HJM-561, a Potent, Selective, and Orally Bioavailable EGFR PROTAC that Overcomes Osimertinib-Resistant EGFR Triple Mutations

This article is featured in Highlights of This Issue, p. 1045

Lung cancer is the leading cause of cancer-related death, claiming 1.8 million lives in 2020 worldwide according to GLOBOCAN data. Active mutations in the EGFR gene, the most common genetic driver in non–small cell lung cancer (NSCLC), are found in about 15% of Caucasian and 30% to 40% of Asian patients with advanced NSCLC (1, 2). EGFR tyrosine kinase inhibitors (TKI) have revolutionized the treatment paradigm for patients with NSCLC. The first and second generation EGFR TKIs, such as gefitinib, erlotinib, and afatinib, provide dramatic clinical benefits for patients with NSCLC with sensitizing EGFR mutations, including exon 19 deletion and L858R. Recently, osimertinib, the third generation EGFR TKI, which potently and irreversibly inhibits both the EGFR T790M mutation and TKI-sensitizing mutations, has displayed impressive clinical activity with a progression-free survival of 18.9 months in EGFR-mutated NSCLC (3–5). However, patients treated with osimertinib inevitably developed a new acquired resistance after an average treatment duration of 19 months (4). The most common mechanism for osimertinib on-target resistance is EGFR C797S-containing triple mutations (Del19/T790M/C797S or L858R/T790M/C797S), which account for 10% to 26% of osimertinib-resistant patients (6–8). Cysteine 797 is the key covalent binding site of osimertinib in mutant EGFR, and therefore, the C797S mutation disrupts the covalent bond and nullifies its pharmacologic effect (9). There are currently no approved therapies for patients with NSCLC progressing on osimertinib treatment, which represents an urgent medical need for new therapies that overcome the EGFR C797S-containing triple mutations.

Proteolysis targeting chimeras (PROTAC) that induce targeted protein degradation by hijacking the ubiquitin–proteasome machinery have emerged as a new and promising therapeutic modality. PROTACs are hetero-bifunctional molecules that contain two linked functional domains, with one binding to a protein of interest and another binding to an E3 ligase (10–14). Unlike traditional occupancy-driven small molecule inhibitors, PROTACs act as an event-driven mechanism and are catalytic in nature. These properties give PROTACs many advantages over traditional inhibitors, such as potentially allowing the targeting of proteins that are considered difficult or undruggable with small molecules (e.g., scaffold proteins, and transcription factors), overcoming drug resistance, and reducing drug doses in clinic. PROTACs– ability to overcome drug resistance has been exemplified by Bruton's Tyrosine Kinase, androgen receptor, and BCR-ABL PROTACs (15–18). Several EGFR mutant-targeting PROTACs have been reported, which degrade EGFR L858R/T790M or Del19/T790M double mutants and exhibit antiproliferation activities in cellular assays (19, 20). Recently, Jang and colleagues reported an allosteric EGFR degrader with selective activity against the EGFR L858R/T790M/C797S mutation (21). Given that the C797S mutation preferentially coexists with exon 19 deletion genotype rather than the L858R mutation, PROTACs targeting both Del19/T790M/C797S and L858R/T790M/C797S mutations are of high clinical significance (22). Herein, we report for first time a novel orally bioavailable EGFR PROTAC molecule, HJM-561, that degrades both EGFR Del19/T790M/C797S and L858R/T790M/C797S mutants, but not wild-type (WT) EGFR, in cellular assays in vitro, and is highly effective in inhibiting tumor growth in osimertinib-resistant CDX and PDX mice models in vivo. HJM-561 has the potential to overcome the acquired on-target resistance of osimertinib, and may provide a therapeutic option for treating osimertinib-resistant patients with NSCLC.

NCI-H2228, A431, 293T, SR, and NCI-H1975 cell lines were from the ATCC. MDA-MB-231 was purchased from Dalian Meilun Biotech. PC-9 was obtained from Wuhai Boster Biological Technology. All cell lines were authenticated by the providers using short tandem repeat profiling. Mycoplasma testing was periodically performed using a PCR detection kit from TransGen Biotech. The full-length EGFR, EGFR Del19/T790M/C797S, and EGFR L858R/T790M/C797S sequences were synthesized and inserted into pLVX-NEO vectors. The Lentivirus was generated by cotransfection of EGFR vectors with helper plasmids (”8.9 and VSVG) into 293T cells. Ba/F3 cells stably expressing EGFR and these mutants were generated by infecting Ba/F3 cells with the virus and selecting with 1 ug/mL puromycin. Pomalidomide and MLN4924 were purchased from Selleckchem (23). Osimertinib and brigatinib were purchased from AbMole BioScience. TQB3804 was obtained from Topscience. HJM-561 and HJM-561NC were synthesized as described in the Supplementary Methods.

Cells were plated into 96-well plate at density of 3,000 cells per well with 95 μL culture medium and incubated at 37°C, 5% CO₂, 95% humidity for 24 hours. Five microliters of culture medium with compounds were transferred into the culture plate. The compound concentration was started from 4 μmol/L with a serial 4-fold dilution for 8 doses. The final DMSO concentration was 0.2%. After incubating at 37°C, 5% CO₂, 95% humidity for 72 hours, the cell viability was measured by CellTiter-Glo assay (Promega, #G7573) on Cytation3 Image Reader (BioTek).

Cells were washed with cold PBS and lysed in RIPA buffer (Beyotime, #P0013B) with protease inhibitors (Thermo Fisher Scientific, #78443). The proteins in supernatant were quantified by bicinchoninic acid (BCA; PIERCE, #23227) and mixed with 4 × SDS loading buffer. Boiled samples were loaded into bis-tris polyacrylamide gels (4%–12%; Genscript, #M00654). Target proteins were transferred to nitrocellulose membranes which were probed with total-EGFR antibody (Cell Signaling Technology, #4267L, 1:1000), phospho-EGFR antibody (Cell Signaling Technology, #3777S, 1:1000), β-actin antibody (Cell Signaling Technology, #5125S, 1:10000), or GAPDH antibody (Cell Signaling Technology, #8884S, 1:10000).

HJM-561 was dissolved in a formulation of 5% DMSO, 10% solutol and 85% saline. CD-1 mice (Shanghai Super B&K Laboratory Animal Corp. Ltd) received a single i.v. bolus administration at 1 mg/kg (n = 3), or oral administration at 10 mg/kg (n = 3). Plasma samples were collected at 0.083 hours, 0.25 hours, 0.5 hours, 1 hours, 2 hours, 4 hours, 8 hour, 12 hours, 24 hours after drug dosing. HJM-561 concentration was determined by LC-MS, and pharmacokinetic (PK) parameters were generated using Phoenix WinNonlin.

All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Shanghai Model Organisms Center. For cell line–derived xenograft (CDX) model, 0.5 million EGFR Del19/T790M/C797S Ba/F3 cells with 50% Matrigel were inoculated subcutaneously into 6 to 8 weeks old female Balb/c nude mice (Vital River Laboratories). When the average tumor size reached about 100 mm³, mice were randomized into four treatment groups: vehicle, 10 mg/kg osimertinib, 20 mg/kg HJM-561, and 40 mg/kg HJM-561. Tumor volume and body weight were measured twice weekly. The efficacy study in osimertinib-resistant EGFR Del19/T790M/C797S patient-derived xenograft (PDX) model LUPF104 was performed at Shanghai Lide Biotech. Briefly, 30 to 50 mg tumor pieces were implanted subcutaneously on the right flank of female Nu/Nu mice. HJM-561 was administered orally once daily when the mean tumor size reached approximately 150 mm³. TQB3804 was used as a reference for this model. In these studies, tumor growth inhibition (TGI) was calculated using the formula: TGI (%) = [1-(Ti-T0)/ (Vi-V0)] ×100 where Ti was the average tumor volume of treatment group at the end of treatment, T0 was the average tumor volume of treatment group at the first day of treatment, Vi was the average tumor volume of vehicle group at the end of treatment, and V0 was the average tumor volume of vehicle group at the first day of treatment.

All data generated in this study are included in this article and its Supplementary Data files.

To identify novel PROTACs targeting EGFR Del19/T790M/C797S and L858R/T790M/C797S triple mutants, we linked brigatinib, an approved anaplastic lymphoma kinase inhibitor with moderate inhibition of the triple-mutant EGFR in vitro with a CRBN ligand lenalidomide (24). Optimization of the linker length, composition, attachment sites, and warhead modification led to the identification of our lead PROTAC molecule HJM-561 (Fig. 1A). The EGFR degradation effect was assessed in Ba/F3 cells expressing the triple mutant or wild-type EGFR by Western blot. As shown in Fig. 1B, HJM-561 potently degraded EGFR Del19/T790M/C797S and L858R/T790M/C797S with a DC₅₀ (50% degradation) value of 9.2 nmol/L and 5.8 nmol/L, respectively, but not WT EGFR. The parental inhibitor brigatinib and the negative control HJM-561NC by methylation of glutarimide ring could not degrade the EGFR triple mutants. Additionally, HJM-561 did not degrade ALK, IGF1R, FLT3, and WT EGFR in NCI-H2228 cells although these proteins were known to bind to brigatinib (Supplementary Fig. S1). These results collectively demonstrate that HJM-561 potently and selectively degrades EGFR triple mutants Del19/T790M/C797S and L858R/T790M/C797S while sparing WT EGFR and brigatinib-related targets. When Del19/T790M/C797S Ba/F3 cells were cotreated with an ubiquitination inhibitor MLN4924 (1 μmol/L), EGFR degradation by HJM-561 was totally rescued (Fig. 1C). In a competitive assay with the cereblon (CRBN) ligand, pomalidomide (10 μmol/L) prevented EGFR degradation in the presence of HJM-561 (Fig. 1C). The results demonstrate that HJM-561–induced EGFR triple mutants degradation is mediated by activating CRBN E3 ligase and the ubiquitin-proteasome system.

To explore the therapeutic potential of HJM-561 to overcome osimertinib resistance, we examined its antiproliferation activity in Ba/F3 cells harboring EGFR C797S triple mutants or WT EGFR. HJM-561 potently inhibited the proliferation of EGFR triple mutants expressing Ba/F3 cells, with an IC₅₀ of 15.6 and 17.0 nmol/L in Del19/T790M/C797S cells and L858R/T790M/C797S cells, respectively, which were 19- to 60-fold more potent than brigatinib (Fig. 2A and B). As anticipated, these EGFR triple-mutant cells were entirely resistant to osimertinib with an IC₅₀ of approximately 1 μmol/L. Similarly, HJM-561 was also highly effective in inhibiting the growth of Ba/F3 cells expressing single- or double-mutant EGFR, especially for osimertinib-resistant Del19/C797S (IC₅₀ = 15.5 nmol/L) and L858R/C797S (IC₅₀ = 49.5 nmol/L) mutations (Supplementary Figs. S2 and S3). In contrast, HJM-561 showed much lower (<1 μmol/L) inhibitory activity in Ba/F3 cells overexpressing WT EGFR and A431 cells which are known to have high expression level of endogenous WT EGFR proteins (Fig. 2C; Supplementary Fig. S4). In addition, in several cell lines such as HEK293T, and MDA-MB-231, where EGFR expression is low, or undetectable, no inhibition on cell proliferation was observed by HJM-561 treatment (IC₅₀ > 3 μmol/L; Supplementary Fig. S4). These results indicated that HJM-561 overcame EGFR C797S-mediated osimertinib resistance while sparing WT EGFR.

We next investigated the PK profile of HJM-561 in mice through intravenous injection and oral administration routes. As shown in Fig. 3, HJM-561 displayed a favorable oral PK properties, with a C_max of 3677.25 ng/mL, an exposure AUC of 1970.2 hours × ng/mL and an oral bioavailability of 62.8% at 10 mg/kg. With the encouraging oral PK parameters in hand, we then examined the in vivo efficacy by oral dosing of HJM-561 in EGFR Del19/T790M/C797S-Ba/F3 cell–derived xenograft model. Osimertinib treatment at 10 mg/kg had no effect on the tumor growth. However, the oral administration of HJM-561 resulted in significant reduction of tumor volume by 58% and 84% at 20 mg/kg and 40 mg/kg doses, respectively (Fig. 4A). As expected, EGFR degradation and inhibition of phospho-EGFR were observed in HJM-561–treated mice (Supplementary Fig. S5). In order to evaluate the potential clinical response of HJM-561 in patients with NSCLC that developed osimertinib resistance, a EGFR Del19/T790M/C797S–driven PDX model was utilized for efficacy evaluation (25, 26). This PDX model was developed from a patient with poorly to moderately differentiated lung adenocarcinoma who progressed on erlotinib and osimertinib treatment and carried an allele frequency of 46% Del19, 17% T790M, and 17% C797S. HJM-561 dose-dependently suppressed the tumor growth in this model, and led to 67% TGI at 40 mg/kg. The antitumor efficacy of HJM-561 was comparable with TQB3804, a potent small molecule EGFR triple-mutant inhibitor currently in phase I clinical trial (Fig. 4C). Body weight of mice treated with HJM-561 was not significant changed in these animal studies (Fig. 4B and D). Taken together, HJM-561 as a novel EGFR mutant PROTAC demonstrated robust oral antitumor activity in EGFR Del19/T790M/C797S-driven, and osimertinib-resistant Ba/F3 CDX and PDX models.

In this study, we describe an orally bioavailable EGFR mutant PROTAC, HJM-561, which potently and selectively degrades osimertinib-resistant EGFR triple mutants (Del19/T790M/C797S and L858R/T790M/C797S). HJM-561 inhibits the proliferation of Ba/F3 cells expressing the EGFR C797S-containing triple mutants while leaving WT EGFR cells unaffected. Notably, oral administration of HJM-561 demonstrates potent antitumor activity in EGFR Del19/T790M/C797S-driven CDX and PDX tumor models. Our data suggests that HJM-561 may be a potential treatment option for patients with NSCLC with osimertinib-resistant EGFR triple mutations.

The C797S-containing mutation is one of the major resistance mechanisms to third-generation EGFR TKIs such as osimertinib, and currently represents the most challenging problem in clinic for treating patients with NSCLC (6–8, 27, 28). A variety of reversible, or allosteric small molecule inhibitors, and/or their combinations with an EGFR antibody were developed to address EGFR C797S-containing triple mutants (29, 30). However, the majority of these inhibitors have not shown adequate efficacy to warrant entering clinical evaluation so far. PROTACs have a significant advantage over traditional small molecule inhibitors in overcoming drug resistance because the removal of mutant proteins may avoid the emergence of resistance mutations arising from the treatment with traditional small molecule inhibitors (31). A great deal of efforts has been made to develop PROTACs to resolve EGFR TKI resistance. However, the PROTACs against EGFR mutations including C797S were reported to have activities only on cellular level (19–21, 32). To the best of our knowledge, HJM-561 is the first and only known PROTAC molecule to inhibit EGFR Del19/T790M/C797S with a strong oral efficacy in vivo. This is of particular importance because this triple mutant is the predominant activating mutation found in osimertinib-resistant patients. SIAIS164018, which was recently reported to degrade ALK and EGFR double mutations (L858R/T790M), is also a brigatinib-based CRBN degrader with subtle structural differences to HJM-561 (33). However, HJM-561 is much more potent than SIAIS164018 in degrading EGFR triple mutants (Supplementary Figs. S6 and S7). This indicates that subtle structural differences in the linker of PROTACs may change the degradation profile significantly.

The exact mechanism of HJM-561 to selectively degrade EGFR triple mutants, but not the WT EGFR is not clear at the moment. HJM-561 is designed to use brigatinib as the warhead of the PROTAC molecule, which is a weak inhibitor of both WT and triple-mutant EGFR in cell proliferation assays (Fig. 2). Our PROTAC compound HJM-561 has, however, significantly higher inhibitory activity for the triple mutants, but not for the WT EGFR in the same cellular assays. Although there is no simple explanation for the mechanism, it is not uncommon for a PROTAC molecule to have a different degradation selectivity profile than the inhibition selectivity profile of the parent small molecule compound. This could be explained by the unique nature of PROTACs, which require a perfect conformational match between E3 ligases and target proteins to facilitate the ubiquitination needed for target degradation to occur. HJM-561 may lead to a preferential or better conformational match between the E3 ligase and the mutant EGFRs than the WT EGFR. Future cocrystallization studies of HJM-561 tertiary complex with WT and mutant EGFRs may shed light on the selectivity.

As reported in the literature, transient and relatively weak binding between PROTACs and the target proteins can induce effective target protein degradation, which endows one PROTAC molecule with the ability to degrade multiple targeted proteins. HJM-561 indeed displays strong degradation activity against EGFR single or double mutations in addition to C797S triple mutations (Supplementary Figs. S2, S3, and S8). That implies that HJM-561 has the potential to counteract drug resistance induced by most EGFR TKIs including osimertinib and the early generations of EGFR TKIs.

HJM-561 displays a high plasma exposure by oral administration, which is sufficient to achieve its effective concentration as observed in cellular assays. The oral bioavailability of HJM-561 in mice is as high as 62.8%, which is remarkable for a PROTAC molecule despite its inobservance of Lipinski rules. The superior oral PK profile and strong in vivo efficacy supports further clinical development of HJM-561 for treating patients with osimertinib-resistant NSCLC.

Based on prior studies reported in the literature, WT EGFR inhibition causes dose-limiting toxicities, such as diarrhea and unbearable rashes (34). HJM-561 shows approximately 100-fold selectivity over WT EGFR in cell proliferation assays, and did not affect the growth of normal cells in our studies (Supplementary Fig. S4). This indicates that HJM-561 is less likely to cause the similar adverse events observed with some prior EGFR small molecule inhibitors in human studies. More extensive preclinical toxicology studies, however, are necessary to assess its overall safety profile and therapeutic index before advancing this PROTAC into clinical studies. Although HJM-561 did not seem to exhibit a clear efficacy advantage in our EGFR triple mutant PDX study over the small molecular inhibitor TQB3804, further work may potentially illustrate the advantages of HJM-561 as a degrader, such as a less frequent dosing scheme, different selectivity/safety profiles, and delay of resistance occurrence in future preclinical and clinical studies.

No disclosures were reported.

Y. Du: Conceptualization, supervision, writing–original draft, writing–review and editing. Y. Chen: Formal analysis, methodology. Y. Wang: Methodology. J. Chen: Formal analysis. X. Lu: Validation. L. Zhang: Methodology. Y. Li: Formal analysis. Z. Wang: Data curation, methodology. G. Ye: Supervision. G. Zhang: Conceptualization, supervision.