Peptide-Based Optical uPAR Imaging for Surgery: In Vivo Testing of ICG-Glu-Glu-AE105

Abstract

Near infrared intra-operative optical imaging is an emerging technique with clear implications for improved cancer surgery by enabling a more distinct delineation of the tumor margins during resection. This modality has the potential to increase the number of patients having a curative radical tumor resection. In the present study, a new uPAR-targeted fluorescent probe was developed and the in vivo applicability was evaluated in a human xenograft mouse model. Most human carcinomas express high level of uPAR in the tumor-stromal interface of invasive lesions and uPAR is therefore considered an ideal target for intra-operative imaging. Conjugation of the flourophor indocyanine green (ICG) to the uPAR agonist (AE105) provides an optical imaging ligand with sufficiently high receptor affinity to allow for a specific receptor targeting in vivo. For in vivo testing, human glioblastoma xenograft mice were subjected to optical imaging after i.v. injection of ICG-AE105, which provided an optimal contrast in the time window 6-24 h post injection. Specificity of the uPAR-targeting probe ICG-AE105 was demonstrated in vivo by 1) no uptake of unconjugated ICG after 15 hours, 2) inhibition of ICG-AE105 tumor uptake by a bolus injection of the natural uPAR ligand pro-uPA, and finally 3) the histological colocalization of ICG-AE105 fluorescence and immunohistochemical detected human uPAR on resected tumor slides. Taken together, our data supports the potential use of this probe for intra-operative optical guidance in cancer surgery to ensure complete removal of tumors while preserving adjacent, healthy tissue.

Fig 1. Structure and binding affinity for ICG-Glu-Glu-AE105.

(A) The chemical structure of ICG-Glu-Glu-AE105 (B) Assessing uPAR binding properties of ICG-Glu-Glu-AE105 by an indirect solution competition for uPA binding by surface plasmon resonance yielding an IC₅₀ value of 134 nM. (C) Absorption spectra of ICG (black, full line) and ICG-Glu-Glu-AE105 (red, full line) measured in HBC solution. Ecitation spectra of ICG (black, broken line) and ICG-Glu-Glu-AE105 (red, broken line) measured in HBC solution. Fluorescence spectra of ICG (blue) and ICG-Glu-Glu-AE105 (green) measured in HBC solution. The noise observed between 860–900 nm in the fluorescence spectra are due to poor detector correction of the instrument in this region.

Fig 2. In vitro expression of uPAR in the cell line U87MG and in vivo TBR values over time (1–72 h).

(A) In vitro flowcytometry confirms the presence of extracellular uPAR on the U87MG cell line with 98.6% positive for uPAR. (B) Dynamic optical imaging of mice with 10 nmol ICG-Glu-Glu-AE105 at the timepoints 1, 2, 4, 8, 12, 24, 48, 72 h. The graph show clear optimum between 6–24 h. Each graph represents one of two tumors per mouse.

Fig 3. Optical images of tumor bearing mice with ICG-Glu-Glu-AE105 or ICG.

(A) Representative optical images of U87MG tumor bearing mice from both groups 15 h post injection (ICG-Glu-Glu-AE105 and ICG) obtained with the IVIS Lumina XR. The images show clear difference in the fluorescent signal from the s.c. tumors with 3.52±0.17 and 1.04±0.04 respectively. The images are shown within the same scalebar to allow for direct comparison. (B) Image from the Fluobeam^®800 and the Fluobeam setup. A Fluobeam image of a representative mouse with s.c U87MG tumors. The difference in intensity of the two tumors is a result of size of the tumors and the optical properties of the camera. The representative image from the fluobeam camera shows similar signal as the black-box imager IVIS Lumina XR. This underlines the translational potential of ICG-Glu-Glu-AE105.

Fig 4. uPAR expression measured by optical signal and ELISA assay.

(A) The mean TBR value was significantly different (1.04±0.04 for ICG and 3.52±0.17 for ICG-Glu-Glu-AE105, p<0.0001), while the uPAR expression per mg tissue was almost identical and supports the hypothesis that the difference in ICG-Glu-Glu-AE105 and ICG signal is due to uPAR binding.

Fig 5. In vivo blocking of ICG-Glu-Glu-AE105 by uPA, the natural ligand.

(A) Representative images obtained by the IVIS Lumina XR at 710 nm showing a mouse receiving uPA simultaneously with ICG-Glu-Glu-AE105 resulting in decreased signal compared to a mouse only receiving ICG-Glu-Glu-AE105. (B) Two groups of mice (n = 4) were dynamically scanned with either ICG-Glu-Glu-AE105 + uPA or ICG-Glu-Glu-AE105. In all timepoints the two groups were significantly different with the group receiving only ICG-Glu-Glu-E105 having a 2 fold higher signal.

Fig 6. Ex Vivo histology and fluorescence images of s.c. U87MG tumor tissue sections.

Shown here is from the top H&E staining (A,E), uPAR staining by immunohistochemistry (B,F), uPAR staining by fluorescence from the injected ICG-Glu-Glu-AE105 (C,G) and merged uPAR IHC and fluorescence imaging (E,H). In the first row panel A shows the H&E staining of the tumor. uPAR IHC staining (B) illustrate two clear islands of uPAR positive cells which are also depicted with fluorescence imaging (C). Co-localization of uPAR expression and ICG-Glu-Glu-AE105 fluorescence is shown I the merged IHC and fluorescence image (D). Additionally in panel B the border between human xenograft tissue and mouse stroma is seen. In the second row another tumor speciment is shown wht H&E staining in panel E. The uPAR staining (F) show heterogeneous uPAR expression and the enlarged image show an island with cells expressing higher amount of uPAR. This is also depicted in panel G where the same island of cells can be located by fluorescence. The merged uPAR IHC and fluorescence image (H) show the co-localization of uPAR expression and fluorescent signal.