Drug delivery with a pH-sensitive star-like dextran-graft polyacrylamide copolymer

The development of precision cancer medicine relies on novel formulation strategies for targeted drug delivery to increase the therapeutic outcome. Biocompatible polymer nanoparticles, namely dextran-graft-polyacrylamide (D-g-PAA) copolymers, represent one of the innovative non-invasive approaches for drug delivery applications in cancer therapy. In this study, the star-like D-g-PAA copolymer in anionic form (D-g-PAAan) was developed for pH-triggered targeted drug delivery of the common chemotherapeutic drugs – doxorubicin (Dox) and cisplatin (Cis). The initial D-g-PAA copolymer was synthesized by the radical graft polymerization method, and then alkaline-hydrolyzed to get this polymer in anionic form for further use for drug encapsulation. The acidification of the buffer promoted the release of loaded drugs. D-g-PAAan nanoparticles increased the toxic potential of the drugs against human and mouse lung carcinoma cells (A549 and LLC), but not against normal human lung cells (HEL299). The drug-loaded D-g-PAAan-nanoparticles promoted further oxidative stress and apoptosis induction in LLC cells. D-g-PAAan-nanoparticles improved Dox accumulation and drugs’ toxicity in a 3D LLC multi-cellular spheroid model. The data obtained indicate that the strategy of chemotherapeutic drug encapsulation within the branched D-g-PAAan nanoparticle allows not only to realize pH-triggered drug release but also to potentiate its cytotoxic, prooxidant and proapoptotic effects against lung carcinoma cells.


Introduction
Lung cancer is a leading cause of cancer-related death, responsible for 2.2 million new cases and 1.8 million deaths in 2020 worldwide. 1 The risk of lung cancer, strongly associated with smoking and air pollution, is anticipated to continue growing. Nowadays, traditional antitumor chemotherapy is used for treatment of most cancers, based on the application of small toxic chemotherapeutic molecules that interact with DNA molecules, modify them and induce cell death in cancer tissues. 2,3 However, conventional chemotherapeutics suffer from a number of limitations, including poor solubility, stability and short blood circulation that result in low drug bioavailability and selectivity. Moreover chemotherapeutic drugs possess high toxicity and damage not only cancer, but also healthy cells, producing diverse side effects. 4,5 The use of modern nanobiotechnology could potentially solve this problem and be more efficient in comparison with traditional anticancer chemotherapeutic drugs. Nanobiotechnology can provide targeted delivery of the drug to cancer tissue and reduce its systemic toxicity. Also it gives a possibility to control the drug concentration, prolong blood circulation and release it targetly on the site of interest. [5][6][7][8][9] Polymeric nanocarriers offer remarkable progress in the anticancer treatment of disease in the near future. [10][11][12][13] A polymer nanocarrier loaded with anticancer drugs overcomes many disadvantages associated with conventional drug delivery. A large surface to volume ratio, encapsulation of hydrophobic molecules, optimization and control of drug dosage, slow and sustained release, targeting of the specic pathological tissues provide major advantages for such systems. Thus, polymers can be classied as smart drug delivery systems which can provide superior biocompatibility and biodegradability and easy preparation. [14][15][16] Moreover, the polymer nanocarrier can be pHor thermosensitive, that expands the use of such nanosystems for solving the specic task of targeted drug delivery in cancer treatment. [16][17][18] Recently a star-shaped dextran-gra polyacrylamide (D-g-PAA) copolymer has been exploited as a potential nanocarrier for photodynamic therapy 19 and chemotherapy. 20 It was proved that the branched star-like copolymer was efficient for the encapsulation of anticancer drugs. The copolymer was absorbed by macrophages (murine macrophage cell line) and was not cytotoxic. 21 A polymer-cisplatin (Cis) nanosystem yielded a dosedependent decrease in the viability of K-562 (human chronic myelogenous leukemia) and U-937 (human histiocytic lymphoma) cell lines at different concentrations from 0.1 to 10 mg mL −1 . Thus, the polymer-Cis nanosystem decreased cell viability to about 22% and 39% at 5 mg mL −1 , respectively. 21 Branched copolymers have been proven to be capable of stabilizing potentially unstable nanosystems. 22 The number of variable parameters affecting the macromolecule structure of branched polyelectrolytes increases substantially, as compared with their nonionic analogs. 23 Conformational changes leading to changes in the size and compactness of macromolecules are additionally controlled by the pH and the ionic strength of the environment, thus expanding the range of their usage as smart materials for pH-responsive targeted drug delivery. A pHtriggered delivery strategy targets the acidic intracellular organelles and extracellular microenvironment of solid tumours. 14,24,25 Thus, pH-responsive polymer nanoparticles can offer a powerful strategy to design smart therapeutic delivery systems.
Here we discuss an application of the anionic analogue of the D-g-PAA copolymer, namely D-g-PAAan as a targeted nanocarrier for anticancer drugs doxorubicin (Dox) and Cis, as well as compare their cytotoxic activity in vitro towards human and mouse lung carcinoma cells (A549 and LLC) and normal human lung cells (HEL299). For this, rst, the size distribution and pHdependent drug release were assessed for the synthesized D-g-PAAan-Dox and D-g-PAAan-Cis followed by cell-based evaluation of its effects on cell viability and morphology, reactive oxygen species (ROS) generation levels and caspase 3/7 activation in the lung cells of different origins.

Dynamic light scattering (DLS) study
The DLS data on the hydrodynamic diameter distribution of the D-g-PAAan nanoparticle aqueous systems quantied their size to be 120 nm with a polydispersity index (PDI) of 0.185 (Fig. 1). Once the nanoparticles encapsulated the drugs a shrinkage was observed in dH 2 O. The size was found to be 80 nm for both Doxand Cis-loaded polymeric nanoparticles (Fig. 1). This was not surprising, because the functional groups of the polymer were partially bonded with molecules of Dox or Cis caused the negative charge shielding in the graed chains resulting in shrinkage of the D-g-PAAan coil.
The PDI value for the Dox-and Cis-loaded D-g-PAAan nanoparticles was estimated at a level of 0.206 and 0.244, respectively, that evidenced the monodisperse size distribution.
The zeta potential is related to the stability of colloidal dispersions. The zeta potential value for the free D-g-PAAan, as well as D-g-PAAan-Dox and D-g-PAAan-Cis aqueous systems was −21.4 mV, −28.6 mV and −23.8 mV, respectively. A high negative surface charge of the individual nanoparticles (or, more strictly, the electrostatic repulsion between the negatively charged aggregates) indicates a very low tendency for them to aggregate over time in an aqueous solution (i.e., a high solute stabilization).
The size distribution and zeta potential of D-g-PAAan, D-g-PAAan-Dox and D-g-PAAan-Cis aqueous systems were stable for 6 months.
To estimate the stability in a cell culture experimental set-up, free D-g-PAAan, D-g-PAAan-Dox and D-g-PAAan-Cis were incubated at 37°C for 48 h in DMEM supplemented with 1% FBS that mimicked the cell-based assays. At 0, 24 and 48 h the hydrodynamic diameter distribution was measured with DLS ( Table 1). The nanoparticle's hydrodynamic diameter distribution in FBS-supplemented cell culture showed that its size was  Table 1 Hydrodynamic size (diameter, nm) of free and Dox-and Cisloaded D-g-PAAan in a 1% FBS-supplemented DMEM medium increased as compared with measurements in dH 2 O. Thus, the free D-g-PAAan nanoparticle size distribution was 130 AE 3 nm. The 10 nm-size increase of the free D-g-PAAan nanoparticle size distribution suggested a protein corona formation on the surface of the nanoparticle in the FBS-supplemented cell culture DMEM medium. The size distributions of D-g-PAAan-Dox and D-g-PAAan-Cis nanoparticles were estimated to be around 150 nm. It is known that the structural organization of nanoparticles, the surface of which is charged, is determined not only by the hydrophobic and van der Waals interactions, but also by the electrostatic interactions and therefore signicantly depends on the presence of electrolytes in the dissolution medium. 26 Both, D-g-PAAan in aqueous solution and bovine serum albumin (BSA), the major FBS component, have small negative zeta-potential values. However, multi-charged cations in the DMEM medium (Ca 2+ , Fe 3+ and Mg 2+ ) are able to internally cross-link PAAan chains by bonding with COO-groups and signicantly decrease the zeta potential of D-g-PAAan. As a result, neutrally charged D-g-PAAan can interact with BSA in a semi-diluted concentration regime and form aggregates. 27 All nanoparticles had no signicant changes in size distributions when measured immediately and during 48 h of incubation under in vitro cell culture experimental conditions ( Table  1).
The maximum detected stability during prolonged incubation indicated that there was no additional aggregation of the nanoparticles during the prolonged incubation in the FBSsupplemented cell culture medium which conrmed their suitability for in vitro studies.

Drug release
The drug release studies from the nanoparticle were conducted in triplicate using dialysis at different pH for 48 h. The concentration of the released drugs was assessed with a HPLC-ESI-MS/MS. The selected pH values were 7.4 and 5.0 that mimicked blood pH and cancer cell pH, which is known to be more acidic. Both drug molecules were released from the nanoparticles at an acidic pH more rapidly than at physiological pH. It can be explained by the fact that at low pH the polyelectrolyte nanocarrier changed its hydrophilic-hydrophobic balance and became more hydrophobic. The COO-groups of the graed hydrolyzed polyacrylamide chains transform to -COOH ones at pH 5.0. Therefore this leads to the release of encapsulated hydrophilic drugs. The Cis molecule is more hydrophilic in comparison with Dox (Fig. 2).
For this reason at pH 5.0 the cumulative Dox release from Dg-PAAan nanoparticles reached 91.9 AE 9.2% at 10 h, whereas at pH 7.4 it was slower reaching the levels of 35.6 AE 5.5 and 61.6 AE 7.2% at 10 and 48 h, respectively (Fig. 2b). Cis was released from D-g-PAAan nanoparticles even faster at pH 5.0, thus reaching the level of 52.6 AE 2.2% at 1 h as compared to the level of 9.0 AE 1.0% of released Dox at pH 7.4. At 48 h Cis release was estimated to be 90.9 AE 7.2 and 46.7 AE 7.3% at pH 5.0 and 7.4, respectively ( Fig. 2d).

Cell viability
The viability of lung carcinoma cells incubated in the presence of increasing concentrations of the investigated drugs or drugpolymer nanoparticles in the drug-equivalent concentrations was estimated with the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) test aer 24 and 48 h of treatment (Fig. 3). The viability of cells incubated without any treatment was taken as 100% (control). No effect of D-g-PAAan nanoparticles on lung cell viability was detected (data are not shown), while the concentration-and time-dependent toxic effects of the free drugs were observed. All studied lung cell lines exhibited  A marked effect of drug encapsulation within D-g-PAAan nanoparticles was revealed on LLC and A549 cell viability. LLC cells were found to be more sensitive to D-g-PAAan-Dox that at a 1.85 mM Dox concentration the viability decreased to 31% at 24 h as compared with the free drug ( Fig. 3c), whereas A549 cell viability was decreased to 34% at 48 h under the action of D-g-PAAan-Dox at a 3.70 mM Dox concentration as compared with that of the free drug (Fig. 4e). The D-g-PAAan nanoparticles had a prolonged effect on the toxicity of Cis against human and murine LLC (Fig. 3c-f). Thus, D-g-PAAan-Cis at 48 h decreased the viability of LLC and A549 cells on 26 and 21% as compared with the free Cis effect correspondingly at a 20 mM equivalent Cis concentration (Fig. 3d and f). Treatment of cells with drug-D-g-PAAan nanoparticles was followed by a signicant drop in IC 50 for the D-g-PAAan-loaded drugs ( Table 2). Visual changes in cell quantity and morphology were also observed with phasecontrast microscopy. As shown in Fig. 4a and b, complexation of Dox and Cis with D-g-PAAan nanoparticles caused a decrease in viable cells at 24 h. These data denote D-g-PAAan nanoparticles' ability to potentiate both Dox and Cis toxic effects against cancer cells.
In contrast, the normal human lung broblast HEL299 cells exhibited similar sensitivity to both free and D-g-PAAan-loaded Dox (Fig. 4a). HEL299 cell viability was decreased to 58 AE 4 and 57 AE 4% under treatment with 3.7 mM free and D-g-PAAanloaded Dox, respectively. The IC 50 value of Dox was not signicantly changed aer its complexation with D-g-PAAan nanoparticles (Table 3). For Cis, D-g-PAAan nanoparticles were found to even protect HEL299 cells against free Cis cytotoxicity (Fig. 4b). Free and D-g-PAAan-loaded Cis at 20 mM decreased the HEL299 cell viability to 23 AE 2 and 54 AE 5%, respectively. Thus, the IC 50 value for the D-g-PAAan-loaded Cis was increased by 2.6 times as compared with that of free Cis (Table 3).
It was shown that the studied D-g-PAAan nanoparticles increased toxic potential of the drugs against human and mouse lung carcinoma cells (Fig. 3c-f), but not against normal human lung cells (Fig. 4). The observed selective biological effects of anticancer drug toxicity upon their encapsulation in the D-g-PAAan nanoparticles suggest different modes of interaction between the developed nanoparticles and cancer cells in comparison to normal cells.

Intracellular ROS generation
Oxidative stress induction presents a promising anticancer strategy due to the high sensitivity of cancer cells to the ROS level increase 28,29 . Extranuclear effects of both Dox and Cis include ROS production induction 30,31 . The intracellular level of generated ROS in LLC cells aer 7, 20 and 27 h of exposure to free and D-g-PAAan-loaded Dox and Cis was estimated using an oxidative-sensitive uorescence dye 2,7-dichlorouorescin diacetate (DCFH-DA) both quantitively and qualitatively (Fig. 5). The ROS generation of the untreated LLC cells was set as 100%.
Increasing the concentrations of both free drugs administered to the cells provoked a time-and dose-dependent increase in intracellular ROS generation in LLC cells (Fig. 5) that denoted their prooxidant activities 32 . Thus, free 2 mM Dox and 20 mM Cis increased ROS levels in LLC cells to 156 AE 11 and 182 AE 13% at 27 h of incubation, correspondingly ( Fig. 5a and b). However, once Dox and Cis were loaded into D-g-PAAan particles, a further strong increase in DCF uorescence intensity revealed the oxidant stress escalation. Thus, 2 mM D-g-PAAan-Dox and 20 mM D-g-PAAan-Cis increased ROS levels in LLC cells to 192 AE 10    Fig. 5a  and b). In parallel the detected increase in the green uorescence intensity was observed with uorescence microscopy (Fig. 5c). Thus, the obtained data suggested that drug-loaded Dg-PAAan-nanoparticles promoted oxidative stress in LLC cells by increasing intracellular ROS generation.

Caspase 3/7 activity
ROS are increasingly recognized as important initiators and mediators of apoptosis 33 suggesting that the developed nanoparticles could nally activate the caspase cascade. Therefore, we determined whether the observed cell viability decrease and oxidative stress in LLC cells upon treatment with drug-loaded Dg-PAAan nanoparticles were associated with caspase-3/7 activation. Caspase 3/7 activity of untreated LLC cells was set as 100%.
The D-g-PAAan nanoparticles themselves had no effect on the caspase-3/7 activity, whereas the treatment with both free drugs resulted in a time-dependent increase of the enzyme activity up to 417 AE 12 and 151 AE 15% for 2 mM Dox and 20 mM Cis at 27 h (Fig. 6). Once the drugs were encapsulated in the D-g-PAAan nanoparticles, a further increase in the caspase-3/7 activity in LLC cells was found. LLC cells were characterized with caspase-3/7 activity at the levels of 496 AE 32 and 287 AE 14% of the control aer 27 h of treatment with D-g-PAAan nanoparticles loaded with 2 mM Dox and 20 mM Cis, respectively (Fig. 6). These nding suggested that both Dox-and Cis possessed higher proapoptotic potential when encapsulated in the D-g-PAAan nanoparticles.
Therapeutic effect on a 3D LLC cell spheroid model  Thus, the behavior of 3D-cultured cells is more reective of in vivo cellular responses and can more accurately produce information about pharmacological activity. 34,35 To investigate a potential correlation of the enhanced toxic effect of D-g-PAAan nanoparticles with a more effective intracellular drug accumulation, the cellular uptake of free Dox and D-g-PAAan-Dox was rst studied. Since Dox possesses strong absorption and uorescence in the visible spectral region, tracking of Dox-loaded nanoparticles is possible with noninvasive direct uorescent-based techniques. 3D LLC cell spheroids were incubated in the presence of 7.4 mM Dox or D-g-PAAan-Dox in a drug-equivalent concentration, examined with phase-contrast and uorescence microscopy. The uorescence microscopy images illustrate that D-g-PAAan-Dox nanoparticles were internalized faster than the free drug as evidenced by brighter intracellular uorescence (Fig. 7a). The observed improved accumulation of Dox suggested the effective accumulation of the D-g-PAAan nanoparticles in LLC spheroids that could be linked to the interplay of the surfaces of the nanoparticles and cells in the 3D environment at a relatively high Dox concentration. The functional groups of the polymer were partially bonded with Dox molecules that could result in charge shielding in the graed chains and shrinking the D-g-PAAan coil at lower pH. The observed accumulation increase of Dox upon D-g-PAAan nanoparticle delivery pointed to the perspective application of the developed nanoparticles for targeted drug delivery and will be investigated further in animal models in the near future.
To assess the possible therapeutic effect on the 3D LLC cell spheroid model cell viability was estimated aer 24 h of treatment with increasing doses of free and D-g-PAAan-Dox. Free Dg-PAAan had no effect on the viability of 3D LLC cell spheroids, while Dox treatment of cell spheroids resulted in dosedependent cell viability decrease (Fig. 7b). Thus, free 1.85, 3.70 and 7.40 mM Dox inhibited the viability of 3D LLC cell spheroids to 99.4 AE 5.0, 68.4 AE 1.6 and 33.6 AE 2.0%, respectively. Once Dox was loaded into D-g-PAAan its higher cytotoxicity was detected as compared with the free drug: 1.85, 3.70 and 7.40 mM D-g-PAAan-Dox inhibited the viability of 3D LLC cell spheroids to 71.7 AE 4.3, 56.1 AE 4.0 and 24.9 AE 1.2%, correspondingly (Fig. 7a). The encapsulation of Cis into D-g-PAAan nanoparticles had a similar effect. The cytotoxicity of Cis towards 3D LLC cell spheroids was increased as compared with that of the free drug at an equivalent concentration (Fig. 7b). The chemotherapeutic drug cytotoxicity modulating effect upon its delivery with D-g-PAAan nanoparticles was proved in a 3D multi-cellular spheroid model that suggested its further promising development for cancer treatment.

Chemicals
Dox and Cis (Sigma-Aldrich, Co, Ltd, USA) were dissolved in distilled water with a maximum concentration of 40 and 500 mg mL −1 , correspondingly.

Polymer nanocarrier synthesis
The D-g-PAA copolymer was synthesized by the radical gra polymerization method using a Ce(IV)/HNO 3 redox system. PAA was graed on certied dextran with a molecular weight of M w ¼ 7 Â 10 4 g mol −1 (produced by Serva). The synthesis and identication of the sample are described in detail in ref. 36 .
The branched anionic polyelectrolyte (D-g-PAAan) used as a carrier for Dox or Cis molecules was obtained by alkaline The scheme of synthesis of the D-g-PAAan copolymer in anionic form is shown in Fig. 8. Stock solution of the D-g-PAAan copolymer (1000 mg mL −1 ) was prepared in distilled water. The D-g-PAAan copolymer and Dox or Cis were mixed in a 1 : 1 volume ratio for obtaining water-soluble D-g-PAAan + Dox (500 + 40 mg mL −1 ) or D-g-PAAan + Cis (500 + 250 mg mL −1 ) nanoparticles, correspondingly. The stock concentrations of Dox and Cis in D-g-PAAan-Dox and D-g-PAAan-Cis nanoparticles were 74 and 830 mM, respectively.

Dynamic light scattering
Particle size distribution was evaluated with a Zetasizer Nano S (Malvern Instruments, UK) equipped with a He-Ne laser (633 nm). Data were recorded at room temperature in backscattering modus at a scattering angle of 2q ¼ 173°. Zeta potential measurements were used for ascertaining the electrokinetic potential of particles in aqueous solution. Measurements were conducted on a Zetasizer Nano-ZS90 (Malvern, Instruments, UK) at room temperature.
Drug release D-g-PAAan-Dox and D-g-PAAan-Cis nanoparticles (1000 mL) were added to a dialysis bag (MWCO 1 kDa, Spectra/Por® Float-A-Lyzer® G2 (Carl Roth GmbH + Co. KG, Karlsruhe, Germany). The dialysis system was suspended in a release volume of 100 mL buffer at 37°C and rotated at 200 rpm (1 : 100 dilution between donor and acceptor compartments). At scheduled intervals, 500 mL of the release medium was collected for the HPLC-ESI-MS/MS analysis.
The same volume of fresh buffer at the same temperature was added immediately to maintain a constant release volume. Acetate and PBS buffers were used for pH 5.0 and 7.4, respectively. The obtained data were normalized with the buffer control and expressed as a % of the respective control sample, analyzed at 0 h. The Dox concentration was assessed with the HPLC-ESI-MS/MS analysis described before, 37 whereas Cis quantication required the establishment of a new method, described below.
Elution and separation of Cis were performed using an Eclipse XDB-C18 column (4,6 Â 100 mm) with a 3 mm particle size under isocratic conditions with a mobile phase of methanol and a 0.1% formic acid water solution. The ow rate was set at 0.5 mL min −1 . The chromatographic reverse phase conditions and optimized MS/MS parameters are presented in Table 4. For identication and quantication, the Cis adduct ion with Na [M + Na] + was chosen (Fig. 9a).
HPLC-ESI-MS/MS analysis was performed in positive mode with the usage of multiple reactions monitoring (MRM) mode,   which provides the best sensitivity and accuracy of measurements. Aer MS/MS optimization, a unique MRM-transition that includes a precursor and two characteristic product ions was acquired and used for further identication and quantication. The ionized Cis adduct with Na ([M + Na] + , 332.9 m/z) was used as a precursor ion with the most abundant fragment ions of 305.9 and 23.05 m/z. Cis calibration standards from 0.1 to 50 mg mL −1 were prepared from a 250 mg mL −1 water stock solution. These standards were stored in the dark at 4°C. Quantication was achieved using the regression curve (Fig. 9b) according to the linear regression eqn (1): The obtained data were normalized with the buffer controls and expressed as a percentage of the respective control sample, analyzed at 0 h.

2D cell culture
LLC cells were kindly supplied by the bank of cell cultures and transplantable experimental tumors of the R.E. Kavetsky Institute of Experimental Pathology, Oncology and Radiobiology, NAS of Ukraine (Kyiv, Ukraine). Lung carcinoma human A549 and normal human lung broblast HEL299 cell lines were purchased from AddexBio Technologies (San Diego, CA, USA). Cells were maintained in a 5 mL DMEM (LLC and A549) or EMEM (HEL 299) medium supplemented with 10% FBS, 1% penicillin/streptomycin and 2 mM glutamine, using 25 cm 2 asks from Sarstedt (Nümbrecht, Germany) at 37°C with 5% CO 2 in a humidied incubator binder (Tuttlingen, Germany). Passaging was performed once cells reached z80%. Treatment with trypsin (1 : 10 in PBS) was used to detach adherent cells. The number of the viable cells was counted with the use of a Roche Cedex XS Analyzer (Basel, Switzerland) aer staining with 0.1% trypan blue.

2D cell viability
Cells (10 4 cells per well), cultured in 96-well cell culture plates by Sarstedt (Nümbrecht, Germany) for 24 h, were treated with a 1% FBS DMEM medium containing 0-7.4 mM Dox and D-g-PAAan-Dox, 0-40 mM Cis and D-g-PAAan-Cis in a drug equivalent concentration. Cells were observed with a Keyence BZ-9000 BIOREVO microscope (Osaka, Japan) in combination with the Keyence BZ-II viewer acquisition soware (Osaka, Japan). Cell viability was determined with an MTT reduction assay 38 at 24 and 48 h. Briey, cells were incubated for 2 h at 37°C in the presence of 0.5 mg mL −1 MTT. Diformazan crystals were dissolved in DMSO and determined at 570 nm with a microplate reader Tecan Innite M200 Pro (Männedorf, Switzerland).

Intracellular ROS generation
To determine ROS production DCFH-DA (Sigma-Aldrich Co., St-Louis, USA) was applied. A 5 mM stock solution of DCFH-DA was prepared in DMSO, stored at −20°C and diluted with PBS immediately before use. LLC cells were seeded into 96-well plates (10 4 cells per well) and incubated for 24 h. Then the medium was exchanged with the one containing free Dox or Dg-PAAan-Dox in 1 and 2 mM Dox-equivalent concentrations and cells were incubated under normal cell culture conditions for 7, 20 and 27 h. At given time points cells were washed once with PBS. DCFH-DA (5 mM) was added and the uorescence intensity (l ex ¼ 488 nm, l em ¼ 520 nm) was recorded at 45 min with the microplate reader Tecan Innite M200 Pro (Männedorf, Switzerland). At 55 min of incubation uorescence images of cells were obtained with the uorescence microscope Keyence BZ-9000 BIOREVO (Osaka, Japan), equipped with a green lter (l ex ¼ 472 nm, l em ¼ 520 nm).

Caspase 3/7 activity
LLC cells were seeded into 96-well plates (10 4 cells per well) and incubated for 24 h. The cells were treated with free Dox or D-g-PAAan-Dox in 1 and 2 mM Dox-equivalent concentrations for 7, 20 and 27 h. The activity of caspase 3/7 was determined using the Promega Caspase-Glo® 3/7 activity assay kit (Madison, USA) according to the manufacturer's instructions. Briey, plates were removed from the incubator and allowed to equilibrate to room temperature for 30 min. Aer treatment, an equal volume of the Caspase-Glo 3/7 reagent containing a luminogenic peptide substrate was added, followed by gentle mixing with a plate shaker at 300 rpm for 1 min. The plate was then incubated at room temperature for 2 h. The luminescence intensity of the products of the caspase 3/7 reaction was measured with the microplate reader Tecan Innite M200 Pro (Männedorf, Switzerland).

3D cell spheroid generation
The LLC cell spheroids were generated in Corning® ultra-low attachment U-bottomed (round) 96-well plates.
Briey, LLC cells were seeded in a serum-free culture DMEM medium (10 4 cells per well), shortly centrifuged and then placed in the incubator Binder (Tuttlingen, Germany) at 37°C, 5% CO 2 , and 95% humidity.

3D cell spheroid viability
Aer spheroid formation (at 24 h aer seeding), the agents under the study were added: D-g-PAAan, 0-7.4 mM Dox and D-g-PAAan-Dox, 0-80 mM Cis and D-g-PAAan-Cis in a drug equivalent concentration. At 24 h the 3D cell viability was estimated with Promega CellTiter-Glo® 3D Cell Viability Assay (Promega, Madison, USA). Briey, the plate was removed from the incubator and allowed to equilibrate to room temperature for 30 min. An equal volume of the CellTiter-Glo® 3D reagent containing a luminogenic peptide substrate was added, followed by gentle mixing with a plate shaker at 300 rpm for 5 min. The plate was then incubated at room temperature for 25 minutes to stabilize the luminescent signal. The luminescence intensity of the products was measured with the microplate reader Tecan Innite M200 Pro (Männedorf, Switzerland).

Fluorescence microscopy
Aer spheroid formation (at 24 h aer seeding), the agents under the study were added: D-g-PAAan, Dox and D-g-PAAan-Dox in a 7.4 mM Dox equivalent concentration. Then, 3D LLC cell spheroid were treated for 7 h and visualized with a uorescence microscope Keyence BZ-9000 BIOREVO (Osaka, Japan) equipped with a red (l ex ¼ 480 nm, l em ¼ 600 nm) lter and the respective acquisition soware Keyence BZ-II Viewer (Osaka, Japan). The overlayed images were processed with the Keyence BZ-II Analyzer Soware (Osaka, Japan).

Statistics
All experiments were carried out with a minimum of four replicates. Data analysis was performed using the GraphPad Prism 7 Soware (GraphPad Soware Inc., USA). Paired Student's t-test was performed. Differences with p-values <0.05 were considered to be signicant.
The half-maximal inhibitory concentration (IC 50 ) value was calculated with specialized soware GraphPad Prism 7 (GraphPad Soware Inc.). Individual concentration-effect curves were generated by tting the logarithm of the compound concentration versus the corresponding normalized cell viability using nonlinear regression.

Conclusions
The polyacrylamide polymer was synthesized by the radical gra polymerization method, graed on a certied dextran polymer and alkaline-hydrolyzed to get a star-like branched anionic dextran-gra-polyacrylamide (D-g-PAAan) copolymer nanoparticle. For this, the commonly used chemotherapeutic drugs Dox and Cis were encapsulated within D-g-PAAan nanoparticles, whose size was estimated to be 80 nm. The obtained Dox-or Cis-containing D-g-PAAan nanoparticles released drugs in response to a pH decrease to 5.0 that pointed towards the opportunity for pH-triggered drug delivery. D-g-PAAan nanoparticles increased the toxic potential of the drugs against human and mouse lung carcinoma cells (A549 and LLC), but not against normal human lung cells (HEL299). The drugloaded D-g-PAAan-nanoparticles promoted further oxidative stress and apoptosis induction in LLC cells by increasing intracellular ROS generation and activation of caspase 3/7. The observed selective cytotoxic effects of anticancer drug toxicity upon their encapsulation in the D-g-PAAan nanoparticles suggested differential interactions between the developed nanoparticles and the cancer or normal cells, respectively, that pointed to a promising approach for targeted cancer treatment.
The data obtained in the study indicate that the strategy of chemotherapeutic drug encapsulation within the D-g-PAAan nanoparticle allows not only to realize pH-triggered drug release but also to potentiate its cytotoxic, prooxidant and proapoptotic effects against lung carcinoma cells in vitro. As a pH decrease is observed in most solid tumors, the proposed drug-delivery polymer nanoparticle responsive to the slightly acidic extracellular pH environment of solid tumors provides a promising approach for cancer treatment.

Conflicts of interest
There are no conicts to declare.