Design and Evaluation of 223Ra-Labeled and Anti-PSMA Targeted NaA Nanozeolites for Prostate Cancer Therapy—Part II. Toxicity, Pharmacokinetics and Biodistribution

Metastatic castration-resistant prostate cancer (mCRPC) is a progressive and incurable disease with poor prognosis for patients. Despite introduction of novel therapies, the mortality rate remains high. An attractive alternative for extension of the life of mCRPC patients is PSMA-based targeted radioimmunotherapy. In this paper, we extended our in vitro study of 223Ra-labeled and PSMA-targeted NaA nanozeolites [223RaA-silane-PEG-D2B] by undertaking comprehensive preclinical in vitro and in vivo research. The toxicity of the new compound was evaluated in LNCaP C4-2, DU-145, RWPE-1 and HPrEC prostate cells and in BALB/c mice. The tissue distribution of 133Ba- and 223Ra-labeled conjugates was studied at different time points after injection in BALB/c and LNCaP C4-2 tumor-bearing BALB/c Nude mice. No obvious symptoms of antibody-free and antibody-functionalized nanocarriers cytotoxicity and immunotoxicity was found, while exposure to 223Ra-labeled conjugates resulted in bone marrow fibrosis, decreased the number of WBC and platelets and elevated serum concentrations of ALT and AST enzymes. Biodistribution studies revealed high accumulation of 223Ra-labeled conjugates in the liver, lungs, spleen and bone tissue. Nontargeted and PSMA-targeted radioconjugates exhibited a similar, marginal uptake in tumour lesions. In conclusion, despite the fact that NaA nanozeolites are safe carriers, the intravenous administration of NaA nanozeolite-based radioconjugates is dubious due to its high accumulation in the lungs, liver, spleen and bones.


Introduction
Metastatic castration-resistant prostate cancer (mCRPC) is a progressive and incurable disease with poor prognosis for patients [1]. Despite several options for the treatment of mCRPC, such as taxanes-based chemotherapy, second-generation antiandrogens, poly-(ADP-ribose)-polymerase (PARP) inhibitors, immunotherapy with sipuleucel-T and therapy with the bone-seeking radium-223 dichloride (Xofigo ® ), the median survival is 30 months with a 30% 5-year survival rate.

Synthesis and Physicochemical Characterization of 223 RaA-Silane-PEG-D2B Radioconjugate
Synthesis and physicochemical characterization of 223 RaA-silane-PEG-D2B and its derivatives were described in detail by Czerwińska et al. [24]. Briefly, these compounds were synthesized from aluminosilicate gel using the hydrothermal method. In the first step, the NaA zeolite nanocarrier was prepared. Next, the NaA nanocarrier was labeled with 223 Ra 2+ radionuclide by exchanging Na + for 223 Ra 2+ cations, and the surface was modified with silane-PEG groups by siloxane bonds formation [ 223 RaA-silane-PEG and NaA-silane-PEG]. In the last step, an anti-PSMA D2B antibody was conjugated to the nanocarrier via the 2-iminothiolane/m-maleimidobenzoyl-N-hydroxysuccinimide ester coupling method [ 223 RaA-silane-PEG-D2B and NaA-silane-PEG-D2B] (Figure 1). The obtained compound was in the form of nanocrystals with a regular cubic-like shape, an average nominal diameter of ~120 nm, and an average hydrodynamic diameter of ~200 nm. The radiolabeled compounds [ 223 RaA-silane-PEG-D2B and 223 RaA-silane-PEG] had a specific activity of 0.65 MBq mg −1 . It was estimated that ~200 silane-PEG groups and ~50 D2B molecules were coupled with one NaA zeolite molecule.

Analysis of Cell Death by the Annexin/Propidium Iodide Assay
The annexin/propidium iodide assay was used to investigate the effect of 50 mcg mL −1 NaA-silane-PEG and NaA-silane-PEG-D2B on the induction of cell death in LNCaP C4-2, DU-145, RWPE-1 and HPrEC cells following 94 h of incubation. Our results revealed that these nanocarriers induced a statistically significant increase of apoptosis and necrosis in HPrEC cells (Figure 2). NaA-silane-PEG elevated the percentage of early apoptotic and late apoptotic/necrotic HPrEC cells to 1.7-fold (1.9% vs. 1.1%) and to 1.5-fold (12.1% vs. 8.1%) of the control cells, respectively. NaA-silane-PEG-D2B increased the percentage of early apoptotic and late apoptotic/necrotic HPrEC cells to 2.3-fold (2.5% vs. 1.1%) and to 1.7-fold (13.5% vs. 8.1%) of the control cells, respectively. No significant difference in percentage of apoptotic and necrotic cells was observed in LNCaP C4-2, DU-145 and RWPE-1 cells.

