Single Chain Antibody Production

Single Chain Antibody Production ABSTRACT The study aimed to characterise His- and Myc- tagged scFv MFE-23 antibodies produced from transformed E.coli cultures using ELISA and immunohistochemistry assays. Anti-His, anti-myc and anti-MFE secondary antibodies were used in the CEA/PBS coated ELISA plate with horseradish peroxidase-OPD chromatic reaction for detection. Culture 1 was identified to produce MFE-His and culture 2 giving MFE-Myc antibodies. The immunohistochemistry assay confirmed the CEA binding profile of scFv MFE-Myc by the comparison between negative controls, positive anti-CEA binding reactions and 4-stage anti-Myc binding of the scFv MFE-Myc examined. The CEA specificity displayed by tagged scFv MFE-Myc can be utilised in antibody-based cancer therapy by targeting tumour antigens specifically. INTRODUCTION Cancer arises due to the defective regulation of normal cell proliferation and homeostasis. This allows tumours to possess the capabilities including self-sufficiency in growth signal, insensitivity to antigrowth signals, avoidance of apoptosis, unlimited replicative potential, sustained angiogenesis and metastasis.1 In additional to the surgical removal of tumours, conventional cancer therapies such as radiation, chemotherapy and immunotherapy have a focus in inducing cytotoxicity against malignant cells. With better understanding of the molecular biology of carcinogenesis, targeted therapies are being developed to achieve lower toxicity to normal tissues and higher clinical efficacy through disrupting pathways that contribute to the tumours proliferative advantage. Attempts are also being made in developing cancer gene therapies to compensate or repair the mutated genes. All tumours express their unique set of antigens on cell surface which can be a result of genetic alterations, upregulated self-antigens or tissue-specific antigens that can be utilized to distinguish cancer cells from normal cells. Antibodies, one of the effectors in immune response, are Y-shaped proteins that each recognizes and binds to specific antigen (Figure 1). The protein consists of two light chains and two heavy chains, with variable (Fab) and constant (Fc) domains on each of the chain. The region of the antibody at which antigen binds is referred as the Complementary Determining Regions (CDR) present at the variable Fab region. The Fc region is responsible in modulating immune response through activation of the complement cascade and Fc receptor mediated activation of effectors such as phagocytes, mast cells, neutrophils and Natural Killer (NK) cells.3 The initial approach of antibody-based cancer therapy is to tag cancerous cells as foreign and eliminate such targets thro ugh cytotoxic effectors of the human immune system. Antibodies against essential growth factors can be synthesized to sequester such molecules from further promoting tumour growth. The high specificity of antibodies to particular antigens also serves as a vehicle in delivering killing machineries to tumour cells. For example, antibodies can be directly conjugated to radionuclides, toxins or cytokines, or indirectly to the surface of liposomes carrying drugs or toxins.4 Antibody-directed enzyme prodrug therapy (ADEPT) is also available in which the antibody targets an enzyme selectively to the tumor where it converts a relatively non-toxic prodrug to a potent cytotoxic drug.4 These various strategies aim to minimize the systemic toxicity afflicted by the cytotoxic agents administered. Monoclonal antibodies are usually preferred compared to polyclonal ones, as they recognize specific epitope of an antigen and hence have greater specificity. Carcinoembryonic antigen (CEA) is first identified as a glycoprotein in the human colon cancer tissue extract and fetal gut, and plays a role in cell adhesion.5 Although CEA can also be detected in normal gastrointestinal tissue, the glycoprotein is overexpressed on the plasma membrane of colon cancer tissues.5 CEA level has also been found to be highly elevated in various cancers of an epithelial origin such as breast, lungs and pancreas.5,6 Moreover, normal CEA is localised on the luminal surface of columnar epithelial cells lining the crypts of the intestine so the glycoproteins are not directly accessible to the blood flow.6 However CEA is usually found on all sides of the cell membranes in tumours. Thus, CEA can be a useful target on cancer cells in immunotherapy using anti-CEA antibodies. The objectives of this