Latonduine-1-Amino-Hydantoin Hybrid, Triazole-Fused Latonduine Schiff Bases and Their Metal Complexes: Synthesis, X-ray and Electron Diffraction, Molecular Docking Studies and Antiproliferative Activity

: A series of latonduine derivatives, namely 11-nitro-indolo[2,3-d ]benzazepine-7-(1-amino-hydantoin) ( B ), triazole-fused indolo[2,3-d ]benzazepine-based Schiff bases HL 1 and HL 2 and metal complexes [M( p -cymene)(HL 1 )Cl]Cl, where M = Ru ( 1 ), Os ( 2 ), and [Cu(HL 2 )Cl 2 ] ( 3 ) were synthesized and characterized by spectroscopic techniques (UV–vis, 1 H, 13 C, 15 N– 1 H HSQC NMR) and ESI mass spectrometry. The molecular structures of B and HL 1 were conﬁrmed by single-crystal X-ray diffraction, while that of 3 by electron diffraction of nanometer size crystalline sample. Molecular docking calculations of species B in the binding pocket of PIM-1 enzyme revealed that the 1-amino-hydantoin moiety is not involved in any hydrogen-bonding interactions, even though a good accommodation of the host molecule in the ATP binding pocket of the enzyme was found. The antiproliferative activity of organic compounds B , HL 1 and HL 2 , as well as complexes 1 – 3 was investigated in lung adenocarcinoma A549, colon adenocarcinoma LS-174 and triple-negative breast adenocarcinoma MDA-MB-231 cells and normal human lung ﬁbroblast cells MRC-5 by MTT assays; then, the results are discussed.


Introduction
In our modern society cancer has become a major threat [1].Even though more and more people survive this severe disease in some particular cases, such as lung cancer, survival rate did not ameliorate much over the years.Cancer is still the second frequent cause for death after cardiovascular diseases [2].The Global Cancer Observatory estimates that the total number of incidences and mortality will rise in the nearest future [3].These predictions make the search for new anticancer drugs even more imperative.
Quite recently we turned our attention to indolo [2,3-d]benzazepine derivatives as isomers of Paullones.Intriguingly, their backbone is derived from the natural alkaloids of Latonduines (Figure 1), which do not show marked cytotoxicity [18,19].However, fusion of an additional indole moiety to the Latonduine skeleton led to an enormous enhancement of antiproliferative activity with IC 50 values in the low to sub-micromolar concentration range [20].The favorable effect of an indole core on the pharmacological profile of other drugs is well-documented in the literature [21][22][23].These indolo [2,3d]benzazepine molecules were discovered to act as microtubule destabilizing agents (MDA) fitting well into the colchicine binding pocket [24].One disadvantage for the development of these molecules as anticancer drugs is their poor aqueous solubility.As for Paullones this drawback could be overcome by their coordination to metals via creation of suitable binding sites attached to the main scaffold (see Figure 2) [4,5,7,8,25].Ru(II)-and Os(II)-arene, as well as Cu(II) complexes with these potential indolo [2,3-d]benzazepine ligands shown in Figure 2 were recently prepared and investigated for their antiproliferative activity [26][27][28].
Inorganics 2023, 11, x FOR PEER REVIEW 2 of 18 underlying the mechanism of antiproliferative activity of these compounds at the molecular level several cancer related kinases were reported, namely GSK3β, CDK2/cyclin A, Lck and Src [9][10][11][12][13][14][15][16][17].Quite recently we turned our attention to indolo [2,3-d]benzazepine derivatives as isomers of Paullones.Intriguingly, their backbone is derived from the natural alkaloids of Latonduines (Figure 1), which do not show marked cytotoxicity [18,19].However, fusion of an additional indole moiety to the Latonduine skeleton led to an enormous enhancement of antiproliferative activity with IC50 values in the low to sub-micromolar concentration range [20].The favorable effect of an indole core on the pharmacological profile of other drugs is well-documented in the literature [21][22][23].These indolo [2,3-d]benzazepine molecules were discovered to act as microtubule destabilizing agents (MDA) fitting well into the colchicine binding pocket [24].One disadvantage for the development of these molecules as anticancer drugs is their poor aqueous solubility.As for Paullones this drawback could be overcome by their coordination to metals via creation of suitable binding sites attached to the main scaffold (see Figure 2) [4,5,7,8,25].Ru(II)-and Os(II)-arene, as well as Cu(II) complexes with these potential indolo [2,3-d]benzazepine ligands shown in Figure 2 were recently prepared and investigated for their antiproliferative activity [26][27][28].Established structure-activity relationships (SARs) revealed very good antiprolifera- ular level several cancer related kinases were reported, namely GSK3β, CDK2/cyclin A, Lck and Src [9][10][11][12][13][14][15][16][17].Quite recently we turned our attention to indolo [2,3-d]benzazepine derivatives as isomers of Paullones.Intriguingly, their backbone is derived from the natural alkaloids of Latonduines (Figure 1), which do not show marked cytotoxicity [18,19].However, fusion of an additional indole moiety to the Latonduine skeleton led to an enormous enhancement of antiproliferative activity with IC50 values in the low to sub-micromolar concentration range [20].The favorable effect of an indole core on the pharmacological profile of other drugs is well-documented in the literature [21][22][23].These indolo [2,3-d]benzazepine molecules were discovered to act as microtubule destabilizing agents (MDA) fitting well into the colchicine binding pocket [24].One disadvantage for the development of these molecules as anticancer drugs is their poor aqueous solubility.As for Paullones this drawback could be overcome by their coordination to metals via creation of suitable binding sites attached to the main scaffold (see Figure 2) [4,5,7,8,25].Ru(II)-and Os(II)-arene, as well as Cu(II) complexes with these potential indolo [2,3-d]benzazepine ligands shown in Figure 2 were recently prepared and investigated for their antiproliferative activity [26][27][28].Established structure-activity relationships (SARs) revealed very good antiproliferative activity of the metal-free Latonduine derivatives modified at the lactam moiety [27,28], while those containing a metal-binding unit at position 11 (Figure 2) are markedly less active, but still with IC50 values in the micromolar concentration range [26].In both Established structure-activity relationships (SARs) revealed very good antiproliferative activity of the metal-free Latonduine derivatives modified at the lactam moiety [27,28], while those containing a metal-binding unit at position 11 (Figure 2) are markedly less active, but still with IC 50 values in the micromolar concentration range [26].In both cases metal complex formation with the respective transition metals enhanced the pharmacological profile, however, distinct differences could still be observed.For Cu(II) complexes the aqueous solubility only slightly increased, while cytotoxicity of lead compounds reached the low nanomolar concentration range [27,28].Cationic Ru(II) and Os(II) complexes became soluble in biocompatible media, but IC 50 values did not ameliorate when compared to those of metal-free derivatives [26].Taking into account the established SARs, and, in particular, that the increase in the cytotoxicity of indolo [2,3-d]benzazepines is observed when structural modifications are performed at the lactam moiety of the backbone (compare with [26,27]), as well as the results reported by Anand et al. [29] on the synthesis of a large library of MST1 kinase inhibitors via coupling of inert organoruthenium complexes with bioactive fragments, we selected one of them, namely, 1-amino-hydantoin (1-AH, see Figure 3 highlighted in blue) as a unit to be attached to our scaffold at the lactam group (see Figure 3).We expected to increase the docking ability of the new hybrid to ATP binding sites of cancer related mammalian kinases, and eventually enhance their antiproliferative activity [29].
fects of hydantoin on antimitotic and antiviral activities of inactive amino acids were reported [31].In addition, this moiety has been proven to have an immunomodulating effect antagonizing carcinogenesis [32] and its derivatives can act as independent EGFR-inhibitors and inducers of apoptosis [33,34] providing further arguments in favor of using this 1-AH unit for the design of new anticancer drugs.
Herein we report on attempts to attach the 1-AH unit at the lactam moiety of 11nitroindolo [2,3-d]benzazepine-7-one and create a suitable binding site at position 11 for accommodation of Ru(II)/Os(II) and Cu(II).The prepared two ligands (HL 1 and HL 2 ) along with Ru(II)-and Os(II)-p-cymene, and Cu(II) complexes 1-3 collected in Figure 3 were spectroscopically ( 1 H, 13 C NMR, 2D NMR, UV-vis) characterized and their structures confirmed by ESI mass spectrometry and single-crystal X-ray diffraction (SC-XRD), as well as by Electron Diffraction (ED), a new technique, which was used for determination of the structure of the nanometer size crystals.All compounds were assessed for antiproliferative activity and structure-activity relationships were discussed.

Synthesis and Spectroscopic Analysis
The synthesis of potential ligands HL 1 and HL 2 started with attempts to couple 1-AH to scaffold A at the lactam moiety.As thiolactam group is considered to be more suitable Transition metal complexes of hydantoin derivatives are known [30] and positive effects of hydantoin on antimitotic and antiviral activities of inactive amino acids were reported [31].In addition, this moiety has been proven to have an immunomodulating effect antagonizing carcinogenesis [32] and its derivatives can act as independent EGFRinhibitors and inducers of apoptosis [33,34] providing further arguments in favor of using this 1-AH unit for the design of new anticancer drugs.
Herein we report on attempts to attach the 1-AH unit at the lactam moiety of 11nitroindolo [2,3-d]benzazepine-7-one and create a suitable binding site at position 11 for accommodation of Ru(II)/Os(II) and Cu(II).The prepared two ligands (HL 1 and HL 2 ) along with Ru(II)-and Os(II)-p-cymene, and Cu(II) complexes 1-3 collected in Figure 3 were spectroscopically ( 1 H, 13 C NMR, 2D NMR, UV-vis) characterized and their structures confirmed by ESI mass spectrometry and single-crystal X-ray diffraction (SC-XRD), as well as by Electron Diffraction (ED), a new technique, which was used for determination of the structure of the nanometer size crystals.All compounds were assessed for antiproliferative activity and structure-activity relationships were discussed.

Synthesis and Spectroscopic Analysis
The synthesis of potential ligands HL 1 and HL 2 started with attempts to couple 1-AH to scaffold A at the lactam moiety.As thiolactam group is considered to be more suitable for condensation reaction with hydrazine monohydrate [27,28], we tried first to thionate species A. The reaction of A with P 4 S 10 afforded the thiolactam derivative in low yield.However, further reaction with 1-AH failed, and species B could not be prepared.When A was allowed to react with an excess POCl 3 at reflux under inert atmosphere a more reactive 7-chloroimine derivative was generated, a species already reported previously for indoloquinolines [6,25,35].Further reaction with excess 1-AH in dry MeCN produced compound B, which was isolated as hydrochloride salt.The hydrochloride in EtOAc was neutralized with a saturated aqueous solution of sodium bicarbonate to give the free base.
Purification on silica resulted in the desired product of excellent purity and moderate yield (60%).Positive ion ESI mass spectrum revealed a peak with m/z 391.17 attributed to [M+H] + (calcd m/z 391.12).A single-crystal X-ray diffraction structure of species B confirmed the attachment of 1-AH moiety to the main scaffold (vide infra).
Further reduction in B with molecular hydrogen in the presence of Pd/C at 3 bar in dry THF [26] yielded pure amine C in almost quantitative yield.Condensation of amine C bearing 1-AH moiety with pyridine-2-carboxaldehyde and 2-formyl-6-morpholinomethylpyridine was expected to afford Schiff bases bearing the same 1-AH moiety.However, spectroscopic ( 1 H, 13 C and 15 N-1 H HSQC NMR) investigation of analytically pure products HL 1 and HL 2 indicated the formation of unexpected species via intramolecular ring closure reaction, which is facilitated in the presence of base.Suspending C in methanol with a small amount of NEt 3 and stirring at reflux for 3 h resulted in cyclization product C* (see Scheme 1).The typical chemical shift for H 5 (~10.9ppm) in Latonduines [26][27][28] in 1 H NMR spectra of B and C disappeared in the case of C* (see Figures S4-S6).Upon Schiffbase condensation of C with the two aldehydes to HL 1 and HL 2 , respectively, this singlet disappeared as well (see Figures S7 and S8), indicating a similar transformation with that of C into C*.In addition, 15 N-1 H HSQC NMR spectra revealed two new protons of the amide group further attesting the identity of HL 1 and HL 2 (see Figure S15).Positive ion ESI mass spectra of HL 1 and HL 2 showed peaks with m/z 472.22 and 549.33 attributed to [M+Na] + and [M+H] + , respectively.The molecular structure of the compounds was further confirmed by single-crystal X-ray diffraction study of HL 1 (vide infra).
species A. The reaction of A with P4S10 afforded the thiolactam derivative in low yield.However, further reaction with 1-AH failed, and species B could not be prepared.When A was allowed to react with an excess POCl3 at reflux under inert atmosphere a more reactive 7-chloroimine derivative was generated, a species already reported previously for indoloquinolines [6,25,35].Further reaction with excess 1-AH in dry MeCN produced compound B, which was isolated as hydrochloride salt.The hydrochloride in EtOAc was neutralized with a saturated aqueous solution of sodium bicarbonate to give the free base.Purification on silica resulted in the desired product of excellent purity and moderate yield (60%).Positive ion ESI mass spectrum revealed a peak with m/z 391.17 attributed to [M+H] + (calcd m/z 391.12).A single-crystal X-ray diffraction structure of species B confirmed the attachment of 1-AH moiety to the main scaffold (vide infra).
Further reduction in B with molecular hydrogen in the presence of Pd/C at 3 bar in dry THF [26] yielded pure amine C in almost quantitative yield.Condensation of amine C bearing 1-AH moiety with pyridine-2-carboxaldehyde and 2-formyl-6-morpholinomethyl-pyridine was expected to afford Schiff bases bearing the same 1-AH moiety.However, spectroscopic ( 1 H, 13 C and 15 N-1 H HSQC NMR) investigation of analytically pure products HL 1 and HL 2 indicated the formation of unexpected species via intramolecular ring closure reaction, which is facilitated in the presence of base.Suspending C in methanol with a small amount of NEt3 and stirring at reflux for 3 h resulted in cyclization product C* (see Scheme 1).The typical chemical shift for H 5 (~10.9ppm) in Latonduines [26][27][28] in 1 H NMR spectra of B and C disappeared in the case of C* (see Figures S4-S6).Upon Schiff-base condensation of C with the two aldehydes to HL 1 and HL 2 , respectively, this singlet disappeared as well (see Figures S7 and S8), indicating a similar transformation with that of C into C*.In addition, 15 N-1 H HSQC NMR spectra revealed two new protons of the amide group further attesting the identity of HL 1 and HL 2 (see Figure S15).Positive ion ESI mass spectra of HL 1 and HL 2 showed peaks with m/z 472.22 and 549.33 attributed to [M+Na] + and [M+H] + , respectively.The molecular structure of the compounds was further confirmed by single-crystal X-ray diffraction study of HL 1 (vide infra).The synthesis of metal(II)-arene complexes 1 and 2 was performed by reactions of HL 1 with [MCl2(p-cymene)]2 (M = Ru, Os) in hot isopropanol as reported for other related complexes elsewhere. 26ESI mass spectra of 1 and 2 revealed peaks with m/z 720.21 and 810.26, respectively, which could be easily assigned to [M(p-cymene)(HL 1 )Cl] + , where M = Ru and Os.The experimental isotopic patterns for these ions were in very good agreement with calculated isotopic distributions (see Supplementary Material).Complex 3 was synthesized by reaction of HL 2 with CuCl2•2H2O in boiling isopropanol as recently reported for similar compounds [27,28].The ESI mass spectrum of 3 showed a peak with m/z 646.20, which could be assigned to [Cu(HL 2 )Cl] + (calcd m/z 646.13).If the structure of diamagnetic complexes 1 and 2 could be confirmed by 1 H and 13 C NMR spectra, the coordination geometry of copper(II) in complex 3 could not be inferred directly from ESI mass spectrometry investigation or elemental analysis.Attempts to obtain single crystals of Xray diffraction quality for routine measurements failed.However, investigation of the powdered sample under electronic microscope (see Experimental part) showed the presence of very small crystals of 100 nm size, which proved to be suitable for collection of electron diffraction (ED) data, structure solution and refinement.The results of ED study The synthesis of metal(II)-arene complexes 1 and 2 was performed by reactions of HL 1 with [MCl 2 (p-cymene)] 2 (M = Ru, Os) in hot isopropanol as reported for other related complexes elsewhere [26].ESI mass spectra of 1 and 2 revealed peaks with m/z 720.21 and 810.26, respectively, which could be easily assigned to [M(p-cymene)(HL 1 )Cl] + , where M = Ru and Os.The experimental isotopic patterns for these ions were in very good agreement with calculated isotopic distributions (see Supplementary Material).Complex 3 was synthesized by reaction of HL 2 with CuCl 2 •2H 2 O in boiling isopropanol as recently reported for similar compounds [27,28].The ESI mass spectrum of 3 showed a peak with m/z 646.20, which could be assigned to [Cu(HL 2 )Cl] + (calcd m/z 646.13).If the structure of diamagnetic complexes 1 and 2 could be confirmed by 1 H and 13 C NMR spectra, the coordination geometry of copper(II) in complex 3 could not be inferred directly from ESI mass spectrometry investigation or elemental analysis.Attempts to obtain single crystals of X-ray diffraction quality for routine measurements failed.However, investigation of the powdered sample under electronic microscope (see Experimental part) showed the presence of very small crystals of 100 nm size, which proved to be suitable for collection of electron diffraction (ED) data, structure solution and refinement.The results of ED study of complex 3 and X-ray diffraction measurements of species B and ligand HL 1 are discussed below.
Elemental analyses of B, HL 1 , HL 2 and 1-3 (Table S1 in the Supplementary Material) confirmed their ≥95% purity required for in vitro testing.

Molecular Structure Description of Species B, HL 1 and 3
The results of X-ray diffraction measurements of species B and potential ligand HL 1 are shown in Figure 4, while those of the electron diffraction structure of complex 3 in Figure 5. Details of data collection and refinement are summarized in Table S2.
of complex 3 and X-ray diffraction measurements of species B and ligand HL 1 are discussed below.
Elemental analyses of B, HL 1 , HL 2 and 1-3 (Table S1 in the Supplementary Material) confirmed their ≥95% purity required for in vitro testing.

