4-Aminopyrazolopyrimidine scaffold and its deformation in the design of tyrosine and serine/threonine kinase inhibitors in medicinal chemistry

Abstract

The 4-Aminopyrazolopyrimidine scaffold has been an interesting pharmacophore since the disclosure of the intimate connection between small-molecule inhibitors and the treatment of diseases. Modification of the 4-aminopyrazolopyrimidine scaffold according to different targets, especially tyrosine kinase and serine/threonine kinase, has resulted in a variety of small-molecule inhibitors. Kinase inhibitors with 4-aminopyrazolopyrimidine derivatives as scaffolds have been widely applied in the treatment of diseases. In this article, we summarized the reports on 4-aminopyrazolopyrimidine as well as its deformation and the application of its derivatives in designing small-molecule inhibitors and the treatment of diseases.

Deformation of the 4-aminopyrazolopyrimidine scaffold in designing small-molecule inhibitors.

1. Introduction

With the development of science and technology, obvious achievements have been made against many diseases. However, cancer, immune-cell-mediated diseases, etc. are still major problems in the medical profession. Researchers had confirmed that a lot of specific proteins were closely related to the occurrence of diseases, such as abnormal activation of the epidermal growth factor receptor (EGFR) in non-small cell lung cancer (NSCLC) cells. EGFR was closely associated with the proliferation, angiogenesis, tumor invasion, metastasis, and inhibition of apoptosis in NSCLC cells.¹ Hence, small molecule inhibitors targeting EGFR were developed to limit the phosphorylation of EGFR and consequently limit the phosphorylation of the downstream signaling pathway. Clinical trials showed that small-molecule inhibitors had the advantages of selectivity, less toxicity, and fewer side effects compared to traditional radiation therapy drugs. As the veil of diseases was slowly being lifted, more specific proteins were found to be inseparable from the occurrence of diseases, such as FGFR4 and HCC and BTK and immunological diseases. These proteins have become hot targets for the treatment of diseases. So far, many small-molecule inhibitors targeting the corresponding targets have been designed, and many of them have already been applied in clinical research.

Protein phosphorylation is the most basic, common, and important mechanism for regulating and controlling the activity and function of proteins. Protein phosphorylation occurs mainly in two amino acids, serine (including threonine) and tyrosine. The protein kinase binds to ATP, autophosphorylates, and induces signal transduction. Hence, ATP competitive inhibitors have become an important medium to disturb the binding of the kinase to ATP. The ATP-binding pocket is roughly divided into three regions: (1) hydrophobic region I; (2) the adenine region; (3) hydrophobic region II. We found that the 3-N and 4-NH₂ of 4-aminopyrazolopyrimidine form an intermolecular hydrogen bond with the amino acid residue of the adenine region. Because of the excellent binding mode of 4-aminopyrazolopyrimidine, combined with the characteristics of the ATP pockets of different targets, 4-aminopyrazolopyrimidine derivatives have been widely applied in designing small-molecule inhibitors. In this article, we summarized the reports on 4-aminopyrazolopyrimidine as well as its deformation and the application of its derivatives in designing small-molecule inhibitors and the treatment of diseases.

According to articles reported, the deformation of 4-aminopyrazolopyrimidine was roughly divided into two categories. As shown in Fig. 1, the first category was mainly the heterocyclic modification of the 4-aminopyrazolopyrimidine scaffold. Something the derivatives had in common was that all of them retained the 3-N and 4-NH₂ moieties which could form an H-bond with the amino acid residue in the adenine region. Seven derivatives had been reported: 7H-pyrrolo[2,3-d]pyrimidin-4-amine (A), 1H-pyrazolo[3,4-d]pyrimidin-4-amine (B), imidazo[1,5-a]pyrazin-8-amine (C), 6-amino-7,9-dihydro-8H-purin-8-one (D), 4-amino-1,6-dihydro-7H-pyrazolo[3,4-d]pyridazin-7-one (E), 4-amino-2,6-dihydro-7H-pyrazolo[3,4-d]pyridazin-7-one (F), and pyrrolo[2,1-f][1,2,4]triazin-4-amine (G).

Fig. 1. The 4-aminopyrazolopyrimidine scaffold and its deformation.

The second category was mainly the split of the 4-aminopyrazolopyrimidine scaffold. The first way was the split of the pyrimidine ring which yielded the 5-amino-1H-pyrazole-4-carboxamide ring (H). The amide moiety could form two intermolecular hydrogen bonds with the amino acid residue in the hinge region. Further optimization of the 5-amino-1H-pyrazole-4-carboxamide ring yielded the 4,5,6,7-tetrahydropyrazolo[1,5-a]pyrimidine-3-carboxamide scaffold (I). The last way was the introduction of a new ring based on the original structure, mainly at the NH-7 and C-6 positions (J and Q). According to the target under research, a structure–activity relationship was developed and the ideal structure was ultimately obtained. Until now, a lot of derivatives of 4-aminopyrazolopyrimidine as well as its different proteins as deformation targets, such as BTK, EGFR, FGFR, PI3K, FLT, IGF-1R, RET, VEGFR, Src, mTOR, and so on, have been reported as effective inhibitors.

2. 4-Aminopyrazolopyrimidine derivatives as kinase inhibitors

The signal transduction process plays an important role in cell proliferation, differentiation, migration, metabolism, and programmed cell death. Tyrosine kinases and serine/threonine kinases are important mediators for extra & intra-cellular signal transduction processes. The tyrosine kinases can be divided into two subgroups: receptor tyrosine kinases and non-receptor tyrosine kinases. Receptor tyrosine kinases (RTK), such as EGFR, FGFR, PDGFR, VEGFR, and IGF-1R, are activated by ligand binding at the extracellular domain. Non-receptor tyrosine kinases (NRTKs), such as Src, ABL, JAK, FAK, ACK, and BTK, exist in the cytoplasm of cells and play important roles in intracellular signal transductions.

2.1. FGFR inhibitors

FGFRs (FGFR1–4) belong to transmembrane receptor tyrosine kinases. The kinase domains of FGFR1–4 share a high structural similarity (especially FGFR1–3). Alterations of FGFRs have been identified in breast tumor, lung cancer, rhabdomyosarcoma, bladder cancer, and so on.² They have emerged as a promising target for the treatment of cancer.

The first FGFR irreversible inhibitor, TAS-120 (Fig. 3), takes 1H-pyrazolo[3,4-d]pyrimidin-4-amine as the scaffold. 3,5-Dimethoxyphenylethynyl was introduced and it could stretch into the back pocket to form an H-bond with Asp-641. Then, the acrylamide moiety could covalently bind to Cys488 (Fig. 2A).³ TAS-120 selectively inhibited FGFR1/2/3/4 with IC₅₀ values of 3.9, 1.3, 1.6, and 8.3 nM. Moreover, TAS-120 was identified to inhibit several drug-resistant FGFR2 mutants (FGFR2^WT, FGFR2^V565L, FGFR2^N550H, and FGFR2^E566G) with great potency (IC₅₀ values less than 4 nM). In the first-in-human study, TAS-120 was affirmed to demonstrate safety, pharmacodynamic activity, and preliminary responses in patients with advanced solid tumors. TAS-120 was tested to have a lower drug-resistant possibility than reversible FGFR inhibitors and exhibit robust activity in FGFR-deregulated cancer lines as well as animal tumor models.^4,5

Fig. 3. Reported 4-aminopyrazolopyrimidine derivatives as FGFR inhibitors.

Fig. 2. (A) Binding mode of TAS-120 with FGFR1 (PDB: 6MZW). (B) Binding mode of compound 5 with FGFR1 (PDB: 6ITJ).

Wang et al. reported compound 2 (FGFR1 IC₅₀ < 0.3 nM; FGFR2 IC₅₀ = 0.8 nM; FGFR3 IC₅₀ = 9.4 nM; FGFR4 IC₅₀ = 5.3 nM; SD rats, 10 mg kg⁻¹, T_max = 2.75 h, C_max = 77.7 ng mL⁻¹, AUC = 277 ng h mL⁻¹, T_1/2 = 1.73 h) as an irreversible FGFR inhibitor with an 7H-pyrrolo[2,3-d]pyrimidin-4-amine scaffold via introducing the 7-methoxy-5-methylbenzo[b]thiophen-2-yl which could point into the back pocket properly (Fig. 3). Compound 2 displayed significantly antiproliferative effects on FGFR-mediated cancer cell lines (NCI-H1581 IC₅₀ < 0.5 nM; SNU-16 IC₅₀ < 0.1 nM; RT-112 IC₅₀ < 0.5 nM; HuH7 IC₅₀ < 1.5 nM). Compound 2 effectively inhibited the phosphorylation of FGFR1/2/3 and the downstream signaling pathways of PLCγ and Erk. In a mouse xenograft tumor model, it showed its potential as a drug candidate (NCI-H1581 100 mg kg⁻¹ TGI = 98.8%; SNY-16 50 mg kg⁻¹ TGI = 85.8%).⁶

Collin et al. adopted the computational chemistry approach to analyze the ATP binding site of FGFR1 and approved the 1H-pyrazolo[3,4-d]pyrimidin-4-amine scaffold compound 3 with an IC₅₀ value of 115 nM against FGFR1.⁷ Further optimization of compound 3 yielded the clinical candidate inhibitor compound 4 (Rogaratinib, FGFR1 IC₅₀ = 15 nM, FGFR2 IC₅₀ < 1 nM, FGFR3 IC₅₀ = 19 nM, FGFR4 IC₅₀ = 33 nM) with good DMPK properties (IV 1 mg kg⁻¹ CL = 0.45 L h⁻¹ kg⁻¹, T_1/2 = 12 h; PO 3 mg kg⁻¹ AUC = 0.78 kg h L⁻¹, T_1/2 = 12 h, F = 22%). Reduced tumor growth was observed in pre-clinical models with different FGFR alterations both in monotherapy and combination therapy (Fig. 3).

In 2019, Wang et al. verified the potential of a series of pyrazolo[3,4-d]pyridazinone compounds to be FGFR inhibitors through a computer-aided drug design strategy (Fig. 2B). The FGFR inhibitor with the most potential, compound 5 (FGFR1 IC₅₀ = 20.4 nM, FGFR2 IC₅₀ = 7.2 nM, SD rats, IV 10 mg kg⁻¹, AUC_0−t = 4412.8 ng h mL⁻¹, T_1/2 = 8.55 h; IP 20 mg kg⁻¹, AUC_0−t = 16790.8 ng h mL⁻¹, T_1/2 = 4.63 h), was more stable and soluble than the others and showed high plasma exposure as well as an acceptable T_1/2 value (Fig. 3). Compound 5 showed IC₅₀ values of >40 μM against HERG which reached the standard of compounds to be a safe drug candidate (standard >10 μM).⁸ In the meantime, Wu et al. shared their attempt at modification of the hit reported by Wang et al. and eventually obtained the preferred compound 6 (FGFR1 IC₅₀ = 3 nM, BAF3/TEL-FGFR1 IC₅₀ < 0.3 nM, IV 1 mg kg⁻¹, AUC_0−t = 281 ng h mL⁻¹, T_1/2 = 1.39 h; IP 3 mg kg⁻¹, AUC_0-t = 704 ng h mL⁻¹, T_1/2 = 3.70 h) which showed high selectivity against other tyrosine kinases with a methyl group anchored to the back pocket. In the FGFR1-amplified NCI-H1581 xenograft model, compound 6 displayed a high potent antitumor efficacy (50 mg kg⁻¹, TGI = 91.6%; 10 mg kg⁻¹ TGI = 74.4%).⁹

Recently, Duan et al. have reported their breakthrough in exploiting FGFR inhibitors via the modification of the purine scaffold and introducing the 5-bromo-7-methoxybenzo[b]thiophene moiety to replace the 1-ethynyl-3,5-dimethoxybenzene moiety of TAS-120. In this research, compound 7 (FGFR1 IC₅₀ < 0.3 nM, FGFR2 IC₅₀ = 1.1 nM, FGFR3 IC₅₀ < 0.3 nM, FGFR4 IC₅₀ = 0.5 nM) which took imidazo[1,5-a]pyrazin-8-amine as the scaffold was the preferred one with good selectivity against other kinases (Fig. 3). Compared to most FGFR inhibitors, compound 7 exhibited a more than 1100-fold selectivity against the recombinant kinase insert domain receptor (KDR), which showed that compound 7 could induce the side effects of multitarget inhibitors. Compound 7 was identified to significantly inhibit FGFR signaling at low nanomolar concentrations. In the FGFR1-amplified NCI-H1581 xenograft model, compound 7 significantly suppressed tumor growth at a dose of 5 mg kg⁻¹ (TGI = 62.6%) or 25 mg kg⁻¹ (TGI = 84.6%). What was strikingly noticeable was that the TGI at 50 mg kg⁻¹ of TAS-120 in this xenograft model was 42%. In the human HCC Hep3B xenograft model, compound 7 showed a significant inhibitory effect (5 mg kg⁻¹ TGI = 97.8%) compared with TAS-120 (5 mg kg⁻¹, 29.8%).¹⁰

2.2. EGFR inhibitors

The epidermal growth factor receptor (EGFR), also called ErbB1/HER1, is a member of the ErbB family which belongs to receptor tyrosine kinases. The ErbB family includes EGFR, ErbB2 (HER2), ErbB3 (HER3), and ErbB4 (HER4). The ErbB family was identified to be closely related to the occurrence of lung cancer (especially NSCLC), breast cancer, gastric carcinoma, and so on.^11,12

Ibrutinib (PCI-32765), an irreversible BTK kinase inhibitor, had been extensively studied in a variety of hematopoietic malignancies, such as mantle cell lymphoma (MCL) and chronic lymphatic lymphoma (CLL). Besides its potent activity against BTK kinase, previous studies had shown that ibrutinib also potently inhibited BMX, BLK, EGFR, and HER2. Wu et al. reported the progress they achieved for ibrutinib in the evaluation of anti-tumor activities in NSCLC.¹³ Ibrutinib strongly affected EGFR mediated signaling pathways and induced apoptosis and cell cycle arrest (G0/G1) in mutant EGFR but not EGFR^WT cells. Ibrutinib exhibited excellent antitumor activities in EGFR mutant NSCLC, and it is now in phase I/II clinical trials for NSCLC (NCT02321540). On the basis of the scaffold of ibrutinib, Wang et al. reported their work on the design and anti-tumor activity evaluation of CHMFL-EGFR-202 (Fig. 5).¹⁴ The modification of the hydrophobic region of ibrutinib yielded the desired compound. CHMFL-EGFR-202 potently inhibited EGFR primary mutants (L858R, del19) and the drug-resistant mutant L858R/T790M. Analysis of the X-ray crystal structure of the compound with EGFR^T790M protein gave a conclusion that compound 8 adopted a “DFG-in-C-helix-out” inactive binding conformation (Fig. 4). Also, compound 8 was verified to inhibit several kinases which were proved to be closely associated with NSCLC, such as MEK, ERBB2, and ERBB4. Then, the antitumor efficacy of compound 8 was evaluated in H1975 (EGFR-L858R/T790M) and PC9 (EGFR-Del19) cell line-inoculated xenograft mouse models. Compound 8 exhibited good antitumor efficacy (H1975, 50 mg kg⁻¹, TGI = 45%; PC9 100 mg kg⁻¹, TGI = 50%) as well as dose-dependent tumor cell proliferation suppression efficacy without obvious toxicity.

Fig. 5. The reported 4-aminopyrazolopyrimidine derivatives as EGFR inhibitors.

Fig. 4. Binding mode of CHMFL-EGFR-202 with EGFR (PDB: 5GNK).

Hasako et al. reported their latest discovery of EGFR inhibitors with a novel scaffold. A new ring was introduced based on the original scaffold, and eventually they obtained the optimal compound 9 (TAS6417, Fig. 5) which could potently inhibit the EGFR exon 20 insertion and selectivity. TAS6417 potently inhibited the EGFR phosphorylation in NIH/3T3 cells which stably expressed exon 20 insertion mutant human EGFRs and Ba/F3 cells driven by various EGFR exon 20 insertion mutations. TAS6417 provided a survival benefit with good tolerability in a lung orthotopic implantation mouse model, which supported the clinical evaluation of TAS6417 as an efficacious drug candidate for patients with NSCLC harboring EGFR exon 20 insertion mutations.¹⁵

Given the reason that the reported HER2 inhibitors showed no selectivity for ERBB family kinases, leading to the resiliency of the HER2/HER3 pathway in HER2-activated cancers, Irie et al. discovered the first covalent inhibitor TAS0728 with high potency and selectivity against HER2, rather than EGFR, which may reduce the toxicity of targeting EGFR in normal tissues. At concentrations of 30 nmol L⁻¹ in HER2-overexpressing SK-BR-3 cells, the phosphorylation of HER2 was completely inhibited and the phosphorylation of EGFR was only weakly suppressed at a concentration of 1000 nmol L⁻¹ in EGFR-overexpressing A-431 cells. In addition, phosphorylation of cells with HER2 mutants was verified to be inhibited equivalently to HER2^WT. In an NCI-N87 HER2-amplified human gastric cancer mouse model, TAS0728 significantly regressed the tumor after 60 mg kg per day dosing. In an NCI-N87 peritoneal dissemination model, any evident toxicity (diarrhea, body weight loss) was observed on long-term dosing of TAS0728, which suggested that TAS0728 (compound 10, Fig. 5) had good tolerability of long-term administration. In summary, TAS0728 was proved to be a promising therapeutic option for the treatment of cancers harboring HER2 gene abnormalities. Clinical trial evaluation of TAS0728 in patients with solid tumors is ongoing (NCT03410927).¹⁶

With the aim of designing selective HER3 inhibitors, high throughput screening was performed by Lim's group. Covalent modification of the lead compound and structure-guided hybridization generated the reported HER3 inhibitor TX1-85-1 (compound 11, Fig. 5) with an IC₅₀ value of 23 nM. TX1-85-1 was the first selective HER3 ligand, which formed a covalent bond with Cys721 located in the ATP-binding site of HER3. However, TX1-85-1 exhibited no inhibition of the downstream pathway of HER3 at a concentration of 5 μM, which suggested that TX1-85-1 was not capable of inhibiting HER3-dependent growth and signaling in the line tested.¹⁷ Based on the structure of TX1-85-1, they introduced the hydrophobic group adamantine to induce proteasomal degradation. Eventually, they obtained the desired compound TX2-121-1 (Fig. 5) with good kinase activity (IC₅₀ = 49 nM). Not only did compound TX1-121-1 attenuate the phosphorylation of downstream effectors of HER3 but it also perturbed the hetero-dimerization of HER3 with either HER2 or c-Met and induced the partial degradation of HER3. The authors first approved the application of this small-molecule inhibitor which could selectively target HER3-dependent functions.¹⁸

To develop EGFR^T790M inhibitors structurally based on 4-aminopyrazolopyrimidine, Engel et al. expounded their work on finding a series of covalent inhibitors, namely, 13 (EGFR^T790M/C797S IC₅₀ < 1 nM), 14 (EGFR^T790M/C797S IC₅₀ = 1.9 nM), and 15 (EGFR^T790M/C797S IC₅₀ = 1.9 nM), which could target the drug resistant T790M in EGFR with a strong lipophilic character in the C-3 position (Fig. 5).¹⁹ Sterically advanced bicyclic moieties were explored to be the optimal functional group which could exploit the maximum tolerable size of lipophilic extensions into the pocket adjacent to the gatekeeper residue Met790. In the drug-resistant NSCLC cell line H1975 which harbored EGFR-L858R/T790M, compounds 13–15 showed excellent inhibitory effects (EC₅₀ = 0.14 μM, 0.49 μM, 0.23 μM) without significantly impacting the EGFR wild-type cell lines (A431, EC₅₀ = 3.6 μM, 29.4 μM, 6.78 μM; A549, EC₅₀ = 11.6 μM, no inhibition, 23.3 μM; H358, EC₅₀ = 25.7 μM, 60.0 μM, 29.5 μM), which ultimately overcame issues of on-target toxicity. Further evaluation of the in vitro ADME profile of the compounds suggested that 15 exhibited promising cellular permeability and aqueous kinetic solubility.

In the meantime, they also reported a series of pyrazolo[3,4-d]-pyrimidine scaffold covalent reversible inhibitors via introducing an electron-withdrawing group on acrylamide to achieve a reversible effect. Unexpectedly, these compounds showed no selectivity against another EGFR mutation. Meanwhile, compound 16 (Fig. 5) showed the most potent inhibitory effect on each EGFR mutant. Further investigations on this covalent-reversible warhead are needed to achieve a better inhibitory effect.²⁰

2.3. FLT3 inhibitors

FMS-like tyrosine kinase 3 (FLT3), which belongs to type III receptor tyrosine kinases, plays a significant role in the differentiation and survival of hematopoietic stem cells in bone marrow. FLT3 has been observed to be overexpressed in AML and ALL. Mutations of FLT3, such as FLT3-internal tandem duplication (FLT3-ITD), FLT3-D835Y/E/V/H, and FLT3-K663Q, had been observed to be associated with AML. It was noted that there would be synthetic lethal myelosuppression toxicity if FLT3 inhibitors also showed good activities against c-KIT.^21,22

Similarly, on the basis of the reported BTK inhibitor ibrutinib, Li et al. deeply expounded their optimization of this structure. The binding mode of the compound with BTK and FLT3 kinases suggested that the 4-aminopyrazolopyrimidine scaffold could bind to the hinge region rationally and benzyl phenyl also adapted to the hydrophobic pocket perfectly. Modification of the hydrophobic pocket generated compound 17 (Fig. 7) with good activities (BaF3-FLT3-ITD GIC₅₀ = 11 nM; BaF3-TEL-c-KIT GIC₅₀ = 1900 nM; FLT3-ITD positive AML cancer cell lines MV4-11 GI₅₀ = 22 nM; MOLM13/14 GI₅₀ = 21 nM/42 nM). In vivo studies showed that compound 17 has good bioavailability (30%) and good distribution but relatively fast clearance. Antitumor efficacy evaluation of compound 17 in an MV4-11 cell inoculated xenograft mouse model (50 mg kg per day, TGI = 88%) showed the potential of this compound to be a drug candidate for FLT3-ITD positive AML.²³

Fig. 7. Reported 4-aminopyrazolopyrimidine derivatives as FLT3 inhibitors.

To overcome secondary mutations, especially of F691L, located at the gatekeeper position in FLT3, they proceeded to excavate the binding mode of ibrutinib with FLT3 kinase. They found that the mutation of Phe691 to L691 led to the loss of beneficial π–π interaction. Hence, a type II binding conformation was approved because of its essential amide or urea bond at this position. In addition, they attempted to flip pyrazolopyrimidine and eventually obtained the reported compound 18 (CHMFL-FLT3-213, Fig. 7) with highly potent inhibitory effects (FLT3^WT IC₅₀ = 91 nM; FLT3-ITD IC₅₀ = 9 nM; BaF3-FLT3-ITD-F691L GI₅₀ = 2 nM; FLT3-ITD positive AML cancer cell lines MV4-11 GI₅₀ < 0.3 nM) (Fig. 6A). Compound 18 completely blocked FLT3 autophosphorylation at the Tyr589/591 site (1 nM) as well as the downstream signaling mediators pSTAT5 (Tyr694), pERK (Thr202/Tyr204), and pAKT (Thr308/Ser473) (10 nM). It was shown that compound 18 was suitable for oral administration (T_1/2 = 3.61 h, F = 18.8%) via in vivo evaluation. Antitumor efficacy evaluation of compound 18 in an MV4-11 cell inoculated xenograft mouse model (15 mg kg per day, TGI = 97%) also showed the potential of this compound to be a drug candidate for FLT3-ITD positive AML. However, what was worth noting was that CHMFL-FLT3-213 almost had no selectivity against c-KIT and CDK8/11 kinases which might induce the myelosuppression toxicity and increase the risk of pleiotropic toxicity.²⁴

Fig. 6. (A) Binding mode of CHMFL-FLT3-213 with FLT3 (PDB: 4XUF). (B) Binding mode of NVP-AEW541 with IGF-1R (PDB: 5HZN).

Because of its low selectivity and relatively low bioavailability, Yuan et al. further analyzed the binding mode of compound 18; they tried to keep the pyrrolo[2,3-d]pyrimidine scaffold reported for compound 17 and modified the ethylmorpholine moiety with the aim of hunting a new selective FLT3 inhibitor. They found that compound 19 (Fig. 7) with allylic substitution exhibited a satisfactory bioavailability (F = 59.5%) and antiproliferation effect (FLT3 IC₅₀ = 7 nM; MV4-11 IC₅₀ = 0.089 nM; Molm-13 IC₅₀ = 0.022 nM; BaF3-FLT3-ITD-F691L IC₅₀ = 12.99 nM). Expectedly, compound 18 was bound well to the DFG-out inactive conformation and exhibited an over 40-fold selectivity for FLT3 relative to c-KIT kinase which was supposed to reduce myelosuppression toxicity. In the oral administration of the AML xenograft model, it also showed a strong inhibitory effect (Molm-13, 3 mg kg⁻¹ TGI = 94.5%, MV4-11, 10 mg kg⁻¹, TGI = 100%).²⁵

2.4. IGF-1R inhibitors

The insulin-like growth factor 1 receptor (IGF-1R), a tetrameric transmembrane receptor tyrosine kinase, plays an important role in cell proliferation, survival, metastasis, and angiogenesis.²⁶ IGF-1R was verified to be associated with acquired resistance of tumor cells. Clinical studies have shown that IGF-1R is closely related to breast cancer, prostate cancer, and lung cancer.

In 2009, Mulvihill et al. reported their progress in modifying IGF-1R inhibitors and put forward the imidazo[1,5-a]pyrazin-8-amine scaffold compound 20 (OSI-906, linsitinib, Fig. 8) as a selective and potent dual inhibitor of IGF-1R (IC₅₀ = 35 nM) and IR (IC₅₀ = 75 nM). The 2-phenylquinolinyl moiety could accurately anchor to the hydrophobic pocket. OSI-906 was identified to be an orally efficacious dual inhibitor (mouse 5–250 mg kg⁻¹ TGI > 84%; rat 100 mg kg⁻¹ TGI = 92%; dog 5 mg kg⁻¹ TGI = 64%). Compound 20 could efficaciously inhibit the growth of the tumor in an LISN xenograft model (75 mg kg⁻¹, TGI = 100%).²⁷

Fig. 8. Reported 4-aminopyrazolopyrimidine derivatives as IGF-1R inhibitors.

In addition, Novartis reported a series of pyrrolo[2,3-d]pyrimidine derivatives as IGF-1R inhibitors, namely, NVP-AEW541 (compound 21, IGF-1R IC₅₀ = 86 nM, IR IC₅₀ = 2300 nM), NVP-ADW742 (compound 22, IGF-1R IC₅₀ = 170 nM, IR IC₅₀ = 2800 nM), and compound 23 (IGF-1R IC₅₀ = 12 nM), and its application in the treatment of multiple myeloma and other solid tumors (Fig. 8). In 2004, NVP-ADW541 and NVP-ADW742 were approved as IGF-1R inhibitors with high selectivity against IR. NVP-ADW541 could inhibit IGF-1R signaling in NWT-21 tumor xenografts and reduce the growth of IGF-1R-driven fibrosarcoma significantly. Meanwhile, NVP-ADW742 was confirmed to be effective in an orthotopic xenograft MM model whether in monotherapy or in combination with chemotherapy.^28,29 Taking NVP-ADW541 as the starting point, they further modified the structure, including the back-pocket and the hydrophobic region II. They first introduced achiral [2.2.1]-bicyclic ether methyl ether to replace the back-pocket moiety of NVP-ADW541 (Fig. 6B). Most of the compounds in this series exhibited good IGF-1R kinase activities, but the low pK_a values associated with the reduced affinity for hERG channels have limited the development of these compounds. The substitution of achiral [2.2.1]-bicyclic ether methyl ether by (S)-2-tetrahydrofuran methyl ether 6-fluorophenyl ether generated the ideal compound 23 with affinities for the hERG channel >30 μM and good rat PK data (CL = 28 mL min⁻¹ kg⁻¹, V_ss = 5.0 L kg⁻¹, AUC = 1395 nmol h L⁻¹, T_1/2 = 3.1 h, F = 100%) compared to NVP-AEW541 (CL = 123 mL min⁻¹ kg⁻¹, V_ss = 19 L kg⁻¹, AUC = 311 nmol h L⁻¹, T_1/2 = 2.0 h, F = 23%). In nude mice bearing NIH3T3 xenotransplants, a clear reduction of tumor volumes was observed (53% regression) for compound 23 when administered orally twice-daily at doses of 40 mg kg⁻¹ for 1 week.³⁰

2.5. BTK inhibitors

BTK (Bruton's tyrosine kinase), one of the nonreceptor tyrosine kinases, belongs to the Tec kinase family. Except for the T cells and terminally differentiated plasma cells, BTK is expressed in all hematopoietic cells. It plays a crucial role in the activation of B-cell antigen receptor-signaling and B-cells. Constitutive activation of the signaling pathway downstream of BCR is a hallmark of B-cell malignancies.^31,32 Hence, BTK is considered as a promising target for the treatment of various diseases involving B cell and/or macrophage activation such as B cell malignancies, asthma, rheumatoid arthritis, systemic lupus erythematosus, and multiple sclerosis.

In 2007, Pan et al. reported the first 4-aminopyrazolopyrimidine inhibitor, compound 28 (Fig. 10), with IC₅₀ values of 8.2 nM against BTK.³³ Kinase screening of compound 28 indicated that it also inhibited certain Tec and Src family kinases. Based on the electrophilic center, Cys481, of BTK, the group designed the first approved Bruton's tyrosine kinase (BTK) inhibitor, ibrutinib (compound 24, Fig. 10), which covalently targeted Cys481 and took 1H-pyrazolo[3,4-d]pyrimidin-4-amine as the scaffold. As shown in Fig. 9A, the primary amine NH₂ formed two hydrogen bonds with the gatekeeper Thr474 hydroxyl and the backbone carbonyl of Glu475, the N-3 nitrogen of the pyrimidine ring interacted with the backbone NH of Met477 at the hinge region, and the diphenyl ether moiety occupied the hydrophobic pocket behind the Thr474 gatekeeper residue and displayed an edge-to-face aromatic interaction with Phe540.³⁴ Ibrutinib was approved for the treatment of MCL, CLL (including relapsed or refractory CLL), WM, and chronic graft-versus-host disease (cGVHD). Expectedly for BTK, ibrutinib also interacted with other proteins, such as EGFR, JAK3, HER2, and ITK, which may lead to some undesired side effects.

Fig. 10. Reported 4-aminopyrazolopyrimidine derivatives as BTK inhibitors.

Fig. 9. (A) Binding mode of ibrutinib with BTK (PDB: 5YU9). (B) Binding mode of zanubrutinib with BTK (PDB: 6J6M); (C) binding mode of compound 34 with BTK (PDB: 3GEN).

Based on the structure of ibrutinib, Acerta Pharma developed another BTK inhibitor, acalabrutinib (compound 25, Fig. 10), with good kinase activity (BTK IC₅₀ = 3 nM) and less inhibitory effects on the kinase activities of EGFR, JAK3, HER2, and ITK. Due to its reduced activity against EGFR and TEC, and higher efficiency against BTK, acalabrutinib may bring fewer side effects. Acalabrutinib was a novel BTK inhibitor which took imidazo[1,5-a]pyrazin-8-amine as the scaffold and but-2-ynamide as the warhead. The 2-butylacetylene warhead could also covalently bind to Cys481 via Michael addition. Acalabrutinib was approved as an orphan drug by the FDA for the treatment of MCL, CLL, and WM.³⁵

For compound 26 (Fig. 10), tirabrutinib, another irreversible second-generation BTK inhibitor, the covalent design of acalabrutinib was followed and 6-amino-7,9-dihydro-8H-purin-8-one was chosen as the new scaffold.³⁶

In 2019, Ran et al. reported compound 27 (Fig. 10) with an IC₅₀ value of 36 nM via a structure-based design. What was different with ibrutinib was that the author put forward a new warhead, chloroacetamide, which covalently bound to Cys481 via nucleophilic substitution. In MCL cell lines, compound 27 (Mino cell line IC₅₀ = 0.4 μM; Jeko-1 cell line IC₅₀ = 0.4 μM; Z138 cell line IC₅₀ = 0.4 μM; Maver cell line IC₅₀ = 0.4 μM) showed robust antiproliferative activities and lower IC₅₀ values compared to ibrutinib (Mino cell line IC₅₀ = 15.7 μM; Jeko-1 cell line IC₅₀ = 1.1 μM; Z138 cell line IC₅₀ = 9.7 μM; Maver cell line IC₅₀ = 7.8 μM). At a concentration of 0.5 μM in Z138 cell lines, compound 26 could completely inhibit the phosphorylation of BTK as well as the PLCγ-2 downstream signaling pathway. Compound 27 exhibited robust antiproliferative effects in both mantle cell lymphoma cell lines and primary patient tumor cells. A longer half-life (16.8 min) and a lower clearance (82.5 μL min⁻¹ mg⁻¹) were observed in the stability estimation of compound 27 compared with ibrutinib (T_1/2 = 1.6 min, CL = 846.7 μL min⁻¹ mg⁻¹).^37,38

Zhang et al. devoted themselves to searching for a novel central core. Taking the ibrutinib scaffold as the foundation, the introduction of a new ring made no sense in the improvement of the kinase activities. Eventually, they found the novel 4-amino-1H-pyrazolo[3,4-d]pyridazin-7(6H)-one scaffold. Further optimization of the linker and the warhead generated the target compound 29 (BTK IC₅₀ = 2.1 nM, Fig. 10). As demonstrated in in vivo pharmacokinetic studies in rats and mice, compared to ibrutinib, it exhibited better plasma exposure (rat, 642 vs. 318 ng h⁻¹ mL⁻¹; mice, 128 vs. 21.4 ng h⁻¹ mL⁻¹), C_max (rat, 466 vs. 150 ng mL⁻¹; mice, 252 vs. 55.3 ng mL⁻¹) and oral bioavailability (rat, 23.7% vs. 10.3%; mice, 11.2% vs. 1.8%). Evaluation of the efficacy of compound 29 in a mouse collagen-induced arthritis model showed an equivalent effect to ibrutinib (10 mg kg⁻¹) at a dose of 3 mg kg⁻¹ with no significant body weight loss.³⁹

Considering that BTK and PI3Kδ kinases were responsible for the transduction of the B-cell signal, Pujala et al. analyzed the structure of the BTK inhibitor PCI-29732 and the PI3Kδ inhibitor SW13. They employed their commonly used 1H-pyrazolo[3,4-d]pyrimidin-4-amine scaffold. Modification of the back pocket and the hydrophilic area yielded the dual inhibitor compound 30 (BTK IC₅₀ = 32 nM, PI3Kδ IC₅₀ = 16 nM, Fig. 10) with a lower systemic clearance (CL = 1.46 L h⁻¹ kg⁻¹) and considerable oral bioavailability (F = 57%).⁴⁰

In 2013, Rankin et al. described the pharmacological properties of their newly found BTK inhibitor, compound 31 (PF-06250112, BTK IC₅₀ = 0.5 nM; BMX IC₅₀ = 0.9 nM; TEC IC₅₀ = 1.2 nM), which adopted 5-amino-1H-pyrazole-4-carboxamide as the scaffold. PF-06250112 could not only inhibit BCR-mediated signaling and proliferation but also FcR-mediated activation (Fig. 10).⁴¹

In 2019, Paekche Shenzhou disclosed the development process of compound 33, which was approved by the FDA for the treatment of MCL, CLL, and SLL. Starting from the 4-aminopyrazolopyrimidine derivatives reported, splitting the 4-aminopyrimidine ring yielded the compound 31 analogue, compound 32 (Fig. 10), which displayed significant potency against BTK (IC₅₀ = 0.17 nM) and EGFR (IC₅₀ = 0.21 nM). Then, innovation was pursued on the scaffold of compound 32. Further optimization of the aromatic character and coplanarity of the bicyclic core gave the ideal 4,5,6,7-tetrahydropyrazolo[1,5-a]pyrimidine-3-carboxamide scaffold (Fig. 9B). Compound 33 (Fig. 10) showed a relatively lower clearance rate (76.8 mL min⁻¹ kg⁻¹) and a higher exposure C_max (235 ng mL⁻¹), AUC (257 ng h mL⁻¹), and bioavailability (23.6%) in rats. Pharmacodynamic characterization of compound 33 in mice indicated that it was about 3-fold more potent than ibrutinib in mouse PD assays. In an OCI-LY10 xenograft model, compound 33 could significantly induce the TGI of female NCG mice (ibrutinib, 2.5 mg kg⁻¹ TGI = 61%; 5 mg kg⁻¹ TGI = 77%, compound 33, 2.5 mg kg⁻¹ TGI = 76%; 5 mg kg⁻¹ TGI = 88%).⁴²

Intriguingly, starting with ibrutinib, Xue et al. tried to use the ring-merging strategy to design novel BTK inhibitors. Combining the pyrazolo[3,4-d]pyrimidine scaffold and the piperidine ring into a tricyclic skeleton eventually yielded compound 34 (Fig. 10) with IC₅₀ values of 0.4 nM against BTK and 16 nM against BTK-dependent TMD8 cells (Fig. 9C). The selectivities of compound 34 against BMX (174.1 vs. 5.8 nM), RET (20.3 vs. 5.2 nM), and ErbB2 (23.2 vs. 1.5 nM) were inferior to those of ibrutinib. In addition, in a TMD8 cell derived animal xenograft model, at a dose of 5 mg kg⁻¹ of oral administration, compound 34 showed almost no antitumor efficacy (relative tumor volume 11.1), but when the dose was increased to 15 mg kg⁻¹, it significantly suppressed the growth of the tumor (RTV 5.3) which was inferior to ibrutinib (25 mg kg⁻¹, RTV 6.6).⁴³

Following the design strategy of compound 32, Ma et al. modified the hinge region of compound 24 and reported compound 35 (Fig. 10) as a highly selective covalent BTK inhibitor with an IC₅₀ value of 2.7 nM and a 317-fold selectivity against BMX. Compound 35 showed better antiproliferative activities than compound 24 in BTK overexpression in human B-cell lymphoma Ramos cells (2.5 μM, 31.3 μM), MOLM13 cells (3.7 μM, 358 μM), and TMD8 cells (0.025 μM, 0.024 μM). The evaluation of drug-like properties showed the favorable PK profile of compound 35 (IV 5 mg kg⁻¹, T_1/2 = 1.32 h, CL = 13.4 mL min⁻¹ kg⁻¹; PO 5 mg kg⁻¹, C_max = 501 ng mL⁻¹, AUC = 1996 ng h mL⁻¹, F = 34%). Compound 35 showed no significant effect on the enzymatic activity of CYP1A2, CYP2C19, CYP2D6, CYP3A4M, and CYP3A4T, which meant that the drug–drug interactions may not have happened. At a dose of 7.5 mg kg⁻¹ for 21 consecutive days (twice daily) in female CB-17 SCID nude mice with human TMD8 tumor xenografts, a better tumor inhibition could be observed with a dose-dependent increase in drug exposure (compound 35 TGI = 96%, AUC_0−8h = 1782 ng h mL⁻¹, C_max = 1003 ng mL⁻¹, compound 24 TGI = 91%); the tumor growth was completely inhibited and the tumor volume was significantly reduced at a dose of 15 mg kg⁻¹ (TGI = 106%).⁴⁴

Given that a small portion of patients with chronic lymphocytic leukemia relapsed due to the mutation of Cys481 to Ser481 under the treatment of ibrutinib. Research had confirmed that compounds that did not covalently bind to Cys481 could be efficacious. Hence, Liu et al. reported a series of 8-aminoimidazo[1,5-a]pyrazine reversible BTK inhibitors (compounds 36–38, Fig. 11) which did not utilize acrylamide. They just introduced the amide carbonyl which was confirmed to form an H-bond with the amide nitrogen of peptides G480 and C481. All three compounds exhibited excellent BTK kinase activities (0.27 nM, 0.32 nM, and 0.31 nM, respectively) as well as selectivity against other kinases (11–1200 fold). In rat pharmacokinetic studies of compounds 36–38, after oral administration, compounds 36 and 37 showed high bioavailabilities (F = 84%, 77%).⁴⁵

Fig. 11. Reported 4-aminopyrazolopyrimidine derivatives as BTK inhibitors.

With the revelation of the fact that it was the drug–target residence time rather than affinity that influenced in vivo pharmacodynamic activity and disease efficacy, Bradshaw et al. attempted to bring in a reversible cyanoacrylamide-based electrophile on the basis of the structure of ibrutinib. With the substitution of cyanoacrylamides with branched-alkyl capping groups, the residence time was tested. And ultimately, they obtained the desired compound 39 (Fig. 11) with an IC₅₀ value of 1.3 nM against BTK. The limitation was that there must be enough room in the cysteine to accommodate small molecular groups.⁴⁶

2.6. Src inhibitors

Src is a non-receptor tyrosine kinase that has been shown to be a key regulator of numerous intracellular signaling pathways. Abnormal activation or amplification of Src has been detected in TNBC and demonstrated to play an important role in the proliferation, migration, and invasion of breast cancer cell lines. Src was seen as a target for the treatment of TNBC.⁴⁷ What was worth noting was that BCR-ABL had been found to be the opposite of the therapy of breast cancer. Hence, the selectivity against ABL had become a significant index for Src inhibitors.

A lot of 4-aminopyrazolopyrimidine derivatives have been identified to be potential inhibitors. Previously, PP1 (compound 40) and PP2 (compound 41) with pyrazolo[3,4-d]pyrimidin-4-amine as the scaffold were reported as Src kinase family inhibitors (Fig. 13). PP1 and PP2 inhibited LCK and Fyn kinase with IC₅₀ values of 3–6 nM.⁴⁸ Dar et al. also reported AD-80 (compound 42) as an RET kinase inhibitor with an IC₅₀ value of 4 nM. Compound 42 also exhibited robust activity against Src (no specific value was given).⁴⁹

Fig. 13. Reported 4-aminopyrazolopyrimidine derivatives as Src inhibitors.

Given that PP2 was less selective for other kinases, Brandvold et al. analyzed the co-crystal structure of PP2 with respect to Src-kinase family kinases. They found a special pocket formed by the P-loop that appeared only in c-Src kinase. Decoration of PP2 yielded the final candidate compound 43 (K_i = 44 nM, K_d = 86 nM) with 1,5-disubstituted benzyl triazoles (Fig. 13). Different from PP2, compound 43 exhibited high selectivity against ABL even though they shared a high homology (69%). However, subsequent studies showed that the P-loop pocket had no effect on the selectivity of compound 43 for ABL kinase. Compound 43 also showed different activities against other Src kinase families – LCK (160 nM), FGR (240 nM), and Yes (720 nM). At a single concentration of 10 μM, mean growth across the 57 cell lines tested was 71%. Seven cell lines showed <50% growth.⁵⁰

Besides that, they found that panobinostat (an HDAC inhibitor) showed high synergism with compound 43 (Fig. 13) in the SK-BR-3 cell line (Her2⁺ breast cancer cell line), as Src levels were down-regulated by inhibiting Src transcription. Hence, they combined the structures of compound 43, vorinostat, and panobinostat, and they obtained the Src and HDAC dual inhibitor compound 49 (Src IC₅₀ = 138 nM; HDAC1 IC₅₀ = 0.26 nM, Fig. 13). Compound 49 was more efficacious in inhibiting the growth of the SK-BR-3 cell line with a cellular therapeutic index of 23.5 (the therapeutic index of compound 43/vorinostat/compound 43 + vorinostat, 0.9/4.8/6/8) which highlighted the advantage of chimeric inhibition compared to dual-agent targeting.⁵¹

In 1999, Missbach et al. described the design of compound 44 (CGP77675, Fig. 13) as a potent inhibitor of Src family kinases (Src IC₅₀ = 20 nM; EGFR IC₅₀ = 150 nM).⁵²

In 2000, Wilson et al. reported the pyrrolo-pyrimidine inhibitor compound 45 (Fig. 13) which was designed to improve potency against Src family kinases (Src IC₅₀ = 9 nM; LCK IC₅₀ < 3 nM; Lyn IC₅₀ < 3 nM). Further study of compound 45 in the treatment of CML confirmed that compound 45 could block Ph⁺ CML cell growth and induce apoptosis in a dose-dependent manner.⁵³

Starting from the Src family kinase inhibitor PP1 (compound 39), Fraser et al. found the novel Src inhibitor compound 46 (ECF-506, Src IC₅₀ < 0.5 nM, ABL IC₅₀ = 479 nM, selectivity >1000) through ligand-based design and phenotypic screening in an iterative manner (Fig. 13). They chose to replace the tert-butyl of PP1 with flexible water-solubilizing groups. Further study of the methoxy group, pyrimidine group, diaminomethyl group, and carbamate group made no sense in the improvement of the potency of compound 46 (Fig. 12A). The most striking difference was that compound 46 was the first small molecule Src inhibitor at subnanomolar concentrations, showing a 1000-fold selectivity against ABL. In Src-sensitive triple-negative breast cancer cell lines (MCF7 and MDA-MB-231), compound 46 induced a very potent antiproliferative effect. Phenotypic screening in zebrafish also showed that compound 46 significantly reduced neuromast migration (>100 μM in average) with a minimal effect on the development of embryos, which suggested that compound 46 was more efficacious than dasatinib (no migration inhibition and induced a patent cardiotoxic phenotype at concentrations of 1–10 μM).⁵⁴

Fig. 12. (A) Binding mode of ECF-506 with Src (PDB: 7NG7). (B) Binding mode of compound 48 with Src (PDB: 3EL7).

Because of the high lethality of TNBC and the lack of a corresponding efficacious drug, Zhang et al. reported the utilization of the 1H-pyrazolo[3,4-d]pyrimidin-4-amine scaffold in designing Src type-II inhibitors based on the structure of ponatinib.⁵⁵ Compound 47 (Fig. 13) was obtained via structure–activity relationship analysis with an excellent Src IC₅₀ value (0.9 nM) as well as perfect anti-cell viability in MDA-MB-231 cell lines (11 nM). Similar to ponatinib, compound 47 was a multikinase inhibitor which showed high potencies against many kinases, especially the Src family kinases (Yes IC₅₀ = 0.8 nM; Fyn IC₅₀ = 5 nM; Blk IC₅₀ = 19 nM). In particular, compound 47 exhibited a strong antiproliferative activity in TNBC high expression cell lines (MDA-MB-231 IC₅₀ = 11 nM, MDA-MB-435 IC₅₀ = 0.5 nM, MDA-MB-436 IC₅₀ = 316 nM, and MDA-MB-453 IC₅₀ = 325 nM) and good pharmacokinetic properties (AUC_0−∞ = 4772.5 μg h mL⁻¹, T_1/2 = 12.76 h, C_max = 214.63 μg L⁻¹, CL = 2.1 L h⁻¹ kg⁻¹, V_ss = 38.75 L kg⁻¹) after a per oral administration of 10 mg kg⁻¹. In an MDA-MB-231 xenograft model, compound 47 could completely prevent tumor growth (tumor inhibitory rate >100%). However, it caused multiple organ toxicities in preliminary acute toxicity tests following a single dose of 100 mg kg⁻¹ in rats. In addition, a lower selectivity of the hERG channel was a risk for cardiotoxicity. Hence, starting from these two points, the group continued to optimize compound 47; aiming to reduce toxicity, remove hERG activity, and improve potency, they finally obtained compound 48 (Fig. 13) as a multikinase inhibitor through a sophisticated process (Src IC₅₀ = 3 nM, Yes IC₅₀ = 1 nM, Lyn IC₅₀ = 4 nM, Fyn IC₅₀ = 6 nM, and KDR IC₅₀ = 32 nM) (Fig. 12B).⁵⁶ Not only did compound 48 exhibit the same potencies against high expression TNBC cell lines as compound 46 but it also inhibited tumor growth with an inhibition rate of 96.2% at 40 mg kg per day. Importantly, compound 48 showed no obvious hERG toxicity and low acute toxicity at a single oral dose of 200 mg kg⁻¹, which fulfilled the purpose of this project. In addition, compound 48 also showed good pharmacokinetic properties (AUC_0−∞ = 2861 μg h mL⁻¹, T_1/2 = 18.9 h, C_max = 143.0 μg L⁻¹). It completely suppressed tumor growth in MDA-MB-231 and MDA-MB-435 xenograft models at a dose of 40 mg per kg q.d.

Gushwa et al. reported their discovery of compounds 50 and 51 (Fig. 13) based on the Src kinase family inhibitor PP1 which could selectivity target Lys295 and Cys277 of the Src kinase ATP pocket in an irreversible way. Enzymatic assays revealed that compounds 50 and 51 had an inhibitory effect on Yes (2.3 nM/1.3 nM), Lyn (8.9 nM/89 nM), ABL (35 nM/182 nM), FGFR3 (261 nM/1.6 nM), and VEGFR2 (87 nM/475 nM), which directly confirmed the permissive conditions for these kinases when designing inhibitors in this way. Importantly, the alkyne tag of compounds 50 and 51 could efficiently label the respective endogenous kinase targets in intact cells.⁵⁷

2.7. PI3K inhibitors

Phosphatidylinositol 3-kinases (PI3Ks) belong to the family of intra-cellular signal transducer enzymes. PI3K has serine/threonine (Ser/Thr) kinase as well as phosphatidylinositol kinase activity. PI3K regulates critical cellular processes including cell growth, proliferation, differentiation, motility, and intracellular trafficking. Deregulation of the phosphoinositide 3-kinase pathway has been identified to be closely associated with cancer, diabetes, thrombosis, and rheumatoid arthritis.⁵⁸ Many 4-aminopyrazolopyrimidine derivatives have been developed as PI3Kδ inhibitors.

Combined with kinase-directed screening, Bristol-Myers Squibb reported the first pyrrolo[2,1-f][1,2,4]triazin-4-amine scaffold, compound 52 (PI3Kα IC₅₀ = 88 nM, PI3Kβ IC₅₀ = 2640 nM, PI3Kγ IC₅₀ = 8.8 nM, PI3Kδ IC₅₀ = 22 nM), as a potent PI3K inhibitor with low selectivity (Fig. 14).⁵⁹ Further evaluation of compound 52 indicated that the 4-pyridyl moiety contributed to its poor selectivity and potent CYP inhibition, while the cyclohexyl moiety led to its poor microsomal stability. Therefore, compound 53 (PI3Kα IC₅₀ = 1330 nM, PI3Kβ IC₅₀ = 1600 nM, PI3Kγ IC₅₀ = 260 nM, and PI3Kδ IC₅₀ = 2 nM) was obtained after three rounds of SAR analyses. The m-phenylbenzylic amines could afford good selectivity to other isoforms (Fig. 16). A mouse PK study revealed the high clearance (82.1 mL min⁻¹ kg⁻¹), high volume of distribution (6.2 L kg⁻¹), good oral bioavailability (46%), and short half-life (1 h) of compound 53.

Fig. 14. Reported 4-aminopyrazolopyrimidine derivatives as PI3K inhibitors.

Fig. 16. Binding mode of compound 53 with PI3Kδ (PDB: 2WXN).

Afterwards, Qin et al. reported compound 54 (PI3Kα IC₅₀ = 795 nM, PI3Kβ IC₅₀ = 1876 nM, PI3Kγ IC₅₀ = 57 nM, and PI3Kδ IC₅₀ = 1.3 nM) as a selective PI3Kδ inhibitor aimed at ameliorating acute toxicity shown in exploratory toxicity studies in mice (300 mg kg per day).⁶⁰

Subsequently, they reported compound 55 as a highly potent PI3Kδ inhibitor with an >1000 fold selectivity against other forms (PI3Kα IC₅₀ > 200 nM, PI3Kβ IC₅₀ > 200 nM, PI3Kγ IC₅₀ > 200 nM, and PI3Kδ IC₅₀ = 0.2 nM). They exploited the potentialities of the trifluoromethyl group to improve the selectivity and activity. What was unfortunate was that compound 55 exhibited poor stability toward liver microsomes where it was determined that the morpholine ring was extensively metabolized and no further progress was made.⁶¹ Due to the poor potency, PK, and cardio safety of compound 55, Liu et al. developed the systematic SAR analysis of reported PI3K inhibitors and synthesized compound 56 (PI3Kα IC₅₀ = 330 nM, PI3Kγ IC₅₀ = 110 nM, and PI3Kδ IC₅₀ = 3 nM), but it showed high cardiovascular liabilities and lower blood pressure, which restricted the clinical development of this compound. In view of this, they re-examined the X-ray cocrystal structure of compound 56 and PI3K kinase and they chose to remove the pyrazole ring (compound 57, PI3Kγ IC₅₀ = 648 nM, PI3Kδ IC₅₀ = 2.4 nM). Although it was selective for other subtypes and showed less cardiovascular liabilities, it also showed low microsomal stability, which interfered with its studies in rodents. To sterically block cytochrome P450s from accessing the metabolic soft spots, small groups were introduced into the phenyl ring, and eventually compound 58 (PI3Kα IC₅₀ = 1200 nM, PI3Kβ IC₅₀ > 5000 nM, PI3Kγ IC₅₀ = 170 nM, and PI3Kδ IC₅₀ = 1.7 nM) with high selectivity, cardio safety, and a good PK value was reported (rat IV 2 mg kg⁻¹ CL = 2.3 mL min⁻¹ kg⁻¹; V_ss = 0.5 L kg⁻¹; F = 71%).⁶² In a mouse collagen-induced arthritis model (2 and 5 mg kg⁻¹ for 42 days), greater than 50% suppression of paw swelling was observed with dose-dependent reduction, even with lower than expected exposures (Fig. 14).

Besides that, many inhibitors which took pyrrolo[2,1-f][1,2,4]triazin-4-amines as the scaffold had been reported in patent WO2016064957 (such as compound 59, Fig. 14).⁶³

In 2001, ICOS reported the first isoform-selective PI3Kδ inhibitor, IC-87114 (compound 60), which showed an IC₅₀ value of 130 nM against PI3Kδ and a >100-fold selectivity against PI3Kα/β/γ (Fig. 15).⁶⁴

Fig. 15. Reported 4-aminopyrazolopyrimidine derivatives as PI3K inhibitors.

Yue et al. identified compound 61 (parsaclisib, PI3Kα IC₅₀ > 20 000 nM, PI3Kβ IC₅₀ > 20 000 nM, PI3Kγ IC₅₀ > 20 000 nM, and PI3Kδ IC₅₀ = 1.1 nM) as a novel PI3Kδ inhibitor via the structure-based design of dezapelisib (Fig. 15). Compound 61 was a highly selective PI3Kδ inhibitor with drug-like ADME properties (low systemic clearance: 26% in rats, 2% in dogs, and 5% in monkeys; moderate steady-state volume of distribution (1.54 L kg⁻¹ in rats and 0.72 L kg⁻¹ in monkeys), high oral exposure, and complete bioavailability (100% in dogs, 79% in monkeys, and 74% in rats)). Oral administration twice daily of compound 61 in a Pfeiffer DLBCL subcutaneous mouse xenograft model showed that it significantly inhibited tumor growth at 1 mg kg⁻¹ which was identified to be related to the profound inhibition of the phosphorylation of AKT at Ser473. In addition, preclinical 28 day IND-enabling toxicology studies in rats and dogs showed no liver toxicities, suggesting that compound 61 was suitable for advancement into clinical trials.⁶⁵

TGR-1202 (umbralisib, compound 62, Fig. 15) is a selective PI3Kδ inhibitor with K_d values of >10 000 nM/>10 000 nM/1400 nM/6.2 nM against PI3Kα/β/γ/δ. Burris et al. reported the first-in-human phase I study of compound 62 to establish its safety and preliminary activity profile in patients with hematological malignancies. The result of this research showed that compound 62 showed noteworthy single-agent activity (85% of evaluable patients with CLL and 53% of those with follicular lymphoma showed an objective response).⁶⁶

Combining the structures of CAL-101 and TGR-1202, Li et al. reported the pyrazolo[3,4-d]pyrimidine compound 63 (PI3Kα IC₅₀ = 1163 nM, PI3Kβ IC₅₀ = 784 nM, PI3Kγ IC₅₀ = 1134 nM, and PI3Kδ IC₅₀ = 1.7 nM) as a potent and selective PI3Kδ inhibitor (Fig. 15). Compound 63 showed poor PK by oral administration but high exposure (5599 ng g⁻¹), high plasma clearance (348.5 mL min⁻¹ kg⁻¹), and low exposure in blood (66 ng mL⁻¹) were retained after the inhalation route in rats, which was an advantage for the treatment of lung disease due to the side-effect caused by oral administration of PI3Kδ inhibitors. Further evaluation in cigarette-smoke- and LPS-induced rodent model mimics showed that compound 63 could improve lung function and reduce the inflammatory pattern characteristics of COPD.⁶⁷

Additionally, patents WO2015069441 (ref. 68) and WO2016058084 (ref. 69) (Fig. 15) reported a series of 1H-pyrazolo[3,4-d]pyrimidin-4-amine PI3Kδ selective inhibitors such as compounds 64 (10 nM), 65 (9.7 nM), 66 (0.27 nM), and 67 (1 nM).

2.8. mTOR inhibitors

The mammalian target of rapamycin (mTOR) is a highly conserved eukaryotic serine/threonine (Ser/Thr) protein kinase that belongs to the phosphatidylinositol 3-kinase-related kinase (PI3K) family. mTOR was known as the central controller of cell growth due to its important role in controlling transcription, translation, and ribosomal biosynthesis. Protein synthesis is crucial to cell division, which provides us an idea on designing inhibitors via inhibiting the mTOR signaling pathway, thereby inducing cell cycle arrest at the G1 phase.⁷⁰

Shokat et al. reported the application of 4-aminopyrazolopyrimidine derivatives in the design of mTOR inhibitors. Torkinib (compound 68), PP30 (compound 69), PP-487 (compound 70), and PP-121 (compound 71) were approved as mTOR inhibitors with IC₅₀ values of 8 nM, 80 nM, 72 nM, and 10 nM, respectively (Fig. 18). Among them, PP232 and PP30 exhibited high selectivity for mTOR, while PP487 and PP121 had high potency against mTOR, PI3K, ABL, Src, and some other tyrosine kinases at nanomolar concentrations.^71,72

Fig. 18. Reported 4-aminopyrazolopyrimidine derivatives as mTOR inhibitors.

Millennium Pharmaceuticals approved the other 4-aminopyrazolo[3,4-d]pyrimidine scaffold (Fig. 18) mTOR/PI3K dual inhibitor 72 (sapanisertib, mTOR IC₅₀ = 1 nM, PI3Kα IC₅₀ = 219 nM, PI3Kγ IC₅₀ = 221 nM, and PI3Kδ IC₅₀ = 230 nM) (Fig. 17). Sapanisertib inhibited mTOR and PI3K as well as mTOR downstream effectors S6, 4E-BP1, and AKT. A phase I study of compound 72 in patients with advanced solid malignancies showed that there was no significant difference in the pharmacokinetics of compound 72 when administered with or 24 h after administration of paclitaxel 67, which indicated the potential of compound 72 in drug combination.⁷³

Fig. 17. Binding mode of compound 83 with mTOR (PDB: 6VGF).

Bhagwat et al. reported the development of an orally available dual mTOR inhibitor 73 (OSI-027, Fig. 18) which took imidazo[5,1-f][1,2,4]triazin-4-amine as the scaffold with a >100 fold selectivity against PI3K. OSI-027 showed concentration-dependent pharmacodynamic effects on the phosphorylation of 4E-BP1 and AKT in tumor tissue, resulting in tumor growth inhibition. In an MDA-MB-231 breast cancer xenograft model, compound 73 showed 100% tumor growth inhibition (TGI) at 24 hour post-dose in the 65 mg kg⁻¹ dose group and 90% TGI on a twice daily schedule of 25 mg kg⁻¹ for the 14 day group. In addition, the effects of OSI-027 were further evaluated in human colorectal xenograft models (GEO, COLO 205) and an mTOR pathway dependent PIK3CA mutant SKOV-3 and KRAS mutant OVCAR-5 xenograft model; OSI-027 exhibited robust anti-tumor activity.^74,75

2.9. The application of 4-aminopyrazolopyrimidine derivatives in PROTACs

Proteolysis-targeting chimera technology is a technology that utilizes small molecular compounds to target protein degradation at the protein level through the rational design of the chain link ligands that bind the target protein and ligands that bind the E3 ubiquitin ligase.⁷⁶ This new bifunctional small molecule can make the distance between the target protein and the E3 ubiquitin ligase close and ubiquitinate the target protein and then degrade it through the ubiquitin–proteasome pathway. The new pharmacological model supports a new research direction for disease therapy, especially proteins that are difficult to target and be acted upon.⁷⁷ Up to now, some 4-aminopyrazolopyrimidine derivative inhibitors have been applied in designing PROTACs to target specific proteins.

Unlike traditional drugs, PROTACs are designed to eliminate dysfunctional proteins, not inhibit their activity. Therefore, the acquired resistance caused by the C481S BTK mutant could be overcome by using PROTAC molecules. Due to the C481S BTK mutation which mainly caused ibrutinib resistance, in 2018, Sun et al. reported the first BTK-targeting degrader using the PROTAC strategy.⁷⁸ These new PROTACs could efficiently degrade ibrutinib-sensitive BTK^WT. Compound 74 (Fig. 19) could induce 73% degradation of BTK at 10 nM and 89% at 100 nM in human Burkitt's lymphoma RAMOS cells. In the BTK^C481S DCBCL cell line used as a model system for addressing ibrutinib resistance, ibrutinib showed no efficacy, and the proliferation of the DLBCL cell line (HBL cells) was significantly inhibited (GI₅₀ = 28 nM). Notably, compound 84 showed no degradation effect against ITK and EGFR, which indicated that the PROTACs could prevent the side effects of ibrutinib.

Fig. 19. Reported 4-aminopyrazolopyrimidine derivatives, PROTACs.

Then, Buhimschi et al. reported another BTK PROTAC, compound 86 (MT802), targeting C481S ibrutinib-resistance mutation (Fig. 19). Compound 75 induced 99% degradation of BTK^WT and BTK^C481S. Due to the poor pharmacokinetic properties of compound 75 (CL = 1662 mL min⁻¹ kg⁻¹, T_1/2 = 0.119 h, C_max = 0.073 μg mL⁻¹, AUC_last = 10.2 ng min mL⁻¹), they further decorated the structure of the linker and E3-recruiting ligand of compound 75 and eventually they obtained the target compound 76 with good pharmacokinetic properties (CL = 102 mL min⁻¹ kg⁻¹, T_1/2 = 1.62 h, C_max = 0.831 μg mL⁻¹, AUC_last = 166 ng min mL⁻¹). Additionally, compound 76 still retained 91% degradation of BTK^WT and BTK^C481S. PROTACs of other kinases with 4-aminopyrazolopyrimidine as the scaffold were less reported. On account of the application of 4-aminopyrazolopyrimidine in PROTACs of BTK, it is supposed that more PROTACs of other targets will be researched and play important roles in medicinal chemistry and disease therapy.^79,80

3. Conclusion

In this review, we systematically introduced the application of the 4-aminopyrazolopyrimidine derivative scaffold and its deformation in designing small-molecule inhibitors of tyrosine kinases and serine/threonine kinases as well as PROTACs, such as FGFR, BTK, EGFR, FLT3, mTOR, and so on. We found that the 4-aminopyrazolopyrimidine derivative scaffold formed an essential H-bond with the amino acid residue in the hinge region, which was an advantage for the compound to be effective. For different targets, great efforts had been made to analyze which substituent group was more effective for improving the kinase activity, selectivity, ADME properties, and cellular potency against subtypes or other kinases. 4-Aminopyrazolopyrimidine derivatives as inhibitors had been confirmed to be effective in the treatment of diseases. In this article, we summarized the different changes in the core of 4-aminopyrazolopyrimidine which were thought to facilitate the development of kinase inhibitors. We believe that the 4-aminopyrazolopyrimidine derivative scaffold and its deformation will be increasingly involved in more and more ideas of designing kinase inhibitors. This article provides useful and practical information for medicinal chemists for the development of 4-aminopyrazolopyrimidine derivatives and its deformation as kinase inhibitors and disease therapy.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.