3.1 Mutation Levels of ANGPTL2 in Different Cancers
DNA modifications are classified into two types: mutations(truncating mutation and missense)and copy number alterations(amplification and deepdeletion).As depicted in Figure 2A,among the topfive cancers in pan-cancer study,endometrial carcinoma and melanoma have mutation rates exceeding 3%,while colorectal cancer mainly exhibits high mutation rates(>2%).Prostate cancer primarily exhibits amplification as the major type of mutation,with a rate exceeding 1.5%.Esophageal and gastric cancers predominantly exhibit high mutation rates(>1%),accompanied by alterations such as amplification and deepdeletion.As shown in Figure 2B,we find that the main mutation mode of ANGPTL2 is missense,followed by truncating mutation.Figure 2C displays the A198V/T alteration in the three-dimensional structure of the ANGPTL2 protein.
3.2 Expression of ANGPTL2 in Pan-Cancer Tissues
We investigated the expression of ANGPTL2 in various tumors from the TCGA cohort using the Sangerbox database.As shown in Figure 2D,ANGPTL2 shows varying expression levels among 31 distinct tumor types.Notably,ANGPTL2 expression is significantly lower than that in adjacent normal tissues in 16 types of cancer,including endometrial carcinoma(UCEC),breast invasive carcinoma(BRCA),cervical squamous cell carcinoma(CESC),lung adenocarcinoma(LUAD),esophageal carcinoma(ESCA),kidney renal papillary cell carcinoma(KIRP),colon adenocarcinoma(COAD),colon adenocarcinoma/rectum adenocarcinoma(COADREAD),prostate adenocarcinoma(PRAD),lung squamous cell carcinoma(LUSC),skin cutaneous melanoma(SKCM),bladder urothelial carcinoma(BLCA),thyroid carcinoma(THCA),ovarian serous cystadenocarcinoma(OV),uterine carcinosarcoma(UCS)and pheochromocytoma and paragangliomas(KICH).On the contrary,ANGPTL2 has been found in GBM(Glioblastoma multiforme),GBMLGG(Glioma),LGG(Brain Lower Grade Glioma),KIPAN(Pan-kidney cohort)(KICH+KIRC+KIRP),STAD(Stomach adenocarcinoma),HNSC(Head and Neck squamous cell carcinoma),KIRC(Kidney)renal clear cell carcinoma,WT,PAAD(Pancreatic adenocarcinoma pancreatic cancer),TGCT(Testicular Germ Cell)Testicular germ cell Tumors),ALL(Acute lymphoblastic leukemia),LAML(Acute Myeloid Leukemia),PCPG(Pheochromocytoma and The expression of Paraganglioma was considerably reduced in 15 types of tumors,including Paraganglioma and paraganglioma,Adrenocortical carcinoma(ACC)and Cholangio carcinoma(CHOL).Given that certain data in normal tissues were unavailable,we obtained RNA sequencing expression data from TCGA,TARGET,and GTEx databases and analyzed ANGPTL2 expression in the 31 tumors mentioned above.As shown in Figure 2E,ANGPTL2 expression differed between 16 tumors.ANGPTL2 expression was considerably higher in eight types of cancers compared to surrounding normal tissues.Including STAD,Thymoma(Thymoma),KIRC,GBM,DLBC(Lymphoid Neoplasm Diffuse Large B-cell Lymphoma),PAAD,LGG,ACC expression up-regulated.However,we also observed downregulation of ANGPTL2 expression in UCS,UCEC,READ,OV,KICH,CESC,BRCA,BLCA and other tumors.
In addition,to assess CNN1 expression in various tumor stages,we also used GEPIA2.0 to explore the relationshipbetween ANGPTL2 expression level and tumor pathological stage.As shown in Figure 2F,ANGPTL2 expression was different in different tumor stages of 5 types of tumors,including STAD,KIRP,TGCT,BRCA,and BLCA.
3.3 Correlation between ANGPTL2 expression and survival prognosis of tumor patients
In the pan-cancer survival analysis of ANGPTL2,we first investigated the correlation between ANGPTL2 expression and overall survival(OS)across multiple cancers using the GEPIA database.As shown in Figure 3A,the expression of ANGPTL2 is correlated with the overall survival of five tumors,namely COAD,LGG,LIHC,LUSC,and PAAD,among which the high expression of ANGPTL2 in COAD,LIHC,LUSC,and PAAD is often significantly correlated with poor prognosis(P<0.05).As shown in Figure 3B,ANGPTL2 expression was correlated with disease-free survival of five tumors,namely COAD,DLBC,KIRC,LGG,and READ,among which high ANGPTL2 expression in COAD,KIRC,and READ was often significantly correlated with poor prognosis(P<0.05).
Next,we analyzed the expression of ANGPTL2 in relation to overall survival in ten different cancers using the Kaplan-Meier plotter online platform.As shown in Figure 3C,the expression of ANGPTL2 correlated with prognosis in UCEC,STAD,PAAD,OV,LUSC,LIHC,KIRP,HNSC,CESC and BLCA.Specifically,high ANGPTL2 expression was significantly associated with poor prognosis in STAD(P<0.001),UCEC(P<0.05),PAAD(P<0.05),OV(P<0.01),LUSC(P<0.05),LIHC(P<0.05),KIRP(P<0.01),CESC(P<0.05)and BLCA(P<0.05).Moving on to the correlation between ANGPTL2 protein expression levels and immune checkpoint genes,as well as immune cell infiltration,we extracted ANGPTL2 gene expression data from TCGA,TARGET,and GTEx databases.Using the SangerBox online platform,we evaluated the relationshipbetween ANGPTL2 expression and 60 genes in the immune checkpoint pathway.Remarkably,ANGPTL2 expression showed strong correlations with immune checkpoint genes in CHOL,PRAD,KICH,OV,PAAD,KIPAB,BLCA,READ,COAD,COADREAD,LGG,GBM and TGCT(Figure 4A).Specifically,ANGPTL2 expression positively correlated with immune checkpoint genes in CHOL,PRAD,KICH,OV,PAAD,KIPA,BLCA,READ,COAD and COADREAD.Pronounced effects were observed in OV、PAAD、KIPA、READ、COAD and COAREADE(Figure 4A).Conversely,the expression of ANGPTL2 is negatively correlated with immune checkpoint genes in LGG,GBMLGG,and TGCT(Figure 4A).These findings suggest that high ANGPTL2 expression is generally associated with a better prognosis.However,in the context of immunotherapy for LGG,GBMLGG and TGCT,ANGPTL2 inhibitors may lead to improved outcomes.
3.4 Correlation of ANGPTL2 protein expression level with immune checkpoint genes and immune infiltration
The results of the estimated immune score in Figure 4B showed that ANGPTL2 expression was positively correlated with the immune score of 23 cancers.Such as STAD (p=1.8e-10),BRCA (P=1.3e-10),LUAD (P=2.3e-8),STES (P=2.8e-17),ESCA (P=1.7e-5),SARC (P=3.0e-6),KIPR (P=9.2e-7),KIPA (P=6).0e-59),COAD (P=2.2e-29),COADREAD (P=7.9e-35),PRAD (P=1.8e-14),HNSC (P=3.6e-5),LIHC (P=2.6e-6),OV (P=2.5e-4),UVW (P=8.7e-4),LUSC (P=3.8e-1),BLCA (P=1.5e-18),READ (P=3.1e-7),PAAD (P=1.7e-6),PCPG (P=1.0e-5),SKCM (P=8.9e-3),THCA (P=3.0e-3),KICH (P=2.5e3),etc.,However,there were four significant negative correlations,such as GBMLGG (P=2.8e-26),LGG (P=1.9e-14),TGCT (P=6.1e-6) and ALL (P=1.6e-4).
3.5 Single-Cell Expression Patterns of ANGPTL2 in Various Cancers
To explore the impact of ANGPTL2 in different cancers,we analysed its correlation with 14 different functional categories in 10 cancer types by using the CancerSEA single-cell database.In retinoblastoma(RB),ANGPTL2 expression was negatively correlated with cell cycle,DNA damage and DNA repair.Conversely,ANGPTL2 expression was positively correlated with angiogenesis,differentiation,inflammation,tumour metastasis,quiescence and stemness.In uveal melanoma(UM),ANGPTL2 expression was negatively correlated with tumour biological behaviour such as cell apoptosis,DNA damage and invasion,while positively correlated with cell cycle,hypoxia and stemness,albeit weakly.Notably,ANGPTL2 showed a stronger association with high-grade glioma(HGG).Specifically,ANGPTL2 was positively correlated with cell cycle,DNA damage,DNA repair and stemness(P<0.05),while negatively correlated with angiogenesis,cell apoptosis,epithelial-mesenchymal transition,hypoxia,inflammation,tumour metastasis and cell proliferation(P<0.01)(Figure 5A,B).Interestingly,ANGPTL2 showed a positive correlation with stemness in most cancers(Figure 5A),but a negative correlation with DNA repair(Figure 5A).
3.6 Functional Enrichment Analysis of ANGPTL2
Next,we performed functional enrichment analysis to evaluate the potential molecular mechanisms of ANGPTL2 in tumour initiation and progression.As shown in Figure 5C,we identified 10 molecules that interact with ANGPTL2 from the BioGRID network tool.Subsequently,we conducted a KEGG-GO enrichment analysis(FIG.5D&E),which showed that ANGPTL2 co-expressed genes played a regulatory role in tyrosine phosphorylation,angiogenesis,vasculature development,and actin cytoskeleton regulation.In addition,ANGPTL2 is closely associated with 12 signalling pathways,including the PI3K-Akt,Jak-STAT and Rap1 pathways.
3.7 HE Staining and Immunohistochemistry of ANGPTL2 in Gastric Cancer Tissues
HE staining results showed that early stage gastric cancer cells had loosely arranged cell layers with minimal inflammatory cells.In contrast,advanced gastric cancer cells showed significant changes,with deepand dense nuclear concentration,serious disordered nuclear arrangement,and obvious heterogeneity .In the advanced stage IV group,the number of stained cells in the center of the tumor tissue was reduced,and the periphery of the tumor tissue was significantly atrophied(Figure 6A).Immunohistochemistry revealed that ANGPTL2 expression was significantly lower in early stage gastric cancer tissues compared to late stage cases(Figure 6B).These findings suggest that ANGPTL2 may serve as a sensitive marker in the tumour microenvironment of gastric cancer patients and warrant further investigation.
3.8 Impact of overexpression of ANGPTL2 on GES-1 cell invasion assay
To further investigate the effects of ANGPTL2 on malignant biological behaviours such as cell proliferation and migration,we performed ANGPTL2 overexpression plasmid transfection in GES-1 gastric mucosal epithelial cells.According to Wound healing experiments and Transwell test,compared with scratch area 0h after the initial test and 48h after the initial test,the cell invasion assay of ANGPTL2 overexpression groupwas significantly higher than that of NC group(P<0.01).In the Transwell experiment,we further confirmed the enhanced migration ability of cells in the ANGPTL2 overexpression group(Figure 6C,D).
The results showed that the number of cells attached to the lower surface of the chamber in the NC groupwas significantly lower than that in the ANGPTL2 overexpression group,indicating that the number of cells migrated in the ANGPTL2 overexpression groupwas significantly higher than that in the NC group(P<0.01),which also proved that the increase of ANGPTL2 expreThe results showed that the number of cells attached to the lower surface of the chamber in the NC groupwas significantly lower than that in the ANGPTL2 overexpression group,indicating that the number of cells migrated in the ANGPTL2 overexpression groupwas significantly higher than that in the NC group(P<0.01),which also proved that the increase of ANGPTL2 expression level could enhance the migration and invasion ability of cells to a certain extent.
3.9 Correlation between ANGPTL2 expression and clinical prognosis
This is a retrospective study in which all patients received continuous follow-up.Table 1 shows the baseline clinicopathological features of patients with gastric cancer.We performed immunohistochemistry on paraffin sections of 60 patients with gastric cancer to assess the expression of ANGPTL2 in cancer tissues.The positive expression sites were quantitatively analyzed using Image J software,and the expression of ANGPTL2 in GC patient tissues was further divided into low and high levels according to the median expression.Then,the relationshipbetween ANGPTL2 expression level and clinicopathological features was analyzed based on the clinical data of patients.In univariate analysis,there were statistical differences in TNM staging between the two groups(P= 0.035,Table 2,Figure 6C),and no significant differences in other clinical features.In order to further assess whether ANGPTL2 levels in patients with gastric cancer can be used as a prognostic predictors of patients and to explore whether there is a correlation between its expression and survival outcomes,we used Cox regression model for clinical pathological factors affecting the prognosis of the single factor analysis,the results showed that ANGPTL2 expression levels(P<0.001,HR=6.118,95%CI:[2.176-17.205]); TNM staging(P<0.001,HR=0.113,95%CI[0.035 -- 0370]); The degree of differentiation(P=0.018,HR=3.780,95%CI :[1.251-11.428])and age(P= 0.002,HR=2.198,95%CI [0.056-0.509])were associated with poorer prognosis(Table 3).
TABLE 1 Clinicopathological features of patients with gastric cancer.
|
|
Patient characteristics
|
Number of pataients (N=60)
|
Percentage(%)
|
|
Gender
|
|
|
|
Male
|
42
|
70
|
|
Female
|
18
|
30
|
|
Age(year)
|
|
|
|
<60
|
43
|
71.67
|
|
≥60
|
17
|
28.33
|
|
Histological type
|
|
|
|
Adenocarcinoma
|
47
|
78.33
|
|
Non-adenocarcinoma
|
13
|
21.67
|
|
Differentiation
|
|
|
|
high differentiation
|
30
|
50
|
|
moderately differentiated
|
25
|
41.67
|
|
poorly differentiated
|
5
|
8.33
|
|
TNM staging
|
|
|
|
Ⅰ
|
10
|
16.67
|
|
Ⅱ
|
27
|
45
|
|
Ⅲ
|
18
|
30
|
|
Ⅳ
|
5
|
8.33
|
|
Depth of intestinal wall infiltration
|
|
|
|
mucous membrane layer
|
3
|
5
|
|
muscle layer
|
11
|
18.33
|
|
serous layer
|
43
|
71.67
|
|
outer layer
|
3
|
5
|
|
Tumour size
|
|
|
|
<5 cm
|
29
|
48.33
|
|
≥5 cm
|
31
|
51.67
|
|
Lymph node metastasis
|
|
|
|
Without
|
39
|
65
|
|
With
|
21
|
35
|
|
CEA
|
|
|
|
<5
|
43
|
71.67
|
|
≥5
|
17
|
28.33
|
|
CA199
|
|
|
|
<37
|
48
|
80
|
|
≥37
|
12
|
20
|
|
Survival state
|
|
|
|
death
|
25
|
41.67
|
|
survival
|
35
|
58.33
|
|
CD11 expression
|
|
|
|
low expression
|
30
|
50
|
|
high expression
|
30
|
50
|
|
TABLE 2 Single-factor Cox regression analysis.
|
Clinicopathologic
|
Hazard
|
95.0% CI for Exp(B)
|
P value
|
characteristics
|
ratio
|
Lower
|
Upper
|
Gender
|
|
|
|
|
Male
|
1
|
0.592
|
4.663
|
0.334
|
Female
|
1.662
|
Age(year)
|
|
|
|
|
<60
|
1
|
0.056
|
0.509
|
0.002**
|
≥60
|
0.169
|
Histological type
|
|
|
|
|
Adenocarcinoma
|
1
|
0.813
|
7.765
|
0.11
|
mucoid adenocarcinoma
|
2.512
|
Depth of tumor invasion
|
|
|
|
|
T1&T2
|
1
|
0.429
|
7.792
|
0.414
|
T3&T4
|
1.829
|
degree of differentiation
|
|
high and moderately differentiation
|
1
|
1.251
|
11.428
|
0.018*
|
poorly differentiated
|
3.78
|
Tumour size
|
|
|
|
|
<5 cm
|
1
|
0.393
|
2.762
|
0.934
|
≥5 cm
|
1.042
|
TNM staging
|
|
|
|
|
Ⅰ&Ⅱ stage
|
1
|
0.035
|
0.37
|
<0.001***
|
Ⅲ&Ⅳ stage
|
0.113
|
Lymph node metastasis
|
|
|
|
|
Without
|
1
|
0.27
|
2.178
|
0.619
|
With
|
0.768
|
CEA
|
|
|
|
|
<5 ng/mL
|
1
|
0.679
|
4.037
|
0.267
|
≥5 ng/mL
|
1.656
|
CA199
|
|
|
|
|
<37 U/mL
|
1
|
0.48
|
0.178
|
1.625
|
≥37 U/mL
|
0.538
|
ANGPTL2
|
|
|
|
|
Low expression
|
1
|
2.176
|
17.205
|
<0.001***
|
High expression
|
6.118
|
TABLE 3 Correlation between ANGPTL2 and clinicopathologic characteristics.
|
|
Clinicopathological paramete
|
N
|
ANGPTL2 Expression
|
P 值
|
|
Low expression
|
High expression
|
|
Gender
|
|
|
|
|
|
Male
|
42
|
21
|
21
|
1
|
|
Female
|
18
|
9
|
9
|
|
Age(year)
|
|
|
|
|
|
<60
|
43
|
19
|
24
|
0.152
|
|
≥60
|
17
|
11
|
6
|
|
Histological type
|
|
|
|
|
|
Adenocarcinoma
|
47
|
23
|
24
|
0.754
|
|
mucoid adenocarcinoma
|
13
|
7
|
6
|
|
Depth of tumor invasion
|
|
|
|
|
|
T1&T2
|
6
|
4
|
2
|
0.389
|
|
T3&T4
|
54
|
26
|
28
|
|
degree of differentiation
|
|
|
|
|
|
high and moderately differentiation
|
30
|
16
|
14
|
0.606
|
|
poorly differentiated
|
30
|
14
|
16
|
|
Tumour size
|
|
|
|
|
|
<5 cm
|
29
|
14
|
15
|
0.796
|
|
≥5 cm
|
31
|
16
|
15
|
|
TNM staging
|
|
|
|
|
|
Ⅰ&Ⅱ stage
|
24
|
16
|
8
|
0.035*
|
|
Ⅲ&Ⅳ stage
|
36
|
14
|
22
|
|
Lymph node metastasis
|
|
|
|
|
|
Without
|
39
|
19
|
20
|
0.787
|
|
With
|
21
|
11
|
10
|
|
CEA
|
|
|
|
|
|
<5 ng/mL
|
43
|
21
|
22
|
0.774
|
|
≥5 ng/mL
|
17
|
9
|
8
|
|
CA199
|
|
|
|
|
|
<37 U/mL
|
48
|
25
|
23
|
0.519
|
|
≥37 U/mL
|
12
|
5
|
7
|
|