Novel approaches to assess cellular interactions and their role in the pathology and treatment of lymphoproliferative disorders

Background: Migration and homing are essential to B-lymphocyte mediated immunity, and are driven by rapid, directed, and appropriate reorganisation of the actin cytoskeleton. Important observations have linked the cytoskeletal-rearrangements made by leukaemic B-lymphocytes of chronic lymphocytic leukaemia (CLL) to disease pathology. In particular, cytoskeletal alterations mediated by B-cell receptor (BCR) engagement or chemokine-binding are recognised to be central to the migration of CLL cells to lymphoid tissues, where they engage in the complex cellular and molecular interactions that underlie their survival, proliferation, and drug resistance. Further emphasising this importance, has been the observation that highly effective small molecule inhibitors that target key components of the BCR signalling machinery, such as Bruton's tyrosine kinase (BTK), disrupt the migratory behaviour of CLL cells, and that this may, at least in part, underlie their clinical effect. Detailed characterisation of the highly dynamic cytoskeletal alterations in CLL may, therefore, inform novel therapeutic interventions, particularly for subsets with unmet clinical needs, such as those with mutations affecting the tumour protein P53 (TP53), ataxia telangiectasia mutated (ATM), or Notch receptor 1 (NOTCH1) genes, which are all more frequent in IGHV-unmutated disease. This work describes the development of techniques to characterise and quantify morphological responses to inhibitors, aiming to produce a quantitative description of cytoskeletal changes relating to specific signalling pathways, and to suggest rational drug combinations in the disorder. Methods: Primary CLL cells were cultured at high density with autologous T-lymphocytes and monocytes in the presence of specific signal inhibitors. The morphological responses of leukaemic cells were determined using a range of microscopic techniques, including scanning electron microscopy, and immunocytofluorescent detection of cytoskeletal and plasma membrane components. Cytoskeletal alterations were evaluated via computer-aided morphometric analyses of cell shape parameter, homotypic cellular interactions, and migration, generating a precise description of changes to the polymerised F-actin cytoskeleton and cell behaviour. Matrigel™ matrix models were combined with transmission electron microscopy to study cellular morphology within a 3D tumour microenvironment (TME)-like setting. Immunogold labelling of specific proteins within neoplastic lymphocytes was performed to allow visualisation of protein localisation changes in response to signal inhibition at the ultrastructure level. Results: This study tested inhibitors targeting different signalling pathways as a 'proof of principle' evaluation to determine whether the morphological and behavioural responses induced could be effectively distinguished from one another and quantitatively described. Inhibition of BTK by ibrutinib resulted in uniform populations of globular cells with retained polarity and, consequently, increased amoeboid motility. BTK blockage is recognised to impair integrin-mediated retention of leukaemic cells within tissue niches, leading to the observed peripheral blood lymphocytosis seen in CLL patients receiving ibrutinib. Reduced integrin-mediated motility was associated with impaired homotypic cellular interactions within IGHV-mutated cases specifically, indicating that this subgroup may have a greater dependency on elongated-type migration for permitting pro-survival cellular contact than their IGHV-unmutated counterparts. Disruption of Rho-associated protein kinase 1 (ROCK1) activity by Y-27632 lead to impaired actomyosin-mediated retraction of cytoskeletal processes. Loss of the ROCK1-induced cytoskeletal asymmetry required for effective cell migration resulted in reduced CLL cellular interactions; however, CXCL12-driven motility was attenuated in IGHV¬-mutated cases alone. The Abelson kinase 1 (ABL1) inhibitor imatinib caused CLL cells to acquire a globular phenotype with frequent microvilli, similar to that of B-lymphocytes isolated directly from the peripheral blood. Transient cellular interactions were markedly reduced by imatinib, whereas elongated-type motility, being a largely ABL1-independent process, was unaltered. The morphological and behavioural responses of CLL cells were compared to those observed in mantle cell lymphoma (MCL) cell lines. These cells lines, which were utilised as a surrogate model for BTK inhibitor sensitivity in CLL, demonstrated that the establishment of anterior-posterior morphology, mediated by the activity of ROCK1 and ABL1, is essential for effective trafficking of B-lymphocytes to protective niches, regardless of ibrutinib sensitivity. Blockage of NOCTH1 signalling by gamma-secretase inhibitors (GSIs) PF-03084014 and R04929097 resulted in varying morphological responses, possibly indicating differences in NOTCH1 activation between CLL cases. Despite chemotaxis being identified as a key NOTCH1-regulated process, CLL and NOTCH1-mutated MCL cells demonstrated enhanced directional transmigration with GSI treatments. MCL cell lines were utilised to model the effects of GSI sensitivity in CLL. In contrast, NOTCH1-unmutated MCL cells displayed unaltered migration with PF-03084014 pre-treatment, consistent with reports of low GSI sensitivity in MCL cells exhibiting unmutated NOTCH1, and reduced chemotaxis with R04929097. The developed 3D ex vivo culture system preserved CLL cell viability, migration, and dynamic cellular interactions, as demonstrated by flow cytometry and time-lapse live-cell imaging. Interrogation with transmission electron microscopy enabled high-resolution visualisation of cell morphology within a TME-like setting; however, further optimisation of immunogold labelling of effector proteins is required. Conclusion: Using novel imaged-based morphometric analyses, distinct signal inhibitor-induced cytoskeletal adaptations were identified in CLL B-lymphocytes. This approach may be applied to prognostically-defined subgroups or resistant cases to provide in-depth characterisation of morphological responses to novel therapeutic agents and to assess treatment responses within the TME. These observations, when combined with transcriptional data, may allow more effective combinational targeting of behavioural signatures unique to the patient and, thus, improve treatment outcomes in the disease.