lab practical are to introduce the applications of antibodies in cancer therapy and diagnosis through immunohistochemical staining of tissues. It also serves to familiarise students with the enzyme linked immunosorbent assay (ELISA). This is achieved by the production of MFE-23 single chain Fc antibody fragment (scFv) against CEA from transformed E.coli cultures. The scFv MFE-23 can be linked to either a His- or a Myc- tag. The unknown antibodies are characterised in the ELISA assay using appropriate detection antibodies, and the chromatic reaction between horseradish peroxidise (HRP) and OPD substrate. The scFv MFE-23 obtained by students is also used in the immunohistochemical characterisation in cryostat sections of normal and cancerous human tissues. MATERIALS METHODS E. Coli growth curves E.coli cultures were transformed with pUC119 containing either His- or Myc- tagged scFv MFE-23 (Culture 1 2) and were incubated overnight. The expression of pUC119 was controlled by the lac operon, which could be induced by either lactose or lactose analogue isopropyl-1-szlig;-D-thiogalactoside (IPTG). The vector also encoded for ampicillin resistance. The E.coli cultures were grown with ampicillin selection and 0.05% glucose. Optical density (OD) readings at 600nm were taken at 30min interval until OD=0.9 when IPTG was added to both cultures to induce pUC119 expression and hence scFv MFE-23 production. A negative control of 2xYT was set up, and all three cultures were incubated at 30? overnight. Full experimental procedures are described in Appendix 1. ELISA assay The ELISA assay aims to characterize the identities of the tags conjugated to scFv MFE-23 obtained in the supernatant of the overnight E.coli cultures. Detailed protocols of the assay can be found in Appendix 1. 36 wells of the 96-well ELISA plate were coated with the CEA antigen or phosphate-buffered saline (PBS) as its negative control. Supernatant of the overnight E.coli cultures 12 from the bacterial growth curve assay was obtained which contain either His- or Myc- tagged scFv MFE-23. The supernatant samples 12, the positive control MFE-his-myc antibodies and the negative control 2xYT growth medium were added to the corresponding wells as indicated in the format diagram in Appendix 2. Secondary antibodies rabbit anti-MFE23 polysera, mouse monoclonal anti-HIS tag (TetraHis, Qiagen) and mouse monoclonal anti-MYC (Sigma) were added to the corresponding wells (Appendix 2) to bind the primary antibodies present. Tertiary horseradish peroxidase (HRP) conjugated antibodies against the secondary antibodies goat anti-rabbit HRP (sigma) and sheep anti-mouse HRP (Sigma) in blocking solution were added to the wells (Appendix 2). OPD substrate buffer was applied to each well to detect the presence of HRP, which should give a yellow-orange product in case of positive result. HCl was added to stop the reaction when colour has developed, and the OD at 490nm for each well was measured. Immunohistochemistry assay Five glass slides, each containing 2 colonic adenocarcinoma, 1 normal colon and 1 normal liver tissue sections were fixed and processed for immunohistochemical staining. Avidin-biotin-peroxidase complex (ABC complex) was added following the application of biotin-labelled antibodies. The localisation of antigens was visualised by the formation of brown pigments, as peroxidise reacts with the diaminobenzidene (DAB) substrate. Lattices of several peroxidase molecules were formed to amplify the binding signal from the biotinylated antibody.7 A summary of the treatment given to each slide was illustrated in Table 1, and the complete protocol can be found in Appendix 1. RESULTS E. Coli growth curves Both E.coli cultures transformed with either His-tagged scFv MFE-23 or the Myc-tagged version in a pUC119 vector follow a similar exponential growth curve, as shown by the plot of OD600 against time (Figure 2). The bacterial cultures were in lag phase at t=0-90, and the log phase from t=90. It took approximately 175minutes for the cultures to reach OD600=0.9. The PBS negative controls in wells A-F/7-12 worked relatively accurately with low OD490 readings. Most readings also corresponded to the negative controls in the CEA coated wells (B4-6, D4-6, F4-6) but with 2xYT added instead of primary antibodies. However, the OD490 readings for wells C10-12 and E10-12 were higher than most readings from the negative controls. The higher OD490 readings obtained in certain wells in comparison of their corresponding negative controls indicated the presence of yellow-orange product formation from reaction between HRP and OPD substrate. These positive outcomes found in certain wells are highlighted in the shaded cells of Table Immunohistochemistry Assay Figure 4 shows the immunohistochemical binding reactions for the negative controls (slides 3-5). Slide 3 acts as a negative control for slide 1, treated with the 4-stage anti-Myc technique but omitting the primary Myc-tagged scFv MFE-23 antibody. Thus we should not be able to visualize the localization of CEA antigens due to the absence of MFE-23 binding on slide 3. Slide 4 is a negative control for slide 2, omitting the mouse monoclonal anti-CEA A5B7 antibody treatment in the 3-stage mouse monoclonal technique. No binding reaction is expected on slide 4 as well. Slide 5 was treated like slide 4 but without initial biotin/avidin blocking. As expected, the colonic adenocarcinoma tissue section of slide 3 does not display brown colouration and hence there is no binding reaction to the cytoplasm of tumour cells and connective tissues. Binding reaction to the cytoplasm of cryptal epithelium is not seen for the normal colonic mucosa, although there are strong brown colourations for a few cells in the lamina propria. The normal liver tissue on slide 3 shows some weak reaction with the parenchymal cells. Similarly, the parenchymal cells of the normal liver tissue on slide 4 do not display brown colouration and hence indicates the absence of binding reaction. On slide 5, the normal parenchymal cells are positive for binding reaction, demonstrating the presence of biotin in normal liver. Figure 5 shows the immunohistochemical binding results of slides 1 and 2. Slide 1 was treated with the 4-stage anti-Myc technique and slide 2 is the positive control treated with the 3-stage mouse monoclonal technique (See Table 1). For the colonic adenocarcinoma tissue on slide 1, strong brown colouration is present in the cytoplasm of tumour cells and the basement membrane of malignant acinar structures. Weak positive binding reactions can also be observed in fibrovascular stroma. The normal colonic mucosa of slide 1 shows strong reactions to the cytoplasm of goblet cells in the cryptal epithelium as well as a few cells in the lamina propria. The normal liver tissue shows only weak positive reactions with the parenchymal cells. Thus the scFv MFE-myc antibody was reactive with both normal colonic epithelium and adenocarcinoma, but not the biotin/avidin blocked liver tissue. The colonic adenocarcinoma tissue on slide 2 shows strong reactions with the tumour cell cytoplasm and the basement membrane of malignant acinar structures similar to the reactions seen in slide 1. Weak positive results are obtained in fibrovascular stroma of the positive control slide. As the binding reaction of slide 1 was similar to that of the positive control, this confirms the CEA reactive profile of the scFv MFE-myc antibody from the E.coli supernatant sample 2. DISCUSSION E.coli Growth Curve Both E.coli cultures followed the exponential growth curve as expected. However, the growth curve was obtained in the absence of a negative control i.e. same volume of 2xYT to be treated in the same way as the two cultures. The lack of a proper negative control means that the possibility of contamination cannot be eliminated. Thus it is unknown whether the increase in OD600 readings was partially attributed to culture contamination. ELISA Assay As mentioned in the results, the OD490 readings for wells C10-12 and E10-12 were higher than most readings from the negative controls. This might indicate contamination of these wells with CEA antigens, or insufficient PBS washing following the application of HRP-conjugated antibodies. Wells A-B/1-6 were applied with anti-His secondary antibodies and so would indicate the presence of His-tag by the production of yellow-orange product. Wells A1-3 were treated with supernatant from bacterial culture 1 and wells A4-6 with that of culture 2. As higher readings of OD490 in the positive control B1-3 and wells A1-3 were obtained compared to the negative control B4-6, the culture 1 supernatant contained the His-tagged scFv MFE-23. Wells C-D/1-6 were applied with anti-Myc secondary antibodies and so would indicate the presence of Myc-tag by the production of yellow-orange product. Wells C1-3 were treated with supernatant from bacterial culture 1 and wells C4-6 with that of culture 2. As higher readings of OD490 in the positive control D1-3 and wells C4-6 were obtained compared to the negative control D4-6, the culture 2 supernatant contained the Myc-tagged scFv MFE-23. The anti-MFE antibodies added to wells E-F/1-12 can bind to both His- or Myc-tagged scFv MFE-23. Thus binding should occur against primary antibodies from both cultures 12 and also the positive MFE-myc-his control, as demonstrated by the higher OD490 readings in wells E1-6 and F1-3 compared to the negative controls F4-6. It can be observed that the anti-His antibody gave a stronger signal than anti-Myc and was due to anti-His binding more strongly to its target than anti-Myc (Unpublished results, Kogelberg, H.). Comparing the OD490 of wells E1-3 and E4-6, absorbance of anti-Myc was slightly higher than anti-His despite the lower binding affinity of anti-Myc. Thus there might be a higher concentration of MFE-Myc in the culture 2 supernatant than MFE-His in culture 1, although technical issues like washing times, salt concentration and pH, or structural characteristics affecting the accessibility of antibody can affect the amount of binding. Immunohistochemistry Assay The results obtained from slides 1 and 2 confirm the CEA reactive profile of scFv MFE-myc antibody, as both slides displayed similar binding reaction patterns in colonic adenocarcinoma. The weak positive signal in normal liver tissue in slide 1 is likely to be caused by cross-reaction of secondary mouse anti-Myc antibodies, as similar result can be observed in the negative control (slide 3). Primary antibodies may also cross react and bind to non-target tissues. This demonstrates the possible cross-reactions with antibodies and hence the importance of negative controls to eliminate such artifactual reactivity. The strong binding to goblet cells cytoplasm in cryptal epithelium of normal colonic mucosa in slide 1 is consistent with the findings of CEA present in normal colonic mucosa.5 Although CEA level is lower on normal colonic mucosa, their presence implies that antibodies against CEA in cancer therapy may target normal cells other than malignant cells. Thus it is important to control anti-CEA antibody concentration used to avoid imposing toxicity to normal tissues yet is effective in producing a clinical response on cancer. The negative controls (slides 3-5) allowed the assessment of the level of binding for secondary, tertiary and quaternary antibodies/reagents. The slides also revealed any endogenous background material that might be confused with specific binding of primary antibodies, as well as information about the basic pathology of tissues. The inclusion of normal liver sections helped illustrate the importance in carefully controlling the specific reactions of test antibodies with potential targeting specificity for other cellular proteins. The ABC complex for detecting biotinylated horse anti-mouse antibodies can also react with biotin present in liver and give a positive non-specific result. This unwanted reaction as seen in slide 5 can be prevented by biotin blocking treatment of tissue samples. Moreover, myeloperoxidase white cells containing endogenous peroxidase can react with the DAB substrate and produce brown colouration even in the absence of ABC complexes. The normal colonic mucosa in slide 3 provided examples of such artifact. The naturally occurring bile pigments in liver seen as green/brown granules under high-power bright field microscopy can also be found occasionally. Careful interpretation of slides is required to avoid false judgments as there can be false positive reactions. In addition, it should be noted that adequate fixation of the sample is important in successful localisation of antigens present and hence an accurate representation of antigen distribution profile. The procedure helps to ensure the preservation of tissue morphology, the immobilisation of antigen and the preservation of antigen immunoreactivity. It is also important to ensure optimal fixation for an adequate permeability of the tissue to the immunochemical reagents. Other than mmunohistochemistry, western blotting and immunoprecipitation (IP) can be carried out to confirm scFv specificity. Antigens transferred to nitrocellulose membrane can be probed by specific antibodies in western blot, or precipitated out of lysate using antibodies in IP. Affinity chromatography can also be used which is a method of separating biochemical mixtures based on highly specific biological interaction. Specific antigen can be covalently coupled to a solid support and allow supernatant with the testing antibody to flow through so that only specific antibodies will be bound to the antigens. The assay used in this study can only detect antigens in a non-quantitative way with chromatic display of antigen localisation. Instead radioimmunoluminography (RILG) can be used for quantitative measurements of antigen concentration along with its distribution in histological sections.8 Radiolabelled antibodies against specific antigen can be applied to tissue sections and bound antibodies are mapped by phosphor imaging. Radioactivity detected in each pixel of the digital image will be proportional to antigen concentration if saturating antibody concentration is used. Future Perspective of Antibody Targeted Cancer Therapy The successful application of scFv MFE-23 with a Myc-tag at its C-terminus for detection in immunohistochemistry proves that the attachment of small molecules to the antibody will not affect its specificity for CEA. Thus the scFv chain can be conjugated to cytotoxic reagents or to be used for ADEPT as mentioned previously. The potential clinical efficacy of scFv MFE-23:enzyme fusion protein has been shown in nude mice with human colon adenocarcinoma xenografts by the Bhatia group.9 Moreover, scFv MFE-23 can be used in radioimmunoguided surgery (RIGS) based on the pre-operative injection of a radiolabelled anti-tumor antibody to detect tumour deposits during surgery. A Phase I clinical trial of RIGS using 125iodine-labelled MFE- 23-his scFv has reported good selective localisation at sites of primary colorectal cancer and metastases.10 As illustrated by MFE-23 scFv fragment, antibody targeting has the potential for selective imaging or delivery of anti-cancer molecules. Antibodies can be engineered to modify their biological properties with increased specificity and functionality. This is achieved by reducing antibody size, altering valency, and fusing to different molecules to improve therapeutic efficiency. Scientists have been trying to produce smaller antibody fragments but retaining specific binding to antigens, in order to minimize immunogenicity and achieve better tumour penetration. Continuous research on the specificity and stability of these fragments, and hunting for more tumour-specific antigens are required to further expand the field of antibody-targeted cancer therapies. Slide 3 is a negative control for slide 1 which was treated with the 4 stage anti-Myc technique. Slide 4 is a negative control for slide 2 which was treated with the 3 stage mouse monoclonal technique. Slide 5 is a negative control for slide 2 as well but without biotin/avidin blocking. Slide 1 was treated with the 4 stage anti-Myc technique, and slide 2 was treated with the 3 stage mouse monoclonal technique. REFERENCES Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. 2000 Jan 7;100(1):57-70. Baron, E. J. 1996. Classification. In S. Baron et al., eds. Barons Medical Microbiology, 4th edition. University of Texas Medical Branch. Gura, T. Therapeutic antibodies:Magic bullets hit the target. Nature 417, 584-586 (6 June 2002) Carter, P. Improving the efficacy of antibody-based cancer therapies Nature Reviews Cancer 1, 118-129 (November 2001) Sten Hammarstrà ¶m. The carcinoembryonic antigen (CEA) family: structures, suggested functions and expression in normal and malignant tissues. Seminars in Cancer Biology. Volume 9, Issue 2, April 1999, Pages 67-81 A. Mayer, K. A. Chester, et al. Taking engineered anti-CEA antibodies to the clinic. Journal of Immunological Methods. Volume 231, Issues 1-2, 10 December 1999, Pages 261-273 Hsu SM, Raine L, Fanger H. Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem. 1981 Apr;29(4):577-80. GBoxer, SStuart-Smith, et al. Radioimmunoluminography: a tool for relating tissue antigen concentration to clinical outcome. British Journal of Cancer (1999) 80, 922-926. Bhatia J, Sharma SK, et al. Catalytic activity of an in vivo tumor targeted anti-CEA scFv:carboxypeptidase G2 fusion protein. Int J Cancer. 2000 Feb 15;85(4):571-7. Mayer, A., et al. Radioimmunoguided Surgery in Colorectal Cancer Using a Genetically Engineered Anti-CEA Single-Chain Fv Antibody. Clinical Cancer Research May 2000 6; 1711. APPENDIX Appendix 1 Experimental Procedures (UCL Cancer MSc Lab Practical 2 Handout, 12-13Nov09) E. Coli growth curves E.coli culture transformed with pUC119 containing either His- or Myc- tagged scFv MFE-23 (Culture 1 2) were incubated overnight. The expression of pUC119 was controlled by the lac operon, which could be induced by either lactose or lactose analogue isopropyl-1-ß-D-thiogalactoside (IPTG). The vector also encoded for ampicillin resistance. 30 µl of the media was added to 15ml growth media with ampicillin selection and 0.05% glucose. Both tubes on loose caps were placed into a 37? shaker at 225rpm and time was taken as t=0. Optical density (OD) of the culture was measured with a spectrophotometer at 600nm from t=0 at a 30min interval until reading reached OD=0.9. 1mM IPTG was added to each culture and the tubes were incubated at 30? overnight with shaking. A negative control was set up using the same volume of growth media 2xYT with 1mM IPTG for overnight incubation. ELISA assay 100 µl of 10 µg/ml CEA antigen was applied to strips 1-6, A-H of the 96-well ELISA plate. 100 µl of PBS was applied to strips 7-12, A-H as negative controls. The plate was covered with plastic film and incubated for 1 hour at room temperature. Each well was rinsed 4 times with PBS and blocked with 200 µl 5% Marvel milk/PBS. The plate was covered and incubated overnight at 4?. The plate was washed with PBS 4 times after the overnight incubation. Overnight bacterial cultures and the negative control were centrifuged at 4000rpm for 20minutes to obtain the supernatant. 100 µl of each supernatant sample and negative control was added to the ELISA plate according to the format diagram as seen in Appendix 2. The covered plate was left at room temperature for an hour. The plate was then rinsed twice with 0.1% Tween-20/PBS and 4 times with PBS. 100 µl of secondary antibodies rabbit anti-MFE23 polysera, mouse monoclonal anti-HIS tag (TetraHis, Qiagen) and mouse monoclonal anti-MYC (Sigma) with a concentration of 1:1000 in 1% blocking solution was added to the wells according to the format diagram: anti-His A1-12 and B1-12; anti-Myc C1-12 and D1-12; anti-MFE E1-12 and F1-12. The covered plate was left at room temperature for an hour. The plate was washed again twice with 0.1% Tween-20/PBS and 4 times with PBS. 100 µl horseradish peroxidase (HRP) conjugated antibodies against the secondary antibodies 1:1000 goat anti-rabbit HRP (sigma) and 1:500 sheep anti-mouse HRP (Sigma) in 1% blocking solution was added to the corresponding wells. The covered plate was left at room temperature for an hour, and then rinsed twice with 0.1% Tween-20/PBS and 4 times with PBS. 100 µl of OPD substrate buffer was added to each well to detect the presence of HRP, giving a yellow-orange product in the case of positive outcome. 100 µl of 4M HCl was added when the colour has developed, and OD was measured at 490nm with an automated plate reader. Immunohistochemistry assay Five glass slides, each containing 2 colonic adenocarcinoma, 1 normal colon and 1 normal liver tissue sections were removed from the freezer and air dried for 5minutes. Slides were fixed in acetone in 10minutes and rinsed in tap water for 2minutes. Slides were flooded with PBS and Avidin blocking solution was applied to slides 1-4 for 10 minutes. Slides 1-4 were rinsed in PBS and applied with Biotin blocking solution for 10minutes followed by PBS wash. PBS was removed from slide 1 and flooded with 1:20 normal horse serum in PBS for 15minutes. Slide 1 was drained and applied with bacterial supernatant containing Myc-tagged scFv MFE-23 antibodies for 45minutes. Slide 1 was then rinsed with PBS thrice over 10minutes, and 1:20 normal horse serum in PBS was added to slides 2-5 for 15minutes. 20 µg/ml of mouse monoclonal anti-Myc antibody (Qiagen) in PBS was added to slides 13 and left for 35minutes. 20 µg/ml of A5B7 anti-CEA antibody was added to slide 2 for 35minutes. Slides 4-5 were washed in PBS. Slides 1-3 were rinsed with PBS thrice over 10minutes. 1:200 biotinylated horse anti-mouse immunoglobulins in PBS with 5% normal human serum was added to all slides and incubated for 35minutes. All slides were rinsed with PBS thrice over 10minutes. The avidin biotin-peroxidase complexes reagent (ABC reagent) was added to all slides, followed by PBS wash 3 times over 10minutes. 0.03% of hydrogen peroxide in 1mg/ml 3,3,diaminobenzidene tetrahydrochloride solution was applied to all slides immediately and left for 5minutes. Slides were washed with tap water. All slides were stained in Harris haematoxylin for 40seconds, rinsed in tap water and left for 5minutes for colour to develop. Slides were then dehydrated through graded alcohols (70/95/100%), cleared in inhibisol and coverslip using DPX mountant. The slides were then ready for observations under the light microscope.