ARTICLE IN PRESS

0142-9612/$ - se

doi:10.1016/j.bi

$A Contrib

Research Grou�Correspond

gineering, The P

Park, PA 16802

E-mail addr

Biomaterials 27 (2006) 3187–3194

www.elsevier.com/locate/biomaterials

Interfacial energetics of blood plasma and serum adsorption to ahydrophobic self-assembled monolayer surface$

Anandi Krishnana, Paul Chab, Yi-Hsiu Liub, David Allarab,c, Erwin A. Voglera,b,�

aDepartment of Bioengineering, The Pennsylvania State University, University Park, PA 16802, USAbDepartment of Materials Science and Engineering, The Pennsylvania State University, University Park, PA 16802, USA

cDepartment of Chemistry, The Pennsylvania State University, University Park, PA 16802, USA

Received 29 August 2005; accepted 7 December 2005

Available online 21 February 2006

Abstract

Interfacial energetics of blood plasma and serum adsorption to a hydrophobic, methyl-terminated self-assembled monolayer (SAM)

surface (solid–liquid SL interface) are shown to be essentially the same as to the buffer–air interface (liquid–vapor LV interface).

Specifically, spreading pressure (Pa) isotherms scaled on a w/v concentration basis constructed from advancing contact angles (ya) ofserially diluted plasma/serum derived from four different mammalian species (bovine, equine, human, and ovine) on the SAM surface are

not resolvable at the 99% confidence level and furthermore are found to be strikingly similar to isotherms of purified human-blood

proteins. Maximum advancing spreading pressures Pamax for protein mixtures fall within a relatively narrow 17oPa

maxo26mN/m band,

mirroring results obtained at the LV surface. These observations lead to the conclusion that neither depletion of coagulation proteins in

the conversion of plasma to serum nor variation in the plasma proteome among species has a substantial affect on adsorption energetics

to these test hydrophobic surfaces. Experimental results are rationalized on the basis that there is a generic mechanism controlling

adsorption of globular-blood proteins to test hydrophobic surfaces. We conclude that this generic mechanism is the hydrophobic effect

by which proteins are expelled from aqueous solution in order to increase hydrogen-bonding (self-association) among water molecules at

the expense of less favorable water–protein interactions. Expelled protein readily displaces water within the hydrophobic-surface region

and becomes adsorbed. The amount of water displaced per gram of adsorbed protein does not vary greatly among globular proteins

because the partial specific volume v0 of globular proteins is quite conserved (0.70pv0p0.75 cm3/g protein). Any single blood protein or

mixture of proteins consequently displaces nearly an equivalent amount of interfacial water and hence adsorption is observed to scale

similarly with solution concentration expressed in w=v units.

r 2006 Elsevier Ltd. All rights reserved.

Keywords: Plasma; Serum; Protein adsorption; Hydrophobic solid–liquid interface; SAM

1. Introduction

Perhaps it was Johlin who made the first clear experi-mental connection between protein adsorption and thebiological response to artificial materials by noting in his1929 classic Interfacial Adsorption as a Factor in the

e front matter r 2006 Elsevier Ltd. All rights reserved.

omaterials.2005.12.032

ution from the Hematology at Biomaterial Interfaces

p.

ing author. Department of Materials Science and En-

ennsylvania State University, 103 Steidle Bldg, University

-5005, USA. Tel.: +1 814 863 7403.

ess: eav3@psu.edu (E.A. Vogler).

Clotting of Blood Plasma [1] that ‘‘yclotting can beinduced by the contact of the plasma with substances thatproduce an adsorbing interfacey’’. If not Johlin, it wascertainly in the literature of this timeframe that we findorigins of a working hypothesis that has emerged over thedecades as a fundamental biomaterials-surface-sciencetenet stating that protein adsorption is the first step in the

biological response to materials. The focus on proteinadsorption has henceforth been unswerving, from thesystematic studies of the adsorption behavior of purifiedproteins by Lyman and Brash [2–6] carried out in the1965–1975 era, through to today’s emphasis on the use of

ARTICLE IN PRESSA. Krishnan et al. / Biomaterials 27 (2006) 3187–31943188

powerful analytical techniques such as ellipsometry [7],internal-reflection spectroscopy [8], and surface-plasmonresonance [9]; to name a few from many. And for allthis effort, the mechanism of protein adsorption tobiomaterial surfaces remains an important and contro-versial topic [10]. Among the important unsolved, out-standing issues are (i) how protein selectively collectsat biomaterial surfaces from multi-component proteinsolutions such as blood; (ii) quantitative structure-property relationships connecting surface chemistry/energy to the extent and specificity of protein adsorption;and (iii) detailed biochemical mechanisms by whichsurface-bound protein directs the biological response toartificial materials [11].

Our attention has been riveted by recent experimentalobservations that the adsorption energetics of bloodplasma and serum to the hydrophobic, aqueous-buffer/air(liquid–vapor, LV) surface is practically identical amongdiverse mammalian species (bovine, equine, human, andovine donors) [12], in spite of substantial differences inplasma proteomes. It has also been found that adsorptionof plasma/serum is very similar to that of purified-plasma-protein constituents [13–15]. These results are quiteunexpected on the widely held basis that differentproteins should adsorb differently to the same adsorbentsurface. Indeed, if different proteins adsorb to the samesurface in significantly differently ways (e.g. more-or-lesssurface concentration, more-or-less strongly, or more-or-less reversibly), then interfacial energetics of adsorptionshould reflect these differences because interfacial ener-getics are exquisitely sensitive to these aspects of adsorp-tion [15–20]. Instead, we find ‘‘generic’’ interfacialenergetics of protein adsorption to the LV surface andinterpret this evidence in terms of a heretofore unresolvedmechanistic commonality underlying protein adsorption[12]. Interesting as this hint of an underlying commonalitymight be, the LV surface is, after all, not entirely germaneto biomaterials and might be unique as an adsorbentsurface.

With all of the above in mind, we have extended ourprotein-adsorption studies to methyl-terminated, self-assembled monolayers (SAMs) supported on gold-coatedsemi-conductor wafers (solid–liquid, SL interface) as amodel hydrophobic solid surface. These studies show thatadsorption energetics of purified-plasma proteinsspanning 3 decades in molecular weight (MW) is verysimilar to that observed at the LV surface [11,16],suggesting that the mechanism of protein adsorption tothis hydrophobic SL surface is not fundamentally differentthan adsorption to the LV surface. Herein we report thatadsorption behavior of these purified proteins cannot beclearly distinguished from the adsorption of plasma andserum, again mirroring results obtained at the LV surfacereported above. Experimental observations are reconciledby asserting that the hydrophobic effect is the mechanisticcommonality underlying protein adsorption to hydropho-bic surfaces.

2. Methods and materials

2.1. Plasma/serum

Human platelet-poor plasma (citrated) was prepared from outdated

(within 2 days of expiration) lots obtained from the Hershey Medical

Center Blood Bank. Human serum was prepared from this plasma in

15mL batches by recalcification with 0.1M CaCl2 at 5:1 v=v plasma:cal-

cium ratio in clean glass scintillation vials for about 15min. Bovine, ovine,

and equine plasma and serum were used as received from Hemaresource

and Supply Inc. (Aurora, OR) and were not subjected to any additional

fractionation/purification steps. Ref. [13] discloses details of protein

solution preparation including serial dilutions that were performed in 96-

well microtiter plates by (typically) 50:50 dilution in phosphate buffered

saline solution (PBS) prepared from powder (Sigma Aldrich) in distilled-

deionized (18MO) water (interfacial tension of PBS and water was

checked periodically by Wilhelmy-balance tensiometry).

2.2. Surfaces

Methyl-terminated, thiol-based self-assembled monolayer surfaces

(SAMs) on gold-coated electronic-grade semiconductors were prepared

according to standard methods of surface engineering as reported

elsewhere [11,16,21–26]. Briefly, silicon wafers were pre-cleaned in hot

1:4 H2O2 (30%)/H2SO4 followed by rinsing with distilled-deionized H2O

and absolute ethanol. Gold-coated wafers were prepared by vapor

deposition of chromium and gold (99.99% purity) from resistively heated

tungsten boats onto clean 3-in. diameter silicon wafers at about

1� 10�8 Torr base pressure in a cryogenically pumped deposition

chamber. The sample was not allowed to rise above �40 1C during the

evaporation. Film thicknesses, monitored with a quartz crystal oscillator,

were typically 15 and 200nm for chromium and gold, respectively.

Chromium was deposited prior to gold to enhance adhesion to the

substrate. After deposition, the chamber was backfilled with research-

grade nitrogen. Gold-coated samples were removed and immersed in 1mM

solutions of 1-hexadecanethiol (CH3(CH2)15SH) in ethanol, contained in

glass jars at ambient temperature, for at least 3 days. The alkanethiol

(Aldrich Chemical Co., Milwaukee, WI) and ethanol (commercial reagent-

grade) were used as-received, without further purification. Samples were

stored in the thiol solution until use, rinsed with ethanol and air dried just

prior to an experiment.

2.3. Tensiometry and goniometry

Liquid–vapor interfacial tensions required by this work were measured

by pendant drop tensiometry (PDT) as described in Refs. [12–15]. Contact

angle and wettability methods applied in this work have been disclosed in

detail elsewhere, including verification that measured advancing angles

(ya) were in statistical agreement with that obtained by Wilhelmy balance

tensiometry [16,21]. Receding angles (yr) were shown to be not as reliable

as ya. Consequently, only ya was analyzed in this work. Briefly, for the

purposes of this paper, ya measurements were made using a commercial

automated tilting-plate goniometer (TPG, First Ten Angstroms Inc.,

Portsmouth VA). The goniometer employed a Tecan liquid-handling

robot to aspirate 12 mL of solutions contained in a 96-well microtiter plate

prepared by the serial-dilution protocol mentioned above. The robot was

used to reproducibly transfer the tip with fluid contents into a humidified

(99+% RH) analysis chamber and dispense10 mL drops of protein

solution onto the surface of test substrata held within the focal plane of a

magnifying camera. These and all other aspects of tilting-plate goniometry

were performed under computer control. Proprietary algorithms supplied

by the vendor were used to deduce contact angles from drop images

captured at a programmed rate by a frame grabber. Typically, 600 images

were captured at a rate of 1 image every 6 s following 20 s delay to permit

vibrations of the expelled drop to dampen. Precision of ya was about 0.51based on repeated measurement of the same drop. The analysis chamber

ARTICLE IN PRESSA. Krishnan et al. / Biomaterials 27 (2006) 3187–3194 3189

was thermostated to a lower-limit of 2571 1C by means of a computer-

controlled resistive heater. Thus, reported ya values were probably not

more accurate than about 11 on an inter-sample basis considering the

small, but measurable, variation of water interfacial tension with

temperature. This range of accuracy was deemed adequate to the

conclusions of this report which do not strongly depend on more accurate

ya that is difficult to achieve on a routine basis.

Test substrata were held on a rotating, tilting-plate platform driven

by stepper motors under computer control. Substrata were allowed to

come to equilibrium within the sample-chamber environment for no less

than 30min before contact angle measurements were initiated.

The platform was programmed to tilt at 11/s from horizontal to 251 after

θ θ a (

deg

rees

)

70

80

90

100

110

θ θ a (

deg

rees

)

70

75

80

85

90

95

100

105

110

InW B (W

B in

PPT)

θ θ a (

deg

rees

)

θ θ a (

deg

rees

)

70

80

90

10

110

100

90

80

70

θ θ a (

deg

rees

)

110

100

90

80

70

θ θ a (

deg

rees

)

110

100

90

80

70

1000

2000

3000

4000 2826

2422

2018

1614

1210

8

InW B (W

B in

PPT)

InW B (W

B in

PPT)

2826

2422

2018

1614

1210

8

2826

2422

2018

1614

1210

8

Time (seconds)

1000

2000

3000

4000

Time (seconds)

1000

2000

3000

4000

Time (seconds)

0

110

(A)

(B)

(C)

Fig. 1. Advancing contact angle isotherms in 3D (ya as a function of analysis

2D (ya as a function of logarithmic solution concentration WB at selected time

and equine (ES, panel C) serum. In each case, solute concentration WB is expre

in 2D panels represent time slices through 3D representations (filled circl

annotations in Panel A indicate maximum and half-maximum contact angle re

attributed to slow hydration of the SAM surface (Panel B, arrow annotation;

the drop was deposited on the surface by the robot. The first

120 s (20 images) monitored evolution of the advancing angle. At the

end of the 1 h ya measurement period, the platform was programmed to

return to horizontal and rotate 151 to the next analysis position along

the periphery of the semiconductor wafer. This process was repeated

for all dilutions of the protein solution under study so that results

reported for each protein or plasma/serum solution were obtained on a

single test surface, eliminating the possibility of substratum-to-substratum

variation within reported results. We observed that the contact angle

of a pure PBS droplet slowly decreased with time from the initial value

of 108�oy�ao106� at t ¼ 0 to 104�oy�ao102� at t ¼ 1 h; where y�a is

the pure-buffer contact angle and the range of reported results

8 10 12 14 16 18 20 22 24 26 28

8 10 12 14 16 18 20 22 24 26 28

InWB (WB in PPT)

InWB (WB in PPT)

8 10 12 14 16 18 20 22 24 26 28

InWB (WB in PPT)

adsorption time

aθ h

'aθ

maxaθ

max1

2aθ

/ 2ln BC Θ

hydration time

time (drop age) and logarithmic (natural) solution concentration WB) and

s) formats comparing human (HS, panel A, Table 1), ovine (OS, panel B),

ssed in v=v units of parts-per-trillion (PPT, mL/1012mL solution). Symbols

e ¼ 0.25 s, open circle ¼ 900 s, triangles ¼ 1800 s, and squares ¼ 3594 s;

duction, ymaxa and 1=2ymax

a respectively. Note also decrease in y0a with time,

see Results section for more discussion).

ARTICLE IN PRESSA. Krishnan et al. / Biomaterials 27 (2006) 3187–31943190

corresponds to all of the 8 methyl-terminated SAM surfaces analyzed

during the course of this work. We attribute this systematically vary-

ing wettability to slow ‘‘hydration’’ of SAM surfaces, which may

include dissolution of thiol from the surface into solution, but have no

specific proof this was the cause. We do not believe this slight

but apparently unavoidable attribute of thiol-SAMs on gold-coated

silicon wafers negatively affects the veracity of conclusions drawn herein

which were based on final, steady-state measurements made at �1 h

analysis time.

2.4. Computation and data representation

Computational, statistical, and theoretical methods used in this work

have been discussed in detail elsewhere [17,19,20]. In brief, time-dependent

ya data corresponding to protein dilutions were recovered from TPG files

and correlated with concentrations, leading to a matrix of results with row

values representing concentration and time (in sec) as column values. It

was generally observed that ‘‘ya isotherms’’ plotting ya against w=v protein

concentration WB were sigmoidal in shape when plotted on logarithmic-

concentration (lnWB) axes [17,19] with well-defined, low-concentration

asymptotes y�a and high-concentration asymptotes y0a (see Fig. 1 for

example). Protein concentration WB was herein expressed in parts-per-

trillion (i.e. PPT, grams solute/1012 g solvent) so that the lnWB scale has

positive values at all concentrations. Successive non-linear least-squares

fitting of a four-parameter logistic equation ½ya ¼ ðy�a � y0aÞ=1þ

ðln WY=2B = ln WBÞ

Mþ y0a� to contact angle isotherms data for each

time within the observation interval quantified y�a and y0a parameters

with a measure of statistical uncertainty. Fitting also recovered a

parameter measuring concentration-at-half-maximal-change in ya,ln W

Y=2B ðwhereY=2 ¼ 1=2Ymax andYmax � y�a � y0aÞ, as well as a para-

meter M that measured steepness of the sigmoidal curve. This multi-

parameter fitting to concentration-dependent ya data was a purely

pragmatic strategy that permitted quantification of best-fit protein and

surfactant characteristics but is not a theory-based analysis [17,19,20].

Three-dimensional (3D) representations of time-and-concentration ya datawere created in Sigma Plot (v8) from the data matrix discussed above

and overlain onto fitted-mesh data computed from least-squares fitting.

Two-dimensional (2D) representations were created from the same data

matrices at selected observation times. Measured ya were converted to

advancing adhesion tension ta ¼ glv cos ya for general interpretation [17];

where glv is the interfacial tension of the contact-angle fluid.

Adhesion tensions t0a ¼ g0lv cos y�a (pure saline) and t0a ¼ g01v cos y0a (at

the minimum contact angle observed y0a) were computed with fitted

parameters g0lv and g0lv reported in Ref. [13]. Smoothed adhesion-tension

isotherms (ta vs. ln WB) were computed from smoothed ya isotherms

above, using smoothed glv values computed from best-fit parameters

reported in Ref. [13]. Likewise, smoothed spreading pressure isotherms

(Pa vs. ln WB) were computed from smoothed ta isotherms, where

Pa � ðta � t0aÞ.

Table 1

steady-state plasma/serum adsorption parameters

Protein source (acronym) y0a (degrees) y0a (degrees) ln CY=2B PPT

Human plasma (HP) 104.270.6 79.671.6 20.170.4

Human serum (HS) 100.470.9 77.771.0 17.570.3

Bovine plasma (BP) 100.570.6 76.671.8 20.370.5

Bovine serum (BS) 101.670.6 86.470.8 17.770.4

Equine plasma (EP) 101.470.5 80.270.5 18.170.2

Equine serum (ES) 100.670.5 79.170.9 18.970.3

Ovine plasma (OP) 100.970.5 86.171.1 21.170.4

Ovine serum (OS) 101.770.5 85.570.9 19.970.3

Listed data result from statistical fit (see Materials and Methods).

3. Results

3.1. Contact-angle isotherms

The principal experimental observations of this workwere time- and concentration-dependent aqueous-buffer/solid (solid–liquid, SL) contact angles ya of human andanimal (bovine, ovine, equine) blood plasma and serum. yaof serially diluted, whole plasma or serum was measuredusing tilting-plate goniometry (TPG), resulting in graphicalconstructions herein referred to as ‘ya isotherms’. Table 1compiles quantitative results of this work. Contact-angleparameters y�a, y0a, ln W

Y=2B and M listed in cols 2–5 of

Table 1 are the mean fitted values corresponding to final 25ya isotherms recorded within the 60-min time frame of theTPG experiment (see Methods and Materials). Listed erroris the standard deviation of this mean. Correspondingadhesion tensions t0a and t0a (Columns 6, 7) were computedfrom y�a and y0a values, respectively, with uncertaintyestimates computed by propagation of error in ya and glvmeasurements (Methods and Materials). Maximumspreading pressure Pmax

a � ðt0a � t0aÞ (Column 8) wascomputed directly from aforementioned ta values andassociated uncertainty estimated by propagation of error.Fig. 1 compiles representative TPG results for human

(HS, panel A), bovine (BS, panel B), and equine bloodserum (ES, panel C). Concentration-dependent ya followeda sigmoidal-like function on logarithmic concentration axiswith characteristic parameters y�a, y0a, ln W

