Design, synthesis and in vitro evaluations of new cyclotriphosphazenes as safe drug candidates

Abstract

Although it is possible to discover new drug candidate molecules using in silico approaches, chemical synthesis followed by screening of their functions is still at the center of bioactive molecule discovery. While determining the potential effects of compounds on target signaling molecules or pathways, assessing their effects on the circadian rhythm is also very important for determining the efficacy of drug candidates because they control most of the signaling pathways. Herein, new members of the biocompatible cyclotriphosphazene family were prepared, and their in vitro biological activities and effects on circadian rhythm were evaluated for the first time. In particular, new cyclotriphosphazene derivatives carrying morpholine, thiomorpholine and triazole groups were designed and synthesized, and their chemical structures were characterized using appropriate spectroscopic methods. Cellular toxicity analyses of the compounds were performed using different biological methods, such as determination of IC₅₀ values, calculation of population doubling times, and colony formation patterns. Subsequently, the effects of the compounds on the cell cycle were analyzed using the flow cytometry technique. Finally, the effects of the synthesized compounds on circadian rhythm were determined using a real-time bioluminescence approach. Based on these studies, it was determined that some compounds demonstrated varying degrees of antiproliferative activity, with the most potent compounds causing G₂/M phase arrest. Additionally, most derivatives had no adverse effects on the circadian rhythm, indicating their potential for safe therapeutic application in targeting cell proliferation. Furthermore, an important pharmacological characteristic of the drug candidate molecules, namely, membrane permeability in terms of log P values, was assessed. In conclusion, these novel cyclotriphosphazene-based compounds are a class of circadian rhythm-safe drug candidate compounds.

Cyclotriphosphazene compounds carrying morpholine, thiomorpholine and triazole groups were synthesized and their biological activities were investigated. These compounds were presented as a class of circadian rhythm-safe drug candidate compounds.

1. Introduction

Versatile nature of the heterocyclic motifs is of importance in inorganic and organic chemistry as they are used as sources for the preparation of new compounds with outstanding activities.¹ They constitute the molecular structure of the most biologically active pharmaceuticals and agrochemicals, where the heterocyclic moieties influence the physicochemical properties of the molecule.² As a biocompatible heterocyclic platform, cyclotriphosphazene ring facilitates easy and versatile functionalization via selective substitutions, which enable synthesis of the tailor-made regio- and stereo-specific derivatives.^3–6 For diverse applications, the phosphazene ring is generally used as an important starting core or intermediate, and the properties of the resulting materials vary with characteristics of the substituted side groups.^5–8 Numerous cyclotriphosphazene derivatives have been prepared, focusing not only on their structural chemistry^4,9–12 but also recently on their potential use as biomaterials,^13,14 anticancer,^15–17 anti-Alzheimer's,¹⁸ and antimicrobial agents.^14,19 An appropriate amphiphilic character is characteristic of meticulously designed therapy agents and poly(ethylene glycol) moieties that imitate biological transport systems,²⁰ and they have been extensively used in phosphazene chemistry to improve the solubility in biological media.^11,21 Previous studies have shown the close relationship of the chemical/physical properties with the nature of the substituted side groups, which has encouraged researchers to modify the known biologically active molecules to obtain more potent, more selective and safer agents.¹⁶

Interest in indole/pyrimidine heterocycles bearing morpholine and/or thiomorpholine has increased because of their valuable physicochemical properties, such as ionization, lipophilicity, polarity, and van der Waals volume, which affect their pharmacokinetic ability.^22,23 Heteroatoms and their moieties on the molecule influence the pK_a values and the ability to form hydrogen bonds.^2,24–26 Numerous morpholine derivatives with different functional groups are anti-tumorigenic, adrenergic, histaminergic or effective for Alzheimer's disease.^1,2,24 Because heterocycles form the majority of the pharmacophores (higher than 90%), it might be valuable to design new hybrid systems bearing heterocyclic components.^27,28 Despite the challenges in the draft, synthesis and optimization phases, such research has advantages over blending strategies. Previously, cyclotriphosphazene-morpholine derivatives were prepared and successfully tested for their physicochemical properties, DNA binding, in vitro antimicrobial, and/or cytotoxic activities.^29–31 However, there are no reports in the literature related to the systematic investigation of the effects of amphiphilic morpholine-containing cyclotriphosphazene derivatives on biological events directly related to human health, such as cell viability, cell differentiation, division and biological clock.

Based on these observations, it is worth investigating the effect of the presence of heterocyclic morpholine or thiomorpholine groups on the regio- and stereo-specific cyclotriphosphazene skeleton with or without triazole moieties on the potential to confer novel effects on a new class of biologically active molecules. The principal reason for selecting triazole moieties to click the morpholine or thiomorpholine was the studies describing the 1,2,3-triazole derivatives as an important class of heterocycles in medicinal chemistry.^25,32,33 Additionally, molecules were designed to have amphiphilic tripodal structures that can improve the interaction with biological media (Fig. 1). Furthermore, cell toxicity and its biological properties have been investigated methodically. Because the majority of metabolism and physiology are under the control of the circadian clock, which generates circadian (approximately 24-hour) rhythms,³⁴ the effect of cyclotriphosphazene on the circadian rhythm was also evaluated for the first time, providing a clue as to whether it could interfere with major metabolic events by disrupting the circadian clock.

Fig. 1. Chemical structures of cyclotriphosphazene compounds.

2. Results and discussion

2.1. Molecular design/synthesis and characterization of cyclotriphosphazene derivatives

It is possible to develop synthetic bioactive compounds based on naturally occurring small molecules or analogues of drug-active compounds. Morpholine is a heterocyclic ring containing oxygen and nitrogen atoms in its structure and is a bioactive molecule that is frequently used in medical applications. Morpholine contributes to the pharmacokinetic profile of the molecules it binds, as it can form hydrogen bonds through the oxygen in its structure and engage in hydrophobic interactions due to its relatively electron-deficient ring. In addition to the antibiotic and antimicrobial properties of morpholine derivatives, many other pharmacological effects, such as anticancer, antioxidant, and anti-inflammatory activities, are also known.^2,22,24,26 Thiomorpholine groups, which share a heterocyclic structure similar to morpholine, have a wide range of applications, including anticancer and antimicrobial properties. Triazole units are a pioneer group in drug design with their five-membered heterocyclic structure containing three nitrogen atoms that are suitable for hydrogen bonding and devoid of pi electrons.²⁵ The triazole ring enhances the resistance of molecules to both chemical and metabolic degradation.^27,28,33 Many anticancer drugs containing triazole units are also known to affect circadian rhythms.²⁷

In this study, heterocyclic motifs containing nitrogen, oxygen and sulfur atoms (morpholine, triazole, thiophene) were selected as bio-active groups to be incorporated into the cyclotriphosphazene core owing to their above-mentioned pharmacological advantages. The synthesis phase started with the reaction of hexachlorocyclotriphosphazene with triethylene glycol monomethyl ether, resulting in the formation of compound 1 with a hydrophilic character.³⁵ Subsequently, compounds 2 and 3 were prepared from the reactions of compound 1 with morpholine and 2-morpholino ethyl amine in THF in the presence of triethylamine, respectively (Scheme 1). Compound 4 was obtained from the reaction of compound 1 with 4-morpholine 4-phenol in THF and in the presence of NaH, while compounds 5 and 6 were prepared from the reactions of compound 1 with thiomorpholine and 2-(1,1-dioxido-4-thiomorpholinyl)ethanamine in THF and in the presence of triethylamine, respectively. After all the obtained compounds were purified, their structures were characterized by applying spectroscopic methods (mass, ³¹P, ¹H and ¹³C NMR). In all the MALDI-TOF analyses, the ion peaks of the molecules were consistent with the expected values. Mass analysis data for compound 1 showed that three chlorine atoms in the cyclotriphosphazene ring were replaced by triethylene glycol moieties, and the ³¹P NMR spectrum of compound 1 with the A₃ spin system supported these data. ³¹P NMR data of compounds 2–6 confirmed the binding of three morpholine units to the cyclotriphosphazene skeleton, in agreement with their mass spectra. Although a single peak was observed in the ³¹P NMR spectra of compounds 2, 3, 5 and 6 in the range of 22.6–19.8 ppm, compound 4 gave a signal at 13.6 ppm owing to the phenoxy group on it. In the ¹³C NMR spectra of the compounds (1, 2, 3, 5 and 6), carbon atoms of glycol chains appeared in the range of approximately 72.0–57.0 ppm, while aliphatic carbons resonated in the range of 58.0–37.0 ppm. The aromatic carbons of compound 4 were observed in the range of 107.5–97.7 ppm.

Scheme 1. Synthesis pathway of cyclotriphosphazene derivatives (1–10).

In the second part of the synthesis step, compound 7, as a starting compound of the second series, was prepared by reacting compound 1 with 2-azidooethanol.³⁶ Morpholine- and thiomorpholine-substituted cyclotriphosphazenes (8–10) with triazole rings were obtained from the click reactions of morpholine, thiomorpholine and commercially available 4-propargylthiomorpholine-1,1-dioxide units with compound 7 in the presence of CuI and diisopropylethylamine (DIPEA). The exact molecular structures of the compounds (8–10) were analyzed, and the molecular ion peaks of cyclotriphosphazene derivatives (8–10) were marked as 1258.033, 1305.826, and 1402.572 Da. When the ³¹P NMR spectra of the compounds were examined, it was observed that they gave a single peak in the range of 17.43–18.91 ppm owing to the chemical environment equivalence of all the phosphorous nuclei. In the ¹³C NMR spectra, the signals of the characteristic triazole carbon atoms of the compounds (8–10) were observed at approximately 177 ppm. Carbon atoms of glycol chains were in the range of 72.0–58.0 ppm, and other aliphatic carbons resonated in the range of 57.0–22.0 ppm. The integral values and chemical shifts in the ¹H NMR spectra of the compounds supported the proposed structures.

2.2. Biology

In the process of drug development from small molecules, initial compounds should be sufficiently optimized to be selective, potent and safe in preliminary in vitro experiments and animal models before being nominated as drug candidates.³⁷ Considering anticancer drug development research, compounds should be selectively cytotoxic to cancerous cells while minimally affecting normal cell models, such as normal fibroblasts. Anticancer drugs inhibit/kill cancer cells through several mechanisms. Therefore, after observing cell toxicity and selective potency, it is important to know the mechanism of the cytotoxicity. Most chemotherapy drugs or anticancer compounds have a direct action on DNA or proteins involved in DNA replication, translation and transcription, and the result is programmed cell death (such as apoptosis) or inhibition of cellular proliferation (by stopping the cell cycle).³⁸ However, circadian rhythms generated by a molecular circadian clock control major signaling pathways in cells, and many drug target proteins are under the control of the circadian clock and show daily oscillations.³⁹ With the appreciation of chronotherapy, it is becoming more important if a new compound would affect circadian oscillation because the interaction of the circadian with the compound may affect the efficacy of the drug candidates at the end.⁴⁰

Therefore, in this study, we systematically analyzed novel compounds owing to their cytotoxic properties and then their effect on the circadian rhythm. As the circadian clock cross talks with many signaling pathways and signals from other pathways may affect the oscillations, such as lowering amplitude and causing a phase shift or even arrhythmicity, the absence of adverse effects on the circadian rhythm can be considered another sign for the safety of the compounds.

Compound screening assays are typically run at 1–10 μM compound concentration⁴¹ even though compounds with IC₅₀ values < 100 mM are considered potentially active in the literature. However, if the final aim is not cytotoxicity, knowing the safety of the compound is also important, especially considering that it can be used as a starting material for non-cytotoxic applications in the future. Therefore, we tested our compounds up to 1 mM (as the highest concentration) to ensure their safety.

In addition to biological activity and screening experiments, bioavailability is a critical factor in drug design. An effective drug must reach its target at sufficient concentrations while maintaining its bioactive form throughout biological processes. Understanding the pharmacokinetic and physicochemical properties of a compound is therefore essential. The SwissADME web server facilitates the computational evaluation of these properties, including absorption, distribution, metabolism, and excretion (ADME).⁴² Although these predictions do not guarantee a molecule's efficacy as a drug, they provide valuable insights into drug-likeness and are widely utilized in the literature. Lipophilicity, expressed as the n-octanol–water partition coefficient (clog P), is a key determinant of passive membrane permeability. Although higher log P values generally enhance permeability, maintaining an optimal range is crucial to balancing solubility and membrane penetration. SwissADME predicts log P using five models: iLOGP, XLOGP3, WLOGP, MLOGP, and Silicos-IT LogP.⁴² Among them, XLOGP3, an atomistic method incorporating corrective factors and a knowledge-based library, is widely used.⁴³ The optimal log P range for drug-like properties is typically considered between −0.7 and +5. In our study, most compounds exhibit XLOGP3 values within this range (Table S1†), suggesting membrane permeability.

2.2.1. Toxic effects of cyclotriphosphazene derivatives in cellular models

Osteosarcoma U2OS and BJ human fibroblast cell lines were used to provide a comprehensive evaluation of the biological activities of cyclotriphosphazene derivatives. The U2OS cell line was chosen owing to its high proliferative capacity, intact cell cycle checkpoints and genetically stable circadian rhythm gene expression. U2OS cells are a widely used model in circadian rhythm research, as they naturally maintain rhythmic expression of the core circadian genes, such as CRYPTOCHROME, PERIOD, BMAL1 and CLOCK. These cells also provide a suitable biological system for studying antiproliferative effects and cell cycle regulation. As the BJ human fibroblast cell line is commonly used as a representative of normal cells by researchers, it was selected as a healthy, non-transformed and genetically stable cell line. Therefore, these two cell lines were used to identify toxicity differences between healthy and cancer cells. Even though fibroblast cells have rhythms, U2OS cells are preferred owing to their continuous proliferation ability and many robust oscillation patterns.⁴⁴ The cells were exposed to increasing concentrations (from 0 to 1 mM) of 10 cyclotriphosphazene derivatives for 24 and 48 hours (Fig. 2). Cell viability was determined using the MTT assay. To accurately determine IC₅₀ values, it was necessary to observe cell viability decreasing from 100% to 0%, which also required testing at very high concentrations. Compound 7 had the highest IC₅₀ (697.7 μM) value confirming its very low cytotoxicity, while the IC₅₀ value of each synthesized compound was lower than that value, indicating that derivatives had higher cytotoxicity compared to compound 7. Because of the lowest toxicity and being the starting material for the synthesis of series 2 compounds, we selected compound 7 as the reference molecule. Notably, compounds 3 and 4, with IC₅₀ values of 80.22 μM and 88.93 μM, respectively, emerged as potential candidates with promising anticancer properties in U2OS cells at lower concentrations (Table 1).

Fig. 2. log(inhibitor) vs. response curves. Dose-dependent inhibitory effects of 10 different compounds in U2OS cancer cells (A and B) and BJ human fibroblast non-cancerous cells (C and D). Compounds were tested at two time points, 24 (A and C) and 48 (B and D) hours, to assess their cytotoxic effects. Graphs were generated using GraphPad, illustrating the differential responses of the cancerous and non-cancerous cell lines across varying concentrations of the compounds.

Table 1. IC₅₀ values (μM) derived from MTT assay after treatment at two different time points (24 and 48 hours) in U2OS and BJ human fibroblast cell lines. Absorbance values were analyzed using GraphPad to calculate the half-maximal inhibitory concentration (IC₅₀). Results are presented as mean IC₅₀ values in μM, indicating the concentration required to reduce cell viability by 50%.

Compound	U2OS		BJ human fibroblast
	24 h	48 h	24 h	48 h
	μM	μM	μM	μM
1	153.1	262.2	357.2	131.6
2	380.3	390.2	653.1	358.1
3	80.2	116.7	318.2	244.1
4	88.9	94.9	248.6	188.0
5	195.5	196.1	201.2	130.8
6	140.6	375.8	720.2	134.9
7a	697.7	957.6	830.9	712.2
8	280.5	96.2	515.0	305.3
9	115.7	203.2	274.7	282.8
10	144.7	153.9	425.9	209.0

^aReference molecule.

2-Morpholinoethylamine and 4-hydroxyphenylmorpholine-functionalized cyclotriphosphazene derivatives exhibit more effective anti-cancer activity compared to other cyclotriphosphazene derivatives with other moieties. Considering their low toxicity to healthy cells compared with cancer cells, these molecules may offer a favorable balance between efficacy and safety for potential therapeutic use. It is even possible that they could be further investigated for their potential in cancer treatment.

2.2.2. Determination of population doubling times in the presence of cyclotriphosphazene derivatives

The proliferation dynamics of U2OS cells after 48 h of exposure to novel cyclotriphosphazene derivatives were determined through population doubling (PD) analyses after determining IC₁₀ values (Table 2).

Table 2. IC₁₀ values for U2OS cells treated with chemicals for 48 hours. The calculation was based on the logarithmic transformation of MTT assay results, followed by data normalization and curve fitting using the cubic spline method. The concentration corresponding to approximately 90% cell viability was identified and selected as the dose for subsequent PD experiments.

Compound	Cell survival (%)	IC10 concentration (μM L−1)
1	90.08	50
2	90.06	50
3	90.06	20
4	90.00	20
5	90.13	70
6	90.11	70
7	90.98	130
8	90.04	40
9	90.92	20
10	90.04	20

Compounds 4 and 6 exhibited markedly lower PD values, suggesting stronger cytotoxic effects (Fig. 3). These findings correlate with its low IC₅₀ values for compound 4 and significant reductions in plating efficiency, highlighting its strong antiproliferative properties. Compound 3, with one of the lowest IC₅₀ values, maintained PD values close to the normal range, suggesting that its cytotoxic effect might be reversible or less detrimental to long-term cell viability. Compounds 1, 7, and 9, with moderate to high IC₅₀, showed PD results close to the control, indicating moderate antiproliferative effects without severely compromising cell viability.

Fig. 3. Logarithmic growth curves (left) and PD rates (right) of U2OS cells treated with cyclotriphosphazene derivatives at various concentrations for 48 hours. PD counts were calculated using the formula PD = log₂(N_f/N_i)/log₂(2) and normalized to the baseline cell numbers. Cells were treated in triplicate (n = 3), and the data were analyzed with GraphPad using a two-tailed Student's t-test for statistical significance. Results are expressed as mean ± SD. *p < 0.05.

2.2.3. Evaluation of the effects of cyclotriphosphazene derivatives on colony formation

To evaluate the effects on colony formation capacity, cyclotriphosphazene derivatives were applied to U2OS cells at two different concentrations based on their IC₅₀ values. Following a 24-hour treatment, the medium was replaced with a fresh medium without any compounds. When the colonies reached an appropriate size (≥50 cells per colony), they were stained with crystal violet (Fig. 4). Then, the number of colonies was counted (Table S2†). Moreover, 4 notably reduced the plating efficiency to 21.63% at the highest concentration. Compounds 2 and 10, despite having higher IC₅₀ values (380.3 μM and 144.7 μM), also demonstrated strong dose-dependent antiproliferative activities, reducing the plating efficiency to 29.23% and 23.25%, respectively, at the highest concentrations. Compounds 1, 5, and 7 also reduced colony formation, less dramatically, suggesting moderate antiproliferative activity, while 6 showed a relatively high IC₅₀ (140.6 μM) and maintained the highest plating efficiency, indicating minimal cytotoxicity and antiproliferative effects. Additionally, compound 7 displayed the highest IC₅₀ value (697.7 μM), indicating lower cytotoxicity, but still exhibited a moderate antiproliferative effect with a reduction in plating efficiency. Overall, these results indicate that compounds 2, 4, and 10 are particularly potent in inhibiting U2OS cell proliferation, with both high cytotoxicity and strong antiproliferative effects. Compound 6, however, appears to be less effective, suggesting the potential for lower therapeutic efficacy. These findings highlight the varying potency of these cyclotriphosphazene derivatives, emphasizing the need for further investigation of their mechanisms of action and potential therapeutic applications.

Fig. 4. Clonogenic survival assay results following the treatment of U2OS cells with 10 different compounds at two different concentrations. Cells were exposed to the compounds for 24 hours, and the images were captured using a digital camera. Experiments were performed in triplicate (n = 3), and results were expressed as mean ± SD. The data were analyzed by one-way ANOVA with Tukey's multiple comparisons test with GraphPad, and the statistical significance was adjusted at *p < 0.05.

2.2.4. Effects of cyclotriphosphazene derivatives on apoptosis

Apoptosis is a form of programmed cell death and the primary mechanism for balancing cell growth and cell division to prevent uncontrolled cell division, which is associated with cancer development. Most anticancer drugs currently used in clinical therapy target intact apoptotic signaling pathways to activate cancer cell death and reduce tumor growth.⁴⁵

The impact of cyclotriphosphazene derivatives on apoptosis in U2OS cells was evaluated by measuring PARP and cleaved-PARP (c-PARP) levels following treatment at various concentrations (Fig. 5). Cisplatin used as a positive control showed a clear dose-dependent increase in c-PARP levels, indicating robust apoptosis induction. The results showed that among the cyclotriphosphazene derivatives, compound 4, which exhibited a notable reduction in colony formation at both low and high concentrations, also increased c-PARP levels to 22.8% at 50 μM. However, the absence of a dose-dependent increase in apoptosis marker (cleaved-PARP) suggests that other forms of cell death, such as necrosis, autophagy or growth inhibition, may be also involved. In contrast, most of the cyclotriphosphazene derivatives, including compounds 1, 2, 5, 6, and 9, did not show consistent dose-dependent increases in c-PARP levels, indicating a lack of potent induction of apoptosis under the conditions tested. This suggests that although these compounds may exhibit antiproliferative activity, as demonstrated in the colony formation assays, their effects on apoptosis are minimal. For instance, compound 2 reduced colony formation efficiency and PD at higher doses but did not significantly elevate c-PARP levels, indicating that its antiproliferative action may involve mechanisms other than apoptosis, such as cell cycle arrest. Similarly, compound 6, which showed significantly low plating efficiency and minimal cytotoxicity in colony assays, showed low c-PARP levels, further supporting the absence of strong apoptotic effects. Even though compound 3 resulted in a relatively low IC₅₀ compared to the other compounds, indicating higher toxicity, the observed responses did not show significant changes in colony-forming efficiency or PD rates. There was no clear dose-dependent increase in c-PARP levels, suggesting that c-PARP does not exert a strong cytostatic or pro-apoptotic effect under the tested conditions. Overall, the results showed that several cyclotriphosphazene derivatives effectively inhibited U2OS cell proliferation, as demonstrated by colony formation assays, while their apoptosis induction potential was limited and inconsistent. This suggests that other mechanisms, such as necrosis or autophagy, might contribute to cytotoxic properties and require further mechanistic studies to fully understand their therapeutic potential.

Fig. 5. Evaluation of apoptosis by measuring PARP and c-PARP levels in U2OS cells treated with novel cyclotriphosphazene derivatives. Cells were treated with varying concentrations of the compounds, and PARP and c-PARP protein levels were analyzed by immunoblotting. Cisplatin used as a positive control induced a dose-dependent increase in the c-PARP ratio to the total PARP level, confirming its apoptotic effect. Band intensity analysis was performed using ImageJ, and the data were analyzed with GraphPad using a two-way ANOVA. Experiments were performed in triplicate (n = 3), and the results were expressed as mean ± SD. *p < 0.01.

2.2.5. Investigation of the effects of cyclotriphosphazene derivatives on the cell cycle

Using U2OS cells, the effects of cyclotriphosphazene derivatives on the cell cycle were evaluated using flow cytometry after staining with 7-AAD (Fig. 6). Compound 4 at 50 μM induced a marked G₂/M phase arrest, with a substantial decrease in G₀/G₁ phase and a corresponding accumulation in G₂/M, highlighting its potential to stop cell division at this checkpoint. Similarly, 10 showed an increase in the G₂/M population at 10 μM concentration owing to its role in cell cycle disruption. Meanwhile, compounds 1, 2, and 6 had moderate effects, causing slight shifts in the G₀/G₁ and S phases without significant G₂/M accumulation. Considering the low c-PARP levels in U2OS cells treated with the compounds, the antiproliferative effects of these derivatives are primarily mediated through cell cycle arrest rather than apoptotic pathways.

Fig. 6. A) Representative histograms of the effects of novel cyclotriphosphazene derivatives on the cell cycle. U2OS cells were treated with various concentrations of compounds and examined using flow cytometry. Cell cycle analysis was performed using BD Accuri™ C6 software, and B) data were analyzed with GraphPad using two-way ANOVA. Experiments were performed in duplicate (n = 2), and the results were expressed as mean ± SD. *p < 0.05.

2.2.6. Effects on circadian rhythm

The effects of novel cyclotriphosphazene derivatives on circadian rhythm were assessed using a real-time bioluminescence assay in U2OS-pLV7-Bsd-P(Per2)-dLuc cells. In these cells, the transcription of destabilized luciferase (dLuc) is under the direct control of the Per2 promoter, and the luciferase activity was recorded for three days after the luciferase substrate was added into the culture medium following the synchronization of the circadian clock of the cells. The addition of DMSO, which was used to dissolve the compounds in the culture medium, did not affect the circadian rhythm compared to untreated cells (ESI†). The cells were then treated with different concentrations of each compound, and the bioluminescent signal from luciferase activity was monitored for over 72 hours (Fig. 7). The results showed that most compounds did not significantly alter the circadian period. However, compound 4 at 20 μM caused a notable lengthening of the period, suggesting a potential disruption of the circadian clock at this concentration. Interestingly, this effect was not dose dependent, as higher concentrations of compound 4 did not lead to further alterations in the circadian period. This lack of a clear dose–response relationship suggests that compound 4 may exert its effects using a threshold mechanism, where a certain concentration is sufficient to impact circadian regulation, but additional increases in dosage do not further enhance this effect. It is also possible that compound 4 affects specific circadian clock-related pathways that are saturated at lower concentrations, limiting its impact at higher doses. Amplitude analysis revealed significant dampening of circadian oscillations by compound 5 at 250 μM, indicating a modulatory effect on Per2 expression. It demonstrated a dose-dependent effect, with increasing concentration, resulting in progressively greater reductions in amplitude values. This suggests that compound 5 may interfere with the molecular components of the circadian oscillator, leading to an observed dampening of the rhythm. The fact that the majority of the compounds did not alter the circadian period suggests that they are relatively safe in terms of circadian rhythm stability, particularly at lower concentrations. This stability could be advantageous for therapeutic applications where maintaining normal circadian function is crucial. However, the non-dose-dependent effect observed with compound 4 and the dose-dependent effect of compound 5 require further investigation to better understand the molecular pathways involved and to assess the safety and efficacy of these compounds in modulating circadian rhythms.

Fig. 7. Real-time bioluminescence graphs representing circadian oscillations in U2OS cells treated with various concentrations of novel cyclotriphosphazene derivatives. Graphs representing the raw data were illustrated using GraphPad. Period data were analyzed using the Kruskal–Wallis test, revealing no significant differences between the groups (p = 0.827), and amplitude data were analyzed by one-way ANOVA with Dunnett's post hoc test. Period and amplitude values are represented as mean ± SD (n = 3). Statistical significance was set at *p < 0.05 and ***p < 0.0001.

3. Conclusion

Within the scope of this study, the biological activities of cyclotriphosphazene-based compounds were systematically investigated for the first time. The target molecules were synthesized using nucleophilic substitution and click reactions, and their chemical structures were characterized using spectroscopic methods (mass, ³¹P, ¹H, and ¹³C NMR). Compounds 3 and 4 exhibited the lowest IC₅₀ values in U2OS cancer cells, while their toxicities were significantly lower in normal BJ cells. Being more toxic to cancerous cells and less toxic to normal cells is a good indication that these compounds are good anticancer candidates. In this respect, one possible explanation for this selectivity is that the compounds target signaling pathways that are active in cancer cells but not in normal cells. Additionally, other factors, such as functional groups, polarity, and lipophilicity, may contribute to the differential uptake by normal and cancer cells. Our study revealed that, among the cyclotriphosphazene derivatives tested, compounds 4 and 6 exhibited the most pronounced antiproliferative effects, as evidenced by their significantly low PD values, correlating with their low IC₅₀ values and substantial reductions in plating efficiency. Compound 3, despite its low IC₅₀, demonstrated PD values close to normal, indicating potentially reversible cytotoxic effects or a milder long-term impact on cell viability. As the compounds become more toxic in BJ fibroblasts, we can assume that there is a cumulative or delayed cytotoxic effect rather than an immediate toxic response. In contrast to U2OS cancer cells, BJ fibroblasts are non-transformed cells with considerably slower cell division rates and are therefore less sensitive to short-term drug exposure. However, as the incubation period is prolonged, even in slowly dividing cells, damage may occur owing to the intracellular accumulation of active compounds and their persistence in the culture medium. Moderately stable and biologically active compounds can accumulate intracellularly over time, leading to stronger toxic effects. Furthermore, BJ fibroblasts may initially delay toxicity by activating stress and survival mechanisms; however, prolonged exposure beyond 48 hours may trigger secondary toxic effects, such as mitochondrial dysfunction or oxidative damage. We did not check the toxicity beyond 48 hours as the main concern was to compare the toxicity between normal and cancerous cells. Although compound 3 showed a modest increase in c-PARP levels at high concentrations, the lack of a clear dose–response relationship suggests that apoptosis may not be the dominant mechanism behind its cytotoxic effects. Cyclotriphosphazene derivatives 4 and 10 exhibited notable antiproliferative effects in U2OS cells potentially through interference with cell cycle progression, as suggested by G₂/M phase arrest observed in cell cycle analyses. However, because apoptosis induction was minimal and other cell death mechanisms were not investigated in this study, it is premature to conclude that their antiproliferative effects are solely owing to cytostatic activity. Further studies are required to explore additional mechanisms of action and evaluate their broader therapeutic potential in targeting cell division processes. Furthermore, although most of the derivatives had minimal impact on circadian rhythms, suggesting safety in maintaining circadian function, compounds 4 and 5 displayed noteworthy modulatory effects on circadian oscillations, highlighting their potential as circadian rhythm modulators. Overall, we cannot present a pattern between the modifications and their effects on cellular events, such as toxicity. We believe that the number of compounds required to reach this conclusion is insufficient. However, the effect of the compounds is not limited to modifications, but the three-dimensional overall structure, including the core, might also have a cumulative effect, which makes it difficult to observe a pattern with such a few compounds. Further modification of the compound in follow-up studies may help in this respect. Our findings collectively suggest that the cyclotriphosphazene derivatives exert their antiproliferative effects through diverse mechanisms, warranting further investigation into their therapeutic potential and molecular targets.

4. Experimental section

4.1. Experimental instruments

All the precursor chemical reagents were procured from commercial suppliers. Analytical thin-layer chromatography (TLC) was performed on silica gel plates. Column chromatography was performed on silica gel (200–400 mesh). Mass spectra were recorded using a Bruker Daltonics Microflex mass spectrometer and Agilent LC-MS/MS. Varian 400 MHz was used for NMR spectra (³¹P, ¹H and ¹³C NMR). Cells were maintained in the NUVE EC 160 model incubator, and cell studies were carried out in the NUVE MN120 Class II laminar flow cabinet. The circadian rhythm was recorded by applying a Synergy H1 multimode microplate reader.

4.2. Chemistry

Synthesis Methods

Compounds 1 and 7 were synthesized according to the literature procedure.^35,36 Compound 2–6 and compound 8–10 are new and designed within the scope of this study.

Synthesis of compound 1

Hexachlorocyclotriphosphazene (6.0 g, 17.3 mmol) was placed into a two-necked round bottom flask, dissolved in 15 mL dry THF under Ar atm and placed in an ice bath. NaH (2.4 g, 58.7 mmol, 60%) was washed with n-hexane, dissolved 15 mL THF and added to the reaction medium. Triethyleneglycol monomethyl ether (8.7 mL, 55.2 mmol) was dissolved in 15 mL dry THF and added dropwise to the medium with the aid of a dropping funnel. The resulting mixture was stirred at room temperature for 5 hours under Ar atm and controlled using a TLC silica plate. After the reaction was completed, the mixture was filtered through a G4 filter to remove the NaCl salts formed. The solvent of the filtrate was evaporated using a rotary evaporator. The reaction mixture was purified by silica gel column chromatography using THF : CH₂Cl₂ (2 : 1) as a mobile phase to produce compound 1 (1.6 g, 13%, oily). MALDI TOF (m/z) calc. for C₂₁H₄₅Cl₃N₃O₁₂P₃: 730.871, found: 730.868 [M]⁺, 753.071 [M + Na]⁺ (Fig. S1†). ³¹P NMR (162 MHz, CDCl₃) δ_P 21.73 (s) ppm (Fig. S2†). ¹H NMR (400 MHz, CDCl₃) δ_H 4.28–4.22 (m, 6H, –POCH₂), 3.75–3.72 (m, 6H, –OCH₂), 3.67–3.65 (m, 6H, –OCH₂), 3.64–3.62 (m, 12H, –OCH₂), 3.53 (dd, J = 5.6, 2.9 Hz, 6H, –OCH₂), 3.36 (s, 9H, –OCH₃) ppm (Fig. S3†). ¹³C NMR (101 MHz, CDCl₃) δ_C 71.87, 70.75, 70.64, 70.54, 69.33, 67.33, 59.01 ppm (Fig. S4†).

Synthesis of compound 2

Morpholine (107 μL, 1.23 mmol) was dissolved in 10 mL dry THF, and triethylamine (328 μL, 2.35 mmol) was added to the reaction flask. Compound 1 (300 mg, 0.41 mmol) dissolved in 10 mL THF was added dropwise to the reaction medium. The resulting mixture was stirred at 55 °C for 4 days with the aid of a magnetic stirrer. The reaction was controlled by applying a TLC silica plate. The reaction mixture was filtered through a G4 filter to remove insoluble impurities. The solvent was evaporated, and the resulting mixture was subjected to silica gel column chromatography using THF as an eluent to produce compound 2 (158 mg, 44%, oily). MALDI TOF (m/z) calc. for C₃₃H₆₉N₆O₁₅P₃: 882.863, found: 882.403 [M]⁺ (Fig. S5†). ³¹P NMR (162 MHz, CDCl₃) δ_P 21.30 (s) ppm (Fig. S6†). ¹H NMR (400 MHz, CDCl₃) δ_H 3.95–3.88 (m, 6H, –POCH₂), 3.63–3.55 (m, 32H, –OCH₂), 3.48 (dd, J = 5.7, 3.5 Hz, 6H, –OCH₂), 3.32 (s, 9H, –OCH₃), 3.02 (broad, 12H, –NCH₂) ppm (Fig. S7†). ¹³C NMR (101 MHz, CDCl₃) δ_C 71.83, 70.53, 70.52, 70.43, 70.17, 66.83, 64.09, 58.94, 44.45 ppm (Fig. S8†).

Synthesis of compound 3

2-Morpholinoethylamine (267 μL, 2.05 mmol) was dissolved in 10 mL dry THF, and triethylamine (183 μL, 1.31 mmol) was added to the reaction flask. Compound 1 (300 mg, 0.41 mmol) dissolved in 10 mL THF was added dropwise to the reaction medium. The resulting mixture was stirred at ambient temperature for 2 days with the aid of the magnetic stirrer. The reaction was controlled by applying a TLC silica plate. The reaction mixture was filtered using a G4 filter to remove insoluble impurities. The solvent was evaporated, and the resulting mixture was subjected to silica gel column chromatography using THF : methanol (5 : 2) as an eluent to produce compound 3 (55 mg, 13%, oily). MALDI TOF (m/z) calc. for C₃₉H₈₄N₉O₁₅P₃: 1012.070, found: 1012.052 [M]⁺ (Fig. S9†). ³¹P NMR (162 MHz, CDCl₃) δ_P 21.76 (s) ppm (Fig. S10†). ¹H NMR (400 MHz, CDCl₃) δ_H 3.95–3.91 (m, 6H, –POCH₂), 3.68–3.53 (m, 36H, –OCH₂), 3.47 (dd, J = 5.7, 3.5 Hz, 6H, –OCH₂), 3.30 (s, 9H, –OCH₃), 2.97–2.91 (m, 3H, –NH), 2.44–2.40 (m, 12H, –NCH₂) ppm (Fig. S11†). ¹³C NMR (101 MHz, CDCl₃) δ_C 71.93, 70.61, 70.55, 70.53, 70.41, 66.81, 63.92, 59.02, 58.91, 53.30, 37.07 ppm (Fig. S12†).

Synthesis of compound 4

4-(4-Hydroxyphenyl)morpholine (235 mg, 1.31 mmol) was dissolved in 5 mL dry THF into a two-necked round bottom flask, and NaH (52 mg, 1.31 mmol, 60%) was washed with n-hexane, dissolved 10 mL THF and added to the reaction medium. Compound 1 (300 mg, 0.41 mmol) was dissolved in 15 mL dry THF and added dropwise to the medium with the aid of a dropping funnel. The resulting mixture was stirred at 60 °C for 3 days under Ar atm and controlled using a TLC silica plate. After the reaction was completed, the mixture was filtered through a G4 filter to remove the NaCl salts formed. The solvent of the filtrate was evaporated using a rotary evaporator. The reaction mixture was purified by silica gel column chromatography using n-hexane : THF (1 : 4) as a mobile phase to produce compound 4 (100 mg, 21%, oily). MALDI TOF (m/z) calc. for C₅₁H₈₁N₆O₁₈P₃: 1158.482, found: 1156.539 [M–2H]⁺, 1179.22 [M–2H + Na]⁺ (Fig. S13†). ³¹P NMR (162 MHz, CDCl₃) δ_P 13.63 (s) ppm (Fig. S14†). ¹H NMR (400 MHz, CDCl₃) δ_H 5.51–5.49 (m, 2H, Ar–CH), 5.45–5.42 (m, 2H, Ar–CH), 5.34–5.31 (m, 8H, Ar–CH), 3.75–3.70 (m, 36H, –POCH₂ + –OCH₂), 3.70–3.50 (m, 12H, –OCH₂), 3.43 (s, 9H, –OCH₃), 1.84–1.73 (m, 12H, –NCH₂) ppm (Fig. S15†). ¹³C NMR (101 MHz, CDCl₃) δ_C 107.56, 107.19, 106.69, 100.80, 100.69, 97.93, 71.60, 70.21, 70.10, 68.83, 67.57, 67.35, 66.98, 61.79, 59.88, 58.64, 49.89 ppm (Fig. S16†).

Synthesis of compound 5

Thiomorpholine (123 μL, 1.23 mmol) was dissolved in 10 mL dry THF, and triethylamine (200 μL, 1.43 mmol) was added to the reaction flask. Compound 1 (300 mg, 0.41 mmol) dissolved in 10 mL THF was added dropwise to the reaction medium. The resulting mixture was stirred at ambient temperature for 5 days with the aid of a magnetic stirrer. The reaction was controlled by applying a TLC silica plate. The reaction mixture was filtered through a G4 filter to remove insoluble impurities. The solvent was evaporated, and the resulting mixture was subjected to silica gel column chromatography using THF : CH₂Cl₂ (1 : 1) as an eluent to give compound 5 (26 mg, 7%, oily). MALDI TOF (m/z) calc. for C₃₃H₆₉N₆O₁₂P₃S₃: 931.046, found: 931.020 [M]⁺ (Fig. S17†). ³¹P NMR (162 MHz, CDCl₃) δ_P 19.75 (s) ppm (Fig. S18†). ¹H NMR (400 MHz, CDCl₃) δ_H 3.89–3.80 (m, 6H, –POCH₂), 3.64–3.55 (m, 24H, –OCH₂), 3.50–3.46 (m, 6H, –OCH₂), 3.31 (s, 9H, –OCH₃), 2.99–2.83 (m, 12H, –NCH₂), 2.82–2.71 (m, 12H, –SCH₂) ppm (Fig. S19†). ¹³C NMR (101 MHz, CDCl₃) δ_C 72.12, 70.83, 70.76, 66.93, 66.36, 59.23, 54.37, 53.04, 46.53 ppm (Fig. S20†).

Synthesis of compound 6

4-(2-Aminoethyl)thiomorpholine-1,1-dioxide (144 mg, 0.81 mmol) was dissolved in 10 mL dry THF, and triethylamine (145 μL, 1.04 mmol) was added to the reaction flask. Compound 1 (200 mg, 0.27 mmol) dissolved in 10 mL THF was added dropwise to the reaction medium. The resulting mixture was stirred at 80 °C for 18 h with the aid of a magnetic stirrer. The reaction was controlled by applying a TLC silica plate. The solvent was evaporated using a rotary evaporator. The crude product was dissolved in DCM and washed twice with water. The organic layer was collected and dried over Na₂SO₄. Then, the solvent was evaporated to give compound 6 (250 mg, 79%, oily). MALDI TOF (m/z) calc. for C₃₉H₈₄N₉O₁₈P₃S₃: 1156.247, found: 1156.470 [M]⁺ (Fig. S21†). ³¹P NMR (162 MHz, CDCl₃) δ_P 22.63 (s) ppm (Fig. S22†). ¹H NMR (400 MHz, CDCl₃) δ_H 3.86–3.76 (m, 6H, –POCH₂), 3.60–3.47 (m, 30H, –OCH₂), 3.46–3.36 (m, 12H, –PNCH₂ + –NCH₂), 3.20 (s, 9H, –OCH₃), 2.95–2.91 (m, 12H, –NCH₂), 2.89–2.81 (m, 12H, –SCH₂) ppm (Fig. S23†). ¹³C NMR (101 MHz, CDCl₃) δ_C 71.70, 70.54, 70.50, 70.44, 70.26, 67.60, 6415, 59.01, 51.11, 50.73, 38.02 ppm (Fig. S24†).

Synthesis of compound 7

Compound 1 (490 mg, 0.67 mmol) was dissolved in 10 mL dry THF into a two-necked round bottom flask and placed into an ice bath. NaH (72 mg, 3.02 mmol, 60%) was washed with n-hexane, dissolved in 10 mL THF and added to the reaction medium. 2-Azidoethanol (270 mg, 3.10 mmol) was dissolved in 5 mL dry THF and added dropwise to the medium with the aid of a dropping funnel. The resulting mixture was stirred at 60 °C for 4 days under Ar atm and controlled using a TLC silica plate. After the reaction was completed, the mixture was filtered through a G4 filter to remove the NaCl salts formed. The solvent of the filtrate was evaporated using a rotary evaporator. The reaction mixture was purified by silica gel column chromatography using n-hexane: THF : CH₂Cl₂ (1 : 2 : 1) as a mobile phase to give compound 7 (100 mg, 17%, oily). ³¹P NMR (162 MHz, CDCl₃) δ_P 17.69 (s) ppm (Fig. S25†). ¹H NMR (400 MHz, CDCl₃) δ_H 4.14–4.05 (m, 12H, –POCH₂), 3.71–3.68 (m, 6H, –OCH₂), 3.67–3.61 (m, 18H, –OCH₂), 3.53 (dd, J = 5.7, 3.4 Hz, 6H, –OCH₂), 3.46 (t, J = 5.0 Hz, 6H, –CH₂N₃), 3.37 (s, 9H, –OCH₃) ppm (Fig. S26†). ¹³C NMR (101 MHz, CDCl₃) δ_C 72.50, 71.89, 70.57, 70.50, 67.61, 65.35, 64.84, 61.68, 59.01 ppm (Fig. S27†).

Synthesis of compound 8

Compound 7 (80 mg, 90.6 μmol) was dissolved in 2 mL dry THF in a round bottom flask. 4-Propargylmorpholine (33 mg, 266.8 μmol), CuI (43 mg, 335.3 μmol) and DIPEA (86 μL, 335.3 μmol) were added to the reaction medium. The resulting mixture was stirred at ambient temperature for 5 days and controlled using a TLC silica plate. After the reaction was completed, the mixture was filtered to remove insoluble impurities. The solvent of the filtrate was evaporated using a rotary evaporator. The crude product was dissolved in CH₂Cl₂ and washed with water. The organic phase was collected and dried over Na₂SO₄. The solvent was evaporated to give compound 8 (90 mg, 78%, oily). MALDI TOF (m/z) calc. for C₄₈H₉₀N₁₅O₁₈P₃: 1258.256, found: 1258.033 [M]⁺ (Fig. S28†). ³¹P NMR (162 MHz, CDCl₃) δ_P 18.91 (s) ppm (Fig. S29†). ¹H NMR (400 MHz, CDCl₃) δ_H 7.65 (s, 3H, –NCH), 4.30–4.25 (m, 6H, –POCH₂), 4.19–4.13 (m, 6H, –POCH₂), 4.12–3.98 (m, 6H, –OCH₂), 3.68–3.58 (m, 24H, –OCH₂), 3.55–3.48 (m, 12H, –NCH₂), 3.31 (s, 9H, –OCH₃), 2.52–2.50 (m, 12H, –OCH₂), 2.45–2.41 (m, 12H, –NCH₂) ppm (Fig. S30†). ¹³C NMR (101 MHz, CDCl₃) δ_C 176.81, 176.79, 71.49, 70.83, 69.54, 69.42, 69.25, 68.85, 65.75, 60.63, 57.97, 51.09, 46.73, 28.65 ppm (Fig. S31†).

Synthesis of compound 9

Compound 7 (55 mg, 62.3 μmol) was dissolved in 2 mL dry THF in a round bottom flask. 4-Propargylthiomorpholine (28 mg, 119.4 μmol), CuI (42 mg, 220.5 μmol) and DIPEA (38 μL, 220.5 μmol) were added to the reaction medium. The resulting mixture was stirred at ambient temperature for 5 days and controlled using a TLC silica plate. After the reaction was completed, the mixture was filtered to remove insoluble impurities. The solvent of the filtrate was evaporated using a rotary evaporator. The crude product was dissolved in CH₂Cl₂ and washed with water. The organic phase was collected and dried over Na₂SO₄. The solvent was evaporated to give compound 9 (35 mg, 43%, oily). MALDI TOF (m/z) calc. for C₄₈H₉₀N₁₅O₁₅P₃S₃: 1305.512, found: 1305.826 [M]⁺ (Fig. S32†). ³¹P NMR (162 MHz, CDCl₃) δ_P 17.43 (s) ppm (Fig. S33†). ¹H NMR (400 MHz, CDCl₃) δ_H 7.67 (s, 3H, –NCH), 4.60–4.50 (m, 6H, –POCH₂), 4.17–4.10 (m, 6H, –POCH₂), 3.94–3.84 (m, 6H, –NCH₂), 3.60–3.58 (m, 6H, –OCH₂), 3.57–3.53 (m, 18H, –OCH₂), 3.47 (dd, J = 5.4, 3.3 Hz, 6H, –OCH₂), 3.31 (s, 6H, –NCH₂), 3.29 (s, 9H, –OCH₃), 2.76–2.71 (m, 12H, –NCH₂), 2.66–2.61 (m, 12H, –SCH₂) ppm (Fig. S34†). ¹³C NMR (101 MHz, CDCl₃) δ_C 177.81, 177.76, 72.54, 71.83, 70.50, 70.43, 70.23, 68.51, 61.59, 58.95, 53.66, 48.01, 27.77, 22.14 ppm (Fig. S35†).

Synthesis of compound 10

Compound 7 (70 mg, 79.3 μmol) was dissolved in 2 mL dry THF in a round bottom flask. 4-Propargylthiomorpholine-1,1-dioxide (40 mg, 230.0 μmol), CuI (52 mg, 277.5 μmol) and DIPEA (48 μL, 277.5 μmol) were added to the reaction medium. The resulting mixture was stirred at ambient temperature for 5 days and controlled using a TLC silica plate. After the reaction was completed, the mixture was filtered to remove insoluble impurities. The solvent of the filtrate was evaporated using a rotary evaporator. The crude product was dissolved in CH₂Cl₂ and washed with water. The organic phase was collected and dried over Na₂SO₄. The solvent was evaporated to give compound 10 (30 mg, 27%, oily). MALDI TOF (m/z) calc. for C₄₈H₉₀N₁₅O₂₁P₃S₃: 1402.433, found: 1402.572 [M]⁺, 1425.484 [M + Na]⁺ (Fig. S36†). ³¹P NMR (162 MHz, CDCl₃) δ_P 17.95 (s) ppm (Fig. S37†). ¹H NMR (400 MHz, CDCl₃) δ_H 7.70 (s, 3H, –NCH), 4.32 (m, 18H, –POCH₂ + –NCH₂), 3.70–3.59 (m, 24H, –OCH₂), 3.52–3.50 (m, 6H, –OCH₂), 3.34–3.31 (m, 6H, –NCH₂), 3.07 (s, 9H, –OCH₃), 2.46 (t, J = 7.9 Hz, 12H, –NCH₂), 2.25–2.22 (m, 12H, –SCH₂) ppm (Fig. S38†). ¹³C NMR (101 MHz, CDCl₃) δ_C 177.91, 177.86, 72.51, 71.82, 70.49, 70.45, 70.40, 68.55, 58.96, 58.92, 51.44, 50.11, 27.78, 22.14 ppm (Fig. S39†).

4.3. In vitro studies

4.3.1. Preparation of cyclotriphosphazene derivatives

All synthesized cyclotriphosphazene derivatives were dissolved in DMSO at a final 100 mM main stock concentration.

4.3.2. Cell culture

The U2OS cell line (ATCC # HTB-96) and BJ cells (ATCC # CRL-2522) were purchased from the American Type Culture Collection (ATCC) (Rockville, MD) and cultured in high-glucose Dulbecco's modified Eagle medium (DMEM) (ThermoFisher Scientific, Gibco, cat. no: 11965092). BJ human fibroblast cells were cultured in low-glucose Dulbecco's modified Eagle medium (DMEM) (ThermoFisher Scientific, Gibco, cat. no: 11885092) supplemented with 10% fatal bovine serum (FBS) (ThermoFisher Scientific, Gibco, cat. no: 10270106), 1% penicillin–streptomycin (ThermoFisher Scientific, Gibco, cat. no: 15140122) at the final 100 U mL⁻¹ concentration and 1% non-essential amino acids. Cells were maintained at 37 °C humidified incubator containing 5% CO₂.

4.3.3. Viability assay

5 × 10³ U2OS cells and 7.5 × 10³ BJ human fibroblast cells per well were seeded into the 96-well tissue culture plates in duplicate, and they were then incubated overnight. The next day, 0.5% DMSO and each cyclotriphosphazene-based compound were applied to each cell line at determined concentrations ranging from 0 to 1 mM, and the cells were incubated for 24 and 48 h. After incubation, the medium in each well was removed, and the cells were washed with 1× phosphate-buffered saline (PBS). Thiazolyl blue tetrazolium bromide (MTT) solution (5 mg mL⁻¹ in 1× PBS) was mixed with a medium at a ratio of 1 : 10, and 100 μL of MTT–DMEM mix was added to each well. Plates were incubated at a 37 °C incubator for 4 h. MTT mix was aspirated, and 100 μL of DMSO was pipetted to each well. Plates were incubated on a shaker at room temperature (RT) for 5–15 min. The optical density was measured at 570 nm.

4.3.4. Clonogenic survival assay

5 × 10² U2OS cells per well were seeded into the 6-well tissue culture plates and incubated overnight. The next day, 0.5% DMSO and each cyclotriphosphazene-based compound were applied to cells at certain concentrations determined according to IC₅₀ values. 24 hours after treatment with the compounds and DMSO, the medium was renewed to be chemical-free. The cells were then incubated for approximately 10–14 days, with the medium being changed every 2–3 days until the colonies grew enough to be counted (≤50 cells). The medium was removed, and the cells were washed with 1× PBS. Colonies were fixed with 1 mL of fixation solution (4% formaldehyde in 1× PBS) at RT for 20 min and dyed with 1 mL of staining solution (1 : 1000 “mass/volume” crystal violet in 1× PBS) at RT for 15 min. The staining solution was removed. The sink was filled with water, and the plates were immersed in water until the background was clear. Visible colonies were manually counted on digital images, and the plating efficiency and survival fractions were calculated.

4.3.5. Measurement of population doubling (PD)

5 × 10⁵ U2OS cells per well were seeded into the 6-well tissue culture plates and incubated overnight. After 16 h, the cells were treated with compounds or 0.5% DMSO for 48 h. Simultaneously, to determine the number of cells adhering to the dishes after cell seeding, U2OS cells that were not treated with any compound or carrier 16 h after seeding were removed with trypsin and counted by hemocytometer (N_i baseline cell number). Cells were removed with trypsin after 48 h of compound treatment and counted using a hemacytometer in the same way. PD value was calculated using the following formula: PD = log₂ (N_f/N_i)/log₂ (2), where N_f is the number of cells at the end of 48 h and N_i represents the initial number of cells at 0 h.

4.3.6. Determination of apoptosis rate

2.5 × 10⁵ U2OS cells per well were seeded into the 6-well tissue culture plates, and they were incubated overnight. The cells were exposed to several concentrations of compounds and 0.5% DMSO for 24. Cells were collected with their medium, spun down, washed with 1× PBS and then lysed with 1× RIPA lysis buffer containing protease inhibitor (Sigma-Aldrich, cat. no: S8830-2TAB). Total protein was measured by applying a BCA protein assay kit (Thermo Scientific, cat. no: 23227). Protein samples were loaded into SDS-PAGE to determine the apoptotic effects of the compounds by anti-cleaved PARP antibody.

4.3.7. Immunoblotting

Protein samples were separated by SDS-PAGE and transferred to a nitrocellulose membrane (Bio-rad, Hercules, CA, cat. no: 162-0115). Membranes were blocked with 5% skimmed-milk solution in 1× TBS-Tween-20 and incubated with primary antibodies; anti-PARP1 (cleaved Asp214) antibody (Thermo Fisher Scientific, Invitrogen, PA5-77850) or anti-Nanog antibody (Abcam, ab109250) overnight at +4 °C. The next day, membranes were washed with 1× TBS-T for 5, 15, 5 and 5 min and incubated with secondary antibodies (anti-actin, Cell Signaling Technology, cat. no: 4967, anti-mouse HRP Cell Signaling Technology, cat. no: 7076S and, anti-rabbit HRP Thermo Fisher Scientific cat. no: 31460). GangNam-Stain Prestained Protein Ladder (Intron Biotechnology cat. no: S24052) was used as a protein marker in SDS-PAGE. The membranes were washed with 1× TBS-T for 5, 15, 5 and 5 min to remove unbound antibodies. A WesternBright Sirius HRP substrate (Advansta, cat. no: K-12043-D20) was used to develop chemiluminescence, and the membranes were visualized using the ChemiDoc XRS + (Bio-Rad) instrument.

4.3.8. Flow Cytometry

5 × 10⁵ U2OS cells per well were seeded into the 6-well tissue culture plates, and they were incubated overnight. After that, several concentrations of each compound were treated, and the cells were incubated at 37 °C in a 5% CO₂ incubator for 48 h. After the treatment, the media was removed, and the cells were washed with 1× PBS. Afterwards, the cells were detached with trypsin and collected by centrifugation. The cells were fixed with 1 mL of ice-cold 70% ethanol overnight at −20 °C. The next day, the fixation solution was removed, and the cells were washed with 1× PBS. The cells were then dyed with 1 mL of 7-AAD solution (0.1% Triton-X-100, 15 μg mL⁻¹ RNase A, 0.5 μg mL⁻¹ of 7-AAD (Thermo Fisher Scientific, cat. no: 00-6993-50) in 1× PBS) and incubated at 37 °C for 1.5 h in the dark. After incubation, the dye was removed, and the cells were dissolved in 1 mL of 1× PBS. The cells were then passed through the BD Accuri C6 flow cytometry (BD Bioscience, Franklin Lakes, NJ, USA) with a cell count setting of 1 × 10⁴ cells in the same population, and the data were analyzed using BD Accuri™ C6 software.

4.3.9. Real-time bioluminescence assay

To follow circadian rhythm, U2OS cells expressing destabilized luciferase enzyme under the Per2 promoter were prepared using pLV7-Bsd-P(Per2)-dLuc plasmid, a kind gift of Andrew C Liu, following a previously published protocol.⁴⁶ 5 × 10⁴ U2OS-pLV7-Bsd-P(Per2)-dLuc cells were seeded into a 96-well whit opaque culture plate. Cells were reset with DMEM containing 0.1 μM dexamethasone for 2 hours. 0.1 mM beetle luciferin (Promega, cat. no: E1602) was freshly added to a recording medium (DMEM powder (Sigma cat. no: D-2902), sodium bicarbonate (Sigma cat. no: S5761), d-(+)-glucose powder (Sigma cat. no: G7021), 1 M HEPES buffer (Gibco cat. no: 15140-122), and 1% pen/strep, 5% FBS, 1% NEAA, Sterile Milli-Q water). Several concentrations of each compound were added into this medium containing luciferin. The dexamethasone was aspirated, and the cells were washed with 1× PBS. Finally, the recording medium with compounds was added into each well, and the plate was placed on Synergy™ H1 Hybrid Multi-Mode Microplate Reader. Device settings were adjusted according to the manufacturer's instructions. Data recording continued for 3 days and was analyzed with GraphPad Prism (Version 8.0.1, GraphPad Software Inc).

4.3.10. Statistical analysis

Data analysis of MTT cell viability assay, clonogenic survival assay and PD time estimation was carried out on GraphPad Prism. SD was calculated by averaging three independent experiments. The statistical analysis performed with IC₅₀ values was calculated as the log₂ normalized dose–response variable slope for the MTT toxicity assay. Clonogenic survival assay results were analyzed using a one-way ANOVA test. PD counts were normalized to the log₂ of the baseline cell numbers, and statistical significance was assessed using a two-tailed Student's t-test. The immunoblotting results of PARP and C-PARP were analyzed with ImageJ and GraphPad with a two-way ANOVA test, respectively. Cell cycle analyses were investigated with BD Accuri™ C6 software, and the data were then analyzed using GraphPad with a two-way ANOVA test and Dunnett's multiple comparison test. Real-time bioluminescence data of U2OS-pLV7-Bsd-P(Per2)-dLuc were visualized using GraphPad. Amplitude and period data were analyzed using one-way ANOVA with Dunnett's multiple comparison test and the Kruskal–Wallis test owing to non-normal distribution (Shapiro–Wilk p-values < 0.05), revealing no significant differences between groups (p = 0.827), respectively.

Data availability

The mass and NMR spectra of the compounds and the comparison of real-time bioluminescence graphs representing circadian oscillations in untreated (UT) and 0.5% DMSO-treated U2OS cells are presented in the ESI† file. XLOGP3 values are listed in Table S1.† Counts of colonies are listed in Table S2.†

Author contributions

Elif Yıldız-Gül: validation, investigation, data curation, conceptualization, visualization, writing – original draft. Büşra Tiryaki: validation, investigation, data curation, conceptualization, visualization, methodology, writing – original draft. Buse Köse: validation, investigation, data curation, conceptualization, visualization. Nuri Öztürk: validation, formal analysis, data curation, writing – original draft, writing – review & editing. Elif Okutan: validation, investigation, data curation, conceptualization, visualization, methodology, writing – original draft. Burcu Dedeoğlu: validation, investigation. Esra Tanrıverdi-Eçik: validation, project administration, resources, funding acquisition, formal analysis, data curation, writing – original draft, writing – review & editing.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.