Langmuir 2006, 22, 6969-6976 6969

Functionalized Self-Assembled Monolayers on Ultraflat Gold as

Platforms for Single Molecule Force Spectroscopy and Imaging
Vamsi K. Yadavalli, Jeffrey G. Forbes, and Kuan Wang*
Muscle Proteomics and Nanotechnology Section, Laboratory of Muscle Biology,
National Institute of Arthritis and Musculoskeletal and Skin Diseases,
National Institutes of Health, DHHS, Bethesda, Maryland 20892

ReceiVed February 1, 2006. In Final Form: May 23, 2006

Single molecule force spectroscopy is a valuable tool for studying unfolding and nanomechanical properties of
proteins. The common practice is to stretch proteins from a surface that was dosed to give a reasonable hit rate and
to analyze the curves that exhibit the expected characteristics of a single polymer. Whether the surface-bound proteins
are indeed single and isolated remains unclear, and the undesirable protein/surface interactions that obscure informative
features of the force curves are implicitly assumed to be absent. In this study, mixed self-assembled monolayers
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

(SAMs) consisting of N-hydroxysuccinimide (NHS) and oligoethylene glycol (OEG) terminated thiols on an ultraflat
gold surface were used to covalently immobilize proteins via lysine residues. By the optimization of attachment sites
via lysine-NHS linkages amidst a protein-resistant layer of the OEG SAM, it was possible to isolate single proteins
Downloaded via WEST VIRGINIA UNIV on September 21, 2018 at 19:25:23 (UTC).

for study in a controlled fashion. The single protein distribution on the surface is clearly demonstrated by atomic force
microscopy (AFM) imaging. The OEG also significantly reduces nonspecific tip-surface interactions between the
cantilever and surface. Stretching covalently attached single proteins produces high-quality and reproducible force-
extension curves. This experimental strategy is an attractive platform with which to study protein structure, interactions,
and nanomechanical properties of single proteins.

Introduction An important consideration in such experiments has been the

The study of single biomolecules enables the understanding
of properties and dynamic behavior that are usually averaged out
in ensemble (bulk) experiments. Unique structural or temporal
states may therefore be identified and studied toward under-
standing complex processes or bulk properties. Such research
involves studying the nanomechanics,1 conformational or struc-
tural changes,2,3 and molecular interactions4,5 of single molecules.
Atomic force microscopy (AFM)6 is an ideal tool not just for
imaging single molecules7,8 but also for studying molecular
interactions and nanomechanics. In particular, force-distance
spectroscopy9 has been utilized for a variety of single molecule
measurements including molecular rupture forces,10 protein
domain unfolding,11,12 and receptor-ligand interactions.5,13-15
* To whom correspondence should be addressed. Telephone: (301) 496-
4097. Fax: (301) 402-8566. E-mail: wangk@exchange.nih.gov.
(1) Rief, M.; Gautel, M.; Oesterhelt, F.; Fernandez, J. M.; Gaub, H. E. Science
1997, 276, 1109-1112.
(2) Kageshima, M.; Lantz, M. A.; Jarvis, S. P.; Tokumoto, H.; Takeda, S.;
Ptak, A.; Nakamura, C.; Miyake, J. Chem. Phys. Lett. 2001, 343 (1-2), 77-82.
(3) Forbes, J. G.; Lorimer, G. H. Science 2000, 288 (5463), 63-64.
(4) Weisel, J. W.; Shuman, H.; Litvinov, R. I. Curr. Opin. Struct. Biol. 2003,
19, 227-235.
(5) Hinterdorfer, P.; Baumgartner, W.; Gruber, H. J.; Schilcher, K.; Schindler,
H. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 3477-3481.
(6) Binnig, G.; Quate, C. F.; Gerber, C. Phys. ReV. Lett. 1986, 56 (9), 930-
933.
(7) Stroh, C.; Wang, H.; Bash, R.; Ashcroft, B.; Nelson, J.; Gruber, H.; Lohr,
D.; Lindsay, S. M.; Hinterdorfer, P. Proc. Natl. Acad. Sci. U.S.A. 2004, 101 (34),
12503-12507.
(8) Hansma, H.; Kim, K.; Laney, D.; Garcia, R.; Argaman, M.; Allen, M.;
Parsons, S. J. Struct. Biol. 1997, 119, 99-108.
(9) Frederix, P.; Akiyama, T.; Staufer, U.; Gerber, C.; Fotiadis, D.; Muller,
D.; Engel, A. Curr. Opin. Chem. Biol. 2003, 7, 641-647.
(10) Grandbois, M.; Beyer, M.; Rief, M.; Clausen-Schaumann, H.; Gaub, H.
E. Science 1999, 283, 1727-1730.
(11) Rounsevell, R.; Forman, J. R.; Clark, J. Methods 2004, 34, 100-111.
(12) Forbes, J. G.; Wang, K. J. Vac. Sci. Technol. 2004, 22 (4), 1-5.
(13) Allen, S.; Davies, J.; Davies, M. C.; Dawkes, A. C.; Roberts, C. J.; Tendler,
S. J. B.; Williams, P. M. Biochem. J. 1999, 341, 173-178.
(14) Lo, Y. S.; Simons, J.; Beebe, T. P. J. Phys. Chem. B 2002, 106 (38),
9847-9852.
immobilization of the molecules involved. These molecules may
be coupled to the AFM tip16 or the surface,17 or both,18 while
maintaining functionality. Another issue is the minimization of
nonspecific adhesion of the tip to the surface.19
In AFM experiments studying proteins on a surface, there is
often a large nonspecific adhesive force between the surface and
the cantilever tip. This results in a cantilever deflection mani-
fested by the presence of a linear nondelayed retraction curve
with the same slope as that of the contact region. A combination
of factors such as hydrophobic, electrostatic, or van der Waals
interactions may result in tip-surface interactions of magnitudes
on the order of binding events between an antibody and antigen
or receptor and ligand.20 Such confounding factors make the
selection and analysis of force-distance spectra considerably
difficult. Traditional approaches to single molecule studies have
consisted of applying low concentrations of proteins to a surface
followed by force-distance spectroscopy. The presence of a
single molecule on the surface is then inferred by fitting the
force data assuming the presence of single, isolated molecules.
More often than not, the proteins tend to aggregate, form films,
or associate on the surface. This results in the vast majority of
such traces being representative of multiple molecules. In addition
to this lack of control on the surface distribution, the presence
of such complexes on the surface exacerbates the tip-surface
adhesion that presents confounding factors in the analysis of
force curves.
(15) Hugel, T.; Seitz, M. Macromol. Rapid Commun. 2001, 22 (13), 989-
1016.
(16) Noy, A.; Vezenov, D. V.; Lieber, C. M. Annu. ReV. Mater. Sci. 1997, 27,
381-421.
(17) Wagner, P.; Hegner, M.; Kemen, P.; Zaugg, F.; Semenza, G. Biophys.
J. 1996, 70, 2052-2066.
(18) Harada, Y.; Kuroda, M.; Ishida, A. Langmuir 2000, 16, 708-715.
(19) Amirgoulova, E. V.; Groll, J.; Heyes, C. D.; Ameringer, T.; Rocker, C.;
Moller, M.; Nienhaus, G. U. ChemPhysChem 2004, 5, 552-555.
(20) Israelachvili, J. Intermolecular and surface forces, 2nd ed.; Academic
Press: San Diego, CA, 1992.

10.1021/la060320h This article not subject to U.S. Copyright. Published 2006 by the American Chemical Society
Published on Web 07/07/2006
6970 Langmuir, Vol. 22, No. 16, 2006
Typically, these regions of the force-distance curve, as well
as the tip-surface force obtained for such a trace, have been
ignored or mathematically eliminated by approximating the point
of contact for each curve.21 Different groups have presented
varied approaches toward dealing with this problem. Best et al.22
proposed criteria to select force-distance curves that would
remove subjectivity in the determination of which aspects of the
spectra are artifacts and which represent “useful” information,
representative of single molecules. Willemsen et al. developed
a model to calculate this adhesion force of a tip with a mica
surface and attempted to minimize the nonspecific adhesion by
imaging in the repulsive regime.23 Brogan et al. investigated the
use of surfactants to reduce the nonspecific binding events.24
Chemical approaches for tackling this problem have included
the use of flexible polyethylene glycol tethers25 and coating
surfaces with dextrans to reduce unwanted interactions.26 The
former approach using tethers may further complicate protein-
stretching data by introducing an additional flexible component
that may present dynamics of its own. The latter approach
involving dextran requires coating the cantilever tip to reduce
nonspecific interactions. In addition, these approaches have not
completely solved the problem of clearly separating single
molecules on surfaces.
Here, we take a surface chemistry approach toward improving
the behavior of the substrate to which the proteins are attached.
This is a fundamental approach for eliminating the source of the
problem by using self-assembled monolayers (SAMs). SAMs
have been long studied for their ability to create precisely
controlled and functionalized surfaces for studying interfacial
phenomena and biomolecular and cellular interactions.27,28 SAMs
of ω-substituted alkanethiols on Au(111) surfaces, in particular,
have attracted a great deal of attention due to their ease of
preparation, their stability, and their ability to create a broad
variety and distribution of surface groups and properties.29,30
Mixed SAMs using binary mixtures of functionalized thiols have
been synthesized by coadsorption to a surface from a mixture
of thiols.31,32 Molecules of interest may then be immobilized at
a single molecule level on this functionalized surface based on
the controllable functional groups.
This paper presents the first systematic approach, to our
knowledge, toward isolating single proteins on surfaces with a
view to studying their unfolding and nanomechanics. In addition,
we present data showing how problematic nonspecific tip-surface
interactions may be avoided by the use of self-assembled
monolayers. We demonstrate that a mixed thiol solution with a
protein-resistant oligoethylene glycol (OEG) thiol SAM along
with sparsely populated N-hydroxysuccinimide (NHS) thiol
(21) Baumgartner, W.; Hinterdorfer, P.; Schindler, H. Ultramicroscopy 2000,
82, 85-95.
(22) Best, R. B.; Brockwell, D. J.; Toca-Herrera, J. L.; Blake, A. W.; Smith,
D. A.; Radford, S. E.; Clarke, J. Anal. Chim. Acta 2003, 479 (1), 87-105.
(23) Willemsen, O. H.; Snel, M. M. E.; Kuipers, L.; Figdor, C. G.; Greve, J.;
De Grooth, B. G. Biophys. J. 1999, 76 (2), 716-724.
(24) Brogan, K. L.; Shin, J. H.; Schoenfisch, M. H. Langmuir 2004, 20, 9729-
9735.
(25) Hinterdorfer, P.; Gruber, H. J.; Kienberger, F.; Kada, G.; Riener, C.;
Borken, C.; Schindler, H. Colloids Surf., B: Biointerfaces 2002, 23, 115-123.
(26) Stevens, M. M.; Allen, S.; Davies, M. C.; Roberts, C. J.; Schacht, E.;
Tendler, S. J. B.; VanSteenkiste, S.; Williams, P. M. Langmuir 2002, 18, (17),
6659-6665.
(27) Ulman, A. Chem. ReV. 1996, 96 (4), 1533-1554.
(28) Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G.
M. Chem. ReV. 2005, 105 (4), 1103-1169.
(29) Bertilsson, L.; Liedberg, B. Langmuir 1993, 9, 141-149.
(30) Bain, C. D.; Troughton, E. B.; Tao, Y.-T.; Evall, J.; Whitesides, G. M.;
Nuzzo, R. G. J. Am. Chem. Soc. 1989, 111, 321-335.
(31) Folkers, J. P.; Laibinis, P. E.; Whitesides, G. M. J. Adhes. Sci. Technol.
1992, 6 (12), 1397-1410.
(32) Laibinis, P. E.; Nuzzo, R. G.; Whitesides, G. M. J. Phys. Chem. 1992,
96 (12), 5097-5105.
YadaValli et al.
tethers33 can be used to create a surface for the covalent attachment
of single protein molecules. The presence of the oligoethylene
glycol prevents the adhesion of the protein of interest to all but
the exposed NHS groups on the surface. Surfaces with an
oligoethylene glycol (OEG) SAM surface have been demonstrated
to resist protein adsorption,34,35 resulting from steric repulsion
as well as the exclusion of proteins by a water interface layer
created by the helical arrays of the OEG chains. It is also shown
that the OEG SAM ensures that there is minimal tip-surface
interaction at points where a molecular recognition event does
not occur.
This approach may be extended further for the creation of
surfaces that may be used to covalently immobilize biomolecules
for devices and biosensors. Other single molecule techniques
such as optical tweezers where nonspecific interactions also
present confounding factors36 could also be potentially aided by
the use of such functionalized platforms.
Experimental Procedures
Materials and Instrumentation. (1-Mercaptoundec-11-yl) hexa-
ethylene glycol (OEG), HS-C11- (EG)6OH, and (1-mercaptohexa-
decanoic acid)-N-succinimidyl ester (NHS-terminated thiol), HS-
C15COO-NHS, were purchased from Prochimia (Gdansk, Poland).
Ethanol (200-proof) was purchased from the Warner-Graham
Company (Cockeysville, MD). Epotek 377 glue was purchased from
Epoxy Technology (Billerica, MA). Ethanolamine and tetrahydro-
furan (THF) were purchased from Sigma-Aldrich (St. Louis, MO).
Phosphate-buffered saline (PBS; 155.17 mM NaCl, 1.06 mM
KH2PO4, 2.97 mM Na2HPO4, pH 7.4) was obtained from Gibco and
sterile filtered prior to use. To minimize contamination, experiments
were performed in Teflon beakers and containers (Savillex Corp.,
Minnetonka, MN). Silicon nitride (Si3N4) microlevers from Veeco
and ultrasharp Type I MAC cantilevers from Molecular Imaging
(Phoenix, AZ) were used for force measurement and imaging.
Proteins. A variety of different proteins of different sizes and
shapes were surveyed. A large spherical protein (ferritin), an
immunoglobulin G (anti-nebulin SH3 antibody), a small intrinsically
disordered nebulin protein fragment (ND66 fragment), and a large
filamentous protein (myosin) were investigated.
To enhance the attachment of single molecules, all purified proteins
were applied to a gel filtration column immediately prior to use and
the monodisperse fractions selected. Rabbit skeletal myosin was
purified and gel filtered as described.37 Rabbit myosin subfragment
S2 was gel filtered on a Superose 6 column in 0.6 M KCl, 20 mM
sodium phosphate buffer (pH 7.0). Cloned nebulin ND66 fragments
were expressed in E. coli as described earlier.38 ND66 was gel filtered
on a Superose 12 column equilibrated with 150 mM NaCl, 5 mM
CaCl2, 10 mM imidazole (pH 7.0). Ferritin (Type I from horse spleen)
was obtained from Sigma-Aldrich (St. Louis, MO). Immediately
prior to use, the ferritin was gel filtered using a Superose 6 column
in the same buffer. Nebulin anti-SH3 antibodies (polyclonal IgG)
were produced in rabbit against commercially synthesized peptide
sequences coupled to a KLH carrier. Serum was affinity purified on
a peptide conjugated agarose column, eluted with glycine, and gel
filtered prior to use using a Superose 6 column in PBS pH 7.4
(Gutierrez and Wang, unpublished data).
Functionalized OEG SAM on Template-Stripped Gold Sur-
faces. Ultraflat gold surfaces were fabricated using the template-
stripping process of Wagner et al.39 Briefly, evaporated gold surfaces
(33) Wagner, P.; Kernen, P.; Hegner, M.; Ungewickell, E.; Semenza, G. FEBS
Lett. 1994, 356, 267-271.
(34) Prime, K. L.; Whitesides, G. M. J. Am. Chem. Soc. 1993, 115, 10714-
10721.
(35) Pale-Grosdemange, C.; Simon, E. S.; Prime, K. L.; Whitesides, G. M. J.
Am. Chem. Soc. 1991, 113, 12-20.
(36) Stout, A. L. Biophys. J. 2002, 80 (6), 2976-2986.
(37) Root, D. D.; Yadavalli, V. K.; Forbes, J. G.; Wang, K. Biophys. J. 2006,
90 (8), 2852-2866.
(38) Wang, K.; Knipfer, M.; Huang, Q. Q.; van Heerden, A.; Hsu, L. C.;
Gutierrez, G.; Quian, X. L.; Stedman, H. J. Biol. Chem. 1996, 271 (8), 4304-14.
SAMs for Single Molecule Force Spectroscopy
Figure 1. Covalent immobilization of single proteins on mixed self-assembled monolayer of (1-mercaptohexadecanoic acid)-N-succinimidyl
ester and (1-mercaptoundec-11-yl) hexaethylene glycol on ultraflat Au(111) surface. Single antibodies are covalently immobilized at different
locations and with varied orientations via the sparsely distributed and exposed NHS groups on the SAM surface.
on mica substrates were glued face down using epoxy glue Epotek
377 on Si(100) wafers or glass slides that were piranha (3:1 v/v
H2SO4:H2O2) cleaned. Proper care was exercised in handling the
extremely corrosive piranha solution. The surfaces were then heat
