Identification of Protein-RNA and Protein-Protein Interactions by the Neuronal HuC Protein of Mus musculus

Thesis submitted in part fulfilment of the requirements for the Degree of M.Sc. in the Discipline of Biochemistry, University of Adelaide

Bradley Simpson, B.Sc. (Mol. Biol.), B.Sc. (Hons.) March 2012

1

Table of Contents: Statement: .......................................................................................................................................... i Abstract: ............................................................................................................................................. ii Abbreviations: ................................................................................................................................. iv

Chapter 1 – Introduction: ........................................................................................ 1 1.1 - Post-Transcriptional Gene Regulation:......................................................................... 2 1.2 – The Hu Family of RNA-Binding Proteins: .................................................................... 3

1.2.1 – Identification and Expression: ................................................................................. 3 1.2.2 – Conserved Domain Structure: ................................................................................. 5 1.2.3 - HuA and Post-Transcriptional Gene Regulation: ........................................... 10 1.2.4 – Neuronal Hu – Role in Neuronal Differentiation:.......................................... 10 1.2.5 – Identifying Putative Neuronal Hu Targets – The CLIP Method: ............... 14 1.2.6 – Protein-Protein Interactions in Post-Transcriptional Gene Regulation: 17 1.3 – Aims & Approaches: ......................................................................................................... 20

1.3.1 – Understanding the Role of the Hu Proteins in Neuronal Differentiation:..................................................................................................... 20 1.3.2 – Identification of Neuronal Hu RNA Targets: ................................................... 20 1.2.3 – Identification of Neuronal Hu Protein Interactors: ....................................... 21

Chapter 2 – Materials & Methods: ....................................................................... 24 2.1 – Materials: .............................................................................................................................. 25

2.1.1 – Bacterial Strains: ........................................................................................................ 25 2.1.2 – Yeast Strains: .............................................................................................................. 25 2

2.1.3 – Plasmid Constructs: .................................................................................................. 26 2.1.4 – Primers: ......................................................................................................................... 27 2.1.5 – Enzymes: ...................................................................................................................... 30 2.1.6 – Antibodies/Antisera: ................................................................................................ 31 2.1.7 – Bacterial Growth Media: ......................................................................................... 32 2.1.9 – Yeast Growth Media: ............................................................................................... 32 2.1.9 – Chemicals & Reagents: ........................................................................................... 33 2.1.10 – Buffers & Solutions: ............................................................................................... 34 2.1.11 – Proprietary Buffers: ................................................................................................ 38 2.1.12 – Proprietary Kits: ....................................................................................................... 38 2.1.13 – DNA Markers: .......................................................................................................... 38 2.1.14 – Protein Markers:...................................................................................................... 38 2.1.15 – Additional Equipment: .......................................................................................... 39 2.2 – Experimental Methods: .................................................................................................... 40

2.2.1 – General Bacterial Procedures: .............................................................................. 40 2.2.2 – General Yeast Procedures: ..................................................................................... 41 2.2.3 – DNA Manipulation: .................................................................................................. 42 2.2.4 – PCR Techniques: ........................................................................................................ 43 2.2.5 – Protein Manipulation:.............................................................................................. 44 2.2.6 – Protein Purification:.................................................................................................. 48 2.2.7 – RNA Manipulation: ................................................................................................... 49 2.2.8 – Co-Immunoprecipitation: ...................................................................................... 52

3

Chapter 3 – Results: ................................................................................................ 55 3.1 – Cloning of pET41a(+)-Strep-mHuCsv1-His 8 : ........................................................... 56

3.1.1 – Selection of Purification Process for Recombinant mHuCsv1: ................. 56 3.1.2 – Replacement of GST Coding Sequence With the Strep Tag ORF: .......... 57 3.2 – Expression and Isolation of Recombinant mHuCsv1: ........................................... 60

3.2.1 – Purification Under Standard Conditions Results in mHuCsv1 Aggregation: ........................................................................................................ 60 3.2.2 – Purification in the Absence of Protease Inhibitors Significantly Reduces Strep-mHuCsv1-His 8 Protein Aggregation: ............................ 64 3.3 – Investigation of Recombinant mHuCsv1 RNA-Binding Activity: ...................... 66

3.3.1 – Confirmation of Recombinant mHuCsv1 RNA-Binding Activity: ............. 66 3.3.2 – CLIP Tag RNA Sequences Do Not Necessarily Represent Sites of Neuronal HuCsv1 Binding:......................................................................... 72 3.3.3 – Neuronal HuCsv1 Preferentially Binds RNA Sequences Possessing Multiple Uridine-Rich Sequence Motifs: .................................................... 85 3.4 – Identification of Neuronal Hu Protein Interactors: ................................................ 89

3.4.1 – Investigation of Neuronal Hu Protein Interactors Previously Identified by Yeast Two-Hybrid: ................................................................... 89 3.4.2 – Co-Immunoprecipitation Fails to Confirm Yeast Two-Hybrid Identified HuC Protein-Protein Interactions: ............................................ 95 3.4.3 – Putative Hu Protein Interactors Fail to Support Growth Under Stringent Selection Conditions: ................................................................... 101

4

Chapter 4 – Discussion: ........................................................................................ 104 4.1 – Expression and Purification of Recombinant mHuCsv1: .................................. 105 4.2 – Investigation of RNA-Binding by Recombinant HuCsv1: ................................. 107 4.3 – Identification of Neuronal Hu Protein Interactors: ............................................. 110 4.4 – Conclusions and Future Directions: ......................................................................... 113

Chapter 5 – Bibliography: .................................................................................... 115

5

Statement: This work contains no material which has been accepted for the award of any other degree or diploma in any university or other tertiary institution to Bradley Simpson and, to the best of my knowledge and belief, contains no material previously published or written by another person, except where due reference has been made in the text.

I give consent to this copy of my thesis, when deposited in the University Library, being made available for loan and photocopying, subject to the provisions of the Copyright Act 1968. I also give permission for the digital version of my thesis to be made available on the web, via the University’s digital research repository, the Library catalogue, the Australasian Digital Theses Program (ADTP) and also through web search engines, unless permission has been granted by the University to restrict access for a period of time.

Bradley Simpson

1i

Abstract: Post-transcriptional gene regulation is an essential process by which all vertebrate organisms regulate gene expression across a wide array of cell types.

Such

regulation is of particular importance within neurons, as the highly polarised nature of such cells requires that gene expression be tightly regulated in a spatiotemporal manner throughout both proximal and distal regions of the cell. RNAbinding proteins have been shown to regulate a wide range of post-transcriptional regulatory processes, and one family of vertebrate RNA-binding proteins, the Hu family, has been implicated in neuronal specification and differentiation. Of the four proteins in the Hu family, two (HuC and HuD) are uniquely neuronal, whilst one (HuB) is expressed both in neurons and the early embryo. The fourth (HuA) is expressed ubiquitously. Each of the four Hu proteins binds RNA in a sequencespecific manner, mediated by three RNA recognition motifs (RRMs) present in each family member, and are believed to regulate mRNA translation, localisation, and/or message stability. However, little information is currently available regarding the molecular mechanisms by which the neuronal Hu proteins mediate their effects. Thus, whilst HuA has been demonstrated to bind to AU-rich elements (AREs) in the 3’UTRs of a number of labile mRNA messages, little information is currently available regarding the specific RNA sequences that mediate RNA binding by the neuronal Hu proteins. Furthermore, a number of studies have demonstrated that the function of HuA relies upon a number of direct protein-protein interactions. However, to date few interactions have been identified for the neuronal Hu proteins. Identification of the RNA sequences bound by the neuronal Hu proteins and examination of the protein cofactors that mediate the function of these proteins in respect to such bound messages is essential in order to develop an understanding, both of the role played by the neuronal Hu proteins during development, and of the mechanisms by which these functions are achieved.

ii2

Herein we describe the development of a protocol for the efficient expression and purification of active recombinant HuC, in addition to the subsequent application of this protein to an investigation of the mRNA sequence motifs that mediate RNA binding by this neuronal Hu family member. Moreover, we demonstrate that HuC preferentially binds to RNAs containing short, interspersed uridine-rich sequence motifs in vitro. Additionally, we present evidence which suggests that Hu-bound RNA sequences identified by the Cross-Linking and Immunoprecipitation (CLIP) method do not necessarily represent sites of direct HuC binding, and thus that further study is required in order to identify the complement of mRNAs bound by each neuronal Hu protein.

Furthermore, we describe the identification of a

number of putative HuC protein-protein interactors by means of yeast two-hybrid analysis, in addition to the development of a protocol for the specific immunoprecipitation of the neuronal Hu proteins. This protocol may be applied both to the confirmation of putative protein interaction partners, and the identification of further protein interactors that mediate the function of the neuronal Hu proteins during neuronal development.

Together, these studies

represent the necessary first steps toward a complete understanding of the molecular mechanisms that underlie the function of these proteins, and provide an essential foundation for further studies on the role of the neuronal Hu family proteins during neuronal development.

3 iii

Abbreviations: AbA

Aureobasidin A

APS

ammonium persulfate

ARE

AU-rich element

ATP

adenosine triphosphate

bp

base pair(s)

ccpm

Cerenkov counts per minute

cDNA

complementary deoxyribonucleic acid

CIP

calf intestinal phosphatise

CLIP

cross-linking and immunoprecipitation

CNS

central nervous system

CPI

complete protease inhibitor

cpm

counts per minute

Cy3

cyanine3

Cy5

cyanine5

DAPI

4’,6-damidino-2-phenylindole

DNA

deoxyribonucleic acid

dNTP

deoxynucleotide triphosphate

DTT

dithiothreitol

ECL

enhanced chemiluminescence

EDTA

ethylene diamine tetraacetic acid

FITC

fluorescein isothiocyanate

GFP

green fluorescent protein

GST

glutathione-S-transferase

HEPES

4-2-hydroxyethyl-1-piperazineethanesulfonic acid

His n

n-histidine polypeptide

hpf

hours post-fertilisation

HRP

horseradish peroxidise 4 iv

IEF

isoelectric focusing

IgG

immunoglobulin G

kb

kilobase(s)

Kd

dissociation constant

kDa

kiloDalton(s)

mHuCsv1

murine HuC Splice Variant 1

min

minute(s)

MOPS

3-(N-morpholino)propanesulfonic acid

mRNA

messenger ribonucleic acid

NHS

normal human serum

nHu

neuronal Hu protein(s)

Ni-IDA

nickel-iminodiacetic acid

NP-40

nonidet P40

nt

nucleotide(s)

NTP

nucleotide triphosphate

OD 600

optical density at 600 nm

ORF

open reading frame

PAGE

polyacrylamide gel electrophoresis

PBS

phosphate buffered saline

PBST

phosphate buffered saline with 0.1% Tween-20

PCR

polymerase chain reaction

pI

isoelectric point

PND

paraneoplastic neurological disorder

PNK

polynucleotide kinase

PNS

peripheral nervous system

PVDF

polyvinylidene fluoride

RNA

ribonucleic acid

RNP

ribonucleoprotein

RRM

RNA recognition motif 5v

SDS

sodium dodecyl sulphate

sv

splice variant

TBE

tris-borate EDTA buffer

TEMED

tetramethylethylenediamine

TEV

TEV protease site

UTR

untranslated region

UV

ultraviolet

V

volt(s)

Standard one-letter abbreviations for nucleotides and three-letter abbreviations for amino acids are also used.

6 vi

Chapter 1 – Introduction:

1

1.1 - Post-Transcriptional Gene Regulation: The flow of genetic information from DNA to RNA and finally to functional protein forms the foundation for the replication and survival of all known organisms, and has over the last fifty years underpinned the study of all biological processes. Whilst it has previously been assumed that DNA transcription is the primary step at which gene expression is regulated, more recent research has unequivocally demonstrated that such regulation can occur at every step in the pathway from DNA to protein. Thus, in addition to transcription it is now clear that the processes of pre-mRNA capping, splicing, and polyadenylation and mRNA nuclear export, sub-cellular localisation, translation, and degradation are all stringently controlled by the cell (for a review see (Glisovic, Bachorik, Yong, & Dreyfuss, 2008)). Whilst such regulation is vital to the survival of all cells, it is of particular importance in a neuronal context, as the highly polarised nature of the neuron requires unique regulatory systems that can accurately direct the localisation and expression of a multitude of mRNAs encoding proteins responsible for directing both the extension and guidance of axons during development (Campbell & Holt, 2001; Yao, Sasaki, Wen, Bassell, & Zheng, 2006), and the ongoing synaptic function of mature neurons (Leung et al., 2006; Miller et al., 2002; Steward & Schuman, 2003). The principal effectors of this regulation are RNA-binding proteins (RBPs), which bind to RNA in either a specific or non-specific manner, and mediate their regulatory effects through a diverse array of molecular mechanisms (reviewed in (Gebauer & Hentze, 2004)). One family of RBPs, the Hu family, has been strongly implicated in the specification and differentiation of neurons during embryonic development, and evidence from an increasing number of studies suggests a role for this family in the post-transcriptional regulation of genes that directly effect neuronal development.

2

1.2 – The Hu Family of RNA-Binding Proteins: 1.2.1 – Identification and Expression: The Hu family consists of four genes in mammals: HuA (also termed HuR), HuB, HuC, and HuD, each of which produces a multitude of RNAs, encoding a wide array of Hu protein variants. Every member of the family shares a high degree of sequence homology with the ELAV (embryonic lethal abnormal vision) protein of

Drosophila melanogaster (Figure 1.1), the expression of which is essential for the proper differentiation and maintenance of neurons in the fly (Robinow & White, 1988).

Correspondingly, three of the Hu proteins, HuB, HuC, and HuD,

demonstrate uniquely neuronal patterns of expression.

The fourth, HuA, is

expressed within all adult tissues (Okano & Darnell, 1997). Additionally, recent work within our laboratory has demonstrated that HuB is also expressed in murine embryonic stem cells (Wiszniak, 2010). The neuronal Hu proteins (HuB, HuC, and HuD) were initially identified as target antigens in the paraneoplastic neurological disorder (PND), Hu Syndrome (King, 1994; Levine, Gao, King, Andrews, & Keene, 1993; Okano & Darnell, 1997; Szabo et al., 1991). PNDs occur when the cells of a malignant tumour (most often small-cell lung carcinoma in Hu Syndrome) ectopically express an antigen that is ordinarily restricted to immune-privileged cells of the central and peripheral nervous systems. This results in an auto-immune response that often leads to regression of the tumour, but also to severe neurodegeneration that manifests itself through multiple, varied neurological symptoms that reflect the particular antigen being targeted (reviewed in (Darnell & Posner, 2003)). Due to the autoimmune nature of these disorders, antisera from Hu syndrome patients harbour exceptionally high titre antibodies that specifically react with the neuronal Hu proteins, and do not react with the ubiquitous HuA protein.

3

elav zHuA zHuC mHuA mHuB mHuC mHuD

MDFIMANTGAGGGVDTQAQLMQSAAAAAAVAATNAAAAPVQNAAAVAAAAQLQQQQVQQAILQVQQQQTQQAVAAAAAAVTQQLQQQQQAVVAQQAVVQQ 100 ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------M 1

elav zHuA zHuC mHuA mHuB mHuC mHuD

QQQQAAAVVQQAAVQQAVVPQPQQAQPNTNGNAGSGSQNGSNGSTETRTNLIVNYLPQTMTEDEIRSLFSSVGEIESVKLIRDKSQVYIDPLNPQAPSKG -------------MSNGYEDH----------------MADELID--SKTNLIVNYLPQNMSQDELRSLFSSIGEVESAKLIRDK-------------VAG -MVT-IISTMETQVSNGPSGTSLPNGP---------VISTNGATDDSKTNLIVNYLPQNMTQEEFKSLFGSIGEIESCKLVRDK-------------ITG -------------MSNGYEDH----------------MAEDCRDDIGRTNLIVNYLPQNMTQEELRSLFSSIGEVESAKLIRDK-------------VAG ---------METQLSNGPTCNNTANGPTTVNNNCSSPVDSGN-TEDSKTNLIVNYLPQNMTQEELKSLFGSIGEIESCKLVRDK-------------ITG -MVTQILGAMESQVGGGPAGPALPNGP---------LLGTNGATDDSKTNLIVNYLPQNMTQDEFKSLFGSIGDIESCKLVRDK-------------ITG RLQNQIISTMEPQVSNGPTSN-TSNGPSSNNRNCPSPMQTGAATDDSKTNLIVNYLPQNMTQEEFRSLFGSIGEIESCKLVRDK-------------ITG

elav zHuA zHuC mHuA mHuB mHuC mHuD

QSLGYGFVNYVRPQDAEQAVNVLNGLRLQNKTIKVSFARPSSDAIKGANLYVSGLPKTMTQQELEAIFAPFGAIITSRILQNAGNDTQTKGVGFIRFDKR HSLGYGFVNYVNPNDAERAISTLNGLRLQSKTIKVSYARPSSDSIKDANLYVSGLPKTMSQKEMEQLFSQYGRIITSRILLDQATG-VSRGVGFIRFDKR QSLGYGFVNYVDPNDADKAINTLNGLKLQTKTIKVSYARPSSASIRDANLYVSGLPKTMSQKDMEQLFSQYGRIITSRILVNQVTG-ISRGVGFIRFDKR HSLGYGFVNYVTAKDAERAISTLNGLRLQSKTIKVSYARPSSEVIKDANLYISGLPRTMTQKDVEDMFSRFGRIINSRVLVDQTTG-LSRGVAFIRFDKR QSLGYGFVNYIDPKDAEKAINTLNGLRLQTKTIKVSYARPSSASIRDANLYVSGLPKTMTQKELEQLFSQYGRIITSRILVDQVTG-ISRGVGFIRFDKR QSLGYGFVNYSDPNDADKAINTLNGLKLQTKTIKVSYARPSSASIRDANLYVSGLPKTMTQKELEQLFSQYGRIITSRILVDQVTG-VSRGVGFIRFDKR QSLGYGFVNYIDPKDAEKAINTLNGLRLQTKTIKVSYARPSSASIRDANLYVSGLPKTMTQKELEQLFSQYGRIITSRILVDQVTG-ISRGVGFIRFDKR

elav zHuA zHuC mHuA mHuB mHuC mHuD

EEATRAIIALNGTTPSSCTDPIVVKFSNTPGSTSKIIQPQLPAFLNPQLVRRIGGAMHTPVNKGLARFSPMAGDMLDVMLPNGLGAAAAAATTLASGPGG AEAEDAIKDLNGQKPPGAAEQMTVKFAASP---NQVKNTQVIPQVYHQQSRRFGGPVHHQAQRF--RFSPMS---VDHMSG------MSGVNVPG-NSSS NEAEEAIKGLNGQKPLGAAEPITVKFANNP---SQKTGQALLTQLYQTAARRYTGPLHHQTQRF--RFSPIT---IDSMTS------LAGVNLTG-PTGA SEAEEAITSFNGHKPPGSSEPITVKFAANP---NQNKNMALLSQLYHSPARRFGGPVHHQAQRF--RFSPMG---VDHMSG------ISGVNVPG-NASS IEAEEAIKGLNGQKPPGATEPITVKFANNP---SQKTNQAILSQLYQSPNRRYPGPLAQQAQRF--RFSPMT---IDGMTS------LAGINIPG-HPGT IEAEEAIKGLNGQKPLGAAEPITVKFANNP---SQKTGQALLTHLYQSSARRYAGPLHHQTQRF--RFSPIA---IDGMSG------LAGVGLSGGAAGA IEAEEAIKGLNGQKPSGATEPITVKFANNP---SQKSSQALLSQLYQSPNRRYPGPLHHQAQRF--RFSPIT---IDGMTS------LVGMNIPG-HTGT

elav zHuA zHuC mHuA mHuB mHuC mHuD

AYPIFIYNLAPETEEAALWQLFGPFGAVQSVKIVKDPTTNQCKGYGFVSMTNYDEAAMAIRALNGYTMGNRVLQVSFKTNKAK-GWCIFIYNLGQDADEGILWQMFGPFGAVTNVKVIRDFNTNKCKGFGFVTMTHYEEAAMAIASLNGYRLGDKILQVSFKTSKSHKGWCIFVYNLSPEADESVLWQLFGPFGAVTNVKVIRDFTTNKCKGFGFVTMTNYDEAAMAIASLNGYRLGDRVLQVSFKTSKQHKA GWCIFIYNLGQDADEGILWQMFGPFGAVTNVKVIRDFNTNKCKGFGFVTMTNYEEAAMAIASLNGYRLGDKILQVSFKTNKSHKGWCIFVYNLAPDADESILWQMFGPFGAVTNVKVIRDFNTNKCKGFGFVTMTNYDEAAMAIASLNGYRLGDRVLQVSFKTNKTHKA GWCIFVYNLSPEADESVLWQLFGPFGAVTNVKVIRDFTTNKCKGFGFVTMTNYDEAAMAIASLNGYRLGERVLQVSFKTSKQHKA GWCIFVYNLSPDSDESVLWQLFGPFGAVNNVKVIRDFNTNKCKGFGFVTMTNYDEAAMAIASLNGYRLGDRVLQVSFKTNKAHKS

RRM1

RRM1

200 56 76 58 77 77 87

RRM2 300 155 175 157 176 186 176

RRM2 400 240 260 242 262 262 271

RRM3 483 324 345 326 347 347 356

Figure 1.1 – Hu family proteins are orthologous to ELAV Multiple sequence alignment of two Hu proteins from Danio rerio (zHuA, zHuC) and all four Hu proteins of Mus musculus (mHuA-mHuD) with the Embryonic Lethal Abnormal Vision protein of Drosophila melanogaster reveals a high degree of sequence conservation. This conservation is most evident within the RNA-Recognition Motifs (RRMs) of each protein. Prepared using ClustalW.

4

These Hu patient antisera have proven to be invaluable in the study of the neuronal Hu proteins, having facilitated identification of the human neuronal Hu proteins HuB (Hel-N1) (Levine, et al., 1993), HuC (ple21) (Sakai et al., 1994), and HuD (Szabo, et al., 1991), as well as the corresponding murine homologues (mHuB, mHuC, and mHuD) (Okano & Darnell, 1997).

The HuC and HuD proteins of

zebrafish (zHuC and zHuD) have subsequently been cloned from cDNA (Park, Hong, et al., 2000), whilst the zebrafish HuB protein was identified bioinformatically within our laboratory (Webb, 2006). The Hu proteins of both mouse and zebrafish share similarly neuronal patterns of expression with their human and mouse orthologues, although the timing of expression varies between species. In addition to the four Hu genes identified in mammals, zebrafish also possess a fourth Hu gene, HuG, which is highly homologous to HuA, and demonstrates a correspondingly ubiquitous pattern of expression (Park, Hong, et al., 2000).

1.2.2 – Conserved Domain Structure: All proteins of the Hu family share a common, evolutionarily conserved domain structure, consisting of three globular RNA-binding domains, termed RNA recognition motifs (RRMs), each of which possess a very high degree of sequence conservation (Figure 1.1 & Figure 1.2A). RRMs are characterised by a βαββαβ secondary structure, in which the four beta strands form an antiparallel beta-sheet that contains two ribonucleoprotein (RNP) motifs, both of which are very highly conserved across species, and are responsible for the RNA-binding activity of each RRM (Wang & Tanaka Hall, 2001). Of the three RRMs in each Hu protein, two are arranged in tandem at the N-terminus of each protein, whilst the third lies at the distal C-terminus, and is linked to the first two by a long, seemingly unstructured spacer domain. In HuA, this spacer domain contains a polypeptide region termed the HuA Nucleocytoplasmic Shuttling sequence (HNS) (Figure 1.2A), which is conserved between all Hu family members and has been shown to directly mediate 5

both the nuclear import and export of HuA, and thus regulate the steady state sub-cellular localisation of this protein (Fan & Steitz, 1998).

Additionally, this

subdomain of HuA has also been shown to interact with a number of protein cofactors which are responsible for mediating this localisation (see section 1.2.6).

6

A.

HuA HuBsv4 HuCsv4 HuDsv4

RRM1

190 209 219 209

RRM2

HNS

RRM3

NKNMALLSQLYHSPARRFGGPVHHQAQRFRFSPMGVDHMSGISGVNVPGNASS KTNQAILSQLYQSPNRRYPGPLAQQAQRFRFSPMTIDGMTSLAGINIPGHPGT KTGQALLTHLYQSSARRYAGPLHHQTQRFRFSPIAIDGMSGLAGVGLSGGAAG KSSQALLSQLYQSPNRRYPGPLHHQAQRFRFSPITIDGMTSLVGMNIPGHTGT

245 264 273 263

B. HuA:

205 RRFGGPVHHQAQRFR---------------------------FSPMGVDHMSGISGVNVP

HuB: sv2: 224 RRYPGPLAQQAQRFRLDNLLNMAYGVKR--------------FSPMTIDGMTSLAGINIP sv4: 224 RRYPGPLAQQAQRFR---------------------------FSPMTIDGMTSLAGINIP HuC: sv1: sv2: sv3: sv4:

224 224 224 224

RRYAGPLHHQTQRFRLDNLLNMAYGVKS-------PLSLIARFSPIAIDGMSGLAGVGLS RRYAGPLHHQTQRFRLDNLLNMAYGVKR--------------FSPIAIDGMSGLAGVGLS RRYAGPLHHQTQRFR--------------------PLSLIARFSPIAIDGMSGLAGVGLS RRYAGPLHHQTQRFR---------------------------FSPIAIDGMSGLAGVGLS

HuD: sv1: 231 RRYPGPLHHQAQRFRLDNLLNMAYGVKRLMSGPVPPSACPPRFSPITIDGMTSLVGMNIP sv2: 231 RRYPGPLHHQAQRFRLDNLLNMAYGVKR--------------FSPITIDGMTSLVGMNIP sv4: 231 RRYPGPLHHQAQRFR---------------------------FSPITIDGMTSLVGMNIP

Figure 1.2 – Alternate splicing within the HNS region of the Hu proteins produces multiple splice variants with interrupted HNS-like sequences A. Every member of the Hu family shares a conserved domain structure, consisting of two RNA-recognition motifs (RRMs) located in tandem at the N-terminus, and linked to a third RRM by a long, unstructured spacer domain. In HuA, this spacer region contains the HuA Nucleocytoplasmic Shuttling sequence (HNS). Whilst the RRMs sequences are generally the most conserved of the Hu proteins, significant sequence conservation is also evident between the HNS region of HuA and at least one splice variant of each neuronal Hu protein. B. Alternative splicing within the HNS-like regions of the neuronal Hu proteins yields numerous splice variants that possess conserved polypeptide insertions within the HNS region.

7

The spacer domains of HuB, HuC, and HuD also possess a HNS-like polypeptide sequence that appears to confer similar nucleo-cytoplasmic shuttling activity upon the neuronal Hu proteins.

However, while at least one splice variant of each

neuronal Hu protein possesses a conserved, uninterrupted HNS-like sequence (Figure 1.2A), the spacer region of these neuronal family members is also subject to extensive alternative splicing, resulting in the production of numerous splice isoforms that possess polypeptide insertions of varying length within these HNSlike regions (Figure 1.2B) (Okano & Darnell, 1997). Transient transfection studies carried out in our lab using myc-tagged forms of each neuronal Hu protein and splice variant have demonstrated a clear relationship between alternative splicing in the HNS region of these proteins and the steady-state sub-cellular localisation of each resultant splice isoform (Hing, 2009). Specifically, splice variants that possess a conserved HNS region similar to that in HuA (the sv4 variants of each nHu protein) localise to the nucleus of transfected cells (Figure 1.3A), whilst those possessing polypeptide insertions within this domain demonstrate a greater degree of cytoplasmic localisation (Figure 1.3B-D).

We hypothesise that the presence of leucine-rich peptide

sequences within each HNS-like region act to enhance CRM1-mediate nuclear export of these isoforms, and thus promote steady-state cytoplasmic localisation. Taken together, these results suggest a conserved role for the HNS-like region in regulating the sub-cellular localisation of the neuronal Hu proteins. Moreover, the diversity of polypeptide sequences that result from alternative splicing with the HNS region suggest that each splice variant may interact with varying combinations of protein co-factors to direct the localisation of each isoform. Furthermore, recent studies have suggested that the spacer domain may also play a role in the specific recognition of RNA sequences (Fialcowitz-White et al., 2007), implying that alternate splice isoforms may bind to unique RNA sequences and thus effect the regulation of distinct subsets of mRNA messages during neuronal development.

8

A

B

HuA

C

D

Figure 1.3 – Alternate splicing within the HNS-like region of the neuronal Hu proteins yields splice variants with distinct sub-cellular localisation patterns: Expression in HeLa cell culture of myc-tagged forms of each murine HuCsv4 (A), HuCsv3 (B), HuCsv2 (C), and HuCsv1 (D) followed by immunofluorescent staining of myc-tagged Hu (red) and cell nuclei (DAPI, blue) reveals that alternate splicing within the HNS region has a profound impact upon the sub-cellular localisation of the resultant splice isoforms. Myctagged murine HuA (inset) is shown for comparison (Hing, 2009).

9

1.2.3 - HuA and Post-Transcriptional Gene Regulation: The ubiquitously-expressed HuA protein has been implicated in a diverse array of post-transcriptional regulatory processes. Numerous studies have established that HuA binds to AU-rich elements (AREs), RNA sequences present in the 3’UTRs of many highly labile mRNA transcripts that mediate the accelerated cytoplasmic turnover of these messages (reviewed in (Bolognani & Perrone-Bizzozero, 2008) and (Khabar, 2010)). Binding of HuA to ARE sequences within the 3’UTRs of the cfos and c-jun transcripts (Peng, Chen, Xu, & Shyu, 1998) results in stabilisation of these mRNAs (Brennan, Gallouzi, & Steitz, 2000), whilst binding to the tissue necrosis factor (TNF), Cyclooxygenase 2 (Cox2), and transforming growth factor β (TGFβ) mRNAs in macrophages is known to mediate translational downregulation of these messages during the inflammatory response (Katsanou et al., 2005). More recently, HuA has been shown to play a role in micro-RNA (miRNA)-mediated translational regulation, whereby binding of HuA to the 3’UTR of the Cationic Amino Acid Transporter 1 (CAT-1) mRNA during serum starvation of Huh7 cells results in relief of miR-122 mediated translational repression, and thus upregulation of CAT-1 expression (Bhattacharyya, Habermacher, Martine, Closs, & Filipowicz, 2006). These and other studies have established a clear role for HuA in post-transcriptional regulation of gene expression.

Given the high degree of

sequence homology between HuA and the neuronal Hu proteins, it seems likely that these proteins perform similar regulatory functions in a neuronal context.

1.2.4 – Neuronal Hu – Role in Neuronal Differentiation: Numerous lines of evidence suggest that the neuronal Hu proteins play a central role in regulating neuronal differentiation (for a review see (Pascale, Amadio, & Quattrone, 2008)).

Overexpression of the longest splice forms of any of the

neuronal Hu proteins in the neurogenic PC12 cell line results in the formation of 10

neurite-like processes of a morphology similar to that observed following treatment with nerve growth factor (NGF) (Kasashima, Sakashita, Saito, & Sakamoto, 2002). Additionally, ectopic introduction into the developing neural tube of E9.5 mouse embryos of mRNAs encoding either mHuB or mHuC results in a correspondingly ectopic expression of the neuronal marker neurofilament-M (Akamatsu et al., 1999). Furthermore, HuB has been shown to directly bind to the 3’UTR of mRNA encoding neurofilament-M, resulting in stabilisation of this transcript (Antic, Lu, & Keene, 1999) whilst mHuD similarly binds and stabilises the mRNAs encoding the neuronal Tau (Aranda-Abreu, Behar, Chung, Furneaux, & Ginzburg, 1999) and GAP43 proteins (Mobarak et al., 2000), and has recently been implicated in localized protein synthesis in developing dendrites (Tiruchinapalli, Ehlers, & Keene, 2008). Moreover, homozygous knockout mice deficient in HuD protein demonstrate delayed neurite outgrowth by ordinarily HuD-positive cranial nerves during embryonic development, whilst cultured neural stem cells isolated from knockout embryos possess increased proliferative capacity and decreased differentiation potential when compared to equivalent cells isolated from wildtype mice (Akamatsu et al., 2005). Consistent with these observations, more recent studies have demonstrated that in vivo ectopic overexpression of miR-375, in adult mouse hippocampus, results in a marked decrease in neurite outgrowth as a direct result of miR-375-mediated HuD downregulation (Abdelmohsen et al., 2010). Taken together, these studies strongly support a role for the neuronal Hu proteins in

the

processes

of

neuronal

specification

and

differentiation

during

embryogenesis. Consistent with this hypothesis, the HuC protein of Danio rerio is one of the earliest markers of neuronal specification in the developing zebrafish (Kim et al., 1996), and demonstrates a correspondingly pan-neuronal pattern of expression in the both the embryonic and adult nervous systems (Park, Kim, et al., 2000). By comparison, neuronal differentiation in mice is initially accompanied by expression of HuB and HuC, with upregulation of HuD occurring later in development (Okano & Darnell, 1997). Additionally, specific subsets of neurons 11

within the developing and adult nervous systems express unique combinations of HuB, HuC, and HuD, suggesting a potential combinatorial role for these proteins in the specification of distinct neural subregions during development.

Interestingly, previous work within our laboratory has demonstrated localisation of the neuronal Hu proteins to the growth cone of the extending axonal projection of cultured E18 mouse hippocampal neurons (Figure 1.4) (Jensen, 2004), whilst studies involving morpholino-based knockdown of zHuD in the developing Danio

rerio embryo have yielded a phenotype characterised by a significant delay in axon growth within a specific subset of motor neurons, the CaP primary motor neurons (Sibbons, 2007). These results are consistent with a potential role for the neuronal Hu proteins in post-transcriptional regulation during axon growth and guidance, a process known to rely upon localised regulation of transcription (Campbell & Holt, 2001; Leung, et al., 2006) (reviewed in (Holt & Bullock, 2009)). Taken together, this evidence supports a role for the neuronal Hu proteins in post-transcriptional gene regulation during neuronal development.

12

Figure 1.4 – Neuronal Hu proteins localise to the growth cone of extending axons: Immunofluorescent staining of neuronal Hu proteins (green) and acetylated α-tubulin (red) in E18 rat hippocampal neurons cultured for 3.5 days in vitro reveals that neuronal Hu localises to the growth cone of the extending axon (Jensen, 2004).

13

1.2.5 – Identifying Putative Neuronal Hu Targets – The CLIP Method: Using the Cross-Linking and Immunoprecipitation (CLIP) method, our lab has identified numerous in vivo mRNA targets of both the murine and zebrafish neuronal Hu proteins. The CLIP method was developed specifically to identify genuine in vivo targets of RNA-binding proteins, and has been successfully applied in the identification of RNA molecules bound by the sequence-specific RNAbinding protein Nova-1, a KH domain RNA binding protein that possesses a domain structure similar to that of the Hu proteins (Ule et al., 2003). Briefly, the CLIP experiment (Figure 1.5) involves UV-mediated covalent cross-linking of and RNA-binding protein of interest to bound RNA sequences in situ, followed by lysis of cross-linked tissue, and subsequent limiting RNase treatment to reduce the average length of crosslinked RNAs to approximately 50-60 nucleotides. Neuronal Hu protein-RNA complexes are then isolated by immunoprecipitation using nHuspecific PND patient antisera.

The protein-bound RNA thus isolated is

radiolabelled and ligated to specific 5’ and 3’ polynucleotide linker sequences. Resultant labelled protein-RNA complexes are then isolated from IP beads, separated from non-crosslinked RNAs by SDS-PAGE and remaining crosslinked protein-RNA transferred to a nitrocellulose filter.

Once isolated from the

nitrocellulose filter, the neuronal Hu in these complexes is removed by treatment with Proteinase K, and the resultant RNA “tags” are directionally cloned and sequenced.

Comparison of sequenced tags to mouse and zebrafish genome

databases allows for the identification both of mRNAs that possess neuronal Hu binding sites, and of the approximate location of such sites within each identified message.

14

Figure 1.5 – The Cross-Linking and Immunoprecipitation (CLIP) Method

15

The CLIP method has been applied in our laboratory using tissue isolated from both 2 hpf and 24 hpf zebrafish embryos.

At 2 hpf, the zebrafish embryo is

essentially undifferentiated, and thus express only HuB, whilst at 24 hpf the principal elements of the CNS have been established, and these embryos correspondingly express both HuC and HuD.

These CLIP experiments have

demonstrated that the neuronal Hu proteins of the zebrafish preferentially bind Urich RNA motifs, as indicated by the presence of long U-rich sequences within the majority of CLIP tag sequences identified to date (Dredge, 2005). By comparison, CLIP tags identified from previous experiments on murine E18 neural tissue more frequently possess C-rich RNA sequences, often with one or more U-rich sequences located adjacent to each identified CLIP tag within the parental mRNA message (Jensen, 2004). Recent RNA-binding studies carried out on mice overexpressing myc-tagged HuD have also identified a marked enrichment of Urich sequences in murine HuD mRNA targets, further supporting a role for these sequences in neuronal Hu binding (Bolognani, Contente-Cuomo, & PerroneBizzozero, 2010). Interestingly, the vast majority of putative neuronal Hu target sequences identified both in this study, and also from CLIP experiments performed within our laboratory, are located within the 3’UTR regions of assayed mRNAs. Given the well-described roles for 3’UTRs in numerous post-transcriptional gene regulatory processes (reviewed in (Gebauer & Hentze, 2004)), these results further support a role for the neuronal Hu proteins in such regulation. Whilst the CLIP method allows for the direct identification of genuine in vivo RNA targets of the neuronal Hu proteins, it is limited by the lack of available antibodies that specifically bind individual neuronal Hu proteins or splice isoforms. Thus, whilst the application of PND patient antisera allows for the identification of RNA regions bound specifically by the neuronal members of the Hu family, it does not provide detailed information regarding the specific neuronal Hu proteins or splice variants that bind to each identified message. One exception to this is the analysis 16

of early zebrafish embryos, as HuB is the only ‘neuronal’ Hu protein expressed during early embryogenesis. Additionally, whilst NMR and crystallographic studies have demonstrated that single Hu proteins bind to no more than 15 consecutive nucleotides on their RNA targets (Fialcowitz-White, et al., 2007; Wang & Tanaka Hall, 2001), CLIP tags must be on the order of 50-70 nucleotides in length in order to facilitate unambiguous bioinformatic identification.

Consequently, more

detailed biochemical studies are required, both to identify specific site of neuronal Hu binding, and also to elucidate the sequence specificity of each neuronal Hu protein and splice variant.

1.2.6 – Protein-Protein Interactions in Post-Transcriptional Gene Regulation: A multitude of studies have illustrated the importance of protein-protein interactions in the mechanisms underlying post-transcriptional gene regulation. Work on the Cytoplasmic Polyadenylation Binding Element protein (CPEB), which is responsible for repressing the translation of mRNAs possessing the Cytoplasmic Polyadenylation Element (CPE) sequence, have demonstrated that this protein mediates its effects through direct recruitment of the Maskin protein to target mRNAs (reviewed in (Mendez & Richter, 2001). Once recruited by CPEB, Maskin binds to the cap-binding eukaryotic translation initiation factor eIF4E, disrupting interaction with its partner initiation factor, eIF4G, and thus blocking recruitment of the 40S small ribosomal subunit, which in turn prevents formation of the 43S ribosomal scanning complex, and results in translational inhibition. Similarly, the Smaug protein of Drosophila mediates its translational repressor function by binding to the 3’UTR of nanos mRNA and subsequently recruiting the Cup protein. The recruited Cup then binds eIF4E, repressing its translational activator activity and thus preventing translation of the nanos mRNA (Nakamura, Sato, & HanyuNakamura, 2004). With regard to neurons, immunoprecipitation studies on the neuronal kinesin heavy chain, KIF5, have demonstrated that this protein interacts 17

with a vast array of RBPs and other proteins to form large RNP complexes which are then subject to kinesin-dependent directional transport within axons (Kanai, Dohmae, & Hirokawa, 2004). Interestingly, the binding of RBPs to transported messages may be responsible for nucleating these large complexes by recruiting other protein factors involved in translational repression and/or transport.

In addition to these and other studies underscoring the widespread role of protein-protein interactions in post-transcriptional gene regulation, several studies have begun to elucidate the roles that such interactions play in effecting the generegulatory functions of the Hu proteins.

As mentioned previously (1.2.2), a

number of protein factors have been identified that interact with the HNS/RRM region of the HuA protein, including the CRM-1 substrates pp32 and Acidic Protein Rich in Leucine (APRIL) (Brennan, et al., 2000), which are believed to be responsible for the nuclear export of HuA via the CRM1 nuclear export pathway, and Transportin 2 (Trn2), which has been shown to mediate import of HuA into the nucleus (Guttinger, Muhlhausser, Koller-Eichhorn, Brennecke, & Kutay, 2004). More recently, HuA has been shown to directly interact with the cell-cycle related RNA-binding protein RNPC1 on the 3’UTR of the p21 mRNA, resulting in stabilisation and translational upregulation of this transcript (Cho, Zhang, & Chen, 2010).

With regard to the neuronal Hu proteins, HuD has been shown to interact with the mRNA nuclear export factor TAP/NXF1 (Saito, Fujiwara, Katahira, Inoue, & Sakamoto, 2004), implying a potential role for HuD in mRNA export, and has more recently been observed to interact with the eukaryotic translation initiation factor, eIF4A, whereby the HNS/RRM region of HuD interacts directly with eIF4A to promote circularisation and translation of HuD-bound mRNAs (Fukao et al., 2009). Interestingly, mutation of the eIF4a binding site abolishes the neurite-inducing capacity of HuD, suggesting that this interaction may be necessary for HuD to 18

promote neuronal differentiation. Taken together, these studies have begun to illustrate the potential roles played by protein-protein interactions in neuronal Hu protein function.

Further research will allow us to identify additional protein-

protein interactions involving the neuronal Hu proteins, and will help elucidate the mechanisms underlying post-transcriptional gene regulation by these proteins during neuronal development.

19

1.3 – Aims & Approaches: 1.3.1 – Understanding the Role of the Hu Proteins in Neuronal Differentiation: A complete understanding of the roles of the Hu proteins in the processes of neuronal specification and differentiation requires a comprehensive study of the molecular mechanisms underlying post-transcriptional gene regulation by the neuronal members of the Hu family. Given the wide array of research describing the importance of both RNA-protein and protein-protein interactions in HuA function, in addition to a small number of studies that have begun to investigate the roles of these processes with respect to the neuronal Hu proteins, it seems likely that such interactions form the basis for nHu function during neuronal specification and differentiation. However, little published data exists regarding the roles of these mechanisms in neuronal Hu function. Consequently, further investigation of both the RNA-protein and protein-protein interactions undertaken by the neuronal Hu proteins in vivo is required before we can fully understand the function of these proteins during development. To this end, the primary aims of this project have been to identify both the specific RNA sequences that are bound by the neuronal Hu proteins, and the protein cofactors that interact with these proteins to mediate their function in vivo.

1.3.2 – Identification of Neuronal Hu RNA Targets: In order to begin elucidating the sequence-specificity of nHu RNA-binding, we have selected a number of CLIP-identified nHu RNA-targets from both mice and zebrafish, and investigated the binding of neuronal Hu to specific subregions within and adjacent to these CLIP-tag sequences by application of in vitro nitrocellulose filter-binding assays.

This allows for the identification of RNA

sequences within each CLIP tag that bind nHu with high affinity, and are thus likely sites of direct nHu binding in vivo. 20

For these experiments, we have chosen to investigate RNA-binding by murine HuCsv1, as previous studies have demonstrated the widespread expression and functional importance of HuC during neuronal development (1.2.4). Furthermore, HuCsv1 demonstrates the greatest degree of cytoplasmic localisation of any HuC splice variant (1.2.2), and it thus seems likely that this splice variant is one of those present at the growth cone of the extending axon during neuronal differentiation (1.2.4). Consequently, a complete understanding of the RNA-binding properties of HuCsv1 will likely yield a significant increase in our understanding of the role of the neuronal Hu proteins during neuronal development.

As discussed previously (1.2.2), both zebrafish and murine HuC share a very high degree of sequence conservation, with 94.8% sequence identity between the two proteins, concentrated within the RRM domains, which mediate RNA binding. As such, it is expected that both proteins are likely to interact with RNA sequences of similar sequence composition.

Consequently, we have applied recombinant

mHuCsv1 to the analysis of binding to RNA sequences identified from CLIP experiments performed on both 512-cell and 24 hpf zebrafish embryos, as well as E18 murine neural tissue.

1.2.3 – Identification of Neuronal Hu Protein Interactors: In addition to studying the RNA-protein interactions of the neuronal Hu proteins, we have also chosen to investigate the protein-protein interactions formed by these proteins. This has been undertaken through a combination of yeast twohybrid studies and neuronal Hu co-immunoprecipitation, which allows us to identify putative nHu interactors in vitro and to then examine these interactions within a more physiologically relevant context.

21

We have selected two bait proteins for two-hybrid analysis, consisting of either the spacer region of HuCsv2 alone, or of the spacer region together with RRM3. As previous studies have demonstrated the inability of the Hu proteins to bind RNA in the absence of the first two RRMs (Fialcowitz-White, et al., 2007), the application of these two baits reduces the likelihood of identifying artificial ‘through-RNA’ interactions that can occur between RNA-binding proteins within a two-hybrid system. mHuCsv2 has been selected for these experiments, as HuC is the most widely-expressed of all neuronal Hu proteins, whilst the spacer region of splice variant 2 demonstrates considerable sequence divergence from HuA within the HNS region (Fig 1.2B). It is thus likely that HuCsv2 interacts with a complement of protein co-factors distinct from those already identified for HuA (1.2.6), and hence yield a significant increase in our understanding of Hu protein function. However, this approach is applicable to all neuronal Hu proteins and splice variants, and future investigation utilising other members of the Hu family will allow for identification of protein-protein interactions unique to each neuronal Hu protein and splice variant. Whilst the yeast two-hybrid experiment allows for the identification of multiple interactions specific to individual Hu proteins and splice variants, the somewhat artificial nature of this system requires that each interaction be confirmed in a more physiological context.

To this end, we have applied neuronal Hu co-

immunoprecipitation from isolated murine neural tissue in order to further examine each putative two-hybrid interaction. patient

antisera

in

these

experiments,

By utilising nHu-specific PND

it

is

possible

to

selectively

immunoprecipitate the neuronal Hu proteins, and thus minimize the likelihood of identifying cofactors that interact only with the ubiquitous HuA protein. Following neuronal

Hu

immunoprecipitation,

examination

of

co-immunoprecipitated

proteins by western blot allows us to investigate the presence or absence of specific putative protein interactors identified by yeast two-hybrid, and thus 22

investigate whether these interactions are likely to be occurring in vivo. Thus, by combining these two techniques we can potentially identify a wide array of putative neuronal Hu interactors in vitro, and subsequently confirmation of each of these interactions in vivo.

23

Chapter 2 – Materials & Methods:

24

2.1 – Materials: 2.1.1 – Bacterial Strains: Strain

Genotype

Application (Source)

E. Coli B F- dcm ompT hsdS(r B - m B -) gal λ(DE3)

BL21(DE3)pLysS

Expression

of

recombinant

proteins encoded by vectors

r

possessing an upstream lac or

[pLysS Cam ]

T7lac promoter including pET,

pGEX (Stratagene) -

-

-

+

E. Coli B F ompT hsdS(r B m B ) dcm Tetr gal λ(DE3) endA The [argU proL Camr][argU ileY leuW Strep/Specr]

BL21(DE3) RIPL

Expression

of

recombinant

proteins encoded by vectors possessing an upstream lac or

T7lac promoter incl. pET, pGEX (Stratagene) espescially those containing high-frequency rare codons

F- φ80lacZ∆(lacZYA-argF)U169 recA1

DH5α

-

+

endA1 hsdR17(r k m k ) phoA supE44 thi-1 gyrA96 relA1 tonA

Routine cloning of plasmid DNA constructs

2.1.2 – Yeast Strains: Strain AH109

Genotype

Application (Source)

MATa, trp1-901, leu2-3, 112,

Expression and screening of

ura3-52, his3-200, gal4Δ, gal80Δ,

selected cDNA libraries by yeast

LYS2::GAL1 UAS -GAL1 TATA -HIS3, GAL2 UAS -GAL2 TATA -ADE2,

two-hybrid.

Allows for typical

nutrient-based selection.

URA3::MEL1 UAS -MEL1 TATA -lacZ, MEL Matchmaker Gold

MATa, trp1-901, leu2-3, 112,

Expression and screening of

ura3-52, his3-200, gal4Δ, gal80Δ,

selected cDNA libraries by yeast

LYS2::GAL1 UAS –Gal1 TATA –His3,

two-hybrid.

Allows for both

GAL2 UAS –Gal2 TATA –Ade2,

nutrient- and Aureobasidin A-

URA3::MEL1 UAS –Mel1 TATA ,

based selection.

AUR1-C MEL

25

2.1.3 – Plasmid Constructs: Plasmid constructs listed herein were either purchased from the company indicated, or were prepared previously by the indicated person according to standard cloning protocols. Instances where no source is listed denote plasmid constructs prepared by the author. Plasmid pET-41a(+)

Description (Source) pET vector for expression of proteins under control of the T7 Promoter. Parental vector expresses GST, a sixhistidine tag, an S-tag, and a further six-histidine tag, with a multiple cloning site allowing addition of an ORF of interest in-frame with one or more of these peptides. Provides kan resistance (Novagen).

pET-41a(+)-TEV-mHuCsv1

As for pET-41a(+), with the coding sequence of mHuCsv1 replacing those of the first six-histidine tag and S-tag.

pET41a(+)-Strep-mHuCsv1-His 8

As for pET-41a(+)-TEV-mHuCsv1, with a Strep Tag coding sequence located directly 5’ to the TEV protease site, and an eight-histidine tag coding sequence located direcly 3’ to the HuCsv1 ORF.

pGBKT7

pGBKT7 vector for bacterial and yeast expression of proteins under control of the T7 and ADH1 promoters, respectively. Parental vector expresses the GAL4 DNAbinding domain, with a multiple cloning site allowing addition of an ORF of interest in-frame for use in yeast two-hybrid screening experiments. Also possesses the TRP1 gene, which provides for biosynthesis of tryptophan in yeast possessing a mutated TRP1 locus (Clontech).

pGBKT7-mHuCsv2spacer

As for pGBKT7, with the coding sequence of the mHuCsv2 spacer domain inserted directly 3’ to the GAL4 DNA-binding domain.

pGBKT7-mHuCsv2spacer_RRM

As for pGBKT7, with the coding sequence of the mHuCsv2 spacer and RRM3 domains inserted directly 3’ to the GAL4 DNA-binding domain.

26

pAD-GAL4-2.1

pAD-GAL4-2.1

vector

for

bacterial

and

yeast

expression of proteins under control of the T7 and ADH1

promoters,

respectively.

Parental

vector

expresses the GAL4 transcriptional activation domain (GAL4-AD), with a multiple cloning site allowing addition of ORFs of interest for use in yeast two-hybrid screening experiments. Also possesses the LEU2 gene, which allows for leucine biosynthesis by yeast cells possessing a mutated LEU2 locus (Clontech).

2.1.4 – Primers: Name StrepTag Fwd

Sequence (5’ to 3’)

Application

GAT CCC TGG AAA TAC AGG TTT TCT TTT Formation of Strep Tag TCG AAC TGC GGG TGG CTC CAC A

ORF with Bam HI and Nde I compatible single strand overhangs (with StrepTag Rev)

StrepTag Rev

TAT GTG GAG CCA CCC GCA GTT CGA AAA Formation of Strep Tag AGA AAA CCT GTA TTT CCA GG

ORF with Bam HI and Nde I compatible single strand overhangs (with StrepTag Fwd)

pET21-F

CGA AAT TAA TAC GAC TCA CTA TAG GGG

Sequencing/colony

PCR

(with 9) from pET vectors. pET21-R

GCT AGT TAT TGC TCA GCG GTG GC

Sequencing/colony

PCR

(with 8) from pET vectors. GST-F

GCT ACC TGA AAT GCT GAA AAT GTT CG

Sequencing/amplification (with

27)

from

vectors

containing the GST coding sequence. HuClinker F

AAG ACA GGG CAG GCC CTG CTC

Colony

PCR

(with

144)

against mHuCsv1 coding sequence. HuCR2

GCT GCT TGC TGG TCT TGA AGG

Sequencing of mHuCsv1 coding sequence.

HuCsvFwd

TCC CAG TTC TGC CTC TAT CC

Sequencing of mHuCsv1 coding sequence.

27

HuCsvRev

CGT TGA GGC TGG CGA TAG

Colony

PCR

(with

115)

against mHuCsv1 coding sequence. HuCsv4F1

TTC TCC CCA ATC GCC ATC

Sequencing of mHuCsv1 coding sequence.

CIRBP CT1

TAA AGC AAA AAG TAA AAA ATA AAA AAG In vitro transcription (with GTC TCC CTA TAG TGA GTC GTA TTA CT

T7Top) of CIRBP CT1 RNA for

nitrocellulose

filter-

binding analyses. CIRBP CT2

AGA ACA TTT CAA CTG GAA CAC AAC GTA In vitro transcription (with AAG CTC CCT ATA GTG AGT CGT ATT ACT

T7Top) of CIRBP CT2 RNA for

nitrocellulose

filter-

binding analyses. CIRBP CT3

GTA AAA AAA GGG GAC TGT CTT TAA TAG In vitro transcription (with AAC CTC CCT ATA GTG AGT CGT ATT ACT

T7Top) of CIRBP CT3 RNA for

nitrocellulose

filter-

binding analyses. hnRNP A/B CT1

AAA ATA AAA CAA AAA CAA AAA ACA AAA In vitro transcription (with

GTA AAA TAA AAC AAA TCT CCC TAT AGT T7Top) of hnRNP A/B CT1 GAG TCG TAT TAC T

RNA

for

nitrocellulose

filter-binding analyses. hnRNP A/B CT2

AAA ATG GAC AAG CTT ACA ATC ATA AAG In vitro transcription (with

AAA TGG GCT CCC TAT AGT GAG TCG TAT T7Top) of hnRNP A/B CT2 TAC T

RNA

for

nitrocellulose

filter-binding analyses. hnRNP A/B CT3

GAG GAA AAA AAA AAC AAA TTA AAA In vitro transcription (with

GAA TAA AAA TGT CAC CCT CCC TAT AGT T7Top) of hnRNP A/B CT3 GAG TCG TAT TAC T

RNA

for

nitrocellulose

filter-binding analyses. hnRNP A/B CT4

CAC AAA GCA CAG GTG ACA CTA GGG In vitro transcription (with

GGA GAA AAG GAA ATC ATC TAG GAA CAC T7Top) of hnRNP A/B CT4 TAC TCC CTA TAG TGA GTC GTA TTA CT

RNA

for

nitrocellulose

filter-binding analyses. Nucleolin CT1

CTA AAG CAA ACT TAA AAA AAC AAA TTA In vitro transcription (with

AAT AAA AAT TGG GTT TGA AAT AAA GTC T7Top) of Nucleolin CT1 TCC CTA TAG TGA GTC GTA TTA CT

RNA

for

nitrocellulose

filter-binding analyses.

28

Nucleolin CT2

AAC AAA AAA GAG AAA TGG GAA AAG In vitro transcription (with

AAA CTG TTC TCC CTA TAG TGA GTC GTA T7Top) of Nucleolin Ct2 TTA CT

RNA

for

nitrocellulose

filter-binding analyses. bCatenin CT1

AAC CGA AGA GGA TGA AAA TAA AGA In vitro transcription (with GTA AAA GAG TTC CGT CTT CAG ACC GAG T7Top)

of

beta-Catenin

AGT CCG TCT CCC TAT AGT GAG TCG TAT CT1 RNA for nitrocellulose TAC T bCatenin CT2

filter-binding analyses.

ACT AAA CAC TAC ACA AAC ACA AAG AAA In vitro transcription (with AGA GCA GGA TAA AAA CAA AGC TCC TCC T7Top) CTA TAG TGA GTC GTA TTA CT

of

beta-Catenin

CT2 RNA for nitrocellulose filter-binding analyses.

zfeIF4G2a CT1

GAG TTT TGT GCT GAA AGG AAC AGC AGA In vitro transcription (with

AAA GGC TCT TGC AAA TAC TCC CTA TAG T7Top) of eIF4G2a CT1 TGA GTC GTA TTA CT

RNA

for

nitrocellulose

filter-binding analyses. zfeIF4G2a CT2

CAG GTT TCT TAA AAT GTA GAA TAA AAA In vitro transcription (with

AAA CAA AAC AAA AAG AGA TGC AGC TCC T7Top) of eIF4G2a CT2 CTA TAG TGA GTC GTA TTA CT

RNA

for

nitrocellulose

filter-binding analyses. Sfrs1l CT1

CTA TGT ATT GAA AAA AAA TGG ATG GGG In vitro transcription (with

GGG AAA AAC AAC TCC CTA TAG TGA GTC T7Top) of sfrs1l CT1 RNA GTA TTA CT

for

nitrocellulose

filter-

binding analyses. Sfrs1l CT2

AGA GGG AGG GAA AAA AAA AAA AAA In vitro transcription (with

GAC AAC ATT CAT AAG ATC AAC TGT CTC T7Top) of sfrs1l CT2 RNA ACG AAC TCC CTA TAG TGA GTC GTA TTA for CT Sfrs1l CT3

nitrocellulose

filter-

binding analyses.

CTG TCT CAC GAA ACA GGG GGT GAT TAA In vitro transcription (with

AAC ATG GGG TTC TCC CTA TAG TGA GTC T7Top) of sfrs1l CT3 RNA GTA TTA CT

for

nitrocellulose

filter-

binding analyses. Sfrs1l CT4

GTA AAG ACT GTA AAA ACA AAA CGA CTC In vitro transcription (with

AAG GGC TCC ACC AAT GAG GTG AAG AGT T7Top) of sfrs1l CT4 RNA GGC TCC CTA TAG TGA GTC GTA TTA CT

for

nitrocellulose

filter-

binding analyses. mCofilin CT1

AGT TGG CAG CAT GGG ATG GGG AGG In vitro transcription (with

GAT ACG GAG TAG GGG TGT CTC CCT ATA T7Top) of mCofilin CT1 GTG AGT CGT ATT ACT

RNA

for

nitrocellulose

filter-binding analyses.

29

mCofilin CT2

GTT AAA AAA AAA AAT ACA GGC TCC CCC In vitro transcription (with

CAA ACT GGG TGC CTA GGA GCT CCC TAT T7Top) of mCofilin CT2 AGT GAG TCG TAT TAC T

RNA

for

nitrocellulose

filter-binding analyses. mHig1 CT1

GGG ATT TTG GGA GCG GGA GGG GGC In vitro transcription (with

AGC GGT GGA AGC TCC CTA TAG TGA GTC T7Top) of mHig1 CT1 RNA GTA TTA CT

for

nitrocellulose

filter-

binding analyses. mHig1 CT2

GGA CGC ATC ACA AAT AGA AGC ATG CAA In vitro transcription (with

CCC AAC TGA AAT GAA CCT CTC CCT ATA T7Top) of mHig1 CT2 RNA GTG AGT CGT ATT ACT

for

nitrocellulose

filter-

binding analyses. T7Top

AGT AAT ACG ACT CAC TAT AGG GAG AGG

In vitro transcription of RNAs

for

nitrocellulose

filter-binding analyses.

2.1.5 – Enzymes: AccuPrime II Supermix:

Invitrogen

CIP:

New England Biolabs

Restriction Endonucleases:

New England Biolabs

RNase A:

Worthington Biochemical Corporation

RNase T1:

Ambion

RQ1 DNase:

Promega

Taq DNA Polymerase:

Invitrogen

T4 DNA Ligase:

New England Biolabs

T4 PNK:

New England Biolabs

T7 RNA Polymerase Plus:

Ambion

30

2.1.6 – Antibodies/Antisera: All patient antisera used are listed by the first two letters of the surname of the patient from which they were obtained, with the antigen recognized by each serum indicated in parentheses. ‘Ro’ PND Patient Antiserum (anti-nHu) ‘Gu’ PND Patient Antiserum (anti-nHu) ‘Fa’ PND Patient Antiserum (anti-cdr2) Mouse anti-HuR (3A2) Monoclonal Antibody:

Santa Cruz Biotechnology

Mouse anti-HuC/D (16A11) Monoclonal Antibody:

Abcam

Mouse anti-CUGBP2 Monoclonal Antibody:

Sigma-Aldrich

Rabbit anti-Maged1 Monoclonal Antibody:

Abcam

Rabbit anti-eIf4a Monoclonal Antibody:

Cell Signalling Technology

Rabbit anti-actin Monoclonal Antibody:

Sigma-Aldrich

Mouse anti-His 6 Monoclonal Antibody:

Cell Signalling Technology

Donkey anti-human IgG, HRP conjugate:

Jackson ImmunoResearch

Donkey anti-mouse IgG, HRP conjugate:

Jackson ImmunoResearch

Donkey anti-rabbit IgG, HRP conjugate:

Jackson ImmunoResearch

Donkey anti-human IgG, Cy5 conjugate:

Jackson ImmunoResearch

Donkey anti-human IgG, Cy3 conjugate:

Jackson ImmunoResearch

Donkey anti-mouse IgG, Cy5 conjugate:

Jackson ImmunoResearch

Donkey anti-mouse IgG, Cy3 conjugate:

Jackson ImmunoResearch

Donkey anti-rabbit IgG, FITC conjugate:

Jackson ImmunoResearch

31

2.1.7 – Bacterial Growth Media: L Agar Plates:

1.5% Bacto-Agar, 25 mL per plate

Luria Bertani Broth:

1% Tryptone Peptone, 0.5% Yeast Extract, 0.5% NaCl, pH 7.5

Luria Broth:

1% Bacto-Tryptone, 0.5% Yeast Extract, 1% NaCl, pH 7.0

SOC + Glucose:

2% Tryptone, 0.5% Yeast Extract, 10 mM NaCl, 2.5 mM KCl, 10 mM MgCl 2 , 10 mM MgSO 4 , 0.2% Glucose

Antibiotics were added to plates at time of pouring, or to liquid media at time of inoculation, as required. In all cases, ampicillin was utilised at a concentration of 100 µg/mL and kanamycin at 50 µg/mL.

2.1.9 – Yeast Growth Media: YPAD Agar Plates:

1% Yeast Extract, 2% Tryptone Peptone, 2% Glucose, 0.1 mg/mL Adenine Hemisulfate, 1.5% Bacto-Agar, 25 mL per plate

YPAD Broth:

1% Yeast Extract, 2% Tryptone Peptone, 2% Glucose, 0.1 mg/mL Adenine Hemisulfate

SC-Agar Plates:

0.67% Yeast Nitrogen Base (without Amino Acids), 2% Glucose, 0.83 mg/mL SC Drop-Out Mix, 1.5% Bacto-Agar, pH 5.6

SC Broth:

0.67% Yeast Nitrogen Base (without Amino Acids), 2% Glucose, 0.83 mg/mL SC Drop-Out Mix, pH 5.6

32

SC Drop-Out Mix:

5.5% Adenine Hemisulfate*, 5.5% Arginine HCl, 5.5% Histidine HCl*, 5.5% Isoleucine, 5.5% Leucine*, 5.5% Lysine HCl, 5.5% Methionine, 8.3% Phenylalanine, 5.5% Serine, 5.5% Threonine, 8.3% Tryptophan*, 5.5% Tyrosine, 3.9% Uracil, 24.5% Valine

*Amino acids omitted as necessary to prepare selective drop-out media

Aureobasidin A (AbA) was added to plates at the time of pouring, as required. In all cases, final AbA concentration was 125 ng/mL.

2.1.9 – Chemicals & Reagents: All chemicals and reagents used herein were purchased from Sigma-Aldrich Inc. with the exception of the following: -

4-12% Bis Tris Nu-PAGE gels:

Invitrogen

Agarose, DNA Grade:

Progen

Complete Protease Inhibitors (CPI):

Roche Diagnostics

EDTA-free CPI:

Roche Diagnostics

Glycine:

Amresco

Hi-Safe Scintillation Fluid:

Wallac Scintillation Products

His-Trap FF Columns:

GE Healthcare Life Sciences

Hybond-LFP Membrane:

GE Healthcare Life Sciences

Immobilon FL PVDF Membrane:

Pall Life Sciences

Ni-IDA Resin:

Scientifix

Nitrocellulose Filters:

Millipore 33

Potassium Acetate:

Fluka

SAC/Zysorbin:

Invitrogen

Sodium Chloride:

Ajax Finechem

Tris Base:

Amresco

Tween 20:

Fluka

Urea:

Merck

Western Lightning Chemiluminescence Reagent Plus:

Perkin-Elmer

2.1.10 – Buffers & Solutions: BB:

200 mM Potassium Acetate, 50 mM Tris-Acetate, 5 mM Magnesium Acetate, pH 7.7

Big Dye Buffer:

200 mM Tris, 5 mM MgCl 2 , pH 9.0

Binding Buffer:

1M Potassium Acetate, 250 mM Tris-Acetate, 25 mM Magnesium Acetate, 50 mM DTT, 5 mg/mL heparin, pH 7.7

Blocking Buffer:

150 mM NaCl, 50 mM Tris-Cl, 1 mM EDTA, 0.1% Triton X-100, 5% (w/v) Skim Milk Powder, pH 7.5

Chloroform/Isoamyl Alcohol (49:1):

Coomassie Staining Solution:

98% Chloroform, 2% Isoamyl Alcohol

0.2% Brilliant Blue Coomassie, 50% Ethanol, 10% Acetic Acid 34

Coomassie Destaining Solution:

5% Ethanol, 5% Acetic Acid

Coomassie Gel Drying Solution:

30% Methanol, 5% Glycerol

Crosslinking Dilution Buffer:

1 mg/mL BSA, 1x PBS

Crosslinking Wash Buffer:

0.2 M Triethanolamine, 1x PBS

DNA Loading Buffer (8x):

19% Ficoll 400, 1x TBE, Xylene Cyanol, Bromophenol Blue

Formamide Loading Buffer (2x):

95% Formamide (deionized), 1x TBE, Xylene Cyanol, Bromophenol Blue

IP Lysis Buffer:

20 mM Tris-HCl pH 8, 137 mM NaCl, 10% Glycerol, 1% NP-40, 2 mM EDTA, 1x CPI

mHuCsv1 Storage Buffer:

150 mM KCl, 50 mM Tris, 0.5% NP-40, 0.15 mM DTT, 20% Glycerol, 0.5 mM EDTA, pH 7.0

MOPS Running Buffer (20x):

1 M MOPS, 1 M Tris, 69.3 mM SDS, 20.5 mM EDTA, pH 7.7

Ni-IDA Wash Buffer:

136.9 mM NaCl, 2.7 mM KCl, 1.5 mM KH 2 PO 4 , 8.1 mM Na 2 HPO 4 , 20 mM Imidazole, 0.5% Tween-20, pH 8.0

35

Ni-IDA Elution Buffer:

136.9 mM NaCl, 2.7 mM KCl, 1.5 mM KH 2 PO 4 , 8.1 mM Na 2 HPO 4 , 250 mM Imidazole, 0.5% Tween-20, pH 8.0

PBS:

136.9 mM NaCl, 2.7 mM KCl, 1.5 mM KH 2 PO 4 , 8.1 mM Na 2 HPO 4 , pH 7.4

PBST:

136.9 mM NaCl, 2.7 mM KCl, 1.5 mM KH 2 PO 4 , 8.1 mM Na 2 HPO 4 , 0.1% Tween-20, pH 7.4

Phenol (DNA):

Pure phenol equilibrated with 0.15 M Sodium Acetate pH 5.2

Phenol (RNA):

Pure phenol equilibrated with 0.1 M Tris pH 8.0

PXL2 (1x):

136.9 mM NaCl, 2.7 mM KCl, 1.5 mM KH 2 PO 4 , 8.1 mM Na 2 HPO 4 , 0.5% Deoxycholate, 0.5% NP-40, EDTA-free CPI (1x), pH 7.4

R Buffer:

150 mM NaCl, 50 mM Tris-Cl, 1 mM EDTA, 0.1% Triton X-100, pH 7.5

RNA Elution Buffer:

1M Sodium Acetate, 1 mM EDTA, pH 5.2

SDS-PAGE Loading Buffer (2x):

20% Glycerol, 4% (w/v) SDS, 100 mM Tris-Cl, 2% β-Mercaptoethanol, 0.2% Bromophenol Blue, pH 6.8

36

SDS-PAGE Running Buffer:

25 mM Tris, 192 mM Glycine, 0.1% SDS

SDS-PAGE Separating Buffer (2x):

0.75 M Tris-Cl, 0.2% SDS, pH 8.6

SDS-PAGE Stacking Buffer (2x):

0.25 M Tris-Cl, 0.2% SDS, pH 6.8

Semi-Dry Transfer Buffer:

24 mM Tris, 192 mM Glycine, 20% (v/v) Methanol

TBE (20x):

1.78 M Tris-Cl, 1.78 M Boric Acid, 50 mM EDTA, pH 8.3

TBE (1x):

89 mM Tris-Cl, 89 mM Boric Acid, 2.5 mM EDTA, pH 8.3

Wash Buffer I:

1 M NaCl, 50 mM Tris-Cl, 1 mM EDTA, 0.1% Triton X-100, pH 7.5

Wash Buffer II:

150 mM NaCl, 50 mM Tris-Cl, 1mM EDTA, 1% Triton X-100, 0.05% SDS, pH 7.5

Yeast Transformation Buffer:

33.3% PEG 3500, 100 mM Lithium Acetate, 6.9 mg/mL salmon sperm DNA, 100 ng plasmid DNA

37

2.1.11 – Proprietary Buffers: NEB Restriction Enzyme Buffers (10x):

New England Biolabs

Novex-LDS Load Buffer:

Invitrogen

One-Phor-All Buffer (10x):

GE Healthcare Life Sciences

T4 DNA Ligase Buffer (10x):

New England Biolabs

T4 PNK Buffer (10x):

New England Biolabs

Taq DNA Polymerase Buffer (10x):

Invitrogen

Transcription Buffer (5x):

Ambion

2.1.12 – Proprietary Kits: DryEase Minigel Drying Kit:

Invitrogen

QIAprep Spin Miniprep Kit:

Qiagen

QIAEX II Gel Extraction Kit:

Qiagen

Silver Staining Kit, Protein:

GE Healthcare Life Sciences

All kits were used according to the manufacturers’ instructions.

2.1.13 – DNA Markers: 2-Log DNA Ladder:

New England Biosciences

2.1.14 – Protein Markers: ECL Plex Fluorescent Markers:

GE Healthcare Life Sciences

Precision Plus Protein Dual Colour Standard:

Bio-Rad

Prestained Protein Marker, Broad Range:

New England Biosciences

SeeBlue Plus 2:

Invitrogen

38

2.1.15 – Additional Equipment: P1 Peristaltic Pump:

GE Healthcare Life Sciences

39

2.2 – Experimental Methods: 2.2.1 – General Bacterial Procedures: Liquid Cultures: Bacterial liquid cultures were inoculated under sterile conditions by addition of single colonies from bacterial growth plates, or from samples of liquid cultures, and were each grown with aeration under appropriate conditions of temperature and duration of growth, as indicated within the text. Where applicable, culture growth was monitored by spectrophotometric measurement of optical density at 600 nm (OD 600 ). Samples of liquid culture to be kept for storage were placed into 40% Glycerol solution and stored at -80oC. Plate Cultures: Bacterial plate cultures were prepared under sterile conditions by application of an appropriate liquid sample of bacterial cells directly to the plate surface, followed by spreading of this sample with a heat-sterilized glass spreading implement. In all cases, plates were dried at room temperature, agar face up, for approximately ten min, and were subsequently grown for 12-16 hours at 37oC, agar face down. Plates were subsequently sealed with parafilm, and stored at 4oC, as required. Bacterial DNA Transformation: Plasmid DNA (not more than 10 µL) was added to a sample of competent cells (50 µL), thawed on ice. The sample of cells was incubated on ice for 30 min. followed by incubation at 42oC for 30 sec, and a further 2 min on ice. SOC + Glucose (450 µL) was added, and the resultant sample incubated at 37oC for 1 hour with shaking.

50 µL of this culture was plated onto an L Agar plate

containing the appropriate antibiotic.

Remaining culture was submitted to

centrifugation (5,900 x g; 1 min), and the resultant cell pellet resuspended in 50 µL SOC + glucose. This sample was plated onto an L Agar plate containing the appropriate antibiotic. Plates were incubated as described (2.2.1).

40

2.2.2 – General Yeast Procedures: Liquid Cultures: Yeast liquid cultures were inoculated under sterile conditions by addition of single colonies from yeast growth plates, or from samples of liquid cultures, as indicated, and were each grown with aeration at 30oC for appropriate duration, as indicated in

the

text.

Where

applicable,

culture

growth

was

monitored

by

spectrophotometric measurement of optical density at 600 nm (OD 600 ). Plate Cultures: Yeast plate cultures were prepared under sterile conditions either by application of an appropriate liquid sample of yeast cells directly to the plate surface, followed by spreading of this sample with a heat-sterilized glass spreading implement, or by streaking of individual yeast colonies from previously prepared yeast growth plates, using a heat-sterilized glass streaking implement. In all cases, plates were dried at room temperature, agar face up, for approximately 10 min, and were subsequently grown for 3-5 days at 30oC, agar face down.

Plates were

subsequently sealed with parafilm and stored at 4oC, as required. Yeast DNA Transformation: YPAD or specific SC drop-out liquid culture media (4 mL each) were prepared and inoculated with parental AH109 or y2hgold yeast strains or pre-transformed yeast strains, respectively.

Resultant cultures were grown overnight at 30oC, OD 600

measured by spectrophotometry, cultures diluted to a final density of OD 600 ~ 0.7 in a final volume of 50 mL, and grown at 30oC until OD 600 ~ 1.5. Yeast were collected by centrifugation (3,000 x g, 21oC, 5 min.), resuspended in 25 mL Milli-Q water, collected again (3,000 x g, 21oC, 5 min.), resuspended in 1 mL Milli-Q water, and collected a final time (16,100 x g, 21oC, 30 sec.). Resultant yeast cell pellet was resuspended in 1 mL Milli-Q water, divided into 100 µL aliquots, spun down (16,100 x g, 21oC, 30 sec.), supernatant discarded, and yeast resuspended in 360 µL 41

Yeast Transformation Mix. Transformation reactions were incubated at 42oC for 30 min., with mixing at 15 min. intervals, yeast collected by centrifugation (16,100 x g, 21oC, 30 sec.), resuspended in 1 mL Milli-Q water, and 200 µL plated onto the appropriate YPAD or SC drop-out selective media. Plates were incubated as described (2.2.2).

2.2.3 – DNA Manipulation: Preparation of Bacterial Plasmid DNA: In all cases, plasmid DNA was isolated from cultures of E. Coli DH5α using a QIAprep Spin Miniprep Kit, according to the manufacturers’ instructions. Preparation of Yeast Plasmid DNA: In all cases, plasmid DNA was isolated from cultures of S. Cerevisiae AH109 using a QIAprep Spin Miniprep Kit (Yeast), according to the manufacturers’ instructions. Determination of DNA Concentrations: DNA sample concentrations were assayed by either spectrophotometric measurement of absorbance at a wavelength of 260 nm (A 260 ), or agarose gel electrophoresis followed by comparison of the relative intensities of sample bands to quantitative marker bands. Restriction Endonuclease Digestion: Endonuclease digestion of plasmid DNA was carried out for 1-2 hours according to the manufacturer’s instructions, and was analysed by agarose gel electrophoresis. Agarose Gel Electrophoresis: DNA

size,

purity,

and/or

concentration

was

analysed

by

agarose

gel

electrophoresis. DNA loading buffer was diluted 1:8 in DNA samples, loaded onto a 1% Agarose gel containing trace amounts of ethidium bromide, and 42

electrophoresis carried out in 1x TBE at 120 V for 0.5-1 hour, as appropriate. Electrophoresed DNA was visualized by UV transillumination. Agarose Gel Purification of DNA: DNA fragments of interest were separated by agarose gel electrophoresis as described (2.2.3), bands excised with a minimum of agarose, and DNA extracted using the QIAEX II Gel Extraction Kit according to the manufacturers’ instructions. Ligation of Endonuclease Digested DNA: Fragments of endonuclease-digested DNA to be ligated were subjected to agarose gel purification as described (2.2.3), and added in a 1:1 molar ratio to the ligation reaction, consisting of T4 DNA Ligase Buffer (1x), 1 µL 400 U/µL T4 DNA Ligase, such that the final DNA concentration was 10 ng/µL, and the total reaction volume 10 µL. Ligation reactions were incubated overnight at 16oC, followed by 8 hours at 4oC, and were transformed into E. Coli DH5α competent cells as described (2.2.1).

2.2.4 – PCR Techniques: General PCR Amplification: PCR amplification was carried out from plasmid DNA template using AccuPrime Supermix II according to the manufacturers’ instructions. Colony PCR: Colony PCR was performed to assay for the presence of desired constructs in individual plate-grown colonies of E. Coli DH5α cells transformed with ligation reaction output DNA as described (2.2.1). Colonies were selected and isolated from L Agar plates, and resuspended in 10 µL Milli-Q water.

1 µL of this

resuspension was added to a 25 µL reaction containing 2.5 µL 10x Taq DNA Polymerase Buffer, 0.5 µL 10 mM each dNTPs, 0.5 µL 10 µM forward primer, 0.5 µL 10 µM reverse primer, and 0.5 µL 5 U/µL Taq DNA Polymerase. 43

PCR reactions were incubated at 94oC for 4 min, followed by 25 cycles of 94oC for 30 sec, 52oC for 30 sec, and 72oC for 30 sec, with a final incubation at 72oC for 5 min. Completed reactions were stored at 4oC prior to analysis by agarose gel electrophoresis as described (2.2.3). Big Dye DNA Sequencing: Plasmid DNA constructs were sequenced using Perkin-Elmer ABI PRISM Big Dye Version 3.

Reactions consisted of 0.5 µL Big Dye, 7.5 µL Big Dye Buffer,

300 ng template DNA, and 3 pmoles of the appropriate primer in a total reaction volume of 20 µL. Sequencing reactions were incubated at 96oC for 2 min, followed by 20 cycles of 96oC for 20 sec, 50oC for 10 sec, and 60oC for 4 min. Completed reactions were stored at 4oC prior to being prepared for sequencing, which consisted of transfer of each reaction to an eppendorf tube, followed by addition of 1 µL of 10 mg/mL glycogen and 80 µL 75% isopropanol and incubation at room temperature for 15 min. Incubated samples were centrifuged (20 min; 16,100 x g) to pellet DNA, supernatant removed, and the DNA pellet washed with 150 µL 75% isopropanol.

After a further centrifugation (5 min; 16,100 x g),

supernatant was removed, and pellets air died.

Precipitated reactions were

sequenced at the Institute of Medical and Veterinary Science Sequencing Center, Adelaide, and the resulting chromatograph files analysed and compared to predicted sequences using MacVector 9.0

2.2.5 – Protein Manipulation: Nu-PAGE Analysis: Tissue lysates or other protein samples were diluted 1:1 with 2x Novex LDS Load Buffer, incubated for 5 min at 95oC, and electrophoresed through a 4-12% Bis-Tris Nu-PAGE gradient polyacrylamide gel in an Invitrogen X-Cell SureLock Mini Cell containing 600 mL MOPS Running Buffer at a constant 150 V for 1.5 hours. 44

SDS-PAGE Analysis: Tissue lysates or other protein samples were diluted 1:1 with 2x SDS-PAGE Loading Buffer, incubated for 5 min at 95oC, and electrophoresed through either a 10% SDS-PAGE polyacrylamide gel, consisting of 2.5 mL 2x Separating Buffer, 1.25 mL 40% Acrylamide (37.5:1), and 1.25 mL Milli-Q water, or a 12% SDS-PAGE polyacrylamide

gel,

consisting

of

2.5

mL

2x

Separating

Buffer,

1.5 mL 40% Acrylamide (37.5:1), and 1 mL Milli-Q water. Both types of gel were overlayed with a low-concentration SDS-PAGE polyacrylamide “stacking” gel, consisting of 1.25 mL 2x Stacking Buffer, 187.5 µL 40% Acrylamide (37.5:1), 1 mL Milli-Q water. All SDS-PAGE gels were induced to undergo polymerisation by addition of 30 µL TEMED and 60 µL 10% APS. Electrophoresis was carried out in an Invitrogen X-Cell SureLock Mini Cell containing 600 mL SDS-PAGE Running Buffer at a constant of 150 V for 1.5 hours. Coomassie Staining: The SDS-PAGE polyacrylamide gel of interest was incubated in 50 mL Coomassie Stain Solution at room temperature for 1 hour (10% SDS-PAGE) or 2 hours (12% SDS-PAGE) with shaking. Stain solution was removed, the gel washed once with 50 mL Coomassie Destain Solution, and incubated in Coomassie Destain Solution at room temperature overnight. Destained gels were dried using the DryEase Minigel Drying Kit according to the manufacturers’ instructions. Silver Staining: Nu-PAGE polyacrylamide gels were analysed by silver staining using the Silver Staining Kit, Protein according to the manufacturers’ instructions. Chemiluminescent Western Blot Analysis: Tissue lysates or other protein samples were separated by SDS-PAGE or Nu-PAGE analysis as described (2.2.5).

Following electrophoresis, low-concentration

polyacrylamide in the upper region of the gel was excised and discarded, and the 45

remaining gel placed into a “sandwich” consisting of, in order, Bio-Rad Trans-Blot SD Anode, three sheets of whatman paper pre-equilibrated in Semi-Dry Transfer Buffer,

Immobilon

FL

PVDF

Membrane

pre-equilibrated

for

5

min

in

100% methanol followed by 15 min in Semi Dry Transfer Buffer, polyacrylamide gel, three sheets of whatman paper pre-equilibrated in Semi-Dry Transfer Buffer, Bio-Rad Trans-Blot SD Cathode.

Air bubbles trapped between layers were

removed before placement of the cathode on the upper face of the “sandwich” by rolling over with a clean falcon tube.

Transfer was carried out at a constant

15 V for 45 min. Once transfer was complete, the PVDF membrane was removed and incubated in 20 mL Blocking Buffer at room temperature for 1 hour or at 4oC overnight, followed by incubation with nutation in 3 mL fresh Blocking Buffer containing the appropriate primary antibody at room temperature for 1 hour or at 4oC overnight. After incubation in antibody, the membrane was submitted to consecutive 5 min washes at room temperature in 20 mL each of Wash Buffer I, Wash Buffer II, Wash Buffer I, Wash Buffer II, and R Buffer. Subsequently, the membrane was incubated with nutation in 3 mL fresh R Buffer containing the appropriate HRP-conjugated secondary antibody at room temperature for 1 hour. The previous scheme of consecutive washes was repeated, and protein bands reactive with the applied primary antibody visualised by chemiluminescence using Western Lightning Chemiluminescence Reagent Plus according to the manufacturer’s instructions. Fluorescent Western Blot Analysis: Tissue lysates or other protein samples were separated by SDS-PAGE or Nu-PAGE analysis as described (2.2.5).

Following electrophoresis, low-concentration

polyacrylamide in the upper region of the gel was excised and discarded, and the remaining gel placed into a “sandwich” consisting of, in order, Bio-Rad Trans-Blot SD Anode, three sheets of whatman paper pre-equilibrated in Semi-Dry Transfer 46

Buffer, Hybond-LFP Membrane pre-equilibrated for 60 sec in 100% Methanol followed by 20 sec in Milli-Q water and 5 min in Semi Dry Transfer Buffer, polyacrylamide gel, three sheets of whatman paper pre-equilibrated in Semi-Dry Transfer Buffer, Bio-Rad Trans-Blot SD Cathode. Air bubbles trapped between layers were removed before placement of the cathode on the upper face of the “sandwich” by rolling over with a clean falcon tube. Transfer was carried out at a constant 15 Volts for 45 min.

Once transfer was complete, the Hybond-LFP membrane was removed and incubated for in 20 mL 1x PBST for 30 min at room temperature, followed by incubation with nutation in 3 mL 1x PBST containing the appropriate primary antibodies at 4oC overnight.

After incubation in antibody, the membrane was

rinsed twice in 20 mL 1x PBST at room temperature, and was subsequently washed twice in 20 mL 1x PBST at room temperature for 5 min per wash, followed by incubation with nutation in 3 mL fresh 1x PBST containing the appropriate fluorophore-conjugated secondary antibodies at room temperature for 1 hour in darkness. The membrane was subsequently rinsed twice in 20 mL 1x PBST, followed by four washes in 20 mL 1x PBST at room temperature for 5 min per wash in darkness.

The membrane was dried in darkness between two sheets of

whatman paper, and fluorescence signal visualised on a Typhoon Trio Variable Mode Phosphorimager (Amersham Biosciences).

47

2.2.6 – Protein Purification: Induction of Protein Expression: Luria Broth cultures (200 mL) supplemented with Kanamycin (50 µg/mL) were prepared and inoculated with a single colony of pET41a(+)-StrepTag-mHuCsv1His 8 -transformed BL21(DE3) RIPL, as described (2.2.1). Cultures were grown overnight at 37oC and 50 mL samples used to inoculate 2 x 500 mL flasks of Luria Broth supplemented with Kanamycin (50 µg/mL). Resultant cultures were grown at 37oC with aeration until OD 600 reached a value of 0.5-0.6 (~ 2 hours), 1 mL of culture removed for later analysis, IPTG added to a final concentration of 0.1 mM, and cultures grown at 30oC for a further 5 hours. Culture Centrifugation: Following growth in the presence of IPTG, bacterial cultures were collected by centrifugation (4,000 x g; 4oC, 10 min), growth media decanted, and bacterial pellets stored at -80oC overnight.

Pellets were thawed, combined, and

resuspended in 25 mL 4oC 1x PBS either in the presence or absence of Complete Protease Inhibitor (CPI), as indicated in the text. Mechanical Lysis of Bacteria: Mechanical lysis of bacteria was carried out using a Branson B-30 Cell Sonifier. Samples of bacteria for lysis were kept on ice and sonicated in five cycles, each consisting of 30 sec sonication followed by 30 sec on ice.

Where indicated,

100 µL lysate was removed for later analysis, and remaining lysate clarified by centrifugation (50,512 x g, 4oC, 30 min). Where indicated, 100 µg clarified lysate was removed for later analysis.

48

Binding of Recombinant Protein to HisTrap FF Column: A Single 5 mL HisTrap FF column was pre-equilibrated by flushing with 10 column volumes 1x PBS at a flow rate of ~20 mL/min. Clarified lysates were supplemented by addition of imidazole to a final concentration of 50 mM and applied directly to the column at a flow rate of ~ 2 mL/min. Following binding, column was washed with 10 column volumes Ni-IDA Wash Buffer at 20 mL/min, and eluted by application of 10 column volumes Ni-IDA Elution Buffer at either 4 mL/min or 2 mL/min, as indicated in the text. Eluted proteins were collected into 500 µL fractions and analysed by SDS-PAGE gel electrophoresis and coomassie staining, as described (2.2.5). Dialysis of Recombinant Protein: Fractions containing the highest concentrations of recombinant protein, as described in the text, were combined, dialysed against 5 L mHuCsv1 Storage Buffer for 24 hours, and stored at -80oC for later use.

2.2.7 – RNA Manipulation: Phenol/Chloroform Extraction of RNA: Purification of plasmid DNA samples was achieved by Phenol/Chloroform extraction. Sample volume was brought up to 300 µL by addition of Milli-Q water, and 300 µL of RNA Phenol added, followed by 100 µL of chloroform/isoamyl alcohol (49:1). Samples were subjected to high-speed vortex mixing, centrifuged (16,100 x g; 5 min), and the aqueous (top) layer collected. 30 µL 3M Sodium Acetate, pH 5.2 and 900 µL 100% ethanol were added, and the samples incubated for 30 min to overnight at -20oC followed by centrifugation (16,100 x g; 4oC; 20 min) to yield an RNA pellet. Supernatant was removed, the pellet washed with 100 µL 75% Ethanol, and centrifuged (16,100 x g; 4oC; 5 min). Supernatant was again removed, and the RNA pellet allowed to air dry at room temperature. RNA was resuspended in an appropriate volume of Milli-Q water.

49

In Vitro Transcription: 10 mM DNA oligonucleotides coding for the desired RNA sequence downstream of the T7-top in vitro transcription start site were mixed in a 1:1 ratio and annealed by heating at 95oC for 5 min, followed by slowly cooling to room temperature. 8 µL annealed oligonucleotides were added to a transcription reaction containing 10 µL 5x transcription buffer, 2.5 µL NTPs (10 mM ea.), 5 µL α-32P-UTP, and 2.5 µL T7 RNA Polymerase Plus in a final volume of 50 µL.

Reactions were

incubated for 2 hours at 37oC, followed by addition of 3 µL 1 U/µL RQ1 DNase and further incubation for 15 min at 37oC. Resultant RNA was separated by denaturing RNA-PAGE analysis, as described (2.2.7). Denaturing RNA PAGE Purification: RNA samples were separated by electrophoresis through a gel consisting of 7M Urea, 2.5 mL 20x TBE, 10 mL 40% Acrylamide (19:1) in a total volume of 50 mL, with gel polymerisation initiated by addition of 500 µL 10% APS, 50 µL TEMED. α-32P-Labelled RNA was visualized by exposure of the gel to Agfa Curix Ortho X-Ray film. Identified bands corresponding to labelled RNA molecules of interest were excised, crushed in RNA Elution Buffer, and incubated for 45 min at 37oC whilst vortexing.

The resulting samples were passed through cellulose

acetate costar

columns with

glass fiber

pre-filters to

remove residual

polyacrylamide.

1 mL 50% ethanol, 50% isopropanol solution was added to

column flow-through, and the solutions incubated for 30 min at -20oC, followed by centrifugation (16,100 x g; 4oC; 30 min) to pellet RNA. Supernatant was removed, the RNA pellet washed with 500 µL 50% ethanol, 50% isopropanol, and air dried. RNA pellets were analysed by scintillation counting, resuspended in Milli-Q water to a final concentration of 10,000 ccpm/µL, and stored at -20oC.

50

RNA-Binding Analysis: Serial three-fold dilutions of purified and dialysed recombinant StrepTagmHuCsv1-His 8 , prepared as described (2.2.6), yielding ranges of protein concentrations of 0.31-676.00 nM (Fig 3.6, 3.7), or 1.30-2,920.00 nM (Fig 3.9-3.13) in 1x Binding Buffer to a final volume of 50 µL were prepared, in addition to a further

two

samples

containing

50

µL

1x

Binding

Buffer

alone.

10 µL α-32P-Labelled RNA prepared as described (2.2.7) was incubated at 60oC for 5 min, chilled on ice, and 10 µL added to each sample of protein and 1x Binding Buffer alone. Resultant samples were incubated at room temperature for 10 min. 45 µL each protein/RNA binding reaction and one sample of 1x Binding Buffer alone were applied to nitrocellulose filters and washed with 5 mL 1x BB. The remaining 1x Binding Buffer alone sample was applied to a nitrocellulose filter without washing. This procedure was repeated for each RNA being analysed for binding. Nitrocellulose filters were placed into plastic scintillation vials containing scintillation fluid, and the total number of counts remaining on each filter analysed by scintillation counting. Resultant binding data was analysed by plotting fraction radiolabelled RNA bound against log 10 protein concentration, and binding curves fit to data using the equation y = (M o /(M 2 +M o )*M 1 . Where indicated, normalised dissociation constants were calculated by multiplying raw dissociation constants by a CIRBP CT1 normalisation factor, calculated as N = CIRBP exp /CIRBP std , where CIRBP exp represents the measured K d for CIRBP CT1 control RNA, and CIRBP std represents the K d measured for CIRBP CT1 initally (Figure 3.7)

51

2.2.8 – Co-Immunoprecipitation: Preparation of Lysates: Halved P4 mouse brain was placed into a 1.5 mL eppendorf tube on ice, and resuspended in an equal volume of 4oC IP Lysis Buffer containing either 1% NP-40, 1% NP-40 with 0.5% Triton X-100, or 1% NP-40 with 5% Deoxycholate, as indicated in the text. Resuspended tissues were lysed by homogenisation with the plunger of a 1 mL syringe, and lysis completed by passing the resultant sample 10x through a G22 needle. Lysates were incubated on ice for 10 min and cleared by centrifugation (16,100 x g; 4oC; 15 min). Clarified lysates were cleared using SAC/Zysorbin resin, as described below. SAC/Zysorbin Clearing of Lysates: Clarified lysate (1 mL) was supplemented with 50 µL normal rabbit serum and incubated on ice for 1 hour.

Following incubation, lysate was applied to

100 µL packed cell volume S. Aureus Cowan I (SAC), pre-washed with 1 mL IP Lysis Buffer. The resultant resuspension was incubated on ice for 30 min, SAC pelleted by centrifugation (10,000 x g; 4oC; 15 min), and lysate removed for immunoprecipitation analysis. Crosslinking of Antisera to Protein A Sepharose: Protein A sepharose (60 µL slurry in 20% ethanol, ~30 uL packed volume) was collected by centrifugation (10,000 x g; 4oC; 30 sec), washed twice with 200 µL 1x PBS and spun down again (10,000 x g; 4oC; 30 sec). Remaining PBS was aspirated and 100 µL Crosslinking Dilution Buffer added. The resultant solution was incubated at 4oC for 10 min, sepharose beads spun down (10,000 x g; 4oC; 30 sec) buffer aspirated, and 10 µL either Ro patient antiserum or normal human serum, diluted to a final volume of 100 µL in Crosslinking Dilution Buffer, added. Following incubation at 4oC for 1 hour, Protein A sepharose was again spun down 52

by

centrifugation

(10,000

x

4oC;

g;

30

sec),

supernatant

aspirated,

100 µL Crosslinking Dilution Buffer added, and the resultant solution incubated at 4oC for a further 5 min.

Following centrifugation (10,000 x g; 4oC; 30 sec)

sepharose beads were washed

once each

with

200

µL 1x

PBS and

200 µL Crosslinking Wash Buffer, supernatant aspirated, beads resuspended in 500 µL Crosslinking Wash Buffer containing 20 mM dimethyl pimelimidate, and the resultant solution incubated at room temperature for 30 min, with nutation. Sepharose beads were collected by centrifugation (10,000 x g; 4oC; 30 sec), washed with 200 µL Crosslinking Wash Buffer, resuspended in Crosslinking Wash Buffer containing 20 mM dimethyl pimelimidate, and incubated for a further 30 min at room temperature, with nutation. Protein A sepharose was again collected by centrifugation (10,000 x g; 4oC; 30 sec), washed with 200 µL Crosslinking Wash Buffer, twice resuspended in 500 µL 20 mM Tris-Cl pH 7.5 and incubated at room temperature for 10 min.

Sepharose beads were spun down (10,000 x g, 4oC;

30 sec), and washed twice with 200 µL IP Lysis Buffer. Binding of Antisera to Protein A Sepharose (Without Crosslinking): Protein A sepharose (60 µL slurry in 20% ethanol, ~30 uL packed volume) was spun down (10,000 x g; 4oC; 30 sec), and washed twice with 200 µL 1x PBS and spun down again (10,000 x g; 4oC; 30 sec). Remaining PBS was aspirated, 10 µL either Ro patient antiserum or normal human serum added, as indicated in the text, and 1x PBS added to a final volume of 100 µL. Resultant solutions were incubated at 4oC for one hour with nutation.

Sepharose beads were spun down by

centrifugation (10,000 x g; 4oC; 30 sec), supernatant aspirated, and beads washed three times with 200 µL IP Lysis Buffer.

53

Co-Immunoprecipitation: Protein concentrations of cleared lysates prepared as described (2.2.8) were assayed by Bradford assay, and lysates diluted to a final concentration of 500 µg/µL in their corresponding lysis buffers. 150 µL each resultant lysate was added to Protein A sepharose-bound antisera also prepared previously (2.2.8), and the resultant solutions incubated at 4oC with nutation for 90 min.

Protein A

sepharose from each sample was collected by centrifugation (10,000 x g, 4oC, 30 sec), supernatant aspirated, and sepharose beads washed twice with 500 µL 1x PBS, twice with µL IP Lysis Buffer, and twice again with 500 µL 1x PBS. Protein A beads were resuspended in 50 µL 1x non-reducing Novex-LDS load buffer, and incubated at 70oC for 10 min with constant vortexing. Protein A sepharose beads were spun down (10,000 x g; 4oC; 30 sec), and 20 µL each supernatant analysed by 12% SDS-PAGE gel electrophoresis, as described (2.2.5).

54

Chapter 3 – Results:

55

3.1 – Cloning of pET41a(+)-Strep-mHuCsv1-His8: 3.1.1 – Selection of Purification Process for Recombinant mHuCsv1: Previous work within our laboratory has explored the purification of bacteriallyexpressed recombinant mHuCsv1 by means of a glutathione-S-transferase (GST) Tag system, whereby mHuCsv1 was expressed from the pET41a(+)-TEV-mHuCsv1 vector to yield a recombinant protein possessing both an N-terminal GST Tag and a C-terminal His 8 tag. However, application of this protocol resulted in the copurification of a ~50 kDa protein contaminant that could not effectively be separated from full-length recombinant mHuCsv1 (Simpson, 2006).

Mass

spectrometric analysis of this contaminant revealed it to be formed by a truncation event within the GST protein. Consequently, it was decided to remove from the pET41a(+)-TEV-mHuCsv1 plasmid and instead pursue a histidine-tag based purification strategy utilising the eight-histidine tag already present at the C-terminus of the mHuCsv1 recombinant protein. Furthermore, the GST coding sequence was replaced with that of the Strep Tag synthetic peptide. The reasons for this replacement were twofold: first, the Strep Tag is much smaller than the GST Tag utilized previously, being only eight residues in length, compared with 220 residues for GST, and is thus less likely to yield the truncation issues observed with GST-tagged mHuCsv1, whilst also being less likely to sterically interfere with

in vitro RNA-binding experiments. Second, similarly to GST the Strep Tag allows application of the resultant protein to in vitro pull-down experiments, and thus allows for the resultant Strep-mHuCsv1-His 8 to be applied to the investigation of protein-protein

interactions

identified

by

either

yeast

two-hybrid

or

co-immunoprecipiation experiments (Zwicker, Adelhelm, Thiericke, Grabley, & Hanel, 1999).

56

3.1.2 – Replacement of GST Coding Sequence With the Strep Tag ORF: As efficient and effective induction of recombinant protein expression from the pET41a(+) vector had previously been demonstrated (Simpson, 2006), this system was applied to the expression of Strep-mHuCsv1-His 8 recombinant protein in order to determine whether a similar approach could be successfully applied to expression of this new fusion protein.

Thus, the GST coding sequence of

pET41a(+)-TEV-mHuCsv1 was replaced with the Strep Tag ORF to form pET41a(+)-Strep-mHuCsv1-His 8 as outlined in Figure 3.1. Briefly, pET41a(+)-TEVmHuCsv1 plasmid DNA was subjected to restriction endonuclease digestion with Bam HI and Nde I to remove the GST open reading frame. Resultant restriction digest products were analysed by agarose gel electrophoresis and the desired linearised plasmid sequence (lacking the ~680 nt GST coding sequence) (Figure 3.2A, arrow) excised and purified. Following purification, this DNA was ligated to an in vitro synthesised DNA oligonucleotide formed by the annealing of equal molar quantities of StrepTag Fwd and StrepTag Rev single-stranded DNA oligonucleotides (2.1.4), to yield a sequence possessing the Strep Tag open reading

frame

flanked

by

5’

and

3’

single-stranded

DNA

overhangs

complementary to those formed by Bam HI and Nde I restriction digest (2.2.1). Ligated DNA was transformed into E. Coli DH5α as described (2.2.1), and transformants possessing the resultant pET41a(+)-Strep-mHuCsv1-His 8 expression construct selected by growth on solid media containing kanamycin (2.2.1). Following selection of individual transformants, plasmid DNA was isolated and analysed by diagnostic digest with Bsr GI and Nde I.

Subsequent analysis of

restriction products by agarose gel electrophoresis (Figure 3.1B) identified four transformants whose plasmid DNA yielded fragments of the expected size, with DNA sequencing revealing one of these to possess the pET41a(+)-Strep-mHuCsv1His 8 plasmid.

57

His-Tag

His-Tag

mHuCsv1 KanR

GST

mHuCsv1

KanR

BamHI pET41a(+)-TEV-mHuCsv1 6,799bp

His-Tag mHuCsv1

Bam HI/Nde I Digest

KanR pET41a(+)-TEV-mHuCsv1 6,799bp

DNA Ligation

Strep-Tag

pET41a(+)-Strep-mHuCsv1-His8 6,799bp

NdeI lacI lacI

lacI

Bam HI/Nde I Digest NdeI

BamHI Strep-Tag (45bp)

Figure 3.1 – Overview of pET41a(+)-Strep-mHuCsv1-His8 Cloning Strategy:

58

pET41a(+)-TEV-mHuCsv1 was subjected to restriction endonuclease digest with Bam HI and Nde I, as described (2.1.3), to excise the GST-coding sequence. The resultant restriction product was purified by 1% agarose gel electrophoresis, and ligated to a Strep-Tag DNA sequence consisting of two complementary DNA oligonucleotide sequences synthesised and annealed in vitro to produce a single linear DNA fragment with overhanging ends complementary to those produced by Bam HI and Nde I. The resulting DNA plasmid, termed pET41a(+)-Strep-mHuCsv1-His8, encodes a recombinant StrepTag-mHuCsv1-His8 protein.

A.

B. kb

kb

10.0

10.0

3.0

3.0

1.0 1.0 0.5 0.5

0.1

0.1 1

2

3

4

Colony No.

Figure 3.2 – Restriction digest of pET41a(+)-TEV-mHuCsv1 and confirmation of successful ligation: A. pET41a(+)-TEV-mHuCsv1 was subjected to restriction endonuclease digestion with Bam HI and Nde I and digested products separated by 1% agarose gel electrophoresis, as described (2.2.3). Indicated bands were excised and purified as described (2.2.3). B. Purified pET41a(+)-TEV-mHuCsv1 BamHI/Nde I digest DNA and StrepTag oligo DNA E. Coli, individual colonies selected, and plasmid DNA were ligated, transformed into isolated, as described (2.2.3). Isolated plasmid DNA was assayed for the presence of StrepTag DNA by restriction endonuclease digest with Bsr GI and Nde I. Restriction digest products were analysed by 2% agarose gel electrophoresis, as described (2.2.3).

59

3.2 – Expression and Isolation of Recombinant mHuCsv1: 3.2.1 – Purification Under Standard Conditions Results in mHuCsv1 Aggregation: As discussed previously, a protocol for the efficient expression and isolation of GST-tagged recombinant mHuCsv1 from the pET41a(+)-TEV-mHuCsv1 vector has previously been developed within our laboratory, with recombinant mHuCsv1 isolated according to this procedure having been shown to demonstrate RNAbinding activity (Simpson, 2006). Given that this protocol was well-established, had been shown to yield active protein, and that the Strep-mHuCsv1-His 8 protein lacked the GST Tag that yielded the major truncation product observed in previous studies, it was applied to the initial bacterial expression of histidine-tagged recombinant HuCsv1.

Comparison by SDS-PAGE of bacterial culture samples collected immediately prior to and following IPTG-induced protein expression demonstrated successful induction of Strep-mHuCsv1-His 8 protein expression under these conditions (lanes 1 & 2, Figure 3.3A), whilst comparison of crude bacterial lysates to those clarified by ultra-centrifugation revealed that a significant quantity of the recombinant protein was present in soluble form (lanes 3 & 4, Figure 3.3A). However, analysis of individual eluate fractions indicated the presence of a protein contaminant of approximately 75 kDa mass (Figure 3.3B, arrow) present at levels sufficiently high to warrant further investigation.

In an effort to confirm the nature of the 39 kDa purification product, and to further investigate the nature of the 75 kDa contaminant, a single eluate fraction from this purification (Lane 8, Figure, 3.3B) was analysed by three-channel fluorescent western blot with either 3A2 anti-HuA monoclonal antibody (Figure 3.4A), 16A11 anti-HuC/D monoclonal antibody (Figure 3.4B), or ‘Ro’ PND patient antiserum (Figure 3.4C).

Consistent with the reported reactivity of these 60

antibodies, all three reacted with the 39 kDa product, confirming its identity as mHuCsv1. Interestingly, each of the three antibodies also showed strong reactivity toward the 75 kDa contaminant, suggesting that this species contains neuronal Hu protein.

61

A.

B.

kDa

kDa

100 75

100 75

50

50

37

25

mHuCsv1

mHuCsv1

37

25 1

2

3

4

5

6

7

8

9

Elution Fractions

Figure 3.3 – Bacterial expression of recombinant Strep-mHuCsv1-His8 results in successful isolation of recombinant mHuCsv1 in the presence of a coeluting 75kDa protein contaminant: Glycerol stock of E. Coli BL21(DE3) RIPL transformed with pET-41a(+)-Strep-mHuCsv1-His8 was used to inoculate 200 mL LB/Kan liquid medium, grown overnight at 37oC, and 25 mL aliquots used to inoculate six 500 mL LB/Kan liquid cultures, grown at 37oC until OD600 = 0.48. Protein expression was induced with 0.1 mM IPTG, induced cultures grown at 30oC for 5 hours, cells lysed by sonication, protease inhibitors (CPI) added, and the resultant lysates clarified by centrifugation (2.2.6). Imidazole was added to the resultant lysate to a final concentration of 20 mM, and the lysate applied to a single 5mL HisTrap FF column, as described (2.2.6). Column was washed with 20 column volumes 20 mM imidazole, 0.05% Tween-20 in 1x PBS, pH 8.0, followed by 20 column volumes 50 mM imidazole, 0.05% Tween-20 in 1x PBS, pH 8.0. Bound proteins were eluted into eighteen 1 mL fractions with four column volumes 250 mM imidazole, 0.05% Tween-20 in 1x PBS, pH 8.0. A. Preinduction and post-induction culture samples (2 µL), and crude and clarified lysate samples (2 µL) were analysed by 12% SDS-PAGE and coomassie staining (2.2.5). B. 10 µL each of the first nine elution fractions were analysed by 12% SDS-PAGE and coomassie staining (2.2.5). Arrow denotes the 75kDa contaminant.

62

A.

B.

kDa

C.

kDa

3A2

kDa

16A11

75

75

75

50

50

50 mHuCsv1

mHuCsv1

‘Ro’

mHuCsv1

35

35

35

30

30

30

Figure 3.4 – 75kDa contaminant in bacterially expressed recombinant Strep-mHuCsv1-His8 is an Hu protein aggregation product: Samples of purified recombinant Strep-mHuCsv-His8 prepared previously (Fig 3.3B, lane 8) were analysed by 12% SDS-PAGE and western blotting with either A. 3A2 monoclonal antiHu antibody, B. 16A11 anti-HuC/D monoclonal antibody, and C. ‘Ro’ PND patient antiserum, as described (2.2.5).

63

3.2.2 – Purification in the Absence of Protease Inhibitors Significantly Reduces Strep-mHuCsv1-His8 Protein Aggregation: Since the bacterial expression and purification of Strep-mHuCsv1-His 8 (Figure 3.3) had been undertaken in the presence of peptide-based protease inhibitors, which may promote protein aggregation under certain conditions (Whelan, 2008), it was hypothesised that the observed contamination product may be the result of Strep-mHuCsv1-His 8 aggregation.

To further investigate this, expression and

purification recombinant mHuCsv1 was repeated in the absence of protease inhibitors. Examination of the resultant eluate fractions (Figure 3.5A) revealed the successful purification of recombinant Strep-mHuCsv1-His 8 in the absence of any significant protein contaminants, confirming the efficacy of this approach with respect to Strep-mHuCsv1-His 8 purification. Analysis of a single eluate fraction from this purification (Lane 9, Figure 3.5A) by Bradford Assay revealed the concentration of purified Strep-mHuCsv1-His 8 to be 0.11 µg/µL (2.5 µM).

However, previous studies utilising nitrocellulose filter-

binding assays to investigate in vitro RNA binding have typically used protein concentrations of 2.5 µM as a minimum in order to achieve saturation of radiolabelled RNA (Devaux, Colegrove-Otero, & Standart, 2006; Dredge & Darnell, 2003; Huang et al., 2005). Thus, in an effort to increase the final concentration of purified Strep-mHuCsv1-His 8 for use in such assays, the purification procedure was repeated with an elution buffer flow rate of 2 mL/min, reduced from the previous rate of 4 mL/min in an effort to increase the quantity of recombinant protein eluted per unit volume of elution buffer.

Examination of resultant individual

elution fractions by SDS-PAGE and coomassie staining (Figure 3.5B), revealed an apparent increase in mHuCsv1 concentration, which was subsequently confirmed by Bradford Analysis, indicating a final Strep-mHuCsv1-His 8 concentration of 0.15 µg/µL (3.4 µM), sufficient for use in in vitro RNA-binding analyses. 64

A. kDa 98 64 50

mHuCsv1

36

1

2

3

4

5

6

7

8

9

Elution Fractions

B. kDa 98 64 50 mHuCsv1

36

1

2

3

4

5

6

7

8

9

Elution Fractions

Figure 3.5 – Bacterial expression and purification of recombinant Strep-mHuCsv1-His8 in the absence of protease inhibitors reduces formation of the 75 kDa aggregation product: Glycerol stock of E. Coli BL21(DE3) RIPL transformed with pET-41a(+)-Strep-mHuCsv1-His8 was used to inoculate 200 mL LB/Kan liquid medium, grown overnight at 37oC, and 25 mL aliquots used to inoculate six 500 mL LB/Kan liquid cultures, grown at 37oC until OD600 = 0.48. Protein expression was induced with 0.1 mM IPTG, induced cultures grown at 30oC for 5 hours, cells lysed by sonication in the absence of peptide-based protease inhibitors and clarified by centrifugation (2.2.6). Imidazole was added to the resultant lysate to a final concentration of 20 mM, and the lysate applied to a single 5mL HisTrap FF column, as described (2.2.6). Column was washed with 20 column volumes 20 mM imidazole, 0.05% Tween-20 in 1x PBS, pH 8.0, followed by 20 column volumes 50 mM imidazole, 0.05% Tween-20 in 1x PBS, pH 8.0. Bound proteins were eluted into eighteen 1 mL fractions with four column volumes 250 mM imidazole, 0.05% Tween-20 in 1x PBS, pH 8.0 at a flow rate of either A. 4 mL/min or B. 2 mL/min. 10 µL each of the first nine elution fractions were analysed by 12% SDS-PAGE and coomassie staining (2.2.5).

65

3.3 – Investigation of Recombinant mHuCsv1 RNA-Binding Activity: 3.3.1 – Confirmation of Recombinant mHuCsv1 RNA-Binding Activity: Having successfully developed a protocol for the bacterial expression and purification of Strep-mHuCsv1-His 8 , the RNA-binding activity of recombinant mHuCsv1 purified in this manner was examined.

To this end, recombinant

Strep-mHuCsv1-His 8 was expressed and purified according to this protocol (3.2.2) and the successful purification of recombinant protein confirmed by SDS-PAGE analysis of individual elution fractions (Figure 3.6A). Following the pooling and dialysis of selected fractions (lanes 3-5, Figure 3.6A), final protein concentration was determined to be 145 µg/mL (3.38 µM) by Bradford Assay (2.2.6). To investigate the RNA-binding activity of this recombinant protein, nitrocellulose filter-binding assays were performed on several RNA sequences selected from the

D. rerio Cold-Inducible RNA-Binding Protein mRNA. CLIP experiments performed on ~512-cell Danio rerio embryos had previously identified five overlapping CLIP tags within the 3’UTR of this mRNA (Dredge, 2005).

Three overlapping

RNA sequences (CIRBP CT1, CT2, and CT3) from the region spanned by these tags were selected for further analysis (Figure 3.7A). These RNAs were synthesised by

in vitro transcription in the presence of α-32P-UTP and purified by denaturing RNAPAGE (Figure 3.7B). Binding of Strep-mHuCsv1-His 8 to each purified RNA was assayed by in vitro nitrocellulose filter-binding assays.

66

A. kDa 100 75

50 mHuCsv1 37

1

2

3

4

5

6

7

8

9

Elution Fractions

B. kDa 75 50 mHuCsv1

37

25

1

2

3

4

5

6

7

8

9

Elution Fractions

Figure 3.6 – Bacterial expression and purification of recombinant Strep-mHuCsv1-His8 for use in nitrocellulose-RNA filter-binding assays: Glycerol stock of E. Coli BL21(DE3) RIPL transformed with pET-41a(+)-Strep-mHuCsv1-His8 was used to inoculate 200 mL LB/Kan liquid medium, grown overnight at 37oC, and 25 mL aliquots used to inoculate six 500 mL LB/Kan liquid cultures, grown at 37oC until OD600 = 0.48. Protein expression was induced with 0.1 mM IPTG, induced cultures grown at 30oC for 5 hours, cells lysed by sonication in the absence of peptide-based protease inhibitors and clarified by centrifugation (2.2.6). Imidazole was added to each resultant lysates to a final concentration of 20 mM, and each lysate applied to a single 5mL HisTrap FF column, as described (2.2.6). Columns were washed with 20 column volumes 20 mM imidazole, 0.05% Tween-20 in 1x PBS, pH 8.0, followed by 20 column volumes 50 mM imidazole, 0.05% Tween-20 in 1x PBS, pH 8.0. Bound proteins were eluted into eighteen 1 mL fractions with four column volumes 250 mM imidazole, 0.05% Tween-20 in 1x PBS, pH 8.0 at a flow rate of 2 mL/min. 10 µL each of the first nine elution fractions from each purification were analysed by 12% SDS-PAGE and coomassie staining (2.2.5).

67

A. 1602

628 m7G

(AAAA)n

CIRBP

762

849

AAGACCUUUUUUAUUUUUUACUUUUUUGCUUUACGUUGUGUUCCAGUUGAAAUGUUCUAUUAAAGACAGUCCCCUUUUUUUACCCGU

CIRBP CT1

B.

CIRBP CT3

CIRBP CT2

C. 1 100

% Total RNA Bound

0.8 80

CIRBP CT1 CIRBP CT2 CIRBP CT3

0.6 60

Kd = 3.7 ± 0.7 nM 0.4 40

Kd = 88.4 ± 13.8 nM

0.2 20

0 0.1

1

10

100

1000

[HuCsv1] (nM)

Figure 3.7 – HuCsv1 binds to a subset of RNA sequences from CLIP tags identified within the 3’UTR of zebrafish Cold-Inducible RNA-Binding Protein (CIRBP) mRNA: A. Location and sequence composition of CLIP tags identified within the 3’UTR of zebrafish Cold-Inducible RNA-Binding Protein mRNA (Dredge, 2010). Five overlapping sequences were identified (black bars), spanning 87nt within the 3’UTR. Three overlapping RNA sequences (CIRBP CT1, CT2, and CT3; red, green, and blue, respectively) from this region were selected for further analysis. B. Radiolabeled RNAs were synthesized by in vitro transcription in the presence of α-32P-UTP, separated by Urea/PAGE gel electrophoresis , and visualized by radiography (2.2.7). Indicated RNA bands were excised and purified as described (2.2.7). C. Nitrocellulose filter-binding assays were performed (2.2.7), using recombinant Strep-mHuCsv1-His8 shown in Figure 3.6A. Curve fits were extrapolated and dissociation coefficients (Kd) calculated as described (2.2.7).

68

Analysis of the binding curves and dissociation constants calculated for each RNA (Figure 3.7C) revealed that CIRBP CT1 was bound by recombinant HuC with high affinity (K d = 3.7 ± 0.7 nM), comparable to previously reported affinities for HuR RNA-binding (Fialcowitz-White, et al., 2007). By comparison, CIRBP CT2 was bound with relatively lower affinity (K d = 88.4 ± 13.8 nM), whilst CIRBP CT3 was not significantly bound. Thus, recombinant Strep-mHuCsv1-His 8 purified under these conditions was capable of binding RNA with high affinity. Moreover, the binding of recombinant mHuC to a subset of selected RNA sequences suggests that this RNA-binding was specific in nature.

To confirm this result, filter-binding assays were performed on a further four RNA sequences,

selected

from

the

3’UTR

of

the

Heterogeneous

Nuclear

Ribonucleoprotein A/B (hnRNP A/B) mRNA (Figure 3.8A). These RNAs represent four distinct RNA sequences (hnRNP A/B CT1-CT4) derived from CLIP tags identified in experiments performed on 24hpf D. rerio embryos (Dredge, 2005). Following synthesis of these RNAs and purification by denaturing RNA-PAGE (Figure 3.8B), binding of each purified RNA by Strep-mHuCsv1-His 8 was again assayed by in vitro nitrocellulose filter-binding assays.

69

A.

hnRNP CT1 GGGCGCAAAUGUCACUUUUCACAUUUGUUUUAUUUUACUUUUGUUUUUUGUUUUUGUUUUAUUUUCGUGUUUU

1958

2399

2216 2269

1816

m7G

2579

hnRNP A/B

(AAAA)n

2280

2399

CACGGUGAGAUGAAAAGGGAUUUCUUUAUGAUUGUAAGCUUGUCCAUUUUGAAAUCUGUUAGCGGCAAGUCUUGGUGACAUUUUUAUUCUUUUAAUUUGUUUUUUUUUUCCUCUCCCUG

hnRNP CT3

hnRNP CT2 2216

2269

UAGUGUUCCUAGAUGAUUUCCUUUUCUCCCCCUAGUGUCACCUGUGCUUUGUG

hnRNP CT4

C. 100

80

% Total RNA Bound

B.

hnRNP A/B CT1 hnRNP A/B CT2 hnRNP A/B CT3 hnRNP A/B CT4

60

Kd = 84.4 ± 7.9 nM 40 Kd = 28.1 ± 6.0 nM

20

0.1

1

10

100

1000

[HuCsv1] (nM)

Figure 3.8 – HuCsv1 binds with moderate affinity to a subset of RNA sequences from CLIP tags identified within the 3’UTR of zebrafish hnRNP A/B mRNA: A. Location and sequence composition of CLIP tags identified within the 3’UTR of zebrafish hnRNP A/B mRNA (Dredge, 2010). Four CLIP tags (black bars) were identified from three distinct regions within the 3’UTR, covering a total of 244nt. Four discrete RNA sequences (hnRNP CT1, CT2, CT3, and CT4; red, green, blue, and orange) from these CLIP tags were selected for further analysis. B. Radiolabeled RNAs were synthesized by in vitro transcription in the presence of α-32P-UTP, separated by Urea/PAGE gel electrophoresis, and visualized by radiography (2.2.7). Indicated RNA bands were excised and purified as described (2.2.7). C. Nitrocellulose filter-binding assays were performed (2.2.7), using recombinant StrepmHuCsv1-His8 shown in Figure 3.6A. Curve fits were extrapolated and dissociation coefficients (Kd) calculated as described (2.2.7).

70

Analysis of the binding curves and dissociation constants for each RNA (Figure 3.8C) revealed that hnRNP A/B CT1 (K d = 28.1 ± 6.0 nM) was bound by recombinant mHuCsv1 with greater affinity than hnRNP A/B CT3 (K d = 84.4 ± 7.9 nM), whilst hnRNP A/B CT1 and hnRNP A/B CT4 were not significantly bound. These results, taken together with those previously observed for CIRBP, confirmed that recombinant Strep-mHuCsv1-His 8 bound RNA selectively and with affinities in the low nanomolar range, comparable to previous reports (Fialcowitz-White, et al., 2007). However, comparison of these dissociation constants to those previously observed for CIRBP CT1 and CIRBP CT2 (Figure 3.7C), reveals that hnRNP A/B CT1 and CT2 are bound by recombinant mHuCsv1 with significantly lower affinities. Given that the recombinant protein used in the analysis of hnRNP A/B RNAbinding had been stored at -80oC for approximately three weeks, this may suggest that prolonged storage of the isolated recombinant protein resulted in a reduction in RNA-binding activity.

Alternatively, these variations in binding affinity may

merely represent an inherent variability in the affinity of mHuC for the assayed RNA sequences.

71

3.3.2 – CLIP Tag RNA Sequences Do Not Necessarily Represent Sites of Neuronal HuCsv1 Binding: Comparison of the sequence composition of each analysed RNA sequence revealed that the highest-affinity RNA interactors from each experiment were those possessing the most-uridine rich RNA sequences.

Thus, CIRBP CT1 was

significantly more uridine rich (70% U) than either CIRBP CT2 (47%) or CIRBP CT3 (43%), whilst hnRNP A/B CT1 and CT3 were also significantly more uridine rich (79% U and 65% U, respectively) than either hnRNP A/B CT2 (50% U) or CT4 (44% U).

More strikingly, both CIRBP CT1 and hnRNP A/B CT1 appeared to

possess a common sequence motif, consisting of short interspersed uridine-rich sequences interrupted by one or more non-uridine nucleotides. Based on these observations, we hypothesised that this motif may represent a funadmental determinant of high-affinity RNA-binding by HuC. To further investigate the role of such interspersed uridine-rich motif with respect to mHuCsv1 binding, a number of additional CLIP-identified RNA sequences were selected for analysis in mHuCsv1 nitrocellulose filter-binding assays (summarized in Table 3.1).

For these analyses, recombinant Strep-mHuCsv1-His 8 was expressed and purified as described previously (3.2.2). Analysis of individual elution fractions by SDSPAGE (Figure 3.6B) revealed the presence of some minor protein contamination, whilst examination of pooled elution fractions (lanes 4-6, Figure 3.6B), by Bradford Assay revealed a final recombinant protein concentration of 145 µg/mL (14.6 µM). The higher apparent concentration of recombinant protein compared to that isolated previously (Figure 3.6A) is most likely due to the presence of higher levels of protein contamination than those observed in previous protein samples (cf. Figure 3.6B, Figure 3.6A).

Based on this variation in recombinant protein

concentration and purity, and the apparent variability in Strep-mHuCsv1-His 8 RNA-binding activity observed in previous filter-binding assays (cf. Fig 3.7C, Fig 72

3.8C), a CIRBP CT1 control RNA was included in each subsequent set of filterbinding assays performed. This control enables dissociation constants measured for each

RNA-protein interaction to be normalized to that of CIRBP CT1

(Figure 3.7), and thus allows for the direct comparison of dissociation constants derived from discrete RNA-binding experiments.

To begin a systematic analysis RNA-binding by recombinant HuC, two RNA sequences were selected from the 3’UTR of the D. Rerio beta-Catenin 2 mRNA (Figure 3.9A), consisting of a single CLIP tag identified in experiments on ~512-cell embryos (beta-Catenin CT1) and an additional uridine-rich sequence (beta-Catenin CT2) located adjacent and 3’ to this CLIP tag.

Such CLIP tag-

adjacent uridine-rich sequences have been observed in CLIP experiments performed on both ~512 cell and 24 hpf zebrafish embryos (Dredge, 2005), as well as those performed on E18 murine neural tissue (Jensen, 2004), and may represent nearby sites of neuronal Hu binding yet to be isolated by CLIP. Thus, in order to investigate the relationship between individual CLIP tags and sites of Hu RNAbinding, in addition to elucidating the sequence determinants of HuC RNA-binding activity, comparison of recombinant mHuCsv1 binding, both to these sequences and adjacent CLIP tags was undertaken. To this end, each RNA was synthesised by

in vitro transcription, purified by denaturing RNA-PAGE (Figure 3.9B), and subjected to nitrocellulose filter-binding assays using previously purified recombinant Strep-mHuCsv1-His 8 protein (Figure 3.6B). Interestingly, comparison of the binding curves and CIRBP CT1-normalized dissociation constants for each RNA-protein interaction (Figure 3.9C) revealed that both RNAs bound recombinant mHuCsv1 with high affinity. Moreover, the CLIP tag adjacent betaCatenin CT2 sequence

(K d = 3.8 ± 1.6 nM) was bound with two-fold higher

affinity than the CLIP tag itself (beta-Catenin 2 CT1, K d = 8.3 ± 3.5 nM), suggesting that additional neuronal Hu binding sites may exist in this and other mRNAs, beyond those already identified by CLIP. 73

A.

β-catenin CT1 ACGGACUCUCGGUCUGAAGACGGAACUCUUUUACUCUUUAUUUUCAUCCUCUUCGGUUAU

2705

2765

3537

2501

β-catenin

m7G

(AAAA)n

2766

2846

UCCUCCUCCUCCUCCUCCUCAUGGUACUGACCCGAGCUUUGUUUUUAUCCUGCUCUUUUCUUUGUUUGUGUAGUGUUUAGU

Β-catenin CT2

B.

C. 100 CIRBP CT1 Control beta-Catenin 2 CT1 beta-Catenin 2 CT2

% Total RNA Bound

80 Kd = 27.7 ± 4.7 nM Kd(norm) ~ 3.8 nM 60

40

Kd = 70.6 ± 7.7 nM Kd(norm) ~ 7.7 nM

20

1

10

100

1000

104

[HuCsv1] (nM)

Figure 3.9 – HuCsv1 binds with high affinity to multiple RNA sequences within and adjacent to a single CLIP tag identified within the 3’UTR of zebrafish β-Catenin mRNA: A. Location and sequence composition of CLIP tags identified within the 3’UTR of zebrafish β-catenin mRNA (Dredge, 2010). A single 60 nt CLIP tag (black bar) was identified within the 3’UTR . Two discrete RNA sequences (β-catenin CT1, CT2; red and green) were selected for further analysis, representing the CLIP tag sequence and a single U-rich RNA sequence directly adjacent and 3’ to the CLIP tag B. Radiolabeled RNAs were synthesized by in vitro transcription in the presence of α-32P-UTP, separated by Urea/PAGE gel electrophoresis, and visualized by radiography (2.2.7). Indicated RNA bands were excised and purified as described (2.2.7). C. Nitrocellulose filter-binding assays were performed (2.2.7), using recombinant Strep-mHuCsv1-His8 shown in Figure 3.6B. Curve fits were extrapolated and dissociation coefficients (Kd) calculated as described (2.2.7). CIRBP CT1 RNA was used as a normalization control. Normalized dissociation coefficients were calculated as described (2.2.7) by comparing the measured dissociation constant for CIRBP CT1 RNA to that measured previously (Fig 3.7).

74

To further examine the occurrence of neuronal Hu binding sites not identified by CLIP, two RNA sequences from the 3’UTR of the eukaryotic Initiation Factor 4G subunit 2a (eIF4G2a) mRNA of D. Rerio were selected for filter-binding analysis (Figure 3.10A), consisting of a single CLIP tag (eIF4G2a CT1) identified in experiments on 24 hpf zebrafish embryos (Dredge, 2005) in addition to a single uridine-rich sequence (eIF4G2a CT2) located directly adjacent and 5’ to this CLIP tag.

Following synthesis by in vitro transcription, and purification by

denaturing RNA-PAGE (Figure 3.10B), these RNAs were subjected to nitrocellulose filter-binding assays using previously purified recombinant Strep-mHuCsv1-His 8 protein (Figure 3.6B).

Surprisingly, comparison of the binding curves and

normalized dissociation constants for each RNA (Figure 3.10C) revealed that whilst the CLIP tag-adjacent eIF4G2a CT2 sequence bound mHuCsv1 with high affinity (K d = 4.8 ± 2.2 nM), the CLIP tag sequence itself (eIf4G2a CT1) did not demonstrate any significant interaction with the recombinant protein.

Taken

together with the previous observations on beta-Catenin, these results suggest not only that additional Hu-bound RNA sequences have yet to be identified, but also that RNA sequences identified by CLIP may not necessarily represent sites of direct neuronal Hu binding, which may in some cases lie adjacent to one or more isolated CLIP tags.

75

A.

eIF4G2a CT1 UAUUUGCAAGAGCCUUUUCUGCUGCUGUUCCUUUCAGCACAAAACUC

3272

3225

4023

2978

m7G

eIF4G2a

(AAAA)n

3175

3224

CUGCAUCUCUUUUUGUUUUGUUUUUUUUAUUCUACAUUUUAAGAAACCUG

eIF4G2a CT2

B.

C. 100 CIRBP CT1 Control eIF4G2a CT1 eIF4G2a CT2

80

% Total RNA Bound

Kd = 40.8 ± 6.0 nM Kd(norm) ~ 4.8 nM 60

40

20

1

10

100

1000

10

4

[HuCsv1] (nM)

Figure 3.10 – HuCsv1 binds with high affinity to an RNA sequence located adjacent to a CLIP tag identified within the 3’UTR of zebrafish eIF4G2a mRNA, but not to the CLIP tag itself: A. Location and sequence composition of CLIP tags identified within the 3’UTR of zebrafish eIF4G2a mRNA (Dredge, 2010). A single 60 nt CLIP tag (black bar) was identified within the 3’UTR. Two discrete RNA sequences (eIF4G2a CT1, CT2; red and green) were selected for further analysis, representing the CLIP tag itself and a single U-rich RNA sequence directly adjacent and 5’ to the CLIP tag B. Radiolabeled RNAs were synthesized by in vitro transcription in the presence of α-32P-UTP, separated by Urea/PAGE gel electrophoresis, and visualized by radiography (2.2.7). Indicated RNA bands were excised and purified as described (2.2.7). C. Nitrocellulose filter-binding assays were performed (2.2.7), using recombinant Strep-mHuCsv1-His8 shown in Figure 3.6B. Curve fits were extrapolated and dissociation coefficients (Kd) calculated as described (2.2.7). CIRBP CT1 RNA was used as a normalization control. Normalized dissociation coefficients were calculated as described (2.2.7) by comparing the measured dissociation constant for CIRBP CT1 RNA to that measured previously (Fig 3.7).

76

In order to examine this hypothesis in more detail, RNA binding by recombinant HuC was analysed in the context of an mRNA possessing multiple CLIP tags within its 3’UTR. Such an mRNA is less likely to possess additional Hu binding sites within its 3’UTR that have yet to be identified by CLIP, and should thus provide a more complete picture of the relationship between specific CLIP tags and sites of neuronal Hu binding. Thus, four RNA sequences were selected from the 3’UTR of the D. rerio sfrs1l mRNA (Figure 3.11A), consisting of three CLIP tags identified from experiments performed on 512-cell (sfrs1l CT2, and CT3) and 24 hpf (sfrs1l CT1) zebrafish embryos, in addition to a single uridine-rich RNA sequence (sfrs1lCT4) located directly adjacent and 5’ to the first CLIP tag within the 3’UTR. Each of these RNAs were synthesised by in vitro transcription, purified by denaturing RNA-PAGE (Figure 3.11B, D), and subjected to nitrocellulose filterbinding assays using previously purified recombinant Strep-mHuCsv1-His 8 protein (Figure 3.6B).

Subsequent analysis of the binding curves and normalized

dissociation constants for each RNA (Figure 3.11C, E) revealed that whilst both sfrs1l CT2 and sfrs1l CT3 bound Strep-mHuCvs1-His 8 with high affinity (K d = 3.6 ± 2.9 nM and 6.5 ± 6.4 nM, respectively), sfrs1l CT1 was not significantly bound by the recombinant protein, despite being directly identified by CLIP. Moreover, sfrs1l CT4, located directly adjacent to the unbound sfrs1l CT1 CLIP tag RNA, was bound with high affinity by recombinant mHuCsv1 (K d = 6.7 ± 5.9 nM), further supporting the hypothesis that CLIP tags do not necessarily represent the site of physical mHuCsv1 binding to a given mRNA.

77

A.

sfrs1l CT2

sfrs1l CT1

AACCCCAUGUUUUAAUCACCCCCUGUUUCGUGAGACAGUUGAUCUUAUGAAUGUUGUCUUUUUUUUUUUUUUUUCCCCUCCCUCU

1000

920

1255

850

m7G

sfrs1l

(AAAA)n

1084

919 1040

856

CCACUCUCACCUCAUUGGUGGAGCCCUUGAGUCGUUUUGUUUUACAGUCUUUACAAAUCACG UUGUUUUUCCCCCCCAUCCAUUUUUUUUUUUUUCAAUACAUAG

sfrs1l CT4 sfrs1l CT3

B.

C. 100 CIRBP CT1 Control sfrs1l CT1

% Total RNA Bound

80

60

40

20

1

10

100

1000

104

[HuCsv1] (nM)

Figure 3.11 – HuCsv1 binds with high affinity both to a subset of RNA sequences from identified zebrafish sfrs1l CLIP tags, and to an adjacent 5’ U-rich sequence Figure continues on next page.

78

E. 100 CIRBP CT1 Control sfrs1l CT2 sfrs1l CT3 sfrs1l CT4

80

% Total RNA Bound

D.

Kd = 87.0 ± 17.9 nM Kd(norm) ~ 3.6 nM 60 Kd = 156.3 ± 43.6 nM Kd(norm) ~ 6.5 nM 40

20 Kd = 161.6 ± 43.7 nM Kd(norm) ~ 6.7 nM 1

10

100

1000

10

4

[HuCsv1] (nM)

Figure 3.11 – HuCsv1 binds with high affinity both to a subset of RNA sequences from identified zebrafish sfrs1l CLIP tags, and to an adjacent 5’ U-rich sequence (cont.): A. Location and sequence composition of CLIP tags identified within the 3’UTR of zebrafish sfrs1l mRNA (Dredge, 2010). A total of three CLIP tags (black bars) were identified within the 3’UTR, covering a total of 124nt. Four discrete RNA sequences (sfrs1l CT1, CT2, CT3, and CT4; red, green, blue, and orange) were selected for further analysis, representing the three CLIP tags and a single U-rich RNA sequence directly adjacent and 5’ to the first CLIP tag B, D. Radiolabeled RNAs were synthesized by in vitro transcription in the presence of α-32P-UTP, separated by Urea/PAGE gel electrophoresis, and visualized by radiography (2.2.7). Indicated RNA bands were excised and purified as described (2.2.7). C, E. Nitrocellulose filter-binding assays were performed (2.2.7), using recombinant StrepmHuCsv1-His8 shown in Figure 3.6B. Curve fits were extrapolated and dissociation coefficients (Kd) calculated as described (2.2.7). CIRBP CT1 RNA was used as a normalization control. Normalized dissociation coefficients were calculated as described (2.2.7) by comparing the measured dissociation constant for CIRBP CT1 RNA to that measured previously (Fig 3.7).

79

As discussed previously (1.2.5), whilst CLIP tags identified from experiments performed on 512-cell and 24hpf zebrafish embryos are typically uridine-rich in nature, those identified from CLIP experiments performed on E18 murine neural tissue are more often cytidine-rich in nature, with one or more uridine-rich RNA sequences frequently located adjacent to each identified CLIP tag. Based on the marked preference for uridine-rich RNA sequences demonstrated by recombinant HuC in RNA-binding analyses performed thus far, in addition to the surprising result that HuCsv1 often binds uridine-rich RNA sequences adjacent to individual CLIP tags, rather than tags themselves, it was hypothesised that cytidine-rich sequences identified from CLIP tag experiments performed on mice may not represent sites of direct Hu binding.

To investigate this hypothesis, recombinant

mHuCsv1 RNA-binding to a number of cytidine-rich CLIP tags and adjacent uridine-rich RNA sequences identified from previous CLIP experiments performed on E18 murine neural tissue (Jensen, 2004) was examined.

To this end, two RNA sequences were selected from the 3’UTR of the Cofilin 1 mRNA of Mus musculus (Figure 3.12A), consisting of a single cytidine-rich CLIP tag (mCofilin CT1) and an additional uridine-rich RNA sequence (mCofilin CT2) located directly adjacent and 5’ to this sequence.

Consistent with previous

observations, analysis of the binding curves and normalized dissociation constants calculated for each RNA (Figure 3.12C, 3.12E) revealed that the uridine-rich mCofilin CT2 sequence was bound by Strep-mHuCsv1-His 8 with high affinity (K d = 7.0 ± 5.6 nM), whilst the CLIP tag sequence itself was not bound by recombinant HuC.

80

A.

mCofilin 1 CT1 ACACCCCUACUCCGUAUCCCUCCCCAUCCCAUGCUGCCAACU

883

842

1078

541

m7G

mCofilin 1

(AAAA)n

796

841

UCCUAGGCACCCAGUUUGGGGGGAGCCUGUAUUUUUUUUUUUAACG

mCofilin 1 CT2

B.

C. 100 CIRBP CT1 Control mCofilin CT1

% Total RNA Bound

80

60

40

20

1

10

100

1000

104

[HuCsv1] (nM)

Figure 3.12 – HuCsv1 binds with high affinity to an RNA sequence located adjacent to a CLIP tag identified within the 3’UTR of murine Cofilin 1 mRNA, but not to theCLIP tag itselff: Figure continues on next page.

81

D.

E. 100 CIRBP CT1 Control mCofilin1 CT2

% Total RNA Bound

80 Kd = 168.0 ± 34.8 nM Kd(norm) ~ 7.01 nM 60

40

20

1

10

100

1000

104

[HuCsv1] (nM)

Figure 3.12 – HuCsv1 binds with high affinity to an RNA sequence located adjacent to a CLIP tag identified within the 3’UTR of murine Cofilin 1 mRNA, but not to theCLIP tag itself (cont.): A. Location and sequence composition of CLIP tags identified within the 3’UTR of murine Cofilin 1 mRNA (Dredge, 2010). A single 41nt CLIP tag (black bar) was identified within the 3’UTR. Two adjacent RNA sequences (mCofilin 1 CT1, CT2; red and green) were selected for further analysis, representing the CLIP tag and a single U-rich RNA sequence directly adjacent and 5’ to the CLIP tag B, D. Radiolabeled RNAs were synthesized by in vitro transcription in the presence of α-32P-UTP, separated by Urea/PAGE gel electrophoresis, and visualized by radiography (2.2.7). Indicated RNA bands were excised and purified as described (2.2.7). C, E. Nitrocellulose filter-binding assays were performed (2.2.7), using recombinant StrepmHuCsv1-His8 shown Figure 3.6B. Curve fits were extrapolated and dissociation coefficients (Kd) calculated as described (2.2.7). CIRBP CT1 RNA was used as a normalization control. Normalized dissociation coefficients were calculated as described (2.2.7) by comparing the measured dissociation constant for CIRBP CT1 RNA to that measured previously (Fig 3.7).

82

In an effort to confirm this result, a further two RNA sequences, from the 3’UTR of the murine Hig1 mRNA, were selected for analysis (Figure 3.13A), consisting of a single cytidine-rich CLIP tag (mHig1 CT1) in addition to a single uridine-rich RNA sequence (mHig1 CT2) located directly adjacent and 5’ to the CLIP tag. Analysis of the binding curves and normalized dissociation constants calculated for these RNAs (Figure 3.13C) demonstrated that mHig1 CT1 was not significantly bound by recombinant mHuCsv1, whilst the adjacent uridine-rich mHig1 CT2 sequence bound with moderate affinity (K d = 26.3 ± 18.8 nM). Taken together with the previous results for sequences from the 3’UTR of the Cofilin 1 mRNA, in addition to those of the previously described RNA binding analyses performed on zebrafish CLIP tags, this strongly implies that such tags do not necessarily represent sites of physical RNA-binding by murine HuCsv1. However, as each of these assays has been performed only once, further study is required to confirm the statistical significance of these results, in addition to investigating whether those CLIP tags that are not bound by HuCsv1 interact with other neuronal Hu proteins and splice variants.

83

A.

mHig1 CT1 CUUCCACCGCUGCCCCCCUCCCGCUCCCAAAAUCCC

532

497

612

353

m7G

mHig1

(AAAA)n

533

577

AGGUUCAUUUCAGUUGGGUUGCAUGCUUCUAUUUGUGAUGCGUCC

mHig1 CT2

B.

C. 100 CIRBP CT1 Control mHig1 CT1 mHig1 CT2

% Total RNA Bound

80

Kd = 257.5 ± 96.8 nM Kd(norm) ~ 26.3 nM

60

40

20

1

10

100

1000

104

[HuCsv1] (nM) Figure 3.13 – HuCsv1 binds with high affinity to an RNA sequence located adjacent to a CLIP tag identified within the 3’UTR of murine Hig1 mRNA, but not to the CLIP tag itself: A. Location and sequence composition of CLIP tags identified within the 3’UTR of murine Hig1 mRNA (Dredge, 2010). A single 37 nt CLIP tag (black bar) was identified within the 3’UTR . Two discrete RNA sequences (mHig1 CT1, CT2; red and green) were selected for further analysis, consisting of the CLIP tag itself and a single U-rich RNA sequence directly adjacent and 3’ to the CLIP tag B. Radiolabeled RNAs were synthesized by in vitro transcription in the presence of α-32P-UTP, separated by Urea/PAGE gel electrophoresis, and visualized by radiography (2.2.7). Indicated RNA bands were excised and purified as described (2.2.7). C. Nitrocellulose filter-binding assays were performed (2.2.7), using recombinant Strep-mHuCsv1-His8 shown in Figure 3.6B. Curve fits were extrapolated and dissociation coefficients (Kd) calculated as described (2.2.7). CIRBP CT1 RNA was used as a normalization control. Normalized dissociation coefficients were calculated as described (2.2.7) by comparing the measured dissociation constant for CIRBP CT1 RNA to that measured previously (Fig 3.7).

84

3.3.3 – Neuronal HuCsv1 Preferentially Binds RNA Sequences Possessing Multiple Uridine-Rich Sequence Motifs:

In addition to examining the relationship between individual CLIP tags and RNA binding by HuC, comparison of binding and sequence data from each filterbinding analysis (summarized in Table 3.1) also allows for an investigation of the sequence-specificity of mHuCsv1 RNA-binding. For these analyses the previously described hnRNP A/B filter-binding assays (Figure 3.8) were excluded, as the lack of a CIRBP CT1 normalization control in these assays makes direct comparison of data between this and other experiments problematic.

Comparison of normalised binding affinities to overall nucleotide frequency revealed a strong preference for uridine-rich sequences (Figure 3.14A), with uridine

nucleotides

comprising

55-70%

of

high-affinity

interactors

(beta-Catenin 2 CT2, sfrs1l CT2, CIRBP CT1, eIF4G2a CT2) compared to approximately 20-35% of non-binding RNAs (eIF4G2a CT1, sfrs1l CT1, mCofilin CT1, mHig1 CT1). However, prevalence of uridine nucleotides by itself was not sufficient to predict RNA-binding, as sfrs1l CT3 (K d = 6.5 ± 6.4 nM), sfrs1l CT4 (K d = 6.7 ± 5.9 nM), and mCofilin1 CT2 (K d = 7.0 ± 5.6 nM), demonstrated similar dissociation constants, despite marked differences in sequence composition (53.5% U, 35.5% U, and 39.1 % U, respectively).

85

RNA Species

Sequence (5’ to 3’)

Kd (nM)

β-Catenin 2 CT2

GAGCUUUGUUUUUAUCCUGCUCUUUUCUUUGUUUGUGUAGUGUUUAGU

~ 3.5

sfrs1l CT2

UUUCGUGAGACAGUUGAUCUUAUGAAUGUUGUCUUUUUUUUUUUUUUUUCCCCUCCCUCU

~ 3.6

CIRBP CT1

ACCUUUUUUAUUUUUUACUUUUUUGCUUUA

eIF4G2a CT2

CUGCAUCUCUUUUUGUUUUGUUUUUUUUAUUCUACAUUUUAAGAAACCUG

~ 4.8

sfrs1l CT3

UUGUUUUUCCCCCCCAUCCAUUUUUUUUUUUUUCAAUACAUAG

~ 6.5

sfrs1l CT4

CCACUCUCACCUCAUUGGUGGAGCCCUUGAGUCGUUUUGUUUUACAGUCUUUACAAAUCACG

~ 6.7

mCofilin 1 CT2

UCCUAGGCACCCAGUUUGGGGGGAGCCUGUAUUUUUUUUUUUAACG

~ 7.0

β-Catenin 2 CT1

ACGGACUCUCGGUCUGAAGACGGAACUCUUUUACUCUUUAUUUUCAUCCUCUUCGGUUAU

~ 8.3

mHig1 CT2

AGGUUCAUUUCAGUUGGGUUGCAUGCUUCUAUUUGUGAUGCGUCC

~ 26.3

hnRNP A/B CT1

AUUUGUUUUAUUUUACUUUUGUUUUUUGUUUUUGUUUUAUUUU

28.1 ± 6.0

Nucleolin CT1

ACUUUAUUUCAAACCCAAUUUUUAUUUAAUUUGUUUUUUUAAGUUUGCUUUAG

64.7 ± 13.9

Nucleolin CT2

AACAGUUUCUUUUCCCAUUUCUCUUUUUUGUU

79.7 ± 9.9

hnRNP A/B CT3

GGUGACAUUUUUAUUCUUUUAAUUUGUUUUUUUUUUCCUC

84.4 ± 7.9

CIRBP CT2

CUUUACGUUGUGUUCCAGUUGAAAUGUUCU

88.4 ± 13.8

3.7 ± 0.7

Table 3.1 – Comparison RNA sequences bound by recombinant mHuCsv1 in vitro reveals a preference for uridine-rich RNA sequences. RNA sequences bound by recombinant mHuCsv1 in vitro are arranged according their measured and normalized (where applicable) dissociation constants.

86

A.

B.

Kd

No. UU Motifs

C.

87

A. RNAs assayed for mHuCsv1 binding (excl. hnRNP A/B) were arranged in order of decreasing binding affinity (Kd, shown in nM, top) and the sequence composition of each RNA plotted as percentage uridine (red), guanine (black), cytosine (blue), and adenine (green). B, C. RNAs assayed for mHuCsv1 binding (excl. hnRNP A/B) were arranged in order of decreasing binding affinity and analysed for the presence of B. UU or C. UUU motifs.

No. UU Motifs

Figure 3.14 – Recombinant mHuCsv1 preferentially binds RNA sequences possessing multiple uridine-rich sequence motifs:

Kd

In addition to being generally uridine-rich, high-affinity mHuCsv1 RNA interactors also commonly possess short, interspersed uridine-rich motifs.

To further

investigate the role of these sequences in determining mHuCsv1 RNA-binding affinity, the number of di- and tri-uridine motifs (UU, and UUU) present in each assayed RNA was quantified. Interestingly, the highest-affinity RNA interactors also possess the highest prevalence of di-uridine (Figure 3.14B) motifs, with betaCatenin 2 CT2, sfrs1l CT2, CIRBP CT1, eIF4G2a CT2 each possessing eight or more such sequences.

By comparison, RNAs not bound by recombinant mHuCsv1

possess the lowest prevalence of such motifs, with eIF4G2a CT1, sfrs1l CT1, mCofilin CT1, mHig1 CT1 all possessing five or fewer.

Correspondingly,

quantification of tri-uridine motifs (Figure 3.14C) reveals a similar trend, with beta-Catenin 2 CT2, sfrs1l CT2, CIRBP CT1, and eIF4G2a CT2 each possessing five or more UUU motifs, whilst eIF4G2a CT1, sfrs1l CT1, mCofilin CT1, mHig1 CT1 each possess three or fewer. Taken together, these observations suggest a role for short uridine-rich motifs in neuronal Hu binding.

This is consistent with the

putative x-U/C-U-x-x-U/C-U-U/C consensus sequence derived from previous x-ray crystallography studies on the first two RRMs of murine HuD (Wang & Tanaka Hall, 2001).

However, identification of a definitive mHuCsv1 RNA-binding consensus

will require the analysis of a significantly larger set of putative target sequences, in addition to the application of binding site boundary-mapping experiments to identify the precise site of neuronal Hu binding within each RNA.

88

3.4 – Identification of Neuronal Hu Protein Interactors: 3.4.1 – Investigation of Neuronal Hu Protein Interactors Previously Identified by Yeast Two-Hybrid: A number of putative neuronal Hu interactors had previously been identified within our laboratory from a yeast two-hybrid assay undertaken with the assistance of the University of Adelaide Biochemistry III practical class. For this assay, the coding sequences corresponding either to the spacer domain of mHuCsv2 alone or the spacer domain together with the third RRM were cloned into the pGBKT7 plasmid to yield the pGBKT7-mHuCsv2spacer and pGBKT7-mHuCsv2spacer_RRM vectors (Dredge, 2009), which express the GAL4 DNA-binding domain fusion proteins described in Figure 3.15. Each of these vectors was transformed into yeast strain AH109 together with a library of murine adult spinal cord cDNAs fused to the open reading frame of the GAL4 transcriptional activation domain, expressed from the pAD-GAL4-2.1 vector (Jensen, 2009). Following growth on media lacking tryptophan, leucine, and histidine to select for transformants possessing putative mHuCsv2 protein interactors, plasmid DNA was isolated and sequenced, and putative interactors identified by BLAT analysis of sequenced DNA. From this screen, a total of seven putative HuCsv2 protein interactors were identified, as summarized in Table 3.2. Two of these proteins, Immt and Gaa, are expressed exclusively in mitochondria (John et al., 2005) and lysosomes (Hoefsloot, Hoogeveen-Westerveld, Reuser, & Oostra, 1990), respectively, and were thus excluded from further analysis. Of the remaining five proteins, four were chosen for further analysis: CUG-Binding Protein 2 (CUG-BP2), which has several reported

roles

in

post-transcriptional

gene

regulation,

including

mRNA

stabilisation and alternative splicing (Dembowski & Grabowski, 2009; Sureban et al., 2007); Hsp70-associated protein Bag4, and Melanoma-Associated Antigen D1 (MAGE-D1), both of which have been described as regulators of apoptosis, and 89

may play roles in neuronal differentiation (Bertrand et al., 2008; Kendall et al., 2005; Takada et al., 2003); and Ribosomal Protein SA, a known component of the 40S small ribosomal subunit, in addition to its previously reported function as a laminin receptor (Nelson et al., 2008).

90

spacer H2N

GAL4 DNA-Binding

RRM3

CO2H

HNS spacer H2N

GAL4 DNA-Binding

CO2H

HNS

Figure 3.15 – Identification of murine HuCsv2 interacting proteins by yeast two-hybrid: Graphical representation of the GAL4 DNA-binding domain fusion proteins used as ‘bait’ in the HuCsv2/adult mouse spinal cord cDNA yeast two-hybrid experiment. Expression constructs were prepared in our laboratory (Dredge, 2009) and used to screen an adult mouse cDNA library within our laboratory, withthe assistance of the Adelaide University Biochemistry III class (Jensen, 2010).

Table 3.2 – List of putative mHuCsv2 interactors identified by yeast two-hybrid: Screen

Gene Symbol

Description

Ensembl

Spacer-RRM3

CUG-BP2

CUG-binding protein 2

ENSMUSG00000002107

Spacer-RRM3

Bag4

Hsp70-associated protein Bag4

ENSMUSG00000037316

Spacer-only

MAGE-D1

Melanoma-associated antigen D1

ENSMUSG00000025151

Spacer-only

RPSA

Ribosomal protein SA

ENSMUSG00000032518

Spacer-only

Bat3

HLA-B-associated transcript 3

ENSMUSG00000024392

Spacer-only

Immt

Inner mitochondrial membrane protein Immt

ENSMUSG00000052337

Spacer-only

Gaa

Lysosomal alpha-glucosidase

ENSMUSG00000025579

91

To begin examining each of these putative interactors in more detail, we sought to confirm the interaction between these proteins and the HuCsv2 spacer-only and/or spacer-RRM3 bait fusion proteins in yeast. To this end, pAD-GAL4-2.1 plasmid constructs expressing each protein fused to the GAL4 transcriptional activation domain were transformed into yeast strain y2hgold, together with pGBKT7 plasmid DNA expressing either the mHuCsv2 spacer alone or spacer-RRM3 baits described previously (Figure 3.15).

Resultant transformants were restruck onto selective

media lacking either histidine or adenine.

Comparison of individual restruck

transformants to positive controls transformed with pGADT7-T, expressing SV40 large T antigen fused to the GAL4 transcriptional activation domain, together with pGBKT7-53, expressing the known SV40 large T-interacting protein p53 fused to GAL4 DNA-binding domain (Iwabuchi, Li, Bartel, & Fields, 1993), revealed that yeast co-transformed with either HuCsv2 spacer-only bait or spacer-RRM3 bait, and either CUGB-BP2, Bag4, RPSA, or Maged1 fusion proteins, showed strong growth on media lacking histidine, comparable to that observed for the SV40/p53 positive control (Figure 3.16). However, examination of negative control yeast, cotransformed with both SV40 large T antigen and laminin protein, also demonstrated growth in the absence of histidine, likely due to basal HIS3 reporter gene expression, as has been reported in previous studies (Iyer & Struhl, 1995). Thus, the observed growth of the experimental bait/prey transformants on media lacking histidine could not be taken to confirm the presence of interaction between HuC and each of the identified interactors. Moreover, the level of growth observed for each of these experimental transformants plated on media lacking both histidine and adenine was comparable to that observed for either the SV40 large T/laminin or bait-only negative control transformants. Consequently, these retransformations failed to conclusively demonstrate any interaction between either the HuCsv2 spacer alone or spacer-RRM3 bait proteins and each of the putative two-hybrid interactors.

92

-Trp -His

-Trp -Ade

p53

-Trp -His

-Trp-Ade

-Trp -Leu -His

-Trp -Leu -Ade

-Trp -His

-Trp -Ade

-Trp -Leu -His

-Trp -Leu -Ade

empty vector

laminin

-Leu -His

-Leu -Ade

-Trp -Leu -His

-Trp -Leu -Ade

empty vector/ CUG-BP2

SV40 large T empty vector/ Bag4

p53/ SV40 large T

empty vector/ RPSA

laminin/ SV40 large T

empty vector/ Maged1

-Trp -His

-Trp -Ade

spacer-RRM3

spacer-only

-Trp -Leu -His

-Trp -Leu -Ade

spacer-only/ CUG-BP2

spacer-RRM3/ CUG-BP2

spacer-only/ Bag4

spacer-RRM3/ Bag4

spacer-only/ RPSA

spacer-RRM3/ RPSA

spacer-only/ Maged1

spacer-RRM3/ Maged1

93

Figure 3.16 – Putative neuronal Hu protein interactors show very weak reporter activation following yeast retransformation: Vectors expressing control proteins (p53, laminin, SV40 large T antigen), HuCsv2 bait proteins (spacer-only, spacer-RRM3), or putative interactor proteins (CUG-BP2, Bag4, RPSA, Maged1) were transformed into competent y2hgold yeast in the indicated combinations, as described (2.2.2). Transformed yeast were grown on SC-Leu, SC-Trp, or SC-Trp-Leu, and single transformants restruck onto SC-Trp, SC-Trp-Leu-His or SC-Trp-Leu-Ade selective dropout media as indicated. Restruck yeast were photographed following incubation at 30oC for 96 hours.

94

3.4.2 – Co-Immunoprecipitation Fails to Confirm Yeast Two-Hybrid Identified HuC Protein-Protein Interactions:

Whilst the yeast two-hybrid system allows for identification of multiple putative protein interactors in a single screen, it does not recapitulate the environment in which mammalian proteins function in vivo. Based on this, in addition to the inconclusive results of the previous interactor retransformations (3.4.1), it was decided to immunoprecipitate neuronal Hu from murine neural tissue using nHu-specific PND patient antisera, along with any interacting proteins, and assay for the presence of each two-hybrid-identified putative interactor, in an effort to examine whether these interactions occur in a more physiological context. To this end, the relative abundance of neuronal Hu proteins present in murine P21 neural tissue lysates prepared in the presence of 1% Nonidet P-40 (NP-40) alone or in combination with either 0.5% Triton X-100 or 5% Deoxycholate (DOC) was examined, in order to identify which of these lysis conditions would provide the highest levels of Hu protein for subsequent immunoprecipitation. Comparison of equal quantities of each lysate, as determined by Bradford Assay, by western blotting with 3A2 anti-HuA monoclonal antibody, revealed the presence of significantly higher quantities of Hu protein in lysates prepared in the presence of 5% DOC (Figure 3.17C) than in those prepared in the presence of either 0.5% Triton (Figure 3.17B), or 1% NP-40 alone (Figure 3.17A).

Furthermore,

subsequent immunoprecipitation of neuronal Hu from each lysate using Ro patient antiseru, revealed approximately equal quantities of neuronal Hu remaining in each of the three lysates following immunodepletion (Figure 3.17A-C). Given the higher levels of Hu protein present prior to immunoprecipitation in lysates prepared with 5% DOC, this suggests that immunoprecipitates from lysates prepared under these conditions should contain the highest concentration of neuronal Hu proteins.

95

A.

B.

1% NP-40

kDa

C.

1% NP-40 0.5% Triton

3A2

1% NP-40 5% DOC

3A2

3A2

80 60 50 40

nHu

30 20 Ab Crosslinking:

-

D.

-

+

-

+

kDa

+

3A2

80 60 50 40

nHu HuA

30 20 Ro: NHS:

+

+ -

+ -

96

Figure 3.17 – Optimization of neuronal Hu Co-IP conditions: P21 mouse brain was lysed in IP lysis buffer containing either A. 1% NP-40, B. 1% NP-40 with 0.5% Triton X-100, or C. 1% NP-40 with 5% Deoxycholate (DOC), as described (2.2.8). Lysates were diluted to a final concentration of 0.5 µg/µL and 200 µL of each lysate incubated with 20 µL Protein A sepharose pre-incubated with either 5 µL Ro patient antiserum (Ro) or 5 µL non-immune human serum (NHS), in the presence or absence of DMP antibody crosslinking (2.2.8). Input and cleared lysates were analysed by 12% SDS-PAGE and western blotting with 3A2 anti-Hu monoclonal antibody (2.2.5). D. P23 mouse brain was lysed in IP lysis buffer containing 1% NP-40 with 5% Deoxycholate (DOC), as described (2.2.8). Lysate was diluted to a final concentration of 0.5 µg/µL and 200 µL of each lysate incubated with 20 µL Protein A sepharose pre-incubated with either 5 µL Ro patient antiserum (Ro) or 5 µL non-immune human serum (NHS). Neuronal Hu proteins were immunoprecipitated as described (2.2.8) and input, cleared lysates, and IP output analysed by 12% SDS-PAGE and western blotting with 3A2 anti-Hu monoclonal antibody (2.2.5).

97

To confirm the efficacy of neuronal Hu immunoprecipitation from lysates prepared in the presence of 5% DOC, murine P23 neural tissues lysed under these conditions were subjected to immunoprecipitation with either Ro patient antisera or normal human sera (NHS). Subsequent comparison of equal quantities of input lysate to each immunodepleted lysate (lanes 1-3, Figure 3.17D) revealed significant immunodepletion of neuronal Hu proteins with Ro antiserum, but not with normal human serum (cf. Lanes 2 & 3).

Moreover, analysis of a sample of post-

immunoprecipitation eluate demonstrated that neuronal Hu, but not the ubiquitous

HuA

was

specifically

precipitated

by

Ro

antiserum

(lane 4, Figure 3.17B), consistent with the previously reported specificity of Hu PND patient antisera (Okano & Darnell, 1997), thus confirming the specificity of this protocol toward the neuronal Hu proteins.

Having established conditions for the successful immunoprecipitation of nHu, this protocol was applied in an effort to confirm the interaction of each previously identified putative interactor with the neuronal Hu proteins. For this experiment, lysates were prepared from p17 murine neural tissue in the presence of 5% DOC, and the neuronal Hu proteins immunoprecipitated.

Examination of equal

quantities of input and cleared lysates by western blotting with 3A2 monoclonal antibody, which reacts with both ubiquitous HuA and all neuronal Hu proteins (Figure 3.18A, lanes 1-3), in addition to comparison with equal quantities of eluates from each immunoprecipitation (Figure 3.18A, lanes 4 & 5), confirmed the specific immunoprecipitation of neuronal Hu by Ro patient antiserum.

However,

further analysis of these input and eluate samples by western blot with antibodies against the 47 kDa known HuD-interactor Eukaryotic Initiation Factor 4a (47 kDa, Figure 3.18B) (Fukao, et al., 2009), revealed the presence of this protein in samples immunoprecipitated with either Ro or normal human serum, suggesting that it was co-immunoprecipitating in a non-specific manner. Moreover, western blotting with antibodies against Maged1 (83 kDa, Figure 3.18C) and CUG-BP2 98

(52 kDa, Figure 3.18D) revealed that neither of these proteins were successfully immunoprecipitated by Ro antiserum, despite the presence of appreciable quantities of both species in input lysates (lane 1, Fig. 3.18C & 3.19D). Both 3A2 and α-CUG-BP2 monoclonal antibodies showed cross-reactivity with Ro IgG light chain present in eluates, as indicated. Whilst these negative results imply that Maged1 and CUG-BP2 may not represent genuine nHu interactors, the inability of the protocol to specifically isolate eIF4a, a known HuD protein interactor, suggests that further optimisation is required before the interaction of each of these putative

interactors

with

Hu

can

be

conclusively

analysed

by

co-

immunoprecipitation.

99

A.

B.

kDa

kDa

40

nHu

80 60 50 40

HuA

30

IgG

eIF4a (47 kDa)

20 3A2

+ -

Ro: NHS:

+

+ -

-

30

+

20 α-eIF4a Ro: NHS:

C.

+ -

+

+ -

+

D. kDa

kDa α-Maged1

80 60

100 80 60 Ro: NHS:

Maged1 (83 kDa)

+ -

+

+ -

-

CUGBP2 (52 kDa)

50 40

30

IgG

+ 20 α-CUGBP2 Ro: NHS:

+ -

+

+ -

+

Figure 3.18 – Putative yeast two-hybrid positives do not specifically co-immunoprecipitate with neuronal Hu proteins: P17 mouse brain was lysed and subjected to immunoprecipitation with either ‘Ro’ patient antiserum or non-immune human serum (NHS), as described (2.2.8). Samples of input lysate, cleared lysate, and IP output were examined by 12% SDS-PAGE and western blotting with A. 3A2 anti-Hu monoclonal antibody, B. anti-eIF4a monoclonal antibody, C. anti-Maged1 monoclonal antibody, or D. anti-CUGBP2 monoclonal antibody.

100

3.4.3 – Putative Hu Protein Interactors Fail to Support Growth Under Stringent Selection Conditions: Based on the previously observed weak growth on media lacking adenine of yeast retransformed with each putative interactor, in addition to the inconclusive results regarding growth of these transformants in the absence of histidine (Figure 3.16) and

subsequent

inconclusive

co-immunoprecipitation

analyses

on

both

CUG Binding Protein 2 and Maged1 (Figure 3.18), it was decided to re-examine the growth of yeast co-transformed with either CUG-BP2 or Maged1, together with each of the two HuCvs2 bait proteins, on selective media lacking either histidine or adenine in an effort to conclusively determine whether these proteins interact within a two-hybrid context. Additionally, it was decided to examine the growth of each of these co-transformants on media supplemented with the anti-fungal compound Aureobasidin A, which allows for the selection of two-hybrid interactions in y2hgold yeast possessing the AUR1-C drug-resistance gene under control of the GAL promoter, and is significantly more stringent than typical nutritional selection (Nakazawa & Iwano, 2004).

Following retransformation, transformants were plated onto each selective medium (Figure 3.19) and the number of colonies on each plate enumerated (summarized in Table 3.3). Subsequent comparison of growth by positive control transformants possessing the p53 and SV4 large T antigen proteins described previously (3.4.2), to that of cells co-transformed with Bag4 and either HuCsv2 spacer-only or spacer-RRM3 bait proteins revealed that both bait/prey combinations supported growth in the absence of histidine, with colony counts equal to or greater than those observed for the positive control. By comparison, CUG Binding Protein 2 supported only weak growth in the presence of co-transformed HuCsv2 spacer-RRM, with few colonies present on media lacking histidine, whilst co-transformation with the spacer-only fusion yielded no growth 101

-Trp -Leu -His

+ve control

-ve control

-ve control

-ve control

p53/ SV40 large T

laminin/ SV40 large T

spacer-only

spacer-RRM3

spacer-only/ CUGBP2

spacer-only/ Bag4

spacer-RRM3/ CUGBP2

spacer-RRM3/ Bag4

-Trp -Leu -Ade

-Trp -Leu +AbA +Xgal

Table 3.3 – Colony counts of two-hybrid bait/prey re-transformants: Transformation

-Trp -Leu -His

-Trp -Leu -Ade

-Trp -Leu +AbA +Xgal

P53/ SV40 large T

76

117

14

Laminin/ SV40 large T

0

0

0

Spacer-only

0

0

0

Spacer-RRM3

0

0

0

Spacer-only/ CUGBP2

0

0

0

Spacer-only/ Bag4

19

0

0

Spacer-RRM3/ CUGBP2

214

0

0

Spacer-RRM3/ Bag4

99

0

0

Figure 3.19 – Putative yeast two-hybrid interactors fail to support growth under either Adenine nutritional or Aureobasidin A drug selection:

102

Vectors expressing control proteins (p53, laminin, SV40 large T antigen), HuCsv2 bait proteins (spacer-only, spacer-RRM3), or putative interactor proteins (CUG-BP2, Bag4, RPSA, Maged1) were transformed into competent AH109 yeast in the indicated combinations, as described (2.2.2). Transformations were plated directly onto the indicated selective media and plates examined following growth at 30oC for approximately 72 hours.

under these conditions.

Importantly, negative controls co-transformed with

laminin and SV40 large T antigen showed no growth under any of the selective conditions, in contrast to previous experiments (cf. Figure 3.16). Moreover, none of the experimental bait/prey combinations supported growth in either the absence adenine or the presence of Aureobasidin A, despite the growth of positive controls under these conditions.

Taken together with the previous results

demonstrating the inability of these two putative interactors to support significant growth in the absence of adenine when co-transformed with either HuCsv2 bait fusion protein (Figure 3.16), these results suggest that interactions occurring between these HuCsv2 bait fusion proteins and CUG Binding Protein 2 or Bag4 may be weak in nature, and thus incapable of inducing sufficient upregulation of reporter gene transcription for survival under more stringent selection conditions. Furthermore, the absence of clear immunoprecipitation of the known HDuinteractor eIF4a (Fukao, et al., 2009) suggests that further optimisation of immunoprecipiation conditions may be required before the interaction of neuronal Hu with each of these putative interactors can be conclusively investigated. Thus, further studies must be undertaken in order to confirm whether these and other putative interactors identified in this two-hybrid screen genuinely interact with HuCsv2, and to identify what role, if any, they play in neuronal Hu function.

103

Chapter 4 – Discussion:

104

4.1 – Expression and Purification of Recombinant mHuCsv1: A comprehensive understanding of the mechanisms underlying neuronal Hu function is dependent on the examination of two fundamental classes of molecular interactions: first, those that occur between individual Hu proteins and their RNA substrates, and second, the protein-protein interactions that mediate the generegulatory functions of the Hu family. Investigation of each of these types of interaction is reliant upon the isolation of recombinant Hu proteins in both high quantity and purity, which can be applied to a range of biochemical experiments that serve to elucidate such interactions. Consequently, we have undertaken the expression and purification of recombinant murine HuCsv1, and have applied this protein to an investigation of the RNA-protein interactions that mediate the function of this protein in vivo. The protocol that had previously been developed within our laboratory applied a glutathione-S-transferase based purification of mHuCsv1, which yielded quantities of recombinant mHuCsv1 sufficient for use in biochemical applications, but failed to isolate this Hu protein with adequate purity for reliable analyses (Simpson, 2006). Subsequent investigation revealed the bulk of this contamination to be the result of degradation of the GST protein during purification, leading us to conclude that an alternate approach was required for the isolation of recombinant mHuCsv1. In an effort to overcome this issue, we removed the GST-tag from the recombinant protein and pursued a histidine-tag based approach, utilising an eight-histidine tag present at the C-terminus of the recombinant protein. Additionally, the Nterminal GST tag was replaced with a streptavidin-based Strep-Tag, allowing for use of the resultant protein in both RNA-protein and protein-protein co-pull down assays (3.1.2).

105

Whilst replacement of the GST tag prevented formation of the protein degradation products observed in previous studies, purification of the new Strep-mHuCsv1His 8 resulted in the formation of a protein contaminant of approximately 75 kDa that had not previously been observed.

Surprisingly, western blot analysis of

individual elution fractions revealed this species to be reactive with a number of Hu-specific antibodies, leading us to hypothesize that it may be formed as a result of mHuCsv1 aggregation during purification (3.2.1). Subsequent purification of recombinant Hu in the absence of peptide-based protease inhibitors resulted in the isolation of mHuCsv1 in the absence of this contaminant, leading us to conclude that these compounds were promoting aggregation of recombinant mHuCsv1 during purification.

Moreover, subsequent purification of Strep-

mHuCsv1-His 8 , utilising a decreased elution buffer flow rate yielded pure protein at concentrations sufficient for in vitro RNA-binding analyses (3.2.2), and subsequent RNA nitrocellulose filter-binding assays against CLIP-identified RNA sequences from the Cold-Inducible RNA-Binding Protein and heterogeneous nuclear Ribonucleoprotein A/B mRNAs of D. rerio confirmed the ability of this protein to bind RNA selectively, and with affinities equivalent to those previously reported for in vitro studies with HuA (3.3.1) (Fialcowitz-White, et al., 2007).

106

4.2 – Investigation of RNA-Binding by Recombinant HuCsv1: Whilst previous studies have examined RNA binding by the ubiquitous HuA protein (Fialcowitz-White, et al., 2007), little published data exists regarding the RNA targets of the neuronal Hu proteins.

In order to form a complete

understanding of the role played by these proteins during neuronal development, it is essential that their RNA substrates be identified, and their regulation by the Hu proteins investigated. Whilst application of the CLIP method to 512-cell and 24hpf

D. rerio embryos and E18 murine neural tissue within our laboratory has facilitated the identification of a large number of specific mRNAs bound by neuronal Hu in each context, as well as the approximate sites of Hu binding to each of these targets, the precise relationship between CLIP tag RNA sequences and sites of direct Hu binding remains unclear.

In an effort to further explore these relationships, we have analysed the binding of recombinant Strep-mHuCsv1-His 8 to a number of CLIP-identified mRNAs by means of in vitro nitrocellulose filter-binding assays (3.3.2). Analysis of Hu binding to CLIP-identified RNA sequences in the 3’UTRs of the Cold-Inducible RNA-Binding Protein and heterogeneous Ribonucleoprotein A/B mRNAs of zebrafish confirmed the efficacy of this approach, with recombinant mHuCsv1 binding to a subset of selected RNA sequences with an affinity comparable to that reported in previous studies with HuA (Fialcowitz-White, et al., 2007). Surprisingly, these assays also revealed that two of the CLIP tags identified within the hnRNP A/B 3’UTR were not bound by the recombinant protein in vitro.

Subsequent examination of

Strep-mHuCsv1-His 8 binding to a number of CLIP tags and adjacent RNA sequences identified within the 3’UTRs of the zebrafish β-Catenin 2, eIF4G2a, and sfrs1l mRNAs revealed the existence of at least one HuCsv1-bound RNA sequence in each that had not been identified by CLIP (Figures 3.9-3.11). Furthermore, we have found that both eIF4G2a and sfrs1l possessed at least one CLIP tag sequence 107

within

their

3’UTRs

that

was

not

bound

by

HuC.

In

both

cases,

Strep-mHuCsv1-His 8 bound with high affinity to an RNA sequence directly adjacent to each unbound CLIP tag. Moreover, analysis of CLIP tags from the 3’UTRs of the murine Cofilin 1 and Hig1 mRNAs, identified in CLIP experiments performed on E18 murine neural tissue, revealed that Hu did not bind to either of these RNA sequences, but rather to adjacent uridine-rich RNA sequences present in each mRNA (Figures 3.12-3.13).

Together, these results strongly suggest that RNA sequences identified by CLIP do not necessarily represent sites of direct binding by HuCsv1. However, each of the described binding assays were performed only once, and thus a definitive analysis of Hu binding by these sequences requires that each experiment be repeated in order to confirm the statistical significance of these results. Additionally, whilst the RNAs used in these assays are each greater than 30 nt in length, previous studies have reported that a single Hu protein binds no more than 15 nt (Fialcowitz-White, et al., 2007; Wang & Tanaka Hall, 2001), and thus a thorough description of the relationships between individual CLIP tags and associated HuCsv1 binding sites will rely upon the application of binding site boundary-mapping experiments. Such assays involve the transcription of RNA sequences of interest in vitro followed by labelling of the resultant RNA molecules at either the 5’ or 3’ end with

32

P. The

resultant RNAs are then subjected to mild alkaline hydrolysis to produce a series of labelled RNAs varying in length from a few nucleotides to the full length of the sequence of interest, with single nucleotide differences between each consecutive hydrolysis product. The subset of labelled RNA sequence to which mHuCsv1 binds can then be isolated by in vitro RNA-binding. Bound RNAs may then be eluted and visualised by SDS-PAGE and autoradiography.

The length of the smallest

bound 5’-labeled RNA then corresponds to the distance from the 5’ end of the longest RNA to the extreme 3’ end of the Hu binding site, whilst the length of the smallest 3’-labeled RNA corresponds to the distance, in nucleotides, from the 3’ 108

end of the longest RNA to the extreme 5’ end of the Hu binding site. Application of such boundary-mapping assays would elucidate the minimum sequences within each mHuCsv1-bound RNA that are required to mediate high-affinity binding. Furthermore, whilst these analyses have only examined RNA binding by a single member of the neuronal Hu protein family, the CLIP method, by virtue of the reactivities of the PND patient antisera used, identifies putative RNA sequences which may be bound by any of the neuronal Hu proteins. Thus, analysis of RNA binding by the remaining Hu proteins may reveal that CLIP tags not bound by mHuCsv1 are bound by other members of the Hu family. Consequently, a more detailed analysis of RNA binding by both mHuCsv1 and the other neuronal Hu proteins is required before the relationship between individual CLIP tags and Hu binding sites can be fully elucidated. In addition to examining HuCsv1 binding to individual CLIP tag sequences, we have also applied the data from these assays in an effort to identify the specific sequence determinants underlying HuCsv1 RNA-binding activity.

Whilst

examination of the sequence composition of each assays RNA revealed a strong preference for uridine-rich RNA sequences, more striking was the high incidence of short interspersed uridine-rich motifs in each of the highest affinity interactors, illustrated by the greater abundance of di- and tri-uridine motifs in high-affinity RNA targets, when compared to those not bound by mHuCsv1 (3.3.3). These

results

are

consistent

with

the

previously

reported

putative

x-U/C-U-x-x-U/C-U-U/C consensus sequence identified in crystallisation studies on RRMs 1 and 2 of HuD (Wang & Tanaka Hall, 2001). However, identification of a definitive consensus HuCsv1-binding sequence will rely upon the identification and bioinformatic comparison of a significantly larger pool of HuC-bound RNA sequences, in addition to the identification of a 13-15nt minimal HuC binding sequence within each RNA, by application of in vitro boundary-mapping studies. 109

4.3 – Identification of Neuronal Hu Protein Interactors: Whilst previous studies have investigated the protein-protein interactions of the ubiquitous HuA protein by means of co-immunoprecipitation from non-neuronal cell lines (Brennan, et al., 2000), and more recent studies have begun to describe the protein-protein interactions behind neuronal Hu function (Fukao, et al., 2009), there is still relatively little published data regarding neuronal Hu protein-protein interactions. This is due in part to the lack of specific antibodies against each of these proteins, which precludes the identification of individual nHu protein cofactors by means of immunoprecipitation.

Thus, the identification of these

proteins must rely upon the application of non-immunological techniques that allow for the examination of specific neuronal Hu proteins and splice variants. Consequently, our laboratory has previously performed a yeast two-hybrid assay in an effort to identify protein interactors to the murine HuCsv2 protein (Jensen, 2009). This screen resulted in the successful identification of a number of putative interactors, including CUG-Binding Protein 2, Melanoma-associated Antigen D1, Ribosomal Protein SA, and Hsp70-associated Protein Bag4, all of which have been previously reported to play roles in processes directly related to neuronal specification and differentiation, including post-transcriptional gene regulation and apoptosis. In an effort to confirm whether these represent genuine Hu protein interactors, we have examined the ability of these proteins to individually support yeast growth under varying stringency nutritional selection when co-transformed with HuCsv2 bait proteins from the initial two-hybrid screen. However, whilst retransformation of each of these putative interactors yielded growth of the resultant co-transformants under mild selective conditions, the growth of negative control transformants under these conditions, in addition to a lack of growth under more stringent selection conditions made interpretation of these results problematic (3.4.1).

110

Thus, we endeavoured to examine several of these putative interactions through application of a co-immunoprecipitation approach, utilising Hu PND patient antisera. To this end, we have successfully developed a protocol for the efficient and specific immunoprecipitation of the neuronal Hu proteins from murine neural tissue (3.4.2). However, application of this protocol to the co-immunoprecipitation of neuronal Hu protein cofactors resulted in the non-specific isolation of the known HuD-interactor, eIF4a, by both PND patient antiserum and non-immune control serum. Additionally, neither Maged1 nor CUG-Binding Protein 2 could be isolated under these conditions. Consequently, further optimisation is required before this protocol can be successfully applied to the confirmation and identification of neuronal Hu protein interactors. Given the observed non-specific precipitation of eIF4a, we hypothesize that an increase in the stringency of washes used to remove non-specifically precipitating proteins is required, in order to prevent non-specific precipitation of eIF4a, as well as other, non-interacting, proteins. Currently, these washes are performed using buffer containing the nonionic detergent Nonidet P-40. Introduction of a small amount of ionic detergent (eg. 0.5% Deoxycholate), may help to reduce non-specific precipitation. Following this,

western

blot

analysis

can

be

applied

to

confirm

the

specific

immunoprecipitation of eIF4a by PND patient antiserum. Once this known protein cofactor can be specifically and reliably isolated, this protocol can then be applied to the confirmation and identification of additional Hu protein interactors.

Following the unsuccessful application of co-immunoprecipitation to the confirmation of these putative two-hybrid interactors, we have further attempted to confirm the ability of these proteins to support yeast growth under varying stringency nutritional selection, as well as under more stringent selection by the anti-fungal compound Aureobasidin A, when co-transformed with HuCsv2 bait fusion proteins from the initial two-hybrid screen (3.4.3). Whilst this successfully demonstrated the ability of the Bag4 and CUG-Binding Protein 2 prey fusion 111

proteins to support growth under mild selection in the absence of histidine, we observed that neither was capable of supporting growth under more stringent selection conditions, implying that interactions between these proteins and HuCsv2 may be weak in nature. Based on the largely inconclusive results of these and previous retransformations, in addition to those of co-immunoprecipitation studies on Maged1 and CUG-Binding Protein 2, further investigation is clearly required before each of the four putative interactors identified by yeast two-hybrid can be confirmed as HuCsv2 protein interactors. The first step in this confirmation is the retransformation of each of these interactor fusion proteins, together with each HuCsv2 bait fusion protein, followed by conclusive demonstration of their ability to support growth under nutritional selection. Following this confirmation, application of an optimised co-immunoprecipitation protocol will allow for definitive confirmation of these interactors, whilst subsequent in vitro Strep-Tag co-pull down studies utilising recombinant HuC will allow for a thorough investigation of the protein subregions with which each of these protein interact.

112

4.4 – Conclusions and Future Directions: The development of a protocol for the efficient expression and purification of active, recombinant HuCsv1 provides an invaluable tool in the ongoing biochemical study of the molecular interactions underlying neuronal Hu function. Application of this protein, both to the identification of specific Hu RNA targets and also to the isolation of the protein cofactors that mediate this protein family’s gene regulatory functions, will allow us to significantly expand our understanding of the role played by Hu during neuronal development. Moreover, application of this protocol to the expression and purification of the remaining neuronal Hu proteins and splice variants will allow us to more fully explore the distinct functions of each of these proteins. Application to in vitro binding assays of recombinant mHuCsv1 isolated by this procedure has allowed us to begin elucidating the sequence determinants underlying RNA-binding by the neuronal Hu proteins, and to explore the relationship between specific sites of HuC RNA-binding and putative binding sites previously identified in CLIP experiments from multiple species and developmental stages. Further binding analyses utilising this recombinant protein will allow us to identify the specific RNA sequences which mediate HuCsv1 binding and their relationship to those identified by CLIP, whilst the application of similar approaches to other neuronal Hu proteins will also allow us to define the RNA sequences that mediate binding by each individual family member and splice variant, and to subsequently begin applying bioinformatic techniques to the identification of potential targets of Hu-mediated post-transcriptional gene regulation.

113

We have also begun investigating the protein-protein interactions that allow the neuronal Hu proteins to mediate their effects on bound mRNAs. Thus, we have developed a procedure for the efficient and specific immunoprecipitation of the neuronal Hu proteins from murine brain tissue.

Further optimisation of this

protocol will allow us to isolate and identify individual proteins that interact with neuronal Hu, and thus to describe the molecular mechanisms underlying the gene regulatory functions of the neuronal Hu proteins. Thus, further investigation, both of the RNA-protein and protein-protein interactions undertaken by the neuronal Hu proteins, in addition to our current knowledge regarding their temporal and spatial expression throughout neuronal development, will allow us to begin elucidating the molecular mechanisms of neuronal Hu function. It will also allow us to begin defining the roles of individual Hu proteins and splice variants during neuronal development. Together, these results will yield significant advances in our understanding of neuronal specification and differentiation during development, and provide insight into the molecular mechanisms which underlie these processes.

114

Chapter 5 – Bibliography:

115

Abdelmohsen, K., Hutchison, E. R., Lee, E. K., Kuwano, Y., Kim, M. M., Masuda, K., et al. (2010). miR-375 inhibits differentiation of neurites by lowering HuD levels. Mol Cell Biol, 30(17), 4197-4210. Akamatsu, W., Fujihara, H., Mitsuhashi, T., Yano, M., Shibata, S., Hayakawa, Y., et al. (2005). The RNA-binding protein HuD regulates neuronal cell identity and maturation. Proc Natl Acad Sci U S A, 102(12), 4625-4630. Akamatsu, W., Okano, H. J., Osumi, N., Inoue, T., Nakamura, S., Sakakibara, S., et al. (1999). Mammalian ELAV-like neuronal RNA-binding proteins HuB and HuC promote neuronal development in both the central and the peripheral nervous systems. Proc Natl Acad Sci U S A, 96(17), 9885-9890. Antic, D., Lu, N., & Keene, J. D. (1999). ELAV tumor antigen, Hel-N1, increases translation of neurofilament M mRNA and induces formation of neurites in human teratocarcinoma cells. Genes Dev, 13(4), 449-461. Aranda-Abreu, G. E., Behar, L., Chung, S., Furneaux, H., & Ginzburg, I. (1999). Embryonic lethal abnormal vision-like RNA-binding proteins regulate neurite outgrowth and tau expression in PC12 cells. J Neurosci, 19(16), 6907-6917. Bertrand, M. J., Kenchappa, R. S., Andrieu, D., Leclercq-Smekens, M., Nguyen, H. N., Carter, B. D., et al. (2008). NRAGE, a p75NTR adaptor protein, is required for developmental apoptosis in vivo. Cell Death Differ, 15(12), 1921-1929. Bhattacharyya, S. N., Habermacher, R., Martine, U., Closs, E. I., & Filipowicz, W. (2006). Relief of microRNA-mediated translational repression in human cells subjected to stress. Cell, 125(6), 1111-1124. Bolognani, F., Contente-Cuomo, T., & Perrone-Bizzozero, N. I. (2010). Novel recognition motifs and biological functions of the RNA-binding protein HuD revealed by genome-wide identification of its targets. Nucleic Acids Res, 38(1), 117-130. Bolognani, F., & Perrone-Bizzozero, N. I. (2008). RNA-protein interactions and control of mRNA stability in neurons. J Neurosci Res, 86(3), 481-489. Brennan, C. M., Gallouzi, I. E., & Steitz, J. A. (2000). Protein ligands to HuR modulate its interaction with target mRNAs in vivo. J Cell Biol, 151(1), 1-14. Campbell, D. S., & Holt, C. E. (2001). Chemotropic responses of retinal growth cones mediated by rapid local protein synthesis and degradation. Neuron, 32(6), 1013-1026. 116

Cho, S. J., Zhang, J., & Chen, X. (2010). RNPC1 modulates the RNA-binding activity of, and cooperates with, HuR to regulate p21 mRNA stability. Nucleic Acids Res, 38(7), 2256-2267. Darnell, R. B., & Posner, J. B. (2003). Paraneoplastic syndromes involving the nervous system. N Engl J Med, 349(16), 1543-1554. Dembowski, J. A., & Grabowski, P. J. (2009). The CUGBP2 splicing factor regulates an ensemble of branchpoints from perimeter binding sites with implications for autoregulation. PLoS Genet, 5(8), e1000595. Devaux, A., Colegrove-Otero, L. J., & Standart, N. (2006). Xenopus ElrB, but not ElrA, binds RNA as an oligomer: possible role of the linker. FEBS Lett, 580(20), 4947-4952. Dredge, B. K. (2005). Unpublished Work. University of Adelaide. Dredge, B. K. (2009). Unpublished Work. University of Adelaide. Dredge, B. K., & Darnell, R. B. (2003). Nova regulates GABA(A) receptor gamma2 alternative splicing via a distal downstream UCAU-rich intronic splicing enhancer. Mol Cell Biol, 23(13), 4687-4700. Fan, X. C., & Steitz, J. A. (1998). HNS, a nuclear-cytoplasmic shuttling sequence in HuR. Proc Natl Acad Sci U S A, 95(26), 15293-15298. Fialcowitz-White, E. J., Brewer, B. Y., Ballin, J. D., Willis, C. D., Toth, E. A., & Wilson, G. M. (2007). Specific protein domains mediate cooperative assembly of HuR oligomers on AU-rich mRNA-destabilizing sequences. J Biol Chem, 282(29), 20948-20959. Fukao, A., Sasano, Y., Imataka, H., Inoue, K., Sakamoto, H., Sonenberg, N., et al. (2009). The ELAV protein HuD stimulates cap-dependent translation in a Poly(A)- and eIF4A-dependent manner. Mol Cell, 36(6), 1007-1017. Gebauer, F., & Hentze, M. W. (2004). Molecular mechanisms of translational control. Nat Rev Mol Cell Biol, 5(10), 827-835. Glisovic, T., Bachorik, J. L., Yong, J., & Dreyfuss, G. (2008). RNA-binding proteins and post-transcriptional gene regulation. FEBS Lett, 582(14), 1977-1986. Guttinger, S., Muhlhausser, P., Koller-Eichhorn, R., Brennecke, J., & Kutay, U. (2004). Transportin2 functions as importin and mediates nuclear import of HuR. Proc Natl Acad Sci U S A, 101(9), 2918-2923. 117

Hing, F. (2009). Unpublished Work. University of Adelaide. Hoefsloot, L. H., Hoogeveen-Westerveld, M., Reuser, A. J., & Oostra, B. A. (1990). Characterization of the human lysosomal alpha-glucosidase gene. Biochem J, 272(2), 493-497. Holt, C. E., & Bullock, S. L. (2009). Subcellular mRNA localization in animal cells and why it matters. Science, 326(5957), 1212-1216. Huang, L., Hwang, J., Sharma, S. D., Hargittai, M. R., Chen, Y., Arnold, J. J., et al. (2005). Hepatitis C virus nonstructural protein 5A (NS5A) is an RNA-binding protein. J Biol Chem, 280(43), 36417-36428. Iwabuchi, K., Li, B., Bartel, P., & Fields, S. (1993). Use of the two-hybrid system to identify the domain of p53 involved in oligomerization. Oncogene, 8(6), 1693-1696. Iyer, V., & Struhl, K. (1995). Mechanism of differential utilization of the his3 TR and TC TATA elements. Mol Cell Biol, 15(12), 7059-7066. Jensen, K. B. (2004). Unpublished Work. University of Adelaide. Jensen, K. B. (2009). Unpublished Work. University of Adelaide. John, G. B., Shang, Y., Li, L., Renken, C., Mannella, C. A., Selker, J. M., et al. (2005). The mitochondrial inner membrane protein mitofilin controls cristae morphology. Mol Biol Cell, 16(3), 1543-1554. Kanai, Y., Dohmae, N., & Hirokawa, N. (2004). Kinesin transports RNA: isolation and characterization of an RNA-transporting granule. Neuron, 43(4), 513-525. Kasashima, K., Sakashita, E., Saito, K., & Sakamoto, H. (2002). Complex formation of the neuron-specific ELAV-like Hu RNA-binding proteins. Nucleic Acids Res, 30(20), 4519-4526. Katsanou, V., Papadaki, O., Milatos, S., Blackshear, P. J., Anderson, P., Kollias, G., et al. (2005). HuR as a negative posttranscriptional modulator in inflammation. Mol Cell, 19(6), 777-789. Kendall, S. E., Battelli, C., Irwin, S., Mitchell, J. G., Glackin, C. A., & Verdi, J. M. (2005). NRAGE mediates p38 activation and neural progenitor apoptosis via the bone morphogenetic protein signaling cascade. Mol Cell Biol, 25(17), 77117724. 118

Khabar, K. S. (2010). Post-transcriptional control during chronic inflammation and cancer: a focus on AU-rich elements. Cell Mol Life Sci, 67(17), 2937-2955. Kim, C. H., Ueshima, E., Muraoka, O., Tanaka, H., Yeo, S. Y., Huh, T. L., et al. (1996). Zebrafish elav/HuC homologue as a very early neuronal marker. Neurosci Lett, 216(2), 109-112. King, P. H. (1994). Hel-N2: a novel isoform of Hel-N1 which is conserved in rat neural tissue and produced in early embryogenesis. Gene, 151(1-2), 261265. Leung, K. M., van Horck, F. P., Lin, A. C., Allison, R., Standart, N., & Holt, C. E. (2006). Asymmetrical beta-actin mRNA translation in growth cones mediates attractive turning to netrin-1. Nat Neurosci, 9(10), 1247-1256. Levine, T. D., Gao, F., King, P. H., Andrews, L. G., & Keene, J. D. (1993). Hel-N1: an autoimmune RNA-binding protein with specificity for 3' uridylate-rich untranslated regions of growth factor mRNAs. Mol Cell Biol, 13(6), 34943504. Mendez, R., & Richter, J. D. (2001). Translational control by CPEB: a means to the end. Nat Rev Mol Cell Biol, 2(7), 521-529. Miller, S., Yasuda, M., Coats, J. K., Jones, Y., Martone, M. E., & Mayford, M. (2002). Disruption of dendritic translation of CaMKIIalpha impairs stabilization of synaptic plasticity and memory consolidation. Neuron, 36(3), 507-519. Mobarak, C. D., Anderson, K. D., Morin, M., Beckel-Mitchener, A., Rogers, S. L., Furneaux, H., et al. (2000). The RNA-binding protein HuD is required for GAP-43 mRNA stability, GAP-43 gene expression, and PKC-dependent neurite outgrowth in PC12 cells. Mol Biol Cell, 11(9), 3191-3203. Nakamura, A., Sato, K., & Hanyu-Nakamura, K. (2004). Drosophila cup is an eIF4E binding protein that associates with Bruno and regulates oskar mRNA translation in oogenesis. Dev Cell, 6(1), 69-78. Nakazawa, N., & Iwano, K. (2004). Efficient selection of hybrids by protoplast fusion using drug resistance markers and reporter genes in Saccharomyces cerevisiae. J Biosci Bioeng, 98(5), 353-358. Nelson, J., McFerran, N. V., Pivato, G., Chambers, E., Doherty, C., Steele, D., et al. (2008). The 67 kDa laminin receptor: structure, function and role in disease. Biosci Rep, 28(1), 33-48. 119

Okano, H. J., & Darnell, R. B. (1997). A hierarchy of Hu RNA binding proteins in developing and adult neurons. J Neurosci, 17(9), 3024-3037. Park, H. C., Hong, S. K., Kim, H. S., Kim, S. H., Yoon, E. J., Kim, C. H., et al. (2000). Structural comparison of zebrafish Elav/Hu and their differential expressions during neurogenesis. Neurosci Lett, 279(2), 81-84. Park, H. C., Kim, C. H., Bae, Y. K., Yeo, S. Y., Kim, S. H., Hong, S. K., et al. (2000). Analysis of upstream elements in the HuC promoter leads to the establishment of transgenic zebrafish with fluorescent neurons. Dev Biol, 227(2), 279-293. Pascale, A., Amadio, M., & Quattrone, A. (2008). Defining a neuron: neuronal ELAV proteins. Cell Mol Life Sci, 65(1), 128-140. Peng, S. S., Chen, C. Y., Xu, N., & Shyu, A. B. (1998). RNA stabilization by the AU-rich element binding protein, HuR, an ELAV protein. EMBO J, 17(12), 3461-3470. Robinow, S., & White, K. (1988). The locus elav of Drosophila melanogaster is expressed in neurons at all developmental stages. Dev Biol, 126(2), 294-303. Saito, K., Fujiwara, T., Katahira, J., Inoue, K., & Sakamoto, H. (2004). TAP/NXF1, the primary mRNA export receptor, specifically interacts with a neuronal RNAbinding protein HuD. Biochem Biophys Res Commun, 321(2), 291-297. Sakai, K., Gofuku, M., Kitagawa, Y., Ogasawara, T., Hirose, G., Yamazaki, M., et al. (1994). A hippocampal protein associated with paraneoplastic neurologic syndrome and small cell lung carcinoma. Biochem Biophys Res Commun, 199(3), 1200-1208. Sibbons, J. (2007). Unpublished Work. University of Adelaide. Simpson, B. (2006). Biochemical and Proteomic Studies on the Neuronal Hu Proteins of Mus Musculus and Danio Rerio. B. Sc. (Hons), University of Adelaide, Adelaide. Steward, O., & Schuman, E. M. (2003). Compartmentalized synthesis and degradation of proteins in neurons. Neuron, 40(2), 347-359. Sureban, S. M., Murmu, N., Rodriguez, P., May, R., Maheshwari, R., Dieckgraefe, B. K., et al. (2007). Functional antagonism between RNA binding proteins HuR and CUGBP2 determines the fate of COX-2 mRNA translation. Gastroenterology, 132(3), 1055-1065. 120

Szabo, A., Dalmau, J., Manley, G., Rosenfeld, M., Wong, E., Henson, J., et al. (1991). HuD, a paraneoplastic encephalomyelitis antigen, contains RNA-binding domains and is homologous to Elav and Sex-lethal. Cell, 67(2), 325-333. Takada, H., Chen, N. J., Mirtsos, C., Suzuki, S., Suzuki, N., Wakeham, A., et al. (2003). Role of SODD in regulation of tumor necrosis factor responses. Mol Cell Biol, 23(11), 4026-4033. Tiruchinapalli, D. M., Ehlers, M. D., & Keene, J. D. (2008). Activity-dependent expression of RNA binding protein HuD and its association with mRNAs in neurons. RNA Biol, 5(3), 157-168. Ule, J., Jensen, K. B., Ruggiu, M., Mele, A., Ule, A., & Darnell, R. B. (2003). CLIP identifies Nova-regulated RNA networks in the brain. Science, 302(5648), 1212-1215. Wang, X., & Tanaka Hall, T. M. (2001). Structural basis for recognition of AU-rich element RNA by the HuD protein. Nat Struct Biol, 8(2), 141-145. Webb, J. (2006). Unpublished Work. University of Adelaide. Whelan, F. (2008). Unpublished Observation. University of Adelaide. Wiszniak, S. (2010). Unpublished Observations. University of Adelaide. Yao, J., Sasaki, Y., Wen, Z., Bassell, G. J., & Zheng, J. Q. (2006). An essential role for beta-actin mRNA localization and translation in Ca2+-dependent growth cone guidance. Nat Neurosci, 9(10), 1265-1273. Zwicker, N., Adelhelm, K., Thiericke, R., Grabley, S., & Hanel, F. (1999). Strep-tag II for one-step affinity purification of active bHLHzip domain of human c-Myc. Biotechniques, 27(2), 368-375.

121