SRSF2 Mutations Impair Nuclear Aggregation and Represent a Novel Mechanism Of Spliceosomal Dysregulation In CMML

SRSF2 is a member of the spliceosome complex that is mutated at, or near, proline 95 in 30-50% of patients with Chronic Myelomonocytic Leukemia (CMML). The molecular consequence of SRSF2 mutation and its contribution to the pathogenesis of CMML is unknown. SRSF2 has two major functions within the nucleus: (1) bind cis elements on pre-mRNA functionally redefining exon-intron boundaries, and (2) interact with other members of the spliceosome complex at interchromatin granules (ICGs, or nuclear speckles). So that these two events occur in an ordered manner, phosphorylation/dephosphorylation of SRSF2 in the nucleus alters the relative affinity for RNA capture versus spliceosomal interaction and nuclear aggregation. We hypothesize that SRSF2 P95 mutations disrupt this delicate balance resulting in altered RNA binding affinity and impaired nuclear aggregation.

An in silico and in vitro approach was used to investigate the effects of the most common SRSF2 mutants (MT) P95H, P95L, and P95R on RNA binding. In silico three dimensional modeling localized P95 to the base of a dynamic “clamp” within the RNA-binding pocket of the known RNA-bound SRSF2 NMR structure. In silico mutation analysis was performed and Poisson-Boltzmann electrostatic surface potentials for each of the SRSF2 MT at 310 K with +/- 4 Å isovalues were calculated. There was a net positive charge in the electrical potential over the entire electrostatic surface map of wild type (WT) SRSF2 consistent with its ability to capture negatively charged RNA. The magnitude of gain in the net positive charge for all SRSF2 MTs was elevated and reached a 2-fold increase for P95R, supporting increased RNA binding capacity of the SRSF2 P95R MT. To experimentally determine the impact of SRSF2 MT, constructs containing GFP alone and the above SRSF2 MTs were generated and transfected into HeLa cells. Cellular lysates from SRSF2 transfected cells were incubated with a biotin-labeled RNA probe containing the cis element required for SRSF2 capture and binding was assessed by chemiluminecent RNA-EMSA and RNA pull-down. These assays demonstrated efficient capture of the SRSF2 MT and WT proteins with a relative increase in SRSF2 P95R RNA probe binding compared to SRSF2 WT lysates consistent with in silico modeling. Surface plasmon resonance analysis is underway using the biotin labeled probe to determine the precise RNA binding kinetics of SRSF2 MT and WT proteins.

To explore the effects of SRSF2 P95 MT on ICG formation, GFP-tagged constructs, as described above, were transfected into HeLa cells. Live-culture fluorescence microscopy (FM) defined the cellular location of SRSF2 by measuring GFP luminescence at 3 minute intervals for a total of 18 hrs. A custom watershed analysis using the Definians Developer XD software suite™ was used to calculate ICG numbers, mean ICG intensity, and mean ICG area of every GFP positive cell. In two separate transfection experiments, approximately 175 SRSF2 transfected cells were tracked each with up to 360 frames of ICG data for a total of up to 63,000 GFP positive events. Morphologic analyses of merged GFP and bright field images demonstrated that the localization of WT and MT SRSF2 was restricted to the nucleus. In contrast, control GFP transfected cells demonstrated diffuse cellular expression of GFP. This data suggests that P95 MT did not alter the predicted nuclear localization of SRSF2. However, two-way ANOVA demonstrates an impairment of maximal ICG number per cell (p<0.001) and maximal ICG positive cells (ICG) (p<0.001) in SRSF2 MTs compared to SRSF2 WT, indicating that P95 MT inhibit ICG (nuclear speckle formation) in transfected HeLa cells. When tracking GFP-SRSF2 by FM, we also observed that SRSF2 transfected cells displayed morphologic evidence of apoptosis compared to GFP only controls. To determine the effects of SRSF2 MT on the rate of apoptosis, we calculated the time from first expression of GFP to first evidence of apoptosis for all GFP positive cells. Using the time-to-event Kaplan-Meier Method, we found that overexpression of WT SRSF2 induced apoptosis relative to GFP only-transfected cells. Compared to WT, all SRSF2 MT constructs partially rescued the rate of apoptosis demonstrating a relevant survival promoting consequence of SRSF2 MT (p<0.001).These data suggest that SRSF2 MT alters both RNA binding and nuclear speckle formation representing a novel molecular abnormality mediating spliceosomal dysregulation in CMML.