CelPress
Heliyon
journal homepage: www.cell.com/heliyon
Heliyon
H
ColPress
Research article
The prognostic value and immunological role of SULF2 in adrenocortical carcinoma
Jiusong Yan ª,1, Xiaodu Xie ª,1, Qinke Lib, Peihe Lianga, Junyong Zhang a,"", Guangyong Xu ª,
a Department of Urology, The Second Affiliated Hospital of Chongqing Medical University, Chongqing 400010, China
b Department of Gynecology and Obstetrics, The Second Affiliated Hospital of Chongqing Medical University, Chongqing 400010, China
Check for updates
ARTICLE INFO
Keywords:
Adrenocortical carcinoma (ACC) Sulfatase2 (SULF2) Prognosis Tumor immune regulation
ABSTRACT
Background: Adrenocortical carcinoma (ACC) represents the rare urological epithelial cancer of urinary tract, which has a large mass and is usually diagnosed at the advanced stage, thus inducing the poor prognosis. As a result, early detection and diagnosis are more important for the prognosis rather than the treatment of ACC. There is evidence supporting the association of Sulfatase2 (SULF2) with bladder cancer. However, the relationships of SULF2 with the clinical features and immune infiltration of ACC remain unclear.
Methods: This work comprehensively investigated the different expression levels of SULF2 within ACC and its prognostic significance through various databases including Gene Expression Profiling Interaction Analysis (GEPIA), Tumor Immune Estimation Resource (TIMER), The Cancer Genome Atlas (TCGA), Gene Expression Omnibus (GEO), Kaplan-Meier (KM) plotter and UAL- CAN. Besides, SULF2 levels within different tumor and paraneoplastic tissues were examined based on Human Protein Atlas (HPA) and TIMER. Afterwards, this study identified differentially expressed genes (DEGs) in high-compared with low-SULF2-expression groups. To predict the possible interaction between SULF2 and its targets, a protein-protein interaction (PPI) network was constructed based on relevant data collected in STRING database. Besides, the SULF2 func- tional annotation was carried out, including Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) and GSEA. In addition, gene mutation analysis was also performed based on the cBioPortal database. The relation of SULF2 with immune infiltration was analyzed from various aspects by using the resources of various databases including TIMER, TISIDB, and GEPIA, which was first reported in this work. Finally, R package was utilized to plot the receiver oper- ating characteristic (ROC) curves of diagnosis, time-dependent survival, and the association of SULF2 with cancer stage and the nomogram model. Finally, CellMiner dataset was adopted for SULF2 correlation as well as drug sensitivity analysis.
Results: Relative to healthy people, SULF2 level markedly elevated within ACC tissues. Besides, SULF2 up-regulation significantly predicted the dismal prognostic outcome, which may be an important prognostic factor. Afterwards, the PPI network was constructed, and the possible link of SULF2 with the corresponding targets was predicted. Besides, up-regulated SULF2 expression was tightly related to immune regulation and tumor-infiltration immune cell (TIICs), including
* Corresponding author.
** Corresponding author.
E-mail addresses: zhangmachine@cqmu.edu.cn (J. Zhang), 300453@hospital.cqmu.edu.cn (G. Xu).
1 These authors contributed equally.
https://doi.org/10.1016/j.heliyon.2023.e13613
Received 19 October 2022; Received in revised form 1 February 2023; Accepted 6 February 2023
CD8+, CD4+ and mast cells. Finally, SULF2 expression was speculated to help determine the sensitivity of certain drugs.
Conclusions: SULF2 may offer a new therapeutic target for ACC patients and become an important potential prognostic biomarker.
1. Introduction
Adrenocortical carcinoma (ACC) represents a rare while insidious endocrine disease, and its annual incidence is about 2 per million. Available data show that ACC cases have a low (<50%) 5-year survival rate [1]. However, there are few available treatment options, and radical surgical resection is the only cure method. Even so, unfortunately, most patients who have undergone radical surgery still develop local recurrence and metastasis [2]. With regard to the prognostic factors of ACC, in addition to tumor stage, the reference value of gene sequencing and other molecular biological examinations is limited although they have been available [3,4].
As a member of the sulfatase family, Sulfatase2 (SULF2) affects heparan sulfate proteoglycan (HSPG) sulfate pattern, thus contributing to cancer progression [5]. In tumor cells, the abnormal expression of SULF2 can induce structural changes in pro- teoglycans, leading to abnormal tumor cell proliferation, enhanced invasion capacity and increased susceptibility to lymph node metastasis (LNM). The role of SULF2 in numerous tumor types, including bladder cancer and pancreatic cancer, has been studied investigated [6,7]. As of yet, no study has explored the relationship between SULF2 and ACC.
In the present research, SULF2 expression together with the relation with ACC patient survival was analyzed by electronic data- bases including Gene Expression Profiling Interaction Analysis (GEPIA), The Cancer Genome Atlas (TCGA), UALCAN datasets, as well as Kaplan-Meier (KM) plotter. Besides, relation between SULF2 and tumor-infiltrating immune cell (TIICs) under different tumor microenvironments (TME) was also evaluated based on TISIDB and Tumor Immune Estimation Resource (TIMER) databases.
2. Methods
2.1. Differential expression of SULF2 in ACC
Genotype-Tissue Expression (GTEx) database and TCGA were used as the source of research data for the web-based database GEPIA (http://gepia.cancer-pku.cn/index.html) [8]. The “DIY Expression” option in GEPIA was selected for investigating SULF2 levels among ACC cases and normal tissues. Then, SULF2 expression in ACC and normal tissues was compared using GSE14922 and GSE12368 from Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo/) [9]. To be specific, the former included 4 pairs of tumor and normal tissue samples, and the latter contained 12 tumor and 6 normal tissue samples. Moreover, SULF2 levels among distinct samples collected in Human Protein Atlas (HPA) (http://www.proteinatlas.org/) and Tumor Immune Estimation Resource (TIMER2.0) (http://cistrome.shinyapps.io/timer/) were compared [10,11]. Besides, differentially expressed genes (DEGs) were identified by R package in high-compared with low-SULF2-expression groups upon the thresholds of p < 0.05 and |log2 fold change (FC)|≥1.5. Finally, R packages “ggplot2” was utilized to illustrate the results in the form of volcano plots.
2.2. UALCAN
UALCAN is a powerful network interaction platform, which allows researchers to gather valuable information and perform multiple bioinformatics analyses [12]. In this study, UALCAN was adopted to compared the expression of SULF2 in ACC according to patient’s gender, individual cancer stages (stage 1, 2, 3, 4), lymph node metastasis (LNM), and TP53 mutation status.
2.3. Functional annotation analysis of protein-protein interaction (PPI) networks
STRING (http://string-db.org) is the powerful platform developed to constructed protein networks, which allows researchers to input a list of proteins by name or amino acid sequence. STRING was utilized in the present work to construct a PPI network of SULF2. Our default minimum required interaction score is 0.4. Besides, GeneMANIA (https://genemania.org) helps researchers integrate biological network to sequence genes and predict gene functions [13]. In this study, GeneMANIA was employed to construct gene interaction networks of SULF2 and predict the related functions. In addition, the relevance of nine genes associated with SULF2 was analyzed in STRING database using R package. Gene Oncology (GO) together with KEGG analysis was subsequently conducted on the nine genes associated with SULF2 and DEGs by using ggplot2 R packages. Additionally, gseKEGG and gsePathway functions in clusterProfiler were also applied in gene set enrichment analysis (GSEA).
2.4. Genetic mutation analysis
SULF2 mutation features were analyzed based on the cBioPortal for Cancer Genomics database (http://cbioportal.org) via genome- wide pan-cancer analysis (ICGC/TCGA, Nature 2020) from pan-cancer studies [14,15].
membranous
Staining: Low
Patient id: 1966
Normal tissue
Adrenal gland
Female, age 71
B
Endometrium
50
100
150
200
A
score
E
C
Location:Cytoplasmic/
Quantity: < 25%
Intensity: Moderate
0
nTPM
0
1
2
3
4
SULF2 Expression Level (log2 TPM)
a
9
Smooth muscle
Ovary
ACC. Tumor (n=79)
Cervix
BLCA. Tumor (n=408)
BLCA.Normal (n=19)
Normal
Esophagus
BRCA.Tumor (n=1093)
(n=4)
Fallopian tube
BRCA.Normal (n=112)
BRCA-Basal. Tumor (n=190)
GSE14922
Gallbladder
T
BRCA-Her2. Tumor (n=82)
BRCA-LumA. Tumor (n=564)
*
Adipose tissue
BRCA-LumB.Tumor (n=217)
Vagina
CESC. Tumor (n=304)
**
CESC.Normal (n=3)
CHOL. Tumor (n=36)
E
Medulla oblongata
Kidney
CHOL.Normal (n=9)
-
COAD. Tumor (n=457)
Pons
Tumor
COAD.Normal (n=41)
(n=4)
DLBC. Tumor (n=48)
Hippocampal formation
Midbrain
T
ESCA.Tumor (n=184)
ESCA.Normal (n=11) GBM. Tumor (n=153)
**
Colon
GBM.Normal (n=5)
HNSC.Tumor (n=520)
HNSC.Normal (n=44)
Rectum
HNSC-HPV+. Tumor (n=97)
HNSC-HPV -. Tumor (n=421)
Breast
KICH. Tumor (n=66)
**
F
KICH.Normal (n=25)
membranous
Location:Cytoplasmic/
Quantity: < 25%
Intensity: Moderate
Staining: Low
Patient id: 2374
Normal tissue
Adrenal gland
Female, age 44
Thyroid gland
Hypothalamus
Appendix
KIRC. Tumor (n=533)
-
score
KIRC.Normal (n=72)
KIRP.Tumor (n=290)
2000
4000
6000
8000
10000
Pituitary gland
KIRP.Normal (n=32)
LAML. Tumor (n=173)
Urinary bladder
LGG.Tumor (n=516)
Heart muscle
Liver
0
LIHC. Tumor (n=371)
LIHC.Normal (n=50)
LUAD. Tumor (n=515)
.
Salivary gland
LUAD.Normal (n=59)
LUSC. Tumor (n=501)
Normal
LUSC.Normal (n=51)
(n=6)
MESO. Tumor (n=87)
T
OV.Tumor (n=303)
GSE12368
Cerebral cortex
Basal ganglia ?
Skin
PAAD. Tumor (n=178)
PAAD.Normal (n=4)
Thalamus
PCPG.Tumor (n=179)
.
**
PCPG.Normal (n=3)
Duodenum
PRAD. Tumor (n=497)
Lymph node
PRAD.Normal (n=52)
(n=12)
Small intestine
READ.Tumor (n=166)
Tumor
READ.Normal (n=10)
SARC. Tumor (n=259)
Placenta
T
SKCM.Tumor (n=103)
SKCM.Metastasis (n=368)
Amygdala
STAD.Tumor (n=415)
STAD.Normal (n=35)
White matter
Prostate
Tonsil
TGCT.Tumor (n=150)
THCA. Tumor (n=501)
THCA.Normal (n=59)
THYM. Tumor (n=120)
Male, age 35
UCEC.Tumor (n=545)
Tongue
The expression of PDCD1 Log2 (TPM+1)
membranous
Location:Cytoplasmic/
Quantity: < 25%
Intensity: Moderate
Staining: Low
Patient id: 2212
Normal tissue
Adrenal gland
Seminal vesicle
G
UCEC.Normal (n=35)
UCS. Tumor (n=57) UVM. Tumor (n=80)
Spinal cord
Testis
0
1
N
3
The expression of SULF2
Choroid plexus
Cerebellum
Stomach
2
0
2
.
.
00
D
Thymus
Log2 (TPM+1)
Lung
4
(num(T)=77; num(N)=128)
Skeletal muscle
Spleen
ACC
Adrenal gland Retina
Bone marrow
6
Parathyroid gland
Epididymis
Pancreas
-0.039
Spearman
9731
8
Human Protein Atlas (HPA) database. (B) Immunohistochemical (IHC) staining of normal adrenal tissue SULF2 in a 71-year-old female from HPA
Analysis (GEPIA). (E, F) SULF2 expression was higher in ACC than in the normal tissue in GSE14922 and GSE12368 ( ** p < 0.01). (G) The cor- 0.01, *** p < 0.001). (D) Increased SULF2 in Adrenocortical carcinoma (ACC) compared with normal tissues in Gene Expression Profiling Interaction database. (C) high expression SULF2 in different tumor types from Tumor Immune Estimation Resource (TIMER2.0) database (*p < 0.05, ** p <
relation analysis between SULF2 and PD1 mRNA level.
2.5. SULF2 and immune infiltration
TIMER2.0 database allows for comprehensively analyzing TIICs levels in different cancer types. By applying the Gene module, users can select one or more genes and see how its or their expression is correlated with TIICs levels. In this work, relation of SULF2 level with gene markers for tumor-infiltrating lymphocytes (TILs), including B cells, CD8+ T cells, Mast cell, M2 macrophages, Tregs and natural killer (NK) cells, was analyzed. TISIDB (http://cis.hku.hk/TISIDB/index.php) has been developed as the integrated web-based database to analyze tumor-immune system interconnectedness, which encompasses an extensive immune data resources [16]. It may help researchers develop new immunotherapeutic targets and forecast immunotherapeutic responses. In our study, TISIDB was chosen to explore the relations of SULF2 with immune-related molecules and cells in ACC.
2.6. Prognostic significance of SULF2 for ACC
The study analyzed whether SULF2 was associated with survival events based on a Cox proportional hazards model that estimated hazard ratios (HRs). Univariate survival analysis was first of all conducted, which obtained HRs, associated 95% confidence intervals (CIs) and p-values upon log-rank test. When the p-value was below 0.05, it indicated statistical significance of our results, which might thus be used for reference. In addition, the present work selected Kaplan-Meier plotter (http://kmplot.com/analysis/) for investi- gating whether SULF2 was of prognostic significance for ACC, like overall survival (OS) as well as disease-free survival (DFS). By using R packages (such as survival packages, pROC, and timeROC), this study also plotted ROC curve and time-dependent curve of diagnosis, and carried out nomogram model analysis. Source data for the above analyses were obtained from TCGA database.
2.7. Drug sensitivity analysis
Data regarding gene expression profiles together with drug sensitivity were collected in the CellMiner dataset. Later, correlation coefficients of SULF2 level with drug sensitivity were determined, and correlation tests were performed using R language. P < 0.05 stood for the significant correlation of the outcomes. A correlation coefficient > 0 indicated the positive gene-drug sensitivity cor- relation, and vice versa.
A
Expression of SULF2 in ACC based on individual cancer stages
250
S1 vs S2 p=0.3034
S1 vs S3 p= 4.182400E-02
200
S1 vs S4 p= 2.023600E-03
Transcript per million
S2 vs S3 p= 0.3129
S2 vs S4 p= 2.435800E-03
S3 vs S4 p= 4.496300E-02
150
100
50
0
Stage1 (n=9)
Stage2 (n=37)
Stage3 (n=16)
Stage4 (n=15)
TCGA samples
B Expression of SULF2 in ACC based on nodal metastasis status
200
NO vs N1 p= 4.488400E-02
150
Transcript per million
100
50
0
NO (n=68)
N1 (n=9)
TCGA samples
C Expression of SULF2 in ACC based on patient’s gender
D Expression of SULF2 in ACC based on TP53 mutation status
300
Male vs Female p=7.801200E-01
300
TP53-Mutant vs TP53-NonMutant p= 7.944200E-03
250
250
Transcript per million
200
Transcript per million
200
150
150
100
100
50
50
0
Male (n=31)
Female (n=48)
0
TP53-Mutant (n=16)
TP53-NonMutant (n=64)
TCGA samples
TCGA samples
3. Results
3.1. High expression of SULF2 in ACC
According to analysis based on HPA database, SULF2 expression was notably lower in adrenal gland than in other tissues. Meanwhile, immunohistochemical (IHC) results of normal adrenal gland from a 71-year-old female in the database also showed the low expression levels of SULF2 (Fig. 1A and B). Thereafter, it was found that SULF2 expression was up-regulated in ACC, kidney renal clear cell carcinoma (KIRCC) and breast cancer (Fig. 1C). Besides, high SULF2 expression was observed in ACC from TCGA database by GEPIA (p<0.05 , Fig. 1D). By analyzing GSE14922 and GSE12368 from GEO database, the same conclusion was made (Fig. 1E and F). On the other hand, the relationship between SULF2 and PD1 expression was also analyzed, unfortunately, the result was not statis- tically significant (Fig. 1G).
3.2. Relationship between SULF2 expression and cancer stage
By applying UALCAN, SULF2 expression was examined in ACC based on patient’s gender, individual cancer stages (stage 1, 2, 3, 4), LNM, and TP53 mutation status. The data showed that SULF2 showed high expression in intermediate-to-advanced cancers, with significant differences (Fig. 2A). Moreover, SULF2 was expressed at similarly increased levels in ACC patients developing LNM relative to those without LNM (Fig. 2B). However, SULF2 expression levels were similar between male and female ACC patients, which was of
A
B
antimicrobial humoral response
20
G protein-coupled receptor signaling pathway, coupled
to cyclic nucleotide second messenger
BP
humoral immune response
-Log 10 (P.adj)
15
9
ion channel complex
synaptic membrane
CC
BP
5
collagen-containing extracellular matrix
CC
10
MF
8
channel activity
KEGG
passive transmembrane transporter activity
MF
5
receptor ligand activity
B
Retinol metabolism
0
Drug metabolism - cytochrome P450
KEGG
-8
-4
0
4
8
Neuroactive ligand-receptor interaction
Log2 (Fold Change)
0 5 101520 -Log 10 (p.adjust)
C
D
KEGG_NEUROACTIVE_LIGAND_RECEPTOR_INTERACTION
REACTOME_G_ALPHA ___ SIGNALLING_EVENTS
0.2
0.2
Enrichment Score
Enrichment Score
0.0
0.0
-0.2
NES =- 1.503
-0.2
NES =- 1.421
p.adj = 0.023
p.adj = 0.023
FDR = 0.018
FDR = 0.018
Ranked list metric
5.0
Ranked list metric
5.0
2.5
2.5
0.0
0.0
-2.5
-2.5
-5.0
-5.0
-7.5
-7.5
10000
20000
30000
10000
20000
30000
Rank in Ordered Dataset
Rank in Ordered Dataset
no statistical difference (Fig. 2C). In addition, SULF2 was expressed at higher levels in patients with TP53 mutations in the tumor than in non-mutated patients (Fig. 2D).
3.3. PPI network and functional annotation
DEGs were identified in high-compared with low-SULF2 expression groups using the R package. There were altogether 1889 DEGs obtained, which included 953 with up-regulation and 936 with down-regulation (Fig. 3A). According to Fig. 3B, such DEGs were enriched in antimicrobial humoral response and humoral immune response in the biological process (BP). Meanwhile, the enriched cellular components (CC) terms included synaptic membrane, collagen-containing extracellular matrix and ion channel complex. Furthermore, the enriched molecular function (MF) terms were channel activity, passive transmembrane transporter activity and receptor ligand activity. According to KEGG enrichment, these DEGs were enriched into retinol metabolism, drug metabolism- cytochrome P450 and neuroactive ligand-receptor interaction. As revealed by GSEA results, low-SULF2-expression patients showed significant enrichment of KEGG neuroactive ligand receptor interaction and reactome g alpha I signaling events (Fig. 3C and D). However, there was no enriched pathway in the high-SULF2-expression group.
STRING database assists in building the PPI network related to SULF2, as a result, ten functional partner genes with high con- nectivity were obtained from the network. In fact, HS2ST1 and ENSP00000359579 are two names of Heparan sulfate 2-O-sulfotrans- ferase 1, therefore, 9 functional partner genes were finally obtained (Fig. 4A). Besides, GeneMANIA database-based gene-gene network
A
GLCE
B
ARSA
GPC2
GPC6
GPC3
Networks
XS3ST3A0
GPC5
HS6ST2
Physical Interactions
HS3ST2
AC093155.3
GPC4
Shared protein domains
HS2ST1
GPC1
SULF2
ENSP00000359579
Co-expression
HS6ST
1
HS3ST1
CHST8
GPC3
GLCE
Genetic Interactions
NDST
ARSG
MS3ST380
CHST10
NOSTI
HS6ST2
Functions
SULF2
aminoglycan biosynthetic process
H93ST5
HS2ST1
HS6ST1
HS6ST3
glycosaminoglycan metabolic process
GPC1
aminoglycan metabolic process
COPE
HS3STO
COPE
ARSG
CHST14
sulfotransferase activity
UST
CHST13
transferase activity, transferring sulfur-containing groups
CHST12
proteoglycan metabolic process
CHST11
CHST9
proteoglycan biosynthetic process
C
SULF2
GPC1
GPC3
HS6ST1
HS2ST1
HS6ST2
NDST1
GLCE
COPE
ARSG
D
SULF2
GPC1
GPC3
HS6ST1
HS2ST1
HS6ST2
NDST1
GLCE
COPE
ARSG
SULF2
1
0.21
0.32
0.28
0.25
0.2
0.39
0.2
0.46
-0.02
SULF2
**
**
*
*
**
**
GPC1
0.21
1
0.28
0.52
0.52
0.14
0.15
0.4
0.29
0.19
GPC1
*
**
**
**
**
* p < 0.05
GPC3
0.32
0.28
1
0.42
0.34
0.23
0.24
0.4
0.31
0.19
Correlation
GPC3
**
*
**
**
**
*
*
**
**
** p < 0.01
HS6ST1
0.28
0.52
0.42
1
0.46
0.21
0.23
0.6
0.27
0.21
1.0
HS6ST1
*
**
**
**
**
*
**
*
Correlation
0.5
HS2ST1
0.25
0.52
0.34
0.46
1
0.15
0.5
0.58
0.38
0.32
HS2ST1
1.0
*
**
**
**
**
**
**
**
0.0
0.5
HS6ST2
0.2
0.14
0.23
0.21
0.15
1
0.3
0.03
0.32
0.41
HS6ST2
*
**
**
**
**
-0.5
0.0
NDST1
0.39
0.15
0.24
0.23
0.5
0.3
1
0.32
0.44
0.36
-1.0
NDST1
**
*
*
**
**
**
**
**
**
-0.5
GLCE
0.2
0.4
0.4
0.6
0.58
0.03
0.32
1
0.37
0.28
GLCE
**
**
**
* *
**
**
*
-1.0
COPE
0.46
0.29
0.31
0.27
0.38
0.32
0.44
0.37
1
0.04
COPE
**
**
**
*
**
**
**
**
**
ARSG
-0.02
0.19
0.19
0.21
0.32
0.41
0.36
0.28
0.04
1
ARSG
**
**
**
*
E
F
G
Gene Ontology
aminoglycan biosynthetic process
KEGG pathway
GSEA analysis
glycosaminoglycan biosynthetic process
proteoglycan metabolic process
5
KEGG NEUROACTIVE LIGAND RECEPTOR INTERACTION
heparan sulfate proteoglycan metabolic process
Proteoglycans in cancer
anchored component of membrane
REACTOME G ALPHA I SIGNALLING EVENTS
Padjust
Golgi lumen
CC
lysosomal bamen
8
KEGG
MP
REACTOME CLASS A 1 RHODOPSIN LIKE RECEPTORS
0.023
anchored component of plasma membrane
Glycosaminoglycan biosynthesis - heparan sulfate / heparin
arylsulfatase activity
transferase activity, transferring sulfur-containing
REACTOME GPCR LIGAND BINDING
groups
sulfotransferase activity
¥
heparan sulfate sulfotransferase activity
0
2
4
6
8
10
WP NUCLEAR RECEPTORS METAPATHWAY
0
5
10
15
-Log 10 (p.adjust)
-6
-4
-Log 10 (p.adjust)
-2
0
analysis revealed interaction of SULF2 with 30 candidate target genes (Fig. 4B). Excitingly, SULF2 was strongly linked with nine functional partner genes, which displayed significantly positive correlation with each other (Fig. 4C and D). As shown in Fig. 4E, the GO analysis results revealed the involvement of SULF2 and its partner in BPs including “glycosaminoglycan biosynthetic process”, “aminoglycan biosynthetic process”, “heparan sulfate proteoglycan metabolic process”, and “proteoglycan metabolic process”. The CC terms enriched included “anchored component of membrane”, “Golgi lumen”, “lysosomal lumen” and “anchored component of plasma membrane”. The enriched MF terms were “arylsulfatase activity”, “heparan sulfate sulfotransferase activity”, “sulfotransferase ac- tivity” and “transferase activity, transferring sulfur-containing groups”. Moreover, as demonstrated by KEGG enrichment, these genes were enriched into proteoglycans in cancer and glycosaminoglycan biosynthesis-heparan sulfate/heparin (Fig. 4F). Our GSEA was conducted by using TCGA-derived RNAseq data and the results showed obvious enrichment of KEGG neuroactive ligand receptor interaction and reactome g alpha I signaling events pathways (Fig. 4G).
3.4. Genetic alteration analysis of SULF2 in ACC
The present work attempted to explore the mutational signature of SULF2 in ACC based on the cBioPortal tool, as a results, the genetic alteration frequency of SULF2 was lower than 6% in ACC (Fig. 5A). Fig. 5B displays the mutation spots of SULF2 in ACC. As shown in Fig. 5C-E, there was no significant difference in OS (p = 0.905), progression-free survival (PFS) (p = 0.624) or disease- specific survival (DSS) (p = 0.766) in SULF2-altered ACC group compared with SULF2-unaltered group. In conclusion, SULF2 gene alterations may not be associated with the development of ACC.
3.5. Immune correlation analysis
Immune infiltration is involved in the occurrence and development of tumors. According to our results, SULF2 was associated with the infiltration levels of TILs based on TISIDB database (Fig. 6A). Besides, as shown in Fig. 6B, SULF2 up-regulation led to the decreased TIL infiltration levels, such as Mast cell (rho = - 0.367) and NK cell (rho = - 0.289). However, SULF2 up-regulation showed positive relation to infiltration levels of CD4+ Th1 cell (rho = 0.451) and B cell (rho = 0.24). The statistical results showed that all P-values were less than 0.05. Therefore, it was reasonable to believe that SULF2 showed close relation with tumor immunity, which might affect ACC
A
6%-
B
Alteration Frequency
# SULF2 Mutations
5-
5%-
KB53T
4%-
0-
Sulfatase
DUF3740
0
200
400
600
800
870aa
3%-
C
2%-
100%
Logrank Test P-Value: 0.905
90%
1%-
80%
Probability of Overall Survival
70%
Structural variant data +
60%
Mutation data +
50%
CNA data +
40%
Adrenocortical Carcinoma
30%
20%
10%
0%
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
Overall Survival (Months)
D
E
Overall
Altered group
Unaltered group
Mutation
Amplification
100%
Logrank Test P-Value: 0.766
100%
Logrank Test P-Value: 0.624
90%-
90%
80%-
80%
70%
70%
Disease-specific
60%-
Progression Free
60%
50%
50%
40%
40%
30%
30%
20%
20%
10%-
10%
0%
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
0%
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
Months of disease-specific survival
Progress Free Survival (Months)
Disease-specific
Progression Free
Altered group
Altered group
Unaltered group
Unaltered group
A
Act CD8
B
SULF2 Expression Level (log2 TPM)
Purity
T cell CD8+_EPIC
T cell CD4+_EPIC
T cell CD4+ (non-regulatory)_XCELL
Rho = - 0.298
Tcm CD8
8
Rho = 0.282
p = 1.498-02
Rho =- 0.231
p = 4.94e-02
p = 1.03e-02
Rho = 0.266
p = 2.29e-02
Tem CD8
6
Act CD4
ACC
Tcm CD4
4.
Tem CD4
Tfh
0.2
0.4
0.6
0.8
1.0 0.00
0.01
0.02
0.03
0.04
0.00
0.05
0.10
0.15
0.20
0.000
0.005
0.010
Purity
Infiltration Level
Infiltration Level
Infiltration Level
Tgd
T cell CD4+ memory resting_CIBERSORT
ell CD4+ memory resting_CIBERSORT-A
T cell CD4+ Th1_XCELL
T cell regulatory (Tregs)_XCELL
Rho = - 0.259
p = 2.67e-02
Rho =- 0.244
p = 3.75e-02
Rho = 0.451
Rho =- 0.243
Th1
p = 6.10c-05
p = 3.81e-02
Th17
Th2
Treg
Act B
Imm B
0.0
0.1
0.2
0.3
0.4
0.00
0.05
0.10
0.15
0.00
0.05
0.10
0.04
Infiltration Level
Infiltration Level
Infiltration Level
0.00
0.01
0.02
0.03
0.05
Infiltration Level
Mem B
B cell_QUANTISEQ
Macrophage M2_QUANTISEQ
Macrophage M2_XCELL
Myeloid dendritic cell_MCPCOUNTER
Rho = 0.24
Rho =- 0.252
NK
p = 4.07e-02
Rho = - 0.421 p = 2.10e-04
p = 3.12e-02
Rho = - 0.286 p = 1.42e-02
CD56bright
CD56dim
MDSC
NKT
Act DC
0.000
0.025
0.050
0.075
0.100
0.00
0.05
0.10
0.00
0.02
0.04
0.06
0
30
60
90
120
1
Infiltration Level
Infiltration Level
Infiltration Level
Infiltration Level
pDC
NK cell_QUANTISEQ
Mast cell_XCELL
Mast cell activated_CIBERSORT-ABS
Rho =- 0.289
Rho = - 0.367
p = 1.41e-03
Rho = - 0.287
iDC
p = 1.30e-02
p = 1.39e-02
Macrophage
Eosinophil
Mast
Monocyte
Neutrophil
1
0.000
0.005
0.010
0.015
0.020
0.02
0.00
0.02
0.04
0.00
0.05
0.10
Infiltration Level
Infiltration Level
Infiltration Level
development. The expression of SULF2 was associated with multiple immune molecules. Based on our research results, SULF2 was related to several kinds of immunoinhibitors, such as PVRL2 (rho = 0.495, p = 4.74e-06), IL10RB (rho = 0.314, p = 0.00507) and CSF1R (rho = - 0.272, p = 0.0157) (Fig. 7A). Besides, SULF2 expression was related to immunostimulators, including PVR (rho = 0.454, p = 3.23e-05), TNFSF13 (rho = - 0.445, p = 4.7e-05), HHLA2 (rho = - 0.424, p = 0.000115) and CD28 (rho = - 0.387, p = 0.000465) (Fig. 7B). Therefore, it was suggested that SULF2 might be involved in facilitating immune surveillance escape of tumors.
3.6. Correlation of SULF2 expression with chemokines and receptors
Chemokines and receptors play important role in tumor immune process. Based on our results, SULF2 was related to chemokines and receptors. For example, SULF2 level was tightly related to CCL8 (rho = - 0.434, p = 7.72e-05), XCL1 (rho = - 0.332, p = 0.0029) and CCL2 (rho = - 0.264, p = 0.019) (Fig. 8A). Meanwhile, SULF2 expression was also closely associated with receptors, including CCR2 (rho = - 0.317, p = 0.0046), CCR6 (rho = - 0.348, p = 0.00179) and CXCR6 (rho = - 0.347, p = 0.00185) (Fig. 8B).
3.7. Diagnostic and prognostic significance of SULF2 for ACC
By integrating age, T stage, and gender, a nomogram model was constructed for predicting 3- and 5-year survival of ACC (Fig. 9A). When SULF2 expression was added to the nomogram model, it was found that the model could be used to guide the prediction of 2-, 3-, and 5-year survival probabilities of tumor patients. And the survival probability was significantly associated with SULF2 expression
A Immunoinhibitor
ACC (79 samples)
ACC (79 samples)
ACC (79 samples)
ACC (79 samples)
ACC (79 samples)
.
2.5
7.5
4
6
6
CD244_exp
CD274_exp
0.0
CSF1R_exp
IL10RB_exp
5.0
IDO1_exp
0
3
-2.5
5
2.5
0
-4
-5.0
0.0
-3
4
4
SULF2_exp
6
8
4
6
8
4
6
8
4
6
8
4
6
SULF2_exp
Spearman Correlation Test: rho = - 0.274, p = 0.0147 ACC (79 samples)
SULF2_exp
8
Spearman Correlation Test: rho = - 0.261, p = 0.0206 ACC (79 samples)
Spearman Correlation Test: rho = - 0.272, p = 0.0157 ACC (79 samples)
SULF2_exp
Spearman Correlation Test: rho = 0.23, p = 0.0421 ACC (79 samples)
SULF2_exp
Spearman Correlation Test: rho = 0.314, p = 0.00507
6
8
9
2.5
3
KDR_exp
0.0
₹6
LAG3_exp
PVRL2_exp
TIGIT_exp
0
2.5
4
-3
-5.0
6
4
6
8
-6
SULF2_exp
4
SULF2_exp
6
8
4
SULF2_exp
6
8
4
Spearman Correlation Test: rho = - 0.242, p = 0.032
Spearman Correlation Test: rho = 0.278, p = 0.0132
Spearman Correlation Test: rho = 0.495, p = 4.74e-06
SULF2_exp
6
8
Spearman Correlation Test: rho = - 0.228, p = 0.0437
B Immunostimulator
ACC (79 samples)
ACC (79 samples)
ACC (79 samples)
ACC (79 samples)
ACC (79 samples)
4
2.5
2-
7.5
0.0
4
CD27_exp
CD28_exp
0
CXCL12_exp
HHLA2_exp
ICOSLG_exp
0.0
5.0
-2.5
-2
2.
-2.5
2.5
-4-
-5.0
-5.0
0
-6-
0.0
i
4
SULF2_exp
6
8
4
Spearman Correlation Test: rho = - 0.237, p = 0.036 ACC (79 samples)
SULF2_exp
6
8
4
Spearman Correlation Test: rho = - 0.387, p = 0.000465 ACC (79 samples)
SULF2_exp
6
8
4
8
4
Spearman Correlation Test: rho = - 0.254, p = 0.0241 ACC (79 samples)
SULF2_exp
6
Spearman Correlation Test: rho = - 0.424, p = 0.000115 ACC (79 samples)
SULF2_exp Spearman Correlation Test: rho = - 0.37, p = 0.000847
6
8
5.0
8-
6
6
2.5
IL2RA_exp
0.0
IL6R_exp
7
TNFSF13_exp
-4
PVR_exp
4
-2.5
2
6
2-
-5.0
0
5
4
SULF2_exp Spearman Correlation Test: rho = - 0.356, p = 0.00138
6
8
4
SULF2_exp
6
8
4
Spearman Correlation Test: rho = - 0.285, p = 0.0111
SULF2_exp Spearman Correlation Test: rho = 0.454, p = 3.23e-05
6
8
4
SULF2_exp Spearman Correlation Test: rho = - 0.445, p = 4.7e-05
6
8
(Fig. 9B). Based on TCGA database, univariate as well as multivariate regression revealed that SULF2 was positively correlated with HRs of ACC, and it was a valuable prognostic factor (Fig. 9C and D). According to the diagnostic ROC curve, SULF2 exhibited a good ability to discriminate tumor from healthy samples (AUC = 0.864) (Fig. 10A). Moreover, the time-dependent survival ROC curve based on SULF2 expression was used for predicting 1-, 3-, and 5-year survival of ACC cases. All these AUC values were found to be above 0.7, suggesting good predictive power of SULF2 expression (Fig. 10B). As revealed by KM curve analysis, ACC patients who had SULF2 up- regulation exhibited the poorer OS (p = 3.3e-05) and DFS (p = 0.00023) (Fig. 10C and D).
3.8. The association of the SULF2 expression level with drug sensitivity
Data regarding gene expression profiles along with drug sensitivity were collected in CellMiner. Thereafter, correlation coefficient of SULF2 level with drug sensitivity was determined by R language. Thereafter, 15 SULF2-related drugs were chosen based on R-value, as a result, SULF2 expression showed positive relation to tumor cell sensitivity to drugs like LY-2835219, Rapamycin, Everolimus, Midostaurin and Idelalisib, but the opposite was true in some drugs such as Fludarabine, DACARBAZINE and Raltitrexed (Fig. 11A-O). Therefore, it was speculated that SULF2 expression might help determine the sensitivity of certain drugs based on these results.
A
Chemokine
ACC (79 samples)
ACC (79 samples)
ACC (79 samples)
ACC (79 samples)
ACC (79 samples)
7.5-
6
7.5
9
5.0
4
2.5
5.0
6
CCL2_exp
CCL3_exp
2
CCL5_exp
CCL8_exp
0.0
CCL14_exp
2.5
2.5
3
0
-2.5
0.0
0.0
0
-2
-5.0-
-2.5
-2.5
4
SULF2_exp Spearman Correlation Test: rho = - 0.264, p = 0.019 ACC (79 samples)
6
8
4
6
SULF2_exp
8
4
SULF2_exp
6
8
-3
4
6
4
6
Spearman Correlation Test: rho = - 0.239, p = 0.0342 ACC (79 samples)
Spearman Correlation Test: rho = - 0.222, p = 0.049 ACC (79 samples)
SULF2_exp
8
Spearman Correlation Test: rho = - 0.434, p = 7.72e-05
SULF2_exp Spearman Correlation Test: rho = - 0.283, p = 0.0118
8
2.5
7.5
0.0
0.0
CXCL12_exp
5.0
XCL1_exp
XCL2_exp
-2.5
-2.5
2.5
-5.0
-5.0
0.0
4
SULF2_exp
6
8
4
SULF2_exp
6
8
4
Spearman Correlation Test: rho = - 0.254, p = 0.0241
Spearman Correlation Test: rho = - 0.332, p = 0.0029
SULF2_exp
6
8
Spearman Correlation Test: rho = - 0.281, p = 0.0123
B Receptor
ACC (79 samples)
ACC (79 samples)
ACC (79 samples)
ACC (79 samples)
ACC (79 samples)
2.5
1
2.5
2.5
4
0.0
0
CCR1_exp
CCR2_exp
CCR6_exp
-1
CXCR2_exp
0.0
CXCR6_exp
0.0
2
-2.5
-2
-2.5
2.5
0
-3
-5.0
-5.0
-5.0
-2
-4
4
SULF2_exp
6
8
4
SULF2_exp
6
8
4
6
8
4
6
8
4
6
8
Spearman Correlation Test:
rho = - 0.273, p = 0.0151
Spearman Correlation Test: rho = - 0.317, p = 0.0046
SULF2_exp
Spearman Correlation Test: rho = - 0.348, p = 0.00179
SULF2_exp
Spearman Correlation Test: rho = - 0.224, p = 0.0478
SULF2_exp
Spearman Correlation Test: rho = - 0.347, p = 0.00185
4. Discussion
Nowadays, there are studies revealing the relationship between genes and tumors development [17,18]. The current research is the first to analyze SULF2 expression and its prognostic value of ACC through a series of bioinformatics approaches. Our study results suggested a possible link between high SULF2 expression and poor prognosis of ACC. In addition, this study also first suggested that SULF2 was strongly related to infiltration level of numerous immune-related molecules and cells in ACC. According to univariate as well as multivariate regression results, SULF2 expression was positively correlated with the HRs of ACC. Therefore, our research indicated the new and important functions of SULF2, which might affect the survival and prognosis of ACC patients by participating in immune infiltration.
SULF2, a member of sulfatase family, targets 6-O-sulfate groups on glucosamine residues in heparan sulfate (HS) chains while regulating a variety of molecular processes in the TME [19]. The signaling ligand-receptor interactions will be altered when the extracellular matrix (ECM) components of HS are affected by SULF2 [20]. Moreover, SULF2 can modify HS-mediated pathway by regulating HSPG expression [21,22]. Previous studies have shown that SULF2 expression increased within different tumor cells like bladder cancer, lung cancer and hepatocellular carcinoma, and that the high SULF2 expression level was related to poor prognosis [6, 23,24]. Until now, whether SULF2 is associated with tumor immunity and its specific significance in ACC remain unclear, which deserves further research.
First, SULF2 levels within ACC and normal tissues were investigated through TCGA, GEO, TIMER and GEPIA databases. As a result, SULF2 showed remarkably differential expression among tumor and healthy tissues in multiple human malignancies. Similarly, SULF2 expression notably increased in ACC compared with para-cancerous samples. Besides, the increased SULF2 expression was consistent with PD1, but unfortunately, there was no statistically significant relationship between them. According to these findings, SULF2 expression might be the possible auxiliary diagnostic reference for ACC. On the other hand, the up-regulated SULF2 expression level was markedly related to the poorer outcome of ACC in stages 1 and 2, 3, 4, N 0 and 1. Furthermore, based on KM plotter analysis, SULF2 up-regulation predicted the poor OS and DFS of ACC. The above results powerfully confirmed our hypothesis that SULF2 was
A
B
Points
0
20
40
60
80
100
Points
0
20
40
60
80
100
T stage
T3&T4
T stage
T3&T4
TI&T2
NI
T1
N stage
N stage
&T2
N1
NO
M
M stage
NO
M1
M stage
MO
MO
Female
Gender
Female
Gender
Age
Male
>50
Male
Age
50
Laterality
⇐ 50
eft
Laterality
-50
Len
SULF2
Right
High
Right
Low
Total Points
Total Points
0
40
80
120
160
200
Linear Predictor
0
40
80
120
160
200
240
280
Linear Predictor
2
1
0
2.5
3-year Survival Probability
1
2
₹
2-year Survival Probability
2.5
-1.5
0.5
0.5
1.5
3-year Survival Probability
0.8
0.6
0.4
0.2
5-year Survival Probability
0.8
0.6
0.4
0.2
0.4
5-year Survival Probability
0.8
0.6
0.4
0.2
0.8
0.6
0.2
0.8
0.6
0.4
0.2
C
D
0
10
20
30
0
5
10
15
20
| Characteristics | Total(N) | HR(95% CI) | P value | Characteristics | Total(N) | HR(95% CI) | P value |
|---|---|---|---|---|---|---|---|
| T stage | 77 | T stage | 77 | ||||
| T1 | 9 | Reference | T1 | 9 | |||
| T2 | 42 | 2.444 (0.305-19.596) | 0.4 | T2 | 42 | 1.897 (0.232-15.531) | 0.551 |
| T3 | 8 | 12.439 (1.300-118.988) | 0.029 | T3 | 8 | 5.778 (0.554-60.221) | 0.142 |
| T4 | 18 | 30.506 (3.595-258.857) | 0.002 | T4 | 18 | 17.511 (1.797-170.675) | 0.014 |
| N stage | 77 | N stage | 77 | ||||
| NO | 68 | Reference | NO | 68 | |||
| N1 | 9 | 2.038 (0.769-5.400) | 0.152 | N1 | 9 | ||
| M stage | 77 | M stage | 77 | ||||
| MO | 62 | Reference | MO | 62 | |||
| M1 | 15 | 6.150 (2.710-13.959) | <0.001 | M1 | 15 | 0.709 (0.249-2.013) | 0.518 |
| Age | 79 | Age | 79 | ||||
| <= 50 | 41 | Reference | <= 50 | 41 | |||
| >50 | 38 | 1.799 (0.846-3.824) | 0.127 | >50 | 38 | ||
| Gender | 79 | Gender | 79 | ||||
| Female | 48 | Reference | Female | 48 | |||
| Male | 31 | 1.001 (0.469-2.137) | 0.999 | Male | 31 | ||
| SULF2 | 79 | 2.015 (1.492-2.722) | <0.001 | SULF2 | 79 | 1.759 (1.237-2.503) | 0.002 |
Fig. 9. Nomogram and Cox hazard analysis of SULF2 in ACC. (A) Nomogram model, based on clinicopathologic factors to predict survival prob- ability at 3-, and 5-years. (B) Nomogram model, integrating clinicopathologic factors and SULF2 level to predict survival probability at 2-, 3-, and 5- years. (C) Single-factor cox analysis of ACC. (D) Multivariate cox analysis of ACC.
likely to be a prognostic biomarker in ACC. SULF2 expression performed well in distinguishing cancer from healthy samples and predicting the long-term survival rates, suggesting that it was the possibly useful biomarker used to diagnose and predict the prog- nostic of ACC.
GPC3, NDST1 and COPE were identified as the functional partner genes for SULF2. As revealed by our GO and KEGG analyses, they were enriched in heparan sulfate proteoglycan metabolic process, glycosaminoglycan biosynthesis-heparan sulfate/heparin, and heparan sulfate sulfotransferase activity. Previous studies have clearly stated that extracellular vesicle-mediated communication is very important for pathological process in tumor [25]. Many important extracellular vesicle-mediated communication processes like uptake and biogenesis, are under the regulation of HSPG [26]. Based on these results, it is reasonable to believe that GPC3, NDST1, COPE and SULF2 have important synergistic roles in the pathogenesis of ACC. Further, our GSEA results showed that neuroactive ligand receptor interaction and reactome g alpha I signaling events pathways were obviously enriched by these genes. Studies reveal that neuroactive ligand receptor interaction is associated with diffuse intrinsic pontine glioma and granulosa cell tumor development [27,28]. Therefore, the neuroactive ligand receptor interaction may be linked with ACC pathogenesis. Additionally, infiltration of immune cells has a key role in carcinogenesis [29]. SULF2, a novel immunomodulators, has positive modulatory effects on antigen delivery and cytophagy of immune cells [30]. Xuping Niu et al. discovered that SULF2 was up-regulated in dermal mesenchymal stem cells, which affected the inflammatory microenvironment via multiple pathways, including regulation of immune cell proliferation, differentiation, migration and recruitment [31,32]. But so far, it remains unknown whether SULF2 expression is involved in immune infiltration in ACC. This study first illustrated that SULF2 was related to immune infiltration level in ACC. Based on our analyses, SULF2 expression was obviously correlated with immune cells like CD4+T cells, CD8+ T cells, and NK cell. Meanwhile, the increased SULF2 expression was related to diverse kinds of immuneinhibitors, immunestimulators, chemokines and receptors suggesting that SULF2 had a critical effect on immune regulation in ACC. According to our PPI network and functional annotation results, SULF2 and interacting genes were related to the underlying tumor biological processes and contributed to tumorigenesis and progression. Besides, PD1, an important immune checkpoint component, has been recognized as an important target for tumor immunotherapy [33,34]. This study disclosed that SULF2 expression was associated with PD1 and immune-related receptors and molecules. Therefore, immunotherapy may be useful for ACC and more efforts are needed to explore the molecular mechanism. On the other hand, we found that TP53 mutation status was related to tumor development and prognostic outcome [35,36]. Moreover, SULF2 levels remarkably increased in ACC patients with TP53 mutation compared with in non-mutated cases, with the difference being statistically significant. It suggests that there may be a link between SULF2 and TP53, which deserves further study. Finally, SULF2 expression was markedly related to several drugs, which provides some guidance for clinical treatment of the disease.
Nonetheless, certain limitations should be noted in this analysis. For example, the data were mostly sourced from online databases
A
B
1.0
1.0
0.8
0.8
Sensitivity (TPR)
Sensitivity (TPR)
0.6
0.6
0.4
0.4
0.2
SULF2
SULF2
0.2
1-Year (AUC = 0.818)
AUC: 0.864
3-Year (AUC = 0.781)
0.0
CI: 0.800-0.928
0.0
5-Year (AUC = 0.844)
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
1-Specificity (FPR)
1-Specificity (FPR)
C
Overall Survival
D
Disease Free Survival
1.0
LOW SULF2 TPM
1.0
High SULF2 TPM
Low SULF2 TPM
Logrank p=7.6e-07
High SULF2 TPM
Logrank p=8.1e-05
0.8
HR(high)=11
p(HR)=3.3e-05
0.8
HR(high)=3.9
n(high)=38
p(HR)=0.00023
Percent survival
Percent survival
n(high)=38
0.6
n(low)=38
0.6
n(low)=38
0.4
0.4
0.2
0.2
0.0
0.0
0
50
100
150
0
50
100
150
Months
Months
and our results might be affected by subsequent data updates, which might possibly introduce bias. Besides, because of the rarity of the disease, not enough cases were obtained to complete the experimental validation.
5. Conclusion
According to our results, SULF2 shows abnormal expression in ACC, which predicts the dismal prognostic outcome. Meanwhile, this study is the first to illustrate that SULF2 expression is closely related to immune system, indicating its possible involvement in immune infiltration of tumors, which sheds novel lights on diagnosing and treating ACC.
Author contribution statement
Jiusong Yan and Xiaodu Xie: Conceived and designed the experiments; wrote the paper. They are co-first authors.
Qinke Li and Peihe Liang: Performed the experiments; analyzed and interpreted the data.
Junyong Zhang and Guangyong Xu: Contributed reagents, materials, analysis tools or data. They were responsible for the final review of the paper who are co-corresponding authors of this paper.
Funding statement
The research was funded by the Natural Science Foundation of Chongqing (2022NSCQ-MSX0283).
A
SULF2, Fludarabine Cor =- 0.382, p=0.003
B
SULF2, LY-2835219 Cor=0.351, p=0.006
C
SULF2, Rapamycin
D SULF2, Everolimus Cor=0.340, p=0.008
Cor=0.346, p=0.007
·
2
.
1.0
··
..
1
2
1
0.5
1
0.0
0
0
0
-1
-0.5
-1
-1.0
-1
-2
0
2
4
6
0
2
4
6
0
2
4
6
0
2
4
6
E
SULF2, Midostaurin
F
SULF2, DACARBAZINE
G
SULF2, Idelalisib
H
SULF2, IPI-145 Cor=0.288, p=0.026
Cor=0.315, p=0.014
Cor =- 0.314, p=0.015
Cor=0.294, p=0.023
2
2
:
3
·
:
2
0
1
2
1
..
1
-2
0
0
0
-1
-1
-1
0
2
4
6
0
2
4
6
0
2
4
6
0
2
4
6
I
SULF2, Copanlisib
SULF2, Abiraterone
Cor=0.277, p=0.032
J
Cor=0.271, p=0.037
K
SULF2, 6-MERCAPTOPURINE Cor =- 0.266, p=0.040
L SULF2, RAPAMYCIN Cor=0.265, p=0.041
··
1.5
1.0
..
..
1
1.0
1
0.5
:
0
0.5
0.0
0
0.0
-1
-0.5
-1
-0.5
-2
-1.0
-2
-1.0
-1.5
0
2
4
6
0
2
-1.5
4
6
0
2
4
6
0
2
4
6
M
SULF2, Cladribine Cor =- 0.259, p=0.046
N
SULF2, Raltitrexed Cor =- 0.256, p=0.048
0
SULF2, Irofulven
Cor=0.256, p=0.049
2
2.
