X-ray spectromicroscopy study of competitive adsorption of protein and
peptide onto polystyrene-poly„methyl methacrylate…
Bonnie O. Leung and Adam P. Hitchcocka?
BIMR, TuesduesMcMaster University, Hamilton, Ontario, Canada L8S 4M1
John L. Brash
School of Biomedical Engineering, McMaster University, Hamilton, Ontario, Canada L8S 4L8
Andreas Scholl and Andrew Doran
Advanced Light Source, Lawrence Berkeley National Laboratory, Berkeley, California 94720
Institut fur Biochemie, Universitatsklinikum Charite, Humboldt-Universitat, 10117 Berlin, Germany
Joerg Overhage, Kai Hilpert, John D. Hale, and Robert E. W. Hancock
Department of Microbiology and Immunology, University of British Columbia, Vancouver, British Columbia,
Canada V6T 1Z4
?Received 29 April 2008; accepted 19 June 2008; published 24 July 2008?
A synchrotron-based x-ray photoemission electron microscope ?X-PEEM? was used to investigate
the coadsorption of a mixture of human albumin serum and SUB-6, a synthetic antimicrobial
peptide, to a phase-segregated polystyrene/poly?methyl methacrylate? ?PMMA? substrate at varying
concentrations and pH. The authors show that X-PEEM could detect the peptide adsorbed from
solution at concentrations as low as 5.5?10−9M and could differentiate the four components via
near-edge x-ray absorption fine structure spectromicroscopy. At neutral pH the SUB-6 peptide
adsorbed preferentially to PMMA. At a pH of 11.8 where the charge on the peptide was neutralized,
there was a more balanced adsorption of both species on the PMMA domains. The authors interpret
these observations as indicative of the formation of an electrostatic complex between positive
peptide and negative protein at pH of 7.0. This solution complex had an adsorption behavior that
depended on the polarity of the substrate domains, and favored adsorption to the electronegative
PMMA regions. At a pH of 11.8 the complex formation was suppressed and a more competitive
adsorption process was observed. © 2008 American Vacuum Society. ?DOI: 10.1116/1.2956637?
Synchrotron-based soft x-ray microscopy1–3using near-
edge x-ray absorption fine structure ?NEXAFS? contrast is a
useful tool for acquiring information on soft matter, in par-
ticular, polymeric and biological materials. In fact, NEXAFS
spectromicroscopy is well suited for visualizing macromo-
lecular subcomponents by differentiating between polysac-
charides, nucleic acids, lipids, and proteins without the use
of external probes.4NEXAFS is also an excellent tool to
differentiate among amino acids and small peptides.5–7How-
ever, distinguishing one type of protein from a multitude of
types of proteins8is difficult due to the fundamental similar-
ity of the NEXAFS spectra of all proteins, which arises from
the strong local sensitivity of the spectroscopy combined
with spectral averaging over relatively similar distributions
of amino acids. Nonetheless, since each amino acid yields a
unique NEXAFS spectrum,5,6proteins or peptides containing
an unusual sequence may be identified in favorable cases.7,8
Recently, a program was developed to predict the experi-
mental carbon, nitrogen, and oxygen 1s NEXAFS spectra of
any protein or peptide by means of spectral summation over
its amino acid sequence.7Using this approach, the arginine-
?RWWKIWVIRWWR-NH2? ?Ref. 9? was identified as a
suitable candidate for mapping against the blood protein hu-
man serum albumin ?HSA?, via a characteristic transition at
289.3 eV from the guanidine group of arginine. Subsequent
scanning transmission x-ray microscopy ?STXM? experi-
ments of SUB-6 dusted on a background of HSA easily iden-
tified and distinguished between the two biocomponents.7
Charged antimicrobial peptides are of particular interest due
to their ability to disrupt cytoplasmic membranes in bacterial
cells. Thus, locating and mapping these peptides among
other peptides and proteins within cells can aid in under-
standing the mechanisms of antimicrobial action and resis-
We are interested in using soft x-ray spectromicroscopy
techniques to visualize proteins or peptides in complex ma-
trices that model environments such as biofilms or initial
interactions of biological components with existing candi-
date biomaterials. Previous studies of blood protein adsorp-
tion to polymeric phase-segregated polystyrene ?PS?—
poly?methyl methacrylate? ?PMMA? thin films demonstrated
that surface sensitive x-ray photoemission electron micros-
copy ?X-PEEM?10–13is complementary to STXM, with the
capability of probing submonolayer protein-surface interac-
tions. Systematic studies of the concentration dependence,11
a?Author to whom correspondence should be addressed; electronic mail:
FB27FB27Biointerphases 3„2…, June 20081934-8630/2008/3„2…/FB27/9/$23.00©2008 American Vacuum Society
exposure time,11and effect of pH ?Ref. 12? on HSA adsorp-
tion to PS/PMMA films suggest that protein adsorption is
strongly influenced by hydrophobic effects. Moreover, a re-
cent X-PEEM study of competitive adsorption of HSA and
the blood-clotting protein fibrinogen ?Fg? to PS/PMMA
found that the NEXAFS spectra of HSA and Fg are too simi-
lar to be distinguished.13
The aim of this work was to investigate whether X-PEEM
can be used to differentiate SUB-6 and HSA in a mixed
adsorption on a phase-segregated polymer surface and, if so,
to examine the competitive adsorption of protein and pep-
tide. Previously, we had identified the hydrophobic effect as
being important in determining adsorption sites of proteins
?Fg and HSA? on the PS/PMMAsubstrate.10–13At neutral pH
HSA in its native form has a substantial negative charge ?
?−15? while SUB-6 is highly positive ?+5?. By changing the
pH, the charge on these two species can be changed thus
providing a direct probe of potential electrostatic interactions
which might be expected to occur between the protein and
peptide in solution, and thus might play a role in adsorption
to polymer surfaces. The present results, combined with
other studies of competitive protein adsorption,13are part of
a larger program to evaluate the strengths and weaknesses of
soft x-ray spectromicroscopy for mapping similar biocompo-
nents on a complicated background matrix.
Polystyrene ?molecular weight ?MW? of 1.07?10−6, ?
1.06? and poly?methyl methacrylate? ?MW of 312 000, ?
1.01? were purchased from Polymer Source Inc. and used as
received. Thin films were prepared by dissolving PS and
PMMA 30:70 ?wt%/wt%? in toluene and then spun cast onto
clean native oxide silicon wafers ?Si?111?? at 4000 rpm for
40 s. Before spin coating, the Si wafers were vigorously
degreased with 10 min sonication with trichloroethylene, ac-
etone, and methanol followed by extensive rinsing with dou-
bly distilled de-ionized water. The PS/PMMA substrates
were annealed below 160 °C for 12 h in a vacuum oven at a
pressure ?10−5Torr, achieved with a cryotrapped turbo
pump. Noncontact mode atomic force microscopy was used
to verify the topology of the thin films and revealed discrete
PMMA domains approximately 500–1000 nm wide within a
continuous PS phase. The total film thickness was ?40 nm
with a measured corrugation of 4–8 nm. As observed previ-
ously, both the PS and PMMA domains contain micro-
domains of the opposite phase that are approximately 50–
200 nm wide.14
Human serum albumin ?HSA? was obtained from Be-
hringwerke AG, Marburg, Germany, and found to be homo-
geneous as judged by sodium dodecyl sulfate polyacrylamide
gel electrophoresis ?SDS-PAGE?. The SUB-6 peptide
?RWWKIWVIRWWR-NH2? was prepared by Henklein with
9-fluorenylmethyl carbamate solid-phase synthesis and puri-
fied by high-performance liquid chromatography and mass
B. Protein/peptide exposure
The PS/PMMA substrate ?8?8 mm2? was placed in a 50
ml beaker, covered with 5 ml of peptide or protein/peptide
mixture solution ?varied concentrations? for 20 min and then
diluted three times with at least 50 ml of doubly de-ionized
?DDI? water. The substrate was removed and then vigorously
rinsed with DDI water. The protein/peptide adsorbed sample
was then carefully dried by touching the edge of the Si wafer
with lens paper.
The X-PEEM data were collected at the Advanced Light
Source ?ALS? on the PEEM-2 microscope at bending magnet
beamline 7.3.1. Detailed accounts of the experimental appa-
ratus, beamline setup, and instrument optics have been pre-
sented previously.16In brief, photoelectrons and secondary
electrons ejected by absorption of 70%–80% right circularly
polarized ?a mask in the beamline is used to select below
plane ?right circularly polarized? radiation?, monochromatic
x rays are accelerated into an electrostatic imaging column,
where the spatial distribution is magnified and detected by a
charge coupled device ?CCD? camera. X-PEEM is a surface
sensitive technique with a sampling depth ?1/e? of 4 nm for
polymers,17with integrated signals sampling the top 10 nm
of the sample. Image sequences ?stacks18? were recorded at
the C 1s and N 1s edge. A 100-nm-thick Ti filter was used to
eliminate second-order light. To minimize radiation damage,
a fast shutter ?0.1s? was used to block the beam except dur-
ing data acquisition. This reduced the x-ray exposure by 50%
by blocking the beam during the time required to transfer
images from the CCD camera and to step the photon energy.
The incident flux was reduced to about 10% of the full in-
tensity by masking upstream of the monochromator, and a
limited number of energies ?23 in C 1s, 46 in N 1s? and a
short exposure time ?1s? per image were used as other ways
to minimize radiation damage. The field of view was ap-
proximately 20 ?m.
All STXM data were collected at beamline 5.3.2 at the
ALS.19,20Typically, solvent-cast samples with thickness be-
low 100 nm were analyzed by collecting micrometer sized
stacks. Image sequences18were used in homogeneous areas
to minimize radiation damage. An image at a damage sensi-
tive energy was recorded after each spectral measurement to
ensure negligible damage.
III. DATA ANALYSIS
A. Reference spectra
High quality NEXAFS reference spectra of PS, PMMA,
HSA, and SUB-6 peptide7were obtained with STXM on
films floated onto ?PS/PMMA? or solvent cast upon ?HSA,
SUB-6? a silicon nitride window. Furthermore, the linear ab-
sorbance was set to an absolute scale by normalizing outside
the edge region to the elemental response of 1 nm at bulk
FB28Leung et al.: X-ray spectromicroscopy study of competitive adsorptionFB28
Biointerphases, Vol. 3, No. 2, June 2008
densities computed from tables of elemental x-ray absorp-
tion.21Although NEXAFS spectra obtained with STXM are
collected in transmission rather than total electron yield
mode, the spectra from both detection schemes exhibit the
same shapes. The STXM reference spectra were used since
they are better resolved ?energy resolution of 0.1–0.2 eV?
compared to X-PEEM ?energy resolution of 0.4–0.5 eV?.
The reference spectra for PS, PMMA, HSA, and SUB-6
are shown in Fig. 1. The spectrum of each species has char-
acteristic features which allow differentiation from each of
the other components. PS is characterized by a C 1s
strong C 1s→?C=O
eV, respectively. The shift to lower energy for the protein
relative to PMMA arises from the less electronegative envi-
ronment of the carbonyl carbon in R-CONH peptide bonds
compared to that in the R-COOMe ester functional group of
PMMA. The C 1s spectrum of SUB-6 is readily identified
from its strong C 1s→?*?C=N? transition at 289.37?3? eV,
as well as significant intensity in the 286–287 eV region
from C 1s?C−R?→?C=C
transition at 285.15?3? eV. PMMA and HSA exhibit
transitions at 288.45?3? and 288.20?3?
B. Chemical mapping and quantitative analysis
Each image sequence was aligned if needed, normalized
to the ring current, and divided by the Iosignal. The Iospec-
trum is that measured from a piece of HF-etched Si?111?,
corrected by the adsorption spectrum of Si and a linear en-
ergy term which accounts for the bolometric response of
PEEM detection. The energy scale was calibrated by assign-
ing the peak of the C 1s→?C=C
The spectrum at each pixel of the image sequence was fit
to linear combinations of reference spectra using singular
value decomposition, which is an optimized method of per-
transition of PS to 285.15
forming least-squares fits in the cases of highly overdeter-
mined data sets.23,24The resulting fit coefficients are col-
lected into component maps which display the spatial
distribution of each chemical component. Illumination cor-
rection was applied by dividing by a heavily smoothed ver-
sion of the sum of all component maps. Typically, the illu-
mination varied by less than 10% over the area studied. The
intensity scale of each component map was set to an absolute
thickness value by dividing by a scale factor determined by
setting the average total thickness ?determined by adding all
component maps? to 10 nm.17The total thickness maps typi-
cally had residual fluctuations of 15%–20% which were larg-
est in the PMMA domains ?see Fig. 2 of Ref. 12?.
Pixels of areas specific to PS or PMMA were selected by
applying a threshold mask to the PS and PMMA component
maps—i.e., only pixels having a signal above a defined value
are included. The interface was specified as areas not present
in either the PS or PMMA masked areas. The spectrum ex-
tracted from each region ?PS, PMMA, and interface? was
further modified by setting the pre-edge region to zero inten-
sity followed by adjustment of the absorption scale to obtain
a total thickness of 10 nm. The background subtracted and
normalized spectrum extracted from each region was then fit
using the same reference spectra as used to derive the com-
ponent maps. Several stacks were obtained from each sample
and the compositional results from these independent repeat
measurements were averaged to yield the final quantitative
results. The uncertainties cited in the tables are the standard
deviations from these multiple determinations.
IV. RESULTS AND DISCUSSION
A. X-PEEM detection limits
The NEXAFS spectrum of SUB-6 is primarily distin-
guished from those of PS, PMMA, and HSA by the C 1s
1. Careful analysis of the curve-fitting results revealed that
the dip at 286.0 eV in the fit reported by the least-squares
refinement was too large, leading to an overestimation of the
peptide thickness value. Thus, baseline control fitting proce-
dures were performed to verify the minimum detectable
amount of SUB-6. Figure 2 plots spectra of the PS, PMMA,
and interface regions of a pure PS/PMMA substrate along
with the result of optimized curve fits to the PS, PMMA, and
SUB-6 reference spectra. Table I summarizes the quantitative
results. The spectrum obtained from the PS region was domi-
nated by the aromatic transition at 285.15 eV and the small
contribution from PMMA yielded a less intense transition at
288.50 eV that masked the SUB-6 contribution from the
transitions at both 288.50 and 289.37 eV. In contrast, the
intensity for the PMMA and interface regions
allowed for more flexible fits of SUB-6 in the region 287.5–
289.5 eV leading to peptide thickness values close to 1.5?5?
