Supplemental Materialx

richessewoozyBiotechnology

Oct 1, 2013 (4 years and 1 month ago)

109 views

1


Supplemental methods and materials


Primer sequences

Sequences of qPCR primers for measuring gene expression, noncoding RNAs, or DNA
enrichment after ChIP, mutagenesis primers for generation of FUS constructs, and T7 promoter
-
conjugated primers for synthesis of
in vitro

transcription substrates for noncoding

RNAs will be
promptly provided upon request.


Additional siRNAs

siCDK9 sequence was
rUrCrCrUrGrArArGrCrUrGrGrCrArGrArCrUrUTT

annealed to its
compliment
. siCDK12 was a SMARTpool™ (Dharmacon).
The second siRNA targeting
FUS, siFUS2, was
rUrGrCrGrCrGrGrArCrA
rUrGrGrCrCrUrCrArATT
.


ChIP
s
equencing

Antibodies used for sequencing include: monoclonal anti
-
FUS (4H11:
sc
-
47711
Santa

Cruz),
monoclonal anti
-
RNAP2 (CTD4H8: Millipore), anti
-
ser2 phospho RNAP2

(ab24758; abcam: with
mouse anti
-
IgM


bridging antibody,

12
-
488; Millipore)
, anti
-
ser5 phospho RNAP2

(ab5131;
abcam)
, and anti
-
CDK9

(H
-
169 sc8338; Santa Cruz)
.


Between 10 and 5
0 million cells were transfected

in 15 cm
2
dishes and harvested 3 days
after transfection. A small
portion

of
the
cells
was

harvested t
o confirm

FUS knockdown

by
realtime PCR. Cells were crosslinked in 1% formaldehyde for 10 min, quenched in 150 mM
glycine and harvested by scraping. Nuclei were isolated by

washing twice on ice in

cold
hypotonic lysis

buffer

(10 mM Tris pH 7.5, 10 mM NaCl,

3 mM MgCl
2
, and 0.5% NP40)

and
centrifuging nuclei at 500g for 4 min
. Cells were lysed in lysis buffer
(1% SDS, 1
0 mM EDTA, 50
2


mM Tris
-
HCL pH 8.0
) with

1x Complete® protease inhibitors (Roche). Cells were sonicated for
10 to 15 min using a Bioruptor™
UCD
-
200 (Diagenode) with 30 sec pulses at maximum power.


Lysates were diluted to 10 ml in IP buffer (0.01% SDS, 1.1% Triton
-
X, 1.2 mM EDTA,
16.7 mM Tris
-
HCL pH 8.0, 167 mM NaCl), and 10 to 20

g of antibodies were added and
incubated overnight at 4 ºC wi
th rotation. Antibodies were immunoprecipitated with protein G
Plus/ protein A™ (Calbiochem) and washed 4 times in
wash buffers: low salt
(0.1% SDS, 1%
Triton X
-
100, 2 mM EDTA, 20 mM Tris
-
HCl pH 8.1, 150 mM NaCl), high salt (0.1% SDS, 1%
Triton X
-
100, 2 mM

EDTA, 20 mM Tris
-
HCl pH 8.1, 500 mM NaCl), LiCl wash (0.25M LiCl, 1%
NP40, 1% deoxycholate w/v, 1 mM EDTA, 10 mM Tris
-
HCl pH 8.1), and then Tris
-
EDTA pH 8.0.



DNA was eluted in 0.1 M NaHCO
3

and 1% SDS and crosslinks were reversed in 200
mM NaCl at 65 ºC
for 2 hours. Samples were Proteinase K and RNase A treated,
phenol:chloroform extracted and ethanol precipitated. Samples were resuspended and
concentration was measured using Qubit™ (Invitrogen).


Between 10 and 200 ng of recovered DNA was used to synthes
ize sequencing libraries
using the ChIP
-
seq Sample Preparation kit (Illumina). Concentration and quality of resulting
libraries were assessed by Qubit™ (Invitrogen), bioanalyzer, and realtime PCR. Between 6 and
10 pmoles were used for sequencing on either
an Illumina GAII or HiSeq2000 sequencer.


