Cytotoxicity investigation of luminescent nanohybrids based on chitosan and carboxymethyl chitosan conjugated with Bi₂S₃ quantum dots for biomedical applications

Abstract

Bioengineered hybrids are emerging as a new class of nanomaterials consisting of a biopolymer and inorganic semiconductors used in biomedical and environmental applications. The aim of the present work was to determine the cytocompatibility of novel water-soluble Bi₂S₃ quantum dots (QDs) functionalized with chitosan and O-carboxymethyl chitosan (CMC) as capping ligands using an eco-friendly aqueous process at room temperature. These hybrid nanocomposites were tested for cytocompatibility using a 3-(4,5-dimethylthiazol-2yl) 2,5-diphenyl tetrazolium bromide (MTT) cell proliferation assay with cultured human osteosarcoma cells (SAOS), human embryonic kidney cells (HEK293T cells) and a LIVE/DEAD® viability-cytotoxicity assay. The results of the in vitro assays demonstrated that the CMC and chitosan-based nanohybrids were not cytotoxic and exhibited suitable cell viability responses. However, despite the “safe by design” approach used in this research, we have proved that the impact of the size, surface charge and biofunctionalization of the nanohybrids on cytotoxicity was cell type-dependent due to complex mechanisms. Thus, these novel bionanocomposites offer promising prospects for potential biomedical and pharmaceutical applications as fluorescent nanoprobes.

Graphical Abstract

Carboxymethyl chitosan-Bi₂S₃ quantum dots nanoconjugates : from polysaccharides to cytocompatible fluorescent biohybrids.

Open in new tabDownload slide

1. Introduction

Quantum dots (QDs) are a novel class of fluorescent inorganic semiconductor nanomaterials exhibiting unique optical, magnetic and electronic properties, which provide distinct advantages over traditional organic dyes in terms of their large extinction coefficients, tunable emission features, narrow emission spectra, high quantum yields and relatively high photochemical stability.^1–3 Unlike bulk materials, the optical and electronic properties of semiconductor QDs are very dependent on the nanocrystal dimensions due to the quantum confinement effect. However, after being formed, the nucleated nanocrystals would tend to grow or agglomerate thermodynamically driven by the reduction in the high surface energy (i.e., decrease the surface area/volume ratio). Therefore, several methods have been used to control the growth of the just formed QDs with tunable optical properties such as in glass matrices, Langmuir–Blodgett films, and colloidal ligands.^1–5 The development of QDs has experienced extraordinary advances in recent decades mostly driven by their enormous potential applications in the biological and medical fields, such as in cell biology research for labeling, tracking and imaging, cancer theranostics, immunoassay, biosensing, and others.^6–10 However, some challenges such as aqueous solubility combined with highly luminescent nanocrystals, surface functionalization and non-toxic characteristics need to be more comprehensively addressed by the research community, focusing on their safe utilization in biomedical, pharmaceutical and environmental applications in vivo.  ^2–5 Much of this research has been performed with the ultimate aim of using QDs in clinical applications, but a great deal of concern has been voiced about the latent hazards of QDs based on reports demonstrating release of toxic heavy metal ions from their degradation in cell culture studies.^11,12 Moreover, the cytotoxicity of QDs can be associated with the extremely high surface area (i.e., a large number of atoms with unsaturated bonds) available to generate reactive oxygen species (ROS) and other free radicals with evidence suggesting the involvement in QD-induced toxicity in vitro and in vivo.  ^4–6,13,14 So far, the majority of the QD cytotoxicity studies has been focused on cadmium-based nanocrystals and much less information can be obtained for other types of QDs.¹⁵ In fact, some researchers are convinced that heavy metal-based QDs (e.g. Cd, Pb) will never reach medical and clinical applications, because they are particularly labile and intrinsically carry a highly toxic core, which would produce unpredictable behaviors in the living body.^3,5,16–18 In that sense, the cytotoxicity of QDs is yet highly controversial, with many questions remaining unanswered. Several interesting papers and comprehensive reviews have recently been published investigating the toxicity of QDs in vitro with different cell lines and in vivo using small animal models.^17,18 Ling Ye and co-workers¹⁹ conducted an elegant groundbreaking study demonstrating that Cd-based QD encapsulated phospholipid micelles did not exhibit evidence of toxicity at 90 days after an intravenous injection in non-human primates. Moreover, systematic studies by Chan's group demonstrated that nanomaterial combinations (e.g. CdS, PbS), size ranges, and surface capping chemistry certainly matter with respect to QD toxicity in vitro and in vivo.  ^17,18 Hence, the overall discrepancy in the toxicological data from the in vitro and in vivo studies are essentially related to the fact that nanomaterials, such as QDs and others, are not uniform, i.e., each design is a unique combination of physico-chemical properties that influence the overall biological activity and toxicity evaluation.²⁰ Therefore, developing functional biomedical devices based on semiconductor nanomaterials requires a deep understanding of the interactions taking place at the material-biointerface and an extensive characterization of the entire system from a broad perspective, as the cells’ behavior is very dependent on the local physico-chemical environment. This approach is the starting point for properly assessing the cytotoxicity and biocompatibility of QDs toward designing and producing biologically and environmentally safe QDs.^3,5 To reduce the potential toxicity associated with the heavy metal content of QDs, the interest in semiconductors made of bismuth chalcogenides (e.g., Bi₂X₃, X = S, Se, Te) has intensified in recent years. Bismuth is an extraordinarily harmless element, which has earned the status of a ‘green element’ and has sparked great attention in areas varying from medicinal to environmental applications, as many bismuth based compounds are even less toxic than common food salts (e.g., NaCl).²¹ However, synthesis of Bi₂S₃ nanomaterials is rarely researched, such as for preparing colloidal QDs.²² Additionally, from the perspective of toxicity, the biofunctionalization of QDs with capping ligands has been used as the most common strategy for rendering them water-soluble and biocompatible for biomedical applications, which may theoretically ‘shield’ the toxic semiconductor core with an organic biocompatible layer.^2,3 Nevertheless, in addition to the aqueous solubility and biocompatibility of QDs for in vitro and in vivo biological applications, the surface functionalization needs to fulfill other important requirements, such as specificity for the target site, biochemical stability, and environmental compatibility.³ These requirements can be achieved by choosing “smart ligands” (i.e., affinity ligands with specific parameters for biomolecular interactions with the target site) with functional groups, such as thiols (–SH), carboxyls (–COOH), amines (–NH₂) and hydroxyls (OH), to bind to the QD surface in order to enable targeted bioimaging and drug delivery in vivo.  ^23–25 In that sense, some biomolecules, such as carbohydrates,^26,27 peptides,²⁸ amino acids,²⁹ enzymes and proteins,³⁰ play a major role in biofunctionalization because they combine the presence of functional groups with the biological affinity required for targeting specific cells and tissues. For example, proteins, antibodies and peptides contain several amine and carboxyl groups that can be coupled to the surface of modified QDs through simple amide bonding and later used as bioconjugates for cell imaging, diagnosis and drug carriers.^3,31 More recently, polysaccharides, such as chitosan and carboxymethyl chitosan, have been investigated as a very attractive alternative for the biofunctionalization of QDs due to their intrinsic biocompatibility, rational water-solubility, chemical stability, environmental compatibility, and natural source abundance, as they are usually extracted from crustacean shells.³² Additionally, chitosan, a polyaminosaccharide derived from the deacetylation of chitin, contains amine and hydroxyl groups, which enable the conjugation of distinct biomolecules for specific biological functions based on affinity interactions. Surprisingly, despite their undisputable flexibility and distinct advantages, the direct biofunctionalization of QDs by chitosan and its derivatives using aqueous processing routes is scarcely reported in the literature.^23,33 No research was found in the consulted literature, which has investigated the synthesis of heavy metal-free QDs (e.g., Cd, Pb) based on Bi₂S₃ functionalized by chitosan and its derivatives strictly using water processing routes.

Thus, in this study, novel nanohybrids were designed and synthesized using the approach of simultaneously combining biocompatibility and environmental compatibility via an aqueous colloidal processing route. Thus, novel water-soluble Bi₂S₃ QDs functionalized with chitosan (CHI) and O-carboxymethyl chitosan (CMC) as the capping ligands were produced using an eco-friendly process at room temperature. The physico-chemical properties of these bio-nanohybrids, such as their surface charges and hydrodynamic diameters, and the in vitro cytotoxicity with two cell-types were extensively investigated aiming at potential luminescent nanoprobes for bioimaging and diagnosis applications.

2. Materials and methods

2.1. Materials

All of the reagents and precursors, including bismuth chloride (Aldrich, USA, ≥98%, BiCl₃), sodium sulfide (Synth, Brazil, >98%, Na₂S·9H₂O), sodium hydroxide (Merck, USA, ≥99%, NaOH), and acetic acid (Synth, Brazil, ≥99.7%, CH₃COOH), were used as received. Chitosan was purchased from Aldrich Chemical, USA (catalog no. 419419; high molecular weight, M  _W = 310 to >395 kDa; degree of deacetylation (DD) ≥75.0%; viscosity 800–2000 cPoise, 1 wt% in 1% acetic acid) and O-carboxymethyl chitosan was synthesized with the degree of substitution of 55 ± 1% using a previously reported method from our group.³⁴ Unless otherwise indicated, deionized water (DI water, Millipore Simplicity™) with a resistivity of 18 MΩ cm was used to prepare the solutions, and the procedures were conducted at room temperature (RT, 23 ± 2 °C).

2.2. Synthesis of Bi₂S₃ nanohybrids

The Bi₂S₃ nanoparticles were synthesized using an aqueous route in a reaction flask at room temperature. The synthesis of the Bi₂S₃ nanoparticles was carried out as follows: 2 mL of a chitosan solution (1% w/v in 2% v/v aqueous solution of acetic acid) or an O-carboxymethyl chitosan solution (1% w/v in water) and 45 mL of DI water were added into the flask reacting vessel. Under moderate magnetic stirring, 3 mL of the S²⁻ precursor solution (Na₂S·9H₂O, 1.0 × 10⁻² mol L⁻¹) and 2 mL of the Bi³⁺ precursor solution (BiCl₃, 1.0 × 10⁻² mol L⁻¹ in acetic acid) were added to the flask and stirred for 60 min. During the addition of Bi³⁺ solution, the pH was measured and adjusted to 3.0 ± 0.5 or 10.0 ± 0.5 with NaOH (1.0 mol L⁻¹). The Bi₂S₃ QD suspensions produced were referred to as “QD_ligand_pH”, where the “ligand” was CHI or CMC and the “pH” was 3.0 or 10.0, as a function of the capping ligand and the pH of the quantum dot synthesis process.

2.3. Physico-chemical characterization of the Bi₂S₃ nanohybrids

UV-Visible spectroscopy (UV-Vis) measurements were conducted using Perkin-Elmer equipment (Lambda EZ-210) in transmission mode in a quartz cuvette over a wavelength range from 1100 nm to 190 nm.

The morphological and structural features of the QDs were characterized by transmission electron microscopy (TEM, Tecnai G2-20-FEI microscope, 200 kV) and using the selected area electron diffraction (SAED) analysis. The QD size and distribution data were obtained from the TEM images by measuring at least 100 randomly selected nanoparticles using an image processing program freeware (ImageJ, version 1.50).³⁵

Dynamic Light Scattering (DLS) and Zeta Potential (ZP, ζ-potential) measurements were performed on a colloidal suspension of QD in water using a ZetaPlus instrument by applying the laser light diffusion method (Brookhaven Instruments).

The photoluminescence (PL) characterization of the nanohybrids was conducted based on the spectra acquired at room temperature using a violet diode laser module at λ  _excitation = 405 nm (150 mW, Roithner LaserTechnik, GmbH) coupled to a USB4000 VIS-NIR spectrophotometer (Ocean Optics). Additionally, QD colloidal media samples were placed inside a “darkroom-chamber” where they were illuminated by using a UV radiation emission bulb (λ  _excitation = 254 nm, 6 W, Boitton Instruments). Digital color images were collected when the QDs fluoresced in the visible range of the radiation spectra.

2.4. Cytotoxicity investigations

2.4.1. Cell culture

2.4.1.1. Human sarcoma cell line (SAOS cells)

The immortalized human osteosarcoma-derived (SAOS) cells were provided by Prof. A. Goes of the Department of Immunology and Biochemistry, UFMG. SAOS cells are broadly accepted as a permanent line of human osteoblast-like cells because they possess several osteoblastic features, they are highly proliferative, and they are a source of bone-related molecules. To be able to better predict the cytocompatibility of the tested samples, it is generally preferable to use cell lines with similar characteristics and phenotypes as the original tissues. All of the reagents for cell culture were supplied by Gibco BRL (NY, USA) unless otherwise noted. The SAOS cells were cultured in DMEM (Dulbecco's Modified Eagle's Medium) with 10% fetal bovine serum (FBS), streptomycin sulfate (10 mg mL⁻¹), penicillin G sodium (10 units mL⁻¹), and amphotericin-b (0.025 mg mL⁻¹) using a humidified atmosphere of 5% CO₂ at 37 °C. The cells were used for the experiments on passage 23.