Y=2B and M

compiled in Table 1. In each case represented in Fig. 1, yawas observed to decrease with time to a final, steady-statey0a value. Dynamics were undoubtedly due to protein mass-transfer and adsorption-kinetic effects. Observation oftime-dependence was important to this particular workonly in so far as data demonstrate that a steady-state(equilibrium) was achieved, or at least asymptoticallyapproached, within the 1 h observation window. Table 1refers only to steady-state measurements. We also observedthe ‘‘SAM hydration’’ phenomenon that slowly shifted thewhole contact angle isotherm (ya vs. concentration) lowerwith time (see annotations in Fig. 1B and Methods andMaterials section). Note that computation of spreading

M (dimensionless) t0a (mN/m) t0a (mN/m) Pmaxa (mN/m)

�9.271.6 �17.670.8 8.871.4 26.471.9

�13.673.2 �12.771.1 10.170.8 22.971.4

�9.871.8 �13.270.8 8.971.5 22.171.7

�12.973.1 �14.370.7 2.670.6 16.970.9

�36.177.3 �13.670.6 8.170.4 21.671.1

�9.471.2 �12.970.4 8.470.6 20.770.7

�20.876.6 �13.670.6 3.070.9 16.671.1

�14.973.3 �14.570.6 3.570.7 17.970.9

ARTICLE IN PRESSA. Krishnan et al. / Biomaterials 27 (2006) 3187–3194 3191

pressure Pa � ðta � t0aÞ discussed below effectively correctsfor the SAM-hydration phenomenon.

3.2. Adhesion-tension and spreading-pressure isotherms

Fig. 2 traces sequential interpretation of steady-state (1 hdrop age), concentration-dependent ya data (Panel A) interms of concentration-dependent ta (Panel B) andspreading pressure Pa (Panel C) for human serum (HS).Smooth curves through the data serve as guides to the eye.Steady-state (equilibrium) spreading pressure isotherms Pa

were used as the basis of comparison of proteins listed in

θ θ a (

deg

rees

)

75

80

85

90

95

100

105

τ τ a (

mN

/m)

-10

-5

0

5

10

ln WB (WB

8 10 12 14 16 18

a (m

N/m

)

0

5

10

15

20

25

'aθ

'aτ

�

(A)

(B)

(C)

Fig. 2. Sequential interpretation of a steady-state (3600 s drop age) contact ang

contact angles, ya; Panel B, advancing adhesion tension, ta; Panel C, advancingto the eye. Annotations identify low- and high-concentration asymptotes for

pressure Pmaxa that are used to characterize isotherms (Table 1).

Table 1. Fig. 3 graphically compares steady-state, concen-tration-dependent Pa (SL interface, upper panel) to glv (LVinterface, lower panel) obtained for human immunoglobu-lin-G (IgG, MW ¼ 160 kDa), human plasma, and serumon a single concentration axis. Lines surrounding the datarepresent 99% confidence intervals around the best-fithuman-plasma data listed in Table 1. Evidently, adsorptionof the purified protein cannot be confidently distinguishedfrom plasma or serum on the basis of this data andadsorption to the SL interface was effectively identical tothe LV interface. Fig. 4 is a similar construction comparingadsorption of human, bovine, ovine, and equine plasma.

in PPT)

20 22 24 26 28

aθ ο

aτ ο

max� = '

τ

τ− οa a a

le adsorption isotherm for human serum (HS, Table 1); Panel A, advancing

spreading pressure Pa. Smoothed curves through the data serve as guides

contact angles (y�a, y0a), adhesion tensions (t0a, t

0a) and maximum spreading

ARTICLE IN PRESS

8 10 12 14 16 18 20 22 24 26 28

γ γ lv (m

N/m

)

45

50

55

60

65

70

ln WB (WB in PPT)

� a

(mN

/m)

0

5

10

15

20

25

30

max~ 25 mN/m�

max~ 25 mN/m�

Fig. 3. Comparison of steady-state, concentration-dependent Pa (upper

panel) and glv (lower panel) isotherms for (human) plasma (HP), serum

(HS), and IgG (hIgG) on a single concentration axis, showing that

adsorption from protein mixtures cannot be confidently distinguished

from a purified protein at hydrophobic surfaces represented by the

aqueous-buffer/air (LV) and aqueous-buffer/methyl-terminated SAM

(SL) interface. Lines surrounding data represent 99% confidence intervals

around best-fit-human plasma data (closed circle ¼ HP, open

square ¼ HS, open triangle ¼ hIgG). Note that maximum spreading

pressure Pamax�25mN/m for both LV and SL interfaces.

ln WB (WB in PPT)

8 10 12 14 16 18 20 22 24 26 28

γ γ lv (

mN

/m)

45

50

55

60

65

70

�

a (m

N/m

)

0

5

10

15

20

25

30

max~ 25 mN/m�

max~ 25 mN/m�

Fig. 4. Comparison of steady-state, concentration-dependent Pa (upper

panel) and glv (lower panel) isotherms for human (HP, filled circle, Table

1), ovine (OP, open triangle), equine (EP, open square) and bovine (BP,

open circle) plasma on a single concentration axis. Lines surrounding data

represent 99% confidence intervals around best-fit-human plasma,

suggesting that species cannot be confidently distinguished from this data.

In each case, solute concentration WB is expressed in v=v units of parts-

per-trillion (PPT, mL/1012mL solution). Note that maximum spreading

pressure Pamax�25mN/m for both LV and SL interfaces for all species.

A. Krishnan et al. / Biomaterials 27 (2006) 3187–31943192

As in Fig. 3, lines surrounding the data represent 99%confidence intervals about the best-fit human-plasma.Evidently, adsorption of plasma derived from thesedifferent mammalian species cannot be confidently distin-guished on the basis of this data, in spite of the differencesin plasma proteome, and adsorption to the SL interfacewas effectively identical to the LV interface. Collectively,Figs. 1–4 show a striking similarity in the adsorption of apurified protein, plasma, and serum at two differenthydrophobic surfaces.

4. Discussion

Similarity in adsorption characteristics of plasma andserum derived from diverse mammalian species (bovine,equine, ovine, human) to hydrophobic LV (aqueous-buffer/air) and SL (aqueous-buffer/SAM) surfaces is veryintriguing in light of the substantial differences in proteincomposition between plasma and serum, as well as the

differences in the plasma proteome among species. Theseresults are all the more compelling in light of our recentfindings that adsorption of purified blood-plasma proteinsto LV and SL surfaces are themselves very similar[11,13,15,16] and, as shown herein, not remarkablydifferent than the adsorption behavior of whole plasmaand serum. These experimental outcomes corroboratework dating back to the early 1900s [27–30] as well as themore modern work of Lyman, Brash, and Andrade[2–6,31], but results are generally inconsistent with theexpectation that has arisen over time within the biomater-ials community that different proteins adsorb differently tothe same surface (see Introduction). Specifically, we findthat interfacial tension ðglvÞ and spreading pressure (Pa)isotherms of plasma, serum, and purified proteins arequalitatively similar (see Fig. 3 comparing human plasma,serum, and purified IgG for example) and that saturating-spreading pressures Pa

max for purified-blood proteinsspanning 3 decades in molecular weight fall within

ARTICLE IN PRESSA. Krishnan et al. / Biomaterials 27 (2006) 3187–3194 3193

10oPamaxo25mN/m [16]. This full range in Pa

max forpurified-blood proteins should be compared to the17oPa

maxo26mN/m range observed for plasma/serumlisted in Table 1. Evidently, neither profound biochemicaldifferences between plasma and serum nor substantialvariability in the proteome of donor species [32] lead to arange of adsorption energetics greater than observedamong purified-blood proteins. Simply summarized, inter-facial energetics of adsorption of plasma and serumproteins to LV and SL test surfaces is very similar to thatof purified-blood constituents.

All taken together, we find this evidence suggestive of aheretofore unresolved mechanistic commonality under-lying protein adsorption to test hydrophobic surfacesconsidered herein. Our recently developed theory ofprotein adsorption to hydrophobic surfaces (liquid–vapor,LV and methyl-terminated SAM, SL) [15] asserts thatthe mechanistic commonality mentioned above is relatedto the behavior of water at interfaces [10,18,33].Theory suggests that interfacial capacity for protein iscontrolled by the extent to which interfacial water can bedisplaced by adsorbing protein molecules (energetics ofsurface dehydration). This maximal interfacial capacitycan be expressed either as a maximal volume fractionor, equivalently, a maximum weight/volume (not molar)concentration denoted W max

I at which surface satura-tion occurs. Calibration of theory to experimental neutron-reflectivity or light scattering data reveals thatWI

max�436mg/mL [15,16], consistent with the conclusion

drawn from a survey of protein-adsorption literature [10]that adsorbed protein concentrations can be surprisinglylarge. Importantly, W max

I is found not to be stronglydependent on globular-protein identity but rather is ageneric, water-limited adsorption capacity for protein.Hence, W max

I can be attained by any mixture of proteins,which is of relevance in the interpretation of the observedsimilarity in adsorption of proteins and protein mixtures asdiscussed further below.

First, fixed W maxI immediately explains the experimental

observation that Pa isotherms for diverse purified proteinsspanning three decades of MW appear more similar thandissimilar when scaled on a w=v basis [15]. Serial dilution ofa protein solution at-or-above surface-saturating concen-trations exhibiting the characteristic spreading pressurePa

max (and adhesion tension t0a) will trace similar concen-tration dependence because ðdPa=d ln WBÞ of individualproteins are very similar. Second, fixed W max

I rationalizesthe experimental observation that Pa isotherms for plasmaand serum are substantially similar, regardless of species.At 60–85mg/mL total protein (including fibrinogen) [34],plasma and serum is well over the �3mg/mL required tosaturate a hydrophobic surface at W max

I , thus achievingPa

max (see Fig. 1A). Depletion of coagulation proteins inthe conversion of plasma to serum certainly changes thenumber and concentration of individual protein constitu-ents in bulk solution, but does not alter W max

I because thebulk-solution concentration is well above that required to

saturate a hydrophobic surface. As a consequence, Pa

isotherms of plasma and serum are nearly identical becauseother protein constituents compete for the interfacialvacancies effectively created by removal of proteinsconsumed in the coagulation process (e.g. conversion offibrinogen to insoluble fibrin). Third, although variationsin plasma proteome among species lead to a differentweight-fraction distribution of proteins in the bulk phaseand surface, these differences do not affect W max

I . And,because ðdPa=d ln WBÞ of individual protein constituentsare very similar, the net observed Pa isotherms for plasmaderived from different species are not remarkably different.Apparently, the widely held expectation that differentproteins adsorb differently to the same surface does nothold true, at least for the test hydrophobic surfacesconsidered herein.

5. Conclusions

The principal conclusion drawn from this and relatedwork [11,12] is that the interfacial energetics of adsorptionof plasma and serum derived from diverse mammalianspecies to hydrophobic LV (aqueous-buffer/air) and SL(aqueous-buffer/methyl SAM) surfaces are very similar.Furthermore, it is found that adsorption of purified-plasmaproteins to these same surfaces cannot be confidentlydistinguished from that of plasma or serum. Results arerationalized by suggesting there is a generic mechanismunderlying adsorption to hydrophobic surfaces that holdsfor purified proteins as well as protein mixtures. Wepropose that this generic mechanism is essentially thehydrophobic effect in which protein is expelled fromaqueous solution in order to maximize self-association[10,33] at the expense of less-favorable water–proteininteractions. Expelled protein accumulates at the hydro-phobic interface, displacing relatively loosely bound water-of-hydration. However, energetics of this surface dehydra-tion place a cap on the amount of water that can bedisplaced by adsorbing protein. This cap on the amount ofwater that can be displaced translates into a maximum w/vconcentration of protein within the interface because thepartial specific volume v0 of globular proteins is quiteconserved (0.70pv0p0.75 cm3/g protein). Thus, on aweight basis, any blood protein or mixture of proteinsdisplaces nearly an equivalent amount of interfacial water.

Acknowledgments

This work was supported, in part, by the NationalInstitute of Health PHS 5 R01 HL 69965-03, by Johnson &

Johnson through the Focused Giving Grant Program, anda Grant with the Pennsylvania Department of Health.The Department Pennsylvania Department of Healthspecifically disclaims responsibility for any analyses,interpretations or conclusions.. Authors appreciate addi-tional support from The Materials Research Institute andDepartments of Bioengineering and Materials Science and

ARTICLE IN PRESSA. Krishnan et al. / Biomaterials 27 (2006) 3187–31943194

Engineering, Penn State University. Authors gratefullyacknowledge the assistance of Dr. Roger Woodward ininstrument design and implementation.

References

[1] Johlin JM. Interfacial adsorption as a function of the concentration

of colloidal solutions. J Biol Chem 1929;84:543–51.

[2] Lyman DJ, Muir WM, Lee IJ. The effect of chemical structure and

surface properties of synthetic polymers on the coagulation of blood.

I. Surface free energy effects. Trans Am Soc Artif Int Organs 1965;

21:301–6.

[3] Lyman DJ, Brash JL, Chaikin SW, Klein KG, Carini M. The effect

of chemical structure and surface properties of synthetic polymers on

the coagulation of blood. II. Protein and platelet interaction with

polymer surfaces. Trans Am Soc Artif Int Organs 1968;14:250–5.

[4] Lyman DJ, Knutson K, McNeill B, Shibatani K. The effect of

chemical structure and surface properties of synthetic polymers on

the coagulation of blood. IV. The relation between polymer

morphology and protein adsorption. Trans Am Soc Artif Int Organs

1975;11:49–54.

[5] Brash JL, Lyman DJ. Adsorption of plasma proteins in solution to

uncharged, hydrophobic polymer surfaces. J Biomed Mat Res 1969;

3:175–89.

[6] Brash J, Lyman D. Adsorption of proteins and lipids to nonbiolo-

gical surfaces. In: Brash JL, editor. The chemistry of biosurfaces.

New York: Marcel Dekker; 1971. p. 177–222.

[7] Tengvall P, Lundstrom I, Liedberg B. Protein adsorption studies on

model organic surfaces: an ellipsometric and infrared spectroscopic

approach. Biomaterials 1998;19:407–22.

[8] Malmsten M, editor. Biopolymers at interfaces. New York: Marcel

Dekker; 1998.

[9] Green RJ, Davies MC, Roberts CJ, Tendler SJB. Competitive protein

adsorption as observed by surface plasmon resonance. Biomaterials

1999;20:385–91.

[10] Vogler EA. Structure and reactivity of water at biomaterial surfaces.

Adv Colloid Interf Sci 1998;74(1-3):69–117.

[11] Krishnan A, Liu Y-H, Cha P, Allara DL, Vogler EA. Scaled

interfacial activity of proteins at a hydrophobic solid/aqueous-buffer

interface. J Biomed Mater Res 2005;75A(2):445–57.

[12] Krishnan A, Wilson A, Sturgeon J, Siedlecki CA, Vogler EA.

Liquid–vapor interfacial tension of blood plasma, serum and purified

protein constituents thereof. Biomaterials 2005;26:3445–53.

[13] Krishnan A, Sturgeon J, Siedlecki CA, Vogler EA. Scaled interfacial

activity of proteins at the liquid–vapor interface. J Biomed Mater Res

2004;68A:544–57.

[14] Krishnan A, Siedlecki CA, Vogler EA. Mixology of protein solutions

and the vroman effect. Langmuir 2004;20(12):5071–8.

[15] Krishnan A, Siedlecki C, Vogler EA. Traube-rule interpretation of

protein adsorption to the liquid–vapor interface. Langmuir 2003;19:

10342–52.

[16] Krishnan A, Liu Y-H, Cha P, Allara DL, Vogler EA. Interfacial

energetics of globular-blood protein adsorption to a hydrophobic

surface from aqueous-buffer solution. J Roy Soc Interf 2005 in press.

[17] Vogler EA. Interfacial chemistry in biomaterials science.

In: Berg J, editor. Wettability. New York: Marcel Dekker; 1993.

p. 184–250.

[18] Vogler EA. How water wets biomaterials. In: Morra M, editor.

Water in Biomaterials Surface Science. New York: Wiley; 2001.

p. 269–90.

[19] Vogler EA. Practical use of concentration-dependent contact angles

as a measure of solid–liquid adsorption I: theoretical aspects.

Langmuir 1992;8:2005–12.

[20] Vogler EA. Practical use of concentration-dependent contact angles

as a measure of solid–liquid adsorption II: experimental aspects.

Langmuir 1992;8:2013–20.

[21] Krishnan A, Liu Y- H, Cha P, Allara DL, Vogler EA. An evaluation

of goniometric methods. J Colloid Interf Sci 2005;43:95–8.

[22] Allara DL, Nuzzo RG. Spontaneously organized molecular assem-

blies. 2. Quantitative infrared spectroscopic determination of

equilibrium structures of solution-adsorbed n-alkan oic acids on an

oxidized aluminum surface. Langmuir 1985;1(1):52–66.

[23] Nuzzo RG, Allara DL. Adsorption of bifunctional organic disulfides

on gold surfaces. J Am Chem Soc 1983;105(13):4481–3.

[24] Nuzzo RG, Fusco FA, Allara DL. Spontaneously organized

molecular assemblies. 3. Preparation and properties of solution

adsorbed monolayers of organic disulfides on gold surfaces. J Am

Chem Soc 1987;109:2358–68.

[25] Nuzzo RG, Dubois LH, Allara DL. Fundamental studies of

microscopic wetting on organic surfaces. 1. Formation and structural

characterization of a self-consistent series of polyfunctional organic

monolayers. J Am Chem Soc 1990;112:558–69.

[26] Porter MD, Bright TB, Allara DL, Chidseyi CED. Spontaneously

organized molecular assemblies. 4. Structural characterization

of n-alkyl thiol monolayers on gold by optical ellipsometry,

infrared spectroscopy, and electrochemistry. J Am Chem Soc 1987;

109:3559–68.

[27] Morgan JLR, Woodward HE. The weight of a falling drop and the

laws of tate. J Am Chem Soc 1913;35:1249–62.

[28] DuNuoy PL. Surface tension of serum. J Exp Med 1925;41:663.

[29] DuNouy PL. Surface tension of serum. XIII. On certain physico-

chemical changes in serum as a result of immunization. J Exp Med

1925;41:779–93.

[30] Harkins HN, Harkins WD. The surface tension of blood serum, and

the determination of the surface tension of biological fluids. J Clin

Invest 1929;7(2):263–81.

[31] Tripp BC, Magda JJ, Andrade JD. Adsorption of globular

proteins at the air/water interface as measured via dynamic

surface tension: concnetration dependence, mass-transfer consi-

derations, and adsorption kinetics. J Colloid Interf Sci 1995;173:

16–27.

[32] Swenson MJ. Physiological properties and cellular and chemical

constituents of blood. In: Swenson MJ, editor. Duke’s physiology of

domestic animals. London: Cornell University Press; 1977.

[33] Vogler EA. Water and the acute biological response to surfaces.

J Biomat Sci Polym Edn 1999;10(10):1015–45.

[34] Adkins JN, Varnum SM, Auberry KJ, Moore RJ, Angell NH, Smith

RD, et al. Toward a human blood serum proteom. Molec Cell

Proteom 2002;1(12):947–55.