Abstract
SLC25A13, a pivotal component of the mitochondrial aspartate-glutamate
carrier, is integral to cellular metabolism and has been linked to
various diseases. However, its role in cancer biology remains largely
unexplored. In this study, we employed multi-omics data to elucidate
the genetic landscape, expression profile, and prognostic value of
SLC25A13 in a pan-cancer context. Additionally, we examined the
correlation between SLC25A13 and the immune microenvironment across
various cancers. By applying multiple machine learning methods, we
identified seven core SLC25A13 co-expressed genes and developed a
nomogram to predict the prognosis of glioma patients, validating its
efficacy across multiple independent datasets. Furthermore, in vitro
and in vivo experiments demonstrated that SLC25A13 is significantly
overexpressed in glioblastoma tissues compared to paraneoplastic
tissues, promoting glioblastoma cell proliferation and migration while
inhibiting apoptosis. Collectively, our study positions SLC25A13 as a
promising biomarker for cancer prognosis and a potential therapeutic
target, particularly in glioma, thereby laying the groundwork for
future research into its therapeutic exploitation in cancer.
Supplementary Information
The online version contains supplementary material available at
10.1186/s12935-025-03696-z.
Keywords: SLC25A13, Pan-cancer, Glioma, Multi-omics, Tumor
microenvironment
Introduction
Cancer is one of the leading causes of human mortality worldwide
[[36]1]. As a heterogeneous group of diseases, different cancer types
exhibit distinct pathogenetic mechanisms, and even within the same
cancer type, manifestations can vary significantly [[37]2, [38]3].
Therefore, integrating multiple cancer types in research is essential
for understanding both the heterogeneity and commonality among cancers,
which can also facilitate the discovery of new therapeutic targets.
Recent advances in sequencing and bioinformatics technologies have
enabled pan-cancer studies that integrate data from multiple cancer
types, yielding numerous new insights into the mechanisms and
characteristics of cancer [[39]4, [40]5]. Cancer metabolism has emerged
as a key focal point in oncological research due to the discovery that
metabolic processes within tumor cells differ fundamentally from those
in normal cells [[41]6]. These metabolic alterations not only
facilitate rapid tumor growth but also present potential avenues for
therapeutic intervention. Among the regulators of metabolic pathways,
the solute carrier (SLC) superfamily plays a critical role by
transporting essential metabolites across cellular membranes [[42]7].
Research on solute carrier proteins has become a hotspot for
identifying new cancer therapeutic targets [[43]8, [44]9].
The solute carrier family 25 (SLC25) comprises 53 members and is the
largest solute transporter family in humans [[45]10–[46]12]. It is
crucial for solute transport across the inner mitochondrial membrane,
directly or indirectly influencing the metabolic processes of fats,
sugars, and amino acids, thereby playing a significant role in cellular
metabolic regulation [[47]10–[48]12]. SLC25A13, a member of the SLC25
family and also known as calcium-regulated mitochondrial
aspartate/glutamate carrier 2, is primarily responsible for the
transport of glutamine and aspartate [[49]11]. It has been implicated
in various diseases, including intrahepatic cholestasis [[50]13],
hepatocellular carcinoma [[51]14], melanoma [[52]15], and citrin
deficiency [[53]16]. While SLC25A13 has been reported to be involved in
the metabolic reprogramming of tumor cells and associated with rapid
tumor cell proliferation [[54]17], its characterization across
different cancers remains unclear, limiting its potential as a
therapeutic target. Therefore, integrated pan-cancer studies targeting
SLC25A13 could help investigate the heterogeneity of SLC25A13 across
different cancer types and provide guidance for therapeutic
applications.
In this study, we conducted a detailed analysis of the genetic
landscape, expression patterns, and prognostic value of SLC25A13 across
multiple tumors by integrating multidimensional pan-cancer data.
Additionally, we explored the cancer-promoting role of SLC25A13 in
glioma and confirmed through in vitro and in vivo experiments that
SLC25A13 promotes the malignant behavior of glioblastoma (GBM) cells,
suggesting its potential as a therapeutic target in glioma.
Materials and methods
Genetic landscape and expression pattern of SLC25A13 in pan-cancer
Methylation data and copy number variation data for SLC25A13 were
obtained from TCGA database
([55]https://www.cancer.gov/ccg/research/genome-sequencing/tcga). The
relationship between gene expression and methylation was assessed using
Spearman correlation analysis of beta values and
[MATH: log2(FPKM+1) :MATH]
. The mutation profile of SLC25A13 across various cancer types was
analyzed using cBioPortal ([56]https://www.cbioportal.org/). Gene
expression data from The Cancer Genome Atlas (TCGA) pan-cancer,
Therapeutically Applicable Research to Generate Effective Treatments
(TARGET), and Genotype-Tissue Expression (GTEx) projects were obtained
from the UCSC Xena database ([57]https://xenabrowser.net/). These
datasets had undergone batch effect removal and normalization.
Quantitative protein expression data for pan-cancer studies were
acquired from the Clinical Proteomic Tumor Analysis Consortium (CPTAC)
database ([58]https://proteomics.cancer.gov/programs/cptac).
Analysis of the potential function of SLC25A13 in pan-cancer and its
relationship with the immune microenvironment
The correlation between SLC25A13 and tumor mutational burden (TMB) and
microsatellite instability (MSI) across various cancers was analyzed
using the TCGAplot R package [[59]18]. The immune microenvironment
composition in pan-cancer was assessed with the CIBERSORT algorithm,
while the ESTIMATE algorithm was employed to calculate the IMMUNE
SCORE, STROMAL SCORE, and ESTIMATE SCORE. Spearman correlation
coefficients were then calculated between SLC25A13 expression and these
scores. Additionally, the protein interaction network of SLC25A13 was
constructed using the ComPPI database [[60]19].
Analysis of the potential function and prognostic value of SLC25A13 in glioma
Sequencing data and corresponding clinical information for glioma
patients were obtained from the TCGA and CGGA
([61]http://www.cgga.org.cn/) databases. The Rembrandt dataset was
acquired from the Gene Expression Omnibus (GEO) database
([62]https://www.ncbi.nlm.nih.gov/geo/). Quantitative protein data for
key pathways were sourced from the TCPA database
([63]https://www.tcpaportal.org/tcpa/). Glioma patients were
categorized into high and low SLC25A13 expression groups based on
median expression levels, and differentially expressed genes were
identified. Gene Set Enrichment Analysis (GSEA) was performed using the
R package clusterProfiler [[64]20]. Fourteen tumor functional states
defined by CancerSEA [[65]21] were assessed using the R package GSVA
[[66]22], and their correlations with SLC25A13 expression were
calculated. Patients were further categorized into quartiles (Q1, Q2,
Q3, and Q4) based on SLC25A13 expression levels, with Q1 representing
the top 25% and Q4 representing the bottom 25% of expression. Immune
response and genomic status scores for each group were calculated based
on previously reported definitions [[67]23]. Meta-analysis of the
results from univariate Cox regression analysis was performed using the
inverse variance method, with the log hazard ratio (HR) as the primary
measure.
Establishment and evaluation of the nomogram
Through the application of multiple machine learning techniques, seven
core genes co-expressed with SLC25A13 were identified. Subsequently, a
multivariate Cox analysis model was developed, from which we derived a
risk score formula based on the product of gene expression levels and
their respective regression coefficients (Risk score
[MATH:
=∑n=1
7expn×coefn
:MATH]
). By integrating this risk score with additional clinicopathological
parameters, we constructed a nomogram to predict the prognosis of
glioma patients. The nomogram's reliability was validated across
multiple datasets. For the analysis of survival and prognostic factors,
we utilized the R packages survival and survminer, while the R package
regplot was employed to construct the nomogram.
Analysis of single-cell and spatial transcriptome data
The expression of SLC25A13 in the glioma single-cell sequencing
datasets was analyzed using data derived from the TISH2 database
[[68]24]. Single-cell sequencing data [69]GSE182109, downloaded from
the GEO database, was utilized for further analysis. Standard analysis
was performed using the R package Seurat [[70]25], and batch effects
between samples were removed with the Harmony algorithm [[71]26]. Copy
number variation (CNV) was inferred using the R package inferCNV
[[72]27] to differentiate between malignant and normal cells. Risk
score for each cell were calculated based on previously established
formula, and cellular communication was analyzed using the R package
CellChat [[73]28]. Metabolic pathway activities of low-risk and
high-risk malignant cells were inferred using the R package
scMetabolism [[74]29]. Spatial transcriptome data were obtained from
10 × Genomics ([75]https://www.10xgenomics.com/), analyzed with Seurat,
and normalized using the sctransform algorithm. The spatial
distribution of cell subpopulations was inferred through the inverse
convolution algorithm, categorizing regions with high malignant cell
content as mixed regions and those with high stromal cell content as
normal regions.
Cell culture
U251 and U87 cell lines were cultured in standard DMEM medium
(Sigma-Aldrich, St. Louis, MO, USA) supplemented with 10% fetal bovine
serum (FBS), 1% penicillin–streptomycin, and 1% glutamine (Gibco). The
cells were maintained at 37 °C in a humidified tissue culture incubator
with 5% CO[2].
Plasmid construction and cell transfection
The SLC25A13 shRNA-Luc sequence was cloned into the pLKO.1-shRNA-Puro
vector. Lentiviral infection was packaged using Lipofectamine 3000
(Invitrogen, USA) with psPAX2 and pMD2.G to transfect 293 T cells with
U87 and U251 cells according to the manufacturer's protocol. Puromycin
(Sigma-Aldrich) was also used for stable transfection.
EdU cell proliferation assay
Cells were cultured in fresh medium containing 10 μM EdU for 3 h,
followed by trypsinization to collect the cells. The cells were then
fixed with 4% formaldehyde in PBS. After permeabilization with 0.3%
Triton X-100, 500 μL of Click-iT^® reaction cocktail (ThermoFisher,
Waltham, MA, USA) was added to each well according to the
manufacturer's instructions. Detection was performed using a flow
cytometer. The average fluorescence intensity of the EdU^+ population
was then analyzed using FlowJo software (version 10, TreeStar). All
experiments were conducted with three biological replicates and
subjected to statistical analysis.
Cell viability assay
The cells were seeded in 96-well plates at a density of 3,000 cells per
well. Afterward, 100 μL of Cell Count Kit-8 (APEBIO, Houston, USA)
reagent was added to each well at 0, 24, 48, and 72 h, followed by
incubation at 37 °C for 3 h. Absorbance was measured at 450 nm using a
Tecan flatbed instrument. Dose–response curves were generated using
GraphPad Prism, and cell viability was calculated. All experiments were
conducted with three biological replicates and subjected to statistical
analysis.
Assessment of cell cycle and apoptosis
To evaluate the cell cycle, cells were stained with 50 μg/mL propidium
iodide (PI) containing 20 μg/mL RNase without DNase. The cells were
then analyzed using a flow cytometer according to the manufacturer's
instructions. Based on DNA content, the G1 phase, S phase (DNA
synthesis phase), G2 phase, and M phase (mitosis phase) were
determined, and the percentage of cells in each phase was measured. All
cellular experiments were conducted in biological triplicates using
cryopreserved cells that were thawed independently at different time
points. Additionally, cells were double-stained with fluorescein
isothiocyanate-conjugated Annexin V (Annexin V-APC) and PI (BD
Pharmingen, San Diego, USA), followed by analysis using a flow
cytometer.
Cell migration assay
To assess cell migration, a scratch assay was performed. A sterile
pipette tip was used to create a scratch in a confluent monolayer of
glioma cells. Images of the scratch were captured at 0, 24, and 36 h
post-scratch to evaluate the migration of cells into the wound area.
All experiments were conducted with three biological replicates and
subjected to statistical analysis.
Colony formation assay
The colony formation assay was performed 48 h after transfection with
SLC25A13 shRNA. Cells were counted and seeded into six-well plates at a
density of 1 × 10^4 cells per well. The cells were then incubated in a
humidified incubator at 37 °C. After an appropriate incubation period,
the cells were fixed with 4% paraformaldehyde for 10 min. Subsequently,
the colonies were stained with 0.5% crystal violet. Finally, the number
of macroscopic colonies was counted under an optical microscope. All
experiments were conducted with three biological replicates and
subjected to statistical analysis.
Western blot assay
Afterward, the harvested cells were lysed using RIPA buffer (50 mM
Tris–HCl, pH 7.5; 150 mM NaCl; 0.1% sodium deoxycholate; 0.1% SDS; 1 mM
EDTA, pH 8.0; 1% NP-40) containing a protease and phosphatase inhibitor
cocktail (Thermo Fisher) and incubated on ice for 10 min. The cell
lysates were then centrifuged at 4 °C using a tabletop centrifuge to
remove cellular debris. Equal amounts of protein extracted from
different samples were loaded onto a sodium dodecyl
sulfate–polyacrylamide gel electrophoresis (SDS-PAGE). Following
electrophoresis, the proteins were transferred onto a polyvinylidene
difluoride (PVDF) membrane. After blocking with 5% bovine serum
albumin, the membrane was incubated with primary antibodies and
horseradish peroxidase-conjugated secondary antibodies (Proteintech,
USA). Detection was performed using the Pierce™ Enhanced
Chemiluminescence Substrate Kit (Thermo Fisher Scientific, Waltham, MA,
USA) and the ChemiDoc™ Touch Imaging System (Bio-Rad Laboratories,
Hercules, CA, USA).
Immunohistochemical assay
Clinical glioma tissue specimens were collected at the Neurosurgery
Department of Wuhan Union Hospital. Ethical approval was obtained from
the Wuhan Union Hospital. All patients provided written informed
consent and did not receive any financial compensation. For
immunohistochemistry, 6 μm formalin-fixed paraffin-embedded sections
were incubated with an anti-SLC25A13 antibody (1:400, CST). A secondary
HRP-conjugated antibody was then applied, followed by visualization
using diaminobenzidine (DAB). Pathological assessment was conducted in
a blinded manner. Protein expression levels were quantified using a
visual grading system based on the extent (percentage of positive tumor
cells) and intensity of staining. Under a microscope, protein
expression was evaluated by scoring the staining intensity and extent.
Staining intensity was rated on a scale of 0–3: 0 = no staining,
1 = weak staining, 2 = moderate staining, and 3 = strong staining.
Additionally, the extent of staining was scored based on the percentage
of positive-stained cells (0 = 0–5%, 1 = 6–25%, 2 = 26–50%, 3 = 51–75%,
4 = 76–100%). The final score for protein expression was obtained by
multiplying the intensity and extent scores.
Xenograft model
Animal experiments were performed per the NIH Guidelines for the Care
and Use of Laboratory Animals and approved by the Animal Care Committee
of Tongji Medical College. Five-week-old female BALB/c nude mice were
used for all xenograft experiments (10 per group). The animals were
weighed and randomly divided into two groups, and 5 μL U87-Luc glioma
cell suspension (3 ×
[MATH: 105 :MATH]
cells) was injected into the mouse brain using a stereotaxic apparatus
at 2 mm lateral and 2 mm anterior to the bregma and at a 3.5 mm depth.
Statistical analysis
Statistical analyses were performed using R software (version 4.4.0)
and GraphPad Prism (version 9.5). The Wilcoxon rank-sum test or
Student’s t-test was employed to assess differences between two groups.
Correlations between variables were evaluated using Spearman's rank
correlation coefficient. Survival analyses were conducted using
Kaplan–Meier curves, with differences assessed by the log-rank test. To
account for multiple comparisons, p-values were adjusted using the
Benjamini–Hochberg method. Statistical significance was set at p < 0.05
for all analyses.
Results
Abnormal expression and genetic alterations of SLC25A13 in multiple cancers
By comparing pan-cancer data extracted from the TCGA TARGET, and GTEx
projects, we identified abnormal expression of SLC25A13 in human
cancers. The analysis revealed that SLC25A13 was significantly
upregulated in multiple tumors, including adrenocortical carcinoma
(ACC), acute lymphoblastic leukemia (ALL), bladder urothelial carcinoma
(BLCA), colon adenocarcinoma (COAD), esophageal carcinoma (ESCA),
glioblastoma multiforme (GBM), head and neck squamous cell carcinoma
(HNSC), acute myeloid leukemia (LAML), lower grade glioma (LGG), lung
adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), pancreatic
adenocarcinoma (PAAD), skin cutaneous melanoma (SKCM), stomach
adenocarcinoma (STAD), stomach and esophageal carcinoma (STES),
testicular germ cell tumors (TGCT), uterine corpus endometrial
carcinoma (UCEC), and uterine carcinosarcoma (UCS). Conversely, lower
expression levels were observed in cholangiocarcinoma (CHOL), kidney
chromophobe (KICH), pan-kidney cohort (KIPAN), kidney renal clear cell
carcinoma (KIRC), and thyroid carcinoma (THCA) (Fig. [76]1A). Using the
CPTAC database, we then analyzed the protein expression of SLC25A13 in
various tumor types. Our findings showed that SLC25A13 protein levels
were significantly elevated in GBM, head and neck squamous cell
carcinoma (HNSCC), lung squamous cell carcinoma (LSCC), and LUAD, while
significantly decreased in clear cell renal cell carcinoma (CCRCC) and
hepatocellular carcinoma (HCC) (Fig. [77]1B). Further analysis of the
relationship between SLC25A13 expression levels and copy number
variations (CNVs) revealed a general upward trend in SLC25A13
expression as copy number increased from homozygous deletion to high
copy number amplification (Fig. [78]1C). Additionally, in cancers such
as cervical squamous cell carcinoma and endocervical adenocarcinoma
(CESC), thymoma (THYM), SKCM, and LGG, a negative correlation was
observed between SLC25A13 expression and promoter methylation levels
(Fig. [79]1D). Examining the genetic landscape of SLC25A13 across
various cancers, we discovered that the primary genetic alteration in
SLC25A13 was amplification (Fig. [80]1E and S1A). These findings
suggest that SLC25A13 may play a significant role in the pathogenesis
of various cancers, influenced by copy number variations, promoter
methylation, and genetic alterations.
Fig. 1.
[81]Fig. 1
[82]Open in a new tab
Abnormal expression and genetic alterations of SLC25A13 in various
cancers. Statistical significance is denoted as follows: ns (not
significant, p > 0.05), *(p < 0.05), **(p < 0.01), ***(p < 0.001), and
**** (p < 0.0001). A SLC25A13 mRNA expression levels in multiple cancer
types compared to normal tissues based on TCGA, TARGET, and GTEx
datasets. Differences between the two groups were assessed using the
Wilcoxon rank-sum test. B Protein expression levels of SLC25A13 across
different cancer types from the CPTAC dataset. Differences between the
two groups were assessed using the Wilcoxon rank-sum test. C
Association between SLC25A13 expression levels and copy number
variations across various cancers. Differences among multiple groups
were assessed using the Kruskal–Wallis H test. D Correlation between
SLC25A13 expression and promoter methylation levels in multiple
cancers. E Genetic alteration of SLC25A13 in various cancers
Association of SLC25A13 with TMB, MSI, and immune profiles in multiple
cancers
Understanding the correlation between gene expression and oncological
markers such as Tumor Mutational Burden (TMB) and Microsatellite
Instability (MSI) is crucial for elucidating the biological behavior of
tumors, predicting therapeutic responses, and devising personalized
treatment plans [[83]30, [84]31]. Our study highlights the significant
positive correlation between SLC25A13 expression and both TMB and MSI
across various cancers, indicating its potential role in tumor
progression and response to therapy (Fig. [85]2A and B). To evaluate
the prognostic significance of SLC25A13 expression in different cancer
types, we stratified patients into high- and low-expression groups
based on the median SLC25A13 expression level. Subsequent univariate
Cox regression analysis revealed that elevated SLC25A13 expression was
associated with poor prognosis in CESC, KICH, LGG, SKCM, and uveal
melanoma (UVM), while it served as a favorable prognostic indicator in
KIRC, LAML, and liver hepatocellular carcinoma (LIHC) (Fig. [86]2C).
Fig. 2.
[87]Fig. 2
[88]Open in a new tab
Correlation of SLC25A13 expression with tumor mutational burden (TMB),
microsatellite Instability (MSI), and immune profiles in diverse
cancers. Statistical significance is denoted as follows: ns (not
significant, p > 0.05), *(p < 0.05), **(p < 0.01), ***(p < 0.001), and
**** (p < 0.0001). A Spearman correlation between SLC25A13 expression
and TMB across multiple cancer types. B Spearman correlation between
SLC25A13 expression and MSI across various cancer types. C Univariate
Cox survival analysis showing hazard ratios for overall survival
associated with SLC25A13 expression in different cancer types. D
Proportion of patients with high and low SLC25A13 expression,
stratified by median SLC25A13 expression levels, across different
immune subtypes as classified by The Immune Landscape of Cancer. E
Spearman correlation between SLC25A13 expression and different immune
cell infiltration fractions calculated by the CIBERSORT algorithm in
multiple cancers. F Spearman correlation between SLC25A13 expression
and various immune suppressive molecules. G Potential protein–protein
interactions involving SLC25A13 identified from the ComPPI database. H
Spearman correlation between SLC25A13 expression and scores calculated
by the ESTIMATE algorithm in multiple cancers
Building on the framework of previous research [[89]23] that identifies
six immune subtypes—wound healing, IFN-γ dominant, inflammatory,
lymphocyte-depleted, immunologically quiet, and TGF-β dominant—our
analysis reveals that high SLC25A13 expression is predominantly
associated with immune subtypes C1 and C2, while low expression
correlates with subtypes C3 and C5 (Fig. [90]2D). Immunoinfiltration
analysis further indicated that SLC25A13 expression correlates with the
infiltration of immune cells across multiple cancer types
(Fig. [91]2E). Additionally, our study identifies significant
associations between SLC25A13 and various immunomodulatory molecules
(Fig. [92]2F and S2A-B). Our analysis reveals both consistent patterns
and tumor-specific heterogeneity in the correlation between SLC25A13
expression and immune parameters. Notably, across most tumor types,
SLC25A13 expression shows a negative correlation with T cell
infiltration, particularly CD8 + T cells. This inverse relationship
suggests that high SLC25A13 expression may be associated with reduced T
cell infiltration and impaired immune-mediated tumor cell elimination.
Furthermore, we observed consistent positive correlations between
SLC25A13 and specific immune checkpoint molecules, particularly KDR and
TGFBR1, across various tumor types. This correlation pattern implies
potential involvement of SLC25A13 in the regulatory pathways of these
immune checkpoint molecules. Together, these findings highlight the
distinct immunomodulatory functions of SLC25A13 in different cancer
contexts. Utilizing the ComPPI database, we identified proteins that
potentially interact with SLC25A13, suggesting intricate networks
influencing tumor behavior (Fig. [93]2G). ESTIMATE analysis revealed a
negative correlation of SLC25A13 expression with stromal scores, immune
scores, and overall ESTIMATE scores across diverse cancers,
underscoring its complex role in modifying the tumor microenvironment
(Fig. [94]2H). Gene Set Enrichment Analysis (GSEA) at the pan-cancer
level demonstrated that SLC25A13 was significantly associated with cell
cycle-related pathways and various metabolic pathways (Figure S3A).
Differential pathway enrichment and prognostic implications of SLC25A13 in
glioma
First, we divided the glioma patients into high and low expression
groups based on the median SLC25A13 expression level. Subsequent
enrichment analysis of differentially expressed genes using hallmark
gene sets revealed significant pathway enrichments in both groups.
Notably, in both LGG and GBM, multiple pathways such as Mitotic
Spindle, G2M Checkpoint, and E2F Targets were significantly enriched in
the high expression group of SLC25A13 (Fig. [95]3A and S4C). Using the
KEGG database, we assessed the activity of various metabolic pathways
in the high and low SLC25A13 expression groups. In the high SLC25A13
expression group, certain metabolic pathways, including tyrosine
metabolism, were significantly suppressed (Figure S4A and B). To
further understand the relevance of SLC25A13 to oncogenic pathways, we
analyzed the correlation between SLC25A13 and quantitative data from
key oncogenic pathways in the TCPA. The results showed that SLC25A13
was positively correlated with the key oncogenic pathway PI3K/AKT in
glioma (Fig. [96]3B). Further enrichment analysis revealed that
SLC25A13 was significantly correlated with pathways involved in cell
growth and death, indicating its potential role in tumorigenesis and
cancer progression (Figure S5A). Scoring of 14 tumor functional states
based on CancerSEA definitions revealed significant positive
correlations between SLC25A13 expression and cell cycle, DNA damage
repair, EMT, migration, invasion, proliferation, and stemness
(Fig. [97]3C).
Fig. 3.
[98]Fig. 3
[99]Open in a new tab
Differential pathway enrichment and prognostic implications of SLC25A13
expression in gliomas. Statistical significance is denoted as follows:
ns (not significant, p > 0.05), *(p < 0.05), **(p < 0.01),
***(p < 0.001), and **** (p < 0.0001). A Hallmark gene set enrichment
analysis for high and low SLC25A13 expression groups in glioblastoma
multiforme (GBM). B Spearman correlation of SLC25A13 expression with
quantification of functional proteins in TCPA-RPPA database at the
pathway level. C Spearman correlation of SLC25A13 with GSVA scores for
14 functional states in gliomas. D Meta-analysis using inverse variance
method for SLC25A13-related overall survival (OS), disease-specific
survival (DSS), and progression-free interval (PFI) in TCGA LGG cohort.
E Association of SLC25A13 expression with different clinicopathologic
features in TCGA pan-glioma cohort. Differences between the two groups
were assessed using the Wilcoxon rank-sum test. F Correlation between
SLC25A13 expression quartiles and immune response scores in gliomas
By performing a meta-analysis of four different survival outcome
metrics—overall survival (OS), disease-specific survival (DSS),
disease-free interval (DFI), and progression-free interval (PFI)—in the
TCGA LGG cohort, we confirmed SLC25A13 as a risk factor for patient
prognosis (Fig. [100]3D). Furthermore, Kaplan–Meier survival curve
analysis revealed that patients with high SLC25A13 expression in both
the pan-glioma and LGG cohorts had significantly worse prognosis
(Figure S4D). Additionally, we analyzed the association of SLC25A13
with other clinical features. The results indicated that SLC25A13
expression increased with WHO grade and was higher in IDH wild-type
patients, elderly patients, and patients without 1p19q co-deletion
(Fig. [101]3E). This suggests that SLC25A13 may be associated with more
malignant clinical features. Moreover, we divided all patients into
four quartiles based on gene expression percentiles: Q1 (highest 25%),
Q2, Q3, and Q4 (lowest 25%). Analysis of the immune response and
genomic status, as proposed by previous studies [[102]23], showed that
the CTA score and stromal fraction were higher in the low SLC25A13
expression group (Fig. [103]3F).
SLC25A13 is preferentially expressed in malignant cells and correlates with
drug sensitivity
Using the TISCH2 database, we obtained single-cell resolution
expression data for SLC25A13 in gliomas. Our analysis revealed that
SLC25A13 was predominantly expressed in oligodendrocytes and malignant
cells (Fig. [104]4A). To further elucidate the role of SLC25A13, we
classified cells into SLC25A13-positive and SLC25A13-negative groups.
In dataset [105]GSE131928, the proportion of MES-like malignant cells
was significantly higher in the SLC25A13-positive group compared to the
SLC25A13-negative group (Fig. [106]4B). Utilizing drug sensitivity data
from the PRISM, CTRP, GDSC1, and GDSC2 databases, we analyzed the
correlation between SLC25A13 expression and responsiveness to various
chemotherapeutic agents. The results indicated that high SLC25A13
expression was associated with resistance to multiple drugs, including
temozolomide, a commonly used first-line therapeutic agent for glioma
patients (Fig. [107]4C). To identify potential therapeutic agents that
could counteract SLC25A13-mediated tumor promotion, we employed the
XSum algorithm to analyze connectivity map (cMAP) drug sensitivity
data. Consistent results were obtained in both lower-grade glioma (LGG)
and glioblastoma multiforme (GBM) datasets, with the drug STOCK1N-35696
receiving the lowest score (Figure S5B and C). This suggests that
STOCK1N-35696 may be a potential therapeutic agent for glioma patients
with high SLC25A13 expression. To further confirm the preferential
expression of SLC25A13, we first performed deconvolution analysis of
spatial transcriptomics data to determine the spatial distribution of
cellular subpopulations (Fig. [108]4D). We then analyzed the
relationship between SLC25A13 expression and spatial cellular
components. The results indicated that SLC25A13 had higher expression
levels in malignant regions and was proportional to the number of
malignant cells (Fig. [109]4E–G). Additionally, analysis of the glioma
single-cell datasets from TISCH2 revealed that SLC25A13 expression was
positively correlated with malignant cells and negatively correlated
with CD8 T cells, but not significantly correlated with M1 macrophages
(Fig. [110]4H–J).
Fig. 4.
[111]Fig. 4
[112]Open in a new tab
Single-cell and spatial transcriptome resolution of SLC25A13 expression
in the glioma tumor microenvironment. A Heatmap showing SLC25A13
expression across various cell types from single-cell resolution data.
B Proportion of different cell types in SLC25A13-positive and
SLC25A13-negative groups from dataset [113]GSE131928. C Spearman
correlation of dose–response curves (area under the curve values) in
the CTRP and PRISM databases and IC50 values in the GDSC1 and GDSC2
databases with SLC25A13 expression. D Spatial distribution of different
cellular components in glioblastoma. E Spatial distribution of SLC25A13
expression in glioblastoma microenvironment. F Correlation of SLC25A13
expression with the content of different cell types in spatial
transcriptome data. G Comparison of SLC25A13 expression levels in spots
defined as malignant mixed niche versus normal niche. Differences
between the two groups were assessed using the Wilcoxon rank-sum test.
H Correlation between SLC25A13 expression and abundance of malignant
cells. I Correlation between SLC25A13 expression and presence of CD8 T
cells. J Correlation between SLC25A13 expression and presence of M1
tumor-associated macrophages (TAMs)
Identification and prognostic evaluation of SLC25A13-related core genes in
gliomas
First, we performed a correlation analysis of SLC25A13 in the TCGA
pan-glioma cohort and identified the top 150 genes with the highest
correlation coefficients as co-expressed genes (Table S1).
Interestingly, we found that all 150 co-expressed genes were associated
with poor prognosis in the TCGA pan-glioma cohort (Table S2). To
further identify the core co-expressed genes, we employed a combination
of multiple machine learning approaches (Fig. [114]5A). Using the
C-index as the evaluation criterion, we selected the overlapping genes
from the top 15 combinations, ultimately identifying seven core genes:
GAS2L3, KIF2C, NDE1, EZH2, SMC4, SLC30A6, and SLC25A13 (Fig. [115]5B).
We obtained risk scores for each sample by performing a multivariate
Cox analysis of the seven core genes. Our analysis revealed that the
SLC25A13-related risk score is significantly associated with several
critical clinical parameters, including age, IDH mutation status,
1p/19q codeletion status, MGMT promoter methylation status, glioma
stage, overall survival, and survival status (Fig. [116]5C).
Furthermore, in both TCGA and CGGA datasets, Kaplan–Meier survival
curve analysis revealed that patients with high-risk scores had a
significantly worse prognosis (Fig. [117]5D).
Fig. 5.
[118]Fig. 5
[119]Open in a new tab
Identification and prognostic evaluation of SLC25A13-related risk score
in gliomas. A Combination of multiple machine learning methods to
screen SLC25A13-related core genes. B Intersection genes for the top 15
machine learning method combinations with the highest C-index. C
Association of SLC25A13-related risk score with clinical parameters
such as gender, age, IDH mutation status, 1p19q co-deletion status,
MGMT promoter methylation status, glioma stage, overall survival, and
survival status. D Kaplan–Meier survival curve analysis of high- and
low-risk scoring groups in the glioma cohorts from TCGA and CGGA
databases
In addition, consistent with our findings in glioma cohorts, patients
in the high-risk group across ACC, KICH, prostate cancer (PRAD), LIHC,
KIRP, LUAD, mesothelioma (MESO), PAAD, and sarcoma (SARC) showed a
significantly shorter overall survival compared to those in the
low-risk group (Figure S6A-I). Intriguingly, in patients with THYM,
those in the high-risk group experienced a longer survival duration
compared to their low-risk counterparts (Figure S6J). This distinct
survival pattern in THYM patients suggests a potential differential
impact of SLC25A13-related risk score in this particular subtype,
warranting further investigation. Next, we evaluated the predictive
capability of the SLC25A13-related risk score for the prognosis of
glioma patients using time-dependent ROC analysis. In the TCGA
pan-glioma dataset, the three-year area under the curve (AUC) value for
the SLC25A13-related risk score was 0.9 (Figure S6K). In the CGGA
dataset, the three-year AUC value was 0.82 (Figure S6L), indicating
that the SLC25A13-related risk score has a strong discriminative
ability for the prognosis of glioma patients.
Development and evaluation of a prognostic model for glioma patients based on
SLC25A13-related risk score
We developed a comprehensive prognostic model for glioma patients that
included the SLC25A13-related risk score, age, IDH mutation status,
1p/19q co-deletion status, MGMT promoter methylation status, and glioma
stage (Fig. [120]6A). To evaluate the prognostic significance of these
variables, we performed both univariate and multivariate regression
analyses. The results indicated that the SLC25A13-related risk score
was an independent prognostic factor for glioma patients
(Fig. [121]6B). Additionally, we validated our glioma prognostic model
using three independent datasets: TCGA, CGGA693, and CGGA325. Receiver
operating characteristic (ROC) curve analysis demonstrated high
discriminative ability of the model, with three-year AUC values of
0.84, 0.95, and 0.84 for the TCGA, CGGA693, and CGGA325 datasets,
respectively (Fig. [122]6C–E). Calibration plots, which compare
predicted survival probabilities with actual outcomes, indicated a high
degree of concordance, suggesting that the model reliably predicts
patient survival (Fig. [123]6F–H). Decision curve analysis (DCA)
further showed that the nomogram offers superior clinical net benefit
compared to other prognostic factors (F[124]ig. [125]6I–K). These
findings underscore the clinical utility of the SLC25A13-related risk
score in improving the prognostication and management of glioma
patients.
Fig. 6.
[126]Fig. 6
[127]Open in a new tab
Development and evaluation of a prognostic model for glioma patients
based on SLC25A13-related risk score. A The nomogram constructed based
on SLC25A13-related risk scores and other clinicopathologic features. B
Univariate and multivariate Cox regression analyses identifying
SLC25A13-related risk score as an independent prognostic factor. C–E
Receiver Operating Characteristic (ROC) analysis illustrating the
model's discrimination ability in predicting glioma patient outcomes
across the TCGA, CGGA693, and CGGA325 datasets. F–H Calibration plots
comparing the predicted probabilities of survival outcomes with the
observed outcomes in glioma patients. I–K Decision curve analysis
indicating greater clinical net benefit of the nomogram compared to
other prognostic factors
Comprehensive single cell analysis of SLC25A13-related risk score in the
tumor microenvironment
First, we selected newly diagnosed glioblastoma multiforme (GBM)
samples from the single-cell RNA sequencing dataset [128]GSE182109.
Subsequently, we analyzed the single-cell data using the Seurat
standard workflow, annotating 29,778 cells based on specific markers.
These cells were categorized into tumor-associated macrophages (TAMs),
malignant cells, T cells, pericytes, oligodendrocytes, and B cells
(Fig. [129]7A and B). Further validation through inferCNV analysis
demonstrated significant copy number variation (CNV) events in tumor
cells compared to TAMs, notably the amplification of chromosome 7 and
the deletion of chromosome 10 (Fig. [130]7C). This result confirms the
accuracy of our annotation of tumor cells. We examined the expression
of the seven core genes at single-cell resolution and found that all
seven core genes were detected with variable expression in malignant
cells (Figure S7A). Subsequently, based on the previously determined
formula, we calculated the risk score for each cell, categorizing
malignant cells with risk scores in the top 30% as high-risk cells,
those in the bottom 30% as low-risk cells, and those in the middle as
medium-risk cells (Fig. [131]7D and E).
Fig. 7.
[132]Fig. 7
[133]Open in a new tab
Single-cell transcriptomic analysis reveals the implications of
SLC25A13-based risk score in the tumor microenvironment. A, B
Single-cell clustering and annotation of cell types derived from the
[134]GSE182109 dataset. C InferCNV analysis demonstrating significant
copy number variation events in tumor cells. D Distribution of
SLC25A13-related risk scores across different cell subtypes. E
Categorization of malignant cells into high, medium, and low-risk
groups based on SLC25A13-related risk scores. F CellChat analysis
revealing differential communication patterns among different risk
groups. G, H Variation in the strength of the PTPR signaling pathway
among malignant cells across different risk groups. I, J CALCR
signaling pathway sent by medium-risk malignant cells to pericytes. K
The scMetabolism analysis showing significant metabolic differences
between high and low-risk malignant cells
To explore whether these cells displayed differential interactions with
other cellular components in tumor microenvironment, we employed the
CellChat algorithm for cell communication analysis. The results
revealed intricate cellular communication networks between different
cell subpopulations (Fig. [135]7F and S8A-B). For instance, the PTPR
signaling pathway was predominantly initiated by low-risk malignant
cells and received by oligodendrocytes, mediated specifically through
the LRRC4C-PTPRF receptor-ligand pair (Fig. [136]7G and H). Similarly,
the CALCR signaling pathway was mainly sent by Medium-risk malignant
cells to pericytes, facilitated by the DM-CALCR receptor-ligand pair
(Fig. [137]7I and J). Additionally, we analyzed metabolic pathways
using scMetabolism, revealing significant metabolic differences between
high-risk and low-risk malignant cells (Fig. [138]7K). This suggests
that there is a notable distinction between malignant cells with
high-risk scores and those with low-risk scores.
Elevated SLC25A13 expression in glioblastoma tissues verified by multi-source
data and experimental analysis
To validate the previous results, we utilized three independent
datasets—[139]GSE50161, Rembrandt, and [140]GSE4290—to examine the
expression of SLC25A13 in glioma tissues and normal brain tissues. The
results confirmed the up-regulation of SLC25A13 expression in tumor
tissues (Fig. [141]8A). In addition, we collected six pairs of
surgically resected glioblastoma samples along with their corresponding
adjacent non-cancerous tissues. To evaluate the expression levels of
SLC25A13, quantitative PCR (qPCR) assays were conducted. The qPCR
results demonstrated significantly elevated expression of SLC25A13 in
glioma tissues compared to the adjacent non-cancerous tissues
(Fig. [142]8B). This observation was further corroborated by
immunohistochemistry (IHC) analysis, which also revealed a high
expression of SLC25A13 protein in the tumor samples (Fig. [143]8C and
D). Western blot (WB) analysis of the paired samples provided
consistent results, confirming that SLC25A13 levels were markedly
higher in the glioma tissues relative to the adjacent non-cancerous
tissues (Fig. [144]8E and F).
Fig. 8.
[145]Fig. 8
[146]Open in a new tab
Elevated SLC25A13 expression in glioblastoma tissues verified by
multi-source data and experimental analysis. Statistical significance
is denoted as follows: ns (not significant, p > 0.05), *(p < 0.05),
**(p < 0.01), ***(p < 0.001), and ****(p < 0.0001). A SLC25A13
expression in glioblastoma tissues versus non-tumour tissues across
[147]GSE50161, Rembrandt, and [148]GSE4290 datasets. Differences
between the two groups were assessed using the Wilcoxon rank-sum test.
B The qPCR analysis showing elevated SLC25A13 mRNA levels in
glioblastoma tissues (n = 6 per group, paired t-test). C, D
Immunohistochemical staining indicating stronger SLC25A13 protein
expression in glioblastoma tissues compared to paraneoplastic tissue.
E, F Western blot assay evaluating SLC25A13 protein expression in tumor
tissues compared to adjacent non-cancerous tissues
SLC25A13 promotes the malignant behavior of glioblastoma cells
To investigate the role of SLC25A13 in glioma cells, we generated an
SLC25A13 knockdown model using shRNA in U87 and U251 cell lines (Figure
S9A). EDU staining revealed that the EDU fluorescence intensity in U87
and U251 cells decreased after SLC25A13 knockdown, indicating a
reduction in cell proliferation capacity (Fig. [149]9A and B). The CCK8
assay further confirmed a decrease in the viability of U87 and U251
cells following SLC25A13 knockdown (Fig. [150]9C). Flow cytometry
analysis of cell cycle distribution demonstrated that SLC25A13
knockdown resulted in a decreased G0/G1 phase ratio and an increased
G2/M phase ratio, suggesting a G2/M phase block (Fig. [151]9D). These
findings imply that SLC25A13 is essential for proper cell cycle
progression in glioma cells. Moreover, the scratch assay indicated
reduced migratory capability in both cell lines after SLC25A13
knockdown (Fig. [152]9E and F), suggesting a role for SLC25A13 in cell
motility. An increase in cell apoptosis was also observed
(Fig. [153]9G), indicating increased programmed cell death, which
likely contributes to the observed reduction in cell viability and
proliferation. Finally, the colony formation assay demonstrated a
significant reduction in clonogenic potential following SLC25A13
knockdown (Fig. [154]9H).
Fig. 9.
[155]Fig. 9
[156]Open in a new tab
Impact of SLC25A13 knockdown on the functionality of glioblastoma
cells. Statistical significance is denoted as follows: ns (not
significant, p > 0.05), *(p < 0.05), **(p < 0.01), ***(p < 0.001), and
**** (p < 0.0001). Data are presented as mean ± SD of three independent
experiments (biological replicates, n = 3). A, B Reduction in EdU
staining intensity following SLC25A13 knockdown. C CCK-8 assay confirms
reduced cell viability following SLC25A13 knockdown. D Changes in cell
cycle distribution with decreased G1 phase percentage and increased G2
phase cells. E, F Diminished migratory capability in glioblastoma cells
after SLC25A13 knockdown. G Increased apoptosis levels observed
following SLC25A13 knockdown. H Significant reduction in clonogenic
potential after SLC25A13 knockdown
SLC25A13 promotes the malignant progression of glioblastoma in vivo
To confirm the tumor-promoting effects of SLC25A13 in vivo, we
established an in situ tumor model in nude mice. In vivo imaging
demonstrated that knockdown of SLC25A13 significantly impaired the
tumorigenic capacity of glioma cells (Fig. [157]10A and S9B).
Hematoxylin and eosin (H&E) staining and Ki67 immunohistochemistry of
tumor tissue sections further confirmed that SLC25A13 knockdown
resulted in a reduced tumor-forming ability (Fig. [158]10B).
Additionally, the knockdown of SLC25A13 prolonged the survival of nude
mice and mitigated weight loss compared to the control group,
suggesting that the malignant progression of glioma cells was
significantly inhibited (Fig. [159]10C and D).
Fig. 10.
[160]Fig. 10
[161]Open in a new tab
In vivo experiments validating the tumor-promoting effects of SLC25A13.
Statistical significance is denoted as follows: ns (not significant,
p > 0.05), *(p < 0.05), **(p < 0.01), ***(p < 0.001), and
****(p < 0.0001). A In vivo imaging experiments illustrating the
tumor-forming ability of both control and experimental groups. B
Hematoxylin and eosin (H&E) staining and Ki67 staining of tumor tissues
from control and experimental groups. C Survival curve analysis of nude
mice in control and experimental groups (n = 8 per group). D Body
weight of nude mice in control and experimental groups over time (n = 8
per group)
Discussion
SLC25A13, a key member of the solute carrier family, has been
predominantly studied in the context of non-tumorigenic diseases
[[162]32–[163]34], with its role in cancer largely underexplored. In
this study, we identified the dysregulation of SLC25A13 expression and
observed genetic alterations across various cancer types. We further
elucidated its relationship with the tumor immune microenvironment and
evaluated its prognostic significance. Importantly, our findings
provide the first evidence that SLC25A13 plays a role in promoting the
malignant progression of glioma, highlighting its potential as a
therapeutic target for this aggressive cancer.
With the publication of data from large international pan-cancer
studies such as TCGA and CPTAC, significant advancements in our
understanding of cancer have emerged. These studies, leveraging
extensive cohorts, have yielded numerous new insights into oncogenesis
[[164]4, [165]35]. In our investigation, we first analyzed the
expression pattern of the SLC25A13 gene across various cancers using
sequencing data from TCGA and normal tissues data from the GTEx
project. We further validated these findings with protein
quantification data from CPTAC. The widespread dysregulation of
SLC25A13 expression observed in these analyses suggests a potentially
crucial role in tumorigenesis. Copy number variation and aberrant
methylation are critical mechanisms by which tumor cells modulate gene
expression to gain survival advantages [[166]36, [167]37]. We observed
consistent aberrations in SLC25A13 expression correlated with copy
number amplification and methylation changes, suggesting that
alterations in SLC25A13 expression at the genetic level may contribute
to malignant transformation. In recent years, immunotherapy has become
a focal point in cancer treatment research. However, the variable
responses among different cancer patients have limited its universal
application [[168]38]. The expression of immune checkpoints and the
composition of the tumor immune microenvironment significantly
influence immunotherapy efficacy [[169]39]. Tumor mutational burden and
microsatellite instability are considered powerful predictors of
immunotherapy response [[170]40, [171]41]. This study investigated the
associations between SLC25A13 expression and multiple immune-related
parameters, including microsatellite instability, tumor mutational
burden, immune cell infiltration patterns, immunophenotype profiles,
and immunomodulatory molecule expression. The results revealed both
conserved patterns and tumor-specific variations in SLC25A13-mediated
immunomodulation across different cancer types, suggesting that
targeted regulation of SLC25A13 could potentially enhance
immunotherapeutic outcomes in specific tumor types.
Gliomas, the most common primary malignant brain tumors in adults, are
classified by the World Health Organization (WHO) into grades I-IV,
with glioblastoma (WHO grade IV) being the most aggressive form and
having a median survival of approximately 15 months [[172]42, [173]43].
Given the abnormally high expression of the SLC25A13 gene in gliomas
and its association with poorer patient prognosis, we conducted an
in-depth analysis of the potential oncogenic role of SLC25A13 in
gliomas. Gene set enrichment analysis consistently revealed a
significant role for SLC25A13 in the cell cycle of glioma cells.
Notably, our bioinformatics analysis results were corroborated by
subsequent cell cycle experiments, which confirmed that SLC25A13
knockdown induced a G2/M phase arrest. Moreover, both ex vivo and in
vivo experiments consistently demonstrated that SLC25A13 was highly
expressed in tumor tissues and facilitated the malignant behavior of
glioblastoma cells. These findings suggest that SLC25A13 may serve as a
potential therapeutic target in glioblastoma.
The identification of cancer biomarkers has been a critical focus in
oncology research, given their importance in early cancer diagnosis,
prognosis prediction, and treatment response assessment [[174]44]. In
this study, we developed an SLC25A13-associated risk score using
multiple machine learning methods and constructed a nomogram by
integrating the risk score with other clinical features to guide
patient prognosis. Multiple independent datasets validated the
reliability of the nomogram for prognostic prediction. Furthermore, our
analysis revealed that glioma patients with elevated SLC25A13
expression are more likely to exhibit resistance to temozolomide.
Interestingly, STOCK1N-35696 emerged as a potentially effective
therapeutic option for this subset of patients. Additionally, higher
SLC25A13 expression correlated with more malignant clinicopathologic
features of gliomas. These findings suggest that SLC25A13 plays a
crucial role in the malignant progression of gliomas and may contribute
to treatment resistance. Consequently, we propose that SLC25A13
represents a promising biomarker for predicting tumor aggressiveness
and patient outcomes in glioma.
In recent years, rapid advancements in single-cell sequencing and
spatial transcriptome technologies have significantly enhanced our
understanding of the tumor microenvironment [[175]45, [176]46].
Motivated by these developments, we first determined the localization
of SLC25A13 within the tumor microenvironment, finding a high
correlation with malignant cells and a negative correlation with
CD8 + T cell infiltration. This suggests that SLC25A13 may play a
crucial role in malignant cells and may be associated with
immunosuppression. Furthermore, our risk score, derived from the
weighted expression of various genes, reflects the expression patterns
of multiple genes. We evaluated these patterns at single-cell
resolution, categorizing malignant cells into high, medium, or low-risk
groups. Notably, low-risk malignant cells exhibited enhanced PTPR
signaling activity. Previous studies have shown that PTPRF functions as
a tumor suppressor in various cancers [[177]47, [178]48], supporting
our hypothesis that low-risk cells are less malignant than high-risk
cells. Medium-risk malignant cells, on the other hand, exhibited
stronger CALCR signaling activity, while CALCRL has been reported to
play a pro-carcinogenic role in tumors [[179]49, [180]50].
Additionally, differences in metabolic pathway activity between the two
groups of cells further confirm that the risk score effectively
distinguishes the heterogeneity of tumor cells. These findings provide
a microscopic explanation for the poorer prognosis observed in
high-risk patients.
In summary, our study presents, for the first time, a comprehensive
pan-cancer analysis of SLC25A13, elucidating its role in promoting the
malignant progression of gliomas. These findings suggest that SLC25A13
may serve as a potential therapeutic target in this aggressive
malignancy. Moreover, our results support the utility of SLC25A13 as a
biomarker for predicting glioma aggressiveness and patient outcomes.
The nomogram developed based on the SLC25A13-related risk score offers
a valuable tool for prognostic stratification in clinical practice.
However, several limitations warrant consideration. The relationship
between SLC25A13 and the tumor immune microenvironment requires further
experimental validation. Additionally, the molecular mechanisms
underlying SLC25A13-mediated promotion of malignant behavior in
glioblastoma cells necessitate more in-depth investigation. We
anticipate that our work will provide a foundation for future studies
aimed at elucidating the specific molecular pathways through which
SLC25A13 influences tumor biology.
Conclusions
In conclusion, our study provides a comprehensive analysis of SLC25A13
across various cancer types. We demonstrated for the first time that
SLC25A13 is highly expressed in glioma tissues and confirmed through
both in vivo and in vitro experiments that SLC25A13 promotes the
malignant progression of glioblastoma. Furthermore, our SLC25A13-based
nomogram effectively guides glioma patient prognosis. Collectively, our
findings suggest that SLC25A13 may serve as a potential therapeutic
target for glioma patients, warranting further investigation into its
clinical applications.
Supplementary Information
[181]12935_2025_3696_MOESM1_ESM.tif^ (1.9MB, tif)
Supplementary Material 1: Figure S1A. Genetic variation patterns of
SLC25A13 across different tumor types.
[182]12935_2025_3696_MOESM2_ESM.tif^ (4.9MB, tif)
Supplementary Material 2: Figure S2. Correlation of SLC25A13 with
immunomodulatory molecules. (A) Correlation of SLC25A13 with
immunostimulatory molecules. (B) Correlation of SLC25A13 with
chemokines.
[183]12935_2025_3696_MOESM3_ESM.tif^ (6.2MB, tif)
Supplementary Material 3: Figure S3. Gene Set Enrichment Analysis
(GSEA) of the HALLMARK and metabolic pathways in SLC25A13 high and low
expression groups. (A) Dot plots illustrating the intensity of
different pathway enrichment in the high SLC25A13 expression group.
[184]12935_2025_3696_MOESM4_ESM.tif^ (3.8MB, tif)
Supplementary Material 4: Figure S4. Impact of SLC25A13 expression on
metabolic pathway strength in gliomas. (A) Scoring of various metabolic
pathways in GBM across high and low SLC25A13 expression groups. (B)
Scoring of various metabolic pathways in LGG across high and low
SLC25A13 expression groups. (C) Hallmark gene set enrichment analysis
for high and low SLC25A13 expression groups in LGG. (D) Survival curve
analysis of high and low SLC25A13 expression groups across different
grades of gliomas.
[185]12935_2025_3696_MOESM5_ESM.tif^ (3.9MB, tif)
Supplementary Material 5: Figure S5. Pathway enrichment analysis and
prediction of drug sensitivity. (A) Enrichment of oncogenic pathways in
the SLC25A13 high-expression group in gliomas. (B-C) Identification of
potential therapeutic agents for patients with high SLC25A13 expression
using the XSum algorithm in LGG and GBM.
[186]12935_2025_3696_MOESM6_ESM.tif^ (2.5MB, tif)
Supplementary Material 6: Figure S6. Broader implications of
SLC25A13-related risk score across various cancers. (A-J) Prognostic
value of risk scores in multiple cancers. (K-L) Time-dependent ROC
curves illustrating the discriminative power of the SLC25A13-related
risk score in the TCGA and CGGA glioma cohorts.
[187]12935_2025_3696_MOESM7_ESM.tif^ (2.4MB, tif)
Supplementary Material 7: Figure S7. Expression of core genes in the
tumor microenvironment of glioblastoma. (A) Expression of seven core
genes across different cellular compositions.
[188]12935_2025_3696_MOESM8_ESM.tif^ (7.3MB, tif)
Supplementary Material 8: Figure S8. Cellular interactions between
malignant cells at varying risk levels in the microenvironment. (A)
Detailed receptor-ligand pairs involved in signaling from malignant
cells at varying risk levels in cellular communication. (B) Detailed
receptor-ligand pairs for signals received by malignant cells at
varying risk levels in cellular communication.
[189]12935_2025_3696_MOESM9_ESM.tif^ (1.9MB, tif)
Supplementary Material 9: Figure S9. (A) Validation of SLC25A13
knockdown efficiency in U87 and U251 cell lines. (B) Quantitative
analysis of in vivo imaging results at different time intervals (n = 5
per group).
[190]Supplementary Material 10.^ (5.9KB, csv)
[191]Supplementary Material 11.^ (7.3KB, csv)
[192]Supplementary Material 12.^ (9.9KB, xlsx)
Acknowledgements