SUPPLEMENTAL INFORMATION The supplemental information includes the following material: Supplemental note: related to supplemental Figure S1 Seven supplemental figures: Supplemental Figure S1: Related to Materials and Methods Supplemental Figure S2: Related to Figure 1 Supplemental Figure S3: Related to Figure 2 Supplemental Figure S4: Related to Figure 3 and Figure 4 Supplemental Figure S5: Related to Figure 4 and Figure 5 Supplemental Figure S6: Related to Figure 5 Supplemental Figure S7: Related to Figure 6 Six supplemental tables and one supplemental data file: Supplemental Table S1: Related to Figure 1 Supplemental Table S2: Related to Figure 2 Supplemental Table S3: Related to Figure 4 Supplemental Table S4: Related to Figure 4 Supplemental Table S5: Related to Figure 4 Supplemental Table S6: Related to Figure 5 1
37
Embed
SUPPLEMENTAL INFORMATIONgenesdev.cshlp.org/.../28.22.2498.DC2/Supp_Information.docx · Web viewAnalysis of Transcripts in Total, polyA+ and polyA- RNA Preparations We subjected RNA
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
SUPPLEMENTAL INFORMATION
The supplemental information includes the following material:
Supplemental note: related to supplemental Figure S1
Seven supplemental figures:
Supplemental Figure S1: Related to Materials and Methods
Supplemental Figure S2: Related to Figure 1
Supplemental Figure S3: Related to Figure 2
Supplemental Figure S4: Related to Figure 3 and Figure 4
Supplemental Figure S5: Related to Figure 4 and Figure 5
Supplemental Figure S6: Related to Figure 5
Supplemental Figure S7: Related to Figure 6
Six supplemental tables and one supplemental data file:
Supplemental Table S1: Related to Figure 1
Supplemental Table S2: Related to Figure 2
Supplemental Table S3: Related to Figure 4
Supplemental Table S4: Related to Figure 4
Supplemental Table S5: Related to Figure 4
Supplemental Table S6: Related to Figure 5
Supplemental Data File: Cufflinks-assembled transcriptome based on RNA-seq
data from control, XRN1-, SMG6/XRN1- and UPF1/XRN1-depleted samples.
Additional supplemental material:
Supplemental Materials and Methods
Supplemental References
1
SUPPLEMENTAL NOTE
Analysis of Transcripts in Total, polyA+ and polyA- RNA Preparations
We subjected RNA purified from control-, XRN1-, SMG6/XRN1- and UPF1/XRN1-
depleted HEK293--39 cells to oligo-dT selection and assessed its efficiency on a
set of transcripts. Spike-in RNAs with polyA-tails of 21-25nt in length and with
mono-phosphates at their 5’-ends were added in equal amounts to total RNA
samples before the selection procedure. We observed an efficient depletion of
~8- to ~18-fold from the polyA- fraction of the endogenous GAPDH mRNA and
the -39 nonsense reporter RNA (Supplemental Fig. S1D, first and second plot
from left in the top panel). The spike-in RNA molecules, represented here by T7-
Luc-pA, were similarly depleted (~10-fold) from the poly(A)- fraction
(Supplemental Fig. S1D, second plot from left in the bottom panel). In
comparison, the non-polyadenylated MALAT1 transcript was approximately 2-
fold depleted from the poly(A)- fraction in all samples (Supplemental Fig. S1D,
first plot from left in the bottom panel). Levels of -39 RNA were almost identical
in the total and polyA+ fractions when measured relative to GAPDH - which was
also seen for the spike-in RNAs (Supplemental Fig. S1D, second plot from the
right in the top and bottom panel). Finally, we measured the relative level of
RNA that had been ligated at the 5’-end. T7-Luc-pA served as a control, and the
relative levels of ligated molecules were virtually the same across all twelve
samples (Supplemental Fig. S1D, first plot from right in the bottom panel).
Importantly, the relative levels of ligated -39 molecules appeared very similar
between the polyA+ and the total RNA sample (and even in the polyA- fraction;
Supplemental Fig. S1D, first plot from right in the upper panel). Based on these
2
observations, we concluded that the polyA+ fraction represents well the total
sample when it comes to measuring decapping levels of nonsense RNAs.
Based on the northern blot in Supplemental Fig. S1B and previous observations
(Eberle et al. 2009) it is also clear that the 3’-fragment produced after SMG6-
catalyzed endocleavage can be enriched in the polyA+ fraction.
3
SUPPLEMENTAL FIGURES
4
Supplemental Figure S1. Overview and quality control of the experimental
procedure, Related to Methods
(A) Schematic overview of the principle behind the massive parallel sequencing
approach used in this study (see also Fig. 1A). (B) Northern blotting analyses
monitoring the various steps of the digitonin extraction procedure of HEK293--
39 cells depleted for the indicated factors; (lanes 1-4) whole-cell total RNA,
(lanes 5-8) total RNA from the pellet (enriched for nuclei and other membrane-
bound organelles), (lanes 9-12) total RNA from the digitonin extract (enriched
with soluble cytoplasm), (lanes 13-16) polyA+ RNA from the digitonin extract
and (lanes 17-20) polyA- RNA from the digitonin extract. Northern membranes
were hybridized with probes directed against the 3’-region of the -39 reporter
RNA, 16S mitchondrial rRNA, U3 snRNA and tRNALys (the latter three to verify the
enrichment of mitochondria/nuclei and cytoplasm in the pellet and the digitonin
extract, respectively). 28S and 18S rRNA levels were detected by methylene blue-
staining of the membrane. (C) Western blotting analyses of cell extracts verifying
depletion of the indicated factors. The upper panel corresponds to the samples
used for massive parallel sequencing. The lower panel is representative of all
other presented depletion experiments. Western membranes were probed with
antibodies recognising XRN1, SMG6 and UPF1. Gel loading was controlled by
probing membranes with an anti-U170K antibody (upper panel) or a -actin
antibody (lower panel). (D) qPCR analyses of the indicated transcripts
performed on reverse-transcribed total, polyA+ or polyA- RNA isolated from the
HEK293--39 cell line either control-depleted (red) or depleted for XRN1 (blue),
SMG6/XRN1 (green) or UPF1/XRN1 (orange). Relative ligation (right panel) is
normalized to the XRN1 sample within each fraction.
5
6
Supplemental Figure S2. PTC-Proximal Endocleavages of Endogenous
Nonsense RNAs, Related to Figure 1
(A-E) Overview of the sequencing data used in the analyses of the endogenous
genes MATR3 (A-B), NOP56 (C-D) and ATF4 (E) as examples of detected NMD-
specific endocleavages (A, C and E are as Fig. 1C and B and D are as Fig. 1F). (F)
Northern blotting analyses of total RNA isolated from the HEK293--39 cell line
depleted for the indicated factors. The northern membranes were hybridized
with probes directed against regions downstream of the endocleavage sites in
the MATR3 (upper panel), ATF4 (middle panel) and NOP56 (lower panel) RNAs,
respectively. GAPDH levels were detected as an internal loading standard. (G)
Northern blotting analyses of cycloheximide pulse/actinomycin D-chase
experiments performed in the context of control- or UPF1-depletion on -39,
SNHG15, NOP56, ATF4, GADD45A, GAS5 and EIF5 nonsense RNAs as indicated.
GAPDH levels were detected as an internal loading standard. The estimated half-
lives are indicated in red below the name of the RNA species (ctrl to left and
UPF1 to the right). Similar results were obtained from actinomycin D chase
experiments (data not shown). (H) Positions of NMD-specific endocleavages
mapped to the ‘NMD reference set’ relative to the annotated stop codons.
7
8
Supplemental Figure S3. NMD-Specific Decapping Sites in Endogenous
Nonsense RNAs, Related to Figure 2
(A-G) Overview of the sequencing data used in the analyses of the genes (A-B)
TRA2B, (D-E) RSRC2 and (G) SNHG15 as representative examples of endogenous
genes expressing RNA undergoing NMD-specific decapping (A, D and G are as Fig.
1C and B and E are as Fig. 1F). (C,F,H) Relative decapping (left panel, as Fig. 2B)
and levels of RNA co-immunoprecipitated with endogenous UPF1 (right panel, as
Fig. 2F) of (B) TRA2B nonsense RNA (upper panel) and mRNA (lower panel), (F)
RSRC2 nonsense RNA (upper panel) and mRNA (lower panel) and (H) SNHG15
nonsense RNA. Decapping levels are relative to levels measured upon XRN1-
depletion. All co-IP values are relative to co-IP’ed levels of -39 under control
conditions. Relative decapping and immunoprecipitation measurements are
from four (n=4) and two (n=2) independent experiments, respectively. Error-
bars depict standard deviations.
9
Supplemental Figure S4. Pipelines For Identification of NMD-Targeted
Transcripts/Genes Based on 5’end-seq and RNA-seq, Related to Figure 3
and Figure 4
(A-B) See text in the figure and main manuscript for details..
10
11
Supplemental Figure S5. Global Overview of NMD-Specific Endocleavage
and Decapping Events in NMD-sensitive Transcript Isoforms, Related to
Figure 4 and Figure 5
(A-C) The same analysis as in Fig. 3A-B and 4A, but focusing on transcripts rather
than genes. (A-B) See legend for Fig. 3A-B. (C) See legend for Fig. 4A. (D) Same as
Fig. 4B., producing an NMD-substrate set based on the combined analysis of
RNA-seq and 5’end-seq, but using ‘relaxed’ criteria for peak calling in the 5’-end-
seq data. This NMD set is listed in Supplemental Table S4. (E) As Fig. 5A, but
based on ‘relaxed’ criteria for identification of NMD-responsive genes. ***
indicates that snoRNA host genes are significantly enriched compared to protein-
coding genes (Fisher two-sided test, P = 8.9·10-12). (F) Density plot (left panel)
illustrating the significant differences between snoRNA host genes (n=173) and
highly expressed protein-coding genes (n=3000; expression distributions are
shown in the right panel), in terms of the fraction of nonsense isoforms out of the
total number of produced isoforms (defined by RNA-seq). Relative densities from
the given groups are calculated by kernel density estimation (KDE) with
Gaussian kernel. (G) Venn diagrams displaying the number of SRSF-protein
coding genes that produce one or more nonsense isoforms identified
independently by RNA-seq and 5’-end-seq by ‘stringent’ (left) and ‘relaxed’
(right) criteria (based on the NMD-substrate sets in Supplemental Table S3). The
numbers below signify the combined number of NMD-substrates based on both
RNA-seq and 5’-end-seq vs the total number of genes in the category. (H) As G,
but for snoRNA host genes.
12
13
Supplemental Figure S6. NMD-sensitive snoRNA Host Genes In Human and
Mouse Cells, Related to Figure 5
(A) As Fig. 5C, but northerns were probed for the following host/snoRNA-pairs: