Access to this full-text is provided by Springer Nature.
Content available from Cancer Cell International
This content is subject to copyright. Terms and conditions apply.
RESEARCH Open Access
© The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use,
sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included
in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The
Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available
in this article, unless otherwise stated in a credit line to the data.
Kang et al. Cancer Cell International (2024) 24:23
https://doi.org/10.1186/s12935-023-03189-x
Introduction
Cervical cancer ranks rst in the female reproductive
tract [1] and consists mainly of squamous and adeno-
carcinoma types of the cervix. Patients diagnosed with
early or locally advanced cervical cancer have achieved
some remission and have experienced high survival rates
with radical resection or concurrent radiotherapy [2, 3],
which have signicantly reduced mortality, especially in
developing countries and poor regions [4]. Nevertheless,
the prognosis and treatment outcomes for patients with
refractory cervical cancer, encompassing those aicted
with recurrent, persistent, or metastatic forms, remain
Cancer Cell International
†Jiawen Kang and Jingwen Jiang have contributed equally to this
work and share rst authorship.
*Correspondence:
Yong Zhang
374794955@qq.com
Jie Tang
tangjie_73@163.com
Lesai Li
LLS0731@126.com
1Department of Gynecologic Oncology, School of Medicine, Hunan
Cancer Hospital/The Aliated Cancer Hospital of Xiangya, Central South
University, Changsha, Hunan, China
2Department of Clinical Medicine, Medical College of Hunan Normal
University, Changsha, Hunan, China
Abstract
Patients with recurrent or metastatic cervical cancer are in urgent need of novel prognosis assessment or treatment
approaches. In this study, a novel prognostic gene signature was discovered by utilizing cuproptosis-related
angiogenesis (CuRA) gene scores obtained through weighted gene co-expression network analysis (WGCNA) of
The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) datasets. To enhance its reliability, the
gene signature was rened by integrating supplementary clinical variables and subjected to cross-validation.
Meanwhile, the activation of the VEGF pathway was inferred from an analysis of cell-to-cell communication, based
on the expression of ligands and receptors in cell transcriptomic datasets. High-CuRA patients had less inltration
of CD8 + T cells and reduced expression of most of immune checkpoint genes, which indicated greater diculty
in immunotherapy. Lower IC50 values of imatinib, pazopanib, and sorafenib in the high-CuRA group revealed the
potential value of these drugs. Finally, we veried an independent prognostic geneSFT2D1was highly expressed
in cervical cancer and positively correlated with the microvascular density. Knockdown ofSFT2D1signicantly
inhibited ability of the proliferation, migration, and invasive in cervical cancer cells. CuRA gene signature provided
valuable insights into the prediction of prognosis and immune microenvironment of cervical cancer, which could
help develop new strategies for individualized precision therapy for cervical cancer patients.
Keywords Cervical cancer, Cuproptosis, Angiogenesis, Prognosis, Immune inltration, Cell communication
Identication of a new gene signature
for prognostic evaluation in cervical
cancer: based on cuproptosis-associated
angiogenesis and multi-omics analysis
JiawenKang1,2†, JingwenJiang2†, XiaoqingXiang2, YongZhang2*, JieTang1* and LesaiLi1*
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 2 of 15
Kang et al. Cancer Cell International (2024) 24:23
disheartening [5–7]. e implementation of novel scien-
tic methodologies and an enhanced comprehension of
tumor pathogenesis are poised to augment our under-
standing of cervical cancer mechanisms and ultimately
ameliorate the prognosis for individuals grappling with
refractory cervical cancer. It is necessary to explore new
ways to improve the prognosis of patients with refractory
cervical cancer.
Tumor angiogenesis, inammatory inltration of the
tumor microenvironment, and programmed cell death
processes have been identied as contributing factors
to tumor metastasis [8]. Tumor cells exhibit elevated
secretion of pro-angiogenic factors, which stimulate the
development of heterogeneous and immature neovascu-
larization. is heteromorphic neovascularization often
leads to a hypoxic microenvironment caused by inade-
quate perfusion, thereby favoring the survival and growth
of more aggressive tumor cells [9]. Simultaneously, the
presence of pro-angiogenic factors within the tumor
microenvironment facilitates the process of angiogenesis
and immunosuppression [10]. Consequently, angiogen-
esis fosters the tumor’s ability to evade the immune sys-
tem and engenders drug resistance. Tumor angiogenesis
stands as a contributing factor to recurrence, prompting
the clinical utilization of anti-angiogenic drugs in the
management of advanced or recurrent cervical cancer,
resulting in notable enhancements in survival rates [11].
However, the limited applicability of targeting mature
stable vessels [12] and the presence of various treatment-
related side eects have necessitated the exploration of
novel therapeutic approaches [13]. Notably, the remark-
able eectiveness of immunotherapy in cervical cancer
has highlighted the signicance of targeting angiogen-
esis in the tumor microenvironment for immunothera-
peutic interventions [14, 15]. Consequently, for patients
with metastatic, persistent, and recurrent cervical cancer
who exhibit PD-L1 positivity, the combination of pabli-
zumab and chemotherapy, with or without bevacizumab,
has emerged as the preferred rst-line treatment option
[6]. Investigating modications in the immune microen-
vironment and immune checkpoint genes within tumors
experiencing varying angiogenic states can provide valu-
able insights into the development of precise combina-
tions of vascular targeting therapy and immunotherapy.
Recent ndings indicate that several well-established
regulators of programmed cell death play a role in pro-
moting angiogenesis [8, 16–20]. Additionally, cupropto-
sis, a distinct form of programmed cell death, is primarily
characterized by the excessive accumulation of intracel-
lular copper, leading to cell death [21]. It has been shown
increased intratumor copper concentrations promote
tumor growth and invasion as well as treatment resis-
tance [22]. Serum copper concentrations have been
found to exhibit a correlation with tumor progression
and morbidity [23]. Additionally, cuproptosis, a recently
identied mode of cell death, has been reported to play
a role in tumor growth, angiogenesis, and tumor metas-
tasis [24, 25]. Studies have demonstrated that copper
facilitates tumor angiogenesis by activating various
angiogenic factors, such as basic broblast growth factor
(bFGF) and vascular endothelial growth factor (VEGF)
[26]. Moreover, copper is implicated in signal transduc-
tion processes within endothelial cells, thereby inuenc-
ing angiogenesis [23]. In summary, copper assumes an
indispensable function in the advancement of tumors
as a trace element crucial for the proliferation of cancer
cells and the formation of blood vessels within tumors.
Further exploration of the correlation between angiogen-
esis and cuproptosis is imperative, as it holds potential
for novel treatment approaches [19, 27]. e association
between angiogenesis and cuproptosis in cervical cancer
has yet to be investigated, thus necessitating a meticu-
lous and proactive investigation employing innovative
methodologies.
e emergence of precision oncology and the integra-
tion of big data have facilitated the utilization of bulk
RNA sequencing to uncover the mean gene expression
in tissues, thereby enabling exploration into the realm
of cognitive dierential gene expression [28]. Further-
more, the progression of technological tools has allowed
for the implementation of single cell sequencing, which
has proven instrumental in discerning dierential gene
expression among cells and investigating intricate cell
populations [29–31]. is technique has signicantly
contributed to the elds of tumor diagnosis, targeted
therapy, and prognosis prediction [32, 33]. e analy-
sis of intercellular communication in cell populations
aids in the elucidation of communication and signaling
mechanisms among diverse cells [34]. Additionally, cor-
relation analysis of receptor-ligand pairs enables a deeper
comprehension of cellular functionality and regulatory
networks. In this study, we have developed a prognos-
tic model for the CuRA gene using both single-cell RNA
sequencing and bulk RNA sequencing. is model holds
promise for the development of innovative prognostic
prediction models and treatment approaches for individ-
uals with cervical cancer.