2.4.1.2. Kidney cell line of a human embryo culture (HEK293T cells)

The human embryonic kidney cell line (HEK293T) was kindly provided by Prof. M.F Leite of the Department of Physiology and Biophysics, UFMG. The cells were cultured in DMEM with 10% FBS, penicillin G sodium (10 units mL⁻¹), streptomycin sulfate (10 mg mL⁻¹), and amphotericin-b (0.025 mg mL⁻¹) under a humidified atmosphere of 5% CO₂ at 37 °C. The HEK293T cells were used for experiments on passage 12.

2.4.2. Toxicity assay by MTT (3-(4,5-dimethylthiazol-2yl) 2,5-diphenyl tetrazolium bromide) assay

All of the biological tests were performed according to ISO standards 10993-5:1999 (Biological evaluation of medical devices; Part 5: tests for in vitro cytotoxicity). All experiments were performed using the direct contact methodology.

The SAOS and HEK293T cells were plated (3 × 10⁵ cells per well) in 96-well plates. The cell populations were synchronized in serum-free media for 24 h, after which the media volume was aspirated and replaced with media containing 10% FBS (Gibco BRL, NY, USA) for 24 h. The Bi₂S₃/polymer nanoconjugated samples were added to individual wells at a concentration of 3.0%. Controls had been used with cells and DMEM medium with 10% FBS, positive control Triton X-100 (1%, Sigma-Aldrich, St Louis, MO, USA) and as negative control chips of sterile polypropylene Eppendorf (1 mg mL⁻¹, Eppendorf, Hamburg, Germany). After 24 h, all media were aspirated and replaced with 60 μL of culture media containing serum to each well and photographed using an inverted optical microscope (Leica DMIL LED, Germany). MTT (5 mg mL⁻¹, Sigma-Aldrich, St Louis, MO, USA) was added to each well and incubated for 4 h in an oven at 37 °C and 5% CO₂. Next, it was placed in 40 μL SDS (Sigma-Aldrich, St Louis, MO, USA) and solution/4% HCl suitable for cell culture (Sigma-Aldrich, St Louis, MO, USA), and the cells were incubated for 16 h in an oven at 37 °C and 5% CO₂. Then, 100 μL were removed from each well and transferred to a 96-well plate. The absorbance was measured at 595 nm on a Thermo-Plate (TP-READER) with a 595 nm filter. Percentage cell viability was calculated to be 100% × (absorbance of the sample and cells)/(absorbance of the control). The values of the controls (wells with cells, and no samples) were set to 100% cell viability.

2.4.3. LIVE/DEAD® cell toxicity assay

The SAOS cells at passage 23 were plated (3 × 10⁵ cells per well) in 96-well plates. The cell populations were synchronized in serum-free medium for 24 h, after which the medium was aspirated and replaced with medium containing 10% FBS. The QD samples, QD_CHI_3.0, QD_CMC_{_}3.0, and QD_CMC_10.0 nanoconjugates, were added to the cells at a concentration of 3% for 24 h. The reference controls were cells cultured in DMEM medium with 10% FBS. After 24 h, all media were aspirated, and the cells were washed two times with 10 mL of phosphate buffered saline (PBS, Gibco BRL, NY, USA). The SAOS cells were treated with the LIVE/DEAD® Viability/cytotoxicity kit (Life Technologies of Brazil Ltd, São Paulo, Brazil) for 30 min, according to the manufacturer's specifications. Images were obtained with an inverted optical microscope (Leica DMIL LED, Germany), and the fluorescent emissions were separately acquired, calcein at 530 ± 12.5 nm, and EthD-1 (ethidium homodimer-1) at 645 ± 20 nm.

2.4.4. Statistical analysis

There are many methods for multiple-comparisons (e.g., Tukey's, Bonferroni's, Dunnett's, Scheffé's) as a follow-up to ANOVA (analysis of variance). No single method of multiple comparisons is uniformly best among all the methods. The Bonferroni method allows many comparison statements to be made (or confidence intervals to be constructed) while still assuring an overall confidence coefficient is maintained. Compared to others, it is very easy, widely applicable, as well as valid for equal and unequal sample sizes.³⁶ In this study, Prism software (GraphPad Software, San Diego, CA, USA) was used for data analysis. Statistical significance was tested using One-way ANOVA followed by Bonferroni's method. A p value < 0.05 was considered statistically significant. The experiments were performed in triplicate (n = 3). Bonferroni's method was applied to compare the means of treatments with all of the samples (a pairwise comparison) using two cell types, the Human sarcoma cell line (SAOS cells) and Kidney cell line of a human embryo culture (HEK 293 T cells).

3. Results and discussion

3.1. Physico-chemical characterization of Bi₂S₃ nanohybrids

The UV-Vis absorption spectra of Bi₂S₃ nanohybrids produced using chitosan (CHI) and O-carboxymethyl chitosan (CMC) as capping ligands are shown in Fig. 1A. The curves exhibit an absorbance onset at approximately λ = 800 nm, which is blue-shifted compared to that of bulk bismuth sulfide (Fig. 1A, arrow at λ = 932 nm).³⁷ The optical band gap (E  _QD) values were determined using the “Tauc relation”³⁸ extracted from the UV-Vis spectra (Fig. 1B). These values provided solid evidence that the Bi₂S₃ QDs were effectively stabilized in aqueous medium at room temperature because the band gap energies of the nanoparticles (E  _QD = 2.60 ± 0.2 eV for QD_CHI_3.0 and QD_CMC_10.0 and E  _QD = 2.82 ± 0.2 eV for QD_CMC_3.0) were significantly larger than that of bulk Bi₂S₃ (E  _g = 1.33 eV). The differences in the bandgap energy of the QDs and the original value of bulk Bi₂S₃ (i.e., ΔE = E  _QD − E  _g = E  _QD − 1.33 eV) > 1.2 eV, known as ‘blue shift’) are strong confirmation of the large quantum confinement of the exciton (h⁺/e⁻) due to the ultra-small size of the nanocrystals.

Fig. 1C shows the photoluminescence spectra of the nanohybrids. An emission band in the blue-orange region was detected for the Bi₂S₃ QDs and it is associated with radiative recombination from the trap states and electronic transitions from Bi³⁺(³P₁ → ¹S₀).³⁹  Fig. 1D shows typical images of the green fluorescence of Bi₂S₃ QDs capped with polysaccharides, chitosan or CMC, upon exposure to ultraviolet radiation (“darkroom-chamber”, light source λ  _excitation = 254 nm) compared with the fluorescence of the nanohybrid colloidal solution under natural light illumination.

Based on TEM images (Fig. 2A), Bi₂S₃ nanoparticles are reasonably dispersed with a spherical-like morphology. The QD nanoparticle sizes (i.e., diameter = 2R) were calculated to be 5.7 ± 0.4 nm, 8.0 ± 1.2 nm, and 3.0 ± 0.6 nm for QD_CHI_3.0, QD_CMC_3.0, and QD_CMC_10.0, respectively. These values are much smaller than the Bohr`s radius (a  _B ∼ 24 nm) for Bi₂S₃,²⁴ validating the results of UV-Vis spectroscopy showing the “quantum confinement regime” of the Bi₂S₃ nanocrystals. The SAED pattern (Fig. 2B) shows the crystalline nature of the synthesized QDs and revealed lattice fringes with an interplanar distance of approximately 0.34 ± 0.01 nm that can be assigned to the (111) plane of the orthorhombic structure (ICCD 43-1471) obtained by X-ray diffraction (Fig. 2C).

From the DLS results, it can be observed that the samples produced at pH = 3.0 with chitosan (QD_CHI_3.0) and CMC (QD_CMC_3.0) presented smaller hydrodynamic diameter (H  _D) values of 25 nm and 85 nm, respectively, than the hybrids prepared at pH = 10 using CMC ligands with H  _D = 185 nm. In addition, zeta potential measurements indicated that under acidic conditions (pH = 3.0) both systems, QD_CHI_3.0 and QD_CMC_3.0, displayed positively charged ζ = +47 mV and ζ = +20 mV structures, respectively, and at alkaline medium (pH = 10.0), the QD_CMC_10.0 nanohybrids displayed negative net charge ζ = −177 mV. The results of the DLS and ZP measurements are summarized in Fig. 3, in which the core–shell nanostructures are schematically depicted.

3.2. Toxicity assay by MTT

In the present study, key processing variables were selected to investigate the cytotoxicity of the nanohybrids made of Bi₂S₃ (core) and polysaccharides (shell), which are: (a) chitosan and CMC as biofunctional ligands; (b) the pH of the synthesis (acidic and alkaline); (c) and two types of cultured cells – SAOS and HEK293T. These parameters are expected to have an important effect on all of the morphological and physico-chemical properties, as well as the biological responses of the systems.

