Adsorption and orientation of the physiological extracellular peptide glutathione disulfide on surface functionalized colloidal alumina particles

Article abstract of DOI:10.1021/ja401590c

Understanding the interrelation between surface chemistry of colloidal particles and surface adsorption of biomolecules is a crucial prerequisite for the design of materials for biotechnological and nanomedical applications. Here, we elucidate how tailoring the surface chemistry of colloidal alumina particles (d₅₀ = 180 nm) with amino (-NH₂), carboxylate (-COOH), phosphate (-PO₃H₂) or sulfonate (-SO₃H) groups affects adsorption and orientation of the model peptide glutathione disulfide (GSSG). GSSG adsorbed on native, -NH₂-functionalized, and -SO ₃H-functionalized alumina but not on -COOH- and -PO₃H ₂-functionalized particles. When adsorption occurred, the process was rapid (≤5 min), reversible by application of salts, and followed a Langmuir adsorption isotherm dependent on the particle surface functionalization and ζ potential. The orientation of particle bound GSSG was assessed by the release of glutathione after reducing the GSSG disulfide bond and by ζ potential measurements. GSSG is likely to bind via the carboxylate groups of one of its two glutathionyl (GS) moieties onto native and -NH₂-modified alumina, whereas GSSG is suggested to bind to -SO₃H-modified alumina via the primary amino groups of both GS moieties. Thus, GSSG adsorption and orientation can be tailored by varying the molecular composition of the particle surface, demonstrating a step toward guiding interactions of biomolecules with colloidal particles.

Full text of DOI:10.1021/ja401590c

Article
pubs.acs.org/JACS
Adsorption and Orientation of the Physiological Extracellular
Peptide Glutathione Disulfide on Surface Functionalized Colloidal
Alumina Particles
Fabian Meder, Henrik Hintz, Yvonne Koehler, Maike M. Schmidt,‡,§ Laura Treccani,†
†
,∥
‡,∥
‡,§
Ralf Dringen,‡,§ and Kurosch Rezwan*,†
†
‡
§
Faculty of Production Engineering, Advanced Ceramics, Center for Biomolecular Interactions Bremen, and Center for
Environmental Research and Sustainable Technology, University of Bremen, Bremen, Germany
*
S Supporting Information
ABSTRACT: Understanding the interrelation between surface
chemistry of colloidal particles and surface adsorption of
biomolecules is a crucial prerequisite for the design of materials
for biotechnological and nanomedical applications. Here, we
elucidate how tailoring the surface chemistry of colloidal alumina
particles (d50 = 180 nm) with amino (−NH ), carboxylate
2
(
−COOH), phosphate (−PO H ) or sulfonate (−SO H) groups
3
2
3
affects adsorption and orientation of the model peptide
glutathione disulfide (GSSG). GSSG adsorbed on native, −NH -
2
functionalized, and −SO H-functionalized alumina but not on
3
−
COOH- and −PO H -functionalized particles. When adsorption occurred, the process was rapid (≤5 min), reversible by
3
2
application of salts, and followed a Langmuir adsorption isotherm dependent on the particle surface functionalization and ζ
potential. The orientation of particle bound GSSG was assessed by the release of glutathione after reducing the GSSG disulfide
bond and by ζ potential measurements. GSSG is likely to bind via the carboxylate groups of one of its two glutathionyl (GS)
moieties onto native and −NH -modified alumina, whereas GSSG is suggested to bind to −SO H-modified alumina via the
2
3
primary amino groups of both GS moieties. Thus, GSSG adsorption and orientation can be tailored by varying the molecular
composition of the particle surface, demonstrating a step toward guiding interactions of biomolecules with colloidal particles.
1
. INTRODUCTION
the particle surface itself with specific properties to induce the
desired interaction (attraction/repulsion of the peptide).
Surface functionalization allows to tailor the surface proper-
19
Inorganic colloidal particles become coated with biomolecules
like peptides and proteins when exposed to biological fluids.
1
,2
ties of colloidal particles and to facilitate the attachment of
How the particle surface chemistry guides the adsorption and
orientation of the biomolecules on the particle surface is yet not
well understood but crucial for the tailored design and
successful application of submicrometer- and nanoparticles in
many biotechnological and nanomedical applications and for
+
−
charged groups (e.g., −NH , −SO ) onto almost any metal
3
3
oxide particle. Such particle surface groups can interact with
oppositely charged carboxylate or amino groups in pep-
9
,20−23
tides.
However, peptide adsorption deviates from
1
−3
predictions by just considering the net charges of particles
toxicological considerations.
1
5,21,23
and peptides.
Due the multifunctionality of peptides
Peptide−particle interactions are a model system for the
(
e.g., coexistence of cationic and anionic surface groups), it is
more complex protein−particle interactions and, moreover,
important for biosensing applications, biomineralization,
antibacterial surface modification, cell-material adhesion,
and controlled drug release. Peptide adsorption onto particle
surfaces are, nevertheless, hard to control as contributions of
either electrostatic, hydrophobic interactions or hydrogen
4
5
7
crucial, but often unclear, which particle functional groups and
peptide residues indeed interact on a molecular level during
peptide-particle adsorption.
6
8
Colloidal alumina particles are an ideal substrate for stable
21,22,24
and tunable surface functionalizations
and thus an
9
−13
appreciated model system to study peptide-particle adsorption
as a function of different surface chemistry. Alumina is
moreover widely used as a substrate material for biomedical
bonding,
the main driving forces of peptide-particle
2
,10,14
interactions, are often not fully enlightened.
is becoming apparent that peptides “recognize” particle surface
properties and bind material specific and selective toward
However, it
2
5
and biotechnological applications, including, e.g., biosensing,
26
27
1
5−17
protein separation and purification systems, biocatalysis, or
different oxide particles.
Such peptides with affinity toward
a specific material can be obtained by phage display
1
7,18
techniques.
In contrast, to promote or avoid binding of a
Received: February 13, 2013
Published: April 8, 2013
given peptide onto an oxide particle, it is necessary to engineer
©
2013 American Chemical Society
6307
dx.doi.org/10.1021/ja401590c | J. Am. Chem. Soc. 2013, 135, 6307−6316

Journal of the American Chemical Society
Article
2
8,29
drug delivery applications.
The alumina particle features
purchased from Krahn Chemie (Germany). For surface functionaliza-
tion we used 3-aminopropyltriethoxysilane (99% in ethanol) for
surface hydroxyl groups (Al−OH) that can be used to
−
NH groups and pyrophosphoric acid (≥90%) for −PO H groups,
functionalize the particle surface with, e.g., −NH or −SO H
2
3
2
2
3
both from Sigma-Aldrich (Germany). 3-(triethoxysilyl)-
groups. On the other hand, the Al−OH groups are itself
propylsuccinicanhydride, >94%, was used to generate −COOH groups
positively charged sites on the particle surface when they
and 3-(trihydroxysilyl)-1-propanesulfonic acid, 30−35% in H O for
+
2
protonate (Al−OH ) below the particles’ isoelectric point
2
−
SO H groups, both purchased from ABCR (Germany). 5,5′-dithio-
22
3
(
IEP) of about 9.
bis(2-nitrobenzoic acid) (DTNB), ethylenediaminetetraacetic acid
EDTA), and GSSG were obtained from Sigma-Aldrich (Germany).
An important physiological peptide is glutathione disulfide
GSSG; γ-GluCysGly disulfide) that derives from the oxidation
of glutathione (GSH; γ-GluCysGly) and is the disulfide of two
GSH molecules. The GSH/GSSG redox pair is a highly
important redox system in eukaryotic cells, and GSH is a
necessary molecule in the defense of cells against oxidative
(
(
Dithiothreitol (DTT), nicotinamide adenine dinucleotide phosphate
(NADPH), and sulfosalicylic acid were obtained from AppliChem
1
1
(
Germany). Glutathione reductase was purchased from Roche
Diagnostics (Germany). All other chemicals were obtained from
Fluka (Switzerland) or Merck (Germany) at analytical grade.
3
0,31
2.2. Methods. 2.2.1. Surface Functionalization of Al O Particles.
2 3
stress and toxins.
Intact mammalian cells contain hardly any
Alumina particles were functionalized as described previously by
GSSG. In defending oxidative stress, GSH is oxidized to GSSG
and then either regenerated by reduction or efficiently exported
by multidrug resistance proteins, as shown for various cell
GSSG export helps to prevent a toxic intracellular
GSSG surge and leads to substantial extracellular GSSG
concentrations, that partially also derive from extracellular
22
Meder et al. Briefly, suspensions of α-alumina particles, prepared by
mixing 15 g particles with 50 mL double-deionized water (ddH O)
2
with an electrical resistance of 18 MΩ × cm (Synergy, Millipore,
Germany), were deagglomerated by sonication for 10 min with an
ultrasound horn Sonifier 450 (Branson, U.S.A., output: 150 W, pulse
1
1,32−34
types.
rate: 0.5 s). The precursors were mixed with 50 mL ddH O to a final
2
3
2
2
concentration of 0.09 M (this corresponds to 26 μmol/m , normalized
to the specific surface area of the unmodified Al O particles).
GSH oxidation. In body fluids, such as blood serum, GSSG
35
2
3
concentrations of up to 150 μM are found. Materials that
come into contact with serum in biomedical and biotechno-
logical applications thus likely encounter extracellular GSSG.
GSSG, as a marker for oxidative stress, is furthermore a targeted
Subsequently, the precursor solution was added to the particle
suspensions and stirred for 60 min at 25 °C and was then heated for
9
0 min at 115 °C. Afterward, the particles were separated by
centrifugation and washed three times in 100 mL ddH O to remove
3
6,37
2
peptide in several biosensor applications.
any unreacted precursors. Then the particles were freeze-dried for 96 h
GSSG is an ideal biomolecule for the investigation of the
adsorption and desorption behavior on surface functionalized
particles. GSSG contains a disulfide bridge in addition to
carboxylate and amino groups and has been therefore
considered as a suitable model peptide for small proteins and
epitopes of larger proteins involved in protein-particle bind-
at −20 °C in vacuum using a freeze-dryer P8KE-80 (Piatkowski
̈
Forschungsgerate, Germany) and subsequent heat treated at 70 °C for
2 h. The native alumina was calcined at 400 °C for 4 h followed by
autoclaving for 15 min at 121 °C and 2.05 bar (Systec 2540ELV,
11
Systec, Germany) as a sterilization step, according to Dringen et al.
.2.2. Particle Characterization. The ζ potential measurements
were conducted with an electroacoustic spectrometer DT1200
Dispersion Technology Inc., U.S.A.). Isoelectric points were
2
1
1
ing. In addition, the availability of a specific enzymatic cycling
assay for the quantification of picomole amounts of GSSG
allows a rapid and highly sensitive determination of even low
quantities of GSSG that have adsorbed or desorbed from
(
determined by pH titrations using 1 M HCl or 1 M KOH. All
measurements were conducted with suspensions containing 1 vol % of
particles. Average and standard deviations were calculated from three
independent measurements. This procedure was also used to
determine the ζ potential and IEP of particles with GSSG.
Specific surface areas were calculated from nitrogen adsorption
isotherms using the Brunauer, Emmet, and Teller (BET) equation.
Adsorption isotherms have been recorded at −196 °C with an
automated surface area analyzer BELSORP-mini (Bel Inc., Japan). All
samples were gassed out at 120 °C under vacuum for 3 h before the
measurement.
Particle size was measured by dynamic light scattering with an
Ultrafine Particle Analyzer UPA 150 (Microtrac Inc., USA) using
aqueous suspensions with 0.1 vol % particle content. The average and
standard deviations were calculated from four to nine single
measurements of the d50 (intensity).
38
moderate amounts of particles. GSSG adsorption onto native
colloidal alumina particles was recently studied by Dringen et
al., who found that GSSG strongly adsorbs onto alumina via
electrostatic interactions between carboxylate groups of one
11
GSSG moiety and the positively charged alumina surface.
In this study, we investigate how the GSSG adsorption and
orientation on alumina particles can be modulated by tailoring
the particle surface with acidic (−COOH, −SO H, −PO H )
3
3
2
and basic (−NH ) surface groups. We have previously shown
2
that such surface functionalizations direct protein adsorption
2
1,22
processes.
This study goes beyond and here, the GSSG-
particle binding mechanisms, contribution of particles and
GSSG functional surface groups, and its orientation on the
particle surface could be elucidated on a molecular level by
studying GSSG adsorption on well-characterized surface
functionalized particles in combination with ζ potential
measurement and analysis of the GSSG disulfide bond
reduction on the particle surfaces. Therefore the results
broaden fundamental understanding of protein−surface inter-
actions and are valuable for developing surface design strategies
for colloidal particles in biomedical and technological
applications.
Elemental analysis was carried out by Mikrolabor Pascher
(
Germany), and element concentrations were used to calculate the
amounts of functional groups related to the specific surface area of
unfunctionalized alumina particles (11.5 m /g). Inductive coupled
2
plasma atomic emission spectroscopy was applied to determine the
sulfur and phosphor content of the functionalized particles after acid
pressure hydrolysis in an ICAP 6500 (Thermo, Germany). To quantify
the amino groups, the nitrogen content was determined using the
3
9
Dumas method. Carboxylate groups were determined by measuring
the carbon concentration by conductometric carbon dioxide
determination after combustion. The lower detection limit of the
elemental analysis was 0.01 wt % for S, P, and C and 0.1 wt % for N.
The analysis was conducted in duplicates. Further characterizations of
the functionalized particles including transmission electron micros-
copy, water vapor, and n-heptane adsorption to determine the
hydrophilic and hydrophobic properties of the particles can be found
2. MATERIALS AND METHODS
2
.1. Materials. Polycrystalline α-alumina particles (Al O ; TM-
2 3
DAR, lot. no. 8086, high-purity alumina >99.99%, d50 = 179 ± 8 nm,
3
2
22
density = 3.98 g/cm , specific surface area = 11.5 m /g) were
in Meder et al.
6
308
dx.doi.org/10.1021/ja401590c | J. Am. Chem. Soc. 2013, 135, 6307−6316

Journal of the American Chemical Society
Article
Table 1. Physicochemical Properties of Functionalized and Unfunctionalized Alumina Particles
2
.2.3. Incubation of GSSG with the Particles. Aqueous particle
2.2.6. Determination of GSSG and GSH. GSSG was quantified as
40,41
suspensions (1 vol %) were prepared by mixing 398 mg particles in 9.9
mL ddH O. The suspensions were then deagglomerated for 10 min
described previously
in microtiter plates according to the
38
colorimetric method originally described by Tietze et al. Briefly, 10
2
using an ultrasound horn Sonifier 450 (Branson, U.S.A., output: 150
W, pulse rate: 0.5 s). The as-prepared suspensions were subsequently
incubated with GSSG under the conditions indicated for the individual
experiments. All incubations were performed at room temperature (25
μL of the supernatant was mixed with 10 μL 1% (w/v) sulfosalicylic
acid and diluted with 80 μL ddH O in a well of a microtiter plate. The
2
reaction was started by the addition of 100 μL reaction mixture (0.06
U glutathione reductase, 0.3 mM DTNB, 0.4 mM NADPH, 1 mM
EDTA in 0.1 M sodium phosphate buffer, pH 7.5), and the increase in
absorbance due to the formation of 5-thio-2-nitrobenzoate was
followed at 405 nm using a microtiter plate reader (Sunrise, Tecan,
Austria). GSSG contents were calculated by comparison to values
obtained for GSSG standards. This assay detects GSH in addition to
°
C). Particle sedimentation was prevented by permanent mixing using
a rotator drive STR4 (Staffordshire, USA) at a speed of 13 rpm
adsorption/desorption experiments) or by using a horizontal shaker
(
Unimax 1011 (Heidolph Instruments, Germany).
For the adsorption experiments, 50 μM GSSG, or the GSSG
concentration stated in the figure, was incubated with 0.1 vol %
3
8
GSSG. Therefore, the amount of GSH that was adsorbed to the
particles after treatment of GSSG-bound particles with DTT was also
quantified by the GSSG assay, considering that a given GSH amount
gives only half of the signal of the respective GSSG amount.
particle suspension in 1 mL ddH O. The pH for the prepared
2
suspensions was 4.8 for Al O , 6.4 for Al O −NH , 4.3 for Al O −
2
3
2
3
2
2
3
COOH, 4.2 for Al O −SO H, and 4.1 for Al O −PO H . After
2
3
3
2
3
3
2
2
.2.7. Adsorption Isotherm Fit. The experimental adsorption
different incubation time intervals (60 min or as stated in the figures),
the samples were centrifuged for 10 min at 12 100 g (Minispin,
Eppendorf, Germany), and the remaining GSSG in the supernatant of
the solution was determined.
isotherm was fitted using the Langmuir equation:
Γ
*c
+ c
max
Γ = K
−
1
To quantify the particle bound GSSG, the GSSG was released from
L
1
1
the particle surface using a procedure adapted from Dringen et al.
where Γ and Γ are the adsorbed and maximal adsorbed quantities,
max
Therefore, the pellet, received in the above-described centrifugation
−1
respectively, c is the GSSG concentration in solution, and K is the
Langmuir constant of the system. KL was then used to calculate the
standard Gibbs energy of adsorption ΔG as previously described
using the equation:
L
step, was washed with 1 mL ddH O and centrifuged again (5 min, 12
−1
2
1
1
00 g). This pellet was then resuspended in 1 mL 10 mM KOH (pH
2.3) by vortexing and directly centrifuged (10 min, 12 100 g).
0
ads
42,43
Incubation of the particles in 10 mM KOH for up to 60 min (as done
11
in the procedure of Dringen et al.) did not further increase the
amount of released GSSG. The KOH treatment was sufficient to
completely release the particle bound GSSG, which was then
quantified.
⎛
⎞
⎟
⎠
0
csolv
ΔG = −RT ln⎜
ads
−1
⎝
K
L
where R is the gas constant, T is the standard ambient temperature,
and csolv is the molar concentration of the solvent water (55.5 mol/L).
2
.2.4. Reduction of Particle Bound GSSG. The disulfide bonds of
particle bound GSSG were reduced using DTT. Therefore, 0.1 vol %
particle suspensions were preincubated for 30 min with 1 mM GSSG.
After centrifugation (10 min at 12 100 g) and washing with 1 mL
ddH O, the particle pellets were dispersed in 1 mL ddH O containing
2
.2.8. Modeling of GSSG Properties. The partial charge distribution
of GSSG as function of the pH was calculated from equilibrium
constants predicted by linear free energy relationships using ACD/
LabsI-Lab 2.0, version 5.0.0.184, Advanced Chemistry Development,
2
2
no DTT or 20 mM DTT. After 60 min of incubation, the particles
were washed, and the bound GSSG or GSH was released with KOH as
described above. The samples for the ζ potential measurements were
prepared likewise using a 1 vol % particle suspension to achieve a
reliable signal. The concentrations of GSSG and DTT were
accordingly upscaled to 10 mM GSSG and 200 mM DTT.
44
Inc., Toronto, ON, Canada, www.acdlabs.com. To calculate the
surface potential distribution, the chemical structure information for
GSSG available in the PubChem substance and compound database
45
through the substance identifier number CID: 44630308 was used,
and Gasteiger partial charges were assigned using Vega ZZ (version
46
3
.0.1). The GSSG surface potential at neutral pH and ionic strength
2
.2.5. Effect of Ions on Particle Bound GSSG. To investigate the
of 1 mM at 25 °C was calculated using the adaptive Poisson−
potential of ions to desorb GSSG from the particles, 1 mM GSSG were
preincubated with 0.1 vol % particle suspensions for 30 min in 1 mL
total volume. After centrifugation (10 min, 12 100 g) the pellet was
washed with 1 mL ddH O, dispersed in 1 mL ddH O or in ddH O
47
Boltzmann solver, and isosurfaces of negative and positive potentials
represent equipotentials of ±30 mV visualized using Visual Molecular
48
Dynamics (version 1.8.7).
2
2
2
containing NaCl, Na SO , or (NH ) SO in concentrations of up to
2
4
4
2
4
3
. RESULTS
100 mM and subsequently incubated for 60 min. The pH of the
ddH O and salt solutions was 6.2. After centrifugation, the supernatant
was harvested, and the desorbed GSSG was determined.
3.1. Characteristics of Surface Functionalized Par-
ticles. Table 1 shows the physicochemical properties of the five
2
6
309
dx.doi.org/10.1021/ja401590c | J. Am. Chem. Soc. 2013, 135, 6307−6316

Journal of the American Chemical Society
Article
types of colloidal alumina particles investigated in this study.
The surface densities of the functional groups after surface
functionalization were found between 2.3 and 4.6 ± 0.2
2
groups/nm as determined by elemental analysis. The hydroxyl
group surface density on native alumina (Al−OH) yields 2.0 ±
2
0
.2 Al−OH/nm determined by potentiometric titration (data
not shown). Comparing the surface densities suggests that on
average one precursor molecule associated with one Al−OH
group in case of the −COOH and −PO H functionalization.
3
2
During −SO H and −NH surface functionalization, a second
3
2
precursor molecule may additionally condense or cross-link
with an already deposited precursor molecule (e.g., via a free
silanol group) leading to a functional group density that is
21
about twice than expected from the Al−OH surface density.
The median particle size (d ) was maintained at about 180 nm
5
0
after surface functionalization (particle size distribution is given
in the Supporting Information). The specific surface area
determined by nitrogen adsorption slightly increased depend-
ing on type and surface density of functional groups attached.
This indicates that the small nitrogen molecules can adsorb on
the molecules attached to the particle surface increasing the
Figure 1. Partial charge distribution of GSSG as function of the pH.
Surface potential distribution is represented by isosurfaces of negative
(red) and positive (blue) potentials calculated for equipotentials of
±30 mV at neutral pH and ionic strength of 1 mM at 25 °C.
21
measured particle surface area. However, these changes of the
surface area are expected to be too small to influence the
adsorption of a biomolecule like GSSG that is about 10 times
larger than a nitrogen molecule. To normalize the GSSG
adsorption on the particle surface area (see section 3.4.), thus
for all particles, the specific surface area of native Al O particles
2
3
was used. The IEP of native alumina particles was found as
49
expected at pH 9.3 and was shifted toward pH 11 after
introduction of the basic −NH groups. In contrast, attachment
2
of acidic −COOH, −PO H , and −SO H groups shifted the
3
2
3
IEP to 3.2, 2.4, and 1.5, respectively. A further increase in
precursor concentration for functionalization did neither
increase the shift of the IEP nor the functional group density
(
data not shown), suggesting that the conditions used caused a
saturation of the particle surface with functional groups. Similar
findings are reported in our previous work which shows by
transmission electron microscopy that the particle morphology
and size are unchanged after surface functionalization and by
water vapor and n-heptane adsorption studies that the particles
have slight differences in their hydrophilic/hydrophobic
Figure 2. (A) Adsorption of GSSG as function of the surface
functionalization of Al O particles: 50 nmol GSSG per sample
2 3
22
properties depending on their surface functionalization.
(corresponding to a concentration of 50 μM) was incubated for 60
min with the particles (0.1 vol %) in ddH O, and the remaining GSSG
3.2. GSSG Structure and pH-Dependent Charge. The
2
content in the supernatant was determined (black bars, left).
GSSG structure is depicted in Figure 1. The charged status of
GSSG is strongly affected by the pH which determines the
extent of deprotonation and protonation of the carboxyl and
amino residues of GSSG (Figure 1). In the pH range of 4−8,
which is the pH of the GSSG adsorption experiments, GSSG
has a constant total charge of −2 due to four deprotonated
Subsequently the particles were washed in ddH O and resuspended
2
in 10 mM KOH (pH 12.3) to release the adsorbed GSSG (white bars,
right). The data represent means ± standard deviation of three
independent experiments. (B) The pH of the as-prepared suspensions
during GSSG incubation and the particles’ ζ potential at this pH (ionic
−4
strength is ≤10 M).
−
carboxylate groups (−COO ) and two protonated amino
+
3
groups (−NH ). Although GSSG is overall negatively charged,
positively charged sites exist, which are illustrated by the surface
potential distribution of the GSSG molecule (Figure 1).
indicating GSSG adsorption on these particles. In contrast,
none or only marginal GSSG adsorption was found after
incubation with Al O −COOH (46.4 ± 1.3 nmol) or Al O −
3.3. Effect of Surface Functionalization on GSSG
2
3
2
3
Adsorption. Figure 2A depicts the influence of the particle
surface functionalization on the amount of GSSG in the
supernatant (Figure 2A, left panel) after the particles were
incubated with 50 μM GSSG (50 nmol per sample). Figure 2B
shows the pH of the as-prepared suspension during GSSG
incubation and the ζ potential of the particles at this pH.
Incubation with Al O , Al O −NH , and Al O −SO H
PO H (46.3 ± 1.2 nmol).
3 2
Treatment of the particles after GSSG adsorption with 10
mM KOH enabled to release and quantify the particle bound
GSSG (Figure 2A, right panel). For Al O −COOH and
2
3
Al O −PO H , no GSSG was released by KOH as expected
2
3
3
2
from the, at best, marginal GSSG loss found for the supernatant
of these particles. For Al O , Al O −NH , and Al O −SO H,
2
3
2
3
2
2
3
3
2
3
2
3
2
2
3
3
significantly reduces the amount of GSSG in the supernatant
to 14.2 ± 1.2, 17.7 ± 0.7, and 32.6 ± 1.3 nmol, respectively,
29.1 ± 0.6, 27.4 ± 0.5, and 10.3 ± 1.0 nmol GSSG were found
adsorbed to the particles, respectively, which corresponded well
6
310
dx.doi.org/10.1021/ja401590c | J. Am. Chem. Soc. 2013, 135, 6307−6316

Journal of the American Chemical Society
Article
Figure 3. (A) Time and concentration dependency of GSSG adsorption to Al O , Al O −NH -, or Al O −SO H-functionalized particles. The pH
2
3
2
3
2
2
3
3
for the adsorption phase was 4.8 for Al O , 6.4 for Al O −NH , and 4.2 for Al O −SO H (at 50 μM GSSG). (B) GSSG adsorption isotherms and
2
3
2
3
2
2
3
3
Langmuir regressions of the concentration-dependent adsorption of GSSG to Al O , Al O −NH , or Al O −SO H particles. (C) Langmuir
2
3
2
3
2
2
3
3
constants and Gibbs energies for the adsorption of GSSG calculated from the values shown in (B). The data shown are the amounts of GSSG that
had been released from the particle surface by KOH. The data represent means ± standard deviation of three independent experiments.
2
with the GSSG loss determined for the supernatant. The KOH
treatment partially lowers the amount of particle surface
functional groups indicated by IEP measurements (Table S2),
and the GSSG release by KOH may be a consequence of
alumina particle surface dissolution, dissolution of the func-
tional groups from the particles, and GSSG desorption
processes. In fact, the GSSG amounts determined in the
KOH supernatants matched the amounts of GSSG that had
at about 39 μM, which corresponds to 852 nmol GSSG/m or
2
1.95 nm /GSSG molecule when the amount of adsorbed GSSG
was normalized to the surface area of the particles. These values
correspond quite well to the concentration of a theoretical
regular monolayer of GSSG which contains 844 nmol/m
assuming that GSSG occupies a circular surface area of about
2
2
50
1.55 nm (maximal GSSG projection area, Figure S3) and a
surface packing density of 78.5% (quadratic packing). In case of
Al O −NH , GSSG adsorption exceeded a theoretical GSSG
disappeared from the ddH O supernatants, confirming that the
2
2
3
2
KOH treatment of the particles is suitable to quantitatively
determine the adsorbed GSSG.
monolayer and saturated at about 76 μM (1660 nmol GSSG/
2
m ). The area that one GSSG molecule occupied on Al O −
2
3
2
The particle ζ potentials determined for the GSSG
adsorption conditions (Figure 2B) indicate that Al O and
NH was about 1.0 nm , which corresponds to the theoretical
2
50
2
minimal projection area of GSSG of 1.0 nm (Figure S3). For
2
3
Al O −NH are, as expected from their IEP (Table 1), net
Al O −SO H, GSSG adsorption saturated already at about 19
2
3
2
2
3
3
2
positively charged under the conditions used for GSSG
adsorption. Interestingly, among the net negatively charged
particles, i.e., Al O −COOH, Al O −SO H, and Al O −
μM (415 nmol GSSG/m corresponding to an adsorption area
2
of 4.0 nm per GSSG molecule), which is below the calculated
theoretical monolayer.
2
3
2
3
3
2
3
PO H , GSSG adsorbed exclusively onto Al O −SO H.
The Langmuir regression provided a good fit for the
adsorption data (Figure 3B). The correlation coefficient and
the characteristic Langmuir constants calculated for the GSSG
adsorption onto Al O , Al O −NH or Al O −SO H particles
3
2
2
3
3
As significant GSSG adsorption was only found on Al O ,
2
3
Al O −NH , and Al O −SO H, we focused for further analysis
2
3
2
2
3
3
on the interaction of GSSG with alumina only on these particle
types.
2
3
2
3
2
2
3
3
are given in Figure 3C. The Gibbs energies of GSSG adsorption
0
ads
3.4. Time and Concentration Dependencies of GSSG
ΔG were calculated from the Langmuir fits of each particle
Adsorption. The adsorption of GSSG was found to be rapid
and after already <5 min incubation, maximal amounts of
GSSG were found adsorbed onto Al O , Al O −NH , and
type and were found similar for Al O , Al O −NH , and
2
3
2
3
2
Al O −SO H.
2
3
3
3.5. Desorption of GSSG by Salt Addition. To estimate
the contribution of electrostatic interactions in the binding of
GSSG to alumina particles, the effect of ions of varying type
and valences on the amount of GSSG bound to Al O , Al O −
2
3
2
3
2
Al O −SO H that had been incubated with 5, 50, or 500 nmol
2
3
3
GSSG (Figure 3A). Longer incubation times of up to 60 min
did not result in higher GSSG adsorption, and 30 min
adsorption time was used to investigate the concentration
dependence of the GSSG binding to Al O and Al O −NH -
2
3
2
3
NH , or Al O −SO H particles was investigated. Particles were
2
2
3
3
preincubated with 1000 nmol GSSG which resulted in maximal
binding of GSSG to the particles (Figure 3). Subsequently, the
GSSG-loaded particles were incubated with salt solutions of pH
2
3
2
3
2
and Al O −SO H-functionalized particles in more detail
2
3
3
(
Figure 3 B). GSSG adsorption saturated on Al O particles
2 3
6
311
dx.doi.org/10.1021/ja401590c | J. Am. Chem. Soc. 2013, 135, 6307−6316

Journal of the American Chemical Society
.2. The relative amount of GSSG desorbed from the particles
Article
6
after incubation at nonreducing or reducing conditions
(absence and presence of DTT, respectively) is shown in
Figure 5A. For Al O and Al O −NH particles, about 75% of
by different ions is depicted in Figure 4. The presence of NaCl
2
3
2
3
2
the initially bound GS-moieties was released by the DTT
treatment compared to the nonreducing conditions. In contrast,
DTT treatment did not liberate any GS moieties from GSSG
that had been adsorbed to Al O −SO H particles, suggesting
2
3
3
that both GS moieties of GSSG are involved in interactions
with the surface of Al O −SO H-functionalized particles.
2
3
3
The ζ potential and IEP measurements were used to confirm
the results of the DTT treatment and to provide information
on the functional groups of adsorbed GSSG or of the particle
surface that contributes to the electrokinetic particle properties.
We therefore compared ζ potential and IEP of the particles
without GSSG, after GSSG adsorption, and after GSSG
adsorption and DTT treatment (Figure 5 B). For native
Al O particles, the GSSG adsorption shifted the IEP from 9.3
2
3
±
0.1 to the more acidic 7.2 ± 0.1, demonstrating the
contribution of adsorbed GSSG to the electrokinetic particle
properties. After DTT treatment, the IEP was shifted back to
more basic 8.8 ± 0.1, which correlated with the loss of GS
moieties. However, the initial IEP of Al O was not recovered,
2
3
indicating that GS moieties are still attached to the particle
surface. For Al O −NH -functionalized particles, GSSG
2
3
2
adsorption shifted the IEP similarly from 10.9 ± 0.3 to the
slightly more acidic 9.5 ± 0.2. After DTT treatment, the IEP of
Al O −NH -functionalized particles was found at 11.0 ± 0.1,
Figure 4. Desorption of GSSG from Al O , Al O −NH -, or Al O −
2
3
2
3
2
2
3
2
3
2
SO H-functionalized particles resuspended in different salt solutions.
3
representing almost the IEP of untreated Al O −NH particles.
2
3
2
The 1000 nmol GSSG was preincubated with 0.1 vol % of the particles
for 30 min. Subsequently, particles were centrifuged, washed with 1
mL ddH O, resuspended in ddH O at pH 6.2 containing NaCl,
However, the ζ potential curve of the treated particles was
between that of bare and GSSG-covered Al O −NH -particles,
2
3
2
2
2
which suggests presence of remaining GS moieties on the
Na SO , or (NH ) SO in the indicated concentrations, and incubated
2
4
4
2
4
for 60 min. The data represent the relative desorbed GSSG amounts.
Means ± standard deviation are of three independent experiments.
particle surface. For Al O −SO H-functionalized particles, the
2
3
3
IEP varied only slightly after GSSG adsorption and DTT
treatment. In addition, no significant variation of the ζ potential
curve was observed, suggesting that the −SO H groups still
leads to a concentration-dependent increase in GSSG
3
dominate the electrokinetic properties of these particles, in
correlation with the below-monolayer amounts of adsorbed
desorption from Al O and Al O −NH particles. After
2
3
2
3
2
incubation with 100 mM NaCl, about 55% (Al O ) and 63%
2
3
GSSG on Al O −SO H-functionalized particles.
(
Al O −NH ) of the bound GSSG had been desorbed. The 1:2
2
3
3
2
3
2
electrolyte Na SO resulted at a concentration of 100 mM in a
2
4
4
. DISCUSSION
stronger desorption reaching 89% GSSG desorption for Al O3
2
and 82% for Al O −NH . In (NH ) SO solution (also 1:2
GSSG adsorption is strongly affected by the surface
functionalization of alumina particles as demonstrated by our
data for particles with −NH , −COOH, −SO H, and −PO H
functional groups, and the sensitive enzymatic cycling assay of
GSSG allowed a precise analysis of the amounts of GSSG
adsorbed and desorbed from the particles under various
experimental conditions.
2
3
2
4 2
4
electrolyte), the GSSG desorption behavior was similar to that
+
+
4
in Na SO solution, thus exchanging Na with the larger NH
2
4
2
3
3
2
ion did not affect the GSSG desorption from Al O and Al O −
2
3
2
3
NH . In contrast, for Al O −SO H, even a 100 mM NaCl
2
2
3
3
concentration did not support GSSG desorption, indicating a
stronger interaction of GSSG with Al O −SO H than with
2
3
3
Al O and Al O −NH . In contrast, presence of either Na SO
We observed high GSSG adsorption onto native alumina
particles. As previously reported, the net negatively charged
GSSG adsorbs effectively onto the positively charged alumina
particles driven by electrostatic interactions between GSSG and
2
3
2
3
2
2
4
or (NH ) SO enabled a concentration-dependent GSSG
4
2
4
desorption from Al O −SO H and 100 mM of these salts
2
3
3
desorbed 68% and 63%, respectively, of the bound GSSG.
.6. Reduction of the Disulfide Bond of Adsorbed
11
3
alumina. Functionalization of the alumina particles with
GSSG. One GSSG molecule contains two glutathionyl (GS)
moieties, which are connected by a disulfide bond (Figure 1).
Reduction of GSSG with DTT will therefore generate two
GSH molecules. This reduction allows to estimate whether one
or both of the two GS moieties of one adsorbed GSSG
molecule contributes to the binding to the particles, following
the approach of Dringen et al. Therefore, 0.1 vol % of the
particles was preincubated for 30 min with 1 mM GSSG to bind
GSSG, unbound GSSG was removed, and the particles were
−NH groups introduced positively charged groups onto the
2
surface, increased the particles IEP, and resulted in enhanced
GSSG adsorption. This was observed although the net ζ
potential was slightly lower for the −NH -modified alumina
2
particles when compared to native alumina particles, which may
be due to the different pH values of the as-prepared and
unbuffered solutions used for the measurements (pH 6.4 for
11
−NH -functionalized alumina and pH 4.8 for native alumina
2
particles). The acidic functionalizations (−COOH, −PO H ,
3
2
subsequently incubated for 60 min with ddH O or with 20 mM
and −SO H functional groups) introduced a negative particle ζ
2
3
DTT in ddH O. The amount of particle adsorbed GS moieties
potential, and according to an electrostatic interaction
2
6
312
dx.doi.org/10.1021/ja401590c | J. Am. Chem. Soc. 2013, 135, 6307−6316

Journal of the American Chemical Society
Article
Figure 5. (A) Effects of dithiothreitol (DTT) on the GSSG adsorbed to Al O , Al O −NH - or Al O −SO H-functionalized particles: 0.1 vol % of
2
3
2
3
2
2
3
3
the particles were preincubated for 30 min with 1 mM GSSG, the unbound GSSG was removed, and the particles were subsequently incubated for
0 min with ddH O or 20 mM DTT in ddH O. For the nonreducing conditions, the amount of adsorbed GSSG is considered to be not affected,
6
2
2
while for the reducing conditions, it is assumed that exposure to DTT will have reduced all disulfide bonds in the adsorbed GSSG. The data
represent the amount of GS-moieties in either GSSG or GSH that remained adsorbed on the particle surface during the main incubation and were
finally released by 10 mM KOH. (B) Influence of the GSSG adsorption and DTT treatment on the ζ potential as function of the pH and the
particles’ IEP. The data represent means ± standard deviation of three independent experiments.
approach, they lowered (Al O −SO H) or even prevented
Native alumina, −NH -, and −SO H-functionalized alumina
2
3
3
2
3
(
Al O −COOH, −PO H ) adsorption of net negatively
particles were investigated in detail to study the molecular
mechanisms involved in their GSSG adsorption behavior.
GSSG adsorption was rapid as demonstrated by maximal
bindings that were achieved within 5 min, and the adsorption
2
3
3
2
charged GSSG to the particles (Figures 1 and 2). Unexpectedly,
some GSSG adsorption was exclusively found on −SO H-
3
functionalized alumina, although −COOH and −PO H -
3
2
functionalized particles have almost the identical ζ potential.
isotherms demonstrated that the GSSG adsorption was
0
ads
It was previously shown that protein adsorption on −SO H-
saturable. The Gibbs energy of adsorption ΔG estimated
3
functionalized alumina also varies from that of −COOH and
from the GSSG adsorption isotherms were in the range of
typical values for electrostatic adsorption processes (<50 kJ/
−
PO H -functionalized alumina particles, and an influence of
3 2
52
the different acidity of the functional groups was suggested to
mol) and similar to those previously found for GSSG and
22
11,13
be responsible for this effect. Indeed, −SO H groups have the
lowest pK (pK = 1.53) value, and the deprotonated state is
GSH adsorption onto alumina.
The negative values of
3
0
ΔG , which were almost identical for Al O , Al O −NH , or
a
a
ads
2
3
2
3
2
most stable compared to the −COOH (pK = 4.13) and
Al O −SO H particles, indicate an exergonic adsorption
a
2
3
3
51
−
PO H (pK = 2.16) groups at the pH of about 4.2 ± 0.1,
behavior of GSSG to all three types of particles, which thus
appears in this regard to be rather independent of the surface
functionalization.
3
2
a
beneficial and necessary for interactions with the positively
charged sites on GSSG. Milani et al. also reported different
protein adsorption on −COOH and −SO H modified
That mainly electrostatic interactions are responsible for the
adsorption of GSSG to the particle surface is further supported
by the ability of salts to desorb GSSG from the three types of
particles investigated. Nonetheless, independent of the particle
surface chemistry, some GSSG remained adsorbed on the
particles after salt application, which was not unexpected, as
electrostatic interactions still occur at the highest ionic
strengths used (Debye lengths are ≈1 nm at 100 mM). In
addition, a contribution of nonelectrostatic interaction forces
(e.g., hydrogen bonding) in GSSG−particle interaction cannot
be totally excluded but is likely to play only a secondary role in
the interaction between GSSG and the alumina particles
investigated. Interestingly, the interaction between GSSG and
3
polystyrene particles and found that the adsorption of
transferrin is stronger on −SO H modified particles than on
3
14
−
COOH modified particles. A further effect may derive from
the close proximity (geminal configuration) of the positively
charged amino and negatively charged carboxylate group in the
GSSG molecule. The binding of GSSG via its positively charged
amino group with a negatively charged particle surface group
might be hindered if a second negatively charge particle surface
group is nearby and repulses the carboxylate group of GSSG.
Such an effect may particularly happen in the case of Al O −
53
2
3
COOH, where the precursor molecule carries two vicinal
carboxylate groups.
6
313
dx.doi.org/10.1021/ja401590c | J. Am. Chem. Soc. 2013, 135, 6307−6316

Journal of the American Chemical Society
SO H-functionalized alumina seems to be more stable against
Article
−
(Figure 5). The loss of GS moieties even exceeds 50%
indicating that some of the GSH moieties generated on the
particles during the reduction of adsorbed GSSG will desorb
from the particles consistent with the previous findings that
binding of GSH is weaker on alumina surfaces than that of
GSSG and bound GSH quickly desorbs from alumina to
3
exposure to different salt solutions than that of GSSG with
native or −NH -functionalized alumina, as the presence of
2
NaCl, even in a concentration of 100 mM, was unable to
desorb additional GSSG from the −SO H-functionalized
3
alumina compared to the respective salt-free control. This
suggests a different binding mechanism and a potentially
different orientation of GSSG on positively and negatively
charged surfaces as discussed in the following.
1
1,13
establish equilibrium between adsorbed and free GSH.
On −SO H-functionalized alumina particles, the experimen-
tally determined GSSG projection area (4.0 nm /GSSG
3
2
The saturation of GSSG adsorption suggests that a limited
number of adsorption sites for GSSG are available on the
particle surface and excludes substantial multilayer formation by
peptide−peptide interactions between GSSG and particle-
preadsorbed GSSG. The GSSG concentrations that enable
maximal GSSG binding were transferred into the surface area
molecule) is higher than the maximal theoretical projection
area of GSSG. This suggests that GSSG is not densely packed
on the particle surface and some particle surface may remain
−
uncovered. The repulsion between negatively charged −SO
3
groups of the particles and carboxylate groups of GSSG could
be a driving force preventing a better coverage of the surface.
2
occupied by one GSSG molecule (1.95 nm per GSSG
Assuming that GSSG covers its maximal projection area (side-
2
2
molecule on native alumina, 1.0 nm on −NH -functionalized
2
on, 1.55 nm /GSSG molecule) on −SO H-modified alumina
3
2
2
alumina, and 4.0 nm on −SO H-functionalized alumina). The
3
(4.6 ± 0.2 SO H/nm ), a GSSG molecule would occupy about
3
results for native alumina and −NH -functionalized alumina are
2
seven −SO H groups. Due to such a high availability of −SO H
3
3
quite alike the projection areas calculated from a rigid structure
groups per GSSG molecule, it is likely that both of the two
2
+
of the GSSG molecule, maximal 1.55 nm (side-on area) and
minimal 1.0 nm (end-on area), respectively (Figure S3).
protonated −NH of the GSSG interact with deprotonated
3
2
50
−
−
SO groups on the particle surface. This hypothesis is
3
Although GSSG is a flexible molecule and conformational
changes are likely to influence the molecule size during
strongly supported by the results of the reduction of the
disulfide bridges of bound GSSG on the particle surface by
DTT, which does not lower the amount of adsorbed GS
moieties on Al O −SO H. As each of the GS moiety of GSSG
9
,11
adsorption,
the strong correlation of the theoretical and
experimentally determined GSSG adsorption area suggests that
GSSG adsorbs in a side-on adsorption mode on native alumina
2
3
3
+
−
contains one −NH , that is likely to interact with the −SO
groups of the particle surface, the reduction of one molecule of
adsorbed GSSG into two molecules of bound GSH will not
3
3
particles and in an end-on adsorption mode on −NH -
2
functionalized alumina particles.
The approach estimates also which functional groups of the
particles and the GSSG molecule can theoretically interact with
lower the amount of adsorbed GS moieties from −SO H-
3
functionalized alumina. Illustration of hypothetical binding
2
each other. A surface area of 1.55 nm on native alumina
modes of GSSG on Al O , Al O −NH -, or Al O −SO H-
2
3
2
3
2
2
3
3
2
particles (2.0 ± 0.2 Al−OH/nm ) contains about 3 Al−OH
functionalized particles is given in Figure 6.
groups which could potentially interact with 1−3 of the 4
carboxylate groups of GSSG in side-on orientation. This is only
possible when all Al−OH groups determined by potentiometric
are sterically accessible and equivalent binding sites for the
GSSG carboxylate groups. That may not be always the case,
and GSSG is indeed suggested to bind, despite side-on
orientation, with only 1−2 carboxylates of one GS moiety as
The electrokinetic analysis of the particles after GSSG
adsorption and disulfide bridge reduction (Figure 5) supports
the proposed adsorption behavior. Assuming that the surface is
completely covered with GSSG in case of native and −NH2-
discussed later. GSSG has a higher affinity for −NH -
2
functionalized alumina particles and is proposed to adsorb on
these particles with its minimal projection area reaching higher
surface packing than on native alumina particles. The
concentration of −NH groups per surface area is <3.4 NH /
2
2
2
nm (although the IEP of these particles clearly demonstrates
the presence of −NH an exact quantification via elemental
2
analysis was not achieved). A GSSG molecule that covers 1.0
2
nm can contact thus maximally three −NH groups. The
2
suggested end-on adsorption mode of GSSG might promote
that only 1−2 of the −COO groups of GSSG interact with the
particle surface. This assumption is further supported by the
−
Figure 6. Illustration of hypothetical binding modes of GSSG on
Al O , Al O −NH -, or Al O −SO H-functionalized particles. GSSG
slightly lower adsorption enthalpy for −NH -functionalized
2
2
3
2
3
2
2
3
3
alumina compared to native or −SO H functionalized particles,
contains two GS moieties that are connected by a disulfide bridge.
3
Reduction of GSSG by DTT will generate two molecules of GSH. In
which is influenced by the number of functional groups/charge
54
case of Al O and Al O −NH about 75% GSH molecules are released
carriers interacting during adsorption. Moreover, the results
of the reduction of the disulfide bridges of bound GSSG on the
particle surface by DTT suggest that GSSG interacts indeed
2
3
2
3
2
by the DTT treatment and around 25% remain bound to the particle
by electrostatic interactions between positively charged groups on the
particle surface and negatively charged carboxylate groups of GSH. In
case of Al O −SO H particles, the DTT treatment does not liberate
with only one of its two GS moieties with native and −NH -
2
2
3
3
functionalized alumina particles (also previously suggested for
any GSH, and it is expected that both GS moieties are bound to the
1
1
native alumina). On both, native and −NH -functionalized
2
particle surface by interactions between negatively charged −SO H
3
particles, the reduction of the disulfide bridge removed >50% of
groups on the particle surface and positively charged protonated
the GS moieties which are not bound to the particle surface
−NH groups of the GSSG.
2
6
314
dx.doi.org/10.1021/ja401590c | J. Am. Chem. Soc. 2013, 135, 6307−6316

Journal of the American Chemical Society
Article
functionalized alumina, only the functional groups of the
adsorbed GSSG will determine the ζ potential and IEP, and the
initial particle surface will lose its influence. For native and
ASSOCIATED CONTENT
Supporting Information
■
*
S
Isoelectric point of Al O , Al O −NH , Al O −COOH,
2
3
2
3
2
2
3
−
NH -functionalized alumina, GSSG adsorption shifted the
2
Al O −SO H, and Al O −PO H after a 10 mM KOH
2
3
3
2
3
3
2
particles IEP to more acidic values, which demonstrates the
contribution of free carboxylate groups of GSSG to the
establishment of the new IEP. Those carboxylate groups that
contribute to the particle IEP are unlikely to be involved in
binding between GSSG and the particle surface.
treatment and GSSG projection areas. This material is available
free of charge via the Internet at http://pubs.acs.org
AUTHOR INFORMATION
Corresponding Author
krezwan@uni-bremen.de
Author Contributions
These authors contributed equally.
Notes
The authors declare no competing financial interest.
■
In addition, the ζ potential and IEP of native and -NH2-
functionalized alumina after GSSG adsorption were not
completely governed by negative carboxylate groups which
would have led to a negatively charged and acidic IEP. This
∥
observation suggests a contribution of unbound protonated
+
−
NH3 groups of GSSG to the electrokinetic properties. The
increase of the IEP after DTT treatment confirms the release of
ACKNOWLEDGMENTS
We thank the EuropeanResearch Council for financial support
within the BiocerEng, project no. 205509.
■
GS moieties from native and −NH -functionalized alumina. In
2
contrast, in case of −SO H-functionalized particles, the GSSG
3
adsorption did not alter the IEP and pH-dependent ζ potential
as expected by a substantial influence of the initial particle
REFERENCES
■
surface of −SO H-functionalized alumina on the IEP and ζ
3
(
1) Monopoli, M. P.; Aberg, C.; Salvati, A.; Dawson, K. A. Nat.
Nanotechnol. 2012, 7, 779.
potential due to only partial GSSG coverage. Due to their likely
involvement in binding of GSSG to the −SO H-functionalized
3
(2) Vallee, A.; Humblot, V.; Pradier, C.-M. Acc. Chem. Res. 2010, 43,
1297.
(3) Nel, A. E.; Maedler, L.; Velegol, D.; Xia, T.; Hoek, E. M. V.;
Somasundaran, P.; Klaessig, F.; Castranova, V.; Thompson, M. Nat.
Mater. 2009, 8, 543.
4) Helms, B.; van Baal, I.; Merkx, M.; Meijer, E. W. ChemBioChem
007, 8, 1790.
5) Banerjee, I. A.; Yu, L.; Matsui, H. Proc. Natl. Acad. Sci. U.S.A.
2003, 100, 14678.
6) Costa, F.; Carvalho, I. F.; Montelaro, R. C.; Gomes, P.; Martins,
M. C. L. Acta Biomater. 2011, 7, 1431.
7) Krauland, E. M.; Peelle, B. R.; Wittrup, K. D.; Belcher, A. M.
Biotechnol. Bioeng. 2007, 97, 1009.
(8) Cheng, R.; Feng, F.; Meng, F.; Deng, C.; Feijen, J.; Zhong, Z. J.
Controlled Release 2011, 152, 2.
9) Soliman, W.; Bhattacharjee, S.; Kaur, K. J. Phys. Chem. B 2010,
14, 11292.
10) Thyparambil, A. A.; Wei, Y.; Latour, R. A. Langmuir 2012, 28,
687.
11) Dringen, R.; Koehler, Y.; Derr, L.; Tomba, G.; Schmidt, M. M.;
+
3
alumina particles, the positively charged −NH groups of
GSSG do not significantly impact the electrokinetic particle
properties, which should also not be affected by reduction of
bound GSSG to GSH by DTT. Indeed, DTT treatment did not
(
2
(
affect the ζ potential and only marginally the IEP of −SO H-
3
functionalized alumina particles that had been preincubated
with GSSG.
(
CONCLUSIONS
■
(
The surface chemistry of colloidal alumina particles tailored
with −NH , −COOH, −PO H , and −SO H functional groups
2
3
2
3
strongly affects the adsorption and the orientation of GSSG on
the particle surface. Although electrostatic interactions are
suggested as a main driving force for the GSSG adsorption, the
type, concentration, and acidity/basicity of the functional
surface groups, rather than just the ζ potential, have to be
considered to predict and control GSSG adsorption. Our data
suggest the following adsorption behavior of GSSG as function
of the particle surface chemistry: GSSG adsorbs on positively
charged native alumina in monolayer concentrations via the
carboxylate groups of one of its GS moieties in a side-on
orientation. GSSG has a higher affinity and denser surface
(
1
(
5
(
Treccani, L.; Colombi Ciacchi, L.; Rezwan, K. Langmuir 2011, 27,
9449.
(
12) Lopes, I. n.; Piao, L.; Stievano, L.; Lambert, J.-F. o. J. Phys. Chem.
C 2009, 113, 18163.
13) Schmidt, M. M.; Koehler, Y.; Derr, L.; Treccani, L.; Rezwan, K.;
Dringen, R. J. Phys. Chem. C 2012, 116, 23136.
14) Milani, S.; Baldelli Bombelli, F.; Pitek, A. S.; Dawson, K. A.;
Radler, J. ACS Nano 2012, 6, 2532.
(
(
packing on positively charged −NH -functionalized alumina,
̈
2
binding in end-on orientation also via the carboxylate groups of
one GS moiety. Among the acidic and negatively charged
particles bearing −PO H , −COOH, and −SO H groups,
(15) Rothenstein, D.; Claasen, B.; Omiecienski, B.; Lammel, P.; Bill,
J. J. Am. Chem. Soc. 2012, 134, 12547.
(
16) Joshi, S.; Ghosh, I.; Pokhrel, S.; Mad
Nano 2012, 6, 5668.
17) Mirau, P. A.; Naik, R. R.; Gehring, P. J. Am. Chem. Soc. 2011,
33, 18243.
18) Seker, U. O. S.; Demir, H. V. Molecules 2011, 16, 1426.
(19) Hoshino, Y.; Koide, H.; Urakami, T.; Kanazawa, H.; Kodama,
T.; Oku, N.; Shea, K. J. J. Am. Chem. Soc. 2010, 132, 6644.
20) Basalyga, D. M.; Latour, R. A. J. Biomed. Mater. Res., Part A
003, 64A, 120.
21) Meder, F.; Brandes, C.; Treccani, L.; Rezwan, K. Acta Biomater.
013, 9, 5780.
22) Meder, F.; Daberkow, T.; Treccani, L.; Wilhelm, M.;
̈
ler, L.; Nau, W. M. ACS
3
2
3
GSSG adsorbs exclusively onto −SO H modified alumina.
3
(
1
(
Here, GSSG is suggested to bind with both GS moieties via its
two protonated amino groups and only partially covers the
particle surface in a side-on orientation. These results represent
a firm base for further studies on GSSG and peptide adsorption
phenomena at physiological conditions or in biological fluids
under competition with other peptides and proteins. Tailored
particle surface functionalization might then be employed to
control peptide and protein adsorption amount and orientation
by a specific design of the materials surface chemistry and its
functional groups.
(
2
(
2
(
Schowalter, M.; Rosenauer, A.; Maedler, L.; Rezwan, K. Acta Biomater.
2012, 8, 1221.
6
315
dx.doi.org/10.1021/ja401590c | J. Am. Chem. Soc. 2013, 135, 6307−6316

Journal of the American Chemical Society
Article
(
(
(
23) Wei, Y.; Latour, R. A. Langmuir 2009, 25, 5637.
24) Neouze, M.-A.; Schubert, U. Monatsh. Chem. 2008, 139, 183.
25) Largueze, J.-B.; Kirat, K. E.; Morandat, S. Colloids Surf., B 2010,
7
(
9, 33.
26) Lazzara, T. D.; Behn, D.; Kliesch, T.-T.; Janshoff, A.; Steinem,
C. J. Colloid Interface Sci. 2012, 366, 57.
27) Li, J.; Wang, J.; Gavalas, V. G.; Atwood, D. A.; Bachas, L. G.
Nano Lett. 2002, 3, 55.
(
(
(
28) Li, H.; Li, Y.; Jiao, J.; Hu, H.-M. Nat. Nanotechnol. 2011, 6, 645.
29) Kossovsky, N.; Gelman, A.; Sponsler, E. E.; Hnatyszyn, H. J.;
Rajguru, S.; Torres, M.; Pham, M.; Crowder, J.; Zemanovich, J.;
Chung, A.; Shah, R. Biomaterials 1994, 15, 1201.
(
(
30) Hirrlinger, J.; Dringen, R. Brain Research Reviews 2010, 63, 177.
31) Schmidt, M.; Dringen, R. In Advances in Neurobiology, Vol. 4:
Neural Metabolism In Vivo; Choi, I.-Y., Gruetter, R., Eds.; Springer:
New York, 2012; Vol. 4, p 1029.
(
32) Park, H. A.; Khanna, S.; Rink, C.; Gnyawali, S.; Roy, S.; Sen, C.
K. Cell Death Differ. 2009, 16, 1167.
(
33) Ellison, I.; Richie, J. P., Jr Biochem. Pharmacol. 2012, 83, 164.
(
34) Minich, T.; Riemer, J.; Schulz, J. B.; Wielinga, P.; Wijnholds, J.;
Dringen, R. J. Neurochem. 2006, 97, 373.
35) Angeli, V.; Chen, H.; Mester, Z.; Rao, Y.; D’Ulivo, A.; Bramanti,
E. Talanta 2010, 82, 815.
36) Monostori, P.; Wittmann, G.; Karg, E.; Tur
009, 877, 3331.
37) Noh, H.-B.; Chandra, P.; Moon, J. O.; Shim, Y.-B. Biomaterials
012, 33, 2600.
(
(
́
i, S. J. Chromatogr., B
2
(
2
(
38) Tietze, F. Anal. Biochem. 1969, 27, 502.
(
39) Saint-Denis, T.; Goupy, J. Anal. Chim. Acta 2004, 515, 191.
(
40) Dringen, R.; Kranich, O.; Hamprecht, B. Neurochem. Res. 1997,
22, 727.
(
(
41) Dringen, R.; Hamprecht, B. J. Neurochem. 1996, 67, 1375.
42) Gomma, G. K.; Wahdan, M. H. Mater. Chem. Phys. 1994, 39,
1
(
42.
43) Serro, A. P.; Degiampietro, K.; Colaco
Surf., B 2010, 78, 1.
44) ACD/Labs, version 5.0.0.184; Advanced Chemistry Develop-
ment, Inc.: Toronto, ON, Canada, 2012; www.acdlabs.com.
45) PubChem Substance Database; National Center for Biotechnol-
̧ , R.; Saramago, B. Colloids
(
(
ogy Information: Bethesda, MD; http://pubchem.ncbi.nlm.nih.gov/
summary/summary.cgi?cid=44630308 (accessed December, 2012).
(
46) Pedretti, A.; Villa, L.; Vistoli, G. J. Comput.-Aided Mol. Des. 2004,
8, 167.
47) Baker, N. A.; Sept, D.; Joseph, S.; Holst, M. J.; McCammon, J.
A. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 10037.
48) Humphrey, W.; Dalke, A.; Schulten, K. J. Mol. Graphics 1996,
4, 33.
49) Rezwan, K.; Meier, L. P.; Gauckler, L. J. Langmuir 2005, 21,
493.
50) Maximal and minimal GSSG projection area estimated by the
1
(
(
1
(
3
(
calculator plugin of Marvin version 5.4.0.0, ChemAxon (http://www.
chemaxon.com) and a rigid conformation of GSSG extracted from the
RCSB Protein Data Bank file (pdb ID: YCK1).
(
51) Lide, D. R. CRC Handbook of Chemistry and Physics, 75th ed.;
CRC Press: Boca Raton, FL, 1994.
52) Bathen, D.; Breitbach, M. Adsorption technology (German ed.);
Springer: Berlin, 2001.
53) van Oss, C. J. Interfacial Forces in Aqueous Media , 2nd ed.; CRC
press: Boca Raton, FL, 2006.
54) Baier, G.; Costa, C.; Zeller, A.; Baumann, D.; Sayer, C.; Araujo,
P. H. H.; Mailander, V.; Musyanovych, A.; Landfester, K. Macromol.
Biosci. 2011, 11, 628.
(
(
(
̈
6
316
dx.doi.org/10.1021/ja401590c | J. Am. Chem. Soc. 2013, 135, 6307−6316