Pan-cancer analysis of the oncogenic effects of G-protein-coupled receptor kinase-interacting protein-1 and validation on liver hepatocellular carcinoma
*Tao Wang1,A, *Kun Su1,C,D, Lianming Wang1,C, Yanmei Shi2,D, Yichun Niu2,8, Yahao Zhou3,B, Ayong Wang3,C, Tao Wu1,E,F
1 Department of Hepatobiliary Surgery, The Second Affiliated Hospital of Kunming Medical University, China
2 Department of Gastroenterology, The Second Affiliated Hospital of Kunming Medical University, China
3 Department of Hepatobiliary Surgery, Puer People’s Hospital, China
A - research concept and design; B - collection and/or assembly of data; C - data analysis and interpretation;
D - writing the article; E - critical revision of the article; F - final approval of the article
Advances in Clinical and Experimental Medicine, ISSN 1899-5276 (print), ISSN 2451-2680 (online)
Adv Clin Exp Med. 2023;32(10):1139-1147
| Address for correspondence Tao Wu | Abstract |
|---|---|
| E-mail: taowubio@outlook.com | Background. Despite G-protein-coupled receptor kinase-interacting protein-1 (GIT1) being recognized as a new promoter gene in some types of cancer, its effect on human pan-cancers and liver hepatocellular |
| Funding sources The study was supported by the major provincial | carcinoma (LIHC) remains unclear. |
| science and technology projects of Yunnan (grant No. 202002AA100007). Conflict of interest None declared | Objectives. To elucidate the molecular mechanisms of GIT1 in pan-cancer and LIHC. |
| Materials and methods. Various bioinformatics approaches were utilized to elucidate the oncogenic effects of GIT1 on human pan-cancers. | |
| *Tao Wang and Kun Su contributed equally to this work. | Results. The GIT1 was aberrantly expressed in pan-cancers and associated with the clinical stage. More- over, the upregulation of GIT1 expression was indicative of poor overall survival (OS) in patients with LIHC, skin cutaneous melanoma (SKCM) and uterine corpus endometrial carcinoma (UCEC), as well as of poor |
| Received on May 18, 2022 Reviewed on June 13, 2022 Accepted on February 12, 2023 | disease-free survival (DFS) in patients with LIHC and UCEC. Furthermore, GIT1 levels were correlated with |
| cancer-associated fibroblasts (CAFs) in adrenocortical carcinoma (ACC), cervical squamous cell carcinoma | |
| (CESC) and LIHC. The analysis of single-cell sequencing data revealed an association of GIT1 levels with | |
| Published online on March 30, 2023 | apoptosis, cell cycle and DNA damage. In addition, multivariate Cox analysis indicated that high GIT1 levels were an independent risk factor for shorter OS in patients with LIHC. Finally, the gene set enrichment analysis revealed INFLAMMATORY_RESPONSE pathway and IL2_STAT5_SIGNALING to be the most enriched in LIHC. |
| Conclusions. Our data demonstrate the oncogenic effects of GIT1 on various cancers. We believe that GIT1 can serve as a biomarker for LIHC. | |
| Key words: pan-cancer analysis, GIT1, oncogene, liver hepatocellular carcinoma |
Cite as
Wang T, Su K, Wang L, et al. Pan-cancer analysis of the oncogenic effects of G-protein-coupled receptor kinase-interacting protein-1 and validation on liver hepatocellular carcinoma. Adv Clin Exp Med. 2023;32(10):1139-1147. doi:10.17219/acem/161157
DOI 10.17219/acem/161157
Copyright
Copyright by Author(s) This is an article distributed under the terms of the Creative Commons Attribution 3.0 Unported (CC BY 3.0) (https://creativecommons.org/licenses/by/3.0/)
Background
Liver cancers are associated with elevated mortality rates across the world,1-3 and while significant advancements have been made in surgical techniques, chemotherapy and other treatment approaches, the 5-year survival rate remains far from satisfactory.4-6 Moreover, liver cancer is the most common type of cancer in China. Specifically, cancer recurrence at the intermediate or advanced stage occurs in approx. half of the patients. Considering the in- crease in the incidence and mortality rates of liver cancer, it is crucial to identify new prognostic biomarkers.
G-protein-coupled receptor kinase-interacting protein-1 (GIT1) has been shown to repress the ß2-adrenergic re- ceptor pathway and stimulate receptor phosphorylation. Many proteins interact with GIT1 via its various domains. Notably, GIT1 is essential for focal cell migration, adhe- sion and the development of lamellipodia. The principal roles of GIT1 include focal adhesion remodeling,7 recep- tor internalization and transmission of cellular signals.8 The GIT1 is widely expressed in the brain, liver, lungs, nerves, and blood vessels.9,10 The expression of GIT1 is up- regulated in breast cancer, while its downregulation has been found to regulate the cell progression of breast can- cer.11 The GIT1 can stimulate tumor development by ac- tivating extracellular signal-regulated kinase signaling in hepatocellular carcinoma.12,13 Moreover, GIT1 partici- pates in epithelial-mesenchymal transition and promotes the invasion of oral squamous cell carcinoma.14 Interest- ingly, this protein is involved in a number of varied cel- lular processes, including enhancing neurite and spine maturation,15 mediating vascular intima and pulmonary vasculature development,16 as well as cell migration and adhesion.17 While the overexpression of GIT1 has been shown to regulate chondrocyte proliferation and apoptosis via integrin-ß1, it also increases autophagy via disrup- tion of the Beclin-1 and Bcl-2 interaction in osteoclast. Mechanistically, GIT1 achieves these outcomes by altering ERK1/2, AKT, NF-KB, and Notch expression, and acceler- ating lung cancer cell migration and metastasis via Rac1/ Cdc42 signal, which further validates its participation in cancer occurrence and development.18-21 A previous study found that the suppression of GIT1 inhibits breast cancer cell invasion and metastasis via the upregulation of miR-149.22 Recently, a report demonstrated that GIT1 is reduced in ER(-) breast cancer when compared to ER(+) cancer, and that higher GIT1 expression implied a better prognosis in ER(-) breast cancer patients.11 Thus, GIT1 appears to have distinct functions in the growth and migration of breast cancer cells. However, its roles and mechanisms in pan-cancer demand further investigations.
Herein, we investigated various cancers for GIT1 expres- sion and patient survival data. To elucidate the mechanisms of GIT1 and the associated proteins, we performed Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway
analysis and Gene Set Enrichment Analysis (GSEA). Fur- thermore, we evaluated the association between GIT1 lev- els and immune infiltration. Finally, single-cell sequencing results were assessed to examine the GIT1 expression in cells in related tumors.
Objectives
This study aimed to measure the expression of GIT1 in various cancers and the association between GIT1 levels and immune infiltration.
Materials and methods
Pattern of GIT1 expression based on the pan-cancer study
The GIT1 level patterns in cancer and corresponding samples were obtained using ONCOMINE (http://www. oncomine.org/resource/login.html) and TIMER2.0 (http:// timer.comp-genomics.org/). For ONCOMINE, the pa- rameters were set as p = 0.001, fold change: 2.0 and gene ranking: top 10%. The GIT1 level patterns in different can- cer stages were acquired using the “Stage plots” module of GEPIA2 (http://gepia2.cancer-pku.cn/#index).
Survival and prognosis
Both overall survival (OS) and disease-free survival (DFS) results were obtained through the GEPIA website.22 High and low GIT1 expression groups were established based on the median level of GIT1. The association between GIT1 levels and pan-cancer survival outcome was detected using the log-rank test. Furthermore, Cox regression examining GIT1 levels and the clinical variables was used to detect the effects of GIT1 on the prognostic value of liver hepato- cellular carcinoma (LIHC) patients. Calibration curves and the concordance index (C-index) were evaluated by com- paring predicted probabilities with the observed events.
GIT1-associated functional enrichment
Proteins interacting with GIT1 were analyzed using the STRING tool (http://string.embl.de/)23 under the set- ting of no more than 100 interactors and low confidence (0.150) to obtain the potent GIT1-binding proteins. Fur- thermore, the top 100 genes demonstrating an expression profile similar to that of GIT1 in various cancers were ana- lyzed with the GEPIA2 tool. Then, Gene Ontology (GO) and KEGG pathway enrichment analyses were performed using proteins interacting with GIT1, together with the top 100 genes, using the DAVID software. A p-value of <0.01 was considered statistically significant.
Immune infiltration
The relationship between GIT1 expression, immune infiltration and cancer-associated fibroblasts (CAFs) was analyzed with TIMER24 using Spearman’s correla- tion based on the ranked values. The p-values and partial correlation values were measured employing the purity- adjusted Spearman’s rank correlation test, and data were visualized with heat maps and scatter plots. Furthermore, the relationship between GIT1 levels and various tumor immune subtypes was investigated through the TISDB tool (http://cis.hku.hk/TISIDB/index), and the distribution of the 6 immune subtypes was determined. The TISDB is an online tool for cross-linking studies of tumors and immunity, which contains data from PubMed, The Cancer Genome Atlas (TCGA) and other public databases.25,26
Single-cell sequencing results
The distinct functional states of various cancer cells at single-cell level,27 and the association of GIT1 levels and pan-cancer functional status were obtained through the “correlation plot” module of CancerSEA (http://biocc. hrbmu.edu.cn/CancerSEA).28 The threshold for the asso- ciation between GIT1 and cancer functional states was set as a correlation strength >0.3 and a p-value <0.05.
GSEA
The GSEA is a method to demonstrate that the expres- sion of a given gene set is overrepresented. The GSEA was employed to evaluate distinct functions among the high- and low-risk score subgroups, using the hallmark gene set h.all.v7.0.symbols.gmt. Gene sets with |normalized enrich- ment score (NES)| > 1, nominal (NOM) p < 0.01 and false discovery rate (FDR) q < 0.25 were considered significant.
Statistical analyses
To assess the different levels of GIT1 in normal and pan- cancer samples, we used Wilcoxon rank-sum test. Cancer pa- tient survival was detected with the Kaplan-Meier curve, and Spearman’s rank correlation coefficient was used to measure the correlation between the 2 groups. The statistical analysis was performed using R software v. 3.6.3 (R Foundation for Statistical Computing, Vienna, Austria) and the ‘edgeR’ pack- age. A value of p < 0.05 was considered statistically significant.
Results
Abnormal expression of GIT1 in different cancers
The GIT1 expression patterns were evaluated in pan- cancer through TIMER2.0, which includes data about gene
expression patterns in normal and pan-cancer samples. We found that GIT1 expression levels were significantly upregulated in various cancers, including LIHC and lung adenocarcinoma (LUAD), among others (Fig. 1A).
Next, we used GEPIA2 to investigate the correlation between GIT1 levels and clinical stage. An association between GIT1 levels and clinical stage for glioblastoma multiforme, head and neck squamous cell carcinoma, kid- ney chromophobe (KICH), LIHC, lung squamous cell car- cinoma, and others was found (Fig. 1B). Collectively, these data indicate that GIT1 expression is upregulated in pan- cancer and that GIT1 can be a promotor of pan-cancers.
Based on the above data, it became evident that GIT1 is involved in both pan-cancer and LIHC development and can thus serve as a potential biomarker.
Relationship between GIT1 levels and patient prognosis
To study the correlation between GIT1 levels and patient prognosis, we used GEPIA2 to conduct a survival investi- gation. The obtained data showed that the overexpression of GIT1 was indicative of poor OS in patients with LIHC (p = 0.002), skin cutaneous melanoma (SKCM) (p = 0.026) and uterine corpus endometrial carcinoma (UCEC) (p = 0.006). Conversely, better OS was found in patients with kidney renal clear cell carcinoma (KIRC) (p = 0.011) and glioma (p < 0.001) (Fig. 2A). Furthermore, the overex- pression of GIT1 was associated with poor DFS in patients with LIHC and UCEC, and improved DFS in those with KIRC, SKCM and glioma (Fig. 2B). These data indicate that there is a close association of GIT1 overexpression with poor survival outcomes in some types of cancers, including LIHC.
Protein-protein interaction and enrichment pathway analyses
Unfortunately, the mechanism underlying GIT1-medi- ated oncogenesis remains unknown. To examine the pro- tein-protein interaction (PPI) network and enrichment signal of GIT1, proteins that bind GIT1 were obtained from the STRING database, and the database was veri- fied using the experimental setup. Eleven proteins were found to interact with GIT1, namely ADRBK1, ARHGEF6, ARHGEF7, CAMK4, ERC2, LPXN, PAK1, PAK2, PPFIA1, PTK2, and PXN (Fig. 3A).
Then, the top 100 proteins that closely interacted with GIT1 were found using GEPIA2, with PXN found to be common to both methods. Furthermore, GO and KEGG pathway enrichment analyses indicated that the above genes were involved in several cellular processes, includ- ing regulation of GTPase activity, microtubule polymer- ization/depolymerization, protein kinase activator activ- ity, and Ras GTPase binding, among others (Fig. 3B,C). In addition, KEGG data revealed that GIT1 participated
A
10
ns
ns **
**
ns
The expression of GIT1 Log2 (TPM+1)
8
4
:
Normal
6
-
E
Tumor
3
=
T
E
-
I
I
C
.
·
ES
U
4
-
·
2
ACC
BLCA
BRCA
CESC
CHOL
COAD
DLBC
ESCA
GBM
HNSC
KICH
KIRC
KIRP
LAML
LGG
LIHC
LUAD
LUSC
MESO
OV
PAAD
PCPG
PRAD
READ
SARC
SKCM
STAD
TGCT
THCA
THYM
UCEC
UCS
UVM
B
ACC ;; GIT1_exp
Pv=2.76e-03 n=CIMP-high 19, CIMP-intermediate 27,
BRCA= GIT1_exp
Pv=3.24e-19
n=Basal 172, Her2 73.
Expression (log2CPM)
COAD : GIT1_exp Pv=2.82e-03 nac N 226, GS 49, HM-SNV 6, HM-indel 60
ESCA :: GIT1_exp
LumA 508, LumB 191, Normal 137
Pv=3.97e-03 RECIN 74. ESCC 90, GS 1.
GBM :: GIT1_exp
PV=3.25e-01
n=Classic-like 47.
CIMP-low 32
HM-SNV 2, HM-indel 2
Expression (log2CPM)
G-CIMP-high 2, G-CIMP-low 5. LGm6-GBM 12
7
Expression (log2CPM)
Expression (log2CPM)
Mesenchymal-like 53
11
Expression (log2CPM)
8
6
H
1
9
7.5
8
H
8
P
8
H
8
H
®
8
H
6
8
5
7
5.0
6
H
H
8
H
5.
CIMP-high
CIMP-intermediate
CIMP-low
5
2.5
4.
4
HM-SNV
HM-indel
HM-SNV
HM-indel
Classic-like
G-CIMP-high
G-CIMP-low
LGm6-GBM
Mesenchymal-like
Basal
Her2
LumA
LumB
Normal
CIN
GS
CIN
ESCC
GS
Subtype
Subtype
Subtype
Subtype
Subtype
HNSC = GIT1_exp
Pys7.fre-05 Paraay Classical 48 Mesenchymal 74
KIRP :: GIT1_exp Pv=8.310-05
LIHC : GIT1_exp
Pv=5.7e-04
LUSC : GIT1_exp Pv=5.450-09
STAD :: GIT1_exp
Pv=1.490-09
med C2b 22.
iCluster:2 55 iCluster:3 63
classical 63
Expression (log2CPM)
C2c-CIMP 9
primitive 26. secretory 39
EBV 30 SEO
Expression (log2CPM)
Expression (log2CPM)
Expression (log2CPM)
HM-SNV 7 HM-indel 73
9
8
9
7
Expression (log2CPM)
8
7
8
6
8
:
N
8
H
-
7
I
Y
7
i
6-
7
:
I
U
5
!
Y
!
H
5
H
6
I
6
H
6
4.
5
5
4
Atypical
Basal
Classical
Mesenchymal
5
4.
4
C1
C2a
C2c-CIMP
3.
iCluster:1
iCluster:2
iCluster:3
basal
classical
primitive
secretory
CIN
EBV
GS
HM-SNV
HM-indel
Subtype
Subtype
Subtype
Subtype
Subtype
A
1.0
KIRC
GIT1
1.0
LIHC
GIT1
1.0
SKCM
GIT1
1.0
UCEC
GIT1
1.0
glioma
GIT1
Low
Low
Low
Low
Low
High
High
High
High
High
Survival probability
0.8
Survival probability
0.8
Survival probability
0.8
Survival probability
0.8
Survival probability
0.8
0.6
0.6
0.6
0.6
0.6
OS
0.4
0.4
0.4
0.4
0.4
0.2
Overall Survival HR = 0.67 (0.50-0.91)
0.2
Overall Survival HR = 1.76 (1.24-2.50)
0.2
Overall Survival HR = 1.36 (1.04-1.79)
0.2
Overall Survival HR = 1.81 (1.19-2.74)
0.2
Overall Survival HR = 0.42 (0.33-0.54)
0.0
P = 0.011
0.0
P = 0.002
0.0
P = 0.026
0.0
P = 0.006
0.0
P < 0.001
0
50
100
150
0
30
60
90
120
0
100
200
300
0
50
100
150
200
0
50
100
150
200
Time (months)
Time (months)
Time (months)
Time (months)
Time (months)
B
1.0 -
KIRC
GIT1
1.0
LIHC
GIT1
1.0 -
SKCM
GIT1
1.0 -
UCEC
GIT1
1.0 -
glioma
GIT1
Low
Low
Low
Low
Low
High
High
High
High
High
Survival probability
0.8
Survival probability
0.8
Survival probability
0.8
Survival probability
0.8
Survival probability
0.8
0.6
0.6
0.6
0.6
0.6
DS
0.4
0.4
0.4
0.4
0.4
0.2
Disease Specific Survival HR - 0.53 (0.35-0.79)
0.2
Disease Specific Survival HR = 1.99 (1.26-3.14)
0.2
Disease Specific Survival HR - 0.53 (0.35-0.79)
0.2
Overall Survival HR = 1.81 (1.19-2.74)
0.2
Disease Specific Survival HR - 0.42 (0.33-0.55)
0.0
P = 0.002
0.0
P = 0.003
0.0
P = 0.002
0.0
P = 0.006
0.0
P < 0.001
0
50
100
150
0
30
60
90
120
0
50
100
150
0
50
100
150
200
0
50
100
150
200
Time (months)
Time (months)
Time (months)
Time (months)
Time (months)
A
TGFB111
PXN
B
PTK2
8
Renal cell carcinoma
PAK1
ARHGER
NICKIPSD
2
Yersinia infection
PAK2
ARHGEF
Regulation of actin cytoskeleton
SDCCAG3
CAMK4
LPXN
SCRIB
GIT2
M
Ras GTPase binding
GIT1
kinase activator activity
BP
PCLO
ADRBK1
protein kinase activator activity
CC
MF
PPFIA4
ERC2
SNX6
ATPase complex
KEGG
f
PPFIA1
SWI/SNF superfamily-type complex
PPFIAZ
lamellipodium
C
microtubule polymerization or depolymerization
regulation of microtubule polymerization or
microtubule depolymerization
lamellipodium
depolymerization
regulation of microtubule polymerization or degos superfamily-type complex
ATPase complex
microtubule polymerization or depolymerization
protein kinase activator activity
kinase activator activity
microtubule depolymerization
Ras GTPase binding
Counts
Regulation of actin cytoskeleton
4
Yersinia infection
6
0
1
2
3
4
Renal cell carcinoma
8
-Log 10 (p.adjust)
10
in tumorigenesis of renal cell carcinoma, regulation of the actin cytoskeleton, focal adhesion, and other sig- naling pathways (Fig. 3B,C). Thus, these data found that GIT1 together with its closely interacting partner proteins correlated with focal adhesion and regulation of the ac- tin cytoskeleton, which implied an increased complexity of the GIT1-mediated signal network.
Relationship between GIT1 levels and tumor microenvironment
To explain the effect of GIT1 expression on the immune microenvironment, TIMER was applied to study the asso- ciation between GIT1 levels and tumor microenvironment (TME) characteristics in various cancers. We found that GIT1 levels were correlated with CAFs in adrenocorti- cal carcinoma (ACC), cervical squamous cell carcinoma (CESC) and LIHC (Fig. 4A).
To further explore the relationship between GIT1 levels and CAFs, we examined biomarkers of CAF levels in differ- ent cancers and found that GIT1 expression was associated with the C1-C6 immune subtypes (Fig. 4B). Interestingly, the GIT1 levels were also connected with those immune subtypes in LIHC.
GIT1 expression pattern at the single-cell level and its relationship with biological functions
We validated GIT1 expression at the single-cell level across pan-cancers and determined its association with biological functions. The GIT1 levels were found to be positively cor- related with acute lymphocytic leukemia (ALL), LUAD and ovarian serous cystadenocarcinoma (OV) apoptosis. Spe- cifically, GIT1 expression was correlated with the LUAD cell cycle and retinoblastoma DNA damage (Fig. 5A).
There was an association between GIT1 levels and pro- liferation, epithelial-mesenchymal transition and metas- tasis in ALL (Fig. 5B). Moreover, t-distributed stochastic neighbor embedding (t-SNE) diagrams revealed GIT1 ex- pression patterns in single cells in ALL, colorectal cancer (CRC), LUAD, and glioma (Fig. 5C). Collectively, these data suggest that GIT1 participates in mediating cancer development.
Cox regression study
A nomogram was developed for internal validation, and a predictive model was prepared (Fig. 6A). We found
A
Cancer associated fibroblast_MCPCOUNTER
☒ p > 0.05
p ≤ 0.05
Cancer associated fibroblast_EPIC
Cancer associated fibroblast_XCELL
Cancer associated fibroblast_TIDE
Partial_Cor
1
0
GIT1 Expression Level (log2 TPM)
Purity
Cancer associated fibroblast_EPIC
GIT1 Expression Level (log2 TPM)
Purity
cer associated fibroblast_MCPCOUNT
GIT1 Expression Level (log2 TPM)
Purity
Cancer associated fibroblast_TIDE
Rno - 8482
-1
= 1.200-01
19 8459
Q
”
-
2-
1.
A
ACC (n=79)
AC
ACC
ACC
ACC
BLCA (n=408)
BRCA (n=1100)
BRCA-Basal (n=191)
:
BRCA-Her2 (n=82)
·
.
*
.
.
+
+
BRCA-LumA (n=568)
0.2
0.4
0.6
0.8
1.0 0.0
0.1
0.2
0.3
0.2
0.4
0.6
0.8
1.0 0
5000
10000
0.2
0.4
0.6
0.8
1.0
-0.1
0.0
0.1
0.2
BRCA-LumB (n=219)
Purity
Infiltration Level
Purity
Infiltration Level
Purity
Infiltration Level
CESC (n=306)
CHOL (n=36)
COAD (n=458)
DLBC (n=48)
X
X
ESCA (n=185)
GIT1 Expression Level (log2 TPM)
Purity
Cancer associated fibroblast_EPIC
GIT1 Expression Level (log2 TPM)
Purity
cer associated fibroblast_MCPCOUNT
GIT1 Expression Level (log2 TPM)
Purity
Cancer associated fibroblast_TIDE
GBM (n=153)
HNSC (n=522)
HNSC-HPV- (n=422)
HNSC-HPV+ (n=98)
KICH (n=66)
CESC
CES
-
CESC
CESC
KIRC (n=533)
KIRP (n=290)
4
LGG (n=516)
LIHC (n=371)
1
*
.
*
+
LUAD (n=515)
0.25
0.50
Purity
0.75
1.000.0
0.1
0.2
0.3
0.25
0.50
5000
10000
Infiltration Level
0.75
1.00
0
Infiltration Level
1500
5
Purity
0.25
0.50
0.75
1.00
-0.2
0.0
0.2
0.4
LUSC (n=501)
Purity
Infiltration Level
MESO (n=87)
XIXIXI
OV (n=303)
PAAD (n=179)
PCPG (n=181)
X
GIT1 Expression Level (log2 TPM)
Purity
Cancer associated fibroblast_EPIC
GIT1 Expression Level (log2 TPM)
Purity
cer associated fibroblast_MCPCOUNT
GIT1 Expression Level (log2 TPM) 1 -
Purity
Cancer associated fibroblast_TIDE
PRAD (n=498)
HhOCH
5841
Aho OCH
1192047
2
Ame- 90 H
p == 1.060-01
:
ATY -80.33
READ (n=166)
8
-
+
1
SARC (n=260)
X
SKCM (n=471)
:
SKCM-Metastasis (n=368)
LIHC
LIHC
4
LIHC
SKCM-Primary (n=103)
LIHC
STAD (n=415)
W
TGCT (n=150)
X
.
$
THCA (n=509)
THYM (n=120)
0.25
0.50
Purity
0.75
1.00 0.0
0.1
0.2
0.3
0.25
0.50
Infiltration Level
Purity
0.75
1.00 0
2500
5000
7500
10001
0.25
0.50
0.75
0.2
UCEC (n=545)
Infiltration Level
Purity
1.00-0.2
Infiltration Level
0.0
UCS (n=57)
X
UVM (n=80)
X
B
ACC :: GIT1_exp
BLCA: GIT1_exp Pv=6.07e-01
BRCA : GIT1_exp Pv=5.2e-14
CESC = GIT1_exp Pv=5.5-4e-01 n=C1 77,C2 217,C4 6
CHOL = GIT1_exp Pvm4.95e-01
COAD = GIT1_exp Pv= 1.22e-01
n=C1 1,C2 1,C3 23,C4 49,C5 3,C6 1
n=C1 173,C2 164,C3 21,C4 36,C6 3
n=C1 369,C2 390,C3 191,C4 92,06 40
n=C1 7,C2 2,C3 17,C4 8,C6 1
n=C1 332,C2 85,C3 9,C4 12,C6 3
Expression (log2CPM)
10
3
Expression (log2CPM)
Expression (log2CPM)
Expression (log2CPM)
8
Expression (log2CPM)
8
Expression (log2CPM)
10.0
9
H
7
7.5
8
8
8
T
6
5
I
8
0
1
7
9
H
H
H
!
8
H
Z
H
6
₿
H
5.0
6
8
6
I
CA
5
2.5
5
4
C1
C2
C3
C4
C5
C6
4
C1
C2
C3
C4
C6
3
C1
C2
C3
C4
C6
C1
C2
C4
C1
C2
C3
C4
C6
C1
C2
C3
C4
C6
Subtvne
Subtvoe
Subtvne
Subtvoe
Subtvne
Subtvoe
Esune exp Py=1.944e-01 nºC1 71,C2 87,C3 7,C4 6,C6 2
GEM : GITT exp Pv=2.48e-01 nºC1 2,C4 150,C5 1
HINSCH OUT exp
KICH = GITI exp Pv=5.66c-02 nºC1 2,C3 38,C4 12,C5 13
KIRCE GITT exp
KIR BUL SEND
nºC1 128,C2 379,C3 2,C4 2,C6 3
nºC1 7,C2 20,C3 445,C4 27,C5 3,C6 13
4 6 2-646-02 n=C1 3,C2 4,C3 202,C4 66,C5 2,C6 2
Expression (log2CPM)
Expression (log2CPM)
8
Expression (log2CPM)
Expression (log2CPM)
8
Expression (log2CPM)
9
Expression (log2CPM)
8
7.5
8
!
Z
7
7
7
8
U
H
-
O
8
7
A
i
₿
-
5.0
6
6
9
6
i
P
HI
H
6
5
5
2.5
5
3
4
4
C1
C2
C3
C4
C6
4
C1
C4
C5
C1
C2
C3
C4
C6
C1
C3
C4
C5
C1
C2
C3
C4
C5
C6
C1
C2
C3
C4
C5
C6
Subtype
Subtype
Subtype
Subtype
Subtype
Subtype
LGG = GITI exp Pv-2.366-09
LIHC :: GIT1 exp Pv-3 840-09
LUAD : GIT1_exp Py-3.11e-04
LUSC = GITI exp Pv-2.470-03
UCEC :: GIT1 exp
STAD = GITI exp Pv=1,036-05
10
n=C3 10,C4 147,C5 356,C6 1
n=C1 22,C2 45,C3 135,C4 159,C6 1
nºC1 83,C2 147,C3 179,C4 20,06 28
n=C1 275,C2 182,C3 8,C4 7,C6 14
Pv=4.27e-01
nºC1 247,C2 212,C3 52,C4 16,C6 1
m=C1 129.C2 210.C3 36,C4 9,C6 7
Expression (log2CPM)
Expression (log2CPM)
Expression (log2CPM)
Expression (log2CPM)
Expression (log2CPM)
9
Expression (log2CPM)
9
7
8
8
9
8
8
6
7
8
!
-
1
H
7
!
