Covalent Proteins as Targeted Radionuclide Therapies Enhance Antitumor Effects

Abstract

Molecularly targeted radionuclide therapies (TRTs) struggle with balancing efficacy and safety, as current strategies to increase tumor absorption often alter drug pharmacokinetics to prolong circulation and normal tissue irradiation. Here we report the first covalent protein TRT, which, through reacting with the target irreversibly, increases radioactive dose to the tumor without altering the drug's pharmacokinetic profile or normal tissue biodistribution. Through genetic code expansion, we engineered a latent bioreactive amino acid into a nanobody, which binds to its target protein and forms a covalent linkage via the proximity-enabled reactivity, cross-linking the target irreversibly in vitro, on cancer cells, and on tumors in vivo. The radiolabeled covalent nanobody markedly increases radioisotope levels in tumors and extends tumor residence time while maintaining rapid systemic clearance. Furthermore, the covalent nanobody conjugated to the α-emitter actinium-225 inhibits tumor growth more effectively than the noncovalent nanobody without causing tissue toxicity. Shifting the protein-based TRT from noncovalent to covalent mode, this chemical strategy improves tumor responses to TRTs and can be readily scaled to diverse protein radiopharmaceuticals engaging broad tumor targets.

Figure 1

Covalent protein radiopharmaceuticals to enhance efficacy and safety for TRT. A schematic comparison of the noncovalent WT Nb_HER2 (A) and the covalent Nb_HER2 (B) in targeted delivery of radionuclide to HER2-expressing cancer cells. The noncovalent Nb_HER2 binds HER2 reversibly allowing dissociation. In contrast, when the covalent Nb_HER2 binds to HER2, the latent bioreactive Uaa FSY reacts with Lys through proximity-enabled SuFEx reaction, resulting in irreversible cross-linking of Nb_HER2 with HER2 and persistent tumoral retention of the attached radionuclide.

Figure 2

Genetically encoding FSY in Nb_HER2 to covalently cross-link HER2 irreversibly in vitro. (A) Crystal structure of Nb_HER2 bound to HER2 ECD (PDB: 5MY6), showing the FSY incorporation site (D54) and the proximal target residue (K150) in HER2. (B) Western blot analysis of Nb_HER2(FSY) production in E. coli with and without 1 mM FSY in growth media. A His6x tag was appended at the C-terminus of Nb_HER2 for detection. (C) Mass spectrum of the intact Nb_HER2(FSY) protein confirming FSY incorporation at position 54 in high fidelity. (D) Nb_HER2(FSY), but not Nb_HER2(WT), cross-linked with HER2 ECD in vitro. Indicated proteins were incubated at 37 °C for 4 h followed with Western blot analysis. (E) Tandem mass spectrum of Nb_HER2(FSY) incubation with HER2 ECD confirmed that FSY (represented by U) of Nb_HER2(FSY) cross-linked with Lys150 of HER2 as designed. (F) Cross-linking of Nb_HER2(FSY) to HER2 ECD occurred efficiently at 10 min and increased with time. (G) Kinetics of Nb_HER2(FSY) cross-linking with HER2 ECD. Nb_HER2(FSY) concentrations in (F) were measured with densitometry and 1/[Nb_HER2(FSY)] was plotted against time. Linear regression of the data yielded a second-order rate constant of 34154 ± 1921 M^–1min^–1 (mean ± s.d.). Error bars represent s.d., n = 3 independent experiments.

Figure 3

Nb_HER2(FSY) covalently cross-linked native HER2 on cancer cells and on tumor in vivo. (A) Nb_HER2(FSY) covalently cross-linked HER2 on NCI-N87 cell surface. Nb_HER2 proteins were incubated with NCI-N87 cells for 3 h followed with Western blot analysis. (B) Cross-linking of Nb_HER2(FSY) with cancer cells were HER2 specific. Cross-linking occurred only on NCI-N87 and SK-OV-3 cells, which have detectable HER2 expression. (C) Nb_HER2(FSY) covalently cross-linked HER2 on NCI-N87 tumor in vivo. Nb_HER2(FSY) or Nb_HER2(WT)was injected into mice xenografted with HER2-expressing NCI-N87 tumor. After 6 h postinjection, the tumor was excised and homogenized, followed with Western blot analysis.

Figure 4

Radiolabeled covalent nanobody ¹²⁴I-Nb_HER2(FSY) prolonged tumor retention, increased tumor accumulation and exhibited low background in mice. (A) Schematic procedures to radiolabel WT and covalent Nb_HER2 with ¹²⁴I by IODO-GEN. Tyrosine is usually labeled at the ortho position with mono- or di-iodination. (B) Iodine labeling did not impair Nb_HER2(FSY) cross-linking with HER2. The cold NaI labeled product I-Nb_HER2(FSY) or the unlabeled Nb_HER2(FSY) was incubated with HER2 ECD for cross-linking, followed with Western blot analysis. (C) The covalent ¹²⁴I-Nb_HER2(FSY) enabled specific and sustained tumor accumulation of ¹²⁴I. Tumors were clearly detectable 24–72 h postinjection for ¹²⁴I-Nb_HER2(FSY) but not ¹²⁴I-Nb_HER2(WT). Representative decay-corrected PET images of mice xenografted with HER2-expressing NCI-N87 tumor and injected with either ¹²⁴I-Nb_HER2(WT) or ¹²⁴I-Nb_HER2(FSY) are shown. The transverse images of mice were taken at 3–72 h postinjection. Color bars indicate percent injected dose per gram (%ID/g). (D) The covalent ¹²⁴I-Nb_HER2(FSY) significantly enhanced tumor accumulation of ¹²⁴I than ¹²⁴I-Nb_HER2(WT). The standardized uptake value (SUV) of ¹²⁴I in tumor was quantified in percent injected dose per cm³ (%ID/cc) and plotted with postinjection time. The increase in tumor uptake by ¹²⁴I-Nb_HER2(FSY) over ¹²⁴I-Nb_HER2(WT) is highlighted in cyan. Error bars represent s.d.; n = 3 mice for ¹²⁴I-Nb_HER2(WT) injection; n = 4 mice for ¹²⁴I-Nb_HER2(FSY) injection; ns, not significant; ** p < 0.01; Student’s t test for statistical analysis. (E) The covalent ¹²⁴I-Nb_HER2(FSY) enabled clear imaging of tumor distinct from the background. 3D PET image reconstruction of mice 24–72 h postinjection of ¹²⁴I-Nb_HER2(WT) or ¹²⁴I-Nb_HER2(FSY) are shown. Color bars indicate %ID/g. n = 3 mice for ¹²⁴I-Nb_HER2(WT) injection; n = 4 mice for ¹²⁴I-Nb_HER2(FSY) injection.

Figure 5

α-Emitter labeled covalent ²²⁵Ac-Nb_HER2(FSY) inhibited tumor growth in mice without tissue toxicity. (A) Schematic procedures to radiolabel WT and covalent Nb_HER2 with ²²⁵Ac. (B) Mass spectrometric analyses confirming successful conjugation of Macropa-PEG₄ on Nb_HER2(WT) (top panel) and Nb_HER2(FSY) (bottom panel). (C) Western blot analysis confirming that Macropa-PEG₄ labeling did not impair Nb_HER2(FSY) cross-linking with HER2. Cross-linking of Macropa-PEG₄-Nb_HER2(FSY) to HER2 ECD occurred efficiently after 10 min incubation and increased with time, while no cross-linking was detected with Macropa-PEG₄-Nb_HER2(WT). (D) Experiment scheme for TRT of NCI-N87 tumor in mice. (E) Growth curves of engrafted NCI-N87 tumors indicate that ²²⁵Ac-Nb_HER2(FSY) inhibited tumor growth, while ²²⁵Ac-Nb_HER2(WT) did not. (F) Weight comparison of dissected tumors showing tumor weight reduction by ²²⁵Ac-Nb_HER2(FSY) treatment. (G) Mice body weight remained stable over the course of the therapy study. For panels E–G, error bars represent SEM; n = 8 mice for ²²⁵Ac-Nb_HER2(FSY) treatment group; n = 7 mice for ²²⁵Ac-Nb_HER2(WT) treatment group; n = 5 mice for vehicle saline control. ns, not significant; *p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001; Student’s t test for statistical analysis. (H) Representative microscopic images of hematoxylin and eosin stained liver, kidneys, heart, and bone marrow for both ²²⁵Ac-Nb_HER2(WT) and ²²⁵Ac-Nb_HER2(FSY) treatment groups. No abnormalities were detected in the tissues. Scale bar, 50 μm.