Organization of motor proteins into functional micropatterns fabricated by a photoinduced Fenton reaction.

Full text

Page 1

Surface Patterning
DOI: 10.1002/anie.200904576
Organization of Motor Proteins into Functional Micropatterns
Fabricated by a Photoinduced Fenton Reaction**
Maniraj Bhagawati, Surajit Ghosh, Annett Reichel, Klaus Froehner, Thomas Surrey, and
Jacob Piehler*
The functional organization of proteins on solid supports is a
key prerequisite for the integration of the powerful capabil-
ities of biomolecules into miniaturized biomedical and
biotechnological devices.[1]Motor proteins are particularly
attractive building blocks for the construction of such devices.
Numerous approaches have been reported for the organiza-
tion of motor proteins into functional micro- and nano-
structures,[2–8]which have inspired the development of novel
bioanalytical devices.[9–13]For these purposes, techniques for
the functional organization of proteins on surfaces into
micrometer-andsubmicrometer-sized
required.
Despite substantial developments in this field, simple and
generic techniques for functional protein patterning are
scarce.[14]A critical prerequisite is that the functionality of
proteins immobilized on the surface must be fully maintained.
As many proteins denature upon interaction with solid
supports, surface modifications, for example, in the form of
thin protein-repellent polymer layers, are required to render
the surface biocompatible. Moreover, suitable, spatially
resolved functionalization of these layers is required for the
site-specific capturing of target molecules onto the surface.[14]
We previously developed multivalent head groups containing
nitrilotriacetic acid (NTA) moieties (such as tris-NTA,
Figure 1a) as generic, high-affinity adapters for oligohisti-
dine-tagged proteins.[15]These multivalent chelators have
proven powerful for stable, yet reversible protein binding in
solution and for immobilization onto various supports.[16–18]
Through theuse of suchchelators in combination with adense
poly(ethylene glycol) (PEG) polymer brush, we demon-
strated the oriented capturing of highly active kinesin on glass
surfaces.[19]Herein, we present a generic method for the
assembliesare
functional micropatterning of such surface architectures. This
method is based on selective photodestruction by a light-
induced Fenton reaction (Figure 1b), whereby suitable tran-
sition-metal ions are complexed by the immobilized NTA
moieties.
For the implementation of this approach, we used both
UV illumination through a mask and the UV laser of a
standard confocal microscope. The principle of the first
technique is depicted schematically in Figure 1c: After
loading of the NTA moieties with CoIIions, the surface is
illuminated with UV light through a mask. All metal ions are
then removed by washing with HCl or ethylenediaminetetra-
acetic acid (EDTA), and the remaining NTA groups are
loaded with NiIIions prior to protein binding.
The protein-binding efficiencies of tris-NTA-functional-
ized surfaces after UV illumination with the NTA moieties
loaded with different transition-metal ions are compared in
Figure 1b. Illumination in the presence of NiIIions did not
affect the binding capacity of the surface; in contrast, no
protein binding was observed after illumination of surfaces
loaded with CoII, CuI, or FeIIions. These three transition-
metal ions mediate photoinduced Fenton reactions,[20–22]
which are probably responsible for the destruction of the
NTA moieties on the surface. The same effect was observed
for surfaces functionalized with mono-NTA. A decrease in
the efficiency of photodestruction on surfaces was observed
when the length of the PEG chain was decreased (see the
Supporting Information). Thus, the surface architecture has
some influence on the destruction process.
To further characterize the photodestruction process, we
quantitatively assessed protein binding to surfaces after
illumination for shorter periods of time than required for
full destruction of the binding capacity. The unstable protein
binding observed under these conditions (see the Supporting
Information) indicated that the tris-NTA groups were parti-
ally destroyed and thus lost binding affinity. Moreover,
leaching of NiIIions was observed, which indicated that the
NTA moieties themselves were decomposed by the Fenton
reaction. This result is in line with studies carried out on the
hydroxyl-radical-mediated oxidation of chelating agents such
as NTA and EDTA. The oxidation of NTA yielded species
with a weaker metal-ion-coordination ability, such as imidodi-
acetic acid, glycolic acid, oxalic acid, and glycine.[23]Since the
active oxidant in the Fenton reaction is also a hydroxyl
radical, a similar oxidation pathway can be assumed. Thus, the
metal-ion-mediated photodestruction appears to selectively
eliminate the transition-metal-ion-binding moieties, but not
the protein-repelling PEG polymer brush. This conclusion is
[*] M. Bhagawati,[+]A. Reichel, Prof. Dr. J. Piehler
Institut f?r Biophysik, Universit?t Osnabr?ck
Barbarastrasse 11, 49076 Osnabr?ck (Germany)
Fax: (+49)541-9692262
E-mail: piehler@uos.de
Homepage: http://www.biologie.uni-osnabrueck.de/Biophysik/
Piehler/
Dr. S. Ghosh,[+]Dr. T. Surrey
Cell Biology and Biophysics Unit, EMBL Heidelberg (Germany)
K. Froehner
NB Technologies, Bremen (Germany)
[+] These authors contributed equally.
[**] We thank Mathias Utz and Gerhard Spatz-K?mbel for technical
assistance, and Christian Hentrich and Dr. Peter Bieling for advice
and reagents. This project was supported by the BMBF (0312034).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200904576.
Angewandte
Chemie
1
Angew. Chem. Int. Ed. 2009, 48, 1–5? 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
These are not the final page numbers! ??

Page 2

supported by the observationof negligible nonspecific protein
binding to surfaces after photodestruction (Figure 1b).
This approach turned out to be
promising for functional surface
micropatterning: after UV illumi-
nation through a mask, the selec-
tive and fully reversible binding of
fluorescence-labeled,
dine-tagged maltose binding pro-
tein (MBP-H10) to the areas that
had not been illuminated was
observed by confocal laser scan-
ning microscopy (Figure 1d). We
explored the specificity of protein
binding as well as the functionality
of the immobilized protein with a
protein–protein interaction assay
(Figure 1e). For this purpose, the
protein interferon a2 labeled with
Oregon Green 488 (OG488IFNa2,
without a His tag) was incubated
on a micropatterned surface with
and without prior incubation of its
receptor IFNAR2 fused to a dec-
ahistidine tag (IFNAR2-H10). The
binding ofOG488IFNa2 into micro-
patterns was observed only in the
presence of IFNAR2-H10. This
experiment thus confirmed the
highly specific protein binding as
well as the functional integrity of
proteins immobilized into such
micropatterns.
In the second approach, we
explored the in situ patterning of
tris-NTA-functionalized
with the 405 nm laser beam of a
confocal laser scanning micro-
scope (Figure 2a). We carried out
thephotodestruction
loaded tris-NTA
selected regions of interest with
the beam of the 405 nm laser
either in the
OG488MBP-H10 for focusing pur-
poses (Figure 2a) or without the
protein. After the removal of CoII,
the surfaces were loaded with NiII
ions, and the binding ofOG488MBP-
H10 was imaged. Selective photo-
destruction occurred in the illumi-
nated areas (Figure 2b). Photo-
destructionby
Fenton reaction was confirmed by
several control experiments (see
the Supporting Information). Var-
iation of the number of iteration
cycles during exposure to the
405 nm laser yielded different den-
decahisti-
surfaces
of
scanning
CoII-
by
presenceof
alight-induced
sities of binding sites (Figure 2c) and thus provided the
possibility to control the surface concentration of immobi-
Figure 1. Photodestruction of NTA-functionalized surfaces by a light-induced Fenton reaction. a) Struc-
ture of NiII-loaded tris-NTA attached to a PEG polymer brush. The free coordination sites for capturing
oligohistidine moieties are indicated by “X”. b) Binding of the His-tagged protein MBP-H10 to tris-
NTA surfaces after UV illumination in the presence of different transition-metal ions and subsequent
loading of the remaining NTA groups with NiIIions. Protein binding was detected by reflectance
interference. The mechanism of the light-induced Fenton reaction is also shown. c) Schematic
illustration of the patterning process: I) UV irradiation through a photomask of surfaces modified with
tris-NTA head groups, which have been loaded with CoIIions; II) tris-NTA head groups exposed to UV
irradiation are destroyed; III) remaining tris-NTA head groups can capture His-tagged proteins after
loading with NiIIions. d) Functional surface micropatterning by illumination through a mask. The
remaining tris-NTA moieties were loaded with fluorescence-labeled MBP-H10 and imaged by confocal
laser scanning microscopy. e) Binding of fluorescence-labeled IFNa2 to micropatterned surfaces with
(top) and without (bottom) prior incubation of its His-tagged receptor IFNAR2. The scale bar is 20 mm
in all images.
Communications
2
?
www.angewandte.org
These are not the final page numbers!
? 2009 Wiley-VCH Verlag GmbH & Co. KGaA, WeinheimAngew. Chem. Int. Ed. 2009, 48, 1–5
?

