Development of selective FGFR1 degraders using a Rapid synthesis of proteolysis targeting Chimera (Rapid-TAC) platform
Graphical abstract
Introduction
Fibroblast growth factor receptors (FGFRs) belong to the class of receptor tyrosine kinases (RTKs) and there are four known members FGFR1-4, which all contain an extracellular domain, a transmembrane domain, and an intracellular tyrosine kinase domain1. Members of FGFRs share high sequence homology (56 % to 71 %) and their active sites are highly conserved2. Fibroblast growth factors (FGFs) can bind to the extracellular domain of FGFRs and activate the FGF/FGFR signaling pathway. The subsequent receptor dimerization then induces phosphorylation and triggers the downstream signaling pathways3, 4 associated with cell proliferation, differentiation, and migration.5, 6, 7, 8, 9, 10 The dysregulation and overexpression of FGFR signaling in many types of cancer has been well documented.9 The aberrations of FGFRs have been detected in 5–10 % of all human cancers and as high as 10–30 % in certain types of cancers, such as urothelial cancer and intrahepatic cholangiocarcinoma (iCCA)11. FGFR1 is frequently amplified in squamous non-small cell lung cancer (20–25 %) and breast cancer (15 %).12, 13 Mutated FGFR1 were also found in 18 % of midline gliomas.14 FGFR2 is mainly activated by gene fusion in iCCA (15 %), and mutations of FGFR2 are also found in 10 % of endometrial tumors.15, 16 Mutations of FGFR3 are frequently found in urothelial carcinomas (∼20 %)17, 18; FGFR3 gene fusions, mainly FGFR3-TACC3, are present in 3 % of glioblastomas and gliomas and 2–3 % of bladder cancer.19, 20.
Not surprisingly, significant amount of efforts have been devoted to the development of small molecule FGFR inhibitors for the treatment of various types of cancers. First-generation FGFR-tyrosine kinase inhibitors (TKIs) (e.g., anoltonib, ponatinib, dovitinib, lucitanib, lenvatinib, and nintedanib) are relatively promiscuous by inhibiting FGFRs and other kinases (VEGFR1/3, KIT, RET among others), which then lead to a variety of undesired adverse effects such as hypertension and bleeding21. To overcome these drawbacks, several second-generation FGFR inhibitors have been developed. They possess high potency and selectivity as pan-FGFR TKIs such as BGJ398 (Infigratinib), JNJ42756493 (Erdafitinib), AZD4547 and CH5183284 (Debio 1347)21. Two FGFR inhibitors, erdafitinib and pemigatinib, have recently been approved by the Food and Drug Administration (FDA) for the treatment of urothelial cancer and cholangiocarcinoma, respectively.22, 23 These inhibitors are selective for FGFRs, but not selective for members of FGFRs, have better toxicity profile than the first-generation multi-target kinase inhibitors. However, it is still challenging to develop isoform-selective FGFR inhibitors due to the high sequence homology among the FGFR members, especially around the active site.
Proteolytic targeting chimeras (PROTACs) are heterobifunctional degraders composed of a targeting ligand for a protein of interest, a linker, and a ligand of E3 ubiquitin ligase. Upon the binding of the target protein and E3 ligase to the degrader, the target protein will be ubiquitinated and degraded by the cell’s native proteasomal degradation machinery.24, 25, 26 PROTACs have many potential advantages. One of them is the possibility for the development of isoform-selective degraders even when the ligand can bind to multiple isoforms because the degradation efficiency is often determined by the formation of the ternary complex.27, 28, 29, 30 In addition, PROTACs can be highly efficient and their effect can last longer because its mode of action is a catalytic rather than occupying a pocket and the re-synthesis of the target protein may take time. At present, PROTACs have been developed for many different families of proteins. It is projected that a dozen of degraders will be progressed into human clinical trials soon.31.
One of the major challenges for the development of PROTACs is the lengthy synthesis and screening of the linker length, linker type, linker position, and ligands of the target and the E3 ligase. Recently, our group developed a novel platform to optimize the linker length/types by using aldehyde-hydrazide coupling chemistry.32 The reaction between aldehyde and hydrazide can proceed in DMSO solution with high conversion and produce water as the only byproduct. It allows us to test the activity of the resulting PROTACs in DMSO solution directly in cell-based assays without further purification32. By using this platform of rapid synthesis of proteolysis targeting chimeras (Rapid-TAC), we quickly prepared a library of potential degraders for estrogen receptor (ER) by simply mixing a hydrazide-containing ligand of ER with a library of E3 ligase ligands bearing various linkers and an aldehyde functional group.32 To further demonstrate the utility of our platform for other therapeutically relevant targets, we herein report the development of highly potent and selective degraders for FGFR1, a challenging membrane protein target, based on a pan FGFR inhibitor AZD4547. The degradation mechanism and anti-proliferation activity of these degraders are also investigated. During our investigation, FGFR1/2 selective degraders were reported.33.
AZD4547 (Fig. 1a) is a pan inhibitor of FGFRs and being investigated in several clinical trials34. However, AZD4547 inhibits all members of FGFR and resistance has occurred in some lung cancer.35, 36 We decide to develop a library of FGFR degraders using AZD4547 as the binder and investigate the degradation activity and selectivity against members of FGFRs. The co-crystal structure (Fig. 1b) indicates that the piperazine group is exposed to the solvent and we hypothesize that we can attach a linker to the nitrogen atom of this group for the development of PROTACs.37.
A hydrazide group can be introduced to the solvent-exposing region of AZD4547 to form hydrazide-1 (Fig. 1c). This building block can react with our previously reported library of E3 ligase ligands bearing various linkers and a terminal aldehyde functional group under miniaturized conditions, which produces 100 μL of 10 mM DMSO solution for each of the resulting acylhydrazone product.32 Using our Rapid-TAC platform, a library of 24 FGFR PROTACs was prepared in 1–2 days by conjugating one hydrazide-containing building block hydrazide-1 to our previously reported von Hippel–Lindau (VHL) and cereblon (CRBN) ligand library (Fig. 1d).
The PROTAC-induced degradation of FGFRs was evaluated in MCF-7 cells (Fig. 2). Because membrane proteins including FGFR often have several different glycosylated forms, multiple bands on the Western blot are typically observed. MCF-7 cells were treated with each of the 24 compounds at 100 nM for 6 h. In the VHL alkyl linker series, compounds 1 and 3 (n = 1 and 4, respectively) are the most potent compounds to induce FGFR1 degradation in MCF-7 cells. In the VHL PEG linker series, compound 9 (n = 3) and 10 (n = 4) induce the most degradation of FGFR1 in MCF-7 cells (Fig. 2a). For CRBN series, compounds 14 and 21 have the optimal linker length for FGFR1 degradation in alkyl and PEG linker series, respectively (Fig. 2b). We then selected three representative compounds from the CRBN-recruiting FGFR PROTACs with a wide range of linker length to test their degradation activity across all four FGFRs. Compounds 13, 15 and 17 all degrade FGFR1 selectively without changing much of the level of FGFR2, 3 and 4 (Fig. 2c). Since AZD4547 has high affinity to FGFR1, 2 and 3 with Kd at 0.2, 2.5 and 1.8 nM, respectively,38 it is not surprising to see the selectivity for the degradation of FGFR1 over others.
Multiple types of E3 ligands and linkers were found for FGFR1 PROTACs from the initial screening of our library of 24 compounds. Since acylhydrazone is known to be hydrolytic labile, we then prepared several stable FGFR PROTACs by simply replacing the acylhydrazone motif by its more stable amide isostere (Fig. 3a). For VHL recruiting PROTACs, alkyl linkers ranging from 1 to 5 units of methylene group were synthesized because compounds 1 and 4 have comparable activity. The FGFR1 degradation activity was then evaluated by Western Blot (Fig. 3b). As shown in Figure 3b, LG1188 (n = 1) is the most potent degrader for FGFR1 in MCF-7 cells. Based on our initial results from the screening of the library of 24 compounds, PROTAC 14 with 2 PEG linker is the most potent degrader in CRBN series, we also synthesized its stable analogue 30 for comparison. The degradation was evaluated by Western Blot and we found that the potency of analogue 30 was comparable to parent compound 14 (Fig. 3c). We then further verified the selective degradation of FGFR1 over FGFR2 and FGFR3 for compound LG1188 and four related stable analogues including 25 with a very short linker between AZD4547 and VHL ligand (Fig. 3d).
FGFR1 amplification is one of the most frequent events in the FGFR oncogenic alteration. FGFR1 amplification is found in approximately 17 % of squamous nonsmall cell lung cancers and approximately 6 % of smallcell lung cancers.1 To further validate the anti-tumor effect in FGFR1-driven cancers, we characterize the FGFR1 PROTAC in two FGFR1 amplification lung cancer cell lines, DMS114 (small-cell lung cancer) and NCI-H1581 (non-small cell lung cancer). The FGFR1 degradation of compounds LG1188 and 30 was first evaluated in these two cell lines. FGFR1 was potently degraded by LG1188 in DMS114 cells within 6 h, while the CRBN-recruiting PROTAC 30 is less potent (Fig. 4a). In the case of NCI-H1581 cells, LG1188 can induce obvious FGFR1 degradation, though it is less effective as compared to DMS114 cells (Fig. 4b). Similarly, compound 30 is also less potent in this cell line. We next focused our studies on LG1188 in DMS114 based on these results.
In DMS114 cells, a dose-dependent FGFR1 degradation was observed for compound LG1188. FGFR1 can be depleted completely in cells treated with 100 nM of LG1188 for 6 h (Fig. 4c). The degradation of FGFR1 occurred as early as 0.5 h in DMS114 cells exposed to 100 nM of LG1188 and quickly reached the maximum effect after 2 h (Fig. 4d).
To verify the engagement of FGFR1, VHL E3 ubiquitin ligase and the proteasome in the degradation of FGFR1, DMS114 cells were pre-incubated with four pathway inhibitors for 1 h followed by the co-treatment with LG1188 (100 nM) for 5 h. As illustrated in Fig. 5a, the FGFR inhibitor AZD4547 which is used as the FGFR PROTAC warhead can partially block the FGFR1 degradation induced by LG1188 in cells co-treated with 10 µM AZD4547. It is likely that the FGFR1 level is not fully recovered by the addition of AZD4547 because 10 µM AZD4547 is not sufficient to block the ternary complex formation. The degradation of FGFR1 was abolished by the addition of 10 µM compound 1198, a ligand of VHL E3 ligase. This result indicates the involvement of VHL E3 ligase in the LG1188 mediated degradation of FGFR1. Additionally, FGFR1 degradation was blocked completely by the addition of MG132 and MLN4924 as shown in Fig. 5a, indicating that LG1188 induced FGFR1 degradation is dependent on the ubiquitin–proteasome system and NEDD8-activating enzyme (NAE). Taken together, these results suggest that LG1188 induced FGFR1 degradation is mediated by VHL E3 ligase engagement and executed by proteasome-ubiquitin system.
The long-term effect was investigated by a wash-out assay. DMS114 cells were treated with 100 nM LG1188 for 12 h. Cells were then washed with PBS and medium without compounds was added. Cells were collected at 12 h, 24 h, 36 h and 48 h post-treatment for Western Blot assay. The potency of a PROTAC is highly dependent on the turnover rate of the protein of interest. The re-synthesis rate of FGFR1 is fast in DMS114 cells due to FGFR1 amplification. FGFR1 level started to recover at 24 h post-treatment and increased over time (Fig. 5b). However, FGFR1 level did not recover completely even after 48 h, indicating that the compound left in the cells can still counteract the effect of FGFR1 re-synthesis.
AZD4547 binds to the tyrosine kinase domain and then prevents the activation of downstream signaling of FGFR stimulated by FGFs.4 We compared the downstream signaling inhibition effect mediated by inhibitor AZD4547 and PROTAC LG1188 in DMS114 cells (Fig. 6). Cells were pre-treated with AZD4547 or LG1188 at the indicated concentrations for 4 h. Upon the stimulation by 50 ng/mL of FGF2 for 30 min, an increased phosphorylation level of FGFR, FRS2α (fibroblast growth factor receptor substrate 2) and, p42/44 (ERK, Extracellular signal-regulated kinase) was detected by Western Blot. FRS2α and FGFR activation transmits the signal to the downstream pathways, such as MAPK/ERK pathways39, 40. ERK is involved in the MAPK/ERK pathways upon the FGFR activation.41, 42 Their phosphorylation was blocked by FGFR inhibitor AZD4547 as shown in Figure 6, indicating the inhibition of FGFR activation. FGFR PROTAC LG1188 degraded FGFR1 and also led to the reduction of phosphorylated downstream effectors.
The two FGFR1 amplified cell lines, DMS114 and NCI-H1581, displayed profound sensitivity to FGFR inhibitor AZD4547 [43]. To confirm whether FGFR1 PROTAC also possesses the anti-proliferation effect, cell viability was evaluated in DMS114 using MTT assay. Cells were treated with indicated compounds at concentrations ranging from 10 µM to 40 nM for 120 h. As demonstrated in Fig. 7, both DMS114 and NCI-H1581 cells were responsive to AZD4547 effectively and the IC50 of AZD4547 in DMS114 and NCI-H1581 is 14 nM and 4 nM, respectively. With regard to the response to LG1188, both cell lines showed an obvious growth inhibition effect with IC50s of 139 nM and 160 nM in DMS114 and NCI-H1581, respectively. To summarize, both FGFR inhibitor and FGFR1 PROTAC are able to inhibit the growth of FGFR1 amplified tumor cells. The more potent anti-proliferation effect of FGFR inhibitor than FGFR1 PROTAC may be attributed to better cell permeability and stability of AZD4547, which has been extensively optimized. The efficacy of FGFR1 degrader can be improved with further optimization.
In conclusion, we are able to prepare 24 potential PROTACs quickly from a hydrazide-containing FGFR inhibitor and our previously established VHL and CRBN ligand library bearing various linkers and an aldehyde functional group. These 24 PROTACs were then directly used for screening in cellular assay for protein degradation. Multiple hits were identified from the initial screening. Based on these results, we then prepared the corresponding stable analogues by replacing the hydrolytic labile acylhydrazone motif with an amide. Among them, PROTAC LG1188 was identified as a potent and selective FGFR1 degrader, though the corresponding warhead AZD4547 inhibits FGFR1/2/3 with high potency. LG1188 can be used as a chemical probe to study the biological functions, especially non-enzymatic functions of FGFR1. We also found that the VHL recruiting FGFR1 PROTACs are more potent than the CRBN recruiting FGFR1 PROTACs from our library in the cell lines we examined. Our mechanistic investigations confirmed the engagement of the E3 ubiquitin ligase, FGFR, and ubiquitin–proteasome system for LG1188-induced FGFR1 degradation. LG1188 also showed expected anti-proliferation effect and the inhibition of downstream signaling pathway despite that PROTAC is at least twice larger than inhibitor.
Section snippets
Declaration of Competing 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.
Acknowledgements
Support for this research was provided by the University of Wisconsin – Madison Office of the Vice Chancellor for Research and Graduate Education with funding from the Wisconsin Alumni Research Foundation through a UW2020 award. This study made use of the Medicinal Chemistry Center at UW-Madison instrumentation, supported by the Lachman Institute for Pharmaceutical Development, University of Wisconsin School of Pharmacy and the University of Wisconsin Carbone Cancer Center Support Grant NIH P30
References (42)
- et al.
Evolution of the Fgf and Fgfr gene families
Trends Genet
(2004) - et al.
Precision medicine for patients with advanced biliary tract cancers: an effective strategy within the prospective MOSCATO-01 trial
Eur J Cancer
(2017) - et al.
Second-line dovitinib (TKI258) in patients with FGFR2-mutated or FGFR2-non-mutated advanced or metastatic endometrial cancer: a non-randomised, open-label, two-group, two-stage, phase 2 study
Lancet Oncol
(2015) - et al.
Preliminary results of a ph2a study to evaluate the clinical efficacy and safety of erdafitinib in Asian patients with biomarker-selected advanced cholangiocarcinoma (CCA)
Ann Oncol
(2018) - et al.
Discovery and optimization of novel pyrazole-benzimidazole CPL304110, as a potent and selective inhibitor of fibroblast growth factor receptors FGFR (1–3)
Eur J Med Chem
(2021) - et al.
Pemigatinib for previously treated, locally advanced or metastatic cholangiocarcinoma: a multicentre, open-label, phase 2 study
Lancet Oncol
(2020) - et al.
Development of the first small molecule histone deacetylase 6 (HDAC6) degraders
Bioorg Med Chem Lett
(2018) - et al.
An overview of the binding models of FGFR tyrosine kinases in complex with small molecule inhibitors
Eur J Med Chem
(2017) - et al.
Fibroblast growth factor receptor inhibitors as a cancer treatment: from a biologic rationale to medical perspectives
Cancer Discov
(2013) - et al.
FGFR-TKI resistance in cancer: current status and perspectives
J Hematol Oncol
(2021)
Fibroblast growth factor receptors (FGFRs): structures and small molecule inhibitors
Cells
Targeted disruption of fibroblast growth factor (FGF) receptor 2 suggests a role for FGF signaling in pregastrulation mammalian development
Proc Natl Acad Sci USA
Effects of FGFR signaling on cell proliferation and differentiation of Apert dental cells
Cells Tissues Organs.
Fibroblast growth factor signalling: from development to cancer
Nat Rev Cancer
FGFR1 expression and role in migration in low and high grade pediatric gliomas
Front Oncol
Inhibition of the fibroblast growth factor receptor (FGFR) pathway: the current landscape and barriers to clinical application
Oncotarget
The FGF family: biology, pathophysiology and therapy
Nat Rev Drug Discov
Fibroblast growth factor receptors in cancer: genetic alterations, diagnostics, therapeutic targets and mechanisms of resistance
Br J Cancer
Frequent and focal FGFR1 amplification associates with therapeutically tractable FGFR1 dependency in squamous cell lung cancer
Sci Transl Med
Molecular characterization of breast cancer with high-resolution oligonucleotide comparative genomic hybridization array
Clin Cancer Res
Actionable targets involving FGF receptors in gliomas: molecular specificities, spatial distribution, clinical outcome and radiological phenotype
J Clin Oncol
Cited by (6)
Recent progress in degradation of membrane proteins by PROTACs and alternative targeted protein degradation techniques
2023, European Journal of Medicinal ChemistryThe application of PROTAC in HDAC
2023, European Journal of Medicinal ChemistryTargeted degradation of extracellular secreted and membrane proteins
2023, Trends in Pharmacological SciencesPROTACS: A technology with a gold rush-like atmosphere
2023, European Journal of Medicinal ChemistryCitation Excerpt :FGFR1/2 degrader (compound 24, DGY-09-192) was first reported in 2021, consisting of pan-FGFR inhibitors linked to ligands recruiting VHL E3 ligase that are highly selective for the induction of FGFR1/2 degradation (Fig. 4) [84]. In 2022, Tang's group discovered the selective FGFR1 degrader LG1188 (compound 25) (Fig. 4) using the previously developed Rapid-TAC platform [85]. Surprisingly, the research on FGFR degraders conducted by our group has made preliminary progress, and further optimization and biological activity evaluation are underway.
Development of Potent and Selective Coactivator-Associated Arginine Methyltransferase 1 (CARM1) Degraders
2023, Journal of Medicinal Chemistry
- 1
These authors contributed equally to this work.
- 2
Current address: Institute for Pharmacology and Toxicology, Chia Tai-Tianqing Pharmaceutical, Nanjing, Jiangsu, 211100, China.