Tiling array design and RNA Hybridization
A custom tiling array (Roche Nimblegen) was designed at 5 basepair resolution across 25kb of the 9p21 region (which encompasses CDKN2a
, and CDKN2b
), as well as from 10kb upstream to 2kb downstream of each TSS from 53 other cell cycle genes including cyclins, CDKs, and CDKIs (Table S1
). In addition, the HOXA
loci were placed on the array as a control. Briefly, RNA was amplified (MessageAmp Kit, Ambion), reverse transcribed (Retroscript Kit, Ambion), labeled, and hybridized according to the standard Nimblegen protocol.
Robust multichip average (RMA) normalized single channel data from each array was subjected to peak calling using the Nimblescan program (Roche Nimblegen) with a window size = 50. Peaks with a peak score > 10 were considered significant transcriptional units. Peak calls from all 55 array samples were clustered using Galaxy 2,48, and only transcripts present in a minimum of 10% of the samples were considered for further analysis. Transcripts were annotated as following – “genomic location (upstream of TSS of cell cycle protein-coding gene = upst; exon of cell cycle protein-coding gene = exon; intron of cell cycle protein-coding gene = int; downstream of cell cycle protein coding gene = dst)” : “gene symbol of nearest mRNA” : “distance from TSS”.
Measuring Protein Coding Potential
To assess the coding potential of the novel transcribed regions, we evaluated the evolutionary signatures in their alignments with orthologous regions in 20 other sequenced placental mammalian genomes using the Codon Substitution Frequencies (CSF) method 30,49,50
, which has also been applied to assess novel transcribed regions in mouse 14
. CSF produces a score for any region in the genome considering all codon substitutions observed within its alignment, based on the relative frequency of similar substitutions in known coding and non-coding regions. Briefly, it performs a statistical comparison between two empirical codon models 51
, one estimated from alignments of known coding regions and the other based on non-coding regions, and reports a likelihood ratio that quantifies whether the protein-coding model is a better explanation, while controlling for the overall level of sequence conservation30
Module Map analysis
We generated a module map of the ncRNAs versus the protein-coding genes by computing the Pearson correlations for all pairwise combinations based on expression across 17 different samples. This map was clustered and visualized using the program Genomica (http://genomica.weizmann.ac.il/
). For each ncRNA, we then defined gene sets of the protein coding genes that had a Pearson correlation that was greater than or less than 0.5 with that ncRNA. To determine functional associations, we then generated a module map of these ncRNA gene sets with Gene Ontology Biological Processes gene sets () and with curated gene sets of metabolic and signaling pathways and biological and clinical states from the Molecular Signatures Database (MSigDB c2 collection) (Fig S4
-value of enrichment was determined by the hypergeometric distribution, and a false discovery rate (FDR) calculation was used to account for multiple hypothesis testing (p<0.05, FDR<0.05).
Tissue samples and cells
Informed consent was obtained for tissue donation as well as approval from institutional review boards. Human primary breast tumors from the Netherlands Cancer Institute 52
, and normal breast tissues and metastatic breast tumors from the Johns Hopkins University Rapid Autopsy Program (Gupta et al., 2010) are as described. Human fetal pancreata were obtained from the Birth Defects Research Laboratory, University of Washington (Seattle, WA). Staged fetal pancreata were processed within 24 hours of receipt, minced, washed and processed for RNA isolation using standard methods. Human fetal lung fibroblasts FL3 (Coriell AG04393) or foreskin fibroblasts (ATCC CRL2091) were cultured in 10% FBS (Hyclone), 1% penicillin-streptomycin (Gibco) at 37C in 5% CO2.
PANDA Cloning and sequence analysis
3′ and 5′ RACE was performed using the FirstChoice RLM-RACE Kit (Ambion). RNA was extracted from 200ng/ml doxorubicin (Sigma) treated human fetal lung fibroblasts , polyA selected using the PolyA Purist-MAg kit (Ambion), and RLM-RACE was performed according to the standard manufacturer’s protocol.
Total RNA was extracted from cells using the Trizol reagent (Invitrogen) and the RNeasy Mini Kit (Qiagen) and genomic DNA was eliminated using Turbo DNA-free (Ambion). RT-PCR using 50-250 ng of total RNA was performed using the One-Step RT-PCR Master Mix (Applied Biosystems) using Taqman Gene Expression Assays and normalized to GAPDH. Strand specific RT-PCR for PANDA was performed using the One Step RT-PCR Master Mix SYBR Green (Stratagene)).
TaqMan® custom ncRNA Assays
A panel of TaqMan® custom ncRNA assays was developed targeting 60 of the 219 novel transcribed regions using “Single-exon” design mode. The transcript specificity and genome specificity of all TaqMan assays were verified using a position specific alignment matrix to predict potential cross-reactivity between designed assays and genome-wide non-target transcripts or genomic sequences. For gene expression profiling of these ncRNAs across different conditions, cDNAs were generated from 50ng of total RNA using the High Capacity cDNA Reverse Transcription Kit (Life Technologies, Foster City, CA). The resulting cDNA was subjected to a 14-cycle PCR amplification followed by real-time PCR reaction using the manufacturer’s TaqMan® PreAmp Master Mix Kit Protocol (Life Technologies, Foster City, CA). Two replicates were run for each gene for each sample in a 384-well format plate on 7900HT Fast Real-Time PCR System (Life Technologies, Foster City, CA). PPIA was used as an endogenous control for normalization across different samples.
5 ug of polyA RNA was obtained using RNeasy Kit (Qiagen) and PolyA Purist Mag (Ambion). Northern Blot was performed using NorthernMax Kit (Ambion) following the standard manufacturer’s protocol. Probes were generated with full length PANDA using the Prime-It RmT Random Primer Labeling Kit (Agilent).
The following antibodies were used for Chromatin Immunoprecipitation Assays: anti-H3K4me3 (Abcam ab8580), anti-H3K35me3 (Abcam ab9050), anti-p53 (Abcam ab28). Western blots were performed using anti-PARP (Cell Signal 9542),anti-B-tubulin (Abcam ab6046), anti LSD1 (ab17721), anti EZH2 (Cell Signal AC22), anti p21 (Santa Cruz Biotech), anti NF-YA (Santa Cruz Biotech H-209).
Human fetal lung fibroblasts were transfected with 50 nM of onTargetPLUS siRNAs (Dharmacon) targeting PANDA
(Supplementary Table 5
). Validated siRNAs for mRNAs were obtained from Ambion (Supplementary Table 5
TUNEL assays were performed using the In Situ Cell Death Detection Kit, TMR Red (Roche). Human fetal lung fibroblasts were cultured on chamber slides (Lab-Tek), treated with 200ng/ml doxorubicin (Sigma) for 24 hours, fixed with methanol at -20C for 10 minutes, and incubated with the TUNEL labeling mixture for one hour at 37C. Slides were then washed with PBS and mounted in Prolong® Gold antifade reagent with DAPI (Invitrogen) and imaged at 20x magnification.
10 million cells were treated with 200ng/ml doxorubicin for 16 hours, trypsinized, and crosslinked with 1% formaldehyde for 10 minutes followed by the addition of .125 M Glycine for 5 minutes. After 2 PBS washes, cells were lysed with 2x volume of Buffer A (10mM HEPES pH 7.5, 1.5 mM MgCl2, 10 mM KCl, .5 mM DTT, 1mM PMSF) for 15 minutes on ice at 150 RPM. NP-40 was added to a final concentration of .25% for 5 minutes on ice. Lysates were centrifuged for 3 minutes at 2000 RPM, and the supernatant (cytosol) was collected. Next, an equal volume of Buffer C as Buffer A was added to the pellet for 20 minutes with frequent vortex (20 mM HEPES pH 7.5, 10% Glycerol, .42M KCl, 4 mM MgCl2, .5 mM DTT, 1 mM PMSF). Nuclear lysates were dounced for 5 seconds using a motorized pestle and sonicated for 7 minutes using a Diagenode Sonicator (30 seconds on, 30 seconds off, power setting H). Nuclear and cytoplasmic lysates were combined and centrifuged for 15 minutes at 13K RPM. Supernatants were transferred into Micro spin columns (Pierce 89879) and 2 ug of antibody was added and incubated overnight. 10 ul of Protein A/G Ultralink Resin (Pierce 53132) was washed 3x with RIP wash buffer (50 mM TrisHcl pH 7.9, 10% glycerol, 100mM KCl, 5mM MgCl2, 10 mM B-me, and .1% NP-40) and added to the Immunoprecipitation reaction for 1 hr at 4C. Samples were washed 4x with RIP wash buffer and 2x with 1M RIPA (50 mM Tris pH 7.4, 1M NaCl, 1 mM EDTA, .1% SDS, 1% NP-40, .5% sodium deoxycholate, .5mM DTT and 1 mM PMSF). Beads were resuspended in 200 ul 150mM RIPA (50 mM Tris pH 7.4, 150 mM NaCl, 1 mM EDTA, .1% SDS, 1% NP-40, .5% sodium deoxycholate, .5mM DTT and 1 mM PMSF) + 5 ul Proteinase K (Ambion) and incubated for 1 hr at 45C. 1 ml of Trizol was added to the sample and RNA was extracted using the RNEasy Mini Kit (Qiagen) with the on column DNAse digest (Qiagen).
RNAse mediated RNA chromatography
RNAse mediated RNA chromatography 41
was performed as previously described with the following modifications: 6 pmols of RNA (PANDA
or a 1.2 KB fragment of LacZ) were used per reaction. RNA was folded (90C for 2 minutes, ice for 2 minutes, supplied with RNA structure buffer (Ambion), and shifted to room temperature for 20 minutes prior to conjugation to beads. RNAse digestion was performed with 5 ul of Rnase A/T1 cocktail (Ambion) and 2 ul of Rnase V1 (Ambion).
Cellular lysates were prepared as follows: 10 million doxorubicin treated cells (16 hrs) were incubated in 200 ul PBS, 600 ul H20, and 200 ul nuclear lysis buffer (1.28 M sucrose; 40 mM Tris-HCl pH 7.5; 20 mM MgCl2; 4% Triton X-100) on ice for 20 min. Nuclei were pelleted by centrifugation at 2,500 G for 15 min. Nuclear pellet was resuspended in 1 ml RIP buffer (150 mM KCl, 25 mM Tris pH 7.4, 0.5 mM DTT, 0.5% NP40, 1 mM PMSF and protease Inhibitor (Roche Complete Protease Inhibitor Cocktail Tablets)). Resuspended nuclei were sheared using a motorized douncer for 5 seconds. Nuclear membrane and debris were pelleted by centrifugation at 13,000 RPM for 10 min.
was performed as previously described53
(ChIP) was performed as previously described54
. Q-PCR primers for FAS and CCNB1 and FAS-control NF-YA binding sites were obtained from Morachis et al.40
Primers for PUMA and BAX were designed to surround the NF-YA consensus motif CCAAT (Supplementary Table 5