Abstract

Cancer cells have the capacity to synthesize nanoparticles (NPs). The detailed mechanism of this process is not very well documented. We report the mechanism of biomineralization of aqueous gold chloride into NPs and microplates in the breast-cancer cell line MCF7. Spherical gold NPs are synthesized in these cells in the presence of serum in the culture media by the reduction of HAuCl4. In the absence of serum,the cells exhibit gold microplate formation through seed-mediate growth albeit slower reduction. The structural characteristics of the two types of NPs under different media conditions were confirmed using scanning electron microscopy (SEM); crystallinity and metallic properties were assessed with transmission electron microscopy (TEM) and x-ray photoelectron spectroscopy (XPS). Gold-reducing proteins, related to cell stress initiate the biomineralization of HAuCl4 in cells (under serum free conditions) as confirmed by infrared (IR) spectroscopy. MCF7 cells undergo irreversible replicative senescence when exposed to a high concentration of ionic gold and conversely remain in a dormant reversible quiescent state when exposed to a low gold concentration. The latter cellular state was achievable in the presence of the rho/ROCK inhibitor Y-27632. Proteomic analysis revealed consistent expression of specific proteins under serum and serum-free conditions. A high-throughput proteomic approach to screen gold-reducing proteins and peptide sequences was utilized and validated by quartz crystal microbalance with dissipation (QCM-D).

Statement of significance

Cancer cells are known to synthesize gold nanoparticles and microstructures, which are promising for bioimaging and other therapeutic applications. However, the detailed mechanism of such biomineralization process is not well understood yet. Herein, we demonstrate that cancer cells exposed to gold ions (grown in serum/serum-free conditions) secrete shock and stress-related proteins with specific goldbinding/reducing polypeptides. Cells undergo reversible senescence and can recover normal physiology when treated with the senescence inhibitor depending on culture condition. The use of mammalian cells as microincubators for synthesis of such particles could have potential influence on their uptake and biocompatibility. This study has important implications for in-situ reduction of ionic gold to anisotropic micro-nanostructures that could be used in-vivo clinical applications and tumor photothermal therapy. Θ 2018 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction

Metal and polymeric nanoparticles (NPs) are utilized in a wide range of biomedical applications such as radiotherapy, drug delivery, x-ray imaging due to their size and shape dependent biophysical and optoelectronic properties [1–3]. Metal NPs suitable for biomedical applications are synthesized using a broad range of microbes, such as actinomycetes, algae, bacteria, fungi, viruses, and yeasts [4–6]. It is shown that these particles can also be synthesized by plants [5] and mammalian cells [7]. Using such cells as biological vehicles for the biosynthesis of NPs is highly beneficial, due to their hierarchical structuring, safety, costeffectiveness, and negligible carbon footprint [8]. Currently, the biological routes for the synthesis of shapeand size-controlled metal NPs are not fully explored. There is documented evidence for bacteria [9], diatoms [10] and calcite [11] serving as ideal platforms for the bulk synthesis of NPs [10,12]. However, there is a paucity of their use for biomedical applications due to difficulties in controlling shape and size, which effects long term cell viability.

The biological platforms mentioned above are capable of creating anisotropic NPs in a controlled environment by utilizing polypeptides. These polypeptides bind specifically to inorganic surfaces as functional molecules and/or cellular components such as cell membrane vesicles [13,14]. The formation of gold (Au) NPs are very well documented in human embryonic kidney (HEK293), neuroblastoma (SKNSH) and cervical cancer (HeLa and SiHa) cell lines [7,15]. While the biosynthetic pathway remains elusive, it is postulated that cellular components such as redox enzymes and carbohydrates are a requirement [15]. In the current research paper, we delineate a detailed mechanism of NPs and anisotropic triangular or prismatic microstructure formation in mammalian cells. The process of gold biomineralization is investigated by utilizing the cancer cell line MCF7 as a model, which secretes specific proteins with gold-binding polypeptides (GBP). Additionally, the impact of the biomineralization process on the overall cellular health is investigated. We demonstrate that facet-specific GBP sequences are used as regulating agents for the highly controlled synthesis of nanocrystals with particular shapes.

2. Materials and methods
2.1. Synthesis of NPs and microplates with cell lines

Human breast adenocarcinoma (MCF7) and mouse myoblast (C2C12) cell lines were used as model cells in this investigation. Both cell lines were maintained in T25 culture flasks containing Dulbecco’s Modified Eagles Medium (DMEM) supplemented with 10% heat inactivated fetal bovine serum (FBS) and 1% antibiotic (Penicillin/Streptomycin) solution at 37 。C (5% CO2). Once cells reached a confluent monolayer, they were incubated with filter sterilized Gold (III) chloride hydrate (HAuCl4, Sigma Aldrich) prepared in DMEM (serum containing conditions) or phosphate buffered saline (PBSserum free conditions) at a final concentration of 1 mM. Four negative controls were used. These consisted of (a) dead cells incubated in PBS only; (b) live cells incubated in PBS and HAuCl4 ; c) live cells incubated in PBS only and d) PBS with HAuCl4 respectively. Cells were incubated between 36 and 48 h, during which anisotropic prisms formed. Under serum containing conditions, the visual hallmark was a change in the color of the cell culture medium (red to dark pink). No color change was observed in the negative controls. In PBS, gold solutions are stable for months and hence the possibility of autoreduction was ruled out. Human umbilical vein endothelial cells (HUVECs) and human adipose derived stem cells (ADSC) were obtained from Lonza, Germany. Cells were cultured in endothelial growth medium (EGM) and Human Adipose Derived Stem Cell growth medium supplemented with their respective growth factor kits (Lonza, Germany).

2.2. X-ray photoelectron spectroscopy

Samples were prepared by fixing the cells in 4% Paraformaldehyde in PBS for 30 min, dehydrating in graded ethanol (1%–25%– 50%–70%–100%) for 10 min each. Samples were then critical point-dried (Leica, EM CPD300, Germany). NPs and microplates suspended in deionized water were drop coated onto silicon wafer chips (5 根 6 mm) and air-dried overnight. Energy calibration, removal of contaminants, and charge compensation were carried out while recording measurements. The XPS spectra were deconvoluted with the Thermo VG Scientific Advantage software version 5.47 (Thermo Fisher Scientific). XPS analysis was performed with a Thermo VG Thetaprobe 300 (Thermo Fisher Scientific) system using monochromatic incident Al Ka radiation (hm = 1486.68 eV; spot size 400 μm; base pressure <10–7 Pa; average detection angle of 53。with respect to the sample surface).

2.3. Fourier transform infrared spectroscopy analysis

Chemical imaging of cells in the presence and absence of HAuCl4 was conducted with a Bruker Tensor II spectrometer over an attenuated total reflectance (ATR) sensor. Cells grown to 100% confluence on flat glass substrate were treated with gold ions for 4 days in a tissue culture incubator. Thereafter, pending media removal cells were washed with PBS and fixed with 4% paraformaldehyde in PBS for 30 min. Prior to IR analysis, the slides were washed with Hank’s buffered saline for 3 times. The sample was placed over the ATR-crystal carefully and pressed with a platinum diamond ATR-accessory to avoid uneven in thickness and facilitate close contact of the sample with the ATR crystal [16]. IR Spectra were recorded into absorbance mode with 4 cm-1 resolution and 1024 scans. Data analysis was performed with OPUS software (Bruker).

2.4. pH and UV–vis spectroscopy analysis

Changes in the pH or time dependent optical properties were assessed periodically after every 24 h via sampling of aliquots (1 mL) of the aqueous component (LAQUA, HORIBA Scientific, Germany). UV–visible spectra of the solution was recorded using an UV–visible spectrophotometer (Synergy ZEN 5, Biotek Instruments).

2.4. Dynamic light scattering and zeta potential measurement

The hydrodynamic diameter of the anisotropic NPs was determined using a light scattering instrument (MÖBIUζ勇 analyzer, Wyatt technology, Germany) with an ATLAS pressurization system. The zeta potential was calculated from measurements of electrophoretic mobility performed by the same instrument.

2.5. Scanning electron microscopy

Scanning electron microscopy (SEM) samples were prepared by incubating cells with 1 mM HAuCl4 for 4 days. Thereafter, cells were fixed in 2.5% glutaraldehyde in PBS for 45 min at 4 。C, rinsed with PBS, then water. The samples were dehydrated in an increasing ethanol gradient (30%, 50%, 70% and 90%) for 5 min followed by a 10 min incubation in 100% ethanol. Residual alcohol was removed using an automated critical point dryer (Leica EM, CPD 300). Silicon wafers with cells were air dried and sputter coated with a 5 nm coating of nickel using a Leica coating system (Leica EM, ACE600) and a Zeiss Ultra 55 Gemini scanning electron microscope equipped with a nitrogen cooled EDAX detector system was used for imaging and element dispersive x-ray spectroscopy (EDX) of MCF7 cells containing NPs/microplates. An accelerating voltage of 5 keV and INLENS detector was used for imaging.

2.6. Transmission electron microscopy

MCF7 cells grown in T25 flasks to confluency were harvested using routine trypsinization, resuspended in PBS and centrifuged (2000 rpm, 5 min thrice). Cell pellets were fixed in 3% glutaraldehyde in HEPES buffer (pH 7.4) and dehydrated with graded ethanol (30%, 50%, 70% and 90%) for 5 min followed by a 10 min incubation in 100% ethanol) prior to embedding in epoxy 618. A microtome (EM-UC6) from Leica Co, Austria was used to cut ultrathin sections of samples onto a glass coverslip and stained with uranyl acetate and lead citrate. For high resolution imaging and electron diffraction, the NPs and microplates were drop coated on a suspended monolayer of graphene on a transmission electron microscope (TEM) grid substrate (Quantifoil gold, Sigma). In-situ TEM was performed using a CM 200 device (Phillips Electronics) operated at 200 kV.

2.7. Irreversible cell senescence analysis with Y-27632 Rho/ROCK inhibitor

Quantitative cell senescence was studied with Live Cell Analysis Assay Kit, which uses a fluorogenic substrate to measure SA-ß-Gal activity (ENZO life sciences, Switzerland). The membrane permeable compound is a non-fluorescent substrate of ß-galactosidase, which after hydrolysis of the galactosyl residues emits green fluorescence and remains confined within the cell. In comparison of the conventional cytochemical assay using X-gal, the use of a fluorogenic substrate greatly enhanced the sensitivity of the assay. Also, the non-toxic fluorogenic substrate allows to study live cells by flow cytometry for more quantitative measurement of SA-ß-Gal activity. In brief, after overnight incubation of MCF7 cells with different concentration of gold ions (2–0.25 mM), the medium from the senescent cells, expressing SA-ß-Gal was aspirated and cells with CeNPs were incubated at 37 。C for 2 h after adding 2 mL of 1X Cell Pretreatment Solution. Subsequently, 10 μL of 200X SA-ß-Gal Substrate Solution was added directly to the cells in 1X Cell Pretreatment Solution; gently mixed and incubated at 37 。C overnight. The stained cells were washed three times with 3 mL 1X PBS and analyzed for cell senescence by epifluorescence microscopy (Excitation: 485 nm/Emission: 520 nm).

2.8. Time-lapse video microscopy

MCF7 cells were seeded at a density of 5 根 104 cells/ mL in 8well glass bottom microwell plates (Ibidi, Germany) and maintained with DMEM culture media as described previously. Once the cells formed confluent monolayers, they were used for TimeLapse Video Microscopy. The Ibidi plate was placed in an incubation chamber (H301-EC-BL, Okolab) and placed on the stage of a Nikon Eclipse Ti Confocal Microscope with a Yokogawa CSU-W1 spinning disk. The temperature was maintained at 37 。C. Prior to starting time-lapse, cells were washed and supplemented with the corresponding media containing gold ions (0.25–1 mM). MCF7 cells with no gold ions served as a control. Bright-field phase contrast images were collected every 10 min for 12 h by a CoolSNAP HQ2 CCD camera (Photometrics, AZ).

2.9. Proteomic analysis and peptide mass finger printing with MALDITOF

The protein coat formed on NP-microplates was analyzed with MALDI-TOF. Sample aliquots containing NP-microplates were incubated for 2 hat 50 。C in 1,4-dithioerythritol (DTE) and 2.5% sodium dodecyl sulphate solution to remove protein from biomineralized Au NPs surface. The protein content in the samples was quantified using a Bicichoninic Assay Kit (Sigma Aldrich, Germany) according to the provided protocol. The quantified proteins samples were suspended in Laemmli buffer (125 mM Tris-HCl pH 6.8, 2% SDS, 5% glycerol, and 0.002% bromophenol blue) at a final loading volume of 15 μL and separated using 8–15% gradient ready Gel勇 Tris-HCl precast gel using Mini-PROTEAN tetra electrophoresis system (8 x 10 cm) at constant voltage of 80 V until the dye front reached the lower end of the gel. After Coomasie staining in 1X dye solution, the gel was imaged with a GS710 Imaging Densitometer (BIORAD), and analyzed with Image Master Platinum 5.0 software (Amersham). The protein bands were extracted by excision from the gel using a scalpel and digested with trypsin. Tryptic digested peptides were concentrated with a POROS勇 XS cation exchange resin and eluted into 2.5-dihydroxybenzoic acid (DHB) matrix (Sigma Aldrich) and spotted on the target plate. The Ion electrospray spectra for MS/PMF was carried out using AXIMA resonance (SHIMADZU Biotech) and analyzed with Launchpad 2.9 software.

2.10. Gold binding studies with quartz crystal microbalance dissipation (QCMD) analysis

The Q-Sense E4 instrument (Quantum design, Germany) was used for QCM-D measurements. Gold-coated AT-cut quartz sensor crystals, with a fundamental frequency of 5 MHz, were purchased from Q-Sense AB. Prior to the experiment, the QCM-D chamber and crystals were pre-conditioned. The first step involved a wash with de-ionised water for 10 mins, followed by a 20–40 min wash with the working solution (PBS) until a stable baseline was obtained. Thereafter the samples of peptide at 4, 6 and 8 μM were run at a flow rate of 50 μL/min for 60 min and proceeded with PBS flow to measure the desorption of the peptide. After completing the experimental run, the QCM-D chamber and crystals were flowed with 2% Hellmanex III solution (Hellma Analytics, Germany) for 20 mins, followed by 0.1 M hydrochloric acid, de-ionised water and air drying. Frequency and dissipation shifts (Δf and ΔD) were recorded at the 3rd, 5th, 7th, 9th and 11th harmonics. As very similar frequency and dissipation data were obtained at the different harmonics, which is characteristic for rigid layers in QCM-D, only the results obtained at the seventh harmonic are presented in the graphs. Frequency shifts were normalized to the fundamental frequency by dividing with the overtone number. The mass deposition of peptide on the gold surface was estimated using the Sauerbrey equation:Δm = CQCMΔfn /n where Δm is the adsorbed mass on the surface, CQCM is the mass sensitivity constant (17.7 ng cm-2 Hz-1, for the 5 MHz quartz sensors used in our study), and Δfn is the change in the resonance frequency at the nth harmonic. Linear regression was performed to calculate the adsorption (ka) and desorption (kd) rates which indicate adsorption/ desorption behavior of the gold-binding peptides from the surface of the gold coated QCM substrate during the rinse with the buffer.

2.11. Statistical analyses

Data is presented by the mean value ± the standard deviation (SD). For senescence, concentration-and-time dependent studies, statistical analysis was carried out using a paired or unpaired two-tailed student’s t-test. Experiments were conducted in triplicate five times (n = 5).For the proteomic analyses and gold binding determination, all statistical tests were performed in log space with the assumption of peptide log-ratios being normally distributed. The average in this context is the geometric mean, and the standard deviation (SD) is the geometric SD for the log-ratio. This is a factor never less than 1. The Shapiro Wilk test was used to test for outliers and report the SD for the protein ratio. It relies on the notion that the null hypothesis of the population is normally distributed. The null hypothesis is rejected if the p value is less than 0.05.

3. Results and discussion

Experiments under serum containing and serum free conditions were conducted for seven days in total. The cell culture media of MCF7 cells changed color upon the addition and mixing of aqueous chloroauric acid (HAuCl4) from red to pale yellow (Fig. S1A), accompanied by a concentration and time dependent decline in pH within 8 h. The lowering of the pH was far more pronounced in serum containing culture (Fig. S2B). In serum free conditions, pH changes were small due to the buffering effects of the phosphate buffered saline (Table S1). In cells grown with serum supplemented media, the drop of pH after two days occurs due to a scarcity of glucose, thereby stalling cellular respiration. The cells then switch to ATP generation by activating the glycolytic pathway (Table S1). The ensuing release of glycolytic metabolites makes the cell culture media acidic [17]. The pH of the medium shifts towards a neutral pH from the fourth day onwards, which is attributed to cellular adaptation whereby the cells release buffering proteins and lipids into the extracellular milieu [18]. This was quantified and recorded in Table S2.

The formation of NPs happens inside cells grown in serum containing media. This process is initiated within 8 h incubation (Fig. 1A). The mechanism of gold prism formation under serum free can be observed in the SEM images (Fig. 1B, C, D). The thickness of the microplates, and also the spherical NPs are shown in the inset in Fig. 1B. It is postulated that the larger structures rely on spherical NPs acting as ‘‘seeds” which are subject to complexation on the live cell membrane [19,20]. The SEM and EDX investigation revealed the microplates formed to be hexagonal or trigonal in shape and composed solely of metallic gold (Figs. 1C, D and S2E). We postulate the formation of NPs and complexation of bigger microplates (in PBS) requires live cell membrane protrusions and fibrillary surfaces (Figs. S2A–D). It is thought the membrane dynamic topography and associated surface proteins of live cells may act as confined diffusion boundaries for reduction, medium spiny neurons nucleation and seed-mediated growth of anisotropic structures [21]. These cells secrete specific proteins and glucose to buffer low pHinduced stress. These stress proteins are able to block lattice planes required for NP assembly and clustering, which are discussed further in this manuscript. Based on attenuated infrared (IR) spectroscopy, we observed the presence of gold ions triggering more versatile secretion of biomolecules in cancer cells, compared with untreated cultures (Table S3). NP synthesis was not observed in the negative control comprising of a monolayer of dead cells.

To understand the impact of the HAuCl4 concentration on biomineralization and the survival of cancer cells, the concentration of gold ions was varied in between 0.25 mM and 2.0 mM. Under serum containing conditions and after addition of 2.0 mM HAuCl4 intermediate NP aggregates formed. Whereas under serumfree conditions quasi-spherical particles were formed (Fig. S3A and B). In the serum-free culture, the cancer cells treated with 2.0 mM HAuCl4 became irreversibly senescent due to sudden shock (compare serum versus serum-free culture in Supporting Movies 1 and 2). The sudden cellular shock induces a senescence associated secretory phenotype in cancer cells. Under these circumstances, the cells secrete amino acids, such as Serine and Threonine [23]. These amino acids reduce Au+3 in the cell culture media to Au0. The XPS analyses of NPs (Fig. S4) and biomass (Fig. S5) after biomineralization show an increase of Au0/Au+ and Clfractions in cells treated with 1.0 mM gold ions (Table S4).

3.1. Electron microscopy investigation and the crystal formation

After addition of the gold salt to the cells maintained in PBS, Au (III) complexes form and attach to the cell walls, proposed as a mechanism for gold mineralization by Reith et al. [6]. Thus, the biomineralization takes place outside of the cells or on their membrane. Hereby the Au(III) complexes undergo reduction to Au(I) complexes on the cell membranes or fibrils, due to the large electronegative potential of Au3+. This reduction process introduces oxidative stress to the cell membrane, thereby leading to pronounced physical stress conditions to the MCF7 cells. In the absence of serum the oxidative stress to the cell is promoted due to stronger electronegative conditions (compare zeta potential in serumand serum-free conditions in Table S5).

Fig. 1. SEM images of (A) MCF7-reduced NPs in serum media and (B) microplate synthesis in serum-free media. Red arrows indicate a thin microplate with attached spherical gold NP clusters that could initiate ‘‘nucleationand seed-”mediated slow growth of anisotropic microplates (inset shows the enlarged view). (C) X-ray analysis of gold elemental mapping over MCF7 cell surfaces, where inset shows strong gold intensity. (D) Entangled spherical NP-decorated cell surfaces (seed), initiating nucleation over filopodia/lamellipodia cell membrane protrusions. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

In this context the amino acids tyrosine and tryptophan are well known as weak reducing agents and electron donors, assisting in reduction of Au (III) to Au (0) as possible redox chemical reaction [22]:AuCl4þ 3eTryptoph!(an/)Tyrosine AuðsÞ þ 4Cl-Conversely, in the presence of serum, the oxidative stress to the cell membrane is reduced, due to the chemical buffering effect of the fetal calf serum (FCS). The ensuing oxidative stress culminates in the cell compensating by secreting peptides. These peptides induce the anisotropic growth of microplates.To be precise, the Au(I) complexes readily associate with the peptides to further reduce into metallic gold. Hereby, due to attachment of the peptides to certain crystal lattice planes, growth of a particle in this direction is hindered. (Table S5 and Fig. 2B, C). These clusters exhibited strong attachment to the fibrils, which occurs due to surface charge effects of fibrils. Red arrows indicate NP-attached fibrils observed by SEM shown in Fig. 2A. Both microplates and clusters are present at the same time, since the amount of peptides being secreted is limited, (see Fig. 2B).

Fig. 2. Scanning and transmission electron microscopy (SEM/TEM) and UV–vis analysis. (A) Spherical gold NPs (arrows) attached to the fibrils can be seen on the cell surface. (B) Gold nanoprisms with fibrils and gold NPs. (C) Highly oriented and buckled gold plate as seen from the top. (D) Layer-by-layer growth of gold nanoprisms. (E-F) The edge of the prisms images by TEM. (G) A high magnification TEM picture, inset: diffraction pattern along the [1 1 1] direction. (H) Time-dependent UV–vis spectra in serum (solid line) and serum-free condition (dotted line).

Change in cellular metabolism due to pH shock triggers the secretion of specific proteins, which blocks the growth of certain crystal lattice planes. Thus, crystal growth can only occur along certain crystallographic directions, leading to the formation of an anisotropic microplate in the process. Hereby, the microplates exhibit large (1 1 1) facets, as can be observed in Fig. 2. F-G. Lateral growth of the prisms reaches a critical size limit, due to the decrease in gold concentration, whereas horizontal growth seems to occur by the attachment of NPs on the upside of the (1 1 1) facet, leading to a rough surface as shown in Figures (2C, D). Further investigations about the attachment mechanism of NPs are still ongoing. This anisotropic growth mediated by the secretion of defense proteins from cancer cells is novel in context with inbuilt microbial resistance mechanism towards heavy metal ions reported previously [23,24]. It is a noteworthy strategy to scale up biosynthesis of these active anisotropic NPs-microplates for downstream biomedical applications.

3.2. TEM and crystal orientation

TEM analysis of the edge of the sheets (Fig. 2E) shows the rough nature of the edges of the gold prisms. Particle attachment seems to occur as well at the edge of the microplates as it does on the (1 1 1) facets, but in a different rate. These gold NPs seem to be in a liquid gold state and align to the single crystalline prism [19,20]. The near perfect single crystal is confirmed by the Moiré pattern (Fig. 2G). In Fig. 2F, a gold crystal (inside red circle) with a different orientation is lying on top of the gold prism. The gold NP is not attached or oriented to the underlying crystal lattice planes, as evident by a different Moiré pattern. The interplanar lattice plane distances could be measured as 0.247 nm. The bigger red line marks the outside of the sheet with an alteration in crystal orientation, attributed to the attached NPs. The electron diffraction pattern exhibits a crystal texture into (1 1 1) direction. This is caused by the large (1 1 1) facet facing upward and being perpendicular to the beam axis. In conclusion, the growth of nanoprisms by an intermediate liquid gold is in agreement with the work proposed by Shankar et al. [20], however, further work is warranted.

3.3. Cells secret diverse biomolecules and undergo motility arrest in response to gold ions-mediated pH shock

Cellular secretions and biogenic factors released into the extracellular media under cellular shock and stress conditions were analyzed. The ultraviolet–visible (UV–Vis) spectroscopic analysis revealed a time dependent increase of NPs into the cellular media in the presence (Fig. 2H, solid lines) and absence of serum (Fig. 2H, dotted line). The presence of a peak in serum condition represents the spherical gold NPs. However, the serum-free media does not exhibit a similar peak due to triangular and hexagonal prismatic structures. In addition, the concentration of spherical seed particles are undetectable due to being too low.MCF7 cancer cells treated with different concentrations of gold ions were further subjected to Fourier transform infrared (FTIR) spectroscopy in attenuated total reflection (ATR) mode. As reported in Table S3, spectral scanning and signature of cell secretion products exhibited very heterogeneous peaks compared to the control experiments (Fig. S6). Cells incubated in PBS release diverse biomolecular secretions to the extracellular media, which yield in the reduction of gold ions and growth of anisotropic structures.Interestingly, as shown in Table S3, the signature peak for DNAassociated phosphate (DNA-P) was found in the control and culture established with serum, where small spherical NPs are formed. The missing DNA-P signature peak (Fig. 3A and Fig. S6) together with the present a 1, 6–glucans peak at around 850–860 nm from serum-free growth of microplate are unique [25]. It is an indication of either no secretion or complete consumption of secreted nucleic acid/glucose content in reduction and stabilization of a formed ‘‘seed” [18,26]. This can be further corroborated by XPS-linked binding energy of surface elements from NPs and microplates (Table S4). These elements, as electron donor present in functional groups of glucose, protein and DNA could contribute towards reducing the Au ions and stabilizing NPs. The FTIR signatures morphological and biochemical MRI were identical for cells incubated with gold ions in serum and serumfree conditions. Broad signature bands for free amine groups (… 3300 nm) and C = CH2 stretching vibration of carbohydrate (… 850 nm) were observed in cells incubated with gold ions in serum containing medium. Under serum free conditions, a collagen band appeared … 1160 nm with large red shifts in serum supplemented media (Table S3), most likely due to gold-binding and capping [27].

3.4. Cells monolayer exhibit motility arrest and pH induced shock freeze phenomenon

As shown in Supplementary Movie 1, cells in the middle panel treated with 0.25 mM gold ions share similar motility and phenotypic characteristics to cells in the left control panel (no gold ion treatment). In both these panels, cells were grown in serum containing media. Cells shown in the right panel were treated with 1.0 mM gold ions. In this case, the monolayer became unstable after 12 h of incubation. Under serum-free conditions, cells became quiescent and adopted a non-motile state upon the addition of 1.0 mM gold (Supplementary Movie 2, right panel). After 16 h exposure to gold ions, the cells in serum-free media demonstrated nanocolloid synthesis and Brownian motion in confined spaces of the cellular membrane in the presence of gold ions (Supplementary Movie 3). At higher concentrations (2.0 mM) large black clusters were reduced to gold NPs, which accumulated over the cell monolayer.

3.5. Cells exhibits reversible senescence and rescued by Y-27632 ROCK inhibitor

The cell lines MCF7, C2C12, human umbilical vein endothelial cells (HUVECs), along with other cell lines, primary cells, as well as stem cells were screened. The results are shown in supporting Table S6. The advantages of using eukaryotic cells to produce gold NPs for nanotherapeutics compare to bacterial mediated synthesis is nontoxic nature of NPs. The outer core of (Gram negative) bacteria are comprised of lipopolysaccharides and endotoxins which could coat the surface of NPs. The coating can interact with mammalian cells and cause inflammation. Therefore the synthesis of NPs using bacteria is not feasible for biomedical applications. The usage of eukaryotic cells circumvents this issue. Furthermore, the stress and defense proteins formed could be used for the in-situ synthesis of gold NPs by injecting gold directly into tumors. These NPs could be used for photothermal therapy. The method of using eukaryotic cells has the potential to create custom NPs for targeting specific cancer cells. In order to demonstrate this concept, MCF7 human breast cancer cells and C2C12 mouse myoblasts were used. In brief, both natively synthesized NPs and commercially available NPs were added to the culture media. The uptake of the NPs were monitored using inductively coupled plasma optical emission spectroscopy (ICP-OES). We observed that cells exhibit significantly more uptake of natively-synthesized NPs compare to commercial NPs (supporting Table S8).All cell types cultured in serum supplemented and serum free media maintained tight monolayers in the presence of gold ions.

Fig. 3. FTIR spectroscopic analysis and live-dead cell assay results. FTIR Spectra of MCF7 cell treated with 1.0 mM gold ions in serum-free culture (A). Live cell permeant Calcein uptake by MCF7 cells in serum (B), serum-free culture (C), gold ion in serum (D), and serum-free (E) conditions. Red dots show dead cell clusters stained with PI. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Cell viability was assessed using calcein and propidium iodide (PI) (Fig. 3B–E). Dye uptake of cells cultured in serum supplemented media and treated with 1.0 mM gold ions was similar to control cells (Fig. 3B, C vs D, E). The cellular response to stress is an alteration in phenotype, motility, viability and perturbation of chromatin [28]. This is also accompanied by replicative senescence of cells which is induced by a drop in pH of the culture media [29]. Further hallmarks of cell stress include plasma membrane permeability, the presence of apoptotic cells, and nuclear chromatin condensation. MCF7 cells grown in serum-containing media with gold ions stained positive with YO-PROTM and were comparable to the control. In serumfree media with gold ions, cells were more prone to apoptosis as indicated by condensed chromatin with HOECHST-33342 positive staining, as observed in Fig. 4A, and 4B. The senescent/quiescent state of the cells was examined using βgalactosidase (SA-β-gal) activity as a marker. Cells exposed to high doses (2.0 mM) of ionic gold in serum-containing and-free media exhibited high SA-β-gal activity [30] as indicated in Fig. 4C and Fig. S7A, B.Thereafter, MCF7 cells supplemented with gold ions were treated overnight with Y-27632, a selective Rho-associated kinase (Rho/ROCK) inhibitor. This compound selectively delays cell senescence, reverses pH induced apoptosis, and restores the cell cytoskeleton [30,31]. Cells grown in serum containing media and 三 0.5 mM gold ions were rescued with ROCK inhibitor treatment. This was also evident with cells grown in PBS supplemented with 0.25 mM gold ions (Fig. 4D and S8). This signifies that addition of HAuCl4 to serum containing media partly inducessenescence/quiescence; however in serum-free culture, it largely induces senescence. Cells treated with higher concentration (more than 1.0 mM) of gold ions could not be restored in terms of normal cell function and phenotype. In the context of cell senescence [32], we are currently conducting further detailed studies to characterize the mechanism of cell senescence.

3.6. Proteomic analysis and peptide mass finger printing to identify gold reducing, and binding polypeptides

Proteomics analysis of NPs/microplates was carried out by stripping off protein coating the surface as described previously [33,34]. SDS-PAGE analysis revealed consistent expression of specific proteins ranging between 10 kDa and 80 kDa (Fig. S9). These bands were extracted from the gel and further processed for proteomic analysis using matrix-assisted laser desorption
/ionization time-of-flight mass spectrometry (MALDI-TOF).Peptide mass finger printing was used to identify the presence of unknown proteins. MASCOT, a generic mass spectrometry based proteomic analysis tool was used to identify Swiss-prot curated protein sequences. The major proteins obtained from MASCOT search with significant hits (score >50 and expect ratio p 三 0.05) are shown in Table 1. Interestingly some of the proteins extracted from the anisotropic prismatic structures in PBS are listed with metal binding biological functions as reported with the UniProt protein database [35]. The MASCOT histogram and RMS error are shown in Fig. 5A, B. The MS-MS analysis of parent ion 1439.84 Dalton was identified as acetyl glucosamine transferase (ALG14-UDPN). Using similar approach, we further investigated the presence of other proteins on NPs or microplates as reported in Table 1. Comparing with heavy metal resistance related gene expression to reductive precipitation of toxic Au(III)-complexes into bacterial systems, the defense protein reported herein with the cancer systems are novel except heat shock protein family (HSP70) [6]. The commonality of this 70 kDa protein to Au(III) reduction into cancer and bacterial cells is no surprise as the HSP70s (or DnaK) are a conserved family of ubiquitously expressed proteins which protect the prokaryotes and eukaryotic cells from the stress [36]. We also compared the gold binding proteins and peptides expressed by the cancer cells as defense mechanism with those reported with the other Proteins and gold-binding peptides identified from the NPs/microplates and separated with SDS-PAGE. The peptides correspond to the sequences those bind (confirmed with QCM-D) and reduce the gold ions.

Fig. 4. Cell senescence analysis. Chromatin condensation analysis by PI, YO-PROTM and HOECHST 33,342 dye (A). Apoptotic cell quantification (B). Measurement of expression level of senescence associated-β-gal expressions, a biochemical marker of irreversible replicative cell senescence (C). Low doses of gold ions induces reversible cell quiescent state which is reverted via treatment of Rho/ROCK pathway inhibitor Y-27632 as shown with the cytoskeleton elements. (Red-Actin/Green-Tubulin/Blue-DAPI stained Nuclei). (D). Visible aggregates in gold ion treated samples are due to dye integrated into intracellular NPs,quartz crystal microbalance with dissipation monitoring (QCM-D). The interactions between gold and gold reducing polypeptides (good versus bad binders) were examined. Specific peptides exhibiting gold-binding/reducing capability are presented in Table 2 with detailed kinetic coefficients. The net mass addition on the gold QCM crystal due to polypeptide adsorption is reflected by a negative Δf (decrease in frequency). It is plotted as a frequency shift of the sensor signal in as shown in Fig. 5C, which shows data obtained from the gold reducing polypeptide sequence CWQPNPR.

The tested gold-reducing peptides exhibited significantly low desorption during the timescale of the rinse phase, indicating strong gold binding. Specific peptides rich in polar amino acid and a net charge at pH 7 (Table S7), collected from culture media containing no serum were observed as good gold binders/reducers as highlighted in bold in Table 1. These peptides readily form NPs and microplates like anisotropic structures (Fig. S10). In previous reports, simulation and modeling studies indicate the specific polypeptides interacting with gold atomic lattices, which exhibit specific repeats of sequences to form an antiparallel β–sheet [40,41]. The close contact between peptide and gold (1 1 1) surfaces mainly involves the polar side chain of serine and threonine, which places a periodic structure of –OH groups into the regular lattice [40]. These peptides play a vital role in the growth of anisotropic structures. Additionally, these GBPs do not bind to other gold facets as tightly as to the (1 1 1) facets. It happens mainly due to the migration of water molecules through the atomic grooves of the crystallographic surface, which decouples the polypeptide from the surface [41]. Growth of the NPs goes on as fast as reductant proteins with GBP are expressed by the cancer cells. Subsequently gold nanoclusters are formed. The attachment of gold nanoclusters occurs with two competing mechanisms. First, fibrils act as template for the attachment of the gold NPs, which was shown in Fig. 1D and is sketched in Fig. 5D (I). Second, in the presence of GBPs the oriented attachment mechanism as schematically depicted in Fig. 5D (II) takes place. This was also corroborated with the zeta potential measurement (Table S5). In GBPs secreted environment (less negative zeta potential), we see less attachment to the fibrils and the oriented attachment mechanism taking place, i.e. the formation of nanoprisms.

4. Conclusion

This study demonstrates the biomineralization of ionic gold to spherical NPs and microplate like anisotropic structures, a process that is dependent on the composition of the cell culture media. We demonstrated that cells undergo a reversible senescence state at low doses of gold ions. At short time scales (0–8 h), sudden pH change in the surrounding media lead to denaturing of serum protein and cell membrane proteins which triggers a sudden reduction of gold Au+3 to spherical NPs as Au0. At longer time scales (up to a week), the composite protein-gold spherical particles act as nucleation sites where reduction of more AuCl4(-) ions on the Au nuclei takes place (Fig. 5D) and ordered agglomeration occurs. We further show, cells in shock and stress conditions, secrete proteins which show binding preferences for specific facets, promoting anisotropic growth.Further extensive investigation of novel gold-binding/reducing polypeptide sequences from these cell membrane and secreted proteins, which promote the anisotropic triangular prismatic structures. This method can be used as bottom-up assembly of the anisotropic microstructures for plasmonic photo-thermal therapy (PPTT) against tumors and for biosensor development [42]. In vitro cell culture models to fabricate versatile anisotropic structures demonstrated here could be extended to directly injecting ionic gold into tumors in vivo to homogeneously diffuse, in situ reduction, and PPTT applications. Mammalian cell based systems for the synthesis of NP are highly advantageous over bacterial based systems. Firstly, they do not form toxic surface coats such as lipopolysaccharide or endotoxins, as seen with bacterial based systems. Secondly, the gold induced expression of stress and defense proteins to reduce gold KC7F2 ions to metallic NPs could be used for in-situ synthesis of plasmonically active NPs directly inside tumors via injecting gold ions. Finally, nanoparticle-microcarrier designed via unconventional processes may be used for creating personalized drugs for targeting specific cancer cells [43–52].