MUC1-targeting small peptide radiopharmaceuticals for breast cancer Authors: Rim Malek, Madeleine Landry, Pierre Cheung, Corinne Beinat Published: 2026 ABSTRACT AbstractBackgroundMucin 1 (MUC1) is a transmembrane glycoprotein overexpressed and underglycosylated in numerous epithelial cancers, including breast cancer. Reduced glycosylation leads to the exposure of the variable number tandem repeat (VNTR) region. To the best of our knowledge, all peptides previously described in the literature target the same epitope sequence of the VNTR. Given the high prevalence of breast cancer and the limited treatment options for the aggressive subtype triple-negative breast cancer (TNBC), due to its lack of oestrogen receptors (ER), progesterone receptors (PR), and human epidermal growth factor receptor 2 (HER2), we sought to develop a small peptide radiopharmaceutical targeting MUC1 by exploring all 3 minimal epitope sequences of the VNTR: RPAPGS, PPAHGVT and PDTRP. We also investigated the influence of linker lipophilicity on the binding affinity to MUC1.ResultsThe reference compound 1 showed the highest cell uptake among all tested compounds. While some statistically significant differences were observed for the cell uptake between the different peptide sequences and the linkers, the uptake was so low that no reliable structure-activity relationships could be established. We then studied the specificity of all compounds for MUC1 by comparing the uptake in MUC1-expressing and MUC1-knockout (KO) cells, and unexpectedly observed no specificity for any compound. A saturation binding assay of several peptides showed their binding was too low to reliably determine their binding affinity (Kd). Surface plasmon resonance (SPR) further confirmed the absence of binding of all [NatGa]- and [NatLu]-labelled peptides tested. Given the discrepancies between our cell data and the previously reported results, we next assessed the specificity of the reference (1) in vivo in mice bearing MUC1-expressing and MUC1-knockdown (KD) tumours, which further proved its non-specificity.ConclusionsWhile MUC1 is a very promising target for the development of breast cancer theranostics, designing peptidomimetics based on its minimal epitopes do not lead to high-affinity binders. Our ongoing efforts involve utilizing phage-display to identify new peptide sequences.Supplementary InformationThe online version contains supplementary material available at 10.1186/s41181-026-00451-1.Keywords: MUC1, Radiopharmaceutical, Epithelial cancer, Breast cancer, Triple-negative breast cancer Background Mucin 1 (MUC1) is a transmembrane glycoprotein formed by an extracellular N-terminal subunit (MUC1-N) and a C-terminal subunit. The latter comprises a cytoplasmic tail, a transmembrane part and a short extracellular domain. MUC1-N is mainly formed by the Variable Number Tandem Repeat (VNTR), a sequence of 20 amino acids (PDTRPAPGSTAPPAHGVTSA) repeated 20–120 times. In healthy epithelial cells, MUC1 is heavily O-glycosylated at the serine and threonine residues of the VNTR, the carbohydrates being responsible for 50–90% of its mass (Nath and Mukherjee 2014). In most human epithelial cancers, such as breast, lung, gastric, pancreatic, ovarian, prostate and colorectal carcinomas, MUC1 is overexpressed and under-glycosylated, leading to the exposure of the core peptide of the VNTR (Ghosh et al. 2013). Breast cancers (BC) are the most commonly diagnosed cancers worldwide with 2.3 million new cases reported in 2020 (Sung et al. 2021). Over the years, different classifications of BC have been developed. The most widely employed one describes four subtypes: three of them are characterized by the presence of oestrogen receptors (ER), progesterone receptors (PR), or human epidermal growth factor receptor 2 (HER2); the fourth subtype is characterized by the lack of all three receptors, known as triple-negative breast cancer (TNBC) (Orrantia-Borunda et al. 2022). It is a particularly aggressive subtype, often affecting younger patients, is usually diagnosed at an advanced clinical stage, has a poor prognosis, high rate of recurrence, limited treatment options, and becomes rapidly resistant to chemotherapy (Yi et al. 2021). Although, immunotherapy targeting programmed cell death 1 (PD-1) pathway has recently been approved by the FDA for TNBC, only a minority of patients benefit and the vast majority suffer from disease