Highly Homogeneous Biotinylated Carbon Nanodots: Red-Emitting

Subscriber access provided by UNIV AUTONOMA DE COAHUILA UADEC

Biological and Medical Applications of Materials and Interfaces

Highly Homogeneous Biotinylated Carbon Nanodots: Red-emitting Nanoheaters as Theranostic Agents Towards Precision Cancer Medicine Cinzia Scialabba, Alice Sciortino, Fabrizio Messina, Gianpiero Buscarino, Marco Cannas, Giuseppina Roscigno, Gerolama Condorelli, Gennara Cavallaro, Gaetano Giammona, and Nicolò Mauro ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.9b04925 • Publication Date (Web): 15 May 2019 Downloaded from http://pubs.acs.org on May 17, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Highly Homogeneous Biotinylated Carbon Nanodots: Red-emitting Nano-heaters as Theranostic Agents Towards Precision Cancer Medicine Cinzia Scialabba,a Alice Sciortino,b Fabrizio Messina,b Gianpiero Buscarino,b Marco Cannas,b Giuseppina Roscigno,cd Gerolama Condorelli,c Gennara Cavallaro,a Gaetano Giammonaa and Nicolò Mauro ad* a Laboratory

of Biocompatible Polymers, Department of “Scienze e Tecnologie Biologiche, Chimiche e

Farmaceutiche” (STEBICEF), University of Palermo, Via Archirafi, 32, 90123 Palermo, Italy. b

Dipartimento di Fisica e Chimica, Università Degli Studi di Palermo, Via Archirafi 36, 90123

Palermo, Italy c

Department of Molecular Medicine and Medical Biotechnology, “Federico II” University of Naples,

Naples, Italy d

Fondazione Umberto Veronesi, Piazza Velasca 5, 20122 Milano, Italy

Keywords Carbon nanodots, biotin, targeted cancer therapy, photothermal therapy, imaging ABSTRACT Very recent red-emissive carbon nanodots (CDs) have shown potential as NIR converting tools to produce local heat useful in cancer theranostics. Besides, CDs seem very appealing for clinical applications combining hyperthermia, imaging and drug delivery in a single platform capable of selectively targeting cancer cells. However, CDs still suffer from dramatic dot-to-dot variability issues, such that a rational design of their structural, optical and chemical characteristics for medical applications has been impossible so far. Herein, we report for the first time a simple and highly controllable layer-by-layer synthesis of biotin-decorated CDs with monodisperse size distribution, well

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 38

established polymeric shell thickness and degree of surface functionalization, endowed with strong red luminescence and the ability to convert NIR-light into heat. These CDs, henceforth named CDs-PEGBT, consist of a carbonaceous core passivated with biotin-terminated PEG2000 chains, which we demonstrate as active targeting groups to recognize cancer cells. The CDs-PEG-BT are designed to efficiently incorporate high amount of anticancer drugs such as irinotecan (16-28 %) and to act as NIRactivated nano-heaters capable of triggering local hyperthermia and massive drug release inside tumors, thus provoking sudden and efficient tumor death. The potential of the irinotecan-loaded CDs-PEG-BT (CDs-PEG-BT @IT) in fluorescence imaging (FL) was studied on 2-D cultures as well as on complex 3-D spheroids mimicking in vivo tumor architectures, showing their capability of selectively entering cancer cells through biotin receptors overexpressed in cell membranes. The efficient anticancer effect of these CDs was thoroughly assessed on multicellular 3-D spheroids and patient organoids (tumor-ona-dish preclinical models) to predict the drug response in humans in view of personalized medicine applications. CDs-PEG-BT@IT have a smart combination of properties which pave the way to their real-world use as anticancer theranostic agents for image-guided photothermal (IG-PTT) applications. 1. INTRODUCTION In recent years, considerable attention is being focused on photothermal nanotheranostics, as a promising route to overcome shortcomings of conventional cancer therapies such as poor bioavailability, severe side effects and multidrug resistance (MDR) due to non-specific drug biodistribution.1,2 In fact, the cutting edge of modern nanomedicine envisions a multi-target approach capable of selectively killing cancer cells through non-invasive imaging-guided photothermal therapy (IG-PTT) as well as monitoring the progression of the treatment.2–4 Typical materials proposed for this application are organic dyes, noble metal and metal oxide nanoparticles, which have strong fluorescence and high photothermal conversion within the biological transparency window (beyond


Page 3 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


620 nm), allowing to reach proper contrast in fluorescence (FL) imaging and light-activated hyperthermia of the tumour mass.5–7 While FL imaging provides a map to track cancer cells, hyperthermia is the most powerful and localized therapeutic way to eradicate a tumour mass, since a focused rise in temperature of a tumour mass (40-45 °C) leads to sudden death of cancer cells or hypersensitization toward xenobiotics, thus circumventing MDR.8–12 However, the potential long-term accumulation, non-degradability, the low thermal- and photo-stability of such nanosystems dictate serious limits to their therapeutic applications in vivo.13,14 On the other hand, near-infrared (NIR)responsive carbon-based theranostic agents, such as graphene oxide and carbon nanotubes, are of keen interest for IG-PTT utilization because of their biodegradability and peculiar physicochemical features.9,15,16 Despite that, the poor water stability and the undefined biocompatibility to chronic exposure impede their potential utilization on humans.15,17 Carbon nanodots (CDs) are an interesting family of emerging zero-dimensional materials endowed with a combination of properties, such as strong and stable fluorescence, high photo/thermostability, huge surface area, low cost, high biocompatibility and efficient photothermal conversion, which makes them extremely promising in optoelectronic, photocatalytic, sensing and biomedical applications.18–23 Very recent work has suggested that CDs could also be employed as theranostic anticancer agents for IG-PTT applications, although several limitations still need to be overcome.8 Despite CDs display excellent optical absorption and emission bands in the blue-green region of the spectrum, modulating these bands to the biological transparency window region (red region) to obtain good performance for FL imaging remains demanding.24–26 In fact, living tissues do not absorb red light, permitting good penetration of signals derived from FL diagnostic agents in vivo. However, only a few reports of redemitting CDs with an acceptable NIR-triggered photothermal effect are available in the literature.27–29 Although such CDs have good NIR fluorescence and efficient photothermal conversion they are quite inhomogeneous and characterized by a large dot-to-dot variability: most CDs consist of nanostructures



Page 4 of 38

with broad size distribution, extremely complex emission bands and uncontrolled surface functionalization, thus impeding a refined modulation of chemical and physicochemical features.30,31 Besides, these samples suffer from a lack of surface cell recognition mechanisms (active targeting) that could be responsible for the weak targeted tissue accumulation observed for these CDs in vivo so far.20,32 Hence, engineering well-defined red-emitting CDs with optimal photothermal conversion, homogeneous size and surface, uniform red emission spectrum and diligently functionalized on the surface to intercept cancer cells would be a significant enhancement to impart tailor-made properties and fabricate smart nanomaterials exploitable in anticancer nanotheranostics. In our previous work, we developed a simple method to select CDs from the solvothermal reaction of urea and citric acid with extremely homogeneous morphological, structural and optical characteristics.20 Herein we extend this method to obtain CDs endowed with

tailor made

photoluminescence in the red region and NIR-triggered photothermal conversion, and we functionalized these dots with biotin-terminated PEG chains to lead to CDs (named CDs-PEG-BT) with cell-recognition capability.9 While PEG passivation yields CDs with isolated red emission, desired in FL imaging,33 the biotin (BT) end-chains play a role of targeting agent toward cancer cells (which normally overexpress BT receptors, BR) and show huge potential to efficiently enter cancer cells, as we directly demonstrate in multicellular 3-D cell line spheroids and patient organoids. The CDs-PEGBT was loaded with irinotecan, used as anticancer drug against breast cancer cells, and employed as “Trojan horse” to release its payload by NIR-triggered photothermal activation once inside tumor mass. The massive release of irinotecan from the CDs-PEG-BT, which occurs during the photothermal ablation of the tumor mass (T = 40-45 °C), serves as local dose for further acting against cancer cells eventually escaped from the primary tumors during its physical disruption, thereby avoiding the outbreak of metastatic stem cells and MDR. On the whole, we designed multifunctional CDs-based


Page 5 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


theranostic agents with highly controlled structure and physicochemical properties, which will provide advanced strategies to recognize, eradicate and monitor cancer by IG-PTT. 2. EXPERIMENTAL SECTION 2.1. Synthesis of the Biotin-functionalized Carbon Nanodots (CDs-PEG-BT). The crude CDs were prepared according to the literature with some additional modification.19 In particular, urea (6 g) and citric acid (3g) were dissolved in DMF (30 mL) and placed under solvothermal condition at 160 °C for 4 h. Then, the reaction was cooled down to r.t. and ethanol (60 mL) was added. The precipitate was retrieved by centrifuging at 10000 rpm for 10 min and washed with ethanol (30 mL) three times. The dark powder was dispersed in water (8 g L-1) by sonicating (15 min x 3) and CDs with different size distribution and absorption bands were separated by size exclusion chromatography (SEC), using a glass column (100 cm length, 1.5 cm diameter) packed in turn with sephadex G25 (15 g), G15 (15 g), and G10 (30 g). Only red-emitting fractions with a strong NIR-sensitive photothermal behavior, henceforth named CDs, were selected for further surface passivation. In a first step, the amino-PEG2000alkyne (NH2-PEG-CC, Figure S 4-5) (250 mg, 7.2x1019 chains) was dissolved in an aqueous dispersion of CDs (10 mL, 2 mg mL-1, 3.4x1018 nanoparticles), the pH was adjusted to 6.4, and EDC (24.92 mg, 0.13 mmol) and NHS (14.96 mg, 0.13 mmol) were added under stirring. The reaction was kept at pH 6.4 for 18 h and purified from the unreacted NH2-PEG-CC and by-products through dialysis. A violet powder of CDs-PEG-CC was obtained after freeze-drying. Yield: 72%. Finally, CDs-PEG-CC (50 mg) and biotin-PEG300-N3 (Sigma Aldrich) were dispersed in water (4 mL) by sonicating (15 min x 3). Then, ascorbic acid (10% cat.) and copper (II) sulfate (10% cat.) were added under nitrogen atmosphere and the reaction was kept under stirring for 18 h. After this time, the reaction was purified by SEC as above described and freeze-dried to get the CDs-PEG-BT . Yield: 88%.



Page 6 of 38

2.2. Characterization of the Carbon Nanodots and Their Derivatives. The diameter distribution of the CDs was evaluated on the basis of their height from the atomic force microscopy (AFM) images. 10 L of aqueous dispersions of either CDs, CDs-PEG or CDs-PEG-BT (0.1 mg L-1) were deposited on a mica substrate and then dried in vacuum (10 mbar). AFM measurements were carried out in air using a Bruker FAST-SCAN microscope equipped with a closed-loop scanner (X, Y, Z maximum scan region: 35 m, 35 m, 3 m, respectively). Scans were obtained in soft tapping mode using a FAST-SCAN-A probe with apical radius of about 5 nm and each image was obtained with a resolution comparable to the tip radius. The structure was characterized using a JEOL JEMS-2100 High Resolution Transmission Electron Microscope (HR-TEM) operating at 200 kV electron energy. Each sample was prepared in ultrapure water (1 mg L-1) and deposited on a 400 m mesh Cu-grid covered by a holey amorphous carbon film, with nominal thickness of 3 nm. The surface functional groups of the CDs and their derivatives were studied by FTIR on a Bruker Alpha II spectrometer using a transmission geometry (scan times: 24; resolution: 4 cm-1). Samples were prepared as KBr pellets and the measurements were collected at room temperature under nitrogen atmosphere. The pKa values and the meq of surface acidic and amine groups were extrapolated employing the de Levie method of acidbased chemical equilibria for polyelectrolytes (supporting information). All samples (30 mg) were dispersed in 0.0036 M HCl (15 mL), diluted with CO2-free water to a total volume of 20 mL, and titrated with 0.01 M CO2-free NaOH at 25±0.1 °C and under nitrogen atmosphere by using an AMEL 631 differential electrometer. The backward titrations were carried out with 0.01 M HCl conducted with a dynamic injection volume ranged from 0.010 to 0.100 mL. To maintain the activity of H+ ions constant, the ionic strengths of the HCl and NaOH solutions were adjusted to 0.10 M by dissolution of NaCl.34 The pH-metric system was calibrated against multiple standard buffer set (pH = 1.629, 4, 7, 9, 10 and 11), exhibiting a calibration fitting with R2 = 0.9997, slope = 56.81 mV and ideality grade of 96.1%. ACS Paragon Plus Environment

Page 7 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2.3. Optical Characterizations. Steady-state absorption reported in Figure 3a have been performed on aqueous solutions prepared dissolving the CDs (20 mg L-1) and the biotinylated CDs (100 mg L-1) in ultrapure water. Steady-state absorption measurements were carried out using a double beam spectrophotometer (JASCO V-560) in the 220–700 nm range in a 1 cm quartz cuvette. The emission spectra have been collected on diluted solutions of CDs in ultrapure water by a JASCO FP-6500 spectrofluorometer in a 1 cm cuvette with a 3 nm resolution bandwidth. To measure the emission quantum yield, samples fluorescence has been compared with that obtained by reference samples (Rhodamine 6G in water, pH = 13) under identical excitation conditions. The emission kinetic traces were recorded exciting the solutions by 5 ns laser pulses of 0.1–0.5 mJ energy provided by a tunable laser consisting of an optical parametric oscillator pumped by the third harmonic of a pulsed Qswitched Nd:YAG laser. The laser provides pulses in the visible (410–700 nm) range. The kinetic traces were recorded by an intensified CCD camera capable of acquiring the emission spectrum within temporal windows of 0.5 ns with controlled delays with respect to the laser pulse. The time resolution of is about 0.2–0.3 ns. 2.4. Preparation of the Irinotecan-loaded CDs-PEG-BT (CDs-PEG-BT@IT). The adsorption of irinotecan hydrochloride (IT) (15 mg) on the CDs-PEG-BT (20 mg in 3 mL water) surface was attained by means of sonication (2 x 15 min). After 16 h of incubation at 25 °C the free drug was removed by dialysis against water and the amount of drug adsorbed was calculated spectrophotometrically measuring the absorbance of the waste water at 366 nm (DL = 16%). The same procedure was adopted for the preparation of the CDs-PEG-CC@IT nanosystem (DL = 28.4%). 2.5. Characterization of CDs-PEG-BT@IT. Either free IT (0.1 mg mL-1) or equivalent amount of CDs-PEG-CC@IT or CDs-PEG-BT@IT was dissolved in PBS pH 7.4 (1 mL) and placed into a dialysis tubing with MWCO 2 kDa. Then it was embedded into PBS (9 mL) and incubated at 37 °C under orbital stirring (100 rpm) for 48 h. At scheduled time intervals the external medium (0.2 mL) was ACS Paragon Plus Environment


Page 8 of 38

withdrawn and replaced with equal amount of fresh medium. The amount of IT released was calculated spectrophotometrically as above described. The photothermal-triggered drug release was evaluated treating a dispersion of CDs-PEG-BT@IT in PBS pH 7.4 (0.1 mg mL-1) for 200 sec with an 810 nm laser (GBox 15A/B by GIGA Laser; power 0.1 W mm-3) before the dialysis process. Then, the drug release after 4 h of incubation in PBS pH 7.4 was assessed to make a comparison with the untreated control. The photothermic effect was determined by exciting an aqueous dispersion of CDs or CDsPEG-BT (10 mL, 0.1 mg mL-1) with an 810 nm laser with the power fitted at 8.0 x 10-3 W mm-3, and monitoring the temperature of the dispersion over exposure time. Data were compared to that of ultrapure water (10 mL). 2.6. In vitro Anticancer Activity of CDs-PEG-BT@IT on 2-D models. The cytotoxicity of the loaded CDs derivatives was assessed by the MTS assay (Promega) on human breast cancer cell lines (MCF-7, MDA-MB-231 by Sigma Aldrich). Cells were seeded in a 96-multiwell plate at a density of 1 × 104 cells/well and grown in supplemented Dulbecco’s Minimum Essential Medium (DMEM) (supporting information). After 24 h, the medium was replaced with 200 μL of fresh culture medium containing either CDs-PEG-CC@IT, CDs-PEG-BT@IT or irinotecan at a drug concentration per well ranging from 25, to 150 μg mL-1. Untreated cells were used as control. After 24 and 48 h, samples were taken away from the wells and substituted by fresh medium (100 μL) and 20 μL of a MTS solution. Cells were incubated for additional 2 h at 37 °C before measuring the absorbance at 492 nm. All culture experiments were performed in triplicates. The 2-D cell uptake of CDs-PEG-BT (0.5 mg mL-1) was evaluated by fluorescence microscopy on MCF-7 and MDA-MB-231 monolayers after 6 h of incubation. Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI). Images were recorded by a fluorescence microscope using an Axio Cam MRm (Zeiss). Untreated cells were used as negative control to set the auto fluorescence.


Page 9 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2.7. In vitro Anticancer Activity of CDs-PEG-BT@IT on 3-D models. The anticancer efficacy of CDs-PEG-BT@IT was also tested on multicellular 3-D spheroids consisting of a core of MDA-MB231 or MCF7 and a shell of HDFa, formed using Perfecta3D 384-Well Hanging Drop Plates and following a procedure reported elsewhere35. Briefly, 20 μL of a cells suspension (HDFa/MDA-MB-231 or MCF7 1:1), at density of 100 cells μL-1 enriched with 2.5% of ECM Gel, was added into each well from the top side. Spheroids were growth for three days adding every day 5 μL of fresh medium. The fourth day, 10 μL of media were replaced with fresh cultured media containing CDs-PEG-BT@IT or irinotecan at a final drug concentration per well within the range of 300 - 25 μg mL-1. Untreated spheroids were used as negative control. After 48 h the cell viability was evaluated by MTS assay as above described. In another experimental set, the NIR-triggered photothermal ablation of the multicellular spheroids by CDs-PEG-BT@IT was evaluated. Spheroids were treated with CDs-PEGBT@IT (150 μg mL-1) for 48 h and then irradiated with an 810 nm laser beam for 30 seconds (4 W cm2).

The photothermal ablation was measured as the reduction of the volume of spheroids calculated by

the formula V = L · W2 / 2 (where L is the longest diameter of the spheroids and W is the longest perpendicular diameter with respect to L).36 The ability of CDs-PEG-BT@IT to enter spheroids and selectively target breast cancer cells was established on three days cultured organoids. Spheroids were transferred into 8 well plate and incubated with CDs-PEG-BT@IT (100 μg mL-1) for 24 h. Then the culture medium was removed, the spheroids were washed with DPBS and the cell nuclei were stained with DAPI. Images were recorded by a fluorescence microscope using an Axio Cam MRm (Zeiss). Untreated spheroids were used as negative control to setup the auto fluorescence. 2.8. Ex-vivo Anticancer Activity of CDs-PEG-BT@IT on 3-D organoids (Tumor-on-a-dish). The breast cancer drug response was also tested on tumor organoids isolated from human biopsies. Tumor organoids will be separated using Hans Clevers protocol. Briefly, biopsies were minced and then digested in 10 mL AdDF (Advanced DMEM/F12, 1x Glutamax, 10mM HEPES, and 1% antibiotics/ ACS Paragon Plus Environment


Page 10 of 38

antimycotic) containing 2 mg mL-1 collagenase (Sigma, C9407) on a shaker at 37 °C for 3 h. Then, the digested suspension was shared using sequentially plastic with less diameter and finally glass Pasteur pipette. After that, the suspension was centrifuged at 400 rcf. If erythrocytes were present, the pellet was suspended in red blood cell lysis buffer (Roche 11814389001) for 5 min. The pellet was washed and centrifuged again at 400 rcf. Finally, the pellet was resuspended in cold Cultex growth factor reduced BME type 2 and 40 l drops were allow to solidify in a 24 multi-well suspension plates (Greiner, M9312) for 20 min at 37°C. Tumor organoids were then cover with 400 µl of organoid medium as describe by Hans Clevers protocol. Medium will be replaced every 3–4 days. Isolated organoid will be propagated over time and splitted enzymatically using Triple Express (Invitrogen, 12605036) for 10 min at 37 °C and glass Pasteur pipette. Then, the organoids were washed with AdDF buffer and centrifuge at 400 rcf. Smaller organoids are resuspended in new BME drops and diluted in a ratio of 1:1-1:4. Organoids were splitted and plated in 100 L of growth media containing 2% BME. Then, organoids suspension was plated in a 96 well coated with 40 L of BME. The day after, irinotecan was dissolved at an initial concentration of 0.35 mg mL-1 and diluted with serial dilutions up to 0.05 mg mL-1. The CDs-PEG-BT@IT was dissolved at an initial concentration of 1.82 mg mL-1 up to 0.26 mg mL-1. Finally, 200 L of drug solution were added on organoids in triplicate. Cell viability was evaluated by MTS assay. 3. RESULTS AND DISCUSSION 3.1. Design and Preparation of CDs-PEG-BT. Carbon nanodots (CDs) were prepared by a simple solvothermal method from citric acid and urea in N-N-dimethylformamide (DMF).19 Then, the use of size exclusion chromatography (SEC) allowed to isolate the most red-luminescent CDs fraction endowed with NIR-triggered hyperthermic effect. This was achieved by using a specific combination of sephadex stationary phase packed with increasing cutoff values (G10G15G25), chosen on the


Page 11 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


basis of previous works indicating the correlation between the hydrodynamic radius (Hr) and the average weight molecular weight (Mw) of statistical polymers.20,37 This expedient allows to convert the rage of exclusion of commercially available resins from Mw to Hr so as to select a stationary phase suitable to separate nanoparticles in the 0.5-8 nm range of diameter. We obtained eight fractions (CDs1CDs8) with distinct optical features, with emission bands ranging from blue to red. Besides, an initial survey of these samples revealed that all of them possess significant NIR-sensitive photothermal effect under excitation with an 810 nm laser diode source. The optical spectra of these fractions, very variable, are reported in Figure S1. On these grounds, the first step of this work was to produce homogenous CDs with optical performance appropriate for bioimaging applications. Hence, we focused our attention only on one particular fraction, which represents about 25% of the crude reaction (Yield = 0.8 g), showing interesting characteristics, including homogeneous morphology (Figure S2) and a fluorescence quantum yield (QY) two times higher than the other fractions, combined with acceptable photothermal effect (Figure S3). Although red emissive CDs with higher QY (~ 18 %)38 have been recently reported, our QY in the red region (4%) is enough high to work well in bioimaging applications.33,39,40 It should be highlighted that fluorescence and photothermal relaxation, both desired for photothermal theranostic agents, are competitive phenomena that occur after photoexcitation of a CDs sample; the higher the photoluminescence, the lower the photothermal conversion.41 Hence, although our QY is lower if compared with such very recent red emissive CDs, it provide sufficient photothermal conversion of the absorbed radiation. This sample, named CDs hereafter, was used as the starting point for further surface functionalization with PEG2000 chains and BT end-groups (Figure 1).



Page 12 of 38

Figure 1. Schematic representation of the synthetic pathway employed for the synthesis of CDs-PEGBT To this purpose we chose to use heterodifunctional PEG with a molecular weight of 2 kDa (Figure S4) to coat CDs surface, because it is biocompatible and protects nanoparticles from opsonization, thereby improving their bioavailability after intravenous administration.42 Indeed, PEG2000 decreases unspecific uptake from the reticuloendothelial system (RES), liver and spleen, by a factor of 6–9 and 2–3, ACS Paragon Plus Environment

Page 13 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


respectively, reducing undesired accumulation.42 Moreover, the pristine CDs have a diameter of 1.5 nm and would be renally excreted (renal cut-off: 5 nm), whereas after surface functionalization with PEG2000 chains the hydrodynamic diameter of the CDs it is expected to increase over 5 nm, thus avoiding fast renal clearance.43 Slow renal excretion of drug delivery systems and reduced RES clearance by PEGylation are encouraging features to reach high concentration of anticancer drugs inside tumors so as to prevent MDR. The adopted synthetic pathway involved two steps depicted in Figure 1. In the first one heterodifunctional PEG bearing amine and alkyne end-groups (Figure S4, S5a), named NH2-PEG-CC, was orthogonally conjugated to the CDs surface by the amide coupling reaction between carboxyl functions at the CDs surface and the amine function of NH2-PEG-CC. This reaction took place using N-(3-Dimethylaminopropyl)-N′-ethylcarbodiimide (EDC) and Nhydroxysuccinimide (NHS) as coupling agents giving rise to, after purification, alkyne-functionalized CDs (CDs-PEG-CC). The feasibility of the surface functionalization was established by FTIR and 1H/13C

NMR confirming the presence of alkyne functions at the CDs surface (Figure S5 b-c). In the

second step, to enhance active targeting effect towards cancer cells, CDs-PEG-CC were further modified with biotin end-groups by means of the Cu (I)-catalized Huisgen 1,3-dipolar cycloaddition between alkyne functions of CDs-PEG-CC and the azide group of BT-terminated PEG-azide (N3-PEGBT) (Figure 1). This reaction was chosen because of its regioselectivity and quantitative yield (>95%), which promise the total conversion of alkyne functions into BT end-chains. High density of BT pendants at the CDs-PEG-BT surface is desired to obtain efficient active targeting toward cancer cells, since as a rule receptor-mediated internalization mechanisms of nanoparticles are ligand concentrationdependent. Hence, the total conversion from alkyne groups to BT functions was easily obtained using slight excess N3-PEG-BT (1.2 eq.), which was removed by SEC after completion. The degree of biotinylation, measured spectrophotometrically by the HABA/avidin assay (Sigma Aldrich), was 0.372 µmol mg-1, roughly corresponding to eighteen PEG-BT chains per dot.



Page 14 of 38

3.2. Characterization of CDs-PEG-BT. Atomic force microscopy (AFM) of bare CDs and the functionalized CDs reveals thousands of isolated objects with very limited tendency to aggregation (Figure S2). From these images we extracted the size distribution reported in Figures 2 a’-c’. High resolution transmission electron microscopy (HR-TEM) was employed to investigate the structure of carbonaceous core of the CDs-PEG-BT. HR-TEM revealed that pristine CDs were homogeneous crystals of a few nanometers with regular structure (Figure 2a). The core of each carbon dot exhibited crystalline lattice fringes with a spacing of 2.252 Å, a little bigger than other recently reported red emissive CDs (0.15-0.22 nm).23,38,40 The fine structure of the carbonaceous core was also preserved after the functionalization, as showed in Figure 2 b-c, implying that the reaction of CDs with PEG-BT only occurred at the dot surface. These data were also corroborated by DSC/TGA analysis, showing that CDs-PEG-BT is essentially composed of CDs (~5% w/w) and PEG-BT (~89% w/w) (Figure S6 a-c). The average diameter of CDs samples, determined by AFM, grew as the molecular weight of the PEG shell increased (Figure 2 a’-c’). In particular, the pristine CDs had a diameter of 1.51±0.25 nm (Figure 2a’), and passed to 5.6±0.9 (Figure 2b’) and 8.1±1.3 nm (Figure 2c’) for the CDs-PEG-CC and CDs-PEG-BT, respectively. It is noteworthy that all CDs samples, despite the surface functionalization, have narrow size distribution if compared with most CDs described so far.40 In particular, using the layer-by-layer approach the typical multimodal size distribution reported for biotinylated graphene quantum dots synthesized by coupling reaction was not observed.44 This hints an efficient purification of the crude reaction coupled with the highly conservative and regioselective functionalization of the CDs surface, the latter being due to the heterodifunctional and orthogonal nature of the PEG end-chains, which does not permit dot-dot crosscoupling reaction. The effective functionalization of the CDs surface with PEG-BT chains was also confirmed by FTIR spectroscopy. As shown in Figure 2d the IR spectra have many diagnostic bands typical of hydroxyl, amine, carboxyl and amide groups, e.g. those relative to O-H stretching (3420 cm-


Page 15 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1),


NH stretching (3200 cm-1), asymmetric (1711 cm-1) and symmetric COOH stretching (1381 cm-1),

and the amide I band (1620 cm-1).

Figure 2. Chemical and physicochemical characterization of the nanosystems. HR-TEM micrographs (a-c) and size distribution obtained by AFM measurements (a’-c’) of the CDs, CDs-PEG-CC and CDsPEG-BT, respectively. (d) IR spectrum of CDs (pink) and CDs-PEG-BT (black). (e) Backward titration of CDs (black) and CDs-PEG-BT (blue): De Levie fitting (red). (f) acid-ionization constants and equivalent of functional groups calculated for CDs and CDs-PEG-BT. There are also bands attributable to C-N (1368 cm-1) and C-O-C-N (1069 cm-1) vibrations. These data indicate that CDs contained several COOH reactive groups amenable to conversion to amide via ACS Paragon Plus Environment


Page 16 of 38

coupling reactions. The appearance of aliphatic bands, attributable to CH vibrations (2891 and 1465 cm-1), and characteristic bands of C-O-C vibrations (1107 and 951 cm-1) in the CDs-PEG-BT spectrum imply that PEGyilation of the CDs core successfully occurred. Furthermore, the depletion of the band at 2108 cm-1 attributable to the alkyne groups of the CDs-PEG-CC precursor pointed out that PEG-BT is covalently bonded by a 1,2,3-triazole bridge. This assumption was also confirmed by the potentiometric titration of surface groups of the virgin CDs and comparing this with results obtained for the CDs-PEG-BT. Salient information on the model used to obtain the pKa values and the equivalent of the surface ionizable groups are reported in supporting information. Given the variety of acidic groups on the CDs surface, this model relies on making a hypothesis on the kind of acidic group of each of the Brønsted–Lowry functions which contribute to the experimental curve, after which one can calculate the corresponding fitting curves. Data reported in Figure 2e-f display for the virgin CDs four acidic groups with pKa values within 3.21-6.55, ascribable to carboxyl groups derived from citric acid portions, and one with pKa of 9.18 typically assigned to primary amines.34 Interestingly, the surface acidic groups of the CDs exhibited pKa values comparable to that of citric acid, hinting that these derived from a conservative structural rearrangement of starting monomers. These groups were still present in the CDs-PEG-BT sample, but the quantitative composition was very different. In particular, carboxyl groups significantly decreased after the coupling process, varied from 5.49 to 1.31 meq g-1, corroborating that we gained a stable covalent surface functionalization of these groups with PEG chains through amidic bonds. 3.3. Optical Characterization. The CDs sample displays a complex optical absorption spectrum with several discernible features across the entire UV/VIS range (Figure 3a). Before PEGylation, this sample shows multiple emissions spanning from blue to the red, similarly to many CDs in literature.45,46 This is shown in Figure 3b, where we report the fluorescence spectra collected upon excitation at three representative wavelengths. In particular, excitation within the absorption features at ACS Paragon Plus Environment

Page 17 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


340, 400 and 550 nm yields three distinct emission bands peaking in the blue, green and red, respectively. Notably, we found that PEGylation leads to isolated red emission, which is the most important for the purposes of this work, and avoid the contribution of the other chromophores. In fact, once the surface of the dots is passivated with PEG2000 chains, a quenching of the blue and of the green emissions occurs (Figure 3b), and only the red fluorescence remains unchanged, with a QY estimated to be 4%. This is self-evident under the white light excitation, which covers the entire spectrum of excitation, where the emission of CDs-PEG-BT appeared red to the naked eye, very different from the pristine CDs (Figure 3 b’-b’’). Comparing in more detail the optical properties, we see that the disappearance of blue/green emissions is accompanied by changes of shape in the absorption spectrum (Figure 3a), in the spectral regions related to the excitation of these two chromophores. On the contrary, the shape of absorption band at 550 nm, that is the one exciting the red emission, does not undergo any variations. Despite the absorption band and the emission QY value of the red chromophore are unchanged upon surface passivation, the emission peak redshifts (Figure 3b), confirming the involvement of the surface in the electronic transition. To explain the origin of the red band and its behaviour upon passivation, it is possible to hypothesize that this electronic transition involves a coupling between the crystalline core and the surface of the dot47. In the raw sample, the absorption transition causes an ultrafast charge separation transferring an electron from the core to the surface, and then the electron exposed to the solvent recombines radiatively with the hole left in the core.47 The presence of PEG chains on the surface of the biotinylated sample probably lowers the energy level of the surface acceptor state, causing a slight redshift of the emission band.



Page 18 of 38

Figure 3. Normalized absorption spectra of aqueous solution of CDs and CDs-PEG-BT(a): the arrows indicate different transitions related to the different chromophores and indicate the excitation wavelengths used to detect the emission spectra in Panel (b). Emission spectra excited at 380 nm, 440 nm and 540 nm of CDs and CDs-PEG-BT (b). Photo of the emission of CDs-PEG-CC (b’) and CDsPEG-BT (b’’) under white light. Decay fluorescence of CDs and CDs-PEG-BT with the respective least-squares fitting curves: λexc 540, λem 630 nm (c). Photothermal effect of the CDs-PEG-BT and the parent compound (0.1 mg mL-1) (d). Cumulative drug release profile of CDs-PEG-CC@IT (red) and CDs-PEG-BT@IT (blue) in PBS pH 7.4 (e): the insert on the right depicts drug release by CDs-PEGBT after 4h of incubation and NIR exposure (200 s, 2 W cm-2). Also the fluorescence decay kinetics is slightly different in the two samples: in fact, we see a small increase of the lifetime value in the biotinylated sample (from 1.2 ns to 1.6 ns, Figure 3c), which suggests that the passivation also introduces small variations in the radiative rate of the emission. Overall, passivation of these CDs with biotinylated PEG chains turns out to be an effective way to


Page 19 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


isolate their red emission capabilities, and to further red-shift the fluorescence without losses in the QY, very conveniently for our application. 3.4. Photothermal effect and Drug Release Studies. Preliminary results lead us to select the CDs fraction with optimal red-emission, needed for deep tissue FL imaging, coupled with a satisfactory photothermal effect required to provide NIR-triggered local hyperthermia (Figure S1 and S3). Although the CDs are endowed with photothermal properties under excitation with an 810 nm laser diode source (Figure 3d), PEGylation might abolish the photothermal conversion of the light into heat. This concern was ruled out comparing the kinetic temperature of the virgin CDs and the CDs-PEG-BT samples. The temperature increase (T) was tested on two samples with a concentration of 0.1 mg mL-1 under 810 nm NIR laser (2 W cm-2, 8.0 x 10-3 W mm-3) and pure water was used as control. Whereas the temperature of pure water increased only of 2 °C after 300 s, virgin CDs provided a T of 24.6 °C after 300 s of exposure. A remarkable photothermal effect (0.1 mg mL-1, T = 18.4 °C) was still measured after the surface functionalization of CDs with PEG-BT. Besides, the photothermal conversion efficiency of the CDs and the CDs-PEG-BT, measured as previously reported48, was 25.8 and 21.2 % respectively, comparable to that calculated for other recently published nano-heaters.28 It seems likely that the NIR absorption from these CDs originates from the long-wavelength tail of their optical absorption band at 540 nm (Figure 3a). It might be noticed that CDs-PEGBT possesses a controllable NIR photothermal effect as a function of the exposure time and power density, which could supply the reported minimum temperature (41 °C) needed to eradicate tumors by hyperthermia,49 and make them potential candidates as nano-heaters for cancer IG-PTT. The virgin CDs have well-defined spherical morphology of 1.5 nm diameter and thus huge surface area to ensure that high amount of drugs can be adsorbed to be delivered at the tumour site. However, being very small, the drug loading (DL) per dot which can be reached and delivered at the site of action is too low and MDR can be promoted. Here, the chemotherapeutic agent named irinotecan (IT), used in ACS Paragon Plus Environment


Page 20 of 38

metastatic breast cancer therapy, was selected as model drug. We observed that CDs alone can adsorb 6.9% of irinotecan (IT) on a weight basis, about 1.3x10-11 pg per particle, while the nanometric PEG shell (3.5 nm thickness) in CDs-PEG-BT allowed reaching a loading capacity four times higher (from 6.9 to 28.4 %), proving that PEGylation is a good strategy to improve the drug loading while maintaining a suitable photothermal effect for IG-PTT applications. This assumption is corroborated by -potential values reported in Table S1, where one can observe a reduction of the -potential from 22.0 to -12.3 mV, owing to strong electrostatic interactions between the cationic drug and negative charges of naked CDs. On the contrary, negligible electrostatic interaction between the drug and the residual charges on the PEGylated CDs surface occurred, since -potential values of the CDs-PEG-BT was comparable to that of CDs-PEG-BT@IT (- vs -5.93 mV, respectively). Drug release kinetics of CDs-PEG-BT were studied in phosphate buffer saline (PBS) pH 7.4 and compared to that obtained for the CDs-PEG-CC at the same drug concentration. Figure 3e shows a slow drug release profile for both CDs sample, suggesting an excellent stability of the drug/CDs complexes in physiological medium, but with some qualification. In particular, CDs-PEG-CC avidly retained its payload over time if compare with the further derivatized CDs-PEG-BT sample (26 vs 48 % drug release in 10 h, respectively). In addition, the biotinylated sample released its residual payload up to 54 % in 48 h as the unbiotinylated CDs reached the equilibrium after 12 h. We speculated that this is due the conformation of PEG chains after biotinylation, which changed from collapsed to the much looser brush-like conformation, thereby enabling a facile drug release across the polymeric shell. Very interesting, the drug payload of the CDs-PEG-BT can be rapidly released (100% release) after irradiating the sample with 810nm laser for 200 s (2 W cm-2, 8.0 x 10-3 W mm-3) (Figure 3e). This behaviour endorses CDs-PEG-BT as a likely candidate to selectively release anticancer chemotherapeutics by photothermal-triggered processes once inside cancer cells. 3.5. Biological Characterization. ACS Paragon Plus Environment

Page 21 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3.5.1. In vitro Anticancer Activity of CDs-PEG-BT@IT on 2-D models. To demonstrate the ability of CDs-PEG-BT of acting as contrast agent in FL imaging, photothermal agent and as carrier to specifically deliver irinotecan into cancer cells by BR, we performed step-by-step experiments on 2-D cell cultures and on 3-D multicellular cultures mimicking the complex architecture of human cancer in vivo. First, FL microscopy (λex = 559) was employed to track our biotinylated CDs in living cells, showing excellent contrast in FL imaging and the ability to distinguish cancer cells from the healthy ones on hierarchically organized 3-D cell cultures named spheroids. Then, we also tested the NIRactivated hyperthermic effect by CDs-PEG-BT on spheroids in order to assess the feasibility of the photothermal eradication of a tumour mass trough IG-PTT. Finally, we carried out cell viability experiments on 3-D patient organoids (tumors-in-a-dish technology) to predict the anticancer efficacy of our biotinylated CDs in vivo.50 This set of preclinical experiments are proposed as a rational approach to evaluate the translational potential of multifunctional nanomedicines to the clinic. First of all, CDs samples were proven biocompatible within the entire concentration range studied (0.01-1 mg mL-1), displaying negligible cytotoxic effects especially for the PEGylated samples (Figure S7). The in vitro cytotoxic effect of CDs-PEG-BT@IT or equivalent amount of either IT or CDs-PEG-CC was carried out on 2-D and 3-D cultures of MCF7 (ER2+, BR+++) and MDA-MB-231 (triple-negative, BR++), two human breast cancer (HBC) cells overexpressing different amounts of biotin receptors (BR).51 They also represent cancers with distinctive invasivity towards pre-metastatic niche, and hence can be used to test the anticancer efficacy of our nanosystem on different kinds of HBCs. As shown in Figure 4a the cell viability of both cells decreased after 48 h of treatment for the free drug and the CDs samples, following the typical dose-response curve. However, on the whole, the susceptibility of MCF7 cells to the drug was greater. The cell viability of the biotinylated CDs was drastically lower if compared with the unbiotinylated CDs-PEG-CC for both cell lines, indicating a crucial rule of BT pendants in determining the higher anticancer performance of the CDs-PEG-BT.



Page 22 of 38

Figure 4. Anticancer activity on 2-D human breast cancer cell cultures (a): MTS assay on MCF7 (dashed lines) and MDA-MB-231 (solid lines) for irinotecan (purple), CDs-PEG-CC@IT (red) and CDs-PEG-BT@IT (blue) after 48h of incubation. Uptake of CDs-PEG-BT after 6 h of incubation on MDA-MB-231 (b-b’’) and MCF7 (c-c’’). This is self-evident comparing the IC50 and Imax values, reported in Table 1, taken as a measure of the pharmacological potency and maximum effect, respectively. The IC50 value of the biotinylated sample was at least two times lower than that calculated for the unbiotinylated one for both cell lines (MCF7: 42 vs 76; MDA-MB-231: >150 ) and comparable to the free drug. Even the Imax values were significantly improved by the presence of BT groups, passing from 63 to 84 % of cell death for MCF7 and 27 to 62 % for MDA-MB-231, confirming the idea that biotinylation may improve the anticancer effect of CDs.9 The uptake of CDs-PEG-BT was tested by fluorescence microscopy on both 2-D cell cultures after 6 h of incubation to explain why biotinylation affects the cytotoxic effect. Figure 4 b-b’’ and 4 c-c’’ shows that the nanosystem can efficiently enter both cell types, but following a different intracellular localization. While CDs-PEG-BT was prevailingly localized inside MDA-MB-231 nuclei (Figure 4 bb’’), after the same incubation time it was localized throughout the cytosol inside endosomes for MCF7 cells (Figure 4 c-c’’), indicating that preferential BR-mediated cell uptake phenomena occurred for MCF7 cells. It might be noticed that, having a diameter of about 8 nm, once CDs-PEG-BT reach the


Page 23 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


cytosolic compartment they could enter MDA-MB-231 nuclei by means of nuclear pore complexes (0.6-10 nm pore size). This was in agreement with the greater expression of BR on the MCF7 cell membranes. It is interesting to notice that the virgin CDs did not display this selective localization during uptake phenomena (Figure S9), fortifying our hypothesis on the rule of biotin to promote the intracellular drug accumulation for a long time. 3.5.2. In vitro Anticancer Activity of CDs-PEG-BT@IT on 3-D models. The potential anticancer effect of the biotinylated CDs was also studied on multicellular 3-D cultures mimicking the typical architecture of solid tumor in vivo, where a robust capsule of fibroblasts surrounds parenchymal cancer cells. In our model we used either MCF7 or MDA-MB-231 in combination with human dermal fibroblasts adult (HDFa) to obtain spheroids as a powerful platform for further investigating drug sensitivity as a key factor of translational medicine. HDFa were used because they are healthy cells which constitute one of the most abundant cells in the stroma and also participate in the formation of sub-epithelial/endothelial basement membranes. We tested the cytotoxic effect of CDs-PEG-BT after an incubation time of 48 h by the MTS assay and compared the values with that obtained for the free IT (Figure 5a). Surprisingly, the biotinylated CDs had a cytotoxic effect comparable to that of the free drug, although a significant increase in the IC50 values was recorder for HDFa/MCF7 spheroids (Table 1; 84 vs 62 g mL-1 for MCF7 and MDA-MB-231, respectively).



Page 24 of 38

Figure 5. Anticancer activity on 3-D spheroids (a): MTS assay on HDFa/MCF7 (open symbol, dashed lines) and HDFa/MDA-MB-231 spheroids (solid symbol, solid lines) for irinotecan (purple) and CDsPEG-BT@IT (blue) after 48h of incubation. Uptake of CDs-PEG-BT@IT after 24 h of incubation on HDFa/MDA-MB-231 (b-b’’) and HDFa/MCF7 spheroids (c-c’’). NIR-triggered photothermal ablation of spheroids after 48 h of incubation followed by the photothermal treatment (30 s, 2 W cm-2, 810nm) (d). Data shown as mean±s.e.m. (n=6, 2 independent replicates). * P

Highly Homogeneous Biotinylated Carbon Nanodots: Red-Emitting

Recommend Documents