With solid-phase supports for protein
immobilization procedures hindering some downstream analyses, magnetic
microparticles are starting to get their moment in the sun. In this
article, the author discusses a magnetic bead-based technique with the
potential to streamline peptide–protein pull-down.
| Aug
1, 2005 |
| By:
Hege
Skjellerudsveen |
| Pharmaceutical
Discovery |
|
Key to successful scientific research in any discipline is obtaining
sufficient sample material. In particular for pharmaceutical research,
protein isolation procedures often need to produce samples that have
high integrity and concentration. Ideally, such a process would be easy
to carry-out and, even more importantly, produce the sample in a format
that is readily applicable to the experiment at hand. An ingenious way
of achieving both sample isolation and practical starting material is to
extract and concentrate the material directly onto a solid-phase
support. Yet, for this to work, the immobilization platform must be very
stable and specific to the sample of interest. Dynal Biotech's Dynabeads®
are being used in a number of protein protocols more commonly associated
with agarose and resin based methods, such as chromatography and
immunoprecipitation. This article examines a peptide-protein pull-down
protocol, which is based on the work of Matthias Mann and coworkers at
the University of Southern Denmark (Odense, Denmark).
Protein Handling with Magnetic
Micro-particles
Protein isolation and manipulation is often achieved using resins and
slurries of molecules as column- or liquid-based solid-phase supports.
As described below, agarose-based supports can also be used for novel
peptide-protein pull-down experiments. These solid-phase technologies
have driven advances in protein isolation chemistries that can
facilitate a plethora of downstream processes, but the basic platforms
have many drawbacks. Chromatography columns are very slow to use, as
most rely on gravitational flow and, therefore, binding, washing and
elution steps can take many hours to complete for even simple
isolations. Agarose bead-based technologies allow in-solution isolation
but can generate significant background contamination in downstream
analysis due to non-specific binding and incomplete washing.
Centrifugation steps used during agarose-based methods are time
consuming and can generate greater sample loss due to incomplete
separation and hazy delineation of the fluid component.
With these solid-phase supports being a hindrance in some downstream
analyses, a cleaner and easier-to-use support based on magnetic
micro-particles have proven their worth. Dynal Biotech's (Oslo, Norway)
Dynabeads® is one example of such magnetic technology.
Careful selection of the correct bead and chemistry enables scientists
not only to perform complex procedures reliant on the solid-phase nature
of the beads, but also to isolate and concentrate cells, cell
components, proteins, DNA or RNA based on very strict interactions.
Dynabeads are super-paramagnetic (i.e., they have magnetic
properties only when in a magnetic field). Once the field is removed
they have no magnetic remanence and therefore are not attracted towards
non-magnetic sources. Furthermore, the uniform monosized beads have a
very slow sedimentation rate. Binding targets such as streptavidin can
be attached easily to the surface of Dynabeads and will isolate target
molecules with biotin-based tags. Mann and colleagues formally used
agarose-based beads, but have reported that they are now using Dynabeads
coupled with streptavidin to immobilize desthiobiotinylated synthesized
peptides used as 'bait' for novel protein pull-down experiments (1).
These experiments have shown that magnetic bead technology greatly
improved their novel peptide-protein pull-down screen.
Streamlining Peptide–Protein
Pull-down
Figure 1. Dynabeads®
MyOne™. These paramagnetic beads are 1µm in diameter and
demonstrate size and structure uniformity.
|
As mentioned above, a commonly
exploited binding interaction is that of biotin-streptavidin.
Streptavidin (a derivate of avidin) on the surface of a solid-phase
support, such as agarose beads or magnetic beads, tightly binds biotin
or desthiobiotin tags on recombinant or synthetic peptides (or DNA). For
agarose-based techniques, clarification and washing of the isolated
peptide is a time consuming process, which can lead to a loss of
integrity and sample. On a magnetic bead-based solid-phase though, the
bound complex can be separated quickly from the rest of the sample and
washed thoroughly by simply holding the beads in a magnetic field. The
resulting purified and concentrated sample has a high integrity and can
be used as the basis for a range of reactions, including those that are
enzymatically or thermally driven. Mann and coworkers have used
streptavidin conjugated Dynabeads (Dynabeads® MyOne™
Streptavidin C1 [Figure 1]) to immobilize desthiobiotinylated
synthesized peptides, with and without modified amino acid residues.
Once immobilized, they use these peptides as 'bait' for their
proprietary peptide-protein pull-down experiments to identify unique
binding interactions.
Figure 2. A simplified schematic
diagram of peptide–protein pull-down used by Mann and
coworkers (1). Two cell culture sources are differentially
labeled using 12C6-Arginine or the stable isotope 13C6-Arginine.
Once fully incorporated into the proteome, the cells are lysed
and contents applied to recombinant peptides based on the
phospho-tyrosine region of the ErbB3 receptor. Different
peptides are produced with different phosphorylation patterns.
Proteins bind to the peptides depending on their native
interaction. The recombinant peptides are held on to Dynabeads
MyOne Streptavidin C1 via a desthiobiotin tag. The Dynabeads act
like a solid-phase support during this process, but due to their
superparamagnetic nature also can be used to isolate the
peptides. Furthermore, this magnetic property of the Dynabeads
is used to separate the protein–peptide complexes formed after
the addition of the cell lysate from the rest of the cellular
components. These two peptide-protein interactions are then
released from the Dynabeads using biotin displacement and are
combined for trypsin digestion prior to LC–MS-MS. Due to the
differential labeling of the proteins from the two different
sources, mass spectrum peaks appear in pairs with intensities
relating to binding preferences (4).
|
Identical proteins from distinct cell
culture sources can be discriminated easily in downstream mass
spectrometry analyses using metabolic labeling (Figure 2). With one cell
culture left unlabeled, others can be labeled with stable-isotope tagged
amino acids such as 13C6-arginine or LeucineD3.
These will produce distinctive mass changes in spectra. Labeling in this
way has been described by the Mann laboratory and is known as SILAC –
stable isotope labeling by amino acids in cell culture (2, 3). Once
complete incorporation into the proteome has occurred, the cells
containing the labeled proteins are lysed and incubated with a synthetic
peptide of choice, and unlabeled lysates are incubated with a control
peptide.
To study the resultant
interactions, the synthetic peptides must be attached to a controllable
solid-phase. Although this is often achieved by using agarose-based
beads, they require considerable processing time and produce a distinct
background contamination during analysis. Recovering bound proteins from
these agarose-bound peptide bait molecules is also time-consuming and
labor-intensive. For example, boiling and sieve separation has to be
performed using 1-D sodium dodecyl sulfate polyacrylamide gel
electrophoresis (SDS-PAGE). Only then can sections of the SDS-PAGE gel
be cut up and treated with trypsin to produce protein fragments for mass
spectrometry analysis. Knowing these difficulties, Mann and his
coworkers started to incorporate Dynabeads MyOne Streptavidin C1 (Figure
1) into their protocols.
The initial isolation of the bait
peptides with Dynabeads does not require any centrifugation, thus
simplifying the procedure. Instead, the beads are drawn rapidly to the
side of the vessel by magnetic force, enabling any fluid to be removed
and replaced with wash buffer or reaction buffer. This ensures the beads
are immediately ready for addition of cell lysates. Once the proteins
within these lysates have had time to interact with the peptides, those
that have bound can be isolated in exactly the same way as the peptides
were immobilized — by applying a magnet and removing the reaction
buffer.
Using desthiobiotin in the
synthetic peptide also means that the peptide–protein interactions can
be recovered more easily from the beads for subsequent mixing and
trypsin digestion. Elution is achieved by the addition of biotin, which
binds to streptavidin more avidly than desthiobiotin and thus displaces
the peptide–protein complexes from the beads. In addition, performing
the trypsin digestion in the liquid phase results in much more efficient
and consistent fragmentation of proteins for mass spectrum analysis.
Applying the Solid-phase Technology
Mass spectrum-based proteomics is
useful for high-throughput screening of protein–protein interactions.
The flexibility of the screen means that processes such as cell
activation can be investigated relatively easily. Therefore, differences
in the protein–protein signalling pathways among different activation
states can be identified and may lead to the rapid identification of
novel drug targets. One of the most recent examples of such work from
Mann's group is the elucidation of phoshoryl group-dependent binding
interactions in the EGF-receptor pathway (1). For this work, the
researchers synthesized a series of peptides, which mirrored the phospho-tyrosine
region of the ErbB3 receptor, as below:
A) a non-phosphorylated 'control'
peptide –YEY-
B) a peptide phosphorylated at the
first tyrosine residue — 'active 1' –pYEY-
C) a peptide phosphorylated at the
second tyrosine residue — 'active 2' –YEpY-
D) a peptide phosphorylated at both
tyrosine residues — 'active 3' –pYEpY-
All four peptides were immobilized
using Dynabeads and a desthiobiotin tag. Three different cell cultures
were grown to at least five cell population doublings in three different
media:
Culture 1. Normal abundance amino
acids
Culture 2. Normal abundance amino
acids - 12C6Arginine replaced by – 13C6Arginine
Culture 3. Normal abundance amino
acids - 12C6Arginine replaced by - 13C615N4-Arginine
Table I. An outline of the
peptide–protein pull-down experiment parameters.
|
On lysis of the cell cultures, lysates
were added to the synthetic bait peptides according to the parameters
outlined in Table I. Bound proteins from the lysates were isolated using
paramagnetic separation and eluted from the Dynabeads MyOne Streptavidin
C1 using 16 mM of biotin. Once eluted, the labeled and unlabeled
fractions were mixed and digested with trypsin and subjected to in-depth
mass spectrometry analysis (Figure 2).
Figure 3. Mass spectra showing the
clarity produced by using Dynabeads® MyOne™:
Interactions with –YEY- motifs of ErbB3. Panel A – Filamin-A,
non-specific. Panel B – Grb2, specific for –YEpY- and –pYEpY-.
Panel C – PI3-Kinase, specific for –pYEY- and –pYEpY-.
|
From the mass spectra generated
(Figure 3), proteins with the same expression in each cell culture
appear as three peaks of equal intensity but with clearly different
masses, due to the isotope labels (Figure 3, panel A). Proteins that
bind specifically or more abundantly to the phosphorylated recombinant
peptides will appear with unequal peak heights, the intensity ratios
directly reflecting the molecular ratios (Figure 3, panels B and C).
Importantly, it is also clear that some proteins bind more specifically
to one or other of the phosphorylated forms.
Figure 3 summarizes the data, with
panel A showing that the cytoskeletal scaffold protein, Filamin-A, binds
non-specifically to the ErbB3 phosphotyrosine region with all peaks of
an equal intensity. Panel B shows that a signalling protein in the
tyrosine kinase pathway, Grb2, is specific for the –YEpY- form and
less so for the doubly phosphorylated –pYEpY-, but does not show any
significant binding to the –YEY- or –pYEY- forms. Panel C shows that
the cell signalling factor phosphatidylinositol 3-kinase (PI3-Kinase),
binds specifically to the –pYEY- and -pYEpY- forms with equal
intensity, but does not show any significant binding to the –YEY- and
–YEpY- forms.
Conclusions
Peptide–protein interaction
screens are an elegant way of elucidating modification-dependent
protein–protein interactions. The method is also flexible, allowing
multifactoral investigations in parallel. The number of samples and the
volume can be scaled up (or down) to make the process suitable for many
purposes, including high-throughput analysis.
For the work carried out by Mann's
group, Dynabeads MyOne Streptavidin C1 were used for the isolation of
desthiobiotin-tagged synthetic peptides. This not only allowed the
isolation and concentration of the peptides directly onto a solid-phase
support but also provided the capacity to isolate binding proteins
without any further processing steps. The procedural changes brought
about using the Dynabeads not only reduced the time required to complete
an experiment, but also significantly reduced background contamination
(4). This was achieved partly by removal of the gel electrophoresis
processing stage, but also the significant reduction in non-specific
binding, due to the nature of Dynabeads. This work shows that even
specialized procedures can benefit from the incorporation of magnetic
bead technologies.
Hege Skjellerudsveen is an
international product manager at Dynal Biotech in Oslo, Norway. She can
be reached at Tel. +47 2206 1211; e-mail hege.skjellerudsveen@dynalbiotech.com
References
1. W. Schulze, J. Olsen, L. Deng
and M. Mann, Profiling of peptide-protein interactions in signalling
pathways of the EGF-receptor family using the new LTQ-FT-ICR mass
spectrometer. Poster abstract at 52nd ASMS Conference in Nashville,
Tennessee, USA (2004).
2. S.E. Ong, B. Blagoev, I.
Kratchmarova et al., Mol. Cell. Proteomics 1, 376–386
(2002).
3. S.E. Ong, I. Kratchmarova and M.
Mann, J. Proteome Res. 2(2), 173–181 (2003).
4. M. Mann and W. Schulze, Personal
communication, 2005.
|
