University of Southern California Dissertations and Theses

Analysis of ELAVL4 RNA interaction using a switchSENSE biosensor

(USC Thesis Other)

Analysis of ELAVL4 RNA interaction using a switchSENSE biosensor

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content 1

ANALYSIS OF ELAVL4 RNA INTERACTION USING A
SWITCHSENSE BIOSENSOR

By
Anusha Muralidhar

A Thesis Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
in Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(BIOCHEMISTRY AND MOLECULAR MEDICINE)
December 2018

2

DEDICATION

To my parents

3

ACKNOWLEDGEMENTS:
Words don’t suffice to express my appreciation and gratitude for the innumerable people who
have played such an important role throughout my master’s years by their constant motivation
and support.
First and foremost, I wish to express my gratitude to the principal investigator and my
backbone in my thesis, Dr. Ite.A.Offringa for not just her constant support and dedication
but also her generosity and ardent efforts in bringing the instrument on site.
I would like to thank Chunli Yan, our lab manager, who has always been by my side for these
two years and helped me complete this task. Also, I would like to thank the members of the
Offringa lab, Evelyn Tran, Theresa Ryan Stueve, Daniel Mullen, Laura St. Pierre, Madhura
Lotlikar, Arthur Sefiani, Tuo Shi and Jonathan Castillo of Dr. Crystal Marconett’s lab for their
valuable inputs and suggestions for the project.
I would also like to thank the scientists at Dynamic Biosensors. Inc. Special thanks to Thomas
Weber, Thomas Welte, Ralf Strasser, Joanna Deek, Kenneth Dickerson, Ulrich Rant for
training me, constantly guiding me and clearing my queries during my experiments. Last, but
not the least, immense amount of gratitude goes towards my committee members, Dr. Ite A.
Offringa, Dr. Ansgar Siemer and Dr. Ian S. Haworth for sparing their time from their very
busy schedule to guide and support my thesis.

4

ABSTRACT

Characterization of RNA/protein interactions is key to identifying the cellular mechanisms
involved in expression of genes. Traditional methods like gel shifts and isothermal titration
calorimetry (ITC) provide quantitative binding data of intermolecular interactions at
equilibrium conditions and surface plasmon resonance (SPR) yields kinetic data, but none of
these methods provide information on molecular shape changes during binding. The recently
developed switchSENSE technology, based on electro-switchable DNA nanolevers, provides
not only high sensitivity kinetics but also information on size, shape and conformational
changes occurring during an interaction, yielding a new depth of understanding of
biomolecular interactions in real time. Here we use switchSENSE technology to study the
interaction between RNA-binding protein HuD (ELAVL4) and RNA. Neuronal ELAVL
proteins are involved in neuron-specific RNA processing and neural development. HuD
consists of two N- terminal RNA-binding domains, a hinge region and a C-terminal RNA-
binding domain. The interaction between HuD and prototype target RNAs of the sequence
UU(AUUU)nAUU has been studied using equilibrium methods and SPR. The data suggested
that HuD might undergo a rearrangement upon RNA binding. We hypothesize that the hinge
region between RNA-binding domains 2 and 3 plays a key role in this conformational change.
If true, this might provide a mechanism to regulate the protein/RNA interaction. The
potential of switchSENSE technology to provide structural as well as kinetic data under
equilibrium as well as dynamic conditions makes this an ideal tool for our investigation. A
5

series of ELAVL4 mutants were generated to investigate the roles of different domains in the
interaction with AU-rich RNA and to test whether and how the tertiary structure of the
protein changes while binding to the RNA. In this study we focused on mainly studying how
the switchSENSE biosensor works by studying the interaction between WT ELAVL4 protein
and 15AU3 RNA. Due to time limitations, the different ELAVL4 domain mutants were not
analyzed. The experiments conducted on the switchSENSE will provide further mechanistic
insight into the dynamic nature of RNA/protein interactions, and how these might be
regulated.

6

TABLE OF CONTENTS

ACKNOWLEDGEMENTS .............................................................................. 3
ABSTRACT ................................................................................................. 4
LIST OF FIGURES ........................................................................................ 9
LIST OF TABLES ......................................................................................... 12
CHAPTER 1: INTRODUCTION ..................................................................... 13
1.1 RNA-binding proteins ............................................................................ 13
1.2 Implication of RNA-binding proteins in human diseases ................................... 17
1.3 RNA binding proteins of the ELAVL family .................................................. 21
1.4 ELAVL4 interacts with AU-rich RNA ......................................................... 23
1.5 Role of the hinge .................................................................................. 28
CHAPTER 2: MATERIALS AND METHODS ..................................................... 30
2.1 Preparation of constructs ........................................................................ 30
2.1.1 Preparation of WT ELAVL4 construct .................................................. 31
2.1.2 Preparation of H74 construct .............................................................. 31
2.1.3. Preparation of K318 construct ........................................................... 32
2.1.4 Preparation of K220 construct ............................................................ 32
2.1.5 Preparation of K208 construct ............................................................ 33
2.1.6 Preparation of J300 construct ............................................................. 33
2.2 Production of recombinant protein ELAVL4: ............................................... 34
2.3 RNA transcripts ................................................................................... 36
2.4 Brief description of the methodology to study the RNA protein interaction using the
switchSENSE biosensor. .............................................................................. 37
2.4.1 General principle ............................................................................ 37
2.4.2 Instrumentation .............................................................................. 39
2.4.3 The sensor chip ............................................................................... 40
2.4.4 Nomenclature of biochip ................................................................... 41
2.4.5 Software ....................................................................................... 42
7

2.4.6 Modes of measurement ..................................................................... 44
2.5 Experimental design: ............................................................................. 46
2.5.1 Setup of the experiment using switchBUILD software ................................ 47
2.5.2 switchCONTROL software ............................................................... 50
2.6 Control experiment with the WT ELAVL4 and 15AU3MUT ............................ 53
CHAPTER 3: RESULTS ................................................................................ 54
3.1 Analysis of WT ELAVL4 with 15AU3 RNA, Experiment 1 .............................. 54
3.1.1 Hybridization data analysis ................................................................. 55
3.1.2 Kinetic data analysis ......................................................................... 56
3.1.3 Sizing data analysis ........................................................................... 59
3.2 Analysis of WT ELAVL4 with 15AU3 RNA, Experiment 2 .............................. 60
3.2.1 Kinetic data analysis ......................................................................... 61
3.2.2. Sizing data analysis .......................................................................... 61
3.2.3 Titration curve analysis ..................................................................... 62
3.3: Analysis of WT ELAVL4 with 15AU3 RNA, Experiment 3. ............................ 63
3.3.1 Kinetic data analysis: ........................................................................ 64
3.3.2 Sizing data analysis: .......................................................................... 65
3.3.3 Titration data analysis ....................................................................... 65
3.4 Analysis of WT ELAVL4 with 15AU3MUT RNA, Experiment 4........................ 66
3.4.1 Kinetic data analysis ......................................................................... 66
3.5 Analysis of WT ELAVL4 with 15AU3MUT RNA in the inverse orientation using the
NTA assay, experiment 5. ............................................................................ 68
3.5.1 General workflow of the NTA assay ..................................................... 68
3.5.2 switchBUILD script for the NTA assay .................................................. 69
3.5.3: switchANALYSIS data ..................................................................... 70
CHAPTER 4: DISCUSSION AND FUTURE DIRECTIONS .................................... 72
4.1 Kinetics of the DNA RNA mix as the ligand and WT ELAVL4 as the analyte in
dynamic response mode. ............................................................................. 72
8

4.2 Kinetics of the DNA RNA mix as the ligand and WT ELAVL4 as the analyte in
Fluorescence proximity sensing mode. ............................................................ 75
4.3 Dissociation constants in kinetic and equilibrium modes .................................. 76
4.4 Affinity of ELAVL4 for 15AU3 RNA. ......................................................... 77
4.4 Inverse assay orientation kinetics with the protein bound on the surface and the RNA
flowing in. ............................................................................................... 77
4.5 Advantages and limitations of switchSENSE technology compared to other methods.
............................................................................................................ 79
4.6 Future directions. ................................................................................. 80
REFERENCES ............................................................................................ 81
APPENDIX ................................................................................................ 85

9

LIST OF FIGURES

Figure 1: RNA-binding proteins and their role in RNA metabolism. .......................... 13
Figure 2: Conserved structure of RNA recognition motif (RRM). .............................. 16
Figure 3: Network of RNA-binding proteins in human diseases.. ............................... 18
Figure 4: RBP-associated mutations may disrupt RNA processing by several mechanisms. . 19
Figure 5: Diverse mechanisms of ELAVL protein functions. ..................................... 21
Figure 6: RNA seq data of ELAVL4 in the body. ................................................... 23
Figure 7: Model depicting the various functions of ELAVL4 in the cell… ..................... 23
Figure 8: Analysis of ELAVL4 binding to different AU rich targets. ............................ 24
Figure 9: Quantitated gel shift data and comparision of equilibrium binding affinities for
different mutants ......................................................................................... 25
Figure 10: Kinetic analysis of mutant ELAVL4 proteins. .......................................... 26
Figure 11: Schematic of the effects of PTM on RNA binding proteins… ...................... 29
Figure 12: Construction of ELAVL4 and mutants .................................................. 30
Figure 13: Schematic representation of the WT ELAVL (amino acids 1-373). ................ 31
Figure 14: Schematic representation of the mutant H74 (amino acids 1-224). ................ 31
Figure 15: Schematic representation of the mutant H74 (amino acids 1-224). ................ 32
Figure 16: Schematic representation of the mutant K220(amino acids 1-220+257-373) ... 32
Figure 17: Schematic representation of the mutant K208 (amino acids 1-208+278-373) ... 33
Figure 18: Schematic representation of the mutant J300 (1-373AA). ........................... 33
Figure 19: SDS PAGE gel of the WT ELAVL4 eluted in 250mM imidazole of two different
preps. ....................................................................................................... 34
Figure 20: Quantification of WT ELAVL4 protein using the standard graph. ................. 35
Figure 21: Schematic of the general principle of the switchSENSE technology ................ 37
Figure 22: Schematic of DNA switching process with alternating electrical voltages. ....... 38
Figure 23: switchSENSE Analyzer DRX 2400 series. .............................................. 39
Figure 24: Schematic of the DNA nanolevers with both the red and green fluorescent dyes.
............................................................................................................... 39
Figure 25: switchSENSE analyzer dual color biochip ............................................... 40
Figure 26: Process of regeneration of the flow channel… ........................................ 41
Figure 28: Schematic of the static measurement mode with constant negative
voltage……………………………………………………………………………...44
Figure 29: Schematic of the dynamic mode of measurement. .................................... 45
Figure 30: Negative potential applied to the nanolevers.... ....................................... 45
Figure 31: Dynamic response signals allow measurement of k
on
and k
off
. ....................... 46
10

Figure 32: Schematic of the nanolever with the hybridized DNA RNA complimentary
strand. ...................................................................................................... 47
Figure 33: Selection of assay type, assay elements, channels and buffers ....................... 48
Figure 34: Selection of the measurement mode, interacting partners, kinetic parameters,
and measurement spots. ................................................................................. 48
Figure 35: Predicted plot of percentage of RNA bound vs time ................................. 49
Figure 36: Volume of buffers, ligand and analyte needed for the experiment along with
position of the vials to be placed in the autosampler. .............................................. 49
Figure 37: Passivation of the chip to prevent any nonspecific binding........................... 51
Figure 38: Increase in fluorescence intensity vs. time during hybridization of the
complimentary nanolevers. ............................................................................. 52
Figure 39: Schematic representation of nanolever with 15AU3MUT ligand and ELAVL4 as
the analyte. ................................................................................................ 53
Figure 40: Importing the measurement data file to the switchANALYSIS software. ......... 55
Figure 41: switchANALYSIS software and evaluation of various assay elements by choosing
the appropriate analysis tool. ........................................................................... 55
Figure 42: Hybridization of the nanolevers. ......................................................... 56
Figure 43: Kinetics assay tool was used to analyze the kinetics of 15AU3 with the WT
ELAVL4. .................................................................................................. 57
Figure 45: Dynamic response kinetic data ........................................................... 58
Figure 46: Fluorescence proximity data .............................................................. 59
Figure 47: Selection of the sizing tool to size analysis of 15AU3 and complex with WT
ELAVl4 ..................................................................................................... 59
Figure 48: Sizing data for the complex formed between 15AU3 and WT ELAVL4 .......... 60
Figure 49: Association data ............................................................................. 61
Figure 50: Sizing data for the complex formed between 15AU3 and WT ELAVL4 ......... 62
Figure 51: Selection of the titration tool to determine the equilibrium dissociation constant
............................................................................................................... 62
Figure 52: Titration curve analysis tool ............................................................... 63
Figure 53: Titration curve for experiment 2 ......................................................... 63
Figure 54: Association data ............................................................................. 64
Figure 55: Sizing data ................................................................................... 65
Figure 55: Titration curve .............................................................................. 66
Figure 57: Dynamic response data for interaction of WT ELAVL4 and 15AU3MUT........ 67
Figure 58: Fluoresence data for interaction of WT ELAVL4 and 15AU3MUT ............... 67
Figure 59: Schematic representation of the inverse orientation with WT ELAVL4 as analyte
and 15AU3 RNA as ligand. ............................................................................ 68
11

Figure 60: Workflow overview with Tris -NTA nanolevers ...................................... 68
Figure 62: Hybridization curve of 300nM ELAVL4 binding to the cNLB-NTA nanolever .. 70
Figure 63: Association and dissociation curves in the dynamic mode ........................... 71
Figure 64: Association and dissociation curves in the Fluorescence proximity sensing mode
............................................................................................................... 71
Figure 65: Effect of ligand density in the switch SENSE biochips ................................ 73
Figure 66: Effect of ligand density and mass transport limitations using a biotin/anti-biotin
mAb model system ...................................................................................... 74

12

LIST OF TABLES

Table 1: Selected RBPs and their association with neurological diseases. ...................... 20
Table 2: Kinetic values for complexes of ELAVL4/HuD with AU-3 RNA. ................... 26
Table 3: BSA standards dissolved in PE40 to obtain the standard graph. ...................... 35

13

CHAPTER 1: INTRODUCTION

1.1 RNA-binding proteins

RNA-binding proteins (often abbreviated as RBPs) are proteins that bind to the RNA (either
double- or single-stranded) in cells and participate in the formation
of ribonucleoprotein complexes.

Figure 1: RNA-binding proteins and their role in RNA metabolism.
(Yeo et al., 2015)

Functionally, RBPs have crucial roles in numerous cellular processes, particularly in RNA
processing (Fig. 1). They regulate every aspect of RNA metabolism and function, including
RNA biogenesis, maturation (including processes like (alternative) splicing
and polyadenylation), transport, cellular localization, translation and stability. Among others,
14

eukaryotic RNA-binding activity and protein–protein interactions enabled eukaryotic cells to
make strong use of alternative splicing, in which RNA exons can be used in numerous
arrangements through the formation of a unique RNP (ribonucleoprotein) for each differential
RNA (Hogan et al., 2008). Post-transcriptional gene regulation is essential to sustain cellular
metabolism, coordinating maturation, transport, stability and degradation of all classes of
RNAs. Mechanistically, each of these events is regulated by the formation of different
ribonucleoprotein (RNP) complexes with RBPs at their core (Stefanie Gerstberger et al.,
2014). The recent development of large scale quantitative methods, especially next
generation sequencing and modern protein mass spectrometry, facilitates not only the
identification of RBPs but also their protein cofactors and RNA targets. Deep sequencing
approaches using immunoprecipitation of RBPs as well as in vitro evolution methods have
revealed the binding ranges of RBPs and showed that many RBPs bind to thousands of
transcripts in cells at defined binding sites (D Marchese et al., 2016)
Central to the function RBPs is their ability to recognize the proper RNA target. RBPs often
exhibit high specificity for their RNA targets by recognizing specific sequences and structures.
The kinetics of these interactions vary depending on function; some RBPs remain bound to an
RNA until degradation whereas others only transiently bind to RNA to regulate RNA splicing,
processing, transport and localization (Stefl et al., 2005). RBPs that are structural elements
of cellular machinery can be very tightly bound in stable complexes, whereas proteins
involved in RNA processing or export must dissociate readily from their targets.
15

RBPs can be classified by their interacting RNA targets. As such they are largely grouped into
binding mRNA, rRNA, tRNA, small nuclear RNA (snRNA), and small nucleolar RNA
(snoRNA), as well as a residual ncRNA binding category. Almost all categories of RBPs are
directly or indirectly invested in the process of protein synthesis. It has been estimated that
692 proteins are mRNA-binding, 169 are ribosomal proteins, and 130 proteins are involved
in the biogenesis and delivery of charged tRNAs to the ribosome. Another 90 proteins are
involved in the biogenesis of snRNAs; while 122 and 41 RBPs take part in the biogenesis of
rRNA and snoRNA, respectively, which may be an underestimation given that rRNA
biogenesis and the functional roles of some ‘orphan’ snoRNAs have yet to be fully
characterized. Another group comprises 122 RBPs that interact with the remaining ncRNAs,
including all remaining ncRNA categories, such as mi RNAs, PIWI interacting RNAs
(piRNAs), long ncRNAs and more (Shazman et al., 2008).

Most RBPs have modular structures, composed of one or more of a limited set of RNA-
binding domains carrying specific motifs. These domains are arranged in varying combinations
to fulfill the need for a diversity of RNA-binding functions (Glisovic et al., 2008). The
sequence variation and organization of the domains determines a given RBP's ability to
recognize a specific RNA. The best-known RNA-binding domains are the RNA recognition
motif (RRM), the dsRNA-binding domain, and the zinc finger domain ( Lunde et al.,
2007). This thesis focuses on a protein carrying three RRM domains.
16

Figure 2: Conserved structure of RNA recognition motif (RRM).
(Nagai et al., 1990)

The RRM is the most common RNA binding motif. Its core is a small protein domain of 75–
85 amino acids that forms a four-stranded anti-parallel β-sheet supported at the back by two
α helices (Fig. 2). Characteristic for this domain are conserved sequences carrying aromatic
residues that project from the β-sheet surface and that can stack on un-base paired RNA bases.
Proteins carrying RRMs play a role in numerous cellular functions, especially in
mRNA/rRNA processing, splicing, translation regulation, RNA export, and RNA stability,
polynucleotide extension, RNA splicing, subcellular RNA localization, cellular export,
translation (initiation elongation and extension), and RNA destruction (K S Manning et al.,
2017)
Members of the RRM family, which consists of over 1200 proteins, are implicated in the
etiology of disease. Therefore, advances in the understanding of these proteins holds
promise to be directly applicable to the treatment of human neurodegenerative diseases,
cancers and developmental disorders (Castello et al., 2013).
17

RBPs were first characterized using biochemical methods, such as gel electrophoresis of
labeled RNAs with native or ultraviolet-crosslinked nuclear extracts or RNA affinity
purification coupled with mass spectrometry and/or immunodetection. Further insights into
the specific RNA targets of RNPs and their protein composition were gained using RNA
immunoprecipitation (RIP) coupled with cDNA array hybridization of recovered RNA and
protein mass spectrometry, respectively (Gerstberger et al., 2014). Surface plasmon
resonance (SPR) is a technique that has been used to study the dynamics of RNA/protein
interactions in real time. (Park-Lee et al., 2003). A recent novel biophysical method called
switchSENSE can also provide insight into the kinetics RNA/protein interactions and in
addition, this technique can help detect structural conformational changes upon RBP binding
to RNA (Clery et al., 2017). This thesis focuses on switchSENSE technology.

1.2 Implication of RNA-binding proteins in human diseases

RNA-binding proteins are key components in RNA metabolism, regulating the temporal,
spatial and functional dynamics of RNAs. Altering the expression or function of RBPs has
profound implications for cellular physiology, affecting RNA processes from pre-mRNA
splicing to protein translation. Recent genetic and proteomic data have revealed that RBPs are
involved in many human diseases ranging from neurologic disorders to cancer.
18

Figure 3: A network of RNA-binding proteins in human diseases. Aberrant expression or
functions of RNA binding proteins (RBPs) (green) have been identified in several major
classes of human diseases (orange) including neurologic disorders, muscular atrophies and
cancer. Aberrations in RBPs are sometimes directly (solid lines) or indirectly (dashed lines)
associated with specific disease (blue). (Lukong et al., 2008).

Neurodegenerative disorders are the principal clinical manifestation of RBP defects,
because of the high prevalence of alternative splicing in the brain, a major function
fulfilled by RBPs.
19

Figure 4: RBP-associated mutations may
disrupt RNA processing by several
mechanisms.
(Kapeli et al., 2017)

Many neurobiological mechanisms are
regulated by Transcriptional and
posttranscriptional regulation of gene
expression. Most of the RBPs (ABOUT
50%) are present in the brain and are
identified as key regulators of
neurodevelopment and neurobehavioural
functions. (Bryant et al., 2016)
Broadly, the diseases related to RBPs are
grouped into RBP loss of function and gain
of function. A loss of function is observed
when genetic changes or autoimmune
antibodies lead to the inactivation of RBPs.
For example, a trinucleotide repeat (CGG) expansion in the 50-untranslated region (UTR) of
the fragile X mental retardation (FMR1) gene results in the loss of function of the FMRP
causing fragile X syndrome (FXS) (Preprah et al., 2013). The toxic RNA gain of function is
usually observed when microsatellite expansion repeats are transcribed into mRNAs resulting
in the entrapment of RBPs that associate with the repeats and interfere with the normal
function of RBPs. This is
20

exemplified by myotonic dystrophy type 1 (DM1), in which a CUG trinucleotide expansion
is present in the 30 UTR of the myotonic dystrophy protein kinase (DMPK) mRNA (Lee et
al., 2013).
Dysfunction or mutation of RBPs can exert global effects on their targetomes that underlie
neurodegenerative disorders such as Alzheimer’s and Parkinson’s diseases as well as
neurodevelopmental disorders, including autism and schizophrenia. (Bryant and Yazdani,
2016). Some of the RBPs in nervous system are shown in Table 1.

Table 1: Selected RBPs and their association with neurological diseases.
(Lin et al., 2014).
21

1.3 RNA binding proteins of the ELAVL family

Figure 5: Diverse mechanisms of ELAVL protein functions include mediating many post-
transcriptional processing events in both nucleus and cytoplasm (Adapted from Hinman et
al., 2008)

ELAVL proteins (Embryonic Lethal, Altered Vision-Like) are named after their
Drosophila ELAV homologues, which show a vision defect when the gene is mutated
(Koushika et al.,1996) They perform a variety of functions as shown in Fig 5. In
mammals, the ELAVL protein family (previously called Hu proteins) includes four
members, ELAVL1 (HuR), ELAVL2 (HuB or Hel-N1), ELAVL3 (HuC) and ELAVL4
(HuD). The neuronal ELAVL proteins (ELAVL2-4) are implicated in neuronal
plasticity and memory, neuronal development, and many other functions. They are
classical RBPs containing three RNA recognition motifs (RRM1, RRM2, RRM3 with
a hinge region between RRM2 and RRM3). The three RRMs cooperate to provide an
22

RNA-binding platform that recognizes and binds to AU rich sequences of mRNA
targets (Park et al., 2000)
ELAVL1, the ubiquitously expressed ELAVL protein, plays roles in DNA damage
response, negative regulation of apoptosis, response to hypoxia, carcinogenesis,
inflammation among others (Hua et al., 2013). The other three ELAVLs are implicated
in neuronal differentiation, maintenance, learning, memory, regulation of neuronal
excitability (Hua et al., 2013) Roles outside the nervous system are limited. The
proteins are expressed in the reproductive system (Good et al., 1995) and ELAVL4 is
implicated in the regulation of insulin in pancreatic beta cells (Margo et al., 2013)

Numerous studies indicate that neuronal ELAVL proteins play an important role in
neuronal development. For example, over-expression of ELAVL4 accelerates neurite
outgrowth in E19 rat cortical neurons, PC12 cells, and retinoic acid-induced
embryonic stem cells (Anderson et al., 2000) while antisense-mediated inhibition of
ELAVL4 expression in PC12 cells results in a failure to grow neurites upon nerve
growth factor stimulation (Mobarak et al., 2000).
ELAVL4, encoded by a gene located on chromosome 1, consists of about 375 amino
acids (there can be alternative translation starts and/or alternative splicing at the N-
terminus and at internal exons). It is found most abundantly in the brain (Fig. 6) and
plays major roles in brain development and plasticity through mRNA-based regulation
23

(Fig.7). ELAVL4 was first identified as a target antigen in paraneoplastic
encephalomyelitis-sensory neuropathy, an autoimmune disease associated with small
cell lung cancer and neuroblastoma (Dalmau et al., 1992).

Figure 6: RNA seq data of ELAVL4 in the body. (Fagerberg L et al., 2013)

Figure 7: Model depicting the various functions of ELAVL4 in the cell. A) In the nucleus,
ELAVL4 regulates alternative splicing and (B) ELAVL4 forms part of an mRNP complex, and
it likely facilitates mRNA export into the cytoplasm. (C) In the cytoplasm, ELAVL4 and
destabilizing RBPs bind competitively or cooperatively to mRNAs. (D) ELAVL4 transports
mRNAs to different compartments of neurons, most notably neurites, along microtubules.
(E) At the synaptic terminal, ELAVL4 may promote or repress translation of transcripts.
(Bronicki et al., 2013)
1.4 ELAVL4 interacts with AU-rich RNA

24

ELAVL4 protein has been shown to bind to AU rich elements in the 3’ untranslated regions
of unstable mRNAs, causing the stabilization of certain transcripts, such as c-FOS (Chung et
al., 1996), cell cycle regulator p21(Joseph et al., 1998), neuron specific mRNAs such as N-
myc (Ross et al.,1999) and GAP-43 (Chung et al., 1997).
Previous studies of ELAVL4/RNA interaction using equilibrium and kinetic methods have
suggested that a single molecule of ELAVL4 requires at least three AUUU repeats to bind
tightly to the RNA. (Park-Lee et al., 2003)

Figure 8: Analysis of ELAVL4 binding to different AU rich targets. The left two panels
represent gel shift assays with increasing concentrations of ELAVL4 equilibrated with MUT
(UUA(UCUA)3UU) and AU-3 (UUA(UUUA)3UU). * complex and P- is free RNA. This data
is quantified and plotted in the right panel as the percentage of RNA bound vs. the protein
concentration. (Park-Lee et al., 2003)

Gel shift analyses have revealed that ELAVL4 binds with AU-3 (sequence: UU(AUUU)
3

AUU), a motif found in many mRNAs such as cytokine mRNAs, immediate early proto-
25

oncogene mRNAs, and cell cycle regulatory mRNAs, with an apparent affinity (K
D
) in nM
range while it shows no binding to a mutant RNA target in which the central U of each repeat
is replaced by a C (Fig. 8).

Figure 9: Left panel: Quantitated gel shift data for the different mutants is plotted as the
percentage of RNA bound vs. the protein concentration. Right panel: Comparison of
equilibrium binding affinities of ELAVL4 and deletion mutants for AU-3 RNA. Clones and
their respective residues are given on the left. KD determined by gel shift assays is given on
the right. (Park et al., 2003)

To determine the function of each of the domains involved in binding of ELAVL4 to the
mRNA prototype for a stable complex formation, a series of deletion mutants were made and
analyzed for their interaction to the prototype mRNA using gel shift assays as shown in Fig 9.
This equilibrium-based data suggested that RRM1is the critical domain, as its deletion severely
affects RNA binding. In contrast, deletion of RRM2 or three only marginally reduced the
af finity for the AU3 target. Individually, RRM2 and RRM3 bind much more weakly than
RRM1alone, supporting the idea that the roles of RRM2 and 3 are secondary to that of RRM1.
Two caveats of gel shift assays are, first, that they depend on detection of complex formations
in equilibrated binding assays, and thus do not provide kinetic data. Secondly, that detection
26

of complexes is dependent on the "caging effect" Most complexes have a dissociation time
much shorter than the gel run but are detectable because the complexes rebind during the gel
run because the molecules are trapped together in the matrix. Complexes that have a fast
dissociation rate and/or a slow association rate are less likely to be detecting in this assay
because dissociated RNA may "escape" before it can be rebound by protein.

Figure 10: Kinetic analysis of mutant ELAVL4 proteins at 1.2nM,3.6nM and 11nM with AU-3
(Park-Lee et al., 2003)

Table 2: Kinetic values for complexes of ELAVL4/HuD with AU-3 RNA. (Park-Lee et
al., 2003)

To follow the kinetics of the ELAVL4/RNA interaction in real time, surface plasmon
resonance (SPR)-based biosensor was used. SPR-based biosensors can be used to monitor
27

binding kinetics in real time. The phenomenon of surface plasmon resonance causes a
reduction in the intensity of light reflected at a certain angle from the sensor surface. As
molecules bind to the surface, the refractive index close to the surface changes, altering the
angle of minimum reflected intensity. The change in the SPR angle is proportional to the mass
of material bound. The result from the change in refractive index is displayed as a sensogram
with the binding responses on the y axis against time on the x axis. Using this method, the
ELAVL4/RNA interaction was studied. The association time for the ELAVL4 wild type and
deletion mutants was recorded for 2min, followed by a 3-min flow of running buffer during
which dissociation could be observed. This data confirmed that RRM1 is a critical domain for
binding to AU rich RNA, the loss of which would lead to a strong loss in affinity. Deletion of
RRM2 or RRM3 alone only marginally reduced affinity, both in both cases, the kinetics were
affected in a very interesting way: deletion of either RRM2 and RRM3 sped up association as
well as dissociation (Table 2). In other words, it causes kinetic changes that partially
compensate for each other, thereby limiting deleterious effects on the apparent binding
affinity. This could mean that RRM2 and 3 are functionally important for stable complex
formation but also in some way inhibit complex formation. This suggests that a conformational
change might occur during binding.

28

1.5 Role of the hinge

The role of the three domains involved in binding to RNA has been studied but the role of
hinge region in contributing to this complex formation has not yet been fully understood.
Though the paper by Park et al. (2000) suggests that the removal of hinge did not affect
equilibrium binding (Fig 9), it led to a further change in the kinetics of binding as shown in
Fig 10 and Table 2. This indicates that the hinge itself may occlude binding as well as stabilize
the RNA/protein complex. If ELAVL4 undergoes a rearrangement upon RNA binding this is
difficult to detect by conventional methods.

Of further interest is that the position of the hinge between the accessory domains 2 and 3
may allow this domain to mediating/affect a change in tertiary structure of the protein that
might occur as all three RRMs accommodate the RNA on their surface, generating a stable
RNA-protein complex. The hinge domain can be alternatively spliced and has been reported
to be post-translationally modified. Post-translational modifications can affect RBP function
(Fig. 11) and in this case, both splice variation and modifications might influence the role of
the hinge and thereby provide a mechanism to modulate RNA binding.
29

Figure 11: Schematic of the effects of PTM on RNA binding proteins. (a) Signal integration.
(b) PTM may activate or deactivate the functions of an RNA binding protein (c) The altered
functions of RNA binding proteins lead to overall differences in the metabolism of RNAs at
every stage of their existence, from transcription through destruction (Lovci et al, 2016).

It is worth considering that ELAVL4 proteins might compete with other proteins for binding
to AU rich sequences, and that the identity of the bound protein might decide the fate of the
mRNA. If such competition exists, it is essential to study not only the equilibrium binding
affinities but also understand the dynamic nature of the interaction.
The recently developed novel biophysical switchSENSE technology can provide structural as
well as kinetic data under equilibrium as well as dynamic conditions, and thus provides an
ideal tool to gain further understanding of this RNA/protein interaction. This thesis focuses
on using a DRX2 switchSENSE biosensor to study the interaction between ALVL4 and AU-
rich RNA.

30

CHAPTER 2: MATERIALS AND METHODS

2.1 Preparation of constructs

To investigate the role of each of the domains and the hinge, different constructs were made,
carrying two tags for maximal flexibility in coating the protein onto the sensor chip. The
constructs were made to have a biotinylatable tag at the N-terminus and a hexa histidine tag
at the C-terminus. The latter serves for purification of recombinant proteins using a nickel
affinity column and can also be used to couple the protein to the sensor chip surface. In that
case, a special nickel affinity chip would be used. The biotinylatable tag consists of a 19-amino
acid sequence MSGLNDIFEAQKIEWHGAP, which can be biotinylated in E. coli when the
biotin ligase BirA is co-expressed. This provides another way to coat the protein on the sensor
chip surface using a streptavidin-linked nanolever.

Figure 12: Construction of ELAVL4 and mutants in the two RRM mutants, key aromatic
residues were mutated to alanine. Each protein also carries a biotinylatable tag at the N
terminus and a hexa histidine tag at the C terminus.

31

The WT ELAVL4, K318, H74, K220 constructs were regrown from stored plasmid aliquots
and sequenced for confirmation. They were digested with Bsu36I (which two codons
downstream of the N-terminal methionine) and NotI (which is at the C-terminus before stop
codon) to obtain the protein-encoding fragments and inserted into the plasmid
(Bsubiotag2/H600) containing the biotinylatable tag at the N terminal and his tag at the C
terminal. J300 and K208 were newly made. The individual mutant preparations are listed
below.
2.1.1 Preparation of WT ELAVL4 construct

Figure 13: Schematic representation of the WT ELAVL (amino acids 1-373).

Full length ELAVL4 (WT) was generated as described in Park et al., 2000. An aliquot of the
DNA construct was used to grow minis and sequenced to verify it end coded the ELAVL4
WT construct with C terminal his tag. A 19 amino acid MSGLNDIFEAQKIEWHGAP
biotinylatable tag on the N terminal end was ligated to produce a doubly-tagged ELAVL4
protein construct named as Bsubio tag2/H600.
2.1.2 Preparation of H74 construct

Figure 14: Schematic representation of the mutant H74 (amino acids 1-224).

32

H74 construct consisting of just RRM1 and RRM2 was generated as described in Sungmin
Park et al., 2003. The construct was double digested with Bsu36I and NotI to get the fragment
containing the RRM1+2 and inserted into the double-digested Bsubio tag2/H600 construct.
2.1.3. Preparation of K318 construct

Figure 15: Schematic representation of the mutant H74 (amino acids 1-224).

K318 construct consisting of mutated RRM1 (at the aromatic residue 52 from tyrosine to
alanine and 91 from phenylalanine to alanine) +RRM2+H+RRM3 was generated as described
in Park-Lee et al., 2003. The construct was double digested with Bsu36 I and NotI to get the
fragment containing the mutated RRM1+RRM2+H+RRM3 and inserted into the double
digested Bsubio tag2/H600 construct.
2.1.4 Preparation of K220 construct

Figure 16: Schematic representation of the mutant K220(amino acids 1-220+257-373)

K220 construct consisting of RRM1+RRM2+ partial hinge deletion (27 amino acids)
+RRM3 was generated as described in Sungmin Park et al., 2003. The construct was double
digested with Bsu36 I and NotI to get the fragment containing the mutated
RRM1+RRM2+partial hinge deletion of 27AA+RRM3 and inserted into the double
digested Bsubio tag2/H600 construct.
33

2.1.5 Preparation of K208 construct

Figure 17: Schematic representation of the mutant K208 (amino acids 1-208+278-373)

The primers used for HDF: TGGTGCATCTTTGTCTAC,
HDR: GTTGTTGGCAAACTTCAC. Using the K220 construct as the template, the mutant
with 70AA hinge deletion was generated using Q5 site directed mutagenesis kit from NEB.
2.1.6 Preparation of J300 construct

Figure 18: Schematic representation of the mutant J300 (1-373AA).

Using the Q5 site directed mutagenesis kit from NEB, with Bsubio tag2/H600 construct as
the template, RRM1+RRM2+H+mutated RRM3 was generated.
The mutations were at the aromatic amino acid residues, tyrosine (Y) and phenylalanine (F)
at positions 289 and 328 respectively to alanine. Using YtoA FP:
5’CATCTTTGTCGCCAACCTGTCCCCCG 3’
YtoA RP: 5’ CACCACCCAGTTCCTGTG 3’ for Y to A mutation and this mutated plasmid
was used as template to mutate the F to A. Using FtoA FP:
5’GGGATTCGGCGCTGTCACCATG 3’ and FtoA RP 5’ TTGCACTTGTTGGTGTTG 3’

34

2.2 Production of recombinant protein ELAVL4:

The recombinant protein was bacterially expressed in Escherichia coli BL21 (J57) cells. J57 is
a plasmid that is pACYC-based, and carries a chloramphenicol resistance gene, two rare
tRNAs (to improve protein production) and an inducible biotin ligase. Proteins were purified
using Nickel beads (Qiagen, Valencia, CA) in sonication buffer (10ml 1M Tris-HCl
(pH=8.0), 20.45g NaCl, 5ml Triton X-100, 0.61g Imidazole) containing 10% glycerol and
increasing concentration of imidazole (10mM, 20mM,40Mm, 250Mm). The protein was run
on an SDS PAGE gel and stained with Coomassie.

Figure 19: SDS PAGE gel of the WT ELAVL4 eluted in 250mM imidazole of two different
preps.

Fraction 1, after buffer exchange using desalting columns from Thermo Fisher Scientific was used for the first
experiment after quantification using Bradford assay but the quantification was incorrect. The standards used
were from a stock of 10mg/ml BSA solution which was not accurate hence the concentration of the unknown
was misinterpreted. However, using the nanodrop, the protein concentration was about 0.21 mg/ml which
when divided by the MW of 42000 gives about 500nM. The experiment was run on the instrument with
different concentrations of the WT protein.
35

Pooling the fractions 2 and 3 together and performing a buffer exchange with PE40 (the standard
recommended switchSENSE running buffer, consisting of 10mM NaHPO4, 40mM NaCl, 0.05% Tween 20,
50uM EDTA, 50uM EGTA], the amount of protein was quantified using BCA assay kit by Thermo Fisher
scientific with the standards table shown in table 3 and the graph in figure 20.
Concentration(ug/ml) BSA STANDARD 1 BSA STANDARD 2
0 0.122 0.109
62.5 0.174 0.18
125 0.238 0.273
250 0.302 0.377
500 0.459 0.736
1000 0.744 1.138
2000 1.268 2.152

Table 3: BSA standards dissolved in PE40 to obtain the standard graph.

Figure 20: Quantification of WT ELAVL4 protein using the standard graph.

By reading the concentration at an absorbance of 0.162 nm, the concentration was estimated
to be 13.3ug/ml, resulting in a molar concentration of 300 nM, (MW ~42000 Dalton).

36

2.3 RNA transcripts

For the assay orientation with RNA as ligand, two DNA-RNA mixed oligonucleotides were
ordered from Integrated DNA Technologies.
1) DNA-RNA 15AU3:
rUrArUrUrUrArUrUrUrArUrUrUrArUATCAGCGTTCGATGCTTCCGACTAATCAGC
CATATCAGCTTACGACTA
2) DNA-RNA 15AU3MUT:
rUrArUrCrUrArUrCrUrArUrCrUrArUATCAGCGTTCGATGCTTCCGACTAATCAGC
CATATCAGCTTACGACTA

37

2.4 Brief description of the methodology to study the RNA protein interaction
using the switchSENSE biosensor.

2.4.1 General principle

Figure 21: Schematic of the general principle of the switchSENSE technology
(www.dynamicbiosensors.com)

switchSENSE is a novel technology that works with electrically switchable DNA nanolevers
on a chip-based gold surface. The general principle is shown in Fig. 21. The negatively charged
DNA gets repeatedly attracted to or repelled from the surface when an alternating electrical
voltage is applied. When the DNA nanolever bearing the fluorophore at its tip encounters the
gold surface, the fluorophore is quenched. In its uncoated state, the chip carries single-
stranded DNA coupled to the gold surface at its 5' end with a fluorophore at its 3' end. A
complementary DNA strand can be "functionalized" by any desired molecule (i.e. chemically
modified so that other molecules can be attached or linked). The complementary nanolever
is then annealed to the DNA on the surface to generate the ligand surface.
38

This process can be monitored by the hybridization of the complimentary nanolever that
increases the fluorescence signal and stabilizes the DNA nanolever. Once the surface is
assembled, the target (analyte) can be injected over the surface, and binding can be detected.
Binding alters the hydrodynamic friction, slowing down the DNA movement, thus, changing
the speed of the nanolevers. The position of the nanolevers relative to the gold surface is
determined by observing the fluorescence emission of the fluorescent dye at the distal end of
the DNA nanolevers.
At positive potentials the DNA nanolevers aligns horizontally to the gold surface and the
fluorescent dye is close to the gold surface resulting in low fluorescence intensity. On
reversing the electrode potential to a negative (repulsive) potential, the DNA nanolevers
gradually move into an upright orientation, which is accompanied by a gradual increase of
fluorescence intensity as shown in figure 22. Depending on the target molecule’s charge,
shape and size, the switching dynamics are changed in a characteristic way.

Figure 22: Schematic of DNA switching process with alternating electrical voltages.
(Source: www.dynamicbiosensors.com)
39

2.4.2 Instrumentation

Figure 23: switchSENSE Analyzer DRX 2400 series. (Source:
www.dynamicbiosensors.com)

Two versions of the switchSENSE analyzer are available: DRX (single color) and DRX2 (dual
color). DRX2 has two light sources and two photon counters optimized for red and green
fluorophores. It can thereby detect both the fluorescent tags on two nanolevers of different
sequences yet simultaneously. This allows for the analysis of two molecules, or the
incorporation of an internal control unmodified surface. Since the red fluorophore is stronger,
it is usually used for coating the experimental sample, while the green-labeled nanolever is
usually used as an internal control.

Figure 24: Schematic of the DNA nanolevers with both the red and green fluorescent dyes.
(Source: www.dynamicbiosensors.com)
40

The instrument consists of three parts:
1. Fluidics
2. Optical detection
3. Autosampler region

2.4.3 The sensor chip

Figure 25: switchSENSE analyzer dual color biochip. (Source: www.dynamicbiosensors.com)

It consists of four flow channels, each with six microelectrodes in series.
One switchSENSE sensor spot (≈0.01 mm
2
) is populated with nanolevers either of one
or two populations (for the two color instrument) at a density of around one million
nanolevers per spot as shown in Fig. 25.

41

Figure 26: Process of regeneration of the flow channel. (Source: www.dynamicbiosensors.com)

One advantage of the technology is that it can be easily fully regenerated for each
measurement as shown in Fig. 26. This is done by stripping the complementary DNA from
the surface using an automated protocol. Having a completely fresh surface for every injection
means that cumbersome optimization of regeneration conditions, required for most surface
plasmon resonance experiments, is not required. This also prevent artefacts in the
experiments that could arise from deterioration of the ligand surface over the course of an
experiment.

2.4.4 Nomenclature of biochip

The chip usually used for all the measurements is the MPC2-48-2-G1R1-S.
MPC2 refers to dual color multipurpose chip, 48 refers to the length of the DNA nanolever
in bases, 2 refers to the number of different DNA sequences on the biochip, G1R1 refers to
42

green and red fluorescently attached nanolevers, S refers to the standard type grade of
biochip.

2.4.5 Software

switchANALYSIS: This software package helps to process and analyze the raw
data from the experiment. It generates the hybridization curves, kinetic curves,
sizing data by determining the hydrodynamic diameter (see below), titration data and
conformational change data.
Hydrodynamic diameter: The size of a molecule appended to the DNA nanolevers can be
described using the hydrodynamic diameter (D
H
), defined as the diameter of a perfect solid
sphere that would exhibit the same hydrodynamic friction as the molecule of interest.
This is based on the lollipop model as shown in Fig. 27, in which the DNA nanolevers as
treated as charged and rigid cylinders that carry a spherical molecule at the distal end.
Thus, the D
h
value reflects primarily the hydrodynamic friction but is usually also a good
estimation of the absolute size of the molecule. The more globular the molecule, the more
accurate the D
h
43

Figure 27: The lollipop model that provides the hydrodynamic diameter of the protein
attached to the DNA nanolever in nanometers. (Source: www.dynamicbiosensors.com)

switchBUILD software is the graphical planning tool used to design and program
the experimental workflows. By dragging and dropping particular assay elements like kinetics,
titration etc. to the workflow, it allows for designing scripts to automate the instrument.
Here the type of chip, measurement channel, buffers used and electrodes for measurement
need to be chosen. It also allows you to input the working concentrations of the ligand and
analyte and predicts the graph of fraction of ligand bound vs. time.
It can also provide an overview for all samples and buffers needed to run an experiment by
clicking the “autosampler” tab. The volumes for each of the samples including the dead volume
is mentioned and it indicates where samples should be placed in the autosampler.
switchCONTROL is a software package that is used to control the instrument by
uploading the appropriate switchBUILD scripts. It converts this task flow into advanced
44

commands for the automatic operation of the instrument. It is in this software that the real
time changes in fluorescence and/ or switching speeds are observed.

2.4.6 Modes of measurement
switchSENSE measures kinetics in both dynamic and equilibrium conditions either in the
fluorescence proximity sensing mode and/or dynamic response mode.

Figure 28: Schematic of the static measurement mode with constant negative voltage. A:
When the ligand is attached to the nanolever, the fluorescence intensity is still high. B: Due
to the binding of the analyte to the ligand the fluorescence intensity reduces. (Source:
www.dynamicbiosensors.com)

In the fluorescence proximity sensing mode, the nanolevers are in an upright position due to
a constant negative voltage as shown in Fig. 28. By detecting the changes in fluorescence, it is
possible to analyze the binding kinetics of the interaction. This mode is suitable for big ligand
molecules interacting with small analytes, something that is not feasible in the same way with
surface plasmon resonance biosensors because the mass change readout is too small.
45

Figure 29: Schematic of the dynamic mode of measurement. A. Represents the
complimentary DNA with the ligand molecule with lower friction and higher switching
speed. B. When an interacting analyte molecule binds to the ligand, the friction increases
and the switching speed decreases. (Source: www.dynamicbiosensors.com)

In the dynamic mode, the nanolevers are constantly attracted to or repelled from the surface
depending on the voltage applied as shown in Fig. 29. It is generally used for small ligand
molecule relative to a larger analyte molecule, as it provides the best signal to noise ratio.
This mode provides information about dynamics, size of interacting molecules, and changes
in fluorescence that occur in the local environment. For ELAVL4 RNA interaction
measurements, the dynamic mode is suitable. In Fig. 30, a sample graph of the expected
effects of fluorescence vs. time with and without protein is shown.

Figure 30: Negative potential applied to the nanolevers in the presence of protein decreases
the switching speed, quantified by dynamic response curve.
46

(Source: www.dynamicbiosensors.com)
By plotting the dynamic response signals in real time, the association and dissociation curves
can be plotted as shown in Fig. 31.

Figure 31: Dynamic response signals allow measurement of k
on
and k
off
parameters, which
can then be used to calculate the K
D
value of protein-RNA complexes.
(Source: www.dynamicbiosensors.com)

2.5 Experimental design:

To study the interaction between WT ELAVL4 and prototype mRNA 15AU3, a long-mixed
DNA-RNA oligonucleotide of the sequence:
5’rUrArUrUrUrArUrUrUrArUrUrUrArUATCAGCGTTCGATGCTTCCGACTAATCAG
CCATATCAGCTTACGACTA 3’ was ordered from IDT so that the DNA region would
hybridize with the nanolever surface while the RNA overhang is free to interact.

47

Figure 32: Schematic of the nanolever with the hybridized DNA RNA complimentary strand.

2.5.1 Setup of the experiment using switchBUILD software

To design optimal parameters for kinetic measurements between WT ELAVL4 and 15AU3,
switchBUILD software was used. A new script was developed by selecting "new" under
"properties". The chip type was chosen as MPC2-48-2 (dual color-48mer chip) and channel
1 was selected as it showed relative fluorescence level to be above 80% (>50% is good,
between 20-50% is moderate but electrodes can still be used) The standard buffer (PE40 pH
7.4) was used for all measurements in the dynamic mode as shown in Fig. 33.
The association parameters were set as 200 ul with 100ul/min for 2 min and the dissociation
parameters 10,000 ul with 1000 ul/min for 10 min as shown in Fig. 34.

The first step is the "passivation" of the channel. "Passivation" is basically running a control
fluid over the surface to minimize non-specific binding in the ensuing injections. It also allows
the electrode status the be determined. By using the + tab, different assay elements like
kinetics were chosen. To ensure that after the run the chip is prepared for storage, the selected
channel is vented, and the surface is reset by removing analytes and ligands and the surface is
re-passivated to determine chip status. By clicking the autosampler tab, the amount of sample
needed and the placement of the tubes in the autosampler can be determined as shown in Fig.
36. The protocol was saved as Kinetics of ELAVL4 RNA interaction.
48

Figure 33: Selection of assay type, assay elements, channels and buffers. (Adapted from
www.dynamicbiosensors.com)

Figure 34: Selection of the measurement mode, interacting partners, kinetic parameters,
and measurement spots. (Adapted from www.dynamicbiosensors.com)

49

Figure 35: Predicted plot of percentage of RNA bound vs time based on the concentrations
entered in Fig. 34. (Adapted from www.dynamicbiosensors.com)

Figure 36: Volume of buffers, ligand and analyte needed for the experiment along with
position of the vials to be placed in the autosampler.
(Adapted from www.dynamicbiosensors.com)

50

2.5.2 switchCONTROL software

switchCONTROL software converts this task flow to advanced commands that will automate
the instrument and display in real time changes in the dynamic response and or fluorescence
changes that occur during the measurement.
To initially set up the instrument for running the experiment, the cleaning chip was removed
by clicking the eject chip holder option from the dropdown menu of ‘fluidics’. The
measurement chip was inserted, and the chip holder was retracted. The chip was aligned, and
a prime routine was performed, which is the liquid handling procedure where all residual
buffers and air from the flow channels are removed. This is particularly important as any air/
residual buffers can affect the fluorescence read-outs.
The task flow from the switchBUILD software is then loaded onto the switchCONTROL
software.
During the first step of passivation, the solution containing a thiol reactive compound is used
for assembling the monolayer on the sensor chip that prevents any non-specific binding of
DNA/ proteins as shown in Fig. 37.
51

Figure 37: Passivation of the chip to prevent any nonspecific binding. (Source:
www.dynamicbiosensors.com)

The nanolevers NL-A48: 5’-TAG TGC TGT AGG AGA ATA TAC GGG CTG CTC GTG
TTG ACA AGT ACT GAT-3’ and NL-B48: 5‘-TAG TCG TAA GCT GAT ATG GCT GAT
TAG TCG GAA GCA TCG AAC GCT GAT-3‘bear the green fluorophore representing the
reference nanolever and red fluorophore representing the test nanolever, respectively. To
hybridize the reference and the test nanolevers, the complimentary nanolever named as cNL-
A48: 5’-ATC AGT ACT TGT CAA CAC GAG CAG CCC GTA TAT TCT CCT ACA GCA
CTA-3’ was used for the reference while the experimental complimentary nanolever
containing the RNA binding element 15AU3 of the sequence,
5’rUrArUrUrUrArUrUrUrArUrUrUrArUATCAGCGTTCGATGCTTCCGACTAATCAG
CCATATCAGCTTACGACTA 3’ was used as the test.
During the hybridization, the single nanolevers bearing the fluorophore bind to the incoming
complimentary sequences thus, stabilizing the nanolevers. This causes an increase in the
fluorescence intensity on both the reference and test nanolever as shown in Fig. 38. This
increase in signal is due to the increased rigidity of the now double-stranded DNA on the
surface.
52

Figure 38: Increase in fluorescence intensity vs. time during hybridization of the
complimentary nanolevers.

This is followed by the kinetics: association of the ligand WT ELAVL4 which binds to
nanolevers containing the RNA overhang.
During association, a decrease the fluorescent intensity as well as the decrease in the dynamic
response (DR) in real time is observed. This is followed by a stopped-flow analysis which
includes sizing of the ligand and analyte complex.
The dissociation leads to an opposite effect where the buffer is flushed through the channels
and the analyte is washed off the surface resulting in a relative increase in dynamic response
due to the increase in switching speed and an increase in the relative fluorescence. This phase
can also be observed in real time in the switchCONTROL software.
53

2.6 Control experiment with the WT ELAVL4 and 15AU3MUT

Figure 39: Schematic representation of nanolever with 15AU3MUT ligand and ELAVL4 as the
analyte.

As a control, the mutant RNA with the central U’s replaced by C with sequence,
5’rUrArUrCrUrArUrCrUrArUrCrUrArUATCAGCGTTCGATGCTTCCGACTAATCAG
CCATATCAGCTTACGACTA3’ was used. This would be expected to show little or no
binding to the WT ELAVL4, based on published experiments (Park et al. 2001?). The same
association and dissociation parameters as that of the 15AU3 ELAVL4 interaction experiment
was used.

54

CHAPTER 3: RESULTS
3.1 Analysis of WT ELAVL4 with 15AU3 RNA, Experiment 1

Using the protein fraction with stock concentration of 500nM estimated by the nanodrop,
serial dilutions were made with a factor of 2 i.e. with concentrations of 250nM,125nM,
62.5nM and 31.3nM. Hybridization of the complimentary strands and kinetics of the
interaction between WT ELAVL4 and 15AU3 RNA were observed in real time. The
measurement data was imported to the switchANALYSIS software and analyzed. The arrow
indicates that measurement data (already sorted as functionalization, association, sizing and
55

dissociation) was added to the panel on the right by simply dragging and dropping as shown
in figure 40.

Figure 40: Importing the measurement data file to the switchANALYSIS software. (Adapted
from www.dynamicbiosensors.com)

3.1.1 Hybridization data analysis

Figure 41: switchANALYSIS software and evaluation of various assay elements by choosing
the appropriate analysis tool. (Adapted from www.dynamicbiosensors.com)

56

By dragging and dropping the functionalization data and clicking the ‘Create new analysis’ tab
on the lower right end, a window pops up with different types of analysis tools. To analyze
functionalization data, the overview tool was selected as shown in Figure 41.
Notice that the curves that are higher represent the complimentary nanolever 15AU3
hybridizing with the NLB48. The lower curves represent the reference complimentary
nanolever cNLA48 hybridizing with NLA48 as shown in figure 42.

Figure 42: Hybridization of the nanolevers. (Adapted from www.dynamicbiosensors.com)

By choosing the kinetics analysis tool, association and dissociation data can be dragged and
dropped simultaneously on the right panel and using the clicking the ‘Create new analysis’
tab on the lower right end. To analyze kinetic data, kinetics tool was selected as shown in
figure 43.
3.1.2 Kinetic data analysis

57

Figure 43: Kinetics assay tool was used to analyze the kinetics of 15AU3 with the WT
ELAVL4. (Adapted from www.dynamicbiosensors.com)

By simply dragging the appropriate reference to the experimental measurement, the
appropriate pair is selected and added into the association or dissociation boxes respectively.
The normalized data is shown in the graph as indicated in Fig. 43. The borders of the curves
can be chosen by dragging the fitting borders on the graph. The fit data tab is chosen to obtain
the global curve fit of the kinetic data as shown in Fig. 44.
58

Figure 44: switchANALYSIS software referencing for the analysis of kinetic data (Adapted
from www.dynamicbiosensors.com)

Figure 45: Dynamic response kinetic data shows the K
on
and K
off
as 4.68± 0.24 X 10
5
and 4.36±
0.5 X 10
-3
respectively with a K
D
of 9.33±1.17nM. (Adapted from www.dynamicbiosensors.com)

This data can also be plotted in the fluorescence proximity sensing mode. The association of
WT ELAVL4 to 15AU3 RNA increases the fluorescence as shown in Fig. 45.

59

Figure 46: Fluorescence proximity data observed for concentrations 500nM, 250nM,125nM,
62.5nM during association. (Adapted from www.dynamicbiosensors.com)

Comparing the two association rates from Fig. 45 and 46, the difference in K
on
is by a factor
of 10 which means that there are fluorescence intensity changes occurring in the local
environment during association, that are much slower compared to the dynamic response
changes.

3.1.3 Sizing data analysis

To obtain the sizing data, a new tab is created by using the + sign and sizing is chosen as
shown in Fig. 47.

Figure 47: Selection of the sizing tool to size analysis of 15AU3 and complex with WT
ELAVL4(Adapted from www.dynamicbiosensors.com)

By dragging and dropping the reference cNLA48 to the test 15AU3, the sizing data was
plotted in the stopped flow as shown in Fig. 48. The graph of normalized fluorescence vs.
time is plotted with the signal and reference curves with fits in blue and orange respectively.
60

Here n=4, taking into consideration the highest four concentrations of 15AU3-ELAVL4
bound complex that show saturation and no unspecific association curves observed.

Figure 48: Sizing data for the complex formed between 15AU3 and WT ELAVL4 is about
3.5±1.35 nm with n=4 (Adapted from www.dynamicbiosensors.com)

3.2 Analysis of WT ELAVL4 with 15AU3 RNA, Experiment 2

This experiment was done using the fraction of ELAVL4 protein with stock concentration of
300nM estimated by the BCA assay. Serial dilutions were made with a factor of 2 i.e. with
concentrations of 150nM, 75nM and 38nM, 19nM and 9.4nM. Hybridization of the
complimentary strands and kinetics of the interaction between WT ELAVL4 and 15AU3 RNA
were observed in real time.
The analysis was done in the same way as mentioned in 3.1.2.
In this experiment the syringe / peristaltic pump controls were mixed up in the script due to
the fact that a loaner DRX2 machine was being used and it did not have the updated software.
Because of this, no buffer at all was injected into the measurement channel during
61

the “dissociation” step, hence there is no dissociation observed. But the association curves to
obtain K
on
were analyzed in the dynamic mode as shown in Fig. 49A as well as the fluorescence
proximity sensing mode in Fig. 49B.
The observed-on rate was about 2.35±0.26 X 10
5
M
-1
s
-1
in the dynamic response mode and a
tenfold change is observed in 4.38 ± 0.28 X 10
5
M
-1
s
-1
.
3.2.1 Kinetic data analysis

Figure 48: Association data for concentrations 300nM,150nM, 75nM and 38nM, 19nM 9.4nM
AND 4.7nM A: Dynamic mode association curve, B: Fluorescence proximity sensing mode
curve (Adapted from www.dynamicbiosensors.com)

3.2.2. Sizing data analysis

Using the sizing tool as described in 3.1.3, the sizing of the entire complex was performed,
and the results are shown in Fig. 49. Here n=4, taking into consideration the highest four
concentrations of 15AU3-ELAVL4 bound complex that show saturation and no unspecific
association curves observed.
62

Figure 50: Sizing data for the complex formed between 15AU3 and WT ELAVL4 is about
4.22± 0.77nm with n=4 (Adapted from www.dynamicbiosensors.com)

3.2.3 Titration curve analysis

A titration curve can be plotted by using the stopped flow measurement data into the titration
analysis tool as shown in Fig. 51 and the ‘fit analysis’ tool can be used to fit the titration curve
as shown in Fig. 52.

Figure 49:

Figure 51: Selection of the titration tool to determine the equilibrium dissociation constant.
(Adapted from www.dynamicbiosensors.com)

63

Figure 52: Titration curves plotted for 300nM,150nM, 75nM and 38nM, 19nM 9.4nM and
4.7nM on electrodes 3 and 5 using the stopped flow measurement data. (Adapted from
www.dynamicbiosensors.com)

Figure 53: Titration curve with a K
D
of about 83.5nM. (Adapted from
www.dynamicbiosensors.com)

3.3: Analysis of WT ELAVL4 with 15AU3 RNA, Experiment 3.

Using the fraction with stock concentration of 300nM estimated by the BCA assay, serial
dilutions of ELEVL4 protein were made with a factor of 2 i.e. with concentrations of 150nM,
75nM and 38nM, 19nM and 9.4nM. Hybridization of the complimentary strands and kinetics
64

of the interaction between WT ELAVL4 and 15AU3 RNA were observed in real time. This
was a replication of experiment 2 with the same kinetic parameters.
3.3.1 Kinetic data analysis:

As in experiment 2, in this experiment the syringe / peristaltic pump was mixed up in the
script, so that no buffer was injected into the measurement channel during
the “dissociation” step, hence no dissociation was observed. But the association curves to
obtain K
on
was analyzed in the dynamic mode as shown in Fig. 54A as well as the fluorescence
proximity sensing mode in Fig. 54B. There was air in the system which is represented by the
sudden jumps in the Fig. 54B. This can be eliminated by priming the system with PE40.
This does not affect the association curve as the jumps are right before the start of association.
The K
ON
in the dynamic response mode was found to be 1.05± 0.03 X 10
5
and in the
fluorescence mode 2.25 ± 0.02 X 10
4
M
-1
s
-1
which is consistent with the results obtained in
the previous two experiments.

Figure 54: Association data for concentrations 300nM,150nM, 75nM and 38nM, 19nM 9.4nM
AND 4.7nM A: Dynamic mode association curve, B: Fluorescence proximity sensing mode
curve (Adapted from www.dynamicbiosensors.com)

65

3.3.2 Sizing data analysis:

Using the sizing tool as described in 3.1.3, the sizing of the entire complex was performed,
and the results are shown in Fig. 55. Here n=6 taking into consideration, highest six
concentrations of 15AU3-ELAVL4 bound complex that show saturation and no unspecific
association curves observed.

Figure 55: Sizing data for the complex formed between 15AU3 and WT ELAVL4 is about
4.5±1.79 nm with n=6 (Adapted from www.dynamicbiosensors.com)

3.3.3 Titration data analysis

Using the titration tool as described in 3.2.3, the titration analysis was performed and is as
shown in Fig. 56.
66

Figure 55: Titration curve with a K
D
of about 145±73nM. (Adapted from
www.dynamicbiosensors.com)

3.4 Analysis of WT ELAVL4 with 15AU3MUT RNA, Experiment 4.

The protocol for this interaction was created as described in Chapter 2, 2.5.1. Only one
concentration of 300nM of WT ELAVL4 was made to interact with 15AU3MUT.

3.4.1 Kinetic data analysis

It is observed that there is no association of the WT ELAVL4 with 15AU3MUT as shown in
Fig. 57.

67

Figure 57: Interaction of WT ELAVL4 and 15AU3MUT showed no association/dissociation in
the dynamic mode (no binding). (Adapted from www.dynamicbiosensors.com)

An interesting observation of the kinetics in the fluorescence mode as shown in Fig. 58,
even if there is no association, there is some signal/change in fluorescence.

Figure 58: Interaction of WT ELAVL4 and 15AU3MUT showed change in fluorescence
intensity during association

The assay orientation with RNA as ligand and protein ELAVL4 as analyte was one of the
many ways this interaction could be studied. Using switchSENSE technology, the
immobilized molecule can also be coupled in an inverse orientation such that the protein
ELAVL4 is the ligand interacting with the analyte 15AU3.

68

3.5 Analysis of WT ELAVL4 with 15AU3MUT RNA in the inverse orientation
using the NTA assay, experiment 5.

In this assay orientation, the WT ELAVL4 with his tag is immobilized on the surface and the
RNA is flowed in to observe the interaction.

Figure 59: Schematic representation of the inverse orientation with WT ELAVL4 as analyte
and 15AU3 RNA as ligand. (Adapted from www.dynamicbiosensors.com)

3.5.1 General workflow of the NTA assay

Figure 60: Workflow overview with Tris -NTA nanolevers
(Adapted from www.dynamicbiosensors.com)

The NTA assay basically includes the following steps as shown in Fig. 60:
1. Functionalization of the biochip with the tris-NTA carrying nanolevers.
69

2. Activation of tris-NTA with Ni2+ ions.
3. Ligand binding.
4. Interaction measurement with the analyte.

3.5.2 switchBUILD script for the NTA assay

The switchBUILD script was modified as shown in Fig. 61. A single concentration of 300nM
WT ELAVL4 was used to test the reverse orientation.
The association parameters were set to 600ul with 200ul/min for 3 min and the dissociation
parameters were set to 2000ul with 200ul/min for 10min.
Figure 61: switchBUILD script for the assay development of ELAVL4 15AU3 RNA
interaction. (Adapted from www.dynamicbiosensors.com)

70

3.5.3: switchANALYSIS data

Using the switchANALYSIS software the data for hybridization, kinetics and sizing were
generated by following 3.1.1,3.1.2 and 3.1.3

3.5.3.1: Hybridization of the cNLB48-NTA with the WT ELAVL4.

The association of cNLB48-NTA with the WT ELAVL4 with a decrease in speed of
nanolevers indicated by the decrease in dynamic response signal is as shown below in figure
62.

Figure 62: Hybridization curve of 300nM ELAVL4 binding to the cNLB-NTA nanolever
(Adapted from www.dynamicbiosensors.com)

3.5.2.2 Kinetic analysis

Kinetic analysis of WT ELAVL4 with the 15AU3 RNA is as shown below with a K
ON
in the
dynamic response mode was found to be 4.47± 7.31 X 10
5
and in the fluorescence mode
2.25 ± 0.02 X 10
4
M
-1
s
-1
which is consistent with the results obtained in the previous two
71

experiments. The dynamic response shows decrease in fluorescence intensity during
association and an opposite trend in dissociation.

Figure 63: Association and dissociation curves in the dynamic mode of 300nM ELAVL4 to
15AU3RNA with a K
ON
of 4.47±7.31 X 10
5
, K
OFF
of 1.38 ±0.3 X 10
-1
and K
D
of 309± 509nM.
(Adapted from www.dynamicbiosensors.com)

The fluorescence proximity sensing curve shows an increase in fluorescence intensity during
association and an opposite trend in dissociation as shown in Fig. 64.

Figure 64: Association and dissociation curves in the Fluorescence proximity sensing mode.
(Adapted from www.dynamicbiosensors.com)

72

CHAPTER 4: DISCUSSION AND FUTURE DIRECTIONS

4.1 Kinetics of the DNA RNA mix as the ligand and WT ELAVL4 as the analyte
in dynamic response mode.

Based on the results obtained for experiments 1,2 and 3, It can be observed that the K
ON
is
about 4.68 ± 0.24 X 10
5

M
-1
s
-1
, 2.35± 0.26 X 10
5

M
-1
s
-1
and 1.05± 0.03 X 10
5

M
-1
s
-1
for
concentrations of 500nM serially diluted by factor of 2 to 31.3nM for experiment 1 and
dilutions by a factor of 2 from 300nM to 4.7nM for experiments 2 and 3.
The SPR data yielded a K
ON
of about 4.21± 0.12 X 10
6
with concentrations 3.6nM, 11nM,
33nM. (Park-Lee et al., 2003)
73

It is observed the K
ON
using SPR is about 10 times faster than that yielded by the switchSENSE
technology. Differences in observed association rates could arise by differences in flow rates,
the effect of ligand density, mass transport limitations etc. The flow rates for the experiments
with SPR were much slower (about 30ul/min for 2 min) and the switchSENSE flow rates
were about 200ul/min for 2 min. Because mass transport limitations are so common in SPR
biosensors (it is difficult to get enough signal if the chip density if too low), the switchSENSE
chip has been designed in such a way as to reduce the effect of mass transport over the chip
surface as shown in Fig. 65. The lower ligand density compared to the SPR biosensor (Source:
www.dynamicbiosensors.com) and the higher flow rate should reduce mass transport
limitations.

Figure 65: Effect of ligand density in the switch SENSE biochips. A). With increase sample
concentration, the mass transport limitation is high. B) With increasing flow rates, the mass
transport limitation is lower compared to A. C) With decreased ligand density and
increased flow rates mass transport limitation is overcome. (Source:
www.dynamicbiosenors.com)

74

An experiment to characterize effect of ligand density and mass transport limitation to
determine difference in association rates (if any) was performed by Dynamic biosensors as
shown in Fig. 66.

Figure 66: Effect of ligand density and mass transport limitations using a biotin/anti-biotin
mAb model system was tested with different densities and flow rates represented in the
graph of K on vs density. A) The graphs on the left panel describe the association flow rates
with different ligand densities, low, medium and high. With low ligand density and high
flow rate, the curve at the bottom left shows a good association curve. B) Green dotted line
represents the 10ul/min association flow rate, red represents the 100ul/min association
flow rate and blue line represents 1000ul/min association flow rate. With higher flow rates
and lower ligand density the association curve yields more reliable K
ON
values. (Source:
www.dynamicbiosenors.com)

If indeed mass transport was limiting in the SPR data, the DRX2 should show faster on rates.
The fact that on rates are slower using the DRX2, with lower surface densities and faster flow
75

rates indicates that something else is affecting the data. Besides a lower surface density and
flow rate, the salt concentration of the buffer used in the DRX2 (40 mM,) is lower than the
150 mM salt buffer used in previous SPR experiments. Low salt conditions could affect the
association of the molecules and this should be investigated in future experiments by repeating
them in buffers with increasing salt concentrations. However, this may destabilize the DNA
nanolever hybrid, which can be mitigated by using the longer 96-mer DNA surface. Another
possibility is that the higher surface density on the SPR biosensor somehow led to
cooperativity in binding of ELAVL4 to multiple RNAs. Lastly, it is also possible that there
may have been differences in the active concentration of the protein preparations. At this point
in time, with the limited number of experiments and protein preparations that were done,
further experiments are needed.

4.2 Kinetics of the DNA RNA mix as the ligand and WT ELAVL4 as the analyte
in Fluorescence proximity sensing mode.

The kinetics graphs with the normalized fluorescence vs. time data in all three experiments
show a decrease in signal during association and an opposite trend during dissociation. In this
case, permanent sensing of its local environment causes the fluorescent dye to emit a specific
intensity of fluorescent light. The association of ELAVL4 influences the local environment,
resulting in a decrease of fluorescence emission.
76

Kinetics of the fluorescence proximity data vs. the dynamic response data is 10 X slower. This
could be explained by a 2-step binding process in which the initial binding induces the quick
reduction in switching speed, followed by a slow process of ELAVL4
reorganization/conformational change which can be seen in the FPS read out (if ELAVL4 is
bent/pulled more towards the fluorescent dye in the process if reorganization, this would
lead to a constant but slower reduction in fluorescence due to the ELAVL4 protein quenching
the dye.

4.3 Dissociation constants in kinetic and equilibrium modes

Dissociation constant in the kinetic mode reveals on off rate of 4.36 ± 0.5 X 10
-3
s
-1
. Off rates
are independent of concentration. This seems to be in good agreement with the SPR data by
Sungmin Park et al., 2003 which is about 3.05± 0.05 X 10
-3
s
-1
. Because mass transport can
also affect dissociation readings, if the SPR data was affected by mass transport, the effect
would be rebinding of molecules due to a too high surface density and a too slow flow rate
(washing away). In the SPR experiments a very low surface density was used to try to avoid
this problem. It is possible that the small increase in off rate seen with the DRX2 is due to loss
of mass transport limitations, but it would be important to repeat the experiment and also
77

examine the interaction in different experimental set ups (RNA on the surface vs. protein on
the surface.
4.4 Affinity of ELAVL4 for 15AU3 RNA.

The equilibrium binding constant plotted as a titration curve yielded a K
D
value of
approximately 80 nM. Using the gel shift analysis, the equilibrium binding constant was about
19±3nM. This 4-fold difference could possibly be explained by differences in active protein
concentration or the effect of using buffers with different salt concentrations. As noted above,
further experiments will be needed to address this question.

4.4 Inverse assay orientation kinetics with the protein bound on the surface
and the RNA flowing in.

The switchSENSE technology offers several alternative experimental setups. For example,
the protein can be immobilized on the surface using amine coupling (to the DNA
nanolever), a biotin streptavidin link (using the biotinylatable tag and a streptavidin linked
nanolever), the hexahistidine tag on the protein using a nickel affinity surface, etc.
In the Dynamic response mode, the k
ON
obtained was about 4.47 ± 7.31 X 10
5
M
-1
s
-1
which is
of the same order as the assay orientation in which the RNA was on the surface. The kOFF was
recorded as 1.38 ± 0.30 M
-1
s
-1
however it was much higher than that of the first assay
78

orientation. The experiment was only done once and will need to be repeated to verify these
results. In principle, the off rate should not be dependent on the assay. However, coupling the
protein to the surface by its C-terminus may affect its activity or conformation, because we
expect RRM2 and 3 to rearrange upon binding. If the tethering orientation prevents the
molecules from achieving their stable bound conformation, this would result in a fast off rate.
The fluorescence proximity sensing mode data shows the opposite effect on fluorescence than
what when the RNA is attached to the surface. In this case, fluorescence appears to increase upon
RNA binding. This is not necessarily surprising since the response of the fluorophore depends on
its environment in the bound and unbound states. One of the strengths of the instrument is its
ability to detect effects on the fluorophore, whatever they may be. Even the binding of small
molecules can cause changes in fluorescence, allowing such interactions to be measured in this
instrument.
The sizing data during the stopped flow measurement with the protein bound to the surface
provided a hydrodynamic change read out which was similar to that seen in the other
orientation. The experiment will need to be repeated to confirm this observation. The sizing
data appears consistent for the complex with all the three domains, hinge and RNA is about
3.5 to 5nm from the three experiments. It should be noted that these measurements assume
the lollipop model, but the ELAVL4-bound RNA is probably more like an open-faced
subway sandwich, in which the domains lie side by side as an RNA platter. The only
structural data available for the ELAVL4/RNA complex is RRM1+2 complexed with c-fos
79

and tumor necrosis factor AU-rich RNA (Wang et al., 2001). Due to the missing hinge and
RRM3, that crystal structure is of limited use to estimate to size of the complex.
4.5 Advantages and limitations of switchSENSE technology compared to other
methods.

In the switchSENSE setup, the surface is fully regenerated before each association. This allows
independent measurements that are not affected by the surface being damaged / contaminated
with the previous run as in SPR.
The possibility to analyze interactions in multiple different ways gives flexibility and allow the
dynamics of the interaction to be examined in different experimental setups. As we have seen,
results obtained in different orientations may differ, and in our case, this will require further
examination.
The ability to measure using dynamics or fluorescence is a strength, and it is very interesting
to see that the kinetic data differs when using the switching speed versus the fluorescence
readout, suggesting conformational change. The same was suggested by SPR data (Park et al.
2001). It will be important to do the sizing experiment with the collection of ELAVL4
mutants lacking one or more domains.
Although there are a lot of advantages, there also are some disadvantages to switchSENSE
technology. The biggest drawback of the repeated complete surface regeneration is that the
experiments take longer, as each binding interaction requires reannealing. In addition, this
80

process consumes more of the ligand. However, since the surface density is low, the amount
of ligand expended is not that large. The recommended buffer is PE40 which has a sub-
physiological salt concentration and may yield data that does not fully reflect the natural
situation. Higher salt can be used, but it is better to use the longer nanolevers, and they are
more expensive.
The major differences between the switchSENSE and SPR are listed in the appendix.

4.6 Future directions.

In this study we focused on mainly studying how the switchSENSE biosensor works by
studying the interaction between WT ELAVL4 protein and 15AU3 RNA. Due to time
limitations, the different ELAVL4 domain mutants were not analyzed. Thus, much exciting
work remains to be done. In addition, the experiments with the WT protein should be
repeated and expanded upon.
The first assay orientation with RNA on the surface and protein as the analyte was studied
with different concentrations. It will be important to use higher salt concentration buffers
(150 mM NaCl) to compare to previously obtained SPR data. In subsequent experiments,
higher protein concentrations should be injected such that the fluoresce data reaches a
saturation during association, which means there are no further changes in the local
environment, confirming to formation of a stable complex. Performing the sizing analysis
right after equilibration will allow a better estimate of the complex size in the bound state.
81

The inverse assay orientation experiment was run with only one single concentration of RNA
and needs to be fully excited. Further experiments will be needed to clarify whether the
dissociation curve in the experimental run is actually the analyte RNA coming off of the
surface or both the RNA and ELAVL4, in other words to test the robustness of NTA/ELAVL4
surface.

switchSENSE provides a variety of ways to immobilize the ligand protein. One of the ways is
biotin streptavidin coupling. Since the protein has a biotinylatable tag, it would be interesting
to also study the assay orientation with streptavidin kit for 48mers. It is like using NTA kit
except that the binding of biotin streptavidin is much stronger.
Another alternative would be to amine couple the ELAVL4 to the surface. This can be done
using the AKTA start by GE Healthcare which is basically a size exclusion chromatography
that links the amine group of the protein to the cNLB nanolever, allowing it to be coupled to
the surface.
REFERENCES

Antar LN, Dictenberg JB, Plociniak M, Afroz R, Bassell GJ. Localization of FMRP ‐ associated
mRNA granules and requirement of microtubules for activity ‐ dependent trafficking in
hippocampal neurons. Genes, Brain and Behavior. 2005 Aug;4(6):350-9.
Berto S, Usui N, Konopka G, Fogel BL. ELAVL2-regulated transcriptional and splicing
networks in human neurons link neurodevelopment and autism. Human molecular genetics.
2016 Jun 3;25(12):2451-64.
82

Bronicki LM, Jasmin BJ. Emerging complexity of the HuD/ELAVl4 gene; implications for
neuronal development, function, and dysfunction. Rna. 2013 Aug 1;19(8):1019-37.
Bryant CD, Yazdani N. RNA ‐ binding proteins, neural development and the addictions.
Genes, Brain and Behavior. 2016 Jan;15(1):169-86.
Castello A, Fischer B, Eichelbaum K, Horos R, Beckmann BM, Strein C, Davey NE,
Humphreys DT, Preiss T, Steinmetz LM, Krijgsveld J. Insights into RNA biology from an
atlas of mammalian mRNA-binding proteins. Cell. 2012 Jun 8;149(6):1393-406.
Chung S, Eckrich M, Perrone-Bizzozero N, Kohn DT, Furneaux H. The Elav-like proteins
bind to a conserved regulatory element in the 3 ′-untranslated region of GAP-43 mRNA.
Journal of Biological Chemistry. 1997 Mar 7;272(10):6593-8.
Cléry A, Sohier TJ, Welte T, Langer A, Allain FH. switchSENSE: a new technology to study
protein-RNA interactions. Methods. 2017 Apr 15;118:137-45.
Dalmau J, Graus F, Rosenblum MK, Posner JB. Anti-Hu--associated paraneoplastic
encephalomyelitis/sensory neuronopathy. A clinical study of 71 patients. Medicine. 1992
Mar;71(2):59-72.
Good PJ. A conserved family of elav-like genes in vertebrates. Proceedings of the National
Academy of Sciences. 1995 May 9;92(10):4557-61.
Hinman MN, Lou H. Diverse molecular functions of Hu proteins. Cellular and molecular life
sciences. 2008 Oct 1;65(20):3168.
Hogan DJ, Riordan DP, Gerber AP, Herschlag D, Brown PO. Diverse RNA-binding proteins
interact with functionally related sets of RNAs, suggesting an extensive regulatory system.
PLoS biology. 2008 Oct 28;6(10):e255.
Kapeli K, Martinez FJ, Yeo GW. Genetic mutations in RNA-binding proteins and their roles
in ALS. Human genetics. 2017 Sep 1;136(9):1193-214.
Kim W, Kyung Lee E. Post-transcriptional regulation in metabolic diseases. RNA biology.
2012 Jun 1;9(6):772-80.
Kissil JL, Wilker EW, Johnson KC, Eckman MS, Yaffe MB, Jacks T. Merlin, the product of
the Nf2 tumor suppressor gene, is an inhibitor of the p21-activated kinase, Pak1. Molecular
cell. 2003 Oct 1;12(4):841-9.
Koushika SP, Lisbin MJ, White K. ELAV, a Drosophila neuron-specific protein, mediates the
generation of an alternatively spliced neural protein isoform. Current Biology. 1996 Dec
1;6(12):1634-41.
83

Lukong KE, Chang KW, Khandjian EW, Richard S. RNA-binding proteins in human genetic
disease. Trends in Genetics. 2008 Aug 1;24(8):416-25.
Lunde BM, Moore C, Varani G. RNA-binding proteins: modular design for efficient function.
Nature reviews Molecular cell biology. 2007 Jun;8(6):479.
Majumder P, Chu JF, Chatterjee B, Swamy KB, Shen CK. Co-regulation of mRNA translation
by TDP-43 and Fragile X Syndrome protein FMRP. Acta neuropathologica. 2016 Nov
1;132(5):721-38.
Manning KS, Cooper TA. The roles of RNA processing in translating genotype to phenotype.
Nature Reviews Molecular Cell Biology. 2017 Feb;18(2):102.
Marchese D, de Groot NS, Lorenzo Gotor N, Livi CM, Tartaglia GG. Advances in the
characterization of RNA ‐ binding proteins. Wiley Interdisciplinary Reviews: RNA. 2016
Nov;7(6):793-810.
Mobarak CD, Anderson KD, Morin M, Beckel-Mitchener A, Rogers SL, Furneaux H, King
P, Perrone-Bizzozero NI. The RNA-binding protein HuD is required for GAP-43 mRNA
stability, GAP-43 gene expression, and PKC-dependent neurite outgrowth in PC12 cells.
Molecular biology of the cell. 2000 Sep 1;11(9):3191-203.
Nagai K, Oubridge C, Ito N, Avis J, Evans P. The RNP domain: a sequence-specific RNA-
binding domain involved in processing and transport of RNA. Trends in biochemical sciences.
1995 Jun 1;20(6):235-40.
Nussbacher JK, Batra R, Lagier-Tourenne C, Yeo GW. RNA-binding proteins in
neurodegeneration: Seq and you shall receive. Trends in neurosciences. 2015 Apr
1;38(4):226-36.
Park S, Myszka DG, Yu M, Littler SJ, Laird-Offringa IA. HuD RNA recognition motifs play
distinct roles in the formation of a stable complex with AU-rich RNA. Mol Cell Biol. 2000
Jul;20(13):4765-72. Erratum in: Mol Cell Biol. 2004 Aug;24(15):6888.
Park-Lee S, Kim S, Laird-Offringa IA. Characterization of the interaction between neuronal
RNA-binding protein HuD and AU-rich RNA. Journal of Biological Chemistry. 2003 Oct
10;278(41):39801-8.
Shazman S, Mandel-Gutfreund Y. Classifying RNA-binding proteins based on electrostatic
properties. PLoS computational biology. 2008 Aug 8;4(8):e1000146.
Stefl R, Skrisovska L, Allain FH. RNA sequence ‐ and shape ‐ dependent recognition by
proteins in the ribonucleoprotein particle. EMBO reports. 2005 Jan 1;6(1):33-8.
84

Vassar R, Bennett BD, Babu-Khan S, Kahn S, Mendiaz EA, Denis P, Teplow DB, Ross S,
Amarante P, Loeloff R, Luo Y. β-Secretase cleavage of Alzheimer's amyloid precursor protein
by the transmembrane aspartic protease BACE. science. 1999 Oct 22;286(5440):735-41.
Zhou HL, Luo G, Wise JA, Lou H. Regulation of alternative splicing by local histone
modifications: potential roles for RNA-guided mechanisms. Nucleic acids research. 2013 Sep
28;42(2):701-13.

85

APPENDIX
86

87

Abstract (if available)

Abstract Characterization of RNA/protein interactions is key to identifying the cellular mechanisms involved in expression of genes. Traditional methods like gel shifts and isothermal titration calorimetry (ITC) provide quantitative binding data of intermolecular interactions at equilibrium conditions and surface plasmon resonance (SPR) yields kinetic data, but none of these methods provide information on molecular shape changes during binding. The recently developed switchSENSE technology, based on electro-switchable DNA nanolevers, provides not only high sensitivity kinetics but also information on size, shape and conformational changes occurring during an interaction, yielding a new depth of understanding of biomolecular interactions in real time. Here we use switchSENSE technology to study the interaction between RNA-binding protein HuD (ELAVL4) and RNA. Neuronal ELAVL proteins are involved in neuron-specific RNA processing and neural development. HuD consists of two N- terminal RNA-binding domains, a hinge region and a C-terminal RNA-binding domain. The interaction between HuD and prototype target RNAs of the sequence UU(AUUU)nAUU has been studied using equilibrium methods and SPR. The data suggested that HuD might undergo a rearrangement upon RNA binding. We hypothesize that the hinge region between RNA-binding domains 2 and 3 plays a key role in this conformational change. If true, this might provide a mechanism to regulate the protein/RNA interaction. The potential of switchSENSE technology to provide structural as well as kinetic data under equilibrium as well as dynamic conditions makes this an ideal tool for our investigation. A series of ELAVL4 mutants were generated to investigate the roles of different domains in the interaction with AU-rich RNA and to test whether and how the tertiary structure of the protein changes while binding to the RNA. In this study we focused on mainly studying how the switchSENSE biosensor works by studying the interaction between WT ELAVL4 protein and 15AU3 RNA. Due to time limitations, the different ELAVL4 domain mutants were not analyzed. The experiments conducted on the switchSENSE will provide further mechanistic insight into the dynamic nature of RNA/protein interactions, and how these might be regulated.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Developing a method for measuring kinetic rates of RNA/protein interaction using switchSENSE® technology on the DRX2

PDF

Development of immunotherapy for small cell lung cancer using novel modified antigens

PDF

Development of immunotherapy for small cell lung cancer using iso-aspartylated antigen

PDF

Do ZFX and ZNF711 regulate the same genes in HEK293T cells?

PDF

Interaction of monoclonal antibodies MW1 and PHP1 with huntingtin exon1 protein

PDF

Defining the functional region of LINC00261 in lung adenocarcinoma

PDF

PARIS with enriched target RNAs

PDF

Evaluating the impact of long non-coding RNAs on tumor mutational burden in cancer

PDF

Testing structure-induced RNA hydrophobicity

PDF

Review of long noncoding RNAs and chromosome structure

PDF

Evaluation of preservatives in blood collection tubes for cell-free RNA transcriptional profiles in human plasma

PDF

Positive regulation of RNA polymerase III-mediated transcription of tRNA genes by the Mediator kinase submodule

PDF

Modeling lung adenocarcinoma progression in vitro using immortalized human alveolar epithelial cells

PDF

Using epigenetic toggle switches to repress tumor-promoting gene expression

PDF

A functional genomic approach based on shRNA-mediated gene silencing to delineate the role of NF-κB and cell death proteins in the survival and proliferation of KSHV associated primary effusion l...

PDF

Impacts of post-translational modifications on interactions between G9a and its N-terminus binding partners

PDF

Detecting joint interactions between sets of variables in the context of studies with a dichotomous phenotype, with applications to asthma susceptibility involving epigenetics and epistasis

PDF

Computational analysis of genome architecture

PDF

Understanding the role of APP and DYRK1A in human brain pericytes

PDF

Modulation of KIF3A function by herpes simplex virus-1 US3 kinase in viral immune evasion

Asset Metadata

Creator Muralidhar, Anusha (author)

Core Title Analysis of ELAVL4 RNA interaction using a switchSENSE biosensor

School Keck School of Medicine

Degree Master of Science

Degree Program Biochemistry and Molecular Medicine

Publication Date 10/12/2018

Defense Date 08/23/2018

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag DRX2,dynamic biosensors,ELAVL4,intermolecular interactions,OAI-PMH Harvest,RNA,switchSENSE

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Offringa, Ite (committee chair)

Creator Email anusha.murali12@gmail.com,anushamurali1205@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-88965

Unique identifier UC11676753

Identifier etd-Muralidhar-6819.pdf (filename),usctheses-c89-88965 (legacy record id)

Legacy Identifier etd-Muralidhar-6819.pdf

Dmrecord 88965

Document Type Thesis

Format application/pdf (imt)

Rights Muralidhar, Anusha

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA