Unmasking Thrombo-inflammation by in vivo Molecular Magnetic Resonance Imaging of von Willebrand Factor – FlaMRIng

The consortium generated two nanobody libraries targeting murine and human VWF using the phage display technique. This method allows selection of phages expressing antibodies specific to an immobilized antigen. Selected phages were tested by ELISA to confirm binding and identify those with the highest affinity. The corresponding nanobodies were then sequenced, expressed in E. coli, and purified.

To circumvent the rapid formation of a protein corona around circulating particles, we developed a bispecific nanobody (KB-V6F6) targeting both fibrinogen and VWF. This nanobody was conjugated to MPIOs. We also studied other critical aspects of thromboinflammation, such as leukocyte adhesion to the vascular wall and endothelial cell activation, as indicated by VCAM-1 expression. To do this, we produced MPIOs covalently conjugated to a commercial anti-VCAM-1 monoclonal antibody and to an anti-CD45 antibody, to target activated endothelial cells and adherent leukocytes, respectively.

All in vivo experiments were conducted on male Swiss mice aged 6–12 weeks. We used murine models of thrombo-inflammation induced by intracerebral injection of LPS or histamine, as well as a subarachnoid hemorrhage (SAH) model generated by prechiasmatic injection of fresh arterial blood.

In vivo molecular MRI was performed using a Pharmascan 7-T/12 cm system with surface coils (Bruker, Germany). Brain images were acquired using T2*-weighted iron-sensitive MRI before and after intravenous injection of MPIOs.

To validate our results, we also performed RT-qPCR analyses of VCAM-1 and VWF gene expression in brain samples collected at various time points, immunofluorescence microscopy on brain sections stained for VCAM-1, VWF, and endothelial markers, and ELISA assays on plasma samples to quantify VWF levels under different experimental conditions.

Initial studies were conducted by conjugating these MPIOs to anti-VWF nanobodies and then evaluated in a thrombo-inflammation context induced by LPS. The results proved negative, as no specific signal was detected. To understand the underlying reasons, we performed comparative experiments with MPIOs coupled to a nanobody directed against VCAM-1, a target validated in our laboratory for inflammation imaging. We observed that MPIOs conjugated to nanobodies were no longer detectable, whereas MPIOs conjugated to full-length antibodies against VCAM-1 allowed visualization of inflammation. Additional competition experiments with antibodies provided supporting evidence for our hypothesis: due to the small size of nanobodies, the rapidly forming fibrinogen layer around MPIOs masked the nanobody binding site, preventing it from reaching its target.

We therefore explored alternative strategies to overcome this methodological limitation by developing a bispecific nanobody that binds independently to VWF and fibrinogen and is able to recognize both antigens simultaneously. In vivo experiments using this MPIOs conjugate enabled detection of a specific signal in a model of cerebral thrombo-inflammation induced by intracerebral histamine injection. However, the amount of bound MPIOs was insufficient to allow MRI imaging of VWF in more clinically relevant models.

Within the scope of this project, we therefore redirected our efforts toward other key players in the thrombo-inflammatory response, particularly adherent leukocytes and activated endothelial cells.

We first developed a method to track adherent leukocytes (CD45+ cells) using superparamagnetic particles phagocytosed in situ. The goal was to target leukocytes that utilize VWF to adhere to the vessel wall, thereby enabling indirect detection of VWF. In vivo results in a model of ischemic stroke suggest that this approach can detect leukocytes adhering to the vessel wall at the subacute stage of ischemic stroke. However, this method proved insufficiently sensitive to detect the thrombo-inflammatory response in experimental models with milder neurovascular lesions, such as subarachnoid hemorrhage.

Finally, we assessed the relevance of immuno-MRI targeting proteins stably exposed on the endothelial surface as markers of thrombo-inflammation. In the subarachnoid hemorrhage model, imaging targeting VCAM-1 showed a marked increase in its expression in the brain 24 hours after the event, followed by a gradual return to baseline levels after 7 days. The increase in VCAM-1 expression is a marker of endothelial activation. Accordingly, VWF brain expression showed a temporal profile similar to that of VCAM-1, with a peak at 24 hours followed by a return to normal levels after 7 days. This increase was also observed in blood samples, and complementary analyses confirmed substantial VWF release by endothelial cells of cerebral microvessels at this early stage.

The development of next-generation molecular imaging tools for thrombo-inflammation must bridge the gap between experimental success and clinical utility. Based on the findings of this project, the following perspectives outline the path toward biodegradable and clinically translatable imaging agents:

- Transitioning to Biodegradable Contrast Agents: A critical hurdle for the clinical translation of immuno-MRI is the long-term persistence of MPIO in the body. While highly effective for preclinical proof-of-concept, traditional MPIOs lack a clear pathway for degradation and excretion.

- M3P Development: Future efforts should prioritize the refinement of microsized matrix-based magnetic particles (M3P). These particles leverage a self-assembly mechanism of catechol-coated magnetite nanocrystals, potentially offering a more biocompatible and biodegradable framework than conventional MPIOs.

- Safety Profiles: Developing probes that can be naturally metabolized or cleared by the reticuloendothelial system without toxic accumulation is essential for regulatory approval and patient safety.

The "masking effect" caused by the rapid formation of a protein corona, specifically the adsorption of fibrinogen, remains a significant technical bottleneck for nanobody-based targeting. Technologies to overcome this limitation could be developed:

- Structural Optimization: While the bispecific KB-V6F6 construct proved that the protein corona can be used as a "molecular anchor," its current sensitivity is insufficient for reliable MRI of VWF.

- Enhanced Display: Future research must focus on optimizing how nanobodies are displayed on the particle surface to ensure they protrude beyond the protein layer. This may involve the use of advanced chemical spacers or multimeric constructs to increase binding avidity.

Lastly, the project established that imaging endothelial activation via VCAM-1 or tracking adherent leukocytes can successfully identify early brain injury in models like subarachnoid hemorrhage (SAH). Because VWF is released almost instantaneously from Weibel-Palade bodies upon injury, it remains the "gold standard" candidate for time-critical clinical imaging. Translating these methods would allow physicians to detect vascular injury before irreversible tissue damage occurs, enabling the early deployment of targeted anti-thrombo-inflammatory therapies.

• Texier, A., Lenting, P. J., Denis, C. V., Roullet, S., & Christophe, O. D. (2023). Angiopoietin-2 binds to multiple interactive sites within von Willebrand factor. Research and Practice in Thrombosis and Haemostasis, 7(7), 102204. doi.org/10.1016/j.rpth.2023.102204
• Lenting, P. J., Texier, A., & Casari, C. (2023). von Willebrand factor: from figurant to main character in the scene of inflammation. Journal of Thrombosis and Haemostasis (JTH), 21(4), 710–713. doi.org/10.1016/j.jtha.2023.01.014
• Martinez de Lizarrondo, S., Jacqmarcq, C., Naveau, M., Navarro-Oviedo, M., Pedron, S., Adam, A., Freis, B., Allouche, S., Goux, D., Razafindrakoto, S., Gazeau, F., Mertz, D., Vivien, D., Bonnard, T., & Gauberti, M. (2022). Tracking the immune response by MRI using biodegradable and ultrasensitive microprobes. Science Advances, 8(28), eabm3596. doi.org/10.1126/sciadv.abm3596
• Texier, A. “La cellule endothéliale au croisement de l’angiogenèse, la transplantation hépatique et l’imagerie de la neuroinflammation.” PhD thesis in Physiology and Pathophysiology, supervised by Peter Lenting and co-supervised by Stéphanie Roullet. Université Paris-Saclay, 2024.

Thrombo-inflammation has emerged as a key pathophysiologic player in a large number of disorders. It is caused by the activation of endothelial cells in injured tissues, leading to secondary microvascular thrombosis and further tissue damage. To date, available imaging methods only allow to detect the consequences of thrombo-inflammation, such as tissue hypoperfusion and necrosis, preventing prophylactic treatment. Thus, the ability to detect the 1st steps of the thrombo-inflammatory reaction using non-invasive imaging could not only improve our pathophysiological understanding of thrombo-inflammatory disorders but also provide a window for patient therapeutic intervention before irreversible injury occurs. In this project, we propose to develop an original method to image thrombo-inflammation using molecular magnetic resonance imaging (MRI). This method will be characterized in thrombo-inflammatory disorders such as stroke, subarachnoid hemorrhage and systemic inflammatory disorders.