Abstract

   Bladder cancer (BCa) is a lethal malignancy of the urinary system and
   exhibits a poor prognosis. Chaperonin-containing tailless complex
   polypeptide 1 subunit 3 (CCT3) acts as an oncogene in various tumors,
   whereas its effect on BCa remains unknown. We identified the
   ferroptosis-associated differentially expressed genes through
   bioinformatic analysis and selected CCT3 for further verification. The
   levels of cell viability, apoptosis, migration, invasion, and
   proliferation were measured to clarify the effect of silencing CCT3 on
   BCa cells. Then we evaluated the role of CCT3 knockdown in vivo.
   Ferroptosis was assessed by the expression detection of the
   ferroptosis-related proteins. The underlying mechanism was predicted by
   RNA sequencing and verified by an agonist for nucleotide-binding and
   oligomerization domain 1 (NOD1). Western blotting was conducted to
   detect the protein expression of NOD1, nuclear factor kappa B (NF-κB)
   inhibitor alpha (IκBα), and phospho-IκBα (p-IκBα). In vitro,
   down-regulation of CCT3 suppressed the cell viability, migration,
   invasion, and proliferation, as well as induced apoptosis of BCa cells.
   In vivo, silencing CCT3 elevated the body weight of mice and suppressed
   the BCa progression. In addition, CCT3 knockdown could induce
   ferroptosis in vitro and in vivo. CCT3 knockdown suppressed the
   expression of NOD1 and p-IκBα/IκBα and the NOD1 agonist could reverse
   the effect of CCT3 suppression on BCa in vitro and in vivo. In summary,
   our findings demonstrate that silencing CCT3 inhibits BCa via induction
   of ferroptosis and suppression of the NOD1-NF-κB pathway.

Supplementary Information

   The online version contains supplementary material available at
   10.1038/s41598-024-76724-z.

   Keywords: Bladder cancer, CCT3, Ferroptosis, The NOD1-NF-κB signaling
   pathway

   Subject terms: Biochemistry, Cancer, Cell biology, Chemical biology,
   Biomarkers, Nephrology

Introduction

   Bladder cancer (BCa) is the most prevailing kind of lethal malignancy
   of the urinary system^[34]1. By 2024, it is projected that 83,190 new
   BCa cases and 16,840 mortality are expected to emerge in the United
   States^[35]1. Based on tumor invasion depth, BCa can be divided into
   the non-muscle invasive BCa, with minimal effects on survival, and the
   muscle-invasive BCa, with serious impairments in survival^[36]2.
   Clinically, transurethral resection is the cornerstone treatment for
   non-muscle invasive BCa^[37]3, while platinum-based neoadjuvant
   chemotherapy is the standard of care for patients with muscle-invasive
   BCa^[38]4. Nevertheless, despite these treatments, the survival rates
   of BC patients have not improved significantly over the past two
   decades^[39]5. Hence, it is of great importance to find effective
   therapeutic targets for BCa.

   Ferroptosis represents a unique form of programmed cell death driven by
   the accumulation of lipid peroxidation in an iron-dependent
   manner^[40]6. Among its primary positive regulators are malondialdehyde
   (MDA) and ferrous iron (Fe^2+), whereas the negative include
   glutathione (GSH), superoxide dismutase (SOD), ferritin heavy
   polypeptide 1 (FTH1), recombinant solute carrier family 7, member 11
   (SLC7A11) and glutathione peroxidase 4 (GPX4)^[41]7. Recent research
   has increasingly illustrated the efficacy of enhancing ferroptosis in
   tumor cells as an anti-tumor strategy, due to its contribution to
   reversing drug resistance^[42]8, tumor microenvironment^[43]9, cancer
   metabolism^[44]10, etc. Abnormal expression levels of various
   ferroptosis-related genes have been observed in BCa specimens^[45]11,
   and inducing ferroptosis has been found to repress the progression of
   BCa and even lead to the elimination of BCa cells^[46]12, [47]13,
   indicating the pivotal role of ferroptosis in BCa. Therefore, it is of
   great importance to find new therapeutic targets for BCa based on the
   regulatory mechanism of ferroptosis in BCa.

   Chaperonin-containing tailless complex polypeptide 1 subunit 3 (CCT3)
   has been implicated in a series of biological processes, as well as the
   initiation and progression of different malignant tumors^[48]14.
   Elevated expression levels of CCT3 have been positively correlated with
   carcinogenesis^[49]15. The knockdown of CCT3 has been shown to
   stimulate apoptotic pathways in breast and prostate cancers^[50]15, and
   to promote cervical cancer cell death via down-regulation of
   fibronectin 1^[51]16. Knockdown of CCT3 represses the progression of
   lung adenocarcinoma by inhibiting the function of glycolysis^[52]17. In
   addition, Wang et al. have reported that the overexpression of CCT3 can
   inhibit ferroptosis in lung adenocarcinoma, leading to tumor
   growth^[53]18. Nevertheless, the role of CCT3 in the progression of BCa
   is unclear.

   Herein, CCT3 was identified as a key ferroptosis-related differentially
   expressed gene (DEG) in BCa by bioinformatics analysis and its effect
   on BCa was explored both in vitro and in vivo by down-regulating its
   expression. Our research, for the first time, revealed the anti-BCa
   potential of silencing CCT3 via inducing ferroptosis through
   suppressing the NOD1-NF-κB signaling pathway, which could offer a novel
   and promising therapeutic target for BCa treatment.

Materials and methods

Data acquisition and processing

   The original gene expression data was sourced from the Gene Expression
   Omnibus (GEO) database ([54]https://www.ncbi.nih.gov/geo/). The search
   was conducted using the term “carcinoma of bladder” as a query, with
   the organism limited to “Homo sapiens”. The inclusion criteria
   specified the follows: (1) the comparison involves human bladder tumor
   samples vs. normal bladder tissue samples; (2) the expression profiling
   method by array or high-throughput sequencing; (3) the sample size in
   each group must be at least 3; (4) raw data is available. Finally, the
   datasets [55]GSE133624 and [56]GSE42089 that met the above criteria
   were selected. [57]GSE42089 comprised 18 samples (10 urothelial
   carcinoma samples and 8 para-carcinoma tissue samples) based on
   [58]GPL9828 [HG-U133_Plus_2] Affymetrix Human Genome U133 Plus 2.0
   Array. Meanwhile, [59]GSE133624 contained 65 samples (36 urothelial
   carcinoma samples and 29 para-carcinoma tissue samples) based on the
   [60]GPL20795 HiSeq X Ten (Homo sapiens) were chosen for further
   analysis. The DEGs between carcinoma samples and normal tissue samples
   (control) were identified by GEO2R
   ([61]https://www.ncbi.nlm.nih.gov/geo/geo2r) and all identified DEGs
   met adjusted p-value ≤ 0.05 and |log[2](Fold-change)| ≥ 0.8.

   The Genecards database ([62]https://www.genecards.org/cgi-bin/carddisp)
   was used to retrieve ferroptosis-related genes. The FDEGs were then
   defined as DEGs from [63]GSE42089 and [64]GSE133624 that overlapped
   with the list of ferroptosis-related genes in the Venn diagram.

Functional and pathway enrichment analysis

   We uploaded the above FDEGs to the Database for Annotation,
   Visualization, and Integrated Discovery (DAVID,
   [65]https://david.ncifcrf.gov/summary.jsp) for Gene Ontology (GO)
   function enrichment analysis, which consisted of biological process,
   cellular component, and molecular function, and the Kyoto Encyclopedia
   of Gene and Genome (KEGG) functional pathway enrichment analysis. The
   enriched GO terms and KEGG pathways with the minimum p-value, were
   shown with histograms and bubble plots.

Construction of the protein-protein interaction (PPI) network and acquisition
of hub genes

   The PPI analysis was performed by the Search Tool for the Retrieval of
   Interacting Genes (STRING, [66]http://string-db.org/) with a confidence
   score of 0.4 and was visualized as a network in the Cytoscape
   ([67]https://www.cytoscape.org/). Subsequently, we used the Cytoscape
   plug-in molecular complex detection (MCODE) to filter the significant
   modules (gene clusters) and selected innovative genes as hub genes.

Hub gene analysis

   A principal component analysis was conducted using the expression
   levels of hub genes from the [68]GSE133624 dataset. Then a ridgeline
   plot was generated to visually represent the abundance and distribution
   of expression data of hub genes within the [69]GSE133624 dataset.
   Furthermore, correlation coefficients between the hub genes were
   calculated, based on which the matrix analysis diagram was plotted,
   revealing the internal correlation of hub genes in [70]GSE133624. The
   Gene Expression Profile Interactive Analysis (GEPIA) database
   ([71]http://gepia.cancer-pku.cn/) was used for hub gene expression
   verification.

Cell culture and treatment

   The human BCa cell lines (T24 and 5637) along with the immortal
   ureteral epithelium cell (SV-HUC-1) were purchased from iCell
   Bioscience Inc (Shanghai, China). T24 and 5637 cell lines were cultured
   in Roswell Park Memorial Institute (RPMI)-1640 medium (Gibco, Grand
   Island, NY, USA) and SV-HUC-1 was in F-12 K medium (Thermo Fisher
   Scientific, Inc.USA). Both RPMI-1640 and F-12 K mediums contained 10%
   fetal bovine serum (FBS, Gibco), 10 U/mL penicillin (Gibco), and
   100 µg/mL streptomycin (Gibco).

