Potential of Substituted Quinazolines to Interact with Multiple Targets in the Treatment of Cancer

Abstract

The efficacy of quinazoline-based antiglioma agents has been attributed to their effects on microtubule dynamics.^{1, 2} The design, synthesis and biological evaluation of quinazolines as potent inhibitors of multiple intracellular targets, including microtubules and multiple RTKs, is described. In addition to the known ability of quinazolines 1 and 2 to cause microtubule depolymerization, they were found to be low nanomolar inhibitors of EGFR, VEGFR-2 and PDGFR-β. Low nanomolar inhibition of EGFR was observed for 1-3 and 9-10. Compounds 1 and 4 inhibited VEGFR-2 kinase with activity better than or equal to that of sunitinib. In addition, compounds 1 and 2 had similar potency to sunitinib in the CAM angiogenesis assay. Multitarget activities of compounds in the present study demonstrates that the quinazolines can affect multiple pathways and could lead to these agents having antitumor potential caused by their activity against multiple targets.

Graphical Abstract

Introduction

Angiogenesis is the formation of new blood vessels from existing vasculature and is an essential process for tumor growth and metastasis.³ Proangiogenic growth factors such as VEGF, PDGF, and EGF are released by tumors under hypoxic conditions. These growth factors bind to and activate their respective RTKs: VEGFR, PDGFR-β and EGFR, leading to a cascade of events that promote angiogenesis and tumor growth, survival and metastasis.^4-6 RTK inhibitors that prevent angiogenesis established a new paradigm in cancer chemotherapy.^{7, 8} The use of multi-RTK inhibitors in cancer chemotherapy offers an advantage in overcoming resistance mechanisms that limit single RTK inhibitors including, point mutations in the ATP binding site and upregulation of additional RTKs.⁹ Thus, the use of multi-RTK inhibitors has emerged as an important approach in cancer chemotherapy and is validated by the approval of several multi-RTK inhibitors such as axitinib^{10, 11} (VEGFR, PDGFR, c-kit), pazopanib¹² (PDGFR, VEGFR, c-kit, FGFR), vandetanib¹³ (VEGFR, EGFR, RET kinase), sunitinib^{14, 15} (VEGFR, PDGFR and c-kit) and sorafenib^{16, 17} (VEGFR-2, PDGFR β and Raf kinase), among others. The cytostatic activity of RTK inhibitors does not kill cancer cells and thus, they need to be combined with cytotoxic chemotherapeutic agents and/or radiation to improve therapeutic outcomes.^18-21

Microtubules are essential cytoskeletal components in eukaryotic cells, and these cytoskeletal components are involved in cellular processes during both interphase and mitosis.²² In interphase, microtubules are involved in intracellular trafficking of proteins and organelles. In dividing cells, they make up the mitotic spindle and are essential for the separation of daughter chromosomes.^{22, 23} Microtubule targeting agents (MTAs) act by interfering with crucial processes involved in cell division and cellular transport in interphase.^22-24

The strategy of using a combination of drugs, each acting at a different biological target, has been successfully applied in the treatment of many diseases.^{25, 26} The basic premise for such an approach is to overcome alternate pathways that may be activated when only one pathway is inhibited. Similarly, combination chemotherapy with antiangiogenic and cytotoxic agents act by simultaneously inhibiting multiple modes, thereby inhibiting the growth of tumors dependent on multiple signaling pathways.^18-21

The vascular normalization theory proposes that the antiangiogenic effects of RTK inhibitors causes pruning of abnormal blood vessels and transiently normalizes the structural and functional integrity of the vasculature.²⁷ This normalization restores blood flow to the tumors, improves delivery of co-administered cytotoxic agents and thus provides an explanation, in part, of the improved efficacy observed with combination chemotherapy with antiangiogenic agents.^{27, 28} Multi-target directed agents are rationally designed analogs that incorporate essential pharmacophores for inhibiting two or more biological targets.²⁹ In cancer chemotherapy, such agents may help in preventing cancer cells from developing resistance, lowering the risk of drug-drug interactions, improving patient compliance and have a lower side effect profile.

Single agents with dual microtubule targeting and antiangiogenic activities are highly desirable because these agents can exert their cytotoxic effects throughout the tumor following normalization of tumor via the antiangiogenic component.³⁰ Clinical evaluation of the safety and efficacy of metronomically dosed cytotoxic agents in conjunction with antiangiogenic agents such as sunitinib and sorafenib has also been explored.^31-33 The dose-limiting cytotoxicity of conventional combination chemotherapy could also potentially be avoided because these agents work by complementary mechanisms.

Multi-target directed ligands with dual cytotoxic and antiangiogenic activities have been shown to significantly reduce tumor growth, metastasis and angiogenesis in tumor xenograft mouse models, without any significant toxicity.^{34, 35} Because of their crucial role in angiogenesis, VEGFR-2, PDGFR-β and EGFR were chosen as targets in the current studies. Due to the clinical utility of cytotoxic tubulin inhibitors, such as paclitaxel, in combination with antiangiogenic agents, tubulin was chosen as the cytotoxic target.^36-38

Microtubule targeting agents, particularly those binding to the colchicine site, have been shown to act on tumor vasculature and several agents were tested in clinical trials for efficacy in glioblastoma multiforme (GBM),³⁹ but none has yet advanced past phase II evaluation. Compound 1 (verubulin, Figure 1) was reported as a potent inducer of apoptosis, binding at the colchicine site and inhibiting tubulin polymerization in cells (T47D EC₅₀ = 2 ± 0.1 nM).² The compound is not a substrate for multidrug resistance transporters such as P-glycoprotein (Pgp) and breast cancer resistance protein (BCRP1). In addition, compound 1 is able to cross the BBB to attain 30-fold higher concentrations in the brain compared to the plasma.² These pharmacokinetic parameters led to the evaluation of 1 in GBM and other cancers, but failed to achieve orphan drug status for GBM.⁴⁰

Figure 1.

Lead compound 1

The quinazoline scaffold has also been widely explored as the basis for RTK inhibitory activity as exemplified by gefitinib (EGFR), erlotinib (EGFR) and vandetanib (VEGFR-2, VEGFR-3 and EGFR) (Figure 2). Structure-activity relationships with the quinazoline scaffold using N⁴-H analogs as EGFR inhibitors suggest that the quinazoline scaffold has bulk tolerance at the 6- and 7-positions.^{41, 42} However, the corresponding N⁴-CH₃ analogs were not explored for their RTK inhibitory activities. Separately, we have reported various scaffolds such as furo[2,3-d]pyrimidines³⁴ and pyrrolo[3,2-d]pyrimidines^{35, 43} with N⁴-CH₃ that are dual acting antitubulin compounds with multi-RTK inhibitory activities. Thus, on the basis of the structural similarities between 1 and the quinazoline based RTK inhibitors (Figure 2), it was of interest to determine the RTK inhibitory properties (if any) of 1 and its antiangiogenic attributes as well.

Figure 2.

Representative quinazoline based clinically used RTK inhibitors

Compound 1 was docked in the binding site of EGFR (Figure 3) and had a docked pose similar to erlotinib. The 2-CH₃ group of 1 makes hydrophobic interactions with Ala719, the N1 and N3 make hinge region binding interactions with Met769 and a water mediated H-bond with Thr766, respectively. The 4-position aniline substitution is oriented in the pocket in a manner similar to the 3-ethynyl substituted aniline in erlotinib and lies in a pocket lined by residues Lys721 and Asp831. The quinazoline scaffold of 1 is stabilized by hydrophobic interactions with Leu820, Leu768 and Met769. The docked score of 1 was −8.68 kcal/mol, close to the −9.29 kcal/mol score of erlotinib.⁴⁵

Figure 3.

Docked pose of lead compound 1 (purple) superimposed on the crystallized ligand erlotinib (cyan) in EGFR (PDB: 1M17)⁴⁴

The docked pose of 1 in the binding site of VEGFR-2 (Figure 4)⁴⁶ shows that the quinazoline scaffold binds to the pocket where the thiobenzamide of axitinib (a potent VEGFR inhibitor) binds and interacts with the backbone of the hinge binding site residue Phe1047. In addition, the quinazoline scaffold makes a cation-pi interaction with the sidechain of Lys868. The 4-position aniline substituent of 1 interacts with Val848, Cys919, Leu1035 and Phe1047 and binds in the pocket where the indazole scaffold of axitinib binds. The docked score of 1 in VEGFR-2 was −10.71 kcal/mol as compared to axitinib, which had a docked score of −13.25 kcal/mol.

Figure 4.

Docked pose of 1 (orange) superimposed on the crystallized ligand axitinib (cyan) in VEGFR-2 (PDB: 4AG8)⁴⁶

With the docked scores reported above, it was important to evaluate 1 as an RTK inhibitor. As expected, a preliminary evaluation of 1 in an in vitro assay for inhibition of EGFR, VEGFR-2 and PDGFR-β resulted in potent inhibitory activities against all three RTKs (Table 1). Compound 1 was more potent than the clinically used agent sunitinib (EGFR ~61-fold, VEGFR-2 ~2-fold and PDGFR-β ~15-fold). Compared to erlotinib, 1 was 2-fold less potent in inhibiting EGFR but −15-fold and ~2-fold more potent in inhibiting VEGFR-2 and PDGFR-β, respectively. Thus, we discovered that, in addition to tubulin inhibition, 1 was a multitargeted RTK inhibitor as well.

Table 1.

EGFR, VEGFR-2 and PDGFR-β inhibitory activity of compound 1

Compound	EGFR kinase inhibition (nM)	Flk-1(VEGFR-2) kinase inhibition (nM)	PDGFR-β kinase inhibition (nM)
1	2.8 ± 1.1	8.4 ± 2.2	5.6 ± 0.4
Sunitinib	172.1 ± 19.4	18.9 ± 2.7	83.1 ± 10.1
Erlotinib	1.2 ± 0.2	124.7 ± 18.2	12.2 ± 1.9

Design of target compounds 2-11.

The reported antitubulin activity of 1, along with its RTK inhibitory activity of 1 (Table 1), provided the impetus for us to design quinazoline based multi-targeted inhibitors with both tubulin and RTK inhibitory properties in single agents. Compounds 2-4, Series I (Figure 5), with a 2-CH₃ substitution, were designed to optimize the substitution at the 4-position for activity with multiple RTKs and at the colchicine site of tubulin. The clinically used quinazoline RTK inhibitors (Figure 2) all had a 2-H substitution and hence compounds in Series II (5-7) were designed to mimic the 2-position substitution of the known RTK inhibitors and to determine its effect, if any, on microtubule polymerization activity. Compounds in Series III were expected to bind at the active sites of EGFR and VEGFR-2 by maintaining the hinge region binding interactions seen for erlotinib in EGFR (Figure 3) and axitinib in VEGFR-2 (Figure 4) while retaining an interaction at the colchicine site of tubulin. In addition to EGFR and VEGFR-2, the designed compounds 2-11 were anticipated to inhibit PDGFR-β analogous to 1.

Figure 5.

Series I (2-4), Series II (5-7), Series III (8-11)

The docked pose of 5 in the colchicine site of tubulin (Figure 6) is very similar to that of lead compound 1. The quinazoline scaffold in 1 and 5 undergoes hydrophobic interactions with Alaβ250, Leuβ248, Leuβ252, Leuβ255, Ileβ318, Alaβ354, and Alaβ316. The N⁴-CH₃ in 1 and 5 are oriented towards and interact with Lysβ254. The N⁴-phenyl group hydrophobically interacts with Lysβ352, Alaαl80 and Valα181. The docked score of 1 in the colchicine site of tubulin was −8.68 kcal/mol and that of 5 was −8.43 kcal/mol. Compounds in Series I (2-4) had docked scores of −8.17 to −9.83 kcal/mol and compounds in Series II (5-7) −8.06 to −8.67 kcal/mol, suggesting that these compounds should have similar or better activity than 1 as MTAs binding to the colchicine site.

Figure 6.

Comparison of the docked poses of 1 (cyan) and 5 (peach) in the colchicine site of tubulin (PDB ID: 4O2B).⁴⁷

In EGFR, compound 5 (Figure 7) also undergoes a hinge region binding interaction with Met769 and a water mediated H-bond with Thr766, showing similar interactions at the binding site of EGFR as observed with 1 (Figure 3). Compounds in Series I docked in a manner similar to that of 1 and had docked scores of −7.38 to −8.90 kcal/mol. Compounds in Series II docked in a manner similar to that of 5 and had docked scores of −8.42 to −9.09 kcal/mol.

Figure 7.

Superimposition of the docked poses of 1 (magenta) and 5 (orange) in EGFR (PDB: 1M17).⁴⁴

Compound 5 (Figure 8), when docked in the X-ray crystal structure of VEGFR-2,⁴⁶ showed two low energy binding modes, which corroborate similar findings reported for a series of furo[2,3-d]pyrimidine analogs.³⁴ In Figure 8A, 5 binds to VEGFR-2 in a binding mode similar to that seen for compound 1 (Figure 4) and has a docked score of −10.66 kcal/mol. In addition, due to the absence of bulk at the 2-position, 5 also binds to VEGFR-2 in an alternate binding mode where the N1 of the quinazoline scaffold in 5 undergoes H-bonding with Cys919 at the hinge region of VEGFR-2 and has a similar docking score of −9.32 kcal/mol. Compounds in Series I docked in VEGFR-2 in a manner similar to that of 1 and had docked scores of −10.71 to −12.15 kcal/mol. Compounds in Series III docked in a manner similar to 5 and had docked scores of −10.66 to −12.00 kcal/mol.

Figure 8.

Superimposition of the docked poses of 1 (orange) and 5 (magenta) in VEGFR-2 in two different binding modes (PDB: 4AG8).⁴⁶

The 2-Cl substituted analog of 1 was reported to be an equipotent inhibitor of cellular microtubule polymerization (T47D EC₅₀ = 2 ± 0.1 nM) as compared with 1² and hence, by varying the bulk and electronics at the 2-position, compounds 8-11 (Series III, Figure 5) were designed to assess the inhibitory effects on multiple RTKs. These compounds docked at the colchicine site of tubulin (docked scores, −8.06 to −9.40 kcal/mol), EGFR (docked scores, −7.59 to −8.46 kcal/mol) and VEGFR-2 (docked scores, −10.72 to −12.42 kcal/mol) in a mode similar to that shown for lead compound 1 and were predicted be active at the colchicine site and with EGFR and VEGFR-2.

Materials and Methods

Compounds 1-11 were synthesized by nucleophilic displacement of 4-chloroquinazolines (12a, 12b and 12c) by the appropriate anilines in isopropyl alcohol/acetonitrile and a drop of conc. HCl at room temperature for 12 h (Scheme 2).

Scheme 2.

Synthesis of target compounds 1-11 (Series I, II and III)

Yields of compounds in Series I were 58% (1, reported yield 71%),² 65% (2, reported yield 79%),⁴⁸ 70% (3, reported yield 41%),² 68% (4). Yields of compounds in Series II were 76% (5, reported yield 79%),¹ 70% (6, reported yield 69%),⁴⁹ and 72% (7). Yields of compounds in Series III were 9% (8, reported yield 87%),¹ 25% (9, reported yield 87%).⁴⁸ The synthesis of 10 was unsuccessful under the reaction conditions described above, but instead dimer 18 was formed (¹H NMR indicated four methyl singlets, two for SCH₃ at 2.48 and 2.49 ppm and two for NCH₃ at 3.29 and 3.57 ppm). Alternatively, a S_NAr reaction in acetonitrile at rt gave 10 in 22% yield. Compound 11 was synthesized over two steps. The first step involved the nucleophilic displacement of quinazoline 12c with 6-aminonaphthalen-1-ol (19) to yield 6-((2-chloroquinazolin-4-yl)amino)naphthalen-1-ol (20). N- and O-methylation of 20 using methyl iodide and sodium hydride yielded the desired compound 11 in 47% yield.

To synthesize N-methylated aniline 16 (Scheme 3), commercially available aniline 21 was formylated using paraformaldehyde in the presence of sodium methoxide in methanol. Subsequent reduction of the N-formyl group using NaBH₄ yielded aniline 16 (39%).⁵⁰ To synthesize the aniline 17 from 6-aminonaphthalen-1-ol (19), both N- and O-methylation, were required, and a different strategy was necessary. 6-Aminonaphthalen-1-ol (19) was first Boc protected to give the N-Boc protected intermediate 22, which was subsequently dimethylated (both N- and O-) with methyl iodide in the presence of sodium hydride⁵¹ and deprotected using trifluoroacetic acid to afford 5-methoxy-N-methylnaphthyl-2-amine (17) (55% over 3 steps).⁵²

Scheme 3.

Synthesis of anilines 16 and 17

Evidence of conformational restriction of 1 by ¹H NMR.

Comparison of the ¹H NMR spectra of compound 1 with the desmethyl analog 1a² in DMSO-d₆ suggested that the N-Me group imparts conformational restriction along the C4-N bond. In 1, the 5-H proton appears at δ 6.88, however in compound 1a, the 5-H proton is more deshielded and appears at δ 8.06 (Figure 9). The shielding effect of the 5-H can be attributed to a diamagnetic anisotropic effect in 1 due to the proximity of the phenyl ring on the 5-H proton in the more favorable conformation. The same effect, however, does not occur in the less favorable conformation since the phenyl ring lies anti to the 5-H and leads to steric clashes between the N-Me and 5-H proton groups. The diamagnetic anisotropy is also not observed for the conformationally flexible analog 1a, and the 5-H proton appears more deshielded.

Figure 9.

Conformational restriction of C4-N observed for 1 due to presence of the N-Me group.

Results and Discussion

Effect on cellular microtubules and cancer cell proliferation.

Compounds 2–11 were evaluated for their ability to depolymerize microtubules in A-10 cells with a phenotypic assay to evaluate microtubule loss. An EC₅₀, the concentration that causes 50% cellular microtubule loss, was determined. The antiproliferative effects were evaluated in MDA-MB-435 melanoma cells using the sulforhodamine B (SRB) assay, and the IC₅₀, the concentration that causes 50% inhibition of proliferation, was determined. The activities of the compounds were compared with 1, paclitaxel and CA-4, a small molecule that binds the colchicine site of β-tubulin at its interface with α-tubulin (Table 2).

Table 2.

IC₅₀ values for inhibition of proliferation of MDA-MB-435 cells and EC₅₀ values for microtubule depolymerization in A-10 cells.

Compound	IC50 ± SD in MDA- 435 Cells (nM)	EC50 (nM) for Microtubule Depolymerization in A·10 Cells	EC50/IC50 Ratio	CLogP
1	1.7 ± 0.1	2.1	1.2	4.2
2	1.1 ± 0.2	3	2.7	4.8
3	1.2 ± 0.2	2	1.7	4.8
4	2.0 ± 0.4	2.4	1.2	5.3
5	7.1 ± 0.8	8.2	1.2	3.7
6	12.4 ± 1.7	23.4	1.9	4.3
7	7.8 ± 0.8	12.3	1.6	4.3
8	0.6 ± 0.1	2	3.3	4.4
9	0.7 ± 0.2	2	2.9	5.1
10	2.7 ± 0.5	2.7	1	5.1
11	1.4 ± 0.1	1.9	1.4	5.6
CA-4	4.4 ± 0.46	9.8	2.2
Paclitaxel	4.5 ± 0.52	-	-

CLogP values were calculated using ChemDraw Professional 2018

In the antiproliferative assay (Table 2), the potency of the compounds followed the trend: 2-Cl (8-11) > 2-CH₃ (1-4) > 2-H (5-7) substitution, suggesting that small hydrophobic groups at the 2-position could contribute to improved cellular permeability as indicated by the CLogP values (Table 2). In addition, small hydrophobic groups at the 2-position were tolerated for binding at the colchicine site. In the microtubule depolymerization assay, compounds from Series I (1-4) and Series III (8-11) were equipotent and were 4- to 12-fold more potent than Series II (5-7) compounds with 2-H substitution. For compounds 1-11, the EC₅₀/IC₅₀ ratios (Table 2) are close to 1.0, suggesting a cytotoxic mechanism of action that is primarily microtubule dependent.

Clinically relevant drug resistance mechanisms associated with most clinically approved microtubule targeting agents, include expression of either the β-III isotype of tubulin or Pgp, an ATP-dependent efflux pump.⁵³ Compounds 1-11 were tested for their efficacy in parental cells and β-III and Pgp expressing cell lines (Table 3). In β-III overexpressing HeLa cells, compounds 1-11 were able to all overcome β-III mediated resistance compared with paclitaxel and were comparable to CA-4 in their activity. The resistance ratio (Rr) values (Table 3) were 0.6-1.2 versus 1.0 for CA-4 and 8.6 for paclitaxel.

Table 3.

Compound activity in parental, β-III and Pgp expressing cell lines

Compound	IC50 ± SD in HeLa (nM)	IC50 ± SD in WT βIII (nM)	Rr Value (WT/HeLa)	IC50 ± SD in SK-OV-3 Cells (nM)	IC50 ± SD in SK-OV- 3-MDR-1- 6/6 Cells (nM)	Rr Value (MDR- 6/6/SK- OV-3)
1	2.0 ± 0.2	1.5 ± 0.2	0.8	2.2 ± 0.1	2.6 ± 0.1	1.2
2	1.2 ± 0.5	1.2 ± 0.4	1.0	1.5 ± 0.3	2.2 ± 0.6	1.5
3	1.3 ± 0.6	1.1 ± 0.4	0.8	1.6 ± 0.2	2.1 ± 1.1	1.3
4	2.5 ± 0.3	1.7 ± 0.2	0.7	2.9 ± 0.3	3.5 ± 0.4	1.2
5	8.6 ± 0.3	7.4 ± 0.7	0.9	10.4 ± 1.1	14.3 ± 0.5	1.4
6	16.8 ± 2.2	13.1 ± 0.4	0.8	18.1 ± 1.8	23.4 ± 4.5	1.3
7	8.5 ± 0.8	7.3 ± 0.4	0.6	13.6 ± 0.9	15.6 ± 1.4	1.1
8	0.6 ± 0.2	0.6 ± 0.1	1.0	0.8 ± 0.1	1.1 ± 0.3	1.4
9	0.6 ± 0.2	0.7 ± 0.2	1.2	0.8 ± 0.2	1.0 ± 0.5	1.3
10	5.0 ± 0.4	3.2 ± 0.2	0.6	3.3 ± 0.1	3.5 ± 0.6	1.1
11	2.8 ± 0.5	1.6 ± 0.3	0.6	2.8 ± 0.4	2.6 ± 0.4	0.9
CA-4	3.3 ± 0.4	3.3 ± 0.3	1.0	5.5 ± 0.5	7.2 ± 1.1	1.3
Paclitaxel	2.8 ± 0.4	24.0 ± 0.2	8.6	5.0 ± 0.6	1200 ± 58	240

Paclitaxel is a known substrate for Pgp, and cells expressing Pgp are resistant to its effects, resulting in a 240-fold loss in activity (Table 3). Compounds 1-11 were evaluated for their ability to circumvent Pgp-mediated drug resistance in the SK-OV-3 isogenic cell line pair. As expected, the compounds showed activity similar to that of CA-4 and were essentially equipotent in the parental and Pgp-expressing cells (Rr values 0.9-1.5). In contrast, paclitaxel had a Rr value of 240, indicating its susceptibility to Pgp-mediated resistance.

Selected quinazoline derivatives, on the basis of their activities in Tables 2 and 3, were also tested for their activity as inhibitors of tubulin assembly and their ability to inhibit the binding of [³H] colchicine to tubulin. The compounds also had potencies similar to that of CA-4 for inhibition of colchicine binding in both the assays.

Evaluation of compounds 1-4, 6, 8-9 in EGFR, VEGFR-2 and PDGFR-β assays suggested that these compounds were potent inhibitors of multiple RTKs (Table 5). With respect to the structure-activity relationship, within Series I, the 4-methoxy-N-methylaniline substitution (1) was equipotent with the tetrahydroquinoline substitution (2) and 3-fold better than 3 and 4 with EGFR. For VEGFR-2, compounds 1-4 had similar inhibitory IC₅₀ values, suggesting the VEGFR-2 binding site is more accommodating towards introduction of bulk at the 4-position of the quinazoline scaffold. With PDGFR-β, 1 and 2 has similar activity, but, introduction of a thiomethyl group in 3 results in ~ 41-fold loss of potency and ~ 5-fold loss in potency for compound 4 with PDGFR-β. Comparing the tetrahydroquinoline compounds 2, 6 and 9 across Series I-III, the 2-H substituted compound 6 resulted in ~7-10-fold loss in potency for all the RTKs and also a similar loss of activity in the cytotoxicity assay, suggesting the presence of bulk at the 2- position is important for activity in RTKs. 2-CH₃ and 2-Cl substitution (2 and 9) did not affect EGFR activity but results in ~6-10-fold loss of potency for VEGFR-2 and PDGFR- β, respectively.

Table 5.

RTK inhibitory activities of quinazolines (1-4, 6, 8-9)

Compound	EGFR kinase inhibition (nM)	Flk-1 (VEGFR-2) kinase inhibition (nM)	PDGFR-β kinase inhibition (nM)	A431 Cytotoxicity (nM)	U251 cytotoxicity (nM)	CAM angiogenesis inhibition (μM)
1	2.8 ± 1.1	8.4 ± 2.2	5.6 ± 0.4	0.9 ± 0.2	2.8 ± 0.6	2.8 ± 1.1
2	5.0 ± 0.4	9.3 ± 3.9	7.8 ± 0.5	2.5 ± 0.2	3.1 ± 0.3	3.1 ± 1.3
3	8.7 ± 1.6	14.7 ± 2.8	231.5 ± 29.9	2.9 ± 0.5	4.9 ± 0.9	9.8 ± 0.7
4	10.6 ± 1.8	16.8 ± 4.1	29.4 ± 5.1	5.3 ± 0.9	4.2 ± 2.0	ND
6	48.9 ± 6.0	111.1 ± 19.2	54 ± 14.1	16.3 ± 2.0	27.8 ± 1.8	ND
8	3.6 ± 0.6	97.5 ± 10.1	76.6 ± 16.2	1.2 ± 0.4	32.5 ± 4.1	130.0 ± 22.2
9	5.4 ± 0.3	55.2 ± 11.1	80.6 ± 12.1	1.8 ± 0.1	14.8 ± 1.8	41.4 ± 6.1
Sunitinib	172.1 ± 19.4	18.9 ± 2.7	83.1 ± 10.1	-	-	-
Erlotinib	1.2 ± 0.2	124.7 ± 18.2	12.2 ± 1.9	-	-	29.1 ± 1.9

ND-Not Determined

In addition, comparing the 2-position quinazoline substitution across Series I, II and III, for EGFR 2-CH₃ and 2-Cl substitution were ~ 10-fold better than the 2-H substitution. For VEGFR-2 binding, 2-CH₃ was 6-fold better than the 2-Cl substitution and ~ 12-fold better than the 2-H substitution. In contrast, for PDGFR-β, the 2-H substitution was 1.5-fold better than the 2-Cl substitution, which in turn was 7-fold less potent than the 2-CH₃ substitution.

Moreover, compounds 1-3 were 3- to 10-fold more potent than erlotinib in the CAM assay for angiogenesis inhibition (Table 5). However, compounds in Series III were 1- to 4-fold less potent than erlotinib in inhibiting angiogenesis.

Conclusion

Substituted quinazolines have been explored as a scaffold for targeting RTKs as well as microtubules. Here we describe the design, synthesis and evaluation of 4-anilino quinazolines with multiple modes of action contributing to their antiproliferative activities in cancer cells. Compounds were synthesized by varying substitutions at the 2- and 4- positions of the quinazoline scaffold. For the MTA effects in general, the 2-H substituted compounds (5-7) were 7- to 10-fold less potent than the 2-CH₃ (1-4) and the 2-Cl (8-9) substituted compounds. This suggests that bulk at the 2-position of the quinazoline scaffold is important for binding to the colchicine site. In the RTK inhibitory assays, the 2-CH₃ and 2-Cl substituted compounds were more potent than the 2-H substituted compounds. For 2-substituted quinazolines (1-4), varying the aniline substitution at the 4-position was well tolerated. The 2-Cl substituted compound 8 was a potent inhibitor of EGFR but resulted in a 10-fold and 15-fold loss of potency with VEGFR-2 and PDGFR-β, respectively, when compared to the corresponding 2-CH₃ compound 1. Our studies establish that the quinazoline based analogs are indeed single agents with multitargeted activities, and that they inhibit both tubulin and RTKs and have potential as anticancer agents.

Experimental Section

Chemistry.

All evaporations were carried out under vacuum with a rotary evaporator. Analytical samples were dried in vacuo in a CHEM-DRY drying apparatus over P₂O₅ at 50 °C. Melting points were determined either using a MEL-TEMP II melting point apparatus with FLUKE 51 K/J electronic thermometer or using an MPA100 OptiMelt automated melting point system and are uncorrected. Thin-layer chromatography (TLC) was performed on Whatman® PE SIL G/UV254 flexible silica gel plates or Sorbetch silica gel TLC plates w/UV254, and the spots were visualized under 254 and 365 nm ultraviolet illumination. Proportions of solvents used for TLC are by volume. All analytical samples were homogeneous on TLC in at least two different solvent systems. Flash chromatography was carried out on the CombiFlash® Rf 200 (Teledyne ISCO) automated flash chromatography system with prepacked RediSep® Rf normal-phase flash columns (230 to 400 mesh) of various sizes. The weight of silica gel for column chromatography was in the range of 50-100 times the weight of the crude compounds being separated. Nuclear magnetic resonance spectra for proton (¹H NMR) were recorded on the Bruker Avance II 400 (400 MHz) or Bruker Avance II 500 (500 MHz) NMR systems and were analyzed using MestReC NMR data processing software. The chemical shift values (δ) are expressed in ppm (parts per million) relative to tetramethylsilane as an internal standard: s, singlet; d, doublet; dd, doublet of doublet; t, triplet; q, quartet; m, multiplet; br, broad singlet; td, triplet of doublet; dt, doublet of triplet; quin, quintet. Elemental analyses were performed by Atlantic Microlab, Inc., Norcross, GA. Element compositions are within ± 0.4% of the calculated values. Fractional moles of water or organic solvents frequently found in some analytical samples could not be prevented despite 24 to 48 hours of drying in vacuo and were confirmed where possible by their presence in the ¹H NMR spectra. All solvents and chemicals were purchased from Sigma-Aldrich or Fisher Scientific and were used as received.

N-Methyl-4-(methylthio)aniline (16)

4-(Methylthio)aniline (21) (1.13 g, 8.12 mmol) was added to a suspension of sodium methoxide (2.19 g, 40.6 mmol) in MeOH (10 mL). This mixture was poured into another suspension of paraformaldehyde (0.340 g, 11.4 mmol) in MeOH (10 mL). The reaction mixture was then stirred at room temperature. After 5 h, sodium borohydride (0.31 g, 8.1 mmol) was added, and the mixture was kept at reflux for 2 h. After the reaction was complete, the solvent was evaporated and 1 M KOH (10 mL) was added. The sample was then extracted 3 times using ether and dried over MgSO₄. The combined organic fractions were evaporated, a silica gel plug was made, and the product was purified by column chromatography using 10% EtOAc in hexanes to yield 16 as a brown liquid (0.49 g, 39%). TLC R_f = 0.76 (Hex: EtOAc; 1:1); ¹H NMR, DMSO-d₆ (400 MHz): 2.33 (s, 3 H, SCH₃), 2.64-2.66 (d, 3 H, J = 5.09, NHCH₃), 5.72-5.73 (m, 1 H, exch, NH), 6.49-6.52 (d, 2 H, J = 8.58, Ar), 7.11-7.13 (d, 2 H, J = 8.57, Ar). ¹H NMR agreed well with the literature⁵⁴ values.

5-Methoxy-N-methylnaphthalen-2-amine (17)

Compound 17 was prepared over 3 steps. 6-Aminonaphthalen-1-ol (19) (1.25 g, 7.85 mmol) was stirred with boc anhydride (5 mL) in ethyl acetate (15 mL) at room temperature. After 12 h, boc anhydride was removed by evaporation, and the residue was dissolved in ethyl acetate and purified using flash chromatography in CHCl₃/MeOH to afford 22 as a brown liquid (1.74 g, 85%). TLC R_f = 0.67 (CH₃OH: CHCl₃; 1:5); ¹H NMR, DMSO-d₆ (400 MHz): 1.51 (s, 9 H, (CH₃)₃), 6.71-6.73 (dd, 1 H, J₁ = 7.13, J₂ = 1.21, Ar), 7.17-7.25 (m, 2 H, Ar), 7.43-7.46 (dd, 1 H, J₁ = 1.97, J₂ = 9.14, Ar), 8.00-8.01 (d, 2 H, J = 8.87, Ar), 9.54 (s, 1 H, exch., OH), 10.00 (s, 1 H, exch., NH). Compound 22 (1.70 g, 6.79 mmol) was dissolved in DMF (30 mL) under argon. Sodium hydride (0.488 g, 20.4 mmol) was added, and the mixture was stirred at 0 °C for 15 min. Methyl iodide (2.12 g, 14.9 mmol) was added, and the reaction mixture was stirred at room temperature for 2 h. At the end of the reaction, solvent was evaporated, silica gel was added with methanol to make a plug, and the mixture was purified using flash chromatography with hexanes/ethyl acetate to afford 23 (1.36 g, 72%) as a yellow-colored liquid. TLC R_f = 0.8 (Hex: EtOAc; 1:1); ¹H NMR, DMSO-d₆ (400 MHz): 1.41 (s, 9 H, (CH₃)₃), 3.28 (s, 3 H, NCH₃), 3.96 (s, 3 H, OCH₃), 6.92-6.94 (m, 1 H, Ar), 7.42-7.43 (m, 2 H, Ar), 7.45-7.46 (d, 1 H, J = 2.19, Ar), 7.72-7.73 (d, 1 H, J = 2.12, Ar), 8.07-8.09 (d, 1 H, J = 9.03, Ar). Boc deprotection of 23 (1.3 g, 4.5 mmol) was carried out in trifluoroacetic acid (TFA, 15 ml) at room temperature for 2 h. At the end of the reaction, TFA was removed by evaporation, and the mixture was neutralized using an aq. solution of K₂CO₃. Extraction with ethyl acetate (20 mL × 3), followed by concentration and drying over Na₂SO₄ yielded 17 (0.810 g, 95%) as a green-colored solid. TLC R_f = 0.77 (Hex: EtOAc; 1:1); ¹H NMR, DMSO-d₆ (400 MHz): 2.75-2.76 (d, 3 H, NCH₃), 3.89 (s, 3 H, OCH₃), 5.98-5.99 (q, 1 H, NHCH₃, exch.), 6.57-6.59 (d, 1 H, J = 7.43, Ar), 6.61-6.62 (d, 1 H, J = 1.87, Ar), 6.88-6.90 (dd, 1 H, J₁ = 1.42, J₂ = 9.05, Ar), 7.15-7.23 (m, 2 H, Ar), 7.83-7.86 (d, 1 H, J = 9.05, Ar). The compound was carried to the next step without further purification.

4-Chloro-2-methylquinazoline (12a)

To a 250 mL round-bottomed flask were added 2-methylquinazolin-4(3H)-one (13a) (1.0 g, 6.2 mmol), trimethylamine (3 mL), and toluene (30 mL). The mixture was cooled to 0 °C, and phosphorus oxychloride (2 mL) was added. The reaction mixture was stirred at room temperature for 1 h. The reaction was then heated to 95 °C for 3 h. After monitoring the reaction by TLC, the reaction mixture was cooled to room temperature and diluted with 30 mL of ethyl acetate. The solution was then washed with 10 mL of ice cold water, 10 mL of saturated NaHCO₃, 10 mL of water, 5 mL of 1 N HCl, 10 mL of water, 10 mL of saturated NaHCO₃ and 10 mL of saturated NaCl. The organic layer was dried over Na₂SO₄, filtered and concentrated to obtain 12a as a yellow solid (0.600 g, 54%). TLC R_f = 0.8 (MeOH:CHCl₃; 1:5); mp, 83.9-85.3 °C (lit.⁵⁵ 85-86 °C); ¹H NMR, DMSO-d₆ (400 MHz): 2.58 (s, 3 H, CH₃), 7.59-7.63 (t, 1 H, J₁ = 7.58, J₂ = 7.58, Ar), 7.78-7.80 (d, 1 H, J = 7.98, Ar), 7.90-7.95 (t, 1 H, J₁ = 7.67, J₂ = 7.67, Ar), 8.13-8.15 (d, 1 H, J = 7.65, Ar).

4-Chloroquinazoline (12b)

To a 250 mL round-bottomed flask were added quinazolin-4(3H)-one (13b) (2.00 g, 13.7 mmol), trimethylamine (3 mL), and toluene (30 mL). The mixture was cooled to 0 °C, and phosphorus oxychloride (2 mL) was added. The reaction mixture was stirred at room temperature for 1 h. The reaction was then heated to 95 °C for 3 h. After monitoring the reaction by TLC, the reaction mixture was cooled to room temperature and diluted with 30 mL of ethyl acetate. The solution was then washed with 10 mL of ice cold water, 10 mL of saturated NaHCO₃, 10 mL of water, 5 mL of 1 N HCl, 10 mL of water, 10 mL of saturated NaHCO₃ and 10 mL of saturated NaCl. The organic layer was dried over Na₂SO₄, filtered and concentrated to obtain 12b as a yellow solid (1.79 g, 80%). TLC R_f = 0.85 (MeOH:CHCl₃; 1:5); mp, 144.2-146.1 °C (lit.⁵⁶ 145-147.5 °C); ¹H NMR, DMSO-d₆ (400 MHz): 7.59-7.63 (dt, 1 H, J₁ = 1.09, J₂ = 7.23, J₃ = 8.01, Ar), 7.76-7.78 (d, 1 H, J = 8.24, Ar), 7.89-7.92 (dt, 1 H, J₁ = 1.47, J₂ = 7.24, J₃ = 8.29, Ar), 8.15-8.17 (d, 1 H, J = 7.03, Ar), 8.58 (s, 1 H, Ar).

General method for the synthesis of compounds 1-10

To a 100 mL round-bottomed flask substituted-4-chloroquinazolines (1 equivalent), appropriate anilines (1.2 equivalents) and isopropanol (10 mL) were added and stirred at room temperature for 12 h. The compounds precipitated as HCl salts, which were filtered and dried in vacuo over P₂O₅ to afford the desired compounds.

N-(4-Methoxyphenyl)-N,2-dimethylquinazolin-4-amine (1)

Using the general method described above, 12a (0.150 g, 0.840 mmol) and 4-methoxy-N-methylaniline (14) (0.138 g, 1.01 mmol) were reacted for 12 h. At the end of the reaction, solvent was evaporated, silica gel and methanol were added, and the methanol evaporated in vacuo to give a dried plug. The reaction mixture was purified by flash chromatography using chloroform/methanol to afford 1 as a brown solid (0.136 g, 58%). TLC R_f = 0.8 (MeOH:CHCl₃; 1:5); mp, 99.1-101.2 °C; ¹H NMR, DMSO-d₆ (400 MHz): 2.66 (s, 3 H, 2-CH₃), 3.60 (s, 3 H, NCH₃), 3.81 (s, 3 H, OCH₃), 6.88-6.90 (d, 1 H, J = 8.35, Ar), 7.04-7.06 (d, 2 H, J = 8.89, Ar), 7.14-7.18 (t, 1 H, J₁ = 7.49, J₂ = 7.49, Ar), 7.30-7.32 (d, 2 H, J = 8.78, Ar), 7.68-7.75 (m, 2 H, Ar). Anal. Calcd. for C₁₇H₁₇N₃O .0.52 HCl: C, 68.46; H, 5.92; N, 14.08. Found C, 68.50; H, 6.24; N, 13.80; Cl, 4.54.

4-(6-Methoxy-3,4-dihydroquinolin-1(2H)-yl)-2-methylquinazoline (2)

Using the general method described above, 12a (0.150 g, 0.840 mmol) and 6-methoxy-1,2,3,4-tetrahydroquinoline (15) (0.164 g, 1.01 mmol) were reacted to afford 2 as a yellow precipitate (0.188 g, 65%). TLC R_f = 0.85 (MeOH:CHCl₃; 1:5); mp, 255.8-257.9 °C; ¹H NMR, DMSO-d₆ (400 MHz): 2.1 (quin, 2 H, CH₂), 2.75 (s, 3 H, 2-CH₃), 2.87-2.90 (t, 2 H, CH₂), 3.79 (s, 3 H, OCH₃), 4.20 (t, 2 H, NCH₂), 6.68-6.71 (dd, 1 H, J₁ = 2.88, J₂ = 8.85, Ar), 7.02-7.03 (d, 1 H, J = 2.81, Ar), 7.05-7.08 (d, 1 H, J = 8.87, Ar), 7.26-7.28 (d, 1 H, J = 8.53, Ar), 7.36-7.40 (dt, 1 H, J₁ = 1.38 , J₂ = 6.83, J₃ = 8.28, Ar), 7.88-7.96 (m, 2 H, Ar). Anal. Calcd. for C₁₉H₁₉N₃O. 1.07HCl: C, 66.26; H, 5.87; N, 12.20. Found C, 66.28; H, 5.75; N, 12.11; Cl, 10.49.

N,2-Dimethyl-N-(4-(methylthio)phenyl)quinazolin-4-amine (3)

Using the general method described above, 12a (0.135 g, 0.756 mmol) and N-methyl-4-(methylthio)aniline (16) (0.135 g, 0.907 mmol) were reacted to afford 3 as a white precipitate (0.175 g, 70%). TLC R_f = 0.9 (MeOH:CHCl₃; 1:5); mp, 235.2-236.4 °C; ¹H NMR, DMSO-d₆ (400 MHz): 2.55 (s, 3 H, 2-CH₃), 2.77 (s, 3 H, SCH₃), 3.74 (s, 3 H, NCH₃), 6.87-6.89 (d, 1 H, J = 8.39, Ar), 7.30-7.34 (dt, 1 H, J₁ = 1.33, J₂ = 7.0, J₃ = 8.48, Ar), 7.43-7.50 (m, 4 H, Ar), 7.85-7.89 (dt, 1 H, J₁ = 1.14, J₂ = 7.05, J₃ = 8.34, Ar), 7.93-7.95 (d, 1 H, J = 8.32, Ar). Anal. Calcd. for C₁₇H₁₇N₃S. 1.34 HCl: C, 59.28; H, 5.37; N, 12.20; S, 9.31. Found C, 59.37; H, 5.62; N, 11.93; S, 9.03; Cl, 10.55.

N-(5-Methoxynaphthalen-2-yl)-N,2-dimethylquinazolin-4-amine (4)

Using the general method described above, 12a (0.135 g, 0.840 mmol) and 5-methoxy-N-methylnaphthalen-2-amine (17) (0.156 g, 0.831 mmol) were reacted for 12 h. At the end of the reaction, solvent was evaporated, and silica gel and methanol were added, and the methanol evaporated in vacuo to give a dried plug. The reaction mixture was purified using flash chromatography using chloroform/methanol to afford 4 as a yellow solid (0.170 g, 68%). TLC R_f = 0.9 (MeOH:CHCl₃; 1:10); mp, 203.0-204.4 °C; ¹H NMR, DMSO-d₆ (400 MHz): 2.70 (s, 3 H, 2-CH₃), 3.72 (s, 3 H, NCH₃), 3.99 (s, 3 H, OCH₃), 6.91-6.93 (d, 1 H, J = 8.26, Ar), 7.00-7.02 (d, 1 H, J = 7.46, Ar), 7.05-7.09 (t, 1 H, J₁ = 7.70, J₂ = 7.70, Ar), 7.40-7.49 (m, 3 H, Ar), 7.66-7.69 (t, 1 H, J₁ = 7.59, J₂ = 7.59, Ar), 7.74-7.76 (d, 1 H, J = 8.33, Ar), 7.84 (s, 1 H, Ar), 8.21-8.23 (d, 1 H, J = 8.99, Ar). Anal. Calcd. for C₂₁H₁₉N₃O. 0.41 HCl: C, 73.22; H, 5.68; N, 12.20. Found C, 73.19; H, 5.82; N, 12.35; Cl, 2.22.

N-(4-Methoxyphenyl)-N-methylquinazolin-4-amine (5)

Using the general method described above, 12b (0.150 g, 0.911 mmol) and N-methyl-4-methoxyaniline (14) (0.150 g, 1.09 mmol) were reacted to afford 5 as a pale-yellow precipitate (0.210 g, 76%). TLC R_f = 0.8 (MeOH:CHCl₃; 1:5); mp, 244.3-245.3 °C; ¹H NMR, DMSO-d₆ (400 MHz): 3.75 (s, 3 H, NCH₃), 3.84 (s, 3 H, OCH₃), 6.84-6.86 (d, 1 H, J = 8.43, Ar), 7.12-7.16 (d, 2 H, J = 8.97, Ar), 7.32-7.36 (dt, 1 H, J₁ = 1.12, J₂ = 6.93, J₃ = 8.45, Ar), 7.46-7.50 (d, 2 H, J = 8.96, Ar), 7.87-7.91 (dt, 1 H, J₁ = 1.05, J₂ = 6.94, J₃ = 8.25, Ar), 7.95-7.97 (d, 1 H, J = 8.15, Ar), 9.06 (s, 1 H, Ar). Anal. Calcd. for C₁₆H₁₅N₃O. 1.10 HCl: C, 62.89; H, 5.31; N, 13.75. Found C, 62.89; H, 5.33; N, 13.75; Cl, 11.43.

4-(6-Methoxy-3,4-dihydroquinolin-1(2H)-yl)quinazoline (6)

Using the general method described above, 12b (0.150 g, 0.911 mmol) and 6-methoxy-1,2,3,4-tetrahydroquinoline (15) (0.179 g, 1.09 mmol) were reacted to afford 6 as a yellow precipitate (0.211 g, 70%). TLC R_f = 0.8 (MeOH:CHCl₃; 1:5); mp, 234.2-235.3 °C; ¹H NMR, DMSO-d₆ (400 MHz): 2.05-2.13 (quint, 2 H, CH₂), 2.87-2.91 (t, 2 H, CH₂), 3.79 (s, 3 H, OCH₃), 4.20-4.21 (t, 2 H, NCH₂), 6.67-6.70 (dd, 1 H, J₁ = 2.91, J₂ = 8.85, Ar), 7.02-7.03 (d, 1 H, J = 2.82, Ar), 7.08-7.10 (d, 1 H J = 8.85, Ar), 7.33-7.35 (d, 1 H, J = 8.50, Ar), 7.42-7.46 (dt, 1 H, J₁ = 1.55, J₂ = 6.67, J₃ = 8.32, Ar), 7.91-7.98 (m, 2 H, Ar), 9.04 (s, 1 H, Ar). Anal. Calcd. for C₁₈H₁₇N₃O. 1.22 HCl: C, 64.38; H, 5.47; N, 12.51. Found C, 64.39; H, 5.47; N, 12.45; Cl, 10.41.

N-Methyl-N-(4-(methylthio)phenyl)quinazolin-4-amine (7)

Using the general method described above, 12b (0.150 g, 0.911 mmol) and N-methyl-4-(methylthio)aniline (16) (0.168 g, 1.09 mmol) were reacted to afford 7 as a pale yellow precipitate (0.209 g, 72%). TLC R_f = 0.8 (MeOH:CHCl₃; 1:5); mp, 233.4-235.9 °C; ¹H NMR, DMSO-d₆ (400 MHz): 2.54 (s, 3 H, SCH₃), 3.75 (s, 3 H, NCH₃), 6.94-6.96 (d, 1 H, J = 8.74, Ar), 7.35-7.39 (t, 1 H, J₁ = 7.69, J₂ = 7.69, Ar), 7.43-7.45 (d, 2 H, J = 8.70, Ar), 7.48-7.50 (d, 2 H, J = 8.68, Ar), 7.88-7.92 (t, 1 H, J₁ = 7.55, J₂ = 7.55, Ar), 7.95-7.97 (d, 1 H, J = 7.68, Ar), 9.08 (s, 1 H, Ar). Anal. Calcd. for C₁₆H₁₅N₃S. 1.13 HCl: C, 59.59; H, 5.04; N, 13.03; S, 9.94. Found C, 59.56; H, 5.13; N, 12.97; S, 10.22; Cl, 10.22.

2-Chloro-N-(4-methoxyphenyl)-N-methylquinazolin-4-amine (8)

Using the general method described above, 12c (0.500 g, 2.51 mmol) and 4-methoxy-N-methylaniline (14) (0.413 g, 3.01 mmol) were reacted for 12 h. At the end of reaction, solvent was evaporated, silica gel and methanol were added, and the methanol evaporated in vacuo to give a dried plug. The reaction mixture was purified using flash chromatography using chloroform/methanol to afford 8 as buff-colored solid (0.070 g, 9%). TLC R_f = 0.95 (MeOH:CHCl₃; 1:5); mp, 146.3-148.7 °C; ¹H NMR, DMSO-d₆ (400 MHz): 3.52 (s, 3 H, NCH₃), 3.81 (s, 3 H, OCH₃), 6.86-6.88 (d, 1 H, Ar), 7.03-7.06 (d, 2 H, J = 8.96, Ar), 7.11-7.16 (m, 1 H, Ar), 7.32-7.35 (d, 2 H, J = 8.95, Ar), 7.66-7.67 (m, 2 H, Ar). Anal. Calcd. for C₁₆H₁₄C1N₃O: C, 64.11; H, 4.71; N, 14.02; Cl, 11.83. Found C, 64.08; H, 4.77; N, 13.93; Cl, 11.97.

2-Chloro-4-(6-methoxy-3,4-dihydroquinolin-1(2H)-yl)quinazoline (9)

Using the general method described above, 12c (0.300 g, 1.51 mmol) and 6-methoxy-1,2,3,4-tetrahydroquinoline (15) (0.295 g, 1.81 mmol) were reacted for 12 h. At the end of the reaction, solvent was evaporated, silica gel and methanol were added, and the methanol evaporated in vacuo to give a dried plug. The reaction mixture was purified by flash chromatography using chloroform/methanol to afford 9 as a yellow solid (0.121 g, 25%). TLC R_f = 0.97 (MeOH:CHCl₃; 1:5); mp, 135.9-137.0 °C (lit.⁴⁸ 136-138 °C); ¹H NMR, DMSO-d₆ (400 MHz): 2.13-2.18 (m, 2 H, CH₂), 2.88-2.91 (t, 2 H, CH₂), 3.85 (s, 3 H, OCH₃), 4.11-4.16 (m, 2 H, CH₂), 6.59-6.62 (dd, 1 H, J₁ = 2.89, J₂ = 8.81, 7´-Ar), 6.75-6.77 (d, 1 H, J = 8.85, 8´-Ar), 6.857-6.863 (d, 1 H, J = 2.86, 5´-Ar), 7.16-7.19 (m, 1 H, J₁ = 1.26, J₂ = 6.96, J₃ = 8.43, Ar), 7.31-7.33 (dd, 1 H, J₁ = 0.90, J₂ = 8.53, Ar), 7.68-7.71 (m, 1 H, J₁ = 1.36, J₂ = 6.97, J₃ = 8.39, Ar), 8.05-8.07 (d, 1 H, Ar). Anal. Calcd. for C₁₈H₁₆ClN₃O . 0.24 (CH₃)₂CHOH: C, 66.08; H, 5.31; N, 12.34; Cl, 10.41. Found C, 66.30; H, 5.19; N, 12.35; Cl, 10.17.

2-Chloro-N-methyl-N-(4-(methylthio)phenyl)quinazolin-4-amine (10)

Using the general procedure described above, 12c (0.300 g, 1.51 mmol) and 16 (0.23 g, 1.5 mmol) were reacted in acetonitrile for 12 h. At the end of reaction, solvent was evaporated, silica gel and methanol were added, and the methanol evaporated in vacuo to give a dried plug. The reaction mixture was purified by flash chromatography using chloroform/methanol to afford 10 (0.1 g, 22%) as a dark yellow solid. TLC R_f = 0.76 (MeOH:CHCl₃; 1:5); mp, 123.6 °C, ¹H NMR, CDCl₃ (400 MHz): 2.55 (s, 3 H, SCH₃), 3.68 (s, 3 H, NCH₃), 6.97-7.02 (m, 1 H, Ar), 7.10 (s, 1 H, Ar), 7.17 (d, 2 H, J = 8.6 Hz, Ar), 7.32 (d, 2 H, J = 8.6 Hz, Ar), 7.64 (s, 1 H, Ar), 7.95 (s, 1 H, Ar). Anal. Calcd. for C₁₆H₁₄ClN₃S: C, 60.85; H, 4.47; N, 13.31; S, 10.15; Cl, 11.22. Found: C, 60.97; H, 4.55; N, 13.19; S, 10.12; Cl, 11.07.

N²,N⁴-Dimethyl-N²,N⁴-bis(4-(methylthio)phenyl)quinazoline-2,4-diamine (18)

Compound 12c (0.200 g, 1.00 mmol) and N-methyl-4-(methylthio)aniline (16) (0.185 g, 1.21 mmol) were stirred at room temperature for 12 h. At the end of the reaction, solvent was evaporated, silica gel and methanol were added, and the methanol evaporated in vacuo to give a dried plug. The reaction mixture was purified using flash chromatography using chloroform/methanol which afforded dimer 18 as a yellow semisolid (0.195 g, 45%). TLC R_f = 0.87 (Hex:EtOAc; 1:1); ¹H NMR, DMSO-d₆ (400 MHz): 2.48 (s, 3 H, SCH₃), 2.49 (s, 3 H, SCH₃), 3.29 (s, 3 H, NCH₃), 3.57 (s, 3 H, NCH₃), 6.76-6.81 (m, 1 H, Ar), 6.88-6.90 (dd, 1 H, J = 0.89, J = 8.43, Ar), 7.14-7.16 (d, 2 H, J = 8.68, Ar) 7.25-7.29 (m, 4 H, Ar), 7.37-7.45 (m, 4 H, Ar).

6-((2-Chloroquinazolin-4-yl)amino)naphthalen-1-ol (20)

Using the general method described above, 12c (0.200 g, 1.00 mmol) and 6-aminonaphthalen-1-ol 19 (0.144 g, 0.904 mmol) were reacted for 12 h. At the end of the reaction, the solvent was evaporated, silica gel and methanol were added, and the methanol evaporated in vacuo to give a dried plug. The reaction mixture was purified using flash chromatography using chloroform/methanol to the afford 20 as white solid (0.070 g, 22%). TLC R_f = 0.67 (MeOH:CHCl₃; 1:5); ¹H NMR, DMSO-d₆ (400 MHz): 6.83-6.87 (t, 1 H, Ar), 7.32-7.33 (d, 2 H, J = 4.55, Ar), 7.67-7.71 (dt, 1 H, J₁ = 1.13, J₂ = 6.96, J₃ = 8.09, Ar), 7.74-7.76 (dd, 1 H, J₁ = 0.79, J₂ = 8.31, Ar), 7.81-7.84 (dd, 1 H, J₁ = 2.11, J₂ = 9.07, Ar), 7.90-7.94 (dt, 1 H, J₁ = 1.13, J₂ = 7.08, J₃ = 8.35, Ar), 8.16-8.18 (d, 1 H, J = 9.02, Ar), 8.24-8.25 (d, 1 H, J = 2.03, Ar), 8.63-8.65 (dd, 1 H, J₁ = 1.35, J₂ = 8.66, Ar), 10.16 (s, 1 H, exch., NH), 10.40 (s, br, 1 H, exch., OH). The compound was used without further characterization.

2-Chloro-N-(5-methoxynaphthalen-2-yl)-N-methylquinazolin-4-amine (11)

Compound 20 (0.070 g, 0.22 mmol) was dissolved in anhydrous DMF (10 mL) in a three necked round-bottomed flask, and the reaction mixture was stirred at 0 °C under argon. Sodium hydride (0.016 g, 0.65 mmol) was added to the above mixture, which was stirred at 0 °C for an additional 15 min. Methyl iodide (0.068 g, 0.40 mmol) was added while maintaining the reaction at room temperature for 4 h. Methanol was added at the end of the reaction, solvent was evaporated, and the reaction mixture was then extracted with water and ethyl acetate (25 mL × 3) and dried over Na₂SO₄. Silica gel was added to the combined fractions, the solvent was removed by evaporation to yield a dry plug, and a column was run using hexanes/ethyl acetate. The fractions containing the product (TLC) were pooled and evaporated to obtain the desired compound 11 as an off-white solid (0.036 g, 47%). TLC R_f = 0.82 (MeOH:CHCl₃; 1:5); mp, 171.9-173.8 °C; ¹H NMR, DMSO-d₆ (400 MHz): 3.64 (s, 3 H, NCH₃), 3.99 (s, 3 H, OCH₃), 6.86-6.88 (d, 1 H, J = 8.54, Ar), 7.00-7.06 (m, 2 H, Ar), 7.40-7.49 (m, 3 H, Ar), 7.63-7.71 (m, 2 H, Ar), 7.88-7.89 (d, 1 H, J = 2.02, Ar), 8.21-8.23 (d, 1 H, J = 8.92, Ar). Anal. Calcd. for C₂₀H₁₆ClN₃O . 0.11 CH₃(CH₂)₄CH₃: C, 69.07; H, 4.93; N, 11.68; Cl, 9.86. Found C, 68.84; H, 5.06; N, 11.69; Cl, 9.64.

Molecular modeling and computational studies

Molecular modeling was performed for all analogs with the tubulin (PDB: 4O2B⁴⁷), VEGFR-2 (PDB: 4AG8⁴⁶) and EGFR (PDB: 1M17⁵⁷) crystal structures using the induced fit docking protocol of Maestro 11.9.⁴⁵ The ligands were prepared using the Ligprep application of Maestro. The docking protocol was validated by re-docking the co-crystallized ligands colchicine, axitinib and erlotinib in tubulin, VEGFR-2, and EGFR, respectively, with RMSD of 0.17 Å, 0.50 Å, and 1.19 Å, respectively. The centroids around the ligands were defined as the ligand binding site. The OPLS3e force field was used, and amino acid residues within 5 Å from the docked poses were allowed to be optimized using prime refinement.

Biological studies

Effects of compounds on cellular microtubules.

A-10 cells were used to evaluate the effects of the compounds on cellular microtubules using indirect immunofluorescence techniques. Cells were treated overnight with the compounds, fixed in cold MeOH, and microtubules were visualized with a β-tubulin antibody (Sigma–Aldrich, St. Louis, MO). EC₅₀ values were calculated as previously described⁵⁸ and represent an average of at least three independent experiments.

Sulforhodamine B (SRB) assay.

The SRB assay was used, as previously described, to evaluate the antiproliferative and cytotoxic effects of the compounds against cancer cells.⁵⁹ MDA-MB-435, SK-OV-3 and HeLa cells were purchased from the American Type Culture Collection (Manassas, VA). Details about the generation of the SK-OV-3-MDR1-6/6 and HeLa WT-βIII cells were previously described.⁵⁹ The IC₅₀ values represent an average of at least 3 independent experiments using triplicate points in each experiment.

Antibodies.

The PY-HRP antibody was obtained from BD Transduction Laboratories (Franklin Lakes, NJ). Antibodies against EGFR, VEGFR-2, and PDGFR-β were purchased from Cell Signaling Technology (Danvers, MA).

Phosphotyrosine ELISA.

A high-throughput phosphotyrosine ELISA was developed for evaluating the effect of compounds on RTKs.⁶⁰ Cells used for these experiments have been shown to overexpress particular RTKs; specifically A431 for EGFR, U251 for VEGFR-2, and SH-SY5Y for PDGFR-β. Briefly, cells at 60–75% confluence are placed in serum-free medium for 18 h to reduce background phosphorylation. Cells were always >98% viable by trypan blue exclusion. Cells were then pretreated for 60 min with 100 nM-1.4 μM concentrations of the compound of interest followed by the addition of 100 ng/mL of purified growth factor (EGF, VEGF, or PDGFR-β) for 10 min. The reaction was stopped, and cells were permeabilized by quickly removing the media and adding ice-cold Tris-buffered saline (TBS) containing 0.05% Triton X-100, protease inhibitor cocktail, and tyrosine phosphatase inhibitor cocktail (Sigma Chemical). The TBS solution was then removed, and cells fixed to the plate for 30 min at 60 °C, with a further incubation in 70% ethanol for an additional 30 min. Cells were then exposed to a blocking solution (TBS with 1% BSA) for 1 h, washed, and then a horseradish peroxidase (HRP)-conjugated phosphotyrosine (PY) antibody was added overnight. The antibody was removed, cells were washed again in TBS, exposed to an enhanced luminol ELISA substrate (Pierce Chemical EMD, Rockford, IL), and light emission was measured using a plate reader (BioTek, Winooski VT). The known RTK-specific inhibitors (semaxanib, sunitinib, erlotinib) were used as positive controls for kinase inhibition. Data were graphed as a percent of cells receiving growth factor alone, and IC₅₀ values were calculated from two to three separate biological replicates (n = 8 technical replicates per biological replicate) using sigmoidal dose-response relations in Prism 8.0 software (GraphPad).

Chorioallantoic membrane (CAM) assay.

The CAM assay is a standard assay for testing antiangiogenic agents. The CAM assay used in these studies was performed as described previously (Zhang, X.; Raghavan, S.; Ihnat, M.; Thorpe, J. E.; Disch, B. C.; Bastian, A.; Bailey-Downs, L. C.; Dybdal-Hargreaves, N. F.; Rohena, C. C.; Hamel, E.; Mooberry, S. L.; Gangjee, A. Bioorg Med Chem 2014, 22, 3753-72) Briefly, fertile leghorn chicken eggs (Ideal Poultry, Cameron, TX) were incubated for 10 days. The proangiogenic factors, human VEGF-165 and bFGF (100 ng each) were then added at saturation to a 6 mm microbial testing disk (BBL, Cockeysville, MD) and placed onto the CAM by breaking a small hole in the superior surface of the egg. Antiangiogenic compounds, at various doses, were added 8 h after the VEGF/bFGF at saturation to the same microbial testing disk, and the embryos were incubated for an additional 40 h. After 48 h, the CAMs were perfused with 2% paraformaldehyde containing 0.025% Triton X-100 for 20 sec, excised around the area of treatment, fixed again in 4% paraformaldehyde for 30 min, placed on Petri dishes, and a digitized image was taken using a dissecting microscope (Leica Z16-APO; Wetzlar, Germany) at 7.5X magnification. A grid was then added to the digital CAM images, and the average number of vessels within 5–7 grids counted as a measure of vascularity. Sunitinib and semaxanib were used as positive controls for antiangiogenic activity. Data were graphed as a percent of CAMs receiving bFGF/VEGF only and IC₅₀ values calculated from two to three separate experiments with 2-5 replicates per experiment using non-linear regression dose-response relation analysis.

Quantitative tubulin studies.

Bovine brain tubulin was purified as described previously.⁶¹ Briefly, 1.0 mg/mL of tubulin (10 μM) was incubated for 15 min with 0.8 M monosodium glutamate (pH of 2 M stock solution adjusted to 6.6 with HCl), varying compound concentrations and 4% (v/v) dimethyl sulfoxide as solvent. After a 15 min preincubation, 0.4 mM GTP was added. The reaction mixtures were transferred to cuvettes held at 0 °C in a cuvette holder equipped with an electronic temperature controller in a recording spectrophotometer. After baselines were established, the temperature was jumped over about 30 s to 30 °C, and changes in turbidity were monitored for 20 min. The compound concentration that caused a 50% reduction in increase in turbidity, interpolated from the values obtained with defined compound concentrations, was defined as the IC₅₀ value. The assay to measure inhibition of [³H]colchicine binding was described in detail previously.⁶² Briefly, 0.1 mg/mL (1.0 μM) tubulin was incubated, at 37 °C, with 5.0 μM [³H]colchicine and potential inhibitors at 1.0 or 5.0 μM, as indicated. Incubation was for 10 min, at which point 40-60% of the maximum colchicine that can be bound without inhibitor is reached. The [³H]colchicine was a product of Perkin-Elmer. CA-4 was a generous gift of Dr. G. R. Pettit, Arizona State University.

Scheme 1.

Synthesis of 4-chloro-2-methylquinazoline 12a and 4-chloroquinazoline 12b

Table 4.

Inhibition of [³H] colchicine binding and inhibition of tubulin assembly

Compound	Inhibition of tubulin assembly	Inhibition of colchicine binding
	IC50 (μM ± SD)	5 μM inhibitor	1 μM inhibitor
3	0.48 ± 0.07	99 ± 0.3	91 ± 0.5
4	0.64 ± 0.01	98 ± 0.2	88 ± 0.2
8	0.47 ± 0.02	99 ± 0.8	92 ± 0.2
CA-4	0.73 ± 0.04	98 ± 0.3	87 ± 0.8

Abbreviations:

RTKs

Receptor tyrosine kinases

VEGF

vascular endothelial growth factor

VEGFR-2

vascular endothelial growth factor receptor-2

PDGFR-β

platelet-derived growth factor receptor-β

EGFR

epidermal growth factor receptor

CA-1

combretastatin A-1

CA-4

combretastatin A-4

CAM

chorioallantoic membrane

Pgp

P-glycoprotein

MTAs

microtubule targeting agents