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


     
 

The Journal of Immunology, 2008, 181, 1282 -1287
Copyright © 2008 by The American Association of Immunologists, Inc.
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 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 Request Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lacy, S. E.
Right arrow Articles by Reilly, E. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lacy, S. E.
Right arrow Articles by Reilly, E. B.

The Potency of Erythropoietin-Mimic Antibodies Correlates Inversely with Affinity

Susan E. Lacy*, Peter J. DeVries{dagger}, Nancy Xie{dagger}, Emma Fung*, Richard R. Lesniewski{dagger} and Edward B. Reilly1,{dagger}

* Global Pharmaceutical Research and Development, Abbott Bioresearch Center, Worcester, MA 01605; and {dagger} Global Pharmaceutical Research and Development, Abbott Laboratories, Abbott Park, IL 60064


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Preclinical animal studies have shown that Ab12.6, an agonistic human Ab targeting the erythropoietin receptor (EPOR), exhibits several potential dosing and safety features that make it an attractive clinical candidate for the treatment of anemia. Ab12.6 was derived by yeast display affinity maturation of parental Ab12, a strategy initially intended to improve off-rate and affinity for EPOR, thereby enhancing erythropoietic activity. Analysis of full-length IgGs derived from yeast clones identified sequences within Ab12 CDRH2 that independently influenced both affinity and potency. The Ab12.6 derivative displayed improved in vitro potency and in vivo efficacy, although its binding affinity to the EPOR was lower than that of the parent Ab12. Additional Ab12 derivatives also exhibited an inverse correlation between affinity and potency. These results suggest that for this class of agonistic Abs, faster off-rates may permit continuous receptor stimulation and more efficient erythropoiesis.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Erythropoietin (EPO),2 a naturally occurring hematopoietic growth factor produced by the kidney, is the primary regulator of erythropoiesis (1). Recombinant human EPO (rHu-EPO) has important clinical uses in patients with anemia associated with renal disease and cancer (2). We have previously shown that a human EPO-mimic Ab, Ab12.6 (also known as ABT007), is a potent stimulator of erythropoiesis in a human EPOR-expressing transgenic mouse model (3). Based on the its ability to support potent, more sustained, and less pulsatile elevation of hematocrit, while requiring less frequent dosing compared with standard doses of rHu-EPO, Ab12.6 may provide both potential medical benefits and improved patient convenience. It is unlikely that Ab12.6 would induce Ab-mediated pure red cell aplasia, a condition associated with some forms of rHu-EPO due to the formation of rHu-EPO-induced neutralizing Abs (4). Although the extended serum half-life of the human Ab may enhance its in vivo properties, the potency of Ab12.6 resides primarily in a novel mechanism of receptor activation based on a unique, Ab-imposed, conformational change (3).

Ab12.6 was originally derived from a parental human Ab, Ab12, which was raised to the extracellular domain (ECD) of EPOR using XenoMax technology (5). Ab12 is capable of stimulating the proliferation of EPO-responsive cells both in vitro and in vivo. This activation is influenced by Ab12/EPOR on-off rate kinetics and binding affinity. Since the Kd of Ab12 for EPOR is 10-fold higher than the 1 nM dissociation constant observed between EPO and the high-affinity EPOR binding site (6), we reasoned that improving the off-rate and, therefore, the overall affinity of Ab12 may enhance its erythropoietic activity. Improvements in Ab affinity often correlate with improvements in potency (7, 8, 9, 10, 11, 12, 13), although the correlation between changes in agonistic Ab affinity and biologic potency is less well characterized. Using yeast display affinity maturation to engineer single-chain variable fragment (scFv) Ab12 variants (14), various derivatives were isolated, constructed as full-length IgG2/{kappa} Abs, and tested for affinity and potency to EPOR. Sequences within CDRH2 were identified that independently affected both affinity and potency. Ab12.6 emerged from these studies as a potential clinical candidate based on enhanced biological efficacy. Despite its improved potency, however, Ab12.6 had a KD value less than that of the parental Ab12, indicating that for this pair of closely related Abs, potency and affinity are inversely correlated.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 References
 
Generation and characterization of Ab12

A soluble form of mature human EPOR ECD representing residues 1–225 was expressed in Escherichia coli, refolded, and purified as described (15). XenoMouse (XenoMouse XG2; Amgen) were immunized with the EPOR ECD, and Ab12 was generated using XenoMax technology as previously described (3, 5, 16). Ab12 and its variants were tested for erythropoietic activity in cell proliferation assays with EPO-responsive cells (17, 18) and by their ability to support the formation of erythroid colonies (CFU-E) from human bone marrow containing CD34+ progenitor cells (3). In vivo efficacy was evaluated in an EPOR–/– human EPOR+ transgenic mouse model (19, 20). All animal studies were conducted in accordance with the guidelines established by the Abbott Laboratories Institutional Animal Care and Use Committee.

Expression of Ab12 scFv on yeast surface

The yeast cell (Saccharomyces cerevisiae) serves as a living scaffold permitting surface display of diverse populations of scFvs as inducible fusion proteins with the Aga2p mating agglutinin protein (14). Two individual DNAs encoding Ab12 VH and VL domains were amplified by PCR with oligonucleotide primers. The VH domain was amplified using forward primer 5107 5'-AGTAACGTTTGTCAGTAATTGC-3' and reverse primer 5'-TCCTGAGGAGACGGTGACCAGGGTTCC-3', and the VL using forward primer 5'-CCCTGGTCACCGTCTCCTCAGGACCCGCCAAGGAGTTGACGCCCCTGAAGGAGGCGAAGGTCAGTGACATCCAGCTGACCCAATCT-3' and reverse primer 5106 5'-GTCGATTTTGTTACATCTACAC-3'. One microgram of each of these PCR products was mixed in a trans-PCR reaction with the outermost forward and reverse primers 5106 and 5107, respectively. The resulting scFv DNA encoded from 5' to 3' an SfiI restriction site, the VH domain, a linker sequence, the VL domain, and a NotI restriction site. The linker (designated linker 40) encoded the amino acid sequence GPAKELTPLKEAKVS. The cDNA was then ligated into the SfiI and NotI sites of a modified vector derived from pYD1 (Invitrogen). The pYD1 vector was modified to encode substitution of the (Gly4Ser)3 (3' of the Aga2p sequence) with linker sequence GENKVEYAPALMALS, designated linker sequence 41. The Ab12 scFv fusion construct also included vector encoded N-terminal Xpress and C-terminal V5 and poly-His epitope tags to allow normalization of scFv surface copy number. In summary, the Ab12 scFv construct encoded Aga2p–linker 40–Ab12 VH–linker 41–Ab12 VL–Xpress epitope–V5 epitope–poly-His epitope. The vector was introduced into S. cerevisiae using a Geitz yeast transformation kit (Tetralink), and transformants were plated on dextrose-containing tryptophan (–) and uracil (–) selective agar plates. Several emergent colonies were expanded into dextrose-containing tryptophan (–) and uracil (–) selective liquid culture. Approximately 1 x 106 yeast cells were transferred to selective media supplemented with galactose to induce scFv expression. After 16–20 h incubation at 30°C, scFV expression was measured by flow cytometric analysis of the FL2 signal generated following incubation with murine anti-V5 mAb (Invitrogen) for 30 min on ice, washing cells three times at 4°C with PBS containing 1 mg/ml BSA, and incubation at 4°C with anti-mouse PE (Jackson ImmunoResearch Laboratories) for an additional 30 min. All samples were analyzed on an Epics XL flow cytometer (Beckman Coulter). Robust scFv expression was noted on 60% of the cells, typical of inducible scFv protein expression on yeast. A single stock clone was selected and expanded for expression studies.

Off-rate analysis of Ab12 scFv on yeast

Off-rate measurements were performed by incubating Ab12 scFv yeast cells with saturating quantities of EPOR ECD followed by addition of a 10,000-fold excess of Ab12 IgG1/{kappa} as a competitor. (Ab12-IgG1 is a derivative of Ab12 that was genetically engineered to encode an IgG1 isotype.) Individual yeast cell sample aliquots, incubated at 37°C, were withdrawn at defined time points and placed on ice until all samples were ready for incubation with detection reagents. The soluble EPOR remaining bound to Ab12 scFV was measured by addition anti-EPOR reactive mAb307 (R&D Systems) followed by anti-mouse PE Ab. All samples were analyzed on an Epics XLI flow cytometer and the mean fluorescence of the initial signal in each sample was then plotted against the time allowed for dissociation to generate an off-rate curve.

Generation of Ab12 scFv CDR mutant libraries

All six CDRs of Ab12 scFv (three in the H chain and three in the L chain) were subjected to a mutagenesis scan by randomization of three amino acids per library, theoretically creating 8000 versions of Ab12 per library. Forty-four single-stranded degenerate oligonucleotides (100 bases each), each encoding randomized versions of three amino acids in the VH or VL of Ab12, were synthesized (Midland Certified Reagent). To create overlapping three amino acid libraries, the three amino acid "window" was shifted 3' by one amino acid in each library. For example, VH library 1 had substitutions in amino acids 1–3, library 2 had substitutions in amino acids 2–4, library 3 had substitutions in amino acids 3–5, and so forth. To create the libraries, the pYD1-Ab12 vector was modified by PCR to contain a gap where the VH or VL degenerate oligonucleotides would overlap by 25 bp on each end of the vector. Thereafter, the gap was replaced by one of 44 degenerate single-stranded oligonucleotides by homologous recombination in yeast. Briefly, 1 µg vector and 16 µg oligonucleotides were combined and added to 1.25 x 108 yeast cells, followed by transformation of the mixture using the Geitz transformation kit (Tetralink). A total of 44 libraries were generated using this method. The libraries were assessed for quality following the selection and sequencing of yeast plasmid DNA encoding the mutant Ab12 scFv sequences. Plasmid was isolated using a Y-PER kit (Pierce), and scFv sequences were amplified by PCR before sequencing.

