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


     

Clinical Chemistry 54: 1008-1017, 2008. First published April 10, 2008; 10.1373/clinchem.2007.097352
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
clinchem.2007.097352v1
clinchem.2007.097352v2
54/6/1008    most recent
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 Web of Science
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 Siegel, R. W.
Right arrow Articles by Tyner, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Siegel, R. W.
Right arrow Articles by Tyner, J.
Related Collections
Right arrow Drug Monitoring and Toxicology
(Clinical Chemistry. 2008;54:1008-1017.)
© 2008 American Association for Clinical Chemistry, Inc.

Drug Monitoring and Toxicology

Affinity Maturation of Tacrolimus Antibody for Improved Immunoassay Performance

Robert W. Siegel1,a, Wade Baugher1, Tanya Rahn1, Susan Drengler1 and Joan Tyner1

1 Core R&D, Diagnostic Division, Abbott Laboratories, Abbott Park, IL.

aAddress correspondence to this author at: Eli Lilly and Company, Lilly Research Laboratories, P.O. Box 708, Greenfield, IN 46140. Fax (317) 433-6963; e-mail siegelro{at}lilly.com.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Background: Organic solvents used for extraction of tacrolimus from whole blood samples lower the apparent affinity of the antibody used in a diagnostic immunoassay, thereby affecting the detection limit.

Methods: We used in vitro recombinant antibody engineering to screen and isolate clones from diverse libraries with mutagenic complementarity regions (CDRs) from tacrolimus 1-60-46 hybridoma cell line, with improved binding to tacrolimus in the presence of 10% methanol organic solvent solution.

Results: We isolated a number of clones with mutations in variable heavy (VH) CDR 2, variable light (VL) CDR 1, and VL CDR 3 with improved binding. Various combinatorial pairings constructed from these individual mutations contained >10-fold improvements in both the dissociation rate and overall equilibrium affinity constants. Selected clones produced as IgG have increased functional sensitivity, with a 3- to 6-fold reduction in the limit of detection relative to the parental tacrolimus 1-60-46 monoclonal antibody in the Architect® Tacrolimus immunodiagnostic assay.

Conclusions: The recent advent of recombinant in vitro antibody display technologies in general, and yeast surface display in particular, allows the flexibility to engineer new or augment specific analytical characteristics, such as affinity, specificity, or stability, into previously isolated and otherwise desirable antibodies to enhance assay performance. These in vitro selections can also be performed under conditions meant to mimic the assay in which the reagent will ultimately be used, to increase the likelihood of successful assay development.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Tacrolimus, a macrolide immunosuppressant produced by the bacterium Streptomyces tsukabaensis(1), is used for the prevention of organ rejection(2)(3) by preventing T-lymphocyte activation(4). Toxicity concerns(5) dictate patient monitoring, and a variety of different diagnostic immunoassays are commercially available(6). Tacrolimus binds to a cytosolic immunophilin protein, FK-506 binding protein(7), and the complex is located in erythrocytes. Organic solvents (e.g., methanol, ethanol, methylene chloride) used for extraction from whole blood samples before measurement often lower the functional activity of the antibody used in the immunoassay, with the potential for reduced accuracy and robustness(8).

The recent advent of in vitro antibody display technologies provides the flexibility to engineer new or augment specific attributes of interest, such as affinity, specificity, or stability, into previously isolated antibody molecules, a process often termed "directed evolution." A central tenet of any in vitro display platform is that the phenotype of the displayed protein must be coupled to the gene encoding the displayed protein. This coupling allows diverse libraries to be sampled and the coding sequence to be recovered from clones identified with desired properties. Yeast display(9) (Fig. 1A ) has proven to be a highly effective platform for various directed evolution applications, including affinity maturation(10)(11)(12) and changes in specificity(13)(14). Single-chain fragment variable (scFv)1 antibodies, in which the antigen-specific immunoglobulin variable heavy (VH) and variable light (VL) domains are linked into a single polypeptide, are displayed on the yeast surface as a fusion protein to a cell wall component. Fluorescence activated cell sorting (FACS) coupled with yeast cell surface display enables monitoring of both antibody expression on the cell surface and the ability of that antibody to bind antigen (Fig. 1B ). Discrete populations differing in affinity and/or epitope specificity can be visualized, and desired clones can be quantitatively recovered(15)(16)(17).


Figure 1
View larger version (17K):
[in this window]
[in a new window]

  		Figure 1. (A) Yeast surface display.

The scFv is displayed on the yeast cell surface as a translational fusion to Aga2 protein. Surface expression can be detected using fluorescence-labeled antibodies binding to the C-terminal epitope tag. Biotinylated antigen binding can be detected using fluorescence-labeled streptavidin. Expression of the translational fusion in transformed yeast from the pYD plasmid is under control of the inducible Gal promoter. (B) Dual-color flow cytometric analysis. scFv expression is shown on the x axis and antigen-binding on the y axis. Top plot is no-antigen control; bottom plot is plus-antigen control. Uninduced cells are located in the bottom left quadrant of each plot.

Improvements in the affinity of the recognition reagents used in any given immunoassay often dictate ultimate performance and utility. Tacrolimus immunoassays with improvements in the limit of quantification (LOQ) or the limit of detection (LOD) will need to be developed as more effective treatment regimens are adopted. We used in vitro directed evolution to increase the binding affinity of the tacrolimus 1-60-46 antibody in the presence of 10% methanol to determine degree of improvement with the Architect Tacrolimus immunoassay.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
identification of tacrolimus 1-60-46 variable genes
We extracted mRNA from tacrolimus 1-60-46 hybridoma cell line (Abbott Laboratories code 93339) using Oligotex direct mRNA isolation kit (Qiagen). We synthesized Ig cDNA using MuIgGVH3'-2 (5'-CCCAAGCTTCCAGGGRCCARKGGATARACIGRTGG-3') and MuIg{kappa}VL3'-1 (5'-CCCAAGCTTACTGGATGGTGGGAAGATGGA-3') primers (Mouse Ig-Primer set; Novagen) with SuperScript III reverse transcriptase (Invitrogen). The Ig cDNA was PCR amplified: VH reactions from IgG cDNA used MuIgVH3'-2 and MuIgVH5' A–F primers (Mouse Ig-Primer set; Novagen) and VL reactions from Ig{kappa} cDNA used MuIg{kappa}VH3'-1 and MuIgkVH5' A–G primers (Mouse Ig-Primer set; Novagen). PCR reactions were ligated into pCR2.1 TOPO vector (Invitrogen), and plasmid DNA was sequenced from a number of transformants.

construction of tacrolimus wild-type SCFV for yeast display
We used tacrolimus VH or VL TOPO plasmid DNA to amplify the appropriate gene with gene-specific 5' and 3' primers. VH reaction used Tacro VH forward (5'-GCGGCCCAGCCGGCCATGGCCGAGGTGGAATTGGTGGAGTCTGGG-3') and Tacro VH reverse (5'- CGCCTCCTTCAGGGGCGTCAACTCCTTGGCGG GACCTGCAGAGACAGTGACCAGAGTCCC-3'). VL reaction used Tacro VL forward (5'-AAGGAGTTGACGCCCCTGAAGGAGGCGAAGGTCTCTGATGTTTTGATGACCCAAACTCCA-3') and Tacro VL reverse (5'-AGACTCGAGGGCGGCCGCCCGTTTCAGCTCCAGCTTGGTCCC-3'). The resulting purified PCR products were used for single overlap extension PCR (SOE-PCR) with Tacro VH forward and Tacro VL reverse primers. The tacrolimus wild-type (WT) scFv gene was cloned into pYD41 yeast display vector (derived from pYD1 vector (Invitrogen) after digestion with SfiI and NotI restriction enzymes, and we verified the sequence.

Tacrolimus WT scFv pYD41 plasmid was transformed into EBY100 yeast (Trp– phenotype) (Invitrogen) using lithium acetate as described(18). Briefly, we mixed approximately 108 yeast with 0.5 µg plasmid DNA in 240 µL 50% (wt/vol) polyethylene glycol, 36 µL 1.0 mol/L LiAc, 25 µL ssDNA (2 g/L) in a total volume of 351 µL at 42 °C for 40 min. Dilutions were spread onto culture medium glucose -Trp-Ura-His selective plates (Technova) and incubated at 30 °C. Individual colonies were used to inoculate SD-CAA [2% glucose/0.5% casamino acids/0.5% (NH4)2SO4/0.16% yeast nitrogen base/62 mmol/L NaH2PO4-H2O/38 mmol/L Na2HPO4-7H2O] cultures and grown at 30 °C for 12–24 h with shaking. scFv expression was induced by transferring midlog growth cells into SG/R-CAA (2% galactose and 1% raffinose replace 2% glucose) media to an initial density of 0.5 A600 cells/mL and grown at 20 °C for 12–24 h with shaking.

binding analysis of SCFV clones
We determined dissociation rates as described(19). Induced cells were first saturated with biotinylated tacrolimus (bt-tacro) antigen at 25 °C, washed, and incubated at 25 °C with 100-fold molar excess unlabeled tacrolimus (Astellas Pharma, Inc.). Reactions were performed in a selection diluent composed of PBS pH 7.4, 1% BSA, and 10% methanol. Individual samples were withdrawn at various time points and analyzed by flow cytometry to determine the amount of bt-tacro remaining after addition of streptavidin-R-phycoerythrin (SA-PE) (1:200) (Invitrogen). We used a first-order exponential decay equation (see Supplemental Methods and Materials in the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/content/vol54/issue6) to fit the antigen-binding mean fluorescence intensity (MFI) obtained from each time point and calculate the dissociation rate constant (koff).

We determined quantitative equilibrium binding using flow cytometry as described(19)(20). Typically, a titration of bt-tacro concentrations ranging from 10-fold above to 5-fold below the equilibrium constant for each clone were incubated with 106 yeast per reaction in the selection diluent containing 10% methanol. Incubation volumes were adjusted to limit antigen depletion (assuming 5 x 104 scFv/cell), and the reactions were allowed to equilibrate overnight. Binding of bt-tacro antigen to the yeast-displayed scFv was detected using SA-PE (diluted 1:200). To calculate the equilibrium affinity constant (Kd), we fitted the MFI of the antigen-binding channel to a nonlinear least-squares regression (see Supplemental Methods and Materials in the online Data Supplement).

mutagenic library construction
Individual libraries, in which 3 successive complementarity determining region (CDR) amino acid positions are mutated, were generated for all 6 CDRs (see Supplemental Methods and Materials in the online Data Supplement). Tacrolimus WT scFv pYD41 plasmid DNA was amplified to introduce a "gap" in each CDR region (primers listed in Supplemental Table 1 in the online Data Supplement). The CDR gap was replaced by a degenerate single-strand oligonucleotide encoding all 20 amino acid possible replacements within the 3–amino acid mutagenic window (Fig. 2 ) (Midland Certified Reagent Co.) flanked by regions of homology to the amplified vector to facilitate recombination. Approximately 1 µg of gapped vector and 16 µg of each individual degenerate single-strand oligonucleotide were cotransformed into yeast as outlined previously(21) to generate 51 individual libraries spanning every CDR residue of the tacrolimus 1-60-46 antibody. Yeast cells containing reconstituted vector after homologous recombination were propagated in SD-CAA media at 30 °C with shaking. We determined transformation efficiency by dilution plating of each library after transformation. Individual libraries within each CDR region were normalized and pooled before selection.


Figure 2
View larger version (13K):
[in this window]
[in a new window]

  		Figure 2. Schematic representation of the tacrolimus 1-60-46 CDR mutagenic libraries.

The amino acid sequences of the tacrolimus 1-60-46 VH (top) and VL (bottom) CDR regions are listed. The name and location of each 3-residue mutagenic window for all 51 libraries is shown below each CDR. The individual libraries within each CDR were pooled to create 6 master CDR libraries for selection.

affinity maturation selections
An overrepresentation (≥10x) of each mutagenic library or sort output from a previous round of selection, induced for scFv expression, was incubated with 0.1 µmol/L bt-tacro and anti-V5 monoclonal antibody (mAb) (2.5 mg/L) (Invitrogen) in the selection diluent (PBS, pH 7.4, 1% BSA, and 10% methanol) at 25 °C to saturate surface binding. Cells were washed and incubated at 25 °C with 100-fold molar excess of unlabeled tacrolimus for a fixed period(15) in the selection diluent [typically 3x–5x the dissociation half-life (t1/2)] (see Supplemental Methods and Materials in the online Data Supplement). We detected the amount of bt-tacro antigen remaining on each cell using SA-PE (1:200 dilution) and normalized binding to the amount of scFv expression using Alexa Fluor 488 goat antimouse (GaM-488, 1:100 dilution; Invitrogen).

Clones with enhanced binding were enriched using FACS on a FACSAria cell sorter (BD Biosciences). Typically the brightest 0.1% to 1.0% of scFv-expressing, antigen-binding cells after dissociation in 10% methanol were sorted. We performed 2 rounds of sorting and regrowth per library to isolate enriched populations of improved mutants. After each round of sorting, we plated an aliquot of cells on SD-CAA agar plates to obtain individual clones for sequence identification and further antigen-binding analysis.

generation of tacrolimus combinatorial mutant clones
Using PCR amplification (see Supplemental Methods and Materials in the online Data Supplement), we constructed combinatorial clones containing various mutations in the H2, L1, and L3 CDR regions. Each clone was amplified with appropriate primers (see Supplemental Table 2 in the online Data Supplement). After amplification of representative portions of each VL mutant containing homology in the framework 3 region, intact VL genes combining the L1 mutation with the L3 mutations were reconstructed and then combined by SOE-PCR using VL gap forward and reverse primers. Intact scFv genes containing the various mutant combinations in both VH and VL domains (see Supplemental Table 3 in the online Data Supplement) were reconstructed by SOE-PCR, using VH gap and VL gap reverse primers. The resulting combinatorial mutant scFv were cloned into the pYD41 vector using homologous recombination after transformation into EBY100 as described(21) using 1 µg of each amplified combinatorial mutant scFv gene and 0.1 µg pYD41 linearized with SfiI and NotI.

IGG expression of tacrolimus mutant antibodies
We converted selected tacrolimus mutant scFv clones into murine Ig2a/{kappa} antibodies by PCR amplification of the variable domains, followed by ligation into appropriate pBOS vector(22). We amplified mutant VH genes using Tacro VH IgG2a forward (5'-TTCTT GTCGCGATTTTAAAAGGTGTCCAGTGCGAGGT GGAATTGGTGGAGTCT-3') and Tacro VH IgG2a reverse (5'-TGTTTTAGCGCTTGCAGAGACAGTGA CCAGAGT-3'). We amplified mutant VL genes using Tacro VL mCk forward (5'-CCCGGCTCGCGATGC GATGTTTTGATGACCCAAACT-3') and Tacro VL Ck reverse (5'-AGCATCAGCGCTCGCCCGTTTCA GCTCCAGCTT-3'). We cotransfected pBOS plasmids into HEK-293 cells and purified the resulting supernatants over protein A column.

affinity determination of tacrolimus 1-60-46 mutant IGG antibodies
We determined the equilibrium dissociation constants for both the Tacrolimus AM 2 IgG and the Tacrolimus WT mAb IgG using kinetic exclusion assay (KinExA®) (Sapidyne Instruments)(23). A constant amount of IgG antibody was incubated with various concentrations (10–8 to 10–13 mol/L) of tacrolimus drug in the selection diluent (PBS, pH 7.4, 1% BSA, and 10% methanol) and allowed to come to equilibrium before sampling. We determined the Kd by analyzing the amount of free binding sites vs the amount of antigen present in the reaction sample and fitting to a 1:1 reversible binding model using KinExA Pro Software (version 1.0.3; Sapidyne Instruments).

tacrolimus mutant IGG immunoassay evaluation
The anti-tacrolimus mutant IgGs were individually immobilized to GaM-coated paramagnetic microparticles. The tacrolimus test sample was mixed with Architect Tacrolimus whole blood precipitation reagent (1L77-55; Abbott Laboratories) and incubated with the microparticles and a tracer reagent (1L77; Abbott Laboratories). The tracer molecule contains tacrolimus attached covalently to acridinium through a linker at position 32. Calibration curves were generated after titration of tacrolimus; data are plotted as ratio of signal for each concentration relative to the 0 µg/L control. Each data point is the mean of 3 replicates. LOD was calculated as the lowest tacrolimus concentration distinguishable from zero with 95% confidence, assuming a constant 3% relative light unit (RLU) CV.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
affinity maturation selections
We isolated VH and VL genes from tacrolimus 1-60-46 hybridoma cell line and used them to construct the WT scFv antibody for display on yeast cell surface. The tacrolimus WT scFv has a koff of 9.4 x 10–4 (±1.8 x 10–4)/s in a selection diluent (composed of PBS, pH 7.4, 1% BSA, and 10% methanol) to mimic tacrolimus whole blood immunoassay conditions. We used a dissociation rate selection strategy to identify tacrolimus variants with improved binding characteristics in the whole blood immunoassay selection diluent. Individual libraries within each CDR region were pooled before selection (e.g., H1 libraries 1–8 were combined to generate the H1 master library); however, each CDR master library was kept separate during the selection process (Fig. 2Up ). Three libraries (H2, L1, and L3) showed improved binding characteristics over the tacrolimus WT scFv clone after 2 rounds of selection (Fig. 3 ).


Figure 3
View larger version (24K):
[in this window]
[in a new window]

  		Figure 3. Tacrolimus FACS selection summary.

Bivariate plots after dissociation in selection diluent containing 10% methanol of the WT scFv clone and 3 libraries that were enriched for improved binders after second-round selection. scFv expression was detected using anti-V5 mAb/GaM-488 and is indicated with an arrow on the x axis. Antigen binding was detected using bt-tacro/SA-PE and is indicated with an arrow on the y axis. The degree of antigen binding after dissociation of the scFv-expressing population for the WT clone is indicated and included in all plots for comparison. The H2, L1, and L3 libraries show an increased level of antigen-binding MFI after 2 rounds of selection denoted by a shift of the population above the "WT" signal gate into the "Improved" signal gate along the y axis. Individual cells from each of these libraries were isolated and characterized.

We sequenced individual clones from each of the 3 libraries with improvements (Table 1 ) and characterized unique clones (Table 2 ). The H2 library had 3 unique clones, all of which contained a consensus tyrosine-to-tryptophan change at residue 62 (IMGT numbering(24)) (Table 1 ). Unexpectedly, 2 of the 3 clones also contained a valine-to-alanine mutation in residue 110 of VL CDR3. Clone H2-1A had the greatest improvement (nearly 6-fold relative to WT), with a koff of 1.6 x 10–4/s in the selection diluent. The L1 library also had 3 unique clones, all of which contained a consensus serine-to-glycine change at residue 28 (Table 1 ). The L1 clones also contained the same unexpected valine-to-alanine mutation in residue 110 of VL CDR3. Clone L1-1B had the greatest improvement (nearly 5-fold relative to WT), with a koff of 1.9 x 10–4/s in the selection diluent. The L3 library had 6 unique clones in which valine was changed to either cysteine or serine at residue 110 (Table 1 ). The best L3-cysteine mutant (L3-1A) was nearly 8-fold improved relative to WT, with a koff of 1.2 x 10–4/s in the selection diluent. The best L3-serine clone (L3-2B) was approximately 3-fold better than WT, with a koff of 3.2 x 10–4/s in the selection diluent.


View this table:
[in this window]
[in a new window]

  		Table 1. Tacrolimus affinity maturation mutants.


View this table:
[in this window]
[in a new window]

  		Table 2. Binding characterization of tacrolimus mutant clones.

analysis of tacrolimus combinatorial mutant clones
We used clones having the greatest improvement in dissociation rate from each master CDR library to construct scFv genes containing different pairings of the individual mutations. Combinatorial clones containing various mutations in the H2 (H2-1A), L1 (L1-1B), and L3 (L3-1A, L3-2B, and L1-1B, which contains V110A mutation) CDR regions were constructed by SOE-PCR. For simplicity, the L3 mutants were labeled as L3 cys, L3 ser, and L3 ala. The names and CDR sequences of the various combinatorial mutants are shown at the bottom of Table 1Up .

The dissociation rates and equilibrium dissociation constants for each combinatorial mutant, determined by flow cytometry in the selection diluent, are listed in Table 2Up . Many of the combinatorial mutants have additive improvements relative to each single mutant. Combinatorial mutant H2-L3 cys has the slowest koff (5.5 x 10–5/s), 17-fold better than the WT clone and >2-fold better than any single mutant clone. This clone has a Kd of 7.4 x 10–11 mol/L for bt-tacro and is 7.5-fold better than WT. Combinatorial mutant H2-L1-L3 ala has the lowest Kd (3.8 x 10–11 mol/L), which is nearly a 15-fold improvement relative to the WT clone and approximately 4-fold better than any single mutant clone. This clone also had >12-fold improved dissociation rate (7.5 x 10–5/s) relative to WT. Conversely, combining the L1 mutation with either the L3 cys or L3 ser mutations caused a reduction in dissociation rate improvement relative to any of the individual mutant clones and an overall decrease in Kd for the L1-L3 cys mutant relative to the WT clone.

functional testing of tacrolimus IGG mutants
Three of the combinatorial mutants were transiently expressed from mammalian cells as murine IgG2a/{kappa} antibodies (denoted tacrolimus AM 1–3) to allow direct comparison to the parental monoclonal antibody currently used in the Architect Tacrolimus immunoassay (Table 1Up ). We determined the solution-based equilibrium affinity of the tacrolimus WT and AM 2 IgG for unlabeled tacrolimus drug in the presence of 10% methanol. The WT IgG has a Kd of 1.5 x 10–10 mol/L for tacrolimus in buffer with 10% methanol. The AM 2 IgG is 12-fold improved relative to the starting hybridoma, with a Kd of 1.3 x 10–11 mol/L for tacrolimus in buffer with 10% methanol.

We tested tacrolimus combinatorial mutants AM 1–3 in an automated tacrolimus assay using a competitive format on the Architect instrument. Calibration curves were generated after titration of tacrolimus and data were plotted as the ratio of signal for each concentration to the signal for the 0 µg/L sample (X/A ratio). The results from an experiment using a set of calibrator samples ranging from 0 to 30 µg/L are shown in Fig. 4A . A tacrolimus concentration of 3 µg/L was able to displace 41% of the tracer with the tacrolimus WT IgG. The same 3-µg/L sample was able to displace 78%, 79%, and 80% for AM 1–3 IgG, respectively. A second experiment was performed with a new set of calibrators ranging from 0 to 5 µg/L; results are shown in Fig. 4B . A tacrolimus concentration of 1 µg/L was able to displace only 15% of the tracer with the tacrolimus WT IgG. Conversely, the same 1-µg/L sample was able to displace 58%, 35%, and 54% for AM 1–3 IgG. The LOD with this prototype assay using the tacrolimus AM IgG was decreased 3- to 6-fold relative to tacrolimus 1-60-46 WT mAb. Tacrolimus AM 1 has a LOD of 80 µg/L, AM 2 has a LOD of 170 µg/L, AM 3 has a LOD of 100 µg/L, and the WT mAb has a LOD of 470 µg/L.


Figure 4
View larger version (11K):
[in this window]
[in a new window]

  		Figure 4. Architect Tacrolimus assay results.

Immunoassay results comparing WT 1-60-46 IgG with various 1-60-46 mutant IgGs (AM 1, AM 2, or AM 3) using Architect Tacrolimus whole blood precipitation reagent. Microparticles coated with the denoted IgGs were incubated with various concentrations of unlabeled tacrolimus antigen: 0–30 µg/L (A) and 0–5 µg/L (B) for 18 min. Unoccupied binding sites were determined after addition of acridinium-labeled tacrolimus conjugate for 4 min. The ratio of the labeled antigen signal (which is shown on the y axis) is plotted against the concentration of unlabeled tacrolimus (which is shown on the x axis). •, WT mAb; {blacksquare}, AM 1; {triangleup}, AM 2; {diamondsuit}, AM 3. The amount of signal displacement for WT and mutants at 3 µg/L (A) and 1 µg/L (B) is denoted with arrows. The LOD was calculated as the lowest tacrolimus concentration distinguishable from zero with 95% confidence, assuming a constant 3% relative light unit (RLU) CV.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
We used yeast display to increase the affinity of the tacrolimus 1-60-46 mAb used in a commercial immunoassay to monitor whole blood concentration of the immunosuppressant drug tacrolimus. We used VH and VL genes isolated from the hybridoma cell line to construct a scFv antibody for in vitro directed evolution. Mutagenic libraries, in which every possible amino acid replacement was sampled at each CDR position, were screened for improved binding kinetics in a selection diluent designed to mimic conditions used for drug extraction from whole blood. Several clones with improved binding were isolated, and individual mutations were then combined for greater binding enhancements (Tables 1Up and 2Up ). Selected scFv clones were converted to full-length IgG and demonstrated significant improvement in the Architect Tacrolimus competitive immunoassay (Fig. 4Up ). As expected, the binding improvements were maintained as the mutant scFv constructs were converted to IgG. This conservation validates the conversion of murine mAb to the scFv format instead of the heterodimeric Fab. Cross-reactivity with tacrolimus metabolites was not tested because complete validation of the derived antibodies is beyond the scope of this work; however, we do not anticipate that the specificity will be altered.

Sequence analysis of the selected mutants shows that only 3 residues, each in a different CDR, are likely responsible for improved dissociation kinetics (Table 1Up ). VH CDR 2 contained a consensus tyrosine-to-tryptophan change at residue 62. VL CDR 1 contained a consensus serine-to-glycine change at residue 28. The valine at residue 110 in VL CDR 3 was changed to alanine, cysteine, or serine. The unique clones from each library contain different sequences that flank each of these 3 critical residues. The heterogeneity of the flanking sequences is likely due to the 3-residue window used to construct the mutagenic libraries (e.g., AWT for H2-1A, KWV for H2-1B, and EWT for H2-3B) (Table 1Up ). However, the local environment of each unique mutation window may have slightly different effects on the overall canonical conformation of the CDR loop and may affect binding(25). The unexpected V110A mutation in VL CDR 3 of the H2 and L1 clones occurred at the same position that was changed in all of the mutants from the L3 library and was likely introduced by PCR during library construction. This congruence further indicates that valine is not well tolerated in this position for this antigen system. Although the single V110A change alone does provide some improved binding relative to the parental clone, this change is not solely responsible for the improved dissociation observed in any of the H2 or L1 mutant clones (data not shown). Many of the combinatorial mutations appeared to be additive; however, combining the S28G mutation in VL CDR 1 with either V110C or V110S mutation in VL CDR 3 negates much of the benefit of any of the clones alone, and overall affinity is equivalent to the WT clone. It is tempting to speculate, but the mechanisms by which any of the single or combinatorial mutations affect antigen binding are difficult to infer without comparing cocrystallographic structures.

Many examples have been published of in vitro directed evolution to improve antibody affinity or specificity(10)(11)(12)(14)(26)(27)(28)(29)(30). These reports have largely centered on improved efficacy of therapeutic agents, mechanistic understanding of the bimolecular interaction under study, or technology development. The degree of affinity improvement obtained with the work presented here is in keeping with previous results obtained with both yeast and other in vitro antibody display systems. Affinity improvements >1000-fold have been reported(10)(30); however, in contrast to the work presented here, those results were obtained after multiple rounds of mutagenesis and screening from parental clones with much lower starting affinities (typically low nmol/L) than that of the tacrolimus 1-60-46 antibody. mAb clones have usually undergone extensive somatic hypermutation during the immunization process unlike clones derived from in vitro antibody discovery efforts. Nevertheless, this iterative in vitro approach is readily applicable to our work if warranted, and its use will ultimately depend on both the starting affinity and desired level of detection.

Although it is generally accepted that improved recognition elements are prerequisites for the development of highly sensitive diagnostic assays, to our knowledge, this is the first example to show enhanced performance with a commercially available immunoassay. The improved LOD of each of the tacrolimus 1-60-46 AM IgGs will allow more robust and sensitivity assays to be developed as more efficacious treatment regimens are adopted. This approach is also easily adaptable to most antibody-antigen systems and provides the ability to improve a wide range of immunoassays.


   Acknowledgments
 
Grant/Funding Support: None declared.

Financial Disclosures: None declared.

Acknowledgments: The authors acknowledge Drs. Jon Belk and Susan Lacy (Abbott Bioresearch Center) for the gracious gift of pYD41 vector plasmid, Don Johnson (Abbott Laboratories) for preparation of biotinylated tacrolimus, and Drs. Susan Brophy and Audrey Bartnicki (Abbott Laboratories) for critical review of the manuscript.


   Footnotes
 
1 Nonstandard abbreviations: scFv, single-chain fragment variable; VH, variable heavy; VL, variable light; FACS, fluorescence activated cell sorting; LOQ, limit of quantification; LOD, limit of detection; SOE, single overlap extension; WT, wild-type; bt-tacro, biotinylated tacrolimus; SA-PE, streptavidin-R-phycoerythrin; MFI, mean fluorescence intensity; koff, dissociation rate constant; Kd, equilibrium affinity constant; CDR, complementarity determining region; mAb, monoclonal antibody; GaM-488, Alexa Fluor 488 goat antimouse.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Kino T, Hatanaka H, Miyata S, Inamura N, Nishiyama M, Yajima T, et al. FK-506, a novel immunosuppressant isolated from a Streptomyces. II. Immunosuppressive effect of FK-506 in vitro. J Antibiot (Tokyo) 1987;40:1256-1265.[Medline] [Order article via Infotrieve]
  2. Fung JJ, Starzl TE. FK506 in solid organ transplantation. Ther Drug Monit 1995;17:592-595.[Web of Science][Medline] [Order article via Infotrieve]
  3. Inamura N, Nakahara K, Kino T, Goto T, Aoki H, Yamaguchi I, et al. Prolongation of skin allograft survival in rats by a novel immunosuppressive agent, FK506. Transplantation 1988;45:206-209.[Web of Science][Medline] [Order article via Infotrieve]
  4. Flanagan WM, Corthesy B, Bram RJ, Crabtree GR. Nuclear association of a T-cell transcription factor blocked by FK-506 and cyclosporin A. Nature (Lond) 1991;352:803-807.[CrossRef][Medline] [Order article via Infotrieve]
  5. Randhawa PS, Shapiro R, Jordan ML, Starzl TE, Demetris AJ. The histopathological changes associated with allograft rejection and drug toxicity in renal transplant recipients maintained on FK506: clinical significance and comparison with cyclosporine. Am J Surg Pathol 1993;17:60-68.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  6. Dietemann J, Berthoux P, Gay-Montchamp JP, Batie M, Berthoux F. Comparison of ELISA method versus MEIA method for daily practice in the therapeutic monitoring of tacrolimus. Nephrol Dial Transplant 2001;16:2246-2249.[Abstract/Free Full Text]
  7. Van Duyne GD, Standaert RF, Karplus PA, Schreiber SL, Clardy J. Atomic structure of FKBP-FK506, an immunophilin-immunosuppressant complex. Science (Wash DC) 1991;252:839-842.[Abstract/Free Full Text]
  8. Melnikova YI, Odintsov SG, Kravchuk ZI, Martsev SP. Antigen-binding activity of monoclonal antibodies after incubation with organic solvents. Biochemistry (Mosc) 2000;65:1256-1265.[Medline] [Order article via Infotrieve]
  9. Boder ET, Wittrup KD. Yeast surface display for screening combinatorial polypeptide libraries. Nat Biotechnol 1997;15:553-557.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  10. Boder ET, Midelfort KS, Wittrup KD. Directed evolution of antibody fragments with monovalent femtomolar antigen-binding affinity. Proc Natl Acad Sci U S A 2000;97:10701-10705.[Abstract/Free Full Text]
  11. Graff CP, Chester K, Begent R, Wittrup KD. Directed evolution of an anti-carcinoembryonic antigen scFv with a 4-day monovalent dissociation half-time at 37 degrees C. Protein Eng Des Sel 2004;17:293-304.[Abstract/Free Full Text]
  12. Razai A, Garcia-Rodriguez C, Lou J, Geren IN, Forsyth CM, Robles Y, et al. Molecular evolution of antibody affinity for sensitive detection of botulinum neurotoxin type A. J Mol Biol 2005;351:158-169.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  13. Weaver-Feldhaus JM, Miller KD, Feldhaus MJ, Siegel RW. Directed evolution for the development of conformation-specific affinity reagents using yeast display. Protein Eng Des Sel 2005;18:527-536.[Abstract/Free Full Text]
  14. Garcia-Rodriguez C, Levy R, Arndt JW, Forsyth CM, Razai A, Lou J, et al. Molecular evolution of antibody cross-reactivity for two subtypes of type A botulinum neurotoxin. Nat Biotechnol 2007;25:107-116.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  15. Boder ET, Wittrup KD. Optimal screening of surface-displayed polypeptide libraries. Biotechnol Prog 1998;14:55-62.[CrossRef][Medline] [Order article via Infotrieve]
  16. Siegel RW, Coleman JR, Miller KD, Feldhaus MJ. High efficiency recovery and epitope-specific sorting of an scFv yeast display library. J Immunol Methods 2004;286:141-153.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  17. VanAntwerp JJ, Wittrup KD. Fine affinity discrimination by yeast surface display and flow cytometry. Biotechnol Prog 2000;16:31-37.[CrossRef][Medline] [Order article via Infotrieve]
  18. Gietz RD, Woods RA. Yeast transformation by the LiAc/SS Carrier DNA/PEG method. Methods Mol Biol 2006;313:107-120.[Medline] [Order article via Infotrieve]
  19. Boder ET, Wittrup KD. Yeast surface display for directed evolution of protein expression, affinity, and stability. Methods Enzymol 2000;328:430-444.[Web of Science][Medline] [Order article via Infotrieve]
  20. Chao G, Lau WL, Hackel BJ, Sazinsky SL, Lippow SM, Wittrup KD. Isolating and engineering human antibodies using yeast surface display. Nat Protoc 2006;1:755-768.[CrossRef][Medline] [Order article via Infotrieve]
  21. Orr-Weaver TL, Szostak JW, Rothstein RJ. Genetic applications of yeast transformation with linear and gapped plasmids. Methods Enzymol 1983;101:228-245.[Web of Science][Medline] [Order article via Infotrieve]
  22. Mizushima S, Nagata S. pEF-BOS, a powerful mammalian expression vector. Nucleic Acids Res 1990;18:5322.[Free Full Text]
  23. Darling RJ, Brault PA. Kinetic exclusion assay technology: characterization of molecular interactions. Assay Drug Dev Technol 2004;2:647-657.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  24. Giudicelli V, Chaume D, Lefranc MP. IMGT/V-QUEST, an integrated software program for immunoglobulin and T cell receptor V-J and V-D-J rearrangement analysis. Nucleic Acids Res 2004;32:W435-W440.[Abstract/Free Full Text]
  25. Al-Lazikani B, Lesk AM, Chothia C. Standard conformations for the canonical structures of immunoglobulins. J Mol Biol 1997;273:927-948.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  26. Groves M, Lane S, Douthwaite J, Lowne D, Gareth Rees D, Edwards B, Jackson RH. Affinity maturation of phage display antibody populations using ribosome display. J Immunol Methods 2006;313:129-139.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  27. Luginbuhl B, Kanyo Z, Jones RM, Fletterick RJ, Prusiner SB, Cohen FE, et al. Directed evolution of an anti-prion protein scFv fragment to an affinity of 1 pM and its structural interpretation. J Mol Biol 2006;363:75-97.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  28. Wang Z, Kim GB, Woo JH, Liu YY, Mathias A, Stavrou S, Neville DM. Improvement of a recombinant anti-monkey anti-CD3 diphtheria toxin based immunotoxin by yeast display affinity maturation of the scFv. Bioconjug Chem 2007;18:947-955.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  29. Wu H, Pfarr DS, Johnson S, Brewah YA, Woods RM, Patel NK, et al. Development of motavizumab, an ultra-potent antibody for the prevention of respiratory syncytial virus infection in the upper and lower respiratory tract. J Mol Biol 2007;368:652-665.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  30. Zahnd C, Wyler E, Schwenk JM, Steiner D, Lawrence MC, McKern NM, et al. A designed ankyrin repeat protein evolved to picomolar affinity to Her2. J Mol Biol 2007;369:1015-1028.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
clinchem.2007.097352v1
clinchem.2007.097352v2
54/6/1008    most recent
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 Web of Science
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 Siegel, R. W.
Right arrow Articles by Tyner, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Siegel, R. W.
Right arrow Articles by Tyner, J.
Related Collections
Right arrow Drug Monitoring and Toxicology

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS