HIGH-RESOLUTION SPATIOTEMPORAL ANALYSIS OF SOMATOSTATIN RECEPTOR TYPE 2 (SSTR2) - FILAMIN A (FLNA) INTERACTION BY SINGLE-MOLECULE IMAGING

DOTTORATO DI RICERCA IN MEDICINA CLINICA E SPERIMENTALE (XXIX ciclo) Abstract ENG Cognome: TREPPIEDI Nome: DONATELLA Matricola n.: R10592 Title: "High-resolution spatiotemporal analysis of Somatostatin Receptor Type 2 (SSTR2) – Filamin A (FLNA) interaction by single-molecule imaging" Background: SSTR2 is the main pharmacological target to treat acromegalic patients harboring GH-secreting pituitary adenomas. However, about 30% of patients displays resistance to somatostatin analogues (SSAs). Recent published data from our group demonstrated that the cytoskeletal protein FLNA plays an essential role in tumor responsiveness by regulating SSTR2 signaling and stability after prolonged stimulation. To date, there are no evidence in the literature describing the dynamic of SSTR2-FLNA interaction at the plasma membrane in vivo. Aim: Aim of my PhD project was to follow the spatiotemporal behavior of SSTR2-FLNA complexes in real time by high resolution strategy. In particular I wanted to investigate the presence of a spatial distribution of SSTR2-FLNA complexes at the plasma membrane, estimate SSTR2 lateral mobility and elucidate a possible involvement of FLNA in regulating this biological phenomenon. A further goal was to evaluate the impact of FLNA-SSTR2 binding on ligand-induced SSTR2 clusters organization and internalization. Materials and Methods: To characterize the dynamics of SSTR2-FLNA complexes in living CHO cells we used the single-molecule imaging approach, a method that combines the labelling of SNAP/CLIP-tagged proteins (SNAP-tagged SSTR2 and CLIP-tagged FLNA) with small organic fluorophores and the use of a total internal reflection fluoresence (TIRF) microscope. To calculate SSTR2 lateral mobility, mean square displacement (MSD) analysis were performed using the u-track algorithm implemented in Matlab. Confocal microscopy was used to evaluate the receptor surface distribution, agonist-mediated clusterization and alignment with actin filaments. Immunofluoresce experiments were assessed to evaluate the colocalization between SSTR2 clusters and AP-2, one of the early endocytosis marker, whereas the overall SSTR2 internalization process was analyzed by both imaging (confocal microscopy) and biochemical (biotinylation assay) strategies. All these SSTR2 aspects were evaluated in the presence or the absence of FLNA interaction. In particular, the overexpression of the dominant negative mutant FLNA 19-20 was used to abolish SSTR2-FLNA binding. Results: First, the motion of freely diffusing SSTR2 particles was observed to slow down in CHO cells treated with 100nM BIM23120 for 5-10min (mean diffusion coefficients from 0,123µm2*s-1 to 0,101µm2*s-1). MSD analysis showed a significant increase in the SSTR2 fraction with diffusion coefficient values ≤ 0.05μm2*s-1 with respect to unstimulated cells (28,1% vs 14,4%, 100nM BIM23120 vs control, respectively, expressed as fraction of total particles, P < 0.05). The presence of the FLNA truncated mutant, that selectively prevents SSTR2-FLNA binding (FLNA 19-20), did not influence the SSTR2 agonist effect on receptor mobility. Such data was further confirmed in melanoma cell lines A7 (FLNA-expressing cells), M2 (FLNA-lacking cells). Then, we described the nature of the interactions between SSTR2 particles and FLNA fibers as extremely dynamic and transient under resting condition, whereas they resulted long-lasting and more stable after100nM BIM23120 exposure. Interestingly, when both FLNA and SSTR2 were expressed at single molecule level, SSTR2-FLNA complexes formation was seen to occur preferentially along actin filaments, in stimulated cells only. Furthermore, when overexpressed and activated by the agonist, SSTR2 was observed to undergo clusters formation, and FLNA-SSTR2 binding was required to preserve SSTR2 clusters alignment on actin structures and colocalization with the clathrin coated pits marker AP-2. In addition, quantitative analysis of the agonist-triggered SSTR2 internalization demonstrated a significant reduction of the internalization rate in the presence of FLNA 19-20 compared to negative control (FLNA 17-18), at all the tested time points (eg. 45,3%±1,4% internalization vs 71,4%±3,1% in FLNA 19-20 vs FLNA 17-18 transfected cells after 30min stimulation with 100nM BIM23120, respectively, P < 0,001), accordingly with biotinylation results. Discussion: The behavior of SSTR2-FLNA interactions were characterized for the first time by means high spatio-temporal resolution strategy. In particular we observed that the SSTR2 agonist is able to modulate the dynamic properties of receptor particles, as demonstrated by a significant increase in the immobile receptor fraction observed upon stimulation, compared to the basal state. Even though FLNA does not contribute to the enrichment of the static receptor population, it seems to act as scaffold platform where ligand-activated receptors preferentially stop, likely to initiate their signaling transduction, then followed by internalization. Moreover, we highlighted a key role of FLNA in anchoring SSTR2 complexes to the cortical actin cytoskeleton. In fact, the abolished SSTR2-FLNA interaction resulted in a remarkably reduced alignment of agonist-induced SSTR2 clusters along actin filaments, revealing that SSTR2 clusters cannot properly associate with the actin cytoskeleton without binding to FLNA. Furthermore, we demonstrated that like for other GPCRs and transmembrane proteins, FLNA interaction is required to initiate and sustain ligand triggered-SSTR2 endocytosis. Indeed, a decrease colocalization between SSTR2 clusters and AP2-defined pits, and a strong impairment of the internalization rate were both associated to the overexpression of FLNA-19-20. Altogether these results let us to speculate that, by tethering SSTR2 to actin filaments, FLNA may facilitate the formation of ligand-inducible complexes with interacting proteins that are necessary for the efficient endocytosis of the receptor into clathrin-coated vesicles. Conclusion: In conclusion, the present work provides evidence for the involvement of FLNA and the cortical actin cytoskeleton in the formation of compartmentalized domains at the plasma membrane, where SSTR2 are first assembled into functional units, and then subjected to endocytosis. Indeed, the deep understanding of these aspects may be useful to clarify the molecular mechanisms by which the cells can modulate the amount of active receptors at their surface, thus determining a variable responsiveness to SSAs, with possible implications in the pharmacological resistance to the clinical management of acromegaly.