[225Ac]Ac- and [111In]In-DOTA-trastuzumab theranostic pair: cellular dosimetry and cytotoxicity in vitro and tumour and normal tissue uptake in vivo in NRG mice with HER2-positive human breast cancer xenografts

Background Trastuzumab (Herceptin) has improved the outcome for patients with HER2-positive breast cancer (BC) but brain metastases (BM) remain a challenge due to poor uptake of trastuzumab into the brain. Radioimmunotherapy (RIT) with trastuzumab labeled with α-particle emitting, 225Ac may overcome this challenge by increasing the cytotoxic potency on HER2-positive BC cells. Our first aim was to synthesize and characterize [111In]In-DOTA-trastuzumab and [225Ac]Ac-DOTA-trastuzumab as a theranostic pair for imaging and RIT of HER2-positive BC, respectively. A second aim was to estimate the cellular dosimetry of [225Ac]Ac-DOTA-trastuzumab and determine its cytotoxicity in vitro on HER2-positive BC cells. A third aim was to study the tumour and normal tissue uptake of [225Ac]Ac-DOTA-trastuzumab using [111In]In-DOTA-trastuzumab as a radiotracer in vivo in NRG mice with s.c. 164/8-1B/H2N.luc+ human BC tumours that metastasize to the brain. Results Trastuzumab was conjugated to 12.7 ± 1.2 DOTA chelators and labeled with 111In or 225Ac. [111In]In-DOTA-trastuzumab exhibited high affinity specific binding to HER2-positive SK-BR-3 human BC cells (KD = 1.2 ± 0.3 × 10–8 mol/L). Treatment with [225Ac]Ac-DOTA-trastuzumab decreased the surviving fraction (SF) of SK-BR-3 cells dependent on the specific activity (SA) with SF < 0.001 at SA = 0.74 kBq/µg. No surviving colonies were noted at SA = 1.10 kBq/µg or 1.665 kBq/µg. Multiple DNA double-strand breaks (DSBs) were detected in SK-BR-3 cells exposed to [225Ac]Ac-DOTA-trastuzumab by γ-H2AX immunofluorescence microscopy. The time-integrated activity of [111In]In-DOTA-trastuzumab in SK-BR-3 cells was measured and used to estimate the absorbed doses from [225Ac]Ac-DOTA-trastuzumab by Monte Carlo N-Particle simulation for correlation with the SF. The dose required to decrease the SF of SK-BR-3 cells to 0.10 (D10) was 1.10 Gy. Based on the D10 reported for γ-irradiation of SK-BR-3 cells, we estimate that the relative biological effectiveness of the α-particles emitted by 225Ac is 4.4. Biodistribution studies in NRG mice with s.c. 164/8-1B/H2N.luc+ human BC tumours at 48 h post-coinjection of [111In]In-DOTA-trastuzumab and [225Ac]Ac-DOTA-trastuzumab revealed HER2-specific tumour uptake (10.6 ± 0.6% ID/g) but spleen uptake was high (28.9 ± 7.4% ID/g). Tumours were well-visualized by SPECT/CT imaging using [111In]In-DOTA-trastuzumab. Conclusion We conclude that [225Ac]Ac-DOTA-trastuzumab exhibited potent and HER2-specific cytotoxicity on SK-BR-3 cells in vitro and HER2-specific uptake in s.c. 164/8-1B/H2N.luc+ human BC tumours in NRG mice, and these tumours were imaged by SPECT/CT with [111In]In-DOTA-trastuzumab. These results are promising for combining [111In]In-DOTA-trastuzumab and [225Ac]Ac-DOTA-trastuzumab as a theranostic pair for imaging and RIT of HER2-positive BC. Supplementary Information The online version contains supplementary material available at 10.1186/s41181-023-00208-0.


Background
There is great interest in treating cancer with biomolecules including monoclonal antibodies (mAbs) and their fragments or peptides labeled with α-particle emitting radionuclides targeted to receptors overexpressed on tumour cells (Aghevlian et al. 2017;Morgenstern et al. 2018).Actinium-225 ( 225 Ac; t 1/2 = 10 d) decays by α-particle emission (Eα = 5.8 MeV) to stable 209 Bi through a series of daughter radionuclides that emit α-particles ( 221 Fr, 211 At, 213 Bi, 213 Po) or β-particles ( 213 Bi, 209 Tl, 209 Pb) (Aghevlian et al. 2017).An attractive property of α-particles, which makes them potent for cancer treatment is their high linear energy transfer (LET = 50-230 keV/µm), since several MeV of α-particle energy are deposited over a range of only 28-100 µm (~ 3-10 cell diameters).In contrast, the β-particles emitted by 177 Lu (Eβ max = 0.5 MeV, t 1/2 = 6.7 d), which has been more widley used for cancer treatment, deposit their energy over a distance up to 1.8 mm (~ 180 cell diameters), resulting in very low LET < 0.1 keV/µm, and consequently lower cytotoxic potency.Furthermore, α-particles directly inflict lethal DNA double-strand breaks (DSBs) in cancer cells, in addition to causing indirect DNA damage mediated by reactive oxygen species (ROS), while β-particles rely on ROS-mediated DNA damage, making treatment with α-particles less susceptible to tumour hypoxia (Seidl 2014;Wulbrand et al. 2013).Thus, α-particles are more powerful and more precise for killing cancer cells than β-particles and may be able to overcome resistance to other types of radiation used for targeted radiotherapy of cancer.Indeed, a patient with widespread metastatic prostate cancer resistant to treatment with β-particle emitting, 177 Lu-labeled PSMA-617, a radiopeptide targeted to prostate specific membrane antigen (PSMA), achieved a complete remission when treated with α-particle emitting, 225 Aclabeled PSMA-617 with disappearance of tumours on positron-emission tomography (PET) scans using [ 68 Ga]Ga-PSMA11 and serum prostate specific antigen (PSA) levels decreasing almost 3000-fold to < 0.1 ng/mL (Kratochwil et al. 2016).
Since radiation kills cancer cells by a mechanism that is independent of inhibition of the growth promoting function of HER2, RIT may overcome resistance to these therapies.Our group reported that trastuzumab modified with nuclear translocation sequence (NLS) peptides and modified with diethylenetriaminepentaacetic acid (DTPA) complexed to the Auger electron-emitting 111 In ([ 111 In]In-DTPA-trastuzumab-NLS) was more effective than trastuzumab for treatment of human HER2-positive BC tumours in mice (Costantini et al. 2010) and killed HER2-positive BC cells in vitro that were resistant to trastuzumab (Costantini et al. 2008).Auger electrons are analogous to α-particles in that they are a high LET form of radiation (LET = 4-26 keV/µm), although these electrons have much lower energy (< 20 keV) and a subcellular range (nanometers to micrometers) (Aghevlian et al. 2017;Ku et al. 2019).The powerful cytotoxic properties of α-particles emitted by [ 225 Ac]Ac-DOTA-trastuzumab may make this agent more effective than trastuzumab for treatment of HER2-positive BC, particularly when delivered to the brain, and may overcome resistance of these tumours to HER2-targeted agents.Two strategies to improve the uptake of [ 225 Ac]Ac-DOTA-trastuzumab in BM are: i) MRI-guided focused ultrasound (MRIg-FUS) which transiently and spatially disrupts the BBB around BM to enhance the delivery of intravenously (i.v.) injected radioimmunoconjugates (RICs) and ii) convection-enhanced delivery (CED) in which RICs are infused directly into BM using a catheter.We recently reported that MRIg-FUS increased the uptake of i.v.injected [ 111 In]In-DTPA-trastuzumab in BM in patients with HER2-positive BC by up to fourfold (Meng et al. 2021).Radiolabeled trastuzumab has not been delivered to BM in patients by CED, but trastuzumab infused into HER2-positive human BC tumours in the brain in athymic rats increased survival by twofold compared to intraperitoneally (i.p.) administered trastuzumab, suggesting CED improved delivery to these tumours (Grossi et al. 2003).The safety of CED delivery of monoclonal antibodies to tumours in the brain is shown by a Phase I study of 124 I-labeled 8H9 antibodies administered by CED in patients with diffuse intrinsic pontine glioma, which found no dose-limiting toxicity (Souweidane et al. 2018).However, the safety of CED for delivery of [ 225 Ac]Ac-DOTA-trastuzumab to BM in patients with HER2-positive BC remains to be evaluated.
Our objectives in this study were to: (i) synthesize and characterize [ 225 Ac]Ac-DOTAtrastuzumab, (ii) determine its ability to decrease the clonogenic survival fraction (SF) of HER2-positive SK-BR-3 human BC cells in vitro compared to trastuzumab and measure DNA DSBs caused by 225 Ac and its daughters, (iii) estimate the cellular dosimetry of [ 225 Ac]Ac-DOTA-trastuzumab and correlate the SF with the radiation absorbed dose in SK-BR-3 cells, and (iv) assess its tumour and normal tissue uptake in vivo in NOD-Rag1 null IL2rg null (NRG) mice with subcutaneous (s.c.) HER2-positive 164/8-1B/ H2N.luc + human BC xenografts.There have been a few previously reported preclinical studies of 225 Ac-labeled trastuzumab for treatment of HER2-positive BC (Ballangrud et al. 2004;Borchardt et al. 2003;Yoshida et al. 2016), but none have reported the absorbed doses in HER2-positive SK-BR-3 BC cells or determined the relationship between dose and the SF in clonogenic survival assays, or measured DNA DSBs caused by emission of α-particles.Moreover, 164/8-1B/H2N.luc+ cells were selected here for tumour and normal tissue uptake studies, because tumour xenografts established from these cells in immunocompromised mice have been shown to have a high affinity for metastasis to the brain (Milsom et al. 2013).This will enable our future studies of [ 225 Ac] Ac-DOTA-trastuzumab for RIT of BM from HER2-positive BC in mouse tumour xenograft models.

