Phosphoinositide 3-Kinase (PI3K) Reactive Oxygen Species (ROS)-Activated Prodrug in Combination with Anthracycline Impairs PI3K Signaling, Increases DNA Damage Response and Reduces Breast Cancer Cell Growth

RIDR-PI-103 is a novel reactive oxygen species (ROS)-induced drug release prodrug with a self-cyclizing moiety linked to a pan-PI3K inhibitor (PI-103). Under high ROS, PI-103 is released in a controlled manner to inhibit PI3K. The efficacy and bioavailability of RIDR-PI-103 in breast cancer remains unexplored. Cell viability of RIDR-PI-103 was assessed on breast cancer cells (MDA-MB-231, MDA-MB-361 and MDA-MB-453), non-tumorigenic MCF10A and fibroblasts. Matrigel colony formation, cell proliferation and migration assays examined the migratory properties of breast cancers upon treatment with RIDR-PI-103 and doxorubicin. Western blots determined the effect of doxorubicin ± RIDR-PI-103 on AKT activation and DNA damage response. Pharmacokinetic (PK) studies using C57BL/6J mice determined systemic exposure (plasma concentrations and overall area under the curve) and T1/2 of RIDR-PI-103. MDA-MB-453, MDA-MB-231 and MDA-MB-361 cells were sensitive to RIDR-PI-103 vs. MCF10A and normal fibroblast. Combination of doxorubicin and RIDR-PI-103 suppressed cancer cell growth and proliferation. Doxorubicin with RIDR-PI-103 inhibited p-AktS473, upregulated p-CHK1/2 and p-P53. PK studies showed that ~200 ng/mL (0.43 µM) RIDR-PI-103 is achievable in mice plasma with an initial dose of 20 mg/kg and a 10 h T1/2. (4) The prodrug RIDR-PI-103 could be a potential therapeutic for treatment of breast cancer patients.

Targeting the PI3K pathway with small molecular weight kinase inhibitors of PI3K, AKT, mTOR, HER2, or anti-HER2 antibodies has improved the outcome for many women

Doxorubicin Induces ROS and Combination of RIDR-PI-103 and Doxorubicin Inhibits Breast Cancer Cell Viability
tions of doxorubicin in combination with RIDR-PI-103 (10-30 µM) based on the IC50 values of doxorubicin in different cancer cell lines. We found that 10, 15 and 30 µM RIDR-PI-103 enhances the cytotoxicity of the anthracycline doxorubicin in MDA-MB-231, MDA-MB-361 and MDA-MB-453 ( Figure 2B-D). We further assessed the effect of the ROS scavenger, N-Acetyl Cysteine (NAC) alone or with RIDR-PI-103 +/-doxorubicin in MDA-MB-231 and MDA-MB-453 cells. We observed that NAC rescued the antiproliferative effects of RIDR-PI-103 and doxorubicin in MDA-MB-231 cells ( Figure S3A) but not in MDA-MB-453 cells ( Figure S3B). Our data indicate that MDA-MB-231 cells contain more endogenous ROS than MDA-MB-453 ( Figure 2A). We speculate that with increased ROS, MDA-MB-231 cells may be more sensitive to treatment with the ROS scavenger NAC, allowing for rescue of treatment with RIDR-PI-103 and doxorubicin. We also analyzed the effect of docetaxel [27], part of the taxane class that act as anti-microtubule agents, in combination with RIDR-PI-103. The data indicated that docetaxel was less effective in combination with RIDR-PI-103 to suppress breast cancer cell growth ( Figure S4A-C). IC50 values of DOXO, DOXO ± 10 µM RIDR, DOXO ± 15 µM RIDR and DOXO ± 30 µM RIDR in MDA-MB-231, MDA-MB-361 and MDA-MB-453 cell lines are provided in Table S2. Notably, in MDA-MB-231 cells the IC50 value was reduced by more than one half comparing doxorubicin and RIDR-PI-103 combination versus doxorubicin alone (Table S2).

Doxorubicin in Combination with RIDR-PI-103 Suppresses Breast Cancer Cell Proliferation
We speculated that the combination of doxorubicin and RIDR-PI-103 could effectively inhibit breast cancer cell proliferation. Accordingly, MDA-MB-231, MDA-MB-361 and MDA-MB-453 cells were treated with doxorubicin (125 nM) and RIDR-PI-103 (10 µM) every alternative day and stained after 10-12 days using a two-dimensional crystal violet assay. The data indicated that the combination of doxorubicin and RIDR-PI-103 significantly reduced cancer cell proliferation versus DMSO or single agent doxorubicin or RIDR-PI-103 in all three cell lines ( Figure 3A-C). for 72 h. The concentration of doxorubicin used was 100-4000 nM for MDA-MB-453. All the cells were treated with MTT (5 mg/mL) for 4 h and absorbance read at 570 nm in a microtiter plate reader (n = 3 independent experiments performed in triplicate ± SEM).

Doxorubicin in Combination with RIDR-PI-103 Suppresses Breast Cancer Cell Proliferation
We speculated that the combination of doxorubicin and RIDR-PI-103 could effectively inhibit breast cancer cell proliferation. Accordingly, MDA-MB-231, MDA-MB-361 and MDA-MB-453 cells were treated with doxorubicin (125 nM) and RIDR-PI-103 (10 µM) every alternative day and stained after 10-12 days using a two-dimensional crystal violet assay. The data indicated that the combination of doxorubicin and RIDR-PI-103 significantly reduced cancer cell proliferation versus DMSO or single agent doxorubicin or RIDR-PI-103 in all three cell lines ( Figure 3A-C).

Doxorubicin in Combination with RIDR-PI-103 Suppresses Matrigel Colony Formation
Examination of cells grown on a basement membrane of matrigel indicated that combination doxorubicin (125 nM) and RIDR-PI-103 (10 µM) inhibit colony formation better than either single agent. Our data showed that the combination of doxorubicin ond day and stained with crystal violet within 7-10 days (panel I). The intensities are represented as mean; Error bars: SEM (n = 3 independent experiments performed in triplicate). * p < 0.05 versus DMSO and **, # p < 0.05 versus individual treatment as indicated (panel II).

Doxorubicin in Combination with RIDR-PI-103 Suppresses Matrigel Colony Formation
Examination of cells grown on a basement membrane of matrigel indicated that combination doxorubicin (125 nM) and RIDR-PI-103 (10 µM) inhibit colony formation better than either single agent. Our data showed that the combination of doxorubicin with RIDR-PI-103 significantly suppressed colony formation on matrigel compared to individual drugs or DMSO

Doxorubicin and RIDR-PI-103 Suppress Breast Cancer Cell Migration
Transwell migration assay was performed to assess the effect of doxorubicin and RIDR-PI-103 on breast cancer cell migration. The data indicated that the combination of

Doxorubicin and RIDR-PI-103 Suppress Breast Cancer Cell Migration
Transwell migration assay was performed to assess the effect of doxorubicin and RIDR-PI-103 on breast cancer cell migration. The data indicated that the combination of doxorubicin and RIDR-PI-103 inhibit breast cancer cell migration better compared to single agent doxorubicin, RIDR-PI-103 or DMSO control. Our cell migration data also showed that in all three breast cancer cell lines co-treatment of doxorubicin and RIDR-PI-103 resulted in statistically significant reduction in migration compared to either single agent ( Figure 5A-C).
doxorubicin and RIDR-PI-103 inhibit breast cancer cell migration better compared to single agent doxorubicin, RIDR-PI-103 or DMSO control. Our cell migration data also showed that in all three breast cancer cell lines co-treatment of doxorubicin and RIDR-PI-103 resulted in statistically significant reduction in migration compared to either single agent ( Figure 5A-C).

The Combination of Doxorubicin and RIDR-PI-103 Results in Enhanced Inhibition of PI3K Signaling and Activates DNA Damage Response
We sought to examine RIDR-PI-103 activity via a series of western blots as shown in Figure 6. RIDR-PI-103 targets PI3K and the downstream AKT signaling pathway. There was no effect of either doxorubicin or RIDR-PI-103 in phosphorylation of AKT at Thr308 (data not shown). In MDA-MB-231 breast cancer cells, a combination of doxorubicin with RIDR-PI-103 suppressed AKTSer473 phosphorylation as compared to single agent doxorubicin or RIDR-PI-103 ( Figure 6). Doxorubicin leads to DNA damage. The protein p53 is an imperative tumor suppressor molecule which plays an important role in DNA damage signaling and various genomic alterations [28]. Activation of p53 leads to cell cycle arrest or apoptosis. DNA damage induces phosphorylation of p53 at Ser15 and Ser20 and leads

The Combination of Doxorubicin and RIDR-PI-103 Results in Enhanced Inhibition of PI3K Signaling and Activates DNA Damage Response
We sought to examine RIDR-PI-103 activity via a series of western blots as shown in Figure 6. RIDR-PI-103 targets PI3K and the downstream AKT signaling pathway. There was no effect of either doxorubicin or RIDR-PI-103 in phosphorylation of AKT at Thr308 (data not shown). In MDA-MB-231 breast cancer cells, a combination of doxorubicin with RIDR-PI-103 suppressed AKTSer473 phosphorylation as compared to single agent doxorubicin or RIDR-PI-103 ( Figure 6). Doxorubicin leads to DNA damage. The protein p53 is an imperative tumor suppressor molecule which plays an important role in DNA damage signaling and various genomic alterations [28]. Activation of p53 leads to cell cycle arrest or apoptosis. DNA damage induces phosphorylation of p53 at Ser15 and Ser20 and leads to a reduced interaction between p53 and its negative regulator, the oncoprotein MDM2 [29,30]. Our data indicated that doxorubicin (125 nM) in combination with RIDR-PI-103 (10 µM) induced p53 phosphorylation at Ser15 in MDA-MB-231 breast cancer cell lines with similar results in MDA-MB-361 and MDA-MB-453 cells. However, we didn't observe p53 posttranslational modification activation in vehicle (DMSO) or doxorubicin treatment alone. We further examined the effect of the combination on cell cycle arrest. CHK1 kinase is downstream of ATM/ATR kinase pathway and plays a crucial role in DNA damage checkpoint control and tumor suppression. Doxorubicin in combination with RIDR-PI-103 induced p-CHK1 compared to either single agent. Checkpoint kinase 2 protein, (CHK2) is also downstream of ATM/ATR. CHK2 contain a series of seven serine/threonine residues (Ser19, Thr26, Ser28, Ser33, Ser35, Ser50, and Thr68) each followed by glutamine (SQ or TQ motif) [31]. Examining the phosphorylation of Thr68 of CHK2, our results indicate that doxorubicin, RIDR-PI-103, and the combination increased phosphorylation of CHK2 at T68 most notably in MDA-MB-361 and MDA-MB-453 versus vehicle control. All original images from western blots are included in Figures S5-S13. to a reduced interaction between p53 and its negative regulator, the oncoprotein MDM2 [29,30]. Our data indicated that doxorubicin (125 nM) in combination with RIDR-PI-103 (10 µM) induced p53 phosphorylation at Ser15 in MDA-MB-231 breast cancer cell lines with similar results in MDA-MB-361 and MDA-MB-453 cells. However, we didn't observe p53 post-translational modification activation in vehicle (DMSO) or doxorubicin treatment alone. We further examined the effect of the combination on cell cycle arrest. CHK1 kinase is downstream of ATM/ATR kinase pathway and plays a crucial role in DNA damage checkpoint control and tumor suppression. Doxorubicin in combination with RIDR-PI-103 induced p-CHK1 compared to either single agent. Checkpoint kinase 2 protein, (CHK2) is also downstream of ATM/ATR. CHK2 contain a series of seven serine/threonine residues (Ser19, Thr26, Ser28, Ser33, Ser35, Ser50, and Thr68) each followed by glutamine (SQ or TQ motif) [31]. Examining the phosphorylation of Thr68 of CHK2, our results indicate that doxorubicin, RIDR-PI-103, and the combination increased phosphorylation of CHK2 at T68 most notably in MDA-MB-361 and MDA-MB-453 versus vehicle control. All original images from western blots are included in Figures S5-S13.

Pharmacokinetics Profile of RIDR-PI-103
In this initial PK study in mice, RIDR-PI-103 (20 mg/kg) was administered intraperitoneally and blood samples were collected over a period of 96 h. The concentration of 20 mg/kg was chosen based on the ability to solubilize RIDR-PI-103 in 40% propylene glycol with 60% injectable saline. The mean plasma concentration-time profile was then analyzed using Phoenix ® WinNonlin ® v8.2 using both a compartmental and non-compartmental analysis (NCA). As shown ( Figure 7A,B), the plasma concentration-time profile fitted a one-compartment model. The key PK parameters are listed in Table 1. As shown, both methods of PK analysis yielded similar results. RIDR-PI-103 has a maximal plasma concentration (Cmax) of 201.5 ng/mL (0.43 µM), which was achieved at the time taken to reach the maximum concentration (Tmax) of 1.44 h. The elimination half-life of RIDR-PI-103 was 9.4 h ( Table 2), aligning with the blood sample schedule employed (up to 96 h, approximately a period equivalent to 10 half-lives) facilitating complete characterization of the elimination profile. Based on these studies, RIDR-PI-103 has a large volume of distribution (Vd), 89 L/kg.

Pharmacokinetics Profile of RIDR-PI-103
In this initial PK study in mice, RIDR-PI-103 (20 mg/kg) was administered intraperitoneally and blood samples were collected over a period of 96 h. The concentration of 20 mg/kg was chosen based on the ability to solubilize RIDR-PI-103 in 40% propylene glycol with 60% injectable saline. The mean plasma concentration-time profile was then analyzed using Phoenix ® WinNonlin ® v8.2 using both a compartmental and noncompartmental analysis (NCA). As shown ( Figure 7A,B), the plasma concentration-time profile fitted a one-compartment model. The key PK parameters are listed in Table 1. As shown, both methods of PK analysis yielded similar results. RIDR-PI-103 has a maximal plasma concentration (C max ) of 201.5 ng/mL (0.43 µM), which was achieved at the time taken to reach the maximum concentration (T max ) of 1.44 h. The elimination half-life of RIDR-PI-103 was 9.4 h ( Table 2), aligning with the blood sample schedule employed (up to 96 h, approximately a period equivalent to 10 half-lives) facilitating complete characterization of the elimination profile. Based on these studies, RIDR-PI-103 has a large volume of distribution (Vd), 89 L/kg.

Discussion
The PI3K/AKT/mTOR pathway is an important therapeutic target for treatment of breast cancer [2,32]. Several PI3K inhibitors are approved by the FDA for different cancers: Alpelisib in HR+ and HER2/neu negative breast cancer [11], idelalisib for leukemia [33] and two types of lymphoma [34,35], duvelisib for chronic lymphocytic leukemia/small lymphocytic lymphoma [36]. Several other PI3K inhibitors are also in different stages of clinical development [37][38][39][40]. Drugs targeting PI3K or mTOR catalytic activity are toxic, due to the physiological roles of PI3K/mTOR signaling in basic cellular processes in tissues throughout the body, including protein translation, intracellular trafficking, autophagy, and metabolism. Toxicity that occurs in patients includes hyperglycemia as genes encoding most glycolytic enzymes are under transcriptional control by PI3K/AKT [7].We designed a novel PI-103 prodrug (RIDR-PI-103) such that the PI3K inhibitor PI-103 would only be released under high oxidative stress conditions found in the tumor milieu. Thus, RIDR-PI-103 is designed to only inhibit PI3K in the tumor microenvironment under high oxidative stress conditions. This would circumvent the toxicity observed with global inhibition of PI3K throughout the body using current PI3K inhibitors. The study herein describes the efficacy of RIDR-PI-103 in breast cancer cells along with initial PK profile in a mouse model.
Enhanced ROS in non-transformed cells or breast cancer cells could have pro-tumorigenic effects via damaging nucleic acids and inducing genomic instability. Breast cancer subtypes demonstrate differential ROS production and susceptibility to antioxidant treatment. TNBC cells have increased ROS levels compared to non-tumorigenic or ER+/luminal breast cancer cells. TNBCs have higher oxidation states marked with enhanced ROS marker (dihydroethidium and MitoSox) staining compared to ER+ and non-tumorigenic control [26]. This is consistent with our findings demonstrating that TNBC MDA-MB-231 cells have high ROS compared to non-tumorigenic MCF10A cells. Doxorubicin produces ROS in vivo. Further, ROS plays a mechanistic role in the cardiotoxicity induced by doxorubicin involved in breast cancer treatment [41,42]. We found that doxorubicin at a concentration 125 nM induced ROS in breast cancer cell lines (MDA-MB-231, MDA-MB-361 and MDA-MB-453) which was higher than that induced in MCF10A cells (Figure 2A).
The PI3K pathway is significant in maintaining genomic stability by involving DNA replication and cell cycle regulation [43]. For example mTOR inhibition, downstream of PI3K, has been shown to activate p-CHK2 T68, an indicator of ATM-CHK2 checkpoint activation [44]. We examined the effect of combination doxorubicin and RIDR-PI-103 in Akt and DNA damage response signaling. Our data showed that the combination of doxorubicin and RIDR-PI-103 activated p-53 in three cancer cell lines. We observed that doxorubicin and RIDR-PI-103 combined upregulated p-CHK1S345 and p-CHK2T68 in MDA-MB-361 and MDA-MB-453 cells. PI3K inhibition can induce DNA damage via nucleoside depletion notably in cells with genetic aberrations in p53 and BRCA1 [45]. MDA-MB-453 cells have been reported to have a mutation in exon 11 of p53. MDA-MB-231 cells contains a R280K p53 mutation and MDA-MB-361 cells contains a p53 mutation in exon 4 [46]. Thus, inhibition of PI3K with RIDR-PI-103 in these breast cancers could reduce nucleotide triphosphates, resulting in DNA damage and activation of DNA damage response. The exact mechanism(s) by which RIDR-PI-103 can activate a DNA damage response as indicated by increased p-p53, p-CHK1 and p-CHK2 remain to be explored in future studies.
Recently, a bioisostere of PI-103 has been developed with a structural modification containing a boronate in place of a phenolic hydroxyl group with the goal to enhance bioavailability [47]. The boron-containing PI-103 bioisostere demonstrated improved bioavailability relative to PI-103. RIDR-PI-103 differs from the boron-containing compound in that RIDR-PI-103 is designed to only release PI-103 under controlled oxidative stress conditions when ROS levels are high, present in the tumor microenvironment milieu. Our goal with the design of RIDR-PI-103 is to circumvent toxicity observed with systemic inhibition of PI3K signaling throughout the body [48].
Herein, we performed an initial PK study meant to estimate some of the drug disposition parameters such as the peak plasma levels, elimination half-life, apparent volume of distributions and total clearance. Since we have not performed a PK study following an intravenous (i.v.) administration, the bioavailability following the intraperitoneal (i.p.) administration cannot be ascertained at this point. The peak plasma concentration at a dose of 20 mg/kg was 0.43 µM. The apparent volume of distribution (89 L) is suggestive of extensive drug distribution while the half-life of approximately 10 h suggesting that the systemic drug elimination is relatively slow. Thus, these derived PK data imply that an efficacious dosing regimen can be developed to yield effective tissue-specific drug levels.
Our study has limitations. The scope of the study focuses on in vitro studies with one in vivo pharmacokinetic study. The PK study was performed in tumor-free mice. In follow up studies, we will first derive a correlation between extracellular (cell culture media) and intracellular concentrations of RIDR-PI-103 which will then reflect on tumor-tissue specific concentrations needed for efficacy in animal models. The PK data derived here can then be employed for designing effective dosing regimen (dose and dosing frequency) by simulating PK profiles assuming linear PK (i.e., no change in the half-life, clearance and volume of distribution). Additional PK studies will measure the amount of PI-103 present, as it will be important to assess the conversion of pro-drug to biologically active drug. Future PK studies will employ a higher dose of RIDR-PI-103 with the initial dose described here using only 20 mg/kg RIDR-PI-103.
Future studies will examine the in vivo efficacy of single-agent RIDR-PI-103 compared to parent drug PI-103 in breast tumor models. There is evidence that inhibition of PI3K

Cell Viabilit
Growth k MB-453 cells w bromide (MTT protects from anthracycline toxicity [49]. Additional in vivo efficacy studies will examine if RIDR-PI-103 has a cardioprotective effect in combination with doxorubicin as PI-103 is a pan-PI3K inhibitor with an IC50 against PI3K Int. J. Mol. Sci. 2021, 22,2088 systemic drug elimination is relatively slow. Thus, the efficacious dosing regimen can be developed to yield e Our study has limitations. The scope of the study f in vivo pharmacokinetic study. The PK study was perfo up studies, we will first derive a correlation between ex intracellular concentrations of RIDR-PI-103 which wil cific concentrations needed for efficacy in animal mod then be employed for designing effective dosing regim simulating PK profiles assuming linear PK (i.e., no cha volume of distribution). Additional PK studies will mea as it will be important to assess the conversion of pro Future PK studies will employ a higher dose of RID scribed here using only 20 mg/kg RIDR-PI-103.
Future studies will examine the in vivo efficacy pared to parent drug PI-103 in breast tumor models. T PI3Kɣ protects from anthracycline toxicity [49]. Addit examine if RIDR-PI-103 has a cardioprotective effect in PI-103 is a pan-PI3K inhibitor with an IC50 against P results are promising and further help in PK modeling cious starting dose for efficacy studies utilizing novel R  [20]. Doxorubicin (Cat# Dwere obtained from LC Laboratories (Woburn, MA, US pared in DMSO solvent.

Cell Viability Assay
Growth kinetics of fibroblasts, MCF10A, MDA-M MB-453 cells was determined by the 4,5-dimethylthi bromide (MTT) assay. Briefly, 2 × 10 4 cells/well were se of 15 nM [50]. Overall, these results are promising and further help in PK modeling and deduction of a safe and efficacious starting dose for efficacy studies utilizing novel RIDR-PI-103 in breast cancer model.  [20]. Doxorubicin (Cat# D-4000) and docetaxel (Cat# D-1000) were obtained from LC Laboratories (Woburn, MA, USA). All these compounds were prepared in DMSO solvent.

Matrigel Colony Formation
Three-dimensional (3D) growth assays were performed in growth factor-reduced matrigel (BD Biosciences, San Jose, CA, USA) where 96 well plates were coated with 80 µL of matrigel/well. MDA-MB-231, MDA-MB-361 and MDA-MB-453 cells (1 × 10 4 /well) were plated and incubated at 37 • C for 24 h. Cells were treated with vehicle (DMSO), 10 µM RIDR-PI-103 or 125 nM doxorubicin or combination of both every alternate day. After 10 days of incubation, colonies were visualized and photographs were captured from 3 random fields under microscope (Nikon, Road Melville, NY, USA) at 10× magnification. The areas of the colonies were measured by ImageJ and represented as mean areas normalized to DMSO control. The experiments were repeated 3 times to confirm the results.

Analytical Methods
A high performance liquid chromatography (HPLC) was used to analyze serial dilutions of RIDR-PI-103 (1000, 500, 250, 125, 12.5 ng/mL) as described previously [28]. Briefly, we employed a Waters column (Milford, MA, USA) and the samples were eluted using mobile phases A: 95% water + 5% acetonitrile and B: 5% water + 95% acetonitrile. The HPLC was performed using a gradient method, with 0% B for 4 min and 95% B for over 15 min. The flow was 1 mL/min. A detection wavelength of 250 nm was used on Waters 2487 Dual Wavelength Absorbance Detector. Once the HPLC method was developed, we then set up a liquid-liquid extraction method for RIDR-PI-103 from mouse plasma. Extractions were performed using standard liquid-liquid extraction process with methyl tert-butyl ether (MTBE) used as extraction solvent. The samples were reconstituted in 100 µL of PBS: DMSO (99:1 v/v) and the pH was adjusted to 2 using 1 M hydrochloric acid (HCl). The extraction efficiency was observed to be~40% with the limit of detection between 1.25 ng/mL and 1250 ng/mL.

Pharmacokinetic Analysis
The plasma concentration-time profile of RIDR-PI-103 was analyzed using Phoenix ® WinNonlin ® v8.2 (Certara L.P. (Pharsight), St. Louis, MO, USA). Regression analysis of the data suggested that a compartment model best fitted the data. This was based on observed goodness of fit (random distribution of the residual around the predicted curve) and the Akaike Information Criterion (AIC), and the Schwarz Criterion (SC). The important PK parameters derived using this approach included elimination half-life (T 1 2 ), time required to achieve peak plasma concentrations (T max ), total Area Under the concentration-time Curve (AUC0-∞) and systemic oral clearance (CL/F) and apparent volume of distribution (Vd/F). "F" is the bioavailability following the intraperitoneal (i.p) administration and cannot be estimated in the absence of intravenous (i.v.) administration. We also employed the non-compartmental analysis (NCA) to verify the PK parameters derived using the one-compartment model. The NCA utilizes the trapezoidal rule for AUC (0-tlast) determination over the blood sample collection period. The regression analysis of the last four concentration-time values (24,48, 72 and 96 h) was done to determine the elimination rate constant (Kel). PK parameters such as AUC (0-∞), Cl/F, Vd/F and the elimination half-life were then computed as shown below: AUC (0-∞) = AUC (0-tlast) + Clast/kel (1) Vd/F = (Cl/F)/kel (3)

Animal Studies
Female C57BL/6J mice at 4 weeks were used with n = 3 per time point. Time points for blood collection were based on previous findings regarding the in vitro microsomal metabolic stability of RIDR-PI-103 [20]. RIDR-PI-103 was formulated using a mixture of 40% propylene glycol with 60% injectable saline in which RIDR-PI-103 was soluble in solution and not a suspension. The stability of RIDR-PI-103 in this formulation was ascertained by measuring drug content over 7 days. RIDR-PI-103 (dose = 20 mg/kg) was injected intraperitoneally in all mice at the start of the experiment. Blood collection was done via cardiac puncture under anesthesia at 0, 0.5, 4, 6, 24, 48, 72, 96 h post injection. Plasma was isolated from the blood samples by centrifugation and was stored at −80 • C until further analysis. Institutional Animal Care and Use Committee (IACUC), University of Cincinnati ethically approved the in vivo mouse experiment under protocol AM02-19-08-28-01, approved on 27 July 2020.

Statistical Analysis
Data are shown as the mean ± standard error of mean (SEM) and representative of at least three independent experiments unless otherwise indicated. Statistical analysis among groups using the two-tailed Student's t-test, one-way analysis of variance, p < 0.05 was considered statistically significant.

Conclusions
We describe a ROS-activated prodrug (RIDR-PI-103) that in combination with doxorubicin inhibits breast tumor growth and migration. Our data indicated that doxorubicin with RIDR-PI-103 downregulated Akt signaling and activated DNA damage response signals.