Underlying every successful clinical application of fluorescent probes is a rigorous, strategic probe validation process. Previous blog posts have discussed the importance of probe specificity, binding affinity, and distribution, and the validation process is where these and other parameters are determined. Given the time and expenses involved in clinical translation, efficient and accurate probe validation is essential.
A Systematic Approach to Developing and Validating Optical Imaging Contrast Agents
In a foundational 2007 article published in Analytical Biochemistry, Kovar et.al. demonstrated the principal steps involved in developing fluorescent optical probes suitable for human clinical use. The authors began with a comprehensive review of NIR fluorochromes, such as IRDye® 800CW, and targets and ligands for fluorescent optical probes, including monoclonal antibodies, tumor surface proteins, peptides, and small molecules. Steps for development and validation described in the remainder of the article are “applicable to any dye-conjugated optical agent,” demonstrating the versatility of this systematic approach [1].
Step 1: Conjugation
The first step in probe development is the conjugation of target and NIR fluorochrome. In this study, the authors conjugated IRDye 800CW to five commercial epidermal growth factor (EGF) sources in equivalent ratios and evaluated signal intensity via the In-Cell Western™ (ICW) assay method. ICW imaging demonstrated variation between the signal strength of each EGF source. Because “Variations in signal strength measured in this fashion have the potential to predict probe performance in vivo,” choosing the correct target for conjugation is a critical first step [1].
Step 2: Specificity and Binding Affinity Validation in vitro
Prior to animal imaging, probe specificity and binding affinity are validated in vitro. Here, the authors again chose the In-Cell Western (cytoblot) method to evaluate IRDye 800CW EGF for binding specificity. Specifically, PC3M-LN4 and 22Rv1 human prostate adenocarcinoma cells were cultured in microtiter plates and were “treated with serial dilutions of labeled EGF to verify a high affinity binding of EGFR-targeted dye” [1]. Specificity was then determined by blocking access of EGF to the EGF receptor with an anti-EGFR monoclonal antibody, and by competition with unlabeled EGF [1]. The authors concluded that “Characterization of the targeting agent in a cell-based assay can simplify probe development,” and that although “success in a cell-based assay format does not guarantee performance in vivo, failure at this step is generally predictive of failure in the animal” [1].
Learn more about the In-Cell Western™ Method.
Step 3: Specificity, Distribution, and Clearance Validation in vivo
Next, the authors validated probe specificity and clearance in vivo in living mice. This is the process of determining probe uptake by the target vs surrounding tissues, the rate at which the probe is expelled from the organism, and the probe to background ratio in the body of mice. First, clearance kinetics of unconjugated IRDye 800CW were established. Then, clearance measurements for IRDye 800CW-anti-EGFR antibody conjugates were established in mice bearing PC3M-LN4 subcutaneous or orthotopic tumors to ensure the conjugate did not accumulate non-specifically in the mouse. This interaction between clearance kinetics and specificity can impact in vivo analysis by falsely indicating tumor tissue in pooled optical probe in the liver or kidneys.
Step 4: Ex vivo Validation in Excised Tissue Samples
Lastly, the PC3M-LN4 tumors were excised and injected intravenously with IRDye 800CW EGF or pre-injected with C225 anti-EGFR monoclonal antibody prior to dosing with IRDye 800CW EGF for ex vivo analysis. After imaging, the distribution of the IRDye probe was assessed and fluorescence signal area was determined against a control, optical agent only, and C225 competition. The authors ultimately concluded that “Fluorochrome-labeled molecular probes are valuable tools for non-invasive longitudinal study of tumorigenesis and metastasis, preclinical studies of the effects of therapeutic agents, and pharmacokinetic and pharmacodynamic studies of drug-target interactions” [1]. Since this paper was published over a decade ago, IRDye labeled molecular probes have been featured in more than 20 clinical trials around the world.
EGFR-Specific Optical Probes Improve EGFRvIII-Targeted Molecular Imaging
In a 2014 study published in Cancer Biology & Therapy, Gong et.al. demonstrated an application of structured probe validation. In this investigative study, the authors created and validated the specificity, binding affinity, distribution, and clearance of three EFGRvIII-targeted fluorescent optical probes. An EFGR-specific affibody, the therapeutic antibody panitumumab, and an EGF ligand were conjugated with IRDye 800CW to create three probes: Aff800, Pan800, and EGF800. The experimental target was rat glioma cell line F98, a known over-expresser of EGFR. A control assay contained EGFR expression-devoid F98 parent (F98-p) cells, and two experimental assays contained F98-derived transgenic cells expressing EGFR or EGFR-vIII.
Each probe was compared with each cell-based assay and imaged for comparison, creating a total of nine experimental conditions. Comparison of specificity and binding affinity between the experimental conditions was performed in cell-based assays using the In-Cell Western assay method. All three probes successfully bound to F98-EGFR, and Pan800 and Aff800 bound to F-98vIII. Signal intensity was also compared in the nine conditions to assess if binding was dose-dependent. The authors concluded “Little signal was detected when Aff800 and Pan800 were incubated with F98-p [the expression-devoid parent cells], indicating that their interactions with F98-EGFR and F98-vIII is highly specific” [2].
Next, probe target specificity to EGFR- and EGFRvIII-expressing tumors and clearance profiles were assessed in vivo. Mice with F98-p, F98-EGFR, and F98-vIII xenograft tumors were injected with the three probes and imaged with the Pearl® Impulse Small Animal Imaging System (LI-COR Biosciences). Fluorescent signal to background ratio for each of the nine probe-tumor conditions were assessed, again revealing highly specific interactions between Aff800 and Pan800 with F98-EGFR and F98-vIII expressing tumors. EGF-800 signal was high in F98-EGFR tumors, corroborating cell based assay results.
Lastly, tumor-containing organs were dissected and imaged ex vivo, validating the previously-measured fluorescence signals and assessing probe distribution in targets. This last step in validation was consistent with in vitro scans, again demonstrating Aff800 and Pan800 affinity to F98-EGFR and F98-vIII tumors. Based on these results, Aff800 and Pan800 may be valuable in “imaging of heterogenous tumors containing both versions of receptors” (EGFR, EGFRvIII) [2]. Alternatively, due to optimal clearance kinetics, Aff800 EGF800 is preferable in scenarios where imaging must be performed within a short time after probe administration” [2].
Conclusion
This example from Gong et.al. demonstrates how different optical probes may be used to assess different tumor properties. Additionally, the authors showed how structured approach to optical probe validation successively builds proof of probe parameters and provides several spots for go/no-go decision-making. Proof of probe parameters are critical for clinical application, and clear decision points provide efficiency and allow for early determination if a probe is worth exploring further.
References:
- Kovar, J. L., Simpson, M. A., Geschwender, A., & Olive, D. M. (2007, August 1). A Systematic Approach to the Development of Fluorescent Contrast Agents for Optical Imaging of Mouse Cancer Models. Analytical Biochemistry, 367(1), 1-12. doi:10.1016/j.ab.2007.04.011
- Gong, H., Kovar, J. L., Cheung, L., Rosenthal, E. L., & Olive, D. M. (2014, February). A Comparative Study of Affibody, Panitumumab, and EGF for Near-Infrared Fluorescence Imaging of EGFR- and EGFRvIII-expressing Tumors. Cancer Biology & Therapy, 15(2), 185-193. doi:10.4161/cbt.26719
Powered by Froala Editor