Coordinated Regulation of Alternative Splicing in Cancer Julian P. Venables, Roscoe Klinck, ChuShin Koh, Julien Gervais-Bird, Anne Bramard, Lyna Inkel, Mathieu Durand, Sonia Couture, Ulrike Froehlich, Elvy Lapointe, Jean-François Lucier, Philippe Thibault, Claudine Rancourt, Karine Tremblay, Panagiotis Prinos, Benoit Chabot and Sherif Abou Elela. Supplementary figures Nature Structural & Molecular Biology: doi:10.1038/nsmb.1608
15
Embed
Coordinated Regulation of Alternative Splicing in Cancer€¦ · Coordinated Regulation of Alternative Splicing in Cancer Julian P. Venables, Roscoe Klinck, ChuShin Koh, Julien Gervais-Bird,
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
Coordinated Regulation of Alternative Splicing in Cancer
Julian P. Venables, Roscoe Klinck, ChuShin Koh, Julien Gervais-Bird, Anne Bramard, Lyna Inkel, Mathieu Durand, Sonia Couture, Ulrike Froehlich, Elvy Lapointe, Jean-François Lucier, Philippe Thibault, Claudine Rancourt, Karine Tremblay, Panagiotis Prinos, Benoit Chabot and
Supplementary Figure 2. Histogram representation of the splicing patterns of selected ASEs. Nor��l tiss�es �re shown in �l�e �nd c�ncer in red. The histogr��s represent �ins o� � ��l�es o� ���,� ��l�es o� ���, ��l�es o� ���, with the smoothened distribution as a dashed line. (a) �ll highly signific�nt (p<10-���� c�ncer-nor��l discriminating ASEs in both breast and ovary (b) All the ASEs, with �e�n � �etween ���-�����, showing no signific�nt shi�t (p>0.1�� �etween nor��l �nd c�ncer in �re�st �nd o��ry.
Supplementary Figure 3. 293 ASEs that shift upon FOX2 knockdown in at least one cell line with at least one siRNA. �xons th�t �re rel�ti�ely excl�ded in the FOX� knockdown �re shown in �l�e �nd those th�t �re incl�ded �re shown in red (see key��. ��t� is shown �or �-�� cell lines (indic�ted�� with � data point for each distinct FOX2 siRN�. ���s which were not detected either �ec��se only one pe�k w�s disco�ered �t >10� o� the tot�l or �ec��se the tot�l �ol�rity w�s <�� nM �re not shown. (a) Cancer-related and apoptotic ASEs from previous studies 9,10,�1 were �ss�yed in �� cell lines. (b) Three cell lines were chosen to assay all detected RefSeq ovary ASEs containing Fox-binding sites. (c) AceView cassette exons from genes expressed in ovarian and breast tissues with proximal Fox-binding sites were �ss�yed in two cell lines. (��� �nd (c�� show d�t� �or pri�er n��es �eginning with ‘Re��eq’ �nd ‘Fox’ respecti�ely in ��pple�ent�ry T��le �.
Supplementary Figure 4. Cartoon representation of 87 FOX2-responsive exons indicating FOX-binding site position. C�rtoon represent�tion o� 8�� FOX�-responsi�e exons showing U�C�U� (�l�e�� �nd �U��C� (red�� sites in the exon �nd the �00 n�cleotides �t the ends o� e�ch intron �nd in �p to 100 n�cleotides o� fl�nking exon. When introns �re <400 n�cleotides in length, the dist�l p�rts �re shortened proportion�tely in the dr�wing. �ene n��es �re to the le�t �nd the shi�t in the FOX� knockdowns is shown to the right as a bar chart and numerically (a) Those ASEs containing FOX-binding sites in the downstre�� 1��0 or �pstre�� 80 n�cleotides th�t �re consistent with the o�ser�ed splicing shi�t (b) The re��ining FOX�-responsi�e exons th�t �re not si�ply expl�ined �y �pstre�� intronic splicing inhi�itors or downstream intronic splicing enhancers. Highlighted genes have cytoskeletal functions.
Supplementary Figure 5. FOX2 knockdown-induced splicing shifts are recapitulated in cancer. The 8�� FOX� hits �re n��ed �t the �otto�. The histogr�� �t the top shows the ��er�ge FOX� knockdown-ind�ced splicing shi�t in the two cell lines �sed (M��-MB-��1 �nd �KOV-�ip1��. �ligned with this d�t�, the he�t��ps show the splicing �eh��io�r o� these 8�� ���s (col��ns�� in the 46 o��ri�n (�pper he�t��p�� �nd 46 �re�st (lower he�t��p�� tiss�es. Uns�per�ised co�plete cl�stering o� the tiss�es (rows�� w�s per�or�ed �sing �y ��clide�n dist�nces on col��n-sc�led � ��l�es. The res�lting� ��l�es. The res�lting dendrograms are shown on the left, and the normal and tumor tissues are indicated by blue and redindicated by blue and red ��rs respecti�ely on the right. �c�led � ��l�es �re shown in colors representing the n���er o� st�nd�rd de�i�tions �ro� the �e�n � ��l�e �or e�ch ��� (Z-score��.
Supplementary Figure 6. Conservation of Fox hits correlates with functionality. Position of all exon-proximal UGCAUG sites is plotted on the x axis. The y axis shows the number of other species’ seq�ences, ��pping to the h���n geno�e �sing the UC�C d�t���se (geno�e.�csc.ed���, th�t h��e � conser�ed site within fi�e n�cleotides o� th�t position. The sites linked to exons th�t shi�t �p (�re incl�ded�� �nd down (excl�ded�� �pon FOX� knockdown �re colored red �nd �l�e respecti�ely. The solid and dashed lines depict the background level of conservation for all possible hexamers around the FOX responsive and non-responsive exons respectively.
Supplementary Figure 7. Expression of FOX2 in cancer. I���nohistoche�istry o� FOX� st�ining in typical examples of normal and cancerous breast and ovary from http://www.proteinatlas.org/. FOX� staining is brown while the nuclear counter stain is blue. Cell types are indicated.
Supplementary Figure 8. 123 PCR-validated Fox-dependent internal human exons. Venn diagram showing the o�erl�p �etween FOX-responsi�e ���s in nine di��erent st�dies indic�ted �y first ��thor (��ll re�erence gi�en �elow��. Where �ore th�n one FOX-dependent exon occ�rs in � gene the exon size is shown in brackets. A lower case letter after the gene symbol indicates the reason why primers were not designed in this study: s exon too small, l exon too large, c ‘complex’ alternative splicing region with three or more possible alternative transcripts, i. exon fl�nked �y short introns (<400��, e expression not detected by microarray, n no FOX-�inding sites within the r�nge (-80 to +1��0�� o� this st�dy, �TP��C1 was not detected in normal ovary. Two ASEs were not detected because of the cell lines used; for MBNL1 (��4�� we o�ser�ed no e��ect in M��-MB-��1 �nd �KOV-�ip1 cells. For the T�C� �ltern�ti�e exon we found no incorporation in the cell line controls so the predicted reduction in exon incorporation could not occur.
• P. B�r�ni�k, J. R. Chen, M. �. ��rci�-Bl�nco, Mol Cell Biol �6, 1�09 (Fe�, �006��.• M. �eg�illien et �l., Blood 98, �809 (�ec 1��, �001��.• Y. Jin et �l., ���o J ��, 90�� (Fe� 1��, �00���.• L. P. Li�, P. �. �h�rp, Mol Cell Biol 18, �900 (J�l, 1998��.• �. N�k�h�t�, �. K�w��oto, N�cleic �cids Res ��, �0��8 (�00����.• J. �. Underwood, P. L. Bo�tz, J. �. �o�gherty, P. �toilo�, �. L. Bl�ck, Mol Cell Biol ���, 1000�� (No�, �00����.• �. W. Yeo et �l., N�t �tr�ct Mol Biol 16, 1�0 (Fe�, �009��.• Zh�ng et �l., �enes �e� ��, �����0 (�ep 1��, �008��.
Supplementary Figure 9. FOX2 binding sites in the vicinity of FOX-regulated exons. Box plot for strong (�0�+��, �edi�� (10-�0�� �nd non-hits (0-10�� showing ‘��� rel�ti�e CLIP t�g density’ c�lc�l�ted from the data of Yeo et al. Nature Structural & Molecular Biology 16, 1�0 - 1��� (�009�� which w�s deposited �s � tr�ck �nder “Reg�l�tion” in the hg1�� �nd hg18 �eno�e Browsers �t http://genome.ucsc.edu/. The ‘��� rel�ti�e CLIP t�g density’ is defined �s the n���er o� FOX� CLIP t�gs per kilo��se between the primers used to study the ASE divided by the number of tags per kilobase distal to those pri�ers. Note the l�rge incre�se in CLIP t�g density in the �icinity o� the �ost strongly reg�l�ted FOX�-dependent exons th�t shi�ted �ore th�n �0� (p=0.000���. �ll the ���s �or which FOX� knockdown d�t� were ���il��le in ��pple�ent�ry T��le � were co�p�red in this �n�lysis.
Supplementary Table 1. Enriched motifs in and adjacent to cancer associated alternative exons. Enriched motifs (p<0.00���� in the first 100 n�cleotides o� downstre�� intron in the strong hit (p<10-���� exons th�t �re pre�erenti�lly incorpor�ted (the �ps�� or excl�ded (the downs�� in o��ri�n c�ncer. ��ccessi�e columns denote 1. The enriched motif, 2. the number of times it occurs in the downs or the number of times it occurs in the ups 3. the number of occurrences in the ups, downs and non hits together, 4. p ��l�e �or the enrich�ent in col��n � co�p�red to col��n � �y Fisher’s ex�ct test.
Supplementary Table 2 available separately as Excel file. ASEs that shift significantly in ovarian or breast cancer or by FOX2 knockdown. Each marker ASE takes one row. Columns contain the following information: (a) Gene symbol (b) Gene name (c) Gene ontology terms (d) TCGA cancer gene category annotations (e) and (f) In-house primer names (primer names not beginning with ‘RefSeq’ or ‘Fox’ are from previous studies 9,10,�1 (g) and (h) Forward and reverse primer sequences (i) and (j) Forw�rd �nd re�erse pri�er chro�oso��l coordin�tes �ro� M�rch �006 �sse��ly (hg18, B�ild �6�� (k) and (l) short �nd long expected PCR prod�ct sizes (m) �ize di��erence �etween the iso�or�s (n) Type of splicing event being distinguished (o-r) Ovarian cancer data (o) Me�n � ��l�es in nor��l (p) Me�n � in c�ncer (q) �plicing shi�t (c�ncer – nor��l�� (r) p-value (s-v) Breast cancer data (s) Me�n � values in normal (t) Me�n � in c�ncer (u) �plicing shi�t (c�ncer – nor��l�� (v) p-value. (w) Me�n � �or M��-MB-��1 �nd �KOV-�ip1 cells (x) Me�n � �or FOX� knockdown (y) Me�n FOX� shi�t (knockdown - control�� (z) Reference splice form accession number (aa) Reference splice form amino acid sequence (ab) Alternative splice form amino acid sequence (ac) Domains affected by the alternative splice (ad) Percent change in protein length (ae) Protein �or� in o��ri�n c�ncer (longer or shorter�� (af) Protein �or� in �re�st c�ncer (longer or shorter�� (ag) Protein �or� in Fox knockdown (longer or shorter��.
Supplementary Table 3. Cancer- and FOX2-regulated ASEs in cytoskeletal genes. Table showing the n��es o� �9 genes with c�ncer �nd/or FOX�-reg�l�ted �� �rr�nged �ccording to the type o� cytoskelet�l fil��ent �nd process they �re �ssoci�ted with; col��n � shows keywords rel�ting to the protein’s specific ��nction. The ��er�ge percent shi�t in splicing is shown �or selected reg�l�ted ���s in the fin�l col��ns. � shi�ts gi�en �re c�ncer-nor��l or knockdown-control.