Development of PVTX-405 as a potent and highly selective molecular glue degrader of IKZF2 for cancer immunotherapy

Abstract

IKZF2 (Helios) is a transcription factor that is selectively expressed by Tregs and is essential for preserving the function and stability of Tregs in the tumor microenvironment (TME), where it suppresses the anti-tumor immune response. Targeted IKZF2 degradation by small molecules represents a promising strategy for the development of a new class of cancer immunotherapy. Herein, we describe the discovery of PVTX-405, a potent, effective, highly selective, and orally efficacious IKZF2 molecular glue degrader. PVTX-405 degrades IKZF2 (DC₅₀ = 0.7 nM and D_max = 91%) while sparing other CRBN neo-substrates. Degradation of IKZF2 by PVTX-405 increases production of inflammatory cytokine IL-2 and reduces the suppressive activity of Tregs, leading to an increase in Teff cell proliferation. Once-daily oral administration of PVTX-405 as single agent significantly delays the growth of MC38 tumors in a syngeneic tumor model using humanized CRBN mice. PVTX-405 in combination with anti-PD1 or anti-LAG3 significantly increases animal survival compared to anti-PD1 or anti-LAG3 alone. Together, these results demonstrate that PVTX-405 is a promising IKZF2 degrader for clinical development for the treatment of human cancers.

Introduction

Immune checkpoint therapies (ICTs), that stimulate a patient’s own immune system to recognize and destroy cancer cells, have revolutionized cancer treatment over the past decade¹. ICTs, such as PD-1/PD-L1 and CTLA-4 blocking antibodies, have shown remarkable success in treating various types of cancers^2,3. However, only a subset of patients respond well to ICTs^2,4. One way tumors evade ICT-induced immune cell killing is by recruiting immune-suppressive CD4⁺/FOXP3⁺/CD25⁺ regulatory T-cells (Treg)^4,5. Treg cells can strongly suppress the expansion and function of effector T (Teff) cells through various mechanisms such as depleting cytokine IL-2 from the tumor microenvironment (TME), secreting immunosuppressive cytokines IL-10, IL-35, and transforming growth factor-β (TGFβ) in the tumor TME, as well as inducing Teff cell senescence^6,7. Therefore, therapies that can precisely target Treg cells in the TME will enhance Teff cell function and may expand the benefit of ICTs to a larger patient population.

IKZF2 (Helios) is a member of the Ikaros transcription factor family that includes IKZF1 (Ikaros), IKZF3 (Aiolos), IKZF4 (Eos) and IKZF5 (Pegasus)⁸. Beyond development, the expression of IKZF2 is restricted in Treg cells and some thymocyte subsets^9,10. Accumulating evidence suggests that IKZF2 is essential for the stability and immune-suppressive activity of Treg cells^11,12,13. Increased IKZF2 expression in Tregs enhanced their immune-suppressive activity, while loss of IKZF2 resulted in destabilization of Tregs, leading to increased expression of pro-inflammatory cytokines^12,13,14,15. Selective knockdown of IKZF2 in CD4 Tregs resulted in significantly improved antitumor immunity in syngeneic xenograft tumor models¹³. Interestingly, selective depletion of IKZF2 in CD4 Tregs caused destabilization of Tregs residing in the TME only, not in peripheral lymphoid tissue. This suggests that IKZF2 depletion may avoid potential side effects caused by systemic immune activation from Treg destabilization in non-tumor tissues^15,16. Overall, these studies suggest that depleting IKZF2 in Treg cells may be an effective immunomodulatory strategy to enhance the efficacy of ICTs without the toxicities associated with systemic immune activation.

IKZF2 is a zinc finger transcription factor and such type of proteins is historically considered as undruggable targets due to their disordered structures and lack of druggable sites^16,17,18,19. However, the immunomodulatory drugs (IMiDs) pomalidomide and lenalidomide induce the degradation of the closely related family members IKZF1 and IKZF3 through a “molecular glue” mechanism^20,21,22,23. Rather than binding directly to IKZF1 and IKZF3, IMiDs first bind to CRBN. The IMiD-bound CRBN then recruits IKZF1 and IKZF3 for polyubiquitination and subsequent proteasome mediated degradation^20,21,22,23. Further structural modifications of IMiDs have been shown to induce the degradation of other zinc finger transcription factors, such as ZBTB16²⁴ and ZMYM2²⁵, and achieve selective degradation of other neo-substrates, such as GSPT1^26,27, suggesting that it is possible to induce selective IKZF2 degradation through structural modifications of the IMiDs. Indeed, the proof-of-concept IKZF2 degradation by a small molecule was demonstrated by Wang et al. who utilized structure-guided design to modify the pan-IKZF1/3 and GSPT1 degrader CC-885 and identified a potent IKZF2 molecular glue degrader, ALV2 (Fig. 1a)²⁸. While ALV2 did not induce degradation of GSPT1, it still caused significant IKZF1/3 degradation. Using a similar approach, Park et al. identified a series of potent dual CK1α and IKZF2 molecular glue degraders based on lenalidomide, exemplified by DEG-35 (Fig. 1a)²⁹. However, both IKZF1/3 and CK1α are important for maintaining the function of normal hematopoietic cells. Degradation of either IKZF1/3 or CK1α is likely to cause on-target hematopoietic toxicities^{29,30,31,32,33}.

Fig. 1: Previously reported molecular glue IKZF2 degraders and new molecular glue degraders of IKZF2.

a Chemical structures of representative, previously reported molecular glue IKZF2 degraders. b Chemical structures of new IKZF2 degraders and IKZF2 degradation data. Data are presented as mean of n = 2 biological replicates. Source data are provided as a Source Data file.

Bonazzi et al. have described the discovery of a selective molecular glue degrader of IKZF2, DKY709 (Fig. 1a), which does not induce the degradation of the other neo-substrates, including IKZF1/3, GSPT1 and CK1α³⁴. Oral administration of DKY709 induced potent and effective IKZF2 degradation³⁴. Two selective IKZF2 degraders, DKY709 and PLX-4545 (Fig. 1a) have entered Phase 1 clinical trials as cancer immunotherapy^34,35.

However, detailed characterization of DKY709 in our study (Table 1) identified several pharmacological properties that warrant optimization: DKY709 significantly degraded the neo-substrate SALL4 (DC₅₀ = 4.9 nM, D_max = 55%), which has been linked to teratogenic effects^36,37,38, and demonstrated moderate hERG inhibition (IC₅₀ = 7.1 µM). These findings highlight the potential to discover a potent and highly selective IKZF2 degrader with improved pharmacology by minimizing the off-target activities associated with DKY-709.

Table 1 Biological Activities and DMPK Profiles for 5 and DKY709

We present herein our design, synthesis and biological evaluation of a series of spiro-iso-indolinone based molecular glue degraders of IKZF2. Our efforts led to the discovery of PVTX-405, a potent, highly selective and orally efficacious IKZF2 degrader. As compared to DKY709, PVTX-405 shows enhanced degradation efficiency, significantly reduced off-target degradation, and much weaker hERG inhibition. PVTX-405 also showed favorable physicochemical properties and excellent DMPK and safety profiles in preclinical evaluations. A single, oral administration of PVTX-405 was capable of inducing profound and persistent depletion of IKZF2 protein in spleen and thymus tissues in humanized CRBN mice³⁹. PVTX-405 demonstrated tumor growth inhibition as single agent and improved survival of animals in combination with immune checkpoint therapies (ICTs) such as anti-PD1 or anti-LAG3. PVTX-405 is a suitable development candidate for evaluating the therapeutic potential of IKZF2 degradation alone or in combination with ICT in patients.

Results and discussion

Design of spiro-iso-indolinones as potent, selective and orally bioavailable IKZF2 degraders

Small molecule CRBN ligands bridge protein–protein interactions (PPI) between the CRL4–CRBN E3 ubiquitin ligase and neo-substrates, leading to polyubiquitination and subsequent proteasome-mediated degradation of IKZF2. While IKZF2 contains four N-terminal zinc finger (ZF1-4) DNA-binding domains and two C-terminal PPI domains (ZF5-6), the ZF2 domain was shown to interact with CRBN for induced IKZF2 degradation by molecular glue degraders³⁴.

We have developed multiple classes of potent CRBN ligands which could be used as “templates” for both molecular glue and heterobifunctional degraders. For example, we have designed compound RR-11055⁴⁰ as CRBN ligand (Fig. 1b) and have successfully employed RR-11055 for the development of a potent and orally efficacious PROTAC degrader of estrogen receptor⁴¹. However, our efforts employing cereblon ligand RR-11055 for the design of IKZF2 degraders have proven to be unsuccessful.

We envisioned that an alternative ring restriction strategy by shifting the 3-alkoxy chain onto the 4-N of the piperazine ring could form a spiro-ring system, which exposed the terminal piperidine ring for tail extension to generate putative IKZF2 degraders (Fig. 1b). We synthesized a series of spirocyclic ring systems on the iso-indolinone scaffold (Fig. 1b). Based upon previously reported IKZF2 degraders^28,29,34, we appended a benzyl group onto the nitrogen atom of the piperazine ring in our initial compounds (compounds 5–12, Fig. 1b). An IKZF2 HiBiT degradation assay was developed in HEK293 cells to quantitatively determine IKZF2 degradation of our synthesized compounds with the data summarized in Fig. 1b.

Gratifyingly, compound 5, which contains a 5,6 spiro-ring system, was found to be a potent and effective IKZF2 degrader, achieving DC₅₀ = 9.7 nM and D_max = 70% in the HiBiT HEK293 cells. In comparison, DKY709 showed DC₅₀ = 1.5 nM and D_max = 73% against IKZF2 under the same assay conditions. To evaluate the importance of the basic nitrogen of the piperazine ring for IKZF2 degradation, we synthesized compounds 6 and 7. Our data showed that 6 devoid of the basic amine on the piperidinyl group completely abolished IKZF2 degradation potency and efficacy. An ether derivative 7 slightly recovered degradation activity (D_max = 38%) from 6. The degradation data for compounds 6 and 7 hence suggested that the basic amino group of 5 is essential for IKZF2 degradation. We synthesized compound 8 with a smaller [5,5]-spiro ring, which was an ineffective IKZF2 degrader (D_max = 33%). Similarly, compound 9 containing a [5,7]-spiro system and compound 10 containing a [6,6]-spiro system were ineffective in IKZF2 degradation (D_max = 17% and D_max = 39%, respectively). Hence, our data on compounds 5, 8, 9 and 10 demonstrated the critical importance of the spiro ring sizes for achieving effective IKZF2 degradation.

DKY709 contains a basic nitrogen in its piperidinyl ring. This basic nitrogen, together with two aromatic groups, constitutes a classical pharmacophore for hERG inhibition⁴². Based upon recent cryo-EM structures of hERG in complex with inhibitors, the basic nitrogen atom in classical hERG inhibitors has cation-π interactions with the electron-rich phenol side chain of tyrosine 652 residue⁴³. In the structure of compound 5 (Fig. 1b), the oxygen atom in the spiro ring system would reduce the basicity of the nitrogen atom through the electron-withdrawing effect, as compared to that for DKY709. We envisioned that such reduction in the basicity of the nitrogen could have a benefit on decreasing the hERG inhibition as compared to DKY709 by reducing the cation-π interactions with hERG. We further examined if IKZF2 degradation activity could be sustained with further reduction of the basicity of the nitrogen atom by direct substitution on the piperidinyl ring. While hydroxylation on the piperidinyl ring (11) resulted in significant loss of IKZF2 degradation activity (DC₅₀ > 1000 nM and D_max = 39%), di-fluorination (12) (DC₅₀ = 5.7 nM and D_max = 68%) was well tolerated.

Encouraged by the initial IKZF2 degradation data for compound 5, we then evaluated its degradation selectivity using HiBiT assays for CRBN neo-substrates developed in the HEK293 cells, with DKY709 included as the control. Compound 5 did not induce any significant degradation of IKZF1/3, GSPT1, and CK1α at up to 10 µM concentrations (Table 1). Compound 5 showed significantly reduced degradation potency and efficacy on SALL4 (DC₅₀ > 1000 nM and D_max = 33%) compared with DKY709 (Table 1).

We next assessed the hERG inhibition, metabolic stability and Caco-2 permeability of 5, with DKY709 as the comparator, as shown in Table 1. In direct comparison with DKY709, compound 5 showed reduced hERG inhibition with IC₅₀ of 11 µM, similarly good metabolic stability in human, rat, and mouse liver microsomes (>60 min), >2-fold improved Caco-2 permeability with P_app value of 19 × 10^-6 cm/s and an efflux ratio (ER) value of 1.1.

Based on the compelling ADME data, we then evaluated the pharmacokinetics (PK) of 5 and DKY709 in mice and rats with the data also summarized in Table 1. In mice, compound 5 displayed an overall good PK profile, with moderate clearance (Cl = 40 mL/min/kg), reasonable half-life (T_1/2 = 2.9 h), high oral plasma exposure at 3 mg/kg (C_max = 398 ng/mL; AUC = 1336 h*ng/mL) and excellent oral bioavailability (F = 108%). Overall, compound 5 had an improved PK profile relative to DKY709 in mice. In rats, compound 5 also showed significantly improved PK properties compared to DKY709, with reduced clearance (Cl = 21 mL/min/kg), >3-fold higher oral plasma exposure at 3 mg/kg (C_max = 263 ng/mL; AUC = 1172 h*ng/mL) and increased oral bioavailability (F = 49%).

Taken together, as compared to DKY709, compound 5 displayed modestly weaker IKZF2 degradation potency and efficiency but much-reduced degradation activity against SALL4, weaker hERG inhibition, increased Caco-2 permeability and overall improved oral exposure with superior oral bioavailability in mice and rats. Hence, compound 5 represented an attractive lead compound for further optimization.

Determination of the ternary complex of CRBN:IKZF2:compound 5

To shed light on the structural basis for the compound 5-mediated recruitment of IKZF2 protein to CRBN E3 ligase, we determined the co-crystal structure for the ternary complex of CRBN:IKZF2:compound 5 at 1.7 Å resolution (Fig. 2, Supplementary Figs. 1–3, PDB ID: 9DOM). The CRBN binding portion of compound 5 engages similar interactions observed for DKY709 in binding to CRBN³⁴, with two hydrogen bond interactions between the glutarimide moiety of 5 and the backbone amide of CRBN H378 and W380 residues, along with the other hydrogen bond interaction between the carbonyl group in the middle isoindolinone ring of compound 5 and the CRBN N351 residue (Fig. 2b). Interestingly, the protonated piperidine nitrogen of 5 loses the charge–charge interaction with the CRBN E377 residue which was observed for DKY709³⁴. Instead, it forms a water-mediated hydrogen bond with the CRBN H353 residue and another water-mediated hydrogen bond with N143 in IKZF2 (Fig. 2b). In addition, the piperidine ring maintains a CH–π interaction with the H141 residue of IKZF2 (ZF2), similar to that of DKY709 (Fig. 2b). Furthermore, the orientation of the phenyl tail in 5 shifts from the right down direction to a right up direction next to the IKZF2 H141 residue, while the phenyl ring was largely exposed to solvent (Fig. 2b).

Fig. 2: X-ray crystal structure of IZKF2:CRBN:compound 5 ternary complex.

a Ternary structure of IKZF2 Zinc finger 2 (ZF2) domain (green) with CRBN Thalidomide-binding domain (TBD, gray) and compound 5 (yellow sticks) at 1.7 Å resolution (PDB ID: 9DOM). b Detailed interactions of IKZF2:CRBN:Compound 5 ternary structure with hydrogen bonds shown in dashed lines.

Further modifications of the initial lead compound 5

Our data showed that compound 5 represented an attractive lead compound for further optimization.

Based upon the co-crystal structure for 5, we replaced the phenyl tail group in 5 with aliphatic or aromatic groups to probe the interactions with the IKZF2 H141 residue. Representative compounds, together with their IKZF2 degradation activity in our HiBiT assay, are shown in Fig. 1b.

Replacing the phenyl group with a cyclohexyl group generated compound 13, which was similarly potent and effective in inducing IKZF2 degradation (DC₅₀ = 16 nM and D_max = 69%) as compared to compound 5. Changing the phenyl group to a 2-pyridinyl group yielded compound 14, which was a modestly weaker and less effective IKZF2 degrader (DC₅₀ = 18 nM and D_max = 64%). Replacing phenyl with an electron-rich thiophene resulted in compound 15, which was moderately more potent and effective in inducing IKZF2 degradation (DC₅₀ = 5.9 nM and D_max = 84%) than compound 5.

We next focused on installation of different substitutions on the phenyl group in compound 5. Our efforts in this line of exploration turned out to be very fruitful and yielded a number of highly potent and effective IKZF2 degraders. Among them, compound 16 (Fig. 1b), containing a 1-methyl-1H-pyrazole substitution at the meta-position of the phenyl ring, was the best and achieved a DC₅₀ value of 0.8 nM and a D_max value of 89% in IKZF2 degradation.

Since 16 was a racemic mixture, it was resolved into both enantiomeric isomers for further evaluation (Table 2). The S-isomer, compound 16a, was the more potent and effective IKZF2 degrader with DC₅₀ = 0.7 nM and D_max = 91%, whereas the R-isomer, compound 16b was a somewhat weaker IKZF2 degrader with DC₅₀ = 4.1 nM and D_max = 89% (Table 2). Accordingly, compound 16a (PVTX-405) was selected for extensive pre-clinical profiling. A comprehensive evaluation of PVTX-405 for CRBN binding activity, potential off-target degradation on several other neo-substrates, and hERG inhibition, in comparison with DKY709 was summarized in Table 2.

Table 2 Comparation of PVTX-405 (16a), 16b and DKY709

PVTX-405 bound to cereblon with IC₅₀ of 99 nM and was five-fold tighter than DKY709. In the IKZF2 HiBiT degradation assay in HEK293 cells, it displayed DC₅₀ of 0.7 nM and D_max of 91%, whereas DKY709 had DC₅₀ of 1.5 nM and D_max of 73%. Hence, PVTX-405 was 2-fold more potent than DKY709 in inducing IKZF2 degradation but much more effective than DKY709 based upon the deeper D_max value. Furthermore, PVTX-405 had much weaker hERG inhibition with IC₅₀ of 48 µM and IC₃₀ of 18 µM in a manual patch-clamp assay, whereas DKY709 displayed IC₅₀ of 9.0 µM and IC₃₀ of 3.0 µM in the same assay.

