Preservation of antibody selectivity on graphene by conjugation to a tripod monolayer

Article abstract of DOI:10.1002/anie.201209149

Three is a magic number: Antibodies adsorbed on bare graphene lose their ability to recognize antigens. However, this obstacle can be overcome by the conjugation of these antibodies to self-assembled monolayers of a tripodal compound (see picture).

Full text of DOI:10.1002/anie.201209149

Angewandte
Chemie
DOI: 10.1002/anie.201209149
Graphene Biofunctionalization
Preservation of Antibody Selectivity on Graphene by Conjugation to
a Tripod Monolayer**
Jason A. Mann, Thomas Alava, Harold G. Craighead,* and William R. Dichtel*
Interfacing biomolecules and cells to electronic materials is
a longstanding challenge to manufacture sensitive biosensors
and inexpensive diagnostic devices.[1] Single-layer graphene
(SLG) has emerged as a promising material for this purpose
because of its transparency, conductivity, robust mechanical
properties, and high surface area.[2] Recent developments in
its synthesis using chemical-vapor-deposition methods[3] pro-
vide large-area, high-quality SLG that is easily transferred to
a variety of substrates. For example, SLG that was supported
on reconstituted silk formed the basis of a wireless biosensor
that was adsorbed onto tooth enamel and detected pathogens
in saliva.[4] The function and specificity of such biosensors rely
on assembling active recognition elements on the SLG
surface, ideally through noncovalent interactions that pre-
serve its superior electronic properties.
Noncovalent-functionalization strategies to attach bio-
molecules to SLG are therefore of significant interest. Small
peptides and nucleic acids bind to graphene and have been
used to detect proteins,[5,6] polynucleotides,[7] and changes in
pH values,[8] nanoparticles,[9] and cells.[4] In contrast, the
activity and selectivity of proteins whose function relies on
a specific tertiary or quaternary structure, including anti-
bodies and enzymes, have not been demonstrated unambig-
uously when they are adsorbed to SLG.[10,11] Furthermore,
both experiments[12] and simulations[13] have indicated signifi-
cant conformational changes of proteins adsorbed to gra-
phene, as well as loss of function.[14] Here we demonstrate that
an anti-E. coli antibody (aEAB) readily adsorbs onto SLG,
but loses its specific recognition ability. In contrast, when the
antibody is supported on a self-assembled monolayer (SAM)
of tripodal graphene-binding molecules, it retains its ability to
recognize E. coli cells. The captured cells divide normally on
the aEAB/tripod SAM and form a biofilm on the SLG
surface. These findings demonstrate the importance of
engineering the graphene–biomolecule interface in order to
preserve protein function and that multivalent binding motifs
provide a solution to this challenge.
The graphene-binding tripod 3, which we used for anti-
body conjugation, is based on a design we have used to
functionalize SLG with redox-active moieties.[15,16] Tripod 3
features a multivalent design in which three pyrene–SLG
interactions provide exceptional SAM stability in both
aqueous and organic solvents. Electrochemical studies of
redox-active tripods with identical binding groups indicated
self-limiting monolayer formation upon exposure to dilute
tripod solutions (mm concentration) and suggested that the
tripods project their active functionality away from the SLG
surface.[16] For biofunctionalization experiments, 3 incorpo-
rates an N-hydroxysuccinimidyl (NHS) ester, which has been
used extensively for protein conjugation at exposed lysine
residues. The core of 3 was established through a Sonogashira
cross-coupling reaction between 4-bromoethyl benzoate and
triarylpropyne
1 (Scheme 1). Subsequent demethylation
[*] J. A. Mann, Prof. Dr. W. R. Dichtel
Department of Chemistry and Chemical Biology
Cornell University, Baker Laboratory
Ithaca, NY 14853-1301 (USA)
E-mail: wdichtel@cornell.edu
Dr. T. Alava, Prof. Dr. H. G. Craighead
School of Applied and Engineering Physics
Cornell University, Clark Hall
Ithaca, NY 14853-1301 (USA)
E-mail: hgc1@cornell.edu
[**] We acknowledge NSF support through use of the Cornell Nano-
fabrication Facility/NNIN and the Cornell Center for Materials
Research facilities. J.A.M. gratefully acknowledges the Integrative
Graduate Education and Research Traineeship (IGERT) Program in
the Nanoscale Control of Surfaces and Interfaces, which is
supported under NSF Award DGE-0654193, the Cornell Center for
Materials Research, and Cornell University. T.A. acknowledges
funding through a grant from the National Science Foundation
under ECCS-1001742 and through support from Analog Devices.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201209149.
Scheme 1. Synthesis of NHS–ester tripod 3 and structure of NHS–
pyrene butyrate 4.
Angew. Chem. Int. Ed. 2013, 52, 3177 –3180
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3177

.
Angewandte
Communications
using BBr3 provided the tris(phenol) derivative 2, onto which
three pyrene moieties were incorporated under Williamson
etherification conditions. Finally, the ethyl ester was saponi-
fied and subsequently elaborated to an NHS ester using
a carbodiimide-mediated esterification procedure to provide
the NHS ester tripod 3. We also evaluated SAMs comprised
of pyrene butyrate NHS ester 4, which has been used
extensively to functionalize SLG,[6,10] graphene oxide,[17] and
carbon nanotubes.[18] As a monovalent pyrene binding group,
4 also serves to evaluate the importance of the multivalent
tripodal design for effective bioconjugation. Despite the
frequent use of 4 for graphene functionalization, the perfor-
mance of aEAB conjugated to its SAMs suggests that
individual pyrene moieties do not prevent antibody denatu-
ration on SLG (see below).
We first compared the adsorption of aEAB onto pristine,
pyrene-modified, and tripod-modified SLG using a graphene-
functionalized quartz-crystal microbalance (GQCM). Anti-
bodies were chosen because of their importance in biosensors
and because they offer highly selective analyte binding
specificity that relies on maintaining their native conforma-
tion. Furthermore, E. coli is a relevant capture/immobiliza-
tion target for biosensors because it is a known foodborne
pathogen. The QCM is a piezoelectric mechanosensor whose
resonant frequency is sensitive to changes in mass adsorbed
on the surface, and whose resistance is related to both
deposition of mass and the type of mechanical coupling
between the surface and the adsorbed mass. The change in
resonant frequency (Df) is related to the mass of the antibody
deposited on the quartz surface (Dm) using the Sauerbrey
equation, Df = ÀCf Dm, in which Cf is a constant dependent
on the quartz properties, 56.6 HzmgÀ1 cm2 for the 5 MHz AT-
cut quartz used here.[19] All three substrates (tripod- and
pyrene-functionalized and bare GQCM) exhibit exponential
frequency decreases upon introduction of the aEAB and
reach an equilibrium Df ꢀ À50 Hz after ten minutes, corre-
sponding to approximately 20 pmolcmÀ2 of adsorbed anti-
body (see Figure 1 and Figure S18 in the Supporting Infor-
mation). There is little difference between the mass change
traces for the pyrene-functionalized, tripod-functionalized,
and bare GQCM surfaces. After the frequency stabilized, the
cell was rinsed with blank PBS buffer for 25 minutes, which
caused almost no desorption of the antibody. Atomic-force
microscopy of each functionalized graphene surface showed
no evidence of aggregation (see the Supporting Information).
As an additional control, we formed a monolayer of a tripod
in which the NHS ester had been displaced by ethanolamine
(Supporting Information, compound S3), thus lacking the
ability to form covalent bonds to aEAB. These monolayers
showed reduced frequency responses to antibody introduc-
tion compared to monolayers of 3, further suggesting that
covalent bond formation occurs between the antibody and 3
(see the Supporting Information, Figure S18). Overall, these
results indicate that similar amounts of aEAB are deposited
on NHS-tripod-functionalized and bare SLG and that both
types of antibody–graphene films are stable to washing.
After formation of the aEAB film formation, E. coli cells
that were resuspended in PBS buffer and introduced to the
GQCM induced an apparent increase in the frequency
Figure 1. A) GQCM trace showing the frequency (blue) and resistance
(red) response to the introduction of aEAB to a surface functionalized
with a SAM of 3 and subsequent introduction of E. coli cells.
B) Corresponding GQCM trace for aEAB adsorbed onto bare SLG.
response, which usually indicates a loss of mass from the
surface. However, the resistance (DR) of the GQCM simul-
taneously increased, which is inconsistent with this interpre-
tation. We attribute these observations to viscoelastic cou-
pling between the bacterial cells and the GQCM, a phenom-
enon that was noted previously for adsorption of bacterial
cells to a SAM-modified QCM.[20] Although it is possible in
principle to quantify cell binding through analysis of DR, such
measurements require careful calibration and are quite
susceptible to environmental noise. We instead quantified
E. coli cell binding directly through fluorescence microscopy
of appropriately stained cells.
The density of bound E. coli cells was determined for the
following three antibody-functionalized graphene surfaces:
aEAB on bare SLG, aEAB conjugated to pyrene butyrate 4,
and aEAB conjugated to tripod 3. We also performed two
additional control experiments: The first utilized a mis-
matched antibody, anti-bovine serum albumin (aBSA), which
does not recognize E. coli cells, conjugated to a SAM of 3
(Figure 2). The second employed the tripod monolayer
incapable of bioconjugation (S3) that had been exposed to
aEAB. Each surface was incubated with a suspension of the
bacterium (108 cfumLÀ1 in lysogeny broth) for ten minutes
and then rinsed to remove weakly bound cells. The remaining
surface-bound cells were stained with propidium iodide and
their density was measured repeatedly using a fluorescence
microscope (Figure 3). Notably, surfaces on which aEAB was
immobilized on the tripodal SAM showed a nearly five-fold
3178
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 3177 –3180

Angewandte
Chemie
Figure 2. Cell count histograms for A) aEAB on tripod SAMs of 3, B) aEAB adsorbed on monopod SAMs of 4, C) aEAB adsorbed on bare SLG,
and D) aBSA conjugated to tripod SAMs of 3.
Figure 3. Captured E. coli binding densities for aEAB on SAMs of 3
(red), on SAMs of 4 (blue), on bare SLG (green), and on SAMs of S3
(purple). E. coli binding density for aBSA on SAMs of 3 (orange).
higher density (7.8 Æ 1 ꢀ 105 cellscmÀ2) than when aEAB was
supported on the monovalent pyrene butyrate SAM (1.4 Æ
0.4 ꢀ 105 cellscmÀ2) or on bare SLG (1.7 Æ 0.3 ꢀ 105 cellscmÀ2).
The aBSA conjugated to the tripod SAM also showed nearly
identical density of bound cells (1.9 Æ 0.3 ꢀ 105 cellscmÀ2)
compared to the aEAB films that were not interfaced to
SAMs of 3. Since aBSA does not recognize E. coli cells, we
attribute the lower density to nonspecific cell adsorption.
These results therefore suggest that aEAB exhibits little or no
specific E. coli cell recognition when it is adsorbed on bare
graphene or conjugated to pyrene 4, yet retains this function
when conjugated to tripod 3. The S3 monolayers that are
incapable of bioconjugation also showed quite low density of
bound cells (1.8 Æ 0.5 ꢀ 104 cellscmÀ2), indicating that cova-
lent attachment between the SAM and antibody are neces-
sary for cell capture. Collectively, these differences in
densities of bound cells suggest that the nature of the SLG–
antibody interface is crucial for antibodies to retain their
desirable specific binding function.
Figure 4. A) Growth curves for E. coli cells in solution (dashed line),
on aEAB on SAMs of 3 (red), and on aEAB on bare SLG (green).
B) Optical images of E. coli cell division on aEAB/tripod SAMs,
indicating the progression of biofilm formation.
forming a graphene-conjugated biofilm (Figure 4A, red). In
contrast, E. coli cells that were immobilized on aEAB on bare
SLG did not remain attached to the graphene surface, and
biofilm coherence was lost above a colony size of two
individuals (Figure 4B, green). E. coli cells responsible for
biofilm growth on tripod-conjugated aEAB divided with
a doubling time of only 50% longer than cells grown in
solution under similar conditions (Figure 4). These observa-
tions further confirm that the nature of the antibody–
graphene interface strongly influences antibody function,
and suggest that immobilizing biomolecules on tripod SAMs
will enable studies of cell growth and differentiation. SLG is
an intriguing analytical platform for such investigations, in
part because it can serve as a transparent barrier material, as
Finally, we confirmed that E. coli cells captured by aEAB
on tripod SAMs are viable by monitoring their division on the
SLG surface (Figure 4). The immobilized cells replicate
exponentially and remain attached to the surface, eventually
Angew. Chem. Int. Ed. 2013, 52, 3177 –3180
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3179

.
Angewandte
Communications
[7] a) E. Dubuisson, Z. Yang, K. P. Loh, Anal. Chem. 2011, 83,
2452 – 2460; b) X. Dong, Y. Shi, W. Huang, P. Chen, L.-J. Li, Adv.
Mater. 2010, 22, 1649 – 1653.
[8] Z. Cheng, Q. Li, Z. Li, Q. Zhou, Y. Fang, Nano Lett. 2010, 10,
1864 – 1868.
was shown by Alivisatos for imaging nanoparticles within
bilayer graphene pockets.[21]
In conclusion, we have shown that similar masses of anti
E. coli antibodies deposit onto both bare SLG and NHS-
functionalized aromatic SAMs. The binding specificity of the
antibody depends strongly on the nature of SLG-functional-
ization method. The antibodies retain their specificity when
conjugated to tripodal SAMs, but exhibit poor E. coli cell
recognition when immobilized onto the monovalent binding
group or bare SLG. Antibody activity on tripodal SAMs was
further confirmed by the observations that captured bacterial
cells readily divide, remain bound to the graphene surface,
and form biofilms. This study answers fundamental questions
about interfacing biomolecules to SLG that are relevant to
both biosensor applications and fundamental studies at the
abiotic–biotic interface. Further studies will elaborate on the
functionality and orientational control that we have demon-
strated through the use of site-selective bioconjugation
strategies and will investigate practical SLG-based biosensors.
[9] Y. Cui, S. N. Kim, S. E. Jones, L. L. Wissler, R. R. Naik, M. C.
McAlpine, Nano Lett. 2010, 10, 4559 – 4565.
[10] a) S. Okamoto, Y. Ohno, K. Maehashi, K. Inoue, K. Matsumoto,
Jpn. J. Appl. Phys. 2012, 51, 06FD08; b) V. K. Kodali, J.
Scrimgeour, S. Kim, J. H. Hankinson, K. M. Carroll, H. W. A.
de Heer, C. Berger, J. E. Curtis, Langmuir 2011, 27, 863 – 865;
c) Y. Huang, X. Dong, Y. Liu, L.-J. Li, P. Chen, J. Mater. Chem.
2011, 21, 12358 – 12362.
[11] a) P. K. Ang, A. Li, M. Jaiswal, Y. Wang, H. W. Hou, J. T. L.
Thong, C. T. Lim, K. P. Loh, Nano Lett. 2011, 11, 5240 – 5246;
b) Y. Ohno, K. Maehashi, Y. Yamashiro, K. Matsumoto, Nano
Lett. 2009, 9, 3318 – 3322.
[12] a) J. Katoch, S. N. Kim, Z. Kuang, B. L. Farmer, R. R. Naik, S. A.
Tatulian, M. Ishigami, Nano Lett. 2012, 12, 2342 – 2346; b) Y.
Zhang, J. Zhang, X. Huang, X. Zhou, H. Wu, S. Guo, Small 2012,
8, 154 – 159; c) M. De, S. S. Chou, V. P. Dravid, J. Am. Chem. Soc.
2011, 133, 17524 – 17527.
[13] a) K. Balamurugan, E. R. A. Singam, V. Subramanian, J. Phys.
Chem. C 2011, 115, 8886 – 8892; b) G. Zuo, X. Zhou, Q. Huang,
H. Fang, R. Zhou, J. Phys. Chem. C 2011, 115, 23323 – 23328;
c) L. Ou, Y. Luo, G. Wei, J. Phys. Chem. B 2011, 115, 9813 – 9822;
d) G. Gianese, V. Rosato, F. Cleri, M. Celino, P. Morales, J. Phys.
Chem. B 2009, 113, 12105 – 12112.
