Evaluation of Tumor Associated Antigens to Optically Label Cutaneous Basal Cell Carcinoma for Surgical Excision
Project ended 30 June 2021
Prof Stefan Barth
- Institute for Infectious Disease and Molecular Medicine, University of Cape Town
Title of the project
Evaluation of tumour-associated antigens to optically label cutaneous basal cell carcinoma for surgical excision.
Basal cell carcinoma is the most common skin cancer worldwide with the second-highest incidence rates in South Africa after Australia. Most cancers develop on highly exposed skin including head as well as neck and are the most suspicious and cosmetically challenging. Mohs surgery (invented in 1930 by Dr Frederic Mohs at the University of Wisconsin) has been accepted as the gold standard to remove BCC. During this micrographic surgery, thin layers of cancer-containing skin are removed under local anaesthetic. Frozen tissue sections are readily analysed by a pathologist in an on-site laboratory. If the remaining cancer cells are detected, the surgeon will remove another layer of tissue. This is a time-consuming, expensive procedure that takes generally 3-4 or more hours if several rounds of excisions are needed. Current total costs for one Mohs surgery treatment are in the range of R45 000 per treatment. The main treatment modality for BCC is tumor excision. The drawback of excision is that the boundary of the tumor margin is not well defined and the tumor can be extended beneath the superficial layer of skin. In consequence, the tumor is not visible when viewed from the top using dermatoscopy. Thus, identifying BCC cells, both in the superficial layer and beneath the layer of an individual patient using a unique antibody format called SNAP-tag technology will determine the success of surgical excision of BCC.
In order to reduce the cost of treatment and allow for a more specific treatment approach, we aim to design, engineer and test a range of SNAP-tag-based antibody fusion proteins to specifically bind and detect BCC cell surface receptors. The SNAP-tag antibody format is based on the genetic fusion of a disease-specific ligand to a protein tag derived from O6-alkylguanine-DNA alkyltransferase allowing covalent auto-labelling of the corresponding antibody-based fusion proteins with benzylguanine-modified (BG) substrates (e.g. fluorophores) under physiological conditions with high efficiency in a 1:1 stoichiometry. The best performing SNAP-tag-based diagnostic antibodies resulting from these studies will be further evaluated in mouse models aiming to reduce the time needed for surgical removal of BCC.
How the project was of value in the struggle against cancer
This project was allowing us to start a project on one of the major cancers with the highest incidences in South Africa. We were able to identify differentially expressed cells surface antigens for selective targeting of BCC. Using SNAP-tag technology, we have successfully generated the first set of recombinant antibody fusion proteins which can easily be chemically modified with any BG-modified substrate in a 1:1 stoichiometry and demonstrated cell-specific binding of fluorophore-conjugated constructs.
Although the pandemic was not allowing us to properly validate the expression of selected targets on FFPE sections of South African BCC patients, we were lucky to have a funded SA MRC project allowing us to establish immunofluorescence imaging as an antibody-based patient stratification platform in precision medicine: Fluorophoreconjugated SNAP tag-based antibody fusion proteins were used for the detection of their corresponding differentially expressed cell surface receptors on FFPE tissue sections of SA triple-negative breast cancer patients that were subjected to standard deparaffinization and antigen retrieval steps. We acquired images from 5 regions of each tissue slide that corresponded to unlabelled tissue (for establishing an autofluorescence baseline), labelled non-tumour tissue, and labelled tumour tissue. Image acquisition was repeated for 5 tissue sections from each patient. Binding data were acquired for each of the 5 biomarkers that were tested. Mean signal intensities from confocal imaging of patient biopsies were pooled (n= 10-18 patients). Our analysis showed significant differences in the expression of ASPH (n=18, p< 0.001), CD44 (n= 12, p= 0.0117), CSPG4 (n= 10, p= 0.0027), EGFR (n= 10, p= 0.0123) and EpCAM (n=11, p= 0.0003).