Design, Synthesis and Biological Activities of New Phthalimide and Thiazolidine Derivatives

The combination of pharmacophoric nuclei with different targets has been a strategy for the development of new drugs aimed at improving cancer treatment. A series of ten novel phthalimido-thiazolidine-2-4-dione derivatives were synthesized by two different synthetic routes. The compounds were tested and evaluated in vitro, through antineoplastic activities against cancer cells. Cell cycle analyzes and clonogenic assay were also performed. In addition to these tests, in silico predictions were performed. The synthesized FT-12 compound (9j) exhibited antiproliferative activity against Panc-1, Sk-mel-28 and PC-3 cells. FT-12 reduced the ability to form new clones, also caused irreversibility in cell cycle, inducing arrest in phase S. Besides, the compound (FT-12) caused necrosis and apoptosis. The results suggest that phthalimido-thiazolidine derivatives may be useful in cancer therapy, highlighting compound FT-12 (9j) as a promising candidate. More studies must be carried out to conrm the viability.


Introduction
The chemical nucleus of phthalimides (-CO-N (R) -CO-) shows that these structures are hydrophobic, which increases their potential to cross biological membranes in vivo [9] (Lamie et al. 2015). Thalidomide is the oldest and most well-known phthalimide derivative. It was initially developed as a sedative hypnotic agent to treat vomiting in pregnancy. Despite its initial successful clinical results, its use has been discontinued because of teratogenic properties. This compound caused severe malformations in the children of women who took the drug during pregnancy [10,11].
Thiazole and thiazolidine nucleus is an important heterocyclic pharmacophore present in several molecules and have signi cance to create privileged chemical structures possessing pharmacological activities [21]. The analogous structure thiazolidine-2,4-dione is characterized by the formation of new compounds from substitutions or additions in the free portions -NH and -CH 2 of its nucleus [22].
For example, scaffolds with possible biological activity were developed by linking the thiazole ring to the phthalimide ring, demonstrating the potential of these hybrid structures [26]. In this article we present the synthesis of hybrid compounds, the union of the thiazolidine nucleus with the phthalimide nucleus.
The series of reagents, intermediate products called IPs (6a-c), produced for the realization of our nal compounds had their methodology approved and with promising results [27,6]. In conclusion, the evaluation of new agents is necessary in order to nd more selective and e cient treatments for different types of tumors [28].
In this article we report synthesis and cytostatic activity of novel hybrid compounds phthalimidothiazolidine-2-4-dione. We examined its e cacies on a panel of human cancer cell lines, and in normal human lymphocytes cells. Finally, with one of our most active compounds (FT-12) (9j), we obtained an important in vitro activity in tumor growth suppression, also in silico analyzes were performed with this compound.
In Fig. 1, the nal compounds were synthesized through the following steps: thiazolidine-2,4-dione (2) was solubilized with acetonitrile. Separately, NaOH, after maceration, was solubilized in methanol and acetonitrile. Then, the two products are placed to react for 30 minutes and N-chloro-methylphthalimide (1) was added. The reaction took place at 65 ºC of temperature for about 2:30h, accompanied by thin layer chromatography (CCD). After the end of the reaction, the compound formed, the intermediate 3-((1,3dioxoisoindolin-2-yl)methyl)thiazolidine-2,4-dione (LPSF/ FT-01) (3), was ltered. The liquid product was placed in the refrigerator, where it stayed for about 72 hours to form white crystals, then ltered. In the second step, LPSF/ IPs (ethyl cyanoacetate esters) (6a-c) were synthesized from ethyl cyanoacetate (4) with aromatic aldehydes (5a-c) using morpholine as catalyst [27]. In the third step, both FT-01 (3) and the speci c synthesized LPSF/ IPs (6a-c) were placed to react in the presence of morpholine (catalyst) at a temperature of 65 C, forming the nal compounds (LPSF / FTs) (7a-c). The reaction times varies slightly depending on the IPs reagents. The products were ltered and washed with ethanol.
A different synthesis route is shown in Fig. 2, in this route were synthesized novel thiazophthalimide derivatives by condensation reaction, LPSF/ FTs (9e-k). In the second step, the 3-((1,3-dioxoisoindolin-2yl)methyl)thiazolidine-2,4-dione(FT-01) (3), compound synthesized in the rst step, was placed to react by condensation reaction following the procedures of Harada et al. using speci c aldehyde (reagents 8e-k of Fig. 4) in the presence of ammonium acetate and acetic acid (solvent), at temperature of 110 ºC [29] ( Harada et al. 2012). At the end of the reaction, the product was ltered and washed with water, generating the nal compounds (LPSF / FTs) (9e-k). The reaction times varies between 3h and 5h.
PBMCs were plated 5x10 5 cells per well, and neoplastic cells were distributed into 96 well plates, cells Loading [MathJax]/jax/output/CommonHTML/fonts/TeX/fontdata.js concentration was according to the NIH recommendations, the number of cells per well/100µl. After 24h, different concentrations of phthalimido-thiazolidine derivatives (1, 10, 50, and 100 µM) in plates were incubated in a humidi ed atmosphere at 37°C and 5% CO 2 for 48h (PBMCs) for 72h. After the incubation period, 20 µL of the 5 mg/mL MTT solution diluted in PBS was added; the plates were then protected from light, and were incubated once more in a humidi ed atmosphere at 37°C and 5% CO2 for a period of 3h. Then it was added 130 µl of 20% SDS per well, and measured optical density were realized after 24 hours at the 560 nm wavelength. The reading was performed on a microplate reader (EL808 -Biotek®). Determination of the cytotoxic potential of the tested compounds was calculated in relation to the control, which was treated with the 0,1% DMSO vehicle. To determine the concentration of the compound, required for 50% inhibition in vitro (IC 50 ), three independent MTT assays were performed. Doxorrubicin Cell cycle analysis. Cells were plated in 6-well plates at a concentration of 5 × 10 5 cells per well and treated at IC 50 concentration. After 48 hours, the cells were trypsinized, washed twice with PBS, and xed in 70% alcohol at 4-6°C; then, the cells were incubated for 24 hours at -20°C. In the next step, the cells were incubated with propidium iodide (5 mg/mL)/RNase A (0.25 mg / mL) in PBS for 30 minutes on ice and were protected from light. Cellular DNA content was quanti ed by ow cytometry using the Accuri C6 ow cytometer, wherein the percentage of DNA in each phase of the cycle was obtained by the C6 software (Becton, Dickinson and Company, USA).
In vitro scratch assay. Cells were plated in 24-well plates at a density of 1x10 5 , and incubated at 37°C under 5% CO 2 atmosphere for 24 hours. After reaching the necessary con uence, the growth medium was aspirated, and two washes with 1X PBS were performed; then, two perpendicular marking lines (striae) were made using a P200 pipette. After the labeling procedure, the cells were treated with complete medium (control), 0.1% DMSO, and FT-12 synthetized, at concentrations previously determined by IC 50 .
They were photographed under the inverted microscope (eclipse Ts2 Nikon) (t = 0) and incubated for 24 hours (t = 0, t = 6, t = 12 and t = 24) under the same conditions as above. Migration assay data were obtained from image analysis through the Image J program (Version 1.49; NIH, USA).
Clonogenic Assay. Cells were initially plated in 24-well cluster plates at a density of 3.0×10 4 cells/well and incubated at 37°C in atmosphere of 5% CO 2 . After 24h and 48h, the medium was changed, the cells were treated with 0.1% DMSO (control group), or with the phthalimido-thiazolidine derivative FT-12 (9j) at concentration previously determined by the IC 50 . After 14 days, the colonies were xed with paraformaldehyde and stained using Methyl Violet (Thermo Fisher Scienti c, Waltham, MA, USA). The stained colonies were counted manually.
Statistical analysis. Three independent experiments were performed in triplicate. IC 50 values and 95% con dence intervals were obtained with nonlinear regression with the OriginPro program (8.0; OriginLab, Northampton, Massachusetts, USA). Statistical signi cance was tested with two-tailed unpaired student's t-tests in relation to the untreated or positive controls, and differences were considered signi cant when p Loading [MathJax]/jax/output/CommonHTML/fonts/TeX/fontdata.js values were less than 0.05. Values were expressed as the mean ± SD of three or more replicate experiments.
In silico predictions. First, we performed computational analysis using PASS online for predicting FT-12 (9j) effects and mechanisms of action. PASS online is a software for evaluating the biological potential of compounds and gives a score for probability to be active (Pa) and inactive (Pi). FT-12 was estimated as biological active for a speci c activity when Pa > Pi and Pa > 0.1 to be considered signi cant. Genes directly related to FT-12 predicted functions were investigated through ENRICHR for gene ontology enrichment analysis of associated biological processes (BP) and KEGG pathways.