cured in an oven for 1-2 h at 150 °C. The template-stripped gold
surfaces were obtained by stripping the mica with tetrahydrofuran.
The resulting gold surfaces were then washed with ethanol and
dried in a stream of purified N2. The RMS (root-mean-square)
roughness of the gold surfaces was measured to be less than 0.5 nm
over 5 µm.
Thiol SAMs were prepared by the incubation of the freshly
prepared gold surfaces in a thiol solution in ethanol for 16-24 h.
After incubation, the gold surfaces were rinsed with ethanol,
dried in a stream of purified N2, and imaged in the buffer. The molar
ratio of the OEG thiol to the NHS thiol was empirically deter-
mined by titration, and an approximately 10 000:1 ratio was used
to obtain well-separated attachment sites on the gold. This also
corresponded to a theoretical estimation assuming a distance be-
tween attachment sites of 200 nm, assuming an area per mole-
cule of 21.4 Å2.40 Figure 1 shows a schematic of the mixed SAM
surface with NHS groups surrounded by protein-resistant oligo-
ethylene glycol thiol (OEG) as prepared by this method. SAM sur-
faces with 100% OEG6 thiol and 100% NHS thiol were studied as
negative and positive controls for the functionalized surface. In both
cases, the functionalized surfaces were incubated with protein in
PBS for 1 h. The protein concentration for all experiments was ∼1
µg/mL with a total protein of ∼500 ng. The unreacted NHS groups
were then quenched for 1 h in 0.1 M ethanolamine solution at pH
7.4. The surfaces were washed with PBS and imaged in noncontact
mode. All experiments were performed at room temperature unless
otherwise specified.
MAC mode AFM cantilever tips for imaging were cleaned by
exposure to ozone in the presence of high-intensity UV light (254
nm) to remove any contamination by organics on the surface, while
preserving the magnetic coating on the cantilevers. Softer cantilever
tips for force-distance measurements were cleaned in a solution of
(39) Wagner, P.; Hegner, M.; Guntherodt, H.-J.; Semenza, G. Langmuir 1995,
11, 3867-3875.
(40) Harder, P.; Grunze, M.; Dahint, R.; Whitesides, G. M.; Laibinis, P. E.
J. Phys. Chem. B 1998, 102, 426-436.
Langmuir, Vol. 22, No. 16, 2006 6971
ethanolic KOH (0.5 M) followed by extensive rinsing in ethanol and
water. Cantilevers were calibrated and their force constants measured
using the thermal fluctuation method.41
Atomic Force Microscopy: Single Molecule Force Curves
and Imaging. For adsorbing the myosin on the gold surface, 190
µL of sterile filtered buffer (0.5 M KCl, 20 mM imidazole, pH 7.0,
1 mM tris(2-carboxyethyl)phosphine hydrochloride (TCEP)) was
placed on the freshly prepared SAM and 10 µL of protein solution
in the same buffer was carefully injected under the buffer surface
and gently mixed. Final concentrations were ∼1 µg/mL. The protein
was incubated for 30 min at room temperature (22 °C), and the
surface was then rinsed three times with the same buffer to remove
any unattached protein. A buffer near physiological ionic strength
(0.1 M KCl, 20 mM imidazole, pH 7.0, 1 mM TCEP, pH 7.0) was
then added (550 µL), and AFM measurements were immediately
taken.
A PicoSPM AFM instrument (Molecular Imaging, Phoenix, AZ)
in both contact and noncontact (MAC) mode was used for force-
distance and imaging measurements. Soft cantilever tips (nominal
force constant ∼0.01 N/m, resonance frequency 7 kHz) were used
for the force-distance analysis. Stiffer cantilevers (nominal force
constant 0.6-0.95 N/m, resonance frequency 75-105 kHz) were
used for imaging in the MAC mode.
Freshly cleaned Si3N4 tips were used to measure the presence and
magnitude of nonspecific adhesion to clean surfaces. Clean gold,
glass, silicon, and mica substrates were investigated as substrates.
Epitaxially deposited gold on mica stored under Ar was cleaned in
UV/ozone. Glass and polished Si(111) were cleaned immediately
prior to use using piranha solution. Surfaces were washed extensively
with water and imaged in PBS.
Force spectra were obtained by moving the tip in contact with the
surface, holding it in place for some time to allow binding to occur
and then retracting in a repeated, cyclic manner. Several hundred
curves were obtained for each experiment by moving the tip to
different points on the surface. The cantilever speed was ∼1 µm/s
and the force of contact was ∼1 nN to avoid damaging the surface
protein. Pull-off forces were analyzed using Scanning Probe Image
Processor (SPIP) software (Image Metrology, Denmark).
(41) Hutter, J.; Bechhoefer, J. ReV. Sci. Instrum. 1993, 64 (7), 1868-1873.
6972 Langmuir, Vol. 22, No. 16, 2006 YadaValli et al.

Molecular Dynamics Simulation Methods. The atomic model

derived from X-ray crystallography of scallop muscle myosin
subfragment-242 was used as a starting point for the molecular
dynamics simulations. The leucine zipper residues were removed
from the atomic model, and hydrogens were added to the remaining
intact residues (843-885) of the myosin subfragment-2 coiled coil.
The molecule was oriented with the center of the coiled coil along
the x-axis and then placed into a 228 Å × 25 Å × 33 Å TIP3 water
box with four sodium ions to neutralize the system. The structure
was manipulated and visualized with VMD.43 Energy minimization
and molecular dynamics were performed using NAMD.44 This
hydrated system was then minimized for 60 000 steps at 0 K and
then equilibrated to 300 K over 20 000 steps (40 ps). The system
was then subjected to 1.5 ns of molecular dynamics. The relaxed
structure at the 1.5 ns simulation was used in subsequent steered
molecular dynamics simulations (SMD). SMD was performed using
NAMD with a spring constant of 10 kcal/(mol Å2) (6.9 N/m) and
a velocity of 0.0001 Å/step (0.05 Å/ps). The simulation was run for
3000 ps for an extension of 15 nm.

Results and Discussion

Force-distance spectra are operationally defined by the
cantilever movement at the surface, reflecting mainly the
stretching of the protein segment between the tip and substrate.
The force-distance curves are subsequently interpreted, e.g., in
the context of the wormlike chain model45 or the freely jointed
chain model15 to obtain nanomechanical information.
Typical force-distance spectra from AFM, however, are
frequently composite curves consisting of contributions from
both tip-surface and the tip-target interaction (Figure 2A).
Figure 2B shows the occurrence of tip-surface adhesion of a
Si3N4 tip interacting with a gold surface. The attractive force
bends the cantilever and delays its release from the surface,
causing the tip-surface spikes indicated on this figure. Typically
such spikes can linger up to tens of nanometers long, obscuring
valuable nanomechanical data as the cantilever retracts from the
surface.
To illustrate the type of nanomechanical information at low
Figure 2. Reduction of tip-surface sticking on functionalized OEG
extensions, the mechanical unfolding of a 42-amino acid S2 SAM. Typical AFM force-distance curves (right panels) and the
fragment of myosin coiled-coil of R-helices was simulated by cantilever deflections (left panels) are shown. (A) Stretching of a
the SMD method. Parts A and B of Figure 3 show the force protein molecule myosin on the nonpassivated gold surface. (B)
spectrum from the SMD simulation in water, with low-pass Tip-surface sticking caused by nonspecific adhesion forces. Convex
filtering to remove the thermal noise that obscures the larger deflection of the cantilever is caused by the tip sticking to the gold
motions of the molecular system. The principal characteristics surface as it retracts. (C) No tip-surface sticking for a passivating
OEG SAM on the surface. Tip-target traces are seen only when a
are a rapid rise to a plateau which varies in force as the turns target protein recognition event occurs. (D) Histogram analysis of
of the helices unfold and then a rapid rise to higher forces as the pull-off forces from stretching ND66 nebulin fragments on either
unfolded polypeptide becomes increasingly stretched. The 42- naked gold surface or on a mixed SAM/gold surface. A substantial
residue polypeptide chains are almost completely unfolded by portion of the pull-off forces on naked gold derived from the
8 nm of extension where the force begins to rise rapidly. As can tip-surface interaction. In contrast, no tip-surface force was
be seen in the simulation of the short coiled-coil (Figure 3B) observed on SAM and the pull-off forces arose from stretching
ND66 molecules. Clean gold surfaces show a marked increase in
(and experimentally for longer coiled-coils37), there is significant
tip-surface sticking even when a low concentration of surface protein
variation in the plateau force as the coiled-coils unfold. This is present.
variation starts shortly after the plateau force is reached, which
is less than 1 nm of extension in this simulation. is therefore of paramount importance that such tip-surface
This simulation demonstrates the fact that there may be interactions be minimized or eliminated in order to extract
significant structural information in the first few nanometers of complete data from the low force and short extension regimes
a force spectrum that can often be obscured by tip-surface of the force curves.
interactions. This is especially true for structures such as the To illustrate this point, myosin subfragment S2 was im-
coiled-coil that has structural changes at low levels of force. It mobilized on the mixed SAM platform as described above and
the force-distance curve was obtained by stretching this protein
(42) Li, Y.; Brown, J. H.; Reshetnikova, L.; Blazsek, A.; Farkas, L.; Nyitray, via AFM (Figure 3C). As described in an earlier work from our
L.; Cohen, C. Nature 2003, 424 (6946), 341-345.
(43) Humphrey, W.; Dalke, A.; Schulten, K. J. Mol. Graphics 1996, 14 (1),
group, the curves show the characteristic rise, plateau, and
33. exponential phases of stretching an R-helical coiled-coil, before
(44) Kale, L.; Skeel, R.; Bhandarkar, M.; Brunner, R.; Gursoy, A.; Krawetz, detaching from the AFM cantilever. The rise length of around
N.; Phillips, J.; Shinozaki, A.; Varadarajan, K.; Schulten, K. J. Comput. Phys.
1999, 151 (1), 283-312. 5 nm and the plateau force of 70-80 pN correspond well to the
(45) Bustamante, C.; Marko, J.; Siggia, E. Science 1994, 265, 1599-1600. reported values for S2.37 In particular, the striking similarity
SAMs for Single Molecule Force Spectroscopy
Figure 3. Nanomechanical events at low force and small extension.
(A, B) Steered molecular dynamics of stretching scallop myosin S2
coiled-coil reveals significant structural (A) and mechanical events
(B) at low force-low extension regions of force-distance curves.
Stretching speed is 5 m/s. (C) AFM stretching of rabbit myosin S2
R-helical coiled-coil on the SAM platform shows a striking similarity
to computationally determined events. Stretching speed is 0.5 µm/s.
A characteristic triphasic extension behavior with rise (R), plateau
(P), and exponential (E) phases is observed (see ref 37 for details).
between the experimentally obtained curves and those obtained
computationally illustrates the power of integrating the current
platform with molecular modeling to gain deeper nanomechanical
insights over a broad range of force and extension. The ability
to resolve force-extension behavior of the S2 subfragment
revealed that this flexible region of the myosin motor is able to
undergo a force-induced unfolding during the acto-myosin
interaction. The principal difference between the simulated and
experimental spectra of coiled-coils is in the force level of the
plateau. In the simulation data shown in Figure 3, the plateau
force is ∼1 nN, which is about 10-fold higher than that reached
experimentally. This is due, in part, to the high velocities necessary
for SMD simulations to be computationally efficient. (Reduction
of the velocity by a factor of 5 reduces the plateau force 2-fold.)
For the experiment, the stretching speed was ∼0.5 µm/s, which
is 6 orders of magnitude less than the simulation.
Elimination of Tip-Surface Adhesion: OEG SAM on
Ultraflat Gold. Mica,46 glass,47 and gold17 have been typical
Langmuir, Vol. 22, No. 16, 2006 6973
substrates for biological applications involving force microscopy.
While mica presents an atomically flat surface, the difficulty of
surface modification limits its usage for single molecule work.
In addition, mica surfaces are also prone to wear at the atomic
level.48 Glass is rough at an atomic level, and discrimination of
nanometer-size features poses problems for imaging single
molecules. Gold surfaces are relatively inert and may be easily
functionalized with thiol molecules having tailorable endgroups.
Epitaxially deposited gold, however, is limited to flat terraces
only a few hundreds of nanometers across. Ultraflat gold prepared
by template stripping39 has been suggested as an attractive
alternative for preparing large-area flat surfaces. Such surfaces,
with very low RMS roughness (<0.5 nm), result in features that
are easily distinguished from the background. The added
advantage of using ultraflat gold is the reduction in the number
of surface artifacts that may confound the interpretation of AFM
imaging data.49 We adapted the strategy of preparing such flat
gold surfaces that could be functionalized with OEG6 thiol to
minimize protein physisorption. NHS thiol groups were used to
covalently immobilize proteins on the surface via lysine groups.
Figure 2B represents a force curve resulting from the stretching
and unfolding of myosin molecules on a gold surface. Such
force-distance traces, often with significant tip-surface con-
tribution, are typically selected for structural interpretation. Figure
2C shows the interaction of a cleaned tip with a surface
functionalized with an OEG SAM, with the protein (see below)
attached via NHS tethers. There was no sticking of the cantilever
to the surface at the point of lift-off. The absence of tip-surface
adhesion ensures a zero baseline that makes it possible to
discriminate unfolding events at low forces near the point of
contact at the beginning of stretch.
To compare our SAM platform with traditionally used
substrates, the magnitude of tip-surface adhesion was measured
by quantifying forces of a tip with clean surfaces of mica, glass,
polished silicon, and gold. The tip-surface force was determined
as the maximum force recorded by the cantilever as it retracts
from the substrate surface. Owing to sticking to the surface,
forces as high as hundreds of piconewtons were observed in
some cases. Freshly cleaned AFM cantilever tips were cycled
for several hundred force-distance curves for each substrate in
0.1 M PBS. Randomly selected traces (n ) 100) were used to
estimate the tip-surface force after subtraction of the baseline
trace. Since the surfaces were clean and devoid of any protein,
the force of adhesion would represent the nonspecific tip-surface
adhesion. Figure 4 shows representative tracings of tip-surface
interactions and a histogram of the frequency of occurrence vs
the tip-surface forces (pN) for the substrates studied. Tip-
surface forces of even several tens of piconewtons may be
overwhelming compared to unbinding forces. Mica surfaces
display a very large degree of tip-surface sticking on the order
of a few hundred piconewtons as the tip retracts from the surface.
Minimal sticking is observed in only 8% of the traces. Piranha-
cleaned polished silicon has 43% of the curves showing little or
no sticking. Glass slides are sticky, with only 5% showing minimal
tip-surface adhesion. The high frequency of unwanted events
for glass and mica render them unsuitable for nanomechanical
studies of single molecules. Unexpectedly, the mica surface under
PBS displayed very little tip sticking, with a majority of the
traces showing no interaction, probably resulting from the
(46) Thundat, T.; Allison, D. P.; Warmack, R. J.; Brown, G. M.; Jacobson,
K. B.; Schrick, J. J.; Ferrell, T. L. Scanning Microsc. 1992, 6 (4), 911-918.
(47) Karrasch, S.; Dolder, M.; Schabert, F.; Ramsden, J.; Engel, A. Biophys.
J. 1993, 65 (6), 2437-2446.
(48) Helt, J. M.; Batteas, J. D. Langmuir 2005, 21 (2), 633-639.
(49) Hansma, H. G. Proc. Natl. Acad. Sci. U.S.A. 1999, 96 (26), 14678-
14680.
6974 Langmuir, Vol. 22, No. 16, 2006 YadaValli et al.

Figure 5. Passivation of gold surface by OEG SAM. AFM analysis

in MAC mode reveals that no protein adheres to the surface, either
before (A) or after protein adsorption and washing (B), demonstrating
the resistance of such a surface. Surface roughness: before (A)
-0.64 nm, and after (B) -0.76 nm over 500 nm.

Figure 4. Tip-surface sticking of commonly used solid surfaces

for single molecule work. Histograms of pull-off forces for Si3N4
tips from different clean substrates under aqueous buffer and in the
absence of protein. The large pull-off forces due to nonspecific
tip-surface adhesion are most evident on mica and glass surfaces.
Clean gold surfaces show a very low level of nonspecific tip-
surface sticking.

lessening of electrostatic interaction between the mica and the

Si3N4 tip by PBS. However, the presence of protein in PBS
resulted in large tip-sticking events similar to that seen for mica
under water. While silicon is relatively easy to functionalize, the
tip interaction is still too high to be suitable for quantitative
nanomechanical studies.
Clean gold and a clean tip show very little tip-surface
interaction, with 93% of the curves displaying none or very little
sticking behavior. However, even the presence of small quantities
of protein makes the gold very adhesive, negating the advantages
of a clean gold surface. Figure 2D shows histogram analysis of
the frequency of occurrence vs the pull-off force for surfaces
with a model disordered protein (ND66 fragment of nebulin38).
The plot shows the pull-off force data from two separate
experiments done with the same type of tips and protein fractions
in parallel on gold surfaces with and without an OEG SAM
layer. A significant decline in tip-surface sticking is seen on
OEG SAM gold, with 88% nonsticky traces, as compared to
clean gold with only 35% nonsticky traces. This method thus
greatly simplifies the analysis of force curves in terms of
nanomechanical events of the target molecules. It is clear that
the ease of preparation of such substrates coupled with their
versatility and atomic scale flatness make them ideal for studying
Figure 6. Densely packed “carpet” of ferritin molecules covalently
immobilized on a SAM surface with 100% (1-mercaptohexadecanoic
acid)-N-succinimidyl ester. The thickness of the carpet was revealed
by scratching out 250 nm patches with the AFM tip at increasing
forces. Forces at spot 1 ) 48 nN, spot 2 ) 70 nN, spot 3 ) 105
nN, and spot 4 ) 128 nN. The removed ferritin (represented by the
asterisk (/)) is piled up at the edge of each excavation site.
single molecule interactions and nanomechanics. The force-
distance curves as well as the preparation of the surface are
repeatable over a large number of measurements. An additional
advantage is that the OEG SAM also prolongs the useful life of
the tip and several hundreds of curves can be taken before the
tip becomes heavily contaminated.
Covalent Immobilization of Single Molecules on NHS-
Functionalized SAM. Surfaces with an oligoethylene glycol
(OEG) SAM surface have been reported to resist protein
SAMs for Single Molecule Force Spectroscopy
Figure 7. Single molecule imaging on functionalized OEG SAM on gold. AFM images in MAC mode demonstrate single protein molecules
of IgG immunoglobulins. Height analysis shows the presence of single molecules with a height of around 5-10 nm, consistent with the
dimensions of IgG at different orientations. Panels A, B, and C represent decreasing concentrations of surface NHS groups and corresponding
density of surface protein over a 500 nm2 area.
adsorption.34,35 A control experiment was performed to investigate
this resistance to adhesion. Figure 5 shows AFM images of a
gold surface with OEG6 thiol before and after incubation with
the anti-nebulin SH3 antibody as a representative antibody. There
was no change in surface properties after incubation (verified by
visual examination and topographical analysis including surface
roughness, mean height, and depth of features), reflecting the
inability of the protein to adhere noncovalently to the surface.
This result is consistent with earlier work demonstrating OEG
surfaces resisting protein adhesion.40,50 In contrast, a surface
with 100% NHS thiol is shown to generate a “carpet” of protein,
resulting from extensive and multiple sites for covalent im-
mobilization of proteins. As a proof-of-concept illustration, the
attachment of ferritin was performed. Ferritin is a spherical
supermolecular assembly (∼12 nm in diameter) consisting of a
protein shell surrounding an iron hydroxide core.51 The shape
of the molecule renders it attractive for preparing ordered surface
lattices.52
To demonstrate the presence of a single layer of ferritin on
the surface, the protein was scratched out via contact mode AFM
by the application of forces in the range of 50-200 pN to clear
out patches on the SAM surface. Figure 6 shows a NHS thiol
SAM surface covered with ferritin. By the application of different
forces, it was possible to excavate the attached protein to different
extents in the form of 250 nm squares. The depth of the patch
varies with the force applied and at the highest force where all
the ferritin molecules were dug out and piled up nearby (Figure
6, spots 1-4), presumably exposing the SAM surface beneath.
The depth is seen to be around 12 nm, corresponding to the
diameter of a single ferritin molecule. Such experiments,
consistent with earlier work in nanografting53 and dip-pen
lithography,54 may demonstrate the ability to “sweep” out
biomolecules by selectively applying the right force to make
surfaces with a high level of control at the nanometer scale.
An important consideration in single molecule studies is the
optimization of the right concentration of the sample on the
surface. The protein concentration used has to be minimized to
reduce the number of multiple pickups while also maintaining
(50) Feldman, K.; Hahner, G.; Spencer, N. D.; Harder, P.; Grunze, M. J. Am.
Chem. Soc. 1999, 121, 10134-10141.
(51) Caruso, F.; Furlong, D. N.; Kingshott, P. J. Colloid Interface Sci. 1997,
186, 129-140.
(52) Yamashita, I. Thin Solid Films 2001, 393, 12-18.
(53) Xu, S.; Miller, S.; Laibinis, P. E.; Liu, G.-y. Langmuir 1999, 15, 7244-
7251.
(54) Lee, K.-B.; Park, S.-J.; Mirkin, C. A.; Smith, J. C.; Mrksich, M. Science
2002, 295 (5560), 1702-1705.
Langmuir, Vol. 22, No. 16, 2006 6975
a sufficiently high hit rate to have a useful experiment. By
controlling the number of exposed NHS groups, it was possible
to reduce the attachment sites for a protein on the surface. Figure
7 demonstrates the attachment of single anti-nebulin SH3 antibody
molecules on a mixed SAM surface by the variation of the NHS
thiol to OEG thiol ratio. The surface profile indicates the presence
of molecules with a height of around 3-5 nm, corresponding to
the reported diameter of immunoglobulin G (IgG).55 Antibodies
with lysine residues distributed on the surface of these non-
spherical molecules are capable of adopting different orientations
on the surface, resulting in a broader height distribution. The
surface clearly shows a number of single protein molecules after
incubation with ∼1 µg/mL antibody. Surfaces with a higher
concentration of NHS groups in the starting thiol mixture produced
a higher concentration of attached antibody molecules on the
surface (Figure 7A). An optimization of the protein concentration
as well as the ratio of the OEG to the NHS thiols thus makes
it possible to tailor the surface to produce the desired distribution
of protein molecules on the surface. Antibody coverage for the
experiment shown in Figure 7B (∼20 molecules/500 nm2) was
obtained by halving both the concentration of the NHS thiol in
the SAM and the protein concentration to ∼0.5 µg/mL (an overall
reduction by a factor of 4). With further optimization it was
possible to obtain a sparse distribution (Figure 7C) of ∼1 antibody
molecule over a 500 nm2 area. The presented platform thus offers
significantly greater control over these parameters, reducing the
amount of trial and error associated with such experiments.
This technique has been applied to the covalent immobilization
and site-directed attachment of native nebulin molecules and
will be published in a subsequent paper. While it may be noted
that this method does not result in a patterned protein array, it
is clear that the attachment sites are generally well distributed
to view single molecules on the surface. The exposed functionality
of the SAM can easily be tailored to enable attachment of a wide
variety of targets, such as cysteine residues of native and
engineered proteins, polyhis-tags, and carbohydrates on glyco-
proteins.
Conclusions
This paper has demonstrated the use of a self-assembled
monolayer approach containing dispersed covalent attachment
sites to create single molecule tethers on an ultraflat and
nonadhesive solid surface. Such sites can be used for determining
(55) Silverton, E. W.; Navia, M. A.; Davies, D. R. Proc. Natl. Acad. Sci.
U.S.A. 1977, 74 (11), 5140-5144.
6976 Langmuir, Vol. 22, No. 16, 2006
the force behavior of single molecules as well as covalently
immobilizing biomolecules for a variety of applications. This
approach significantly lowered or abolished nonspecific tip-
surface interaction for force-distance spectroscopy. The reduction
in such nonspecific behavior allowed the visualization of key
nanomechanical parameters over a broad force-extension regime
for the stretching of proteins. In addition, by careful control of
the surface properties and altering imaging conditions, it is
possible to apply a size discrimination approach to detect single
molecules. Such an approach is clearly well suited for a broad
range of biophysical and molecular imaging applications by
YadaValli et al.
eliminating guesswork, both by the establishment of clearly
controlled single protein immobilization on a surface and by the
reduction in the arbitrary cutoff for nonspecific adhesive forces.
Acknowledgment. We thank Gustavo Gutierrez-Cruz and
Wanxia L. Tsai for protein preparation. This research was
supported by the Intramural Research Program of the NIH,
NIAMS. This study utilized the high-performance computational
capabilities of the Biowulf PC/Linux cluster at the National
Institutes of Health, Bethesda, MD (http://biowulf.nih.gov).
LA060320H