Selection of Ab12 scFv libraries by FACS

All 44 Ab12 scFv libraries and wild-type Ab12 scFv yeast were subjected to individual off-rate FACS analysis on a MoFlo high-speed cell sorter (DakoCytomation). Approximately 6 x 106 cells from each library were incubated with 0.5 µM EPOR ECD at 37°C until equilibrium was reached (2 h). Cells were then chilled on ice, washed three times with BSS at 4°C, and a 10,000-fold molar excess (5 µg/ml) of Abl2 IgG1 prewarmed to 37°C was added. After incubation for 20 min at 37°C, cells were chilled, washed, and labeled with mAb307 plus rabbit anti-6-His Ab (Research Diagnostics) for 30 min on ice. Cells were washed three times with BSS at 4°C and then incubated with a mixture of anti-mouse PE Ab and goat anti-rabbit FITC (Southern Biotechnology Associates). Individual control samples were also prepared to set MoFlo compensation and to ensure that no nonspecific background staining existed.

For the round 1 off-rate FACS, each library was analyzed on the MoFlo for FL1 (expression of scFv) and FL2 (bound soluble EPOR). For all 44 libraries, the brightest 1% of cells in the FL2 axis were co-gated and collected into a single tube containing drug-selective media. These cells were then grown 10 generations in selective media and called the round 1 output. These cells were then processed and stained for the round 2 FACS exactly as the cells were treated for round 1. The brightest 0.1% of cells in the FL2 axis were collected and amplified as for round 1. Plasmid DNA was isolated from individual yeast colonies and sequenced to identify regions containing mutations.

Cloning and expression of yeast display-derived Abs

Selected scFvs were converted into full-length IgG2/{kappa} Abs by PCR amplification of the mutant Ab12 VH or VL domains. These products were ligated into an intact IgG2 constant region or {kappa} region present in the vector pBOS (21). pBOS plasmids encoding both H and L chain regions were co-transfected transiently into COS cells, and resulting supernatants from 3-day cell cultures were purified over a protein A-Sepharose column (Amersham Biosciences). Purified Abs were dialyzed into PBS and concentrations were determined using absorbance at 280 nm.

Surface plasmon resonance analysis of Abs derived from yeast display

Real-time binding interactions between Abs captured on a biosensor matrix with goat anti-human IgG and soluble EPOR were measured by surface plasmon resonance using the Biacore 3000 system (Biacore Life Sciences) according to the manufacturer’s instructions. Briefly, soluble EPOR was diluted in HBS running buffer and 50-µl aliquots were injected through the immobilized protein matrixes at a flow rate of 5 µl/min. The concentrations of EPOR used were 2.5, 5, 10, 20, 25, 40, 50, 80, 100, and 200 nM. Biacore kinetic evaluation software (version 4.0.1) was used to determine the Kd, Ka, and KD.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 References
 
Expression and characterization of Ab12 scFv on the surface of yeast

Ab12 was one of several recombinant Abs identified from the XenoMax approach based on its ability to stimulate the proliferation of EPO-responsive cells (3). Affinity measurements of Ab-12 bound to sEPOR indicated an on-rate of 1.4 x 105/M/s, an off-rate of 1.3 x 10–3/s, and an overall KD of 9 nM (Table I). The Kd of Ab12 for EPOR is 10-fold higher than the 1 nM dissociation constant observed between EPO and EPOR (6). Since improvements in Ab affinity often correlate with enhanced potency, it was hypothesized that an Ab with higher affinity may result in increased erythropoietic activity. Ab12 was converted into a single-chain format and expressed in S. cerevisea via the Aga2 tether protein. The Ab12 scFv construct also contained a V5 epitope tag fused in-frame with the scFv. Expression of Ab12 scFv was induced and detected by incubating cells with an anti-V5 Ab followed by an anti-mouse PE secondary Ab. Equilibrium titration data generated with increasing concentrations of EPOR ECD provided an approximate KD value of 4 nM for the Ab12 scFv-soluble EPOR interaction (data not shown), consistent with the 9 nM KD value for the intact Ab12-sEPOR interaction. An off-rate determination for yeast-displayed Ab12 scFv was performed by measuring remaining prebound sEPOR dissociation over time. The design was such that as EPOR ECD dissociated from Ab12 scFv, it would immediately bind to Ab12 IgG1 competitor mAb (present at a saturating concentration) and would therefore no longer be detected on the surface of yeast. After 20 min of competition at 37°C, no further dissociation appeared to be taking place, consistent with the 10–3 off-rate measured for Ab12 by BIAcore (Table I).


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

  		Table I. Affinity measurement of Ab12 variants

 
Ab12 scFv CDR mutant libraries and identification of variants with improved binding to EPOR

Forty-four libraries containing randomized mutations within the CDRs were subjected to off-rate analysis (Fig. 1). Eleven of these libraries, including eight CDR H2 libraries, two CDR L2 libraries, and one CDR L3 library (highlighted in gray in Fig. 1), displayed the highest fluorescent shift indicative of EPOR binding. The round 1 and round 2 off-rate FACS outputs for library H2-1–3, displaying the highest fluorescent shift, is shown in Fig. 2. Only two rounds of sorting were necessary to generate a 10,000-fold enrichment from the starting library due to the low diversity contained in each of the libraries and the percentage of cells that were gated in each round. Individual clones from these outputs were recovered following plating on selective media and plasmid DNA isolation. The scFv region of each cloned plasmid was amplified by PCR and sequenced to identify the nucleotide substitutions.


Figure 1
View larger version (32K):
[in this window]
[in a new window]

  		FIGURE 1. Schematic representation of Ab12 scFv CDR VH and VL yeast libraries. Each library comprised a three amino acid window in which all of the residues were randomized. All libraries were subjected to a single round of off-rate FACS analysis and, based on increased fluorescent binding, 18 libraries were selected for round 2 FACS. Eleven libraries (highlighted in gray) displaying the highest fluorescent shift, indicative of EPOR binding, following round 2 FACS were selected and sequenced to identify the Ab scFV amino acid mutations resulting in improved off-rates.

 

Figure 2
View larger version (27K):
[in this window]
[in a new window]

  		FIGURE 2. Off-rate FACS of mutagenic Ab12 scFv library H2-1–3. A, The Ab12 scFv control indicates the number and fluorescence of cells binding soluble EPOR after a 20-min competition (FL2 axis) as well as the Ab12 scFv expression level (FL1). B, Round 1 FACS output of library H2-1–3 following a 20-min competition with Ab12 IgG1. Cells contained in the indicated gate were collected, expanded, and subjected to a second round of FACS. C, Round 2 off-rate FACS of the H2-1–3 library showing an overall enrichment of clones displaying improved binding to EpoR following a 20-min competition with Ab12 IgG1. Cells contained in the indicated gate were collected, and a fraction of the output cells was plated for colony growth.

 
Sequence analysis and affinity measurements of Ab12 variants

Overall, clones containing mutations in CDR H2 displayed the highest overall EPOR binding following a 20-min competition with Ab12 IgG1 competitor mAb, and representatives of these were selected for conversion into full-length IgG2/{kappa} mAbs to match the original isotype of Ab12. The mAbs were designated Ab12.6, Ab12.17, Ab12.25, Ab12.56, Ab12.61, Ab12.70, and Ab12.76. A schematic of the selected clones and the exact locations of their CDR H2 amino acid mutations (in the context of Ab12) are shown in Fig. 3. Essentially two regions in CDR H2 were represented in mutant Ab12 scFvs displaying improved off-rates on the surface of yeast. These regions spanned amino acids YYS and TNY, and in each case two or more of these amino acids were substituted. To determine whether these mutations correlated with improved off-rates in the context of a full-length mutant Ab12 IgG2/{kappa} mAb, expressed and purified Abs were subjected to affinity measurements using EPOR ECD as the test Ag analyte. Ab on- and off-rates were measured, and the improvement in overall KD was calculated (Table I). Ab variants having CDRH2 mutations substituting the TNY region of Ab12 (Fig. 3) had improved off-rates corresponding to a 9–27-fold improvement in overall KD (Ab12.17, Ab12.76, Ab12.25, and Ab12.61). These data suggest that the TNY region in CDRH2 is critical for high-affinity interactions with soluble EPOR. In contrast, the variants having CDRH2 mutations substituting the YYS region of Ab12 (Fig. 3) displayed off-rates that were comparable to or up to 3-fold faster than Ab12 (Ab12.6, Ab12.56, and Ab12.70). Since all of the above Abs were initially converted from characterized and sequenced yeast clones that showed both improved off-rates and overall fluorescence, it appears that, at least in the case of the YYS mutants, off-rate kinetic attributes were not retained upon conversion to full-length IgGs.


Figure 3
View larger version (8K):
[in this window]
[in a new window]

  		FIGURE 3. Schematic representation of Ab12 CDR H2 variants. All clones represented were expressed and purified as full-length IgG2/{kappa} Abs. Changes to CDRH2 are highlighted in bold. All other amino acids in each Ab12 variant were identical to parental Ab12, including the Ab12 {kappa} sequence.

 
Erythropoietic activity of Ab12.6 vs Ab12

Preliminary results indicated that Ab12.6, despite having a faster off-rate for the EPOR compared with Ab12, retained significant potency. To further evaluate functional differences between these Abs, their erythropoietic potency was evaluated in an in vitro cell proliferation assay using the F36E EPO-dependent cell line. As shown in Fig. 4A, the maximal proliferative response for Ab12.6 exceeded that for Ab12, indicating that Ab12.6 is a more potent stimulator of F36E cells. The erythropoietic activity of Ab12.6 appeared comparable to the rHu-EPO control, indicating that the YYS mutation in CDRH2, while negatively affecting binding affinity, improved overall agonistic activity. A similar pattern of stimulation was also seen with UT-7, another EPO-responsive cell line (data not shown).


Figure 4
View larger version (19K):
[in this window]
[in a new window]

  		FIGURE 4. Ab12.6 is more effective than Ab12 at stimulating in vitro erythropoiesis. A, The effect of increasing concentrations of Abs 12 and 12.6 on the proliferation of EPO-responsive F36E human erythroleukemic cells was measured after 72 h in culture using MTS reagent (Promega) readout. B, CFU-E colony formation from hematopoietic precursor cells following addition of increasing concentrations of Ab12, Ab12.6, and an isotype control Ab. The colonies were red in color and identified microscopically. rHu-EPO (Epogen; Amgen) was used as a positive control. Error bars represent SD calculated from the average of duplicate counts.

 
EPO supports the growth and differentiation of erythroid progenitor cells. Ab12 and Ab12.6 were therefore also evaluated for their ability to support the formation of erythroid colonies from human bone marrow containing CD34+ progenitor cells. Consistent with the results observed with EPO-dependent cell lines (Fig. 4A), Fig. 4B indicates that Ab12.6 is more effective at stimulating erythropoiesis, as evidenced by the increased number of erythroid colonies.

Erythropoietic activity of Ab12 variants

The eight full-length IgG2/{kappa} Ab 12 derivatives (Fig. 3) were evaluated to further determine the effect of the two CDRH2 substituted regions on potency. F36E proliferation assay results shown in Fig. 5 indicate that the activities of the Abs could be divided into two main classes dependent on the region in CDRH2 that was mutated. Ab12.56 and Ab12.70, like Ab12.6, displayed maximal proliferative activity at least as comparable to or greater than that for Ab12 (red line graph in Fig. 5A), indicating that overall agonistic activity could be positively impacted by amino acid substitutions of the YYS region. In contrast, proliferation curves from TNY variants 12.17, 12.25, 12.61, and 12.76 shifted primarily to the right of Ab12 (red line graph in Fig. 5B), consistent with reduced potency. Since the YYS and TNY region variants also correlate with opposing off-rate kinetics (Table I and Fig. 3), collectively, these results substantiate the observations made with Ab12 and Ab12.6 that improvements in agonistic activity are inversely correlated with overall Ab affinity.


Figure 5
View larger version (28K):
[in this window]
[in a new window]

  		FIGURE 5. EPO-dependent cell proliferation activity of Ab12 variants correlates inversely with Kd. A, Ab variants with faster off-rates (Abs 12.6, 12.70, and 12.56) stimulate the proliferation of EPO-responsive F36E human erythroleukemic cells more effectively than does Ab12. B, Ab variants with slower off-rates (Abs 12.61, 12.17, 12.25, and 12.76) are less effective than Ab12 in stimulating proliferation of F36E cells. rHu-EPO was used as a positive control; error bars represent SD calculated from the average of duplicate counts.

 
A similar pattern of reactivity was also seen following assessment of selected Abs for their ability to support the formation of erythroid colonies from human bone marrow. As shown in Fig. 6, YYS variant Abs 12.6, 12.56, and 12.70 were all more effective, to varying extents, in inducing colony formation than was Ab12. In contrast, the TNY variant, Ab12.17, despite having the greatest binding affinity for EPOR within the existing Ab panel (Table I), was ineffective in stimulating formation of erythroid colonies.


Figure 6
View larger version (11K):
[in this window]
[in a new window]

  		FIGURE 6. CFU-E colony forming activity of Ab12 variants correlates inversely with Kd. Compared with Ab12, Abs with faster off-rates (12.6, 12.70, and 12.56) are more effective, whereas Ab12.17 with a slower off-rate is less effective at supporting the formation of erythroid colonies from hematopoietic precursor cells. rHu-EPO was used as a positive control. The number of colonies on top of each bar represents the average from duplicate experiments. Error bars represent SD calculated from the average of duplicate counts.