Synthesis
The structures of new compounds LPSF/FTs synthesized were con rmed by spectroscopic techniques such as nuclear magnetic resonance ( 1 H NMR) model Varian 300 or 400 MHz, using solvent DMSO-d 6 and CDCL 3 . The peaks of the RMN signals were designated s -singlet; d -doublet; t -triplet; m -multiplet.
Melting points were determinated in a capillary tube using Buchi melting point M-565. Infrared spectra (IR) were recorded on a Prestige-21, Shimadzu model 01801. For Mass Spectra (MS), Bruker Daltonics, modelo auto ex III smart beam was utilized.
For example, the nal compounds were proven by the presence of the CH 2 group (simplet) attached to the nitrogen atom of thiazolidine in the range of 5.6 ppm in the nuclear magnetic resonance spectrum. In the compound 9k some hydrogens in the aromatic ring of the radical, with the presence of a nitrogen, were less protected, appearing in the lower eld of the spectrum than the hydrogens present in the aromatic phthalimidic ring.

In vitro cytotoxic activity in tumor cells and clonogenic assays
Initially, the eleven compounds synthesized were tested against peripheral blood mononuclear cells (PBMCs). All the synthesized compounds were assessed for in vitro cytotoxic effect in PBMCs using MTT assay, as a primary step for investigating their cytotoxic activity. It was observed that phthalimidothiazolidine derivatives were nontoxic (IC 50 > 100 µM) in this type of cells. The synthesized compounds were tested in different hematopoietic cells lines and solid tumors, only FT-9 (9i) and FT-12 (9j) showed antineoplastic activity as shown in Table 1. The compound FT-9 shows activity in two tumor lines (K-562 and Molt-4), while FT-12 in three (PC-3, Panc-1, SK-mel-28).
Loading [MathJax]/jax/output/CommonHTML/fonts/TeX/fontdata.js To evaluate the cytotoxic potential, the highest and lowest IC 50 concentrations were chosen. Initially the inhibition of clone formation was evaluated at 24 and 48 hours of treatment. Results in Fig. 3 showed that the number of viable PC-3 cell colonies was signi cantly decreased after FT-12 (9j) treatment. Both in the 24-hour (p = 0,0153) and 48-hour (p = 0,0644), there was a reduction in the formation of new clones.
We also evaluated the ability to inhibit the formation of colonies in Panc-1 cells (Fig. 4). Both, in the 24hour (p = 0,0372) and 48-hour (p = 0,096), there was a reduction in the formation of new clones. Against PC-3 strain, the FT-12 derivative exhibited a greater inhibitory effect with less treatment time. However, against the Panc-1 strain, the inhibitory effect is observed in both treatment times.
Loading [MathJax]/jax/output/CommonHTML/fonts/TeX/fontdata.js Effects of LPSF/FT-12-9 j on the cell cycle and on the induction of apoptosis The FT-12 derivative effects on cell cycle in Panc-1 cells was evaluated. This derivative did not alter the cell cycle progression signi cantly (Table 2). To evaluate if FT-12 (9j) derivative induces other cells to death processes, we tested for analysis by 7-AAD (permeabilidade de mebrana marker) and annexin V (apoptosis). The derivative induced a signi cant increase in necrosis (p = 0.0202), also induced by doxorrubin treatment (p = 0.0016; Fig. 5). Furthermore, FT-12 treatment resulted in a signi cant increase in apoptosis (p = 0.0017).
To investigate apoptosis molecular pathway, we next evaluated the expression of the PARP, cell apoptosis-related protein Poly (ADP-ribose) polymerase in PANC-1 cells. FT-12 treatment did not induce a signi cant increase (17.35%) in cleaving PARP compared to untreated cells (2.35%), as shown in Fig. 6.

Bioinformatics studies
We chose the most promising synthesized compound to run some in silico tests. Through PASS online screening was performed. Some genes (shown in Table 3) were predicted to be targeted by FT-12 in some way (e.g Mcl-1 receptor antagonist).  *Pa (probability "to be active"); **Pi (probability "to be inactive")

Discussion
The new intermediate obtained FT-01 (3), since the rst reactions presented satisfactory yields, between 62-81%. The synthesis is also relatively fast, two hours of reaction. Therefore, it was made on a larger scale of production, also obtaining good yields. The main di culty is the fact that FT-1 (3) is not very soluble, making it di cult to use as a reagent.
Loading [MathJax]/jax/output/CommonHTML/fonts/TeX/fontdata.js Seven compounds (FT-2, FT-3, FT-4, FT-6, FT-9, FT-12, FT-15), 9e-k, were obtained by the condensation methodology [29]. This second methodology (Fig. 2) was used due to the di culty of synthesis by the rst methodology. For the rst methodology is necessary to synthesize one more intermediate step (the IPs) to reach the nal compound, what makes this synthesis more laborious and costly (Fig. 1). The compounds FT-7, FT-8 and FT-18, 7a-c, were synthesized by the rst and more expensive methodology, following similar steps carried out previously [27]. Research done on the SciFinder website, did not present in the literature compounds with similarity su cient for a good comparison with the obtained thiazophthalimides.
Most of the nal reactions did not take much time to be completely nished, the times varied between 1h and 5h. All the nal compounds FTs (7a-c and 9e-k) had yields varying between 22.5% -95%, this variation demonstrates that the aldehydes, intermediates of the reaction, have a great in uence on the reactivity for the formation of new compounds.  (Fig. 7).
The FT-12 derivative owns a group quinoline. Quinine and camptothecin are important quinoline alkaloids with important drugs antimalarial and anticancer. Many molecules with a broad range of bioactivities, including antitumor, antimalarial, antibacterial and antifungal, antiparasitic and insecticidal, antiviral, antiplatelet, anti-in ammatory, herbicidal, antioxidant and other activities were studied [31]. Our results demonstrated that the 2-chloro-quinoline ring system had an antiproliferative activity more prominent than others substitutes. Pun et al. evaluated an novel quinoline effect in proliferation of esophageal cancer model in vivo, they showed that this compound reduced the tumor size in nude mice xenograft [32]. Mphahlele et al. synthesized new thienoquinolines, which showed cytotoxic effects against human breast adenocarcinoma cell line (MCF-7cells) with IC 50 between 0.014 and 1.84 mg/ml [33]. Magar et al. showed chlorine group at the meta or para-position of the 2-phenyl ring, displayed highly potent topoisomerase II inhibitory activity [34].
We performed analysis to identify speci c biological processes that could be affected by the compound FT-12. This included regulation of cell death, regulation of apoptosis, DNA intercalator, programmed cell death, and signal transduction. FT-12 (9j) induced a decrease in the formation of clones in both Panc-1 and PC-3. This assay can be used to determine the cytotoxic effects of various treatments, including chemotherapeutic and ionizing radiation [35]. Belluti et al. demonstrated that phthalimide-based Loading [MathJax]/jax/output/CommonHTML/fonts/TeX/fontdata.js curcumin derivatives exerted inhibitory effects of the migration capacity of PC-3 and Du-145 cells, the clonogenic inhibition effect with decrease in colony formation reached up to 95% following K3F21 treatment at 5 and 10 µM in PC-3 and DU-145 cells [36].
We decided to evaluate the cell death mechanism induced by FT-12 compound in cell Panc-1. The FT-12 derivative induced necrosis and apoptosis, besides inducing cleavage of PARP. Rohitkumar et al. showed that thieno(2,3-b)quinoline (BPTQ) derivatives induced arrest cell cycle in phase S and cleaved Parp in MOLT-4 cells leukemia [37]. Santos et al. showed 1,3-thiazole cause irreversible cancer cell damage by inducing necrosis and apoptosis in HT-29 cancer cells [38].

Conclusion
In summary, synthesized compounds derived from phthalimide and thiazolidine were made by two different synthetic routes, some of them induced cytotoxic activity against tumor cells. The synthesized compound FT-12 (9j) also had clonogenic inhibition effect, decreasing colony formation and inhibiting migration in prostate and pancreatic cells. FT-12 (9j) derivative also induced necrosis/apoptosis in pancreatic cells. In silico evaluation demonstrated that phthalimide-thiazolidine nucleus has the potential to act in several therapeutic targets. More studies must be carried out to con rm the viability of the new compounds.   Colony formation assay for analyzing the effect of FT-12 derivative in Panc-1 cells.

Figure 5
Effect in Panc-1 of FT-12 derivative and doxorubicin in apoptosis/necrosis by ow cytometry.