2
2
1
1
0
0
0
-2
-1
..
-4
-1
0
2
4
6
0
2
4
6
0
2
4
6
Data availability statement
Data can be obtained from corresponding authors upon on request.
Declaration of interest statement
All authors declared no competing interest.
References
[1] M. Ettaieb, et al., Past, present and future of epigenetics in adrenocortical carcinoma, Cancers 12 (5) (2020).
[2] M. Fassnacht, et al., Adrenocortical carcinomas and malignant phaeochromocytomas: ESMO-EURACAN Clinical Practice Guidelines for diagnosis, treatment and follow-up, Ann. Oncol. 31 (11) (2020) 1476-1490.
[3] Z. Zhang, et al., Expression and clinical significance of VISTA and PD-L1 in adrenocortical carcinoma, Endocr. Relat. Cancer 29 (7) (2022) 403-413.
[4] G. Cantini, et al., Circulating fascin 1 as a promising prognostic marker in adrenocortical cancer, Front. Endocrinol. 12 (2021), 698862.
[5] I. Maltseva, et al., The SULFs, extracellular sulfatases for heparan sulfate, promote the migration of corneal epithelial cells during wound repair, PLoS One 8 (8) (2013), e69642.
[6] J. Huang, et al., SULF2 is a novel diagnostic and prognostic marker for high-grade bladder cancer with lymphatic metastasis, Ann. Transl. Med. 9 (18) (2021) 1439.
[7] R. He, et al., SULF2 enhances GDF15-SMAD axis to facilitate the initiation and progression of pancreatic cancer, Cancer Lett. 538 (2022), 215693.
[8] Z. Tang, et al., GEPIA: a web server for cancer and normal gene expression profiling and interactive analyses, Nucleic Acids Res. 45 (W1) (2017) W98-w102.
[9] T. Barrett, et al., NCBI GEO: mining tens of millions of expression profiles-database and tools update, Nucleic Acids Res. 35 (Database issue) (2007) D760-D765.
[10] K. Colwill, S. Gräslund, A roadmap to generate renewable protein binders to the human proteome, Nat. Methods 8 (7) (2011) 551-558.
[11] T. Li, et al., TIMER2.0 for analysis of tumor-infiltrating immune cells, Nucleic Acids Res. 48 (W1) (2020) W509-w514.
[12] D.S. Chandrashekar, et al., UALCAN: an update to the integrated cancer data analysis platform, Neoplasia 25 (2022) 18-27.
[13] D. Warde-Farley, et al., The GeneMANIA prediction server: biological network integration for gene prioritization and predicting gene function, Nucleic Acids Res. 38 (2010) W214-W220 (Web Server issue).
[14] E. Cerami, et al., The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data, Cancer Discov. 2 (5) (2012) 401-404.
[15] J. Gao, et al., Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal, Sci. Signal. 6 (269) (2013) pl1.
[16] B. Ru, et al., TISIDB: an integrated repository portal for tumor-immune system interactions, Bioinformatics 35 (20) (2019) 4200-4202.
[17] Y. Feng, et al., ARID1A is a prognostic biomarker and associated with immune infiltrates in hepatocellular carcinoma, Chin. J. Gastroenterol. Hepatol. 2022 (2022), 3163955.
[18] L. Lu, et al., Methylation and expression of the exercise-related TLR1 gene is associated with low grade glioma prognosis and outcome, Front. Mol. Biosci. 8 (2021), 747933.
[19] E. Hammond, et al., The role of heparanase and sulfatases in the modification of heparan sulfate proteoglycans within the tumor microenvironment and opportunities for novel cancer therapeutics, Front. Oncol. 4 (2014) 195.
[20] C.B. Kirn-Safran, S.S. D’Souza, D.D. Carson, Heparan sulfate proteoglycans and their binding proteins in embryo implantation and placentation, Semin. Cell Dev. Biol. 19 (2) (2008) 187-193.
[21] W. Chen, Y. Hu, D. Ju, Gene therapy for neurodegenerative disorders: advances, insights and prospects, Acta Pharm. Sin. B 10 (8) (2020) 1347-1359.
[22] S.D. Rosen, H. Lemjabbar-Alaoui, Sulf-2: an extracellular modulator of cell signaling and a cancer target candidate, Expert Opin. Ther. Targets 14 (9) (2010) 935-949.
[23] H. Lemjabbar-Alaoui, et al., Sulf-2, a heparan sulfate endosulfatase, promotes human lung carcinogenesis, Oncogene 29 (5) (2010) 635-646.
[24] J.P. Lai, et al., The oncogenic effect of sulfatase 2 in human hepatocellular carcinoma is mediated in part by glypican 3-dependent Wnt activation, Hepatology 52 (5) (2010) 1680-1689.
[25] M. Tkach, C. Théry, Communication by extracellular vesicles: where we are and where we need to go, Cell 164 (6) (2016) 1226-1232.
[26] M. Cerezo-Magaña, A. Bång-Rudenstam, M. Belting, The pleiotropic role of proteoglycans in extracellular vesicle mediated communication in the tumor microenvironment, Semin. Cancer Biol. 62 (2020) 99-107.
[27] N. Ni, et al., Transcriptomic profiling of gene expression associated with granulosa cell tumor development in a mouse model, Cancers 14 (9) (2022).
[28] L. Wei, et al., Bioinformatics analysis of microarray data to reveal the pathogenesis of diffuse intrinsic pontine glioma, Biol. Res. 51 (1) (2018) 26.
[29] G. Curigliano, Gyneco-oncological genomics and emerging biomarkers for cancer treatment with immune-checkpoint inhibitors, Semin. Cancer Biol. 52 (Pt 2) (2018) 253-258.
[30] H.J. Kim, H.S. Kim, Y.H. Hong, Sulfatase 1 and sulfatase 2 as novel regulators of macrophage antigen presentation and phagocytosis, Yeungnam Univ. J. Med. 38 (4) (2021) 326-336.
[31] X. Niu, et al., Dermal mesenchymal stem cells: a resource of migration-associated function in psoriasis? Stem Cell Res. Ther. 10 (1) (2019) 54.
[32] Y.S. Lee, et al., Human umbilical cord blood-derived mesenchymal stem cells ameliorate psoriasis-like skin inflammation in mice, Biochem. Biophys. Rep. 9 (2017) 281-288.
[33] J. Bedke, et al., Updated European association of urology guidelines on renal cell carcinoma: nivolumab plus cabozantinib joins immune checkpoint inhibition combination therapies for treatment-naïve metastatic clear-cell renal cell carcinoma, Eur. Urol. 79 (3) (2021) 339-342.
[34] A.S. Truong, et al., Entinostat induces antitumor immune responses through immune editing of tumor neoantigens, J. Clin. Invest. 131 (16) (2021).
[35] Y. Zhang, et al., TP53 loss-of-function causes vulnerability to autophagy inhibition in aggressive prostate cancer, Int. J. Urol. 29 (9) (2022) 1085-1094.
[36] Y. Guo, et al., Three prognostic biomarkers correlate with immune checkpoint blockade response in bladder urothelial carcinoma, Int. J. Genomics 2022 (2022), 3342666.