Single step synthesized sulfur and nitrogen doped carbon nanodots from whey protein: nanoprobes for longterm cell tracking crossing the barrier of photo-toxicity

Abstract

Long-term cell tracking is a research interest for biological scientists across disciplines and applications. However, long-term cell tracking experiments are often limited due to photobleaching and phototoxicity. In the current study, a carbonaceous nanoprobe was developed using a single step microwave assisted degradation of whey protein in the aqueous phase. The CNDs were characterized via UV-Vis spectroscopy, fluorescence spectroscopy, HRTEM, DLS and FTIR. Due to choice of the precursor, the CNDs were observed to be doped with sulfur and nitrogen. The CNDs were capable of bioimaging. In a 2D cell culture system (culture flask), the cells retained fluorescence for up to five passages. In a 3D microenvironment, cell tracking was also successful for up to 10 days. The CNDs were observed to be capable of scavenging superoxides and hydroxyl radicals in vitro. The CNDs were also observed to save cells from phototoxicity and UV exposure via cytotoxicity, microscopy and nanoindentation analysis.

---

1. Introduction

Cell tracking is a technology of interest over the past two decades for biological scientists. The biological study of molecular pathways, cellular differentiation, cancer biology, disease prognosis, developmental biology and many other diverse areas involve cell tracking.¹ With the discovery of time-lapse microscopy and advanced fluorescence microscopic techniques, cell biologists started getting into cellular differentiation and fate exploring experiments.² Over the previous decades, cell tracking experiments in vitro have provided the scientific community with immense information regarding multiple “missing links” and unanswered questions regarding biological science. It has unveiled mysteries related to hemogenic endothelium development,^3,4 cytokine signaling in differentiation,⁵ neural and hematopoietic stem cell differentiation patterns and the effect of cytoskeleton proteins upon these differentiations.^6–10

Fluorescence microscopy based live cell bioimaging has been observed to be one of the major steps regarding cell-tracking studies. However, fluorescence microscopy related biophysical research has continuously dealt with the problem related to photobleaching, the depth of signal acquisition and phototoxicity.¹¹ There are multiple commercially available bio-chemical probes and organic dyes that have been employed for cell tracking studies. However most of these agents are observed to lose their efficiency beyond a few generations owing to rapid photobleaching.^12,13

In recent years, different fluorescent probes have been explored as fluorescent fusion proteins and quantum dots. The fluorescent protein can be easily targeted for bio conjugation, organelle specific binding, and immuno-fluorescence such as green fluorescent protein (GFP) and red fluorescent protein (RFP). However, in long term imaging experiments, fluorescent proteins suffer from bleaching and low intensity. Moreover, in long-term studies, the proteins were observed to obtain oligomerise and aggregate resulting in reducing their efficiency.¹⁴

Owing to their high quantum yield and low bleaching properties, quantum dots were observed to be significantly suitable for long-term cellular imaging. However, most quantum dots are synthesized using toxic metals such as cadmium, lead, indium and arsenic. Recently, polymeric capping on bare semiconductor quantum dot crystals has been observed to reduce the immediate cytotoxicity in both in vitro and in vivo models.^15,16 However, heavy metal originated quantum dots have been observed to cause the generation of free radicals, resulting in cytotoxicity and developmental defects.^17,18

In long-term cell, tracking experiments, UV and visible light is applied to excite the fluorochrome tagged to the cell. This process results in the generation of reactive oxygen species (ROS) causing apoptosis and the mutation of cells. This phototoxicity of cells is amplified and unavoidable during multiple imaging studies during prolonged culture. As it has been always mentioned by biophysicists, an image of a healthy cell captured at poor resolution is always preferable than a high-resolution image of a dying cell.³ Therefore, this drawback has to be minimized without compromising the quality of the experiments. In addition, components of the cell culture media and intracellular compounds are also reported to generate ROS in the presence of UV and visible light.^19,20 Therefore, this multiple ROS generation process needs to be controlled and mitigated as well.

In recent times, carbon nanodots (CNDs) have been reported in the literature as novel alternative nanoprobes for bio imaging. CNDs are ultra-small nanostructures with C–C linkages all over the backbone and have significant photoluminescence properties.

Initially, CNDs were synthesized via organic phase refluxing, GPC, HPLC and long hour autoclaving from precursors, such as graphite, carbon soot, carbon nanotube, and organic molecules such as sucrose and synthetic polymers.^21–25 Recent publications on CNDs based on microwave irradiation and the application of biomass as a precursor have also been observed to be lucrative for the mass scale production of CNDs.²⁶ Orange juice, soymilk, willow bark, amino acids, ground coffee and many other sources of biomass have been extensively used as a precursor for the synthesis of CNDs.^27–31 CNDs have been reported to have antioxidative properties both in ex vivo and in vitro systems.^32,33 Therefore, CNDs can be utilized for long-term cell tracking studies and minimize phototoxicity. Doping of CNDs with different elements has been shown to improve their properties such as quantum yield, surface functionalization and catalysis. Due to doping, changes or stresses in the lattice of the carbon dots and changes in the electronic arrangements mostly evoke such properties. This can create electron hole pairs in the lattice, which can improve the properties of the nanostructure. Furthermore, CNDs due to doping can become more useful for bioconjugation and metal sensing. There are significant reports on the doping of CNDs with nitrogen, boron, phosphorus, and sulfur.^34–36

Whey protein is a component of milk, which is a by-product of the cheese industry. It is rich in proteins like beta- and alpha-lactoglobulin and all the essential amino acids.⁴² Whey protein has been utilized for making films for biomedical applications. In vivo studies show whey protein not only to be biocompatible but also non-immunogenic as a biomaterial.³⁷ More importantly, the major proteins of whey protein are rich in amino acids especially in cysteine and methionine. Therefore, this precursor can be useful in the single step doping of nitrogen and sulfur into carbon nanodots.

In the current study, a microwave-assisted synthesis of CNDs from whey protein is reported. The study not only incorporates the bioimaging and long term cell tracking capability of CNDs, but also explores the therapeutic application of CNDs.

2. Materials and methods

All experimental chemicals were used as obtained from Merck, India and used without further purification unless mentioned. Deionized water (Milli-Q 18.2 MΩ cm) was utilized in all the reactions. Whey protein was obtained from Myprotein, UK. 3-[4,5-Dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) salt and nitroblue tetrazolium (NBT) were obtained from SRL chemicals, Mumbai. DCF-DA and poly-L-lysine solution were obtained from Sigma-Aldrich. Modified essential media-alpha (MEM-α), and fetal bovine serum (FBS) were obtained from Gibco and an antibiotic solution obtained from HiMedia was utilized in the cell culture studies.

2.1 CND synthesis

A 0.25% (wt/vol) whey protein solution was prepared in deionized (DI) water and the pH was increased to 12 by adding 1 M sodium hydroxide (NaOH) solution. It was immediately transferred to a domestic microwave (1000 W, 2.45 GHz). The time of irradiation was optimized via irradiating for different time intervals (∼2 minutes in 30 s intervals). The fluorescence intensity measurement was carried out in a plate type fluorimeter (BioRad). The synthesized CNDs were filtered using a 0.2 μM filter paper, neutralized using 1 M HCl solution and finally dialyzed against DI water using a membrane with a cut-off of 12–14 kDa (Hi Media) overnight. The obtained solution was dried in a vacuum oven and the obtained powder was used for all the biological experiments suspended in PBS (concentration 100 μg mL⁻¹).

2.2 Characterization

The CNDs were characterized using UV-Vis spectroscopy (Shimadzu), fluorescence spectroscopy (Varian Cary), X-ray diffraction (Bruker, India, CuKα – 0.154 nm), high resolution transmission electron microscopy (HRTEM, JEM-2100, JEOL, 200 kV), field emission scanning electron microscopy (FESEM, Carl Zeiss, 30 kV), EDAX analysis (Oxford Instruments), Raman spectroscopy (Jobin Yvon Horiba), Fourier transform infra-red spectroscopy (FTIR, Perkin Elmer). A Malvern Zetasizer machine was used to perform dynamic light scattering (DLS) analysis and zeta potential measurements. The fluorescence quantum yield was measured with respect to quinine sulfate dissolved in 0.1 M H₂SO₄ (1 mg mL⁻¹). Using same UV-Vis spectroscopy and fluorescence spectroscopy parameters (same Laser, slit number and cuvette path length) a CND suspension at a similar concentration was used for quantum yield measurement. A single point measurement was performed using the protocol described by Jana et al. (2015).³⁸ The formula used for measuring the quantum yield was as follows:

Q_ysample = Q_yref(F_sample/F_ref)(η_sample²/η_ref²)(A_ref/A_sample)

where Q_y is quantum yield. F represents the fluorescence intensity. The refractive index of the solvents is represented by η. A is absorbance of the suspensions. Herein, the sample and the reference are the CND suspension and quinine sulfate solution of same concentration (1 mg mL⁻¹), respectively.

2.3 Cell culture study

Wharton's jelly derived mesenchymal stem cells (MSC) (isolated under the Institute's Ethical Clearance from Indian Institute of Technology, Kharagpur and Midnapore Medical College, India) was utilized for the cell culture studies. Umbilical cord was collected in a sterile environment from pregnant women delivering babies from the hospital. Before collection, written consent was obtained from the subject. Furthermore, the cell isolation from Wharton's jelly was conducted in a sterile environment in the laboratory. A Carl Zeiss inverted fluorescence microscope was utilized for microscopy. Fluorescence microscopy was performed by culturing 10⁴ cells on poly-L-lysine coated coverslips. After 72 h of culturing, cells were fixed using 4% paraformaldehyde solution and incubated with the CND solution for 30 minutes. Samples were washed with PBS and then counterstained with 4′,6-diamidino-2-phenylindole (DAPI, Life Technologies) and imaging was performed using an inverted fluorescence microscope (Carl Zeiss). Two optical filters of excitation (365 nm band pass 25 nm and 470 nm band pass 50 nm) and emission (450 nm band pass 25 nm and 540 nm band pass 25 nm) were used for capturing the images. Along with fluorescence imaging, differential interference contrast (DIC) mode image was captured and merged with the fluorescent micrograph.

The semi-quantitative cytocompatibility of the cells treated with CNDs was determined via an MTT assay.³⁹ Different increasing concentrations of CNDs were used in the study. Samples were incubated with MTT solution in PBS (5 mg mL⁻¹) after day 1 and 3 and in triplicate (n = 3). The mean and standard deviation were calculated for reporting. After 4 h of incubation, the solution was discarded following washing in PBS. Finally, the samples were incubated with DMSO (HiMedia) and after 30 minutes, the optical density was measured at 590 nm in a plate reader (RMS healthcare system).

2.4 Cell tracking study

Cell-tracking studies were carried out via two different methods. Initially, a 5% (vol/vol) CND suspension was prepared in cell culture media and the pH was adjusted. Furthermore, the MSCs were cultured in CND containing media for one generation. The consecutive passages (each passage for 5 days) were maintained in CND free cell culture media and cells were imaged after each passage. The cell tracking study requires not only imaging capability in a 2D culture system but also in 3D. Therefore, sintered 3D micro-patterned alumina scaffolds (non-bioactive surface) were fabricated and seeded with the CND tracked cells. The scaffolds were washed and fixed using 4% paraformaldehyde solution on day 3, 5 and 10. Furthermore, scaffolds were counter-stained with DAPI and fluorescence microscopic imaging was performed. The imaging was carried out in a Carl Zeiss inverted fluorescence microscope. For the 3D samples, images were captured in different Z plane (height) and finally stacked to a merged micrograph via Z-stacking using Zen-2012 software (Carl Zeiss). In most of the figures, the optimal section depth was taken to be 4.5 μm.

2.5 In vitro free radical scavenging properties

The in vitro superoxide scavenging properties of the prepared CNDs were studied via an NBT assay.³³ For the NBT assay, the MSCs were seeded on poly-L-lysine coated coverslips in a petridish at a concentration of 10⁴ cells per sample. In each sample, 100 μL of 6 mM NBT solution (dissolved in media) was added. The samples were protected from direct white light. The CND suspension was added after 24 h of culturing. After 72 h, the media was discarded and samples were washed with sterile PBS. The samples were visualized under a Carl Zeiss inverted microscope in bright field mode using a color camera. Furthermore, cells were lysed using DMSO and 1 M KOH solution and the optical density was measured at 610 nm for quantitative measurement. Different concentrations of L-ascorbic acid (1 mg mL⁻¹ stock solution) were used as a standard. All the samples were conducted in triplicate (n = 3) and mean and standard deviation determined.

The in vitro hydroxyl radical scavenging properties were investigated using a di-chloro di-hydrofuran fluorescein di-acetate (DCFH-DA) assay.⁴⁰ The MSCs were cultured in poly-L-lysine coated cell culture plates at a concentration of 10⁴ cells per well. Different concentrations of CNDs and L-ascorbic acid were added and incubated for 72 h. All the samples were kept in triplicate for statistical significance. The mean and standard deviation were measured for further reporting. Furthermore, the media was discarded and the samples were washed with sterile PBS. The samples were incubated with 1 mM methanolic DCFH-DA for 1 h followed by measurement of the fluorescence intensity with excitation at 485 nm and emission at 530 nm in a fluorimeter (Biorad).

2.6 Phototoxicity assessment

Phototoxicity assessment of the cells treated with CNDs was performed by culturing 10⁴ cells per mL using spectroscopy (MTT Assay), microscopy (ethidium bromide staining) and cellular biomechanics study (nanoindentation). All the experiments were performed for 72 h. In all cases, the samples were exposed to 1 h of UV irradiation followed by 24 h of dark over two consecutive days before analysis.

The semi-quantitative cytocompatibility of the cells exposed to UV was determined via an MTT assay.³³ Different increasing concentrations of the CNDs were used in the study. Samples were incubated with an MTT solution in PBS (5 mg mL⁻¹) after day 1 and 3. After 4 h of incubation, the solution was discarded followed by washing with PBS. Finally, the samples were incubated with DMSO (HiMedia) and after 30 minutes, the optical density was measured at 590 nm in plate reader (RMS healthcare system).

The phototoxicity and apoptosis rate of the CND treated cells with respect to the control were also measured via microscopic analysis using ethidium bromide staining. Ethidium bromide was used to selectively stain dead cells. One plate was incubated with 5% CND suspension in cell culture media, whereas the other was kept as a control with CND-free media overnight. Then, the samples were subjected to UV exposure. Furthermore, media was discarded and the samples were incubated with 1 mM ethidium bromide (SRL Chemicals, Mumbai) dissolved in PBS. The samples were washed with PBS and observed under a microscope both in DIC and fluorescence mode (excitation 545 nm and emission 610 nm).

The cell membrane stiffness is a significant marker of oxidative stress induced oxidation of the membrane lipids and related apoptosis and was measured via nano-indentation⁴¹ using a nano indentation system TI 950 TriboIndenter, Hysitron Inc., USA. Post CND treatment and UV exposure, the 72 h cells were fixed using 4% paraformaldehyde solution after which the samples were dried and used for nano-indentation. Ten different spots were used for each sample during the nano indentation study. The study was conducted in constant displacement mode with a displacement of 100 nm.