Page 3

lized proteins in a spatially resolved manner. The two
approaches for surface patterning can be combined readily,
as depicted in Figure 2d: Patterns created by illumination
through a mask were subsequently modified further by
confocal scanning with a 405 nm laser beam as described
above.
Next, we used our photochemical surface-patterning
method to immobilize molecular motors into microstructures
and tested whether they supported guided filament transport.
For this purpose, we used a recombinant dimeric kinesin with
a C-terminal decahistidine tag (kinesin-H10) that had been
shown before to support microtubule gliding on unpatterned
tris-NTA surfaces.[19]We coimmobilized decahistidine-tagged
green fluorescent protein (GFP-H10) as a fluorescent marker
to visualize the tris-NTA lines. Alexa568-labeled microtu-
bules landed selectively on the kinesin/GFP lines in random
orientations (Figure 3a,b). In the presence of adenosine
triphosphate (ATP), all microtubules in contact with the
lines were transported by the motors, thus confirming that
photochemical patterning followed by subsequent selective
tris-NTA-mediated motor immobilization yields highly active
protein patterns. Microtubules which had landed perpendic-
ularly to the line diffused into solution after having been
transported out of the line region.
Microtubules with a parallel orientation with respect to
the line were observed to glide along the line (Figure 3c; see
movie 1 in the Supporting Infor-
mation); in most cases they glided
distancesseveral
length (Figure 3c). Quantitative
analysis revealed that the distance
traveled on individual lines was
influencedby the characteristics of
the pattern. Microtubules were
observed to systematically travel
slightly longer distances on thin-
ner lines (Figure 3d). The very
small difference in the velocities
of the microtubules gliding on the
two types of lines studied (360?
60 nms?1on 1 mm wide lines and
370?60 nms?1
lines) confirmed that the motor
proteins on both lines had similar
densities.[19]
These observations
suggest that further reduction of
the thickness of these structures to
submicron dimensions could lead
to even higher microtubule-trans-
port fidelity.
In summary, we have devel-
oped a versatile method for func-
tional protein patterning. In con-
trast to conventional methods,
such as the microcontact printing
of proteins, our new method is
optimized for highly active protein
patterns, because the microstruc-
turing reaction takes place at the
timestheir
on5 mmwide
level of chemistry, and the biomolecules are captured under
physiological conditions. This approach introduces the flex-
ibility to separate patterning and protein immobilization and
thus reduce the danger of protein inactivation during
patterning. We have used this method to organize motor
proteins into highly functional microstructures. Motility
assays established that microtubule-transport characteristics
depend on geometrical features of the tracks that are defined
by the patterning process. The possibility to modify mask-
based protein structures by using the additional flexibility of
confocal laser scanning provides a means for generating
complex patterns, which will enable further exploration of the
foundations of microtubule transport by micropatterned
motor proteins. However, the generic application of this
technique for the functional micropatterning of proteins,
including recombinant antibodies or protein ligands for
localized cell-surface-receptor activation, is ensured by the
widespread use of the His tag for affinity purification.
Received: August 17, 2009
Published online: && &&, 2009
.
protein micropattern · protein–protein interaction
Keywords: immobilization · microarrays · molecular devices ·
Figure 2. In situ patterning of tris-NTA surfaces. a) Schematic illustration of the process: I) photo-
destruction of CoII-loaded tris-NTA (to which a protein may be bound); II) removal of CoIIions (and
the protein, if applicable) with imidazole and EDTA; III) loading of proteins onto intact areas after
loading with NiIIions. b) Lines of different width were scanned with the 405 nm laser in the presence
of NTA-complexed CoIIions. c) Protein binding to surfaces with rectangular structures scanned in the
presence of NTA-complexed CoIIions with the 405 nm laser for different numbers of iterations (from
top: 10, 50, 100, 200). d) Pattern obtained by illumination through a mask and further modification of
the square in the upper right corner by confocal laser scanning. Scale bar is 20 mm in all images.
Angewandte
Chemie
3
Angew. Chem. Int. Ed. 2009, 48, 1–5? 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
These are not the final page numbers! ??

Page 4

[1] M. G. van den Heuvel, C. Dekker, Science 2007, 317, 333.
[2] J. Clemmens, H. Hess, R. Doot, C. M. Matzke, G. D. Bachand, V.
Vogel, Lab Chip 2004, 4, 83.
[3] H.Hess, C. M.Matzke, R. K. Doot,J.Clemmens,G. D.Bachand,
B. C. Bunker, V. Vogel, Nano Lett. 2003, 3, 1651.
[4] J. Clemmens, H. Hess, C. Matzke, G. Bachand, B. Bunker, V.
Vogel, Biophys. J. 2003, 84, 293A.
[5] H. Hess, J. Clemmens, D. Qin, J. Howard, V. Vogel, Nano Lett.
2001, 1, 235.
[6] C. Reuther, L. Hajdo, R. Tucker, A. A. Kasprzak, S. Diez, Nano
Lett. 2006, 6, 2177.
[7] L. Ionov, M. Stamm, S. Diez, Nano Lett. 2006, 6, 1982.
[8] Y. M. Huang, M. Uppalapati, W. O. Hancock, T. N. Jackson,
Biomed. Microdevices 2007, 9, 175.
[9] T. Kim, L. J. Cheng, M. T. Kao, E. F. Hasselbrink, L. Guo, E.
Meyhofer, Lab Chip 2009, 9, 1282.
[10] L. Rios, G. D. Bachand, Lab Chip 2009, 9, 1005.
[11] T. Fischer, A. Agarwal, H. Hess, Nat. Nanotechnol. 2009, 4, 162.
[12] A. Goel, V. Vogel, Nat. Nanotechnol. 2008, 3, 465.
[13] C. T. Lin, M. T. Kao, K. Kurabayashi, E. Meyhofer, Nano Lett.
2008, 8, 1041.
[14] C. You, M. Bhagawati, A. Brecht, J. Piehler, Anal. Bioanal.
Chem. 2009, 393, 1563.
[15] S. Lata, A. Reichel, R. Brock, R. Tamp?, J. Piehler, J. Am. Chem.
Soc. 2005, 127, 10205.
[16] S. Lata, M. Gavutis, J. Piehler, J. Am. Chem. Soc. 2006, 128, 6.
[17] A. Tinazli, J. Tang, R. Valiokas, S. Picuric, S. Lata, J. Piehler, B.
Liedberg, R. Tampe, Chemistry 2005, 11, 5249.
[18] S. Lata, J. Piehler, Anal. Chem. 2005, 77, 1096.
[19] P. Bieling, I. A. Telley, J. Piehler, T. Surrey, EMBO Rep. 2008, 9,
1121.
[20] P. Ciesla, P. Kocot, P. Mytych, Z. Stasicka, J. Mol. Catal. A 2004,
224, 17.
[21] N. K. Urbanski, A. Beresewicz, Acta Biochim. Pol. 2000, 47, 951.
[22] S. Leonard, P. M. Gannett, Y. Rojanasakul, D. Schwegler-Berry,
V. Castranova, V. Vallyathan, X. Shi, J. Inorg. Biochem. 1998, 70,
239.
[23] C. Emilio, R. Gettar, M. Litter, J. Appl. Electrochem. 2005, 35,
733.
Figure 3. Transport of microtubules by kinesin-H10 immobilized selec-
tively on tris-NTA line patterns. a,b) Fluorescence images of Alexa568-
labeled microtubules (red) on kinesin-H10 captured on tris-NTA lines
(green) with widths of 5 mm (a) and 1 mm (b). c) Images from a time-
lapse movie showing a labeled microtubule (red) on a kinesin-H10–
tris-NTA line with a width of 1 mm (green) at the times indicated.
d) Percentage of microtubules (MTs) that traveled further than the
indicated number of times their average length along lines with widths
of 5 mm (red) and 1 mm (black). More than 85% and more than 90%
of the microtubules traveled more than three times (vertical line) their
average length along lines with widths of 5 mm and 1 mm, respectively.
Communications
4
?
www.angewandte.org
These are not the final page numbers!
? 2009 Wiley-VCH Verlag GmbH & Co. KGaA, WeinheimAngew. Chem. Int. Ed. 2009, 48, 1–5
?

Page 5

Communications
Surface Patterning
M. Bhagawati, S. Ghosh, A. Reichel,
K. Froehner, T. Surrey,
J. Piehler*
&&&&—&&&&
Organization of Motor Proteins into
Functional Micropatterns Fabricated by a
Photoinduced Fenton Reaction
Walking the line: Selective photodestruc-
tion of nitrilotriacetic acid moieties on a
poly(ethylene glycol) polymer brush by a
light-induced Fenton reaction enabled the
functional organization of motor proteins
into micropatterns (see schematic illus-
tration). Microtubules were selectively
captured on the structures, and adeno-
sine triphosphate dependent transport
along lines was observed.
Angewandte
Chemie
5
Angew. Chem. Int. Ed. 2009, 48, 1–5? 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
These are not the final page numbers! ??