PEDS Advance Access originally published online on June 7, 2006
Protein Engineering Design and Selection 2006 19(8):359-367; doi:10.1093/protein/gzl020
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Development of tumor targeting anti-MUC-1 multimer: effects of di-scFv unpaired cysteine location on PEGylation and tumor binding
Department of Internal Medicine, Division of Hematology/Oncology, Section of Radiodiagnosis and Therapy, University of California Davis Cancer Center 1508 Alhambra Boulevard, Sacramento, CA 95816, USA
1To whom correspondence should be addressed. Radiodiagnosis and Therapy, Molecular Cancer Institute, University of California Davis Medical Center, 1508 Alhambra Boulevard, Room 3100, Sacramento, CA 95816, USA E-mail: sjdenardo{at}ucdavis.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
MUC1 mucin expressed in epithelial cancer, such as prostate and breast, is aberrantly glycosylated providing unique targets for imaging and therapy. In order to create a broadly applicable construct to target these unique epitopes on metastatic cancer, we selected an antibody fragment (scFv) that binds both synthetic MUC1 core peptide and epithelial cancer cell-expressed MUC1, and developed a recombinant bivalent molecule (di-scFv). Genetically engineered modifications of the di-scFv were constructed to create five molecular versions, each having a free cysteine (di-scFv-c) at different locations for site-specific conjugation. The effects of the engineered cysteine in the varied sites were studied relative to tumor binding and polyethylene glycol-maleimide (PEG-Mal) conjugation (PEGylation). Escherichia coli production as well as binding to MUC1 core peptide, human tumor cell lines and human tumor biopsies, were comparable. However, the location of the engineered cysteine in these di-scFv-c did influence PEGylation efficiency of this free thiol; higher PEGylation efficiency occurred with this cysteine in the inter-scFv linkage. Di-scFv-c PEG, with the cysteine engineered after the fifth amino acid in the linker, was used as an example to demonstrate comparable antigen-binding to non-PEGylated di-scFv-c. In summary, novel anti-MUC1 di-scFv-c molecules can be efficiently produced, purified and conjugated by site-specific PEGylation without loss of immunoreactivity, thus providing flexible multidentate constructs for cancer-targeted imaging and therapy.
Keywords: di-scFv/MUC-1/PEGylation/pretargeting
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Normal epithelial mucin 1 (MUC1) is a heavily glycosylated glycoprotein that is expressed in most human glandular epithelium and plays a role in cell adhesion, cell signaling and immune responses (Lloyd et al., 1996
). The MUC1 has cytoplasm and transmembrane regions, and an extracellular region in which the core peptide is comprised of tandem repeats of 20 amino acids (PDTRPAPGSTAPPAHGVTSA). Considerable attention has been given to MUC1 as a cancer-associated antigen because it is expressed by most human epithelial cancers, including breast and prostate cancers (Burchell et al., 1987; Braun et al., 1999; Kirschenbaum et al., 1999; Schut et al., 2003). In contrast to hyperglycosylation of MUC1 in normal tissue, the extracellular domain of the protein is hypoglycosylated in cancers, and is characterized by truncated or incomplete oligosaccharide side chains (Gendler et al., 1990), resulting in exposure of cryptic peptide epitopes. Additionally, normal gland architecture is lost and thus MUC1 is ubiquitously expressed over the cell surfaces (Masaki et al., 2004). In combination with its 200–500 nm extension above the cell surface (Bramwell et al., 1986), these features make the aberrant MUC1 epitopes easily recognized, thus providing unique targets for imaging or therapy of many cancers.
During the past two decades, a number of anti-MUC1 monoclonal antibodies (MAbs) have been produced; the majority of them recognize epitopes in the tandem repeat region of the extracellular domain and several have been used in radioimmunotherapy patient trials (DeNardo, 1999). These MAb studies can be exemplified by the results with radiolabeled anti-MUC1 MAb, BrE3 which, at low Yttrium-90 dose levels in patients with advanced metastatic breast cancer, induced transient and partial responses. However, therapy with these MAbs was limited by development of human antibodies against this large mouse protein (HAMA) (DeNardo et al., 1997; Kramer et al., 1998), and marrow toxicity due to prolonged circulation of the intact radioactive antibody (Reilly et al., 1995; DeNardo, 1999; Richman and DeNardo, 2001). Antibody fragments have been proposed to address these problems that are inherent in systemic delivery of most radiolabeled MoAb therapy for solid tumors. The single-chain antibody fragment (scFv) consisting of the immunoglobulin heavy chain variable region (VH) and the light chain variable region (VL) connected by an amino-acidic linker can provide a 25–27 kDa binding unit. ScFv to several well-characterized tumor antigens have been developed and studied (Wu et al., 1996; Denton et al., 1997; Larson et al., 1997; Adams et al., 1998; Colcher et al., 1998; Desai et al., 1998; Henderikx et al., 1999; Winthrop et al., 1999; Takemura et al., 2000; Winthrop et al., 2003). Although better tumor penetration and decreased immunogenicity have been reported (Larson et al., 1997; Adams et al., 1998; Desai et al., 1998; Adams et al., 2001), rapid blood clearance and univalent antigen binding of scFv have limited tumor uptake and retention (Colcher et al., 1998). To overcome these limitations, both covalent dimeric scFv (di-scFv) and non-covalent diabody constructs have been studied; as bivalent molecules, they exhibited greater tumor targeting and retention than the smaller univalent scFv fragments (Wu et al., 1996; Denton et al., 1999; Santimaria et al., 2003; Arndt et al., 2004; Robinson et al., 2005) and may provide useful molecular modules for imaging and therapy.
PEGylation (covalent scFv conjugation to polyethylene glycol polymers) modification strategies have more recently been considered to overcome the effects of size and increase circulation time to enhance in vivo targeting and pretargeting. PEGylated scFv have increased blood residence time (Molineux, 2002) and reduced immunogenicity in vivo (Molineux, 2003). A recent pharmacokinetic study in mice of scFv PEGylated by random conjugation demonstrated up to 100-fold prolongation of blood residence time when compared with non-PEGylated scFv (Yang et al., 2003). Further investigations of scFv PEGylation have shown that the addition of an unpaired cysteine to scFv (scFv-c) for site-specific conjugation is more likely to preserve the active binding sites of the protein (Albrecht et al., 2004; Deiters et al., 2004; Natarajan et al., 2005).
Our previous studies revealed that an unpaired cysteine could be engineered at the C-terminus of scFv without impairing antigen binding or production (Albrecht et al., 2004). These scFv-c, when concentrated, generate covalent dimeric scFv by inter scFv-c disulfide linkage or when conjugated with bifunctional linkers such as maleimide-PEG-maleimide (PEG-(Mal)2) scFv-PEG-scFv was produced (Natarajan et al., 2005). Herein, we describe the use of an unpaired engineered cysteine for site-specific PEGylation of recombinantly linked anti-MUC1 di-scFv-c. These di-scFv-c can provide bivalent binding units having site-specific chelate conjugates for radioimaging or modules for the development of larger multivalent bispecific constructs. We report the selection of a high-affinity anti-MUC1 scFv, the generation of di-scFv-c and the effect of the location of the engineered cysteine on PEGylation efficiency and antigen-binding.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Synthetic MUC1 peptide and MUC1 expressed on cell lines
Synthetic 100mer MUC1 core peptide (MW 9335) corresponding to five tandem repeats was generously provided by Dr Olivera Finn (University of Pittsburgh, PA). The peptide was dissolved in H2O in a concentration of 50 mg/ml and stored at 4°C. MUC1-positive human cell lines MCF-7 from breast cancer and DU145 from prostate cancer were obtained from the American Type Culture Collection (ATCC, Manassas, VA). MCF-7 cells were grown in Iscove's Modified Dulbecco's medium with 5% fetal bovine serum (FBS) and DU145 cells were grown in RPMI 1640 medium with 10% FBS.
Antibodies as controls
The MAb B2729 (Biodiagnostics Inc., Edmonton, Canada) reacts with glycosylated MUC1 and recognizes a region of MUC1 core protein (Schut et al., 2003). Because this antibody binds with MUC1 antigen expressed in breast cancer (Grinstead et al., 2003) and in prostate cancer (Schut et al., 2003), it was used as a positive control in both ELISA and immunohistochemistry (IHC) in which breast and prostate cancer cells and tissues were analyzed. The MAb Lym-1 that was kindly provided by Dr Epstein (Peregine Pharmaceuticals Inc., Tustin, CA) was generated with nuclear material from Raji (human Burkett's lymphoma) cells (Epstein et al., 1987; Ottonello et al., 1996). Lym-1 does not react with MUC1 peptide or epithelial cancers and was used as negative control.
ELISA
ELISA was performed on MUC1 peptides, cell lysates and live cells. The 96-well plates were coated with the MUC1 peptide (1 µg/100 µl/well) or DU145 and MCF-7 cell lysates containing the cell membranes (100 µl/well). The latter was prepared as follows: cells in log-phase were harvested and homogenized using a homogenizer. The concentration of cell lysates was adjusted to 1 mg protein per ml. For ELISA on live cells, cells in log-phase were harvested, re-suspended in fresh medium and added into 96-well plate (1 x 106 cells/well). Medium was changed to BSA–PBS solution (PBS buffer containing 1.5% BSA) and all reagent and wash changes performed by spinning the cells down, maintaining cell integrity. All ELISA assays were performed at least three times in triplicate, following the approach described previously (Winthrop et al., 2003). Anti-DOTA scFv-c (C6-c) was used as the negative control and the signal induced by the di-scFv-c studied was expressed relative to the signal from that control.
Selection of anti-MUC1 scFv
Anti-MUC1 scFv clones were selected from an anti-MUC1 scFv phage library developed previously in our laboratory from immunized mice (Winthrop et al., 1999). Briefly, the phage were subjected to four rounds of affinity panning with decreasing amounts (100, 50, 10 and 1 µg) of the MUC1 peptide conjugated to biotinylated BSA–magnetic streptavidin beads in PBST-milk solution (0.2% Tween-20 plus 3% dry non-fat milk in PBS buffer). Phages selected in each round were re-infected into Escherichia coli strain TG1. After the fourth panning, the MUC1-bound phages were further selected using ELISA against the MUC1 peptide and against cell lysates. DNA sequencing was performed for scFv-phage clones with binding to both the peptide and cell lysates. The affinity of a selected scFv (D5) was measured using the BIACORE system (BIACORE Inc., Piscataway, NJ) with biotin-BSA-MUC1 peptide compared to biotin-BSA (negative control) immobilized on a streptavidin-functionalized sensor chip, BIACORE Sensor Chip SA. This was performed by injection at a protein concentration of 0.1 µg/ml for sufficient time to reach immobilization levels of 267 and 298 Resonance Units (RU) for biotin-BSA-MUC1 and biotin-BSA, respectively. ScFv was then injected over the functionalized surfaces at varying concentrations in a 2-fold dilution series starting at the highest concentration of 1.485 µg/ml (equivalent to 55 nM).
Recombination of anti-MUC1 di-scFv-c genes
The scFv D5 was selected for the construction of di-scFv-c molecules. The primers used for constructing five different anti-MUC1 di-scFv-c genes are as follows: MDP1, 5'-CAACGTGAAAAAATTATTATTCGC; MP2, 5'-GTAAATGAATTTTCTGTATGAGG; MDP3, 5'-TATAATTAGGTCTCgCCGCTGCCACCTCCGCCACATGAACCGCCTCCACCCCGTTTTATTTCCAACTTTG; MDP4, 5'-TATAATTAGGTCTCaGCGGCGGTGGCGGATCGGGTGGAGGCGGTTCAATGGCCCAGGTGAAACTGCAG; MDP5, 5'-TATAATTAGGTCTCcACAGCTGCCACCTCCGCCTGAACCGCCTCCACCCCGTTTTATTTCCAACTTT; MDP6, 5'-TATAATTAGGTCTCgCTGTGGCGGTGGCGGATCGGGTGGAGGCGGTTCAATGGCCCAGGTGAAACTGCAGC; MDP7, 5'-TATAATTAGGTCTCgCCGCTGCCACCTCCGCCTGAACCGCCTCCACCCCGTTTTATTTCCAACTTTG; MDP8, 5'-TATAATTAGGTCTCaGCGGCGGTGGCGGATCGTGTGGTGGAGGCGGTTCAATGGCCCAGGTGAAACTGCAG; MDP9, 5'-ATTAATATGCGGCCGCAGAGCCACCTCCGCCTGAACCGCCTCCACCCCGTTTTATTTCCAACTTTG.
These primers were designed based on the DNA sequences of the expression plasmids, the D5 scFv and/or the (GGGGS)4 linker, as well as the inserting sites of a TGT triplet encoding a cysteine. The reverse primers amplifying the 5'-terminal scFv gene contain a portion of linker-encoding sequences and a BsaI site, while the forward primers amplifying the 3'-terminal scFv gene contain the remaining portion of the linker sequences and the BsaI site. For each di-scFv-c construction, two intact D5 scFv gene fragments were amplified from the D5 scFv phage clone using corresponding primer pairs and Pfu DNA polymerase (Stratagene, La Jolla, CA). The amplified D5 scFv genes were digested using the restriction endonuclease BsaI. After purification, the restricted fragments were ligated to form the D5 di-scFv-c genes that were then cloned as SfiI/NotI fragments into the pCANTAB 5E vector, or the pCANTAB 5E-cysteine expression vector that had been engineered for the addition of a C-terminal cysteine to any protein expressed (Albrecht et al., 2004). These vectors introduced a C-terminal E Tag used for both immunoassays and protein purification.
Expression of anti-MUC1 di-scFv-c
The five D5 di-scFv-c genes were separately transfected into the E.coli strain HB2151. Colonies expressing the different di-scFv-c were selected using both PCR and ELISA. PCR of primers R1 and R2 (Amersham, Piscataway, NJ) was used to determine the size of di-scFv-c genes, and ELISA of unpurified protein from supernatant of culture was used to assess the binding of di-scFv-c proteins to MUC1 antigen. The colonies showing both correct DNA size and the best binding to MUC1 peptide were selected. Protein was expressed in larger amounts in shaker flasks following the procedure described previously (Lu et al., 2001). In brief, periplasmic extract was prepared by re-suspending bacterial pellet in TES solution (0.2 M Tris, 0.5 mM EDTA and 20% sucrose, pH 8.0) and gently shaking at 4°C for 45 min. After centrifugation at 10 000 g for 30 min, soluble di-scFv-c protein was purified from supernatant by an anti-E Tag affinity chromatography using the RPAS Purification Module (Amersham, Piscataway, NJ). The purified di-scFv-c proteins were analyzed in both 4–12% SDS–PAGE gel and 4–12% native gel electrophoresis followed by Coomassie blue staining. Western blot analysis (Takemura et al., 2000) was performed with HRP-conjugated anti-E Tag antibody. Further purification was performed by 1 x 32 cm Sephadex G75 in PBS column chromatography.
IHC
Formalin-fixed, paraffin-embedded human prostate cancer tissues were sectioned and slides were deparaffinized. Cultured cells were smeared onto slides, air dried and fixed in 10% formalin. Endogenous peroxidase in both fixed tissues and cells was quenched by submerging slides in 0.3% H2O2 in methanol for 15 min. For fixed tissues, antigen retrieval was performed using a microwave approach in sodium citrate buffer (10 mM, pH 6.0). After rinsing in PBS, all slides to be stained were blocked for 30 min with 10% goat serum in PBS. Then, D5 di-scFv-c proteins to be tested, MAb B2729 as positive control or Lym-1 as negative control, were separately added to corresponding slides followed by incubation for 3 h at 37°C. After rinsing, HRP-conjugated anti-E Tag antibody or HRP-conjugated anti-mouse IgG was added and incubated for 1 h at room temperature followed by rinsing in PBS. The peroxidase reaction was developed with 3,3'-diaminobenzidine (DAB) reagent (Vector Laboratories, Burlingame, CA). After rinsing, sections were counterstained with hematoxylin. Before photography, images of the slides were reviewed with a pathologist, Dr Linlang Gun, photographed and electronically saved.
PEGylation of di-scFv-c
PEGylation was performed following the method described previously (Natarajan et al., 2005). Briefly, the unpaired cysteine at the linker region or the C-terminus of di-scFv-c protein was reduced using tris(2-carboxyethy)phosphine (TCEP) prior to reaction with different masses of PEG-(Mal)2 (2 or 3.4 kDa) (Nektar Therapeutics, San Carlos, CA), or methoxy-PEG-maleimide (M-PEG-Mal) (5, 10, 20 or 40 kDa) (Sunbio PEG-Shop, Anyang City, South Korea). The reduced di-scFv-c proteins were diluted to 1.0 mg/ml using PEGylation reaction buffer (100 mM sodium phosphate, pH 7.0 and 2 mM EDTA). PEGylation was conducted overnight in N2 atmosphere at a reaction molar ratio of 5:1 (PEG:protein). The PEGylated di-scFv-c products were analyzed in SDS–PAGE gel, and quantitated using a Molecular Dynamics PD-SI laser scanner (Molecular Dynamics, PDSI, CA). Purification of selected PEGylated di-scFv-c was performed by Sephadex G75 column chromatography and the product analyzed by HPLC (SEC 3000).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Characterization of anti-MUC1 scFv immunoreactivity
In ELISA studies the scFv D5 demonstrated binding to MUC1 peptide and MCF-7 and DU145 cell lysates, with 10- to 14-fold increased signal over the negative control (Figure 1). Further confirmation of D5 scFv binding to MUC1 peptide was provided by BIACORE analysis in which studies on immobilized biotin-BSA showed no evidence of scFv protein binding, while on immobilized biotin-BSA-MUC1, D5 scFv binding data demonstrated dose–response curves analyzed by fitting to Langmuir interaction binding model, giving: kon = 4.76 x 105 M/s, koff = 7.08 x 10–4/s and binding affinity equilibrium constant (KD) = 1.49 nM. These parameters are similar to those reported for other scFv proteins (Goel et al., 2000; Peter et al., 2003).
|
DNA sequencing revealed that the D5 scFv gene is comprised of a VH chain of 118 residues, a (GGGGS)3 linker and a VL chain of 109 residues. VH and VL were separately compared with sequences of Ig genes posted in the National Center for Biotechnology Information (NCBI) database. Of 46 170 Ig genes, the DNA sequences identical to the D5 VH or VL chain were not found. Also, amino acid sequences identical to the D5 scFv were not found in the database of 2 259 584 proteins or peptides. However, three Ig heavy chains and eight Ig light chains share identical framework regions (FWRs) 1–3 and complementary-determining regions (CDR) 1 and 2 with the D5 VH chain and the D5 VL chain, respectively. Major divergences between the D5 scFv and these heavy and light chains posted in this database are in the CDR3 regions and demonstrate that D5 scFv is a unique scFv.
Design of anti-MUC1 di-scFv-c genes
In order to generate a bivalent, activity-retained construct with an unpaired cysteine for efficiently site-specific conjugation, five anti-MUC1 di-scFv-c genes were constructed from the D5 scFv using a PCR-based recombinant approach. These di-scFv-c gene variants were each designed to carry a TGT triplet encoding a free cysteine at different locations, allowing site-specific PEGylation to that region of the di-scFv-c. To distinguish the various locations of the unpaired cysteine insertion, they were called D5c5D5, D5c10D5, D5c15D5, D5D5/c and D5D5tail/c. Since an antigen-binding site may constitute about one-third of the Fv surface area, random PEG conjugation is likely to result in steric hindrance of that site; cysteine was therefore placed at various locations within the (GGGGS)4 linker that connects two scFvs or at the c-end of di-scFv protein. As illustrated in Figure 2A, D5c5D5, D5c10D5 and D5c15D5 carry the cysteine behind the first, second and third GGGGS of the (GGGGS)4 linker, respectively; D5D5/c contains the cysteine at its C-terminus; while D5D5tail/c has an insertion of (GGGGS)2-cysteine structure in its C-terminus. These di-scFv-c genes were inserted separately into the expression vectors for preparation of individual anti-MUC1 di-scFv-c clones. ELISA of at least 40 colonies from each construct was performed to select the clones that express di-scFv-c proteins with an affinity similar to or higher than that of the D5 scFv to synthetic and cell-expressed MUC1 (data not shown). Restriction digestion of plasmids extracted from these ELISA-selected clones revealed that the di-scFv genes have correct size of 1500 bp (Figure 2B).
|
Expression and purification of anti-MUC1 di-scFv-c protein
These D5 di-scFv-c proteins were expressed in the E.coli HB2151 and purified using anti-E Tag affinity chromatography. SDS–PAGE of these anti-E Tag purified di-scFv-c proteins without TCEP demonstrated some 98 kDa protein along with 52 kDa di-scFv-c; after TCEP treatment, the 98 kDa molecules were all reduced to di-scFv-c of 52 kDa, doubling the size of D5 scFv protein (Figure 3A).
|
Soluble D5 di-scFv-c protein purified from the bacterial periplasmic extracts by the anti-E Tag affinity chromatography provided
1–2 mg/l of bacterial culture for all five clones. Since some bacteria-expressed scFvs are reported to form aggregates which negatively influence their application in vivo (Le Gall et al., 2004; Kontermann, 2005), native gel analysis was performed to determine if there was aggregation of these purified di-scFv-c proteins. Similar to the SDS–PAGE analysis, all di-scFv-c proteins exhibit 52 kDa protein bands on the native gel in the presence of TCEP; no evidence of aggregation was observed (Figure 3B). It was noted that these affinity-purified di-scFv-c proteins were contaminated with two lesser proteins of 31 and 45 of kDa. These lower molecular weight proteins proved to be E Tag negative, as they are not seen in an anti-E Tag western blot (Figure 3A). They were, however, efficiently removed by means of G75 column chromatography. Figure 3C demonstrates Sephadex G75-purified D5c5D5 di-scFv-c protein with >95% molecular purity. Immunoreactivity of anti-MUC1 di-scFv-c protein
The ability of the di-scFv-c to bind with MUC1 peptide was determined using ELISA. In order to demonstrate that di-scFv-c proteins bind the extracellular domain of MUC1, assays with live DU145 and MCF-7 were performed. The di-scFv-c variants exhibited different binding activity. As shown in Figure 4, B2729 demonstrated the expected strong immunoactivity to MCF-7 cells and relative weak binding to MUC1 peptide and DU145 cells; all di-scFv-c were able to bind with synthetic or cell-expressed MUC1. D5c5D5 and D5D5/c had 5-fold more signal on MUC1 peptide than D5c15D5 and D5D5tail/c, but comparable results were obtained with the two cell lines with all five di-scFv-c and B2729. These results suggest that these di-scFv-c proteins can bind these cells comparable to the intact antibody. Additionally, we compared the immunoactivity of D5 scFv and D5c5D5 di-scFv-c binding to MUC1 peptide. The dimeric molecules exhibited increased reactivity (data not shown), reflecting a gain of immunoactivity due to binding to two epitope simultaneously.
|
The immunoactivity of anti-MUC1 di-scFv-c by IHC analysis on DU145 and MCF-7 cells is shown, Figure 5A; all the di-scFv-c bound well to both cancer cell lines, with strong brown staining of the cell membrane. In these experiments, the positive control antibody B2729 strongly stained the prostate and breast cancer cells, while the negative control antibody Lym-1 did not stain these tumor cells (Figure 5B). Additionally, five formalin-fixed prostate specimens with grade-3 tumor were immunostained using these di-scFvs (Figure 5A). Cancer cells in these prostate tissues were stained strongly, well above a weak background staining. The IHC results from cell lines and clinical prostate tissue indicate that the di-scFv-c proteins specifically bind to cancer cell membranes.
|
PEGylation of di-scFv-c
These five di-scFv-c proteins were conjugated with PEG-(Mal)2 (2 and 3.4 kDa) or M-PEG-Mal (5, 10, 20 and 40 kDa), and the variation in efficiency of PEGylation is shown in Table I. Representative results are illustrated in Figure 6A and B. It was observed that di-scFv-c prior to PEGylation had apparent molecular weight of 52 kDa, while the PEGylated di-scFv-c exhibit an increase in their molecular weight in accordance with the PEG mass conjugated to di-scFv-c. Conjugation of PEG-(Mal)2 of 2 kDa mass to di-scFv-c results in minor increase in the size of protein, while a protein band with an apparent 98 kDa molecular weight was visualized in the conjugate of di-scFv-c and M-PEG-Mal of 40 kDa mass. HPLC analysis of PEGylated D5c5D5 di-scFv-c was performed on D5c5D5 protein conjugated with 5 kDa M-PEG-Mal; a majority of di-scFv-c was PEGylated as demonstrated in the Figure 6A PAGE analysis of D5c5D5 reacted with 5 kDa M-PEG-Mal and the HPLC analysis of the same material in Figure 6C. PEGylated D5c5D5 protein separated from non-PEGylated protein by Sephadex G75 chromatography analyzed on SDS–PAGE revealed a highly pure di-scFv-PEG product (Figure 6D).
|
|
Notably higher PEGylation efficiency was obtained in di-scFv-c proteins having the engineered cysteine in the linker region rather than the C-terminus. This suggests that the location of the engineered cysteine in these molecules influences the conjugation. PEG(Mal)2 conjugated D5c5D5, D5c10D5 and D5c15D5 completely (Table I) but a second di-scFv-c is only very seldom conjugated to the same PEG(Mal)2 (<5%) in spite of the second Mal functional group (Figure 6A and B). In the D5c5D5 preparation, a smaller protein species (<5%) also formed a PEGylated product and thus was considered to be a breakdown product containing a free cysteine.
The immunoactivity of PEGylated di-scFv-c proteins was assessed using the highly PEGylated D5c5D5 in ELISA. PEGylated D5c5D5 proteins conjugated with the various PEG compounds were able to bind with synthetic MUC1 peptide and DU145 and MCF-7 cell lysates similar to that of non-PEGylated di-scFv-c (Figure 7). Additionally, IHC of both DU145 and MCF-7 cells using PEGylated D5c5D5 demonstrated that the membranes of these tumor cells were strongly stained (Figure 5B).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
In this study we report the development of di-scFv-c with an engineered specific site for conjugation by way of an unpaired cysteine at one of several locations. These di-scFv-c molecules were developed using a high-affinity antibody fragment, D5 scFv, selected from our anti-MUC1 scFv libraries (Winthrop et al., 1999
). The D5 scFv demonstrated specific binding to synthetic MUC1 with the KD of 1.49 nM, and excellent binding to cancer cells expressing MUC1. Adams et al. (1998, 2001) reported that antibody fragments having a threshold affinity of between 10–7 and 10–8 M meet the requirement for tumor targeting and penetration, as they bind tumor cells in vivo but are not restrictively retained by their initial encounter with perivascular tumor cells; Begent et al. (1996) and Larson et al. (1997) have used scFv with similar affinities to CEA or TAG 72 to image tumors in patients. D5 scFv, having appropriate affinity for MUC1 as well as easily demonstrable binding to MUC1 positive cell lines, was considered a good candidate for MUC1 tumor targeting. However, since univalent scFv fragments generally have rapid blood clearance, relatively poor tumor retention compared to MAb and unacceptably high kidney uptake (Buchegger et al., 1990; Li et al., 2002) limiting their clinical application, we have modified the D5 scFv into a larger unit with the potential for site-specific conjugation. Two monovalent scFv genes were linked so as to recombinantly make bivalent scFv dimers from which we constructed five D5 di-scFv-c gene variants differing only by the site of a single-engineered cysteine residue, thus enabling future site-specific modification or conjugation.
The most common scFv dimeric form previously studied has been the diabody form produced by shortening the length of the linker peptide connecting the VH and VL domains. Short linkers consisting of 3–12 residues result in failure to fold into functional scFv modules, but enhance the self-association (multimerization) of these non-functional scFv fragments into non-covalent, bivalent dimers (Todorovska et al., 2001). These diabodies, including the anti-MUC1 C595 (Denton et al., 1999) demonstrate antigen binding characteristics similar to the parental antibodies and/or improved tumor retention (Nielsen et al., 2000; Kipriyanov et al., 2002). However, this multimerization of short linker scFv is quite variable (Michaela et al., 2004) and such molecules often exhibit poor stability or aggregations (Willuda et al., 1999, 2001; Tahtis et al., 2001; Le Gall et al., 2004; Huang et al., 2005). In the present study, we used the 20-residue linker (GGGGS)4 to connect two anti MUC1 scFv fragments with or without a cysteine insertion, thus making covalently recombinant di-scFv. Linkers containing glycine (G) and serine (S) residues are common due to their flexibility, neutral change and protease resistance (Huston et al., 1988) and an association between their use and improvement of recombinant product yields has been suggested (Tang et al., 1996; Turner et al., 1997). We therefore assumed that the GGGGS linkers would not interfere with appropriate folding of di-scFv proteins within the bacterial periplasm, thus allowing formation of functional dimeric structures. Indeed, employing this linker led to effective expression of soluble D5 di-scFv proteins that do not appear to form non-covalent aggregates, as demonstrated on native gel and HPLC analyses. Furthermore, all di-scFv-c protein products were able to bind with MUC1 expressing cancer cells and synthetic peptide, demonstrating that the unpaired cysteine inserted in the linker or at the C-terminus does not inhibit binding of these anti-MUC1 di-scFv proteins.
All D5 di-scFv-c proteins were PEGylated with the panel of PEG-Mal, specifically conjugating the unpaired cysteine. The size increase of PEGylated di-scFv proteins was consistent with the mass addition of di-scFv-c with each specific Mal-PEG, and utilizing a 5:1 PEG:di-scFv ratio, 30–80% of the di-scFv-c were PEGylated with the uni-functional Mal-PEG; with the use of di-functional PEG-(Mal)2 and the same 5:1 ratio, 60–100% of the di-scFv-c was PEGylated, but <5% was conjugated as PEG-(di-scFv)2. Additionally, the location of unpaired cysteine affected the PEGylation efficiency. Insertion of cysteine in the linker yielded the best PEGylation efficiencies with the differences probably reflecting the accessibility of the engineered thiol. Unlike the four fully buried native cysteine residues, that provide two disulfide bonds for stabilizing the configuration of VL and VH (Padlan, 1994), the positions of the ‘unprotected’ engineered cysteine was selected to allow it to be exposed on the surface of molecule (Tsutsumi et al., 2000). Although in these di-scFv-PEG molecules, the thioether linkage between PEG-Mal and di-scFv is not buried in the protein, PEG-Mal PEGylated scFv have demonstrated prolongation of plasma half-lives in vivo when compared to non-PEGylated molecules, as well as increased plasma stability (Yang et al., 2003; Jo et al., 2006).
Since the engineered cysteine in the linker provided substantially better PEGylation efficiency, D5c5D5 was selected to determine whether such PEG conjugation to the linker of di-scFv-c influences the immunoactivity; the six variations of PEGylated D5c5D5 di-scFv-c were therefore analyzed for their MUC1-binding ability. These proteins demonstrated similar antigen binding compared with their non-PEGylated counterparts (Figure 7). PEGylation with up to 40 kDa PEG did not significantly reduce the immunoactivity of di-scFv-c, which was unexpected evidence that this engineered cysteine is positioned in an opportune location and at a sufficient distance to minimize PEG blocking the antigen-binding site. Thus, in these molecular configurations, the preferred attachment site appeared to be at this location within the (GGGGS)4 linker.
In summary, we have generated five di-scFv-c formats using an anti-MUC1 scFv molecule, and modified them by site-specific PEGylation. In vitro studies demonstrated the retention of binding of the PEGylated D5c5D5 with synthetic MUC1 peptides and tumor cells having abundant hypoglycosylated MUC1 expression. These studies of di-scFv-c and site-specific di-scFv-c PEGylation should provide useful information for development of new antitumor imaging and therapy agents. This specific novel antibody construct will be further studied as a modular component in a pretargeting strategy for radionuclide therapy.
|
Footnotes |
|---|
Edited by Paul Carter
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
National Cancer Institute Grant PO1 CA47829 supported this work. We thank Dr Linlang Gun for his help in evaluating the IHC in this study.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Adams G.P., Schier R., Marshall K., Wolf E.J., McCall A.M., Marks J.D., Weiner L.M. (1998) Cancer Res. 58:485–490.
Adams G.P., Schier R., McCall A.M., Simmons H.H., Horak E.M., Alpaugh R.K., Marks J.D., Weiner L.M. (2001) Cancer Res. 61:4750–4755.
Albrecht H., Burke P.A., Natarajan A., Xiong C.Y., Kalicinsky M., DeNardo G.L., DeNardo S.J. (2004) Bioconjug. Chem. 15:16–26.[CrossRef][ISI][Medline]
Arndt M.A., Krauss J., Rybak S.M. (2004) FEBS Lett. 578:257–261.[Medline]
Begent R.H., et al. (1996) Nat. Med. 2:979–984.[CrossRef][ISI][Medline]
Bramwell M.E., Wiseman G., Shotton D.M. (1986) J. Cell Sci. 86:249–261.
Braun S., Hepp F., Sommer H.L., Pantel K. (1999) Int. J. Cancer 84:1–5.[CrossRef][ISI][Medline]
Buchegger F., Pelegrin A., Delaloye B., Bischof-Delaloye A., Mach J.P. (1990) J. Nucl. Med. 31:1035–1044.
Burchell J., Gendler S., Taylor-Papadimitriou J., Girling A., Lewis A., Millis R., Lamport D. (1987) Cancer Res. 47:5476–5482.
Colcher D., Pavlinkova G., Beresford G., Booth B.J., Choudhury A., Batra S.K. (1998) Q. J. Nucl. Med. 42:225–241.[ISI][Medline]
Deiters A., Cropp T.A., Summerer D., Mukherji M., Schultz P.G. (2004) Bioorg. Med. Chem. Lett. 14:5743–5745.[Medline]
DeNardo S.J. (1999) Adv. Oncol. 15:23–29.
DeNardo S.J., Kramer E.L., O'Donnell R.T., Richman C.M., Salako Q.A., Shen S., Noz M., Glenn S.D., Ceriani R.L., DeNardo G.L. (1997) J. Nucl. Med. 38:1180–1185.
Denton G., Sekowski M., Spencer D.I.R., Hughes O.D.M., Murray A., Denley H., Tendler S.J.B., Price M.R. (1997) Br. J. Cancer 76:614–621.[Medline]
Denton G., Brady K., Lo B.K., Murray A., Graves C.R., Hughes O.D., Tendler S.J., Laughton C.A., Price M.R. (1999) Cancer Immunol. Immunother. 48:29–38.[CrossRef][ISI][Medline]
Desai S.A., Wang X., Noronha E.J., Kageshita T., Ferrone S. (1998) Cancer Res. 58:2417–2415.
Epstein A.L., Marder R.J., Winter J.N., Stathopoulos E., Chen F.M., Parker J.W., Taylor C.R. (1987) Cancer Res. 47:830–840.
Gendler S.J., Lancaster C.A., Taylor-Papadimitriou J., Duhig T., Peat N., Burchell J., Pemberton L., Lalani E.-N., Wilson D. (1990) J. Biol. Chem. 265:15286–15293.
Goel A., Colcher D., Baranowska-Kortylewicz J., Augustine S., Booth B.J.M., Pavlinkova G., Batra S.K. (2000) Cancer Res. 60:6964–6971.
Grinstead J.S., Schuman J.T., Campbell A.P. (2003) Biochemistry 42:14293–14305.[CrossRef][Medline]
Henderikx P., Kandilogiannaki M., Petrarca C., von Mensdorff-Pouilly S., Hilgers J.H., Krambovitis E., Arends J.W., Hoogenboom H.R. (1999) Cancer Res. 58:4324–4332.
Huang B.C., Foote L.J., Lankford T.K., Davern S.M., McKeown C.K., Kennel S.J. (2005) Biochem. Biophys. Res. Commun. 327:999–1005.[CrossRef][Medline]
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. (1988) Proc. Natl Acad. Sci. USA 85:5879–5883.
Jo Y., Youn Y., Lee S., Kim B., Kang S., Yoo M., Chol E., Lee K. (2006) Int. J. Pharm. 309:87–93.[Medline]
Kipriyanov S.M., Cochlovius B., Schafer H.J., Moldenhauer G., Bahre A., Le Gall F., Knaackmuss S., Little M. (2002) J. Immunol. 169:137–144.
Kirschenbaum A., Itzkowitz S.H., Wang J.P., Eliashvili M., Levine A.C. (1999) Mol. Urol. 3:163–168.[Medline]
Kontermann R.E. (2005) Acta Pharmacol. Sin. 26:1–9.[CrossRef][ISI][Medline]
Kramer E.L., Liebes L., Wasserheit C., Noz M.E., Blank E.W., Zabalegui A., Melamed J., Furmanski P., Peterson J.A., Ceriani R.L. (1998) Clin. Cancer Res. 4:1679–1688.[Abstract]
Larson S.M., et al. (1997) Cancer Suppl. 80:2458–2468.
Le Gall F., Reusch U., Little M., Kipriyanov S.M. (2004) Protein Eng. Des. Sel. 17:357–366.
Li L., Olafsen T., Anderson A.L., Wu A., Raubitschek A.A., Shively J.E. (2002) Bioconjug. Chem. 13:985–995.[CrossRef][ISI][Medline]
Lloyd K.O., Burchell J., Kudryashov V., Yin B.W.T., Taylor-Papadimitriou J. (1996) J. Biol. Chem. 271:33325–33334.
Lu D., Jimenez X., Zhang H., Wu Y., Bohlen P., Witte L. (2001) Cancer Res. 61:7002–7008.
Masaki I., Hee-Young I., James S., Norris S.T. (2004) Immunology 111:291–297.[CrossRef][Medline]
Michaela A.E., Arndt J.K., Susanna M.R. (2004) FEBS Lett. 538:257–261.
Molineux G. (2002) Cancer Treat. Rev. 28:13–16.[Medline]
Molineux G. (2003) Pharmacotherapy 23:3S–8S.[CrossRef][ISI][Medline]
Natarajan A., Xiong C.Y., Albrecht H., DeNardo G.L., DeNardo S.J. (2005) Bioconjug. Chem. 16:113–121.[CrossRef][Medline]
Nielsen U.B., Adams G.P., Weiner L.M., Marks J.D. (2000) Cancer Res. 60:6434–6440.
Ottonello L., Morone P., Dapino P., Dallegri F. (1996) Blood 87:5171–5178.
Padlan E.A. (1994) Mol. Immunol. 31:169–217.[CrossRef][ISI][Medline]
Peter J.C., Eftekhari P., Billiald P., Wallukat G., Hoebeke J. (2003) J. Biol. Chem. 278:36740–36747.
Reilly R.M., Sandhu J., Alvarez-Diez T.M., Gallinger S., Kirsh J., Stern H. (1995) Clin. Pharmacokinet. 28:126–142.[ISI][Medline]
Richman C.M. and DeNardo S.J. (2001) Crit. Rev. Oncol. Hematol. 38:25–35.[Medline]
Robinson M.K., Doss M., Shaller C., Narayanan D., Marks J.D., Adler L.P., Gonzalez Trotter D.E., Adams G.P. (2005) Cancer Res. 65:1471–1478.
Santimaria M., et al. (2003) Clin. Cancer Res. 9:571–579.
Schut I.C., Waterfall P.M., Ross M., Miller W.R., Habib F.K., Bayne C.W. (2003) Br. J. Urol. Int. 91:278–283.
Tahtis K., Lee F.T., Smyth F.E., Power B.E., Renner C., Brechbiel M.W., Old L.J., Hudson P.J., Scott A.M. (2001) Clin. Cancer Res. 7:1061–1072.
Takemura S.I., Asano R., Tsumoto K., Kodama H., Yoshida H., Suzuki H., Imai K., Matsuno S., Kudo T., Kumagai I. (2000) Protein Eng. 13:583–588.
Tang Y., Jiang N., Parakh C., Hilvert D. (1996) J. Biol. Chem. 271:15682–15686.
Todorovska A., Roovers R.C., Dolezal O., Kortt A.A., Hoogenboom H.R., Hudson P.J. (2001) J. Immunol. Methods 248:47–66.[CrossRef][ISI][Medline]
Tsutsumi Y., Onda M., Nagata S., Lee B., Kreitman R.J., Pastan I. (2000) Proc. Natl Acad. Sci. USA 97:8548–8553.
Turner D.J., Ritter M.A., George A.J.T. (1997) J. Immunol. Methods 205:43–54.[CrossRef][ISI][Medline]
Willuda J., Honegger A., Waibel R., Schubiger P.A., Stahel R., Zangemeister U., Pluckthun A. (1999) Cancer Res. 59:5758–5767.
Willuda J., Kubetzko S., Waibel R., Schubiger P.A., Zangemeister-Wittke U., Pluckthun A. (2001) J. Biol. Chem. 17:14385–14392.
Winthrop M.D., DeNardo S.J., DeNardo G.L. (1999) Clin. Cancer Res. 10:3088–3094.
Winthrop M.D., DeNardo S.J., Albrecht H., Mirick G.R., Kroger L.A., Lamborn K.R., Venclovas C., Colvin M.E., Burke P.A., DeNardo G.L. (2003) Clin. Cancer Res. 9:3845s–3853s.[Medline]
Wu A.M., Chen W., Raubitschek L., Williams L., Neumaier M., Fischer R., Hu S.Z., Odom-Maryon T., Wong J., Shively J.E. (1996) Immunotechnology 2:21–36.[CrossRef][ISI][Medline]
Yang K., et al. (2003) Protein Eng. 16:761–770.
Received October 31, 2005; revised March 21, 2006; accepted March 27, 2006.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||