Abstract
1146
Objectives Increased activity of matrix metalloproteases (MMPs) -2 and -9 is associated with poor prognosis and metastasis processes in cancer tissues. Aiming to produce a MMP tumour cell-selective cytotoxin, Leppla et al. genetically modified the two protein components of the AB-type lethal toxin (LT) from Bacillus anthracis. The celullar binding component of this toxin, Protective Antigen (PA), was designed to form an oligomeric pore in cell membranes only when cleaved by MMPs (PA-L1). The second part, Lethal Factor (LF), which binds to the PA-pore, had its N-terminal translocation domain (LFn) fused to exotoxin-A domain (fusion toxin - FP59) to increase its cytotoxic effect when delivered to cancerous cells. The current work describes the development of MMP imaging using PA-L1 and LFnCys (LFn containing a C-terminal cysteine residue) radiolabelled with In-111.
Methods Using the MTT assay we evaluated PA-L1 MMP-mediated activation and translocation of FP59 toxin into a set of cancer cell lines. Gelatinase (MMP-2 and -9) zymograms were used to confirm these results. Since LFnCys is known to be thermolabile, CD analysis reproducing standard radiolabelling conditions were used to evaluate secondary structure changes. The metal ion chelators DTPA-maleimide and DOTA-GA-maleimide were conjugated to LFnCys to allow radiolabelling with In-111. Since In-111 incorporation into DOTA-GA requires heating, a strategy was developed where In-111 was chelated to DOTA-GA-maleimide first and then conjugated to LFnCys under mild conditions. The MMP-positive cell lines HT-1080 (gelatinase positive) and MCF7 (negative) were used to compare internalisation of In-111-DOTA-GA-LFn with In-111-DTPA-LFn in the presence of PA-L1. For cross-validation by confocal microscopy, LFnCys was site-specifically conjugated to AlexaFluor488-maleimide.
Results MTT tests showed that MCF7 cells did not activate PA-L1 pore formation, while B8484, CAPAN-1, MDA-MB-231 and HT-1080 cells were sensitive to FP59, suggesting activation by MMPs. Zymography gels demonstrated varying levels of MMP-2 and -9 activity in the different cells lines, correlating well with the MTT assay results. While buffer composition did not affect the secondary structure of LFnCys, temperature increases over 95°C for 15 to 60 minutes significantly disturbed its overall secondary structure. Radiolabelling yields for 111In-DTPA-LFn was 98.9%, while 92.4% was obtained for In-111-DOTA-GA-LFn using the pre-labelling strategy. Uptake of both radiolabelled forms of LFnCys in gelatinase-positive HT-1080 cells was significantly higher in the presence of PA-L1 when compared to cells not exposed to PA-L1 or to MCF-7 cells. These results were corroborated by confocal imaging using fluorescently labelled LFnCys.
Conclusions Our results indicate that radiolabelled forms of genetically engineered anthrax lethal toxin could be developed as a promising MMP activity imaging agent for non-invasive imaging of cancer.