Development of Anti-MUC1 Di-scFvs for Molecular Targeting of Epithelial Cancers, Such As Breast and Prostate Cancers

Written By: Sam Savage

Published Date: January 12, 2008 Last Edited: January 12, 2008

By Albrecht, H Denardo, G L; Denardo, S J

Pretargeted radioimmunotherapy (RIT) is a promising approach to increase the therapeutic index of RIT for malignant solid tumors. For pretargeted RIT of epithelial cancers, such as breast and prostate, mucin 1 (MUC1), the epithelial mucin, was chosen as a target antigen (Ag). Overexpression, hypoglycosylation and loss of apical distribution on the cellular membrane distinguish the tumor associated MUC1 from normal MUC1. These characteristics of MUC1, best known in breast cancer, were validated in prostate cancer. The multivalent bispecific MUC1 pretargeting molecule under development consists of a tumor binding module and a radioactive hapten capturing module. The building blocks of each module were chosen as single chain antibody fragments (scFv) to be covalently attached to a multifunctional polyethylene glycol (PEG) scaffold. PEGylation studies with scFvs selected from anti-MUC1 libraries and engineered with a free thiol for site-specific conjugation showed that highest reaction yields were obtained with short monofunctional PEG molecules. To accommodate the use of a bifunctional PEG for covalent assembly of binding and capturing modules, the MUC1 binding module was developed into a di-scFv-SH format and optimized for linker length and location of the free thiol in respect to Ag binding and site-specific conjugation. Approaches under study to improve PEGylation yields with bifunctional PEG molecules include alkyne- azide cycloaddition. Assembly efficiencies, through PEGylation, of the binding and capturing modules and pharmacokinetics will influence the final valency of the MUC1 pretargeting molecule: anti- MUC1 di-scFv-PEG- anti-radioactive hapten scFv or di-scFv-PEG-anti- radioactive hapten di-scFv. KEY WORDS: Drug labelling – Breast neoplasms – Prostatic neoplasms.

Although the first radioimmunotherapy (BIT) experiment was conducted in the early 1950’s with ^sup 131^I-labeled rabbit antiserum in rats bearing osteosarcoma xenografts,1 clinical interest in the use of antibodies increased with the development of the hybridoma technique by Kohler etal. in 1975.2 The ability to produce gram amounts of pure murine monoclonal antibodies (MAbs), that target a single epitope, greatly facilitated the use of antibodies and contributed to the identification of tumor associated antigens (TAA). The targeted delivery of cytotoxic radioactive doses to tumor tissues, using systemically delivered MAbs, known as RIT, appeared as a novel and promising cancer therapy in the 1980’s. Following the early clinical use of radiolabeled MAbs in imaging studies,3 the efficacy of RIT in the treatment of cancer has been widely investigated in animal models and patients.4-5 Such trials brought forward two major limitations of RIT: 1) development of an immune response, in some patients, against MAbs of mouse origin, known as human anti-mouse antibodies; and 2) normal tissue radiation toxicity, particularly bone marrow, as a consequence of the long circulation time of radiolabeled MAbs.

The immune response triggered by MAbs of mouse origin was bypassed by their substitution with chimeric, humanized and fully human antibodies, as well as recombinant antibody fragments. This was made possible through advances in molecular immunology and the development and adaptation of a plethora of molecular biology techniques to genetically engineer antibodies. The adaptation of the peptide phage display technique to display single chain (sc)Fva6 was a major breakthrough in recombinant antibody technology. The power of phage display, and subsequently developed display techniques, resides in the fact that they provide a direct link between phenotype and genotype. This greatly facilitates the selection and cloning of an antibody fragment against a given antigen (Ag). Two routes are currently used for the production of human antibodies: cloning of human Fv fragments onto human constant regions of immunoglobulins (Ig) and immunization of transgenic mice engineered to express human IgGs.7

The efficacy of RTT in the treatment of non-Hodgkin’s lymphoma has led to approval by the Food and Drug Administration of two radiolabeled anti-CD20 MAbs: Zevalin ([90Y]ibritumomab) (Idee Pharmaceuticals, San Diego, CA, USA) and Bexxar ([^Ijtositumomab) (Corixa, Seattle, WA, USA). In contrast to solid tumors, where radiolabeled antibodies have to penetrate into tissue, hematological malignancies provide better access of the radiolabeled antibodies to tumor cells and are more radiosensitive. Therefore, the radioactive dose delivered to the tumor is greater than that received by normal tissues/organs and results in a high therapeutic index (TO, defined as radioactivity to tumor divided by radioactivity to normal tissues. Therefore, successful RIT of solid tumors requires a higher TI. A number of strategies, including the optimization of the pharmacokinetics of targeting antibodies,8 cleavable linkers 9 and combined modality RIT10 have been devised to achieve this goal. One of the most promising approaches is pretargeting, in which the pharmacokinetics of the targeting antibody are dissociated from those of the radionuclide through separate injections.11 Greatly enhanced TI has been shown in animal models and in patients for two extensively studied pretargeting systems: 1) antibody fragment- streptavidin (SA) fusion protein with radiolabeled biotin; and 2) bispecific anti-target Ag and antiradioactive hapten antibodies.12, 13 The advantages of the SA/biotin based system are twofold: 1) high functional affinity for the target Ag is achieved through SA tetramerization; and 2) high affinity of biotin for SA provides an efficient capture of the radionuclide. However, SA and its analogues are immunogenic. Bispecific antibodies, on the other hand, humanized or human, have been developed in various formats, sizes and valencies,14-15 that contribute to the optimization of their pharmacokinetics.

Modular designs appear as a good strategy to overcome low production yields of multivalent bispecific antibodies. Here, we report the experimental path followed for the development of modular multivalent bispecific pretargeting molecules for RIT of metastatic breast and prostate cancers.

MUC1 as a target antigen

Mucin 1 (MUC1), also known as polymorphic epithelial mucin, polymorphic urinary mucin, episialin, DF3 Ag, epithelial membrane antigen and CA15-3, is a member of the mucin family that is represented by large molecular weight (MW) glycoproteins. According to their gene sequences, mucins fall into one of two groups: secreted or transmembrane mucins. The MUC1 gene encodes a transmembrane mucin characterized by the presence of 3 domains: extracellular, transmembrane and cytoplasmic.16 Beside genetic, MUC1 polymorphism is created at the messenger ribonucleic acid (mRNA) level through alternative splicing and at the protein level through glycosylation.

MUC1, initially synthesized as a single polypeptide, functions as a non-covalent heterodimer.17 The N terminal subunit consists of the extracellular domain, composed of a variable number (20-100) of 20 aa tandem repeats (VNTR).18 The VNTR motif, GVTSAPDTRPAPGSTAPDAH, can carry up to 5 Oglycosyl chains since serine and threonine residues are the targets of O-glycosylation.19 The MUC1 C terminal subunit consists of a short extracellular segment followed by a transmembrane domain and cytoplasmic tail.16 Dimerization of the MUC1 N and C terminal subunits leads to tethering to the cell membrane of the large MUC1 extracellular domain 17 (Figure 1). Its rodlike structure extends over 100-200 nm above the cell surface which exceeds by 5 to 10 fold the length of most membrane proteins.19 In normal glandular tissues, MUC1 is expressed, as a heavily glycosylated protein, on the apical borders of normal secretory epithelial cells.20 By contrast in neoplastic tissue, MUC1 is overexpressed and underglycosylated and, as a consequence of the loss of gland structure, MUC1 expression is no longer restricted to the apical borders of epithelial cells 21 (Figure 1).

The tumor associated MUC1 fulfills many characteristics of an ideal TAA:22 it is expressed on almost all human epithelial cell adenocarcinomas (breast, pancreas, ovary, lung, urinary bladder, prostate and endometrium), thus in nearly 80% of all human tumors; TAA MUC1 is distinct from MUC1 present on normal cells; TAA MUC1 is more abundant than normal MUC1.

Assessment of target antigen in prostate cancer

Expression of hypoglycosylated MUC1 epitopes (less O-glycans) is best documented in breast cancer, where in comparison to normal tissue, more staining was observed with anti-MUC1 MAbs recognizing hypoglycosylated forms of MUC1.23-24 Not all cancers express high levels of hypoglycosylated MUC1 epitopes and cancer progression is not always associated with this MUC1 form.25, 26

In primary prostate cancer, gene expression profiling showed that 3 cancer subtypes can be distinguished and that MUC1 is expressed in the two most aggressive subtypes.27 No differential expression of the MUC1 gene was reported upon comparison of gene expression in primary and metastatic prostate cancer tissues.28 The detection of MUC1 in normal and malignant prostate epithelia with a MAb recognizing a cytoplasmic epitope led to the conclusion that MUC1 expression is heterogeneous in both tissues.29 However, a correlation between the levels of sialylated MUC1 and the histological grade and clinical stage of the prostate cancer was observed by using a MAb specific for sialylated MUC1.3[degrees] In a study comparing the reactivity of normal, primary and metastatic prostate cancer tissues with a MAb binding to hypoglycosylated MUC1, an increase of MUC1 epitopes and hypoglycosylated MUC1 forms was found on higher grade prostate cancer.31 In order to obtain a better insight on MUC1 expression in prostate cancer, epitope mapping of the MUC1 extracellular domain was carried out, but unlike previous studies that used one or two MAbs, 7 well characterized anti-MUC1 MAbs were used. Immunohistochemistry (IHC), performed on normal and prostate cancer tissues with increasing Gleason grades present on a microarray, showed an increase of MUC1 epitopes and hypoglycosylated forms of MUC1 with increasing Gleason grade (Figure 2).32 IHC data also showed that most, but not all, prostate cancer samples were positive for MUC1, in agreement with previous findings.27- 3i Thus, according to our data and that of others,*1 the targeting of hypoglycosylated forms of MUC1 for imaging and pretargeted RIT of aggressive prostate cancer is appropriate. Figure 1.- Characteristics of normal and aberrant mucin 1 (MUC1). Schematic representation of MUC1 protein expressed on normal and malignant cells of epithelial origin. 1: extracellular domain of MUC1 protein with O-glycosylations on serine (S) and threonine (T) residues of the variable number tandem repeats motif. Note hypoglycosylation on aberrant MUC1; 2: transmembrane domain of MUC1 protein; and 3: cytoplasmic domain of MUC1 protein.

Figure 2.-Immunohistochemistry (IHC) on prostate tissue with antimucin 1 (MUC1) single chain antibody fragments (scFvs). Reactivity of two anti-MUC1 scFvs and two well characterized anti- MUC1 murine monoclonal antibodies (MAbs) (BrE3 recognizes the TRP epitope on moderately and hypoglycosylated MUC1 and B27.29 reacts preferentially with hyperglycosylated/normal MUC1) tested by EHC on normal to benign (A), prostatic intraepithelial and Gleason grade 1 to 2 (B) and Gleason grade 3 to 5 (C) prostate tissue cores (0.6 mm).

Selection of anti-MUC1 scFvs

Numerous anti-MUC1 MAbs have been generated against Ags, including human milk fat globule (HMFG), membranes of tumor cells, peptides and oligosaccharides. The investigation of the reactivity and specificity of 56 MAbs against the MUC1 glycoprotein led to the following conclusions: 34 MAbs defined epitopes located within the VNTR motif; 16 MAbs showed evidence for involvement of carbohydrate residues in their epitopes; no obvious relationship was found between the type of immunogen and the specificity of each antibody; the hydrophilic sequence of PDTRPAP was always present either in part or full in epitopes within the MUC protein core.33 The presence of the PDTRPAP sequence in peptidic epitopes is in agreement with the immunodominant peptide region in each MUC1 VNTR motif that was identified by nuclear magnetic resonance and referred to as the “PDTR” knob.34 Another anti-MUCl MAb, BrE3, raised against HMFG and reactive with the TRP epitope 35 has been used in breast cancer patients for imaging and RIT.36

Based on this information, a MUC1 peptide containing 4 VNTR was chosen as an Ag for the preparation of hyperimmune anti-MUCl scFv phage display libraries. The choice of the scFv format for the phage display library was guided by the fact that a monovalent 25-30 KD scFv appeared as a suitable module for engineering multivalent targeting molecules and that the technology to construct such a library was available.37

Two anti-MUC1 scFv phage display libraries were constructed by using the RPAS mouse scFv module (Amersham Biosciences Corp., Piscataway, NJ, USA) and mRNAs extracted from the spleens of BALB/c or NZB immunized mice (cell membrane enriched lysates from MUC1 expressing MCF7 and HBT 3344 tumor cells at a ratio of 10:1 followed by 3 injections at 3 week intervals of the 80 mer MUC1 peptide conjugated to KLH). The initial BALB/c scFv phage library contained 107 clones with an estimated diversity of 105.38

Cloning of VH and VL domains of Igs with the RPAS mouse scFv module results in the insertion of DNA sequences corresponding to scFvs in the VH-VL orientation with a (G^sub 4^S)^sub 3^ peptide linker into the pCANTAB 5E expression vector (Figure 3). ScFvs expressed from this vector carry a N terminal signal peptide for periplasmic export and a C terminal E Tag for purification by immunoaffinity chromatography.

Anti-MUC1 scFvs were selected by 3 rounds of affinity selection with decreasing amounts (100, 50 and 10 nM) of MUC1 peptide conjugated to biotinylated bovine serum albumin and captured with magnetic SA coated beads. Further selection was carried out in ELISA assays with scFv containing periplasmic extracts of individual clones. ScFvs that bound to the MUC1 synthetic peptide and lysates of MCF7 (human breast cancer) and DU145 (human prostate cancer) cells were subjected to DNA sequencing. Sequence analyses showed that the selected anti-MUC1 scFvs share 70% or more sequence homology at the amino acid level. Production yields of soluble scFvs purified from periplasmic extracts varied from 0.1 mg/L to 1.5 mg/ L. Affinity constants of these scFvs, for binding to the MUC1 peptide, determined by Scatchard and linear regression ranged between 1.7 x 10^sup 8^ and 8.2 x 10^sup 8^ M^sup -1^.39 Such scFv binding affinities are satisfactory for targeting Ags associated with solid tumors knowing that threshold affinities of 10^sup 7^- 10^sup 8^ M^sup -1^ are required for visualization of ^sup 125^I- scFv tumor uptake and that no gain in tumor accumulation is observed with affinities above 10^sup 9^ M^sup -1^.40

Insertion of a cysteine for site-specific PEGylation

The attachment of scFv modules to a scaffold like PEG can be random or site-specific. Conjugation by random targeting of primary amines of lysine residues or free thiols introduced via the use of 2 iminothiolane has been satisfactory with large molecules, such as antibodies. For smaller proteins and peptides, random conjugation often results in reduction or loss of biological activity.41 Furthermore, random conjugation also generates a heterogeneous final product, an undesirable quality for a pharmaceutical. On the other hand, conjugation at a specific site, remote from the molecule’s active site, limits interferences with biological activity and leads to homogeneous conjugates. Site-specific conjugation based on thiol chemistry appeared applicable to scFvs for the following reasons: most scFvs are devoid of free thiols and the intradomain disulfide bridges formed between the 2 cysteines present in each VH and VL domains are fully buried 42 and a free thiol can be provided through insertion of a cysteine at any chosen location by insertinal DNA mutagenesis. Because we were interested in adding a free thiol to more than 1 scFv, a modification of the expression vector seemed more appropriate. Thus, N and C termini of a scFv remained as the only two possible locations for the insertion of an extra-cysteine. The presence in a scFv of unstructured hydrophilic N and C termini, as confirmed by 3D modeling,39 supported accessibility to the free SH at either scFv end. However, the scFv C terminus appeared as the better choice since it already contains an additional E Tag sequence and other scFvs with a C terminal cysteine insertion have retained activity.43-45 A cysteine specifying codon (TGT) was added by PCR directed mutagenesis to the pCANTAB 5E vector backbone, downstream of the Sfil/NotI restriction sites used for scFv cloning and upstream of the E tag sequence used for scFv purification by affinity chromatography (Figure 3); this version of pCANTAB 5E, engineered for the addition of a C terminal cysteine to any scFv expressed from it, was named pCANTAB 5E Cys.46 In comparison to non modified scFvs, production yields in shake flasks of soluble scFv cys purified from bacterial periplasmic extracts were usually decreased by one half, whereas their Ag binding activities were increased twofold. The tendency of scFv cys to leak into the culture medium accounts for their yield reductions and their improved Ag binding activities reflect the increased functional affinity resulting from the formation of covalent scFv homodimers or (scFv’)^sub 2^. In the absence of a reducing agent, such as tris(2- carboxyethy)phosphine (TCEP), monomers and dimers of scFv cys coexist at approximately similar ratios.45, 46 Each of the dual forms of scFv cys has an application. In the absence of a reducing agent, scFv-S-S-scFv dimers with enhanced Ag binding activity provide higher sensitivity to Ag binding assays; this is particularly useful for immunohistochemistry.46 For site-specific conjugation, 100% of the free thiols are made available through conversion of dimers to scFv-SH monomers by addition of a reducing agent.46

Figure 3.-Schematic illustration of the steps involved in the development of multivalent pretargeting molecules.

Format of MUC1 binding modules: covalent and bivalent di-scFvs

Initial attempts for site-specific conjugation of scFv cys modules onto a bifunctional PEG scaffold were met with 45% of (scFv)^sub 2^-PEG product. Site-specific PEGylation yields with monofunctional PEG up to 80% have also been reported by others.48 In addition to PEGylation yield decrease with increased numbers of functional groups per PEG molecule, the synthesis of homogeneous multifunctional PEG is also difficult. Thus, although attractive in theory, the multimerization of 3 or 4 scFv cys modules through PEGylation remains elusive in practice. Since pharmacokinetic and optimal tumor Ag binding considerations contributed to the original design of the MUC1 pretargeting molecule, its design could only change in terms of building blocks, but not of mode of action. In other words, the MUC1 pretargeting molecule should retain a MW >70 KD to circumvent glomerular filtration and provide bivalent binding to gain functional affinity for the targeted tumor Ag. Hence, instead of scFv cys modules and trior tetra-functionalized PEG, the format of the tumor binding module was revised to accommodate the use of a bifunctional PEG (Figure 3). The most popular bivalent scFv format is the diabody, a non-covalent dimer obtained by shortening the length (

Figure 4.-Immunoreactivity of anti-mucin 1 (MUC1) di-single chain antibody fragments (scFv)-SH and di-scFv-PEG. A) Immunoreactivity of anti-MUC1 di-scFv-SH (D5c5D5) tested by immunohistochemistry (IHC) on MUC1 expressing breast (MCF7) and prostate (DU145) cancer cells. B) Immunoreactivity of anti-MUC1 di-scFv-PEG conjugate (D5c5D5-PEG) tested by IHC on MUC1 expressing breast (MCF7) and prostate (DU145) cancer cells in comparison to that of a MUC1 negative control MAb (Lym-1 reacts with HLA-DR determinants). Bars represent 20 mm.

Immunoreactivity of anti-MUC1 di-scFvs

Immunoreactivity of an anti-MUC1 scFv in the selected format (scFv-G^sub 4^S-C-(G^sub 4^S)^sub 3^-scFv) for the di-scFv-SH module was assessed in vitro and in vivo. In vitro binding to MUC1 peptide and lysates of MUC1 expressing cells was demonstrated in ELISA and IHC experiments showed that the di-scFv-SH (D5c5D5) bound to membranes of MUC1 expressing tumor cells (Figure 4A). For in vivo tumor binding evaluation, the di-scFv-SH (D5c5D5) was conjugated to [niIn]DOTA-maleimide and the purified [111In]DOTA-D5c5D5 conjugate, with an apparent MW of 52 KD, was injected into a mouse carrying bilateral subcutaneous DU145 tumor xenografts on its abdominal wall. Tumor targeting was visualized by y camera imaging (Figure 5).

Immunoreactivity of PEGylated MUC1 binding module was assessed in vitro. The anti-MUCl di-scFv-SH (D5c5D5) was PEGylated with a 5 KD linear bifunctional (methoxy and maleimide active groups) PEG and the di-scFv-PEG product, purified by gel exclusion chromatography, was tested for its MUC1 binding activity. ELISA assays, with MUC1 peptide, MCF7 and DU145 cell lysates as Ags, showed that the immunoreactivity of the di-scFv-PEG was comparable to that of di- scFv-SH, the non-PEGylated counterpart. This result was substantiated by IHC on MUC1 expressing tumor cells (Figures 4A, B).

Figure 5.-In vivo tumor imaging with an anti-mucin 1 (MUC1) di- single chain antibody fragment (scFv). DU145 tumor xenograft gamma camera image. The arrows point to [111n]DOTA-anti-MUC1 di-scFv-SH (D5c5D5) uptake by tumors bilaterally implanted in the abdominal wall of a nude mouse, as visualized 1 day after tail vein injection. The DU145 tumor xenograft best visualized was smaller (about 250 mg) in size than the larger (about 450 mg) necrotic tumor, nearly undetected at 24 h postinjection. Strong kidney and bladder uptakes are also visualized.

Discussion

The choice to target a MUC1 peptidic epitope, carried by the VNTRs of the MUC1 extracellular domain, is justified by the presence of abundant hypoglycosylated MUC1 forms on epithelial cancers, such as breast and prostate cancers.32,56

Assembly of the MUC1 pretargeting molecule for imaging and pretargeted RIT of metastatic breast and prostate cancers requires three major components: a tumor Ag binding module, a radioactive hapten capturing module and a bifunctional PEG for covalent attachment of the binding and catching modules.

The MUC1 binding module, developed as a di-scFv-SH and optimized to scFv-G^sub 4^S-C-(G^sub 4^S)^sub 3^-scFv, has an apparent MW of 52 KD and is ready for use. The radioactive hapten capturing module, consisting of an anti-DOTA (radioactive metal) scFv, has been developed in two formats: scFv-SH and di-scFv-SH. While the design of the MUC1 pretargeting molecule is definitively set for bivalent binding to tumor Ags, capture of the radioactive hapten will be either mono or bivalent. The choice between a scFv-SH or di-scFv-SH capturing module will ideally be based on in vivo performances of bispecific (scFv)^sub 2^-PEG-scFv and (scFv)^sub 2^-PEG-(scFv)^sub 2^ MUC1 pretargeting molecules. Steric hindrance, created by the hydrophilic bulky PEG, that accounts for the well known benefits of PEG conjugates, such as reduced immunogenicity, increased half-life and solubility and protease resistance, also limits PEGylation yields by reducing protein accessibility.57 Because steric hindrance should be reduced by the use of shorter PEG molecules,57 the PEG scaffold in the MUC1 pretargeting molecule should not exceed 10 KD in size. Therefore, the MW of such a molecule will be 75-85 KD for (scFv)^sub 2^-PEG-scFv and 100-110 KD for (scFv)^sub 2^-PEG- (scFv)^sub 2^ and in either case large enough to prevent rapid elimination from the circulation through glomerular filtration.

Various approaches to increase PEGylation efficiencies of scFv- SH and di-scFv-SH modules are under investigation. One such approach, consists of using 1, 2, 3 trizole ligation (“click chemistry”) for covalent attachment to a bifunctional (azide/tri- alkyne) PEG scaffold of the binding and capturing modules. This approach has yielded up to 74% scFv-PEG-scFv molecules and provides control over the composition of the product.58 Because attachment of smaller molecules to a bifunctional and small size PEG is more efficient, the likelihood to increase the assembly yield of (scFv)^sub 2^-PEG-scFv molecules is higher than that of (scFv)^sub 2^-PEG-(scFv)^sub 2^ molecules. On the other hand, a significant yield improvement for (scFv)^sub 2^-PEG-(scFv)^sub 2^ might be achieved by using a bifunctional short PEG, such as a 2 KD PEG- Mal^sub 2^, with bispecific, anti-MUC1 and anti-radioactive hapten, di-scFv-SH.

The fact that valency prevails over MW for superior tumor retention 59 might not be critical for the radionuclide capture module of the MUC1 pretargeting molecule, suggesting that the goal to be reached, TI improvement of RIT, can be achieved with bispecific tri- or tetravalent pretargeting molecules. Recently, a carcinoembryonic antigen pretargeting molecule, as a bispecific trivalent Fab construct assembled through the dock and lock method, showed high tumor localization in a human colonic carcinoma model.60

In summary, the preclinical development of MUC1 pretargeting molecules for RIT of metastatic breast and prostate cancers is in its final stages. PEGylation efficiencies of the binding and capturing modules will influence the final valency of these molecules: bispecific, tri- or tetravalent scFvs. Presentation at the 18th IRIST meeting, London, July 12-14, 2006.

Funding.-This research was supported by National Cancer Institute Grant PO1 CA47829, U.S. Department of Energy Grant DE-FG01- 00NE22944 and Department of Defense Grant DAMD17-01-0177.

References

1. Pressman D, Korngold L. The in vivo localization of anti- Wagner-osteogenic sarcoma antibodies. Cancer 1953;6:619-23.

2. Kohler G, Milstein C. Continuous culture of fused cells secreting antibody of predefined specificity. Nature 1975;256:495- 7.

3. Mach JP, Chatal JF, Lumbroso JD, Buchegger F, Forni M, Ritschard J et al. Tumor localization in patients by radiolabeled monoclonal antibodies against colon carcinoma. Cancer Res 1983;43:5593-600.

4. Buchegger F, Halpern SE, Sutherland RM, Schreyer M, Mach JP. In vitro and in vivo tumor models for studies of distribution of radiolabeled monoclonal antibodies and fragments. Nuklearmedizin 1986;25:207-9.

5. DeNardo SJ, DeNardo GL. Targeted radionuclide therapy for solid tumors: an overview. Int J Radiat Oncol Biol Phys 2006;66:S89- S95.

6. McCafferty J, Griffiths AD, Winter G, Chiswell DJ. Phage antibodies: filamentous phage displaying antibody variable domains. Nature 1990;348:552-4.

7. Weiner LM. Fully human therapeutic monoclonal antibodies. J Immunother 2006;29:1-9.

8. Buchegger F, Pelegrin A, Delaloye B, Bischof-Delaloye A, Mach JP. Iodine-131-labeled MAb F(ab’)2 fragments are more efficient and less toxic than intact anti-CEA antibodies in radioimmunotherapy of large human colon carcinoma grafted in nude mice. J Nucl Med 1990;31:1035-44.

9. Kukis DL, Novak-Hofer I, DeNardo SJ. Cleavable linkers to enhance selectivity of antibody-targeted therapy of cancer. Cancer Biother Radiopharm 2001;16:457-67.

10. DeNardo SJ, Kukis DL, Kroger LA, O’Donnell RT, Lamborn KR, Miers LA et al. Synergy of Taxol and radioimmunotherapy with yttrium- 90-labeled chimeric L6 antibody: efficacy and toxicity in breast cancer xenografts. Proc Natl Acad Sci U S A 1997;94:4000-4.

11. Gruaz-Guyon A, Raguin O, Barbet J. Recent advances in pretargeted radioimmunotherapy. Curr Med Chem 2005;12:319-38.

12. Goldenberg DM, Sharkey RM, Paganelli G, Barbet J, Chatal JF. Antibody pretargeting advances cancer radioimmunodetection and radioimmunotherapy. J Clin Oncol 2006;24:823-34.

13. Urbano N, Papi S, Ginanneschi M, De Sands R, Pace S, Lindstedt R et al. Evaluation of a new biotin-DOTA conjugate for pretargeted antibody-guided radioimmunotherapy (PAGRIT((R))). Eur J Nucl Med Mol Imaging 2007;34:68-77.

14. Carter, P. Bispecific human IgG by design. J Immunol Methods 2001;248:7-15.

15. Tomlinson I, Holliger P. Methods for generating multivalent and bispecific antibody fragments. Methods Enzymol 2000;326:461-79.

16. Gendler SJ, Lancaster CA, Taylor-Papadimitriou J, Duhig T, Peat N, Burchell J et al. Molecular cloning and expression of human tumor-associated polymorphic epithelial mucin. J Biol Chem 1990;265:15286-93.

17. Ligtenberg MJ, Kruijshaar L, Buijs F, van Meijer M, Litvinov SV, Hilkens J. Cell-associated episialin is a complex containing two proteins derived from a common precursor. J Biol Chem 1992;267:6171- 7.

18. Gendler S, Taylor-Papadimitriou J, Duhig T, Rothbard J, Burchell J. A highly immunogenic region of a human polymorphic epithelial mucin expressed by carcinomas is made up of tandem repeats. J Biol Chem 1988;263:12820-3.

19. Taylor-Papadimitriou J, Burchell J, Miles DW, Dalziel M. MUC1 and cancer. Biochim Biophys Acta 1999;1455:301-13.

20. Brayman M, Thathiah A, Carson D. MUC1: a multifunctional cell surface component of reproductive tissue epithelia. Reprod Biol Endocrinol 2004;2:4-12.

21. Zotter S, Hageman C, Lossnitzer A, Mooi WJ, Hilgers J. Tissue and tumor distribution of human polymorphic epthelial mucin. Cancer Rev 1988;11-12:55-101.

22. Barratt-Boyes SM. Making the most of mucin: a novel target for tumor immunotherapy. Cancer Immunol Immunother 1996;43:142-51.

23. Burchell J, Gendler S, Taylor-Papadimitriou J, Girling A, Lewis A, Millis R et al. Development and characterization of breast cancer reactive monoclonal antibodies directed to the core protein of the human milk mucin. Cancer Res 1987;47:5476-82.

24. Ceriani RL, Peterson JA, Blank EW, Derek TA. Epitope expression on the breast epithelial mucin. Breast Cancer Res Treat 1992;24:103-13.

25. ten Berge RL, Snijdewint FGM, von Mensdorff-Pouilly S, Poort- Keesom RJJ, Oudejans JJ, Meijer JWR et al. L. MUC1 (EMA) is preferentially expressed by ALK positive anaplastic large cell lymphoma, in the normally glycosylated or only partly hypoglycosylated form. J Clin Pathol 2001;54:933-9.

26. Cao Y, Blohm D, Ghadimi M, Stosiek P, Xing PX, Karsten U. Mucins (MUC1 and MUC3) of gastrointestinal and breast epithelial reveal different and heterogeneous tumor-associated aberrations in glycosylation. J Histochem Cytochem 1997;45:1547-57.

27. Lapointe J, Li C, Higgins JP, van de Rijn M, Bair E, Montgomery K et al. Gene expression profiling identifies clinically relevant subtypes of prostate cancer. Proc Natl Acad Sci U S A 2005;101:811-6.

28. Best CJM, Gillespie JW, Yi Y, Chandramouli GVR, Perlmutter MA, Gathright Y et al. Molecular alterations in primary prostate cancer after androgen ablation therapy. Clin Cancer Res 2005;11:6823- 34.

29. O’Connor JC, Julian J, Lim SD, Carson DD. MUC1 expression in human prostate cancer cell lines and primary tumors. Prostate Cancer Prostatic Dis 2005;8:36-44.

30. Arai T, Fujita K, Fujime M, Irimura T. Expression of sialylated MUC1 in prostate cancer: Relationship to clinical stage and prognosis. Int J Urol 2005;12:654-61.

31. Zhang S, Zhang HS, Reuter VE, Slovin SF, Scher HI, Livingston PO. Expression of potential target antigens for immunotherapy on primary and metastatic prostate cancers. Clin Cancer Res 2001;4:295- 302.

32. Burke PA, Gregg JP, Bakhtiar B, Beckett LA, DeNardo GL, Albrecht H et al. Characterization of MUC1 glycoprotein on prostate cancer for selection of targeting molecules. Int J Oncol 2006;29:49- 55.

33. Price MR, Rye PD, Petrakou E, Murray A, Brady K, Imaj S et al. Summary report on the ISOBM TD-4 Workshop: analysis of 56 monoclonal antibodies against the MUC1 mucin. Tumor Biol 1998;19:1- 20.

34. Fontenot JD, Tjandra N, Bu D, Ho C, Montelaro RC, Finn OJ. Biophysical characterization of one-, two-, and three-tandem repeats of human mucin (MUC-1) protein core. Cancer Res 1993;53:5386-94.

35. Peterson JA, Zava DT, Duwe AK, Blank EW, Battifora H, Ceriani RL. Characterization of antigens preferentially expressed on the surface and cytoplasm of breast carcinoma cells identified by monoclonal antibodies against the human milk fat globule. Hybridoma 1990;9:221-35.

36. Kramer EL, DeNardo SJ, Liebes L, Kroger LA, Noz ME, Mizrachi H et al. Radioimmunolocalization of breast carcinoma using BrE-3 monoclonal antibody: phase I study. J Nucl Med 1993;34:1067-74.

37. Hoogenboom HR. Designing and optimizing library selection strategies for generating high-affinity antibodies. Trends Biotechnol 1997;15:62-70.

38. Winthrop MD, DeNardo SJ, DeNardo GL. Development of a hyperimmune anti-MUC-1 single chain antibody phage display library for targeting breast cancer. Clin Cancer Res 1999;10:3088-94.

39. Winthrop MD, DeNardo SJ, Albrecht H, Mirick GR, Kroger LA, Lamborn KR et al. Selection and characterization of anti-MUC-1 scFvs intended for targeted therapy. Clin Cancer Res 2003;9:3845s-3853s.

40. Adams GP, Schier R, McCall AM, Crawford SM, Wolff EJ, Weiner M et al. Prolonged in vivo tumor retention of a human diabody targeting the extracellular domain of human HER2/neu. Br J Cancer 1998;77:1405-12.

41. Chinol M, Casalini P, Maggiolo M, Canevari S, Omodeo ES, Caliceti P et al. Biochemical modifications of avidin improve pharmacokinetics and biodistribution, and reduce immunogenicity. Br J Cancer 1998;78:189-97.

42. Padlan EA. Anatomy of the antibody molecule. Mol Immunol 1994;31:169-217.

43. Adams GP, McCartney JE, Tai MS, Oppermann H, Huston JS, Stafford WF et al. Highly specific in vivo tumour targeting by monovalent and divalent forms of 741F8 anti-c-erbB-2 single-chain Fv. Cancer Res 1993;53:4026-34.

44. Kipriyanov SM, Dubel S, Breitling F, Konterman RD, Little M. Recombinant single-chain Fv fragments carrying C-terminal cysteine residues: production of bivalent and biotinylated miniantibodies. Mol Immunol 1994;31:1047-58.

45. Wang D, Berven E, Li Q, Uckun F, Kersey JH. Optimization of conditions for formation and analysis of anti-CD19 FVS191 single- chain Fv homodimer (scFv’)^sub 2^. Bioconjug Chem 1997;8:64-70.

46. Albrecht H, Burke PA, Natarajan A, Xiong CY, Kalicinsky M, DeNardo GL et al. J. Production of soluble ScFvs with C-terminal- free thiol for site-specific conjugation or stable dimeric ScFvs on demand. Bioconjug Chem 2004;15:16-26.

47. Natarajan A, Xiong CY, Albrecht H, DeNardo GL, DeNardo SJ. Characterization of site-specific ScFv PEGylation for tumor- targeting pharmaceuticals. Bioconjug Chem 2005;16:113-21.

48. Yang K, Basu A, Wang M, Chintala R, Hsieh MC, Li S et al. Tailoring structure-function and pharmacokinetic properties of single-chain Fv proteins by site-specific PEGylation. Protein Eng 2003;16:761-70.

49. Holliger P, Prospero T, Winter G. ‘Diabodies’: small bivalent and bispecific antibody fragments. Proc Natl Acad Sci U S A 1993;90:6444-8.

50. Cochlovius B, Kipriyanov SM, Stassar MJJS, Schumacher J, Benner A, Moldenhauer G et al. Cure of Burkitt’s lymphoma in severe combined immunodeficiency mice by t cells, tetravalent CD3xCD19 tandem diabody, and CD28 costimulation. Cancer Res 2000;60:4336-41.

51. Olafsen T, Cheung CW, Yazaki PJ, Li L, Sundaresan G, Gambhir SS et al. Covalent disulfide-linked anti-CEA diabody allow site- specific conjugation and radiolabeling for tumor targeting applications. Protein Eng Des Sel 2004;17:21-7. 52. Mack M, Riethmuller G, Kufer P. A small bispecific antibody construct expressed as a functional single-chain molecule with high tumor cell cytotoxicity. Proc Natl Acad Sci U S A 1995;92:7021-5.

53. Goel A, Colcher D, Baranowska-Kortylewicz J, Augustine S, Booth BJM, Pavlinkova G et al. Genetically engineered tetravalent single-chain Fv of the pancarcinoma monoclonal antibody CC49: improved biodistribution and potential for therapeutic application. Cancer Res 2000;60:6964-71.

54. Albrecht H, DeNardo GL, DeNardo SJ. Monospecific bivalent scFv-SH: effects of linker length and location of an engineered cysteine on production, antigen binding activity and free SH accessibility. J Immunol Methods 2006;310:100-16.

55. Xiong CY, Natarajan A, Shi XB, DeNardo GL, DeNardo SJ. Development of tumor targeting anti-MUC-1 multimer: effects of di- scFv unpaired cysteine location on PEGylation and tumor binding. Protein Eng Des Sel 2006;19:359-367.

56. Burchell JM, Mungul A, Taylor-Papadimitriou J. O-linked glycosylation in the mammary gland: changes that occur during malignancy. J Mammary Gland Biol Neoplasia 2001;6:355-64.

57. Kubetzko S, Sarkar CA, Pluckthun A. Protein PEGylation decreases observed target association rates via a dual blocking mechanism. Mol Pharmacol 2005;68:1439-54.

58. Natarajan A, Du W, Xiong CY, DeNardo GL, DeNardo SJ, Gervay- Hague J. Construction of di-scFv through a trivalent alkyne-azide 1,3 dipolar cycloaddition. R Soc Chem Chem Comm. In press 2007.

59. Adams GP, Tai MS, McCartney JE, Marks JD, Stafford WF III, Houston LL et al. Avidity-mediated enhancement of in two tumor targeting by single-chain Fv dimers. Clin Cancer Res 2006;12:1599- 605.

60. Rossi EA, Goldenberg DM, Cardillo TM, McBride WJ, Sharkey RM, Chang CH. Stably tethered multifunctional structures of defined composition made by the dock and lock method for use in cancer targeting. PNAS 2006;103:6841-6.

H. ALBRECHT, G. L. DENARDO, S. J. DENARDO

Davis Medical Center

University of California, Sacramento, CA, USA

Address reprint requests to: S. DeNardo, MD, Radiodiagnosis and Therapy, University of California, Davis Medical Center, 1508 Alhambra Boulevard, Room 3100, Sacramento, CA 95816, USA.

E-mail: [email protected]

Comments

comments