B
H
H
B
5
H
6
T
H
6
4
7
H
T
H
U
1
*
7-
H
H
6
H
A
6
4
S
4
6
5
4
0
3
4
5
C3
C4
C5
C6
C1
C2
C3
C4
C6
C1
C2
C3
C4
C6
C1
C2
C3
C4
C6
C1
C2
C3
C4
C6
C1
C2
C3
C4
C6
Subtype
Subtype
Subtype
Subtype
Subtype
Subtype
ACC - adrenocortical carcinoma; LIHC - liver hepatocellular carcinoma; CESC - cervical squamous cell carcinoma.
that the C-index of the nomogram was 0.669 (95% confi- dence interval (95% CI): 0.637-0.701), and the calibration curve displayed the nomogram’s desirable prediction for 1-5-year clinical consequences (Fig. 6B). Altogether, these data indicate that GIT1 may be a potent biomarker for LIHC.
GSEA
The GSEA revealed that the INFLAMMATORY_RE- SPONSE pathway and IL2_STAT5_SIGNALING were the most enriched in LIHC (Fig. 6C,D). Taken together, obtained data revealed that GIT1 expression was associated with specific gene signatures of these key pathways in LIHC.
A
B
Angiogenesis
Inflammation
=
Apoptosis
CellCycle
Differentiation
DNAdamage
DNArepair
EMT
Hypoxia
Invasion
Metastasis
Proliferation
Quiescence
Stemness
geneExp
4
No
signifitam datasets
positive negtive
Correlation
Pvalue
S
Proliferation
-0.40
0
AML
-1.0
1.0
Blood
ALL
-0.8
0.8
-0.5
0.5
CML
-0.3
0.3
EMT
-0.37
**
1
CEM
0.0
.
0.0
Correlation
Clioma
CNS/brain
AST
Metastasis
-0.32
**
HCG
ODG
C
Expression distribution with t-SNE plot
Expression distribution with t-SNE plot
15
25
LUAD
Lung
M
5
NSCLC
3
25
Skin
MEL
·
Kidney
RCC
4
Breast
BRCA
-
1
A
Head and neck
HNSCC
-
-UP
-5
4
5
18
15
-75
-5
-25
4
25
$
75
Ovary
OV
Bowel
CRC
Expression distribution with t-SNE plot
Expression distribution with t-SNE plot
M
RB
Eye
100
2
UM
10
. L
M
5
-30
4
-
-20
-310
-
.
50
150
-
-10
4
.
5
10
15
A
B
40
60
BO
100
1.0
Points
0
20
T2
Observed fraction survival probability
T stage
T1
N1
T3&T4
0.8
N stage
M stage
NO
M1
MO
Lov
0.6
ERBB2
High
TP53
5
1
5
0.4
Total Points
0
40
80
120
160
200
Linear Predictor
1-Year
-1
0.6
-0.2
0.6
1.4
0.2
3-Year
3-year Survival Probability
0.2
1
5-Year
5-year Survival Probability
0.8
0.6
0.4
0.2
Ideal line
0.8
0.6
0,4
02
0.2
0.4
0.6
0.8
1.0
Nomogram predicted survival probability
C
D
0.0
Enrichment Score
HALLMARK INFLAMMATORY RESPONSE
02
HALLMARK 12 STAS SIGNALING
0.4
HALLMARK ALLOGRAFT REJECTION
0.6
ــ ·
-HALLMARK INFLAMMATORY_RESPONSE
HALLMARK COMPLEMENT
Ranked list metric
2
25
-20
-45
-10
2
2
10000
20000
30000
Rank in Ordered Dataset
NES - normalized enrichment score; NOM - nominal; FDR - false discovery rate.
Discussion
Previous studies have reported a close relationship between aberrant gene expression and the development of pan-cancers. Investigations into pan-cancer provide deep insights into the molecular mechanism underly- ing different malignancies and are useful to identify new therapeutic markers for cancer treatment.29 Therefore, we investigated the expression and predictive significance of GIT1 in different tumors.
Firstly, we found that GIT1 is aberrantly expressed in different cancers, including LIHC. To further examine the prognostic value of GIT1, a Kaplan-Meier survival study was performed, which revealed an association be- tween high GIT1 levels and the poor outcomes associ- ated with pan-cancers, including LIHC. Thus, we found that the overexpression of GIT1 could be an independent indicator of poor prognosis in patients with LIHC and other cancers. Moreover, Cox regression analysis verified that GIT1 overexpression may be a risk factor for LIHC. Thus, our data suggested that GIT1 is a pro-oncogene in pan-cancers.
Recently, Chen et al. described a prognostic model for OS which included age and other factors for pan-cancer based on GIT1 expression.30 In accordance with that model, we developed a prognostic nomogram model in- cluding clinical stage and GIT1 levels, which may increase
the accuracy of classifying high-risk cases. This model further assessed the relationship between clinical features and GIT1 levels in cases of LIHC, and demonstrated that increased GIT1 levels were associated with the clinical stage. The results revealed that GIT1 could act as a potent biomarker for different cancers, especially LIHC.
Additionally, the enrichment analyses revealed that GIT1 may impact cancer development through the regu- lation of focal adhesions and the actin cytoskeleton, to- gether with their associated pathways. Chen et al. have shown that these signals have a key role in the development of pan-cancers.31
The TME has been shown to promote crosstalk be- tween cancer cells and other cell types. In fact, CAFs have been reported to have a functional role in stimu- lating tumorigenesis. Thus, the signature of pan-CAF is associated with poor survival in cancer. Interestingly, other studies have suggested that CAFs inhibit cancer development, which implies that they have an antitumor effect.32-34 Our results indicated an association between GIT1 levels and CAFs in different cancers, and therefore we believe that GIT1 mediates the development of pan- cancers. However, the molecular mechanism regarding how GIT1 modulates CAFs warrants further investiga- tion. The well-defined immune subtype in various can- cers could improve the effectiveness of targeted immune treatment. We found that GIT1 is aberrantly expressed in various immune subtypes of pan-cancer, which poten- tially makes it an important target in immune therapies aimed at various cancers.
Considering the complex nature of cancer cells, the uti- lization of single-cell transcriptomic data is a valuable method of examining various types of cancers. To elu- cidate the effect of GIT1 on pan-cancer progression, the CancerSEA website was used. The GIT1 expression was found to be positively associated with ALL, LUAD and OV apoptosis, and specifically positively associated with the LUAD cell cycle. Furthermore, an association was found between GIT1 levels and cell proliferation, epi- thelial-mesenchymal transition, and metastasis in ALL. However, the mechanism underlying GIT1 in pan-cancer warrants further investigation.
Finally, GSEA results indicated that GIT1 was associ- ated with the inflammatory response pathway and IL2/ STAT5 signaling in LIHC. These signals have been shown to be actively involved in the development of pan-cancers, including LIHC.
Recently, advances in the prediction abilities of com- putational biology began to offer new understanding of biomarkers and non-coding RNAs connected to pan- cancers, including ceRNA network prediction. A previ- ous report presented data highlighting that GIT1 was involved in the ceRNA network, and, consequently, more research is necessary to investigate the role of GIT1 in ceRNA interaction.
Limitations
Our findings were mostly obtained from online tools, and more results based on clinical cases are needed to fur- ther authenticate our findings. Furthermore, in vitro and in vivo analyses should be performed to confirm the role of GIT in LIHC progression.
Conclusions
We comprehensively investigated the effects of GIT1 on various cancers. Our findings revealed that GIT1 was overexpressed in different cancers, including LIHC, which was in turn associated with a poor prognosis. Furthermore, GIT1 was shown to mediate pan-cancer development, namely LIHC progression, through the regulation of focal adhesion and the actin cytoskeleton, inflammatory response pathways, and IL2/STAT5 signaling. Further studies are needed to elucidate the molecular mechanisms underlying GIT1, which appears valuable for cancer-targeted therapy.
Availability of data and materials
The datasets generated and/or analyzed in this study are available in the TCGA database (https://portal.gdc. cancer.gov/).
ORCID iDs
Tao Wang @ https://orcid.org/0009-0005-3799-6782 Kun Su ® https://orcid.org/0009-0006-2387-1441 Lianming Wang @ https://orcid.org/0009-0009-3706-0510 Yanmei Shi @ https://orcid.org/0009-0005-3990-6998 Yichun Niu @ https://orcid.org/0009-0008-0242-8496 Yahao Zhou @ https://orcid.org/0009-0005-2388-9994 Ayong Wang ® https://orcid.org/0009-0006-8482-1691 Tao Wu ® https://orcid.org/0009-0002-1260-3645
References
1. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOB- OCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71(3):209-249. doi:10.3322/ caac.21660
2. Ferlay J, Colombet M, Soerjomataram I, et al. Cancer statistics for the year 2020: An overview. Int J Cancer. 2021;149(4):778-789. doi:10.1002 /ijc.33588
3. Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer Statistics, 2021. CA Cancer J Clin. 2021;71(1):7-33. doi:10.3322/caac.21654
4. Chen J, Liang H, Lu J, He Y, Lai R. Probiotics improve postoperative adaptive immunity in colorectal cancer patients: A systematic review and meta-analysis. Nutr Cancer. 2022;74(8):2975-2982. doi:10.1080/ 01635581.2022.2056619
5. Hong R, Sun H, Li D, et al. A review of biosensors for detecting tumor markers in breast cancer. Life (Basel). 2022;12(3):342. doi:10.3390/life 12030342
6. Zhang X, Yang Z, An Y, et al. Clinical benefits of PD-1/PD-L1 inhibitors in patients with metastatic colorectal cancer: A systematic review and meta-analysis. World J Surg Onc. 2022;20(1):93. doi:10.1186/s12957- 022-02549-7
7. Zhang N, Cai W, Ma Y, Yin G, Nagel D, Berk B. GIT1 is a novel MEK1- ERK1/2 scaffold that localizes to focal adhesions. Cell Biol Int. 2009; 34(1):41-47. doi:10.1042/CBI20090016
8. Segura I, Essmann CL, Weinges S, Acker-Palmer A. Grb4 and GIT1 transduce ephrinB reverse signals modulating spine morphogenesis and synapse formation. Nat Neurosci. 2007;10(3):301-310. doi:10.1038 /nn1858
9. Li L, Tang P, Zhou Z, et al. GIT1 regulates angiogenic factor secretion in bone marrow mesenchymal stem cells via NF-KB/Notch signalling to promote angiogenesis. Cell Prolif. 2019;52(6):e12689. doi:10.1111/ cpr.12689
10. Kim H, Kim JI, Kim H, Kim JW, Kim BN. Interaction effects of GIT1 and DRD4 gene variants on continuous performance test variables in patients with ADHD. Brain Behav. 2017;7(9):e00785. doi:10.1002/ brb3.785
11. Zhang S, Miyakawa A, Wickström M, et al. GIT1 protects against breast cancer growth through negative regulation of Notch. Nat Commun. 2022;13(1):1537. doi:10.1038/s41467-022-28631-y
12. Wang G, Bai X, Jiang G, et al. GIT1 overexpression promotes epithe- lial-mesenchymal transition and predicts poor prognosis in hepa- tocellular carcinoma. Bioengineered. 2021;12(1):30-43. doi:10.1080/ 21655979.2020.1855914
13. Chen J, Yang P, Yang J, Wen Z, Zhang B, Zheng X. GIT1 is a novel prog- nostic biomarker and facilitates tumor progression via activating ERK/MMP9 signaling in hepatocellular carcinoma. Onco Targets Ther. 2015;8:3731-3742. doi:10.2147/OTT.S96715
14. Huang WC, Chan SH, Jang TH, et al. miRNA-491-5p and GIT1 serve as modulators and biomarkers for oral squamous cell carcinoma inva- sion and metastasis. Cancer Res. 2014;74(3):751-764. doi:10.1158/0008 -5472.CAN-13-1297
15. Hong ST, Mah W. A critical role of GIT1 in vertebrate and inverte- brate brain development. Exp Neurobiol. 2015;24(1):8-16. doi:10.5607/ en.2015.24.1.8
16. Pang J, Hoefen R, Pryhuber GS, et al. G-protein-coupled receptor kinase interacting protein-1 is required for pulmonary vascular devel- opment. Circulation. 2009;119(11):1524-1532. doi:10.1161/CIRCULA- TIONAHA.108.823997
17. Zhu J, Zhou Q, Xia Y, et al. GIT/PIX condensates are modular and ideal for distinct compartmentalized cell signaling. Mol Cell. 2020;79(5):782. e6-796.e6. doi:10.1016/j.molcel.2020.07.004
18. Xu R, Xu R, Wang Y, Wang W, Jiang L, Gong S.G-protein-coupled recep- tor kinase-interacting protein 1 (GIT1) promotes head and neck squa- mous cell carcinoma metastases via activating the PI3K/AKT/mTOR signal pathway. Comput Math Methods Med. 2022;2022:6881932. doi:10.1155/2022/6881932
19. Ye X, Chen X. miR-149-5p inhibits cell proliferation and invasion through targeting GIT1 in medullary thyroid carcinoma. Oncol Lett. 2019;17(1):372-378. doi:10.3892/ol.2018.9628
20. Jiao Y, Zhang L, Li J, He Y, Zhang X, Li J. Exosomal miR-122-5p inhib- its tumorigenicity of gastric cancer by downregulating GIT1. Int J Biol Markers. 2021;36(1):36-46. doi:10.1177/1724600821990677
21. Chang JS, Su CY, Yu WH, et al. GIT1 promotes lung cancer cell metas- tasis through modulating Rac1/Cdc42 activity and is associated with poor prognosis. Oncotarget. 2015;6(34):36278-36291. doi:10.18632/ oncotarget.5531
22. Yang L, Zhang W, Li M, et al. Evaluation of the prognostic relevance of differential claudin gene expression highlights claudin-4 as being suppressed by TGFbeta1 inhibitor in colorectal cancer. Front Genet. 2022;13:783016. doi:10.3389/fgene.2022.783016
23. Ukey S, Ramteke A, Choudhury C, Purohit P, Sharma P. Differential expres- sion of zinc-dependent HDAC subtypes and their involvement in unique pathways associated with carcinogenesis. Asian Pac J Cancer Prev. 2022;23(3):877-883. doi:10.31557/APJCP.2022.23.3.877
24. Guo Y, Shang A, Wang S, Wang M. Multidimensional analysis of CHMP family members in hepatocellular carcinoma. Int J Gen Med. 2022; 15:2877-2894. doi:10.2147/IJGM.S350228
25. Wang Y, Zhang H. FAT10 is a prognostic biomarker and correlated with immune infiltrates in skin cutaneous melanoma. Front Mol Biosci. 2022;9:805887. doi:10.3389/fmolb.2022.805887
26. Wu H, Duan Y, Gong S, Zhu Q, Liu X, Liu Z. An integrative pan-can- cer analysis of kinesin family member C1 (KIFC1) in human tumors. Biomedicines. 2022;10(3):637. doi:10.3390/biomedicines10030637
27. Wang Y, Song M, Gao B. EF-hand domain-containing protein D2 (EFHD2) correlates with immune infiltration and predicts the prog- nosis of patients: A pan-cancer analysis. Comput Math Methods Med. 2022;2022:4878378. doi:10.1155/2022/4878378
28. Ma J, Jin Y, Gong B, Li L, Zhao Q. Pan-cancer analysis of necropto- sis-related gene signature for the identification of prognosis and immune significance. Discov Onc. 2022;13(1):17. doi:10.1007/s12672- 022-00477-2
29. Liu Y, Lin L, Zou R, Wen C, Wang Z, Lin F. MSC-derived exosomes pro- mote proliferation and inhibit apoptosis of chondrocytes via lncRNA- KLF3-AS1/miR-206/GIT1 axis in osteoarthritis. Cell Cycle. 2018;17(21-22): 2411-2422. doi:10.1080/15384101.2018.1526603
30. Chen J, Yang P, Yang J, Wen Z, Zhang B, Zheng X. GIT1 is a novel prog- nostic biomarker and facilitates tumor progression via activating ERK/MMP9 signaling in hepatocellular carcinoma. Onco Targets Ther. 2015;8:3731-3742. doi:10.2147/OTT.S96715
31. Chen S, Hourwitz MJ, Campanello L, Fourkas JT, Losert W, Parent CA. Actin cytoskeleton and focal adhesions regulate the biased migration of breast cancer cells on nanoscale asymmetric sawteeth. ACS Nano. 2019;13(2):1454-1468. doi:10.1021/acsnano.8b07140
32. Mizutani Y, Kobayashi H, Iida T, et al. Meflin-positive cancer-associ- ated fibroblasts inhibit pancreatic carcinogenesis. Cancer Res. 2019; 79(20):5367-5381. doi:10.1158/0008-5472.CAN-19-0454
33. Ishibashi Y, Mochizuki S, Horiuchi K, et al. Periostin derived from can- cer-associated fibroblasts promotes esophageal squamous cell car- cinoma progression via ADAM17 activation. Biochim Biophys Acta Mol Basis Dis. 2023;1869(5):166669. doi:10.1016/j.bbadis.2023.166669
34. Dong J, Wang F, Gao X, et al. Integrated analysis of genome-wide DNA methylation and cancer-associated fibroblasts identified prog- nostic biomarkers and immune checkpoint blockade in lower grade gliomas. Front Oncol. 2023;12:977251. doi:10.3389/fonc.2022.977251