Synthesis and characterization of immunoconjugates
Trastuzumab (Herceptin ® ; Hoffman La Roche, Mississauga, ON, Canada) was purchased from the Princess Margaret Cancer Centre (Toronto, ON, Canada) pharmacy and reconstituted with the supplied Bacteriostatic Water for Injection, USP to 21 mg/ mL following the manufacturer's directions.Reconstituted trastuzumab was bufferexchanged into 0.1 M NaHCO 3 buffer, pH 8.2 at room temperature (RT) on an Amicon Ultra 0.5 ultracentrifugal filtration device (30 kDa MW cut-off ) and the protein concentration measured by determining the absorbance at 280 nm (A 280 of a 1 mg/mL solution = 1.47).Buffer-exchanged trastuzumab (15-20 mg/mL) was then conjugated to 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA) by reaction with a 10-, 30 or 60-fold molar excess of DOTA N-hydroxysuccinimide ester (DOTA-NHS; Macrocyclics, Dallas, TX) for 2 h at RT. DOTA-trastuzumab was purified and buffer-exchanged into 0.1M NH 4 Ac buffer, pH 5.5 by transferring the reaction mixture to an Amicon Ultra 0.5 ultracentrifugal device, diluting with 0.1M NH 4 Ac buffer, pH 5.5 to a volume of 500 µL and centrifuging at 7000×g for 5 min.The retentate was rediluted with 300 µL of 0.1M NH 4 Ac buffer, pH 5.5 and ultracentrifugation was repeated a total of 6 times.Purified DOTA-trastuzumab was recovered and the protein concentration measured and adjusted to 20-30 mg/mL with 0.1M NH 4 Ac buffer, pH 5.5.Irrelevant human IgG 1 (Sigma-Aldrich Product No. 15154) was conjugated to DOTA by reaction of 5 mg (200 µL) in 0.1 M NaHCO 3 buffer, pH 8.2 with a 30-fold molar excess of DOTA-NHS (Macrocyclics) for 2 h at RT, and purifying the DOTA-IgG 1 immunoconjugates on an Amicon Ultra 0.5 ultracentrifugal device (Amicon).
The purity and homogeneity of DOTA-trastuzumab was assessed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) on a 4-20% Mini-Protean Tris/ glycine mini-gel (Bio-Rad, Hercules, CA, USA).The bands were stained with Coomassie Brilliant Blue G-250 (BioRad).Broad range MW markers (BLUeye Prestained Protein Ladder, FroggaBio, Concord, ON, Canada) were electrophoresed to calibrate the gel.In addition, size-exclusion HPLC (SE-HPLC) was performed on a BioSep SEC-S2000 column (Phenomenex, Torrance, CA, USA) eluted with 0.1 M NaH 2 PO 4 buffer, pH 7.0 at a flow rate of 0.8 mL/min with UV detection at 280 nm.The number of DOTA conjugated per trastuzumab molecule after reaction with a 30-fold excess of DOTA-NHS was determined by analysing the mass spectra of trastuzumab and DOTA-trastuzumab by matrixassisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) at the AIMS Mass Spectrometry Laboratory in the Department of Chemistry, University of Toronto (Toronto, ON, Canada).The mass difference for DOTA-trastuzumab compared to trastuzumab determined by the m/z value of the single-protonated (1 + charge state) mass spectral peaks was divided by the mass of DOTA (387.41 mass units) to calculate the number of DOTA conjugated to trastuzumab.The number of DOTA conjugated per trastuzumab was determined for reaction of trastuzumab with a 10-fold, 30-fold and 60-fold excess of DOTA-NHS by measuring the conjugation efficiency (CE) following trace labeling with 111 In by determining the proportion of [ 111 In]In-DOTAtrastuzumab and unconjugated [ 111 In]In-DOTA by instant thin layer-silica gel chromatography (ITLC-SG) developed in 0.1 M sodium citrate buffer, pH 5.5 as described in the next section.This CE was multiplied by the molar ratio for the reaction to calculate the number of DOTA per trastuzumab.

Cell binding of radioimmunoconjugates (RICs)
The HER2 binding properties of [ 111 In]In-DOTA-trastuzumab were assessed in a direct (saturation) radioligand cell-binding assay.The HER2-binding of [ 225 Ac] Ac-DOTA-trastuzumab was not determined since it was assumed that substitution of 225 Ac for 111 In would not change these properties.Briefly, 1 × 10 6 SK-BR-3 cells suspended in phosphate buffered saline (PBS), pH 7.4 in 1.5 mL Microtubes (Diamed Lab Supplies, Mississauga, ON, Canada) were incubated with increasing concentrations (0.073 to 300 nmoles/L) of [ 111 In]In-DOTA-trastuzumab in PBS, pH 7.4 in the absence or presence of a 50-fold molar excess of trastuzumab for 3.5 h at 4 °C to measure total binding (TB) and non-specific binding (NSB), respectively.The tubes were mixed by gentle agitation every 30 min to minimize adsorpton of [ 111 In]In-DOTA-trastuzumab onto the walls of the tubes.The tubes were centrifuged at 2000 rpm (400×g) for 5 min on an Eppendorf Centrifuge Model 5424 (Thermo Fisher Scientific), and the supernatant was collected.Cell pellets were then rinsed with PBS and centrifuged again to collect the supernatant.This procedure was repeated once more.Once the supernatant from all rinses and the cell pellet was isolated, the combined supernatant containing unbound activity was transferred using a micropipettte to a γ-counting tube.The cells were resuspended in 500 µL of ice-cold PBS, pH 7.4 and recentrifuged.The supernatant was removed and added to the previously recovered supernatant.This process was repeated a total of two times.The combined supernatants and cell pellets were measured in a γ-counter.NSB was subtracted from TB to obtain specific binding (SB).[ 111 In]In-DOTA-trastuzumab specifically bound to SK-BR-3 cells (pmoles) was plotted versus the concentration of free (unbound) [ 111 In]In-DOTA-trastuzumab (nmoles/L) and the resulting curve fitted to a one-site-receptor-binding model using Prism Ver.8.0 software (GraphPad, San Diego, CA, USA) to estimate the dissociation constant (K D ) and maximum number of binding sites per cell (B max ).