 
In vivo potency of Ab12.6 vs Ab12

To determine whether the inverse correlation between affinities and erythropoietic activity is also reflected in vivo, Abs 12 and 12.6 were evaluated for the ability to elevate hematocrit in animals. In these experiments the erythropoietic activities of the Abs were compared with that of NESP (Amgen), a long-acting, hyperglycosylated analog of rHu-EPO currently in clinical use for anemia treatment as darbepoetin alfa (22). A transgenic mouse model, generated by rescuing genetic knockout mice lacking the murine EPOR gene with the human EPOR transgene (19, 20), was used since Ab12 and Ab12.6 do not recognize rodent EPOR (3). A single administration of either Ab at concentrations ranging from 0.4 to 1.6 mg/kg resulted in a dose-dependent rise in hematocrit measured on day 21 (Fig. 7). Additional results indicated that these hematocrit increases are sustained beyond 3 wk following dosing with Ab12.6 (3). At every dose tested, Ab12.6 was more effective at elevating hematocrit than was the equivalent dose of Ab12. Additionally, the hematocrit increase observed on day 21 following a single dose of Ab12.6 was at least equivalent to that observed with a clinically relevant dose of NESP (3.75 µg/kg) administered on day 0 and day 14 (23). These results confirm that the in vivo efficacy for Ab12 and Ab12.6 also inversely correlates with affinity.


Figure 7
View larger version (12K):
[in this window]
[in a new window]

  		FIGURE 7. Ab12.6 is more effective than Ab12 at elevating hematocrit. Male transgenic mice were dosed s.c. either with NESP on days 0 and 14 or Abs (Ab12, Ab12.6, and isotype control) on day 0 only at the concentrations indicated. Blood samples were collected on day 21, and hematocrit was measured using a Cell Dyn 3700 hematology analyzer (Abbott Laboratories). Error bars represent ± SE of 5 mice per treatment group. The NESP dose used translates into a typical dose used in clinical practice of 3 µg/kg every 2 wk (23 ).

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 References
 
Unlike other murine EPO-mimic Abs that have been described (24, 25), Ab12 is a human EPO agonistic Ab that is capable of stimulating and sustaining erythropoiesis in vivo. We sought to enhance the agonistic activity of Ab12 even further by improving the off-rate and, therefore, the affinity of the Ab to EPOR using yeast display affinity maturation. Yeast-displayed Ab12 variants containing CDRH2 amino acid substitutions were isolated at a high frequency and displayed improved off-rate kinetics on the surface of yeast. Several Ab12 variants were subsequently constructed as full-length IgG2/{kappa} Abs and were tested for affinity and potency to the EPOR.

Essentially two regions in CDRH2 were interrogated following the generation of full-length IgG: the YYS and TNY regions. Abs containing amino acid changes in the YYS region (Abs 12.6, 12.56, and 12.70) did not demonstrate improved off-rates, contrary to what was observed for the scFv version of these Abs on the surface of yeast. In fact, the affinities of Ab12.6 and Ab12.56 were significantly less than that of parental Ab12, while that of Ab70 was comparable. It is not clear why selected scFv mutants having CDRH2 mutations in the YYS region displayed improved off-rate kinetics on the surface of yeast, yet the corresponding full-length IgG2/{kappa} versions did not have improved affinity. One possibility is that these scFvs bound and orientated EPOR ECD in such a way as to make it more accessible to the fluorescent-tagged detection reagents, resulting in higher signals mistakenly interpreted as increases in the duration of the off-rate. These findings indicate that the selection criteria used to identify yeast clones with enhanced fluorescent binding may not always yield improved kinetics and ultimately translate into full-length Abs with higher affinity binding.

When constructed as full-length IgGs, however, Ab12.6 did display enhanced agonistic potency as measured by the ability to support both proliferation of EPO-dependent cell lines and formation of erythroid colonies. Importantly, in vivo results also indicated that Ab12.6 efficacy had been positively impacted by the amino acid substitutions in the YYS region of CDRH2. Other Abs containing substitutions in this same region also displayed increased maximal proliferative activity and potential to support greater erythroid colony formation than did Ab12. Conversely, the TNY Ab12 mutants identified as having slower off-rates of 10–30-fold (Ab12.17, Ab12.61 and Ab12.76) are less efficient in stimulating erythropoiesis, as measured in both in vitro assay formats. Collectively, these data suggest that improvements in agonistic activity for this class of Abs are inversely correlated with overall Ab affinity, as tight binding of Ab to receptor leads to reduced potency. Faster Ab off-rates from the receptor and subsequent reengagement may permit continuous, more efficient erythropoiesis.

Although improvements in Ab affinity generally correlate with improved function (7, 8, 9, 10, 11, 12, 13), we show herein that, at least for an agonistic EPOR Ab, the reverse is true. However, examples of antagonistic Abs that retain improved potency despite a faster off-rate have been described. Certain CD20-reactive human Abs are more effective at inducing complement-mediated lysis than the therapeutic rituximab, a chimeric Ab used in the treatment of non-Hodgkins lymphoma (26), despite faster dissociation rates (27). Interestingly, biological activity of these Abs is linked to recognition of a unique epitope. Crystal structure resolution of the EPOR ECD complexed to the Ab12.6 Fab fragment has also identified a unique binding site that is consistent with a novel mechanism of receptor activation based on a unique Ab-imposed conformational change (3). Recognition of novel epitopes may augment off-rates in determining the activity of some Abs.

