Introduction

Colorectal cancer (CRC) is currently the most common cause of malignancy, and the third leading cause of cancer-related deaths around the world. By 2040, 3.2 million additional cases and 1.6 million deaths from CRC malignancies are expected [1]. Current treatments for CRC usually involve radiation or surgical removal of the tumor followed by adjuvant chemotherapy with 5-Fluorouracil (5-FU) and (5-FU)-based combinations such as 5-FU/folinic acid/irinotecan (FOLFIRI), 5-FU/folinic acid/irinotecan/oxaliplatin (FOLFIRINOX), as well as irinotecan/capecitabine (XELIRI). Despite these chemotherapeutics have proven to be effective in the treatment of CRC, high doses, low tumor-specific selectivity, drug resistance and chronic side-effects involving severe gastrointestinal, neurological disorders, and myelosuppression, leads to the therapy discontinuation [2,3,4]. Because of these facts, the urgent need is to create new, safer, and more efficient pharmaceutical agents for chemoprevention of CRC. As a result of these efforts, in the last decade, two bioactive heterocyclic scaffolds (quinazolinone-based and melatonin) have attracted attention because their encouraging cytotoxic activity, hence these nuclei and its conjugates could provide promising prototypes for the development of improved CRC drugs.

The 2,3-dihydroquinazolinone and its derivatives not only have emerging as “privileged scaffolds” in medicinal chemistry, but currently a panel of marketed 2,3-dihydroquinazolinone-based drugs are prescribed for the supportive therapy of various diseases, including cancer (Fig. 1A) [5]. Of those, nolatrexed and raltitrexed are two potent recognized marketed drugs with antiproliferative and tumor inhibiting activity in liver cancer, as well as adjuvants in combination with the 5-FU+ leucovorin regime in the management of patients with metastatic colorectal cancer (CRC). Additionally, ispinesib, have been proven to be effective in the second-line treatment of patients with colorectal cancer, after the successful completion of phase II oncology trial [6]. Moreover, diverse natural products incorporating the 2,3-dihydroquinazolinone moiety have been reported, highlighting tryptanthrin (TR), an alkaloid isolated from the roots and leaves of the Isatis taxa, which has attracted increasing attention for its antitumor effects in human hepatocellular/breast adenocarcinoma cells in both in vitro and in vivo models mainly via regulating tumor inflammatory microenvironment, and the MAPK signaling pathway [7, 8]. In this case, (Fig. 1B), it was recently reported mostotrin, a new water-soluble and less toxic tryptanthrin derivative which exhibited 5–10 times higher cytotoxicity than TR against colon cancer cells (HCT‑116), a human breast carcinoma cell line (МСF‑7) and the human lymphoblastoid cell line (K‑562) [9]. In addition, halofuginone (HF), a small molecule alkaloid-derived, has received special attention for effectively suppresses tumor progression and metastasis of 5-FU-resistant human colorectal cancer cells [10]. Besides these marketed drugs and naturally occurring alkaloids, a large number of quinazolinone-derived compounds has been widely reported for diverse in vitro and in vivo pharmacological and medicinal properties including antihypertensive, insomnia, antihypertonic diuretic, antihistamine, antipyretic, vasodilating, analgesic, antidepressant and antitumor activities [11, 12]. In addition, regarding the activity against potentially malignant disorders, 2,3-dihydroquinazolinone have exhibited a powerful antiproliferative, antimetastatic, and anti-angiogenic response in both in vitro and in vivo models, particularly in human colorectal carcinomas (Fig. 1B). Thus, for example, Ha and colleagues showed that MJ-33 induces autophagy‑associated apoptosis via AKT/mTOR in colorectal cancer cells resistant to 5FU [13], while the bis-quinazolinone Cpd 5 was designed as a potent selective human tankyrase inhibitor for CRC [14]. In the same line, there were also reported QHMEM and a series of N3-substituted quinazolinone derivatives which suppress the proliferation of human colon cancer HCT-116 cells through induction of apoptosis [15, 16].

Fig. 1
figure 1

A 2,3-dihydroquinazolinone-related cancer drugs. B Naturally occurring alkaloids and analogs incorporating the 2,3-dihydroquinazolinone core with potent antiproliferative and tumor inhibiting activity against CRC. C FDA approved anticancer agents bearing MLT

Full size image

Mortazavi and colleagues designed and synthesized a series of quinazolinone hydrazide triazole derivatives, besides, they evaluated their potential against a human lung squamous cell carcinoma cell line (EBC-1), adenocarcinoma human alveolar basal epithelial cells (A549), a human colorectal adenocarcinoma cell line (HT29) and a human glioblastoma cell line (U87MG). These authors found that quinazolinone derivatives induced antiproliferative activity and apoptosis in these cancer cells, highlighting the compound CM9 for possessing the ability to inhibit different protein kinases involved in carcinogenesis [17].

On the other hand, melatonin (5-methoxy-N-acetyltryptamine, MLT) an endocrine hormone released by the pineal gland in the brain, have raised the attention in recent years due to their therapeutic biological activities, especially in preventing, treating, and delaying tumor development. Concerning cancer, exogenous MLT can inhibit several human malignancies in vitro and in vivo via diverse oncostatic mechanisms that include cell cycle arrest, promotion of cell apoptosis, reduction in tumor growth, antioxidant activity and inhibition of angiogenesis and metastases [18, 19]. Besides, it was further proved that concomitantly administration of melatonin during adjuvant chemotherapy and radiotherapy regimens results in a remarkable reduction of detrimental side-effects in patients [20,21,22]. In particular, there is increasing evidence that MLT could play an important role in chemoprevention of colorectal cancer [23,24,25,26,27]. In this scenario, direct in vivo evidence has shown that MLT could affect numerous intracellular pro-apoptotic signaling pathways involved in the development and progression of cancer [28, 29]. Of these, MLT modulates free radical scavenger action [30], decreases the production of pro-inflammatory cytokines [31], modulates the tumor necrosis factor (TNF-α), promotes the production of interleukins (IL)-1, -6 and -12 [32,33,34], and triggers upregulation of the microRNA-34a/449a cluster [35]. MLT may regulate the matrix metalloproteinase 9 (MMP9) activity through the histone acetylation, in addition to altering the status of DNA methylation patterns which plays a vital role in invasion and metastasis in human cancer cells [36, 37]. Likewise, MLT induce apoptosis in human colorectal cancer cells (HT-29) by blocking the MT3 receptor (quinone oxidoreductase-2), a receptor involved in regulating the homeostasis of copper and zinc, which in turn is strongly associated with the cell growth and proliferation in human carcinogenic cells [38]. Lastly, the available evidence showed that proapoptotic action of melatonin could be also strongly related to HDAC4 nuclear import mediated by CaMKII inactivation, as well by modulating the Bax/Bcl-2 protein expression, and inhibition of caspase-9 activity in tumoral cells [33, 39,40,41]. Despite these valuable antitumoral properties mentioned above, MLT has significant pharmacokinetics limitations thereby restringing its use in conventional drug formulations [42, 43]. To overcome these issues, recent progress towards the discovery of orally active melatonin-based compounds have finalized in various oncological drugs approved by the FDA (Fig. 1C). Thus, for example, it was recently approved a melatonin-analog (trademarked as rucaparib) as a chemotherapeutic agent to treat recurrent ovarian and prostate cancer in adults [44]. Moreover, it was demonstrated that rucaparib plus irinotecan could be an effective combinatorial strategy for more effective management of CRC [45]. Further, two new marketed drugs bearing melatonin-based motifs (in red) were also approved: dacinostat (to treat breast and prostate cancer) and panobinostat (for the management of different myeloma). More recently it was reported that the antidepressant melatonin-derived drug agomelatine exerts a persistent antitumor effect in various in vitro and in vivo models of CRC through cell cycle arrest and a caspase-dependent apoptosis mechanism [46].

The promising therapeutic potential of MLT, as well as 2,3-dihydroquinazolinone derivatives for many types of malignances, especially CRC has drawn therefore our attention. Because of that, in this work, we aimed to utilize a hybridization strategy for the construction of a new anti-colorectal cancer scaffold by merging 2,3-dihydroquinazolin-4-one and melatonin backbones in a unique molecular entity (Fig. 2). Further, aiming to examine the structure–activity relationships (SAR), a substituted aryl group containing electron-donating and -withdrawing substituents was attached at the C-2 position of the 2,3-dihydroquinazolin-4(1H)-one portion with concomitant introduction of chirality. We reasoned that the new designed hybrids based on quinazolinone/MLT could exert a remarked cytotoxic effect in human colorectal malignant cells, making them potential drug candidates for CRC interventions.

Fig. 2
figure 2

Our design strategy for novel quinazolinone-MLT hybrids as anti-CRC agents by incorporating both 2,3-dihydroquinazolin-4-one (in black) and melatonin (in red) cores

Full size image

Results and discussion

Chemistry

In this work, a series of novel quinazolinone/MLT conjugates were designed, synthesized and biologically evaluated for their anti-colorectal cancer activity.

Preparation of quinazolinone/MLT conjugates 1a–l

A very simple, cost effective, and environmentally benign methodology has been used for the synthesis of novel quinazolinone/MLT hybrids using a modified protocol from the previous report by Karimi et al. [47]. The modified protocol involved a simple one-pot three-component approach that using the commercially available isatoic anhydride (1.0 equiv.), 5-methoxytryptamine (1.1 equiv., 5-MeOT) and diverse aromatic aldehydes (1.0 equiv.), catalyzed by either acetic acid under reflux conditions, as illustrated in Fig. 3. This modified one-pot three-component strategy allowed successfully afford the expected quinazolinone/MLT 1a–l in good to excellent yields (63–93%) after 1.5 h. It is also worth note that new hybrids were isolated pure from the crude reaction mixture after treatment with crushed-ice followed by adjusted to pH 8-9 with dilute potassium carbonate, respectively. Besides, target hybrids were prepared by using relatively inexpensive and commercially available starting materials, as well as an eco-friendly, safe, high-yielding and experimentally simple methodology.

Fig. 3
figure 3

One-pot three-component synthesis of novel quinazolinone-MLT hybrids 1a–l

Full size image

All the newly synthesized hybrids were characterized by 1H-NMR, 13C-NMR, as well as electrospray ionization mass spectrometry (MS-ESI), whose spectra showed m/z values of the molecular ion peak [M + H]+ which are in good agreement to the their molecular weights. 1H-NMR spectra revealed the presence of a peak at around 5.0 ppm (singlet) corresponding to the aliphatic methine proton (C-H) of the chiral center, which confirmed the quinazolinone portion into the hybrid structures. Moreover, four or three independent signals observed between 3.30 and 2.70 ppm are ascribed to the diastereotopic two protons belonging to each methylene group along the melatonin moiety, which are also adjacent to the chiral center. These diastereotopic CH2 groups have markedly distinct chemical shifts. Aromatic hydrogens in both quinazolinone and melatonin features were distinguished in the range of δ = 8.26 − 6.70 ppm. Furthermore, the 13C-NMR spectra of hybrids 1a–l displayed a signal around 169 ppm assigned to the quinazolinone carbonyl group. Besides, 13C-NMR spectral data revealed a signal corresponding to the methine carbon (C-H) at 57 ppm which also confirmed the presence of the quinazolinone skeleton in the designed compounds. In addition, 13C-NMR spectra show two picks around 42 and 21 ppm, respectively, confirming the two methylene groups in the melatonin backbone. In short, spectral data unequivocally confirms that novel quinazolinone/MLT conjugates 1a–l were prepared by the proposed one-pot methodology.

Biological assays

Cell growth inhibition

Using the sulforhodamine B (SRB) assay, the cytotoxic potential of the quinazolinone/melatonin hybrids 1a–l was evaluated against the nonmalignant human colon mucosal epithelial cell line (NCM460) and human adenocarcinoma cells (SW480). To verify the efficacy of molecular hybridization, this study included the parental compound melatonin as control. We also included a vehicle control to establish a baseline and the gold standard of first-line treatment for CRC (5-fluorouracil, 5-FU). The cytotoxicity of compounds is expressed as the 50% inhibitory concentration (IC50), that is the concentration of compound required to reduce cell viability by 50%. In addition, considering that safety is a key concern in drug discoveries, we also report the selectivity index (SI), to probe if the newly evaluated hybrids exhibit better selectivity towards malignant cells.

As indicated in Table 1, most of the synthesized compounds exert good cytotoxic activity against malignant SW480 cells 48 h post treatment, with IC50 values ranging from 0.3 to 37.3 µM. It is remarkable the behavior of compounds 1i–l through the time, which were significantly more active than the parental compound and the reference drug, being also 60- and 128-fold more selective than melatonin and 5-FU, respectively, at 48 h of treatment (Fig. 4). Besides, it is important to highlight the behavior of compounds 1j and 1k, which started to induce effect on cancer cells after 24 h of treatment. Considering these results and the important improvement we evidenced in the activity over time, as evidenced in the reduction of the IC50 values at 48 h post treatment, we suggest that our hybrids could exert cytotoxic and/or antiproliferative activity in SW480 cancer cells in a time- and concentration (Fig. 5) dependent manner. These changes were also evidenced by microscopical observations, including the presence of vacuoles and intracytoplasmic granules, together with changes in shape and size (Fig. 6). This emphasizes the significance of molecular hybridization since the parental compound did not exhibit any activity under the evaluated conditions (IC50 > 40 µM). Our data point to the possibility that quinazolinone/melatonin hybrids could be significant candidates since they reduce the toxicity over the non-malignant cell line NCM460 preserving their activity on the colorectal cancer cells evaluated.

Table 1 Cytotoxic effects and selectivity index for novel quinazolinone/melatonin conjugates 1a–l
Full size table
Fig. 4
figure 4

Comparison of the selectivity index (SI) values for the quinazolinone/melatonin hybrids 24 and 48 h post treatment. Melatonin: the parental compound. 5-FU: the reference drug. Selectivity index was calculated as the ratio of the IC50 for the normal cell line (NCM460) to the IC50 for cancer cells (SW480). Values higher than one (dotted line) indicate that the molecules is more selective for cancer cells. **p < 0.01; ****p < 0.0001

Full size image
Fig. 5
figure 5

Effect on cell viability after the treatment with the most active compounds (1i–1l). Results after 48 h of treatment. Cells treated with DMSO was used as control

Full size image
Fig. 6
figure 6

Representative images of SW480 48 h after treatment with different concentrations of the most active hybrids (1i–1l) and the vehicle control (DMSO 0.5%). Magnification: 40×

Full size image

In addition, a more recent study reported the synthesis and biological evaluation of a new series of quinazolinone–pyrimidine hybrids. Authors tested the new compounds against human lung adenocarcinoma cells (A549), the human colorectal cancer cell line SW480 and human breast cancer cells (MCF-7), obtaining one compound with antiproliferative effect against these malignant cells. Besides, the same hybrid induced cell cycle arrest at the S phase and apoptosis in A549 cells in a dose dependent manner [48]. On the other hand, regarding the second precursor melatonin, our group previously identified some chalcone-melatonin hybrids with important cytotoxic and antiproliferative effect against SW480 human colon adenocarcinoma cells [27]. Besides, different authors recently reported a napthalene analog of melatonin named Agomelatine, which induced apoptosis dependent on caspase activities and arrest of the cell cycle on colon cancer cells [46]. All these findings clearly complement our results, suggesting that our compounds may act as cytotoxic or cytostatic agents and could be significant candidates for further studies in the search of new therapeutic alternatives against colorectal cancer, either as monotherapy or as adjuvant therapy.

Structure–activity relationship of cytotoxic response for hybrids 1a–l

A structure-activity relationship (SAR) was rationalized by analyzing the substituents effects and it revealed key differences between those evaluated hybrids 1a–l (Fig. 7): (1) when the aromatic ring was not functionalized (hybrid 1a) we did not detect any cytotoxic activity towards the SW480 cancer cells. (2) In those hybrids containing a mono-substitution on C-3 position, we observed that the presence of weak electron- withdrawing groups is seen to slightly enhance cytotoxic potential compared to that bearing electro-donating substituents, but strong electron- withdrawing appears does not favors the cytotoxic response. (3) Interestingly, the activity increased when methoxy group was substituted at both C-3- and C-5-positions, with inhibitory potency around 0.3 to 2.2 µM. (4) We also note that the activity diminished when methoxy-group is attached at both C-2 and C-3 position (hybrid 1e) compared to that of 2,4-dimethoxy-substituted positional analog (hybrid 1f) (5) Furthermore, the presence of the hydroxyl or methoxy group at the C-4 position in that most active hybrid 3,5-dimethoxy substituted (hybrid 1j) decreased considerably the cytotoxic effect of it.

Fig. 7
figure 7

Structure activity-relationship (SAR) for cytotoxic effect of hybrids 1a–l against SW-480 tumoral cells

Full size image

Theoretical ADME-tox analysis for promising hybrids 1i and 1j

When developing novel anti-cancer candidates, features like drug-likeness, pharmacokinetics, and physicochemical qualities can be effectively utilized as filters. Early biopharmaceutical profiling forecasts could increase the likelihood of success in drug discovery settings, which would save a substantial amount of time and money, thereby deciding whether a lead candidate is ready for preclinical and clinical experiments. Findings from experimental assays showed that among all the synthesized compounds, 1i (3,4-dimethoxy) and 1j (3,5-dimethoxy) showed great ability to induce cytotoxicity in human SW480 colon cancer cells with remarkable selectivity (IS > 57), therefore, in the course of this section we would focus our study only on hybrids 1i and 1j. Thus, 11 drug-like indices were calculated for the most promising hybrids 1i and 1j and then compared against those of 95% of approved drugs (Table 2). Our findings reveal suitable pharmacokinetics indices for both hybrids similar to major modern marketed oral drugs, showing its therapeutic potential as orally administrable drug-candidates for further in pre-clinical experiments. In particular, the degree of lipophilicity (calculated as logPo/w) was predicted to be about 4.76 for both, falling within the ideal the optimal range for lipid-based formulations (–2.0 to 6.0) [49].

Table 2 In silico Biopharmaceutic score for the most potent conjugates 1i and 1j
Full size table

In addition to this crucial permeability parameter, we also calculated the PSA, which similar to logP is well-known to correlate with passive membrane permeability [50]. Notably, hybrids showed an optimal PSA score of 75.82 which together with the logPo/w value would suggest good oral absorption and bioavailability. Beyond the rule of 5, the aromatic ring count (#ArRNG) and the proportion of carbon atoms that are sp3 hybridized (Fsp3), two important novel drug-likeness metrics related to the potential for lability or movement of a molecule over a biological barrier, were also examined [51,52,53,54,55]. Major commercially available drugs have Fsp3 ranging from 1 to 0.25. In this context, an optimal 0.22 of Fsp3 value was found for hybrids. Further, an aromatic/heteroaromatic rings count was conducted (#ArRNG). Current evidence has showed that drug molecules with > 3 aromatic rings is adequate (∼95% of oral approved drugs meet this criterion) have better chance during drugs developability. At this respect, our top-two hybrids bearing three aromatic rings would have an optimal developability profile.

Next to it, we computed the passive transmembrane permeation using MDCK cells or Caco-2 cell monolayers as models. In drug development, both models are frequently suggested as a simplified in vitro model of intestine absorption following oral administration. [56,57,58]. Thus, promising 1i and 1j would possess a permeability of 2200 and 3183 nm/s, respectively, across the intestinal Caco-2 cell monolayers model, and an apparent permeability (Papp) of 1160 and 1729 nm/s across a monolayer of MDCK cells. These in silico permeability results means that these hybrids would be suitable to test in further studies for oral dosing. For 1i and 1j, a different method that forecasted the drug candidates’ capacity to bind blood plasma proteins was computed. The most crucial factor influencing the distribution and transportation of anti-cancer formulations in the systemic circulation and a crucial factor in the initial stages of drug discovery is binding to human serum albumin (logKHSA) [59, 60]. According to the predictive model, compounds with positive values are expected to have a higher affinity for binding HSA, whereas compounds with negative values may exhibit a lower affinity for binding HSA. For the top-two hybrids, the computed binding affinities to HSA were positive values of 0.762 (for 1i) and 0.803 (for 1j), with good fitting within the advised range for potential oral drug prototypes (−1.5 to 1.5).

Bioavailability radar study

In the context of modern drug discovery, the bioavailability radar chart is a useful tool which have been frequently used in biomedical research to select potential drug scaffolds for oral formulations [61] This protocol provides clear evidence if a potential drug candidate would have similar pharmacokinetics attributes to those oral FDA-approved medications. As depicted in Fig. 8, eight crucial biopharmaceutical drug-likeness parameters were graphically inspected for lead-hybrids 1i and 1j by using a radar chart. The grey area represents the optimal range for each parameter, while the computed properties for hybrid appears in the red area. Of note, the computed values for 1i and 1j, fit better within the gray area, thereby suggesting that this compound would have great chance to advance to oral lipid-based formulations, in the future.

Fig. 8
figure 8

Bioavailability radar diagram for the promising hybrids 1i and 1j. The grey area depicting the optimal range for each parameter. The red area represents the computed drug-like properties for hybrid

Full size image

In silico toxicological study

In addition to pharmacokinetic studies, twelve toxicological endpoints have also been estimated for the top-two hybrids 1i and 1j by employing several open-sources chemoinformatic servers such as OSIRIS, TEST, ProTox-II, Pred-hERG, pkCSM, SwissADME, ToxTree, and ADMET-SAR (Table 3). These parameters are closely associated to adverse effects in the progress of a lead molecule, and their early estimation through artificial intelligence tools offer an attractive and rapid low-cost approach toward the design of safer pharmaceutical lead candidates [62, 63]. Regarding their in silico toxicological diagnosis, promising 1i and 1j would have no apparent warnings, precautions and adverse events as mutagenic, tumorigenic, irritant, hepato/nephro/neuro/cardio/ or inmunotoxic. Besides, apparently 1i and 1j would not have any reproductive toxic effect, any structural alert for covalent DNA binding, and any conspicuous oral toxicity, as well, not alerts for pan-assay promiscuity (PAINS).

Table 3 In silico toxicological screening for the most promising conjugates 1i and 1j
Full size table

All together, we demonstrated that the new fusing of rings of quinazolinone and melatonin into a unique molecular framework provide potent cytotoxic compounds with a probable optimal ADME-tox profile. Thereon, we encourage believe that this novel scaffold should be taken into consideration in further investigations addressed to develop novel drug candidates to fight CRC.

Conclusions

Due to the great potential of melatonin and the 2,3-quinazolinone core in anticancer drug discovery, these two pharmacophore units were fused to build a new quinazolinone/MLT scaffold. They were evaluated to determine the chemopreventive potential on colorectal cancer cells. Hybrids 1i–k displayed the highest activity with important selectivity against malignant cells, involving either a cytotoxic and/or antiproliferative activity. Most of these compounds demonstrated potent cytotoxic response, highlighting 1i and 1j for exhibiting ~0.7 and ~0.3 µM of inhibition concentration with low toxicity in non-malignant cells. These compounds were notably more active than the reference compounds 5-FU and MLT. Afterward, computer-aided prediction on drug-like, pharmacokinetics and toxicological profile for leads 1i and 1j suggest that this class of compounds would tend to have favorable and safer biopharmaceutical indices, thereby eligible for entry into more robust pre-clinical oncology studies. All together, these results showed that this class of hybrids emerge as a new chemotherapeutical scaffold that could be taken in account in future colorectal cancer drug developments.

Materials and methods

Chemistry

All chemical reagents were of analytical grade, purchased from Sigma Aldrich and Alfa Aesar (USA), and used without further purification. TLC can be used to monitor the progress of a reactions, on the silica gel pre-coated chromatographic plates 60 F254 (Merck). Stuart Digital Melting Point apparatus (SMP10) and an open capillary was used to determine melting points (uncorrected). The 1H and 13C NMR (CD3OD, CDCl3 or DMSO-d6, 300 MHz, 75.0 MHz) spectra were recorded in a Varian instrument at 25 °C using tetramethylsilane (TMS) as the internal standard. An electrospray ionization mass spectrometry (ESI-HRMS) was a choice method used to record high-resolution mass spectra of all hybrids.

General procedure for the synthesis of quinazolinone-MLT hybrids 1a−l

For the one-pot synthesis of hybrids 1a−l a modified protocol to a published method was used [47]. Commercially available isatoic anhydride (1 mmol), aldehyde (1 mmol) and 5-methoxytryptamine (1.1 mmol), were dissolved in acetic acid (5 mL). Subsequently, the reaction mixture was refluxed ( ~ 120 °C) under vigorous stirring for 1.5 h. After reaction was completed (TLC monitoring), crushed ice was added into the reaction vessel and stirred for 1 h, then aqueous potassium carbonate solution was added, and stir until the pH is adjusted to 8-9. The precipitate formed was collected by vacuum filtration, washed three times with water and oven-drying (45 °C, 3 h) to afford pure the desired hybrids 1a−l as a racemic mixture in good to excellent yields (63–93%).

3-(2-(5-methoxy-1H-indol-3-yl)ethyl)-2-phenyl-2,3-dihydroquinazolin-4(1H)-one (1a)

White solid; yield: 63%, mp: 240–242 °C. 1H NMR (300 MHz, CD3OD): δ 7.90–7.72 (m, 5H, Ar-H), 7.53–7.43 (m, 3H, Ar-H), 7.38 (d, J = 7.5 Hz, 1H, Ar-H), 7.16 (d, J = 8.8 Hz, 1H, Ar-H), 7.02 (d, J = 2.5 Hz, 1H, Ar-H), 6.78 (dd, J = 8.8, 2.5 Hz, 1H, Ar-H), 5.60 (s, 1H, C-Hquinazolinone), 3.84 (s, 3H, OCH3), 3.56–3.45 (m, 1H, N-CH2-A), 3.42–3.34 (m, 1H, N-CH2-B), 3.18–3.06 (m, 1H, N-CH2-CH2-A), 3.06–2.95 (m, 1H, N-CH2-CH2-B). 13C NMR (75 MHz, CD3OD): δ 169.6 (C = O amide), 154.0 (Ar-O), 137.0, 134.0, 132.9, 132.0, 130.3, 129.2, 129.1, 128.8, 126.6, 122.7, 111.8, 111.6, 107.9, 99.7, 57.2 (C-Hquinazolinone), 54.9 (OCH3), 41.1 (N-CH2-CH2), 19.5 (N-CH2-CH2). HRMS (ESI): m/z [M + H]+ calcd for C25H23N3O2: 398.3165; found: 398.3168.

3-(2-(5-methoxy-1H-indol-3-yl)ethyl)-2-(3-methoxyphenyl)-2,3-dihydroquinazolin-4(1H)-one (1b)

Pale brown solid; yield: 78%, mp: 97–100 °C. 1H NMR (300 MHz, CDCl3): δ 7.94–7.86 (m, 2H, Ar-H), 7.82–7.77 (m, 2H, Ar-H), 7.60 (s, 1H, Ar-H), 7.34–7.26 (m, 1H, Ar-H), 7.16 (d, J = 8.7 Hz, 1H, Ar-H), 7.04 (d, J = 2.4 Hz, 1H, Ar-H), 6.84 (dd, J = 8.8, 2.5 Hz, 1H, Ar-H), 5.23 (s, 1H, C-Hquinazolinone), 3.91 (s, 3H, OCH3), 3.80 (s, 3H, OCH3), 3.53–3.36 (m, 1H, N-CH2-A), 3.27–3.11 (m, 1H, N-CH2-B), 3.02–2.91 (m, 1H, N-CH2-CH2-A), 2.91–2.79 (m, 1H, N-CH2-CH2-B). 13C NMR (75 MHz, CDCl3): δ 169.6 (C = O amide), 168.4 (Ar-O), 160.0 (Ar-O), 154.1, 142.7, 134.8, 134.3, 132.7, 131.1, 129.9, 127.7, 123.6, 120.8, 114.0, 113.9, 111.7, 109.9, 100.5, 57.9 (C-Hquinazolinone), 56.0 (OCH3), 55.3 (OCH3), 42.6 (N-CH2-CH2), 22.2 (N-CH2-CH2). HRMS (ESI): m/z [M + H]+ calcd for C26H25N3O3: 428.1962; found: 428.1963.

2-(3-chlorophenyl)-3-(2-(5-methoxy-1H-indol-3-yl)ethyl)-2,3-dihydroquinazolin-4(1H)-one (1c)

Pale brown solid; yield: 77%, mp: 137–140 °C. 1H NMR (300 MHz, CDCl3): δ 7.93–7.85 (m, 2H, Ar-H), 7.83–7.73 (m, 2H, Ar-H), 7.36–7.28 (m, 4H, Ar-H), 7.27–7.22 (m, 1H, Ar-H), 7.16 (d, J = 8.7 Hz, 1H, Ar-H), 7.05 (d, J = 2.5 Hz, 1H, Ar-H), 6.85 (dd, J = 8.7, 2.5 Hz, 1H, Ar-H), 5.18 (s, 1H, C-Hquinazolinone), 3.91 (s, 3H, OCH3), 3.47–3.32 (m, 1H, N-CH2-A), 3.23–3.10 (m, 1H, N-CH2-B), 3.02–2.89 (m, 1H, N-CH2-CH2-A), 2.88–2.77 (m, 1H, N-CH2-CH2-B). 13C NMR (75 MHz, CDCl3): δ 168.6 (C = O amide), 154.1 (Ar-O), 143.8, 134.7, 134.3, 134.3, 132.7, 131.1, 130.1, 128.6, 128.4, 127.6, 126.8, 123.6, 111.8, 111.7, 111.6, 110.3, 100.6, 57.4 (C-Hquinazolinone), 56.0 (OCH3), 42.4 (N-CH2-CH2), 22.4 (N-CH2-CH2). HRMS (ESI): m/z [M + H]+ calcd for C25H22ClN3O2: 432.1460; found: 432.1465.

3-(2-(5-methoxy-1H-indol-3-yl)ethyl)-2-(3-nitrophenyl)-2,3-dihydroquinazolin-4(1H)-one (1d)

Yellow solid; yield: 80%, mp: 95–97 °C. 1H NMR (300 MHz, CD3OD): δ 8.26–8.12 (m, 3H, Ar-H), 7.86–7.76 (m, 3H, Ar-H), 7.69 (d, J = 7.7 Hz, 1H, Ar-H), 7.63–7.52 (m, 2H, Ar-H), 7.12 (d, J = 8.7 Hz, 1H, Ar-H), 6.99 (d, J = 2.5 Hz, 1H, Ar-H), 6.74 (dd, J = 8.7, 2.5 Hz, 1H, Ar-H), 5.29 (s, 1H, C-Hquinazolinone), 3.83 (s, 3H, OCH3), 3.27–3.14 (m, 1H, N-CH2-A), 3.15–3.01 (m, 1H, N-CH2-B), 2.98–2.86 (m, 1H, N-CH2-CH2-A), 2.83–2.72 (m, 1H, N-CH2-CH2-B). 13C NMR (75 MHz, CD3OD): δ 169.5 (C = O amide), 153.7 (Ar-O), 148.4 (Ar-NO2), 144.0, 134.8, 133.9, 133.2, 132.9, 131.8, 129.3, 127.2, 123.3, 122.7, 122.4, 111.3, 111.1, 109.2, 99.8, 56.6 (C-Hquinazolinone), 54.9 (OCH3), 41.3 (N-CH2-CH2), 21.4 (N-CH2-CH2). HRMS (ESI): m/z [M + H]+ calcd for C25H22N4O4: 443.1957; found: 443.1961.

2-(2,3-dimethoxyphenyl)-3-(2-(5-methoxy-1H-indol-3-yl)ethyl)-2,3-dihydroquinazolin-4(1H)-one (1e)

Pale brown solid; yield: 68%, mp: 119–121 °C. 1H NMR (300 MHz, CD3OD): δ 7.89–7.74 (m, 4H, Ar-H), 7.62–7.52 (m, 1H, Ar-H), 7.12 (d, J = 8.7 Hz, 1H, Ar-H), 7.04–6.99 (m, 2H, Ar-H), 6.97 (d, J = 2.5 Hz, 1H, Ar-H), 6.71 (dd, J = 8.7, 2.5 Hz, 1H, Ar-H), 6.64 (t, J = 4.6 Hz, 1H, Ar-H), 5.47 (s, 1H, C-Hquinazolinone), 3.88 (s, 3H, OCH3), 3.83 (s, 3H, OCH3), 3.74 (s, 3H, OCH3), 3.31–3.20 (m, 1H, N-CH2-A), 3.13–3.01 (m, 1H, N-CH2-B), 2.92–2.85 (m, 1H, N-CH2-CH2-A), 2.84–2.77 (m, 1H, N-CH2-CH2-B). 13C NMR (75 MHz, CD3OD): δ 169.5 (C = O amide), 153.6 (Ar-O), 152.9 (Ar-O), 147.3 (Ar-O), 134.2, 133.9, 132.9, 131.5, 127.3, 123.7, 122.7, 121.4, 112.4, 111.2, 110.6, 108.2, 99.6, 59.7 (C-Hquinazolinone), 54.9 (OCH3), 54.8 (OCH3), 52.3 (OCH3), 41.3 (N-CH2-CH2), 21.4 (N-CH2-CH2). HRMS (ESI): m/z [M + H]+ calcd for C27H27N3O4: 458.2061; found: 458.2065.

2-(2,4-dimethoxyphenyl)-3-(2-(5-methoxy-1H-indol-3-yl)ethyl)-2,3-dihydroquinazolin-4(1H)-one (1f)

Pale brown solid; yield: 78%, mp: 97–99 °C. 1H NMR (300 MHz, CD3OD): δ 7.88–7.75 (m, 4H, Ar-H), 7.13 (d, J = 8.7 Hz, 1H, Ar-H), 6.96 (d, J = 2.4 Hz, 1H, Ar-H), 6.77–6.68 (m, 3H, Ar-H), 6.61 (d, J = 2.4 Hz, 1H, Ar-H), 6.40 (dd, J = 8.4, 2.4 Hz, 1H, Ar-H), 5.52 (s, 1H, C-Hquinazolinone), 3.88 (s, 3H, OCH3), 3.82 (s, 3H, OCH3), 3.77 (s, 3H, OCH3), 3.20–3.08 (m, 1H, N-CH2-A), 3.07–2.96 (m, 1H, N-CH2-B), 2.88–2.76 (m, 2H, N-CH2-CH2-A + N-CH2-CH2-B). 13C NMR (75 MHz, CD3OD): δ 169.6 (C = O amide), 161.0 (Ar-O), 158.4 (Ar-O), 153.6 (Ar-O), 133.9, 132.9, 131.6, 130.1, 127.3, 122.6, 121.2, 111.1, 110.5, 108.6, 103.7, 99.6, 98.1, 54.9 (OCH3), 54.6 (OCH3), 54.4 (OCH3), 50.4 (C-Hquinazolinone), 40.0 (N-CH2-CH2), 21.3 (N-CH2-CH2). HRMS (ESI): m/z [M + H]+ calcd for C27H27N3O4: 458.2067; found: 458.2072.

2-(3-hydroxy-4-methoxyphenyl)-3-(2-(5-methoxy-1H-indol-3-yl)ethyl)-2,3-dihydroquinazolin-4(1H)-one (1g)

Pale brown solid; yield: 77%, mp: 137–139 °C. 1H NMR (300 MHz, CD3OD): δ 7.88–7.72 (m, 4H, Ar-H), 7.66–7.49 (m, 1H, Ar-H), 7.12 (d, J = 8.7 Hz, 1H, Ar-H), 6.96 (d, J = 2.4 Hz, 1H, Ar-H), 6.92 (d, J = 8.1 Hz, 1H, Ar-H), 6.80–6.73 (m, 2H, Ar-H), 6.71 (dd, J = 8.7, 2.4 Hz, 1H, Ar-H), 5.10 (s, 1H, C-Hquinazolinone), 3.84 (s, 3H, OCH3), 3.83 (s, 3H, OCH3), 3.30–3.20 (m, 1H, N-CH2-A), 3.16–2.99 (m, 1H, N-CH2-B), 2.98–2.75 (m, 2H, N-CH2-CH2-A + N-CH2-CH2-B). 13C NMR (75 MHz, CD3OD): δ 169.6 (C = O amide), 153.6 (Ar-O), 147.7 (Ar-O), 146.3 (Ar-OH), 134.0, 132.9, 129.6, 129.4, 122.7, 120.0, 115.4, 111.2, 110.7, 108.3, 99.7, 56.9 (C-Hquinazolinone), 55.0 (OCH3), 54.9 (OCH3), 41.3 (N-CH2-CH2), 21.2 (N-CH2-CH2). HRMS (ESI): m/z [M + H]+ calcd for C26H25N3O4: 444.1174; found: 444.1179.

2-(4-hydroxy-3-methoxyphenyl)-3-(2-(5-methoxy-1H-indol-3-yl)ethyl)-2,3-dihydroquinazolin-4(1H)-one (1h)

Pale orange solid; yield: 81%, mp: 211–213 °C. 1H NMR (300 MHz, DMSO-d6): δ 7.85–7.76 (m, 3H, Ar-H), 7.10 (d, J = 8.7 Hz, 1H, Ar-H), 6.92–6.82 (m, 3H, Ar-H), 6.71 (d, J = 8.1 Hz, 1H, Ar-H), 6.63 (dd, J = 8.7, 2.2 Hz, 3H, Ar-H), 4.96 (s, 1H, C-Hquinazolinone), 3.74 (s, 3H, OCH3), 3.72 (s, 3H, OCH3), 3.20–3.01 (m, 1H, N-CH2-A), 2.97–2.82 (m, 1H, N-CH2-B), 2.77–2.55 (m, 2H, N-CH2-CH2-A + N-CH2-CH2-B). 13C NMR (75 MHz, DMSO-d6): δ 169.8 (C = O amide), 153.4 (Ar-O), 147.8 (Ar-O), 146.3 (Ar-OH), 137.2, 134.8, 134.4, 133.1, 131.4, 127.6, 123.4, 121.1, 115.4, 113.0, 112.1, 110.5, 108.2, 100.2, 57.3 (C-Hquinazolinone), 56.0 (OCH3), 55.8 (OCH3), 42.2 (N-CH2-CH2), 22.8 (N-CH2-CH2). HRMS (ESI): m/z [M + H]+ calcd for C26H25N3O4: 444.1911; found: 444.1916.

2-(3,4-dimethoxyphenyl)-3-(2-(5-methoxy-1H-indol-3-yl)ethyl)-2,3-dihydroquinazolin-4(1H)-one (1i)

Pale brown solid; yield: 75%, mp: 114–116 °C. 1H NMR (300 MHz, CD3OD): δ 7.87–7.74 (m, 5H, Ar-H), 7.10 (d, J = 8.7 Hz, 1H, Ar-H), 6.95 (d, J = 1.9 Hz, 1H, Ar-H), 6.89 (d, J = 7.3 Hz, 1H, Ar-H), 6.88 (s, 1H, Ar-H), 6.74 (dd, J = 8.2, 1.9 Hz, 1H, Ar-H), 6.63 (dd, J = 8.7, 2.5 Hz, 1H, Ar-H), 5.00 (s, 1H, C-Hquinazolinone), 3.74 (s, 3H, OCH3), 3.73 (s, 3H, OCH3), 3.71 (s, 3H, OCH3), 3.16–3.05 (m, 1H, N-CH2-A), 2.99–2.84 (m, 1H, N-CH2-B), 2.75–2.66 (m, 1H, N-CH2-CH2-A), 2.66–2.55 (m, 1H, N-CH2-CH2-B). 13C NMR (75 MHz, CD3OD): δ 169.8 (C = O amide), 153.4 (Ar-O), 149.0 (Ar-O), 148.5 (Ar-O), 137.0, 136.0, 134.8, 133.1, 131.4, 127.6, 123.4, 120.8, 112.6, 112.1, 111.8, 110.6, 108.3, 100.2, 57.2 (C-Hquinazolinone), 56.0 (OCH3), 55.9 (OCH3), 55.8 (OCH3), 42.1 (N-CH2-CH2), 22.7 (N-CH2-CH2). HRMS-ESI (m/z): 458.2051 [M + H]+ calcd. for C27H27N3O4: 458.2054.

2-(3,5-dimethoxyphenyl)-3-(2-(5-methoxy-1H-indol-3-yl)ethyl)-2,3-dihydroquinazolin-4(1H)-one (1j)

Pale brown solid; yield: 93%, mp: 150–152 °C. 1H NMR (300 MHz, DMSO-d6): δ 7.89–7.71 (m, 4H, Ar-H), 7.12 (d, J = 8.7 Hz, 1H, Ar-H), 6.89 (s, 1H, Ar-H), 6.64 (dd, J = 8.7, 2.5 Hz, 1H, Ar-H), 6.53–6.30 (m, 4H, Ar-H), 4.98 (s, 1H, C-Hquinazolinone), 3.74 (s, 3H, OCH3), 3.70 (s, 6H, 2 x OCH3), 3.19–3.02 (m, 1H, N-CH2-A), 2.99–2.82 (m, 1H, N-CH2-B), 2.78–2.55 (m, 2H, N-CH2-CH2-A + N-CH2-CH2-B). 13C NMR (75 MHz, DMSO-d6): δ 170.0 (C = O amide), 160.7 (2 x Ar-O), 153.4 (Ar-O), 145.9, 136.5, 134.7, 133.2, 131.4, 127.6, 123.3, 112.1, 110.7, 108.3, 106.9, 100.2, 99.3, 57.4 (C-Hquinazolinone), 55.8 (OCH3), 55.6 (2 x OCH3), 42.1 (N-CH2-CH2), 22.7 (N-CH2-CH2). HRMS-ESI (m/z): 458.2066 [M + H]+ calcd. for C27H27N3O4: 458.2069.

3-(2-(5-methoxy-1H-indol-3-yl)ethyl)-2-(3,4,5-trimethoxyphenyl)-2,3-dihydroquinazolin-4(1H)-one (1k)

Pale brown solid; yield: 81%, mp: 92–94 °C. 1H NMR (300 MHz, DMSO-d6): δ 7.87–7.78 (m, 4H, Ar-H), 7.12 (d, J = 8.7 Hz, 1H, Ar-H), 6.89 (d, J = 2.4 Hz, 1H, Ar-H), 6.69–6.64 (m, 3H, Ar-H), 6.62 (d, J = 2.4 Hz, 1H, Ar-H), 5.00 (s, 1H, C-Hquinazolinone), 3.74 (s, 3H, OCH3), 3.72 (s, 6H, 2 x OCH3), 3.65 (s, 3H, OCH3), 3.26–3.10 (m, 1H, N-CH2-A), 3.00–2.86 (m, 1H, N-CH2-A), 2.80–2.67 (m, 1H, N-CH2-CH2-A), 2.65–2.55 (m, 1H, N-CH2-CH2-B). 13C NMR (75 MHz, DMSO-d6): δ 169.9 (C = O amide), 153.4 (Ar-O), 153.1 (2 x Ar-O), 139.1 (Ar-O), 137.1, 136.8, 134.8, 133.1, 131.4, 127.6, 123.4, 112.1, 110.6, 108.2, 105.9, 100.2, 60.4 (C-Hquinazolinone), 58.0 (OCH3), 56.2 (2 x OCH3), 55.8 (OCH3), 42.7 (N-CH2-CH2), 22.7 (N-CH2-CH2). HRMS (ESI): m/z [M + H]+ calcd for C28H29N3O5: 488.2076; found: 488.2079.

2-(4-hydroxy-3,5-dimethoxyphenyl)-3-(2-(5-methoxy-1H-indol-3-yl)ethyl)-2,3-dihydroquinazolin-4(1H)-one (1l)

Pale red solid; yield: 82%, mp: 161–163 °C. 1H NMR (300 MHz, DMSO-d6): δ 7.89–7.78 (m, 5H, Ar-H), 7.11 (d, J = 8.7 Hz, 1H, Ar-H), 6.89 (d, J = 2.5 Hz, 1H, Ar-H), 6.63 (dd, J = 8.7, 2.5 Hz, 1H, Ar-H), 6.60–6.56 (m, 2H, Ar-H), 4.96 (s, 1H, C-Hquinazolinone), 3.74 (s, 3H, OCH3), 3.70 (s, 6H 2 x OCH3), 3.26–3.13 (m, 1H, N-CH2-A), 2.98–2.84 (m, 1H, N-CH2-B), 2.78–2.63 (m, 1H, N-CH2-CH2-A), 2.66–2.54 (m, 1H, N-CH2-CH2-B). 13C NMR (75 MHz, DMSO-d6): δ 169.8 (C = O amide), 153.4 (Ar-O), 148.2 (2 x Ar-O), 137.2 (Ar-OH), 135.3, 134.8, 133.4, 133.1, 131.4, 127.7, 123.4, 112.1, 110.5, 108.1, 106.3, 100.3, 57.9 (C-Hquinazolinone), 56.4 (2 x OCH3), 55.8 (OCH3), 42.7 (N-CH2-CH2), 22.7 (N-CH2-CH2). HRMS (ESI): m/z [M + H]+ calcd for C27H27N3O5: 474.1918; found: 474.1922.

Biological assays

Cell lines and culture medium

Human colorectal adenocarcinoma cells (SW480) and a normal human colon mucosal epithelial cell line (NCM460) from the European Collection of Authenticated Cell Cultures (ECACC, UK) were used to study the chemopreventive potential of the new quinazolinone/melatonin hybrids. The cells were maintained in Dulbecco’s Modified Eagle Medium (DMEM), which was supplemented with 10% horse serum that had been previously inactivated at 56 °C, 1% non-essential amino acids, and 1% antibiotics (streptomycin and penicillin; Gibco Invitrogen, Carlsbad, CA, USA). Horse serum was lowered to 3% for each experiment, and ITS defined medium (Gibco, Invitrogen, Carlsbad, CA, USA) with 10 mg/mL insulin, 5 mg/mL transferrin, and 5 ng/mL selenium was added to the culture media as well [64].

Cell viability

The effect of the hybrids on cell viability was assessed using the Sulforhodamine B (SRB) colorimetric technique. Before the monolayer reached 100% confluence, the cell density was adjusted to 10,000 cells/well in 96-well tissue culture plates, which had previously been standardized to guarantee the cells’ exponential phase during the experiment. Following this, the cells were incubated for a full day before being treated every day for the next 24 to 48 h with progressively higher concentrations (0.6 to 40 µM) of the various compounds. We used different controls in this investigation in addition to the hybrid molecules. The first one (vehicle control) established the baseline for the tests by solubilizing the chemicals in DMSO (0.5%). Furthermore, we employed the parental compound (melatonin) and the reference drug (5-FU). The ideal parameters were 5% CO2 and 37 °C. After each treatment, culture plates were left at 4 °C for one hour with 50% v/v of cold trichloroacetic acid (MERCK), to fix the cells. Next, the cell protein content was stained using 0.4% (w/v) SRB (Sigma-Aldrich, St. Louis, MO, USA). Five washes using 1% acetic acid were employed to remove any unbound SRB. Before solubilizing in 10 mM Tris-base and measuring the absorbance at 492 nm using a Mindray MR-96A microplate reader, protein-bound SRB was allowed to air dry. Each experiment was conducted thrice [65, 66].

Statistical analysis

The experimental design included at least three repetitions for each measurement. The information was displayed as mean ± SE, or standard error. Following the two-way ANOVA statistical analysis, Tukey’s test was conducted. Values of p < 0.05 were considered significant. GraphPad Prism version 8.0.1 for (Graph Pad Software, San Diego, CA, USA) was used to analyze the data.

Theoretical ADME-tox screening

The most promising hybrids 1i (3,4-dimethoxy) and 1j (3,5-dimethoxy) were screened for their ADME-tox profile using the opensource cheminformatics toolkits SwissADME [67]. For 1i and 1j, eleven biopharmaceutical relevant properties were accessed: Topographical polar surface area (TPSA), MW and rotatable bonds, logPo/w, the number of H-donors/H-acceptors, the fraction of sp3 carbon atoms, and the number of aromatic/heteroaromatic rings. Finally, SwissADME was also used for investigating substructural alerts to identify pan-assay interference compounds (PAINS). Note that, these important biopharmaceutical parameters govern oral exposure, as well as permeability, absorption, motion and action of potential drug candidates. Further, we explored the toxicological profile for 1i and 1j using eight in silico freeware platforms such as SwissADME [67], OSIRIS [68], T.E.S.T Version 5.1.2 [69], ProTox-II [70], Pred-Herg [71], pkCSM [72], ToxTree [73], and ADMET-SAR [74].

Supporting information

Supplementary material provides copy of the NMR spectra (1H-NMR and 13C-NMR) and ESI-HRMS of quinazolinone-MLT conjugates 1a–l.