Imported: 13 Feb '17 | Published: 18 Jan '11
USPTO - Utility Patents
A novel single domain antibody AFAI and fragments thereof which has specific affinity for binding to carcinoma, and especially lung carcinoma. This antibody, and portions thereof, can be used, inter alia in the diagnosis and treatment of carcinoma.
This application is a continuation-in-part of international application Number PCT/CA2004/001488 filed 17 Aug. 2004 and a continuation-in-part of U.S. application Ser. No. 10/547,528 filed 31 Aug. 2005 now abandoned which entered the national phase based upon PCT/CA2004/000309 filed Mar. 2, 2004.
It is widely expected that proteomic research will greatly facilitate the discovery of novel tumor targets. Major advances have been made in the identification of targets for diagnostic purposes. However, limitations of the present technologies have hindered identification of new therapeutic targets. The techniques commonly employed in proteomics, such as two-dimensional gel electrophoresis, Liquid Chromatography Tandem Mass Spectrometry (LC/MS/MS), Matrix-Assisted Laser Desorption Ionization/Mass Spectrometry (MALDI-MS), and the yeast two-hybrid system have not met the demand for “drugable” targets, such as cell surface markers.
Tumor targeting antibodies and peptides can be isolated by library display approaches (e.g. Aina, 2002; Hoogenboom, 1998). This is usually accomplished by screening a phage display library, or libraries with other display formats, against purified tumor specific or tumor associated antigens. However, tumor targeting antibodies and peptides have also been isolated by panning libraries against tumor cells or tumor tissues without prior information on the molecular targets. Noteworthy advantages of the latter approach are: (i) the isolated antibodies/peptides bind to native forms of their antigens/ligands on the cell surface whereas purified tumor antigens are often recombinant in nature and lack post-translational modification, (ii) the antigens are accessible to the isolated antibodies/peptides whereas those isolated by panning against pure antigens may recognize epitopes which are naturally buried in the membrane or blocked by carbohydrate modification. However, antibodies/peptides isolated with this method usually have a low to moderate affinity to their antigens/ligands.
Since each M13 phage particle presents five copies of the minor coat protein pIII, a phage particle displaying an antibody fragment on all copies of pIII can be considered a pentavalent antibody. This multivalent display of antibody fragments on phage greatly increases the avidity of the antibody and facilitates both screening and evaluation of phage antibodies. Isolated antibody fragments (scFvs or sdAbs) or peptides bind antigen much less efficiently since they exist primarily in a monovalent form and lack avidity.
An antibody fragment oligomerization strategy that permits pentavalency as in pIII phage display is the subject of PCT/CA02/01829 (MacKenzie and Zhang). Fusion of a single domain antibody (sdAb) to the homo-pentamerization domain of the B subunit of verotoxin (VT1B) results in the simultaneous pentamerization of the sdAb. The pentavalent sdAbs, termed pentabodies, bind much more strongly to immobilized antigen than their monomeric counterparts. In the instance of peptide hormone-binding sdAb, pentamerization resulted in 103 to 104-fold improvement in binding to immobilized antigen.
It is an object of the invention to provide a single-domain antibody with affinity for lung carcinoma.
There is provided herein a novel single domain antibody AFAI and fragments thereof which has specific affinity for binding to carcinoma, and especially lung carcinoma. This antibody, and portions thereof, can be used, inter alia in the diagnosis and treatment of carcinoma.
Proteomics research has delivered many novel tumor targets. However, due to some limitations, it has been difficult to identify targets that are most accessible for drug application. A novel tumor antigen discovery platform based on screening a single domain antibody (sdAb) library against tumor cells and subsequently identifying the corresponding antigens of the isolated antibodies is described herein. A specific sdAb, AFAI, specific for non-small cell lung carcinoma (A549 cell line) was isolated from a phage library derived from the heavy chain antibody repertoire of a llama as described in Example 1. The homopentamerization property of a non-toxic verotoxin B-subunit was exploited to make the ES1 pentabody, the pentameric forms of AFAI. Pentamerization dramatically improved the binding of the AFAI to A549 cells. Immunohistostaining showed that ES1 is highly specific for lung carcinoma.
