Enzymes in the Environment: Activity, Ecology and Applications - Chapter 11 ppsx

11
Enzyme Adsorption on Soil Mineral
Surfaces and Consequences for the
Catalytic Activity
Herve
´
Quiquampoix
Institut National de la Recherche Agronomique, Montpellier, France
Sylvie Servagent-Noinville and Marie-He
´
le
`
ne Baron
Centre National de la Recherche Scientifique, Universite
´
Paris VI, Thiais, France
I. INTRODUCTION
Soil enzymes are either actively secreted by living microorganisms and plant roots or
released after the death of soil biota by cell lysis. One class of enzymes, the hydrolases,
have a very important role in the biogeochemical cycles of major elements (C, N, P, and
S) since their substrates in soil are mainly in a polymerized form. Usually microorganisms
(and plant roots) cannot take up macromolecules directly from the external medium. There
are few exceptions to this, such as the uptake of fragments of deoxyribonucleic acid (DNA)
or plasmids, that lead to transformation in bacteria. In general, the membrane transport
systems are specific and recognize the universal biological monomers, such as amino
acids, sugars, and nucleotides; low number oligomers, such as cellobiose or maltose; or
mineral ionic groups that can be released by enzymatic hydrolysis of organic molecules,
such as orthophosphate and sulfate. Extracellular enzymes perform three main functions
in soil: (1) they reach substrates in pores with dimensions roughly 100 times smaller
than those of bacteria; (2) they hydrolyze these substrates and make them soluble and
consequently able to diffuse back to the microorganisms or plant roots; and at the same

time (3) they transform polymers into monomers or oligomers that can be recognized and
taken up by membrane transport systems to undergo intracellular metabolism.
As enzymes are proteins, they share with this class of macromolecules their strong
affinity with interfaces. Proteins are potentially flexible polypeptide chains, even if they
have a stable folded configuration in solution, with individual amino acids whose lateral
chains have various physicochemical properties: hydrophilic or hydrophobic; negatively,
neutrally, or positively charged. From a thermodynamic point of view, these properties
give rise to both enthalpic (related to intermolecular forces) and entropic (related to the
spatial arrangement of the molecules) contributions to the interactions with surfaces. The
strong and often largely irreversible adsorption of enzymes on the mineral phase of the soil
Copyright © 2002 Marcel Dekker, Inc.
has important consequences, not only for their mobility, but also for their survival and
catalytic activity.
The most well-known effect of the adsorption of enzymes on negatively charged
surfaces, such as clay minerals, is a shift of their optimal catalytic activity towards a higher
pH range (1–4). A second general effect, which can be extended to all proteins, is that
the maximal adsorption is observed near the isoelectric point (i.e.p.) of the enzyme.
McLaren et al. described these two properties in the 1950s (5–7). However, until the
beginning of the 1990s, no effort, supported by independent structural study of the ad-
sorbed enzymes or proteins by modern physical methods, was made to propose a mecha-
nism explaining both observations.
II. NATURE OF THE DRIVING FORCES LEADING TO ADSORPTION
OF ENZYMES ON SURFACES
Studies on the quantity of protein adsorbed on surfaces cannot be separated from the study
of their conformation on these surfaces. The reason is that a modification toward a more
disordered structure contributes to the driving forces of adsorption, since it increases the
entropy of the system and thus decreases the Gibbs energy. The modification of conforma-
tion also can have an effect on the maximal quantity of protein adsorbed, since conforma-
tional changes may affect the area occupied by each single protein on the surface.
The spontaneous adsorption of proteins at constant temperature and pressure leads

to a decrease of the Gibbs energy of the system, according to the second law of thermody-
namics (8–13). The Gibbs energy, G, depends on enthalpy, H, which is a measure of the
potential energy (energy that has to be supplied to separate the molecular constituents
from one another), and entropy, S, which is related to the disorder of the system:
∆
ads
G ϭ ∆
ads
H Ϫ T∆
ads
S Ͻ 0
where T is the absolute temperature and ∆
ads
is the change in the thermodynamic functions
resulting from adsorption.
Some difficulties arise in the analysis of these processes because enthalpic effects,
related to intermolecular forces, and entropic effects, related to the spatial arrangements
of molecules, are not totally independent. Intermolecular forces influence the distribution
of molecules, and the potential energy also is also dependent on the molecular structure of
the system.
A. Enthalpic Effects
1. Coulombic Interactions
The electrical charge of proteins results from the ionization of the carboxylic, tyrosyl,
amine, imine, and imidazole groups of the side chains of some amino acids. The electrical
charge of mineral surfaces can result from pH-independent isomorphic substitutions in
the crystal lattice, as in some clays (basal surfaces of illite or montmorillonite), or from
pH-dependent ionization of hydroxyls (edge sites of clays, oxyhydroxides). Coulombic
forces are very strong and long-range intermolecular forces. They can be screened by the
ions in the solution. As all electrical charges have to be compensated by an equal number
of electrical charges of the opposite sign, a diffuse double layer is established around the

Copyright © 2002 Marcel Dekker, Inc.
macromolecules and mineral surfaces. The electrostatic interactions between proteins and
surfaces thus can be analyzed as an overlap of their electrical double layers (8,10). The
electrostatic part of the Gibbs energy is given by the isothermal and isobaric work of
charging the electrical double layer:
G
el
ϭ
Ύ
σ
0
φ(σ)dσ
where φ is the variable electrostatic potential and σ is the variable surface density during
the charging process.
2. Lifshitz–van der Waals Interactions
Contrary to the Coulombic interactions, van der Waals forces act on all molecules, even
if they are electrically neutral. They are short-range forces and are composed of three
different components. The main component are the dispersion (or London) forces, which
originate from the instantaneous dipolar moment resulting from the fluctuation of the
electrons around the nuclear protons. The electric field created induces, by polarization,
a dipole moment in nearby molecules, which, in turn, creates an instantaneous attractive
interaction. The two other components are the induction (or Debye) forces, related to the
interaction between a polar molecule and a nonpolar molecule, and the orientation (or
Keesom) forces, related to the interaction between two polar molecules.
B. Entropic Effects
1. Hydrophobic Interactions
The stability of proteins in solution results mainly from the shielding of amino acids with
a hydrophobic side chain in the core of the protein from contact with water. It is due to
the hydrophobic effect that causes water molecules around a nonpolar group to establish
more hydrogen bonds among themselves than around a polar group. This process maxi-

mizes the mutual association of water molecules by hydrogen bonds and results in an
increased order of the surrounding water, and thus a favorable decrease in entropy of the
system.
Sometimes hydrophobic interactions are involved in interactions of proteins with
hydrophilic mineral surfaces. An example is given by the higher affinity of a hydrophobic
methylated derivative of bovine serum albumin (BSA) for the hydrophilic montmorillonite
surface than the native, less hydrophobic BSA (14). The explanation is that the adsorption
of proteins is accompanied by the exchange of charge-compensating cations on the clay,
which are the true hydrophilic centers, leaving a hydrophobic siloxane layer (15). Thus,
the montmorillonite surface presents hydrophilic properties for molecules whose adsorp-
tion does not result in the removal of charge-compensating cations, and hydrophobic prop-
erties for molecules that replace the charge-compensating cations.
2. Modifications in Protein Molecular Structure
The entropic contribution to adsorption also can result from a modification of the confor-
mation of the protein. This phenomenon is related to an increase of the rotational freedom
of the peptide bonds engaged in secondary structures, such α helices and β sheets. The
ordered secondary structures are an important part of the densely packed hydrophobic
core of proteins. After adsorption, internal hydrophobic amino acids can reach more exter-
nal positions in contact with the surface, since the amino acids remain shielded from
Copyright © 2002 Marcel Dekker, Inc.
contact with the water molecules of the surrounding solvent phase. If a decrease of internal
ordered secondary structures accompanies this process, it results in an increase of confor-
mational entropy. The gain of conformational entropy, S
conf
, can be calculated from the
assumption that four different conformations are possible for peptide units in random
structures as compared with only one in α helices and β sheets:
∆
ads
S

conf
ϭ R ln 4
n
where R is the molar gas constant and n is the number of peptide units involved in the
transfer from an ordered secondary structure to a random secondary structure (9,10).
III. EXPERIMENTAL EVIDENCE FOR pH-DEPENDENT CHANGES IN
THE STRUCTURE OR ORIENTATION OF ADSORBED PROTEINS
The previous thermodynamical considerations indicate that modifications in conformation
are an important parameter to consider in the adsorption of proteins, since the entropic
effect related to these structural changes is itself a factor of adsorption. These three-dimen-
sional changes affect the catalytic activity of the adsorbed enzymes.
A major difficulty in the evaluation of the extent of such events is that no experimen-
tal method allows the direct measurement of the conformation of proteins in an adsorbed
state. Only two methods are suitable for the determination of the tertiary structure of the
proteins, and neither can be employed when proteins are adsorbed. One method, X-ray
diffraction, necessitates the preparation of protein crystals, which is impossible for ad-
sorbed proteins. The other, nuclear magnetic resonance (NMR) spectroscopy, is confined
to molecules with a sufficiently high tumbling rate to obtain spectra with narrow linewidth
peaks, a condition not compatible with the adsorption on a surface of larger dimension than
the protein itself since even a small adsorbed molecule experiences the slower rotational
movement of the mineral colloid.
Without dramatic advances in solid-state NMR spectroscopy, information on the
conformation can be deduced only from lower levels of structural information than the
tertiary structure, such as the secondary structure and the specific interfacial area occupied
by adsorbed proteins. Two spectroscopical approaches that permit such investigations are
now discussed.
A. Study of the Interfacial Area of Protein-Surface Contact by
Nuclear Magnetic Resonance Spectroscopy
The study of adsorption isotherms of proteins on clay mineral surfaces has been disap-
pointing with regard to the interpretation of the adsorbed enzyme activity. The idea that

can be advanced is that the quantity of protein adsorbed is by itself insufficient to describe
a complex phenomenon that can involve at least four other parameters: (1) the orientation
of the protein on the surface (this is important as proteins are rarely perfect spheres and
are more often described as ellipsoids with a long and a large axis; thus an end-on adsorp-
tion involves a higher quantity of protein adsorbed than a side-on adsorption); (2) a possi-
ble unfolding of the protein on the surface changing the interfacial area between individual
protein and surface and the quantity of protein adsorbed at saturation; (3) the surface
coverage at saturation, which could be less than 100% for packing reasons, if the adsorp-
tion is irreversible; and (4) the possibility of a multilayer adsorption. Thus, it is always
Copyright © 2002 Marcel Dekker, Inc.
possibletofindseveralexplanationstointerpretaproteinadsorptionisotherm,withno
experimentalevidenceavailabletochooseamongthem.TheadvantageoftheNMR
methodisthatitsimultaneouslygivesthequantityofadsorbedprotein,thesurfacecover-
ageofthesolidbytheprotein,andthemonolayerormultilayermodeofadsorption(16).
Onlyknowledgeofthesethreefactorsallowsapossibleunfoldingoftheproteinsonthe
claysurfacestobedetectedandquantified.
1.NuclearMagneticResonanceDetectionoftheExchangeofa
ParamagneticCationonProteinAdsorptiononClays
Theprincipleofthemethod(16)isbasedonthefactthattheadsorptionofproteinson
clayscausesthereleaseofcharge-compensatingcations(7,17).Italsousesthesensitivity
oftherelaxationtimesT
1
andT
2
ofnuclearspinstoparamagneticcationsinNMRspectros-
copy(18,19).
Asmallquantity(between3and20µMdependingonthepH)ofaparamagnetic
cation,Mn
2ϩ
,isaddedtoasodium-saturatedmontmorillonitesuspension(1gL

Ϫ1
)with
a10-mMconcentrationoforthophosphate.Thesuspensionisstudiedby
31
PNMRspec-
troscopy.Aninterestingphenomenonisobserved:(1)theMn
2ϩ
cationsthatareadsorbed
ontheclaysurfacedonotinteractatallwiththeorthophosphate,asshownbythecompari-
sonbetweentheclaysuspensionandsupernatantafterremovaloftheclaybycentrifuga-
tion;and(2)theMn
2ϩ
cationsinsolutioninteractwiththeorthophosphate,leadingtoa
linearincreaseofthelinewidthathalfheight,∆ν
1/2
,oftheorthophosphatepeakonthe
NMRspectrum.Thislasteffectistheresultoftheparamagneticcontributiontothede-
creaseofthespin–spinrelaxationtime,T
2
,oftheorthophosphatesignal.Whenagiven
quantityofproteinisintroducedintothissuspension,itdisturbstheequilibriumbetween
theparamagneticMn
2ϩ
adsorbedontheclaysurfaceandthatinsolution.Theanalysisof
theresultinglinewidthoftheorthophosphosphatesignalgivesthequantityofcations
exchangedonadsorption.
Witha300-MHzNMRspectrometer,themeasurementtakesafewminutes;with
a500-MHzspectrometer,1minissufficient(evenlessifhigherconcentrationsofortho-
phosphateareused).Asnocentrifugationisrequiredwiththismethod,thisshorttimeof
signalacquisitioniscompatiblewithkineticstudies.Theresultsareexpressedas∆ν

P
,
whichisthedifferencebetween∆ν
1/2
inthesystemwithparamagneticcationsand∆ν
1/2
inacontrolofthesamecomposition,(butwithoutparamagneticcations)dividedbythe
concentrationofparamagneticcations.Thesurfacecoverageoftheclaybytheprotein
canbededucedfromthefractionofMn
2ϩ
released.Theknowledgeofboththequantity
ofproteinadsorbedandthesurfacecoverageofthesolidallowsthecalculationofthe
interfacialareaofcontactbetweenasingleproteinmoleculeandtheclaysurfaceatdiffer-
entpHandionicstrengths.
2.ConformationalChangesonAdsorptionofaSoftProtein,Bovine
SerumAlbumin
a.DescriptionoftheProgressiveSurfaceCoverageoftheClayFigure1shows
the evolution of ∆ν
p
, i.e., the release of the paramagnetic cation Mn
2ϩ
, when the total
quantity of bovine serum albumin (BSA) introduced in the clay suspension increases, and
at a pH corresponding to the i.e.p. of the BSA (pH 4.7). The increase is linear, followed
by a plateau. The plateau corresponds to the saturation of the montmorillonite surface, as
shown by the comparison with the measurement, by UV adsorption (A
279
nm), of the BSA
in the supernatant solution after centrifugation. The linear increase of ∆ν
p

before the pla-
Copyright © 2002 Marcel Dekker, Inc.
Figure1EffectoftheadditionofbovineserumalbuminonthereleaseofMn
2ϩ
,asdetectedby
itsline-broadeningeffect∆νponorthophosphatebyNMR,andontheUVabsorptionA
279
nmof
theprotein.Whenpresent,themontmorillonitesuspensionisat1gdm
Ϫ3
,pH4.65.(Adaptedfrom
Ref.16.)
teauindicatesthattheadsorbedproteinsalwayshavethesameinterfacialareaofcontact
withtheclaysurface,whateverthesurfacecoverage.Nochange,fromaside-ontoan
end-onstate,orfromanunfoldedtoamorenativestate,resultingfromanincreaseof
lateralrepulsionswithpackingcanbeinvoked.Ifthereweresuchachangeinthemode
ofadsorption,theamountofparamagneticcationreleasedperunitmassofproteinwould
begreateratlowsurfacecoveragethanathighsurfacecoverage,andthiswouldbeseen
asaconvexityratherthanalinearityofthecurve.
b.MonolayerModeofAdsorptionThecomparisonbetweentheMn
2ϩ
exchange
dataandthedepletiondatainFigure1showsalsothatthemaximaladsorptionofBSA
correspondstoamonolayer.Indeed,onlythecontactoftheproteinwiththeclaysurface
canleadtotheexchangeofthecharge-compensatingcations.Asecondlayerwouldin-
volveaprotein–proteincontact,withnoreleaseofMn
2ϩ
.Thus,theoccurrenceofthe
breaksinbothcationexchangeandproteindepletioncurvesatthesameproteinconcentra-
tioniscompatibleonlywithamonolayerofproteinontheclaysurface.

c.MaximumofAdsorptionatthei.e.p.Thedata,suchasthosepresentedin
Fig.1,havebeencollectedoveralargepHrange.Theyallhavethesamegeneralaspect;
onlytwoparametersvary.Fig.2showstheevolutionwithpHofthesetwoparameters:
the plateau amount of BSA adsorbed on montmorillonite, measured by either NMR or
the depletion method, and the maximal fraction of Mn
2ϩ
that is displaced. As often is
observed, the maximal amount of protein adsorbed occurs near the i.e.p. of the protein,
which is 4.7 for BSA. Several hypotheses have been advanced to explain this phenomenon.
One class of hypotheses is based on the same (symmetrical) mechanisms above and below
the i.e.p. to explain the decrease of adsorption. They can be based on the effect of lateral
electrostatic repulsions between the adsorbed proteins, which increase as the pH is more
Copyright © 2002 Marcel Dekker, Inc.
Figure 2 Effect of pH on the maximal amount of bovine serum albumin adsorbed on montmoril-
lonite and on the clay surface coverage followed by the release of Mn
2ϩ
on protein adsorption.
(Adapted from Ref. 16.)
distant from the i.e.p. (20–22). Alternatively, they can be based on a decrease in the
structural stability of the protein when the net electric charge increases, leading to an
unfolding of the protein (7,8,10). In addition, different (asymmetrical) mechanisms can
be invoked above and below the i.e.p., and the comparison of the NMR exchange data
and the protein adsorption data shows that such an asymmetrical mechanism is involved
in the adsorption of BSA on montmorillonite, as explained later.
d. Repulsive Electrostatic Interactions and Protein Unfolding Figure 2 shows
that, above the i.e.p., the maximal amount of BSA adsorbed and the fraction of Mn
2ϩ
exchanged decrease in exactly the same proportion. This good correlation between the
quantity of protein adsorbed and the surface coverage of the clay surface supports a mecha-
nism based on an increase of the electrostatic repulsion between the protein, whose net

negative charge increases with pH above the i.e.p., and the montmorillonite, which carries
a permanent negative charge. Figure 2 also shows that, below the i.e.p., the maximal
amount of BSA adsorbed decreases, but the fraction of Mn
2ϩ
exchanged remains nearly
constant. Below the i.e.p. of the BSA (pH 4.7) there is no important variation in the
number of positively charged side chains of the protein because the pKa of histidine (His),
lysine (Lys), and arginine (Arg) is approximately pH 7, 10, and 12, respectively. The
constant proportion of the cation exchanged by the BSA, despite a decreasing quantity
adsorbed, can be explained only by an unfolding of the protein on the clay surface, moving
more positively charged side chains of His, Lys, and Arg to near the surface. The increase
of the specific interfacial area of the protein resulting from this unfolding is compatible
with a smaller quantity adsorbed at constant surface coverage. It can be calculated from
the data reported in Fig. 2 that the interfacial surface area occupied by a molecule of BSA
on montmorillonite is 60 nm
2
at pH 4.5 and increases to 120 nm
2
at pH 3.0. Again, electro-
static interactions appear to be the main driving force, but here they are attractive since
the protein becomes more positively charged as the pH decreases below the i.e.p. and the
montmorillonite remains negatively charged. BSA is a soft protein since, even at the i.e.p.,
the interfacial area of 60 nm
2
is higher than would be expected from the X-ray structure
Copyright © 2002 Marcel Dekker, Inc.
oftheanalogoushumanserumalbumin,whichhastheshapeofanequilateraltriangle
withsidesof8nmandadepthof3nm(23,24).Ifnomodificationofconformationhad
occuredonadsorptionatthei.e.p.,theinterfacialareaofcontactshouldhavebeen28
nm

2
foraside-onadsorption.ThisishalfofthevalueobtainedbyNMRatpH4.5.
B.StudyoftheModificationinSecondaryStructuresbyFourier
TransformInfraredSpectroscopy
Althoughthedeterminationofthetertiarystructureofadsorbedproteinsisimpossible
withthepresentstateofscientificknowledge,thestudyofchangesintherepartitionof
thedifferentsecondarystructuresonadsorptionispossible.Thisdeterminationallowsthe
deductionoftheoccurrenceofamodifiedconformation,anditalsoallowsdirectcalcula-
tionofthecomponentoftheGibbsenergyofadsorptionthatisrelatedtothesestructural
changes.Circulardichroismandinfraredspectroscopycanbeusedtoinvestigatethesec-
ondarystructureofproteins.Circulardichroismhas,nevertheless,limitationsinturbid
suspensionsbecauseoflight-scatteringeffectsandcanbeappliedonlytoparticleswith
asizebelow30nm(25–27).Fouriertransforminfraredspectroscopy(FTIR)doesnot
havethisdisadvantageandcanbeappliedtomoreturbidsuspensionsofclaysofagreater
size.
1.FourierTransformInfraredSpectralAnalysis
TransmissionFTIRspectrainthe1800-to1500-cm
Ϫ1
regiongiveinformationonthe
protonationstate,thesecondarystructure,andthesolvationoftheprotein.Allsamples
werepreparedin
2
H
2
Omediuminordertoshiftthespectralabsorptiondomainofwater
molecules,boundtothepolypeptidebackboneofthestudiedprotein,outoftheAmideI
andIIspectralrange.Aphosphatebuffer(Na
2
H
2

PO
4
)wasusedatafinalconcentration
of0.055molL
Ϫ1
in
2
H
2
O.Severalp
2
Hvaluesinthe4–12rangewereobtainedbyadding
2
HClorNaO
2
H(11).Thestateoftheproteininsolutionisobtainedfromthespectral
differencebetweentheproteininsolutionandthecorrespondingbuffer;thespectraldiffer-
encebetweenthesolidprotein–claymixtureandthecorrespondingclaysuspensionreveals
thestateoftheadsorbedprotein.Spectraldecompositioncouldbeachievedbysecond-
derivative,curvatureanalysis,orself-deconvolutionprocedures.Thesamenumberofprin-
cipalcomponentsoftheoverallspectrumrange(1500–1800cm
Ϫ1
)wasobtainedatsimilar
wavenumbers(Ϯ1cm
Ϫ1
orless).Aleast-squareiterativecurve-fittingprogram(Leven-
berg–Marquardt)wasappliedtofittheoverallspectrumwiththefoundnumberofprinci-
palcomponents.Thefixedparameters(frequency,IRbandprofile)foranyspectraldecom-
positionallowsacomparativequantitativeanalysisofintensitychangesforeach
componentfromonespectrumtoanother.Examplesofinitialdifferencespectraandspec-

traldecompositionofBSAinsolutionandadsorbedonmontmorillonitearereproduced
inFig.3.
The assignments specific for the Amide I′/I region are deduced from the literature
and our own experiments on model amides, polypeptides, and proteins (28–40). The area
of each Amide I component is expressed as a percentage of the sum of the areas of all
Amide I components. Intensities (percentage peptide CO) are used to deduce the propor-
tion of peptide units involved in the various solvated structural domains of the polypeptide
backbone.
The solvation parameter is given by the percentage of N
2
H. The level of exchange
at a given time depends on the rate at which water molecules gain access to internal
Copyright © 2002 Marcel Dekker, Inc.
Figure3FTIR-vibrationalabsorptionspectra(1750–1500cm
Ϫ1
)andcomputeddecomposition
ofspectralprofilesforBSAinsolution(left)oradsorbedonmontmorillonite(right)atp
2
Hϭ5.6.
(AdaptedfromRef.43.)
peptidegroupsintheproteincore(28–32,41).Theprotonationparameterisgivenby
COO
Ϫ
fractions(percentage)deducedfrommeasurementsoftheareaoftheν(CO)
COOH
absorptionfortheremainingCOOHspecies(withrespecttotheoverallAmideIintensity
atagivenp
2
H).Theserelativeareasareexpressedwithrespecttothecorresponding
relativeareas(percentage)obtainedatlowp

2
HwhenAspandGlusidechainsareall
fullyprotonated(100%COOH).
2.ConformationalChangesonAdsorptionofaSoftProtein,Bovine
SerumAlbumin
AdsorptiononmontmorillonitesurfacesimpliespH-dependentchangesinBSAsolvation,
unfoldingofhelicaldomains,aswellaschangesinhydrationandself-associateddomains
(42,43).
a.BSASolvationFigure4showstheadsorptioneffectsinducedbythenega-
tivelychargedmontmorillonitesurfaceontheBSAsolvationwithp
2
H,proteinconcentra-
tion,andtime.Theadsorptioneffectsalreadyareestablishedafter10min;therelative
intensityissimplymorepronouncedat2hours.Foracidicp
2
H,theweakerexchange
afteradsorptionsuggeststhattheelectronegativesurfaceprotectssomedomainofthe
protein.Incontrast,inthei.e.p.range,adsorptionincreasestheNH/N
2
Hexchange,and
athigherp
2
H,therateofwaterdiffusionisnolongerinfluencedbyadsorption.
b.BSAProtonationBSAadsorptiononmontmorilloniteleadstotheproton-
ationoftheionizablecarboxylicgroupsoftheprotein,asparticacid(Asp),andglutamic
acid(Glu),atleasttop
2
H6.5(Fig.5).Variousreasonsmayexplainsuchashiftofthe
apparent pK
a

of Asp and Glu. In the primary structure of BSA, some Asp and Glu side
chains are adjacent to R
ϩ
functions. Embedded among positively charged side chains inter-
acting with the electronegative clay or embedded in self-associated domains, external Asp
and Glu side chains are assumed to become indifferent to buffer. Moreover, the electroneg-
ative charge of the clay surface could favor a protonation of the Asp and Glu carboxylates
to decrease the coulombic repulsion between the protein and the surface, as observed by
titration on other systems (8,10).
Copyright © 2002 Marcel Dekker, Inc.
Figure4p
2
H-DependentNH/N
2
Hexchange(expressedasN
2
H%)ofBSAin
2
H
2
Oat10min
and2h.∆N
2
H%representsthechangeinproteinsolvationforBSAadsorbedonmontmorillonite
withrespecttothesolution.(AdaptedfromRef.43.)
c.HelixUnfoldingMontmorilloniteinducesimportantunfoldingofhelicaldo-
mainsofBSA(Fig.6).Afteradsorptiononmontmorillonite,thelargelyp
2
H-independent
external helix unfolding is related to new orientations for the Lys

ϩ
and Arg
ϩ
side chains
forced close to the negative clay surface. In contrast, unfolding in internal and packed
helices is largely p
2
H-dependent. This change could be related to the disruption of some
Figure 5 Effect of BSA adsorption on montmorillonite on the p
2
H-dependent Asp and Glu depro-
tonation. (Adapted from Ref. 43.)
Copyright © 2002 Marcel Dekker, Inc.
Figure 6 Effect of BSA adsorption on montmorillonite on the p
2
H-dependent secondary struc-
tures; open symbols, BSA in solution; closed symbols, BSA adsorbed on montmorillonite. Abbrevia-
tions are for the environment of the CO peptide groups (free CO in polar or hydrophobic environ-
ments; H-bonded CO in bundled or external helices, in bents or in protein self-association; hydrated
CO). (Adapted from Ref. 43.)
(Asp
Ϫ
/Glu
Ϫ
)–His
ϩ
salt bridges that enhance helix formation in solution. Adsorption in-
volves protonated His
ϩ
, as proved by the recovery of helices when all His side chains

become deprotonated. At low p
2
H, the protein molecules adsorbed on montmorillonite
spread over the entire mineral surface, as is in agreement with the instability of BSA
structure in solution. It should be noted that the p
2
H-dependent profiles for bundled helices
follow those observed for the decreases of the Amide II bands for both adsorbed and
solution states. Unfolding of bundled/internal helical domains increases water diffusion
inside the core of the protein (Fig. 6).
d. Hydrated and Self-Associated Domains Helix unfolding increases the amount
of self-association, free polar CO, and hydrated peptide CO. Among the peptide units that
Copyright © 2002 Marcel Dekker, Inc.
areunfolded,thoseembeddedinhydrophobicregionsprobablyareresponsibleforthe
increaseinproteinself-association;theothersinpolarenvironmentsshouldbecomehy-
drated.Onaverage,adsorptiononmontmorilloniteentailsalargedegreeofproteinself-
association(Fig.6).Thelevelofproteinself-associationonmontmorilloniteisevenmore
importantatlowp
2
H.
e.EntropyofConformationChangesThelossoforderedsecondarystructures,
suchasinternalandexternalαhelices,iscompensatedbytheincreaseofunorderedstruc-
turessuchashydratedandself-associateddomains.Aspreviouslyemphasized,thisrepre-
sentsanimportantcontributionoftheentropyofconformationtotheadsorptionprocess,
sincealossofapproximatively20%oforderedsecondarystructuresnearthei.e.p.repre-
sentsforBSAanentropiccontributionofϪ400kJmol
Ϫ1
tothedecreaseofGibbsenergy
accompanyingadsorption.Thus,BSAcanbeconsideredasasoftprotein,asthestudyof
thevariationoftheinterfacialareaonadsorptionbyNMRspectroscopyalreadyhasshown.

3.OrientationChangesonAdsorptionoftheHardProtein,Bovine
Pancreaticα-Chymotrypsin
a.EnzymeActivityTheadsorptionofchymotrypsinonmontmorilloniteresults
inacompleteinhibitionbelowpH7,aprogressiverecoveryoftheactivityfrompH7to
pH9,andanactivityquitesimilartothatobservedinsolutionabovepH9(Fig.7)(42).
b.PreservationoftheSecondaryStructureonAdsorptionFigure8showsthat
the adsorption of α-chymotrypsin on montmorillonite has only a very small effect on the
secondary structure of this protein (32). Only 10 to 20 peptide units in peripheral β sheets
are lost on adsorption below pH 7, and this value decreases above this pH. The contrast
with the low structural stability of the secondary structure of BSA is thus well marked.
These weak structural changes should not perturb the structure of α-chymotrypsin at the
level of the enzymatic site. The transmission-FTIR analysis of solutions also confirms
that the optimal catalytic activity near pH 8 results from the convergence of several param-
eters: (1) the deprotonation of the carboxylic side chains; (2) the deprotonation of two
His side chains, increasing both protein flexibility and hydration; and (3) a local β sheet
folding that results from the formation of a salt bridge between the Ile
ϩ
-16 (isoleucine)
end chain aminium group and the Asp-194 side chain carboxylate. For pH Ͼ 10, the
Figure 7 Catalytic activity of α-chymotrypsin in the presence of montmorillonite. (a) experimen-
tal data; (b) data normalized with respect to the maximal value. (Adapted from Ref. 42.)
Copyright © 2002 Marcel Dekker, Inc.
Figure8Secondarystructureofα-chymotrypsininsolutionoradsorbedonmontmorillonite.
AbbreviationsasinFigure6.(AdaptedfromRef.32.)
complete inhibition of the catalytic activity should result not only from peripheral second-
ary structure unfolding caused by external Lys or Tyr (tyrosine) deprotonations, but also
from internal Tyr deprotonations entailing excessive internal hydration in the vicinity of
the catalytic center.
c. Orientation of the Catalytic Site of the Enzyme The pH dependence of the
adsorbed enzyme catalytic activity shows that electrostatic interactions are involved. Nev-

ertheless, the dynamic structural transition of the protein, resulting from both flexibility
and hydration that are slightly enhanced with respect to solution phase, cannot explain
the inactivation of the enzyme in the 5–9 pH range by these weak structural changes alone.
If the tertiary structure of α-chymotrypsin, as determined by X-ray diffraction studies, is
taken as relatively invariant on adsorption and if the time dependence of the Amide II
intensity is analyzed for varying pH, information on the pH dependence of the α-chymo-
trypsin orientation on the montmorillonite surface can be obtained (32). The kinetics of
the NH/N
2
H exchange measured by the Amide II intensity indicates which class of amino
Copyright © 2002 Marcel Dekker, Inc.
acids are protected from water contact. The analysis of the results shows that most of the
inhibition would aries from a steric hindrance by the clay of the substrate access to the
α-chymotrypsin catalytic site. This is due to an interaction involving positively charged
His
ϩ
-40 and His
ϩ
-57 imidazole and Ala
ϩ
-149 (alanine) end chain aminium that control
the initial specific recognition of the substrate by the enzyme. At pH higher than 8.5, when
His-40, His-57, and Ala-149 are deprotonated, the enzyme is adsorbed with a different
orientation, which allows a recovery of activity, similar to that measured in solution in
the same pH range, since the catalytic site now is exposed to the solvent (Fig. 9). Even
if adsorption on montmorillonite implies weak effects on the structure of α-chymotrypsin,
which can be considered as a hard protein, whereas major ones are observed for BSA, a
soft protein, a pH-dependent orientation effect nevertheless can affect the catalytic activity
of α-chymotrypsin.
Figure 9 Schematic representation of an orientation for α-chymotrypsin adsorbed on montmoril-

lonite below pH 7 (A) and above pH 8 (B). See text. (Adapted from Ref. 32.)
Copyright © 2002 Marcel Dekker, Inc.
IV. EFFECT ON CATALYTIC ACTIVITY OF ENZYME ADSORPTION
ON MINERAL SURFACES
A. Models of Interaction
Four different mechanisms currently are invoked to explain the modification of properties
with pH between the free and adsorbed enzymes; they are based either on the activity of
the protons or the substrates in the microenvironment of the enzyme (interfacial pH effects,
diffusional effects) or on the state of the adsorbed enzyme itself (orientation effects, con-
formational effects).
1. Interfacial pH Effects
Although increasingly contested, this hypothesis is the most frequently invoked to explain
the pH shift in the optimal catalytic activity when an enzyme is adsorbed on an electrically
charged surface. It originates from the double diffuse layer theory and assumes that there
is a difference in pH between the bulk of the solution (pH
b
) and the liquid layer near the
solid surface (pH
2
) (6,44–46) due to the increased concentration of protons on electronega-
tive surfaces such as clays. Thus, the active site of an adsorbed enzyme on a clay surface
could be at a pH lower than that measured in the bulk of the solution, and the apparent
pH
b
for maximal activity of a bound enzyme could appear to be higher than that of a free
enzyme.
2. Diffusion-Limited Reactions
The apparent enzyme activity could decrease as a result of limitations in the diffusion of
substrates toward adsorbed enzymes (47–50). The concentration of substrate can be lower
near the surface than in the bulk of the solution as the result of its consumption by the

adsorbed enzyme. A concentration gradient is thus established. If diffusion is slow with
respect to substrate consumption, a steady state is eventually established such that the rate
of diffusion of the substrate in the unstirred layer equals the consumption rate of the
substrate. It should be noted that as the rate of consumption of the substrate is the only
factor that depends on the enzyme in this process, this phenomenon should not cause a
shift in the pH of the optimal catalytic activity. The enzyme activity versus pH follows
a bell-shaped curve, and thus, there are always two pH values, below and above the optimal
pH for activity, where the enzyme activity is similar. Consequently, the decrease in cata-
lytic activity, as a result of a limitation in diffusion of the substrate, should be the same.
The expected effect should be symmetrical with respect to the optimal pH, and with no
pH shift.
3. Orientation Effects
Adsorption of an enzyme with its active site facing the mineral surface obviously limits
the access of the substrate to this site. Such a case has been described for the interaction
of α-chymotrypsin with montmorillonite (32).
4. pH-Dependent Modifications of Conformation
In contrast to the three preceding models, which assume that the enzymes retain the same
conformation in the adsorbed state and in solution, another model is based on a pH-depen-
dent unfolding of the enzyme on the surface. The mechanism could be analogous to the
modification of conformation observed on BSA adsorbed on montmorillonite (43).
Copyright © 2002 Marcel Dekker, Inc.
B.ExperimentalStudyoftheInteractionof␤-
D
-Glucosidaseswith
MineralSurfaces
1.ExperimentalApproach
Theinteractionbetweentheβ-d-glucosidasesandthemineralsurfacesatdifferentpH
valueswasstudiedbythreeprocedures(Fig.10).ProcedureAisameasurementofthe
enzymeactivityintheabsenceofadsorbentsurfaces.ProcedureBisameasurementof
theenzymeactivityinthepresenceoftheadsorbentsurface.ProcedureCisdesignedto

determinethecatalyticactivityofthenonadsorbedfractioninprocedureB.Inotherwords,
itisameasurementoftheenzymeactivityinthesupernatantsolutionaftercentrifugation
oftheenzyme-adsorbentsurfacesuspension.IfA,B,andCarethevaluesofthecatalytic
activitymeasuredbytherespectiveprocedures,twoparameterswellsuitedtoaphysico-
chemicalanalysisoftheinteractionofenzymeswithsolidsurfacesmaybecalculated
(Fig.11).ThefirstisF,theproportionofnonadsorbedenzyme,whichisinverselyrelated
to the affinity for the surface:
F ϭ C/A
The second is the relative activity, R, a structure-related parameter that is defined by the
ratio of the catalytic activity resulting from the enzyme fraction adsorbed (B Ϫ C) and
that of an equal quantity of enzyme in solution (A Ϫ C):
R ϭ (B Ϫ C)/(A Ϫ C)
2. Interactions with Montmorillonite
The β-d-glucosidase of Aspergillus niger (51) has an i.e.p. of 4.0. At a pH above the
i.e.p., the nonadsorbed fraction, F, increases progressively from pH 4 until at pH 6 no
enzyme is adsorbed (Fig. 11). This shows the electrostatic repulsions between the net
Figure 10 Effect of pH on the activity of Aspergillus niger β-d-glucosidase in solution (A); in
the presence of montmorillonite (B), where the activities of both the free and the bound enzyme
are measured; and in the supernatant (C), where the bound fraction has been eliminated by centrifu-
gation. (Adapted from Ref. 51.)
Copyright © 2002 Marcel Dekker, Inc.
Figure11EffectofpHontherelativecatalyticactivityRintheadsorbedstateandontherelative
quantityFinthenonadsorbedstateoftwoβ-d-glucosidasesfromAspergillusnigerandsweetal-
mond.(AdaptedfromRef.51.)
negativelychargedenzymeandtheelectronegativeclaysurface.Incontrast,whenthepH
decreasesbelowthei.e.p.,therelativeactivity,R,oftheenzymedecreases.Itcanbe
theresultofaprogressiveunfoldingcausedbytheelectrostaticattractionsbetweenthe
electronegativesurfaceandthepositivelychargedenzymeinthisrangeofpH,asshown
withBSAbytheNMRandFTIRapproach(16,43).Alternatively,itcanbetheresultof
achangeoforientationoftheenzymewithpH,asshownwithα-chymotrypsinbythe

comparisonofcatalyticactivityandFTIRdata(32,42).Although,inouropinion,the
modificationofconformationhypothesishastheadvantagethatitalwayscanexplainan
alkalineshiftoftheoptimalcatalyticactivityofenzymesadsorbedonelectronegative
surface,itisnotreallypossibletochoosebetweenthesetwohypothesesinthiscase.
3.IrreversibleEffectsonAdsorbedEnzymeActivity
Thesweetalmondβ-d-glucosidaseadsorptiononmontmorillonitepresentscharacteristics
thataremoreinfavorofamodificationofconformation(52).Figure11showsresults
qualitativelysimilartotheβ-d-glucosidaseofA.niger.Buttheresultsofanotherexperi-
mentcanbeexplainedonlybystructuralchangesofthesweetalmondβ-d-glucosidase.
Indeed,irreversibleeffectsinfluencetheactivityofthisenzymewhenadsorbedatapH
thatisdifferentfromasubsequentpHofcatalyticreaction(52).Thesweetalmondβ-d-
glucosidasefirstwasincubatedwithmontmorilloniteatthepHofadsorptionfor2h
(pH
ads
).After2h,thepHwaschanged(pH
react
),andthecatalyticactivitywasmeasured
(Fig.12).ForagivenpH
react
, the catalytic activity decreased when the pH
ads
was decreased.
At a pH
ads
of 3.6, no catalytic activity is detected over the entire range of pH
react
, despite
complete stability of the enzyme in solution in this pH range. This behavior suggests two
remarks. First, an irreversible effect would not be expected if surface pH was involved,
as the 2 h of equilibration should be sufficient to reach a new repartition of the protons

in the double diffuse layer. Thus, no lasting effect of pH
ads
should have been observed.
Second, irreversible effects are common with polymers adsorbed on surfaces and are ex-
plained by irreversible changes of conformation (8,10,53). If the enzyme is unfolded, the
number of points of contact with the surface increases and the energy necessary for
Copyright © 2002 Marcel Dekker, Inc.
Figure12EffectofthepHofadsorptiononmontmorillonite(numbersoncurves)onthesubse-
quentpHprofileofactivityofsweetalmondβ-d-glucosidasewhenthepHisadjustedafterthe
adsorptionstagetogivethepHofthecatalyticreactiononthexaxis.(AdaptedfromRef.52.)
thereversaloftheunfoldingoftheadsorbedenzymethereforewouldbegreaterthanthe
thermalenergy,kT,availabletothesystem.
4.InadequacyoftheInterfacialpHHypothesis
TheirreversibleeffectofpHchangesonadsorbedenzymeactivityisnottheonlyfact
thatmitigatesagainsttheinterfacialpHinterpretationofthepHshiftofthecatalyticactiv-
ity.AnimportantconsequenceofalocalpHmechanismisthatthecatalyticactivityof
theadsorbedenzymeshouldbehigherthanthatoftheenzymeinsolutioninthealkaline
pHrange.Theresultsobtainedforα-chymotrypsin(42)showthatitisnotthecasewhen
absolutevaluesareconsidered(Fig.7a),contrarytoconclusionsdrawnconsideringvalues
normalizedwithrespecttotheirmaximum(Fig.7b).Inaddition,evenanincreaseinionic
strengthincreasecannoteliminatethepHshifteffect(54).Finally,itshouldbeenpointed
outthatalthoughtheconcentrationofprotonsisdifferentnearanelectricallycharged
surfaceandinthebulkofthesolution,theirelectrochemicalpotentialsmust,nevertheless,
beidentical,sincetheyareinthermodynamicequilibrium.Becausetheelectrochemical
potentialdeterminestheworkavailableforareaction,higherprotonactivitynearasurface
doesnotmeanhigherreactivityoftheseprotons.
5.ComparisonoftheEffectofDifferentMineralSurfaces
Althoughelectrostaticinteractionsbetweensurfacesandenzymesexplainmostofthe
observedconformationalchanges,othertypesofinteractionsometimesneedtobetaken
intoaccount.Figure13comparestheeffectsoftalc,goethite,andmontmorilloniteonthe

relativecatalyticactivityofsweetalmondβ-d-glucosidase.Thehydrophilicsurfaceof
goethite,whichhadalowelectricalchargeasaresultofcomplexationofoxyanionsunder
theconditionsoftheexperiment(52),hasthelowerdestabilizingeffect.Thehydrophobic
talc,whichalsohasnoelectricalchargeonitsbasalsurface,hasamorepronounced
destabilizingeffect,confirmingtheinterventionofhydrophobicinteractions.
6.InteractionwithOrganomineralSurfaces
Organiccoatingsonclayscanhaveaprotectiveeffectontheadsorbedenzymeactivity(54).
Figure14showstheeffectofdifferentnaturalandartificialclay–polymercomplexesonthe
Copyright © 2002 Marcel Dekker, Inc.
Figure 13 Effect of different mineral surfaces on the relative activity R of adsorbed sweet almond
β-d-glucosidase. (Adapted from Ref. 52.)
relative catalytic activity of sweet almond β-d-glucosidase. A soil fraction rich in organic
matter has a pronounced protective effect, at least above pH 4. A polyethylene glycol–montm-
orilonite complex has a similar effect. In this case, it could be explained by a protective effect
of this hydrophilic polymer. The lack of protective effect below pH 4 could be due to an
exchange mechanism between the positively charged enzyme at this pH and the neutral poly-
mer. This mechanism is supported by the absence of inhibition of the β-d-glucosidase at pH
4 by a lysozyme–montmorillonite complex. In this case, it is the strong interaction of the
lysozyme with the clay surface, the high i.e.p. of the protein, and its positive charge over the
entire pH range studied that prevents exchange with the β-d-glucosidase.
Figure 14 Effect of different organic coatings on mineral surfaces on the relative activity R of
adsorbed sweet almond β-d-glucosidase. (Adapted from Ref. 54.)
Copyright © 2002 Marcel Dekker, Inc.
IV. CONCLUSIONS
The adsorption of enzymes on mineral surfaces is a complex phenomenon that involves
both enthalpic and entropic effects. An important and difficult challenge is the determina-
tion of possible changes in conformation of the adsorbed enzyme. NMR anf FTIR spectro-
scopies are useful tools to answer this question since they, respectively, give information
on the interfacial area of the surface in contact with the protein and on the secondary
structure of adsorbed proteins. It has been shown that both pH-dependent modification of

conformation and pH-dependent orientation of the catalytic site of the enzyme can explain
the alkaline pH shift of the enzyme activity on electronegative soil mineral surfaces. Soft
proteins, such as BSA, are more prone to the first mechanism, whereas hard proteins, such
as α-chymotrypsin, are more prone to the second. On the other hand, it has been shown,
from experimental observations as well as for theoretical reasons, that a surface pH effect
cannot explain adequately the shift in the pH of the optimal activity when an enzyme is
adsorbed on clays. In addition to electrostatic forces, hydrophobic interactions are implied
in the interaction of proteins with clays. An important aspect is the interplay of different
driving forces in adsorption. For example, the hydrophobic interactions with clays can
result from an electrostatic exchange of the hydrophilic counterions on the clay surface
that leaves a hydrophobic siloxane surface. The rearrangement of the protein structure on
the clay surface subsequently can be facilitated when hydrophobic amino acids come in
contact with the hydrophobic siloxane layer and remain shielded from the water molecules
of the solution. If this rearrangement is accompanied by a decrease of ordered secondary
structures, it results in an additional increase in conformational entropy, lowering the
Gibbs energy of the system. The combination of all these different subprocesses is respon-
sible for the irreversible aspects of the modifications of conformation of enzymes on clay
surfaces.
REFERENCES
1. AD McLaren. Concerning the pH dependence of enzyme reactions on cells, particulates and
in solution. Science 125:697, 1957.
2. P Monsan, G Durand. Pre
´
paration d’invertase insolubilise
´
e par fixation sur bentonite. FEBS
Lett 16:39–42, 1971.
3. J Skujins
ˇ
, A Pukite, AD McLaren. Adsorption and activity of chitinase on kaolinite. Soil Biol

Biochem 6:179–182, 1974.
4. P Douzou, P Maurel. Ionic control of biochemical reactions. Trends Biochem Sci 2:14–17, 1977.
5. AD McLaren. The adsorption and reactions of enzymes and proteins on kaolinite. J Phys Chem
58:129–137, 1954.
6. AD McLaren, EF Estermann. Influence of pH on the activity of chymotrypsin at a solid-liquid
interface. Arch Biochem Biophys 68:157–160, 1957.
7. AD McLaren, GH Peterson, I Barshad. The adsorption and reactions of enzymes and proteins
on clay minerals. IV. Kaolinite and montmorillonite. Soil Sci Soc Am Proc 22:239–244, 1958.
8. W Norde. Adsorption of proteins from solution at the solid-liquid interface. Adv Colloid Inter-
face Sci 25:267–340, 1986.
9. W Norde, J Lyklema. Why proteins prefersurfaces. J Biomater Sci PolymerEdn 2:183–202, 1991.
10. CA Haynes, W Norde. Globular proteins at solid/liquid interfaces. Colloids Surfaces B Bioint-
erfaces 2:517–566, 1994.
11. W Norde, J Lyklema. Thermodynamics of protein adsorption: Theory with special reference
Copyright © 2002 Marcel Dekker, Inc.
to the adsorption of human plasma albumin and bovine pancreas ribonuclase at polystyrene
surfaces. J Colloid Interface Sci 71:350–366, 1979.
12. J Lyklema. Proteins at solid-liquid interfaces: A colloid-chemical review. Colloids Surfaces
10:33–42, 1984.
13. W Norde. Energy and entropy of protein adsorption. J Dispersion Sci Technol 13:363–377,
1992.
14. S Staunton, H Quiquampoix. Adsorption and conformation of bovine serum albumin on mont-
morillonite: Modification of the balance between hydrophobic and electrostatic interactions
by protein methylation and pH variation. J Colloid Interface Sci 166:89–94, 1994.
15. P Chassin, C Jouany, H Quiquampoix. Measurement of the surface free energy of calcium-
montmorillonite. Clay Miner 21:899–907, 1986.
16. H Quiquampoix, RG Ratcliffe. A
31
P NMR study of the adsorption of bovine serum albumin
on montmorillonite using phosphate and the paramagnetic cation Mn

2ϩ
: modification of con-
formation with pH. J Colloid Interface Sci 148:343–352, 1992.
17. JT Albert, RD Harter. Adsorption of lysozyme and ovalbumin by clay: Effect of clay suspen-
sion pH and clay mineral type. Soil Sci 115:130–136, 1973.
18. H Quiquampoix, G Bacic, BC Loughman, RG Ratcliffe. Quantitative aspects of the
31
P-NMR
detection of manganese in plant tissues. J Exp Bot 44:1809–1818, 1993.
19. H Quiquampoix, BC Loughman, RG Ratcliffe. A
31
P-NMR study of the uptake and compart-
mentation of manganese by maize roots. J Exp Bot 44:1819–1827, 1993.
20. TJ Su, JR Lu, BK Thomas, ZF Cui, J Penfold. The effect of solution pH on the structure of
lysozyme layers adsorbed at the silica-water interface studied by neutron reflection. Langmuir
14:438–445, 1998.
21. S Duinhoven, R Poort, G Van der Voet, WGM Agterof, W Norde, J Lyklema. Driving forces
for enzyme adsorption at solid-liquid interfaces. 1. The serine protease savinase. J Colloid
Interface Sci 170:340–350, 1995.
22. S Duinhoven, R Poort, G Van der Voet, WGM Agterof, W Norde, J Lyklema. Driving forces
for enzyme adsorption at solid-liquid interfaces. 2. The fungal lipase lipolase. J Colloid Inter-
face Sci 170:351–357, 1995.
23. XM He, DC Carter. Atomic structure and chemistry of human serum albumin. Nature 358:
209–214, 1992.
24. DC Carter, JX Ho. Structure of serum albumin. Adv Protein Chem 45:155–203, 1994.
25. A Kondo, K Higashitani. Adsorption of model proteins with wide variation in molecular prop-
erties on colloidal particles. J Colloid Interface Sci 150:344–351, 1992.
26. A Kondo, F Murakami, K Higashitani. Circular dichroism studies on conformational changes
in protein molecules upon adsorption on ultrafine polystyrene particles. Biotechnol Bioeng
40:889–894, 1992.

27. A Kondo, F Murakami, M Kawagoe, K Higashitani. Kinetic and circular dichroism of enzymes
adsorbed on ultrafine silica particles. Appl Microbiol Biotechnol 39:726–731, 1993.
28. L Boulkanz, N Balcar, MH Baron. FTIR analysis for structural characterisation of albumin
adsorbed on reversed phase support. Appl Spectrosc 49:1737–1746, 1995.
29. B De Collongue, B Sebille, MH Baron. Chromatography of the Interferon γ and the Analogue
II : FTIR Analysis. Biospectros 2:101–111, 1996.
30. L Boulkanz, C Vidal-Madjar, N Balcar, MH Baron. Adsorption mechanism of human serum
albumin on a reversed-phase support by kinetic, chromatographic and FTIR methods. J Colloid
Interface Sci 188:58–67, 1997.
31. A Pantazaki, MH Baron, M Revault, C Vidal-Madjar. Characterisation of human serum albu-
min adsorbed on a porous anion-exchange support. J Colloid Interface Sci 207:324–331, 1998.
32. MH Baron, M Revault, S Servagent-Noinville, J Abadie, H Quiquampoix. Chymotrypsin ad-
sorption on montmorillonite: Enzymatic activity and kinetics FTIR structural analysis. J Col-
loid Interface Sci 214:319–332, 1999.
33. JLR Arrondo, A Muga, J Castresana, FM Goni. Quantitative studies of the structure of proteins
Copyright © 2002 Marcel Dekker, Inc.
in solution by Fourier transform infrared spectroscopy. Prog Biophys Mol Biol 59:23–56,
1993.
34. DM Byler, H Susi. Application of computerized infrared and Raman spectroscopy to confor-
mation studies of casein and other food proteins. J Ind Microbiol 3:73–88, 1988.
35. D Byler, H Susi. Examination of the secondary structure of proteins by deconvolved FTIR
spectra. Biopolymers 25:469–487, 1986.
36. WK Surewicz, HH Mantsch. New insight into protein secondary structure from resolution-
enhanced infrared spectra. Biochim Biophys Acta 952:115–130, 1988.
37. F Fillaux, C de Loze. Spectroscopic study of monosubstituted amides. II. Rotation isomers
in amides substituted by aliphatic side-chain models. Biopolymers 11:2063–2077, 1972.
38. C de Loze, MH Baron, F Fillaux. Interactions of the CONH group in solution. Interpretation
of the infrared and Raman spectra in relationship to secondary structures of peptides and pro-
teins. J Chimie Physique 75:631–647, 1978.
39. MH Baron, C de Loze, C Toniolo, GD Fasman. Structure in solution of protected homo-

oligopeptides of L-valine, L-isoleucine and L-phenylalanine: An infrared absorption study.
Biopolymers 17:2225–2239, 1978.
40. M Hollo
`
si, ZS Majer, AZ Ro
`
nai, A Magyar, K Medzihradszky, S Holly, A Perczel, GD Fas-
man. CD and FTIR spectroscopic studies of peptides. II. Detection of βturns in linear peptides.
Biopolymers 34:117–185, 1994.
41. RB Gregory, A Rosenberg. Protein conformational dynamics measured by hydrogen isotope
exchange techniques. Methods Enzymol 131:448–511, 1986.
42. H Quiquampoix, J Abadie, MH Baron, F Leprince, PT Matumoto- Pintro, RG Ratcliffe, S
Staunton. Mechanisms and consequences of protein adsorption on soil mineral surfaces. In:
TA Horbett, JL Brash, eds. Proteins at Interfaces II, ACS Symposium Series 602. Washington,
DC: American Chemical Society, 1995, pp 321–333.
43. S Servagent-Noinville, M Revault, H Quiquampoix, MH Baron. Conformational changes of
bovine serum albumin induced by adsorption on different clay surfaces: FTIR analysis. J Col-
loid Interface Sci 221: 273–283, 2000.
44. G Durand. Modification de l’activite
´
de l’ure
´
ase en pre
´
sence de bentonite. C R Acad Sci 259:
3397–3400, 1964.
45. RA Aliev, VS Gusev, DG Zvyagintsev. Influence of adsorbents on the optimum pH of catalase.
Vestn Moskov Univ Serie 6 Biol Pochvoved 31:67–70, 1976.
46. P Douzou, GA Petsko. Proteins at work: ‘‘Stop-action’’ pictures at subzero temperatures. Adv
Protein Chem 36:245–361, 1984.

47. J Konecny, W Voser. Effects of carrier morphology and buffer diffusion on the expression
of enzymic activity. Biochim Biophys Acta 485:367–378, 1977.
48. M Arrio-Dupont, JJ Be
´
chet. Diffusion limited kinetics of immobilised myosin ATPase. Biochi-
mie 71:833–838, 1989.
49. V Bille, D Plainchamp, S Lavielle, G Chassaing, J Remacles. Effect of the microenvironment
on the kinetic properties of immobilized enzymes. Eur J Biochem 180:41–47, 1989.
50. CA Ku, BB Lentrichia. Effects of diffusion limitation on binding of soluble ligand to avidin-
coupled latex particles. J Colloid Interface Sci 132:578–584, 1989.
51. H Quiquampoix, P Chassin, RG Ratcliffe. Enzyme activity and cation exchange as tools for
the study of the conformation of proteins adsorbed on mineral surfaces. Prog Colloid Polym
Sci 79:59–63, 1989.
52. H Quiquampoix. A stepwise approach to the understanding of extracellular enzyme activity
in soil. I. Effect of electrostatic interactions on the conformation of a β-D-glucosidase on
different mineral surfaces. Biochimie 69:753–763, 1987.
53. J Israelachvili. Intermolecular and Surface Forces. 2nd ed. London: Academic Press, 1991.
54. H Quiquampoix. A stepwise approach to the understanding of extracellular enzyme activity
in soil. II. Competitive effects on the adsorption of a β-D-glucosidase in mixed mineral or
organo-mineral systems. Biochimie 69:765–771, 1987.
Copyright © 2002 Marcel Dekker, Inc.