JBC Biosymposia, Inc.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Tang, P. M.
Right arrow Articles by Schueler, P. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tang, P. M.
Right arrow Articles by Schueler, P. A.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Volume 270, Number 14, Issue of April 7, 1995 pp. 7829-7835
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
A High Affinity Digoxin-binding Protein Displayed on M13 Is Functionally Identical to the Native Protein (*)

(Received for publication, September 16, 1994; and in revised form, December 16, 1994)

Pauline M. Tang,  Lisa A. Foltz,  Walter C. Mahoney,  Paula A. Schueler (§)

From the From Molecular Diagnostics, Research and Development, Boehringer Mannheim Corporation, Indianapolis, Indiana 46250

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES

ABSTRACT

Phage display of peptides and proteins has successfully been employed to produce binding molecules of altered affinity. Little is known, however, regarding the impact on affinity measurements of phage-displayed molecules compared to their native freely soluble configuration. That identical affinities can be obtained was shown by Scatchard analysis of the native antibody, its single chain derivative (scFv), and its phage-displayed single chain counterpart for the ligand digoxin. No significant difference, within one standard deviation, was detected in affinity for digoxin when the phage-displayed scFv was compared to either its soluble scFv form or the purified antibody. In addition, no change in binding specificity was detected, within two standard deviations, when the binding proteins were challenged with two commonly cross-reactive compounds (dihydrodigoxin and digitoxin). That phage-display can be employed for molecules having high binding affinities (K of 6 times 10M) is also shown.

INTRODUCTION

A wide variety of genes including those encoding human growth hormone(1) , Escherichia coli alkaline phosphatase(2) , rat anionic trypsin(3) , a library of protease substrates(4) , elastase inhibitors(5) , portions of the human IgE receptor(6) , Staphylococcus aureus Protein A(7) , a library of zinc finger DNA-binding proteins(8) , ricin B chain(9) , and a myriad of peptide, antibody variable domain, and Fab (^1)libraries (10, 11, 12) have been genetically fused to a gene encoding one of the coat proteins of the filamentous bacteriophage (fd or M13) and presented on the phage surface for selection. The phage containing the genetic material for the molecule displayed on its surface can be recovered following the specific selection for appropriate binding or enzymatic activity. Phage display is a powerful methodology that is capable of speeding the search and isolation of candidate binding elements by tying together phenotypic positive selection with ready access to the specific coding DNA sequence. In the case of antibodies, phage-displayed libraries of antibody fragments can provide a way to bypass hybridoma technology and even animal immunization (for review, see (13) ), by tapping directly into the vast multitude of germline sequences. For phage display to be successfully exploited, however, it is essential that the molecule displayed be presented in a manner identical to that of the native molecule. Antibody heavy and light chain variable regions chosen randomly from a library of such fragments and displayed on phage as a single chain antibody (scFv) are selected based on the binding activity of the displayed scFv. The scFv-binding site must be presented in such a way that it accurately represents the ligand binding site of the intact antibody, both with respect to affinity for the ligand and in the fine specificity of the binding site. If not, the selection of the scFv using phage display could result in the choice of a binding protein without the desired characteristics when the scFv is produced in a soluble form, no longer attached to the phage surface. Numerous cases of affinity differences between soluble scFv and their respective whole antibody or Fab counterparts(14, 15, 16, 17, 18, 19, 20, 21, 22) have been reported. In many cases these differences can be attributed to the additive impact on affinity measurements of two binding sites versus the presence of a single site, in antibodies and scFv, respectively. Additionally, alterations in protein folding of these molecules have also been invoked to account for the observed differences; however, there have been few experiments to support this contention.

The first phage-displayed scFv (11) which was directed against hen egg white lysozyme showed specificity by a lack of binding to turkey egg white lysozyme in an enzyme-linked immunosorbant assay (ELISA); however, no affinity measurements were made on the displayed scFv. Garrard and coworkers (23) attempted to separate phage-displayed wild type Fab fragments (K = 3.3 nM) of the humanized 4D5 antibody from three lower affinity variants (K between 0.05-1.1 µM). Interestingly, efficient selection of the higher affinity Fab phage did not correlate strictly with the affinity of the variants.

Both rat anionic trypsin (3) and bacterial (E. coli) alkaline phosphatase (2) have been displayed as phage enzymes (phage-zymes). In each example, while the catalytic activity of the phage-zyme was unaltered, both proteins demonstrated a reduced affinity for their respective substrates. Since alkaline phosphatase is a dimer in its native conformation, not only is correct folding critical but correct dimerization as well. This additional constraint hampers the interpretation of the alkaline phosphatase results, however, since trypsin is a simpler protein; the difference encountered between native and phage-displayed enzyme may be primarily due to the effect of being phage-bound.

Human growth hormone (hGH) gene III fusion protein was shown to be folded correctly on phage by reactivity with a series of six conformationally sensitive hGH monoclonal antibodies(1) . Mutational variants of hGH were phage displayed and selected for high affinity binding to hGH receptor coated on beads. The affinity of these variant hormones for the receptor was determined after release from the phage surface, and in some cases hGH variants with lower affinities were isolated more frequently than higher affinity variants(24) . Using a similar approach, phage-displayed I-domain derived from leukocyte integrins was shown to bind ligand in an ELISA and react with a panel of conformationally sensitive monoclonal antibodies(25) .

The widespread use of phage display for selection of peptide/protein fragments makes essential a controlled comparison of a binding element expressed on phage to the same binding element in its native environment. At equilibrium under identical conditions, we have compared the binding characteristics of a well-defined antibody directed against the hapten digoxin, its scFv counterpart, and this same scFv molecule displayed as a gene III fusion on phage. Digoxin, a cardiac glycoside hapten of approximate molecular dimensions 31 times 8 times 9 Å(26) , is very similar in size to that of an antibody combining site (34 times 12 times 7 Å; 27, 28). The rigid, noncharged structure of digoxin's steroid backbone does not allow for major conformational changes of the hapten even when functional groups are substituted at various positions on the multiring structure. The crystal structure (at 2.7Å) of the high affinity anti-progesterone Fab` DB3 complexed with the steroid progesterone showed the ligand (and various progesterone analogs) to be almost completely buried within the very hydrophobic binding pocket(29, 30) . The total buried surface area of the steroid and that of the antibody binding pocket was very close, 241 and 270 Å^2, respectively. This close size approximation and rigidity of the steroid nucleus make possible the examination of the topographical features of the antibody binding site by analyzing both fine specificity (with structural analogs) and affinity for the primary ligand.

EXPERIMENTAL PROCEDURES

Hybridoma Cells and Antibody Purification

The murine anti-digoxin hybridoma DigV11C12 was obtained from Cindy Vistica (Microgenics, Concord, CA) and was generated from digoxin-(3)carboxymethyloxime-keyhole limpet hemocyanin-immunized spleen cells from BALB/c mice fused with the myeloma line 8.653. Antibody was purified from ascites by passing through a protein A column (Repligen, Cambridge, MA) and eluting with 100 mM citric acid and 200 mM (NH(4))(2)SO(4) at pH 3.0. Neutralized fractions were pooled and dialyzed against 10 mM potassium phosphate and 150 mM NaCl buffer at pH 7.4. The purified antibody was lyophilized and stored at -20 °C.

mRNA Isolation and Preparation of scFv Phagemid

mRNA was isolated from a hybridoma culture (IgG1, kappa) containing approximately 3 times 10^7 cells using the Fast Track mRNA isolation kit (Invitrogen, San Diego, CA). cDNA was prepared from the isolated mRNA using GeneAmp RNA PCR kit (Perkin-Elmer) and a Perkin Elmer Cetus DNA thermal cycler 480. The kappa light chain and the Fd region of heavy chain were amplified by polymerase chain reaction (PCR) using the following primers synthesized on an Applied Biosystems 380B DNA synthesizer (Foster City, CA): 5`-GGGAATTCTGAGGTGA(C)AGCTGCAGGAGTCTG-3`, 5`-CCGCTCGAGTCAAATTTTCTTGTCCACCTTGGTGCT-3`, 5`-GGGAATTCCGAT(C)ATTGTGC(A)TGACACAA(G)TCTCAA-3`, 5`-CCGCTCGAGTCAACACTCATTCCTGTTGAAGCA(T)CTT-3`, and purified using reversed phase high performance liquid chromatography. The crude oligo mixture was loaded onto a column (Brownlee Aquapore-Butyl; 25 times 10 mm) equilibrated with 12% acetonitrile in 100 mM triethylammonium acetate, pH 7. Elution was accomplished with a linear gradient over 5 min from 12 to 17% acetonitrile in 100 mM triethylammonium acetate, pH 7, followed by a fast ramp to 50% acetonitrile in 100 mM triethylammonium acetate, pH 7. Purified oligos usually eluted approximately 11 min into the run and were detected at 280 nm.

The PCR-amplified heavy or light chain fragment were ligated to the Uni-ZAP XR vector (Stratagene, La Jolla, CA) at EcoRI-XhoI sites. After phage packing and in vivo excision, the pBluescript phagemids containing the inserts were isolated and several clones sequenced. To construct the scFv, the primer mixture of the Recombinant Phage Antibody System from Pharmacia LKB Biotechnol was used to amplify the V(H) and V regions. The carboxyl-terminal of V(H) was linked to the amino-terminal of V by a 15-amino-acid linker, (Gly(4)Ser)(3). A SfiI and a NotI site were introduced by PCR to the 5` and 3` ends of the scFv fragment, respectively. The resulting scFv fragment was subsequently ligated to a pCANTAB 5 (Pharmacia) or a pCANTAB-5 his6 c-myc phagemid (Dr. Greg Winter, MRC, Cambridge, United Kingdom). This latter phagemid contains a 6-histidine tag and a c-myc tag followed by an amber stop codon prior to the gene III protein. The phagemid containing the scFv was transformed into E. coli TG1 cells (supE hsdDelta5 thi Delta(lac-proAB) F` [traD36proABlacI^qlacZDeltaM15]. These pCANTAB-5 vectors allow the expression (under the lac promoter) and transport to the periplasm of the scFv fused to the gene III protein (g3p) of M13. Upon rescue with a helper phage that carries the rest of the M13 structural and replication proteins, the phagemid is packaged as a recombinant M13 phage displaying on its surface one or more copies of the antibody scFv fusion-g3p along with the native g3p. The procedures of rescuing phagemid with helper phage M13K07 and the production of scFv phage antibodies were done as specified in the Pharmacia kit. Phagemid infectivity titers were based on colony forming units (cfu) selected on ampicillin containing plates (100 µg/ml).

Enrichment of Digoxin-binding Phage

The digoxin-binding phage antibodies were first selected from Digoxigenin-BSA (25:1)-coated polystyrene T-flasks. The scFv phage were eluted by a 10-min incubation with 100 mM triethylamine (Aldrich) at pH 12.1, and the neutralized phages were reinfected into TG1 cells. Individual ampicillin-resistant colonies were picked and grown in 100 µl of 2 times YT medium containing ampicillin and 2% glucose in a 96-well microtiter plate overnight at 30 °C. A replica of this 96-well plate was rescued with helper phage, and the phagemid-containing supernatants were tested for digoxin reactivity by ELISA using Digoxin-BSA-coated plates and horseradish peroxidase-conjugated sheep anti-M13 IgG for detection. Each supernatant was screened for concomitant binding to BSA alone. Clone 3H was picked for two more rounds of enrichment on Digoxin-BSA (10:1, York Biological, Stony Brook, NY) covalently linked to carboxylated magnetic particles (PerSeptive Diagnostics Cambridge, MA). The triethylamine-eluted scFv phage were reinfected to TG1 cells and cloned a second time on a 96-well microtiter plate. One clone, 3H-3H, showing a strong positive signal for digoxin in ELISA, was chosen for large scale phage preparation using a helper phage multiplicity of infection of 15 and a growth temperature of 30 °C to increase phage yield. Phage antibodies prepared from 400 ml of cell culture were precipitated by one-fifth volume of 20% polyethylene glycol 8000 and 2.5 M NaCl (Sigma) and resuspended in 8 ml of phosphate-buffered saline (PBS, 2.7 mM KCl, 1.2 mM KH(2)PO(4), 138 mM NaCl, and 8 mM Na(2)HPO(4).7H(2)O, pH 7.1, Life Technologies, Inc.). Phagemid particles concentrated using this method routinely showed a titer of approximately 1 times 10 cfu/ml.

Preparation of Soluble 3H-3H-scFv

In the pCANTAB-5 his6 c-myc phagemid, an amber codon located between the cloned scFv sequence and the g3p sequence allows soluble scFv to be produced. In TG1 cells (a supressor strain, SupE), soluble scFv can be produced by driving the lac promotor with a small amount of isopropylthio-beta-galactoside (IPTG), thus overcoming the amber supressor and allowing scFv to be transported to the periplasmic space, and upon extended incubation, to leak into the medium. The clone 3H-3H was grown in 400 ml of 2 times YT medium containing 100 µg/ml ampicillin and 0.1% glucose to an OD of 0.9; IPTG was added to a final concentration of 1 mM and allowed to grow an additional 22 h at 22 °C. Bacteria were removed by centrifugation for 15 min at 6000 revolutions/min in a Sorvall GSA (Dupont) rotor. The supernatant was then subjected to two PEG precipitations as described previously to remove any residual phage. Finally, the pH of the scFv-containing supernatant was adjusted to 7.8 and added to 3 ml of a 50% slurry Ni-NTA resin (Sepharose CL-6B) (Qiagen, Chatsworth, CA, catalog no. 30230) pre-equilibrated in binding buffer (50 mM sodium phosphate, pH 7.7, 200 mM NaCl) per Qiagen instructions. The mixture was gently rotated for 20 h at 4 °C, spun at 400 times g in a Sorvall RT6000B at 4 °C for 5 min, and the supernatant discarded. The pellet was resuspended in 400 ml of wash buffer (50 mM sodium phosphate, pH 6.0, 150 mM NaCl) and rotated for 45 min at 4 °C, spun, and supernatant discarded. The pellet was resuspended in 400 ml of final buffer (50 mM sodium phosphate, pH 8.0, 150 mM NaCl, 0.5% bovine serum albumin, 0.1% Tween 20) and rotated for 45 min at 4 °C, spun at 800 times g for 5 min, and supernatant discarded. The scFv-resin pellet was resuspended in final buffer to a volume of 10 ml and immediately tested for digoxin binding activity.

DNA Sequencing

The sequence of scFv clone 3H-3H and those of the original cDNA clones of heavy and light chains were confirmed by sequencing from both directions utilizing oligonucleotide primers by the dideoxy chain terminating method(31) . Sequences were repeated between two and three times for verification. Either Sequenase (United States Biochemicals Corp., Cleveland, OH), the Tth DNA polymerase of the Cycle Sequencing Kit (Pharmacia), or the PRISM Ready Reaction DyeDeoxy Terminator Cycle Sequencing Kit (Applied Biosystems, Perkin Elmer) were used according to the kit manuals. Automated sequencing was performed using a model 373A-Stretch DNA sequencer from Applied Biosystems (Divison of Perkin-Elmer Corporation).

ELISA

High binding microtiter plates (Costar, Cambridge, MA) were coated with (200 µl/well) either Digoxin-BSA (10:1) or BSA alone at 10 µg/ml and allowed to incubate at 4 °C overnight. Plates were washed with 0.1% Tween 20/PBS (Pierce) three times. Phage samples were diluted serially (1-10^5, varying with different samples) with 2% nonfat dry milk (Bio-Rad) in PBS. 195 µl of each sample were added to washed plate wells and allowed to incubate at 34 °C for 2 h. M13mp18 bacteriophage (Pharmacia's Detection Module no. 27-9402-01) was used as a positive control. Plates were washed with 0.1% Tween/PBS and then incubated with horseradish peroxidase-conjugated sheep anti-M13 IgG (1:5000) for 1 h at room temperature. Substrate 2`,2`-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) diammonium was then added to the washed plate wells, and absorbance was measured at 405 nm using a microtiter plate reader (Molecular Devices Corp., Menlo Park, CA).

Affinity Measurement

A competitive radioimmunoassay was set up using either 3H-3H scFv-resin, PEG-concentrated 3H-3H scFv phage, or purified antibody as the binding partner. The same human serum-based calibrators and mass of I-digoxin tracer were used regardless of the binding partner assayed except that the method of separation of bound tracer differed. The assay buffer varied only for the scFv in that the Tris had to be replaced with phosphate since a tertiary amine can reduce the Ni and affect the scFv binding to the resin. Lyophilized purified antibody was reconstituted in 50 mM sodium phosphate buffer, pH 7.4, containing 0.1% carrier BSA to a give a final concentration of 1 mg/ml. Dilutions were made from this stock with assay buffer (100 mM Tris-Cl, pH 8.0, 150 mM NaCl, 0.5% BSA, 0.1% Tween 20). The assay consisted of 25 µl of assay buffer, 50 µl of diluted antibody, 25 µl of the appropriate digoxin calibrator, and 50 µl of I-digoxin (approximately 64,000 counts/min, 2,500 µCi/µg, Monobind, Costa Mesa, CA) in a 12 times 75-mm polypropylene tube or 0.5-ml microfuge tube. The mixture was vortexed and incubated overnight (at least 16 h) at 4 °C with constant gentle agitation. Human serum-based digoxin calibrators were obtained from MonoBind, Costa Mesa, CA and values assigned independently with the Baxter Stratus II Digoxin assay (0, 0.2, 0.4, 0.8, 1.9, 3.7, and 6 ng/ml). An additional solution of digoxin in the zero digoxin serum calibrator (500 ng/ml) was prepared to measure nonspecific binding. Bound I-digoxin was separated from free using BioMag particles (no. 8-4349D, Perseptive Diagnostics) coupled with goat anti-mouse IgG (H&) for the purified antibody, PEG precipitation for the scFv phage, and Ni-NTA resin for the scFv. Fifty µg of buffer washed BioMag particles were added to each equilibrated assay tube and allowed to incubate 2 h with gentle rocking at 4 °C. The particles were separated in a Corning magnetic separation unit, washed, and counted in a Packard B5002 Cobra Auto-gamma counter (Downers Grove, IL).

The final assay volume for the scFv and the scFv phage was adjusted to 300 µl instead of the 150 µl used for the purifed antibody; this allowed a more efficient separation of bound from free I-digoxin. Both were incubated overnight at 4 °C with constant gentle agitation as specified above. The agitation was that sufficient to keep the scFv resin evenly suspended throughout the incubation period. The scFv phage antibody was precipitated with one-fifth volume of 20% polyethylene glycol and 2.5 M NaCl on ice for 2 h, and then spun at top speed (2,200 times g) in a Sorvall RT 6000B (DuPont) for 1 h. The scFv assay was done in a 0.5-ml microfuge tube, and separation was accomplished by spinning at top speed (16,000 times g) in an Eppendorf 5415C at 4 °C for 45 min. The radioactivity in the pellets was counted in a gamma counter as specified. Data were fitted using a four-parameter logistic curve(32) . Varying dilutions of purified antibody, scFv, and scFv phage were tested, and affinities were determined and compared at a dilution of each that gave an ED (effective dose at 50%) close to 1 ng/ml. The association constants were determined according to Scatchard (33) , and assays were routinely done in triplicate.

Specificity Determination

Stock solutions of two digoxin analogs, digitoxin (Sigma) and dihydrodigoxin (Boehringer Mannheim GmbH, Mannheim), were made to 0.1 mg/ml in absolute ethanol and further diluted in the zero digoxin serum calibrator to 3.3, 4, 5, 10, 20, 40, and 100 ng/ml for digitoxin and 18.3, 55, 110, 220, and 1100 ng/ml for dihydrodigoxin. The binding specificities were determined by the 50% displacement method(34) , in which, the dose of digoxin (D) and cross-reactant (CR) necessary to displace 50% of the bound labeled tracer are compared in the following calculation: cross-reactivity % = D/CR times 100.

RESULTS

Assembly of Digoxin scFv and Enrichment of the Digoxin-specific scFv Phage

Amplified V(H) and V portions were joined by a 15 amino acid linker using Pharmacia's Recombinant Phage Antibody System, ligated into the pCANTAB-5 his6 c-myc phagemid and transformed into TG1 cells. Rescued recombinant phage antibodies were enriched on digoxigenin-BSA-coated polystyrene T-flasks. Those recombinant phage antibodies that bound were reinfected into TG1 cells and individual colonies picked for rescue. Approximately one-third of the colonies tested (30 out of 96 clones) showed a digoxin-specific positive reaction in ELISA (data not shown). Recombinant phage antibodies from well 3H showed a high rate of substrate turnover and were selected for further enrichment on digoxin-BSA covalently linked to magnetic particles. In order to increase the titer of digoxin binding phage to allow us to set up I-labeled digoxin binding assays two additional rounds of enrichment and reinfection into TG1 cells were done. The resulting [3H]phagemid PEG concentrate had a titer of 1.2 times 10 cfu/ml and gave positive ELISA signal at a dilution of greater than 10^4. These phagemids were highly enriched for digoxin binding as evidenced by another round of individual colony rescue followed by a digoxin-specific ELISA where all the wells tested were reactive with digoxin. One of these recombinant antibody phages designated 3H-3H was chosen to study the high affinity digoxin-binding site displayed on phage.

Binding Affinity and Specificity

Competition radioimmunoassay dose-response curves were generated for multiple dilutions of the purified digoxin antibody, the 3H-3H scFv phage antibody, and the 3H-3H scFv. Dilutions of each that gave an ED close to 1 ng/ml of digoxin and similar slopes at this point were chosen for comparison of the binding sites. Fig. 1(A-C) shows representative standard curves for 150 µl of a 10% solution of 3H-3H scFv resin, approximately 10 cfu of 3H-3H scFv phage antibody, and 2.5 ng of the purified whole antibody. Fig. 1(D-F) presents the Scatchard analysis of the radioimmunoassay affinity determinations from the standard curves shown in A-C, respectively. The slope of the line is the affinity constant (K(a)) and is 1.18 times 10M, 1.16 times 10M, and 2.22 times 10M for the scFv, the scFv phage, and the purified antibody, respectively. The mean and standard deviation of multiple affinity determinations for these digoxin binding partners are given in Table 1. The original hybridoma cell supernatant also showed a similar affinity for digoxin when tested under comparable assay conditions (data not shown).

Figure 1: Panels A-C, representative competitive I-digoxin binding curves for the 3H-3H scFv, the 3H-3H scFv phage, and purified anti-digoxin monoclonal antibody DIGVIIC12, respectively. Radioimmunoassay was as described under ``Experimental Procedures.'' Panels D-F, Scatchard analysis of the binding data from the 3H-3H scFv shown in panel A, the 3H-3H scFv phage shown in panel B, and the purified antibody shown in panel C, respectively.


To access the fine specificity of the binding site, multiple concentrations of two digoxin analogs with only minor differences in their structure were allowed to compete with the I-labeled digoxin under the same assay conditions described under ``Experimental Procedures.'' The percentage cross-reactivity was calculated by solving the equation for the line drawn between the actual concentrations of analog and their digoxin doses read from the standard curve. The percentage cross-reactivity was determined at the 50% inflection point of the digoxin standard curve. The structures of digoxin and the two analogs digitoxin and dihydrodigoxin are given in Fig. 2. Digitoxin is identical to digoxin except it is missing only one hydroxyl group at carbon-12 of the steroid backbone, while in dihydrodigoxin only the double bond between carbons 20-22 of the lactone ring becomes reduced. Table 1gives the mean and standard deviation of multiple cross-reactivity determinations for the purified antibody, the 3H-3H scFv displayed on phage, and the 3H-3H scFv using these two digoxin analogs.

Figure 2: Structures of digoxin (digoxigenin tridigitoxose), dihydrodigoxin, and digitoxin.


DNA Sequencing

The 3H-3H scFv is 720 bp and using the SfiI and NotI sites was cloned into the pCANTAB-5 his6 c-myc phagemid vector between the gene3/pelB signal sequence and the DNA sequences encoding the 6-histidines, c-myc tag, amber codon, and gene 3 structural sequence. From the 5` end, the scFv consists of 354 bp coding for the V(H) region followed by 45 bp encoding the (Gly(4)Ser)(3) linker and 321 bp encoding the Vkappa region. The DNA sequence of the 3H-3H scFv was determined for comparison to the DNA sequence of the variable regions of the kappa light chain and Fd regions originally cloned from the DIGVIIC12 monoclonal antibody cell line (Fig. 3).

Figure 3: DNA sequence of 3H-3H scFv. Sequencing was done as described under ``Experimental Procedures.'' The V(H) region extends from nucleotide 1 to 354, the linker from 355 to 399 (underlined), and the V from nucleotide 400 to 720. The nucleotide sequence of the originally cloned V(H) and V from the DIGVIIC12 monoclonal cell line is the same as the scFv except where designated below the DNA sequence. The inferred amino acid sequence of the scFv is also given; with the differences in the originally cloned variable regions given below (in parenthesis) for comparison. The nucleotide change at position 380 in the scFv sequence probably resulted from misincorporation by the AmpliTaq polymerase and led to a glycine to aspartic acid substitution in the linker; all other changes were due to degeneracy in the primer mixes used in the scFv construction.


Within the V(H) and Vkappa regions of the 3H-3H scFv, there were 17 nucleotide differences from the sequence of the V(H) and Vkappa counterparts from the original clones. All 17 of these discrepancies were caused by the family-specific oligonucleotide primer mixes used in the construction of the scFv. The primers used to construct the scFv according to the Pharmacia kit are a family-specific mixture designed to amplify all murine variable regions. These differences resulted in two amino acid substitutions in the framework 1 region (positions 1 and 3) and two substitutions in the framework 4 region (positions 113 and 114) of the V(H). The Vkappa region of the 3H-3H scFv shows three substitutions in the framework 1 region only (positions 3, 4, and 8). The proline at position 8, however, is probably correct and present in the original cell line but was introduced by the primer (which coded for glutamine) used in the original PCR cloning. Proline is found with high frequency in this position in all seven mouse kappa chain subgroups(35) . Only Kabat subgroup V shows 16 out of 324 sequences with glutamine at this position. An additional base pair change, probably due to misincorporation by the AmpliTaq polymerase, was found in the 3H-3H scFv sequence and resulted in an amino acid substitution. This occurred in the linker region between the V(H) and Vkappa regions causing the coded linker sequence to become GGGGSGGGDSGGGGS (an A instead of a G at base no. 380 causing amino acid change from glycine to aspartic acid).

DISCUSSION

The present results demonstrate that the antibody combining site of the 3H-3H anti-digoxin scFv displayed on the head of M13 phage is indistinguishable in its hapten binding characteristics from that of the entire antibody molecule (IgG(1), kappa) and the same scFv expressed without phage. This conclusion is based on the following observations. First, multiple affinity determinations showed an average affinity constant of 1.6 ± 0.34 times 10 liters/mol for the phage displayed binding site, which is not significantly different at one standard deviation from the affinity constant determined both for the scFv (1.2 times 10 liters/mol) and for the purified whole antibody (2.1 ± 0.35 times 10 liters/mol) studied under the same conditions. This high affinity for digoxin, a hapten that itself approximates the antibody combining site in size, suggests extensive complementarity (multiple contacts) between the binding site and the hapten. Therefore, any alterations of the uncharged steroid nucleus should help discriminate these two sites. Second, the reactivity of these binding sites for the two digoxin analogs, dihydrodigoxin and digitoxin, also did not differ significantly at two standard deviations. These two analogs differ from digoxin only slightly, either the lack of the hydroxyl at carbon-12 or saturation of the double bond in the lactone ring, yet the conformation of the binding site remains as discriminatory on the phage as it is in the native structure.

An IgG antibody has two antigen-binding sites/molecule. Since phage have five copies of the gene III protein, scFv-gene III fusions could result in five scFvs displayed per phage. Due to tight control on the promotor, statistically it is more likely that two or less of these gene III proteins are actually displaying scFvs/phage molecule(36) . The affinity constants and analog competitiveness were determined from assays wherein the binding reaction was allowed to reach equilibrium. This was done in order to eliminate any complication of multiple binding partners displayed on a single phage. Antibody selection of phage displayed random peptides or ``epitope libraries'' even under equilibrium binding conditions (37, 38) have shown that the effect of multiple binding partners on a single phage complicates the selection of high affinity ligands. In these instances, all five copies of the gene III coat protein were decorated with a random peptide, making the selection of a ligand, based on affinity for a bivalent antibody, complicated by avidity effects.

Neither the presence of the linker itself, nor the unintentional aspartic acid residue within it, appears to have any detrimental effect on the binding functionality of the site. Although, in previous reports the presence of scFv linkers have been implicated in the decreased affinity of various soluble scFv as compared to whole molecule or Fab fragments(16, 17, 20, 21) . Primer-induced sequence differences found within the 3H-3H scFv itself also did not alter the binding affinity or specificity of the site. Two substitutions were found within the first three amino-terminal residues of the framework 1 region and two more in the framework 4 region of the V(H); in the Vkappa three additional changes were found in the amino-terminal region of framework 1. In the Vkappa, the proline at position 8 is most likely the residue within the native antibody since upon investigation of the mouse Vkappa protein groups, proline is the most common residue in all 7 Vkappa protein groups. Glutamine in this position is seen in only 16 of the 324 sequences catalogued in protein group V(35) . An erronous primer choice in the original PCR cloning led to the glutamine in position 8 that was confirmed in the DNA sequencing of the original cDNA clones. Reamplification of the V(H) and V(L) from these clones with the family-specific primer mix reverted the glutamine in position 8 back to a proline.

In order to do the I-digoxin binding assay with the scFv as the binding partner and to keep the same basic assay format for comparison of all the binding partners, it was necessary to concentrate the scFv-containing culture supernatant. Digoxin immunoreactive material could be detected in the culture supernatant prior to concentration only using the enzyme-amplified detection in an ELISA format (developed with anti-mouse IgG-horseradish peroxidase). Protein A was investigated as a method of concentration since by sequence comparison the 3H-3H scFv was shown to express a V(H) gene from a V(H) family, J606(35) , known to contain members that bind Protein A alternatively(39, 40) ; however, this binding proved minimal. Due to the 6-histidine tag on the scFv, Ni-NTA resin could be used for concentration and as a separation step in the assay. With the scFv bound to the surface of the resin particles, the binding activity could be titered as in solution provided the resin particles remained evenly suspended(41, 42) . This allowed a direct comparison of the 3H-3H scFv and the 3H-3H scFv phage.

The functional qualities of the 3H-3H scFv while displayed on the surface of M13 bacteriophage are native in character and importantly validate the usefulness of the phage display technology for selection of functional molecules from complex mixtures such as libraries. Critical to the selection of compounds from a library displayed on phage is that the qualities chosen during the selection protocols be maintained in the final product. The data presented herein illustrate that phage display of an antibody can provide a protein that mirrors the native antibody with respect to affinity and specificity.

FOOTNOTES

*
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked ``advertisement'' in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank(TM)/EMBL Data Bank with accession number(s) U20617[GenBank].

§
To whom correspondence should be addressed: Molecular Diagnostics, R & D, Boehringer Mannheim Corp., 9115 Hague Rd., Indianapolis, IN 46250. Tel.: 317-576-3448 Fax: 317-576-4426.

(^1)
The abbreviations used are: Fab`, antigen binding fragment; scFv, single chain variable fragment of an antibody; hGH, human growth hormone; ELISA, enzyme-linked immunosorbant assay; V(H), variable region of heavy chain; V(L), variable region of light chain; V, variable region of kappa light chain; Fd, variable region and first constant region of heavy chain; g3p, gene III minor coat protein; PCR, polymerase chain reaction; cfu, colony forming unit; PBS, phosphate buffered saline; PEG, polyethylene glycol; IPTG, isopropylthio-beta-galactoside; BSA, bovine serum albumin; bp, base pair.

ACKNOWLEDGEMENTS

We thank Scott Davidson for synthesis and purification of the oligonucleotide primers, Cindy Vistica for kindly providing the DIGVIIC12 cell line, Dr. Greg Winter for the pCANTAB-5 his6 c-myc phagemid vector, and Dr. Scott Eisenbeis for reviewing the manuscript. We also acknowledge the support of Drs. J. William Freytag and Albert A. Luderer.

REFERENCES

  1. Bass, S., Greene, R., and Wells, J. A. (1990) Proteins Struct. Funct. Genet. 8, 309-314 [CrossRef][Medline] [Order article via Infotrieve]
  2. McCafferty, J., Jackson, R. H., and Chiswell, D. J. (1991) Protein Eng. 4, 955-961 [Abstract/Free Full Text]
  3. Corey, D. R., Shiau, A. K., Yang, Q., Janowski, B. A., and Craik, C. S. (1993) Gene (Amst.) 128, 129-134 [CrossRef][Medline] [Order article via Infotrieve]
  4. Matthews, D. J., and Wells, J. A. (1993) Science 260, 1113-1117 [Abstract/Free Full Text]
  5. Roberts, B. L. Markland, W., Ley, A. C., Kent, R. B., White, D. W., Guterman, S. K., and Ladner, R. C. (1992) Proc. Natl. Acad Sci. U. S. A. 89, 2429-2433 [Abstract/Free Full Text]
  6. Robertson, M. W. (1993) J. Biol. Chem. 268, 12736-12743 [Abstract/Free Full Text]
  7. Djojonegoro, B. M., Benedik, M. J., and Willson, R. C. (1994) BioTechnology 12, 169-172 [CrossRef][Medline] [Order article via Infotrieve]
  8. Rebar, E. J., and Pabo, C. O. (1994) Science 263, 671-673 [Abstract/Free Full Text]
  9. Swimmer, C., Lehar, S. M., McCafferty, J., Chiswell, D. J., Blattler, W. A., and Guild, B. C. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 3756-3760 [Abstract/Free Full Text]
  10. Smith, G. P. (1991) Curr. Opin. Biotech. 2, 668-673
  11. McCafferty, J., Griffiths, A. D., Winter, G., and Chiswell, D. J. (1990) Nature 348, 552-554 [CrossRef][Medline] [Order article via Infotrieve]
  12. Kang, A. S., Barbas, C. F., Janda, K. D., Benkovic, S. J., and Lerner, R. A. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 4363-4366 [Abstract/Free Full Text]
  13. Winter, G., Griffiths, A. D., Hawkins, R. E., and Hoogenboom, H. R. (1994) Annu. Rev. Immunol. 12, 433-455 [Medline] [Order article via Infotrieve]
  14. Bird, R. E., Hardman, K. D., Jacobson, J. W., Johnson, S., Kaufman, B. M., Lee, S.-M., Lee, T., Pope, S. H., Riordan, G. S., and Whitlow, M. (1988) Science 242, 423-426 [Abstract/Free Full Text]
  15. Bedzyk, W. D., Weidner, K. M., Denzin, L. K., Johnson, L. S., Hardman, K. D., Pantoliano, M. W., Asel, E. D., and Voss, E. W., Jr. (1990) J. Biol. Chem. 265, 18615-18620 [Abstract/Free Full Text]
  16. Huston, J. S., Levinson, D., Mudgett-Hunter, M., Tai, M.-S., Novotny', J., Margolies, M. N., Ridge, R, J., Bruccoleri, R. E., Haber, E., Crea, R., and Oppermann, H. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 5879-5883 [Abstract/Free Full Text]
  17. Barry, M. M., and Lee, J. S. (1993) Mol. Immunol. 30, 833-840 [CrossRef][Medline] [Order article via Infotrieve]
  18. Malby, R. L., Caldwell, J. B., Gruen, L. C., Harley, V. R., Ivancic, N., Kortt, A. A., Lilley, G. G., Power, B. E., Webster, R. G., Colman, P. M., and Hudson, P. J. (1993) Proteins Struct. Funct. Genet. 16, 57-63 [CrossRef][Medline] [Order article via Infotrieve]
  19. Rumbley, C. A., Denzin, L. K., Yantz, L., Tetin, S. Y., and Voss, E. W. Jr. (1993) J. Biol. Chem. 268, 13667-13674 [Abstract/Free Full Text]
  20. Savage, P., Rowlinson-Busza, G., Verhoeyen, M., Spooner, R. A., So, A., Windust, J., Davis, P. J., and Epenetos, A. A. (1993) Br. J. Cancer 68, 738-742 [Medline] [Order article via Infotrieve]
  21. Wels, W., Harwerth, I.-M., Zwickl, M., Hardman, N., Groner, B., and Hynes, N. E. (1992) BioTechnology 10, 1128-1132 [CrossRef][Medline] [Order article via Infotrieve]
  22. Anthony, J., Near, R., Wong, S.-L., Iida, E., Ernst, E., Wittekind, M., Haber, E., and Ng, S.-C. (1992) Mol. Immunol. 29, 1237-1247 [CrossRef][Medline] [Order article via Infotrieve]
  23. Garrard, L. J., Yang, M., O'Connell, M. P., Kelley, R. F., and Henner, D. J. (1991) BioTechnology 9, 1373-1377 [CrossRef][Medline] [Order article via Infotrieve]
  24. Lowman, H. B., Bass, S. H., Simpson, N., and Wells, J. A. (1991) Biochemistry 30, 10832-10838 [CrossRef][Medline] [Order article via Infotrieve]
  25. Miceli, R. M., DeGraaf, M. E., Fairbanks, M., Goodman, T., Bajt, M. L., and Fischer, H. D. (1994) FASEB J. 8, A1408
  26. Go, K., Kartha, G., and Chen, J. P. (1980) Acta Crystallogr. B36, 1811-1819 [CrossRef]
  27. Kabat, E. A. (1966) J. Immunol. 97, 1-11 [Abstract/Free Full Text]
  28. Amzel, L. M., Poljak, R. J., Saul, F., Varga, J. M., and Richards, F. F. (1974) Proc. Natl. Acad. Sci. U. S. A. 71, 1427-1430 [Abstract/Free Full Text]
  29. Arevalo, J. H., Stura, E. A., Taussig, M. J., and Wilson, I. A. (1993) J. Mol. Biol. 231, 103-118 [CrossRef][Medline] [Order article via Infotrieve]
  30. Arevalo, J. H., Taussig, M. J., and Wilson, I. A. (1993) Nature 365, 859-863 [CrossRef][Medline] [Order article via Infotrieve]
  31. Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. Acad. Sci. U. S. A. 74, 5463-5467 [Abstract/Free Full Text]
  32. Rodbard, D., and Hutt, D. M., (1974) Proceedings Series: Radioimmunoassay and Related Procedures in Medicine , Vol. 1, pp. 165-192, International Atomic Energy Agency, Vienna
  33. Scatchard, D. (1949) Ann. N. Y. Acad. Sci. 51, 660-672 [CrossRef]
  34. Miller, J. J., and Valdes, R., Jr. (1992) J. Clin. Immun. 15, 97-107
  35. Kabat, E. A., Wu, T. T., Perry, H. M., Gottesman, K. S., and Foeller, C. (1991) Sequences of Proteins of Immunological Interest , 5th Ed., United States Department of Health and Human Services, Washington, D. C.
  36. Marks, J. D., Hoogenboom, H. R., Griffiths, A. D., and Winter, G. (1992) J. Biol. Chem. 267, 16007-16010 [Free Full Text]
  37. Cwirla, S. E., Peters, E. A., Barrett, R. W., and Dower, W. J. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 6378-6382 [Abstract/Free Full Text]
  38. Scott, J. K., and Smith, G. P. (1990) Science 249, 386-390 [Abstract/Free Full Text]
  39. Seppala, I., Kaartinen, M., Ibrahim, S., and Makela, O. (1990) J. Immunol. 145, 2989-2993 [Abstract]
  40. Ibrahim, S., Kaartinen, M., Seppala, I., Matoso-Ferreira, A., and Makela, O. (1993) Scand. J. Immunol. 37, 257-264 [CrossRef][Medline] [Order article via Infotrieve]
  41. Hechemy, K., and Michaelson, E. (1984) Lab Manage. 6, 27-40
  42. Hechemy, K., and Michaelson, E. (1984) Lab Manage. 7, 26-35
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.


Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Tang, P. M.
Right arrow Articles by Schueler, P. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tang, P. M.
Right arrow Articles by Schueler, P. A.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 1995 by the American Society for Biochemistry and Molecular Biology.