Nanomaterials, particularly metal nanoparticles (NPs) (Di Pietro et al., 2016; Liu et al., 2021; Nejabat et al., 2023), have garnered significant interest in recent years for their potential as therapeutic and diagnostic tools (theranostics). Their dimensions, typically ranging from 1 to 100 nm, align closely with the scale of cellular organelles, thereby enabling unique interactions with various biological pathways (Villanueva-Flores et al., 2020). Noble metal NPs, for instance, possess extraordinary physicochemical properties (e.g., optical, electrical, photothermal, and high surface area) that can be exploited for different applications in sustainable environmental management as well as in biomedicine, such as carriers for drug delivery, agents for photothermal therapy, imaging agents, or even radiosensitizers in radiation, proton, and photodynamic therapy (Anselmo and Mitragotri, 2016; Naletova et al., 2019).

Among the myriad of NPs investigated in the last decades as potential nanomedicine tools, gold nanorods (AuNR) are exceptional due to their anisotropic nature, which results in two distinct localized surface plasmon resonance (LSPR) bands. These bands are generated by light absorption along their short (transverse) and long (longitudinal) axes, with the longitudinal plasmon resonance band being modulable by altering the aspect ratio of nanoparticles (Bouhelier et al., 2005) to reach the near-infrared (NIR) region, i.e., in overlap with the optical window of biological tissues, which is very convenient for thermal ablation of cancer cells (Nejabat et al., 2023).

Gold nanoconstructs are slowly but steadily transitioning into clinical trials (Zhang et al., 2022); however, for their effective and safe translation from bench-to-bedside, still some burdens need to be wrecked, including long-term and accumulation of toxicity (in the case of prolonged treatments), pharmacokinetic behaviors (i.e., half-life and clearance rate), and tissue distribution (Yao et al., 2023). Therefore, it becomes essential to tailor size, shape, and surface functionalities, which in turn affect the charge and cellular uptake of nanoparticles, to fulfill the requested criteria for such applications. For example, it is critically important to have a high absorption cross-section, which results in effective photothermal heating close to the nanoparticle within a wavelength range that is compatible with physiological conditions (Pérez-Juste et al., 2005; Cole et al., 2009).

Many synthetic approaches have been extensively studied and tuned for construction of anisotropic nanomaterials to achieve certain morphologies and optimized aspect ratios. However, conventional seed-mediated synthesis still suffers a certain number of limitation/flaws, including the reproducibility, poor yield of ionic-to-metallic gold conversion (Vigderman and Zubarev, 2013), and toxicity. The chemical route of synthesis particularly for rod-shaped Au NPs requires the use of toxic surfactants such as cetyltrimethylammonium bromide (CTAB) in excess stoichiometry, which provides a micellar environment that promotes the anisotropic growth of Au seeds in certain directions (Tomasella et al., 2020; Wittmann et al., 2023). CTAB toxicity is mainly due to the impairment of the cell membrane and inhibition of ATP synthase, both leading to cell death (Johnson et al., 2002). An established strategy to remove CTAB from the suspension is to centrifugate and wash the NRs, with the limitations of of potential aggregation and no effective removal from the nanorod surface. Other established methods for CTAB removal from the AuNR surface include ligand exchange with biocompatible molecules, such as polyethylene glycol (PEG) (Niidome et al., 2006) and thiolated polyamidoamine (PAMAM) (Li et al., 2009). A facile synthesis for CTAB removal was proposed by He J. et al., consisting in a first addition of sodium borohydride to remove CTAB, given its great affinity to gold surface, followed by ligand replacement with mercaptoundecanoic acid (MUA) (He et al., 2018).

Our approach extends upon the framework proposed by He et al., by introducing adaptations tailored to optimize the biocompatibility, biodegradation, moisture retention, and anti-inflammatory properties of AuNRs. For this purpose, the ligand exchange reaction with sodium borohydride was followed by the addition of hyaluronan (HA), also called hyaluronic acid or hyaluronate. HA is the main non-protein glycosaminoglycan component of the extracellular matrix (ECM) (Garantziotis and Savani, 2019), with anti-inflammatory properties that, thanks to its structure, i.e., repeating glucuronic acid and N-acetylglucosamine units, enhance moisture retention and biocompatibility (Della Sala et al., 2022). HA also has cancer-targeting capability, mediated by the CD44 cell-surface receptor (Luo et al., 2019), which has been shown to play a key role in prostate cancer metastasis, migration, and invasion (Senbanjo et al., 2019). HA has been used as a bioactive scaffold in nano-biointerface studies (Greco et al., 2020; Palomba et al., 2020; Sanfilippo et al., 2020).

We investigated the removal of CTAB from AuNRs and their concurrent biofunctionalization with HA in terms of the plasmonic properties and resulting changes in the photothermal response of the nanoparticles. Proof-of work in vitro cellular experiments of cytotoxicity and cell migration were carried out on prostate cancer cells (PC-3 line) in comparison with fibroblast cells (NT-3 line), highlighting the very promising potential of this method for effective application of gold nanorods in nanomedicine.

The LSPR peak features of AuNRs before and after CTAB removal were investigated by UV–visible spectroscopy in the comparison with TEM analyses (Figure 1). The spectra in Figure 1A for the CTAB-capped AuNR show the expected features of rod-shaped nanoparticles, with evidence of two absorption bands, the most intense one in the NIR region (∼800 nm), due to electron oscillations along the long axis, referred to as the longitudinal band (λ_L), and a lower band in the visible region (∼512 nm), generated by electron oscillation along the transverse axis (λ_T), with absorbance maximum at a wavelength comparable to that of spherical gold approximately 10 nm in diameter (Foti et al., 2023). In the inset of Figure 1A is displayed the magnified view of longitudinal band shifts. The spectral parameters of the plasmon bands, i.e., wavelengths, and the correspondent absorbance values are reported in Table 1, together with the calculated aspect ratio and nanoparticle concentration, according to Eqs 1, 2 (see Supplementary Methods). The data in Table 1 evidence a relatively high stability of the nanorods upon the purification process done by centrifugation, with the transversal plasmonic band being practically unmoved and a blue shift for the longitudinal band of ∼20 nm, which still falls in the NIR region. As per the purified CTAB-capped AuNR (pellet 2 sample), the calculated aspect ratio from the LSPR peak analysis (AR = 3.7) is comparable to the value measured by TEM (AR = 4), corresponding to the average size of 9 (±2) nm and 33 (±3) nm for the short and long axes, respectively.

The spectra in Figure 1B allow following the process of ligand exchange during the treatment with NaBH₄ up to 30 min from the beginning of the reaction. In particular, hypsochromic shifts of ∼20 nm and ∼5 nm for λ_L and λ_T, respectively, were observed (Table 2), in agreement with the evident smaller features observed in the TEM micrographs of CTAB-free (Figure 1B) than CTAB-capped AuNR (Figure 1A). The average transversal and longitudinal dimensions calculated from TEM for the CTAB-free nanorods are 6 (±1) nm and 26 (±3) nm. Such a “shrinking” of the NPs can be explained considering the mechanism of CTAB-assisted growth of Au seeds, involving a rod-shaped micellar bilayer formed by the surfactant, where the polar heads on the one end adsorb on gold and on the other end interact with water (Meena and Sulpizi, 2013). Molecular dynamics simulations discriminated between the arrangements of CTAB adsorbing onto flat or curved Au surfaces, with a lower CTAB coverage density and larger inter-micellar channels found on curved surfaces (representing the tip facets of growing AuNR) compared to those on a flat surface (representing the lateral facets of growing AuNR) (da Silva et al., 2019). By inferring a similar mechanism for CTAB removal (i.e., desorption process from the gold surface), one can hypothesize that the decreased CTAB concentration affects tip facets more than the lateral facets, resulting in decreased aspect ratio. Furthermore, CTAB removal from the AuNR surface leads to a reduction in the refractive index (1.44 for CTAB and 1.33 for water), which is in agreement with the blue shift observed in the longitudinal band (Casas et al., 2013; Movsesyan et al., 2019).

The XPS, Raman, and FTIR spectra of the CTAB-capped and CTAB-free AuNR (Figure 2) confirm the effective removal of the surfactant by treatment with NaBH₄. Figure 2A shows that the Br 3d_5/2 signal, centered at a binding energy of 64.3 eV, which points to the presence of Br⁻ species, is strongly decreased in CTAB-free AuNR compared to CTAB-capped nanorods. Moreover, the Au 4f_7/2 signal, centered at 84.0 eV in both spectra, confirms the formation of metallic gold (Oliveira et al., 2021). The Au/Br atomic ratio was evaluated to be 0.1 and 0.66 for CTAB-capped and CTAB-free AuNR, respectively, thus confirming the effective removal of the surfactant by treatment with NaBH₄. In Figure 2B, the spectrum of AuNR (+CTAB) (red line) in the 1,040–1780 cm⁻¹ range shows strong CTAB-characteristic Raman vibrational bands (gray line) at 1,057 cm⁻¹ (C-C stretch), 1,124 and 1,293 (C-C stretching and CH₂ twist), and 1,425 and 1,478 cm⁻¹ (the combination of CH₂ scissors and CH₃ deformations) (Gökce and Bahçeli, 2013). In the spectrum of AuNR (-CTAB) (blue line), such bands totally disappear, confirming the formation of an almost CTAB-free surface suitable for subsequent functionalization with hyaluronic acid. Accordingly, the ATR-FTIR spectra (Figure 2C) display for the CTAB-free AuNR an evident decrease of the strong absorptions at 2,924 and 2,854 cm⁻¹ due to C−H stretching vibration of methyl and methylene groups of CTAB, respectively (Su et al., 2015).

Concerning the HA-functionalized AuNR, both AuNR (+CTAB)/HA and AuNR (-CTAB)/HA samples, the UV–visible spectra (Figure 3) show evidence of effective polymer coating, with similar bathochromic shifts of the plasmon bands (Δλ_T = + 3 nm; Δλ_L = + 12–13 nm) found for both CTAB-capped (Figure 3A) and CTAB-free AuNR (Figure 3B), as expected for the plasmonic signals of nanoparticles that increase their optical size due to the physisorption of the hyaluronan molecules (Sanfilippo et al., 2020). Normal Raman and SERS spectra of HA-functionalized nanoparticles were collected and compared to those of pure HA for comparison and peak assignment following the literature (Cannavà et al., 2022). A sole band around 1,040 cm⁻¹ resulting from overlap of the C-C and C-O stretching modes is observed in both HA and HA-functionalized AuNR samples, therefore confirming the success of the interaction (Supplementary Figure S2).

NTA analyses, which allow particle-by-particle analysis, provided the NP concentration of the sample after CTAB removal and successively after HA addition (Table 3). The similar values found of 4.0 10⁹ and 1.2 10⁹ NP mL⁻¹ demonstrate that the treatment does not induce drastic changes, such as precipitation, in the solution. Table 2 reports also the values of hydrodynamic diameter and reveals that the particles without CTAB are smaller than the particles coated with CTAB, according to TEM analyses; furthermore, the presence of HA increases the particle dimension. A further confirmation of effective CTAB removal from the surface of gold nanorods and NP bioconjugation with HA was also obtained from the ZP values from the surface charge analysis of the nanorods, pointing to the neutralization of charges in AuNR (+CTAB)/HA and the negative surface charge on AuNR (-CTAB)/HA, as depicted in Figure 4.

The thermoresponsive properties of AuNR (±CTAB)/±HA were investigated by irradiation with a 671-nm continuous-wave laser (CW), and the trend is shown in Figure 5 (thermographic images of the samples recorded before and after 4 min of irradiation at 671 nm are reported in Supplementary Figure S3). The AuNR samples present a rapid increase in temperature during the 4-min irradiation, except the AuNR (+CTAB)/HA, which precipitated during the experiments and HA alone, showing no response to laser irradiation. In particular, ΔT = 27°C for AuNR (+CTAB) (Figure 5A) and ΔT = 17°C for AuNR (-CTAB)/HA (Figure 5B) were observed. Thus, the hybrid system with HA is less responsive than the CTAB-free AuNR alone but reaches a temperature increase like the CTAB-capped sample, suggesting that it is possible to take advantage of the HA desirable bioproperties while preserving photothermal properties almost identical to those observed with the toxic CTAB-capped AuNRs.

The cytotoxicity and oxidative stress of CTAB-capped and CTAB-free AuNR hybrids with HA were tested in vitro in fibroblasts (3T3 cell line) and prostate cancer (PC-3, tumor cell line) by MTT and MitoSOX assays, respectively (Figure 6).

In the cell viability assay, for AuNR (+CTAB), a dose-dependent toxic effect of CTAB was evident, as expected from literature data (Jia et al., 2020), with IC₅₀ values reaching for the highest treatment dose (i.e., 50x dilution of the AuNR pellet 2) for both PC-3 (Figure 6A) and 3T3 (Figure 6C) cell lines. In the case of AuNR (+CTAB)/HA, a similar trend in the dose–response effect is visible, with a general reduction in toxicity (IC₅₀ values above the dose treatment with 50x dilution only for the non-tumoral, CD44-negative cells), which is mitigated by the beneficial presence of the hyaluronan. Noticeably, the AuNR (-CTAB) and AuNR (-CTAB)/HA samples show null or minimal toxicity (∼20% reduction), with the dose–response effect visible only for the non-tumoral cells, confirming the success of CTAB removal procedure discussed above.

For ROS production, while no significant oxidative stress induction was detected in 3T3 (Figure 6D), PC-3 cells were found to be the significantly affected by the treatments (Figure 6B). Again, CTAB-capped nanoparticles produce a higher level of oxidative stress than CTAB-free nanoparticles, in agreement with literature data concerning the toxic effects of AuNR that can be attributed to mitochondrial impairment (Bidian et al., 2023). Indeed, AuNRs have been demonstrated to selectively accumulate in the tumoral cell mitochondria, causing their death, while this effect was null on normal cells (3T3, in our case) (Zhang et al., 2013).

The LSM analysis of PC-3 cells upon treatment with the different samples, although not allowing for the visualization of nanoparticles uptake, for instance, nanoparticle aggregates into the cytoplasm clearly visible from the bright field micrographs, gives, however, clear evidence of cellular changes in terms of both cell morphology as well as perturbation of intracellular organelles (Figure 7). Indeed, a significant decrease in the mean fluorescence intensity of stained lysosomes was found in cells incubated with CTAB-capped nanoparticles, while no significant changes were detected in the emission of dye-labeled mitochondria. According to the short incubation time (2 h) used in LSM imaging experiments, our findings are well in agreement with literature data about the special sensitivity of cancer cells to CTAB, which, as a quaternary ammonium cation, can destroy the lysosomal membrane and be further released into the cytoplasm for further targeting of the mitochondria (Wang et al., 2011). On the other hand, the PC-3 cells incubated with AuNR (-CTAB) showed a significant increase in both lysosomal and mitochondrial emission, thus confirming that cellular trafficking of AuNRs in cancer cells is strongly affected by the CTAB concentration. Interestingly, the treatment of cells with HA, which decreases the average emission of stained lysosomes and especially that of the mitochondria, according to the activation of the CD44 receptor by HA and apoptosis regulation by the mitochondrial pathway (Michalczyk et al., 2022), is also observed in cells treated with AuNR (-CTAB)/HA, with comparable decreases in the fluorescence intensity with respect to the negative control of stained lysosomes and the mitochondria of approximately 50%–70% and 5%–6%, respectively.

Figure 8; Supplementary Figure S4 present data from the scratch assay conducted on PC-3 cells over a 48-h timeframe period to evaluate the effects of AuNR (±CTAB)/±HA on cell migration. While both negative and positive control scratches reach 100% of wound closure after 24 h, all the treatments with the gold nanorods inhibit cancer cell migration. As expected, the two samples with CTAB prevent cellular migration, paralleling the toxic effect evidenced by the MTT assay. On the other hand, the cell treatments with CTAB-free nanorods showed a reduction in cell migration, which is not as high as that observed in the CTAB-capped nanorods, confirming that AuNRs possess intrinsic anti-tumoral properties, even if the toxic surfactant is removed from the surface, in accordance with the cytotoxicity tests. These findings are particularly promising since controlled inhibition of cancer cell migration is crucial to prevent metastasis formation (Yamaguchi et al., 2005) as well as for inhibiting the development of hypertrophic scar tissue due to the over-production and migration of fibroblasts (Wang et al., 2008).

In this study, we exploited the processes of CTAB removal from gold nanorods and the concurring nanoparticle functionalization with a bioactive low-molecular weight (200 kDa) hyaluronan. The hybrid nanoconstructs of AuNR (-CTAB)/HA maintained a significant photothermal responsivity as bare nanorods, while nullifying the cytotoxic effect exhibited in prostate cancer cells and fibroblasts by the corresponding CTAB-capped analog nanoparticles. This approach for the post-synthesis treatment and biofunctionalization of gold nanorods is very promising for nanomedicine applications in the photothermal therapy of tumors as well as in the control of scar tissue formation in wound healing.