Profiling of degradation of other neo-substrates revealed that PVTX-405 had minimal effect in reducing levels of IKZF1, IKZF3, GSPT1 and CK1α proteins with D_max values of <20% at up to 10 µM concentrations (Table 2). While DKY709 had a potent degradation activity on SALL4 with DC₅₀ of 4.9 nM and D_max of 55%, PVTX-405 showed a very weak activity (DC₅₀ > 1000 nM, D_max = 33%) (Table 2).

Further evaluation in Jurkat cells using FACS assay showed that PVTX-405 demonstrated potent IKZF2 degradation with DC₅₀ of 6.3 nM and D_max of 65% (Fig. 3a, Supplementary Fig. 4a). We next evaluated IKZF2 degradation kinetics for PVTX-405 and DKY709 with IKZF2 HiBiT assay in HEK293 cells (Fig. 3b). Both PVTX-405 and DKY709 displayed time-dependent IKZF2 degradation. However, PVTX-405 showed more rapid degradation kinetics than DKY709 and achieved maximal degradation by merely 3 h while DKY709 required >6 h to reach a D_max plateau.

Fig. 3: PVTX-405 rapidly degrades IKZF2 with high potency in a CRBN-dependent manner.

a FACS analysis of IKZF2 degradation in Jurkat cells after PVTX-405 treatment. Data presented as mean ± SD, with n = 2 biological replicates for each concentration. b Kinetics of IKZF2 degradation by PVTX-405 and DKY709 in HEK293 IKZF2 HiBiT cell line. Data presented as mean ± SD, with n = 2 biological replicates for each concentration. c Western blotting analysis shows that CRISPR/Cas9 mediated KO of CRBN in Jurkat cells completely abrogates IKZF2 degradation by PVTX-405. Representative blots from two independent experiments. Source data are provided as a Source Data file.

We next confirmed CRBN-dependent degradation of IKZF2 by PVTX-405. CRISPR/Cas9 mediated CRBN KO in Jurkat cell line completely abrogated PVTX-405 induced IKZF2 degradation, as demonstrated by western blot assay (Fig. 3c). In summary, PVTX-405 is a rapid, selective, potent and effective CRBN-dependent glue degrader of IKZF2 with superior neo-substrate selectivity and alleviated hERG inhibition as compared to DKY709.

Effect of PVTX-405 on T-cells in vitro

IKZF2 and IKZF1 are both involved in the inhibition of interleukin-2 (IL-2) transcription, a key inflammatory cytokine¹⁴. PVTX-405 and DKY709 treatment of Jurkat T cells stimulated with anti-CD3/CD28 antibody cocktail at the bead-to-cell ratio of 1:4 led to an increase in IL-2 concentration in the supernatant (Fig. 4a). These data showed that PVTX-405 effectively regulates an established IKZF2 transcriptional target in vitro in Jurkat T cells. We next examined the impact of PVTX-405 and DKY709 on primary human Treg cells in vitro. A dose-dependent degradation of IKZF2 (Fig. 4b, Supplementary Fig. 4b), and not IKZF1 (Supplementary Fig. 4b and Supplementary Fig. 5), was observed in human primary CD25-enriched FOXP3 + CD4 T cells upon treatment with PVTX-405 and DKY709. PVTX-405 (DC₅₀ = 6 nM and D_max = 92%) demonstrated more rapid and potent degradation of IKZF2 in human Treg cells in vitro as compared to DKY709 (DC₅₀ = 12 nM and D_max = 73%) after a 6 h treatment (Fig. 4b). Treg cells stimulated and expanded in vitro in the presence of PVTX-405 showed significantly decreased capacity to suppress Teff cell proliferation as compared to DKY709 in a Treg:Teff cell co-culture assay performed using cells derived from six different human donors (Fig. 4c, Supplementary Fig. 4c). These data demonstrated that through potent and rapid degradation of IKZF2, PVTX-405 modulates Treg cell function in vitro.

Fig. 4: PVTX-405 effectively degrades IKZF2, destabilizes human T_regs ex vivo, and induces the proliferation of T_eff cells.

a IL-2 levels secreted by Jurkat cells after activation with CD3/CD28 beads at a beads-to-cell ratio of 1:4 and treatment with increasing levels of PVTX-405 or DKY709 for 24 h. Data presented as mean ± SD, with n = 3 biological replicates for each concentration. b PVTX-405 shows significantly higher potency and maximal degradation of IKZF2 compared to DKY709 after 3 and 6 h treatment in primary human Tregs ex vivo. Data presented as mean ± SD, with n = 2 biological replicates for each concentration. c Suppression of Teff proliferation by primary human Treg cells expanded in the presence of PVTX-405 or DKY709 or DMSO. Data are shown as six independent experiments/donors, with median ± interquartile range at 1:4 Treg/Teff coculture ratio. Data are normalized to the proliferation of Teff alone for each donor. *p < 0.05. Groups were compared using two-tailed paired parametric t test. Source data are provided as a Source Data file.

Preclinical DMPK and in vitro toxicity profiling of PVTX-405

We extensively evaluated PVTX-405 for its physiochemical properties, preclinical DMPK properties and potential in vitro toxicity, with the data summarized in Table 3.

Table 3 Preclinical DMPK and in vitro toxicity profiling of PVTX-405

PVTX-405 exhibited a high solubility in simulated gastric fluid (SGF), simulated intestinal fluid (SIF), fed state simulated intestinal fluid (FeSSIF), and fasted state simulated gastric fluid (FaSSGF). PVTX-405 showed an excellent metabolic stability in human, rat and mouse liver microsomes with T_1/2 > 120 min, and a moderate metabolic stability in monkey liver microsome with T_1/2 = 36 min. PVTX-405 displayed an excellent plasma stability with T_1/2 > 4 h for all the 4 species evaluated (human/mouse/rat/monkey). It demonstrated good Caco-2 permeability (P_app = 5.9 × 10^-6 cm/s) and a moderate efflux ratio (ER = 4.4). In addition, PVTX-405 showed a relatively high but consistent plasma protein binding in human, monkey, rat and mouse plasma with 5.7-8.5% unbounded, free fraction of the drug. In the CYP inhibition assay, PVTX-405 exhibited a clean profile with IC₅₀ > 10 µM against all the CYP enzymes evaluated, including 3A4, 1A2, 2B6, 2C8, 2C19 and 2D6.

PVTX-405 was evaluated for pharmacokinetics in mice, rats and monkeys with intravenous and oral administration. It showed an excellent PK profile in mice, characterized with a moderate clearance (Cl = 19 mL/min/kg), a favorable volume of distribution (V_ss = 2.7 L/kg), a long half-life (T_1/2 = 3.8 h), a high oral plasma exposure at 3 mg/kg (C_max = 600 ng/mL, AUC = 1997 h*ng/mL), and an excellent overall oral bioavailability (F = 76%). In rats, PVTX-405 also had an excellent PK profile and an oral bioavailability of 118%. Although PVTX-405 exhibited a moderate metabolic stability in monkey liver microsome, it still displayed a good PK profile in cynomolgus monkey, with a moderate clearance (Cl = 19 mL/min/kg), a high volume of distribution (V_ss = 8.1 L/kg), a long half-life (T_1/2 = 10 h), a high oral plasma exposure at 3 mg/kg (C_max = 111 ng/mL, AUC = 1289 h*ng/mL) and a good oral bioavailability (F = 50%). Furthermore, PVTX-405 had no significant PXR (Pregnane X Receptor) activation, no detected reactive metabolites, no hepG2 cytotoxicity (IC₅₀ > 50 µM), no significant inhibition on aldehyde oxidase and BSEP-mediated taurocholate transporter and was negative in the mini-Ames assay. Hence, PVTX-405 showed a very clean toxicity profile in all in vitro preclinical safety evaluations. Taken together, PVTX-405 demonstrated very favorable physicochemical, DMPK and safety profiles in its extensive preclinical evaluations.

In vivo PK/PD evaluation of PVTX-405

Previous studies have demonstrated that a single valine-to-isoleucine amino acid difference between human and murine CRBN in the CRBN-neo-substrate contacted region (Thalidomide Binding Domain, or TBD) is responsible for the lack of IMiD-induced degradation of neo-substrates in murine models^39,44,45. Furthermore, a simple change of the Ile391 in mouse CRBN to the corresponding human valine amino acid restored the IMiD-induced degradation of IKZF1/3 and CK1α^39,44,45 and the ALV2-induced IKZF2 degradation in humanized Crbn^I391V mice²⁸. We thus employed the humanized Crbn^I391V mouse model to evaluate the in vivo PK/PD and efficacy of PVTX-405.

Because the spleen and thymus are two major immune-related organs abundant with Treg cells and other T cells with high IKZF2 protein expression⁴⁶, we assessed the PK/PD effect of PVTX-405 in these organs, with the data summarized in Fig. 5 and Supplementary Fig. 6. PVTX-405 showed a dose-dependent and profound IKZF2 degradation in both spleen and thymus tissues (Fig. 5a). A single oral administration of PVTX-405 at 10, 30 and 100 mg/kg depleted IKZF2 protein by 70%, 84%, and 89%, respectively, in spleen, and 62%, 65% and 73%, respectively, in thymus at 6 h (Fig. 5a). Furthermore, PVTX-405 demonstrated sustained depletion of IKZF2 protein at 12 and 24 hour time points in both tissues. Analysis of the drug concentrations in these tissues revealed a dose-dependent increase in PVTX-405 concentration at the 6-h time point. At 10, 30, and 100 mg/kg doses, PVTX-405 concentrations in the spleen were 186, 1926, and 7177 ng/g, respectively, and in the thymus were 76, 413, and 2854 ng/g, respectively at 6 h time-point (Fig. 5b). The drug concentration in both tissues significantly decreased (>90%) at 24 h time-point (Fig. 5b). Of note, the depletion of IKZF2 protein by PVTX-405 in spleen was more pronounced than that in thymus (Fig. 5a), consistent with the higher drug concentration in spleen than in thymus (Fig. 5b).

Fig. 5: PK/PD evaluation of PVTX-405 in immune-competent and gene-edited C57BL/6 mice harboring humanized Crbn^I391V gene.

Crbn^I391V mice were administered with vehicle control or PVTX-405 via oral gavage using 100% PEG200 as the dosing vehicle. Mice were sacrificed at indicated time-points, and blood samples, spleen tissues and thymus tissues were harvested for analysis. At each time point, 4 mice were euthanized for each dosing group. a IKZF2 degradation induced by PVTX-405 in spleen and thymus tissues in mice. A method of one-way ANOVA was used for P value calculation: “**”P < 0.01; “****”P < 0.0001. The p values for all PVTX-405 treated groups compared with their corresponding vehicle control are <0.0001, except a p value of 0.0088 for the group (10 mg/kg PVTX-405, Thymus, 24 h) compared with the vehicle control of thymus tissue. Data presented as mean ± SEM, with n = 4 animals for each treatment group. b Concentrations of PVTX-405 in plasma, spleen and thymus at 6, 12, and 24 h time-points and 10, 30, and 100 mg/kg doses. “ND”: not determined. Data presented as mean ± SEM, with n = 4 animals for each treatment group. Source data are provided as a Source Data file.

In vivo efficacy study of PVTX-405

Mice with IKZF2 deficiency in CD4 T cells demonstrated enhanced antitumor immunity in syngeneic mouse tumor models¹³. However, it is unknown that if pharmacological depletion of IKZF2 using small molecule degraders will recapitulate the same anti-tumor activity. One of the primary limitations in the field so far has been lack of availability of syngeneic mouse tumor models where CRBN-based glues can demonstrate degradation in mouse cells. To overcome this limitation, we developed a syngeneic tumor model of MC38 in humanized CRBN mice (Crbn^I391V mice). We hypothesized that since MC38 tumors⁴⁷ and humanized CRBN mice³⁹ are developed using C57BL/6 mice, humanized Crbn^I391V mice will not mount allogenic response to MC38 tumor cells. Our tumor engraftment study of MC38 tumors in Crbn^I391V mice validated our hypothesis (Fig. 6a).

Fig. 6: PVTX-405 shows anti-cancer efficacy in the MC38 mouse tumor xenograft model in immune-competent Crbn^I391V C57BL/6 mice.

PVTX-405 or vehicle control was administrated to 10 mice in each group via oral gavage using 100% PEG200 as the dosing vehicle. a Tumor growth for each group. A method of one-tailed t test with Welch’s correction was used for p value calculation. “*”p < 0.05. Data presented as mean ± SEM, with n = 10 animals for each treatment group. b Animal body weight change for each group. Data presented as mean ± SEM, with n = 10 animals for each treatment group. Source data are provided as a Source Data file.

Single oral administration of PVTX-405 at 30 mg/kg effectively induced IKZF2 depletion in Crbn^I391V mouse (Fig. 5). We thus evaluated the antitumor activity of PVTX-405 in the MC38 mouse tumor syngeneic model with a 30 mg/kg daily oral dose, with obtained data summarized in Fig. 6. Oral administration of 30 mg/kg PVTX-405 significantly suppressed MC38 tumor growth (P = 0.022), with 68% of tumor growth inhibition at the end of treatment (day 22) (Fig. 6a). PVTX-405 was well tolerated in Crbn^I391V mice and did not induce any weight loss (Fig. 6b) and/or other signs of toxicity during the duration of dosing.

Weak tumor-associated inflammatory responses and the immune-suppressive function of Tregs can be major hurdles to broad effectiveness of the immune checkpoint therapies (ICTs)^4,5. We therefore determined whether reduction of Treg suppressive activity by PVTX-405 could improve efficacy of ICTs such as anti-PD1 or anti-LAG3 antibody by unleashing anti-tumor activity of Teff cells. PVTX-405 was evaluated in combination with anti-PD1 or anti-LAG3 antibody in our established MC38 syngeneic tumor model using Crbn^I391V mice with obtained data summarized in Fig. 7. Daily oral administration of PVTX-405 at 30 mg/kg in combination with once-weekly administration of anti-PD1 antibody at 1 mg/kg significantly delayed the time required for tumors to reach 2000 mm³ as compared to once-weekly anti-PD1 antibody alone at 1 mg/kg alone (p = 0.029) (Fig. 7a). In addition, PVTX-405 treatment in combination with anti-PD1 antibody showed more complete responses compared to anti-PD1 treatment alone (Fig. 7b). Similar to PVTX-405 and anti-PD1 antibody combination study, daily oral administration of PVTX-405 at 30 mg/kg in combination with once-weekly administration of anti-LAG3 antibody at 10 mg/kg significantly delayed the time required for tumors to reach 2000 mm³ as compared to once-weekly anti-LAG3 antibody at 10 mg/kg alone (p = 0.025) (Fig. 7c). PVTX-405 treatment in combination with anti-LAG3 antibody also showed more complete responses as compared to anti-LAG3 treatment alone (Fig. 7d). Both combination therapies were well tolerated in Crbn^I391V mice (Fig. 7e, f). Overall, our data provide compelling evidence that degradation of IKZF2 by PVTX-405 destabilizes Tregs in vitro and ex vivo and inhibits their immune suppressive function (Fig. 4). Importantly, PVTX-405 exhibits robust immunomodulatory activity in vivo, effectively inhibiting MC38 tumor growth in mice alone and in combination with ICTs.

Fig. 7: PVTX-405 in combination with either Anti-PD1 or Anti-LAG3 shows synergistic anti-cancer efficacy in the MC38 mouse tumor xenograft model in immune competent Crbn^I391V C57BL/6 mice.

a Kaplan-Meier survival curve for time it requires animal to reach tumor volume of 2000 mm³. PVTX-405 induces significantly better tumor growth reduction in combination with PD1 blockade as compared to PD1 blockade alone (p = 0.029, log-rank test, two sided). b Spider plot showing tumor growth kinetics for each animal in various treatment groups for the duration of the study. PVTX-405 shows more complete responses in combination with anti-PD1 than anti-PD1 alone. c Kaplan-Meier survival curve for time it requires animal to reach tumor volume of 2000 mm³. PVTX-405 induced significantly better tumor growth reduction in combination with LAG3 blockade compared to LAG3 alone (p = 0.025, log-rank test, two sided). d Spider plot showing tumor growth kinetics for each animal in various treatment groups for the duration of the study. PVTX-405 shows a significant increase in complete responses in combination with anti-LAG3 than anti-LAG3 alone. e Animal body weight change for the group of PVTX-405 + Anti-PD1 combination study. Data presented as mean ± SD, with n = 10 animals for each treatment group. f Animal body weight change for the group of PVTX-405 + Anti-LAG3 combination study. Data presented as mean ± SD, with n = 10 animals for each treatment group. Source data are provided as a Source Data file.

Summary

IKZF2 is a transcription factor and a promising target for cancer immunotherapy. While IKZF2 molecular glue degraders have been reported, previous degraders have significant limitations for clinical development in order to fully evaluate IKZF2 degradation as a new therapeutic strategy. For example, DKY709, a previously reported potent and selective IKZF2 degrader, which has been advanced into clinical development, has significant off-target activity.

In the present study, we described our development of potent, selective, and oral bioavailable IKZF2 molecular glue degraders. Our structure-guided medicinal chemistry campaign led to the discovery of PVTX-405, a tetracyclic spiro-iso-indolinone-based IKZF2 molecular glue degrader. In direct comparison to DKY709, PVTX-405 displayed superior IKZF2 degradation potency, efficacy, selectivity over other neo-substrates, significantly reduced hERG inhibition, and an excellent PK profile in mice, rats, and monkeys.

Our mechanistic data demonstrated that targeted degradation of IKZF2 by PVTX-405 significantly increased IL-2 production in vitro and reduced the suppressive activity of Tregs ex vivo. PVTX-405 significantly delayed the growth of MC38 tumors in humanized CRBN mice with once-daily oral administration. Of significance, PVTX-405, when used in combination with ICTs such as anti-PD1 or anti-LAG3 antibody, markedly enhanced animal survival and induced durable tumor regressions compared with ICTs alone. Collectively, our data highlight the strong synergy between ICTs and Treg destabilization through pharmacological degradation of IKZF2, presenting a promising combination strategy for cancer immunotherapy. Taken together, PVTX-405 is a potent, highly selective, and orally efficacious molecular glue degrader of IKZF2. PVTX-405 has been selected as an advanced preclinical candidate to be developed as a new immunomodulatory agent for the treatment of human cancers alone and in combination with ICTs.