Cytotoxicity and DNA DSBs in vitro
The cytotoxicity of [ 225 Ac]Ac-DOTA-trastuzumab on HER2-overexpressing BC cells in vitro was determined in a clonogenic survival assay.SK-BR-3 cells were selected for these assays due to their ability to form colonies that have strong adherence to 6-and 24-well plates during methylene blue staining making them suitable for clonogenic survival assays, while 164/8-1B/H2N.luc+ human BC cells which were used to estabish tumour xenografts in NRG mice did not adhere well to plates, and were not practical for conducting clonogenic survival assays.Briefly, 1.0 × 10 5 SK-BR-3 cells were cultured overnight in wells in a 24-well plate (Sarstedt, Nümbrecht, Germany), then treated for 3 h with 20 nmoles/L (3.0 µg/mL) of [ 225 Ac]Ac-DOTA-trastuzumab (SA = 0.148-1.665kBq/µg) or DOTA-trastuzumab in 300 µL of medium.To assess HER2-specific cytotoxicity, cells were exposed to [ 225 Ac]Ac-DOTA-trastuzumab (SA = 1.665 kBq/µg) in the presence of a 100-fold molar excess of DOTA-trastuzumab to block HER2.Then 1,800-6,000 cells were seeded into wells in 6-well culture plates (Sarstedt) containing 3 mL of fresh medium supplemented with 10% FBS per well to obtain a measurable number of colonies after culturing for 12 d at 37 °C and 5% CO 2 .Colonies were stained and fixed with methylene blue (1% in a 1:1 mixture of ethanol and water) and imaged on a ChemiDoc gel imaging system (Bio-Rad, Mississauga, ON, Canada).Surviving colonies (≥ 50 cells) were counted using ImageJ software (U.S. National Institutes of Health, Bethesda, MD, USA) and a customized macro (Cai et al. 2011).The plating efficiency (PE) was calculated by dividing the number of colonies by the number of cells seeded.The surviving fraction (SF) was calculated as the PE of treated cells divided by the PE of untreated cells.The SF was plotted versus SA of [ 225 Ac]Ac-DOTA-trastuzumab or absorbed dose (D) in the cell (Gy) calculated as described under Cellular Dosimetry.The SF versus dose curve was fitted to a linear quadratic equation: SF = e −αD−βD 2 , where α and β are curve fitting parameters.

Cellular dosimetry
The radiation absorbed dose in the cell from treatment of SK-BR-3 cells in vitro in clonogenic survival and γ-H2AX assays with [ 225 Ac]Ac-DOTA-trastuzumab was estimated based on the cellular uptake of 225 Ac, which was derived by measurement of the cell uptake of [ 111 In]In-DOTA-trastuzumab.Briefly, 1 × 10 5 SK-BR-3 cells were seeded into wells in 24-well plates in 300 μL of medium and cultured overnight.The medium was then replaced with 300 μL of fresh medium containing 20 nmoles/L (3.0 µg/mL) of [ 111 In] In-DOTA-trastuzumab at a SA of 37, 74 and 111 kBq/µg, and cells were incubated for 60, 120 or 180 min at 37 °C and 5% CO 2 .The culture medium containing unbound activity was removed.The wells were rinsed twice with cold PBS, pH 7.4.The cells were then lysed with 300 µL/well of 0.1 M NaOH at RT for 15 min and the cell lysates transferred to γ-counting tubes.The wells were rinsed twice with double-distilled water (ddH 2 O) and the rinses were added to the cell lysates.Cell bound activity in the lysates was measured in a γ-counter.Uptake of [ 111 In]In-DOTA-trastuzumab (74 kBq/µg) by SK-BR-3 cells incubated with a 100-fold molar excess of trastuzumab was determined to assess HER2-specific cell uptake.Cells incubated with medium alone were detached at 60, 120 or 180 min by trypsinization and counted by a TC20 automated cell counter (Bio-Rad, Mississauga, ON, Canada) to estimate the number of cells per well for incubation with the RICs.The cellular uptake of [ 111 In]In-DOTA-trastuzumab at 60, 120 or 180 min was expressed as percent incubated activity [%IA(t)].Assuming that [ 225 Ac]Ac-DOTA-trastuzumab exhibits the same uptake kinetics as [ 111 In]In-DOTA-trastuzumab at the same mass concentration (20 nmoles/L; 3.0 µg/mL), 225 Ac activity in SK-BR-3 cells (A Cell ; Bq/ cell) or in the surrounding medium (A M ; Bq) at each incubation time point was calculated using Eq. 1 and 2, respectively: where IA was the 225 Ac activity incubated with the cells (Bq), IA% is the percentage of activity taken up by cells at the selected incubation times, t (min), and k is the decay constant of 225 Ac (4.81 × 10 -5 min −1 ).
The time-integrated 225 Ac activity in monolayer SK-BR-3 cells (Bq × min/cell) and in the surrounding medium during the 180-min incubation period (Ã ML, Cell, 0-180 min , Ã M, 0-180 min , respectively) were calculated from the area under curve (AUC) of a plot of A Cell (t)/cell or A M (t) versus time (min).The time-integrated activity of 225 Ac in SK-BR-3 cells for the 12-d colony formation period (Ã Cell180min-12d ; Bq × min/cell) was calculated using Eqs.3-5: (1) where, T b the biological half life, equivalent to the doubling time of SK-BR-3 cells (1.55 d = 2246 min) and T p is the physical half life of 225 Ac (10 d = 14,400 min).Thus, the effective half life (T e ) and decay constant (k e ) were calculated as 1,943 min and 3.568 × 10 -4 min −1 .Ã in Bq × min was converted to Bq × s by multiplying by the unit conversion factor = 60 s/min.
During the colony formation period, since the cells were seeded in fresh medium, Ã M, 3 h-12d was assumed as zero.
The radiation absorbed dose in SK-BR-3 cells (modeled as 18 µm diameter spheres closely packed as a monolayer) in wells in a 24-well plate incubated with [ 225 Ac]Ac -DOTA-trastuzumab (20 nmoles/L, 3.7 µg/mL, 0-111 kBq/µg) in 300 µL of medium (modeled as cylinder of 1.54 cm diameter and 0.161 cm high) for 180 min, followed by 12 d colony formation from a single seeded cell (modeled as 18 µm diameter sphere) was estimated by cellular dosimetry.S values of cell to cell (S C←C ), cell monolayer to cell (S C←CML ) and medium to cell (S C←M ) were calculated using Monte Carlo N-Particle (MCNP) code Ver.6.1 (Los Alamos National Laboratory, Los Alamos, NM, USA) (Cai et al. 2017).It was assumed that 225 Ac was homogenously distributed in the cell, the cell monolayer or cylinder, and all 225 Ac decay daughters remained at the same location as 225 Ac.The α-particle emission spectra for 225 Ac and its daughters were obtained from the Medical Internal Radiation Dose (MIRD) decay schemes (Eckerman and Endo 2008) and used as particle input in the simulation.5 × 10 6 α-particles were launched for each simulation and the energies deposited in a single cell, monolayer cells and medium were tallied to calculate values of S C←C , S C←CML and S C←M in Gy × Bq −1 s −1 .Neither photon nor electron emissions were included in the calculation for simplicity, since the energy of α-particle emission per nuclear transition of 225 Ac including its daughters (27.996MeV) is 900 times greater than the energy of all emitted electrons and photons (0.0311 MeV) (Eckerman and Endo 2008).The doses (D SF in Gy) in the cell (target) from three sources (medium, cell monolayer, cell) during the clonogenic assay were calculated using Eq. ( 6): where the units of Ã were Bq × s.
The doses (D γ-H2AX in Gy) in the cell (target) from two sources (medium, cell monolayer) during the γ-H2AX assay were calculated using Eq. ( 7): (3) ) 3 mice per group were imaged by SPECT/CT, then all mice were sacrificed under 2% isoflurane in O 2 anaesthesia and the tumour and samples of blood and other normal tissues obtained, weighed and 111 In measured in a γ-counter.Tissue uptake of 111 In was expressed as percent injected dose/g (%ID/g).Mice were anaesthetized using 2% isoflurane in O 2 and were imaged in a prone position on a trimodality NanoScan ® SPECT/CT/PET system (Mediso, Budapest, Hungary) equipped with 4 NaI (Tl) detectors.SPECT employed a 40 s/frame acquisition time resulting in a scan duration of 45 min.CT images were acquired using parameters of 50 kVp X-rays, 980 μA and 300 ms exposure time, isotropic voxel size of 125 μm and maximum field-of-view with 1:4 binning.Images were reconstructed using TeraTomo 3D Normal Dynamic Range Monte Carlo-based reconstruction protocol with a 128 × 128 reconstruction matrix, with three subsets of data undergoing 48 iterations applied with CT-based attenuation and scatter correction.SPECT and CT images were co-registered by InterView Fusion software (Ver.3.09; Mediso).

Statistical analysis
Data were expressed as mean ± SD.Statistical analyses were performed using Prism Ver.9.5 software (GraphPad) with significance tested by Welch's t-test or one-way ANOVA (P < 0.05).

Radioimmunoconjugates (RICs)
Reaction of trastuzumab (30 mg/mL) with DOTA-NHS at a 30:1 molar excess resulted in conjugation of 12.7 ± 1.2 DOTA per trastuzumab molecule measured by MALDI-TOF analysis.Reaction of trastuzumab at a 10:1, 30:1 and 60:1 molar excess resulted in 1.9 ± 0.7.5.5 ± 2.2 and 10.1 ± 1.9 DOTA per trastuzumab molecule measured by ITLC-SG after trace labeling with 111 In (Additional file 1: Fig. S1).DOTA-trastuzumab and trastuzumab migrated as a single major protein band by SDS-PAGE (MW ~ 185 kDa) (Fig. 1a).SE-HPLC with UV detection at 280 nm showed a single peak for trastuzumab (Fig. 1b) and DOTA-trastuzumab (Fig. 1c) with similar retention times (t R ) of 14.90 and 14.86 min, respectively.Immunoconjugates synthesized by reaction of trastuzumab (15 mg/mL) with a 60-fold molar excess of DOTA-NHS were used to study the effects of incubation temperature and time, SA and protein concentration on LE with 111 In.There was a trend towards higher LE with 111 In at increasing temperature with the highest LE (95.2 ± 1.2%) obtained at 1.5 h and 40 °C and SA = 0.1 MBq/µg (Additional file 1: Fig. S2a).There were no significant differences (P > 0.05) in 111 In LE for 0.5, 1, 1.5, 2 or 3 h incubation times at 40 °C and SA = 0.1 MBq/µg (Additional file 1: Fig. S2b).There were no significant differences in LE with 111 In over a SA range of 0.1 MBq/µg to 0.25 MBq/ µg at 40 °C for 0.5 h (Additional file 1: Fig. S2c).Protein concentration did not affect LE with 111 In over a concentration range of 2.5-15.0mg/mL at 40 °C for 0.5 h (Additional file 1: Fig. S2d).Based on these results, DOTA-trastuzumab synthesized by reaction of trastuzumab (30 mg/mL) with a 30:1 molar excess of DOTA-NHS resulting in 12.7 ± 1.2 DOTA conjugated per trastuzumab molecule and was labeled with 111 In by incubation for 0.5 h at 40 °C at a SA = 0.1-0.35MBq/µg and protein concentration = 15 mg/mL for all subsequent studies.The LE and final RCP of [ 111 In]In-DOTA-trastuzumab labeled under these conditions was 94.3 ± 8.1%.The LE of DOTA-trastuzumab prepared at 10:1, 30:1 or 60:1 molar excess of DOTA-NHS:trastuzumab under these same conditions except for a longer 1.5 h incubation time was 90.7 ± 0.1%, 93.8 ± 1.6% and 95.2 ± 1.2%, respectively (Additional file 1: Fig. S3).The conditions chosen for 225 Ac labeling of DOTA-trastuzumab (37 °C for 2 h at a protein concentration of 5 mg/mL) were based on the protocol described by Macguire et al. (Maguire et al. 2014).Under these conditions, the LE of DOTA-trastuzumab with 225 Ac at a SA = 5.3 kBq/µg was 76.4 ± 3.1%; (n = 3).Following purification by ultrafiltration on an Amicon Ultra 0.5 device, the final RCP of [ 225 Ac]Ac-DOTA-trastuzumab was 95.9 ± 0.9 (n = 3).Irrelevant human IgG 1 was  111 In or 225 Ac to a final RCP of 95.7 ± 0.1% and 94.4 ± 0.2%, respectively.

Biodistribution and SPECT/CT imaging studies
The  5a. 111 In was used to measure the tumour and normal tissue uptake of the RICs due to the very low amount (4 kBq) of 225 Ac that could be safely injected, assuming that coadministered 111 In and 225 Ac-labeled RICs distribute equivalently.The γ-photons of 111 In [Eγ = 171 keV (90.7%) and 245 keV (94.1%) were used to acquire SPECT/CT images (Fig. 5b).The tumour uptake of [ 111 In]In-DOTA-trastuzumab was 2.5-fold significantly greater than irrelevant [ 111 In]In-DOTA-IgG 1 (10.6 ± 0.6% ID/g vs. 4.3 ± 0.7% ID/g respectively, P < 0.0001).The highest normal tissue uptake was in the spleen but this was 3.5-fold significantly lower for [ 111 In]In-DOTA-trastuzumab than [ 111 In]In-DOTA-IgG 1 (28.9 ± 7.4 vs. 100.6± 36.2%ID/g; P = 0.003).Liver uptake of [ 111 In]In-DOTA-trastuzumab was 1.4-fold significantly lower than [ 111 In]In-DOTA-IgG 1 (9.2 ± 0.8 vs. 12.6 ± 1.2% ID/g; P = 0.001).Blood and heart activity were 4.1-fold and 2.0-fold significantly higher for [ 111 In]In-DOTA-trastuzumab than [ 111 In]In-DOTA-IgG 1 , respectively (11.1 ± 2.7 vs. 1.9 ± 1.9% ID/g; P < 0.001 and 3.7 ± 0.7 vs. 1.9 ± 0.5; P = 0.002, respectively).Tumours   S1), but MALDI-TOF is considered to be more accurate, since it directly measures the increase in molecular weight associated with DOTA conjugation of trastuzumab.Despite this relatively high level of DOTA modification, the immunoconjugates migrated as a single band on SDS-PAGE and a single peak by SE-HPLC (Fig. 1).In addition following labeling with 111 In, high affinity specific binding of [ 111 In]In-DOTAtrastuzumab to HER2-positive SK-BR-3 cells was observed (Fig. 2).The K D of [ 111 In] In-DOTA-trastuzumab (1.2 ± 0.3 × 10 -8 mol/L) was approximately two-fold greater than previously reported for trastuzumab (5 × 10 -9 mol/L).DOTA-NHS reacts with the ε-amino group on lysine amino acids or the N-terminal amines of trastuzumab.There are four N-terminal amines -one on each of the two heavy and light chains and 88 lysines in the trastuzumab molecule, but only one lysine is present in the complementarity-determining region (CDR) at heavy chain position K 65 (Chen et al. 2016).This may explain the minimal impact of trastuzumab modification with multiple DOTA chelators on HER2-binding affinity.Guleria M et al. measured a similar K D (1.36 × 10 -8 mol/L) for binding of [ 177 Lu]Lu-DOTA-trastuzumab conjugated with 6 DOTA to HER2-positive SK-OV-3 human ovarian cancer cells (Guleria et al. 2021).DOTA is often used to complex 225 Ac to biomolecules, since DOTA forms stable complexes with tripositive metal ions (ie.Ac 3+ ) (Viola-Villegas and Doyle, 2009).However, due to the lower stability of DOTA complexes with metals that have a large ionic radius (e.g.Ac 3+ ), novel chelators for 225 Ac have recently been reported (e.g.macropa, H 2 bispa2, H 4 octapa) (Hu and Wilson 2022;Ramogida et al. 2019).It would be interesting to compare these chelators to DOTA for 225 Ac-labeling of trastuzumab in the future.The protocol described by Maguire et al. was followed for 225 Ac labeling of DOTA-trastuzumab (Maguire et al. 2014).The LE of DOTA-trastuzumab (5 mg/mL) with 225 Ac at a SA of SA = 5.3 kBq/µg after an incubation time of 2 h at 37 °C was 76.4 ± 3.1%.The LE of DOTA-trastuzumab with 111 In at SA = 0.1 MBq/µg (89.8 ± 4.8%; Additional file 1: Fig. S1D) was higher under similar conditions (5 mg/mL, 0.5 h), but purification of [ 225 Ac]Ac-DOTA-trastuzumab by ultrafiltration increased the final RCP of to 95.9 ± 0.9%.
Treatment of SK-BR-3 cells with [ 225 Ac]Ac-DOTA-trastuzumab (20 nmoles/L; 3.7 µg/ mL) potently decreased their SF in vitro dependent on SA (0.148-1.665 kBq/µg; Fig. 3a).At a SA = 0.754 kBq/µg, the SF was < 0.001 and SA = 1.10 or 1.665 kBq/µg, there were no surviving colonies detected.The cytotoxicity of [ 225 Ac]Ac-DOTA-trastuzumab was HER2-specific, as SF was restored to 0.63 ± 0.13 at a SA = 1.665 kBq/µg by co-incubation of SK-BR-3 cells with a 100-fold excess of DOTA-trastuzumab to block HER2.Yoshida et al. reported that the SF of HER2-positive SUM225-Luc + ductal carcinoma in situ of the breast (DCIS) cells was decreased to 0.15 by exposure to 3.7 kBq/mL of [ 225 Ac]Ac-DOTA-trastuzumab (SA = 37 kBq/µg) compared to 0.91 for exposure to 3.7 kBq/mL of irrelevant [ 225 Ac]Ac-DOTA-rituximab, demonstrating HER2 specific cell killing (Yoshida et al. 2016).In our study, co-incubation of [ 225 Ac]Ac-DOTA-trastuzumab with a 100-fold excess of trastuzumab to determine HER2-specific killing of SK-BR-3 cells may further represent treatment of these cells with trastuzumab.This may explain the residual cytotoxicity observed even in the presence of HER2 blocking (ie.SF = 0.63 ± 0.13).Yoshida reported that trastuzumab treatment (1 µg/mL) of SUM225-Luc+ cells decreased the SF to 0.65-0.7,but [ 225 Ac]Ac-DOTA-trastuzumab was much more potent (SF = 0.1) (Yoshida et al. 2016).At a similar concentration as used in our study (4 µg/mL), Rasaneh et al. found using the MTT cell viability assay that treatment of SK-BR-3 cells in vitro with [ 177 Lu]Lu-DOTA-trastuzumab (SA = 0.0925 MBq/ µg) decreased the number of viable cells to 10%, while trastuzumab decreased the cell number to 41% (Rasaneh et al. 2010).Since the SF of SK-BR-3 cells exposed to [ 225  The cytotoxicity of [ 225 Ac]Ac-DOTA-trastuzumab was further assessed by plotting the SF versus dose in SK-BR-3 cells (Fig. 3b).The dose was estimated from the timeintegrated activity in SK-BR-3 cells derived from the cell uptake of [ 111 In]In-DOTAtrastuzumab (Additional file 1: Fig. S4 and Table S1).The dose required to decrease the clonogenic survival of SK-BR-3 cells to 0.10 (D 10 ) was 1.10 Gy for [ 225 Ac]Ac-DOTAtrastuzumab versus D 10 = 4.855 Gy reported for exposure of SK-BR-3 cells to γ-radiation (Cai et al. 2011).Relative biological effectiveness (RBE) is defined as the ratio of the dose of the reference radiation versus the test radiation required to produce an equivalent biologic effect (i.e.SF = 0.10).Based on a comparison with γ-radiation, the calculated RBE of [ 225 Ac]Ac-DOTA-trastuzumab was 4.4 (Chen 2004).No previous studies have reported the RBE for [ 225 Ac]Ac-DOTA-trastuzumab for killing HER2-positive BC cells.However, Nayak et al. estimated a RBE = 3.4 for Capan-2 human pancreatic cancer cells treated in vitro with α-particle emitting, [ 213 Bi]Bi-DOTATOC versus γ-radiation, while the RBE of [ 177 Lu]Lu-DOTATOC was 1.0 (Nayak et al. 2007).
