SEPTEMBEWOCTOBER 1995 Volume 6, Number 5 0Copyright 1995 by the American Chemical Society
Bioconjugate Chemistry
COMMUNICATIONS Surface-ModifiedDiamond Nanoparticles as Antigen Delivery Vehicles Nir Kossovsky,*bt Andrew Gelman,; H. J a m e s Hnatyszyn,' Samir Rajguru,' Robin L. Garrell,*,t Shabnam Torbati,$ Siobhan S. F. Freitas,t a n d Gan-Moog Chows Biomaterials Bioreactivity Characterization Laboratory, University of California, Los Angeles School of Medicine, Los Angeles, California 90024-1732, Department of Chemistry & Biochemistry, University of California, Los Angeles, California 90024-1569, and Laboratory for Molecular Interfacial Interactions, Center for BiolMolecular Science and Engineering, Code 6930, Naval Research Laboratory, Washington, DC 20375. Received June 26, 1995@
Recognition of antigens by immunocompetent cells involves interactions that are specific to the chemical sequence and conformation of the epitope (antigenic determinant). Adjuvants that are currently used to enhance immunity to antigens tend to either alter the antigen conformation through surface adsorption or shield potentially critical determinants, e.g., functional groups. It is demonstrated here that surface-modified diamond nanoparticles (5-300 nm) provide conformational stabilization, as well as a high degree of surface exposure to protein antigens. By enhancing the availability and activity of the antigen in vivo, a strong, specific immune response can be elicited. Results are demonstrated for mussel adhesive protein (MAP), a substance for which conventional adjuvants have proven only marginally successful in evoking a n immune response. Surface-modified diamond nanoparticles as antigen delivery vehicles are a novel example of the exciting marriage of materials science, chemistry, and biology.
The ability to generate antigen-specific antibodies has led to significant advances in molecular localization (11, molecular recognition and catalysis (21, biosensors and immunoassays (3), vaccine development, and the characterization of conformational changes induced by molecular interactions (4). Nevertheless, certain fields of research have been hampered or misguided by the inability to evoke reactive antibodies in a mammalian
* Authors to whom correspondence should be addressed. University of California, Los Angeles School of Medicine. University of California, Los Angeles. 4 Naval Research Laboratory. t 3 Abstract published in Advance ACS Abstracts, September 1, 1995.
host. For example, when Freund's adjuvant is used, mussel adhesive protein (MAP) elicits only a weak immune response in New Zealand white rabbits (5). This has hindered the development of a simple, antibodybased purification for MAP, which is potentially useful as a corrosion inhibitor and surgical adhesive (6). The apparently weak immunogenicity of MAP has also led to the inference that MAP is likely to be a safe material for use i n vivo (5),which in turn has led to tests of MAPbased adhesives in animal models (7). We show here that surface-modified diamond nanoparticles can be used as very effective antigen delivery vehicles. The coated particles, which consist of a diamond substrate, a glassy carbohydrate film, and an immunologically active surface
1043-1802/95/2906-0507$09.00/0 0 1995 American Chemical Society
Kossovsky et al.
508 Bioconjugate Chem., Vol. 6, No. 5, 1995
molecule in an aqueous dispersion, represent a novel marriage of materials science, surface chemistry, and immunology. It will be shown that these antigen delivery vehicles evoke a strong immune response to antigens such as MAP. These results indicate the need to reevaluate the apparent biocompatibility of MAP-based materials. More generally, they suggest that modified diamond nanoparticles have great potential as alternatives to conventional adjuvants. Our results demonstrate the feasibility of optimizing antigen-substrate interactions to enhance antigen presentation and immunogenicity. Antigen carriers and adjuvants are selected ostensibly to enhance the immunogenicity of protein antigens; they play a major role in controlling the conformation of the antigen by virtue of their close physical association with it (8).Although B-cells usually recognize protein antigens in their native conformational state (91,conventional methods for raising antibodies tend to either alter the antigen through surface adsorption or shield potentially critical determinants (8). Several years ago, we developed a solid-phase, high surface-area nanocrystalline antigen carrier that restricts antigen penetration into the surface of the carrier, thereby preventing determinant shielding (10).In vivo experiments with surface-modified tin oxide yielded antibodies that exhibited in vitro viral neutralization. Because of the desirability of using a high surface energy ceramic, but concerned over the potential toxicity of tin, we subsequently began experiments with carbon ceramic (diamond) nanoparticles. Figure 1 is a transmission electron micrograph showing acid- and water-washed nanocrystalline diamond particles (General Electric, Worthington, OH) that were dried and suspended on a carbon-coated TEM grid. The morphology of the particles was assessed by bright field imaging, which showed that the particles are multifaceted with varying aspect ratios. Dark field imaging revealed a mixture of crystallite sizes ranging from 5 to 10 nm, with rare 100-300 nm polygons. The crystalline nature of the particles is seen extending to the surface. The actual antigen carrier consisted of the diamond particles coated with cellobiose, a disaccharide. Because diamond is a high-surface energy material, it was anticipated that thermodynamics would favor adsorption and adhesion of celloboise onto the diamond particles. This would create a colloid surface capable of hydrogenbonding to the proteinaceous antigen that would subsequently be adsorbed. Furthermore, the disaccharide could act as a dehydroprotectant and help minimize surface-induced denaturation of the subsequently adsorbed antigen (11). Figure 2 is a HRTEM image of the diamond particles modified with cellobiose. The glassy (amorphous) cellobiose coating was visible only by HRTEM and was between 4 and 6 nm thick. The coating was generally not uniform and focally seemed to consist of two layers. The nonuniformity may be related to the aspect ratio of the particles, since the shape of the One g of diamond powder (General Electric, Worthington OH) was cleaned by 400 W sonication at 4 "C in 12 N HCl for 16 h and subsequently washed with ultrafiltered water until the pH was near 7. The resulting opaque dispersion was layered over glass plates and baked in a vacuum oven for 2 days at 185 "C. The dried diamond was then rehydrated and acid washed as described above. The clean, activated diamond dispersions in ultrafiltered deionized water were diluted to 1.0 mg/mL and then added to 250 mM cellobiose [Sigma, St. Louis, M o l and lyophilized for 24 h. Unadsorbed cellobiose was removed by ultrafiltration dialysis against sterile water in a 100 kD nominal molecular-weight-cutoff stir cell [Filtron, Northborough, MA] a t room temperature.
,-"
nr(...w--
r
. ' -
Figure 1. Transmission electron micrograph of acid- and water-washed and sonicated nanocrystalline diamond particles. Images obtained by (top) conventional microscopy with a JEOL JEM 200CX electron microscope operated a t 200 keV and (bottom) high-resolution transmission electron microscopy (HRTEM) with a Hitachi H-9000UHR microscope a t 300 keV. The lattice fringes for the two particles in this field correspond to the (111)planes.
diamond crystals may affect the adsorption onto their their surfaces. The equilibrium shape of a crystal is such that the total surface free energy is a minimum, and it can be geometrically constructed by Wulff s theorem (12). For a crystal with a polyhedral shape, the surface energy is different for different faces, and adsorption onto the facets will therefore also differ, leading to nonuniform coverage and thickness of adsorbed layers. For diamond, the surface energy for the (111)plane is much smaller than that of the (100)plane (i.e., 5400 ergs/cm2and 9400 ergdcm2, respectively) (Harkins, ref 12). The antigen used in these studies was MAP. It is an unusual protein, consisting largely of a repeating (con2 MAP was purified according to the procedure developed by Waite (15).The final purification step was by gel filtration on a Sephadex G-150 column. Fractions were analyzed by W-vis spectroscopy to find the fraction with maximum absorbance (usually the first fraction after the void volume). Identity and purity were assessed by PAGE (slab gels, Coomassie Blue and NitroBlue Tetrazolium (NBT) stain, which is Dopa-specific) and HPLC against a n authentic sample (courtesy of J. H. Waite).
Bioconjugate Chem., Vol. 6, No. 5, 1995 509
Communications
-5
P
a) EUSA of Rabbn anti-MAP Antibody Activity (PS plate) 1.4
1.2
8 1
g
0.6 0.2
4
0
1
0
--c
immunized serum binding to MAP
A
immunized serum binding to BSA
naive serum binding to MAP
naive serum binding to BSA
b) ELISA of Rabbit anti-MAP Antibody Activity on PS and PDMS-coated PS
5 4 ' i r i
0 polystyrene PDUScoated polystyrene
Figure 2. HRTEM image of cellobiose-modified diamond nanoparticles.
K
immunized serum binding to MAP
immunized serum binding to BSA
naive serum binding to MAP
naive serum binding to BSA
Figure 4. Rabbit serum antibody avidity to surface immobilized MAP as measured by ELISA. (a) Binding of immunized serum and naive (nonimmunized) serum to MAP (specific binding) and BSA (nonspecific binding) immobilized on standard polystyrene ELISA plates. Each error bar represents the standard deviation in five measurements. (b) Comparison of antibody binding (specific and nonspecific) to proteins immobilized on polystyrene and poly(dimethylsi1oxane)-coated ELISA plates (PDMS = poly(dimethylsi1oxane)). Ser
Tyr
PH
-N-&IIC--N-CCH-C-NH-CH-C-NH-CH-C-NH-CH-C
II
HYP
II
HYP
\I
II Dopa
Thr
t
! Lys
II Ala
Lys
Pro
Thr
Tyr
Lys
Figure 3. MAP from Mytilus edulis, MW -130 kD,consists primarily of repeating deca- and hexapeptide sequences, in which the 0's indicate sites of additional hydroxyl groups on some of the repeats. Figure based on ref 13.
sensus) decapeptide in which hydroxyl or amine groups are present on virtually every residue (Figure 3) (13). Previous reports indicated only poor or marginal success in generating antibodies to MAP using Freund's adjuvant (5),which is mineral-oil based. The lack of success may be partly attributable to the hydrophilicity of the protein. MAP may be sequestered in aqueous microdroplets within the adjuvant bolus, denatured at the oil-intercellular fluid interface, or complexed with trehalosedimycolate moieties (14) at the surface of the bolus in a way that diminishes MAP'S immunogenicity. To adsorb MAP onto the modified diamond particles, 1.0mg of MAP was first solubilized in 2.5 mL of 0.10 M acetic acid (pH 4.7) by gentle agitation. This was added to 1.0 mL of 1.0 mg/mL cellobiose-coated diamond in water in a 100 kD stir cell a t 4 "C and then dialyzed against 150 mL of 20 mM phosphate buffer (pH 7.4). Under these conditions, MAP precipitates irreversibly (6).
To prepare the MAP-diamond nanoparticle couple, highly purified MAP2 was mixed with ca. 4-20 nmdiameter cellobiose-modified diamond particles and dial y ~ e d .Adsorption ~ of l" onto the cellobiose-coated particles was assessed by electrophoretic light scattering, which revealed colloidally dispersed solid aggregates with a mean diameter of 300 nm.4 The product, an aqueous MAP adsorption onto the cellobiose-modified diamond was assayed by measuring changes in the nanoparticle electrophoretic mobility (DELSA 440, Coulter Electronics Inc., Hialeah, FL) as described previously (16). Prior to measurement, all samples were diluted 150 in one of several pH buffer solutions prepared from monobasic and dibasic sodium phosphates spanning the range from pH 5.8 to 9.4. Samples were positioned so that mobility readings were taken at the solvent's stagnant point. Light scattering was used to assess the particles' terminal velocity (mobility) under the influence of an electric field, which is proportional to the 1; potential or surface charge. The electrophoretic mobility of the cellobiose-coated diamond particles did not change significantly over a period of 8 weeks in cold storage or at room temperature. As a further test of the stability of the cellobiose coating, a suspension of 1.0 mg/mL cellobiose-coated diamond that was aged at least 2 weeks at room temperature and centrifuged at 60000g to precipitate the particles. The supernatant was analyzed by HPLC with a Waters SC-1011column and refractive index detection. The concentration of dissolved cellobiose was only 6.3 pM, which corresponds to a loss of only 0.2% of the original mass of the suspended coated particles. Together, the electrophoretic mobility and HPLC results indicate that the sugar coating does not redissolve appreciably. The MAP-modified particles are stable in cold storage for at least 1 month, although aggregation and other changes occur within a few days a t room temperature. Five 2-month-old female New Zealand white rabbits were injected intramuscularly in the right hind leg with approximately 500 pg of the MAP-diamond conjugate carried in 1.0 mL of 20 mM phosphate buffer (pH 7.4). After 2 weeks, the rabbits were boosted with an additional 100 pg of suspended MAP conjugate. The animal was exsanguinated by intracardial puncture 2 weeks after the boost, and anti-MAP antibodies were quantified by enzyme-linked immunoassay (ELISA).
Kossovsky et al.
510 Bioconjugate Chem., Vol. 6, No. 5, 1995
colloid comprised of a three-layered solid phase (diamond/ cellobiose/MAP), was fluid and readily injectable. Antibodies were raised against the surface-immobilized MAP in New Zealand white rabbit^,^ and their specificity against MAP and its various conformational epitopes was measured by ELISA.6 In contrast to earlier efforts in which anti-MAP antibodies were raised only with difficulty (5),the presentation of antigen on surface-modified diamond nanocrystalline particulates yielded a strong and specific antibody response (Figure 4). Figure 4a shows the actual ELISA data for immunized serum and naive serum, binding to MAP (specific binding) and BSA (nonspecific binding) immobilized on standard polystyrene (PS)ELISA plates. Each error bar represents the standard deviation in five measurements. The binding avidity of rabbit anti-MAP antibodies to the immobilized MAP is substantially greater than for immobilized bovine serum albumin (BSA). The latter response (nonspecific binding activity) is comparable to that measured for naive serum. The standard deviation in the antibody activity of serum from four different animals was 0.014 absorbance units, less than the standard deviation in three successive activity measurements on the same serum (0.040). This demonstrates that the effectiveness of cellobiose-coateddiamond particles as antigen delivery vehicles is reproducible in different animals. Figure 4b shows the same data as Figure 4a, along with data for plates that were coated with PDMS prior to protein adsorption. The absorbances have been normalized relative to the value measured for immunized serum binding to MAP. Antibodies raised against the aqueous conformation of MAP bind avidly to MAP immobilized on the more hydrophilic surface (standard The binding avidity of rabbit IgG to MAP was assayed by conventional ELISA protocol using purified MAP a s the test antigen, bovine serum albumin (BSA) a s the negative antigen control, and naive rabbit serum as a negative serum control. The conformational specificity of the antibody binding avidity was measured against MAP immobilized on polystyrene surface and MAP immobilized on a siliconized surface. Briefly, standard polystyrene 96-well microtiter plates (Falcon 3913 microtest 111, Becton Dickinson, Franklin, NJ) were cleaned with 1.2 N HCl for 1 h to remove all surface contaminants. The plates were neutralized by washing three times with water and then were either left untreated (noncoated) or were filled with 300 pL per well poly(dimethylsi1oxane)(PDMS)oil (Dow Corning 200 fluid, 100 cSt) and then emptied gravitometrically upside down a t 1000 rpm for 10 min. MAP (1mg in 3 mL 100 mM acetic acid, pH 4.7) or BSA (1 mg in 3 mL 20 mM potassium phosphate buffer, pH 7.3) was then added to completely fill each well and allowed to bind overnight a t 4 "C, after which the wells were emptied. A blocking solution of 250 pL of 2% BSA in PBS was added to each well and incubated for 3 h a t room temperature. The plates were washed with three successive 250 pL aliquots of wash buffer (1.91 mM KH2POd8.08 mM Na2HPOd150 mM NaCV0.5% v/v Tween-20, a t pH 7.4). Two hundred fifty pL of sera from the rabbits were added in triplicate (diluted 1:lOO in phosphate-buffered saline per well per plate) and incubated overnight a t 4 "C. The following morning, the plates were washed three times with wash buffer, after which 125 pL of secondary goat anti-rabbit Ig alkaline phosphate conjugate (Sigma, St. Louis, MO), diluted 1:lOOO in 50 mM pH 7.4 Tris, was added to each well and incubated 45 min a t 25 "C. Following three additional washes with wash buffer, 125 pL of freshly prepared pNPP substrate (p-nitrophenol phosphate tablets, Sigma, St. Louis, MO) (5 mg tablets dissolved in 1.0 M diethanolaminel0.5 mM MgC12/0.02% NaNs to a concentration of 1 mg/mL, pH 9.8) was added to each well and allowed to incubate a t 25 "C in the dark for 10 min. The colorimetric reaction was quantified by visible absorption 405 nm (BioRad Model 3550 Microplate Reader, Hercules, CA).
n
-I,
50 - 300 nm
I- ,
-1
5 -10 nmI-,
Figure 5. Schematic representation of the structure of the -300 nm diamond-cellobiose-MAP antigen delivery system, consisting of 5- 10 nm diamond particles coated with adsorbed layers of cellobiose and MAP. The construct is held together by van der Waals interactions and hydrogen bonds.
treated polystyrene) but substantially less avidly to MAP immobilized on the very hydrophobic (PDMS-treated) surface. The latter is known to alter protein conformations (9, 17). These data provide evidence for the conformational specificity of the elicited antibodies. The diamond-cellobiose-MAP antigen delivery system is illustrated conceptually in Figure 5. We believe that the effectiveness of using saccharide-coated diamond particles as antigen delivery vehicles is attributable to the nature of the cellobiose matrix and the proteincellobiose interactions at the surface of the particles. Because cellobiose binds water (18),the sugar is in a hydrated glassy state that, when dried onto the diamond particles, lacks long range order (19). This macroscopic thermodynamic description is supported by the HRTEM images (Figure 2). The sugar bonds to the diamond surface because the high interfacial energy of diamond is reduced by the adsorbed film, thus lowering the free energy of the system (12).Hydrogen bonding within the glassy sugar matrix confers three-dimensional stabililty and retards dissolution. The driving forces for protein adsorption are the formation of protein-cellobiose hydrogen bonds and the enthalpically-favorable release of water from the surfaces of the cellobiose and protein to the bulk aqueous phase (20). Because the protein is adsorbed, rather than embedded in the matrix (as with conventional adjuvants), most of the protein surface can probably retain its mobility, remaining hydrated and accessible to antibodies. These are essential aspects for strong immunological reactivity; the most potent antigenic regions of proteins are characterized by high mobility and low packing density, which allow localized conformational rearrangements and induced fit to the antibody (21). Because of our success in raising antibodies to MAP using modified diamond as the antigen carrier, we have been able to develop a new, antibody-based purification method for MAP (22). More generally, the immunogenicity of MAP demonstrated here suggests the need for further immunological studies of the utility of MAP as a potential surgical adhesive. Through a combination of materials science, surface chemistry, and biology, we have demonstrated the efficacy of a new organically-modified ceramic antigen delivery vehicle. The coated nanoparticulate carrier concept is not limited to proteinaceous antigens and is, therefore, likely to prove widely applicable to problems in antigen delivery, molecular immobilization, and mo-
Communications
lecular recognition. The same design principles are currently being applied in our laboratories to the development of new antiviral drug delivery systems and acoustic sensor-based assays. ACKNOWLEDGMENT
This work was supported by a research contract from Structured Biologicals, Inc. (N.K.), by a grant from the National Science Foundation Divisions of Chemistry, Materials Research, and Biological Sciences (R.L.G.) (CHE-9204081), and by the Office of Naval Research/ Naval Research Laboratory core program (G.M.C.). We thank Bobbette Frye and the Waters Corporation for use of an SC-1011 column for the HPLC analysis of cellobiose. The views expressed here are those of the authors, and do not represent those of the U. S. Navy, Army, or Department of Defense. LITERATURE CITED (1) Burnens, A., Demotz, S.,Corradin, G., Binz, H., and Bosshard, H. R. (1987) Epitope mapping by chemical modification of free and antibody-bound protein antigen. Science 235, 780. Jemmerson, R., and Paterson, Y. (1986) Mapping epitopes on a protein antigen by the proteolysis of antigenantibody complexes. Science 232, 1001. (2) Schultz, P. G., and Lerner, R. A. (1993) Antibody catalysis of difficult chemical transformations. Acc. Chem. Res. 26,391. Burton, D. R. (1993) Monoclonal antibodies from combinatorial 1ibraries.Acc. Chem. Res. 26,405. Lerner, R. A,, Benkovic, S.J., and Schultz, P. G. (1991) At the crossroads of chemistry and immunology: catalytic antibodies. Science 252, 659. Berzofsky, J. (1985) Intrinsic and extrinsic factors in protein antigenic structure. Science 229, 932 . (3) Wise, D. L., Ed. (1989) Applied Biosensors Butterworths, Boston, MA. Scheller, F., and Schubert, F. (1989)Biosensoren, Akademie-Verlag, Berlin. Vanderlaan, M., Stanker, L. H., Watkins, B. E., and Roberts, D. W., Eds. (1991) Immunoassays for Trace Chemical Analysis, ACS Symposium Series 451, American Chemical Society, Washington, DC. (4) Rini, J. M., Schelze-Gahmen, U., and Wilson, I. A. (1992) Structural evidence for induced fit as a mechanism for antibody-antigen recognition. Science 255,959. Darst, S.A., Robertson, C. R., and Berzofsky, J. A. (1988) Adsorption of the protein antigen myoglobin affects the binding of conformation-specific monoclonal antibodies. Biophys. J. 5, 533. ( 5 ) Benedict, C., Waite, J. H. (1986) Location and analysis of byssal structural proteins of Mytilus edulis. J . Morphol. 189, 171. Saez, C., Pardo, J., Gutierrez, E., Brito, M., and Burzio, L. 0. (1991) Immunological studies of the polyphenolic proteins of mussels. Comp. Biochem. Physiol. 98B, 569. ( 6 ) Waite, J. H. (1991) Mussel beards: A coming of age. Chem. Znd. 607. Waite, J. H. (1990) Marine adhesive proteins: Natural composite thermosets. Int. J . Biol. Macrom. 2, 139. Waite, J. H. (1987) Nature’s underwater adhesive specialist. Int. J . Adhes. Adhes. 7, 9. (7) Fulkerson J. P., Norton, L. A., Gronowicz, G., Picciano, P., Massicotte, J. M., and Nissen, C. W. (1990) Attachment of epiphyseal cartilage cells and 17/28 rat osteosarcoma osteoblasts using mussel adhesive protein. J . Orthop. Res. 8 , 793. Pitman, M. I., Menche, D., Song, E. K., Ben-Yishay, A., Gilbert, D., and Grande, D. A. (1989) The use of adhesives in chondrocyte transplantation surgery: In-vivo studies. Bull. Hosp. Joint Dis. Orthop. Znst. 49, 213. Robin, J. B., Picciano, P., Kusleika, R. S., Salazar, J., and Benedict, C. (1988) Preliminary evaluation of the use of mussel adhesive protein in experimental epikeratoplasty. Arch. Opthalmol. 106, 973.
Bioconjugafe Chem., Vol. 6,No. 5, 1995 511 (8) Unanue, E. R. (1992) Antigen Processing. Encyclopedia of Immunology (I. M. Roitt, and P. J. Delves, Eds.) pp 116-118, Academic Press, New York. Goodlick, L., and Braun, J. (1994) Revenge of the microbes: superantigens of the T and B cell lineage. Am. J . Path. 144, 623. Harding, C. V., and Unanue, E. R. (1990) Cellular mechanisms of antigen processing and the function of class I and I1 major histocompatibility complex molecules. Cell Regulat. I , 499. Brodsky, F. (1991) The cell biology of antigen processing and presentation. Ann. Rev. Zmmunol. 9, 707. (9) Sela, M. (1992) Antigens. Encyclopedia of Immunology (I. M. Roitt, and P. J. Delves, Eds.) pp 128-133, Academic Press, New York: Sela, M. (1989) Antigenicity: Some molecular aspects. Science 166, 1365. Kossovsky, N., and Freiman, C. (1994) Silicone breast implant pathology: Clinical data and immunological consequences. Arch. Path. Lab. Med. 118,686. Chedid, L. (1987) Synthetic Vaccines (R. Arnon, Ed.) Vol. 1, pp 93-103, CRC Press, Boca Raton, FL. (10) Kossovsky, N., Gelman, A., Sponsler, E., and Millett, D. (1991) Nanocrystalline Epstein-Barr virus decoys. J . Appl. Biomat. 2, 251. Kossovsky, N., and Bunshah, R. F. U.S. Pat. 5,178,882. (11) Kossovsky, N., Nguyen A., Sukiassians, K., Festekjian, A., Gelman, A., and Sponsler E. (1994) Secondary structure of albumin acquired rapidly by modified conventional ATR-FTIR is comparable to CD spectral data. J . Colloid Interface Sci. 166, 350. Kossovsky, N., and Bunshah, R. F. U.S. Pat. 5,219,477. (12) Herring, C. (1951) Some theorems on the free energies of crystal surfaces. Phys. Rev. 82, 87. Harkins,W. D. (1942) Energy relations of the surface of solids: I. Surface energy of the diamond J . Chem. Phys. 10, 268. (13) Waite, J. H., Housley, T. J., and Tanzer, M. L. (1985) Peptide repeats in a mussel glue protein: Theme and variations. Biochemistry 24, 5010. (14) Retzinger, G. S., Meredith, S.C., Takayam, K., Hunter, R. L., and Kezdy, F. J. (1981) The role of surface in the biological activities of trehalose 6,6’-dimycolate. J . Biol. Chem. 256, 8208. (15) Waite, J. H. Redox-Active Amino Acids in Biology. Methods in Enzymology (J. P. Klinman, Ed.) Academic Press, New York (in press). (16) Woodle, M. C., Collins, L. R., Sponsler, E., Kossovsky, N., Papahadjopoulos, D., and Martin, F. J. (1992) Sterically stabilized liposomes-reduction in electrophoretic mobility but not electrostatic surface potential. Biophys. J . 61, 902. Tatsumi, N., Tsuda, I., Masaoka, M., and Imai, K. (1992) Measurement of the zeta potential of human platelets by the used of laser-light scattering. Thromb. Res. 65, 585. (17) Kossovsky, N., Zeidler, M., Chun, G., Nugyen,A., Rajguru, S., Papasian, N., Gelman, A,, and Sponsler E. (1993) Surface dependent antigens identified by high binding avidity of serum antibodies in a subpopulation of patients with breast prostheses. J . Appl. Biomat. 4, 281. (18) Hardy, B. J., and Sarko, A. (1993) Molecular dynamics simulation of cellobiose in water. J . Comput. Chem. 14, 848. (19) Green, J. L., and Angell, C. A. (1989) Phase relations and vitrification in saccharide-water solutions and the trehalose anomaly. J. Phys. Chem. 93, 2880. (20) Lemieux, R. U. (1993) How proteins recognize and bind oligosaccharides. Carbohydrate Antigens ACS Symposium Series 519, (P.J. Garegg, A. A. Lindberg, Eds.), Chapter 2, p 5 , American Chemical Society, Washington, DC. (21) Geysen, H. M., Tainer, J. A,, Rodda, S. J., Mason, T. J., Alexander, H., Getzoff E. D., and Lerner, R. A. (1987) Chemistry of antibody binding to a protein. Science 235, 1184. (22) Garrell, R. L. et al. Unpublished results. BC9500608
