Drug-Triggered Self-Assembly of Linear Polymer into Nanoparticles for Simultaneous Delivery of Hydrophobic and Hydrophilic Drugs in Breast Cancer Cells

ABSTRACT: Breast cancer is the most devastating disease among females globally. Conventional chemotherapeutic regimen relies on the use of highly cytotoxic drugs as monotherapy and combination therapy leading to severe side effects to the patients as collateral damage. Moreover, combining hydrophobic and hydrophilic drugs create erratic biodistribution and suboptimal medicinal outcome. Hence, packaging multiple drugs of diverse mechanisms of action and biodistribution for safe delivery into tumor tissues with optimal dosages is indispensable for next- generation breast cancer therapy. To address these, in this report, we describe a unique cisplatin-triggered self-assembly of linear polymer into 3D-spherical sub 200 nm particles. These nano- particles comprise a hydrophobic (paclitaxel) and hydrophilic drug (cisplatin) simultaneously in a single particle. Molecular dynamics
simulation revealed hydrophilic−hydrophilic interaction and interchain H-bonding as underlying mechanisms of self-assembly. Confocal microscopy studies evidently demonstrated that these novel nanoparticles can home into lysosomes in breast cancer cells, fragment subcellular nuclei, and prevent cell division, leading to improved breast cancer cell death compared to free drug combination. Moreover, 3D-breast tumor spheroids were reduced remarkably by the treatment of these nanoparticles within 24 h. These dual-drug-loaded self-assembled polymeric nanoparticles have prospective to be translated into a clinical strategy for breast cancer patients.

1.INTRODUCTION
In recent years, breast cancer has emerged as the most frequently diagnosed cancer and foremost reason of casualties among females, with ∼1.7 million new cases and 0.6 million deaths per year globally.1 Traditional treatments involve surgical removal of tumor (or breast) along with radiation therapy, hormonal therapy, and chemotherapy.2 In adjuvant, neo-adjuvant chemotherapy and advanced stages of breast cancers, several small molecule cytotoxic drugs [paclitaxel (PTX), cisplatin (CDDP), 5-fluorouracil, and doxorubicin] are widely used in clinics.3−8 Unfortunately, because of tumor heterogeneity and drug resistance mechanisms (intrinsic and extrinsic), most of the cancer cells evade single drug treatment, leading to resort on combination therapy for improved efficacy.9−12 Drug combination regimens are exploitedextensively in clinics for the treatment of breast cancer.13−16However, cytotoxic drug combinations generate severe augmented dose-limiting toxic side effects to the patients as collateral damage. Moreover, combination of drugs with entirely different water solubility (hydrophobic and hydrophilicdrugs) leads to inconsistent biodistribution, hence poor accumulation in appropriate dose in the cancerous tissue preventing desired therapeutic outcome. Nanotechnology- based tools exhibit the promise to address these issues.In the last decade, nanoscale platforms have changed the direction of cancer chemotherapy.17−20 Myriads of different nanovectors have been developed to package multiple therapeutic materials (small molecule drugs, antibodies, siRNAs, mRNAs, and proteins).

Nanoscale platforms can specifically accumulate into tumor tissues by unique dysfunctional leaky vasculature as well as receptor-mediated active targeting.28,29 Several nanovectors containing cytotoxic drugs are already in clinics or in clinical trials for the treatment of different types of cancers including breast cancer.20,30 Despite having tremendous advancement in nanotechnology- based tool kits for monodrug or combination drug delivery,amalgamating hydrophobic and hydrophilic drugs in a single nanoplatform with controlled loading and release profile remained a major challenge to overcome the erratic biodistribution and improve therapeutic efficacy.31−34To address this, herein, we illustrate a simple and robustsynthesis of CDDP-mediated self-assembled poly(isobutylene- alt-maleic anhydride) (PMAn) nanoparticles to inhibit cancer cell division leading to a significant reduction in 3D-breast cancer spheroids. These polymeric nanoparticles can comprise the hydrophobic, microtubule-stabilizing drug PTX by ester linkage. However, upon reaction with hydrophilic CDDP [Food and Drug Administration (FDA)-approved anticancer drug], a remarkable morphological conversion from 2D- structures into 3D-spherical nanoparticles was observed. Molecular dynamics (MD) simulation confirmed that thehydrophilic−hydrophilic interaction and interpolymer chain H- bonding triggered CDDP-mediated self-assembly of linear polymers into spherical nanoscale particles. The hydro- phobic−hydrophilic dual-drug-loaded polymeric NPs were compartmentalized into subcellular lysosomes followed by nuclear fragmentation and stalled cell division directing to a significant reduction in breast cancer spheroid formation. In this study, we have chosen PTX and CDDP because of their (i)highly hydrophobic and hydrophilic nature, respectively, causing different biodistribution, (ii) extensive use as monotherapy and combination therapy in clinics because of FDA approval despite having severe toxic side effects to the patients, and (iii) different mechanisms of action.

2.RESULTS AND DISCUSSION
Synthesis of the Polymer−Drug Conjugate and Self-Assembly. Sequential conjugation of hydrophobic− hydrophilic drugs to polymer and self-assembly are depictedin Scheme 1a,b. First, PMAn35,36 (1) was completely hydrolyzed into poly(isobutylene-alt-maleic acid) (PMAc) (2) followed by conjugation of hydrophobic drug PTX (PMAc/PTX = 1:5 molar ratio) through ester linkage with the 2′-OH group of PTX to form the PMA−PTX conjugate (3) (Scheme1a). CDDP was further conjugated with PMA−PTX (PMA− PTX/CDDP = 1:20 molar ratio) to obtain the PMA−PTX− CDDP conjugate (4). The hydrolyzed PMAc (2), PMA−PTX conjugate (3), and PMA−PTX−CDDP conjugate (4) were characterized by 1H NMR spectroscopy (Figures S1−S3). We further calculated the number of PTX molecules conjugated ineach polymer chain by 1H NMR spectroscopy. The number of CH3 protons in the polymer and PTX and the ortho-protons in aromatic ester and aromatic amide moieties in PTX were calculated from the 1H NMR spectra in Figure S3. The ratio of CH3 protons and aromatic protons confirmed that nearly five PTX molecules were conjugated in each polymer chain. PMA−PTX−CDDP conjugate (4) was further confirmed by 195Pt NMR having a characteristic peak at δ = −2572.6 ppm (FigureS4).To visualize the shape and morphology, the PMA−PTX conjugate (3) was subjected to field-emission scanning electron microscopy (FESEM).

From the FESEM image in Figure 1a, it was confirmed that the PMA−PTX conjugate exhibited a polymeric 2D-sheet-like structure. Interestingly, a remarkable transformation of morphology was observed in the PMA− PTX−CDDP conjugate (4). Electron microscopy [FESEM, atomic force microscopy (AFM), and transmission electron microscopy (TEM)] images (Figures 1b,c and S5c) clearly demonstrated that the reaction with the hydrophilic drug CDDP to PMA−PTX transformed its shape into spherical nanoparticles of sub 200 nm diameter. We further confirmed the self-assembled particle nature of the PMA−PTX−CDDP conjugate in water by dynamic light scattering (DLS) and the Tyndall effect (Figure S5a,b). The critical aggregation concentration (CAC) at which the PMA−PTX−CDDP conjugate self-assembled into nanoparticles was determined by conventional fluorescence emission spectroscopy of pyrene encapsulation and was calculated to be 60 μg/mL (Figure 1d). The loading of PTX and CDDP in the nanoparticle was determined by UV−vis spectroscopy through the absorbance versus concentration calibration graph at characteristic λmax = 273 and 706 nm, respectively. PTX and CDDP loading wasfound to be 202.9 ± 8 μM and 1563.0 ± 3 μM, respectively (Figure S5d). Finally, the presence of CDDP in PMA−PTX− CDDP-NP was further validated by energy-dispersion X-ray spectroscopy (EDXS) (Figure S6).CDDP-Induced Self-Assembly.

Electron Mi- croscopy (FESEM and AFM). To understand the role of CDDP in inducing the self-assembly, we reacted PMAc (2) with aquated CDDP in a ratiometric manner (PMAc/CDDP = 1:5, 1:10, and 1:20 molar ratio) to obtain PMAc−CDDP conjugates with different CDDP contents (Figure 2a). The morphologicaltransformation of PMAc−CDDP conjugates were further visualized by FESEM and AFM. FESEM images in Figure 2b evidently confirmed that CDDP induced the self-assembly of linear polymer PMAc into spherical shaped nanoparticles in PMAc/CDDP = 1:10 molar ratio. The same observation was further validated by AFM images (Figure 2c) which confirmed that CDDP is the responsible agent for self-assembly of polymeric PMAc. The presence of CDDP in PMA−CDDP- NPs was further confirmed by EDXS (Figure S7).Molecular Dynamics Simulation. To evaluate the mechanism of self-assembly, MD simulation was performed on the PMA−PTX−CDDP polymer chain having the PTX/ CDDP molar ratio of 1:4 using GROMACS-4.6.3 package.37 The initial energy minimized structure and self-assembled structures after 500 ns of simulation (Figure 3a,b) showed that PTX, CDDP, and polymeric carboxylic acid (−COOH) groups were distributed throughout the whole self-assembled structure. To understand the structural arrangements between hydro- phobic PTX, the center of masses (COM) between PTX monomers were calculated over last 10 ns of simulation trajectory.

The small peak near 1.3 nm (Figure S8) indicated that PTX units tend to aggregate near each other, although because of other predominant interactions (H-bonding), hydrophobic aggregation was not enhanced as expected.To evaluate the interaction between hydrophobic (PTX) andhydrophilic (CDDP, COOH) residues near each other, the radial distribution function (RDF) between similar types as well as different types of residues were calculated (Tables S1−S3). Figure 3c clearly delineated that polymeric COOH groups remained closest to each other (within 0.5 nm), whereas CDDP residues remained more distant from each other (∼0.6 nm and higher). By contrast, as expected hydrophobic PTX residues were not found to be aggregated. On the other hand, RDF between different types of residues showed sharp peaks between hydrophilic CDDP and COOH groups (Figure 3d).Additionally, hydrophobic PTX showed interactions to a lesser extent with hydrophilic COOH and CDDP residues because of their hydrophobic mismatch and steric bulk. Furthermore, the hydrogen bonding interaction between polymer chains may play a role in the process of self-assembly. The different types of hydrogen bond donor and acceptor sites like −OH, −NH, and−CO are present in the polymer chain. This gives rises tothe possibility of formation of hydrogen bonds in between different monomers of the same polymer chain (intra- molecular) and in between different chains (intermolecular). The distance distribution was calculated between different possible H-bonding donor and acceptor sites (−OH, −NH−,and −CO−), which showed that H-bonds formed only inbetween −OH and −CO− groups (Figure S9). Further calculation over 10 ns of simulation time revealed six interchains, an insignificant number of intrachains, and a highnumber of H-bonds to be formed with water molecules (Figure 3e,f).

This MD simulation study illustrated the self-assembly of the PMA−PTX−CDDP polymer into nanoparticles through the interpolymer chain H-bonding and hydrophilic−hydro- philic interaction between COOH and CDDP residues.For effective delivery ofhydrophobic and hydrophilic drugs together, the nanovector needs to be internalized inside the cancer cells. To visualize the self-assembled polymeric nanoparticles inside the cells, red fluorescent rhodamine-isothiocyanate (RITC) was tagged with PMAn through the ethylenediamine (ED) linker. First, the anhydride moiety of PMAn (1) was opened up using the ED linker [PMAn/ED = 1:5 molar ratio] to obtain the PMAn−ED conjugate (5) (Figure S10). RITC (6) was further reacted with free amine moiety of the PMAn−ED conjugate (5) to obtainthe RITC-labeled PMAn−RITC conjugate (7) (PMAn−ED/RITC = 1:5 molar ratio). Both conjugates 5 and 7 were characterized by 1H NMR spectroscopy (Figures S11 and S12). The anhydride moieties of the PMAn−RITC conjugate (7) were opened up by using dimethyl formamide (DMF)/water mixture (1:1) at 60 °C for 48 h to obtain the PMAn−RITC conjugate (8), which was further conjugated with PTX and CDDP sequentially (PTX/CDDP = 1:4 molar ratio) to afford the PMAc−RITC−PTX−CDDP conjugate (9) (Figure S10).Expectedly, the PMA−RITC−PTX−CDDP conjugate (9) self-assembled into the nanoparticles, which was confirmed by FESEM and AFM images along with energy dispersive analysis of X-rays for the confirmation of CDDP in the nanoparticles (Figures S13 and S14). MCF-7 breast cancer cells were treated with PMA−RITC−PTX−CDDP-NPs in a time-dependent manner (0, 3, 6 h) followed by staining nuclei and lysosomes with Hoechst 33342 (blue) and LysoTracker DND-26 (green), respectively.

Confocal laser scanning microscopy (CLSM) images clearly showed that internalization of red fluorescent PMA−RITC−PTX−CDDP-NPs was almost negligible in MCF7 cells at 0 h (Figure 4, topmost panel) having an undetectable red fluorescence signal. However, with time, PMA−RITC−PTX−CDDP-NPs internalized into MCF7 cells and localized into lysosomes in 3 and 6 h yielding merged yellow regions from LysoTracker green and red fluorescent nanoparticles observed in CLSM images (Figure 4, middle and lowermost panels). Hence, from these CLSM images it was confirmed that RITC-labeled PMA−PTX−CDDP-NPs were taken up by the breast cancer cells within 3 h and homed into acidic lysosomes. We further quantified the red fluorescence signals inside the cells at 0, 3, and 6 h using confocal microscopy. The quantification revealed that the red fluorescence intensity increased significantly at 3 h compared to 0 h (Figure S15). However, we found a negligible change in the subcellular red fluorescence intensity at 6 h compared to 3h. This quantification corroborated that PMA−RITC−PTX−CDDP-NPs internalized into MCF7 cells within 3 h.The acidic environment inside lysosomes would lead to release the active hydrophobic and hydrophilic drugs by cleavage of the acid labile ester and Pt−O coordination chemical linkages in PTX and CDDP, respec- tively.38,39 To evaluate the release of active drugs, the nanoparticles were incubated into pH = 5.5 buffer (lysosome mimic) and dual drug release was quantified by UV−vis spectroscopy in different time points at characteristic λmax = 273 and 706 nm for PTX and CDDP, respectively, from the absorbance versus concentration calibration graph. It was observed that 79.9 ± 4.1% and 54.9 ± 6.5% of PTX and CDDP were released slowly from the nanoparticles after 72 h, respectively (Figure 5a). Ideally, the nanoparticle should not release its payload under physiological conditions before reaching the targeted tumor tissues. To evaluate the dual drug release under physiological conditions, we incubated PMA−PTX−CDDP-NPs into phosphate buffer saline (PBS,pH = 7.4) and quantified the release of PTX and CDDP by UV−vis spectroscopy.

It was observed that only 29.5 ± 2.4% and 50.0 ± 6% of CDDP and PTX were released even after 72 h at pH = 7.4 (Figure 5b). From these release studies, it was evident that PMA−PTX−CDDP-NPs released payload in the acidic environment in much improved quantities compared to physiological conditions in a slow and controlled manner over 3 days, which would be ideal for successful delivery of PTX and CDDP into tumor tissues for augmented therapeutic outcome.Targeting Nucleus and Microtubules. 2.4.1. Nuclear Fragmentation. Acidic environment-mediated cleavage of PTX and CDDP from nanoparticles would target subcellularmicrotubules and DNA residing in nuclei, respectively.40−43 We evaluated the ability of PMA−PTX−CDDP-NPs to damage subcellular nuclei. MCF-7 cells were treated with PMA−PTX−CDDP-NPs temporally (6, 24, and 48 h), followed by staining tubulin and nuclei with α-tubulin antibody(green) and Hoechst 33342 (blue), respectively. As control, MCF-7 cells were treated with free PTX and CDDP cocktail having the same ratio in nanoparticles. CLSM images in Figure6 clearly revealed that PMA−PTX−CDDP-NPs induced nuclear fragmentation in a time-dependent manner (Figure6). In comparison, the free drug cocktail also fragmented the nuclei of MCF7 cells in a manner very similar to the nanoparticle treatment (Figure S16). We further quantified the fragmented cellular nuclei induced by the free drug cocktail or nanoparticles using confocal microscopy in differentincubation times. It was observed that PMA−PTX−CDDP- NPs induced a similar nuclear damage (20.4 ± 8.8% and 43.4 ± 2.8%, respectively) compared to free drug cocktail treatments(16.7 ± 5.4% and 34.36 ± 5.8%, respectively) (Figure S17) at24 and 48 h postincubation.Stalled Cell Division by the Microtubule Damage. Moreover, PTX binds with microtubules to stabilize them leading to the inhibition of cell division in the mitosis stage.44,45 To evaluate the effect of PMA−PTX−CDDP-NPs on cell division, MCF-7 cells were treated with nanoparticles at 6 and 24 h.

Nucleus and tubulin were stained with Hoechst 33342 (blue) and α-tubulin antibody (green), respectively, followed by visualization through fluorescence confocal microscopy.CLSM images in Figure 7 showed the characteristic damaged microtubule and stalled cell division in the mitosis stage leading to the accumulation of genomic materials in the nucleus after treatment with the nanoparticles at both 6 and 24 h. We have observed a similar microtubule damage and accumulation of genomic materials in the central part of cells in free PTX and CDDP cocktail treatment (Figure S18). We further quantified the number of cells with stalled cell division using confocal microscopy. It was revealed that 24 h of nanoparticle treatmentinduced stalled division in 22% of the cells whereas only 7% of cells that were treated with free PTX and CDDP cocktail exhibited stalled division (Figure S19).The remarkable increase in the stalled cell division upon nanoparticle treatment compared to free drug combination can be attributed to the simultaneous improved cellular internalization of PTX and CDDP through nanoparticles. By contrast, free PTX and CDDP have vastly different aqueous solubility leading to the erratic cellular internalization in right dosages to interact with their respective subcellular targets.Reduction of 3D-Breast Cancer Spheroids. Nano- particle-mediated fragmentation of nucleus and inhibition of cell division lead to cellular death.

To assess the effect of nanoparticles on the cancer cell death, MCF-7 cells were incubated with PMA−PTX−CDDP-NPs in a dose-dependent manner for 24 h, and cell viability was measured by the MTT assay. As control, MCF7 cells were treated with free PTX and CDDP combination. Interestingly, PMA−PTX−CDDP-NPs induced cell death with IC50 = 0.29 μM (Figure S20a). By contrast, free drug combination showed much higher IC50 =4.06 μM compared to nanoparticle treatment. For successful translation of the nanoplatforms having multiple drugs as payload, the vector should not show any toxicity profile itself. To investigate the toxicity profile of our polymer vector PMAn, we treated MCF7 cells with PMAn in a dose-dependent manner for 24 h and evaluated the cell viability with the MTT assay. Interestingly, PMAn showed negligible cytotoxicity in MCF-7 cells even at 10 μM concentration after 24 h (Figure S20b). This cell viability assay clearly indicated that PMAn has potential for further translation to clinics.Finally, we evaluated the effect of the nanoparticles in 3- dimensional cultures as in vivo mimic. MCF7 cells were grown over Matrigel to develop 3-dimensional spheroids over 8 days. 3D-MCF7 spheroids were treated with PMA−PTX−CDDP- NPs for 24 h, and the spheroids were allowed to grow for 16 more days. Finally, the nuclei and actin in 3D-MCF7 spheroids were stained with Hoechst 33342 (blue) and Alexa Fluor- labeled phalloidin 568 (red), respectively. The spheroids were visualized by CLSM. Figure 8 evidently showed that PMA−PTX−CDDP-NPs reduced the size of the MCF7 spheroidssignificantly compared to non-nanoparticle-treated MCF7 spheroids. We further quantified the 3D-MCF7 breast tumorspheroids by measuring the surface area and volume of the acini. It was observed that PMA−PTX−CDDP-NPs remark- ably reduced the surface area and volume of 3D-MCF7- spheroids (Figure S21). The reduction in the size of the acini can be attributed to either cell death or inhibition of cell division. Thus, taking into consideration the results of the cytotoxicity assay as well as the 3D spheroid assay, it can be concluded that PMA−PTX−CDDP-NPs killed the MCF7 breast cancer cells extraordinarily with no significant toxicity for the starting polymeric vector used for dual drug conjugation.

3.CONCLUSIONS
In conclusion, this present work demonstrates the unique CDDP-induced self-assembly of linear polymers into spherical nanoparticles which can encompass hydrophobic and hydro- philic drugs simultaneously. The essential mechanism for self- assembly was determined by MD simulation and found to be the interaction between hydrophilic moieties as well as interpolymer chain H-bonding. These hydrophobic−hydro- philic drug-loaded polymeric nanoparticles were taken up by the breast cancer cells into lysosomes, leading to nuclear fragmentation and stalled cell division in mitosis by inhibiting microtubule formation. The nanoparticles demonstrated a remarkable cell death in vitro as well as a 3D-tumor spheroid model. We foresee that our new approach of polymeric nanoparticles has an immense potential for future translation into clinics for combination therapy in breast cancer.

4.EXPERIMENTAL SECTION
PMAn, CDDP, anhydrous DMF, silver nitrate, o-phenylenediamine, N-(3-dimethylaminopropyl)-N′- ethylcarbodiimide hydrochloride (EDC), N,N-dimethyl amino pyridine (DMAP), pyrene, ethylenediamine, rhodamine B isothiocyanate, dimethyl sulfoxide (DMSO-d6), methanol-d4, and silicon wafer for FESEM were bought from Sigma-Aldrich. PTX was purchased from Selleck Chemical. Dialysis mem- branes (3.5 kDa) were purchased from Spectrum Labs. MCF7 cells were procured from ECACC. Dulbecco’s modified Eagle’s medium (DMEM), fetal bovine serum (FBS), LysoTracker Green DND-26, Hoechst 33342, SlowFade Gold antifade, and
Alexa Fluor-conjugated phalloidin 568 were purchased from Invitrogen. MTT reagent and tissue culture grade DMSO were purchased from Sigma-Aldrich. 96-well flat-bottomed tissue- culture plates were obtained from Corning. Synthesis of PMAc 2. PMAn (20 mg) was dissolved in a 5 mL DMF/water (1:1) mixture and stirred at 60 °C until the turbid solution turns into a clear solution. This solution was cooled to room temperature and subjected to dialysis [molecular weight cut-off (MWCO): 3.5 kDa] against water for 24 h. The dialyzed solution was lyophilized to get dry PMAc. Polymer 2 (10 mg, 0.00149 mmol) was dissolved into 1 mL of dry DMF in a round-bottom flask under inert atmosphere. EDC (4.4 mg, 0.022 mmol) and DMAP (0.9 mg, 0.00745 mmol) were added into polymer 2 followed by stirring at room temperature for 10 min. PTX (6.36 mg, 0.00746 mmol) Paclitaxel was added into activated polymer 2, and the reaction mixture was stirred for 48 h. The reaction was quenched by adding 0.1 N HCl, and PMA−PTX was dialyzed (MWCO = 3.5 kDa) against water for 48 h to remove organic solvents and reagents used. The pure PMA− PTX conjugate was lyophilized.