Check for updates
Hindawi
Retraction
Retracted: Pancancer Analysis of Neurovascular-Related NRP Family Genes as Potential Prognostic Biomarkers of Bladder Urothelial Carcinoma
BioMed Research International
Received 12 March 2024; Accepted 12 March 2024; Published 20 March 2024
Copyright @ 2024 BioMed Research International. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
This article has been retracted by Hindawi following an investigation undertaken by the publisher [1]. This investi- gation has uncovered evidence of one or more of the follow- ing indicators of systematic manipulation of the publication process:
(1) Discrepancies in scope
(2) Discrepancies in the description of the research reported
(3) Discrepancies between the availability of data and the research described
(4) Inappropriate citations
(5) Incoherent, meaningless and/or irrelevant content included in the article
(6) Manipulated or compromised peer review
The presence of these indicators undermines our confi- dence in the integrity of the article’s content and we cannot, therefore, vouch for its reliability. Please note that this notice is intended solely to alert readers that the content of this arti- cle is unreliable. We have not investigated whether authors were aware of or involved in the systematic manipulation of the publication process.
Wiley and Hindawi regrets that the usual quality checks did not identify these issues before publication and have since put additional measures in place to safeguard research integrity.
We wish to credit our own Research Integrity and Research Publishing teams and anonymous and named external researchers and research integrity experts for con- tributing to this investigation.
The corresponding author, as the representative of all authors, has been given the opportunity to register their agreement or disagreement to this retraction. We have kept a record of any response received.
References
[1] C. Deng, H. Guo, D. Yan, T. Liang, X. Ye, and Z. Li, “Pancancer Analysis of Neurovascular-Related NRP Family Genes as Poten- tial Prognostic Biomarkers of Bladder Urothelial Carcinoma,” BioMed Research International, vol. 2021, Article ID 5546612, 31 pages, 2021.
Hindawi
Research Article
Pancancer Analysis of Neurovascular-Related NRP Family Genes as Potential Prognostic Biomarkers of Bladder Urothelial Carcinoma
Chao Deng, Hang Guo, Dongliang Yan D, Tao Liang, Xuxiao Ye, and Zuowei Li
Department of Urology, Shanghai Jiao Tong University Affiliated Sixth People’s Hospital, Shanghai, China
Correspondence should be addressed to Dongliang Yan; dly1919@126.com
Chao Deng and Hang Guo contributed equally to this work.
Received 2 February 2021; Revised 8 March 2021; Accepted 20 March 2021; Published 15 April 2021
Academic Editor: Qian Wang
Copyright @ 2021 Chao Deng et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background. Neurovascular-related genes have been implicated in the development of cancer. Studies have shown that a high expression of neuropilins (NRPs) promotes tumourigenesis and tumour malignancy. Method. A multidimensional bioinformatics analysis was performed to examine the relationship between NRP genes and prognostic and pathological features, tumour mutational burden (TMB), microsatellite instability (MSI), and immunological features based on public databases and find the potential prognostic value of NRPs in pancancer. Results. Survival analysis revealed that a low NRP1 expression in adrenocortical carcinoma (ACC), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), low-grade glioma (LGG), and stomach adenocarcinoma (STAD) was associated with poor prognosis. A high NRP2 expression in bladder urothelial carcinoma (BLCA), kidney renal papillary cell carcinoma (KIRP), and mesothelioma (MESO) was associated with poor prognosis. Moreover, NRP1 and NRP2 were associated with TMB and MSI. Subsequent analyses showed that NRP1 and NRP2 were correlated with immune infiltration and immune checkpoints. Genome-wide association analysis revealed that the NRP1 expression was strongly associated with kidney renal clear cell carcinoma (KIRC), whereas the NRP2 expression was closely associated with BLCA. Ultimately, NRP2 was found to be involved in the development of BLCA. Conclusions. Neurovascular-related NRP family genes are significantly correlated with cancer prognosis, TME, and immune infiltration, particularly in BLCA.
1. Introduction
The growth and development of neovascular tissue or angio- genesis are critical for normal physiological processes. There- fore, dysregulation of the angiogenic process has been linked to tumour development and progression [1]. The vascular endothelial growth factor (VEGF) is a key factor involved in angiogenesis. VEGF messenger RNA (mRNA) is widely over- expressed in tissues and is associated with metastasis, recur- rence, and prognosis [2]. In recent years, several drugs that inhibit the VEGF signaling pathway have been designed to treat cancer, including anti-VEGF monoclonal antibodies [3-6]. And neurovascular-related genes have been implicated
in cancer development. There is a strong link between neural stem/progenitor cells (NSPCs) and endothelial cells (ECs) [7].
Evidence suggests that neuropilins (NRPs), the VEGF receptors, are involved in tumourigenesis [8, 9]. NRPs partic- ipate in the development of the nervous system by function- ing as receptors for axon guidance factors [10]. Several signaling pathways regulate neuronal development by target- ing NRPs [11]. High expression of NRPs is closely associated with tumourigenesis and malignancy [12].
NRP1 and NRP2 are two isoforms of NRPs in mammals; studies have demonstrated their cancer-promoting potential [13]. For example, NRP2 is highly expressed in triple-negative breast cancers [14]. In prostate cancer, NRP2 expression is
Pan-cancer analysis of neurovascular disease-related NRP family genes (NRP1 and NRP2)
RNA-sequence data
Differential expression of NRP1 and NRP2 in pan-cancer (Figure 2)
Survival and clinicopathological data
Survival analysis of pan-cancer (Figure 3)
KIRC
D)
Clinicopathology analysis (Figure 8)
NRP1
Somatic mutation data
Association between NRP family genes and TMB, MSI in pan-cancer (Figure 4)
LGG
Correlation analysis (Figure 9)
TCGA
Relationship between NRP1 and NRP2 expression and immunity in pan-cancer
Immune checkpoint (Figure 5)
BLCA
NRP2
Immune cell infiltration (Figure 6)
Relationship between NRP family genes and tumour microenvironment (Figure 7)
positively correlated with the Gleason grade [15]. In the bladder cancer, high expression of NRP2 is associated with chemoresis- tance and epithelial-to-mesenchymal transition and poor patient prognosis [16]. However, the expression and function of NRPs in different cancers are not fully known.
Herein, we comprehensively analysed the correlation of NRP expression with prognosis and tumour microenviron- ment landscape in 33 cancer types. Our findings reveal that NRPs may be a potential prognostic marker associated with immune infiltration, tumour mutations, and tumour microenvironment, particularly in bladder urothelial carci- noma (BLCA).
2. Materials and Methods
2.1. Analysis of Differential NRP1 and NRP2 Gene Expression in Human Cancer. RNA sequences, somatic mutations, and clinicopathological features of 33 cancers were downloaded from The Cancer Genome Atlas (TCGA) database. The data included 10,953 patients (10,967 samples). A pancancer anal- ysis was performed on NRP1 and NRP2 mRNA expression levels in the Oncomine database (http://www.ONCOMINE .org). The threshold was set at p value < 0.05 and |fold change |>1.5. In addition, changes in NRP1 and NRP2 expression in different cancer types were determined using the R package “ggpubr” and the cBioPortal database (https://www.cbioportal .org). All data analyses were performed using version 4.0.3 of the R language package (https://www.r-project.org/).
2.2. Survival Analysis. The association of NRP1 and NRP2 with survival was assessed with the Kaplan-Meier method and log-rank test (p < 0.05). Patients were divided into high- and low-risk groups based on median expression levels of NRP1 and NRP2. Survival curves were created using “surv- miner” and “survivor” packages of R. Cox analysis was per- formed to explore the association of NRP1 and NRP2 with the prognosis of different cancers. A “forestplot” function was used to draw a forest plot whereas the “ggplot2” function was used to analyse clinicopathological features.
2.3. Association of NRP Family Genes with Tumour Mutational Burden (TMB) and Microsatellite Instability (MSI) in Various Cancers. TMB was derived from a study published by Gentles et al. [17], and MSI was obtained from a study published by Bonneville et al. [18]. As in previous studies [19-21], statistical analyses were performed using the rank-sum test, and p values less than 0.05 were consid- ered statistically significant; R software was used for plotting.
2.4. Association of NRP1 and NRP2 Expression with Immune Checkpoint-Related Genes in Different Cancers. As described in previous studies [22-27], the xCell method was used to perform immune score assessment. The immune checkpoint genes, pDCD1, SIGLEC15, HAVCR2, IDO1, CD274, LAG3, CTLA4, and PDCD1LG2, were analysed to examine the asso- ciation of NRP1 and NRP2 with expression of immune checkpoint-related genes.
2.5. DNAss, RNAss, StromalScore, and ImmuneScore among Subgroups. The differentiated phenotype was rapidly lost during cancer progression, and progenitor and stem-cell- like characteristics were acquired [28]. RNAss based on mRNA expression and DNAss based on DNA methylation were utilized to measure the tumour stemness [29]. The ESTI- MATE algorithm in the R language ESTIMATE package was used to estimate the ratio of immune to stromal components in the TME for each sample and is presented as two scores: ImmuneScore and StromalScore, which are positively corre- lated with immune and stromal components, respectively.
2.6. Integrative Data Visualization. The correlation of NRP1 and NRP2 with other genes was mapped using Cancer Regu- lome Tools (http://explorer.cancerregulome.org/). A p value > -log100 was considered statistically significant.
3. Results
3.1. NRP1 and NRP2 mRNA Levels in Pancancers. The flow chart of this study is shown in Figure 1. NRP1 and NRP2 were found to be widely expressed in human tissues (Figure 2(a)). The overall expression level of NRP1 did not
Interactive bodymap
6
The median expression of tumour and normal samples in bodymap
5
Gene expression
NRP1
NRP2
4
3
2
1
0
NRP1
NRP2
Log2 (TPM + 1) scale
Log2 (TPM + 1) scale
(a)
(b)
NRP1
Analysis type by cancer
Bladder cancer
Brain and CNS cancer
Breast cancer
Cervical cancer
Colorectal cancer
Esophageal cancer
Gastric cancer
Head and neck cancer
Kidney cancer
Leukemia
RE
Liver cancer
Lung cancer
Lymphoma
Melanoma
Myeloma
Other cancer
Ovarian cancer
Pancreatic cancer
Prostate cancer
Sarcoma
Significant unique analyses
Total unique analyses
1
5
10
10
5
1
%
(c)
| Cancer VS. normal | Cancer vs. cancer | ||||
|---|---|---|---|---|---|
| Cancer histology | Multi-cancer | ||||
| 1 | 3 | 3 | 1 | ||
| 5 | 4 | 6 | 6 | ||
| 2 | 8 | 3 | 1 | 2 | |
| 1 | 1 | 1 | |||
| 1 | 3 | 2 | 2 | 2 | 3 |
| 2 | 2 | 2 | 2 | ||
| 4 | 1 | 3 | 3 | 2 | |
| 1 | 1 | ||||
| 3 | 4 | 3 | 6 | 9 | |
| 2 | 1 | 9 | 5 | 9 | |
| 2 | 2 | 3 | 1 | ||
| 4 | 6 | 6 | 3 | ||
| 6 | 5 | 3 | 8 | ||
| 1 | 3 | ||||
| 2 | 1 | 1 | 3 | ||
| 4 | 1 | 3 | 2 | 3 | |
| 3 | 4 | 3 | 2 | 1 | |
| 1 | 4 | ||||
| 1 | 1 | ||||
| 6 | 2 | 12 | 15 | 2 | |
| 37 | 32 | 61 | 61 | 34 | 29 |
| 417 | 698 | 263 | |||
FIGURE 2: Continued.
ED
THCA
UCEC
RE
6
$
NRP1 expression
4
2
0
BLCA
BRCA
CHOL
COAD
ESCA
GBM
HNSC
KICH
KIRC
KIRP
LIHC
LUAD
LUSC
PRAD
READ
STAD
Cancer type
Type
Normal
Tumour
(d)
NRP2
Analysis type by cancer
Bladder cancer
Brain and CNS cancer
Breast cancer
Cervical cancer
Colorectal cancer
Esophageal cancer
Gastric cancer
Head and neck cancer
Kidney cancer
Leukemia
Liver cancer
Lung cancer
Lymphoma
Melanoma
Myeloma
Other cancer
Ovarian cancer
Pancreatic cancer
Prostate cancer
Sarcoma
Significant unique analyses
Total unique analyses
1
5
10
10
5
1
%
(e)
| Cancer | Cancer vs. cancer | ||||
|---|---|---|---|---|---|
| VS. | |||||
| normal | Cancer histology | Multi-cancer | |||
| 6 | 3 | 3 | 1 | 1 | |
| 7 | 2 | 4 | 5 | 6 | 1 |
| 5 | 3 | 4 | 2 | 4 | |
| 1 | 1 | ||||
| 6 | 4 | 3 | 5 | ||
| 1 | 1 | 1 | 1 | ||
| 4 | 3 | 5 | 3 | 1 | |
| 8 | 2 | ||||
| 3 | 1 | 5 | 8 | 3 | |
| 1 | 1 | 8 | 4 | 6 | |
| 1 | |||||
| 1 | 2 | 1 | 4 | ||
| 5 | 1 | 5 | 3 | 2 | 3 |
| 1 | 1 | 1 | 12 | ||
| 1 | 1 | 1 | 1 | ||
| 5 | 1 | 1 | 1 | ||
| 1 | 6 | 11 | 2 | 3 | |
| 2 | 2 | 1 | |||
| 6 | 1 | 1 | 2 | ||
| 4 | 1 | 12 | 18 | 1 | |
| 53 | 29 | 59 | 65 | 33 | 30 |
| 424 | 718 | 263 | |||
FIGURE 2: Continued.
6
NRP2 expression
ED
4
2
m
0
BLCA
BRCA
CHOL
COAD
ESCA
GBM
HNSC
KICH
KIRC
KIRP
LIHC
LUAD
LUSC
PRAD
READ
STAD
THCA
UCEC
Cancer type
Type
Normal
Tumour
(f)
significantly differ from that of NRP2 in human tissues (Figure 2(b)), suggesting good concordance between NRP1 and NRP2 expression in humans. Results of NRP1 and NRP2 mRNA levels in the Oncomine database are shown in Figures 2(c) and 2(d). We further assessed the expression of NRP1 and NRP2 in different cancers by analysing 730 nor- mal samples and 10,327 fractional tumour samples in TCGA data sets (Figures 2(e) and 2(f)). Overall, whether NRP1 and NRP2 are highly or lowly expressed in tumour tissue was difficult to establish. The expression of NRP1 and NRP2 was different between normal tissue and tumour tissues in the brain and central nervous system cancers. Of note, the expression of NRP1 and NRP2 genes in some cancers was inconsistent in different databases. These inconsistencies may be caused by different gene extraction methods and bio- logical mechanisms. These results demonstrate that NRP1 and NRP2 are differentially expressed in different tissues, suggesting they may have distinct roles in different tissues.
3.2. Prognostic Value of NRP1 and NRP2 in Various Cancers. Next, we explored the prognostic value of NRP1 and NRP2 in various cancers in the TCGA database. We found that NRP1 and NRP2 expression was associated with the prognosis of various cancers. NRP1 was found to be a risk factor in differ- ent cancers, including ACC (HR 1.027, 95% CI 1.014-1.040, p <0.001), CESC (HR 1.021, 95% CI 1.007-1.035, p<0.003), GBM (HR 1.014, 95% CI 1.004-1.025, p = 0.009), LGG (HR 1.038, 95% CI 1.024-1.053, p < 0.0001), LIHC (HR 1.009, 95% CI 1.003-1.016, p = 0.0053), MESO (HR 1.011, 95% CI 1.003-1.020, p=0.0062), and STAD (HR 1.018, 95% CI
1.010-1.026, p<0.0001) (Figure 3(a)). In contrast, NRP1 was a protective factor in KIRC (HR 0.995, 95% CI 0.992- 0.997, p < 0.0001). Further analysis showed that NRP2 was a risk factor in different cancers such as BLCA (HR 1.012, 95% CI 1.003-1.021, p = 0.0093), KICH (HR 1.178, 95% CI 1.008-1.375, p = 0.0390), KIRP (HR 1.048, 95% CI 1.015- 1.081, p = 0.0040), LAML (HR 1.127, 95% CI 1.031-1.232, p = 0.0086), LGG (HR 1.012, 95% CI 1.002-1.021, p = 0.0168), LIHC (HR 1.015, 95% CI 1.001-1.029, p = 0.0400), MESO (HR 1.012, 95% CI 1.006-1.019, p = 0.0003), PAAD (HR 1.017, 95% CI 1.006-1.029, p= 0.0027), and STAD (HR 1.009, 95% CI 1.001-1.018, p = 0.0282) (Figure 3(b)). Survival analysis suggested that low NRP1 expression in adrenocorti- cal carcinoma (ACC), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), low-grade glioma (LGG), and stomach adenocarcinoma (STAD) was associated with poor patient prognosis. However, high NRP1 expression in kidney renal clear cell carcinoma (KIRC) predicted good prognosis (Figures 3(c)-3(g)). High NRP2 expression in BLCA, kidney renal papillary cell carcinoma (KIRP), and mesothelioma (MESO) was associated with poor prognosis (Figures 3(h)-3(j)).
3.3. Association of NRP1 and NRP2 Expression with TMB and MSI in Different Cancers. A high TMB influences immuno- therapy sensitivity [28, 29]. Thus, we assessed the relation- ship between NPR2 expression levels and BLCA, kidney chromophobe (KICH), KIRP, acute myeloid leukemia (LAML), LGG, liver hepatocellular carcinoma (LIHC), MESO, pancreatic adenocarcinoma (PAAD), and STAD.
-
0.925 1 1.05 1.1 1.15 1.2 1.25 1.3
Hazard ratio
(a)
FIGURE 3: Continued.
| Cancer | p value | NRP1 Hazard ratio (95% CI) | |
|---|---|---|---|
| ACC | < 0.0001 | 1.027 (1.014, 1.04) | |
| BLCA | 0.1289 | 1.005 (0.999, 1.011) | |
| BRCA | 0.0937 | 1.007 (0.999, 1.015) | |
| CESC | 0.003 | 1.021 (1.007, 1.035) | |
| CHOL | 0.5013 | 1.006 (0.989, 1.023) | |
| COAD | 0.1877 | 1.009 (0.996, 1.023) | |
| DLBC | 0.5041 | 1.041 (0.925, 1.173) | |
| ESCA | 0.489 | 1.005 (0.991, 1.02) | |
| GBM | 0.009 | 1.014 (1.004, 1.025) | |
| HNSC | 0.2342 | 1.006 (0.996, 1.015) | |
| KICH | 0.5708 | 1.015 (0.964, 1.069) | |
| KIRC | < 0.0001 | 0.995 (0.992, 0.997) | |
| KIRP | 0.6805 | 1.002 (0.993, 1.011) | TRACTED |
| LAML | 0.4065 | 0.984 (0.947, 1.022) | |
| LGG | < 0.0001 | 1.038 (1.024, 1.053) | |
| LIHC | 0.0053 | 1.009 (1.003, 1.016) | |
| LUAD | 0.919 | 1 (0.995, 1.004) | |
| LUSC | 0.6307 | 1.002 (0.994, 1.011) | |
| MESO | 0.0062 | 1.011 (1.003, 1.02) | |
| OV | 0.1705 | 1.01 (0.996, 1.024) | |
| PAAD | 0.0604 | 1.01 (1, 1.02) | |
| PCPG | 0.2534 | 0.973 (0.928, 1.02) | |
| PRAD | 0.4625 | 0.986 (0.948, 1.024) | |
| READ | 0.3013 | 1.02 (0.983, 1.058) | |
| SARC | 0.2035 | 1.002 (0.999, 1.006) | |
| SKCM | 0.3529 | 0.998 (0.992, 1.003) | |
| STAD | < 0.0001 | 1.018 (1.01, 1.026) | |
| TGCT | 0.9949 | 1 (0.932, 1.072) | |
| THCA | 0.3754 | 1.011 (0.987, 1.036) | |
| THYM | 0.6983 | 1.003 (0.99, 1.016) | |
| UCEC | 0.1204 | 0.988 (0.972, 1.003) | |
| UCS | 0.4564 | 1.005 (0.992, 1.017) | |
| UVM | 0.1001 | 1.122 (0.978, 1.288) |
Cancer
p value
NRP2
Hazard ratio (95% CI)
ACC
0.4764
0.989 (0.958, 1.02)
1
BLCA
0.0093
1.012 (1.003, 1.021)
BRCA
0.5143
1.004 (0.992, 1.016)
CESC
0.3377
1.008 (0.992, 1.025)
CHOL
0.1554
1.038 (0.986, 1.094)
4
COAD
0.2383
1.012 (0.992, 1.032)
DLBC
0.4527
0.945 (0.815, 1.096)
4
ESCA
0.0974
0.982 (0.961, 1.003)
GBM
0.8297
1.001 (0.993, 1.009)
HNSC
0.2699
1.003 (0.997, 1.009)
KICH
0.0390
1.178 (1.008, 1.375)
KIRC
0.8833
0.999 (0.987, 1.011)
KIRP
0.0040
1.048 (1.015, 1.081)
H
LAML
0.0086
1.127 (1.031, 1.232)
1
4
LGG
0.0168
1.012 (1.002, 1.021)
LIHC
0.0400
1.015 (1.001, 1.029)
LUAD
0.3321
1.004 (0.996, 1.011)
LUSC
0.4776
0.998 (0.991, 1.004)
MESO
0.0003
1.012 (1.006, 1.019)
OV
0.4996
1.002 (0.996, 1.009)
PAAD
0.0027
1.017 (1.006, 1.029)
PCPG
0.5363
0.983 (0.93, 1.039)
+
4
PRAD
0.8259
0.979 (0.811, 1.182)
1
4
READ
0.3286
1.024 (0.976, 1.075)
F
4
SARC
0.8890
1 (0.995, 1.004)
SKCM
0.4413
0.999 (0.997, 1.001)
STAD
0.0282
1.009 (1.001, 1.018)
TGCT
0.6160
0.99 (0.952, 1.03)
1
THCA
0.3399
0.992 (0.976, 1.008)
THYM
0.7439
0.998 (0.985, 1.011)
UCEC
0.4796
1.004 (0.992, 1.016)
UCS
0.1919
0.988 (0.971, 1.006)
UVM
0.1425
0.985 (0.966, 1.005)
0.811
0.9
0.975 1.075 1.175 1.275 1.375
Hazard ratio
(b)
Cancer: ACC
NRP1
Cancer: CESC
NRP1
1.00
1.00
Overall survival
0.75
Overall survival
0.75
0.50
0.50
0.25
p = 0.008
0.25
p = 0.019
0.00
0.00
0
2
4
6
8
10
12
0
2
4
6
8
10
12
14
16
18
20
Time (years)
Time (years)
NRP1 levels
High Low
39 40
NRP1 levels
28
11
4
12
1
0
0
30
19
7
4
2
High
148 147
68
21
13
10
5
3
0
0
Low
0
0
74
41
22
13
11
8
4
2
0
0
0
2
4
6
8
10
12
0
2
4
6
8
10
12
14
16
18
20
Time (years)
Time (years)
NRP1 levels
NRP1 levels
+
High
+
High
+
Low
+
Low
(c)
(d)
Cancer: KIRC
NRP1
Cancer: LGG
NRP1
1.00
1.00
Overall survival
0.75
Overall survival
0.75
0.50
0.50
0.25
p = 0.004
0.25
p = 0.046
0.00
0.00
0
2
4
6
8
10
12
0
2
4
6
8
10
12
14
16
18
20
Time (years)
Time (years)
NRP1 levels
High Low
265 266
NRP1 levels
181
20
0
High Low
32
16
179
114
104
48
51
21
6
7
262
121
52
14
6
4
1
262
133
43
24
11
1
0
0
5
3
1
0
0
0
0
2
4
6
8
10
12
0
2
4
6
8
10
12
14
16
18
20
Time (years)
High Low (e) RE
Time (years)
NRP1 levels
NRP1 levels
+
High
+
Low
(f)
FIGURE 3: Continued.
Cancer: STAD
NRP1
Cancer: BLCA
NRP2
1.00
4
1.00
Overall survival
0.75
Overall survival
0.75
0.50
0.50
0.25
0.25
p = 0.006
p = 0.003
0.00
0.00
0
2
4
6
8
12
0
2
4
6
8
10
12
14
Time (years)
Time (years)
NRP1 levels
High Low
NRP2 levels
175
9
2
175
47
1
53
14
0
5
2
1
High
203
203
71
34
18
33
10
6
Low
0
68
9
3
0
3
0
0
0
2
4
6
8
10
0
2
4
6
8
10
12
14
Time (years)
Time (years)
NRP1 levels
NRP2 levels
+
High
+
High
+
Low
+
Low
(g)
(h)
Cancer: KIRP
NRP2
Cancer: MESO
NRP2
1.00
1.00
J
Overall survival
0.75
Overall survival
0.75
0.50
0.50
0.25
p = 0.009
0.25
p = 0.048
0.00
0.00
0
2
4
6
8
10
12
14
16
0
2
4
6
8
Time (years)
Time (years)
NRP2 levels
High
Low
143
143
76
74
35
41
16 20
NRP2 levels
6
2
2
0
0
0
1
1
High Low
42
42
11
19
1
0
0
7
1
7
3
0-00
r
Y
0
2
4
6
8
10
12
14
16
0
2
4
6
8
Time (years)
Time (years)
NRP2 levels
NRP2 levels
+
High
+
High
+
Low
+
Low
(i)
(j)
This is because the expression of NRP1 and NRP2 correlated with the overall survival of such cancers (according to the results of one-way Cox and Kaplan-Meier survival analyses). The results showed that NRP1 expression was positively cor- related with TMB in ACC and LGG but negatively correlated with the TMB of MESO, LIHC, and STAD expression (Figure 4(a)). NRP2 expression was positively correlated with the TMB of LAML and PAAD but negatively correlated with the TMB of MESO, KIRP, STAD, and LIHC (Figure 4(c)).
In further analyses, it was found that NRP1 expression was significantly positively correlated with MSI in MESO but negatively correlated with MSI in STAD (Figure 4(b)). NRP2 expression was also significantly positively correlated with MSI in KIRC but negatively correlated with MSI in STAD (Figure 4(d)).
3.4. Coexpression of Immune Checkpoint Genes with NRP1 and NRP2 in Different Cancers. A coexpression analysis was performed to explore the correlation of NRP1 and NRP2 expression with immune checkpoint genes. In most cancers, NRP1 and NRP2 expression was found to be positively corre- lated with immune checkpoint genes (CD274, CTLA4, HAVCR2, LAG3, PDCD1, PDCDILG2, SIGLEC15, and TIGIT) (Figures 5(a) and 5(b)). In BLCA, the NRP1 and NRP2 expres- sion was negatively correlated with the SIGLEC15 expression. In MESO, the SIGLEC15 expression was negatively correlated with the NRP2 expression. In KIRC, LAG3 and PDCD1 expres- sion levels were positively correlated with the NRP1 expression.
3.5. Association of NRP1 and NRP2 Expression with Immune Infiltration. Previously, we showed that low NRP1 expression
NRP1-TMB
-log10 (p value)
THYM
6
ACC
LGG
SARC
COAD
LAML
PRAD
UCEC
LUAD
OV
SKCM
CESC
4
KICH
KIRC
UCS
KIRP
DLBC
GBM
LUSC
BLCA
READ
CHOL
2
HNSC
MESO
TGCT
BRCA
PAAD
ESCA
PCPG
LIHC
STAD
THCA
UVM
-0.25
0.00
0.25
Correlation (TMB)
Correlation
0.1
0.3
0.2
0.4
(a)
FIGURE 4: Continued.
NRP1-MSI
-log10 (p value)
READ
ED
COAD
6
MESO
UVM
GBM
SARC
PAAD
CESC
SKCM
OV
LAML
LIHC
UCEC
4
ACC
KIRC
THYM
KIRP
ESCA
PCPG
TGCT
BLCA
BRCA
PRAD
2
LGG
LUAD
KICH
LUSC
THCA
CHOL
HNSC
UCS
STAD
DLBC
-0.4
-0.2
0.0
0.2
Correlation (MSI)
Correlation
0.1
0.2
0.3
(b)
NRP2-TMB
-log10 (p value)
THYM
LAML
DLBC
PAAD
SKCM
UCS
LGG
KICH
6
TGCT
BRCA
GBM
UVM
READ
UCEC
BLCA
HNSC
4
KIRC
PCPG
COAD
LUAD
OV
CHOL
SARC
LUSC
CESC
2
THCA
PRAD
MESO
ACC
KIRP
ESCA
STAD
LIHC
-0.2
0.0
0.2
Correlation (TMB)
Correlation
0.1
0.2
0.3
(c)
NRP2-MSI
-log10 (p value)
READ
COAD
KIRC
LUSC
OV
MESO
TGCT
SARC
PRAD
3
SKCM
PAAD
ACC
UVM
BRCA
BLCA
UCEC
LGG
2
KICH
THCA
THYM
GBM
LUAD
KIRP
LAML
LIHC
1
PCPG
ESCA
CESC
HNSC
STAD
CHOL
UCS
DLBC
-0.4
-0.2
0.0
Correlation (MSI)
Correlation
0.1
0.2
0.3
(d)
in ACC, CESC, LGG, and STAD was associated with poor prognosis, whereas high NRP1 expression in KIRC predicted good prognosis. Moreover, high NRP2 expression in BLCA, KIRP, and MESO was associated with poor prognosis.
Hence, the xCell approach was used to comprehensively assess the association of NRP family genes with immune infil- tration (Figures 6(a) and 6(b)). We found that the NRP1 and NRP2 expression correlated significantly negatively with the
2738, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/5546612 by National Library Of Medicine, Wiley Online Library on [05/04/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Correlation
Correlation
UVM
**
**
**
**
**
*
**
UVM
*
UCS
+
+
+
UCS
+
+
UCEC
**
UCEC **
*
*
**
**
THYM
..
..
THYM
**
**
**
..
THCA
0.6
.
THCA
0.50
TGCT
*
TGCT
.
..
STAD
**
**
**
**
**
STAD **
**
**
**
**
**
SKCM
++
SKCM
.
SARC
+
..
SARC
..
.
..
.
..
READ
**
..
..
..
READ **
**
..
**
**
..
PRAD
0.4
**
**
..
.
**
**
PRAD
0.25
PCPG
..
.
*
.
PCPG
**
**
..
.
**
**
**
PAAD
**
**
..
..
**
**
.
**
PAAD
OV
**
**
**
**
*
**
OV
..
..
.
..
..
MESO
MESO
.
.
.
.
LUSC **
..
**
**
0.2
LUSC
..
**
**
NRP1
0.00
LUAD
NRP2
**
..
**
LUAD
LIHC
LIHC
**
..
**
**
LGG
..
LGG
..
..
..
.+
..
..
LAML
.
..
..
LAML
**
**
.
..
KIRP
.
..
..
KIRP
..
**
..
..
..
..
..
KIRC
..
..
..
0.0
KIRC
.
-0.25
KICH
**
**
**
**
**
**
KICH
++
+
HNSC
..
HNSC
..
..
..
GBM
..
..
.
..
GBM
*
ESCA
**
**
..
..
**
**
ESCA
..
..
..
DLBC
DLBC
**
..
**
**
COAD
..
..
..
..
**
-0.2
COAD
-0.50
CHOL
**
..
CHOL
..
.
..
CESC
**
**
**
**
CESC
.
BRCA
++
++
++
+
++
++
++
++
BRCA
++
++
++
+
++
++
++
++
BLCA
**
**
..
..
**
..
..
BLCA
++
ACC
.
ACC
**
.
*
CD274
CTLA4
HAVCR2
LAG3
PDCD1
PDCD1LG2
SIGLEC15
TIGIT
CD274
CTLA4
HAVCR2
LAG3
PDCD1
PDCD1LG2
SIGLEC15
TIGIT
*
P < 0.05
*
P < 0.05
**
P < 0.01
** P < 0.01
(a)
(b)
R
T cell CD4+ Th1 expression in almost all of the cancer types. Infiltration of mast cells was positively correlated with the NRP1 expression in most of the cancer types. The high NRP1 expression in ACC, CESC, GBM, LGG, MESO, and STAD was associated with poor prognosis, suggesting that mast cell infiltration may be associated with NRP1 expression.
In addition, high NRP1 expression was associated with higher stroma, microenvironment, and immune scores, as well as more endothelial cell infiltration in most tumours. A high NRP2 expression in BLCA and KIRP was associated with poor patient prognosis, while a high NRP2 expression in BLCA and KIRP implied depletion of T cell CD4+ central memory.
NRP 1
Correlation
stroma score
microenvironment score
immune score
T cell regulatory (Tregs)
T cell gamma delta
0.4
T cell NK
T cell CD8+ naive
T cell CD8+ effector memory
T cell CD8+ central memory
T cell CD8+
T cell CD4+ naive
T cell CD4+ memory
T cell CD4+ effector memory
T cell CD4+ central memory
T cell CD4+ Th2
T cell CD4+Thì
T cell CD4+ (non-regulatory)
xCELL
Plasmacytold dendritic cell
Neutrophil
NK cell
0.0
Myeloid dendritic cell activated
Myeloid dendritic cell
Monocyte
Mast cell
Macrophage M2
Macrophage M1
Macrophage
Hematopoietic stem cell
Granulocyte-monocyte progenitor
Eosinophil
Endothelial cell
Common myeloid progenitor
Common lymphoid progenitor
Class-switched memory B cell
-0.4
B cell plasma
B cell naive
B cell memory
B cell
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
⁎
P < 0.05
P < 0.01
P < 0.001
(a)
NRP 2
Correlation
stroma score
microenvironment score
immune score
0.6
T cell regulatory (Tregs)
T cell gamma delta
T cell NK
T cell CD8+ naive
T cell CD8+ effector memory
T cell CD8+ central memory
T cell CD8+
T cell CD4+ naive
0.3
T cell CD4+ memory
T cell CD4+ effector memory
T cell CD4+ central memory
T cell CD4+ Th2
T cell CD4+Thl
xCELL
T cell CD4+ (non-regulatory) Plasmacytold dendritic cell
Neutrophil
NK cell
Myeloid dendritic cell activated
0.0
Myeloid dendritic cell
Monocyte Mast cell
Macrophage M2
Macrophage M1
Macrophage
Hematopoletic stem cell
Granulocyte-monocyte progenitor
-0.3
Eosinophil
Endothelial cell
Common myeloid progenitor
Common lymphoid progenitor Class-switched memory B cell
B cell plasma
B cell naive
B cell memory
B cell
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
P < 0.05
⁎
P < 0.01
P < 0.001
(b)
Overall, these results suggest that the NRP1 and NRP2 expres- sion is associated with alterations in immune gene expression and infiltration in different cancers.
3.6. Association of NRP1 and NRP2 Expression with the TME in Various Cancers. The heterogeneity of TME across differ- ent cancers affects tumour drug resistance and modulates
Cancer: LGG
NRP1
NRP2
R= = 0.13, p =0.0032
R == 0.16,p=3e.04
RN Ass
R = 0.12, p = 0.0049
R == 0.11, p=0.0087
V
DNAss
·2:
R= 0.34, p = 1.2e-15
R = 0.16, p =0.00019
StromalScore
R = 0.22, p = 5.3e-07
R = 0.15, p= 0.00062
ImmuneScore
R= 0.27;p=2.62-10
R = 0.16, p=0.00024
RET
ESTIMATEScore
Gene expression
(a)
FIGURE 7: Continued.
Cancer: BLCA
NRP1
NRP2
R = - 0.3, p = 7.6e-10
R = 0.49, p .< 2.2e-16
RNAss
:R =- 0.31 p = 2.8e-10.
.R =:- 0.32, p .= 3.8e-11
DNAss
R = 0.54, p < 2.2e-16
R = 0.81, p < 2.2e-16
StromalScore
R = 0.47, p < 2.2e-16.
R = 0:58, p< 2.2e-16
ImmuneScore
:R = 0.54, p < 2.2e-16
R = 0:74, p ≤ 2.2e-16
RET
ESTIMATEScore
Gene expression
(b)
FIGURE 7: Continued.
Cancer: ACC
| NRP1 | NRP2 | |
|---|---|---|
| R =- 0.036, p= 0.76 | ·RF-0.33, ₺= 0.0032 RNAss | |
| ·R == 0.15, p = 0.2 | :R = 0.35, p = 0.0019 | |
| R = 0.093, p = 0.42 | DNAss R = 0.23, p =0.042 | |
| R= - 0.013, p =0.91 | StromalScore R = 0.26, p = 0.022 | |
| ImmuneScore | ||
| R = 0.034, p = 0.77 | ·R = 0.26, p -= 0.023 | |
| RETI | ESTIMATEScore |
Gene expression (c)
FIGURE 7: Continued.
Cancer: CESC
| NRP1 | NRP2 |
|---|---|
| R =- 0.34, p= 2e-09- | R =- 0:37, p == 1.6e-11- RNAss |
| R == 0.019, p = 0.74 R = 0.38; p .= 4.8e-12 R = 0.1, p = 0.075 | R == 0.05, p=0.38 DNAss R = 0.29, p .= 2.8e-07 StromalScore R == 0.023, p=0.68. ImmuneScore |
Gene expression (d) FIGURE 7: Continued. RETR
R = 0.24, p=1.8e-05
R=0.13, p=0.022
ESTIMATEScore
Cancer: KIRC
| NRP1 | NRP2 |
|---|---|
| ·R = = Q.43, p = 4.2e=15 | R =0.25, p = 7.32-06. |
| RNAss | |
| R =0.076, p= 0.18 | R =- 0.17, p= 0.002 |
| R =. 0.52, p .<: 2.2e-16 | DNAss R = 0.49, p.k 2.2e-16 StromalScore |
R =- 0.045, p -= 0.43
-R =- 0.18, p= 0.0011
ImmuneScore
R = 0.19, p = 0.001
R=0.37, p= 2.5e 11
RET
ESTIMATEScore
Gene expression (e)
FIGURE 7: Continued.
Cancer: KIRP
| NRP1 | NRP2 | |
|---|---|---|
| .R = 0.47, p .= 2.8e:16 | R =: 0.26,p .=. 1.7e-05 RNAss | |
| R = 0.042, p = 0.49 | R = 0.3, p =7.4e-07 | |
| DNAss | ||
| R = 0.089, p = 0.15: | R = 0.66, p < 2.2e-16 | |
| R == 0.04, p= 0.52. | StromalScore R = 0.58, p .< 2.2e-16 | |
| ImmuneScore | ||
| R = 0.017, p = 0.78. | R =. 0.65, p .< 2.2e-16 ESTIMATEScore | |
RET
Gene expression (f)
FIGURE 7: Continued.
2738, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/5546612 by National Library Of Medicine, Wiley Online Library on [05/04/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Cancer: MESO
NRP1
NRP2
R =- 0.19, p = 0.083:
R =- 0.32, p = 0.0028
RNAss
R = 0.15; p = 0.18
R= 0.31; p = 0.0037
DNAss
R = 0.56, p = 1.5e-08
R= 0.46, p= 7.6e-06
StromalScore
R = 0.14, p = 0.19
R = 0.14, p = 0.19
ImmuneScore
R =. 0.38, p= 0.00033
R=0.33, p= 0.0019
RET
ESTIMATEScore
Gene expression
g
FIGURE 7: Continued.
Cancer: STAD
NRP1
NRP2
R+=0.58, p< 2.2-16
R:G.66, $ 522e-16
RNAss
R =— 0.3,p =3:5e-08
R =:- 0.31,p = 6.62-09.
DNAss
R =. 0.73, p .< 2.2e-16
R =. 0.73, p .< 2.2e=16
StromalScore
R = 0.48, p < 2.22-16
R = 0.28, p = 4,2e-07.
ImmuneScore
R = 0.66, p < 2.2e-16
R = 0.54, p < 2.2e=16
ESTIMATEScore
Gene expression
(h)
cancer progression and metastasis [30, 31]. Here, we further explored the association of NRP1 and NRP2 expression with the immune microenvironment of some cancers (LGG, BLCA, ACC, CESC, KIRC, KIRP, MESO, and STAD). The ESTI- MATE algorithm was used to calculate, among other things, stem cell and immune cell indices in tumour cells. The expres- sion of NRP family genes in BLCA and LGG was found to be correlated most significantly with RNAss, DNAss, Stromal- Score, ImmuneScore, and ESTIMATEScore (Figures 7(a) and 7(b)). Overall, the NRP1 and NRP2 expression was positively correlated with StromalScore, ImmuneScore, and ESTIMATEScore in most prognosis-related cancers (Figures 7(a)-7(h)). Conversely, the correlation of the NRP1 and NRP2 expression with RNAss and DNAss was heteroge- neous across cancer types. In conclusion, expression of NRP family genes is associated with the TME of various cancers.
3.7. Association of NRP1 and NRP2 Expression with Clinicopathological Features in Various Cancers. Further anal- ysis demonstrated that the NRP1 and NRP2 expression was correlated with clinicopathological features of several cancers (KIRC, LGG, STAD, BLCA, and KIRP) (Figures 8(a)-8(e)). In patients with KIRC and STAD, NRP1 expression was signifi-
cantly correlated with ethnicity. The degree of NRP1 expres- sion was higher in Blacks and Asians. In BLCA, NRP2 expression was higher in Asian populations compared to Cau- casians. A high NRP1 and NRP2 expression was also found to be correlated with tumour diameter. In KIRC, a high NRP1 expression was associated with a larger tumour size, higher risk of distant metastases, and worse stage staging and grade staging. Similarly, a high NRP1 expression in STAD implied a worse grade staging. However, in LGG, a high NRP1 expres- sion implied a better grade staging. Furthermore, in BLCA, NRP2 expression was associated with tumour size, stage stag- ing, and worse grade staging. In KIPR, the NRP2 expression was higher in male patients.
3.8. Genome-Wide Association of NRP1 and NRP2 mRNA in Various Cancers. The previous results revealed that NRP1 might play important roles in KIRC and LGG, whereas NRP2 might play important roles in BLCA. Therefore, we analysed the association of KIRC, LGG, and BLCA with NRP1 and NRP2 in human genomic models (including gene expression, DNA methylation, somatic copy number, micro- RNA expression, somatic mutation, and protein level RPPA). The results showed that NRP1 was associated with genome-
KIRC-NRP1
KIRC-NRP1
C3
0.54
0.45
0
C3
0.72
0.69
0
C2
1.46 (*)
0
0.45
C2
2.53 (*)
0
0.69
C1
0
1.46 (*)
0.54
C1
0
2.53 (*)
0.72
Percentage (%)
100
Percentage (%)
100
75
75
50
50
25
25
0
0
C1
C2
C3
C1
C2
C3
Asian
T1
T3
Black
T2
T4
White
KIRC-NRP1
KIRC-NRP1
C3
0.94
0.86
0
C3
0.52
0.5
0
C2
2.16 (*)
0
0.86
C2
2.2 (*)
0
0.5
C1
0
2.16 (*)
0.94
C1
0
2.2 (*)
0.52
Percentage (%)
100
Percentage (%)
100
75
75
50
50
25
25
0
0
C1
C2
C3
C1
C2
C3
M0
III
M1
I
IV
RET
II
KIRC-NRP1
C3
1.05
1.05
0
3.64 (*)
0
1.05
0
3.64 (*)
1.05
Percentage (%)
75
50
25
0
C1
C2
C3
G1
G3
G2
G4
(a)
LGG-NRP1
C3
2.27 (*)
2.24 (*)
0
C2
5.87 (*)
0
2.24 (*)
C1
0
5.87 (*)
2.27 (*)
Percentage (%)
100
75
50
25
0
C1
C2
C3
G2
G3
(b)
STAD-NRP1
STAD-NRP1
C3
0.4
0.31
0
C3
1.17
1.1
0
C2
1.33 (*)
0
0.31
C2
3.39 *)
0
1.1
C1
0
1.33 (*)
0.4
C1
0
3.39 (*)
1.17
Percentage (%)
100
100
75
C3 ISLANDER WHITE (c) RETR
Percentage (
75
50
50
25
25
0
0
C1
C2
C1
C2
C3
ASIAN
G1
BLACK
G2
G3
FIGURE 8: Continued.
BLCA-NRP2
BLCA-NRP2
C3
1.02
1.01
0
C3
1.33 (*)
0.95
0
C2
3.54 (*)
0
1.01
C2
2.79 (*)
0
0.95
C1
0
3.54 (*)
1.02
C1
0
2.79 (*)
1.33 (*)
Percentage (%)
100
Percentage (%)
100
75
75
50
50
25
25
0
0
C1
C2
C3
C1
C2
C3
T1
T3
T1
T3
T2
T4
T2
T4
BLCA-NRP2
BLCA-NRP2
C3
1.87 (*)
1.33 (*)
0
C3
1.09
0.88
0
C2
4.63 (*)
0
1.33 (*)
C2
3.83 (*)
0
0.88
C1
0
4.63 (*)
1.87 (*)
C1
0
3.83 (*)
1.09
Percentage (%)
100
75
Percentage (%)
100
75
50
50
25
25
0
0
C1
C2
C3
C1
C2
C3
Asian
III
Black
I
IV
White
II
(d)
C3
0.56
0.6
0
C2
1.35 (*)
0
0.6
KIRP-NRP2
C1
0
1.35 (*)
0.56
Percentage (%)
100
G1: High expression of NRP1 or NRP2
75
G2: Low expression of NRP1 or NRP2
50
G3: Expression of NRP1 or NRP2 in the overall sample
25
0
C1
C2
C3
Female
Male
(e)
wide features in KIRC and LGG (Figures 9(a) and 9(b)), while NRP2 was broadly associated with genome-wide fea- tures in BLCA (Figure 9(c)).
4. Discussion
Data obtained from pancancer analysis has the potential to guide tumour control strategies and design of therapies
[32]. In recent years, genome-wide pancancer analysis has revealed mutations, RNA expression profiles, and immune profiles associated with tumour development. This has pro- vided numerous biomarkers for the diagnosis and treatment of tumours [33].
In this study, we used different tools to analyse the expression of NRPs in different tumours and its association with mutations, TME, immune landscape, and prognosis.
X
Y
1
22
21
20
2
.9
18
15
3
16
15
KIRC-NRP1
4
14
13
5
12
6
1
10
7
9
8
Variable types
Gene expression
MicroRNA expression
DNA methylation
Somatic mutation
Somatic copy number
Protein level-RPPA
(a)
X
Y
1
22
21
20
19
2
18
17
3
16
15
LGG-NRP1
4
14
13
5
12
6
11
10
7
9
8
Variable types
Gene expression
MicroRNA expression
DNA methylation
Somatic mutation
Somatic copy number
Protein level-RPPA
(b)
FIGURE 9: Continued.
X
Y
1
22
21
20
2
19
18
17
3
16
.5
BLCA-NRP2
4
14
13
5
12
6
11
10
7
9
8
Variable types
Gene expression
MicroRNA expression
DNA methylation
Somatic mutation
Somatic copy number
Protein level-RPPA
(c)
We found that neurovascular-associated NRPs can predict the prognosis of many cancers. Moreover, NRP1 and NRP2 were differentially expressed levels in different tissues. This suggests that they may play distinct roles in different cancers. Survival analysis demonstrated that a low NRP1 expression in ACC, CESC, LGG, and STAD was associated with poor patient prognosis, whereas a high NRP1 expression in KIRC predicted good prognosis. A high NRP2 expression in BLCA, KIRP, and MESO was associated with poor patient progno- sis. Further analysis revealed that NRP1 and NRP2 were significantly associated with TMB and MSI in various cancers. Moreover, the NRP1 and NRP2 expression was pos- itively correlated with the expression of immune checkpoint genes and immune infiltration. The expression level of NRPs was associated with the TME and clinicopathological features of cancers. Finally, genome-wide association analysis sug- gested that the NRP1 expression was closely associated with KIRC, whereas the NRP2 expression was closely associated with BLCA. Together with previous studies, we suggest that NRP2 may be involved in the development of various cancers, particularly BLCA.
NRPs are highly conserved, multifunctional transmem- brane proteins that are unique to vertebrates and are involved in various physiological and pathological processes in the body [34, 35]. In mammals, there are two isoforms of NRPs (NRP1 and NRP2) that are functionally distinct and comple- mentary. These genes are involved various biological pro-
cesses such as neuroangiogenesis, cell migration, and immune regulation [36, 37].
A high NRP1 expression has been reported to be closely associated with tumourigenesis and progression, which is consistent with our findings [38, 39]. Using NRP1 antago- nists, several studies have demonstrated the therapeutic potential of NRP1 in cancers [40]. Previous studies have also revealed that NRP1 modulates the function of various immune cells. In recent studies, NRP1 was found to regulate the stability and function of Tregs. It has also been reported to function as an antitumour immune inhibitor [41]. Anti- NRP1 treatment improved the efficacy of anti-PD-1 immu- notherapy. This indicates that immunotherapy targeting NRP1 may have good clinical outcomes [42]. NRP1 has also been previously found to promote tumour angiogenesis, tumour proliferation, and migration [43-48]. Anti-NRP1 therapy can block tumour angiogenesis and upregulate the antitumour immune response [49-52]. Currently, anti- NRP1 therapy is used as a potential antitumour treatment option [42, 53]. In conclusion, the results of our study reveal that anti-NRP1 therapy has good clinical benefits.
A high NRP2 expression in BLCA, KIRP, and MESO was associated with poor prognosis. Similar to our study, a high NRP2 expression in the bladder has been associated with chemoresistance and epithelial-to-mesenchymal transition [16]. In addition, a higher NRP2 expression has been reported in triple-negative breast cancers indicating that the
NRP2 expression depends on the type of breast cancer [14]. Moreover, the NRP2 expression in prostate cancer is posi- tively correlated with the Gleason grading [15]. NRP2 is closely related to the immune system [12]. The xCell algo- rithm was to first provide indirect data on the expression pattern of NRP2 in B cells, NPRs, natural killer cells, and T cells. Recent studies have shown that NRP2 regulates various processes such as cell migration and antigen migration in the immune system [12]. Similarly, this study reveals that NRP2 influences immune processes. NRP2 has also been found to be closely associated with metastasis and BRAFV600E in thyroid cancer [54]. Downregulation of NRP2 has been shown to influence epithelial-mesenchymal transition by affecting phosphorylation signaling pathways [54]. This suggests a potential association of NRP2 expression with the TME and gene mutations.
Energy metabolism is interconnected, coupled to insulin signaling, and linked to the release of metabolic hormones from adipose tissue. Understanding the diverse roles of energy metabolism should prevent and treat various human diseases such as diabetes, obesity, and cancer [55]. Previous studies have found that NRP1/2 may be involved in energy metabolism [56, 57]. Diabetes is an energy metabolism- related disease that can lead to multiple systemic pathologies [58-61]. And diabetes is closely associated with neurovascu- lar disease [62-65]. Therefore, we propose the bold hypothe- sis that NRP1/2 may also influence tumour prognosis through energy metabolism-related pathways.
However, there are limitations to this study that warrant further exploration. Firstly, the present study does not demonstrate how NRPs influence tumour growth and devel- opmental processes by affecting the immune microenviron- ment or the TME, as well as other pathways. Secondly, in vivo and in vitro experiments should be performed to sub- stantiate our results and clarify the impact of NRP expression on tumourigenesis development. Further studies at cellular and molecular levels would be beneficial to elucidate the specific functional mechanisms of NRPs in different cancer types. Thirdly, future well-designed studies are needed such as single-cell RNA sequencing. Further improvements in precision would be beneficial to prevent systematic bias at the cellular level. Therefore, future cohort studies and population-based case-control studies are necessary to exam- ine the mechanisms involved.
5. Conclusion
In conclusion, neurovascular-related NRP family genes are significantly correlated with the prognosis, TME, and immune profiles of tumours, especially in BLCA. Therefore, NRPs may be used as a marker for predicting the prognosis of various tumours. Besides, NRPs hold great promise as a potential target for tumour therapy.
Abbreviations
ACC: Adrenocortical carcinoma
BLCA: Bladder urothelial carcinoma
BRCA: Breast invasive carcinoma
CESC: Cervical squamous cell carcinoma
CHOL: Cholangiocarcinoma
COAD: Colon adenocarcinoma
DLBC: Lymphoid neoplasm diffuse large B-cell lymphoma
ESCA: Esophageal carcinoma
GBM: Glioblastoma multiforme
LGG: Brain lower grade glioma
HNSC: Head and neck squamous cell carcinoma
KICH: Kidney chromophobe
KIRC: Kidney renal clear cell carcinoma
KIRP: Kidney renal papillary cell carcinoma
LAML: Acute myeloid leukemia
LIHC: Liver hepatocellular carcinoma
LUAD: Lung adenocarcinoma
LUSC: Lung squamous cell carcinoma
MESO: Mesothelioma
OV: Ovarian serous cystadenocarcinoma
PAAD: Pancreatic adenocarcinoma
PCPG: Pheochromocytoma and paraganglioma
PRAD: Prostate adenocarcinoma
READ: Rectum adenocarcinoma
SARC:
Sarcoma
SKCM: Skin cutaneous melanoma
STAD: Stomach adenocarcinoma
TGCT: Testicular germ cell tumours
THCA: Thyroid carcinoma
THYM: Thymoma
UCEC: Uterine corpus endometrial carcinoma
UCS: Uterine carcinosarcoma
UVM: Uveal melanoma.
Data Availability
All data was obtained from the public database described in Materials and Methods.
Conflicts of Interest
No competing interests exist.
Authors’ Contributions
Chao Deng and Hang Guo contributed equally to this work.
References
[1] R. S. Apte, D. S. Chen, and N. Ferrara, “VEGF in signaling and disease: beyond discovery and development,” Cell, vol. 176, no. 6, pp. 1248-1264, 2019.
[2] R. S. Kerbel, “Tumor angiogenesis,” The New England Journal of Medicine, vol. 358, no. 19, pp. 2039-2049, 2008.
[3] N. Ferrara, “VEGF and intraocular neovascularization: from discovery to therapy,” Translational Vision Science & Technol- ogy, vol. 5, no. 2, p. 10, 2016.
[4] N. Ferrara and A. P. Adamis, “Ten years of anti-vascular endo- thelial growth factor therapy,” Nature Reviews. Drug Discov- ery, vol. 15, no. 6, pp. 385-403, 2016.
[5] G. C. Jayson, R. Kerbel, L. M. Ellis, and A. L. Harris, “Antian- giogenic therapy in oncology: current status and future direc- tions,” Lancet, vol. 388, no. 10043, pp. 518-529, 2016.
[6] S. Soker, S. Takashima, H. Q. Miao, G. Neufeld, and M. Klagsbrun, “Neuropilin-1 is expressed by endothelial and tumor cells as an isoform-specific receptor for vascular endo- thelial growth factor,” Cell, vol. 92, no. 6, pp. 735-745, 1998.
[7] X. Wang, S. Li, Y. Ma et al., “Identification of miRNAs as the crosstalk in the interaction between neural stem/progenitor cells and endothelial cells,” Disease Markers, vol. 2020, Article ID 6630659, 29 pages, 2020.
[8] S. Soker, M. Kaefer, M. Johnson, M. Klagsbrun, A. Atala, and M. R. Freeman, “Vascular endothelial growth factor-mediated autocrine stimulation of prostate tumor cells coincides with progression to a malignant phenotype,” The American Journal of Pathology, vol. 159, no. 2, pp. 651-659, 2001.
[9] A. L. Elaimy and A. M. Mercurio, “Convergence of VEGF and YAP/TAZ signaling: implications for angiogenesis and cancer biology,” Science Signaling, vol. 11, no. 552, article eaau1165, 2018.
[10] H. Chen, A. Chédotal, Z. He, C. S. Goodman, and M. Tessier- Lavigne, “Neuropilin-2, a novel member of the neuropilin family, is a high affinity receptor for the semaphorins Sema E and Sema IV but not Sema III,” Neuron, vol. 19, no. 3, pp. 547-559, 1997.
[11] T. Takahashi, A. Fournier, F. Nakamura et al., “Plexin-neuro- pilin-1 complexes form functional semaphorin-3A receptors,” Cell, vol. 99, no. 1, pp. 59-69, 1999.
[12] S. Schellenburg, A. Schulz, D. M. Poitz, and M. H. Muders, “Role of neuropilin-2 in the immune system,” Molecular Immunology, vol. 90, pp. 239-244, 2017.
[13] H. L. Goel and A. M. Mercurio, “VEGF targets the tumour cell,” Nature Reviews. Cancer, vol. 13, no. 12, pp. 871-882, 2013.
[14] H. L. Goel, B. Pursell, C. Chang et al., “GLI1 regulates a novel neuropilin-2/a6ß1 integrin based autocrine pathway that con- tributes to breast cancer initiation,” EMBO Molecular Medi- cine, vol. 5, no. 4, pp. 488-508, 2013.
[15] H. L. Goel, C. Chang, B. Pursell et al., “VEGF/neuropilin-2 reg- ulation of Bmi-1 and consequent repression of IGF-IR define a novel mechanism of aggressive prostate cancer,” Cancer Dis- covery, vol. 2, no. 10, pp. 906-921, 2012.
[16] A. Schulz, I. Gorodetska, R. Behrendt et al., “Linking NRP2 with EMT and chemoradioresistance in bladder cancer,” Fron- tiers in Oncology, vol. 9, p. 1461, 2020.
[17] V. Thorsson, D. L. Gibbs, S. D. Brown et al., “The immune landscape of cancer,” Immunity, vol. 51, no. 2, pp. 411-412, 2019.
[18] R. Bonneville, M. A. Krook, E. A. Kautto et al., “Landscape of microsatellite instability across 39 cancer types,” JCO Precision Oncology, vol. 2017, 2017.
[19] F. G. Frost, P. F. Cherukuri, S. Milanovich, and C. F. Boerkoel, “Pan-cancer RNA-seq data stratifies tumours by some hall- marks of cancer,” Journal of Cellular and Molecular Medicine, vol. 24, no. 1, pp. 418-430, 2020.
[20] V. Izzi, M. N. Davis, and A. Naba, “Pan-cancer analysis of the genomic alterations and mutations of the matrisome,” Can- cers, vol. 12, no. 8, p. 2046, 2020.
[21] Q. Zhang, R. Huang, H. Hu et al., “Integrative analysis of hypoxia-associated signature in pan-cancer,” iScience, vol. 23, no. 9, p. 101460, 2020.
[22] G. Sturm, F. Finotello, F. Petitprez et al., “Comprehensive eval- uation of transcriptome-based cell-type quantification methods for immuno-oncology,” Bioinformatics, vol. 35, no. 14, pp. i436-i445, 2019.
[23] G. Sturm, F. Finotello, F. Petitprez et al., “Comprehensive anal- yses of tumor immunity: implications for cancer immunother- apy,” Genome Biology, vol. 17, no. 1, p. 174, 2016.
[24] D. Aran, Z. Hu, and A. J. Butte, “xCell: digitally portraying the tissue cellular heterogeneity landscape,” Genome Biology, vol. 18, no. 1, p. 220, 2017.
[25] J. Wang, J. Sun, L. N. Liu et al., “TIMER2.0 for analysis of tumor-infiltrating immune cells,” Nucleic Acids Research, vol. 48, no. W1, pp. W509-W514, 2020.
[26] J. Wang, J. Sun, L. N. Liu et al., “Siglec-15 as an immune sup- pressor and potential target for normalization cancer immu- notherapy,” Nature Medicine, vol. 25, no. 4, pp. 656-666, 2019.
[27] D. Zeng, M. Li, R. Zhou et al., “Tumor microenvironment characterization in gastric cancer identifies prognostic and immunotherapeutically relevant gene signatures,” Cancer Immunology Research, vol. 7, no. 5, pp. 737-750, 2019.
[28] X. Zhang, B. Klamer, J. Li, S. Fernandez, and L. Li, “A pan- cancer study of class-3 semaphorins as therapeutic targets in cancer,” BMC Medical Genomics, vol. 13, Suppl.5, p. 45, 2020.
[29] T. M. Malta, A. Sokolov, A. J. Gentles et al., “Machine learning identifies stemness features associated with oncogenic dedif- ferentiation,” Cell, vol. 173, no. 2, pp. 338-354.e15, 2018.
[30] Z. R. Chalmers, C. F. Connelly, D. Fabrizio et al., “Analysis of 100,000 human cancer genomes reveals the landscape of tumor mutational burden,” Genome Medicine, vol. 9, no. 1, p. 34, 2017.
[31] M. Yarchoan, A. Hopkins, and E. M. Jaffee, “Tumor muta- tional burden and response rate to PD-1 inhibition,” The New England Journal of Medicine, vol. 377, no. 25, pp. 2500- 2501, 2017.
[32] F. X. Schaub, V. Dhankani, A. C. Berger, M. Trivedi, A. B. Richardson, R. Shaw et al., “Cancer Genome Atlas Network. Pan-cancer alterations of the MYC oncogene and its proximal network across The Cancer Genome Atlas,” Cell Systems, vol. 6, no. 3, pp. 282-300.e2, 2018.
[33] T. Schlomm, “Ergebnisse des “ICGC/TCGA pan-cancer anal- ysis of the whole genomes”(PCWAG)-konsortiums,” Urologe A., vol. 59, no. 12, pp. 1552-1553, 2020.
[34] X. Li, S. Fan, X. Pan et al., “Nordihydroguaiaretic acid impairs prostate cancer cell migration and tumor metastasis by sup- pressing neuropilin 1,” Oncotarget, vol. 7, no. 52, pp. 86225- 86238, 2016.
[35] N. Gioelli, F. Maione, C. Camillo et al., “A rationally designed NRP1-independent superagonist SEMA3A mutant is an effec- tive anticancer agent,” Science Translational Medicine, vol. 10, no. 442, article eaah4807, 2018.
[36] J. Wang, Y. Huang, J. Zhang et al., “NRP2 in tumor lymphan- giogenesis and lymphatic metastasis,” Cancer Letters, vol. 418, pp. 176-184, 2018.
[37] W. P. Li, H. Zhao, X. Zhang et al., “Study on the white matter neuronal integrity in amnestic mild cognitive impairment based on automating fiber-tract quantification,” Zhonghua Yi Xue Za Zhi, vol. 100, no. 3, pp. 172-177, 2020.
[38] L. E. Jimenez-Hernandez, K. Vazquez-Santillan, R. Castro- Oropeza et al., “NRP1-positive lung cancer cells possess tumor-initiating properties,” Oncology Reports, vol. 39, no. 1, pp. 349-357, 2018.
[39] H. Al-Shareef, S. I. Hiraoka, N. Tanaka et al., “Use of NRP1, a novel biomarker, along with VEGF-C, VEGFR-3, CCR7 and SEMA3E, to predict lymph node metastasis in squamous cell
carcinoma of the tongue,” Oncology Reports, vol. 36, no. 5, pp. 2444-2454, 2016.
[40] K. Jung, J. A. Kim, Y. J. Kim et al., “A neuropilin-1 antagonist exerts antitumor immunity by inhibiting the suppressive func- tion of intratumoral regulatory T cells,” Cancer Immunology Research, vol. 8, no. 1, pp. 46-56, 2020.
[41] C. Liu, A. Somasundaram, S. Manne et al., “Neuropilin-1 is a T cell memory checkpoint limiting long-term antitumor immu- nity,” Nature Immunology, vol. 21, no. 9, pp. 1010-1021, 2020.
[42] M. Leclerc, E. Voilin, G. Gros et al., “Regulation of antitumour CD8 T-cell immunity and checkpoint blockade immunother- apy by neuropilin-1,” Nature Communications, vol. 10, no. 1, p. 3345, 2019.
[43] Y. Hori, K. Ito, S. Hamamichi et al., “Functional characteriza- tion of VEGF- and FGF-induced tumor blood vessel models in human cancer xenografts,” Anticancer Research, vol. 37, no. 12, pp. 6629-6638, 2017.
[44] K. Appiah-Kubi, Y. Wang, H. Qian et al., “Platelet-derived growth factor receptor/platelet-derived growth factor (PDGFR/PDGF) system is a prognostic and treatment response biomarker with multifarious therapeutic targets in cancers,” Tumour Biology, vol. 37, no. 8, pp. 10053-10066, 2016.
[45] Y. Ding, J. Zhou, S. Wang et al., “Anti-neuropilin-1 monoclo- nal antibody suppresses the migration and invasion of human gastric cancer cells via Akt dephosphorylation,” Experimental and Therapeutic Medicine, vol. 16, no. 2, pp. 537-546, 2018.
[46] Y. Li, J. T. Luo, Y. M. Liu, and W. B. Wei, “miRNA- 145/miRNA-205 inhibits proliferation and invasion of uveal melanoma cells by targeting NPR1/CDC42,” International Journal of Ophthalmology, vol. 13, no. 5, pp. 718-724, 2020.
[47] Z. Ding, J. Zhu, Y. Zeng et al., “The regulation of neuropilin 1 expression by miR-338-3p promotes non-small cell lung can- cer via changes in EGFR signaling,” Molecular Carcinogenesis, vol. 58, no. 6, pp. 1019-1032, 2019.
[48] M. P. Barr, S. G. Gray, K. Gately et al., “Correction to: vascular endothelial growth factor is an autocrine growth factor, signal- ing through neuropilin-1 in non-small cell lung cancer,” Molecular Cancer, vol. 19, no. 1, p. 16, 2020.
[49] S. Rizzolio, G. Cagnoni, C. Battistini et al., “Neuropilin-1 upregulation elicits adaptive resistance to oncogene-targeted therapies,” The Journal of Clinical Investigation, vol. 128, no. 9, pp. 3976-3990, 2018.
[50] W. Hansen, “Neuropilin 1 guides regulatory T cells into VEGF-producing melanoma,” Oncoimmunology., vol. 2, no. 2, article e23039, 2013.
[51] W. Pang, M. Zhai, Y. Wang, and Z. Li, “Long noncoding RNA SNHG16 silencing inhibits the aggressiveness of gastric cancer via upregulation of microRNA-628-3p and consequent decrease of NRP1,” Cancer Management and Research, vol. - Volume 11, pp. 7263-7277, 2019.
[52] C. Teijeiro-Valiño, R. Novoa-Carballal, E. Borrajo et al., “A multifunctional drug nanocarrier for efficient anticancer ther- apy,” Journal of Controlled Release, vol. 294, pp. 154-164, 2019.
[53] H. Benachour, A. Sève, T. Bastogne et al., “Multifunctional peptide-conjugated hybrid silica nanoparticles for photody- namic therapy and MRI,” Theranostics., vol. 2, no. 9, pp. 889-904, 2012.
[54] G. Lee, Y. E. Kang, C. Oh et al., “Neuropilin-2 promotes growth and progression of papillary thyroid cancer cells,” Auris, Nasus, Larynx, vol. 47, no. 5, pp. 870-880, 2020.
[55] S. Marshall, “Role of insulin, adipocyte hormones, and nutrient-sensing pathways in regulating fuel metabolism and energy homeostasis: a nutritional perspective of diabetes, obe- sity, and cancer,” Science’s STKE, vol. 2006, no. 346, p. re7, 2006.
[56] A. A. van der Klaauw, S. Croizier, E. Mendes de Oliveira et al., “Human semaphorin 3 variants link melanocortin circuit development and energy balance,” Cell, vol. 176, no. 4, pp. 729-742.e18, 2019.
[57] C. King, D. Wirth, S. Workman, and K. Hristova, “Interactions between NRP1 and VEGFR2 molecules in the plasma mem- brane,” Biochimica et Biophysica Acta - Biomembranes, vol. 1860, no. 10, pp. 2118-2125, 2018.
[58] A. Falkowska, I. Gutowska, M. Goschorska, P. Nowacki, D. Chlubek, and I. Baranowska-Bosiacka, “Energy metabolism of the brain, including the cooperation between astrocytes and neurons, especially in the context of glycogen metabolism,” International Journal of Molecular Sciences, vol. 16, no. 11, pp. 25959-25981, 2015.
[59] Y. Xu, Q. Wang, Z. Wu et al., “The effect of lithium chloride on the attenuation of cognitive impairment in experimental hypo- glycemic rats,” Brain Research Bulletin, vol. 149, pp. 168-174, 2019.
[60] Y. S. Chen, X. R. Kang, Z. H. Zhou et al., “MiR-1908/EXO1 and MiR-203a/FOS, regulated by scd1, are associated with fracture risk and bone health in postmenopausal diabetic women,” Aging (Albany NY), vol. 12, no. 10, pp. 9549-9584, 2020.
[61] Y. Xu, Q. Wang, D. Li et al., “Protective effect of lithium chlo- ride against hypoglycemia-induced apoptosis in neuronal PC12 cell,” Neuroscience, vol. 330, pp. 100-108, 2016.
[62] A. Gasecka, D. Siwik, M. Gajewska et al., “Early biomarkers of neurodegenerative and neurovascular disorders in diabetes,” Journal of Clinical Medicine, vol. 9, no. 9, p. 2807, 2020.
[63] Z. Miao, X. Tang, M. Schultzberg, Y. Zhao, and X. Wang, “Plasma resolvin D2 to leukotriene B4 ratio is reduced in dia- betic patients with ischemic stroke and related to prognosis,” BioMed Research International, vol. 2021, Article ID 6657646, 8 pages, 2021.
[64] D. Xu, X. Chu, K. Wang et al., “Potential factors for psycholog- ical symptoms at three months in patients with young ische- mic stroke,” BioMed Research International, vol. 2021, Article ID 5545078, 7 pages, 2021.
[65] X. Chu, J. Zhang, B. Zhang, and Y. Zhao, “Analysis of age and prevention strategy on outcome after cerebral venous throm- bosis,” BioMed Research International, vol. 2020, Article ID 6637692, 6 pages, 2020.