Identification and Biological Characterization of the Pyrazolo[3,4-d]pyrimidine Derivative SI388 Active as Src Inhibitor

Src is a non-receptor tyrosine kinase (TK) whose involvement in cancer, including glioblastoma (GBM), has been extensively demonstrated. In this context, we started from our in-house library of pyrazolo[3,4-d]pyrimidines that are active as Src and/or Bcr-Abl TK inhibitors and performed a lead optimization study to discover a new generation derivative that is suitable for Src kinase targeting. We synthesized a library of 19 compounds, 2a-s. Among these, compound 2a (SI388) was identified as the most potent Src inhibitor. Based on the cell-free results, we investigated the effect of SI388 in 2D and 3D GBM cellular models. Interestingly, SI388 significantly inhibits Src kinase, and therefore affects cell viability, tumorigenicity and enhances cancer cell sensitivity to ionizing radiation.


Introduction
Src is a non-receptor tyrosine kinase (TK) belonging to the Src family kinases (SFKs), which comprises different members, including Fyn. Src involvement in cancer has been extensively demonstrated; indeed, Src was the first TK and the first oncogene ever identified [1][2][3]. Several studies have shown Src kinase deregulation in different tumors; interestingly, Src hyperactivation is rarely due to mutations or gene duplication, while it is mainly linked to the aberrant activation of upstream receptor tyrosine kinases (RTKs) including EGFR, PDGFR, and MET. Src constitutive activation results in aberrant cell proliferation, survival, and migration and sustains angiogenesis, metastasis development, and resistance to therapy [4].
In several cases, the most active compounds were decorated with an anilino group at the C4 position. For instance, derivative 1a (SI83) was active in a xenograft mouse model of osteosarcoma [9], compound 1b (SI223) reduced more than 50% of the tumor volume in Starting from these data, we decided to synthesize a library of 19 compounds, 2a-s, to extend the structure-activity relationship (SAR) of this class of inhibitors. Compounds 2a-s (Table 1) were designed by partially combining the structural features of the lead compounds 1a-c and evaluated in enzymatic assays against Src, Abl and Fyn, based on the activity of in-house pyrazolo [3,4-d]pyrimidines that were previously synthesized. Interestingly, this study led to the identification of 2a (SI388), a quite potent Src inhibitor endowed with increased activity compared the lead compound 1a.  Starting from these data, we decided to synthesize a library of 19 compounds, 2a-s, to extend the structure-activity relationship (SAR) of this class of inhibitors. Compounds 2a-s (Table 1) were designed by partially combining the structural features of the lead compounds 1a-c and evaluated in enzymatic assays against Src, Abl and Fyn, based on the activity of in-house pyrazolo [3,4-d]pyrimidines that were previously synthesized. Interestingly, this study led to the identification of 2a (SI388), a quite potent Src inhibitor endowed with increased activity compared the lead compound 1a. in mice inoculated with 32D-T315I CML cells [5], and compound 1c (SI306) showed activity in in vivo models of NB and GBM [6,7] (Figure 1). Starting from these data, we decided to synthesize a library of 19 compounds, 2a-s, to extend the structure-activity relationship (SAR) of this class of inhibitors. Compounds 2a-s (Table 1) were designed by partially combining the structural features of the lead compounds 1a-c and evaluated in enzymatic assays against Src, Abl and Fyn, based on the activity of in-house pyrazolo [3,4-d]pyrimidines that were previously synthesized Interestingly, this study led to the identification of 2a (SI388), a quite potent Src inhibitor endowed with increased activity compared the lead compound 1a. Based on these cell-free results, we decided to investigate the effect of SI388 using 2D and 3D GBM cellular models in which Src activity has been previously reported to be aberrantly activated [10,11].
We provide evidence of the ability of SI388 to affect cell viability and tumorigenicity via Src inhibition. Furthermore, we demonstrate that SI388 increases cell sensitivity to ionizing radiation.

Chemistry
We designed the new library of compounds 2a-s (Table 1) to expand our SAR knowledge and identify new molecules endowed with improved activity profiles compared to the previously synthesized derivatives. With this aim, we decorated the pyrazolo [3,4d]pyrimidine scaffold by introducing a few substituents present in our lead compounds 1a-c and combined them with new chains, which had never introduced before. Herein, we report the C6 substituted compounds.
In detail, we planned to explore the chemical space around the C6 position by synthesizing some C6 thioalkyl derivatives (compounds 2a-g and 2p) as analog compounds of 1a and 1b and introducing different alkyl chains that had never been investigated before (compounds 2h-o and 2q-s). In particular, among the sulfur-containing compounds, we prepared the thiomethyl derivatives 2a and 2b to evaluate the effect of moving the chlorine atom from the meta (compound 1a) to the ortho (2a) and para (2b) positions. Furthermore, we investigated the effect of the presence of an unsubstituted (2c, h, j, k, n, p-s) or a substituted (2d-g, i, l, m, o) anilino ring at the C4 position. In particular, we decorated the aromatic ring in C4 with different halogen atoms and, in a few cases, with a hydroxyl group since this polar group was recently shown to increase the enzymatic activity of our in-house 4-amino pyrazolo [3,4-d]pyrimidines [12,13]. Finally, we functionalized unsubstituted anilino derivatives 2p-s with a halogen atom on the para position of the N1 2-chloro-2-phenylethyl side chain, affording structural analogs of compound 1c.
The synthesis of compounds 2a-s was performed following a multistep route reported in Schemes 1 and 2.
For the preparation of C6-thioalkyl derivatives 2a-g and 2p, we exploited a general chemical route, already reported by us for the synthesis of analog compounds [8,9]. In the last step, we reacted the 4-chloro pyrazolo [3,4-d]pyrimidines 3a-e [8,9,14,15] with the suitable aniline in the presence of absolute ethanol at reflux to afford the corresponding amino derivatives via a nucleophilic substitution (Scheme 1).

Chemistry
We designed the new library of compounds 2a-s (Table 1) to expand o knowledge and identify new molecules endowed with improved activity compared to the previously synthesized derivatives. With this aim, we decora pyrazolo [3,4-d]pyrimidine scaffold by introducing a few substituents present in o compounds 1a-c and combined them with new chains, which had never intr before. Herein, we report the C6 substituted compounds.
In detail, we planned to explore the chemical space around the C6 pos synthesizing some C6 thioalkyl derivatives (compounds 2a-g and 2p) as compounds of 1a and 1b and introducing different alkyl chains that had nev investigated before (compounds 2h-o and 2q-s). In particular, among the containing compounds, we prepared the thiomethyl derivatives 2a and 2b to eval effect of moving the chlorine atom from the meta (compound 1a) to the ortho ( para (2b) positions. Furthermore, we investigated the effect of the presenc unsubstituted (2c, h, j, k, n, p-s) or a substituted (2d-g, i, l, m, o) anilino ring a position. In particular, we decorated the aromatic ring in C4 with different haloge and, in a few cases, with a hydroxyl group since this polar group was recently sh increase the enzymatic activity of our in-house 4-amino pyrazolo [3,4-d]pyri [12,13]. Finally, we functionalized unsubstituted anilino derivatives 2p-s with a atom on the para position of the N1 2-chloro-2-phenylethyl side chain, affording st analogs of compound 1c.
The synthesis of compounds 2a-s was performed following a multiste reported in Schemes 1 and 2.

Enzymatic Activity
Pyrazolo [3,4-d]pyrimidines 2a-s were evaluated in enzymatic assays against Src, Bcr-Abl, and Fyn, using compound 1a as the reference. The percentage of inhibition was calculated at different concentrations (100, 10, and 1 µM) (Table S1, in the Supporting Information), and the K i was determined for the most promising derivatives (Table 2). Interestingly, compound 2a, possessing K i values of 0.423 and 0.419 µM for Src and Abl, respectively, emerged as a valuable dual Src/Bcr-Abl inhibitor [9,16]. On the other hand, the corresponding para substituted derivative 2b was inactive.
Unfortunately, other derivatives also generally showed low activity in cell-free assays. The introduction of an alkyl chain at C6 (compounds 2h-o and 2q-s) did not afford potent SFK/Bcr-Abl inhibitors. This SAR could be explained by the fact that alkyl groups cannot form polar interactions with the target and stabilize the complex. Furthermore, it is worth pointing out that the polar thioethylmorpholinyl chain (compounds 2e-g) also did not lead to effective Src/Fyn/Bcr-Abl inhibitors. This result is quite unexpected since other sets of 6-thioethylmorpholinyl pyrazolo [3,4-d]pyrimidines recently published by us [12,13] showed an activity profile towards Src and/or Bcr-Abl in the nanomolar range. In this case, we can hypothesize that the activity loss is due to the different substituents on the anilino ring. A careful comparison of previous and current data allowed us to shed light on this behavior and expand our SAR knowledge. Many of the best in-house thioethylmorpholinyl derivatives that are active as Src inhibitors bear a 2-chloro-5-hydroxy anilino group at C4 [12]. Modeling studies highlighted that the OH was able to form a hydrogen bond with a residue of glutamate of the catalytic pocket of Src, enhancing the affinity between the kinase and the inhibitor [7]. The current study further confirmed these in silico results: compound 2e was inactive towards all of the tested kinases, while its hydroxylated analog compound possessed a K i value of 190 nM towards Src [12]. Furthermore, data reported in Table 1 show that introducing a fluorine atom in the ortho or ortho/para positions of the anilino ring (compounds 2f and 2g) led to inactive compounds. We already reported that a 2-fluoro-3-hydroxy anilino derivative was less active than the corresponding chloro analog compound [12]; in the study, we also demonstrated that moving the fluorine to the para position or the presence of two fluorine atoms did not lead to positive effects on the activity.
Overall, this study allowed us to identify a low micromolar Src/Bcr-Abl inhibitor (2a). Because in-house C6 substituted pyrazolo-pyrimidines generally showed activity in tumors characterized by Src overexpression, we decided to investigate the activity of 2a in GBM cellular models.

SI388 Inhibits Src Kinase Activity in GBM Cellular Models
To test the ability of compound 2a (SI388) to inhibit Src activity in vitro, we took advantage of glioblastoma cellular models. Glioblastoma, indeed, was the first cancer type to be systematically studied by The Cancer Genome Atlas Research Network (TCGA) [17], who identified the RTK/Ras/PI3K pathway as having the most frequently deregulated signaling [4,18], with Src being a common node downstream of different RTKs [4,10].
U251 and T98G glioblastoma cell lines were treated with two ascending drug concentrations (10 nM and 1 µM) or DMSO for 24 h and Src activity was revealed by immunoblotting with a specific anti-phosho Tyr416-Src (anti-pY416Src) antibody. As shown in Figure 2 (panels A and B), we observed a reduction in Src phosphorylation on Tyr416 (Y416) after treatment with 2a in a dose-dependent manner. Similar results were also obtained upon treatment with 1a (SI83) (Figure 2, panels C and D), its analogue compound, and dasatinib (DAS) (Figure 2, panels E and F), a well-known commercially available Src inhibitor.

SI388 Inhibits Src Kinase Activity in GBM Cellular Models
To test the ability of compound 2a (SI388) to inhibit Src activity in vitro, we took advantage of glioblastoma cellular models. Glioblastoma, indeed, was the first cancer type to be systematically studied by The Cancer Genome Atlas Research Network (TCGA) [17], who identified the RTK/Ras/PI3K pathway as having the most frequently deregulated signaling [4,18], with Src being a common node downstream of different RTKs [4,10].
U251 and T98G glioblastoma cell lines were treated with two ascending drug concentrations (10 nM and 1 µM) or DMSO for 24 h and Src activity was revealed by immunoblotting with a specific anti-phosho Tyr416-Src (anti-pY416Src) antibody. As shown in Figure 2 (panels A and B), we observed a reduction in Src phosphorylation on Tyr416 (Y416) after treatment with 2a in a dose-dependent manner. Similar results were also obtained upon treatment with 1a (SI83) (Figure 2, panels C and D), its analogue compound, and dasatinib (DAS) (Figure 2, panels E and F), a well-known commercially available Src inhibitor.

SI388 Reduces Cell Growth and Viability in GBM Cell Lines
To further characterize the ability of SI388 to inhibit tumor growth, we performed clonogenic assays using T98G and U251 cell lines. Cells were starved and treated for 24 h with increasing concentration (10 nM, 1 µM, and 25 µM) of SI388 or DAS. The ability of single cells to proliferate and produce colonies was evaluated after 15 days, both in control conditions (treated with DMSO) and after treatment with different Src inhibitors. As shown in Figure 3 (panels A and B), both SI388 and DAS impaired cell growth and proliferation in a dose-dependent manner in both cell lines.

SI388 Reduces Cell Growth and Viability in GBM Cell Lines
To further characterize the ability of SI388 to inhibit tumor growth, we performed clonogenic assays using T98G and U251 cell lines. Cells were starved and treated for 24 h with increasing concentration (10 nM, 1 µM, and 25 µM) of SI388 or DAS. The ability of single cells to proliferate and produce colonies was evaluated after 15 days, both in control conditions (treated with DMSO) and after treatment with different Src inhibitors. As shown in Figure 3 (panels A and B), both SI388 and DAS impaired cell growth and proliferation in a dose-dependent manner in both cell lines.
Next, the ability of SI388 to inhibit cell viability was evaluated using an MTS assay. As shown in Figure 3C, SI388 significantly impaired cell viability after 72 h of drug treatment in a dose-dependent manner both is T98G and in U251 cells.  Next, the ability of SI388 to inhibit cell viability was evaluated using an MTS assay. As shown in Figure 3C, SI388 significantly impaired cell viability after 72 h of drug treatment in a dose-dependent manner both is T98G and in U251 cells.

SI388 Inhibits Src Kinase Activity in GBM Patient-Derived Cancer Stem Cells and Reduces Their Ability to form Neurospheres
Recently, gene expression subtyping highlighted that the mesenchymal GBM subtype shows higher Src pathway activation, resulting in enhanced sensitivity to dasatinib treatment compared to other GBM subtypes [19]. For this reason, we next evaluated the ability of SI388 to affect cell growth in a patientderived mesenchymal GBM cancer stem cells, named GBMSC83, cultured in non-adherent conditions to form neurospheres.
Firstly, we observed a significant reduction in phosphorylation of Src on Tyr416 upon a 1 µM SI388 treatment in this system; although, the inhibitory effect of SI388 is weaker compared to a 1 µM DAS treatment ( Figure 4A). We then demonstrated that both treatments significantly decreased the growth of spheres, resulting in an overall reduced diameter of the spheres compared to the control condition ( Figure 4B) and a reduced average number of cells per sphere ( Figure 4C). *** p < 0.001; **** p < 0.0001; ns, not significant.

SI388 Inhibits Src Kinase Activity in GBM Patient-Derived Cancer Stem Cells and Reduces Their Ability to form Neurospheres
Recently, gene expression subtyping highlighted that the mesenchymal GBM subtype shows higher Src pathway activation, resulting in enhanced sensitivity to dasatinib treatment compared to other GBM subtypes [19].
For this reason, we next evaluated the ability of SI388 to affect cell growth in a patientderived mesenchymal GBM cancer stem cells, named GBMSC83, cultured in non-adherent conditions to form neurospheres.
Firstly, we observed a significant reduction in phosphorylation of Src on Tyr416 upon a 1 µM SI388 treatment in this system; although, the inhibitory effect of SI388 is weaker compared to a 1 µM DAS treatment ( Figure 4A). We then demonstrated that both treatments significantly decreased the growth of spheres, resulting in an overall reduced diameter of the spheres compared to the control condition ( Figure 4B) and a reduced average number of cells per sphere ( Figure 4C).

SI388 Increases Cell Sensitivity to Ionizing Radiation (IR) Treatment
To further strengthen our findings, we next investigated if SI388 treatment could increase cell sensitivity to radiotherapy of GBM cells.
To this aim, T98G and U251 cells were pre-incubated or not with SI388 for 24 h and then irradiated (10 Gy). MTS assays revealed that SI388 pre-treatment increases the sensitivity to IR in both GBM cell lines; although, this result was only statistically significant in T98G cells ( Figure 5A). Figure 5A).

T98G cells (
GBMSC83 cells were pretreated with 1 µM SI388 or 1 µM DAS, or DMSO as a control, for 24 h and then irradiated (10 Gy). After 48 h from IR, cytofluorimetric analysis of An-nexinV-PI staining was performed. Interestingly, SI388, as well as dasatinib, sensitized cells to IR and significantly increased the percentage of cell death compared to the control samples ( Figure 5, panels B and C).  GBMSC83 cells were pretreated with 1 µM SI388 or 1 µM DAS, or DMSO as a control, for 24 h and then irradiated (10 Gy). After 48 h from IR, cytofluorimetric analysis of AnnexinV-PI staining was performed. Interestingly, SI388, as well as dasatinib, sensitized cells to IR and significantly increased the percentage of cell death compared to the control samples ( Figure 5, panels B and C).

Chemistry
All commercially available chemicals were used as purchased. DCM was dried over calcium hydride. Anhydrous reactions were run under a positive pressure of dry N 2 or argon. TLC was carried out using Merck TLC plates silica gel 60 F254. Chromatographic purifications were performed on columns packed with Merk 60 silica gel, 23-400 mesh, for flash technique. 1 H NMR and 13 C NMR spectra were recorded on a Brucker Avance DPX400 (at 400 MHz for 1 H and 100 MHz for 13 C) or using a Varian Gemini 200 (200 MHz for for 1 H) in DMSO-d 6 , CDCl 3 , or acetone-d6 as solvents as indicated. Chemical shifts (δ) were expressed in parts per million (ppm) relative to tetramethylsilane (TMS), which was used as the internal standard. Data are shown as follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, quint = quintet, sx = sextet, dd = doublet of doublets, tt = triple of triplets, bs = broad singlet), coupling constant (J) in Hertz (Hz), and integration. Elemental analysis for C, H, N, and S was determined using Thermo Scientific Flash 2000 and results were within ±0.4% of the theoretical value. All target compounds possessed a purity of ≥95%, which was verified by elemental analysis. Microwave irradiation experiments were conducted using a CEM Discover synthesis unit (CEM Corp., Matthews, NC, USA). The machine consists of a continuous focused microwave power delivery system with operator selectable power output from 0 to 300 W. The temperature of the contents of the vessels was monitored using a calibrate infrared temperature control mounted under the reaction vessel. All of the experiments were performed using a stirring option whereby the contents of the vessel are stirred by means of a rotating magnetic plate located below the floor of the microwave cavity and a Teflon-coated magnetic stir bar in the vessel.

General Procedure for the Synthesis of Final Compounds 2f,g
The appropriate 3-aminophenol derivative (5 mmol) was added to a solution of 4-chloro- (1 mmol, 438 mg) in absolute ethanol (10 mL), and the mixture was refluxed for 3−5 h. After cooling to room temperature, the solvent was evaporated under reduced pressure and the crude was solved in ethyl acetate (10 mL), washed with 0.1 N HCl solution (2 × 10 mL), 1 N NaOH solution (10 mL) and brine (2 × 10 mL), dried (MgSO4), filtered, and concentrated under reduced pressure to give a brown oil which crystallized by adding a 1:1 mixture of diethyl ether/petroleum ether (bp 40−60 • C) affording yellow solid.  13
To avoid peptide adsorbing to the plastic surface, protein low-binding tubes were used. ADP-Glo kinase assay (Promega) was then used to detect kinase activity according to the manufacturer's instructions with minor modifications. In detail, reactions were transferred to white 384 well-plates and stopped by adding 10 µL of ADP-Glo reagent (Promega) for 50 min at room temperature. A total of 20 µL of detection reagent (Promega) was then added for 30 min and the reaction read using a GloMax Discover microplate reader (Promega). Data were plotted using GraphPad Prism 5.0. IC 50 values were obtained according to Equation (1): where v is the measured reaction velocity, V is the apparent maximal velocity in the absence of inhibitor, I is the inhibitor concentration, and IC 50 is the 50% inhibitory dose. Compounds tested were assumed to act as fully ATP-competitive inhibitors. Therefore, K i values were calculated according to Equation (2): where K i is the affinity of the inhibitor to the enzyme, S is the ATP concentration, and K m is the affinity of ATP calculated according to the Michaelis-Menten equation.

Clonogenic Survival Assay
For colony formation, control and irradiated (5 Gy) cells were seeded at the concentration of 1000 cells/dish in 6-well culture plates and incubated at 37 • C, 5% CO 2 for 10-15 days.
Crystal violet solution (10% (v/v) methanol and 0.5% Crystal Violet) was used to fix and stain colonies, upon 20 min incubation. The stained colonies (>50 cells) were counted under a microscope or using Image "Colonyarea" plugin. Data were expressed as the mean and SD of three independent experiments.

MTS Assay
Cells were plated in 96 multiwells (1000 cells/well) in 100 µL of the complete medium. The following day, after 24 h of starvation, the cells were treated with DMSO, dasatinib or SI388 at different concentrations (10 nM, 1 µM, 25 µM). After 72 h, cells were treated with tetrazolium (Promega) for 1-2 h. The quantity of formazan product is directly proportional to the number of living cells in the culture. The plate is then read with a plate reader measuring the absorbance of 490 nm for the different conditions. 3.3.6. Neurosphere Formation, Cell Counting and Sphere Size Measurement GBMSC83 cells were seeded at the density of 1 × 10 4 cells/well in a 96-well ultralow attachment plate, and the following day they were treated with SI388 1 µM or dasatinib 1 µM. After 72 h from treatment, cells were stained with Hoechst 33,342 (Thermo Fisher Scientific) and images were acquired with Fluorescence microscopy (ZEISS) and processed with Fiji version 2.3. Hoechst positive nuclei were counted by particle detection instrument of ComDet Plugin of Fiji. The bright fields were acquired and processed by Cell Profiler version 2.1.1 to measure spheres' diameter and reported as the fold change from the control condition.
3.3.7. Cell Death Analysis 1 × 10 6 cells were analyzed for cell death 48 h from ionizing radiation (10 Gy) upon staining with Annexin V-APC-propidium iodide (PI) kit, according to manufacturer instructions (eBioscience™ Annexin V apoptosis detection kits, ThermoFisher Scientific, Waltham, MA, USA). Unstained samples were used as the control. CytoFLEX S (Beckman Coulter, Milan, Italy) instrument was used to quantify the double staining. Quality control was evaluated using CytoFLEX Daily QC Fluorospheres (Beckman Coulter). CytExpert version 2.2 software (Beckman Coulter) was used to analyze FCS files. Dead cells were graphed as fold change to control conditions.

Statistical Analyses
All data were analyzed and presented as mean ± S.E.M. The significance of the differences between populations of data were assessed according to the paired or unpaired Student's two-tailed t-test with a level of significance of at least p ≤ 0.05.

Conclusions
We performed a lead optimization study that afforded the discovery of the compound SI388 (2a), endowed with an improved activity compared to the parent compound SI83 (1a). This promising cell-free data prompted us to investigate the efficacy of SI388 in 2D and 3D tumor models characterized by Src overexpression. For this purpose, we decided to perform our biological experiments using GBM cellular models, which are known to be characterized by Src hyperactivation.
In this work, we tested SI388 compound on two commercial GBM cell lines (T98G and U251 cells) and on mesenchymal patient-derived cancer stem cells (GBMSC83 cells). Interestingly, SI388 can efficiently target Src and its functionality both in T98G and U251 cell lines as well as in GBMSC83 patient-derived neurospheres. Importantly, SI388 can sensitize patient derived GBM cells to IR treatment, similarly to dasatinib. Overall, this study identifies SI388 as a promising candidate for Src kinase inhibition and as a potential new tool to enhance responses to radiotherapy, which is worthy of further investigation.