The main focus of our recently published paper (Szenes, E., Härzschel, A. et al. Leukemia, 2020) is on VLA-4, an integrin heterodimer comprised of an alpha (CD49d) and beta (CD29) subunit.VLA-4 plays a major role in homing and retention of leukemic and other cells in the tissue, such as the bone marrow and the secondary lymphoid organs. Malignant cells in chronic lymphocytic leukemia (CLL) are quiescent in the circulation; they need signals from the microenvironment to survive and proliferate. One mechanism of action of ibrutinib, the Bruton’s tyrosine kinase (BTK) inhibitor that has greatly improved treatment options and outcome for many CLL patients, is to block the interaction of the leukemic cells with tissue environments, flushing them out into the blood.
A previously proposed mechanism of this re-localization of CLL cells into the blood is inhibition of the interaction between the BCR and VLA-4, with BCR signals activating the high-affinity, adhesive states of VLA-4 via a process called inside-out signaling. Therefore, this is a story of cooperation: on one hand, between the VLA-4 integrin and the BCR, mediating the interaction between CLL cells and their microenvironment; and on the other hand, between researchers: the extensive network of the ICA graduate school (funded by the Austrian Science Fund), the research groups of Tanja Hartmann (formerly Salzburg, Austria, now Freiburg, Germany) and Valter Gattei (Aviano, Italy), as well as between the two first authors (Eva Szenes-Nagy and Andrea Härzschel).
Previously, we have observed that CLL patients expressing high levels of CD49d have inferior response to ibrutinib (Tissino et al., J Exp Med 2018). As technology is not advanced enough to provide complex cell culture systems mimicking whole organisms, we still depend on animal models to test new treatments, especially when considering the complex tumor-microenvironment interactions. Eµ-TCL1-transgenic (TCL1-tg) mice represent a widely used mouse model for high-risk CLL, and we aimed to investigate whether they resemble the CD49d-high CLL cohort.
First, we measured CD49d expression of leukemic cells from TCL1-tg mice and found that it resembles that of CD49d-high CLL cells. Next, we analyzed BCR-induced VLA-4 activation in TCL1-tg mice. In human CLL, measuring VLA-4 activation by the conformation-sensitive anti-CD29 antibody (HUTS-21) is easily achievable, however, no such antibody exists for murine cells. Therefore, we used a flow cytometric method, adopted from Alexandre Chigaev, Larry Sklar and their team (Albuquerque, US), using the small molecule VLA-4 ligand LDV. We found that BCR activation increases the affinity of VLA-4 to its ligand VCAM-1 on murine leukemic cells, and we also showed that this could not be fully inhibited by ibrutinib.
Last, we investigated the in vivo role of VLA-4 in tumor development and distribution. We performed two sets of experiments, using the small molecule CD49d inhibitor firategrast, and the rat-anti-mouse-CD49d antibody PS/2. Although the two studies were performed in different laboratories (Salzburg and Freiburg, respectively), the results align perfectly. Spleen tumor load immediately decreased upon treatment, while bone marrow infiltration responded at a slightly later time point. Follicles in the white pulp of the spleen may play a more important role in the early development of this murine leukemia, while the bone marrow microenvironment has a greater role at late stages.
In summary, the TCL1-tg model represents the VLA-4 expressing subgroup of human CLL, including the main molecular mechanisms of VLA-4 inside-out activation. We confirmed that in this mouse model, like in humans, BTK inhibition does not completely prevent BCR-mediated VLA-4 activation. We hope that these findings will help to design animal trials leading to an improvement of outcome for VLA-4 positive CLL patients. We hope that our research contributes to avoiding the phenomenon that has too often been observed and criticized in the past, that we cure mice, but not men.