Pd2Spermine Complex Shows Cancer Selectivity and Efficacy to Inhibit Growth of Triple-Negative Breast Tumors in Mice

Pd2Spm is a dinuclear palladium(II)-spermine chelate with promising anticancer properties against triple-negative breast cancer (TNBC), a breast carcinoma subset with poor prognosis and limited treatment options. The present study evaluated the in vitro and in vivo anticancer effects of Pd2Spm compared to the reference metal-based drug cisplatin. Triple-negative breast cancer MDA-MB-231 cells, non-cancerous MCF-12A breast cells and chorioallantoic membrane (CAM) assay were used for antiproliferative, antimigratory and antiangiogenic studies. For an in vivo efficacy study, female CBA nude mice with subcutaneously implanted MDA-MB-231 breast tumors were treated with Pd2Spm (5 mg/kg/day) or cisplatin (2 mg/kg/day) administered intraperitoneally during 5 consecutive days. Promising selective antiproliferative activity of Pd2Spm was observed in MDA-MB-231 cells (IC50 values of 7.3–8.3 µM), with at least 10-fold lower activity in MCF-12A cells (IC50 values of 89.5–228.9 µM). Pd2Spm inhibited the migration of MDA-MB-231 cells, suppressed angiogenesis in CAM and decreased VEGF secretion from MDA-MB-231 cells with similar potency as cisplatin. Pd2Spm-treated mice showed a significant reduction in tumor growth progression, and tumors evidenced a reduction in the Ki-67 proliferation index and number of mitotic figures, as well as increased DNA damage, similar to cisplatin-treated animals. Encouragingly, systemic toxicity (hematotoxicity and weight loss) observed in cisplatin-treated animals was not observed in Pd2Spm-treated mice. The present study reports, for the first time, promising cancer selectivity, in vivo antitumor activity towards TNBC and a low systemic toxicity of Pd2Spm. Thus, this agent may be viewed as a promising Pd(II) drug candidate for the treatment of this type of low-prognosis neoplasia.


Introduction
Female breast cancer is currently the most commonly diagnosed cancer and the leading type of cancer death in women worldwide, with 2.3 million new cases having been diagnosed in 2020 [1]. Breast cancer is a highly heterogeneous disease that presents several subtypes with different treatment success and survival prognosis. Triple-negative

Pd 2 Spm Synthesis and Formulation
The Pd 2 Spm complex was synthesized according to published procedures [20] optimized by the authors [21]: 2 mmol of K 2 PdCl 4 was dissolved in a minimal amount of water and 1 mmol of spermine aqueous solution was added dropwise under continuous stirring. After 24 h, the resulting orange powder was filtered and washed with acetone, and the yellow-orange needle-shaped crystals were recrystallized from water. The composition and purity of the synthesized compound were fully obtained by elemental analysis and vibrational spectroscopy (FTIR, Raman and inelastic neutron scattering), which were compared with the previously calculated vibrational profiles (by ab initio methods) [21,22]. Yield: 68%. Elemental analysis (Pd 2 (C 10 N 4 H 26 )Cl 4 ): Found-C: 21.2%; H: 4.7%; N: 9.6%, Cl: 25.9%; Calculated-C: 21.5%; H: 4.6%; N: 9.9%, Cl: 25.6%. A Pd 2 Spm 0.28 mg/mL solution for in vivo administration was freshly prepared by dissolving an appropriate quantity of drug in phosphate-buffered saline (PBS) (H 2 PO 4 1.5 mM, Na 2 HPO 4 4.3 mM, KCl 2.7 mM, NaCl 150 mM, pH 7.4) containing 0.5% of DMSO and sterile-filtered. A solution of cisplatin 0.35 mg/mL was prepared in PBS and sterile-filtered.

Cell Cultures
The human triple-negative breast cancer cell line MDA-MB-231 (ATCC HTB-26) (absence of estrogen and progesterone receptors, HER2 overexpression) and the non-cancerous breast cell line MCF-12A (ATCC CRL-10782) were purchased from ATCC (Manassas, VA, USA). MDA-MB-231 cells were cultured in DMEM-HG cell growth medium supplemented with 10% (v/v) FBS. MCF-12A cells were cultured in DMEM/F12 medium supplemented with 20 ng/mL hEGF, 100 ng/mL cholera toxin, 0.01 mg/mL bovine insulin, 500 ng/mL hydrocortisone and 5% (v/v) horse serum. Cells were cultured in monolayers in a sterile environment at 37 • C with 5% CO 2 humidified atmosphere. Under these conditions, the population doubling time was 25.5 ± 0.9 h and 20.6 ± 3.1 h for MDA-MB-231 and MCF-12A cells, respectively. The cell cultures were routinely screened for mycoplasma contamination, yielding negative results.

Cell Proliferation Assay
Cells were seeded in 96-well microplates at the cell density 1.5 × 10 4 cells/cm 2 (final volume 200 µL/well) and left 24 h to attach. Afterwards, the growth medium was replaced with the growth medium containing Pd 2 Spm (1-100 µM) or cisplatin (0.1-100 µM). Label-free kinetic live monitoring of cell proliferation was performed using LionheartFX automated microscope (BioTek, Winooski, VT, USA) with direct image acquisition of cells in microplates at 0, 24, 48 and 72 h post-addition of the tested compounds. Acquired 4X images were processed using Gen 5 Image Analysis software (BioTek, Winooski, VT, USA) that allows for identification and counting of individual cells per image. The normalized cell growth (%) was calculated using the following formula:

Cell Migration-Wound Healing Assay
The effect of compounds on cell migration was studied using wound healing assay (or scratch assay). Triple-negative breast cancer MDA-MB-231 cells were seeded in 24-well microplate at the cell density 12.5 × 10 4 cells/cm 2 (1 mL final volume) and left 24 h to attach and create a confluent monolayer. The cell monolayer was scratched with 200 µL pipette tip, producing a vertical straight wound in the middle of the well; growth medium was removed and wells were washed twice with growth medium to remove all detached cells. Growth medium containing Pd 2 Spm (12.5-200 µM) or cisplatin (6.25-200 µM) was added to the wells, and microplates were imaged using LionheartFX automated microscope at 0, 3, 18, 20, 24 and 48 h. Wound closure was analyzed as a confluence of the initial cell-free area covered by migrating cells at the indicated time points measured with Gen 5 Image Analysis software. Confluence was calculated using the following formula: where C(t) is the percent wound confluence over time, Object Sum Area(t) is the area covered by cells at each time point, Object Sum Area(0) is the area covered by cells at time 0 h and I(A) is the total area of the 4X image. The area under the curve (AUC) of kinetic time-confluence profiles obtained within 0-18 h for each concentration was calculated (AUC 0-18h ), plotted against the drug concentration and analyzed with nonlinear regression to obtain the half maximal effective concentration (EC50) for Pd 2 Spm and cisplatin.

Quantitative Analysis of Vascular Endothelial Growth Factor (VEGF) Production
MDA-MB-231 cells were seeded in 96-well microplates at the cell density 1.5 × 10 4 cells/cm 2 (final volume 200 µL/well) and left 24 h to attach. The growth medium was replaced with growth medium containing Pd 2 Spm (5-20 µM) or cisplatin (5-20 µM), and cells were incubated for 12 h. VEGF concentration in cell culture medium was determined in undiluted samples in duplicate using the Human VEGF Quantikine ELISA Kit (R&D Systems Inc., Minneapolis, MN, USA) according to the manufacturer's protocol. In brief, 200 µL of conditioned media was added to a 96-well microplate pre-coated with monoclonal antibody for human VEGF and the plate was incubated for 2 h. After washing with wash buffer, enzyme-linked polyclonal antibody specific for human VEGF was added, followed by incubation for 2 h. The plate was washed to remove unbound antibody-enzyme reagent, substrate solution was added and incubated for 30 min and the reaction was terminated by the addition of the stop solution, producing a yellow solution. The optical density at 450 nm with wavelength correction at 540 nm was measured using a Synergy HT microplate reader (BioTek, Winooski, VT, USA). The concentration of VEGF was interpolated from the standard curve (15.6-1000 pg/mL) prepared in parallel.

Mice
Female CBA nude (N:NIH(S)II-nu/nu) mice (6-7 weeks old, 21 animals in total) with combined immunodeficiency [23] were purchased from i3S Animal Facility (Porto, Portugal). Animals were acclimatized at ICBAS-UP Rodent Animal House Facility (Porto, Portugal) at least one week prior to the experiments, were randomly distributed into groups of three per individually ventilated cage and were housed under controlled specific-pathogen-free (SPF) environmental conditions (temperature 22.5 ± 1.5 • C; relative humidity 50 ± 10%; 12 h light/dark cycle) with ad libitum access to water and standard pellet food (4RF21, Mucedola, Italy). Environmental enrichment included corncob bedding, paper roll tube and one large sheet of tissue paper for nesting. Animals were monitored daily for health status and welfare.

Subcutaneous In Vivo Breast Cancer Xenograft Study
Mice were subcutaneously implanted in the left flank with breast cancer MDA-MB-231 cells (25G needle, 5 × 10 6 cells in 150 µL of PBS). At day 25 post-implantation, when tumors reached a mean volume of~250 mm 3 , mice were randomly allocated into three groups (7 animals per group) using computer-generated randomization sequence followed by random group allocation to the treatment with either (A) Pd 2 Spm, (B) cisplatin or (C) vehicle. The Pd 2 Spm (5 mg/kg/day), cisplatin (2 mg/kg/day) or vehicle (PBS + 0.5% DMSO) were administered via intraperitoneal injection (500 µL injection volume) during five consecutive days in the A) Pd 2 Spm, B) cisplatin and C) vehicle group, respectively. The animals were monitored daily for physical activity, wellbeing and measurements of body weight. Tumor measurements were performed by two independent researchers using a digital caliper in two perpendicular diameters of the implant. Tumor volumes were calculated using the Carlsson formula [24]: tumor volume (mm 3 ) = 0.5 × largest diameter (mm) × smallest diameter 2 (mm). Researchers were blinded to treatment allocation while performing outcome measurements. At day 39 post-implantation (end of the study), animals were euthanized with isoflurane followed by cardiac puncture for blood collection. Blood was collected into K3EDTA tubes for hematological and biochemical analyses. Brain, heart, lungs, liver, kidneys, spleen, inguinal lymphatic nodes and tumor were excised, washed in PBS and weighed. Two animals from the vehicle group developed ulcerated tumors during the treatment period (day 28 post-implantation); thus, these animals were euthanized and excluded from the study.

Histological Analysis and Immunohistochemistry
Animal tissues were fixed in 10% buffered formalin, routinely processed and embedded in paraffin wax. Serial sections with 2 µm thickness were cut from each paraffin block; one section was stained with hematoxylin-eosin for histological examination, and the others were used for immunohistochemistry. Histological grading of the mammary carcinomas was performed according to the Nottingham histological grading method [25] based on a semi-quantitative assessment of three histological parameters: tubular formation, nuclear pleomorphism and mitotic figures count. The score of these parameters is used to assign the tumor grade: grade I (well-differentiated tumors), grade II (moderately differentiated tumors) or III (poorly differentiated tumors).
For the immunohistochemical study, tissue sections were dewaxed and rehydrated. Slides were submitted to proteolytic digestion by immersion in 10% retrieval solution (Dako) and kept in a water bath at 100 • C for 20 min. After blocking of non-specific staining, slides were incubated with primary antibody for Ki-67 (1:50 dilution, clone MIB-1, Dako) overnight at 4 • C. Slides were then stained with Novolink Polymer Detection Systems (Leica Biosystems Inc., Buffalo Grove, IL, USA) following manufacturer's instructions, counterstained with hematoxylin and permanently mounted. The nuclear Ki-67 staining was analyzed using Fiji's high-throughput automatic quantitation with Andy's Algorithm pipeline, as described elsewhere [26]. The immunoreactivity was evaluated, counting at least 600 nuclei/field from 5 representative fields of the lesion at high magnification, avoiding necrotic areas. The ratio of Ki-67 positive nuclei per all counted nuclei was expressed as a percentage (Ki-67 proliferation index).

Detection of Apoptosis with Labeling of DNA Strand Breaks (TUNEL Assay)
The paraffin-embedded tumor sections were deparaffinized, rehydrated through a graded series of ethanol and double distilled water and then stained with In Situ Cell Death Detection Kit (Roche Diagnostics GmbH, Mannheim, Germany) as per manufacturer's instructions. Apoptosis was quantified at single-cell level by counting the number of cells positive for terminal deoxynucleotidyl transferase (TdT) dUTP nick-end labeling (TUNELpositive cells). Briefly, tumor sections were incubated with 20 µg/mL proteinase K solution (in 10 mM Tris/HCl, pH 7.4-8) for 30 min at 37 • C. Slides were washed and stained with TUNEL reaction mixture (mixture of terminal deoxynucleotidyl transferase with labeled nucleotides) for 60 min at 37 • C in the dark followed by three washes. Slides were mounted with VECTASHIELD Antifade Mounting Medium with DAPI, and coverslips were sealed. At least twenty nonoverlapping high-power microscope fields of each tumor section were captured while avoiding areas with central necrosis. Fluorescence images were acquired using the LionHeart automated microscope at Ex: 520-560 nm and Em: 570-620 nm.

In Vivo CAM Assay
The chicken embryo chorioallantoic membrane (CAM) assay was used to study the effect of tested drugs on neovascularization, as described elsewhere [27]. Briefly, fertilized chicken eggs were incubated at 37.5 • C in a humidified atmosphere with agitation. After 3 days of incubation, 2.5 mL of albumen was removed to detach developing CAM from the shell. A window in the eggshell was cut off to expose the embryo, and then, the eggshell was sealed with paraffin. Incubation of fertilized eggs continued until day 9, when the windows were unsealed and four paper disks previously sterilized and pre-treated with hydrocortisone (a cyclooxygenase inhibitor to avoid inflammatory responses) were placed in direct contact with CAM in each egg (3 PBS-soaked disks and 1 VEGF-soaked (10 ng/mL) disk), and windows were re-sealed with paraffin. On day 11, two of the PBS-soaked disks were treated with Pd 2 Spm (2 to 8 µM) or cisplatin (2 to 8 µM). The remaining two disks were used as negative/vehicle (treated with PBS) and positive (treated with VEGF) controls. The eggs were further incubated two more days, and at day 13, the disks bound to the CAM were removed, placed in PBS and imaged using a contrast-phase microscope (Motic ® AE200 inverted microscope, Spectra Services, VWR International) (with a 4X magnification) coupled to a Moticam 5 digital camera. Images were analyzed using the Angiogenesis Analyzer for Fiji [28].

Ethical Considerations
Handling and care of animals were conducted according to Portuguese (Decreto-Lei n.

Statistical Data Analysis
Data are expressed as mean ± standard error of the mean (SEM) or medians with corresponding quartiles. Statistical analysis was performed using two-tailed Student's t-test or one-way ANOVA followed by Dunnett's multiple comparisons for analysis of means. Changes over the time were analyzed with two-way ANOVA followed by Tukey's multiple comparisons test. Medians were compared with nonparametric Kruskal-Wallis test followed by Dunn's multiple comparisons test. GraphPad Prism 7 Software (San Diego, CA, USA) was used. A p-value < 0.05 was considered statistically significant. Data are expressed as mean ± standard error of the mean (SEM) or medians with corresponding quartiles. Statistical analysis was performed using two-tailed Student's ttest or one-way ANOVA followed by Dunnett's multiple comparisons for analysis of means. Changes over the time were analyzed with two-way ANOVA followed by Tukey's multiple comparisons test. Medians were compared with nonparametric Kruskal-Wallis test followed by Dunn's multiple comparisons test. GraphPad Prism 7 Software (San Diego, CA, USA) was used. A p-value < 0.05 was considered statistically significant.

Pd2Spm Showed Selective Antiproliferative Activity for MDA-MB-231 Breast Cancer Cells
The effects of Pd2Spm and cisplatin on the proliferation of MDA-MB-231 and MCF-12A cells at 24, 48 and 72 h are shown in Figure 1 as dose-response curves. The full range of antiproliferative activity of Pd2Spm and cisplatin on net cell growth (0-100%) was achieved with the tested concentrations (1 to 100 µM). For MCF-12A cells, the highest concentration of Pd2Spm (100 µM) did not produce a maximum growth inhibition when compared to the same dosage of cisplatin. A clear separation of dose-response curves obtained for MDA-MB-231 and MCF-12A cells was observed for Pd2Spm at 24, 48 and 72 h incubation periods. For cisplatin, an overlap between the dose-response curves for MDA-MB-231 and MCF-12A was detected at these time points. Nonlinear regression analysis of dose-response curves was used to compute the half maximal inhibitory concentration (IC50) for Pd2Spm and cisplatin (Table 1). For Pd2Spm, the IC50 values obtained in MDA-MB-231 cells were significantly lower than the IC50 values determined in MCF-12A cells, suggesting lower sensitivity of non-cancerous MCF-12A to Pd2Spm. For cisplatin, on the other hand, the IC50 obtained for MDA-MB-231 and MCF-12A cells did not show significant differences, which is indicative of non-discriminatory antiproliferative effects elicited by cisplatin.    Table 2, showing no significant difference in the potency of Pd 2 Spm versus cisplatin in inhibiting the migration of MDA-MB-231 cells. Data are expressed as mean logEC 50 ± SEM (EC 50 in µM), n = 5.

Pd 2 Spm Administration Inhibited the Growth of MDA-MB-231 Xenografts in Female Nude Mice
To confirm the in vitro findings, we further studied the in vivo effects of intraperitoneal Pd 2 Spm administration on the growth of MDA-MB-231 cancer cells subcutaneously implanted in female nude mice. The cisplatin-treated mice were used as a reference treatment group. The selection of doses of Pd 2 Spm and cisplatin used in vivo took into consideration the doses we used previously for pharmacokinetic and tissue biodistribution studies [15]. The average tumor volumes in mice treated with either Pd 2 Spm (5 mg/kg/day) or cisplatin (2 mg/kg/day) for 5 consecutive days were significantly lower compared to vehicle-treated control mice (p < 0.05; Figure 3). At the end of the experiment, the tumor growth reduction (versus the vehicle group) was 43.3% and 56.6% for Pd 2 Spm-and cisplatin-treated animals, respectively.  eration the doses we used previously for pharmacokinetic and tissue biodistribution studies [15]. The average tumor volumes in mice treated with either Pd2Spm (5 mg/kg/day) or cisplatin (2 mg/kg/day) for 5 consecutive days were significantly lower compared to vehicle-treated control mice (p < 0.05; Figure 3). At the end of the experiment, the tumor growth reduction (versus the vehicle group) was 43.3% and 56.6% for Pd2Spmand cisplatin-treated animals, respectively.  Tumors treated with the vehicle showed highly dense cancer cells with atypical nuclei, evidencing intense mitosis. The mitotic count of vehicle-treated animals was significantly increased compared to cisplatin-and Pd 2 Spm-treated groups (Figure 4b). In order to evaluate whether Pd 2 Spm-mediated suppression of tumor growth occurred with a decrease in cell proliferation and/or DNA damage induction, the Ki-67 proliferation index (a wellaccepted marker of cellular proliferation, [30]) and the TUNEL assay (marker of DNA damage) were used. Immunoreactivity for Ki-67 in tumor sections from the groups of animals under study (Pd 2 Spm, cisplatin or vehicle) are shown in Figure 4a. The Ki-67 index of tumors was significantly decreased in animals treated with Pd 2 Spm when compared to animals treated with the vehicle. Tumors from cisplatin-treated animals also showed a significant decrease in Ki-67 relative to the vehicle-treated mice (Figure 4c). Analysis of DNA damage showed a significant increase in TUNEL-positive cells in tumors obtained from animals treated with Pd 2 Spm and cisplatin when compared to tumors from vehicletreated animals (Figure 4d).  Overall, these results evidenced that Pd 2 Spm administration inhibited the growth of MDA-MB-231 xenografts in vivo by reducing tumor cell proliferation and inducing DNA damage and cell death. Our data also showed that 5-day treatment of tumor-bearing mice with 5 mg/kg/day of Pd 2 Spm or 2 mg/kg/day of cisplatin results in comparable anticancer effects when (i) tumor volume reduction, (ii) tumor proliferative activity decrease (Ki-67 and mitotic count) and (iii) tumor DNA damage increase are compared in Pd 2 Spmversus cisplatin-treated animals. This may be explained in part by the similarity in the pharmacokinetics of Pd 2 Spm and cisplatin [15], as well as by the similarity in the amounts of administered metals (1.91 mg/kg/day of palladium administered from 5 mg/kg/day of Pd 2 Spm, and 1.3 mg/kg/day of platinum administered from 2 mg/kg/day of cisplatin).

Impact of Pd 2 Spm on the Weight of MDA-MB-231 Breast Tumor-Bearing Mice
As body weight can be considered a marker of general systemic toxicity, we performed a daily monitoring of the mice's weight in all groups. A five-day treatment administration induced notable changes in mice's body weight (Figure 5a): during the initial three days of the treatment, a body weight loss was observed in all study groups, and this trend continued in cisplatin-treated animals also after the administration period. A progressive recovery of the body weight was observed in all groups, and at the end of the experiment, the body weight of all animals was higher compared to the starting value. The cisplatin group demonstrated a slower body weight recovery compared to Pd 2 Spm-and vehicletreated groups. There was no significant difference in body weight between groups at the end of the procedure.
In addition to body weight decrease, cisplatin-treated animals showed signs of general toxicity, such as decreased movement/activity, transient dehydration and diarrhea. In turn, no such adverse effects were observed in animals treated with either Pd 2 Spm or the vehicle.
The treatment of animals with Pd 2 Spm, cisplatin or vehicle did not impact the relative weight of the liver, heart, lungs, spleen or brain, with no significant differences being detected among the studied groups (Table 3). On the other hand, the relative weight of the kidneys in cisplatin-treated animals was lower compared to animals treated with the vehicle or Pd 2 Spm. Histopathological evaluation showed a decreased perirenal adipose tissue in cisplatin-treated compared to vehicle-treated animals, which may further support a reduction in the relative weight of the kidneys in cisplatin-treated animals (Figure 5b). No other histopathological changes were observed during histological analysis of the kidneys (data not shown). Regarding the liver, histopathological analysis allowed us to conclude that cisplatin-treated animals exhibited a swollen, pale and finely vacuolated cytoplasm due to intracellular accumulation of glycogen (Figure 5b), as confirmed by periodic acid-Schiff (PAS). This glycogen accumulation was not observed in the livers of vehicle-and Pd 2 Spm-treated animals.

Impact of Pd 2 Spm on the Hematologic and Biochemical Parameters of MDA-MB-231 Breast Tumor-Bearing Mice
Hematologic analysis (RBC, Hb, Htc, MCV, MCH) of animals treated with Pd 2 Spm revealed no significant deviations from the animals treated with the vehicle (Figure 6a). Cisplatin-treated animals, in turn, showed a decrease in red blood cells, hemoglobin and hematocrit, accompanied by an increase in the mean corpuscular volume, as compared to vehicle-treated animals. Plasma biochemical analysis (urea, creatinine, glucose, AST, ALT, total proteins, triglycerides and total cholesterol) in Pd 2 Spm-treated animals did not reveal changes relative to vehicle-treated animals (Figure 6b). For cisplatin treatment, a decrease in total proteins and an increase in triglycerides were observed as compared to the vehicle-treated group.

Pd2Spm Suppressed Neovascularization in Chorioallantoic Membrane and Reduced VEGF Production in MDA-MB-231 Cells
As angiogenesis plays an important role in the carcinogenesis of TNBC [31], the antiangiogenic potential of Pd2Spm was investigated and compared to cisplatin. We determined the effect of Pd2Spm treatment on the neovascularization steps using the CAM as-

Pd 2 Spm Suppressed Neovascularization in Chorioallantoic Membrane and Reduced VEGF Production in MDA-MB-231 Cells
As angiogenesis plays an important role in the carcinogenesis of TNBC [31], the antiangiogenic potential of Pd 2 Spm was investigated and compared to cisplatin. We determined the effect of Pd 2 Spm treatment on the neovascularization steps using the CAM assay. Increasing concentrations of Pd 2 Spm (2 to 8 µM) and of cisplatin (2 to 8 µM) were able to decrease, in a dose-dependent manner, the number of nodes, junctions and segments, which are parameters associated with the early steps of angiogenesis (Figure 7a). Similarly, Pd 2 Spm and cisplatin, in the concentrations tested, reduced, in a dose-dependent manner, the total length, total branching length and total segment length, which are linked to the expansion of newly formed vessels in later steps of angiogenesis (Figure 7b). Thus, these data indicate that both Pd 2 Spm and cisplatin have the ability to inhibit angiogenesis, in both the initial and later steps, with a similar potency (p > 0.05).
Biomedicines 2022, 10, x FOR PEER REVIEW 17 of 22 dependent manner, the total length, total branching length and total segment length, which are linked to the expansion of newly formed vessels in later steps of angiogenesis ( Figure 7b). Thus, these data indicate that both Pd2Spm and cisplatin have the ability to inhibit angiogenesis, in both the initial and later steps, with a similar potency (p > 0.05). Since VEGF plays a crucial role in angiogenesis, we have explored the hypothesis that Pd2Spm may impact the secretion of VEGF by MDA-MB-231 cells. As shown in Figure  8 Since VEGF plays a crucial role in angiogenesis, we have explored the hypothesis that Pd 2 Spm may impact the secretion of VEGF by MDA-MB-231 cells. As shown in Figure 8, both Pd 2 Spm (5-20 µM) and cisplatin (5-20 µM) inhibited VEGF secretion in cell culture medium in a dose-dependent manner. Compared to cisplatin, Pd 2 Spm incubation resulted in a similar decrease in VEGF secretion by MDA-MB-231.
Together these results indicated that Pd 2 Spm may possess anti-angiogenic potential to be further explored and that this effect can be ascribed, at least in part, to its ability to reduce the secretion of VEGF. (red) or cisplatin (blue). Data are shown as means ± SEM (n = 2) and analyzed with one-way ANOVA followed by Dunnett's multiple comparison test. Significant differences from the control: ** p < 0.01, *** p < 0.001, **** p < 0.0001.
Together these results indicated that Pd2Spm may possess anti-angiogenic potential to be further explored and that this effect can be ascribed, at least in part, to its ability to reduce the secretion of VEGF.

Discussion
The in vivo anticancer effects of Pd2Spm treatment in mice were studied for the first time, unraveling the promising potential of this dinuclear Pd(II)-polyamine complex regarding its selectivity, efficacy and tolerability as compared to the clinically used Pt(II) drug cisplatin.
The treatment of mice bearing TNBC tumors with 5 mg/kg/day of Pd2Spm for five days (cumulative dose of 25 mg/kg) induced significant effects: (i) decrease in tumor volume to a similar extent to that observed in animals treated with 2 mg/kg/day of cisplatin (cumulative dose of 10 mg/kg); (ii) reduction in the number of mitoses in the tumor; (iii) decrease in the proliferative index measured with immunohistochemical marker Ki67; and (iv) increase in TUNEL-positive cells (in the tumor), which are indicative of increased DNA damage. The anti-proliferative and selective effects of Pd2Spm towards TNBC cells was evidenced by the significantly higher IC50 values obtained for non-cancerous MCF-12A cells (89.5-228.9 µM) as compared to MDA-MB-231 cancer cells (7.3-8.3 µM). This ca. 10-fold higher selectivity of Pd2Spm contrasts with the unselective activity of cisplatin, since this conventional Pt drug shows an equivalent potency towards both MDA-MB-231 and MCF-12A cells. Cisplatin's non-selective antiproliferative activity presently observed corroborates previous reports [32] and may indicate that cisplatin's non-discriminatory effects are responsible for the appearance of off-target toxicity and adverse effects during chemotherapy. For Pd2Spm, the established selectivity further confirms previous studies performed in human fibroblasts [9] and is a particularly promising result that may lead to a considerable reduction in adverse events during chemotherapy. Additionally, Pd2Spm consistently shows IC50 values within a low micromolar range in MDA-MB-231 [9,11] and MDA-MB-468 TNBC cells [33], which is a requirement for the successful course towards pre-clinical development and further translation into the clinics. Altogether, these in vitro studies sustain the evidence that TNBC is expected to present greater susceptibility to Pd2Spm than non-TNBC phenotypes, though the molecular mechanisms for these differential effects are yet to be unraveled. . Data are shown as means ± SEM (n = 2) and analyzed with one-way ANOVA followed by Dunnett's multiple comparison test. Significant differences from the control: ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Discussion
The in vivo anticancer effects of Pd 2 Spm treatment in mice were studied for the first time, unraveling the promising potential of this dinuclear Pd(II)-polyamine complex regarding its selectivity, efficacy and tolerability as compared to the clinically used Pt(II) drug cisplatin.
The treatment of mice bearing TNBC tumors with 5 mg/kg/day of Pd 2 Spm for five days (cumulative dose of 25 mg/kg) induced significant effects: (i) decrease in tumor volume to a similar extent to that observed in animals treated with 2 mg/kg/day of cisplatin (cumulative dose of 10 mg/kg); (ii) reduction in the number of mitoses in the tumor; (iii) decrease in the proliferative index measured with immunohistochemical marker Ki67; and (iv) increase in TUNEL-positive cells (in the tumor), which are indicative of increased DNA damage. The anti-proliferative and selective effects of Pd 2 Spm towards TNBC cells was evidenced by the significantly higher IC 50 values obtained for non-cancerous MCF-12A cells (89.5-228.9 µM) as compared to MDA-MB-231 cancer cells (7.3-8.3 µM). This ca. 10-fold higher selectivity of Pd 2 Spm contrasts with the unselective activity of cisplatin, since this conventional Pt drug shows an equivalent potency towards both MDA-MB-231 and MCF-12A cells. Cisplatin's non-selective antiproliferative activity presently observed corroborates previous reports [32] and may indicate that cisplatin's non-discriminatory effects are responsible for the appearance of off-target toxicity and adverse effects during chemotherapy. For Pd 2 Spm, the established selectivity further confirms previous studies performed in human fibroblasts [9] and is a particularly promising result that may lead to a considerable reduction in adverse events during chemotherapy. Additionally, Pd 2 Spm consistently shows IC 50 values within a low micromolar range in MDA-MB-231 [9,11] and MDA-MB-468 TNBC cells [33], which is a requirement for the successful course towards pre-clinical development and further translation into the clinics. Altogether, these in vitro studies sustain the evidence that TNBC is expected to present greater susceptibility to Pd 2 Spm than non-TNBC phenotypes, though the molecular mechanisms for these differential effects are yet to be unraveled.
The selectivity and higher tolerability of Pd 2 Spm treatment relative to cisplatin was also established from the analysis of the animal welfare, body weight, relative organ weight, histopathological features, hematology and serum biochemistry parameters. Animals treated with Pd 2 Spm did not show changes in body weight, gastrointestinal reactivity (diarrhea) or dehydration, as opposed to animals treated with cisplatin. Moreover, Pd 2 Spm administration did not result in significant variations regarding hematological and biochemical parameters when compared to cisplatin-treated animals which evidenced signs of macrocytic anemia (decrease in red blood cells, hemoglobin and hematocrit together with an increase in mean corpuscular volume). The cisplatin-induced changes in body weight, diarrhea and hematotoxicity were previously described [34,35] and are generally considered as surrogates for the drug's systemic toxicity. Furthermore, these results support the reduced off-target effects and low toxicity elicited by Pd 2 Spm treatment compared to cisplatin. Interestingly, besides a decrease in the total serum proteins and an increase in serum triacylglycerides, no other parameters related to renal or hepatic functions were altered in cisplatin-treated animals. In contrast, Pd 2 Spm treatment did not trigger any of these changes.
Since angiogenesis and cell migration are significant contributors to TNBC pathogenesis, particularly regarding its metastatic potential and aggressive biology, the development of agents able to suppress cancer cell proliferation, migration and angiogenesis are highly desired. In this study, Pd 2 Spm-mediated effects against cancer cell invasion yielded promising results since Pd 2 Spm: (i) inhibited the migration of MDA-MB-231 cells; (ii) inhibited angiogenesis in CAM; and (iii) suppressed VEGF secretion by MDA-MB-231 cells with similar potency to cisplatin. These results are in line with previous reports that revealed that cisplatin's anti-migratory and antiangiogenic effects, although the underlying mechanism(s) are still largely unknown [36,37]. Regarding the effects of Pd(II) compounds, there is limited information in the literature describing their ability to suppress angiogenesis and cell migration. So far, studies with the mononuclear palladium (II) saccharinate complex of terpyridine [PdCl(terpy)](sac)·2H 2 O have shown inhibition of the migration of endothelial cells (HUVEC) in the range of 12.5-50 µM and have also shown inhibition of angiogenesis in CAM assay at a concentration of 5 mg/mL [38]. In the current study, Pd 2 Spm significantly inhibited angiogenesis at an 8 µM concentration (corresponding to 4.46 µg/mL of Pd 2 Spm), impacting both early and late steps of angiogenesis and further supporting the previous reports [11]. Nevertheless, the mechanism behind this activity is still unexplored and requires further study.
The in vivo anticancer activity of Pd(II)-based compounds against TNBC is still largely unexplored, and to date, there are only a few in vivo studies in rodents regarding Ehrlich ascites carcinoma (EAC), which is an undifferentiated murine mammary adenocarcinoma sensitive to chemotherapy and not classified as TNBC [39,40]. The Pd complex of terpyridine with saccharinate, also known as [Pd(sac)(terpy)](sac)·4H 2 O, administered intraperitoneally at a dose of 2 mg/kg (two injections per week for 2 weeks, up to a cumulative dose of 8 mg/kg) was found to reduce tumor volume by 67.5%, but one mouse died during treatment (1 of 10), indicating the toxicity and poor tolerability of this compound [39]. Another study focused on the similar derivative [PdCl(terpy)](sac)·2H 2 O found that at a dose of 3 mg/kg (two injections per week for 2 weeks, up to a cumulative dose of 12 mg/kg), it reduced tumor volume by 69.8%, and one mouse died during treatment (1 of 9) [40]. Additionally, mice treated with 4 mg/kg of cisplatin (two injections per week for 2 weeks up to a cumulative dose of 16 mg/kg) showed 36.4% tumor volume reduction in a first study with two associated deaths [39] and a in 46.7% tumor volume reduction in a second study with no associated deaths [40]. When compared to the present results, it is clear that Pd 2 Spm shows a good ability to effectively reduce the growth of TNBC tumors by 43.3% at a cumulative dose of 25 mg/kg (5 mg/kg/day of Pd 2 Spm for 5 consecutive days) with no associated deaths, while cisplatin reduced the tumor growth by 56.6% at a cumulative dose of 10 mg/kg (2 mg/kg/day of cisplatin for 5 consecutive days). The selection of the 5 mg/kg/day dose of Pd 2 Spm accounted for Pd 2 Spm's pharmacokinetics and differential accumulation of palladium in cancer cells that were explored in previous studies [15]. Moreover, the dose selection also reflected the different molar ratios of palladium and platinum atoms in Pd 2 Spm and cisplatin molecules, respectively (i.e., 5 mg/kg/day of Pd 2 Spm corresponded to 1.91 mg/kg/day of the administered dose of palladium, and 2 mg/kg of cisplatin corresponded to 1.3 mg/kg/day of the administered dose of platinum) [15]. Finally, given the fact that the dose of Pd 2 Spm used in this study yielded a promising efficacy and good tolerability, future studies may be focused on a further increase in therapeutic outcomes using different treatment schedules and/or higher Pd 2 Spm concentrations.

Conclusions
This study constitutes the first proof-of-concept regarding the in vivo activity of Pd 2 Spm as a potential candidate chemotherapeutic for TNBC treatment. The in vivo administration of Pd 2 Spm in TNBC tumor-bearing mice resulted in tumor growth inhibition to a similar extent to that in animals treated with cisplatin. Furthermore, the systemic toxicity (hematotoxicity, weight loss, diarrhea and biochemical alterations) observed in cisplatin-treated animals was not verified in Pd 2 Spm-treated mice. Here we have provided encouraging evidence of the selective activity of Pd 2 Spm towards TNBC with minimal effects to non-cancerous breast cells, both in vitro and in vivo. Therefore, Pd 2 Spm may become a promising potential Pd(II) drug candidate for the treatment of TNBC.