2.1. AuNPs

AuNPs were synthesized as described previously [10, 11]. Briefly, to a boiling solution (95 mL) of HAuCl₄ (1 mM), 5 mL of disodium citrate (1% w/v) was added dropwise. The solution mixture was stirred continuously for 15 min. The mixture was then allowed to cool down to room temperature taking about 2 h. And the AuNP solution was stored at 4°C away from light for further usage. The characteristics of AuNPs have been reported in detail in our previous publications [10, 11]. The mean size of the AuNPs was 62.2 ± 6 nm [10, 11].

2.2. Cell Line Culture

LNCaP (HSPC), PC3 (CRPC), and human monocyte THP-1 cells were cultured in a conventional condition. THP-1 cells were used to establish M0 macrophages by adding 250 ng/mL phorbol ester (PMA), as described previously [18]. For coculturing cancer cells and macrophages, LNCaP or PC3 was seeded on the upper chamber of Corning Costar Transwell inserts (0.4 μm pores; Sigma–Aldrich, St. Louis, MO, USA), which contain membranes permeable to liquids but not cells. The inserts were then placed in compatible wells and plated onto M0 THP-1 cells.

2.3. Molecular-Level Detection

Cytokines were measured using enzyme-linked immunosorbent assay (ELISA) kits with purified coatings and detected using biotinylated antibodies. Other proteins were measured by western blotting assay, while mRNA levels were determined by real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) assay.

2.4. Autophagic Vesicle (AV) Counting

The cells were fixed with 25 g/L glutaraldehyde, dehydrated, soaked, embedded, and sliced according to the standard procedure for TEM specimen preparation [19]. The slice thickness was set to 60 nm. Ultrathin slices were stained with a uranium acetate solution for 20 min and treated with a lead solution for 5–6 min. TEM was used to analyze autophagosome and autolysosome formation at different stages. The methods for counting AVs (autophagosomes and autolysosomes) were as described previously [19].

2.5. Luciferase Assay

The wild-type (WT) sequence of 3′ untranslated regions (3′-UTRs) of ATG5 and the A-to-T mutant sequence of the specific m⁶A site were synthesized by FitGene (Guangzhou, China). These sequences were inserted into a pLuciferase reporter vector (pLuc). The 293T cells were seeded in 12-well plates at a density of 10⁵ cells/well. Subsequently, cells in designated wells were transfected with 10 ng of the pLuc vector and 1 ng of the control luciferase vector using 0.08 μL of GeneJuice transfection reagent (Merck Millipore, Darmstadt, Germany). After 48 h, luciferase activity was assessed using the Dual-Luciferase Reporter (DLR) Assay (Promega, Madison, WI, USA).

2.6. Methylated RNA Immune Precipitation (MeRIP)–qPCR

MeRIP–qPCR analysis was performed as previously described [20]. Briefly, mRNA was fragmented using an RNA fragmentation reagent (Invitrogen, Waltham, MA, USA) and was then immunoprecipitated with anti-m⁶A antibody coupled to Dynabeads (Invitrogen). The m⁶A-containing mRNAs were precipitated with five glycogen (AM9510; Life Technologies, Carlsbad, CA, USA). The m⁶A enrichment was determined using qPCR.

2.7. Quantification of Global Methylation

Global m⁶A and m⁵C levels were quantified using EpiQuik m⁶A RNA Methylation Quantification Kit (EpigenTek, Farmingdale, NY, USA) and MethylFlash 5-mC RNA Methylation ELISA Easy Kit (EpigenTek).

2.8. Xenograft Mice Model

In vivo experiments were performed in the animal laboratory of Beijing Langke Biotechnology Co., Ltd. (Beijing, China) and were approved by the Institutional Animal Care and Use Committee (IACUC) of the institution (approval number: IACUC-20230111-1). Male severe combined immune deficiency (SCID) mice were propagated in Charles River (MA, USA) and used to construct xenograft models as described previously. Briefly, 6- to 8-week-old male SCID mice were used to establish models of PCa xenografts via surgical orthotopic tumor transplantation. For each mice experiment, at the endpoint, CO₂ euthanasia was performed on mice gradually until the animals lose consciousness and passed away peacefully.

Approximately 100 μL of the LNCaP cell or PC3 cell suspensions were pulled up into the syringe, and then the syringe was capped with the needle. Subsequently, an amount of 100 μL of Matrigel was drawn up through the needle and was mixed up with the cells by moving the plunger up and down. The mixture (final dose, 5 × 10⁶ cells/mouse) was injected into the prostate capsule until the membrane bulged and separated from the prostate surface. After inoculation, tumor growth and infection were observed daily. The tumor size was monitored weekly. Once the tumor volume reached 100 µL, the animals were categorized into different subgroups according to the treatment method. At the endpoint, blood samples were used to determine cytokine levels, and tumor tissues were collected for the immunofluorescence assay.

Inductively coupled plasma mass spectrometry (ICP-MS) was used to quantify the amount of gold taken up by the tumors, as described previously [8]. Parts of the tumor tissues from the AuNP-treated mice were homogenized and lysed, and the product was used for ICP-MS analysis.

2.9. Bone Marrow–Derived Macrophage (BMDM) Cultures and Infusion into the Xenograft Mice

Bone marrow cells were isolated from the mice and inducted into BMDMs as described previously [16]. Subsequently, the cells were stimulated with IL-4 and IL-13 (20 ng/mL) for 48 h to polarize them into M2 BMDM. Next, the cells were harvested and infused into the mice at a concentration of 5 × 10⁶ cells/mouse every second day for 14 days.

2.10. Immunofluorescence Assay

The samples were blocked, and the mouse anti-CD68 primary antibodies (Abcam, Cambridge, UK) were applied, followed with CY3-conjugated antimouse IgG (Abcam) treatment. Subsequently, the rabbit anti-CD163 antibodies (Abcam) were introduced, followed with CY5-conjugated antirabbit IgG (Abcam). Immunofluorescence images were captured to identify the regions with double-positive staining.

2.11. Statistical Analysis

The results were expressed as means ± SD from at least three separate experiments performed in triplicate. The statistical significance of differences between groups was determined using a two-tailed Student’s t-test. p values less than 0.05 were considered statistically significant. All analyses were performed using SPSS software (IBM Corp., Armonk, NY, USA).

3.1. AuNPs Inhibited M2 Polarization of TAM in Both HSPC and CRPC

To investigate the role of AuNPs in TAM polarization, M0 macrophages derived from THP-1 cells were cocultured with LNCaP and PC3 cells, with or without 50 mg/L AuNPs, for 24 h to stimulate TAMs. AuNPs decreased the M2 markers, including IL-10, TGF-β, and Arg 1 mRNA (Figure 1A–C) and increased the M1 markers, including IL-6, TNF-α, and iNOS mRNA (Figure 1D–F) in both LNCaP- and PC3-cocultured TAMs. However, the flow cytometry results showed that the proportion of CD163+ (M2 marker) was decreased by AuNPs in both LNCaP- and PC3-cocultured TAMs (Figure 1G, Supporting Information 1: Figure S1). These results indicate that AuNPs inhibited the M2 polarization of TAM in both HSPC and CRPC.

3.2. AuNPs Inhibited Autophagy of TAMs in Both HSPC and CRPC

Autophagy is strongly associated with TAM polarization [21–23]. Therefore, we investigated whether AuNPs alter the autophagy levels in TAMs. First, western blotting results indicated that the LC3-II/GAPDH ratio (autophagy-positive marker) was increased, whereas the SQSTM1/GAPDH ratio (autophagy-negative marker) was decreased in PC3-cocultured TAMs compared with LNCaP-cocultured TAMs. AuNP treatment significantly decreased the LC3-II/GAPDH ratio and increased the SQSTM1/GAPDH ratio (Figure 2A). The number of AVs reflects the degree of autophagy [17]. The number of AVs increased in PC3-cocultured TAMs compared with LNCaP-cocultured TAMs. AuNP treatment significantly decreased the number of AVs in both LNCaP- and PC3-cocultured TAMs (Figure 2B). Furthermore, several autophagy-related genes, including ATG5, ATG7, ATG12, and BECN1, were significantly downregulated by AuNPs, with ATG5 showing the greatest level of inhibition (Figure 2C,D). Therefore, we explored the regulatory mechanisms of AuNPs on ATG5-associated autophagy.

3.3. AuNPs Inhibited ATG5 Levels in an m⁶A-Dependent Manner

mRNA may be induced transcriptionally or post-transcriptionally [24]. AuNPs regulate some key targets through post-transcriptional regulation [25]. Therefore, we investigated whether AuNPs regulate ATG5 levels post-transcriptionally. In AuNP-treated LNCaP- and PC3-cocultured TAMs, ATG5 mRNA decayed significantly under actinomycin D (ActD, transcription inhibitor, 5 mg/L) treatment, and AuNPs significantly accelerated this decay (Figure 3A,B), confirming their post-transcriptional regulatory effects on ATG5.

The m⁶A methylation in 3′-UTR is a recently discovered mechanism of post-transcriptional regulation of RNA stability [26, 27]. Next, we focused on the m⁶A methylation-dependent regulation of ATG5 by AuNPs for the following reasons: (1) It has been reported that AuNP treatment can affect DNA methylation [28]; (2) we found that AuNP treatment could decrease global m⁶A methylation levels but had little impact on global m⁵C methylation levels (Supporting Information 2: Figure S2); and (3) we found that METTL3, which is considered to be the most important m⁶A writer [26, 27], was downregulated by AuNP treatment (Supporting Information 3: Figure S3).