Here, the cytotoxicity of core–shell Bi₂S₃/chitosan and Bi₂S₃/CMC nanoconjugates was assessed using the enzyme-based MTT assay with samples synthesized under acidic (i.e., pH = 3.0, QD_CHI_3.0 and QD_CMC_3.0) and alkaline conditions (i.e., pH = 10.0, QD_CMC_10.0). In addition, two cell lines were used, SAOS and HEK293T cells. Briefly, the enzyme-based MTT method relies on a reductive, colored reagent and the presence of dehydrogenase in the viable cells to determine their viability based on mitochondrial function using a colorimetric method. This method is far superior to other similar methods because it is easy-to-use, safe, highly reproducible and widely used in both cell viability and cytotoxicity tests.

When analyzing the results of SAOS cells in contact with all the nanohybrid samples, no significant differences in the cell viability and non-toxic effects compared to the control group were detected within the statistical range of variation (Fig. 4). In addition, no alterations in the morphological and shape features of the SAOS cells were observed, presenting 100% cell confluence. In an initial analysis, as expected, they all have presented similar cell viability values of over 90% (i.e., non-cytotoxic), when considering the main concept of “safe by design” used in this research. Thus, these nanohybrids were designed and synthesized with a heavy metal-free QD core (Bi₂S₃) associated with the polysaccharide biocompatible shell (chitosan or CMC) via an eco-friendly aqueous processing route. Hence, based on the results, it can be assumed that the changes of the ligands using chitosan or CMC and the differences in the pH from acidic to alkaline conditions, did not show any evidence of having affected the overall biocompatibility and cytotoxicity of the novel nanohybrids toward the SAOS cell line.

Analogously, the MTT assay was performed with HEK293T cell culture and the results are shown in Fig. 5. Although all of the samples presented cell viability responses above 65%, indicating the non-toxic characteristics of the systems, the results clearly indicated that there were very significant differences regarding the parameters of the synthesis and the structure of the nanohybrids, which deserve a more in-depth analysis. It is broadly recognized that cellular interactions with nanoparticles and their uptake utilize rather complex mechanisms, which are usually governed by the size and surface characteristics, such as hydrophobicity, hydrophilicity and charges.⁴⁰ Here, it can be observed that the physico-chemical and morphological characteristics of the nanohybrids considerably affected the cell viability results of the HEK293T cells in the MTT assays. The most striking effect was observed with the chitosan-based nanoconjugates that were synthesized at pH = 3.0 (QD_CHI_3.0), which produced an estimated decrease in cell viability of approximately 33% compared with the control. Next, the system with CMC at pH = 3.0 (QD_CMC_3.0) showed a similar reduction of approximately 30%. On the other hand, remarkably, no significant reduction (i.e., no statistical difference) in the cell viability results (∼90% cell viability) was observed for the system prepared in alkaline medium (pH = 10.0) using CMC as the capping ligand (QD_CMC_10.0) compared with the control reference. In order to enlighten these findings, it is vital to rely on the physico-chemical features evaluated by the DLS method and ζ-potential analysis (Fig. 3). In that sense, this notable difference in the dimensions of the core–shell nanostructures may be accountable for the reduction in the cell viability response of the QD_CHI_3.0 samples, as similar effects have been widely reported for other polymer-based nanoparticles.^41,42 In general, typically for nanoparticles with sizes <100 nm, the smaller the nanoparticulate system, the higher is the cellular uptake and, therefore, the possibility of cell endocytosis mechanisms.^43,44 Thus, in this study, the drastic increase in the H  _D values from 25 nm for QD_CHI_3.0 to 185 nm for QD_CMC_10.0 (Fig. 3) was mostly attributed to the sum of the effects of the electrostatic charges and the presence of chemical functional groups in the polymer ligands, which are pH-dependent. On the other hand, no clear relationship was observed for the sizes of the Bi₂S₃ nanocrystal core (5.7 nm and 8.0 nm, respectively) with the H  _D values. Therefore, chitosan, a well-known cationic biopolymer, is fully protonated at pH = 3.0 due to the presence of amine groups in the polymer chain.⁴² Analogously, under acidic conditions (pH = 3.0), CMC also has fully protonated amine groups, but the presence of the inserted hydrophilic carboxyl groups in the polymer backbone favored further interactions with water molecules within the organic shell, leading to an increase of the colloidal volume of solvation (i.e., H  _D), as determined by DLS. This same effect was observed to a greater extent on the QD_CMC_10.0 samples, which had the highest H  _D of 125 nm due to the formation of negative carboxylate species at pH = 10, further improving the solvation of the water molecules in the charged polymer chain. Using the same rational approach, the overall balance of charges in each system was different, as the measurements of ζ-potential analysis (Fig. 3) have noticeably shown. As discussed above, the ζ-potential results are attributed to the chemical balance of protonation/deprotonation of chemical species in the polymer chains, where at pH = 3.0 amines are protonated (NH₂ + H⁺ → NH₃  ⁺) in chitosan and CMC, resulting in overall net positively charged nanohybrids. On the other hand, under alkaline conditions (at pH = 10.0, chitosan is insoluble in water at pH > 6.5), the CMC-based nanoconjugates display the opposite trend, causing the deprotonation of amines (NH₃  ⁺ → NH₂ + H⁺) and formation of carboxylates (COOH → COO⁻ + H⁺), leading to net negatively charged colloidal nanostructures. Based on these results, it is suggested that, despite the non-cytotoxic response of the systems for HEK293T cells, the relative reduction of cell viability was caused by the combination of both factors, dimension and charges, as it is supported by the literature on the theme. Although positive charges on the nanomaterials appear to improve the efficacy of imaging, gene transfer, and drug delivery, increased cytotoxicity of such constructs has been reported.^40,45 Cationic nanoparticles can cause a more pronounced disruption of plasma-membrane integrity, greater mitochondrial and lysosomal damage than anionic nanoparticles.⁴⁰ Moreover, regarding cytotoxic action, both charge density and charge polarity play a relevant role. Cationic polymers bind to the negatively charged plasma membrane of the target cells to a higher degree than negatively charged or neutral molecules, which may cause focal dissolution of the plasma membrane, hole formation and perturbation of the internal membrane structure.^40,45 However, in this study, the differences in the cell viability results and, therefore, the relative cytotoxicity of the nanohybrids for the two cell lines, SAOS and HEK293T, are not sufficiently supported by only considering the physico-chemical characteristics, i.e., the hydrodynamic sizes and zeta potential values. Our results indicate that the polymer ligands and size of nanohybrids as well as the target cell type are critical determinants of the degree of cell viability and potentially triggering mechanisms of toxicity. Undeniably, despite the intensive research efforts, the published reports on the cellular responses to nanomaterials are very often inconsistent and even contradictory. Moreover, the relationships between the responding cell type and nanomaterial properties are not well understood when cytotoxicity is monitored through a reduction in mitochondrial activity.⁴⁴ Therefore, it is necessary to understand how different cells respond to nanomaterials, as they can induce cell type-specific responses, resulting in variable toxicity and subsequent changes in the cell fate, based on the type of exposed cell.⁴⁴ Nevertheless, it would be over simplistic to explain the observed differences by only the surface charges and particle sizes without attempting to discuss the specificities of each cell line used to perform the cytotoxicity analysis. Thus, from the cell-type perspective, it can be affirmed that both the SAOS and HEK293T cell lines share some similarities, as they are widely used as cell models, making them a common choice of cultures for biological research. In addition, compared with normal cells from adult tissues, embryonic (HEK293) and tumor (SAOS) cells are more easily cultured because they have a higher growth capability and adapt more readily to variations in external factors, making them suitable for pharmaceutical and biomedical research.^46,47 In particular, HEK293T is a permanent cell line established from primary embryonic human kidney, which is very useful for transfection experiments, as it has a higher transfection efficiency than other cell lines. However, it is not a good model for typical human cells or other unaltered mammalian cells. Thus, based on the results of our study, it is conceivable to suggest that due to the permeability of the membranes of HEK293T cells, they are more susceptible to be affected by the physico-chemical characteristics of the nanohybrids (i.e., charge and size) than the SAOS cells. This assumption is supported by the published literature of similar systems and studies conducted with HEK293T cells, where cytotoxic behavior toward some compounds was caused by disruption of the cell membrane and/or membrane pore formation.⁴⁸ In contrast, SAOS is a cell line derived from the primary osteosarcoma cancerous bone tumor, which assigns high chemical and biological resistance, making it less susceptible to the toxic effects of compounds.⁴⁹ A schematic representation of the interactions of the QD/polysaccharide nanohybrids with different surface charges and the HEK293T cells at the biointerfaces is depicted in Fig. 5 (top inset). It should be highlighted that it is a simplified approach of the complex phenomena involving the nanomaterials and cell biointerfaces and is a preliminary attempt to initially clarify the differences observed in the cell viability results of the SAOS and HEK293T cells in the MTT assay.

3.3. LIVE/DEAD® cell toxicity assay

In the current study, the LIVE/DEAD® assay was used to validate the toxicity results of cell viability assay by MTT in SAOS cells. To avoid redundancy, the LIVE/DEAD® test was performed only with SAOS cells and the cell count was made using freely available image processing software (ImageJ, v1.50). Essentially, LIVE/DEAD® cytotoxicity assay provides a cell viability assay of two color fluorescence, which is based on the simultaneous determination of live cells (stained in green) and dead cells (stained in red). Two probes are used, calcein AM that penetrates the cell membrane of living cells and an ethidium homodimer (EthD-1) that can only pass through the disrupted membrane areas of dead cells. Since both Calcein and EthD-III dyes can be excited at λ = 490 nm, a simultaneous monitoring of viable and dead cells is possible by fluorescence microscopy. Hence, it can be observed in the images presented in Fig. 6 that the cells in contact with the QD_CHI_3.0, QD_CMC_3.0 and QD_CMC_10.0 nanohybrid samples had fluorescence patterns similar to the control group, i.e., high green fluorescence (viable cells) and little or no red fluorescence (dead cells). Therefore, these results of the LIVE/DEAD® assay qualitatively and quantitatively confirm the results showed in cell viability assay by MTT without toxicity response toward SAOS cell-type. According to the outstanding results obtained from the in vitro cytocompatibility assays, it is suggested that the nanohybrids made of Bi₂S₃ and polysaccharides (chitosan or CMC) can offer a promising platform for future biomedical applications, such as fluorescent nanoprobes for diagnostic and bioimaging studies.

Certainly, future studies are needed to further investigate the intriguing aspects involving the toxicity due to the interactions and specificities of nanomaterials with cells. The results demonstrated conclusively that the nanohybrids produced in the present research using chitosan and CMC have favorable characteristics as not cytotoxic, nonetheless the overall cell viability responses are cell-line and physico-chemical dependent.

4. Conclusion

Novel core–shell nanohybrids were produced using a “safe by design” approach using chitosan and carboxymethyl chitosan as the organic biofunctional shell and Bi₂S₃ semiconductor quantum dots as the fluorescent inorganic core. In addition, these nanohybrids were synthesized using a one-step, environmentally friendly process in aqueous colloidal medium at room temperature. The results clearly demonstrated that the pH of the solution, from acidic to alkaline conditions, and the polysaccharide capping ligand (CHI or CMC) used during the synthesis significantly affected the physico-chemical and morphological characteristics of the Bi₂S₃/polymer hybrids. The average hydrodynamic diameters of the nanohybrids assessed by DLS measurements varied from the smallest value of 25 nm under acidic conditions (pH = 3.0) with chitosan to the highest value of 185 nm when the nanohybrids were produced under alkaline (pH = 10.0) conditions with CMC. Thus, in this study, it is suggested that the pH-dependent chemical functionalities in the CHI and CMC polymer chains, mainly the amino and carboxylic groups, have played a major role in the mechanism of stabilization of the Bi₂S₃ nanoparticles in aqueous colloidal dispersions via electrostatic charge interactions. Because of the overall contribution of the nanohybrid sizes and net surface charges, either positive at low pH or negative at high pH, the results of the in vitro cell viability assays evidenced clear cell-type dependence with distinct cytotoxicity responses of the SAOS and HEK293T cells. Nonetheless, all the nanohybrid systems investigated were found non-cytotoxic for both cell cultures under the evaluated conditions. Finally, these nanohybrids were photoluminescent, water-soluble and “heavy metal-free”, which may offer a broad range of possibilities to be exploited as narrow-bandgap fluorescent bioprobes for biomedicine and as nanoplatforms for eco-friendly applications.

Conflict of interest

The authors declare that they have no conflict of interest.

Acknowledgements

The authors acknowledge the financial support from the following Brazilian research agencies: CAPES – Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (PROEX-433/2010; PNPD), FAPEMIG – Fundação de Amparo à Pesquisa do Estado de Minas Gerais (PPM-00202-13;BCN-TEC 30030/12), CNPq – Conselho Nacional de Pesquisa (PQ1B-306306/2014-0; UNIVERSAL-457537/2014-0), and FINEP – Financiadora de Estudos e Projetos (CTINFRA-PROINFRA 2008/2010). The authors express their gratitude to the staff at the Microscopy Center at UFMG for their help with the TEM-EDX analysis.

References