Analysis of Cell Death by the Annexin/Propidium Iodide Assay
The annexin/propidium iodide assay was used to investigate the effect of 50 mcg mL −1 NaA-silane-PEG and NaA-silane-PEG-D2B on the induction of cell death in LNCaP C4-2, DU-145, RWPE-1 and HPrEC cells following 94 h of incubation. Our results revealed that these nanocarriers induced a statistically significant increase of apoptosis and necrosis in HPrEC cells ( Figure 2). NaA-silane-PEG elevated the percentage of early apoptotic and late apoptotic/necrotic HPrEC cells to 1.7-fold (1.9% vs. 1.1%) and to 1.5-fold (12.1% vs. 8.1%) of the control cells, respectively. NaA-silane-PEG-D2B increased the percentage of early apoptotic and late apoptotic/necrotic HPrEC cells to 2.3-fold (2.5% vs. 1.1%) and to 1.7-fold (13.5% vs. 8.1%) of the control cells, respectively. No significant difference in percentage of apoptotic and necrotic cells was observed in LNCaP C4-2, DU-145 and RWPE-1 cells. Figure 2. The effect of NaA-silane-PEG and NaA-silane-PEG-D2B on the proportion of viable cells (annexin V-negative and propidium iodide-negative cells), early apoptotic cells (annexin V-positive cells and propidium iodide-negative cells) and late apoptotic/necrotic cells (annexin V-positive and propidium iodide-positive cells). LNCaP C4-2 cells (A), DU-145 cells (B), RWPE-1 cells (C) and HPrEC cells (D) were exposed to 50 mcg mL −1 of NaA-silane-PEG or NaA-silane-PEG-D2B for 96 h. Data are expressed as the mean ± standard deviation (SD) from three independent experiments. * p < 0.05 versus control group.

Analysis of Apoptosis by Caspase 3/7 Assay
To confirm the results obtained with the annexin/propidium iodide assay, we determined the percentage of caspase 3/7 positive cells, representing apoptotic cells. Our results were nearly identical to those observed by the annexin/propidium iodide assay and revealed that NaA-silane-PEG and NaA-silane-PEG-D2B significantly induced apoptosis of HPrEC cells, while no differences were observed in LNCaP C4-2, DU-145 and RWPE-1 cells (Figure 3). NaA-silane-PEG significantly elevated the percentage of apoptotic HPrEC cells to 1.8-fold (2.5% vs. 1.4%) of the control cells, whereas NaA-silane-PEG-D2B significantly increased the percentage of apoptotic cells to 2.6-fold (3.7% vs. 1.4%) of the control cells. HPrEC cells (D) were exposed to 50 mcg mL −1 of NaA-silane-PEG or NaA-silane-PEG-D2B for 96 h. Data are expressed as the mean ± standard deviation (SD) from three independent experiments. * p < 0.05 versus control group.

Analysis of Apoptosis by Caspase 3/7 Assay
To confirm the results obtained with the annexin/propidium iodide assay, we determined the percentage of caspase 3/7 positive cells, representing apoptotic cells. Our results were nearly identical to those observed by the annexin/propidium iodide assay and revealed that NaA-silane-PEG and NaA-silane-PEG-D2B significantly induced apoptosis of HPrEC cells, while no differences were observed in LNCaP C4-2, DU-145 and RWPE-1 cells (Figure 3). NaA-silane-PEG significantly elevated the percentage of apoptotic HPrEC cells to 1.8-fold (2.5% vs. 1.4%) of the control cells, whereas NaA-silane-PEG-D2B significantly increased the percentage of apoptotic cells to 2.6-fold (3.7% vs. 1.4%) of the control cells.

Gene Expression Profiling
NaA-silane-PEG and NaA-silane-PEG-D2B were analyzed by real-time PCR for their effects on the expression of 84 key genes involved in autoimmune and inflammatory responses and 84 key genes involved in NF-κB signaling in LNCaP C4-2, DU-145, RWPE-1 and HPrEC cells. The detailed results are presented in Supplementary Materials (Supplementary Tables S1-S8), while the number of genes significantly affected by these nanocarriers is presented in Venn diagrams ( Figure 4).  . The effect of NaA-silane-PEG and NaA-silane-PEG-D2B on the proportion of caspase 3/7 positive cells. LNCaP C4-2 cells, DU-145 cells, RWPE-1 cells and HPrEC cells were exposed to 50 mcg mL −1 of NaA-silane-PEG or NaA-silane-PEG-D2B for 96 h. Data are expressed as the mean ± standard deviation (SD) from three independent experiments. * p < 0.05 versus control group.

Gene Expression Profiling
NaA-silane-PEG and NaA-silane-PEG-D2B were analyzed by real-time PCR for their effects on the expression of 84 key genes involved in autoimmune and inflammatory responses and 84 key genes involved in NF-κB signaling in LNCaP C4-2, DU-145, RWPE-1 and HPrEC cells. The detailed results are presented in Supplementary Materials (Supplementary Tables S1-S8), while the number of genes significantly affected by these nanocarriers is presented in Venn diagrams ( Figure 4). For NaA-silane-PEG, only 2 out of 84 genes involved in autoimmune and inflammatory responses were deregulated in LNCaP C4-2 cells, 3 genes in DU-145 cells, 5 genes in RWPE-1 cells and 8 genes in HPrEC cells ( Figure 4A). For NaA-silane-PEG, only 1 out of 84 genes involved in NF-κB signaling was affected in LNCaP C4-2 cells, 11 genes in DU-145 cells, 1 gene in RWPE-1 cells and 12 genes in HPrEC cells ( Figure 4B). As shown in For NaA-silane-PEG, only 2 out of 84 genes involved in autoimmune and inflammatory responses were deregulated in LNCaP C4-2 cells, 3 genes in DU-145 cells, 5 genes in RWPE-1 cells and 8 genes in HPrEC cells ( Figure 4A). For NaA-silane-PEG, only 1 out of 84 genes involved in NF-κB signaling was affected in LNCaP C4-2 cells, 11 genes in DU-145 cells, 1 gene in RWPE-1 cells and 12 genes in HPrEC cells ( Figure 4B). As shown in Figure 4C, NaAsilane-PEG-D2B deregulated 5 out of 84 genes involved in autoimmune and inflammatory responses in LNCaP C4-2 cells, 5 genes in DU-145 cells, 15 genes in RWPE-1 cells and 27 genes in HPrEC cells. For NaA-silane-PEG-D2B, 6 out of 84 genes involved in NF-κB signaling was affected in LNCaP C4-2 cells, 10 genes in DU-145 cells, 12 genes in RWPE-1 cells and 38 genes in HPrEC cells ( Figure 4D). The overlap in expression of genes between cell lines was very small. Since the vast majority of genes reported in Figure 4 presented very low fold changes (<1.5) despite statistical significance, we focused on further analysis of the statistically significant results with cut-off value for gene expression fold changes >1.5. The results revealed that NaA-silane-PEG deregulated expression of 4 genes in DU-145 cells, 1 gene in RWPE-1 cells and 3 genes in HPrEC cells ( Figure 5). There were no significant differences in gene expression in LNCaP C4-2 cells. However, NaA-silane-PEG-D2B deregulated expression of 2 genes in LNCaP C4-2 cells, 3 genes in DU-145 cells and 7 genes in RWPE-1 cells ( Figure 6A). The effects of NaA-silane-PEG-D2B on gene expression were much more pronounced in primary HPrEC cells. As shown in Figure 6B Figure 6A). The effects of N PEG-D2B on gene expression were much more pronounced in primary HPrEC shown in Figure 6B, NaA-silane-PEG-D2B upregulated expression of 22 genes a regulated expression of 7 genes.

Toxicity In Vivo
The nanocarrier and radioconjugate in vivo toxicity was tested on a homogeneous group of BALB/c mice, using unlabeled (NaA-silane-PEG and NaA-silane-PEG-D2B) and radiolabeled ( 223 RaA-silane-PEG and 223 RaA-silane-PEG-D2B) conjugates, 1 and 7 days after a single intravenous injection of the dose of c.a 6.4-8 mcg kg −1 b.w. and 49-51 kBq per mouse. The injected doses were well tolerated by all animals without treatment-related lethality. Moreover, no changes were observed in body weight, blood morphology, biochemical parameters and during histological examination of the liver, kidneys and femur 1 day after injection for all compounds, as compared to the control group (PBS). However, 7 days of exposure to 223 RaA-silane-PEG and 223 RaA-silane-PEG-D2B induced toxic effects, manifested by body weight loss ( Figure 7a) and an abnormally low level of platelets, significantly different from the control mice and mice exposed to unlabeled NaA-silane-PEG and NaA-silane-PEG-D2B compounds (Figure 7c). In addition, the increased serum concentrations of ALT and AST enzymes were observed in mice exposed to all compounds, as compared to the control group ( Figure 7b). Furthermore, the results showed that mice exposed to 223 RaA-silane-PEG and 223 RaA-silane-PEG-D2B had significantly lower red blood cell (RBC) and white blood cell (WBC) counts, as well as hematocrit (HCT) and hemoglobin (HGB) levels, at 7 days after injection, compared to the results at 1 day after injection (Figure 7d,e, Table 1 and Appendix A (Tables A3-A7).
centrations of ALT and AST enzymes were observed in mice exposed to all compounds, as compared to the control group ( Figure 7b). Furthermore, the results showed that mice exposed to 223 RaA-silane-PEG and 223 RaA-silane-PEG-D2B had significantly lower red blood cell (RBC) and white blood cell (WBC) counts, as well as hematocrit (HCT) and hemoglobin (HGB) levels, at 7 days after injection, compared to the results at 1 day after injection (Figure 7d,e, Table 1 and Appendix (Tables A3-A7).

Figure 7.
Influence of unlabeled and radiolabeled compounds on hematological and biochemical parameters of blood and body weight 7 days after a single injection of the tested compounds in BALB/c mice. Body weight (a), concentration of ALT and AST enzymes (b), concentration of platelets (c), hematological parameters in blood of mice exposed to NaAsilane-PEG (d) and NaA-silane-PEG-D2B (e). Data are expressed as the mean ± standard deviation (SD) from 10 mice/group. * p < 0.05, ** p < 0.01, *** p < 0.001 versus control group.
The mean value of RBC (red blood cell) counts never decreased below the lower reference limit (6.93 × 10 3 mcL) [23], but 7 days after injection, the concentration of platelets Figure 7. Influence of unlabeled and radiolabeled compounds on hematological and biochemical parameters of blood and body weight 7 days after a single injection of the tested compounds in BALB/c mice. Body weight (a), concentration of ALT and AST enzymes (b), concentration of platelets (c), hematological parameters in blood of mice exposed to NaA-silane-PEG (d) and NaA-silane-PEG-D2B (e). Data are expressed as the mean ± standard deviation (SD) from 10 mice/group. * p < 0.05, ** p < 0.01, *** p < 0.001 versus control group.
The mean value of RBC (red blood cell) counts never decreased below the lower reference limit (6.93 × 10 3 mcL) [23], but 7 days after injection, the concentration of platelets ( Figure 7c) decreased for both radiolabeled compounds, and this value was significantly different with respect to the 1 day after injection value (p = 0.030). At the same time, we observed the significant decrease of HCT (hematocrit), below the lower reference limit (42.1%), and HGB (hemoglobin). Particular attention should be paid to the strong leukopenia in both groups of animals ( 223 RaA-silane-PEG and 223 RaA-silane-PEG-D2B) 7 days after injection. The mean value of WBC significantly decreased below the lower reference limit (3.48 G L −1 ) and was 10-fold lower that the value at 1 day. Detailed analysis of the individual fractions of WBC revealed significant differences: neutrophil (NEUT) counts-over 20-fold decrease, lymphocyte (LYMPH) counts-10-fold decrease and monocyte (MONO) counts-17-fold decrease.
Examination of hematoxylin-eosin-stained mice tissues collected 1 day after injection revealed no changes in the liver, kidneys, and bone marrow for unlabeled (NaA-silane-PEG and NaA-silane-PEG-D2B) and radiolabeled ( 223 RaA-silane-PEG and 223 RaA-silane-PEG-D2B) compounds. Additionally, histopathological examination of these tissues collected 7 days after injection from mice treated with unlabeled NaA-silane-PEG and NaA-silane-PEG-D2B compounds showed no changes, as compared with tissues from the control mice. However, 7 days of exposure to 223 RaA-silane-PEG and 223 RaA-silane-PEG-D2B induced bone marrow fibrosis, without any changes in the liver and kidneys ( Figure 8).
PEG and NaA-silane-PEG-D2B) and radiolabeled ( 223 RaA-silane-PEG and 223 RaA-silane-PEG-D2B) compounds. Additionally, histopathological examination of these tissues collected 7 days after injection from mice treated with unlabeled NaA-silane-PEG and NaAsilane-PEG-D2B compounds showed no changes, as compared with tissues from the control mice. However, 7 days of exposure to 223 RaA-silane-PEG and 223 RaA-silane-PEG-D2B induced bone marrow fibrosis, without any changes in the liver and kidneys ( Figure 8).   Biodistribution studies of 223 RaA-silane-PEG and 223 RaA-silane-PEG-D2B in BALB/c mice were planned on the basis of the results dealing with pharmacokinetics and the biodistribution of 133 BaA-silane-PEG and 133 BaA-silane-PEG-D2B in BALB/c mice (Appendix A Table A1, Figure A1). 133 Ba radionuclide is γ-emitter and is particularly feasible as a diagnostic match to the therapeutic α-emitter 223 Ra radionuclide [21].
In the first set of experiments, the biodistribution of 223 RaA-silane-PEG and 223 RaAsilane-PEG-D2B was evaluated in BALB/c mice 1 and 7 days after injection. Mice were exposed to a dose of c.a 8 mcg kg −1 b.w. and 38 and 48 kBq per mouse, respectively. Side-by-side comparison is shown in Table 2. To investigate the efficacy of 223 RaA-silane-PEG-D2B penetration into the prostate tumor, we used BALB/c Nude mice with a subcutaneous prostate tumor xenograft derived from the PSMA-positive LNCaP C4-2 prostate cancer cell line. Biodistribution studies were conducted 4 h, 24 h, 72 h and 7 days after injection. 223 RaA-silane-PEG was used as a control. Both radioconjugates were applied at in average dose of c.a 12 mcg kg −1 b.w. and 50 kBq per mouse. Side-by-side comparison is shown in Table 3.
Statistically significant differences were observed between radioactivity (%ID g −1 ) in the lungs, spleen, bones and blood of healthy BALB/c mice and tumour bearing BALB/c Nude mice. Radioactivity of 223 RaA-silane-PEG and 223 RaA-silane-PEG-D2B in the lungs and bones of BALB/c mice was 2-fold lower than in tumour-bearing mice. Both radioconjugates were excreted more slowly by tumour-bearing mice than healthy mice. However, 223 RaA-silane-PEG-D2B accumulated at significantly higher levels in the spleen, liver and lungs of tumour-bearing mice at 7 day, as compared to the 223 RaA-silane-PEG-injected animals. The bones exhibited similar radioactivity in both 223 RaA-silane-PEG-D2B-and 223 RaA-silane-PEG-exposed animals. Surprisingly, we observed very low uptake of 223 RaAsilane-PEG-D2B in tumour lesions (below 1 %ID g −1 ), despite the presence of a targeting antibody. The value of accumulation was higher for 223 RaA-silane-PEG than for 223 RaAsilane-PEG-D2B, and this was related to the blood flow through the tumour blood vessels but not to specific localization. The lower tumour accumulation value for 223 RaA-silane-PEG-D2B corresponded to lower blood concentration of the radioactivity. After 24 h, both 223 RaA-silane-PEG and 223 RaA-silane-PEG-D2B were still present in the liver and spleen with high %ID g −1 .

Discussion
In the present study, we aimed to extent our previous findings, which revealed that the radiobioconjugate 223 RaA-silane-PEG-D2B was very stable in human blood serum in vitro, bound specifically and internalized into PSMA-expressing LNCaP C4-2 cells, as well as demonstrating potent radiotoxicity in LNCaP C4-2 cells [24].
Since one of the major challenges in the application of nanomaterials as carriers to deliver therapeutics for targeted therapy is their safety, the first goal of this study was to evaluate the in vitro and in vivo toxicity of an antibody-free nanocarrier (NaA-silane-PEG) and antibody-functionalized nanocarrier (NaA-silane-PEG-D2B). Our results concluded that both nanocarriers express no cytotoxic activity against established normal and cancer cell lines and very low cytotoxic activity in extremely sensitive primary cells, which are ill adapted to the two-dimensional culture and changes in their environment [26]. These results are concordant with our previous findings, which showed that NaA-silane-PEG-D2B had no effect on the metabolic activity of LNCaP C4-2 and DU145 cells at a dose of 100 mcg mL −1 following 96 h of treatment time [24]. We also previously reported that NaA nanozeolites coated with long PEG molecules (MW1000 and MW2000), as well as unmodified NaA nanozeolite, were nontoxic in HeLa and HEK239 cells [27]. Moreover, no significant cytotoxic activity was also reported for nanozeolites A and Y in macrophages, epithelial and endothelial cells [28], and for pure silica nanozeolites in HeLa cells [29].
Apart from the determination of cytotoxicity of nanocarriers, the assessment of their immunotoxicity is an important component of safety evaluation [30]. It is well known that nanoparticles can interact with various components of the immune system, potentially leading to undesirable immunosuppression, hypersensitivity, immunogenicity, and autoimmunity, involving both innate and adaptive immune responses [31]. Moreover, previous studies demonstrated that accumulation of nanoparticles in the body can modulate pro-inflammatory cytokine production, which in turn is involved in the regulation of cellular events in prostate carcinogenesis and metastasis [32]. One of the valuable tools in screening nanoparticle immunotoxicity is the measurement of expression of immunerelated genes [33]. Therefore, we conducted a comprehensive analysis of immune gene profiling, which involved an analysis of the expression of 84 genes involved in the NF-κB signaling, which plays a causative role in inflammatory processes, controls the transcription of cytokines and genes that regulate various aspects of innate and adaptive immune responses [34] and the expression of 84 genes involved in the inflammatory response and autoimmunity. We found that antibody-free nanocarrier and antibody-functionalized nanocarrier triggered very week response of a small number of genes in vitro. Most of the genes deregulated by NaA-silane-PEG and NaA-silane-PEG-D2B encode proteins involved in both stimulation and suppression of pro-inflammatory responses [35][36][37][38][39]. Simultaneous deregulation of genes encoding proteins that are key to both pro-inflammatory and antiinflammatory responses in the same cell line does not allow a definitive statement about the immunotoxicity of these carriers. Nevertheless, we cannot exclude the possibility that prolonged exposure to these carriers due to their accumulation in the body may contribute to the development of inflammation, responsible for spread of prostate cancer [32,38].
Since the in vitro behavior of nanocarriers frequently does not correlate with their in vivo responses [40], we performed in vivo toxicity studies in a homogeneous group of BALB/c mice. Our results revealed no changes in body weight or the level of platelet and blood morphology 7 days after a single administration of NaA-silane-PEG and NaA-silane-PEG-D2B. However, we observed increased serum levels of aspartate aminotransferase (AST) and alanine aminotransferase (ALT), which are the most sensitive liver enzymes (Figure 7) [41]. We suppose that in our study the elevated concentration of these enzymes was a consequence of nanocarrier accumulation in the liver. These results are consistent with other studies that demonstrated elevated ALT and AST levels due to a high accumulation of different types of nanocarriers in the liver [42].
Taking into account no obvious symptoms of toxicity of both carriers revealed by in vitro and in vivo studies, we investigated biodistribution and radiotoxicity of 223 Ralabeled conjugates at 24 h and 7 days post injection in healthy BALB/c mice. Critical organs for 223 RaA-silane-PEG and 223 RaA-silane-PEG-D2B toxicity were the liver, spleen and the lungs. We assume that the observed accumulation of radioactivity in these organs was related to the phagocytosis of 223 RaA-silane-PEG and 223 RaA-silane-PEG-D2B by Kupffer cells in the liver, pulpa macrophages in the spleen and alveolar macrophages in the lungs. The highest accumulation of 223 RaA-silane-PEG-D2B was observed in the spleen, which was 2-fold higher compared with 223 RaA-silane-PEG. This might be due to the reduction of spleen mass, observed in our study. The result is in line with findings, which showed a radiation-induced progressive spleen volume decrease [43]. During the next six days, the accumulation in the spleen decreased to reach a similar value for both radiolabelled compounds. The increased accumulation of nanoparticle-based conjugates in organs of the reticuloendothelial system was reported by others in healthy mice [44,45]. In our opinion, the biodistribution pattern of 223 RaA-silane-PEG and 223 RaAsilane-PEG-D2B is regulated mainly by their hydrodynamic size, the presence of antibodies, and probably on the hydrophilicity of the particle surface. The 223 RaA-silane-PEG-D2B molecules presented~50 D2B antibodies on the coat, with a final size of nanoparticles~250 nm, whereas the antibody-free 223 RaA-silane-PEG molecules had a diameter~196 nm [24].
It is well known that once nanomaterials are introduced into biological environments, such as blood, serum, or intracellular fluid, they adsorb on their surfaces the so-called "biocorona", containing many proteins and lipids [46]. Bio-corona formation depends on physicochemical and biological properties of nanoparticles, which in turn influences their uptake by cells, biodistribution and circulation half-time [46,47].
In addition, we observed the high radioactivity level in bone tissue for both radioconjugates at 72 h post injection. We suppose that this effect was related to the leakage of 223 Ra radionuclide and its decay products from nanozeolites, and to the natural affinity of 223 Ra radionuclide with bones [48]. Our in vivo results are in contrast to the in vitro findings, which revealed only~5% leakage of 223 Ra and~6% leakage of its decay products from bioconjugate in human blood serum after 12 days [24]. However, it should be noted that the in vitro stability results could not be transferred to in vivo models, where blood flow may easily dislocate the decay products from the surface of the nanozeolite particles and reduce the resorption probability [45,49]. Referring to the limited stability of radiolabeled nanozeolites, in our opinion, the suitability of NaA nanozeolite as the 223 Ra delivery system is dubious from a clinical point of view. A comparison of the biodistribution of 223 Ra-silane-PEG-D2B and 223 Ra-silane-PEG with the biodistribution of 223 RaCl 2 , based on data from the Alpharadin (Xofigo) Assessment Report [50], showed that accumulation of 223 RaCl 2 in the bones, liver and spleen was significantly lower than 223 Ra-silane-PEG-D2B and 223 Ra-silane-PEG (Appendix A Figure A2, Table A2). The difference in the accumulation of the tested compounds in the small intestine, large intestine and kidney were not statistically significant. This is most likely related to the size of the tested molecules compared with 223 RaCl 2 . The much larger diameter of 223 Ra-silane-PEG-D2B and 223 Ra-silane-PEG suggest that retention of these molecules in the liver or spleen may be higher than 223 Ra ions alone. On the other hand, we cannot rule out that the tested compounds are not stable in the living organism, and thus dissociation of 223 Ra from the nanoconjugates and redistribution may occur. This is evidenced by the gradually increasing accumulation of radioactivity in the bone.
The increased accumulation of 223 Ra-labeled conjugates in bone tissue led to bone marrow fibrosis, observed 7 days post injection, which was accompanied by an abnormally low level of platelets and white blood cells. Similar findings with other radio-conjugates were also observed in preclinical in vivo studies and in patients receiving targeted radioligand therapy [51,52]. In addition, we observed the increased serum concentrations of ALT and AST enzymes. This observation is probably a reflection of the increased accumulation of 223 RaA-silane-PEG-D2B and 223 RaA-silane-PEG in the liver. However, we did not detect any sign of histological damage in this tissue. These observations were accompanied by a significant decrease in the body weight of the animals in both groups of mice ( 223 Rasilane-PEG and 223 Ra-silane-PEG-D2B). Therefore, due to animal welfare considerations, no further follow-up beyond 7 days was conducted. The significant toxicity observed after administration of radioactivity in the formulations of approximately 2 MBq/kg, was similar to the acute toxicity results reported in the literature in preclinical studies of 223 RaCl 2 in mice, where the NOAEL for 223 RaCl 2 was found to be <1250 kBq/kg bw [50,53].
To investigate the efficacy of 223 RaA-silane-PEG-D2B in its effective delivery to the prostate tumor, we used BALB/c Nude mice with a subcutaneous prostate tumor xenograft derived from the PSMA-positive LNCaP C4-2 prostate cancer cell line. Surprisingly, we observed very low uptake of 223 RaA-silane-PEG-D2B to tumour mass (below 1 %ID g −1 ), despite the presence of a targeting antibody. This observation clearly showed that the injected 223 RaA-silane-PEG-D2B was not preferentially transported to the prostate tumor xenograft. The reasons for the lack of effective delivery to the prostate tumor are not fully understood, especially since our in vitro studies demonstrated the highly specific binding of 223 RaA-silane-PEG-D2B at the PSMA-expressing LNCaP C4-2 cell surface and its fast cellular internalization. Moreover, the D2B antibody alone has excellent in vivo tumor targeting characteristics, showing the high LNCaP xenograft uptake at doses from 0.1 to 3 mcg/mouse (~50 %ID g −1 ) [54]. The dose of D2B antibody used in our study was within this range (~1 mcg/mouse), indicating that this factor was probably not the reason for low tumor uptake. The most likely explanation for the observed low tumor deposition of 223 RaA-silane-PEG-D2B is the relatively high hydrodynamic diameter of 223 RaA-silane-PEG-D2B (~250 nm) and low Enhanced Permeability and Retention effect (EPR), which is responsible for extravasation and retention of nanocarriers in tumors [55]. Though tumors growing in the subcutaneous microenvironment have a functional pore cut-off size ranging from approximately 10 nm to 1000 µm [55][56][57], the size of endothelial fenestrae and vascular permeability in tumors is highly heterogeneous and depends on tumor type, development and growth [58]. It was reported that slow-growing tumors, such as the prostate tumor, have a decreased EPR effect and are usually difficult to treat with nanomedicine [59]. Similar findings were reported for different tumor xenografts [45,60,61].

Synthesis and Physicochemical Properties of 223 RaA-Silane-PEG-D2B Radioimmunoconjugates
The synthesis and physicochemical properties of 223 RaA-silane-PEG-D2B radioconjugate were described in detail by Czerwińska et al. [24]. Briefly, the NaA nanozeolites were synthesized using the hydrothermal method. Sodium hydroxide was dissolved in distilled water, and afterwards, TMAOH was added. This sodium hydroxide solution was divided into two equal volumes, and aluminum isopropoxide was added to one half of the solution. Silicate solution was prepared by dissolving colloidal silica in the other half of the solution and was combined with the aluminate solution to obtain a gel. Stirring was performed in an ice bath for 96 h (aging time). After this time, the mixture was placed in a high-pressure autoclave and heated at 100 • C for 24 h. Finally, the obtained material was washed and dried. In order to remove the template, the resulting product was calcined for 72 h at 600 • C. The 223 Ra-labeled NaA nanozeolite was prepared by exchanging Na + for 223 Ra 2+ cations in RaCl 2 solution (activity~0.6 MBq). The 133 Ba-labeled NaA nanozeolite was prepared by exchanging Na + for 133 Ba 2+ cations. The functionalization of the surface was carried out by using silane coupling agent with three ethoxy groups and PEG molecules. The anti-PSMA monoclonal antibody D2B (IgG1) was prepared as previously described by Frigeiro et al. [62]. Conjugation of the D2B antibody with the obtained NaA-silane-PEG nanozeolite consisted of three steps. In the first step the antibody D2B-SH derivative was formed, using 2-iminothiolane (2-IT). In the next step, m-maleimidobenzoyl-N-hydroxysuccinimide ester (MBS) was added to the silane-PEG-NH 2 functionalized nanozeolite. Finally, the obtained NaA nanozeolite-silane-PEG-MBS was conjugated with Abs-D2B-2-IT. In addition to the final compound 223 RaA-silane-PEG-D2B, its three derivatives, NaA nanozeolite modified with silane-PEG groups [NaA-silane-PEG], NaA nanozeolite modified with silane-PEG groups and functionalized with anti-PSMA D2B antibodies [NaA-silane-PEG-D2B] and NaA nanozeolite labeled with 223 Ra radionuclide and modified with silane-PEG groups [ 223 RaA-silane-PEG], were synthesized and characterized.

Analysis of Cell Death by the Annexin/IP Assay
Annexin/IP assay was performed according to the manufacturer's instructions (BD Biosciences). Briefly, cells were plated at 80,000 cells per well in a 12-well plate, incubated overnight, and then treated with 50 mcg mL −1 of NaA-silane-PEG or NaA-silane-PEG-D2B for 96 h. After that time, cells were gently harvested by tripsin-EDTA, washed twice with cold PBS, and resuspended in 1× binding buffer at 1 × 10 6 cells mL −1 . Then, FITCconjugated annexin V and PI were added to each sample. Samples were incubated for 15 min in the dark, and 400 mcL of 1× binding buffer was added to each tube. Samples were run on a LSR II flow cytometer (BD Biosciences, San Diego, CA, USA) using FACSDiva software.

Detection of Caspase 3/7 Activities
CellEvent™ Caspase-3/7 Green Flow Cytometry assay was performed according to the manufacturer's instructions (Thermo Fisher, Vienna, Austria). Briefly, LNCaP C4-2, DU-145, RWPE-1 and HPrEC cells were plated at 50,000 cells per well in a 12-well plate, incubated overnight, and then treated with 50 mcg/mL of NaA-silane-PEG or NaA-silane-PEG-D2B for 96 h. After that time, cells were gently harvested by tripsin-EDTA and washed twice with cold PBS. Then, cells were incubated with CellEvent Caspase-3/7 Green Detection Reagent to a final concentration of 2 µM for 30 min. During the last 5 min of incubation, 1 µL of SYTOX AADvanced dead cell stain solution was added to each sample. Samples were run on a LSR II flow cytometer (BD Biosciences, San Diego, CA, USA) using FACSDiva software.

RNA Isolation, Reverse Transcription, and Real-Time PCR
LNCaP C4-2, DU-145, RWPE-1 and HPrEC cells were plated at 70,000 cells per well in a 12-well plate and allowed to adhere. After 24 h, cell culture media were removed and new culture media with 50 mcg mL −1 of NaA-silane-PEG or NaA-silane-PEG-D2B were added. After a 24 h treatment period, cells were harvested and immediately frozen in liquid nitrogen until RNA isolation. Total RNA was extracted from cell pellets using the ReliaPrep RNA Cell Miniprep System (Promega, Madison, WI, USA) according to manufacturer's protocol. RNA concentration was measured using Quantus Fluorometer (Promega, Madison, WI, USA) and the QuantiFluor RNA System (Promega, Madison, WI, USA). RNA integrity was tested by agarose gel electrophoresis. For PCR array analysis, 1 mcg of total RNA was converted to complementary DNA (cDNA) in a 20-mcL reaction volume using RT 2 First Strand Kit (Qiagen, Hilden, Germany). The cDNA was diluted with 91 mcL distilled water and used for expression profiling using the human NF-κB Signaling Targets PCR Array (Qiagen, Hilden, Germany, cat. no. PAHS-225Z) and the human Inflammatory Response and Autoimmunity PCR Array (Qiagen, Hilden, Germany, cat. no. PAHS-077Z) according to manufacturer's instructions. Briefly, a total volume of 25 mcL of PCR reaction mixture, which included 12.5 mcL of RT 2 SYBR Green/ROX qPCR Master Mix (containing HotStart DNA Taq polymerase, SYBR Green dye and the ROX reference dye), 11.5 mcL of double-distilled H 2 O, and 1 mcL of diluted template cDNA, was used for each primer set in each well of the PCR array. One technical replicate was performed for each sample. PCR amplification was carried out using a 7500 Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA, USA) with an initial 10-min step at 95 • C followed by 40 cycles of 95 • C for 15s and 60 • C for 1 min. Relative gene expression was calculated using the ∆∆Ct method with ACTB, B2M, GAPDH, HPRT1, and RPLP0 as reference controls. Calculations were carried out using the Relative Quantification Software version 2019.2.7-Q2-19-build3 (Thermo Fisher Cloud, Thermo Fisher Scientific, Waltham, MA, USA). Statistical differences were examined by Student's t-test with p < 0.05 considered to be statistically significant.

Experimental Animals
BALB/c male mice (5-6-week-old, mean body mass of 20 g) were purchased from the M. Mossakowski Institute of Experimental and Clinical Medicine at the Polish Academy of Sciences in Warsaw (Poland). BALB/c Nude male mice (7-week-old, mean body mass of 20 g) were purchased from the Charles River Laboratories (Sulzfeld, Germany). On arrival, animals were housed for five days in groups of five in standard cages (BALB/c) and IVS cages (BALB/c Nude) in the animal facility of the Radioisotope Centre POLATOM (Otwock, Poland). They were housed in a quiet room under constant conditions (22 • C, 50% relative humidity, 12-h light/dark cycles with dark period from 7 p.m. to 7 a.m.) with free access to standard food and water. Veterinarian staff and investigators observed the mice daily to ensure animal welfare and determine if humane endpoints were reached (e.g., hunched and ruffled appearance, apathy, ulceration, severe weight loss, tumor burden). Experimental procedures were carried out in conformity with the National Legislation and the Council Directive of the European Communities on the Protection of Animals Used for Experimental and Other Scientific Purposes (2010/63/UE) and the "ARRIVE guidelines for reporting animal research" [63]. The POLATOM protocol was approved by the 1st Local Animal Ethics Committee in Warsaw (authorization 429/2017, approval date 22 November 2017).

Experimental Groups and Treatment
Mice were divided into nine groups (G1-G9). G1 group: n = 35 BALB/c mice, treated with a single intravenous injection of 133 BaA-silane-PEG (0.1 mL, 280-320 mcg, 185-209 kBq) into the lateral tail vein. Average dose was 12.3 ± 2.0 mcg kg −1 body weight. Mice were randomized into five groups, from G3 to G7 (10 mice per group). Before injection, 223 RaA-silane-PEG and 223 RaA-silane-PEG-D2B were diluted in PBS and then sonicated for 5 min to break up aggregates of micron-sized colloidal particles. At 24 h and 7 days post injection, venous blood was drawn from the submandibular vein while the mouse was restrained but not anesthetized. Mice were put to death by cervical dislocation at 24 h and seven days post injection.

Measurements of Biochemical Parameters
The vein blood (ca 0.3 mL) was collected in blood collection serum tubes coated with a clot activator on the inner wall of the tubes (Microvette ® 200 Sarstedt, Marfour, Polska, cat. no. 20.1292). The blood test was carried out up to three hours after blood taking. For the biochemical profile, we used the Olympus AU680 Clinical Chemistry Analyzer and calorimetric method at 37 • C. We checked the following parameters and methods: enzymes (AST (aspartate aminotransferase), ALT (alanine aminotransferase)-IFCC Reference method with Pyrioxal Phosphate; Ap-IFCC Reference Method 2011), albumin-Bromocresol Green, urea-Urease/GLDH, creatinine-Jaffe IDMS and c-reactive protein.

Histopathology
The bone, liver and kidneys were fixed in 4% formalin in PBS. Organ samples were trimmed and embedded in paraffin. Then, histological sections were prepared, stained with hematoxylin-eosin and assessed under a light microscope. 4.14. Biodistribution of 223 RaA-Silane-PEG and 223 RaA-Silane-PEG-D2B in LNCaP C4-2 Tumor-Bearing BALB/c Nude Mice LNCaP C4-2 cells were grown to 80-90% confluence before trypsinization and formulation in Matrigel™ Basement Membrane Matrix (Corning, Bedford, MA, US). BALB/c Nude mice were subcutaneously injected in the shoulder with 200 mcL of bolus containing a suspension of 5 × 10 6 freshly harvested LNCaP C4-2 cells (100 mcL) in Matrigel™ (100 mcL) under anesthesia with 2% isoflurane. Mice were kept under pathogen-free conditions. Experiments were performed~4-5 weeks later, when the tumor reached a volume of approximately 100 ± 40 mm 3 . Then, mice were randomized into two groups (10 mice per group): G8-treated with a single intravenous injection of 223 RaA-silane-PEG and G9-treated with a single intravenous injection of 223 RaA-silane-PEG-D2B. Mice were euthanized 4 h, 24 h, 72 h and 7 days after injection by an appropriate method and dissected. Selected organs and tissues were weighed and assayed for their radioactivity. Ex-vivo biodistribution was then calculated and expressed as the percentage of administrated radioactivity found in each organ or tissue per gram (%IDg −1 ).

Statistical Analysis
Statistical analysis of the in vitro data (with the exception of the PCR Array data) was performed using the Statistica 7.1 software (Stat Soft. Inc., Tulsa, OK, USA). The data were expressed as mean ± standard deviation (SD) of at least three independent experiments. Data were evaluated by the Kruskal-Wallis one-way analysis of variance on ranks (ANOVA) followed by the post-hoc Fisher's test. Differences were considered statistically significant when the p value was less than < 0.05. Venn diagrams were drawn using the online web tool [64].
The results of in vivo biochemical parameters, concentration of platelets, and physiological distribution were expressed as a percentage of the dose administered per gram of tissue (%ID g −1 ) and presented in the form of an average with standard deviation (mean ± SD), with n representing the number of samples or animals per group. Data were statistically analyzed using GraphPad Prism version 8.0.0 for Windows and tested for normal distribution with the Kolmogorov-Smirnov test. In case of normal distribution, results were assessed by two-tailed, unpaired Student's t-tests. Otherwise, results were assessed by two-way ANOVA. A p value of < 0.05 with two-tailed testing was considered statistically significant. For blood activity data, a one-phase exponential decay model was used to model the percentage of remaining activity (%ID/g) as a function of time post-injection (t): where: %ID g −1 (0) is the %ID g −1 value when T (time) is zero plateau is the %ID g −1 value at infinite times K is the rate constants Nonlinear least-squares regression was used to estimate the half-live of the exponential functions:   week) were exposed to a single intravenous dose of 223 RaCl2 of ~25 µL (20 g body weight with solution activity at 500 kBq/mL on day 0).    week) were exposed to a single intravenous dose of 223 RaCl2 of ~25 µL (20 g body weight with solution activity at 500 kBq/mL on day 0). Figure A2. Comparative analysis of biodistribution of 223 Ra-silane-PEG-D2B and 223 Ra-silane-PEG with biodistribution of 223 RaCl 2 (Xofigo) on 7th day. BALB/c mice (weight: 19.1-23.6 g; age 6 week) were exposed to a single intravenous dose of 223 RaCl2 of~25 µL (20 g body weight with solution activity at 500 kBq/mL on day 0).  Table A4. Hematological parameters of blood of BALB/c mice at 1 and 7 days post injection with NaA-silane-PEG.  Table A6. Hematological parameters of blood of BALB/c mice at 1 and 7 days after injection with NaA-silane-PEG-D2B.