Group Members: Jacqueline Testa, Philippe Valadon, Han Nguyen
Mass spectrometry-based methods have yielded a wealth of information describing the proteins present at the lung endothelial cell surface, but we simply cannot map every tissue of the body and every disease state using mass spectrometry-based analysis of isolated endothelial cell membranes. Antibody probes provide essential and complementary proteomic analysis because they can be used to validate proteins identified through mass spectrometry and, perhaps more importantly, to assess target expression in a much broader range of tissues. They also provide key additional information about where antigens are expressed, not only at the tissue level, but also at the cellular and subcellular level, as we’ve done in the past. Antibody-based proteomic analysis requires generation of novel, high affinity probes. We have developed multiple methods to produce antibodies against known proteins identified through mass spectrometry-based methods. Additonally, antibodies against unknown proteins can be generated by directly injecting target tissue, cells, or even tissue subfractions as an immunogen. Monoclonal antibodies can then be used to screen expression in cells and tissue as well as to immunoprecipitate the protein of interest for sequencing and identification. The goals of the Antibody Development and Engineering Facility are to extend knowledge of the topological expression of proteins on vascular endothelial cells, to create high-affinity probes that can target the vascular endothelium, and to increase the speed at which antibodies can be tested and incorporated into larger complexes such as nanoparticles.
The Antibody Development and Engineering Facility currently consists of a Beckman J2 High Speed Centrigure, a PCR machine, an AKTA FPLC 900, a Alpha Inotek gel documentation system, a UV-VIS Spectrophotometer from Beckman Coulter, an HPLC, multiple gel electrophoresis boxes, tissue culture hoods and incubators, and a VersaMax tunable plate reader. This facility also has full access to the complete PRISM lab space, which includes cold rooms, radioactive labeling preparation and analysis rooms, communal instrument rooms, scintillation counters, refrigerators, chemistry hoods ultracentrifuges and key rotors, ultra-low freezers, a MilliQ water purification system, and general-use computers.
Generation of caveolae- and tissue-specific monoclonal antibodies:
To assess the existence of tissue-specific targets on the endothelium and its caveolae, we have produced monoclonal antibodies using isolated rat lung endothelial cell membrane as the immunogen in mice. Over 100 hybridomas have been raised that recognize rat lung endothelial cell membrane by ELISA. Mice are also immunized with recombinant proteins identified through mass spectrometry-based methods. Candidate monoclonal antibodies are further screened with Western analysis to ensure that they recognize a single protein, are only found in a single organ, and are enriched at the luminal surface of endothelial cells compared to the entire tissue homogenate. Further analysis can reveal whether these antibodies target proteins enriched in caveolae. We have used these methods to create an array of lung-specific proteins that target proteins on the luminal surface of ECs both within caveolae and outside of caveolae.
Phage display libraries are another way to discover novel targets and generate affinity probes. George Smith first suggested that bacteriophages could be used to display polypeptide fragments in 1985. Because the approach can use a large, random library of peptides, it is unbiased and unknown targets can be identified. In a process called panning, phages that bind to a target (antibody, protein, or even tissue) are repeatedly isolated and selectively amplified. Phage can be purified and used as probes themselves to isolate the binding partner for identification. Phage can also be injected intravenously, to circulate and presumably to bind a single protein at the endothelial surface. They can then be isolated from each organ or tissue of interest. However, the liver and spleen rapidly scavenge phage from the blood, mostly before they have a chance to circulate through each organ and bind sufficiently. Additionally, short peptides can lack specificity and may bind a large range of proteins in a multitude of organs, requiring additional ex vivo validation. These problems can be partially overcome by avoiding direct in vivo panning but rather screen antibody phage libraries on key membranes and then creating antibody-like fusion proteins, which unlike the phage, are not rapidly removed from the circulating blood and can indeed immunotarget successfully in vivo. Phage display libraries have revealed some promising targets, but currently this method may not yet be optimally suited for the high-throughput needed to comprehensively map protein expression and identify tissue-specific proteins.
We built a library containing 3x108 full-length, in-frame, recombinant clones from the spleen cell mRNA of a mouse immunized with rat lung luminal EC plasma membranes. To test the ability to generate binders to lung EC surface proteins, we pre-adsorbed this library on EC plasma membranes isolated from liver to subtract out phage that bound both non-specifically and phage that bound to liver proteins. We then screened the library against EC membranes from lung three times, each time isolated the phage that bound tightly to membrane and amplifying them. Ultimately, this yielded several promising targets that specifically bound lung endothelial membranes with high affinity.
This work is a fundamental advance that allows high throughput screening of a random antibody library. Given the small size of the repertoire analyzed in classic hybridoma technology, the greater immune repertoire incorporated into phage libraries may be of considerable benefit in yielding a greater pool of targeting probes. Additionally, repeated panning against endothelial membranes (to isolate antibodies that bind to unknown antigens) or against recombinant proteins (to isolate antibodies that bind to known antigens) allows rapid and rigorous purification.
We have designed a cloning and expression platform using the Invitrogen Gateway system for converting polypeptide sequences identified by mass spectrometry and potentially of high interest into readily expressed proteins that can be used to immunize mice and generate mAbs. This platform bypasses the use of purified endothelial membranes and offers a way to translate AVATAR data into tissue-specific targeting and effective delivery of imaging or therapeutic agents. Selected sequences from the AVATAR database are analyzed in silico and accessible, folded protein domains are identified. The corresponding cDNAs are amplified by PCR from commercially available sources such as the NIH Image clone collection. Entry clones can be introduced by enzymatic DNA recombination into varied expression acceptor plasmids, either for bacterial expression or transient expression in 293 cells in suspension culture. We usually add a biotin-acceptor tag to generate biotinylated proteins that are easy to detect and quantify by Western analysis together with a HIS tag for purification. Our bacterial expression vector has HIS tags at both the N- and C-termini, whereas our eukaryotic expression vector has both HIS tags at the C-terminus. The mammalian expression system has proven efficient in generating noticeable amounts of folded, soluble proteins that are quite easy to purify from the culture supernatant using an IMAC column and our FPLC system (a GE Healthcare AKTA purifier). The biotin tag is also helpful for ELISA after capture on streptavidin-coated plates; we also plan to use it for capturing the proteins on streptavidin beads that would serve to screen large naïve antibody libraries and automate the process of antibody generation. Currently we have developed the acceptor vector for both expression systems and have more than 12 proteins in the pipeline and five ongoing immunizations.
Nanostreptabodies:
We have recently created a multifunctional antibody targeting system that we call nanostreptabodies. A manuscript describing this process is under review. In this system, the DNA for the antibodies is cloned so that recombinant antibodies can be created. A 15-amino acid biotin-acceptor sequence is incorporated into full length antibodies or into Fab’ antibodies, allowing for efficient and specific enzymatic biotinylation in vivo to create bisbiotinylated antibodies and monobiotinylated Fab’ fragments. We have created biotinylated Fab’ fragments of several lung-specific antibodies, some of which target caveolae and some of which bind outside of caveolae. To assemble nanostreptabodies, streptavidin is first mixed with biotinylated antibodies. Because streptavidin can bind four biotin residues, a diverse array of complexes can be formed by attaching other moieties to the streptavidin. Adding a second type of antibody produces a bivalent antibody; adding an imaging and/or drug components produces multifunctional complexes. When injected intravenously, these auto-assembled complexes could target the lung in vivo and were rapidly transcytosed across the endothelial cell barrier. Targeting antibody and biotinylated nanoparticles, probes, imaging agents, viruses and even biotinylated siRNAs can be readily combined to provide a flexible, rapid, and efficient way to assemble antibody-based targeting platforms.
