CUDC-907 was purchased from TargetMol. The drug powder was dissolved into a 10 mg/ml storage solution with an appropriate amount of dimethyl sulfoxide (DMSO), stored at −80°C. In vivo, CUDC-907 was diluted into suspension by normal saline using an ultrasonic processor (QSONICA SonicatorQ700).

All cell lines, including SK-N-SH, SK-N-BE(2), SH-SY5Y, SK-N-AS and IMR32, were purchased from Cobioer Biosciences Co., Ltd. The SK-N-SH and IMR32 cells were cultured in minimum essential medium (MEM; Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.) and 1% 1 mM sodium pyruvate (Gibco; Thermo Fisher Scientific, Inc.) and 1% MEM non-essential amino acids (MEM NEAA; Gibco; Thermo Fisher Scientific, Inc.). The SK-N-BE(2) and SH-SY5Y cells were cultured in MEM/F12 (1:1) (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% FBS, 1% 1 mM sodium pyruvate and 1% MEM NEAA. The SK-N-AS cells were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% FBS. All culture mediums were supplemented with 1% penicillin-streptomycin solution (Gibco; Thermo Fisher Scientific, Inc.). All cells were cultured in a 5% CO₂ and humidified incubator maintained at 37°C. All cell lines had been authenticated by STR profiling.

CUDC-907 at the initial maximum concentration (2 mM) was diluted at a gradient ratio of 1:10 in 96-well plates with adherent cells. The blank control group received DMSO instead of CUDC-907. In addition, each well was provided with three vice-holes. Moreover, 10 μl CCK-8 (ApexBio) was added to each well of the plate. The absorbance at 450 nm was measured using a microplate reader (Tecan Spark 10M, Tecan Group, Ltd.) after incubating at 37°C for 2 h.

The cells were prepared into a cell suspension at a concentration of 2×10³/ml. After 24 h, an appropriate amount of PBS was added for washing. Medium containing various concentrations (2, 4, 8 and 16 nM) of CUDC-907 was then added, and each well was provided with three vice-holes. The cells were incubated at 37°C in the 5% CO₂ humidified incubator for 10-14 days. Following removal from the incubator, the cells were fixed with 10% formalin (Biosharp Life Sciences) for 20 min, and then stained with crystal violet (MilliporeSigma) for 10 min, with both procedures conducted at room temperature.

The apoptosis of NB cells treated with or without CUDC-907 was analyzed using flow cytometry. All the procedures were performed following the manufacturer's instructions (Annexin V-FITC/PI Kit, 4A Biotech, Co., Ltd.). Briefly, a total of 5×10⁵ NB cells were prepared, washed in PBS twice, and then resuspended with 500 μl binding buffer. The cells were then incubated with 5 μl Annexin V-FITC for 15 min in the dark at room temperature, and 5 μl propidium Iodide (PI) was then added to each tube. The stained cells were then analyzed using a flow cytometer (CytoFLEX, Beckman Coulter, Inc.) within 1 h.

Whole-cell lysates were generated using RIPA lysis buffer (Thermo Fisher Scientific, Inc.), and the protein concentration was detected using a BCA kit (Thermo Fisher Scientific, Inc.). An equal amount of protein was separated using 10% SDS-PAGE and then transferred onto polyvinylidene fluoride (PVDF) membranes. After blocking in 5% skim milk (EpiZyme) for 1 h at room temperature, the PVDF membranes were incubated overnight with primary antibodies at 4°C, followed by incubation with secondary antibodies for 2 h at room temperature. All the antibodies used are listed in Table SI. Signal detection was conducted using the ECL chemiluminescence detection system (Bio-Rad Laboratories, Inc.).

Total RNA was extracted from the cultured NB cells using TRIzol reagent (Ambion; Thermo Fisher Scientific, Inc.). RNA was reverse transcribed using the RT001 Fast Reverse Transcription kit (ES-RT001, ES Science). qPCR was then performed on the Bio-Rad CFX96 (Bio-Rad Laboratories, Inc.) with SYBR-Green Master Mix (ES-QP002, ES Science). The primers used are listed in Table SII. The amplification reactions were conducted using the following cycling parameters 95°C for 10 min, followed by 40 cycles of 95°C for 10 sec, and 60°C for 30 sec. The 2^−ΔΔCq method was used to calculate the levels of gene expression (19).

Transwell inserts (BD Biosciences) coated with 50 μl Matrigel were used for the Matrigel invasion assay as previously described (20).

The NB cells at a density of 10⁵ cells/ml were cultured in 6-well plates with scratch plug-in components. The protocol of the wound healing assay was as previously described (20).

FBS-free DMEM/F12 supplemented with 2% B27 (Invitrogen; Thermo Fisher Scientific, Inc.), 20 ng/ml human recombinant EGF (Invitrogen; Thermo Fisher Scientific, Inc.) and 20 ng/ml bFGF (Invitrogen; Thermo Fisher Scientific, Inc.) were used for the sphere formation experiment. A total of 300 SK-N-BE(2) cells suspended in tumor sphere medium were seeded into each ultra-low attachment 24-well plate for 14 days. After forming spheres, the cells were transferred to another new 24-well plate for further culture and different generations of sphere-forming cells were collected for RT-qPCR assays.

A total of 55 patients included in the present study were newly diagnosed with NB in the Sun Yat-sen University Cancer Center between April, 2009 to June, 2016. These patients were between 10 to 179 months old, and the median age was 45 months, with 25 females and 30 males. The selection criteria for enrolling patients were as follows: i) A pathological diagnosis of NB; ii) an age <18 years; iii) tissue specimens were obtained prior to the initiation of treatment; and iv) a complete and detailed treatment process and follow-up data. The specimens from patients with NB were obtained by needle biopsy or open surgery. The research protocol was approved by the Institutional Review Board (IRB) of the Sun Yat-sen University Cancer Center (SYSUCC; Guangzhou, China; approval no. B2021-274-01). Informed written consent was obtained from the parents of each patient involved in the study.

The NB tissue specimens were fixed in formalin, embedded in paraffin blocks, and sectioned at a thickness of 4 μm. IHC was performed following standard protocols. The sections were blocked with 5% goat serum (C0265, Beyotime Institute of Biotechnology) at room temperature for 30 min. They were then incubated with primary antibodies against Ki67 (1:500; cat. no. ab92742, Abcam), HDAC1 (1 μg/ml; cat. no. ab19845, Abcam), HDAC2 (1:1,000; cat. no. ab32117, Abcam), HDAC3 (1:500; cat. no. ab32369, Abcam), CD44 (1:200; cat. no. sc-7297, Santa Cruz Biotechnology, Inc.) at 4°C. The following day, the sections were washed in PBS three times and then incubated with goat anti-rabbit secondary antibody (HRP-conjugated, 1:200; cat. no. CW0103S, CWBio) or goat anti-mouse secondary antibody (HRP-conjugated, 1:200; cat. no. CW0102S, CWBio) for 2 h at room temperature. The results were evaluated according to the staining intensity and the proportion of tumor cells with an unequivocal positive reaction. The intensity was scored as follows: 0, negative; 1, weak; 2, moderate; and 3, strong. The frequency of positive cells was defined as follows: 0, <5%; 1, 5-25%; 2, 26-50%; 3, 51-75%; and 4, >75%. The composite score was the product of the two scores. The composite scores of 0 to 7 were considered a low expression, and 8 to 12 were considered a high expression. A fluorescence microscope (Olympus BX61, Olympus Corporation) was used for image acquisition and two independent pathologists reviewed the slides and evaluated the scores.

Human CD44 expression plasmids purchased from Guangzhou iGene Biotechnology Co., Ltd. were transfected into SK-N-BE(2) cells using Lipofectamine 3000^® (Thermo Fisher Scientific, Inc.) for 48 h. The cells were then treated with culture medium containing CUDC-907 for 24 h. Subsequent procedures of assay were conducted using the Dual-Luciferase Reporter Gene Assay kit (DL101-01, Vazyme Biotech Co., Ltd.) following the manufacturer's protocol. A total of 20 μl cell extract was mixed with 100 μl luciferase assay reagent at room temperature and the reaction was detected with a microplate reader (Tecan Spark 10M, Tecan Group, Ltd.). The ratio of Firefly to Renilla Luciferase activity was calculated for each hole.

All the animal experiments were carried out in accordance with the Animal Care and Use Committee of the Sun Yat-sen University Cancer Center, and were approved by the Animal Ethics Committee of Sun Yat-sen University (Approval no. L102042020120P). A mixture of SK-N-BE(2) cell suspension (8×10⁶ cells) and thawed Matrigel (Corning, Inc.) was injected subcutaneously into the right inguinal of 14 female NOD/SCID mice, aged 3-4 weeks (GemPharmatech), weighing between 15 to 20 g, in a SPF environment (room temperature, 20-26°C; relative humidity, 40-70%; alternating time for the light/dark cycle, 12/12 h; food and water were regularly provided by the feeders). Subcutaneous tumor formation was palpable in 12 mice after 1 week, and the mice were then randomly divided into a control group and a treatment group, with 6 mice in each group. The treatment group was administered with 100 μl CUDC-907 solution (25 mg/kg) via gavage for 5 days, and the treatment was continued for 5 days after an interval of 2 days. The control (vehicle) group was administered with 100 μl normal saline via gavage at the same time. During this period, the body weight and tumor volumes of the mice were measured and recorded every 2 days, and tumor volumes were calculated using the formula V=(short diameter² × long diameter)/2. Following a total of 10 days of treatment, the mice were euthanized with carbon dioxide using the displacement rate at 50% volume per minute (21), and the subcutaneous tumors were removed to measure their size and weight. When the tumor diameter of any mice reached 2 cm, this was regarded as the humane endpoint in this experiment.

The detailed procedures for siRNA transfection and lentiviral transduction of the NB cells were as previously described (20). Human siRNA oligos targeting PTX3 were purchased from Shanghai GenePharma Co., Ltd, at a working concentration of 50 nM. The siRNA sequences were as follows: siNC, UUC UCC GAA CGU GUC ACG UTT (5′-3′) and TTA AGA GGC UUG CAC AGU GCA (3′-5′); siPTX3#1, GCA CAA AGA GGA AUC CAU ATT (5′-3′) and UAU GGA UUC CUC UUU GUG CTT(3′-5′); siPTX3#2, GGG AUA GUG UUC UUA GCA ATT (5′-3′) and UUG CUA AGA ACA CUA UCC CTT (3′-5′). The PTX3 overexpression plasmids and lentiviral vectors were purchased from GeneCopoeia.

The cells were treated with DMSO or CUDC-907 at 25 nM for 24 h, and total RNA was extracted using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.). The cDNA library building and RNA-seq were performed using a commercially available service (service ID F21FTSSCWLJ1037, BGI, Huada Biotechnology). RNA-seq reads were aligned to the hg38 genome. Differentially expressed genes (DEGs) were identified using |log2FC| ≥1 and FDR ≤0.001.

All the experiments were performed in triplicate. Data are expressed as the mean ± standard deviation, and comparisons between groups were performed using an unpaired Student's t-test or one-way ANOVA followed by Tukey's post hoc test. Fisher's exact test was used to assess the association between HDACs or the CD44 expression level and clinicopathological variables. Kaplan-Meier curves were used for survival analysis following the log-rank test. Overall survival (OS) was defined as the endpoint of the study as the period from the date of initial diagnosis to mortality or the last follow-up. Event-free survival (EFS) was defined as the period from the date of the initial diagnosis to the date of recurrence, progression, mortality, or a second malignancy. P<0.05 was considered to indicate a statistically significant difference. All the data was analyzed using SPSS 25.0 software (IMB Corp.) or GraphPad Prism 8.0 software (Graphpad Software, Inc.).

CUDC-907 is a dual target inhibitor of HDAC and PI3K. The present study evaluated the effects of CUDC-907 on the viability of five NB cell lines, including MYCN non-amplified NB cell lines (SK-N-SH, SH-SY5Y and SK-N-AS) and MYCN-amplified NB cell lines [SK-N-BE(2) and IMR32]. The results revealed that CUDC-907 inhibited the viability of the five NB cell lines in a concentration-dependent manner within 72 h, and the half-maximal inhibitory concentration (IC50) values calculated by GraphPad Prism ranged from 5.53 to 46.22 nM (Fig. 1A and B). The colony formation assay revealed that cell proliferation was significantly inhibited by CUDC-907 (Fig. 1C). Subsequently, flow cytometric analysis was performed and it was found that CUDC-907 significantly induced apoptosis (Fig. S1A). The results of western blot analysis demonstrated an increased expression of Bax, Bak and cleaved caspase-3, and a decreased protein expression of Bcl-2 in the cells treated with CUDC-907 (Fig. 1D). Furthermore, the inhibitory effects of CUDC-907 on the migratory ability of NB cells were also demonstrated by the wound healing assay (Fig. S1B) and the Matrigel invasion assay (Fig. S1C).

The potential anti-NB effect of CUDC-907 was explored in vivo. The detailed mode of administration is illustrated in Fig. 2A. Following treatment for 10 days, CUDC-907 significantly inhibited NB tumor growth and weight compared with the vehicle control (Fig. 2B, D and E), while there was no significant difference in the body weight of the mice in the two groups (Fig. 2C). The results of western blot analysis (Fig. 2F) revealed that CUDC-907 markedly increased H3K27ac expression, whereas it inhibited the expression of phosphorylated (p-)AKT. IHC staining revealed that CUDC-907 downregulated the expression of the proliferative marker, Ki67, compared to the control tissues (Fig. 2G and H).

MYCN amplification is associated with the poor prognosis of patients with NB (22); thus, the present study investigated whether CUDC-907 downregulates the expression of MYCN. The two MYCN-amplified cell lines, SK-N-BE(2) and IMR32, exhibited higher MYCN mRNA and protein levels compared to the MYCN non-amplified cell lines, SK-N-SH, SH-SY5Y and SK-N-AS (Fig. 3A and B). The SK-N-BE(2) or IMR32 cells were treated with CUDC-907 at the indicated concentrations (0, 10, 25, 50 and 100 nM) for 24 h, or with a fixed concentration of 50 nM for various periods of time (0, 2, 4, 6, 16, 24 and 48 h) to observe the changes in MYCN protein levels. The results revealed that CUDC-907 downregulated MYCN expression in a concentration- and time-dependent manner in the MYCN-amplified NB cell lines (Fig. 3C-F).

Furthermore, the downstream targets directly regulated by CUDC-907 were verified. The results revealed that H3K27ac expression was markedly increased in the NB cell lines, SK-N-SH, SK-N-BE(2) and IMR32, following exposure to CUDC-907 for 24 h. Moreover, the expression of p-AKT, which is downstream of activated PI3K, was also obviously decreased (Fig. 3G). Activating mutations in the RAS-MAPK-ERK pathway are known to occur at a high frequency in relapsed NB (23,24). Thus, the present study examined the changes in the downstream proteins of MAPK, p-MEK/MEK and p-ERK/ERK, and found that the p-ERK expression level was markedly decreased by CUDC-907 (Fig. 3G).

Cancer stem cells (CSCs) lead to post-transplant relapse and are associated with poor survival outcomes of patients with high-risk NB (25,26). Herein, to explore whether CUDC-907 affects the stem cell-like properties of NB, a sphere formation assay was performed using the SK-N-BE(2) cells exposed to the indicated concentrations (0, 2, 4, 8 and 16 nM) of CUDC-907. Following treatment of the SK-N-BE(2) cells with 4, 8 or 16 nM CUDC-907, clonogenic formation exhibited a marked decrease compared to the controls in concentration-dependent manner (Fig. 4A and B). Furthermore, both the mRNA and protein levels of several stem cell markers, including CD44, SOX2, OCT4 and BMI-1, were significantly downregulated by CUDC-907 (Fig. 4C and D).

To further explore the mechanisms through which CUDC-907 affects the stem-like properties of NB cells, RNA-seq analysis was performed using two NB cell lines [SK-N-BE(2) and SK-N-SH]. According to the DEGs painted into a heatmap, 20 DEGs were identified (Fig. S1D), of which PTX3 was the only downregulated gene. CUDC-907 significantly inhibited the expression of PTX3 at the mRNA level in NB cells (Fig. 5A). PTX3 is a ligand and upstream protein of CD44, which has been identified as a CSC marker in several types of cancer, including NB (27-29). PTX3 has also been reported to promote the stemness of breast cancer cells by activating the downstream ERK1/2, AKT and NF-κB pathways (30); however, its role in NB has not been reported to date, at least to the best of our knowledge. In the present study, PTX3 siRNA was transfected into SK-N-BE(2) cells to knockdown PTX3 expression. It was observed that PTX3 knockdown significantly decreased the expression of CD44 and that of a series of other stemness-associated genes, as well as tumor sphere forming ability of the cells (Fig. 5B-E). Even though previous research has indicated that PTX3 activates the AKT pathway (30), the present study revealed that the level of p-AKT did not exhibit any obvious change following the knockdown of PTX3. This discrepancy may be attributed to the different functions of PTX3 in various cell types; thus, these findings need to be verified in future studies. In addition, it was found that the reduction of sphere formation and CD44 mRNA/protein expression by CUDC-907 was reversed by exogenous PTX3 overexpression (Fig. 5F-I). These findings indicated that CUDC-907 inhibited the stem-like properties of NB cells and CD44 expression by inhibiting PTX3.

To observe whether the target proteins of CUDC-907 are associated with the prognosis of patients with NB, IHC of HDAC1, 2, 3 and CD44 was performed on paraffin-embedded sections of 55 patients with NB from Sun Yat-sen University Cancer Center (Fig. 6A). The median follow-up time was 38.9 months (range, 9.6 to 124.6 months). Up to the final follow-up date, 17 patients (30.9%) succumbed and 33 patients (60.0%) suffered disease progression or recurrence.

The association between HDAC1, 2, 3 expression and the clinical characteristics of patients with NB was analyzed (Tables SIII-SV). The results revealed that the expression level of HDAC3 was related to MYCN amplification (OR, 15.789; P=0.002), the International Neuroblastoma Staging System (INSS) staging system (OR, 6.400; P=0.043) and the Children's Oncology Group (COG) risk group (OR, 7.243; P=0.017) (Table SV). Kaplan-Meier survival analysis indicated that the upregulated mRNA levels of HDAC1, HDAC2 and HDAC3 were significantly associated with a poor OS (HDAC1, P=0.0209; HDAC2, P=0.0143; HDAC3, P=0.0063) and EFS (HDAC1, P=0.0023; HDAC2, P=0.0026; HDAC3, P=0.0309) (Fig. 6B-G). It was also found that the high expression of CD44 was significantly associated with a poor OS (P=0.0239) and EFS (P=0.0477) of patients with NB (Fig. 6H and I). However, the high expression of CD44 was found not to be significantly associated with the clinical characteristics of patients with NB (Table SVI).