In conclusion, we identified distinct CDRH2 sequences that independently and inversely affected affinity and potency. Identification of mutations in CDRH2 that influence the orientation of Ab12 on the EPOR complex may present an unexpected strategy for the design of EPO-mimic Abs with enhanced therapeutic value.


   Acknowledgments
 
We thank Dr. Constance Noguchi for providing breeding pairs of mouse EPOR–/–, human EPOR+ transgenic mice. We also acknowledge that the parental anti-EPOR Ab12 was generated at Amgen (Freemont, CA; formerly Abgenix). We thank Michael Roguska for guidance on yeast display, Jonathan Belk for assistance in generating Ab mutants, Nicolette Zielinski for participation in design of animal experiments, and Lisa A. Baker for assay support.


   Disclosures
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
References
 
All authors are employed by Abbott Laboratories.


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

1 Address correspondence and reprint requests to Dr. Edward B. Reilly, Abbott Laboratories, Department 4CD AP-31-4, 200 Abbott Park Road, Abbott Park, IL 60064. E-mail address: ed.reilly{at}abbott.com Back

2 Abbreviations used in this paper: EPO, erythropoietin; ECD, extracellular domain; EPOR, erythropoietin receptor; rHu-EPO, recombinant human erythropoietin; scFv, single chain variable fragment. Back

Received for publication December 27, 2007. Accepted for publication May 15, 2008.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 

  1. Graber, S. E., S. B. Krantz. 1978. Erythropoietin and the control of red blood cell production. Annu. Rev. Med. 29: 51-66. [Medline]
  2. Tong, E. M., A. R. Nissenson. 2001. Erythropoietin and anemia. Semin. Nephrol. 21: 190-203. [Medline]
  3. Liu, Z., V. S. Stoll, P. J. DeVries, C. G. Jakob, N. Xie, R. L. Simmer, S. E. Lacy, D. A. Egan, J. E. Harlan, R. R. Lesniewski, E. B. Reilly. 2007. A potent erythropoietin-mimicking human antibody interacts through a novel binding site. Blood 110: 2408-2413. [Abstract/Free Full Text]
  4. Verhelst, D., J. Rossert, N. Casadevall, A. Kruger, K. U. Eckardt, I. C. Macdougall. 2004. Treatment of erythropoietin-induced pure red cell aplasia: a retrospective study. Lancet 63: 1768-1771.
  5. Babcook, J. S., K. B. Leslie, O. A. Olsen, R. A. Salmon, J. W. Schrader. 1996. A novel strategy for generating monoclonal antibodies from single, isolated lymphocytes producing antibodies of defined specificities. Proc. Natl. Acad. Sci. USA 93: 7843-7848. [Abstract/Free Full Text]
  6. Philo, J. S., K. H. Aoki, T. Arakawa, L. O. Nahri, J. Wen. 1996. Dimerization of the extracellular domain of the erythropoietin (EPO) receptor by EPO: one high-affinity and one low-affinity interaction. Biochemistry 35: 1681-1691. [Medline]
  7. Maynard, J. A., C. B. Maassen, S. H. Leppla, K. Brasky, J. L. Patterson, B. L. Iverson, G. Georgiou. 2002. Protection against anthrax toxin by recombinant antibody fragments correlates with antigen affinity. Nat. Biotechnol. 20: 597-601. [Medline]
  8. Johnson, S., S. D. Griego, D. S. Pfarr, M. L. Doyle, R. Woods, D. Carlin, G. A. Prince, S. Koenig, J. F. Young, S. B. Dillon. 1999. A direct comparison of the activities of two humanized respiratory syncytial virus monoclonal antibodies: MEDI-493 and RSHZl9. J. Infect. Dis. 180: 35-40. [Medline]
  9. Barbas, C. F., III, D. Hu, N. Dunlop, L. Sawyer, D. Cababa, R. M. Hendry, P. L. Nara, D. R. Burton. 1994. In vitro evolution of a neutralizing human antibody to human immunodeficiency virus type 1 to enhance affinity and broaden strain cross-reactivity. Proc. Natl. Acad. Sci. USA 91: 3809-3813. [Abstract/Free Full Text]
  10. Wu, H., G. Beuerlein, Y. Nie, H. Smith, B. A. Lee, M. Hensler, W. D. Huse, J. D. Watkins. 1998. Stepwise in vitro affinity maturation of Vitaxin, an {alpha}vβ3-specific humanized mAb. Proc. Natl. Acad. Sci. USA 95: 6037-6042. [Abstract/Free Full Text]
  11. Adams, G. P., R. Schier, K. Marshall, E. J. Wolf, A. M. McCall, J. D. Marks, L. M. Weiner. 1998. Increased affinity leads to improved selective tumor delivery of single-chain Fv antibodies. Cancer Res. 58: 485-490. [Abstract/Free Full Text]
  12. Viti, F., L. Tarli, L. Giovannoni, L. Zardi, D. Neri. 1999. Increased binding affinity and valence of recombinant antibody fragments lead to improved targeting of tumoral angiogenesis. Cancer Res. 59: 347-352. [Abstract/Free Full Text]
  13. Wark, K. L., P. J. Hudson. 2006. Latest technologies for the enhancement of antibody affinity. Adv. Drug Delivery Rev. 58: 658-669.
  14. Boder, E. T., K. D. Wittrup. 1997. Yeast surface display for screening combinatorial polypeptide libraries. Nat. Biotechnol. 15: 553-557. [Medline]
  15. Johnson, D. L., S. A. Middleton, F. McMahon, F. P. Barbone, D. Kroon, E. Tsao, W. H. Lee, L. S. Mulcahy, L. K. Jollife. 1996. Refolding, purification, and characterization of human erythropoietin binding protein produced in Escherichia coli. Protein Expr. Purif. 7: 104-113. [Medline]
  16. Kumar, A., R. Arora, P. Kaur, V. S. Chauhan, P. Sharma. 1992. "Universal" T helper cell determinants enhance immunogenicity of a Plasmodium falciparum merozoite surface antigen peptide. J. Immunol. 148: 1499-1505. [Abstract]
  17. Chiba, S., F. Takaku, T. Tange, K. Shibuya, C. Misawa, K. Sasaki, K. Miyagawa, Y. Yazaki, H. Hirai. 1991. Establishment and erythroid differentiation of a cytokine-dependent human leukemic cell line F-36: a parental line requiring granulocyte-macrophage colony-stimulating factor or interleukin-3, and a subline requiring erythropoietin. Blood 78: 2261-2268. [Abstract/Free Full Text]
  18. Komatsu, N., M. Yamamoto, H. Fujita, A. Miwa, K. Hatake, T. Endo, H. Okano, T. Katsube, Y. Fukumaki, S. Sassa, Y. Miura. 1993. Establishment and characterization of an erythropoietin-dependent subline, UT-7/Epo, derived from human leukemia cell line, UT-7. Blood 82: 456-464. [Abstract/Free Full Text]
  19. Yu, X., C. S. Lin, F. Costantini, C. T. Noguchi. 2001. The human erythropoietin receptor gene rescues erythropoiesis and developmental defects in the erythropoietin receptor null mouse. Blood 98: 475-477. [Abstract/Free Full Text]
  20. Liu, Z. Y., K. Chin, C. T. Noguchi. 1994. Tissue specific expression of human erythropoietin receptor in transgenic mice. Dev. Biol. 166: 159-169. [Medline]
  21. Mizushima, S., S. Nagata. 1990. pEF-BOS, a powerful mammalian expression vector. Nucleic Acids Res. 18: 5322[Free Full Text]
  22. Egrie, J. C., J. K. Browne. 2001. Development and characterization of novel erythropoiesis stimulating protein (NESP). Br. J. Cancer 84: 3-10. [Medline]
  23. Glaspy, J. A., J. S. Jadeja, G. Justice, J. Kessler, D. Richards, L. Schwartzberg, N. S. Tchekmedyian, S. Armstrong, J. O'Byrne, G. Rossi, A. B. Colowick. 2002. Darbepoetin alfa given every 1 or 2 weeks alleviates anaemia associated with cancer chemotherapy. Br. J. Cancer 87: 268-276. [Medline]
  24. Schneider, H., W. Chaovapong, D. J. Matthews, C. Karkaria, R. T. Cass, H. Zhan, M. Boyle, T. Lorenzini, S. G. Elliott, L. B. Giebel. 1997. Homodimerization of erythropoietin receptor by a bivalent monoclonal antibody triggers cell proliferation and differentiation of erythroid precursors. Blood 89: 473-482. [Abstract/Free Full Text]
  25. Elliott, S., T. Lorenzini, D. Yanagihara, D. Chang, G. Elliott. 1996. Activation of the erythropoietin (EPO) receptor by bivalent anti-EPO receptor antibodies. J. Biol. Chem. 271: 24691-24697. [Abstract/Free Full Text]
  26. Johnson, P., M. Glennie. 2003. The mechanisms of action of rituximab in the elimination of tumor cells. Semin Oncol. 30: (1 Suppl. 2):3-8. [Medline]
  27. Teeling, J. L., W. J. Mackus, L. J. Wiegman, J. H. van den Brakel, S. A. Beers, R. R. French, T. van Meerten, S. Ebeling, T. Vink, J. W. Slootstra, et al 2006. The biological activity of human CD20 monoclonal antibodies is linked to unique epitopes on CD20. J. Immunol. 177: 362-371. [Abstract/Free Full Text]




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 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 Request Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lacy, S. E.
Right arrow Articles by Reilly, E. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lacy, S. E.
Right arrow Articles by Reilly, E. B.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS