Loading [Contrib]/a11y/accessibility-menu.js
Research article

Global analysis of carbon disulfide (CS2) using the 3-D chemistry transport model STOCHEM

  • Received: 13 March 2017 Accepted: 17 May 2017 Published: 22 May 2017
  • Carbon disulfide (CS2), a precursor to the long-lived carbonyl sulphide (OCS) is one of the main contributors to the atmospheric sulfate layer. The annual fluxes from its sources and sinks are investigated using a 3-D chemistry transport model, STOCHEM-CRI. In terms of the flux analysis, the oxidation of CS2 by OH is found to be the main removal process (76–88% of the total loss) and the dry deposition loss contributes 11–24% to the total loss of CS2. The global burden of CS2 was calculated, varying between 6.1 to 19.2 Tg and the lifetime of CS2 was determined to be within the range of 2.8–3.4 days. The global distribution of CS2 found the Northern Hemisphere (NH) continental landmasses to be the areas of concentration maxima with peak concentrations reaching up to 20 ppt during June-July-August (J-J-A) season and 40 ppt during December-January-February (D-J-F) season in anthropogenic source regions. Oceanic regions returned low CS2 levels of less than 2 ppt. The vertical profile of CS2 shows higher levels up to 3 ppt at 30°N–45°N during J-J-A and up to 4 ppt at 30°N–55°N during D-J-F. The oxidation of CS2 by OH can produce a substantial amount (0.58 Tg/yr) of atmospheric OCS and the annual average surface distribution of this flux shows up to 6 Gg/yr OCS formed in the regions with highest anthropogenic pollution (e.g., South east Asia). In general, the model-measurement comparison reveals an underprediction of model CS2 compared with measured CS2 for most of the regions. It is likely that the emissions of CS2 are being underestimated and there are likely much larger emission sources of atmospheric CS2 than previously suggested.

    Citation: Anwar Khan, Benjamin Razis, Simon Gillespie, Carl Percival, Dudley Shallcross. Global analysis of carbon disulfide (CS2) using the 3-D chemistry transport model STOCHEM[J]. AIMS Environmental Science, 2017, 4(3): 484-501. doi: 10.3934/environsci.2017.3.484

    Related Papers:

    [1] Dong-feng Li, Aisikeer Tulahong, Md. Nazim Uddin, Huan Zhao, Hua Zhang . Meta-analysis identifying epithelial-derived transcriptomes predicts poor clinical outcome and immune infiltrations in ovarian cancer. Mathematical Biosciences and Engineering, 2021, 18(5): 6527-6551. doi: 10.3934/mbe.2021324
    [2] Xin Lin, Xingyuan Li, Binqiang Ma, Lihua Hang . Identification of novel immunomodulators in lung squamous cell carcinoma based on transcriptomic data. Mathematical Biosciences and Engineering, 2022, 19(2): 1843-1860. doi: 10.3934/mbe.2022086
    [3] Jun Wang, Mingzhi Gong, Zhenggang Xiong, Yangyang Zhao, Deguo Xing . Immune-related prognostic genes signatures in the tumor microenvironment of sarcoma. Mathematical Biosciences and Engineering, 2021, 18(3): 2243-2257. doi: 10.3934/mbe.2021113
    [4] Jihong Yang, Hao Xu, Congshu Li, Zhenhao Li, Zhe Hu . An explorative study for leveraging transcriptomic data of embryonic stem cells in mining cancer stemness genes, regulators, and networks. Mathematical Biosciences and Engineering, 2022, 19(12): 13949-13966. doi: 10.3934/mbe.2022650
    [5] Kaiyu Shen, Shuaiyi Ke, Binyu Chen, Tiantian Zhang, Hongtai Wang, Jianhui Lv, Wencang Gao . Identification and validation of biomarkers for epithelial-mesenchymal transition-related cells to estimate the prognosis and immune microenvironment in primary gastric cancer by the integrated analysis of single-cell and bulk RNA sequencing data. Mathematical Biosciences and Engineering, 2023, 20(8): 13798-13823. doi: 10.3934/mbe.2023614
    [6] Xu Guo, Yuanming Jing, Haizhou Lou, Qiaonv Lou . Effect and mechanism of long non-coding RNA ZEB2-AS1 in the occurrence and development of colon cancer. Mathematical Biosciences and Engineering, 2019, 16(6): 8109-8120. doi: 10.3934/mbe.2019408
    [7] G. V. R. K. Vithanage, Hsiu-Chuan Wei, Sophia R-J Jang . Bistability in a model of tumor-immune system interactions with an oncolytic viral therapy. Mathematical Biosciences and Engineering, 2022, 19(2): 1559-1587. doi: 10.3934/mbe.2022072
    [8] Xiaowei Zhang, Jiayu Tan, Xinyu Zhang, Kritika Pandey, Yuqing Zhong, Guitao Wu, Kejun He . Aggrephagy-related gene signature correlates with survival and tumor-associated macrophages in glioma: Insights from single-cell and bulk RNA sequencing. Mathematical Biosciences and Engineering, 2024, 21(2): 2407-2431. doi: 10.3934/mbe.2024106
    [9] Tingting Chen, Wei Hua, Bing Xu, Hui Chen, Minhao Xie, Xinchen Sun, Xiaolin Ge . Robust rank aggregation and cibersort algorithm applied to the identification of key genes in head and neck squamous cell cancer. Mathematical Biosciences and Engineering, 2021, 18(4): 4491-4507. doi: 10.3934/mbe.2021228
    [10] Linqian Guo, Qingrong Meng, Wenqi Lin, Kaiyuan Weng . Identification of immune subtypes of melanoma based on single-cell and bulk RNA sequencing data. Mathematical Biosciences and Engineering, 2023, 20(2): 2920-2936. doi: 10.3934/mbe.2023138
  • Carbon disulfide (CS2), a precursor to the long-lived carbonyl sulphide (OCS) is one of the main contributors to the atmospheric sulfate layer. The annual fluxes from its sources and sinks are investigated using a 3-D chemistry transport model, STOCHEM-CRI. In terms of the flux analysis, the oxidation of CS2 by OH is found to be the main removal process (76–88% of the total loss) and the dry deposition loss contributes 11–24% to the total loss of CS2. The global burden of CS2 was calculated, varying between 6.1 to 19.2 Tg and the lifetime of CS2 was determined to be within the range of 2.8–3.4 days. The global distribution of CS2 found the Northern Hemisphere (NH) continental landmasses to be the areas of concentration maxima with peak concentrations reaching up to 20 ppt during June-July-August (J-J-A) season and 40 ppt during December-January-February (D-J-F) season in anthropogenic source regions. Oceanic regions returned low CS2 levels of less than 2 ppt. The vertical profile of CS2 shows higher levels up to 3 ppt at 30°N–45°N during J-J-A and up to 4 ppt at 30°N–55°N during D-J-F. The oxidation of CS2 by OH can produce a substantial amount (0.58 Tg/yr) of atmospheric OCS and the annual average surface distribution of this flux shows up to 6 Gg/yr OCS formed in the regions with highest anthropogenic pollution (e.g., South east Asia). In general, the model-measurement comparison reveals an underprediction of model CS2 compared with measured CS2 for most of the regions. It is likely that the emissions of CS2 are being underestimated and there are likely much larger emission sources of atmospheric CS2 than previously suggested.


    The tumor microenvironment (TME) is a complex ecosystem comprising numerous cells, including immune cells, stromal cells and immunosuppressive cells [1]. Tumour endothelial cells (TECs) are a prominent component of TME and TECs, which have critical functions on tumor growth, progression and metastasis [2]. As tumor angiogenesis is essential for tumor growth and metastasis, inducing tumor-associated angiogenesis is a promising tactic in cancer progression [3]. Angiogenesis is now accepted as an important hallmark of cancer [4] and endothelial cells (which form tight adhesions to ensure vessel integrity) are essentials in inducing angiogenesis in TME [3]. In TME, TECs interact with tumor cells via juxtacrine and paracrine signaling during tumor intravasation and metastasis [5].

    Colorectal cancer (CRC) is the fourth most common cancer and a leading cause of cancer mortality worldwide [6]. Recently stated that distinct stromal transcriptional and interactions are altered in colon cancer development [7] and stromal cells contribute to the CRC transcriptome [8]. It was also stated that stromal gene expression defines poor prognosis in colorectal cancer [9] and induces poor-prognosis subtypes in colorectal cancer [10]. Altogether, stromal cells have substantial influence and regulatory roles in colon cancer and endothelial cells (ECs) may cause colon carcinogenesis as a part of the stromal component.

    So, analyzing the transcriptomes of colon TECs (CTECs) in TME reveals new targets and avenues and explores more uncontrolled factors in colon cancer research. We present a comprehensive analysis through bioinformatic tools and identified molecular and genetic alterations in TECs. We identified DEGs from microarray datasets of gene expression omnibus (GEO), hub genes from the interactions of deregulated genes and find out significant KEGG pathways. We also identified significant hub genes that are linked with poor survival of patients. Moreover, we found that hub genes are associated with immune cells infiltration and positively correlated with immune suppressive markers. This integrated analysis provides vital molecular insights into CTECs characterization, which may directly affect treatment recommendations for colon cancer patients. It provides opportunities for genome-guided clinical trials as well as drug development for the treating of colon cancer patients.

    We searched the NCBI gene expression omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/) using the keywords "colon cancer endothelial cells, " "endothelial cells, " and "normal endothelial cells, " and identified one CTECs gene expression dataset GSE89287 (n = 24) (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE89287) [11]. This dataset included different cell types (endothelial cells, macrophages and epithelial cells) which were isolated from non-paired primary normal colon tissues and colorectal carcinomas and subsequently, RNA was isolated. In this study, we included only endothelial cells and excluded the other cell types from our study. Endothelial cells were isolated using the facs-sorting method. The dataset included 24 samples of endothelial cells, including 10 normal colon samples and 14 colon tumor samples. The platform of this data is GPL4133 (Agilent-014850 whole human genome microarray 4x44K G4112F) with a feature number version. Gene expression profiling interactive analysis (GEPIA) (http://gepia2.cancer-pku.cn/#index) dataset (n = 316) was used for the study of prognostic and expression levels of hub genes [12]. In addition, the correlation between the expression of hub genes and tumor-infiltrating immune cells was explored via gene modules in the TIMER dataset (n = 457) (https://cistrome.shinyapps.io/timer/) [13]. For identifying the gene-gene correlation, we used the TCGA-COAD dataset (n = 287) (https://gdc-portal.nci.nih.gov/) which was normalized into base log2.

    We used the web tool Network Analyst [14]to identify the significant DEGs between CTECs and CNECs. Dataset was normalized by quantile normalization and the R package "limma" was utilized to identify the DEGs between CTECs and CNECs. We utilized the Network Analyst tool to identify the average expression level of single gene having multiple probes in this gene expression study. We identified the DEGs with a threshold of |LogFC| > 0.585 and adjusted P value < 0.05.

    We performed gene-set enrichment analysis of the DEGs by GSEA [15]. We identified the KEGG [16] pathways that were significantly associated with the up-regulated and the down-regulated DEGs, respectively. We identified the significant pathways with a threshold of false discovery rate (FDR) < 0.05.

    To better know the relationship among these screened DEGs, the PPI network was established using the STRING database [17]. The hub genes in the PPI network were identified according to a degree using the node explorer module of Network Analyst software [14]. A hub gene was defined as a gene that was connected to no less than 10 other DEGs. A cytoscape plug-in molecular complex detection (MCODE) was employed to detect the modules from the PPI network [18]. We identified the significant modules based on the MCODE score and node number. The threshold of the MCODE was Node Score Cut-off: 0.2, Haircut: true, K-Core: 2 and maximum depth from Seed: 100.

    We compared the overall survival (OS) and the disease-free survival (DFS) of colon cancer patients classified based on gene expression levels (expression levels > median versus expression levels < median). Kaplan-Meier survival curves were used to show the survival differences and the log-rank test was utilized to evaluate the significance of survival differences. The prognostic roles of screened hub genes in COAD were analyzed using expression profiling interactive analysis (GEPIA) databases [12]. Cox P < 0.05 was considered as significant between two groups of patients.

    The expression levels of top hub genes were further validated using the gene expression profiling interactive analysis (GEPIA), a newly developed interactive web server for analyzing the RNA sequencing data [12]. We used TCGA COAD tumor data with matched normal. Hub genes with |log2FC| > 0.585 and P < 0.05 were considered statistically significant between the two groups.

    We utilized the "ESTIMATE" R package to quantify the immune scores (predict the immune cell quantity) and stromal scores (predict the stromal cells quantity) for each of the tumor samples [19]. Then we calculated Spearman's correlations between the expression levels of prognostic hub genes and immune and stromal scores. The threshold value of correlation is R > 0.30 and P-value is less than 0.001 (Spearman's correlation test).

    We analyzed the significant correlations of hub genes with the abundance of immune infiltrates, including B cells, CD4+ T cells, CD8+ T cells, neutrophils, macrophages and dendritic cells, via gene modules in TIMER [13]. In addition, Spearman's correlations between the expression level of hub genes and the immune inhibitory marker genes of tumor-infiltrating immune cells were explored via correlation modules in TIMER [13]. Moreover, we identified pearson's correlations between the expression level of hub genes and the marker genes of inhibitory immune cells in TCGA-COAD datasets using R software. The R package "ggplot2" was employed for drawing a correlation graph [20]. The gene markers of tumor-infiltrating immune cells included markers of monocytes, TAMs, M2 macrophages, T-helper 1 (Th1) cells, Tregs and exhausted T cells. These gene markers are illustrated in prior studies [12,21,22,23,24].

    We used R programming software for the statistical analysis of this study. For the differential expression analysis of gene expression data, only genes with |logFC| > 0.585 and adjusted P-value < 0.05 were considered statistically significant between the two groups. In the survival analysis, Cox P < 0.05 was considered as statistically significant. Spearman's correlation test between the ssGSEA scores and the expression level of specific prognostic genes was performed because these data were not normally distributed (P < 0.05) [25]. For analyzing the correlations between the expression levels of hub genes with the expression levels of other marker genes, we employed Pearson's correlation test because these data were normally distributed [25].

    Based on the log2FC and adjusted P-value, we identified 362 differentially expressed genes (DEGs) in CTECs relative to the CNECs, which included 117 up-regulated genes (Supplementary Table S1) and 245 down-regulated (Supplementary Table S2) in the CTECs samples. Table 1 demonstrates the top ten up-regulated genes (SPARC, IGFBP7, COL4A2, SELE, CXCL1, VWF, CCDC3, FSTL1, VWA1 and IFITM3) with significant statistical description. The most up-regulated gene SPARC is up-expressed in the stromal portion of CRC tissues [26]. The extracellular matrix-associated gene, COL4A2, is linked with cancer stemness [27]. Another up-regulated gene, CXCL1, is associated with increasing the metastatic ability of colon cancer by enhancing cell migration, matrix metalloproteinase-7 expression and EMT [28]. VWF, another up-regulated gene, is a plasma marker for the early detection of adenoma and colon cancer [29]. Table 2 illustrated the top ten down-regulated genes (IGKV1-5, IGLV1-44, IGKC, IGHV3-30, KRT81, IGHG1, IGHM, POU2AF1, IGKV2-24 and IGLJ3) with all significant statistical descriptions. Interestingly, most of the top down-regulated genes are associated with immunological activities. Some other down-regulated genes of CTECs have been demonstrated to be under-expressed in CRC. For example, the expression of POU2AF1 is down-regulated in the rat colon and in turn, reveals proximal-distal differences in the process of histone modifications and proto-oncogene expression [30]. Altogether, abnormally expressed numerous genes in CTECs compared to NTECs identified by the bioinformatic analysis have been associated with CRC pathology and carcinogenesis.

    Table 1.  List of top ten up-regulated genes.
    Entrez ID Symbols logFC P-Value Adjusted P-Value Name of gene
    6678 SPARC 2.21 0.0009 0.029 Secreted protein, acidic, cysteine-rich (osteonectin)
    3490 IGFBP7 2.19 0.0008 0.026 Insulin-like growth factor-binding protein 7
    1284 COL4A2 2.19 0.0002 0.016 Collagen, type IV, alpha 2
    6401 SELE 1.97 0.0017 0.038 Selectin E
    2919 CXCL1 1.85 0.0003 0.018 Chemokine (C-X-C motif) ligand 1 (melanoma growth stimulating activity, alpha)
    7450 VWF 1.85 0.0023 0.044 Von Willebrand factor
    83643 CCDC3 1.82 0.0004 0.019 Coiled-coil domain containing 3
    11167 FSTL1 1.68 0.0005 0.022 Follistatin-like 1
    64856 VWA1 1.62 0.0002 0.016 Von Willebrand factor A domain containing 1
    10410 IFITM3 1.59 0.0003 0.018 Interferon-induced transmembrane protein 3

     | Show Table
    DownLoad: CSV
    Table 2.  List of top ten down-regulated genes.
    Entrez ID Symbols logFC P-Value Adjusted P-Value Name of gene
    28299 IGKV1-5 −3.30 0.00058 0.024 Immunoglobulin kappa variable 1−5
    28823 IGLV1-44 −3.28 0.00004 0.009 Immunoglobulin lambda variable 1−44
    3514 IGKC −3.23 0.00020 0.016 Immunoglobulin kappa constant
    28439 IGHV3-30 −3.19 0.00002 0.008 Immunoglobulin heavy variable 330
    3887 KRT81 −3.19 0.00006 0.009 Keratin 81
    3500 IGHG1 −3.07 0.00004 0.009 Immunoglobulin heavy constant gamma 1 (G1m marker)
    3507 IGHM −2.98 0.00017 0.016 Immunoglobulin heavy constant mu
    5450 POU2AF1 −2.91 0.00007 0.010 POU class 2 associating factor 1
    28923 IGKV2-24 −2.75 0.00094 0.029 Immunoglobulin kappa variable 2−24
    28831 IGLJ3 −2.74 0.00032 0.018 Immunoglobulin lambda joining 3

     | Show Table
    DownLoad: CSV

    The functional enrichment analysis identifying the enriched up-regulated and down-regulated biological pathways that are associated with significant DEGs (Figure 1). GSEA pathway analysis revealed nine up-regulated pathways (focal adhesion, ECM-receptor interaction, pathways in cancer, small cell lung cancer, adherens junction, arrhythmogenic right ventricular cardiomyopathy (ARVC), pathogenic escherichia coli infection, cytokine-cytokine receptor interaction and proximal tubule bicarbonate reclamation) (Figure 1A) in CTECs. In addition, we also identified ten down-regulated pathways (protein export, cell adhesion molecules (CAMs), arrhythmogenic right ventricular cardiomyopathy (ARVC), N-Glycan biosynthesis, the intestinal immune network for IgA production, primary immunodeficiency, hypertrophic cardiomyopathy (HCM), ECM-receptor interaction, dilated cardiomyopathy and vasopressin-regulated water reabsorption) in CTECs (Figure 1B). Recently, it was stated that focal adhesion, ECM-receptor interaction, pathways in cancer, small cell lung cancer, adherens junction and cytokine-cytokine receptor interaction pathways are dysregulated in colon tumors stroma [7].

    Figure 1.  KEGG pathway enrichment analysis of significant DEGs. A. Significantly up-regulated pathways in colon tumor endothelial cells. B. Significantly down-regulated pathways in colon tumor endothelial cells. FDR: false discovery rate.

    Using all 362 DEGs, PPI networks were constructed using STRING software and hub genes were identified using the node explorer of network analyst software. Many significant DEGs are involved in the networks (Supplementary Table S3). The identification of hub genes explored that FN1, GAPDH, COL1A1, COL1A2, CTNNB1, LAMB1, LAMC1, SPARC, IGFBP3 and COL4A1 are top up-regulated hub genes and HSP90B1, PDIA6, ITGA4, DDOST, SDC1, CKAP4, CD38, MANF, ITGB7 and CCR2 are top down-regulated hub genes in CTECs respectively (Figure 2 and Table 3). Interestingly, we found that top-up-regulated (Table 1) SPARC also acts as a hub gene in the PPI network. FN1 suppressed apoptosis and promoted viability, invasion and migration of tumor cells in CRC [31]. It was showed that COL1A1 and COL1A2 are associated with colon cancer [32,33]. The high expression levels of CTNNB1 positively correlated with metastasis of colon cancer [34]. Deregulation of SDC1 contributes to cancer progression by promoting cell proliferation, metastasis, invasion and angiogenesis [35]. Altogether, it may be stated that CTECs-derived these hub genes may contribute to colon cancer pathogenesis.

    Figure 2.  The functional enrichment analysis of MCODE derived clusters. A. PPI interaction of Cluster 1. B. Enrichment of the KEGG pathways in Cluster 1. C. PPI interaction of Cluster 3. D. Enrichment of the 11 KEGG pathways in Cluster 3. E. Interaction of 23 nodes in Cluster 4. F. The enrichment of 11 KEGG pathways in Cluster 4.
    Table 3.  The top ten up-regulated and top ten down-regulated hub genes are associated with PPI.
    Regulatory status Symbols Degree logFC Adjusted P-Value Name of gene
    Up-regulated FN1 65 1.14 0.020 Fibronectin 1
    GAPDH 58 0.73 0.010 Glyceraldehyde-3-phosphate dehydrogenase
    COL1A1 39 1.06 0.041 Collagen, type I, alpha 1
    COL1A2 32 1.01 0.032 Collagen, type I, alpha 2
    CTNNB1 32 0.79 0.007 Catenin (cadherin-associated protein), beta 1, 88kda
    LAMB1 31 1.10 0.021 Laminin, beta 1
    LAMC1 29 0.76 0.031 Laminin, gamma 1 (formerly LAMB2)
    SPARC 27 2.20 0.028 Secreted protein, acidic, cysteine-rich (osteonectin)
    IGFBP3 24 1.47 0.008 Insulin-like growth factor-binding protein 3
    COL4A1 24 1.10 0.017 Collagen, type IV, alpha 1
    Down-regulated HSP90B1 42 −0.83 0.008 Heat shock protein 90kda beta (Grp94), member 1
    PDIA6 32 −0.730 0.041 Protein disulfide isomerase family A, member 6
    ITGA4 25 −0.72 0.020 Integrin, alpha 4 (antigen CD49D, alpha 4 subunit of VLA-4 receptor)
    DDOST 22 −0.78 0.015 Dolichyl-diphosphooligosaccharide-protein glycosyltransferase
    SDC1 21 −1.09 0.009 Syndecan 1
    CKAP4 20 −0.74 0.020 Cytoskeleton-associated protein 4
    CD38 20 −1.48 0.012 CD38 molecule
    MANF 18 −0.67 0.018 Mesencephalic astrocyte-derived neurotrophic factor
    ITGB7 18 −1.89 0.017 Integrin, beta 7
    CCR2 17 −0.73112 0.022 Chemokine (C-C motif) receptor 2

     | Show Table
    DownLoad: CSV

    Moreover, MCODE identified 15 clusters from the original PPI networks. The description of MCODE derived clusters is illustrated in Table 4. The top significant cluster 1 contained 15 nodes and 105 edges (Figure 2 and Table 4). We identified the functional enrichment of KEGG pathways for all clusters by using the GSEA. Interestingly, we found that eight of the clusters out of 15 are associated with the enrichment of KEGG pathways. Gene set of Cluster 1 is related to the enrichment of 4 pathways: ECM-receptor interaction, small cell lung cancer, pathways in cancer and focal adhesion (Figure 2A, B). Cluster 3 (Figure 2C, D), Cluster 4 (Figure 2E, F) and Cluster 9 (Cell adhesion molecules and natural killer cell-mediated cytotoxicity) are associated with the enrichment of immune, stromal and cancer-associated pathways. Some of the enriched pathways in Cluster 3 included ECM-receptor interaction, focal adhesion, small cell lung cancer, cell adhesion molecules, the intestinal immune network for IgA production, pathways in cancer and hematopoietic cell lineage. In addition, some of the pathways enriched in Cluster 4 are the chemokine signaling pathway, adherens junction, cytokine-cytokine receptor interaction, focal adhesion, NOD-like receptor signaling pathway, pathways in cancer, leukocyte trans-endothelial migration and tight junction. Protein export is enriched in Cluster 2 and the B cell receptor signaling pathway is enriched in Cluster 5. Cluster 13 (Fatty acid metabolism and PPAR signaling pathway) and cluster 15 (Glycosphingolipid biosynthesis-Lacto and neolacto series and N-Glycan biosynthesis) are mainly associated with the enrichment of metabolic pathways. Uddin et al. also identified some of these pathways, including pathways in cancer, small cell lung cancer, ECM-receptor interaction, focal adhesion, natural killer cell-mediated cytotoxicity and cell adhesion molecules, are enriched in colon tumor stroma [7,9]. Altogether, it can be stated that this hub PPI from CTECs may contribute to colon carcinogenesis.

    Table 4.  MCODE identified 15 clusters from the PPI networks.
    Cluster Score Nodes Edges Node Ids
    1 15 15 105 FSTL1, IGFBP5, IGFBP3, APLP2, HSP90B1, PDIA6, FN1, PRSS23, TMEM132A, VWA1, CKAP4, LAMB1, IGFBP7, LAMC1, CYR61
    2 9.556 10 43 DDOST, SSR3, SEC61B, SSR4, SPCS2, SPCS1, SPCS3, TRAM1, OSTC, SEC11C
    3 9.2 11 46 ITGA8, ITGA5, COL4A2, LAMA4, COL12A1, COL4A1, SPARC, ITGB7, COL1A1, ITGA4, COL1A2
    4 5.364 23 59 CCR10, SERP1, OS9, DNAJB11, ABCG2, SDF2L1, HERPUD1, SEL1L, ACTG1, MET, GNG7, CCR2, CTNNB1, SDC1, XBP1, MMP1, DNAJB9, APLN, PNOC, CXCL1, TIMP2, CXCL2, SNAI1
    5 4.857 8 17 CD27, CD79A, UCHL1, IGLL5, SLAMF1, SLAMF7, CD81, CD38
    6 4.5 5 9 TUBA1B, CCT2, CCT3, DNAJA4, CCT4
    7 4.5 5 9 CLCA4, ZG16, SLC26A3, GUCA2A, GUCA2B
    8 4 5 8 DERL2, DERL3, ERLEC1, CRELD2, UBE2J1
    9 4 4 6 SPN, CD48, GZMB, CD4
    10 3 3 3 MZB1, TNFRSF17, POU2AF1
    11 3 3 3 FCGBP, CLCA1, RETNLB
    12 3 3 3 PLAC8, RNASET2, CECR1
    13 3 3 3 HMGCS2, ACSL1, ACADVL
    14 3 3 3 IFITM3, IFITM2, ISG20
    15 2.667 4 4 B4GALT3, ST3GAL6, MGAT1, MANEA

     | Show Table
    DownLoad: CSV

    We considered the top ten up-regulated and top ten down-regulated genes from the original PPI network for survival analysis. We selected the TCGA-COAD database for screening OS and DFS. We got COL1A1, COL1A2, IGFBP3, SPARC and DDOST hub genes are significantly involved with patient survival time (Figure 3). We found that the higher expression level of up-regulated COL1A1, COL1A2, IGFBP3 and SPARC are worse for patient's DFS and COL1A2 is also associated with poor OS. In addition, the low expression level of down-regulated DDOST is more inferior for patient's OS and DFS. It showed that higher expressions of COL1A1 were related to poor survival in colon cancer patients [33]. COL1A2 gene is also associated with the prognosis of IIA stage colon cancer [32]. Stromal SPARC had a pro-metastatic impact in vitro and was a characteristic of aggressive tumors with a poor prognosis in CRC patients [26]. Collectively, hub genes of CTECs are prognostic markers in COAD and may be contributed to colon carcinogenesis.

    Figure 3.  Survival analysis of individual hub genes. Up-regulated COL1A1, COL1A2, IGFBP3 and SPARC and down-regulated DDOST have been associated with the poor prognosis in COAD.

    Interestingly, we found that all four up-regulated hub genes (COL1A1, COL1A2, IGFBP3 and SPARC) were consistently up-regulated in TCGA COAD samples versus normal samples (P < 0.05) (Figure 4). It indicates that CTECs may have a contribution to the COAD-derived transcriptomes in colon cancer. However, one of the down-regulated hub genes, DDOST, showed no significant expression differences between TCGA-COAD samples versus normal samples (P < 0.05), suggesting that this hub gene is expressed explicitly in CTECs.

    Figure 4.  Validation of the expression levels of prognostic hub genes between colon cancer and normal tissue (P < 0.05, |Log FC| > 0.585).

    Tumor-infiltrating lymphocytes are independent prognostic factors for better survival and sentinel lymph node status in cancers [36]. Genes highly expressed in the microenvironment are expected to negatively associate with tumor purity, while the opposite is expected for genes highly expressed in the tumor cells [13]. So, we assessed whether the expression of hub genes was correlated with immune infiltration levels in COAD. We investigated the correlations of prognostic hub genes (COL1A1, COL1A2, IGFBP3, SPARC and DDOST) with immune scores and stromal scores in COAD. Interestingly, we found that the expression levels of COL1A1, COL1A2, IGFBP3 and SPARC positively correlated with the immune scores (Figure 5A) and stromal scores (Figure 5B) (Spearman's correlation test, R > 0.30, P < 2.2e−16). This result indicated that the prognostic hub genes are associated with the regulation of the tumor microenvironment.

    Figure 5.  Up-regulated prognostic hub genes are associated with the regulation of the tumor microenvironment in COAD. The expression level of COL1A1, COL1A2, IGFBP3 and SPARC positively correlated with the immune scores (A) and stromal scores (B).

    In addition, we identified the correlations between the expression of hub genes and the immune infiltration levels in COAD by using the TIMER tool. The results discovered that up-regulated COL1A1, COL1A2, IGFBP3 and SPARC expression have significant (P ≤ 0.05) correlations with tumor purity in COAD but down-regulated DDOST not correlated with tumor purity (Figure 6). In addition, COL1A1 expression has significant correlations with infiltrating levels of CD4+ T cells, macrophage, neutrophil and dendritic cells. We also found that a significant positive correlation between COL1A2, IGFBP3 and SPARC expression and infiltration of all five immune cells (CD4+ T cells, CD8+ T cells, Neutrophils, Macrophages and Dendritic cells) in COAD. Besides, we found significant negative correlations between DDOST expression and infiltration of all five immune cells (CD4+ T cells, CD8+ T cells, Neutrophils, Macrophages and Dendritic cells) (Figure 6). Recently it was stated that macrophages and neutrophils are associated with immunosuppressive inflammation to modulate anti-tumor immunity [37]. DCs can promote tumor metastasis by increasing Treg cells and reducing CD8+ T cell cytotoxicity [38]. These results suggest that the expression of COL1A1, COL1A2, IGFBP3, SPARC and DDOST has potential roles in tumor purity and immune cell infiltrations in colon cancer.

    Figure 6.  Hub genes are correlated with immune infiltration level and tumor purity. Expression of COL1A1, COL1A2, SPARC, IGFBP3 and DDOST are significantly negatively correlated to tumor purity and have significant positive correlations with infiltrating levels of CD4+ T cells, macrophages, neutrophils and dendritic cells in COAD. The correlation analyses were performed using TIMER [13]. RSEM: RNA-Seq by Expectation Maximization [39].

    To elucidate the relationship between up-regulated hub genes COL1A1, COL1A2, IGFBP3 and SPARC and the diverse immune infiltrating cells, we find out the correlations between these genes and immune marker sets of various immune cells including monocytes, TAMs, M2 macrophages, Th1, Tregs and T cell exhaustion (Table 5). Surprisingly, we got that the hub genes COL1A1, COL1A2, IGFBP3 and SPARC are positively correlated (Pearson correlation test, P ≤ 0.001) with the immune markers of monocytes, TAMs, M2 macrophages, Th1, Tregs and T cell exhaustion. These findings suggest that the high expression level of hub genes plays an essential role in the infiltration of monocytes, TAMs, M2 macrophages, Th1, Tregs and T cell exhaustion in COAD. It was recently reported that LAYN acts as a prognostic biomarker for determining prognosis and immune infiltration in gastric and colon cancers [21]. Potent immunosuppressive T-regulatory cells (Tregs) are found in a vast array of tumor types and tumor-infiltrating Tregs are often associated with poor clinical outcomes. Tregs also promote cancer progression through their ability to limit anti-tumor immunity and promote angiogenesis [40]. Another marker, FOXP3, plays an important role in Treg cells which leads to the suppression of cytotoxic T cells attacking tumor cells [40]. TIM-3 is a crucial surface protein on exhausted T cells, a critical gene that regulating T cell exhaustion [41].

    Table 5.  Correlation analysis between survival-associated hub genes and markers of immune-inhibitory cells in TCGA-COAD data.
    Immune cell Gene markers SPARC COL1A1 COL1A2 IGFBP3
    R P R P R P R P
    Monocyte CD86 0.69 1.49E−42 0.65 2.71E−35 0.67 3.02E−39 0.55 2.39E−24
    CSF1R 0.71 1.22E−44 0.68 5.80E−41 0.72 1.60E−46 0.59 9.29E−28
    TAM CCL2 0.76 3.55E−56 0.69 1.00E−41 0.72 2.51E−47 0.58 2.49E−27
    CD68 0.49 8.62E−19 0.48 5.34E−18 0.49 1.16E−18 0.46 1.25E−16
    IL10 0.57 1.86E−26 0.52 4.70E−21 0.54 2.65E−23 0.42 1.87E−13
    M2 macrophage CD163 0.71 4.72E−46 0.71 4.11E−46 0.73 7.07E−49 0.52 1.88E−21
    VSIG4 0.73 8.85E−50 0.70 2.16E−43 0.71 6.58E−46 0.56 7.05E−25
    MS4A4A 0.70 2.77E−44 0.65 1.77E−35 0.67 7.62E−39 0.54 6.22E−23
    Th1 T-bet (TBX21) 0.36 3.16E−10 0.38 1.66E−11 0.38 3.46E−11 0.37 8.76E−11
    IFN-γ (IFNG) 0.18 0.001 0.21 0.00040 0.20 0.000834 0.16 0.00589
    TNF-α (TNF) 0.40 3.37E−12 0.38 1.65E−11 0.40 3.16E−12 0.21 0.00030
    Treg FOXP3 0.56 6.11E−25 0.55 3.52E−24 0.58 1.49E−27 0.50 1.01E−19
    CCR8 0.56 1.13E−24 0.54 4.14E−23 0.58 2.02E−27 0.48 1.23E−17
    TGFβ (TGFB1) 0.74 1.08E−50 0.73 1.99E−49 0.75 1.61E−52 0.62 6.64E−32
    T cell exhaustion PD-1 (PDCD1) 0.32 3.27E−08 0.32 2.35E−08 0.33 9.61E−09 0.35 8.10E−10
    CTLA4 0.42 1.40E−13 0.42 1.10E−13 0.44 7.38E−15 0.40 3.58E−12
    LAG3 0.25 1.56E−05 0.29 6.10E−07 0.28 1.94E−06 0.27 3.07E−06
    TIM-3(HAVCR2) 0.70 1.48E−43 0.67 2.34E−38 0.69 3.05E−41 0.54 1.58E−23
    TIGIT 0.39 4.84E−12 0.41 7.99E−13 0.42 5.46E−14 0.40 2.49E−12
    CXCL13 0.35 1.82E−09 0.36 5.46E−10 0.35 8.72E−10 0.31 1.34E−07
    LAYN 0.89 2.42E−97 0.79 1.39E−63 0.83 4.05E−75 0.67 2.35E−38
    *Note: R is Pearson correlation and P is p-value in Pearson's correlation test.

     | Show Table
    DownLoad: CSV

    Since the elevated expression of prognostic hub genes were positively associated with the immune inhibitory markers PD-1 and TGFB1 (Table 5), we expect the expression levels of PD-L1 (CD274), PD-L2 (PDCD1LG2) and TGFBR1 could be positively associated with the prognosis-associated hub genes. Interestingly, we found that PDL1 (CD274), PDL2 (PDCD1LG2) and TGFBR1 are moderate to highly correlated with all four-prognosis related highly expressed hub genes: COL1A1, COL1A2, SPARC and IGFBP3 (Figure 7). T cell immunity recovered by PD-1/PD-L1 immune checkpoint blockade has been demonstrated to be a promising cancer therapeutic strategy [42]. The expression of PD-L2 was observed in tumors, stroma and endothelial cells and PD-L2 expression is relevant to anti-PD-1 therapy in cancer [43]. TGFBR1 is one of the receptors for TGF-β ligands. The TGF-β signaling pathway is crucially associated with stimulation, induction and maintenance of EMT in cancers [44]. TGFBR1 is up-regulated in multiple malignancies and dysregulation of TGFBR1 leads to tumorigenesis by controlling cellular signaling, which in turn is associated with colorectal cancer risk [45]. Altogether, it suggests that the prognostic associated hub genes may regulate the immune-suppressive activities of TME through the interactions with the immune marker in colon cancer.

    Figure 7.  Prognostic hub genes are significantly positively correlated with immune suppressive PD-L1, PD-L2 and TGFBR1. Up-regulated four hub genes (SPARC, COL1A2, COL1A2 and IGFBP3) are significantly correlated with three prominent immune-inhibitory markers (PD-L1, PD-L2 and TGFBR1). The correlation analyses were performed using R software. The Pearson's correlation test p-values (P) and correlation coefficients (R) are shown in the figure.

    TECs are the substantial component of tumor stroma in TME and these cells are associated with tumor malignancies, progression and metastasis [5]. This present study conducted a differential expression analysis using gene expression profiling data from the GEO database. We identified 362 DEGs in CTECs, including 117 up-regulated genes (SPARC, COL4A2, CXCL1 and VWF) and 245 down-regulated genes. Interestingly, we found that many immunological genes are down-regulated in CTECs (Table 2). Subsequent pathway enrichment analysis revealed that up-regulated DEGs were significantly enriched with cancer, cellular development and immune regulation. Down-regulated DEGs were enriched considerably with mostly immunodeficiency and metabolism-related pathways (Figure 1). These results revealed the abnormal cellular growth, immune regulatory, cancerous and metabolism-associated pathways in CTECs.

    Next, we employed DEGs to construct a PPI network and extracted significant clusters from the original PPI network (Figure 2). Finally, we identified hub genes (degree > 10) that are dysregulated in CTECs. We identified CTECs-derived five hub genes (COL1A1, COL1A2, SPARC, IGFBP3 and DDOST) whose expression was significantly associated with the poor prognosis in TCGA-COAD data (Figure 3). These hub genes are mainly involved in protein digestion and absorption (COL1A1 and COL1A2), cellular signaling (SPARC and IGFBP3) and enzymatic action of metabolism (DDOST), suggesting that the deregulation of these cellular functions in CTECs may contribute to the altered prognosis in colon cancer. In addition, the expression analysis of these prognostic hub genes in colon cancer was further evaluated using the GEPIA database [12]. We found that four hub genes (COL1A1, COL1A2, SPARC and IGFBP3) are also significantly up-regulated in TCGA-COAD samples. Still, another down-regulated hub gene (DDOST) is not considerably down-regulated in TCGA-COAD samples (Figure 4). It indicates that CTECs-associated gene signature substantially contributed to colon carcinogenesis.

    Furthermore, our study showed that the immune infiltration levels and diverse immune marker sets are correlated with the expression levels of COL1A1, COL1A2, SPARC and IGFBP3 hub genes. Our analysis demonstrated that a significant positive correlation between the COL1A1, COL1A2, SPARC and IGFBP3 expression levels and infiltration level of CD4+ T cells, macrophages, neutrophils and DCs in COAD (Figure 6). These results are suggesting that the potential immunological roles of hub genes in colon TME. In addition, our study indicated that COL1A1, COL1A2, SPARC and IGFBP3 could activate Tregs and induce T cell exhaustion. The increase in up-regulated four hub genes expression positively correlates with Treg and T cell exhaustion markers (Table 5). Other markers of monocytes, TAM, M2 macrophage, Th1, are also positively correlated with the expression of these hub genes (Table 5). Some other crucial immune inhibitory markers, PD-L1, PD-L2 and TGFBR1, are significantly correlated with the expression level of COL1A1, COL1A2, SPARC and IGFBP3 (Figure 7). Altogether, the prognostic hub genes are associated with immunosuppressive colon TME.

    Since our analysis identified numerous CTECs-derived transcriptomes that are critically associated with colon cancer, we speculated that the endothelial cells might contribute to the other vascular pathologies, including the central nervous system (CNS) and other diseases. It was stated that cerebral endothelial cells play an active role in the pathogenesis of CNS inflammatory diseases [46]. Endothelial cells are associated with the leukocyte migration in CNS and regulating the leukocyte-endothelial cell interactions and the crosstalk between endothelial cells and glial cells or platelets in CNS [46]. The major vascular anomalies of the nervous system included cerebral cavernous malformations (CCMs) and arteriovenous malformations (AVMs) [47,48]. These CNS-associated vascular diseases are characterized by genetic alterations, such as gene polymorphisms and mutations [47,48,49,50,51]. For example, germline mutation enrichment in pathways controlling the endothelial cell homeostasis in patients with brain arteriovenous malformation [48]. Altogether, our findings may provide insights to find the roles of endothelial cells in diseases, including cancer, CNS-associated disease and other vascular pathologies.

    This study identified numerous deregulated transcriptomes in the CTECs that could be used as biomarkers for the diagnosis and prognosis of colon cancer and may provide therapeutic targets for colon cancer. However, to translate these findings into clinical application, further experimental and clinical validation would be necessary.

    In summary, we identified potential CTECs-derived significantly deregulated transcriptomes that are involving with the pathogenesis, prognosis and immune inhibition within the tumor microenvironment of colon cancer. This study reveals a potential regulatory mechanism of CTECs in the tumor microenvironment of colon cancer and may contribute to revealing CTECs-colon cancer cellular cross-talk.

    The authors declare no conflict of interest.

    We thank all laboratory staff and technicians for their participation in this study. The present study was supported by the project from Xinjiang Health Poverty Alleviation Work Appropriate Technology Promotion (Grant no. SYTG-Y202035).

    Jie Wang and Md. Nazim Uddin contributed equally to this work. Jie Wang and Md. Nazim Uddin conceived the research, designed analysis strategies. Rehana Akter wrote the manuscript and Yun Wu conceived the study and wrote the manuscript. All the authors read and approved the final manuscript.

    [1] Wine PH, Chameides WL, Ravishankara AR (1981) Potential role of CS2 photooxidation in tropospheric sulfur chemistry. Geophys Res Lett 8: 543-546. doi: 10.1029/GL008i005p00543
    [2] Andreae MO (1990) Ocean-atmosphere interactions in the global biogeochemical sulfur cycle. Mar Chem 30: 1-29. doi: 10.1016/0304-4203(90)90059-L
    [3] Khalil MAK, Rasmussen RA (1984) Global sources, lifetimes and mass balances of carbonyl sulphide (OCS) and carbon disulfide (CS2) in the Earth's atmosphere. Atmos Environ 18: 1805-1813. doi: 10.1016/0004-6981(84)90356-1
    [4] Leck C, Rodhe H (1991) Emissions of marine biogenic sulfur to the atmosphere of northern Europe. J Atmos Chem 12: 63-86. doi: 10.1007/BF00053934
    [5] Kim K-H, Andreae MO (1992) Carbon disulfide in the estuarine, coastal, and oceanic environments. Marine Chem 40: 179-197. doi: 10.1016/0304-4203(92)90022-3
    [6] Xie H, Moore RM, Miller WL (1998) Photochemical production of carbon disulphide in seawater. J Geophys Res 103: 5635-5644. doi: 10.1029/97JC02885
    [7] Xie H, Moore RM, Miller WL (1999) Carbon disulfide in the North Atlantic and Pacific oceans. J Geophys Res 104: 5393-5402. doi: 10.1029/1998JC900074
    [8] Kettle AJ, Rhee TS, von Hobe M, et al. (2001) Assessing the flux of different volatile sulfur gases from the ocean to the atmosphere. J Geophys Res 106: 12193-12209. doi: 10.1029/2000JD900630
    [9] Chin M, Davis DD (1993) Global sources and sinks of OCS and CS2 and their distributions. Global Biogeochem. Cycles 7: 321-337. doi: 10.1029/93GB00568
    [10] Watts SF (2000) The mass budgets of carbonyl sufide, dimethyl sulphide, carbon disulfide and hydrogen sulphide. Atmos Environ 34: 761-779.
    [11] Blake NJ, Streets DG, Woo JH, et al. (2004) Carbonyl sulphide and carbon disulfide: large-scale distributions over the western Pacific and emissions from Asia during TRACE-P. J Geophys Res 109: D15.
    [12] Ren YL (1999) Is carbonyl sulfide a precursor for carbon disulfide in vegetation and soil? Interconversion of carbonyl sulfide and carbon disulfide in fresh grain tissues in vitro. J Agric Food Chem 47: 2141-2144.
    [13] Steinbacher M, Bingemer HG, Schmidt U (2004) Measurements of the exchange of carbonyl sulfide (OCS) and carbon disulfide (CS2) between soil and atmosphere in a spruce forest in central Germany. Atmos Environ 38: 6043-6052. doi: 10.1016/j.atmosenv.2004.06.022
    [14] Stickel RE, Chin M, Daykin EP, et al. (1993). Mechanistic studies of the OH-initiated oxidation of CS2 in the presence of O2. J Phys Chem 97: 13653-13661. doi: 10.1021/j100153a038
    [15] Seinfeld JH, Pandis SN (2006) Atmospheric Chemistry and Physics: From Air Pollution to Climate Change, 2 Eds., John Wiley and Sons, Inc., New Jersey: 1-1203.
    [16] Crutzen PJ (1976) The possible importance of CSO for the sulfate layer of the stratosphere. Geophys Res Lett 3: 73-76. doi: 10.1029/GL003i002p00073
    [17] Hofmann DJ (1990) Increase in the stratospheric background sulfuric acid aerosol mass in the past 10 years. Science 248: 996-1000. doi: 10.1126/science.248.4958.996
    [18] Taubman SJ, Kasting JF (1995). Carbonyl sufide: No remedy for global warming. Geophys Res Lett 22: 803-805.
    [19] Chin M, Davis DD (1995) A reanalysis of carbonyl sulphide as a source of stratospheric background sulfur aerosol. J Geophys Res Atmos 100: 8993-9005. doi: 10.1029/95JD00275
    [20] Bruhl C, Lelieveld J, Tost H, et al. (2015). Stratospheric sulfur and its implications for radiative forcing simulated by the chemistry climate model EMAC. J Geophys Res-Atmos 120: 2103-2118.
    [21] Xu X, Bingemer HG, Schmidt U (2002) The flux of carbonyl sulphide and carbon disulfide between the atmosphere and a spruce forest. Atmos Chem Phys 2: 171-181.
    [22] Taylor Jr GE, McLaughlin Jr SB, Shriner DS, et al. (1983). The flux sulfur-containing gases to vegetation. Atmos Environ 17: 789-796.
    [23] De Bruyn WL, Swartz E, Hu JH, et al. (1995) Henry's law solubilities and setchenow coefficients for biogenic reduced sulfur species obtained from gas-liquid uptake measurements. J Geophys Res 100: 7245-7251. doi: 10.1029/95JD00217
    [24] Berglen TF, Berntsen TK, Isaksen SA, et al. (2004) A global model of the coupled sulphur/oxidant chemistry in the troposphere: The sulphur cycle. J Geophys Res 109: D19310. doi: 10.1029/2003JD003948
    [25] Kloster S, Feichter J, Maier-Reimer E, et al. (2006) DMS cycle in the marine ocean-atmosphere system- a global model study. Biogeosciences 3: 29-51. doi: 10.5194/bg-3-29-2006
    [26] Kettle AJ, Kuhn U, Von Hobe M, et al. (2002) Global budget of atmospheric carbonyl sulfide: Temporal and spatial variations of the dominant sources and sinks. J Geophys Res 107: D22.
    [27] Suntharalingam P, Kettle AJ, Monzka SM, et al. (2008) Global 3-D model analysis of the seasonal cycle of atmospheric carbonyl sulfide: Implications for terrestrial vegetation uptake. Geophys Res Lett 35: L19801. doi: 10.1029/2008GL034332
    [28] Berry J, Wolf A, Campbell JE, et al. (2013) A coupled model of the global cycles of carbonyl sulfide and CO2: A possible new window on the carbon cycle. J Geophys Res-Biogeo 118: 842-852. doi: 10.1002/jgrg.20068
    [29] Kuai L, Worden JR, Campbell JE, et al. (2015) Estimate of carbonyl sulfide tropical oceanic surface fluxes using Aura Tropospheric Emission Spectrometer observations. J Geophys Res-Atmos 120: 11012-11023.
    [30] Glatthor N, Höpfner M, Baker IT, et al. (2015) Tropical sources and sinks of carbonyl sulfide observed from space. Geophys Res Lett 42: 10082-10090. doi: 10.1002/2015GL066293
    [31] Kremser S, Thomason LW, von Hobe M, et al. (2016) Stratospheric aerosol-Observations, processes, and impact on climate. Rev Geophys 54: 278-335.
    [32] Lennartz ST, Marandino CA, von Hobe M, et al. (2017) Direct oceanic emissions unlikely to account for the missing source of atmospheric carbonyl sulfide. Atmos Chem Phys 17: 385-402.
    [33] Pham M, Müller J-F, Brasseur GP, et al. (1995) A three-dimensional study of the tropospheric sulfur cycle. J Geophys Res 100: 26061-26092.
    [34] Weisenstein DK, Yue GK, Ko MKW, et al. (1997) A two-dimensional model of sulfur species and aerosols. J Geophys Res 102: 13019-13035. doi: 10.1029/97JD00901
    [35] Kjellström E (1998) A three-dimensional global model study of carbonyl sulphide in the troposphere and the lower stratosphere. J Atmos Chem 29: 151-177. doi: 10.1023/A:1005976511096
    [36] Stevenson DS, Collins WJ, Johnson CE, et al. (1998) Intercomparison and evaluation of atmospheric transport in a Lagrangian model (STOCHEM), and an Eulerian model (UM), using 222Rn as a short-lived tracer. Quat J Royal Meteorol Soc 124: 2477-3492.
    [37] Stevenson DS, Dentener FJ, Schultz MG, et al. (2006) Multimodel ensemble simulations of present-day and near-future tropospheric ozone. J Geophys Res 111: D08301.
    [38] Collins WJ, Stevenson DS, Johnson CE, et al. (1997) Tropospheric ozone in a Global-Scale Three-Dimensional Lagrangian Model and its response to NOx emission controls. J Atmos Chem 26: 223-274.
    [39] Utembe SR, Cooke MC, Archibald AT, et al. (2010) Using a reduced Common Representative Intermediates (CRI v2-R5) mechanism to simulate tropospheric ozone in a 3-D Lagrangian chemistry transport model. Atmos Environ 13: 1609-1622.
    [40] Derwent RG, Collins WJ, Jenkin ME, et al. (2003) The global distribution of secondary particulate matter in a 3-D Lagrangian chemistry transport model. J Atmos Chem 44: 57-95. doi: 10.1023/A:1022139814102
    [41] Stevenson DS, Johnson CE, Highwood EJ, et al. (2003) Atmospheric impact of the 1783–1784 Laki eruption: Part I chemistry modelling. Atmos Chem Phys 3: 487-507.
    [42] Derwent RG, Stevenson DS, Doherty RM, et al. (2008) How is surface ozone in Europe linked to Asian and North American NOx emissions? Atmos Environ 42: 7412-7422.
    [43] Jenkin ME, Watson LA, Utembe SR, et al. (2008) A Common Representative Intermediate (CRI) mechanism for VOC degradation. Part-1: gas phase mechanism development. Atmos Environ 42: 7185-7195.
    [44] Watson LA, Shallcross DE, Utembe SR, et al. (2008) A Common Representative Intermediate (CRI) mechanism for VOC degradation. Part 2: gas phase mechanism reduction. Atmos Environ 42: 7196-7204.
    [45] Utembe SR, Watson LA, Shallcross DE, et al. (2009) A Common Representative Intermediates (CRI) mechanism for VOC degradation. Part 3: Development of a secondary organic aerosol module. Atmos Environ 43: 1982-1990.
    [46] Utembe SR, Cooke MC, Archibald AT, et al. (2011) Simulating secondary organic aerosol in a 3-D Lagrangian chemistry transport model using the reduced Common Representative Intermediates mechansim (CRI v2-R5). Atmos Environ 45: 1604-1614. doi: 10.1016/j.atmosenv.2010.11.046
    [47] Collins WJ, Stevenson DS, Johnson CE, et al. (2000) The European regional ozone distribution and its links with the global scale for the years 1992 and 2015. Atmos Environ 34: 255-267. doi: 10.1016/S1352-2310(99)00226-5
    [48] Olivier JG, Bouwman AF, Berdowski JJ, et al. (1996) Description of EDGAR Version 2.0: A set of global emission inventories of greenhouse gases and ozone-depleting substances for all anthropogenic and most natural sources on a per country basis and on 1 degree ´ 1 degree grid. Technical report, Netherlands Environmental Assessment Agency.
    [49] Olivier JGJ, Berdowski JJM (2001) Global emissions sources and sinks. Berdowski JJM, Guicherit R, Heij BJ (Eds.). The Climate System, Swets and Zeitlinger Publishers, Lisse, Netherlands.
    [50] Granier C, Lamarque JF, Mieville A, et al. (2005) POET, a database of surface emissions of ozone precursors. Available from: http://accent.aero.jussieu.fr/database_table_inventories.php.
    [51] Atkinson R, Baulch DL, Cox RA, et al. (2004) Evaluated kinetic and photochemical data for atmospheric chemistry: Volume I-gas phase reactions of Ox, HOx, NOx and SOx species. Atmos Chem Phys 4: 1461-1738. doi: 10.5194/acp-4-1461-2004
    [52] Khan MAH, Cooke MC, Utembe SR, et al. (2015) A study of global atmospheric budget and distribution of acetone using global atmospheric model STOCHEM-CRI. Atmos Environ 112: 269-277.
    [53] Sander SP, Friedl RR, Golden DM, et al. (2006) Chemical kinetics and photochemical data for use in atmospheric studies. Evaluation number 15, JPL publication 06-2, Jet Propulsion Laboratory, Pasadena, CA.
    [54] Lee CL, Brimblecombe P (2016) Anthropogenic contributions to global carbonyl sulfide, carbon disulfide and organosulfides fluxes. Earth-Sci Rev 160: 1-18. doi: 10.1016/j.earscirev.2016.06.005
    [55] Majozi T, Veldhuizen P (2015) The chemicals industry in South Africa. American Inst. Chem. Eng (AlChE) July: 46-51. Available from: https://www.aiche.org/sites/default/files/cep/20150746.pdf
    [56] Kim KH, Swan H, Shon ZH, et al. (2004) Monitoring of reduced sulfur compounds in the atmosphere of Gosan, Jeju Island during the Spring of 2001. Chemosphere 54: 515-526. doi: 10.1016/j.chemosphere.2003.07.003
    [57] Guo H, Simpson IJ, Ding AJ, et al. (2010) Carbonyl sulphide, dimethyl sulphide and carbon disulfide in the Pearl River Delta of southern China: Impact of anthropogenic and biogenic sources. Atmos Environ 44: 3805-3813.
    [58] Pal R, Kim KH, Jeon EC, et al. (2009) Reduced sulfur compounds in ambient air surrounding an industrial region in Korea. Environ Monit Assess 148: 109-125. doi: 10.1007/s10661-007-0143-z
    [59] Yujing M, Hai W, Zhang X, et al. (2002) Impact of anthropogenic sources on carbonyl sulphide in Beijing City. J Geophys Res 107: D24.
    [60] Maroulis PJ, Bandy AR (1980) Measurements of atmospheric concentrations of CS2 in the eastern United States. Geophys Res Letts 7: 681-684. doi: 10.1029/GL007i009p00681
    [61] Jones BMR, Cox RA, Penkett SA (1983) Atmospheric chemistry of carbon disulphide. J Atmos Chem 1: 65-86. doi: 10.1007/BF00113980
    [62] Thornton DC, Bandy AR (1993). Sulfur dioxide and dimethyl sulfide in the central pacific troposphere. J Atmos Chem 17: 1-13. doi: 10.1007/BF00699110
    [63] Inomata Y, Hayashi M, Osada K, et al. (2006) Spatial distributions of volatile sulfur compounds in surface seawater and overlying atmosphere in the northwestern Pacific Ocean, eastern Indian Ocean, and Southern Ocean. Global Biogeochem Cycles 20: GB2022.
    [64] Cooper DJ, Saltzman ES (1993) Measurements of atmospheric dimethylsulphide, hydrogen sulphide, and carbon disulfide during GTE/CITE 3. J Geophys Res 98: 23397-23409. doi: 10.1029/92JD00218
    [65] Inomata Y, Iwasaka Y, Osada K, et al. (2006) Vertical distributions of particles and sulfur gases (volatile sulfur compounds and SO2) over East Asia: comparison with two aircraft-borne measurements under the Asian continental outflow in spring and winter. Atmos Environ 40: 430-444.
    [66] Bandy AR, Thornton DC, Johnson JE (1993) Carbon disulfide measurements in the atmosphere of the western north Atlantic and the northwestern south Atlantic oceans. J Geophys Res-Atmos 98: 23449-23457. doi: 10.1029/93JD02411
  • This article has been cited by:

    1. Jun Li, Dawei Chen, Minhong Shen, Tumor Microenvironment Shapes Colorectal Cancer Progression, Metastasis, and Treatment Responses, 2022, 9, 2296-858X, 10.3389/fmed.2022.869010
    2. Jie Wang, Rehana Akter, Md. Fahim Shahriar, Md. Nazim Uddin, Cancer-Associated Stromal Fibroblast-Derived Transcriptomes Predict Poor Clinical Outcomes and Immunosuppression in Colon Cancer, 2022, 28, 1532-2807, 10.3389/pore.2022.1610350
    3. Zhaoxiang Wang, Yue Xia, Yi Pan, Li Zhang, Fengyan Tang, Xiawen Yu, Zhongming Yang, Dong Wang, Ling Yang, Jue Jia, Guoyue Yuan, Weighted Gene Co-Expression Network Analysis of Immune Infiltration in Nonalcoholic Fatty Liver Disease, 2023, 23, 18715303, 10.2174/1871530323666221208105720
    4. Xianhua Zhuo, Cheng Huang, Liangping Su, Faya Liang, Wenqian Xie, Qiuping Xu, Ping Han, Xiaoming Huang, Ping-Pui Wong, Identification of a distinct tumor endothelial cell-related gene expression signature associated with patient prognosis and immunotherapy response in multiple cancers, 2023, 149, 0171-5216, 9635, 10.1007/s00432-023-04848-2
    5. Yuxiong Wang, Yishu Wang, Bin Liu, Xin Gao, Yunkuo Li, Faping Li, Honglan Zhou, Mapping the tumor microenvironment in clear cell renal carcinoma by single-cell transcriptome analysis, 2023, 14, 1664-8021, 10.3389/fgene.2023.1207233
    6. Zhiqiang Huang, Lu Huang, Lili Li, Chunming Xiang, Xin Xiong, Yongxiu Lu, Identification of Key Genes Involved in Carcinogenesis and Progression of Colon Cancer Based on Bioinformatics, 2023, 19, 1550-7033, 1279, 10.1166/jbn.2023.3640
  • Reader Comments
  • © 2017 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(6539) PDF downloads(1135) Cited by(15)

Figures and Tables

Figures(6)  /  Tables(3)

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog