Discovery of New 1,4,6-Trisubstituted-1H-pyrazolo[3,4-b]pyridines with Anti-Tumor Efficacy in Mouse Model of Breast Cancer

Purine analogues are important therapeutic tools due to their affinity to enzymes or receptors that are involved in critical biological processes. In this study, new 1,4,6-trisubstituted pyrazolo[3,4-b]pyridines were designed and synthesized, and their cytotoxic potential was been studied. The new derivatives were prepared through suitable arylhydrazines, and upon successive conversion first to aminopyrazoles, they were converted then to 1,6-disubstituted pyrazolo[3,4-b]pyridine-4-ones; this served as the starting point for the synthesis of the target compounds. The cytotoxic activity of the derivatives was evaluated against several human and murine cancer cell lines. Substantial structure activity relationships (SARs) could be extracted, mainly concerning the 4-alkylaminoethyl ethers, which showed potent in vitro antiproliferative activity in the low μM level (0.75–4.15 μΜ) without affecting the proliferation of normal cells. The most potent analogues underwent in vivo evaluation and were found to inhibit tumor growth in vivo in an orthotopic breast cancer mouse model. The novel compounds exhibited no systemic toxicity; they affected only the implanted tumors and did not interfere with the immune system of the animals. Our results revealed a very potent novel compound which could be an ideal lead for the discovery of promising anti-tumor agents, and could also be further explored for combination treatments with immunotherapeutic drugs.


Introduction
Nitrogen containing fused heterocyclic rings that present a structural analogy to adenine are usually considered privileged scaffolds, since they are often found in many biologically active and clinically useful derivatives. Among them, pyrazolopyrimidine and the closely related pyrazolopyridine pharmacophores are present in druggable small molecule entities, marketed drugs, such as Allopurinol, Zaleplon, Sildenafil, Ibrutinib, and compounds such as Etazolate, that undergo clinical evaluation ( Figure 1).
Several pyrazolo [3,4-b]pyridine-based derivatives that bear diverse substitutions were reported to exhibit potent antiviral [1,2] and antibacterial [3,4] properties; to inhibit important enzymes, such as phosphodiesterase-4 [5], or neutrophil elastase [6]; and to serve as ligands for A1-adenosine [7] or prostaglandin E2 receptor 1 [8]. The anti-cancer potential of this class of compounds is of particular interest as they show antiproliferative activity, Our long-term aims are the discovery of novel purine analogues with in vitro and in vivo cytotoxic activity [16,17]. In this respect, we have previously reported on numerous pyrazolo [3,4-c]pyridines with potent antiproliferative activity, bearing suitable substituents at critical positions of the central scaffold [18,19]. We observed that the most interesting compounds possessed a 3-aryl group and were also substituted at positions 5 and/or 7 of the original pyrazolo [3,4-c]pyridine nucleus. As a continuation of this effort we have designed, synthesized, and evaluated the cytotoxic activity of a new series of compounds which bear the pyrazolo [3,4-b]pyridine ring system as their central core. Based on our earlier findings, we have inserted a variety of substituents in specific sites of this scaffold, taking care to preserve the above mentioned substitution pattern, in order to investigate whether the new structural analogues maintain the cytotoxic activity, thereby assisting in the extraction of helpful structure activity relationships. We have also evaluated the efficacy of our newly synthesized analogues in vitro, and the three most promising among the new compounds were subsequently evaluated in an orthotopic syngeneic mouse model of breast cancer, prompted by our in vitro results.

Synthetic Procedures and Analytical Data
Melting points were determined on a Büchi apparatus and are uncorrected. 1 H NMR spectra and 2D spectra were recorded on a Bruker Avance III 600 or a Bruker Avance DRX 400 instrument, whereas 13 C NMR spectra were recorded on a Bruker Avance III 600 spectrometer in deuterated solvents and were referenced to TMS (δ scale). The signals of 1 H and 13 C spectra were unambiguously assigned by using 2D NMR techniques: 1 H 1 H COSY, NOESY, HMQC, and HMBC. Mass spectra were recorded with a LTQ Orbitrap Discovery instrument, possessing an Ionmax ionization source. The purity of the key compounds (>95%) was determined on a Thermo Finnigan ® HPLC System (P4000 Pump, AS3000 Autosampler, UV Spectra System UV6000LP detector, Chromquest™ 4.1 Software); Fortis ® UniverSil HS-C18 (150 mm, 4.6 mm, 5 um); mobile phase 1% acetic acid in water/acetonitrile; flow rate 1 mL/min; column temperature 25 • C; injection volume 10 µL; absorbance at 254 nm). Flash chromatography was performed on Silicagel ACROS OR-GANICS 40-60 µm and 60-200 µm, 60A. Analytical thin layer chromatography (TLC) was carried out on precoated (0.25 mm) Merck silica gel F-254 plates.

In Vitro Cytotoxicity (MTT) Assays
The human HCT116 colon cancer cell line and PC-3 prostate cancer cell line were obtained from the American Type Cell Culture (ATCC, Bethesda, MD, USA). HCT116 and PC-3 cell lines were grown at 37 • C in 5% CO 2 using Roswell Park Memorial Institute 1640 medium (RPMI 1640) and Dulbecco's modified Eagle's medium F/12 (DMEM/F-12) containing 10% fetal bovine serum (FBS). MEFs and KPC cells were cultured in a 5% CO 2, with DMEM medium containing 10% fetal bovine serum (FBS), 100 U/mL Penicillin and Pharmaceutics 2023, 15, 787 9 of 18 100 mg/mL Streptomycin at 37 • C. EO771 cells were purchased from CH3 BioSystems and cultured in complete RPMI 1640 medium supplemented with 10% fetal calf serum, 100 U/mL Penicillin and 100 mg/mL Streptomycin. To test the inhibitory activities of compounds using a cell-based assay, MTT assays were performed for cell viability. Briefly, in 96-well plates, HCT116 cells were plated at a density of 1500 per well, PC-3 cells were plated at a density of 750 per well; KPC and MEFs were plated at a density of 3000 cells per well; and EO771 were seeded at a density of 2500 cells per well. The differences in the initial seeding cell numbers reflect the differences in the doubling time of cells, given that we sought to have wells that were close to but not confluent at the end of the experiment. After 24 h, cells were treated with the indicated compounds in a dose-dependent manner for 72 h and 96 h. Viable cell numbers were determined by tetrazolium conversion to its formazan dye. All the experiments were performed three times, and the tested concentrations in each experiment were evaluated in quadruplicated wells. Mouse embryonic fibroblasts (MEFs) and KPC pancreatic cells (Kras(mut); Pdx Cre) are cell lines derived from our C57/Bl6 mice and the double transgenic mouse pancreatic cancer model, respectively [21][22][23].

Mouse Models
All studies were approved by the National Hellenic Research Foundation Animal Care and Use Committee. The study protocol was approved by the local ethics committee (Athens Prefecture Veterinarian Service; (431956/17-05-2022.) Animal care was provided in accordance with the procedures outlined in the "Guide for Care and Use of Laboratory Animals (National Research Council; 1996; National Academy Press; Washington, DC, USA). C57BL/6 female mice were purchased from Jackson Laboratory. Mice at 6-8 weeks of age were randomly assigned to treatment or control groups. As indicated by the performed power analysis (clinical calculator/clincalc.com), 5 mice per group were used in each of the in vivo experiments. We repeated the in vivo experiments three times and found similar results. In the results section, we present our data from one experiment. The EO771 were cultured in complete RPMI 1640 medium supplemented with 10% fetal calf serum, 50 mM 2-mercaptoethanol, 100 U/mL Penicillin, and 100 mg/mL Streptomycin. Murine EO771 (5 × 10 5 ) were orthotopically inoculated at the fourth mammary fat pad of 6-8 weeks old mice. The cells were resuspended in PBS. Matrigel (Corning Inc., Corning, NY, USA) was added at 1:3 dilution to facilitate the inoculation process. Matrigel, an extract of basement membrane proteins, was used as a cell carrier medium for cell transplantation studies by forming a 3D gel at 37 • C that facilitated the inoculation. Tumor size was measured using a digital caliper, and tumor volume (mm 3 ) was calculated by the following equation: L*W*H*π/6. The mice were monitored on a daily basis for any signs of discomfort and the orthotopic tumors were also monitored routinely for any signs of ulcerations or any other type of wounds. The mice were also weighted once a week. Per our guidelines, any mouse found with an ulcerated wound, 30% weight loss, or visual signs of discomfort (slow reflexes, not walking normally, hunched back, or rough coat) was immediately excluded from the experiment and euthanized. Despite the strict rules, no mice were excluded from our experiments. All the treatments were well tolerated by the mice, showing no signs of toxicity.

Treatment of EO771 Tumor-Bearing Mice
Treatment was initiated when tumors reached~20 mm 3 . Animals were treated with 9b, 9c and 9e peritumorally by administration of 100 µg of the aforementioned compounds in Matrigel (Corning Inc.), used in 1:4 dilution, every 4 days. The mice were sacrificed when the primary tumor reached a 2 cm diameter, or at any other humane endpoints as listed in the ACUC-approved animal protocol, such as 20% weight loss or acute morbidity.

Histology and Immunohistochemistry Staining
Tissue samples, including tumors, were fixed in 10% neutral buffered formalin (NBF, Sigma), then routinely processed and paraffin embedded. Tumor and lung sections were dewaxed and rehydrated, then stained with hematoxylin and eosin (H&E). For immunohistochemistry, sections were antigen-retrieved with the heat-induced or enzymatic method. Peroxidase activity was blocked using 1.5% hydrogen peroxide. Sections were blocked with different blocking protocols, depending on the antibody. Staining was performed using the following anti-mouse antibodies: anti-Ki67 (Cell Signaling, 9449) (1:1000 dilution) and anti-Caspase 3 (Cell Signaling, 9661) (1:800 dilution). A polymer-based detection kit, which consists of horseradish peroxidase-conjugated polymers was used for detection. To determine proliferation indices, Ki67-positive and Ki67-negative cells were counted using ImageJ software (US National Institutes of Health) in 8-10 representative fields of all tumors (on average,~3000 nuclei were counted per specimen). A similar approach was followed to evaluate the % percentage of apoptotic cells.

Statistical Analysis
Statistical analyses and graph generation were performed with GraphPad Prism 9.2.0 (San Diego, CA, USA). Tumor areas were plotted as mean ± standard error of the mean (SEM) for each data point, and tumor growth curves were compared using mixed effects ANOVA. Differences were evaluated by 1-way ANOVA or unpaired parametric Student's t test. The p-values were calculated for multiple comparisons using Tukey's multiple comparisons test.

Synthesis of the Novel Compounds
Upon a literature search for the optimum way to accomplish the synthesis of pyrazolo [3,4b]pyridines [24,25], we have decided to prepare the target compounds by using commercial phenylhydrazine (1a) or 3-fluorophenylhydrazine (1b) as starting materials (Scheme 1). Each hydrazine was converted to the corresponding pyrazole-4-carbonitriles (2a,b) [26,27], and then, upon hydrolysis and subsequent decarboxylation, converted to the pyrazoles 3a,b following a previously reported methodology [20,28]. The above mentioned pyrazoles were then treated with the Meldrum's acid dimethythiomethylene-derivative 4 [29] to provide the intermediate dioxanediones 5a,b. The remaining methylthio group of compounds 5a,b were then displaced upon reaction with aniline, and the derived dioxanediones 6a,b were subjected to thermal cyclization and converted to the pyrazolopyridinols 7a,b.
The pyridinones 7a,b were treated with K 2 CO 3 and suitable bromides to provide the corresponding ethers through their enolates 8a-f. The unambiguous confirmation for the O-alkylation was evidenced from 2D NMR data. Thus, from the HMBC spectrum, a crosspeak between the most downfield ethyloxy side chain methylene with C-4 was obvious, and additionally, data from the nOe spectrum showed that the above-mentioned methylene correlates to both H-5 and H-3 as well. Additionally, the bromides 8c,f were used for the introduction of selected amines, namely aniline, 4-methylpiperazine, or cyclohexylamine, which resulted in the aminoderivatives 9a-f. Along with the aim of expanding the structural diversity of this scaffold, compounds 7a,b were allowed to react with Lawesson's reagent, and the resulting thiols 10a,b were reduced by using Raney nickel as the catalyst to provide the corresponding 1,6-disubstituted pyrazolopyridines 11a,b. An alternative pathway for the synthesis of a 7a modified analogue is also presented in the Supporting Information section in Supplementary Materials (Scheme S1). The pyridinones 7a,b were treated with K2CO3 and suitable bromides to provide the corresponding ethers through their enolates 8a-f. The unambiguous confirmation for the O-alkylation was evidenced from 2D NMR data. Thus, from the HMBC spectrum, a crosspeak between the most downfield ethyloxy side chain methylene with C-4 was obvious, and additionally, data from the nOe spectrum showed that the above-mentioned

Evaluation of the Antiproliferative Efficacy of the New pyrazolo[3,4-b]pyridine Derivatives In Vitro
The cytotoxic activity of all target compounds, 7a,b, 8a-f, 9a-f, 10a,b, and 11a,b, was initially evaluated against the prostatic PC-3 and the colon HCT116 human cancer cell lines. This screening provided very interesting information concerning the structural requirements for activity. The majority of the compounds did not exhibit significant activity; they possessed IC 50 values greater than 10 µM, with the remarkable exception of two couples of analogously substituted derivatives, specifically, 9b,c and 9e,f. The latter were endowed with low µM activity in the IC 50 range of 0.75-4.55 µM ( Table 1). All four compounds are 4-alkylaminoethoxy derivatives, providing an indication that the presence of a 4-arylaminoether, a 4-arylalkyl or 4-alkyl ether do not encourage activity. This is also true in the case of pyrazolopyridinones 7a,b, the corresponding thiones 10a,b, and the  4-unsubstituted analogues 11a,b. These active analogues were subsequently tested against two additional cancer cell lines and one non-cancerous murine cell line. The first murine cancer cell line, EO771, is a breast cancer cell line, syngeneic to C57/Bl6; the second cell line, KPC, is a pancreas cancer cell line, syngeneic to C57/Bl6. The non-cancerous murine cell line consists of mouse embryonic fibroblasts (MEFs) and is also derived from C57/Bl6 mice. Cell viability of the four derivatives against normal mouse embryonic fibroblasts (MEFs) was also examined. Three compounds were found non-toxic (9b, 9c and 9e, presenting IC 50 >200 µM), and the fourth showed severe cytotoxicity with an IC 50 = 11.2 µM.  Figure 2D) in mouse embryonic cells and EO771 mouse breast cancer cell. The difference in the cell viability between those cell lines indicate possible anti-breast cancer efficacy of these compounds and minimal off target effects. It is noteworthy that those compounds were not as effective in reducing the cell proliferation of the KPC mouse pancreatic cell line, suggesting possible breast cancer specificity. Based on our in vitro data, we decided to investigate the efficacy of 9b, 9c, and 9e in vivo in our tumor bearing EO771 mouse model. Only 9f was excluded from the subsequent in vivo experiments because of its low IC 50 values on the MEFs, so that potential increased toxicity or lethality to the experimental animals could be avoided.

Evaluation of the Anti-Cancer Efficacy of the Most Active Analogues In Vivo
We next sought to investigate the efficacy of the novel compounds in mouse syngeneic models of cancer, enabling us to monitor the effects of the treatment not only to the tumor growth itself, but also to the immune system of the mouse, which better simulated the interactions we often monitor in clinical settings. For subsequent in vivo analysis, we chose EO771 cells as they showed very good IC 50 values and specificity in comparison to the pancreatic cell line or the non-cancerous fibroblasts. We thus established mouse tumors by using EO771 cells and orthotopically inoculated into the mammary fat pad of the mouse. At 6-7 days after inoculating 5 × 10 5 cells into the mammary fad pad, the mice develop palpable tumors; immediately after this observation, we initiated our treatment with the selected compounds. Based on our previous in vivo studies [21], and the fact that all the in vivo evaluated agents showed minimal toxicity on the mouse embryonic fibroblasts, we initiated a treatment protocol of intraperitoneal delivery of 100 µg every four days until day 26 post-inoculation. As presented in Figure 3, all the tested compounds resulted in a strong inhibition of tumor growth ( Figure 3A). The compound-specific tumor growth curves revealed that 9b was more potent in inhibiting tumor growth in comparison to the other analogues. (Figure 3B-D). The treatment schedule is depicted in Figure 3E.

Evaluation of the Anti-Cancer Efficacy of the Most Active Analogues In Vivo
We next sought to investigate the efficacy of the novel compounds in mouse syngeneic models of cancer, enabling us to monitor the effects of the treatment not only to the tumor growth itself, but also to the immune system of the mouse, which better simulated the interactions we often monitor in clinical settings. For subsequent in vivo analysis, we chose EO771 cells as they showed very good IC50 values and specificity in comparison to the pancreatic cell line or the non-cancerous fibroblasts. We thus established mouse tumors by using EO771 cells and orthotopically inoculated into the mammary fat pad of the mouse. At 6-7 days after inoculating 5 × 10 5 cells into the mammary fad pad, the mice develop palpable tumors; immediately after this observation, we initiated our treatment

Histological Analysis of the Tumors
The histology of the tumors revealed that all treatments affected both the proliferation of the cancer cells and their overall survival. Representative pictures of the Ki67 staining are depicted in Figure 4A. The proliferation index, measured by Ki67 immunohistochemistry (Figure 4B), showed that the control tumors had increased proliferation in contrast to the treated tumors (76.74% control vs. 41.95% 9b, 43.47% 9c and 41.08% 9e). The reduced proliferation rates of all treated groups were statistically significant when compared to the untreated controls (p = 1.23 × 10 −12 , p = 3.3 × 10 12 and p = 1.53 × 10 −13 for 9b, 9c and 9e respectively. growth curves revealed that 9b was more potent in inhibiting tumor growth in comparison to the other analogues. (Figure 3B-D). The treatment schedule is depicted in Figure  3E.  The histology of the tumors revealed that all treatments affected both the proliferation of the cancer cells and their overall survival. Representative pictures of the Ki67 staining are depicted in Figure 4A. The proliferation index, measured by Ki67 immunohistochemistry ( Figure 4B), showed that the control tumors had increased proliferation in contrast to the treated tumors (76.74% control vs 41.95% 9b, 43.47% 9c and 41.08% 9e). The reduced proliferation rates of all treated groups were statistically significant when compared to the untreated controls (p = 1.23 × 10 −12 , p = 3.3 × 10 12 and p = 1.53 × 10 −13 for 9b, 9c and 9e respectively. We next decided to analyze the apoptotic rate of cancer cells by using cleaved caspase-3 ( Figure 5). For that purpose, we evaluated the staining in the periphery of the tumors rather than the staining on the central necrotic part. (Figure 5A). We observed an increase of the apoptotic cells in all the treated groups. The cleaved caspase-3 staining revealed a significant increase of the apoptotic cells (p = 1.32 × 10 −5 , p = 0.03, p = 0.002) in 9b, 9c, and 9e, respectively. The mean percentages of apoptosis were 6.1% for the controls, 9.9% for 9b treated, 7.68% for 9c, and 8.37% for 9e respectively. Most importantly, the 9b treated tumors showed statistically significant increase in apoptosis in comparison to the other treated groups (p = 0.009 with 9c and p = 0.049 with 9e), which is also in line with our in vitro experiments. We next decided to analyze the apoptotic rate of cancer cells by using cleaved caspase-3 ( Figure 5). For that purpose, we evaluated the staining in the periphery of the tumors rather than the staining on the central necrotic part. (Figure 5A). We observed an increase of the apoptotic cells in all the treated groups. The cleaved caspase-3 staining revealed a significant increase of the apoptotic cells (p = 1.32 × 10 −5 , p = 0.03, p = 0.002) in 9b, 9c, and 9e, respectively. The mean percentages of apoptosis were 6.1% for the controls, 9.9% for 9b treated, 7.68% for 9c, and 8.37% for 9e respectively. Most importantly, the 9b treated tumors showed statistically significant increase in apoptosis in comparison to the other treated groups (p = 0.009 with 9c and p = 0.049 with 9e), which is also in line with our in vitro experiments.

Conclusions
In this work, a number of substituted pyrazolo[3,4-b]pyridines were prepared and tested for their antiproliferative activity. We decided not to alter the nature of the substituent in positions 1 and 6 of the scaffold; thus, we inserted 1-phenyl or 1-(3-fluorophenyl) group together with 6-phenylamino group in all target compounds. In total, we used a variety of 4-substituents and have studied their impact on the biological activity. Interest- Figure 5. Panel (A) shows representative IHC images of cleaved caspase-3 staining of control and treated tumors. In (B), the actual percentages of cleaved caspase-3 per field were calculated. Scale bar 300 µm. To determine apoptotic cells, cleaved caspase-3 positive and negative cells were counted using ImageJ software (US National Institutes of Health) in 8-10 representative fields from each sample. All the samples were included in the analysis.

Conclusions
In this work, a number of substituted pyrazolo [3,4-b]pyridines were prepared and tested for their antiproliferative activity. We decided not to alter the nature of the substituent in positions 1 and 6 of the scaffold; thus, we inserted 1-phenyl or 1-(3-fluorophenyl) group together with 6-phenylamino group in all target compounds. In total, we used a variety of 4-substituents and have studied their impact on the biological activity. Interestingly, the in vitro evaluation revealed that only 4-alkylaminoethoxy derivatives possessed strong cytotoxicity against the four cancer cell lines tested in the low µM range, while the normal cell line (normal mouse embryonic fibroblasts) remains practically unaffected. This finding prompted us to study the in vivo efficacy of the three most active analogues in a mouse breast cancer model. All compounds, and more profoundly, the 4-(4-methylpiperazin-1-yl)ethoxy derivative 9b, presented a strong inhibition of tumor growth and induced apoptosis. It should be noted that these compounds do not show systemic toxicity and did not interfere with the immune system of the animals. More comprehensive structureactivity relationships, e.g., the necessity of the 1-phenyl (or fluorophenyl) substituent, as well as of the 6-phenylamino group, is currently under active investigation in our laboratories, with the aim of assisting in structure optimization and elucidating the molecular mode of action of this class of compounds.