Improved biostability of gold nanostars for enhanced intercellular interactions

Abstract

Gold nanostars (AuNS) have emerged as promising platforms for biosensing, bioimaging, and therapeutic applications due to their strong plasmonic enhancement and tunable optical properties. However, their applications are hindered by stability issues in complex biological environments, affecting their interactions with biomolecules and cells. To address this challenge, we employ advanced characterization techniques, including surface-enhanced Raman spectroscopy (SERS) to track the stability and intracellular fate of AuNS. We found that AuNS functionalized with 4-mercaptobenzoic acid, acting as SERS nanotags, exhibit exceptional biostability in cell culture media, both uncoated and precoated with bovine serum albumin (BSA). High-resolution 2D and 3D SERS imaging revealed cell-type-dependent uptake, with immune cells displaying greater internalization than cancer cells, while protein precoating had minimal influence. Cytocompatibility assays confirmed low toxicity and suitability of AuNS for biological applications. Our findings highlight the potential of improved biostability of AuNS for precise intracellular probing, paving the way for advanced nanoparticle-based applications that would enable spatially resolved biomolecular analysis in live cells.

---

Introduction

Despite the vast potential of plasmonic nanoparticles in cellular imaging and biosensing, their translation into practical applications remains challenging due to the complex interactions with biological media.^1–3 Nanoparticles introduced into biological fluids encounter a dynamic and heterogeneous environment consisting of proteins, lipids, and other biomolecules, which can alter their physicochemical properties, influence aggregation tendencies, and ultimately affect their cellular uptake and distribution.^4–6

One of the key limitations in nanoparticle-based biological applications is the formation of a biomolecular corona – layers of biomolecules most often proteins – which can significantly modify surface characteristics, reducing reproducibility and specificity in targeting. The efficiency of nanoparticle delivery to target sites, such as cellular uptake in vitro, is influenced by these stability issues.^7–9 The biomedical studies and translation of plasmonic nanoparticles into clinic practice have been further complicated by the validity of the enhanced permeability and retention effect. Therefore, understanding cellular uptake is essential for maximizing the efficacy and safety of nanoparticle-based biomedical applications.

While spherical nanoparticles have been extensively studied for clinical assays,^10–13 anisotropic nanostructures such as gold nanostars (AuNS) are not well explored due to the limit understanding of their behaviour in complex biological systems. AuNS offer promising alternatives due to their enhanced stability and resistance to protein fouling as well as their unique plasmonic properties and geometry advantages.^6,14 These advantages suggest that AuNS may serve as robust platforms for intracellular applications, maintaining their integrity and function within complex biological systems.^6,14–19

In this study, we systematically investigate the stability and cellular interactions of AuNS in the presence of cell culture media (CCM), both with and without bovine serum albumin (BSA) precoating. By examining their optical and functional properties under biologically relevant conditions, we aim to determine the extent to which AuNS retain their biostability and ability to facilitate cellular uptake. Additionally, we explore the intracellular distribution of AuNS-functionalized with 4-mercaptobenzoic acid (MBA), named as SERS nanotags, leveraging high-resolution Raman 2D and 3D mapping to provide spatially resolved insights into nanoparticle fate within cells. This work highlights the potential of biostable AuNS as effective intracellular imaging agents, offering a more reliable and structurally consistent platform for biosensing applications.^20–23 By demonstrating that AuNS maintain their functional integrity in biologically complex environments, our findings contribute to the ongoing efforts to optimize plasmonic nanomaterials for advanced biomedical applications, particularly in cellular uptake studies and intracellular diagnostics.²⁴

Results and discussion

Characterization of AuNS in different media

AuNS were synthesized using a well-established one-pot method,^14,25 followed by functionalization with BSA, MBA, or sequentially with both MBA and BSA. MBA was chosen for AuNS functionalisation due to its strong Raman signal and ability to form a stable self-assembled monolayer via thiol–gold bonding.^14,25 The terminal carboxylic group also enhances electrostatic repulsion and hydrophilicity, improving colloidal stability of AuNS@MBA. Additional BSA coating provides steric stabilisation and mimics physiological conditions, further preventing aggregation. While other thiolated reporters and targeting ligands exist, this study focused on evaluating nanoparticle stability and cellular interactions in CCM, with and without BSA, without the influence of targeting ligands. Initial characterization of the AuNS was conducted using various techniques including UV-visible spectroscopy, TEM, NTA, and surface charge measurements. These analyses were carried out in both water and CCM to assess the properties and morphology of the AuNS.

The surface plasmon resonance (SPR) bands of the AuNS were determined using UV-visible spectroscopy. The main SPR band for AuNS was identified at 715 nm. This band exhibited a shift towards the near-infrared (NIR) region upon functionalization, indicating the formation of molecular layers around the AuNS and the subsequent alteration of the refractive environment surrounding nanoparticles (Fig. 1A). TEM images were obtained to evaluate the shape and morphology of the AuNS in both water and after incubation in CCM (Fig. 1B and C). The images revealed that plasmonic nanostructures preserve spiky morphology. The TEM analysis demonstrated well-dispersed AuNS with an average size of approximately 100 nm, consisting of a core around 50 nm and tips ranging from 20 to 40 nm.

Fig. 1 Characteristics of the gold nanostars (AuNS): (A) normalised extinction spectra of AuNS upon BSA and MBA absorption and change of surrounding media; (B) TEM images of AuNS in water and (C) in cell culture media; (D) size of AuNS after different types of functionalization; (E) surface charge of the AuNS after different types of functionalization and control samples (BSA in water, CCM and BSA in CCM). Error bars indicate standard deviation of 3 replicate measurements.

NTA measurements indicated increase in size (diameter) following nanoparticle functionalization, with an average particle size increase of 20 nm relative to unfunctionalized nanostars (Fig. 1D). It is important to note that the particle size variation arises from structural heterogeneity intrinsic to nanostars, rather than sample inconsistency. AuNS are characterized by a relatively uniform core size but variable tip lengths. This anisotropic structure contributes to a broader size distribution when assessed by techniques sensitive to hydrodynamic diameter or overall particle dimensions (e.g., NTA, DLS). Importantly, the variability within individual NTA measurements remains small (on the order of a few nm), indicating reproducibility and batch-to-batch uniformity. This is further supported by NTA data provided in the Table S1. The surface charge of the AuNS was also assessed (Fig. 1E). Initially, the water solution of AuNS exhibited a negative surface (z-potential) charge of −38.1 mV, attributed to the presence of ascorbate capping ions. Upon redispersion in CCM, the net charge increased (became less negative, −4.8 mV), likely due to the adsorption of cationic proteins from FBS and of other ions present in the CCM. For AuNS functionalized with BSA (AuNS@BSA), a significant increase in surface charge to 10.16 mV was observed in water. However, redispersion in CCM resulted in a reduction of the surface charge to −12.7 mV, due to the presence of salt ions in the medium. The AuNS functionalized with MBA (AuNS@MBA) maintained a net negative charge of −22.5 mV due to the presence of carboxylate (–COO⁻) ions. Resuspension of AuNS@MBA in CCM led to a slight increase in surface charge to −6.3 mV, likely due to interactions with proteins. Similar changes in surface charge were exhibited after the sequentially functionalized AuNS@MBA@BSA samples were dispersed in water and CCM.

Overall, the exposure of SERS nanotags (AuNS@MBA) to CMM, both with and without protein coating, did not lead to particle aggregation. The primary properties of the SERS nanotags, including their structural integrity and functional capabilities, are preserved under these conditions. We have examined the enhancement factor (EF) of SERS nanotags, which did not undergo significant changes upon dispersion in CCM (Fig. S1 and Table S2). This stability is critical for their effective application in biological environments, ensuring consistent performance in cellular uptake studies and minimizing variability due to aggregation effects. It should be noted that varying BSA concentration during the pre-coating step does not significantly affect the optical properties or colloidal stability of AuNS, as we have previously demonstrated.⁶

Cell viability after exposure to SERS nanotags

We further tested cellular interaction of SERS nanotags with and without protein coating (AuNS@MBA@BSA, AuNS@MBA). Cell viability was assessed at 24 and 48 h post-exposure to the AuNS using the CCK-8 cell viability assay. J774 and MCF7 cells were cultured and incubated in a 96-well plate, with each sample tested in triplicate. The experimental setup included a negative control with untreated cells, a positive control with cells treated with 20% DMSO, and three wells where AuNS were incubated without cells (to remove possible absorption background of AuNS).

In the CCK-8 assay, WST-8 (water-soluble tetrazolium salt) is reduced by dehydrogenases in live cells, resulting in the formation of an orange-coloured product (formazan). The amount of formazan produced is proportional to the number of live cells.²⁶

As indicated in Fig. 2, the positive control (20% DMSO treatment) significantly decreased cell viability, with J774 cells exhibiting approximately 20% viability. When J774A cells were exposed to AuNS at various concentrations, both with and without BSA, cell viability ranged from 35% to 38%, with no statistically significant differences observed between the samples (p > 0.05 with and without BSA, Fig. 2A). In contrast, the MCF7 cells maintained a higher viability of 70% to 90% when incubated with both protein-coated and uncoated SERS nanotags (Fig. 2B, also without significant difference between SERS nanotags samples). This suggests that immune cells (J774) were more adversely affected by the presence of SERS nanotags compared to cancer cells (MCF7).

Fig. 2 Effect of nanoparticles on cell viability after 24 h and 48 h of exposure using CCK-8 assays: (A) J774 cells incubated with nanoparticles for 24 h; (B) MCF7 cells incubated with nanoparticles with a range of concentration 0.25–1 × 10⁸ particles per well for 24 h; (C) J774 cells incubated with nanoparticles for 48 h; (D) MCF7 cells incubated with nanoparticles for 48 h; each bar indicates the mean  ±  error standard from 3 biological replicates. 70% cell viability relative to control level is annotated with dashed line. Both one-way ANOVA and Tukey tests results showed no significant difference in cell viability between results for AuNS and AuNS + BSA treatments in cases of J774 or MCF7 cells both after 24 and 48 h incubation periods (p  >  0.05).

