Fig 2: FIGURE 3: FLCN, but not FNIP1, is required for shear stress-induced autophagy.(A-D) HK2 cells were transfected either with siRNA targeting FLCN (siFLCN) or control siRNA (siCTRL) 72 h later they were subjected to fluid flow (shear) for the indicated times or not (static). (A) Levels of FLCN, LC3I and LC3II were analyzed by western blot and LC3 II/actin ratio was quantified (B). (C) Cells were fixed, labeled with DAPI, immunostained for LC3 and then analyzed by fluorescence microscopy. (D) LC3 dots were quantified from experiments shown in (C). (E-H) HK2 cells were transfected either with siRNA targeting FNIP1 (siFNIP1) or control siRNA (siCTRL). 72 h later they were subjected to fluid flow for 4 days (shear 4D) or not (static 4D). (E) Levels of FNIP1, FLCN, LC3I and LC3II were analyzed by western blot and LC3 II/actin ratio was quantified (F). (G) Cells were fixed, labeled with DAPI, immunostained for LC3 and then analyzed by fluorescence microscopy. (H) LC3 dots were quantified from experiments shown in (G). Scale bars in (C) and (G) = 5μm.

Fig 3: FIGURE 4: FLCN, but not FNIP1, is required for shear stressinduced cell size regulation.(A-B) HK2 cells were transfected either with siRNA targeting FLCN (siFLCN) or control siRNA (siCTRL) 72 h later they were subjected to fluid flow for 4 days (shear 4D) or not (static 4D). (A) Cells were fixed, labeled with DAPI and phalloidin to reveal F-actin and cell boarder and then analyzed by fluorescence microscopy. (B) Cells areas were quantified from experiments shown in (A). (C-D) HK2 cells were transfected with siRNA targeting FNIP (siF-NIP1) or control siRNA (siCTRL)). The experiments were performed and quantified as in (A-B). Scale bars in (A) and (C) = 10µm.

Fig 4: FNIP knockdown results in increased mTOR lysosomal localization and diffusion of FLCN in HeLa cellsA. Immunoblot (IB) analysis of whole cell lysates derived from HeLa cells stably expressing shRNA against GFP (control), FNIP1, FNIP2, or FNIP1/FNIP2. After being serum and amino acid-starved for 3 h, cells were treated with fresh 10% FBS DMEM and harvested at the indicated time points. B-E. Confocal images of HeLa cells presented in (A). DAPI-loaded HeLa cells were analyzed for co-localization of FLCN (B and D) (red) or mTOR (C and E) (red) with a lysosomal marker, LAMP1 (green). Y (yellow) indicates predominant localization of FLCN or mTOR in the lysosome. Scale bars, 20 μm (5 μm in the enlarged images).

Fig 5: Inactivation of the FLCN-FNIP1/2 axis activates STAT2 in renal cells.(A) qRT-PCR levels of genes with ISRE or E-box motif in FNIP1POS/FNIP2POS and FNIP1NEG/FNIP2NEG RPTEC cells reveal that the identified FLCN-dependent gene signature is also induced upon loss of FLCN interacting proteins FNIP1 and FNIP2. Results shown are representative for two independent experiments with three technical replicates. To determine quantitative gene expression data levels were normalized to the geometric mean of two housekeeping genes. See Figure 7—source data 1 for raw qRT-PCR values and fold change calculations. (B) qRT-PCR levels of genes with ISRE or E-box motif in FLCNPOS and FLCNNEG retinal pigment epithelial cells (RPE/TERT1 tet on Cas9 TP53KO) reveal that the identified FLCN dependent gene signature is absent in an epithelial cell type of another tissue origin. Results shown are representative for two independent experiments. To determine quantitative gene expression data levels were normalized to the geometric mean of two housekeeping genes. *OAS2 level in FLCNNEG RPE was too low to detect using qRT-PCR. See Figure 7—source data 1 for raw qRT-PCR values and fold change calculations. (C) Spearman correlation analysis reveals overlapping FLCN-dependent RNA and protein data. FLCN differential mRNAs and proteins (FDR < 0.01, n = 181, red line) showed a higher correlation than the overlap of all identified mRNAs and proteins in our datasets (blue line). Statistical methods are described in Materials and methods section. (D) iRegulon analysis of differentially expressed genes (FDR < 0.01) with highest correlation with differentially expressed proteins (r > 0.8, n = 49) reveal STAT1, STAT2, IRF1, and IRF3 as most obvious upstream transcriptional regulators. Only regulons displaying normalized enrichment scores (NES) > 4.5 are shown. STAT2 appears twice due to the fact that iRegulon ranks this transcription factor to be the most likely upstream regulator for two sets of targets genes, containing slightly different ISRE-motifs upstream from the transcription start site (TSS). (E) Reintroducing FLCN (overexpression, OE) or siRNA-mediated knock down of STAT1/STAT2 (10 nM, 72 hr) revert the IFN expression gene signature induced by FLCN loss in RPTEC FLCNNEG C2. FLCN OE also lowers the enhanced expression of E-box-associated target genes but knock down of STAT1/2 has no effect on E-box-associated genes. Expression levels were determined by qRT-PCR and are representative of two independent experiments. To determine quantitative gene expression data levels were normalized to the geometric mean of two housekeeping genes. See Figure 7—source data 1 for raw qRT-PCR values and fold change calculations. (F) Western blots of subcellular fractionated samples show higher expression of STAT1 and STAT2 in FLCNNEG RPTEC as compared to FLCNPOS RPTEC. STAT2 was also detected in both cytoplasmic and nuclear fractions. Tubulin and histone H3 levels were used as loading control and to distinguish each fraction (N=nuclear, C=cytoplasmic). Results shown are representative of two independent fractionations. (G) Western blot of subcellular fractionated samples shows enhanced STAT2 DNA binding in FLCNNEG RPTEC. Results shown are representative of three independent fractionations.Figure 7—source data 1.Raw qRT-PCR values and fold change calculations belonging to 7A, 7B and 7E.