progression, making sustained disease control rare (Cortes et al. 2020; Schmid et al. 2018). While most BC are curable if diagnosed at an early stage, unfortunately metastatic disease still has a poor prognosis (Harbeck et al. 2019). Theranostics are compounds designed to deliver a radionuclide selectively to targeted cells. They are designed by combining a targeting moiety, a chelator, and a radionuclide. The targeting moiety (e.g. peptide or antibody) binds specifically to receptors or antigens overexpressed on cancer cells. The chelator (e.g. DOTA, or NODAGA) allows stable radiolabelling. The radionuclide, depending on its properties, could either be for imaging if it emits positrons (e.g. [68Ga]) or therapy if it emits β particles (e.g. [177Lu]). Theranostics allow to diagnose patients by Positron Emission Tomography (PET), identify those who would benefit from therapy, treat them with minimal side effects using radionuclide therapy, and image tumour response to therapy. MUC1 is overexpressed in over 90% of BC and 94% of TNBC (Singh and Bandyopadhyay 2007). Since the core peptide of the VNTR is only exposed in cancer cells, targeting this region should allow to selectively reach cancer cells while sparing healthy tissues. Moreover, MUC1 overexpression is associated with metastatic disease and poor prognosis (Jing et al. 2019). Hence MUC1 is a promising target for the development of a theranostic for BC including TNBC. Kelly et al. developed a MUC1-targeting antibody-based theranostic for TNBC. It showed good tumour accumulation, reduction in tumour volume and improved survival in mice (Kelly et al. 2020). However, peptides are preferred as a delivery vehicle for radiopharmaceuticals due to their ease of synthesis, greatly improved pharmacokinetics profiles, and low immunogenicity making them well suited for clinical translation (Fichna and Janecka 2003). Aljammaz and Okarvi developed many MUC1-targeting peptide radiotracers, all based on the PDTRP epitope of the VNTR, labelled with different imaging radionuclides such as [68Ga]-NODAGA-KE-PDTRP (1) and a reported binding affinity Kd on cells of 26.18 ± 5.78 nM (Al Jammaz et al. 2021; Alhabardi et al. 2021; Okarvi and Al Jammaz 2016; Okarvi and AlJammaz 2019; Okarvi et al. 2009). Their most promising compound, a PDTRP-folic acid derivative, dual-targeting MUC1 and folate receptor alpha, showed an excellent tumour uptake of 12% ID/g at 1 h post-injection, in SCID mice bearing MCF7 xenograft (Al Jammaz et al. 2021). We sought to expand upon Aljammaz and Okarvi’s MUC1-targeting peptides, by improving the binding affinity, tumour uptake, and retention. Von Mensdorff-Pouilly and al. identified the three most frequent minimal epitopes of MUC1 as: RPAPGS, PPAHGVT and PDTRP (von Mensdorff-Pouilly et al. 2000). To the best of our knowledge, only the PDTRP sequence was previously explored. Herein we investigate all three sequences and assess their specificity for MUC1 in vitro and in vivo in a mouse model of TNBC. Methods General information and materials All reagents and solvents were purchased form Sigma-Aldrich or Fisher Scientific and used without further purification. Reversed phase HPLC was performed on a Dionex UltiMate 3000 system with 220 and 254 nm UV detection, using a semi-preparative Phenomenex Jupiter 4 μm Proteo 90 Å (250 × 10 mm) column at a flow rate of 4 mL/min, and an analytical Phenomenex Jupiter 4 μm Proteo 90 Å (250 × 4.6 mm) column at a flow rate of 1 mL/min. H2O with 0.1% TFA and MeCN with 0.1% TFA were used as solvents. T47D and HCC70 cell lines were purchased from ATCC and maintained in the recommended growth media. All in vitro and in vivo experiments were conducted with the specified number of replicates. Peptides synthesis All peptides were prepared following standard solid-phase Fmoc peptide chemistry on a Rink Amide ChemMatrix® resin using a Biotage® Initiator+ Alstra™. For the final deprotection and cleavage, the resin was suspended in 5 mL of a mixture of TFA: H2O: TIPS (95, 2.5, 2.5%) and shaken at room temperature for 2 h. After evaporation, the peptide was triturated twice with Et2O (2 × 8 mL), then purified by semi-preparative HPLC starting with 2 min at 5% MeCN with 0.1% TFA, then increasing to 29% organic phase over 13 min. The peptide mass was confirmed by matrix-assisted laser desorption ionization (MALDI), and the purity determined by analytical HPLC. The first analytical HPLC method started with 2 min at 5% of MeCN with 0.1% TFA in water with 0.1%TFA, then increasing the organic phase to 40% over 13 min, followed by an increase to 60% over 3 min, maintaining 60% of organic phase for 3-min followed by 4 min at 5%. This method was later shortened by 3 min while maintaining the same retention time of the compounds by increasing the organic phase from 40% at 15 min to 90% in 0.5 min maintaining it for 3 min, before decreasing to 5% for 4 min. All MALDI spectra and analytical HPLC chromatograms are shown in the Supporting Information (SI) Figs. S1–S9 and S10–S18 respectively. Radiolabelling with [68Ga] [68Ga] was obtained from an IGG100 [68Ge]/[68Ga] generator (Eckert & Ziegler) following the procedure described by Mueller et al. (Mueller et al. 2012). The generator was eluted with ultrapure HCl (0.1 M, 6 mL). The eluate was passed through a strong cation exchange (SCX) cartridge (Agilent, Bond Elut, 100 mg, 1 mL, 40 μm). [68Ga] (~ 370 MBq) was eluted from the cartridge with 300 µL of a mixture of NaCl solution (5 M, 292.7 µL) with HCl (5.5 M, 7.3 µL). NaOAc buffer (250 µL, pH = 5.5) and NODAGA-KEPDTRP (30 µg, 25 nmol, 67.4 µL aqueous solution) were added. The mixture was heated at 90 °C for 10 min. The product was obtained in over 90% yield determined by radio-TLC using a 1:1 mixture of MeOH/NH4OAc (1 M). A radio-TLC chromatogram is shown in the SI Fig. S19. Radiolabelling with [177Lu] [177Lu] was purchased as a non-carrier-added [177Lu]LuCl solution in 0.04 M HCl from ITM Isotope Technologies Munich SE. NaOAc buffer (1 M, pH = 5.5, 10 µL), was mixed with an aqueous solution of the peptide (1 nmol, 1 µL), and [177Lu]LuCl (7.4 MBq, 14 µL). The mixture was then heated at 90 °C for 10 min. The products were obtained in 100% yields (Figs. S20–S27) determined by radio-TLC using citrate (0.1 M in H2O) as eluant. MUC1 knockdown and knockout RNP complexes were made by diluting Cas9 (IDT s.p. Cas9) in duplex buffer (2 µL of duplex per 1 µL of Cas9), with 1.5 µL total per cell line. 1.5 µL of Cas9 mix was added to 1.5 µL of sgRNA (Synthego) and the complex allowed to form for 10 min. HCC70 and T47D cells were cultured, 1–2 million harvested, and media/trypsin spun out. Cells were resuspended in 18 µL of P3 buffer. 18 µL of cells and 2 µL of RNP complex were added to the microcuvette for electroporation. Cells were electroporated using Lonza 4D Nucleofector. The nucleofected cells were transferred into a T25 flask. Cells were expanded, then stained with anti-MUC1-PE antibody (HMFG1 clone), and sorted on the negative population. The cell lines were then analysed with another antibody clone (HMFG2). They were first stained with HMFG2 primary antibody (diluted 1:100 in PBS with 2% FBS for 500 K cells), then washed 3 times with PBS, then stained with secondary FITC-labelled antibody, then washed 3 times with PBS before analysis on a cytometer. Cell uptake studies MUC1-expressing T47D cells and MUC1-KO T47D cells (1 × 105) were placed in test tubes and washed with 40 mM HEPES. [68Ga]-labelled reference peptide and [177Lu]-labelled peptides were diluted in HEPES and 100 µL (1 µM) added per tube. Cells were incubated at room temperature for 30 min, then media was removed and cells washed twice with HEPES. Cells were lysed in radioimmunoprecipitation assay (RIPA) buffer containing protease inhibitors (100 µL). A portion of the cell lysates (80 µL) were used to determine the amount of decay-corrected radioactivity on a gamma counter (Wizard 2 2480 automatic gamma counter, PerkinElmer). The total amount of decay-corrected radioactivity per tube was then calculated and normalized to 106 cells. Saturation binding cell study NODAGA-KE-PDTRP (1), DOTA-PEG4-PDTRP (6), and DOTA-PEG4-RPAPGS (8) were radiolabelled with a specific activity of 1 nmol/37 MBq at 95 °C for 5–10 min yielding 100% purity. T47D and T47D MUC1 KO cells were seeded at 300,000 per well in a 24-well plate. 24 h after seeding, plates were washed with ice cold HBSS (1 mL per well), then 250 µL of tracer was added per well with 8 × 2-fold serial dilutions, with a maximum concentration of 500 nM. Plates were incubated on ice for 2 h, then washed 2 times with 1 mL ice cold HBSS, then lysed with 110 µL RIPA containing protease inhibitor per well. 80 µL were counted on a Perkin Elmer Wizard2 gamma counter, and 20 µL used to quantify protein per well by the BCA assay. CPM values were normalized to the protein content in each well. Surface plasmon resonance NODAGA-KE-PDTRP (1), DOTA-PEG4-PDTRP (6), and DOTA-PEG4-RPAPGS (8) were labelled with natural gallium or lutetium by heating 1 mM with 2.5 equiv. of LuCl or Ga(NO3)3.xH2O respectively at 90 °C for 25 min. Biotinylated MUC1 60-mer synthetic peptide (3 repeats of the 20-mer VNTR sequence) was purchased from LifeTein. It was captured on a Cytiva CAPture chip at a flow rate of 10 µL/min at 200 nM for 25 s to achieve 93 response units (RU). [NatGa]- and [NatLu]-labelled peptides were serially diluted 2-fold in the running buffer (1x PBS with 0.05% Tween20) with a maximum concentration of 100 µM. 5 concentrations of each peptide were injected in a single cycle at flow rate of 50 µL/min with an association time of 60 s, and a dissociation time of 150 s. HMFG2 anti-MUC1 antibody (recognizing the PDTR epitope) was injected as a positive control with a maximum concentration of 100 nM and 5 concentrations injected in single cycle. In vivo tumour models All in vivo experimental procedures were conducted with approval from the Stanford University Institutional Animal Care and Use Committee (APLAC). MUC1-positive and MUC1-knockdown (KD) HCC70 cells (3 × 106, mixed with cultrex 1:1, total volume 100 µl) were injected subcutaneously respectively on the right and left flank of female NU/NU mice (n = 3, age ~ 4 months). Tumours were grown to approximately 220 mm3 (~ 9 weeks after implantation), for PET imaging. PET imaging PET imaging studies were completed using a Bruker Albira PET/SPECT/CT scanner. Mice (n = 3) were anesthetized with isoflurane in oxygen (3% for induction and 1–2% for maintenance). They were injected via their tail vain a net dose of 5.5 ± 0.09 MBq of the reference radiotracer 1, diluted in saline (total volume ~ 100 µL). A dynamic 50.2 min PET scan started just prior to the injection, using 1 × 10, 10 × 300 s frames for reconstruction, followed by a 10 min CT scan. Ex-vivo biodistribution At the end of the PET/CT, mice were sacrificed, the organs collected, weighted, and the activity measured in a gamma counter (Wizard 2 2480 automatic gamma counter, PerkinElmer). The radioactivity in each tissue was expressed as % Injected Dose (ID/g). Statistical analyses Data were expressed as mean ± SD. Statistical significance was determined using one-way ANOVA test or unpaired t test with Welch correction, with P values of less than 0.05 being considered significant. Results Chemistry We synthesized the reference peptide 1 (Table 1) developed by Aljammaz group, (Alhabardi et al. 2021) and 8 new peptides (2–9) by automated solid phase peptide synthesis based on classical Fmoc-chemistry. The identity of all peptides was confirmed by MALDI (Figs. S1–S9) and analytical HPLC showed over 93% purity for all of them (Figs. S10–S18). Table 1Structure of the reference 1 and new compounds 2–9ReferenceProduct structureChelatorLinkerPeptide1NODAGAKEPDTRP2DOTAβAPDTRP3DOTAβAAPDTRP4DOTAβARPAPGS5DOTAβAPPAHGVT6DOTAPEG4PDTRP7DOTAPEG4APDTRP8DOTAPEG4RPAPGS9DOTAPEG4PPAHGVT Structure of the reference 1 and new compounds 2–9 Radiolabelling To maintain the exact published structure, peptide 1 was radiolabelled with [68Ga] and obtained in 90–98% radiopurity, with a molar activity ranging from 8.6 to 12.7 GBq/µmol. A representative radio-TLC is shown in Fig. S19. As our aim was to develop radiopharmaceuticals with potential therapeutic application, we radiolabelled our new peptides 2–8 with [177Lu]. They were 100% radio-pure (Figs. S20–S27), with a molar activity of 7.4 GBq/µmol. Cell uptake studies We first sought to investigate the impact of the different peptide sequences (based on the three minimal epitopes of MUC1) and the linker’s hydrophilicity on cellular uptake. Cell uptake studies showed low cellular uptake for all compounds including the reference product 1. However, it had the highest uptake with a value of 0.063 ± 0.004% uptake/106 cells (Fig. 1). Our newly synthesized products 2–9 showed uptakes ranging from 0.007 ± 0.001 to 0.018 ± 0.005% uptake/106 cells. Regarding the linker’s influence on the uptake, for the APDTRP and PPAHGVT sequences, compounds with a PEG4 linker (7 and 9) demonstrated higher uptake than products with a βA linker; for example, for the PPAHGVT sequence product 9 shows 0.014 ± 0.002% uptake/106 cells whereas 5 has an uptake of 0.008 ± 0.001%. This trend was maintained for the PDTRP sequence but the difference between 2 and 6 was not statistically significant. Regarding the influence of the peptide sequence on the uptake of products with the βA linker, compound 2 with the PDTRP sequence demonstrated higher uptake than 3 with the APDTRP peptide. Moreover, compound 4 comprising RPAPGS showed improved uptake than 3 and 5 with the APDTRP and PPAHGVT sequences with uptake values respectively of 0.012 ± 0.001, 0.07 ± 0.001 and 0.08 ± 0.001% uptake/106 cells. For the PEG4 linker while compound 6 including the PDTRP sequence seems to have higher uptake than the other peptides, the differences were not statistically significant. Cellular uptake of the [68Ga]-labelled reference product 1 and [177Lu]-labelled products 2–9 in MUC1-expressing T47D breast cancer cells. Data of compounds with the same linker were analysed using one-way ANOVA test. Data of products with the same peptide sequence were analysed using unpaired t test with Welch correction Given the overall low uptake across all peptides, we next sought to determine the specificity of uptake to determine whether further medicinal chemistry optimization was warranted. To assess the specificity of the compounds for MUC1, we generated MUC1-knockout (KO) T47D cells. The flow cytometric analysis of the parent and KO cells is shown in Fig. S28A. We compared the uptake of the radiolabelled products 1–9 in the wild type (WT) and KO cells (Fig. 2). All products showed similar uptakes in both cell lines, indicating they are not specific for MUC1. Surprisingly some compounds, such as the published reference 1, showed a slightly higher uptake in the KO cell line compared to the WT with values of 0.072 ± 0.011 and 0.063 ± 0.004% uptake/106 cells. However, for all tested products, none of the differences between the uptake in the WT versus KO cells were statistically significant. Cellular uptake of the [68Ga]-labelled reference product 1 and [177Lu]-labelled products 2–9 in the wild type (WT) and MUC1-knockout (KO) T47D breast cancer cells Binding affinity Given the lack of adequate cellular uptake, we next attempted to determine the binding affinity of representative peptides through a saturation binding assay in T47D parent and KO cells for [177Lu]-NODAGA-KE-PDTRP, [177Lu]-DOTA-PEG4-PDTRP and [177Lu]-DOTA-PEG4-RPAPGS. Similar to the cellular uptake results, all tested compounds, when radiolabelled with the same isotope, demonstrated minimal binding in the T47D parent cells and higher binding in the KO cells (SI, Fig. S29). We similarly attempted to measure the binding affinity by SPR for the representative peptides labelled with non-radioactive gallium or lutetium, [NatGa/NatLu]-NODAGA-KE-PDTRP, [NatGa/NatLu]-DOTA-PEG4-PDTRP, [NatGa/NatLu]-DOTA-PEG4-RPAPGS, but none of the tested compounds, irrespective of the chelated metal, showed any detectable binding at concentrations up to 100 µM (SI, Fig. S30). We confirmed that the epitope was available for binding on the surface of the chip by testing anti-MUC1 HMFG2 clone, which recognizes the PDTR unglycosylated epitope (SI, Fig. S30). In vivo biodistribution Based on the inconsistencies between the published literature data and our cellular uptake results, we lastly sought to determine whether specific uptake was observed in vivo. Given the higher cellular uptake of the reference compound 1 compared to our newly synthesized products, we selected this compound to assess its specificity in nude mice bearing MUC1-positive tumour on the right flank and MUC1 KD tumour on the left flank. HCC70 cells were designated as MUC1 KD based on the double population of cells in the KD population overlapping with MUC1 positive cells (Fig. S28B). Mice (n = 3) were administered a net dose of 5.5 ± 0.09 MBq of the [68Ga]-labelled reference product 1 and a dynamic PET scan acquired for 50.2 min. The images (Fig. 3A) showed that the tracer is predominantly cleared through the renal pathway. No visible tumour uptake was observed for the MUC1 WT or KD tumours. A Representative coronal PET/CT image after intravenous injection of a net dose of 5.59 MBq of the [68Ga]-labelled reference compound 1, in a nude mouse bearing subcutaneous MUC1-positive and MUC1-KD tumours, respectively shown by white arrows on the right and left flanks. Images are summed 0–50.2 min. B Ex-vivo biodistribution following PET/CT imaging after injection of a net dose of 5.5 ± 0.09 MBq (n = 3) in nude mice bearing subcutaneous MUC1-positive and MUC1-KD tumours.Tracer uptake is expressed as % Injected Dose (ID)/g Ex vivo biodistribution Ex vivo biodistribution studies further confirmed that the tracer had no specificity for our target MUC1 as the MUC1 expressing and MUC1-KD tumours had uptakes of 0.41 ± 0.07 and 0.66 ± 0.32%ID/g, respectively as represented in Fig. 3B. It also confirmed the renal clearance of the tracer with a kidney uptake of 9.5 ± 3.0%ID/g. It revealed a high blood uptake (2.56 ± 2.7%ID/g), and a lung and liver uptakes (0.37 ± 0.06 and 0.25 ± 0.13%ID/g) as comparable to the tumour uptake. Discussion Despite the progress made over the past decades, there is still an unmet medical need for safe, efficient, targeted treatment for BC and even more so for TNBC. Indeed, it is usually diagnosed late with a particularly poor prognosis, and due to its lack of ER, PR and HER2, it remains a challenge to develop targeted treatments. Since MUC1 is overexpressed and under-glycosylated in numerous cancers including BC, our aim through this work was to develop a peptide-radiopharmaceutical for the imaging and treatment of BC including TNBC. Based on the promising tumour-targeting PDTRP derivatives developed by Okarvi and Al Jammaz, (Al Jammaz et al. 2021; Alhabardi et al. 2021; Okarvi and Al Jammaz 2016; Okarvi and AlJammaz 2019; Okarvi and Jammaz 2009) we sought to further build upon this work and improve the tumour uptake and retention. Since the VNTR has two β-turns, namely PDTR and PGST, (Apostolopoulos et al. 2015) we hypothesized that the PDTRP peptides were binding to MUC1 through β-turn mimicry. We designed our peptides based on the three minimal epitopes of MUC1, namely, RPAPGS, PPAHGVT and PDTRP (von Mensdorff-Pouilly et al. 2000). Given that it is more likely to form a β-turn conformation when proline is in the second position, we also investigated APDTRP as a potential β-turn mimetic (Hutchinson and Thornton 1994). We incorporated two different linkers, β-alanine and PEG4 to study whether a more lipophilic or hydrophilic chain was preferred as a spacer between the peptide and the chelator. We used DOTA as the chelator, since it is the most commonly used chelator, and was successfully applied for the chelation of [68Ga] and [177Lu] for the FDA-approved theranostic, DOTATATE (Lutathera®) for the treatment of neuroendocrine tumours. Aiming to develop radiopharmaceuticals with potential theranostic application, we labelled our peptides with [177Lu]. The cellular uptakes in MUC1-expressing T47D cells seem to indicate that a more hydrophilic linker (PEG4) allows a higher uptake for the PDTRP and APDTRP sequence. The results also show that for compounds with the lipophilic linker (β-Ala) the PDTRP peptide has a better uptake than the APDTRP sequence, and the RPAPGS has a better uptake than both APDTRP and PPAHGVT. However, given the low overall uptake for all compounds, more studies are required to establish reliable structure-activity relationships. Furthermore, all compounds showed a similar uptake in MUC1-expressing and MUC1-KO cells, indicating the absence of specificity for the target. Our attempts to determine the binding affinity of several peptides incorporating both the gallium and lutetium metals and including the reference compound only showed their lack of affinity for MUC1 and emphasize that a lack of binding is not the result of incorporation of different metals or use of different chelators. Surprised by these results, as numerous articles report different PDTRP derivatives specifically targeting MUC1 with high affinity in vitro and in vivo, (Al Jammaz et al. 2021; Alhabardi et al. 2021; Okarvi and Al Jammaz 2016; Okarvi and AlJammaz 2019; Okarvi and Jammaz 2009) we decided to further investigate the reference compound 1 in mice bearing subcutaneous tumours of MUC-expressing and MUC1-KD HCC70 cells. In vivo studies further confirmed the absence of specificity of the reference product 1 for MUC1, as both tumours had comparable and overall minimal tracer uptake. Our flow and SPR data demonstrate that antibodies do bind to MUC1 indicating the accessibility of the VNTR. The binding assays in cells with variable glycosylation patterns and SPR with an unglycosylated 60-mer peptide representing the core protein indicate the lack of peptide binding is not due to the difference in glycosylation patterns. Taken together, our results indicate that the low cellular/tumour uptake of the radiopeptides is most likely due to the lack of affinity of the peptides for MUC1 and not the variable glycosylation patterns or VNTR accessibility. One limitation of our study is the knockdown of HCC70 cells instead of a complete knockout, so we cannot rule out the influence of the remaining MUC1-positive cells in the KD tumours on the results. However, given the low overall uptake, absence of selectivity in vitro in KO cells, low binding in T47D cells and absence of binding by SPR, this very likely does not change the results of the study. Hence despite the published results with the reference compound 1, (Alhabardi et al. 2021) our extensive in vitro and in vivo investigations using both MUC1-expressing and MUC1-KD/KO cells, demonstrate that this radiopeptide had no significant or specific tracer uptake. Conclusions Herein we have investigated different minimal epitopes of MUC1 for the development of a radiopharmaceutical with potential for theranostic application for BC. Despite promising previously published results, our cell uptake studies showed that all studied radiopharmaceuticals, including the reference compound described in the literature, had low cellular uptake and no specificity for MUC1. Our attempts at determining the binding affinities of different peptides only confirmed their poor binding. In vivo studies of the reference further proved that it had no specificity for our target in a mice model of TNBC. These results suggest that designing peptidomimetics for MUC1 based on its epitopes will likely not yield a high affinity binder. Our ongoing efforts focus on using peptide phage display to identify novel peptides with affinity for this promising target. Electronic Supplementary Material Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements We thank the Cyclotron and Radiochemistry Facility and the Small Animal Imaging facility at Stanford. We dedicate this manuscript to the loving memory of our dear colleague Ananth Srinivasan. Abbreviations Breast cancers Computed tomography Oestrogen receptors Human epidermal growth factor receptor 2 High performance liquid chromatography Equivalent Injected dose Knockdown Knockout Matrix-assisted laser desorption ionization Mucin 1 N-terminal subunit of MUC1 Programed cell death 1 Polyethylene glycol 4 Positron emission tomography Progesterone receptors Radioimmunoprecipitation assay Ribonucleoprotein Strong cation exchange Surface plasmon resonance Trifluoroacetic acid Triisopropylsilane Thin layer chromatography Triple-negative breast cancer Variable number tandem repeat Wild type Author contributions RM contributed to the design of the study, synthesised and radiolabelled the compounds, contributed to the in vivo and ex vivo studies, analysed the data and wrote the manuscript. ML contributed to the design of the study, executed the in vitro studies, contributed to the in vivo and ex vivo studies, analysed the results, contributed to the manuscript. PC realized the in vivo and ex vivo studies. CB designed and supervised the study, analysed the results and reviewed the manuscript. Funding This work was supported by the Stanford Brain Metastasis Consortium (SBMC) Seed Grant Program, funded by the NIH NCI U54 Metastasis Network program (U54CA261717) and R37CA278744 (CB). This work was supported by the National Institutes of Health grant 1S10OD034265 and Stanford Community of Shared Research Platforms (C-ShaRP) award for the Bruker Albira Si PET/SPECT/CT. Data availability All data generated or analysed during this study are included in this published article and its supplementary information files. Declarations Ethics approval and consent for participation All animal experiments were realized following the guidelines for the Care and Use of Laboratory Animals of Stanford University and approved by The Administrative Panel on Laboratory and Animal Care (APLAC)/Institutional Animal Care and Use Committee (IACUC) of Stanford University. This research does not involve human subjects and this is therefore not applicable. No client owned animals were used in this study. Consent for publication Not applicable. Competing interests The authors declare that they have no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement All data generated or analysed during this study are included in this published article and its supplementary information files.