Protein Reduction and Dialysis-Free Work-Up through Phosphines Immobilized on a Magnetic Support: TCEP-Functionalized Carbon-Coated Cobalt Nanoparticles

Article abstract of DOI:10.1002/chem.201701162

Tris(2-carboxyethyl)phosphine (TCEP) is an often-used reducing agent in biochemistry owing to its selectivity towards disulfide bonds. As TCEP causes undesired consecutive side reactions in various analytical methods (e.g., gel electrophoresis, protein labeling), it is usually removed by means of dialysis or gel filtration. Here, an alternative method of separation is presented, namely the immobilization of TCEP on magnetic nanoparticles. This magnetic reagent provides a simple and rapid approach to remove the reducing agent after successful reduction. A reduction capacity of 70 μmol per gram of particles was achieved by using surface-initiated atom transfer polymerization.

Full text of DOI:10.1002/chem.201701162

A Journal of
Accepted Article
Title: Protein Reduction and Dialysis-Free Work-Up Through
Phosphines Immobilized on a Magnetic Support: TCEP-
Functionalized Carbon-Coated Cobalt Nanoparticles
Authors: Adrian Zwyssig, Elia M Schneider, Martin Zeltner, Balder
Rebmann, Vladimir Zlateski, Robert N Grass, and Wendelin
Jan Stark
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201701162
Link to VoR: http://dx.doi.org/10.1002/chem.201701162
Supported by

10.1002/chem.201701162
Chemistry - A European Journal
COMMUNICATION
Protein Reduction and Dialysis-Free Work-Up Through
Phosphines Immobilized on a Magnetic Support: TCEP-
Functionalized Carbon-Coated Cobalt Nanoparticles
Adrian Zwyssig,[a] Elia M. Schneider,[a] Martin Zeltner,[a] Balder Rebmann,[b] Vladimir Zlateski,[a] Robert
N. Grass,[a] and Wendelin J. Stark*[a]
Tris(2-carboxyethyl)phosphine (TCEP) is an often used reducing
agent in biochemistry due to its selectivity towards disulfide
bonds. As TCEP causes undesired consecutive side reactions in
various analysis methods (i.e. gel electrophoresis, protein
immobilized on polyethylene glycol (PEG) beads, but has only
been shown effective for small peptides (e.g. mesethericin
3.8 kDa) and under microwave irradiation (at potentially
problematic higher temperatures).[8] The commercially available
labeling), it is usually removed via dialysis or gel filtration. Herein, TCEP bound on agarose Gel (Pierce™ Immobilized TCEP
an alternative method of separation is presented, namely the
immobilization of TCEP on magnetic nanoparticles. This
magnetic reagent provides a simple and rapid approach to
Disulfide Reducing Gel)[9] has recently been introduced as a
promising reducing agent also for larger proteins (at room
temperature, using excess amounts of reducing agent). However,
the loading capacity was relatively low (> 8 μmol mL-1) and the
necessity of using centrifuges or spin-columns resulted in loss of
remove the reducing agent after successful reduction.
A
reduction capacity of 70 μmol per gramm of particles was
achieved by using surface-initiated atom transfer polymerization. sample (especially positively charged proteins)[10] or dilution of the
sample due to washing of the gel. Here, we investigate an
alternative support, based on magnetic nanoparticles. They
exhibit a beneficial surface-to-volume ratio, and less diffusion
limitations than resins (e.g. polyethylene based resins). The
Polyacrylamide gel electrophoresis (PAGE) is a frequently used
analytic method in biochemistry. For the majority of those PAGE-
type methods, bio-molecules (e.g. proteins) have to be denatured
magnetism permits simple separation by application of
a
or unfolded. For denaturation of proteins sodium dodecyl sulfate
(SDS) or lithium dodecyl sulfate (LDS) are often used (especially
for denaturation of secondary and non-disulfide-containing
tertiary structures).[1] Trialkylphosphines on the other hand are
highly effective agents for the selective reduction of disulfide
bonds in proteins, peptides and other disulfide bond-containing
molecules.[2] The solubility issue of alkylphosphines has been
tackled by the development of tris(2-carboxyethyl)phosphine
(TCEP) which is active over a broad pH-range (pH 1.5 – 8.5).[3]
TCEP is stable in aqueous solutions and is not prone to rapid
oxidation that often occurs with other common disulfide reducing
agents, such as dithiotreitol (DTT) and β-mercaptoethanol
(BME).[2a] Residual TCEP causes loss of focus in two-dimensional
electrophoresis[4] and interferes with protein labeling and
modification agents (e.g. maleimides and iodoacetic acid) under
conditions relevant for protein chemistry.[5] Additionally, as TCEP
can lead to cleavage of the protein backbone (in case of exposure
over prolonged time),[6] the demand for reliable removal methods
of TCEP before subsequent modification reactions (e.g. labeling,
bio-conjugate crosslinking) is high.[7] Dialysis, the conventional
TCEP removal procedure, is time-consuming, has to be
performed under non-oxidating conditions and causes loss of
valuable proteins. Therefore, efforts for alternative solid support
removal methods were considered: TCEP has previously been
commercial permanent magnet (see Fig. 5).[11] Therefore, this
kind of supporting material avoids common work-up procedures
(e.g. filtration, centrifugation), resulting in a separation step that is
faster overall and does not result in loss or dilution of the sample.
Thus, combination of the reducing capacities of TCEP with the
attributes of functionalized magnetic nanoparticles (magnetic
separation[12] and high possible loading capacity[13]) offers a
substantial improvement in the field of immobilized reducing
agents.
Herein, we present the synthesis and use of a magnetically
recoverable TCEP reagent (Scheme 1). Carbon-coated magnetic
nanoparticles (e.g. graphene-coated cobalt nanoparticles (C/Co))
have already been modified with numerous desired moieties such
[a]
[b]
A. Zwyssig, E. Schneider, M. Zeltner, V. Zlateski, R. Grass, W. Stark
D-CHAB, ICB, ETH Zurich,
Vladimir-Prelog-Weg 1, 8093 Zurich, Switzerland,
E-mail: wendelin.stark@chem.ethz.ch
Scheme 1. Magnetic tris(2-carboxyethyl)phosphine (TCEP) for a facile
reduction of disulfide bonds in biomolecules is presented. This magnetic
reagent enables site selective modification or analysis of the biomolecule
without tedious reagent removal procedures.
B. Rebmann
Faculty of Biology, and Centre for Biological Signalling Studies,
University of Freiburg,
Schaenzlestrasse 18, 79104 Freiburg, Germany
Supporting information for this article can be found under:
http://xxx
This article is protected by copyright. All rights reserved.

10.1002/chem.201701162
Chemistry - A European Journal
COMMUNICATION
Figure 1. Synthetic route for the production of magnetic tris(2-carboxyethyl)phosphine (TCEP). The Initiator is attached via diazonium chemistry, followed by
esterification (C/Co@Initiator 1, n1= 0.1 mmol g-1 particles). Poly(ethylene glycol) methacrylate is polymerized using surface-initiated ATRP (C/Co@PEGMA 2,
n2 = 6 PEGMA units on average). TCEP is linked via a modified Steglich esterification (C/Co@PEGMA-TCEP 3). The successful modification was monitored
by microanalysis (bottom) and transmission electron microscopy (middle). The dispersion stability in water changes during modification due to the surface
alternation (pictures of capped vials).
as polymers, catalysts, ligands, chelating agents or enzymes.[14]
The graphene-like coating protects the magnetic metal core from
oxidation and allows chemical modification of the surface (i.e.
covalent attachment of various chemical functional groups).[15] If
these particles are applied in aqueous solution, the modification
of the hydrophobic graphene is necessary in bifocal perspective.
Turning the intrinsically hydrophobic surface into a hydrophilic
surface leads to improved dispersion stability of the particles in
aqueous solutions, hence optimal exploitation of the large surface
area.[16] Furthermore, the hydrophobic graphene surface is prone
to non-selective physisorption of complex, hydrophobic structures
(bio-fouling),[17] that is tremendously undesired in most
biochemical uses. Altering the surface properties to avoid bio-
fouling (i.e. unspecific adsorption of proteins) is an irrevocable
requirement of such a supporting material.[18] Well-established
neutral non-biofouling polymer brushes comprise 2-hydroxyethyl
methacrylate (HEMA) or poly(ethylene glycol) methacrylate
(PEGMA).[19] Therefore, the initiator (1, Figure 1) and the
poly(ethylene glycol) methacrylate (PEGMA) brushes were
covalently attached to the graphene-like surface as described in
previous work via surface-initiated atom transfer radical
moiety was attached by a modified Steglich esterification reaction
with the aid of 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide
hydrochloride (EDC) and N,N-dimethylpyridine-4-amine (DMAP),
where only one carboxylic group of the TCEP is activated and
bound to 2, yielding TCEP-bearing magnetic nanoparticles (3,
Figure 1). Cross-linking of more than one carboxylic group had to
be avoided, as this negatively influenced the dispersion stability
of the particles (see Figure S1 and S2 in the ESI). For detailed
Figure 2. Non-biofouling control using tetramethylrhodamine attached to
albumin from bovine serum (BSA, 0.007 mg of labeled protein with 0.7 mg
of particles in 1 ml water). Unfunctionalized C/Co particles 1 (max. loading
0.03 mg protein mg-1 particles): blue circles, magnetic reduction agent 3: red
triangles.
polymerization (SI-ATRP),[16] providing magnetic nanoparticles
20]
with non-biofouling properties.[12a,
After polymerization, the
graphene layers were still intact (see Figure S3). The TCEP-
This article is protected by copyright. All rights reserved.

10.1002/chem.201701162
Chemistry - A European Journal
COMMUNICATION
the polymer shell is visible (Figure 1, middle). Dynamic light
scattering (DLS) and Lumisizer particle-size distribution
measurement C/Co@PEGMA 2 resulted in an average particle
diameter of 558 nm (Figure S11) and 466 nm (Table S1),
respectively. Compared to the TEM micrographs, the difference
in size is most likely due to agglomeration of particles and the
hydrophilic nature of the polymer brushes, which increase the
hydrothermal radius in aqueous solution substantially. The
polymerization led to a significant increment of the carbon content
(ΔC = 13.16 %), which translates to an average of six PEGMA
units per starter moiety. The phosphor content was then
increased by attachment of TCEP (from 0.04 % for 2 up to 1.23 %
for 3). Thermogravimetric analysis before and after
polymerization (Figure S12) confirmed successful polymerization.
Generally, a polymer content of between 12 to 20 [C] % on the
particles proved to be ideal for optimal dispersion-stability, non-
biofouling properties and short separation time. The non-
biofouling property of the final magnetic reagent 3 was tested via
adsorption of commercially available tetramethylrhodamine
labeled albumin from bovine serum (0.007 mg of labeled protein
with 0.7 mg of particles in 1 ml water). 6.7 % of protein adsorbed
after 8 days stirring with 3 (Figure 2), while non-modified particles
already adsorb 95 % after 1 hour. Besides the proper
characterization of the material (see ESI for full characterization),
the performance of magnetic TCEP (i.e. reducing activity) was of
eminent importance. Therefore, a standard performance test was
chosen, namely the cleavage of dithionitrobenzoic acid (DTNB;
Ellman’s reagent), which can be observed by UV/VIS-
spectroscopy (Figure 3a).[21] For this test, the nearly transparent
DTNB (5 mM) was dissolved in Tris-HCl buffer and 0.1 mg of
C/Co@PEGMA-TCEP was added and dispersed by
ultrasonication for seven minutes at room temperature. After
separation of magnetic TCEP, the newly generated yellow
nitrothiobenzoic acid (NTB) was detected by measuring the
absorbance at 412 nm. The test showed a discrepancy between
the theoretical loading, which was calculated by phosphor content,
and the reduction capacity towards DTNB (400 μmol g-1 particles
Figure 3. a) Reduction of dithionitrobenzoic acid to nitrothiobenzoic acid to
determine the reducing capacity of nanoparticles C/Co@PEGMA-TCEP 3.
b) Washing protocol to ensure covalent connection of TCEP and the
magnetic support. No DTNB reduction occurred in the supernatant of buffer
exposed to the magnetic reagent (control experiments B and C), while the
reagent 3 itself reduces DTNB in 5 minutes (positive control).
experimental conditions concerning the synthesis, see supporting
information. The successful modification of the carbon-coated
magnetic nanoparticles was monitored by elemental
microanalysis (EMA, Figure 1, bottom), infrared spectroscopy
(see Figure S4 and S5 in the ESI) and transmission electron
microscopy (TEM), where for PEGMA modified nanoparticles 2
Figure 4. a) Reduction of bovine insulin using an excess of reducing agent (3-fold) and prolonged reaction times (17 hours) at 60°C. UV-spectra of LC-MS
measurement of the reduction using C/Co@PEGMA-TCEP 3 (gray line), TCEP (black dashes) and no reducing agent (grey dots). A and B-chains were identified
using MS. b) Reductive cleavage of human IgG Fab fragment; Lithium dodecyl sulfate polyacrylamide gel electrophoresis (LDS PAGE). L = Mark12TM unstained
standard, A = Fab fragment from human IgG, B = Fab + C/Co@PEGMA, C = Fab + 720-fold excess of C/Co@PEGMA-TCEP, D = Fab + TCEP 470-fold excess,
E = Fab + Agarose-TCEP 942-fold excess.
This article is protected by copyright. All rights reserved.

10.1002/chem.201701162
Chemistry - A European Journal
COMMUNICATION
selective coupling. In order to set the magnetic TCEP in a
comparative context, an already applied and commercial
available bound TCEP (Pierce™ Immobilized TCEP Disulfide
Reducing Gel, loading capacity of > 8 μmol mL-1)[9] was also used
in this experimental series (column E). The agarose bound TCEP
was suspended in the protein/buffer solution. After the reaction,
the sample had to be centrifuged and decanted or centrifuged
using
a spin column. The chemical performance of this
commercial available TCEP is comparable to the magnetic TCEP.
However, in contrast to the agarose species, magnetic TCEP
could be separated within seconds by application of a permanent
magnet (Figure 5), thus minimizing the time dependent
reoxidation in air. Also, it should be noted that there is always loss
of protein when conducting centrifugation or dialysis (after
reduction with agarose TCEP), especially if no buffer is used (40-
60 % loss compared to <10% with C/Co@PEGMA-TCEP).[10]
In conclusion, a promising and fast magnetic alternative to
common and commercial methods for disulfide reduction is
presented. Magnetic nanoparticles were modified with a non-
biofouling PEGMA layer and TCEP. Performance of the TCEP
functionalized particles was successfully tested for bovine insulin
and the Fab fragment from human IgG. The former resulted in full
reduction (and partial reoxidation after removal of the reducing
agent) while the latter resulted in a partially reduction of the Fab
fragment. Probable future applications of this magnetic reagent
will be mainly in the field of partial reduction, where sterically
hindered reducing agents are essential (e.g. site-specific
modifications of antibodies).
Figure 5. Magnetic tris(2-carboxyethyl)phosphine 3 (left, black) and agarose
tris(2-carboxyethyl)phosphine (right, white) in dispersion respectively
suspension (a). Fast separation (5 s) from the solution by use of a permanent
magnet for 3 and centrifugation (1 min at 10k RPM & 5 min handling) for the
agarose gel (b).
calculated via P-content versus 70 μmol g-1 particles via reduction
capacity). Thus, it could be estimated that a part of the
immobilized TCEP was inactive due to oxidation or inaccessible
due to agglomeration of the magnetic particles. However, 17.5 %
of the calculated TCEP was active and remained active if
appropriately stored (-20°C in EtOH under argon atmosphere,
Figure S6, see ESI). Within this test series, it could also be shown
that the non-biofouling layer is of huge importance, as TCEP
directly attached to the carbon surface of C/Co (C/Co@TCEP)
had no visible activity (see ESI, -1 μmol g-1) at all because the
hydrophobic surface immediately adsorbs the generated products.
Covalent linking of TCEP to the magnetic support was proven by
using an intensive washing protocol (Figure 3b). If TCEP was only
physisorbed, one could detect reduced DTNB in samples B and
C. However, not even trace amounts of free, unbound TCEP were
detected. Now, the target proteins for PAGE exhibit far more
complex structures than the small organic molecules such as
Ellman’s reagent (DTNB). For this reason, the magnetic TCEP
was applied on more relevant substrates, namely on bovine
insulin (5.8 kDa) and on the Fab fragment from human IgG (55
kDa). After optimization of the reduction conditions concerning the
use of magnetic TCEP, bovine insulin could successfully be
reduced using equivalent amounts of magnetic reducing agent
(observed by LCMS, Figure 4a). After removal of the magnetic
reagent partial intramolecular reoxidation occurred (possibly
during the HPLC measurement) and several peaks corresponding
to the A- and B-chain of insulin were detected (Figure 4a),
implying different chain conformations. However, the use of
unbound TCEP resulted in similar spectra. For a more detailed
explanation of the subject, see ESI, Figure S7-S10. An even more
challenging task involved the reduction of the more complex Fab
fragment from human IgG, where for example the free thiol groups
can be used for sophisticated antibody modification.[22] The
performance within this setup was observed by LDS-Page
(Figure 4b). As expected, C/Co@PEGMA without attached TCEP
did not show any reducing capacity (Column B). Free TCEP fully
reduced the protein (column D), while magnetic TCEP partially
reduced the disulfide bonds of the Fab fragment from human IgG
(column C). However, the magnetic approach showed promising
potential, especially since partial reduction due to sterical
hindrance is an important feature in biochemistry to obtain site
Acknowledgements
The authors would like to thank Dr. Frank Krumeich for TEM
micrographs and Anna Beltzung for DLS measurements.
Keywords: nanoparticles • magnetic • reduction • separation •
TCEP
[1]
[2]
[3]
a) K. Weber, M. Osborn, J. Biol. Chem. 1969, 244, 4406-4412; b) U. K.
Laemmli, Nature 1970, 227, 680-685.
a) J. C. Han, G. Y. Han, Anal. Biochem. 1994, 220, 5-10; b) T. L. Kirley,
Anal. Biochem. 1989, 180, 231-236.
a) J. A. Burns, J. C. Butler, J. Moran, G. M. Whitesides, J. Org. Chem.
1991, 56, 2648-2650; b) M. E. Levison, A. S. Josephson, D. M.
Kirschenbaum, Experientia 1969, 25, 126.
[4]
[5]
[6]
G. A. P. Khoudoli, I.M.; Blow, J.J.; Swedlow, J.R., Proteome Sci. 2004,
2, 1-12.
E. B. Getz, M. Xiao, T. Chakrabarty, R. Cooke, P. R. Selvin, Anal.
Biochem. 1999, 273, 73-80.
P. R. Liu, B. W. O'Mara, B. M. Warrack, W. Wu, Y. P. Huang, Y. H. Zhang,
R. L. Zhao, M. Lin, M. S. Ackerman, P. K. Hocknell, G. D. Chen, L. Tao,
S. Rieble, J. Wang, D. B. Wang-Iverson, A. A. Tymiak, M. J. Grace, R. J.
Russell, J. Am. Soc. Mass Spectrom. 2010, 21, 837-844.
D. E. Shafer, J. K. Inman, A. Lees, Anal. Biochem. 2000, 282, 161-164.
G. Miralles, P. Verdie, K. Puget, A. Maurras, J. Martinez, G. Subra, ACS
Comb. Sci. 2013, 15, 169-173.
[7]
[8]
[9]
For additional information of TCEP@Agarose see:
https://tools.thermofisher.com/content/sfs/manuals/MAN0011439_Immo
bil_TCEP_Disulfide_Reducing_Gel_UG.pdf.
This article is protected by copyright. All rights reserved.

10.1002/chem.201701162
Chemistry - A European Journal
COMMUNICATION
[10] S. H. N. Lisa C. Shriver-Lake, and Chris Rowe Taitt, BioTechniques 2013,
55, 292 -294.
[15] R. N. Grass, E. K. Athanassiou, W. J. Stark, Angew. Chem.-Int. Edit.
2007, 46, 4909-4912.
[11] A. Schatz, O. Reiser, W. J. Stark, Chem.-Eur. J. 2010, 16, 8950-8967.
[12] a) E. M. Schneider, M. Zeltner, V. Zlateski, R. N. Grass, W. J. Stark,
Chem. Commun. 2016, 52, 938-941; b) E. M. Schneider, R. A. Raso, C.
J. Hofer, M. Zeltner, R. D. Stettler, S. C. Hess, R. N. Grass, W. J. Stark,
J. Org. Chem. 2014, 79, 10908-10915.
[16] M. Zeltner, R. N. Grass, A. Schaetz, S. B. Bubenhofer, N. A. Luechinger,
W. J. Stark, J. Mater. Chem. 2012, 22, 12064-12071.
[17] Y. Liu, X. Dong, P. Chen, Chem. Soc. Rev. 2012, 41, 2283-2307.
[18] I. Banerjee, R. C. Pangule, R. S. Kane, Adv. Mater. 2011, 23, 690-718.
[19] R. Barbey, L. Lavanant, D. Paripovic, N. Schuwer, C. Sugnaux, S. Tugulu,
H. A. Klok, Chem. Rev. 2009, 109, 5437-5527.
[13] C. S. Gill, W. Long, C. W. Jones, Catal. Lett. 2009, 131, 425-431.
[14] a) S. Fernandes, C. M. Eichenseer, P. Kreitmeier, J. Rewitzer, V. Zlateski,
R. N. Grass, W. J. Stark, O. Reiser, RCS Adv. 2015, 5, 46430-46436; b)
C. J. Hofer, V. Zlateski, P. R. Stoessel, D. Paunescu, E. M. Schneider,
R. N. Grass, M. Zeltner, W. J. Stark, Chem. Commun. 2015, 51, 1826-
1829; c) M. Zeltner, A. Schatz, M. L. Hefti, W. J. Stark, J. Mater. Chem.
2011, 21, 2991-2996; d) V. Zlateski, R. Fuhrer, F. M. Koehler, S. Wharry,
M. Zeltner, W. J. Stark, T. S. Moody, R. N. Grass, Bioconjugate Chem.
2014, 25, 677-684.
[20] K. Matyjaszewski, H. Dong, W. Jakubowski, J. Pietrasik, A. Kusumo,
Langmuir 2007, 23, 4528-4531.
[21] a) C. K. Riener, G. Kada, H. J. Gruber, Anal. Bioanal. Chem. 2002, 373,
266-276; b) P. D. Tzanavaras, C. Mitani, A. Anthemidis, D. G. Themelis,
Talanta 2012, 96, 21-25.
[22] A. Maruani, M. E. B. Smith, E. Miranda, K. A. Chester, V. Chudasama,
S. Caddick, Nat. Commun. 2015, 6, 6645.
This article is protected by copyright. All rights reserved.

10.1002/chem.201701162
Chemistry - A European Journal
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
Layout 1:
COMMUNICATION
Text for Table of Contents
Author(s), Corresponding Author(s)*
Page No. – Page No.
Title
((Insert TOC Graphic here))
Layout 2:
COMMUNICATION
Author(s), Corresponding Author(s)*
((Insert TOC Graphic here))
Page No. – Page No.
Title
This article is protected by copyright. All rights reserved.