A ²¹¹At-labelled mGluR1 inhibitor induces cancer senescence to elicit long-lasting anti-tumor efficacy

Summary

Metabotropic glutamate receptor 1 (mGluR1), a key mediator of glutamatergic signaling, is frequently overexpressed in tumor cells and is an attractive drug target for most cancers. Here, we present a targeted radiopharmaceutical therapy strategy that antagonistically recognizes mGluR1 and eradicates mGluR1⁺ human tumors by harnessing a small-molecule alpha (α)-emitting radiopharmaceutical, ²¹¹At-AITM. A single dose of ²¹¹At-AITM (2.96 MBq) in mGluR1⁺ cancers exhibits long-lasting in vivo antitumor efficacy across seven subtypes of four of the most common tumors, namely, breast cancer, pancreatic cancer, melanoma, and colon cancers, with little toxicity. Moreover, complete regression of mGluR1⁺ breast cancer and pancreatic cancer is observed in approximate 50% of tumor-bearing mice. Mechanistically, the functions of ²¹¹At-AITM are uncovered in downregulating mGluR1 oncoprotein and inducing senescence of tumor cells with a reprogrammed senescence-associated secretory phenotype. Our findings suggest α-radiopharmaceutical therapy with ²¹¹At-AITM can be a useful strategy for mGluR1⁺ pan-cancers, regardless of their tissue of origin.

Graphical abstract

Highlights

• •Most cancer patients, across 32 different tumor types, aberrantly express mGluR1
• •α-Radiopharmaceutical ²¹¹At-AITM binds to mGluR1-expressing multiple human cancers
• •A single dose of ²¹¹At-AITM elicits durable antitumor effect in mGluR1⁺ pan-cancers
• •²¹¹At-AITM decreases mGluR1 and modulates cancer senescence with reprogrammed SASP

Xie et al. report a targeted radiopharmaceutical therapy strategy that antagonistically recognizes mGluR1, a key mediator of glutamatergic signaling, and eradicates mGluR1⁺ human tumors by harnessing an α-radiopharmaceutical, ²¹¹At-AITM. A single injection of ²¹¹At-AITM in mGluR1⁺ pan-cancers elicits long-lasting antitumor efficacy and even cures, regardless of their tissue origins.

Introduction

Metabolic reprogramming has been recognized as one of the 14 hallmarks of cancer.¹ Cancer cells alter their glucose and glutamine metabolism to maintain bioenergetics, redox status, cell signaling, and biosynthesis.^2,3,4,5 The discovery of the Warburg effect, increasing aerobic glycolysis, led to the extensive use of ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) as the main positron emission tomography (PET) tracer to map glucose metabolism for cancer detection, staging, and response monitoring in the clinic.^6,7,8 In addition to glucose metabolism, proliferating cancer cells rely on increased glutamine metabolism (glutaminolysis) to maintain the functioning of the tricarboxylic acid cycle and produce glutathione as a precursor for nucleotide and lipid synthesis.^2,9 Notably, glutamate is an immediate product of glutamine metabolism catalyzed by glutaminase.¹⁰ The central role of glutamine/glutamate metabolism in cancer bioenergetics and biosynthesis has reinforced the interest in key receptors of glutamatergic signaling.

Indeed, the overexpression of glutamate receptors, triggered by glutamate, has been detected in several cancer types and contributes to tumor cell growth via paracrine or autocrine signaling.^11,12,13 Among these tumor-associated glutamate receptors, metabotropic glutamate receptor 1 (protein: mGluR1; gene: GRM1), a seven-transmembrane domain G protein-coupled receptor,¹⁴ has been shown to be overexpressed in a variety of cancers (such as melanoma,¹¹ breast cancer,¹⁵ prostate cancer,¹⁶ chondromyxoid fibroma,¹⁷ and lung cancer¹⁸). mGluR1 is normally expressed in the CNS and is involved in memory and learning.¹⁹ However, its aberrant overexpression led to neoplastic transformation in vitro and tumorigenesis in vivo.²⁰ Binding of glutamate, the endogenous ligand of mGluR1, activates the mitogen-activated protein kinase (MAPK), phosphatidylinositol-3 kinase/Ak strain transforming (AKT), and AKT/mammalian target of rapamycin (mTOR)/hypoxia-inducible factor 1 pathways and is involved in numerous aspects of tumor cell growth, metastasis, angiogenesis, and survival.^21,22 The oncogenic effect of mGluR1 in mammary tumors was revealed when full-length wild-type GRM1 exogenously introduced in melanocytes and epithelial cells independently drove carcinogenesis with 100% penetrance.^11,23

Genetic and pharmacological interventions of mGluR1 have been performed and have resulted in clinically tested drug candidates.^22,24,25 Genetically, mGluR1 silencing using GRM1 small hairpin RNA diminished cell proliferation, migration, and invasion in triple-negative breast cancers (TNBCs).²³ Pharmacologically, the inhibition of mGluR1 activity using antagonists such as riluzole,²⁶ LY367385,²⁷ and BAY36-7620¹⁵ inhibited tumor growth, decreased blood vessel density, and induced DNA damage in mGluR1-overexpressing (mGluR1⁺) melanoma and TNBCs, but not in mGluR1-negative (mGluR1^–) cancers. A phase 2 clinical trial of riluzole for antitumor therapy in stage III and IV melanoma patients yielded short-term stable disease in 30%–46% of patients.²² Nevertheless, the single agent riluzole is unlikely to have a long-lasting benefit in melanoma patients. Combinatorial therapies with riluzole and γ-irradiation²⁸ or immunotherapy²⁹ have also been studied; however, these approaches failed to induce durable remission in any type of cancer.

As the majority of conventional mGluR1-targeting strategies are devoted to transiently modulating mGluR1 activity through antagonistic binding, the subsequent consequence is usually revealed to be moderate and susceptible to metabolic compensatory regulation. To overcome this bottleneck associated with mGluR1 antagonists, there is an urgent need for therapeutic strategies that can hinder the proliferation potential of mGluR1⁺ tumor cells via direct DNA double-strand breaks (DSBs) and reduce the oncoprotein expression to simultaneously block tumorigenesis. Tumor targeted alpha (α) radiopharmaceutical therapy (α-RPT), which refers to the targeted delivery of lethal α-particles to cancerous cells mediated by specific tumor-binding ligands, thus eliminating tumor cells by inducing DNA DSBs, has been demonstrated to be a remarkable therapeutic approach.^30,31,32 In this context, we envisioned that mGluR1 ligand-based RPT might represent a remedy for conventional molecular targeted therapy that fails in early clinical trials. For targeted α-RPT, whether the oncoprotein phenotype would be altered after radiation has been scarcely studied. Moreover, tumor cell responses against radiation exposure are usually cellular and genetic type and dosimetry dependent. To develop effective targeted RPT, it is essential to fully understand these issues, especially what will occur to the oncoprotein and the tumor cells after radiation and why it occurs.

Herein, we report a targeted α-RPT strategy for mGluR1⁺ human tumors using a small molecule-based α-emitting radiopharmaceutical (termed ²¹¹At-AITM) (Figures 1A and 1B).³³ We verified the therapeutic efficacy of ²¹¹At-AITM in a variety of human tumors with different tissue origins and distinct mGluR1 expression levels, including breast cancer, pancreatic cancer, melanoma, and colon cancer. Furthermore, we analyzed the phenotypic variations and exact cell fate of tumor cells after ²¹¹At-AITM exposure. We found functions of ²¹¹At-AITM in downregulating mGluR1 expression and inducing senescence of tumor cells with a reprogrammed senescence-associated secretory phenotype (SASP), except for its classical role in causing DNA DSBs. Consequently, the divergent outcomes showed that synergism in expanding the therapeutic index and eliciting a durable damage signal led to long-lasting benefit and even cure in multiple human solid cancers. Taken together, ²¹¹At-AITM based RPT represents a two-birds-one-stone strategy (²¹¹At-AITM is the stone, which causes phenotypic and genetic variations) for tackling mGluR1⁺ pan-cancers.

Figure 1.

Construction of a mGluR1 targeted ²¹¹At-AITM RPT strategy for the tumor-agnostic treatment of cancers

(A) Left: Chemical structure of ²¹¹At-AITM. Right: Radio-characteristics of ²¹¹At-AITM.

(B) ²¹¹At-AITM docking to the human mGluR1 seven-transmembrane crystal structure. See also Figure S2.