3. Results and discussions

3.1 Synthesis of the CNDs

Whey protein is a milk-isolated product formed when making cheese, which is rich in globular proteins such as the lactoglobulins. Therefore, whey protein can be considered as a rich source of amino acids, which is already explored in the literature as a source for synthesizing CNDs.²⁸ It can also incorporate nitrogen and sulfur into the carbon backbone, which can enhance the properties of the synthesized CNDs. A 0.25% solution of whey protein in deionized water was prepared. The pH of the solution was increased to around 12 by adding 1 M NaOH solution. The solution was immediately transferred to a domestic microwave (1000 W, 2.45 GHz) and irradiated for different time durations with 30 s interval differences. The dielectric heating of microwave can induce stress at the CND precursor chain and can degrade the precursor molecule to ultra-small CND formation. Whey protein also has a tendency for gelation at high temperature.³⁹ Therefore, an optimization was needed to find a balance between the two reactions, as gelation might cause precipitation of the precursor and alter the formation of the CNDs. Therefore, the irradiation time was optimized for CND synthesis by irradiating the samples at different time intervals. The results were evaluated via both visual inspection and fluorescence intensity measurements.

As displayed in the Fig. 1a, the fluorescence intensity of the samples was observed to increase upon increasing the irradiation time. The quantitative analysis via fluorometry, as represented in Fig. 1b, also followed a similar trend. As evaluated from Fig. 1, from 120 s of irradiation onwards the sample displayed formation of aggregates along with a decrease in fluorescence. Therefore, for the synthesis, the optimized time for irradiation was observed to be 90 s. The aggregate free CND suspension was isolated via filtration and used for further experiments. It was oven dried and for the cell culture study, a 100 μg mL⁻¹ CND suspension prepared in sterile phosphate buffered saline (PBS) was utilized as a stock suspension.

Fig. 1 (a) Visual inspection during the optimization of the irradiation time under both visible and optical light and (b) the fluorescence intensity of the samples post different irradiation times (n = 4).

3.2 Physico-chemical characterization

The synthesized CNDs were characterized via physico-chemical analysis tools to determine the CNDs' multiple properties for further exploration of their biological applications.

The UV-Vis spectra of the CNDs have certain specific characteristics. CNDs are mostly formed with the skeleton of graphite-like C–C linkages. In certain cases, even some unsaturations are also observed in the structures. Therefore, n–p* and p–p* linkages are characteristic of the CNDs. These specific bonds absorb light in the UV region and two specific notches at around 290 nm and 345 nm (Fig. 2a) are observed in the obtained spectra. Therefore, it is a preliminary confirmation for the synthesis of CNDs.

Fig. 2 (a) UV-Vis spectrum of the CNDs, (b) fluorescence spectra of the CNDs, (c), and (d) Raman spectra of CNDs.

The photoluminescence of CNDs is major characteristic and reason for the research behind CND synthesis; there are multiple theories behind the photoluminescence of CNDs such as quantum confinement, surface states absorption, and carbon core state formation,⁴⁰ although, the exact reasons behind these properties are not clear. However, for bio-labeling applications, the fluorescence properties need to be measured. In the current study, the fluorescence spectra of the CNDs are reported in Fig. 2b. The excitation maxima were observed to be around 395 nm. The excitation spectrum was observed to significantly sharp, which could be useful in bio-imaging applications. However, the emission spectrum has an emission maximum around 462 nm. A notch around 486 nm was observed in the emission spectra, which can be further investigated for the possibility of FRET (fluorescence resonance energy transfer). The donor and acceptor could be where different populations of CNDs with different size distribution present at energy transferrable molecular distance. The emission spectrum was comparatively flat and leptokurtic in nature with respect to the excitation spectrum. The quantum yield was observed to be around 0.57 with respect to quinine sulfate.

The Raman spectroscopy results were useful for confirming the formation of carbon nanodots and their surface groups (Fig. 2c). Two major peaks for the D and G bands of graphite-like carbon structures were observed at around 1315 cm⁻¹ and 1530 cm⁻¹, respectively. This is a characteristics signature for CNDs. Along with this, the intensification of the peaks around 1700 cm⁻¹ and 2210 cm⁻¹ proves formation or enhancement of C=N and C≡N linkages. This can be considered as evidence for the incorporation for nitrogen atoms in the carbon backbone. Similarly, the peaks near 2500 cm⁻¹ (SH) and 1250 cm⁻¹ (C=S) can also be considered as a signature for sulfur doping. The FTIR spectroscopy of the CNDs and whey protein was both conducted to find out the changes in both the backbone and side chains during the reaction to form the CNDs from whey protein. FTIR spectroscopy was used to analyze the vibrational modes of the surface groups (Fig. 2d). In the spectrum of whey protein, the major peaks were observed for C–H stretching (2938 cm⁻¹), S–H groups (2500 cm⁻¹), CO₂ (2343 cm⁻¹), unconjugated C=O (1834 cm⁻¹), N–H bending (1643 cm⁻¹), C–C aromatic stretching (1415 cm⁻¹), C–N stretching (1059 cm⁻¹), N–H wagging (925 cm⁻¹) and alkyl halides (595 cm⁻¹). The major vibrational states in whey protein were due to the different amino acids, their side chains, amino groups, carboxyl groups, disulfide linkages and aromatic amino acids. The alkyl halides' peak was observed due to the salting out of the protein. However, most of the mentioned peaks were observed in FTIR spectrum of the CNDs as well. In addition, certain peaks for C–C linkages were observed and designated to the formation of the CNDs. These peaks include C–C stretching (1567 cm⁻¹), C=CH₂ (875 cm⁻¹), S–H groups (2500 cm⁻¹), C–H rocking (704 cm⁻¹), and C–H out-of-plane bending (780 cm⁻¹). In the spectrum of the CNDs, the groups involved in C–C related linkages were found to be stronger.

High-resolution transmission electron microscopy of the CNDs was performed to understand the exact size and phase of the synthesized CNDs (Fig. 3a). The CNDs were observed as black tiny dots in the micrograph. The average size of the CNDS were observed to be around 5–7 nm. Nevertheless, certain aggregates were also observed, which may be due to aggregation of molecules of whey protein. The selective area electron diffraction (SAED) pattern displayed a ring-like pattern with bright spots, which proved the nano-crystallinity of the CNDs (Fig. 3b). The ring diameter also corresponded to the inter-crystal planar distance of the graphite-like carbon structure (JCPDS file no. 74-2330) with the lattice plane [002]. Furthermore, elemental analysis of the CNDs was carried out using nanoprobe-EDAX analysis (Fig. 3c). The spectra were collected directly from tiny black spots (CNDs) using the nanoprobe. The elemental analysis suggests incorporation of nitrogen and sulfur into the carbon nanostructures (4.9 and 0.3 weight%).

Fig. 3 (a) TEM bright field micrograph of the CNDs, (b) SAED pattern of the CNDs, (c) EDAX spectra obtained from the TEM nanoprobe, (d) XRD analysis of the CNDs, (e) FESEM micrograph of the CNDs, and (f) DLS analysis of the CNDs.

These results were further verified by XRD analysis (Fig. 3d). The nature of the peaks confirmed the crystalline nature of the sample. Moreover, the SAED results were compared with the XRD and significant similarities were observed. The major angles 18.366, 37.226, 42.394, 42.515, 42.873, 43.466, 47.996, 55.437, 57.671, 60.040 were observed, which match with the corresponding JCPDS file. However, as CNDs are different from graphite there are structural differences as well. Therefore, in certain cases, there were certain shifts in the angles from the original graphite angles. To compare with the TEM results, FESEM imaging of the CND film developed via drop-casting on glass slide was performed (Fig. 3e). The FESEM image confirms the formation of carbon nanostructures in the order of 10–15 nm with certain larger aggregates with sizes around 30 nm.

The zeta potential and DLS studies (Fig. 3f) were conducted to check the colloidal stability, surface charge and hydrodynamic radius of the samples. The zeta potential of the CNDs was observed to be around +13.2 mV. However, this result was attributed to the colloidal instability and a tendency to form aggregates. The presence of positive surface charge on the CNDs at the medium interface might be useful for electrostatic interactions with a negatively charged cell surface. The cell membrane is mostly negatively charged due to activity of ion channels. The DLS results showed that the average hydrodynamic radius of the CNDs was around 50 nm. This can be correlated to the poor zeta potential and possible aggregation of certain CNDs in the aqueous suspension. Therefore, the CNDs can enter inside a cell due to their small size even without bio-conjugation wherein the electrical charge gradient becomes the major driving force.

3.3 Bio-imaging study

CNDs being non-bleachable and significantly photoluminescent in nature have been extensively explored for bioimaging in the literature.³³ The bioimaging capability of the CNDs synthesized in the current study was also evaluated in a cellular system.

The image of the cells with CNDs (Fig. 4a) was observed to be bright, where CNDs were well distributed in the cytoplasm (as the nucleus is observed as unstained dark spots). The cell nucleus image obtained via staining with DAPI (Fig. 4b) was found not to overlap with Fig. 4a. The tendency of CNDs to be restricted within the cytoplasm decreased the possibility of mutagenesis. With the specificity of binding and bright luminescence as observed in Fig. 4, it can be concluded that the synthesized CNDs may be useful as a bio-labeling agent and as good as any commercial organic dye. Certain large aggregates were also visible, which were fluorescent in the image. Those aggregates were mostly attributed to the combination of dead cellular organelles and the CNDs.

Fig. 4 The fluorescent image of the MSCs after staining with (a) CNDs, (b) DAPI and (c) DIC mode. (d) The merged image of (a)–(c) and (e) the time and dose dependent cytotoxicity analysis (n = 3).

The time and dosage dependent cytotoxicity study was conducted to check the biological safety of the CNDs (Fig. 4e). With different dilutions of the CNDs in cell culture media the cells were grown, after the initial 24 h the cell death in the samples were observed to be more with respect to the control. This initial higher cell death could mostly be attributed to the change in osmolarity and the change in the local microenvironment of the cells. However, after a longer duration (72 h) in the study, the cells were observed to recover after the shock and grow more prevalently in samples with respect to the control.

3.4 Long-term cell tracking study

Long-term cell tracking is highly important for biological study in the field of developmental biology, cancer biology, stem cell biology and many other important fields. Therefore, the major objective of the current study was application of the synthesized CNDs for long-term cell tracking. Noteworthy, cell tracking should not only work in 2D culture systems, but also in the 3D cell culture niche. Therefore, this current study was evaluated in both the systems towards the objective.

Fig. 5a–f shows the fluorescence and DIC merged micrograph of the MSCs in 5 consecutive passages after incubation with the CNDs. In the figures, it was clearly visible that for 5 generations the cells retained the CNDs and the fluorescence signal coming from the cells were not bleached with time. Therefore, for long-term tracking the CNDs are efficient. Moreover, because a single passage took ∼5 days, the tracking can be performed for almost a month efficiently. However, the MSCs of passage 5 were used for starting the experiment so after five consecutive passages (originally which is the 10^th passage) the MSCs started losing their shape. Nevertheless, the fluorescence property was maintained, which is required for a long-term tracking study.

Fig. 5 The fluorescent images of the MSCs after treating with CND suspension mixed media in the 2D culture (culture flask): (a) passage 5, (b) passage 6, (c) passage 7, (d) passage 8, (e) passage 9 and (f) passage 10.

The cell tracking study also requires evaluation of the cellular growth and differentiation kinetics monitored in a 3D microenvironment. Especially, for tissue engineering and regenerative medicine, in vitro oncogenesis and microtissue development, the time-lapse fluorescence microscopy of a 3D microenvironment is essential.¹ Micropatterned sintered alumina scaffolds were used for this experiment to mimic a 3D microenvironment. The results are reported in Fig. 6a–f.

Fig. 6 (a, d and g) The fluorescence micrographs of the cells stained with CNDs in a 3D microenvironment at 3, 5 and 10 days. (b, e and h) The micrographs of same frames stained with DAPI. (c, f and i) The merged micrographs of the previous two staining experiments.

It was observed that the CNDs treated cells were able to grow in the 3D environment (3 days) grew along the patterns (5 days) and with the day of culture the cells formed a sheet-like morphology around the scaffold (10 day). The fluorescence signal was not bleached with time and was as bright as the nucleus counter stain, which was applied just before the imaging. Therefore, the CNDs are useful for long-term tracking in both 2D and 3D systems.

3.5 In vitro ROS scavenging study

As long-term cell tracking experiments are always affected by oxidative stress; therefore, the CNDs antioxidant properties in a cellular microenvironment were evaluated. The inhibition capability of the CNDs towards two major free radicals, hydroxyl radical and superoxide, was assessed. A standard antioxidant L-ascorbic acid was used as the control for comparison.

As the NBT post reaction with superoxide forms black precipitates (formazan), by counting the number of cells with dark black spots in Fig. 7a and b it can be clearly concluded that the number of dark cells were much more in Fig. 5a. Therefore, after applying the CNDs, a definite amount of superoxide was reduced. However, when comparing the performance of the CNDs to quench superoxide with respect to L-ascorbic acid via spectroscopy it was observed to be inferior in superoxide scavenging (Fig. 7c). The IC₅₀ for L-ascorbic acid was observed to be around 120 μL mL⁻¹ whereas the IC₅₀ for the CNDs was observed to be around 165 μL mL⁻¹. However, the scenario is completely reversed when hydroxyl radical scavenging was conducted via a DCFH-DA assay (Fig. 7d). In this assay, the IC₅₀ value for the CNDs was observed to be around 50 μL mL⁻¹ and the IC₅₀ for L-ascorbic acid was observed to be around 146 μL mL⁻¹. Therefore, it was clearly observed that the CNDs were more efficient for scavenging hydroxyl radicals with respect to superoxide radicals.

Fig. 7 NBT assay via microscopy for the (a) control and (b) CNDs. (c) The NBT assay via spectroscopy and (d) the DCFH-DA assay via spectroscopy.

3.6 Phototoxicity analysis

It was observed that the CNDs were capable of scavenging free radicals from a cellular microenvironment; however, the real problem of long-term cell tracking is known to be photo-bleaching and photo-toxicity.² The first is addressed in the case of the synthesized CNDs in the current study and previously discussed results. However, phototoxicity is considered as one of the major limitations for long term cellular monitoring via optical microscopy. Therefore, a set of experiments were designed in the present study to mimic a similar microenvironment via irradiating the control (without CND) and experimental (with CND) samples cells with UV light.

Ethidium bromide selectively stains dead cells among a population of a living and dead cells. From Fig. 8a–f it can be clearly observed post UV treatment, the number of cells undergone apoptosis was much more in the control sample with respect to the CND treated samples as red spotted cells were more abundant in the control. A similar result was observed in the MTT assay of cells after UV exposure (Fig. 8g). Upon increasing the CND concentration, the cell viability started to increase, whereas in the control sample (untreated), the viability was observed to be decreased. From the abovementioned results, it can be clearly observed that via both microscopy and semi quantitative cell viability assay, phototoxicity generated apoptosis was more prevalent in the control samples with respect to the CND treated samples. As phototoxicity produced free radicals oxidize cell membrane lipids weakening the cell membrane and causing cell death, the cell membrane biomechanics can be considered as a direct marker for the phototoxicity induced oxidative stress a cell has gone through. The mechanical properties of the cells were measured via an nanoindentation study (Fig. 8h). The hardness of the CND treated cell membrane was observed to be higher (0.3 GPa) with respect to the untreated cells (0.17 GPa). However, the cell membrane is an irregular surface; therefore, a huge standard deviation in the measurement was observed. However, the average hardness of the CND treated cells with respect to the untreated cells was greater, which can be considered as an effect of reduced membrane lipid oxidation.

Fig. 8 (a) Ethidium bromide stained control, (b) control in DIC mode, (c) control merged image, (d) ethidium bromide stained CND treated, (e) CND treated in DIC mode, (f) CND treated merged image, (g) cell viability post UV exposure and (h) nanoindentation study post UV exposure.

Cell tracking has been a great research interest over the previous decades for multiple biological fields. It includes regenerative medicine, cancer biology, developmental biology, stem cell research, and reproductive medicine.^2,43 Even the regenerative medicine researchers have also incorporated long-term cell tracking study in the in vivo tissue repair experiments.^44,47 The development of efficient instruments in terms of microscopy, camera and many other cell incubating system has improvised this system drastically. Researchers also developed multiple fluorescent dye and fusion proteins for this area of research.

However, photo-bleaching and photo-toxicity are known to be the major limitations to the results obtained in the study. Photo-bleaching causes weakening of the signals coming from the signal, which requires a longer duration of light exposure. This increases the phototoxic effects upon the cells. The free radicals generated due to photo-toxicity modify the membrane lipids via oxidation and can cause apoptosis.⁴⁶ It also reacts with the DNA, causes mutation and changes in gene expression.⁴⁵ All these phenomena not only induce stress in the cellular microenvironment but also cause changes in gene expression due to mutation causing the results obtained in developmental studies or stem cell tracking experiments to be erroneous. The CNDs reported in the current experiment are observed to be effective against this problem. A brief mechanism has also been proposed to show how CNDs protect cells against photo-toxicity (Fig. 9).

Fig. 9 The mechanism of CNDs to save cells from photo-toxicity.

As explained in our previous reports,³³ CNDs can protect cells from oxidative stress mostly via scavenging the free radicals generated. During optical and fluorescent microscopic experiments, light exposure induces reactive oxygen species in the cellular microenvironment. This commonly affects the DNA in the nucleus, cell membrane lipid rafts and many other organelles. The cell is often equipped with intracellular ROS scavengers such as superoxide dismutase. However, in many cases the amount of ROS generated cannot be mitigated via cellular antioxidants. In that context, the CNDs can take an efficient role. The free radicals mostly have singlet electrons, which are transferred to the side chain of the CNDs. These electrons either oxidize the side chain or can be transferred inside the CND core to react with the C–C linkages. In both cases, the electron is consumed to form an acidic ion (mostly HCO₃–), which is released into the media and can be taken care by the buffer present in it. Thus, the CNDs can consume free radicals and save cells from photo-toxicity in long-term tracking studies.

4. Conclusions

In the current study, a novel set of CNDs was synthesized from a newly explored biomass and characterized for their biological applications. The application of whey protein as a precursor helped the method to incorporate the synthesis and dual doping in a single step. Due to the single precursor, the possibility of uniformity in doping increases significantly. The doping can increase its application in the future for biosensing and bioconjugation studies. The CNDs were observed to be effective for bio-imaging studies. Furthermore, the CNDs were explored successfully for long-term cell tracking studies in both 2D and 3D cell culture systems. However, long photo-bleaching and phototoxic effects often limit long-term cell tracking studies. The cells treated with CNDs were observed to retain their fluorescence for long durations without being bleached. The CNDs were also observed to save the cells from photo-toxicity. Therefore, the synthesized doped CNDs may be useful for not only long-term cell tracking (almost a month), but also saving the cells in situ from phototoxic effects.

Acknowledgements

The authors acknowledge the MHRD, DRDO and DST, Govt of India for funding. The authors also acknowledge the central research facility and department of chemistry, Indian Institute of Technology, Kharagpur for research facilities.

References

1. B. Das, S. Dhara and P. Dadhich, Biomass based CND system, Indian Patent No: 439/KOL/2015, 2015.
2. D. L. Coutu and T. Schroeder, Cell. Sci., 2013, 126, 1–11.
3. T. Schroeder, Nat. Methods, 2011, S31–S35.
4. H. M. Eilken, S. Nishikawa and T. Schroeder, Nature, 2009, 457, 896–900.
5. M. A. Rieger, P. S. Hoppe, B. M. Smejkal, A. C. Eitelhuber and T. Schroeder, Science, 2009, 325, 217–218.
6. S. G. Megason and S. E. Fraser, Cell, 2007, 130, 784–795.
7. M. A. Rieger and T. Schroeder, J. Cell. Biochem., 2009, 108, 343–352.
8. M. Baker, Nature, 2010, 466, 1137–1140.
9. A. R. Cohen, F. L. A. Gomez, B. Roysam and M. Cayouette, Nat. Methods, 2010, 7, 213–218.
10. M. R. Costa, F. Ortega, M. S. Brill, R. Beckervordersandforth, C. Petrone, T. Schroeder, M. Gotz and B. Berninger, Development, 2011, 138, 1057–1068.
11. T. Schroeder, Nature, 2008, 453, 345–351.
12. Tracking cells through time and space. Bioprobe, 2011, vol. 66, pp. 30–32.
13. K. A. Muirhead, J. D. Tario and P. K. Wallace, Cell tracking with lipophilic membrane dyes, Spring, Biowire, 2011.
14. J. Wiedenmann, F. Oswald and G. U. Nienhaus, IUBMB Life, 2009, 61, 1029–1042.
15. H. S. Han, E. Niemeyer, Y. Huang, W. S. Kamoun, J. D. Martin, J. Bhaumik, Y. Chenb, S. Roberge, J. Cui, M. R. Martin, D. Fukumura, R. K. Jain, Mo G. Bawendi and D. G. Duda, Proc. Natl. Acad. Sci. U. S. A., 2015, 112, 1350.
16. M. Cao, L. Yu, P. Zhang, H. Xiong, Y. Jin, Y. Lu and L.-Q. Wang, Colloids Surf., B, 2013, 109, 154.
17. K. C. Nguyen, W. G. Willmore and A. F. Tayabali, Toxicology, 2013, 306, 114–123.
18. F. Ye, C. C. White, Y. Jin, X. Hu, S. Hayden, X. Zhang, X. Gao, T. J. Kavanagh and D. T. Chiu, Nanoscale, 2015, 7, 10085–10093.
19. J. W. Dobrucki, D. Feret and A. Noatynska, Biophys. J., 2007, 93, 1778–1786.
20. A. Grzelak, B. Z. Rychlik and G. Bartosz, Free Radical Biol. Med., 2001, 30, 1418–1425.
21. S. N. Baker and G. A. Baker, Angew. Chem., Int. Ed., 2010, 49, 6726–6744.
22. X. Xu, R. Ray, Y. Gu, H. J. Lehn, L. Gearheart, K. Raker and W. A. Scrivens, J. Am. Chem. Soc., 2004, 126, 12736–12737.
23. R. Zhang, Y. Liu, L. Yu, Z. Li and S. Sun, Nanotechnology, 2013, 24, 225601.
24. Y. Liu, C. Y. Liu and Z. Y. Zhang, Synthesis and surface photochemistry of graphitized Carbon quantum dots, J. Colloid Interface Sci., 2011, 356, 416–421.
25. S. Zhu, Q. Meng, L. Wang, J. Zhang, Y. Song, H. Jin, K. Zhang, H. Sun, H. Wang and B. Yang, Angew. Chem., Int. Ed., 2013, 125, 4045–4049.
26. X. Wang, K. Qu, B. Xu, J. Rena and X. Qu, J. Mater. Chem., 2011, 21, 2445–2450.
27. Z. Zhu, J. F. Zhai and S. J. Dong, Chem. Commun., 2011, 48, 9367–9369.
28. X. Y. Qin, W. B. Lu, A. M. Asiri, A. O. Al-Youbi and X. P. Sun, Catal. Sci. Technol., 2013, 3, 1027–1035.
29. S. Sahu, B. Behera, T. K. Maiti and S. Mohapatra, Chem. Commun., 2012, 48, 8835–8837.
30. J. Jiang, Y. He, S. Li and H. Cui, Chem. Commun., 2012, 48, 9634–9636.
31. P. C. Hsu, Z. Y. Shih, C. H. Lee and H. T. Chang, Green Chem., 2012, 14, 917–920.
32. M. D. Purkayastha, A. K. Manhar, V. K. Das, A. Borah, M. Mandal, A. J. Thakur and C. L. Mahanta, J. Agric. Food Chem., 2014, 62, 4509–4520.
33. B. Das, P. Dadhich, P. Pal, P. K. Srivas, K. Bankoti and S. Dhara, J. Mater. Chem. B, 2014, 2, 6839–6847.
34. Q. Xu, P. Pu, J. Zhao, C. Dong, C. Gao, Y. Chen, J. Chen, Y. Liu and H. Zhou, J. Mater. Chem. A, 2015, 3, 542–547.
35. Z. Q. Xu, L. Y. Yang, X. Y. Fan, J. C. Jin, J. Mei, W. Peng, F. L. Jiang, Q. Xiao and Y. Liu, Carbon, 2014, 66, 351–360.
36. M. K. Barman, B. Jana, S. Bhattacharyya and A. Patra, J. Phys. Chem. C, 2014, 118, 20034–20041.
37. M. Rouabhia, V. Gilbert, H. Wang and M. Subirade, Biomed. Mater., 2007, 2, S38–S44.
38. J. Jana, M. Ganguly, B. Das, S. Dhara, Y. Negishi and T. Pal, Talanta, 2016, 150, 253–264.
39. T. J. Mosmann, J. Immunol. Methods, 1983, 65, 55–63.
40. S. M. Driever, M. J. Fryer, P. M. Mullineaux and N. R. Baker, Plant Signal Transduction, Humana Press, 2009,  DOI:10.1007/978-1-59745-289-2_7.
41. Y. T. Yang, J. D. Liao, C. Ching, K. Lin, C. T. Chang, S. H. Wang and M. S. Ju, Soft Matter, 2012, 8, 682–687.
42. S. M. Fitzsimons, D. M. Mulvihill and E. R. Morris, Food Hydrocolloids, 2007, 638–644.
43. S. Zhu, Y. Song, X. Zhao, J. Shao, J. Zhang and B. Yang, Nano Res., 2015, 8, 355–381.
44. P. Thevenot, A. Nair, J. Dey, J. Yang and L. Tang, Tissue Eng., Part C, 2008, 14, 319–330.
45. J. M. Leonardo, P. R. Reynolds and R. W. Tuveson, Mutat. Res., 1984, 126, 1–8.
46. S. Sun, S. Wong and A. Mak, J. Biomech., 2014, 47, 3650–3656.
47. J. M. Lee, B. S. Kim, H. Lee and I. I. Gun, Mol. Ther., 2012, 20, 1434–1442.

---