Applied Microbiology and Biotechnology

, Volume 62, Issue 2, pp 226–232

Display of a functional hetero-oligomeric catalytic antibody on the yeast cell surface

Authors

  • Y. Lin
    • Laboratory of Applied Biological Chemistry, Department of Synthetic Chemistry and Biological Chemistry, Graduate School of EngineeringKyoto University
    • Department of BiotechnologySouth China University of Technology
  • T. Tsumuraya
    • Biomolecular Engineering Research Institute and Protein Engineering Research Institute
  • T. Wakabayashi
    • Biomolecular Engineering Research Institute and Protein Engineering Research Institute
  • S. Shiraga
    • Laboratory of Applied Biological Chemistry, Department of Synthetic Chemistry and Biological Chemistry, Graduate School of EngineeringKyoto University
  • I. Fujii
    • Biomolecular Engineering Research Institute and Protein Engineering Research Institute
  • A. Kondo
    • Department of Chemical Science and Engineering, Faculty of EngineeringKobe University
    • Laboratory of Applied Biological Chemistry, Department of Synthetic Chemistry and Biological Chemistry, Graduate School of EngineeringKyoto University
Original Paper

DOI: 10.1007/s00253-003-1283-x

Cite this article as:
Lin, Y., Tsumuraya, T., Wakabayashi, T. et al. Appl Microbiol Biotechnol (2003) 62: 226. doi:10.1007/s00253-003-1283-x

Abstract

A functional hetero-oligomeric protein was, for the first time, displayed on the yeast cell surface. A hetero-oligomeric Fab fragment of the catalytic antibody 6D9 can hydrolyze a non-bioactive chloramphenicol monoester derivative to produce chloramphenicol. The gene encoding the light chain of the Fab fragment of 6D9 was expressed with the tandemly-linked C-terminal half of α-agglutinin. At the same time, the gene encoding the Fd fragment of the heavy chain of the Fab fragment was expressed as a secretion protein. The combined Fab fragment displayed and associated on the yeast cell surface had an intermolecular disulfide linkage between the light and heavy chains. This protein fragment catalyzed the hydrolysis of a chloramphenicol monoester derivative and exhibited high stability in binding with a transition-state analog (TSA). The catalytic reaction was also inhibited by the TSA. The successful display of a functional hetero-oligomeric catalytic antibody provides a useful model for the display of hetero-oligomeric proteins and enzymes.

Introduction

The catalytic antibody 6D9 (Miyashita et al. 1997) catalyzes the hydrolysis of a non-bioactive chloramphenicol monoester derivative to produce chloramphenicol (Fig. 1). Antibody 6D9 was one of the antibodies prepared by immunization of a haptenic phosphonate, the transition-state analog (TSA) (Pollack et al. 1986; Tramontano et al. 1986; Takahashi et al. 2001), exhibiting higher binding affinity with TSA and lower catalytic activity than those of the natural enzyme. Phage-display technology was used to evolve the catalytic antibody in vitro on the basis of the evolutionary dynamics of enzymes (Takahashi et al. 2001). A combinatorial library of 6D9 light-chain CDR1 (L-CDR1) was prepared in the pCom3 system (Fujii et al. 1998) and the libraries displayed on the surface of phage particles. The library of phage clones was screened for binding to the hapten, which was attached to a solid phase. Positive mutants were then constructed and expressed in Escherichia coli MC1061. Prior to further activity assays, it was necessary to purify the protein from the culture supernatant. Single mutations can dramatically influence the expression/yield of correctly folded antibody fragment in E. coli (Ito et al. 1993; Duenas et al. 1995; Knappik and Pluckthum 1995; Ulrich et al. 1995), which may account for the failure of mutagenesis to improve antibody properties.
Fig. 1.

6D9 antibody-catalyzed prodrug-activation and catalytic reaction, and chemically synthesized compounds. The hydrolysis of the chloramphenicol monoester derivative (1) was catalyzed by the catalytic antibody 6D9, which was generated by immunization of a hapten, the transition-state analog (TSA) (3), to produce the bioactive chloramphenicol (2). Substrate (4) was used for the assay of catalytic reaction. Fluorescein isothiocyanate (FITC)-conjugated TSA (5) was chemically synthesized via (6) and (7), and (8) from (3)

In contrast to phage display, yeast surface display by fusion with a cell wall protein allows the displayed protein to be measured directly for soluble ligand-binding kinetics and biological activity in the displayed format (Ueda and Tanaka 2000a; Boder and Wittrup 2001; Zou et al. 2002). As yeast possesses typical eukaryote-specific post-translational modification mechanisms such as proteolytic processing, folding, glycosylation, efficient disulfide isomerization and secretory machinery homologous to mammalian cells, yeast surface display is well suited to expressing eukaryotic proteins, and has been developed to express several single-chain antibodies (scFv) (Keike et al. 1997, 1999; Boder et al. 2000). In addition, libraries have been constructed by yeast display and screened directly by flow cytometry.

The scFv fragment of 6D9 presented lower catalytic activity than Fab in our previous research (Miyashita et al. 1997). It is reported that the C-terminal extension of the light chain and Fd of the heavy chain do not hamper hetero-dimerization and secretion of L:Fd hetero-dimer in mammalian cells (Schoonjans et al. 2000). The secretion of the Fab fragment of a functional mouse-human chimeric antibody has been further confirmed in yeast (Horwitz et al. 1988). In the present study, the Fab fragment of 6D9 was constructed with a hetero-oligomeric form and displayed on the yeast cell surface by combination with the secretion system. The results show that a functional catalytic antibody formed as a hetero-oligomer can be successfully and effectively displayed on the yeast cell surface.

Materials and methods

Construction of the 6D9 Fab gene

DNA fragments encoding Lc and Fd genes were amplified from pComb3-6D9 (Fujii et al. 1998) by PCR. The primers used were Lc-f 5′-CATGGCCGCGGGGTGATGACCCAGACTCC-3′, Lc-r 5′-CTCGACCATGGGGCGAGCCACCGCCACCACACTCATTCCTGTTGAAGCTCTTGAC-3′, Fd-f 5′-CCAGGCCGCGGTGCTTGAATCTGGGGGAG-3′ and Fd-r 5′-GAGCCGGTACCACCTTAAGTACAATCCCTGGGCAC-3′. Lc was fused with a gene encoding the 320 amino acids of the C-terminal half of α-agglutinin using a linker encoding (Gly)4Ser at the 3′-terminus of the α-agglutinin-encoding gene, and placed downstream of the secretion signal sequence of glucoamylase under the control of the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) promoter (Murai et al. 1998). The fusion gene was inserted into a SacI-KpnI section of the yeast integration vector pRS406 (Sikorski and Hieter 1989) to construct pIRS-Lc, carrying the URA3 selectable marker. Fd was placed downstream of the glucoamylase secretion signal sequence controlled by the same promoter, and then inserted into a SacI-KpnI section of the yeast integration vector pRS404 (Sikorski and Hieter 1989) to obtain pICAS-Fd, carrying the TRP1 selectable marker. The two plasmids constructed were digested with ApaI and XbaI, respectively, and co-transformed into Saccharomyces cerevisiae strain MT8-1 for chromosomal integration (Ito et al. 1983).

A DNA fragment encoding the tandemly aligned Vl (variable region of light chain) and Vh (variable region of heavy chain) fusion gene (Miyashita et al. 1997) was inserted into the pICAS1 vector (Murai et al. 1998). The primers used were Vh-f 5′-CCAGGCCGCGGTGCTTGAATCTGGGGAG-3′,Vh-r 5′-GTCGTCCATGGTGAGGAGACGGTGACCGA-3′, Vl-f 5′-ACTCCCATGGGGTGGCGGTGGCTCGGAACTCGTGATGACCCAGACTCC-3′ and Vl-r 5′-TGGTGCTCGAGAGCCCGTTTGATTTCCAAC-3′. The resulting plasmid, designated pICAS-scFv69, was transformed into MT8-1 to display the scFv of 6D9 on the yeast cell surface.

PCR analysis of the integrated gene

The genomic DNA of recombinant yeast cells was isolated (Adam et al. 1997) and used as the template for PCR. PCR (32 cycles) was performed in a 50 μl reaction mixture with 250 μM dNTPs, 100 nM primers, 5 μl 10× PCR buffer (Applied Biosystems, Foster City, Calif.), and 1 U AmpliTaq DNA polymerase. Aliquots (20 μl) of the reaction were analyzed by agarose gel electrophoresis (1%).

Display of 6D9 Fab fragment on the yeast cell surface

Cells were grown in 10 ml SD medium with shaking at 30°C for ~36 h to OD600=1.5 to check Fab display and binding affinity with TSA. Cells were precultivated in 5 ml SD medium at 30°C for ~24 h to OD600=1.0 and transferred to 100 ml SD medium for large-scale cultivation at 30°C for ~48 h to check catalytic activity.

Immunofluorescence assay

Cells were harvested and washed with 25 mM PBS buffer (pH 7.4) containing 0.1% bovine serum albumin (BSA), and labeled with anti-mouse IgG (Fab-specific) biotin conjugate antibody (1:300) (Sigma, Saint Louis, Mo.) as the primary antibody and ExtrAvidin-fluorescein isothiocyanate (FITC) (1:200) (Sigma) as the second reagent at room temperature. Reaction time was 1 h in each case. After reaction, the cells were washed with PBS buffer and observed by fluorescence microscopy (BH-BFL, Olympus Optical, Japan).

Hydrolytic activity assay

The catalytic activity of Fab was examined by hydrolysis of the substrate (Fig. 1) in 170 μl cells (OD600=560) in 50 mM Tris (pH 8.0) at 30°C for 20 min. The reaction was initiated by adding 10 μl Tris buffer and 20 μl of a stock solution (2 mM) of substrate in dimethyl sulfoxide to 170 μl Fab-displaying cells (OD600=560). Hydrolysis rates were measured by HPLC with injection of 10 μl of the reaction solution. Analytical HPLC was performed on a Waters 600 unit equipped with a Waters 490 multi-wavelength detector using a YMC ODS A303 column (YMC, Kyoto, Japan) eluted with CH3CN/0.1% aqueous trifluoroacetic acid (TFA) at a flow-rate of 1.0 ml/min with detection at 278 nm. The observed rate was corrected using the uncatalyzed rate of hydrolysis in MT8-1/pICAS control cells without antibody. Inhibition of the catalytic reaction was performed in the same system, substituting 10 μl TSA stock solution (4 mM) for the 10 μl Tris buffer.

Assay of binding affinity

After cultivation, cells harboring pIRS-Lc and pICAS-Fd simultaneously or separately were washed with PBS buffer containing 0.1% BSA. The cell pellet (OD600=15) was added to 100 μl FITC-conjugated TSA (10 μg/ml in 20 mM PBS, pH 7.4), and incubated for 1 h at room temperature. Immunofluorescence-labeled yeasts were observed by fluorescence microscopy. Relative fluorescence units (RFU) were measured with a Fluoroskan Ascent fluorometer (Labsystems, Helsinki, Finland). The binding of displayed Fab to TSA was monitored in a competition format. Cells displaying 6D9 Fab were incubated for 1 h at room temperature in the presence of FITC-labeled TSA at various concentrations. Data from measurements of RFU by fluorometer are the average of three separate experiments.

Chemical synthesis of FITC-conjugated TSA

FITC-conjugated TSA (5) (Fig. 1) is synthesized via 6, 7, and 8 from 3 as follows.

Compound 6

Ethyl isocyanatoacetate (45 μl, 0.40 mmol) and DBU (1,8-diazabicyclo[5,4,0]undec-7-ene) (60 μl, 0.40 mmol) were added to a stirred solution of (3) (39.4 mg, 0.067 mmol) in acetonitrile (2 ml) at room temperature,. After stirring for 2 h, the reaction mixture was acidified with TFA and purified by HPLC (YMC AM-323: C-18 reverse-phase column, ∅10 mm ×250 mm, acetonitrile/0.1% aqueous TFA=30:70 linearly changed to 50:50 at 10 min, thereafter kept at 50:50, 3.0 ml/min, 254 nm, retention time 17.0 min). Acetonitrile and TFA were removed in vacuo, and the water was removed by lyophilization to give (6) (14.1 mg, 29%). 1H NMR (300 MHz, CD3OD): δ 8.11 (d, J=8.7 Hz, 2H), 7.52–7.48 (m, 4H), 7.28 (dd, J=2.2, 8.6 Hz, 2H), 6.21 (s, 1H), 5.67 (dd, J=3.7, 9.4 Hz, 1H), 4.46 (m, 1H), 4.27–4.07 (m, 4H), 3.86, 3.79 (ABq, J=17.9 Hz, 2H), 3.20 (m, 2H), 1.25 (t, J=7.2 Hz, 3H); 13C NMR (75 MHz, CD3OD): δ 171.8, 166.5, 158.4, 156.7 (JCF=37 Hz), 149.1, 146.6, 136.3 (JCP=3.5 Hz), 131.5 (JCP=3.6 Hz), 128.9, 124.3, 122.1 (JCP=2.4 Hz), 117.4 (JCF=287 Hz), 76.0 (JCP=6.4 Hz), 67.3, 64.5, 62.3, 55.6 (JCP=5.3 Hz), 43.4, 34.9 (JCP=136 Hz), 14.5; HRMS (ESI+): calcd for C25H25O11N435Cl2F3PNa2 [M+2Na-H]+ 761.0382; obsd 761.0397.

Compound 7

Two hundred microliters 1 N NaOH was added to a stirred solution of 6 (16.7 mg, 0.023 mmol) in acetonitrile (1 ml) at 0°C. After stirring for 1 h, the reaction mixture was acidified with TFA and purified by HPLC (YMC AM-323: C-18 reverse-phase column, ∅10 mm ×250 mm, acetonitrile/0.1% aqueous TFA=30:70 linearly changed to 50:50 at 10 min, thereafter kept at 50:50, 3.0 ml/min, 254 nm, retention time 13.5 min). The acetonitrile and TFA were removed in vacuo, and the water was removed by lyophilization to give (7) (12.9 mg, 80%). 1H NMR (300 MHz, CD3OD): δ 8.12 (d, J=8.7 Hz, 2H), 7.53–7.49 (m, 4H), 7.29 (dd, J=2.4, 8.6 Hz, 2H), 6.21 (s, 1H), 5.69 (dd, J=3.8, 9.4 Hz, 1H), 4.467 (m, 1H), 4.24–4.07 (m, 2H), 3.85, 3.78 (ABq, J=17.9 Hz, 2H), 3.23 (m, 2H); 13C NMR (75 MHz, CD3OD): δ 173.4, 166.6, 158.4, 149.2, 146.2, 136.5, 131.6 (JCP=6.9 Hz), 130.9 (JCP=9.4 Hz), 128.9, 124.4, 122.2 (JCP=2.9 Hz), 76.3 (JCP=6.7 Hz), 67.3, 64.3, 55.4, 43.1, 34.6 (JCP=137 Hz); HRMS (ESI+): calcd for C23H21O11N435Cl2F3PNa2 [M+2Na-H]+ 733.0069; obsd 733.0081.

Compound 8

1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (15.4 mg, 0.080 mmol) was added to a stirred solution of (7) (24.2 mg, 0.035 mmol), t-butyl N-(6-aminohexyl)carbamate (10.8 mg, 0.050 mmol) and N-hydroxysuccinimide (5.2 mg, 0.045 mmol) in acetonitrile (2 ml) at room temperature. After stirring for 3 h, water (0.2 ml) and TFA (1 ml) were added to the reaction mixture. After stirring overnight, the reaction mixture was concentrated in vacuo, and the obtained residue was purified by HPLC (YMC AM-323: C-18 reverse-phase column, ∅10 mm ×250 mm, acetonitrile/0.1% aqueous TFA=30:70 linearly changed to 50:50 at 10 min, thereafter kept at 50:50, 3.0 ml/min, 254 nm, retention time 11.3 min). Acetonitrile and TFA were removed in vacuo, and water by lyophilization to give (8) (20.9 mg, 66%). 1H NMR (300 MHz, CD3OD): δ 8.11 (d, J=8.6 Hz, 2H), 7.51–7.43 (m, 4H), 7.24 (d, J=6.9 Hz, 2H), 6.23 (s, 1H), 5.63 (d, J=8.0 Hz, 1H), 4.45–4.34 (m, 2H), 4.07 (m, 1H), 3.79, 3.62 (ABq, J=17.1 Hz, 2H), 3.21 (t, J=6.6 Hz, 2H), 3.07 (m, 2H), 2.88 (t, J=7.5 Hz, 2H), 1.65–1.50 (m, 4H), 1.37–1.35 (m, 4H); 13C NMR (75 MHz, CD3OD): δ 172.2, 166.8, 158.4, 156.7 (JCF=37 Hz), 149.0, 147.2, 136.1 (JCP=3.4 Hz), 132.5 (JCP=8.5 Hz), 131.4 (JCP=6.6 Hz), 128.8, 124.3, 122.0, 117.4 (JCF=287 Hz), 75.5 (JCP=6.3 Hz), 67.3, 65.3, 55.6, 45.1, 40.5, 40.0, 35.6 (JCP=135 Hz), 30.0, 28.4, 27.0, 26.8; HRMS (ESI+): calcd for C29H36O10N635Cl2F3PNa [M+Na]+ 809.1457; obsd 809.1484.

Compound 5

Triethylamine (10 μl) was added to a stirred solution of (8) (8.8 mg, 0.0098 mmol) and FITC (5.2 mg, 0.013 mmol) in acetonitrile (1 ml) and N,N-dimethylformamide (0.5 ml) at room temperature. After stirring for 30 min, the reaction mixture was acidified with TFA and purified by HPLC (YMC AM-323: C-18 reverse-phase column, ∅10 mm×250 mm, acetonitrile/0.1% aqueous TFA=30:70 linearly changed to 50:50 at 10 min, thereafter kept at 50:50, 3.0 ml/min, 254 nm, retention time 18.7 min). Acetonitrile and TFA were removed in vacuo, and water by lyophilization to give (5) (9.0 mg, 78%). 1H NMR (300 MHz, CD3OD): δ 8.27 (d, J=1.4 Hz, 1H), 8.09 (d, J=8.7 Hz, 2H), 7.84 (dd, J=1.7, 8.2 Hz, 1H), 7.51–7.45 (m, 4H), 7.27–7.20 (m, 3H), 7.01 (d, J=8.9 Hz, 2H), 6.90 (d, J=2.2 Hz, 2H), 6.77 (dd, J=2.2, 8.9 Hz, 2H), 6.22 (s, 1H), 5.67 (dd, J=2.8, 9.6 Hz, 1H), 4.47 (m, 1H), 4.35 (dd, J=5.3, 11.0 Hz, 1H), 4.02 (dd, J=8.0, 11.0 Hz, 1H), 3.82–3.61 (m, 4H), 3.24–3.09 (m, 4H), 1.65–1.29 (m, 8H); 13C NMR (75 MHz, CD3OD): δ 172.1, 170.0, 166.8, 165.2, 158.4, 156.4, 149.0, 146.8, 143.0, 136.3 (JCP=3.8 Hz), 131.7, 131.6, 131.5, 130.1, 128.83, 128.79, 127.6, 124.3, 122.0, 117.4 (JCF=287 Hz), 116.2, 113.9, 103.5, 75.7, 67.3, 65.1, 55.5, 45.4, 45.1, 40.2, 35.2 (JCP=136 Hz), 30.2, 29.8, 27.5, 27.4; HRMS (ESI+): calcd for C50H47O15N735Cl2F3PSNa [M+Na]+ 1198.1815; obsd 1198.1838.

Evaluation of disulfide linkage in Fab

Cells displaying Fab were washed twice with 0.15 M NaCl and shaken for 1 h at 4°C in 25 mM PBS (pH 7.4) containing different concentrations of dithiothreitol (DTT). Then, 100 μl of the cells was added to the FITC-conjugated TSA solution (final concentration 10 g/ml) and incubated for 1 h. The cells were washed and the RFU measured by fluorometer.

Results

Display of 6D9 Fab on the yeast cell surface

Using a yeast cell surface engineering system (Murai et al. 1997; Ueda and Tanaka 2000a, 2000b), the 6D9 light chain (Lc) fused to the C-terminal half of α-agglutinin with a (Gly)4Ser linker at the C-terminal end is anchored on the yeast cell wall; the Fd fragment (Fd) of the heavy chain is assembled to the light chain via a disulfide linkage, allowing the 6D9 Fab fragment to be displayed on the yeast cell surface (Fig. 2A, B). Two plasmids, pIRS-Lc and pICAS-Fd, were constructed and co-transformed into S. cerevisiae MT8-1 to obtain strain MT8-1-Lc-Fd. Strains individually transformed with pIRS-Lc and pICAS-Fd were named MT8-1-Lc and MT8-1-Fd. Integration of the Lc and 3′-half of the α-agglutinin-encoding gene (1,974 bp) and Fd-encoding gene (663 bp) into the yeast genome was confirmed by PCR analysis using genomic DNA as the template and antibody gene-specific primers (Fig. 2C). Cells were grown to the initial stationary phase in SD medium. MT8-1-Lc-Fd was immunofluorescently labeled with anti-mouse IgG-Fab antibody; control MT8-1 cells harboring pICAS1 were unlabeled (Fig. 2D). The results confirmed that MT8-1-Lc-Fd displayed the Fab fragment of the catalytic antibody 6D9 on the cell surface.
Fig. 2A–D.

6D9 Fab display vector. A The 6D9 Lc-encoding gene was inserted in the multi-cloning site of the yeast cell surface display vector together with the 3′-half of α-agglutinin (Murai et al. 1997). The 6D9 Fd fragment-encoding gene was placed downstream of the secretion signal sequence. The two constructed genes were co-expressed in yeast. B The 6D9 Fab fragment, consisting of the Fd and Lc fragments fused with α-agglutinin as illustrated, can be displayed on the yeast cell surface. C Integration of the constructed genes was confirmed by PCR. Lanes: 1 Molecular size marker (1 kb ladder), 2 Lc and α-agglutinin-encoding fusion gene integrated in chromosome of MT8-1-Lc, 3 Fd-encoding gene integrated in chromosome of MT8-1-Fd, 4 Lc and α-agglutinin-encoding fusion gene integrated in chromosome of MT8-1-Lc-Fd, 5 Fd-encoding gene integrated in chromosome of MT8-1-Lc-Fd. 6, 7 results from MT8-1 control cells using two pairs of primers in each case. D Labeling of the 6D9 Fab fragment displayed on the yeast cell surface was performed with an anti-mouse IgG (Fab specific) biotin-conjugate antibody as the primary epitope tag and the FITC-conjugated avidin as the secondary reagent. a, b Immunofluorescence micrographs; c, d phase-contrast micrographs. Cells displaying 6D9 Fab fragments are shown in panels a and c, and control cells displaying α-agglutinin only in panels b and d

Binding affinity of 6D9 Fab fragment on the yeast cell surface

To confirm the correct folding and function of the catalytic antibody, a direct and competitive binding assay with the TSA must be performed. TSA (as shown in Fig. 1) has a trifluoroacetyl group buried in the antigen-combining site, where it functions as an important epitope for the overall binding affinity (Fujii et al. 1995). TSA was used for measuring the binding affinity. The cell with surface-displayed Fab bound to FITC-conjugated TSA, but MT8-1 cells harboring the control plasmid pICAS1 did not; both MT8-1-Lc and MT8-1-Fd cells showed almost no binding (Fig. 3A), with only a low level of the RFU detected (Fig. 3B). A single chain Fv fragment of 6D9 was also constructed on the yeast surface and its binding affinity examined: almost none was detected (Table 1). In a competition assay against unlabeled TSA, 50% binding inhibition (IC50) was achieved at a concentration of 7.1 μM (5.9 μM in the case of the purified 6D9 Fab as reported by Miyashita et al. 1994) (Fig. 4A), and complete inhibition at approximately 200 μM. The results demonstrated that the 6D9 Fab displayed on the yeast surface forms a correct antigen-binding cavity via disulfide linkage between the light chain and Fd fragment, and possesses high binding affinity with TSA (Kd=1.41×105 M−1; Kd (purified 6D9 Fab)=1.7×105 M−1) (Miyashita et al. 1994).
Fig. 3.

A Fluorescence imaging of catalytic antibody fragment on cell surface bound with FITC-labeled TSA. Cells were grown in 10 ml SD medium for 24 h (OD600=1.1). Fab fragment (Fab), Fd fragment (Fd), light chain (Lc) and 3′-half of α-agglutinin (Agg) as control were displayed on the respective cell surfaces. Binding with FITC-labeled TSA was observed under fluorescence microscopy. B Binding affinity was measured using a fluorometer and represented as relative fluorescence units (RFU) on the cell surface using a fluorometer. (cells detected in 1 ml PBS buffer pH 7.4, OD600=10)

Table 1.

Comparison of activity between scFv and Fab displayed on yeast surface. Catalytic reaction was performed with 170 μl cells (OD600=560, see Materials and methods). RFU Relative fluorescence units

Antibody fragment on yeast

RFUa

Product (μM)

Productivity (μM/OD600)

scFv

0.16

NDb

ND

Fab

1.78

1.72

1.08×10−3

aDetermined in 1 ml cells with fluorometer, OD600≈10

bNot detected

Fig. 4A, B.

Effect of inhibitor TSA on binding and catalytic reaction. A Cells were incubated with various concentrations of unlabeled TSA before addition of FITC-labeled TSA (10 μg/ml). Inhibition of binding was measured by a fluorometer. IC50 is the hapten concentration required for 50% inhibition of maximal binding in reaction; Kd=1/IC50. B Fab displayed on cells was added to TSA (200 μM) prior to catalytic reaction for substrate (4) (◯), and compared with catalytic reaction in inhibitor-free mixture (●) and control cells (◆). All data are mean of three experiments

Treatment of cells displaying Fab with DTT prior to the binding reaction (Zou et al. 2000) decreased the formation of 6D9 Fab on the cell surface, as well as binding to FITC-labeled TSA , through reduction of the specific disulfide linkage between the light chain and Fd fragment (Fig. 5). Upon treatment of cells displaying Fab with 100 mM DTT, the binding affinity of the antibody was reduced by over 50%. The incomplete loss of binding affinity may be explained by a structural hindrance to dissociation of the two chains at the cell wall. Fab is already formed inside the cell before being transported to the cell surface; Fab bound to FITC-labeled TSA can be observed within the cell by fluorescence microscopy (data not shown).
Fig. 5.

Effect of dithiothreitol (DTT) on binding affinity of Fab displayed on yeast cell surface. Cells (OD600=5.4) were treated with various concentrations of DTT in PBS (pH 7.4) for 1 h at 4°C. The binding affinity of Fab is expressed in RFU

Catalytic activity of the 6D9 Fab fragment on the yeast cell surface

The hydrolytic activity of 6D9 Fab on the cell surface for substrate (4) and the competitive inhibition of TSA were examined. Equal amounts of cells displaying 6D9 Fab were added to 200 μM substrate or substrate plus inhibitor. The antibody-catalyzed reactions were performed at 30°C in a system containing 10% DMSO/50 mM Tris (pH 8.0), and followed by HPLC (see Materials and methods). The antibody-catalyzed reaction was completely inhibited by the addition of 200 μM TSA to the reaction mixture (Fig. 4B). The results show that 6D9 scFv has no activity (Table 1), and a hetero-dimeric form was active on the yeast surface. It was concluded that the hetero-dimer displaying Fab had high stability at 30°C in antibody-catalyzed reactions.

Discussion

When a gene encoding a heterologous protein is inserted between sequences encoding the N-terminal secretion signal sequence and the C-terminal half of α-agglutinin containing the glycosylphosphatidylinositol anchor attachment signal sequence in yeast, a fusion protein can be expressed and anchored on the cell wall in an active form (Murai et al. 1997; Ueda and Tanaka 2000a; 2000b). This cell surface engineering system is very useful for displaying monomeric proteins. In this study, the combination of such a yeast cell surface display system and an extra-cellular secretion system allowed a hetero-oligomeric protein of the Fab fragment of catalytic antibody 6D9 to be displayed on the yeast cell surface. Our hetero-dimer displaying Fab had high stability and reactivity in antibody-catalyzed reactions. The incomplete loss of binding affinity seen in Fig. 5 might be due to a structural hindrance to dissociation of the two chains at the cell wall. The Fd fragment of the heavy chain fused to the C-terminal half of α-agglutinin could not be successfully displayed due to intra-cellular cleavage by an unidentified protease (data not shown).

The successful display of a functional hetero-oligomeric catalytic antibody will provide a useful means not only to display hetero-oligomeric proteins, but also to construct a combinatorial library of catalytic antibodies on the yeast cell surface as whole-cell biocatalysts. This system allows a larger protein library to be constructed and attractive clones to be easily selected from a mutagenized protein pool using a fluorescently-labeled hapten. It can also be expected to allow construction of a combinatorial library of catalytic antibodies with simultaneous mutations on double (heavy and light) chains and generation of novel catalytic antibodies on the yeast cell surface.

Copyright information

© Springer-Verlag 2003