Methods

All the animal studies comply with all relevant ethical regulations. All the in vivo PK/PD and efficacy animal studies were performed under University of Michigan IACUC approval of animal protocols: PRO00009463 and PRO00011174 (P.I. Shaomeng Wang). All the PK studies were performed in Shanghai Medicilon Inc. Shanghai, China, in compliance with institutional ethical guidelines. No human study was involved.

Chemistry

The synthesis of PVTX-405 and other IKZF2 degraders and their chemical data are described in Supplementary Information.

HTRF cereblon binding assay

The binding potency was determined using HTRF assay technology (Perkin Elmer). Compounds were serially diluted in DMSO and 0.2 µL volume was transferred to white 384-well plate. The reaction was conducted in total volume of 20 µL with the addition of 2 nM His tagged CRBN + DDB-DLS7 + CXU4 (Wuxi, catalog #RP210521GA) to compounds followed by the addition of 60 nM Fluorescent probe Cy5-labeled Thalidomide (Tenova Pharma, catalog # T52461), and 0.4 nM of MAb Anti-6HIS Tb cryptate Gold (Cisbio, catalog # 61HI2TLA in the assay buffer (50 mM HEPES pH 7.5, 1 mM TCEP, 0.01% Brij-35, 50 mM NaCl, and 0.1% BSA). After one hour incubation at room temperature, the HTRF signals were read on Envision reader (Perkin Elmer). Data was analyzed using XLfit using four parameters dose–response curve to determine IC₅₀ values.

HiBiT degradation assay

IKZF1-HiBiT, IKZF2-HiBiT, IKZF3-HiBiT, IKZF4-HiBiT, SALL4-HiBiT, CK1α-HiBiT, and GSPT1-HiBiT expressing cells were generated with the FLP-IN system (Invitrogen). HEK293 cells containing the FRT recombination site were purchased from Invitrogen. IKZF1-HiBiT, IKZF2-HiBiT, IKZF3-HiBiT, IKZF4-HiBiT, SALL4-HiBiT, CK1α-HiBiT, and GSPT1-HiBiT cDNA in pcDNATM5/FRT constructs were generated by inserting IKZF1-HiBiT, IKZF2-HiBiT, IKZF3-HiBiT, IKZF4-HiBiT, SALL4-HiBiT, CK1α-HiBiT, and GSPT1-HiBiT cDNA (Generay Biotech and Azenta Lifesciences) in pcDNATM5/FRT (Invitrogen) construct. IKZF1-HiBiT, IKZF2-HiBiT, IKZF3-HiBiT, IKZF4-HiBiT, SALL4-HiBiT, CK1α-HiBiT, and GSPT1-HiBiT were inserted to the genomic recombination site by FLP recombinase enzyme using pcDNATM5/FRT constructs. All stable cell lines were selected for hygromycin B resistance and consequently cultured in Dulbecco’s modified Eagle’s medium (DMEM) + 10% FBS + 1% P/S + 100 mg/ml hygromycin B (Invitrogen). Integration of IKZF1-HiBiT, IKZF2-HiBiT, IKZF3-HiBiT, IKZF4-HiBiT, SALL4-HiBiT, CK1α-HiBiT, and GSPT1-HiBiT pcDNA™5/FRT constructs into the FRT site was verified by Nano-Glo® HiBiT Lytic Detection System (Promega) using Envision plate reader (PerkinElmer).

IKZF1-HiBiT, IKZF2-HiBiT, IKZF3-HiBiT, IKZF4-HiBiT, SALL4-HiBiT, CK1α-HiBiT, and GSPT1-HiBiT cells were treated at desired compound concentrations (0.017 nM to 10 µM) and DMSO as vehicle control for 6 h. After 6 h of treatment HiBiT signal was detected using Nano-Glo® HiBiT Lytic Detection System (Promega) on Envision plate reader (PerkinElmer). The data was further analyzed using the Spotfire program, employing a four-parameter dose-response curve to determine DC₅₀ and D_max.

Western blotting analysis

Cells were maintained in the appropriate culture medium with 10% FBS at 37 °C and an atmosphere of 5% CO₂. All the cell lines were used within three months of thawing fresh vials. Jurkat CRBN KO cells were generated by transfecting Jurkat cells with TruCut Cas9 Protein V2 (Thermo Fisher Scientific) and TruGuide Synthetic sgRNA against CRBN (Thermo Fisher Scientific, CRISPR668069_SGM). Cells were lysed 1X Cell Lysis Buffer (Cell Signaling Technology, #9803), resolved by SDS-PAGE NuPAGE gel (Thermo Fisher Scientific), and transferred to a PVDF membrane (Millipore). Membranes were blocked using Odyssey TBS Blocker Buffer (LI-COR). IRDye 680RD and 800CW Dye-labeled secondary antibodies (LI-COR) were used. The washed membranes were scanned using Odyssey CLx imager (LI-COR). The intensity of Western blot signaling was quantitated using the Odyssey software. Primary antibodies used are: Helios (D8W4X) XP® Rabbit mAb (Cell Signaling Technology, #42427), CRBN Rabbit polyclonal antibody (Millipore Sigma, HPA045910), β-Actin (13E5) Rabbit mAb (Cell Signaling Technology, #4970), and GAPDH mouse monoclonal antibody (Santa Cruz Biotechnology, sc-47724). The uncropped and unprocessed scans of the blots for Fig. 3c and Supplementary Fig. 6 are provided in the Source Data file.

IKZF2 FACS assay

Jurkat cells (ATCC, Cat # HB-8065) were cultured in RPMI1640 + 10% FBS + 1% P/S. Cells were treated at desired compound concentrations (0.05 nM to 10 μM) and DMSO as vehicle control for 24 h. After 24 h of drug treatment cells were washed, fixed (3.7% PFA), and permeabilized with perm buffer (0.3% Triton X-100 in 1% BSA Solution). Subsequently, cells were stained with IKZF2 (1:100, Cell signaling) primary antibody and Alexa 488-labelend anti-rabbit IgG (1:200, Cell Signaling) secondary antibodies in staining buffer (1% BSA in PBS). Cells were images on iQue Flow cytometer and IKZF2 levels were quantified using iQue software. The data was further analyzed using the Spotfire program, employing a four-parameter dose–response curve to determine DC₅₀ and D_max. The gating strategy is illustrated in Supplementary Fig. 4a.

Frozen PBMCs from healthy donors were purchased from Miaoshun Biotechnology Co. (Shanghai) in accordance with their respective IRBs. Informed consent was obtained from all participants and all ethical regulations were respected. PBMC cells were thawed and resuspended in complete T-cell media (RPMI + 10% FBS + 1% P/S). Cells were treated at desired compound concentrations (0.05 nM to 1 μM) and DMSO as vehicle control for 3 or 6 h. After desired time of drug treatment cells were washed and first stained with extracellular staining master mix containing anti-CD45 (1:20, Biolegend), anti-CD3 (1:50, BD), anti-CD4 (1:50, BD), and anti-CD25 (1:10, Biolegend). Subsequently cells were fixed (3.7% PFA), and permeabilized with perm buffer (0.3% Triton X-100 in 1% BSA Solution). After fixation and permeabilization cells were stained using intracellular staining master mix containing anti-FoxP3 (1:20, Biolegend), anti-IKZF2 (1:20, Biolegend), and anti-IKZF1 (1:20, Biolegend). Cells were analyzed using an Thermo Attune NxT Flow cytometer (ThermoFisher), and results were graphed using GraphPad Prism. The gating strategy is illustrated in Supplementary Fig. 4b.

IL-2 production in Jurkat cells

Jurkat cells were plated at 2 × 10⁵ cells/mL in 96-well round bottom plates in 50 mL volume. The compound was added at decreasing dilution from 1000 nM down 10-fold, 4-points, in triplicates. After 18 h Jurkat cells were stimulated with anti-human CD3/CD28 beads at bead-to-cell ratio of 1:4 for 24 h. After 24 hours, supernatants were collected and concentration of IL-2 was measured using Human IL-2 ELISA Kit (Cat. #DY202, R&D). Data were normalized to the DMSO-treated wells.

Treg isolation and expansion

Frozen PBMCs from healthy donors was purchased from Miaoshun Biotechnology Co. (Shanghai) in accordance with their respective IRBs. Informed consent was obtained from all participants and all ethical regulations were respected. Treg cells were isolated from PBMCs using Human CD4⁺CD127^lowCD25⁺ Regulatory T Cell Isolation Kit (Stemcell Cat.#18063) according to manufacturer’s instructions. Isolated Tregs were expanded for 8-14 days in the presence of DKY709 or DMSO, using T-cell activator beads (ThermoFisher, cat.#11131D) at a bead-to-cell ratio of 3:1, in the presence of 10 ng/mL rhIL-2 (Peprotech Cat.#200-02-100).

Treg suppression assay

CD3⁺ T-Cells were isolated from donor matched PBMCs using Human T Cell Isolation Kit (Stemcell Cat.#17951). Isolated CD3⁺ T were stained with CellTrace Violate Dye as per manufacturer instruction (ThermoFisher Cat.#C34571). Expanded Treg cells were dispensed in co-culture with CellTrace Violate-labeled CD3 + T-Cells at various Treg:T cell ratios in the presence of T-cell activator beads (ThermoFisher, cat.#11131D) at the bead-to-cell ratio of 1:6. After 5 days of incubation, proliferation of CD8⁺ effector T cell (PE anti-human CD8 Ab, Biolegend Cat.#344706) was assessed by analyzing CFSE dye dilution using flow cytometry. Analysis was performed using an Thermo Attune NxT Flow cytometer (ThermoFisher), and results were graphed using GraphPad Prism. Teff cells (CD8⁺) that had proliferated during the co-culture were identified as having diluted CellTrace Violate and data were plotted as the proportion of CellTrace Violate low, proliferated cells in the final culture. Gating strategies are illustrated in Supplementary Fig. 4c.

Metabolic stability studies in liver microsomes

Pooled human, rat and mouse microsomes (10 μL aliquot) were prepared and stored at −80 °C prior to use. Master-mix containing microsome, phosphate buffer and test compound solution was made as follows: (1) 10 μL of microsome (20 mg/mL) was diluted with 330 μL 0.1 M Phosphate buffer (3.3 mM MgCl₂); (2) about 3.3 mg of NADPH was dissolved in 200 μL of 0.1 M phosphate buffer (3.3 mM MgCl₂); (3) 40 μL of 10 μM PBS solution of tested compound, was added to microsome; (4) the master solution was pre-warmed at 37 °C for 5 min. NADPH (20 μL) was added to the above master solution to initiate the reaction. The final concentration of tested compounds in the reaction system was 1 μM. Aliquot of 40 μL was pipetted from the reaction solution and stopped by the addition of 160 μL cold MeCN containing 25 nM of CE302 as an internal standard at the designated time points (0, 5, 10, 15, 30, 45, and 60 min). The incubation solution was vortexed-mixed (60 g/10 min) and centrifuged at 1100 g for 10 min to precipitate proteins. The supernatant was collected and used for the LC/MS/MS analysis. The natural log peak area ratio (compound peak area/internal standard peak area) was plotted against time and the gradient of the line was determined.

Permeability and efflux ratio determination in Caco-2 cells

The Caco-2 permeability evaluation of selected compounds was conducted in Shanghai Medicilon Inc. Shanghai, 201200, China. Caco-2 cell were seeded onto polyethylene membranes (PET) in 96-well Falcon insert systems at 2 × 105 cells/ cm² and cultured for 21-28 days for confluent cell monolayer formation. The cell culture media was changed every 3-4 days. Test compounds were diluted with the transport buffer (HBSS or HBSS with BSA) from a 10 mM stock solution to a concentration of 10 µM and applied to the apical or basolateral side of the cell monolayer. Permeation of the test compounds from A to B direction or B to A direction was determined in duplicate over a 120 min incubation at 37 °C and 5% CO₂ with a relative humidity of 95%. In addition, the efflux ratio of each compound was also determined. Test and reference compounds were quantified by LC-MS/MS analysis based on the peak area ratio of analyte/IS. The apparent permeability coefficient P_app (cm/s) was calculated using the equation: P_app = (dC_r/dt) × V_r / (A × C₀), where dC_r/dt is the cumulative concentration of compound in the receiver chamber as a function of time (S); V_r is the solution volume in the receiver chamber (0.1 mL on the apical side, 0.25 mL on the basolateral side); A is the surface area for the transport, i.e., 0.0804 cm² for the area of the monolayer; C₀ is the initial concentration in the donor chamber. The efflux ratio was calculated using the equation: Efflux Ratio = P_app (BA) / P_app (AB). LC/MS/MS Condition is as follows: detection method, LC-MS/MS-20 (TQ-6500 + ); matrix, HBSS; internal standard, Tolbutamide; MS conditions, positive ion ESI for Atenolol & Propranolol & tested compound and negative ion ESI for Digoxin & Tolbutamide; mobile phase, A = 0.1% FA in H₂O, B = 0.1% FA in MeCN; column, ACQUITY UPLC HSS T3 1.8 μm (50 mm × 2.10 mm).

hERG channel inhibition assay

The hERG inhibition data shown in Table 1 were obtained in Medicilon Preclinical Research LLC, Shanghai, China. The in vitro effects of tested compounds on ionic currents in voltage-clamped human embryonic kidney cells (HEK-293) that stably express the human ether-à-go-go-related gene (hERG) were determined. Five concentrations of compounds (0.3, 1, 3, 10 and 30 μM) were evaluated at room temperature.

The hERG Channel Inhibition data shown in Table 2 were obtained in Pharmaron, Beijing, China. The potential inhibitory effect of test articles on human Ether-à-go-go related gene (hERG) channel were evaluated by using manual patch-clamp system. HEK293 cell line stably transfected with hERG gene was employed in this study. Five concentrations of compounds (1, 3, 10, 30, 100 μM) were evaluated.

PK studies in rats, mice and monkeys

Pharmacokinetic (PK) studies were performed in Shanghai Medicilon Inc. Shanghai, 201200, China. All animal PK studies were performed in compliance with institutional ethical guidelines.

For the PK studies in rats and mice, male Sprague Dawley (SD) rats and male ICR mice were purchased from Sino-British SIPPR/BK Lab Animal Ltd., Shanghai, China. One group of three mice or rats was dosed intravenously (IV) with a dose level of 1 mg/kg, and a second group of three mice or rats was dosed orally with a dose level of 3 mg/kg. The animals were fasted prior to oral administration and food supply to the animals dosed orally were resumed 4 hours post-administration. The drug solution was freshly prepared before administration. For the IV route, each compound was formulated in 100% PEG200 or 5% DMSO + 10% Solutol+85% Saline as a clear solution, a dosage volume of 5 mL/kg, and a theoretical concentration of 0.2 mg/mL. For the oral route, each was formulated in 100% PEG200 or 5% DMSO + 10% Solutol + 85% Saline as a clear solution, a dosage volume of 10 mL/kg, and a theoretical concentration of 0.3 mg/mL. Blood samples were collected at the following time points: 5 min, 15 min, 30 min, 1 h, 2 h, 4 h, 6 h, 8 h and 24 h post dose administration. 200 µL (for rat) or 30 µL (for mouse) of blood was collected and the samples were be placed in tubes containing heparin sodium and stored on ice. The samples were centrifuged at ~6800 g for 6 min at 2–8 °C and the resulting plasma was transferred to appropriately labeled tubes within 1 h of blood collection/centrifugation (1500 g) then stored frozen at −80 °C.

The monkey PK studies of PVTX-405 were conducted in male Non-Naïve Cynomolgus Monkey (Medicilon Ltd., Shanghai China, #999M-014). One group of three monkeys was dosed intravenously (IV) with a dose level of 1 mg/kg, and a second group of monkeys was dosed orally with a dose level of 3 mg/kg. The drug solution was freshly prepared before administration. For the IV route, PVTX-405 was formulated in 5% DMSO + 10% Solutol+85% Saline as a clear solution, a dosage volume of 2 mL/kg, and a theoretical concentration of 0.5 mg/mL. For the oral route, PVTX-405 was formulated in 5% DMSO + 10% Solutol + 85% Saline as a clear solution, a dosage volume of 5 mL/kg, and a theoretical concentration of 0.6 mg/mL. Blood samples were collected at the following time points: 5 min, 15 min, 30 min, 1 h, 2 h, 4 h, 6 h, 8 h and 24 h post dose administration. ~1 mL of Blood was collected and the samples were be placed in tubes containing heparin sodium and stored on ice. The blood samples will be centrifuged at approximately 2200 g for 10 min at 2–8 °C and the resulting plasma will be transferred to appropriately labeled tubes within 1 h of blood collection/centrifugation and stored frozen at approximately −70 °C.

Method development and biological sample analysis for the test articles (Sodium heparin anticoagulant) were performed by Testing Facility by means of LC-MS/MS. The analytical results were confirmed using quality control samples for intra-assay variation. The accuracy of >66.7% of the quality control samples was between 80-120% of the known value(s). Standard set of parameters including T_1/2 (elimination half-life), AUC_(0-t) (area-under-the-curve), V_ss (volume of distribution at steady state), Cl (clearance), C_max (maximum drug concentration), F (oral bioavailability) were calculated using Phoenix WinNonlin 7.0 (Pharsight, USA) by the Study Director.

Plasma stability studies

The plasma stability of PVTX-405 was assessed using the following protocol. (1) Preheat 100 mM K-buffer with 5 mM MgCl₂ pH 7.4. (2) Plasma preparation: Thaw frozen plasma by placing at 37 °C quickly. (3) Test compounds and reference compounds spiking solution: For 0.5 mM compound spiking solution A, add 10 µL of 10 mM compound stock solution to 190 µL MeCN; for 0.01 mM compound spiking solution B, add 20 µL of spiking solution A to 980 µL of 0.1 M K/Mg-buffer. (4) Pre-warm the plasma and spiking solution B (see step 3.2) at 37 °C for 5 min. (5) Add 90 µL of pre-warmed plasma into the wells designated for all the time points (0, 5, 15, 30, 60,120 min). (6) For 0 min, add 10 µL of spiking solution B and then add 400 µL of MeCN containing IS to the wells of 0 min plate. (7) For other time points, add 10 µL of pre-warmed spiking solution B into the wells designated for the time points (5, 15, 30, 60,120 min), and start timing. (8) At 5, 15, 30, 60,120 min, add 400 µL of MeCN containing IS to the wells of corresponding plates, respectively, to stop the reaction. (9) After quenching, shake the plates for 5 min (30 g) and store at −20 °C (if necessary) until analyze. (10) Before LC/MS/MS analysis, thaw the samples at RT and centrifuge at 3200 g for 15 min. (11) Transfer 100 μL of the supernatant from each well into a 96-well sample plate containing 100 μL of pure water for LC/MS analysis.

Plasma protein binding

The plasma protein binding of PVTX-405 was assessed using the following protocol. (1) Preheat 0.1 M sodium phosphate and 0.15 M NaCl buffer, pH 7.4. (2) Plasma preparation: Thaw frozen plasma by placing at 37 °C; Centrifuge plasma at 13000 g for 5 min to remove clots, pipette and pool the supernatant as the plasma to be used in the experiment. (3) Membrane pretreated. Dialysis membrane strips were soaked in distilled water for an hour. Add 20% by volume ethanol and soak for a further 20 min. The membrane strips were then rinsed in distilled water 3 times before use. (4) Spiking solutions of test and reference compounds. Solution A (0.5 mM): Add 5 µL of 10 mM stock solution into 95 µL of DMSO. Solution B (0.02 mM): Add 8 µL of solution A into 192 µL of 0.1 M sodium phosphate buffer. (5) Preparation of test and reference compounds in plasma. Preload a 96-well plate with 380 µL aliquots of plasma in the wells designated for plasma, respectively. Spike 20 µL of Solution C (0.02 mM of test and reference compounds) into the pre-loaded plasma in the 96-well plate. The final test concentration is 1 µM. (6) Dialysis Sample Loading. Preparing plasma against buffer system (duplicate): Apply aliquots of 100 µL of blank dialysis buffer to the receiver side of dialysis chambers. Then apply aliquots of 100 µL of the plasma spiked with test and reference compounds to the donor side of the dialysis chambers (always add the blank buffer to the receiver first, clearly mark the buffer and plasma chamber holes to avoid cross contamination). Aliquot 25 µL of the plasma spiked with test and reference compounds into a 96-well sample preparation plate as T0 samples and store the sample plate in a freezer (−20 °C). Mix the plasma samples (from step 5) with same volume of blank buffer (25:25, v/v). Quench the samples with 200 μL of acetonitrile containing internal standard (IS). Vortex the samples at 30 g for 10 min and cover the plate and store it in a freezer (−20 °C). Cover the dialysis block with a plastic lid and place the entire apparatus in a shaker for 5 h at 37 °C. Preparing dialyzed samples after 5-hour incubation: Aliquot 25 µL from both the donor sides and receiver sides of the dialysis apparatus into new sample preparation plates and mix the aliquots with same volume of opposite matrixes (blank buffer to Plasma and vice versa); Quench the samples with 200 μL MeCN containing internal standard (IS). Vortex all the samples (from 0 h and 5 h) at 30 g for 10 min followed by centrifugation at 3200 g for 15 min. Transfer 100 μL of the supernatant from each well into a 96-well sample plate containing 100 μL of ultra-pure water for LC/MS analysis.

Cytochrome P450 (CYP) enzyme inhibition assay

The CYP inhibition of PVTX-405 was profiling in Shanghai Medicilon Inc. Shanghai, 201200, China. The protocol for the experiment is as follows. (1) Preheat 0.1 M K-Buffer with 5 mM MgCl₂ (K/Mg-buffer), pH 7.4; (2) Prepare serial dilution for test compound and reference inhibitors in a 96-well plate: (a) Transfer 8 μL of 10 mM test compounds to 12 μL of MeCN; (b) Prepare individual inhibitor spiking solution for CYPs 1A2, 2B6, 2C19, 2C9, 2D6 and 3A4 from 8 μL of DMSO stock to 12 μL of MeCN; (c) Perform 1:2 serial dilutions in DMSO:MeCN mixture (v/v: 40:60). (3) Prepare NADPH cofactor (66.7 mg NADPH in 10 mL 0.1 M K/Mg-buffer, pH7.4). (4) Prepare substrate (2 mL for each enzyme isoform) as indicated (add HLM where required on ice). (5) Prepare 0.2 mg/mL HLM solution (10 µL of 20 mg/mL to 990 µL of 0.1 M K/Mg-buffer). (6) Add 400 μL of 0.2 mg/mL HLM to the assay wells and then add 2 μL of test compound set (serially diluted, see step 2.1) into the designated wells. (7) Add 200 μL of 0.2 mg/mL HLM to the assay wells and then add 1 μL of serially diluted reference inhibitor solution (see step 2a and 2b) into the designated wells. (8) Add following solutions (in duplicate) in a 96-well assay plate on ice: (a) Add 30 μL of test compound and reference compound in 0.2 mg/mL HLM solution (see step 6 and 7); (b) Add 15 μL of substrate solution (see step 4). (9) Pre-incubate the 96-well assay plate and NADPH solution at 37 °C for 5 min. (10) Add 15 µL of pre-warmed 8 mM NADPH solution to into the assay plates to initiate the reaction (See step 3). (11) Incubate the assay plate at 37 °C. 5 min for CYP3A4, 10 min for CYP1A2, CYP2B6 and CYP2C9, 20 min for CYP2D6, and 45 min for CYP2C19. (12) Stop the reaction by adding 180 µL of acetonitrile containing IS. (13) After quenching, shake the plates for 10 min (30 g/min) and then centrifuge at 3200 g for 15 min. (14) Transfer 80 μL of the supernatant from each well into a 96-well sample plate containing 120 μL of ultra-pure water for LC/MS analysis. The LC-MS analysis was conducted using LC-MS/MS-20(TQ-6500+) ACQUITY UPLC HSS T3 1.8 μm (50 mm × 2.10 mm) or LC-MS/MS-11(8050) ACQUITY UPLC BEH C18 1.7 μm (50 mm × 2.10 mm). Perform curve-fitting to calculate IC₅₀ using a Sigmoidal (non-linear) dose-response model (GraphPad Prism 5.0 or Xlfit model 205) based on data calculation using the formula below: Y = Bottom + (Top-Bottom)/(1 + 10^((LogEC₅₀ - X) × HillSlope)), where X is the logarithm of concentration, Y is the response starting from Bottom to Top in a sigmoid shape in response to inhibitor concentration from high to low. The results generated for the reference compounds were consistent with the historic values.

PK/PD and efficacy studies in mice

All in vivo PK/PD and efficacy studies were performed under animal protocol (PRO00011174 and PRO00009463) approved by the Institutional Animal Care & Use Committee (IACUC) of the University of Michigan, in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health.

Crbn^I391V mice^39,44,45 breeding pairs were purchased from the Jackson Laboratory (Strain # 032487) under the Leap License Agreement. Crbn^I391V mice for all the studies described below were obtained from the University of Michigan breeding colony.

PK/PD studies

The Crbn^I391V mice were administered with vehicle control or PVTX-405 via oral gavage using 100% PEG200 as the dosing vehicle (dosing volume/mouse weight = 10 µL/g). Mice were sacrificed at indicated time-points, and blood samples and some organ tissues (spleen and thymus) were harvested for analysis. At each time point, 4 mice were euthanized with CO₂ and 250-300 µL of blood were collected by cardiac puncture. The blood samples were put into 1.5 mL microfuge tubes containing Heparin sodium and placed on wet ice, and then centrifuged at 20,000 g for 10 min. A minimum of 100 µL of blood plasma was collected from the upper layer, leaving the blood cells behind in the microfuge tube. The plasma was transferred into a fresh 1.5 mL microfuge tube and kept frozen at −80 ^oC for drug concentration analysis. The tissue samples from each mouse were divided into two parts. One part was immediately frozen in liquid nitrogen, ground into fine powder, placed on dry ice and stored in −80 ^oC for Western blot analysis. Western blots were performed as detailed in the previous section. Another part was placed in tared Precellys® 2 mL Hard Tissue tubes with Homogenizing Ceramic Beads 16859 (Cayman Chem), weighed, snap frozen in LN2 and stored at −80 ^oC for drug concentration analysis.

To prepare organ tissue samples for LC-MS analysis, mixed ultrapure water, and MeCN solution (4:1) were added to the defrosted tissue samples 5:1, v/w, in order to facilitate homogenization with a Precellys evolution homogenizer at 4 °C. The homogenized tissues solution was denatured using cold MeCN (1:3, v/v) with vortex and centrifuged at 15,000 g, 4 °C for 10 min. Following protein precipitation, the final supernatants were collected for LC-MS analysis.

To determine drug concentrations in plasma and tissue samples, an LC-MS/MS method was developed and validated. This method consisted of a Shimadzu HPLC system, and chromatographic separation of a test compound was achieved using a Waters column (ACQUITY UPLC BEH C18 1.7 µm, 2.1 × 50 mm). An AB Sciex QTrap 5500 mass spectrometer equipped with an electrospray ionization source (Applied Biosystems, Toronto, Canada) in the positive-ion multiple reaction monitoring mode was used for detection. For example, the precursor/product ion transitions were monitored at m/z 526.2/171 for PVTX-405 and 766.3/328.2 for an internal standard, respectively, in the positive electrospray ionization mode. The mobile phases used on HPLC were 0.1% formic acid in purified water (A) and 0.1% formic acid in MeCN (B). The flow rate was set at 0.4-0.6 mL/min and injection volume was 5–10 µL.

Efficacy studies

MC38 cells were purchased from ATCC. Crbn^I391V mice were implanted with 5 × 10⁶ MC38 cells subcutaneously to establish tumor xenograft. Tumor-bearing animals were randomized into treatment and vehicle control groups once the average tumor volume reached approximately ~80 mm³, and treatment was subsequently initiated. Each group was comprised of 10 mice. Crbn^I391V mice were housed in a controlled environment with a 12 h light/dark cycle, with ambient temperature maintained between approximately 20–22 °C and relative humidity between 30–70%.

For the single-agent efficacy study, animals were treated daily either with 100% PEG200 as a vehicle control or with 30 mg/kg PVTX-405 administered by oral gavage for up to 3 weeks. For the anti-PD1 combination study, animals were treated with either 100% PEG200 as a vehicle control, or 30 mg/kg PVTX-405 by daily oral gavage, or 1 mg/kg isotype control antibody (BioXcell, Cat. #BE0089) by intraperitoneal (i.p.) injection once a week, or 1 mg/kg anti-PD1 antibody (BioXcell, Cat. #BE0146) by i.p. injection once a week, or a combination of PVTX-405 and the isotype control antibody, or a combination of PVTX-405 and the anti-PD1 antibody for up to 3 weeks. For the anti-LAG3 combination study, animals were treated with either 100% PEG200 as a vehicle control, or 30 mg/kg PVTX-405 by daily oral gavage, or 10 mg/kg isotype control antibody (BioXcell, Cat. #BE0088) by i.p. injection once a week, or 10 mg/kg anti-LAG3 antibody (BioXcell, Cat. #BE0174) by i.p. injection once a week, or a combination of PVTX-405 and the isotype control antibody, or a combination of PVTX-405 and the anti-LAG3 antibody for up to 3 weeks.

For single agent efficacy study, mice were sacrificed when tumor volume exceeds 2000 mm³ or at the end of the treatment (whichever occurs first). Tumor sizes and animal weights were measured 2–3 times per week. Tumor volume (mm³) was calculated as (length × width²)/2. Tumor growth inhibition was calculated as TGI (%) = [1-(V_t – V₀’)/(V_c – V₀)] × 100, where V_c and V_t are the mean tumor volume of the vehicle control and treated group at the end of treatment, respectively, and V₀ and V₀’ are the mean tumor volume of the vehicle control and treated group at the start, respectively. The tumor volumes at the end of treatment were statistically analyzed using t test. For combination study, animals were treated for up to 3 weeks. Mice were sacrificed when tumor volume exceeds 2000 mm³ or at the end of study, which was 50 days after the start of the treatment (whichever occurred first) to estimate the survival of the animals. Tumor sizes and animal weights were measured 2–3 times per week. Tumor volume (mm³) was calculated as (length × width²)/2. Statistical analysis for survival was performed using log-rank test.

Structural biology

CRBN thalidomide binding domain (CRBN-TBD, Uniprot accession number Q96SW2, residues 318-426) was cloned into a pET28a vector with a N-terminal 6His-GST tag containing a PreScission protease site. The CRBN-TBD sequence was cloned with several point mutations to help aid crystallization (C322S, C343S, C366S). CRBN-TBD was expressed in Escherichia coli BL21 (Gold) cells using LB media supplemented with 50 µM ZnCl₂. Cells were induced with 0.2 mM IPTG at an optical density OD600 of 0.8 and grown overnight at 16 °C. Cells were pelleted by centrifugation at 5000 g for 5 min at 4 °C and stored at −20 °C. Cell pellets were then resuspended in a lysis buffer containing 20 mM Tris, pH 8.0, 500 mM NaCl, 0.1 mM TCEP, 5 mM ZnCl₂ and 1 mM PMSF at a ratio of 5 mL/g cell pellet. Resuspend cells were sonicated on ice for 15 min of on-time (5 s on, 5 s off) and centrifuged at 13,000 g for 30 min at 4 °C. Clarified lysate was purified with Ni-NTA affinity resin (Qiagen) that was equilibrated with lysis buffer and then washed and eluted with lysis buffer containing 20 mM, 50 mM, and 250 mM imidazole. The N-terminal 6His-GST was cleaved with PreScission protease and dialyzed overnight in a buffer containing 20 mM Tris, pH 8.0, 500 mM NaCl, 0.1 mM TCEP, and 5 mM ZnCl₂. Cleaved 6His-GST tag and protease were then removed with an additional Ni-NTA affinity purification step and the flow through was collected and diluted to lower the NaCl concentration to 50 mM. The CRBN-TBD protein was then purified with a HiTrap SP HP cation exchange column (Cytiva) equilibrated with buffer containing 20 mM HEPES, pH 7.5, 50 mM NaCl, 1 mM TCEP, and 5 µM ZnCl₂. The HiTrap SP column was then eluted with a gradient of buffer containing 20 mM HEPES, pH 7.5, 1 M NaCl, 1 mM TCEP, and 5 µM ZnCl₂. Peak fractions containing the CRBN-TBD protein were collected and further purified using a size exclusion chromatography column (Cytiva, Superdex 75 160/600) equilibrated with buffer containing 50 mM Hepes, pH 7.4, 200 mM NaCl, and 0.1 mM TCEP. Peak fractions containing CRBN-TBD were pooled and then concentrated to 39.3 mg/mL before flash freezing 25 µL aliquots in liquid nitrogen and storing at −80 °C.

IKZF2 zinc finger 2 domain (IKZF2-ZF2, Uniprot accession number Q9UKS7, residues 136-164) was cloned into a pET21a vector containing an N-terminal 6His-GST tag containing a TEV protease site. IKZF2-ZF2 was expressed in Escherichia coli BL21-CodonPlus(DE3)-RIL cells using LB media supplemented with 50 µM ZnCl₂. Cells were induced with 0.2 mM IPTG at an optical density OD600 of 0.8 and grown overnight at16 °C. Cells were pelleted by centrifugation at 5000 g for 5 min at 4 ^oC and stored at −20 °C. Cells were then resuspended in lysis buffer containing 50 mM Sodium Phosphate, pH 8.0, 500 mM NaCl, 1 mM TCEP, 5 μM ZnCl₂, 50 µL Benzonase nuclease, and 1 mM PMSF at a ratio of 5 mL/g of cell pellet. Resuspended cells were sonicated on ice for 15 min of on-time (5 s on, 5 s off) and centrifuged at 13,000 g for 30 min at 4 °C. Clarified lysate was purified with Ni-NTA affinity resin (Qiagen) that was equilibrated with lysis buffer and then washed and eluted with lysis buffer containing 20 mM, 50 mM, and 250 mM imidazole. The N-terminal 6His-GST tag was cleaved with TEV protease and dialyzed overnight in a buffer containing 50 mM Sodium Phosphate, pH 8.0, 500 mM NaCl, 1 mM TCEP, and 5 μM ZnCl₂. The 6His-GST tag and protease were then removed using a second Ni-NTA affinity purification step. Ni-NTA affinity resin (Qiagen) was equilibrated with dialysis buffer and then washed with dialysis buffer containing 20 mM and 250 mM imidazole. The flow through was collected and then diluted to reduce the NaCl concentration to 50 mM. The sample was then purified using a HiTrap SP HP cation exchange column (Cytiva) equilibrated with buffer containing 50 mM Sodium Phosphate, pH 7.4, 50 mM NaCl, 1 mM TCEP and 5 μM ZnCl₂. The cation exchange column was then eluted with a gradient using a buffer containing 50 mM Sodium Phosphate, pH 7.4, 1 M NaCl, 1 mM TCEP and 5 μM ZnCl₂. Peak fractions containing IKZF2-ZF2 were then purified using a size exclusion chromatography column (Cytiva, Superdex 75 10/300) equilibrated with buffer containing 25 mM Hepes, pH 7.4, 250 mM NaCl, 0.5 mM TCEP, 5 μM ZnCl₂, and 5% glycerol. Peak fractions were collected and then concentrated 3.15 mg/mL before snap freezing 25 µL aliquots in liquid nitrogen and storing at −80 °C.

Ternary complex was prepared for crystallization by mixing CRBN-TBD, IKZF2-ZF2, and compound 5 to a final concentration of 0.4 mM, 0.4 mM, and 0.8 mM, respectively. Ternary complex mixture was incubated on ice for 1 h before setting up crystallization screening at 18 °C using hanging-drop vapor diffusion using a 1:1 ratio of mother liquor solution containing 0.4 M Magnesium chloride, 0.1 M Tris, pH 7.5, and 27% PEG 4000. Crystals were optimized using seeding and then frozen using liquid nitrogen. Diffraction data was collected from a single crystal at the Diamond Beamline I03. The CRBN:IKZF2:compound 5 structure was solved by molecular replacement and refined using CCP4. Crystallography statistics are summarized in Supplemental Fig. 3.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.