CLIP sequencing

Our CLIP protocol was modified from previously published protocols
(Hafner et al., 2010; Ule et
al., 2005)
. 80 million HEK293T cells were seeded on twenty 15 cm
2
dishes and grown to
confluence. Cel
ls were harvested by scraping, resuspended in 10 ml of cold PBS in a 15 cm
2

dish and UV crosslinked twice with 400 mJ/cm
2
.


Nuclei were isolated as for ChIP
-
seq experiments (described above), lysed, and treated
with DNase

I (Worthington Inc) for 20 min at 37 ºC, followed by 20 min treatment with 1 U of
RNase T1 at 37 ºC. After treatment with RNase T1, beads that had been pre
-
incubated with
3


anti
-
FUS antibody (4H11; Santa Cruz) for 1 hour were added and incubated with the ly
sate for 2
hours with rotation at 4 ºC. Beads were washed as described
(Ule et al., 2005)
. RNA bound to
proteins on the beads was radiolabeled with gamma
-
32
P
-
ATP and PNK as described and
without CIP treatment
(Hafner et al., 2010)
. Crosslinked protein
-
RNA
complexes were eluted by
adding NuPAGE™ loading buffer (Invitrogen) to beads and separated by SDS
-
PAGE
electrophoresis using 4
-
12% Bis
-
T
ris NuPAGE™ gels (Invitrogen). Complexes were transferred
to Hybond
-
ECL (Amersham) and exposed to a phosphor screen and
imaged after 1 to 8 hours
on a Typhoon Trio Imager (Amersham Biosciences). Bands were excised and RNA was
recovered as described
(Ule et al., 2005)
.


Recovered RNA was converted to double
-
stranded cDNA by first strand synthesis using
random primers and th
e Universal cDNA synthesis kit (ABI). Second strand synthesis was
performed in a 200
µ
l reaction by the addition of 2 U RNase H, 10 U DNA Ligase (NEB), and 40
U of DNA Polymerase I (NEB) in 1X reaction buffer (NEB) and incubating at 16 ºC for 2 hours.
The
reaction was finished by adding 10 U of T4 DNA polymerase for the last 5 min. Resulting
double
-
stranded cDNA was cleaned up using the Minelute™ PCR purification kit (Qiagen) and
then sequencing libraries were prepared using the ChIP
-
seq Sample Preparation
kit (Illumina).


RNA
-
seq

3 wells of 6
-
well dishes were treated with NEG siRNA and 3 treated with siFUS and cells were
harvested 3 days after transfection. Knockdown was confirmed by qPCR. Polyadenylated RNA
was purified from 5 ug of RNA and double stranded

cDNA libraries were prepared. Double
stranded cDNA was fragmented using DNaseI according to Illumina specifications. Adaptors
were ligated and sequencing libraries amplified. Paired end sequencing with 90 nt long reads
was performed to maximize coverage f
or analysis of changes in splicing.

4



Data analysis

Sequences were checked for quality and processed for alignment using the FastX
-
Toolkit
(
http://hannonlab.cshl.edu/fastx_toolkit/index.htm
l
)
.
Sequences acquired were aligned against
human genome hg18 and hg19 using the software Bowtie
(Langmead et al., 2009)
. Data were
subsequently manipulated using tools in software suites SAMtools
(Li et al., 2009)

and
BEDTools
(Quinlan and Hall, 2010)
.


Peak Calling:

We used the program FINDPEAKS to annotate peaks in our data using the default settings. We
then isolated peaks that intersected with the +/
-

300 nt window near the TSS, which we used for
our analysis. We used FINDPEAKS monte carlo simulation
to measure false discovery rates at
each peak height averaged across the genome. This relationship was fitted with a polynomial
curve and the FDR for each peak within the region of our analysis was calculated based on the
peak height.


Normalization:


All

data shown were multiplied by a normalization factor, N, using the total number of
aligned reads as described below. Although not shown in the manuscript, each CHIPseq and
RNAseq datasets treated with NEG control or siFUS were normalized independently usi
ng three
criteria to ensure trends and changes measured were insensitive to these procedural biases.
First, datasets were normalized by the total number of reads aligned in each experiment, which
was the only criteria used in data for figures in this manus
cript: N = 1/Number of Aligned Reads
in Experiment. Second, datasets were compared using number of reads normalized to the
average or median number of reads for each dataset: N = 1 / Median Reads per Kb in
Experiment. Third, datasets were compared using nu
mber of reads normalized to the number of
5


reads mapping to control genes
GAPDH, ACTB, PPARG, SDHA
, and
TBP
:
N = 1 / Sum of
Reads Mapped to Control Genes in Experiment. Without exception, comparison of datasets
using each of the three normalization criteria

yielded identical trends and differences.


Binning:


Data used to generate plots of individual genes were extracted from Bedgraph files
produced by the genomecoverageBed program within BEDTools. Coordinates for REFSEQ
genes were downloaded

from the UCSC genome browser. For metagene analysis, sequences
were binned and summed relative to coordinates for REFSEQ genes using the intersectBed
program within BEDTools. Binning summed the total number of reads for each 100 nt window
between 4000 nt
upstream of the TSS and 500 nt downstream, then 500 nt upstream of the 3’
end of the gene to 1500 nt downstream. Within the gene, the density of reads per 100 nt was
measured for each 10
th

of the gene to correct for the range of gene lengths in the human
g
enome. For these plots, no noise subtraction was used. Where indicated in the manuscript,
when total read counts differed between experiments, each bin was multiplied by a
normalization factor as described above.



RNA
-
seq analysis:


RNA
-
seq data
were mapp
ed to the UCSC hg19 genomic reference/annotation using
Tophat 2.0.4

and then processed
in two ways. First, a
lternative isoform analysis was performed
with MISO

0.4.6

using a ΔΨ

cutoff of either >0.20 or >0.10 and a minimum Bayes factor of 10

and with the hg19 alternative event annotation
found at

http://genes.mit.edu/burgelab/miso/docs/

. Secondly
, differential isoform expression using
Cufflinks v2.0.2. The Volcano plot and isoform bar c
harts were generated using CummeRbund
v1.2.



6


Ontology

Genes with FUS bound were selected for those whose mRNA levels were 1.2 higher or 0.8 fold
lower with a p
-
value less than 0.05. REFSEQ numbers were submitted to FuncAssociate 2.0
from the Roth Lab
(Ber
riz et al., 2009)
.


Realtime PCR (qPCR)

RNA was recovered using Trizol™ (Invitrogen) according to the manufacturer’s instructions.
cDNA libraries were synthesized using the High Capacity cDNA Reverse Transcription kit
(Applied Biosystems, Inc). qPCR

was performed using either iSybr (Biorad) or Lightcycler® 480
Probes Master (Roche) on a Roche Lightcycler® 480. Primers targeting mRNA were designed
to span exon boundaries and 2 to 3 primers were screened for efficient amplification. Primers
targeting g
enes promoters were designed with up to 14 primers screened for efficient
amplification.


Protein purification

His
-
tagged FUS was expressed from a bacmid (gift from the T. Tuschl lab, Rockefeller U.) in
Sf9 insect cells at the
Tissue Culture Core Facility

(UC Denver Cancer Center) facility in
Denver
. 0.5 to 2 L of cells were pelleted and stored at
-
80 ºC. Pellets were thawed and
resuspended in lysis buffer (40 mM Tris
-
HCL pH 8.0, 5 mM MgCl
2
, 1 M KCl, 1 mM imidazole, 5
mM

-
mercaptoethanol, and 0.05% NP
-
40)

and sonicated for 1.5 min on ice (Misonix Sonicator
3000). Lysate was cleared by centrifugation at 13000 rpm for 15 min. Supernatant was
incubated with 1 to 4 ml of washed High Capacity Ni
-
Sepharose beads (Qiagen) for at least 20
min. Beads were washed in

batch 3 times in 5x volume of lysis buffer without NP
-
40. Beads
were then washed over a column in 3x volume of lysis buffer without NP
-
40 and with 25 mM
imidazole. Finally protein was eluted from beads with lysis buffer without NP
-
40 and with 250
7


mM imida
zole. Protein was cleared by centrifugation at 13000 rpm for 20 min and then injected
onto a HiLoad Superdex™ 200 16/60 column. Protein was run in 50 mM Tris
-
HCl pH 8.1, 5 mM
MgCl
2
, 100 mM KCl and eluted between 65 and 75 ml consistent with a monomer of mo
lecular
weight near 70 kDa.


GST
-
FUS constructs were a gift from the M. G. Rosenfeld lab (UC San Diego) and were
transformed into BL
-
21 Gold cells (Agilent). 1 to 6 L were grown, induced at OD600 0.8 and
grown for 3 hours at 30 ºC. GST
-
FUS was purified e
ssentially as described above but using
glutathione
-
sepharose beads and eluted with lysis buffer without NP
-
40 and with 30 mM
glutathione. Full length GST
-
CTD, P
-
TEFb, and TFIIH were purified as described
(Knuesel et
al., 2009; Tahirov et al., 2010)
.


Pull
-
down assays

FUS (0.3

g), GST
-
CTD (0.1

g), and P
-
TEFb (0.08

g) were incubated in 50

l pull down
buffer (20 mM Tris
-
HCL pH 7.0, 150 mM KCl, 2 mM MgCl
2
, 2 mM DTT, 0.15 mg/mL BSA) for 30
min at room temperature. 50

l of washed and equilibrated glutathione sepharose bead solution
was added to each sample and incubated with gentle shaking for 30 min at 4 ºC. Beads were
washed in twice with 400

l pull down buffer without BSA, twice with 400

l high salt pull down
buffe
r (20 mM Tris
-
HCl pH 7.0, 500 mM KCl, 2 mM MgCl
2
, 2 mM DTT), once with 400

l
pulldown buffer containing 0.1% NP40, and once more with 400

l pulldown buffer. Proteins
were eluted with 40

l elution buffer (20 mM Tris
-
HCl pH 7.0, 150 mM KCl, 2 mM MgCl
2
, 2
mM
DTT, 30 mM glutathione) at 4 ºC with gentle shaking for 30 min. Eluted proteins were resolved
by western on a 4
-
20% NuPAGE bisTris gel (Invitrogen). For testing effects of DNA on FUS
recognition of RNAP2
-
CTD, these samples were treated with RNase A (Amb
ion).

8



Purification of CDK12


HeLa nuclear lys
ates were prepared as described (Knuesel et al., 2009). 8

g of
antibody against CDK12 provided by the Arno Greenleaf lab was

prebound to
protein G Plus/
protein A™ (Calbiochem)
beads and

incubated with
1 mL of

lysates under two separate buffer
conditions. Condition 1, the antibody was added to the lower salt buffer of the lysate (
80 mM
Tris, 0.1 mM EDTA, 10% Glycerol, 0.02% NP
-
40, 150 mM KCl
, 2 mM DTT and protease
inhibitors
). Condition 2, the antibody was
diluted into a higher salt buffer (500mM NaCl, 0.1%
TritonX, 25mM KCl, 5% glycerol, 20mM HEPES pH 7.6 with protease inhibitors and
2 mM
DTT)
.
Beads were

allow to incubate with
lysates for 2 hours. Beads were washed 4 times in high salt
(500 mM KCl) and twi
ce in low salt (150 mM KCl) buffers. Beads were then used immediately or
stored at
-
80 in low salt buffer with 30% glycerol. Either condition resulted in similar levels of
activity, but the condition 1 yielded CDK12 with additional bands detected by wester
n blot with
the same CDK12 antibody. CDK12 from condition 2 was used in figures in this paper. Beads
were washed and resuspended in kinase buffer before use.


Kinase assays

Active P
-
TEFb was purchased from Millipore or purified by coexpression in Sf9 insec
t cells as
previously described
(Tahirov et al., 2010)
. 0.1

g GST
-
CTD or CTD cleaved away from the
GST tag by
thrombin

protease was preincubated with or without FUS added for 20 min at room
temperature in kinase buffer (20 mM Tris
-
HCl pH 7.0, 150 mM KCl,
2 mM MgCl
2
, 2 mM DTT,
0.15 mg/mL BSA, 0.1

M ATP,

-
32
P
-
ATP). After pre
-
incubation, 40 ng of P
-
TEFb complex was
added and incubated for 20 min at room temperature. Reaction was stopped by addition of
9


NuPAGE loading buffer (Invitrogen) and heating at 95 ºC
for 4 min. Radiolabeled GST
-
CTD was
resolved on a 4
-
12% NuPAGE bisTris gel (Invitrogen). Gels were dried, exposed to a phosphor
screen, and imaged on a Typhoon Trio imager. TFIIH kinase assays were performed in the
same fashion as P
-
TEFb using 4

L of the
active fraction of TFIIH purified as described
(Knuesel et al., 2009)
.


Purification of CTD
-
interacting cofactors and RNA Polymerase II

CTD
-
interacting cofactors were purified with GST
-
CTD (52 repeats) immobilized to
Glutathione
-
Sepharose

beads (GE Life Sciences). After binding, the resin was washed five
times with 20 column volumes (CV)
of
0.5 M KCl HEGN (20 mM Hepes, pH 7.6; 0.1 mM EDTA;
10% Glycerol; 0.1% NP
-
40 alternative) and one time with 20 CV
of
0.15 M KCl HEGN (0.02%
NP
-
40 alterna
tive). Bound proteins were eluted for mass spectrometry with 2% (w/v) Sarcosyl in
Tris elution buffer (80 mM Tris, 0.1 mM EDTA, 10% Glycerol, 0.02% NP
-
40, 150 mM KCl).

RNA Polymerase II complexes were immunoprecipitated from HeLa nuclear extract
using 8WG1
6 antibody immobilized to Protein G Sepharose (Amersham Biosciences).
Immobilized antibody was incubated with HeLa nuclear extract at 4 °C. The antibody resin was
then washed three times with 20 CV
of
0.5 M KCl HEGN and twice with 20 CV
of
0.15 M KCl
HEGN.

8WG16
-
bound proteins were eluted with 0.1M Glycine, pH 2.7 for mass spectrometry
analyses. Silver stain and immunoblot analyses were peptide eluted two times with 2mg/mL
elution peptide (YSPTSPSYSPTSPSYSPTSPSYSPTSPS).

P
-
TEFb

complexes were
immunoprecipita
ted nearly identically but using

anti
-
Cyclin T1

antibody

(Santa Cruz T
-
18)

and
eluted with 2% (w/v) Sarcosyl in Tris elution buffer.

To test RNA
-
dependence for FUS binding RNAP2, (1) no nuclease, (2) 0.25U/µL
Benzonase (Sigma ) or (3) 20µg/mL RNAse A were
added to HeLa nuclear extract prior to
10


8WG16 immunoprecipitation. RNAP2 complexes were purified and eluted with peptide for silver
stain and immunoblotting as above.



Mass spectrometry

The purified RNAP2 complex and CTD
-
interacting cofactors were precipit
ated at 4ºC using 20%
(v/v) TCA, 0.067mg/mL insulin and 0.067% (w/v) deoxycholate. Precipitated samples were
washed twice with
-
20ºC Acetone and air dried. Purified protein samples were digested into
peptides using a modified Filter
-
Aided Sample Prep (FASP
) protocol
(Wisniewski et al., 2009)
.
Briefly, protein pellets were suspended with 4% (v/v) SDS, 0.1M Tris pH 8.5, 10mM Tris(2
-
Carboxyethyl)
-
Phosphine HCl (TCEP) and incubated 30 minutes ambient for reduction.
Reduced samples were then diluted with 8M Urea
, 0.1M Tris pH 8.5 and iodoacetamide was
added to 10 mM and incubated 30 minutes in total darkness. Reduced and alkylated samples
were then transferred to a Microcon YM
-
30 spin concentrator and washed three times with 8M
Urea, 0.1M Tris pH 8.5. Following t
hree washes with 2M Urea, 0.1M Tris pH 8.5, samples were
digested with Lys
-
C rocking at ambient overnight, followed by trypsin and 2 mM CaCl
2

at 37
°
C
for another six hours. Digested peptides were eluted and acidified with 5% (v/v) formic acid and
desalted
online followed by fractionation with a Phenomenex Jupiter 5µm C18 (300
Å;

0.25 x
~150mm) column using a two dimensional LC/MS/MS method (Agilent 1100). Seven steps of
increasing acetonitrile (3, 6, 9, 12, 16, 20 and 100% B; A: 20 mM ammonium formate pH 10
, 4%
(v/v) acetonitrile; and B: 10 mM ammonium formate pH 10, 65% (v/v) acetonitrile) at 5 µL/minute
eluted peptides for a second dimension analysis on a Phenomenex Jupiter 4 µm Proteo (90
Å
;
0.075 x ~250 mm) running a gradient at 0.2µL/minute from 5 to 25%

B in 100 minutes for steps
one through six and 10 to 30% B in 100 minutes for step seven (A: 4% acetonitrile and B: 80%
acetonitrile, both with 0.1% formic acid pH ~2.5). Eluted peptides were detected with an Agilent
11


MSD Trap XCT (3D ion trap) mass spectr
ometer. All spectra were searched with Mascot v2.2
(Matrix Sciences) against the International Protein Index (IPI) database version 3.65 with two
missed cleavages and mass tolerances of m/z ±2.0 Da for parent masses and ±0.8 Da for
MS/MS fragment masses. P
eptides were accepted above a Mascot ion score corresponding to
a 1% false discovery rate (1% FDR) determined by a separate search of a reversed IPI v3.65
database. Peptides were then filtered and protein identifications were assembled using in
-
house
softw
are as described
(Old et al., 2005; Resing et al., 2004; Meyer
-
Arendt et al., 2011)
.


Berriz, G.F., Beaver, J.E., Cenik, C., Tasan, M., and Roth, F.P. 2009. Next generation software
for functional trend analysis.
Bioinformatics

25
:

3043
-
3044.


Hafner, M.,
Landthaler, M., Burger, L., Khorshid, M., Hausser, J., Berninger, P., Rothballer, A.,
Ascano, M., Jr., Jungkamp, A.C., Munschauer, M.
, et al.

2010
. Transcriptome
-
wide
identification of RNA
-
binding protein and microRNA target sites by PAR
-
CLIP.
Cell

141
:

12
9
-
141.


Knuesel, M.T., Meyer, K.D., B
ernecky, C., and Taatjes, D.J. 2009
. The human CDK8
subcomplex is a molecular switch that controls Mediator coactivator function.
Genes Dev

23
:

439
-
451.


Langmead, B., Trapnell, C
., Pop, M., and Salzberg, S.L. 2009
. Ultrafast and memory
-
efficient
alignment of short DNA sequences to the human genome.
Genome Biol

10
:

R25.


Li, H., Handsaker, B., Wysoker, A., Fennell, T., Ruan, J., Homer, N., Marth, G.
, Abecasis, G., and
Durbin, R. 2009
. The Sequence Alignment/Map form
at and SAMtools.
Bioinformatics

25
:

2078
-
2079.


Meyer
-
Arendt K, Old WM, Houel S, Renganathan K, Eichelberger B, Resing KA, Ahn NG.
2011.
IsoformResolver: A peptide
-
centric algorithm for protein inference.
Journal of Proteome
Research.
10
:
3060
-
3075.


Old WM, Meyer
-
Arendt K, Aveline
-
Wolf L, Pierce KG, Mendoza A, Sevinsky JR, Resing KA, Ahn
NG.
2005. Comparison of label
-
free methods for quantifying human proteins by shotgun
proteomics.
Molecular and Cellular Proteomics.
4
:1487

1502


Resing KA, Meyer
-
Aren
dt K, Mendoza AM, Aveline
-
Wolf LD, Jonscher KR, Pierce KG, Old WM,
Cheung HT, Russell S, Wattawa JL, Goehle GR, Knight RD, Ahn NG..
2004. Improving
12


reproducibility and sensitivity in identifying human proteins by shotgun proteomics.
Analytical
Chemistry
76
:
3556

3568


Quinlan, A.R., and Hall, I.M. 2010
. BEDTools: a flexible suite of utilities for comparing
genomic features.
Bioinformatics

26
:

841
-
842.


Tahirov, T.H., Babayeva, N.D., Varzavand, K., Cooper, J.J.,

Sedore, S.C., and Price, D.H. 2010
.
Crystal st
ructure of HIV
-
1 Tat complexed with human P
-
TEFb. Nature

465
:

747
-
751.

Ule, J., Jensen, K
., Mele, A., and Darnell, R.B. 2005
. CLIP: a method for identifying protein
-
RNA interaction sites in living cells.
Methods

37
:

376
-
386.


i niewski R, ougman A,
Nagaraj N, Mann M.
2009
.
Universal sample preparation method
for proteome analysis.
Nature Methods
.
6
:
359
-
362.



Supplemental Table
2
:

List of alternative splice isoform changes called by MISO 0.4.6 using
the
ΔΨ

cutoff of either >0.20

(
right
)

or >0.10

(
left
).


Supplemental Table 3:
List of alternative splicing events called by Cufflinks v2.0.2.


Supplemental Table 4:

List of significant changes in gene expression called by CuffDiff v2.0.2.