Materials and methods
Data acquisition
e cervical cancer patient dataset (the TCGA-CESC
cohort) was obtained from the Cancer Genome Atlas
as the training group (TCGA, https://portal.gdc.cancer.
gov/). For external validation, we used 55 cervical cancer
patients from the GSE52903 dataset in Gene Expression
Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/)
databases. We also included single-cell sequencing datas-
ets GSE168652 in cervical cancer using “Seurat” package
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 3 of 15
Kang et al. Cancer Cell International (2024) 24:23
[35] for calculating genetic correlations to score cells and
patients. e scores of CuRA gene-sets were calculated
by applying the “Percentage FeatureSet” function. e
GeneCards website (https://www.genecards.org/) was
used to obtain 1245 angiogenesis-related genes using the
“angiogenesis” keyword with a correlation > 1. R pack-
age “limma” [36] was used to obtain cuproptosis-related
angiogenesis (CuRA) genes. Genes with dierential
expression in normal and cervical tissues were obtained
from the GEPIA website (http://gepia.cancer-pku.cn/).
WGCNA
Single-sample gene set enrichment analysis (ssGSEA),
which estimated the relative enrichment of a particular
gene set in each sample by comparing the gene expression
data of that sample with that set. We performed weighted
gene co-expression network analysis (WGCNA) using
the “WGCNA” package [37]. e threshold of clustering
cut tree was set to 210, and the minimum threshold was
set to 80. We merged the modules with threshold < 0.5.
en, we performed analysis and included all genes in
modules of CuRA phenotypes with P < 0.05 for subse-
quent analysis.
Construction of the CuRA model
We congured the alpha parameter of the elastic network
to 0.5 and computed the errors for ridge regression, lasso
regression and elastic network regression. e model
regression was constructed using the “glmnet” package
[38]. e “timeROC” package [39] and “survivalROC”
package [40] were performed to plot ROC curves for
survival outcomes at dierent time points. Nomogram
based on logistic regression and Cox regression was con-
structed using the “rms” package [41].
Immune inltration analysis
“IOBR” package [42] was used to immune inltration
analysis. Six algorithm CIBERSORT, EPIC, MCP, XCELL,
TIMER, QUANTISEQ was performed to compare the
dierences between the high and low-CuRA groups. Dif-
ferences in the expression of immune checkpoint genes
were also compared. e relevant mutation data were
obtained from Cbioportal (https://www.cbioportal.org/
datasets). e “maftools” package [43] was performed for
visualization.
Cell communication analysis
We performed cell communication analysis using the R
package “CellChat” [34]. We ltered out cell communi-
cation with less than 10 cells and obtained the cell com-
munication relationship between each cell. We inferred
cell-to-cell communication at the pathway level, deduced
pathway-level interaction networks, and obtained the
interaction relationship between receptor-ligand pairs
and cell communication.
GSEA
e GSEA software (version 3.0) was downloaded from
the GSEA (http://software.broadinstitute.org/gsea/
index.jsp) website, divided the samples into high and
low expression groups based on the expression levels
ofSFT2D1. APvalue of < 0.05 and an FDR value of < 0.25
were considered statistically signicant. e correspond-
ing data was listed in Additional le 1: Table S5.
Drug sensitivity analysis
We searched the GDSC database to predict drug sensi-
tivity by comparing the IC50 of drugs among dierent
groups based on the CuRA scores. By analyzing in the
Drug Signatures Database (DSigDB, http://tanlab.ucden-
ver.edu/DSigDB), we listed corresponding small mol-
ecule drugs of relevant modeling genes (Additional le 1:
Table S4).
Cell culture
e ECT1/E6E7 cell line (ATCC: CRL-2614™), the
SiHa cell line (ATCC: HTB-35™), the CaSki cell line
(ATCC: CRM-CRL-1550™) were obtained from Ameri-
can Type Culture Collection (ATCC). Ect1/E6E7 cells,
SiHa cells were cultured in DMEM medium (Procell,
Wuhan,China) containing 10% fetal bovine serum (Pro-
cell, Wuhan, China), 100 U/mL penicillin and 100 µg/mL
streptomycin, and incubated at 37 °C under conditions
of 5% CO2. CaSki cells were cultured with the same con-
ditions in 1640 medium (Procell, Wuhan, China). ese
cells were transfected with synthetic small interfering
RNAs (GenePharma, Shanghai, China) by Lipo8000™
Transfection Reagent (Beyotime, Shanghai, China), and
the siRNA sequences targeting SFT2D1 gene are pro-
vided in the Additional le 1: Table S6.
Real-time uorescence quantitative PCR
We extracted total RNA of cells using TRIZOL reagent
(Vazyme, China), followed by adding chloroform for cen-
trifugation. e supernatant was collected and mixed
with isopropanol. e RNA pellet was washed with 75%
ethanol and air-dried. e purity of RNA was measured
using a spectrophotometer (ermo Fisher Scientic,
USA). e cDNA was synthesized using reverse tran-
scription reagent (TransGen, China) for uorescence
quantication (RT-qPCR).
Immunohistochemistry
We collected normal cervical tissue and cervical can-
cer tissue at the Hunan Provincial Cancer Hospital for
immunohistochemical staining, and it has been reviewed
and approved by the Ethics Committee of Hunan
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 4 of 15
Kang et al. Cancer Cell International (2024) 24:23
Cancer Hospital. After dewaxing of sections, heat antigen
retrieval was performed. e primary antibodies SFT2D1
(Immunoway, USA, 1:100) and CD31 (ZenBio, China,
1:100) were incubated overnight at 4 °C. e secondary
antibodies were incubated for 20 min using the PV-9000
kit (ZSGB-BIO, China). DAB reagent (ZSGB-BIO, China)
was used for antibody staining, with brown-yellow indi-
cating positive signal areas. Cell nuclei were stained blue
with hematoxylin (Servicebio, China). Images were cap-
tured using microscope (Zeiss, Germany) and analyzed
using Image J software (1.53, USA).
Western blotting
We added a mixture of cell lysis buer (Servicebio,
China) and protease inhibitor PMSF to the cells. e
sample was then denatured by adding SDS-loading buer
and subjected to electrophoresis. e PVDF membrane
(Millipore, USA) was wet-transferred at a constant cur-
rent. After blocking with skim milk at room tempera-
ture for 2 h, the primary antibody SFT2D1 (Immunoway,
USA, 1:1000) was incubated at 4℃ overnight. e sec-
ondary antibody (bioworld, USA, 1:10000) was incubated
at room temperature for 1 h, followed by detection with a
developing solution.
CCK-8 assay
After transfection, the appropriate amounts of resus-
pended cervical cancer cells in the logarithmic phase of
growth were added in 96-well plates with trypsin diges-
tion down and set up 5 sub-wells per group (NC, si-
SFT2D1). When the cells were adhered to the wall, the
solution in Cell Counting Kit-8 (CCK-8, APE, USA) was
added after replacing the fresh medium, and the absor-
bance value was measured at 450 nm after incubation
with the cells for 2 h at 37 ℃ in an MicroplateReader
Instrument (Biotek, USA). First data were grouped into
the 0 h group. And the readings of 0 h, 24 h, 48 h, 72 and
96 h were recorded to calculate the proliferative capacity
of the cells.
