by Takao Hamakubo1, Osamu Kusano-Arai1,2, Hiroko Iwanari1
1Department of Quantitative Biology and Medicine Research Center for Advanced Science and Technology, The University of Tokyo
2Institute of Immunology Co. Ltd.
Membrane proteins are a major drug target in cancer immunotherapy. Monoclonal antibodies (mAbs) against membrane proteins on target cells have attracted growing attention as a specific probe for drug delivery systems to these proteins [1, 2].
The major cancer immunotherapy target membrane proteins are thus far, cell surface receptors or adhesion molecules . There is one mAb against a G protein-coupled receptor (GPCR) on the market for the treatment of leukemia. However, most others are single-pass membrane receptors, and there are no ion channels or transporter proteins at present in either a phase 3 trial or in clinical use . A mAb kills cancer cells by inducing antibody-dependent cellular cytotoxicity (ADCC) and by complement-mediated cytotoxicity (CMC). Both require a large amount of antigen on the cell surface. Recently, “armed” mAb therapeutics, such as antibody-drug conjugates (ADC), radioimmunoconjugates, bispecific antibodies with immunoadaptive recognition sites and also chimeric antigen receptor (CAR) T cells [4-6] have come to market. As these approaches have enhanced capacity for killing target cells, it is expected that less abundant proteins such as GPCRs or other multi-pass membrane proteins will become new target for cancer therapeutics. Thus, the efficient generation of mAb against membrane proteins is suggested to be of growing importance. Nevertheless, there are still difficulties associated with the generation of effective mAbs against multi-pass membrane proteins.
We would like to introduce here our strategy for a baculovirus display technology that generates mAbs against membrane proteins.
2. Membrane protein display on the baculovirus envelope
The problems for targeting multi-pass membrane proteins include: (1) the difficulty of the preparation of a large amount conformationally correct proteins , (2) the immunological tolerance that arises due to the high sequence homology between species in many biologically important proteins, and (3) the difficulty of obtaining high affinity antibodies due to the lack of an extracellular loop for the purpose of an antigenic epitope. The use of peptide fragments or small domains as immunogens frequently proves to be unsuccessful means of obtaining specific antibodies which recognize cell surface antigens. This is due to the conformational difference between the partial peptide and the whole protein.
In recent years, the method of membrane protein preparation for the purpose of structural study has been improved such that various multi-pass membrane proteins, including GPCRs, have come to be prepared in a sufficiently large amount for crystallography . Once the solubilization method is established, membrane proteins are reconstituted into phospholipid vesicles in which the adjuvant molecule is also incorporated for immunization . However, this method is only suitable when a large amount of purified protein is available.
In the course of expressing membrane proteins of endoplasmic reticulum origin using a baculovirus expression system, we noticed a relatively large amount of membrane proteins were displayed on the particles of budded baculovirus (BV)  (Fig.1). We confirmed the expression of several known ER proteins, including sterol responsive element binding protein 2 (SREBP-2) and SREBP cleavage activating protein (SCAP) on BV . We then investigated other membrane proteins and found this BV display system to be useful for antibody generation against multi-pass membrane proteins which are difficult because they require the purification of a large amount of material . In addition to this whole protein display method, there is a baculovirus display technique using fragment peptide as a fusion protein with the viral membrane protein gp64 [11, 12].
3. Multiple membrane proteins displayed in functional complexes on BV
The expression of functional membrane proteins on BV particles with the β2-adrenergic receptor, a member of GPCR family, was first reported by Loisel et al . We have come across a similar phenomenon with membrane proteins of endoplasmic reticulum origin . We thus checked the expression of GPCR on the BV using leukotriene B4 (LTB4) receptor (BLT1) . BLT1 couples to the Gi isoform of a trimeric G-protein which inhibits adenylyl cyclase. We were only able to recover the highly sensitive LTB4 binding activity on the BV when recombinant baculoviruses harboring trimeric Gi-protein subunit genes were co-infected  (Figure 2). The effector protein adenylyl cyclase, a multi-pass membrane protein, was also expressed at the same time. As a result, the whole signaling complex was reconstituted on BV .
We have confirmed a couple of dozen GPCRs are suitable for BV display, although the expression level differed by an order of magnitude between some of them. Interestingly, even the odorant receptors, which are well known GPCRs that are difficult to be expressed on the surface of mammalian cells, are capable of being displayed on BV in functional complexes . Thus not only is BV able to display single pass membrane proteins, but also multiple proteins, so that the functional protein complex can be reconstituted on the virus. The most dramatic example is the γ-secretase complex, which is only active when all four membrane protein components are assembled .
We have shown the BV display of multi-pass membrane proteins other than receptors, such as transporters (pepT1 ), channels (aquaporin 4 , band 3 ), and enzymes (DHHC palmitoylation enzymes ). All of the membrane proteins tested to date, with very few exceptions such as vesicle transfer proteins, are functionally displayed on the viral particles. This functional expression of membrane proteins is highly useful for the construction of a variety of assay systems .
4. Immunization of displayed membrane proteins
Once the recombinant virus with the target membrane protein gene is established, it is easy to obtain a sufficient amount of BV as the antigen for immunization. It is preferable to prepare the BV sample just before the treatment, or ethylenediaminetetraacetic acid (EDTA) and E-64 are added for one to two months with preservation at 4 °C. We usually use pertussis toxin for the adjuvant because other lipid-based adjuvants may cause the destruction of the envelope or result in the unwanted solubilization of the displayed membrane protein .
BV immunization evokes a potent immune reaction against its envelope protein, gp64. We found that too many clones reacted with gp64, which made it difficult to get the reactive clones to the site of the target protein. To avoid such excess reactivity to gp64, we generated transgenic mouse harboring the gp64 gene. These gp64-transgenic mice exhibited the expression of large amount of the gp64 protein and became tolerant to gp64 . The hybridoma clones obtained from the gp64-transgenic mice exhibited much less reactivity to gp64. Thus, the monoclonal antibodies with ADCC and complement-dependent cytotoxicity (CDC) activities against peptide transporter 1 (PepT1) expressing pancreatic tumor cells were generated by using this technique .
5. Use of monoclonal antibodies as a probe for cancer diagnosis and therapy
In search of better antibody therapeutics for cancer, second-generation mAb technology has successfully found several way to move forward. Antibody drug conjugate (ADC) has been reported to exhibit effective clinical results [4, 24]. Furthermore, radioisotope labelled immunotherapy and imaging are reportedly on their way [25, 26]. In another vein, the retargeting of cellular immunity with a bispecific antibody or CAR T cells also demonstrated powerful cytotoxic activity against cancer cells [6, 27]. These are among the strategies for using mAbs as a specific probe to antigens, and can be explored in combination with other techniques, such as nanoparticle drug delivery systems (Figure 3), in the future. Specific mAbs against cell surface antigens should also prove to be useful for collecting circulating tumor cells or exosomes as a liquid biopsy for personalized medicine (Figure 3).
We have been exploring the possibility of radioimmunotherapy (RIT) with a mAb against Robo1. Robo1 is an axonal guidance receptor with five immunoglobulin-like domains and three fibronectin-like domains in its extracellular portion, we have obtained reactive antibodies to Robo1 by immunizing the BV-displayed whole protein. Robo 1 is over-expressed on liver cancer cells and is considered a good candidate for the immunotherapy . The monoclonal antibodies generated with BV immunization belong mostly to the IgG2 subclass, so the ADCC activity is relatively weak. However, the 90Y radioisotope-labelled form exhibited good anti-tumor activity against HepG2 cells and also against the NCI-H69 cells, a small cell lung carcinoma, a xenograft tumor implanted in immunologically deficient mice [29, 30]. These results indicate the BV display technique and the monoclonal antibody generated by this method should prove to be useful for cancer treatment.
We are now developing a Kinetic Exclusion Assay (KinExA) for BV display. The kinetics of KinExA provides a precise evaluation of the affinity of the antibody and the amount of antigen. This will greatly expand the possibility of generating high affinity therapeutic mAbs, not only to cancer targets, but also many other targets, including viral antigens.
We thank Dr Kevin Boru of Pacific Edit for review of the article. This work was supported by the Program for Development of New Functional Antibody Technologies of the New Energy and Industrial Technology Development Organization (NEDO) of Japan, the Funding Program for World-Leading Innovative R&D on Science and Technology (FIRST program) of Japan Society for the Promotion of Science.
About the Authors
Dr. Takao Hamakubo is a professor of Department of Quantitative Biology and Medicine of Research Center for Advanced Science and Technology at The University of Tokyo, Japan. He was trained and obtained PhD from Kyoto University, and then conducted research on experimental hypertension in Vanderbilt University, USA. He started research in RCAST, University of Tokyo (UT). in 1996 and became an endowed chair professor of FUJIFILM from 2013. He has invented the technique of membrane protein display on baculovirus and is now advancing the development of therapeutic antibodies for cancers. He is one of the founders and now an outside board member of Perseus Proteomics Inc.
Dr. Hiroko Iwanari is a specially appointed associated professor of Department of Quantitative Biology and Medicine of Research Center for Advanced Science and Technology at The University of Tokyo, Japan. She was received MA from Kyoto University and PhD from The University of Tokyo. After working at the Institute of Immunology Co., Ltd., Tokyo, she started developing antibody production methods using baculovirus and developed antibodies against proteins, including nuclear receptors, related to many diseases in a NEDO project from 2000. She is now developing antibodies against polytopic transmembrane proteins, such as cancer-specific membrane proteins, GPCRs and transporters.
Mr. Osamu Kusano-Arai is a doctoral course student of the Department of Advanced Interdisciplinary Studies, Graduate School of Engineering, The University of Tokyo. He received MPharm and BPharm from Tokyo University of Pharmacy and Life Sciences. Since 2008, he has been working at RCAST, the University of Tokyo, as a member of NEDO antibody project on loan from the Institute of Immunology Co., Ltd., Tokyo. He is working on the production of antibodies against disease-related proteins using baculovirus display with the aim of developing antibodies for the diagnosis and therapy of intractable diseases such as cancer.
- Glassman PM, Balthasar JP. Mechanistic consideration for the use of monoclonal antibodies for cancer therapy. Cancer Biol Med 2014; 11:20-33.
- Reichert JM. Antibody-based therapeutics to watch in 2011. mABs 2011; 3:76-99.
- Hamakubo T, Kusano-Arai O, Iwanari H. Generation of antibodies against membrane proteins. Biochim Biophys Acta 2014; 1844:1920-4.
- Zolot RS, Basu S, Million RP. Antibody-drug conjugates. Nat Rev Drug Discovery 2013; 12:259-60.
- Pardoll DM. The blockade of immune checkpoints in cancer immunotherapy. Nat Rev Cancer 2012; 12:252-64.
- Weiner GJ. Building better monoclonal antibody-based therapeutics. Nat Rev Cancer 2015; 13:361-70.
- Junge F, Schneider B, Reckel S, et al. Large-scale production of functional membrane proteins. Cell Mol Life Sci 2008; 65:1729-55.
- Carpenter EP, Beis K, Cameron AD, et al. Overcoming the challenges of membrane protein crystallography. Curr Opin Struct Biol 2008; 18:581-6.
- Hino T, Iwata S, Murata T. Generation of functional antibodies for mammalian membrane protein crystallography. Curr Opin Struct Biol 2013; 23:563-8.
- Urano Y, Yamaguchi M, Fukuda R, et al. A novel method for viral display of ER membrane proteins on budded baculovirus. Biochem Biophys Res Commun 2003; 308:191-6.
- Boublik Y, Bonito PD, Jones IM. Eukaryotic virus display: engineering the major surface glycoprotein of the Autographa californica nuclear polyhedrosis virus (AcNPV) for the presentation of foreign proteins on the virus surface. Biotechnology 1995; 13:1079–84.
- Grabherr R, Ernst W, Doblhoff-Dier O, et al. Expression of foreign proteins on the surface of Autographa Californica nuclear polyhedrosis virus. Biotechniques 1997; 22:730–5.
- Loisel TP, Ansanay H, St-Onge S, et al. Recovery of homogeneous and functional beta 2-adrenergic receptors from extracellular baculovirus particles. Nat Biotechnol 1997; 15:1300–4.
- Masuda K, Itoh H, Sakihama T, et al. A combinatorial G protein-coupled receptor reconstitution system on budded baculovirus: evidence for Gαi and Gαo coupling to a human leukotriene B4 receptor. J Biol Chem 2003; 278:24552–62.
- Sakihama T, Masuda K, Sato T, et al. Functional reconstitution of G-protein-coupled receptor-mediated adenylyl cyclase activation by a baculoviral co-display system. J Biotechnol 2008; 135:28-33.
- Mitsui K, Sakihama T, Takahashi K, et al. Functional reconstitution of olfactory receptor complex on baculovirus. Chem Senses 2012; 37:837-47.
- Hayashi I, Urano Y, Fukuda R, et al. Selective reconstitution and recovery of functional gamma-secretase complex on budded baculovirus particles. J Biol Chem 2004; 279:38040-6.
- Saitoh R, Ohtomo T, Ito Y, et al. Recovery of functional peptide transporter PepT1 in budded baculovirus fraction. Protein Expr Purif 2006; 46:130-5.
- Miyazaki K, Abe Y, Iwanari H, et al. Establishment of monoclonal antibodies against the extracellular domain that block binding of NMO-IgG to AQP4. J Neuroimmunol 2013; 260:107-16.
- Arakawa T, Kobayashi-Yurugi T, Alguel Y, et al. Crystal structure of the anion exchanger domain of human erythrocyte band 3. Science 2015; 350:680-4.
- Noritake J, Fukata Y, Iwanaga T, et al. Mobile DHHC palmitoylating enzyme mediates activity-sensitive synaptic targeting of PSD-95. J Cell Biol 2009; 186:147-60.
- Sakihama T, Sato T, Iwanari H, et al. A simple detection method for low-affinity membrane protein interactions by baculoviral display. PLoS One 2008; 3:e4024.
- Saitoh R, Ohtomo T, Yamada Y, et al. Viral envelope protein gp64 transgenic mouse facilitates the generation of monoclonal antibodies against exogenous membrane proteins displayed on baculovirus. J Immunol Methods 2007; 322:104-17.
- Sievers EL, Senter PD. Antibody-drug conjugates in cancer therapy. 2013; 64:15-29.
- Larson SM, Carrasquillo JA, Cheung N-K V, et al. Radioimmunotherapy of human tumors. Nature Rev Cancer 2015; 15:347-70.
- Wu AM. Engineered antibodies for molecular imaging of cancer. Methods 2014; 65:139-47.
- Nunez-Prado N, Compte M, Harwood S, et al. The coming of age of engineered multivalent antibodies. Drug Discovery Today 2015; 20:588-94.
- Ito H, Funahashi S, Yamauchi N, et al. Identification of ROBO1 as a novel hepatocellular carcinoma antigen and a potential therapeutic and diagnostic target. Clin Cancer Res 2006; 12:3257-64.
- Fujiwara K, Koyama K, Suga K, et al. A 90Y-labelled anti-ROBO1 monoclonal antibody exhibits antitumour activity against hepatocellular carcinoma xenografts during ROBO1-targeted radioimmunotherapy. EJNMMI Res 2014; 4:29.
- Fujiwara K, Koyama K, Suga K, et al. 90Y-labelled anti-ROBO1 monoclonal antibody exhibits antitumor activity against small cell lung cancer xenografts. PLoS One 2015 10:e0125468.