(C) The percentage of GRM1⁺ in human cancers from different tissues of origin. Data were collected from TCGA. CESC, cervical squamous cell carcinoma and endocervical adenocarcinoma. See also Figure S1.

(D) Validation of mGluR1 protein expression levels in seven representative cancer cell lines from four solid cancer types using immunofluorescence staining. Green, Alexa Fluor 488-labeled anti-mGluR1 antibody. Scale bar, 50 μm. Representative images are shown from three independent experiments.

(E) Quantification of GRM1 mRNA expression using qRT-PCR.

(F) Specific binding ability of ²¹¹At-AITM in a panel of human cancers with and without mGluR1 expression after 1, 3, and 21 h of treatment.

(G) Pearson correlation analysis between the level of GRM1 expression and the uptake of ²¹¹At-AITM.

(H) Biospecificity of ²¹¹At-AITM based on competitive co-incubation with the unlabeled mGluR1 antagonist FITM (10 μmol/L) for 3 h.

(I) Cellular localization of ²¹¹At-AITM action after 1, 3, and 21 h of treatment. (E, F, H, and I) Results are expressed as mean ± SEM, n = 3–4. All comparisons were performed using an unpaired two-tailed Student’s t test or two-way ANOVA followed by Tukey’s multiple comparison test.

Results

GRM1 is widely expressed in human cancers

Pan-cancer analysis of GRM1 expression in The Cancer Genome Atlas (TCGA) database data revealed that GRM1 expression was universally present in 27.03%–100% of human cancers across distinct histological subtypes (to date, 32 major tumor types covering 10,967 clinical samples have been identified) (Figures 1C, S1A, and S1B). A high percentage of GRM1 positivity was observed in most common solid cancers, such as breast cancer (96.13% [1,042/1,084]), pancreatic cancer (88.04% [162/184]), melanoma (97.32% [436/448] in skin melanoma and 87.50% [70/80] in uveal melanoma), and colon cancer (63.80% [379/594]). Mutation at Thr815^7.38 and Pro756^5.43 of GRM1 only represented 0.23% (1/442) of all human tumor samples, which were previously reported to reduce the affinity and potency of small-molecule mGluR1 antagonists, including 4-fluoro-N-[4-[6-(isopropylamino)pyrimidin-4-yl]-1,3-thiazol-2-yl]-N-methylbenzamide (FITM).¹⁴ Our findings indicate that most cancer patients, across 32 different tumor types, aberrantly express mGluR1 and could be good candidates for the mGluR1-targeting α-RPT.

In this study, we profiled and treated a panel of the most common solid cancers, including seven cancer cell lines of four cancer types (MDA-MB231 TNBC; MIA PaCa2 and PANC1 pancreatic cancers; A375, Bowes and A2058 melanomas; and DLD1 colon cancer). mGluR1 expression in these cell lines was assessed at the protein level using immunocytochemistry (Figure 1D) and confirmed at the mRNA level using quantitative real-time reverse transcriptase-polymerase chain reaction (qRT-PCR) (Figure 1E). Five tumor cell lines (MDA-MB231, MIA PaCa2, Bowes, A375, and DLD1) exhibited differential mGluR1 expression (Figures 1D and 1E). In comparison, mGluR1 expression was negligible in A2058 and PANC1 cell lines.

²¹¹At-AITM specifically binds to mGluR1-expressing multiple human cancer cells

Ligand receptor-mediated targeting is a facile and effective approach for radiopharmaceutical design. Based on this approach, we previously designed the α-emitting radiopharmaceutical ²¹¹At-AITM (Figures 1A and 1B),³³ which was derived from a high-affinity, small-molecule mGluR1 antagonist, FITM (half-maximal inhibitory concentration = 5.1 nM),³⁴ whereas the two compounds are distinguished by the type of halogen atoms. ²¹¹At-AITM retains high binding affinity and selectivity to murine and human mGluR1 as revealed by molecular dynamics simulations (Figures S2A–S2D). Notably, the bulky ²¹¹At atom can prevent ²¹¹At-AITM from entering the mGluR1-rich brain,^33,35,36,37 thus avoiding detrimental radiation to the CNS, which is a major concern associated with the clinical translation of mGluR1-targeted radiotherapies.

The binding characteristics of ²¹¹At-AITM were measured by incubating a subset of the human cancer cells with ²¹¹At-AITM for 1, 3, and 21 h. ²¹¹At-AITM displayed a high binding ability to mGluR1⁺ MDA-MB231, MIA PaCa2, Bowes, A375, and DLD1 cells after 1 h and maintained consistently high uptake at 3 and 21 h, while exhibiting only low uptake in mGluR1^– A2058 and PANC1 cells at all time points (Figure 1F). A close positive correlation was found between the GRM1 expression and the uptake of radioactivity at 1, 3, and 21 h after ²¹¹At-AITM incubation (Figure 1G). To clarify the specificity of ²¹¹At-AITM for mGluR1, we used excess unlabeled FITM (10 μmol/L), which competes for the same receptor, in these cancer cells. As expected, the uptake of ²¹¹At-AITM was significantly decreased in mGluR1⁺ MDA-MB231 and MIA PaCa2 cells but not in mGluR1^– A2058 and PANC1 cells (Figure 1H). The cellular localization of ²¹¹At-AITM action was also investigated³⁸; more than half of the radioactivity was anchored on the membrane of MDA-MB231 cells (68.71%–73.25% CD; the value is quantified for a percentage of cell-associated radioactivity dose, similarly hereinafter) and MIA PaCa2 cells (49.42%–61.45% CD), while the remaining radioactivity was rapidly internalized by cancer cells (Figure 1I). These results indicate that ²¹¹At-AITM specifically enters cancer cells expressing mGluR1 at high and stable concentrations and could act on the membrane, as well as in the interior of cancer cells, thus potentially providing double benefits for targeted α-RPT.

²¹¹At-AITM induces mGluR1 downregulation and locks cancer cells into senescence

Given the reported oncogenic role of mGluR1 when aberrantly expressed in cells, we next assessed the effects of ²¹¹At-AITM on mGluR1-expressing cancer cells. We treated mGluR1⁺ MDA-MB231 cells with different doses of ²¹¹At-AITM (Figures S3A and S3B), then chose a dose of 18.5 kBq/mL for subsequent cell culture experiments. MDA-MB231 and MIA PaCa2 cancer cells were incubated with ²¹¹At-AITM for 1 day (day 1) followed by 7 days of culture without ²¹¹At-AITM (days 2–7). Real-time cell survival rate measurements showed that ²¹¹At-AITM did not trigger immediate and massive cell death (Figures 2A and S3A). Most cancer cells remained viable on day 7 post-therapy (Figure 2A); however, cell proliferation was obviously impaired as the cell number doubling time was significantly prolonged in MDA-MB231 (vehicle vs. treatment = 38.22 ± 0.75 h vs. 71.18 ± 3.97 h; p < 0.0001) and MIA PaCa2 (vehicle vs. treatment = 27.54 ± 0.63 h vs. 37.31 ± 0.95 h; p < 0.0001) cells after ²¹¹At-AITM treatment. Cell cycle analysis showed that treatment with ²¹¹At-AITM in MDA-MB231 and MIA PaCa2 cancer cells induced a rapid shift from the G0/G1 phase and S phase to the G2/M phase and caused cell arrest in the G2/M phase over a long period of time after α-irradiation (Figures 2B, 2C, S3B–S3D). As shown in Figures 2D and 2G, a few multinucleated giant cells were observed, which is likely the result of mitotic catastrophe,³⁹ consistent with a small fraction of cells dying after ²¹¹At-AITM exposure (Figure 2A).

Figure 2.

²¹¹At-AITM downregulates mGluR1 and induces senescence in mGluR1⁺ cancer cells

(A) Survival rate of mGluR1⁺ MDA-MB231 and MIA PaCa2 cells. Cancer cells were treated with ²¹¹At-AITM or vehicle (PBS) for 1 day (day 1), incubated for 7 days without drug (days 2–7), and pretreated on day 0. Survival rate (to vehicle) of cells measured using Muse Cell Analyzer for up to 7 days.

(B and C) Cell cycle state distribution. Data show the results from cells in (A). See also Figures S3C and S3D.

(D) Representative fluorescence images of mGluR1 (green) and DNA (blue) in MDA-MB-231 and MIA PaCa2 cells. Scale bar, 50 μm.

(E and F) Quantification of mGluR1 expression.