Wound scratch experiment
After transfection, resuspended cells were added in
6-well plates by trypsin digestion. e cells incubated
under conditions of constant temperature and constant
CO2, a straight line was drawn vertically in the center of
the 6-well plate with the pipette tip, and then the width of
the straight line was photographed and recorded under
the microscope. Replace the medium with serum-free
medium to continue incubation for 24-48 h, and then
take pictures with the microscope to record the growth
of cells. Cell migration rate = (0 h scratch width - scratch
width after incubation)/0 h scratch width × 100%, which
was analyzed by ImageJ software (1.53, USA) to calculate
the migration ability of cells.
Transwell migration and invasion assay
e chambers were hydrated with serum-free DMEM
medium for 30 min. After aspirating the medium a cell
suspension mixed with appropriate amount of serum-
free medium was added to the upper chamber, and the
lower chamber was incubated with medium containing
10% serum for 24 h. After xation in methanol and stain-
ing with crystal violet, the cells that did not pass through
the upper chamber were wiped away, and the cells that
passed through the lower chamber were observed and
counted under the microscope, and the migration abil-
ity of the cells was judged according to the number of
cells. e cells were observed and counted in the lower
chamber under the microscope. e matrix gel was pur-
chased from Corning (USA). e gel was spread on the
upper chamber surface, and after 2 h, the gel was allowed
to solidify and then hydrated with medium, and the same
procedure was followed to determine the invasion ability
of the cells according to the number of cells.
Statistical analyses
We used R software (version 4.2) and GraphPad prism
(version 8.3.0) for relative analyses and drawings.
T-test was performed to analysis dierences between
two groups. ANOVA was used to analysis dierences
between three or more groups.P < 0.05 was considered as
statistically dierent.
Results
Flow chart
e ow chart was shown in Fig. 1.
Identication of phenotype -related dierent CuRA genes
by WGCNA
19 cuproptosis genes was shown in Additional le 1:
Table S1. en we downloaded angiogenesis-associated
genes with correlation > 1 from the GENECARD website.
Based on the gene expression of TCGA-CESC patients,
as a screening condition of |cor|>0.3 andP < 0.05, nally
533 CuRA genes were included. At the same time, we
performed univariate cox analysis to obtain 66 prognos-
tic CuRA genes (Additional le 1: Figure S1) from 533
genes. To explore genes that are dierentially expressed
between the normal cervix and cervical cancer, we
downloaded 6057 dierential genes from the GEPIA
website(|log2FC|>1). We took the intersection of 6057
dierential genes with the 66 prognostic CuRA genes
(Fig. 2A) above and nally obtained 20 CuRA genes with
signicant dierences (Fig. 2B, Additional le 1: Table
S2). Based on the scores of 20 CuRA gene-sets, each cell
was divided into high-CuRA and low-CuRA cells groups
according to the median value in GSE168652 dataset (Fig.
2C, D). In the TCGA-CESC cohort, we quantied and
visualized the level of immune inltration in dierent
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 5 of 15
Kang et al. Cancer Cell International (2024) 24:23
patients by ssGSEA (Fig. 2E). Meanwhile, we performed
CuRA scores for each patient by the ssGSEA algorithm,
then plotted circle plots to see the dierences in scoring
for each patient (Fig. 2F). WGCNA was performed to
obtain phenotype-related modules for CuRA genes (Fig.
2G). We included all genes (P < 0.05) in non-grey mod-
ules (e grey module contained genes that couldn’t be
classied as any module): pink, brown, magenta, purple,
and yellow modules for the follow-up study (Fig. 2H).
Construction of a CuRA prognostic model
We obtained CuRA intersection genes by intersecting the
dierent genes of high-CuRA and low-CuRA cells groups
of GSE168652 with the WGCNA phenotype-related
modular genes of patients. We congured the alpha
parameter of the elastic network to 0.5 and computed the
errors for primary methods. e results revealed that the
error for ridge regression is 3.055991, for lasso regression
it is 0.0002452, and for elastic network it is 0.0002547035.
After comparing the errors of the regression methods, we
opted for lasso regression to construct the model (Addi-
tional le 1: Figure S2). en a prognostic model was
constructed based on 10 CuRA genes by lasso regres-
sion according to optimal lambda value (Additional le
1: Figure S3A, B). CuRA modeling scores = 0.001007113
*IRF6+ 0.001993273*THBD+ 0.007235823*EFEMP2+ 4.
45E-04*SNX9+ 0.012465781*PCDH18+ 7.23E-05*MFA
P4+ 0.005812479*ADAM9+ 0.004983918*EHBP1+ 0.05
5328306*AVL9+ 0.00770984*SFT2D1. We included the
GSE52903 as validation set. en we analyzed the dier-
ences between TCGA patients (Additional le 1: Figure
S3C) and GSE52903 patients (Additional le 1: Figure
S3D) according to CuRA modeling scores by PCA. Fur-
ther, we assessed ecacy of the model to predict progno-
sis. Heatmaps and point chart of risk scores were drawn
showing dierences in expression of CuRA model genes
Fig. 1 Flow chart of the full text. CuRA, cuproptosis-related angiogenesis gene. DEGs, dierential genes. WGCNA, weighted correlation network analysis.
ROC, receiver operating characteristic
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 6 of 15
Kang et al. Cancer Cell International (2024) 24:23
in TCGA patients (Fig. 3A) and GEO patients (Fig. 3D.
In TCGA dataset (Fig. 3B) and GEO dataset (Fig. 3E),
patients in the high-CuRA group had signicantly lower
survival than those in the low-CuRA group. In the ROC
curves, the 1, 2, 3, and 5-year AUC values were 0.653,
0.759, 0.748, and 0.799 for TCGA patients, respectively
(Fig. 3C). e AUC values for 2, 3, and 5-year survival for
GEO patients were 0.653, 0.646, and 0.604, respectively
(Fig. 3F). Results indicated the model had better predic-
tive eect on prognosis in both TCGA and GEO datasets.
Meanwhile, we included clinical data of TCGA patients
and then combined T1-2 patients into early stage and
T3-4 stage patients into late stage of T-stage for univari-
ate COX analysis, and we found CuRA modeling scores,
N_stage were risk factors (Additional le 1: Figure S3E).
Further, we performed a multivariate COX analysis and
the results showed CuRA modeling scores, N_stage, and
T_stage were independent risk factors (Additional le 1:
Figure S3F).
Immune inltration landscape and mutational landscape
Based on the modeling gene scores, we explored the dif-
ferences in immune inltration and tumor mutations
of high-CuRA and low-CuRA groups. We introduced
Fig. 2 Selection of CuRA phenotype-related genes by WGCNA. (A) Venn diagram of intersection of 6057 dierential genes with 66 prognostic genes. (B)
Histogram of 20 CuRA genes validated at the GEPIA website. (C) Scores of CuRA genes in GSE168652, with the darker the purple, the higher the scores.
(D) Grouping into high-CuRA and low-CuRA groups, red indicating high-CuRA group, blue indicating low-CuRA group. (E) Quantication of immune
inltration levels by ssGSEA. (F) Circle plot of CuRA genes scores of patients. (G) Waterfall plot of WGCNA. (H) Heatmap of phenotype-related modules
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 7 of 15
Kang et al. Cancer Cell International (2024) 24:23
six algorithms to perform a comprehensive analysis of
immune cell inltration in two groups (Additional le
1: Figure S5). e six algorithms showed inltration of
the overall immune cells was signicantly lower in the
high-CuRA group than in the low-CuRA group. Among
them, it showed a signicant dierence between the two
groups by xCell algorithm. Next, we compared the 10
types of major immune cells, and found CD8 + T cells
were signicantly reduced in patients in the high-CuRA
group (Fig. 4A). Also, we analyzed the expression of com-
mon immune checkpoint genes, and we found the clini-
cally common immune checkpoint genesCD274(PD-L1)
and CTLA4 did not dier signicantly between two
groups (Fig. 4B). erefore, we speculate patients clas-
sied into high-CuRA and low-CuRA groups may not
dier in treatment benet by applyingPD-L1inhibitors
orCTLA4 inhibitors. Also, the results showed most of
the immune checkpoint genes were signicantly less
expressed in the high-CuRA group than in the low-
CuRA group. We also found IL10RB, KDR, TGFB1,
and TGFBR1 genes were signicantly more expressed
in the high-CuRA group than in the low-CuRA group.
Results suggested patients in the high-CuRA group may
get better therapeutic outcomes by using inhibitors tar-
geting these 4 genes. Next, we analyzed the mutation
landscape in both groups (Fig. 4C, D), and found the top
3 genes with the highest mutation frequencies were PI
K3CA(31%), TTN(31%),SYNE1(18%) in 127 patients of
the high-CuRA group, while in the low-CuRA group, the
top 3 genes with the highest mutation frequencies were
TTN(33%),PIK3CA(27%),KMT2C(22%) in 113 patients.
We foundSYNE1showed a higher mutation frequency
(23 cases) in the high-CuRA group and a lower muta-
tion frequency (11 cases) in the low-CuRA group with
signicant dierence between the two groups (odds ratio
(OR) = 0.412) (Additional le 1: Figure S6). Also, other
genes such as RELN, SPATA31D1, TCOF1 had higher
mutations in the high-CuRA group.
Analysis of clinical characterization and construction of
nomogram
We analyzed prognosis with dierent clinical character-
istics, and constructed a nomogram based on the CuRA
model. Results showed the N_stage and CuRA model-
ing scores contributed signicantly to the model. e
predicted mortality of the patient was 0.571, 0.98, and
0.997 at 1, 3, and 5-years, respectively (Fig. 5A), and
its odds ratio of status was 8.19 (Fig. 5B). ROC curves
were performed to predict the accuracy of nomogram.
AUC values at 1, 3, and 5-years were 0.71, 0.78, and
0.83, respectively, which indicated the nomogram had
good predictive accuracy (Fig. 5C). Finally, we plotted
Fig. 3 Construction and verication of a CuRA prognostic model. Heatmap and point chart of TCGA patients (A) and GEO patients (D). Survival curve plot
of TCGA patients (B) and GEO patients (E). ROC curve of TCGA patients (C) and GEO patients (F)
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 8 of 15
Kang et al. Cancer Cell International (2024) 24:23
the DCA decision curves, and results showed the model
had a strong predictive eect on the survival rate at 1, 3,
and5-years (Fig. 5D).
Drug sensitivity prediction
To explore the dierent drugs that may be eective
against treating the high- and low-CuRA groups, we used
GDSC database by predicting the IC50 to determine the
dierences in drug sensitivity between the high- and
low-CuRA groups. Results showed the IC50 of Imatinib,
Pazopanib, and Sorafenib was signicantly lower for the
high-CuRA group than for the low-CuRA group, suggest-
ing they may have better ecacy when applied with the
high-CuRA patient group (Fig. 6A–C). Similarly, for the
low-CuRA group, the application of AMG.706, CEP.701,
Sunitinib, ABT.888 (Veliparib), AZD.2281 (Olaparib),
and MS.275 (Entinostat) may lead to better therapeutic
remission (Fig. 6D–I).
Cell communication analysis
We performed cell communication analysis by single-cell
sequencing dataset GSE168652 from GEO. We grouped
the cells into 25 clusters and divided the cell clusters into
8 types based on annotation, which are: endothelial cells,
FDX1 + tumor/epithelial cells, broblasts, lymphocytes,
macrophages, smooth muscle cells, tumor/epithelial cells
(other types), and VEGFA + tumor/epithelial cells (Fig.
7A). We observed the expression and localization of 10
modeling genes in GSE168652 (Additional le 1: Figure
S4). SinceFDX1was a representative gene for cupropto-
sis, we labeled FDX1-positive or VEGFA-positive tumor/
epithelial cells here in the hope of exploring the relation-
ship between cuproptosis and angiogenesis-associated
tumor/epithelial cells in intercellular communication.
Multidirectional cell communication was discovered in
each cell subpopulation (Fig. 7B). en we identied the
cell-extrinsic communication patterns. We analyzed the
signaling pathways of both incoming and outgoing sig-
nals in the samples. Our results revealed the main outgo-
ing signals of FDX1 + tumor/epithelial cells were PDGF,
WNT, CD46, MHC-1, MIF, and MK pathways, while the
primary incoming signals were IFN-II and other path-
ways. As for VEGFA + tumor/epithelial cells, the major
outgoing signals were WNT, EGF, and VEGF, while the
main incoming signals involved numerous signaling
pathways, including COLLAGEN (Fig. 7C). Specically,
Fig. 4 Immune inltration and tumor mutation analysis. (A) Dierences in inltration of 10 types of immune cells between two CuRA groups. (B) Dif-
ferences in expression of immune checkpoints gene between the two groups. (C) Tumor mutation characteristics in the high-CuRA group. (D) Tumor
mutation characteristics in the low-CuRA group
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 9 of 15
Kang et al. Cancer Cell International (2024) 24:23
we focused on the involvement of VEGF pathway in cell
communication as the key pathway of angiogenesis. We
observed higher expression of the key gene VEGFA in the
VEGFA + tumors/epithelial cells, tumor/epithelial cells,
and FDX1 + tumors/epithelial cells in the VEGF signaling
pathway (Fig. 7D). In signal transduction, by calculating
the network centrality indices for each cell population,
we found VEGFA + tumor/epithelial cells were the domi-
nant signaler in the intercellular communication net-
work, endothelial cells were the main receivers and
inuencer, and FDX1 + tumors/epithelial cells also played
an important role in inuencer (Fig. 7E). In the visual cir-
cular and hierarchical plots of the VEGF signaling path-
way (Additional le 1: Figure S9), our analysis revealed
VEGFA + tumor/epithelial cells and FDX1 + tumor/epi-
thelial cells had the most signicant eect on endothelial
cells. In addition, we analyzed relevant receptor-ligand
pairs in FDX1 + or VEGFA + tumors/epithelial cells com-
municating with other cells respectively (Additional le
1: Figure S10).
Critical functional role of SFT2D1 in cervical cancer
We selected prognosis-related model genes
among 10 model genes (Additional le 1: Figure
S7).ADAM9,EHBP1, andSFT2D1gene were shown sig-
nicantly aecting the survival of patients. Meanwhile,
we also identied SFT2D1 as an independent risk fac-
tor by multivariate analysis along with clinical features
(Additional le 1: Table S3). en we performed GSEA
of SFT2D1, and results showed SFT2D1 was mainly
involved in the regulation of autophagy, glycosaminogly-
can degradation, RNA degradation, riboavin metabo-
lism, mTOR signaling pathway. We then investigate the
eect of SFT2D1 on immune microenvironment. e
total scores, stromal scores and immune scores were
signicantly lower in the high-SFT2D1 group than in
the low-SFT2D1group, and the immune related scores
were negatively correlated with SFT2D1 expression
(Additional le 1: Figure S8). Meanwhile, we performed
immunohistochemical analysis of SFT2D1 and the neo-
vascularization marker CD31 in paran sections of cer-
vical cancer patients. e results showed both SFT2D1
and CD31 were expressed up-regulated in cervical can-
cer tissues (Fig. 8A). Correlation analysis also showed
Fig. 5 Construction and validation of nomogram. (A) Cox regression to construct nomogram. (B) Logistic regression to construct nomogram. (C) Curves
of ROC for 1, 3, and 5-years. (D) DCA decision curves for 1, 3, and 5-years
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 10 of 15
Kang et al. Cancer Cell International (2024) 24:23
positive correlation in SFT2D1 and CD31 (Fig. 8B). en
we performed RT-qPCR to verify SFT2D1 was highly
expressed in cervical cancer cell lines compared to nor-
mal cervical cell lines ECT1/E6E7 (Fig. 8C). Meanwhile,
western blotting showed that SFT2D1 was upregulated
in SiHa, CaSki cervical cell lines (Fig. 8D). Since the role
ofSFT2D1in cervical cancer has not yet been explored,
we authenticated the eect ofSFT2D1on the function of
SiHa and CaSki cells by in vitro experiments. RT-qRCR
showed that four siRNAs signicantly suppressed the
expression ofSFT2D1in transfected SiHa and CaSki cells
(Fig. 8E, F). We selected the two siRNAs with the high-
est knockdown eciency among them: si-SFT2D1-2,
si-SFT2D1-3 for subsequent experiments. CCK-8 analy-
sis showed that knockdown of SFT2D1 signicantly
inhibited the proliferative ability of SiHa and CaSki cells
(Fig. 8G, H). Wound scratch assay showed that knock-
down ofSFT2D1signicantly inhibited the migration of
cervical cancer cells (Fig. 8I, J). Transwell assay showed
that knockdown of SFT2D1 signicantly inhibited the
migratory and invasive abilities of SiHa and CaSki cells
(Fig. 8K, L).
Fig. 6 Drug sensitivity analysis of the high and low-CuRA group.(A-I)The IC50 of Imatinib, Pazopanib, and Sorafenib was lower in the high-CuRA group
than for the low-CuRA group contrary to AMG.706, CEP.701, Sunitinib, ABT.888, AZD.2281, and MS.275. The vertical coordinates displayed as drug names,
demonstration of drugs with statistical signicance
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 11 of 15
Kang et al. Cancer Cell International (2024) 24:23
Discussion
In this study, we developed and validated a prognostic
model based on CuRA genes in cervical cancer. We also
analyzed clinical characteristics and the immune micro-
environment between the high and low-CuRA patient
groups. Based on the modeling scores, we analyzed drug
sensitivity of the high or low-CuRA patient group to pro-
vide guidance on drug administration.SFT2D1, as a key
gene involved in the progression of cervical cancer, it was
associated with the cuproptosis-dependent angiogenesis
pathway.
Cuproptosis and tumor angiogenesis are closely linked
in the tumor microenvironment. CuRA genes may help
explain the potential link between cuproptosis and angio-
genesis, which could improve the prognosis of cervi-
cal cancer patients. We constructed and validated our
model based on CuRA signatures using patient data from
TCGA and GEO. Several web tools enable us to extract
prognostic variable characteristics from multi-omics data
by selecting clinical variables or subgroup variables (lasso,
elastic network regularization, and network regularized
high-dimensional Cox regression) [44]. is implies the
necessity of choosing the optimal regression method for
subsequent studies. After comparing the errors of these
three regression methods, we observed that lasso regres-
sion minimized the error. Consequently, we have opted
for the lasso regression method to construct the CuRA
model. e AUC values for 1, 2, 3, and 5-year survival
in TCGA patients were 0.653, 0.759, 0.748, and 0.799,
respectively, while the AUC values for 2, 3, and 5-year
survival in GEO patients were 0.653, 0.646, and 0.604,
respectively. e shorter survival time of patients in the
high CuRA group may indicate that tumor cells pro-
mote tumor progression through cuproptosis-associated
angiogenesis. In the analysis of clinical characteristics,
CuRA modeling scores, N_stage, and T_stage were inde-
pendent risk factors, suggesting that modeling scores
Fig. 7 Analysis of cellular communication network. (A) Cell annotation of 8 types of cells. (B) Number of interactions and interaction weights of samples
in GSE168652. (C) Schematic diagram of the incoming and outgoing signals of samples. (D) Visualization of the expression of key genes in the VEGF signal-
ing pathway in 8 types of cells. (E) Visualization of cells involved as senders, receivers, mediators and inuencers in the VEGF pathway
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 12 of 15
Kang et al. Cancer Cell International (2024) 24:23
could independently contribute to cervical cancer pro-
gression as a risk factor. Our CuRA model is represented
by 10 genes involved in programmed death-related path-
ways, membrane vesicle transport, and tumorigenesis
and progression. Among them, SFT2 Domain Containing
1 (SFT2D1) is involved in protein and vesicle-mediated
translocation and is also associated with poor survival in
patients with high-risk neuroblastoma [45]. GSEA path-
way analysis results suggest thatSFT2D1plays an impor-
tant role in tumor-related pathways and is associated
with the invasion and progression of cervical cancer. e
immune microenvironment was scored forSFT2D1, and
patients with highSFT2D1expression had a higher CuRA
score and a worse prognosis, showing a more aggressive
immunosuppressive phenotype. We veriedSFT2D1was
signicantly upregulated in cervical cancer cells by west-
ern blotting, RT-qPCR, and immunohistochemistry.
erefore,SFT2D1, a CuRA modeling gene, may serve as
a marker gene and provide a new reference for the treat-
ment of cervical cancer patients.
Antitumor strategies targeting angiogenesis have been
used in the clinical management of patients with meta-
static or recurrent cervical cancer. However, improve-
ments in overall survival (OS) and progression-free
survival (PFS) times for patients are still limited. Com-
bining immune checkpoint inhibitors (ICIs) with
vascular targeting therapy has demonstrated synergis-
tic sensitization in the treatment of various tumors,
Fig. 8 SFT2D1was involved in the progression of cervical cancer. (A) SFT2D1 and CD31 were highly expressed in cervical cancer by IHC. (B) Correlation
analysis between SFT2D1 and CD31. (C) Validation ofSFT2D1expression by RT-qPCR. (D) Validation of SFT2D1 expression by western blotting. Four siRNAs
suppressed the expression ofSFT2D1in SiHa (E)and CaSki (F) cells by RT-qPCR. CCK-8 assay in in SiHa (G) and CaSki (H) cells. Wound scratch assay in SiHa
(I) and CaSki (J) cells. Transwell assay in SiHa (K) and CaSki (L) cells
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 13 of 15
Kang et al. Cancer Cell International (2024) 24:23
including hepatocellular carcinoma [46, 47], Non-small
cell lung cancer [48], gastric cancer [49]. In cervical
cancer, a phase III randomized controlled trial showed
increased overall survival in patients with recurrent or
metastatic cervical cancer after treatment with pabli-
zumab combined with chemotherapy and bevacizumab
[50]. erefore, researching the dierences in immune
microenvironment in cervical cancer patients with dier-
ent CuRA scores can help promote the precise combina-
tion of immunotherapy and vascular targeting therapy,
and enable personalized treatment selection of immune
checkpoint inhibitors for patients with CuRA-related
cervical cancer. Results showed that the group of patients
with high-CuRA scores exhibited immunosuppression,
while patients with low-CuRA scores may have a more
signicant therapeutic eect with immune agents target-
ing CD8 + T cells compared to patients in the high-CuRA
scores group. Patients with high-CuRA scores had lower
expression in most immune checkpoint genes. However,
a minority of immune checkpoint genes presented high
expression in the high-CuRA group, which suggests that
treatment with anti-IL10RB, anti-KDR, anti-TGFB1, and
anti-TGFBR1 may be considered to improve prognosis of
patients in the high-CuRA group.
Cuproptosis plays an important role in tumor cell pro-
liferation and angiogenesis. Cell-cell communication
analysis based on single-cell sequencing helps to reveal
the tumor immune microenvironment and changes
in the tumor itself [51]. Previous studies have shown
thatFDX1, as a key gene in cuproptosis, is involved in the
progression of hepatocellular carcinoma [52] and glioma
[53] as well as tumor immunity and drug sensitivity [54].
To investigate the relationship between angiogenesis and
cuproptosis in cervical cancer, we identied two types of
cells, FDX1 + tumor/epithelial cells and VEGFA + tumor/
epithelial cells, and conducted cell-cell communica-
tion analysis. We found that the genesVEGFAandPGF,
which promote angiogenesis, are highly expressed in
these two types of cells. Meanwhile, both types of cells
are sent as signals to endothelial cells, indicating that
they can aect the process of transendothelial migra-
tion and promote angiogenesis. In addition, both types
of cells can transmit cell signals to macrophages to aect
the macrophage migration inhibitory factor (MIF) path-
way. Previous studies have shown that MIF is involved in
multiple immune processes and mediates immune escape
leading to tumor metastasis [55–58]. Our results suggest
that VEGFA + tumor/epithelial cells and FDX1 + tumor/
epithelial cells may play irreplaceable roles in the tumor
immune microenvironment.
We conducted drug sensitivity analysis for dierent
CuRA groups of patients, which may lay the foundation
for personalized treatment. e GDSC provides informa-
tion on the sensitivity of tumor cell lines to drugs, with
a smaller IC50 indicating greater sensitivity of the bio-
logical system to the compound. Our analysis showed
that among small molecule tyrosine kinase inhibitors,
imatinib, pazopanib, and sorafenib may provide bet-
ter ecacy for patients with high CuRA scores, while
AMG.706, CEP.701, and sunitinib may provide better
ecacy for patients with lower CuRA scores. Receptor
tyrosine kinase inhibitors have signicant anti-angiogenic
eects and have made great progress in the treatment of
gynecological tumors [59]. For example, pazopanib has
improved PFS and OS in patients with advanced or recur-
rent cervical cancer, while AMG.706 (motesanib) inhibits
angiogenesis in recurrent ovarian cancer and CEP.701
(lestaurtinib) inhibits the growth of cervical cancer cells
[60–62]. Studies have shown dierences in IC50 values
of drugs such as pazopanib, sorafenib, sunitinib, and
imatinib among dierent CuRA subgroups of clear cell
renal cell carcinoma, bladder cancer, and triple-negative
breast cancer [63–65]. Our study also suggested that in
the high CuRA score group, IC50 values of PARP inhibi-
tors ABT.888 (Veliparib), AZD.2281 (Olaparib), and
MS.275 (Entinostat) were higher, indicating that these
drugs were eective bythe cuproptosis-dependent angio-
genesis pathway ese studies supported our results of
drug sensitivity analysis and suggested that research on
anti-cuproptosis-related angiogenesis targeted drugs may
bring new treatment ideas and expand the application of
drugs in the current dilemma.
In conclusion, we have conducted a novel cuproptosis-
related angiogenesis (CuRA) gene signature using single-
cell RNA sequencing and bulk RNA sequencing data,
which provides signicant predictive value for patients
with cervical cancer. Of course, there are limitations to
our results, and future relevant mechanisms need to be
further validated in in vivo and in vitro experiments.
Supplementary Information
The online version contains supplementary material available at https://doi.
org/10.1186/s12935-023-03189-x.
Supplementary Material 1:Figure S1: Forest plot of 66 CuRA prognostic
genes by univariate COX analysis. Figure S2: Comparison of errors in ridge
regression, lasso regression and elastic network regression and compre-
hensive comparison of the errors in three regression algorithms.Figure
S3: Construction and verication of a CuRA prognostic model. Figure
S4: Localization and validation of 10 modeling genes in the GSE168652
dataset. Figure S5: Immune inltration landscape in high and low CuRA-
groups calculated by CIBERSORT, EPIC, MCP, Quanti-seq, TIMER, xCell
algorithms. Figure S6: Dierences in mutation frequency between high
and low CuRA groups. Figure S7: Survival analysis of 10 modeling genes.
Figure S8: Assessment of regulatory pathways and immune microen-
vironment of SFT2D1. Figure S9: Visual circular and hierarchical plots
showing cellular communication in the VEGF pathway. Figure S10: Cell
Communication Analysis
Supplementary Material 2:Table S1: Genelist of Cuproptosis-Related
Gene.Table S2: Genelist of Cuproptosis-Related Angiogenesis Gene
(CuRA).Table S3: Univariate Analysis and Multivariate Analysis of SFT2D1
and Clinical Characteristics.Table S4: Prediction of small molecule drugs
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 14 of 15
Kang et al. Cancer Cell International (2024) 24:23
by modeling genes.Table S5: GSEA to assess SFT2D1-related path-
ways.Table S6: Oligonucleotides used in research
Acknowledgements
Not applicable.
Author contributions
Conceptualization, Yong Zhang and Lesai Li; Data curation, Jiawen
Kang; Investigation, Jingwen Jiang; Methodology, Yong Zhang; Project
administration, Jie Tang; Validation, Xiaoqing Xiang; Writing – original draft,
Jiawen Kang; Writing – review & editing, Lesai Li.
Funding
This research was funded by Natural Science Foundation of Hunan Province
(Grant. 2022JJ30415), Hunan Cancer Hospital Climb Plan (ZX2020004-3).
Data availability
The data and materials in the current study are available under the permission
of author.
Declarations
Ethics approval and consent to participate
The study was conducted in accordance with the Declaration of Helsinki, and
approved by the Ethics Committee of Hunan Cancer Hospital (No. SBQLL-
2021-289). All patients gave informed consent.
Consent for publication
All authors of this paper consent for publishing the manuscript and gures in
the journal.
Competing interests
The authors declare no competing interests.
Received: 30 September 2023 / Accepted: 23 December 2023
References
1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray
F. Global Cancer statistics 2020: GLOBOCAN estimates of incidence and
Mortality Worldwide for 36 cancers in 185 countries. CA Cancer J Clin.
2021;71(3):209–49.
2. Aviki EM, Chen L, Dessources K, Leitao MM Jr., Wright JD. Impact of hospital
volume on surgical management and outcomes for early-stage Cervical
cancer. Gynecol Oncol. 2020;157(2):508–13.
3. Abu-Rustum NR, Yashar CM, Bean S, Bradley K, Campos SM, Chon HS, Chu C,
Cohn D, Crispens MA, Damast S, et al. NCCN guidelines insights: Cervical Can-
cer, Version 1.2020. J Natl Compr Cancer Network: JNCCN. 2020;18(6):660–6.
4. Salehiniya H, Momenimovahed Z, Allahqoli L, Momenimovahed S, Alkatout
I. Factors related to Cervical cancer screening among Asian women. Eur Rev
Med Pharmacol Sci. 2021;25(19):6109–22.
5. Adiga D, Eswaran S, Pandey D, Sharan K, Kabekkodu SP. Molecular landscape
of recurrent Cervical cancer. Crit Rev Oncol/Hematol. 2021;157:103178.
6. Mutlu L, Tymon-Rosario J, Harold J, Menderes G. Targeted treatment options
for the management of metastatic/persistent and recurrent Cervical cancer.
Expert Rev Anticancer Ther. 2022;22(6):633–45.
7. Li H, Wu X, Cheng X. Advances in diagnosis and treatment of metastatic
Cervical cancer. J Gynecologic Oncol. 2016;27(4):e43.
8. Su Z, Yang Z, Xu Y, Chen Y, Yu Q. Apoptosis, autophagy, necroptosis, and
cancer Metastasis. Mol Cancer. 2015;14:48.
9. Viallard C, Larrivée B. Tumor angiogenesis and vascular normalization: alterna-
tive therapeutic targets. Angiogenesis. 2017;20(4):409–26.
10. Albini A, Bruno A, Noonan DM, Mortara L. Contribution to Tumor Angiogen-
esis from Innate Immune cells within the Tumor Microenvironment: implica-
tions for Immunotherapy. Front Immunol. 2018;9:527.
11. Tewari KS, Sill MW, Long HJ 3rd, Penson RT, Huang H, Ramondetta LM,
Landrum LM, Oaknin A, Reid TJ, Leitao MM, et al. Improved survival with
bevacizumab in advanced Cervical cancer. N Engl J Med. 2014;370(8):734–43.
12. Carmeliet P, Jain RK. Principles and mechanisms of vessel normaliza-
tion for cancer and other angiogenic Diseases. Nat Rev Drug Discovery.
2011;10(6):417–27.
13. Alameddine RS, Yakan AS, Skouri H, Mukherji D, Temraz S, Shamseddine A.
Cardiac and vascular toxicities of angiogenesis inhibitors: the other side of
the coin. Crit Rev Oncol/Hematol. 2015;96(2):195–205.
14. Chitsike L, Duerksen-Hughes P. The potential of Immune Checkpoint Block-
ade in Cervical Cancer: can combinatorial regimens maximize response? A
review of the literature. Curr Treat Options Oncol. 2020;21(12):95.
15. Mortara L, Benest AV, Derosa L, Chouaib S, Ribatti D. Editorial: the intricate
innate immune-cancer cell relationship in the context of Tumor angio-
genesis, immunity and microbiota: the angiogenic switch in the Tumor
microenvironment as a key target for immunotherapies. Front Immunol.
2022;13:1045074.
16. Gong C, Bauvy C, Tonelli G, Yue W, Deloménie C, Nicolas V, Zhu Y, Domergue
V, Marin-Esteban V, Tharinger H, et al. Beclin 1 and autophagy are required
for the tumorigenicity of Breast cancer stem-like/progenitor cells. Oncogene.
2013;32(18):2261–72. 2272e.2261 – 2211.
17. Tisch N, Freire-Valls A, Yerbes R, Paredes I, La Porta S, Wang X, Martín-Pérez
R, Castro L, Wong WW, Coultas L, et al. Caspase-8 modulates physiological
and pathological angiogenesis during retina development. J Clin Investig.
2019;129(12):5092–107.
18. Zhang F, Li Y, Tang Z, Kumar A, Lee C, Zhang L, Zhu C, Klotzsche-von Ameln
A, Wang B, Gao Z, et al. Proliferative and survival eects of PUMA promote
angiogenesis. Cell Rep. 2012;2(5):1272–85.
19. Tisch N, Ruiz de Almodóvar C. Contribution of cell death signaling to blood
vessel formation. Cell Mol Life Sci. 2021;78(7):3247–64.
20. Yang L, Joseph S, Sun T, Homann J, Thevissen S, Oermanns S, Strilic B. TAK1
regulates endothelial cell necroptosis and Tumor Metastasis. Cell Death Dier.
2019;26(10):1987–97.
21. Tsvetkov P, Coy S, Petrova B, Dreishpoon M, Verma A, Abdusamad M,
Rossen J, Joesch-Cohen L, Humeidi R, Spangler RD, et al. Copper induces
cell death by targeting lipoylated TCA cycle proteins. Sci (New York NY).
2022;375(6586):1254–61.
22. Lelièvre P, Sancey L, Coll JL, Deniaud A, Busser B. The multifaceted roles of
copper in Cancer: a Trace Metal element with Dysregulated Metabolism, but
also a target or a bullet for Therapy. Cancers 2020, 12(12).
23. Das A, Ash D, Fouda AY, Sudhahar V, Kim YM, Hou Y, Hudson FZ, Stans-
eld BK, Caldwell RB, McMenamin M, et al. Cysteine oxidation of copper
transporter CTR1 drives VEGFR2 signalling and angiogenesis. Nat Cell Biol.
2022;24(1):35–50.
24. Chi H, Peng G, Wang R, Yang F, Xie X, Zhang J, Xu K, Gu T, Yang X, Tian G.
Cuprotosis Programmed-Cell-Death-Related lncRNA Signature Predicts
Prognosis and Immune Landscape in PAAD Patients. Cells 2022, 11(21).
25. Ruiz LM, Libedinsky A, Elorza AA. Role of copper on mitochondrial function
and metabolism. Front Mol Biosci. 2021;8:711227.
26. Li Y. Copper homeostasis: emerging target for cancer treatment. IUBMB Life.
2020;72(9):1900–8.
27. Naito H, Iba T, Wakabayashi T, Tai-Nagara I, Suehiro JI, Jia W, Eino D, Sakimoto
S, Muramatsu F, Kidoya H, et al. TAK1 prevents endothelial apoptosis and
maintains Vascular Integrity. Dev Cell. 2019;48(2):151–166e157.
28. Li X, Wang CY. From bulk, single-cell to spatial RNA sequencing. Int J Oral Sci.
2021;13(1):36.
29. Hwang B, Lee JH, Bang D. Single-cell RNA sequencing technologies and
bioinformatics pipelines. Exp Mol Med. 2018;50(8):1–14.
30. Sun D, Guan X, Moran AE, Wu LY, Qian DZ, Schedin P, Dai MS, Danilov AV,
Alumkal JJ, Adey AC, et al. Identifying phenotype-associated subpopula-
tions by integrating bulk and single-cell sequencing data. Nat Biotechnol.
2022;40(4):527–38.
31. Li X, Sun Z, Peng G, Xiao Y, Guo J, Wu B, Li X, Zhou W, Li J, Li Z, et al. Single-cell
RNA sequencing reveals a pro-invasive cancer-associated broblast sub-
group associated with poor clinical outcomes in patients with gastric cancer.
Theranostics. 2022;12(2):620–38.
32. Lei Y, Tang R, Xu J, Wang W, Zhang B, Liu J, Yu X, Shi S. Applications of single-
cell sequencing in cancer research: progress and perspectives. J Hematol
Oncol. 2021;14(1):91.
33. Kim C, Gao R, Sei E, Brandt R, Hartman J, Hatschek T, Crosetto N, Foukakis
T, Navin NE. Chemoresistance Evolution in Triple-negative Breast Cancer
delineated by single-cell sequencing. Cell. 2018;173(4):879–893e813.
34. Jin S, Guerrero-Juarez CF, Zhang L, Chang I, Ramos R, Kuan CH, Myung P,
Plikus MV, Nie Q. Inference and analysis of cell-cell communication using
CellChat. Nat Commun. 2021;12(1):1088.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 15 of 15
Kang et al. Cancer Cell International (2024) 24:23
35. Mangiola S, Doyle MA, Papenfuss AT. Interfacing Seurat with the R tidy
universe. Bioinf (Oxford England). 2021;37(22):4100–7.
36. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, Smyth GK. Limma powers
dierential expression analyses for RNA-sequencing and microarray studies.
Nucleic Acids Res. 2015;43(7):e47.
37. Langfelder P, Horvath S. WGCNA: an R package for weighted correlation
network analysis. BMC Bioinformatics. 2008;9:559.
