Hindawi Publishing CorporationInternational Journal of GenomicsVolume 2013 Article ID 480534 10 pageshttpdxdoiorg1011552013480534
Research ArticleIntegrated Analysis of Long Noncoding RNA and Coding RNAExpression in Esophageal Squamous Cell Carcinoma
Wei Cao1 Wei Wu12 Fachun Shi3 Xiaobing Chen1 Lihua Wu1 Ke Yang1 Fu Tian1
Minghui Zhu1 Guoyong Chen1 WeiWei Wang1 Fred G Biddle4 and Jianqin Gu3
1 Clinical Research Center Peoplersquos Hospital of Zhengzhou 33 Yellow River Road Zhengzhou Henan 45003 China2Department of Pathology and Experimental Medicine University of Calgary Calgary AB Canada T2N 4N13 Science and Education Department Health Bureau of Zhengzhou China4Departments of Medical Genetics and Biological Sciences University of Calgary Calgary AB Canada T2N 4N1
Correspondence should be addressed to Wei Cao caoweiyuhotmailcom
Received 28 May 2013 Accepted 26 August 2013
Academic Editor Soraya E Gutierrez
Copyright copy 2013 Wei Cao et alThis is an open access article distributed under the Creative Commons Attribution License whichpermits unrestricted use distribution and reproduction in any medium provided the original work is properly cited
Tumorigenesis is a complex dynamic biological process that includes multiple steps of genetic and epigenetic alterations aberrantexpression of noncoding RNA and changes in the expression profiles of coding genes We call the collection of those perturbationsin genome space the ldquocancer initiatomerdquo Long noncoding RNAs (lncRNAs) are pervasively transcribed in the genome andthey have key regulatory functions in chromatin remodeling and gene expression Spatiotemporal variation in the expression oflncRNAs has been observed in development and disease states including cancer A few dysregulated lncRNAs have been studiedin cancers but the role of lncRNAs in the cancer initiatome remains largely unknown especially in esophageal squamous cellcarcinoma (ESCC)We conducted a genome-wide screen of the expression of lncRNAs and coding RNAs from ESCC andmatchedadjacent nonneoplastic normal tissues We identified differentially expressed lncRNAs and coding RNAs in ESCC relative to theirmatched normal tissue counterparts and validated the result using polymerase chain reaction analysis Furthermore we identifieddifferentially expressed lncRNAs that are co-located and co-expressed with differentially expressed coding RNAs in ESCC and theresults point to a potential interaction between lncRNAs and neighboring coding genes that affect ether lipid metabolism and theinteraction may contribute to the development of ESCC These data provide compelling evidence for a potential novel genomicbiomarker of esophageal squamous cell cancer
1 Introduction
Esophageal squamous cell carcinoma (ESCC) is one of themost common types of cancer and it ranks among the maincauses of cancer deaths worldwide [1] There are markedregional variation and exceptionally high incidence in certainareas of China Despite advances in multidisciplinary treat-ment of ESCC 5-year survival rate remains poor The ini-tiatome [2] of ESCC is a complex dynamic biological processin genome space and it may include multiple steps of geneticand epigenetic alterations [3] aberrations in expression ofnoncoding RNA (eg microRNAs) [4] and changes in theexpression profile of coding genes [5 6] In past decadesexpression profiling of coding genes has defined important
signaling pathways involved in tumorigenesis The latestknowledge of actively transcribed long noncoding RNAs(lncRNAs) from high-throughput sequencing is revealingan even greater complexity about cancer genome regulatorynetworks
LncRNAs are endogenous cellular RNA transcripts rang-ing from 200 to 100000 nucleotides in length and theylack an open reading frame of significant length (less than100 amino acids) [7] LncRNAs are generally expressed ata lower level than protein-coding genes and they displaymore tissue-specific and cell-specific expression patterns [89] LncRNAs were previously believed to be transcriptionalnoise but now they have critical roles in development anddifferentiation as well as in the proliferation and progress
2 International Journal of Genomics
of disease including cancer [10] Mechanisms of action oftranscribed lncRNAs are described as modifying chromatinarchitecture and regulating gene expression in a cis or transmanner For example H19 lncRNA cis-regulates IGF2 geneexpression at the same genomic locus and HOTAIR lncRNAis transcribed on Chr 12 and it transregulates HoxD geneon Chr 2 Additionally lncRNAs have also been reportedto coordinate the regulation of neighboring coding genesthrough a ldquolocus controlrdquo process [11] which mediates thelocalization of genes within nuclear regions to favor theirtranscription through the formation of domains of histonemodification and intra- or interchromosomal loops [12]Dysregulated lncRNAs have been identified with differentscreening methodologies in various types of cancer Forexample the cancer-related lncRNA metastasis-associatedlung adenocarcinoma transcript 1 (MALAT-1) was identifiedby subtractive hybridization during screening for early non-small cell lung cancer with metastasis [13] Overexpression ofMALAT-1 is highly predictive of poor prognosis and short-ened survival time in early stage lung cancer OverexpressionofHOTAIR lncRNAwas found in several solid tumors [14ndash17]in association with cancermetastasis and increasedHOTAIRexpression in breast cancer is transcriptionally induced byestradiol [18] Prostate cancer associated lncRNA PCGEM1[19] andPCAT-1 [20] appear to be prostate-specific regulatorsof cell apoptosis and proliferation Recently AFAP1-AS1lncRNA was reported to be overexpressed in esophagealadenocarcinoma [21]
The handful of dysregulated lncRNAs in different cancerssuggests that lncRNAs are an enigmatic component of thewhole transcriptome which may participate in tumorige-nesis invasion and metastasis Efforts are being made toexplore the ldquolncRNAomerdquo of various cancers with advancedhigh-throughput RNA sequencing technologies [8 9] anddynamic changes in lncRNA expression have been observedin cancer cells during different stages of cancer developmentand during treatment [22] However our understanding ofthe role of lncRNAs in cancer biology is still in an early stageand a clearly defined predictive set of biological functionsfor lncRNAs is lacking in cancer biologyTherefore thoroughsearches and analyses of the interactions between lncRNAand coding genes may help to infer their potential biologicalroles
In order to understand the role of lncRNAs in ESCC wereport a pilot study of the profiles of differentially expressedlncRNAs and coding RNAs from tumor and adjacent normaltissue of individual patients with ESCC We assessed thewhole transcriptomic landscape for potential interactionsbetween lncRNAs and coding-gene expression In particularwe evaluated the coding genes that are co-located and co-expressed with the differentially expressed lncRNAs duringthe genesis of ESCC
2 Results and Discussion
21 Transcriptomic Landscape of ESCC Our genome-widegene expression profiling of both lncRNAs and coding genesfrom ESCC and adjacent nonneoplastic tissue was conducted
to detect possible associations of lncRNAs with ESCC Wefirst asked whether these transcripts of 7419 noncoding and27958 coding RNAs could distinguish ESCC from normaltissues Figure 1(a) shows that the four ESCC samples areclustered together in one group and clearly separated fromthe samples of normal tissue Next we examined the wholetranscriptomic pattern (lncRNAs + coding RNAs) fromeach sample and the landscapes of the whole transcriptome(represented by heatmaps in Figure 1(a)) of normal tissuesdiffer from those of ESCC that exhibit more heterogeneousalterations The overall changes from a respective normalto cancer state were also seen separately as a differencein expression profile of either the lncRNA or the codingRNA These observations suggest that a potential dynamicinteraction between lncRNAs and coding RNAs may bereshaping the landscape of the whole transcriptome duringESCC development
To gain a detailed understanding of the biological themesof all RNA transcripts we further identified those transcriptsthat are significantly and differentially expressed (DE) inESCC tissue compared to matched normal tissue based onthe criteria described in the methods There are 410 DE-lncRNAs and 1219 DE-mRNAs that represent about 5 ofthe transcripts in the respective microarrays (SupplementaryTables S1 and S2 available online at httpdxdoiorg1011552013480534) DE-lncRAs distinguish a cancer cell from itsnormal cell state with three times fewer transcripts thanDE-mRNAs (Figures 1(b) and 1(c)) suggesting that the DE-lncRNA profile is more informative and potentially a morefaithful indicator of a specific cell state
Enrichment analysis of DE-mRNAs demonstrated thatthe respective genes are involved in cancer-related pathways(Figure 1(d)) Since expression profiling of coding-RNA hasbeen intensively studied in esophageal cancer we validated10 genes whose expression level in other studies [23ndash25] issignificantly changed (119875 lt 005) by at least 2-fold relative tonormal tissues (Table 1)
22 Expression of lncRNAs in ESCC LncRNAs are emergingas a novel class of noncoding RNAs that are pervasivelytranscribed in the genome but there is limited functionalknowledge about them High-throughput screening of lncR-NAs from ESCC has been poorly studied except for a recentreport of overexpressed lncRNA AFAP1-AS1 in esophagealadenocarcinoma [21] In our study a total of 7419 intergeniclncRNAs and other transcripts of uncertain coding potentialwere examined and we identified 410 DE-lncRNAs in ESCCrelative to adjacent normal esophageal tissues We namedthe anonymous lncRNAs ESCC Associated Long noncodingRNAs (ESCCAL Supplementary Table S1) Expression ofHOTAIR lncRNA is increased in various cancers [14ndash1726] and it is also significantly increased in our analysisof ESCC (Figure 2(a)) In addition we confirmed anothertwo upregulated lncRNAs that are differentially expressed inESCC and that we have named ESCCAL-1 and ESCCAL-5 The increased and differential expression of HOTAIRESCCAL-1 and ESCCAL-5 in ESCC tissue relative to adja-cent nonneoplastic tissue was independently assessed with
International Journal of Genomics 3
Wholetranscriptome
lncRNAs mRNAs
3N
4N
1N 2N
1T
3T
2T 4T
(a)
3N4N
1N2N
1T3T2T 4T
DE-
lncR
NA
s
minus2 0 +2
(b)
DE-
mRN
As
3N4N
1N 2N 1T 3T
2T4T
minus2 0 +2
(c)
Pathways in cancer
Pancreatic cancer
Renal cell carcinomaErbB signaling pathway
Wnt signaling pathwayMAPK signaling pathway
Nonsmall cell lung cancer
p53 signaling pathway
(d)
Figure 1 Transcriptomic landscape of esophageal squamous cell cancer (ESCC) (a) Whole transcriptome of tumor (T) and adjacent normaltissue (N) of four patients with ESCCwere detected using a microarray with 7419 long noncoding RNAs (lncRNAs) and 27958 coding RNAsTwo main clusters (Ts and Ns) were generated using unsupervised clustering methods Then a self-organizing map (SOM) of either wholetranscriptome (both lncRNAs and mRNAs) or lncRNAs or mRNA was produced from each sample (see legend in up-left corner of thisfigure and the arrows are meant to indicate the potential interaction) using gene expression dynamic inspector (GEDI) Mosaic patternsare pseudocolored SOMs to show integrated biological entity in each sample Red through blue color indicates high to low expression level(b) and (c) Differentially expressed lncRNAs (DE-lncRNAs) and coding RNAs (DE-mRNAs) in ESCC Hierarchical clustering analysis of 410DE-lncRNAs (b) and 1219 DE-mRNAs (c) between ESCC tissue and adjacent normal tissue (fold change gt or lt 2-fold and 119875 lt 005) Redand green colors indicate high and low expression respectively In the heatmap columns represent samples and rows represent each geneThe scale of expression level is shown on the horizontal bar (d) KEGG functional analysis of DE-mRNA networks in ESCCThe DE-mRNAgenes are involved in cancer-related signaling functions and a detailed list of significant GO terms is shown in Figure S1 and its associatedlegend in Supplementary Information
4 International Journal of Genomics
100
80
60
40
20
0N T N T N T
Nor
mal
ized
inte
nsity
in m
icro
arra
y
Nor
mal
ized
inte
nsity
Nor
mal
ized
inte
nsity
in m
icro
arra
y
in m
icro
arra
y
HOTAIRlowast
lowastlowast
lowastlowast
ESCCAL-1 ESCCAL-52500
2000
1500
1000
500
0
700
600
500
400
300
200
100
0
(a)
N T N T N T N T N T N T N T N T N T N T N T N T
GA
PDH
HO
TAIR
Patient 1 Patient 2 Patient 3
ESCC
AL-
1
ESCC
AL-
5
GA
PDH
HO
TAIR
ESCC
AL-
1
ESCC
AL-
5
GA
PDH
DN
A m
arke
r
HO
TAIR
ESCC
AL-
1
ESCC
AL-
5(b)
Figure 2 Long noncoding RNAs (lncRNAs) expression in esophageal squamous cell carcinoma (ESCC) (a) Three differentially expressedlncRNAs HOTAIR ESCCAL-1 and ESCCAL-5 from microarray detection The average intensity of expression in normal tissues (N) andtumors (T) is plotted with their standard deviations (b) Validation ofHOTAIR ESCCAL-1 and ESCCAL-5with independent patient samplesby PCR analysis The amplicons were separated with 2 agarose gel GAPDH was used as an internal control Significance is lowast119875 lt 005lowastlowast
119875 lt 001
Table 1 Validation of selected differential expression of mRNAs in esophageal squamous cell carcinoma in independent studies
Probe name P value FC Regulation Gene symbol Genbank accession Independent study ReferenceA 33 P3232692 0005838984 8301461 Up IL24 NM 001185156
Microarray [20]
A 24 P411121 000055 5329484 Up TNFRSF18 NM 148901A 23 P169097 881119864 minus 05 4466178 Up WISP1 NM 080838A 23 P304304 0004822649 3944957 Down ARSF NM 004042A 24 P56363 0003573538 3323955 Down CAB39L NM 030925A 23 P419760 0001041661 3270335 Down CRISP3 NM 006061A 23 P413923 0002921898 454899 Down DMRTA1 NM 022160A 23 P56978 0002093997 5438183 Down PTK6 NM 005975 RNA-seq [21]A 23 P115091 0005171322 3289834 Down RAB25 NM 020387 Q-RT-PCR [22]A 33 P3258542 0001039129 2036035 Down SPINK8 NM 001080525 Microarray [20]
International Journal of Genomics 5
PCR methods in matched-pair tissue samples from threeadditional ESCC patients and the results are consistent withthemicroarray analysis (Figure 2(b)) Interestingly except forHOTAIR other previously reported lncRNAs (ie MALAT-1 PCAT-1 and AFAP1-AS1) are not differentially expressed inour analysis of ESCC Therefore the DE-lncRNAs that wehave identifiedmay be a unique property of ESCC andwe arecurrently using a population-based analysis to characterizethese DE-lncRNAs as potential genomic biomarkers andregulatory elements in the dynamic process leading to ESCC
23 LncRNAs Co-located and Co-expressed with Coding Genesin ESCC LncRNAs have been reported to coordinate theregulation of neighboring coding genes through a ldquolocuscontrolrdquo process [11] We wondered whether such a ldquolocuscontrolrdquo process could operate in ESCC development andtherefore we searched neighboring genes of the 410 DE-lncRNAs in the genomeThemajority (988) of the 410 DE-lncRNAs harbor neighboring coding genes whose genomiclocations are within sim5 kb upstream and sim1 kb downstreamof the lncRNA and may extend to 1000 kb in both directions(Figure 3(a)) Interrogation of 538 coding genes that areneighbors of these DE-lncRNAs (DE-lncRNAs co-locatedgenes) revealed predicted functions in 9 common pathwayssuch as the AP1 transcription factor network integrin-linkedkinase signaling several signaling pathways in adherensjunctions and FOXO family signaling (Figure 3(b))
We asked whether any DE-lncRNAs co-located genes arealso differentially expressed in ESCC Analysis of the DE-lncRNAs co-located genes withDE-mRNAdata set identified76 genomically co-located and differentially co-expressedgenes (Figure 3(c) and Table 2) Strikingly the co-locatedand co-expressed genes with DE-lncRNAs may be involvedin ether lipid metabolism pathways by the participation ofthe LPCAT1 gene encoding lysophosphatidylcholine acyl-transferase1 and the PLD1 gene encoding phospholipase D1(Figure 3(c)) The lncRNA ESCCAL-337 (chr3171506370-171528740) was downregulated in ESCC and located at22068 bp downstream of the PLD1 gene whose expres-sion was also decreased in ESCC In contrast the lncRNAESCCAL-356 (chr51544500-1567142 reverse strand) wasdownregulated in ESCC and located at 21250 bp upstreamof LPCAT1 whose expression was upregulated in ESCC(Figure 3(c)) LPCAT1 modulates phospholipid compositionby catalyzing lysophosphatidylcholine into phosphatidyl-choline and overexpression of LPCAT1was reported to createfavorable conditions for cancer cell proliferation [27 28]Therefore at least two of the DE-lncRNAs have the potentialto contribute to ESCC by a ldquolocus controlrdquo process withneighboring coding genes
In conclusion we performed a genome-wide survey ofthe expression of lncRNAs and coding mRNAs from pairedsamples of primary neoplastic tissue and adjacent non-neoplastic normal tissue from four individuals The overalltranscriptomic landscape (both lncRNAs andmRNAs) is ableto distinguish malignant from normal tissue in each personWe discovered a set of differentially expressed lncRNAsand their co-located and co-expressed coding mRNAs and
demonstrated that lncRNAs may be involved in ether lipidmetabolism in ESCC Our study provides genomic supportfor a model of a ldquolocus controlrdquo process in ESCC and aframework for further experimental study
3 Materials and Methods
31 Specimens Written informed consent was obtained frompatients before surgery and the study protocol was approvedby the Institutional Review Board for the use of human sub-jects at Zhengzhou Hospital Primary tumors and adjacentnonneoplastic tissues were obtained frompatients with ESCCwhounderwent surgical treatment at LinxianHospital inMay2012 All tissues were frozen in liquid nitrogen immediatelyafter surgical resection None of the patients had priorchemotherapy or radiotherapy nor did they have any otherserious diseases All ESCC tissues were histopathologicallydiagnosed by at least two independent senior pathologists
32 Microarray Hybridization Total RNAs were extractedusing Trizol reagent following manufacturerrsquos instructions(Invitrogen Carlsbad CA USA) The quality of RNAswas measured with a 2100 Bioanalyzer (Agilent technologyUSA) Input of 100 ng of total RNA was used to generateCyanine-3 labeled cRNA according to theAgilentOne-ColorMicroarray-Based Gene Expression Analysis Low for InputQuick Amp Labeling kit (v60) Samples were hybridizedon Agilent SurePrint G3 Human GE 8 times 60K Microarray(Design ID 028004) Arrays were scanned with the AgilentDNAMicroarray Scanner at a 3120583m scan resolution and datawere processed with Agilent Feature Extraction 11011 Themicroarray data discussed in this article have been depositedin National Center for Biotechnology Information (NCBI)Gene ExpressionOmnibus (GEO) and are accessible through(GEO) Series accession number GSE45350 (httpwwwncbinlmnihgovgeoqueryacccgiacc=GSE45350)
33 Validation by Polymerase Chain Reaction (PCR) PCRanalysis was performed on additional matched ESCC andadjacent nonneoplastic tissues for selected lncRNAs Theprimer sequences for PCR are as follows HOTAIR forward51015840-GGTAGAAAAAGCAACCACGAAGC-31015840 and reverse51015840-ACATAAACCTCTGTCTGTGAGTGCC-31015840 ESSCAL-1(chr876121095-76189420 reverse strand) forward 51015840-CCA-GACAGCAGCAAAGCAAT-31015840 and reverse 51015840-GGAAGC-AGCAAATGTGTCCAT-31015840 ESSCAL-5 (chr2216585154-216585719 forward strand) forward 51015840-TACCAACATTGT-CCACCGGG-31015840 and reverse 51015840-GCTGATGACAGTCCC-TTGCT-31015840 GAPDH was used as a control forward 51015840-CCG-GGAAACTGTGGCGTGATGG-31015840 and reverse 51015840-AGG-TGGAGGAGTGGGTGTCGCTGTT-31015840 The thermocycleconditions are as follows initial denaturation at 95∘C for10 minutes followed by 94∘C for 45 seconds 65∘C for 30seconds and 72∘C for 1 minute for 15 cycles Then theannealing temperature was reduced by 05∘Ccycle for thenext 14 cycles and the amplification was finished withanother 24 cycles with the annealing temperature at 58∘C
6 International Journal of Genomics
lncRNAs
Gene b Gene c
20
40
60
80
100
0lncRNAs
Gene a
lncRNAs
Number of associated genes per region0 1 2G
enom
ic re
gion
s of l
ncRN
As (
)
(a)
AP-1 transcription factor network
Integrin-linked kinase signalingRegulation of CDC42 activity
Posttranslational regulation of adherens junction stability and disassemblyN-cadherin signaling events
E-cadherin signaling in the nascent adherens junctionStabilization and expansion of the E-cadherin adherens junction
E-cadherin signaling eventsFOXO family signaling
1176
1105904
511466
448448440
422
(b)
DE-lncRNAsco-located genes
DE-mRNAs
538 330776
ESCCAL-356
ESCCAL-337
Chr3
Chr5
Lipid metabolism
PLD1
LPCAT1
(c)
Figure 3 Identification of lncRNAs co-located and co-expressed neighboring genes in esophageal squamous cell carcinoma (ESCC)(a) Identification of neighboring genes of the DE-lncRNAs The genomic coordinate information of 410 DE-lncRNAs was used to searchneighboring genes whose genomic locations are within sim5 kb upstream and sim1 kb downstream of the lncRNA and may extend to 1000 kbin both directions using GREAT software (httpbejeranostanfordedugreatpublichtmlindexphp) The percentage of DE-lncRNAsharboring zero one or two neighboring genes is presented (b) Gene Ontology (GO) enrichment analysis of lncRNAs co-located genesIdentified gene enriched pathwaysterms are listed on the left the length of horizontal bars and the numbers on the right indicate thepercentage of genes involved in each pathwayterm (c) LncRNAs co-located and co-expressed coding mRNAs Overlap of 538 DE-lncRNAco-located genes with 3307 DE-mRNAs in microarrays identified 76 lncRNAs co-located and co-expressed coding mRNAs (list in Table 2)GO enrichment analysis suggests phospholipase D1 (PLD1) and lysophosphatidylcholine acyltransferase1 (LPCAT1) are involved in ether lipidmetabolism pathway Genomic location shows that PLD1 is located at minus22068 bp upstream of ESCCAL-337 lncRNA on Chr 3 and LPCAT1is at minus21250 bp upstream of ESCCAL-356 lncRNA on Chr 5
International Journal of Genomics 7
Table2Listof
identifi
edco-lo
catedandco-expressed
genesw
ithdifferentially
expressedlncR
NAsinES
CC
LncR
NA
nam
eG
enom
ic co
ordi
nate
Ln
cRN
Aex
pres
sion
LncR
NA
asso
ciat
edge
nes
Gen
eexp
ress
ion
Gen
esym
bol
Gen
bank
acce
ssio
nES
CCA
L-177
chr1
205404138
-205404079
Up
CDK18
(minus69575
)LE
MD
1(minus12895
)U
pLE
MD1
NM001001552
ESC
CAL-348
chr1
222587441
-222587382
Up
DU
SP10
(minus671896
)H
HIP
L2(+134032
)U
pD
USP10
NM007207
ESC
CAL-31
chr1
225237693
-225237752
Up
DN
AH14
(+120367
)LB
R(+378092
)U
pD
NA
H14
NM001373
ESC
CAL-159
chr1
225240300
-225240359
Up
DN
AH14
(+122974
)LB
R(+375485
)U
pD
NA
H14
NM001373
ESC
CAL-327
chr1
89887111
-89887052
Up
LRRC8
B(minus103315
)G
BP6
(+57646
)D
own
GBP6
NM198460
XLO
C000915
chr1
91295528
-91295469
Up
BARH
L2(minus112705
)ZN
F644
(+192313
2N
M020063
ESC
CAL-65
chr11
2017146
-2017205
Up
MRP
L23
(+48674
)IG
F2(+145165
)U
pIG
F2N
M000612
ESC
CAL-19
chr12
66204406
-66204347
Up
HM
GA
2(minus13863
)M
SRB3
(+531610
)U
pH
MG
A2
NM003484
XLO
C011548
chr15
81953024
-81952965
Up
TMC3
(minus286577
)M
EX3
B(+385366
)U
pM
EX3
BN
M032246
ESC
CAL-102
chr17
6766871
-6766930
Up
ALO
X12
(minus132483
)TE
KT1
(minus31841
)12
NM000697
ESC
CAL-342
chr17
78302158
-78302217
Up
END
OV
(minus86779
)RN
F213
(+67521
)U
pRN
F213
NM020954
ESC
CAL-342
chr17
78302158
-78302217
Up
END
OV
(minus86779
)RN
F213
(+67521
)001164638
ESC
CAL-99
chr2
102034125
-102034066
Up
CREG
2(minus30131
)RF
X8(+57069
)U
pCR
EG2
NM153836
ESC
CAL-5
chr2
216585455
-216585514
Up
FN1
(minus284694
)M
REG
(+292861
)D
own
MRE
GN
M018000
ESC
CAL-5
chr2
216585455
-216585514
Up
FN1
(minus284694
)M
REG
(+292861
)U
pFN1
NM054034
ESC
CAL-288
chr2
27789661
-27789602
Up
ZNF512
(minus16261
)G
CKR
(+69926
)D
own
GCK
RN
M001486
ESC
CAL-344
chr2
37327299
-37327358
Up
CCD
C75
(+15735
)EI
F2A
K2
(+56861
)U
pEI
F2A
K2N
M001135652
ESC
CAL-10
chr22
48086469
-48086528
Up
FAM
19A
5(minus798789
)TB
C1D22
A(+927981
)U
pFA
M19
A5
NM015381
ESC
CAL-81
chr5
141710377
-141710436
Up
SPRY4
(minus5787
)FG
F1(+355246
)U
pFG
F1N
M000800
ESC
CAL-81
chr5
141710377
-141710436
Up
SPRY4
(minus5787
)FG
F1(+355246
)U
pSP
RY4
NM030964
ESC
CAL-204
chr6
126699769
-126699828
Up
RSPO3
(minus740249
)CE
NPW
(+38546
)U
pCE
NPW
NM001012507
ESC
CAL-106
chr6
21822910
-21822969
Up
SOX4
(+228968
)PR
L(+480142
)U
pSO
X4N
M003107
ESC
CAL-239
chr8
104258684
-104258625
Up
FZD
6(minus52006
)BA
ALC
(+105734
)U
pBA
ALC
NM024812
ESC
CAL-239
chr8
104258684
-104258625
Up
FZD
6(minus52006
)BA
ALC
(+105734
)U
pFZ
D6
NM003506
ESC
CAL-300
chr1
180918852
-180918793
Dow
nST
X6(+73223
)XP
R1(+317677
)U
pXP
R1N
M004736
XLO
C000515
chr1
200384539
-200384598
Dow
nZN
F281
(minus5403
)KI
F14
(+205293
)U
pKI
F14
NM014875
XLO
C000515
chr1
200384539
-200384598
Dow
nZN
F281
(minus5403
)KI
F14
(+205293
)U
pZN
F281
NM012482
ESC
CAL-77
chr1
201592869
-201592810
Dow
nCS
RP1
(minus116873
)N
AV1
(minus24610
)U
pN
AV1
NM020443
ESC
CAL-76
chr1
90091283
-90091342
Dow
nLR
RC8
C(minus7331
)LR
RC8
B(+100916
)U
pLR
RC8
CN
M032270
ESC
CAL-209
chr10
14549236
-14549177
Dow
nFR
MD4
A(minus176341
)CD
NF
(+330776
)D
own
Dow
n
Dow
nD
own
Dow
n
CDN
FN
M001029954
ESC
CAL-284
chr10
44848467
-44848526
P1(minus562632
)CX
CL12
(+32048
)CX
CL12
NM199168
ESC
CAL-256
chr11
117671760
-117671701
Dow
nD
SCA
ML1
(minus3755
)D
SCA
ML1
NM020693
ESC
CAL-254
chr11
126219961
-126220020
Dow
nST3
GA
L4(minus5549
)D
CPS
(+46344
)ST3
GA
L4N
M006278
ESC
CAL-307
chr11
14975924
-14975983
Dow
nCY
P2R1
(minus62203
)CA
LCA
(+17878
)CY
P2R1
NM024514
ESC
CAL-105
chr11
17366661
-17366720
Dow
nB7
H6
(minus6618
)N
UCB2
(+68405
)D
own
NU
CB2
NM005013
ESC
CAL-232
chr12
131245982
-131245923
Dow
nRI
MBP2
(minus243491
)ST
X2(+77858
)D
own
RIM
BP2
NM015347
ESC
CAL-24
chr12
132824534
-132824475
Dow
nG
ALN
T9(+81400
)N
OC4
L(+195512
)U
pG
ALN
T9N
M021808
Dow
nH
NRN
PA3
)D
own
BARH
L
Dow
nA
LOX
Dow
nEN
DO
VN
MN
M
8 International Journal of Genomics
Table2Con
tinued
LncR
NA
nam
eG
enom
ic co
ordi
nate
sLn
cRN
Aex
pres
sion
LncR
NA
asso
ciat
edge
nes
Gen
eexp
ress
ion
Gen
esym
bol
Gen
bank
acce
ssio
nES
CCA
L-122
chr12
2952238
-2952297
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
n
Dow
n
Dow
n
Dow
nD
own
Dow
n
Dow
n
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
NRI
P 2(minus8047
)FO
XM1
(+34053
)FO
XM1
NM202002
ESC
CAL-338
chr12
32101291
-32101232
BICD1
(minus158923
)H3
F3C
( minus156087
)BI
CD1
NM001714
ESC
CAL-275
chr14
71180731
-71180790
TTC9
(+72257
)M
AP3
K9(+95127
)M
AP3
K 9N
M033141
ESC
CAL-275
chr 14
71180731
-71180790
TTC9
(+72257
)M
AP 3
K 9(+95127
)TT
C9N
M015351
ESC
CAL-121
chr15
78088340
-78088281
LIN
GO1
( minus163602
)TB
C1D2
B(+281683
)LI
NG
O1
NM032808
ESC
CAL-231
chr 16
88381397
-88381338
ZNF469
( minus112511
)BA
NP
(+396330
)ZN
F469
NM001127464
ESC
CAL-74
chr2
47548478
-47548537
CALM2
(minus144768
)EP
CAM
(minus47779
)EP
CAM
NM002354
ESC
CAL-152
chr20
56839403
-56839344
PMEP
A1(minus554343
)PP
P 4R1
L(+45121
)PM
EPA1
NM020182
ESC
CAL-337
chr 3
171506465
-171506406
TNIK
(minus328239
)PL
D1
(+22068
)PL
D1
NM002662
ESC
CAL-80
chr3
177935638
- 177935579
KCN
MB2
(minus318615
)KC
NM
B 2N
M181361
ESC
CAL-90
chr 3
195441846
-195441787
MU
C20
(minus5936
)SD
HA
P 2(+56907
)M
UC20
NM001098516
ESC
CAL-59
chr 3
64855048
-64855107
AD
AM
TS9
(minus181713
)A
DA
MTS9
NM182920
ESC
CAL-72
chr 4
74922502
-74922443
CXCL3
(minus17983
)CX
CL2
(+42524
)CX
CL3
NM002090
ESC
CAL-79
chr4
79626460
- 79626401
BMP2
K(minus71101
)A
NX
A3
(+153689
)A
NX
A3
NM005139
ESC
CAL-253
chr4
8357097
-8357038
ACO
X3(+85384
)H
TRA3
(+85579
)AC
OX 3
NM003501
ESC
CAL-179
chr4
8359416
- 8359357
ACO
X3(+83065
)H
TRA3
(+87898
)H
TRA3
NM053044
ESC
CAL-11
chr4
84299039
- 84299098
HPS
E(minus43035
)H
ELQ
(+77956
)H
PSE
NM006665
ESC
CAL-41
chr4
8512901
- 8512960
GPR78
(minus69286
)M
ETTL19
(+70399
)G
PR78
NM080819
ESC
CAL-353
chr 5
1175322
-1175263
SLC12
A7
(minus63121
)SL
C6A19
(minus26417
)SL
C12
A7
NM006598
ESC
CAL-260
chr5
131808618
-131808677
IRF 1
(+17817
)SL
C22
A5
(+103247
)IR
F 1N
M002198
ESC
CAL-132
chr5
134578804
- 134578863
PITX1
(minus208870
)H2
AFY
(+156094
)PI
TX1
NM002653
ESC
CAL-4
chr5
141732627
-141732568
SPRY4
(minus27978
)FG
F1(+333055
)SP
RY4
NM030964
ESC
CAL-356
chr 5
1545355
-1545296
LPCA
T1( minus21250
)M
RPL36
(+254630
)LP
CAT 1
NM024830
XLO
C004881
chr5
72570680
-72570621
TMEM174
(+101628
)FO
XD1
(+173701
)up
FOXD1
NM004472
ESC
CAL-36
chr6
106899513
-106899572
ATG5
( minus125848
)A
IM1
(minus59762
)A
IM1
NM001624
XLO
C005849
chr6
138145112
-138145053
OLI
G3
( minus329552
)TN
FAIP
3(minus43498
)TN
FAIP3
NM006290
ESC
CAL-262
chr6
2283830
-2283771
GM
DS
(minus37933
)M
YLK4
(+467353
)G
MD
SN
M001500
ESC
CAL-120
chr6
29716817
-29716758
LOC554223
(minus42895
)H
LA-F
(+25671
)H
LA-F
NM018950
ESC
CAL-257
chr 6
29988410
-29988469
ZNRD1
( minus40596
)H
LA-J
(+14223
)H
LA-J
NR024240
ESC
CAL-73
chr6
36126663
-36126722
BRPF3
( minus37857
)M
APK13
(+28431
)M
APK13
NM002754
ESC
CAL-97
chr 6
40305634
-40305575
MO
CS1
(minus410150
)LR
FN2
(+249521
)LR
FN2
NM020737
ESC
CAL-333
chr 6
72018004
-72018063
RIM
S1(minus578616
)O
GFR
L1(+19557
)O
GFR
L1N
M024576
ESC
CAL-115
chr7
12593405
-12593464
VW
DE
(minus149583
)SC
IN(minus16768
)SC
INN
M033128
ESC
CAL-68
chr7
139487166
-139487225
TBX
AS1
(minus41756
)H
IPK2
(minus9503
)H
IPK2
AF207702
ESC
CAL-87
chr7
19958800
- 19958741
TMEM196
(minus146367
)M
ACC1
(+298242
)M
ACC1
NM182762
ESC
CAL-58
chr8
16355082
-16355141
MSR1
(minus304812
)FG
F 20
(+504562
)M
SR1
NM002445
ESC
CAL-130
chr8
37189082
-37189141
ZNF703
(minus364189
)KC
NU1
(+547270
)ZN
F703
NM025069
ESC
CAL-199
8
chr 8
38623612
-38623671
RNF 5
P 1(minus164867
)TA
CC 1
(minus21080
)TA
CC1
BC041391
ESC
CAL-
chr9
34668311
-34668252
CCL27
(minus5593
)C
CL19
(+22992
)C
CL19
NM006274
NotesR
ed-highlighted
genesa
rewho
seexpressio
nsaresig
nificantly
changedin
Esop
hagealSquamou
sCellC
arcino
ma(ESC
C)G
reen
color-high
lighted
rowsgenesinvolvedin
lipid
metabolism
predictedwith
GOenric
hmentanalysis
Yellowcolor-high
lighted
rowsTh
eexpressionof
lncR
NAES
CCAL-5was
valid
ated
byPC
R
International Journal of Genomics 9
Final extension was at 72∘C for 10 minutes The ampliconswere resolved in 2 agarose gel
34 Bioinformatic Analysis Intensity data were exportedto GeneSpring 120 (Agilent Technologies Santa ClaraCA USA) for quantile normalization and the analysis ofdifferentially expressed long noncoding RNAs and codingRNAs Paired 119905-test analysis was used to obtain probe setswhose magnitude of change in expression of RNAs betweenESCC tissue and adjacent normal esophageal tissuewas eithergreater or less than 20 fold and 119875 value lt 005 (119875 values werecorrected formultiple testing using themethod of Benjamini-Hochberg) The normalized data containing 42544 probeswere further analyzed using the R program All controlprobes were removed We then defined the coding (ldquoNM rdquoldquoXM rdquo) and noncoding (ldquolincRNArdquo ldquoNR rdquo and ldquoXR rdquo) genesin the normalized data according to the definition of RefSeqaccession format (httpwwwncbinlmnihgovprojectsRefSeqkeyhtml) Differentially expressed long noncodingRNAs (DE-lncRNAs) and coding RNAs (DE-mRNAs)were further identified The landscapes of the wholetranscriptome (lncRNAs + coding RNAs) or all lncRNAsor all coding RNAs were analyzed with gene expressiondynamic inspector (GEDI)
35 Co-Location and Co-Expression Analysis between DE-lncRNAs and DE-mRNAs Genomic coordinates of DE-lncRNAs were imported to GREAT software (httpbejeranostanfordedugreatpublichtmlindexphp) for co-locationanalysis Neighboring coding genes were then matched withDE-mRNAs to obtain a co-expression dataset Three sub-groups of genes (DE-lncRNA co-located genes DE-mRNAsand co-expressed genes) were used for gene expressionnetwork analysis using Cytoscape software (v283)
Abbreviations
AFAP1-AS1 Actin filament-associated protein 1 antisenseRNA
Wei Cao and Wei Wu contributed equally to this project
Acknowledgments
The National Natural Science Foundation of China (no81171992) and the Zhengzhou Science and Technology
Programme (121PPTGG494-8) supported this work Theauthors thank Anne Haegert atThe Laboratory for AdvancedGenome Analysis at the Vancouver Prostate Centre Van-couver Canada for expert technical support They alsothank GenomeSky Inc Canada for providing computationalconsultation
References
[1] J Ferlay H R Shin F Bray D Forman C Mathers and DM Parkin ldquoEstimates of worldwide burden of cancer in 2008GLOBOCAN2008rdquo International Journal of Cancer vol 127 no12 pp 2893ndash2917 2010
[2] W Wu and J A Chan ldquoUnderstanding the role of longnoncoding RNAs in the cancer genomerdquo in Next GenerationSequencing in Cancer Research-Decoding Cancer Genome WWu and H Choudhry Eds pp 199ndash215 Springer New YorkNY USA 2013
[3] J J Hao T Gong Y Zhang et al ldquoCharacterization of generearrangements resulted from genomic structural aberrationsin human esophageal squamous cell carcinoma KYSE150 cellsrdquoGene vol 513 no 1 pp 196ndash201 2013
[4] M Kano N Seki N Kikkawa et al ldquoMiR-145 miR-133aand miR-133b tumor-suppressive miRNAs target FSCN1 inesophageal squamous cell carcinomardquo International Journal ofCancer vol 127 no 12 pp 2804ndash2814 2010
[5] H SuNHuHH Yang et al ldquoGlobal gene expression profilingand validation in esophageal squamous cell carcinoma and itsassociation with clinical phenotypesrdquo Clinical Cancer Researchvol 17 no 9 pp 2955ndash2966 2011
[6] D M Greenawalt C Duong G K Smyth et al ldquoGeneexpression profiling of esophageal cancer comparative analysisof Barrettrsquos esophagus adenocarcinoma and squamous cellcarcinomardquo International Journal of Cancer vol 120 no 9 pp1914ndash1921 2007
[7] I Dunham A Kundaje S F Aldred et al ldquoAn integratedencyclopedia of DNA elements in the human genomerdquo Naturevol 489 no 7414 pp 57ndash74 2012
[8] E A Gibb C J Brown and W L Lam ldquoThe functional roleof long non-coding RNA in human carcinomasrdquo MolecularCancer vol 10 article 38 2011
[9] A L Brunner A H Beck B Edris et al ldquoTranscriptionalprofiling of lncRNAs and novel transcribed regions across adiverse panel of archived human cancersrdquo Genome Biology vol13 no 8 article R75 2012
[10] M Guttman and J L Rinn ldquoModular regulatory principles oflarge non-coding RNAsrdquo Nature vol 482 no 7385 pp 339ndash346 2012
[11] K C Wang Y W Yang B Liu et al ldquoA long noncodingRNA maintains active chromatin to coordinate homeotic geneexpressionrdquo Nature vol 472 no 7341 pp 120ndash126 2011
[12] A Dean ldquoOn a chromosome far far away LCRs and geneexpressionrdquo Trends in Genetics vol 22 no 1 pp 38ndash45 2006
[13] P Ji S Diederichs W Wang et al ldquoMALAT-1 a novel noncod-ing RNA and thymosin 1205734 predict metastasis and survival inearly-stage non-small cell lung cancerrdquo Oncogene vol 22 no39 pp 8031ndash8041 2003
[14] R A Gupta N Shah K C Wang et al ldquoLong non-codingRNA HOTAIR reprograms chromatin state to promote cancermetastasisrdquo Nature vol 464 no 7291 pp 1071ndash1076 2010
10 International Journal of Genomics
[15] R Kogo T Shimamura K Mimori et al ldquoLong noncodingRNAHOTAIR regulates polycomb-dependent chromatinmod-ification and is associated with poor prognosis in colorectalcancersrdquo Cancer Research vol 71 no 20 pp 6320ndash6326 2011
[16] K Kim I Jutooru G Chadalapaka et al ldquoHOTAIR is anegative prognostic factor and exhibits pro-oncogenic activityin pancreatic cancerrdquo Oncogene vol 32 pp 1616ndash1625 2013
[17] D Li J Feng T Wu et al ldquoLong intergenic noncoding RNAHOTAIR is overexpressed and regulates PTEN methylation inlaryngeal squamous cell carcinomardquo The American Journal ofPathology vol 182 no 1 pp 64ndash70 2013
[18] A Bhan I Hussain K I Ansari S Kasiri A Bashyal and S SMandal ldquoAntisense transcript long noncoding RNA (lncRNA)HOTAIR is transcriptionally induced by estradiolrdquo Journal ofMolecular Biology vol 425 no 19 pp 3707ndash3722 2013
[19] D Cejka D Losert and V Wacheck ldquoShort interfering RNA(siRNA) tool or therapeuticrdquo Clinical Science vol 110 no 1pp 47ndash58 2006
[20] J R Prensner and A M Chinnaiyan ldquoThe emergence oflncRNAs in cancer biologyrdquo Cancer Discovery vol 1 pp 391ndash407 2011
[21] W Wu T D Bhagat X Yang et al ldquoHypomethylation ofnoncoding DNA regions and overexpression of the long non-coding RNA AFAP1-AS1 in Barrettrsquos esophagus and esophagealadenocarcinomardquoGastroenterology vol 144 no 5 pp 956ndash9662013
[22] E Ozgur U Mert M Isin M Okutan N Dalay and U GezerldquoDifferential expression of long non-coding RNAs duringgenotoxic stress-induced apoptosis in HeLa and MCF-7 cellsrdquoClinical and Experimental Medicine vol 13 no 2 pp 119ndash1262013
[23] J D Li Q C Feng and J S Li ldquoDifferential gene expressionprofiling of oesophageal squamous cell carcinoma by dnamicroarray and bioinformatics analysisrdquo Journal of Interna-tional Medical Research vol 38 no 6 pp 1904ndash1912 2010
[24] S Ma J Y Bao P S Kwan et al ldquoIdentification of PTK6via RNA sequencing analysis as a suppressor of esophagealsquamous cell carcinomardquo Gastroenterology vol 143 pp 675ndash686 2012
[25] M Tong K W Chan J Y Bao et al ldquoRab25 is a tumorsuppressor gene with antiangiogenic and anti-invasive activitiesin esophageal squamous cell carcinomardquo Cancer Research vol72 no 22 pp 6024ndash6035 2012
[26] Y J Geng S L Xie Q Li J Ma and G Y Wang ldquoLargeintervening non-coding RNA HOTAIR is associated withhepatocellular carcinoma progressionrdquo Journal of InternationalMedical Research vol 39 no 6 pp 2119ndash2128 2011
[27] X Zhou T J Lawrence Z He C R Pound J Mao andS A Bigler ldquoThe expression level of lysophosphatidylcholineacyltransferase 1 (LPCAT1) correlates to the progression ofprostate cancerrdquo Experimental andMolecular Pathology vol 92no 1 pp 105ndash110 2012
[28] Y Morita T Sakaguchi K Ikegami et al ldquoLysophosphatidyl-choline acyltransferase 1 altered phospholipid composition andregulated hepatoma progressionrdquo Journal of Hepatology vol 59no 2 pp 292ndash299 2013
of disease including cancer [10] Mechanisms of action oftranscribed lncRNAs are described as modifying chromatinarchitecture and regulating gene expression in a cis or transmanner For example H19 lncRNA cis-regulates IGF2 geneexpression at the same genomic locus and HOTAIR lncRNAis transcribed on Chr 12 and it transregulates HoxD geneon Chr 2 Additionally lncRNAs have also been reportedto coordinate the regulation of neighboring coding genesthrough a ldquolocus controlrdquo process [11] which mediates thelocalization of genes within nuclear regions to favor theirtranscription through the formation of domains of histonemodification and intra- or interchromosomal loops [12]Dysregulated lncRNAs have been identified with differentscreening methodologies in various types of cancer Forexample the cancer-related lncRNA metastasis-associatedlung adenocarcinoma transcript 1 (MALAT-1) was identifiedby subtractive hybridization during screening for early non-small cell lung cancer with metastasis [13] Overexpression ofMALAT-1 is highly predictive of poor prognosis and short-ened survival time in early stage lung cancer OverexpressionofHOTAIR lncRNAwas found in several solid tumors [14ndash17]in association with cancermetastasis and increasedHOTAIRexpression in breast cancer is transcriptionally induced byestradiol [18] Prostate cancer associated lncRNA PCGEM1[19] andPCAT-1 [20] appear to be prostate-specific regulatorsof cell apoptosis and proliferation Recently AFAP1-AS1lncRNA was reported to be overexpressed in esophagealadenocarcinoma [21]
The handful of dysregulated lncRNAs in different cancerssuggests that lncRNAs are an enigmatic component of thewhole transcriptome which may participate in tumorige-nesis invasion and metastasis Efforts are being made toexplore the ldquolncRNAomerdquo of various cancers with advancedhigh-throughput RNA sequencing technologies [8 9] anddynamic changes in lncRNA expression have been observedin cancer cells during different stages of cancer developmentand during treatment [22] However our understanding ofthe role of lncRNAs in cancer biology is still in an early stageand a clearly defined predictive set of biological functionsfor lncRNAs is lacking in cancer biologyTherefore thoroughsearches and analyses of the interactions between lncRNAand coding genes may help to infer their potential biologicalroles
In order to understand the role of lncRNAs in ESCC wereport a pilot study of the profiles of differentially expressedlncRNAs and coding RNAs from tumor and adjacent normaltissue of individual patients with ESCC We assessed thewhole transcriptomic landscape for potential interactionsbetween lncRNAs and coding-gene expression In particularwe evaluated the coding genes that are co-located and co-expressed with the differentially expressed lncRNAs duringthe genesis of ESCC
2 Results and Discussion
21 Transcriptomic Landscape of ESCC Our genome-widegene expression profiling of both lncRNAs and coding genesfrom ESCC and adjacent nonneoplastic tissue was conducted
to detect possible associations of lncRNAs with ESCC Wefirst asked whether these transcripts of 7419 noncoding and27958 coding RNAs could distinguish ESCC from normaltissues Figure 1(a) shows that the four ESCC samples areclustered together in one group and clearly separated fromthe samples of normal tissue Next we examined the wholetranscriptomic pattern (lncRNAs + coding RNAs) fromeach sample and the landscapes of the whole transcriptome(represented by heatmaps in Figure 1(a)) of normal tissuesdiffer from those of ESCC that exhibit more heterogeneousalterations The overall changes from a respective normalto cancer state were also seen separately as a differencein expression profile of either the lncRNA or the codingRNA These observations suggest that a potential dynamicinteraction between lncRNAs and coding RNAs may bereshaping the landscape of the whole transcriptome duringESCC development
To gain a detailed understanding of the biological themesof all RNA transcripts we further identified those transcriptsthat are significantly and differentially expressed (DE) inESCC tissue compared to matched normal tissue based onthe criteria described in the methods There are 410 DE-lncRNAs and 1219 DE-mRNAs that represent about 5 ofthe transcripts in the respective microarrays (SupplementaryTables S1 and S2 available online at httpdxdoiorg1011552013480534) DE-lncRAs distinguish a cancer cell from itsnormal cell state with three times fewer transcripts thanDE-mRNAs (Figures 1(b) and 1(c)) suggesting that the DE-lncRNA profile is more informative and potentially a morefaithful indicator of a specific cell state
Enrichment analysis of DE-mRNAs demonstrated thatthe respective genes are involved in cancer-related pathways(Figure 1(d)) Since expression profiling of coding-RNA hasbeen intensively studied in esophageal cancer we validated10 genes whose expression level in other studies [23ndash25] issignificantly changed (119875 lt 005) by at least 2-fold relative tonormal tissues (Table 1)
22 Expression of lncRNAs in ESCC LncRNAs are emergingas a novel class of noncoding RNAs that are pervasivelytranscribed in the genome but there is limited functionalknowledge about them High-throughput screening of lncR-NAs from ESCC has been poorly studied except for a recentreport of overexpressed lncRNA AFAP1-AS1 in esophagealadenocarcinoma [21] In our study a total of 7419 intergeniclncRNAs and other transcripts of uncertain coding potentialwere examined and we identified 410 DE-lncRNAs in ESCCrelative to adjacent normal esophageal tissues We namedthe anonymous lncRNAs ESCC Associated Long noncodingRNAs (ESCCAL Supplementary Table S1) Expression ofHOTAIR lncRNA is increased in various cancers [14ndash1726] and it is also significantly increased in our analysisof ESCC (Figure 2(a)) In addition we confirmed anothertwo upregulated lncRNAs that are differentially expressed inESCC and that we have named ESCCAL-1 and ESCCAL-5 The increased and differential expression of HOTAIRESCCAL-1 and ESCCAL-5 in ESCC tissue relative to adja-cent nonneoplastic tissue was independently assessed with
International Journal of Genomics 3
Wholetranscriptome
lncRNAs mRNAs
3N
4N
1N 2N
1T
3T
2T 4T
(a)
3N4N
1N2N
1T3T2T 4T
DE-
lncR
NA
s
minus2 0 +2
(b)
DE-
mRN
As
3N4N
1N 2N 1T 3T
2T4T
minus2 0 +2
(c)
Pathways in cancer
Pancreatic cancer
Renal cell carcinomaErbB signaling pathway
Wnt signaling pathwayMAPK signaling pathway
Nonsmall cell lung cancer
p53 signaling pathway
(d)
Figure 1 Transcriptomic landscape of esophageal squamous cell cancer (ESCC) (a) Whole transcriptome of tumor (T) and adjacent normaltissue (N) of four patients with ESCCwere detected using a microarray with 7419 long noncoding RNAs (lncRNAs) and 27958 coding RNAsTwo main clusters (Ts and Ns) were generated using unsupervised clustering methods Then a self-organizing map (SOM) of either wholetranscriptome (both lncRNAs and mRNAs) or lncRNAs or mRNA was produced from each sample (see legend in up-left corner of thisfigure and the arrows are meant to indicate the potential interaction) using gene expression dynamic inspector (GEDI) Mosaic patternsare pseudocolored SOMs to show integrated biological entity in each sample Red through blue color indicates high to low expression level(b) and (c) Differentially expressed lncRNAs (DE-lncRNAs) and coding RNAs (DE-mRNAs) in ESCC Hierarchical clustering analysis of 410DE-lncRNAs (b) and 1219 DE-mRNAs (c) between ESCC tissue and adjacent normal tissue (fold change gt or lt 2-fold and 119875 lt 005) Redand green colors indicate high and low expression respectively In the heatmap columns represent samples and rows represent each geneThe scale of expression level is shown on the horizontal bar (d) KEGG functional analysis of DE-mRNA networks in ESCCThe DE-mRNAgenes are involved in cancer-related signaling functions and a detailed list of significant GO terms is shown in Figure S1 and its associatedlegend in Supplementary Information
4 International Journal of Genomics
100
80
60
40
20
0N T N T N T
Nor
mal
ized
inte
nsity
in m
icro
arra
y
Nor
mal
ized
inte
nsity
Nor
mal
ized
inte
nsity
in m
icro
arra
y
in m
icro
arra
y
HOTAIRlowast
lowastlowast
lowastlowast
ESCCAL-1 ESCCAL-52500
2000
1500
1000
500
0
700
600
500
400
300
200
100
0
(a)
N T N T N T N T N T N T N T N T N T N T N T N T
GA
PDH
HO
TAIR
Patient 1 Patient 2 Patient 3
ESCC
AL-
1
ESCC
AL-
5
GA
PDH
HO
TAIR
ESCC
AL-
1
ESCC
AL-
5
GA
PDH
DN
A m
arke
r
HO
TAIR
ESCC
AL-
1
ESCC
AL-
5(b)
Figure 2 Long noncoding RNAs (lncRNAs) expression in esophageal squamous cell carcinoma (ESCC) (a) Three differentially expressedlncRNAs HOTAIR ESCCAL-1 and ESCCAL-5 from microarray detection The average intensity of expression in normal tissues (N) andtumors (T) is plotted with their standard deviations (b) Validation ofHOTAIR ESCCAL-1 and ESCCAL-5with independent patient samplesby PCR analysis The amplicons were separated with 2 agarose gel GAPDH was used as an internal control Significance is lowast119875 lt 005lowastlowast
119875 lt 001
Table 1 Validation of selected differential expression of mRNAs in esophageal squamous cell carcinoma in independent studies
Probe name P value FC Regulation Gene symbol Genbank accession Independent study ReferenceA 33 P3232692 0005838984 8301461 Up IL24 NM 001185156
Microarray [20]
A 24 P411121 000055 5329484 Up TNFRSF18 NM 148901A 23 P169097 881119864 minus 05 4466178 Up WISP1 NM 080838A 23 P304304 0004822649 3944957 Down ARSF NM 004042A 24 P56363 0003573538 3323955 Down CAB39L NM 030925A 23 P419760 0001041661 3270335 Down CRISP3 NM 006061A 23 P413923 0002921898 454899 Down DMRTA1 NM 022160A 23 P56978 0002093997 5438183 Down PTK6 NM 005975 RNA-seq [21]A 23 P115091 0005171322 3289834 Down RAB25 NM 020387 Q-RT-PCR [22]A 33 P3258542 0001039129 2036035 Down SPINK8 NM 001080525 Microarray [20]
International Journal of Genomics 5
PCR methods in matched-pair tissue samples from threeadditional ESCC patients and the results are consistent withthemicroarray analysis (Figure 2(b)) Interestingly except forHOTAIR other previously reported lncRNAs (ie MALAT-1 PCAT-1 and AFAP1-AS1) are not differentially expressed inour analysis of ESCC Therefore the DE-lncRNAs that wehave identifiedmay be a unique property of ESCC andwe arecurrently using a population-based analysis to characterizethese DE-lncRNAs as potential genomic biomarkers andregulatory elements in the dynamic process leading to ESCC
23 LncRNAs Co-located and Co-expressed with Coding Genesin ESCC LncRNAs have been reported to coordinate theregulation of neighboring coding genes through a ldquolocuscontrolrdquo process [11] We wondered whether such a ldquolocuscontrolrdquo process could operate in ESCC development andtherefore we searched neighboring genes of the 410 DE-lncRNAs in the genomeThemajority (988) of the 410 DE-lncRNAs harbor neighboring coding genes whose genomiclocations are within sim5 kb upstream and sim1 kb downstreamof the lncRNA and may extend to 1000 kb in both directions(Figure 3(a)) Interrogation of 538 coding genes that areneighbors of these DE-lncRNAs (DE-lncRNAs co-locatedgenes) revealed predicted functions in 9 common pathwayssuch as the AP1 transcription factor network integrin-linkedkinase signaling several signaling pathways in adherensjunctions and FOXO family signaling (Figure 3(b))
We asked whether any DE-lncRNAs co-located genes arealso differentially expressed in ESCC Analysis of the DE-lncRNAs co-located genes withDE-mRNAdata set identified76 genomically co-located and differentially co-expressedgenes (Figure 3(c) and Table 2) Strikingly the co-locatedand co-expressed genes with DE-lncRNAs may be involvedin ether lipid metabolism pathways by the participation ofthe LPCAT1 gene encoding lysophosphatidylcholine acyl-transferase1 and the PLD1 gene encoding phospholipase D1(Figure 3(c)) The lncRNA ESCCAL-337 (chr3171506370-171528740) was downregulated in ESCC and located at22068 bp downstream of the PLD1 gene whose expres-sion was also decreased in ESCC In contrast the lncRNAESCCAL-356 (chr51544500-1567142 reverse strand) wasdownregulated in ESCC and located at 21250 bp upstreamof LPCAT1 whose expression was upregulated in ESCC(Figure 3(c)) LPCAT1 modulates phospholipid compositionby catalyzing lysophosphatidylcholine into phosphatidyl-choline and overexpression of LPCAT1was reported to createfavorable conditions for cancer cell proliferation [27 28]Therefore at least two of the DE-lncRNAs have the potentialto contribute to ESCC by a ldquolocus controlrdquo process withneighboring coding genes
In conclusion we performed a genome-wide survey ofthe expression of lncRNAs and coding mRNAs from pairedsamples of primary neoplastic tissue and adjacent non-neoplastic normal tissue from four individuals The overalltranscriptomic landscape (both lncRNAs andmRNAs) is ableto distinguish malignant from normal tissue in each personWe discovered a set of differentially expressed lncRNAsand their co-located and co-expressed coding mRNAs and
demonstrated that lncRNAs may be involved in ether lipidmetabolism in ESCC Our study provides genomic supportfor a model of a ldquolocus controlrdquo process in ESCC and aframework for further experimental study
3 Materials and Methods
31 Specimens Written informed consent was obtained frompatients before surgery and the study protocol was approvedby the Institutional Review Board for the use of human sub-jects at Zhengzhou Hospital Primary tumors and adjacentnonneoplastic tissues were obtained frompatients with ESCCwhounderwent surgical treatment at LinxianHospital inMay2012 All tissues were frozen in liquid nitrogen immediatelyafter surgical resection None of the patients had priorchemotherapy or radiotherapy nor did they have any otherserious diseases All ESCC tissues were histopathologicallydiagnosed by at least two independent senior pathologists
32 Microarray Hybridization Total RNAs were extractedusing Trizol reagent following manufacturerrsquos instructions(Invitrogen Carlsbad CA USA) The quality of RNAswas measured with a 2100 Bioanalyzer (Agilent technologyUSA) Input of 100 ng of total RNA was used to generateCyanine-3 labeled cRNA according to theAgilentOne-ColorMicroarray-Based Gene Expression Analysis Low for InputQuick Amp Labeling kit (v60) Samples were hybridizedon Agilent SurePrint G3 Human GE 8 times 60K Microarray(Design ID 028004) Arrays were scanned with the AgilentDNAMicroarray Scanner at a 3120583m scan resolution and datawere processed with Agilent Feature Extraction 11011 Themicroarray data discussed in this article have been depositedin National Center for Biotechnology Information (NCBI)Gene ExpressionOmnibus (GEO) and are accessible through(GEO) Series accession number GSE45350 (httpwwwncbinlmnihgovgeoqueryacccgiacc=GSE45350)
33 Validation by Polymerase Chain Reaction (PCR) PCRanalysis was performed on additional matched ESCC andadjacent nonneoplastic tissues for selected lncRNAs Theprimer sequences for PCR are as follows HOTAIR forward51015840-GGTAGAAAAAGCAACCACGAAGC-31015840 and reverse51015840-ACATAAACCTCTGTCTGTGAGTGCC-31015840 ESSCAL-1(chr876121095-76189420 reverse strand) forward 51015840-CCA-GACAGCAGCAAAGCAAT-31015840 and reverse 51015840-GGAAGC-AGCAAATGTGTCCAT-31015840 ESSCAL-5 (chr2216585154-216585719 forward strand) forward 51015840-TACCAACATTGT-CCACCGGG-31015840 and reverse 51015840-GCTGATGACAGTCCC-TTGCT-31015840 GAPDH was used as a control forward 51015840-CCG-GGAAACTGTGGCGTGATGG-31015840 and reverse 51015840-AGG-TGGAGGAGTGGGTGTCGCTGTT-31015840 The thermocycleconditions are as follows initial denaturation at 95∘C for10 minutes followed by 94∘C for 45 seconds 65∘C for 30seconds and 72∘C for 1 minute for 15 cycles Then theannealing temperature was reduced by 05∘Ccycle for thenext 14 cycles and the amplification was finished withanother 24 cycles with the annealing temperature at 58∘C
6 International Journal of Genomics
lncRNAs
Gene b Gene c
20
40
60
80
100
0lncRNAs
Gene a
lncRNAs
Number of associated genes per region0 1 2G
enom
ic re
gion
s of l
ncRN
As (
)
(a)
AP-1 transcription factor network
Integrin-linked kinase signalingRegulation of CDC42 activity
Posttranslational regulation of adherens junction stability and disassemblyN-cadherin signaling events
E-cadherin signaling in the nascent adherens junctionStabilization and expansion of the E-cadherin adherens junction
E-cadherin signaling eventsFOXO family signaling
1176
1105904
511466
448448440
422
(b)
DE-lncRNAsco-located genes
DE-mRNAs
538 330776
ESCCAL-356
ESCCAL-337
Chr3
Chr5
Lipid metabolism
PLD1
LPCAT1
(c)
Figure 3 Identification of lncRNAs co-located and co-expressed neighboring genes in esophageal squamous cell carcinoma (ESCC)(a) Identification of neighboring genes of the DE-lncRNAs The genomic coordinate information of 410 DE-lncRNAs was used to searchneighboring genes whose genomic locations are within sim5 kb upstream and sim1 kb downstream of the lncRNA and may extend to 1000 kbin both directions using GREAT software (httpbejeranostanfordedugreatpublichtmlindexphp) The percentage of DE-lncRNAsharboring zero one or two neighboring genes is presented (b) Gene Ontology (GO) enrichment analysis of lncRNAs co-located genesIdentified gene enriched pathwaysterms are listed on the left the length of horizontal bars and the numbers on the right indicate thepercentage of genes involved in each pathwayterm (c) LncRNAs co-located and co-expressed coding mRNAs Overlap of 538 DE-lncRNAco-located genes with 3307 DE-mRNAs in microarrays identified 76 lncRNAs co-located and co-expressed coding mRNAs (list in Table 2)GO enrichment analysis suggests phospholipase D1 (PLD1) and lysophosphatidylcholine acyltransferase1 (LPCAT1) are involved in ether lipidmetabolism pathway Genomic location shows that PLD1 is located at minus22068 bp upstream of ESCCAL-337 lncRNA on Chr 3 and LPCAT1is at minus21250 bp upstream of ESCCAL-356 lncRNA on Chr 5
International Journal of Genomics 7
Table2Listof
identifi
edco-lo
catedandco-expressed
genesw
ithdifferentially
expressedlncR
NAsinES
CC
LncR
NA
nam
eG
enom
ic co
ordi
nate
Ln
cRN
Aex
pres
sion
LncR
NA
asso
ciat
edge
nes
Gen
eexp
ress
ion
Gen
esym
bol
Gen
bank
acce
ssio
nES
CCA
L-177
chr1
205404138
-205404079
Up
CDK18
(minus69575
)LE
MD
1(minus12895
)U
pLE
MD1
NM001001552
ESC
CAL-348
chr1
222587441
-222587382
Up
DU
SP10
(minus671896
)H
HIP
L2(+134032
)U
pD
USP10
NM007207
ESC
CAL-31
chr1
225237693
-225237752
Up
DN
AH14
(+120367
)LB
R(+378092
)U
pD
NA
H14
NM001373
ESC
CAL-159
chr1
225240300
-225240359
Up
DN
AH14
(+122974
)LB
R(+375485
)U
pD
NA
H14
NM001373
ESC
CAL-327
chr1
89887111
-89887052
Up
LRRC8
B(minus103315
)G
BP6
(+57646
)D
own
GBP6
NM198460
XLO
C000915
chr1
91295528
-91295469
Up
BARH
L2(minus112705
)ZN
F644
(+192313
2N
M020063
ESC
CAL-65
chr11
2017146
-2017205
Up
MRP
L23
(+48674
)IG
F2(+145165
)U
pIG
F2N
M000612
ESC
CAL-19
chr12
66204406
-66204347
Up
HM
GA
2(minus13863
)M
SRB3
(+531610
)U
pH
MG
A2
NM003484
XLO
C011548
chr15
81953024
-81952965
Up
TMC3
(minus286577
)M
EX3
B(+385366
)U
pM
EX3
BN
M032246
ESC
CAL-102
chr17
6766871
-6766930
Up
ALO
X12
(minus132483
)TE
KT1
(minus31841
)12
NM000697
ESC
CAL-342
chr17
78302158
-78302217
Up
END
OV
(minus86779
)RN
F213
(+67521
)U
pRN
F213
NM020954
ESC
CAL-342
chr17
78302158
-78302217
Up
END
OV
(minus86779
)RN
F213
(+67521
)001164638
ESC
CAL-99
chr2
102034125
-102034066
Up
CREG
2(minus30131
)RF
X8(+57069
)U
pCR
EG2
NM153836
ESC
CAL-5
chr2
216585455
-216585514
Up
FN1
(minus284694
)M
REG
(+292861
)D
own
MRE
GN
M018000
ESC
CAL-5
chr2
216585455
-216585514
Up
FN1
(minus284694
)M
REG
(+292861
)U
pFN1
NM054034
ESC
CAL-288
chr2
27789661
-27789602
Up
ZNF512
(minus16261
)G
CKR
(+69926
)D
own
GCK
RN
M001486
ESC
CAL-344
chr2
37327299
-37327358
Up
CCD
C75
(+15735
)EI
F2A
K2
(+56861
)U
pEI
F2A
K2N
M001135652
ESC
CAL-10
chr22
48086469
-48086528
Up
FAM
19A
5(minus798789
)TB
C1D22
A(+927981
)U
pFA
M19
A5
NM015381
ESC
CAL-81
chr5
141710377
-141710436
Up
SPRY4
(minus5787
)FG
F1(+355246
)U
pFG
F1N
M000800
ESC
CAL-81
chr5
141710377
-141710436
Up
SPRY4
(minus5787
)FG
F1(+355246
)U
pSP
RY4
NM030964
ESC
CAL-204
chr6
126699769
-126699828
Up
RSPO3
(minus740249
)CE
NPW
(+38546
)U
pCE
NPW
NM001012507
ESC
CAL-106
chr6
21822910
-21822969
Up
SOX4
(+228968
)PR
L(+480142
)U
pSO
X4N
M003107
ESC
CAL-239
chr8
104258684
-104258625
Up
FZD
6(minus52006
)BA
ALC
(+105734
)U
pBA
ALC
NM024812
ESC
CAL-239
chr8
104258684
-104258625
Up
FZD
6(minus52006
)BA
ALC
(+105734
)U
pFZ
D6
NM003506
ESC
CAL-300
chr1
180918852
-180918793
Dow
nST
X6(+73223
)XP
R1(+317677
)U
pXP
R1N
M004736
XLO
C000515
chr1
200384539
-200384598
Dow
nZN
F281
(minus5403
)KI
F14
(+205293
)U
pKI
F14
NM014875
XLO
C000515
chr1
200384539
-200384598
Dow
nZN
F281
(minus5403
)KI
F14
(+205293
)U
pZN
F281
NM012482
ESC
CAL-77
chr1
201592869
-201592810
Dow
nCS
RP1
(minus116873
)N
AV1
(minus24610
)U
pN
AV1
NM020443
ESC
CAL-76
chr1
90091283
-90091342
Dow
nLR
RC8
C(minus7331
)LR
RC8
B(+100916
)U
pLR
RC8
CN
M032270
ESC
CAL-209
chr10
14549236
-14549177
Dow
nFR
MD4
A(minus176341
)CD
NF
(+330776
)D
own
Dow
n
Dow
nD
own
Dow
n
CDN
FN
M001029954
ESC
CAL-284
chr10
44848467
-44848526
P1(minus562632
)CX
CL12
(+32048
)CX
CL12
NM199168
ESC
CAL-256
chr11
117671760
-117671701
Dow
nD
SCA
ML1
(minus3755
)D
SCA
ML1
NM020693
ESC
CAL-254
chr11
126219961
-126220020
Dow
nST3
GA
L4(minus5549
)D
CPS
(+46344
)ST3
GA
L4N
M006278
ESC
CAL-307
chr11
14975924
-14975983
Dow
nCY
P2R1
(minus62203
)CA
LCA
(+17878
)CY
P2R1
NM024514
ESC
CAL-105
chr11
17366661
-17366720
Dow
nB7
H6
(minus6618
)N
UCB2
(+68405
)D
own
NU
CB2
NM005013
ESC
CAL-232
chr12
131245982
-131245923
Dow
nRI
MBP2
(minus243491
)ST
X2(+77858
)D
own
RIM
BP2
NM015347
ESC
CAL-24
chr12
132824534
-132824475
Dow
nG
ALN
T9(+81400
)N
OC4
L(+195512
)U
pG
ALN
T9N
M021808
Dow
nH
NRN
PA3
)D
own
BARH
L
Dow
nA
LOX
Dow
nEN
DO
VN
MN
M
8 International Journal of Genomics
Table2Con
tinued
LncR
NA
nam
eG
enom
ic co
ordi
nate
sLn
cRN
Aex
pres
sion
LncR
NA
asso
ciat
edge
nes
Gen
eexp
ress
ion
Gen
esym
bol
Gen
bank
acce
ssio
nES
CCA
L-122
chr12
2952238
-2952297
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
n
Dow
n
Dow
n
Dow
nD
own
Dow
n
Dow
n
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
NRI
P 2(minus8047
)FO
XM1
(+34053
)FO
XM1
NM202002
ESC
CAL-338
chr12
32101291
-32101232
BICD1
(minus158923
)H3
F3C
( minus156087
)BI
CD1
NM001714
ESC
CAL-275
chr14
71180731
-71180790
TTC9
(+72257
)M
AP3
K9(+95127
)M
AP3
K 9N
M033141
ESC
CAL-275
chr 14
71180731
-71180790
TTC9
(+72257
)M
AP 3
K 9(+95127
)TT
C9N
M015351
ESC
CAL-121
chr15
78088340
-78088281
LIN
GO1
( minus163602
)TB
C1D2
B(+281683
)LI
NG
O1
NM032808
ESC
CAL-231
chr 16
88381397
-88381338
ZNF469
( minus112511
)BA
NP
(+396330
)ZN
F469
NM001127464
ESC
CAL-74
chr2
47548478
-47548537
CALM2
(minus144768
)EP
CAM
(minus47779
)EP
CAM
NM002354
ESC
CAL-152
chr20
56839403
-56839344
PMEP
A1(minus554343
)PP
P 4R1
L(+45121
)PM
EPA1
NM020182
ESC
CAL-337
chr 3
171506465
-171506406
TNIK
(minus328239
)PL
D1
(+22068
)PL
D1
NM002662
ESC
CAL-80
chr3
177935638
- 177935579
KCN
MB2
(minus318615
)KC
NM
B 2N
M181361
ESC
CAL-90
chr 3
195441846
-195441787
MU
C20
(minus5936
)SD
HA
P 2(+56907
)M
UC20
NM001098516
ESC
CAL-59
chr 3
64855048
-64855107
AD
AM
TS9
(minus181713
)A
DA
MTS9
NM182920
ESC
CAL-72
chr 4
74922502
-74922443
CXCL3
(minus17983
)CX
CL2
(+42524
)CX
CL3
NM002090
ESC
CAL-79
chr4
79626460
- 79626401
BMP2
K(minus71101
)A
NX
A3
(+153689
)A
NX
A3
NM005139
ESC
CAL-253
chr4
8357097
-8357038
ACO
X3(+85384
)H
TRA3
(+85579
)AC
OX 3
NM003501
ESC
CAL-179
chr4
8359416
- 8359357
ACO
X3(+83065
)H
TRA3
(+87898
)H
TRA3
NM053044
ESC
CAL-11
chr4
84299039
- 84299098
HPS
E(minus43035
)H
ELQ
(+77956
)H
PSE
NM006665
ESC
CAL-41
chr4
8512901
- 8512960
GPR78
(minus69286
)M
ETTL19
(+70399
)G
PR78
NM080819
ESC
CAL-353
chr 5
1175322
-1175263
SLC12
A7
(minus63121
)SL
C6A19
(minus26417
)SL
C12
A7
NM006598
ESC
CAL-260
chr5
131808618
-131808677
IRF 1
(+17817
)SL
C22
A5
(+103247
)IR
F 1N
M002198
ESC
CAL-132
chr5
134578804
- 134578863
PITX1
(minus208870
)H2
AFY
(+156094
)PI
TX1
NM002653
ESC
CAL-4
chr5
141732627
-141732568
SPRY4
(minus27978
)FG
F1(+333055
)SP
RY4
NM030964
ESC
CAL-356
chr 5
1545355
-1545296
LPCA
T1( minus21250
)M
RPL36
(+254630
)LP
CAT 1
NM024830
XLO
C004881
chr5
72570680
-72570621
TMEM174
(+101628
)FO
XD1
(+173701
)up
FOXD1
NM004472
ESC
CAL-36
chr6
106899513
-106899572
ATG5
( minus125848
)A
IM1
(minus59762
)A
IM1
NM001624
XLO
C005849
chr6
138145112
-138145053
OLI
G3
( minus329552
)TN
FAIP
3(minus43498
)TN
FAIP3
NM006290
ESC
CAL-262
chr6
2283830
-2283771
GM
DS
(minus37933
)M
YLK4
(+467353
)G
MD
SN
M001500
ESC
CAL-120
chr6
29716817
-29716758
LOC554223
(minus42895
)H
LA-F
(+25671
)H
LA-F
NM018950
ESC
CAL-257
chr 6
29988410
-29988469
ZNRD1
( minus40596
)H
LA-J
(+14223
)H
LA-J
NR024240
ESC
CAL-73
chr6
36126663
-36126722
BRPF3
( minus37857
)M
APK13
(+28431
)M
APK13
NM002754
ESC
CAL-97
chr 6
40305634
-40305575
MO
CS1
(minus410150
)LR
FN2
(+249521
)LR
FN2
NM020737
ESC
CAL-333
chr 6
72018004
-72018063
RIM
S1(minus578616
)O
GFR
L1(+19557
)O
GFR
L1N
M024576
ESC
CAL-115
chr7
12593405
-12593464
VW
DE
(minus149583
)SC
IN(minus16768
)SC
INN
M033128
ESC
CAL-68
chr7
139487166
-139487225
TBX
AS1
(minus41756
)H
IPK2
(minus9503
)H
IPK2
AF207702
ESC
CAL-87
chr7
19958800
- 19958741
TMEM196
(minus146367
)M
ACC1
(+298242
)M
ACC1
NM182762
ESC
CAL-58
chr8
16355082
-16355141
MSR1
(minus304812
)FG
F 20
(+504562
)M
SR1
NM002445
ESC
CAL-130
chr8
37189082
-37189141
ZNF703
(minus364189
)KC
NU1
(+547270
)ZN
F703
NM025069
ESC
CAL-199
8
chr 8
38623612
-38623671
RNF 5
P 1(minus164867
)TA
CC 1
(minus21080
)TA
CC1
BC041391
ESC
CAL-
chr9
34668311
-34668252
CCL27
(minus5593
)C
CL19
(+22992
)C
CL19
NM006274
NotesR
ed-highlighted
genesa
rewho
seexpressio
nsaresig
nificantly
changedin
Esop
hagealSquamou
sCellC
arcino
ma(ESC
C)G
reen
color-high
lighted
rowsgenesinvolvedin
lipid
metabolism
predictedwith
GOenric
hmentanalysis
Yellowcolor-high
lighted
rowsTh
eexpressionof
lncR
NAES
CCAL-5was
valid
ated
byPC
R
International Journal of Genomics 9
Final extension was at 72∘C for 10 minutes The ampliconswere resolved in 2 agarose gel
34 Bioinformatic Analysis Intensity data were exportedto GeneSpring 120 (Agilent Technologies Santa ClaraCA USA) for quantile normalization and the analysis ofdifferentially expressed long noncoding RNAs and codingRNAs Paired 119905-test analysis was used to obtain probe setswhose magnitude of change in expression of RNAs betweenESCC tissue and adjacent normal esophageal tissuewas eithergreater or less than 20 fold and 119875 value lt 005 (119875 values werecorrected formultiple testing using themethod of Benjamini-Hochberg) The normalized data containing 42544 probeswere further analyzed using the R program All controlprobes were removed We then defined the coding (ldquoNM rdquoldquoXM rdquo) and noncoding (ldquolincRNArdquo ldquoNR rdquo and ldquoXR rdquo) genesin the normalized data according to the definition of RefSeqaccession format (httpwwwncbinlmnihgovprojectsRefSeqkeyhtml) Differentially expressed long noncodingRNAs (DE-lncRNAs) and coding RNAs (DE-mRNAs)were further identified The landscapes of the wholetranscriptome (lncRNAs + coding RNAs) or all lncRNAsor all coding RNAs were analyzed with gene expressiondynamic inspector (GEDI)
35 Co-Location and Co-Expression Analysis between DE-lncRNAs and DE-mRNAs Genomic coordinates of DE-lncRNAs were imported to GREAT software (httpbejeranostanfordedugreatpublichtmlindexphp) for co-locationanalysis Neighboring coding genes were then matched withDE-mRNAs to obtain a co-expression dataset Three sub-groups of genes (DE-lncRNA co-located genes DE-mRNAsand co-expressed genes) were used for gene expressionnetwork analysis using Cytoscape software (v283)
Abbreviations
AFAP1-AS1 Actin filament-associated protein 1 antisenseRNA
Wei Cao and Wei Wu contributed equally to this project
Acknowledgments
The National Natural Science Foundation of China (no81171992) and the Zhengzhou Science and Technology
Programme (121PPTGG494-8) supported this work Theauthors thank Anne Haegert atThe Laboratory for AdvancedGenome Analysis at the Vancouver Prostate Centre Van-couver Canada for expert technical support They alsothank GenomeSky Inc Canada for providing computationalconsultation
References
[1] J Ferlay H R Shin F Bray D Forman C Mathers and DM Parkin ldquoEstimates of worldwide burden of cancer in 2008GLOBOCAN2008rdquo International Journal of Cancer vol 127 no12 pp 2893ndash2917 2010
[2] W Wu and J A Chan ldquoUnderstanding the role of longnoncoding RNAs in the cancer genomerdquo in Next GenerationSequencing in Cancer Research-Decoding Cancer Genome WWu and H Choudhry Eds pp 199ndash215 Springer New YorkNY USA 2013
[3] J J Hao T Gong Y Zhang et al ldquoCharacterization of generearrangements resulted from genomic structural aberrationsin human esophageal squamous cell carcinoma KYSE150 cellsrdquoGene vol 513 no 1 pp 196ndash201 2013
[4] M Kano N Seki N Kikkawa et al ldquoMiR-145 miR-133aand miR-133b tumor-suppressive miRNAs target FSCN1 inesophageal squamous cell carcinomardquo International Journal ofCancer vol 127 no 12 pp 2804ndash2814 2010
[5] H SuNHuHH Yang et al ldquoGlobal gene expression profilingand validation in esophageal squamous cell carcinoma and itsassociation with clinical phenotypesrdquo Clinical Cancer Researchvol 17 no 9 pp 2955ndash2966 2011
[6] D M Greenawalt C Duong G K Smyth et al ldquoGeneexpression profiling of esophageal cancer comparative analysisof Barrettrsquos esophagus adenocarcinoma and squamous cellcarcinomardquo International Journal of Cancer vol 120 no 9 pp1914ndash1921 2007
[7] I Dunham A Kundaje S F Aldred et al ldquoAn integratedencyclopedia of DNA elements in the human genomerdquo Naturevol 489 no 7414 pp 57ndash74 2012
[8] E A Gibb C J Brown and W L Lam ldquoThe functional roleof long non-coding RNA in human carcinomasrdquo MolecularCancer vol 10 article 38 2011
[9] A L Brunner A H Beck B Edris et al ldquoTranscriptionalprofiling of lncRNAs and novel transcribed regions across adiverse panel of archived human cancersrdquo Genome Biology vol13 no 8 article R75 2012
[10] M Guttman and J L Rinn ldquoModular regulatory principles oflarge non-coding RNAsrdquo Nature vol 482 no 7385 pp 339ndash346 2012
[11] K C Wang Y W Yang B Liu et al ldquoA long noncodingRNA maintains active chromatin to coordinate homeotic geneexpressionrdquo Nature vol 472 no 7341 pp 120ndash126 2011
[12] A Dean ldquoOn a chromosome far far away LCRs and geneexpressionrdquo Trends in Genetics vol 22 no 1 pp 38ndash45 2006
[13] P Ji S Diederichs W Wang et al ldquoMALAT-1 a novel noncod-ing RNA and thymosin 1205734 predict metastasis and survival inearly-stage non-small cell lung cancerrdquo Oncogene vol 22 no39 pp 8031ndash8041 2003
[14] R A Gupta N Shah K C Wang et al ldquoLong non-codingRNA HOTAIR reprograms chromatin state to promote cancermetastasisrdquo Nature vol 464 no 7291 pp 1071ndash1076 2010
10 International Journal of Genomics
[15] R Kogo T Shimamura K Mimori et al ldquoLong noncodingRNAHOTAIR regulates polycomb-dependent chromatinmod-ification and is associated with poor prognosis in colorectalcancersrdquo Cancer Research vol 71 no 20 pp 6320ndash6326 2011
[16] K Kim I Jutooru G Chadalapaka et al ldquoHOTAIR is anegative prognostic factor and exhibits pro-oncogenic activityin pancreatic cancerrdquo Oncogene vol 32 pp 1616ndash1625 2013
[17] D Li J Feng T Wu et al ldquoLong intergenic noncoding RNAHOTAIR is overexpressed and regulates PTEN methylation inlaryngeal squamous cell carcinomardquo The American Journal ofPathology vol 182 no 1 pp 64ndash70 2013
[18] A Bhan I Hussain K I Ansari S Kasiri A Bashyal and S SMandal ldquoAntisense transcript long noncoding RNA (lncRNA)HOTAIR is transcriptionally induced by estradiolrdquo Journal ofMolecular Biology vol 425 no 19 pp 3707ndash3722 2013
[19] D Cejka D Losert and V Wacheck ldquoShort interfering RNA(siRNA) tool or therapeuticrdquo Clinical Science vol 110 no 1pp 47ndash58 2006
[20] J R Prensner and A M Chinnaiyan ldquoThe emergence oflncRNAs in cancer biologyrdquo Cancer Discovery vol 1 pp 391ndash407 2011
[21] W Wu T D Bhagat X Yang et al ldquoHypomethylation ofnoncoding DNA regions and overexpression of the long non-coding RNA AFAP1-AS1 in Barrettrsquos esophagus and esophagealadenocarcinomardquoGastroenterology vol 144 no 5 pp 956ndash9662013
[22] E Ozgur U Mert M Isin M Okutan N Dalay and U GezerldquoDifferential expression of long non-coding RNAs duringgenotoxic stress-induced apoptosis in HeLa and MCF-7 cellsrdquoClinical and Experimental Medicine vol 13 no 2 pp 119ndash1262013
[23] J D Li Q C Feng and J S Li ldquoDifferential gene expressionprofiling of oesophageal squamous cell carcinoma by dnamicroarray and bioinformatics analysisrdquo Journal of Interna-tional Medical Research vol 38 no 6 pp 1904ndash1912 2010
[24] S Ma J Y Bao P S Kwan et al ldquoIdentification of PTK6via RNA sequencing analysis as a suppressor of esophagealsquamous cell carcinomardquo Gastroenterology vol 143 pp 675ndash686 2012
[25] M Tong K W Chan J Y Bao et al ldquoRab25 is a tumorsuppressor gene with antiangiogenic and anti-invasive activitiesin esophageal squamous cell carcinomardquo Cancer Research vol72 no 22 pp 6024ndash6035 2012
[26] Y J Geng S L Xie Q Li J Ma and G Y Wang ldquoLargeintervening non-coding RNA HOTAIR is associated withhepatocellular carcinoma progressionrdquo Journal of InternationalMedical Research vol 39 no 6 pp 2119ndash2128 2011
[27] X Zhou T J Lawrence Z He C R Pound J Mao andS A Bigler ldquoThe expression level of lysophosphatidylcholineacyltransferase 1 (LPCAT1) correlates to the progression ofprostate cancerrdquo Experimental andMolecular Pathology vol 92no 1 pp 105ndash110 2012
[28] Y Morita T Sakaguchi K Ikegami et al ldquoLysophosphatidyl-choline acyltransferase 1 altered phospholipid composition andregulated hepatoma progressionrdquo Journal of Hepatology vol 59no 2 pp 292ndash299 2013
Figure 1 Transcriptomic landscape of esophageal squamous cell cancer (ESCC) (a) Whole transcriptome of tumor (T) and adjacent normaltissue (N) of four patients with ESCCwere detected using a microarray with 7419 long noncoding RNAs (lncRNAs) and 27958 coding RNAsTwo main clusters (Ts and Ns) were generated using unsupervised clustering methods Then a self-organizing map (SOM) of either wholetranscriptome (both lncRNAs and mRNAs) or lncRNAs or mRNA was produced from each sample (see legend in up-left corner of thisfigure and the arrows are meant to indicate the potential interaction) using gene expression dynamic inspector (GEDI) Mosaic patternsare pseudocolored SOMs to show integrated biological entity in each sample Red through blue color indicates high to low expression level(b) and (c) Differentially expressed lncRNAs (DE-lncRNAs) and coding RNAs (DE-mRNAs) in ESCC Hierarchical clustering analysis of 410DE-lncRNAs (b) and 1219 DE-mRNAs (c) between ESCC tissue and adjacent normal tissue (fold change gt or lt 2-fold and 119875 lt 005) Redand green colors indicate high and low expression respectively In the heatmap columns represent samples and rows represent each geneThe scale of expression level is shown on the horizontal bar (d) KEGG functional analysis of DE-mRNA networks in ESCCThe DE-mRNAgenes are involved in cancer-related signaling functions and a detailed list of significant GO terms is shown in Figure S1 and its associatedlegend in Supplementary Information
4 International Journal of Genomics
100
80
60
40
20
0N T N T N T
Nor
mal
ized
inte
nsity
in m
icro
arra
y
Nor
mal
ized
inte
nsity
Nor
mal
ized
inte
nsity
in m
icro
arra
y
in m
icro
arra
y
HOTAIRlowast
lowastlowast
lowastlowast
ESCCAL-1 ESCCAL-52500
2000
1500
1000
500
0
700
600
500
400
300
200
100
0
(a)
N T N T N T N T N T N T N T N T N T N T N T N T
GA
PDH
HO
TAIR
Patient 1 Patient 2 Patient 3
ESCC
AL-
1
ESCC
AL-
5
GA
PDH
HO
TAIR
ESCC
AL-
1
ESCC
AL-
5
GA
PDH
DN
A m
arke
r
HO
TAIR
ESCC
AL-
1
ESCC
AL-
5(b)
Figure 2 Long noncoding RNAs (lncRNAs) expression in esophageal squamous cell carcinoma (ESCC) (a) Three differentially expressedlncRNAs HOTAIR ESCCAL-1 and ESCCAL-5 from microarray detection The average intensity of expression in normal tissues (N) andtumors (T) is plotted with their standard deviations (b) Validation ofHOTAIR ESCCAL-1 and ESCCAL-5with independent patient samplesby PCR analysis The amplicons were separated with 2 agarose gel GAPDH was used as an internal control Significance is lowast119875 lt 005lowastlowast
119875 lt 001
Table 1 Validation of selected differential expression of mRNAs in esophageal squamous cell carcinoma in independent studies
Probe name P value FC Regulation Gene symbol Genbank accession Independent study ReferenceA 33 P3232692 0005838984 8301461 Up IL24 NM 001185156
Microarray [20]
A 24 P411121 000055 5329484 Up TNFRSF18 NM 148901A 23 P169097 881119864 minus 05 4466178 Up WISP1 NM 080838A 23 P304304 0004822649 3944957 Down ARSF NM 004042A 24 P56363 0003573538 3323955 Down CAB39L NM 030925A 23 P419760 0001041661 3270335 Down CRISP3 NM 006061A 23 P413923 0002921898 454899 Down DMRTA1 NM 022160A 23 P56978 0002093997 5438183 Down PTK6 NM 005975 RNA-seq [21]A 23 P115091 0005171322 3289834 Down RAB25 NM 020387 Q-RT-PCR [22]A 33 P3258542 0001039129 2036035 Down SPINK8 NM 001080525 Microarray [20]
International Journal of Genomics 5
PCR methods in matched-pair tissue samples from threeadditional ESCC patients and the results are consistent withthemicroarray analysis (Figure 2(b)) Interestingly except forHOTAIR other previously reported lncRNAs (ie MALAT-1 PCAT-1 and AFAP1-AS1) are not differentially expressed inour analysis of ESCC Therefore the DE-lncRNAs that wehave identifiedmay be a unique property of ESCC andwe arecurrently using a population-based analysis to characterizethese DE-lncRNAs as potential genomic biomarkers andregulatory elements in the dynamic process leading to ESCC
23 LncRNAs Co-located and Co-expressed with Coding Genesin ESCC LncRNAs have been reported to coordinate theregulation of neighboring coding genes through a ldquolocuscontrolrdquo process [11] We wondered whether such a ldquolocuscontrolrdquo process could operate in ESCC development andtherefore we searched neighboring genes of the 410 DE-lncRNAs in the genomeThemajority (988) of the 410 DE-lncRNAs harbor neighboring coding genes whose genomiclocations are within sim5 kb upstream and sim1 kb downstreamof the lncRNA and may extend to 1000 kb in both directions(Figure 3(a)) Interrogation of 538 coding genes that areneighbors of these DE-lncRNAs (DE-lncRNAs co-locatedgenes) revealed predicted functions in 9 common pathwayssuch as the AP1 transcription factor network integrin-linkedkinase signaling several signaling pathways in adherensjunctions and FOXO family signaling (Figure 3(b))
We asked whether any DE-lncRNAs co-located genes arealso differentially expressed in ESCC Analysis of the DE-lncRNAs co-located genes withDE-mRNAdata set identified76 genomically co-located and differentially co-expressedgenes (Figure 3(c) and Table 2) Strikingly the co-locatedand co-expressed genes with DE-lncRNAs may be involvedin ether lipid metabolism pathways by the participation ofthe LPCAT1 gene encoding lysophosphatidylcholine acyl-transferase1 and the PLD1 gene encoding phospholipase D1(Figure 3(c)) The lncRNA ESCCAL-337 (chr3171506370-171528740) was downregulated in ESCC and located at22068 bp downstream of the PLD1 gene whose expres-sion was also decreased in ESCC In contrast the lncRNAESCCAL-356 (chr51544500-1567142 reverse strand) wasdownregulated in ESCC and located at 21250 bp upstreamof LPCAT1 whose expression was upregulated in ESCC(Figure 3(c)) LPCAT1 modulates phospholipid compositionby catalyzing lysophosphatidylcholine into phosphatidyl-choline and overexpression of LPCAT1was reported to createfavorable conditions for cancer cell proliferation [27 28]Therefore at least two of the DE-lncRNAs have the potentialto contribute to ESCC by a ldquolocus controlrdquo process withneighboring coding genes
In conclusion we performed a genome-wide survey ofthe expression of lncRNAs and coding mRNAs from pairedsamples of primary neoplastic tissue and adjacent non-neoplastic normal tissue from four individuals The overalltranscriptomic landscape (both lncRNAs andmRNAs) is ableto distinguish malignant from normal tissue in each personWe discovered a set of differentially expressed lncRNAsand their co-located and co-expressed coding mRNAs and
demonstrated that lncRNAs may be involved in ether lipidmetabolism in ESCC Our study provides genomic supportfor a model of a ldquolocus controlrdquo process in ESCC and aframework for further experimental study
3 Materials and Methods
31 Specimens Written informed consent was obtained frompatients before surgery and the study protocol was approvedby the Institutional Review Board for the use of human sub-jects at Zhengzhou Hospital Primary tumors and adjacentnonneoplastic tissues were obtained frompatients with ESCCwhounderwent surgical treatment at LinxianHospital inMay2012 All tissues were frozen in liquid nitrogen immediatelyafter surgical resection None of the patients had priorchemotherapy or radiotherapy nor did they have any otherserious diseases All ESCC tissues were histopathologicallydiagnosed by at least two independent senior pathologists
32 Microarray Hybridization Total RNAs were extractedusing Trizol reagent following manufacturerrsquos instructions(Invitrogen Carlsbad CA USA) The quality of RNAswas measured with a 2100 Bioanalyzer (Agilent technologyUSA) Input of 100 ng of total RNA was used to generateCyanine-3 labeled cRNA according to theAgilentOne-ColorMicroarray-Based Gene Expression Analysis Low for InputQuick Amp Labeling kit (v60) Samples were hybridizedon Agilent SurePrint G3 Human GE 8 times 60K Microarray(Design ID 028004) Arrays were scanned with the AgilentDNAMicroarray Scanner at a 3120583m scan resolution and datawere processed with Agilent Feature Extraction 11011 Themicroarray data discussed in this article have been depositedin National Center for Biotechnology Information (NCBI)Gene ExpressionOmnibus (GEO) and are accessible through(GEO) Series accession number GSE45350 (httpwwwncbinlmnihgovgeoqueryacccgiacc=GSE45350)
33 Validation by Polymerase Chain Reaction (PCR) PCRanalysis was performed on additional matched ESCC andadjacent nonneoplastic tissues for selected lncRNAs Theprimer sequences for PCR are as follows HOTAIR forward51015840-GGTAGAAAAAGCAACCACGAAGC-31015840 and reverse51015840-ACATAAACCTCTGTCTGTGAGTGCC-31015840 ESSCAL-1(chr876121095-76189420 reverse strand) forward 51015840-CCA-GACAGCAGCAAAGCAAT-31015840 and reverse 51015840-GGAAGC-AGCAAATGTGTCCAT-31015840 ESSCAL-5 (chr2216585154-216585719 forward strand) forward 51015840-TACCAACATTGT-CCACCGGG-31015840 and reverse 51015840-GCTGATGACAGTCCC-TTGCT-31015840 GAPDH was used as a control forward 51015840-CCG-GGAAACTGTGGCGTGATGG-31015840 and reverse 51015840-AGG-TGGAGGAGTGGGTGTCGCTGTT-31015840 The thermocycleconditions are as follows initial denaturation at 95∘C for10 minutes followed by 94∘C for 45 seconds 65∘C for 30seconds and 72∘C for 1 minute for 15 cycles Then theannealing temperature was reduced by 05∘Ccycle for thenext 14 cycles and the amplification was finished withanother 24 cycles with the annealing temperature at 58∘C
6 International Journal of Genomics
lncRNAs
Gene b Gene c
20
40
60
80
100
0lncRNAs
Gene a
lncRNAs
Number of associated genes per region0 1 2G
enom
ic re
gion
s of l
ncRN
As (
)
(a)
AP-1 transcription factor network
Integrin-linked kinase signalingRegulation of CDC42 activity
Posttranslational regulation of adherens junction stability and disassemblyN-cadherin signaling events
E-cadherin signaling in the nascent adherens junctionStabilization and expansion of the E-cadherin adherens junction
E-cadherin signaling eventsFOXO family signaling
1176
1105904
511466
448448440
422
(b)
DE-lncRNAsco-located genes
DE-mRNAs
538 330776
ESCCAL-356
ESCCAL-337
Chr3
Chr5
Lipid metabolism
PLD1
LPCAT1
(c)
Figure 3 Identification of lncRNAs co-located and co-expressed neighboring genes in esophageal squamous cell carcinoma (ESCC)(a) Identification of neighboring genes of the DE-lncRNAs The genomic coordinate information of 410 DE-lncRNAs was used to searchneighboring genes whose genomic locations are within sim5 kb upstream and sim1 kb downstream of the lncRNA and may extend to 1000 kbin both directions using GREAT software (httpbejeranostanfordedugreatpublichtmlindexphp) The percentage of DE-lncRNAsharboring zero one or two neighboring genes is presented (b) Gene Ontology (GO) enrichment analysis of lncRNAs co-located genesIdentified gene enriched pathwaysterms are listed on the left the length of horizontal bars and the numbers on the right indicate thepercentage of genes involved in each pathwayterm (c) LncRNAs co-located and co-expressed coding mRNAs Overlap of 538 DE-lncRNAco-located genes with 3307 DE-mRNAs in microarrays identified 76 lncRNAs co-located and co-expressed coding mRNAs (list in Table 2)GO enrichment analysis suggests phospholipase D1 (PLD1) and lysophosphatidylcholine acyltransferase1 (LPCAT1) are involved in ether lipidmetabolism pathway Genomic location shows that PLD1 is located at minus22068 bp upstream of ESCCAL-337 lncRNA on Chr 3 and LPCAT1is at minus21250 bp upstream of ESCCAL-356 lncRNA on Chr 5
International Journal of Genomics 7
Table2Listof
identifi
edco-lo
catedandco-expressed
genesw
ithdifferentially
expressedlncR
NAsinES
CC
LncR
NA
nam
eG
enom
ic co
ordi
nate
Ln
cRN
Aex
pres
sion
LncR
NA
asso
ciat
edge
nes
Gen
eexp
ress
ion
Gen
esym
bol
Gen
bank
acce
ssio
nES
CCA
L-177
chr1
205404138
-205404079
Up
CDK18
(minus69575
)LE
MD
1(minus12895
)U
pLE
MD1
NM001001552
ESC
CAL-348
chr1
222587441
-222587382
Up
DU
SP10
(minus671896
)H
HIP
L2(+134032
)U
pD
USP10
NM007207
ESC
CAL-31
chr1
225237693
-225237752
Up
DN
AH14
(+120367
)LB
R(+378092
)U
pD
NA
H14
NM001373
ESC
CAL-159
chr1
225240300
-225240359
Up
DN
AH14
(+122974
)LB
R(+375485
)U
pD
NA
H14
NM001373
ESC
CAL-327
chr1
89887111
-89887052
Up
LRRC8
B(minus103315
)G
BP6
(+57646
)D
own
GBP6
NM198460
XLO
C000915
chr1
91295528
-91295469
Up
BARH
L2(minus112705
)ZN
F644
(+192313
2N
M020063
ESC
CAL-65
chr11
2017146
-2017205
Up
MRP
L23
(+48674
)IG
F2(+145165
)U
pIG
F2N
M000612
ESC
CAL-19
chr12
66204406
-66204347
Up
HM
GA
2(minus13863
)M
SRB3
(+531610
)U
pH
MG
A2
NM003484
XLO
C011548
chr15
81953024
-81952965
Up
TMC3
(minus286577
)M
EX3
B(+385366
)U
pM
EX3
BN
M032246
ESC
CAL-102
chr17
6766871
-6766930
Up
ALO
X12
(minus132483
)TE
KT1
(minus31841
)12
NM000697
ESC
CAL-342
chr17
78302158
-78302217
Up
END
OV
(minus86779
)RN
F213
(+67521
)U
pRN
F213
NM020954
ESC
CAL-342
chr17
78302158
-78302217
Up
END
OV
(minus86779
)RN
F213
(+67521
)001164638
ESC
CAL-99
chr2
102034125
-102034066
Up
CREG
2(minus30131
)RF
X8(+57069
)U
pCR
EG2
NM153836
ESC
CAL-5
chr2
216585455
-216585514
Up
FN1
(minus284694
)M
REG
(+292861
)D
own
MRE
GN
M018000
ESC
CAL-5
chr2
216585455
-216585514
Up
FN1
(minus284694
)M
REG
(+292861
)U
pFN1
NM054034
ESC
CAL-288
chr2
27789661
-27789602
Up
ZNF512
(minus16261
)G
CKR
(+69926
)D
own
GCK
RN
M001486
ESC
CAL-344
chr2
37327299
-37327358
Up
CCD
C75
(+15735
)EI
F2A
K2
(+56861
)U
pEI
F2A
K2N
M001135652
ESC
CAL-10
chr22
48086469
-48086528
Up
FAM
19A
5(minus798789
)TB
C1D22
A(+927981
)U
pFA
M19
A5
NM015381
ESC
CAL-81
chr5
141710377
-141710436
Up
SPRY4
(minus5787
)FG
F1(+355246
)U
pFG
F1N
M000800
ESC
CAL-81
chr5
141710377
-141710436
Up
SPRY4
(minus5787
)FG
F1(+355246
)U
pSP
RY4
NM030964
ESC
CAL-204
chr6
126699769
-126699828
Up
RSPO3
(minus740249
)CE
NPW
(+38546
)U
pCE
NPW
NM001012507
ESC
CAL-106
chr6
21822910
-21822969
Up
SOX4
(+228968
)PR
L(+480142
)U
pSO
X4N
M003107
ESC
CAL-239
chr8
104258684
-104258625
Up
FZD
6(minus52006
)BA
ALC
(+105734
)U
pBA
ALC
NM024812
ESC
CAL-239
chr8
104258684
-104258625
Up
FZD
6(minus52006
)BA
ALC
(+105734
)U
pFZ
D6
NM003506
ESC
CAL-300
chr1
180918852
-180918793
Dow
nST
X6(+73223
)XP
R1(+317677
)U
pXP
R1N
M004736
XLO
C000515
chr1
200384539
-200384598
Dow
nZN
F281
(minus5403
)KI
F14
(+205293
)U
pKI
F14
NM014875
XLO
C000515
chr1
200384539
-200384598
Dow
nZN
F281
(minus5403
)KI
F14
(+205293
)U
pZN
F281
NM012482
ESC
CAL-77
chr1
201592869
-201592810
Dow
nCS
RP1
(minus116873
)N
AV1
(minus24610
)U
pN
AV1
NM020443
ESC
CAL-76
chr1
90091283
-90091342
Dow
nLR
RC8
C(minus7331
)LR
RC8
B(+100916
)U
pLR
RC8
CN
M032270
ESC
CAL-209
chr10
14549236
-14549177
Dow
nFR
MD4
A(minus176341
)CD
NF
(+330776
)D
own
Dow
n
Dow
nD
own
Dow
n
CDN
FN
M001029954
ESC
CAL-284
chr10
44848467
-44848526
P1(minus562632
)CX
CL12
(+32048
)CX
CL12
NM199168
ESC
CAL-256
chr11
117671760
-117671701
Dow
nD
SCA
ML1
(minus3755
)D
SCA
ML1
NM020693
ESC
CAL-254
chr11
126219961
-126220020
Dow
nST3
GA
L4(minus5549
)D
CPS
(+46344
)ST3
GA
L4N
M006278
ESC
CAL-307
chr11
14975924
-14975983
Dow
nCY
P2R1
(minus62203
)CA
LCA
(+17878
)CY
P2R1
NM024514
ESC
CAL-105
chr11
17366661
-17366720
Dow
nB7
H6
(minus6618
)N
UCB2
(+68405
)D
own
NU
CB2
NM005013
ESC
CAL-232
chr12
131245982
-131245923
Dow
nRI
MBP2
(minus243491
)ST
X2(+77858
)D
own
RIM
BP2
NM015347
ESC
CAL-24
chr12
132824534
-132824475
Dow
nG
ALN
T9(+81400
)N
OC4
L(+195512
)U
pG
ALN
T9N
M021808
Dow
nH
NRN
PA3
)D
own
BARH
L
Dow
nA
LOX
Dow
nEN
DO
VN
MN
M
8 International Journal of Genomics
Table2Con
tinued
LncR
NA
nam
eG
enom
ic co
ordi
nate
sLn
cRN
Aex
pres
sion
LncR
NA
asso
ciat
edge
nes
Gen
eexp
ress
ion
Gen
esym
bol
Gen
bank
acce
ssio
nES
CCA
L-122
chr12
2952238
-2952297
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
n
Dow
n
Dow
n
Dow
nD
own
Dow
n
Dow
n
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
NRI
P 2(minus8047
)FO
XM1
(+34053
)FO
XM1
NM202002
ESC
CAL-338
chr12
32101291
-32101232
BICD1
(minus158923
)H3
F3C
( minus156087
)BI
CD1
NM001714
ESC
CAL-275
chr14
71180731
-71180790
TTC9
(+72257
)M
AP3
K9(+95127
)M
AP3
K 9N
M033141
ESC
CAL-275
chr 14
71180731
-71180790
TTC9
(+72257
)M
AP 3
K 9(+95127
)TT
C9N
M015351
ESC
CAL-121
chr15
78088340
-78088281
LIN
GO1
( minus163602
)TB
C1D2
B(+281683
)LI
NG
O1
NM032808
ESC
CAL-231
chr 16
88381397
-88381338
ZNF469
( minus112511
)BA
NP
(+396330
)ZN
F469
NM001127464
ESC
CAL-74
chr2
47548478
-47548537
CALM2
(minus144768
)EP
CAM
(minus47779
)EP
CAM
NM002354
ESC
CAL-152
chr20
56839403
-56839344
PMEP
A1(minus554343
)PP
P 4R1
L(+45121
)PM
EPA1
NM020182
ESC
CAL-337
chr 3
171506465
-171506406
TNIK
(minus328239
)PL
D1
(+22068
)PL
D1
NM002662
ESC
CAL-80
chr3
177935638
- 177935579
KCN
MB2
(minus318615
)KC
NM
B 2N
M181361
ESC
CAL-90
chr 3
195441846
-195441787
MU
C20
(minus5936
)SD
HA
P 2(+56907
)M
UC20
NM001098516
ESC
CAL-59
chr 3
64855048
-64855107
AD
AM
TS9
(minus181713
)A
DA
MTS9
NM182920
ESC
CAL-72
chr 4
74922502
-74922443
CXCL3
(minus17983
)CX
CL2
(+42524
)CX
CL3
NM002090
ESC
CAL-79
chr4
79626460
- 79626401
BMP2
K(minus71101
)A
NX
A3
(+153689
)A
NX
A3
NM005139
ESC
CAL-253
chr4
8357097
-8357038
ACO
X3(+85384
)H
TRA3
(+85579
)AC
OX 3
NM003501
ESC
CAL-179
chr4
8359416
- 8359357
ACO
X3(+83065
)H
TRA3
(+87898
)H
TRA3
NM053044
ESC
CAL-11
chr4
84299039
- 84299098
HPS
E(minus43035
)H
ELQ
(+77956
)H
PSE
NM006665
ESC
CAL-41
chr4
8512901
- 8512960
GPR78
(minus69286
)M
ETTL19
(+70399
)G
PR78
NM080819
ESC
CAL-353
chr 5
1175322
-1175263
SLC12
A7
(minus63121
)SL
C6A19
(minus26417
)SL
C12
A7
NM006598
ESC
CAL-260
chr5
131808618
-131808677
IRF 1
(+17817
)SL
C22
A5
(+103247
)IR
F 1N
M002198
ESC
CAL-132
chr5
134578804
- 134578863
PITX1
(minus208870
)H2
AFY
(+156094
)PI
TX1
NM002653
ESC
CAL-4
chr5
141732627
-141732568
SPRY4
(minus27978
)FG
F1(+333055
)SP
RY4
NM030964
ESC
CAL-356
chr 5
1545355
-1545296
LPCA
T1( minus21250
)M
RPL36
(+254630
)LP
CAT 1
NM024830
XLO
C004881
chr5
72570680
-72570621
TMEM174
(+101628
)FO
XD1
(+173701
)up
FOXD1
NM004472
ESC
CAL-36
chr6
106899513
-106899572
ATG5
( minus125848
)A
IM1
(minus59762
)A
IM1
NM001624
XLO
C005849
chr6
138145112
-138145053
OLI
G3
( minus329552
)TN
FAIP
3(minus43498
)TN
FAIP3
NM006290
ESC
CAL-262
chr6
2283830
-2283771
GM
DS
(minus37933
)M
YLK4
(+467353
)G
MD
SN
M001500
ESC
CAL-120
chr6
29716817
-29716758
LOC554223
(minus42895
)H
LA-F
(+25671
)H
LA-F
NM018950
ESC
CAL-257
chr 6
29988410
-29988469
ZNRD1
( minus40596
)H
LA-J
(+14223
)H
LA-J
NR024240
ESC
CAL-73
chr6
36126663
-36126722
BRPF3
( minus37857
)M
APK13
(+28431
)M
APK13
NM002754
ESC
CAL-97
chr 6
40305634
-40305575
MO
CS1
(minus410150
)LR
FN2
(+249521
)LR
FN2
NM020737
ESC
CAL-333
chr 6
72018004
-72018063
RIM
S1(minus578616
)O
GFR
L1(+19557
)O
GFR
L1N
M024576
ESC
CAL-115
chr7
12593405
-12593464
VW
DE
(minus149583
)SC
IN(minus16768
)SC
INN
M033128
ESC
CAL-68
chr7
139487166
-139487225
TBX
AS1
(minus41756
)H
IPK2
(minus9503
)H
IPK2
AF207702
ESC
CAL-87
chr7
19958800
- 19958741
TMEM196
(minus146367
)M
ACC1
(+298242
)M
ACC1
NM182762
ESC
CAL-58
chr8
16355082
-16355141
MSR1
(minus304812
)FG
F 20
(+504562
)M
SR1
NM002445
ESC
CAL-130
chr8
37189082
-37189141
ZNF703
(minus364189
)KC
NU1
(+547270
)ZN
F703
NM025069
ESC
CAL-199
8
chr 8
38623612
-38623671
RNF 5
P 1(minus164867
)TA
CC 1
(minus21080
)TA
CC1
BC041391
ESC
CAL-
chr9
34668311
-34668252
CCL27
(minus5593
)C
CL19
(+22992
)C
CL19
NM006274
NotesR
ed-highlighted
genesa
rewho
seexpressio
nsaresig
nificantly
changedin
Esop
hagealSquamou
sCellC
arcino
ma(ESC
C)G
reen
color-high
lighted
rowsgenesinvolvedin
lipid
metabolism
predictedwith
GOenric
hmentanalysis
Yellowcolor-high
lighted
rowsTh
eexpressionof
lncR
NAES
CCAL-5was
valid
ated
byPC
R
International Journal of Genomics 9
Final extension was at 72∘C for 10 minutes The ampliconswere resolved in 2 agarose gel
34 Bioinformatic Analysis Intensity data were exportedto GeneSpring 120 (Agilent Technologies Santa ClaraCA USA) for quantile normalization and the analysis ofdifferentially expressed long noncoding RNAs and codingRNAs Paired 119905-test analysis was used to obtain probe setswhose magnitude of change in expression of RNAs betweenESCC tissue and adjacent normal esophageal tissuewas eithergreater or less than 20 fold and 119875 value lt 005 (119875 values werecorrected formultiple testing using themethod of Benjamini-Hochberg) The normalized data containing 42544 probeswere further analyzed using the R program All controlprobes were removed We then defined the coding (ldquoNM rdquoldquoXM rdquo) and noncoding (ldquolincRNArdquo ldquoNR rdquo and ldquoXR rdquo) genesin the normalized data according to the definition of RefSeqaccession format (httpwwwncbinlmnihgovprojectsRefSeqkeyhtml) Differentially expressed long noncodingRNAs (DE-lncRNAs) and coding RNAs (DE-mRNAs)were further identified The landscapes of the wholetranscriptome (lncRNAs + coding RNAs) or all lncRNAsor all coding RNAs were analyzed with gene expressiondynamic inspector (GEDI)
35 Co-Location and Co-Expression Analysis between DE-lncRNAs and DE-mRNAs Genomic coordinates of DE-lncRNAs were imported to GREAT software (httpbejeranostanfordedugreatpublichtmlindexphp) for co-locationanalysis Neighboring coding genes were then matched withDE-mRNAs to obtain a co-expression dataset Three sub-groups of genes (DE-lncRNA co-located genes DE-mRNAsand co-expressed genes) were used for gene expressionnetwork analysis using Cytoscape software (v283)
Abbreviations
AFAP1-AS1 Actin filament-associated protein 1 antisenseRNA
Wei Cao and Wei Wu contributed equally to this project
Acknowledgments
The National Natural Science Foundation of China (no81171992) and the Zhengzhou Science and Technology
Programme (121PPTGG494-8) supported this work Theauthors thank Anne Haegert atThe Laboratory for AdvancedGenome Analysis at the Vancouver Prostate Centre Van-couver Canada for expert technical support They alsothank GenomeSky Inc Canada for providing computationalconsultation
References
[1] J Ferlay H R Shin F Bray D Forman C Mathers and DM Parkin ldquoEstimates of worldwide burden of cancer in 2008GLOBOCAN2008rdquo International Journal of Cancer vol 127 no12 pp 2893ndash2917 2010
[2] W Wu and J A Chan ldquoUnderstanding the role of longnoncoding RNAs in the cancer genomerdquo in Next GenerationSequencing in Cancer Research-Decoding Cancer Genome WWu and H Choudhry Eds pp 199ndash215 Springer New YorkNY USA 2013
[3] J J Hao T Gong Y Zhang et al ldquoCharacterization of generearrangements resulted from genomic structural aberrationsin human esophageal squamous cell carcinoma KYSE150 cellsrdquoGene vol 513 no 1 pp 196ndash201 2013
[4] M Kano N Seki N Kikkawa et al ldquoMiR-145 miR-133aand miR-133b tumor-suppressive miRNAs target FSCN1 inesophageal squamous cell carcinomardquo International Journal ofCancer vol 127 no 12 pp 2804ndash2814 2010
[5] H SuNHuHH Yang et al ldquoGlobal gene expression profilingand validation in esophageal squamous cell carcinoma and itsassociation with clinical phenotypesrdquo Clinical Cancer Researchvol 17 no 9 pp 2955ndash2966 2011
[6] D M Greenawalt C Duong G K Smyth et al ldquoGeneexpression profiling of esophageal cancer comparative analysisof Barrettrsquos esophagus adenocarcinoma and squamous cellcarcinomardquo International Journal of Cancer vol 120 no 9 pp1914ndash1921 2007
[7] I Dunham A Kundaje S F Aldred et al ldquoAn integratedencyclopedia of DNA elements in the human genomerdquo Naturevol 489 no 7414 pp 57ndash74 2012
[8] E A Gibb C J Brown and W L Lam ldquoThe functional roleof long non-coding RNA in human carcinomasrdquo MolecularCancer vol 10 article 38 2011
[9] A L Brunner A H Beck B Edris et al ldquoTranscriptionalprofiling of lncRNAs and novel transcribed regions across adiverse panel of archived human cancersrdquo Genome Biology vol13 no 8 article R75 2012
[10] M Guttman and J L Rinn ldquoModular regulatory principles oflarge non-coding RNAsrdquo Nature vol 482 no 7385 pp 339ndash346 2012
[11] K C Wang Y W Yang B Liu et al ldquoA long noncodingRNA maintains active chromatin to coordinate homeotic geneexpressionrdquo Nature vol 472 no 7341 pp 120ndash126 2011
[12] A Dean ldquoOn a chromosome far far away LCRs and geneexpressionrdquo Trends in Genetics vol 22 no 1 pp 38ndash45 2006
[13] P Ji S Diederichs W Wang et al ldquoMALAT-1 a novel noncod-ing RNA and thymosin 1205734 predict metastasis and survival inearly-stage non-small cell lung cancerrdquo Oncogene vol 22 no39 pp 8031ndash8041 2003
[14] R A Gupta N Shah K C Wang et al ldquoLong non-codingRNA HOTAIR reprograms chromatin state to promote cancermetastasisrdquo Nature vol 464 no 7291 pp 1071ndash1076 2010
10 International Journal of Genomics
[15] R Kogo T Shimamura K Mimori et al ldquoLong noncodingRNAHOTAIR regulates polycomb-dependent chromatinmod-ification and is associated with poor prognosis in colorectalcancersrdquo Cancer Research vol 71 no 20 pp 6320ndash6326 2011
[16] K Kim I Jutooru G Chadalapaka et al ldquoHOTAIR is anegative prognostic factor and exhibits pro-oncogenic activityin pancreatic cancerrdquo Oncogene vol 32 pp 1616ndash1625 2013
[17] D Li J Feng T Wu et al ldquoLong intergenic noncoding RNAHOTAIR is overexpressed and regulates PTEN methylation inlaryngeal squamous cell carcinomardquo The American Journal ofPathology vol 182 no 1 pp 64ndash70 2013
[18] A Bhan I Hussain K I Ansari S Kasiri A Bashyal and S SMandal ldquoAntisense transcript long noncoding RNA (lncRNA)HOTAIR is transcriptionally induced by estradiolrdquo Journal ofMolecular Biology vol 425 no 19 pp 3707ndash3722 2013
[19] D Cejka D Losert and V Wacheck ldquoShort interfering RNA(siRNA) tool or therapeuticrdquo Clinical Science vol 110 no 1pp 47ndash58 2006
[20] J R Prensner and A M Chinnaiyan ldquoThe emergence oflncRNAs in cancer biologyrdquo Cancer Discovery vol 1 pp 391ndash407 2011
[21] W Wu T D Bhagat X Yang et al ldquoHypomethylation ofnoncoding DNA regions and overexpression of the long non-coding RNA AFAP1-AS1 in Barrettrsquos esophagus and esophagealadenocarcinomardquoGastroenterology vol 144 no 5 pp 956ndash9662013
[22] E Ozgur U Mert M Isin M Okutan N Dalay and U GezerldquoDifferential expression of long non-coding RNAs duringgenotoxic stress-induced apoptosis in HeLa and MCF-7 cellsrdquoClinical and Experimental Medicine vol 13 no 2 pp 119ndash1262013
[23] J D Li Q C Feng and J S Li ldquoDifferential gene expressionprofiling of oesophageal squamous cell carcinoma by dnamicroarray and bioinformatics analysisrdquo Journal of Interna-tional Medical Research vol 38 no 6 pp 1904ndash1912 2010
[24] S Ma J Y Bao P S Kwan et al ldquoIdentification of PTK6via RNA sequencing analysis as a suppressor of esophagealsquamous cell carcinomardquo Gastroenterology vol 143 pp 675ndash686 2012
[25] M Tong K W Chan J Y Bao et al ldquoRab25 is a tumorsuppressor gene with antiangiogenic and anti-invasive activitiesin esophageal squamous cell carcinomardquo Cancer Research vol72 no 22 pp 6024ndash6035 2012
[26] Y J Geng S L Xie Q Li J Ma and G Y Wang ldquoLargeintervening non-coding RNA HOTAIR is associated withhepatocellular carcinoma progressionrdquo Journal of InternationalMedical Research vol 39 no 6 pp 2119ndash2128 2011
[27] X Zhou T J Lawrence Z He C R Pound J Mao andS A Bigler ldquoThe expression level of lysophosphatidylcholineacyltransferase 1 (LPCAT1) correlates to the progression ofprostate cancerrdquo Experimental andMolecular Pathology vol 92no 1 pp 105ndash110 2012
[28] Y Morita T Sakaguchi K Ikegami et al ldquoLysophosphatidyl-choline acyltransferase 1 altered phospholipid composition andregulated hepatoma progressionrdquo Journal of Hepatology vol 59no 2 pp 292ndash299 2013
Figure 2 Long noncoding RNAs (lncRNAs) expression in esophageal squamous cell carcinoma (ESCC) (a) Three differentially expressedlncRNAs HOTAIR ESCCAL-1 and ESCCAL-5 from microarray detection The average intensity of expression in normal tissues (N) andtumors (T) is plotted with their standard deviations (b) Validation ofHOTAIR ESCCAL-1 and ESCCAL-5with independent patient samplesby PCR analysis The amplicons were separated with 2 agarose gel GAPDH was used as an internal control Significance is lowast119875 lt 005lowastlowast
119875 lt 001
Table 1 Validation of selected differential expression of mRNAs in esophageal squamous cell carcinoma in independent studies
Probe name P value FC Regulation Gene symbol Genbank accession Independent study ReferenceA 33 P3232692 0005838984 8301461 Up IL24 NM 001185156
Microarray [20]
A 24 P411121 000055 5329484 Up TNFRSF18 NM 148901A 23 P169097 881119864 minus 05 4466178 Up WISP1 NM 080838A 23 P304304 0004822649 3944957 Down ARSF NM 004042A 24 P56363 0003573538 3323955 Down CAB39L NM 030925A 23 P419760 0001041661 3270335 Down CRISP3 NM 006061A 23 P413923 0002921898 454899 Down DMRTA1 NM 022160A 23 P56978 0002093997 5438183 Down PTK6 NM 005975 RNA-seq [21]A 23 P115091 0005171322 3289834 Down RAB25 NM 020387 Q-RT-PCR [22]A 33 P3258542 0001039129 2036035 Down SPINK8 NM 001080525 Microarray [20]
International Journal of Genomics 5
PCR methods in matched-pair tissue samples from threeadditional ESCC patients and the results are consistent withthemicroarray analysis (Figure 2(b)) Interestingly except forHOTAIR other previously reported lncRNAs (ie MALAT-1 PCAT-1 and AFAP1-AS1) are not differentially expressed inour analysis of ESCC Therefore the DE-lncRNAs that wehave identifiedmay be a unique property of ESCC andwe arecurrently using a population-based analysis to characterizethese DE-lncRNAs as potential genomic biomarkers andregulatory elements in the dynamic process leading to ESCC
23 LncRNAs Co-located and Co-expressed with Coding Genesin ESCC LncRNAs have been reported to coordinate theregulation of neighboring coding genes through a ldquolocuscontrolrdquo process [11] We wondered whether such a ldquolocuscontrolrdquo process could operate in ESCC development andtherefore we searched neighboring genes of the 410 DE-lncRNAs in the genomeThemajority (988) of the 410 DE-lncRNAs harbor neighboring coding genes whose genomiclocations are within sim5 kb upstream and sim1 kb downstreamof the lncRNA and may extend to 1000 kb in both directions(Figure 3(a)) Interrogation of 538 coding genes that areneighbors of these DE-lncRNAs (DE-lncRNAs co-locatedgenes) revealed predicted functions in 9 common pathwayssuch as the AP1 transcription factor network integrin-linkedkinase signaling several signaling pathways in adherensjunctions and FOXO family signaling (Figure 3(b))
We asked whether any DE-lncRNAs co-located genes arealso differentially expressed in ESCC Analysis of the DE-lncRNAs co-located genes withDE-mRNAdata set identified76 genomically co-located and differentially co-expressedgenes (Figure 3(c) and Table 2) Strikingly the co-locatedand co-expressed genes with DE-lncRNAs may be involvedin ether lipid metabolism pathways by the participation ofthe LPCAT1 gene encoding lysophosphatidylcholine acyl-transferase1 and the PLD1 gene encoding phospholipase D1(Figure 3(c)) The lncRNA ESCCAL-337 (chr3171506370-171528740) was downregulated in ESCC and located at22068 bp downstream of the PLD1 gene whose expres-sion was also decreased in ESCC In contrast the lncRNAESCCAL-356 (chr51544500-1567142 reverse strand) wasdownregulated in ESCC and located at 21250 bp upstreamof LPCAT1 whose expression was upregulated in ESCC(Figure 3(c)) LPCAT1 modulates phospholipid compositionby catalyzing lysophosphatidylcholine into phosphatidyl-choline and overexpression of LPCAT1was reported to createfavorable conditions for cancer cell proliferation [27 28]Therefore at least two of the DE-lncRNAs have the potentialto contribute to ESCC by a ldquolocus controlrdquo process withneighboring coding genes
In conclusion we performed a genome-wide survey ofthe expression of lncRNAs and coding mRNAs from pairedsamples of primary neoplastic tissue and adjacent non-neoplastic normal tissue from four individuals The overalltranscriptomic landscape (both lncRNAs andmRNAs) is ableto distinguish malignant from normal tissue in each personWe discovered a set of differentially expressed lncRNAsand their co-located and co-expressed coding mRNAs and
demonstrated that lncRNAs may be involved in ether lipidmetabolism in ESCC Our study provides genomic supportfor a model of a ldquolocus controlrdquo process in ESCC and aframework for further experimental study
3 Materials and Methods
31 Specimens Written informed consent was obtained frompatients before surgery and the study protocol was approvedby the Institutional Review Board for the use of human sub-jects at Zhengzhou Hospital Primary tumors and adjacentnonneoplastic tissues were obtained frompatients with ESCCwhounderwent surgical treatment at LinxianHospital inMay2012 All tissues were frozen in liquid nitrogen immediatelyafter surgical resection None of the patients had priorchemotherapy or radiotherapy nor did they have any otherserious diseases All ESCC tissues were histopathologicallydiagnosed by at least two independent senior pathologists
32 Microarray Hybridization Total RNAs were extractedusing Trizol reagent following manufacturerrsquos instructions(Invitrogen Carlsbad CA USA) The quality of RNAswas measured with a 2100 Bioanalyzer (Agilent technologyUSA) Input of 100 ng of total RNA was used to generateCyanine-3 labeled cRNA according to theAgilentOne-ColorMicroarray-Based Gene Expression Analysis Low for InputQuick Amp Labeling kit (v60) Samples were hybridizedon Agilent SurePrint G3 Human GE 8 times 60K Microarray(Design ID 028004) Arrays were scanned with the AgilentDNAMicroarray Scanner at a 3120583m scan resolution and datawere processed with Agilent Feature Extraction 11011 Themicroarray data discussed in this article have been depositedin National Center for Biotechnology Information (NCBI)Gene ExpressionOmnibus (GEO) and are accessible through(GEO) Series accession number GSE45350 (httpwwwncbinlmnihgovgeoqueryacccgiacc=GSE45350)
33 Validation by Polymerase Chain Reaction (PCR) PCRanalysis was performed on additional matched ESCC andadjacent nonneoplastic tissues for selected lncRNAs Theprimer sequences for PCR are as follows HOTAIR forward51015840-GGTAGAAAAAGCAACCACGAAGC-31015840 and reverse51015840-ACATAAACCTCTGTCTGTGAGTGCC-31015840 ESSCAL-1(chr876121095-76189420 reverse strand) forward 51015840-CCA-GACAGCAGCAAAGCAAT-31015840 and reverse 51015840-GGAAGC-AGCAAATGTGTCCAT-31015840 ESSCAL-5 (chr2216585154-216585719 forward strand) forward 51015840-TACCAACATTGT-CCACCGGG-31015840 and reverse 51015840-GCTGATGACAGTCCC-TTGCT-31015840 GAPDH was used as a control forward 51015840-CCG-GGAAACTGTGGCGTGATGG-31015840 and reverse 51015840-AGG-TGGAGGAGTGGGTGTCGCTGTT-31015840 The thermocycleconditions are as follows initial denaturation at 95∘C for10 minutes followed by 94∘C for 45 seconds 65∘C for 30seconds and 72∘C for 1 minute for 15 cycles Then theannealing temperature was reduced by 05∘Ccycle for thenext 14 cycles and the amplification was finished withanother 24 cycles with the annealing temperature at 58∘C
6 International Journal of Genomics
lncRNAs
Gene b Gene c
20
40
60
80
100
0lncRNAs
Gene a
lncRNAs
Number of associated genes per region0 1 2G
enom
ic re
gion
s of l
ncRN
As (
)
(a)
AP-1 transcription factor network
Integrin-linked kinase signalingRegulation of CDC42 activity
Posttranslational regulation of adherens junction stability and disassemblyN-cadherin signaling events
E-cadherin signaling in the nascent adherens junctionStabilization and expansion of the E-cadherin adherens junction
E-cadherin signaling eventsFOXO family signaling
1176
1105904
511466
448448440
422
(b)
DE-lncRNAsco-located genes
DE-mRNAs
538 330776
ESCCAL-356
ESCCAL-337
Chr3
Chr5
Lipid metabolism
PLD1
LPCAT1
(c)
Figure 3 Identification of lncRNAs co-located and co-expressed neighboring genes in esophageal squamous cell carcinoma (ESCC)(a) Identification of neighboring genes of the DE-lncRNAs The genomic coordinate information of 410 DE-lncRNAs was used to searchneighboring genes whose genomic locations are within sim5 kb upstream and sim1 kb downstream of the lncRNA and may extend to 1000 kbin both directions using GREAT software (httpbejeranostanfordedugreatpublichtmlindexphp) The percentage of DE-lncRNAsharboring zero one or two neighboring genes is presented (b) Gene Ontology (GO) enrichment analysis of lncRNAs co-located genesIdentified gene enriched pathwaysterms are listed on the left the length of horizontal bars and the numbers on the right indicate thepercentage of genes involved in each pathwayterm (c) LncRNAs co-located and co-expressed coding mRNAs Overlap of 538 DE-lncRNAco-located genes with 3307 DE-mRNAs in microarrays identified 76 lncRNAs co-located and co-expressed coding mRNAs (list in Table 2)GO enrichment analysis suggests phospholipase D1 (PLD1) and lysophosphatidylcholine acyltransferase1 (LPCAT1) are involved in ether lipidmetabolism pathway Genomic location shows that PLD1 is located at minus22068 bp upstream of ESCCAL-337 lncRNA on Chr 3 and LPCAT1is at minus21250 bp upstream of ESCCAL-356 lncRNA on Chr 5
International Journal of Genomics 7
Table2Listof
identifi
edco-lo
catedandco-expressed
genesw
ithdifferentially
expressedlncR
NAsinES
CC
LncR
NA
nam
eG
enom
ic co
ordi
nate
Ln
cRN
Aex
pres
sion
LncR
NA
asso
ciat
edge
nes
Gen
eexp
ress
ion
Gen
esym
bol
Gen
bank
acce
ssio
nES
CCA
L-177
chr1
205404138
-205404079
Up
CDK18
(minus69575
)LE
MD
1(minus12895
)U
pLE
MD1
NM001001552
ESC
CAL-348
chr1
222587441
-222587382
Up
DU
SP10
(minus671896
)H
HIP
L2(+134032
)U
pD
USP10
NM007207
ESC
CAL-31
chr1
225237693
-225237752
Up
DN
AH14
(+120367
)LB
R(+378092
)U
pD
NA
H14
NM001373
ESC
CAL-159
chr1
225240300
-225240359
Up
DN
AH14
(+122974
)LB
R(+375485
)U
pD
NA
H14
NM001373
ESC
CAL-327
chr1
89887111
-89887052
Up
LRRC8
B(minus103315
)G
BP6
(+57646
)D
own
GBP6
NM198460
XLO
C000915
chr1
91295528
-91295469
Up
BARH
L2(minus112705
)ZN
F644
(+192313
2N
M020063
ESC
CAL-65
chr11
2017146
-2017205
Up
MRP
L23
(+48674
)IG
F2(+145165
)U
pIG
F2N
M000612
ESC
CAL-19
chr12
66204406
-66204347
Up
HM
GA
2(minus13863
)M
SRB3
(+531610
)U
pH
MG
A2
NM003484
XLO
C011548
chr15
81953024
-81952965
Up
TMC3
(minus286577
)M
EX3
B(+385366
)U
pM
EX3
BN
M032246
ESC
CAL-102
chr17
6766871
-6766930
Up
ALO
X12
(minus132483
)TE
KT1
(minus31841
)12
NM000697
ESC
CAL-342
chr17
78302158
-78302217
Up
END
OV
(minus86779
)RN
F213
(+67521
)U
pRN
F213
NM020954
ESC
CAL-342
chr17
78302158
-78302217
Up
END
OV
(minus86779
)RN
F213
(+67521
)001164638
ESC
CAL-99
chr2
102034125
-102034066
Up
CREG
2(minus30131
)RF
X8(+57069
)U
pCR
EG2
NM153836
ESC
CAL-5
chr2
216585455
-216585514
Up
FN1
(minus284694
)M
REG
(+292861
)D
own
MRE
GN
M018000
ESC
CAL-5
chr2
216585455
-216585514
Up
FN1
(minus284694
)M
REG
(+292861
)U
pFN1
NM054034
ESC
CAL-288
chr2
27789661
-27789602
Up
ZNF512
(minus16261
)G
CKR
(+69926
)D
own
GCK
RN
M001486
ESC
CAL-344
chr2
37327299
-37327358
Up
CCD
C75
(+15735
)EI
F2A
K2
(+56861
)U
pEI
F2A
K2N
M001135652
ESC
CAL-10
chr22
48086469
-48086528
Up
FAM
19A
5(minus798789
)TB
C1D22
A(+927981
)U
pFA
M19
A5
NM015381
ESC
CAL-81
chr5
141710377
-141710436
Up
SPRY4
(minus5787
)FG
F1(+355246
)U
pFG
F1N
M000800
ESC
CAL-81
chr5
141710377
-141710436
Up
SPRY4
(minus5787
)FG
F1(+355246
)U
pSP
RY4
NM030964
ESC
CAL-204
chr6
126699769
-126699828
Up
RSPO3
(minus740249
)CE
NPW
(+38546
)U
pCE
NPW
NM001012507
ESC
CAL-106
chr6
21822910
-21822969
Up
SOX4
(+228968
)PR
L(+480142
)U
pSO
X4N
M003107
ESC
CAL-239
chr8
104258684
-104258625
Up
FZD
6(minus52006
)BA
ALC
(+105734
)U
pBA
ALC
NM024812
ESC
CAL-239
chr8
104258684
-104258625
Up
FZD
6(minus52006
)BA
ALC
(+105734
)U
pFZ
D6
NM003506
ESC
CAL-300
chr1
180918852
-180918793
Dow
nST
X6(+73223
)XP
R1(+317677
)U
pXP
R1N
M004736
XLO
C000515
chr1
200384539
-200384598
Dow
nZN
F281
(minus5403
)KI
F14
(+205293
)U
pKI
F14
NM014875
XLO
C000515
chr1
200384539
-200384598
Dow
nZN
F281
(minus5403
)KI
F14
(+205293
)U
pZN
F281
NM012482
ESC
CAL-77
chr1
201592869
-201592810
Dow
nCS
RP1
(minus116873
)N
AV1
(minus24610
)U
pN
AV1
NM020443
ESC
CAL-76
chr1
90091283
-90091342
Dow
nLR
RC8
C(minus7331
)LR
RC8
B(+100916
)U
pLR
RC8
CN
M032270
ESC
CAL-209
chr10
14549236
-14549177
Dow
nFR
MD4
A(minus176341
)CD
NF
(+330776
)D
own
Dow
n
Dow
nD
own
Dow
n
CDN
FN
M001029954
ESC
CAL-284
chr10
44848467
-44848526
P1(minus562632
)CX
CL12
(+32048
)CX
CL12
NM199168
ESC
CAL-256
chr11
117671760
-117671701
Dow
nD
SCA
ML1
(minus3755
)D
SCA
ML1
NM020693
ESC
CAL-254
chr11
126219961
-126220020
Dow
nST3
GA
L4(minus5549
)D
CPS
(+46344
)ST3
GA
L4N
M006278
ESC
CAL-307
chr11
14975924
-14975983
Dow
nCY
P2R1
(minus62203
)CA
LCA
(+17878
)CY
P2R1
NM024514
ESC
CAL-105
chr11
17366661
-17366720
Dow
nB7
H6
(minus6618
)N
UCB2
(+68405
)D
own
NU
CB2
NM005013
ESC
CAL-232
chr12
131245982
-131245923
Dow
nRI
MBP2
(minus243491
)ST
X2(+77858
)D
own
RIM
BP2
NM015347
ESC
CAL-24
chr12
132824534
-132824475
Dow
nG
ALN
T9(+81400
)N
OC4
L(+195512
)U
pG
ALN
T9N
M021808
Dow
nH
NRN
PA3
)D
own
BARH
L
Dow
nA
LOX
Dow
nEN
DO
VN
MN
M
8 International Journal of Genomics
Table2Con
tinued
LncR
NA
nam
eG
enom
ic co
ordi
nate
sLn
cRN
Aex
pres
sion
LncR
NA
asso
ciat
edge
nes
Gen
eexp
ress
ion
Gen
esym
bol
Gen
bank
acce
ssio
nES
CCA
L-122
chr12
2952238
-2952297
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
n
Dow
n
Dow
n
Dow
nD
own
Dow
n
Dow
n
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
NRI
P 2(minus8047
)FO
XM1
(+34053
)FO
XM1
NM202002
ESC
CAL-338
chr12
32101291
-32101232
BICD1
(minus158923
)H3
F3C
( minus156087
)BI
CD1
NM001714
ESC
CAL-275
chr14
71180731
-71180790
TTC9
(+72257
)M
AP3
K9(+95127
)M
AP3
K 9N
M033141
ESC
CAL-275
chr 14
71180731
-71180790
TTC9
(+72257
)M
AP 3
K 9(+95127
)TT
C9N
M015351
ESC
CAL-121
chr15
78088340
-78088281
LIN
GO1
( minus163602
)TB
C1D2
B(+281683
)LI
NG
O1
NM032808
ESC
CAL-231
chr 16
88381397
-88381338
ZNF469
( minus112511
)BA
NP
(+396330
)ZN
F469
NM001127464
ESC
CAL-74
chr2
47548478
-47548537
CALM2
(minus144768
)EP
CAM
(minus47779
)EP
CAM
NM002354
ESC
CAL-152
chr20
56839403
-56839344
PMEP
A1(minus554343
)PP
P 4R1
L(+45121
)PM
EPA1
NM020182
ESC
CAL-337
chr 3
171506465
-171506406
TNIK
(minus328239
)PL
D1
(+22068
)PL
D1
NM002662
ESC
CAL-80
chr3
177935638
- 177935579
KCN
MB2
(minus318615
)KC
NM
B 2N
M181361
ESC
CAL-90
chr 3
195441846
-195441787
MU
C20
(minus5936
)SD
HA
P 2(+56907
)M
UC20
NM001098516
ESC
CAL-59
chr 3
64855048
-64855107
AD
AM
TS9
(minus181713
)A
DA
MTS9
NM182920
ESC
CAL-72
chr 4
74922502
-74922443
CXCL3
(minus17983
)CX
CL2
(+42524
)CX
CL3
NM002090
ESC
CAL-79
chr4
79626460
- 79626401
BMP2
K(minus71101
)A
NX
A3
(+153689
)A
NX
A3
NM005139
ESC
CAL-253
chr4
8357097
-8357038
ACO
X3(+85384
)H
TRA3
(+85579
)AC
OX 3
NM003501
ESC
CAL-179
chr4
8359416
- 8359357
ACO
X3(+83065
)H
TRA3
(+87898
)H
TRA3
NM053044
ESC
CAL-11
chr4
84299039
- 84299098
HPS
E(minus43035
)H
ELQ
(+77956
)H
PSE
NM006665
ESC
CAL-41
chr4
8512901
- 8512960
GPR78
(minus69286
)M
ETTL19
(+70399
)G
PR78
NM080819
ESC
CAL-353
chr 5
1175322
-1175263
SLC12
A7
(minus63121
)SL
C6A19
(minus26417
)SL
C12
A7
NM006598
ESC
CAL-260
chr5
131808618
-131808677
IRF 1
(+17817
)SL
C22
A5
(+103247
)IR
F 1N
M002198
ESC
CAL-132
chr5
134578804
- 134578863
PITX1
(minus208870
)H2
AFY
(+156094
)PI
TX1
NM002653
ESC
CAL-4
chr5
141732627
-141732568
SPRY4
(minus27978
)FG
F1(+333055
)SP
RY4
NM030964
ESC
CAL-356
chr 5
1545355
-1545296
LPCA
T1( minus21250
)M
RPL36
(+254630
)LP
CAT 1
NM024830
XLO
C004881
chr5
72570680
-72570621
TMEM174
(+101628
)FO
XD1
(+173701
)up
FOXD1
NM004472
ESC
CAL-36
chr6
106899513
-106899572
ATG5
( minus125848
)A
IM1
(minus59762
)A
IM1
NM001624
XLO
C005849
chr6
138145112
-138145053
OLI
G3
( minus329552
)TN
FAIP
3(minus43498
)TN
FAIP3
NM006290
ESC
CAL-262
chr6
2283830
-2283771
GM
DS
(minus37933
)M
YLK4
(+467353
)G
MD
SN
M001500
ESC
CAL-120
chr6
29716817
-29716758
LOC554223
(minus42895
)H
LA-F
(+25671
)H
LA-F
NM018950
ESC
CAL-257
chr 6
29988410
-29988469
ZNRD1
( minus40596
)H
LA-J
(+14223
)H
LA-J
NR024240
ESC
CAL-73
chr6
36126663
-36126722
BRPF3
( minus37857
)M
APK13
(+28431
)M
APK13
NM002754
ESC
CAL-97
chr 6
40305634
-40305575
MO
CS1
(minus410150
)LR
FN2
(+249521
)LR
FN2
NM020737
ESC
CAL-333
chr 6
72018004
-72018063
RIM
S1(minus578616
)O
GFR
L1(+19557
)O
GFR
L1N
M024576
ESC
CAL-115
chr7
12593405
-12593464
VW
DE
(minus149583
)SC
IN(minus16768
)SC
INN
M033128
ESC
CAL-68
chr7
139487166
-139487225
TBX
AS1
(minus41756
)H
IPK2
(minus9503
)H
IPK2
AF207702
ESC
CAL-87
chr7
19958800
- 19958741
TMEM196
(minus146367
)M
ACC1
(+298242
)M
ACC1
NM182762
ESC
CAL-58
chr8
16355082
-16355141
MSR1
(minus304812
)FG
F 20
(+504562
)M
SR1
NM002445
ESC
CAL-130
chr8
37189082
-37189141
ZNF703
(minus364189
)KC
NU1
(+547270
)ZN
F703
NM025069
ESC
CAL-199
8
chr 8
38623612
-38623671
RNF 5
P 1(minus164867
)TA
CC 1
(minus21080
)TA
CC1
BC041391
ESC
CAL-
chr9
34668311
-34668252
CCL27
(minus5593
)C
CL19
(+22992
)C
CL19
NM006274
NotesR
ed-highlighted
genesa
rewho
seexpressio
nsaresig
nificantly
changedin
Esop
hagealSquamou
sCellC
arcino
ma(ESC
C)G
reen
color-high
lighted
rowsgenesinvolvedin
lipid
metabolism
predictedwith
GOenric
hmentanalysis
Yellowcolor-high
lighted
rowsTh
eexpressionof
lncR
NAES
CCAL-5was
valid
ated
byPC
R
International Journal of Genomics 9
Final extension was at 72∘C for 10 minutes The ampliconswere resolved in 2 agarose gel
34 Bioinformatic Analysis Intensity data were exportedto GeneSpring 120 (Agilent Technologies Santa ClaraCA USA) for quantile normalization and the analysis ofdifferentially expressed long noncoding RNAs and codingRNAs Paired 119905-test analysis was used to obtain probe setswhose magnitude of change in expression of RNAs betweenESCC tissue and adjacent normal esophageal tissuewas eithergreater or less than 20 fold and 119875 value lt 005 (119875 values werecorrected formultiple testing using themethod of Benjamini-Hochberg) The normalized data containing 42544 probeswere further analyzed using the R program All controlprobes were removed We then defined the coding (ldquoNM rdquoldquoXM rdquo) and noncoding (ldquolincRNArdquo ldquoNR rdquo and ldquoXR rdquo) genesin the normalized data according to the definition of RefSeqaccession format (httpwwwncbinlmnihgovprojectsRefSeqkeyhtml) Differentially expressed long noncodingRNAs (DE-lncRNAs) and coding RNAs (DE-mRNAs)were further identified The landscapes of the wholetranscriptome (lncRNAs + coding RNAs) or all lncRNAsor all coding RNAs were analyzed with gene expressiondynamic inspector (GEDI)
35 Co-Location and Co-Expression Analysis between DE-lncRNAs and DE-mRNAs Genomic coordinates of DE-lncRNAs were imported to GREAT software (httpbejeranostanfordedugreatpublichtmlindexphp) for co-locationanalysis Neighboring coding genes were then matched withDE-mRNAs to obtain a co-expression dataset Three sub-groups of genes (DE-lncRNA co-located genes DE-mRNAsand co-expressed genes) were used for gene expressionnetwork analysis using Cytoscape software (v283)
Abbreviations
AFAP1-AS1 Actin filament-associated protein 1 antisenseRNA
Wei Cao and Wei Wu contributed equally to this project
Acknowledgments
The National Natural Science Foundation of China (no81171992) and the Zhengzhou Science and Technology
Programme (121PPTGG494-8) supported this work Theauthors thank Anne Haegert atThe Laboratory for AdvancedGenome Analysis at the Vancouver Prostate Centre Van-couver Canada for expert technical support They alsothank GenomeSky Inc Canada for providing computationalconsultation
References
[1] J Ferlay H R Shin F Bray D Forman C Mathers and DM Parkin ldquoEstimates of worldwide burden of cancer in 2008GLOBOCAN2008rdquo International Journal of Cancer vol 127 no12 pp 2893ndash2917 2010
[2] W Wu and J A Chan ldquoUnderstanding the role of longnoncoding RNAs in the cancer genomerdquo in Next GenerationSequencing in Cancer Research-Decoding Cancer Genome WWu and H Choudhry Eds pp 199ndash215 Springer New YorkNY USA 2013
[3] J J Hao T Gong Y Zhang et al ldquoCharacterization of generearrangements resulted from genomic structural aberrationsin human esophageal squamous cell carcinoma KYSE150 cellsrdquoGene vol 513 no 1 pp 196ndash201 2013
[4] M Kano N Seki N Kikkawa et al ldquoMiR-145 miR-133aand miR-133b tumor-suppressive miRNAs target FSCN1 inesophageal squamous cell carcinomardquo International Journal ofCancer vol 127 no 12 pp 2804ndash2814 2010
[5] H SuNHuHH Yang et al ldquoGlobal gene expression profilingand validation in esophageal squamous cell carcinoma and itsassociation with clinical phenotypesrdquo Clinical Cancer Researchvol 17 no 9 pp 2955ndash2966 2011
[6] D M Greenawalt C Duong G K Smyth et al ldquoGeneexpression profiling of esophageal cancer comparative analysisof Barrettrsquos esophagus adenocarcinoma and squamous cellcarcinomardquo International Journal of Cancer vol 120 no 9 pp1914ndash1921 2007
[7] I Dunham A Kundaje S F Aldred et al ldquoAn integratedencyclopedia of DNA elements in the human genomerdquo Naturevol 489 no 7414 pp 57ndash74 2012
[8] E A Gibb C J Brown and W L Lam ldquoThe functional roleof long non-coding RNA in human carcinomasrdquo MolecularCancer vol 10 article 38 2011
[9] A L Brunner A H Beck B Edris et al ldquoTranscriptionalprofiling of lncRNAs and novel transcribed regions across adiverse panel of archived human cancersrdquo Genome Biology vol13 no 8 article R75 2012
[10] M Guttman and J L Rinn ldquoModular regulatory principles oflarge non-coding RNAsrdquo Nature vol 482 no 7385 pp 339ndash346 2012
[11] K C Wang Y W Yang B Liu et al ldquoA long noncodingRNA maintains active chromatin to coordinate homeotic geneexpressionrdquo Nature vol 472 no 7341 pp 120ndash126 2011
[12] A Dean ldquoOn a chromosome far far away LCRs and geneexpressionrdquo Trends in Genetics vol 22 no 1 pp 38ndash45 2006
[13] P Ji S Diederichs W Wang et al ldquoMALAT-1 a novel noncod-ing RNA and thymosin 1205734 predict metastasis and survival inearly-stage non-small cell lung cancerrdquo Oncogene vol 22 no39 pp 8031ndash8041 2003
[14] R A Gupta N Shah K C Wang et al ldquoLong non-codingRNA HOTAIR reprograms chromatin state to promote cancermetastasisrdquo Nature vol 464 no 7291 pp 1071ndash1076 2010
10 International Journal of Genomics
[15] R Kogo T Shimamura K Mimori et al ldquoLong noncodingRNAHOTAIR regulates polycomb-dependent chromatinmod-ification and is associated with poor prognosis in colorectalcancersrdquo Cancer Research vol 71 no 20 pp 6320ndash6326 2011
[16] K Kim I Jutooru G Chadalapaka et al ldquoHOTAIR is anegative prognostic factor and exhibits pro-oncogenic activityin pancreatic cancerrdquo Oncogene vol 32 pp 1616ndash1625 2013
[17] D Li J Feng T Wu et al ldquoLong intergenic noncoding RNAHOTAIR is overexpressed and regulates PTEN methylation inlaryngeal squamous cell carcinomardquo The American Journal ofPathology vol 182 no 1 pp 64ndash70 2013
[18] A Bhan I Hussain K I Ansari S Kasiri A Bashyal and S SMandal ldquoAntisense transcript long noncoding RNA (lncRNA)HOTAIR is transcriptionally induced by estradiolrdquo Journal ofMolecular Biology vol 425 no 19 pp 3707ndash3722 2013
[19] D Cejka D Losert and V Wacheck ldquoShort interfering RNA(siRNA) tool or therapeuticrdquo Clinical Science vol 110 no 1pp 47ndash58 2006
[20] J R Prensner and A M Chinnaiyan ldquoThe emergence oflncRNAs in cancer biologyrdquo Cancer Discovery vol 1 pp 391ndash407 2011
[21] W Wu T D Bhagat X Yang et al ldquoHypomethylation ofnoncoding DNA regions and overexpression of the long non-coding RNA AFAP1-AS1 in Barrettrsquos esophagus and esophagealadenocarcinomardquoGastroenterology vol 144 no 5 pp 956ndash9662013
[22] E Ozgur U Mert M Isin M Okutan N Dalay and U GezerldquoDifferential expression of long non-coding RNAs duringgenotoxic stress-induced apoptosis in HeLa and MCF-7 cellsrdquoClinical and Experimental Medicine vol 13 no 2 pp 119ndash1262013
[23] J D Li Q C Feng and J S Li ldquoDifferential gene expressionprofiling of oesophageal squamous cell carcinoma by dnamicroarray and bioinformatics analysisrdquo Journal of Interna-tional Medical Research vol 38 no 6 pp 1904ndash1912 2010
[24] S Ma J Y Bao P S Kwan et al ldquoIdentification of PTK6via RNA sequencing analysis as a suppressor of esophagealsquamous cell carcinomardquo Gastroenterology vol 143 pp 675ndash686 2012
[25] M Tong K W Chan J Y Bao et al ldquoRab25 is a tumorsuppressor gene with antiangiogenic and anti-invasive activitiesin esophageal squamous cell carcinomardquo Cancer Research vol72 no 22 pp 6024ndash6035 2012
[26] Y J Geng S L Xie Q Li J Ma and G Y Wang ldquoLargeintervening non-coding RNA HOTAIR is associated withhepatocellular carcinoma progressionrdquo Journal of InternationalMedical Research vol 39 no 6 pp 2119ndash2128 2011
[27] X Zhou T J Lawrence Z He C R Pound J Mao andS A Bigler ldquoThe expression level of lysophosphatidylcholineacyltransferase 1 (LPCAT1) correlates to the progression ofprostate cancerrdquo Experimental andMolecular Pathology vol 92no 1 pp 105ndash110 2012
[28] Y Morita T Sakaguchi K Ikegami et al ldquoLysophosphatidyl-choline acyltransferase 1 altered phospholipid composition andregulated hepatoma progressionrdquo Journal of Hepatology vol 59no 2 pp 292ndash299 2013
PCR methods in matched-pair tissue samples from threeadditional ESCC patients and the results are consistent withthemicroarray analysis (Figure 2(b)) Interestingly except forHOTAIR other previously reported lncRNAs (ie MALAT-1 PCAT-1 and AFAP1-AS1) are not differentially expressed inour analysis of ESCC Therefore the DE-lncRNAs that wehave identifiedmay be a unique property of ESCC andwe arecurrently using a population-based analysis to characterizethese DE-lncRNAs as potential genomic biomarkers andregulatory elements in the dynamic process leading to ESCC
23 LncRNAs Co-located and Co-expressed with Coding Genesin ESCC LncRNAs have been reported to coordinate theregulation of neighboring coding genes through a ldquolocuscontrolrdquo process [11] We wondered whether such a ldquolocuscontrolrdquo process could operate in ESCC development andtherefore we searched neighboring genes of the 410 DE-lncRNAs in the genomeThemajority (988) of the 410 DE-lncRNAs harbor neighboring coding genes whose genomiclocations are within sim5 kb upstream and sim1 kb downstreamof the lncRNA and may extend to 1000 kb in both directions(Figure 3(a)) Interrogation of 538 coding genes that areneighbors of these DE-lncRNAs (DE-lncRNAs co-locatedgenes) revealed predicted functions in 9 common pathwayssuch as the AP1 transcription factor network integrin-linkedkinase signaling several signaling pathways in adherensjunctions and FOXO family signaling (Figure 3(b))
We asked whether any DE-lncRNAs co-located genes arealso differentially expressed in ESCC Analysis of the DE-lncRNAs co-located genes withDE-mRNAdata set identified76 genomically co-located and differentially co-expressedgenes (Figure 3(c) and Table 2) Strikingly the co-locatedand co-expressed genes with DE-lncRNAs may be involvedin ether lipid metabolism pathways by the participation ofthe LPCAT1 gene encoding lysophosphatidylcholine acyl-transferase1 and the PLD1 gene encoding phospholipase D1(Figure 3(c)) The lncRNA ESCCAL-337 (chr3171506370-171528740) was downregulated in ESCC and located at22068 bp downstream of the PLD1 gene whose expres-sion was also decreased in ESCC In contrast the lncRNAESCCAL-356 (chr51544500-1567142 reverse strand) wasdownregulated in ESCC and located at 21250 bp upstreamof LPCAT1 whose expression was upregulated in ESCC(Figure 3(c)) LPCAT1 modulates phospholipid compositionby catalyzing lysophosphatidylcholine into phosphatidyl-choline and overexpression of LPCAT1was reported to createfavorable conditions for cancer cell proliferation [27 28]Therefore at least two of the DE-lncRNAs have the potentialto contribute to ESCC by a ldquolocus controlrdquo process withneighboring coding genes
In conclusion we performed a genome-wide survey ofthe expression of lncRNAs and coding mRNAs from pairedsamples of primary neoplastic tissue and adjacent non-neoplastic normal tissue from four individuals The overalltranscriptomic landscape (both lncRNAs andmRNAs) is ableto distinguish malignant from normal tissue in each personWe discovered a set of differentially expressed lncRNAsand their co-located and co-expressed coding mRNAs and
demonstrated that lncRNAs may be involved in ether lipidmetabolism in ESCC Our study provides genomic supportfor a model of a ldquolocus controlrdquo process in ESCC and aframework for further experimental study
3 Materials and Methods
31 Specimens Written informed consent was obtained frompatients before surgery and the study protocol was approvedby the Institutional Review Board for the use of human sub-jects at Zhengzhou Hospital Primary tumors and adjacentnonneoplastic tissues were obtained frompatients with ESCCwhounderwent surgical treatment at LinxianHospital inMay2012 All tissues were frozen in liquid nitrogen immediatelyafter surgical resection None of the patients had priorchemotherapy or radiotherapy nor did they have any otherserious diseases All ESCC tissues were histopathologicallydiagnosed by at least two independent senior pathologists
32 Microarray Hybridization Total RNAs were extractedusing Trizol reagent following manufacturerrsquos instructions(Invitrogen Carlsbad CA USA) The quality of RNAswas measured with a 2100 Bioanalyzer (Agilent technologyUSA) Input of 100 ng of total RNA was used to generateCyanine-3 labeled cRNA according to theAgilentOne-ColorMicroarray-Based Gene Expression Analysis Low for InputQuick Amp Labeling kit (v60) Samples were hybridizedon Agilent SurePrint G3 Human GE 8 times 60K Microarray(Design ID 028004) Arrays were scanned with the AgilentDNAMicroarray Scanner at a 3120583m scan resolution and datawere processed with Agilent Feature Extraction 11011 Themicroarray data discussed in this article have been depositedin National Center for Biotechnology Information (NCBI)Gene ExpressionOmnibus (GEO) and are accessible through(GEO) Series accession number GSE45350 (httpwwwncbinlmnihgovgeoqueryacccgiacc=GSE45350)
33 Validation by Polymerase Chain Reaction (PCR) PCRanalysis was performed on additional matched ESCC andadjacent nonneoplastic tissues for selected lncRNAs Theprimer sequences for PCR are as follows HOTAIR forward51015840-GGTAGAAAAAGCAACCACGAAGC-31015840 and reverse51015840-ACATAAACCTCTGTCTGTGAGTGCC-31015840 ESSCAL-1(chr876121095-76189420 reverse strand) forward 51015840-CCA-GACAGCAGCAAAGCAAT-31015840 and reverse 51015840-GGAAGC-AGCAAATGTGTCCAT-31015840 ESSCAL-5 (chr2216585154-216585719 forward strand) forward 51015840-TACCAACATTGT-CCACCGGG-31015840 and reverse 51015840-GCTGATGACAGTCCC-TTGCT-31015840 GAPDH was used as a control forward 51015840-CCG-GGAAACTGTGGCGTGATGG-31015840 and reverse 51015840-AGG-TGGAGGAGTGGGTGTCGCTGTT-31015840 The thermocycleconditions are as follows initial denaturation at 95∘C for10 minutes followed by 94∘C for 45 seconds 65∘C for 30seconds and 72∘C for 1 minute for 15 cycles Then theannealing temperature was reduced by 05∘Ccycle for thenext 14 cycles and the amplification was finished withanother 24 cycles with the annealing temperature at 58∘C
6 International Journal of Genomics
lncRNAs
Gene b Gene c
20
40
60
80
100
0lncRNAs
Gene a
lncRNAs
Number of associated genes per region0 1 2G
enom
ic re
gion
s of l
ncRN
As (
)
(a)
AP-1 transcription factor network
Integrin-linked kinase signalingRegulation of CDC42 activity
Posttranslational regulation of adherens junction stability and disassemblyN-cadherin signaling events
E-cadherin signaling in the nascent adherens junctionStabilization and expansion of the E-cadherin adherens junction
E-cadherin signaling eventsFOXO family signaling
1176
1105904
511466
448448440
422
(b)
DE-lncRNAsco-located genes
DE-mRNAs
538 330776
ESCCAL-356
ESCCAL-337
Chr3
Chr5
Lipid metabolism
PLD1
LPCAT1
(c)
Figure 3 Identification of lncRNAs co-located and co-expressed neighboring genes in esophageal squamous cell carcinoma (ESCC)(a) Identification of neighboring genes of the DE-lncRNAs The genomic coordinate information of 410 DE-lncRNAs was used to searchneighboring genes whose genomic locations are within sim5 kb upstream and sim1 kb downstream of the lncRNA and may extend to 1000 kbin both directions using GREAT software (httpbejeranostanfordedugreatpublichtmlindexphp) The percentage of DE-lncRNAsharboring zero one or two neighboring genes is presented (b) Gene Ontology (GO) enrichment analysis of lncRNAs co-located genesIdentified gene enriched pathwaysterms are listed on the left the length of horizontal bars and the numbers on the right indicate thepercentage of genes involved in each pathwayterm (c) LncRNAs co-located and co-expressed coding mRNAs Overlap of 538 DE-lncRNAco-located genes with 3307 DE-mRNAs in microarrays identified 76 lncRNAs co-located and co-expressed coding mRNAs (list in Table 2)GO enrichment analysis suggests phospholipase D1 (PLD1) and lysophosphatidylcholine acyltransferase1 (LPCAT1) are involved in ether lipidmetabolism pathway Genomic location shows that PLD1 is located at minus22068 bp upstream of ESCCAL-337 lncRNA on Chr 3 and LPCAT1is at minus21250 bp upstream of ESCCAL-356 lncRNA on Chr 5
International Journal of Genomics 7
Table2Listof
identifi
edco-lo
catedandco-expressed
genesw
ithdifferentially
expressedlncR
NAsinES
CC
LncR
NA
nam
eG
enom
ic co
ordi
nate
Ln
cRN
Aex
pres
sion
LncR
NA
asso
ciat
edge
nes
Gen
eexp
ress
ion
Gen
esym
bol
Gen
bank
acce
ssio
nES
CCA
L-177
chr1
205404138
-205404079
Up
CDK18
(minus69575
)LE
MD
1(minus12895
)U
pLE
MD1
NM001001552
ESC
CAL-348
chr1
222587441
-222587382
Up
DU
SP10
(minus671896
)H
HIP
L2(+134032
)U
pD
USP10
NM007207
ESC
CAL-31
chr1
225237693
-225237752
Up
DN
AH14
(+120367
)LB
R(+378092
)U
pD
NA
H14
NM001373
ESC
CAL-159
chr1
225240300
-225240359
Up
DN
AH14
(+122974
)LB
R(+375485
)U
pD
NA
H14
NM001373
ESC
CAL-327
chr1
89887111
-89887052
Up
LRRC8
B(minus103315
)G
BP6
(+57646
)D
own
GBP6
NM198460
XLO
C000915
chr1
91295528
-91295469
Up
BARH
L2(minus112705
)ZN
F644
(+192313
2N
M020063
ESC
CAL-65
chr11
2017146
-2017205
Up
MRP
L23
(+48674
)IG
F2(+145165
)U
pIG
F2N
M000612
ESC
CAL-19
chr12
66204406
-66204347
Up
HM
GA
2(minus13863
)M
SRB3
(+531610
)U
pH
MG
A2
NM003484
XLO
C011548
chr15
81953024
-81952965
Up
TMC3
(minus286577
)M
EX3
B(+385366
)U
pM
EX3
BN
M032246
ESC
CAL-102
chr17
6766871
-6766930
Up
ALO
X12
(minus132483
)TE
KT1
(minus31841
)12
NM000697
ESC
CAL-342
chr17
78302158
-78302217
Up
END
OV
(minus86779
)RN
F213
(+67521
)U
pRN
F213
NM020954
ESC
CAL-342
chr17
78302158
-78302217
Up
END
OV
(minus86779
)RN
F213
(+67521
)001164638
ESC
CAL-99
chr2
102034125
-102034066
Up
CREG
2(minus30131
)RF
X8(+57069
)U
pCR
EG2
NM153836
ESC
CAL-5
chr2
216585455
-216585514
Up
FN1
(minus284694
)M
REG
(+292861
)D
own
MRE
GN
M018000
ESC
CAL-5
chr2
216585455
-216585514
Up
FN1
(minus284694
)M
REG
(+292861
)U
pFN1
NM054034
ESC
CAL-288
chr2
27789661
-27789602
Up
ZNF512
(minus16261
)G
CKR
(+69926
)D
own
GCK
RN
M001486
ESC
CAL-344
chr2
37327299
-37327358
Up
CCD
C75
(+15735
)EI
F2A
K2
(+56861
)U
pEI
F2A
K2N
M001135652
ESC
CAL-10
chr22
48086469
-48086528
Up
FAM
19A
5(minus798789
)TB
C1D22
A(+927981
)U
pFA
M19
A5
NM015381
ESC
CAL-81
chr5
141710377
-141710436
Up
SPRY4
(minus5787
)FG
F1(+355246
)U
pFG
F1N
M000800
ESC
CAL-81
chr5
141710377
-141710436
Up
SPRY4
(minus5787
)FG
F1(+355246
)U
pSP
RY4
NM030964
ESC
CAL-204
chr6
126699769
-126699828
Up
RSPO3
(minus740249
)CE
NPW
(+38546
)U
pCE
NPW
NM001012507
ESC
CAL-106
chr6
21822910
-21822969
Up
SOX4
(+228968
)PR
L(+480142
)U
pSO
X4N
M003107
ESC
CAL-239
chr8
104258684
-104258625
Up
FZD
6(minus52006
)BA
ALC
(+105734
)U
pBA
ALC
NM024812
ESC
CAL-239
chr8
104258684
-104258625
Up
FZD
6(minus52006
)BA
ALC
(+105734
)U
pFZ
D6
NM003506
ESC
CAL-300
chr1
180918852
-180918793
Dow
nST
X6(+73223
)XP
R1(+317677
)U
pXP
R1N
M004736
XLO
C000515
chr1
200384539
-200384598
Dow
nZN
F281
(minus5403
)KI
F14
(+205293
)U
pKI
F14
NM014875
XLO
C000515
chr1
200384539
-200384598
Dow
nZN
F281
(minus5403
)KI
F14
(+205293
)U
pZN
F281
NM012482
ESC
CAL-77
chr1
201592869
-201592810
Dow
nCS
RP1
(minus116873
)N
AV1
(minus24610
)U
pN
AV1
NM020443
ESC
CAL-76
chr1
90091283
-90091342
Dow
nLR
RC8
C(minus7331
)LR
RC8
B(+100916
)U
pLR
RC8
CN
M032270
ESC
CAL-209
chr10
14549236
-14549177
Dow
nFR
MD4
A(minus176341
)CD
NF
(+330776
)D
own
Dow
n
Dow
nD
own
Dow
n
CDN
FN
M001029954
ESC
CAL-284
chr10
44848467
-44848526
P1(minus562632
)CX
CL12
(+32048
)CX
CL12
NM199168
ESC
CAL-256
chr11
117671760
-117671701
Dow
nD
SCA
ML1
(minus3755
)D
SCA
ML1
NM020693
ESC
CAL-254
chr11
126219961
-126220020
Dow
nST3
GA
L4(minus5549
)D
CPS
(+46344
)ST3
GA
L4N
M006278
ESC
CAL-307
chr11
14975924
-14975983
Dow
nCY
P2R1
(minus62203
)CA
LCA
(+17878
)CY
P2R1
NM024514
ESC
CAL-105
chr11
17366661
-17366720
Dow
nB7
H6
(minus6618
)N
UCB2
(+68405
)D
own
NU
CB2
NM005013
ESC
CAL-232
chr12
131245982
-131245923
Dow
nRI
MBP2
(minus243491
)ST
X2(+77858
)D
own
RIM
BP2
NM015347
ESC
CAL-24
chr12
132824534
-132824475
Dow
nG
ALN
T9(+81400
)N
OC4
L(+195512
)U
pG
ALN
T9N
M021808
Dow
nH
NRN
PA3
)D
own
BARH
L
Dow
nA
LOX
Dow
nEN
DO
VN
MN
M
8 International Journal of Genomics
Table2Con
tinued
LncR
NA
nam
eG
enom
ic co
ordi
nate
sLn
cRN
Aex
pres
sion
LncR
NA
asso
ciat
edge
nes
Gen
eexp
ress
ion
Gen
esym
bol
Gen
bank
acce
ssio
nES
CCA
L-122
chr12
2952238
-2952297
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
n
Dow
n
Dow
n
Dow
nD
own
Dow
n
Dow
n
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
NRI
P 2(minus8047
)FO
XM1
(+34053
)FO
XM1
NM202002
ESC
CAL-338
chr12
32101291
-32101232
BICD1
(minus158923
)H3
F3C
( minus156087
)BI
CD1
NM001714
ESC
CAL-275
chr14
71180731
-71180790
TTC9
(+72257
)M
AP3
K9(+95127
)M
AP3
K 9N
M033141
ESC
CAL-275
chr 14
71180731
-71180790
TTC9
(+72257
)M
AP 3
K 9(+95127
)TT
C9N
M015351
ESC
CAL-121
chr15
78088340
-78088281
LIN
GO1
( minus163602
)TB
C1D2
B(+281683
)LI
NG
O1
NM032808
ESC
CAL-231
chr 16
88381397
-88381338
ZNF469
( minus112511
)BA
NP
(+396330
)ZN
F469
NM001127464
ESC
CAL-74
chr2
47548478
-47548537
CALM2
(minus144768
)EP
CAM
(minus47779
)EP
CAM
NM002354
ESC
CAL-152
chr20
56839403
-56839344
PMEP
A1(minus554343
)PP
P 4R1
L(+45121
)PM
EPA1
NM020182
ESC
CAL-337
chr 3
171506465
-171506406
TNIK
(minus328239
)PL
D1
(+22068
)PL
D1
NM002662
ESC
CAL-80
chr3
177935638
- 177935579
KCN
MB2
(minus318615
)KC
NM
B 2N
M181361
ESC
CAL-90
chr 3
195441846
-195441787
MU
C20
(minus5936
)SD
HA
P 2(+56907
)M
UC20
NM001098516
ESC
CAL-59
chr 3
64855048
-64855107
AD
AM
TS9
(minus181713
)A
DA
MTS9
NM182920
ESC
CAL-72
chr 4
74922502
-74922443
CXCL3
(minus17983
)CX
CL2
(+42524
)CX
CL3
NM002090
ESC
CAL-79
chr4
79626460
- 79626401
BMP2
K(minus71101
)A
NX
A3
(+153689
)A
NX
A3
NM005139
ESC
CAL-253
chr4
8357097
-8357038
ACO
X3(+85384
)H
TRA3
(+85579
)AC
OX 3
NM003501
ESC
CAL-179
chr4
8359416
- 8359357
ACO
X3(+83065
)H
TRA3
(+87898
)H
TRA3
NM053044
ESC
CAL-11
chr4
84299039
- 84299098
HPS
E(minus43035
)H
ELQ
(+77956
)H
PSE
NM006665
ESC
CAL-41
chr4
8512901
- 8512960
GPR78
(minus69286
)M
ETTL19
(+70399
)G
PR78
NM080819
ESC
CAL-353
chr 5
1175322
-1175263
SLC12
A7
(minus63121
)SL
C6A19
(minus26417
)SL
C12
A7
NM006598
ESC
CAL-260
chr5
131808618
-131808677
IRF 1
(+17817
)SL
C22
A5
(+103247
)IR
F 1N
M002198
ESC
CAL-132
chr5
134578804
- 134578863
PITX1
(minus208870
)H2
AFY
(+156094
)PI
TX1
NM002653
ESC
CAL-4
chr5
141732627
-141732568
SPRY4
(minus27978
)FG
F1(+333055
)SP
RY4
NM030964
ESC
CAL-356
chr 5
1545355
-1545296
LPCA
T1( minus21250
)M
RPL36
(+254630
)LP
CAT 1
NM024830
XLO
C004881
chr5
72570680
-72570621
TMEM174
(+101628
)FO
XD1
(+173701
)up
FOXD1
NM004472
ESC
CAL-36
chr6
106899513
-106899572
ATG5
( minus125848
)A
IM1
(minus59762
)A
IM1
NM001624
XLO
C005849
chr6
138145112
-138145053
OLI
G3
( minus329552
)TN
FAIP
3(minus43498
)TN
FAIP3
NM006290
ESC
CAL-262
chr6
2283830
-2283771
GM
DS
(minus37933
)M
YLK4
(+467353
)G
MD
SN
M001500
ESC
CAL-120
chr6
29716817
-29716758
LOC554223
(minus42895
)H
LA-F
(+25671
)H
LA-F
NM018950
ESC
CAL-257
chr 6
29988410
-29988469
ZNRD1
( minus40596
)H
LA-J
(+14223
)H
LA-J
NR024240
ESC
CAL-73
chr6
36126663
-36126722
BRPF3
( minus37857
)M
APK13
(+28431
)M
APK13
NM002754
ESC
CAL-97
chr 6
40305634
-40305575
MO
CS1
(minus410150
)LR
FN2
(+249521
)LR
FN2
NM020737
ESC
CAL-333
chr 6
72018004
-72018063
RIM
S1(minus578616
)O
GFR
L1(+19557
)O
GFR
L1N
M024576
ESC
CAL-115
chr7
12593405
-12593464
VW
DE
(minus149583
)SC
IN(minus16768
)SC
INN
M033128
ESC
CAL-68
chr7
139487166
-139487225
TBX
AS1
(minus41756
)H
IPK2
(minus9503
)H
IPK2
AF207702
ESC
CAL-87
chr7
19958800
- 19958741
TMEM196
(minus146367
)M
ACC1
(+298242
)M
ACC1
NM182762
ESC
CAL-58
chr8
16355082
-16355141
MSR1
(minus304812
)FG
F 20
(+504562
)M
SR1
NM002445
ESC
CAL-130
chr8
37189082
-37189141
ZNF703
(minus364189
)KC
NU1
(+547270
)ZN
F703
NM025069
ESC
CAL-199
8
chr 8
38623612
-38623671
RNF 5
P 1(minus164867
)TA
CC 1
(minus21080
)TA
CC1
BC041391
ESC
CAL-
chr9
34668311
-34668252
CCL27
(minus5593
)C
CL19
(+22992
)C
CL19
NM006274
NotesR
ed-highlighted
genesa
rewho
seexpressio
nsaresig
nificantly
changedin
Esop
hagealSquamou
sCellC
arcino
ma(ESC
C)G
reen
color-high
lighted
rowsgenesinvolvedin
lipid
metabolism
predictedwith
GOenric
hmentanalysis
Yellowcolor-high
lighted
rowsTh
eexpressionof
lncR
NAES
CCAL-5was
valid
ated
byPC
R
International Journal of Genomics 9
Final extension was at 72∘C for 10 minutes The ampliconswere resolved in 2 agarose gel
34 Bioinformatic Analysis Intensity data were exportedto GeneSpring 120 (Agilent Technologies Santa ClaraCA USA) for quantile normalization and the analysis ofdifferentially expressed long noncoding RNAs and codingRNAs Paired 119905-test analysis was used to obtain probe setswhose magnitude of change in expression of RNAs betweenESCC tissue and adjacent normal esophageal tissuewas eithergreater or less than 20 fold and 119875 value lt 005 (119875 values werecorrected formultiple testing using themethod of Benjamini-Hochberg) The normalized data containing 42544 probeswere further analyzed using the R program All controlprobes were removed We then defined the coding (ldquoNM rdquoldquoXM rdquo) and noncoding (ldquolincRNArdquo ldquoNR rdquo and ldquoXR rdquo) genesin the normalized data according to the definition of RefSeqaccession format (httpwwwncbinlmnihgovprojectsRefSeqkeyhtml) Differentially expressed long noncodingRNAs (DE-lncRNAs) and coding RNAs (DE-mRNAs)were further identified The landscapes of the wholetranscriptome (lncRNAs + coding RNAs) or all lncRNAsor all coding RNAs were analyzed with gene expressiondynamic inspector (GEDI)
35 Co-Location and Co-Expression Analysis between DE-lncRNAs and DE-mRNAs Genomic coordinates of DE-lncRNAs were imported to GREAT software (httpbejeranostanfordedugreatpublichtmlindexphp) for co-locationanalysis Neighboring coding genes were then matched withDE-mRNAs to obtain a co-expression dataset Three sub-groups of genes (DE-lncRNA co-located genes DE-mRNAsand co-expressed genes) were used for gene expressionnetwork analysis using Cytoscape software (v283)
Abbreviations
AFAP1-AS1 Actin filament-associated protein 1 antisenseRNA
Wei Cao and Wei Wu contributed equally to this project
Acknowledgments
The National Natural Science Foundation of China (no81171992) and the Zhengzhou Science and Technology
Programme (121PPTGG494-8) supported this work Theauthors thank Anne Haegert atThe Laboratory for AdvancedGenome Analysis at the Vancouver Prostate Centre Van-couver Canada for expert technical support They alsothank GenomeSky Inc Canada for providing computationalconsultation
References
[1] J Ferlay H R Shin F Bray D Forman C Mathers and DM Parkin ldquoEstimates of worldwide burden of cancer in 2008GLOBOCAN2008rdquo International Journal of Cancer vol 127 no12 pp 2893ndash2917 2010
[2] W Wu and J A Chan ldquoUnderstanding the role of longnoncoding RNAs in the cancer genomerdquo in Next GenerationSequencing in Cancer Research-Decoding Cancer Genome WWu and H Choudhry Eds pp 199ndash215 Springer New YorkNY USA 2013
[3] J J Hao T Gong Y Zhang et al ldquoCharacterization of generearrangements resulted from genomic structural aberrationsin human esophageal squamous cell carcinoma KYSE150 cellsrdquoGene vol 513 no 1 pp 196ndash201 2013
[4] M Kano N Seki N Kikkawa et al ldquoMiR-145 miR-133aand miR-133b tumor-suppressive miRNAs target FSCN1 inesophageal squamous cell carcinomardquo International Journal ofCancer vol 127 no 12 pp 2804ndash2814 2010
[5] H SuNHuHH Yang et al ldquoGlobal gene expression profilingand validation in esophageal squamous cell carcinoma and itsassociation with clinical phenotypesrdquo Clinical Cancer Researchvol 17 no 9 pp 2955ndash2966 2011
[6] D M Greenawalt C Duong G K Smyth et al ldquoGeneexpression profiling of esophageal cancer comparative analysisof Barrettrsquos esophagus adenocarcinoma and squamous cellcarcinomardquo International Journal of Cancer vol 120 no 9 pp1914ndash1921 2007
[7] I Dunham A Kundaje S F Aldred et al ldquoAn integratedencyclopedia of DNA elements in the human genomerdquo Naturevol 489 no 7414 pp 57ndash74 2012
[8] E A Gibb C J Brown and W L Lam ldquoThe functional roleof long non-coding RNA in human carcinomasrdquo MolecularCancer vol 10 article 38 2011
[9] A L Brunner A H Beck B Edris et al ldquoTranscriptionalprofiling of lncRNAs and novel transcribed regions across adiverse panel of archived human cancersrdquo Genome Biology vol13 no 8 article R75 2012
[10] M Guttman and J L Rinn ldquoModular regulatory principles oflarge non-coding RNAsrdquo Nature vol 482 no 7385 pp 339ndash346 2012
[11] K C Wang Y W Yang B Liu et al ldquoA long noncodingRNA maintains active chromatin to coordinate homeotic geneexpressionrdquo Nature vol 472 no 7341 pp 120ndash126 2011
[12] A Dean ldquoOn a chromosome far far away LCRs and geneexpressionrdquo Trends in Genetics vol 22 no 1 pp 38ndash45 2006
[13] P Ji S Diederichs W Wang et al ldquoMALAT-1 a novel noncod-ing RNA and thymosin 1205734 predict metastasis and survival inearly-stage non-small cell lung cancerrdquo Oncogene vol 22 no39 pp 8031ndash8041 2003
[14] R A Gupta N Shah K C Wang et al ldquoLong non-codingRNA HOTAIR reprograms chromatin state to promote cancermetastasisrdquo Nature vol 464 no 7291 pp 1071ndash1076 2010
10 International Journal of Genomics
[15] R Kogo T Shimamura K Mimori et al ldquoLong noncodingRNAHOTAIR regulates polycomb-dependent chromatinmod-ification and is associated with poor prognosis in colorectalcancersrdquo Cancer Research vol 71 no 20 pp 6320ndash6326 2011
[16] K Kim I Jutooru G Chadalapaka et al ldquoHOTAIR is anegative prognostic factor and exhibits pro-oncogenic activityin pancreatic cancerrdquo Oncogene vol 32 pp 1616ndash1625 2013
[17] D Li J Feng T Wu et al ldquoLong intergenic noncoding RNAHOTAIR is overexpressed and regulates PTEN methylation inlaryngeal squamous cell carcinomardquo The American Journal ofPathology vol 182 no 1 pp 64ndash70 2013
[18] A Bhan I Hussain K I Ansari S Kasiri A Bashyal and S SMandal ldquoAntisense transcript long noncoding RNA (lncRNA)HOTAIR is transcriptionally induced by estradiolrdquo Journal ofMolecular Biology vol 425 no 19 pp 3707ndash3722 2013
[19] D Cejka D Losert and V Wacheck ldquoShort interfering RNA(siRNA) tool or therapeuticrdquo Clinical Science vol 110 no 1pp 47ndash58 2006
[20] J R Prensner and A M Chinnaiyan ldquoThe emergence oflncRNAs in cancer biologyrdquo Cancer Discovery vol 1 pp 391ndash407 2011
[21] W Wu T D Bhagat X Yang et al ldquoHypomethylation ofnoncoding DNA regions and overexpression of the long non-coding RNA AFAP1-AS1 in Barrettrsquos esophagus and esophagealadenocarcinomardquoGastroenterology vol 144 no 5 pp 956ndash9662013
[22] E Ozgur U Mert M Isin M Okutan N Dalay and U GezerldquoDifferential expression of long non-coding RNAs duringgenotoxic stress-induced apoptosis in HeLa and MCF-7 cellsrdquoClinical and Experimental Medicine vol 13 no 2 pp 119ndash1262013
[23] J D Li Q C Feng and J S Li ldquoDifferential gene expressionprofiling of oesophageal squamous cell carcinoma by dnamicroarray and bioinformatics analysisrdquo Journal of Interna-tional Medical Research vol 38 no 6 pp 1904ndash1912 2010
[24] S Ma J Y Bao P S Kwan et al ldquoIdentification of PTK6via RNA sequencing analysis as a suppressor of esophagealsquamous cell carcinomardquo Gastroenterology vol 143 pp 675ndash686 2012
[25] M Tong K W Chan J Y Bao et al ldquoRab25 is a tumorsuppressor gene with antiangiogenic and anti-invasive activitiesin esophageal squamous cell carcinomardquo Cancer Research vol72 no 22 pp 6024ndash6035 2012
[26] Y J Geng S L Xie Q Li J Ma and G Y Wang ldquoLargeintervening non-coding RNA HOTAIR is associated withhepatocellular carcinoma progressionrdquo Journal of InternationalMedical Research vol 39 no 6 pp 2119ndash2128 2011
[27] X Zhou T J Lawrence Z He C R Pound J Mao andS A Bigler ldquoThe expression level of lysophosphatidylcholineacyltransferase 1 (LPCAT1) correlates to the progression ofprostate cancerrdquo Experimental andMolecular Pathology vol 92no 1 pp 105ndash110 2012
[28] Y Morita T Sakaguchi K Ikegami et al ldquoLysophosphatidyl-choline acyltransferase 1 altered phospholipid composition andregulated hepatoma progressionrdquo Journal of Hepatology vol 59no 2 pp 292ndash299 2013
Integrin-linked kinase signalingRegulation of CDC42 activity
Posttranslational regulation of adherens junction stability and disassemblyN-cadherin signaling events
E-cadherin signaling in the nascent adherens junctionStabilization and expansion of the E-cadherin adherens junction
E-cadherin signaling eventsFOXO family signaling
1176
1105904
511466
448448440
422
(b)
DE-lncRNAsco-located genes
DE-mRNAs
538 330776
ESCCAL-356
ESCCAL-337
Chr3
Chr5
Lipid metabolism
PLD1
LPCAT1
(c)
Figure 3 Identification of lncRNAs co-located and co-expressed neighboring genes in esophageal squamous cell carcinoma (ESCC)(a) Identification of neighboring genes of the DE-lncRNAs The genomic coordinate information of 410 DE-lncRNAs was used to searchneighboring genes whose genomic locations are within sim5 kb upstream and sim1 kb downstream of the lncRNA and may extend to 1000 kbin both directions using GREAT software (httpbejeranostanfordedugreatpublichtmlindexphp) The percentage of DE-lncRNAsharboring zero one or two neighboring genes is presented (b) Gene Ontology (GO) enrichment analysis of lncRNAs co-located genesIdentified gene enriched pathwaysterms are listed on the left the length of horizontal bars and the numbers on the right indicate thepercentage of genes involved in each pathwayterm (c) LncRNAs co-located and co-expressed coding mRNAs Overlap of 538 DE-lncRNAco-located genes with 3307 DE-mRNAs in microarrays identified 76 lncRNAs co-located and co-expressed coding mRNAs (list in Table 2)GO enrichment analysis suggests phospholipase D1 (PLD1) and lysophosphatidylcholine acyltransferase1 (LPCAT1) are involved in ether lipidmetabolism pathway Genomic location shows that PLD1 is located at minus22068 bp upstream of ESCCAL-337 lncRNA on Chr 3 and LPCAT1is at minus21250 bp upstream of ESCCAL-356 lncRNA on Chr 5
International Journal of Genomics 7
Table2Listof
identifi
edco-lo
catedandco-expressed
genesw
ithdifferentially
expressedlncR
NAsinES
CC
LncR
NA
nam
eG
enom
ic co
ordi
nate
Ln
cRN
Aex
pres
sion
LncR
NA
asso
ciat
edge
nes
Gen
eexp
ress
ion
Gen
esym
bol
Gen
bank
acce
ssio
nES
CCA
L-177
chr1
205404138
-205404079
Up
CDK18
(minus69575
)LE
MD
1(minus12895
)U
pLE
MD1
NM001001552
ESC
CAL-348
chr1
222587441
-222587382
Up
DU
SP10
(minus671896
)H
HIP
L2(+134032
)U
pD
USP10
NM007207
ESC
CAL-31
chr1
225237693
-225237752
Up
DN
AH14
(+120367
)LB
R(+378092
)U
pD
NA
H14
NM001373
ESC
CAL-159
chr1
225240300
-225240359
Up
DN
AH14
(+122974
)LB
R(+375485
)U
pD
NA
H14
NM001373
ESC
CAL-327
chr1
89887111
-89887052
Up
LRRC8
B(minus103315
)G
BP6
(+57646
)D
own
GBP6
NM198460
XLO
C000915
chr1
91295528
-91295469
Up
BARH
L2(minus112705
)ZN
F644
(+192313
2N
M020063
ESC
CAL-65
chr11
2017146
-2017205
Up
MRP
L23
(+48674
)IG
F2(+145165
)U
pIG
F2N
M000612
ESC
CAL-19
chr12
66204406
-66204347
Up
HM
GA
2(minus13863
)M
SRB3
(+531610
)U
pH
MG
A2
NM003484
XLO
C011548
chr15
81953024
-81952965
Up
TMC3
(minus286577
)M
EX3
B(+385366
)U
pM
EX3
BN
M032246
ESC
CAL-102
chr17
6766871
-6766930
Up
ALO
X12
(minus132483
)TE
KT1
(minus31841
)12
NM000697
ESC
CAL-342
chr17
78302158
-78302217
Up
END
OV
(minus86779
)RN
F213
(+67521
)U
pRN
F213
NM020954
ESC
CAL-342
chr17
78302158
-78302217
Up
END
OV
(minus86779
)RN
F213
(+67521
)001164638
ESC
CAL-99
chr2
102034125
-102034066
Up
CREG
2(minus30131
)RF
X8(+57069
)U
pCR
EG2
NM153836
ESC
CAL-5
chr2
216585455
-216585514
Up
FN1
(minus284694
)M
REG
(+292861
)D
own
MRE
GN
M018000
ESC
CAL-5
chr2
216585455
-216585514
Up
FN1
(minus284694
)M
REG
(+292861
)U
pFN1
NM054034
ESC
CAL-288
chr2
27789661
-27789602
Up
ZNF512
(minus16261
)G
CKR
(+69926
)D
own
GCK
RN
M001486
ESC
CAL-344
chr2
37327299
-37327358
Up
CCD
C75
(+15735
)EI
F2A
K2
(+56861
)U
pEI
F2A
K2N
M001135652
ESC
CAL-10
chr22
48086469
-48086528
Up
FAM
19A
5(minus798789
)TB
C1D22
A(+927981
)U
pFA
M19
A5
NM015381
ESC
CAL-81
chr5
141710377
-141710436
Up
SPRY4
(minus5787
)FG
F1(+355246
)U
pFG
F1N
M000800
ESC
CAL-81
chr5
141710377
-141710436
Up
SPRY4
(minus5787
)FG
F1(+355246
)U
pSP
RY4
NM030964
ESC
CAL-204
chr6
126699769
-126699828
Up
RSPO3
(minus740249
)CE
NPW
(+38546
)U
pCE
NPW
NM001012507
ESC
CAL-106
chr6
21822910
-21822969
Up
SOX4
(+228968
)PR
L(+480142
)U
pSO
X4N
M003107
ESC
CAL-239
chr8
104258684
-104258625
Up
FZD
6(minus52006
)BA
ALC
(+105734
)U
pBA
ALC
NM024812
ESC
CAL-239
chr8
104258684
-104258625
Up
FZD
6(minus52006
)BA
ALC
(+105734
)U
pFZ
D6
NM003506
ESC
CAL-300
chr1
180918852
-180918793
Dow
nST
X6(+73223
)XP
R1(+317677
)U
pXP
R1N
M004736
XLO
C000515
chr1
200384539
-200384598
Dow
nZN
F281
(minus5403
)KI
F14
(+205293
)U
pKI
F14
NM014875
XLO
C000515
chr1
200384539
-200384598
Dow
nZN
F281
(minus5403
)KI
F14
(+205293
)U
pZN
F281
NM012482
ESC
CAL-77
chr1
201592869
-201592810
Dow
nCS
RP1
(minus116873
)N
AV1
(minus24610
)U
pN
AV1
NM020443
ESC
CAL-76
chr1
90091283
-90091342
Dow
nLR
RC8
C(minus7331
)LR
RC8
B(+100916
)U
pLR
RC8
CN
M032270
ESC
CAL-209
chr10
14549236
-14549177
Dow
nFR
MD4
A(minus176341
)CD
NF
(+330776
)D
own
Dow
n
Dow
nD
own
Dow
n
CDN
FN
M001029954
ESC
CAL-284
chr10
44848467
-44848526
P1(minus562632
)CX
CL12
(+32048
)CX
CL12
NM199168
ESC
CAL-256
chr11
117671760
-117671701
Dow
nD
SCA
ML1
(minus3755
)D
SCA
ML1
NM020693
ESC
CAL-254
chr11
126219961
-126220020
Dow
nST3
GA
L4(minus5549
)D
CPS
(+46344
)ST3
GA
L4N
M006278
ESC
CAL-307
chr11
14975924
-14975983
Dow
nCY
P2R1
(minus62203
)CA
LCA
(+17878
)CY
P2R1
NM024514
ESC
CAL-105
chr11
17366661
-17366720
Dow
nB7
H6
(minus6618
)N
UCB2
(+68405
)D
own
NU
CB2
NM005013
ESC
CAL-232
chr12
131245982
-131245923
Dow
nRI
MBP2
(minus243491
)ST
X2(+77858
)D
own
RIM
BP2
NM015347
ESC
CAL-24
chr12
132824534
-132824475
Dow
nG
ALN
T9(+81400
)N
OC4
L(+195512
)U
pG
ALN
T9N
M021808
Dow
nH
NRN
PA3
)D
own
BARH
L
Dow
nA
LOX
Dow
nEN
DO
VN
MN
M
8 International Journal of Genomics
Table2Con
tinued
LncR
NA
nam
eG
enom
ic co
ordi
nate
sLn
cRN
Aex
pres
sion
LncR
NA
asso
ciat
edge
nes
Gen
eexp
ress
ion
Gen
esym
bol
Gen
bank
acce
ssio
nES
CCA
L-122
chr12
2952238
-2952297
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
n
Dow
n
Dow
n
Dow
nD
own
Dow
n
Dow
n
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Up
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
Dow
n
Dow
nD
own
Dow
nD
own
Dow
nD
own
NRI
P 2(minus8047
)FO
XM1
(+34053
)FO
XM1
NM202002
ESC
CAL-338
chr12
32101291
-32101232
BICD1
(minus158923
)H3
F3C
( minus156087
)BI
CD1
NM001714
ESC
CAL-275
chr14
71180731
-71180790
TTC9
(+72257
)M
AP3
K9(+95127
)M
AP3
K 9N
M033141
ESC
CAL-275
chr 14
71180731
-71180790
TTC9
(+72257
)M
AP 3
K 9(+95127
)TT
C9N
M015351
ESC
CAL-121
chr15
78088340
-78088281
LIN
GO1
( minus163602
)TB
C1D2
B(+281683
)LI
NG
O1
NM032808
ESC
CAL-231
chr 16
88381397
-88381338
ZNF469
( minus112511
)BA
NP
(+396330
)ZN
F469
NM001127464
ESC
CAL-74
chr2
47548478
-47548537
CALM2
(minus144768
)EP
CAM
(minus47779
)EP
CAM
NM002354
ESC
CAL-152
chr20
56839403
-56839344
PMEP
A1(minus554343
)PP
P 4R1
L(+45121
)PM
EPA1
NM020182
ESC
CAL-337
chr 3
171506465
-171506406
TNIK
(minus328239
)PL
D1
(+22068
)PL
D1
NM002662
ESC
CAL-80
chr3
177935638
- 177935579
KCN
MB2
(minus318615
)KC
NM
B 2N
M181361
ESC
CAL-90
chr 3
195441846
-195441787
MU
C20
(minus5936
)SD
HA
P 2(+56907
)M
UC20
NM001098516
ESC
CAL-59
chr 3
64855048
-64855107
AD
AM
TS9
(minus181713
)A
DA
MTS9
NM182920
ESC
CAL-72
chr 4
74922502
-74922443
CXCL3
(minus17983
)CX
CL2
(+42524
)CX
CL3
NM002090
ESC
CAL-79
chr4
79626460
- 79626401
BMP2
K(minus71101
)A
NX
A3
(+153689
)A
NX
A3
NM005139
ESC
CAL-253
chr4
8357097
-8357038
ACO
X3(+85384
)H
TRA3
(+85579
)AC
OX 3
NM003501
ESC
CAL-179
chr4
8359416
- 8359357
ACO
X3(+83065
)H
TRA3
(+87898
)H
TRA3
NM053044
ESC
CAL-11
chr4
84299039
- 84299098
HPS
E(minus43035
)H
ELQ
(+77956
)H
PSE
NM006665
ESC
CAL-41
chr4
8512901
- 8512960
GPR78
(minus69286
)M
ETTL19
(+70399
)G
PR78
NM080819
ESC
CAL-353
chr 5
1175322
-1175263
SLC12
A7
(minus63121
)SL
C6A19
(minus26417
)SL
C12
A7
NM006598
ESC
CAL-260
chr5
131808618
-131808677
IRF 1
(+17817
)SL
C22
A5
(+103247
)IR
F 1N
M002198
ESC
CAL-132
chr5
134578804
- 134578863
PITX1
(minus208870
)H2
AFY
(+156094
)PI
TX1
NM002653
ESC
CAL-4
chr5
141732627
-141732568
SPRY4
(minus27978
)FG
F1(+333055
)SP
RY4
NM030964
ESC
CAL-356
chr 5
1545355
-1545296
LPCA
T1( minus21250
)M
RPL36
(+254630
)LP
CAT 1
NM024830
XLO
C004881
chr5
72570680
-72570621
TMEM174
(+101628
)FO
XD1
(+173701
)up
FOXD1
NM004472
ESC
CAL-36
chr6
106899513
-106899572
ATG5
( minus125848
)A
IM1
(minus59762
)A
IM1
NM001624
XLO
C005849
chr6
138145112
-138145053
OLI
G3
( minus329552
)TN
FAIP
3(minus43498
)TN
FAIP3
NM006290
ESC
CAL-262
chr6
2283830
-2283771
GM
DS
(minus37933
)M
YLK4
(+467353
)G
MD
SN
M001500
ESC
CAL-120
chr6
29716817
-29716758
LOC554223
(minus42895
)H
LA-F
(+25671
)H
LA-F
NM018950
ESC
CAL-257
chr 6
29988410
-29988469
ZNRD1
( minus40596
)H
LA-J
(+14223
)H
LA-J
NR024240
ESC
CAL-73
chr6
36126663
-36126722
BRPF3
( minus37857
)M
APK13
(+28431
)M
APK13
NM002754
ESC
CAL-97
chr 6
40305634
-40305575
MO
CS1
(minus410150
)LR
FN2
(+249521
)LR
FN2
NM020737
ESC
CAL-333
chr 6
72018004
-72018063
RIM
S1(minus578616
)O
GFR
L1(+19557
)O
GFR
L1N
M024576
ESC
CAL-115
chr7
12593405
-12593464
VW
DE
(minus149583
)SC
IN(minus16768
)SC
INN
M033128
ESC
CAL-68
chr7
139487166
-139487225
TBX
AS1
(minus41756
)H
IPK2
(minus9503
)H
IPK2
AF207702
ESC
CAL-87
chr7
19958800
- 19958741
TMEM196
(minus146367
)M
ACC1
(+298242
)M
ACC1
NM182762
ESC
CAL-58
chr8
16355082
-16355141
MSR1
(minus304812
)FG
F 20
(+504562
)M
SR1
NM002445
ESC
CAL-130
chr8
37189082
-37189141
ZNF703
(minus364189
)KC
NU1
(+547270
)ZN
F703
NM025069
ESC
CAL-199
8
chr 8
38623612
-38623671
RNF 5
P 1(minus164867
)TA
CC 1
(minus21080
)TA
CC1
BC041391
ESC
CAL-
chr9
34668311
-34668252
CCL27
(minus5593
)C
CL19
(+22992
)C
CL19
NM006274
NotesR
ed-highlighted
genesa
rewho
seexpressio
nsaresig
nificantly
changedin
Esop
hagealSquamou
sCellC
arcino
ma(ESC
C)G
reen
color-high
lighted
rowsgenesinvolvedin
lipid
metabolism
predictedwith
GOenric
hmentanalysis
Yellowcolor-high
lighted
rowsTh
eexpressionof
lncR
NAES
CCAL-5was
valid
ated
byPC
R
International Journal of Genomics 9
Final extension was at 72∘C for 10 minutes The ampliconswere resolved in 2 agarose gel
34 Bioinformatic Analysis Intensity data were exportedto GeneSpring 120 (Agilent Technologies Santa ClaraCA USA) for quantile normalization and the analysis ofdifferentially expressed long noncoding RNAs and codingRNAs Paired 119905-test analysis was used to obtain probe setswhose magnitude of change in expression of RNAs betweenESCC tissue and adjacent normal esophageal tissuewas eithergreater or less than 20 fold and 119875 value lt 005 (119875 values werecorrected formultiple testing using themethod of Benjamini-Hochberg) The normalized data containing 42544 probeswere further analyzed using the R program All controlprobes were removed We then defined the coding (ldquoNM rdquoldquoXM rdquo) and noncoding (ldquolincRNArdquo ldquoNR rdquo and ldquoXR rdquo) genesin the normalized data according to the definition of RefSeqaccession format (httpwwwncbinlmnihgovprojectsRefSeqkeyhtml) Differentially expressed long noncodingRNAs (DE-lncRNAs) and coding RNAs (DE-mRNAs)were further identified The landscapes of the wholetranscriptome (lncRNAs + coding RNAs) or all lncRNAsor all coding RNAs were analyzed with gene expressiondynamic inspector (GEDI)
35 Co-Location and Co-Expression Analysis between DE-lncRNAs and DE-mRNAs Genomic coordinates of DE-lncRNAs were imported to GREAT software (httpbejeranostanfordedugreatpublichtmlindexphp) for co-locationanalysis Neighboring coding genes were then matched withDE-mRNAs to obtain a co-expression dataset Three sub-groups of genes (DE-lncRNA co-located genes DE-mRNAsand co-expressed genes) were used for gene expressionnetwork analysis using Cytoscape software (v283)
Abbreviations
AFAP1-AS1 Actin filament-associated protein 1 antisenseRNA
Wei Cao and Wei Wu contributed equally to this project
Acknowledgments
The National Natural Science Foundation of China (no81171992) and the Zhengzhou Science and Technology
Programme (121PPTGG494-8) supported this work Theauthors thank Anne Haegert atThe Laboratory for AdvancedGenome Analysis at the Vancouver Prostate Centre Van-couver Canada for expert technical support They alsothank GenomeSky Inc Canada for providing computationalconsultation
References
[1] J Ferlay H R Shin F Bray D Forman C Mathers and DM Parkin ldquoEstimates of worldwide burden of cancer in 2008GLOBOCAN2008rdquo International Journal of Cancer vol 127 no12 pp 2893ndash2917 2010
[2] W Wu and J A Chan ldquoUnderstanding the role of longnoncoding RNAs in the cancer genomerdquo in Next GenerationSequencing in Cancer Research-Decoding Cancer Genome WWu and H Choudhry Eds pp 199ndash215 Springer New YorkNY USA 2013
[3] J J Hao T Gong Y Zhang et al ldquoCharacterization of generearrangements resulted from genomic structural aberrationsin human esophageal squamous cell carcinoma KYSE150 cellsrdquoGene vol 513 no 1 pp 196ndash201 2013
[4] M Kano N Seki N Kikkawa et al ldquoMiR-145 miR-133aand miR-133b tumor-suppressive miRNAs target FSCN1 inesophageal squamous cell carcinomardquo International Journal ofCancer vol 127 no 12 pp 2804ndash2814 2010
[5] H SuNHuHH Yang et al ldquoGlobal gene expression profilingand validation in esophageal squamous cell carcinoma and itsassociation with clinical phenotypesrdquo Clinical Cancer Researchvol 17 no 9 pp 2955ndash2966 2011
[6] D M Greenawalt C Duong G K Smyth et al ldquoGeneexpression profiling of esophageal cancer comparative analysisof Barrettrsquos esophagus adenocarcinoma and squamous cellcarcinomardquo International Journal of Cancer vol 120 no 9 pp1914ndash1921 2007
[7] I Dunham A Kundaje S F Aldred et al ldquoAn integratedencyclopedia of DNA elements in the human genomerdquo Naturevol 489 no 7414 pp 57ndash74 2012
[8] E A Gibb C J Brown and W L Lam ldquoThe functional roleof long non-coding RNA in human carcinomasrdquo MolecularCancer vol 10 article 38 2011
[9] A L Brunner A H Beck B Edris et al ldquoTranscriptionalprofiling of lncRNAs and novel transcribed regions across adiverse panel of archived human cancersrdquo Genome Biology vol13 no 8 article R75 2012
[10] M Guttman and J L Rinn ldquoModular regulatory principles oflarge non-coding RNAsrdquo Nature vol 482 no 7385 pp 339ndash346 2012
[11] K C Wang Y W Yang B Liu et al ldquoA long noncodingRNA maintains active chromatin to coordinate homeotic geneexpressionrdquo Nature vol 472 no 7341 pp 120ndash126 2011
[12] A Dean ldquoOn a chromosome far far away LCRs and geneexpressionrdquo Trends in Genetics vol 22 no 1 pp 38ndash45 2006
[13] P Ji S Diederichs W Wang et al ldquoMALAT-1 a novel noncod-ing RNA and thymosin 1205734 predict metastasis and survival inearly-stage non-small cell lung cancerrdquo Oncogene vol 22 no39 pp 8031ndash8041 2003
[14] R A Gupta N Shah K C Wang et al ldquoLong non-codingRNA HOTAIR reprograms chromatin state to promote cancermetastasisrdquo Nature vol 464 no 7291 pp 1071ndash1076 2010
10 International Journal of Genomics
[15] R Kogo T Shimamura K Mimori et al ldquoLong noncodingRNAHOTAIR regulates polycomb-dependent chromatinmod-ification and is associated with poor prognosis in colorectalcancersrdquo Cancer Research vol 71 no 20 pp 6320ndash6326 2011
[16] K Kim I Jutooru G Chadalapaka et al ldquoHOTAIR is anegative prognostic factor and exhibits pro-oncogenic activityin pancreatic cancerrdquo Oncogene vol 32 pp 1616ndash1625 2013
[17] D Li J Feng T Wu et al ldquoLong intergenic noncoding RNAHOTAIR is overexpressed and regulates PTEN methylation inlaryngeal squamous cell carcinomardquo The American Journal ofPathology vol 182 no 1 pp 64ndash70 2013
[18] A Bhan I Hussain K I Ansari S Kasiri A Bashyal and S SMandal ldquoAntisense transcript long noncoding RNA (lncRNA)HOTAIR is transcriptionally induced by estradiolrdquo Journal ofMolecular Biology vol 425 no 19 pp 3707ndash3722 2013
[19] D Cejka D Losert and V Wacheck ldquoShort interfering RNA(siRNA) tool or therapeuticrdquo Clinical Science vol 110 no 1pp 47ndash58 2006
[20] J R Prensner and A M Chinnaiyan ldquoThe emergence oflncRNAs in cancer biologyrdquo Cancer Discovery vol 1 pp 391ndash407 2011
[21] W Wu T D Bhagat X Yang et al ldquoHypomethylation ofnoncoding DNA regions and overexpression of the long non-coding RNA AFAP1-AS1 in Barrettrsquos esophagus and esophagealadenocarcinomardquoGastroenterology vol 144 no 5 pp 956ndash9662013
[22] E Ozgur U Mert M Isin M Okutan N Dalay and U GezerldquoDifferential expression of long non-coding RNAs duringgenotoxic stress-induced apoptosis in HeLa and MCF-7 cellsrdquoClinical and Experimental Medicine vol 13 no 2 pp 119ndash1262013
[23] J D Li Q C Feng and J S Li ldquoDifferential gene expressionprofiling of oesophageal squamous cell carcinoma by dnamicroarray and bioinformatics analysisrdquo Journal of Interna-tional Medical Research vol 38 no 6 pp 1904ndash1912 2010
[24] S Ma J Y Bao P S Kwan et al ldquoIdentification of PTK6via RNA sequencing analysis as a suppressor of esophagealsquamous cell carcinomardquo Gastroenterology vol 143 pp 675ndash686 2012
[25] M Tong K W Chan J Y Bao et al ldquoRab25 is a tumorsuppressor gene with antiangiogenic and anti-invasive activitiesin esophageal squamous cell carcinomardquo Cancer Research vol72 no 22 pp 6024ndash6035 2012
[26] Y J Geng S L Xie Q Li J Ma and G Y Wang ldquoLargeintervening non-coding RNA HOTAIR is associated withhepatocellular carcinoma progressionrdquo Journal of InternationalMedical Research vol 39 no 6 pp 2119ndash2128 2011
[27] X Zhou T J Lawrence Z He C R Pound J Mao andS A Bigler ldquoThe expression level of lysophosphatidylcholineacyltransferase 1 (LPCAT1) correlates to the progression ofprostate cancerrdquo Experimental andMolecular Pathology vol 92no 1 pp 105ndash110 2012
[28] Y Morita T Sakaguchi K Ikegami et al ldquoLysophosphatidyl-choline acyltransferase 1 altered phospholipid composition andregulated hepatoma progressionrdquo Journal of Hepatology vol 59no 2 pp 292ndash299 2013
Final extension was at 72∘C for 10 minutes The ampliconswere resolved in 2 agarose gel
34 Bioinformatic Analysis Intensity data were exportedto GeneSpring 120 (Agilent Technologies Santa ClaraCA USA) for quantile normalization and the analysis ofdifferentially expressed long noncoding RNAs and codingRNAs Paired 119905-test analysis was used to obtain probe setswhose magnitude of change in expression of RNAs betweenESCC tissue and adjacent normal esophageal tissuewas eithergreater or less than 20 fold and 119875 value lt 005 (119875 values werecorrected formultiple testing using themethod of Benjamini-Hochberg) The normalized data containing 42544 probeswere further analyzed using the R program All controlprobes were removed We then defined the coding (ldquoNM rdquoldquoXM rdquo) and noncoding (ldquolincRNArdquo ldquoNR rdquo and ldquoXR rdquo) genesin the normalized data according to the definition of RefSeqaccession format (httpwwwncbinlmnihgovprojectsRefSeqkeyhtml) Differentially expressed long noncodingRNAs (DE-lncRNAs) and coding RNAs (DE-mRNAs)were further identified The landscapes of the wholetranscriptome (lncRNAs + coding RNAs) or all lncRNAsor all coding RNAs were analyzed with gene expressiondynamic inspector (GEDI)
35 Co-Location and Co-Expression Analysis between DE-lncRNAs and DE-mRNAs Genomic coordinates of DE-lncRNAs were imported to GREAT software (httpbejeranostanfordedugreatpublichtmlindexphp) for co-locationanalysis Neighboring coding genes were then matched withDE-mRNAs to obtain a co-expression dataset Three sub-groups of genes (DE-lncRNA co-located genes DE-mRNAsand co-expressed genes) were used for gene expressionnetwork analysis using Cytoscape software (v283)
Abbreviations
AFAP1-AS1 Actin filament-associated protein 1 antisenseRNA
Wei Cao and Wei Wu contributed equally to this project
Acknowledgments
The National Natural Science Foundation of China (no81171992) and the Zhengzhou Science and Technology
Programme (121PPTGG494-8) supported this work Theauthors thank Anne Haegert atThe Laboratory for AdvancedGenome Analysis at the Vancouver Prostate Centre Van-couver Canada for expert technical support They alsothank GenomeSky Inc Canada for providing computationalconsultation
References
[1] J Ferlay H R Shin F Bray D Forman C Mathers and DM Parkin ldquoEstimates of worldwide burden of cancer in 2008GLOBOCAN2008rdquo International Journal of Cancer vol 127 no12 pp 2893ndash2917 2010
[2] W Wu and J A Chan ldquoUnderstanding the role of longnoncoding RNAs in the cancer genomerdquo in Next GenerationSequencing in Cancer Research-Decoding Cancer Genome WWu and H Choudhry Eds pp 199ndash215 Springer New YorkNY USA 2013
[3] J J Hao T Gong Y Zhang et al ldquoCharacterization of generearrangements resulted from genomic structural aberrationsin human esophageal squamous cell carcinoma KYSE150 cellsrdquoGene vol 513 no 1 pp 196ndash201 2013
[4] M Kano N Seki N Kikkawa et al ldquoMiR-145 miR-133aand miR-133b tumor-suppressive miRNAs target FSCN1 inesophageal squamous cell carcinomardquo International Journal ofCancer vol 127 no 12 pp 2804ndash2814 2010
[5] H SuNHuHH Yang et al ldquoGlobal gene expression profilingand validation in esophageal squamous cell carcinoma and itsassociation with clinical phenotypesrdquo Clinical Cancer Researchvol 17 no 9 pp 2955ndash2966 2011
[6] D M Greenawalt C Duong G K Smyth et al ldquoGeneexpression profiling of esophageal cancer comparative analysisof Barrettrsquos esophagus adenocarcinoma and squamous cellcarcinomardquo International Journal of Cancer vol 120 no 9 pp1914ndash1921 2007
[7] I Dunham A Kundaje S F Aldred et al ldquoAn integratedencyclopedia of DNA elements in the human genomerdquo Naturevol 489 no 7414 pp 57ndash74 2012
[8] E A Gibb C J Brown and W L Lam ldquoThe functional roleof long non-coding RNA in human carcinomasrdquo MolecularCancer vol 10 article 38 2011
[9] A L Brunner A H Beck B Edris et al ldquoTranscriptionalprofiling of lncRNAs and novel transcribed regions across adiverse panel of archived human cancersrdquo Genome Biology vol13 no 8 article R75 2012
[10] M Guttman and J L Rinn ldquoModular regulatory principles oflarge non-coding RNAsrdquo Nature vol 482 no 7385 pp 339ndash346 2012
[11] K C Wang Y W Yang B Liu et al ldquoA long noncodingRNA maintains active chromatin to coordinate homeotic geneexpressionrdquo Nature vol 472 no 7341 pp 120ndash126 2011
[12] A Dean ldquoOn a chromosome far far away LCRs and geneexpressionrdquo Trends in Genetics vol 22 no 1 pp 38ndash45 2006
[13] P Ji S Diederichs W Wang et al ldquoMALAT-1 a novel noncod-ing RNA and thymosin 1205734 predict metastasis and survival inearly-stage non-small cell lung cancerrdquo Oncogene vol 22 no39 pp 8031ndash8041 2003
[14] R A Gupta N Shah K C Wang et al ldquoLong non-codingRNA HOTAIR reprograms chromatin state to promote cancermetastasisrdquo Nature vol 464 no 7291 pp 1071ndash1076 2010
10 International Journal of Genomics
[15] R Kogo T Shimamura K Mimori et al ldquoLong noncodingRNAHOTAIR regulates polycomb-dependent chromatinmod-ification and is associated with poor prognosis in colorectalcancersrdquo Cancer Research vol 71 no 20 pp 6320ndash6326 2011
[16] K Kim I Jutooru G Chadalapaka et al ldquoHOTAIR is anegative prognostic factor and exhibits pro-oncogenic activityin pancreatic cancerrdquo Oncogene vol 32 pp 1616ndash1625 2013
[17] D Li J Feng T Wu et al ldquoLong intergenic noncoding RNAHOTAIR is overexpressed and regulates PTEN methylation inlaryngeal squamous cell carcinomardquo The American Journal ofPathology vol 182 no 1 pp 64ndash70 2013
[18] A Bhan I Hussain K I Ansari S Kasiri A Bashyal and S SMandal ldquoAntisense transcript long noncoding RNA (lncRNA)HOTAIR is transcriptionally induced by estradiolrdquo Journal ofMolecular Biology vol 425 no 19 pp 3707ndash3722 2013
[19] D Cejka D Losert and V Wacheck ldquoShort interfering RNA(siRNA) tool or therapeuticrdquo Clinical Science vol 110 no 1pp 47ndash58 2006
[20] J R Prensner and A M Chinnaiyan ldquoThe emergence oflncRNAs in cancer biologyrdquo Cancer Discovery vol 1 pp 391ndash407 2011
[21] W Wu T D Bhagat X Yang et al ldquoHypomethylation ofnoncoding DNA regions and overexpression of the long non-coding RNA AFAP1-AS1 in Barrettrsquos esophagus and esophagealadenocarcinomardquoGastroenterology vol 144 no 5 pp 956ndash9662013
[22] E Ozgur U Mert M Isin M Okutan N Dalay and U GezerldquoDifferential expression of long non-coding RNAs duringgenotoxic stress-induced apoptosis in HeLa and MCF-7 cellsrdquoClinical and Experimental Medicine vol 13 no 2 pp 119ndash1262013
[23] J D Li Q C Feng and J S Li ldquoDifferential gene expressionprofiling of oesophageal squamous cell carcinoma by dnamicroarray and bioinformatics analysisrdquo Journal of Interna-tional Medical Research vol 38 no 6 pp 1904ndash1912 2010
[24] S Ma J Y Bao P S Kwan et al ldquoIdentification of PTK6via RNA sequencing analysis as a suppressor of esophagealsquamous cell carcinomardquo Gastroenterology vol 143 pp 675ndash686 2012
[25] M Tong K W Chan J Y Bao et al ldquoRab25 is a tumorsuppressor gene with antiangiogenic and anti-invasive activitiesin esophageal squamous cell carcinomardquo Cancer Research vol72 no 22 pp 6024ndash6035 2012
[26] Y J Geng S L Xie Q Li J Ma and G Y Wang ldquoLargeintervening non-coding RNA HOTAIR is associated withhepatocellular carcinoma progressionrdquo Journal of InternationalMedical Research vol 39 no 6 pp 2119ndash2128 2011
[27] X Zhou T J Lawrence Z He C R Pound J Mao andS A Bigler ldquoThe expression level of lysophosphatidylcholineacyltransferase 1 (LPCAT1) correlates to the progression ofprostate cancerrdquo Experimental andMolecular Pathology vol 92no 1 pp 105ndash110 2012
[28] Y Morita T Sakaguchi K Ikegami et al ldquoLysophosphatidyl-choline acyltransferase 1 altered phospholipid composition andregulated hepatoma progressionrdquo Journal of Hepatology vol 59no 2 pp 292ndash299 2013
Final extension was at 72∘C for 10 minutes The ampliconswere resolved in 2 agarose gel
34 Bioinformatic Analysis Intensity data were exportedto GeneSpring 120 (Agilent Technologies Santa ClaraCA USA) for quantile normalization and the analysis ofdifferentially expressed long noncoding RNAs and codingRNAs Paired 119905-test analysis was used to obtain probe setswhose magnitude of change in expression of RNAs betweenESCC tissue and adjacent normal esophageal tissuewas eithergreater or less than 20 fold and 119875 value lt 005 (119875 values werecorrected formultiple testing using themethod of Benjamini-Hochberg) The normalized data containing 42544 probeswere further analyzed using the R program All controlprobes were removed We then defined the coding (ldquoNM rdquoldquoXM rdquo) and noncoding (ldquolincRNArdquo ldquoNR rdquo and ldquoXR rdquo) genesin the normalized data according to the definition of RefSeqaccession format (httpwwwncbinlmnihgovprojectsRefSeqkeyhtml) Differentially expressed long noncodingRNAs (DE-lncRNAs) and coding RNAs (DE-mRNAs)were further identified The landscapes of the wholetranscriptome (lncRNAs + coding RNAs) or all lncRNAsor all coding RNAs were analyzed with gene expressiondynamic inspector (GEDI)
35 Co-Location and Co-Expression Analysis between DE-lncRNAs and DE-mRNAs Genomic coordinates of DE-lncRNAs were imported to GREAT software (httpbejeranostanfordedugreatpublichtmlindexphp) for co-locationanalysis Neighboring coding genes were then matched withDE-mRNAs to obtain a co-expression dataset Three sub-groups of genes (DE-lncRNA co-located genes DE-mRNAsand co-expressed genes) were used for gene expressionnetwork analysis using Cytoscape software (v283)
Abbreviations
AFAP1-AS1 Actin filament-associated protein 1 antisenseRNA
Wei Cao and Wei Wu contributed equally to this project
Acknowledgments
The National Natural Science Foundation of China (no81171992) and the Zhengzhou Science and Technology
Programme (121PPTGG494-8) supported this work Theauthors thank Anne Haegert atThe Laboratory for AdvancedGenome Analysis at the Vancouver Prostate Centre Van-couver Canada for expert technical support They alsothank GenomeSky Inc Canada for providing computationalconsultation
References
[1] J Ferlay H R Shin F Bray D Forman C Mathers and DM Parkin ldquoEstimates of worldwide burden of cancer in 2008GLOBOCAN2008rdquo International Journal of Cancer vol 127 no12 pp 2893ndash2917 2010
[2] W Wu and J A Chan ldquoUnderstanding the role of longnoncoding RNAs in the cancer genomerdquo in Next GenerationSequencing in Cancer Research-Decoding Cancer Genome WWu and H Choudhry Eds pp 199ndash215 Springer New YorkNY USA 2013
[3] J J Hao T Gong Y Zhang et al ldquoCharacterization of generearrangements resulted from genomic structural aberrationsin human esophageal squamous cell carcinoma KYSE150 cellsrdquoGene vol 513 no 1 pp 196ndash201 2013
[4] M Kano N Seki N Kikkawa et al ldquoMiR-145 miR-133aand miR-133b tumor-suppressive miRNAs target FSCN1 inesophageal squamous cell carcinomardquo International Journal ofCancer vol 127 no 12 pp 2804ndash2814 2010
[5] H SuNHuHH Yang et al ldquoGlobal gene expression profilingand validation in esophageal squamous cell carcinoma and itsassociation with clinical phenotypesrdquo Clinical Cancer Researchvol 17 no 9 pp 2955ndash2966 2011
[6] D M Greenawalt C Duong G K Smyth et al ldquoGeneexpression profiling of esophageal cancer comparative analysisof Barrettrsquos esophagus adenocarcinoma and squamous cellcarcinomardquo International Journal of Cancer vol 120 no 9 pp1914ndash1921 2007
[7] I Dunham A Kundaje S F Aldred et al ldquoAn integratedencyclopedia of DNA elements in the human genomerdquo Naturevol 489 no 7414 pp 57ndash74 2012
[8] E A Gibb C J Brown and W L Lam ldquoThe functional roleof long non-coding RNA in human carcinomasrdquo MolecularCancer vol 10 article 38 2011
[9] A L Brunner A H Beck B Edris et al ldquoTranscriptionalprofiling of lncRNAs and novel transcribed regions across adiverse panel of archived human cancersrdquo Genome Biology vol13 no 8 article R75 2012
[10] M Guttman and J L Rinn ldquoModular regulatory principles oflarge non-coding RNAsrdquo Nature vol 482 no 7385 pp 339ndash346 2012
[11] K C Wang Y W Yang B Liu et al ldquoA long noncodingRNA maintains active chromatin to coordinate homeotic geneexpressionrdquo Nature vol 472 no 7341 pp 120ndash126 2011
[12] A Dean ldquoOn a chromosome far far away LCRs and geneexpressionrdquo Trends in Genetics vol 22 no 1 pp 38ndash45 2006
[13] P Ji S Diederichs W Wang et al ldquoMALAT-1 a novel noncod-ing RNA and thymosin 1205734 predict metastasis and survival inearly-stage non-small cell lung cancerrdquo Oncogene vol 22 no39 pp 8031ndash8041 2003
[14] R A Gupta N Shah K C Wang et al ldquoLong non-codingRNA HOTAIR reprograms chromatin state to promote cancermetastasisrdquo Nature vol 464 no 7291 pp 1071ndash1076 2010
10 International Journal of Genomics
[15] R Kogo T Shimamura K Mimori et al ldquoLong noncodingRNAHOTAIR regulates polycomb-dependent chromatinmod-ification and is associated with poor prognosis in colorectalcancersrdquo Cancer Research vol 71 no 20 pp 6320ndash6326 2011
[16] K Kim I Jutooru G Chadalapaka et al ldquoHOTAIR is anegative prognostic factor and exhibits pro-oncogenic activityin pancreatic cancerrdquo Oncogene vol 32 pp 1616ndash1625 2013
[17] D Li J Feng T Wu et al ldquoLong intergenic noncoding RNAHOTAIR is overexpressed and regulates PTEN methylation inlaryngeal squamous cell carcinomardquo The American Journal ofPathology vol 182 no 1 pp 64ndash70 2013
[18] A Bhan I Hussain K I Ansari S Kasiri A Bashyal and S SMandal ldquoAntisense transcript long noncoding RNA (lncRNA)HOTAIR is transcriptionally induced by estradiolrdquo Journal ofMolecular Biology vol 425 no 19 pp 3707ndash3722 2013
[19] D Cejka D Losert and V Wacheck ldquoShort interfering RNA(siRNA) tool or therapeuticrdquo Clinical Science vol 110 no 1pp 47ndash58 2006
[20] J R Prensner and A M Chinnaiyan ldquoThe emergence oflncRNAs in cancer biologyrdquo Cancer Discovery vol 1 pp 391ndash407 2011
[21] W Wu T D Bhagat X Yang et al ldquoHypomethylation ofnoncoding DNA regions and overexpression of the long non-coding RNA AFAP1-AS1 in Barrettrsquos esophagus and esophagealadenocarcinomardquoGastroenterology vol 144 no 5 pp 956ndash9662013
[22] E Ozgur U Mert M Isin M Okutan N Dalay and U GezerldquoDifferential expression of long non-coding RNAs duringgenotoxic stress-induced apoptosis in HeLa and MCF-7 cellsrdquoClinical and Experimental Medicine vol 13 no 2 pp 119ndash1262013
[23] J D Li Q C Feng and J S Li ldquoDifferential gene expressionprofiling of oesophageal squamous cell carcinoma by dnamicroarray and bioinformatics analysisrdquo Journal of Interna-tional Medical Research vol 38 no 6 pp 1904ndash1912 2010
[24] S Ma J Y Bao P S Kwan et al ldquoIdentification of PTK6via RNA sequencing analysis as a suppressor of esophagealsquamous cell carcinomardquo Gastroenterology vol 143 pp 675ndash686 2012
[25] M Tong K W Chan J Y Bao et al ldquoRab25 is a tumorsuppressor gene with antiangiogenic and anti-invasive activitiesin esophageal squamous cell carcinomardquo Cancer Research vol72 no 22 pp 6024ndash6035 2012
[26] Y J Geng S L Xie Q Li J Ma and G Y Wang ldquoLargeintervening non-coding RNA HOTAIR is associated withhepatocellular carcinoma progressionrdquo Journal of InternationalMedical Research vol 39 no 6 pp 2119ndash2128 2011
[27] X Zhou T J Lawrence Z He C R Pound J Mao andS A Bigler ldquoThe expression level of lysophosphatidylcholineacyltransferase 1 (LPCAT1) correlates to the progression ofprostate cancerrdquo Experimental andMolecular Pathology vol 92no 1 pp 105ndash110 2012
[28] Y Morita T Sakaguchi K Ikegami et al ldquoLysophosphatidyl-choline acyltransferase 1 altered phospholipid composition andregulated hepatoma progressionrdquo Journal of Hepatology vol 59no 2 pp 292ndash299 2013
Final extension was at 72∘C for 10 minutes The ampliconswere resolved in 2 agarose gel
34 Bioinformatic Analysis Intensity data were exportedto GeneSpring 120 (Agilent Technologies Santa ClaraCA USA) for quantile normalization and the analysis ofdifferentially expressed long noncoding RNAs and codingRNAs Paired 119905-test analysis was used to obtain probe setswhose magnitude of change in expression of RNAs betweenESCC tissue and adjacent normal esophageal tissuewas eithergreater or less than 20 fold and 119875 value lt 005 (119875 values werecorrected formultiple testing using themethod of Benjamini-Hochberg) The normalized data containing 42544 probeswere further analyzed using the R program All controlprobes were removed We then defined the coding (ldquoNM rdquoldquoXM rdquo) and noncoding (ldquolincRNArdquo ldquoNR rdquo and ldquoXR rdquo) genesin the normalized data according to the definition of RefSeqaccession format (httpwwwncbinlmnihgovprojectsRefSeqkeyhtml) Differentially expressed long noncodingRNAs (DE-lncRNAs) and coding RNAs (DE-mRNAs)were further identified The landscapes of the wholetranscriptome (lncRNAs + coding RNAs) or all lncRNAsor all coding RNAs were analyzed with gene expressiondynamic inspector (GEDI)
35 Co-Location and Co-Expression Analysis between DE-lncRNAs and DE-mRNAs Genomic coordinates of DE-lncRNAs were imported to GREAT software (httpbejeranostanfordedugreatpublichtmlindexphp) for co-locationanalysis Neighboring coding genes were then matched withDE-mRNAs to obtain a co-expression dataset Three sub-groups of genes (DE-lncRNA co-located genes DE-mRNAsand co-expressed genes) were used for gene expressionnetwork analysis using Cytoscape software (v283)
Abbreviations
AFAP1-AS1 Actin filament-associated protein 1 antisenseRNA
Wei Cao and Wei Wu contributed equally to this project
Acknowledgments
The National Natural Science Foundation of China (no81171992) and the Zhengzhou Science and Technology
Programme (121PPTGG494-8) supported this work Theauthors thank Anne Haegert atThe Laboratory for AdvancedGenome Analysis at the Vancouver Prostate Centre Van-couver Canada for expert technical support They alsothank GenomeSky Inc Canada for providing computationalconsultation
References
[1] J Ferlay H R Shin F Bray D Forman C Mathers and DM Parkin ldquoEstimates of worldwide burden of cancer in 2008GLOBOCAN2008rdquo International Journal of Cancer vol 127 no12 pp 2893ndash2917 2010
[2] W Wu and J A Chan ldquoUnderstanding the role of longnoncoding RNAs in the cancer genomerdquo in Next GenerationSequencing in Cancer Research-Decoding Cancer Genome WWu and H Choudhry Eds pp 199ndash215 Springer New YorkNY USA 2013
[3] J J Hao T Gong Y Zhang et al ldquoCharacterization of generearrangements resulted from genomic structural aberrationsin human esophageal squamous cell carcinoma KYSE150 cellsrdquoGene vol 513 no 1 pp 196ndash201 2013
[4] M Kano N Seki N Kikkawa et al ldquoMiR-145 miR-133aand miR-133b tumor-suppressive miRNAs target FSCN1 inesophageal squamous cell carcinomardquo International Journal ofCancer vol 127 no 12 pp 2804ndash2814 2010
[5] H SuNHuHH Yang et al ldquoGlobal gene expression profilingand validation in esophageal squamous cell carcinoma and itsassociation with clinical phenotypesrdquo Clinical Cancer Researchvol 17 no 9 pp 2955ndash2966 2011
[6] D M Greenawalt C Duong G K Smyth et al ldquoGeneexpression profiling of esophageal cancer comparative analysisof Barrettrsquos esophagus adenocarcinoma and squamous cellcarcinomardquo International Journal of Cancer vol 120 no 9 pp1914ndash1921 2007
[7] I Dunham A Kundaje S F Aldred et al ldquoAn integratedencyclopedia of DNA elements in the human genomerdquo Naturevol 489 no 7414 pp 57ndash74 2012
[8] E A Gibb C J Brown and W L Lam ldquoThe functional roleof long non-coding RNA in human carcinomasrdquo MolecularCancer vol 10 article 38 2011
[9] A L Brunner A H Beck B Edris et al ldquoTranscriptionalprofiling of lncRNAs and novel transcribed regions across adiverse panel of archived human cancersrdquo Genome Biology vol13 no 8 article R75 2012
[10] M Guttman and J L Rinn ldquoModular regulatory principles oflarge non-coding RNAsrdquo Nature vol 482 no 7385 pp 339ndash346 2012
[11] K C Wang Y W Yang B Liu et al ldquoA long noncodingRNA maintains active chromatin to coordinate homeotic geneexpressionrdquo Nature vol 472 no 7341 pp 120ndash126 2011
[12] A Dean ldquoOn a chromosome far far away LCRs and geneexpressionrdquo Trends in Genetics vol 22 no 1 pp 38ndash45 2006
[13] P Ji S Diederichs W Wang et al ldquoMALAT-1 a novel noncod-ing RNA and thymosin 1205734 predict metastasis and survival inearly-stage non-small cell lung cancerrdquo Oncogene vol 22 no39 pp 8031ndash8041 2003
[14] R A Gupta N Shah K C Wang et al ldquoLong non-codingRNA HOTAIR reprograms chromatin state to promote cancermetastasisrdquo Nature vol 464 no 7291 pp 1071ndash1076 2010
10 International Journal of Genomics
[15] R Kogo T Shimamura K Mimori et al ldquoLong noncodingRNAHOTAIR regulates polycomb-dependent chromatinmod-ification and is associated with poor prognosis in colorectalcancersrdquo Cancer Research vol 71 no 20 pp 6320ndash6326 2011
[16] K Kim I Jutooru G Chadalapaka et al ldquoHOTAIR is anegative prognostic factor and exhibits pro-oncogenic activityin pancreatic cancerrdquo Oncogene vol 32 pp 1616ndash1625 2013
[17] D Li J Feng T Wu et al ldquoLong intergenic noncoding RNAHOTAIR is overexpressed and regulates PTEN methylation inlaryngeal squamous cell carcinomardquo The American Journal ofPathology vol 182 no 1 pp 64ndash70 2013
[18] A Bhan I Hussain K I Ansari S Kasiri A Bashyal and S SMandal ldquoAntisense transcript long noncoding RNA (lncRNA)HOTAIR is transcriptionally induced by estradiolrdquo Journal ofMolecular Biology vol 425 no 19 pp 3707ndash3722 2013
[19] D Cejka D Losert and V Wacheck ldquoShort interfering RNA(siRNA) tool or therapeuticrdquo Clinical Science vol 110 no 1pp 47ndash58 2006
[20] J R Prensner and A M Chinnaiyan ldquoThe emergence oflncRNAs in cancer biologyrdquo Cancer Discovery vol 1 pp 391ndash407 2011
[21] W Wu T D Bhagat X Yang et al ldquoHypomethylation ofnoncoding DNA regions and overexpression of the long non-coding RNA AFAP1-AS1 in Barrettrsquos esophagus and esophagealadenocarcinomardquoGastroenterology vol 144 no 5 pp 956ndash9662013
[22] E Ozgur U Mert M Isin M Okutan N Dalay and U GezerldquoDifferential expression of long non-coding RNAs duringgenotoxic stress-induced apoptosis in HeLa and MCF-7 cellsrdquoClinical and Experimental Medicine vol 13 no 2 pp 119ndash1262013
[23] J D Li Q C Feng and J S Li ldquoDifferential gene expressionprofiling of oesophageal squamous cell carcinoma by dnamicroarray and bioinformatics analysisrdquo Journal of Interna-tional Medical Research vol 38 no 6 pp 1904ndash1912 2010
[24] S Ma J Y Bao P S Kwan et al ldquoIdentification of PTK6via RNA sequencing analysis as a suppressor of esophagealsquamous cell carcinomardquo Gastroenterology vol 143 pp 675ndash686 2012
[25] M Tong K W Chan J Y Bao et al ldquoRab25 is a tumorsuppressor gene with antiangiogenic and anti-invasive activitiesin esophageal squamous cell carcinomardquo Cancer Research vol72 no 22 pp 6024ndash6035 2012
[26] Y J Geng S L Xie Q Li J Ma and G Y Wang ldquoLargeintervening non-coding RNA HOTAIR is associated withhepatocellular carcinoma progressionrdquo Journal of InternationalMedical Research vol 39 no 6 pp 2119ndash2128 2011
[27] X Zhou T J Lawrence Z He C R Pound J Mao andS A Bigler ldquoThe expression level of lysophosphatidylcholineacyltransferase 1 (LPCAT1) correlates to the progression ofprostate cancerrdquo Experimental andMolecular Pathology vol 92no 1 pp 105ndash110 2012
[28] Y Morita T Sakaguchi K Ikegami et al ldquoLysophosphatidyl-choline acyltransferase 1 altered phospholipid composition andregulated hepatoma progressionrdquo Journal of Hepatology vol 59no 2 pp 292ndash299 2013
[15] R Kogo T Shimamura K Mimori et al ldquoLong noncodingRNAHOTAIR regulates polycomb-dependent chromatinmod-ification and is associated with poor prognosis in colorectalcancersrdquo Cancer Research vol 71 no 20 pp 6320ndash6326 2011
[16] K Kim I Jutooru G Chadalapaka et al ldquoHOTAIR is anegative prognostic factor and exhibits pro-oncogenic activityin pancreatic cancerrdquo Oncogene vol 32 pp 1616ndash1625 2013
[17] D Li J Feng T Wu et al ldquoLong intergenic noncoding RNAHOTAIR is overexpressed and regulates PTEN methylation inlaryngeal squamous cell carcinomardquo The American Journal ofPathology vol 182 no 1 pp 64ndash70 2013
[18] A Bhan I Hussain K I Ansari S Kasiri A Bashyal and S SMandal ldquoAntisense transcript long noncoding RNA (lncRNA)HOTAIR is transcriptionally induced by estradiolrdquo Journal ofMolecular Biology vol 425 no 19 pp 3707ndash3722 2013
[19] D Cejka D Losert and V Wacheck ldquoShort interfering RNA(siRNA) tool or therapeuticrdquo Clinical Science vol 110 no 1pp 47ndash58 2006
[20] J R Prensner and A M Chinnaiyan ldquoThe emergence oflncRNAs in cancer biologyrdquo Cancer Discovery vol 1 pp 391ndash407 2011
[21] W Wu T D Bhagat X Yang et al ldquoHypomethylation ofnoncoding DNA regions and overexpression of the long non-coding RNA AFAP1-AS1 in Barrettrsquos esophagus and esophagealadenocarcinomardquoGastroenterology vol 144 no 5 pp 956ndash9662013
[22] E Ozgur U Mert M Isin M Okutan N Dalay and U GezerldquoDifferential expression of long non-coding RNAs duringgenotoxic stress-induced apoptosis in HeLa and MCF-7 cellsrdquoClinical and Experimental Medicine vol 13 no 2 pp 119ndash1262013
[23] J D Li Q C Feng and J S Li ldquoDifferential gene expressionprofiling of oesophageal squamous cell carcinoma by dnamicroarray and bioinformatics analysisrdquo Journal of Interna-tional Medical Research vol 38 no 6 pp 1904ndash1912 2010
[24] S Ma J Y Bao P S Kwan et al ldquoIdentification of PTK6via RNA sequencing analysis as a suppressor of esophagealsquamous cell carcinomardquo Gastroenterology vol 143 pp 675ndash686 2012
[25] M Tong K W Chan J Y Bao et al ldquoRab25 is a tumorsuppressor gene with antiangiogenic and anti-invasive activitiesin esophageal squamous cell carcinomardquo Cancer Research vol72 no 22 pp 6024ndash6035 2012
[26] Y J Geng S L Xie Q Li J Ma and G Y Wang ldquoLargeintervening non-coding RNA HOTAIR is associated withhepatocellular carcinoma progressionrdquo Journal of InternationalMedical Research vol 39 no 6 pp 2119ndash2128 2011
[27] X Zhou T J Lawrence Z He C R Pound J Mao andS A Bigler ldquoThe expression level of lysophosphatidylcholineacyltransferase 1 (LPCAT1) correlates to the progression ofprostate cancerrdquo Experimental andMolecular Pathology vol 92no 1 pp 105ndash110 2012
[28] Y Morita T Sakaguchi K Ikegami et al ldquoLysophosphatidyl-choline acyltransferase 1 altered phospholipid composition andregulated hepatoma progressionrdquo Journal of Hepatology vol 59no 2 pp 292ndash299 2013
![Page 1: Research Article Integrated Analysis of Long Noncoding RNA ... · Down PTK NM RNA-seq [ ] A P .. Down RAB NM Q-RT-PCR [ ] A P . . Down SPINK NM Microarray [ ] InternationalJournalofGenomics](https://reader033.documents.pub/reader033/viewer/2022060100/60ae752445969c4a7b2cba6a/html5/thumbnails/1.jpg)
![Page 2: Research Article Integrated Analysis of Long Noncoding RNA ... · Down PTK NM RNA-seq [ ] A P .. Down RAB NM Q-RT-PCR [ ] A P . . Down SPINK NM Microarray [ ] InternationalJournalofGenomics](https://reader033.documents.pub/reader033/viewer/2022060100/60ae752445969c4a7b2cba6a/html5/thumbnails/2.jpg)
![Page 3: Research Article Integrated Analysis of Long Noncoding RNA ... · Down PTK NM RNA-seq [ ] A P .. Down RAB NM Q-RT-PCR [ ] A P . . Down SPINK NM Microarray [ ] InternationalJournalofGenomics](https://reader033.documents.pub/reader033/viewer/2022060100/60ae752445969c4a7b2cba6a/html5/thumbnails/3.jpg)
![Page 4: Research Article Integrated Analysis of Long Noncoding RNA ... · Down PTK NM RNA-seq [ ] A P .. Down RAB NM Q-RT-PCR [ ] A P . . Down SPINK NM Microarray [ ] InternationalJournalofGenomics](https://reader033.documents.pub/reader033/viewer/2022060100/60ae752445969c4a7b2cba6a/html5/thumbnails/4.jpg)
![Page 5: Research Article Integrated Analysis of Long Noncoding RNA ... · Down PTK NM RNA-seq [ ] A P .. Down RAB NM Q-RT-PCR [ ] A P . . Down SPINK NM Microarray [ ] InternationalJournalofGenomics](https://reader033.documents.pub/reader033/viewer/2022060100/60ae752445969c4a7b2cba6a/html5/thumbnails/5.jpg)
![Page 6: Research Article Integrated Analysis of Long Noncoding RNA ... · Down PTK NM RNA-seq [ ] A P .. Down RAB NM Q-RT-PCR [ ] A P . . Down SPINK NM Microarray [ ] InternationalJournalofGenomics](https://reader033.documents.pub/reader033/viewer/2022060100/60ae752445969c4a7b2cba6a/html5/thumbnails/6.jpg)
![Page 7: Research Article Integrated Analysis of Long Noncoding RNA ... · Down PTK NM RNA-seq [ ] A P .. Down RAB NM Q-RT-PCR [ ] A P . . Down SPINK NM Microarray [ ] InternationalJournalofGenomics](https://reader033.documents.pub/reader033/viewer/2022060100/60ae752445969c4a7b2cba6a/html5/thumbnails/7.jpg)
![Page 8: Research Article Integrated Analysis of Long Noncoding RNA ... · Down PTK NM RNA-seq [ ] A P .. Down RAB NM Q-RT-PCR [ ] A P . . Down SPINK NM Microarray [ ] InternationalJournalofGenomics](https://reader033.documents.pub/reader033/viewer/2022060100/60ae752445969c4a7b2cba6a/html5/thumbnails/8.jpg)
![Page 9: Research Article Integrated Analysis of Long Noncoding RNA ... · Down PTK NM RNA-seq [ ] A P .. Down RAB NM Q-RT-PCR [ ] A P . . Down SPINK NM Microarray [ ] InternationalJournalofGenomics](https://reader033.documents.pub/reader033/viewer/2022060100/60ae752445969c4a7b2cba6a/html5/thumbnails/9.jpg)
![Page 10: Research Article Integrated Analysis of Long Noncoding RNA ... · Down PTK NM RNA-seq [ ] A P .. Down RAB NM Q-RT-PCR [ ] A P . . Down SPINK NM Microarray [ ] InternationalJournalofGenomics](https://reader033.documents.pub/reader033/viewer/2022060100/60ae752445969c4a7b2cba6a/html5/thumbnails/10.jpg)
![Page 11: Research Article Integrated Analysis of Long Noncoding RNA ... · Down PTK NM RNA-seq [ ] A P .. Down RAB NM Q-RT-PCR [ ] A P . . Down SPINK NM Microarray [ ] InternationalJournalofGenomics](https://reader033.documents.pub/reader033/viewer/2022060100/60ae752445969c4a7b2cba6a/html5/thumbnails/11.jpg)