   To knock down the expression of CCT3, the T24 and 5637 cell lines were
   transfected with small interfering RNA (siRNA) targeting CCT3
   (si-CCT3-1/2–3) using the Lipofectamine 3000 kit (Thermo Fisher
   Scientific). The si-NC was transfected into cells using the same
   methodology. The primer sequences were listed in Supplementary Table 1.
   In addition, T24 cells were exposed to 5 µg/mL
   N-acetyl-muramyl-L-ala-y-D-glutamyl-meso-diaminopimelic acid (M-TriDAP,
   InvivoGen, Toulouse, France), an agonist for nucleotide-binding and
   oligomerization domain 1 (NOD1), for 24 h to activate the NOD1-nuclear
   factor kappa B (NF-κB) signaling pathway^[72]19. Therefore, the cells
   were divided into the following 3 groups: si-NC, si-CCT3, and
   si-CCT3 + M-TriDAP groups.

Real-time quantitative polymerase chain reaction (RT-qPCR)

   Firstly, we extracted total RNAs from cells or tumor tissue samples of
   different groups by TRIZOL reagent (Invitrogen Carlsbad, CA, USA). RNAs
   with an optical density ratio at 260/280 nm ranging between 1.9 and 2.0
   were selected to reverse transcribed into cDNAs. Then RT-qPCR was
   performed on StepOnePlus Real-Time PCR System (Applied Biosystems,
   Foster City, CA, USA), adhering to the instructions. The reaction
   conditions were set as follows: at 95℃ for 30 s initial denaturation,
   followed by 40 amplification cycles (95℃, 10 s, and 60℃, 30 s). The
   sequences of primers were supplied in Supplementary Tables 1 and the
   mRNA relative expression levels were calculated using the 2^−∆∆Ct
   formula using GAPDH as an internal standard.

Cell viability determination

   T24 and 5637 cells from different groups were seeded into 96-well
   plates (3 × 10^3 cells/well). The cell count kit-8 (CCK-8) reagents
   were added to each well after 0, 24, 48, and 72 h of culture, followed
   by an incubation period at 37℃ for 2 h. Then the optical density of
   each well was recorded at 450 nm wavelength.

Apoptosis assay

   T24 and 5637 cells from different groups were inoculated into a
   six-well plate. After 24 h of incubation, the cells were exposed to
   fluorescein isothiocyanate-conjugated anti-annexin V antibody
   (Solarbio, Beijing, China) and propidium iodide (Solarbio) for 15 min.
   With a flow cytometer, we measured the fluorescence intensity which was
   quantified using FlowJo software (Becton Dickinson, Ashland, OR, USA).

Wound healing assay

   T24 and 5637 cells from different groups were inoculated into a
   six-well plate. A scratch was then introduced on the surface of the
   cell layer once the cell confluence reached 80%, followed by 24 h of
   culture. At 0 h and 24 h, with a microscope (Olympus, Tokyo, Japan), we
   captured the cell migration status which was then processed by Image-J
   software.

Transwell assay

   Briefly, T24 and 5637 cells (5 × 10^4/chamber) were seeded onto the
   upper chamber (Corning, Kennebunk, ME, USA), which has been pre-coated
   with a 1: 8 diluted Matrigel. The lower chamber was filled with 500 µL
   of RPMI-1640 medium supplemented containing 20% FBS. After 24 h, the
   invasive cells in the bottom of the upper chamber were fixed with 4%
   paraformaldehyde and stained with 0.1% crystal violet. Finally, cells
   in the bottom of the upper chamber were counted under a microscope
   (Olympus).

Determination of MDA, SOD, GSH, and Fe2+

   Measurements of GSH, SOD, and MDA were conducted with the help of
   corresponding Enzyme-linked immunosorbent assay (ELISA) kits (Esebio,
   Shanghai, China). Additionally, the iron assay kits (Abcam, Cambridge,
   USA) were applied to detect the ferrous iron levels. All procedures
   were conducted in strict accordance with the instructions in the
   manual.

Western blotting

   Total proteins extracted from T24 and 5637 cells or tumor tissues were
   separated by 10% sodium dodecyl sulfate-polyacrylamide gel
   electrophoresis and transferred onto a polyvinylidene fluoride membrane
   (Millipore, Danvers, MA, USA). The membrane with the proteins was
   blocked with 5% skim milk and then incubated with primary antibodies,
   including anti-FTH1 (ab65080; 1:1,000; Abcam), anti-SLC7A11 (ab307601;
   1:1,000; Abcam), anti-GPX4 (ab125066; 1:2,000; Abcam), anti-NOD1
   (ab189409; 1:1,000; Abcam), anti-NF-κB inhibitor alpha (IκBα; ab32518;
   1:2,000; Abcam), anti-phospho-IκBα (p-IκBα; ab92700; 1:1,000; Abcam),
   and anti-GAPDH (ab181602; 1:10,000; Abcam) antibodies at 4℃ overnight.
   After washing, the membrane was incubated with a horseradish
   peroxidase-conjugated anti-rabbit secondary antibody (ab288151;
   1:5,000; Abcam) at room temperature. Finally, the protein bands were
   visualized using a Tanon 5200 chemiluminescence imaging system (Tanon,
   Shanghai, China).

Proliferation assay

   A 5-ethynyl-2-deoxyuridine (Edu) staining proliferation kit (Abcam) was
   utilized to detect the proliferation level of T24 cells. Edu,
   incorporated into newly synthesized DNA, was subsequently combined with
   a green fluorescent called iFluor 488 azide. DAPI (Abcam) was also
   added to stain DNA blue. Images were captured using fluorescence
   microscopy, wherein an increased relative expression of green
   fluorescence indicated a heightened capacity for cell proliferation.

RNA- sequencing and data analysis

   RNA samples extracted from cells of the si-NC and si-CCT3 groups
   underwent transcriptome sequencing, with the aim of identifying DEGs
   between the above groups. DEGs were defined as |log[2](Fold-change)| >
   1 with adjusted p-value ≤ 0.05. The heatmap of cluster analysis of DEGs
   was mapped using the R package “Pheatmap”, which relied on measures of
   similarity, complete linkage, and the Euclidean distance. Using the R
   language ggplots2 software package, a volcano plot of DEGs was drawn to
   show gene distribution and differences in gene expression levels using
   the R package “ggplots2”. Finally, to predict the downstream regulatory
   pathway of silencing CCT3, we subjected the DEGs to GO functions and
   KEGG pathways enrichment analysis.

Establishment of BCa mice model

   A total of 18 female Balb/c nude mice (6-week-old, 18–20 g) were
   obtained from SPF (Beijing) Biotechnology Co., Ltd. (Beijing, China).
   Lentiviruses harboring either empty vectors (LV-NC) or interfering RNA
   lentiviral vectors (LV-CCT3) were purchased from GeneChem Co., Ltd.
   (Shanghai, China). A total of 4 × 10^7 T24 cells, either infected with
   corresponding lentiviral vectors or treated with M-TriDAP, were diluted
   in 80 µL of sterile PBS. These cells were then subcutaneously injected
   into the right flank of mice, which were divided into 3 groups, namely
   LV-NC, LV-CCT3, and LV-CCT3 + M-TriDAP groups (n = 6)^[73]20. The body
   weight of mice was detected every 7 days. After 4 weeks, all mice were
   anesthetized by inhalation of isoflurane and then sacrificed by
   cervical dislocation for collection of tumor tissues^[74]20. All mouse
   experiments were approved by the local Use and Care of Animals
   Committee.

Hematoxylin and Eosin (H&E) staining and immunohistochemical staining

   The tumor tissue samples, which were initially fixed in 4%
   paraformaldehyde and then embedded in paraffin, were sliced into 4 μm
   sections. These sections underwent H&E staining to assess the extent of
   inflammatory infiltration of tumor tissues. In addition, the sections
   were exposed to the rabbit anti-Ki67 antibody (ab16667; 1:100; Abcam)
   at 37℃ overnight, followed by incubation with anti-rabbit IgG
   (ab150077; 1:200; Abcam). Then, the sections reacted with
   3.3-diaminobenzidine. Finally, all the sections were observed by a
   microscope (Olympus).

Statistical analysis

   Data were presented as multiple groups of repeated data or as
   means ± standard deviation and were analyzed using GraphPad Prism 7.0
   statistical software (GraphPad, San Diego, CA, USA). For comparisons
   between two groups, Student’s t-test was employed, whereas for
   comparisons involving multiple groups, one-way ANOVA with Tukey’s post
   hoc analysis was conducted. A p-value less than 0.05 was deemed
   statistically significant.

Results

Identification of DEGs

   We selected the original gene expression profiles from the [75]GSE42089
   and [76]GSE133624 datasets for further screening. There were 5,944 DEGs
   in the [77]GSE133624, including 2,793 up-regulated and 3,151
   down-regulated genes in tumor samples compared to normal samples.
   Similarly, there were 2,159 DEGs in the [78]GSE42089, including 1,035
   up-regulated and 1,124 down-regulated genes. Volcano plots were
   conducted to visualize the DGEs (Fig. [79]1A). Additionally, the top 15
   up-regulated and down-regulated DEGs in the [80]GSE133624 and
   [81]GSE42089 datasets were shown using heatmaps (Supplementary
   Fig. 1A). Then we achieved data correction and standardization for the
   selected samples from both datasets using boxplots (Supplementary
   Fig. 1B) to ensure data consistency.

Figure 1.

   [82]Figure 1
   [83]Open in a new tab

   Identification of differentially expressed genes (DEGs). (A) The
   volcano plots of DEGs in [84]GSE42089 and [85]GSE133624; red points
   presented up-regulated DEGs, and blue points presented down-regulated
   DEGs. (B) The top 3 pathways of gene set enrichment analysis of DEGs in
   [86]GSE42089 and [87]GSE133624. (C) The Veen diagrams of
   ferroptosis-related DEGs (FDEGs), which were defined as DEGs from
   [88]GSE42089 and [89]GSE13362 overlapped with the list of
   ferroptosis-related genes.

   The results of GSEA were listed in Fig. [90]1B. In addition, a total of
   30 up-regulated and 23 down-regulated FDEGs were obtained after Venn
   diagram analysis (Fig. [91]1C), which were selected for GO and KEGG
   analysis. The top 10 GO terms and KEGG pathways with the minimum
   p-value were shown with histograms and bubble plots (Supplementary
   Fig. 2A-2D).

PPI network construction and hub gene selection

   The interaction information of 30 up-regulated and 23 down-regulated
   FDEGs was obtained by the STRING database (Supplementary Fig. 3A and
   3 C). We imported the above FDEGs into Cytoscape and identified 2 gene
   clusters by the plug-in MCODE (Supplementary Fig. 3B and 3D). On the
   basis of innovation, we selected CCT3, cell division cycle 25 A
   (CDC25A), protein regulator of cytokinesis 1 (PRC1), diaphanous 3
   (DIAPH3), and kruppel-like factor 2 (KLF2) as hub genes for the
   following analysis.

Hub gene analysis

   The raw expression data of the 5 hub genes within the [92]GSE133624
   dataset were used for hub gene analysis. A total of 2 principal
   component axes (PC1 and PC2) were generated after principal component
   analysis, collectively explaining 87.7% of the variance between the
   Control group and the Tumor group (Fig. [93]2A). The ridge plot of hub
   gene expression demonstrated that CCT3 possessed the highest expression
   level and the most tightly distributed expression data (Fig. [94]2B).
   The correlation among hub genes was shown in Fig. [95]2C, with CCT3
   exhibiting the lowest correlation with the remaining four genes. Then
   we validated the expression of CCT3 in tumor samples compared with
   normal samples using GEPIA2 and found that its expression level was
   higher in BCa tissues compared to normal tissues (p < 0.05,
   Fig. [96]2D). Subsequently, RT-qPCR displayed that CCT3, CDC25A, PRC1,
   and DIAPH3 had significantly upregulated expression whereas KLF2 had
   significantly downregulated expression in human BCa cell lines, T24 and
   5637, compared to the normal bladder cell line SVHUC-1 (all p < 0.001,
   Fig. [97]3). Then based on our results, corroborated with existing
   literature, we chose CCT3 for experimental validation.

Figure 2.

   [98]Figure 2
   [99]Open in a new tab

   Hub gene analysis. (A) Principal component analysis of hub genes. The
   axes PC1 and PC2 in the figure were the first and second principal
   components. Dots represented samples, and different colors represented
   different groups. (B) A ridge diagram of hub genes. The horizontal axis
   represented the gene expression, the shape of the peak represented the
   dispersion among data, and the height stands for the number of samples
   corresponding to the gene expression. (C) Matrix correlation analysis
   chart. Red represented positive correlations, and blue represented
   negative correlations. (D) The validation of the expression of
   chaperonin-containing tailless complex polypeptide 1 subunit 3 (CCT3)
   in tumor samples compared with normal samples using the Gene Expression
   Profile Interactive Analysis database.

Figure 3.

   [100]Figure 3
   [101]Open in a new tab

   The mRNA expression level of hub genes in SV-HUC-1, T25, and 5637
   cells.^***p < 0.001 vs. SV-HUC-1 group.

Silencing CCT3 inhibits cell growth, invasion, and migration and promotes
apoptosis of BCa cells

   The transfection efficiency of si-CCT3-1/2/3 in T24 and 5637 was
   assessed through RT-qPCR and the results showed that si-CCT3-1/2/3
   transfection could significantly reduce the mRNA level of CCT3 in both
   T24 and 5637 cells (all p < 0.001, Fig. [102]4A). Given its superior
   transfection efficiency, si-CCT3-1 was selected for further
   experimentation. The cell viability, invasion ability, and migration
   rate of T24 and 5637 were significantly reduced whereas the apoptosis
   rate was significantly increased in the si-CCT3 group compared to the
   si-NC group (p < 0.001 or p < 0.01, Fig. [103]4B-[104]4E).

Figure 4.

   [105]Figure 4
   [106]Open in a new tab

   Silencing CCT3 inhibited cell growth, invasion, and migration and
   promoted apoptosis of BCa cells. (A) The transfection efficiency of
   si-CCT3-1/2/3 was determined by RT-qPCR. (B) The viability of treated
   cells was detected by cell counting kit-8 assay. (C) The apoptosis rate
   of treated cells was detected by a flow cytometer. (D) The invasion of
   treated cells was measured by transwell assay. (E) The migration of
   treated cells was detected by wound healing assay. ^***p < 0.001,
   ^**p < 0.01 vs. si-NC group.

Silencing CCT3 induces ferroptosis in BCa cells

   Compared with the si-NC group, increased levels of MDA and Fe^2+
   alongside decreased levels of GSH and SOD were found in both T24 and
   5637 cells of the si-CCT3 group (all p < 0.001, Fig. [107]5A and B).
   The results of western blotting revealed that the protein expression
   levels of FTH1, GPX4, and SLC7A11 in T24 and 5637 were significantly
   reduced by down-regulation of CCT3 (all p < 0.001, Fig. [108]5C).

Figure 5.

   [109]Figure 5
   [110]Open in a new tab

   Silencing CCT3 induced ferroptosis in BCa cells. (A) Content detection
   of malondialdehyde (MDA), glutathione (GSH), and superoxide dismutase
   (SOD) by corresponding Enzyme-linked immunosorbent assay (ELISA) kits.
   (B) The iron assay kits were applied to detect the ferrous iron (Fe^2+)
   levels. (C) The protein expression levels of ferritin heavy polypeptide
   1, recombinant solute carrier family 7, member 11 (SLC7A11), and
   glutathione peroxidase 4 (GPX4). ^***p < 0.001 vs. si-NC group.

Selection of the NOD-like receptor signaling pathway

   The DEGs between the si-NC and si-CCT3 groups were identified through
   transcriptome sequencing, and a heatmap displaying the cluster analysis
   of these DEGs was present in Fig. [111]6A. As evident from the volcano
   plot, cells with silenced CCT3 expression exhibited 77 up-regulated
   DEGs and 273 significantly down-regulated DEGs (Fig. [112]6B). The top
   10 up-regulated and down-regulated DEGs were listed in Supplementary
   Table 2. Then GO and KEGG enrichment analyses were conducted to
   calculate the distribution of these DEGs. The top 10 GO terms with the
   minimum p-value were shown in Fig. [113]6C, demonstrating that
   silencing of CCT3 mostly influenced the extracellular region,
   carboxylic acid transmembrane transporter activity, and positive
   regulation of cholesterol transport. Furthermore, the top 20 KEGG
   pathways with the minimum p-value were present im Fig. [114]6D, from
   which the NOD-like receptor signaling pathway was selected for further
   validation.

Figure 6.

   [115]Figure 6
   [116]Open in a new tab

   Selection of the nucleotide-binding and oligomerization domain
   (NOD)-like receptor signaling pathway. (A) The heatmap of cluster
   analysis of DEGs between the si-NC and si-CCT3 groups identified by
   transcriptome sequencing. (B) The volcano plot of DEGs between the
   si-NC and si-CCT3 groups. (C) The GO enrichment analysis and (D) The
   KEGG pathway analysis of DEGs between the si-NC and si-CCT3 groups.

Activating the NOD1-NF-κBsignaling pathway reverses the effect of CCT3
knockdown on BCa cells

   The NOD-like receptor pathway was validated using the NOD1 agonist
   M-TriDAP. Compared with the si-CCT3 group, the mRNA expression level of
   CCT3 in T24 was significantly elevated in the si-CCT3 + M-TriDAP group
   (p < 0.001, Fig. [117]7A). Based on CCK-8 assay results, we found that
   M-TriDAP promoted the cell viability of T24, which had been suppressed
   by CCT3 silencing(p < 0.05 or p < 0.001, Fig. [118]7B). The cells in
   the si-CCT3 + M-TriDAP group exhibited a lower apoptosis rate and
   higher proliferation, migration, and invasion rates compared to the
   si-CCT3 group (p < 0.001 or p < 0.01, Fig. [119]7C-[120]7F). In
   addition, compared to the si-CCT3 group, reduced levels of MDA and
   Fe^2+, alongside elevated levels of GSH, SOD, GPX4, SLC7A11, and NOD1,
   were noted in the si-CCT3 + M-TriDAP group (p < 0.001 or p < 0.01,
   Fig. [121]7G-[122]7I), indicating that NOD1 activation suppressed
   ferroptosis which had been induced by silencing CCT3. In addition, we
   detected the expression levels of IκBα and p-IκBα, two key proteins of
   the NF-κB signaling pathway mediated by NOD1. Compared to the si-NC
   group, the p-IκBα/IκBα ratio in the si-CCT3 group was significantly
   down-regulated (p < 0.001, Fig. [123]7I), consistent with NOD1
   expression level, which indicated the inhibition of the NOD1-NF-κB
   signaling pathway. However, upon M-TriDAP treatment, the p-IκBα/IκBα
   ratio was up-regulated (p < 0.001, Fig. [124]7I). Therefore, we
   suggested that silencing CCT3 promoted ferroptosis by inhibiting the
   NOD1-NF-κB signaling pathway, which suppressed cell growth, invasion,
   and migration and promoted apoptosis of BCa cells.

Figure 7.

   [125]Figure 7
   [126]Open in a new tab

   Activating the NOD1-nuclear factor kappa B (NF-κB) signaling pathway
   reversed the effect of CCT3 knockdown on BCa cells. (A) The mRNA
   relative expression level of CCT3 was determined by RT-qPCR. (B) The
   viability of treated cells was detected by cell counting kit-8 assay.
   (C) The apoptosis rate of treated cells was detected by a flow
   cytometer. (D) The proliferation ability of treated cells was detected
   by a 5-ethynyl-2-deoxyuridine staining proliferation kit. (E) The
   migration of treated cells was detected by wound healing assay. (F) The
   invasion of treated cells was measured by transwell assay. (G) Content
   detection of MDA, GSH, and SOD by corresponding ELISA kits. (H) The
   iron assay kits were applied to detect the Fe^2+ levels. (I)The protein
   expression levels of NOD1, NF-κB inhibitor alpha (IκBα), phospho-IκBα
   (p-IκBα), SLC7A11, and GPX4. ^***p < 0.001 vs. si-NC group;
   ^###p < 0.001, ^##p < 0.01, ^#p < 0.05 vs. si-CCT3 group.

Silencing CCT3 induces ferroptosis to suppress BCa progression in mice by
inhibiting the NOD1-NF-κB signaling pathway

   To further confirm the effect and regulatory mechanism of CCT3 in BCa,
   we established a subcutaneous xenograft tumor model. Compared to the
   LV-NC group, the LV-CCT3 group displayed an obvious increase in body
   weight, with a substantial decrease in tumor weight and volume
   (p < 0.001, Fig. [127]8A-[128]8C). Notably, these effects were reversed
   by M-TriDAP activating the NOD1-NF-κB signaling pathway (p < 0.05 or
   p < 0.001, Fig. [129]8A-[130]8C). RT-qPCR analysis showed that the mRNA
   level of CCT3 was reduced in the LV-CCT3 group compared to the LV-NC
   group, whereas it was upregulated in the LV-CCT3 + M-TriDAP group
   compared to the LV-CCT3 group (all p < 0.001, Fig. [131]8D). The images
   from H&E staining showed alleviation of inflammatory infiltration in
   the tumor tissue of the LV-CCT3 group, which was reversed in the
   LV-CCT3 + M-TriDAP group (Fig. [132]8E). Moreover, Ki67 staining
   demonstrated that the Ki67 positive rate decreased with the
   down-regulation of CCT3, but increased by M-TriDAP activating the
   NOD1-NF-κB signaling pathway (p < 0.01 or p < 0.05, Fig. [133]8F).
   Western blotting further confirmed the down-regulation of CCT3 protein
   in the LV-CCT3 group compared to the LV-NC group and its up-regulation
   in the LV-CCT3 + M-TriDAP group compared to the LV-CCT3 group (all
   p < 0.001, Fig. [134]8G). In addition, enhanced expression levels of
   MDA and Fe^2+ and reduced expression levels of GSH, SOD, GPX4, SLC7A11,
   NOD1, and p-IκBα/IκBα were found in mice after knockdown of CCT3 (all
   p < 0.001, Fig. [135]8H-[136]8J), which were all reversed in the
   LV-CCT3 + M-TriDAP group (p < 0.001 or p < 0.01 or p < 0.05,
   Fig. [137]8H-[138]8J).

Figure 8.

   [139]Figure 8
   [140]Open in a new tab

   Silencing CCT3 induced ferroptosis to suppress BCa progression in mice
   by inhibiting the NOD1-NF-κB pathway. (A) Body weight of mice. (B)
   Tumor weight. (C) The image of tumors from different groups. (D) The
   relative mRNA expression level of CCT3 was determined by RT-qPCR. (E)
   The results of hematoxylin-eosin staining (100×, scale bars = 200 μm).
   (F) The results of Ki67 staining (100×, scale bars = 200 μm). (G) The
   relative protein expression level of CCT3 was detected by western
   blotting. (H) Content detection of MDA, GSH, and SOD by corresponding
   ELISA kits. (I) The iron assay kits were applied to detect the Fe^2+
   levels. (J) The protein expression levels of NOD1, IκBα, p-IκBα,
   SLC7A11, and GPX4. ^***p < 0.001 vs. LV-NC group; ^###p < 0.001,
   ^##p < 0.01, ^#p < 0.05 vs. LV-CCT3 group.

Discussion

   BCa, especially muscle-invasive BCa, impairs the survival and quality
   of life of patients seriously^[141]21. Thus, it is necessary to develop
   a novel anti-BCa target. Recently, ferroptosis has emerged as a
   potential approach to oncotherapy^[142]22. In this study, we first
   obtained DEGs in BCa which were related to ferroptosis, and screened
   out 5 hub genes by bioinformatics analysis. Then, CCT3 was further
   selected for verification. The results demonstrated that silencing CCT3
   could suppress the progression of BCa by stimulating ferroptosis via
   inhibiting the NOD-like signaling pathway.

   Currently, bioinformatics methods exhibit myriad possibilities for the
   acquisition of potential therapeutic targets for many diseases,
   including BCa^[143]23, [144]24. In our study, a total of 5 hub genes
   were screened out by bioinformatics methods, namely CCT3, CDC25A, PRC1,
   DIAPH3, and KLF2. Fan et al. have reported that CDC25A overexpression
   is involved in the process of phospholipase C epsilon promoting
   tumorigenesis in BCa^[145]25. Additionally, PRC1 suppression also can
   reduce BCa cell viability and induce cell cycle arrest at the G2/M
   phase by inhibiting the cyclin-dependent kinases1/ Cyclin B1
   axis^[146]26. DIAPH3 has been reported as an inducer of tumor growth
   and cisplatin resistance in muscle-invasive BCa, linked to the NF-κB
   and epithelial-mesenchymal transition pathways^[147]27. Small nucleolar
   RNA host gene 1 improves the progression of BCa by inhibiting KLF2
   transcription^[148]28, thereby suggesting KLF2 as an antioncogene in
   BCa. Moreover, down-regulation of CCT3 has exhibited anti-tumor
   potential in increasing types of tumors, such as liver cancer^[149]29
   and hepatocellular carcinoma^[150]30. However, the role of CCT3 in BCa
   remains unclear now. Therefore, we selected CCT3 for further research
   and discovered that silencing CCT3 suppressed BCa progression both in
   vitro and in vivo, exhibiting the inhibitory effect of CCT3 knockdown
   in BCa.

   Accumulating evidence has suggested that inducing ferroptosis
   represents a potential approach to oncotherapy, accompanied by a lethal
   accumulation of lipid peroxides caused by iron^[151]22. Downregulation
   of phosphoglycerate dehydrogenase promotes ferroptosis, leading to the
   suppression of malignant progression in BCa^[152]31. Lung
   cancer-associated transcript 1 has emerged as a novel oncogenic player
   due to its ability to suppress ferroptosis in BCa^[153]32. Conversely,
   overexpression of fibronectin leucine-rich transmembrane protein 2
   suppresses tumor growth and metastasis by promoting lipid peroxidation
   and ferroptosis in BCa^[154]12. Liu et al. have reported arachidonate
   5-lipoxygenase deficiency facilitates BCa resistance to ferroptosis,
   thereby contributing to BCa progression^[155]33. Additionally, Poly C
   binding protein 1 has been found to facilitate BCa growth by inhibiting
   ferroptosis^[156]34. In this study, we found that silencing CCT3
   induced ferroptosis in BCa cells, suggesting that down-regulation of
   CCT3 inhibited BCa progression via the induction of ferroptosis.

   As an important member of the NOD-like receptors, NOD1 is involved in
   diverse aspects of tumor development, such as tumorigenesis,
   metastasis, chemoresistance, and stemness. For instance, Bacteroides
   fragilis enterotoxin-1 directly binds to NOD1, stabilizing its
   expression and thereby improving breast cancer cell stemness and
   chemoresistance^[157]35. Conversely, membrane-associated ring-CH-type
   finger 7 reduces NOD1 by promoting its ubiquitin-proteasome
   degradation, thereby inhibiting the stem-like capacities of bladder
   tumor cells^[158]36. Myeloid-intrinsic NOD1 sustains intra-tumor
   arginase-1 levels to promote the formation of tumor immunosuppressive
   microenvironment which contributes to tumorigenesis^[159]37. The NF-κB
   signaling pathway is one of the downstream effectors of NOD1.
   Activation of the NOD1 signaling facilitates the proliferation and
   invasion of ovarian cancer cells by promoting the NF-κB signaling
   pathway^[160]38. Suppression of the NOD1-NF-κB signaling pathway
   inhibits the proliferation and metastasis of triple-negative breast
   cancer^[161]39. Additionally, upregulation of NOD1 triggers the NF-κB
   signaling pathway, leading to the tumorigenicity and metastasis of
   human squamous cervical cancer^[162]40. In this study, CCT3 knockdown
   suppressed the expression levels of both NOD1 and p-IκBα/IκBα ratio.
   Notably, NOD1 agonist M-TriDAP could reverse the effect of CCT3
   suppression on tumor growth and ferroptosis in BCa, demonstrating that
   silencing CCT3 suppressed BCa by inhibiting the NOD1-NF-κB signaling
   pathway.

   In conclusion, CCT3 was identified as an oncogenic gene in BCa by
   bioinformatics methods, and CCT3 knockdown effectively repressed the
   progression of BCa by inducing ferroptosis and suppressing the
   NOD1-NF-κB signaling pathway. This discovery uncovered the anti-BCa
   potential of CCT3 suppression, suggesting CCT3 as a promising
   therapeutic target for BCa.

Supplementary Information

   [163]Supplementary Material 1^ (725.5KB, docx)
   [164]Supplementary Material 2^ (15.7KB, docx)
   [165]Supplementary Material 3^ (1.8MB, pdf)

Acknowledgements