38. Engebretsen S, Bohlin J. Statistical predictions with glmnet. Clin Epigenetics.
2019;11(1):123.
39. Wang QW, Lin WW, Zhu YJ. Comprehensive analysis of a TNF family based-sig-
nature in diuse gliomas with regard to prognosis and immune signicance.
Cell Communication and Signaling: CCS. 2022;20(1):6.
40. Huang R, Chen Z, Li W, Fan C, Liu J. Immune system–associated genes
increase malignant progression and can be used to predict clinical outcome
in patients with hepatocellular carcinoma. Int J Oncol. 2020;56(5):1199–211.
41. Sun D, Zhu Y, Zhao H, Bian T, Li T, Liu K, Feng L, Li H, Hou H. Loss of ARID1A
expression promotes lung adenocarcinoma Metastasis and predicts a poor
prognosis. Cell Oncol (Dordrecht). 2021;44(5):1019–34.
42. Zeng D, Ye Z, Shen R, Yu G, Wu J, Xiong Y, Zhou R, Qiu W, Huang N, Sun L, et al.
IOBR: Multi-omics Immuno-Oncology Biological Research to Decode Tumor
Microenvironment and signatures. Front Immunol. 2021;12:687975.
43. Mayakonda A, Lin DC, Assenov Y, Plass C, Koeer HP. Maftools: ecient
and comprehensive analysis of somatic variants in cancer. Genome Res.
2018;28(11):1747–56.
44. Pak K, Oh SO, Goh TS, Heo HJ, Han ME, Jeong DC, Lee CS, Sun H, Kang J, Choi
S, et al. A User-Friendly, web-based Integrative Tool (ESurv) for Survival Analy-
sis: Development and Validation Study. J Med Internet Res. 2020;22(5):e16084.
45. Ognibene M, Morini M, Garaventa A, Podestà M, Pezzolo A. Identication of a
minimal region of loss on chromosome 6q27 associated with poor survival of
high-risk neuroblastoma patients. Cancer Biol Ther. 2020;21(5):391–9.
46. Finn RS, Qin S, Ikeda M, Galle PR, Ducreux M, Kim TY, Kudo M, Breder V, Merle
P, Kaseb AO, et al. Atezolizumab plus Bevacizumab in Unresectable Hepato-
cellular Carcinoma. N Engl J Med. 2020;382(20):1894–905.
47. Ren Z, Xu J, Bai Y, Xu A, Cang S, Du C, Li Q, Lu Y, Chen Y, Guo Y, et al. Sintilimab
plus a bevacizumab biosimilar (IBI305) versus sorafenib in unresectable
hepatocellular carcinoma (ORIENT-32): a randomised, open-label, phase 2–3
study. Lancet Oncol. 2021;22(7):977–90.
48. Reckamp KL, Redman MW, Dragnev KH, Minichiello K, Villaruz LC, Faller B,
Al Baghdadi T, Hines S, Everhart L, Highleyman L, et al. Phase II random-
ized study of Ramucirumab and Pembrolizumab Versus Standard of Care
in Advanced Non-small-cell Lung Cancer previously treated with immuno-
therapy-Lung-MAP S1800A. J Clin Oncology: Ocial J Am Soc Clin Oncol.
2022;40(21):2295–306.
49. Song X, Qi W, Guo J, Sun L, Ding A, Zhao G, Li H, Qiu W, Lv J. Immune check-
point inhibitor combination therapy for gastric cancer: Research progress.
Oncol Lett. 2020;20(4):46.
50. Schmidt MW, Battista MJ, Schmidt M, Garcia M, Siepmann T, Hasenburg A,
Anic K. Ecacy and Safety of Immunotherapy for Cervical Cancer-A System-
atic Review of clinical trials. Cancers 2022, 14(2).
51. Liu Y, Shou Y, Zhu R, Qiu Z, Zhang Q, Xu J. Construction and validation of a
Ferroptosis-Related Prognostic Signature for Melanoma based on single-cell
RNA sequencing. Front cell Dev Biology. 2022;10:818457.
52. Zhang Z, Zeng X, Wu Y, Liu Y, Zhang X, Song Z. Cuproptosis-related risk score
predicts prognosis and characterizes the Tumor Microenvironment in Hepa-
tocellular Carcinoma. Front Immunol. 2022;13:925618.
53. Lu H, Zhou L, Zhang B, Xie Y, Yang H, Wang Z. Cuproptosis key gene FDX1 is
a prognostic biomarker and associated with immune inltration in glioma.
Front Med. 2022;9:939776.
54. Yang L, Zhang Y, Wang Y, Jiang P, Liu F, Feng N. Ferredoxin 1 is a cuproptosis-
key gene responsible for Tumor immunity and drug sensitivity: a pan-cancer
analysis. Front Pharmacol. 2022;13:938134.
55. Sumaiya K, Langford D, Natarajaseenivasan K, Shanmughapriya S. Macro-
phage migration inhibitory factor (MIF): a multifaceted cytokine regulated by
genetic and physiological strategies. Pharmacol Ther. 2022;233:108024.
56. Noe JT, Mitchell RA. MIF-Dependent Control of Tumor Immunity. Front Immu-
nol. 2020;11:609948.
57. Rendon BE, Willer SS, Zundel W, Mitchell RA. Mechanisms of macrophage
migration inhibitory factor (MIF)-dependent Tumor microenvironmental
adaptation. Exp Mol Pathol. 2009;86(3):180–5.
58. Nishihira J, Ishibashi T, Fukushima T, Sun B, Sato Y, Todo S. Macrophage migra-
tion inhibitory factor (MIF): its potential role in Tumor growth and tumor-
associated angiogenesis. Ann N Y Acad Sci. 2003;995:171–82.
59. Qin S, Li A, Yi M, Yu S, Zhang M, Wu K. Recent advances on anti-angiogenesis
receptor tyrosine kinase inhibitors in cancer therapy. J Hematol Oncol.
2019;12(1):27.
60. Wei XW, Zhang ZR, Wei YQ. Anti-angiogenic Drugs currently in phase II
clinical trials for gynecological cancer treatment. Expert Opin Investig Drugs.
2013;22(9):1181–92.
61. Monk BJ, Mas Lopez L, Zarba JJ, Oaknin A, Tarpin C, Termrungruanglert W,
Alber JA, Ding J, Stutts MW, Pandite LN. Phase II, open-label study of pazo-
panib or lapatinib monotherapy compared with pazopanib plus lapatinib
combination therapy in patients with advanced and recurrent Cervical
cancer. J Clin Oncology: Ocial J Am Soc Clin Oncol. 2010;28(22):3562–9.
62. Miller SC, Huang R, Sak amuru S, Shukla SJ, Attene-Ramos MS, Shinn P, Van
Leer D, Leister W, Austin CP, Xia M. Identication of known Drugs that act as
inhibitors of NF-kappaB signaling and their mechanism of action. Biochem
Pharmacol. 2010;79(9):1272–80.
63. Shen J, Wang L, Bi J. Bioinformatics analysis and experimental validation of
cuproptosis-related lncRNA LINC02154 in clear cell renal cell carcinoma. BMC
Cancer. 2023;23(1):160.
64. Song Q, Zhou R, Shu F, Fu W. Cuproptosis scoring system to predict the
clinical outcome and immune response in Bladder cancer. Front Immunol.
2022;13:958368.
65. Sha S, Si L, Wu X, Chen Y, Xiong H, Xu Y, Liu W, Mei H, Wang T, Li M. Prognostic
analysis of cuproptosis-related gene in triple-negative Breast cancer. Front
Immunol. 2022;13:922780.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional aliations.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com