(G and I) Representative fluorescence images and quantification of γH2AX (red) and DNA (blue) to assess DNA DSBs. Scale bar, 50 μm.

(J) Representative fluorescence images of SA-β-gal (green) to assess senescence. Scale bar, 50 μm.

(K and L) Percentages of cells that showed SA-β-gal positive activity.

(M and N) Pearson correlation analysis of the mGluR1 expression and the marker of DNA DSBs γH2AX or the marker of senescence SA-β-Gal in MDA-MB231 and MIA PaCa2 cells. (A–C, E, F, H, I, K, and L) Results are expressed as mean ± SEM, n = 4. All comparisons were performed using an unpaired two-tailed Student’s t test or two-way ANOVA followed by Tukey’s multiple comparison test. Asterisks indicate statistical significance (^∗p < 0.05, ^∗∗p < 0.01), ns indicates no significance. (D, G, and J) Representative images are shown from three independent experiments.

To assess why ²¹¹At-AITM triggered long-term growth arrest of cancer cells, we first evaluated the mGluR1 protein expression associated with oncogenic activity and confirmed the mRNA level. We observed a significant and stable reduction in mGluR1 protein expression in both MDA-MB231 and MIA PaCa2 cells after 1–7 days (Figures 2D–2F). Compared with the observed following vehicle treatment, the GRM1 mRNA expression was not detectable in either MDA-MB231(Figure S3E) or MIA PaCa2 cells (Figure S3F) after 2–7 days of treatment, although there was no significant GRM1 mRNA change between vehicle or treated cell samples at the 1-day timepoint. These results indicated that ²¹¹At-AITM treatment induced the mGluR1 downregulation not only at the protein level (Figures 2D–2F), but also at the mRNA level (Figures S3E and S3F) in mGluR1⁺ cancer cells.

Next, we assessed the protein expression associated with DNA damage after treatment with ²¹¹At-AITM. The results showed that the signal of γH2AX, a marker of DNA DSBs, was notable and not lost over time after α-irradiation, indicating that ²¹¹At-AITM caused clustered DNA lesions and persistent DNA damage response (DDR) activation in most MDA-MB231 and MIA PaCa2 cancer cells, which was not observed in vehicle-treated cells (Figures 2G and 2I). Then, we assessed senescence-associated beta-galactosidase (SA-β-gal), a biomarker of senescence, in the treated cancer cells. The SA-β-gal⁺ cells were enriched in MDA-MB231 and MIA PaCa2 cells on day 1 and robustly upregulated on days 2–7 after ²¹¹At-AITM treatment, but not in the vehicle-treated cells (Figures 2J–2L). Simultaneously, the cell morphology apparently changed as the cells became enlarged and flattened with increasing SA-β-gal concentration (Figure 2J). Figures 2M and 2N showed scatterplots and fitted curves of mGluR1 expression versus γH2AX or SA-β-Gal positive cells on days 1–7 after ²¹¹At-AITM treatment. Significant correlations were indicated between the mGluR1 expression and the γH2AX-positive cells, and between the mGluR1 expression and the SA-β-Gal⁺ cells for each timepoint in MDA-MB231 and MIA PaCa2 cancer cells. Overall, our results demonstrate that ²¹¹At-AITM downregulated mGluR1 expression, induced DSBs, and triggered cell senescence, which cooperatively governed cancer cell fates.

²¹¹At-AITM is a safe and effective precision radiopharmaceutical for mGluR1-expressing cancers

To explore whether the ²¹¹At-AITM characteristics observed in vitro could be translated into therapeutic efficacy in vivo, we carried out ²¹¹At-AITM therapeutic experiments in multiple mouse xenograft models. First, the pharmacokinetics of intravenously injected ²¹¹At-AITM were evaluated in human cancer-bearing mice (Figures 3A and 3B). ²¹¹At-AITM showed peak tumor accumulation of radioactivity with 4.05%–5.01% ID/g (the value is quantified for a mean percentage of the injected radioactivity dose per gram of tissue, similarly hereinafter) at 1 h post-injection, which decreased to 3.66%–4.57% ID/g at 3 h and 0.22%–0.57% ID/g at 21 h post-injection in mGluR1⁺ human cancers. In contrast, ²¹¹At-AITM constantly showed low accumulation with 1.80%–2.44% ID/g at 1 h and 0.45%–1.05% ID/g at 3 and 21 h in mGluR1^– human cancers (Figure 3A). A correlation analysis identified a significantly positive association between the mGluR1 expression and the uptake in vivo at 1 and 3 h, but no association at 21 h post-injection in tumor xenografts (Figure S4A). The radiation doses absorbed were 3.79–4.62 Gy/MBq for mGluR1⁺ human cancers and 1.16–2.05 Gy/MBq for mGluR1^– human cancers (Figure 3C), based on a standard method using the Medical Internal Radiation Dose formula.⁴⁰ The biological effective doses to mGluR1⁺ human cancers were 18.95–23.09 Sv/MBq and to mGluR1^– human cancers were 5.8–10.26 Sv/MBq (Figure 3C), based on the quality factor of 5.^41,42 Minimal radioactivity was observed in the thyroid (<0.13% ID) at the three time points, thus indicating high in vivo stability of ²¹¹At-AITM with a low level of deastatination.⁴³ A high accumulation with 10.17–6.28 %ID/g was observed in the stomach, probably because of the physiological mGluR1 expression in the tissue.⁴⁴

Figure 3.

²¹¹At-AITM accumulates rapidly into mGluR1⁺ tumor xenografts in human cancer-bearing mice

(A) Quantitative measurement of radioactivity in human cancers following ex vivo biodistribution analysis at 1, 3, and 21 h after ²¹¹At-AITM injection.

(B) Quantitative measurement of radioactivity in major organs of the representative MDA-MB231-bearing mice after ²¹¹At-AITM injection. Note that thyroid values are presented as a percentage of the injected radioactivity dose (% ID).

(C) Absorbed doses and biological effective doses in human cancers. Absorbed doses were estimated in accordance with the standard method using the medical internal radiation dose formula, expressed as the mean dose per unit of injected activity (gray per MBq; Gy/MBq).⁴⁰ The biological effective dose was calculated by using a quality factor of 5, expressed as sievert per MBq (Sv/MBq).^41,42

(D) Absorbed doses in major organs of the representative MDA-MB231-bearing mice.

(E) Biological effective doses in major organs of the representative MDA-MB231-bearing mice. (A and B) Data are expressed as mean percentage of the injected radioactivity dose per gram of tissue (% ID/g) ± SEM, n = 3 for each time point. Comparisons were performed using an unpaired two-tailed Student’s t test or two-way ANOVA followed by Tukey’s multiple comparison test.

As expected, extremely low radioactivity was observed in the brain of not only 8-week-old adult mice, but also 115-week-old aged mice after ²¹¹At-AITM injection (Figures 3B and S4C), although normal brain tissue exhibits high mGluR1 expression; therefore, we concluded that chemical modifications and the use of the biggest halogen atom ²¹¹At minimized the radiation exposure of the brain. Although high radioactivity was observed in the lung (22.39% ID/g) and spleen (5.86% ID/g) at 1 h (Figures 3B, 3D, and 3E), rapid clearance of radioactivity from blood and these tissues began at 3 h after injection. At 21 h post-injection, except for a low level of radioactivity in the targeted tumor tissues, no considerable amounts of radioactivity were detected in the nontargeted tissues of mice. The good pharmacokinetic profile of ²¹¹At-AITM ensures minimized radiation exposure in healthy off-target organs.

Next, the in vivo antitumor efficacy of ²¹¹At-AITM was tested in xenograft mice bearing seven subtypes of cells characteristic to four types of solid tumors. Given the complex relationship between p53 status and radiation damage responses in tumors,⁴⁵ we defined the p53 status of the panel of human solid cancer cells mutant-type p53 (p53^mt) or wild-type p53 (p53^wt) allele using the PubMed database. The xenograft mice were treated with a single injection of ²¹¹At-AITM (2. 96 MBq), a dose that was chosen based on previous studies³³ and our pilot data (Figures S5A and S5B). In contrast with its minimal effect against mGluR1⁻ cancers (A2058 with p53^mt and PANC1 with p53^mt), a single administration of ²¹¹At-AITM in mGluR1⁺ (MDA-MB231 with p53^mt, MIAPaCa2 with p53^mt, Bowes with p53 ^wt, A375 with p53 ^wt, and DLD1 with p53^mt) cancers exhibited unequivocal and durable antitumor efficacy across breast cancer, pancreatic cancer, melanoma, and colon cancer, regardless of their tissue of origin (Figures 4A–4C). A significant inverse correlation was verified between the mGluR1 expression and the tumor volumes on day 30 post-injection (Figure S4B) (Pearson r = −0.6379, p = 0.0019). Furthermore, the antitumor efficacies of ²¹¹At-AITM were independent of the p53 status, unlike conventional radiotherapy.⁴⁵ Notably, complete regression was observed in 42.11% (8/19) of MDA-MB231 TNBC-bearing mice between 13 and 28 days and 54.55% (6/11) of MIA PaCa2 pancreatic cancer-bearing mice between 13 and 16 days after a single ²¹¹At-AITM treatment (Figures 4A and 4B). No recurrence was observed until the end of this investigation (Figures 4A and S6A–S6C). To clarify the specificity of therapeutic effect of ²¹¹At-AITM in vivo, the xenograft mice bearing mGluR1⁺ MIA PaCa2 cancers were treated with unlabeled FITM (1 mg/kg) or different doses of ²¹¹At-AITM. No significant anti-tumor effect was observed in the FITM-treated group compared with the vehicle-treated group (Figure S5C). However, we found that a single dose of ²¹¹At-AITM at 0.37 MBq resulted in a 22.48% reduction in tumor volume, 82.19% at 1.48 MBq, and 95.53% at 2.96 MBq, compared with the 0 MBq (saline)-treated group at 30 days after administration (Figure S5D). The results confirmed the specificity and dose-response effects of ²¹¹At-AITM RPT from α-particles of ²¹¹At in the mGluR1⁺ xenograft mice.

Figure 4.

A single treatment with ²¹¹At-AITM suppresses tumor growth in multiple models of mGluR1⁺ cancers

(A) Individual tumor growth curves of mice treated with vehicle (saline) or ²¹¹At-AITM. Multiple human cancer cells were grown as tumor xenografts in male and female BALB/c nude mice. Longitudinal progression of tumor volume in tumor-bearing mice treated with a single ²¹¹At-AITM or vehicle (injection = day 0).

(B) Mouse tumor volumes on day 30 post-injection. MDA-MB231: n = 8 for vehicle group, n = 19 for ²¹¹At-AITM group (male: 16, female: 3); MIA PaCa2: n = 6 for vehicle group, n = 11 for ²¹¹At-AITM group (male: 8, female: 3); Bowes: n = 6 for vehicle group, n = 4 for ²¹¹At-AITM group; A375: n = 5 for vehicle group, n = 5 for ²¹¹At-AITM group; DLD1: n = 4 for vehicle group, n = 4 for ²¹¹At-AITM group; A2058: n = 5 for vehicle group, n = 7 for ²¹¹At-AITM group; PANC1: n = 5 for vehicle group (male: 4, female: 1), n = 7 for ²¹¹At-AITM group (male: 4, female: 3). All comparisons were performed using an unpaired two-tailed Student’s t test.

(C) Changes in body weight. Results are expressed as mean ± SEM, n = 4–11.

Finally, the potential toxicity of intravenously injected ²¹¹At-AITM was investigated in multiple human cancer-bearing mice. There was no apparent weight loss (Figure 4C), and no clinical signs of toxicity were observed throughout the examination period. Leukocyte, erythrocyte, and platelet counts were transiently decreased at 5–10 days (Figure 5A). Although blood cell counts recovered to baseline after day 15, management of hematologic functions after ²¹¹At-AITM therapy should be considered in clinical trials. Liver enzyme (glutamic oxaloacetic transaminase and glutamic pyruvic transaminase) levels (Figure 5B) and kidney function (blood urea nitrogen and creatinine) index (Figure 5C) did not show significant changes on days 2, 7, and 30 after exposure to ²¹¹At-AITM. Histological analyses did not show noticeable histological and structural lesions in the liver, kidney, lung, spleen, and stomach at 30 days after a single dose of ²¹¹At-AITM, compared with the vehicle-treated mice (Figure 5D). These results were consistent with the results of enzyme assays (Figures 5B and 5C). ²¹¹At-AITM physiologically accumulated in the stomach, leading to a high stomach-absorbed dose with 12.38 Gy/MBq (Figure 3D). In patients, the maximum tolerated dose in the stomach is postulated to be 50 Gy.⁴⁶ Although stomach histology showed no significant differences in vehicle vs treated mice, stomach protectants during ²¹¹At-AITM RPT are advised, especially when high treatment activities are used. These data suggest that the mGluR1-targeted α-RPT strategy for treating multiple human cancers is safe, can be effective, and can induce complete tumor regression after a single treatment.

Figure 5.

²¹¹At-AITM is a safe radiopharmaceutical in human cancer-bearing mice

(A) Changes in leukocyte, erythrocyte, and platelet counts. Results are expressed as mean ± SEM, n = 6 per group.

(B) The values of glutamic oxaloacetic transaminase (GOT) and glutamic pyruvic transaminase (GPT).

(C) The values of creatinine and blood urea nitrogen (BUN). (B and C) Results are expressed as mean ± SEM, n = 3 per group. All comparisons were performed using an unpaired two-tailed Student’s t test or two-way ANOVA followed by Tukey’s multiple comparison test.

(D) Histological analysis of liver, kidney, lung, spleen, and stomach on 30 days post-injection. Tissue sections (5 μm) were stained with hematoxylin-eosin staining. Scale bars, 50 μm. Representative images are shown from three independent experiments.

²¹¹At-AITM downregulates mGluR1 and modulates cancer senescence with reprogrammed SASP

To gain insight into the mechanism underlying the antitumor effects of ²¹¹At-AITM, we performed multiple analyses to dissect the phenotypic changes in tumor tissues after radiation. The expression levels of mGluR1 (Figures 6A and 6B), the cell proliferation marker Ki67 (Figures 6A and 6C), the senescence marker SA-β-gal (Figures 6D, 6E, and 6M), and the tumor neovascularization marker CD31 (Figure 6N) were assessed on days 2, 7, and 30 after ²¹¹At-AITM treatment. Immunohistological staining of tumor sections showed that ²¹¹At-AITM effectively decreased mGluR1 and Ki67 protein levels in both MDA-MB231 and MIA PaCa2 tumors on day 2 and retained the low levels on day 7 post-injection compared with those observed in the vehicle-treated xenografts (Figures 6A–6C). High Ki67 expression and negligible mGluR1 expression were observed in A2058 and PANC1 xenografts pre- or post-treatment, which is consistent with the result of mGluR1⁺-selective antitumor efficacy of ²¹¹At-AITM. Moreover, ²¹¹At-AITM-treated MDA-MB231 and MIA PaCa2 tumors showed increased numbers of senescent SA-β-gal⁺ cells on day 2, constitutively upregulated on day 7 (Figures 6D and 6E), and still higher levels on days 30 and 45 post-injection compared with the tumors treated with the vehicle (Figures 6M, S6D, and S6E). In contrast, A2058 and PANC1 tumors under the same conditions did not display significant SA-β-gal staining. The results indicated that ²¹¹At-AITM treatment induces stable senescence of mGluR1⁺ cancer cells in vivo, which were consistent with the long-lasting antitumor consequences.

Figure 6.

²¹¹At-AITM therapy reduces mGluR1 expression and triggers cancer senescence by reprogramming the SASP variants

(A) Representative fluorescence images of mGluR1 (green), Ki67 (red), and DAPI (blue) performed on frozen sections from mGluR1⁺ (MDA-MB231 and MIA PaCa2) and mGluR1^- (A2058 and PANC1) xenografts treated either with vehicle (saline) or ²¹¹At-AITM for 2 and 7 days. Scale bar, 50 μm.

(B and C) Quantification of mGluR1 and Ki67 expression.

(D and E) Representative images and quantification of SA-β-gal (green) expression. Scale bar, 50 μm.

(F and G) Quantification of the trigger of cell senescence 53BP1 and CDKN1A using qRT-PCR.

(H–L) Quantification of SASP components using qRT-PCR.

(M) Representative images (left) and quantification (right) of SA-β-gal (green) expression in the representative mGluR1⁺ MIA PaCa2 and mGluR1^- PANC1 cancers on 30-day after ²¹¹At-AITM treatment. Scale bar, 50 μm.

(N) Representative images (left) and quantification (right) of CD31 (green) and DNA (blue) on 30-day following ²¹¹At-AITM treatment to assess tumor angiogenesis. Scale bar, 50 μm.

(O and P) Quantification of VEGFA and MET using qRT-PCR on day 30 post-injection. (B, C, F, G, and E–P) Results are expressed as mean ± SEM, n = 4 per group. All comparisons were performed using an unpaired two-tailed Student’s t test or two-way ANOVA followed by Tukey’s multiple comparison test. (A, D, M, and N) Representative images are shown from three independent experiments.

We also assessed the gene expression associated with DDR after treatment with ²¹¹At-AITM. The expression of 53BP1 (a biomarker of DDR) and CDKN1A (the gene encoding p21) was significantly upregulated on day 2 and persisted at high levels 7 days after ²¹¹At-AITM treatment in mGluR1⁺ cancers compared with that in mGluR1^– cancers (Figures 6F and 6G), although p53 is mutated in both MDA-MB231 and MIA PaCa2 cancer cells. These results suggested that DDR represents a trigger of cell senescence after ²¹¹At-AITM treatment, and persistent DDR activation maintains long-term and stable cancer senescence by upregulating p21 expression.

Activated MAPK and mTOR pathways, which are modulated by mGluR1, play important roles in stimulating the production of SASP factors.^21,22,47 We hypothesized that reduced mGluR1 expression elicits an extensive range of effects on SASP components. To verify this hypothesis, the expression levels of several indicators of SASP, including pro-inflammatory interleukin-6 (IL-6), CXC chemokine ligand 8 (CXCL8), tumor necrosis factor alpha (TNF-α), vascular endothelial growth factor A (VEGFA), and MET, were measured using qRT-PCR in mGluR1⁺ (MDA-MB231 and MIA PaCa2) and mGluR1^– (A2058 and PANC1) xenografts after ²¹¹At-AITM treatment. Consistent with the senescence expansion, the expression levels of the key SASP components IL-6, CXCL8 and TNF-α were significantly increased in ²¹¹At-AITM-treated MDA-MB231 and MIA PaCa2 tumors, but not in ²¹¹At-AITM-treated mGluR1^– tumors (Figures 6H–6J). Surprisingly, a significant decrease in the expression of pro-angiogenic VEGFA and pro-metastatic MET was observed on days 2 and 7 (Figures 6K and 6L), and constitutively low levels on day 30 after α-irradiation in ²¹¹At-AITM-treated mGluR1⁺ cancers, compared with vehicle-treated cancers and ²¹¹At-AITM-treated mGluR1^– cancers (Figures 6O and 6P). The SASP variants had not been identified in conventional X-ray treatment (Figures S7A–S7J). Furthermore, we investigated tumor angiogenesis by immunohistological staining with an antibody against CD31 in 30 and 45-day tumor xenografts after ²¹¹At-AITM RPT.⁴⁸ The CD31-stained blood vessels showed much fewer signals in MIA PaCa2 and MDA-MB231 tumor tissues than their corresponding control tumors (Figures 6N and S6F–S6G), suggesting that reprogrammed SASP variants suppressed tumor angiogenesis and remodeled the tumor microenvironment into angiogenesis antagonizing and metastasis antagonizing. Overall, these results suggested that ²¹¹At-AITM therapy reprogrammed the phenotypes of mGluR1⁺ cancers, including but not limited to downregulating mGluR1 expression, inducing cell senescence, secreting angiogenesis-antagonizing as well as metastasis-antagonizing SASP components, and suppressing tumor angiogenesis. In coordination with its DNA DSB effects, ²¹¹At-AITM and SASP work together to govern the fate of tumor cells.

Discussion

The success of metabolic imaging with ¹⁸F-FDG in clinical oncology spurred the exploration of therapeutic strategies targeting mGluR1, a key mediator of glutamatergic signaling, for tumor-agnostic treatment of cancers. However, mGluR1 antagonists are usually associated with drug resistance and unsatisfactory clinical outcomes; thus, we sought to develop alternative therapeutic strategies that can overcome the bottleneck of conventional molecular targeted therapy. Receptor-targeted α-RPT, which is impervious to conventional cellular resistance mechanisms,⁴⁹ is considered an ideal approach for mGluR1-targeted therapy owing to the metabolism-independent cell-killing effects of α-rays.

In this study, we showed that a single intravenous administration of ²¹¹At-AITM resulted in unequivocal and durable tumor growth inhibition in multiple models of mGluR1⁺ human cancers with little toxicity, regardless of their tissue of origin. Moreover, mGluR1-targeting ²¹¹At-AITM therapy led to complete regression of mGluR1⁺ breast cancer and pancreatic cancer in approximate 50% of tumor-bearing mice. Interestingly, mechanistic studies uncovered the functions of ²¹¹At-AITM in reshaping the phenotypic patterns of tumor cells, such as decreasing mGluR1 expression, inducing cell senescence, releasing angiogenesis-antagonizing as well as metastasis-antagonizing SASP variants, and suppressing tumor angiogenesis, beyond its classical function in DNA DSBs. These divergent effects of ²¹¹At-AITM cooperatively governed tumor cell fate (Figure 7). Thus, our results facilitate the understanding of the tumor cells cascade responses against α-radiation and pave the way for developing precise α-RPT for mGluR1⁺ pan-cancers.

Figure 7.

Schematic representation of the potential mechanism through which ²¹¹At-AITM kills cancer cells

It is proposed that ²¹¹At-AITM RPT inhibits mGluR1 function by decreasing its expression levels, which results in downregulation of the MAPK, phosphatidylinositol-3 kinase (PI3K)/AKT, and AKT/mTOR/hypoxia-inducible factor 1 (HIF1) pathways, leading to the suppression of mGluR1-mediated tumor growth, angiogenesis, and metastasis. Furthermore, the α-particles can trigger irreparable DSBs in targeted cancer cells, which activate a persistent DDR that drives cancer senescence. Lastly, mGluR1 ablation and senescent cancer co-reprogram the SASP components into angiogenesis-antagonizing and metastasis-antagonizing variants to improve the long-term outcome of patients and even cure the disease by co-promoting cancer cell death.

In RPT, radiation is delivered systemically to cancer cells using high-affinity ligands as carriers of radionuclides—primarily β-emitters (β-RPT) or high LET α-emitters (α-RPT).⁵⁰ The α-RPT is currently attracting interest owing to the superior biological radiation effects of α-rays compared to those of β-rays. The ²²³Ra dichloride (Xofigo), the first α-radiopharmaceutical approved by the US Food and Drug Administration,³² is used for treating bone metastases in patients with advanced castration-resistant prostate cancer. The mGluR1-targeting α-RPT is an example of targeting cancer metabolism-associated targets for cancer treatment. Notably, ¹⁸F-FITM is an appropriate companion diagnostic tracer for ²¹¹At-AITM,⁵¹ as the two radiopharmaceuticals share similar pharmacokinetics and in vivo biodistribution profiles. For example, both showed rapid tumor uptake and non-target normal tissue clearance. In the scenario of imaging-guided α-RPT, the target engagement of ²¹¹At-AITM could be accurately predicted using ¹⁸F-FITM PET to provide valuable information for dosimetry and treatment planning. This precision treatment mode provides substantial advantages over conventional antagonistic therapy.

As an oncoprotein that is aberrantly overexpressed on the cell membranes of most cancers, mGluR1 shows only limited background expression in normal peripheral organs,^11,19,51 thus making it a potential target for α-RPT. As a result, an elevated administration dose could be considered a feasible method to improve therapeutic outcomes, largely expanding the therapeutic window of tumors with relatively low mGluR1 protein expression. Notably, even though mGluR1 is highly expressed in normal brain tissue, ²¹¹At-AITM showed little blood-brain barrier penetrance and extremely low brain activity accumulation. This mitigates any concern of radiation toxicity to the normal human brain. Another safety concern of translating ²¹¹At-AITM α-RPT to the clinic is the exposure of patients’ healthy off-target organs to α-radiation. In this regard, it should be noted that the temporarily high radioactivity in the blood, lung, spleen, stomach, liver, and kidneys did not induce body weight reduction, severe hematologic toxicity, apparent liver and kidney functional side effects, or obvious histological and structural lesions of the major organs during the ²¹¹At-AITM treatment. Because of the remarkable α-particle flux emitted by ²¹¹At (t_1/2 = 7.21 h), ²¹¹At-AITM led to irreparable DSBs in targeted cancer cells, and effective cell killing can be expected even in the hypoxic regions of solid cancers.⁵² Furthermore, ²¹¹At-AITM used for treating multiple human solid cancers achieved consistently high response rates. Moreover, in MDA-MB231 and MIA PaCa2 tumor-bearing mice, ²¹¹At-AITM cured approximate 50% of mice and no recurrence was observed until the end of the investigation. Therefore, the outstanding therapeutic efficacy and favorable safety profile of ²¹¹At-AITM highlight its remarkable clinical and translational value associated with systemically targeted therapy of a wide range of human cancers, regardless of their tissue of origin.

Based on the classical theory of radiation biology, α-particles directly or indirectly cause DNA DSBs to induce cell death. In addition to the classical function, our mechanistic investigation into the cellular response against ²¹¹At-AITM exposure led to the identification of two roles of ²¹¹At-AITM through which it affects tumor cell fate. Specifically, ²¹¹At-AITM treatment reshaped the phenotypic features of tumor cells. One is the reduction in mGluR1 expression, while the other is the induction of cancer senescence with reprogrammed SASP variants. The phenotypic variations synergize with genetic alterations to cooperatively govern cell fate.

In summary, we have demonstrated that mGluR1-targeted small-molecule α-RPT could deliver long-lasting therapeutic benefits and even cure a wide range of human cancers, regardless of their tissue of origin. Furthermore, mechanistic studies revealed two roles of ²¹¹At-AITM RPT in reshaping the phenotype of tumor cells, including the reduction in mGluR1 expression and induction of cancer senescence with programmed SASP variants. Given the prevalence of glutamine/glutamate dysregulation and mGluR1 overexpression in cancers, α-RPT with ²¹¹At-AITM may have broad application prospects and bring us one step closer to winning the fight against cancers.

Limitations of the study

There are several limitations of this study. First, our understanding of the molecular mechanisms underlying the biological effects of ²¹¹At-AITM remains insufficient. In the future, research efforts should focus on more extensive in-parallel dosimetry control experiments and longitudinal monitoring of changes in mGluR1 expression within tumor tissues. Second, whether α-radiation or reduced mGluR1 expression alone or both coordinately contribute to the reprogramming of SASP in senescent cancers is currently unclear. Third, the changes in glutamate metabolism induced by mGluR1 downregulation remain unexplored. Furthermore, the correlation between glutamine metabolism variation and senescence in our study was unclear. It should be emphasized that these questions are critical for further clinical application of ²¹¹At-AITM RPT in patients with cancer that will, in turn, provide valuable insights for the development of effective combinations of α-RPT with immunotherapies or senolytic therapies against cancer. Following this proof-of-concept study, further studies on the exact mechanism and the translational clinical trial of ²¹¹At-AITM RPT with long-term follow-up for the treatment of multiple cancers are ongoing.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Rabbit anti-mGluR1	LSBio	Cat#LS-A882; RRID:AB_592206
Mouse anti-Ki67	BD Pharmingen	Cat#550609; RRID:AB_393778
Rat anti-CD31	BD Pharmingen	Cat#550274; RRID:AB_393571
Alexa Fluor® 488 goat anti-rat IgG	Invitrogen	Cat#A-11006; RRID:AB_141373
Alexa Fluor® 488 goat anti-rabbit IgG	Invitrogen	Cat#A-11008; RRID:AB_143165
Alexa Fluor® 546 goat anti-mouse IgG	Invitrogen	Cat#A-11003; RRID:AB_141370
Chemicals, peptides, and recombinant proteins
211At-AITM	Previously developed in our Lab (Xie et al.33)	N/A
DAPI	Vector Laboratories	Cat#H-1200
DMEM, with glucose, L-glutamine, sodium pyruvate, and sodium bicarbonate	Sigma-Aldrich	Cat#D6429
MEM	Sigma-Aldrich	Cat#M4655
RPMI-1640	Sigma-Aldrich	Cat#R8758
Trypsin-EDTA solution	Gibco	Cat# 25300054
Fetal bovine serum	Sigma-Aldrich	Cat#12103C
Penicillin/streptomycin	Gibco	Cat#15140122
DAPI	Vector Laboratories	Cat#H-1200
Matrigel	Corning	Cat#356234
FITM	MedChemExpress	Cat#932737-65-0
Sepasol-RNA I Super	Nacalai Tesque	Cat#09379-84
Critical commercial assays
γH2AX DNA Damage Detection Kit	Dojindo	Cat#G266
Senescence-associated β-galactosidase staining kits	Dojindo	Cat#SG02
Muse® Count & Viability Kit	Luminex	Cat#MCH100102
Muse® Cell Cycle kit	Luminex	Cat#MCH100106
RNeasy Mini Kit	Qiagen	Cat#74904
Protein assay kit	Bio-Rad	Cat#5000111
DNA-free™ DNA Removal kit	Invitrogen	Cat#AM1906
PrimeScript™ RT reagent Kit	Takara	Cat#RR037A
FUJI DRI-CHEM SLIDE GOT/AST-PIII	Fujifilm	Cat#4902520-77484-2
FUJI DRI-CHEM SLIDE GPT/ALT-PIII	Fujifilm	Cat#4902520-77485-9
FUJI DRI-CHEM SLIDE CRE-PIII	Fujifilm	Cat#4902520-77481-1
FUJI DRI-CHEM SLIDE BUN-PIII	Fujifilm	Cat#4902520-89062-7
Deposited data
TCGA RNA –Seq	This paper	Mendeley Data (https://data.mendeley.com/datasets/xfh84npb2h/2) with the database https://doi.org/10.17632/xfh84npb2h.2.
Experimental models: Cell lines
MDA-MB231	ATCC	HTB-26
MIA PaCa2	RIKEN	RCB2094
PANC1	RIKEN	RCB2095
DLD1	ATCC	CCL-221
A375	ATCC	CRL-1619
Bowes	ATCC	CRL-9607
A2058	JCRB	IFO50276
Experimental models: Organisms/strains
Mouse: Balb/c-nu/nu	Japan SLC	N/A
Oligonucleotides
Q-PCR prime: human GRM1	Applied Biosystems	Hs00168250_m1
Q-PCR prime: human 53BP1	Applied Biosystems	Hs00996827_m1
Q-PCR prime: human CDKN1A	Applied Biosystems	Hs00355782_m1
Q-PCR prime: human IL-6	Applied Biosystems	Hs00174131_m1
Q-PCR prime: human CXCL8	Applied Biosystems	Hs00174103_m1
Q-PCR prime: human TNF-α	Applied Biosystems	Hs00174128_m1
Q-PCR prime: human VEGFA	Applied Biosystems	Hs00900055_m1
Q-PCR prime: human MET	Applied Biosystems	Hs01565584_m1
Q-PCR prime:18S ribosomal RNA	Applied Biosystems	Hs99999901_s1
Software and algorithms
GraphPad Prism 8	GraphPad Software	https://www.graphpad.com/scientific-software/prism/
Molecular Operating Environment	Chemical Computing Group	https://www.chemcomp.com/index.htm
Visual Molecular Dynamics	University of Illinois	https://www.ks.uiuc.edu/Research/vmd/
AMBER18	HPC Systems	https://www.hpc.co.jp/chem/software/amber/
PyMol	Schrödinger	https://pymol.org/2/
Specialized Hybrid Cell Count software	Keyence	https://www.keyence.co.jp/
Medical internal radiation dose formula	Stabin40	https://jnm.snmjournals.org/content/37/3/538.long

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Dr. Ming-Rong Zhang (zhang.ming-rong@qst.go.jp).

Materials availability

This study did not generate new unique reagents. The radiopharmaceutical in this study would be available from the lead contact without restriction.

Experimental model and subject details

Cell lines

The following cell lines were used in this study: MDA-MB231 human TNBCs (ATCC, Manassas, VA, USA), MIA PaCa2 (RIKEN, Tsukuba, Japan) and PANC1 (RIKEN) human pancreatic cancers, DLD1 human colon cancers (ATCC), A375 (ATCC), Bowes (ATCC), and A2058 (Japanese Collection of Research Bioresources, Osaka, Japan) human melanomas. MDA-MB231, MIA PaCa-2, A375, and A2058 cells were maintained and passaged in Dulbecco’s modified Eagle’s medium (DMEM; Sigma-Aldrich, St. Louis, MO, USA) supplemented with 10% fetal bovine serum (FBS; Sigma-Aldrich) and 1% penicillin/streptomycin (50 U/mL; Gibco, Grand Island, NY, USA). Bowes cells were maintained and passaged in minimum Eagle’s medium (MEM; Sigma-Aldrich) supplemented with 10% FBS and 1% penicillin/streptomycin. PANC1 and DLD1 cells were maintained and passaged in RPMI-1640 medium (Sigma-Aldrich) supplemented with 10% FBS and 1% penicillin/streptomycin. All cells were cultured at 37°C in a humidified atmosphere with 5% CO₂ and were regularly tested to ensure the absence of mycoplasma contamination.

Mice and mouse models

Animal experiments were performed on 6−8-week-old male and female Balb/c-nu/nu mice (body weight, 21.18 ± 0.25 g) (Japan SLC, Shizuoka, Japan). The xenograft tumor models were created by injecting a single-cell suspension of human cancer cells (5×10⁶) into the left flanks of mice with Matrigel (Corning, Teterboro, NJ, USA). All animal experiments were approved by the Animal Ethics Committee of National Institutes for Quantum Science and Technology (QST). Animals were maintained and handled under a SPF condition (12-h light/dark cycle, 50% relative humidity, between 25°C and 27°C) with free access to food and tap water, in accordance with the recommendations of the National Institute of Health and the institutional guidelines of QST.

Method details

Study design

The major objective of this study was to evaluate the therapeutic efficacy of a mGluR1-based RPT strategy using a small-molecule α-emitting radiopharmaceutical (²¹¹At-AITM) against multiple human cancers and to define the mechanisms underlying the antitumor effects of ²¹¹At-AITM. The binding and specific studies of ²¹¹At-AITM for mGluR1 were carried out using parallel control experiments including comparative experiments in multiple mGluR1⁺ and mGluR1⁻ human cancer cells and competition experiments with unlabeled mGluR1-competitive antagonist FITM. In vitro effects of ²¹¹At-AITM on mGluR1-expressing cancer cells were tested using two mGluR1⁺ human cancer cells treated with the testing α-radiopharmaceutical (²¹¹At-AITM) or vehicle (PBS). End points for cell morphology, viability and phenotypic analyses were selected on the basis of a time course ranging from 1 to 7 days after addition of ²¹¹At-AITM to cells in culture. All cellular experiments were conducted in triplicates (technical replicates) and repeated independently at least three times unless otherwise noted.

In vivo studies were performed to evaluate the effects of ²¹¹At-AITM in xenograft mice bearing 7 subtypes from 4 solid tumors with different tissue origins and distinct mGluR1 expression levels, as well as its pharmacokinetics, radiation dosimetry, biological effective dose, safety, and phenotypic changes of cancer tissues after intravenous injection of ²¹¹At-AITM. Xenotransplantation studies were carried out using more than 4 animals per group unless otherwise specified. Animals were transplanted by injecting a single-cell suspension of human cancer cells (5×10⁶) into the left flanks of mice and randomly assigned to experimental groups as defined in the Figure legends. All groups were dosed with a single intravenous administration of ²¹¹At-AITM, FITM, or vehicle (saline). The end point of these in vivo therapeutic experiments was set at 30 days after injection and body weight loss of more than 20%. Exclusion criteria included unspecified death. All experiments were carried out as unblinded studies.

GRM1 expression data in human cancer tissues

To analyze GRM1 expression in human cancer samples, data were downloaded from cBioPortal (www.cbioportal.org/), and raw sequencing data from TCGA were imported (https://cancergenome.nih.gov/). All available GRM1 mRNA data as of March 2022 were downloaded and transformed into a plot using GraphPad Prism 8 (GraphPad Software, San Diego, CA, USA). Patient samples covering 10,967 across 32 tumor types were analyzed and grouped into two subpopulations according to the expression status of GRM1 (GRM1⁺ cancers and GRM1^- cancers).

Molecular docking study

The three-dimension FITM-mGluR1 complex structure (PDB ID: 4OR2) was used for molecular modeling. The initial structure of AITM complex was generated using MOE (Molecular Operating Environment; Chemical Computing Group, Montreal, QC, Canada). The FITM and AITM complexes were added by lipid bilayer and water molecules around the complexes using VMD (Visual Molecular Dynamics). For the FITM complex, the energy minimized (MM) calculations and molecular dynamics (MD) simulations (5 ns) were performed using AMBER18/SANDER. The conditions of MD simulations were constant temperature (300 K) and pressure (1 atm). On the other hand, the MM calculations were performed for the AITM complex using MOE. The final structures of FITM or AITM complexes were visualized using PyMol.

Radiosynthesis of ²¹¹At-AITM

²¹¹At-AITM was synthesized by reacting a stannyl precursor (100 μg) with ²¹¹At in the presence of an oxidizing agent. ²¹¹At-AITM was prepared as an injectable solution according to our previously reported method.^{33 211}At-AITM was obtained with 36–118 MBq in 45.7 ± 6.5% radiochemical yield (n = 20, no decay-corrected, based on the total radionuclides used) with > 99% radiochemical purity, which was used for all experiments.

Cell binding assay

Human cancer cells (2 × 10⁵ cells) were seeded into 24-well plates and allowed to adhere though incubation overnight for all further experiments. Cells were incubated in triplicate with 18.5 kBq/mL ²¹¹At-AITM for 1, 3, and 21 h at 37°C. Excess unlabeled mGluR1-competitive antagonist FITM (10 μmol/L; MedChemExpress, Monmouth Junction, NJ, USA) was used to determine nonspecific binding.¹⁴ The percentage of internalization and membrane-bound ²¹¹At-AITM was determined using a previously published method.³⁸ Briefly, MAD-MB231 and MIA PaCa2 cells, after incubation with ²¹¹At-AITM, were washed with medium and treated with a mild acid buffer (50 mmol/L glycine +150 mmol/L NaCl, pH 2.8) at 4°C for 5 min. The buffer was collected, and cells were further washed with the same amount of buffer. The pooled washes were considered to represent membrane-bound radioactivity, while the unreleased radioactivity was considered internalized and was collected by treating the cells with 0.2 mol/L NaOH. Protein quantification was performed using a protein assay kit (Bio-Rad, Hercules, CA, USA), while radioactivity uptake was assessed using a γ-counter (PerkinElmer, Waltham, MA, USA). Cellular radioactivity uptake was calculated as a percentage of the incubated radioactivity dose normalized per mg of protein (% ICD/mg protein). The ratio between internalized and membrane-bound radioactivity was expressed as a percentage of cell-associated radioactivity dose (% CD).

Immunohistochemistry

To investigate ²¹¹At-AITM-induced damage events, MDA-MB231 (2 × 10⁵ cells) and MIA PaCa2 cells (2 × 10⁵ cells) were exposed to 18.5 kBq/mL ²¹¹At-AITM or vehicle (PBS) for 21 h in triplicate in 24-well plates. Alterations in cell morphology, viability, and cell cycle distribution were subsequently analyzed until 7 days or otherwise indicated. Animals were anesthetized with isoflurane and then sacrificed. Tissue samples of xenograft tumors were harvested and frozen at 2, 7, and 30 days after ²¹¹At-AITM injection. Tumor sections (5 μm) were prepared using a Microm HM560 microtome (Carl Zeiss Jena, Germany).

mGluR1, Ki67 and CD31 staining

For cell immunohistochemistry, cancer cells after the indicated treatment were fixed with 4% paraformaldehyde (PFA) solution, incubated overnight at 4°C with primary rabbit anti-human mGluR1 antibody (1:200; LS-A882, LSBio, Seattle, WA, USA), incubated with secondary Alexa Fluor® 488 goat anti-rabbit IgG antibodies (1:500; A-11008, Invitrogen, Carlsbad, CA, USA) for 1 h at room temperature (RT), and mounted using mounting medium with DAPI (Vector Laboratories, Burlingame, CA, USA). For tissue immunohistochemistry, the tumor sections treated as indicated were subjected to immunofluorescence staining with rabbit anti-human mGluR1 antibody (1:200), mouse anti-human Ki67 antibody (1:50; 550609, BD Pharmingen, San Diego, CA, USA), or rat anti-mouse CD31 antibody (1:100; 550274, BD Pharmingen), after which they were incubated with Alexa Fluor® 488 goat anti-rabbit IgG (1:500; Invitrogen), Alexa Fluor® 546 goat anti-mouse IgG (1:500; A-11003, Invitrogen), or Alexa Fluor® 488 goat anti-rat IgG (1:500; A-11006, Invitrogen) and mounted using mounting medium with DAPI.

γH2AX staining

At the indicated time points after ²¹¹At-AITM treatment, cells were fixed with 4% PFA and used as a substrate for a γH2AX DNA Damage Detection Kit (G266, Dojindo, Kumamoto, Japan) according to the manufacturer’s instructions. Briefly, cells were washed twice with PBS and permeabilized for 30 min in 1% Triton X-100 at RT. Next, cells were incubated overnight at 4°C with primary mouse anti-γH2AX (1:50), incubated with goat anti-mouse-Red (1:50) secondary antibody for 1 h at RT, and mounted using mounting medium with DAPI.

Senescence-associated β-galactosidase assay

Commercially available senescence-associated β-galactosidase (SA-β-gal) staining kits (SG02, Dojindo) were used to detect senescence in cells and tissues as instructed by the manufacturer. Briefly, tumor cells and tissue sections were fixed with 4% PFA, washed with PBS and stained for 30 min with fresh SA-β-gal solution (10 μmol/L, pH 6.0) at 37°C in an incubator without CO₂. Samples were mounted using mounting medium with DAPI.

Immunohistochemistry images were acquired using a Keyence BZ-X710 microscope (Keyence, Osaka, Japan). The fluorescence intensity, frequency and area of positively stained cells were measured automatically by using specialized Hybrid Cell Count software (Keyence). The results were expressed as the fluorescence intensity per cell and the percentage of total cells or tissue areas that exhibited positive staining. Negative control slides were processed in the absence of primary antibody, secondary antibody, or isotype control IgG to ensure specificity. For comparisons between images from the same experiment, all parameters were adjusted equally, and the ratio between the parameters was not altered.

Cell viability assay

A Muse® Cell Analyzer (Luminex, Austin, TX, USA) was used to determine the viability of tumor cells after ²¹¹At-AITM treatment using the Muse® Count & Viability Kit (MCH100102, Luminex) as instructed by the manufacturer. Briefly, at the indicated time points after ²¹¹At-AITM or vehicle (PBS) treatment, MDA-MB231 and MIA PaCa2 cells were harvested in triplicate as single-cell suspensions through digestion with Gibco trypsin-EDTA (0.05%) at 37°C for 5 min. After the incubation, cells were washed and mixed in fresh PBS. Next, 20 μL of this cell suspension was mixed with 380 μL of Count & Viability reagent. The suspension was incubated for 5 min at RT and thereafter examined for cell count and viability using the Muse® Cell Analyzer.

Cell cycle assay

The effect of ²¹¹At-AITM on the human tumor cell cycle was determined using a Muse® Cell Analyzer and a Muse® Cell Cycle kit (MCH100106, Luminex) according to the manufacturer’s instruction. Briefly, at the indicated time points after ²¹¹At-AITM or vehicle (PBS) treatment, MDA-MB231 and MIA PaCa2 cells were harvested in triplicate as single-cell suspensions and washed with fresh PBS, followed by spinning at 3000 rpm for 5 min. The cell pellet was fixed in 1 mL of 70% ice-cold ethanol and stored at −20°C until used. Next, 200 μL of fixed cells was aliquoted, centrifuged at 3000 rpm for 5 min, and washed with PBS solution. To the fixed cells, 200 μL of Muse® Cell Cycle reagent was mixed and incubated for 30 min in the dark at RT. Samples were examined using the Muse® Cell Analyzer.

In vivo performance, radiation dosimetry and biological effective dose

The in vivo effect of ²¹¹At-AITM was calculated using biodistribution data in corresponding tumor-bearing mice. A therapeutic amount of ²¹¹At-AITM (2.96 MBq/100 μL) was administered intravenously to mice, which were subsequently sacrificed by cervical dislocation at 1, 3, and 21 h post-injection. The tumor and major organs were promptly excised, harvested, as well as weighed. The radioactivity was measured using a 2470 Wizard automatic γ-counter (PerkinElmer). A time profile of the radioactivity in each tumor and tissue was generated by normalizing the decay-corrected radioactivity time to a percentage of the injected dose per gram of wet tissue (%ID/g). The dosimetry of ²¹¹At-AITM absorbed by each organ and tumor was estimated in accordance with the standard method using the medical internal radiation dose formula,⁴⁰ and the results are expressed as the mean dose per unit of injected activity (gray per MBq; Gy/MBq). The biological effective dose was calculated by using a quality factor of 5, expressed as sievert per MBq (Sv/MBq).^41,42

Therapeutic study of ²¹¹At-AITM in xenograft-bearing mice

Radiotherapy experiments were performed on 6−8-week-old male and female Balb/c-nu/nu mice (body weight, 21.18 ± 0.25 g) bearing 7 different human tumor cell types (MDA-MB231, MIA PaCa2, PANC1, A375, Bowes, A2058, and DLD1). Tumor-bearing mice with a tumor volume of ∼50 mm³ (approximately 1 week after tumor inoculation) were randomly assigned to treatment and vehicle groups. The therapeutic efficacy was evaluated with a single intravenous administration of 2.96 MBq of ²¹¹At-AITM, 1 mg/kg of FITM or vehicle (saline) in 100 μL to each tumor-bearing mouse. Tumor dimensions were measured twice weekly with digital calipers in a blinded fashion, and tumor volumes were calculated using the following formula: (width² × length)/2. The tumor xenografts were examined at 2, 7, and 30 days after ²¹¹At-AITM injection by immunohistochemistry and gene analysis.

Safety assessment

Changes in the body weight of mice were evaluated as an indicator of the radiation-related side effects of ²¹¹At-AITM. The hematology and liver and kidney chemistry were evaluated in human cancer-bearing mice administered 2.96 MBq of ²¹¹At-AITM or vehicle (saline). Hematological analyses included leukocyte, erythrocyte and platelet counts were performed using a Celltack F Automated Hematology Analyzer (Nihon Kohden, Tokyo, Japan). Liver and kidney analyses included determinations of glutamic oxaloacetic transaminase (GOT), glutamic pyruvic transaminase (GPT) values, creatinine and blood urea nitrogen (BUN) values, which were conducted using Japan Society of Clinical Chemistry reference methods according to the manufacturer’s instructions. Histological analysis of the major organs (liver, kidney, lung, spleen, and stomach) was performed. The samples were collected from the xenograft mice treated with ²¹¹At-AITM or vehicle (saline) at 30 days post-injection, fixed with 10% PFA and embedded in paraffin blocks. Tissue sections (5 μm) were stained with hematoxylin-eosin staining, investigated using a Keyence BZ-X710 microscope.

RNA preparation and qRT-PCR analysis

Total RNA was extracted from tumor cells using the RNeasy Mini Kit (Qiagen, Valencia, CA, USA) and from frozen tumor tissues using Sepasol-RNA I Super (Nacalai Tesque, Kyoto, Japan) according to the manufacturer’s protocol and subsequently purified with the DNA-free™ DNA Removal kit (Ambion, Austin, TX, USA). The quality of the total RNA was verified using the 260/280 nm ratio and a NanoDrop 2000 (Thermo Fisher Scientific, Wilmington, DE, USA). qRT-PCR was performed using a TaqMan system on an Applied Biosystems StepOne™ machine (Carlsbad, CA, USA) according to the manufacturer’s instruction. Target-specific primers and probes for human GRM1 (Hs00168250_m1), 53BP1 (Hs00996827_m1), CDKN1A (Hs00355782_m1), IL-6 (Hs00174131_m1), CXCL8 (Hs00174103_m1), TNF-α (Hs00174128_m1), VEGFA (Hs00900055_m1), MET (Hs01565584_m1) and 18S ribosomal RNA (18S rRNA, Hs99999901_s1) were purchased from Applied Biosystems. The normalized Ct value of each gene was obtained by subtracting the Ct value for 18S rRNA.

Quantification and statistical analysis

Quantitative data are presented as mean ± standard error of the mean (SEM). Intergroup comparisons were performed using an unpaired two-tailed Student’s t-test or two-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison test in GraphPad Prism8. Pearson’s correlation analysis was used to estimate the relationship between tumor mGluR1 expression and cytotoxicity, ²¹¹At-AITM uptake, or therapeutic efficacy in vitro and in vivo. The threshold for statistical significance was set at P < 0.05.

Published: March 31, 2023

Data and code availability

All data generated from this study are included in the paper and the supplementary information. The RNA-seq data downloaded from TCGA have been deposited to Mendeley Data (https://data.mendeley.com/datasets/xfh84npb2h/2) with the database https://doi.org/10.17632/xfh84npb2h.2. This paper does not report original code. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.