After 48 h of incubation with AuNS, a significant increase in cell viability was observed in both cell lines, reaching nearly 100% (Fig. 2C and D). We hypothesize that this increase in viability is due to the reduction in the number of cells containing AuNS. As the surviving cells proliferate, they do not necessarily excrete and reuptake the SERS nanotags,²⁷ meaning that newly formed cells are not exposed to the nanoparticles or exposed to much less of them. The extent of AuNS uptake appears to be the primary factor contributing to the observed cellular viability. It is possible that the amount of AuNS uptake by MCF7 cells during the 24-hour was insufficient to induce the strong toxic effects that were observed in J774 cells.

Cellular uptake of SERS nanotags probed by confocal Raman mapping

2D Raman mapping. Next, we investigated the cellular uptake of SERS nanotags using confocal Raman mapping. It is important to note that the analysis of SERS nanotags uptake was based on the intensity of the signal assuming that the observed signal results from the scattering of individual nanoparticles, without accounting for possible aggregation, which could cause hot-spot effects and signal amplification. This assumption is justified by our preliminary characterization of nanotags in CCM (Fig. 1A), which did not indicate any aggregation events.

For the incubation with cells, we selected the lowest concentration tested in the cell viability assays (0.25 × 10⁸ particles per well), resulting in approximately 4 × 10³ particles per cell. In this study, the selected concentration range was deemed sufficient to demonstrate the onset of cytotoxicity and to identify a dose suitable for downstream uptake studies. The lowest tested concentration was specifically chosen to remain within a viability range while still eliciting a measurable reduction in cell viability. For the uptake analysis, this minimally toxic concentration was selected intentionally to avoid confounding effects related to cell death or stress responses. Confocal Raman mapping of single cells was performed using a Raman spectrometer within a scan area of 50 × 50 μm with the scan step size set to 2 μm, ensuring that each data point represented a 2 μm square within the mapped region (Fig. 3A and B). Raman intensity data for each pixel was specifically acquired at the peak position of 1079 cm⁻¹, which corresponds to the most prominent vibrational signal of MBA, ring-breathing coupled with the –S–C vibrational modes. SERS maps were analysed, and the SERS signal intensity was calculated for each cell line as an average of three separate measurements (Fig. 3C). For the purpose of intensity analysis, these Raman intensity values were depicted in an 8-bit grayscale format, allowing for precise quantification and comparison of the vibrational signal strength across the scanned area. The SERS intensity of each pixel at the peak position of 1079 cm⁻¹ was mapped and the sum of SERS intensity from all pixels in the area indicates the SERS signal strength in the selected area and correlates with the amount of SERS nanotags present. Examples of the spectra, SERS intensity and optical images are presented in Fig. S2. According to our 2D Raman mapping results, J774 cells exhibited significantly higher SERS signal intensities than MCF7 cells after 24 hours of incubation, indicating more efficient internalization of the nanotags. After 48 hours, a decrease in SERS intensity was observed in both cell lines, reflecting a reduction in the amount of SERS nanotags retained in the cells over time. The trend in cellular uptake closely mirrors the viability results (Fig. 2): J774 cells showed higher toxicity, likely due to their greater nanoparticle internalization, whereas MCF7 cells were less affected. Due to this inherent phagocytic activity, it is expected that J774 cells would internalize a significantly greater number of nanoparticles compared to non-phagocytic MCF7 breast cancer cells, resulting in the observed differences in uptake and associated cytotoxicity.²⁸ This reinforces that cell-type-dependent uptake efficiency plays a critical role in determining the cytotoxic effects of SERS nanotags.²⁹ This reduction may be attributed to a shift in the intracellular distribution of SERS nanotags, concomitant with the proliferation of viable cells. As the cell population expands, the overall quantity of nanoparticles per cell decreases.

Fig. 3 Confocal Raman mapping for single cells after 24 h and 48 h of exposure to SERS nanotags: (A) J774 cells incubated with AuNS@MBA and AuNS@MBA@BSA for 24 h and 48 h; (B) MCF7 cells incubated with AuNS@MBA and AuNS@MBA@BSA for 24 and 48 h; each measurement was done in 3 replicates (i), (ii) and (iii) on 3 different cells of each type. The scale bar on the right represents the SERS intensity calibration scale, with intensity values assigned to corresponding intensity grades. (C) Average sum of relative SERS intensity from 1079 cm⁻¹ peak position (average sum of SERS intensity divided by the enhancement factor of nanotags, each form 3 biological replicates), which indicate SERS signal strength from single cell mapping area. Statistical significance was evaluated via one-way ANOVA and Tukey tests (*p  <  0.05, **p  <  0.05, ***p  <  0.0005, ****p  <  0.0001, ns – p > 0.05).

Multiphoton imaging. To support our findings, we performed multiphoton microscopy of the cells after 24 hours of incubation with nanostars. Cells were labelled with CellMask™ Deep Red dye, which was excited at 1100 nm, and its emission was collected between 650–700 nm. AuNS were excited at 800 nm, and their emission was collected in the 440–490 nm window, enabling simultaneous visualization of cell morphology and nanoparticle distribution. Image analysis was conducted by quantifying nanostar-associated signals located on the cell surface and within the cytoplasm. Signal intensity was normalized to cell number and expressed as pixels per cell (arbitrary units). The images indicate the intracellular presence of nanostars. The results (Fig. 4) support our observation that the cellular uptake of anisotropic AuNS is more strongly influenced by the type of cell rather than by the presence or absence of a protein coating on the nanotag, with immune cells (J774) exhibiting increased uptake.

Fig. 4 Cellular uptake of AuNS in J774 and MCF7 cells after exposure to SERS nanotags. Relative particle signal per cell was measured after 24 h incubation with AuNS@MBA or AuNS@MBA@BSA in J774 and MCF7 cells. Untreated cells were used as controls. (A) Single channel with AuNS-only visualised; (B) CellMask™ Deep Red dye and AuNS channels visualised simultaneously; (C) J774 cells showed significantly higher uptake compared to MCF7 cells for both formulations. BSA coating further increased uptake in J774 cells but had no significant effect in MCF7 cells. All AuNS-treated groups displayed significantly higher signals than their respective controls. Data are shown as mean ± SD (n = 2). Statistical analysis was performed using unpaired two-tailed t-tests. *p < 0.05, ***p < 0.001, ****p < 0.0001, ns = not significant.

3D Raman mapping. To verify that the measured Raman signal originates from the cells, 3D Raman mapping was performed on the sample cells (MCF7 and J774) incubated with AuNS@MBA and AuNS@MBA@BSA for 24 h. The results, presented in Fig. 5, indicate that the nanoparticles responsible for generating the MBA signal (Fig. 5A) are either internalized within the cells or localized on the cellular membrane. Furthermore, nanoparticles that were not internalized were effectively removed by flushing the slides, confirming the specificity of the signal to intracellular or membrane-associated nanoparticles. The small portions of signal that appear outside the visible cellular boundary are likely attributable to thin or transparent cellular extensions (e.g., filopodia, lamellipodia) or cell-associated debris that are not easily visible in bright field images, particularly after chemical fixation. Since the cells were fixed and adhered to the substrate, some structures may become flattened or optically faint yet still contribute to the Raman signal detected during mapping.³⁰

Fig. 5 (A) Representative Raman spectra of background, cell without SERS nanotags, and cell with SERS nanotags in 3 different spots (after initial baseline correction). Grey stripe indicates the MBA peak (at 1079 cm⁻¹), that was used to map SERS nanotags position inside the cells; cells contours were mapped using spectra recognition pattern installed in control Four software (WITec, Germany). (B) 3D Raman maps of the cells and SERS nanotags in yz and xy projections; and bright field images of cells with selected mapping area (green square). Microscope image of the mapping area (45 × 45 μm) obtained with a 100× objective lens (scale bar 20 μm).

Overall, these results suggest that the cellular uptake and subsequent toxicity of star-shaped SERS nanotags are influenced more by the type of cell line than by the presence of a protein coating. The increased uptake of SERS nanotags by immune cells is expected, as these cells function as frontliners and are predisposed to interact with potential external pathogens. Conversely, cancer cells exhibited relatively low uptake of SERS nanotags. BSA coating seem to have no significant improve on cellular uptake of star-shaped SERS nanotags by MCF7 cells. This indicates that SERS nanotags may require a targeting ligand for efficient delivery to cancerous cells. Additionally, although sub-micron precision allows for the discrimination of signal origin across cell compartments, including the plasma membrane and cytoplasmic regions, definitive intracellular localization (e.g., endosomal vs. membrane-bound) would require complementary imaging techniques, such as co-staining with organelle-specific dyes or colocalization by confocal microscopy.

Conclusions

In summary, we have investigated the biostability of gold nanostars in biological environments for their cellular uptake under biologically relevant conditions. The intracellular distribution of AuNS functionalized with 4-mercaptobenzoic acid (SERS nanotags) was explored using high-resolution Raman 2D and 3D mapping to provide spatial insights into nanoparticle fate within cells. Our results demonstrate that the exceptional biostability and cytocompatibility of AuNS in biological environments reinforces their potential for intracellular biosensing applications. Our findings suggest that cellular uptake of anisotropic AuNS is more influenced by the type of cell than by the presence or absence of protein coatings on the nanotag. Immune cells (J774) showed increased uptake of AuNS, which is consistent with their roles as frontline protectors which are predisposed to interact with potential external pathogens. In contrast, cancer cells (MCF7) exhibited relatively low uptake of AuNS, regardless of whether they were protein-coated or not. The differential uptake between immune and cancer cells underscores the importance of tailored nanoparticle delivery strategies to optimize their efficacy in diverse cellular contexts. Moreover, the ability of high-resolution SERS imaging to provide spatially resolved insights into nano–cell interactions highlights its value for advanced biosensing. Expanding these capabilities to real-time 3D Raman mapping could further elucidate intracellular processes, advancing the development of nanoparticle-based diagnostic and therapeutic platforms.

Experimental

Nanostars (AuNS) preparation

Chemicals including 1% chloroauric acid (HAuCl₄), silver nitrate (AgNO₃), L-ascorbic acid (C₆H₈O₆), hydrochloric acid (HCl) and nitric acid (HNO₃) solutions, were purchased from Sigma-Aldrich (Sydney, Australia). Ultra-pure water (Milli-Q, 18.2 MΩ cm at 25 °C) was used for all solutions.

To synthesize AuNS, 10 mL of water and 20 μL of 10 mM of AgNO₃ were mixed with 360 μL of 10 mM of HAuCl₄ and stirred for 30 s. Subsequently, 80 μL of 100 mM C₆H₈O₆ were added dropwise while stirring continuously for another 30 s. The solution turned blue, indicating the formation of nanostars. The synthesized particles were stored at 4 °C in a refrigerator overnight before further use.

To coat AuNS with bovine serum albumin (AuNS@BSA), 1 mL of as prepared AuNS (approximately 1 × 10⁹ particles per mL – determined by NTA) were incubated with 0.5 mL of 4.5 mg mL⁻¹ BSA (Sigma, Sydney, Australia, A9647, heat shock fraction) solution for 30 min at room temperature. After incubation, samples were centrifuged at 5000 rpm at 4 °C for 10 min. The supernatant (containing free protein) was carefully removed and discarded. The pellets were suspended in 1 mL of water and briefly sonicated for a few seconds. The centrifugation–resuspension cycle was repeated twice. The prepared samples were stored at 4 °C.

To coat AuNS with a Raman reporting molecule – 4-mercaptobenzoic acid (AuNS@MBA) – 5 μL of 1 mM MBA/ethanol solution was added to 1 mL of AuNS (1 × 10⁹ particles per mL) and incubated for 2 h under constant shaking at 40 rpm. The samples were centrifuged and resuspended in water to remove unbound molecules. The centrifugation–resuspension cycle was repeated twice. AuNS@MBA@BSA samples were prepared by incubating AuNS@MBA samples with BSA as described above.

After the final centrifugation–resuspension cycle, samples were either resuspended in Milli-Q water or cell culture media (CCM) (DMEM, Dulbecco's Modified Eagle Medium + 10% fetal bovine serum + 1% penicillin/streptomyosin, supplied from Thermofisher, Australia).

UV-visible spectroscopy

Optical properties of AuNS were assessed with UV-visible spectroscopy. Spectra were obtained over a wavelength range of 500–900 nm using a JASCO V-760 UV-vis Spectrometer (Jasco, Hachioji, Japan). For each measurement, 700 μL of the sample was placed in a plastic disposable semi-micro cuvette (BrandTech, US).

Transmission electron microscopy (TEM)

The morphology of AuNS was evaluated using a Philips CM10 TEM (Eindhoven, The Netherlands). TEM samples were prepared by placing a 10 μL drop of nanoparticle suspension on the carbon side of a carbon-coated copper grid (Zhongjingkeyi Film Technology, Beijing, China). After allowing the drop to settle for 10 min, the excess liquid was removed with filter paper. The sample was left to dry overnight before the analysis.

Nanoparticle tracking analysis (NTA)

The concentration and size of the nanoparticles were measured using the NanoSight NS300 device (Malvern Panalytica, Malvern, UK). The samples were diluted to 1/100 of their original concentration and then introduced to the detector at a flow rate of 100 units (instrument software units) on the syringe pump, at a temperature of 21 °C. A 405 nm laser was used to irradiate the particles during analysis. The results are presented as an average of 3 runs of the same sample.

Surface charge: zeta potential

The surface charge of the nanoparticles was determined using a Zetasizer ZS (Malvern Panalytica, Malvern, UK). For the zeta potential measurements, 1000 μL of each sample were placed in disposable folded capillary cells (DTS1070). Each sample was tested three times, with 15 runs per test.

Surface-enhanced Raman spectroscopy (SERS)

SERS spectra were acquired using a portable IM-52 Raman microscope (Snowy Range Instruments, USA) with a 785 nm incident laser power of 100 mW and an integration time of 1 s. For each sample, 60 μL was placed in a Starna 18B/Q/10 quartz cuvette, and 10 spectra were recorded and averaged for the result. The spectra were baseline-corrected using Peak software (Snowy Range Instruments, Laramie, WY, USA).

Cells culturing

Two cell lines were used: J774-A.1 (mouse macrophages, referred to as J774), which were isolated from a tumour in a female BALB/c mouse, and MCF7, which is human breast cancer cell line that has glucocorticoid, progesterone, and oestrogen receptors. The cells were purchased from American Type Culture Collection (ATCC).

RPMI (#R8758, Sigma Aldrich) or DMEM (#12491015 Thermofisher, Australia) medium combined with 10% foetal bovine serum (FBS#10099141, Thermofisher, Australia) and 1% of penicillin–streptomycin (#P4333, Sigma Aldrich) were used for culturing the cells. Cells were incubated in 50 mL falcon flasks at the 37 °C and 5% CO₂ environment. The cells were passaged every 2–3 days after they reached 80% confluency. To passage the MCF7 cells, the cells were washed by gently flushing 5 mL of Dulbecco's phosphate-buffered saline (DPBS, # 14190144 Thermofisher, Australia) over them. After washing, 1 mL of pre-warmed to 37 °C dissociation reagent TrypLE (Thermofisher, Australia), was added to the side of the flask.

For J774 cells passaging we used a slightly modified protocol to avoids enzymatic damage by trypsin, preserving the cells’ functional integrity. Thus, for cells detachment we used sterile cells scraper. The total number of cells and percent viability were determined using an automated cell counter TC20 (BioRad, Australia), and trypan blue dye within Cell Counting Kit (dual-chambered slides with trypan blue dye #1450003, BioRad, Australia).

Cell viability assay

The cytocompatibility of AuNS and composites were assessed via a cell viability colorimetric assay. After passage the cells were transferred from falcon tubes to 96 well plates (Greiner, clear wells, flat bottom, 3 × 10³ cells per well) and incubated for 24 h. AuNS in CCM with concentrations of 1 × 10⁸, 0.5 × 10⁸, and 0.25 × 10⁸ particles per well, were added to replace the previous medium 24 h after the cells had been seeded in the wells with the requisite cell density. Nanoparticles concentration was evaluated with NTA and redispersed in CCM accordingly.

The cytotoxicity of AuNS to the J774 and MCF7 cell lines was examined using a Cell Counting Kit-8 (CCK-8) (Dojindo Laboratories, Kumamoto, Japan). The water-soluble tetrazolium salt (WST-8), which is the active component of CCK-8 assay, is enzymatically broken down to produce a water-soluble formazan dye. The number of live cells, which can be determined using absorbance at 450 nm, are proportional to the amount of the formazan dye produced by dehydrogenases in cells. Only live cells will carry the enzyme to break down the tetrazolium salt to produce formazan dye. Depending on the experiment, the cells were incubated for a further 24 or 48 h after the addition of AuNS. Each experiment was performed in triplicate. In all CCK-8 tests, 20% dimethyl sulfoxide (DMSO; Sigma-Aldrich) was employed as the negative control. Untreated cells were employed as a positive control.

The CCK-8 reagent was added to each well and incubated for 2 hours. Using a multi-well plate reader Synergy H1 apparatus (BioTech Instruments, USA) and Gene 5 software (BioTek Instruments, USA), the absorbance at a wavelength of 450 nm was measured after two hours. Since AuNS also absorb light in the visible spectrum, we subtracted their absorbance (at different concentrations for each sample, to match the amount of AuNS added to cells).

The cell viability was determined using the formula:

where A_S is the absorbance of AuNS + cells; A_P is the absorbance of the AuNS in media without cells; A_B is the absorbance of the untreated cells.

Statistical analysis

Statistical analysis was performed using GraphPad Software (GraphPad Software, San Diego, CA, USA). All data were acquired using three biological replicates. For normally distributed data, a one-way analysis of variance (ANOVA) was conducted, followed by Tukey's post hoc test for multiple comparisons. For non-normally distributed data, the Kruskal–Wallis test was used. Data are presented as mean ± standard deviation (SD). A p-value <0.05 was considered statistically significant.

Cellular uptake and 2D and 3D Raman mapping

To investigate the cellular uptake of AuNS, J774 or MCF7 cells were seeded in a Lab-Tek II system chamber slides 8 chamber (#154534, Thermofisher, Australia) at a density of 6000 cells per well (determined via cell counting kit with Typan blue #1450003, Bio-rad, Australia) in 0.5 mL of DMEM medium. The cells were incubated at 37 °C with 5% CO₂ for 24 h before introduction of SERS nanotags. Following this, the cells were treated with AuNS@MBA or AuNS@MBA@BSA at the lowest concentration used in cell viability tests, specifically 0.25 × 10⁸ particles per well, for either 24 or 48 h. The cell culture media was then discarded, and the cells were washed three times with DPBS. The cells were then incubated with 4% formaldehyde solution for 10 min and further washed twice with DPBS and water. These slides with cells were subsequently used for confocal Raman mapping of single cells to analyze the cellular uptake of AuNS.

Confocal Raman mapping of single cells was conducted using a LabRAM HR Evolution Raman Spectrometer (Horiba Scientific, Japan) equipped with a 50× objective lens and Duoscan mode. A 50 × 50 μm scan area was selected for analysis using a 633 nm excitation laser set to 15 mW power. The scan step size was set at 2 μm, with each square point measuring 4 × 4 μm², resulting in a total of 156 data points (pixels). Raman detection for each pixel was performed with an integration time of 2 s, 1 accumulation, and a Raman shift range of 400–1800 cm⁻¹. This Raman mapping method of visualizing SERS nanotag inside cells was also employed previously reported.³¹ SERS spectra acquired during Raman mapping were baseline-corrected prior to signal intensity analysis. The SERS intensity of each pixel at the peak position of 1079 cm⁻¹ (attributed to the strongest vibrational signal of MBA, ring-breathing coupled with the –S–C vibrational modes) was taken and normalised by the enhancement factor (SI, Table S2) of SERS nanotags in CCM. The average sum of SERS intensities from all pixels in the mapping area represents the SERS signal intensity in the selected region and correlates with the amount of SERS nanotags present.

3D Raman mapping of cells were acquired using an upright Confocal Raman microscope (alpha300R, WITec GmbH, Germany) with a 100× objective lens. The sample was excited with a 633 nm laser with the power set to 25 mW and an integration time of 1 s. 3D Raman mapping was performed with a spatial resolution of 90 points per line and 90 lines per image, yielding 8100 points per 2D scan over a 45 × 45 μm area. Depth profiling included 6 layers spanning a stack depth of 6 μm. The laser spot size of 306 nm ensured sub-micrometer precision, resulting in a detailed 3D dataset of 48 600 points. The z-stag spectral profiles were analysed in Control Four (WITec, Germany) acquisition software and visualized using Fiji and Blender 4.3 software.

Multiphoton imaging

J774 and MCF7 cells were prepared and incubated with nanoparticles as per protocol described in “Cellular uptake and 2D and 3D Raman mapping” section. To visualize the cell membrane, cells were incubated with CellMask™ Deep Red (5 μg mL⁻¹ in serum-free medium; Thermo Fisher Scientific) for 10 minutes at 37 °C. After staining, cells were rinsed twice with PBS, fixed with 4% paraformaldehyde for 15 minutes at room temperature, and washed again with PBS before mounting.

Imaging was performed using a multiphoton microscope (STELLARIS DIVE, Leica Microsystems). Nanostars were excited at 800 nm. Although nanostar scattering signals were broadly detectable across the 400–700 nm range, the 440–490 nm emission window was selected for signal collection as it was distinct from the emission spectrum of CellMask™ Deep Red cell membrane stain (excitation at 1100 nm, and emission between 650–700 nm), enabling simultaneous visualization of cell morphology and nanoparticle distribution.

ImageJ was used for image analysis, including cell counting and AuNS signal quantification. Brightness and contrast were uniformly adjusted for display purposes only, all statistical analyses were performed on raw data. Images were analyzed to quantify cell-associated nanostars including that located on the cell membrane and intracellularly. Signal intensity was normalized to cell number and expressed as pixels per cell (arbitrary units, based on laser settings). Statistical comparisons between groups were performed using an unpaired two-tailed Student's t-test.

Author contributions

AT contributed to this work with design of experiments, experiments conduction, data analysis, writing of the original draft preparation. SH contributed to cellular experiment and data analysis, AR and YW contributed to this work with project supervision, data analysis and writing – revision and editing of the draft.

Conflicts of interest

There are no conflicts to declare.

Data availability

The authors confirm that the data supporting the findings of this study are available within the article [and/or] its SI.

Supplementary information including Tables for NTA size evaluation, Calculated EF and Resolution parameters for both 2D and 3D Raman mapping; as well as Figures for SERS spectra of AuNS and AuNS based tags in different media, SERS spectra and Raman mapping of a single cell, is available. See DOI: https://doi.org/10.1039/d5tb00523j

Acknowledgements

A. T. acknowledges funding support from the International Macquarie University Research Excellence Scholarship (iMQRES). A. R. and Y. W. acknowledge the funding support from the Australian Research Council (ARC) for the Discovery Project (DP220102086). Y. W. acknowledges funding support from ARC Future Fellowship (FT210100737). M. L. acknowledges funding support from ARC Future Fellowship (FT220100092). This work was supported by the Australian Research Council Industrial Transformation Training Centre for Facilitated Advancement of Australia's Bioactives (IC210100040). The authors would like to acknowledge the Macquarie University Faculty of Science and Engineering Microscopy Unit for access to its instrument and staff. The authors acknowledge the technical and scientific assistance of Sydney Microscopy & Microanalysis, the University of Sydney node of Microscopy Australia. The authors also thank the UNSW Katharina Gaus Light Microscopy Facility (KGLMF) and their staff for providing access to the multiphoton microscope and for their helpful consultation.

References

1. L. Lin, X. Bi, Y. Gu, F. Wang and J. Ye, J. Appl. Phys., 2021, 129, 191101.
2. K. Park, J. Controlled Release, 2019, 305, 221.
3. V. C. Yang, J. Controlled Release, 2019, 311–312, 322.
4. Z. Ban, P. Yuan, F. Yu, T. Peng, Q. Zhou and X. Hu, Proc. Natl. Acad. Sci. U. S. A., 2020, 117, 10492.
5. C. Carnovale, G. Bryant, R. Shukla and V. Bansal, ACS Omega, 2019, 4, 242.
6. A. Tukova, Y. Nie, M. Tavakkoli Yaraki, N. T. Tran, J. Wang, A. Rodger, Y. Gu and Y. Wang, Aggregate, 2023, 4, e323.
7. S. Sindhwani, A. M. Syed, J. Ngai, B. R. Kingston, L. Maiorino, J. Rothschild, P. MacMillan, Y. Zhang, N. U. Rajesh, T. Hoang, J. L. Y. Wu, S. Wilhelm, A. Zilman, S. Gadde, A. Sulaiman, B. Ouyang, Z. Lin, L. Wang, M. Egeblad and W. C. W. Chan, Nat. Mater., 2020, 19, 566.
8. C. Lopez-Chaves, J. Soto-Alvaredo, M. Montes-Bayon, J. Bettmer, J. Llopis and C. Sanchez-Gonzalez, Nanomedicine, 2018, 14, 1.
9. P. Foroozandeh and A. A. Aziz, Nanoscale Res. Lett., 2018, 13, 339.
10. N. A. Monteiro-Riviere, M. E. Samberg, S. J. Oldenburg and J. E. Riviere, Toxicol. Lett., 2013, 220, 286.
11. S. G. Elci, Y. Jiang, B. Yan, S. T. Kim, K. Saha, D. F. Moyano, G. Yesilbag Tonga, L. C. Jackson, V. M. Rotello and R. W. Vachet, ACS Nano, 2016, 10, 5536.
12. J. Piella, N. G. Bastús and V. Puntes, Bioconjugate Chem., 2017, 28, 88.
13. G. Bashiri, M. S. Padilla, K. L. Swingle, S. J. Shepherd, M. J. Mitchell and K. Wang, Lab Chip, 2023, 23, 1432.
14. A. Tukova, I. C. Kuschnerus, A. Garcia-Bennett, Y. Wang and A. Rodger, Nanomaterials, 2021, 11, 2565.
15. A. M. Fales and T. Vo-Dinh, J. Mater. Chem. C, 2015, 3, 7319.
16. B. Khlebtsov and N. Khlebtsov, Nanomaterials, 2020, 10, 2228.
17. B. Andreiuk, F. Nicolson, L. M. Clark, S. R. Panikkanvalappil, Kenry, M. Rashidian, S. Harmsen and M. F. Kircher, Nanotheranostics, 2022, 6, 10.
18. I. B. Becerril-Castro, I. Calderon, N. Pazos-Perez, L. Guerrini, F. Schulz, N. Feliu, I. Chakraborty, V. Giannini, W. J. Parak and R. A. Alvarez-Puebla, Analysis Sensing, 2022, 2, e202200005.
19. M. Sivis, N. Pazos-Perez, R. Yu, R. Alvarez-Puebla, F. J. García de Abajo and C. Ropers, Commun. Phys., 2018, 1, 1.
20. S. Tanwar, J. H. Kim, J. W. M. Bulte and I. Barman, Wiley Interdiscip. Rev.: Nanomed. Nanobiotechnol., 2022, 14, e1802.
21. J. Kneipp, in Principles and Clinical Diagnostic Applications of Surface-Enhanced Raman Spectroscopy, ed. Y. Wang, Elsevier, 2022, pp. 303–325.
22. W. Zhang, L. Jiang, J. A. Piper and Y. Wang, J. Anal. Test., 2018, 2, 26.
23. M. T. Yaraki, A. Tukova and Y. Wang, Nanoscale, 2022, 14, 15242.
24. T. J. Moore, A. S. Moody, T. D. Payne, G. M. Sarabia, A. R. Daniel and B. Sharma, Biosensors, 2018, 8, 46.
25. S. He, M. W. C. Kang, F. J. Khan, E. K. M. Tan, M. A. Reyes and J. C. Y. Kah, J. Opt., 2015, 17, 114013.
26. G. Jiao, X. He, X. Li, J. Qiu, H. Xu, N. Zhang and S. Liu, RSC Adv., 2015, 5, 53240.
27. E. Lenzi, M. Henriksen-Lacey, B. Molina, J. Langer, C. D. L. de Albuquerque, D. Jimenez de Aberasturi and L. M. Liz-Marzán, ACS Sens., 2022, 7, 1747.
28. Y. Wu, S. Wan, S. Yang, H. Hu, C. Zhang, J. Lai, J. Zhou, W. Chen, X. Tang, J. Luo, X. Zhou, L. Yu, L. Wang, A. Wu, Q. Fan and J. Wu, J. Nanobiotechnol., 2022, 20, 542.
29. Y. Wang, J. L. Seebald, D. P. Szeto and J. Irudayaraj, ACS Nano, 2010, 4, 4039.
30. S. Zhu, M. A. Eldeeb and S. W. Pang, Lab Chip, 2020, 20, 2188.
31. N. Lyu, B. Sun, A. Tukova, Q. Zhang, Z. Gu and Y. Wang, Mater. Today Chem., 2023, 33, 101698.

---