It is possible in light of the disclosure herein to chemically syntesize the whole gene for ES1 and express it in E. coli or another suitable organism according to standard techniques. The gene of AFAI SEQ ID NO.3 can be synthesized from ES1, which is the binding entity to its antigen. It is possible, in light of the disclosure herein, to make dimeric, trimeric, tetrameric, other pentameric and other multivalent forms of AFAI SEQ ID NO.3. Such products can be useful in methods and compositions relating to the present invention, as can DNA and clone material providing variants of such products which provide specific binding to a malignant tissue or cells of interest, as described herein.
ES1, AFAI and/or variants thereof showing similar binding specificity (“suitable variants”) are useful in the diagnosis and treatment of lung carcinoma. Diagnostic methods with which ES1, AFAI and/or suitable variants thereof can be used include: immunohistochemical methods; labeling the molecule(s) (ES1, AFAI, variants) with radio isotopes and detecting with tumor imaging tools such as positron emission tomography and MRI; analysis of blood samples (and detecting binding using standard techniques).
Diagnostic kits comprising ES1, AFAI and/or a suitable variant thereof and instructions for their use are specifically contemplated.
Therapeutic methods and compositions relating to ES1 and AFAI include employing AFAI, ES1, or a suitable variant thereof and, for example: labeling them with radio isotope and applying the molecules to a patient; conjugating them to one or more conventional therapeutics and applying the conjugate to a patient; conjugating them to one or more toxins and applying the conjugates to patients; expressing nucleic acid molecules encoding ES1, AFAI and/or a suitable variant thereof in a gene therapy vector and applying the vectors to patients.
The non-small cell lung carcinoma cell line A549 was purchased from ATCC (Manassas, Va.) and maintained in DMEM (Gibco, Rockville, Mass.) supplemented with 5% FBS (Gibco) and 1% Antibiotic-Antimycotic (Gibco). Primary human dermal fibroblast cells were kindly provided by Dr. J. Xu (Apotex Research Inc. Ottawa, ON). Polyclonal rabbit anti-verotoxin antiserum was kindly provided by Dr. Clifford Lingwood (Univ. of Toronto).
Isolation of sdAb AFAI which Binds to Non-Small Cell Lung Carcinoma Cell Line A549
A naïve llama single domain antibody library (Tanha et al, 2002) served as the source of an antibody fragment specific for tumor cells, in this instance the non-small cell lung carcinoma cell line A549. The isolation of phage antibodies that bind to A549 cells, termed cell panning, was performed with A549 cells with pre-adsorption of the library on human fibroblasts at each round of panning.
An sdAb phage display library (Tanha et al, 2002) was panned with A549 as the target cells and human dermal fibroblasts as subtracting cells. The panning was performed as described in Becerril et. al. (1999) with slight modifications. For the first round of panning, 1013 pfu were incubated with the subtracting fibroblast cells to remove fibroblast-binding phage. Phage particles remaining in the supernatant were incubated with A549 cells cultured in a 5 cm petri dish at room temperature for 1 hr. The A549 cells were washed 5 times, 1 minute each, with PBS and 5 times, 10 minutes each, with stripping buffer (50 mM glycine, pH 2.8, 0.5 M NaCl, 2 M urea, 2% polyvinylpyrolidone) and then lysed with 100 mM triethylamine. The cell lysate was neutralized by the addition of 100 μl of 1 M Tris (pH 7.0). Phage in the neutralized cell lysate were amplified in E. coli TG1 cells. The amplified sub-library was subjected to the next round of panning employing the same method. Individual phage clones were selected after four rounds of panning and the DNA sequences encoding the displayed antibodies were determined.
Individual phage clones were isolated after four rounds of panning and the cell binding activities of the phage clones were examined by ELISA. Of 94 clones, 25 clones tested positive. Sequence analysis of the ELISA-positive phage clones showed that all 25 positive phage clones displayed the same sdAb. This antibody was designated AFAI because the CDR3 region of the antibody is the tetrapeptide Ala-Phe-Ala-Ile (SEQ ID NO:3; FIG. 1).
Cell Staining Part A: Cell Staining with Phage Displayed AFAI
When A549 cells were immunostained with AFAI phage as the first antibody followed by an anti-M13 monoclonal antibody and Fluor 546 labeled goat anti-mouse IgG it was observed that very intense fluorescent signals were associated with a cell sub-population (FIG. 2). The staining pattern of the positive A549 cells suggested that AFAI binds to an abundant membrane antigen.
To investigate whether the binding of AFAI phage to A549 cells is cell type specific, a human bronchial epithelial cell line, HBE4, a human prostate cell line, PREP and a primary human fibroblast cell line were chosen as controls for immunocytochemical staining. Under the same conditions employed for A549 immunocytochemistry, no staining was observed with human fibroblasts and only very weak staining was observed with HBE4 and PREP.
Cell Staining Part B: Production of Monomeric and Pentameric AFAI sdAbs and Cell Staining with the sdAbs
For further evaluation and characterization of AFAI, monomeric and pentameric AFAI were expressed and purified. The gene encoding AFAI was amplified by PCR and inserted into an E. coli expression vector, generating clone pAFAI (FIG. 1B). To exploit the high avidity effect of pentabodies, a pentameric form of AFAI, designated ES1, was constructed (FIG. 1B and FIG. 1C). The yields of purified AFAI and ES1 (FIG. 1D) from 1 liter flask cultures of E. coli TG1, without fermentation optimization, were 6 mg and 20 mg, respectively.
Briefly, DNA encoding sdAb AFAI was cloned into the BbsI/BamHI sites of plasmid pSJF2 (Tanha, 2003) and BbsI/ApaI sites of plasmid pVT2 to generate expression vectors for monomeric and pentavalent AFAI, respectively. The obtained E. coli clones were designated pAFAI (monomer) and pES1 (pentamer). Proteins AFAI and ES1 were produced as described in Tanha et. al. (2003) with the modification of protein extraction from E. coli cells by cell lysis instead of osmotic shock. Briefly, the pAFAI and pES1 clones were inoculated into 100 ml M9 medium (0.2% glucose, 0.6% Na2HPO4, 0.3% KH2PO4, 0.1% NH4Cl, 0.05% NaCl, 1 mM MgCl2, 0.1 mM CaCl2) supplemented with 0.4% casamino acids, 5 mg/l vitamin B1 and 200 μg/ml ampicillin and shaken overnight at 37° C. Thirty ml of the overnight M9 culture were transferred into 1 liter of M9 medium with the same supplements and shaken at 37° C. for 24 hours. Induction of gene expression was initiated by the addition of 100 ml 10× TB nutrients (12% Tryptone, 24% yeast extract, 4% glycerol), 2 ml of 100 mg/ml ampicillin and 1 ml of 1 M IPTG and the cultures were shaken at room temperature for 48 to 72 hours. E. coli cells were harvested by centrifugation and lysed with an Emulsiflex™ Cell Disruptor (Avestin Inc. Ottawa, ON). The cell lysate was centrifuged, the obtained clear supernatant was loaded onto a Hi-Trap™ Chelating Affinity Column (Amersham Biosciences, Piscataway, N.J.) and proteins containing His5 tag were purified following the manufacturer's instructions.
Immunochemical staining of A549 cells was performed with both monomeric (AFAI) and pentameric (ES1) antibodies.
Standard immunochemical methods, with slight modifications, were employed in cell staining with AFAI phage and with monomeric and pentameric AFAI. Cells were grown on slide cover slips to approximately 70% confluence and fixed for 10 minutes with 4% formaldehyde in PBS. Permeabilization was carried out for 30 minutes at room temperature in 0.05% NP-40 (Bio-Rad, Hercules, Calif.) followed by three washes with PBS containing 0.05% Tween-20 (PBST). For cell staining with AFAI phage, 2×1011 pfu of AFAI phage (in A549 medium) were incubated with fixed cells for 18 hours at 4° C. and washed three times, 5 minutes each, with PBST. For cell staining with monomeric or pentameric AFAI, 100 μg/ml of AFAI and ES1 (in A549 medium) were incubated with the cells for 2 hours at room temperature and washed three times with PBST. Secondary antibodies, monoclonal anti-M13 IgG (Amersham Biosciences) for M13 phage or the 9E10 anti c-myc IgG (ATCC) for the ES1 pentabody were applied at a 1:100 dilution for 30 minutes at room temperature followed by three washes with PBST. The third antibody, Alexa Fluor 546-labeled goat anti-mouse IgG™ (Molecular Probes, Inc. Eugene, Oreg.), was diluted 1:100 and applied in the same way as the secondary antibodies. Contrast staining was performed with DAPI and (DiOC5)3 (Molecular Probes). Following immunochemical staining cover slips were mounted using an Prolong Antifade Kit (Molecular Probes) and observed under an Olympus BX51™ fluorescent microscope and images were recorded.
