PYK2 is overexpressed in chronic lymphocytic leukaemia: A potential new therapeutic target

Abstract Chronic Lymphocytic Leukaemia (CLL) is the most common adult B‐cell leukaemia and despite improvement in patients' outcome, following the use of targeted therapies, it remains incurable. CLL supportive microenvironment plays a key role in both CLL progression and drug resistance through signals that can be sensed by the main components of the focal adhesion complex, such as FAK and PYK2 kinases. Dysregulations of both kinases have been observed in several metastatic cancers, but their role in haematological malignancies is still poorly defined. We characterized FAK and PYK2 expression and observed that PYK2 expression is higher in leukaemic B cells and its overexpression significantly correlates with their malignant transformation. When targeting both FAK and PYK2 with the specific inhibitor defactinib, we observed a dose–response effect on CLL cells viability and survival. In vivo treatment of a CLL mouse model showed a decrease of the leukaemic clone in all the lymphoid organs along with a significant reduction of macrophages and of the spleen weight and size. Our results first define a possible prognostic value for PYK2 in CLL, and show that both FAK and PYK2 might become putative targets for both CLL and its microenvironment (e.g. macrophages), thus paving the way to an innovative therapeutic strategy.

cells in the PB become quiescent, B lymphocytes in tissues find a supporting niche and actively proliferate. 4 In the specific case of CLL, main actors and mechanisms through which the TME regulates the leukaemic clone progression still need to be fully elucidated. 5 We and others previously demonstrated that the trafficking and homing of malignant B cells is tightly regulated by cytoskeletal reorganization 6 and by different cells in the TME that contribute to CLL survival and progression. 7 In particular, the survival of CLL cells is supported by the direct contact with nurse-like cells (CLL-specific tumour associated macrophages 5 ) and targeting macrophages lead to indirect CLL cells apoptosis and inhibition of disease progression. 8 Along this line, recent evidence shows that there is a complex interaction between cancer cells and the TME through an exchange of mechanical perturbations sensed by surface 'sensor' receptors, which translate these inputs into chemical signals. The Focal Adhesion complex is one of these sensors, and its main components are the Focal Adhesion Kinase (FAK) and its related homologue protein Proline-rich Tyrosine Kinase 2 (PYK2) also known as PTK2 and PTK2B, respectively. These are non-receptor protein kinases, which play an important role in several cellular processes, beside mechanosensing and cell adhesion or migration, including proliferation and apoptosis. While FAK is expressed in almost all tissues, PYK2 is expressed mainly in haematopoietic cells. 9 Both FAK and PYK2 dysregulations have been observed in different metastatic cancers, but their role and expression in haematological malignancies and specifically in leukaemia, is not well defined yet. 10,11 In a genome-wide study, PYK2 has been found to correlate with improved outcomes in CLL patients under chemoimmunotherapy regimen. 12 More recently FAK has been proposed as a modulator of migration and invasion in CLL. 13 In a pre-clinical setting, it has been previously demonstrated that targeting FAK and/or PYK2 could bring some benefits in some haematological cancers such as Multiple Myeloma, 14 Myelodysplastic Syndromes, 15 Acute Myeloid Leukaemia, 16 BCR/ ABL-transformed model of Chronic Myeloid Leukaemia 17

and in
Mantle Cell Lymphoma (MCL). In particular, it has been demonstrated that MCL cells, upon culture with stromal cells, show an activation of pathways such as NF-kB, Akt, c-Myc, Cyclin-D and p42/44, all associated with cancer progression, and FAK inhibition abrogates the activation of all these effectors. The same happens after the occurrence of resistance to the BTK inhibitor ibrutinib, providing a rationale for the administration of specific FAK inhibitors in combination with ibrutinib in MCL. 18 Inhibitors exist and target focal adhesion tyrosine kinases on ATP binding sites inhibiting their catalytic activity to repress the auto-phosphorylation activity in Tyr397 on FAK and Tyr402 on PYK2. A second-generation inhibitor VS-6063 (i.e. defactinib or PF-04554878) by Verastem Oncology, with the same mechanism of action, has already been proven safe and effective in patients in several clinical trials for solid tumours. [19][20][21][22] As of October 2022, 7 clinical trials are currently ongoing for defactinib testing, for a total of more than 20 studies overall, but only one is investigating defactinib effects on haematological malignancies. 23 In this work, we aim at defining FAK and PYK2 expression in CLL cells and exploring the effect of defactinib administration to CLL, both in vitro (2D and 3D culture) and in vivo, to understand how leukaemic cells could exploit mechano-transduction pathways like FAs recruitment via FAK and PYK2 activation, to sustain their development, progression and resistance to therapy.

| Immunofluorescence (IF) microscopy
Glass slides were coated with Poly-ornithine for 20 min at 37°C.

| 3D scaffolds preparation, co-culture and treatments in RCCS™ bioreactor
For CLL 3D co-culturing conditions, the protocol previously adapted to CLL in our lab has been followed. 25 Briefly: Gelatine disks of 4 mm of diameter and 3 mm height were cut from Spongostan™ gelatine sheets (Ethicon) using a sterile biopsy punch and then pre-seeded with BM derived stromal cells HS5 (200.000/scaffold) for 4 h in 96well suspension culture plate (Greiner Bio-one). Scaffolds were then transferred to 10 mL High Aspect Ratio Vessels (HARV) in 600 μL DMEM culture medium supplemented with 10% v/v FBS and cultured overnight in the RCCS™ bioreactor. The day after, CLL cells were added to the vessels, following the optimal CLL cells:stromal cells ratio, set in previous experiments. 25 After 5 h, vessels were filled with growth medium (RPMI 1640 culture medium supplemented with 20%). After 48 h of 3D dynamic culture in the bioreactor, supernatants were withdrawn from the vessels and centrifuged at 1500 rpm for 5 min. Recovered cells were counted. Clarified supernatants were put again in the vessels with or 4 μM defactinib (Verastem Oncology). Cultures were stopped after 4 h of treatment and cells in the supernatants and in the scaffolds were recovered and submitted to Trypan Blue exclusion test for viability and cell count.

| Flow cytometry analysis
For cell culture flow cytometry analysis cells were recovered, washed with 3 mL of either PBS (Euroclone) or binding buffer (PBS + 5%FBS) and centrifuged at 1500 rpm for 5 min. We next added the following antibodies: Annexin V-FITC (Thermo Fisher, 5 μL/tube), PI (Thermo Fisher, 10 μL/tube). We incubated for 20 minutes at room temperature in the dark. We then washed with 3 mL of PBS or binding buffer and resuspended the sample in 500 μL PBS. Samples were read at the flow cytometer (Navios; Beckman Coulter). Data were analysed by FCS express software. For in vivo mice experiments. Cells from PB, SPs, BM, lymph nodes and peritoneal washes from each mouse were divided into three tubes, they were washed with 3 mL of PBS (Euroclone) and centrifuged at 1500 rpm for 5 min. Staining continued as follows: CD19-PC7 (BioLegend, 1 μL), CD5-APC (BioLegend, 1 μL), CD11b-PE-Cyanine7 (BD, 1 μL). Samples were read at the flow cytometer (Navios; Beckman Coulter). Data were analysed by FCS express software.

| Cell titre viability assay
The assay was performed according to the manufacturer kit (Cell Titre Glo Luminescent Cell Viability Assay, Promega). CLL cells were plated 3 × 10 6 /mL in their culture medium within a 96-opaque-walled plate in a volume of 100 μL. Control wells were prepared adding only medium.
Then, an equal amount of Cell Titre Glo Reagent was added to each well and the two parts were mixed for 2 min on a shaker. The multiwell was then left 10 min at room temperature to stabilize the signal.
Finally, luminescence was recorded with a Luminometer. The results were compared to an ATP standard curve previously generated.

| 2D co-cultures and cells treatments
Pre-coating with either 200 × 10 3 HS5 stromal cells/well was done by seeding cells the day before the start of the co-cultures. 100 × 10 3 CLL/mL were added in suspension onto plates, stromal cells, endothelial cells, with 4 μM defactinib (Verastem Oncology) for 48 h. At the indicated time-points, CLL cells were recovered and analysed by flow cytometry for viability.

| Mouse model and treatments
Eight-week-old C57BL/6 female mice (Charles River Laboratories) were injected intraperitoneally with 10 7 cells purified from SP of a 12-month-old Eμ-TCL1 leukaemic mouse (on a C57BL

| Human and mouse ethics statements
Patients with CLL were diagnosed according to the updated National Cancer Institute Working Group (NCIWG) guidelines. 26

| FAK and PYK2 are differentially expressed in healthy B and CLL cells
We investigated FAK and PYK2 mRNA expression levels in B cells obtained from the peripheral blood (PB) of patients with CLL (n = 36, Table 1) or healthy donors (n = 10; Figure 1A Figure 2B,C). 30 For FAK, we observed that in MBL cases its expression is close to healthy B cells ( Figure 2B), while for PYK2 we showed that in MBL cases its expression is significantly higher with respect to healthy B cells (healthy B vs. MBL p = 0.0116, Figure 2C).
These results suggest that a dysregulated balance of expression of the two kinases can occur during disease progression. We further investigated FAK and PYK2 localization and expression in CLL cells by immunofluorescence analysis (4 patients, 10 images each, Figure 2D). As expected, we found both kinases expressed mainly in the cytoplasm of CLL cells. In parallel, we observed a constitutive activation of both kinases, in particular pFAK (Y397) was observed also in the nuclei of some CLL cells while p-PYK2 (Y402) was more punctuated in podosome-like structures 31 ( Figure 2D) possibly suggesting a different role or regulation for the two proteins.

| Defactinib treatment of CLL cells induces apoptosis in vitro and downregulation of survival associated pathways
Based on our observations of PYK2 overexpression and activation in CLL cells, we aimed at testing whether the pharmacological inhibition of PYK2, might have any relevant therapeutic effect. To this aim, we exploited the potent and selective ATP-competitive FAK/PYK2 inhibitor defactinib. [19][20][21][22] Firstly, we performed a viability assay testing different doses of the drug, ranging from 2 to 10 μM, on primary PB-derived CLL cells

| Defactinib treatment in a transplanted Eμ-TCL1 mice model shows a reduction of CLL cells in the lymphoid tissues
To investigate in vivo the role of FAK and PYK2 and the effects of defactinib in CLL, we utilized the transplantable CLL mouse model Eμ-TCL1. 28 Diseased mice were randomized in two groups to receive defactinib (n = 18) or vehicle as control group (n = 15) as depicted in  Figure 4C).
Finally, we were interested in evaluating the macrophage population in mice treated with defactinib compared to the controls, as it is known that macrophages support CLL progression 8

| DISCUSS ION
The role of the microenvironment in CLL is of particular importance for both the pathogenesis of the disease and the emergence of drug resistance, also against novel more effective targeted therapies. that little is known about the CLL cells mechanism of retention and release from the protective niche present in the lymphoid tissues.
Along these lines, we found a significant correlation between FAK downregulation and progressive CLL cases and we observed  we also observed a high variability of response between patients that was not associated to the different level of expression and activation of the kinases, suggesting a more complex regulation. We also hypothesized that the inhibitor could have a mobilization effect on CLL cells impacting on their adhesion capacity to microenvironmental cells but we could not detect a clear effect in our 3D model that e previously demonstrated to be able to measure this effect for instance when cells are exposed to ibrutinib. Lastly, we exploited an in vivo CLL mouse model, where we confirmed defactinib proapoptotic effect on CLL cells and interestingly we found that the macrophage population is a potential new target for the drug in this leukaemia, also knowing their crucial role in supporting CLL cells viability and protection from apoptosis. Collectively, these results

CO N FLI C T O F I NTE R E S T S TATE M E NT
Authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.