University of Southern California Dissertations and Theses

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

(USC Thesis Other)

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

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content 1

Developing a method for measuring kinetic rates
of RNA/protein interaction using switchSENSE
®

technology on the DRX2

By

Arthur Sefiani

A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(Biochemistry and Molecular Medicine)
August 2019

2

Table of Contents

Acknowledgements 5
Chapter 1: Introduction 6
1.1 U1A and U1hpII 6
1.2 Introduction to Surface Plasmon Resonance Technology 8
1.3 Previous Kinetic Analyses Between U1A and U1hpII 9
1.4 Introduction to switchSENSE
®
Technology 11
1.5 Current Use of switchSENSE
®
Technology 14
Chapter 2: Optimizing U1A RRM1 Protein Production 17
2.1 Introduction 17
2.2 Cloning 18
2.3 Cell Type 19
2.4 Wash Protocol 20
2.5 Ni-NTA Agarose Volume 22
2.6 Lysing Method 24
2.7 Biotinylation 26
2.8 Buffer Exchange 29
2.9 Final Protocol 30
2.9.1 Reagents 30
2.9.2 Transformation 30
2.9.3 Protein Induction and Lysing 31
2.9.4 Protein Purification 31
Chapter 3: Protein Quantification 33
3.1 Introduction 33
3

3.2 Bradford Assay 33
3.3 BCA Assay 34
3.4 SDS-PAGE Comparison 36
3.5 Discussion 38
Chapter 4: RNA Production 39
4.1 Introduction 39
4.2 In Vitro Transcription 39
4.2.1 Transcribing from Linearized Plasmids 40
4.2.2 Transcribing from Short Oligos 42
4.2.3 Purification 43
4.3 RNA from IDT 43
4.4 Quantitation 45
4.5 Discussion 46
Chapter 5: Kinetic Analyses 47
5.1 Introduction 47
5.1.1 Preparation and Experimental Overview 47
5.1.2 Analyte Dilutions 48
5.1.3 Analysis Modes 49
5.2 Creation of the Ligand Surface 50
5.2.1 U1hpII RNA as the Ligand 50
5.2.2 Consistency of U1hpII Hybridization 52
5.2.3 U1A as the Ligand using Amine Coupling 54
5.2.4 Biotin/Streptavidin Coupling of U1A onto the Surface 56
5.3 Measurement of Association Rates 58
5.3.1 U1hpII Attached to the Surface 59
4

5.4 Measurement of Dissociation Rates 60
5.4.1 U1hpII Attached to the Surface 60
5.5 Discussion 61
Chapter 6: Future Directions 64
6.1 Introduction 64
6.2 RNAstructure 64
6.3 Protein Linkers 65
6.4 RNA Pull-down Assay 66
References 67

5

Acknowledgements

I would like to thank my committee members Dr. Ian S. Haworth and Dr. Ansgar B. Siemer
for their support. I would like to thank Dr. Ite A. Offringa for being an amazing mentor and helping
every step of the way. I would like to thank all Offringa Lab and Marconett Lab members for their
constant support and assistance through the entirety of this project and for sharing equipment. I
would like to thank Dr. Judd Rice and Monica Pan for their advice and support throughout the
Master's degree. I would also like to thank Dynamic Biosensors’ staff members Thomas Weber
and Joanna Deek for their support and technical assistance. Furthermore, I would like to thank my
family for their lifetime of selflessness. I appreciate all they have sacrificed to help me accomplish
my goals.

6

Chapter 1: Introduction
1.1 U1A and U1hpII
U1 small nuclear ribonucleoprotein A (U1A) is part of the U1 small nuclear
ribonucleoprotein (snRNP) and is found in the nucleus. Once U1 snRNP recognizes and binds to
the 5’ pre-mRNA splice-site, U2 snRNP will bind followed by the
binding of U4/U5/U6 snRNP (UniProtKB - P09012). These
snRNPs are protein-RNA complexes found only in the nucleus of
eukaryotic cells that combine to form the spliceosome. The
spliceosome is needed for most proteins inside a eukaryotic cell
and therefore an integral part of gene regulation. The U7 snRNP
is unique from the other snRNPs because of its involvement in
histone RNA processing. U7 snRNP levels are significantly
increased during the S-phase and theorized to be involved in the
replication of DNA by regulating histone gene expression (Keall
et al., 2007).
U1A is a 250 amino acid long protein with 2 domains which are connected by a hinge. The
first domain is the RNA recognition motif 1 (RRM1) which includes amino acids 18 to 97. This
domain is responsible for interacting with the AUUGCAC RNA sequence (Law et al., 2006). The
second domain is RNA recognition motif 2 (RRM2) which includes amino acids 177 to 250
(UniProtKB - P09012). The function of this domain is currently unknown but it may bind to other
proteins to help form a spliceosome. The hinge is important for connecting RRM1 to RRM2 but
is not needed for RRM1 to interact with its target RNA U1hpII (Law et al., 2006). Furthermore,
the hinge dimerizes when 2 U1A proteins bind to the polyadenylation inhibition element.
Figure 1. Crystal Structure of Human U1
snRNA. The 4 stem-loop structures are
SL1 (red), SL2 or U1hpII (blue), SL3 (green)
and SL4 (cyan). The U1A protein binds to
7 key residues in SL2. (Krummel et al.,
2009).
7

The U1 hairpin II RNA (U1hpII) is a 26 nucleotide long section of RNA from the stem-
loop domain 2 (SL2) from the U1 snRNA (Piekna-Przybylska et al., 2007). The U1 snRNP is
conserved in mammals and is formed through the attachment of U1A, U1C, U170K and 7 Sm
proteins to the U1 snRNA (Krummel et al., 2009). It is unknown what specific role U1A plays,
but it is believed that once U1A binds to the U1 hairpin II stem-loop, it helps form the U1 snRNP
complex.
Due to the importance of U1 snRNP, each part of the complex is highly regulated to ensure
proper cell function. For example, the U1A polyadenylation inhibition element (PIE) autoregulates
the concentration of U1A protein in the nucleus. PIE is located
in the U1A mRNA 3' untranslated region and simultaneously
binds to 2 U1A proteins which subsequently inhibits poly(A)
polymerase (PAP) from adding the 3' polyadenine tail to U1A
(Varani et al., 2000).
Over expression of U1 snRNA can occur in disease
states and has shown to have detrimental effects on the human
brain. Increased U1 snRNA concentration can cause
autophagic-lysosomal dysfunction which augments the
development of Alzheimer’s disease. Furthermore, over expression of U1 snRNA increases
reactive oxygen species by decreasing cell superoxide dismutase levels and has shown to increase
malondialdehyde levels in both humans and mice (Cheng et al., 2018). Also, the abnormal
accumulation of U1 snRNAs in the nuclei of motor neurons has been greatly associated with the
development of amyotrophic lateral sclerosis (Tsuiji et al., 2013).

Figure 2. PIE RNA interacting with U1A.
The RRM1 of 2 U1A proteins bind to the PIE
RNA which then will bind to PAP to inhibit
its function. (Gunnewiek et al., 2000).
8

1.2 Introduction to Surface Plasmon Resonance Technology
Surface Plasmon Resonance (SPR) technology is essentially used to detect infinitesimal
changes in mass. A thin gold film is produced with a prism over one surface and a ligand of choice
attached on the opposite surface. Analyte is injected over the ligand attached surface to first bind
then dissociate from the ligand. Binding of the analyte to the ligand increases the overall mass of
the gold surface and dissociation of the analyte decreases it. From the speed at which the mass of
the gold surface changes, assuming a fixed concentration of analyte is being used, one can calculate
the association (ka or kON) and dissociation
rates (kd or kOFF).
During the binding of the analyte to
the ligand attached to the gold surface, light is
shot through the prism on the opposite side
and is reflected off the gold surface onto a
detector. At a certain angle, called the
resonance angle, the electrons in the gold film
absorb the light and resonate. The resonance is caused by the oscillation of free electrons via the
Kretschmann configuration (Biosensing Instrument). This absorption of light at the resonance
angle is seen as a dark band or dip in intensity when comparing to all angles of reflection. The size
and location at which this intensity dip occurs can then be used to make inferences about what is
happening on the gold surface (BiosensingUSA). The resonance angle is correlated with the mass
of the gold surface. Machines that utilize SPR technology record the resonance angle over time.
The change and rate of change of this resonance angle, known as the angular shift, can then be
used to calculate kinetic parameters such as the dissociation constant (KD).
Figure 3. BIAcore 2000 utilizing SPR technology. The analyte (red)
is flown through a channel to bind to the ligand (blue) which is
attached to a gold surface. Light is then shot into a prism above
the gold surface and the angle of resonance (black arrow) is
recorded by the detector. (GE Healthcare Life Sciences).
9

SPR technology does have a few drawbacks. Since the technology is mass based, it is very
difficult to analyze interactions between small molecules because of the limited sensitivity of the
detector to sense minute changes in resonance angle. Also, since the technology is mass based,
there is no built in mechanism to detect conformational changes or to do sizing measurements.
1.3 Previous Kinetic Analyses Between U1A and U1hpII
The Offringa Lab used the BIAcore 2000, which utilizes SPR technology, to measure the
kinetic parameters of the interaction between U1A and U1hpII. The gold surface opposite to the
prism was coated with streptavidin to attach biotinylated molecules. A 2 nM biotinylated U1hpII
solution dissolved in Tris running buffer was injected over the gold surface at a flow rate of 10 μl
min
−1
at 20°C to bind to the streptavidin coated surface and serve as the ligand. U1A, the analyte,
was diluted in running buffer at various known concentrations and injected over the surface at a
flow rate of 50 μl min
−1
at 20°C. The data was processed using Scrubber and analyzed using
CLAMP (Law et al., 2006).
Since only the first RNA recognition motif of U1A is responsible for interacting with
U1hpII, only the first 101 amino acids of U1A are used in Dr. Offringa’s papers. The KD of wild-
type U1A is approximately 30 pM. To study the mechanism responsible for U1A’s attraction and
stable binding to U1hpII, they created many different constructs of U1A, replacing key amino
acids to determine the mechanism of attraction and stabilization.
When the positively charged side chain of Lys20 or Lys22 or Lys23 is neutralized by
replacing it with the side chain of Gln, there is less than an 8-fold increase in KD. When
Lys20,22,23 are all replaced with Gln, the KD increases by 11000. This effect cannot be replicated
when replacing Lys20,22,23 with Arg, which also has a positively charged side chain. It is
10

hypothesized that these positively charged side chains interact with the stem of U1hpII to attract
and stabilize the interaction (Law et al., 2006).
When Tyr13 is replaced with an amino acid without an
aromatic side chain, the association rate decreases by at least 14-
fold and the dissociation rate increases by at least 3200-fold. The
aromatic side chain of Tyr is believed to have a minor role in
association but a major role in complex stability through aromatic
stacking between itself and C5. A very similar pattern is seen
when replacing Phe56 with an amino acid without an aromatic
side chain which disrupts its interaction with A6. Although Gln54
does not have a charged or aromatic side
chain, similar effects are seen when replacing it with Asn, an amino acid
with the same side chain minus a methyl group. It is believed that the larger
side chain of Gln allows it to form a hydrogen bond with Tyr13 and G4,
which allows it to reposition Lys50 and Arg52, moving the β2–β3 loop
through the U1hpII hairpin loop, further stabilizing the interaction (Law et
al., 2005).
Together, these findings suggest that the positively charged side
chains of Lys and Arg lure in U1hpII. Subsequently, additional interaction,
including the aromatic stacking from the side chains of Phe and Tyr and
hydrogen bonds from Gln cause a rapid induced fit event to take place which stabilizes the
interaction and locks U1A onto U1hpII. This is the “lure and lock” theory (Katsamba et al., 2005).
Figure 5. Sequence of
U1hpII used for kinetic
analysis. The nucleotides in
the red box are the key loop
residues crucial for U1A
binding. The numbers are for
referencing purposes. (Law
et al., 2006).
Figure 4. The role of electrostatic forces
in luring U1hpII towards U1A. The
positively charged side chains of Lys20
and Lys22 (blue) of U1A (cyan) are
interacting with the phosphate backbone
(red) of U1hpII (purple). (Katsamba et al.,
2001).
11

The key loop residues shown inside the red box in figure 5 are essential for the binding of
U1A. Replacing U8 with C or adding extra Cs after the key loop residues causes no significant
change in KD. Also, lengthening or shortening the sequence of
the double-stranded stem seems to have no effect on binding
either. This makes sense, since RNA-binding proteins cannot
recognize the sequence of double-stranded RNA due to a very
narrow major groove. Thus, the stem can only aid binding
through electrostatic interactions between the negatively
charged phosphate backbone of U1hpII and the positively
charged side chains of U1A. Changing U1hpII into a single
stranded RNA increases the dissociation rate by 590-fold and
detaching A1 from C-1 increases the dissociation rate by 2900-
fold. This suggests that the loop needs to be constrained for optimal binding (Law et al., 2006).
1.4 Introduction to switchSENSE
®
Technology
Dynamic Biosensors’ switchSENSE technology utilizes fluorescence spectroscopy to
measure the kinetic parameters of an interaction. First, a gold film is produced over a
microelectrode. Then the 5’ of short single stranded DNA oligos are covalently attached onto this
gold surface with rhodium based fluorophores attached onto the 3’ end. These fluorophore-labeled
DNA oligos are usually 48-96 nucleotides long and are called nanolevers. The gold surface is
excellent at quenching these fluorophores and the absolute intensity and rate of change in intensity
of these fluorophores is measured by a detector, which is then quantitated to calculate kinetic
parameters. Each biochip contains 4 flow channels and each flow channel contains 6 gold covered
Figure 6. Cocrytsal structure of the
interaction between U1A and U1hpII.
Arg52 (purple) stabilizes the interaction
between U1A (black) and U1hpII (gray) by
hydrogen bonding to A1 (yellow) and G11
(green). (Law et al., 2006).
12

microelectrodes in series. There are many different biochips to choose from depending on the
ligand, number of different nanolever sequences, nanolever length and nanolever density.
There are several options that can be used to
measure the kinetic parameters. The first is Static
Mode, which sends a constant negative voltage across
the gold surface to push the negatively charged
nanolever backbone away from the surface. In this
mode, the nanolevers are always pushed away from the
gold surface, therefore can only be quenched by bound
analytes. Once the ligand is bound to the nanolevers through annealing of a complementary DNA
strand, the analyte is injected over the surface and the fluorescence intensity is recorded over time
as the analyte binds and dissociates from the ligand. In this mode, the data usually has a higher
signal to noise ratio and higher sensitivity to changes in fluorescence intensity. For each
biomolecular interaction, it has to be empirically determined whether and how binding of the
analyte will change the fluorescence. Depending of the position and properties of the analyte, the
fluorescence intensity could increase or decrease with binding.
The second option is Dynamic Mode, which sends an alternating current across the gold
surface at a frequency of up to 10kHz. In Dynamic Mode, the nanolevers are pushing away and
pulling towards the gold surface at a known rate, known as the switching speed, all while the ligand
is attached to the nanolevers and the analyte is being injected over the gold surface to interact with
the ligand. As the interaction occurs, the switching speed slows down. The actual speed itself is
dependent on many factors such as the size and shape of the ligand and analyte. This change in
speed is measured through changes in fluorescence intensity, which is being recorded over time
Figure 7. Different nanolever density options on
gold covered microelectrodes. The average distance
between nanolevers (orange) compared to the
number of nanolevers per μm
2
(blue) compared to
the biochip label (black). (Technology, Dynamic
Biosensors).
13

just like in Static Mode. The difference is that data is extrapolated differently. Since the gold
surface quenches the fluorophores, we know the position of the nanolevers at any given time which
can be used to calculate the switching speed and the kinetic parameters.
The change in nanolevers’ switching speed results from hydrodynamic friction. Since the
nanolevers are moving in an aqueous environment, they must push water molecules to the side to
reach the gold surface which increases in difficulty by the drag created by bound ligand and
analyte. The hydrodynamic friction created by the bound analyte can also be used to calculate the
diameter of an analyte to the nearest tenth of a nanometer assuming a globular shape. Besides
extrapolating more data, another advantage of running experiments in Dynamic Mode is that when
the bound analyte undergoes a conformational change after binding to the ligand, a change in
hydrodynamic friction will be measured and the conformational change will be reported (FAQs,
Dynamic Biosensors).
Figure 8. Negatively charged gold covered microelectrode pushing nanolevers away from the surface at different rates
during Dynamic Mode. The illustration shows how the switching speed differs between nanolevers bound by different
size analytes (in an actual binding experiment, analyte size should be uniform). The nanolever (silver) on the right only
has a ligand (blue) bound to it, therefore is able to be quickly pushed away from the gold surface. The nanolever on the
left has a very massive analytye (red) bound to its ligand, increasing hydrodynamic friction, slowing the speed at which it
is pushed away from the gold surface. The closer the fluorophore (green) is to the gold surface, the more quenched it is.
The nanolever in the middle illustrates that the switching speed and quenching of the fluorophore are both on a spectrum.
(Nanosystems Initiative Munich, Dynamic Biosensors GmbH).
14

The Offringa Lab operates the Norris Comprehensive Cancer Center's switchSENSE
instrument, a DRX2, which includes a dual fluorescence detection system. With this model, one
can utilize the 2 separate nanolevers. One nanolever usually has a red fluorophore attached and the
other a green fluorophore with a different DNA sequence. Having 2 nanolevers with different
DNA sequences and fluorophores allows one to be used as a reference nanolever or to measure 2
separate interactions simultaneously.
There are 3 programs required to operate the DRX2. The first is switchBUILD which
allows the user to create experiments and plan subsequent operations for the machine to conduct.
Then the operator must use switchCONTROL, which translates said experiments into executable
code to run the experiments on the DRX2. Finally, the user can use switchANALYSIS to analyze
the data. This software is capable of conducting interaction analysis such as kinetics, avidity and
inhibition effects. It can also conduct biophysical analysis to determine the melting point, extract
thermodynamic energies, measure molecule size and quantify conformation changes (Software,
Dynamic Biosensors).
1.5 Current Use of switchSENSE
®
Technology
There are many publications based on using switchSENSE technology to study
biomolecular interactions. Some study proteins interacting with their RNA targets. One example
is studying the N-terminal extension close to the YTH domain on Mmi1 and its function in binding
to and degrading mRNA by utilizing fluorescently-labeled RNA oligonucleotides (Stowell et al.,
2018). Another example is showing how Pab1 interacts with Ccr4-Not to stimulate deadenylation
and differentiates the roles of Ccr4-Not by using an RNA strand with the sequence of interest
covalently attached to a DNA sequence that is complementary to the nanolever (Webster et al.,
2018). There are also many studies involving tRNA such as the structural effects of Charcot-Marie-
15

Tooth disease caused mutations in tyrosyl-tRNA synthetase which utilizes thiol-coupling kits to
attach the protein of interest onto DNA oligos to hybridize onto nanolevers (Blocquel et al., 2017).
Each study provides insight into possible mechanisms for studying a protein’s interaction with
RNA. A different mechanism will be discussed in this dissertation.
A former MS student of Dr. Offringa, Anusha Muralidhar, used the DRX2 to study the
interaction between ELAVL4 and its RNA target. It has previously been shown that each RNA
recognition motif of ELAVL4 plays a distinct role in its interaction with the RNA target, but this
experiment was conducted on the BIAcore 2000 (Park et al., 2000). Anusha was successfully able
to measure the kinetic parameters of this interaction on the DRX2 by using an RNA strand with
the sequence of interest covalently attached to a DNA sequence that is complementary to the
nanolever. This will allow a stable platform for the RNA sequence of interest to serve as the ligand.
All of Dr. Offringa’s experiments, both on the BIAcore 2000 and DRX2, use RNA as the ligand.
Dr. Allain’s laboratory was able to successfully hybridize RNA onto nanolevers without
the need to covalently attach it to DNA. The RNA was made such that the 3’ end of the RNA
strand was complementary to the nanolever and the 5’ end contained the sequence of interest
(Krepl et al., 2017). However, very limited experimental details were provided. Furthermore, our
experiment requires the use of a hairpin target rather than a short RNA oligo which further
complicates matters due to the secondary structure of the RNA.
While the above cited manuscripts describe some approaches to study a protein’s
interaction with RNA, other options are available. In this dissertation, I will focus on 2 different
methods. The first method uses RNA as the ligand, expanding on the technique used by Dr.
Allain’s laboratory and optimizing the protocol to work effectively for interactions with very small
dissociation constants and hairpins. The second is conjugating protein onto the nanolevers and
16

using the protein as the ligand instead. The very small (30 pM) dissociation constant of the robust
interaction between U1A and U1hpII makes it a desirable contestant for developing protocols
around because the globular RRM1 domain is robust with a high affinity interaction and well-
studied by a variety of methods including SPR. Furthermore, since this interaction has been heavily
tested using SPR technology, we will have a precise reference point to compare our switchSENSE
data to. The goal is to develop protocols that produce consistent and accurate kinetic measurements
each and every time that can also be applied to other biomolecular interactions.

17

Chapter 2: Optimizing U1A RRM1 Protein Production
2.1 Introduction
In order to conduct kinetic measurements, we need ample amounts of well-quantitated U1A
protein. The sequence for U1A RRM1 (amino acids 1-101) is inserted into a pET3d plasmid to
express recombinant protein. The pET3d plasmid contains an N-terminal T7 promoter sequence
along with ampicillin resistance. I added a hexahistidine (His) tag at the C-terminus of U1A and a
biotinylatable tag at the N-terminus. The His tag is used for purification on nickel affinity beads,
a widely used method because it has many benefits
such as conserving protein activity, very low
immunogenicity and cost, and high versatility (Ni-
NTA Agarose, Cube Biotech).
The pET3d U1A plasmid was transformed
into ClearColi™ BL21(DE3) bought from Lucigen.
The ClearColi™ version of BL21(DE3) cells lack
outer membrane agonists for hTLR4/MD-2
activation which causes their endotoxic activity to be
drastically reduced (VWR ID:89428-536). Briefly,
the cells were then lysed and incubated with Ni-NTA agarose beads from Qiagen to separate the
His tagged U1A protein from the cell lysate. After washing the solution with a low imidazole
buffer several times, U1A was eluted by a high imidazole buffer. The imidazole is a competitive
agent, competing with His to bind to nitrilotriacetic acid (NTA) which elutes the His-tagged U1A.
J57, a pACYC184 plasmid with chloramphenicol resistance, is co-transformed with the
U1A plasmid into the ClearColi™ BL21(DE3) cells. The J57 plasmid was constructed by Dr.
Figure 9. The pET3d U1A plasmid. The T7 promoter (gray)
is upstream of U1A (left purple arrow). The plasmid also
includes the AmpR gene (right purple arrow) for ampicillin
resistance. The plasmid is 4812 bp in length (SnapGene
Viewer).
18

Offringa to include the E. coli BirA gene and E. coli Arg and Pro tRNAs. The BirA gene is a biotin
operon repressor and a biotin ligase (UniProtKB - P06709), which is used for the biotinylation of
U1A. The plasmid is also meant to increase the concentration of normally rare Arg and Pro tRNAs
inside the E. coli cells to increase protein production potential, especially since human proteins
tend to have higher Arg and Pro amino acid content than bacterial proteins.
2.2 Cloning
Dr. Offringa’s laboratory created 2 different U1A constructs in the pET3d plasmid
backbone; K680 and K833. K833 contains an N-terminal His tag followed by the wild-type U1A
RRM1 sequence. After the expression of K833 U1A construct, there were dual bands on the SDS-
PAGE gel right above and below the 11kDa protein marker. This led us to believe that partially
translated U1A was being eluted right along with the fully translated U1A because both peptides
will contain the N-terminal His tag, thus binding to the Ni-
NTA beads. The K680 construct contained a C-terminal His
tag but 2 mutated residues (Lys20,22). The C-terminal His
tag would only allow the elution of fully translated U1A
RRM1. K680 showed only a single band on the SDS-PAGE
gel right above the 11kDA protein marker, supporting our
hypothesis and therefore was used as the starting point for
future cloning experiments.
The K680 construct needed to be mutated back to carrying the wild-type U1A RRM1
sequence. The plasmid was double digested with MluI and SacI restriction enzymes and a double
stranded oligo with the wild-type sequence was inserted. I used SnapGene Viewer and Serial
Cloner to assist in designing cloning experiments. Furthermore, there were 26 extra amino acids
Figure 10. Final U1A sequence used on all
proceeding experiments. The first 19 amino
acids (red) is the biotinylatable tag at the N-
terminus followed by the first 101 amino acids
of wild-type U1A (black) followed by a protein
spacer (blue) and His tag (green).
19

at the C-terminus which were not part of U1A RRM1 that were cut out. Finally, a biotin tag needed
to be inserted for the biotinylation of U1A for future experiments. For the biotin tag, I used the 19
amino acid sequence previously used to biotinylate recombinant proteins in our laboratory (Park-
Lee et al., 2003). The final expected size of the U1A recombinant protein expressed using my
construct is 14.83kDa (calculated using Protein Molecular Weight by bioinformatics.org) and 129
amino acids long.
2.3 Cell Type
The BL21 (DE3) E. coli strain is ideal for protein expression for several reasons. First, it
is deficient in Lon protease which degrades foreign proteins (Gottesman et al., 1996). It is also
deficient in OmpT protease which degrades extracellular
proteins (Grodberg and Dunn et al., 1988). Therefore, there is
less chance of foreign protein being degraded intracellularly
and after cell lysing. The reason I use the DE3 strain is because
it has the T7 RNAP gene containing λDE3 prophage inserted
into the chromosome of the E. Coli (Rosano et al., 2014). This
is important because our plasmids contain T7 promoter sites
upstream of the recombinant protein of choice. Protein
expression is induced using a 1 mM final concentration of
Isopropyl β-D-1-thiogalactopyranoside (IPTG).
I tried using 2 different cell types from 3 sources.
Electrocompetent ClearColi™ BL21 (DE3) from Lucigen and chemically competent BL21 (DE3)
from both New England Labs and Invitrogen™. Although I didn’t need the ClearColi’s™ ability
to lower endotoxicity, I wanted to see its effect on the quantity of protein expressed. The
Figure 11. Comparing the effectiveness of 3
different BL21 (DE3) cells. 14% SDS-Page
gel of U1A produced using ClearColi™ BL21
(DE3) from Lucigen, BL21 (DE3) from
Invitrogen™, and BL21 (DE3) from New
England Labs (NEB).
20

electrocompetent BL21 cells were placed in CCMB80 transformation buffer (10 mM CH3CO2K,
10% glycerol, 80 mM CaCl2, 20 mM MnCl2, 10 mM MgCl2) to become chemically competent
(Cold Spring Harbor Protocols Rec:102707) and were transformed using the same protocol used
for the chemically competent BL21 cells. The transformation protocol will be listed at the end of
this chapter along with the final protein production protocol.
A 14% SDS-PAGE gel was created and 20µl solution (10µl of U1A, 10µl of loading dye)
was pipetted in each lane. All 3 lanes in figure 11 have a prominent band at the bottom of the
correct size (U1A) and a lot of contamination above it. I am interested in the BL21 cell that will
give us the highest U1A to contamination ratio. Lane Lucigen seems to have a significantly higher
U1A concentration than lane Invitrogen and NEB. Lane Lucigen and Invitrogen seem to have
similar amounts of contamination, both of which being less than lane NEB. Therefore, I conclude
that ClearColi Lucigen™ BL21 (DE3) from Lucigen is the best out of the 3 to use from here on
out.
2.4 Wash Protocol
As the supernatant of lysed E. coli solution incubates with the Ni-NTA beads, U1A and
other contaminants bind to the beads. Non-specific binding is a big problem for the His tag
purification system. Since Ni-NTA beads bind to any chain of His peptides, native proteins with a
high His concentration bind to the beads as well. Therefore, the imidazole concentration of the
buffer is gradually increased to elute the non-specific binders until mostly U1A is bound to the
beads. Once the beads are washed and most of the contaminants are eluted, a 250 mM imidazole
elution buffer is used to elute the U1A.
The negatively charged nitrogen on the side chain of His is able to form an ionic interaction
with Ni which forms an ionic interaction with NTA. Imidazole is a competitive binder to Ni-NTA
21

and therefore elutes His tagged proteins when imidazole concentrations are increased. Interestingly
enough, the side chain of His is an imidazole group; both theoretically should have very similar
dissociation constants.
Each wash cycle involves
incubating the U1A bound Ni-NTA
beads inside a 10 mL solution of a
known imidazole concentration for
several minutes as the solution is
rotating, after which the beads are
centrifuged and the protein eluted. When
visualizing the elution of proteins after
each wash cycle, it is clear that most
contaminants are eluted after using a 10
mM imidazole buffer. A significant amount of contaminants are eluted after using a 30 mM
imidazole buffer as well with minimal elution of U1A. After washing with the 50 mM imidazole
buffer, there is a significant but small amount
of U1A being eluted with a relatively small
number of contaminants. After washing with
the 75 mM and 100 mM imidazole buffers,
there are no visible contaminants being
eluted; only U1A is being eluted at increasing
amounts in relation to the imidazole concentration.
Figure 12. Imaging the proteins eluted after each wash cycle. 14% SDS-
Page gel of the wash solution after the U1A bound Ni-NTA beads were
extracted out of the wash solution. Either a 10 mM imidazole buffer (1),
a 10 mM and 30 mM imidazole buffer (2), a 10 mM, 30 mM and 50 mM
imidazole buffer (3), a 10 mM, 30 mM, 50 mM and 75 mM imidazole
buffer (4), or a 10 mM, 30 mM, 50 mM, 75 mM and 100 mM imidazole
buffer (5) was used to wash the U1A bound Ni-NTA beads before the wash
buffer with the highest concentration of imidazole was loaded onto the
gel.
Table 1. Different wash protocols. I used 4 different wash
protocols to test the effectiveness of each protocol in producing
the highest U1A to contaminant ratio.
22

When comparing different wash protocols, it
is clear that protocol 4 has the lowest contamination
level but also the lowest U1A concentration. Protocol
1 has the highest contamination level but the U1A
concentration does not seem to be significantly
increased. Since we are interested in the highest U1A
to contamination ratio, it seems that protocol 4 would
be a better option than protocol 1. Protocol 2 and 3
seem to have a slightly higher U1A and contaminant
concentration than protocol 4, a difference not prominently visible. I do believe the U1A to
contaminant ratio of protocol 2 to be the highest because in figure 12, the 75 mM and 100 mM
imidazole concentration (used only in protocol 3 and 4) only eluted U1A. I will be using wash
protocol 2 from here on out.
2.5 Ni-NTA Agarose Volume
The Ni-NTA complex is an affinity chromatography matrix with nitrilotriacetic acid (NTA)
covalently bound to silica beads. The negatively charged oxygen and nitrogen groups of NTA form
an ionic interaction with Ni which also forms an ionic interaction with the negatively charged
nitrogen groups of the His side chain (Magdeldin et al., 2012). These silica beads have an average
diameter of 105µm and a binding capacity of 50µg/µL (Qiagen ID:30210).
It is important to know the right amount of Ni-NTA agarose beads to use for the expression
of a given recombinant protein to not only maximize the yield, but also minimize the non-specific
binding. In theory, the less beads used, the less His tagged protein will be purified from the cell
Figure 13. Comparing different wash protocols. 14%
SDS-Page gel of U1A elutions after using wash protocol
1 (1), wash protocol 2 (2), wash protocol 3 (3), or wash
protocol 4 (4) to wash the U1A bound Ni-NTA beads.
23

lysate. The more beads used, the more
interaction sites there will be to bind
nonspecifically to the contaminants. The
beads are washed with a 10 mM imidazole
buffer before incubating with the
supernatant of the cell lysate to decrease
non-specific binding by inhibiting free Ni
ions from binding to contaminants.
In figure 14, the lane with 25µL of beads has the least amount of U1A and contaminants
bound which is as expected. Surprisingly, the lane with 100µL of beads not only has the most U1A
bound, but slightly less contaminants than the 400µL and 1000µL lanes. It is clear that the best
option of the 4 volumes is either 25µL or 100µL. It is unknown
whether 25µL or 100µL of Ni-NTA agarose will lead to a higher
U1A to contaminant ratio. I chose to conducted further experiments
with 100µL of Ni-NTA agarose because there is a significantly
higher absolute yield of U1A which will be easier to quantitate and
store without denaturing.
To test whether the reason for the high contamination levels was a
batch effect of the Ni-NTA beads being used, a second bottle of
agarose with a different lot number was ordered from Qiagen. In the
14% SDS-PAGE gel, there seems to be no significant change in U1A concentration or
contamination levels between the old and new lots, therefore both lots are being used for future
experiments.
Figure 15. Comparing different
batches of Ni-NTA agarose beads.
14% SDS-Page gel of U1A eluted
using a new lot and an old lot of Ni-
NTA agarose.
Figure 14. Comparing the effectiveness of different volumes of Ni-
NTA agarose. 14% SDS-Page gel of U1A eluted using the volume of
Ni-NTA agarose written below each lane.
24

2.6 Lysing Method
The Ni-NTA beads are first
washed with the sonication buffer
before being mixed with the supernatant
of the cell lysate. The sonication buffer
contains imidazole which forms an
ionic bond with the Ni and lowers
nonspecific binding. The sonication buffer used for all previous experiments contained a final
imidazole concentration of 10 mM, but all experiments from here on out use Ni-NTA beads
washed in a final imidazole concentration of 20 mM. 20 mM imidazole is recommended when
dealing with proteins that bind well to the Ni-NTA
column (A handbook for high-level expression and
purification of 6xHis-tagged proteins,
QIAexpressionist). The imidazole concentration was
increased in an effort to increase the U1A to
contamination ratio by lowering nonspecific binding.
Since the sonication buffer already contains a 20 mM
imidazole concentration, there is no need to include a
10 mM imidazole wash; the wash protocol was changed
from a 10 mM, 30 mM and 50 mM imidazole wash to
only a 20 mM and 50 mM imidazole wash.
In order to extract the recombinant protein from the cytosol of the E. coli, I must first lyse
the cells. After lysing, the cell lysate solution is centrifuged to separate the lipids and unbroken
Figure 16. Comparing the effectiveness of different
lysing protocols. 14% SDS-Page gel of U1A elutions
after lysing with protocol 1 (1), protocol 2 (2),
protocol 3 (3), or protocol 4 (4).
Table 2. Different lysing protocols. I used 4 different lysing protocols to
test the effectiveness of each protocol in producing the highest U1A to
contaminant ratio.
25

cells from the recombinant protein which should be in the supernatant if the protein is hydrophilic
(Protein Purification Extraction and Clarification, EMBL). I experimented with 2 different lysing
methods. The first was sonication, which uses sound waves to disrupt cellular membranes and
release intracellular content. I used 2 different Fisher Scientific sonicators; the 550 Sonic
Dismembrator and the 60 Sonic Dismembrator. Both sonicators were tuned beforehand to ensure
proper function. The second lysing method used Bacterial Protein Extraction Reagent (B-PER) by
Thermo Scientific. B-PER is a detergent in Tris buffer that has both enzymatic and chemical
components to help release intracellular content.
I used 4 different lysing protocols to test the effectiveness of each method to increase U1A
concentration in the supernatant of cell lysates. First, the LB solution containing E. coli was
pelleted and resuspended in10 mL of sonication buffer (10 mM Tris-HCl (pH 8.0), 350 mM NaCl,
0.5% Triton X-100, 20 mM imidazole) for protocols not needing B-PER. For protocols needing
B-PER, the pellet was resuspended with B-PER. The resuspended cells were placed on ice during
all sonication times to prevent the overheating and denaturing of U1A.
In figure 16 comparing the effectiveness of different lysing protocols, it is not evident
which method is best. Although the contamination levels seem to be consistent throughout, the
contamination pattern of protocol 2 is similar to protocol 3 and the contamination pattern of
protocol 1 is similar to protocol 4. B-PER might contain foreign enzymes that bind to Ni-NTA
which gives lane 2 and 3 distinct contamination patterns. Protocol 4 seems to produce a higher
U1A concentration than the rest and protocol 2 seems to produce a lower U1A concentration than
the rest. There does not seem to be much difference between the protocols, but the overall
contamination levels are significantly lower when comparing to previous experiments. I
26

hypothesize that increasing the imidazole concentration in the sonication buffer used to wash the
Ni-NTA beads is the reason.
Another way to measure the effectiveness of a lysing protocol is to measure the U1A
concentration in the supernatant and pellet after lysing and centrifugation. U1A is a hydrophilic
protein which should reside in the supernatant. Theoretically, the higher the U1A concentration is
in the supernatant, the more cells have lysed and released their extracellular content. When
comparing the location of U1A in the SDS-PAGE gel, it is evident that only Protocol 4 produces
a similar U1A concentration in the supernatant compared to the pellet. All other protocols produce
a higher U1A concentration in the pellet. Since only the supernatant is collected, Protocol 4 will
be used as the lysing protocol.
2.7 Biotinylation
One method of studying the interaction between U1A and U1hpII is coating the
microelectrodes with U1A. To do this, I must first hybridize streptavidin modified DNA oligos
Figure 17. Comparing the location of U1A to different lysing protocols. 14% SDS-Page gel of the supernatant (1S) and
pellet (1P) using protocol 1, supernatant (2S) and pellet (2P) using protocol 2, supernatant (3S) and pellet (3P) using
protocol 3, and supernatant (4S) and pellet (4P) using protocol 4 after cell lysate is centrifuged. Immediately after lysing
and centrifugation, the pellet was resuspended in 10mL of sonication buffer to accurately compare protein
concentrations to the 10mL cell lysate supernatant solution.
27

onto the nanolevers and then inject biotinylated U1A over the surface to form a stable interaction
with the DNA oligo. The biotin on the U1A will interact with streptavidin with a KD of 40 fM, the
strongest known non-covalent biological interaction (Holmberg et al., 2005). The biotinylatable
tag is fused to the N-terminus of the U1A construct. U1A is then co-expressed with the BirA ligase
during protein induction (induced by IPTG) in an LB broth solution containing 200 μM biotin for
3 hours at 37
o
C. The BirA ligase will then ligate biotin onto the Lys of the biotin tag (Fairhead et
al., 2015).
I conducted 2 tests to ensure that the U1A was biotinylated. First, I used Dynabeads™ M-
270 Streptavidin which has a monolayer of streptavidin covalently attached onto the surface of
magnetic beads 2.8 µm in diameter (Invitrogen™). U1A
was incubated with the Dynabeads™ for 30 minutes.
Then the beads were washed 5 times with Dulbecco’s
phosphate-buffered saline (DPBS). The supernatant from
the first wash was loaded on the SDS-PAGE gel in lane
Not Bound in figure 18. Afterwards, I eluted the bound
U1A from the beads by incubating it in a 10% β-
mercaptoethanol solution for 10 minutes at 100
o
C. The
supernatant was then loaded onto the SDS-PAGE gel in
lane Bound of figure 18. The binding capacity of the
beads is estimated to be about 10 fold less than the amount of biotinylated U1A it was incubating
in. Therefore, lane Bound only contains a small percentage of the total U1A in the solution just as
expected. There were visible contaminants in lane Bound as well, which is unexpected but
nonspecific binding can also occur with Dynabeads™ as well.
Figure 18. Testing the biotinylation of U1A with
streptavidin-coupled Dynabeads
TM
. 14% SDS-Page
gel of proteins bound and not bound to the
streptavidin-coupled Dynabeads
TM
.
28

The second method utilizes the HABA/Avidin reagent by Sigma-Aldrich. This avidin has
4 4-Hydroxyazobenzene-2-carboxylic acid (HABA) molecules attached to it which is quickly
displaced by biotin. The displaced HABA changes the color of the solution which is quantitated
on a spectrophotometer by absorption of 500 nm
wavelength. The change in the absorption of 500 nm
wavelength (ΔA500) is the unit of measurement of the
concentration of biotin in the solution. There is a
significant difference between the ΔA500 of both
biotinylated U1A samples compared to the non-
biotinylated U1A. The expected ΔA500 for a 100µL
sample of an 80µm solution is 0.100-0.400 (Sigma-
Aldrich H2153). Therefore, the difference between the
first and second sample of biotinylated U1A is within the
expected error range for a 100µL sample of a 17µm
solution.
Because of the nature of each experiment, it is evident that the biotin is capable of
interacting with avidin while bound to U1A. In the HABA/avidin reagent test, the biotin was able
to displace HABA effectively at the expected amount for the concentration of U1A in solution.
And the experiment with Dynabeads™ M-270 Streptavidin showed that the biotin was in fact
attached to U1A because of the amount of U1A eluted off of the beads only after being introduced
to harsh denaturing conditions even after the beads were washed several times. Both methods of
validation show positive results and therefore supportive of the notion that this biotinylation
method is effective.
Table 3. Testing the biotinylation of U1A with
HABA/Avidin reagent. First, 100µL of each
sample was incubated with 900µL of the
HABA/Avidin reagent at 20
o
C for 5 minutes. Then
the spectrophotometer was blanked with 100µL
PE140 buffer and 900µL of HABA/Avidin reagent.
The ΔA 500 was calculated automatically when
measuring the absorption of the other samples as
it was blanked with PE140 buffer.
29

2.8 Buffer Exchange
U1A was eluted in 250 mM imidazole wash buffer which contains 10% glycerol. Glycerol
enhances protein stability and prevents aggregation (Vagenende et al., 2009). U1A is therefore
stored in a 10% glycerol solution at -80
o
C until needed for experimental purposes. The viscosity
of glycerol is much greater than water, which changes the dynamic behavior of nanolevers. The
glycerol will cause the nanolevers to move at different speeds through its aqueous environment
than expected for a similar sized molecule which will interfere with the measurements of
hydrodynamic friction created by the U1A and U1hpII interaction. Since the measurement of
hydrodynamic friction is crucial for the calculation of kinetic parameters, we avoid using glycerol
in the running buffer during the measurement of kinetic parameters.
In order to do a buffer exchange on the U1A elution, I use the Zeba™ Spin Desalting
Column with a 7kDa molecular cut off which is about half the size of my protein. These columns
have size-exclusion chromatography resin which allows for the capture of small molecules and
elution of U1A (Thermo Scientific™ ID:89882). The 250 mM imidazole wash buffer is exchanged
to the PE140 buffer (140 mM NaCl, 10 mM Na2HPO4/NaH2PO4, 0.05% Tween20, 50µM EGTA,
50µM EDTA) which is also the running buffer on the DRX2 (Consumables, Dynamic Biosensors).
The Tween20 helps prevent nonspecific binding (Johnson et al., 2013), EDTA and EGTA inhibit
calcium-dependent proteasomes (Cardoso et al., 2011), and the phosphate and sodium chloride
provide proteins with a physiological like habitat to fold and function properly.
To exchange buffers, the resin is first washed with PE140 buffer several times then the
U1A elution is loaded onto the resin and centrifuged. This allows for the elution of U1A into a
PE140 buffer solution. The HABA/avidin reagent is sensitive to glycerol (Sigma-Aldrich H2153),
therefore buffer exchange needs to be done before testing the biotinylation of U1A with the
30

HABA/avidin reagent. The HABA/avidin reagent reacted within the expected range for a given 17
μM solution (Sigma-Aldrich H2153) as mentioned in section 2.7 which is suggestive that the buffer
exchange was successful. Also, the U1A is diluted at least 300 fold with PE140 buffer before
experimentation, therefore it is very unlikely that any leftover glycerol will affect the measured
kinetic parameters.
2.9 Final Protocol
The protocol below is the final recombinant protein production protocol used for all
following experiments. The protocol was created by Dr. Offringa’s laboratory and optimized by
me for the expression of recombinant U1A protein.
2.9.1 Reagents:
1. Sonication Buffer: 10 mM Tris-HCl, 350 mM NaCl, 0.5% Triton X-100, 20 mM imidazole
2. 20 mM imidazole Wash Buffer: 90% Sonication Buffer, 10% Glycerol, 1 mM TCEP, 20
mM imidazole
3. 50 mM imidazole Wash Buffer: 90% Sonication Buffer, 10% Glycerol, 1 mM TCEP, 50
mM imidazole
4. 250 mM imidazole Wash Buffer: 90% Sonication Buffer, 10% Glycerol, 1 mM TCEP, 250
mM imidazole
5. Pertaining to all buffers containing imidazole: cool mixture to 4°C and change pH to 7.5-
8.0 (with HCl or NaOH) and store leftover buffers at 4°C
2.9.2 Transformation
1. Thaw 50 µL of chemically competent ClearColi™ BL21 (DE3) cells
2. As soon as the cells thaw, add 200 ng of pET3d plasmid containing recombinant protein
of interest with a T7 promoter site and 200 ng of the pACYC184 plasmid containing the
BirA gene
3. Keep the solution on ice for 30 min
4. Heat shock the solution in the 42°C water bath for 60 sec
5. Add 500 µL of LB without antibiotics to the solution, then shake and incubate at 37°C for
60 min
6. Meanwhile, prepare agar plates with a final concentration of 100 µg/mL Ampicillin and
100 µg/mL Chloramphenicol
7. Take sample from step 5 and centrifuge at 8000 rpm for 1 min
8. Pipet out 450 µL of the supernatant and use the rest of the supernatant to gently resuspend
the pellet and spread that resuspended solution on the agar plate
31

9. Incubate agar plates at 37°C for 16-24 hours until colonies are visible
2.9.3 Protein Induction and Lysing
1. For each culture, make a 1 L LB solution with a final concentration of 100 µg/mL
Ampicillin and 100 µg/mL Chloramphenicol and pour that solution into a 4 L flask
2. Pick a colony with an autoclaved toothpick and drop the toothpick into the flask
3. Incubate the solution at 37°C while shaking at 225 rpm until the OD600 reaches 0.6 which
takes approximately 16 hours
4. Add 0.24 g of IPTG (1 mM final concentration) and 0.05 g of biotin (200 µM final
concentration) to each 1 L solution and incubate the solution at 37°C for 3 hours while
shaking at 225 rpm
5. Shake the Ni-NTA resin bottle to homogenize the beads
6. For each sample, aliquot 100 µL Ni-NTA agarose resin into a 15 mL falcon tube
7. Centrifuge the aliquot for 2 min at 4000 rpm
8. Discard the supernatant and add 800 µL of sonication buffer
9. Invert tube to resuspend the beads and clamp the tube on a rotator at 4°C for 10 min
10. Centrifuge the tubes for 2 min at 4000 rpm
11. Discard the supernatant and add 800 µL of sonication buffer
12. Invert tube to resuspend the beads and clamp the tube on a rotator at 4°C for 10 min
13. Centrifuge the tubes for 2 min at 4000 rpm
14. Discard supernatant and add 800 µL sonication buffer and store tubes at 4°C until step 21
15. After the 3-hour protein expression period is over, pellet the cells using the SLA-3000 rotor
at 6000 rpm for 10 min at 4°C
16. Resuspend each pellet in 10 mL of sonication buffer for each 1 L culture
17. Dissolve one ROCHE cOmplete™ EDTA-free protease inhibitor pill in each 10 mL
resuspended solution
18. Keeping the solutions on ice at all times, sonicate each 10 mL resuspended solution using
the 60 Sonic Dismembrator at setting 10 for 15 seconds followed by a 30 second pause,
repeat 5 times
19. Pass the entire sample through an 18 gauge needle 5 times to shear DNA
20. Centrifuge each sonicated solution at 13,000 rpm at 4°C
21. Save the supernatant in the 15 mL Falcon with prepared Ni-NTA beads
22. Invert the tube multiple times to resuspend the solution
23. Clamp the tubes on a rotator and rotate at 4°C for 1 hour
2.9.4 Protein Purification
1. Perform all subsequent steps inside the 4°C cold room
2. Centrifuge the 15 mL falcon tube with Ni-NTA beads using the SS34 rotor at 4000 rpm for
3 min
3. Pipet and discard the supernatant
4. Add 10 mL of 20 mM imidazole wash buffer to the tube with beads
5. Rotate for 3 min and centrifuge for 3 min at 4000 rpm
32

6. Pipet and discard the supernatant
7. Add 10 mL of 50 mM imidazole wash buffer to the tube with beads
8. Rotate for 3 min and centrifuge for 3 min at 4000 rpm
9. Pipet and discard the supernatant
10. Add 300 µL of 250 mM imidazole wash buffer and resuspend the beads then transfer the
entire solution into a 1.5 mL tube
11. Invert tube repeatedly for 3 min and centrifuge 3 min at 4000 rpm on a tabletop centrifuge
12. Pipet out and save supernatant as this contains the recombinant protein of interest
13. Add 300 µL of 250 mM imidazole wash buffer into the 1.5 mL tube with the beads inside
and invert tube repeatedly for 3 min and centrifuge 3 min at 4000 rpm
14. Pipet out and save supernatant as this contains the recombinant protein of interest
15. Add 300 µL of 250 mM imidazole wash buffer into the 1.5 mL tube with the beads inside
once again and invert tube repeatedly for 3 min and centrifuge 3 min at 4000 rpm
16. Pipet out and save supernatant as this contains the recombinant protein of interest
17. Store all protein samples at -80°C

Protein quantitation is described in detail in the next chapter.

33

Chapter 3: Protein Quantification
3.1 Introduction
The association rate is the rate at which the analyte binds to its ligand. To measure the
association rate, the DRX2 measures the concentration of bound analyte over time which is why
the unit for ka is per molar per second. The rate at which the analyte is able to bind to its ligand is
directly dependent on the concentration of analyte for a given concentration of ligand. Therefore,
it is the responsibility of the operator to input the concentration of the analyte into the
switchANALYSIS software accurately. For this reason, it is of upmost importance to quantitate
the U1A solution precisely. I explored 3 different methods of quantitation. For the quantitation of
the U1hpII solution, I used the Implen P300 NanoPhotometer
®
to measure the absorbance at a
wavelength of 260 nm which is the wavelength of maximal absorption for RNA (Ganske et al.,
2014). In chapter 4 I will discuss an alternative method for quantitating RNA.
3.2 Bradford Assay
The Bradford assay utilizes the Coomassie brilliant blue G-250 protein-binding dye to
change the color of a protein solution which is measured on a spectrophotometer by its absorbance
at a wavelength of 595 nm (A595). The Coomassie dye most prominently binds to the side chain of
Arg but also interacts with the side chains of His, Lys, Try, Tyr, and Phe with less affinity
(Compton et al., 1985). Those side chains form an ionic interaction with the negatively charged
sulfonate group on the Coomassie dye.
I first created standards using bovine serum albumin (BSA) dissolved in PE140 buffer at
various concentrations to compare the U1A solution to. Then I aliquoted 1000 μL of the Bradford
Protein Assay Dye reagent from Bio-Rad into plastic cuvettes. I then pipetted equal volumes of
each BSA standard and U1A solution into the cuvettes. I incubated the cuvettes at 20
o
C for 10
34

minutes and then measured the A595. I used Microsoft Excel to plot each standard with its
corresponding A595 to calculate the concentration of U1A. The equation
A595=0.8092(concentration(μg/μL))+0.0371 was developed based on figure 19 and the U1A
concentration was calculated to be 0.91 μg/μL.
After conducting a regression analysis, it
calculated the coefficient of determination to be
0.976 which means that the concentration of BSA
is strongly correlated with the A595 with little
variation. The Bradford assay was repeated using
lysozyme as the standard but the Bradford reagent
was not responsive to the lysozyme standards used.
One major drawback of the Bradford assay is that it is dependent on ionic interactions,
which means the calculated concentration is heavily dependent on the amino acid composition.
BSA contains on average 1 Arg for every 23.3 amino acids compared to 16.8 amino acids for my
U1A construct. Furthermore, detergents, such as the Tween20 found in PE140, are known to
interfere with the Bradford assay (Bio-Rad ID:4110065). Therefore, the concentration of U1A
needs to be measured using other methods to validate this calculated concentration.
3.3 BCA Assay
The Bicinchoninic Acid (BCA) assay utilizes the reduction of Cu
2+
to Cu
+
to change the
color of a solution and absorbance is measured at a wavelength of 562 nm (A562). The reduction
of copper mainly occurs through its interaction with the side chains of Cys, Trp, and Tyr but also
with the peptide backbone (He et al., 2011). Since the peptide backbone is universal, the BCA
Figure 19. Bradford assay using BSA as the standard. The
x-axis is the A 595. The y-axis is the concentration of BSA in
units of μg/μL. Each black diamond is a data point
comparing the concentration of each BSA standard to its
A 595.
35

assay is less dependent on the actual amino acid composition than other quantitation methods. The
quantitation methodology is very similar to the Bradford assay.
To conduct the BCA assay, I used the Pierce™ BCA Protein Assay Kit. First I created
standards using lysozyme dissolved in PE140
buffer at various concentrations to compare the
U1A solution to. Then I made the BCA solution
which is 8 parts reagent A (bicinchoninic acid)
and 1 part reagent B (cupric sulfate) (Thermo
Scientific™ ID:23225). Then I aliquoted 200
μL of the BCA solution into each well on a 96
well plate. Then I pipetted 25 μL of each
lysozyme standard and U1A solution into the
respective wells and incubated the plate at 37
o
C for 30 minutes. I then used a BioTek™ microplate
spectrophotometer to measure the A562. The assay was conducted 4 times and the data in figure 20
represents the average with its corresponding 95% confidence interval. I plotted the lysozyme
standards with its corresponding A562 to calculate the U1A concentration. The equation
A562=1.8878(concentration(μg/μL))+0.2483 was developed based on figure 20 and the U1A
concentration was calculated to be 0.42 μg/μL. After conducting a regression analysis, it calculated
the coefficient of determination to be 0.994 which means that the concentration of lysozyme is
even more strongly correlated with the respective absorbance than the BSA in the Bradford assay.
The BCA assay also has a similar drawback to the Bradford assay in the sense that it is
affected by the amino acid composition, and the extent of the effect itself depends on the amino
acid composition. The BCA assay is not affected by detergents but is affected by reducing agents.
Figure 20. BCA assay using lysozyme as the standard. The x-
axis is the A 595. The y-axis is the concentration of lysozyme in
units of μg/μL. Each blue diamond is a data point with the
95% confidence interval in black, comparing the
concentration of each lysozyme standard to its A 595.
36

Fortunately, PE140 buffer does not contain any reducing agents. Unfortunately, the concentration
calculation does not coincide with the calculation from the Bradford assay. One additional problem
is that the assays will not separate out U1A from contaminants, therefore SDS-PAGE gels must
should also be run to determine which assay is more accurate and can help compare relative
concentrations.
3.4 SDS-PAGE Comparison
The sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) gels allow
one to separate proteins by size by electrophoresis. SDS is an anionic detergent that denatures
proteins and gives them a net negative charge. This allows the protein to migrate through the gel
uniformly, allowing for the separation of proteins by size with no effect from the protein’s charged
amino acid side chains. To denature the protein, loading dye (125 mM Tris-Cl, 5% SDS, 10% β-
mercaptoethanol, 25% glycerol, 0.1% bromophenol blue) is mixed with the protein solution before
heating the solution to 100
o
C for 10min. After running
the gel for 90 min at 20mA, the gel is washed with H2O,
stained with the dye used in the Bradford assay
(Coomassie brilliant blue G-250), and imaged with the
Bio-Rad ChemiDoc™ XRS+ imager.
The most important benefit of quantitating with
SDS-PAGE gels is visualizing the protein of interest
independent from the contaminants in solution. The 2
other quantitation methods are unprejudiced against
contaminants which adds an extra variable of inconsistency between the quantitation of different
batches of recombinant protein produced. With an SDS-PAGE gel, you can quantitate only the
Figure 21. First U1A quantitation with SDS-PAGE. In
each well, there is 10 μL of loading dye and 10 μL of
0.40 μg/μL lysozyme standard, U1A, 0.50 μg/μL
lysozyme standard, 0.80 μg/μL lysozyme standard,
0.90 μg/μL lysozyme standard. The gel was running
at 20mA for 90min.
37

protein of interest and also visualize the extent of contamination. This is assuming that there are
no contaminants of equal size to the protein of interest. Lysozyme, a 14.3kDa protein, is used as
the standard once again because of the close resemblance in size to my U1A construct. The reason
size is important is because the length affects a protein’s migration across a gel and different size
proteins are more difficult to compare.
In figure 21, it is evident that the actual concentration of U1A is much less than the
calculated concentration from the Bradford assay and also less than the BCA assay. More gels
were run including figure 22 which shows the concentration of U1A to be between 0.20-0.30
μg/μL. Many more gels were run to confirm the estimation of the concentration being 0.25 μg/μL.
Unfortunately, quantitating with gels also has its own drawbacks, the first being the contrast
limitations of the human eye. Another is the imaging system used. In my gels, there is a white
circle in the top center. This is caused by the light source used in the imager, which is why the gels
were placed in a certain fashion to avoid the overlap of the white spot with the bands on the bottom.

Figure 22. Second U1A quantitation with SDS-PAGE. In each well, there is 10 μL of loading dye and 10
μL of AccuRuler RGB Plus Prestained Protein Marker, 0.25 μg/μL lysozyme standard, U1A, 0.20 μg/μL
lysozyme standard, U1A, 0.30 μg/μL lysozyme standard, U1A, 0.35 μg/μL lysozyme standard. The gel
was run at 20mA for 90min. The size of the smallest 4 protein markers is noted left of the gel in kDa.
38

3.5 Discussion
There are many different ways to quantitate protein. The method used depends highly on
the amino acid composition as seen with the mechanism of action of both the Bradford and BCA
assay. The Bradford assay calculated a very high concentration (0.91 μg/μL) while the BCA assay
calculated a much lower concentration (0.42 μg/μL). The concentration calculated with the BCA
assay was much closer to what was visualized on the SDS-PAGE gels. SDS-PAGE gels do not
calculate a concentration per se, but it is evident from the gels that the concentration of U1A is
comparable to the lysozyme standard of 0.25 μg/μL. The lysozyme concentration is based on
purchased protein dissolved to the correct concentration. Therefore, 0.25 μg/μL or 17 μM is the
concentration I will be using to base all further kinetic measurements on. It is a given that the exact
concentration is impossible to measure but we are aiming for a consistent protein concentration.
When different mutants are compared in the future, the goal is for the variability in concentration
to be smaller than the difference in the kinetic parameters of the mutants. That way, differences in
association rates cannot be attributed to differences in concentration.

39

Chapter 4: RNA Production
4.1 Introduction
The method of attaining RNA for protein/RNA interaction studies is important due to its
cost and instability. RNase is abundant in laboratory settings and easily contaminates RNA
samples. Just 3 seconds of human contact to the RNA samples can increase the RNase
concentration high enough to cause degradation. Even at -80
o
C, RNA degradation is inevitable,
with signs of it within the first month (Mathay et al., 2012). Therefore, I explored multiple methods
to produce large quantities of RNA directly before conducting the experiments to ensure minimal
degradation.
I need to produce 3 different RNA strands. The first 2 are used to coat the gold covered
surface with RNA and the third is used to bind to the U1A coated surface. To coat the gold covered
surface, the RNA must contain the U1hpII hairpin sequence plus a 3’ sequence complementary to
the nanolever it is attaching to. I tested 2 different RNA constructs to coat the surface. The first
construct, named –Spacer, is 76 nucleotides long and only contains the 28 nucleotide U1hpII
sequence and a 48 nucleotide 3’ sequence which is complementary to the 48 nucleotide DNA
nanolevers. The second construct, named +Spacer, contains the exact same sequence as –Spacer
except an additional 4 adenosine spacer between the U1hpII sequence and the complementary 3’
end. This RNA is therefore 80 nucleotides in length. To measure the kinetic parameters between
U1hpII and the surface-bound U1A, I only need the 28 nucleotide U1hpII RNA sequence which
is the third RNA sequence simply named U1hpII.
4.2 In Vitro Transcription
For the purposes of saturating the nanolevers with RNA, we use approximately 50 μg of
RNA per run. For optimization purposes, we conducted many runs, therefore decided to try the
40

HiScribe™ T7 High Yield RNA Synthesis Kit by New England Labs. It is advertised to produce
180 μg of RNA per reaction and only requires an overnight incubation. The cost per reaction is
only $4, therefore it is also a cost-effective method. To transcribe RNA in vitro, I used 2 different
DNA templates, a plasmid and short oligos.
4.2.1 Transcribing from Linearized Plasmids
The first method for in vitro transcription uses a linearized plasmid. The plasmid must
contain a T7 promoter site directly upstream of the RNA sequence of choice because the kit utilizes
the T7 RNA polymerase. I used the same pET3d plasmid that the U1A sequence is inserted into
and digested it with BglII and MluI. Then I inserted an oligo between the two sites that contains
the exact same sequence up until the end of the T7 promoter site and then the sequence of the RNA
I needed to transcribe. The plasmid is constructed in a way that when the plasmid is digested with
MluI, the plasmid is linearized with one end containing the T7 promoter site followed by the exact
RNA sequence needed. The reason why MluI restriction enzyme was chosen is because it creates
a 5’ overhang, which is suitable for RNA transcription. It is hypothesized that not linearizing a
plasmid with a 5’ overhang causes the T7 RNA polymerase to attach onto the anti-sense strand
and continue transcription until it falls off creating a much longer RNA sequence. Once the DNA
plasmid is linearized, T7 RNA polymerase along with RNA nucleotides are mixed with the
plasmid and incubated at 37
o
C for 14 hours. This is the standard method used for RNA
transcription for all following experiments.
The RNA is run on TBE Urea-PAGE gels which denatures the RNA strands and prevents
the formation of secondary structures. The RNA solution is also mixed with a loading dye (47.5%
formamide, 0.01% SDS, 0.01% bromophenol blue, 0.005% xylene cyanol, 0.5 mM EDTA) that
augments the denaturing of RNA and also prevents the degradation of RNA (NEB ID:B0363S).
41

The RNA solution is first heated to 100
o
C for 5 min then loaded into the wells, ran at 20mA for
60 min and then washed with TBE buffer and stained with ethidium bromide before being imaged
on the Bio-Rad ChemiDoc™ XRS+ imager.
In figure 23, it is evident that the production of RNA using a linearized plasmid as a
template is effective. The bands in lanes -Spacer and +Spacer are of expected size. For lane U1hpII,
the size is near the 21-nucleotide dsRNA marker and since the marker is for dsRNA and not
ssRNA, I cannot confirm the exact size. The IDT ordered HPLC purified RNA oligo in lane U1hpII
(IDT) is definitely smaller than the transcribed strand in lane U1hpII (both of which are 28
nucleotides long) leading us to believe that the IDT ordered RNA oligo has degraded. Also, the
bands above lanes -Spacer, +Spacer and U1hpII stuck in the wells are the 4.7kb linearized plasmid
templates.
Figure 23. 15% TBE Urea-PAGE gel of transcribed RNA using a linearized plasmid as the template. In each well
there is 20 μL of RNA loading dye and 20 μL Low Range ssRNA Ladder from NEB, 2000 ng of –Spacer, 2000 ng of
+Spacer, 700 ng of U1hpII, 2000 ng of U1hpII produced by IDT, 20 μL dsRNA Ladder from NEB. On the left hand
side, the numbers signify the length in nucleotides of the respective ssRNA markers. On the right hand side, the
numbers signify the length in nucleotides of the respective dsRNA markers.
42

The one major drawback with using the T7 RNA polymerase is that the RNA sequence
must start with ‘GGG’ to ensure optimal transcriptional efficiency. The third nucleotide can be
replaced with adenosine with decreased efficiency (Ikeda et al., 1991), but I decided to just use the
recommended starting sequence of ‘GGG’. The U1hpII sequence used for all experiments contains
3 guanosine residues at the edge of its stem which is paired with 3 cytosines on the opposite side.
The modification of the U1hpII sequence should not cause a change in the kinetic parameters as
the key loop residues are unchanged (Law et al., 2006).
4.2.2 Transcribing from Short Oligos
Short DNA oligos can be used as a template as well as long as the T7 promoter region is
double stranded. I ordered 2 complementary DNA oligos from Integrated DNA Technologies
(IDT) and hybridized the oligos together creating a blunt end on both sides. The double stranded
oligo was then mixed with the same reagents as the linearized plasmid to transcribe RNA. The
sequence of this oligo contains the T7 promoter site, the RNA sequence
of choice and 14 basepairs upstream of the T7 promoter site to increase
transcription efficiency. In fact, by just including 9 basepairs upstream of
the T7 promoter site, RNA transcription increases by 3-fold (Baklanov et
al., 1996).
This method of trasncription seemed to be working well until a
TBE Urea-PAGE gel was run. The RNA transcription was repeated 5
times with identical results each time. In figure 24, the 15% TBE Urea-
PAGE gel was run directly after RNA transcription to decrease the
probability of the RNA degrading. Identical quantity of RNA was loaded
into each lane to show the expected intensity of the band near the top of
Figure 24. 15% TBE Urea-
PAGE gel of transcribed RNA
using 2 different methods. In
each well, there is 1300ng of
+Spacer transcribed using a
linearized plasmid, or short
DNA oligos as the template.
43

figure 24. The band near the top in lane Linearized Plasmid is significantly more intense than lane
Short Oligos while lane Linearized Plasmid has significantly less distinct bands in total than lane
Short Oligos. This smearing is common in degraded RNA samples, but the likelihood of
degradation in every sample directly after RNA transcription is very slim. I hypothesize that the
T7 RNA polymerase keeps detaching from the short oligos during the transcription process,
causing the production of much shorter RNA fragments rather than a concentrated RNA solution
of a single length. It is also possible that the IDT short oligos are contaminated with RNase.
4.2.3 Purification
The solution RNA is transcribed in contains buffers and enzymes that might hinder the
RNA’s ability to interact with U1A. Therefore, the RNA solution is purified using Monarch
®
RNA
Cleanup Kit which is advertised to cleanup solutions with up to 500 μg of RNA for oligos longer
than 15 nucleotides. The kit utilizes a column with a resin at the bottom to capture the RNA. First
the RNA is precipitated using a binding buffer with ethanol. Then the resin is washed several times
with ethanol and eluted using H2O. The kit consistently purifies my RNA samples at high purity
(260/280 ratio of >1.7 and 260/230 ratio of >1.9). Absorbance at a wavelength of 280 nm signifies
protein contamination and at a wavelength of 230 signifies organic compound contamination
(Assessment of Nucleic Acid Purity, Thermo Scientific™). Therefore, it is essential that this kit is
consistently producing RNA solutions with high 260/280 and 260/230 ratios.
4.3 RNA from IDT
As an alternative to transcribing RNA, I sought to buy the RNA oligos from IDT expecting
a much higher quality product and increased consistency between batches of RNA. First, I ordered
an RNA/DNA mixed oligo which contained the U1hpII 28 nucleotide RNA covalently attached to
a 48 nucleotide DNA that is complementary to the nanolevers, the same sequence as the –Spacer.
44

Its ability to saturate the surface and increase fluorescence during the hybridization process was
marvelous but it did not seem to interact with U1A when measuring the kinetic parameters on the
DRX2. After being stored at -80
o
C for 6 months, the oligo was run on a TBE Urea-PAGE gel.
There was a lot of smearing between the 50 and 80 nucleotide ssRNA markers which is expected
because the DNA strand was 48 nucleotides long and the RNA attached was 28 nucleotides long.
I hypothesize that only the RNA portion was degraded leaving oligos 48-76 nucleotides long
behind causing the smearing between the 50 and 80 nucleotide ssRNA markers.
Considering the expense involved in ordering long RNA oligos, I decided to only order the
28 nucleotide U1hpII RNA which was desalted by IDT before being shipped to us. I ordered the
oligos right before using them to mitigate the degradation. This time I encountered complications
which could have arised from the impure RNA or the conjugation method used to attach U1A onto
the surface which will be explained in chapter 6. Representatives from IDT suggested using HPLC
purified RNA which they provided to us free of charge to try.
High-performance liquid chromatography (HPLC) injects samples at high pressure through
a chromatographically packed column to separate out the individual compounds in solution
(HPLC, Linde). This purification method should increase RNA purity and therefore stability
(Weissman et al., 2013). The HPLC RNA sample was run on lane U1hpII (IDT) on the TBE Urea-
PAGE gel in figure 23 after incubating at -80
o
C for just a week. Most of the RNA sample is below
the expected size (using the RNA in lane U1hpII as reference). This HPLC RNA was run on the
DRX2 to measure its association rate to U1A which was approximately 100-fold lower
(6.31±0.83E+4M
-1
s
-1
) than the expected kON value seen in previous publications. The dissociation
rate could not be calculated which is believed to be caused by the lack of RNA bound to nanolevers
which causes a change in florescence too small for the DRX2 to detect. Because of the high price
45

and lack of stability of IDT ordered RNA oligos, I decided to continue using in vitro transcription
kits for all subsequent experiments.
4.4 Quantitation
When measuring the association rate using U1A as the ligand, U1hpII must be injected
over the surface as the analyte. Therefore, the kON is calculated using our given concentrations of
U1hpII instead of U1A. Consequently, I must accurately quantitate U1hpII. After measuring the
absorbance of U1hpII at a wavelength of 260 nm (A260) on the Implen P300 NanoPhotometer
®
,
the calculated concentration was 47 ng/μL (5.2 μM). The drawback of using the A260 for
quantitation purposes is that the DNA plasmid, which was used as a template to transcribe the
RNA from, is also absorbed at this wavelength. Therefore, I run TBE Urea-PAGE gels to compare
the band intensity to known standards to measure its concentration more accurately. I considered
adding DNase to the RNA solution before purification to rid the solution of the DNA template but
any active DNases in the RNA solution will contaminate the biochips and degrade the nanolevers.
Therefore, I was reluctant to try adding DNase to prevent irreversible damage to the DRX2.
Figure 25. 15% TBE Urea-PAGE gel used to quantitate U1hpII. In each well there is 10 μL of Loading Dye and 10
μL of 0.1 μM DNA Standard, 0.3 μM DNA Standard, 0.5 μM DNA Standard, U1hpII, 0.7 μM DNA Standard, 1.0 μM
DNA Standard, U1hpII, 2.0 μM DNA Standard, 3.0 μM DNA Standard.
46

The standard I used this time for comparison is a 28 nucleotide long DNA oligo that
contains the exact same sequence as U1hpII except replacing uracil nucleotides with thymine. I
followed the same methodlogy for determining the concentration of U1hpII as I did for U1A in
chapter 3. In figure 25, the instensity of the U1hpII bands are somewhere between 1 μM and 2
μM, much lower than the Implen P300 NanoPhotometer
®
calculated concentration of 5.2 μM.
Furthermore, there are several bands above and below the expected band size. Smaller band sizes
can be caused by RNA degradation. Larger band sizes can be caused by fragments of the plasmid
or aggregation of U1hpII.
4.5 Discussion
After exploring multiple RNA production methods, the NEB in vitro transcription kit
seems to be the more cost effective and efficient method compared to ordering from IDT. For the
in vitro transcription kit, it is imperative to use a linearized plasmid instead of short oligos to
consistently produce RNA of an expected length. TBE Urea-PAGE gels are indispensable for the
quantitation of RNA as spectrophotometers do not account for contamination and aggregation. I
will be using the 80 nucleotide +Spacer RNA sequence (25.84kDa) as the ligand to attach to the
nanolevers which will be further discussed in the next chapter. The 28 nucleotide U1hpII sequence
(9.05kDa) will be used as the analyte for experiments using U1A as the ligand.

47

Chapter 5: Kinetic Analyses
5.1 Introduction
All of the kinetic analyses were performed on the DRX2. Before starting, the instrument
must be prepared for the experiment with the respective solutions. Analyses of the results were
conducted on proprietary software. I did these analyses with Anusha Muralidhar, who is the
resident DRX2 expert. We analyzed the hybridization of the ligand to the nanolevers and the
kinetic parameters of the interaction between U1A and U1hpII.
5.1.1 Preparation and Experimental Overview
The DRX2 is stored with a 70% ethanol solution in the fluidics system which is the
preferred storage condition, among others to prevent the aggregation of biological compounds in
the fluidics system. We used the MPC2-48-2-G1R1-S biochip for all experiments; it is a multi-
purpose chip at standard nanolever density of 400 nanolevers per μm
2
with 2 sets of 48-nucleotide
oligos with different sequences attached onto the surface. These 48 nucleotide oligos either have
a red or a green fluorophore attached to their 3’ end. The biochip was aligned with the DRX2 chip
place holder to make sure the fluidics system is attached properly. The entire system was then
primed to clear the tubes of the ethanol solution and replace it with the buffer used during the
kinetic measurements (analyte buffer).
A passivation solution, containing a thiol-reactive compound, was injected over the biochip
to react with any unreacted thiol groups on the surface to prevent nonspecific binding of organic
molecules to the gold surface (FAQs, Dynamic Biosensors). The condition of the microelectrodes
was then viewed to determine which of the 6 microelectrodes in a given flow channel had good
electrode viability. The detector was then focused onto our chosen microelectrode. The
fluorescence intensity should only be recorded from a single microelectrode rather than all 6 and
48

if multiple microelectrodes are chosen to be recorded from, the fluorescence detector must move
its aim during the experiment which decreases the number of data points collected.
Afterwards, switchBUILD was used to create a task. Then switchCONTROL was used to
execute said task on the DRX2. All experiments started with the functionalization protocol to ready
the biochip for kinetic measurements. First, a regeneration solution, a denaturing solution with a
basic pH value, was injected over the surface to release any nanolever-bound molecules. The
ligand solution was then injected over the surface to hybridize to the nanolever of choice. The
analyte solution was then injected over the surface to interact with the ligand to measure the
association rate. Subsequently, just the analyte buffer was injected over the surface to cause
dissociation of the analyte from the ligand to allow measurement of the dissociation rate. Our
experiments usually contained 5 different concentrations of the analyte in a single experiment and
each concentration of analyte was tested independent of the order placed and previous kinetic
measurements. Because the surface is completely regenerated for each binding cycle by repeating
the functionalization step between each kinetic measurement of analyte solution, the surface is
void of previously bound molecules. This is a benefit compared to how SPR experiments are done.
However, there could still be changes over time on the chip surface, such as loss of fluorescence,
so the different analyte concentrations should be tested in random order to avoid any biases.
5.1.2 Analyte Dilutions
To avoid inconsistencies and loss of protein activity, fresh dilutions were made
immediately before each run. The analyte stock solution, whether U1A or U1hpII, was in the low
μM range and was diluted to the nM and pM range. First, 10 μL of the analyte was mixed with 90
μL of analyte buffer to create a 10-fold dilution. Then 900 μL of analyte buffer was added to the
solution to create a 100-fold dilution. 200 μL of the 100-fold dilution was then added to 400 μL of
49

analyte buffer to create a 300-fold dilution. Serial dilutions were made repeating the last step to
create a 900-fold, 2700-fold, 8100-fold and 24300-fold dilution. The exact same dilution method
was repeated for most experiments to decrease inconsistencies between runs.
5.1.3 Analysis Modes
At the end of the experiment, the data was exported into switchANALYSIS to analyze the
data. There were two types of data we were interested in for this particular experiment:
functionalization and kinetics. The functionalization data provides information about the
hybridization of the ligand to the nanolever, i.e. the creation of the ligand surface. This should be
monitored carefully, because a consistent surface is needed to get reliable kinetic data. The kinetic
data provides information about the interaction between the ligand and analyte.
There are 2 nanolevers on our biochips: A48 and B48. A48 contains a green fluorophore
and a unique DNA sequence; it’s used as a reference for our experiments. B48 contains a red
fluorophore and a different unique sequence that binds to our ligand of interest. We used the
intensity of the green fluorophore for referencing purposes. Every experiment started with
measuring the interaction between just the analyte buffer and both nanolevers. Then we measured
the interaction between all the analyte solutions with both nanolevers. This allowed us to have an
analyte negative control (interaction between analyte buffer and B48), a ligand negative control
(interaction between analyte solution and A48), and a double negative control (interaction between
analyte buffer and A48).
Once the data was gathered, the fluorescence intensity measurements from the interactions
between an analyte solution and B48 were normalized to the analyte negative control. Then a
global mono-exponential fit was created to produce the association and dissociation curves. Mono-
exponential fits are used for data that exponentially increase or decrease. If the protein undergoes
50

conformational changes during the interaction, a global bi-exponential fit can be used to analyze
the data. Once the kinetic parameters are calculated, switchANALYSIS will output the kON and
kOFF rates with an error rate. The error rate provides the 95% confidence interval for the calculation
based on the experiment done. The standard errors are calculated based off of the regression
analysis of a monophasic fit model (Weber, Dynamic Biosensors).
There is a subjective aspect to analyzing the kinetic data. The timeline for the experiment
begins from the time the analyte is injected over the surface (time 0). The user can artificially
adjust time 0 to any other time thereafter. Sometimes air bubbles cause an erratic change in
fluorescence intensity. Since this affects the kinetic calculations, the erratic change in fluorescence
intensity should be avoided, but if present, we adjusted time 0 accordingly. We refrained from
adjusting time 0 unless absolutely necessary.
5.2 Creation of the Ligand Surface
In order to analyze the interaction between an analyte and a ligand, the ligand must first
be hybridized onto the B48 nanolevers. Ideally, the ligand should fully saturate the B48 nanolevers.
We explored different methods for hybridizing both the U1hpII RNA and U1A protein onto DNA
oligos.
5.2.1 U1hpII RNA as the Ligand
The first orientation we used for measuring the kinetic parameters was hybridizing the
U1hpII RNA onto the DNA nanolevers. The U1hpII ligand solution contains 50% U1hpII solution
in H20, 25% CNL-A48 (complementary DNA oligo to the A48 nanolever) in PE40 and 25% PE40.
PE40 is identical to PE140 except contains 40 mM NaCl instead of 140 mM NaCl. The final ligand
solution contains 5 mM Na2HPO4/NaH2PO4, 20 mM NaCl, 0.025% Tween20, 25 µM EDTA and
25 µM EGTA. I used 2 different U1hpII RNAs. The first, called –Spacer (76 nucleotides long),
51

has the 28 nucleotide U1hpII sequence and a 3’ end complementary to the B48 nanolever. The
second, called +Spacer (80 nucleotides long), has the exact same sequence as –Spacer with the
addition of a 4-adenosine spacer in between the U1hpII and the 3’ B48 complementary end.
The hybridization curves in figure 26 illustrate the saturation of the nanolevers. The
nanolevers are lying close to the positively charged gold surface before hybridizing to DNA or
RNA. Since the gold surface quenches fluorophores, the fluorescence intensity is very low. As
DNA or RNA hybridizes to the nanolevers, they force the nanolever into a rigid state perpendicular
to the gold surface, unquenching the fluorophores and increasing the fluorescence intensity
(measured in kilo counts per second). This change in fluorescence intensity is graphed versus time
to confirm hybridization.
Figure 26. Comparison of different RNA constructs and concentrations used to saturate the gold surface. The green line
represents CNL-A48 binding to A48 nanolevers. The red line is A) an 8 μM –Spacer solution B) an 8 μM +Spacer solution C) a 32
μM –Spacer solution D) a 32 μM +Spacer solution binding to B48 nanolevers.
52

CNL-A48 (green line) is the control oligo. In all 4 graphs, it consistently rapidly increases
the fluorescence intensity and quickly plateaus suggesting that most if not all the nanolevers are in
a rigid formation therefore hybridized to the control oligo. The more nanolevers hybridized to
U1hpII, the more interactions are essentially being analyzed, which theoretically should decrease
the error rate and increase consistency between experiments. We expect the RNA (red line) to
behave similarly to CNL-A48. Figure 26 suggests that increasing the U1hpII concentration and
adding a 4-adenosine spacer both increases the rate of increase in fluorescence intensity and
saturation of the nanolevers. It appears that the fluorophore at the end of the nanolever and the
RNA stem interfere with each other, necessitating the spacer. Therefore, we only use the +Spacer
U1hpII construct for subsequent experiments. The concentration of RNA transcribed fluctuates
between 5 μM and 70 μM but in every case, the highest possible RNA concentration was used to
hybridize the ligand onto the nanolevers.
We tested whether hybridization potential could be optimized by heating the U1hpII
solution to 90
o
C for 5 minutes before placing it on ice to quickly cool the solution down. It was an
attempt to force the entire U1hpII solution into the most thermodynamically stable configuration,
which is the secondary structure of choice. This method was abandoned because it did not make
the surface coating more consistent.
5.2.2 Consistency of U1hpII Hybridization
When an experiment utilizes 5 different analyte solutions to determine the kinetic rates of
an interaction, it is important that the hybridization of the ligand is consistent throughout the
experiment. This can be done by overlaying the ligand annealing from each injection. Ideally, the
curves should overlap. However, this was not the case in our first experiment (figure 27A).
53

During the experiment, the U1hpII ligand solution incubates in the autosampler, waiting to
be injected over the gold surface for each new set of injections. The autosampler is left at 20
o
C
because decreasing the temperature of the autosampler tends to cause condensation within the
DRX2. In figure 27A, we see a smaller increase in fluorescence intensity the longer the U1hpII
ligand solution is left in the autosampler. This could be caused by RNA degradation. After
incubating at 20
o
C for just 285 minutes (pink line), almost no increase in fluorescence intensity is
detected. Furthermore, the DRX2 was not able to detect an interaction between U1A and the
hybridization that is represented by the pink line.
We repeated the experiment using RNA dissolved in autoclaved PE40. In figure 27B, we
see consistent hybridization curves through the entirety of the experiment; curves overlay quite
nicely. An added difference in the experiment in figure 27B is that the experiment was much
shorter than 27A, because we rearranged the experimental procedure. In figure 27A, a dissociation
test was conducted after every association test, compared to figure 27B where only 1 dissociation
test was conducted after all of the association tests ended. Association tests last about 5 minutes
while dissociation tests last an hour due to the high stability of the U1A/U1hpII complex. By
Figure 27. Change in U1hpII hybridization curve during subsequent runs. A) The U1hpII ligand solution was incubated at 20
o
C
for 15 minutes (red), 105 minutes (orange), 195 minutes (yellow) and 285 minutes (pink) before being injected onto the gold
surface to bind to the nanolevers. B) Using autoclaved PE40 instead, CNL-A48 (green/blue/purple/black) and U1hpII
(red/orange/yellow/pink) were incubated at 20
o
C for 15 minutes (red/green), 45 minutes (orange/blue), 75 minutes
(yellow/purple) and 105 minutes (pink/black).
54

decreasing the number of dissociation tests conducted in figure 27B, we were able to shorten the
experimental time dramatically. The shorter incubation time may also prevent degradation.
5.2.3 U1A as the Ligand Using Amine Coupling
To with test the suitability of using the U1A protein as the ligand, the protein needs to be
somehow attached to the nanolevers. The 2 most common attachment mechanisms are covalent
coupling to a nanolever complementary oligo, and non-covalent linking to a molecule attached to
a nanolever complementary oligo (for example using the tight biotin/streptavidin interactions or
antibodies).
The first method we tried was covalently coupling U1A onto a DNA oligo complementary
to the B48 nanolever (CNL-B48) using amine coupling. U1A is first incubated in a solution
designed to covalently bind a primary amine group onto the CNL-B48 oligo. According to
Dynamic Biosensors, the kit is specifically designed to target the N-terminal amine. However, in
retrospect the company engineers were not so sure of this, and our data (figure 28) suggests this
may not be the case. After the amine coupling reaction, the protein is purified using the ÄKTA
start protein purification system. ÄKTA conducts ion exchange chromatography and utilizes a
triple wavelength UV detection system to show which fractions contain biomolecules (GE
Healthcare Life Science). Fractions that contain protein produce a clear peak on the UV absorption
monitor. There were 2 peaks of interest that might contain the U1A protein covalently bound to
CNL-B48. We conducted a kinetic analysis on both peaks to determine which fraction contained
U1A covalently bound to CNL-B48. In figure 28, only the first peak seemed to interact with
U1hpII. The observed association rate was ~1.5 x 10
5
M
-1
s
-1
, and dissociation rate ~6 x 10
-3
s
-1
.
The association rate was approximately 60-fold slower and dissociation rate was approximately
20-fold faster compared to previously published data which utilized SPR technology (Law et al.,
55

2005). This resulted in an affinity that was ~3 orders of magnitude weaker compared to previously
published data which utilized SPR technology (Law et al., 2005). This could be due to damaged
protein or inaccurate quantitation; a more likely cause is the effect of the amine coupling.
Figure 28. Kinetic analysis of 2 experiments using amine coupling mechanism to hybridize U1A onto nanolevers. A) The
association graph and B) dissociation graph of peak 1. C) The association graph and D) dissociation graph of peak 2. The
concentration of the U1hpII analyte solution and its respective color on the graph is shown on the upper right hand side of the
respective graph. The calculated association and dissociation rates are above the respective graphs. E) The graph produced by
the ÄKTA start protein purification system comparing the fraction number to absorbance using a UV detection system. The peaks
mentioned are labeled above fraction 10 and fraction 13.
56

Amine coupling tends to occur on a primary amine group
which is located on the side chains of Lys and the N-terminal
amino acid. Although the company initally stated that their kit
specifically targeted the N-terminal amine, when questioned, they
conceded that a lysine-rich protein might present too much of a
challenge. The more Lys present in a given region, the higher the
probability the reaction will occur in that location. U1A has a Lys
dense region which is shown in blue in the center of figure 29. If
the amine coupling occurs in this region, the effect of the positively
charged Lys side chains may be neutralized by the covalently
attached negatively charged CNL-B48 and the DNA attachment may sterically hinder the
interaction with the RNA. Our lab showed that replacing Lys20,22,23 with the negatively charged
Glu, the KD increased by 11000-fold (Law et al., 2006). Therefore, it is possible that some or all
of the CNL-B48 is covalently attached to the Lys dense region of U1A. Therefore, the amine
coupling mechanism might not be a viable option for studying the U1A/U1hpII interaction. We
therefore turned to a more site-specific method
5.2.4 Biotin/Streptavidin Coupling of U1A onto the Surface
The next method for coating U1A onto the nanolevers makes use of the high affinity
biotin/streptavidin interaction. To this end, we ordered CNL-B48 with streptavidin already
covalently attached to the 5’ end (CK-SA-1-B48) so that U1A carrying a N-terminally biotinylated
tag could be presented as the ligand. Because attached biotin on the N-terminal biotinylatable tag
is away from kinetically important regions of U1A, this method should provide more reliable data
than chemical coupling.
Figure 29. Lys20,22,23 are important
amino acids for the U1A/U1hpII
interaction. The positively charged
side chains of Lys20,22,23 interact
with the negatively charged stem of
U1hpII to help stabilize the interaction
after initial contact (Law et al., 2006).
57

Two experiments were conducted. Both experiments were conducted on a new flow
channel of the same biochip used for the experiments where the RNA was put on the surface. Each
flow channel should lose fluorescence independent of the other flow channels (fluorescence is lost
over time when surfaces are exposed to the detector), therefore a new flow channel was used to
provide consistency when comparing analyses between different orientations. The recommended
Dynamic Biosensors protocol was used to coat the biotintylated U1A protein onto the surface.
First, a solution containing 200 nM CNL-A48 and 200 nM CK-SA-1-B48 diluted in PE140 was
injected over the biochip surface to generate the streptavidin surface. This process was not without
problems, judging from the fact that different injections to hybridize the SA-linked oligo did not
overlay (figure 30A). Of the three injections, only the third one appeared to give an optimal pattern
(slow and highest level of saturation). Because of these inconsistencies, it was not unexpected that
Figure 30. The 2-step procedure for coating the biochip surface with U1A. A) A 200 nM CNL-A48 (green/blue/purple) and 200
nM CK-SA-1-B48 (red/orange/brown) solution was incubated at 20
o
C for 15 minutes (red/green), 60 minutes (orange/blue) and
90 minutes (brown/purple) before being injected onto the gold surface to bind to the nanolevers followed by B) the association
of a 1 μM U1A ligand solution shown in green, red and blue in that order. C) A 200 nM CNL-A48 (green/blue) and 200 nM CK-SA-
1-B48 (red/orange) solution was incubated at 20
o
C for 15 minutes (red/green) and 70 minutes (orange/blue) before being
injected onto the gold surface to bind to the nanolevers followed by D) the association of a 1 μM U1A ligand solution shown in
green and red in that order.

58

the biotinylated U1A did not bind equally to the surface each time when a 1μM biotinylated U1A
solution diluted in PE140 was injected. Based on the tight interaction between SA and biotin, a
very rapid association should be expected. This is seen for two of the injections (figure 30B).
However, the surface was not stable, exhibiting a slow drift back to baseline. One possibility is a
problem with the DRX2 fluidics system, resulting in only a small portion of the analyte actually
being injected over the biochip surface. A service visit has been scheduled. Because of this, these
biotin-streptavidin experiments were not continued.
5.3 Measurement of Association Rates
The association or kON rate is the rate at which protein binds to its target per mole per
second. All of the experiments are conducted in dynamic mode which allows us to analyze the
association rates using 2 different analysis tools. The first is dynamic response, which measures
changes in fluorescence, influenced by the hydrodynamic friction created by bound proteins. The
behavior of the dynamic response is more predictable than fluorescence proximity sensing (FPS)
as there is a direct correlation between size and hydrodynamic friction. The protein diameter can
also be calculated by measuring the hydrodynamic friction created by a bound molecule. The Y-
axis will be Normalized Dynamic Response Upward because it measures the speed at which the
nanolevers are being pushed away from the gold surface as a negative charge runs through the
surface of the biochip. The speed is measured by measuring the time taken for the nanolever to be
in an upright position by measuring relative fluorescence intensities. That speed is then normalized
to the speed of the analyte negative control group.
The second method utilizes FPS, which is the change in fluorescence intensity caused by
the binding of the analyte. Each analyte will interact differently with the fluorophores; therefore,
magnitude and the direction of change in fluorescence intensity is dependent on the bound
59

molecule. The Y-axis will be Normalized Fluorescence because the fluorescence intensity of the
interaction will be normalized to the analyte negative control.
5.3.1 U1hpII Attached to the Surface
For the kinetic analysis of the U1A/U1hpII interaction, an experiment using U1A as the
analyte was conducted. An important aspect of developing a method is making sure that it
consistently produces accurate results. This kinetic experiment was done with a stock 18 μM
U1hpII RNA (freshly made) and 17 μM U1A solution. The U1A dilutions were also freshly made
since protein is known to lose activity when stored diluted. The experiment used 5 U1A analyte
concentrations (690 pM, 2 nM, 6 nM, 19 nM, 56 nM) diluted in PE140. The rate the U1A analyte
is injected over the surface for these experiments is 50 μL/min for 5min.
The analysis begins where the X-axis is at 0 but there is recorded data before that time
point in figure 31. The recording of fluorescence intensity is measured as soon as the experiment
begins but it takes time for the U1A analyte solution to move through the fluidics system onto the
biochip. Once the analyte solution interacts with the ligand, a steep curve is seen on the association
curve, especially when injecting relatively high protein concentrations (pink line in figure 31A).
Figure 31. Association curves using U1A as the analyte. The experiment was conducted using a U1hpII ligand solution (18 μM)
with U1A analyte dilutions made immediately before the respective run. The U1A analyte solutions were run in the order of
highest to lowest concentration and analyzed in both A) dynamic response and B) fluorescence proximity sensing mode. The
concentration of the U1A analyte solution and its respective color on the graph is shown on the upper right-hand side of the
respective graph. The calculated association rates are above the respective graphs.
60

The analysis software requires the user to choose the correct time 0 which is chosen at the point
the exponential curve begins.
While noisy, the first set of curves (figure 31A) looks exactly as a good data set should,
with a nice distribution of concentrations and the highest protein concentrations reaching
saturation. The kinetic analysis in figure 31A yielded an association rate of ~7.7 x 10
6
M
-1
s
-1
,
which is quite close to the rate measured by SPR (9.4 x 10
6
M
-1
s
-1
, Law et al., 2005). Although the
kON rates calculated in figure 31 are very similar to previously published data using SPR
technology, they will need to be repeated once the instrument has been serviced.
5.4 Measurement of Dissociation Rates
The dissociation or kOFF rate is the rate at which protein dissociates from its target per
second. The dissociation occurs by injecting the analyte buffer over the surface of the biochip and
calculating the dissociation rate using the dynamic response or FPS analysis tool. The Y-axis is
the same as explained in section 5.3 except we would expect the change to occur in the opposite
direction from that observed with association.
5.4.1 U1hpII Attached to the Surface
The experiments shown in figure 32 are part of the same experiments conducted in section
5.3.1. Once all 5 analyte solutions in a given experiment are flown over the surface to measure
their association rates, the last analyte solution is kept on the surface of the biochip with analyte
buffer flown over the biochip surface to measure the dissociation rate. The rate the analyte buffer
is injected over the surface for these experiments is 100 μL/min for 47.5 min. Since it is known
that the U1A/U1hpII interaction is very stable (Katsamba et al., 2001), a long dissociation time is
needed to detect significant changes in fluorescence intensity or nanolever switching speed. Figure
32 shows the dissociation measured using the dynamic mode and the fluorescence mode. However,
61

as we saw in the association, the dynamic mode has too much noise, so the fluorescence mode was
used. The dissociation rate was ~5 x 10
-4
s
-1
which is in the range of the dissociation rate measured
by SPR (~3 x 10
-4
s
-1
, Law et al., 2005). Combined with the kON rate from figure 31A, the affinity
is ~60 pM, similar to that measured by SPR. As mentioned above, this experiment will need to be
repeated when the instrument has been serviced. In addition, the assay using biotinylated U1A
should be redone.
5.5 Discussion
Before the topic of accuracy comes into question, we must first be able to measure
precisely. Precision is based on consistently reproducing results. Even though we replicated
previously published kinetic rates, we were not able to reproduce these measurements. It seems
likely that this was due to instrument malfunction, as the curves do not look correct. Until the data
can be replicated, it is difficult to draw any firm conclusions.
Before optimizations can begin, every aspect of the DRX2 must be independently tested.
We must test the efficacy of the biochips and each flow channel in relation to length of use and
type of experiment. We must test for the variabilities between DRX2 machines. We much ensure
that the fluidics system is working properly along with the various sensors and detectors. The
testing will be done using a gold standard. This gold standard should be a protein that is relatively
Figure 32. Dissociation curves using U1A as the analyte. The experiment was conducted using a 690 pM U1A analyte solution
and analyzed in both A) dynamic response and B) fluorescence proximity sensing mode. The calculated disassociation rates are
above the respective graphs.
62

stable for a long duration (at least months) with a resilient target. This gold standard must go
through rigorous testing to ensure that the quantitation of both the protein and target are accurate,
and the protein retains its binding capabilities during testing. U1A is a great candidate for such a
gold standard because it can retain binding capabilities over a long duration and can bind to DNA
and RNA oligos. We would use a DNA oligo as the gold standard to avoid the degradation issues
associated with RNA.
I am currently testing the interaction between U1A and the DNA version of the U1hpII
RNA. I am using an electrophoretic mobility shift assay to approximate the KD of this interaction.
I have optimized the protocol to produce clear bands on a non-denaturing TBE-PAGE gel with
minimal amount of DNA per well. The data suggests that U1A binds to a DNA target with an
affinity of approximately 40 nM. Electrophoretic mobility shift assays are great for quickly
determining the viability of both the protein and its target and therefore can be used to test the
quality of the gold standard being used. Running a quality control check each step of the way can
substantially decrease the difficulty in identifying the problem. Once it is established that the
Figure 33. Gel-shift assay approximating the KD of the U1A/U1hpII DNA interaction. 15% TBE-PAGE gel was run with 20 μL of
solution per well which contains a final concentration of 250 nM U1hpII DNA and 0-1525 nM U1A protein (labeled below the
respective well) in PE140 buffer. The smaller band sizes represent free U1hpII DNA and the larger band sizes represent U1A bound
U1hpII DNA. Stained with ethidium bromide and imaged on the ChemiDoc™ XRS+ System.
63

DRX2 and components are in working order, we can continue developing different methods for
analyzing RNA/protein interactions.

64

Chapter 6: Future Directions
6.1 Introduction
After the development of a method for measuring kinetic rates of RNA/protein interactions
using switchSENSE
®
technology on the DRX2, we can use those methods to study the effects of
multiple hairpin loops interacting with multiple RNA recognition motifs. I have designed 5
different RNA sequences containing up to 5 hairpin loops and 9 different protein constructs
containing up to 3 RNA recognition motifs which can be used by our laboratory to test their effects
on binding.
6.2 RNAstructure
RNAstructure is a program developed by Dr. David H. Mathews from the University of
Rochester to predict the secondary structures of given RNA sequences. It calculates useful
parameters such as predicting the probability of a nucleotide being located in a certain position
and calculating the Gibbs free energy. Theoretically, the lower the Gibbs free energy is the more
stable the RNA structure is, but stability is also based on many other factors therefore experiences
will differ in vivo (Layton et al., 2005).
Figure 34 shows all 5 designs. 4 of the designs are based off of the U1A polyadenylation
inhibition element (PIE) mentioned in the introduction. The U1A PIE is very efficient at binding
to multiple U1A proteins within a very short sequence. As the number of binding loops increase,
the number of nucleotides needed to create a stable predictable RNA structure increases
exponentially. Complications increase with the length of RNA as well. One example is hybridizing
nanolevers with very long oligos; the oligos interact with each other rather than the protein of
interest. When neighboring oligos interact, they influence fluorophores’ activities which will affect
the measured kinetic parameters. For this reason, Dynamic Biosensors offers special biochips with
65

lower nanolever densities which increases the distance between each nanolever. Unfortunately, as
nanolever densities decrease, error rates increase because fewer interactions are being used to
calculate the kinetic parameters.
6.3 Protein Linkers
To attach multiple RNA recognition motifs subsequently, we needed a protein linker
between the motifs to allow each motif to independently interact with the hairpin loop. I have
already created U1A constructs with up to 3 U1hpII specific RNA recognition motifs which are
connected with various linkers. The first linker has just ‘AAA’ inserted between the motifs which
Figure 34. All 5 hairpin structures designed using RNAstructure. A) 1-hairpin loop (U1hpII) B) 2-hairpin loop C) 3-hairpin loop D)
4-hairpin loop and E) 5-hairpin loop RNA designs. Thick black lines signify hydrogen bonding.
66

was only inserted for cloning purposes but also creates a rigid linker that has been shown to lead
to more helical conformations and hydrogen bonding (Li et al., 2016). The next 2 linkers were
‘SGSGGAAA’ and ‘SGGGGSGGGGSAAA’. Linkers containing Ser and Gly are flexible
potentially allowing the protein to more readily interact with its enviroment (Chen et al., 2012).
Since it is unkown how far the hairpin loops will be from each other in solution, 2 different sizes
of linkers are being tested.
6.4 RNA Pull-down Assay
Once it is estabilished which combination of U1A construct and RNA sequence creates the
strongest interaction, it can be used to create an RNA pull-down assay. The assay will consist of
the U1A construct of choice being attached to magnetic beads either covalently or
biotin/streptavidin coupling. Then the beads will incubate with a cell lysate mixutre to capture the
RNA of choice, consisting of one or more U1hpII motif attached to the RNA of choice. The
purpose of this assay is to study the interaction between a specific RNA and all bound
biomolecules. The elution of the biomolecules will be achieved with a simple high salt solution,
preventing the denaturation of attached biomolecules. For example, long non-coding RNAs
involved with lung cancer are gaining popularity as potentiol biomarkers and targets for drug
therapy (Roth et al., 2016). This assay can be an effective tool to isolate those long non-coding
RNAs of interest to examine which oncogenic transcription factors they are interacting with.

67

References
Aligam, E. Mathews lab. https://rna.urmc.rochester.edu/RNAstructure.html
Amine-Reactive Crosslinker Chemistry. ThermoFisher Scientific.
https://www.thermofisher.com/us/en/home/life-science/protein-biology/protein-biology-
learning-center/protein-biology-resource-library/pierce-protein-methods/amine-reactive-
crosslinker-chemistry.html
Applications. https://www.dynamic-biosensors.com/switchsense/applications/
Avidin-Biotin Interaction. https://www.thermofisher.com/us/en/home/life-science/protein-
biology/protein-biology-learning-center/protein-biology-resource-library/pierce-protein-
methods/avidin-biotin-interaction.html
Avvakumova, S., & Prosperi, D. (2018). Streptavidin.
https://www.sciencedirect.com/topics/neuroscience/streptavidin
B-PER Bacterial Protein Extraction Reagent - Thermo Fisher Scientific.
https://www.thermofisher.com/order/catalog/product/78248
Baklanov, M. M., Golikova, L. N., & Malygin, E. G. (1996, September 15). Effect on DNA
transcription of nucleotide sequences upstream to T7 promoter.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC146148/
Biolabs, N. E. HiScribe™ T7 High Yield RNA Synthesis Kit.
https://www.neb.com/products/e2040-hiscribe-t7-high-yield-rna-synthesis-kit#Protocols,
Manuals & Usage_Manuals
Biolabs, N. E. RNA Loading Dye, (2X). https://www.neb.com/products/b0363-rna-loading-dye-
2x#Product Information
Biosensing Instrument. http://biosensingusa.com/
Blocquel, David, Li, Sheng, Wei, Daub, Ge. (2017, May 22). Alternative stable conformation
capable of protein misinteraction links tRNA synthetase to peripheral neuropathy.
https://academic.oup.com/nar/article/45/13/8091/3844761
Cardoso, J., Lima, C. D., Leal, T., Gradia, D. F., Fragoso, S. P., Goldenberg, S., Krieger, M. A.
Analysis of Proteasomal Proteolysis during the In Vitro Metacyclogenesis of Trypanosoma
cruzi. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0021027
Castro, E. D. Prosite. https://prosite.expasy.org/cgi-
bin/prosite/ScanView.cgi?scanfile=9451047055.scan.gz
CCMB80 Transformation Buffer. (1970, January 01).
http://cshprotocols.cshlp.org/content/2018/3/pdb.rec102707.short
68

Cheng, Z., Du, Z., Zhai, B., Yang, Z., & Zhang, T. (2018, November). U1 small nuclear RNA
overexpression implicates autophagic-lysosomal system associated with AD.
https://www.ncbi.nlm.nih.gov/pubmed/29395359
Christoph F. Kurat1, J. L., Dewald, Tsui1, K., Friesen1, H., Pinay, Nislow1, C., Fillingham6, J.
(1970, January 01). Restriction of histone gene transcription to S phase by phosphorylation
of a chromatin boundary protein. http://genesdev.cshlp.org/content/25/23/2489.long
ClearColi™ BL21(DE3) Electrocompetent Cells, Lucigen.
https://us.vwr.com/store/product/11182344/clearcolitm-bl21-de3-electrocompetent-cells-
lucigen
Cléry, A., Sohier, T. J., Welte, T., Langer, A., & Allain, F. H. (2017, April 15). SwitchSENSE:
A new technology to study protein-RNA interactions.
https://www.ncbi.nlm.nih.gov/pubmed/28286323
Compton, S., & Jones, S. (1985). Mechanism of dye response and interference in the Bradford
protein assay. Analytical Biochemistry. doi:151(2):369-74.
DNA RNA Hybridization. https://www.sciencedirect.com/topics/medicine-and-dentistry/dna-
rna-hybridization
Drescher, D. G., Ramakrishnan, N. A., & Drescher, M. J. (2009). Surface plasmon resonance
(SPR) analysis of binding interactions of proteins in inner-ear sensory epithelia.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2864718/
Dynabeads M-270 Streptavidin - Thermo Fisher Scientific.
https://www.thermofisher.com/order/catalog/product/65305
Edmonds, M., Abrams, R. (1960, April). Polynucleotide biosynthesis: Formation of a sequence
of adenylate units from adenosine triphosphate by an enzyme from thymus nuclei.
https://www.ncbi.nlm.nih.gov/pubmed/13819354
ELAVL4 ELAV like RNA binding protein 4 [Homo sapiens (human)] - Gene - NCBI.
https://www.ncbi.nlm.nih.gov/gene/1996
Embl. European Molecular Biology Laboratory.
https://www.embl.de/pepcore/pepcore_services/protein_purification/extraction_clarification/
centrifugation/
EMBL-EBI, I. InterPro. https://www.ebi.ac.uk/interpro/entry/IPR034409
EschColi_K12-tRNAs. http://lowelab.ucsc.edu/GtRNAdb/Esch_coli_K12/eschColi_K12-
tRNAs.fa
European Bioinformatics InstituteProtein Information ResourceSIB Swiss Institute of
Bioinformatics. (2019, February 13). U1 small nuclear ribonucleoprotein A.
https://www.uniprot.org/uniprot/P09012
69

European Bioinformatics InstituteProtein Information ResourceSIB Swiss Institute of
Bioinformatics. (2019, February 13). Bifunctional ligase/repressor BirA.
https://www.uniprot.org/uniprot/P06709
Fairhead, M., & Howarth, M. (2015). Site-specific biotinylation of purified proteins using BirA.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4304673/
Fried, M. G. (1989). Measurement of protein-DNA interaction parameters by electrophoresis
mobility shift assay.
https://www.ncbi.nlm.nih.gov/pubmed/2670548?dopt=Abstract&holding=npg
Fusion protein linkers: Property, design and functionality. (2012, September 29). Science Direct.
https://www.sciencedirect.com/science/article/pii/S0169409X12003006
G, W., Kastner, R, L., & MC, W. Department of Cellular Biochemistry, Max-Planck-Institute for
Biophysical Chemistry, Göttingen, Germany. http://europepmc.org/abstract/MED/21113136
GE Healthcare Life Science. ÄKTA pure chromatography system.
https://www.gelifesciences.com/en/au/shop/chromatography/chromatography-systems/akta-
pure-p-
05844?gclid=CjwKCAjw7MzkBRAGEiwAkOXexO25WX_YZxDqHa8CQTn5uRYqBQOO
lpE6y-gLYmMXp1fYH-inYX1pgRoCdVEQAvD_BwE
González‐Díaz, J. B., García‐Martín, A., García‐Martín, J. M., Cebollada, A., Armelles, G.,
Sepúlveda, B., Käll, M. (2008, January 14). Plasmonic Au/Co/Au Nanosandwiches with
Enhanced Magneto‐optical Activity.
https://onlinelibrary.wiley.com/doi/abs/10.1002/smll.200700594
Granqvist, N., Yliperttula, M., Välimäki, S., Pulkkinen, P., Tenhu, H., & Viitala, T. (2014,
March 18). Control of the morphology of lipid layers by substrate surface chemistry.
https://www.ncbi.nlm.nih.gov/pubmed/24564782
Grodberg, J., & Dunn, J. J. (1988, March). OmpT encodes the Escherichia coli outer membrane
protease that cleaves T7 RNA polymerase during purification.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC210899/
Guzman, Ghaemi, & M, G. Department of Biochemistry, and Department of Chemistry,
Department of Physics, Center for the Physics of Living Cells, and Center for Biophysics and
Quantitative Biology, University of Illinois , Urbana, Illinois 61801, United States.
http://europepmc.org/abstract/MED/25659967
HABA/Avidin Reagent H2153.
https://www.sigmaaldrich.com/catalog/product/sigma/h2153?lang=en®ion=US
High performance liquid chromatography (HPLC). http://hiq.linde-
gas.com/en/analytical_methods/liquid_chromatography/high_performance_liquid_chromatog
raphy.html

70

His-Tag Purification. http://www.bio-rad.com/featured/en/his-tag-purification.html
His-tagged Proteins–Production and Purification. Thermo Fisher Scientific.
https://www.thermofisher.com/us/en/home/life-science/protein-biology/protein-biology-
learning-center/protein-biology-resource-library/pierce-protein-methods/his-tagged-proteins-
production-purification.html
Holmberg, A., Blomstergren, A., Nord, O., Lukacs, M., Lundeberg, J., & Uhlén, M. (2005,
February). The biotin-streptavidin interaction can be reversibly broken using water at
elevated temperatures. https://www.ncbi.nlm.nih.gov/pubmed/15690449
How Does Surface Plasmon Resonance Work? http://biosensingusa.com/technologies/surface-
plasmon-resonance/surface-plasmon-resonance-work/
How to quality control check your RNA samples. (2015, June 17).
https://bitesizebio.com/13527/how-to-quality-control-check-your-rna-samples/
Howarth, M., Chinnapen, D. J., Gerrow, K., Dorrestein, P. C., Grandy, M. R., Kelleher, N. L.,
Ting, A. Y. (2006, April). A monovalent streptavidin with a single femtomolar biotin binding
site. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2576293/
Hyre, David E, et al. “Cooperative Hydrogen Bond Interactions in the Streptavidin-Biotin
System.” Protein Science: A Publication of the Protein Society, Cold Spring Harbor
Laboratory Press, Mar. 2006, www.ncbi.nlm.nih.gov/pmc/articles/PMC2249767/.
Ideue, T., Adachi, S., Naganuma, T., Tanigawa, A., Natsume, T., & Hirose, T. (2012, April 10).
U7 small nuclear ribonucleoprotein represses histone gene transcription in cell cycle-arrested
cells. https://www.pnas.org/content/109/15/5693
Ikeda, R. A., Lin, A. C., & Clarke, J. (1992, February 05). Initiation of transcription by T7 RNA
polymerase as its natural promoters. https://www.ncbi.nlm.nih.gov/pubmed/1733960
Johnson, M. (2019, January 24). Detergents: Triton X-100, Tween-20, and More.
https://www.labome.com/method/Detergents-Triton-X-100-Tween-20-and-More.html
Katsamba, P. S., Myszka, D. G., & Laird-Offringa, I. A. (2001, June 15). Two functionally
distinct steps mediate high affinity binding of U1A protein to U1 hairpin II RNA.
https://www.ncbi.nlm.nih.gov/pubmed/11297556
Kattah, N. H., Kattah, M. G., & Utz, P. J. (2010, January). The U1-snRNP complex: Structural
properties relating to autoimmune pathogenesis in rheumatic diseases.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3074261
Keall, R., Whitelaw, S., Pettitt, J., & Müller, B. (2007, June 14). Histone gene expression and
histone mRNA 3' end structure in Caenorhabditis elegans.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1924863/
Kimple, M. E., Brill, A. L., & Pasker, R. L. (2013, September 24). Overview of affinity tags for
protein purification. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4527311/
71

Klein Gunnewiek, J. M., Hussein, R. I., Van Aarssen, Y., Palacios, D., De Jong, R., Van
Venrooij, W. J., & Gunderson, S. I. (2000, March). Fourteen residues of the U1 snRNP-
specific U1A protein are required for homodimerization, cooperative RNA binding, and
inhibition of polyadenylation. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC110837/
Kozlowski, L. P. Isoelectric calculator. http://isoelectric.org/calculate.php
Krepl, Miroslav, Blatter, Markus, Antoine, Damberger, Jiri. (2017, May 13). Structural study of
the Fox-1 RRM protein hydration reveals a role for key water molecules in RRM-RNA
recognition. https://academic.oup.com/nar/article/45/13/8046/3823296
Krummel, D., Oubridge, C., Leung, A., Li, J., & Nagai, K. (2009). Crystal structure of human
spliceosomal U1 snRNP at 5.5 Å resolution (Vol. 458, Pages 475–480). Nature.
https://www.nature.com/articles/nature07851.
Laird-Offringa, I. A., & Belasco, J. G. (1995, December 05). Analysis of RNA-binding proteins
by in vitro genetic selection: Identification of an amino acid residue important for locking
U1A onto its RNA target. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC40502/
Law, M. J., Rice, A. J., Lin, P., & Laird-Offringa, I. A. (1970, January 01). The role of RNA
structure in the interaction of U1A protein with U1 hairpin II RNA.
https://rnajournal.cshlp.org/content/12/7/1168
Law, M. J., Chambers, E. J., Katsamba, P. S., Haworth, I. S., & Laird-Offringa, I. A. (2005, May
24). Kinetic analysis of the role of the tyrosine 13, phenylalanine 56 and glutamine 54
network in the U1A/U1 hairpin II interaction.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1140079/
Law, M. J., Rice, A. J., Lin, P., & Laird-Offringa, I. A. (2006, July). The role of RNA structure
in the interaction of U1A protein with U1 hairpin II RNA.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1484440/
Law, M. J., Linde, M. E., Chambers, E. J., Oubridge, C., Katsamba, P. S., Nilsson, L., Laird-
Offringa, I. A. (2006, January 10). The role of positively charged amino acids and
electrostatic interactions in the complex of U1A protein and U1 hairpin II RNA.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1326249/
Law, M. J., Lee, D. S., Lee, C. S., Anglim, P. P., Haworth, I. S., & Laird-Offringa, I. A. (2013,
August). The role of the C-terminal helix of U1A protein in the interaction with U1hpII
RNA. https://www.ncbi.nlm.nih.gov/pubmed/23703211
Layton. (2005, January 01). Statistical analysis of RNA folding algorithms through
thermodynamic parameter perturbation.
https://academic.oup.com/nar/article/33/2/519/2549105
Li, G., Huang, Z., & Chong. (2015, September 22). Construction of a linker library with widely
controllable flexibility for fusion protein design.
https://link.springer.com/article/10.1007/s00253-015-6985-3
72

Lu, Z. J., Turner, D. H., & Mathews, D. H. (2006, October). A set of nearest neighbor parameters
for predicting the enthalpy change of RNA secondary structure formation.
http://www.ncbi.nlm.nih.gov/pubmed/16982646
Lutz, C. S., McClain, M. T., Harley, J. B., & James, J. A. (2002). Anti-U1A monoclonal
antibodies recognize unique epitope targets of U1A which are involved in the binding of U1
RNA. https://www.ncbi.nlm.nih.gov/pubmed/12203842
Magdeldin, S. (2012, March). Affinity Chromatography: Principles and Applications. doi:
10.5772/39087.
Mathay, C., Yan, W., Chuaqui, R., Skubitz, A. P., Jeon, J., Fall, N., ISBER Biospecimen Science
Working Group. (2012, December). Short-term stability study of RNA at room temperature.
https://www.ncbi.nlm.nih.gov/pubmed/24845140
Monitoring of drug release kinetics from thin polymer films by multi-parametric surface plasmon
resonance. (2015, August 28).
https://www.sciencedirect.com/science/article/pii/S037851731530168X?via=ihub
Ni-NTA Agarose. https://cube-biotech.com/us/products/purification-resins/his-affinity-resins/ni-
nta-agarose/
Ni-NTA Agarose. https://www.qiagen.com/us/shop/sample-technologies/protein/expression-
purification-detection/ni-nta-agarose/#technicalspecification
Novagen's pET-3a-d vectors carry an N-terminal T7•Tag® sequence and BamH I cloning site.
March 2, 2019, from http://www.emdmillipore.com/US/en/product/pET-3d-DNA-
Novagen,EMD_BIO-69421
Oubridge, C., Ito, N., Evans, P. R., Teo, C. H., & Nagai, K. (1994, December 01). Crystal
structure at 1.92 A resolution of the RNA-binding domain of the U1A spliceosomal protein
complexed with an RNA hairpin.
https://www.ncbi.nlm.nih.gov/pubmed/7984237?dopt=Abstract
Park, S., Myszka, D. G., Yu, M., Littler, S. J., & Laird-Offringa, I. A. (2000, July). HuD RNA
recognition motifs play distinct roles in the formation of a stable complex with AU-rich
RNA. https://www.ncbi.nlm.nih.gov/pubmed/10848602
Park-Lee, S., Kim, S., & Laird-Offringa, I. A. (2003, October 10). Characterization of the
interaction between neuronal RNA-binding protein HuD and AU-rich RNA.
https://www.ncbi.nlm.nih.gov/pubmed/12900401
Phillips, C., Pachikara, N., & Gunderson, S. I. (2004, July). U1A inhibits cleavage at the
immunoglobulin M heavy-chain secretory poly(A) site by binding between the two
downstream GU-rich regions. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC434241/
Piekna-Przybylska, D., Liu, B., & Fournier, M. J. (2007). The U1 snRNA hairpin II as a RNA
affinity tag for selecting snoRNP complexes.
https://www.ncbi.nlm.nih.gov/pubmed/17673090
73

Pierce BCA Protein Assay Kit - Thermo Fisher Scientific.
https://www.thermofisher.com/order/catalog/product/23225
Protein Research - SPR Systems. https://www.gelifesciences.com/en/us/solutions/protein-
research/products-and-technologies/spr-systems
Protein Structure. https://www.particlesciences.com/news/technical-briefs/2009/protein-
structure.html
Protocol - Improve Research Reproducibility. https://bio-protocol.org/bio101/e44
Protocols Nuc-MW. http://www.encorbio.com/protocols/Nuc-MW.htm
Qiagen. Ni-NTA Superflow protein purification - Fast, high-yield His-tag technology.
https://www.qiagen.com/qdm/aw/pro/ni-nta?cmpid=QVenGANi-NTA&kwid= ni nta
beads&gclid=CjwKCAiA8OjjBRB4EiwAMZe6yweishsDxaUhak1SAajmpugzXTm7L2BVS
3gTMd06Ox8LRT_8FYDpmRoCoh4QAvD_BwE
Qiagen. The QIAexpressionist - (EN).
https://www.qiagen.com/dk/resources/resourcedetail?id=79ca2f7d-42fe-4d62-8676-
4cfa948c9435&lang=en
Reagents. https://www.dynamic-biosensors.com/reagents/#buffers-solutions
Rich, A. (2006, March 24). Discovery of the Hybrid Helix and the First DNA-RNA
Hybridization. http://www.jbc.org/content/281/12/7693
Rosano, G. L., & Ceccarelli, E. A. (2014, April 17). Recombinant protein expression in
Escherichia coli: Advances and challenges.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4029002/
Rosmalen, Krom, Merkx. (2017). Tuning the Flexibility of Glycine-Serine Linkers To Allow
Rational Design of Multidomain Proteins . Biochemistry. 56 (50), 6565-6574 doi:
10.1021/acs.biochem.7b00902
Roth, A., & Diederichs, S. (2016). Long Noncoding RNAs in Lung Cancer.
https://www.ncbi.nlm.nih.gov/pubmed/26037047
Rueda, F. O., Bista, M., Newton, M. D., Goeppert, A. U., Cuomo, M. E., Gordon, E., Taylor, B.
J. (2017, November 20). Mapping the sugar dependency for rational generation of a DNA-
RNA hybrid-guided Cas9 endonuclease. https://www.nature.com/articles/s41467-017-01732-
9
Sargo, C. R., Campani, G., Silva, G. G., Giordano, R. C., Da Silva, A. J., Zangirolami, T. C.,
Rocha, I. (2015). Salmonella typhimurium and Escherichia coli dissimilarity: Closely related
bacteria with distinct metabolic profiles. https://www.ncbi.nlm.nih.gov/pubmed/26097206
Schümperli, D., & Pillai, R. S. (2004, October). The special Sm core structure of the U7 snRNP:
Far-reaching significance of a small nuclear ribonucleoprotein.
https://www.ncbi.nlm.nih.gov/pubmed/15526162
74

SDS-PAGE Gel. https://www.ruf.rice.edu/~bioslabs/studies/sds-page/gellab2.html
Single-Molecule Fluorescence Spectroscopy of Photosynthetic Systems.
https://pubs.acs.org/doi/abs/10.1021/acs.chemrev.6b00195?journalCode=chreay
Software. https://www.dynamic-biosensors.com/software/
Stothard, P. Protein Molecular Weight. https://www.bioinformatics.org/sms/prot_mw.html
Stowell1, J. A., Wagstaff1, J. L., Hill1, C. H., McLaughlin1, S. H., & Stefan. James A.W.
Stowell. http://www.jbc.org/content/early/2018/04/25/jbc.RA118.002291
Streptavidin-coupled Dynabeads, Streptavidin Magnetic Beads. Thermo Fisher Scientific.
https://www.thermofisher.com/us/en/home/brands/product-brand/dynal/streptavidin-coupled-
dynabeads.html#selection
Surface Plasmon Resonance. https://www.sciencedirect.com/topics/biochemistry-genetics-and-
molecular-biology/surface-plasmon-resonance
SwitchSENSE®. https://www.dynamic-biosensors.com/switchsense/
This is NIM (Area I, III, V). https://www.nano-initiative-munich.de/
Tsuiji, H., Iguchi, Y., Furuya, A., Kataoka, A., Hatsuta, H., Atsuta, N., Yamanaka, K. (2013,
February). Spliceosome integrity is defective in the motor neuron diseases ALS and SMA.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3569639/
UV absorbance DNA quantitation. https://www.bmglabtech.com/uv-absorbance-dna-
quantitation/
Vagenende, V., Yap, M. G., & Trout, B. L. (2009, November 24). Mechanisms of protein
stabilization and prevention of protein aggregation by glycerol.
https://www.ncbi.nlm.nih.gov/pubmed/19817484/
Varani, L., Gunderson, S. I., Mattaj, I. W., Kay, L. E., Neuhaus, D., & Varani, G. (2000, April
01). The NMR structure of the 38 kDa U1A protein – PIE RNA complex reveals the basis of
cooperativity in regulation of polyadenylation by human U1A protein.
https://www.nature.com/articles/nsb0400_329
Vector Database. https://www.addgene.org/vector-database/2641/
Vo, M., Terrey, M., Lee, J. W., Roy, B., Moresco, J. J., Sun, L., Ackerman, S. L. (2018, May
16). ANKRD16 prevents neuron loss caused by an editing-defective tRNA synthetase.
https://www.nature.com/articles/s41586-018-0137-8
Webster, M. W., Chen, Y., Stowell, J. A., Alhusaini, N., Sweet, T., Graveley, B. R., . . .
Passmore, L. A. (2018, June 21). MRNA Deadenylation Is Coupled to Translation Rates by
the Differential Activities of Ccr4-Not Nucleases.
https://www.ncbi.nlm.nih.gov/pubmed/29932902
75

Weissman, D., Pardi, N., Muramatsu, H., & Karikó, K. (2013). HPLC purification of in vitro
transcribed long RNA. https://www.ncbi.nlm.nih.gov/pubmed/23296926
What is Suface Plasmon Resonance (SPR)? https://www.reichertspr.com/about/what-is-suface-
plasmon-resonance-spr/
What is Surface Plasmon Resonance? - Nicoya. (2019, March 15).
https://nicoyalife.com/blog/what-is-surface-plasmon-resonance/
What is switchSENSE®? How does it work? https://www.dynamic-biosensors.com/faqs/
Workman, Veith, & DJ, B. From the Departments of Molecular and Cellular Biochemistry and.
http://europepmc.org/abstract/MED/24362020
Zeba Spin Desalting Columns, 7K MWCO, 0.5 mL - Thermo Fisher Scientific.
https://www.thermofisher.com/order/catalog/product/89882
Zeng, S., & Dominique Baillargeat, H. H. (2014, February 18). Nanomaterials enhanced surface
plasmon resonance for biological and chemical sensing applications.
https://pubs.rsc.org/en/Content/ArticleLanding/2014/CS/c3cs60479a#!divAbstract

Abstract (if available)

Abstract The U1A protein is part of the U1 small nuclear ribonucleo protein complex (snRNP) involved in nuclear pre-mRNA splicing. U1A interacts with U1hairpin II (U1hpII) in the U1snRNA, specifically recognizing an AUUGCAC sequence motif in the loop part of the hairpin. Using surface plasmon resonance (SPR), the affinity of this interaction has been determined to be ∼30 pM. SPR has a mass-based readout and cannot detect conformational changes of molecules as they undergo binding interactions. A new technology (switchSENSE) has become available that uses both hydrodynamic friction and fluorescence as readouts for binding. This would be a very valuable tool to add to the repertoire of methods to analyze RNA/protein interactions. The binding surface consists of DNA nanolevers that carry a fluorescence group at their tip, and to which complementary DNA or RNA can by annealed to create a variety of specific binding surfaces. Here we propose to use the robust U1A/U1hpII interaction to develop a method for measuring kinetic rates of RNA binding proteins using switchSENSE® technology on a DRX2 instrument. We will examine the utility of attaching either the RNA target or the U1A protein to the biosensor chip surface. Focusing first on coating RNA, our preliminary studies indicate that we need to include a 4-nucleotide spacer between the RNA target and the part of the RNA that hybridizes to the nanolever. We are currently testing coating biotinylated U1A protein on streptavidin-carrying nanolevers to generate a protein surface. Once proper surfaces are established, we will test binding of injected RNA or protein, with the objective to achieve consistent kinetic measurements throughout. Future aims include studying the relationship between multiple linked RNA recognition motifs binding simultaneously to multiple linked RNA targets.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Analysis of ELAVL4 RNA interaction using a switchSENSE biosensor

PDF

The kinetic study of engineered MBD domain interactions with methylated DNA: insight into binding of methylated DNA by MBD2b

PDF

Protein-protein interaction and protein-lipid interaction of membrane proteins

PDF

Structure and kinetics of the Orb2 functional amyloid

PDF

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

PDF

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

PDF

Overexpression and interaction of Orb2 in S2 insect cells

PDF

Defining the functional region of LINC00261 in lung adenocarcinoma

PDF

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

PDF

Oligomer formation of functional amyloid protein - Orb2A

PDF

LINC00261 induces a G2/M cell cycle arrest and activation of the DNA damage response in lung adenocarcinoma

PDF

The folding, misfolding and refolding of membrane proteins and the design of a phosphatidylserine-specific membrane sensor

PDF

Establishing the non-coding RNA LINC00261 as a tumor suppressor in lung adenocarcinoma

PDF

Generation of an epigenetic toggle switch to test LINC00261 function on lung adenocarcinoma cellular response to the chemotherapeutics oxaliplatin and carboplatin

PDF

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

PDF

Developing peptide and antibody-mimetic ligands for the cell surface receptors β2AR and DC-SIGN

PDF

Structural and biochemical determinants of APOBEC1 substrate recognition and enzymatic function

PDF

Regulating functional amyloid formation: the promiscuous behavior of the Orb2A N-terminal amphipathic region

PDF

Structural characterization of the functional amyloid Orb2A using EPR spectroscopy

PDF

Characterization of a new chromobox protein 8 (CBX8) antagonist in a model of human colon cancer

Asset Metadata

Creator Sefiani, Arthur (author)

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

School Keck School of Medicine

Degree Master of Science

Degree Program Biochemistry and Molecular Medicine

Publication Date 06/13/2019

Defense Date 06/13/2019

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

Tag DRX2,kinetics,OAI-PMH Harvest,protein,RNA,switchSENSE,U1A

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Offringa, Ite (committee chair), Haworth, Ian (committee member), Siemer, Ansgar (committee member)

Creator Email artsef7@gmail.com,sefiani@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-171989

Unique identifier UC11660977

Identifier etd-SefianiArt-7472.pdf (filename),usctheses-c89-171989 (legacy record id)

Legacy Identifier etd-SefianiArt-7472.pdf

Dmrecord 171989

Document Type Thesis

Format application/pdf (imt)

Rights Sefiani, Arthur

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