-
INVESTIGATION
Activity-Dependent Human BrainCoding/Noncoding Gene Regulatory
Networks
Leonard Lipovich,* Fabien Dachet,† Juan Cai,† Shruti Bagla,†
Karina Balan,† Hui Jia,* and Jeffrey A. Loeb*,†,1
*Center for Molecular Medicine and Genetics and †Department of
Neurology, Wayne State University School of Medicine,
Detroit,Michigan 48202
ABSTRACT While most gene transcription yields RNA transcripts
that code for proteins, a sizable proportion of the genome
generatesRNA transcripts that do not code for proteins, but may
have important regulatory functions. The brain-derived neurotrophic
factor(BDNF ) gene, a key regulator of neuronal activity, is
overlapped by a primate-specific, antisense long noncoding RNA
(lncRNA) calledBDNFOS. We demonstrate reciprocal patterns of BDNF
and BDNFOS transcription in highly active regions of human
neocortex removedas a treatment for intractable seizures. A
genome-wide analysis of activity-dependent coding and noncoding
human transcription usinga custom lncRNA microarray identified 1288
differentially expressed lncRNAs, of which 26 had expression
profiles that matchedactivity-dependent coding genes and an
additional 8 were adjacent to or overlapping with differentially
expressed protein-codinggenes. The functions of most of these
protein-coding partner genes, such as ARC, include long-term
potentiation, synaptic activity, andmemory. The nuclear lncRNAs
NEAT1, MALAT1, and RPPH1, composing an RNAse P-dependent
lncRNA-maturation pathway, werealso upregulated. As a means to
replicate human neuronal activity, repeated depolarization of SY5Y
cells resulted in sustained CREBactivation and produced an inverse
pattern of BDNF-BDNFOS co-expression that was not achieved with a
single depolarization. RNAi-mediated knockdown of BDNFOS in human
SY5Y cells increased BDNF expression, suggesting that BDNFOS
directly downregulatesBDNF. Temporal expression patterns of other
lncRNA-messenger RNA pairs validated the effect of chronic neuronal
activity on thetranscriptome and implied various lncRNA regulatory
mechanisms. lncRNAs, some of which are unique to primates, thus
appear tohave potentially important regulatory roles in
activity-dependent human brain plasticity.
THE availability of mammalian genome sequences hasmade it
possible to delineate the boundaries and struc-tures of all genes
in a genome and has demonstrated anabundance of non-protein-coding
transcriptional units thatrivals the numbers of known
protein-coding genes (reviewedin Carninci and Hayashizaki 2007).
Complex and potentiallyfunctional regulatory relationships between
protein-codingand noncoding genes, including noncoding RNA genes
thatare poorly conserved across different species, have
recentlybeen delineated (Katayama et al. 2005; Engstrom et
al.2006). These long noncoding RNA (lncRNA) genes can bedefined by
four fundamental criteria: encoding transcriptsthat lack any open
reading frames (ORFs) .100 amino
acids or possessing protein database homologies (Dingeret al.
2008); being within the known range of lengths ofmammalian mRNAs;
support by transcript-to-genome align-ments from complementary DNA
(cDNA) data; and absenceof matches to any known noncoding-RNA
classes. Function-ally, lncRNAs can have regulatory effects on
coding mRNAsthrough a number of mechanisms, including those
involvingendogenous antisense lncRNA transcripts that repress
theirsense-strand protein-coding partners (Katayama et al. 2005;Yu
et al. 2008).
Endogenous lncRNAs can also have catalytic roles, asexemplified
by the TERC telomerase RNA, and by the RNAseP and MRP RNAs required
for processing of other RNAs.lncRNAs essential to nuclear
architecture include NEAT1and NEAT2. Nuclear hormone receptors,
homeobox tran-scription factors, tumor suppressors, and immune
regulatorsare all endogenously modulated by lncRNAs (reviewed
inLipovich et al. 2010). Numerous lncRNAs are transcribed inthe
vicinity of known protein-coding genes and regulate
Copyright © 2012 by the Genetics Society of Americadoi:
10.1534/genetics.112.145128Manuscript received February 9, 2012;
accepted for publication August 24, 2012Supporting information is
available online at
http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.145128/-/DC1.1Corresponding
author: Department of Neurology, Wayne State University, 421E.
Canfield St., 3122 Elliman, Detroit, MI 48201. E-mail:
[email protected]
Genetics, Vol. 192, 1133–1148 November 2012 1133
http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.145128/-/DC1http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.145128/-/DC1mailto:[email protected]
-
those known genes through epigenetic mechanisms. Regu-lation of
protein-coding genes by overlapping, or nearbyencoded, lncRNAs is
central in cancer, cell cycle, and reprog-ramming (reviewed in
Lipovich et al. 2010; Loewer et al.2010; Orom et al. 2010). lncRNAs
encoded in an antisenseorientation to, and overlapping with, known
protein-codinggenes are particularly abundant, and the small
numberof antisense lncRNAs characterized to date is repletewith
novel functions. Endogenous antisense lncRNAs areessential in
mammalian X-inactivation (Tian et al. 2010);can directly regulate
tumor suppressors; function throughdicer-independent mechanisms;
and may be rapidly evolv-ing or not conserved, raising the
potential for new regula-tion of old genes over evolutionary time
(Lipovich et al.2010). RNA interference (RNAi) and overexpression
oflncRNAs in cell lines generate reproducible phenotypes, aswe and
others have shown (Bernard et al. 2010; SheikMohamed et al. 2010).
Hundreds of human lncRNAs bindthe polycomb repressor complex 2
(PRC2), a key epigeneticnegative regulator (Khalil et al. 2009). In
addition to high-throughput evidence of interactions with
epigenetic factors,specific epigenetic roles of lncRNAs are
beginning to be de-fined. Antisense lncRNAs actively and
specifically modulategene expression by serving as effectors of
epigenetic changesat target loci (Yu et al. 2008). These changes
include anti-sense lncRNA-mediated epigenetic silencing of the
sense-strand protein-coding gene promoter; such silencing can
beabrogated by Argonaute-2-dependent, small-RNA-mediatedsuppression
of the antisense lncRNA, resulting in “RNA ac-tivation” of the
sense gene (Morris et al. 2008). Promoter-overlapping antisense
lncRNAs can also be targeted by ex-ogenous short RNAs that regulate
sense gene expression,also via Argonaute (Schwartz et al. 2008).
Despite thesepromising examples, a majority of the thousands of
otherlncRNAs evident in transcriptome data still remain devoid
ofassigned functions.
This abundance of lncRNAs, many of which are primate-specific,
warrants a systematic assessment of whether theyhave functional,
regulatory roles. Perhaps nowhere mightthis be more important than
in the human brain that iscomposed of a diverse set of cell types
connected throughcomplex synaptic arrangements. The degree of
synapticactivity in the brain can be translated into functional
andstructural changes through activity-dependent changes ingene
expression (Katz and Shatz 1996). Although thesechanges can be
effected through direct activation of synapticgenes, they can also
be achieved through the release ofneurotrophic factors such as
brain-derived neurotrophic fac-tor (BDNF) that have direct effects
on synaptic architectureand indirect effects by producing changes
in gene expression(Isackson et al. 1991; Binder et al. 2001). BDNF,
a memberof the nerve growth factor family, regulates the survival
anddifferentiation of neuronal populations, axonal growth
andpathfinding, and dendritic growth and morphology and hasbeen
linked to many human brain disorders (reviewed inBibel and Barde
2000; Binder and Scharfman 2004; Hu
and Russek 2008). BDNF messenger (mRNA) and proteinare
upregulated by seizure activity in animal models of ep-ilepsy as
well as in human brain tissues that display in-creased epileptic
activities (Ernfors et al. 1991; Lindvallet al. 1994; Nibuya et al.
1995; Beaumont et al. 2012).The genomic locus encoding BDNF is
structurally complexand also encodes BDNFOS, a primate-specific
lncRNA thatis antisense to the coding BDNF gene (Liu et al. 2006;
Aidet al. 2007; Pruunsild et al. 2007). BDNF and BDNFOSform
double-stranded duplexes, suggesting a potential forBDNFOS to
post-transcriptionally regulate BDNF (Pruunsildet al. 2007).
Antisense knockdown of BDNFOS, in fact, hasrecently been shown to
increase BDNF expression in HEK293cells and promotes neuronal
outgrowth in vitro (Modarresiet al. 2012)
BDNF binding to its receptors results in a diverse array
ofdownstream signaling pathways including the activation ofcyclic
adenosine monophosphate response element bindingprotein (CREB),
which, in turn, can also regulate BDNF bybinding to a cognate site
within the BDNF gene (Tao et al.1998; Spencer et al. 2008).
Activation of CREB by phosphor-ylation at serine 106 as a result of
neuronal activity leads tochanges in gene expression that cause
reinforcement andstabilization of more active neuronal circuits
(reviewed inHerdegen and Leah 1998; Kandel 2001; Matynia et al.
2002;West et al. 2002). Downstream from phosphorylated CREB(pCREB),
immediate early genes have been shown to medi-ate long-lasting
changes in neuronal structure and excitabil-ity. Upstream of CREB
activation, several known signalingpathways are rapidly activated
in response to neuronal ac-tivity (Kandel 2001; reviewed in West et
al. 2002), includingCaMKinase IV, protein kinase A, and MAPK. We
have re-cently observed a pattern of transcriptional activation
inhuman brain regions where seizures start that stronglyimplicates
sustained MAPK/CREB activation and down-stream coding gene
activations that could underlie layer-specific changes in synaptic
architecture that makes theseregions prone to seizures (Rakhade et
al. 2005; Barkmeieret al. 2012; Beaumont et al. 2012).
Given that human lncRNA genes tend to be less well-conserved
than protein-coding genes, and can give rise tounique transcripts
not found in other species, we sought outa uniquely human system to
examine activity-dependentgene expression for both coding and
noncoding RNAs usinga pairwise comparison of human cortical regions
displayingvariable degrees of epileptic activities. These brain
regionswere removed as part of surgical treatment for
intractableseizures. We show that regions of human neocortex
thatdisplay increased activity and BDNF expression have re-duced
BDNFOS expression and that BDNFOS directly down-regulates BDNF in
vitro in a neuronal cell line. We developeda custommicroarray
platform to perform a transcriptome-widediscovery of other
regulatory lncRNAs and matched these tonearby or overlapping,
differentially expressed protein-codinggenes to develop a
genome-wide list of lncRNA–mRNA genepairs. Many of the coding mRNAs
identified in this way are
1134 L. Lipovich et al.
-
known to modulate activity-dependent gene expression inthe human
brain, suggesting that these lncRNA–mRNA pairsform a newly revealed
regulatory network of human brainplasticity.
Materials and Methods
Human brain tissue
Informed consent was obtained from seven patients whounderwent
surgery for medically intractable epilepsy. Ex-treme care was taken
to ensure that our study did notinfluence surgical decision making.
All patients underwentpresurgical evaluation and identification of
epileptic andcontrol regions as previously described (Rakhade et
al. 2005;Beaumont et al. 2012). To localize epileptic brain regions
thatdisplayed both clinical seizures and interictal
epileptiformdischarges (spikes), a two-stage surgical approach
usingsubdural electrodes with continuous brain-surface record-ings
(electrocorticography) and video monitoring was un-dertaken for 2–5
days. For these studies, paired tissuesamples from neocortex within
each patient displaying highand low amounts of interictal (between
seizures) spikingwere used to compare differential gene expression
as a func-tion of brain activity (Loeb 2010). To avoid
introducingadditional variables into the analysis, each block of
tissuewas examined histopathologically and demonstrated a nor-mal
six-layered neocortical structure without lesions. Thepaired
analysis of high- and low-spiking neocortex withineach patient is
also critical to isolate the variable understudy, which is the
degree of activity. Total RNA was pre-pared using a modification of
the protocol described previ-ously (Beaumont et al. 2012). The
difference was that onlygray matter was used by pooling two to
three nearby stripsof gray matter that extended from the pial
surface to thewhite matter from each block of tissue corresponding
toa given electrode location. This pooling method helps cor-rect
for differences in dissections that could lead to over-
orunder-representation of specific cortical layers.
Cell cultures, transfections, and depolarizations
The SH-SY5Y cell line (ATCC) was maintained in
Dulbecco’smodified Eagle’s medium supplemented with 10% FBS andused
for experiments. Cells between 17 and 25 passageswere transfected
with BDNFOS-targeting and BC013641-targeting small interfering RNAs
(siRNAs) by electropora-tion according to the manufacturer’s
instructions at �80%confluence (Neon electroporation system,
Invitrogen). Theelectroporation conditions used for SH-SY5Y cell
transfec-tion were the following: 1200 V; pulse width: 20 ms;
andnumber of pulses: 2. Prior to the experiments, these condi-tions
had been optimized using a condition matrix, a controlsiRNA, and
fluorescent reporters (data not shown). Singleand multiple
depolarizations of cells were induced by add-ing 100 mM KCl (final
concentration) to the medium atdifferent time points as indicated
in the Figure 5 legend.
Quantitative PCR, siRNAs, and primers
Total RNA from cultured SH-SY5Y cells was isolated with
anRNAeasy mini kit according to the manufacturer’s instruc-tions
(QIAgen). The first-strand cDNA was prepared usingSuperScript
First-Strand cDNA kit (Invitrogen), and mRNAand lncRNA expression
levels were determined by Taqmanquantitative real-time PCR (Taqman
qPCR). BDNFOS siRNAsdesignated S1, S2, S3, and S4 were
custom-designed andsynthesized by Invitrogen. The BDNFOS Taqman
primer/probe combos were custom-designed by uploading FASTA-format
sequences of preferred amplicon regions to the ABITaqman
custom-design website and were purchased fromABI/Life Technologies.
This vendor does not release the ac-tual primer and probe sequences
of custom-designed ampli-cons to the users. An siRNA against the
housekeeping geneglyceraldehyde 3-phosphate dehydrogenase (GAPDH)
wasused to rule out nonspecific effects. While this siRNAknocked
down GAPDH, it had no effects on BDNF, BDNFOS,and Lin7C at 24 and
48 hr (see supporting information,Figure S1 and File S4).
Western blot analysis
Cell lysates were prepared with SDS sample buffer (Sigma)and
subjected to Western blotting to measure CREB phos-phorylation as
described (Beaumont et al. 2012). Briefly,proteins separated on
4–20% gradient sodium dodecyl sulfate-polyacrylamide gel were
electrically transferred onto nitro-cellulose membrane. After
blocking with 5% (v/v) skim milkin TBS containing 0.05% Tween-20
for 1 hr at room temper-ature (RT), the membrane was incubated with
rabbit poly-clonal antibody against pCREB (Cell Signaling) at a
dilutionof 1:1000 for 1 hr at RT and then with specific
secondaryantibody coupled with HRP (1:5000) for 1 hr at RT.
pCREBwas visualized with ECL detection system (Pierce). The
mem-brane was then stripped and reprobed with CREB antibody(Cell
Signaling) at (1:1000) to measure total CREB.
Custom microarrays
Seven 60-mer probes per gene, unambiguously mapping byBLAT (Kent
2002) to a single genomic location and free ofinterspersed and
simple repeats, were designed using theAgilent Technologies
OpenGenomics eArray interface for5586 of the 6736 lncRNA genes from
Jia et al. (2010).The remaining lncRNA genes had been excluded
becauseof eArray failure to yield seven probes per gene or
becausethe eArray-designed probes failed our subsequent check
forgenomic uniqueness and absence of repeats. As a positivecontrol,
we also included seven probes each for 111 of the137 previously
determined protein-coding epileptic genes(Beaumont et al. 2012) and
for six housekeeping controlgenes. The eArray Fill Array feature
was used to randomlyselect control protein-coding gene probes to
fill all featuresthat would have otherwise remained vacant (,2% of
totalfeatures on a 44,000-feature, i.e., “44k,” array cell).
Theentire probe set was printed in quadruplicate on each slideusing
the Agilent 4 · 44,000 high-density oligonucleotide
lncRNA Networks in Human Brain 1135
http://www.genetics.org/content/suppl/2012/09/06/genetics.112.145128.DC1/genetics.112.145128-1.pdfhttp://www.genetics.org/content/suppl/2012/09/06/genetics.112.145128.DC1/genetics.112.145128-4.xlshttp://www.genetics.org/content/suppl/2012/09/06/genetics.112.145128.DC1/genetics.112.145128-3.pdf
-
microarray platform. Our custom lncRNA microarray con-tained the
combined resulting set of probes (File S2).
Our Agilent 60mer probes are longer than the 25mersused on the
Affymetrix platform, and more importantly, wetested each probe
(after the probe was proposed by theAgilent EArray design software)
for genomic mappinguniqueness [by University of California at Santa
Cruz (UCSC)BLAT] and for repetitive element overlaps (by
RepeatMasker).Only repeat-free and uniquely-mapping 60mer probes
wereincluded. Specificity is assured by these sequence qualities
ofthe probes as well as by our strategy of profiling each
lncRNAgene with seven unique probe sequences (not seven
replicatesof the same probe).
A dye-flip quadruplicate two-color microarray experimentwas
performed on each within-patient pair of high-spikingand
low-spiking surgically resected samples on both theAgilent human
genome-wide array (G4112A) and ourcustom lncRNA array as described,
but using a differentlabeling method (Beaumont et al. 2012). We
used the Epi-centre protocol to generate aminoallyl-amplified RNA
(aRNA)for subsequent amplification and labeling with either
cyanineor Alexa dyes. For our custom lncRNA arrays, we used
label-ing with Alexa dyes (Alexa-647 and Alexa-555,
Invitrogen)within the flip-dye design, as described by the
manufacturer(SuperScript Indirect RNA Amplification System,
Invitro-gen) (Holloway et al. 2008). For every patient, each of
thequadruplicates was hybridized on four separate slides.
Fourslides of 4 · 44,000 Agilent arrays (4 arrays, each composedof
the same set of �44,000 probes) were used to screenseven patients.
All slides were scanned as described previ-ously (Beaumont et al.
2012).
Because our lncRNA custom microarray platform is new,we also
used qPCR to validate a representative subset ofdifferentially
expressed lncRNAs. We considered which spe-cific probes were
responsible for the differential expression ofeach coding and
noncoding gene observed across all sevenpatients and used probes to
target only the region of eachtranscript that was overlapped by the
differentially expressedprobes. Positive correlation coefficients
were seen in all cases,ranging from 0.61 to 0.96 (File S1) between
the array andqPCR results within each patient; all protein-coding
gene dif-ferential expression results were from the G4112A or F
cata-log protein-coding microarray, and all lncRNA results werefrom
our lncRNA custom microarray (Figure S4 in File S4).
All lncRNA–mRNA overlaps in this work are in the anti-sense
orientation. For all lncRNA–mRNA neighbor pairs,there is a spacer
between the two genes along the genome,regardless of strand. All
probes on our microarray arestrand-specific and, therefore, even in
the case of anlncRNA–mRNA antisense pair, will exclusively profile
eitheronly the lncRNA (on the custom array described in
thisarticle) or only the mRNA (on the Agilent catalog array).
Microarray statistical methods
To identify those differentially expressed lncRNAs thatmay be
directly regulating their overlapping or neighboring
protein-coding genes, we integrated our custom lncRNAexpression
microarray data with our conventional mRNAexpression microarray
data for the in vivo high-/low-activitycortical sample pairs from
all seven patients analyzed withboth array types. For each epilepsy
patient, we had a within-patient sample pair of a high-spiking and
a low-spiking re-gion. This within-patient sample pair was
analyzed, usingthe same dye-flip quadruplicate strategy, for both
the catalogcoding (G4112A) and the custom lncRNA microarray.
Dif-ferentially expressed genes were identified from both
micro-array platforms but using the same strategy.
Consistencybetween arrays was first examined by correlating the
foldchange of all protein-coding control genes common to
botharrays, which was possible because our 111 “epileptic
tran-scriptome” genes from the prior protein-coding array
work(Beaumont et al. 2012) were used as controls on the
lncRNAarray. We used the average value of the seven probes
corre-sponding to each control gene on the lncRNA custom array.For
140 catalog (Agilent G4112A) coding-array probes cor-responding to
these 111 genes, Pearson’s correlation coeffi-cient was 0.90,
attesting to very high reproducibility betweenthe coding array and
the noncoding custom array.
Scanned microarray images from coding and noncodingmicroarrays
were analyzed by the software Agilent FeatureExtraction (Agilent,
V10.3.1) with the default protocolGE2_107_Sep09. A fluorescent
correction factor was de-termined using both qRT-PCR and Agilent
Spike-IN probes.This correction factor was then applied on the
fluorescenceintensity (fluorescence at exponent 1.125) and
improvedthe fold change prediction. The fluorescence
distributioninside each repetition of the microarray experiments
wasnormalized by R V2.11 (R Development Core Team 2010)using the
library “limma” (Smyth and Speed 2003) in a two-step process: (i)
normalization of the intensity of fluores-cence between dyes using
a Loess correction (iterations:50; span: 0.05) and (ii) independent
scaling of fluorescenceintensity on the same range across all the
arrays for each dyeusing quantile normalization. The quality of the
normaliza-tion process of the microarray fluorescence was
validatedusing MA plot density and distribution analysis.
Normalitywas asserted using the Anderson–Darling test from the
li-brary Nortest (Gross 2006). For each array, the backgroundlevel
was globally computed using the median of the fluo-rescence
intensity of the negative control probes and sub-tracted from the
signal of each probe.
Once normalized, the microarrays were further analyzedusing
standard statistical methods (Kerr and Churchill2001b; Wolfinger et
al. 2001). The differentially expressedgenes between high and low
spiking were determined usinga two-step mixed model analysis of
variance (Jin et al. 2001)with the library LME4 (Bates et al.
2009). This mixed modelapproach has been used to compute the fitted
effect and therandom effects simultaneously (Littell et al. 1996).
To im-prove the sensibility of the analysis (Kerr et al. 2000;
Jinet al. 2001), computation did not use the ratio but insteadused
dye fluorescence intensity indexed by the type of RNA
1136 L. Lipovich et al.
http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.145128/-/DC1/genetics.112.145128-3.xlshttp://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.145128/-/DC1/genetics.112.145128-1.xlshttp://www.genetics.org/content/suppl/2012/09/06/genetics.112.145128.DC1/genetics.112.145128-3.pdf
-
(Tanaka et al. 2000) (RNA from the high-spiking area orRNA from
the low-spiking area). The false discovery rate(FDR) and corrected
P-value for each gene was computedwith “R” using the library
“fdrtool” (Strimmer 2009). Thedifferentially expressed genes were
detected using fold changeand significance simultaneously (Tanaka
et al. 2000) and weredetermined as significantly differentially
expressed if theirfold change, for at least one probe per gene,
was$1.4 and iftheir FDR was #5%.
In addition to a number of custom approaches to identify8
cis-acting coding lncRNA pairs, 26 trans-acting lncRNAswere
identified as significant and activity-dependent by theirtight
correlation (Pearson’s correlation coefficient minimum of0.9) to a
well-known group of 13 activity-dependent protein-coding genes
(Rakhade et al. 2007; Beaumont et al. 2012),which themselves had
been co-expressed with a Pearson’scorrelation coefficient of 0.95.
These results were displayedgraphically using Cytoscape (Smoot et
al. 2011). To includea trans-acting lncRNA in this group, at least
one probe (ofthe seven available probes) representing the lncRNA
genehad to meet this statistical requirement.
To study the genes represented both on the catalog arrayand on
our custom array, we used genomic localization of alltranscripts
along the same human genome assembly, hg19,to find differentially
expressed transcripts from the customnoncoding and the G4112A
catalog coding array that wereclose to each other along the genome
or overlapped in anantisense orientation within a genomic region.
The genomicposition, strand, and exon/intron location information
foreach transcript is contained in the all_mrna BED file of theUCSC
Genome Database.
Results
Reciprocal patterns of BDNF and BDNFOS expression inelectrically
active human brain
Patients who fail to respond to medical management of
theirseizures can greatly benefit from a two-stage surgical
pro-cedure where long-term in vivo brain-surface recordings areused
to identify and remove epileptic brain regions. We haveused this
human system recently to identify a relativelysmall group of genes,
including BDNF that are differentiallyexpressed in regions of the
human neocortex where seizuresstart (Rakhade et al. 2005; Beaumont
et al. 2012). Whileremoving seizure-onset regions is key to a good
outcome forimproved seizure control, seizures from these brain
regionsare relatively infrequent compared to the small, but
ex-tremely frequent “interictal,” epileptic discharges that
canoccur almost constantly between seizures in some brainregions
(Staley et al. 2005). In fact, several of the genesinduced at
seizure-onset zones correlate precisely with inter-ictal spiking
rather than with seizure frequency (Rakhadeet al. 2007), suggesting
that interictal spiking may be thedriving force behind this altered
expression pattern. Consis-tently, an animal model of interictal
spiking without seizureswas sufficient to produce neuronal
layer-specific changes in
these genes (Barkmeier et al. 2012). Here we have focusedon
brain regions with different levels of interictal spiking
toidentify the relationships between coding and
noncodingtranscripts in the in vivo human brain.
Figure 1A shows a table of seven patients used for thepresent
study, together with quantified in vivo spike frequen-cies, tissue
locations, and pathological descriptions. Patientsvaried in both
sex and age, but were chosen because of theavailability of both
high- and low-interictalspiking neocorti-cal brain samples from
nearby brain regions for each patientthat were removed as part of
their seizure surgery treat-ment. Figure 1B shows how each of these
pairs was selectedwith a short sample of the electroencephalogram
recordingthat illustrates the large difference in interictal
spiking. It isimportant to emphasize that, because of genetic
differences,medication effects, and effects of tissue processing,
our in-ternally controlled experimental design is crucial
(Rakhadeet al. 2005; Beaumont et al. 2012). Although patients
arelisted with different pathological diagnoses from
multipleneocortical regions, only tissue samples that showed a
nor-mal cortical architecture were used so as not to influence
themajor variable of interest: increased brain activity.
Because of the potential regulatory relationship of tran-scripts
that code for BDNF with those that encode thepartially antisense
BDNFOS, as a first step we compared therelative expression levels
of BDNF and BDNFOS betweenpaired high- and low-spiking regions of
human neocortexusing qPCR for each patient (Figure 1C). In most
patients,BDNF expression was higher in more electrically
activeregions, whereas BDNFOS lncRNA levels were
significantlyreduced in the high-spiking regions. We used EGR1
expres-sion as a positive control for high-spiking human
corticalbrain regions as its expression has been shown to be
directlyproportionate to interictal brain activity (Rakhade et
al.2007). These results raise the possibility that increasedBDNF
levels could in part be regulated by a decrease ofthe antisense
BDNFOS RNA.
BDNFOS is a negative regulator of BDNF in an in vitrohuman cell
culture system
The genomic antisense orientation of BDNF and BDNFOSis shown in
Figure 2A, where both overlapping and non-overlapping regions are
delineated. We have previouslydemonstrated that perturbation of
lncRNA levels at multiplecis-antisense lncRNA–mRNA pairs affects
levels of the cog-nate mRNAs (Katayama et al. 2005). To distinguish
whetherthe lncRNA BDNFOS directly regulates BDNF mRNA levels,we
custom-designed three siRNAs targeting human BDNFOS(Figure 2A) and
used qPCR to interrogate BDNFOS lncRNAand BDNF mRNA levels after
the siRNA transfections. BDNFOSsiRNAs were individually transfected
into the human neuro-blastoma cell line SH-SY5Y by electroporation
and caused re-producible BDNFOS knockdown at 24 hr (all three
siRNAs) andat 48 hr (only S2). Two of the siRNAs led to knockdown
ofBDNFOS by .70% (Figure 2B). BDNFOS knockdown by
thesedouble-stranded RNAs (dsRNAs) consistently led to an
increase
lncRNA Networks in Human Brain 1137
-
in BDNF mRNA levels (between 1.5- and
3.5-fold-change),suggesting that the cis-antisense BDNFOS RNA
functions asa negative regulator of human BDNF (Figure 2B).
lncRNA genes in gene chains—loci where three or moregenes are
joined through shared antisense overlaps and bidirec-tional
promoters—are a general property of the mammaliangenome (Engstrom
et al. 2006). Human BDNFOS is part ofa three-gene genomic
positional chain: it shares a putative bi-directional promoter with
the LIN7C gene at its 59 end whilealso encompassing an exonic
cis-antisense overlap with BDNFexonic sequences at its 39 end.
BDNFOS knockdown by the samedsRNAs also increased the mRNA levels
of LIN7C, suggestingthat BDNFOS may negatively regulate other genes
at its locus.
Transcriptome-wide profiling of all known humanprotein-coding
and lncRNAs reveals activity-dependentregulatory pairs and
networks
Our functional validation of the primate-specific BDNF/BDNOS
pair suggests the potential for many more coding/noncoding
regulatory relationships across the human ge-nome that may vary as
a function of brain activity. Here weutilized these same paired RNA
samples from the sameseven patients to identify the
activity-dependent coding/noncoding interactome. To achieve this,
we developed
a custom lncRNA microarray, which allowed us to
comparetranscriptional profiles of lncRNAs to coding RNAs froma
commercial genome-wide coding array (Figure 3). Thisnew custom
array is based on our previously defined andcharacterized human
lncRNA gene catalog from experimen-tal transcriptome data
represented by cDNA and EST se-quences in GenBank, totaling 6736
lncRNA transcriptionalunits (Jia et al. 2010). Our human lncRNA
gene catalog ismostly nonredundant with respect to other recently
pub-lished human lncRNA collections (Figure S2 in File S4).
Incontrast to our custom lncRNA array, current commercialmicroarray
platforms do not adequately represent manygenomically complex loci,
including those encoding lncRNAgenes and sense–antisense pairs
(Orlov et al. 2007; Jia et al.2010).
Both platforms utilized a dye-flip (Kerr and Churchill2001a)
quadruplicate experimental design to obtain themost accurate
statistical comparison of each pair of tissuesamples from each
patient (Yao et al. 2004; Rakhade et al.2005; Beaumont et al.
2012). Each within-patient samplepair was analyzed, using the same
dye-flip quadruplicatestrategy, for both the catalog coding
(G4112A) and the cus-tom lncRNA microarray. Differentially
expressed genes wereidentified from both microarray platforms, but
using the
Figure 1 Reciprocal pattern of BDNFand BDNFOS gene expression in
electri-cally active human neocortex. (A) Summaryof human epilepsy
patients showing theratios of electrical discharges, regions
ofneocortex sampled for each, and histo-pathology. All tissue
sampled for geneexpression changes had a normal histo-logical
structure, even in the presence ofnearby structural abnormalities.
(B) Long-term brain-surface recordings obtainedprior to tissue
resection were used to dif-ferentiate electrode locations with
high-and low-spiking interictal activities foreach patient. (C)
While both the activity-dependent immediate early gene EGR1and BDNF
are constitutively upregulatedin high-spiking cortex, BDNFOS was
con-sistently downregulated in the same sam-ples. The
downregulation was significant(P = 0.016, Wilcoxon’s test; BDNFOS
fold-change , 21.1, 95% C.I.). Bars repre-sent average values for
all seven patientsshown with the color of the circles
corre-sponding to the patients shown in A.
1138 L. Lipovich et al.
http://www.genetics.org/content/suppl/2012/09/06/genetics.112.145128.DC1/genetics.112.145128-3.pdf
-
same strategy. Consistency between arrays was first exam-ined by
correlating the fold-change of all differentiallyexpressed
protein-coding control genes common to botharrays, which was
possible because our 111 epileptic tran-scriptome genes from the
prior protein-coding array workwere used as controls on the lncRNA
array. We used theaverage value of the seven probes corresponding
to eachcontrol gene on the lncRNA custom array. For 140
catalog(Agilent G4112A) coding-array probes corresponding to
these111 genes, Pearson’s correlation coefficient was 0.90,
attest-ing to very high reproducibility between the coding array
andthe noncoding custom array.
To define a gene as differentially expressed, we requiredat
least one microarray oligonucleotide probe correspondingto that
gene to be $1.4-fold differentially expressed withFDR #5% in a
groupwise analysis of all seven patients.These thresholds were
selected based on a power analysisusing this flip-dye quadruplicate
design (Loeb and Beaumont2009). Using this criterion, we identified
4004 protein-codinggenes from the catalog array (1944 upregulated
and 2060downregulated in high-activity areas; File S1). On
thelncRNA arrays, 86 of the 111 positive control genes
wereupregulated, and 1288 lncRNA genes were differentiallyexpressed
between high-activity and low-activity neocortical
regions (698 upregulated lncRNA genes and 590 downregu-lated
lncRNA genes in high-activity areas; File S2). BDNFwas represented
on both the coding microarray and, asa brain-expressed known
control gene, on the lncRNA mi-croarray. BDNF was upregulated in
high-activity tissue fromall seven patients according to both our
array platforms:coding microarray, median 3.6-fold change; lncRNA
micro-array, median 2.8-fold change.
To integrate the coding and noncoding transcriptomes ofthe human
neocortex (File S3), we then determined whichof the differentially
expressed protein-coding genes wereencoded by genomic loci
overlapping, or adjacent to, theloci which also encoded
differentially expressed lncRNAgenes as outlined in Figure 3A.
Here, we analyzed the entireextent of differential expression for
lncRNAs that participatein sense–antisense pairs and in
non-overlapping gene neigh-bor pairs such that one gene in the pair
was protein-codingwhereas the other gene encoded the lncRNA.
Specifically, weidentified all cis-encoded gene pairs in which both
a protein-coding gene and a noncoding (lncRNA) gene were
expressedfrom the same locus. We refer to these pairs as
coding–noncodinggene pairs. We then separated the pairs into two
categories—antisense and neighbor—both of which carry the
potentialfor mRNA regulation by a paired lncRNA (Jia et al.
2010;Lipovich et al. 2010). We defined a cis-antisense gene pair
astwo genes transcribed from the opposite strands of the samelocus
in a configuration such that at least some sequence inat least one
exon overlaps one exon of the other gene. Wedefined a neighbor-gene
pair as any gene pair such thatthe nearest boundaries of two
nearby, but nonoverlap-ping, genes are ,10 kb away from one
another. In thisstudy, “cis” therefore refers to any same-locus
(not neces-sarily same-allele) regulatory mechanisms, which
includeantisense-mediated regulation by lncRNAs of
protein-codinggenes that are encoded in the same locus. “Trans”
refers toany regulation involving genes encoded at multiple
distinctgenomic loci.
Of our lncRNAs differentially expressed at high-activityregions,
290 were members of sense–antisense gene pairs(File S3). We define
codifferential expression as a differen-tial expression profile of
two genes such that the differentialexpression of one gene is
either inversely or directly corre-lated with the differential
expression of the other geneacross multiple sample pairs, each of
which originates froma different patient and all of which are
statistically signifi-cant. Only 4 of these 290 mRNA–lncRNA
cis-antisense pairswere codifferentially expressed in all seven
patients (Figure3B). Only 1 of the 4 pairs (BDNFOS/BDNF) featured
aninverse differential expression profile. The other 3 pairs allhad
a positive, direct correlation. This is, in fact, not surpris-ing,
given the prevalence of synergistic, as opposed to in-verse,
expression patterns in mammalian cis-antisense pairsin response to
a stimulus or to a knockdown (Katayama et al.2005). These 3 pairs
featured lncRNAs cis-antisense to MAP-K1IP1L (MAP Kinase 1
Interacting Protein 1-like, potentiallya modulator of MAP Kinase 1,
whose role centers on the
Figure 2 Downregulation of BDNFOS induces BDNF and LIN7C
expres-sion in SH-SY5Y cells. (A) The BDNFOS/BDNF gene locus shows
antisenseoverlap between BDNF and BDNFOS (UCSC Genome Browser)
(Kent2002). Three siRNAs were generated from a non-overlapping
BDNFOSexon (S1, S2, S3). (B) Downregulation of BDNFOS lncRNA using
each ofthese siRNAs produced a corresponding increase in BDNF and,
to a lesserextent, in LIN7C mRNA levels at 24 and 48 hr. Expression
level changesare relative to a mock-electroporation negative
control. Standard errorbars are displayed. No further BDNFOS
knockdown or BDNF rescue per-sisted at the 48-hr time point for s1
and s3 (data not shown).
lncRNA Networks in Human Brain 1139
http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.145128/-/DC1/genetics.112.145128-1.xlshttp://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.145128/-/DC1/genetics.112.145128-3.xlshttp://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.145128/-/DC1/genetics.112.145128-2.xlshttp://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.145128/-/DC1/genetics.112.145128-2.xls
-
CREB activation pathway upstream of brain activity-dependentgene
expression); PURB (purine-rich element-binding pro-tein, a gene
expression regulator); and C11ORF96, whichwe have shown by
bioinformatic analysis to be a humanhomolog of the rat AG2 gene
(Matsuo et al. 2000), inducedas a consequence of sustained
long-term potentiation in vivoin rat hippocampus and therefore
implicated in neuronalplasticity. In summary, the protein-coding
genes at 3 of the4 codifferentially expressed lncRNA–mRNA
cis-antisensepairs have known neuronal functions centered on
synapticplasticity. The remaining 286 lncRNA–mRNA
cis-antisensepairs did not have any correlation between the mRNA
and
the lncRNA, within each antisense gene pair, across theseven
patients.
A total of 276 lncRNAs differentially expressed at high-activity
brain regions resided at genomic loci where a protein-coding gene
and an lncRNA gene were non-overlapping butwithin 10 kb of each
other along the genome (Jia et al.2010). However, only four
mRNA–lncRNA neighbor-genepairs were codifferentially expressed in
the groupwise anal-ysis of the seven patients (Figure 3B). These
four codiffer-entially expressed neighbor-gene pairs contained
lncRNAgenes neighboring the protein-coding genes ARC
(activity-regulated cytoskeleton-associated), a key regulator of
neuronal
Figure 3 Genome-wide analysis of human cortex reveals
activity-dependent coding–noncoding gene pairs and stand-alone
lncRNAs. (A) This experi-mental design of paired high- and
low-spiking brain samples from the seven patients shown in Figure
1A was used to interrogate both coding andnoncoding gene
transcription as a function of brain activity. A flip-dye,
quadruplicate microarray design was used with both a genome-wide
coding arrayand a novel custom lncRNA array encompassing 5586
lncRNA genes with seven probes per gene. Based on a rigorous
statistical cutoff, a total of 4044protein-coding and 1288 lncRNA
genes were identified for these seven patients (.1.4-fold; FDR ,5%
for each probe). lncRNA genes were furthersubdivided based on known
cis-antisense partners of protein-coding genes, lncRNAs located,10
kb from any known gene, or stand-alone lncRNAs.10kb from any known
gene. Due to gene chains (Engström et al. 2006), some lncRNAs
belonged simultaneously to the first two of these three categories.
(B)Pairs of differentially expressed protein-coding and lncRNA
genes that were either in an antisense overlap or ,10-kb neighbors
are shown together withstand-alone lncRNAs. Many of the affected
protein-coding genes have known functions in synaptic plasticity.
The green arrows to the left of the genepairs have been validated
by qPCR. *BDNFOS-BDNF was discovered by targeted qPCR and did not
meet statistical significance on the microarrays.
1140 L. Lipovich et al.
-
receptor endocytosis required for both synaptic plasticityand
long-term memory; L-plastin (LCP1), relevant to
theactivity-dependent MAPK/CREB activation by its placementwithin
the human MAPK interactome (Bandyopadhyay et al.2010); SMEK2, a
regulatory subunit of Ser/Thr phosphatase4; and CYR61, a secreted
protein that associates with theextracellular matrix and the cell
surface, regulates Akt acti-vation (Goodwin et al. 2010), and is
differentially expressedin autism (Garbett et al. 2008).
Of the 1288 lncRNA genes determined by our custommicroarray to
be differentially expressed at high-activityareas of the human
neocortex, 846 remain largely refractory tofunctional
interpretation as they were not genomically near,or antisense to,
any known protein-coding genes. These includethe lncRNA MALAT-1,
originally discovered as a predictor ofmetastasis and survival in
lung cancer (Ji et al. 2003) and nowknown to be a regulator of
several synaptic genes (Bernard et al.2010; Lipovich et al. 2010).
Some are key components of specificnuclear bodies, while other
lncRNAs regulate imprinting genesand still others perform essential
catalytic roles (Bernard et al.2010; reviewed in Lipovich et al.
2010). Differential expressionof 5 of these known nuclear RNAs
(Figure 3B, bottom) wassignificant. MIAT, the sole member of this
group that was down-regulated in the more active areas delineates a
novel neuronalnuclear domain (Sone et al. 2007) shown to be both a
directtarget and a putative co-activator of the transcription
factor Oct4(Sheik Mohamed et al. 2010). In contrast, the levels of
the other4 known nuclear lncRNAs were increased in the more
activeareas. These lncRNAs included: KCNQ1OT1, which may regu-late
imprinting by recruiting the DNAmethyltransferase DNMT1to
differentially methylated regions (Mohammad et al. 2010);RPPH1, the
catalytic-RNA component of RNase P, essential for
tRNA 59 end maturation and for regulating Pol III-dependenttRNA
transcription (Jarrous and Reiner 2007); NEAT1, an es-sential
component of nuclear paraspeckles that suppresses
thenucleocytoplasmic export of Alu-containing RNAs; and
NEAT2(MALAT-1), an essential component of nuclear speckles anda
regulator of synaptic genes (Bernard et al. 2010).
We also used a second unbiased approach to
identifyactivity-dependent lncRNAs with potential importance
insynaptic plasticity transcriptional regulatory networks. Wehave
shown that a number of coding genes, including EGR1,EGR2, and FOS,
are expressed in human brain in directrelation to the degree of
epileptic discharges (Rakhade et al.2007). Using co-expression
clustering of protein-codinggenes, we identified these and 8
additional genes (13 total)that have the same pattern of expression
across the sevenpatients and then identified 26 lncRNAs whose
patternof expression correlated with this group of coding
genes.Figure 4, constructed from our coding/noncoding
transcrip-tome quantitation integration by Cytoscape software
(Smootet al. 2011), illustrates co-expression of these 26
differen-tially expressed lncRNA genes with the 13
differentiallyexpressed protein-coding genes. In Figure 4, genes
thatare closer together are more tightly linked. While it
appearsthat there are two linked clusters of coding vs.
noncodinggenes, this apparent separation may in fact be due to
theoverall lower level of expression of the lncRNAs comparedto the
mRNAs (by a factor of almost 80-fold). Some of theselncRNAs, such
as NEAT1, are not at or near known coding-gene loci (Figure 3),
while others are adjacent to knowngenes that are not differentially
expressed. IL8RBP, a com-plex mosaic lncRNA transcript that
combines unique up-stream exons with an IL8RB pseudogene downstream
exon,
Figure 4 Parallel patterns of activity-dependent coding and
noncoding genes.As a means to identify activity-dependentlncRNAs
with potential roles in synapticplasticity, we probed the
expression pat-terns of 13 known activity-dependentcoding genes
against the entire dataset of lncRNAs for parallel patterns
ofexpression. This figure shows all signifi-cant relationships
between these 13genes and 26 lncRNAs identified usingan R . 0.90
cutoff. Each line representsa significant correlation and the
proxim-ity of the genes is directly proportionalto this
significance. The length of eachline is inversely proportional to
the cor-relation coefficient that is based on theaverage of
correlations from probesabove the 0.90 cutoff. The width ofeach
line is directly proportional to thenumber of probes above the 0.90
cut-off. Coding genes are shown in bluewhile lncRNAs are pink. This
figure wasprepared using Cytoscape (http://www.cytoscape.org).
lncRNA Networks in Human Brain 1141
http://www.cytoscape.orghttp://www.cytoscape.org
-
is differentially expressed, although its parental gene IL8RBand
the RUFY4 known gene that the pseudogene overlaps,are not
detectable above background in the same sampleson our
protein-coding gene arrays.
Time-dependent patterns of lncRNA–mRNAcodifferential expression
with chronic depolarization ofcultured human neuronal cells
To facilitate the study of primate-specific,
activity-dependent,coding–noncoding regulatory networks, we
developed anin vitro system of repeated depolarization using the
humanSH-SY5Y neuroblastoma cell line. Depolarization with
supra-physiological concentrations of KCl has been used
extensivelyas a means to study CREB activation and downstream
tran-scription in neuronal cells in culture (Sheng et al.
1990;Connolly and Kingsbury 2010). Figure 5A shows that, whilea
single treatment of these cells with 100 mM KCl leads only
to transient CREB activation (CREB phosphorylation, detect-able
by the pCREB Ser106 antibody), repeated 5-min expo-sures with KCl
separated by 2-hr intervals lead to moresustained CREB activation,
similar to that observed in highlyspiking human neocortex (Beaumont
et al. 2012) and in ananimal model of frequent interictal spiking
(Barkmeier et al.2012) (Figure 5B).
We then compared gene expression over a 48-hr timecourse
together with BDNF and BDNFOS expression byqPCR. Repeated
depolarization leads to a marked and moresustained increase in EGR1
activation (a marker of epilepticactivity in the brain) (Figure 5B,
bottom panel). Eventhough EGR1 goes up within 4 hr, both BDNF and
BDNFOSare initially downregulated. However, whereas BDNF andBDNFOS
almost return to baseline levels 24 hr after a
singledepolarization, cultures that were repeatedly depolarizedshow
a small but significant increase in BDNF expression,
Figure 5 Repeated depolarization in vitro can replicate patterns
of coding–noncoding gene transcription. (A) While transient CREB
phosphorylation (topand middle) is induced with a single
depolarization of Sy5Ycells with 100 mM KCl producing
downregulation of BDNF and BDNFOS (bottom), (B)repeated
depolarization produces more sustained CREB activation (top),
accompanied by a clear oscillation of the pCREB:CREB ratio with
pCREBmaximum peaks clearly following each depolarization treatment
(middle), and results in a reciprocal pattern of BDNF/BDNFOS
transcription at 24 and48 hr (shown by the brace) (bottom).
*Reciprocal expression with upregulation of BDNF at 24 hr (P =
0.027) and at 48 hr (P = 0.019) and down-regulation of BDNFOS at 24
ht (P = 0.004) and at 48 hr (P = 0.013) were observed. (A and B,
middle panels) Quantification of triplicate Western blots. (Aand B,
bottom panels) Triplicate Taqman qRT-PCR results for EGR1 as a
positive control for induced activity-dependent transcription,
together with BDNFand BDNFOS at each time point. (C) The expression
of three cis-encoded lncRNA–mRNA pairs and one known functional
lncRNA (NEAT1) was examinedin the same SY5Y repeated depolarization
time course as in B, showing multiple distinct patterns of
expression of both coding–noncoding pairs and thestand-alone
lncRNA.
1142 L. Lipovich et al.
-
while BDNFOS remains downregulated at 24 hr. Impor-tantly, the
48-hr time point extends this trend, maintainingthe sustained
increase in BDNF expression and demonstrat-ing an even stronger
downregulation of BDNFOS. While thisis likely an oversimplification
of a dynamic and complex setof regulatory interactions unrelated to
BDNFOS, the chronicdepolarization-induced reciprocal BDNFOS
depletion andBDNF mRNA increase in culture parallel the inverse
rela-tionship between BDNFOS and BDNF in high-activity areasof the
human brain shown in Figure 1. Therefore, we ap-plied this cell
culture system to interrogate the activity-dependent expression
patterns of other lncRNAs and theircis-encoded partner mRNAs from
Figure 3.
In Figure 5C, chronic depolarization of SH-SY5Y cellsrevealed a
number of distinct time-dependent patterns. Un-like the
cis-antisense BDNF/BDNFOS pair, the AF086035-MAPK1P1L cis-antisense
pair had an opposite effect ofdecreasing lncRNA and increasing mRNA
levels early inthe time course, although by 24 hr the lncRNA level
showeda slight increase. The BC013641-ARC neighbor-gene
pairdisplayed increased expression of both genes at the 4-hrtime
point in the depolarization-treated Sh-sy5y cells, mir-roring the
increased expression of both genes in the high-activity human
brain. At subsequent time points, ARC mRNAlevels decreased back to
the pretreatment levels althoughwe observed a sustained elevated
level of the BC013641lncRNA encoded near the ARC gene along the
genome (Fig-ure 5C). Since the BC013641 gene is located �6 kb
fromARC with a divergent genomic orientation relative to ARCwe
designed two custom siRNA oligonucleotides targetingBC013641.
Although both siRNAs knocked down BC013641,only one led to a
1.25-fold increase in the mRNA level of theneighboring ARC gene,
suggesting that reciprocal gene ex-pression directionality at
lncRNA–mRNA pairs may occur atneighbor-gene loci such as
BC013641-ARC, and not solely atsense–antisense loci such as
BDNFOS-BDNF (data notshown). The BC047792-PURB cis-antisense pair
showed in-creased expression of both genes, which mirrored the
co-ordinate increase observed in high-spiking brain
regions;however, in contrast to BC013641-ARC, expression of
bothtranscripts was maximal at the 8-hr time point and returnedto
baseline at 24 hr, showing no sustained increase inlncRNA
expression. Finally, we looked at the time courseof one stand-alone
nuclear lncRNA, NEAT1, which has a po-tentially far-reaching
regulatory role. NEAT1 goes up within4 hr, returns to baseline at 8
hr, but shows some chronicelevated expression at 24 hr.
Discussion
A primate-specific lncRNA regulatory mechanismfor BDNF
A striking feature of the BDNF/BDNFOS locus is the com-plexity
of its genomic landscape, which is highly representa-tive of the
genomic properties observed at lncRNA-encodingloci throughout
mammalian genomes. In addition to residing
in a three-gene chain, with LIN7C sharing its
bidirectionalpromoter and BDNF overlapping its 39 end, BDNFOS
mayhave emerged in recent mammalian evolution after
theprimate–rodent divergence. A possible recent origin for
thislncRNA gene is supported by two lines of evidence.
First,several splice sites of BDNFOS are poorly conserved outsideof
primates (data not shown), suggesting that the genomicstructure of
BDNFOS either arose or was modified specifi-cally in primate
evolution. This is consistent with our recentfinding that lncRNA
genes may comprise a majority ofprimate-specific genes (Tay et al.
2009). Second, there is noevidence for a BDNFOS-like gene between
the mouse Lin7c andBdnf genes in any public mouse cDNA and EST
sequence datarepresented by UCSC Genome Browser
transcript-to-genomealignments (not shown); however, a recently
identifiednon-orthologous postional equivalent Bdnf antisense
tran-script was found in the mouse, suggesting an
evolutionarilydistinct, but similar, mechanism for lncRNA-dependent
reg-ulation of Bdnf (Engstrom et al. 2006; Modarresi et al.2012).
This genomic and evolutionary complexity of theBDNF/BDNFOS locus
suggests that functional lncRNAs inthe human brain may be
characterized by interspecies non-conservation or high divergence
of their gene structures.This is of particular interest because of
the persistent inversecodifferential expression of the BDNF/BDNFOS
gene pair asa function of human brain activity shown here together
withthe observed increase in BDNF mRNA levels followingknockdown of
BDNFOS. Our rescue of LIN7C by BDNFOSRNAi indicates that BDNFOS
function may, in part, benuclear and epigenetic. This would be
consistent with therecent demonstration that an antisense lncRNA
acts epi-genetically by modulating target transcription (Tay et
al.2009). A possible explanation for the upregulation of bothBDNF
and LIN7C via BDNFOS RNAi might involve BDNFOS-mediated PRC2
recruitment to this locus. PRC2 associationwith lncRNAs (Khalil et
al. 2009) makes BDNFOS-PRC2binding a distinct possibility.
Knockdown of BDNFOS, pre-venting BDNFOS-mediated PRC2 targeting at
this locus,would then result in increased LIN7C and BDNF mRNA
lev-els, which is what we observed. BDNF is known to be
underepigenetic control: it is activated by acetylation of
multipleH3 lysine residues in its promoter chromatin (Tian et
al.2010) and is repressed in vivo by H3K27Me2 (Tsankovaet al.
2006), a direct PRC2-catalyzed modification (reviewedin Margueron
and Reinberg 2011). Our model for BDNFOS-mediated PRC2 recruitment
in LIN7C and BDNF suppressiondoes not contradict the concurrent
possibility of cytoplasmic,post-transcriptional BDNFOS-BDNF
regulation; in fact, theefficiency of our RNAi knockdowns of BDNFOS
implies cy-toplasmic localization, since RNA-induced silencing
complex(RISC) activity is cytoplasmic. Future work in this
fieldshould also clarify whether BDNF mRNA has a
reciprocalregulatory impact on BDNFOS.
To determine whether BDNFOS may regulate BDNFtranscription, as
opposed to splicing or RNA stability, weperformed Taqman qPCR with
a custom amplicon spanning
lncRNA Networks in Human Brain 1143
-
part of the last exon and part of the last intron of
BDNF(hg19::chr11:27,680,112-27,680,229). While there was
littlechange in our ability to detect this amplicon with BDNFOSRNAi
S1 and S2 treatment, there was an increase with S3,raising the
possibility that the BDNFOS effect is at the levelof new
transcription (data not shown). However, more workis needed to
address the question of whether BDNFOSdirectly regulates BDNF
transcription.
In summary, our findings, which center on the poorlyconserved
but functional lncRNA BDNFOS, provide a uniquelyhuman view of
activity-dependent transcriptional regulatorynetworks in the brain
whose endogenous components cannotbe modeled in rodents or other
nonprimate species. A set ofdifferentially expressed microRNAs,
including miR-30a-5p,act as post-transcriptional inhibitors of BDNF
in the prefrontalcortex (Mellios et al. 2008). Our demonstration of
the primate-specific lncRNA BDNFOS as an inhibitor of BDNF
comple-ments this earlier miRNA work, suggesting that BDNF
istargeted by multiple RNA-mediated regulatory mechanismsinvolving
short and long, ancient as well as evolutionarilyyoung, noncoding
RNAs.
This is the first study where reciprocal lncRNA–mRNA reg-ulation
is inferred from the in vivo human brain in a groupwiseanalysis of
multiple living patients and then validated byRNAi in a human
neuronal cell line. Moreover, reciprocalregulation in
sense–antisense pairs is an exception ratherthan the rule (Katayama
et al. 2005). Here, we pinpoint anlncRNA, BDNFOS, which, through
its ability to regulateBDNF, may be a key novel contributor to
epileptogenesis ina locus where future mechanistic analysis is
warranted.
Genome-wide integration of coding/noncoding RNAs asa function of
human brain activity
Recently, we generated a stringently filtered catalog ofhuman
lncRNAs and described the genomic positionalrelationships between
these lncRNAs and protein-codinggenes, providing insights into
lncRNA functions (Jia et al.2010). Despite their prominence in the
transcriptome, mostlncRNAs remain poorly understood, although
lncRNAs maycontribute to the biological complexity of regulatory
net-works (reviewed in Mattick and Makunin 2006). Becauseof their
abundance, lncRNAs may be even more importantthan microRNAs.
microRNAs function as post-transcriptionalrepressors, but lncRNAs
have additional mechanisms to pos-itively and negatively regulate
cotranscriptional and post-transcriptional alterations in gene
expression. Here we haveused these insights to develop a custom
lncRNA microarrayto provide the first genome-wide analysis of human
brainlncRNA-based regulatory networks as a function of
electricalbrain activity. Several co-expressed lncRNA/coding gene
pairsidentified here have important roles in
activity-dependentsynaptic plasticity either directly, such as BDNF
and others in-volved in the MAPK/CREB signaling, or indirectly
through theexpression of regulatory lncRNAs such as MALAT-1
(Bernardet al. 2010). Our focus on the relationship between cod-ing
mRNAs and lncRNAs with respect to brain activity is
complementary to other human brain transcriptome stud-ies such
as those focusing on developmental, regional, anddisease-related
gene expression patterns (Johnson et al.2009; Voineagu et al.
2011), but significantly expands uponthose studies through our
annotation of the human lncRNAtranscriptome and its
expression-level relationships withspecific protein-coding
genes.
This genome-wide lncRNA expression survey of electricallyactive
human neocortex has uncovered hundreds of lncRNAsdifferentially
expressed between more and less electrophys-iologically active
areas of the human neocortex. Of theselncRNAs, 26 are expressed
directly in proportion to knownactivity-dependent genes (Figure 4),
and therefore theselncRNAs could represent biomarkers and drug
targets for hu-man brain diseases, such as epilepsy (Loeb 2011).
The co-expression clustering topology (Figure 4) suggests a
networkwhere mRNAs and lncRNAs are linked by previously
unchar-acterized lncRNA nodes (such as BC009864) as hubs withspoke
edges extending simultaneously to multiple mRNAsand other lncRNAs.
We also observed eight lncRNA–mRNAcis-antisense and neighbor-gene
pairs characterized by coor-dinated differential expression of both
genes in each coding–noncoding pair, suggesting lncRNA-mediated
regulation ofprotein-coding gene expression in the brain, and the
evenmore intriguing reciprocal possibility that some mRNAs
mayfunction at the RNA level to regulate lncRNA expression or
inbidirectionally regulated feedback loops in cis. Other
lncRNAssuch as NEAT1 were detected only by the
trans-regulationanalysis, where we searched for lncRNAs whose
expressionwas highly correlated with protein-coding genes
regardless ofthe genomic mapping location of those coding genes.
Al-though the role of the NEAT2 (MALAT-1) lncRNA from nu-clear
speckles in synaptic gene regulation is now known(Bernard et al.
2010), our study complements that work byimplicating NEAT1, the
lncRNA from nuclear paraspecklesthat is encoded near the NEAT2
locus, in regulatory interac-tions with activity-dependent genes in
the brain. Our threelines of evidence for activity-dependent NEAT1
function inthe neocortex are our detection of NEAT1 as a
differentiallyexpressed lncRNA on the custom microarray analysis of
hu-man brain samples, our demonstration of activity-dependentNEAT1
expression in depolarized human SY5Y cell culture,and the
assignment of NEAT1 as a central node to a co-expression cluster of
specific coding and noncoding RNAs(Figure 4). Our cis-regulation
and trans-regulation analysesuncovered different, nonredundant sets
of lncRNAs, suggest-ing that specific lncRNAs are involved in both
types of regu-lation, which for any given lncRNA may be mutually
exclusive.These results represent the first functional evidence for
a re-markably diverse pattern of lncRNA expression in the hu-man
brain.
Functionality of coding/noncoding RNA regulatorynetworks in the
human brain
Our microarray results and our qPCR analysis of both theepilepsy
patient samples and the recurrent-depolarization
1144 L. Lipovich et al.
-
SH-SY5Y cell culture time course jointly represent the
firstdemonstration that known lncRNAs are activity-dependentboth in
vivo and in cell culture. The complex, but similar,pattern of
lncRNA–mRNA expression in activated humanbrain and in a chronically
depolarized human neuronal cellline enables the temporal
characterization of these regula-tory pathways and provides a new
system in which to studythese complex, primate-specific
transcriptional regulatorynetworks. We previously performed
time-course analysis ofmammalian cis-antisense pairs in cell
culture subjected toa specific stimulus, such as lipopolysaccharide
induction ofmacrophages, revealing a wide diversity of temporal
differ-ential expression patterns. These patterns, with
concordantor synergistic regulation of the two paired genes, were
ob-served at most of the differentially expressed
cis-antisensepairs (similar to our results in Figure 5C). Less
frequently,the patterns showed inverse or reciprocal regulation of
thetwo paired genes at a locus, similar to our results in Figure2B
(Katayama et al. 2005).
Fundamental functional roles have been previously estab-lished
for a relatively few lncRNAs. We show upregulation ofthree nuclear
RNAs—RNase P (RPPH1), NEAT1, and MALAT-1—in high-activity areas of
the neocortex. The catalytic-RNAcomponent of RNAse P is essential
for the 39-end cleavage ofboth NEAT1 (Sunwoo et al. 2009) and
NEAT2/MALAT-1(Wilusz et al. 2008). Therefore, these three lncRNAs
maycompose an lncRNA-mediated lncRNA maturation networkin highly
active brain regions (Figure S3 in File S4). Thefunction of this
induced network may be to modulate theexpression of synaptic genes,
such as those whose mRNAlevels are regulated by MALAT-1 (Bernard et
al. 2010). ThisRNA-mediated regulatory network may function either
inde-pendently from, or synergistically with, the MAPK/CREBpathway
to regulate activity-dependent gene expression.
While BDNF regulation by BDNFOS bolsters previousprecedents for
reverse-genetic approaches to functionalvalidation of lncRNA genes
(Sheik Mohamed et al. 2010),further mechanistic studies of the
novel regulatory lncRNAswill be needed to delineate the functions
of these widelyheterogeneous lncRNAs. For example, it is important
to dis-tinguish nuclear epigenetic from post-transcriptional and
cy-toplasmic regulatory mechanisms of lncRNAs. Such studiesshould
be aided by structural insights into the mammalianlncRNAome,
following in the footsteps of existing whole-transcriptome
empirical RNA secondary structure delinea-tion methodologies
(Kertesz et al. 2010). It has also becomeincreasingly evident that
human lncRNAs perform diverseyet crucial functions, one of which is
to regulate mRNAstability on a transcriptome-wide scale through
repetitiveelements embedded in exons of many lncRNAs (Gong
andMaquat 2011). Ribonucleoprotein complexes that enablelncRNA
function and complexes that facilitate lncRNA-mediated regulation
of mRNAs in sense–antisense pairscan be identified by affinity
columns and mass spectrometricanalysis. This identification will
allow therapeutic targetingof lncRNA–mRNA regulatory relationships.
Finally, integrated
differential expression analysis of the protein-coding and
thelong noncoding transcriptome represents only a limited en-try
point into transcriptional regulatory networks
underlyingactivity-dependent gene expression. A comprehensive
as-sessment of this network is possible only if all
transcriptclasses, including mRNAs, lncRNAs, microRNAs, and the
re-cently discovered endo-siRNAs (Smalheiser et al. 2011),
areprofiled jointly.
The genomic complexity of the human AK093366-AG2lncRNA–mRNA
cis-antisense pair (Figure S5 in File S4) isreminiscent of that
observed in the BDNFOS-BDNF lncRNA–mRNA cis-antisense pair. Both
the BDNFOS lncRNA gene andthe AK093366 lncRNA gene contained
primate-specificsplice sites. The splice donor of AK093366’s sole
intron isprimate-specific because it is harbored within an AluJb
re-peat. Alu repeats are the best-known class of primate-specific
interspersed repeats (Greally 2002), and thereforekey gene
structure elements, including splice sites, con-tained within Alu
repeats provide direct evidence that thecorresponding gene
structures either arose or were modifiedafter the mammalian
radiation, specifically in the primatelineage. Notably,
EST-supported cis-antisense lncRNA tran-scription of the
Alu-containing AK093366 transcriptionalunit extends substantially
beyond the UCSC C11ORF96(AG2) gene model and well into the C11ORF96
ORF. Thisunderscores the utility of EST data, much of which
remainsunincorporated into reference gene models and annotationsin
delineating the boundaries of lncRNA genes, includingthose involved
in putative regulatory relationships with pro-tein-coding
counterparts. While Alu-containing lncRNAs haverecently been
implicated in the in trans post-transcriptionalregulation of gene
expression via effecting mRNA decay (Gongand Maquat 2011), our
analysis suggests distinct cis-regulatoryroles in overlapping-gene
regulation for certain Alu-containinglncRNAs—specifically,
AK093366. Two of the four codiffer-entially expressed lncRNA–mRNA
cis-antisense pairs in thehuman neocortex, BDNFOS-BDNF and
AK093366-AG2, thusfeature primate-specific sequence at lncRNA gene
splicejunctions. The protein-coding genes BDNF and AG2
areoverlapped by endogenous antisense lncRNAs containingexonic Alu
repeats, and these gene pairs are codifferen-tially expressed in
active areas of the human epilepticneocortex.
BDNFOS-mediated regulation of BDNF provides initialevidence that
primate-specific regulation of conserved protein-coding genes by
cis-antisense lncRNAs takes place in epi-lepsy, a complex human
brain disorder. The co-expressionof mRNAs and nonconserved lncRNAs
at loci other thanBDNF, including AG2, suggests that
primate-specific regula-tion of conserved genes by nonconserved
lncRNAs in thehuman brain may not be unique to the BDNF locus.
Acknowledgments
We thank J. B. Brown, Department of Statistics, Universityof
California, Berkeley, for helpful comments on the manu-
lncRNA Networks in Human Brain 1145
http://www.genetics.org/content/suppl/2012/09/06/genetics.112.145128.DC1/genetics.112.145128-3.pdfhttp://www.genetics.org/content/suppl/2012/09/06/genetics.112.145128.DC1/genetics.112.145128-3.pdf
-
script. This work was funded by National Institutes of
Health(NIH)/National Institute of Neurological Disorders andStroke
grants R01NS045207 and R01NS058802 (J.A.L.)and by internal funds
from Wayne State University (L.L.).Microarray development was
supported by NIH/NationalInstitute on Drug Abuse grant
1R03DA026021-01 (L.L.).Microarray scanning was performed by the
Core Facility ofthe Department of Pediatrics, School of Medicine,
WayneState University.
Literature Cited
Aid, T., A. Kazantseva, M. Piirsoo, K. Palm, and T. Timmusk,2007
Mouse and rat BDNF gene structure and expression re-visited. J.
Neurosci. Res. 85: 525–535.
Bandyopadhyay, S., C. Y. Chiang, J. Srivastava, M. Gersten,
S.White et al., 2010 A human MAP kinase interactome. Nat.Methods 7:
801–805.
Barkmeier, D. T., D. Senador, K. Leclercq, D. Pai, J. Hua et
al.,2012 Electrical, molecular and behavioral effects of
interictalspiking in the rat. Neurobiol. Dis. 47: 92–101.
Bates, D., M. Maechler, and B. Bolker, 2009 lme4: Linear
mexed-ef-fects models using S4 classes. Available at:
https://r-forge.r-project.org/R/?group_id=60. Accessed: September,
2012.
Beaumont, T., B. Yao, A. Shah, G. Pai, J. Kapatos, and J. A.
Loeb,2012 Layer-specific CREB target gene induction in
humanneocortical epilepsy. J. Neurosci. (in press).
Bernard, D., K. V. Prasanth, V. Tripathi, S. Colasse, T.
Nakamuraet al., 2010 A long nuclear-retained non-coding RNA
regulatessynaptogenesis by modulating gene expression. EMBO J.
29:3082–3093.
Bibel, M., and Y. A. Barde, 2000 Neurotrophins: key regulators
ofcell fate and cell shape in the vertebrate nervous system.
GenesDev. 14: 2919–2937.
Binder, D. K., and H. E. Scharfman, 2004 Brain-derived
neuro-trophic factor. Growth Factors 22: 123–131.
Binder, D. K., S. D. Croll, C. M. Gall, and H. E. Scharfman,2001
BDNF and epilepsy: Too much of a good thing? TrendsNeurosci. 24:
47–53.
Carninci, P., and Y. Hayashizaki, 2007 Noncoding RNA
transcriptionbeyond annotated genes. Curr. Opin. Genet. Dev. 17:
139–144.
Connolly, S., and T. J. Kingsbury, 2010 Caffeine modulates
CREB-dependent gene expression in developing cortical neurons.
Bio-chem. Biophys. Res. Commun. 397: 152–156.
Dinger, M. E., K. C. Pang, T. R. Mercer, and J. S. Mattick,2008
Differentiating protein-coding and noncoding RNA: chal-lenges and
ambiguities. PLOS Comput. Biol. 4: e1000176.
Engstrom, P. G., H. Suzuki, N. Ninomiya, A. Akalin, L. Sessa et
al.,2006 Complex loci in human and mouse genomes. PLoSGenet. 2:
e47.
Ernfors, P., J. Bengzon, Z. Kokaia, H. Persson, and O.
Lindvall,1991 Increased levels of messenger RNAs for
neurotrophicfactors in the brain during kindling epileptogenesis.
Neuron 7:165–176.
Garbett, K., P. J. Ebert, A. Mitchell, C. Lintas, B. Manzi et
al.,2008 Immune transcriptome alterations in the temporal cortexof
subjects with autism. Neurobiol. Dis. 30: 303–311.
Gong, C., and L. E. Maquat, 2011 lncRNAs transactivate
STAU1-mediated mRNA decay by duplexing with 39 UTRs via Alu
ele-ments. Nature 470: 284–288.
Goodwin, C. R., B. Lal, X. Zhou, S. Ho, S. Xia et al., 2010
Cyr61mediates hepatocyte growth factor-dependent tumor cellgrowth,
migration, and Akt activation. Cancer Res. 70: 2932–2941.
Greally, J. M., 2002 Short interspersed transposable
elements(SINEs) are excluded from imprinted regions in the human
ge-nome. Proc. Natl. Acad. Sci. USA 99: 327–332.
Gross, J., 2006 nortest: Tests for Normality. R package
version1.0. Available at:
http://cran.r-project.org/web/packages/nortest/.Accessed:
September, 2012.
Herdegen, T., and J. D. Leah, 1998 Inducible and
constitutivetranscription factors in the mammalian nervous system:
controlof gene expression by Jun, Fos and Krox, and CREB/ATF
pro-teins. Brain Res. Brain Res. Rev. 28: 370–490.
Holloway, M. G., G. D. Miles, A. A. Dombkowski, and D. J.
Waxman,2008 Liver-specific hepatocyte nuclear factor-4alpha
defi-ciency: greater impact on gene expression in male than in
fe-male mouse liver. Mol. Endocrinol. 22: 1274–1286.
Hu, Y., and S. J. Russek, 2008 BDNF and the diseased
nervoussystem: a delicate balance between adaptive and
pathologicalprocesses of gene regulation. J. Neurochem. 105:
1–17.
Isackson, P. J., M. M. Huntsman, K. D. Murray, and C. M.
Gall,1991 BDNF mRNA expression is increased in adult rat fore-brain
after limbic seizures: temporal patterns of induction dis-tinct
from NGF. Neuron 6: 937–948.
Jarrous, N., and R. Reiner, 2007 Human RNase P: a
tRNA-process-ing enzyme and transcription factor. Nucleic Acids
Res. 35:3519–3524.
Ji, P., S. Diederichs, W. Wang, S. Boing, R. Metzger et al.,2003
MALAT-1, a novel noncoding RNA, and thymosin beta4predict
metastasis and survival in early-stage non-small cell lungcancer.
Oncogene 22: 8031–8041.
Jia, H., M. Osak, G. K. Bogu, L. W. Stanton, R. Johnson et
al.,2010 Genome-wide computational identification and
manualannotation of human long noncoding RNA genes. RNA
16:1478–1487.
Jin, W., R. M. Riley, R. D. Wolfinger, K. P. White, G.
Passador-Gurgel et al., 2001 The contributions of sex, genotype
andage to transcriptional variance in Drosophila melanogaster.Nat.
Genet. 29: 389–395.
Johnson, M. B., Y. I. Kawasawa, C. E. Mason, Z. Krsnik, G.
Coppolaet al., 2009 Functional and evolutionary insights into
humanbrain development through global transcriptome analysis.
Neu-ron 62: 494–509.
Kandel, E. R., 2001 The molecular biology of memory storage: a
di-alogue between genes and synapses. Science 294: 1030–1038.
Katayama, S., Y. Tomaru, T. Kasukawa, K. Waki, M. Nakanishiet
al., 2005 Antisense transcription in the mammalian tran-scriptome.
Science 309: 1564–1566.
Katz, L. C., and C. J. Shatz, 1996 Synaptic activity and the
con-struction of cortical circuits. Science 274: 1133–1138.
Kent, W. J., 2002 BLAT: the BLAST-like alignment tool.
GenomeRes. 12: 656–664.
Kerr, M. K., and G. A. Churchill, 2001a Experimental design
forgene expression microarrays. Biostatistics 2: 183–201.
Kerr, M. K., and G. A. Churchill, 2001b Statistical design and
theanalysis of gene expression microarray data. Genet. Res.
77:123–128.
Kerr, M. K., M. Martin, and G. A. Churchill, 2000 Analysis
ofvariance for gene expression microarray data. J. Comput. Biol.7:
819–837.
Kertesz, M., Y. Wan, E. Mazor, J. L. Rinn, R. C. Nutter et
al.,2010 Genome-wide measurement of RNA secondary structurein
yeast. Nature 467: 103–107.
Khalil, A. M., M. Guttman, M. Huarte, M. Garber, A. Raj et
al.,2009 Many human large intergenic noncoding RNAs associatewith
chromatin-modifying complexes and affect gene expres-sion. Proc.
Natl. Acad. Sci. USA 106: 11667–11672.
Lindvall, O., Z. Kokaia, J. Bengzon, E. Elmer, and M.
Kokaia,1994 Neurotrophins and brain insults. Trends Neurosci.
17:490–496.
1146 L. Lipovich et al.
http://https://r-forge.r-project.org/R/?group_id=60http://https://r-forge.r-project.org/R/?group_id=60http://cran.r-project.org/web/packages/nortest/
-
Lipovich, L., R. Johnson, and C. Y. Lin, 2010 MacroRNA
under-dogs in a microRNA world: evolutionary, regulatory, and
bio-medical significance of mammalian long non-protein-codingRNA.
Biochim. Biophys. Acta 1799: 597–615.
Littell, R., G. Milliken, W. Stroup, and R. Wolfinger, 1996
SASSystem for Mixed Models. SAS Institute, Carey, NC.
Liu, Q. R., L. Lu, X. G. Zhu, J. P. Gong, Y. Shaham et al.,2006
Rodent BDNF genes, novel promoters, novel splice var-iants, and
regulation by cocaine. Brain Res. 1067: 1–12.
Loeb, J. A., 2010 A human systems biology approach to
discovernew drug targets in epilepsy. Epilepsia 51(Suppl. 3):
171–177.
Loeb, J. A., 2011 Identifying targets for preventing epilepsy
usingsystems biology. Neurosci. Lett. 497: 205–212.
Loeb, J. A., and T. L. Beaumont, 2009 What goes in is what
comesout: the design and analysis of microarray experiments, p.
213in Bioinformatics for Systems Biology, edited by S. Krawetz.
Hu-mana Press, New York.
Loewer, S., M. N. Cabili, M. Guttman, Y. H. Loh, K. Thomas et
al.,2010 Large intergenic non-coding RNA-RoR modulates
re-programming of human induced pluripotent stem cells. Nat.Genet.
42: 1113–1117.
Margueron, R., and D. Reinberg, 2011 The Polycomb complexPRC2
and its mark in life. Nature 469: 343–349.
Matsuo, R., A. Murayama, Y. Saitoh, Y. Sakaki, and K.
Inokuchi,2000 Identification and cataloging of genes induced by
long-lasting long-term potentiation in awake rats. J. Neurochem.
74:2239–2249.
Mattick, J. S., and I. V. Makunin, 2006 Non-coding RNA. Hum.Mol.
Genet. 15(Spec. No. 1): R17–R29.
Matynia, A., S. A. Kushner, and A. J. Silva, 2002 Genetic
ap-proaches to molecular and cellular cognition: a focus on LTPand
learning and memory. Annu. Rev. Genet. 36: 687–720.
Mellios, N., H. S. Huang, A. Grigorenko, E. Rogaev, and S.
Akbarian,2008 A set of differentially expressed miRNAs, including
miR-30a-5p, act as post-transcriptional inhibitors of BDNF in
pre-frontal cortex. Hum. Mol. Genet. 17: 3030–3042.
Modarresi, F., M. A. Faghihi, M. A. Lopez-Toledano, R. P.
Fatemi,M. Magistri et al., 2012 Inhibition of natural antisense
tran-scripts in vivo results in gene-specific transcriptional
upregula-tion. Nat. Biotechnol. 30: 453–459.
Mohammad, F., T. Mondal, N. Guseva, G. K. Pandey, and C.
Kan-duri, 2010 Kcnq1ot1 noncoding RNA mediates transcriptionalgene
silencing by interacting with Dnmt1. Development 137:2493–2499.
Morris, K. V., S. Santoso, A. M. Turner, C. Pastori, and P. G.
Haw-kins, 2008 Bidirectional transcription directs both
transcrip-tional gene activation and suppression in human cells.
PLoSGenet. 4: e1000258.
Nibuya, M., S. Morinobu, and R. S. Duman, 1995 Regulation ofBDNF
and trkB mRNA in rat brain by chronic electroconvulsiveseizure and
antidepressant drug treatments. J. Neurosci. 15:7539–7547.
Orlov, Y. L., J. Zhou, L. Lipovich, A. Shahab, and V. A.
Kuznetsov,2007 Quality assessment of the Affymetrix U133A&B
probe-sets by target sequence mapping and expression data
analysis.In Silico Biol. 7: 241–260.
Orom, U. A., T. Derrien, M. Beringer, K. Gumireddy, A. Gardiniet
al., 2010 Long noncoding RNAs with enhancer-like functionin human
cells. Cell 143: 46–58.
Pruunsild, P., A. Kazantseva, T. Aid, K. Palm, and T.
Timmusk,2007 Dissecting the human BDNF locus: bidirectional
tran-scription, complex splicing, and multiple promoters.
Genomics90: 397–406.
R Development Core Team, 2010 R: a language and environmentfor
statistical computing. R: a language and environment forstatistical
computing. R Foundation for Statistical Computing,
Vienna. Available at: http://www.r-project.org/. Accessed:
Sep-tember, 2012.
Rakhade, S. N., B. Yao, S. Ahmed, E. Asano, T. L. Beaumont et
al.,2005 A common pattern of persistent gene activation in hu-man
neocortical epileptic foci. Ann. Neurol. 58: 736–747.
Rakhade, S. N., A. K. Shah, R. Agarwal, B. Yao, E. Asano et
al.,2007 Activity-dependent gene expression correlates with
in-terictal spiking in human neocortical epilepsy. Epilepsia
48(Suppl. 5): 86–95.
Schwartz, J. C., S. T. Younger, N. B. Nguyen, D. B. Hardy, B.
P.Monia et al., 2008 Antisense transcripts are targets for
activat-ing small RNAs. Nat. Struct. Mol. Biol. 15: 842–848.
Sheik Mohamed, J., P. M. Gaughwin, B. Lim, P. Robson, and
L.Lipovich, 2010 Conserved long noncoding RNAs transcription-ally
regulated by Oct4 and Nanog modulate pluripotency inmouse embryonic
stem cells. RNA 16: 324–337.
Sheng, M., G. McFadden, and M. E. Greenberg, 1990
Membranedepolarization and calcium induce c-fos transcription via
phos-phorylation of transcription factor CREB. Neuron 4:
571–582.
Smalheiser, N. R., G. Lugli, J. Thimmapuram, E. H. Cook, and
J.Larson, 2011 Endogenous siRNAs and noncoding RNA-de-rived small
RNAs are expressed in adult mouse hippocampusand are up-regulated
in olfactory discrimination training. RNA17: 166–181.
Smoot, M. E., K. Ono, J. Ruscheinski, P. L. Wang, and T.
Ideker,2011 Cytoscape 2.8: new features for data integration
andnetwork visualization. Bioinformatics 27: 431–432.
Smyth, G. K., and T. Speed, 2003 Normalization of cDNA
micro-array data. Methods 31: 265–273.
Sone, M., T. Hayashi, H. Tarui, K. Agata, M. Takeichi et
al.,2007 The mRNA-like noncoding RNA Gomafu constitutesa novel
nuclear domain in a subset of neurons. J. Cell Sci.120:
2498–2506.
Spencer, T. K., W. Mellado, and M. T. Filbin, 2008 BDNF
activatesCaMKIV and PKA in parallel to block MAG-mediated
inhibitionof neurite outgrowth. Mol. Cell. Neurosci. 38:
110–116.
Staley, K., J. L. Hellier, and F. E. Dudek, 2005 Do interictal
spikesdrive epileptogenesis? Neuroscientist 11: 272–276.
Strimmer, K., 2009 fdrtool: estimation and control of (local)
falsediscovery rates, R package version 1.2.6. Available at:
http://strimmerlab.org/software.html. Accessed: September,
2012.
Sunwoo, H., M. E. Dinger, J. E. Wilusz, P. P. Amaral, J. S.
Matticket al., 2009 MEN epsilon/beta nuclear-retained
non-codingRNAs are up-regulated upon muscle differentiation and are
es-sential components of paraspeckles. Genome Res. 19: 347–359.
Tanaka, T. S., S. A. Jaradat, M. K. Lim, G. J. Kargul, X. Wang
et al.,2000 Genome-wide expression profiling of mid-gestation
pla-centa and embryo using a 15,000 mouse developmental
cDNAmicroarray. Proc. Natl. Acad. Sci. USA 97: 9127–9132.
Tao, X., S. Finkbeiner, D. B. Arnold, A. J. Shaywitz, and M.
E.Greenberg, 1998 Ca2+ influx regulates BDNF transcriptionby a CREB
family transcription factor-dependent mechanism.Neuron 20:
709–726.
Tay, S. K., J. Blythe, and L. Lipovich, 2009 Global discovery
ofprimate-specific genes in the human genome. Proc. Natl. Acad.Sci.
USA 106: 12019–12024.
Tian, D., S. Sun, and J. T. Lee, 2010 The long noncoding
RNA,Jpx, is a molecular switch for X chromosome inactivation.
Cell143: 390–403.
Tsankova, N. M., O. Berton, W. Renthal, A. Kumar, R. L. Neve et
al.,2006 Sustained hippocampal chromatin regulation in a mousemodel
of depression and antidepressant action. Nat. Neurosci.9:
519–525.
Voineagu, I., X. Wang, P. Johnston, J. K. Lowe, Y. Tian et
al.,2011 Transcriptomic analysis of autistic brain reveals
conver-gent molecular pathology. Nature 474: 380–384.
lncRNA Networks in Human Brain 1147
http://www.r-project.org/http://strimmerlab.org/software.htmlhttp://strimmerlab.org/software.html
-
West, A. E., E. C. Griffith, and M. E. Greenberg, 2002
Regulationof transcription factors by neuronal activity. Nat. Rev.
Neurosci.3: 921–931.
Wilusz, J. E., S. M. Freier, and D. L. Spector, 2008 39 end
process-ing of a long nuclear-retained noncoding RNA yields a
tRNA-likecytoplasmic RNA. Cell 135: 919–932.
Wolfinger, R. D., G. Gibson, E. D. Wolfinger, L. Bennett, H.
Hamadehet al., 2001 Assessing gene significance from cDNA
microarrayexpression data via mixed models. J. Comput. Biol. 8:
625–637.
Yao, B., S. N. Rakhade, Q. Li, S. Ahmed, R. Krauss et al.,2004
Accuracy of cDNA microarray methods to detect smallgene expression
changes induced by neuregulin on breast epi-thelial cells. BMC
Bioinformatics 5: 99.
Yu, W., D. Gius, P. Onyango, K. Muldoon-Jacobs, J. Karp et
al.,2008 Epigenetic silencing of tumour suppressor gene p15 byits
antisense RNA. Nature 451: 202–206.
Communicating editor: M. Johnston
1148 L. Lipovich et al.
-
GENETICSSupporting Information
http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.145128/-/DC1
Activity-Dependent Human BrainCoding/Noncoding Gene Regulatory
Networks
Leonard Lipovich, Fabien Dachet, Juan Cai, Shruti Bagla,Karina
Balan, Hui Jia, and Jeffrey A. Loeb
Copyright © 2012 by the Genetics Society of AmericaDOI:
10.1534/genetics.112.145128
-
L. Lipovich et al. 2 SI
Files S1 – S3
Supporting Data
Available for download at
http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.145128/-‐/DC1.
File S1 Dataset of coding
mRNA gene expression changes between
high and low spiking human
cortex for 7 patients. Each
pair of samples was measured in
quadruplicate using a flip-‐dye
design as described in the
methods. Those genes reaching
significance are indicated as a
“1.”
File S2 Dataset of long
non-‐coding RNA gene expression
changes between high and low
spiking human cortex for 7
patients. Each pair of samples
was measured in quadruplicate using
a flip-‐dye design with 7
unique probes for each gene as
described in the methods.
Those genes reaching significance are
indicated as a “1.”
File S3 Integration of
coding / non-‐coding transcriptomes.
Here, we determined which of
the differentially expressed
protein-‐coding genes were encoded by
genomic loci overlapping, or adjacent
to, the loci which also encoded
differentially expressed lncRNA genes
as outlined in figure 3a.
-
L. Lipovich et al. 3 SI
File S4
Supporting Figures
Figure S1 Control
experiment to rule out non-‐specific
effects of siRNA conditions.
An siRNA against GAPDH was used
on SY5Y cells in order to
be certain that non-‐specific or
off-‐target effects are not
responsible for the results shown
in Figure 2 of the manuscript.
While GAPDH was significantly
knocked down at 24 and 48
hours, there was no effect on
BDNFOS, BDNF, or Lin7C.
Results shown were each done in
triplicate real-‐time qPCR with
duplicates at 24 hours and a
single sample at 48 hours.
-
L. Lipovich et al. 4 SI
Figure S2 The majority
of lncRNAs in our dataset are
nonredundant by genomic position
relative to lincRNAs and long
ncRNAs published by two other
groups.
-
L. Lipovich et al. 5 SI
Figure S3 Three nuclear
lncRNAs whose levels are increased
in more electrophysiologically active
areas of the human neocortex
form a network: RNAse P is
essential for maturation of both
NEAT1 and NEAT2 (MALAT1).
-
L. Lipovich et al. 6 SI
Figure S4 Taqman
qRTPCR results closely parallel
microarray results for lncRNA and
mRNA differential expression at
lncRNA-‐mRNA cis-‐pairs across the
within-‐patient sample pairs of
high-‐ and low-‐activity neocortical
regions. MALAT-‐1 is not shown
because of the discrepancy between
its microarray probeset coverage and
its Taqman amplicons coverage.
-
L. Lipovich et al. 7 SI
Figure S5 Genomic
complexity of the human AK093366-‐AG2
lncRNA-‐mRNA cis-‐antisense pair which
is co-‐differentially-‐expressed in human
neocortical epilepsy. Red vertical
line: lncRNA splice acceptor within
an Alu repeat. Magenta arrow:
extent of EST-‐supported cis-‐antisense
lncRNA transcription overlapping the
AG2 protein-‐coding gene. Genomic
diagram was generated using the
UCSC Genome Browser (Kent 2002).
-
L. Lipovich et al. 2 SI
Files S1 – S3
Supporting Data
Available for download at
http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.145128/-‐/DC1.
File S1 Dataset of coding
mRNA gene expression changes between
high and low spiking human
cortex for 7 pa