexes demonstrated a degree of selectivity towards the HT29 colon, Du145 prostate and the cisplatin-resistant ADDP cell lines relative to cisplatin (Table 6).This also corresponds to the low GI 50 values elicited by the complexes in these cell lines (Table 4).In the HT29 colon cell line, P-5B3A and 5-5B3A had a calculated SCI of 1.40 and 1.06, respectively, while for 56-5B3A, the calculated SCI was 0.75.In the same cell line, cisplatin was less selective with a calculated SCI of 0.46.Moreover, while it was expected that cisplatin had a lower SCI (0.19) in the cisplatin-resistant ADDP cell line, the SCI of P-5B3A, 5-5B3A and 56-5B3A ranged between 0.55 and 0.85 (Table 6).Interestingly, cisplatin had a higher SCI (4.33) compared to the complexes (1.47-2.75) in the Du145 prostate cell line, as shown in Table 6.What is also noteworthy is that cisplatin had higher SCI in most of the reported cell lines (Table 6).
Despite the impressive in vitro cytotoxicity exhibited by P-5B3A, 5-5B3A and 56-5B3A in most cell lines, it is important to note that these complexes are not selective (Table 6).For example, in the MCF10A normal breast cell line, the complexes were significantly more toxic than in the MCF-7 cancerous breast cell line (Table 4).Additionally, the prodrugs, and their corresponding platinum(IV) scaffolds and platinum(II) precursors, exhibited decreased cytotoxicity in the BE2-C neuroblastoma and SJ-G2 glioblastoma cell lines compared to the other cell lines tested (Tables 4 and 5).This decreased cytotoxicity may be associated with efflux resistance mechanisms by the cell lines.P-glycoproteins are multidrug resistance proteins that are upregulated in the BE2-C neuroblastoma and SJ-G2 glioblastoma cell types [76], and these may prevent the adequate accumulation of the complexes in these cell lines.Nevertheless, the complexes demonstrated a degree of selectivity towards the HT29 colon, Du145 prostate and the cisplatin-resistant ADDP ovarian variant cell lines, which has been a consistent trend observed with similar complexes in the literature [59,61,65,66,69].Their capacity to overcome resistance was confirmed based on their responsiveness to the cisplatin-resistant ADDP ovarian variant cell line.The coordination of 5B3A to the platinum(IV) cores of PHENSS, 5MESS and 56MESS enhanced the biological activity of the resultant complexes (P-5B3A, 5-5B3A and 56-5B3A) in most cell lines tested (Tables 4 and 5).

Drug-Induced ROS Production as a Treatment Strategy
ROS embody a variety of chemically reactive free and non-free radicals which are byproducts of aerobic metabolism [77].These metabolic intermediates are mostly generated in the mitochondria [78,79].ROS are highly versatile for having important functions in cell signaling and other physiological cascades that contribute to maintaining cellular life [80,81].Additionally, ROS levels are tightly regulated and constantly maintained at basal levels in the normal cell environment [82,83].Regulation is critical to prevent excessive release of ROS from inducing oxidative stress, causing damage to cell membranes, proteins, ribonucleic acid, and DNA [84,85].Other sources that can also elevate levels of ROS are exogenous substances such as drugs, xenobiotics, and radiation pollutants [85].
Elevated ROS levels are disadvantageous for normal healthy cells because oxidative stress would transpire.In contrast, this is advantageous for cancer cells to an extent as cancer cells are equipped with greater levels of antioxidant enzymes to neutralise excessive amount of ROS during proliferation [86].Elevated levels of ROS can promote cancer development and progression.However, cancer cells also become vulnerable to different death-inducing pathways when there is a disproportional increase in ROS levels especially when exposed for prolonged periods [87].This can be achieved by introducing anticancer drugs.Most chemotherapeutics can increase ROS intracellularly, and this treatment approach has the potential to enhance cancer-killing effects [87][88][89][90].We hypothesised that P-5B3A, 5-5B3A and 56-5B3A would produce higher levels of ROS than their equivalent platinum(II) precursors (PHENSS, 5MESS and 56MESS) and platinum(IV) scaffolds (PHENSS(IV)(OH) 2 , 5MESS(IV)(OH) 2 and 56MESS(IV)(OH) 2 ) because of the coordinated 5B3A axial ligand.The indole core of 5B3A has garnered synthetic interest because of the numerous pharmacological potencies it offers [52,53,55], and one of them is the generation of ROS [91,92].
The ROS generation capacity of P-5B3A, 5-5B3A and 56-5B3A, 5B3A, and cisplatin was determined in HT29 colon cells at different time intervals: 24, 48 and 72 h (Table 7 and Figures 9, S39 and S40).All values are expressed as relative fluorescence units (RFU).With reference to the control, a greater RFU corresponds to elevated ROS production (increased fluorescence of DCF).Previously published RFU values of the platinum(II) precursors (PHENSS, 5MESS and 56MESS) and platinum(IV) scaffolds (PHENSS(IV)(OH) 2 , 5MESS(IV)(OH) 2 and 56MESS(IV)(OH) 2 ) were also included in Table 7 and Figure 9, for reference [59,61,65].The HT29 colon cell line was nominated for testing because P-5B3A, 5-5B3A and 56-5B3A displayed a degree of selectivity in this cell line (Tables 4 and 6).Relative to the control, the synthesised platinum(IV) complexes, P-5B3A, 5-5B3A and 56-5B3A caused a notable increase in ROS levels by 24 h, and progressively continued until completion of the experiments at 72 h in the HT29 colon cell line (Table 7 and Figures 9 and S39).P-5B3A and 5-5B3A displayed similar RFU values at 24 and 48 h.At 24 h, P-5B3A incurred a slightly lower RFU value (156) compared to 5-5B3A (160) (Table 7).Subsequently, at 48 h, the RFU value reported for P-5B3A increased to 214, which is slightly lower than the RFU value reported for 5-5B3A (230), as shown in Table 7.At 72 h, the RFU values reported for P-5B3A and 5-5B3A considerably increased to 315 and 474, respectively.The most potent platinum(IV) derivative, 56-5B3A caused the greatest ROS production compared to P-5B3A and 5-5B3A (Table 7 and Figures 9, S39 and S40 (compared to cisplatin)).At 24 and 48 h, 56-5B3A elicited RFU values of 274 and 418, respectively, which are greater than the RFU values reported for P-5B3A and 5-5B3A in that time period, as shown in Table 7. Peak ROS production was demonstrated at 72 h for 56-5B3A, with an RFU value of 474 (Table 7).The 5B3A ligand alone generated ROS (RFU values of 138, 153 and 164) (Table 7 and Figures 9, S39 and S40), but to a lesser extent when compared to P-5B3A, 5-5B3A and 56-5B3A.It is also noteworthy that there may be a link between ROS production at 72 h (Table 7 and Figures 9, S39 and S40) and in vitro cytotoxicity, in the HT29 colon cell line (Table 4).At 72 h, the least potent platinum(IV) derivative (P-5B3A) incurred an RFU value of 246, while the more potent platinum(IV) derivative (5-5B3A) incurred an RFU value of 315 (Table 7).Moreover, the most potent platinum(IV) derivative (56-5B3A) caused the greatest ROS production at 72 h with a reported RFU value of 474, as shown in Table 7.This trend was also observed with similar complexes in the literature [59,65].It is noteworthy that the GI 50 values for each complex were used for the ROS experiments, so the same concentration of each complex is not directly compared.With the exception of 56-5B3A, the reduced effect of the prodrugs, P-5B3A and 5-5B3A, as well as the platinum(II) precursors at 72 h may have been due to their cytostatic effects.Overall, the findings indicate that P-5B3A, 5-5B3A and 56-5B3A are effective at causing elevated levels of ROS for up to 72 h.While there were no significant differences observed between these platinum(IV) complexes cisplatin in terms of ROS production (Figure S43), the increase in ROS from 24, 48 to 72 h for 56-5B3A was sill greater than that of cisplatin.Nevertheless, through this mechanism, the reported platinum(IV) complexes are still expected to trigger oxidative stress inside cancer cells, hence effectively ceasing cancer development.Our findings are also comparable to the results obtained in a similar study undertaken by Tolan et al. [92], where they have demonstrated the coordination of a ROS-inducing indole-based ligand to the core of a platinum complex (cisplatin), produced prodrugs with enhanced anticancer killing effects via the effective production of ROS.The mitochondria are a key feature of in cell types and are primarily responsible for producing energy required for metabolism [93,94].While it is often described as a powerhouse, the mitochondria are also susceptible to oxidative stress [95,96].Several drug classes, including chemotherapy drugs have been identified to promote disturbance in mitochondrial functions [97,98].When this occurs, the mitochondrial membrane undergoes permeabilisation, which in turn creates an imbalance in MtMP [99,100].MtMP is a parameter that reflects mitochondrial homeostasis, clearly because it is involved in the energy storage process during oxidative phosphorylation [101,102].Fluctuations in MtMP are a consequence of normal physiological activity; however, prolonged fluctuations, either a significant increase or decrease in MtMP, prove to be deleterious to the mitochondria as this effect can trigger the release of apoptotic factors that lead to cell death [103][104][105].
Because the majority of ROS is generated in the mitochondria [78,106], it was appropriate to assess any mitochondrial activity changes upon treatment with the synthesised platinum(IV) complexes, P-5B3A, 5-5B3A and 56-5B3A.This was achieved by determining if the overproduction of ROS induced by these complexes impacts mitochondrial activity by altering MtMP.Briefly, there is evidence showing a strong correlation between increased levels of ROS and mitochondrial impairment [106][107][108][109][110]. We determined MtMP changes induced by P-5B3A, 5-5B3A and 56-5B3A, 5B3A, the platinum(II) precursors (PHENSS, 5MESS and 56MESS), the platinum(IV) scaffolds (PHENSS(IV)(OH) 2 , 5MESS(IV)(OH) 2 and 56MESS(IV)(OH) 2 ), and cisplatin in HT29 colon cells at different time intervals: 24, 48 and 72 h (Table 8 and Figures 10 and S41).The MtMP changes recorded are measured as RFU.Relative to the control, a lower or greater RFU is indicative of mitochondrial dysfunction caused by unstable MtMP.
In summary, the platinum(IV) complexes, P-5B3A, 5-5B3A, 56-5B3A and the 5B3A ligand, together with their equivalent platinum(II) precursors (PHENSS, 5MESS and 56MESS) and platinum(IV) scaffolds (PHENSS(IV)(OH) 2 , 5MESS(IV)(OH) 2 and 56MESS(IV)(OH) 2 ), and cisplatin caused a progressive decline in mitochondrial activity for up to 72 h, as reflected by the incurred RFU values (Table 8 and Figures 10 and S41).Interestingly, the pattern observed here may be linked to the elevated ROS levels detected for these complexes and the 5B3A ligand (Table 7 and Figures 9, S39 and S40).Our findings serve as another layer of evidence supporting the notion that elevated levels of ROS can promote mitochondrial impairment [106][107][108][109][110]. Furthermore, the RFU values reported for P-5B3A, 5-5B3A and 56-5B3A, and their equivalent platinum(II) precursors (PHENSS, 5MESS and 56MESS) were almost equivalent at 24, 48 and 72 h (Table 8 and Figures 10 and S41).This suggests that the coordinated 5B3A ligand did not play a significant role in altering MtMP.The changes in MtMP detected were predominantly caused by the core platinum(II) precursors (PHENSS, 5MESS and 56MESS) of the complexes.The results indicate that the platinum(II) precursors of the studied platinum(IV) prodrugs produced more pronounced impairment in the mitochondria.This phenomenon may be linked to the reduction potential of the prodrugs in the mitochondria and to their mechanism of action with a comparable delayed response to that of their platinum(II) precursors.A more pronounced effect of the prodrugs may be observed at a timeframe of 96 h, possibly reaching the same potential as their platinum(II) precursors in causing mitochondrial impairment.

Histone Deacetylase (HDAC) Inhibition Studies
Histone acetyltransferases (HATs) and histone deacetylases (HDACs) are critical transcriptional enzymes that play a role in regulating gene expression [111].Histones are chromatin-regulating proteins which serve as scaffolding that maintain the structural integrity of chromosomal DNA [112].HATs are responsible for adding acetyl (−COCH 3 ) to the highly charged lysine residues of histones [113].This process eliminates the charge of the lysine residues, which decreases the affinity between histones and the DNA.As a result, the chromatin conforms into a relaxed state and this promotes gene activation by enhancing transcription [114].Conversely, the primary function of HDACs is to remove acetyl from lysine residues of hyperacetylated histones and this forces the chromatin to condense, thus preventing further transcription [115].

Compounds and
Surprisingly, 5MESS and 56MESS displayed the greatest HDAC inhibition amongst all the platinum complexes in this study with reported RFU values of 2124 and 1847 (Table 9), respectively.It is not clear why the other platinum(II) derivative, PHENSS was not able to cause the same degree of HDAC inhibition.However, it may be implied that the methylation at the heterocyclic ligands of 5MESS and 56MESS may have aided the complexes to have better inhibitory response towards HDAC.We also believe that this phenomenon may also be related to the mode of action of the complexes.Although the mode of action of 5MESS has not been investigated before, we can only speculate that it parallels the proposed mode of action of 56MESS due to their structural similarity.56MESS is reported to interfere with cytoskeletal networks [36,37].Briefly, there are 18 HDACs present in humans [115], and one of them is a cytoplasmic enzyme called HDAC-6 that is responsible for regulating cytoskeletal assembly [136][137][138].HDAC-6 is overexpressed in HT29 colon cells [139,140].Based on this information, it may be concluded that the strong inhibitory response of 56MESS towards HDAC is a consequence of its ability to inhibit the cytoskeleton.Of further note, while it was anticipated that the platinum(IV) complexes, 5-5B3A and 56-5B3A would display better HDAC inhibition than their platinum(II) counterparts this was not reflected in the results.The weaker HDAC inhibitory response of 5-5B3A and 56-5B3A may be due to non-additive or antagonistic effect with the coordinated ligand, 5B3A.Peak HDAC inhibition was demonstrated by the potent HDAC inhibitor, TSA with a reported RFU value of 749 (Table 9 and Figures 11 and S42).The studied platinum(IV) complexes, P-5B3A, 5-5B3A and 56-5B3A, the 5B3A ligand, the equivalent platinum(II)
3.2.Chemistry 3.2.1.Synthesis Route of NHS-5B3A NHS-5B3A was prepared according to established protocols [60].5B3A was reacted with 1 mol eq. of NHS and DCC in C 3 H 6 O.The reaction solution was left stirring at room temperature for 24 h.The precipitated by-product, dicyclohexylurea (DCU), was removed through syringe filtration.The filtrate was collected and reduced to dryness through rotary evaporation.The NHS ester was used without further purification.
NHS-5B3A-Yield: 320 mg; 84%.2+ were synthesised according to established protocols [61].The appropriate precursor, [Pt IV (H L )(A L )(OH) 2 ] 2+ , was reacted with 2 mol eq. of NHS ester of 5B3A in DMSO (1-2 mL) for 72 h at room temperature, in the dark.The reaction solution was washed with excess Et 2 O and vigorously mixed using a plastic pipette, followed by centrifugation to afford a colourless supernatant and an oily brown layer.The colourless supernatant was discarded while the oily brown layer was collected and dissolved in methanol (1-2 mL), followed by the addition of excess Et 2 O to induce precipitation.Centrifugation was undertaken to collect the final precipitate.Excess C 3 H 6 O was mixed with the precipitate and sonicated, affording a pure and solidified yellow or green precipitate that was collected through centrifugation.

Laboratory Instrumentation
Equipment utilised to verify the purity and confirm the chemical structures of the studied platinum(IV) complexes including relevant protocols are outlined below.

Flash Chromatography
A Biotage Isolera TM One flash chromatography system (Shimadzu, Sydney, NSW, Australia) equipped with a Biotage ® Sfär C18 D (Duo 100 Å 30 µm 30 g) (Shimadzu, Sydney, NSW, Australia) was employed to purify the platinum(IV) complexes.The mobile phase consisted of solvents, A (d.i.H 2 O) and B (MeOH).The samples were dissolved in d.i.H 2 O/MeOH (50:50) and eluted through the column with a 0-40% linear gradient for 40 min with a flow rate of 4 mL.min−1 , collected within the set wavelengths of 200-400 nm.

High-Performance Liquid Chromatography (HPLC)
An Agilent (Melbourne, VIC, Australia) Technologies 1260 Infinity instrument equipped with a Phenomenex Onyx TM Monolithic C 18 reverse-phase column (100 × 4.6 mm, 5 µm pore size) (Sydney, NSW, Australia) was utilised for the complexes.The mobile phase consisted of solvents, A (0.06% TFA in d.i.H 2 O) and B (0.06% TFA in CH 3 CN/d.i.H 2 O (90:10)).An injection volume of 5 µL was utilised and eluted with a 0-100% linear gradient over 15 min with a flow rate of 1 mL.min −1 , at the set wavelengths of 214 and 254 nm.An Agilent ZORBAX RX-C 18 column (100 × 4.6 mm, 3.5 µm pore size) (Sydney, NSW, Australia) was utilised for the NHS ester of 5B3A using the same method described above.

Nuclear Magnetic Resonance (NMR) Spectroscopy
1 H-NMR, 2D-COSY, 1 H-195 Pt-HMQC and 1D-195 Pt-NMR were carried out on a 400 MHz Bruker (Melbourne, VIC, Australia) Avance spectrometer at 298 K.All complexes were prepared to a concentration of 10 mM in 600 µL using D 2 O. DMSO-d 6 was utilised for 5B3A, NHS ester of 5B3A, and the platinum(IV) complexes, P-5B3A, 5-5B3A and 56-5B3A. 1 H-NMR was set to 10 ppm and 16 scans with a spectral width of 8250 Hz and 65,536 data points.2D-COSY was acquired using a spectral width of 3443 Hz for both 1 H nucleus, F1 and F2 dimensions, with 256 and 2048 data points, respectively. 1H-195 Pt-HMQC was carried out using a spectral width of 214,436 Hz and 256 data points for 195 Pt nucleus, F1 dimension, also a spectral width of 4808 Hz with 2048 data points for 1 H nucleus, F2 dimension.1D-195 Pt was measured using a spectral width of 85,470 Hz and 674 data points.All resonance recorded, were presented as chemical shifts in parts per million (δ ppm) with J-coupling constants reported in Hz.For spin multiplicity: s (singlet); d (doublet); dd (doublet of doublets); t (triplet); q (quartet) and m (multiplet).All spectroscopic data gathered were generated and plotted using TopSpin 4.1.3analysis software.

Ultraviolet-Visible (UV) Spectroscopy
An Agilent (Melbourne, VIC, Australia) Technologies Cary 3500 UV-Vis Multicell Peltier spectrophotometer was utilised to perform the UV spectroscopy experiments.UV spectroscopic experiments were completed at room temperature in the range of 200-400 nm with a 1 cm quartz cuvette.All complexes were prepared in d.i.H 2 O, while 5B3A was prepared in CH 3 CN.A stock solution of each complex or compound (1 mM) was prepared, and absorption spectra were recorded at a series of different concentrations by titrating 9 × 3 µL aliquots into a cuvette containing d.i.H 2 O (3000 µL).Experiments were repeated in triplicate.All spectra were baseline corrected by the instrument-a baseline containing d.i.H 2 O was acquired first, and automatically subtracted from each experiment.Average extinction coefficients (ε) were determined with standard deviation and errors based on the generated plot curves.

Circular Dichroism (CD) Spectroscopy
A Jasco (Easton, PA, USA) J-810 CD spectropolarimeter was used to measure the CD spectra of the complexes.The samples were prepared in d.i.H 2 O, with similar concentrations (~0.05 mM) using a 1 mm optical glass cuvette or 1 cm quartz cuvette.CD experiments were undertaken at room temperature in the wavelength range of 200-400 nm (30 accumulations) with a bandwidth of 1 nm, data pitch of 0.5 nm, a response time of 1 s and a 100 nm.min −1 scan speed.The flowrate of nitrogen gas was 6 L.min −1 .The HT (photomultiplier) level remained below 500 V for all experiments.A CD simulation tool (CDToolX-Windows 10 Version) was used to process the spectra.

Elemental Microanalysis
Elemental analysis was performed on Elementar Vario MICRO (Elementar Analysensysteme GmbH, Frankfurt, Germany).The instrument hardware was configured for the analysis of 4 elements (C, H, N and S).An amount of 1.2 ± 0.2 mg of sample material was loaded into a tin foil crucible and combusted at 1150 • C with oxygen dosing time of 80 s and total O 2 flow rate of 30 mL/min.Ultra-high purity grade helium (BOC, 99.999%) and oxygen (BOC, 99.995%) were employed as working fluids in all cases.Pure sulfanilamide was employed as a standard with quality control samples added into the workflow every 10-20 runs.The follow-up data analysis was performed using a custom peak picking and integration algorithm written in Python 3.11.

Solubility
The solubility of P-5B3A, 5-5B3A and 56-5B3A was tested in d.i.H 2 O at room temperature.Small aliquots of d.i.H 2 O were titrated into an Eppendorf tube containing 1 mg of each complex until full dissolution.For every titration, each sample solution was vortexed and sonicated.

Stability
The stability of P-5B3A, 5-5B3A and 56-5B3A were measured at 254 nm.Each complex was incubated in aqueous phosphate-buffered saline (PBS) solution (pH ~7.4) (total sample concentration, ~10 mM) at 37 • C for one week.Analytical HPLC was utilised to monitor the experiments.

Drug-Induced Reactive Oxygen Species (ROS) Production
To evaluate the presence of ROS in treated cells, a DCFDA/H 2 DCFDA-cellular ROS Assay Kit (Abcam, Cambridge, MA, USA) was used, as previously described [61,141,142].A total of 25,000 cells/mL of HT29 cells in DMEM were seeded in 96-well plates.Cells were washed with 1X kit buffer, and then stained with 25 µM 2 ′ ,7 ′ -dichlorofluorescein diacetate (DCFH-DA) and incubated for 45 min.DCFH-DA was then removed, and cells were then re-washed with 1X kit buffer, after which phenol red free media was added.Cells were then treated with a GI 50 drug concentration for each complex.The plates were directly scanned to measure fluorescence (relative fluorescence units (RFU)) at the time points indicated the Glo-Max ® -Multimode microplate reader (Promega Corporation, Alexandra, VIC, Australia) at an excitation/emission of 485/535 nm, which is directly correlated to the DCFH-DA that is deacetylated by cellular esterases and then oxidised by ROS to the fluorescent compound DCF.To generate the positive control (20 µM tert-butyl hydroperoxide (TBHP)), cells were washed with 1X kit buffer, and stained with DCFDA (25 µM) for 45 min; this was removed and TBHP was added in phenol red free media, and the resulting solution was scanned as described above.

Mitochondrial Membrane Potential (MtMP) Changes
To investigate the MtMP changes in HT29 treated cells, a TMRE-MtMP Assay Kit (Abcam, Cambridge, MA, USA) was used, as previously described [36,61].In 96-well plates, 25,000 cells/mL of HT29 colon cells in DMEM were seeded.Cells were treated with a GI 50 drug concentration for each complex.At 24, 48 or 72 h, cells were washed with PBS and stained with tetramethylrhodamine, ethyl ester (TMRE) (1 µM) and incubated for 30 min.TMRE was then removed, and cells were then re-washed with PBS (0.2% BSA), after which phenol red free media was added.The plates were directly scanned to measure fluorescence (expressed as RFU), using the Glo-Max ® -Multimode microplate reader (Promega Corporation, Alexandra, VIC, Australia) at an excitation/emission of 549/575 nm.For the positive control, 20 µM carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP)) was added to the cells and incubated for 10 min.The cells were washed with PBS (0.2% BSA) and stained with TMRE (1 µM) for 30 min; this was removed, and phenol red free media was added, and the final solution was scanned as mentioned above.

Histone Deacetylase (HDAC) Inhibition Measurements
To study the HDAC activity in HT29 treated cells, a Fluorometric HDAC Activity Assay Kit ab156064 (Abcam, Cambridge, MA, USA) was used and all sample preparations were made based on the manufacturer's protocol.A total of 10 6 cells/mL of HT29 cells were seeded in DMEM with serum deprivation for 18 h to synchronise the cells.Then, cells were treated with GI 50 drug concentration of each complex or compound for 72 h.Cells were then washed with cold PBS and resuspended in lysis buffer (10 mM Tris HCl (pH 7.5), 10 mM NaCl, 15 mM MgCl 2 , 250 mM sucrose, 0.5% NP-40 and 0.1 mM EGTA) and lysed on ice for 20 min.Cells were centrifuged through 4 mL of sucrose cushion (30% sucrose, 10mM Tris HCl (pH 7.5), 10 mM NaCl and 3mM MgCl 2 ) at 1300× g for 10 min at 4 • C. Supernatant was discarded and the pellet was washed and resuspended in extraction buffer (50 mM HEPES-KOH (pH 7.5), 420 mM NaCl, 0.5 mM EDTA 2Na, 0.1 mM EGTA and 10% glycerol) and then sonicated for 30 s. Nuclei were lysed for 30 min on ice and centrifuged at 20,000× g for 30 min.The supernatant was collected, and the samples were prepared for the fluorometric assay according to the manufacturer's protocol (Histone Deacetylase (HDAC) Activity Assay Fluorometric Kit ab156064, Abcam).The changes in the relative fluorescence intensity (expressed as RFU) were recorded each minute over 60 min using the Glo-Max ® -Multimode microplate reader (Promega Corporation, Alexandra, VIC, Australia) at an excitation/emission of 355/460 nm.

Conclusions
Three platinum(IV) complexes incorporating the indole-based ligand, 5B3A (P-5B3A, 5-5B3A and 56-5B3A) were successfully synthesised in good yields.HPLC, NMR, UV and CD, as well as elemental microanalysis experiments confirmed the purity and chemical structures of the resultant complexes.The reduction behaviour of the complexes in PBS and AsA were monitored via NMR at 37 • C, with results indicating the ability to reduce to the equivalent platinum(II) precursors.From this, it is proposed that they can be defined as prodrugs.Additionally, the complexes were adequately stable in PBS for up to one week at 37 • C, suggesting their resistance towards significant hydrolysis in aqueous solutions.The rank of increasing lipophilicity of the complexes, as determined by analytical HPLC, was P-5B3A < 5-5B3A < 56-5B3A.
The MTT assay was used to measure the anticancer activity of P-5B3A, 5-5B3A and 56-5B3A.The findings indicated that the complexes were considerably more potent than cisplatin, oxaliplatin and carboplatin in most cell lines tested.This means that the coordination of the 5B3A ligand to our platinum(IV) scaffolds enhanced the biological activity of the resultant platinum(IV) complexes to an extent.The rank of increasing cytotoxicity was determined as P-5B3A < 5-5B3A < 56-5B3A.This also relates with the rank of increasing lipophilicity.Notably, the most lipophilic complex, 56-5B3A displayed the greatest growth inhibitory response in the Du145 prostate and HT29 colon cell lines with reported GI 50 values of 1.2 ± 0.6 nM and 4.4 ± 1.6 nM, respectively.Unfortunately, P-5B3A, 5-5B3A and 56-5B3A were found to be significantly toxic towards the MCF10A normal breast cell line, denoting their limited selectivity.While this is the case, it should be noted that this outcome may not be translated in vivo.Despite this, the complexes were still significantly more potent than cisplatin in the cisplatin-resistant ADDP ovarian variant cell lines, thus confirming their ability to overcome resistance mechanisms by cancer cells.
A sustained increase in ROS production was also observed in HT29 colon cells, upon treatment with P-5B3A, 5-5B3A and 56-5B3A for up to 72 h.The order by which the complexes displayed cumulative ROS production at 24, 48 and 72 h was determined to be P-5B3A < 5-5B3A < 56-5B3A.This trend in ROS production also relates to the trend in cytotoxicity, further confirming that the presence of the 5B3A ligand in the axial position aided the complexes to produce ROS.Moreover, the observed ROS production was also in agreement with the decrease in mitochondrial activity induced by the complexes.P-5B3A, 5-5B3A and 56-5B3A caused a progressive decrease in MtMP in HT29 colon cells for up to 72 h, which translates to mitochondrial dysfunction.These results strongly suggest that ROS production and mitochondrial dysfunction via decreasing MtMP may be additional cytotoxic elements that contribute to the overall potency of the complexes.
The HDAC inhibition of P-5B3A, 5-5B3A and 56-5B3A was also investigated in HT29 colon cells.We anticipated that the 5B3A ligand would enhance the inhibitory response of the resultant complexes towards HDAC, because of its indole pharmacophore which has been reported to inhibit HDAC.Unexpectedly, the results indicated otherwise.P-5B3A proved to be the least active against HDAC.While the other derivatives, 5-5B3A and 56-5B3A displayed a better inhibitory response against HDAC, the degree of inhibition was significantly inferior compared to TSA.Clearly, the 5B3A ligand did not enhance the activity of P-5B3A, 5-5B3A and 56-5B3A against HDAC.For the first time, we also reported the HDAC inhibition of their equivalent platinum(II) precursors (PHENSS, 5MESS and 56MESS) and platinum(IV) scaffolds (PHENSS(IV)(OH) 2 , 5MESS(IV)(OH) 2 and 56MESS(IV)(OH) 2 ) in the same cell line.Amongst the complexes, 5MESS and 56MESS proved somewhat active against HDAC.Particularly for 56MESS, we speculate that its ability to inhibit HDAC significantly is closely associated with one of its mode of actions, that is its interference with the cytoskeleton.

Figure 1 .
Figure 1.Molecular structures of standard platinum(II) drugs with worldwide approval for cancer treatment.

Figure 2 .
Figure 2. Molecular structures of third-generation platinum(II) complexes that were clinically investigated.

Figure 1 .
Figure 1.Molecular structures of standard platinum(II) drugs with worldwide approval for cancer treatment.

Figure 1 .
Figure 1.Molecular structures of standard platinum(II) drugs with worldwide approval for cancer treatment.

Figure 2 .
Figure 2. Molecular structures of third-generation platinum(II) complexes that were clinically investigated.

Figure 2 .
Figure 2. Molecular structures of third-generation platinum(II) complexes that were clinically investigated.

Figure 3 .
Figure 3. Molecular structures of unconventional platinum(II) complexes featuring a bulky heterocyclic ligand and a chiral ancillary ligand.

Figure 3 .
Figure 3. Molecular structures of unconventional platinum(II) complexes featuring a bulky heterocyclic ligand and a chiral ancillary ligand.

Figure 5 .
Figure 5. Synthetic scheme of P-5B3A, 5-5B3A and 56-5B3A.2.1.1.NMR Assignments NMR techniques including 1 H-NMR, 2D-COSY and 1 H-195 Pt-HMQC were applied to confirm the chemical structures of the synthesised platinum(IV) complexes, P-5B3A, 5-5B3A and 56-5B3A (Figures S9-S20).The 1 H-NMR and 1 H-195 Pt-HMQC assignments of P-5B3A, 5-5B3A and 56-5B3A are summarised in the Materials and Methods section.Because DMSO-d 6 was used for the NMR experiments, amine proton resonances were observed in the aromatic region (8-10 ppm) (Figures S9-S11).The acquired NMR data are comparable to published data[40,[61][62][63].As depicted in the 1 H-NMR spectrum of P-5B3A (FigureS9), the protons of the heterocyclic ligand, phen, resonated between 8 and 10 ppm.The most deshielded or downfield phen protons, H2 and H9, were assigned to the doublets at 9.40 and 9.38 ppm, which exhibit an equal J-coupling constant of 5.4 Hz.The doublet of doublets at 9.06 ppm was assigned to H4 and H7 protons, and both have J-coupling constants of 4.8 and 8.1 Hz, respectively.Subsequently, a multiplet was recorded at 8.32 ppm which was induced by H3 and H8 protons.The remaining protons in the phen ligand, H5 and H6, were assigned to the sharp singlet at 8.26 ppm.It is also noteworthy that the broad signals adjacent to the signals of H4, H7, H3, H8, H2 and H9 protons were caused by the amines of the chiral ancillary ligand, SS-DACH (FigureS9).When comparing the multiplicities demonstrated for P-5B3A in the aromatic region with the multiplicities for its corresponding platinum(II) precursor and platinum(IV) scaffold, PHENSS and PHENSS(IV)(OH) 2 , slight variations are evident as shown in Figure6.Upon coordinating 5B3A to the platinum(IV) scaffold, PHENSS(IV)(OH) 2 , an upfield shift was detected for the following phen protons: H4, H7, H3, H8, H5 and H6 (Figure6).This phenomenon may have been caused by the possible π-π stacking interactions between the

Figure 8 .
Figure 8.The expanded 1 H-NMR spectra of P-5B3A treated with PBS and AsA in DMSO-d6 at 37 °C, highlighting the variations in chemical multiplicity in the aromatic region depicted by the arrows.T is time in min.The expanded 1 H-NMR spectrum of the untreated complex (control) was also included for comparison.

Figure 8 .
Figure 8.The expanded 1 H-NMR spectra of P-5B3A treated with PBS and AsA in DMSO-d 6 at 37 • C, highlighting the variations in chemical multiplicity in the aromatic region depicted by the arrows.T is time in min.The expanded 1 H-NMR spectrum of the untreated complex (control) was also included for comparison.

8 *
Data are the mean ± SEM of three independent experiments, each performed in triplicate.

Table 2 .
Calculated log k w values of P

Table 3 .
The approximate time points in min for P
* Data are the mean ± SEM of three independent experiments, each performed in triplicate.
* Data are the mean ± SEM of three independent experiments, each performed in triplicate.