Abstract

   The current work has further elucidated the expression and functional
   implication of cysteine desulfurase (NFS1) in gastric cancer (GC), the
   prognostic value, and therapeutic target because of the interaction
   with tumor immune infiltration and ferroptosis. Transcriptomic data
   from TCGA and GTEX were analyzed to assess mRNA expression and survival
   correlation with NFS1 among GC patients. A total of 152 GC cases were
   retrospectively analyzed. The level of NFS1 expression was upregulated
   in GC tissues compared to non-tumor gastric tissues, which was related
   to clinical characteristics and poor prognosis. Downregulation of the
   NFS1 protein in the GC cell line had an adverse effect on the
   migration, invasion, and proliferation of cells. In addition, NFS1 and
   immune correlation analysis showed that the level of NFS1 expression
   was related to a variety of immune cells and characteristics of the
   immune microenvironment. Based on functional enrichment analysis, NFS1
   may have a role in ferroptosis and the tumor microenvironment (TME),
   such as epithelial–mesenchymal transition control, and the stromal and
   immunologic responses. NFS1 is a potential diagnostic and prognostic
   biomarker linked to ferroptosis and the TME, and provides a novel
   target for cancer treatment and immunotherapy.

   Keywords: NFS1, ferroptosis, gastric cancer, immune infiltration

1. Introduction

   Gastric cancer (GC) is still one of the most common and lethal types of
   cancers worldwide [[37]1]. The burden from this disease is highest in
   China, accounting for approximately one-half of the new cases worldwide
   (total = 478,000) [[38]2,[39]3]. Most patients are diagnosed at a late
   stage and therefore miss the optimal surgical window for disease
   intervention [[40]4,[41]5]. Consequently, the detection of
   characteristic biomarkers at an early stage is essential for improving
   prognosis.

   Cysteine desulfurase (NFS1), a ferroptosis suppressor, is the product
   of a gene located on human chromosome 20q11.22, which helps catalyze
   the release of sulfur from cysteine to produce an iron–sulfur cluster
   (ISC) assembly [[42]6,[43]7]. NFS1 is involved in the biosynthesis of
   ISCs, which are essential in various cellular processes, including
   maintaining iron homeostasis. Previous research has demonstrated that
   the ISC enzyme, NFS1, functions as a ferroptosis suppressor involved in
   ferroptosis and tumor growth [[44]8,[45]9,[46]10,[47]11,[48]12].
   Targeting NFS1 to induce ferroptosis has been under investigation as a
   potential therapeutic strategy in several types of cancers [[49]13].

   Because NFS1 can inhibit ferroptosis, targeting NFS1 might enhance
   treatment efficiency by increasing cancer cell death. Recent studies
   have shown that NFS1 is highly expressed in GC tissues and is related
   to a poor prognosis [[50]14]. NFS1 inhibits ferroptosis in GC by
   upregulating the STAT3 signaling pathway. This inhibition helps GC
   cells survive and proliferate, making NFS1 a potential target for
   therapeutic strategies aimed at inducing ferroptosis to combat GC
   [[51]14]. Knockdown of NFS1 increases the sensitivity of colorectal
   cancer cells to the ferroptosis inducer, RSL3, and the chemotherapeutic
   drug, oxaliplatin, enhancing treatment outcomes [[52]15,[53]16].
   Previous research has highlighted that high expression of NFS1 is
   related to poor prognosis [[54]6,[55]17,[56]18,[57]19], which makes
   NFS1 a potentially good target for early diagnosis and therapeutic
   intervention in GC.

   Few studies have investigated NFS1 in GC patients. The current study
   was designed to determine the expression of NFS1 in GC tissues, observe
   the impact of NFS1 downregulation on GC cells, and analyze the
   relationship with clinical features and the prognostic value.

2. Materials and methods

2.1. NFS1 mRNA expression analysis

   Transcriptomic data, including mRNA and clinical information, were
   obtained from the Cancer Genome Atlas database (TCGA
   [[58]https://portal.gdc.cancer.gov]) and the Genotype-Tissue Expression
   database (GTEX [[59]https://gtexportal.org/home/]). The Kaplan–Meier
   Plotter ([60]https://kmplot.com/analysis/) was used to determine the
   relationship between NFS1 mRNA expression and overall survival (OS) in
   GC patients. mRNA sequencing data from level 3 in the HTSeq-TPM format
   were converted to the TPM format for further analysis. Statistical
   analysis and visualization were performed using R (version 4.2.1) and
   the ggplot2 package.

2.2. Tissue microarrays and immunohistochemical analysis

   A total of 152 (46 males and 106 females) GC and paired adjacent
   tissues were obtained from patients who received treatment in the
   Affiliated Hospital of Nantong University (Nantong, China) between 2005
   and 2010 to evaluate NFS1 expression in relation to clinicopathologic
   features. Patients were divided into two groups (high and low/no
   expression) based on the status of NFS1 expression. The following
   information was collected: gender, age, and tumor-node-metastasis (TNM)
   stage (tumor invasion [T], lymph node involvement [N], and metastasis
   [M]). Patients were grouped by age (<60 years [n = 37] and ≥60 years (n
   = 115). The TNM staging classified the patients into TNM stages I/II (n
   = 55) and III/IV (n = 97). Tumor invasion was as follows: Tis/T1/T2 (n
   = 40) and T3/T4 (n = 112). Lymph node involvement was categorized as N0
   (n = 35), N1a (n = 31), N1b (n = 37), and N2a/b (n = 49). Metastasis
   was classified as M0 (n = 118) and M1 (n = 34). None of the patients
   had preoperative radiotherapy, chemotherapy, or immunotherapy.

   Clinical data and follow-up information were retrieved from medical
   records. Tissue specimens of gastric tissues were paraffin-embedded,
   and tissue specimens were constructed. Immunohistochemical staining was
   performed to detect NFS1 expression. Tissue specimen chips were dewaxed
   and dehydrated, and treated with antigen retrieval using 10 mM sodium
   citrate buffer (pH 6.0) in a microwave. Inhibition of endogenous
   peroxidase activity was achieved using 3% H[2]O[2] for 10 min. Sections
   were incubated at 4°C overnight with primary antibodies to NFS1
   (MG815491S, 1:150 dilution; ABmart), followed by incubation with the
   secondary antibodies. The nuclei were stained with hematoxylin. NFS1
   expression was evaluated using a Vectra Automated Quantitative
   Pathology Imaging System (version 3.0; PerkinElmer, USA). The degree of
   staining was rated as follows: 0 (no staining), 1 (pale yellow), 2
   (brown-yellow), and 3 (tan). The percentage of positively stained cells
   was multiplied by the intensity score to provide a final
   immunohistochemistry score ranging from 0 to 300. Three biological and
   technical replicates were used for each assay.

   Informed consent: Informed consent has been obtained from all
   individuals included in this study.

   Ethical approval: The research related to human use has been complied
   with all the relevant national regulations, institutional policies and
   in accordance with the tenets of the Helsinki Declaration, and has been
   approved by the authors' institutional review board or equivalent
   committee and has been approved by the Ethics Committee of the
   Affiliated Hospital of Nantong University (Approval Number: 2022-L025).

2.3. Cell culture and NFS1 knockdown

   The HGC-27, AGS, MKN-1, and BGC-823 human GC cell lines and the normal
   gastric epithelial cell line, GES-1, were obtained from the Chinese
   Academy of Sciences (Shanghai, China). The cells were maintained in
   RPMI-1640 medium (Corning, Manassas, VA, USA), supplemented with 10%
   fetal bovine serum (FBS; Gibco, Grand Island, NY, USA), and incubated
   in an atmosphere of 5% CO[2] at 37°C.

   Cells were transfected with siRNAs targeting NFS1 for NFS1 knockdown
   (designed and synthesized by GenePharma). The siRNA sequences targeting
   NFS1 were as follows: si-NFS1-1 (5′-CCCTTACCTAATCAACTACTATG-3′) and
   si-NFS1-2 (5′-ATCCAACAACATAGCAATTAAGG-3′). GC cells were seeded at
   50–60% confluence in 6-well plates before transfection using
   Lipofectamine 3000 (Thermo Fisher Scientific), according to the
   manufacturer’s instructions. The cells were subsequently subjected to
   cell collection for use in western blot analysis, as well as invasion
   and migration assays.

2.4. Cell proliferation and colony formation assays

   The Cell Counting Kit-8 (CCK-8) was used to evaluate the number of
   proliferating viable cells. The transfected cells were seeded into
   96-well plates at 1 × 10^3 cells/well and incubated for the specified
   length of time. CCK-8 solution (Beyotime, Shanghai, China) was added to
   each well, and the cells were subsequently incubated at 37°C. Two hours
   after administering the CCK-8 reagent, every 24 h thereafter, cell
   vitality was evaluated at 450 nm (OD450). Transfected cells were seeded
   onto 6-well plates at a density of 400 cells/well in RPMI-1640 medium
   with 10% FBS for colony formation experiments. After 2 weeks, the cells
   were rinsed three times with phosphate-buffered saline, fixed in
   methanol, and stained with 0.1% crystal violet. Visible colonies were
   manually counted.

2.5. Migration and invasion assays

   Transwell inserts (8 μm pore size; Millipore, NY, USA) were used for
   the migration tests. The top chambers of 24-well plates were filled
   with 100 μl of the transfected cells, which had been resuspended in
   serum-free RPMI-1640 medium at a density of 2 × 10^5 cells/ml. The top
   chambers were pre-coated with diluted Matrigel (BD) for invasion
   experiments. The bottom chambers were filled with 500 μl/well of
   RPMI-1640 medium, which included 20% FBS. Cells that had moved or
   invaded the bottom surface of the membrane were fixed, stained, and
   photographed following incubation (16 h for invasion and 24 h for
   migration). Three randomly chosen fields of vision were used to count
   the cells. Fifty microliters of Matrigel (30 mg/well; BD Biosciences,
   San Jose, CA, USA) were used to pre-coat the Transwell membranes.

2.6. NFS1 functional enrichment analysis

   A network of protein–protein interactions (PPIs) was created based on
   STRING. The following Metascape analyses
   ([61]https://metascape.org/gp/index.html#/) were performed on
   co-expressed genes for gene ontology (GO) enrichment: cellular
   components (CCs), molecular functions (MFs), and biological pathways.
   The R packages, “ggplot2” and “clusterProfiler,” were utilized for the
   abovementioned analyses.

2.7. Gene set enrichment analysis (GSEA)

   Using the Kyoto Encyclopedia of Genes and Genomes (KEGG) gene sets (c2
   v7.5.1), Hallmark (v7.5.1), GO (c5 v7.5.1) BPs, CCs, MFs, and pathway
   enrichment analysis, an effort was made to demonstrate the role of NFS1
   expression in biological processes and signaling pathways in GC. The R
   package “clusterProfiler” was used to perform the statistical analysis
   and display the results.

2.8. Immune infiltration analysis

   The ssGSEA method in the GSVA package (v1.46.0) was used to assess
   Spearman correlations between NFS1 and immune cells for the TCGA
   datasets. The Tumor Immune Estimation Resource 2 portal (TIMER2
   [[62]https://timer.cistrome.org]) provided the purity-adjusted Spearman
   correlations between NFS1 and tumor-infiltrating immune cells.
   Correlation scatter plots were created to visualize results. The
   stromal, immune, and ESTIMATE scores were calculated using SangerBox3.0
   and the ESTIMATE package.

2.9. Statistical analysis

   X-tile software (v3.6.1) was used to determine the cut-off values for
   low and high NFS1 expression and statistical analysis of OS (Rimm Lab,
   Yale School of Medicine, New Haven, CT, USA). A χ ^2 test was used to
   analyze the correlations between clinical parameters and NFS1
   expression. Survival curves were plotted using the Kaplan–Meier
   technique and compared using the log-rank test. The Cox regression
   method was adopted in both univariate and multivariate analyses to
   estimate the prognostic value of NFS1 expression. Statistical analyses
   were performed using the SPSS software (v22.0; IBM SPSS Statistics, IBM
   Corporation, Armonk, NY, USA). All continuous data are presented as the
   mean ± SD or mean ± 95% CI. A P-value <0.05 was considered
   statistically significant.

3. Results

3.1. Identification of NFS1 mRNA expression and the diagnostic significance

   NFS1 mRNA expression was analyzed in various types of tumors and normal
   tissues using TCGA and GTEx. NFS1 expression was higher in most types
   of tumors, including stomach adenocarcinoma ([63]Figure 1a).
   Specifically, NFS1 mRNA was significantly higher in GC tissue samples
   compared to matched adjacent normal tissues ([64]Figure 1b and c; P <
   0.001). Furthermore, in support of these findings, upregulation of NFS1
   in GC was confirmed when analyzing the [65]GSE54129 dataset ([66]Figure
   1d; P < 0.001). Then, receiver operating characteristic (ROC) curve
   analysis was performed to further determine the diagnostic value of
   NFS1. The area under the ROC (AUC) was 0.882 ([67]Figure 1e; 95% CI =
   0.828–0.936), suggesting that NFS1 has the potential to serve as a
   diagnostic biomarker in GC.

Figure 1.

   [68]Figure 1
   [69]Open in a new tab

   NFS1 mRNA expression in tissues. (a) NFS1 mRNA expression in different
   tumor types was obtained from the UCSCXenaShiny database. (b) and (c)
   NFS1 mRNA expression in GC patients and matched adjacent normal samples
   in The Cancer Genome Atlas databases. (d) NFS1 expression in GC was
   determined by the GEO dataset. (e) ROC curve analysis for NFS1
   expression in GC patients. *P < 0.05.

3.2. Correlation of NFS1 protein expression with clinicopathologic features
in GC patients

   Immunohistochemical labeling was used on a tissue microarray that had
   152 pairs of GC and nearby normal tissues because post-transcriptional
   regulation of mRNA expression is not necessarily indicative of protein
   levels in vivo [[70]20]. Tumor tissues had substantially greater NFS1
   protein levels than peritumoral tissues based on mRNA expression
   patterns ([71]Figure 2a and b). The association between NFS1 expression
   and clinicopathologic features in patients with GC is shown in
   [72]Table 1. NFS1 expression and the TNM stage were significantly
   correlated (P = 0.002); patients with advanced stages (III and IV)
   exhibited greater NFS1 expression. Similarly, because greater
   expression is more common in patients without metastases (M0), there
   was a significant correlation between the M stage and NFS1 expression
   (P = 0.033). NFS1 expression, however, did not significantly correlate
   with age (P = 0.953), gender (P = 0.599), nerve/vascular invasion (P =
   0.104), T stage (P = 0.138), or N stage (P = 0.465). These findings
   implied that NFS1 expression is impacted by disease severity,
   especially cancer stage, although other clinical characteristics,
   including age and gender, do not appear to have a major impact on NFS1
   expression.

Figure 2.

   [73]Figure 2
   [74]Open in a new tab

   Histochemical analysis and survival data of NFS1 expression in GC
   tissues. (a) and (b) Histochemical staining results of NFS1 in GC
   tissues and para-cancerous tissues taken with original 40×
   magnification (bar = 50 μm) and 400× magnification (bar = 500 μm),
   respectively. (c) OS curves of NFS1 mRNA expression in GC. (d) High
   NFS1 protein expression was correlated with a poor prognosis in the
   clinical GC patient cohort (n = 152).

Table 1.

   Relationship between NFS1 expression and clinicopathological
   characteristics in GC patients
   Characteristic n Low and no expression High expression Pearson χ ^2 P
   Total 152
   Gender 0.0277 0.599
   Male 46 17(37.0) 29(63.0)
   Female 106 44(41.5) 62(58.5)
   Age (years) 0.003 0.953
   <60 37 15 (40.5) 22(59.5)
   ≥60 115 46(40.0) 69(60.0)
   TNM stage 9.451 0.002
   I + II 55 31(56.4) 24(43.6)
   III + IV 97 30(30.9) 67(69.1)
   T 2.2 0.138
   Tis + T1 + T2 40 20(50.0) 20(50.0)
   T3, 4b 112 41(36.6) 71(63.4)
   N 2.560 0.465
   N0 35 18(51.4) 17(48.6)
   N1a 31 11(35.5) 20(64.5)
   N1b 37 13(35.1) 24(64.9)
   N2a,b 49 19(38.8) 30(61.2)
   M 4.522 0.033
   M0 118 42(35.6) 76(64.4)
   M1 34 19(55.9) 15(44.1)
   Nerve/vascular invasion 4.521 0.104
   Positive 104 45(43.3) 59(56.7)
   Negative 27 12(44.4) 15(55.6)
   Unknown 21 4(19.0) 17(81.0)
   [75]Open in a new tab

   Tis: Carcinoma in situ.

3.3. Prognostic potential of NFS1 protein expression in GC

   NFS1 expression (HR = 5.546, 95% CI: 2.723–11.293; P < 0.001), T stage
   (HR = 2.142, 95% CI: 1.085–4.230; P = 0.028), N stage (HR = 1.329, 95%
   CI: 1.056–1.673; P = 0.015), and M stage (HR = 0.329, 95% CI:
   0.141–0.766; P = 0.01) were significantly associated with survival
   ([76]Table 2). Age (HR = 0.824; P = 0.499) and gender (HR = 0.931; P =
   0.797) were not significantly relevant to OS based on univariate
   analysis (P > 0.05; [77]Table 2).

Table 2.

   Univariate analysis of prognostic factors for OS in GC
   Factors                            Univariate analysis
                                    HR    P        95% CI
   NFS1 expression                  5.546 <0.001 2.723 11.293
   High vs low and no
   Age (years)                      0.824 0.499  0.469 1.445
   <60 vs ≥60
   Gender                           0.931 0.797  0.540 1.604
   Male vs female
   TNM stage                        4.391 <0.001 2.156 8.942
   I and II vs III and IV
   T: Tis + T1 + T2 vs T3 + T4      2.142 0.028  1.085 4.230
   N: N0 vs N1a vs N1b and N2a + 2b 1.329 0.015  1.056 1.673
   M: M0 vs M1                      0.329 0.01   0.141 0.766
   [78]Open in a new tab

   HR: hazard ratio; CI: confidence interval; Tis: carcinoma in situ.

   High NFS1 expression represented a powerful independent prognostic
   factor for poor OS in GC based on multivariate analysis (HR = 4.607,
   95% CI: 2.251–9.462; P < 0.001; [79]Table 3). Advanced TNM stage
   (III/IV) was significantly associated with an increased risk of
   mortality (HR = 3.463, 95% CI: 1.692–7.088; P < 0.001; [80]Table 3).
   The same finding was further verified by the Kaplan–Meier analysis, in
   which overexpression of NFS1 was significantly related to poor OS in GC
   patients ([81]Figure 2c and d).

Table 3.

   Multivariate analysis of prognostic factors for OS in GC
   Factors                            Multivariate analysis
                                    HR     P        95% CI
   NFS1 expression                  4.607  <0.001 2.251 9.462
   High vs low and no
   TNM stage                        3.463  <0.001 1.692 7.088
   I and II vs III and IV
   T: Tis + T1 + T2 vs T3 + T4
   N: N0 vs N1a vs N1b and N2a + 2b
   M: M0 vs M1                      0.3362 0.019  0.155 0.848
   [82]Open in a new tab

3.4. Upregulated NFS1 expression in GC cell lines

   NFS1 expression was assessed with a reverse transcription-polymerase
   chain reaction and western blot analysis in four human GC cell lines
   (HGC-27, AGS, MKN-1, and BGC-823) and one normal gastric epithelial
   cell line (GES-1) to determine the phenotypic consequence of NFS1 in
   GC. NFS1 was highly expressed in GC cell lines compared to GES-1
   ([83]Figure 3a–c). Specifically, HGC-27 and AGS had the highest NFS1
   levels but the NFS1 levels were low for MKN-1 and BGC-823. Thus, high
   expression NFS1 and the cell lines with high NFS1 expression (HGC-27
   and AGS) were selected. Construction of stable NFS1 knockdown was
   performed in cells to ascertain knockdown efficacy by qPCR and western
   blot analysis ([84]Figure 3d–i). The original western blot results are
   shown in the Supplemental file, WB-NFS1.

Figure 3.

   [85]Figure 3
   [86]Open in a new tab

   NFS1 expression in GC cell lines. (a) Relative NFS1 expression in GC
   cell lines (HGC-27, AGS, MKN-1, and BGC-823) compared to a normal
   gastric epithelial cell line (GES-1). (b) and (c) Western blot showing
   levels of NFS1 in GC cell lines with GAPDH as a loading control.
   (d)–(f) qRT-PCR and western blot analysis of knockdown and
   overexpression efficiency in HGC-27 cells. (g)–(i) qRT-PCR and western
   blot analysis of knockdown and overexpression efficiency in AGS cells.

3.5. NFS1 promoted cell proliferation, migration, and invasion in vitro

   Cell viability was measured using CCK-8 and colony-forming assays to
   determine the role of NFS1 in GC cell proliferation. siRNA-mediated
   knockdown of NFS1 significantly repressed the growth of HGC-27 and AGS
   cells compared to the negative control, as shown in [87]Figure 4a and b
   (P < 0.05). Similarly, colony formation was significantly reduced in
   such cells following NFS1 knockdown (P < 0.05; [88]Figure 4c and d).
   Then, we determined whether NFS1 had a potential role in GC metastasis
   by using the Transwell assay to assess cell migration and invasion. We
   demonstrated that NFS1 knockdown significantly reduced the migratory
   and invasive capabilities of GC cells ([89]Figure 4e–h). These results
   point to a critical role for NFS1 in GC tumorigenesis and progression.

Figure 4.

   [90]Figure 4
   [91]Open in a new tab

   NFS1 promotes proliferation and inhibits apoptosis of GC cells in
   vitro. (a) and (b) CCK-8 assays were used to determine the cell
   viability of GC cells transfected with si-RNA or si-NC with relative
   control. (c) and (d) Representative images of colony formation induced
   by si-NC and si-RNA cells. (e)–(h) Migration and invasion assays were
   used to investigate the vertical migration and invasion abilities with
   NFS1 knockdown in GC cells.

3.6. NFS1 expression related to TME score and immune cell infiltration

   The amount and proportion of tumor-infiltrating immune cells are major
   determinants in patient selection for immunotherapy [[92]21]. The
   ESTIMATE algorithm was used to calculate the immune, stromal, and
   estimated scores of GC patients to further elucidate the association
   between NFS1 expression and tumor microenvironment (TME). The above
   scores were negatively correlated with NFS1 expression, suggesting that
   tumor immune activity may be more powerful in patients with lower NFS1
   expression, as shown in [93]Figure 5a–c. Moreover, the ssGSEA algorithm
   was used to investigate the relationship between NFS1 and 24 types of
   immune cells in GC. NFS1 had a significant correlation with most types
   of immune cells and may be implicated in the modulation of immune cell
   infiltration into the TME ([94]Figure 5d).

Figure 5.

   [95]Figure 5
   [96]Open in a new tab

   Influence of NFS1 on infiltration of stromal and immune cells. (a)–(c)
   Relationship between NFS1 expression and immune score, estimated score,
   and stromal score in GC. (d) Differences of 24 immune cells between the
   NFS1 high- and low-expression groups. (e)–(g) NFS1 expression
   associated with 12 kinds of immune cells in TIMER2.0.

3.7. Correlation between NFS1 expression and tumor-infiltrating immune cells

   The TIMER2.0 database was analyzed to provide a deeper understanding of
   the relationship between the level of NFS1 expression and
   tumor-infiltrating immune cells, and the tumor purity had a significant
   positive correlation (r = 0.124; P = 1.55 × 10^−2). NFS1 was positively
   correlated with myeloid-derived suppressor cells (MDSCs) (r = 0.331; P
   = 3.64 × 10^−11; [97]Figure 5e), Th1 cells (r = 0.256; P = 4.33 ×
   10^−7; [98]Figure 5e), and Tfh cells (r = 0.209; P = 4.01 × 10^−5;
   [99]Figure 5f). Stromal cell infiltration, including cancer-associated
   fibroblasts (CAFs) and endothelial cells, had a negative correlation
   trend with NFS1 expression but without adequate statistical support
   (CAFs [r = −0.237; P = 3.06 × 10^−6; [100]Figure 5f] and endothelial
   cells [r = −0.158; P = 2.03 × 10^−3; [101]Figure 5e]). In contrast, the
   negative relevance of NFS1 expression was observed with immune cell
   types, such as CD8+ T cells (r = −0.19; P = 2.06 × 10^−4; [102]Figure
   5g), CD4+ T cells (r = −0.133; P = 9.33 × 10^−3; [103]Figure 5g), and
   myeloid dendritic cells (r = −0.102; P = 4.63 × 10^−2; [104]Figure 5g).
   These results suggested that NFS1 expression might influence the tumor
   immune microenvironment by altering the infiltrating density of immune
   and stromal cells.

3.8. Functional enrichment analysis of NFS1

   A functional enrichment analysis was performed to determine the
   potential functions and pathways related to NFS1 in GC. The STRING
   database was used to construct the PPI network, and 20 genes interacted
   with NFS1. The correlations of NFS1 with these genes are represented as
   a heat map in [105]Figure 6a and b. GO enrichment analysis showed that
   NFS1 took part in BP, like “ISC assembly,” “metal–sulfur cluster
   assembly,” “protein maturation by ISC transfer,” “ISC binding,” and
   “sulfur transferase activity.” The MFs enriched significantly included
   “iron ion binding.” At the CC level, high NFS1 expression was highly
   associated with “mitochondrial matrix” and “mitochondrial outer
   membrane translocase complex” ([106]Figure 6c). All these observations
   suggest that NFS1 primarily participates in ISC binding and assembly,
   mitochondrial function, and metabolic processes linked to ferroptosis.

Figure 6.

   [107]Figure 6
   [108]Open in a new tab

   Pathway enrichment analysis of NFS1. (a) PPI network of NFS1 with its
   20 interacting proteins. (b) A heatmap of the correlation between NFS1
   and the co-expressed genes. ***P < 0.001. (c) GO enrichment analysis of
   NFS1 and co-expressed genes. (d) Enrichment analysis of the NFS1 GSEA
   gene in the TCGA database. P < 0.05 was considered a meaningful
   pathway. Red and blue indicate immune- and ferroptosis-related
   metabolic pathways, respectively.

3.9. NFS1-related signaling pathways and the possible role in ferroptosis and
the TME

   We also conducted GSEA to explore the biological processes and
   signaling pathways associated with NFS1 expression. In hallmark gene
   sets, NFS1 expression was positively correlated with cell cycle-related
   pathways, such as “E2F targets,” “G2/M checkpoint,” and “mitotic
   spindle.” In contrast, NFS1 expression was negatively correlated with
   KRAS signaling, TGF-α signaling, EMT, and inflammatory responses. The
   identified DEGs related to NFS1 in GO and KEGG gene sets were involved
   in the pathway related to the cell cycle checkpoint, signaling that
   included sister chromatid segregation, cancer pathways, focal adhesion,
   immune responses, and B- and T-cell-mediated immunity. In addition, a
   significant positive correlation between NFS1 expression and metabolic
   processes involving ferroptosis was detected (steroid metabolism,
   steroid biosynthesis, and terpenoid backbone biosynthesis; [109]Figure
   6d). These enrichment results suggested that NFS1 exerts influence on
   the modulation of GC biological processes through involvement in
   ferroptosis and TME.

4. Discussion

   The current study demonstrated significantly upregulated NFS1
   expression in GC tissues compared to non-tumor stomach tissues.
   Downregulation of the NFS1 protein in the GC cell line had a
   suppressive effect on the migration, invasion, and proliferation of
   cells. In addition, it was observed that the level of NFS1 expression
   was related to a variety of immune cells and characteristics of the
   TME.

   In contrast to normal and benign tissues, the level of NFS1 expression
   in GC tissues was much higher (χ ^2 = 9.451; P = 0.002) and the
   proportion of patients with high expression (n = 67 [69.1%]) in
   patients with TNM stages III + IV was considerably greater than
   patients with low expression (n = 30 [30.9%]). These results suggested
   that NFS1 may have a role in tumor development [[110]22]. The
   biological behavior of the tumor likely changes as cancer progresses to
   a more advanced stage, leading to increased NFS1 expression, which
   might have a role in tumor migration, resistance, or proliferation
   [[111]14,[112]15,[113]16,[114]17,[115]18,[116]19]. The level of NFS1
   expression was inversely correlated with the tumor stage and OS in GC
   patients. Knockdown of NFS1 greatly decreased invasion, migration, and
   proliferation of the GC cell line according to in vitro experiments,
   which is consistent with previous results in other cancer types and
   indicates the role of NFS1 in carcinogenesis
   [[117]14,[118]15,[119]16,[120]17,[121]18,[122]19]. These results
   suggested that during the early and late stages of tumors, NFS1 might
   serve as an important marker for the development and progression of
   cancer.

   GO enrichment analysis suggested that NFS1 is involved in the
   biological processes of ISC binding and assembly, cysteine metabolism,
   and pathways of ferroptosis metabolism. GSEA based on NFS1 expression
   is related to the cell cycle, cancer, stroma, inflammatory responses,
   and ferroptosis metabolic pathways. These findings suggested that the
   potential pathways of NFS1 affecting GC biological processes could be
   related to ferroptosis and TME-related pathways
   [[123]6,[124]7,[125]15,[126]16,[127]17].

   NFS1 is the key participant in ISC biosynthesis, according to earlier
   research, and forms ISCs by binding to certain molecular conformations
   [[128]23]. ISCs have a role in energy conversion, protein translation,
   and DNA maintenance. When NFS1 is depleted, the iron starvation
   response is triggered, which causes ferroptosis [[129]6]. Unlike
   apoptosis, necroptosis, autophagy, pyroptosis, and necrosis, the
   process-named ferroptosis by Dixon et al. presented a new form of lipid
   peroxidation-driven regulatory iron-dependent cell death [[130]24].
   Ferroptosis has recently gained much attention. Either direct
   inhibition of lipid peroxidation or iron deprivation prevents
   iron-dependent phospholipid peroxidation, which is an imbalance between
   redox homeostasis and cell metabolism [[131]25]. Recent studies have
   shown that iron death is implicated in various physiologic processes of
   illness, including organ damage [[132]26], infectious diseases
   [[133]27], autoimmune disorders [[134]28], and tumorigenesis [[135]29].
   Tumor immune activity was higher in patients with low expression of
   NFS1 than in patients with high expression of NFS1, according to the
   association between NFS1 expression and immunologic, estimation, and
   stromal scores in GC. Therefore, tumor cells might develop strategies
   to inhibit immune function in patients with elevated NFS1 expression.
   This finding was consistent with previous studies, indicating that NFS1
   could be important for immune regulation in the TME [[136]30].

   By targeting the OPA3–NFS1 axis, suppression of iron death for
   protecting cardiac toxicity induced by dox [[137]31,[138]32] and
   depletion of NFS1 while targeting CAIX enhanced ferroptosis and
   significantly inhibited tumor growth [[139]12,[140]33]. Based on this
   study, differential expression of NFS1 in GC tissues may be associated
   with the ferroptosis pathway. The involvement of NFS1 in the iron
   metabolism pathway may modulate ISC synthesis, affecting cell
   sensitivity to ferroptosis. Further investigation is needed to give a
   more detailed explanation of the mechanism underlying NFS1-regulated
   ferroptosis in GC.

   The TME influences the growth of tumors and is made up of stromal
   cells, fibroblasts, endothelial cells, immune cells, and various
   chemokines. The tumor immunotherapy therapeutic response and patient
   prognosis are determined by the TME [[141]34]. Using the TIMER
   database, we showed that NFS1 expression had a positive correlation
   with Th1 cells, Tfh cells, MDSCs, and M0 macrophages but a negative
   correlation with infiltration of the majority of immune cells,
   including CD8+ T cells. Numerous investigations have shown that
   ferroptosis may be one of the ways tumor suppressors partially carry
   out tumor-suppressive activity. For example, CD8+ T lymphocytes
   suppress SLC7A11 in tumor cells and secrete IFN-γ to induce ferroptosis
   [[142]35]. Th1 cells produce IFN-γ, IL-2, and TNF-α, playing a major
   role in tumor immunity [[143]36]. In liver cancer, the combination of
   ferroptosis induction and MDSC blockade was shown to be beneficial for
   effective tumor suppression [[144]37,[145]38]. Additionally, targeting
   the selenium-GPX4-ferroptosis axis can regulate TFH cell homeostasis,
   enhancing humoral immunity [[146]38]. These results indicate that NFS1
   expression may be an important biomarker for prognosis and is closely
   related to clinical parameters in GC patients. These findings still
   need further clinical validation in patients with GC.

   Understanding how NFS1 modulates ferroptosis could lead to novel
   ferroptosis-based therapies, enhancing GC cell sensitivity to this
   process and improving treatment outcomes [[147]19,[148]39,[149]40].
   NFS1 has emerged as an important protein in different cancers and hence
   represents a promising target for therapeutic intervention. In lung
   cancer, depletion of NFS1 in combination with cysteine transport
   inhibition disrupts ISCs, inducing ferroptosis and impairing tumor
   growth, and may represent a potential therapeutic intervention by
   targeting NFS1 [[150]41]. High NFS1 expression is associated with poor
   prognosis, increased tumor aggressiveness, and resistance to
   ferroptosis in breast cancer, highlighting the role in tumor
   progression [[151]42]. Similarly, elevated NFS1 expression correlates
   with poor outcomes in prostate cancer, and targeting NFS1 may enhance
   the sensitivity of cancer cells to ferroptosis-based therapies
   [[152]18]. Knockdown of NFS1 increases sensitivity to the ferroptosis
   inducer, RSL3, and to the chemotherapy drug, oxaliplatin, which leads
   to better treatment outcomes [[153]43]. These findings showed the broad
   impact of NFS1 in various cancers. The current study revealed that NFS1
   knockdown suppresses GC cell proliferation, migration, and invasion,
   which is consistent with previous studies
   [[154]18,[155]19,[156]40,[157]41,[158]42,[159]43]. A significant
   association was demonstrated between NFS1 expression and TNM and M
   stages in this study, with higher expression in advanced stages (III
   and IV) and patients without metastases. These findings suggested that
   a relationship between NFS1 and tumor progression is required to
   determine the possible effects of NFS1 in GC patients. The lack of
   significant correlation with other factors, such as gender, age, and T
   or N stages, implies that the NFS1 role in the GC population warrants
   additional study.

   The association between NFS1 and the TME was evaluated using the
   ESTIMATE, stromal, and immune scores to determine the role of NFS1 in
   the immune microenvironment of GC. The results indicated that NFS1 is
   negatively correlated with the ESTIMATE, stromal, and immune scores,
   suggesting that the tumor immune activity of patients with low NFS1
   expression was stronger than patients with high NFS1 expression. The
   ssGSEA algorithm was used to study the relationship between the high
   and low NFS1 expression groups in the TCGA database and 24 immune cell
   subsets in GC. More B cells, CD8+ T cells, cytotoxic cells, dendritic
   cells, eosinophils, immature dendritic cells, macrophages, mast cells,
   neutrophils, natural killer cells, plasmacytoid dendritic cells, T
   cells, effector memory T cells, Tfh cells, Th1 cells, and Th2 cells
   were present in the high NFS1 expression group than in the low NFS1
   expression group. The above results indicate that NFS1 may be involved
   in the TME and affect tumor growth by influencing the infiltration of
   immune cells. NFS1 appeared to regulate both immune and stromal cell
   infiltration within TME, with high NFS1 expression associated with
   lower immune, stromal, and TME scores, suggesting suppressed immune
   infiltration and poor tumor prognosis. Then, the relationship between
   NFS1 expression and immune cell infiltration was studied using the
   TIMER2.0 database. NFS1 was positively correlated with tumor purity (r
   = 0.124; P = 1.55 × 10^−2) and negatively correlated with CD8+ T cells
   (r = −0.19; P = 2.06 × 10^−4), CD4+ T cells (r = −0.133; P = 9.33 ×
   10^−3), myeloid dendritic cells (r = −0.102; P = 4.63 × 10^−2), and
   activated NK cells (r = −0.238; P = 2.79 × 10^−6). The results
   indicated that NFS1 was positively correlated with MDSC (r = 0.331; P =
   3.64 × 10^−11), Th1 cells (r = 0.256; P = 4.33 × 10^−7), and Tfh cells
   (r = 0.209; P = 4.01 × 10^−5). The infiltration of stromal cells was
   evaluated from the TIMER portal, mainly including CAF and endothelial
   cells. NFS1 was negatively correlated with CAFs (r = −0.237; P = 3.06 ×
   10^−6) and endothelial cells (r = −0.158; P = 2.03 × 10^−3) in GC
   ([160]Figure 5e–g). Based on the above analysis, the level of NFS1
   expression positively correlated with Th1 and Tfh cells, and negatively
   correlated with CD8+ T cells, mast cells, Th17 cells, B cells, and
   pDCd. These findings indicated that anti-tumor immunity may have a role
   in NFS1-mediated carcinogenesis in GC and positioned NFS1 as a
   potential target of immunotherapy.

   The present study provided valuable insight into the relationship
   between NFS1, immune cell infiltration, and prognosis. However, the
   present study had some limitations. First, the role of NFS1 in
   ferroptosis was only inferred from bioinformatics analysis, and the
   investigation of the underlying mechanisms may be insufficient. Second,
   immune cell profiling (ssGSEA and TIMER2.0) showed a correlation, but
   the biological implications remain vague. Further experiments, such as
   BODIPY-C11 staining, lipid ROS, iron accumulation, and GPX4 levels
   through in vitro ferroptosis assays, are required to illustrate how
   NFS1 regulates immune and stromal cell infiltration. Third, the current
   study had a limited sample of GC tissues from a retrospective review of
   patients. Adding more samples and diversity in further research could
   affect generalization and the reliability of results, and increasing
   both would strengthen the power of the study. Testing cytokine profiles
   or immune checkpoint molecules in vitro or from public datasets can be
   considered in the next step. Fourth, ROC curves and the AUC lacked
   comparison with standard GC biomarkers. Multivariate ROC curves or
   combination marker analysis highlighted the diagnostic potential.
   Finally, clinical validation was lacking in this study, which limits
   the applicability of the results. Clinical trials to verify the effects
   in actual patients would give further credence to the conclusions of
   this research.

5. Conclusion

   For the first time, the current study showed that NFS1 may function as
   a possible biomarker for diagnosis and prognosis associated with
   ferroptosis and TME. NFS1 may also offer a novel target for
   immunotherapy and medication therapy based on ferroptosis in GC. There
   were certain restrictions. The direct mechanisms by which NFS1
   influences cancer development and prognosis by controlling ferroptosis
   and TME need to be further studied using more fundamental studies with
   additional clinical samples.

Footnotes

   Funding information: Authors state no funding involved.

   Author contributions: Zhiyun Mao: guarantor of integrity of the entire
   study, definition of intellectual content, data acquisition,
   statistical analysis, manuscript preparation, and manuscript editing.
   Zhongmei Shi: study design, data analysis, statistical analysis, and
   manuscript preparation; Ming Cui: literature research, data analysis,
   and manuscript editing. Xiaohong Ma: data acquisition. Yan Wan:
   literature research. Rongrong Jing: guarantor of integrity of the
   entire study, study design, and manuscript review. Jingchun Wang: study
   concepts, definition of intellectual content, and manuscript review.
   All authors approved the final version of the manuscript.

   Conflict of interest: Authors state no conflict of interest.

   Data availability statement: The datasets generated during and/or
   analyzed during the current study are available from the corresponding
   author on reasonable request.

References