Vascular antibody-mediated pharmaceutically induced tumour resection

Final Report Summary - VAMPIRE (Vascular antibody-mediated pharmaceutically induced tumour resection)

The VAMPIRE consortium was created to advance cancer therapies in the field of tumour vascular targeting. Animal models in the 1990's had shown that when you damage the tumour vasculature this brings about tumour collapse. Progress since then had been slow due to the inability to identify suitable targets on the vessels in human tumours. In recent years rapid advances in genomics and obtaining isolates of pure cell populations from tumours had led us to the identification of several potential tumour vascular targets. These studies had been carried out on RNA isolates (gene array analysis or deep sequencing or RNAseq) as these technologies are far in advance of those used to analyse proteins. Consequently, the next step in the analysis was to examine the corresponding proteins for their expression in tumours and healthy tissues. To perform this, there is an absolute need for either anti-sera or monoclonal antibodies. To progress this, VAMPIRE had two main aims. The first was to generate and preliminary characterise novel antibodies to the new targets. The second was to use these antibodies to determine where they homed to when delivered in vivo to tumour bearing mice, the so-called bio-distribution. Each of these aims constituted one of the scientific work-packages of the consortium. SomantiX is a company in the Netherlands whose purpose is to advance tumour vascular targeted therapies. The VAMPIRE EID was comprised of the University of Birmingham and SomantiX BLV.
The targets to be studied had all been identified within the academic environment of the University of Birmingham. The lead target was the C-type lectin CLEC14A. However, more recent studies had identified other potential targets including MCAM in renal carcinoma and several novel lung targets including protocadherin 7 (PCDH7) an STEAP1 (six transmembrane epithelial antigen of the prostate 1). The first task (WP1) was to generate monoclonal antibodies to the targets. The novelty of the targets at the time meant that antibodies were either not available (CLEC14A, PCDH7) or either prohibitively expensive for our needs (MACM) or unavailable to us (STEAP1). Another critical consideration was that we required not only antibodies that would recognise mouse protein to enable validation in mouse cancer models but also that they would recognise the native folded protein. To obtain mouse antibodies to a mouse protein requires breaking tolerance to self, long known to be a difficult and unpredictable task. Nevertheless in on-going studies we had found that we could elicit an immune response to mouse CLEC14A by using a protein conjugate of mouse extracellular CLEC14A fused to human Fc delivered with a strong adjuvant (Freund's complete adjuvant). In this immunisation the mouse sees this entire conjugate as foreign (because of the human part of the conjugate) and in consequence the T helper cells initiated to the human part assist the development of the B cells producing antibodies to the mouse protein. An immunisation and fusion was carried out and generated five mouse monoclonal antibodies that recognised mouse CLEC14A.
A major recent advance in cancer immunotherapy has been the development of CAR T cell modified therapies. The T cell receptor is genetically modified by insertion of antibody variable regions and then retrovirally transduced into a patients T cells which when re-introduced to the patient kill cells expressing the antibody target protein. With Dr Steve Lee (Birmingham) we obtained a Medical Research Council Confidence in Concept grant (£100 k) to explore the possibility of using our antibodies to develop a CLEC14A targeted CAR T cell therapy. The results were very promising and formed the basis for a £1.5 million grant from The Cell Therapy Catapult (press announcement). The agreement with The Cell Therapy Catapult has led to the creation of a new spin off company called Chimeric Therapeutics Ltd that is equally owned by University of Birmingham, The Cell Therapy Catapult and Cancer Research Technology.
The identification of MMRN2 as the ligand for CLEC14A and the fact that CLEC14A continued to show the most perfect expression profile to date of any tumour vascular target prompted us to examine MMRN2 further as an alternative vascular target but lying within the CLEC14A axis. Unlike CLEC14A, MMRN2 is an exceptionally large (>200 kDa) extracellular protein. Were this to be used as an immunogen it would generate many monoclonal antibodies but identifying those of functional significance could be a major effort. Alternatively we could identify the active regions that constitute the binding site to CLEC14A and then use these as immunogen and have a greatly improved chance of obtaining biologically active antibodies. To enable this we used molecular biology (protein truncation and Far Western blotting) to identify the small region of MMRN2 that binds to CLEC14A.
The second target where we have made exciting progress is MCAM. We had previously shown that MCAM is extremely highly expressed on the vessels of renal cancer. We showed expression to be induced by VEGF that is found at exceptionally high levels in renal cancer due to mutation in the vhl gene. Using commercial anti-sera we had shown that antibodies selectively localise to vessels in renal carcinoma but are absent from vessels in healthy tissue. To proceed further in terms of mono clonal antibodies or CAR modified T cell therapy access to our own hybridomas was essential. Unfortunately the generation of monoclonal antibodies remains an unpredictable art. The same approach was taken that had been so successful with CLEC14A, thus, mouse extracellular MCAM was cloned and fused to human Fc. In this case, immunisation of mice was poor and attempted fusions yielded no antibody producing hybridomas that recognised the mouse MCAM. To overcome these problems it was decided to immunise rats with the mouse protein. This led to the successful generation of two new anti-mouse rat monoclonals and the antibodies characterised. The anti-MCAM antibodies showed anti-angiogenic activity in vitro and, most importantly, as with the commercial polyclonal anti-sera, were shown to home exclusively to the renal carcinoma tumour vessels and not to vessels in healthy tissue. The extracellular domain of MCAM was also studied in the in vitro angiogenesis assays and shown to have anti-angiogenic activity. The generation of our own hybridomas expressing anti-MCAM monoclonals had enabled us to secure a second Medical Research Council Confidence in Concept grant in this case to develop CAR modified T cells targeting MCAM (with Drs Steve Lee and Victoria Heath, University of Birmingham, £80 k).
For the last two targets studied, namely PCDH7 and STEAP1, attempts to raise monoclonal anti bodies proved more problematic. In both cases mouse and human chimeric constructs with human Fc were constructed and protein expression studied. With PCDH7 the human protein expressed well but all attempts to express the mouse protein failed. Despite a substantial effort exploring many different expression systems we were never able to obtain mouse protein. The reason for this is not known. In the absence of the mouse protein we decided to attempt to ascertain whether PCDH7 played a functional role in the endothelial cell and angiogenesis. We showed that the extracellular domain of human PCDH7 abrogated angiogenesis in 2D and 3D models of vessel formation. Deletional analysis allowed identification of the functional fold of the protein in this anti-angiogenic effect. These studies have laid the foundation for future studies. Thus, it is clear which part of PCDH7 is biologically active in angiogenesis. It is possible to chemically synthesise this peptide region of the mouse protein. The question then is whether it can be re-folded so that antibodies can be obtained that recognis the native protein.
Future work. STEAP1 also proved difficult to work with. STEAP1 is a small protein but with six intracellualar domains. Indeed a significant proportion of the protein is buried within the plasma membrane. While we could express mouse STEAP1 with and Fc tag and purify this using detergents we were orders of magnitude away from obtaining sufficient protein to be able to effectively immunise rodents.
In terms of dissemination and future work, eight papers of relevance to VAMPIRE have been published to date and at least four more are in preparation and soon to be submitted. Our work on CLEC14 has led to the creation of the new company Chimeric Therapeutics Ltd to develop our CLEC14A antibody variable regions as a CAR modified T cell therapy. To this end, and also to develop the MCAM targeting antibodies developed entirely within VAMPIRE, we have secured £1.5 million of funding from The Cell Therapy Catapult and almost £200 k funding from the Medical Research Council of the United Kingdom. Other research fellows funded by various sources also continue to progress the project.

More information about the VAMPIRE project can be obtained by emailing the Co-ordinator Roy Bicknell at r.bicknell@bham.ac.uk. The VAMPIRE website can be found at https://www.birmingham.ac.uk/generic/vampire/about/index.aspx

Final Report Summary - VAMPIRE (Vascular antibody-mediated pharmaceutically induced tumour resection)

Diese Seite teilen

Herunterladen