Mild hyperthermia via gold nanoparticles and visible light irradiation for enhanced siRNA and ASO delivery in 2D and 3D tumour spheroids

Efficient delivery of therapeutic nucleic acids, such as small interfering RNA (siRNA) and antisense oligonucleotides (ASO), into cells is a crucial aspect of gene therapy

Due to their unique optical properties, AuNPs have already demonstrated the potential to be used as photothermal agents using incident laser light sources, including visible and near-infrared light (NIR) (Amendoeira et al. 2020;Hu et al. 2020;Oliveira et al. 2023;Riley & Day 2017).The electromagnetic energy transformed into heat as a result of electron excitation and relaxation by AuNPs is one of their unique physicochemical properties (Xiaohua Huang and El-Sayed 2010; Krajczewski et al. 2017;Kumar et al. 2023).For instance, Braun et al. reported that hollow gold nanoshells conjugated with siRNA and Tat peptides for cellular uptake were trapped within endosomes, which could escape upon exposure to a pulsed NIR laser, allowing the release of siRNA into the cytosol (Braun et al. 2009).Our group has demonstrated the photothermal potential of spherical AuNPs upon visible laser irradiation at the surface plasmon (SPR) peak (Mendes et al. 2017;Pedrosa et al. 2017Pedrosa et al. , 2018;;Roma-Rodrigues et al. 2020).
Localised mild hyperthermia shows the possibility of modulating the cell membrane behaviour (e.g., fluidity) and increasing the susceptibility of cells to improve the internalisation of AuNPs functionalised with oligonucleotides for gene therapy.The efficacy of combining spheric functionalised AuNPs with visible laser irradiation to enhance therapeutic outcomes in vitro and in vivo models has been reported (Heinemann et al. 2013;Pedrosa et al. 2017;Roma-Rodrigues et al. 2020).
3D models recapitulate the in vivo tumour microenvironment (TME) to closely resemble tumour growth and progression, which are necessary to better evaluate therapeutic responses (Pinto et al. 2020;Valente et al. 2023).The formation of cell-to-cell and cellmatrix interactions in three-dimensional (3D) cell cultures (e.g., multicellular tumour spheroids and organoids) is a pivotal link between 2D cultures and animal models (Jensen and Teng 2020;Zanoni et al. 2020).Spheroids are excellent in vitro 3D models that can mimic the structure of tumours owing to the formation of a gradient of nutrients, oxygen, pH, and metabolism products, and can also recreate some of their mechanisms, such as hypoxia and acidosis.
Herein, we use light irradiation of AuNPs for the controlled increased delivery of oligonucleotides into 2D cell cultures and 3D tumour spheroids for targeted gene silencing-see Fig. 1.

AuNP synthesis and functionalisation
AuNPs were synthetised resorting to the citrate reduction method described by Lee and Meisel (Lee and Meisel 1982), and subsequently functionalised with polyethylene glycol (PEG) for improved stability in biological media as previously reported (Baptista et al. 2013)-100% PEG ensures stability in complex media; the 30% coverage allows for "space" for functionalisation with the oligonucleotide, which compensates for the lower % of PEG at the surface in terms of stability.Briefly, to achieve 100% coverage of the AuNPs' surface, 10 nM of citrate-capped AuNP were incubated with 0.028% SDS (w/v) (Sigma-Aldrich, Switzerland) and 0.01 mg/mL of thiolated PEG (M W :356 Da, Sigma-Aldrich, Switzerland) for 16 h under agitation at room temperature.The excess PEG was removed by centrifugation at 14,000 g for 30 min at 4 ℃ (three times), and the free thiol-PEG was quantified in the supernatants via the Ellman's Assay.A 100% coverage was considered when the supernatant showed no traces of free thiol, which indicated that all thiol-PEG had been bonded to the AuNP surface.
In addition, AuNPs with a 30% coverage of PEG (AuNP@PEG 30%) were prepared and subsequently functionalised with a thiolated stem-looped ASO complementary to the c-MYC transcript (GenBank NM_002467.5;AuNP@c-MYC) and a scramble oligo (AuNP@scramble) as a control, as previously described by Baptista et al.ASO was added at a 1:150 AuNP: oligonucleotide proportion, incubated for 16 h, and the excess oligonucleotides were removed by centrifugation at 15,500 g for 1 h at 4 ℃ and washed twice with diethylpyrocarbonate (DEPC)-treated water (Sigma-Aldrich, Switzerland).The number of ASO functionalised to the surface was inferred from the quantification of single-stranded DNA in the supernatants using a Nanodrop spectrophotometer (ND-1000, NanoDrop Technologies, USA), considering the amount of initially added ASO.
AuNP and Au-nanoconjugates were assessed by ultraviolet-visible absorption spectroscopy (UV-Vis), dynamic light scattering (DLS), zeta potential, and transmission electron microscopy (TEM).UV-Vis spectra were acquired on a UV-Vis spectrophotometer (UV mini-1240 spectrophotometer, Shimadzu, Germany) in the range 400-800 nm using a 1 cm path quartz cuvette (Quartz SUPRASIL, Hellma Analytics, Germany) at room temperature (RT).The hydrodynamic diameter, polydispersity index, and zeta potential were measured by DLS resorting to a Malvern Zetasizer Nano ZS at 25 ℃, scattering angle 173° and laser wavelength 633 nm (Malvern, Zetasizer Nano series, United Kingdom).TEM analysis was provided as a service by Instituto Gulbenkian de Ciência (IGC) and the core diameters were calculated using the open-access software platform FIJI-Image J (National Institutes of Health, Bethesda, United States).
Characterisation of AuNP and Au-nanoconjugates can be found in Additional file 1: Table S1, Figures S1 and S2.

Irradiation and photothermal characterisation
To determine the photothermal capability of the Au-nanoconjugates, these were irradiated at different concentrations (0.25 nM, 0.5 nM, 0.75 nM, and 1 nM) in water in a 96 well-plate.For temperature measurements, a thermocouple was inserted in the wells before and immediately after visible light laser irradiation, as previously described (Mendes et al. 2017;Pedrosa et al. 2018).Irradiation was performed with a continuous wave (CW) 532 nm green diode-pumped solid-state laser (DPSS) (Changchun New Industries Optoelectronics Tech.Co., LTD, Changchun, China) with a diode intensity (LDI) of 2.37 W/cm 2 coupled (Mendes et al. 2017) to an optical fibre for different exposure times (30 s, 60 s, 90 s, and 120 s).

Plasmid transfection with lipofectamine LTX and plus reagent
HCT116 2D cells were placed at a density of 2 × 10 4 cells/well in a 96-well plate and cultured in DMEM for 24 h at 37 ℃, under standard cell culture conditions, prior to transfection.A plasmid that encodes a GFP (λ ex 475 nm/λ em 505 nm) optimised for high expression in mammalian cells (pAcGFP1-Nuc Vector 4.8 kb, Takara Bio Company, USA), was added (100 ng/well) to cells using a 1:2 (plasmid: Lipofectamine) ratio and incubated in DMEM (0% FBS and without antibiotics) for approximately 4 h at 37 ℃, according to the Lipofectamine LTX and Plus Reagent (Invitrogen, ThermoFisher Scientific, USA) manufacturer's recommendation.AcGFP1 expression was confirmed after 24 h by fluorescence microscopy (Ti-U Eclipse inverted microscope; FITC filter (excitation at 480/30 nm and emission at 535/40 nm)) and images were analysed using Image J software.

Quantification of AuNP internalisation into cells
The internalisation of the AuNP@PEG 100% in irradiated and non-irradiated HCT116 2D cells was analysed via inductively coupled plasma atomic emission spectroscopy (ICP-AES).For this experiment, HCT116 2D cells were washed with phosphatebuffered saline (PBS), trypsinised with TrypLE ™ Express, and centrifuged at 750 g for 5 min.The supernatant was removed and stored at 4 ℃, whereas the cell pellet was stored at − 20 ℃.One day before analysis, 1 mL of freshly prepared aqua regia (HCl: HNO 3 = 3:1) was added to the samples.A standard curve for gold was recorded to quantify the amount of intracellular gold (LAQV/REQUIMTE, Laboratório de Análises).

Laser irradiation in 2D cells: HCT116 and MCF-7/GFP
HCT116 and MCF-7/GFP 2D cells were seeded at a density of 2 × 10 4 /well and 1 × 10 4 / well respectively, in a 96-well plate and incubated for 24 h at 37 ℃ in a 99% humidified atmosphere and 5% (v/v) CO 2 .After 24 h, HCT116 2D cells were transfected with 100 ng of pAcGFP1 Nuc Vector using Lipofectamine LTX and Plus Reagent.After 4 h of incubation with Lipo + pGFP complexes, the culture medium was replaced with DMEM without phenol red, and cells were challenged (or not) with 10 nM of AuNP@PEG 100% for 4 h.Next, the cell culture medium was replaced with DMEM without phenol red (at 37 ℃) containing 20 nM of anti-AcGFP1.For MCF-7/GFP: following 24 h, cells were challenged (or not) with 10 nM of AuNP@PEG 100% for 4 h, and then cell culture medium was replaced with DMEM without phenol red (at 37 ℃) containing 20 nM of anti-copGFP.
Cells with AuNP were then irradiated with a continuous wave 532 nm green DPSS laser coupled to an optical fibre (LDI 2.37 W/cm 2 ) for 60 s to achieve mild hyperthermia.Controls without irradiation, with or without AuNP, were also prepared.For temperature measurements, a thermocouple was inserted in the wells (in contact with the cell culture medium) before and immediately after visible light irradiation.
After 6 h and 24 h for copGFP and 24 h for AcGFP1, GFP expression was confirmed by fluorescence microscopy and MTS assay was performed.In addition, total RNA was extracted from irradiated and non-irradiated cells, and the effect of visible irradiation on the transfection of anti-AcGFP1 and anti-copGFP silencing was evaluated by realtime quantitative PCR (RT-qPCR) and compared to a commercial transfection reagent, Lipofectamine RNAiMAX reagent (Invitrogen, ThermoFisher Scientific, USA), using the manufacturer's instructions.

Cells transfection with lipofectamine RNAiMax reagent
HCT116 and MCF7/GFP 2D cells were seeded at a density of 2 × 10 4 /well and 1 × 10 4 / well respectively, in a 96-well plate and incubated for 24 h at 37 ℃ under standard cell culture conditions.After 24 h, 20 nM of oligonucleotide was mixed with the cationic lipid reagent Lipofectamine RNAiMax in DMEM without phenol red (0% FBS and without antibiotics) and incubated for 5 min in a microcentrifuge tube according to the manufacturer's recommendation.Next, this mixture was added to cells in DMEM without phenol red (Final 8% FBS) for 24 h in HCT116 2D cells; and 6 h and 24 h in MCF-7/GFP 2D cells.
Post-transfection incubation, total RNA was extracted, and the efficiency of transfection for AcGFP1 and copGFP silencing was evaluated by RT-qPCR, and fluorescence microscopy for copGFP protein expression.

c-MYC silencing in HCT116 2D cells
HCT116 2D cells were plated at a density of 2 × 10 4 cells/well in a 96-well plate and incubated for 24 h at 37 ℃ in a 99% humidified atmosphere and 5% (v/v) CO 2 .After 24 h, the cells were challenged with AuNP@ASO and AuNP@PEG in DMEM without phenol red to evaluate c-MYC expression at different time points: 6 h, 9 h, 12 h, and 24 h.The concentration of ASO was 50 nM (0.39 nM of AuNP@c-MYC and 0.45 nM of AuNP@scramble), using 0.39 nM and 0.45 nM of AuNP@PEG 30% as controls, respectively.The cells were then irradiated after 4 h of AuNP incubation with the set laser conditions mentioned above.Total RNA was extracted, and an MTS assay was performed.The effect of visible irradiation in the internalisation of AuNP@ASO for c-MYC silencing was evaluated via RT-qPCR and Immunofluorescence assay.Following the initial assessment, the 9 h incubation period was selected for further experiments.

c-MYC silencing in HCT116 3D spheroids
HCT116 2D cells were seeded with a density of 5 × 10 3 /well in an ultra-low attachment 96-well plate to produce 3D spheroids and incubated for 3 and 7 days at 37 ℃, under standard culture conditions.The 3-and 7-day 3D spheroids were challenged with 25 nM and 60 nM of ASO, respectively (0.19 nM and 0.47 nM of AuNP@c-MYC; as controls, 0.23 nM and 0.54 nM of AuNP@scramble, respectively) in DMEM without phenol red.AuNP@PEG 30% was used for control of AuNP@c-MYC (0.19 nM and 0.47 nM) and AuNP@scramble (0.23 nM and 0.54 nM).Upon 4 h, HCT116 3D spheroids were irradiated as described above for HCT116 2D cells.Following 9 h of incubation, total RNA was extracted, and LDH assay and LIVE/DEAD Viability/Cytotoxicity assays were performed.The effect of visible irradiation in the internalisation of AuNP@c-MYC in silencing was evaluated with RT-qPCR and Immunofluorescence assay.

c-MYC immunofluorescence in HCT116 2D cells/3D spheroids
HCT116 2D cells were fixed with 4% formaldehyde for 20 min, then washed three times with PBS and permeabilised with 0.1% Triton-X 100 in PBS for 5 min.HCT116 3D spheroids were fixed with 4% formaldehyde for 1 h, then washed three times with PBS and subsequently permeabilised using 0.1% Triton-X 100 in PBS for 30 min.Cells were incubated with 7.5 μg/mL of Hoechst 33,258 for 15 min and the spheroids were incubated for 45 min, then washed three times with PBS and the blocking was performed with 1% bovine serum albumin (BSA) in PBS with 0.1% Tween 20 (PBST) for 1 h.Next, cells and spheroids were incubated overnight at 4 ℃ with the primary antibody-anti-rabbit c-MYC antibody (1:200 dilution; ab32072; Abcam, UK).After incubation, cells and spheroids were washed three times with PBST, and cells were incubated for 1 h at RT with the secondary antibody-FITC-conjugated anti-rabbit antibody (1:500 dilution; ab6717; Abcam, UK), whereas the spheroids were incubated for 3 h.The cells and spheroids were subsequently washed three times with PBST and visualised in PBST using a Ti-U Eclipse inverted microscope.Images were acquired using the DAPI emission filter (excitation at 360/40 nm and emission at 460/50 nm) and FITC emission filter (excitation at 480/30 nm and emission at 535/40 nm).Fluorescence microscopy images were analysed using Image J software.

Membrane integrity and cell viability assays
HCT116 2D cells membrane integrity was evaluated by Trypan Blue exclusion assay.Briefly, irradiated, and non-irradiated cells, previously incubated or not with 10 nM of functionalised AuNP for 4 h were immediately incubated with Trypan Blue dye for 10 min and washed three times with PBS to remove the excess of the dye.For the analysis of membrane integrity of irradiated and non-irradiated cells, with or without AuNP, several images were acquired with a bright field inverted microscope (Nikon TMS, Tokyo, Japan).
The membrane integrity of HCT116 2D cells/3D spheroids (irradiated or not, in the presence or absence of Au-nanoconjugates) was also evaluated by resorting to the release of intracellular lactate dehydrogenase (LDH) in cells supernatant by LDH assay (CytoTox 96 Non-Radioactive Cytotoxicity Assay, Promega, UK).As a positive control for membrane rupture, the application of LDH lysis solution diluted in DMEM (1:10) was prepared 1 h before the procedure and incubated in cells or spheroids.Following the HCT116 2D cells irradiation or 9 h of 3D spheroids challenge with Au-nanoconjugates, 50 μL of cells or spheroids supernatant are incubated with 50 μL of CytoTox 96 Reagent (1:1 ratio) in a 96-well plate for 30 min at RT, protected from light.Next, 50 μL of Stop Solution is added to the wells and incubated for 1 h at RT, in the dark.The absorbance was measured at 490 nm, on a microplate reader Infinite M200 (Tecan, Switzerland) and cell cytotoxicity was normalised to positive control.
2D cell viability (irradiated and non-irradiated cells, with or without AuNP) was evaluated via the MTS assay.Briefly, cells were incubated with the MTS solution (1:5 v/v in DMEM; CellTiter 96 AQueous One Solution Cell Proliferation Assay Kit, Promega, UK) for 1 h, at 37 ℃ under standard cell culture conditions and the absorbance was measured at 490 nm using a microplate reader.Cell viability was normalised to control cells (cells with culture medium) (Lenis-Rojas et al. 2017;Silva et al. 2013).
LIVE/DEAD Viability/Cytotoxicity assay (Invitrogen, ThermoFisher Scientific, USA) was performed to assess the impact of visible irradiation after the c-MYC silencing experiment in HCT116 3D spheroids.Following the 9 h of silencing, the irradiated and non-irradiated spheroids were washed with PBS and incubated with 1 μM of Calcein AM and 2 μM of Ethidium homodimer-1 (EthD-1) in PBS for approximately 45 min, at 37 ℃ in a 99% humidified atmosphere and 5% (v/v) CO 2 .Fluorescence microscopy images were acquired by a Ti-U Eclipse inverted microscope using an FITC emission filter (excitation at 480/30 nm and emission at 535/40 nm) and G2A emission filter (excitation at 535/50 nm and emission > 590 nm).Fluorescence microscopy images were analysed using Image J software, and the ratio between Calcein AM and EthD-1 was performed to evaluate cell viability/cytotoxicity.
Total RNA extraction, cDNA synthesis, and RT-qPCR assessment of gene silencing Total RNA was extracted from 2D cultures using the Nzyol reagent (NZYTech, Portugal) according to the manufacturer's guidelines.RNA extracted from HCT116 2D cells transiently expressing AcGFP1 or MCF-7 constitutively expressing copGFP was reverse transcribed to cDNA using the NZY M-MuLV First-Strand cDNA Synthesis kit (NZYTech, Portugal).RT-qPCR was performed using NZYSupreme qPCR Green Master Mix (2x) (NZYTech, Portugal) according to the manufacturer's protocol in a Qiagen Rotor-Gene Q cycler (Qiagen, Germany).The following conditions were used for AcGFP1: initial denaturation at 95 ℃ for 5 min; 30 cycles of 95 ℃ 30 s, Tm 52 ℃ for 30 s, 72 ℃ for 45 s; and a final extension step at 72 ℃ for 7 min; for copGFP: initial denaturation at 95 ℃ for 5 min; 30 cycles of 95 ℃ for 30 s, Tm 53 ℃ for 30 s, 72 ℃ for 30 s; and a final extension step at 72 ℃ for 5 min.
Total RNA was also extracted from HCT116 2D cells and six 3D spheroids after c-MYC silencing and RT-qPCR was performed using a one-step NZY RT-qPCR Green kit (NZYTech, Portugal) in a Qiagen Rotor-Gene Q cycler.The set conditions for reverse transcription were 50 ℃ for 20 min, and qPCR with initial denaturation at 95 ℃; 30 cycles of 95 ℃ 15 s, Tm, and extension of 60 ℃ for 15 s.
Gene expression was evaluated according to the 2 −ΔΔCt method (Livak & Schmittgen 2001), using the 18S ribosomal gene as reference.All primers and antisense oligonucleotides were purchased from STAB VIDA, Lda (Portugal); all sequences can be found in Additional file 1: Tables S2 and S3.

Statistical analysis
Data were analysed using GraphPad Prism 8.0 (GraphPad Software, San Diego, USA).One-way and two-way ANOVA with Tukey's multiple comparison test, and unpaired parametric t test with Welch's correction were used to evaluate differences between groups.They were considered statistically significant at p value < 0.05.Data are the mean value of at least three independent assays with at least two technical replicates, and the errors are calculated by the standard error mean.

Synthesis and characterisation of Au-nanoconjugates
Citrate-capped AuNP characterisation by UV-Vis, DLS, and TEM shows a maximum Localised Surface Plasmon Resonance (LSPR) peak at 519 nm, a hydrodynamic diameter of 16.4 (± 0.1) nm by DLS, and an average spherical shape of 12.1 (± 1.5) nm by TEM.The AuNP@PEG showed a slight red shift of the LSPR peak (520 nm) and a slight increase in the hydrodynamic diameter 21.1 (± 0.3) and 18.5 (± 0.1) nm for 30% and 100% PEG coverage, respectively.Functionalisation with the ASO resulted in an additional slight right shift of the LSPR peak to 522 nm and average hydrodynamic diameters for AuNP@c-MYC and AuNP@scramble were 35.7 (± 0.2) and 35 (± 0.2) nm, respectively (see Additional file 1: Table S1 and Figures S1, S2).

Laser irradiation
The photothermal effect of the Au-nanoconjugates was assessed by irradiating a range of concentrations of pegylated AuNP: 0.25 nM, 0.5 nM, 0.75 nM, and 1 nM; at different exposure times: 30 s, 60 s, 90 s and 120 s.The temperature variation (ΔT) was calculated by subtracting the final temperature from the initial temperature in each assay, excluding the temperature of the irradiated water.The thermal capacity of water 4.18 J.g −1 .K −1 (q) and the irradiated mass (m) of 0.1 g were used to estimate the heat via Eq. 1 (Huang et al. 2012;Kim et al. 2023)-Fig.2a, b: For mild hyperthermia a target temperature of 41-42 ℃ is desirable (Mendes et al. 2017), corresponding to the ΔT values of 4-5 ℃ attained for both types of pegylated AuNP upon 60 s laser irradiation.Under these conditions, the photothermy effect was determined to be 7.8 × 10 -13 W per particle for AuNP@PEG 30% (Fig. 2c and 9.5 × 10 -13 W per particle for AuNP@PEG 100% (Fig. 2d).Altogether, these settings were considered to achieve a mild hyperthermia effect in cells.

Cytotoxic and viability assays in HCT116 2D cells
To assess the effect of the selected conditions for mild hyperthermia in cells, HCT116 colorectal carcinoma 2D cells were incubated for 4 h with 10 nM of AuNP@PEG 30% and 10 nM AuNP@PEG 100%, with or without irradiation.The irradiated and (1) Q = q × T × m Fig. 2 Heat capacity and heat generated per second of Au-nanoconjugates.Heat capacity of (a) AuNP@PEG 30% and (b) AuNP@PEG 100% for different exposure times and concentrations of AuNP in water irradiated at 2.37 W/cm 2 .Four different concentrations of (c) AuNP@PEG 30% and (d) AuNP@PEG 100% were irradiated at 2.37 W/cm 2 for 60 s in water and the temperature variation was measured to calculate heat generated per second as a function of the number of particles irradiated.The slope of the curve gives the heat generated per nanoparticle per second non-irradiated 2D cells were then incubated with Trypan Blue 0.4% (m/v) solution and the images acquired in a bright field inverted microscope.Trypan blue is a dye impermeable to live cells which can penetrate necrotic cells or cells with permeable compromised membranes (Strober 2015).Our data show for the used conditions that HCT116 2D cells did not stain with trypan blue (see Additional file 1: Figure S3), despite the slight increase of temperature, showing that the integrity of cell membrane was not compromised by the selected LDI condition: 2.37 W/cm 2 for 60 s.
HCT116 2D cell viability was further assessed by the MTS assay following the 4 h of incubation with 10 nM of AuNP@PEG 30% and 10 nM AuNP@PEG 100%, with or without irradiation.The MTS assay specifically relies on the activity of mitochondrial enzymes, such as dehydrogenases, which are indicative of the functional and metabolically active cells (Cory et al. 1991;Stoddart 2011).Again, the MTS data indicate that pegylated AuNPs, with or without laser irradiation, do not have an impact on cell viability (Additional file 1: Figure S4a. To complete the cell viability assessment, we used the LDH assay to simultaneously evaluate cells' membrane permeability and integrity.This assay measures cytotoxicity by quantifying the LDH release in the cell medium and could indicate permeability and/ or cell membrane damage (Decker & Lohmann-Matthes 1988;Stoddart 2011).Results show that AuNP@PEG 100% combined with laser irradiation increases 2D cells' membrane permeability, but without compromising membrane integrity as corroborated by the trypan blue exclusion assay (Additional file 1: Figure S4b).

Laser irradiation-induced Au-nanoconjugates uptake
TEM and ICP-AES were used to assess Au-nanoconjugates uptake by cells upon laser irradiation-Fig.3. TEM images of HCT116 2D cells incubated for 4 h with AuNP@ PEG 100% show internalisation of AuNPs in small clusters (Fig. 3a.This clustering is commonly observed for internalisation of the AuNPs (Carnovale et al. 2019;Janic et al. 2018).
Then, we determined the amount of gold inside the cells by ICP-AES in cells that were exposed to 10 nM of AuNP@PEG 100% for 4 h either irradiated or non-irradiated.
Fig. 3 TEM images of Au-nanoconjugate internalisation in HCT116 2D cells and Au uptake.a Images of AuNP@PEG 100% clusters (indicated with a black arrow) inside HCT116 2D cells, observing more than four small dark clusters near to cell membrane due to the high electronic density of gold.Scale bars correspond to 1 μm and 200 nm, respectively.b ICP-AES analysis shows an increase of Au uptake in irradiated cells exposed to 10 nM of AuNP@PEG 100% for 4 h.Statistical differences were not observed between AuNP@PEG and AuNP@PEG + Laser samples.Data represent the mean value ± the standard error mean of an independent experiment with three technical replicates for each Data show that irradiated cells (AuNP@PEG + Laser) internalised approximately 1.4fold more Au when compared to non-irradiated cells (AuNP@PEG)-see Fig. 3b.We hypothesise that the localised heat produced by irradiated AuNP@PEG somehow destabilises the lipid bilayer, increasing fluidity and permeability, which allows an increase in the uptake of pegylated-AuNP by cells.The creation of hotspots on the cell membrane has been observed since that influences the permeability of the membrane (Beola et al. 2020;Kang et al. 2019;Madrid 2018).

Mild photothermy for GFP silencing in HCT116 2D cells
To demonstrate that mild photothermy may be used to improve the internalisation of silencing oligonucleotides (antisense siRNA or ASOs) in HCT116 2D cells, AuNP@PEG and an anti-GFP were used (Fig. 4a.First, a transient transfection model of HCT116 2D cells harbouring a pAcGFP1-Nuc plasmid that encodes for GFP from Aequoera coerulescens with three copies of the nuclear localisation signal (see Additional file 1: Figure S5) was used.After 4 h of plasmid-lipofectamine incubation, the cell culture medium was replaced by DMEM without phenol red, with or without 10 nM of AuNP@PEG for another 4 h at 37 ℃.
Then, the cell medium was replaced with DMEM without phenol red at (37 ℃), and 20 nM of anti-AcGFP1 was added before laser irradiation at 2.37 W/cm 2 , for 60 s.The silencing efficiency between the commercially available Lipofectamine RNAiMax and the mild photothermy mediated by the Au-nanoconjugates was compared for irradiated and non-irradiated cells by assessing AcGFP1 expression via RT-qPCR and fluorescence microscopy.A control with non-irradiated cells incubated with AuNP@PEG with the anti-AcGFP1 was used to verify if the presence of AuNP@PEG could promote the intracellular delivery of the silencing oligonucleotide-Fig. 4.
HCT116 cells transiently expressing AcGFP1 protein were observed in a fluorescence microscope after 24 h (Fig. 4b).RT-qPCR data confirmed the silencing of AcGFP1 for all conditions tested (Fig. 4c).Indeed, irradiated cells with prior incubation with AuNP@ PEG (AuNP@PEG + Laser + ASO), showed 70% lower AcGFP1 expression compared to Control (cells without ASO, 2 ΔΔCt = 1).This silencing effect was slightly more pronounced than that obtained for the LipofectamineRNAiMax reagent (approximately 65%).In addition, non-irradiated cells with prior incubation of AuNP@PEG (AuNP@ PEG + ASO) also showed a significant decrease (approximately 45%) of AcGFP1 expression compared to control.Irradiation might be causing the increased uptake of AuNPs, which in turn allows for a "dragging" effect of the oligonucleotide into the cell (Elsner 2012;Stewart et al. 2018).Still, laser irradiation has proven a critical trigger for the uptake of the silencing oligonucleotide since laser irradiation with AuNP@PEG (AuNP@ PEG + Laser + ASO) reduces the expression of AcGFP1 1.8-fold more than without laser irradiation.
None of these conditions showed any impact on cell viability as analysed by the MTS assay (Fig. 4d).

Mild photothermy for copGFP silencing in MCF-7/GFP cells
We then confirmed the robustness of our approach in the MCF-7 breast adenocarcinoma cell line, which has been reported to be more difficult to transfect with siRNA Gene expression levels were normalised to the cells without anti-AcGFP1 (Control).Black asterisks indicate statistical differences between samples and Control (*p < 0.05; **p < 0.01; Unpaired parametric t test with Welch's correction).Data represents the mean value ± the standard error mean of at least three biologically independent experiments with two technical replicates for each.d Cell viability analysis of the different samples tested after 24 h of AcGFP1 silencing via MTS assay.The three conditions tested were normalised to Control.Statistical differences were not observed between all the samples tested with cell control without anti-AcGFP1 (Control).Data represents the mean value ± the standard error mean of two biologically independent experiments with two technical replicates for each than the colorectal model cell line HCT116 (Mokhtary et al. 2018;Yu et al. 2016).In addition, these cells constitutively express GFP and do not require prior use of transfection agents (like lipofectamine) for GFP expression, thus constituting a more challenging model to transfect the siRNA.Cells were incubated with 10 nM of AuNP@PEG for 4 h at 37 ℃, then the cells' culture medium was replaced with DMEM without phenol red at (37 ℃) containing 20 nM of anti-copGFP and irradiated at 2.37 W/cm 2 for 60 s.Irradiated and non-irradiated cells were allowed to recover for 6 h or 24 h, and copGFP expression was analysed by RT-qPCR and fluorescence microscopy at each time point (see Additional file 1: Figure S6).
Mild photothermy significantly reduced copGFP expression by 50% at 6 h and 24 hsee Fig. 5a.This silencing efficacy is comparable to that attained for the commercial Lipofectamine reagent.The quantitative analysis of corrected total cell fluorescence (CTCF) of microscopy images at 6 h and 24 h post-transfection corroborates the silencing of gene expression.A similar reduction of copGFP protein expression may be observed at 6 h and 24 h for AuNP@PEG + Laser + ASO and Lipofectamine + ASO-Fig.5b.Several reports describe the cytotoxicity of cationic lipid reagents (Kasai et al. 2019;Tenchov Fig. 5 RT-qPCR analysis, CTCF evaluation of the fluorescence microscopy images of MCF-7/GFP and cell viability assay.a copGFP gene expression in different conditions tested at 6 h and 24 h.Gene expression levels were normalised to the cells without anti-copGFP (Control).Black asterisks denote statistical differences between samples and the Control (*p < 0.05; **p < 0.01; ***p < 0.001; two-way ANOVA-Mixed-effects analysis with Tukey's multiple comparison test).Data represents the mean value ± the standard error mean of at least three biologically independent experiments with two technical replicates for each.b CTCF analysis to evaluate copGFP protein expression at 6 h and 24 h post-transfection.Statistical differences were not observed between all the samples tested with control without anti-copGFP (Control).Data represent the mean value ± the standard error mean of two biologically independent experiments with three technical replicates each.c Cell viability analysis of the different samples tested after 6 h and 24 h of copGFP silencing via MTS assay.The three conditions tested were normalised to control.Statistical differences were observed between the cells treated with lipofectamine and cells without anti-copGFP (Control).Data represent the mean value ± the standard error mean of at least three biologically independent experiments with two technical replicates each et al. 2012).However, our mild photothermy approach for oligonucleotide transfection does not have an impact on the cells' viability contrary to Lipofectamine-see Fig. 5c.

Mild photothermy for silencing c-MYC proto-oncogene
We then evaluated the efficacy of the mild photothermy combined with AuNP@c-MYC in silencing the frequently overexpressed c-MYC oncogene.Indeed, c-MYC is a suitable model for gene silencing concepts for eventual gene therapy approaches since this protooncogene is a well-known transcription factor involved in cell cycle homeostasis and often dysregulated in cancer cells (Duffy et al. 2021;Strippoli et al. 2020), and several studies have reported its efficacious silencing by ASO vectorised by AuNPs (Conde et al. 2012;Huo et al. 2019;Susnik et al. 2023).
We first optimised the c-MYC silencing at different time points (6 h, 9 h, 12 h, and 24 h) in HCT116 2D cells using AuNP@c-MYC, using AuNP@PEG 30% and AuNP@scramble as controls.An effective decrease of c-MYC expression of approximately 70% was observed at 9 h, which was then selected for subsequent experiments (see Additional file 1: Figure S7).Cells were challenged for 9 h with the Au-nanoconjugates and laser irradiation was conducted 4 h post Au-nanoconjugate challenge.Cell viability was assessed by the MTS Assay, c-MYC expression was analysed by RT-qPCR, and the c-MYC protein expression was evaluated via Immunofluorescence assay (see Additional file 1: Figures S8 and S9).
Figure 6 shows the significant reduction of c-MYC expression (25% silencing compared to untreated cells Control, 2 ΔΔCt = 1).This silencing effect was higher than that obtained for AuNP@c-MYC alone (approximately 20%).The quantitative analysis of CTCF of fluorescence microscopy images shows a reduction of c-MYC protein expression corroborating the data attained for the mild photothermy using AuNP@c-MYC.
A small decrease in cell viability upon 9 h of silencing compared to untreated cells is observed which might be attributed to the downregulation of c-MYC itself, which would have an impact on cell proliferation (Russo et al. 2003;Santo et al. 2023).
Spheroids are 3D cell models that more closely resemble the in vivo behaviour of cell growth than the standard 2D monolayer culture (Abbas et al. 2023;Jubelin et al. 2022;Pinto et al. 2020).As such, we used HCT116 3D spheroids models for a final proof-ofconcept for c-MYC silencing.Following spheroid's growth for 3 or 7 days, they were challenged with Au-nanoconjugates and c-MYC silencing was evaluated as described above for 2D cell cultures (see Additional file 1: Figures S10-S13).The impact on the spheroids' viability was assessed using the LIVE/DEAD Viability/Cytotoxicity assay (see Additional file 1: Figures S14 and S15) and the LDH assay (see Additional file 1: Figure S16).
Figure 7 highlights the effective silencing of the c-MYC for the mild photothermy conditions in 7-day spheroids (AuNP@c-MYC + Laser), with a 30% reduction in the expression of c-MYC.In addition, statistically significant differences between AuNP@c-MYC + Laser and AuNP@c-MYC highlight the effect of laser irradiation (photothermy) to potentiate c-MYC silencing.Interestingly, our mild photothermy did not show the same pronounced reduction of c-MYC expression in the 3-day spheroids.This might be because 3-day spheroids are actively proliferating, where c-MYC is continuously being positively induced (Białkowska et al. 2020;Nayak et al. 2023;Valente et al. 2023).Still, there is a reduction of 5% in AuNP@c-MYC + Laser samples in 3-day spheroids (Fig. 7a).
The impact of c-MYC silencing on cell viability in non-irradiated and irradiated spheroids was assessed by LDH assay and LIVE/DEAD Viability/Cytotoxicity assay (Fig. 9).The ratio between Calcein AM and EthD-1 was calculated since the living cells active intracellular esterase to cleavage the cell-permeant Calcein AM to intensely fluorescent Calcein (Live cells), and EthD-1 is a cell-impermeant that enters in cells with compromised membranes (Dead cells).Irradiated 3-day spheroids show a compromise cell membrane as indicated by the statistically significant decrease to the Calcein AM/ EthD-1 ratio, 1.4-fold, and twofold for AuNP@PEG + Laser and AuNP@c-MYC + Laser more than Control, respectively-see Fig. 9a.In 7-day spheroids, a slight reduction to the Calcein AM/EthD-1 ratio in AuNP@PEG + Laser and AuNP@c-MYC + Laser (around 1.2-fold more) compared to Control is observed although without statistical significance-Fig.9b.These observations are corroborated by those attained by the LDH assay, where 3-day spheroids showed a threefold and 3.3-fold higher LDH activity for AuNP@PEG + Laser and AuNP@c-MYC + Laser compared to Control (see Additional file 1: Figure S16) and in 7-day spheroids were observed approximately 1.7-fold more than Control.These observations are in agreement with previous reports using AuNP or other photothermal agents with laser irradiation (Bromma et al. 2023;Camarero et al. 2023;Pereira et al. 2017).
Few studies have reported the combination of laser irradiation and AuNP for gene silencing in a 3D spheroids model (Huo et al. 2019;Lazzari et al. 2017).Small gold nanoparticles seem to penetrate more than larger AuNPs, indicating that internalisation into spheroids depends on the size of AuNP (Huang et al. 2012;Oliveira et al. 2024).Huang et al. developed a pulsed NIR light-based technique for the spatiotemporal control of gene silencing in 3D-cultured human embryonic stem cells, using siRNA covalently attached to hollow gold nanoshells (Xiao Huang et al. 2016).In addition, 3-day spheroids were more sensitive to localised heat compared to 7-day spheroids, which reinforces the need for the characterisation of spheroid size and structure, cells-cells contact, and TME modifications to allow comparison between gene silencing studies with AuNP@ASO combined with photo-irradiation.

Conclusions
Herein, we demonstrate for the first time the effective silencing of c-MYC via mild photothermy in colorectal carcinoma 3D spheroids.The use of low concentrations of Aunanoconjugates combined with an LDI 2.37 W/cm 2 for 60 s shows no impact on cell viability but allows for improved uptake into the cell.The mild photothermy might trigger modification of the fluidity and permeability of the cell membrane, potentiating uptake.Our mild photothermy approach was able to attain comparable silencing to that of a commercial Lipofectamine reagent but with less cytotoxic effect on the MCF-7/GFP cells.Moreover, our results highlight the enhancement of c-MYC silencing in HCT116 2D cells and 7-day 3D spheroids due to the proposed mild photothermy to improve the internalisation of AuNP@c-MYC.The complexity of 3D cell growth showed that there is a clear difference in sensitivity of 3-day spheroids to localised heat compared to 7-day

Fig. 1
Fig. 1 Overview of the general concept for intracellular delivery of silencing nucleic acids in 3D spheroid.The localised irradiation promotes NP uptake and effective delivery of ASO for gene silencing-triggered by mild hyperthermia (Created with BioRender.com)

Fig. 4
Fig. 4 Overall concept of laser-induced transfection, fluorescence microscopy images of HCT116, RT-qPCR analysis, and cell viability assay.a Overall concept of the transfection of the silencing oligonucleotide mediated by mild photothermy-upon irradiation, AuNPs induce cell uptake and the silencing oligonucleotide (siRNA/ASO) enters the cells (Created with BioRender.com).b Fluorescence microscopy images of HCT116 transiently expressing AcGFP1 protein 24 h after silencing experiment in the different conditions tested.Scale bars correspond to 50 μm.c AcGFP1 gene expression in different conditions tested.Gene expression levels were normalised to the cells without anti-AcGFP1 (Control).Black asterisks indicate statistical differences between samples and Control (*p < 0.05; **p < 0.01; Unpaired parametric t test with Welch's correction).Data represents the mean value ± the standard error mean of at least three biologically independent experiments with two technical replicates for each.d Cell viability analysis of the different samples tested after 24 h of AcGFP1 silencing via MTS assay.The three conditions tested were normalised to Control.Statistical differences were not observed between all the samples tested with cell control without anti-AcGFP1 (Control).Data represents the mean value ± the standard error mean of two biologically independent experiments with two technical replicates for each

Fig. 6
Fig. 6 RT-qPCR analysis and CTCF evaluation of c-MYC protein expression by Immunofluorescence assay of HCT116 2D cells.a c-MYC gene expression at 9 h of gene silencing.Gene expression levels were normalised to respective Controls.Black asterisks indicate statistical differences between samples and Control (**p < 0.01; ns-not statistically significant; Unpaired parametric t test with Welch's correction).Data represent the mean value ± the standard error mean of at least three biologically independent experiments with two technical replicates each.b CTCF analysis via Immunofluorescence assay to evaluate c-MYC protein expression at 9 h of gene silencing.Black asterisks indicate statistical differences between the samples and Control (*p < 0.05; **p < 0.01; Unpaired parametric t test with Welch's correction).Data represents the mean value ± the standard error mean of two biologically independent experiments with three technical replicates each.c Fluorescence microscopy images of 9 h after silencing experiment in Control, AuNP@c-MYC + Laser, and AuNP@c-MYC samples.Scale bars correspond to 20 μm

Fig. 8
Fig. 8 CTCF evaluation of c-MYC protein expression in 3-and 7-day HCT116 spheroids.CTCF analysis of Immunofluorescence assay evaluates c-MYC protein expression at 9 h after gene silencing in (a) 3-day spheroids and (b) 7-day spheroids.Statistical differences were not observed between the samples tested with Control (ns not statistically significant; Unpaired parametric t test with Welch's correction).Data represent the mean value ± the standard error mean of two biologically independent experiments with three technical replicates for each.c Fluorescence microscopy images of Control, AuNP@c-MYC, and AuNP@c-MYC + Laser sample in 3-day spheroids.Scale bars correspond to 300 μm.d Same as for c) in 7-day spheroid

Figure S10 .
RT-qPCR analysis of c-MYC silencing and CTCF evaluation of c-MYC protein expression in 3-day spheroids.Figure S11.RT-qPCR analysis of c-MYC silencing and CTCF evaluation of c-MYC protein expression in 7-day spheroids.Figure S12.Fluorescence microscopy images in 3-day spheroids by Immunofluorescence assay.Figure S13.Fluorescence microscopy images in 7-day spheroids by Immunofluorescence assay.Figure S14.Fluorescence microscopy images in 3-day spheroids by Live/DEAD Viability/ Cytotoxicity assay.Figure S15.Fluorescence microscopy images in 7-day spheroids by Live/DEAD Viability/ Cytotoxicity assay.Figure S16.LDH activity analysis after 9 h of c-MYC silencing via LDH Assay in 3D spheroids.