The MeRIP–qPCR results revealed that ATG5 mRNA had lower m⁶A levels following AuNP treatment, confirming our hypothesis that AuNPs can regulate m⁶A methylation-dependent modification of ATG5 mRNA (Figure 3C). We constructed a luciferase reporter containing WT ATG5 3′-UTR or the ATG5 3′-UTR with specific m⁶A site mutations (MUTs), as described previously [20], and transiently transfected them into HEK293T cells. The luciferase level of WT group of ATG5 3′-UTR was significantly decreased in response to AuNPs, whereas it was not affected by AuNPs in the MUT group of ATG5 3′-UTR (Figure 3D), indicating that m⁶A regulation was the main cause of AuNP downregulation of ATG5 RNA stability.

Next, we added METTL3 overexpression (METTL3-OV group) adenovirus or negative control (METTL3-NC group) adenovirus to M0 THP-1 cells and incubated them for 48 h. ATG5/GAPDH and LC3-II/GAPDH were both significantly enhanced in the METTL3-OV group compared with the METTL3-NC group, confirming that METTL3 enhances ATG5/autophagy levels (Supporting Information 4: Figure S4). Subsequently, M0 macrophages in the METTL3-NC and METTL3-OV groups were cocultured with LNCaP and PC3 cells, with or without 50 mg/L AuNPs, for 24 h to stimulate TAMs. The protein levels of ATG5 were decreased by AuNPs in the METTL3-NC group, and this effect was reversed in the METTL3-OV group (Figure 3E). Collectively, these results demonstrate that AuNPs inhibited ATG5 levels in an METTL3/m⁶A-dependent manner.

3.4. ATG5-Associated Autophagy Is Responsible for Mediating the AuNP Effect on TAM M2 Polarization

Next, we analyzed the role of ATG5-associated autophagy in AuNP-mediated M2 polarization. We added ATG5 overexpression (ATG5-OV group) adenovirus or negative control (ATG5-NC group) adenovirus to M0 THP-1 cells for 48 h. Western blotting results showed that LC3-II/GAPDH increased significantly in the ATG5-OV group compared with the ATG5-NC group (Supporting Information 5: Figure S5).

Subsequently, M0 THP-1 cells in the ATG5-NC and ATG5-OV groups were cocultured with LNCaP and PC3 cells, with or without 50 mg/L AuNPs for 24 h to simulate TAMs. ATG5-OV enhanced the IL-10 (Figure 4A), TGF-β (Figure 4B), and CD163+ proportions, confirming that ATG5/autophagy is the positive regulatory factor for TAM M2 polarization. AuNPs decreased the IL-10 (Figure 4A), TGF-β (Figure 4B), and CD163+ proportions (Figure 4C) in both LNCaP and PC3 cocultured macrophages, and ATG5-OV partially reversed this downregulation compared with the ATG5-NC group (Figure 4A–C). These results indicate that ATG5-associated autophagy is at least partially responsible for mediating the effect of AuNPs on M2 polarization of TAMs.

3.5. AuNPs Regulated M2 Polarization in an Autophagy-Dependent Manner in Vivo

Next, we analyzed the role of autophagy in AuNP-regulated M2 polarization in vivo using a AuNPs treated murine xenograft model. The concentration of AuNPs in tumor tissues every treatment cycle (every 48 h) was shown in Figure 5A. There were no significant differences in the concentrations of AuNPs among the different treatment groups (Figure 5A).

At the endpoint on day 14, we found that AuNP treatment decreased the M2 markers, IL-10, and TGF-β levels in blood cytokines of both LNCaP and PC3 xenograft mice, whereas rapamycin cotreatment reversed this effect (Figure 5B,C).

Next, we used double immunofluorescence staining to detect the number of M2 macrophages. There were fewer CD68+ CD163+ cells (M2 macrophages) in the AuNP-treated mice than in the control group, and this function was reversed in the AuNP + RAP group (Figure 5D). Collectively, these results indicate that AuNPs regulate M2 polarization in an autophagy-dependent manner in vivo.

3.6. AuNPs Inhibit Tumor Growth In Vivo At Least Partially Through Targeting M2 TAM

Next, we explored the role of AuNP-regulated M2 TAM on tumor growth in vivo. AuNPs significantly decreased the tumor volume compared with that in the NC group. The therapeutic effects of AuNPs were reversed in the AuNP + M2 group (Figure 6A,B). The survival time of mice was subsequently observed within 80 days, without any additional treatment, and we found that the 2-week treatment with AuNPs significantly increased the overall survival rate, while this was reversed in the AuNP + M2 group (Figure 6C,D). These results indicate that AuNPs inhibit tumor growth in vivo, at least partially through targeting M2 TAM.