Molecular Structure Description of Species B, HL 1 and 3
The results of X-ray diffraction measurements of species B and potential ligand HL 1 are shown in Figure 4, while those of the electron diffraction structure of complex 3 in Figure 5. Details of data collection and refinement are summarized in Table S2.
Species B crystallized in the monoclinic centrosymmetric space group C2/c, HL 1 in the centrosymmetric triclinic space group P1 ̅ , while complex 3 in the monoclinic centrosymmetric space group P21/n.Species B crystallized in the monoclinic centrosymmetric space group C2/c, HL 1 in the centrosymmetric triclinic space group P1, while complex 3 in the monoclinic centrosymmetric space group P2 1 /n.
Complex 3 adopts a slightly distorted square-pyramidal coordination geometry (τ 5 = 0.07) [36].The Schiff base acts as a neutral tridentate ligand HL 2 binding to copper(II) via azomethine nitrogen atom N19, the pyridine nitrogen N22 and morpholine nitrogen atom N28 and together with the chlorido co-ligand Cl1 forms the base of the pyramid.The apical position is occupied by the second chlorido co-ligand Cl2.Complex 3 adopts a slightly distorted square-pyramidal coordination geometry (τ5 = 0.07) [36].The Schiff base acts as a neutral tridentate ligand HL 2 binding to copper(II) via azomethine nitrogen atom N19, the pyridine nitrogen N22 and morpholine nitrogen atom N28 and together with the chlorido co-ligand Cl1 forms the base of the pyramid.The apical position is occupied by the second chlorido co-ligand Cl2.

Stability in Solution
To confirm solution stability of the compounds to be tested in vitro for their antiproliferative activity 1% aqueous DMSO solutions of complexes 1-3 were monitored by UVvis spectroscopy over 72 h.Figures S1 and S2 indicate that complexes 1 and 2 remain intact in solution, while Figure S3 shows decrease in intensity over time for complex 3 due to precipitation of the compound.Moreover, no isosbestic points were observed indicating that complex 3 is stable in biocompatible media over 72 h.These data are in line with recent findings that metal-arene complexes with this type of coordination mode are resistant to co-ligand(s) exchange for DMSO [26].

Antiproliferative Activity
The antiproliferative activity of organic species B, HL 1 and HL 2 and complexes 1-3 was investigated in three human tumor cell lines (lung adenocarcinoma A549, colon adenocarcinoma LS-174 and triple-negative breast adenocarcinoma MDA-MB-231) and one normal human lung fibroblast cell line MRC-5.Cisplatin was used as reference compound.The IC50 values (µ M) after 72 h of continuous drug action are summarized in Table 1 and a bar graph plot is shown in Figure S16 (Supplementary Material).All organic compounds are almost nontoxic in A549 cells and revealed low toxicity in LS-174 and MDA-MB-231 cells, which is comparable with that in noncancerous cells MRC-5.Comparison of the IC50 values for HL 1 and HL 2 in cancer cell lines suggests that morpholine moiety does not have any effect on the cytotoxicity in cancer cells.Complex formation enhances the antiproliferative activity of HL 1 and HL 2 in all cell lines tested.Copper(II) complex 3 exhibited the highest antiproliferative activity in the tested cell lines with lowest IC50 value of 10.9 ± 0.8 µ M in LS-174 cells, which is even more effective than cisplatin and showed significant selectivity for this cells line when compared to that in noncancerous cell line (Selectivity Index 2.9), which was superior to that of the reference drug (0.62).Increased cytotoxicity of transition metal complexes in comparison with that of organic ligands is probably due to enhanced mechanisms of cellular uptake via metal-specific transporter

Stability in Solution
To confirm solution stability of the compounds to be tested in vitro for their antiproliferative activity 1% aqueous DMSO solutions of complexes 1-3 were monitored by UV-vis spectroscopy over 72 h.Figures S1 and S2 indicate that complexes 1 and 2 remain intact in solution, while Figure S3 shows decrease in intensity over time for complex 3 due to precipitation of the compound.Moreover, no isosbestic points were observed indicating that complex 3 is stable in biocompatible media over 72 h.These data are in line with recent findings that metal-arene complexes with this type of coordination mode are resistant to co-ligand(s) exchange for DMSO [26].

Antiproliferative Activity
The antiproliferative activity of organic species B, HL 1 and HL 2 and complexes 1-3 was investigated in three human tumor cell lines (lung adenocarcinoma A549, colon adenocarcinoma LS-174 and triple-negative breast adenocarcinoma MDA-MB-231) and one normal human lung fibroblast cell line MRC-5.Cisplatin was used as reference compound.The IC 50 values (µM) after 72 h of continuous drug action are summarized in Table 1 and a bar graph plot is shown in Figure S16 (Supplementary Material).All organic compounds are almost nontoxic in A549 cells and revealed low toxicity in LS-174 and MDA-MB-231 cells, which is comparable with that in noncancerous cells MRC-5.Comparison of the IC 50 values for HL 1 and HL 2 in cancer cell lines suggests that morpholine moiety does not have any effect on the cytotoxicity in cancer cells.Complex formation enhances the antiproliferative activity of HL 1 and HL 2 in all cell lines tested.Copper(II) complex 3 exhibited the highest antiproliferative activity in the tested cell lines with lowest IC 50 value of 10.9 ± 0.8 µM in LS-174 cells, which is even more effective than cisplatin and showed significant selectivity for this cells line when compared to that in noncancerous cell line (Selectivity Index 2.9), which was superior to that of the reference drug (0.62).Increased cytotoxicity of transition metal complexes in comparison with that of organic ligands is probably due to enhanced mechanisms of cellular uptake via metal-specific transporter proteins, such as transferrin for Ru(II) [37][38][39], and Ctr1 or ATOX1 for Cu(II) [40][41][42][43].In addition, the higher antiproliferative activity of copper(II) complex 3 in comparison to that of the corresponding ligand, as well as to that of ruthenium(II)-p-cymene-and osmium(II)-p-cymene complexes 1 and 2 is their potential ability of generation ROS via Fenton-like reactions as was observed for other copper(II) complexes with related tridentate ligands [44].To find an explanation for the low cytotoxicity of B we looked at docking capability of B to PIM-1 enzyme binding pocket.PIM-1 was chosen, since Meggers et al. co-crystallized selected compounds with structures related to Paullones with this protooncogene [29], and due to recently published studies, where indolo[2,3-c]quinolines and indolo [2,3-d]benzazepines modified at the lactam moiety showed strong binding to its ATP-binding pocket [28].

Molecular Docking of Species B to PIM-1
Structures A and B were docked to the binding sites of PIM-1 (PDB ID: 1YXX, resolution 2.00 Å) [45] kinase.The robustness of the docking scaffold has been previously reported [28].The GoldScore (GS) [46] and ChemScore(CS) [47,48] ChemPLP (Piecewise Linear Potential) [49] and ASP (Astex Statistical Potential) [50] scoring functions were implemented to predict the binding modes and relative energies of the ligands using the GOLD (v2020.2.0) software suite.The binding scores, for the ligands are given in Table S3 and they have similar scores to the co-crystallized ligand LI7 [45] indicating reasonable binding.Derivative B, with the hydantoin unit, has better scores than species A except for CS.
Analysis of the docking revealed that all of the scoring functions predicted the same conformation in the binding pocket except CS for structure B. In the case of A ASP and PLP predicted the same conformation as for B whereas the other two were quite different.Finally, the co-crystallized ligand LI7 was overlapped with the dominant conformation of B as shown in Figure 6A.A very good fit is predicted as seen in Figure 6A as the B fully occupies the binding pocket without any clashes with the adjacent amino acid residues.However, no hydrogen bonding is found for the dominant conformation to the neighboring amino acids as shown in Figure 6B.The other conformations were predicted to bind to PIM-1 via hydrogen bonds but consequently had a poorer fit in the pocket.

Chemical Space
The calculated molecular descriptors MW (molecular weight), log P (water-octanol partition coefficient), HD (hydrogen bond donors), HA (hydrogen bond acceptors), PSA (polar surface area) and RB (rotatable bonds) are given in Table S4 derived using the QikProp software [51].The values for the ligands' descriptors lie mostly within lead-and drug-like chemical space with the notable exception that PSA for species B, which is outside the Known Drug Space (KDS) (for the definition of lead-like, drug-like and KDS regions see [52] and Table S5).
The Known Drug Indexes (KDIs) for the species B and A were calculated to gauge the balance of the molecular descriptors (MW, logP, HD, HA, PSA and RB).This method is based on the analysis of drugs in clinical use, i.e., the statistical distribution of each descriptor is fitted to a Gaussian function and normalized to 1 resulting in a weighted index.Both the summation of the indexes (KDI 2a ) and multiplication (KDI 2b ) methods were used [53] as shown for KDI 2a in Equation ( 1) and for KDI 2b in Equation ( 2 The numerical results are given in Table S6 in the Supplementary Material.

Chemical Space
The calculated molecular descriptors MW (molecular weight), log P (water-octanol partition coefficient), HD (hydrogen bond donors), HA (hydrogen bond acceptors), PSA (polar surface area) and RB (rotatable bonds) are given in Table S4 derived using the QikProp software [51].The values for the ligands' descriptors lie mostly within lead-and drug-like chemical space with the notable exception that PSA for species B, which is outside the Known Drug Space (KDS) (for the definition of lead-like, drug-like and KDS regions see [52] and Table S5).
The Known Drug Indexes (KDIs) for the species B and A were calculated to gauge the balance of the molecular descriptors (MW, logP, HD, HA, PSA and RB).This method is based on the analysis of drugs in clinical use, i.e., the statistical distribution of each descriptor is fitted to a Gaussian function and normalized to 1 resulting in a weighted index.Both the summation of the indexes (KDI2a) and multiplication (KDI2b) methods were used [53] as shown for KDI2a in Equation ( 1) and for KDI2b in Equation ( 2  The KDI 2a values for the ligands are 4.92 (B) and 5.19 (A) with a theoretical maximum of 6 and the average of 4.08 (±1.27) for known drugs.The KDI 2b are 0.20 (B) and 0.40 (A), with a theoretical maximum of 1 and with KDS average of 0.18 (±0.20).This means that the molecular descriptors' balance is good resulting in good biocompatibility.
Species B contains an imine group, which is quite electron rich rendering it susceptible to an electrophilic attack and a nitro group susceptible for reduction.To test this the ionization potential (one-electron oxidation) and electron affinity (one-electron reduction) were derived for B using the density functional theory (DFT) and compared to the statistical distribution of known drugs [54].The ionization potential is 7.2 eV and 95% of drugs lie in the 6.0-9.0 eV range; the electron affinity is -1.5 eV with drugs in the -1.5-2.0 eV range [54].Thus, B is within the ranges of known drugs albeit at the edge of one-electron reduction.Furthermore, B has one N-N single bond, which is relatively weak compared to their C-C counterparts.Therefore, the bond dissociation energy (BDE) for the N-N single bond was also derived using DFT, resulting in 62.7 kcal/mol, which is substantially higher than the average for drugs with N-N bonds (53.9 kcal/mol, n = 23) [55].

1-((11-nitro-5,8-dihydroindolo[2,3-d]benzazepine-7(6H)-ylidene)amino)imidazolidine-2,4-dione (B)
A suspension of A (300 mg, 1.02 mmol) in POCl 3 (30 mL) under inert atmosphere was stirred at 130 • C for 3 h and then cooled to room temperature.The solvent was removed and to the residue dry acetonitrile (90 mL) was added.The suspension was cooled to 0 • C and added dropwise to a slurry of 1-aminohydantoine in dry acetonitrile (90 mL) at 0 • C. The resulting mixture was stirred at 0 • C for 1 h, then at room temperature for 1 h and at 50 • C overnight.The precipitate was filtered off, treated with EtOAc and saturated aqueous solution of NaHCO 3 (100 mL).The organic phase was separated, while the aqueous phase was extracted with EtOAc (2 × 50 mL) The combined organic phases were dried over MgSO 4 and purified on silica by using DCM:MeOH 93:7 as eluent to give a yellow solid.Yield: 237 mg (60%). 1
To species C (200 mg, 0.58 mmol) in anoxic ethanol (2 mL) 2-formylpyridine (61 µL, 0.64 mmol) was added.The resulting solution was stirred at 85 • C for 16 h.The precipitate was filtered off after cooling the reaction mixture to room temperature and washed with EtOH (2 mL) and Et 2 O (2 mL).The product was dried at 50 • C in vacuo to give a light-orange solid.Yield: 180 mg (69%). 1

3•H 2 O•0.3 i PrOH
To a solution of HL 2 (109 mg, 0.2 mmol) in i PrOH (60 mL) at 70 • C CuCl 2 •2H 2 O (34 mg, 0.2 mmol) in MeOH (200 µL) was added.The resulting suspension was stirred at reflux for 45 min and left to stand at 4 • C overnight.The precipitate was filtered off and washed with i PrOH (3 mL).The product was dried at 50 • C in vacuo to give a light-brown solid.Yield 120 mg (89%) For electron diffraction measurements, nanometer size crystalline complex 3 was generated by slow diffusion of acetonitrile into a concentrated solution of crude 3 in DMF.

Crystallographic Structure Determination
The measurements were performed on a Bruker D8 Venture diffractometer.Single crystals were positioned at 40 and 35 mm from the detector, and 717 and 1320 frames were measured, each for 10 and 3 s over 0.360 • scan, respectively.The data were processed using SAINT software [59].Crystal data, data collection parameters, and structure refinement details are given in Table S2.The structures were solved by direct methods and refined by full-matrix least-squares techniques.Non-H atoms were refined with anisotropic displacement parameters.H atoms were inserted in calculated positions and refined with a riding model.The following computer programs and hardware were used: structure solution, SHELXS and refinement, SHELXL [60]; molecular diagrams, ORTEP [61]; computer, Intel CoreDuo.CCDC 2219174 (B•CH 3 OH), 2219175 (HL 1 •2DMF) and 2218795 (3).

Sample Preparation and Data Collection for Complex 3 by Electron Diffraction
A few dry grains of 3 were provided in a glass vial with about 11 mm diameter (see Figure 7).A TEM copper grid with 3.05 mm diameter and coated with lacey carbon (Ted Pella) was added to the vial.The grid was not glow-discharged before use but used as purchased.The vial was vortexed for 1 min at 3000/s.The grid was inserted into a Gatan cryo-transfer holder ELSA 698 at room temperature and inserted into a transmission electron microscope JEM2100Plus, JEOL Ltd.Electron source was a LaB6 emitter operated at 200 keV.Data were collected with a JUNG-FRAU detector, 1024 × 512 pixel [62].At first, condenser lens aperture 3 and spot size 5 were used.This lead to a degradation of the diffraction power during data collection.Consequently, the beam intensity was reduced by setting to the smallest available condenser lens aperture (50 µ m diameter), and the second to weakest spot size (spot size 4 out of 5).Data were collected with a rotation velocity of nominally 0.05°/s.Data were collected from 15 crystals (see Figures 8 and 9).Four data sets could be indexed and thus processed (see Figure 10).Data were processed with XDS [63].Scaling was switched off during data integration [64] and only applied during scaling with XSCALE.Resolution The grid was inserted into a Gatan cryo-transfer holder ELSA 698 at room temperature and inserted into a transmission electron microscope JEM2100Plus, JEOL Ltd. (Akishima, Japan).Electron source was a LaB6 emitter operated at 200 keV.Data were collected with a JUNGFRAU detector, 1024 × 512 pixel [62].At first, condenser lens aperture 3 and spot size 5 were used.This lead to a degradation of the diffraction power during data collection.Consequently, the beam intensity was reduced by setting to the smallest available condenser lens aperture (50 µM diameter), and the second to weakest spot size (spot size 4 out of 5).Data were collected with a rotation velocity of nominally 0.05 • /s.Data were collected from 15 crystals (see Figures 8 and 9).Four data sets could be indexed and thus processed (see Figure 10).Data were processed with XDS [63].Scaling was switched off during data integration [64] and only applied during scaling with XSCALE.Resolution was cut to 0.7 Å based on I/σ(I) > 1 and CC1/2 > 30% [65].Unit cell parameters suggested a monoclinic lattice, systematic absences suggested space group P2 1 /n.The structure was solved with SHELXD 2013/2 [60] and refined with SHELXL 2019/1 [66] in kinematic approximation.Scattering factors were fitted in Cromer-Mann parametrization f(s) = sum (i = 1...4) a_i exp(-b_i*sˆ2) + c against the values tabulated in Table 4.2.6.8 of the International Tables of Crystallography, Volume C [67], starting from the parameters published by Peng 1999 [68].Hydrogen atoms were placed in riding positions at inter-nuclear distances [69].Geometric 1,2-and 1,3-distance restraints were generated with OPENBABEL [70] and the GRADEserver [71].Only the ligand was restrained, and the distances listed in the legend to Figure 5 are unrestrained.Unit cell dimensions very optimized against the geometric restraints with CELLOPT [72].
The grid was inserted into a Gatan cryo-transfer holder ELSA 698 at room temperature and inserted into a transmission electron microscope JEM2100Plus, JEOL Ltd.Electron source was a LaB6 emitter operated at 200 keV.Data were collected with a JUNG-FRAU detector, 1024 × 512 pixel [62].At first, condenser lens aperture 3 and spot size 5 were used.This lead to a degradation of the diffraction power during data collection.Consequently, the beam intensity was reduced by setting to the smallest available condenser lens aperture (50 µ m diameter), and the second to weakest spot size (spot size 4 out of 5).Data were collected with a rotation velocity of nominally 0.05°/s.Data were collected from 15 crystals (see Figures 8 and 9).Four data sets could be indexed and thus processed (see Figure 10).Data were processed with XDS [63].Scaling was switched off during data integration [64] and only applied during scaling with XSCALE.Resolution was cut to 0.7 Å based on I/σ(I) > 1 and CC1/2 > 30% [65].Unit cell parameters suggested a monoclinic lattice, systematic absences suggested space group P21/n.The structure was solved with SHELXD 2013/2 [60] and refined with SHELXL 2019/1 [66] in kinematic approximation.Scattering factors were fitted in Cromer-Mann parametrization f(s) = sum (i = 1...4) a_i exp(-b_i*s^2) + c against the values tabulated in Table 4.2.6.8 of the International Tables of Crystallography, Volume C [67], starting from the parameters published by Peng 1999 [68].Hydrogen atoms were placed in riding positions at inter-nuclear distances [69].Geometric 1,2-and 1,3-distance restraints were generated with OPENBABEL [70] and the GRADE-server [71].Only the ligand was restrained, and the distances listed in the legend to Figure 5 are unrestrained.Unit cell dimensions very optimized against the geometric restraints with CELLOPT [72].The grid was inserted into a Gatan cryo-transfer holder ELSA 698 at room temperature and inserted into a transmission electron microscope JEM2100Plus, JEOL Ltd.Electron source was a LaB6 emitter operated at 200 keV.Data were collected with a JUNG-FRAU detector, 1024 × 512 pixel [62].At first, condenser lens aperture 3 and spot size 5 were used.This lead to a degradation of the diffraction power during data collection.Consequently, the beam intensity was reduced by setting to the smallest available condenser lens aperture (50 µ m diameter), and the second to weakest spot size (spot size 4 out of 5).Data were collected with a rotation velocity of nominally 0.05°/s.Data were collected from 15 crystals (see Figures 8 and 9).Four data sets could be indexed and thus processed (see Figure 10).Data were processed with XDS [63].Scaling was switched off during data integration [64] and only applied during scaling with XSCALE.Resolution was cut to 0.7 Å based on I/σ(I) > 1 and CC1/2 > 30% [65].Unit cell parameters suggested a monoclinic lattice, systematic absences suggested space group P21/n.The structure was solved with SHELXD 2013/2 [60] and refined with SHELXL 2019/1 [66] in kinematic approximation.Scattering factors were fitted in Cromer-Mann parametrization f(s) = sum (i = 1...4) a_i exp(-b_i*s^2) + c against the values tabulated in Table 4.2.6.8 of the International Tables of Crystallography, Volume C [67], starting from the parameters published by Peng 1999 [68].Hydrogen atoms were placed in riding positions at inter-nuclear distances [69].Geometric 1,2-and 1,3-distance restraints were generated with OPENBABEL [70] and the GRADE-server [71].Only the ligand was restrained, and the distances listed in the legend to Figure 5 are unrestrained.Unit cell dimensions very optimized against the geometric restraints with CELLOPT [72].The grid was inserted into a Gatan cryo-transfer holder ELSA 698 at room temperature and inserted into a transmission electron microscope JEM2100Plus, JEOL Ltd.Electron source was a LaB6 emitter operated at 200 keV.Data were collected with a JUNG-FRAU detector, 1024 × 512 pixel [62].At first, condenser lens aperture 3 and spot size 5 were used.This lead to a degradation of the diffraction power during data collection.Consequently, the beam intensity was reduced by setting to the smallest available condenser lens aperture (50 µ m diameter), and the second to weakest spot size (spot size 4 out of 5).Data were collected with a rotation velocity of nominally 0.05°/s.Data were collected from 15 crystals (see Figures 8 and 9).Four data sets could be indexed and thus processed (see Figure 10).Data were processed with XDS [63].Scaling was switched off during data integration [64] and only applied during scaling with XSCALE.Resolution was cut to 0.7 Å based on I/σ(I) > 1 and CC1/2 > 30% [65].Unit cell parameters suggested a monoclinic lattice, systematic absences suggested space group P21/n.The structure was solved with SHELXD 2013/2 [60] and refined with SHELXL 2019/1 [66] in kinematic approximation.Scattering factors were fitted in Cromer-Mann parametrization f(s) = sum (i = 1...4) a_i exp(-b_i*s^2) + c against the values tabulated in Table 4.2.6.8 of the International Tables of Crystallography, Volume C [67], starting from the parameters published by Peng 1999 [68].Hydrogen atoms were placed in riding positions at inter-nuclear distances [69].Geometric 1,2-and 1,3-distance restraints were generated with OPENBABEL [70] and the GRADE-server [71].Only the ligand was restrained, and the distances listed in the legend to Figure 5 are unrestrained.Unit cell dimensions very optimized against the geometric restraints with CELLOPT [72].2).The cells were grown at 37 • C in a humidified atmosphere containing 5% CO 2 .

MTT-Assay
The cytotoxic activity of the investigated complexes and their corresponding ligands was analyzed in comparison to cisplatin, cis-diamminedichloridoplatinum(II) (CDDP) as reference compound, using the 3-(4,5-dymethylthiazol-yl)-2,5-diphenyltetrazolium bromide (MTT, Sigma) assay as previously described [73,74].Cells were seeded in 96-well culture plates (Thermo Scientific Nunc™, Rochester, NY, USA) at cell densities of 5000 c/w, in a total volume of 100 µL of nutrient medium, and then cultured in 37 • C incubator, in a humidified atmosphere of 5% CO 2 .Briefly, 24 h after seeding, cells were exposed to the investigated compounds.The complexes and ligands were dissolved in dimethyl sulfoxide (DMSO) at a concentration of 10 mM immediately prior the use, and afterwards diluted in culture medium to the desired concentrations.The final DMSO concentration never exceeded 1% (v/v).After an incubation period of 72 h, 20 µL of MTT solution (5 mg/mL in phosphate buffer, pH 7.2) was added to each well.Samples were incubated at 37 • C for 4 h, in 5% CO 2 incubator, and then 100 µL of 10% sodium dodecyl sulfate (SDS) was added.Absorbance was recorded after 24 h, on an enzyme linked immunosorbent assay (ELISA) reader (Thermo Labsystems Multiskan EX 200-240 V), at the wavelength of 570 nm.The IC 50 value, defined as the concentration of the compound causing 50% cell growth inhibition, was determined from the dose response curves.
The QikProp v6.2 [51] software package was used to calculate the molecular descriptors of the molecules.The reliability of it QikProp established for the calculated descriptors [82].Furthermore, the Scigress version FJ 2.6 program [77] was used to calculate the molecular descriptors for the complexes.The Known Drug Indexes (KDI) were calculated from the molecular descriptors as described by Eurtivong and Reynisson [53] For application in Excel, columns for each property were created and the following equations used do derive the KDI numbers for each descriptor: The Gaussian 16 software suite [83] was used with unrestricted DFT.The B3LYP functional hybrid approach was employed [84][85][86] and standard 6-31+G(d,p) diffused basis set [87,88] was used for geometry optimization and frequency analysis (keywords: opt freq).The zero-point vibrational energies (ZPE) were scaled according to Wong (0.9804) [89].In all cases, normal modes revealed no imaginary frequencies indicating that they represent minima on the potential energy surface.The subsequent energy calculations were then performed with the larger 6-311+G(2df, p) basis set.Adiabatic ionization potentials (IP) and adiabatic electron affinities (EA) were calculated as described in Forseman and Frisch [90].The bond dissociation energies were calculated as in Yu and Reynisson [55].The energies and ZPA are given in Table S4.

Conclusions
Attachment of 1-AH moiety to a latonduine backbone was expected to enhance the docking ability to PIM-1 enzyme by exploiting its potential ability to form hydrogen bonds.Molecular docking calculations revealed that despite good accommodation of species B in the binding pocket of PIM-1 the 1-AH moiety is not involved in any hydrogen bonding interactions.Attempts to create a metal-binding site at the backbone of new hybrid molecule via Schiff base condensation reactions resulted in intramolecular cyclization with formation of a triazole ring and a dangling amide functional group with isolation of HL 1 and HL 2 .By further exploiting their complex formation ability ruthenium(II)-and osmium(II)-arene complexes 1 and 2, as well as copper(II) complex 3 were synthesized and characterized.For the first time we used electron diffraction for elucidation of molecular structure of a metal complex, which could not be crystallized to give single crystals suitable for routine X-ray diffraction study.Several nanometer size crystals provided electron diffraction data for determination of molecular structure of complex 3.All the organic compounds tested have not revealed marked cytotoxicity in cancer cell lines.Low antiproliferative activity of species B is likely due to large PSA (polar surface area) value lying outside drug-like space.In contrast to previously reported data, indicating that chemical modification of the lactam group at the indolobenzazepine backbone has a favorable effect on cytotoxicity, intramolecular cyclization with formation of a triazole ring fused to the core structure resulted in a marked drop of antiproliferative activity.Coordination of HL 1 and HL 2 to ruthenium(II)-, osmium(II)-arene and copper(II), respectively, enhanced markedly their antiproliferative activity.The IC 50 values of complexes 1-3 follow the order 2 > 1 > 3. The most cytotoxic in all three cancer cell lines complex 3, when compared to 1 and 2, in addition to the highest antiproliferative activity showed selectivity towards LS-174 cells (when compared to normal cells MRC-5) superior to that of clinical drug cisplatin.

Figure 2 .
Figure 2. From left to right: Ru(II)/Os(II) complexes with Paullone and Latonduine derivatives with different binding sites, and a Cu(II) complex with a Latonduine modified at the lactam moiety.

Figure 2 .
Figure 2. From left to right: Ru(II)/Os(II) complexes with Paullone and Latonduine derivatives with different binding sites, and a Cu(II) complex with a Latonduine modified at the lactam moiety.

Figure 2 .
Figure 2. From left to right: Ru(II)/Os(II) complexes with Paullone and Latonduine derivatives with different binding sites, and a Cu(II) complex with a Latonduine modified at the lactam moiety.

Figure 3 .
Figure 3. Core structure B (1-AH moiety is highlighted in blue), ligands HL 1 and HL 2 , and their Ru(II)-and Os(II)-arene, as well as Cu(II) complexes 1-3.Underlined numbers indicate compounds studied by SC-XRD or ED methods.

Figure 3 .
Figure 3. Core structure B (1-AH moiety is highlighted in blue), ligands HL 1 and HL 2 , and their Ru(II)-and Os(II)-arene, as well as Cu(II) complexes 1-3.Underlined numbers indicate compounds studied by SC-XRD or ED methods.

Inorganics 2023 , 18 Figure 6
Figure6(A) The docked pose of B using the ChemPLP scoring function in the PIM-1's binding site, the co-crystallized ligand LI7 is shown as green sticks, its hydrogens are not shown for clarity.The protein surface is rendered; blue color depicts regions with a partial positive charge on the surface; red color depicts regions with a partial negative charge and grey color shows neutral areas; (B) The predicted binding of species B, no hydrogen bonding interactions were predicted.The amino acids making up the docking scaffold are LEU44, GLY45, SER46, PHE49, ILE104, LEU120, ARG122, PRO123, ILE185 and ASP186 within 5 Å .

Figure 6 .
Figure 6.(A)The docked pose of B using the ChemPLP scoring function in the PIM-1's binding site, the co-crystallized ligand LI7 is shown as green sticks, its hydrogens are not shown for clarity.The protein surface is rendered; blue color depicts regions with a partial positive charge on the surface; red color depicts regions with a partial negative charge and grey color shows neutral areas; (B) The predicted binding of species B, no hydrogen bonding interactions were predicted.The amino acids making up the docking scaffold are LEU44, GLY45, SER46, PHE49, ILE104, LEU120, ARG122, PRO123, ILE185 and ASP186 within 5 Å.

Figure 8 .
Figure 8. Partial view of the crystal no.08 at 25,500-fold magnification.Field of view has a diameter of 1.9 µ m.

Figure 9 .
Figure 9. Partial view of crystal no. 13 at 25,500-fold magnification.Field of view has a diameter of 1.9 µ m.

Figure 8 .
Figure 8. Partial view of the crystal no.08 at 25,500-fold magnification.Field of view has a diameter of 1.9 µM.

Figure 8 .
Figure 8. Partial view of the crystal no.08 at 25,500-fold magnification.Field of view has a diameter of 1.9 µ m.

Figure 9 .
Figure 9. Partial view of crystal no. 13 at 25,500-fold magnification.Field of view has a diameter of 1.9 µ m.Figure 9. Partial view of crystal no. 13 at 25,500-fold magnification.Field of view has a diameter of 1.9 µM.

Figure 9 .
Figure 9. Partial view of crystal no. 13 at 25,500-fold magnification.Field of view has a diameter of 1.9 µ m.Figure 9. Partial view of crystal no. 13 at 25,500-fold magnification.Field of view has a diameter of 1.9 µM.

Figure 8 .
Figure 8. Partial view of the crystal no.08 at 25,500-fold magnification.Field of view has a diameter of 1.9 µ m.

Figure 9 .
Figure 9. Partial view of crystal no. 13 at 25,500-fold magnification.Field of view has a diameter of 1.9 µ m.

Figure 10 .
Figure 10.Diffraction image for crystal no.08.Effective detector distance is 665 mm.

Table 1 .
Cytotoxic activity of compounds B, HL1, HL 2 and 1-3 vs. cisplatin in terms of IC 50 values (µM), expressed as mean ± SD of at least two independent experiments performed in triplicates.IC 50 values of tested complexes and ligands were determined from dose-response curves after 72 h continuous drug action by MTT assay.