No obvious staining was observed when AFAI was employed (FIG. 3), probably because of the low binding affinity of monomeric AFAI. The ability of AFAI antibody to stain A549 cells was, however, observed when the pentameric form, ES1, was employed (FIG. 3). As observed with AFAI phage, ES1 stains only a sub-population of A549 cells.
Determination of Specificity of ES1 to Tumor Tissues
To determine the tissue specificity of ES1, immunohistochemical staining of a broad range of tissues on a tissue microarray using ES1 as primary antibody was performed. The results showed that ES1 recognized most lung adenocarcinomas by displaying a moderate to strong immunoreactivity. None of colon adenocarcinoma displayed strong immunoreactivity for ES1 however a focal weak to moderate immunoreactivity was observed in a few cases. Non cancerous lung and colon tissues were not immunoreactive (FIG. 4, Table 1).
To determine the tissue specificity of ES1, immunohistochemical staining of a broad range of tissues was performed using ES1 as primary antibody.
Immunostaining of human tissues using ES1 as the primary antibody was performed using the avidin-biotin peroxidase complex (ABC) method with an ABC kit (Vector Laboratories, Burlingame, Calif., USA) on four micron-thick sections cut from the paraffin blocks.
Immunostaining of human tissues using MIB1 (Dako, dilution 1:100) and TTF1 (Dako, dilution 1:50), two broadly used antibodies in lung carcinoma detection was performed using peroxidase-antiperoxidase technique.
The immunoreactivity for ES1 was assessed by two pathologists and was scored as moderately or strongly positive staining when there was a continuous membranous and/or cytoplasmic staining pattern and as weakly positive when there was discontinuous membrane or weak cytoplasmic staining. The moderate or strong staining pattern was further scored as 3 in cases showing staining in more than 50%, 2 in more than 10% and 1 in up to 10% of cells. Cases with a discrepant score were reviewed.
One hundred-forty three resection or biopsy specimens containing tumors of lung, colon, breast stomach, pancreas, prostate, endometrium, ovary, thyroid and mesothelium were obtained (Table 1). For each case, one sample of representative tumor tissue, 2 mm in diameter, was removed from the paraffin block and re-embedded with other tumor samples to produce a tissue microarray paraffin block that contained at least 15 different tissues. Normal tissue was also sampled from non tumoral tissue distant from the tumor and from the normal autopsy lung tissue. Cases of lung, colon, breast stomach pancreas urinary bladder, gall bladder, esophagus and ovary with microarray tissue displaying negative, weak or focal immunoreactivity were re-submitted for immunostaining for ES1 using large tissue sections.
Table 2 compares the immunostaining results of the group of non-squamous large cell lung carcinomas with the combined group of colonic, mammary, urothelial carcinomas and other mucus-secreting adenocarcinomas. Excluding other types of carcinomas which showed weak or negative ES1 immunoreactivity, the sensitivity and the specificity of ES1 immunoreactivity for lung non-squamous large cell carcinomas were 97 and 45% respectively. The positive predictive value was 54%.
The results showed that ES1 displayed moderate to strong immunoreactivity with most lung adenocarcinomas. None of the colon adenocarcinomas displayed strong immunoreactivity with ES1; however, weak to moderate focal immunoreactivity was observed in a few cases. Non cancerous lung and colon tissues were not immunoreactive.
The reactivity was often stronger on the membrane (FIGS. 5,6,7) than in the cytoplasm (FIG. 8). The pattern of positive immunostaining varied from diffuse (FIGS. 5,6) to focal staining in portions of tumor or individual cells (FIGS. 7,8). Of 93 cases with negative, weak or focal immunoreactivity in microarray sections, there were eleven large sections showing a score of 1 or 2 immunoreactivity. Table 1 summarizes the final findings on immunostaining of all specimens.
Thirty-five non-squamous large cell carcinomas of the lung including seven bronchiolo-alveolar carcinomas and 22 well to poorly-differentiated adenocarcinomas, six large cell undifferentiated carcinomas showed scores of 1, 2 and 3 immunoreactivity in 7, 17 and 10 tumors respectively (FIGS. 5 to 8). The remaining carcinoma was an undifferentiated large cell carcinoma with some features of squamous differentiation showing negative immunoreactivity. The tumors with focal moderate to strong immunoreactivity (less than 10% immunoreactive cells) were well-differentiated non-mucinous tumors. Two adeno-squamous carcinomas also displayed scores of 1 and 2 immunoreactivity in the adenocarcinoma component (FIG. 9). Four of five atypical adenomatous hyperplasias of the lung showed focal or weak immunoreactivity (FIG. 10). All pure squamous carcinomas, carcinoid tumors, normal and reactive lung parenchyma with or without accompanying carcinomas were not reactive for the antibody.
For 53 non-lung and mucus-secreting tumor including 15 adenocarcinoma colonic adenocarcinomas, 8 breast carcinomas, 4 urothelial carcinomas and adenocarcinomas of the pancreas (six), stomach (six) and gallbladder (one), esophagus (three), urinary bladder (two), ovary (three) and trachea (one), the immunoreactivity was scored as 1, 2 , 3, weak and negative in 15, 11, 3, 12 and 12, respectively (FIGS. 11,12). Colonic adenocarcinomas formed the subgroup of tumor with more diffuse and strong immunoreactivity. Colonic adenomas only displayed focal or weak immunoreactivity.
For 32 non lung and non-mucinous tumors, the immunoreactivity was negative or weak. Normal tissues from lung, liver, pancreas, kidney, urinary bladder, endometrium, thyroid, esophagus and ovary from areas surrounding cancer or from organs not harboring cancer were not reactive for cancer with exception of one case of breast tissue (surrounding duct carcinoma) showing moderate positivity in occasional acini.
Immunostaining for MIB1 performed on 24 large sections of 24 non-squamous large cell lung carcinomas with score 1 or 2 immunoreactivity showed a remarkable increase in proliferativity of tumor cells in areas of large numbers of ES1-immunoreactive cells as compared to areas with negative, weak or focal ES1-immunoreactive cells. The immunostaining for TTF1 was performed on microarray tissue. All non-lung and non-thyroid tissue showed negative nuclear immunoreactivity. For 35 lung non-squamous large cell carcinomas, TTF1 immunoreactivity was negative in 8 cases including 5 undifferentiated large cell carcinomas, two poorly-differentiated adenocarcinomas and one mucinous bronchiolo-alveolar carcinoma. The other types of microarray tissue were not immunoreactive.
In this study, ES1 immunoreactivity was almost completely limited to malignant tumors, particularly lung adenocarcinomas. There was a tendency for the immunoreactivity to be stronger in lung tumors with a mucinous component as non-mucinous bronchiolo-alveolar carcinomas showed only focal immunoreactivity. ES1 immunoreactivity was positive in a number of colonic adenocarcinomas with weaker and more focal staining than in lung adenocarcinomas. Of interest, ES1 immunostaining remained scoring 2 or 3 for undifferentiated large cell lung carcinoma in the primary as well as in the metastatic sites as compared to low sensitivity of TTF1 in the immunostaining of the undifferentiated large cell lung carcinoma (*). Furthermore colonic adenomas showed only focal weak or negative immunoreactivity. Mucinous adenocarcinoma from other organs displayed score 2, focal or weak immunoreactivity in a small number of cases. The immunoreactive changes identified in the lung adenocarcinomas in this study likely correspond to an up-regulation of the AFAI antigen. This impression is supported by the finding of positive ES1 immunoreactivity in areas of carcinoma of the lung with increased proliferativity as demonstrated by the immunoreactivity for MIB1.
Non-Limiting Discussion of Variations and Uses of AFAI
The AFAI antigen, appears to be up-regulated in lung adenocarcinoma, even in less differentiated tumors. ES1 is likely a more sensitive marker for lung poorly differentiated lung adenocarcinoma than TTF1. Since most normal tissue tested were not ES1-immunoreactive, ES1 is suitable for use in the development of a screening test for lung adenocarcinomas and a number of colon and breast carcinomas.
In an embodiment of the invention there is provided an amino acid sequence of AFAI as shown in FIG. 1 or an amino acid sequence at least 90%, 95% or 98% identical to it. Examples of variant amino acid sequences of interest are shown in Table III in which unchanged residues are indicated by a hyphen. It will be appreciated that AFAI may be mutated at any position which does not interfere with antigen binding or specificity. Sites of particular interest include those for which some possible mutations are shown in Table III. While only some possible variants are shown, it will be appreciated that all functional variants, including all variants of AFAI differing from SEQ. ID. No. 1 by one or more of the amino acid changes depicted in Table III are specifically contemplated and fall within the scope of the invention.
In an embodiment of the invention there is provided an amino acid sequence having complete sequence identity to the underlined (CDR) regions of AFAI as shown in FIG. 1 and having at least 40%, 60%, 80%, or 90% sequence identity to the remaining portions of that sequence. Also provided are nucleic acid sequences encoding such amino acid sequences.
In an embodiment of the invention there is provided nucleic acid sequences encoding AFAI as disclosed in FIG. 1, or an amino acid sequence at least 90% or 95% identical to it. In an embodiment of the invention there is provided a nucleic acid sequence encoding a protein which has at least 70%, 80%, 90% 95%, or 98% sequence identity to the sequence of AFAI or ES1 as depicted in Table 2, or to a continuous 250 nucleic acid region thereof, or being complementary to any such nucleic acid sequence. In an embodiment of the invention there are provided PCR primers suitable for the amplification of a nucleic acid encoding AFAI or a portion thereof. In some instances the portion will include at least one CDR region.
In an embodiment of the invention there is provided a polypeptide sequence comprising at least 90 amino acids including at least one of the following three contiguous amino acid sequence: KNLMG SEQ ID No.4 TISGSGGTNYASSVEG SEQ ID NO.6, and AFAI SEQ ID NO.3.
In an embodiment of the invention there is provided the use of an AFAI-derived polypeptide and/or a polypeptide having at least 90% identity to SEQ ID NO. 1 and/or a portion thereof in forming a conjugate by grafting the polypeptide to an antigen binding fragment. In some instances one or more of the AFAI CDR's is grafted onto an antigen binding fragment including, for example, a VHH, VH or VL framework (scaffold) or an immunoglobin and/or fragment thereof (e.g. Fab, scFv). One or more AFAI CDR's may also be used to produce a fusion protein wherein the second polypeptide sequence provides a useful functionality or property. In some instances it may be desired to produce humanized variants of the AFAI antibody using techniques known in the art. Such humanized antibodies are specifically contemplated herein. In some instances it will be desired to conjugate AFAI or a portion thereof to self assembly molecules to allow for the formation of multi meric complexes having enhanced antigen-binding properties.
In an embodiment of the invention there is provided a conjugate of a polypeptide containing at least one of the three contiguous amino acid sequences and a cargo molecule or molecules. The cargo molecule may be useful for diagnosis or treatment of carcinoma. For example, it may be an enzyme or radioisotope useful in the identification and localization of cells of interest in tissue or it may be a cytotoxic agent such as a drug, further strong antigen, apoptosis inducer or radioisotope useful in reducing the viability or ability to proliferate of a carcinoma cell.
The inclusion of a reference is not an admission or suggestion that it is relevant to the patentability of anything disclosed herein.
A person understanding this invention may now conceive of alternative structures and embodiments or variations of the above all of which are intended to fall within the scope of the invention as defined in the claims that follow.