Affinity maturation of Fab antibody fragments by fluorescent-activated cell sorting of yeast-displayed libraries

We report for the first time the affinity maturation of Fab antibody fragments using fluorescent-activated cell sorting (FACS) of yeast-displayed repertoires. A single yeast display vector which enables the inducible expression of an anchored heavy chain and a soluble light chain has been constructed. The assembly and functional display on the yeast cell surface of Fab antibodies specific for different protein targets has been demonstrated by flow cytometry and immunofluorescence microscopy. We have affinity matured a Fab antibody specific for the tetravalent antigen streptavidin using FACS of yeast-displayed repertoires diversified by error-prone polymerase chain reaction. A panel of variants with up to 10.7-fold improvement in affinity was obtained after selection. Two leading clones, R2H10 (3.2 nM) and R3B1 (5.5 nM), had mutations in light chain complementarity determining region 1 LC-CDR1 (H34R) and LC-CDR3 (Y96H or Y96F) and gave a 10.7-fold and 6.3-fold affinity improvement over the starting antibody, respectively. The ability to efficiently affinity mature Fab antibodies is an important component of the antibody development pipeline and we have shown that yeast display is an efficient method for this purpose.     

Phage-display libraries of murine and human antibody Fab fragments

We provide efficient and detailed procedures for construction, expression, and screening of comprehensive libraries of murine or human antibody Fab fragments displayed on the surface of filamentous phage. In addition, protocols for producing and using ultra-electrocompetent cells, for producing Fab phages from libraries, and for selecting antigen binders by panning are presented. The latter protocol includes a procedure for trypsin elution of bound phage.     

Directed evolution of stabilized IgG1-Fc scaffolds by application of strong heat shock to libraries displayed on yeast

We have constructed IgG1-Fc scaffolds with increased thermal stability by directed evolution and yeast surface display. As a basis a new selection strategy that allowed the application of yeast surface display for screening of stabilizing mutations in proteins of already high intrinsic thermal stability and T(m)-values up to 85°C was developed. Besides library construction by error prone PCR, strong heat stress at 79°C for 10min and screening for well-folded proteins by FACS, sorting rounds had to include an efficient plasmid DNA isolation step for amplification and further transfection. We describe the successful application of this experimental setup for selection of 17 single, double and triple IgG1-Fc variants of increased thermal stability after four selection rounds. The recombinantly produced homodimeric proteins showed a wild-type-like elution profile in size exclusion chromatography as well as content of secondary structures. Moreover, the kinetics of binding of FcRn, CD16a and Protein A to the engineered Fc-molecules was very similar to the wild-type protein. These data clearly demonstrate the importance and efficacy of the presented strategy for selection of stabilizing mutations in proteins of high intrinsic stability within reasonable time.     

Directed evolution of stabilized IgG1-Fc scaffolds by application of strong heat shock to libraries displayed on yeast

We have constructed IgG1-Fc scaffolds with increased thermal stability by directed evolution and yeast surface display. As a basis a new selection strategy that allowed the application of yeast surface display for screening of stabilizing mutations in proteins of already high intrinsic thermal stability and T_m-values up to 85 °C was developed. Besides library construction by error prone PCR, strong heat stress at 79 °C for 10 min and screening for well-folded proteins by FACS, sorting rounds had to include an efficient plasmid DNA isolation step for amplification and further transfection. We describe the successful application of this experimental setup for selection of 17 single, double and triple IgG1-Fc variants of increased thermal stability after four selection rounds. The recombinantly produced homodimeric proteins showed a wild-type-like elution profile in size exclusion chromatography as well as content of secondary structures. Moreover, the kinetics of binding of FcRn, CD16a and Protein A to the engineered Fc-molecules was very similar to the wild-type protein. These data clearly demonstrate the importance and efficacy of the presented strategy for selection of stabilizing mutations in proteins of high intrinsic stability within reasonable time.     

Construction of a selective cleavage system for a protein displayed on the cell surface of yeast

We constructed a novel protein-purification system in which Saccharomyces cerevisiae with a protein displayed on the cell surface is harvested and the displayed protein is then cleaved from the cell surface. GFPuv was used as a model protein in this cell surface engineering experiment. In this system, the C-terminal 320 amino acids of α-agglutinin were bound to the C-terminal of GFPuv for display on the cell surface. In this novel system, the insertion of the recognition sequence-encoding gene of protease factor Xa between GFPuv and α-agglutinin was successfully carried out. The GFPuv, displayed by the insertion, was successfully cleaved from yeast cell surface by treatment with factor Xa, and could be easily recovered. By removing such a protease with well-known properties, the displayed protein could be isolated and purified with relative ease.     

Shuffled antibody libraries created by in vivo homologous recombination and yeast surface display

Homologous recombination in yeast can be exploited to reliably generate libraries of >10⁷ transformants from a pool of PCR products and a linearized plasmid vector. Homology in the PCR insertion products drives shuffling of these genes in vivo by yeast homologous recombination. Two scFvs that share 89.8% homology were shuffled in vivo by homologous recombination, and chimeric genes were generated regardless of whether or not one of the scFv PCR products lacked 5′ homology to the cut vector and the second scFv PCR product lacked 3′ homology to the cut vector, or both PCR products had both 5′ and 3′ homology to the cut vector. A majority of the chimeras had single crossovers; however, double and triple crossovers were isolated. Crossover points were evenly distributed in the hybrids created and homology of as little as two nucleotides was able to produce a chimeric clone. The numbers of clones isolated with a given number of crossovers was approximated well by a Poisson distribution. Transformation efficiencies for the chimeric libraries were of the order of 10⁴–10⁵ transformants per microgram of insert, which is the same order of magnitude as when a single PCR product is inserted alone into the display vector by homologous recombination. This method eliminates ligation and Escherichia coli transformation steps of previous methods for generating yeast-displayed libraries, requires fewer PCR cycles than in vitro DNA shuffling and, unlike site-specific recombination methods, allows for recombination anywhere that homology exists between the genes to be recombined. This simple technique should prove useful for protein engineering in general and antibody engineering, specifically in yeast.     

Design of a serine protease-like catalytic triad on an antibody light chain displayed on the yeast cell surface

Lc-WT, the wild-type light chain of antibody, and Lc-Triad, its double mutant with E1D and T27aS designing for the construction of catalytic triad within Asp1, Ser27a, and original His93 residues, were displayed on the cell surface of the protease-deficient yeast strain BJ2168. When each cell suspension was reacted with BODIPY FL casein and seven kinds of peptide-MCA substrates, respectively, a remarkable difference in hydrolytic activities toward Suc-GPLGP-MCA (succinyl-Gly-Pro-Leu-Gly-Pro-MCA), a substrate toward collagenase-like peptidase, was observed between the constructs: Lc-Triad-displaying cells showed higher catalytic activity than Lc-WT-displaying cells. The difference disappeared in the presence of the serine protease inhibitor diisopropylfluorophosphate, suggesting that the three amino acid residues, Ser27a, His93, and Asp1, functioned as a catalytic triad responsible for the proteolytic activity in a similar way to the anti-vasoactive intestinal peptide (VIP) antibody light chain. A serine protease-like catalytic triad (Ser, His, and Asp) is considered to be directly involved in the catalytic mechanism of the anti-VIP antibody light chain, which moderately catalyzes the hydrolysis of VIP. These results suggest the possibility of new approach for the creation of tailor-made proteases beyond limitations of the traditional immunization approach.     

Comparative thermodynamic analyses of the Fv, Fab* and Fab and Fab fragments of anti-dansyl mouse monoclonal antibody

In order to investigate the role of the constant domainson the antigen-binding property of the variable domains, we have carried out a comparative thermodynamic study of the anti-dansyl Fv, Fab* and Fab fragments that possess the identical amino acid sequence of the variable domains. The thermodynamic analyses have shown that binding constants, enthalphy changes and entropy changes are similar for the three antigen-binding fragments, whereas the thermal stability of Fab is much higher than that of Fv and Fab*. We have concluded that (i) the variable domains of the three antigen-binding fragments possess identical intrinsic capability for antigen binding and (ii) the two constant domains serve to improve the stability of the variable domains.     

Cell-free synthesis of functional and endotoxin-free antibody Fab fragments by translocation into microsomes

A eukaryotic cell-free system based on Spodoptera frugiperda cells was developed for the convenient synthesis of Fab antibody fragments and other disulfide bridge containing proteins. The system uses (i) a cell lysate that is mildly prepared under slightly reduced conditions, thus maintaining the activity of vesicles derived from the endoplasmic reticulum, (ii) signal peptide dependent translocation into these vesicles, and (iii) a redox potential based on reduced and oxidized glutathione. Monomeric heavy and light immunoglobulin chains are almost completely converted to highly active dimeric Fab joined by intermolecular disulfide bridges without supplementation of chaperones or protein disulfide isomerase. The applicability of the system is demonstrated by the synthesis of anti-lysozyme and anti-CD4 Fab antibody fragments yielding approximately 10 μg Fab per milliliter reaction mixture. The lack of endotoxins in this system is a prerequisite that synthesized Fab can be applied directly using whole synthesis reactions in cell-based assays that are sensitive to this substance class. Moreover, the system is compatible with PCR-generated linear templates enabling automated generation of antibody fragments in a high-throughput manner, and facilitating its application for screening and validation purposes.     

Development of an optimized expression system for the screening of antibody libraries displayed on the Escherichia coli surface.

Polypeptide library screening technologies are critically dependent upon the characteristics of the expression system employed. A comparative analysis of the lpp-lac, tet and araBAD promoters was performed to determine the importance of tight regulation and expression level in library screening applications. The surface display of single-chain antibody (scFv) in Escherichia coli as an Lpp-OmpA' fusion was monitored using a fluorescently tagged antigen in conjunction with flow cytometry. In contrast to the lpp-lac promoter, both tet and araBAD promoters could be tightly repressed. Tight regulation was found to be essential for preventing rapid depletion of library clones expressing functional scFv and thus for maintaining the initial library diversity. Induction with subsaturating inducer concentrations yielded mixed populations of uninduced and fully induced cells for both the tet and araBAD expression systems. In contrast, homogeneous expression levels were obtained throughout the population using saturating inducer concentrations and could be adjusted by varying the induction time and plasmid copy number. Under optimal induction conditions for the araBAD system, protein expression did not compromise either cell viability or library diversity. This expression system was used to screen a library of random scFv mutants specific for digoxigenin for clones exhibiting improved hapten dissociation kinetics. Thus, an expression system has been developed which allows library diversity to be preserved and is generally applicable to the screening of E. coli surface displayed libraries.     

Fab is the most efficient format to express functional antibodies by yeast surface display

Multiple formats are available for engineering of monoclonal antibodies (mAbs) by yeast surface display, but they do not all lead to efficient expression of functional molecules. We therefore expressed four anti-tumor necrosis factor and two anti-IpaD mAbs as single-chain variable fragment (scFv), antigen-binding fragment (Fab) or single-chain Fabs and compared their expression levels and antigen-binding efficiency. Although the scFv and scFab formats are widely used in the literature, 2 of 6 antibodies were either not or weakly expressed. In contrast, all 6 antibodies expressed as Fab revealed strong binding and high affinity, comparable to that of the soluble form. We also demonstrated that the variations in expression did not affect Fab functionality and were due to variations in light chain display and not to misfolded dimers. Our results suggest that Fab is the most versatile format for the engineering of mAbs.     

Cytochemical evaluation of localization and secretion of a heterologous enzyme displayed on yeast cell surface

A starch-utilizing Saccharomyces cerevisiae strain was constructed by cell surface engineering. Distribution of the heterologous glucoamylase-alpha-agglutinin fusion protein on the yeast cell was analyzed by indirect fluorescence microscopy using an anti-glucoamylase antibody. Most of the intense fluorescence was first localized in the small bud, then observed on the entire cell wall of the daughter and mother cells. Fluorescence also accumulated at the neck region. These observations suggest that the display of the heterologous protein on the cell surface is carried with other cell wall components to the areas in which the cell wall is newly synthesized; the distribution is controlled by the cell cycle. Then, the heterologous protein-encoding gene was expressed in a sec1 mutant, in which secretory vesicles accumulate under restrictive temperature, and the produced protein was detected by immunoelectron microscopy. Most of the gold particles that reacted with the fusion protein were not localized in vesicles but in expanding endoplasmic reticulum. This phenomenon may be due to overproduction of the heterologous protein which was designed to be displayed on the cell wall. Artificial production of heterologous protein may have caused a relative shortage of glycosyl phosphatidylinositol anchors.     

Construction of a starch-utilizing yeast by cell surface engineering.

We have engineered the cell surface of the yeast Saccharomyces cerevisiae by anchoring active glucoamylase protein on the cell wall, and we have endowed the yeast cells with the ability to utilize starch directly as the sole carbon source. The gene encoding Rhizopus oryzae glucoamylase with its secretion signal peptide was fused with the gene encoding the C-terminal half (320 amino acid residues from the C terminus) of yeast alpha-agglutinin, a protein involved in mating and covalently anchored to the cell wall. The constructed plasmid containing this fusion gene was introduced into S. cerevisiae and expressed under the control of the glyceraldehyde-3-phosphate dehydrogenase promoter from S. cerevisiae. The glucoamylase activity as not detected in the culture medium, but it was detected in the cell pellet fraction. The glucoamylase protein transferred to the soluble fraction from the cell wall fraction after glucanase treatment but not after sodium dodecyl sulfate treatment, indicating the covalent binding of the fusion protein to the cell wall. Display of the fused protein was further confirmed by immunofluorescence microscopy and immunoelectron microscopy. The transformant cells could surely grow on starch as the sole carbon source. These results showed that the glucoamylase was anchored on the cell wall and displayed as its active form. This is the first example of an application of cell surface engineering to utilize and improve the metabolic ability of cells.     

Affinity selection of target cells from cell surface displayed libraries: a novel procedure using thermo-responsive magnetic nanoparticles

Biotinylated thermo-responsive magnetic nanoparticles for use in affinity selection from yeast cell surface display libraries were prepared by coating magnetite nanoparticles with a thermo-responsive polymer consisting of N-isopropyl acrylamide and a biotin derivative. These particles showed a reversible transition between flocculation and dispersion at around the lower critical solution temperature of 30 degrees C, above which the flocculated particles--which absorbed a large amount of avidin due to their large surface area--were quickly separable by magnet. The model library was constructed by mixing control yeast cells with target yeast cells co-displaying IgG binding protein (ZZ) and enhanced green fluorescence protein. Biotinylated IgG and avidin were subsequently added to the model library, and target cells were efficiently enriched with the biotinylated magnetic nanoparticles by avidin-biotin sandwich and ZZ-IgG interaction. The few target cells (0.001%) in the model library were enriched by up to 100% in only 5 days by an affinity selection procedure repeated four times. This novel method based on magnetic nanoparticles and a yeast cell surface display system could fulfill a wide range of applications in the analysis of protein-protein interactions and rapid isolation of novel biomolecules.     

Kinetic exclusion assay of monoclonal antibody affinity to the membrane protein Roundabout 1 displayed on baculovirus

The reliable assessment of monoclonal antibody (mAb) affinity against membrane proteins in vivo is a major issue in the development of cancer therapeutics. We describe here a simple and highly sensitive method for the evaluation of mAbs against membrane proteins by means of a kinetic exclusion assay (KinExA) in combination with our previously developed membrane protein display system using budded baculovirus (BV). In our BV display system, the membrane proteins are displayed on the viral surface in their native form. The BVs on which the liver cancer antigen Roundabout 1 (Robo1) was displayed were adsorbed onto magnetic beads without fixative (BV beads). The dissociation constant (K_d, ∼10⁻¹¹ M) that was measured on the Robo1 expressed BV beads correlated well with the value from a whole cell assay (the coefficient of determination, R² = 0.998) but not with the value for the soluble extracellular domains of Robo1 (R² = 0.834). These results suggest that the BV–KinExA method described here provides a suitably accurate K_d evaluation of mAbs against proteins on the cell surface.     

Construction, affinity maturation, and biological characterization of an anti-tumor-associated glycoprotein-72 humanized antibody

The tumor-associated glycoprotein (TAG)-72 is expressed in the majority of human adenocarcinomas but is rarely expressed in most normal tissues, which makes it a potential target for the diagnosis and therapy of a variety of human cancers. Here we describe the construction, affinity maturation, and biological characterization of an anti-TAG-72 humanized antibody with minimum potential immunogenicity. The humanized antibody was constructed by grafting only the specificity-determining residues (SDRs) within the complementarity-determining regions (CDRs) onto homologous human immunoglobulin germ line segments while retaining two mouse heavy chain framework residues that support the conformation of the CDRs. The resulting humanized antibody (AKA) showed only about 2-fold lower affinity compared with the original murine monoclonal antibody CC49 and 27-fold lower reactivity to patient serum compared with the humanized antibody HuCC49 that was constructed by CDR grafting. The affinity of AKA was improved by random mutagenesis of the heavy chain CDR3 (HCDR3). The highest affinity variant (3E8) showed 22-fold higher affinity compared with AKA and retained the original epitope specificity. Mutational analysis of the HCDR3 residues revealed that the replacement of Asn(97) by isoleucine or valine was critical for the affinity maturation. The 3E8 labeled with (125)I or (131)I showed efficient tumor targeting or therapeutic effects, respectively, in athymic mice with human colon carcinoma xenografts, suggesting that 3E8 may be beneficial for the diagnosis and therapy of tumors expressing TAG-72.     

B cell selection and affinity maturation during an antibody response in the mouse with limited B cell diversity

The quasi-monoclonal mouse has limited B cell diversity, whose major (approximately 80%) B cell Ag receptors are comprised of the knockin V(H) 17.2.25 (V(H)T)-encoded H chain and the lambda1 or lambda2 L chain, thereby being specific for 4-hydroxy-3-nitrophenylacetyl. The p-nitrophenylacetyl (pNP) was found to be a low affinity analog of nitrophenylacetyl. We examined affinity maturation of anti-pNP IgG by analyzing mAbs obtained from quasi-monoclonal mice that were immunized with this low affinity Ag. The results are: 1) Although V(H)T/lambda1 and V(H)T/lambda2 IgM were equally produced, V(H)T/lambda2 IgG almost exclusively underwent affinity maturation toward pNP. 2) A common mutation in complementarity-determining region 3 of V(H)T (T313A) mainly contributed to generating the specificity for pNP. 3) Because mutated V(H)T-encoded gamma-chains could form lambda1-bearing IgG in Chinese hamster ovary cells, apparent absence of V(H)T/lambda1 anti-pNP IgG may not be due to the incompatibility between the gamma-chains and the lambda1-chain, but may be explained by the fact that V(H)T/lambda1 B cells showed 50- to 100-fold lower affinity for pNP than V(H)T/lambda2 B cells. 4) Interestingly, a pNP-specific IgM mAb that shared common mutations including T313A with high affinity anti-pNP IgG was isolated, suggesting that a part of hypermutation coupled with positive selection can occur before isotype switching. Thus, even weak B cell receptor engagement can elicit an IgM response, whereas only B cells that received signals stronger than a threshold may be committed to an affinity maturation process.     

Somatic mutation, affinity maturation and the antibody repertoire: a computer model

Somatic mutation has been implicated as a significant and possibly primary factor in the maturation of antibody affinity in the humoral immune response. B cells stimulated by antigen experience a hyper-mutation in the gene segments that code for the antigen-binding site of the antibody, creating antibody specificities that did not exist at the time of immunization. Although most of the mutations are likely to be disadvantageous, new specificities with a higher affinity for the antigen are sometimes created. These higher-affinity cells are preferentially selected for proliferation and eventual antibody secretion, resulting in a progressively higher average affinity over time. In this paper we present the results of an investigation of somatic mutation through the use of a computer model. At the basis of the model is a large repertoire of discrete antibodies and antigens, having three-dimensional structures, that exhibit properties similar to those of the real populations. The key factor is that the binding strength between any antibody/antigen pair can be calculated as a function of the complementarity of the (a) size, (b) shape and (c) functional groups that comprise the two structures. The created repertoires are imbedded in a dynamical system model of the immune response to directly evaluate the affect of somatic mutation on affinity maturation. We also present an expanded hypothesis of clonal selection and development to explain how the mutational restrictions imposed by the genetic code and the structure of the antibody repertoire, along with antigen concentration, affinity, and probabilistic factors may interact and contribute to the expansion of specific clones as the response develops over time.     

Discovery of high affinity anti-ricin antibodies by B cell receptor sequencing and by yeast display of combinatorial VH:VL libraries from immunized animals

Ricin is a toxin that could potentially be used as a bioweapon. We identified anti-ricin A chain antibodies by sequencing the antibody repertoire from immunized mice and by selecting high affinity antibodies using yeast surface display. These methods led to the isolation of multiple antibodies with high (sub-nanomolar) affinity. Interestingly, the antibodies identified by the 2 independent approaches are from the same clonal lineages, indicating for the first time that yeast surface display can identify native antibodies. The new antibodies represent well-characterized reagents for biodefense diagnostics and therapeutics development.     

Thermodynamic characterization of affinity maturation: the D1.3 antibody and a higher-affinity mutant

Understanding the structural and dynamic determinants of binding free energy in the antigen-antibody bond is of great interest. Much work has focused on selective mutations in order to locate key interaction residues, but this generally results in reduced affinity. The present work instead examines a higher-affinity mutant to characterize the thermodynamic pathway of the affinity maturation process. We have compared the antigen binding energetics of scFv D1.3, an anti-hen egg lysozyme single chain antibody, with a higher-affinity mutant (Hawkins, R. E., Russell, S. J., Baier, M. and Winter, G. (1993). J. Mol. Biol. 234, 958-964). The mutant has five-fold higher affinity for lysozyme but nearly the same enthalpy and heat capacity change upon binding, as measured by isothermal titration calorimetry. Thus, much of the binding free energy difference can be attributed to entropic effects. Fluorescence quenching with acrylamide indicates that this more favorable entropy change may result from a more flexible mutant-lysozyme complex and thus be a configurational entropy effect.