Advertisement

Reduction of Irreversible Protein Adsorption on Solid Surfaces by ProteinEngineering for IncreasedStability*

  • Martin Karlsson
    Affiliations
    IFM-Department of Chemistry,Linköping University, SE-581 83 Linköping, Sweden and
    Search for articles by this author
  • Author Footnotes
    § Present address: Eka Chemicals, Separation Products, SE-445 80 Bohus,Sweden.
    Johan Ekeroth
    Footnotes
    § Present address: Eka Chemicals, Separation Products, SE-445 80 Bohus,Sweden.
    Affiliations
    IFM-Department of Chemistry,Linköping University, SE-581 83 Linköping, Sweden and
    Search for articles by this author
  • Hans Elwing
    Affiliations
    Department of Cell and Molecular Biology,Göteborg University, Box 462, SE-40530, Sweden
    Search for articles by this author
  • Uno Carlsson
    Correspondence
    To whom correspondence should be addressed. Tel.: 46-13-281714; Fax:46-13-281399;
    Affiliations
    IFM-Department of Chemistry,Linköping University, SE-581 83 Linköping, Sweden and
    Search for articles by this author
  • Author Footnotes
    § Present address: Eka Chemicals, Separation Products, SE-445 80 Bohus,Sweden.
    * This work was supported by the Swedish Research Council (U. C.) andStiftelsen Lars Hiertas Minne (M. K.). The costs of publication of thisarticle were defrayed in part by the payment of page charges. This articlemust therefore be hereby marked “advertisement” inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Open AccessPublished:April 26, 2005DOI:https://doi.org/10.1074/jbc.M503665200
      The influence of protein stability on the adsorption and desorptionbehavior to surfaces with fundamentally different properties (negativelycharged, positively charged, hydrophilic, and hydrophobic) was examined bysurface plasmon resonance measurements. Three engineered variants of humancarbonic anhydrase II were used that have unchanged surface properties butlarge differences in stability. The orientation and conformational state ofthe adsorbed protein could be elucidated by taking all of the followingproperties of the protein variants into account: stability, unfolding,adsorption, and desorption behavior. Regardless of the nature of the surface,there were correlation between (i) the protein stability and kinetics ofadsorption, with an increased amplitude of the first kinetic phase ofadsorption with increasing stability; (ii) the protein stability and theextent of maximally adsorbed protein to the actual surface, with an increasedamount of adsorbed protein with increasing stability; (iii) the proteinstability and the amount of protein desorbed upon washing with buffer, with anincreased elutability of the adsorbed protein with increased stability. All ofthe above correlations could be explained by the rate of denaturation and theconformational state of the adsorbed protein. In conclusion, proteinengineering for increased stability can be used as a strategy to decreaseirreversible adsorption on surfaces at a liquid-solid interface.
      The adsorption of proteins at a liquid-solid surface interface has been ofincreasing interest because of its implications in medicine, biotechnology,and the food industry (
      • Hlady V.
      • Buijs J.
      ,
      • Nakanishi K.
      • Sakiyama T.
      • Imamura K.
      ). There are instances whenprotein adsorption is an unwanted effect, such as the irreversible proteinadsorption that leads to biofouling of implantable biosensors and inbiotechnological processes. There are also examples in which proteinadsorption is a desired feature, as long as this can be done in a controlledmanner, so that the structural and functional integrity of the protein ismaintained. Examples of such areas are the production of combined, adsorbedvaccines (
      • Matheis W.
      • Zott A.
      • Schwanig M.
      ) and thedevelopment of chromatography material(
      • McNay J.
      • Fernandez E.
      ,
      • McNay J.
      • O'Connell J.
      • Fernandez E.
      ).
      The areas in which protein adsorption has a large impact can roughly bedivided into two parts. The first area is when a specific surface is exposedto a large number of proteins, which includes bioimplantable materials andequipment for crude separation of proteins. In this case, the surface has tobe engineered to avoid or govern the adsorption process(
      • Wisniewski N.
      • Reichert M.
      ,
      • Chapman R.G.
      • Ostuni E.
      • Takayama S.
      • Holmlin R.E.
      • Yan L.
      • Whitesides G.M.
      ,
      • Ostuni E.
      • Chapman R.G.
      • Holmlin R.E.
      • Takayama S.
      • Whitesides G.M.
      ).The second area is when a specific protein comes into contact with a largenumber of surfaces, e.g. in the downstream processing, polishing,storage, and final use of proteins produced for biotechnological or medicalapplications. In this case, it is unlikely that all of the surfaces (glass,plastic, stainless steel, rubber, and so forth) that a specific protein isexposed to during production and use can be made resistant to proteinadsorption, i.e. it could be more feasible to engineer the protein tomake it less prone to adsorb onto surfaces. Although this case has not beengiven as much attention, it is interesting to note that the downstreamprocessing of therapeutic antibodies accounts for an estimated 50–80% ofthe total manufacturing cost(
      • Roque C.
      • Lowe C.
      • Taipa A.
      ). Reduction of irreversibleprotein adsorption is likely to result in less biofouling and higher yields ofprotein and thus reduce the cost of downstream processing. Furthermore, it hasbeen shown that protein adsorption phenomena can lead to a large loss ofactivity of therapeutically formulated protein solutions(
      • McLeod A.G.
      • Walker I.R.
      • Zheng S.
      • Hayward C.P.M.
      ,
      • Tzannis S.
      • Hrushesky W.
      • Wood P.
      • Przybycien T.
      ), i.e. there isboth an economic and a safety incentive to reduce the adsorption of a specificprotein that needs to be produced to meet biotechnological or medicalneeds.
      It has been proposed that the stability of a protein is one of thedeterminants of adsorption behavior and can even be one of the driving forcesfor protein adsorption (
      • Norde W.
      ).The reason for this is that the conformation of folded proteins is quiterestricted, i.e. the entropy is relatively low. However, if theprotein, upon adsorption, tends to unfold to various degrees, this may lead toa conformational entropy gain, which can act as a driving force for adsorptioneven at hydrophilic surfaces under electrostatic repulsion. In terms of thedegree of adsorption, one can thus make a distinction between hard (stable)and soft (less stable) proteins. Therefore, one possible way to reduce proteinadsorption could be to increase the thermal stability of a protein. In arecent study, this has also been shown to be the case, and sugar excipientswere shown to decrease protein adsorption by stabilization of the proteinnative state in solution (
      • Wendorf J.
      • Radke C.
      • Blanch H.
      ).Furthermore, conformational stability of proteins is an important determinantof the structure of the adsorbed protein(
      • Billsten P.
      • Freskgård P-O.
      • Carlsson U.
      • Jonsson B.-H.
      • Elwing H.
      ,
      • Malmsten M.
      ), and earlier workemploying point-mutated protein stability variants has described how thestability of a protein mainly influences with what rate the adsorbed proteinundergoes conformational changes after adsorption(
      • Karlsson M.
      • Mårtensson L-G.
      • Jonsson B.-H.
      • Carlsson U.
      ). Thus, upon adsorption, aless stable protein will adopt various conformational states ranging fromnative, through molten globule, to a fully denatured state, with each stepgiving rise to an increased number of interaction points between the proteinand the surface and between the adsorbed proteins themselves(
      • Moulin A.M.
      • O'Shea S.J.
      • Badley R.A.
      • Doyle P.
      • Welland M.E.
      ). Hence, theirreversibility of protein adsorption will become more pronounced the moredenatured the protein becomes when adsorbed to a surface. One possible way toreduce the irreversible adsorption of a specific protein could thus be toincrease the thermodynamic stability of the protein and effectively make theprotein more resistant to denaturation on the surface. In an earlier study(
      • McGuire J.
      • Wahlgren M.
      • Arnebrant T.
      ), it was concluded thatmore stable variants of the same protein displayed a higher elutability uponelution with a detergent. However, the use of detergents is incompatible withthe storage and use of therapeutical proteins. Instead, in this study, weassess the influence that protein stability, as an intrinsic property of theprotein, has on adsorption and desorption to four surfaces with fundamentallydifferent properties (negatively charged, hydrophilic, hydrophobic, andpositively charged). This will enable us to evaluate (i) whether there is acorrelation between stability and adsorption/desorption behavior and(ii) whether this is independent of the properties of the surface. Wehave made use of three variants of human carbonic anhydrase II that differ inthermodynamic stabilities. Furthermore, the mutation sites for these threestability variants are made on the inside of the protein; other propertiesthat can also influence the adsorption behavior such as size(
      • Ball V.
      • Bentaleb A.
      • Hemmerle J.
      • Voegel J-C.
      • Schaaf P.
      ), surface potentials(
      • Bower C.K.
      • Sananikone S.
      • Bothwell M.K.
      • McGuire J.
      ,
      • Ramsden J.J.
      • Roush D.J.
      • Gill D.S.
      • Kurrat R.
      • Willson R.C.
      ), hydrophobic patches(
      • Malmsten M.
      • Burns N.
      • Veide A.
      ), and secondary structure(
      • Burkett S.R.
      • Read M.J.
      ) are identical in thethree variants. Thus, any differences in adsorption and desorption behaviororiginate in the stability of the protein variants and the differentfunctionalities of the surfaces only.
      The different surfaces were made up of self-assembled monolayers. In orderto obtain well-defined surfaces, the monomers used to build up theself-assembled monolayer surfaces had the same “scaffoldstructure,” onto which different functionalities were added. Thisminimizes any effects that might arise from variation in chain length,conformational flexibility(
      • Carignano M.A.
      • Szleifer I.
      ), monolayer order(
      • Petrash S.
      • Cregger T.
      • Zhao B.
      • Pokidysheva E.
      • Foster M.D.
      • Brittain W.J.
      • Sevastianov V.
      • Majkrzak C.F.
      ), and so forth. We havemonitored the surface adsorption and desorption of the protein variants bysurface plasmon resonance(SPR),
      The abbreviations used are: SPR, surface plasmon resonance; RSA, randomsequential adsorption; MOPS, 4-morpholinepropanesulfonic acid.
      1The abbreviations used are: SPR, surface plasmon resonance; RSA, randomsequential adsorption; MOPS, 4-morpholinepropanesulfonic acid.
      an opticalmethod that can be used to register the protein mass adsorbed.

      MATERIALS AND METHODS

      Surface Preparation: Synthesis of Functionalized Thiols—Thesynthesis of alcohol (
      • Ekeroth J.
      • Borgh A.
      • Konradsson P.
      • Liedberg B.
      )- andsulfate (
      • Ekeroth J.
      • Konradsson P.
      • Höök F.
      )-terminated thiolshas been described previously. The methyl- and amino-terminated thiols wereprepared according to the method for the alcohol-terminated thiol(
      • Ekeroth J.
      • Borgh A.
      • Konradsson P.
      • Liedberg B.
      ), substituting theaminoethanol for ethylamine hydrochloride and N-Boc-diaminoethane,respectively. The Boc protection group on the amine was removed usingtrifluoroacetic acid prior to thioester cleavage.
      Gold Substrates—Gold substrates used for infraredreflection-absorption spectroscopy, ellipsometry, and contact angle goniometrywere prepared as follows: silicon wafers (100) were cut into appropriate sizesand washed in TL2, a mixture of water (Milli-Q), 30% hydrogen peroxide(Merck), and concentrated hydrochloric acid (Merck) (6:1:1), at 80 °C for10 min. The substrates were then mounted in a UHV evaporation system (BalzersUMS 500 P) and primed with a 25 Å layer of titanium followed by 2000Å of gold. The pressure was <2 × 10–8 mbarduring evaporation, and the evaporation rate was 2 Å/s for titanium and10 Å/s for gold.
      Assembly—Prior to assembly, the gold substrates were cleanedin TL1, a mixture of water, 30% hydrogen peroxide, and 25% ammonia (Merck)(5:1:1), at 80 °C for 10 min. The monolayers were prepared from ethanol(Kemetyl, Haninge, Sweden) solutions (thiol concentration, 1 mm) byplacing the cleaned, gold surface in the solution for at least 15 h. Thesubstrates were removed from the solution and gently rinsed in ethanol,followed by a 5-min ultrasonication in ethanol. Prior to use in experiments,the surfaces were dried under a stream of nitrogen.
      Infrared Reflection-Absorption Spectroscopy—Infraredreflection-absorption spectra were recorded on a Bruker IFS66 Fouriertransform spectrometer equipped with a grazing angle of incidence reflectionaccessory aligned at 85°. A liquid nitrogen cooled MCT detector was used.Interferograms were apodized with a three-term Blackmann-Harris functionbefore Fourier transformation. The spectra were recorded by averaging 3000interferograms (10 min) at 2 cm–1 resolution.
      Ellipsometry—Single-wavelength ellipsometry was performed ona Rudolph AutoEL ellipsometer. Light source was a He-Ne laser with wavelength632.8 nm, at an angle of incidence of 70°. The fresh gold plates weremeasured prior to incubation with thiol to obtain reference values of cleangold. As a model, ambient/organic film/gold, assuming an isotropic transparentorganic layer with n = 1.5, was used. The film thickness wascalculated by the AutoEL ellipsometer software as an average of three spots oneach substrate.
      Contact Angle Goniometry—Contact angles were measured with aRamé-Hart NRL 100 goniometer with no control of ambient humidity. Ascontact liquids, freshly deionized water from a Milli-Q unit and hexadecane(Merck) were used, respectively. Two separate measurements were performed oneach plate.
      Protein Variants—For the assessment of the influence ofprotein stability to the adsorption and desorption behavior at surfaces, threestability variants of human carbonic anhydrase II were used. Apseudo-wild-type of HCA II with the mutation C206S (HCA IIpwt) wasused as a template for the S56N (HCA IIpwt with a S56N mutation)and A23C/L203Cox (HCA IIpwt with a A23C and a L203Cmutation in the oxidized state) mutants. HCA IIpwt isindistinguishable from the HCA II wild type with regard to structure,activity, and stability (
      • Mårtensson L-G.
      • Jonsson B.-H.
      • Andersson M.
      • Kihlgren A.
      • Bergenhem N.
      • Carlsson U.
      ,
      • Mårtensson L-G.
      • Jonsson B.-H.
      • Freskgård P.-O.
      • Kihlgren A.
      • Svensson M.
      • Carlsson U.
      ). The production andproperties of the S56N, HCA IIpwt, and A23C/L203Coxvariants have been described elsewhere(
      • Karlsson M.
      • Mårtensson L-G.
      • Jonsson B.-H.
      • Carlsson U.
      ,
      • Mårtensson L-G.
      • Karlsson M.
      • Carlsson U.
      ,
      • Karlsson M.
      • Mårtensson L-G.
      • Olofsson P.
      • Carlsson U.
      ,
      • Karlsson M.
      • Mårtensson L-G.
      • Karlsson C.
      • Carlsson U.
      ).
      Determination of Protein Melting Temperatures(Tm)—Solutions of 0.85 μm protein in 0.1m MOPS, pH 7.5, were prepared and kept on ice. 1.6 ml aliquots weretransferred from the solution to a thermostatted, 1-cm quartz cuvette. Thetime to reach unfolding equilibrium was established by monitoring the changein fluorescence intensity at 336 nm at each temperature, and for eachtemperature a fresh sample was incubated in order to prevent any problems withaggregation during incubation. After the sample had equilibrated, threefluorescence spectra (310–410 nm) of each sample were recorded.Measurements were performed on a Fluoromax-2 spectrofluorometer (Jobin-YvonInstruments). The excitation wavelength was 295 nm, and the excitation andemission slits were 3 and 4 nm, respectively. The fraction of unfolded proteinwas calculated from the fluorescence data and plotted as a function oftemperature. The transition curves were obtained from nonlinear least-squaresanalysis (
      • Santoro M.M.
      • Bolen D.W.
      ), using theprogram TableCurve 2D (Jandel Scientific).
      Adsorption Experiments—All experiments were performed in 10mm sodium borate buffer, pH 8.5, prepared with ultra-pure water(Milli-Q-plus system; Millipore, Bedford, MA). All the buffers wereultrasonically degassed under vacuum before use. In all experiments, theprotein concentration was 50 μg/ml. SPR surfaces were cleaned prior tosurface modifications in a UV/ozone chamber for 10–20 min. This wasfollowed by immersion of the surfaces in a 1:1:6 mixture ofH2O2 (30%) (J. T. Baker), NH3 (25%) (Merck),and Milli-Q water (Millipore) for 10 min at 65 °C. The surfaces weretransferred to a beaker of 95% ethanol and cleaned from residual contaminantsby ultrasonication for 1 min. The surfaces were then rinsed with ethanolbefore transfer to the respective thiol solutions. The surfaces were immersedin the respective thiol solutions overnight. SPR measurements were performedon a Biacore 2000 system (Biacore AB, Uppsala, Sweden) providing a laminarflow through the measuring chamber. The flow rate was 25 μl/min, and thetemperature was kept constant at 22 °C. Gold substrates were gold-coveredglass plates supplied by Biacore AB treated as described under“Assembly” before mounting in the Biacore-chip holder. The proteinwas introduced to the surface by passing a solution of the protein (50μg/ml) in buffer over the surface for 13 min followed by automatic rinsingwith buffer. The total measuring time was at least 43 min for duplicate ortriplicate runs. Data collection was done using the Biacore 2000 software. Forestimation of the surface-coupled mass to the flat surface,ΔmSPR (ng × cm–2), we usedthe relationship ΔmSPR =CSPRΔRU, where ΔRU is the difference in thedimensionless SPR response units, and the constant CSPR is6.5 × 10–2 ng × cm–2(
      • Höök F.
      • Kasemo B.
      • Nylander T.
      • Fant C.
      • Sott K.
      • Elwing H.
      ).
      Treatment of Data for Evaluation of Adsorption Kinetics—Allaverage adsorption isotherms were imported to the program TableCurve 2D(Jandel Scientific) for kinetic evaluation. Three exponential terms wereneeded to make a good fit to the adsorption isotherms with, in some cases, anadditional linear term.
      Theoretically Adsorbed Mass—The experimentally determinedmasses adsorbed were compared with theoretical values in an attempt to gainadditional information concerning the arrangement of the adsorbed protein. Thevalues from the SPR measurements were compared with theoretical valuescalculated according to (i) a model assuming a close packed monolayer and (ii)a model based on random sequential adsorption (RSA).
      In calculating the theoretical, close-packed monolayer for side-on andend-on adsorption, the area occupied by each protein molecule is assumed tocorrespond to the circular footprint formed by rotating the long axis (55Å for side-on and 42 Å for end-on) of the respective orientation.The RSA model assumes that the proteins approach the surface sequentially, areadsorbed irreversibly, and do not diffuse laterally on the surface. Thetheoretical surface mass of such a model is about 54.7% of a close-packedmonolayer of spheres (
      • Välimaa L.
      • Petterson K.
      • Rosenberg J.
      • Karp M.
      • Lövgren T.
      ).Furthermore, the same calculations have been employed to calculate thetheoretical monolayer coverage for the HCA II protein in its molten globulestate by expanding each radius by 10%(
      • Uversky V.N.
      ).

      RESULTS AND DISCUSSION

      Protein Variants—HCA II is a monomeric enzyme, with amolecular mass of 29.3 kDa, consisting of 259 amino acids. The protein isroughly ellipsoidal, with the dimensions 55 × 42 × 39 Å, andis largely composed of β-sheets. The protein has a pI of 7.3(
      • Jonsson M.
      • Pettersson E.
      ) and is expected to have aweak negative net charge at the pH chosen for the adsorption experiments (pH8.5). The main features of the three protein variants used are summarized inTable I, where it can be notedthat all the mutations are situated inside the protein at locations that haveeither no or very low fractional surface accessibility, i.e. thesurface properties of all the variants should be identical. This ensures thatwe have basically identical proteins that differ only in stability. The S56Nvariant has a lower thermodynamic stability, as compared with the HCAIIpwt, whereas the oxidized variant of A23C/L203C(A23C/L203Cox) has a higher thermodynamic stability. HCA II and allthe variants that we have previously produced denature with a molten globuleintermediate that is stable at equilibrium. Such intermediates arecharacterized by a loss of tertiary structure, an increase in diameter(∼10%) (
      • Uversky V.N.
      ), and exposureof hydrophobic patches (
      • Ptitsyn O.B.
      • Bychkova V.E.
      • Uversky V.N.
      ).Interestingly, the stabilized variant A23C/L203Cox has such anincreased stability that its first transition from the native to the moltenglobule state almost coincides with the second transition (the molten globuleto the denatured state), giving an apparent two-state transition(Fig. 1). This results in amarkedly reduced tendency to form a molten globule state forA23C/L203Cox. ANS binding studies during denaturation ofA23C/L203Cox in Gu-HCl revealed that only about 10% molten globuleaccumulated in solution, as compared with HCA IIpwt(
      • Mårtensson L-G.
      • Karlsson M.
      • Carlsson U.
      ). In an earlier study(
      • Billsten P.
      • Freskgård P-O.
      • Carlsson U.
      • Jonsson B.-H.
      • Elwing H.
      ) we showed that, uponadsorption to silica nanoparticles, HCA IIpwt and destabilizedvariants thereof formed a molten globule-like state. This was later shown tobe the case with other protein/surface systems(
      • Engel M.F. M
      • van Mierlo C.P.M.
      • Visser A.J.W.G.
      ). Furthermore, in the caseof HCA II, it was shown that the more destabilized the protein, the fasterthis transition was from the native to the molten globule state afteradsorption (
      • Karlsson M.
      • Mårtensson L-G.
      • Jonsson B.-H.
      • Carlsson U.
      ). Although theA23C/L203Cox variant denatures in accordance with an apparenttwo-state mechanism, all other properties resemble those of HCAIIpwt. Thus, the near-UV CD spectra are almost identical,indicating a very similar tertiary structure, and the enzymatic activity isabout half that of HCA IIpwt(
      • Mårtensson L-G.
      • Karlsson M.
      • Carlsson U.
      ). The activity of S56N isvery close to that of HCA IIpwt, and the mutation had no effect onthe tertiary structure, as judged from CD analysis(
      • Karlsson M.
      • Mårtensson L-G.
      • Jonsson B.-H.
      • Carlsson U.
      ).
      Table ICharacteristics of the protein variants
      ProteinFractional surfaceaccessibility
      The fractional accessibility of the wild-type amino acids to bereplaced.
      PositionCm,NI
      Cm, NI, Cm,IU, and Cm, NU represent thetransition midpoint concentrations from the native (N) to the intermediate (I)state, from the intermediate (I) to the unfolded state (U), and from thenative (N) to the unfolded (U) state upon denaturation in increasingconcentrations of guanidinium chloride, respectively.
      Cm,IU
      Cm, NI, Cm,IU, and Cm, NU represent thetransition midpoint concentrations from the native (N) to the intermediate (I)state, from the intermediate (I) to the unfolded state (U), and from thenative (N) to the unfolded (U) state upon denaturation in increasingconcentrations of guanidinium chloride, respectively.
      Cm,NU
      Cm, NI, Cm,IU, and Cm, NU represent thetransition midpoint concentrations from the native (N) to the intermediate (I)state, from the intermediate (I) to the unfolded state (U), and from thenative (N) to the unfolded (U) state upon denaturation in increasingconcentrations of guanidinium chloride, respectively.
      StabilityTmCO2 hydration activity
      mmmkcal/mol°C%
      S56N
      All data except melting point are from Ref.16.
      0β-strand 20.41.82.9
      For the N⇔I transition.
      46
      For the N⇔I transition.
      81
      HCAIIpwt
      All data except melting point are from Ref.16.
      0.01β-strand 71.01.97.9
      For the N⇔I transition.
      59
      For the N⇔I transition.
      100
      A23C/L203Cox
      All data except melting point are from Ref.30.
      0.08/0310-helix/turn1.811.0
      For the N⇔U transition.
      74
      For the N⇔U transition.
      55
      a The fractional accessibility of the wild-type amino acids to bereplaced.
      b Cm, NI, Cm,IU, and Cm, NU represent thetransition midpoint concentrations from the native (N) to the intermediate (I)state, from the intermediate (I) to the unfolded state (U), and from thenative (N) to the unfolded (U) state upon denaturation in increasingconcentrations of guanidinium chloride, respectively.
      c All data except melting point are from Ref.
      • Karlsson M.
      • Mårtensson L-G.
      • Jonsson B.-H.
      • Carlsson U.
      .
      d For the N⇔I transition.
      e All data except melting point are from Ref.
      • Mårtensson L-G.
      • Karlsson M.
      • Carlsson U.
      .
      f For the N⇔U transition.
      Figure thumbnail gr1
      Fig. 1Protein stability curves for HCA II variants, based on values obtainedby tryptophan fluorescence measurements in various concentration ofGu-HCl. ○, S56N; •, HCA IIpwt; ▵,A23C/L203Cox. The HCA IIpwt and S56N curves were fittedto a three-state transition (N→ I→ U), and theA23C/L203Cox curve was fitted to a two-state function (N→ U).Data for HCA IIpwt and S56N were taken from Ref.
      • Karlsson M.
      • Mårtensson L-G.
      • Jonsson B.-H.
      • Carlsson U.
      , and data for A23C/L203Cwere taken from Ref.
      • Mårtensson L-G.
      • Karlsson M.
      • Carlsson U.
      . Theparameters monitored are summarized in.
      Surfaces—In Fig.2, infrared reflection-absorption spectra are shown for thesurface analogues assembled on gold. In addition to bands shown inFig. 2, several bands werepresent at 2800–3100 cm–1 (C-H, several differentlyfunctionalized hydrocarbons). The amide-I, -II, and -III bands are visible forall surfaces at 1650, 1550, and 1260 cm–1, respectively. Forthe terminal amine, the amide-I band coincides with bands fromC-NH+3. Characteristic sulfate bands are visible at1250, 1080, and 1030 cm–1. The ellipsometrically determinedthickness is in good agreement with results from molecular modeling(Table II). Values from contactangle measurements are also shown in TableII. Values are within the ranges expected for the presentfunctionalities. Measurements from infrared reflection-absorption spectroscopyand ellipsometry showed that monolayers with the correct thickness for a 100%dense layer of the four analogues are formed on gold. The low absorptiondisplayed for the amide-I stretch indicates that the transition dipole isoriented perpendicular to the surface normal, thus having a molecularorientation of the carbon chain that is parallel to the surface normal(
      • Valiokas R.
      • Svedhem S.
      • Svensson S.C.T.
      • Liedberg B.
      ). The difference inintensity of the amide-I stretch for the sulfate functionalized thiol is oflow significance. The difference, however, indicates a slightly largerdeviation from the surface normal orientation than for the correspondingmethyl- and hydroxyl-functionalized molecules. Moreover, values from contactangle goniometry indicate that the derivatized group of the various assembledthiols (Fig. 2) are facedtoward the environment. For convenience, we will refer to the surfaces asnegatively charged (-SO4), hydrophilic (-OH),hydrophobic (-CH3), or positively charged(-NH+3).
      Figure thumbnail gr2
      Fig. 2Infrared reflection-absorption spectra of the methyl-, aminohydroxy-,and sulfate-terminated thiols assembled on gold. The region shown is thefingerprint region of the spectra.
      Table IISurface thickness and contact angles
      Surface functionalityd(Å)
      Values are given with a 95% confidence interval. The number of measurementsis given in parentheses.
      d (Å)estimated
      From molecular modeling.
      θ
      Contact angles. Values are given with a 95% confidence interval, measuredon five separate samples.
      a
      Advancing contact angle.
      ,w
      Water.
      θ
      Contact angles. Values are given with a 95% confidence interval, measuredon five separate samples.
      a
      Advancing contact angle.
      ,hd
      Hexadecane.
      θ
      Contact angles. Values are given with a 95% confidence interval, measuredon five separate samples.
      r
      Retracting contact angle.
      ,w
      Water.
      θ
      Contact angles. Values are given with a 95% confidence interval, measuredon five separate samples.
      r
      Retracting contact angle.
      ,hd
      Hexadecane.
      Negative (-SO4-)11.3 ± 1.0 (7)11.4<10<10<10<10
      Hydrophilic(-OH)
      Values presented are from previous work(26).
      9.0 ± 0.4 (5)9.126 ± 5<10<10<10
      Hydrophobic (-CH3)9.1 ± 0.8 (7)8.582 ± 124 ± 166 ± 1<10
      Positive (-NH3+)9.3 ± 0.9 (7)9.137 ± 1<10<10<10
      a Values are given with a 95% confidence interval. The number of measurementsis given in parentheses.
      b From molecular modeling.
      c Contact angles. Values are given with a 95% confidence interval, measuredon five separate samples.
      d Advancing contact angle.
      e Water.
      f Hexadecane.
      g Retracting contact angle.
      h Values presented are from previous work(
      • Ekeroth J.
      • Borgh A.
      • Konradsson P.
      • Liedberg B.
      ).
      Kinetics of Initial Adsorption—There appears to be no largedifferences in adsorption kinetics for the protein variants(Fig. 3,A–D). This is what is to be expectedbecause, regardless of protein variant, surface potentials and so forth areidentical between the three protein variants used. Some features of theadsorption kinetics are also common for all surfaces when analyzing the data.Firstly, as adsorption begins, the surface can be considered to be“infinite,” and the kinetics during the initial adsorption ispseudo first order. The fitting of the data reveals that, regardless ofsurface, the amplitude of the initial phase is always lowest for thedestabilized variant S56N, followed by HCA IIpwt and the morestable A23C/L203Cox (Fig.3 and Table III).This can be explained by the faster rate at which the less stable S56N variantchanges its native state to a more denatured, expanded state after adsorptiononto the respective surface. This has also been observed in an earlier study(
      • Karlsson M.
      • Mårtensson L-G.
      • Jonsson B.-H.
      • Carlsson U.
      ), in which the rate ofconformational changes was studied by CD upon adsorption of stability variantsof HCA II to negatively charged silica nanoparticles. Because of the fasterspreading of the less stable S56N, a larger surface is occupied within ashorter time, leading to lower amplitude of the initial phase. The more stableA23C/L203Cox variant, on the other hand, has the largest amplitudeof the three protein variants on all surfaces, implying that the structuralintegrity of the protein is maintained for a longer time after adsorption.Secondly, there is also a correlation between the stability of the proteinvariant and the rate constant of the initial adsorption. Although thedifferences are fairly small, the rate constants for the more stableA23C/L203Cox variant are lower, in all cases, than those for S56Nand HCA IIpwt, during the initial adsorption, despite the fact thatthe amplitude for the initial phase of adsorption is the largest forA23C/L203Cox and that the attractive forces between the respectivesurface and each protein variant are identical. Hence, the differences in therate constants of the initial adsorption are supportive of the notions that anincrease in entropy of the adsorbed state could be a driving force foradsorption and that a hard (stable) protein is less prone to adsorb to a solidsurface at a solid surface-liquid interface(
      • Norde W.
      ). Furthermore, because thestructure of A23C/L203Cox is tethered by a disulfide bridge, evenless entropy gain can be expected for this variant, even if it loses itsnative conformation after adsorption.
      Figure thumbnail gr3
      Fig. 3Isotherms of the adsorption and desorption of protein variants.A, negatively charged surface; B, hydrophilic surface;C, hydrophobic surface; D, positively charged surface. S56N,cyan; HCA IIpwt, black; A23C/L203Cox,red. The horizontal bars denote the amplitude of the firstkinetic phase of the respective protein variant (equivalent to the values in). Each isoterm is theaverage of duplicate or triplicate SPR measurements.
      Table IIIKinetic data associated with the initial adsorption step
      SurfaceS56NHCA IIpwtA23C/L203Cox
      k1
      The rate constants and amplitudes were calculated using a nonlinear fitprogram (see “Materials and Methods”).
      A1
      The rate constants and amplitudes were calculated using a nonlinear fitprogram (see “Materials and Methods”).
      r2
      The goodness of fit was calculated by the r2coefficient of determination.
      k1
      The rate constants and amplitudes were calculated using a nonlinear fitprogram (see “Materials and Methods”).
      A1
      The rate constants and amplitudes were calculated using a nonlinear fitprogram (see “Materials and Methods”).
      r2
      The goodness of fit was calculated by the r2coefficient of determination.
      k1
      The rate constants and amplitudes were calculated using a nonlinear fitprogram (see “Materials and Methods”).
      A1
      The rate constants and amplitudes were calculated using a nonlinear fitprogram (see “Materials and Methods”).
      r2
      The goodness of fit was calculated by the r2coefficient of determination.
      min-1min-1min-1
      Negative28820.99722910.996181410.992
      Hydrophilic34660.99539700.98725810.997
      Hydrophobic37450.99930540.99429620.999
      Positive3360.99936120.99916220.999
      a The rate constants and amplitudes were calculated using a nonlinear fitprogram (see “Materials and Methods”).
      b The goodness of fit was calculated by the r2coefficient of determination.
      Adsorbed Mass and Orientation—Under the experimentalconditions used, all adsorption isotherms(Fig. 3,A–D) display a continuous adsorptionwithout any overshoot. The presence of an overshoot in adsorption isothermshas previously been explained by reorientation of adsorbed molecules at thesurface (
      • Wertz C.F.
      • Santore M.M.
      ) or a displacementof adsorbed proteins by “competitive spreading”(
      • Cohen Stuart M.A.
      ). The adsorbed massdiffers substantially between the different surfaces and differs less withinthe protein stability variants. In most of the cases, the more stableA23C/L203Cox variant, with a melting point of 74 °C (data notshown) and a reduced tendency to form a molten globule state under denaturingconditions in solution (
      • Mårtensson L-G.
      • Karlsson M.
      • Carlsson U.
      ),adsorbs to the largest extent. This is likely to be due to the fact that thisvariant undergoes fewer conformational changes during initial phases ofadsorption, leaving more space for molecules, which adsorb subsequently. Thisalso leads to the conclusion that a higher thermodynamic stability of aprotein does not mean that such a protein will adsorb to any smaller extent ascompared with a less stable protein. On the other hand, the least stablevariant, S56N, also appears to adsorb to a large extent. However, as isevident from the kinetic analysis and the adsorption isotherms, this is mostprobably due to the formation of a second or additional layers of proteinsforming on the surface of the already adsorbed protein molecules. This can beseen as a slow linear phase of continuous adsorption(Fig. 3,A–C), which is not present in theadsorption isotherms of HCA IIpwt and A23C/L203Cox,which reach a plateau value and stay constant after a few minutes of exposureto the protein solution on all surfaces except for the positively charged one(Fig. 3,A–D). The slow linear phase displayed byS56N can be explained by the fact that S56N has a rather low stability insolution, with a melting point of 46 °C and an unfolding transition thatstarts at ∼10 °C below this value (data not shown). Thus, some of theprotein molecules in solution are likely to form aggregates with the adsorbedS56N, which, in the adsorbed state, rapidly forms molten globule-likestructures (
      • Karlsson M.
      • Mårtensson L-G.
      • Jonsson B.-H.
      • Carlsson U.
      ). Indeed, suchmolten globule states of HCA II have been shown to be very prone to aggregate(
      • Hammarström P.
      • Persson M.
      • Freskgård P-O.
      • Mårtensson L.-G.
      • Andersson D.
      • Jonsson B.-H.
      • Carlsson U.
      ). This starts theformation of an additional layer of proteins, which makes up the final, slowlinear phase. If it were not for the additional protein layers formed, it isreasonable to believe that the least stable variant, S56N, would adsorb to thesmallest extent during the time span observed. HCA IIpwt, with theintermediate stability of the three variants, also forms a molten globulestate, although not as fast as the S56N variant(
      • Karlsson M.
      • Mårtensson L-G.
      • Jonsson B.-H.
      • Carlsson U.
      ). The thermostability ofHCA IIpwt is considerably higher than that for S56N, with a meltingpoint of ∼59 °C (data not shown), and as can be seen inTable III, the kineticadsorption amplitude of the HCA IIpwt variant has an intermediatevalue for the first phase, indicating a slower structural alteration ascompared with S56N, making the first kinetic phase last longer. The additionallayer formation of S56N most probably leads to an overestimation of themaximum value of adsorbed protein to the respective functionalized surface aspresented in Table IV for thisvariant. Because of this behavior, the following discussion regarding adsorbedmass and orientation is limited to the HCA IIpwt andA23C/L203Cox variants.
      Table IVAdsorbed protein mass as measured with SPR immediately before and afterthe washing step
      SurfaceS56NHCA IIpwtA23C/L203Cox
      Maximumadsorbed
      Values are the mean of duplicate or triplicate SPR measurements ±S.E.
      Afterwashing
      Values are the mean of duplicate or triplicate SPR measurements ±S.E.
      % desorbedMaximumadsorbed
      Values are the mean of duplicate or triplicate SPR measurements ±S.E.
      Afterwashing
      Values are the mean of duplicate or triplicate SPR measurements ±S.E.
      % desorbedMaximumadsorbed
      Values are the mean of duplicate or triplicate SPR measurements ±S.E.
      Afterwashing
      Values are the mean of duplicate or triplicate SPR measurements ±S.E.
      % desorbed
      ng/cm2ng/cm2ng/cm2ng/cm2ng/cm2ng/cm2
      Negative165 ± 3.999 ± 2.040145 ± 0.035 ± 0.076161 ± 0.720 ± 2.388
      Hydrophilic111 ± 0.072 ± 0.035102 ± 0.033 ± 0.068113 ± 3.024 ± 4.579
      Hydrophobic108 ± 4.592 ± 4.11591 ± 2.055 ± 3.040102 ± 9.548 ± 0.753
      Positive64 ± 1.559 ± 1.0853 ± 4.134 ± 4.03671 ± 1.837 ± 3.248
      a Values are the mean of duplicate or triplicate SPR measurements ±S.E.
      Adsorption to the Negatively Charged Surface—The proteinsare adsorbed in the largest quantities to the negatively charged surface(Fig. 3A). Despitethis, the adsorbed mass does not come close to any of the theoretical coveragevalues for a close packed monolayer or a model based on the random sequentialadsorption (Tables IV andV). However, we have recentlybeen able to determine the adsorption orientation of HCA II in the nativestate to negatively charged silica nanoparticles by the use of fluorescentprobes (
      • Karlsson M.
      • Carlsson U.
      ). This studyrevealed a strong pH dependence on the adsorption orientation, and it alsorevealed that at pH 8.5, the protein adsorbed in neither fully side-on norfully end-on orientations, but rather in orientations that lead to an occupiedsurface area that is an approximate average of these two extremes. In terms ofmass adsorbed, this would thus give a number that is the average between fullyside-on and fully end-on (∼280 ng/cm2 for a close packedmonolayer and 150 ng/cm2 for RSA). As is evident from theexperimentally determined masses (TableIV), the adsorption to negatively charged surfaces comes veryclose to a theoretical value for adsorption in the native state, in a mixedend-on and side-on orientation in a random sequential mode(Table V). This is alsosupported by the lack of an overshoot or lag phase in the adsorptionisotherms, which indicates that no reorientation takes place during or afterthe adsorption step. The fact that the negatively charged surface is also thesurface that displays the largest desorption upon washing with buffer furthersupports the notion that the protein (especially the A23C/L203Coxvariant) is in its native state.
      Table VTheoretically adsorbed masses
      State and orientationSize of proteinSurface contact areaTheoretical close-packed monolayer coverageTheoretical monolayer coverage (RSA)
      nm2ng/cm2ng/cm2
      HCA II native side-on55 × 42 × 3923.8205112
      HCA II native end-on42 × 3913.8351192
      HCA II molten globule side-on63 × 46 × 4331.215685
      HCA II molten globule end-on46 × 4316.6293160
      Adsorption to the Hydrophilic Surface—The hydrophilicsurface comes second with regard to the amount of adsorbed protein(Fig. 3B). Also, theoverall appearance of adsorption and desorption very closely resembles that ofthe negatively charged surface. The adsorbed protein mass is, in this case,close to the value for an RSA model with native protein adsorbing in a side-onmanner (Tables IV andV). Although we lackexperimental data regarding the absolute orientation at hydrophilic surfaces,the large amount of desorbed protein upon washing from this surface alsoimplies that the protein (A23C/L203Cox) has, to a large extent,remained in its native state, supporting side-on orientation adsorption.
      Adsorption to the Hydrophobic Surface—The adsorption at thehydrophobic surface is somewhat smaller than the adsorption at the hydrophilicsurface (Fig. 3C). Itis known that hydrophobic surfaces are often denaturing because of thetendency to force proteins to expose internal, hydrophobic residues;therefore, it is interesting to compare the adsorption with a theoreticalmonolayer based on the molten globule state of the protein(Table V). In this case, themaximally adsorbed protein mass (TableIV) does in fact come fairly close to the theoretical RSA valuefor the protein adsorbing in a side-on manner that subsequently adopts amolten globule state. This is especially evident with the HCA IIpwtvariant, which reaches the molten globule state faster than the more stableA23C/L203Cox variant. Moreover, the substantially decreased proteindesorption upon washing indicates that the adsorbed proteins are more tightlybound to the hydrophobic surface.
      Adsorption to the Positively Charged Surface—The adsorptionbehavior of the proteins to the positively charged surface is significantlydifferent as compared with that for the other surfaces, with a very low massadsorbed, despite the fact that the protein is expected to carry a netnegative charge at pH 8.5 (Fig.3D). However, this is most probably due to adoption ofconformations that are even less ordered than the molten globule state. Thisinterpretation is also supported by the very low amplitude of the initialphase upon adsorption and the low desorption of protein upon washing, withvirtually no release of S56N and only a minor fraction of desorbing HCAIIpwt. This indicates that the surface is highly denaturing,leading to a substantial increase of interactions between the surface and theprotein with time. It is also noteworthy that adsorption isotherms of thissurface show that none of the protein variants reaches a plateau value duringthe experimental time.
      Correlation between Protein Stability and Desorption—We havechosen not to make an attempt to analyze the desorption kineticsquantitatively because the desorption occurs from different types ofinteractions; for example, the more stable A23C/L203Cox woulddesorb from the actual surface, whereas the less stable S56N variant woulddisplay desorption from a partial additional layer. The most striking featureof the four differently functionalized surfaces and the three stabilityvariants is the difference in the amount of desorbed protein. Firstly, thereis a strong correlation between how much protein is released upon washing withbuffer and how much protein can be maximally adsorbed to the surface; thus,the more denaturing the surface (low release of adsorbed protein), the lowerthe adsorbed amount, which can be explained in terms of both the fasterdenaturation rate of the protein at the surface and the degree ofdenaturation. Secondly, as can be seen inTable IV andFig. 3, there is a very strongcorrelation between protein stability and the amount of protein desorbed onall surfaces. The correlation between increased protein stability andincreased desorption upon washing with buffer becomes even more evident whenthe percentage of desorbed protein is plotted as a function of proteinstability, expressed as kcal/mol from chemical denaturation(
      • Karlsson M.
      • Mårtensson L-G.
      • Jonsson B.-H.
      • Carlsson U.
      ,
      • Mårtensson L-G.
      • Karlsson M.
      • Carlsson U.
      )(Fig. 4A) or asTm (Fig.4B). On the least denaturing surface (negativelycharged), ∼90% of the stable A23C/L203Cox variant is desorbedby washing with buffer, whereas only 40% of the least stable S56N variant isdesorbed. On the most denaturing surface (positively charged), <10% of thedestabilized S56N variant is desorbed, whereas about half of the stabilizedA23C/L203Cox variant is desorbed, and the value for HCAIIpwt is intermediate between the two. This pattern is evident,from the least denaturing surface to the most denaturing surface.
      Figure thumbnail gr4
      Fig. 4Percentage of desorbed protein after washing with buffer for 30 minplotted against the (A) thermodynamic stability and (B)melting point (Tm) of investigated protein variants foreach surface. ▾, negative surface; ○, hydrophilic surface;•, hydrophobic surface; +, positive surface.

      CONCLUSIONS

      The negatively charged and hydrophilic surfaces are the least denaturingsurfaces of the four surfaces tested. There are also indications that theadsorption to the surfaces occurs according to an RSA model with specificorientations, with the negatively charged surface adsorbing mainly nativeprotein in a mixed side-on and end-on manner, in agreement with earlierfindings (
      • Karlsson M.
      • Carlsson U.
      ), whereas thehydrophilic surface adsorbs in a side-on manner. The more denaturing,hydrophobic surface has a lower maximal surface load because the protein islikely to adopt a more expanded structure on this surface as compared withless denaturing surfaces. Furthermore, the adsorbed mass is, in this case,close to the theoretical value for the protein being in the molten globulestate that has adsorbed in a side-on orientation under random sequentialadsorption. The adsorption to the positively charged surface is more difficultto interpret because both the adsorbed and desorbed masses are very low. It isevident from Fig. 3 andTable IV that the stability ofthe protein does not influence the amount of adsorbed protein such that a morestable protein will adsorb to a less extent than a destabilized protein. Onthe contrary, the more stable A23C/L203Cox actually adsorbs withthe highest surface load on several of the surfaces. The interpretation ofthese findings can be explained in terms of the denaturing behavior of thethree protein variants, for which we have shown previously that S56N, whenadsorbed to negatively charged silica nanoparticles, is completely convertedto a molten globule form within about 15 min, whereas HCA IIpwtonly slowly goes through this transformation. These different behaviors of thethree protein stability variants not only influence the total amount ofadsorbed protein but also influence the kinetics of adsorption.
      Most interestingly, the most stable protein variant,A23C/L203Cox, is the one that most readily desorbs from the varioussurfaces, despite being the protein that adsorbs to the largest extent. Toconclude, this also emphasizes the importance of having a well-defined systemwhen comparing the behavior of different proteins at surfaces because only asingle parameter (in this case, stability) gives such significant differencesin adsorption/desorption behavior. Furthermore, it is only by knowing thestability and unfolding behavior of the protein variants and by monitoringboth the adsorption and desorption characteristics that we have been able todraw conclusions about the state and orientation of the adsorbed proteins.
      Finally, our results also demonstrate that using protein engineering toincrease protein stability with the aim of reducing irreversible proteinadsorption can be added to the list of rationales for proteinstabilization.

      References

        • Hlady V.
        • Buijs J.
        Curr. Opin.Biotechnol. 1996; 7: 72-77
        • Nakanishi K.
        • Sakiyama T.
        • Imamura K.
        J. Biosci. Bioeng. 2001; 91: 233-244
        • Matheis W.
        • Zott A.
        • Schwanig M.
        Vaccine. 2002; 20: 67-73
        • McNay J.
        • Fernandez E.
        Biotechnol. Bioeng. 2001; 76: 224-232
        • McNay J.
        • O'Connell J.
        • Fernandez E.
        Biotechnol. Bioeng. 2001; 76: 233-240
        • Wisniewski N.
        • Reichert M.
        Colloids Surf. B Biointerfaces. 2000; 18: 197-219
        • Chapman R.G.
        • Ostuni E.
        • Takayama S.
        • Holmlin R.E.
        • Yan L.
        • Whitesides G.M.
        J. Am. Chem.Soc. 2000; 122: 8303-8304
        • Ostuni E.
        • Chapman R.G.
        • Holmlin R.E.
        • Takayama S.
        • Whitesides G.M.
        Langmuir. 2001; 17: 5605-5620
        • Roque C.
        • Lowe C.
        • Taipa A.
        Biotechnol. Prog. 2004; 20: 639-654
        • McLeod A.G.
        • Walker I.R.
        • Zheng S.
        • Hayward C.P.M.
        Haemophilia. 2000; 6: 89-92
        • Tzannis S.
        • Hrushesky W.
        • Wood P.
        • Przybycien T.
        Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 5460-5465
        • Norde W.
        Adv. Colloid InterfaceSci. 1986; 25: 267-340
        • Wendorf J.
        • Radke C.
        • Blanch H.
        Biotechnol. Bioeng. 2004; 87: 565-573
        • Billsten P.
        • Freskgård P-O.
        • Carlsson U.
        • Jonsson B.-H.
        • Elwing H.
        FEBS Lett. 1997; 402: 67-72
        • Malmsten M.
        J. Colloid InterfaceSci. 1998; 207: 186-199
        • Karlsson M.
        • Mårtensson L-G.
        • Jonsson B.-H.
        • Carlsson U.
        Langmuir. 2000; 16: 8470-8479
        • Moulin A.M.
        • O'Shea S.J.
        • Badley R.A.
        • Doyle P.
        • Welland M.E.
        Langmuir. 1999; 15: 8776-8779
        • McGuire J.
        • Wahlgren M.
        • Arnebrant T.
        J. Colloid Interface Sci. 1995; 170: 182-192
        • Ball V.
        • Bentaleb A.
        • Hemmerle J.
        • Voegel J-C.
        • Schaaf P.
        Langmuir. 1996; 12: 1614-1621
        • Bower C.K.
        • Sananikone S.
        • Bothwell M.K.
        • McGuire J.
        Biotechnol. Bioeng. 1999; 64: 373-376
        • Ramsden J.J.
        • Roush D.J.
        • Gill D.S.
        • Kurrat R.
        • Willson R.C.
        J. Am. Chem. Soc. 1995; 117: 8511-8516
        • Malmsten M.
        • Burns N.
        • Veide A.
        J. Colloid Interface Sci. 1998; 204: 104-111
        • Burkett S.R.
        • Read M.J.
        Langmuir. 2001; 17: 5059-5065
        • Carignano M.A.
        • Szleifer I.
        Colloids Surf. B Biointerfaces. 2000; 18: 169-182
        • Petrash S.
        • Cregger T.
        • Zhao B.
        • Pokidysheva E.
        • Foster M.D.
        • Brittain W.J.
        • Sevastianov V.
        • Majkrzak C.F.
        Langmuir. 2001; 17: 7645-7651
        • Ekeroth J.
        • Borgh A.
        • Konradsson P.
        • Liedberg B.
        J. Colloid Interface Sci. 2002; 254: 322-330
        • Ekeroth J.
        • Konradsson P.
        • Höök F.
        Langmuir. 2002; 18: 7923-7929
        • Mårtensson L-G.
        • Jonsson B.-H.
        • Andersson M.
        • Kihlgren A.
        • Bergenhem N.
        • Carlsson U.
        Biochim.Biophys. Acta. 1992; 1118: 179-186
        • Mårtensson L-G.
        • Jonsson B.-H.
        • Freskgård P.-O.
        • Kihlgren A.
        • Svensson M.
        • Carlsson U.
        Biochemistry. 1993; 32: 224-231
        • Mårtensson L-G.
        • Karlsson M.
        • Carlsson U.
        Biochemistry. 2002; 41: 15867-15875
        • Karlsson M.
        • Mårtensson L-G.
        • Olofsson P.
        • Carlsson U.
        Biochemistry. 2004; 43: 6803-6807
        • Karlsson M.
        • Mårtensson L-G.
        • Karlsson C.
        • Carlsson U.
        Biochemistry. 2005; 44: 3487-3493
        • Santoro M.M.
        • Bolen D.W.
        Biochemistry. 1988; 27: 8063-8068
        • Höök F.
        • Kasemo B.
        • Nylander T.
        • Fant C.
        • Sott K.
        • Elwing H.
        Anal. Chem. 2001; 73: 5796-5804
        • Välimaa L.
        • Petterson K.
        • Rosenberg J.
        • Karp M.
        • Lövgren T.
        Anal. Biochem. 2004; 331: 376-384
        • Uversky V.N.
        Biochemistry. 1993; 32: 13288-13298
        • Jonsson M.
        • Pettersson E.
        ActaChem. Scand. 1968; 22: 712-713
        • Ptitsyn O.B.
        • Bychkova V.E.
        • Uversky V.N.
        Philos. Trans. R. Soc. Lond. B Biol.Sci. 1995; 348: 35-41
        • Engel M.F. M
        • van Mierlo C.P.M.
        • Visser A.J.W.G.
        J. Biol. Chem. 2002; 277: 10922-10930
        • Valiokas R.
        • Svedhem S.
        • Svensson S.C.T.
        • Liedberg B.
        Langmuir. 1999; 15: 3390-3394
        • Wertz C.F.
        • Santore M.M.
        Langmuir. 2002; 18: 1190-1199
        • Cohen Stuart M.A.
        Malmsten M. Biopolymersat Interfaces. Marcel Dekker, Inc., New York1998: 1-25
        • Hammarström P.
        • Persson M.
        • Freskgård P-O.
        • Mårtensson L.-G.
        • Andersson D.
        • Jonsson B.-H.
        • Carlsson U.
        J. Biol. Chem. 1999; 274: 32897-32903
        • Karlsson M.
        • Carlsson U.
        Biophys. J. 2005; 88: 3536-3544