Energy calibration plays a key role in accurate determina-
tion of the protein and peptide components. If the energy
transition at 289.37 eV and also via a small
contribution at 286.0 eV,8shown in Fig.
FIG. 1. C 1s x-ray absorption spectra of polystyrene ?PS, red?, poly?methyl
methacrylate? ?PMMA, green?, human serum albumin ?HSA, blue?, and the
peptide, SUB-6 ?pink? recorded from pure materials. The spectra are plotted
on an absolute linear absorbance scale.
FB29Leung et al.: X-ray spectromicroscopy study of competitive adsorptionFB29
Biointerphases, Vol. 3, No. 2, June 2008
scale is miscalibrated to a lower energy by even 0.05 eV,
then the amount of protein is overestimated. If it is miscali-
brated to a higher energy by a similar amount, the amount of
PMMA is overestimated. Details of this potential systematic
error are presented in Fig. 3. Nonetheless, if appropriate
measures are taken it is possible to reliably calibrate the en-
ergy scale using the 285.15 eV PS peak to better than
0.02 eV. Based on these studies the minimum detectable
amount of HSA was conservatively estimated to be 1.0?5?
nm ?Table I?.
B. Concentration dependence of the specific
adsorption of SUB-6
The adsorption of SUB-6 on the PS/PMMA surface was
probed as a function of concentration to determine the spe-
cific adsorption of the cationic peptide. Three concentrations
were investigated: 1?10−2, 1?10−4, and 1?10−5mg/ml in
de-ionized water, in each case using a fixed adsorption time
of 20 min. Color-coded composite maps are presented in Fig.
4 with two different presentations of the same data. The
maps on the right were rescaled such that the intensity of
each component was mapped separately to the full range of
its color ?0–255?, thereby highlighting the spatial distribution
of the peptide relative to the polymer substrate. The limits of
the R ?PS?, G?PMMA?, B?SUB-6? color bars indicated the
range for each component. The maps on the right are dis-
played on a common thickness scale ?0–10 nm?, which pre-
served the absolute thickness information.
The rescaled maps indicate that the peptide selectively
adsorbed to the interface between PS and PMMA, which was
FIG. 2. Curve fits to the average C 1s spectra extracted from the PS, PMMA,
and interface regions of a pure PS/PMMA blend ?no adsorbed protein or
peptide?. Image sequence recorded using the ALS polymer STXM.
TABLE I. Analysis of C 1s spectra of a clean PS/PMMA substrate with
inclusion of SUB-6 and HSA components to evaluate the minimum detect-
able amount of protein/peptide. Uncertainty is ?0.5 nm.
Thickness of SUB-6 or HSA on PS/PMMA
FIG. 3. Left column is the result of curve fits to the PS/PMMA spectra with
the energy scale shifted 0.05 eV lower than the correct values, while the
right column is the result for an energy scale shifted +0.05 eV. The amounts
of HSA and SUB-6 derived from this analysis represent a worst case result
from energy miscalibration causing a misassignment of adsorbed protein or
peptide, and thus correspond to a conservative evaluation of the detection
FB30Leung et al.: X-ray spectromicroscopy study of competitive adsorptionFB30
Biointerphases, Vol. 3, No. 2, June 2008
previously shown to be the preferred site of adsorption, and
thus the area of the lowest free energy.10The three maps
were virtually identical with the exception of the width of
adsorbed protein at the interface which became successively
wider as the concentration increased. The absolute maps
demonstrate that more peptide was adsorbed at higher con-
centration on the PMMA domains, which are significantly
bluer than the domains at lower concentration.
These results were verified by the quantitative analysis of
spectra extracted from the images by threshold masking
?Table II?. Over the three concentrations sampled, the peptide
exhibited preferential adsorption to both the interface and
PMMA regions over the PS region, which remained strik-
ingly bare. These trends were reversed in comparison to the
adsorption of the negatively charged HSA, which showed
greater preference for the nonpolar PS compared to polar
PMMA.11Since SUB-6 has a positive ?+5? charge ?if all
amino groups are protonated?, it should be significantly more
attracted to the electronegative ester groups of PMMA, thus
explaining the pronounced selectivity for SUB-6 adsorption
on the PMMA domains.
SUB-6 showed the strongest preference for both the inter-
face and the PMMA regions. As the concentration was low-
ered, the adsorption decreased across all regions, especially
on the PS domains. This is not surprising since the free en-
ergy of the system would be decreased more by adsorption
on the interdomain interfaces than on the domains them-
selves, and the highly positive SUB-6 would be strongly at-
tracted to the polar carbonyl groups of PMMA.
C. HSA and SUB-6 competitive adsorption
Figures 5 and 6 show the rescaled and absolute color-
coded component maps of HSA and SUB-6 adsorption on
PS/PMMA. In each map PS is coded in red, PMMA is coded
in green, and either the protein or peptide is coded in blue.
The rescaled maps ?Fig. 5? demonstrate that both HSA and
SUB-6 showed preferential adsorption to the interface,
which was highlighted in blue under all conditions. Compari-
son of the PS regions for the three concentrations revealed a
much more “pink” color arising from the maps of HSA in
comparison to maps of SUB-6, indicative of protein prefer-
ence in this region. At higher peptide concentrations, the
FIG. 4. Color-coded component maps ?left, absolute; right, rescaled? for a
PS/PMMA blend exposed for 20 min to three different concentrations of
SUB-6—?top? 1?10−2mg/ml, ?middle? 1?10−4mg/ml, and ?bottom? 1
TABLE II. Thickness of SUB-6 ?nm/pixel? on the PS and PMMA domains
and at the PS/PMMA interface for adsorption of SUB-6 on a PS/PMMA
blend at three concentrations. Uncertainty is ?0.5 nm.
SUB-6 concentration ?mg/ml?
FIG. 5. Color-coded rescaled component maps for the SUB-6/HSA mix ad-
sorbed to PS/PMMA at ?top? SUB-6 concentration of 1?10−2mg/ml;
?middle? SUB-6 concentration of 1?10−4mg/ml; ?bottom? SUB-6 concen-
tration of 1?10−5mg/ml. Concentration of HSA was unchanged at 0.05
FB31Leung et al.: X-ray spectromicroscopy study of competitive adsorption FB31
Biointerphases, Vol. 3, No. 2, June 2008
domains of PMMA were strongly colored in blue; however,
as the concentration decreased, the color of the PS/PMMA/
HSA and PS/PMMA/SUB-6 maps were comparable, illus-
trating similar site preference for both protein and peptide.
The absolute maps ?Fig. 6? also revealed strong adsorp-
tion by both HSA and SUB-6 at the interface. Most striking
are the very blue domains found for SUB-6 at higher con-
centration. As the concentration decreased, the color on the
domains became lighter, yet preference for PMMA was still
seen. The PS region was contrasted sharply between the pink
color arising from the protein and the very “red” color from
the peptide. These images clearly illustrate the site prefer-
ence of SUB-6 and HSA for the PMMA and PS regions,
Adsorption in the HSA/SUB-6 system is likely to have
been complicated by possible ionic complex formation be-
tween the negatively charged HSA and positively charged
SUB-6. These interactions can lead to protein unfolding and
conformational change.25In fact, dynamic light scattering of
a HSA/amphiphilic drug complex revealed the presence of
micelles in solution.25However, it seems unlikely that the
hydrophilic SUB-6 peptide will form micelles with HSA due
to the evenly spaced positive charge along the peptide chain.
Also, at peptide concentrations below 1?10−4mg/ml
SUB-6, the number of HSA molecules greatly exceeded the
number of SUB-6 molecules ?in solution the concentration of
HSA was 7.6?10−7M while SUB-6 was only 5.5?10−9M?.
Nonetheless, SUB-6 remained highly surface active, as
shown in Sec. IV B. Thus, for the mixture of coadsorbing
HSA/SUB-6 two competing processes can occur: ?1? indi-
vidual adsorption of the protein or peptide molecules and ?2?
adsorption of a protein-peptide complex formed in solution
by electrostatic attraction.
This complexity was most evident in the PMMA region
where the two processes likely occurred simultaneously
TABLE III. Thickness of SUB-6 and HSA ?nm/pixel? on the PS and PMMA
domains and at the PS/PMMA interface for adsorption from a mixed SUB-
6/HSA solution on a PS/PMMA blend at three SUB-6 concentrations. HSA
concentration held constant at 0.05 mg/ml. Uncertainty is ?0.5 nm.
Concentration of SUB-6 ??HSA?=0.05 mg/ml?
PS PS 7.5
FIG. 6. Color-coded nonrescaled ?absolute? component maps for the SUB-
6/HSA mix adsorbed to PS/PMMA at ?top? SUB-6 concentration of 1
?10−2mg/ml; ?middle? SUB-6 concentration of 1?10−4mg/ml; ?bottom?
SUB-6 concentration of 1?10−5mg/ml. Concentration of HSA was un-
changed at 0.05 mg/ml
FIG. 7. Curve fits of the C 1s spectra extracted from the PS, PMMA, and
interface regions for adsorption from the mixture of 0.05 mg/ml HSA with
0.01 mg/ml SUB-6 at pH ?PS in red, PMMA in green, HSA in blue, and
SUB-6 in pink?
FB32 Leung et al.: X-ray spectromicroscopy study of competitive adsorptionFB32
Biointerphases, Vol. 3, No. 2, June 2008
?Table III, Fig. 7?. In this region, substantial amounts of both
HSA and SUB-6 were detected in a competitive ?individual?
adsorption-type process; however, the thickness of HSA ad-
sorbed from the biomixture was markedly lower than when
the polymer surface was exposed to HSA alone. SUB-6 was
highly surface active in the PMMA region which may ac-
count for lower surface area availability, leading to decreased
HSA adsorption. In contrast, adsorption of SUB-6 from the
mixture was sharply elevated compared to adsorption of
SUB-6 by itself. Thus, the quantitative analyses suggest that
an electrostatic complex must be present, and that the com-
plex had a high affinity for the PS/PMMA surface to account
for the higher thickness of adsorbed peptide. In this case,
based on charge considerations, a ?HSA?SUB−6?3? complex
is likely to have formed, leading to a greater thickness of
peptide. Studies are underway to characterize this postulated
complex via gel electrophoresis ?SDS-PAGE?, mass spec-
trometry, and high field nmr.
The interpretation of the adsorption on the PS region was
simpler. Since SUB-6 adsorption to PS was low, the domi-
nant mechanism is likely adsorption by the protein/peptide
complex. This explanation was supported by the lower ad-
sorption thickness of HSA from the biomixture compared to
adsorption of HSA itself and also slightly higher adsorption
values for SUB-6 from the mixture compared to the pure
peptide solution. Since a charge neutralizing complex would
be formed, the negative charge of HSA was decreased.
Hence, the adsorbed thickness of approximately 1.6?5? nm
was much closer to the values recorded for the protein at the
isoelectric point ?at pH 4, HSA thickness ?1.4?5? nm?.12In
addition, since the complex formation likely led to confor-
mational change in the protein, the new HSA/SUB-6 species
was probably more hydrophilic, also leading to less protein
adsorption on the hydrophobic PS region.
At the interdomain interface, the dominant mechanism is
considered to be individual peptide/protein competitive ad-
sorption which arises from favorable thermodynamic and ki-
netic considerations. Both HSA and SUB-6 selectively ad-
sorb to the interdomain region; however, this region is small
and easily saturated.10Quantitative analyses indicate that the
electrostatic complex was a minor adsorbed component in
this region since the thickness of SUB-6 adsorbed from the
biomixture was much closer to the value for adsorption from
the pure peptide solution than was the case at the PMMA
region. As the concentration of peptide was decreased, the
thickness of protein systematically increased until the thick-
ness was comparable to the pure protein solution adsorption
Although the adsorption of the biocomponents in this sys-
tem is proposed to be primarily driven by hydrophobic inter-
actions, since electrostatic double layers are weak surround-
ing the uncharged substrate, there is some evidence for small
contributions to the preferential adsorption from charged in-
teractions of the peptide For example, the positively charged
SUB-6 showed a much higher preference for the polar car-
bonyl groups of PMMA. In contrast, the adsorption behavior
of the “soft” HSA is known to be driven by gains in confor-
Due to the combination of electrostatic attraction and the
unique aspects of the NEXAFS spectrum of SUB-6,
X-PEEM was capable of detecting the peptide at concentra-
tions as low as 5?10−9M ?5 nM or 0.1 ng/ml?. This ex-
tremely low detection limit exceeds the current detection
limits of several surface sensitive techniques such as
MALDI-MS ?0.6 ng/ml?,27surface plasmon resonance ?50
ng/ml?,28and resonance light scattering ?15 ng/ml?.29It is
worth noting that X-PEEM provides this very low averaged
detection limit while at the same time providing high ?50
nm? lateral spatial resolution in the surface distributions.
NEXAFS is also capable of detecting the four chemical com-
ponents of this system ?PS, PMMA, HSA, and SUB-6? si-
multaneously in a quantitative fashion as is clearly demon-
strated to the least-squares fits to the extracted spectra ?Fig.
7?. Careful analyses of these spectral fits reveal that removal
of either the peptide or the protein component leads to a
visible misfit and substantial increase in the residual spec-
D. Effect of pH
1. Changes in SUB-6 adsorption
To evaluate the electrostatic effect of SUB-6 adsorption,
the pH was raised to neutralize the positive charge on the
peptide. The acid dissociation constant for the guanidine
group of arginine ranges from log K1?10.78 to 15.0, de-
pending on the ionic strength.30At extremely alkaline pH,
the PS/PMMA film deteriorates; thus, all measurements were
carried out at pH 11.8.
Figure 8 compares the color-coded composite maps de-
rived from 1?10−2mg/ml SUB-6 adsorbed onto the PS/
PMMA blend surface at neutral and alkaline pH. The res-
FIG. 8. Color-coded component maps ?left, rescaled; right absolute? for ad-
sorption on a PS/PMMA blend surface from a solution of pure SUB-6
?1?10−2mg/ml? at ?top? neutral pH ?7.2? and ?bottom? alkaline pH ?11.8?.
FB33Leung et al.: X-ray spectromicroscopy study of competitive adsorptionFB33
Biointerphases, Vol. 3, No. 2, June 2008
caled images demonstrate that the interface was still the site
of preferential protein adsorption at alkaline pH; however,
the absolute images clearly reveal that at basic pH the
amount of SUB-6 on the PS regions had increased substan-
tially, since the PS regions shows a much more pink color
than at neutral pH. In contrast, the amounts of SUB-6 on the
PMMA domains and at the interface remained rather similar.
The quantitative spectral analyses support the results from
the mapping ?Table IV?. At alkaline pH, the thickness of
SUB-6 on the PS region had substantially increased to 2.2?5?
nm, in contrast to the 0.8?5? nm thickness at neutral pH.
Interestingly, the adsorbed thickness on both the PMMA and
interface regions remained the same within the limits of un-
certainty. As expected, the preferential adsorption at the in-
terdomain region prevailed regardless of charge. A slight de-
crease in adsorbed PMMA thickness was expected; however,
at pH 11.8 not all of the protons from the guanidine groups
would be removed and thus it is proposed that a fraction of
the positively charged peptide adsorbed at the PMMA re-
2. Redistribution of SUB-6 and HSA
The coadsorption of SUB-6 ?1?10−2mg/ml? and HSA
?0.05 mg/ml? was also examined at alkaline pH ?pH=11.8?.
It is assumed that at this pH most of the protons from the
guanidine group of arginine are removed, suppressing the
formation of the electrostatic complex. Thus, at high pH, the
mode of competitive adsorption was based mainly on the
individual properties of the HSA protein and SUB-6 peptide.
The color-coded composite maps are shown with both res-
caled and absolute presentations in Fig. 9. The rescaled maps
reveal the preference of both SUB-6 and HSA for the inter-
face. However, the absolute maps show that there was a de-
crease in protein/peptide adsorption on the PMMA domains,
hence there was a redistribution of the biocomponents.
The results from the spectral fits showed an increase in
the thickness of SUB-6 adsorbed at the PS region, which
paralleled the increased adsorption of SUB-6 on the PS re-
gion for the pure peptide alkaline solution. The correspond-
ing HSA thickness on PS was similar to the thickness ad-
sorbed at neutral pH ?Table IV?. More importantly, this
thickness value was within the uncertainty of measurements
of the adsorption of pure HSA at pH 10.0 to PS/PMMA at
the PS region.11
At the PMMA and interdomain regions, the charge neu-
tralizing effect resulting in the elimination of the electrostatic
complex was evident. At the PMMA domains the thickness
of adsorbed peptide was similar to the amount of peptide
adsorbed from the pure SUB-6 solution. Nonetheless, HSA
adsorption was substantially lower at alkaline pH than at
neutral pH which suggested that the PMMA domains were
primarily covered by the peptide, leaving a smaller amount
of surface area for HSA adsorption.
As previously discussed, the interdomain region is easily
saturated; therefore, the dominant adsorption mechanism was
individual competitive adsorption. As illustrated from the
quantitative analyses, the adsorption of both HSAand SUB-6
decreased slightly as the pH was elevated; however, since
TABLE IV. Thickness of SUB-6 and SUB-6/HSA mix ?nm/pixel? on the PS
and PMMA domains and at the PS/PMMA interface at alkaline pH. Uncer-
tainty is ?0.5 nm.
SUB-6 ?pH 11.8?
SUB-6/HSA mix ?pH 11.8?
FIG. 9. Color-coded component maps ?upper, rescaled; lower, absolute? for
adsorption on a PS/PMMA blend surface from solutions of a mixture of
SUB-6 ?1?10−2mg/ml? and HSA ?0.05 mg/ml? at ?top? neutral pH ?7.2?
and ?bottom? alkaline pH ?11.8?.
FB34Leung et al.: X-ray spectromicroscopy study of competitive adsorption FB34
Biointerphases, Vol. 3, No. 2, June 2008
the HSA adsorbed thickness did not drastically decrease ?re- Download full-text
moval of protein-peptide attraction?, the previous conclusion
that complex formation is only a minor component was sup-
We have shown that X-PEEM is capable of differentiating
and mapping HSA protein and SUB-6 peptide adsorbed on a
phase-segregated PS/PMMA polymer blend surface from a
mixture at very low concentrations, as low as 5?10−9
mol/l or 0.1 ng/ml in the SUB-6 peptide. The preferential
adsorption of the peptide was elucidated both in pure SUB-6
solutions and in a mixture of protein and peptide at neutral
and alkaline pH. The results were explained in terms of for-
mation of an electrostatic complex between the positively
charged peptide and the negatively charged protein, which
has an adsorption behavior which depends on the polarity of
the different domains of the PS/PMMA substrate. As the pH
was raised to neutralize the charge on the peptide, complex
formation was suppressed and a more competitive-type ad-
sorption process was observed.
This research was supported by the Natural Science and
Engineering Research Council ?NSERC, Canada?, AFMNet
and the Canada Research Chair programs. X-ray microscopy
was carried out using PEEM2 at the ALS. The ALS is sup-
ported by the U.S. Department of Energy under Contract No.
1H. Ade, Experimental Methods in the Physical Sciences ?Academic, New
York, 1998?, Chap. 32.
2H. Ade and S. G. Urquhart, Chemical Applications of Synchrotron Radia-
tion ?World Scientific, Singapore, 2002?, pp. 285–355.
3H. Ade and A. P. Hitchcock, Polymer 49, 643 ?2008?.
4J. R. Lawrence, G. D. Swerhone, G. G. Leppard, T. Araki, X. Zhang, M.
M. West, and A. P. Hitchcock, Appl. Environ. Microbiol. 69, 5543
5K. Kaznacheyev, A. Osanna, C. Jacobsen, O. Plashkevych, O. Vahtras, H.
Ågren, V. Carravetta, and A. P. Hitchcock, J. Phys. Chem. A 106, 3153
6Y. Zubavichus, A. Shaporenko, M. Grunze, and M. Zharnikov, J. Phys.
Chem. A 109, 6998 ?2005?.
7J. Stewart-Ornstein, A. P. Hitchcock, D. Hernandez-Cruz, P. Henklein, J.
Overhage, K. Hilpert, J. D. Hale, and R. E. W. Hancock, J. Phys. Chem.
B 111, 7691 ?2007?.
8Y. Zubavichus, A. Shaporenko, M. Grunze, and M. Zharnikov, J. Phys.
Chem. C ?to be published?.
9K. Hilpert, R. Volkmer-Engert, T. Walter, and R. E. W. Hancock, Nat.
Biotechnol. 23, 1008 ?2005?.
10C. Morin, A. P. Hitchcock, R. M. Cornelius, J. L. Brash, S. G. Urquhart,
A. Scholl, and A. Doran, J. Electron Spectrosc. Relat. Phenom. 137–140,
11L. Li, A. P. Hitchcock, N. Robar, R. Cornelius, J. L. Brash, A. Scholl, and
A. Doran, J. Phys. Chem. B 110, 16763 ?2006?.
12L. Li, A. P. Hitchcock, R. Cornelius, J. L. Brash, A. Scholl, and A. Doran,
J. Phys. Chem. B 112, 2150 ?2008?.
13L. Li, J. L. Brash, R. Cornelius, A. Scholl, A. Doran, and A. P. Hitchcock
14C. Morin et al., J. Electron Spectrosc. Relat. Phenom. 121, 203 ?2001?.
15D. Romeo, B. Skerlavaj, M. Bolognesi, and R. Gennaro, J. Biol. Chem.
263, 9573 ?1988?.
16S. Anders et al., Rev. Sci. Instrum. 70, 3973 ?1999?.
17J. Wang, L. Li, C. Morin, A. P. Hitchcock, A. Doran, and A. A. Scholl, J.
Electron Spectrosc. Relat. Phenom. ?to be published?.
18C. J. Jacobsen, S. Wirick, G. Flynn, and C. Zimba, J. Microsc. 197, 173
19T. Warwick et al., J. Synchrotron Radiat. 9, 254 ?2002?.
20A. L. D. Kilcoyne et al., J. Synchrotron Radiat. 10, 125 ?2003?.
21B. L. Henke, E. M. Gullikson, and J. C. Davis, At. Data Nucl. Data Tables
54, 181 ?1993?.
22aXis2000 is free for noncommercial use. It is written in Interactive Data
23G. Strang, Linear Algebra and its Applications ?Harcourt Brace Jovano-
vitch, San Diego, 1988?.
24I. N. Koprinarov, A. P. Hitchcock, C. T. McCrory, and R. F. Childs, J.
Phys. Chem. B 106, 5358 ?2002?.
25D. Leis, S. Barbosa, D. Attwood, P. Toboada, and V. Mosquera, J. Phys.
Chem. B 106, 9143 ?2002?.
26K. Nakanishi, T. Sakiyama, and K. Imamura, J. Biosci. Bioeng. 91, 233
27Y. Xu, J. Throck Watson, and M. L. Bruening, Anal. Chem. 75, 185
28M. Suzuki, F. Ozawa, W. Sugimoto, and S. Aso, Anal. Bioanal. Chem.
372, 301 ?2002?.
29L. Wang, H. Chen, L. Li, T. Xia, L. Dong, and L. Wang, Spectrochim.
Acta, Part A 60, 747 ?2004?.
30O. Yamauchi and A. Odani, Pure Appl. Chem. 68, 469 ?1996?.
FB35 Leung et al.: X-ray spectromicroscopy study of competitive adsorption FB35
Biointerphases, Vol. 3, No. 2, June 2008