Sage Journals: Discover world-class research

Abstract

Surgery is the most effective treatment for breast cancer patients. However, some patients developed recurrence and distant metastasis after surgery. Adjuvant therapy is considered for high-risk patients depending on several prognostic markers, and lymphovascular invasion has become one of such prognostic markers that help physicians to identify the risk for distant metastasis and recurrence. However, the mechanism of lymphovascular invasion in breast cancer remains unknown. This study aims to unveil the genes and pathways that may involve in lymphovascular invasion in breast cancer. In total, 108 breast cancer samples were collected during surgery and microarray analysis was performed. Significance analysis of the microarrays and limma package for R were used to examine differentially expressed genes between lymphovascular invasion–positive and lymphovascular invasion–negative cases. Network and pathway analyses were mapped using the Ingenuity Pathway Analysis and the Database for Annotation, Visualization and Integrated Discovery. In total, 86 differentially expressed genes, including 37 downregulated genes and 49 upregulated genes were identified in lymphovascular invasion–positive patients. Among these genes, TNFSF11, IL6ST, and EPAS1 play important roles in cytokine–receptor interaction, which is the most enriched pathway related to lymphovascular invasion. Moreover, the results also suggested that an imbalance between extracellular matrix components and tumor micro-environment could induce lymphovascular invasion. Our study evaluated the underlying mechanisms of lymphovascular invasion, which may further help to assess the risk of breast cancer progression and identify potential targets of adjuvant treatment.

Keywords

Breast cancer microarray analysis bioinformatics analysis lymphovascular invasion cytokine interaction pathway

Introduction

Breast cancer is the second most common cancer with estimated 1.67 million incident cases during the year 2012 worldwide. Surgery is an effective treatment for early breast cancers.¹ However, some breast cancer patients developed recurrence or distant metastasis after surgery.^2,3 Adjuvant therapy for breast cancer is determined according to prognostic factors and the risk of recurrence or distant metastasis. Well-established prognostic factors include age, histological grade, tumor size, axillary nodal status, and the presence of hormone receptor.^4–6 Previous studies indicated that patients with small tumors had favorable overall survival than patients with larger tumors.⁷ Furthermore, tumor size was correlated with the time to develop metastasis.⁸ The presence of estrogen receptor (ER) and/or progesterone receptor (PR) is considered for receiving adjuvant hormone manipulation therapy such as tamoxifen to improve survival.^9,10 Human epidermal growth factor receptor 2 (HER2) is also considered a prognostic as well as a predictive marker for trastuzumab responsiveness.⁴

Regional lymph node involvement means that a cancer already has the ability to develop distant metastasis, and it is one of the indications for adjuvant treatment.^11,12 Lymphovascular invasion (LVI) is an important and early step of tumor metastasis and is defined as the presence of carcinoma cells in lymphatic and blood vessels. A previous study showed that patients with LVI had higher mortality rates than patients without LVI.¹³ Several studies indicated that LVI is associated with the risk of axillary lymph node involvement, distant metastases, and recurrence.^{11,12,14–17} However, the underlying mechanism of LVI in breast cancer is unknown. In this study, we attempted to identify crucial molecules and pathways associated with the presence of LVI in breast cancer.

Materials and methods

Breast cancer samples and microarray analysis

Breast cancer tissues were collected and processed as previously described Huang,18 and was reviewed by and approved by institutional review board of Cathay General Hospital. We obtained informed consent from all patients after explanation of the study. Breast cancer samples from pretreated invasive breast cancer patients (<70 years old) were collected during surgery and processed with RNAlater reagent (Qiagen, Germantown, MD) and stored at −80°C liquid nitrogen. Samples with <10% non-malignant constituents were selected for microarray experiment. Immunohistochemical (IHC) test was performed to determine ER positivity status. HER-2 overexpression (IHC 2+/3+) were determined according to ASCO and CAP guidelines by fluorescence in situ hybridization. Grading was performed based on modified Bloom-Richardson (Nottingham) system. In addition, LVI information was obtained under microscopic inspections. After performing microarray experiment using Affymetrix GeneChip® Human Genome U133 plus 2.0 (Affymetrix, Santa Clara, CA), the data was normalized using robust multichip average (RMA) method Irizarry.¹⁹

Data processing and gene expression calculations

All breast cancer patients were divided into two groups according to clinical LVI-positive (LVI+) and LVI-negative (LVI−) phenotypes. Differentially expressed genes (DEGs) were identified using the significance analysis of the microarrays (SAM) method²⁰ and the limma package in R (version. 2.15.1).²¹ For SAM, statistical comparisons were conducted using a two-sided Student’s t test (p < 0.05) and comparisons of multiples of change (>1.5-fold). The percentage of these significant genes identified by chance was referred to as the false discovery rate (FDR). The FDR was set to 0.05, and a q value was provided as the score of the FDR for each significant gene. For the limma analysis, after normalization and quality control of a gene set, we performed a DEG analysis and selected genes with p < 0.05. Then, common genes between these two statistical methods were selected for further study.

Network construction and pathway analysis

The Ingenuity Pathway Analysis (IPA; Ingenuity^® Systems, Mountain View, CA; available at https://www.ingenuity.com) is a web-based software application for integrating, analyzing, and understanding data derived from gene expression measurement. The gene network analysis is used to cluster genes based on their molecular functions and revealed their correlations. We constructed gene networks and relevant metabolic pathways underlying LVI by mapping our genes of interest to the database and subjecting them to a core analysis. In addition, we also consulted Database for Annotation, Visualization and Integrated Discovery (DAVID) bioinformatics tool to find common pathways with the IPA and map the corresponding genes accordingly.

Gene set enrichment analysis and gene ontology

To confirm our finding, gene set enrichment analysis (GSEA) was used to analyze three databases including Kyoto Encyclopedia of Genes and Genomes (KEGG) database, Gene Ontology Biological Processes (GO-BP) database, and Reactome database. This method focuses on groups of genes that share common biological function, chromosomal location, or regulation.²² We used R clusterProfiler package²³ for analysis. In addition, GO terms were performed to classify their biological processes, molecular functions, or cellular components. We used the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING)²⁴ database to access the GO. We mapped all DEGs and selected GO terms which met statistical significance at p < 0.05 level.

Results

Clinical characteristics of the 108 breast cancer patients

Clinical data of all 108 breast cancer patients are shown in Table 1. Among these cases, most patients showed ER positivity (n = 70, 64.8%) and most of them were diagnosed with pathological stage II (n = 46, 42.6%). At diagnosis, 79 patients (73.1%) were presented with LVI, whereas 29 patients (26.9%) were diagnosed without the presence of LVI and 1 patient (0.9%) had no record for LVI status. Gene expression data were deposited in National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO; submission to GEO in progress).

Table 1.

Clinical characteristics of 108 breast cancer patients.

Clinical variable	Categories	Number (%)^a		Fisher’s p
Clinical variable	Categories	LVI+ (n = 79)	LVI− (n = 29)	Fisher’s p
Events^b	Yes	12 (15.2)	2 (6.9)	0.344
Events^b	No	67 (84.8)	27 (93.1)	0.344
Clinical stage	0	0 (0.0)	6 (20.7)	2.34 × 10⁻⁸
	I	10 (12.7)	14 (48.3)
	II	38 (48.1)	8 (27.6)
	III	28 (35.4)	1 (3.4)
	IV	3 (3.8)	0 (0.0)
Grade	I	2 (2.5)	4 (13.8)	0.069
	II	38 (48.1)	14 (48.3)
	III	39 (49.4)	11 (37.9)
Status	Death	71 (89.9)	27 (93.1)	1.000
Status	Alive	8 (10.1)	2 (6.9)	1.000
ER	Yes	51 (64.6)	19 (65.5)	1.000
ER	No	28 (35.4)	10 (34.5)	1.000
HER2	Yes	34 (43.0)	19 (65.5)	0.510
HER2	No	45 (57.0)	10 (34.5)	0.510
PNI	Yes	24 (30.4)	4 (13.8)	0.135
PNI	No	55 (69.6)	23 (79.3)	0.135

LVI: lymphovascular invasion; ER: estrogen receptor; HER2: human epidermal growth factor receptor 2; PNI: perineural invasion.

The percentages were calculated according to the total number of LVI+ and LVI−.

Events included recurrence, metastasis, and death.

Genes involved in LVI

We used SAM and the limma package to identify DEGs between LVI+ and LVI− samples. The intersection of both statistical methods identified 86 common genes: 37 were downregulated and 49 were upregulated (Supplement Figure 1 and Supplement Table 1).

Protein–protein interaction networks

IPA was used to construct the interaction network of DEGs. We selected human species, breast cancer cell lines, and direct interactions to identify interaction networks. After a core analysis, we found that six networks are involved with filtered DEGs (Supplement Table 3). The connective tissue disorder network was the most significant network related to LVI (Figure 1). In addition, we produced an overlapping network and found that the ubiquitin C (UBC) gene had the highest degree of connectivity to other genes in the network including gene developmental disorders (Figure 2) and nucleic acid metabolism networks (Figure 3).

Figure 1.

Network of connective tissue disorder–related genes (red: upregulated; green: downregulated).

Figure 2.

Network of developmental disorder–related genes.

Figure 3.

Network of nucleic acid metabolism network–related genes.

Pathways associated with LVI

After the IPA core analysis, we found that five significant pathways are related to LVI (Figure 4). The oncostatin M (OSM) signaling pathway was the most enriched pathway associated with LVI. This pathway functions as a part of the cytokine–receptor complex. Interleukin-6 signal transducer (IL6ST) and endothelial PAS domain protein 1 (EPAS1) are two genes that belong to this pathway (Supplement Table 4). Furthermore, we also mapped genes using the KEGG¹⁸ and found similar results of cytokine–receptor interactions being the most enriched pathway associated with LVI (Table 2 and Figure 5). To support our discovery, we further analyzed using GSEA and found that cytokine system was indicated to be the significant biological process associated with LVI in three different databases (Figure 6). In addition to the effect of cytokine–receptor interaction on LVI, we noted that collagen- and lipid-related pathways are also relevant to LVI (Table 2 and Supplement Table 3).

Figure 4.

Lymphovascular invasion–related pathways.

Table 2.

Lymphovascular invasion–related pathways using the KEGG analysis.

Pathway	p value	n (genes)
Cytokine–receptor interactions	5.53 × 10⁻³	2 (IL6ST, TNFSF11)
ECM–receptor interactions	9.63 × 10⁻³	3 (COL4A1, COL11A1, LAMA3)
Ether lipid metabolism	1.23 × 10⁻²	2 (PPAPDC1B, PLA2G7)
Spliceosomes	1.54 × 10⁻²	3 (SNRPE, TRA2A, RBM)

KEGG: Kyoto Encyclopedia of Genes and Genomes; IL6ST: interleukin-6 signal transducer; TNFSF11: tumor necrosis factor (ligand) superfamily, member 11; ECM: extracellular matrix; COL4A1: collagen alpha-1(IV) chain; COL11A1: collagen alpha-1(XI) chain; LAMA3: laminin subunit alpha 3; PPAPDC1B: phosphatidic acid phosphatase type 2 domain containing 1B; PLA2G7: phospholipase A2 group VII; SNRPE: small nuclear ribonucleoprotein E; TRA2A: transformer 2 alpha homolog; RBM: RNA-binding motif.

Figure 5.

Cytokine–cytokine receptor interaction pathway using the KEGG analysis.

Figure 6.

GSEA-identified cytokine system was the significance-enriched pathway associated with LVI.

Subgroup analysis of TNFSF11, IL6ST, and EPAS1

We further assessed the difference in expression levels of tumor necrosis factor (ligand) superfamily, member 11 (TNFSF11), IL6ST, and EPAS1 between LVI+ and LVI− samples in different subgroups. Based on ER status, HER2 status, lymph node metastasis and perineural invasion, permutation analysis was conducted (Supplement Table 2). We found that TNFSF11 and EPAS1 showed a significant correlation with LVI in ER-negative and HER2-positive breast cancer patients. IL6ST and EPAS1 showed a significant correlation with LVI in samples without lymph node metastasis and perineural invasion. These results thus suggest the potential roles of TNFSF11, IL6ST, and EPAS1 as subtype-specific LVI markers.

GO analysis

GO analysis provides information about the functions of the DEGs by classifying them by their biological processes, molecular functions, and cellular components of DEGs. With STRING analysis, we found that biological processes of hemidesmosome assembly, molecular function of receptor binding, and the cellular component of the extracellular region were the most significant biological processes associated with LVI. In addition, development of blood vessels and structural constituents of the cytoskeleton may underpin the phenomenon of LVI (Table 3).

Table 3.

GO enrichment in LVI-positive patients.

GO	p value	Number of molecules
GO cellular processes
Hemidesmosome assembly	9.29 × 10⁻⁶	3
Blood vessel development	2.40 × 10⁻⁵	9
Glomerulus vasculature development	2.53 × 10⁻⁵	3
Vasculature development	3.10 × 10⁻⁵	9
Renal system development	3.56 × 10⁻⁵	7
GO molecular function
Receptor binding	2.96 × 10⁻⁴	14
Structural constituent of cytoskeleton	6.93 × 10⁻⁴	4
Intermediate filament binding	9.37 × 10⁻⁴	2
Histone acetyltransferase binding	3.16 × 10⁻³	2
Cell adhesion molecule binding	4.76 × 10⁻³	4
GO cellular component
Extracellular region	9.26 × 10⁻²	29

GO: gene ontology; LVI: lymphovascular invasion.

Discussion

The aim of this study was to identify genes and pathways associated with LVI in breast cancer. Our study performed a gene expression analysis of Taiwanese breast cancers database and found that the OSM signaling pathway was the most enriched pathway related to LVI in breast cancer. OSM belongs to the interleukin 6 (IL-6) cytokine family and regulates the acute response to infection and injury.²⁵ Previous studies indicated that OSM plays many crucial roles in cell growth,^26–29 bone metabolism,^30,31 cholesterol metabolism,³² and angiogenesis.³³

OSM has been evaluated in several human malignancies. Lilja et al.³⁴ found that OSM had higher expression in brain tumors than in normal brain tissues. Another study by Fossey et al.³⁵ showed that OSM induced signal transducer and activator of transcription 3 (STAT3) activation resulting in stimulation of matrix metalloproteinase 2 (MMP-2) and vascular endothelial growth factor (VEGF), finally enhancing invasive behavior and tumor angiogenesis in osteosarcoma. In addition, OSM stimulated mammary tumor metastasis to bone and activated osteoclast differentiation.³⁶

From IPA analysis, two genes were related to this pathway: IL6ST and EPAS1 or hypoxia-inducible factor 1 (HIF-1). The IL6ST (gp130) functions as a part of the cytokine–receptor complex associated with the IL-6 receptor (gp80).³⁷ A previous study showed that IL6ST is expressed by human endothelial progenitor cells (EPCs) and is activated by the IL-6 cytokine. The results suggested that IL-6 induced EPC migration, proliferation, and matrigel tube formation via gp80/gp130 activation and STAT3 phosphorylation signaling pathway.³⁸ Moreover, Yao et al.³⁹ showed that IL-6 stimulated VEGF-induced angiogenic activity and extracellular signal–regulated protein kinases (ERK1/2) phosphorylation in human cerebral endothelial cells.

EPAS1 showed some identical sequences to HIF-1α, which respond to hypoxia and regulate vascularization.⁴⁰ Ke and Costa⁴¹ found that EPAS1 translocates to nuclei and activates genes involved in angiogenesis. Bárdos and Ashcroft⁴² also indicated that EPAS1 stimulates the production of VEGF, which is a regulator of tumor angiogenesis, invasion, and metastasis. Moreover, KEGG analysis from our study showed similar results of cytokine–receptor interactions including IL6ST and TNFSF11 being the most enriched pathways associated with LVI.

TNFSF11 is also known as receptor activator of nuclear factor kappa-B ligand (RANKL). RANKL enhanced mammary cell invasion and migration and bone metastasis.⁴³ In addition, Sfiridaki et al.⁴⁴ demonstrated that serum RANKL level was elevated in multiple myeloma and was positively correlated with the potent angiogenic cytokines of hepatocyte growth factor (HGF), VEGF, and IL-6.

Angiogenesis is strongly correlated with tumor progression, invasion, and metastasis.^45–47 We surmised that expression alterations of genes in the cytokine–receptor pathway could result in angiogenesis and eventually presented as LVI pathologically.

In addition to cytokine–receptor interactions, we also found that extracellular matrix (ECM) components such as collagens were associated with LVI. Collagens are ECM proteins that play important roles in cell adhesion⁴⁸ and stabilization of hemidesmosome assembly.⁴⁹ Our study suggested that collagen alpha-1(IV) chain (COL4A1) and collagen alpha-1(XI) chain (COL11A1) were associated with the ECM–receptor pathway. In contrast, previous studies showed that some ECM proteases such as MMP-2 were involved in the progression and lymph node metastasis of gastric cancer,⁵⁰ and MMP-9 was involved in the process of neovascularization and angiogenesis.^51,52 Our results implied that an imbalance between ECM components and the tumor micro-environment could induce LVI.

From interaction analysis, three networks including connective tissue disorders, developmental disorders, and nucleic acid metabolism were revealed given the filtered genes. UBC was a major protein common to all three networks and had the highest connectivity to others. Although in our dataset, UBC gene did not show the differential expression. We assume that the protein level changes may cause these interactions. Ubiquitin is a small protein covalently attached to the target⁵³ and generates many thousands of ubiquitinated products.⁵⁴ Ubiquitin also plays multiple important roles in cellular processes relevant to cancer.^55,56 Ubiquitination can result in several cellular processes including protein degradation⁵⁷ and control protein–protein interaction (PPI).⁵⁸ Our results suggested that the ubiquitin system interacted with all the networks and may contribute to LVI.

In conclusion, our results with comprehensive bioinformatics approaches including GSEA, GO, PPI, and pathway analyses indicated that the underlying processes of LVI resulted from dysregulations of these three networks. Cytokine–receptor and ECM–receptor interactions are the two major pathways associated with LVI. This study also identified some candidate genes related with LVI which can be potential targets for breast cancer treatment.

Footnotes

Acknowledgements

S.K. and H.S.-C.W. contributed equally to this study.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The work was supported by grants from Cathay Medical Research Institute CGH-MR-A-10312; CGH-MR-10208, from the Excellence for Cancer Research Center Grant, funded by the Ministry of Health and Welfare (MOHW105-TDU-B-212-134007) and Cathay General Hospital, Taipei Medical University (No. 105CGH-TMU-05).

Supplementary materials

Supplementary materials can be found at .

References

Albrand

Terret

. Early breast cancer in the elderly: assessment and management considerations. Drugs Aging 2008; 25: 35–45.

Touboul

Buffat

Belkacemi

. Local recurrences and distant metastases after breast-conserving surgery and radiation therapy for early breast cancer. Int J Radiat Oncol Biol Phys 1999; 43: 25–38.

Langstein

Cheng

Singletary

. Breast cancer recurrence after immediate reconstruction: patterns and significance. Plast Reconstr Surg 2003; 111: 712–720; discussion 721–712.

Schnitt

. Traditional and newer pathologic factors. J Natl Cancer Inst Monogr 2001; 30: 22–26.

Lal

Tan

Chen

. Correlation of HER-2 status with estrogen and progesterone receptors and histologic features in 3,655 invasive breast carcinomas. Am J Clin Pathol 2005; 123: 541–546.

Cianfrocca

Goldstein

. Prognostic and predictive factors in early-stage breast cancer. Oncologist 2004; 9: 606–616.

Carter

Allen

Henson

. Relation of tumor size, lymph node status, and survival in 24,740 breast cancer cases. Cancer 1989; 63: 181–187.

Koscielny

Tubiana

. Breast cancer: relationship between the size of the primary tumour and the probability of metastatic dissemination. Br J Cancer 1984; 49: 709–715.

Bardou

Arpino

Elledge

. Progesterone receptor status significantly improves outcome prediction over estrogen receptor status alone for adjuvant endocrine therapy in two large breast cancer databases. J Clin Oncol 2003; 21: 1973–1979.

10.

Crowe

Jr Gordon

Hubay

. Estrogen receptor determination and long term survival of patients with carcinoma of the breast. Surg Gynecol Obstet 1991; 173: 273–278.

11.

Davis

Gelber

Goldhirsch

. Prognostic significance of peritumoral vessel invasion in clinical trials of adjuvant therapy for breast cancer with axillary lymph node metastasis. Hum Pathol 1985; 16: 1212–1218.

12.

Woo

Silberman

Nakamura

. Lymph node status combined with lymphovascular invasion creates a more powerful tool for predicting outcome in patients with invasive breast cancer. Am J Surg 2002; 184: 337–340.

13.

Lee

Pinder

Macmillan

. Prognostic value of lymphovascular invasion in women with lymph node negative invasive breast carcinoma. Eur J Cancer 2006; 42: 357–362.

14.

Yip

Taib

Tan

. Predictors of axillary lymph node metastases in breast cancer: is there a role for minimal axillary surgery? World J Surg 2009; 33: 54–57.

15.

Song

Shin

Cho

. The role of lymphovascular invasion as a prognostic factor in patients with lymph node-positive operable invasive breast cancer. J Breast Cancer 2011; 14: 198–203.

16.

Hahn

Lee

Kim

. Lymphovascular space invasion is highly associated with lymph node metastasis and recurrence in endometrial cancer. Aust N Z J Obstet Gynaecol 2013; 53: 293–297.

17.

Wong

O’Neill

Recht

. The relationship between lymphatic vessell invasion, tumor size, and pathologic nodal status: can we predict who can avoid a third field in the absence of axillary dissection? Int J Radiat Oncol Biol Phys 2000; 48: 133–137.

18.

Huang

Lien

. Concurrent gene signatures for Han Chinese breast cancers. PLoS One 2013; 8: e76421.

19.

Irizarry

Hobbs

Collin

. Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 2003; 4: 249–264.

20.

Tusher

Tibshirani

Chu

. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci U S A 2001; 98: 5116–5121.

21.

Gentleman

Carey

Bates

. Bioconductor: open software development for computational biology and bioinformatics. Genome Biol 2004; 5: R80.

22.

Subramanian

Tamayo

Mootha

. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 2005; 102: 15545–15550.

23.

Wang

Han

. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 2012; 16: 284–287.

24.

von Mering

Huynen

Jaeggi

. STRING: a database of predicted functional associations between proteins. Nucleic Acids Res 2003; 31: 258–261.

25.

Heinrich

Behrmann

Haan

. Principles of interleukin (IL)-6-type cytokine signalling and its regulation. Biochemical J 2003; 374: 1–20.

26.

Douglas

Goss

Sutherland

. Expression and function of members of the cytokine receptor superfamily on breast cancer cells. Oncogene 1997; 14: 661–669.

27.

Liu

Spence

Wallace

. Oncostatin M-specific receptor mediates inhibition of breast cancer cell growth and down-regulation of the c-myc proto-oncogene. Cell Growth Differ 1997; 8: 667–676.

28.

Nishimoto

Ogata

Shima

. Oncostatin M, leukemia inhibitory factor, and interleukin 6 induce the proliferation of human plasmacytoma cells via the common signal transducer, gp130. J Exp Med 1994; 179: 1343–1347.

29.

Nakamura

Nonaka

Saito

. Hepatocyte proliferation and tissue remodeling is impaired after liver injury in oncostatin M receptor knockout mice. Hepatology 2004; 39: 635–644.

30.

Walker

McGregor

Poulton

. Oncostatin M promotes bone formation independently of resorption when signaling through leukemia inhibitory factor receptor in mice. J Clin Invest 2010; 120: 582–592.

31.

Heymann

Rousselle

. gp130 Cytokine family and bone cells. Cytokine 2000; 12: 1455–1468.

32.

Grove

Mazzucco

Radka

. Oncostatin M up-regulates low density lipoprotein receptors in HepG2 cells by a novel mechanism. J Biol Chem 1991; 266: 18194–18199.

33.

Vasse

Pourtau

Trochon

. Oncostatin M induces angiogenesis in vitro and in vivo. Arterioscler Thromb Vasc Biol 1999; 19: 1835–1842.

34.

Lilja

Nordborg

Brun

. Expression of the IL-6 family cytokines in human brain tumors. Int J Oncol 2001; 19: 495–499.

35.

Fossey

Bear

Kisseberth

. Oncostatin M promotes STAT3 activation, VEGF production, and invasion in osteosarcoma cell lines. BMC Cancer 2011; 11: 125.

36.

Racz

Teri

Javor

. [Transmucosal potential difference in gastric ulcer patients]. Orvosi Hetilap 1986; 127: 2059–2062.

37.

Hibi

Murakami

Saito

. Molecular cloning and expression of an IL-6 signal transducer, gp130. Cell 1990; 63: 1149–1157.

38.

Fan

Shen

. Interleukin-6 stimulates circulating blood-derived endothelial progenitor cell angiogenesis in vitro. J Cereb Blood Flow Metab 2008; 28: 90–98.

39.

Yao

Zhai

Young

. Interleukin-6 triggers human cerebral endothelial cells proliferation and migration: the role for KDR and MMP-9. Biochem Biophys Res Commun 2006; 342: 1396–1404.

40.

Tian

McKnight

Russell

. Endothelial PAS domain protein 1 (EPAS1), a transcription factor selectively expressed in endothelial cells. Genes Dev 1997; 11: 72–82.

41.

Costa

. Hypoxia-inducible factor-1 (HIF-1). Mol Pharmacol 2006; 70: 1469–1480.

42.

Bardos

Ashcroft

. Hypoxia-inducible factor-1 and oncogenic signalling. Bioessays 2004; 26: 262–269.

43.

Blake

Tometsko

Miller

. RANK expression on breast cancer cells promotes skeletal metastasis. Clin Exp Metastasis 2014; 31: 233–245.

44.

Sfiridaki

Pappa

Tsirakis

. Angiogenesis-related cytokines, RANKL, and osteoprotegerin in multiple myeloma patients in relation to clinical features and response to treatment. Mediators Inflamm 2011; 2011: 867576.

45.

Longatto

Filho A

Lopes

Schmitt

. Angiogenesis and breast cancer. J Oncol 2010; 2010: 576384.

46.

Weidner

Semple

Welch

. Tumor angiogenesis and metastasis–correlation in invasive breast carcinoma. N Engl J Med 1991; 324: 1–8.

47.

Folkman

. What is the evidence that tumors are angiogenesis dependent? J Natl Cancer Inst 1990; 82: 4–6.

48.

Heino

. The collagen family members as cell adhesion proteins. Bioessays 2007; 29: 1001–1010.

49.

Nishie

Kiritsi

Nystrom

. Dynamic interactions of epidermal collagen XVII with the extracellular matrix: laminin 332 as a major binding partner. Am J Pathol 2011; 179: 829–837.

50.

Monig

Baldus

Hennecken

. Expression of MMP-2 is associated with progression and lymph node metastasis of gastric carcinoma. Histopathology 2001; 39: 597–602.

51.

Forsyth

Wong

Laing

. Gelatinase-A (MMP-2), gelatinase-B (MMP-9) and membrane type matrix metalloproteinase-1 (MT1-MMP) are involved in different aspects of the pathophysiology of malignant gliomas. Br J Cancer 1999; 79: 1828–1835.

52.

Shipley

Bergers

. MMP-9/gelatinase B is a key regulator of growth plate angiogenesis and apoptosis of hypertrophic chondrocytes. Cell 1998; 93: 411–422.

53.

Komander

Rape

. The ubiquitin code. Annu Rev Biochem 2012; 81: 203–229.

54.

Weissman

. Themes and variations on ubiquitylation. Nat Rev Mol Cell Biol 2001; 2: 169–178.

55.

Shen

Schmitt

Buac

. Targeting the ubiquitin-proteasome system for cancer therapy. Expert Opin Ther Targets 2013; 17: 1091–1108.

56.

Park

Lee

. Downregulation of ubiquitin level via knockdown of polyubiquitin gene Ubb as potential cancer therapeutic intervention. Sci Rep 2013; 3: 2623.

57.

Glickman

Ciechanover

. The ubiquitin-proteasome proteolytic pathway: destruction for the sake of construction. Physiol Rev 2002; 82: 373–428.

58.

Schnell

Hicke

. Non-traditional functions of ubiquitin and ubiquitin-binding proteins. J Biol Chem 2003; 278: 35857–35860.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.42 MB

Identification of genes and pathways related to lymphovascular invasion in breast cancer patients: A bioinformatics analysis of gene expression profiles