research article unravelling the rna-binding properties...

Research ArticleUnravelling the RNA-Binding Properties of SAFB Proteins inBreast Cancer Cells

Elaine Hong1 Andrew Best2 Hannah Gautrey1 Jas Chin1 Anshuli Razdan1

Tomaz Curk3 David J Elliott2 and Alison J Tyson-Capper1

1 Institute of Cellular Medicine Faculty of Medical Sciences Newcastle University Framlington PlaceNewcastle upon Tyne NE2 4HH UK2Institute of Genetic Medicine Faculty of Medical Sciences Newcastle University Central ParkwayNewcastle upon Tyne NE1 3BZ UK3Faculty of Computer and Information Science University of Ljubljana Traska Cesta 25 51-1000 Ljubljana Slovenia

Correspondence should be addressed to Alison J Tyson-Capper alisontyson-cappernewcastleacuk

Received 29 January 2015 Accepted 15 April 2015

Academic Editor Claudia Ghigna

Copyright copy 2015 Elaine Hong et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Scaffold attachment factor B1 (SAFB1) and SAFB2 proteins are oestrogen (ER) corepressors that bind to and modulate ER activitythrough chromatin remodelling or interaction with the basal transcription machinery SAFB proteins also have an internalRNA-recognition motif but little is known about the RNA-binding properties of SAFB1 or SAFB2 We utilised crosslinkingand immunoprecipitation (iCLIP) coupled with high-throughput sequencing to enable a transcriptome-wide mapping of SAFB1protein-RNA interactions in breast cancer MCF-7 cells Analysis of crosslinking frequency mapped to transcript regions revealedthat SAFB1 binds to coding and noncoding RNAs (ncRNAs) The highest proportion of SAFB1 crosslink sites mapped to ncRNAsfollowed by intergenic regions open reading frames (ORFs) introns and 31015840 or 51015840 untranslated regions (UTR) Furthermore wereveal that SAFB1 binds directly to RNA and its binding is particularly enriched at purine-rich sequences not dissimilar to theRNA-binding motifs for SR proteins Using RNAi we also show for the first time that single depletion of either SAFB1 or SAFB2leads to an increase in expression of the other SAFB protein in both MCF-7 and MDA-MD231 breast cancer cells

1 Introduction

Thegrowing interest in SAFB1 and SAFB2 proteins in relationto cancer is generated from their well described ability tobind to and modulate ER-120572 a central player in breast cancerdevelopment Moreover a role for SAFB1 in RNA splicingand metabolism has also been proposed Nayler et al [1] firstdescribed interactions between SAFB1 with RNA polymeraseII and a subset of serinearginine-richRNAprocessing factors(SR proteins) suggesting that SAFB1 serves as a molecularbase for the assembly of a transcriptome complex that coupleschromatin organising SMARs elements with transcriptionand pre-mRNA processing [1] Protein-protein interactionsbetween SAFB1 and a range of RNA-binding proteins includ-ing hnRNP A1 hnRNP D hnRNP G SR splicing regulatoryprotein 86 (SRrp86) SR protein kinase 1 (SRPK1) and Src-associated substrate in mitosis of 68 kDa (Sam68) provide

reasonable evidence to implicate a role in alternative splicing[2ndash6] However it is still not known whether these SAFBproteins exert their effects on pre-mRNA splicing throughdirect RNA interaction or by tethering to other splicingfactors

SAFB1 and SAFB2 proteins share a highly conservedRNA-recognition motif (RRM) with 98 similarity in thecentral region although until now their direct RNA-bindingpotential has remained unclear SAFB1 has also been labelledas a novel hnRNP protein due to its similarity to the highlyconserved RBD found in the hnRNP protein family [6]Subsequent studies have implicated both SAFB proteins inalternative splicing as overexpression of SAFB1 and SAFB2inhibits splicing of a TRA2B variable exon [5 7] Howeverfurther investigation usingmutants lacking the RRM domainrevealed that SAFB1rsquos ability to inhibit TRA2B exon skippingwas independent of its RNA-binding ability [7]This evidence

Hindawi Publishing CorporationBioMed Research InternationalVolume 2015 Article ID 395816 9 pageshttpdxdoiorg1011552015395816

2 BioMed Research International

suggests that SAFB1 may not bind directly to TRA2B pre-mRNA to regulate exon skipping but could possibly mediatean indirect effect through its interaction with various splicingfactors [2 4ndash6] In an unrelated study in vitro evidence hasshown that the RRM domain of SAFB1 was able to bind RNAisolated fromMCF-7 breast cancer cells although the identityof the RNA targets was not described [8] The current studywas designed to establish whether SAFB proteins exert theirRNA processing functions through direct RNA interaction aswell as by tethering to other protein factors

2 Material and Methods

21 iCLIP CLIP with individual nucleotide resolution(iCLIP) was performed for SAFB1 usingMCF-7 breast cancercells based on a published protocol [9] In brief MCF-7 cellswere irradiated with 150mJcm2 of UV at 254 nm and cellpellets resuspended in lysis buffer treated with Turbo DNaseI (Ambion) and high (1 10 dilution) or low (1 500 dilution)RNase I (Ambion) Dynabeads Protein A or DynabeadsProtein G (Invitrogen) were resuspended in lysis buffer con-taining 5120583g SAFB1 antibody (Sigma-Aldrich) and preclearedlysate was added to themagnetic beads for immunoprecipita-tion at 4∘C for 2 hours RNA 31015840 ends were dephosphorylatedand RNA linkers ligated Magnetic beads were then resus-pended in PNK mix containing 32P-120574-ATP to radioactivelylabel the RNA 51015840 ends as previously described [9] Protein-RNA complexes were isolated following electrophoresis (seeSupplementary Figure 1 in Supplementary Material availableonline at httpdxdoiorg1011552015395816) PrecipitatedRNA was reverse transcribed in RNAprimer mix containingdifferent Rclip primers with individual barcode sequencesfor each replicate Three gel fragments corresponding tocDNA size were cut at 120ndash200 nucleotides (high) 85ndash120 nucleotides (medium) and 70ndash85 nucleotides (low)(Supplementary Figure 1(B)) Three independent biologicalreplicates were prepared for sequencing using the TruSeqSample Preparation kit (Illumina) and sequenced on theGenomeAnalyser II system (GAIIx Illumina) Bioinformaticanalyses were performed on the web-based iCount software(httpicountbiolabsi) Mapping of SAFB1 crosslink sitesto regions of respective genes was visualised in UCSCGenome Browser (httpgenomeucscedu) and a graphi-cal representation of the novel SAFB1 consensus bindingmotif was designed using the web-based WebLogo soft-ware (httpweblogoberkeleyedu) Potential target genesthat contain SAFB1 binding sites were selected for furthervalidation using qRT-PCR with TaqMan gene expressionassays

211 Transient Transfections MCF-7 andMDA-MB-231 cellswere reverse transfected with two independent sets ofSilencer Select siRNA for SAFB1 and SAFB2 Silencer Neg-ative Control siRNA Silencer Select GAPDH and 120573-actinPositive Control siRNA (Life Technologies) using INTER-FERin Transfection Agent (Polyplus transfection) accordingto the manufacturerrsquos instructions Following optimisationcells were seeded at 35 times 104 cellswell in 24-well platesor 175 times 105 cellswell into 6-well plates After 24 hours

cells were transfected with 5 nM siRNA with 4120583LmL ofINTERFERin transfection agent RNA was collected 48 and72 hours after transfection and protein collected 72 hoursafter transfection In all experiments levels of knockdown byRNAi were assessed at the RNA and protein level by PCR andimmunoblotting

212 RNA Isolation and PCR Total RNA from culturedcells was extracted with RNeasy spin columns (Qiagen) orthe SV Total RNA Isolation System (Promega) accordingto manufacturerrsquos instructions One 120583g of total RNA wasreverse transcribed using oligo (dT) primers and SuperscriptII Reverse Transcriptase (RT) (Invitrogen Life Technologies)following manufacturerrsquos instructions Conventional PCRwas performed using PCR master mix (Promega) qPCR wasperformed using TaqMan gene expression assays (Life Tech-nologies) according to the manufacturerrsquos instructions Val-idated TaqMan probes were selected to target specific genesas follows SAFB1 (Hs01561652 gl) SAFB2 (Hs01006796 g1)ITGB4 (Hs00236216 m1) SHF (Hs00403125 m1) MALAT-1 (Hs00273907 s1) and 120573-actin (Hs99999903 m1) Dataanalysis was performed using the comparative Ct methodnormalised against 120573-actin expression Experiments wereperformed in triplicate and statistical analysis was performedusing Studentrsquos 119905-test or Repeated Measures ANOVA withDunnettrsquos Multiple Comparison Test All effects at 119875 lt 005are reported as significant

22 RNA Immunoprecipitation MCF-7 cells were washed inice cold PBS and then collected in lysis buffer (10mM Tris-HCl (pH 75) 150mM NaCl 05 NP-40 and 1 TritonX-100) containing protease inhibitors (Sigma-Aldrich) andRNase OUT (Invitrogen) Equal amounts of cell lysate werethen incubated for 3 hours at 4∘C with 2120583g of either rabbitanti-SAFB1 antibody (Genetex) or rabbit IgG control (SantaCruz) Cell lysates and antibodies were then incubated withDynabeads Protein G (Invitrogen) for a further hour at 4∘CBeads were then washed five times with lysis buffer beforethe immunoprecipitated RNA was collected in Trizol reagent(Invitrogen) according to manufacturerrsquos instructions TheRT-PCR was performed as before but half of the RNAobtained from the immunoprecipitation was used in eachreaction Fold enrichment of target mRNA was determinedafter normalization to the input and rabbit IgG controls

3 Results

31 Identification of RNA-Binding Sites for SAFB1 in BreastCancer Cells Although the role of SAFB proteins in RNAprocessing has been speculated the function of their highlyhomologous internal RRM has not been examined Wesought to identify possible direct RNA targets for SAFB1 inbreast cancer cells using iCLIP technology [9ndash12] combinedwith high-throughput sequencing and mapping to generate atranscriptome-wide binding map for SAFB1 SAFB1 protein-RNA complexes were successfully generated by immuno-precipitation and RNA recovered and purified from 3 inde-pendent iCLIP replicates (Supplementary Figure 1) High-throughput sequencing and bioinformatics generated a total

BioMed Research International 3

SAFB1 crosslink sites

ncRNAIntergenicORF

Intron3UTR5UTR

(a)

SAFB1 crosslink enrichment sites

ncRNAIntergenicORF

Intron3UTR5UTR

(b)

Distribution within ncRNA subclasses

ncRNA subclasses

snRN

A

Mt R

NA

snoR

NA

rRN

A

miR

NA

lincR

NA

Oth

er R

NA

0

7

14

ncRN

A (

)

SAFB1 iCLIP

(c)

Figure 1 Distribution of significant SAFB1 crosslink sites within RNA segment types (a) The proportion of cDNAs mapped to differenttranscript regions relative to the total number of cDNA reads revealed that the highest percentage of cDNAs was mapped to ncRNA (4708)followed by intergenic regions (2324) ORFs (1938) introns (523) 31015840 UTRs (383) and 51015840 UTRs (086) (b) The fold enrichmentof cDNA density in different types of RNAs relative to cDNA density in the whole genome highest density enrichment in ncRNAs (c) Thedistribution of SAFB1 crosslink sites within different ncRNA subclasses revealed significant abundance in snRNA Mt RNA and snoRNAldquoOther RNArdquo consists of pseudogenes and processed transcripts with no known ORF or function

of 1145271 unique cDNA reads with single-hits mapping tothe human genome which were subsequently filtered downto 587119 significant unique cDNAs distributed over 127308binding sites in 25207 SAFB1 crosslink clusters (FDR lt 005)A snapshot of the view for SAFB1 crosslink sites on theUCSC Genome Browser (httpgenomeucscedu) is shownin Supplementary Figure 1(C)

iCLIP identified binding sites for SAFB1 across thewhole transcriptome where 100 of significant cDNA readsmapped to the sense orientation in annotated genes This

confirms the high strand specificity of iCLIP also observedin other studies [11 13] Analysis of crosslinking frequencymapped to transcript regions revealed that SAFB1 binds tocoding and noncoding RNAs (ncRNAs) Notably the highestproportion of 127308 SAFB1 crosslink sites from significantclusters map to ncRNAs followed by intergenic regions openreading frames (ORFs) introns and 31015840 or 51015840 untranslatedregions (UTRs) (Figure 1(a)) When the cDNA density foreach transcript region was analysed relative to the cDNAdensity in the whole genome the highest density enrichment


Pentamer

GAAAAAGAAAAAACAAACAAAAAACATGAAAAGAAGAAGAAAAGATGAAAAACCAACAACAATGAATCAAACCAAACAAACAGAACAAATCAAAATGAAG

3464431201305333007229955298572898826496261442585025815256532552225514254502522724147239172366023323

Frequency

(a)

1 2 3 4 5

5998400 3998400

weblogoberkeleyedu

0 = crosslink nucleotide

1= first nucleotide of cDNA

(b)

Base position

Base

A

C

G

T

1 2 3 4 5

65 60 60 75 80

15 15 20 10 10

10 15 15 15 5

10 10 5 0 5

(c)

Figure 2 In vivo consensus binding motif of SAFB1 (a) The frequency of pentamers surrounding SAFB1 crosslink sites was determinedAdenine represents 68of the 20 pentamers that has the highest frequencies (b)WebLogo showing base frequencies of each base at respectivepositions of the pentamer SAFB1 binds to adenine-rich motifs (c) The frequency of each base relative to its position within the pentamerwas summarised in this table The highest frequency of adenine was observed at base position 5 thymine was excluded at base position 4 ofthe consensus binding motif This consensus binding motif was predicted from iCLIP cDNA libraries therefore the uracil base is referred toas thymine in these sequences

was detected in ncRNAs (Figure 1(b)) The distribution ofSAFB1 crosslink sites within ncRNA subclasses was alsoanalysed SAFB1 crosslink sites were most abundant in smallnuclear RNA (snRNA) mitochondrial RNA (Mt RNA) andsmall nucleolar RNA (snoRNA) (Figure 1(c))

32 Identification of an RNA-Binding Motif for SAFB1 The invivo binding specificity of SAFB1 is still currently unknownThe advantage of single nucleotide resolution provided byiCLIPmethod enabled the assessment of sequence specificityfor SAFB1 binding To derive whether a consensus bindingmotif exists for SAFB1 enriched pentamer sequences sur-rounding the crosslink sites were identified The frequenciesof each pentamer were analysed to determine the top 20pentamers for SAFB1 Strikingly adenine appeared as themost frequent nucleotide in the top 20 pentamers andrepresents 68 of the enriched pentamers (Figure 2(a)) Thepredicted SAFB1 consensus binding motif contains adenine-rich sequences derived from the pentamers (Figure 2(b))When the frequency of each nucleotide in the cDNA libraries

was analysed relative to its base position a strong inclusion ofadenine at base position 5 was observed (80) while thymine(uracil inRNA)was excluded at base position 4 of the putativeRNA-binding motif (Figure 2(c)) The consensus bindingmotif for SAFB1 has not been described before thereforethis novel finding is likely to be of significance to further ourcurrent understanding of SAFB1 RNA-binding specificity

33 Identification of Novel RNA Targets from Data Generatedby iCLIP Data analysis of bound RNAs revealed the numberof SAFB1 crosslink sites within each RNA target When thetop 10 RNA targets with the largest number of crosslink siteswere listed according to each RNA segment the positionof SAFB1 binding within each gene was visualised usingthe UCSC Genome Browser (Supplementary Figure 2) Thisenabled the identification of several interesting RNA targetsthat were selected for validation Further experimentationswere performed using qRT-PCR or conventional PCR onRNAi transfectedMCF-7 andMDA-MB-231 cells to verify theeffect of loss of SAFB1 on the expression of these selectedRNA


0

02

04

06

08

1

12

14

16

ctrl SAFB1 SAFB2 SAFB12

Relat

ive g

ene e

xpre

ssio

n

Relat

ive g

ene e

xpre

ssio

n

SAFB2

002040608

112141618


SAFB1

(a)

ctrl

SAFB

1

SAFB

2

SAFB

1+SA

FB2

SAFB1

SAFB2

175

175

42

(kDa)

120573-actin

(b)

0

05

1

15

2

SHF

relat

ive e

xpre

ssio

n (fo

ld ch

ange

)

MCF-7MDA-MB-231

ctrl SAFB1 SAFB2 SAFB1+ SAFB2

lowast

lowast

(c)

0

05

1

15

2

25

ITG

B4 re

lativ

e exp

ress

ion

(fold

chan

ge)

MCF-7MDA-MB-231


lowast

lowast

(d)

Figure 3 Loss of SAFB proteins affects expression of SHF and ITGB4 Knockdown of SAFB1 in MCF-7 cells increases mRNA and proteinexpression of SAFB2 and similarly knockdown of SAFB2 leads to increased expression of SAFB1 MC7-7 cells were transiently transfectedwith negative SAFB1 SAFB2 or SAFB1 and SAFB2 siRNA mRNA levels were measured by qRT-PCR Data represents the average of threebiological replicates plusmn SD Statistical significance of mRNA expression was calculated using Studentrsquos 119905-test lowast = 119875 lt 005 (b) Protein levelswere analysed by immunoblotting using SAFB1 and SAFB2 antibodies (c) The effect of SAFB knockdown on SHF and ITGB4 expression byqRT-PCR using validated TaqMan probes specifically targeting SHF (c) or ITGB4 (d) Data represents the average of three biological replicatesplusmn SD Statistical significance of mRNA expression was calculated using Studentrsquos 119905-test 119875 lt 005

targets Since SAFB2 shares 98 sequence homology to theRRM of SAFB1 these cells were also depleted of SAFB2 anddouble knockdown of SAFB1 and SAFB2 was also includedinterestingly data shows that when MCF-7 cells are reducedof SAFB1 by RNAi the levels of SAFB2 mRNA and proteinincrease (Figure 3(a)) Likewise levels of SAFB1 increase afterknockdown of SAFB2 (Figure 3(b)) A similar pattern wasobserved when the breast cancer MDA-MB-231 cells wereused (Supplementary Figure 3)

Analysis of the RNA map revealed a large number ofSAFB1 binding sites on the SHF mRNA particularly accu-mulated around the alternative promoter (SupplementaryFigure 3) the use of alternative promoters plays a significantrole in gene expression control (reviewed in [14ndash16]) Moreimportantly the aberrant use of alternative promoter hasbeen linked to a number of diseases including cancer [17]Therefore identification of SHF as a potential RNA target forSAFB1 warrants further investigation In MCF-7 noninvasive

breast cancer cells a reduction in SAFB1 did not appear tosignificantly alter SHF mRNA expression whereas in MDA-MB-231 invasive breast cancer cells there was a significantincrease in SHFmRNA expression when SAFB1 was reduced(Figure 3(c)) Loss of SAFB2 and both SAFB proteins byRNAi increased SHF expression again supporting their roleas transcriptional repressors Another potential RNA targetfor SAFB proteins is ITGB4 The observed loss of SAFB1in both breast cancer cell lines does not have an effecton ITGB4 mRNA expression whereas loss of SAFB2 andboth SAFB proteins significantly increased ITGB4 expression(Figure 3(d))

34 Malat-1 A ncRNA Target for SAF2 Another interestingobservation from the iCLIP dataset revealed significantSAFB1 binding sites to metastasis associated lung adeno-carcinoma transcript 1 (MALAT-1) MALAT-1 is a highlyconserved long ncRNA enriched in nuclear speckles that


20

15

10

05

00

MALAT1lowastlowast

Relat

ive m

RNA

expr

essio

n

SAFB1 SAFB2 SAFB12 NC

(a)

1

inpu

t

IgG

SAFB

1Ab

1

inpu

t (RT

minus)

IgG

(RTminus

)

SAFB

1Ab

(RTminus

)

300200100

100

bp la

dder

MAL

AT1

(b)

000

100

200

300

SAFB1 AbIgG

Fold

enric

hmen

t

(c)

SAFB1

SRSF2

DAPI

DAPI

(d)

Figure 4 SAFB2 regulates expression of MALAT-1 (a) Expression of MALAT-1 was measured by qRT-PCR using RNA from MCF-7 andMDA-MB-231 cells transfectedwith negative SAFB1 SAFB2 or SAFB1 and SAFB2 siRNAusing validatedTaqManprobes specifically targetingMALAT-1 Data represents the average of three biological replicates plusmn SD Repeated Measures ANOVA with Dunnettrsquos Multiple ComparisonTest statistical significance of mRNA expression was calculated using Studentrsquos 119905-test119875 lt 001 Enrichment ofMALAT1RNA by conventionalPCR (b) and qPCR (c) after RNA immunoprecipitation with anti-SAFB1 in MCF-7 cells No enrichment was observed using IgG (d)Intranuclear distribution of SAFB1 and SRSF2 inMCF-7 cells by immunofluorescent staining Confocal laser microscopy revealed a punctatepattern for SAFB1 and SRSF2 in nuclear speckles

regulates alternative splicing by modulating splicing factorphosphorylation [18] MALAT-1 is overexpressed in manydifferent cancers including breast and is considered anoncogenic long ncRNA [19 20] We show that in MCF-7cells loss of SAFB2 resulted in an increase in the levels ofMALAT-1 expression (Figure 4(a)) Moreover we also testedthe ability of SAFBI to immunoprecipitateMALAT-1 RNA inMCF-7 cells enrichment ofMALAT-1RNAwas observed anddetermined by conventional PCR and qPCR (Figures 4(b)and 4(c))

4 Discussion

The presence of the highly conserved RRMs within SAFB1and SAFB2 proteins has been a subject of interest sincetheir discovery especially in relation to their RNA-bindingpotential Despite the fascination very little has been under-taken until now to describe their RNA-binding capabilitiesInitial in vitro evidence showed that the RRM of SAFB1 isable to bind RNA when glutathione S-transferase- (GST-)tagged SAFB1 protein combined with total RNA fromMCF-7cells generated a PCR product when reverse transcribed andPCR amplified [8] However the identity of the RNA targetswas not described and important questions with respect to

the role of SAFB1 and SAFB2 in RNA processing remainedunanswered

iCLIP has been proven as a powerful method to deter-mine protein-RNA interactions in vivo on a global scale andidentify the positions of crosslink sites at nucleotide reso-lution [11] The random barcode incorporated to individualcDNA molecules addresses the problem of PCR artifactsfaced by all high-throughput sequencing methods iCLIP hasgenerated a huge dataset and this is an initial analysis of theRNA-binding data for SAFB1 In this study a global viewcomparison of the complete dataset from each individual bio-logical replicate showed that all datasets generated consistentand reproducible results underlining the high quality iCLIPdata achieved by high stringency purification and librarypreparation

The identification of in vivo targets by iCLIP enabledthe mapping of transcript regions and RNA classes boundby SAFB1 An overview of the iCLIP results showed thatthe important class of RNAs bound by SAFB1 was ncRNAsInterestingly this binding distribution of SAFB1 is similar tothe RNA-binding distribution of splicing factors SRSF3 andSRSF4 rather than hnRNP C protein SAFB1 was initiallyclassified as a novel member of the hnRNP protein family


[6] Recent work by Anko et al utilised iCLIP to reveal thatconcentrated SRSF3 and SRSF4 binding sites were also inncRNAs [21] while Konig et al [11] showed that hnRNPC binding sites were most abundant within introns [11]This observation raises the possibility that SAFB1 proteinmay have similar characteristics to SR proteins rather thanhnRNP protein members although at this stage this is onlyspeculative

The term ncRNA is commonly used for RNA that doesnot encode a protein but appears to comprise internal signalsthat control various levels of gene expression includingchromatin organisation transcription RNA splicing editingtranslation and turnover (reviewed in [22]) Consistent withalready known functions of SAFB proteins concentratedSAFB1 binding in ncRNAs observed from the iCLIP datacould possibly contribute to its various role in chromatinorganisation transcription and RNA metabolism Analysisof SAFB1 distribution within ncRNA subclasses revealedmost abundant SAFB1 binding in snRNAs snRNAs are aclass of small RNA molecules found to be uridylate-richand localised within the nucleus [23] The most commonmembers of snRNAs are the U1 U2 U4 U5 and U6snRNAs that form the spliceosome along with many otherprotein factors and primarily function in pre-mRNA splicing(reviewed in [24]) The high distribution of SAFB1 bindingsites in snRNAs observed in this study supports previouslyidentified interactions between SAFB1 with various RNAprocessing factors and splicing machinery [1 4 6 25]

The genome-wide single nucleotide resolution of iCLIPdata enabled the prediction of in vivo consensus bind-ing sequences for SAFB1 based on the enriched pentamersequences surrounding the crosslink sites This study is thefirst to report that SAFB1 binds a consensus adenine-richsequence in vivo Closer examination of the putative con-sensus binding sequence revealed the exclusion of thymine(uracil in RNA) at base position 4 and a strong inclusion ofadenine at base position 5 Interestingly the predicted SAFB1-binding motif is not dissimilar to purine-rich sequencesfound inRNA-bindingmotifs for other SR proteins (reviewedin [26])

When analysing SAFB1 crosslink sites within protein-coding transcripts SAFB1 binding density was also enrichedin regions encompassing the ORF and 31015840 and 51015840 UTR Thelist of RNA targets was filtered according to the region anddensity of SAFB1 binding to identify targets that are relevantto tumourigenesis Several interesting genes were highlightedin this study including SHF ITGB4 andMALAT-1

SHF is a member of a family of adaptor protein charac-terised by their ability tomediate protein-protein interactionsthrough their Src homology 2 domain [27 28] Althoughthe function of SHF is not fully understood evidence hasshown that overexpression of SHF significantly decreases therate of growth factor-induced apoptosis in neuroblastomacells [27] Subsequently Ohira et al showed that SHFmRNAwas highly expressed in nonmetastatic neuroblastoma com-pared to metastatic tumour samples [29] Another recentstudy provided evidence that loss of SHF increased cellularmobility and the invasive capability of neuroblastoma cells[30]

Initial iCLIP data from this study revealed enrichedSAFB1 binding sites at the alternative promoter of SHF Asthe aberrant expression of alternative promoters is linkedto cancer SAFB1 binding surrounding this region gatheredan interest for further examination Interestingly the knock-down of SAFB1 inMCF-7 cells did not significantly alter SHFexpression while the knockdown of SAFB2 or both SAFBproteins significantly increased SHF expressionThis suggeststhat direct SAFB1 binding to the alternative promoter did notaffect the expression of this gene MDA-MB-231 cells wereincluded in this part of the study for comparison and in thiscell type increased SHF expression was observed in the lossof SAFB1 or SAFB2 and both SAFB proteins

Multiple alternatively spliced transcript variants encod-ing distinct isoforms have been found for ITGB4 althoughthe full function of most variants remains to be defined[31ndash33] Alternative splicing mechanism has been indicatedto subtly regulate the ligand binding and signalling activityof many integrin subunits (reviewed in [34]) Although themechanism and significance of alternative splicing in ITGB4have not been elucidated the discovery of SAFB1 bindingsites in its exonic regions may provide a new perspective tofurther understand the mRNA processing of ITGB4

We also identified another potential novel target forSAFB2mdashMALAT-1 Previous work shows that MALAT-1colocalises with SRSF2 in nuclear speckles [35] Furthermoreother splicing factors that localise in nuclear speckles suchas SRSF1 SRSF3 and SRSF4 also bind to MALAT-1 [2136] We and others [1 5] observe that SAFB1 distributionhas a similar punctate pattern to SRSF2 (Figure 4(d)) it istherefore conceivable that SAFB1 may possess other typicalcharacteristics of a splicing factorwhich supports its observedfunction in pre-mRNA splicing [1 5 7]

Interestingly single depletion of either SAFB1 or SAFB2led to an increase in expression of the other this patternwas mirrored at both mRNA and protein levels Our studyalso suggests that SAFBI and SAFB2 may themselves havedifferent and overlapping RNA targetsThis observation sup-ports previous speculations regarding the distinct molecularroles between SAFB1 and SAFB2 [5 37 38] We concludethat SAFB proteins may share multiple similarities in RNA-binding pattern and characteristics with SR proteins Anal-ysis of SAFB1 crosslink regions and RNA targets confirmsprevious reports regarding its interaction with other RNAprocessing machinery and function Further work will nowbe undertaken to define whether SAFB1 and SAFB2 functionsynergistically or compensatory as RNA-binding proteins inbreast cancer cells

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

Elaine Hong was sponsored by Dr William Harker Founda-tion PhD Studentship Additional funding was provided bythe Royal Victoria Infirmary Breast Cancer Research Appeal


References

[1] O Nayler W Stratling J-P Bourquin et al ldquoSAF-B proteincouples transcription and pre-mRNA splicing to SARMARelementsrdquoNucleic Acids Research vol 26 no 15 pp 3542ndash35491998

[2] YArao RKuriyama F Kayama and SKato ldquoAnuclearmatrix-associated factor SAFB-B interacts with specific isoforms ofAUF1hnRNP Drdquo Archives of Biochemistry and Biophysics vol380 no 2 pp 228ndash236 2000

[3] E Nikolakaki R Kohen A M Hartmann S Stamm E Geor-gatsou and T Giannakouros ldquoCloning and characterization ofan alternatively spliced form of SR protein kinase 1 that interactsspecifically with scaffold attachment factor-Brdquo The Journal ofBiological Chemistry vol 276 no 43 pp 40175ndash40182 2001

[4] J Li I C Hawkins C D Harvey J L Jennings A J Link andJ G Patton ldquoRegulation of alternative splicing by SRrp86 andits interacting proteinsrdquoMolecular and Cellular Biology vol 23no 21 pp 7437ndash7447 2003

[5] K A Sergeant C F Bourgeois C Dalgliesh J P Venables JStevenin and D J Elliott ldquoAlternative RNA splicing complexescontaining the scaffold attachment factor SAFB2rdquo Journal ofCell Science vol 120 no 2 pp 309ndash319 2007

[6] F Weighardt F Cobianchi L Cartegni et al ldquoA novel hnRNPprotein (HAPSAF-B) enters a subset of hnRNP complexes andrelocates in nuclear granules in response to heat shockrdquo Journalof Cell Science vol 112 no 10 pp 1465ndash1476 1999

[7] P Stoilov R Dauod O Nayler and S Stamm ldquoHuman tra2-beta1 autoregulates its protein concentration by influencingalternative splicing of its pre-mRNArdquoHumanMolecular Genet-ics vol 13 no 5 pp 509ndash524 2004

[8] S M Townson K Kang A V Lee and S OesterreichldquoStructure-function analysis of the estrogen receptor 120572 core-pressor scaffold attachment factor-B1 identification of a potenttranscriptional repression domainrdquo The Journal of BiologicalChemistry vol 279 no 25 pp 26074ndash26081 2004

[9] J Konig K Zarnack G Rot et al ldquoiCLIPmdashtranscriptome-wide mapping of protein-RNA interactions with individualnucleotide resolutionrdquo The Journal of Visualized Experimentsno 50 Article ID e2638 2011

[10] F Acconcia P Ascenzi A Bocedi et al ldquoPalmitoylation-dependent estrogen receptor alpha membrane localizationregulation by 17beta-estradiolrdquoMolecular Biology of theCell vol16 no 1 pp 231ndash237 2005

[11] J Konig K Zarnack G Rot et al ldquoICLIP reveals the function ofhnRNPparticles in splicing at individual nucleotide resolutionrdquoNature Structural and Molecular Biology vol 17 no 7 pp 909ndash915 2010

[12] J Ule K B Jensen M Ruggiu A Mele A Ule and R BDarnell ldquoCLIP identifies Nova-regulated RNA networks in thebrainrdquo Science vol 302 no 5648 pp 1212ndash1215 2003

[13] Z Wang M Kayikci M Briese et al ldquoiCLIP predicts the dualsplicing effects of TIA-RNA interactionsrdquo PLoS Biology vol 8no 10 Article ID e1000530 2010

[14] T A Y Ayoubi and W J M van de Ven ldquoRegulation of geneexpression by alternative promotersrdquo The FASEB Journal vol10 no 4 pp 453ndash460 1996

[15] R V Davuluri Y Suzuki S Sugano C Plass and T H-MHuang ldquoThe functional consequences of alternative promoteruse in mammalian genomesrdquo Trends in Genetics vol 24 no 4pp 167ndash177 2008

[16] F Koch F Jourquin P Ferrier and J-CAndrau ldquoGenome-wideRNA polymerase II not genes onlyrdquo Trends in BiochemicalSciences vol 33 no 6 pp 265ndash273 2008

[17] G A C Singer J Wu P Yan C Plass T H M Huang andR V Davuluri ldquoGenome-wide analysis of alternative promotersof human genes using a custom promoter tiling arrayrdquo BMCGenomics vol 9 article 349 2008

[18] V Tripathi J D Ellis Z Shen et al ldquoThe nuclear-retainednoncoding RNA MALAT1 regulates alternative splicing bymodulating SR splicing factor phosphorylationrdquoMolecular Cellvol 39 no 6 pp 925ndash938 2010

[19] L Li T Feng Y Lian G Zhang A Garen and X Song ldquoRoleof human noncoding RNAs in the control of tumorigenesisrdquoProceedings of the National Academy of Sciences of the UnitedStates of America vol 106 no 31 pp 12956ndash12961 2009

[20] D S Perez T R Hoage J R Pritchett et al ldquoLong abundantlyexpressed non-coding transcripts are altered in cancerrdquoHumanMolecular Genetics vol 17 no 5 pp 642ndash655 2008

[21] M-L Anko M Muller-McNicoll H Brandl et al ldquoThe RNA-binding landscapes of two SR proteins reveal unique functionsand binding to diverse RNA classesrdquo Genome Biology vol 13no 3 article R17 2012

[22] J S Mattick and I V Makunin ldquoNon-coding RNArdquo HumanMolecular Genetics vol 15 no 1 pp R17ndashR29 2006

[23] H Busch R Reddy L Rothblum and Y C Choi ldquoSnRNAsSnRNPs and RNA processingrdquo Annual Review of Biochemistryvol 51 pp 617ndash654 1982

[24] M C Wahl C L Will and R Luhrmann ldquoThe spliceosomedesign principles of a dynamic RNPmachinerdquo Cell vol 136 no4 pp 701ndash718 2009

[25] J Rappsilber U Ryder A I Lamond and M Mann ldquoLarge-scale proteomic analysis of the human spliceosomerdquo GenomeResearch vol 12 no 8 pp 1231ndash1245 2002

[26] J C Long and J F Caceres ldquoThe SR protein family of splicingfactors master regulators of gene expressionrdquo BiochemicalJournal vol 417 no 1 pp 15ndash27 2009

[27] C K Lindholm J D Frantz S E Shoelson and M WelshldquoShf a Shb-like adapter protein is involved in PDGF-120572-receptorregulation of apoptosisrdquo Biochemical and Biophysical ResearchCommunications vol 278 no 3 pp 537ndash543 2000

[28] M Welsh J Mares T Karlsson C Lavergne B Breant and LClaesson-Welsh ldquoShb is a ubiquitously expressed Src homology2 proteinrdquo Oncogene vol 9 no 1 pp 19ndash27 1994

[29] MOhira AMorohashi H Inuzuka et al ldquoExpression profilingand characterization of 4200 genes cloned from primary neu-roblastomas identification of 305 genes differentially expressedbetween favorable and unfavorable subsetsrdquo Oncogene vol 22no 35 pp 5525ndash5536 2003

[30] D Takagi Y Tatsumi T Yokochi et al ldquoNovel adaptor proteinShf interacts with ALK receptor and negatively regulates itsdownstream signals in neuroblastomardquoCancer Science vol 104no 5 pp 563ndash572 2013

[31] A S Clarke M M Lotz and A M Mercurio ldquoA novel struc-tural variant of the human 1205734 integrin cDNArdquo Cell Adhesionand Communication vol 2 no 1 pp 1ndash6 1994

[32] R N Tamura C Rozzo L Starr et al ldquoEpithelial integrin 12057261205734complete primary structure of 1205726 and variant forms of 1205734rdquoJournal of Cell Biology vol 111 no 4 pp 1593ndash1604 1990

[33] M R van Leusden I Kuikman and A Sonnenberg ldquoTheunique cytoplasmic domain of the human integrin variant1205734E is produced by partial retention of intronic sequencesrdquo


Biochemical and Biophysical Research Communications vol 235no 3 pp 826ndash830 1997

[34] A A de Melker and A Sonnenberg ldquoIntegrins alternativesplicing as a mechanism to regulate ligand binding and integrinsignaling eventsrdquo BioEssays vol 21 no 6 pp 499ndash509 1999

[35] J NHutchinson AW Ensminger CM Clemson C R LynchJ B Lawrence and A Chess ldquoA screen for nuclear transcriptsidentifies two linked noncoding RNAs associated with SC35splicing domainsrdquo BMC Genomics vol 8 article 39 2007

[36] J R Sanford X Wang M Mort et al ldquoSplicing factor SFRS1recognizes a functionally diverse landscape of RNA transcriptsrdquoGenome Research vol 19 no 3 pp 381ndash394 2009

[37] S Hammerich-Hille B A Kaipparettu A Tsimelzon et alldquoSAFB1 mediates repression of immune regulators and apop-totic genes in breast cancer cellsrdquo The Journal of BiologicalChemistry vol 285 no 6 pp 3608ndash3616 2010

[38] M Ivanova K M Dobrzycka S Jiang et al ldquoScaffold attach-ment factor B1 functions in development growth and repro-ductionrdquoMolecular andCellular Biology vol 25 no 8 pp 2995ndash3006 2005

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of


Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology


Molecular Biology International

GenomicsInternational Journal of


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


BioinformaticsAdvances in

Marine BiologyJournal of



Signal TransductionJournal of


BioMed Research International

Evolutionary BiologyInternational Journal of



Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Genetics Research International


Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational



Enzyme Research



Microbiology


suggests that SAFB1 may not bind directly to TRA2B pre-mRNA to regulate exon skipping but could possibly mediatean indirect effect through its interaction with various splicingfactors [2 4ndash6] In an unrelated study in vitro evidence hasshown that the RRM domain of SAFB1 was able to bind RNAisolated fromMCF-7 breast cancer cells although the identityof the RNA targets was not described [8] The current studywas designed to establish whether SAFB proteins exert theirRNA processing functions through direct RNA interaction aswell as by tethering to other protein factors

2 Material and Methods

21 iCLIP CLIP with individual nucleotide resolution(iCLIP) was performed for SAFB1 usingMCF-7 breast cancercells based on a published protocol [9] In brief MCF-7 cellswere irradiated with 150mJcm2 of UV at 254 nm and cellpellets resuspended in lysis buffer treated with Turbo DNaseI (Ambion) and high (1 10 dilution) or low (1 500 dilution)RNase I (Ambion) Dynabeads Protein A or DynabeadsProtein G (Invitrogen) were resuspended in lysis buffer con-taining 5120583g SAFB1 antibody (Sigma-Aldrich) and preclearedlysate was added to themagnetic beads for immunoprecipita-tion at 4∘C for 2 hours RNA 31015840 ends were dephosphorylatedand RNA linkers ligated Magnetic beads were then resus-pended in PNK mix containing 32P-120574-ATP to radioactivelylabel the RNA 51015840 ends as previously described [9] Protein-RNA complexes were isolated following electrophoresis (seeSupplementary Figure 1 in Supplementary Material availableonline at httpdxdoiorg1011552015395816) PrecipitatedRNA was reverse transcribed in RNAprimer mix containingdifferent Rclip primers with individual barcode sequencesfor each replicate Three gel fragments corresponding tocDNA size were cut at 120ndash200 nucleotides (high) 85ndash120 nucleotides (medium) and 70ndash85 nucleotides (low)(Supplementary Figure 1(B)) Three independent biologicalreplicates were prepared for sequencing using the TruSeqSample Preparation kit (Illumina) and sequenced on theGenomeAnalyser II system (GAIIx Illumina) Bioinformaticanalyses were performed on the web-based iCount software(httpicountbiolabsi) Mapping of SAFB1 crosslink sitesto regions of respective genes was visualised in UCSCGenome Browser (httpgenomeucscedu) and a graphi-cal representation of the novel SAFB1 consensus bindingmotif was designed using the web-based WebLogo soft-ware (httpweblogoberkeleyedu) Potential target genesthat contain SAFB1 binding sites were selected for furthervalidation using qRT-PCR with TaqMan gene expressionassays

211 Transient Transfections MCF-7 andMDA-MB-231 cellswere reverse transfected with two independent sets ofSilencer Select siRNA for SAFB1 and SAFB2 Silencer Neg-ative Control siRNA Silencer Select GAPDH and 120573-actinPositive Control siRNA (Life Technologies) using INTER-FERin Transfection Agent (Polyplus transfection) accordingto the manufacturerrsquos instructions Following optimisationcells were seeded at 35 times 104 cellswell in 24-well platesor 175 times 105 cellswell into 6-well plates After 24 hours

cells were transfected with 5 nM siRNA with 4120583LmL ofINTERFERin transfection agent RNA was collected 48 and72 hours after transfection and protein collected 72 hoursafter transfection In all experiments levels of knockdown byRNAi were assessed at the RNA and protein level by PCR andimmunoblotting

212 RNA Isolation and PCR Total RNA from culturedcells was extracted with RNeasy spin columns (Qiagen) orthe SV Total RNA Isolation System (Promega) accordingto manufacturerrsquos instructions One 120583g of total RNA wasreverse transcribed using oligo (dT) primers and SuperscriptII Reverse Transcriptase (RT) (Invitrogen Life Technologies)following manufacturerrsquos instructions Conventional PCRwas performed using PCR master mix (Promega) qPCR wasperformed using TaqMan gene expression assays (Life Tech-nologies) according to the manufacturerrsquos instructions Val-idated TaqMan probes were selected to target specific genesas follows SAFB1 (Hs01561652 gl) SAFB2 (Hs01006796 g1)ITGB4 (Hs00236216 m1) SHF (Hs00403125 m1) MALAT-1 (Hs00273907 s1) and 120573-actin (Hs99999903 m1) Dataanalysis was performed using the comparative Ct methodnormalised against 120573-actin expression Experiments wereperformed in triplicate and statistical analysis was performedusing Studentrsquos 119905-test or Repeated Measures ANOVA withDunnettrsquos Multiple Comparison Test All effects at 119875 lt 005are reported as significant

22 RNA Immunoprecipitation MCF-7 cells were washed inice cold PBS and then collected in lysis buffer (10mM Tris-HCl (pH 75) 150mM NaCl 05 NP-40 and 1 TritonX-100) containing protease inhibitors (Sigma-Aldrich) andRNase OUT (Invitrogen) Equal amounts of cell lysate werethen incubated for 3 hours at 4∘C with 2120583g of either rabbitanti-SAFB1 antibody (Genetex) or rabbit IgG control (SantaCruz) Cell lysates and antibodies were then incubated withDynabeads Protein G (Invitrogen) for a further hour at 4∘CBeads were then washed five times with lysis buffer beforethe immunoprecipitated RNA was collected in Trizol reagent(Invitrogen) according to manufacturerrsquos instructions TheRT-PCR was performed as before but half of the RNAobtained from the immunoprecipitation was used in eachreaction Fold enrichment of target mRNA was determinedafter normalization to the input and rabbit IgG controls

3 Results

31 Identification of RNA-Binding Sites for SAFB1 in BreastCancer Cells Although the role of SAFB proteins in RNAprocessing has been speculated the function of their highlyhomologous internal RRM has not been examined Wesought to identify possible direct RNA targets for SAFB1 inbreast cancer cells using iCLIP technology [9ndash12] combinedwith high-throughput sequencing and mapping to generate atranscriptome-wide binding map for SAFB1 SAFB1 protein-RNA complexes were successfully generated by immuno-precipitation and RNA recovered and purified from 3 inde-pendent iCLIP replicates (Supplementary Figure 1) High-throughput sequencing and bioinformatics generated a total



ncRNAIntergenicORF

Intron3UTR5UTR

(a)


ncRNAIntergenicORF

Intron3UTR5UTR

(b)


ncRNA subclasses

snRN

A

Mt R

NA

snoR

NA

rRN

A

miR

NA

lincR

NA

Oth

er R

NA

0

7

14

ncRN

A (

)

SAFB1 iCLIP

(c)






Pentamer


3464431201305333007229955298572898826496261442585025815256532552225514254502522724147239172366023323

Frequency

(a)

1 2 3 4 5

5998400 3998400

weblogoberkeleyedu



(b)

Base position

Base

A

C

G

T

1 2 3 4 5

65 60 60 75 80

15 15 20 10 10

10 15 15 15 5

10 10 5 0 5

(c)







0

02

04

06

08

1

12

14

16


Relat

ive g

ene e

xpre

ssio

n

Relat

ive g

ene e

xpre

ssio

n

SAFB2

002040608

112141618


SAFB1

(a)

ctrl

SAFB

1

SAFB

2

SAFB

1+SA

FB2

SAFB1

SAFB2

175

175

42

(kDa)

120573-actin

(b)

0

05

1

15

2

SHF

relat

ive e

xpre

ssio

n (fo

ld ch

ange

)

MCF-7MDA-MB-231


lowast

lowast

(c)

0

05

1

15

2

25

ITG

B4 re

lativ

e exp

ress

ion

(fold

chan

ge)

MCF-7MDA-MB-231


lowast

lowast

(d)







20

15

10

05

00

MALAT1lowastlowast

Relat

ive m

RNA

expr

essio

n


(a)

1

inpu

t

IgG

SAFB

1Ab

1

inpu

t (RT

minus)

IgG

(RTminus

)

SAFB

1Ab

(RTminus

)

300200100

100

bp la

dder

MAL

AT1

(b)

000

100

200

300

SAFB1 AbIgG

Fold

enric

hmen

t

(c)

SAFB1

SRSF2

DAPI

DAPI

(d)



4 Discussion

















Acknowledgments



References
















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology



ncRNAIntergenicORF

Intron3UTR5UTR

(a)


ncRNAIntergenicORF

Intron3UTR5UTR

(b)


ncRNA subclasses

snRN

A

Mt R

NA

snoR

NA

rRN

A

miR

NA

lincR

NA

Oth

er R

NA

0

7

14

ncRN

A (

)

SAFB1 iCLIP

(c)






Pentamer


3464431201305333007229955298572898826496261442585025815256532552225514254502522724147239172366023323

Frequency

(a)

1 2 3 4 5

5998400 3998400

weblogoberkeleyedu



(b)

Base position

Base

A

C

G

T

1 2 3 4 5

65 60 60 75 80

15 15 20 10 10

10 15 15 15 5

10 10 5 0 5

(c)







0

02

04

06

08

1

12

14

16


Relat

ive g

ene e

xpre

ssio

n

Relat

ive g

ene e

xpre

ssio

n

SAFB2

002040608

112141618


SAFB1

(a)

ctrl

SAFB

1

SAFB

2

SAFB

1+SA

FB2

SAFB1

SAFB2

175

175

42

(kDa)

120573-actin

(b)

0

05

1

15

2

SHF

relat

ive e

xpre

ssio

n (fo

ld ch

ange

)

MCF-7MDA-MB-231


lowast

lowast

(c)

0

05

1

15

2

25

ITG

B4 re

lativ

e exp

ress

ion

(fold

chan

ge)

MCF-7MDA-MB-231


lowast

lowast

(d)







20

15

10

05

00

MALAT1lowastlowast

Relat

ive m

RNA

expr

essio

n


(a)

1

inpu

t

IgG

SAFB

1Ab

1

inpu

t (RT

minus)

IgG

(RTminus

)

SAFB

1Ab

(RTminus

)

300200100

100

bp la

dder

MAL

AT1

(b)

000

100

200

300

SAFB1 AbIgG

Fold

enric

hmen

t

(c)

SAFB1

SRSF2

DAPI

DAPI

(d)



4 Discussion

















Acknowledgments



References
















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


Pentamer


3464431201305333007229955298572898826496261442585025815256532552225514254502522724147239172366023323

Frequency

(a)

1 2 3 4 5

5998400 3998400

weblogoberkeleyedu



(b)

Base position

Base

A

C

G

T

1 2 3 4 5

65 60 60 75 80

15 15 20 10 10

10 15 15 15 5

10 10 5 0 5

(c)







0

02

04

06

08

1

12

14

16


Relat

ive g

ene e

xpre

ssio

n

Relat

ive g

ene e

xpre

ssio

n

SAFB2

002040608

112141618


SAFB1

(a)

ctrl

SAFB

1

SAFB

2

SAFB

1+SA

FB2

SAFB1

SAFB2

175

175

42

(kDa)

120573-actin

(b)

0

05

1

15

2

SHF

relat

ive e

xpre

ssio

n (fo

ld ch

ange

)

MCF-7MDA-MB-231


lowast

lowast

(c)

0

05

1

15

2

25

ITG

B4 re

lativ

e exp

ress

ion

(fold

chan

ge)

MCF-7MDA-MB-231


lowast

lowast

(d)







20

15

10

05

00

MALAT1lowastlowast

Relat

ive m

RNA

expr

essio

n


(a)

1

inpu

t

IgG

SAFB

1Ab

1

inpu

t (RT

minus)

IgG

(RTminus

)

SAFB

1Ab

(RTminus

)

300200100

100

bp la

dder

MAL

AT1

(b)

000

100

200

300

SAFB1 AbIgG

Fold

enric

hmen

t

(c)

SAFB1

SRSF2

DAPI

DAPI

(d)



4 Discussion

















Acknowledgments



References
















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


0

02

04

06

08

1

12

14

16


Relat

ive g

ene e

xpre

ssio

n

Relat

ive g

ene e

xpre

ssio

n

SAFB2

002040608

112141618


SAFB1

(a)

ctrl

SAFB

1

SAFB

2

SAFB

1+SA

FB2

SAFB1

SAFB2

175

175

42

(kDa)

120573-actin

(b)

0

05

1

15

2

SHF

relat

ive e

xpre

ssio

n (fo

ld ch

ange

)

MCF-7MDA-MB-231


lowast

lowast

(c)

0

05

1

15

2

25

ITG

B4 re

lativ

e exp

ress

ion

(fold

chan

ge)

MCF-7MDA-MB-231


lowast

lowast

(d)







20

15

10

05

00

MALAT1lowastlowast

Relat

ive m

RNA

expr

essio

n


(a)

1

inpu

t

IgG

SAFB

1Ab

1

inpu

t (RT

minus)

IgG

(RTminus

)

SAFB

1Ab

(RTminus

)

300200100

100

bp la

dder

MAL

AT1

(b)

000

100

200

300

SAFB1 AbIgG

Fold

enric

hmen

t

(c)

SAFB1

SRSF2

DAPI

DAPI

(d)



4 Discussion

















Acknowledgments



References
















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


20

15

10

05

00

MALAT1lowastlowast

Relat

ive m

RNA

expr

essio

n


(a)

1

inpu

t

IgG

SAFB

1Ab

1

inpu

t (RT

minus)

IgG

(RTminus

)

SAFB

1Ab

(RTminus

)

300200100

100

bp la

dder

MAL

AT1

(b)

000

100

200

300

SAFB1 AbIgG

Fold

enric

hmen

t

(c)

SAFB1

SRSF2

DAPI

DAPI

(d)



4 Discussion

















Acknowledgments



References
















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology













Acknowledgments



References
















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


References
















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology















Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

research article unravelling the rna-binding properties...

Documents