[ 225 Ac]Ac-DOTA-trastuzumab caused multiple DNA DSBs in SK-BR-3 cells assessed by immunofluorescence confocal microscopy probing for γ-H2AX (Cai et al. 2009) (Fig. 4a).The integrated density of γ-H2AX foci per cell nucleus area was directly dependent on the SA (0.148-1.665 kBq/µg) of [ 225 Ac]Ac-DOTA-trastuzumab; Fig. 4b) and the absorbed dose in the cells (0-0.5 Gy, Fig. 4c).DNA DSBs were HER2-specific since coincubation of SK-BR-3 cells with [ 225 Ac]Ac-DOTA-trastuzumab combined with a 100-fold excess of trastuzumab to block HER2 significantly decreased the integrated γ-H2AX foci density per nucleus area by 5.5-fold compared to [ 225   10.6 ± 0.6% ID/g vs. 4.3 ± 0.7% ID/g, Fig. 5a), demonstrating HER2-specific tumour uptake.However, spleen uptake in mice coinjected with [ 111 In]In-DOTA-trastuzumab and [ 225 Ac]Ac-DOTA-trastuzumab was high (28.9 ± 7.4% ID/g) and was even higher in mice coinjected with [ 111 In]In-DOTA-IgG 1 and [ 225 Ac]Ac-DOTA-IgG 1 (100.6 ± 36.2%ID/g).Sharma et al. (2018) similarly reported high spleen uptake of i.v.injected 89 Zr-labeled mAbs in immunocompromised strains of mice that do not have B-cells (e.g.scid, NOD scid and NSG), but not in athymic (nu/nu) mice which have B-cells but not T-cells.Notably, spleen uptake in NSG mice with s.c.SKOV-3 human ovarian cancer xenografts injected i.v. with 0.9-1.1 MBq (4.4-5.2 µg) of 89 Zr-labeled trastuzumab was very high (> 300% ID/g) at 144 h p.i., causing low tumour uptake (~ 3% ID/g).This phenomenon was avoided by coinjecting 89 Zrlabeled trastuzumab with an excess of isotype-matched IgG (220 µg).NRG mice have no B-cells and this may explain the high spleen uptake in mice coinjected with [ 111 In] In-DOTA-trastuzumab and [ 225 Ac]Ac-DOTA-trastuzumab.However, we injected a 8-10-fold higher total mass of RICs than reported by Sharma et al. (2018) (40 µg vs. 4.4-5.2µg) which may have mitigated spleen sequestration resulting in higher tumour uptake in our study.The phenomenon of high spleen uptake of RICs may not occur in humans due to higher normal levels of circulating immunoglobulins.Nonetheless, high spleen sequestration of [ 225 Ac]Ac-DOTA-trastuzumab may have toxicity implications for studying α-particle RIT in tumour-bearing NRG mice, Spleen toxicity may present in the form of reduced spleen size and weight, loss of integrity, orientation, or structure, and evidence of cellular necrosis (Cortez et al. 2020;Davis et al. 1999;Maguire et al. 2014).Increasing the administered mass of [ 225  The ability to image tumours with [ 111 In]In-DOTA-trastuzumab by SPECT/CT and treat tumours by RIT with [ 225 Ac]Ac-DOTA-trastuzumab presents an opportunity for a theranostic strategy for imaging and treatment of HER2-positive BC.Moreover, since 164/8-1B/H2N.luc+ tumours in immunocompromised mice metastasize to the brain (Milsom et al. 2013), these results support our planned future studies to examine the effectiveness of [ 225 Ac]Ac-DOTA-trastuzumab for RIT of HER2-positive BM in this mouse tumour xenograft model.

Fig. 1 a
Fig. 1 a SDS-PAGE of trastuzumab (lane 1) and DOTA-trastuzumab (lane 2) under non-reducing conditions on a 4-20% Mini-Protean Tris/glycine mini-gel.Bands were stained with Coomassie Brilliant Blue G-250.MW: Broad range MW markers.SE-HPLC analysis of b trastuzumab or c DOTA-trastuzumab on a BioSep SEC-S2000 column eluted with 0.1 M NaH 2 PO 4 buffer, pH 7.0 at a flow rate of 0.8 mL/min with UV detection at 280 nm

Fig. 2
Fig. 2 Binding of [ 111 In]In-DOTA-trastuzumab to HER2-positive SK-BR-3 human BC cells in the absence (TB) or presence (NSB) of a 50-fold molar excess of trastuzumab.The SB was calculated by subtracting NSB from TB. Fitting the SB curve to a 1-site receptor binding model estimated the K D = 1.2 ± 0.3 × 10 -8 mol/L and B max = 4.2 ± 0.2 × 10 6 receptors/cell