[14] T. Alava, J. A. Mann, C. Theodore, J. J. Benitez, W. R. Dichtel,
J. M. Parpia, H. G. Craighead, Anal. Chem. 2013, DOI: 10.1021/
ac303268z.
[15] J. Rodrꢂguez-Lꢃpez, N. L. Ritzert, J. A. Mann, C. Tan, W. R.
Dichtel, H. D. AbruÇa, J. Am. Chem. Soc. 2012, 134, 6224 – 6236.
[16] J. A. Mann, J. Rodrꢂguez-Lꢃpez, H. D. AbruÇa, W. R. Dichtel, J.
Am. Chem. Soc. 2011, 133, 17614 – 17617.
[17] a) F. Liu, K. S. Choi, T. J. Park, S. Y. Lee, T. S. Seo, BioChip J.
2011, 5, 123 – 128; b) W. Song, D.-W. Li, Y.-T. Li, Y. Li, Y.-T.
Long, Biosens. Bioelectron. 2011, 26, 3181 – 3186.
[18] a) D. Tasis, N. Tagmatarchis, A. Bianco, M. Prato, Chem. Rev.
2006, 106, 1105 – 1136; b) R. J. Chen, Y. Zhang, D. Wang, H. Dai,
J. Am. Chem. Soc. 2001, 123, 3838 – 3839.
[19] D. A. Buttry in Electroanalytical Chemistry: A Series of Advan-
ces, Vol. 17 (Ed.: A. J. Bard), Marcel Dekker, New York, 1990,
pp. 1 – 82.
[20] a) C. Poitras, N. Tufenkji, Biosens. Bioelectron. 2009, 24, 2137 –
2142; b) R. D. Vaughan, C. K. OꢄSullivan, G. G. Guilbault,
Enzyme Microb. Technol. 2001, 29, 635 – 638.
[21] J. M. Yuk, J. Park, P. Ercius, K. Kim, D. J. Hellebusch, M. F.
Crommie, J. Y. Lee, A. Zettl, A. P. Alivisatos, Science 2012, 336,
61 – 64.
Received: November 15, 2012
Published online: February 5, 2013
Keywords: antibodies · biofunctionalization · graphene ·
.
self-assembly
[1] M. Medina-Sꢁnchez, S. Miserere, A. Merkoci, Lab Chip 2012,
12, 1932 – 1943.
[2] a) A. Bonanni, A. H. Loo, M. Pumera, TrAC Trends Anal.
Chem. 2012, 37, 12 – 21; b) Y. Liu, X. Dong, P. Chen, Chem. Soc.
Rev. 2012, 41, 2283 – 2307; c) S. Song, Y. Qin, Y. He, Q. Huang, C.
Fan, H.-Y. Chen, Chem. Soc. Rev. 2010, 39, 4234 – 4243; d) W.
Yang, K. R. Ratinac, S. P. Ringer, P. Thordarson, J. J. Gooding, F.
Braet, Angew. Chem. 2010, 122, 2160 – 2185; Angew. Chem. Int.
Ed. 2010, 49, 2114 – 2138.
[3] a) S. Bae, et al., Nat. Nanotechnol. 2010, 5, 574 – 578; b) X. Li,
et al., Science 2009, 324, 1312 – 1314; c) M. P. Levendorf, C. S.
Ruiz-Vargas, S. Garg, J. Park, Nano Lett. 2009, 9, 4479 – 4483;
d) A. Reina, X. Jia, J. Ho, D. Nezich, H. Son, V. Bulovic, M. S.
Dresselhaus, J. Kong, Nano Lett. 2009, 9, 30 – 35.
[4] M. S. Mannoor, H. Tao, J. D. Clayton, A. Sengupta, D. L.
Kaplan, R. R. Naik, N. Verma, F. G. Omenetto, M. C. McAlpine,
Nat. Commun. 2012, 3, 763.
[5] L. Wang, C. Zhu, L. Han, L. Jin, M. Zhou, S. Dong, Chem.
Commun. 2011, 47, 7794 – 7796.
[6] Y. Ohno, K. Maehashi, K. Matsumoto, J. Am. Chem. Soc. 2010,
132, 18012 – 18013.
3180
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 3177 –3180