INTRODUCTION
P terygium is a general ocular surface and degenerative disease characterized by conjunctival fibrovascular proliferation and invasion of the peripheral cornea. Pterygium is composed of the head that invades the cornea, the neck that includes the superficial limbus and the body that overlie the sclera[1]. Diverse biological pathways including increase of mitotic proteins and protein implicated in tissue invasion were remarkably affected in pterygium. Additionally,fibrosis, inflammation, angiogenesis, collagen deposition and extracellular matrix breakdown are associated with pterygium[2]. Pterygium is noted to be more prevalent than 20% of some populations and typically affects the younger population and can potentially lead to blindness[3].
Different processes have been reported as part of the pathology of pterygium, including cytokines, growth factors, viral infections, genetic predisposition, anti-apoptotic mechanisms,immunological mechanisms and extracellular matrix remodeling[4]. Up to now, the major treatment for pterygium is surgical removal. However, it recurs aggressively after surgical excision, and further treatment modalities are wide excision,adjunctive radiotherapy and antimitotic chemotherapy[5].Therefore, understanding the underlying molecular mechanism in pterygium is vital to design a strategy for the non-surgical therapy of pterygium and the prevention of disease recurrence after surgery[6].
Here, we obtained differentially expressed genes (DEGs)in pterygium. The functional annotation was performed to uncover the biological function of DEGs. The protein-protein interaction (PPI) network and transcription factors (TFs)-target gene regulatory network were constructed. The expression of candidate DEGs were validated by the quantitative real-time polymerase chain reaction (qRT-PCR). Our study was useful in identifying pathogenic genes in the diagnosis or therapy of pterygium.
SUBJECTS AND METHODSEthical Approval The study was conducted in accordance with the Declaration of Helsinki and was approved by the Research Ethics Committee of the First Affiliated Hospital of Bengbu Medical College. Informed written consent was obtained from all participants, and research protocols were approved by the ethical committee of our hospital.
Table 1 Three datasets in the GEO dataset
GEO: Gene Expression Omnibus database. ahttps://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE83627; bhttps://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE51995.
?
Gene Expression Profiles Herein, we found datasets from the GEOdatabase (http://www.ncbi.nlm.nih.gov/geo/) with the keywords pterygium “[MeSH Terms] OR pterygium [All Fields]) AND “Homo sapiens”[porgn] AND “gse”[Filter].The type of study was characterized as “expression profiling by array”. Datasets were genome-wide expression data of pterygium tissues and/or normal conjunctiva tissues. The datasets should be normalized or original, and three sets of mRNA data of pterygium were selected. Three datasets were displayed in Table 1.
Identification of Differentially Expressed Genes The DEGs was calculated using Limma and metaMA packages and then p-values were adjusted for multiple comparisons by the Benjamini and Hochberg method. P＜0.05 was considered as DEGs.
Functional Annotation of Differentially Expressed Genes
The GSEABase package in R language was performed for GO(http://www.geneontology.org/) analyses and KEGG (http://www.genome. jp/kegg/pathway.html) pathway enrichment of DEGs. P＜0.01 was considered as statistically significant.
Protein-Protein Interaction The database of BioGRID(http://thebiogrid.org) was performed to construct PPI network between proteins encoded by top 200 DEGs and other proteins.Then, a PPI network was built by the Cytoscape (http://cytoscape.org/).
Analysis of Potential Transcription Factors to Target Differentially Expressed Genes The corresponding promoters of the top 40 DEGs were obtained by UCSC (http://www.genome.ucsc.edu/cgi-bin/hgTables). The TFs involved in regulating these DEGs were derived from the match tools in TRANSFAC.The Cytoscape software was used to build TFs network.
Quantitative Real-time Polymerase Chain Reaction Here,eight pterygium patients and three normal individuals were included in our study from the Department of Ophthalmology,the First Affiliated Hospital of Bengbu Medical College on 2016. Both pterygium tissues and corresponding normal conjunctiva tissues were obtained and frozen in liquid nitrogen immediately for further experiment.
Total RNA was extracted using TRizol reagent (Invitrogen,CA, USA). RNA was reverse-transcribed using a SuperScript III Reverse Transcription Kit (Invitrogen, CA, USA). Quantitative real-time PCR were conducted using SYBR Green PCR
Table 2 Top 20 DEGs in pterygium
DEGs: Differentially expressed genes.
Gene ID Gene symbol P Up-regulated genes 54829 ASPN 1.64E-12 3853 KRT6A 9.04E-12 4016 LOXL1 2.42E-09 11081 KERA 9.25E-09 1474 CST6 1.16E-08 4222 MEOX1 1.33E-08 131578 LRRC15 1.77E-08 19 ABCA1 1.90E-08 7348 UPK1B 2.48E-08 11098 PRSS23 4.35E-08 Down-regulated genes 3912 LAMB1 9.49E-05 1364 CLDN4 1.50E-04 3189 HNRNPH3 2.11E-04 54434 SSH1 2.83E-04 10432 RBM14 2.93E-04 55423 SIRPG 3.27E-04 9620 CELSR1 3.39E-04 8125 ANP32A 3.54E-04 91355 LRP5L 3.94E-04 9929 JOSD1 5.15E-04
Master Mix (Applied Biosystems, Foster City, CA, USA) on Applied Biosystems 7300 (Applied Biosystems, Foster City,CA, USA). The human GAPDH were used as endogenous controls for gene expression in analysis. Relative expression of DEGs was analyzed using the 2−ΔΔCT.
RESULTS
Differentially Expressed Genes Analysis and Functional Analyses After chip data processing, a total of 557 genes (321 up-regulated and 236 down-regulated) were identified. Top 20 genes were listed in Table 2. And the heat map of all DEGs was demonstrated in Figure 1. To investigate the biological function of the identified DEGs in pterygium, all DEGs were performed to GO term and KEGG analyses. In top 15 GO term analyses, single-organism developmental process, movement of cell or subcellular component, localization of cell and cell motility were the significant enrichment in biological process(Figure 2). Receptor binding, protein complex binding, cell adhesion molecule binding and calcium ion binding were the obviously enrichment in molecule function (Figure 3). Plasma membrane region, extracellular matrix component, condensed chromosome and chromosome, centromeric region were the most remarkably enrichment in cellular component (Figure 4).Additionally, proximal tubule bicarbonate reclamation, protein digestion and absorption were the significant enrichment signal pathway, which was shown in Table 3 and Figure 5,respectively.
Figure 1 The heat map of top 100 DEGs Diagram presents the result of a two-way hierarchical clustering of top 100 DEGs and samples. Row and column represented DEGS and GEO data, respectively. The color scale represented the expression levels.
Figure 2 Top 15 significant enrichment biological processes of DEGs.
Figure 3 Top 20 significant enrichment molecular functions of DEGs.
Protein-Protein Intercation Network PPI network was built and visualized by Cytoscape, which was shown in Figure 6. The network consisted of 163 nodes and 191 edges.The proteins with high degrees were fibronectin 1 (FN1,degree=23), replication protein A2 (RPA2, degree=18), WNT signaling pathway regulator (APC, degree=15), estrogen receptor 1 (ESR1, degree=13), karyopherin subunit beta 1(KPNB1, degree=8), damage specific DNA binding protein 1(DDB1, degree=7), neurofibromin 2 (NF2, degree=7), mitotic checkpoint serine/threonine kinase (BUB1, degree=7) and mitotic checkpoint protein (BUB3, degree=7).
Figure 4 Top 20 significant enrichment cellular components of DEGs.
Figure 5 The KEGG enrichment analysis of DEGs.
Figure 6 The PPI networks of top 200 DEGs.
Table 3 KEGG enrichment analysis
KEGG: Kyoto Encyclopedia of Genes Genomes; ECM: The extracellular matrix.
ID Term P Count Symbols 4512 ECM-receptor interaction 3.33E-05 12 COL1A1;COL5A2;COL6A3;COMP;FN1;TNC;ITGB5;LAMA5;LAMB1;SPP1;VWF;TNN 4510 Focal adhesion 1.92E-04 18 ACTN1;CAPN2;COL1A1;COL5A2;COL6A3;COMP;CRKL;FN1;TNC;ITGB5;LAMA5;LAMB1;PPP1R12A;PAK1;RAC3;SPP1;VWF;TNN 5146 Amoebiasis 2.92E-04 12 ACTN1;ARG2;CD14;COL1A1;COL5A2;FN1;LAMA5;LAMB1;SERPINB13;SERPINB4;TG FB2; TGFB3 4810 Regulation of actin cytoskeleton 4.18E-04 18 ACTN1;APC;BDKRB1;BDKRB2;CD14;CRKL;F2R;FGFR3;FN1;GSN;ITGB5;MSN;PPP1R12A;PAK1;RAC3;GNA13;SSH1;MYH14 4110 Cell cycle 3.85E-03 11 BUB1;BUB1B;CCNA2;GADD45A;RB1;TGFB2;TGFB3;TTK;CDC45;BUB3;ORC6 5210 Colorectal cancer 5.52E-03 7 APC;BIRC5;BAX;MSH2;RAC3;TGFB2;TGFB3 4974 Protein digestion and absorption 6.97E-03 8 ATP1B3;COL1A1;COL5A2;COL6A3;PRSS2;SLC1A1;SLC15A1;SLC36A1 4964 Proximal tubule bicarbonate reclamation 7.59E-03 4 AQP1;ATP1B3;CA2;GLUD2
All the nodes are proteins encoded by DEGs and the blue and red borders ellipse represent proteins encoded by top 200 DEGs, respectively; The diamond represents the proteins encoded by other DEGs.
Transcription Factors Network We utilized TRANSFAC to obtain TFs regulating top 20 DEGs. In the end, we obtained TFs network consisting of 250 pairs of TFs-genes involving 36 TFs (Figure 7). The top 9 downstream genes covered by most TFs were slingshot protein phosphatase 1 (SSH1, degree=10),protease, serine 23 (PRSS23, degree=10), LDL receptor related protein 5 like (LRP5L, degree=8), mesenchyme homeobox 1(MEOX1, degree=7), RNA binding motif protein 14 (RBM14,degree=7), ATP binding cassette subfamily A member 1(ABCA1, degree=7), Josephin domain containing 1 (JOSD1,degree=6), keratin 6A (KRT6A, degree=6) and uroplakin 1B(UPK1B, degree=6). Six hub TFs were Oct-1, Pax-4, Pax-6,Nkx2-5, AP-1 and COMP1 (Table 4).
Figure 7 The TFs network The ellipse and diamond represent TFs and top 20 DEGs, respectively; the red and blue represent the up- and down-regulated DEGs, respectively.
Table 4 Top 6 TFs that covered the most target genes
TFs: Transcription factors.
TFs Gene number Target genes 1-Oct 13 UPK1B, LAMB1, ASPN, RBM14, PRSS23, SSH1, KRT6A, KERA, LOXL1, ANP32A, MEOX1, SIRPG, LRP5L Pax-4 13 LRRC15, UPK1B, CLDN4, ABCA1, RBM14, PRSS23, CST6, SSH1, KRT6A, MEOX1, SIRPG, LRP5L, JOSD1 Pax-6 9 LAMB1, ABCA1, HNRNPH3, RBM14, CST6, SSH1, SIRPG, LRP5L, JOSD1 Nkx2-5 7 UPK1B, RBM14, CST6, SSH1, SIRPG, JOSD1, LRP5L AP-1 5 LRRC15, CLDN4, ABCA1, PRSS23, CST6 COMP1 5 ABCA1, HNRNPH3, PRSS23, SSH1, LRP5L
Validation of Quantitative Real-time Polymerase Chain Reaction The expression of DEGs was quantified by qRTPCR in blood samples of eight pterygium patients and three normal conjunctiva tissues. Eight DEGs including keratin 6A (KRT6A), ATP binding cassette subfamily A member 1(ABCA1), mesenchyme homeobox 1 (MEOX1), protease,serine 23 (PRSS23), epithelial cell transforming 2 (ECT2),fibronectin 1 (FN1), matrix metallopeptidase (MMP3) and secreted protein acidic and cysteine rich (SPARC) were selected as candidate genes. As shown in Figure 8, FN1 (P＜0.05), PRSS23(P＜0.01), ABCA1 (P＜0.05), KRT6A (P＜0.05), ECT2 (P＜0.05)and SPARC (P＜0.01), were significantly up-regulated in the pterygium compared to those of normal controls; The expression level of MEOX1 and MMP3 had no significance between pterygium and normal controls, but had the upregulation tendency in pterygium. The expression pattern of all above DEGs was consistent with our integrated analysis.
DISCUSSION
It is significant to uncover the pathological mechanism of pterygium, which may provide a novel field in diagnosis and therapy of pterygium. In this study, we found several DEGs based on the GEO dataset, which was helpful in understanding the potential mechanism of pterygium and provide a new field in finding diagnosis biomarker and drug target of pterygium.
FN1 was found with the highest degree in the PPI network of pterygium. It is one component of the extracellular matrix and is found up-regulated in whole retina of extensive nerve injury[8]. In addition, FN1 is increased in pterygia formation and recurrence[9]. Herein, we found up-regulated expression of FN1 in pterygium. And qRT-PCR also validated the upregulated tendency in pterygium. This suggested that the role of FN1 in the process of pterygium and might constitute promising new target for pterygium.
KPNB1 is a differentially methylated gene in age-related macular degeneration[10]. In addition, the expression of KPNB1 is up-regulated in optic nerve heads with elevated-intraocular pressure-induced injury of glaucoma[11]. DDB1, also called XPE, plays a crucial role in nucleotide excision global genome repair pathway. Defect of DDB1 may be lead to pure xeroderma pigmentosum[12]. It is pointed out that patients with xeroderma pigmentosum are likely to develop ocular surface squamous neoplasia[12]. The inactivation of tumor-suppressor gene NF2 is related to the formation of retinal and optic nerve lesions in neurofibromatosis type 2 patients[13], which is a clinically variable disease caused by NF2 gene mutations. Additionally,it is reported that NF2 is associated with dry eye disease[14].BUB3 is a human retinal endothelial protein identified by 2 or more peptides using multidimensional protein identification technology and plays a key role in the development of retinal[15]. Thus it can be seen that PNB1, DDB1, NF2 and BUB3 play a key role in the development of eye. Herein, we found that they were down-regulated in pterygium, which suggested the role in the process of pterygium. Further research is needed to validate their expression.
Figure 8 Validation DEGs in the pterygium by qRT-PCR
aP＜0.05, bP＜0.001 was considered as statistical significance.
PRSS23 is found differentially expressed in ipsilateral retinal ganglion cells[16]. Furthermore, the expression of PRSS23 is increased in the glaucomatous canine retina[17]. MEOX1 is implicated in differentiation of mesoderm[18]. In mice,homozygous mutation of MEOX1 gene has abnormalities in sclerotome and its derivatives[19]. Moreover, MEOX1 plays an essential function in maintaining of sclerotome polarity and axial skeleton formation[19]. ABCA1 is a cholesterol metabolic marker and mediates cholesterol efflux. The homozygous recessive mutation in the ABCA1 gene may be result in the accumulation of lipid and cholesterol in cornea and reticuloendothelial system[20]. ABCA1 is thought to play essential roles in degenerative retinal diseases[21]. Chen W and Neale BM et al identified ABCA1 as a risk factor for agerelated macular degeneration[22]. In addition, it is significantly up-regulated in pterygium[23]. KRT6A gene encodes for proteins implicated in wound healing. In a genome wide expression data, the expression of KRT6A is remarkably increased by at least 2-fold in pterygium[23]. In addition, Tong et al[23]found that KRT6A is significantly up-regulated in pterygium compared to conjunctiva in microarray analysis. In this study,we found that PRSS23, MEOX1, ABCA1 and KRT6A were all up-regulated in bioinformatics analysis and qRT-PCR validation. This suggested the association between PRSS23,MEOX1, ABCA1, KRT6A and pterygium and provided a new field for understanding the pathology mechanism of pterygium.It is reported that SSH1 plays a role in Wnt5a-enhanced migration in human corneal endothelial cells[24]. It has been demonstrated that RBM14 is down-regulated in primary visual cortex in dementia with lewy bodies[25]. UPK1B is an epithelial differentiation gene and functions in the cornea[26]. It is found up-regulated in the retinoblastoma tumors[27]. In this study, we found the down-regulated expression of SSH1 and RBM14 and up-regulated expression of UPK1B in pterygium. We speculated that SSH1, RBM14 and UPK1B played pivotal roles in the visual function of pterygium.
ECT2 has been associated with tumor progression and is detected in a number of retinal progenitor cells[28]. Moreover,ECT2 is expressed in response to chemotherapy induced stress in the retinoblastoma therapy[29]. Additionally, ECT2 is overexpressed in the immortalized human corneal epithelial cell lines[30]. MMPs plays a crucial role in various eye diseases.And disordered expression of MMPs in the eye has been involved in different disorders, such as pterygium[31]. MMP3,one member of MMPs, is an attractive candidate for the pathogenesis and development of pterygium. Seet et al[1] and Di Girolamo et al[31] found that MMP3 was up-regulated in pterygium. In addition, it is suggested that inhibitors of MMP3 have a protective role against extracellular matrix degradation,which is a manifestation of pterygium. In our study, we found that MMP3 was up-regulated in pterygium, which was in line with previous reports. This further indicated the role of MMP3 in the pathology process of pterygium. SPARC is a matrix protein that plays roles in various biological processes including epithelial-mesenchymal transition, tissue fibrosis and angiogenesis in primary pterygium[9]. Additionally, SPARC is a well-known mediator in collagen production and deposition[32].It is reported that the expression of SPARC is related to aberrant extracellular matrix deposition in pterygium[1]. In this study, SPARC was up-regulated in pterygium, which was consistent with previous reports. This suggested that SPARC may be complicated in the pathogenesis of pterygium.
In addition, there was no study reported that the association of other five DEGs (RPA2, APC, ESR1, LRP5L and JOSD1)with eye disease, especially pterygium, their potential roles in pterygium needs further research.
In summary, several DEGs including FN1, KPNB1, DDB1,NF2, BUB3, PRSS23, MEOX1, ABCA1, KRT6A, SSH1,RBM14 and UPK1B play a pivotal role in the pathology of pterygium. Our findings may contribute to the identification of early diagnosis biomarker or drug target for pterygium.
There are limitations to this study. Firstly, the intensive molecular function of above genes wasn't investigated in this study. The in vivo and in vitro experiments are necessary for exploring the molecular function of these genes in the process of pterygium in the future work. Secondly, the sample size for qRT-PCR validation was small, additional studies with larger cohorts of pterygium are needed to validate and demonstrate the expression pattern of identified genes.
ACKNOWLEDGEMENTS
Authors’ contributions: Yue XL was a major contributor in writing the manuscript. Yue XL and Gao ZQ conducted the data analysis.
Foundation: Supported by Science and Technology Development Fund of Bengbu Medical College (No.BYFY1785).
Conflicts of Interest: Yue XL, None; Gao ZQ, None.
1 Seet LF, Tong L, Su R, Wong TT. Involvement of SPARC and MMP-3 in the pathogenesis of human pterygium. Invest Ophthalmol Vis Sci 2012;53(2):587-595.
2 Chui J, Di Girolamo N, Wakefield D, Coroneo MT. The pathogenesis of pterygium: current concepts and their therapeutic implications. Ocul Surf 2008;6(1):24-43.
3 Luthra R, Nemesure BB, Wu SY, Xie SH, Leske MC, Barbados Eye Studies Group. Frequency and risk factors for pterygium in the Barbados Eye Study. Arch Ophthalmol 2001;119(12):1827-1832.
4 Di Girolamo N, Chui J, Coroneo MT, Wakefield D. Pathogenesis of pterygia: role of cytokines, growth factors, and matrix metalloproteinases.Prog Retin Eye Res 2004;23(2):195-228.
5 Sánchez-Thorin JC, Rocha G, Yelin JB. Meta-analysis on the recurrence rates after bare sclera resection with and without mitomycin C use and conjunctival autograft placement in surgery for primary pterygium. Br J Ophthalmol 1998;82(6):661-665.
6 Hirst LW. The treatment of pterygium. Surv Ophthalmol 2003;48(2):145-180.
7 Wong YW, Chew J, Yang H, Tan DT, Beuerman R. Expression of insulin-like growth factor binding protein-3 in pterygium tissue. Br J Ophthalmol 2006; 90(6):769-772.
8 Guo Y, Cepurna WO, Dyck JA, Doser TA, Johnson EC, Morrison JC. Retinal cell responses to elevated intraocular pressure: a gene array comparison between the whole retina and retinal ganglion cell layer.Invest Ophthalmol Vis Sci 2010;51(6):3003-3018.
9 Engelsvold DH, Utheim TP, Olstad OK, Gonzalez P, Eidet JR, Lyberg T, Trøseid AM, Dartt DA, Raeder S. miRNA and mRNA expression profiling identifies members of the miR-200 family as potential regulators of epithelial-mesenchymal transition in pterygium. Exp Eye Res 2013;115:189-198.
10 Oliver VF, Jaffe AE, Song J, Wang G, Zhang P, Branham KE,Swaroop A, Eberhart CG, Zack DJ, Qian J, Merbs SL. Differential DNA methylation identified in the blood and retina of AMD patients.Epigenetics 2015;10(8):698-707.
11 Johnson EC, Jia L, Cepurna WO, Doser TA, Morrison JC. Global changes in optic nerve head gene expression after exposure to elevated intraocular pressure in a rat glaucoma model. Invest Ophthalmol Vis Sci 2007;48(7):3161-3177.
12 Suarez MJ, Rivera-Michlig R, Dubovy S, Rodriguez FJ.Clinicopathological features of ophthalmic neoplasms arising in the setting of xeroderma pigmentosum. Ocul Oncol Pathol 2015;2(2):112-121.
13 Chan CC, Koch CA, Kaiser-Kupfer MI, Parry DM, Gutmann DH,Zhuang ZP, Vortmeyer AO. Loss of heterozygosity for the NF2 gene in retinal and optic nerve lesions of patients with neurofibromatosis 2. J Pathol 2002;198(1):14-20.
14 Ren G, Li H, Qiao W, Shen H, Zhuang Y, Shao T, Hu H, Zhang X.Possible association of killer cell immunoglobulin-like receptor genotypes and haplotypes with dry eye disease in a Han Chinese population. Mol Vis 2015;21:948-954.
15 Bharadwaj AS, Appukuttan B, Wilmarth PA, Pan Y, Stempel AJ,Chipps TJ, Benedetti EE, Zamora DO, Choi D, David LL, Smith JR. Role of the retinal vascular endothelial cell in ocular disease. Prog Retin Eye Res 2013;32:102-180.
16 Wang Q, Marcucci F, Cerullo I, Mason C. Ipsilateral and contralateral retinal ganglion cells express distinct genes during decussation at the optic chiasm. eNeuro 2016;3(6):ENEURO.0169-ENEURO.0116.2016.
17 Jiang B, Harper MM, Kecova H, Adamus G, Kardon RH, Grozdanic SD, Kuehn MH. Neuroinflammation in advanced canine glaucoma. Mol Vis 2010;16:2092-2108.
18 Yazar S, Mishra A, Ang W, Kearns LS, Mountain JA, Pennell C,Montgomery GW, Young TL, Hammond CJ, Macgregor S, Mackey DA, Hewitt AW. Interrogation of the platelet-derived growth factor receptor alpha locus and corneal astigmatism in Australians of Northern European ancestry: results of a genome-wide association study. Mol Vis 2013;19:1238-1246.
19 Skuntz S, Mankoo B, Nguyen MT, Hustert E, Nakayama A, Tournier-Lasserve E, Wright CV, Pachnis V, Bharti K, Arnheiter H. Lack of the mesodermal homeodomain protein MEOX1 disrupts sclerotome polarity and leads to a remodeling of the cranio-cervical joints of the axial skeleton. Dev Biol 2009;332(2):383-395.
20 Brooks-Wilson A, Marcil M, Clee SM, Zhang LH, Roomp K, van Dam M, Yu L, Brewer C, Collins JA, Molhuizen HO, Loubser O, Ouelette BF,Fichter K, Ashbourne-Excoffon KJ, Sensen CW, Scherer S, Mott S, Denis M, Martindale D, Frohlich J, Morgan K, Koop B, Pimstone S, Kastelein JJ, Genest J Jr, Hayden MR. Mutations in ABC1 in Tangier disease and familial high-density lipoprotein deficiency. Nat Genet 1999;22(4):336-345.
21 Cvekl A, Wang WL. Retinoic acid signaling in mammalian eye development. Exp Eye Res 2009;89(3):280-291.
22 Chen W, Stambolian D, Edwards AO, et al. Genetic variants near TIMP3 and high-density lipoprotein-associated loci influence susceptibility to age-related macular degeneration. Proc Natl Acad Sci U S A 2010;107(16):7401-7406.
23 Tong L, Chew J, Yang H, Ang LP, Tan DT, Beuerman RW. Distinct gene subsets in pterygia formation and recurrence: dissecting complex biological phenomenon using genome wide expression data. BMC Med Genomics 2009;2:14.
24 Lee JG, Heur M. Interleukin-1β-induced Wnt5a enhances human corneal endothelial cell migration through regulation of Cdc42 and RhoA.Molecular Cell Biol 2014;34(18):3535-3545.
25 Khundakar AA, Hanson PS, Erskine D, et al. Analysis of primary visual cortex in dementia with Lewy bodies indicates GABAergic involvement associated with recurrent complex visual hallucinations. Acta Neuropathol Commun 2016;4:66.
26 Adachi W, Okubo K, Kinoshita S. Human uroplakin Ib in ocular surface epithelium. Invest Ophthalmol Vis Sci 2000;41(10):2900-2905.
27 Houston SK, Pina Y, Clarke J, Koru-Sengul T, Scott WK, Nathanson L, Schefler AC, Murray TG. Regional and temporal differences in gene expression of LH(BETA)T(AG) retinoblastoma tumors. Invest Ophthalmol Vis Sci 2011;52(8):5359-5368.
28 Sano M, Genkai N, Yajima N, Tsuchiya N, Homma J, Tanaka R, Miki T, Yamanaka R. Expression level of ECT2 proto-oncogene correlates with prognosis in glioma patients. Oncol Rep 2006;16(5):1093-1098.
29 Nalini V, Segu R, Deepa PR, Khetan V, Vasudevan M, Krishnakumar S.Molecular insights on post-chemotherapy retinoblastoma by microarray gene expression analysis. Bioinform Biol Insights 2013;7:289-306.
30 Greco D, Vellonen KS, Turner HC, Häkli M, Tervo T, Auvinen P,Wolosin JM, Urtti A. Gene expression analysis in SV-40 immortalized human corneal epithelial cells cultured with an air-liquid interface. Mol Vis 2010;16:2109-2120.
31 Di Girolamo N, McCluskey P, Lloyd A, Coroneo MT, Wakefield D. Expression of MMPs and TIMPs in human pterygia and cultured pterygium epithelial cells. Invest Ophthalmol Vis Sci 2000;41(3):671-679.32 Bradshaw AD. The role of SPARC in extracellular matrix assembly. J Cell Commun Signal 2009;3(3-4):239-246.