Water-Soluble Blue Fluorescent Nonconjugated Polymer Dots from Hyaluronic Acid and Hydrophobic Amino Acids.

Fluorescent polymers have been increasingly investigated to improve their water solubility and biocompatibility to enhance their performance in drug delivery and theranostic applications. However, the environmentally friendly synthesis and dual functionality of such systems remain a challenge due to the complicated synthesis of conventional fluorescent materials. Herein, we generated a novel blue fluorescent polymer dot through chemical conjugation of hydrophobic amino acids to hyaluronic acid (HA) under one-pot green chemistry conditions. These nonconjugated fluorescent polymer dots (NCPDs) are water soluble, nontoxic to cells, have high fluorescence quantum yield, and can be used for in vitro bioimaging. HA-derived NCPDs exhibit excitation wavelength-dependent fluorescent properties. In addition, the NCPDs also show enhanced doxorubicin loading and delivery in naive and drug-resistant breast cancer cells in 2D and 3D tumor cellular systems. These results demonstrate the potential for successful synthetic scale-up and applications for HA-derived NCPDs.

Introduction

Developmepan class="Chemical">nt of functional fluorescent polymeric nanpan>omaterials pan> class="Chemical">has gained interest for a range of applications, including bioimaging, drug delivery, bioanalysis, and chemical sensing due to their improved biocompatibility and functionalization.[1,2] Conventional conjugated fluorescent nanomaterials often involve sophisticated and time-consuming synthetic steps. Once the materials are synthesized, complex purification techniques, poor water solubility, and use of harsh nonenvironmentally friendly organic solvents prevent pharmaceutical scale-up processes for commercial drug product development. These drawbacks have spurred researchers to explore alternative strategies to generate nontoxic fluorescent nanomaterials. Therefore, several nonconjugated polymer dots (NCPDs) have recently been developed by implementing techniques such as polymerization, cross-linking, hydrothermal synthesis, and physical loading methodologies. These methods resulted in the generation of water-soluble fluorescent nanomaterials.[3]

Npan>CPDs do not typically containpan> a conjugated fluorescenpan>t dye, but inpan>stead possess numerous subfluorophore chemical groups (e.g., C=pan> class="Chemical">N and N=O) with intrinsic weak fluorescent properties. Immobilization of these groups in a polymeric network results in fluorescence generation. These heteroatom-containing double bonds are traditionally not regarded as fluorescent chromophores. A fluorescent NCPD can be generated by a hydrothermal reaction between a large nonfluorescent polymer and a small nonfluorescent electron-rich molecule. This mechanism for fluorescence generation is referred to as cross-linked-enhanced emission (CEE),[3] which is different from aggregation-induced emission.[4,5] CEE functions when a small, amine-rich organic molecule is reacted with a large polymer under controlled hydrothermal conditions and leads to the formation of NCPDs. The small molecule attachment leads to a decrease in its bond vibrations and rotations, and architectural confinement of electron-rich moieties (delocalized n and π electrons), ultimately leading to fluorescence emission.[6]

Recently, several groups papan class="Chemical">n class="Chemical">have reported the use of hydrothermal treatmenpan>t of natural materials, such as chitosanpan>,[7] natural materials modified with pan> class="Chemical">polyethylene imine,[8−10] and silk,[11] among others[12,13] to prepare water-soluble fluorescent NCPDs. During this process, the amine-rich precursor and large saccharide molecules form NCPDs due to condensation reactions that yield fluorescent aggregated moieties.

Previously, our research group hapan>s reported the use of hyaluronic acid (pan> class="Chemical">HA)-based systems for near infrared fluorescent tumor imaging, biosensing, and drug delivery.[14−18] Here, we further the applications of HA by conjugating it to hydrophobic amino acids, which drives self-assembly of HA conjugates into blue-emitting NCPDs, as schematically depicted in Figure A,B. Water-soluble NCPDs were formed under mild acidic conditions without the need for any organic solvents. Furthermore, the fluorescence origin of HA-derived NCPDs has been discussed. We have created a new class of blue fluorescent nanomaterials from HA using green chemistry methods, and have demonstrated their potential applications for in vitro bioimaging and drug delivery of doxorubicin (DOX) in 2D and 3D tumor models.

Figure 1

(A) Synthetic scheme for generation of fluorescent NCPDs from HA and precursor amino acids under mild acidic conditions. (B) Schematic for self-assembly of HA NCPDs. (C) DLS size distribution analysis of each NCPD. (D) Representative TEM images of the NCPDs. The scale bar represents 100 nm.

(A) Synthetic scheme for gepan class="Chemical">neration of fluorescent NCPDs from pan> class="Chemical">HA and precursor amino acids under mild acidic conditions. (B) Schematic for self-assembly of HA NCPDs. (C) DLS size distribution analysis of each NCPD. (D) Representative TEM images of the NCPDs. The scale bar represents 100 nm.

Materials and Methods

Sopan class="Chemical">dium hyaluronate, npan> class="CellLine">Mn = 10 kDa, was purchased from Lifecore Biomedical (Chaska, MN). All water was purified with a Barnsteadt Nanopuret Diamond system (Thermo Scientific; Waltham, MN). 96-well tissue culture plates (Falcon), 12-well tissue culture plates, Desalting PD10 columns and dialysis tubing (MWCO = 3500), Nunc Glass Bottom Dishes (12 mm), and Nunc Lab-Tek II Chamber Slide System were purchased from Fisher Scientific (Pittsburgh, PA). DOX was purchased from Ark Pharma (Catalog #AK-72874), and chlorpromazine hydrochloride (Catalog # 69-09-0), methyl-β-cyclodextrin (Catalog #332615), quinine sulfate (Catalog #1597005), and bovine serum albumin (BSA) from Sigma-Aldrich (St. Louis, MO). L-Tryptophan (Catalog # A10230), l-tryptophan benzyl ester (Catalog #H63385), and l-proline benzyl ester (Catalog #L15618) were purchased from Alfa Aesar. Ethanol was purchased from the Warner-Graham Company (Cockeysville, MD). The anti-CD44 antibody and PE mouse IgG2b K Isotype Control were purchased from BD Pharmigen BD Biosciences. NMR was performed on a 500 MHz Bruker (1H) and a 600 MHz Varian (13C) system using a 5 mm probe at room temperature (rt). Deuterated water (D2O, 99.9% D) was purchased from Cambridge Isotope Laboratories. FTIR measurements were performed on a Nicolet IR200 FT-IR instrument using a single-reflection ZnSe ATR crystal. Penicillin/streptomycin (100× solution) was purchased from Corning. Human breast cancer cell lines (MDA-MB-231 and MCF10A) were obtained from American Type Culture Collection (Manassas, VA) and were grown in RPMI-1640 (HyClone, GE Healthcare Life Sciences) with 10% fetal bovine serum and 1% penicillin/streptomycin (P/S). Murine breast carcinoma CI66 and CI66-DOX-resistant cell lines were obtained from Dr. Rakesh Singh at the University of Nebraska Medical Center. All cells were incubated at 37 °C in a humidified incubator with 5% CO2. Matrigel Basement Membrane Matrix, LDEV-free (Catalog #356234), was purchased from Corning.

Preparation of the Hyaluronic Acid-Derived NCPDs (HA-NCPDs)

pan class="Chemical">Disease">HA-NCPDs were synthesized by dissolvinpan>g HA (25 mg) in 25 mL ultrapure water and stirring continuously for 30 min at rt. Next, either 30 wt % tryptophan, tryptophan benzyl ester, or proline benzyl ester was added and stirred for another 30 min at rt. The pH of the reaction mixture was maintained between 6.0 and 6.5 using dilute HCl. Once the reaction mixture became clear, the mixture was placed in an oil bath maintained at 60–80 °C with constant stirring for 5 h. Subsequently, the reaction mixture was allowed to cool overnight to rt. The reaction contents were then dialyzed against ultrapure water with a total of eight changes over 48 h to remove all unreacted reactants. The dialyzed product was collected and lyophilized to yield a white fluffy product (78% yield for HA–tryptophan, 81% HA–tryptophan benzyl ester, and 73% HA–proline benzyl ester NCPDs), and stored for further analysis at −20 °C.

Preparation of DOX-Loaded HA-NCPDs (DOX-HA-NCPDs)

pan class="Chemical">Disease">HA-NCPDs (20 mg) of each type were dissolved inpan> ultrapure npan> class="Chemical">water and stirred for 30 min to allow complete dissolution. DOX was dissolved in ultrapure water (100 μg/mL) and added dropwise to the aqueous solution containing the HA conjugate to 20 wt %, and was stirred for 24 h protected from light. The reaction mixture was dialyzed against ultrapure water with a total of 8 changes over 48 h to remove free DOX. After dialysis, residual DOX was removed via a PD10 column with ultrapure water as the mobile phase. The DOX-loaded fraction was collected and was lyophilized to obtain a light red fluffy product (HA–tryptophan DOX = 42% yield, HA–tryptophan benzyl ester DOX = 51%, and HA–proline benzyl ester DOX = 49%). The drug-loading contents were quantified using UV–vis absorption spectroscopy (ThermoFisher Evolution 220) by generating a calibration curve of DOX concentrations (0.001–0.5 mg/mL). The loading capacity was calculated as described by Zhang et al.[19]

Physicochemical Characterization of HA-NCPDs and DOX-HA-NCPDs

The morphology of the Npan>CPDs and pan> class="Chemical">DOX-HA-NCPDs was visualized by TEM using an FEI Tecnai G2 Spirit microscope and 2% aqueous methylamine vanadate at pH 8. Unloaded and loaded HA-NCPDs at 1 mg/mL concentration in ultrapure water were placed on formavar/silicone monoxide-coated 200 mesh copper grids using NanoVan negative stain for 33–34 s and imaged using a transmission electron microscope. The chemical structures of all HA-NCPDs were analyzed by 1H- and 13C NMR spectroscopy (Bruker Avance-III HD, 500 and 600 MHz, respectively). The IR spectra were recorded on the Nicolet IR200 FT-IR instrument using a single-reflection ZnSe ATR crystal. The X-ray photoelectron spectra (XPS) were obtained on a Kartos Axis Ultra Imaging spectrometer. The spectra of C(1s) (275–295 eV binding energy), O(1s) (525–545 eV binding energy), and N(1s) (380–420 eV binding energy) as well as survey scans (0–1100 eV) were recorded with a tilt angle of 45°. The atomic compositions were corrected for atomic sensitivities and measured from high-resolution scans. XPS and FTIR measurements were acquired in the dry state. Unloaded and loaded NCPD samples were prepared for optical characterization by dissolving samples in ultrapure water (1 mg/mL) and filtered through a 0.45 μm syringe filter. Absorption and fluorescence spectra were recorded on a ThermoFisher Evolution 220 UV–vis spectrophotometer and a Horiba Jobin Yvon FluoroMax 4 spectrofluorometer, respectively, in 10 mm quartz cells with a slit width of 5 nm. Colloidal properties were studied using a Malvern Zetasizer Nano ZS90 dynamic light scattering (DLS) instrument. The quantum yield of the fluorescent HA-NCPDs was obtained by comparing fluorescence emission of the reference dye (quinine sulfate in 0.5 M H2SO4, Φ = 0.54).

Determination of in Vitro DOX Release

Three freshly prepared papan class="Chemical">n class="Chemical">DOX-pan> class="Disease">HA-NCPDs and free DOX were assayed for DOX release using a dialysis bag method in phosphate-buffered saline (PBS; 10 mM phosphate, 140 mM NaCl) at pH 7.4 and 4.5, for 72 h. DOX-HA-NCPDs (4 mL) with a constant amount of 50 μg of DOX were transferred to a dialysis bag (MWCO = 3500 Da) and dialyzed against 10 mL PBS in a 15 mL conical tube, sealed, and placed on a rocker. At specific time intervals, 600 μL of the release media was removed and was replenished with fresh PBS. Each sample was subjected to UV–vis spectroscopy to quantify the amount of DOX that had been released based on an established calibration curve. All triplicate experiments were performed independently.

Stability Assessment of HA-NCPDs

The fluorescence ipan class="Chemical">ntensity of the Disease">HA-NCPDs was evaluated inpan> a ranpan>ge of pH values (1–14) by measurinpan>g the emission inpan>tensity at λmax for each pan> class="Disease">HA-NCPD. Photobleaching experiments were performed using a 365 nm UV lamp, a 4 mW/cm2 radiation source, by measuring HA-NCPD emission intensity over 10 h. Salt stability was assessed by dissolving each HA-NCPD in 0–2 M NaCl buffer solution and measuring HA-NCPD fluorescence emission at λmax. To identify any visually observable stability outcomes, HA–tryptophan NCPDs were dissolved in water at 20 mg/mL and placed on the laboratory counter-top for a 10-week period.

Cell Viability

The cell viability of unloapan class="Chemical">ded Disease">HA-NCPDs anpan>d pan> class="Chemical">DOX-loaded HA-NCPDs was evaluated using the CCK-8 assay (Dojindo). Breast cancer cell lines, MDA-MB-231, MCF10A, CI66, and CI66-DOX-resistant, were seeded (5000 cells/well) in a 96-well plate with serum-containing media and allowed to adhere to the bottom of the well overnight. Next, cells were incubated with different concentrations of HA-NCPDs and DOX-HA-NCPDs for either 24 or 72 h. After incubation, the cells were exposed to the CCK-8 reagent (1:10 dilution) and incubated for 1–4 h at 37 °C. The absorbance at 450 nm was read in each well using a Synergy HTX multimode plate reader (BioTek). The absorbance readouts obtained were directly proportional to the relative number of metabolically live/viable cells. Six replicates were used for each concentration, and individual experiments were repeated in triplicate. Relative cell survival was expressed as absorbance of treated cells relative to untreated cells.

Apoptosis Assay

To evaluate the therapeutic potential of papan class="Chemical">n class="Chemical">DOX-loaded pan> class="Disease">HA-NCPDs, CI66 and CI66-DOX-resistant cells were seeded (3 × 105 cells/well) in 12-well plates and allowed to adhere overnight. Cells were then treated for 12 h at 37 °C with empty and DOX-loaded HA-NCPDs. After incubation with the NCPDs or free DOX, cells were washed with 1× PBS thrice, trypsinized, and centrifuged at 3500 rpm for 5 min. The pellet was resuspended in 100 μL Annexin-V-FITC binding buffer. Subsequently, Annexin-FITC (5 μL) and propidium iodide (5 μL) were added in FACS tubes, vortexed for 5 min, and placed for 30 min at rt in the dark as per manufacturer’s recommendations. Stained cells were analyzed using a BD LSRII flow cytometer in the UNMC Flow Cytometry Research Facility. A total of 15,000 gated events were acquired per sample, and the mean fluorescence intensity was plotted in a histogram-based graphical representation. Each data point is representative of the mean of three independent measurements on the flow cytometer. Data were analyzed with FlowJo 10.7.1 (Tree Star) software.

Confocal Microscopy

CI66 anpan class="Chemical">d CI66-DOX-resistanpan>tcells were seeded (1 × 105 cells/well) on inpan>dividual 12 mm Nunc Glass Bottom Dishes (Invitrogen, ThermoFisher Scientific, USA) and allowed to adhere for 24 h. The cells were then treated with unloaded and DOX-loaded HA-NCPDs in serum-free DMEM media and incubated at 37 °C for 1 h. After incubation, cells were washed thrice with 1× PBS, and then fixed with 4% paraformaldehyde solution for 15 min at 37 °C. Next, cells were treated with 0.25% Triton-X-100 in 1× PBS to permeabilize the cell membrane, followed by blocking with 1% BSA in 1× PBS. Next, the Rab5 endosomal marker primary antibody (Rabbit polyclonal IgG, Santa Cruz Biotechnology Inc., Dallas, TX) was added in a 1% BSA 1× PBS solution and incubated with the cells overnight at 4 °C. Cells in each well were incubated with the secondary antibody (FITC-conjugated antirabbit Millipore AP132F) in 1% BSA at rt for 1 h. Cells were then washed thrice with 1× PBS, and stained with HCS NuclearMask Deep Red Stain (ThermoFisher Scientific, Catalog# H10294) for 15 min. Cells were washed and stored at 4 °C until confocal laser scanning microscopy (CSLM) imaging. Confocal images were collected using a Carl Zeiss 800 confocal laser scanning microscope, at 63× magnification with 1 μm cell slices. DAPI, Alexa Flour 488, and Alexa Fluor 647 filters were used to detect blue, green, and NIR signals from the cells.

Development of the CI66 and CI66-DOX-Resistant 3D Tumor Spheroid Model for NCPD Uptake Analysis

Matrpan>igel was thawed at 4 °C and added to a 8-well Nunc Lab-Tek II Chamber Slide System (50 μL/well) to cover the bottom of each well to serve as the ECM matrix containing laminin, collagen, heparin sulfate proteoglycans, entactin, and several soluble factors. Next, CI66 and CI66-DOX-resistant cells (4000/well) were mixed with thawed matrigel along with media and added to matrigel-coated wells. The growth of the spheroids was monitored using an Olympus CKX41 Bright Field, Infinity 1 Luminera camera over a period of 2 weeks until the spheroids reached 500 μm in size. Next, free (100 μg/mL concentration) and drug-loaded HA-NCPDs (DOX equivalent to 10 μg/mL) were added to the tumor spheroids and incubated for 12 h. After washing thrice with 1× PBS and fixing in 4% paraformaldehyde, the empty HA-NCPD (blue fluorescence) and DOX-HA-NCPD (DOX fluorescence) emission in spheroids was measured with a confocal laser scanning microscope using a 20× objective lens with 1.4 N.A., λex/λem = 488 nm/520 nm, and an XYZ-stack with 10 μm intervals at 512 × 512 pixels with imaging acquisitions at 33 Hz. Images captured were analyzed and processed using the Carl Zeiss LSM software (Jena, Germany) v.6.0. The spheroids were then trypsinized to obtain a single-cell suspension, washed with 1× PBS, and subjected to FACS analysis for quantitative uptake analysis of the HA-NCPDs and the free drug DOX.

Endocytosis Evaluation

The effects of several membrane epan class="Chemical">ndocytosis mechanisms were inpan>vestigated onclass="Chemical">pan> HA-derived NCPDs by incubating CI66 cells in a 12-well plate at a density of (105 cells/well) and allowed to adhere overnight to the bottom of each well. After 24 h, various cellular uptake inhibitors/ligands, including 150 mM ethyl isopropyl amiloride [EIPA] (an inhibitor of micropinocytosis), 1.8 mM methyl-β-cyclodextrin (an inhibitor of caveolae-independent endocytosis), 10 μg/mL chlorpromazine (an inhibitor of clathrin-mediated endocytosis), and 5 mg/mL of HA (an inhibitor of CD44-mediated uptake) were preincubated with the cells for 1 h at 37 °C. Next, similar concentration ranges used in confocal/cytotoxicity assessments of HA-NCPDs were used to evaluate the uptake by FACS analysis to detect fluorescence (ex/em 488 nm/520 nm) using a FACS LSRII-green flow cytometer (BD). A total of 10,000 gated events were acquired per sample and the mean fluorescence intensity was plotted in a histogram-based graphical representation. Each data point is representative of the mean of three independent measurements on the flow cytometer. Data were analyzed with FlowJo (Tree Star) software.

Results and Discussion

Preparation and Characterization of HA-NCPDs

We report for the first time generatiopan class="Chemical">n of a novel fluorescent HA–aminpan>o acid conjugate system by a controlled hydrothermal reactionclass="Chemical">pan> under mild acidic conditions. pan> class="Chemical">HA is a nonsulfated, nontoxic glycosaminoglycan biopolymer that consists of alternating (1–3)-β-linked N-acetyl-d-glucosamine and (1–4)-β-linked d-glucuronic acid, and presents abundant conjugatable groups making it well suited for the design of functional nanomaterials.[15,20−22] Due to their diverse chemical properties and ability to structure proteins,[23−25] amino acids can be used as functional groups for the formation of various amphiphilic nanoparticles. Hydrophobic amino acids such as tryptophan, phenylalanine, and proline can be conjugated to a hydrophilic polymer to modulate the amphiphilicity of self-assembling nanoparticles.[26] In this context, amino acids are ideal hydrophobic moieties to drive the amphiphilicity for the formation of NCPDs. In this approach, nonfluorescent HA reacts with nonfluorescent amino acids under hydrothermal conditions to form blue fluorescent NCPDs. The high water solubility (20 mg/mL) of the reactants and products eliminates many of the challenges encountered during a multistep synthetic procedure. This suggests that the HA amino acid conjugate system is a strong example of a green approach for large-scale production of NCPDs.

We hypothesized tpapan class="Chemical">n class="Chemical">hat a Schiff bonclass="Chemical">pan>d is formed via an pan> class="Chemical">amine-carboxyl rearrangement reaction.[27,28] The Schiff base formation occurs between an active aldehydic group and a nucleophilic amine forming a −C=N (imine) bond in a conjugated product. The imine bond functions as a primary subfluorescent center interlocked within the polymeric matrix.[29] However, the Schiff condensation reaction is just one of many possible routes for the formation of NCPDs. The hydrothermal reaction facilitates increased collision among reactants due to elevated temperatures, opening many reaction pathways to enable the formation of NCPDs. Due to this phenomenon, HA could be depolymerized into smaller molecular weight fragments,[30] which could potentially form electrostatic/covalent conjugates driven by hydrophobicity of the amino acids to form the NCPDs. As previously reported,[31] HA at temperatures between 80 and 100 °C cleaves across β-glycosidic and C–O equatorial bonds to expose a reactive C=O (Figure S1a). These reactive carbonyl groups are primary electrophilic sites for a nucleophilic attack from reactive amines of the amino acids to form amide linkages. Due to this, there are multiple subfluorophore species (C=N, C=O) formed within the NCPDs leading to the fluorescence generation. Because there are multiple modes and mechanisms proposed for the formation of NCPDs, a higher-order structural information is warranted using excited state dynamics and based on HDX to determine the exact mechanism,[32] which is beyond the scope of the current work.

pan class="Chemical">Disease">HA-NCPDs emit blue fluorescenpan>ce unpan>der a UV excitation at 365 nm. The observed blue fluorescenpan>ce canpan> be attributed to the new absorption banpan>d inpan> the UV region betweenpan> 280 anpan>d 320 nm cpan> class="Chemical">haracteristic of a n → π* transition for C=N bonds.[33] Several IR absorption bands at 1120–1135 cm–1 indicate C–N stretching and bending vibrations; a broad −OH band at 3320–3340 cm–1, C=N/C=O stretching vibrations at particularly 1620–1640 cm–1, and C–O/C=O stretching characteristic between 1020 and 1060 cm–1 were observed for all HA-NCPDs (Figure S1A). 1H NMR spectral analysis highlighted overlapping regions between 7 and 8.00 ppm belonging to hydrophobic groups of amino acids and N=C–H protons (Figure S1B). We also performed confirmatory 13C NMR to detect peaks between 170 and 190 ppm and attributed them to both −C=N and C=O groups in the formation of fluorescent HA NCPDs (Figures S2–S4).[34]

X-ray photoelectron spectroscopy (XPS) was usepan class="Chemical">d to investigate the surface state and composition of NCPDs for C, pan> class="Chemical">N, and O elements (Figures S1D, S5). Analyzing the high-resolution C 1s spectra, peaks at 283.7, 282.9, and 285.9 eV were assigned to C–C, −C–O, and −C=N/C=O bond signatures, respectively (Figure S1D). Due to lower abundance of imine bonds, there could be a possible overlap of C=O and C=N bond energies. These results demonstrate the successful formation of a covalent bond −C=N and −C=O in HA-NCPDs resulting from conjugation between HA and amino acids.

Amphiphilic waterpan>-soluble spherical NCPDs were formed from the hydrophilic HA backbone and conjugated hydrophobic amino acids, along with pockets of C=N and C=O units. The HA-NCPDs were found to have a negative zeta potential of −17.4 to −19.7 mV that was due to the negatively charged carboxylic groups (−COOH) of HA (Figure S1C).[35] Hydrodynamic size distribution ranged between 200 and 400 nm measured by DLS and TEM (Figure C–D). Overall, the preparation of HA-NCPDs was straightforward and environmentally friendly due to the absence of any organic solvents, making this system a particularly attractive candidate in the field of nanobiomaterials.

Both starting materials lack π-copan class="Chemical">njugated systems as the main source of fluorescence.[36] The exact photophysical mechanism is still under rigorous inpan>vestigation; however, inpan> published literature, it is shown tpan> class="Chemical">hat small molecular rotors (−C=N and −C=O) have hindered bond vibrations and rotations upon immobilization. Restriction of bond movement increases local electron densities in the pockets in the polymer matrix, leading to a decrease in radiationless decay.[37] The buildup of electron density around the molecular rotors ultimately leads to covalent-bond-enhanced emission(CEE) that makes HA-NCPDs fluorescent. The fluorescence enhancement is expected from π–π interactions between the hydrophobic groups along with contribution of the hindered conjugated single-bond rotation of −C=N and −C=O.

Reducepan class="Chemical">d intramolecular motion of conjugated hydrophobic −C=N anpan>d −C=O inpan> the pan> class="Chemical">polymer matrix is also possible due to the hydrophobic interactions in the inner core of the NCPDs. Excitation wavelength, tested between 280 and 420 nm, resulted in a broad emission peak with no dramatic change in the location of the peak wavelength (Figures A–B and S7). HA-NCPDs also exhibited excitation-dependent fluorescent properties, indicating the presence of multiple excited sites, similar to other reports on NCPDs. In summary of a mechanistic understanding of HA-derived NCPDs, we hypothesize that the blue fluorescence origin is a combination of the presence of conjugated −C=N, −C=O groups and multiple fluorescent excited states within the core-polymeric structure.[3,29,38,39]

Figure 2

Absorbance and fluorescence spectra of NCPDs in comparison with (A) precursor amino acids, for example, tryptophan, (B) HA-derived NCPD, and (C) after NaBH4 reduction.

Absorbance apan class="Chemical">nd fluorescence spectra of NCPDs inpan> comparisonpan> with (A) precursor aminpan>o acids, for example, npan> class="Chemical">tryptophan, (B) HA-derived NCPD, and (C) after NaBH4 reduction.

To determipan class="Chemical">ne if a large macromolecule/polymer is necessary to form npan> class="Chemical">NCPDs, we reacted citric acid with tryptophan to investigate if NCPDs were generated. No fluorescence or NCPDs were detected under similar reaction conditions used for NCPDs derived from HA and amino acids. However, if carbonization temperatures (150–200 °C) were achieved, formation of carbon dots was observed with size ranges between 5 and 20 nm due to possible aggregation. We observed that on TEM images and in 1H NMR spectra, individual carbon dots can be seen with sizes of 2–10 nm, and NMR peaks were present (Figures S6–S7) similar to those previously reported in the literature.[40] These results highlight the importance of using a polymeric material and a small amine-rich nucleophilic molecule to generate fluorescent NCPDs.

The specific role of −C=Npan> and −C=O bonds inpan> the fluorescenpan>ce originpan> was further inpan>vestigated by reduction with pan> class="Chemical">sodium borohydride that ultimately leads to the formation of single −NH–CH2- and −CH–OH bonds.[41] The reduced form possesses more freedom for molecular rotations over single bonds, increasing its radiationless deactivation and leading to elimination of any fluorescence properties. HA-based NCPDs were reduced with 0.1 M NaBH4, which resulted in loss of fluorescence. This suggests that the double bonds (−C=N and −C=O) act as primary fluorescent moieties in the HA-NCPDs. Loss of the UV bump between 280 and 320 nm and fluorescence emission between 380 and 550 nm was observed (Figures C and S9). In a previous report, fluorescent PEI-formaldehyde polymer particles[29] displayed increased fluorescence upon reduction with NaBH4 instead of disappearance. This phenomenon was attributed to the formation of π–π* transition from C=C bonds in a double Schiff bond formation (−CH=CH–CH=N–CH2–CH=CH−). This suggests that HA-NCPDs possess only single Schiff bond elements similar to other reports of PEI-glucose[9] and PEI-starch[10]-derived NCPDs.

Stability assessments were performepan class="Chemical">d under different conditions, including a range of pH buffered solutions, salt concenpan>trations, anpan>d under constanpan>t UV irradiation (Figure S12). pan> class="Chemical">No drastic change in maximum fluorescence intensity or photobleaching was observed in solutions with pH ranges (4.0–9.0), different NaCl concentrations (0–2 M), nor upon irradiation with UV light (365 nm, 4 mW/cm2 radiation source) for a period of 10 h. However, at extreme low and high pH, complete disappearance of fluorescence was observed, which could be attributed to destabilization of the chemical structure of HA or increased concentrations of H+ and OH– interrupting/preventing electronic transitions to excited states.[42] Furthermore, no apparent visual precipitation, discoloration, or change in fluorescence intensities at rt was observed during long-term storage for 10 weeks (Figure S13). The above results demonstrate that the HA-derived NCPDs possess an acceptable stability profile and can be expected to remain stable when stored either at rt or at 4 °C.

Cytotoxicity, the Cellular Uptake Mechanism, and Cellular Imaging Application of NCPDs

It is essential tpapan class="Chemical">n class="Chemical">hat imaginpan>g probes used for cellular labeling applications exhibit low pan> class="Disease">toxicity. Thus, we investigated the cytotoxicity profiles of the prepared HA-NCPDs via a CCK-8 assay in two breast cancer cell lines, MDA-MB-231 and CI66, and a nonmalignant breast cell line MCF10A. Viability was greater than 98% for HA-NCPDs (120 μg/mL) through 48 h, as shown in Figure S14, regardless of the HA amino acid conjugate. Because the HA-NCPDs displayed negligible cytotoxicity, a bioimaging study was performed in CI66 cells to compare naïve and DOX-resistant cells. HA-NCPDs were internalized by the cancer cells and localized in the endosomes, as determined by CSLM. Merged images in Figure A show blue fluorescence from HA-NCPDs, Rab5 endosomal marker (green), and nuclear stain (cyan). The HA-NCPDs appeared to primarily remain in the endosomal space and did not enter the nucleus.

Figure 3

(A) CSLM images of CI66 cells incubated with HA-derived NCPDs for 4 h. Blue: NCPDs, green: endosomal marker, and pink: nuclear stain. (B) 3D CSLM images for CI66 spheroids incubated with a library of HA-derived NCPDs (C) Penetration of NCPDs in CI66-derived tumor spheroids, and accumulation of NCPDs in CI66 cells from spheroids indicated as mean fluorescence intensity after flow cytometry analysis. The results are shown as mean ± SD.

(A) CSLM images of CI66 cells incubatepan class="Chemical">d with HA-derived pan> class="Chemical">NCPDs for 4 h. Blue: NCPDs, green: endosomal marker, and pink: nuclear stain. (B) 3D CSLM images for CI66 spheroids incubated with a library of HA-derived NCPDs (C) Penetration of NCPDs in CI66-derived tumor spheroids, and accumulation of NCPDs in CI66 cells from spheroids indicated as mean fluorescence intensity after flow cytometry analysis. The results are shown as mean ± SD.

We investigatepan class="Chemical">d the endocytotic mechanisms responsible for uptake of pan> class="Disease">HA-NCPDs by FACS analysis using specific ligands,[43] which included the following: methyl-β-cyclodextrin (an inhibitor of caveolae-independent endocytosis), ethyl isopropyl amiloride (an inhibitor of micropinocytosis), chlorpromazine (an inhibitor of clathrin-mediated endocytosis), and HA (an inhibitor for CD44-mediated uptake). Uptake of NCPDs was mainly governed through CD44 and clathrin-mediated endocytosis (Figure S15) across all derivatives, which is consistent with other NPs.[44−46] High CD44 expression is expected to facilitate HA-NCPD uptake due to HA-CD44 binding, which is consistent with prior studies.[20,35,47−50]

Traditiopan class="Chemical">nal cell culture methods use a 2-dimensional (2D) monolayer of Disease">tumor cells to study pan> class="Disease">tumor cell biology. 2D systems are limited by the inability to accurately recapitulate the in vivo architecture and microenvironment, the growth and morphology of tumor cells, cell–matrix interconnections, signal transductions, and other aspects. In order to improve the understanding of these cellular phenomena, 3D cell culture, or tumor spheroids, provide a useful platform to highlight key points of translational medicine. 3D cell culture may better bridge the deficits between traditional 2D systems and in vivo experimentation.[51] Thus, we prepared a multicellular 3D tumor spheroid model from CI66 breast cancer cells to evaluate the penetration of HA-NCPDs using CSLM. The data in Figure B show that HA-NCPDs could penetrate to approximately four to seven layers of cells into the spheroid, which was about 40–50 μm. Quantitative measurements of the uptake of HA-NCPDs in the spheroids were performed by single-cell FACS analysis (Figure C) by digesting the spheroids with trypsin to produce single-cell suspension. CSLM images and FACS uptake analysis indicate that HA-NCPDs are beneficial as a cellular imaging platform. The penetration depth of the NCPDs in 3D spheroids is promising but can be further improved by using specific targeting peptides/antibodies for future investigations for improved delivery to multiple cancer in vitro systems. Results from unloaded HA-NCPDs in both 2D and 3D in vitro models highlight the possible application of HA-NCPDs as a delivery vehicle given its nontoxic properties.

HA-NCPD Drug Delivery Properties in a 3D Tumor Model

The possibility of utilizing papan class="Chemical">n class="Chemical">NCPDs as a vehicle for potenpan>tial image-guided drug delivery was investigated usinpan>g pan> class="Chemical">DOX as a model drug. DOX causes dissipation of the mitochondrial membrane, activation of p53, generation of reactive oxygen species, DNA fragmentation triggering apoptosis, and potential necrosis, all leading to sequential cell death.[52] Moreover, due to its intrinsic red fluorescence and its overlapping absorption spectra with the emission spectra of the HA-NCPDs, it is ideal for efficient energy transfer via the Förster resonance energy transfer (FRET).[53] DOX absorbance was used to quantify drug loading based on a calibration curve. The concentration loaded into HA-NCPDs was calculated as follows: 9.3 wt % for HA–tryptophan, 9.8 wt % for HA–tryptophan benzyl ester, and 6.7 wt % for HA–proline benzyl ester NCPDs, which are consistent with other polymeric delivery systems (Figure ).[54]

Figure 4

(A) Schematic representation of DOX loading onto HA-derived NCPDs. (B) Fluorescence spectra of HA–tryptophan DOX-loaded NCPDs representing quenching of the parent blue signal and appearance of DOX fluorescence (520–620 nm).

(A) Schematic representatiopan class="Chemical">n of DOX loadinpan>g onto npan> class="Chemical">HA-derived NCPDs. (B) Fluorescence spectra of HA–tryptophan DOX-loaded NCPDs representing quenching of the parent blue signal and appearance of DOX fluorescence (520–620 nm).

We observed a higher pan class="Chemical">drug loading for tryptophan anpan>d pan> class="Chemical">tryptophan benzyl ester derivatives compared to proline benzyl ester. We hypothesize that the higher DOX loading in tryptophan conjugates was due to additional π–π and electrostatic interactions between the planar anthraquinone ring, benzene ring, and indole ring, which is missing in the proline benzyl moiety. The cooperative effect of ionic interaction (pH and solvent ion effect) and hydrophobic effects facilitates the enhanced DOX loading. Drug loading increased the hydrodynamic diameter of the HA-NCPDs by ∼20–30 nm compared to empty HA-NCPDs as indicated by DLS and TEM (Figure S11). Using different excitation wavelengths between 300 and 400 nm, we observed partial quenching of blue fluorescence from HA-NCPDs and increasing fluorescence at 510–620 nm due to DOX, which is indicative of FRET in the drug-loaded HA-NCPDs.

In vitro pan class="Chemical">drug (DOX) release from the three pan> class="Disease">HA-NCPDs was studied in PBS under two different pH conditions (7.4 and 4.5). The results are shown in Figure , where loaded DOX was released at a higher rate at pH 4.5 (representing the cellular environment) than at pH 7.4 (pH of the blood) for all the three HA-NCPDs. This observation is consistent with the pH stability of HA (Figure S12), which is known to degrade at lower pH. The release of DOX from the HA-NCPDs at pH 4.5 was more gradual than the free drug, suggesting the role of a packed/cross-linked system. The release of DOX from HA-NCPDs was even lower at pH 7.4, suggesting reduced premature release in the bloodstream. The DOX release from the tryptophan conjugates was higher than the proline conjugate at pH 4.5 implying higher drug loading, which correlated with the loading capacities. As shown in Figure , there was no significant difference in release of DOX from tryptophan and tryptophan benzyl ester at both pH 4.5 and 7.4 with a maximum cumulative release of 62%. The free DOX exhibited a burst release profile of 100% under both pH conditions during 1 h of incubation. DOX release profile analysis from all the HA-NCPDs showed that about 40–65% DOX was released even after 48 h.

Figure 5

DOX release curve from HA-NCPDs and free DOX under different pH conditions, (A) 4.5 and (B) 7.4. Release profiles of tryptophan, tryptophan benzyl ester, and proline benzyl ester are shown. The cumulative release profile was obtained by taking samples from the release medium of the DOX-loaded NCPDs samples at specific time intervals.

pan class="Chemical">DOX release curve from npan> class="Disease">HA-NCPDs and free DOX under different pH conditions, (A) 4.5 and (B) 7.4. Release profiles of tryptophan, tryptophan benzyl ester, and proline benzyl ester are shown. The cumulative release profile was obtained by taking samples from the release medium of the DOX-loaded NCPDs samples at specific time intervals.

To explore pan class="Chemical">Disease">cancer cell-killinpan>g efficiency of pan> class="Chemical">DOX-loaded HA-NCPDs as a drug delivery vehicle, we determined the viability of CI66 cancer cells using CCK-8 and apoptosis assays. We observed IC50 values of 1.82, 1.42, 1.92, and 2.24 μM for HA–tryptophan, HA–tryptophan benzyl ester, HA–proline benzyl ester, and free DOX, respectively (Figure A). These results suggest that loading DOX in HA-NCPDs increases the cytotoxic effect of the drug. Furthermore, DOX-loaded HA-NCPDs increased apoptosis and necrosis as demonstrated by a higher percentage of apoptotic/necrotic cells in the Q2 compartment (Annexin V and PI positive) compared to free DOX alone. These results are consistent with IC50 analysis from cytotoxicity assays (Figure C). Interestingly, the CSLM data indicate that DOX was predominantly localized in the nucleus after the 4 h incubation period. We suggest that DOX could be released from the HA-NCPDs at low pH (pH 4.5–5.5) as confirmed by the release study in Figure due to instability of HA at pH 4.5. Due to instability of HA, DOX is readily released in the endosomal-lysosomal compartment. DOX then diffuses to the nucleus as the site of action (Figure A-i).

Figure 6

IC50 assessment on (A) CI66 and (B) CI66-DOX-resistant cells at three different time points (n = 6). (C) CI66 and (D) CI66-DOX-resistant cells were treated as indicated and stained with AnnexinV-FITC and Propidium Iodide (PI) using the FITC Annexin V Apoptosis Detection Kit I (BD Biosciences) and quantified with FACS. Representative scatter plots from each treatment are shown. The percent of late apoptotic/necrotic (Annexin V+/PI+) cells is quantified in each plot.

Figure 7

(i) CI66 and (ii) CI66-DOX res: (A) CSLM images acquired on cells incubated with DOX-loaded HA-NCPDs for 4 h. Green, endosomal marker Rab5 protein and cyan-pink, nuclear stain. Combined images were analyzed to visualize uptake of DOX-loaded HA-NCPDs. Yellow arrows indicate localization of DOX through the red channel. (B) Representative CSLM images of tumor spheroids treated with different DOX-loaded HA-NCPDs for 12 h. Red: DOX-loaded HA-NCPDs. (C) Accumulation of DOX-loaded NCPDs in cells from trypsin-treated spheroids indicated as mean fluorescence intensity after FACS analysis (n = 3). The results are shown as mean ± SD.

IC50 assessment opan class="Chemical">n (A) CI66 and (B) CI66-DOX-resistanpan>t cells at three differenpan>t time poinpan>ts (n = 6). (C) CI66 anpan>d (D) CI66-DOX-resistant cells were treated as indicated and stained with AnnexinV-FITC and Propidium Iodide (PI) using the FITC Annexin V Apoptosis Detection Kit I (BD Biosciences) and quantified with FACS. Representative scatter plots from each treatment are shown. The percent of late apoptotic/necrotic (Annexin V+/PI+) cells is quantified in each plot.

(i) CI66 anpan class="Chemical">d (ii) CI66-DOX res: (A) CSLM images acquired on cells inpan>cubated with npan> class="Chemical">DOX-loaded HA-NCPDs for 4 h. Green, endosomal marker Rab5 protein and cyan-pink, nuclear stain. Combined images were analyzed to visualize uptake of DOX-loaded HA-NCPDs. Yellow arrows indicate localization of DOX through the red channel. (B) Representative CSLM images of tumor spheroids treated with different DOX-loaded HA-NCPDs for 12 h. Red: DOX-loaded HA-NCPDs. (C) Accumulation of DOX-loaded NCPDs in cells from trypsin-treated spheroids indicated as mean fluorescence intensity after FACS analysis (n = 3). The results are shown as mean ± SD.

pan class="Chemical">DOX-loaded npan> class="Disease">HA-NCPD uptake was also studied in a 3D CI66 tumor spheroid model by CSLM as described above with HA-NCPDs. NCPDs confirmed internalization in the spheroids to about five to eight layers of cells, which were about 50–70 μm (Figure B-i). Quantitative FACS assessment (Figure C-i) indicated that the tryptophan derivatives had the highest uptake compared to the proline derivative translating to a higher drug loading and DOX release using the tryptophan derivatives.

The uptake anpan class="Chemical">d antiDisease">cancer effects of pan> class="Chemical">DOX-loaded HA-NCPDs were further investigated in DOX-resistant cells to examine the intracellular drug accumulation and distribution. Viability of CI66-DOX-resistant cancer cells treated with DOX-loaded HA-NCPDs was assessed by the CCK-8 and apoptosis assay. We observed that tryptophan NCPDs exhibited a lower half-maximal inhibitory concentration (IC50) compared to both free DOX and proline derivative (Figure B), but we were not able to obtain absolute IC50 values due to curve flattening. Analysis of FACS, CSLM, apoptosis, and 3D spheroid uptake data in comparison to the IC50 values suggests that the lower IC50 values for the tryptophan derivatives may be indicative of an increased drug payload and thus improved killing efficiency. The results obtained from the apoptosis assay in both nonresistant and resistant cells (Figure ) imply that treatment with the DOX-loaded NCPDs resulted in a higher percentage of apoptotic/necrotic cells in the Q2 compartment (Annexin V and PI positive) compared to free DOX alone, which was consistent with IC50 analysis.

Cellular uptake of pan class="Chemical">DOX was also inpan>vestigated by 2D-CSLM (Figure ii-A). Compared with the cells treated with free DOX, the DOX fluorescence intensity was higher in cells treated with the tryptophan derivatives. Furthermore, higher nuclear DOX localization was observed for DOX-HA-NCPDs compared to cells treated with free DOX. A diffused lower signal for all the derivatives (tryptophan and proline derivative) was observed in the cytoplasm, indicating that the NCPDs might have some affinity toward cellular organelles to facilitate sustained drug release (Figure i-A). These results suggest that the DOX-resistant cells internalize the DOX-HA-NCPDs, slowly release the drug in the cytoplasm, allowing the drug to localize in the nuclei. The decrease in overall DOX fluorescence is due to the dynamic balance of uptake of the NCPDs and efflux of DOX (Figure ). Similar reports were also observed for DOX-loaded carbon dots derived from carbonization of milk with increased delivery to the ACC-2 adenoid cystic carcinoma cell line.[55] DOX uptake was studied in a multicellular 3D CI66-DOX-resistant tumor spheroid model by CSLM to better understand HA-NCPD-mediated delivery in a 3D environment. NCPDs internalized in the spheroids to about five to eight layers of cells, which were about 40–50 μm (Figure ii-B). Relatively higher DOX fluorescence intensities from FACS were observed for the tryptophan and tryptophan benzyl ester NCPDs compared to the drug alone under similar conditions (Figure ii-C). These results indicate that tryptophan derivatives were able to deliver a higher payload and retain in the cell for a longer duration. DOX formulated as a NCPD is expected to evade the efflux by upregulated multidrug-resistant proteins on the cell surface as quickly as free DOX.[56]

The differepan class="Chemical">nces in the IC50 values and the DOX contenpan>t inpan> both CI66 anpan>d CI66-pan> class="Chemical">DOX-resistant cell lines strongly suggest that drug-loaded NCPDs elicit a higher antitumor efficiency and extended therapeutic effect in vitro compared to free DOX. This can be attributed to the higher payload delivery of the drug, which is conducive to a slower prolonged drug release (Figure ), providing a valuable platform for targeted cancer chemotherapy. In both nonresistant and resistant cancer cells, incubation with DOX-loaded NCPDs not only resulted in significant cell death but also a higher degree of apoptosis, in comparison to free DOX.

Conclusions

We developepan class="Chemical">d a new fluorescent nonconjugated HA-derived nonconjugated pan> class="Chemical">polymeric dots (NCPDs) from nonfluorescent precursors via a hydrothermal green chemistry process. Advantages and characteristics of this system include high water dispersibility, ease of synthesis, blue excitation-dependent fluorescence, good stability profiles, and low toxicity. All of this makes the NCPDs promising materials for bioimaging and drug delivery applications. Fundamental properties of fluorescence arising from the formation of Schiff base, and self-assembly properties from tuning hydrophobicity by choosing various amino acids have been discussed. Future studies in this project will entail studying the antitumor efficacy in different in vivo cancer models, understanding biodistribution, and survival analysis..