Protein Fibril-Templated Biomimetic Synthesis of Highly Fluorescent Gold Nanoclusters and Their Applications in Cysteine Sensing.

Biomimetic synthesis of multifunctional fluorescent gold nanoclusters (Au NCs) is of great demand because of their ever-increasing applications. In this study, we have used self-assembled bovine serum albumin (BSA) amyloid-like nanofibers as the bioinspired scaffold for the synthesis of Au NCs. The amyloid fibril stabilized gold nanocluster (Fib-Au NC) has been found to have appreciable enhancement of fluorescence emission and a large 25 nm red shift in its emission maxima when compared to its monomeric protein counterpart (BSA-Au NC). The underlying mechanism accountable for the fluorescence behavior and its spectral shift has been thoroughly investigated by a combined use of spectroscopic and microscopic techniques. We have subsequently demonstrated the use of Fib-Au NCs for cysteine (Cys) sensing both in vitro and inside live cells. Additionally, cellular uptake and postpermeation effect of Fib-Au NCs have also been ascertained by detailed flow cytometry analysis, viability assay, and real-time apoptotic gene expression profiling.

Introduction

Biomimicking synthesis of fluorescent nanoscale materials has generated widespread interest in the research ecosystem of material science and engineering.[1−3] Proteins are often used as templates for biomimicking synthesis of fluorescent noble metal nanoclusters[4−10] with defined properties and applications in contemporary biomedicinal research.[11−13] The pioneer work of Xie et al.[4] first described the synthesis of bovine serum albumin (BSA)-encapsulated fluorescent gold nanoclusters (Au NCs). Different protein systems have been explored for this purpose, including the fibrous proteins,[14,15] cage proteins,[16−18] heat shock proteins,[19,20] and others.

A number of proteins are known to form highly organized fibrillar cross-β-sheet structures, popularly known as amyloid fibrils. These fibrils are typically self-assembled robust nanostructures held together by weak non-covalent forces between β-sheets. The formation of amyloid fibrils is implicated in the pathology of several neurodegenerative diseases, including Alzheimer’s disease, Parkinson’s disease, ALS, Creutzfeldt–Jakob, prion diseases, and type II diabetes. Although disease causing amyloid fibrils have attracted major attention, nontoxic and functional amyloid fibrils are also known, which are shown to have defined functions in multiple systems, ranging from bacteria to humans. Some of these functions include mammalian skin pigmentation,[21] catabolism,[22] hormone storage,[23] epigenetic inheritance, and memory formation.[24] Amyloid fibrils have also been used as building blocks for functional materials such as conductive nanowires,[25] photovoltaic devices,[26] biosensors,[27] hybrids,[28] cell scaffolds,[29] optoelectronic devices,[30] artificial bones,[31] underwater adhesives,[32] liquid-crystal formation,[33] and so forth. More specifically, amyloid fibers developed from lysozyme were found to electrostatically direct the assembly of gold nanoparticles along the fibers into arrays with tunable particle spacing.[34] Such applications are only possible because of precisely assembled fibrillar structures, tunable molecular functionalities, binding affinities, excellent mechanical strength, and high stability over a wide range of harsh conditions including chemical alteration at the nanoscale range. Self-assembled protein fibrils may act as a “bioamplifier” in the context to fabrication of nanoscale materials, and a number of recent publications have been successful in this regard. One example is shown by Garcia et al., where they have reported the synthesis of fluorescent gold nanoclusters using human insulin fibril with emission maxima at 620 nm.[35] In this study, we have used BSA fibrils as the synthesis template to develop Fib-Au NCs. This methodology is also advantageous as a “green” synthetic model with cost-effective devise arrangements.

BSA is a globular plasma protein, which constitutes ∼60% of the total protein pool. It contains 583 amino acid residues in three similar structural domains (predominantly α-helical) adapting a heart-shaped configuration. It has 17 disulfide bonds and a free unpaired cysteine (Cys) at the 34th position. This Cys residue assists in the dimerization and subsequent higher-order self-association. We have found that the synthesis of Au NCs within BSA fibrils (Fib-Au NCs) not only amplifies the fluorescence intensity, but also results in a distinct 25 nm red shift. In addition, Fib-Au NC has been found to offer important applications in Cys biosensing.

Cys plays crucial roles in retaining a cellular antioxidant immune system, biocatalysis, posttranslational modifications, cellular metabolism, and detoxification.[36] Variation in Cys concentration thus affects biological processes within cells. Its deficiency leads to diseases like haematopoieses, leucocyte loss, hair depigmentation, psoriasis, and so forth, whereas neurotoxicity, cardiovascular, and Alzheimer’s diseases are linked to its elevated levels.[37] Among various analytical techniques, high-performance liquid chromatography and postcolumn derivatization and a spectrophotometric assay using Ellman’s reagent are the most common for estimation of Cys in biological samples.[36,37] However, the technique involves skilled manpower, expensive equipment, and time-consuming processes. In biological fluids, two other sulfhydryl (−SH) compounds, namely homocysteine (Hcy) and glutathione (GSH), are available along with Cys. As a result, specific quantification of Cys adds to further complications. In this work, we show that Fib-Au NCs detect and monitor Cys at a very low level limit and with significant specificity when compared to its detection of Hcy and GSH. The synthesized nanocluster probe can also be used conveniently inside live cells for cellular imaging, in which the presence of Cys alters emission output.

Live cell imaging has been carried out using synthesized NCs, where the probe has shown significant localized intracellular red fluorescence using excitation at 488 nm. The Fib-Au NC is found to show considerable cell permeability. More importantly, cells have not shown any sign of postpermeation stress. Our ensemble in cell studies involving flow cytometry and MTT assays over a gradient of NC concentrations reveals an efficient and rapid cytosolic permeation with very high cell viability and minimum apoptotic and necrotic events. Furthermore, as suggested by our real time polymerase chain reaction (PCR) studies, the NC treatment is found to alter the gene expression levels of apoptotic genes. These strings of in-cell studies suggest that NC can be a potential synthetic species for bioimaging purposes with no concurrent deleterious effects on live cells.

Results and Discussion

Synthesis and Characterization of BSA Fibrils Stabilized Gold Nanoclusters (Fib-Au NCs)

The synthesis of BSA fibril-stabilized gold nanoclusters involved two steps. In the first step, BSA fibrils were prepared, which were subsequently used in the second step as the template for the development of fibril-stabilized gold nanoclusters (Fib-Au NC). Fibrillation of BSA was carried out using available literature studies by incubating the protein samples at 75 °C at pH 3.[38] Fluorescence-based ThT assay was used for the initial characterization of BSA fibrils. A significant enhancement of ThT fluorescence intensity was observed for the BSA fibrils (λem of 485 nm for λex of 440 nm) (Figure a). The formation of the fibrillar protein resulted in a 16% drop in the ellipticity as judged by far-UV CD (Figure b). Fourier transform infrared (FTIR) measurements were subsequently used to quantify the secondary structural components of BSA in its native and fibril states. For the native BSA protein, four major absorption bands corresponding to β-sheet (1631 cm–1), α-helix (1651 cm–1), loops and turns (1663 cm–1), and β-turn (1683 cm–1) were considered. The native protein could be seen to be majorly α-helical in nature (55% α-helix, 17% β-sheet) (Figure c), an estimation supported by previous results and also by the crystal structure analysis of BSA.[39] When the protein was subjected to fibrillation, some additional IR absorption bands were observed, which corresponded to cross-β (1619 cm–1), loops and turns (1692 cm–1), and antiparallel β-sheet (1677 cm–1). To summarize, the fibrillation of BSA was accompanied by a clear decrease in α-helical content (from 55% in the native state to 35% in the fibrillar state) and an increase in beta sheet content (from 17% in the native state to 29% in the fibrillar state) (Figure d). As compared to the FTIR spectrum of native BSA protein, fibrillated BSA (Figure S1) showed intensity reduction of the amide I and II bands. The CH2 in-plane and out-of-plane bending bands were dominant, including 1442 and 991 cm–1, for BSA which became negligible when fibrillated. Atomic force microscopy (AFM) images (Figures a–d and S2) confirmed protein self-assembly. The average height of BSA fibrils was observed to be around 20 nm.

Figure 1

(a) Fluorescence ThT assay confirm the formation of BSA fibrils; (b) CD spectra showing decrease in the helicity percent of the protein BSA; deconvoluted FT-IR spectra of (c) native BSA protein and (d) BSA fibril. Panels (c,d) clearly show a significant decrease of α-helix (1651 cm–1, from 55 to 35%) with a subsequent increase in β-sheet (1632 cm–1, from 17 to 29%).

Figure 2

Panels (a,b) show AFM images (4 μm × 4 μm scan) of BSA fibrils in aqueous solution after 100 h of incubation at 75 °C. One of the rod-shaped aggregates from (a,b) is magnified (2 μm × 2 μm scan) and shown as (c,d). (The side bar is for references to color with respect to height). The height vs distance profiles are also shown in the figure.

Subsequently, self-assembled BSA nanofibers thus prepared was employed as a template for the synthesis of fluorescent gold nanoclusters (Fib-Au NCs) (Scheme ). Primary characterization of the nanocluster was carried out by studying optical spectroscopy. The UV–vis absorption spectrum of the Fib-Au NCs did not exhibit any characteristic surface plasmon band (Figure a, inset), indicating that gold nanoparticles with core diameters larger than 2 nm did not form.[40] Instead, the strong fluorescence emission was originated as expected from the gold nanoclusters. The as-prepared Fib-Au NCs exhibited red fluorescence at λem = 675 nm when excited at 520 nm (Figure b). The fluorescence characteristic of gold nanoclusters prepared by using native protein (BSA-Au NCs) was found different from the fibril-stabilized nanoclusters. A significant 25 nm red shift in fluorescence emission maxima was observed for Fib-Au NCs, compared to BSA-Au NCs (Figure c, Table ). To the best of our knowledge, such protein self-association-induced fluorescence enhancement along with bathochromic spectral shift is observed for the first time. Increased beta sheet content was reported to provide extra rigidity to overall protein structure.[41] The increased beta-sheet percentage obtained here may thus be responsible for the decrease in nonradiative decay of Fib-Au NCs, resulting in the observed enhancement of fluorescence intensity.[42] Moreover, the aggregated BSA template would also provide different electronic environments (difference in charge density) surrounding NCs, which would further alter the energy gap between the highest and lowest molecular orbitals, thereby causing the red shift. This can be supported by our previous observation, in which we showed using time-dependent density functional analysis that different chemical environments surrounding a Cys residue in a recombinant protein influence directly the Au–S bonding interaction. It was complemented nicely by experimental emission maxima (HOMO–LUMO energy gap).[43]

Scheme 1

Schematic Presentation of Synthesis of Highly Fluorescent Gold Nanocluster within BSA Fibrils

Figure 3

(a) Fluorescence excitation spectrum (λem = 675 nm) of Fib-Au NCs; inset of figure (a) is the absorption spectrum of the same; (b) emission spectra of Fib-Au NCs, λex = 520 nm (red), λex = 325 nm (black) at room temperature; (c) comparison between the emission spectra (λex = 520 nm) of BSA-Au NCs (black) and Fib-Au NCs (red).

Table 1

Summary Table Showing the Basic Differences between the Two Nanoclusters

	BSA-Au nanocluster	Fib-Au nanocluster
λem (nm)	650	675
TEM size (nm)	1.7 ± 0.3	1.6 ± 0.4
XPS	Au 4f7/2—83.1 eV	Au 4f7/2—84.01 eV
	Au 4f5/2—86.7 eV	Au 4f5/2—87.81 eV
FTIR	1651 cm–1—48%	1651 cm–1—37%
	1631 cm–1—19%	1631 cm–1—28%

Spectral red shift is associated with almost 40% increment in the fluorescence intensity of Fib-Au NCs (Figure c). The measured relative quantum yield of Fib-Au NCs was found to be 3.6%.

Analysis of the excitation spectrum of Fib-Au NCs (λem = 675 nm) showed the presence of two peaks positioned at about ∼325 and ∼520 nm (Figure a). Excitation at 325 nm produced emission maxima at 415 nm. It has been shown by ultrafast spectroscopy that the emission band at 415 nm originates from the icosahedral Au13 core of the nanoclusters.[44] In contrast, the near-infrared emission around 700 nm is due to the passivating monolayer involving a relaxation of the core excited states to S–Au–S–Au–S semiring states. The presence of surface ligands can modulate the fluorescence of metal NCs in two different ways: (i) charge transfer from the ligands to the metal core via the Au–S bonds and (ii) direct donation of delocalized electrons from electron-rich groups of the ligands to the metal core.[45]

The size of the gold nanoclusters was further investigated by transmission electron microscopy (TEM) (Figure ). Bright-field TEM micrograph in Figure a,b represented size distributions as well as lattice fringes of the nanoclusters synthesized within BSA protein. Their size distributions resulted in an average diameter of 1.7 (±0.3) nm. Almost a similar average diameter of 1.6 (±0.4) nm was obtained for BSA fibril stabilized nanoclusters (Figure c,d). Accumulation of the gold nanocluster on the protein fibril was clearly seen by TEM imaging (Figure c). The high-resolution TEM (HRTEM) image in Figure d revealed nanoclusters with lattice fringes, corresponding to the (111) plane of gold nanoclusters.

Figure 4

HRTEM images of BSA-Au NCs ((a,b) 1.7 ± 0.3 nm) and Fib-Au NCs ((c,d) 1.6 ± 0.4 nm). Consecutive lines in the inset image (yellow color in b and red color in d) show the planes (111).

X-ray photoelectron spectroscopy (XPS) was used to determine the chemical and oxidation state of gold (Figure a) and sulfur atoms (Figure b) in Fib-Au NCs. The Au core level for Fib-Au NCs showed distinct Au 4f7/2 and Au 4f5/2 components at binding energies (BE) of 83.1 and 86.7 eV, respectively (Figure a, Table ). In contrast, the reported values of BE of the Au 4f7/2 and Au 4f5/2 levels for BSA-Au NCs have been found to be at 84.01 and 87.81 eV, respectively.[46] The binding energy shifts between Fib-Au NCs and BSA-Au NCs may be related to differential effect of the protein environments on the electronic state of the Au NCs.[47] For sulfur, the peaks for BE of 2p3/2 at 161.7 and 167.2 eV (Figure b) can be attributed to the oxidized states of sulfur. The peak at 161.7 eV corresponded to the covalent interaction of gold nanoclusters with the sulfur groups of the Cys residue of protein.[48]

Figure 5

XPS spectra of Fib-Au NCs, (a) 4f7/2 and 4f5/2 peaks show the formation of Au(0); (b) two peaks of S 2p3/2 confirm the presence of different oxidized states of sulfur.

The formation of BSA-Au NCs changes overall helical content of the protein (from 55% in native BSA to 48% in BSA-Au NCs) (Figures c and 6a, Table ), a result observed before by others.[49] In contrast, the extent of α-helix component in the Fib-Au NC sample was found to increase slightly (37% as opposed to 35% in BSA fibrils) (Figures and 6b, Table ). A sharp, moderately intense FTIR band was observed at ∼880 cm–1 for both BSA-Au NCs and Fib-Au NCs, which is a key marker of tyrosine–tyrosine cross-linking.[50] This cross-linking is a consequence of Au3+ ion reduction by tyrosine residues of the protein (Figure S3a,b). In addition, enhanced signals of −COstr (1640–1690 cm–1, amide I region) and in plane NHbend (1510–1580 cm–1, amide II region) bands were observed for Fib-Au NCs. The −CH2 bending band at 1442 cm–1 also became dominant (Figure S3b and Table S1) after Fib-Au NC formation. In contrast, BSA-Au NCs showed significantly reduced signals for all the above mentioned FTIR bands (Figure S3a and Table S1).

Figure 6

(a) Deconvoluted FT-IR spectra of BSA-Au NCs; (b) deconvoluted FT-IR spectra of Fib-Au NCs.

Variation of Synthesis Parameters

The synthesis parameters were optimized for the best yield of the desired red emitting Fib-Au NCs. Concentration of protein was varied between 5 and 50 mg/mL. A significant decrease in fluorescence intensity of Fib-Au NCs was observed when the protein concentration used was beyond 20 mg/mL (Figure a). Interestingly, the change in fluorescence intensity was insignificant for the BSA-Au NC for the entire range of concentrations (from 5 to 50 mg/mL).

Figure 7

(a) Comparison between the fluorescence emission intensities of the BSA-Au NCs (black) and Fib-Au NCs (red) as a result of increasing BSA monomer concentrations; (b) fluorescence spectra obtained from the synthesis of Au25 by fibrils at different temperatures; (c) time dependent growth of the nanoclusters; (d) stability of the formed Au NCs over a period of a month.

The temperature effect on the synthesis of Au nanoclusters within BSA protein fibrils was also investigated, and optimum yield was observed at 37 °C (Figure b). Synthesis of Fib-Au NCs was repeated at five different temperatures, namely 8, 25, 37, 45, and 70 °C. The yield was insignificant at a low (8 °C) and a high temperature (70 °C), whereas comparable yields were obtained at 25, 37, and 45 °C. Subsequently, the time dependence of nanocluster synthesis was investigated by measuring the fluorescence intensity at 675 nm (λex of 520 nm) (Figure c). The long term stability of the prepared Fib-Au NCs was also checked over a period of 1 month (Figure d). When kept in aqueous solution at 4 °C for a month, Fib-Au NCs were found stable with a ∼20% decrease in fluorescence. However, the lyophilized powder of Fib-Au NCs can be kept at 4 °C for as long as 1 year without any loss of fluorescence intensity, which was checked by redissolving the powder in 20 mM sodium phosphate buffer at pH 7.4 (Figure S4a). In addition, Fib-Au NCs were found stable after one cycle of freeze-thawing (from 25 to −80 to 25 °C) (Figure S4b). Moreover, the zeta potential value of the Fib-Au NCs solution was found to be −21.3 eV, indicating sufficient stability of the amyloid-stabilized gold nanoclusters.

Application in the Detection of Cys

Interestingly, our as-prepared Fib-Au NCs exhibited selective fluorescence sensitivity toward Cys detection. It was found that the fluorescence intensity (λem = 675 nm) of Fib-Au NCs was sufficiently quenched by the gradual addition of Cys (Figure a,b). Hcy and glutathione (GSH) are the most common analytes, which interfere with Cys detection. Thus, for the sensing studies, a series of aqueous solutions of different analytes (all amino acids, Hcy, GSH) were used and their concentration was initially kept at 500 nM, which was gradually increased up to 2 mM keeping all other experimental conditions identical. The addition of Cys to Fib-Au NCs results in a large fluorescence quenching (Figure b). Other analytes did not show significant quenching, confirming the specificity of Fib-Au NCs. Furthermore, the detection ability of Fib-Au NCs follows the following order: Cys ≫ Hcy > GSH > other amino acids. The limit of detection (LOD) for the Cys sensing was found to be 76 × 10–9 M (76 nM) (linear range 76 nM to 300 μM). The concentration level of Cys in blood plasma of a healthy person typically ranges from 240 to 360 μM and normal content of Cys ranges from 30 to 200 μM in cells. Thus, Fib-Au NCs can be considered as sensitive enough for analysis of Cys in real biological samples.

Figure 8

(a) Barplots representing the extent of quenching of Fib-Au NCs in the presence of 2 mM aqueous solution of different analytes. F0 and F are the fluorescence intensities of Fib-Au NCs in the absence and presence of 2 mM analytes, respectively; (b) emission spectra of Au NCs in the presence of increasing concentrations of Cys from 500 nM to 2 mM; (c) relative fluorescence intensity of Fib-Au NCs vs of Cys concentration (symbols) and the corresponding linear regression (line); (d) modified Stern–Volmer plot and the corresponding linear regression; (e) effect of pH on the ΔF of the Au NCs-Cys system.

For unraveling the possible mechanistic pathways of the interaction of Au NCs with Cys molecules, Stern–Volmer analysis has been carried out. From the Stern–Volmer data, it is evident that the interaction is very strong and spontaneous. Stern–Volmer’s equation (eq ) can describe the quenching of the Fib-Au NCs aswhere F0 and F are the fluorescence intensities of the Fib-Au NCs in the absence and in the presence of a quencher (Cys), respectively, and [Q] is the concentration of the quencher (Cys). The value of Ksv is found to be 1.09 × 104 M–1. The straight line observed in Figure c suggested primarily static quenching of Fib-Au NC by Cys, which occurs due to their complex formation driven by the thiol group. Furthermore, we had estimated the binding constant (K = 67.72 M–1) and number of binding sites (n = 0.47) of the Cys molecules on the surface of Fib-Au NCs using a modified version of the Stern–Volmer equation (eq )

The associated thermodynamics of binding can also be estimated by measuring the free energy of binding (ΔGbinding) given by eq :

ΔGbinding was found to be −2.512 kcal mol–1. Next, we investigated the effects of pH on the ΔF(F0 – F) for the detection of Cys. As is observed from Figures e and S5, the ΔF decreased in the range of pH 3–6. Within pH range 6–8, the value of ΔF remained similar but with further increase in the pH up to pH 11 again showed increase in the ΔF of the system (Figure S5). The quenching experiment was repeated with BSA-Au NCs also to compare the sensitivity between Fib-Au NC and BSA-Au NC. Fluorescence intensity of BSA-Au NCs also decreased as a result of increasing Cys concentrations (Figure S6). BSA-Au NCs showed less binding (K of 33.8 M–1) compared to Fib-Au NCs (K of 67.72 M–1). In addition, the LOD value was also found high for BSA-Au NCs (50 μM) when compared to Fib-Au NCs (76 nM).

In this context, it is important to compare the sensitivity and specificity of Fib-Au NCs with previously reported literature. The present LOD value is significantly low when compared to several fluorescence-based small molecule sensors.[51,52] In addition, the LOD of Fib-Au NCs was found less than a number of composite nanoparticle-based sensors of Cys’s.[53,54] Comparable LOD in the range of nM was observed with several other reports.[55,56]

MTT Assay and Fluorescence Activated Cell Sorting

It is also important to understand the nanocluster effect on cellular viability prior to propose them as a convenient candidate for in-cell studies. Therefore, MTT assay was conducted with HeLa cells to evaluate the effects of Fib-Au NCs on cell viability over a broad range of concentrations, ranging from 500 nM to 1600 μM. The cellular viability assay reflected over 98% cell viability when the cluster concentration was kept at 500 nM. At concentrations of 50 and 100 μM, the viability of cells was observed to be 90%. The viability assay data indicated negligible clusters induced dose-dependent cytotoxicity (Figure a). To have a direct quantification of Fib-Au NC permeation, flow cytometry by FACS (fluorescence activated cell sorting) was performed. Flow cytometry results confirmed a significant uptaking of Fib-Au NCs at a concentration of 50 and 100 μM (as compared to 25 μM) within a time frame of 20 min (Figure b). To probe into the mechanisms, which lead to a loss of viability of cells, we resorted to flow cytometry-based investigations. The study was done to unveil if the cell viability is getting affected by apoptosis or if they are treading the necrotic pathway. This stands significant, as death by the necrotic mechanism is highly toxic for cells in the immediate vicinity. On the contrary, apoptosis is programmed internal machinery, whereby the cell organelles are packaged in vesicular structures and there is no spillage of lytic proteins in the vicinity. Interestingly, our flow cytometry results revealed early apoptotic rates of 8 and 10% when treated by 50 and 100 μM, respectively (Figure c,d). Thus, correlating all the above mentioned findings, it can be inferred that at the concentration of 50 μM and a treatment time frame of 20 min Fib-Au NCs can be effectively used for bioimaging purposed with no observable cell damage.

Figure 9

(a) MTT assay results showing percentage viability of cells; (b) bar plot showing the extent of permeation of Fib-Au NC at different concentrations; (c) plot of flow cytometry data indicating gated cell population considered for control set, Fib-Au NC treated sets of 50 and 100 μM; (d) viable, apoptotic, and necrotic cell population percentage plot as obtained for control and treated (20 min incubation) sets with 50 and 100 μM Fib-Au NCs; (e) plot showing real time gene expression levels of apoptotic genes on incubation with Fib-Au NCs.

Real-time PCR Analysis

Novel compounds on cytosolic permeation may induce epigenetic changes. To probe any probable epigenetic modulation on apoptotic cascade, which in turn may affect cell health and viability we scored Bcl2 Bax expression by real-time PCR. Treatment of the HeLa cells showed ∼3 fold increase in the gene expression of the anti-apoptotic marker Bcl2 and a concomitant decrease in the expression of the pro-apoptotic marker Bax by ∼3 folds (Figure e). This observation complies with our flow cytometry results, which revealed a very negligible apoptotic population and a vast majority of viable cells. The real-time PCR data suggest that the Fib-Au NC does not induce any epigenetic effect on the apoptotic propensity of the cells. On the contrary, its permeation orchestrates the expression of the anti-apoptotic factor, further inhibiting the apoptotic propensity.

Detection of Exogenous Cys in Live Cells

Prior to imaging exogenous Cys, we first assured intracellular uptaking of Fib-Au NCs within cell cytosol. Live cell confocal imaging revealed significant extent of Fib-Au NC permeation upon treatment of HeLa cells with 50 μM of Fib-Au NCs for 20 min. The permeation stands out evident with respect to the control set as postpermeation bright fluorescence was observed after an excitation at 488 nm (Figure a,b). As quantified from pixel distributions and 3D interactive surface plots of Image J ensemble, the cells with Fib-Au NCs showed an average fluorescence intensity of 25 units (arb unit) (Figure c). The postpermeation residence of Fib-Au NCs inside the cell does not result in cell stress. Fib-Au NCs were observed in the cell cytosol with no marked alteration on the cell shape.

Figure 10

Confocal fluorescence imaging of HeLa cells: (a) control set and (b) fluorescence image of Fib-Au NCs at 50 μM for 20 min incubation; (c) 3D interactive surface plot of Fib-Au NC permeation, that is, heat map.

To image exogenous Cys, N-ethylmaleimide (NEM) precultured HeLa cells were taken because NEM can efficiently block intracellular sulfhydryl compounds.[57] These NEM precultured HeLa cells were first treated with same Fib-Au NCs and the NCs showed their own bright red fluorescence (excitation at 488 nm) after uptaking. However, further treatment of NEM precultured HeLa cells with exogenous Cys causes significant quenching of the nanocluster fluorescence (Figure a–d).

Figure 11

Confocal fluorescence imaging of (a) cells pretreated with NEM and (b) cells pretreated with NEM, washed and treated with Cys; 3D interactive surface plot of Fib-Au NC permeation in (c) cells pretreated with NEM and (d) cells pretreated with NEM, washed and treated with Cys.

Conclusions

A simple and green approach for the synthesis of the intensely red-emitting BSA fibril-directed Au NCs has been presented in this manuscript. The highly water-soluble clusters possess excellent stability for more than 1 month. The Fib-Au NC has been subsequently used as a sensitive and selective probe for the label-free, sensitive, and selective detection of Cys (LOD = 76 nM) even in the presence of Hcy or GSH, which often interferes in its detection. The as-prepared Fib-Au NC is also an efficient candidate for imaging exogenous Cys within live cells. The nanomaterial itself exhibits excellent photostability and appreciable biocompatibility within cell cytosol. The red fluorescence emitting Fib-Au NCs also show excellent localization and distribution in the cytoplasm with the significant fraction entering into the nucleus of HeLa cells. As a concluding remark, we are proposing this Fib-Au NC as a potential nanoprobe for Cys sensing along with bioimaging property.

Experimental Procedure

Materials

Auric chloride (HAuCl4) was obtained from Spectrochem (India). BSA and NaOH pellets were obtained from Amresco, Ohio, U.S. Sodium chloride, sodium phosphate monobasic, sodium phosphate dibasic, glutathione, Hcy, and the amino acids [l-alanine (C3H7NO2), l-arginine (C6H14N4O2), l-asparagine monohydrate (C4H8N2O3), l-aspartic acid (C4H7NO4), l-Cys (C3H7NO2S), l-glutamic acid (C5H9NO4), l-glutamine (C5H10N2O3), l-glycine (C2H5NO2), l-histidine (C6H9N3O2), l-isoleucine (C6H13NO2), l-leucine (C6H13NO2), l-lysine (C6H14N2O2), l-methionine (C5H11NO2S), l-phenylalanine (C9H11NO2), l-proline (C5H9NO2), l-serine (C3H7NO3), l-threonine (C4H9NO3), l-tryptophan (C11H12N2O2), l-tyrosine (C9H11NO3), and l-valine (C5H11NO2)] were purchased from Sigma (St. Louis, MO) and used without any further purification. The fluorescent probe thioflavin T (ThT) was also obtained from Sigma. All the buffer solutions were prepared using millipore water. The buffer was prepared freshly prior to every experiment. Filters (0.22 μm) were supplied from Millipore (Ireland).

Synthesis of the BSA Fibrils and Fibril-Assisted Fluorescent Gold Nanoclusters

The synthesis of BSA fibrils was done according to the earlier reported methods with slight modifications.[38] All the glassware were thoroughly cleaned before use with freshly prepared aqua regia, rinsed with methanol and ultrapure H2O, and then dried in an oven at ∼100 °C for 1–2 h. In the first step of synthesis, BSA fibrils were produced by dissolving BSA (concentration varying from 1 to 50 mg/mL) in phosphate buffer in a round bottom flask and were incubated at a specified temperature (60–75 °C)[38] in a water bath without agitation for 100 h. In the next step, 500 μL of aqueous HAuCl4 solution (10 mM) was added within previously prepared BSA fibril solution and stirred vigorously at 37 °C. About 5 min later, 1 M aqueous solution of NaOH was added dropwise so that the solution would reach pH 12. After that, the reaction was kept in incubation at 37 °C under constant stirring for 24 h. The color of the solution changed to deep red. The resultant BSA fibril encapsulated Au NC solution was filtered through a 0.22 μm filter, lyophilized, and stored in refrigerator at 4 °C for further use.

Characterizations of the Au NCs

BSA fibrils were primarily characterized by AFM using a Pico Plus 5500 AFM system (Agilent Technologies, USA) in AAC mode. Images were processed by flattening using PicoView software (Agilent Technologies, USA) and the width of protein fibrils were measured manually.

A thermoscientific UV-10 spectrometer was used for UV–visible absorption spectroscopy measurements. Absorbance scans (200–750 nm) were taken using a quartz cuvette of 1 cm path length. All fluorescence measurements were made using a PTI fluorescence spectrometer. The emission spectra were recorded after excitation at 520 nm. The slits were set at 5 nm for both excitation and emission. All measurements were done repeatedly to obtain reproducible results.

A JEOL JEM-2100F transmission electron microscope was used for obtained TEM images. The sample was placed on a carbon-coated Cu grid and analyzed. Far-UV CD (between 190 and 250 nm) spectra of the samples were recorded using a cuvette of 1 mm path length in a JASCO J720 spectropolarimeter. The concentration of protein was kept at 10 μM.[58] The bandwidth was set at 1 nm. Five CD spectra were taken and then averaged for each sample.

FT-IR spectra were taken using a Bruker 600 series FT-IR spectrometer. The deconvolution of raw FT-IR spectra from 1700 to 1600 cm–1 (belonging to amide I region) was done by fitting the curves to Gaussian line shapes. Then, the peaks were assigned to components associated with different secondary structures according to previous reports.[59,60] XPS measurements were done using Omicron Nanotechnology instrument. The ζ-potential was measured using Zetasizer Nano-ZS (Malvern Instruments, U.K.) at 25 °C.

Cys Detection Assay

All the analytes (all amino acids, Hcy, GSH) were taken and a 5 mM stock solution for each of them was prepared. Then, the analyte solutions were gradually added to the nanoclusters and the consequent change in fluorescence intensity was monitored. Effect of pH on nanocluster mediated Cys sensing was further monitored.

Cytotoxicity (MTT) Assay

For estimating the optimal concentration of Fib-Au NCs for in-cell studies, MTT (3-(4, 5-dimethylthiazolyl-2)-2, 5-diphenyltetrazolium bromide) assay was done. The assay primarily measures the proliferation rate of the cells. Only the viable cells reduce MTT into formazan crystals making use of mitochondrial dehydrogenase enzyme. Human cervical cancer (HeLa) cells were seeded in 96-well plates keeping an optimal count of around 103 viable cells per well. The cells were incubated for 1 day. The viability assay was performed by incubating the cells with a series of Fib-Au NCs concentrations, ranging from 500 nM to 1600 μM. After 20 min of treatment of cells with Fib-Au NCs, the media was replaced with the MTT containing media (1 mg/mL). The cells were thereafter incubated for a time frame of 4 h at an ambient temperature of 37 °C. The medium was finally removed and the cells were diluted in 100 μL of 0.04 N HCl/isopropyl alcohol. The relative formazan formation in each well was monitored by determination of absorbance at 540 nm in a microplate reader (MultiSkan, Thermo Scientific, USA). The resulting absorbance values were converted to percentage viability with respect to the untreated control cells.

Permeation Extent Analysis by Flow Cytometry

Au NC cellular permeation was also assessed by resorting to flow cytometry (FACS). A density of 107 cells was followed for FACS measurements. As the MTT assay showed above 90 percent viability of cells at Au NC concentrations of 25, 50 and 100 μM, we restricted FACS measurements to only these concentrations of Au NCs. Treatment was made for 20 min and the incubation was carried out at 37 °C. FACS was carried out in BD FACS Diva. For all FACS-based permeation measurements, 405 nm line was used for excitation from a 15 mW argon ion laser and the blue fluorescence channel (FL1).

Apoptosis Analysis by Flow Cytometry

The probable apoptotic events of HeLa cells upon Au NC uptake were probed by using the Annexin V-FITC apoptosis detection kit. The 12-well plates were used to seed HeLa cells at a density of 107 cells. After washing with phosphate buffer saline, the cells were resuspended in 500 μL of binding buffer. Thereafter, 5 μL of Annexin V-FITC followed 5 μL of propidium iodide were added to each of the centrifuge tube followed by thorough mixing so as to scour for necrotic cell population. Incubation in the dark for 10 min was done before the analysis with BD FACS Diva (Gallios flow cytometer, Beckman-Coulter).[61] The estimations were made with respect to control untreated cell sets.

RNA Isolation and Quantitative PCR

For quantitative PCR, the total cellular RNA was extracted using Tripure reagent (Sigma). During the extraction, standard methods described in the kit protocol were followed. Using iScript reverse transcriptase (Bio-Rad) and referring to manufacturer’s instructions, cDNA was synthesized. qPCR was carried out in a 7500 fast real time PCR system (Applied Biosystems) using SYBR reagent (Bio-Rad). In brief, 1 μg of RNA was reverse transcribed followed by qPCR. The reaction conditions followed were primary denaturation step (95 °C for 5 min) and cycling step (denaturation at 94 °C for 15 s, annealing at a specific temperature for each set of primers for 30 s, extension at 72 °C for 30 s repeated for 35 cycles), followed by melting curve analysis (45–90 °C). In the experimental exhibit, 18s rRNA as the endogenous control was used. The relative fold change was measured and has been shown with representative bar diagram. The experiment was repeated three times and the values represent mean ± S.E.M. (*p < 0.005).

Cell Imaging Study

For live cell experiments, cervical cancer cell line (HeLa) was used and the cells were grown and maintained in Dulbecco’s modified Eagle’s medium supplemented with 10% heat-inactivated fatal bovine serum, 110 mg/L sodium pyruvate, 4 mM glutamine, 100 units/mL penicillin, and 100 μg/mL streptomycin. Incubation was made in humidified air containing 5% CO2 at 37 °C. For imaging, the seeding was done in a 35 mm poly-d-lysine coated plate (MatTek Corporation, Ashland, MA). The cells were allowed to grow ∼75% confluency is reached. Thereafter, these cells were treated with Fib-Au NC solution at a stock concentration of 1 mM, incubated for 20 min in humidified air containing 5% CO2 at 37 °C, and subjected for imaging. For the control experimental set, the cells were cultured with NEM (500 μM).

The time frame followed was 30 min with an ambient temperature of 37 °C to remove the intracellular biothiols. After thorough and gentle washing with phosphate-buffered saline (PBS) buffer, the cells were thereafter incubated with 1 μM of Fib-Au NCs for 30 min at 37 °C. Instant confocal imaging was performed thereafter. For the confocal imaging of exogenous Cys-treated cells, HeLa cells were initially pretreated with NEM (500 μM) for a time frame of 30 min at 37 °C. They were washed with PBS and then treated with probe compound (1 μM) for a time frame of 30 min at 37 °C. Finally, the resulting HeLa cells were incubated with Cys (50 μM) for 30 min at 37 °C. The imaging of cells was then carried out after washing with PBS buffer.

Imaging experiments were carried out using a Zeiss 510 META Confocor3 LSM set up (Carl Zeiss, Evotech, Jena, Germany) equipped with C-Apochromat 40X (NA = 1.20, water immersion) objective and confocal images were acquired with 512 × 512 pixel (pinhole aperture ≈ 1 airy units). The nanocluster was excited using a diode laser at 488 nm.

With an objective to gain a quantitative insight into the Fib-Au NCs permeation within cell cytoplasm, we used Java based image analysis bundle of Image J. Spots showing fluorescence were randomly selected for obtaining a preliminary insight of Fib-Au NC permeation. The pixel densities of these spots were compared with respect to the control set where no treatment was made and hence not showing any fluorescence. Furthermore, a surface plot analysis output was obtained resorting to the Surface Plot plugin of Image J. The average post uptake fluorescence intensity after excitation at 405 nm was found to range from 12 to 16 arb unit. The height for the plot represents the luminance of each pixel which constitutes the image. The 3D visualization resorts to JRenderer3D plugin. The final rendition offers a quantitatively precise postuptake residence rate of the Fib-Au NC.