Abstract Background and Aims Focal segmental glomerulosclerosis (FSGS) is a common glomerular injury and represents a leading cause of ESRD. Studies with genetic FSGS show that a majority of genes are associated with regulation of podocyte cytoskeletons. INF2 is an actin assembly factor that mediates both polymerization and depolymerization and is the most frequent cause of dominant inherited FSGS. INF2 mutations located in the proximal half of DID (residues Leu57 to Glu184) typically cause CMT with early-onset FSGS, whereas that situating in the distal half (residues Glu184 to Leu245) cause FSGS alone. Previous studies disclosed that INF2 mutations primarily disorganize actin network and alter mitochondria shapes and vesicle motility. However, it remains elusive how the cytoskeletal network is concerted with the organelle morphology and functions in the disease cells. This study aims to characterize the interaction between cytoskeleton and organelle in cells expressing INF2 FSGS or CMT/FSGS variants. Method We characterized the cytoskeletal-organelle interaction in live HeLa or COS-7 cells expressing pathogenic INF2 variants in our cohort of FSGS. We performed a triple labeling by co-transfection of GFP-INF2, cytoskeletal markers (Lifeact-mCherry or EMTB-mScarletI), and by the uptake of organelle markers (Lysotracker or Mitotracker). To clarify the detailed pattern and dynamics of cytoskeletons and organelles, we employed the Spinning Disk Confocal Microscope DragonFly, which allows real-time super-resolution imaging with 70nm resolution. Results Wild-type-INF2 (WT-INF2) cells showed an ER pattern with perinuclear, Golgi accumulation. Mild INF2 variants causing FSGS alone (T161N, N202S) preserved the subcellular distribution essential similar to WT-INF2, while showing focal irregularity of ER networks. In contrast, severe INF2 variants leading to CMT/FSGS (G73D) exhibited a diffuse coarse ER pattern with few Golgi clusters and occasional peripheral accumulation in both cell poles. In live imaging of cells coexpressing GFP-INF2 and Lifeact-mCherry, actin stress fibers were pronouncedly reduced in CMT/FSGS variant (G73D) cells, but were mildly decreased FSGS variant (T161N and N202S) cells, compared with WT-INF2 cells. The filopodia in cells expressing T161N variant were less mobile (n = 10, P< 0.01) but were longer (n = 10, P< 0.0001) than those in cells expressing WT-INF2. T161N variant reduced expression of cortactin that stabilizes actin-branching and narrowed area of cortical actin meshwork compared to the WT-INF2 (n = 11, P< 0.01), suggesting that INF2 could serve the cortical actin formation. HeLa cells coexpressing T161N variant and EMTB showed a bipolar cell elongation along the longitudinal axis (n≥30, P< 0.0001) and disarranged microtubule bundles aligned in parallel, compared with those expressing WT-INF2. A time-lapse imaging of living COS-7 cells revealed that T161N variant increases mitochondrial fragmentation as well as mitochondria-ER interfaces more prominent than WT-INF2. Moreover, T161N variant induced the mitochondria misdistribution at the cell periphery (filopodia and cell edge) in both cell poles. Such aberrant patterns frequently coexisted with disarranged microtubules. Moreover, in lysotracker-labeled living COS-7, T161N variant reduced the motility of the lysosomal vesicles in cortical actin meshwork area, compared with WT-INF2 (n = 100, P< 0.0001). These observations suggest that INF2 variants perturbate the trafficking of organelles through disorganization of actin-microtubule network. Conclusion Our results indicate that INF2 CMT/FSGS variants cause more global and severe organelle dysfunction by disrupting the organelle-cytoskeletal interaction, than INF2 FSGS variants. The microtubule disarrangement might be due to an insufficient capping mechanism by which the microtubules are stably tethered to the cortical actin meshwork. Elucidation of cell-specific factors will help better understanding why podocyte and Schwann cell are susceptible to the cytoskeletal disarrangement.
Read full abstract