Cowan, Qian Zhou, Yu Han, David L. Spector, Deyou Zheng and Joseph M. Miano
Robert D. Bell, Xiaochun Long, Mingyan Lin, Jan H. Bergmann, Vivek Nanda, Sarah L.
Long Noncoding RNA
Identification and Initial Functional Characterization of a Human Vascular Cell
Print ISSN: 1079-5642. Online ISSN: 1524-4636
Copyright © 2014 American Heart Association, Inc. All rights reserved.
Greenville Avenue, Dallas, TX 75231
is published by the American Heart Association, 7272
Arteriosclerosis, Thrombosis, and Vascular Biology
2014;34:1249-1259; originally published online February 27,
Arterioscler Thromb Vasc Biol.
World Wide Web at:
The online version of this article, along with updated information and services, is located on the
Data Supplement (unedited) at:
Arteriosclerosis, Thrombosis, and Vascular Biology
Information about subscribing to
Information about reprints can be found online at:
document. Question and Answer
Permissions and Rightspage under Services. Further information about this process is available in the
which permission is being requested is located, click Request Permissions in the middle column of the Web
Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for
can be obtained via RightsLink, a service of the
Arteriosclerosis, Thrombosis, and Vascular Biology
Requests for permissions to reproduce figures, tables, or portions of articles originally published
by guest on May 14, 2014 by guest on May 14, 2014 by guest on May 14, 2014 by guest on May 14, 2014 by guest on May 14, 2014 by guest on May 14, 2014 by guest on May 14, 2014 by guest on May 14, 2014 by guest on May 14, 2014 by guest on May 14, 2014 by guest on May 14, 2014 by guest on May 14, 2014 by guest on May 14, 2014http://atvb.ahajournals.org/http://atvb.ahajournals.org/http://atvb.ahajournals.org/http://atvb.ahajournals.org/http://atvb.ahajournals.org/http://atvb.ahajournals.org/http://atvb.ahajournals.org/http://atvb.ahajournals.org/http://atvb.ahajournals.org/http://atvb.ahajournals.org/ http://atvb.ahajournals.org/http://atvb.ahajournals.org/http://atvb.ahajournals.org/Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from
similar number of protein-coding genes, a fact seemingly in
support of an abundance of junk DNA within our genome.1
Two major discoveries during the past 10 years challenge this
decades-old concept. First, genome-wide RNA expression stud-
ies show widespread transcription across the mouse and human
genomes with roughly equal amounts of polyadenylated and
nonpolyadenylated RNA.2–7 Second, the combined efforts of
the ENCyclopedia Of DNA Elements (ENCODE) Consortium
and many other laboratories have revealed the existence of mil-
lions of codes that punctuate the human genome, most notably
codes for transcription factor binding.8–12 These findings, cou-
pled with the notion that much of the human genome is func-
tional with 50% to 90% comprising transcribed sequences,13,14
debunk the concept of junk DNA and point to a genome replete
with information essential for human life.
lthough the human genome is 30× larger than that of
Caenorhabditis elegans, each species is endowed with a
See accompanying editorial on page 1124
Much of the noncoding RNA (ncRNA) in a cell func-
tions to orchestrate basic translation (transfer and ribosomal
RNA); however, 2 broad classes of ncRNA expanded greatly
at the turn of the millennium, primarily as a result of large-
scale transcriptomics projects.2,3,15 These ncRNAs are classi-
fied subjectively as either short (processed transcript length
<200 nucleotides) or long (processed transcript length >200
nucleotides). Short ncRNAs include small nucleolar RNA and
their derivatives that act as guide RNAs to modify ribosomal
and transfer RNAs,16 as well as microRNA, small interfering
RNA, and PIWI-interacting RNA that use Argonaute proteins
to mediate endonucleolytic cleavage of target RNAs.17
Long ncRNAs (lncRNAs) function in a myriad of biological
processes and may be classified loosely based on their physical
location in the genome. Long intervening ncRNAs (lincRNAs)
© 2014 American Heart Association, Inc.
Arterioscler Thromb Vasc Biol is available at http://atvb.ahajournals.org DOI: 10.1161/ATVBAHA.114.303240
Objective—Long noncoding RNAs (lncRNAs) represent a rapidly growing class of RNA genes with functions related
primarily to transcriptional and post-transcriptional control of gene expression. There is a paucity of information about
lncRNA expression and function in human vascular cells. Thus, we set out to identify novel lncRNA genes in human
vascular smooth muscle cells and to gain insight into their role in the control of smooth muscle cell phenotypes.
Approach and Results—RNA sequencing (RNA-seq) of human coronary artery smooth muscle cells revealed 31
unannotated lncRNAs, including a vascular cell–enriched lncRNA (Smooth muscle and Endothelial cell–enriched
migration/differentiation-associated long NonCoding RNA [SENCR]). Strand-specific reverse transcription polymerase
chain reaction (PCR) and rapid amplification of cDNA ends indicate that SENCR is transcribed antisense from the 5′ end
of the FLI1 gene and exists as 2 splice variants. RNA fluorescence in situ hybridization and biochemical fractionation
studies demonstrate SENCR is a cytoplasmic lncRNA. Consistent with this observation, knockdown studies reveal little
to no cis-acting effect of SENCR on FLI1 or neighboring gene expression. RNA-seq experiments in smooth muscle cells
after SENCR knockdown disclose decreased expression of Myocardin and numerous smooth muscle contractile genes,
whereas several promigratory genes are increased. Reverse transcription PCR and Western blotting experiments validate
several differentially expressed genes after SENCR knockdown. Loss-of-function studies in scratch wound and Boyden
chamber assays support SENCR as an inhibitor of smooth muscle cell migration.
Conclusions—SENCR is a new vascular cell–enriched, cytoplasmic lncRNA that seems to stabilize the smooth muscle cell
contractile phenotype. (Arterioscler Thromb Vasc Biol. 2014;34:1249-1259.)
Key Words: cell migration ◼ endothelial cells ◼ myocytes, smooth muscle ◼ RNA, long noncoding
◼ RNA sequence
Received on: January 11, 2014; final version accepted on: February 12, 2014.
From the Aab Cardiovascular Research Institute, University of Rochester School of Medicine and Dentistry, Rochester, NY (R.D.B., X.L., V.N., S.L.C.,
Q.Z., Y.H., J.M.M.); Department of Genetics (M.L., D.Z.) and Departments of Neurology and Neuroscience (D.Z.), Albert Einstein College of Medicine,
Bronx, NY; and Cold Spring Harbor Laboratory, Cold Spring Harbor, NY (J.H.B., D.L.S.).
*These authors share equal authorship.
The online-only Data Supplement is available with this article at http://atvb.ahajournals.org/lookup/suppl/doi:10.1161/ATVBAHA.114.303240/-/DC1.
Correspondence to Joseph M. Miano, PhD, Aab Cardiovascular Research Institute, University of Rochester School of Medicine and Dentistry, 601
Elmwood Ave, Rochester, NY 14642. E-mail firstname.lastname@example.org; or Deyou Zheng, PhD, Department of Neurology, Albert Einstein College of
Medicine, 1300 Morris Park Ave, Bronx, NY 10461. E-mail email@example.com
Identification and Initial Functional Characterization of a
Human Vascular Cell–Enriched Long Noncoding RNA
Robert D. Bell,* Xiaochun Long,* Mingyan Lin,* Jan H. Bergmann, Vivek Nanda,
Sarah L. Cowan, Qian Zhou, Yu Han, David L. Spector, Deyou Zheng, Joseph M. Miano
by guest on May 14, 2014http://atvb.ahajournals.org/Downloaded from
1250 Arterioscler Thromb Vasc Biol June 2014
are a subclass of lncRNAs found between 2 transcription units,
and they exhibit similar active chromatin signatures as those
found around protein-coding genes.18–20 LincRNAs may display
tissue-specific patterns of expression and function principally as
scaffold or guide RNAs that facilitate chromatin remodeling in
cis or trans to directly influence gene transcription (nuclear lin-
cRNAs) or effect changes in mRNA stability/protein translation
(cytoplasmic lincRNAs).20–22 Examples of lincRNAs include
the abundantly expressed MALAT1 that functions in process-
ing of mRNAs23 and the epidermal prodifferentiating TINCR.24
A recent report defined very long intervening ncRNAs whose
expression correlates with malignancy; these transcripts may
encompass previously annotated lincRNAs.25 LincRNAs may
also overlap transcriptional enhancers to effect cis-mediated
changes in gene expression.26,27
Intragenic lncRNAs represent another subclass of RNA
genes that reside on the sense or antisense strand of an over-
lapping gene. Sense lncRNAs have been reported only spo-
radically,28 although a recent report contends there exists a
large number of ill-defined sense ncRNAs within introns.29
Antisense lncRNAs occur in a significant number of
protein-coding genes and may overlap the 5′ or 3′ end of a gene,
occur entirely within an intron, or overlap multiple exons.30–32
Antisense lncRNAs whose exons overlap protein-coding (or
ncRNA) exons are known as natural antisense transcripts and
these can function in cis or trans to negatively or positively
regulate gene expression through RNA interactions with chro-
matin remodeling factors.33 Examples of natural antisense
transcripts include the X chromosome inactivating XIST34 and
the cell cycle regulator ANRIL.35 Some processed antisense
lncRNAs do not overlap sense exons and thus may have unex-
pected functions (below). The number of human lncRNAs
is soaring with the current catalogue of LNCipedia36 listing
>32 000 (http://www.lncipedia.org/), a number that exceeds
all protein-coding genes. Thus, lncRNAs embody a rapidly
growing class of genes with functions related primarily to the
regulation of gene/protein expression.
Cellular differentiation requires the coordinated activation
of unique gene sets through transcription factors in associa-
tion with cofactors over discrete cis elements. For example,
vascular smooth muscle cell (SMC) differentiation is chiefly
a function of ubiquitously expressed serum response fac-
tor37 binding a cardiovascular-restricted cofactor called
myocardin (MYOCD)38 over CArG elements located in the
proximal promoter region of many SMC-associated genes.39
Similarly, endothelial cell (EC) differentiation proceeds, in
part, through the FOXC240 and ETV241 transcription factors
binding a composite cis element, the FOX-ETS motif, found
in promoter/enhancer sequences of several EC-specific
genes.42 Normal differentiated properties of SMC and EC
further require fine tuning of gene expression through the
action of microRNAs.43 Because lncRNAs are prevalent
and play key roles in modulating gene expression,44 they
too may have functions linked to vascular cell phenotype.
Little is known, however, about the expression or function
of lncRNAs in vascular cells,45–49 and there is nothing known
about human-specific, vascular cell–selective lncRNAs.
Accordingly, we performed RNA-seq in human coronary
artery smooth muscle cell (HCASMC) as a first step toward
understanding the potential role of lncRNAs in human SMC
phenotypic control. Here, we report on the identification of
31 lncRNAs, including 1 named SENCR (for Smooth muscle
and Endothelial cell–enriched migration/differentiation-
associated long NonCoding RNA). We have characterized
the expression, splicing, and localization of SENCR and
have identified unique gene signatures on its knockdown in
SMC. SENCR seems to play a role in maintaining the normal
SMC differentiated state as its attenuated expression leads to
reduced MYOCD and contractile gene expression with ele-
vations in migratory genes that foster a hyper-motile state.
This report outlines the first foray into lncRNA discovery in
human vascular cells and establishes a foundation for further
inquiry into SENCR biology, as well as the identification,
expression, and function of other human vascular-selective
lncRNAs under normal and pathological cell states.
Materials and Methods
Materials and Methods are available in the online-only Supplement.
Identification and Validation of
lncRNAs in HCASMC
We have developed a rigorous workflow for the identifica-
tion and study of lncRNAs in primary-derived HCASMC
using RNA-seq methodology (Figure I in the online-only
Data Supplement). A total of 79.41% of filtered reads could
be aligned to the human reference genome. Thirty-one
lncRNAs met our strict inclusion criteria (Methods in the
online-only Data Supplement) with the majority (22/31)
falling into the lincRNA subclass (Table II in the online-
only Data supplement). Conventional reverse transcription
polymerase chain reaction (RT-PCR) showed detectable
expression of 21 of 31 lncRNAs in a panel of human cell
types, including HCASMC and human umbilical vein EC
(HUVEC; Figure 1A). Sequence analysis of the PCR prod-
ucts confirmed the identity of each lncRNA (not shown).
The majority of HCASMC lncRNAs are distributed widely
across human tissues with several detected in dated human
plasma (Figure 1B and 1C). One of the lncRNAs (lncRNA9)
exhibited a selective pattern of expression in cell lines
(Figure 1A and 1D) and human tissues (Figure 1A and 1B).
Nonstandard Abbreviations and Acronyms
ENCyclopedia Of DNA Elements
Friend leukemia virus integration 1
human coronary artery smooth muscle cell(s)
human umbilical vein endothelial cell(s)
long intervening noncoding RNA
long noncoding RNA
reverse transcription polymerase chain reaction
smooth muscle and endothelial cell enriched migration/differ-
entiation-associated long noncoding RNA
smooth muscle cell
by guest on May 14, 2014http://atvb.ahajournals.org/Downloaded from
Bell et al Novel lncRNA Discovery in Vascular Smooth Muscle Cells 1251
We refer to this lncRNA as SENCR because of its enriched
expression in both smooth muscle and ECs (Figure 1A and
1D) and its proposed function (below).
SENCR Is a Vascular Cell–Selective
RNA-seq alignment, 5′ RACE, and RT-PCR with oligo-dT
and strand-specific primers established that SENCR com-
prises 3 exons and is transcribed in the antisense orienta-
tion from within the first intron of Friend leukemia virus
integration 1 (FLI1), an important transcription factor pro-
gramming EC and blood cell formation50 (Figure 2A). There
is no overlap between SENCR and FLI1 exonic sequences,
indicating that SENCR is not a natural antisense transcript33
(Figure 2A). The longest open reading frame flanked by start
and stop codons is 61 amino acids; however, analysis of this
and other predicted open reading frames in SENCR failed to
reveal any known protein-coding domains, suggesting that
this transcript has no or low protein-coding potential (not
shown). Primers to exons 1 and 3 of SENCR showed the
presence of 2 distinct PCR products (Figure 2B). Sequence
analysis confirmed these products as full length (SENCR_
V1) and an alternatively spliced variant (SENCR_V2) of the
SENCR gene (Figure 2A and 2B). These sequences have been
deposited in GenBank under accession numbers KF806591
and KF806590, respectively. We used specific primer pairs
to examine SENCR isoform expression in a panel of human
tissues and cell lines. Results showed SENCR_V1 to be
more broadly expressed than SENCR_V2 (Figure 2C and
2D). In general, there was coincident expression of SENCR
with FLI1, suggesting that these transcripts may be under
similar transcriptional control processes (Figure 2E and 2F).
Quantitative RT-PCR analysis suggested the FLI1 transcript
to have higher expression than SENCR (Figure II in the
online-only Data Supplement).
Exon 1 of FLI1 shows high conservation across 46 mam-
malian species; however, much less conservation exists across
the 3 exons of SENCR (Figure 3), consistent with the fact that
no orthologous SENCR transcripts have yet been found out-
side human/chimp lineages. Interestingly, exons 2 and 3 of
SENCR harbor single nucleotide polymorphisms, suggesting
potential deleterious effects on SENCR function (Figure 3).
Analysis of ENCODE data on the UCSC Genome Browser
(http://genome.ucsc.edu/) supports the enriched expression
of SENCR in HUVEC with lower levels in other cell types.
Further, there is a prominent HUVEC-associated H3K4me3
mark near exon 1 of SENCR, suggesting the presence of
an active promoter (Figure 3). As a first step toward delin-
eating SENCR transcription, we cloned and tested several
luciferase reporter constructs. Luciferase assays showed
little to no detectable SENCR promoter activity in HUVEC
unless sequences encompassing the 5′ FLI1 promoter region
were included, although even these reporters showed much
lower activity than a control promoter construct (not shown).
Dated human plasma
Relative lncRNA9 RNA
Figure 1. Validation of Long noncoding RNA (lncRNA) expression in human cells and tissues. Reverse transcription polymerase chain
reaction (RT-PCR) analysis of 21 lncRNAs (arbitrarily numbered) in indicated human cells (A) and tissues (B). Bold lncRNA9 and asterisk
denote SENCR. C, RT-PCR of indicated lncRNAs in dated human plasma. All reactions were done using the same PCR parameters.
D, Quantitative RT-PCR of lncRNA9 in the indicated human cell types. HCASMC indicates human coronary artery smooth muscle cell;
HF, human fibroblasts; HUVEC, human umbilical vein endothelial cell; and SKLMS, uterine leiomyosarcoma cell line.
by guest on May 14, 2014http://atvb.ahajournals.org/Downloaded from
1252 Arterioscler Thromb Vasc Biol June 2014
Collectively, these results define an alternatively spliced, vas-
cular cell–enriched antisense lncRNA that overlaps the 5′ end
of the FLI1 transcription factor yet, in its mature form, does
not harbor exonic sequences that could undergo Watson–Crick
base-pairing with corresponding exonic sequences in FLI1.
SENCR Is a Cytoplasmic lncRNA
Quantitative RT-PCR showed SENCR RNA to be most abun-
dant in HUVEC with undetectable transcripts in HeLa cells
(Figure 1D). We used high-resolution RNA fluorescence in
situ hybridization51 in these 2 cells types to unambiguously dis-
cern the intracellular compartment where SENCR transcripts
reside. Consistent with quantitative RT-PCR, no SENCR tran-
scripts were seen in individual HeLa cells (Figure 4A, bottom).
However, we observed variably low numbers of SENCR RNA
molecules in the cytoplasm of individual HUVEC (Figure 4A,
top and middle). We sometimes observed SENCR RNA in the
nucleus although this probably reflects either active transcription
or unprocessed RNA. The cytoplasmic, low-level expression of
SENCR RNA contrasts with the higher-level nuclear accumula-
tion of NEAT1 lncRNA as well as cytoplasmic PP1B mRNA
(Figure 4A). Biochemical fractionation followed by RT-PCR
further documented cytoplasmic localization of SENCR in both
HUVEC and HCASMC. In contrast, the lncRNAs NEAT1 and
XIST show predominantly nuclear accumulation in these cell
types (Figure 4B; Figure III in the online-only Data Supplement).
We next used 2 distinct probe sets to SENCR in HUVEC treated
with a control dicer substrate RNA or 2 dicer substrate RNAs
targeting different regions of SENCR to further demonstrate the
specificity of the signal (Figure 4C). Quantitative analysis of
coincident hybridization of each probe set demonstrated a likely
underestimate of ≈0.8 copies of SENCR per cell, a value that
was approximately halved on SENCR knockdown (Figure 4D).
These results establish the cytoplasmic localization of SENCR
and indicate its relatively weaker level of expression as com-
pared with housekeeping mRNA molecules (PP1B) and at least
one other lncRNA (NEAT1).
SENCR Knockdown Exerts Little Effect
on FLI1 mRNA in Vascular Cells
Many lncRNAs that overlap protein-coding genes in the anti-
sense orientation exert cis or trans effects on gene expression
through the recruitment of chromatin remodeling factors.52
However, no uniform cis-acting effect on FLI1 or neighbor-
ing gene expression was observed on knocking down SENCR
with multiple dicer substrate RNAs in HCASMC (Figure 5A–
5C) or HUVEC (Figure 5D and 5E), consistent with its cyto-
plasmic localization. There was also little effect of SENCR
knockdown on the nuclear accumulation of FLI1 protein or
steady-state FLI1 protein levels (Figure 6G and 6H). Further,
knockdown of FLI1 effected no significant change in levels
of SENCR RNA (Figure 5F). We occasionally observed mild
variation in FLI1 mRNA expression (either up or down) with
some dicer substrate RNAs in certain isolates of vascular
cells; however, these changes were sporadic and not repro-
ducible when tested by multiple investigators. We therefore
conclude that reducing SENCR RNA has little to no cis-acting
effect on local gene expression.
- 1 kb
- 0.5 kb
Figure 2. SENCR gene structure and iso-
form expression. A, Schematic of SENCR
and FLI1 (partial) gene loci. Arrows denote
the transcription start sites and bent lines
in SENCR indicate splicing patterns. B,
Reverse transcription polymerase chain
reaction (RT-PCR) of SENCR with primers
to exons 1 and 3 showing the presence of
2 transcripts reflecting full length (V1) and
alternately spliced (V2) SENCR. RT-PCR
of 2 SENCR isoforms and FLI1 in vari-
ous human tissues (C) and cell lines (D).
Quantitative RT-PCR of SENCR and FLI1 in
select human tissues (E) and cell lines (F).
Bars here and below represent the SD of 1
experiment with 3 biological replicates. All
expression data here and below represent
≥2 (more typically multiple) independent
studies performed by >1 author. Unless
indicated otherwise, SENCR expression
here and below reflects both isoforms using
primers to a common exon. HCASMC
indicates human coronary artery smooth
muscle cell; HF, human fibroblasts; HUVEC,
human umbilical vein endothelial cell; and
SKLMS, uterine leiomyosarcoma cell line.
by guest on May 14, 2014 http://atvb.ahajournals.org/Downloaded from
Bell et al Novel lncRNA Discovery in Vascular Smooth Muscle Cells 1253
SENCR Knockdown Alters the Normal
Contractile Gene Program in HCASMC
Several cytoplasmic lncRNAs effect changes in a cell’s tran-
scriptome through post-transcriptional control processes.53 As
an initial step toward understanding the function of SENCR,
we performed RNA-seq in HCASMC after knockdown of
SENCR to assess changes in the transcriptome. Most sequenc-
ing reads were aligned to the reference genome and scat-
terplots of replicates showed similar transcript profiles (not
shown). Statistical analysis of each set of replicates revealed
hundreds of genes that were significantly induced or repressed
on SENCR knockdown (Figure 6A; Table III in the online-
only Data supplement). Strikingly, many SMC contractile
genes showed significant reduction in mRNA expression with
SENCR knockdown (Figure 6B; Table III in the online-only
Data supplement). Gene ontology analysis using DAVID
revealed biological processes associated with this reduced
contractile gene signature (Table IV in the online-only Data
supplement). Of note, the key transcriptional switch for SMC
contractile gene expression, MYOCD,39 was also reduced with
SENCR knockdown (Figure 6B), and several dicer substrate
RNAs to SENCR validated such downregulation in HCASMC
(Figure 6D). We also confirmed reduced expression of sev-
eral of the SMC contractile genes at both the mRNA level
(Figure 6E) and the protein level (Figure 6G). Although
the SMC contractile program was reduced with SENCR
knockdown, several genes associated with cell migration
were induced (Figure 6C; Table III in the online-only Data
Supplement). DAVID analysis supported biological processes
linked to cellular locomotion with SENCR knockdown (Table
V in the online-only Data Supplement). We validated 2 migra-
tory genes (MDK and PTN) at the mRNA level in HCASMC
(Figure 6F) and HUVEC (Figure IV in the online-only Data
Supplement). Collectively, these data show that reduced
SENCR expression compromised the SMC contractile pheno-
type and promoted a promigratory gene signature.
Attenuated SENCR Expression Confers a
Hyper-Motile Phenotype in HCASMC
To ascertain whether the increase in promigratory gene
expression on knockdown of SENCR translates into a func-
tional phenotype, we performed 2 independent measures of
cell migration. Using a scratch wound assay, we observed
hyper-motile HCASMC with SENCR knockdown (Figure 7A
and 7B). Many of these cells exhibited reorganization of the
actin cytoskeleton with formation of lamellipodia, consis-
tent with a migratory cell phenotype (Figure 7Af, arrows).
Importantly, the increase in HCASMC migration could be
completely rescued on simultaneous knockdown of either of
2 promigratory genes shown to be induced on knockdown
Figure 3. UCSC genome browser track of the human FLI1-SENCR sense–antisense gene pair. The SENCR gene comprises 3 exons
(shown as dark rectangles) and 2 introns. The first exon of SENCR initiates on the opposite strand ≈1.5 kb downstream from the first
exon of FLI1. The dotted vertical lines serve to highlight several features, including (from the top) mammalian conservation (Mammal
Cons), reference single nucleotide polymorphisms (rs numbers), H3K4Me3, and RNA-seq (Transcription) in Tier 1 and Tier 2 cells from
ENCODE. Note the lower conservation and selective transcription in human umbilical vein endothelial cell (HUVEC; blue peaks at bottom)
for SENCR as compared to FLI1. See Results for more details.
by guest on May 14, 2014http://atvb.ahajournals.org/Downloaded from
1254 Arterioscler Thromb Vasc Biol June 2014
of SENCR (Figure 7C; Figure V in the online-only Data
Supplement). To further confirm this accentuated cell migra-
tion phenotype on knockdown of SENCR, we used a modified
Boyden chamber assay. Consistent with the scratch wound
assay, we noted that HCASMC migration was elevated with
SENCR knockdown although not as much as that observed
with the potent migratory stimulus, PDGF-BB (Figure 7D
and 7E). We also observed augmented PDGF-BB–induced
Relative RNA level
Relative RNA Level
Figure 5. Effect of knocking down SENCR
on local gene expression. A, Dicer substrate
control (ds-Ctrl) or dicer substrate SENCR
RNA was transfected into human coronary
artery smooth muscle cell (HCASMC) for
72 hours and then total RNA isolated for
conventional (top) or quantitative (bottom)
reverse transcription polymerase chain reac-
tion (RT-PCR). Dicer substrate RNA to vari-
ous regions of SENCR are abbreviated here
and below as ds followed by a number (see
Table I in the online-only Data Supplement
for details). Quantitative RT-PCR of FLI1
mRNA (B) or flanking genes around FLI1 (C)
after 3 days transfection with indicated dicer
substrate RNAs. D, Conventional RT-PCR
of SENCR and FLI1 in human umbilical vein
endothelial cell (HUVEC) after transfec-
tion with indicated dsRNAs. E, Quantitative
RT-PCR of SENCR and FLI1 after transfec-
tion with indicated dsRNAs in HUVEC. F,
Quantitative RT-PCR of FLI1 and SENCR
after knockdown of FLI1 mRNA in HCASMC.
Data are representative of multiple indepen-
dent experiments performed by indepen-
dent investigators using several isolates of
HCASMC or HUVEC.
C N CN
Average # costained foci/cell
Figure 4. Localization of SENCR RNA. A, RNA fluorescence in situ hybridization (RNA FISH) analysis of SENCR vs cytoplasmic PP1B mRNA
and nuclear NEAT1 RNA in indicated cell types. Arrows point to single molecules of SENCR RNA in the cytoplasm of human umbilical vein
endothelial cell (HUVEC). Scale bars are all 5 μm. The broken white rectangle at upper right is shown at higher magnification in Figure III in the
online-only Data Supplement. B, Reverse transcription polymerase chain reaction analysis of 2 SENCR isoforms vs other long noncoding RNA
(lncRNA) genes from polyA+RNA isolated from the cytoplasmic (C) or nuclear (N) fractions of indicated cell types. C, Application of 2 probe sets
to SENCR RNA (see Methods in the online-only Data Supplement). Arrows point to coincident localization of 2 fluorescently tagged probe sets
(yellow) targeting different regions of the SENCR transcript. D, Quantitative measures of coincident localization of SENCR probes in HUVEC
transfected with a dicer substrate control RNA (ds-Ctrl) or 2 dicer substrate RNAs targeting different regions of SENCR (ds-3 and ds-4). The y
axis indicates the average number of costained foci/cell. DAPI indicates 4′,6-diamidino-2-phenylindole.
by guest on May 14, 2014http://atvb.ahajournals.org/Downloaded from
Bell et al Novel lncRNA Discovery in Vascular Smooth Muscle Cells 1255
cell migration on concomitant knockdown of SENCR (Figure
VI in the online-only Data Supplement). Taken together,
these results strongly support a role for SENCR in the regu-
lation of HCASMC differentiation and cellular motility.
Contrary to the historical notion of pervasive junk DNA,1
most of the human genome is transcribed signifying a
treasure-trove of previously unrecognized functional DNA
sequences. These include tens of millions of regulatory
elements as well as the expansive class of lncRNA genes.
LncRNA genes already outnumber protein-coding genes
and they exhibit diverse functions related to gene expression
and splicing; protein translation, activity, and trafficking;
as well as the formation of specialized microenvironmen-
tal niches.54,55 Here, we present the first RNA-seq study
in a human vascular cell type for the specific discovery of
lncRNA genes. We used strict criteria and discovered 31
previously unannotated lncRNAs, 21 of which we vali-
dated in human cell lines and human tissues. In addition, we
detected a few lncRNAs in dated human plasma, suggest-
ing that these may have potential utility as biomarkers of
clinical disease.56 One of the lncRNA genes, named here as
SENCR, shows a selective pattern of expression in cells and
tissues with highest levels in human vascular SMC and ECs.
We discovered that SENCR undergoes alternative splicing,
consistent with widespread splicing of transcripts across the
human genome.57 SENCR overlaps the 5′ end of the FLI1
transcription factor in the antisense orientation, but does not
seem to regulate local gene expression in cis. Indeed, our
extensive RNA fluorescence in situ hybridization and bio-
chemical fractionation studies clearly indicate SENCR to be
a cytoplasmic lncRNA supporting an extranuclear function.
Using RNA-seq after knockdown of SENCR, we observed
uniform decreases in expression of SMC contractile–asso-
ciated genes as well as attenuated expression of the major
transcriptional switch (Myocardin) for the differentiation of
vascular SMC.39 However, knockdown of SENCR augments
a promigratory gene signature that facilitates heightened
SMC migration. Thus, we have uncovered a new vascular
cell–enriched lncRNA that seems to function in the mainte-
nance of a normal, nonmotile SMC phenotype.
An analysis of 707 sense–antisense gene pairs annotated
in the UCSC genome browser58 shows diversity in struc-
tural orientation, with most lncRNAs representing natural
antisense transcripts (47.0%), followed by intronic (18.8%),
Relative MYOCD mRNA
-5.0-2.50 2.5 5.0
Figure 6. Effect of SENCR knockdown on human coronary artery smooth muscle cell (HCASMC) transcriptome. A, Volcano plot depict-
ing changes in gene expression with SENCR knockdown. The red dashed line indicates genes (in red) whose changes in expression were
statistically significant. Sample smooth muscle cell (SMC) contractile genes (B) and promigratory genes (C) exhibiting either reduced (B)
or increased (C) expression with dicer substrate (ds)-SENCR knockdown. See Table III in the online-only Data Supplement for a complete
listing of all genes showing significant up- or downregulation with SENCR knockdown. Quantitative reverse transcription polymerase
chain reaction (RT-PCR) validation of reduced MYOCD mRNA (D) and SMC contractile genes (E) in HCASMC after knockdown of SENCR
with various dsRNAs. F, Quantitative RT-PCR validation of upregulation of 2 promigratory genes on knockdown of SENCR in HCASMC.
G, Western blot validation of upregulated (ANPEP) and downregulated (SMC contractile) proteins in HCASMC 72 hours after indicated
transfection with dsRNA. Similar findings were observed in an independent experiment. H, Immunofluorescence microscopy of FLI1
protein in the nucleus of HCASMC after 3 days transfection with indicated dsRNAs. Results are representative of multiple experiments
performed by independent investigators. Scale bar, 10 μm for both images.
by guest on May 14, 2014http://atvb.ahajournals.org/ Downloaded from
1256 Arterioscler Thromb Vasc Biol June 2014
divergent (16.4%), completely overlapping (7.4%), 5′ over-
lapping (7.1%), and 3′ overlapping (3.4%) lncRNAs (Table VI
in the online-only Data Supplement). Much of what is known
about sense–antisense gene pairs relates to natural antisense
transcripts and effects on local gene expression through such
processes as transcriptional interference, double-stranded
RNA-mediated events, or the guidance of chromatin remod-
eling complexes that repress or enhance protein-coding gene
expression in cis or trans.33,53,59 SENCR falls within the sub-
class of 5′ overlapping lncRNAs whose exons do not overlap
with those of the sense protein-coding (or noncoding) gene.
The terminal portion of intron 1 of SENCR overlaps a region
of high homology, likely representing conserved sequences
corresponding to the proximal 5′ promoter of FLI1. There is
another island of homology within intron 2 of SENCR, sug-
gesting SENCR could be a precursor for conserved small RNA
molecules. Although the second and third exons of SENCR
overlap the 5′ promoter region of FLI1, there is compara-
tively weak sequence conservation suggesting SENCR does
not sponge critical DNA-binding transcription factors neces-
sary for FLI1 mRNA expression (Figure 3). In fact, SENCR
and FLI1 seem to be coexpressed in several cells and tissues,
including vascular SMC. This is entirely congruent with our
inability to show a consistent effect of knocking down either
SENCR or FLI1 on the other gene’s level of expression. It is
interesting to note that there are little, if any, data on expres-
sion of FLI1 mRNA and protein in vascular SMC. Further, the
functionality of FLI1 in vascular SMC has not been assessed
although an EC-specific knockout of Fli1 showed reduced
pericytes and vascular SMC investing the dermal microvas-
culature.60 In light of FLI1 expression in vascular SMC as
reported here, it will be important to directly assess the role
of FLI1 in vascular SMC differentiation and function through
conditional gene ablation studies.
We know little as to how sense–antisense gene pairs involv-
ing lncRNAs are transcriptionally controlled. Presumably,
divergent (head to head) sense–antisense pairs share a com-
mon promoter as has been described for many bidirectionally
transcribed protein-coding genes.61 However, it is completely
unclear how other sense–antisense pairs may be transcribed,
particularly a lncRNA that is coexpressed with the sense
mRNA as shown in this report. Simultaneous expression
of FLI1 and SENCR would seem unlikely because of tran-
scriptional collision.62 How then might SENCR and FLI1
be transcribed? Perhaps there are shared promoter elements
that facilitate alternating transcription between SENCR and
FLI1. Consistent with this idea, no SENCR promoter activ-
ity was detected unless sequences encompassing the FLI1 5′
Migration index (%)
Migration index (%)
# of migrated cells
Figure 7. Effect of SENCR knockdown in a scratch wound assay of cell migration. A, Human coronary artery smooth muscle cells (HCASMCs)
were transfected with ds-Ctrl (a–c) or ds-SENCR-3 (d–f) for 72 hours after which a scratch wound was created and cell migration assessed
in quiescent (a, d) HCASMC or in similar cells stimulated for 12 hours with 10% FBS (b, c, e, and f). Cells were stained with phalloidin (green)
and 4′,6-diamidino-2-phenylindole (blue). Bars are 25 μm in a, b, d, and e and 10 μm in c and f. Arrows in f denote lamellipodia. B, Quantita-
tive measure of the area of the wound occupied by dsRNA-transfected HCASMC 12 hours after serum stimulation. C, Same experiment as in
B only HCASMC were transfected simultaneously with a control siRNA (siCtrl) or an siRNA to 1 of 2 promigratory genes. Each siRNA reduced
level of MDK or PTN mRNA by >80% (Figure V in the online-only Data Supplement). D, Effect of SENCR knockdown on HCASMC migration
in a Boyden chamber. Cells were transfected with ds-Ctrl (a), ds-3 (b), or 25 ng/mL PDGF-BB (c) for 6 hours and the fold change in number of
cells migrating through the porous membrane quantitated (E). The data reflect cell counts from 5 independent fields.
by guest on May 14, 2014 http://atvb.ahajournals.org/Downloaded from
Bell et al Novel lncRNA Discovery in Vascular Smooth Muscle Cells 1257
region were included, although the level of activity remained
much lower when compared with an EC-restricted promoter
(DLL4; not shown). Interestingly, a previous report showed
undetectable activity of the FLI1 promoter in cells expressing
high levels of FLI1 mRNA.63 This could imply there exists
a remotely acting enhancer element critical for alternating
transcription of SENCR and FLI1. Another possibility is that
SENCR and FLI1 are monoallelically expressed in a mutually
exclusive manner.64 Recently, single-cell RNA-seq analy-
sis demonstrated that as much as 24% of autosomal genes
exhibit monoallelic expression thus providing support for
this hypothesis.65 Clearly, a major task for future investiga-
tive work will be to elucidate the transcriptional control of
SENCR and other lncRNAs during vascular cell differentia-
tion or pathological conditions.
Elucidating the function of lncRNAs has been hampered
by the absence of any obvious lncRNA sequence code. One
approach to begin understanding lncRNA function is to
reduce the level of lncRNA expression and then evaluate
the transcriptome of a cell type.66 In this study, we knocked
down SENCR in HCASMC and found that the contractile
phenotype of these cells was attenuated with concomitant
increases in several promigratory genes leading to enhanced
cell motility. The mechanism for such changes in cell phe-
notype is unknown at this time; however, because SENCR
is localized to the cytoplasm it seems unlikely that it acts
through direct interaction with DNA or the recruitment of
chromatin-modifying complexes to target genes as shown for
many nuclear lncRNAs.53,67 It is more probable that SENCR
functions in some post-transcriptional capacity to effect the
observed changes in gene expression. Because all SMC con-
tractile genes were attenuated with SENCR knockdown, a
post-transcriptional mechanism would likely involve the tar-
geting of a protein or RNA that is antecedent to the SMC con-
tractile gene program. One possibility would be that SENCR
sponges a low abundant microRNA that otherwise would
function to mute the SMC contractile gene program, similar
to what has been shown for linc-MD1 in skeletal muscle.68
Other potential post-transcriptional mechanisms of action for
SENCR include stabilization, destabilization, or enhanced
ribosomal translation of pivotal RNA transcripts, as proposed
for other recently defined lncRNAs.24,69–71 The results of this
study provide a foundation for exploration of these and other
possible mechanisms of SENCR activity using emerging bio-
chemical tools to analyze lncRNA interactions with other
macromolecules in the cytoplasm.72
The explosive rise of lncRNAs in human and mouse
genomes has profound implications for future research in
vascular biology. First, unlike microRNAs, which number
≈1000 and almost universally function through a predictable
and well-defined process, lncRNAs number in the tens of
thousands and their functions and mechanisms of action will
be, arguably, as diverse as those for protein-coding genes.
This will necessitate a global effort to define all lncRNAs in
the vasculature (especially nonpolyadenylated) under normal
and stress-induced conditions and delineate their mode of
regulation and function. Second, lncRNAs such as SENCR
are poorly conserved and lack easily defined sequences that
would imply a clear function in blood vessels. The apparent
lack of orthologous mouse lncRNA genes such as SENCR
constrains the extent to which experimental analyses can be
done in a rigorous and controlled manner to gain functional
insights. However, mouse-specific lncRNAs may have limited
translational relevance to the study of human development
and disease. Structural similarity between lncRNAs having
little sequence homology may, nevertheless, exhibit compa-
rable functions across species.73,74 In this context, there is a
pressing need to gain insight into the structure of lncRNAs to
develop lncRNA codes that would facilitate functional classi-
fication across species. As a first approximation of the struc-
ture of SENCR, we used mFold (http://mfold.rna.albany.edu)
and RNAfold (http://rna.tbi.univie.ac.at/cgi-bin/RNAfold.
cgi) and found it to exhibit a stable RNA structure (Figure
VII in the online-only Data Supplement) with minimum free
energies of −486 and −470 kcal/mol, respectively. Another
implication of widespread lncRNA genes will be the need for
extreme caution and strategic design in the creation of geneti-
cally altered mice, especially when targeting the 5′ end of a
gene where inadvertent disruption of other sequences such
as lncRNAs is likely to occur. The emergence of precision-
guided genome editing (eg, CRISPR/Cas9) will be of great
value in this context.75 Finally, most genetic variation occurs
in non–protein-coding sequence space,76 which is interposed
with transcription factor binding sites such as CArG boxes11
and lncRNAs such as ANRIL.35 Historically, there has been a
notable lack of understanding as to how noncoding sequence
variations associated with disease perturb function in a cell.
Now, with increasing efforts devoted to understanding non-
coding sequences, there will be an effort to model human
single nucleotide polymorphisms associated with vascular
disease through, for example, CRISPR/Cas9-mediated point
mutations in the mouse genome. In this context, it will be
important to know whether the sequence variants in exons 2
and 3 of SENCR confer differential expression, localization,
or function in a disease setting. Altered lncRNA expression
of TIE1-AS146 and ANRIL48 has already been noted in human
In summary, we have developed a rigorous experimental
pipeline for the discovery and study of lncRNAs in human
vascular cells (Figure I in the online-only Data Supplement).
This approach uncovered many previously unrecognized
lncRNAs, including the human-specific, vascular cell–
selective SENCR, which we show is an alternatively spliced
and weakly expressed cytoplasmic 5′ overlapping antisense
lncRNA. Loss-of-function studies support the concept of
SENCR acting as a fine-tuner of the vascular SMC phenotype.
Of note, SENCR is one of the first 5′ overlapping antisense
lncRNAs (as defined here in Table VI in the online-only Data
Supplement) to be studied in detail. Future work should aim
to elucidate the regulatory control and function of SENCR in
models of human vascular SMC and EC development as well
as disease-associated processes.
We gratefully acknowledge the expertise of the University of
Rochester Genomics Research Center for performing the RNA-seq
experiments and analyzing differences in protein-coding genes as
well as generating the volcano plot.
by guest on May 14, 2014http://atvb.ahajournals.org/Downloaded from
1258 Arterioscler Thromb Vasc Biol June 2014
Sources of Funding
This work was supported by grants from the National Institutes of
Health (HL62572 and HL091168 to J.M. Miano; MH099452 to D.
Zheng and partially by HL111770; and 5P01-CA013106 to D.L.
Spector) and the American Heart Association (10SDG3670036 to
X. Long and 12POST11950002 to R.D. Bell). J.H. Bergmann was
funded by a DAAD postdoctoral fellowship.
1. Ohno S. So much “junk” DNA in our genome. Brookhaven Symp Biol.
2. Okazaki Y, Furuno M, Kasukawa T, et al.; FANTOM Consortium; RIKEN
Genome Exploration Research Group Phase I & II Team. Analysis of the
mouse transcriptome based on functional annotation of 60,770 full-length
cDNAs. Nature. 2002;420:563–573..
3. Kapranov P, Cawley SE, Drenkow J, Bekiranov S, Strausberg RL, Fodor
SP, Gingeras TR. Large-scale transcriptional activity in chromosomes 21
and 22. Science. 2002;296:916–919.
4. Carninci P, Kasukawa T, Katayama S, et al.; FANTOM Consortium;
RIKEN Genome Exploration Research Group and Genome Science
Group (Genome Network Project Core Group). The transcriptional land-
scape of the mammalian genome. Science. 2005;309:1559–1563..
5. Cheng J, Kapranov P, Drenkow J, et al. Transcriptional maps of 10 human
chromosomes at 5-nucleotide resolution. Science. 2005;308:1149–1154.
6. Mercer TR, Gerhardt DJ, Dinger ME, Crawford J, Trapnell C, Jeddeloh
JA, Mattick JS, Rinn JL. Targeted RNA sequencing reveals the deep com-
plexity of the human transcriptome. Nat Biotechnol. 2012;30:99–104.
7. Livyatan I, Harikumar A, Nissim-Rafinia M, Duttagupta R, Gingeras TR,
Meshorer E. Non-polyadenylated transcription in embryonic stem cells
reveals novel non-coding RNA related to pluripotency and differentiation.
Nucleic Acids Res. 2013;41:6300–6315.
8. Gerstein MB, Kundaje A, Hariharan M, et al. Architecture of the human reg-
ulatory network derived from ENCODE data. Nature. 2012;489:91–100.
9. Neph S, Vierstra J, Stergachis AB, et al. An expansive human regulatory
lexicon encoded in transcription factor footprints. Nature. 2012;489:83–90.
10. Zhang X, Odom DT, Koo SH, et al. Genome-wide analysis of cAMP-response
element binding protein occupancy, phosphorylation, and target gene activa-
tion in human tissues. Proc Natl Acad Sci USA. 2005;102:4459–4464.
11. Benson CC, Zhou Q, Long X, Miano JM. Identifying functional single
nucleotide polymorphisms in the human CArGome. Physiol Genomics.
12. Johnson R, Richter N, Bogu GK, Bhinge A, Teng SW, Choo SH, Andrieux
LO, de Benedictis C, Jauch R, Stanton LW. A genome-wide screen for
genetic variants that modify the recruitment of REST to its target genes.
PLoS Genet. 2012;8:e1002624.
13. The Encode Project Consortium. An integrated encyclopedia of DNA ele-
ments in the human genome. Nature. 2012;489:57–74.
14. Belinky F, Bahir I, Stelzer G, Zimmerman S, Rosen N, Nativ N, Dalah I, Iny
Stein T, Rappaport N, Mituyama T, Safran M, Lancet D. Non-redundant
compendium of human ncRNA genes in GeneCards. Bioinformatics.
15. Lagos-Quintana M, Rauhut R, Lendeckel W, Tuschl T. Identification of
novel genes coding for small expressed RNAs. Science. 2001;294:853–858.
16. Falaleeva M, Stamm S. Processing of snoRNAs as a new source of regula-
tory non-coding RNAs: snoRNA fragments form a new class of functional
RNAs. Bioessays. 2013;35:46–54.
17. Ghildiyal M, Zamore PD. Small silencing RNAs: an expanding universe.
Nat Rev Genet. 2009;10:94–108..
18. Guttman M, Amit I, Garber M, et al. Chromatin signature reveals over a
thousand highly conserved large non-coding RNAs in mammals. Nature.
19. Khalil AM, Guttman M, Huarte M, Garber M, Raj A, Rivea Morales
D, Thomas K, Presser A, Bernstein BE, van Oudenaarden A, Regev A,
Lander ES, Rinn JL. Many human large intergenic noncoding RNAs asso-
ciate with chromatin-modifying complexes and affect gene expression.
Proc Natl Acad Sci USA. 2009;106:11667–11672.
20. Cabili MN, Trapnell C, Goff L, Koziol M, Tazon-Vega B, Regev A, Rinn JL.
Integrative annotation of human large intergenic noncoding RNAs reveals
global properties and specific subclasses. Genes Dev. 2011;25:1915–1927.
21. Lee JT. Epigenetic regulation by long noncoding RNAs. Science. 2012;338:
22. Ulitsky I, Bartel DP. lincRNAs: genomics, evolution, and mechanisms.
23. Gutschner T, Hämmerle M, Diederichs S. MALAT1—a paradigm for long
noncoding RNA function in cancer. J Mol Med (Berl). 2013;91:791–801.
24. Kretz M, Siprashvili Z, Chu C, et al. Control of somatic tissue differentia-
tion by the long non-coding RNA TINCR. Nature. 2013;493:231–235.
25. St Laurent G, Shtokalo D, Dong B, Tackett MR, Fan X, Lazorthes
S, Nicolas E, Sang N, Triche TJ, McCaffrey TA, Xiao W, Kapranov P.
VlincRNAs controlled by retroviral elements are a hallmark of pluripo-
tency and cancer. Genome Biol. 2013;14:R73.
26. De Santa F, Barozzi I, Mietton F, Ghisletti S, Polletti S, Tusi BK, Muller
H, Ragoussis J, Wei CL, Natoli G. A large fraction of extragenic RNA pol
II transcription sites overlap enhancers. PLoS Biol. 2010;8:e1000384.
27. Ørom UA, Derrien T, Beringer M, Gumireddy K, Gardini A, Bussotti G, Lai F,
Zytnicki M, Notredame C, Huang Q, Guigo R, Shiekhattar R. Long noncod-
ing RNAs with enhancer-like function in human cells. Cell. 2010;143:46–58.
28. Tahira AC, Kubrusly MS, Faria MF, Dazzani B, Fonseca RS,
Maracaja-Coutinho V, Verjovski-Almeida S, Machado MC, Reis EM.
Long noncoding intronic RNAs are differentially expressed in primary
and metastatic pancreatic cancer. Mol Cancer. 2011;10:141.
29. St Laurent G, Shtokalo D, Tackett MR, Yang Z, Eremina T, Wahlestedt C,
Urcuqui-Inchima S, Seilheimer B, McCaffrey TA, Kapranov P. Intronic
RNAs constitute the major fraction of the non-coding RNA in mammalian
cells. BMC Genomics. 2012;13:504.
30. Chen J, Sun M, Kent WJ, Huang X, Xie H, Wang W, Zhou G, Shi RZ,
Rowley JD. Over 20% of human transcripts might form sense-antisense
pairs. Nucleic Acids Res. 2004;32:4812–4820.
31. Katayama S, Tomaru Y, Kasukawa T, et al.; RIKEN Genome Exploration
Research Group; Genome Science Group (Genome Network Project Core
Group); FANTOM Consortium. Antisense transcription in the mammalian
transcriptome. Science. 2005;309:1564–1566..
32. Derrien T, Johnson R, Bussotti G, et al. The GENCODE v7 catalog of
human long noncoding RNAs: analysis of their gene structure, evolution,
and expression. Genome Res. 2012;22:1775–1789.
33. Magistri M, Faghihi MA, St Laurent G 3rd, Wahlestedt C. Regulation of
chromatin structure by long noncoding RNAs: focus on natural antisense
transcripts. Trends Genet. 2012;28:389–396.
34. Lee JT, Bartolomei MS. X-inactivation, imprinting, and long noncoding
RNAs in health and disease. Cell. 2013;152:1308–1323.
35. Pasmant E, Sabbagh A, Vidaud M, Bièche I. ANRIL, a long, noncoding
RNA, is an unexpected major hotspot in GWAS. FASEB J. 2011;25:444–448.
36. Volders PJ, Helsens K, Wang X, Menten B, Martens L, Gevaert K,
Vandesompele J, Mestdagh P. LNCipedia: a database for annotated
human lncRNA transcript sequences and structures. Nucleic Acids Res.
37. Norman C, Runswick M, Pollock R, Treisman R. Isolation and properties
of cDNA clones encoding SRF, a transcription factor that binds to the c-fos
serum response element. Cell. 1988;55:989–1003.
38. Wang D, Chang PS, Wang Z, Sutherland L, Richardson JA, Small E, Krieg
PA, Olson EN. Activation of cardiac gene expression by myocardin, a tran-
scriptional cofactor for serum response factor. Cell. 2001;105:851–862.
39. Chen J, Kitchen CM, Streb JW, Miano JM. Myocardin: a component of a
molecular switch for smooth muscle differentiation. J Mol Cell Cardiol.
40. Kume T. Foxc2 transcription factor: a newly described regulator of angio-
genesis. Trends Cardiovasc Med. 2008;18:224–228.
41. Lammerts van Bueren K, Black BL. Regulation of endothelial and hema-
topoietic development by the ETS transcription factor Etv2. Curr Opin
42. De Val S, Chi NC, Meadows SM, Minovitsky S, Anderson JP, Harris IS,
Ehlers ML, Agarwal P, Visel A, Xu SM, Pennacchio LA, Dubchak I, Krieg PA,
Stainier DY, Black BL. Combinatorial regulation of endothelial gene expres-
sion by ets and forkhead transcription factors. Cell. 2008;135:1053–1064.
43. Small EM, Olson EN. Pervasive roles of microRNAs in cardiovascular
biology. Nature. 2011;469:336–342.
44. Rinn JL, Chang HY. Genome regulation by long noncoding RNAs. Annu
Rev Biochem. 2012;81:145–166.
45. Han DK, Khaing ZZ, Pollock RA, Haudenschild CC, Liau G. H19, a
marker of developmental transition, is reexpressed in human atheroscle-
rotic plaques and is regulated by the insulin family of growth factors in
cultured rabbit smooth muscle cells. J Clin Invest. 1996;97:1276–1285.
46. Li K, Blum Y, Verma A, et al. A noncoding antisense RNA in tie-1 locus
regulates tie-1 function in vivo. Blood. 2010;115:133–139.
by guest on May 14, 2014http://atvb.ahajournals.org/Downloaded from
Bell et al Novel lncRNA Discovery in Vascular Smooth Muscle Cells 1259
47. Congrains A, Kamide K, Katsuya T, Yasuda O, Oguro R, Yamamoto K,
Ohishi M, Rakugi H. CVD-associated non-coding RNA, ANRIL, modu-
lates expression of atherogenic pathways in VSMC. Biochem Biophys Res
48. Motterle A, Pu X, Wood H, Xiao Q, Gor S, Ng FL, Chan K, Cross F,
Shohreh B, Poston RN, Tucker AT, Caulfield MJ, Ye S. Functional analy-
ses of coronary artery disease associated variation on chromosome 9p21
in vascular smooth muscle cells. Hum Mol Genet. 2012;21:4021–4029.
49. Leung A, Trac C, Jin W, Lanting L, Akbany A, Sætrom P, Schones DE,
Natarajan R. Novel long noncoding RNAs are regulated by angiotensin II
in vascular smooth muscle cells. Circ Res. 2013;113:266–278.
50. Liu F, Walmsley M, Rodaway A, Patient R. Fli1 acts at the top of the tran-
scriptional network driving blood and endothelial development. Curr Biol.
51. Battich N, Stoeger T, Pelkmans L. Image-based transcriptomics in thou-
sands of single human cells at single-molecule resolution. Nat Methods.
52. Ponting CP, Oliver PL, Reik W. Evolution and functions of long noncod-
ing RNAs. Cell. 2009;136:629–641.
53. Fatica A, Bozzoni I. Long non-coding RNAs: new players in cell differen-
tiation and development. Nat Rev Genet. 2014;15:7–21..
54. Batista PJ, Chang HY. Long noncoding RNAs: cellular address codes in
development and disease. Cell. 2013;152:1298–1307.
55. Kung JT, Colognori D, Lee JT. Long noncoding RNAs: past, present, and
future. Genetics. 2013;193:651–669.
56. Arita T, Ichikawa D, Konishi H, Komatsu S, Shiozaki A, Shoda K,
Kawaguchi T, Hirajima S, Nagata H, Kubota T, Fujiwara H, Okamoto K,
Otsuji E. Circulating long non-coding RNAs in plasma of patients with
gastric cancer. Anticancer Res. 2013;33:3185–3193.
57. Pan Q, Shai O, Lee LJ, Frey BJ, Blencowe BJ. Deep surveying of alterna-
tive splicing complexity in the human transcriptome by high-throughput
sequencing. Nat Genet. 2008;40:1413–1415.
58. Kent WJ, Sugnet CW, Furey TS, Roskin KM, Pringle TH, Zahler
AM, Haussler D. The human genome browser at UCSC. Genome Res.
59. Lapidot M, Pilpel Y. Genome-wide natural antisense transcription: cou-
pling its regulation to its different regulatory mechanisms. EMBO Rep.
60. Asano Y, Stawski L, Hant F, Highland K, Silver R, Szalai G, Watson DK,
Trojanowska M. Endothelial Fli1 deficiency impairs vascular homeostasis:
a role in scleroderma vasculopathy. Am J Pathol. 2010;176:1983–1998.
61. Wakano C, Byun JS, Di LJ, Gardner K. The dual lives of bidirectional
promoters. Biochim Biophys Acta. 2012;1819:688–693..
62. Hobson DJ, Wei W, Steinmetz LM, Svejstrup JQ. RNA polymerase II col-
lision interrupts convergent transcription. Mol Cell. 2012;48:365–374.
63. Barbeau B, Bergeron D, Beaulieu M, Nadjem Z, Rassart E. Characterization
of the human and mouse Fli-1 promoter regions. Biochim Biophys Acta.
64. Raslova H, Komura E, Le Couédic JP, Larbret F, Debili N, Feunteun J,
Danos O, Albagli O, Vainchenker W, Favier R. FLI1 monoallelic expres-
sion combined with its hemizygous loss underlies Paris-Trousseau/
Jacobsen thrombopenia. J Clin Invest. 2004;114:77–84.
65. Deng Q, Ramsköld D, Reinius B, Sandberg R. Single-cell RNA-seq
reveals dynamic, random monoallelic gene expression in mammalian
cells. Science. 2014;343:193–196.
66. Panzitt K, Tschernatsch MM, Guelly C, Moustafa T, Stradner M, Strohmaier
HM, Buck CR, Denk H, Schroeder R, Trauner M, Zatloukal K. Characterization
of HULC, a novel gene with striking up-regulation in hepatocellular carci-
noma, as noncoding RNA. Gastroenterology. 2007;132:330–342.
67. Yang L, Froberg JE, Lee JT. Long noncoding RNAs: fresh perspectives
into the RNA world. Trends Biochem Sci. 2014;39:35–43.
68. Cesana M, Cacchiarelli D, Legnini I, Santini T, Sthandier O, Chinappi
M, Tramontano A, Bozzoni I. A long noncoding RNA controls muscle
differentiation by functioning as a competing endogenous RNA. Cell.
69. Faghihi MA, Modarresi F, Khalil AM, Wood DE, Sahagan BG, Morgan
TE, Finch CE, St Laurent G 3rd, Kenny PJ, Wahlestedt C. Expression
of a noncoding RNA is elevated in Alzheimer’s disease and drives rapid
feed-forward regulation of beta-secretase. Nat Med. 2008;14:723–730..
70. Gong C, Maquat LE. lncRNAs transactivate STAU1-mediated mRNA decay
by duplexing with 3’ UTRs via Alu elements. Nature. 2011;470:284–288.
71. Carrieri C, Cimatti L, Biagioli M, Beugnet A, Zucchelli S, Fedele S, Pesce
E, Ferrer I, Collavin L, Santoro C, Forrest AR, Carninci P, Biffo S, Stupka
E, Gustincich S. Long non-coding antisense RNA controls Uchl1 transla-
tion through an embedded SINEB2 repeat. Nature. 2012;491:454–457.
72. Zhu J, Fu H, Wu Y, Zheng X. Function of lncRNAs and approaches to
lncRNA-protein interactions. Sci China Life Sci. 2013;56:876–885.
73. Ulitsky I, Shkumatava A, Jan CH, Sive H, Bartel DP. Conserved function
of lincRNAs in vertebrate embryonic development despite rapid sequence
evolution. Cell. 2011;147:1537–1550.
74. Torarinsson E, Sawera M, Havgaard JH, Fredholm M, Gorodkin J.
Thousands of corresponding human and mouse genomic regions unalign-
able in primary sequence contain common RNA structure. Genome Res.
75. Wang H, Yang H, Shivalila CS, Dawlaty MM, Cheng AW, Zhang F, Jaenisch
R. One-step generation of mice carrying mutations in multiple genes by
CRISPR/Cas-mediated genome engineering. Cell. 2013;153:910–918.
76. Maurano MT, Humbert R, Rynes E, et al. Systematic localization of com-
mon disease-associated variation in regulatory DNA. Science. 2012;337:
For the first time, RNA-seq has been performed in human coronary artery smooth muscle cell for the discovery of long noncoding RNA genes.
We report the gene structure, expression, splicing, and spatial localization of a new vascular cell–selective long noncoding RNA we call
SENCR. Although SENCR has no apparent cis effect on gene expression, there is a compromise in the smooth muscle cell contractile gene
program on its knockdown with elevations in many promigratory genes. Accordingly, these cells exhibit a hyper-motile phenotype, which
can be reversed by knocking down 2 promigratory genes that are induced with SENCR knockdown. These results report the first novel long
noncoding RNA gene selectively expressed in human vascular cells and provide a framework for further study of long noncoding RNA genes
during vascular cell development and in disease processes.
by guest on May 14, 2014 http://atvb.ahajournals.org/Downloaded from
Supplemental Figure I
oligo-dT primed, single end, 20 million reads
(q)RT-PCR, RACE, Sequencing
(q)RT-PCR, exon-specific hybridize
PCR products, sequencing
miR sites, 2o structure
Cell Fractionation, RNA-FISH
Assess cis Effects
SENCR knockdown, (q)RT-PCR
SENCR knockdown > RNA-seq
- validate (RT-PCR/Western)
- GO analysis (DAVID)
- Migration assay (Rescue)
Promoter activity (Luciferase)
Non-annotated, non-protein coding,
> 2 exons, > 200 bp, > 0.7 FPKM
TopHat, Scripture, Cufflinks, Cuffdiff
Supplemental Figure I. Summary of experimental workflow. We developed this workflow for the study of
SENCR and other unannotated or uncharacterized lncRNAs. Primary cultures of HCASMC were chosen that
express contractile proteins (such as CNN1 in red) as well as the SRF transcription factor (shown in green). See
Materials and Methods and Results for further details.
Supplemental Figure II
Supplemental Figure II. Relative level of SENCR versus FLI1 in HCASMC. One divided by the
delta Ct value for SENCR (n=9) and FLI1 (n=9) RNA expression in HCASMC.
Supplemental Figure III
C C C N N N
Supplemental Figure III. Localization of SENCR. (A) Higher magnification image of boxed region in Figure 4A.
Arrows indicate SENCR transcripts localized to the cytoplasm of two HUVEC. Omission of labeled probes
revealed no background fluorescence (not shown). (B) SENCR versus XIST RNA localization in cytoplasmic (C)
or nuclear (N) fractions of the indicated cell types.
Relative RNA Level
Supplemental Figure IV
Supplemental Figure IV. Effect of SENCR knockdown on expression of pro-migratory genes in HUVEC.
Quantitative RT-PCR of indicated genes following transfection with either control dicer substrate RNA or either of
two dicer substrate RNAs that target different regions of SENCR. Note obvious reductions in SENCR upon its
targeted knockdown with minimal effects on FLI1, but associated induction of PTN and MDK.
Supplemental Figure V
Supplemental Figure V. siRNA knockdown of PTN and MDK. Quantitative RT-PCR analysis of PTN (left) and
MDK (right) mRNA levels after siRNA knockdown. The y-axis represents the normalized levels of each transcript
with si-Controls set to 1.
PDGF-BB + ds-Ctrl PDGF-BB + ds-SENCR-3
Supplemental Figure VI
Supplemental Figure VI. Effect of combined PDGF-BB treatment and SENCR knockdown on HCASMC
migration. Boyden chamber assay with HCASMC transfected with either control (A) or SENCR (B) dicer substrate
RNA followed by 6 hr stimulation with 25 ng/ml PDGF-BB. Note accentuated HCASMC migration with combined
Supplemental Figure VII
Supplemental Figure VII. Predicted secondary structure of SENCR_V1. This secondary structure was
generated with the program RNAFold. See Discussion for more details.
Materials and Methods
Cells and Tissues – Several independent isolates of primary HCASMC and human
umbilical vein endothelial cells (HUVEC) were maintained in growth medium supplied by the
manufacturer (Invitrogen). HUVEC were obtained from the Cell Culture Core in the Aab CVRI
and plated onto gelatin-coated plates/chambers. HeLa, HEK293, SKLMS (a human
leiomyosarcoma cell line of uterine SMC origin), LnCAP, and MCF7 cells were grown in medium
as specified by the manufacturer (ATCC). Human tissue RNA samples were obtained from a
commercial source (Zyagen). Dated human plasma was obtained through the University of
Rochester Medical Center Blood Bank.
RNA-Sequencing Analysis – Total RNA was isolated from HCASMC using RNeasy
extraction kit (Qiagen) under normal growth conditions or where SENCR was knocked down for
3 days with 25 nM of either a dicer substrate RNA to exon 2 (ds-SENCR-5, Table I in the online
only Data supplement) or a control dicer substrate RNA. Following bioanalyzer quality control
confirmation, RNA-seq was performed on the polyadenylated fraction using Illumina Genome
Analyzer IIx platform at the University of Rochester Medical Center Genomics Research Center
(http://www.urmc.rochester.edu/fgc/). Single-end sequencing was done at a depth of 20 million
reads per replicate (n=3). Pre-processing of raw sequence reads included demultiplexing with
CASAVA 1.8.2, transcript trimming of contaminating sequences with Sequence Cleaner
(http://sourceforge.net/projects/seqclean/), removal of vector sequences with UniVec database
(http://www.ncbi.nlm.nih.gov/VecScreen/UniVec.html), and FASTQ quality trimming using the
FASTX Toolkit (http://cancan.cshl.edu/labmembers/gordon/fastx_toolkit/index.html).
SHRiMP2.2.3 was used to align sequence reads to annotated transcripts on the UCSC
Reference Genome (build GRCh37/hg19). Quantitative analysis, including the statistical
analysis of differentially expressed genes was done with Cufflinks 2.0.2 and Cuffdiff2
(http://cufflinks.cbcb.umd.edu). For the SENCR knockdown RNA-seq experiment, the
Benjamini-Hochberg method was applied for multiple test correction (FDR < 0.05). Data output
files such as Volcano plots and scatterplots were generated with cummerbund
(http://compbio.mit.edu/cummeRbund/). Gene ontology (GO) analysis was done using DAVID 1.
All RNA-seq data were deposited into NCBI’s Gene Expression Omnibus (GSE51878).
Bioinformatics Methods for Identifying Novel LncRNAs – RNA-seq reads were aligned to
the human genome (hg19) using TopHat 1.4 2. In this analysis, two iterations of TopHat
alignment were performed in order to maximize the chance of identification of exon-exon
junctions. The alignment data were used to define novel lncRNAs following the method
described for lncRNA identification 3. The aligned data for each sample were used
independently by two complementary programs, Scripture 4 and Cufflinks 5, for assembling
transcripts independent of gene annotation. We determined the threshold for the read coverage
of each transcript across all samples by optimizing the sensitivity and specificity for identifying
full length versus partial length transcripts of protein coding genes in RefSeq 3. In the end, we
kept assembled transcripts present in both Scripture and Cufflinks outputs, and with ≥2 exons,
≥200 bp and ≥ 2.7 read coverage as reads below this threshold were deemed unreliable in
predicting exon structure. Next, we eliminated all transcripts that had an exon overlapping in
the same strand with known transcripts from available databases. We then computed the
coding potential of all remaining putative novel transcripts using PhyloCSF 6 and removed any
transcripts containing an open reading frame with PhyloCSF score ≥ 100 across any of three
reading frames. We further removed transcripts that were homologous to known protein coding
domains in the Pfam database (release 26; both PfamA and PfamB) 7 using the program
HMMER-3 (e-value = 10) 8. Lastly, we computed expression values (in FPKM, fragments per
kilobase of exon per million fragments mapped) of all remaining transcripts, together with all
coding and known non-coding genes. The final list of lncRNAs (Table II in the online only Data
supplement) comprises transcripts with FPKM >0.7.
Dicer-Substrate RNA Knockdown – Several dicer-substrate RNA (dsRNA) molecules to
different exons of SENCR or control dsRNA (ds-Ctrl) were synthesized (Integrated DNA
Technologies) and pre-tested in HCASMC and HUVEC (see Table I in the online only Data
supplement for list of all DNA molecules used in this study). The ds-Ctrl does not target known
transcribed sequences in human, mouse, or rat genomes and thus serves as a negative control
for dsRNA transfections. Briefly, cells were Lipofectamine-transfected with each dsRNA (20-30
nM) for three days and then total RNA or protein was isolated for further analysis. Results were
confirmed using at least two independent dsRNA constructs to SENCR, in up to five
independent isolates of HCASMC and HUVEC, often times by multiple investigators.
Gene Expression Assays – Total RNA was isolated using the RNeasy kit (Qiagen). RNA
integrity was assessed by spectrophotometery (NanoDrop, Thermo Scientific) and agarose gel
electrophoresis. cDNA was synthesized from 1g of total RNA using iScript (Bio-Rad) plus
random decamers and/or an oligodT primer. RT-PCR was performed using Platinum PCR
Supermix (Invitrogen) with a MyCycler thermocycler (BioRad) and PCR products were resolved
in a 1% agarose gel. Some gels shown were adjusted uniformly in Photoshop using the “invert”
function. For lncRNA validations, we included a no RT step that revealed little to no products
indicating authentic polyA+ RNAs were amplified as opposed to contaminating genomic DNA
(data not shown). Quantitative RT-PCR was performed using IQ SYBR Green Supermix with a
MyiQ single color real-time PCR detection system (BioRad). Experiments shown are
representative of multiple independent experiments using different lots of HCASMC and
HUVEC, performed by separate investigators to ensure quality control and accurate
interpretation of observed changes in gene expression.
Western Blotting – Total protein was isolated from HCASMC following ds-SENCR or ds-
Ctrl knockdown using RIPA buffer (50 mM Tris, pH 8.0, 150 mM NaCI, 0.1% SDS, 1.0% NP-40,
0.5% sodium deoxycholate and Roche protease inhibitor cocktail) and resolved in acrylamide
gels for Western blotting as previously done 9. Antibodies were LMOD1 (ProteinTech, 1:2000),
TAGLN (Abcam, 1:4000), CNN1 (DAKO, 1:2000), ANPEP (R&D, 1:1000), FLI1 (Santa Cruz,
1:250) and PPIA (Santa Cruz, 1:2000), used as an internal control.
RNA Fluorescence In Situ Hybridization (FISH) – RNA FISH with single-molecule
sensitivity was performed using QuantiGene® (QG) ViewRNA ISH Cell Assay reagents
(Affymetrix) based on branched DNA technology 10. Custom probe oligonucleotide pair pools
specific for SENCR long and short isoforms were designed and synthesized by Affymetrix as
“Type 6” (50 pairs targeting all 3 exons; product ID, VA6-14704) and “Type 4” (19 pairs
excluding exon 2; product ID, VA4-14958), respectively. A probe pair pool specific for human
PP1B housekeeping mRNA (VA1-10148, “Type 1”) and the lncRNA, NEAT1 (VA1-12621, “Type
1”) were used as cytoplasmic and nuclear controls, respectively, to assist in interpreting spatial
localization of SENCR RNA. HUVEC (± SENCR knockdown) were grown on acid-washed #1.5
glass cover slips (Thermo Scientific) to 70%-80% confluence, washed with PBS, and fixed for
30 min in fresh 0.45µm-filtered 4% paraformaldehyde (Electron Microscopy Sciences) dissolved
in Dulbecco’s PBS without CaCl2 and MgCl2 (Invitrogen). RNA-FISH was performed with minor
modifications from the manufacturer’s protocol as follows: After permeabilization using QG
Detergent Solution, cells were treated with 0.5% Triton X-100/PBS for 5 min at room
temperature. Partial protease digestion was carried out with a 1:6,000 dilution of QG Protease
K for 10 min at room temperature. Coverslips were incubated with primary probe pair sets (3-
color multiplexing) or QG Probe Set Diluent as negative control at 40oC for 3 hr. Pre-amplifiers
were incubated for an extended period of 1 hr. Between probe set incubations, cells were
washed 4 times each in QG Wash Buffer for a total of 10 min. After counter-staining with DAPI,
coverslips were mounted in home-made anti-fade mounting medium
(www.spectorlab.cshl.edu/protocols) and sealed with nail polish.
Image Acquisition and Analysis – Cells were imaged on a DeltaVision Core system
(Applied Precision) based on an inverted IX-71 microscope stand (Olympus) equipped with a
60x U-PlanApo 1.40 NA oil immersion lens (Olympus). Images were captured using a
CoolSNAP HQ CCD camera (Photometric) as 10µm image stacks with a z-spacing of 0.2µm at
a 1x1 binning. Stage, shutter and exposure were controlled through SoftWorx (Applied
Precision). Image deconvolution was performed using SoftWorx. Parameters for acquisition
and post-acquisition processing were identical for all coverslips. Analysis was done on
individual image stacks in 3D space by counting the number of SENCR hybridization signals
divided by the number of cells in each field of view (≥50 cells in ≥10 randomly chosen fields per
experiment). In some experiments, we employed two fluorescently-tagged probe sets (above)
in the absence or presence of SENCR knockdown to further confirm spatial localization. Only
signals that showed overlap of QG “Type 4” and “Type 6” probe sets were considered, thus
minimizing potential false-positive signal counts when using single color analysis.
Luciferase Assay – The putative promoter of SENCR was defined through 5’ RACE
(Ambion). Several constructs of varying 5’ and 3’ length were PCR amplified from HCASMC
genomic DNA, cloned into the pGL3 Basic Vector (Promega), and sequence confirmed (URMC
Genomics Research Center). HUVEC were plated in 12-well dishes and grown to ~60%-70%
confluency and transfected with various SENCR promoter constructs or a DLL4 reporter gene
as a positive control. Lipofectamine was used in transfections and the normalized average
luciferase activity calculated for each reporter plasmid.
Migration Assays – HCASMC were plated onto coverslips and transfected with either ds-
Ctrl or dsRNAs targeting non-overlapping sequences in SENCR. Sixty hr after transfection,
cells were “scratched” with a sterile P200 pipette tip and the culture medium immediately
changed to DMEM containing 10% FBS. 12 hr after scratch wounding, the cells were fixed and
stained with Alexa Fluor® 660 Phalloidin and DAPI (Invitrogen) according to manufacturer’s
instructions. The cells were then imaged by confocal microscopy (Olympus FV1000) and the
migratory index (percentage of cells that migrated into the time 0 wound area) was calculated
using NIH Image J software. An independent assay for migration was done using a modified
Boyden chamber (Corning). Briefly, HCASMC were transfected for three days with either ds-
Ctrl or ds-SENCR (25 nM) and then seeded into a 24-well Boyden chamber plate. Cells were
then serum-deprived overnight and subsequently treated either with PDGF-BB (25 ng/ml) or
vehicle for 6 hr. Cells were then fixed, stained with hematoxylin, and imaged with an inverted
phase contrast microscope. Migration assays are representative of multiple experiments
performed independently by two authors (RDB and XL).
Statistical Analysis – Student’s t-test or one way ANOVA followed by Tukey’s post-hoc
test were used to determine statistical significance of the means (± standard deviation) and
graphs were plotted (Graph-Pad Prism 5.0). Statistical significance was assumed at p < 0.05.
1. Dennis G, Jr., Sherman BT, Hosack DA, Yang J, Gao W, Lane HC, Lempicki RA. David:
Database for annotation, visualization, and integrated discovery. Genome Biology.
Trapnell C, Pachter L, Salzberg SL. Tophat: Discovering splice junctions with rna-seq.
Cabili MN, Trapnell C, Goff L, Koziol M, Tazon-Vega B, Regev A, Rinn JL. Integrative
annotation of human large intergenic noncoding rnas reveals global properties and
specific subclasses. Genes Dev. 2011;25:1915-1927
Guttman M, Garber M, Levin JZ, Donaghey J, Robinson J, Adiconis X, Fan L, Koziol MJ,
Gnirke A, Nusbaum C, Rinn JL, Lander ES, Regev A. Ab initio reconstruction of cell
4 Download full-text
type-specific transcriptomes in mouse reveals the conserved multi-exonic structure of
lincrnas. Nature Biotechnology. 2010;28:503-510
Trapnell C, Williams BA, Pertea G, Mortazavi A, Kwan G, van Baren MJ, Salzberg SL,
Wold BJ, Pachter L. Transcript assembly and quantification by rna-seq reveals
unannotated transcripts and isoform switching during cell differentiation. Nature
Lin MF, Jungreis I, Kellis M. Phylocsf: A comparative genomics method to distinguish
protein coding and non-coding regions. Bioinformatics. 2011;27:i275-282
Finn RD, Tate J, Mistry J, Coggill PC, Sammut SJ, Hotz HR, Ceric G, Forslund K, Eddy
SR, Sonnhammer EL, Bateman A. The pfam protein families database. Nucleic acids
Finn RD, Clements J, Eddy SR. Hmmer web server: Interactive sequence similarity
searching. Nucleic acids research. 2011;39:W29-37
Nanda V, Miano JM. Leiomodin 1: A new serum response factor-dependent target gene
expressed preferentially in differentiated smooth muscle cells. The Journal of biological
Battich N, Stoeger T, Pelkmans L. Image-based transcriptomics in thousands of single
human cells at single-molecule resolution. Nature methods. 2013;10:1127-1133