Human folliculin delays cell cycle progression through late S and G2/M-phases: effect of phosphorylation and tumor associated mutations

Abstract

The Birt-Hogg-Dube disease occurs as a result of germline mutations in the human Folliculin gene (FLCN), and is characterized by clinical features including fibrofolliculomas, lung cysts and multifocal renal neoplasia. Clinical and genetic evidence suggest that FLCN acts as a tumor suppressor gene. The human cell line UOK257, derived from the renal cell carcinoma of a patient with a germline mutation in the FLCN gene, harbors a truncated version of the FLCN protein. Reconstitution of the wild type FLCN protein into UOK257 cells delays cell cycle progression, due to a slower progression through the late S and G2/M-phases. Similarly, Flcn (-/-) mouse embryonic fibroblasts progress more rapidly through the cell cycle than wild type controls (Flcn (flox/flox)). The reintroduction of tumor-associated FLCN mutants (FLCN ΔF157, FLCN 1-469 or FLCN K508R) fails to delay cell cycle progression in UOK257 cells. Additionally, FLCN phosphorylation (on Serines 62 and 73) fluctuates throughout the cell cycle and peaks during the G2/M phase in cells treated with nocodazole. In keeping with this observation, the reintroduction of a FLCN phosphomimetic mutant into the UOK257 cell line results in faster progression through the cell cycle compared to those expressing the wild type FLCN protein. These findings suggest that the tumor suppression function of FLCN may be linked to its impact on the cell cycle and that FLCN phosphorylation is important for this activity. Additionally, these observations describe a novel in vitro assay for testing the functional significance of FLCN mutations and/or genetic polymorphisms.

Figure 1. FLCN delays progression through the late S and the G2/M phases of the cell cycle.

(A) Representative cell cycle profiles of UOK257 cells infected with retroviruses expressing the vector only or WT FLCN. Progression through the cell cycle was measured by propidium iodide (PI) labeling of synchronized cells collected at 0, 4, 8 and 12 hours after release from thymidine block. (B) FLCN-deficient UOK257 cells (squares and dashed line) and their isogenic counterparts reconstituted with wild type (WT) FLCN (triangles and solid line) were synchronized by double thymidine block and their progression through cell cycle after release was analyzed by FACS. Vertical axis indicates the percent of all cells registering in that specific phase of cell cycle. Horizontal axis indicates hours post release from thymidine. The average of three experiments is presented (n = 3), * indicates a statistically significant difference (t test, p<0.05), bars correspond to standard error of the mean (SEM).

Figure 2. MEF cells lacking the Flcn gene progress more rapidly through the cell cycle.

MEF cells with a floxed copy of the Flcn gene (Flcn ^flox/flox; circles and solid line) and MEF cells null for Flcn following Cre recombinase mediated excision and recombination (Flcn ^–/–; squares and dashed line) were synchronized by double thymidine block. FACS analysis was used to determine the percentage of cells in each stage of the cell cycle (y axis) over the course of 12 hours (x axis). Error bars indicate SEM from three experiments (n = 3), * indicates a statistically significant difference (t test, p<0.05).

Figure 3. Tumor-associated FLCN mutations fail to delay cell cycle progression.

(A) Expression of exogenous FLCN in UOK257 cells infected with retroviruses harboring empty vector control, FLCN WT, FLCN ?F157, FLCN K508R, or FLCN 1–469. (B) Cell cycle profiles of the UOK257 cells and the isogenic cell lines expressing empty vector control, FLCN WT, FLCN ΔF157, FLCN K508R, or FLCN 1–469, demonstrating the amount of cells in the G1, S and G2/M stages as determined by PI staining. The experiment was repeated three times (n = 3), but only one representative image of the cell cycle profile, 12 hours after release from thymidine block, is shown. (C) Tumor-associated FLCN truncated mutant 1–469 fails to delay cell cycle progression. FLCN deficient UOK257 cells (black dashed line) and their isogenic counterpart reconstituted with FLCN WT (red line) or FLCN 1–469 mutant (blue line) were synchronized by double thymidine block and released. Progression through the cell cycle was analyzed by FACS. Vertical axis indicates the percentage of all cells registering in the specific phase of cell cycle. Horizontal axis indicates hours post thymidine release. (D) Tumor-associated FLCN missense mutant K508R fails to delay cell cycle progression. Same as in (C), except that the reconstituted mutant is FLCN K508R (green line). (E) Tumor-associated FLCN deletion mutant ΔF157 fails to delay cell cycle progression. Same as in (C) and (D), except that the reconstituted mutant is FLCN ?F157 (purple line). The average of three experiments is presented (n = 3) in each panel; * indicates a significant difference between UOK257+ vector only compared to UOK257+ FLCN WT; # indicates a difference between UOK257 FLCN WT and the UOK257+ tumor-associated mutant (1–469, K508R, or ?F157); $ indicates a difference between UOK257+ vector only cells and the UOK257 cells expressing a tumor-associated mutant (One way ANOVA, Tukey's post-test, p<0.05). Error bars correspond to SEM.

Figure 4. FLCN is hyperphosphorylated during cell cycle progression from G2 to M.

(A) A slower migrating FLCN band is detected in cells arrested at the G2/M boundary and is likely the result of phosphorylation. The expression of endogenous human FLCN in nocodazole arrested, mitotic (M) U20S cells was analyzed by FLCN immunoblot after a 2 hour incubation with vehicle only (M+V), phosphatase (M+ Ph), or in the presence of phosphatase along with phosphatase inhibitors (M+ Ph +I). U20S cells growing asynchronously (Asy) were used as a control (hypophosphorylated). (B) FLCN phosphorylation decreases upon exit from mitosis. U2OS cells arrested by nocodazole in G2/M were collected by mitotic shake and replated. FLCN expression was detected by Western blot at increasing time intervals post-replating (time indicated on the horizontal axis). Actin is used as a loading control.

Figure 5. Identification of FLCN phosphorylation sites at the G2/M boundary.

(A) U2OS cells were cultured in the presence (lane 2) or absence (lane 1) of nocodazole overnight and immunoprecipitated FLCN protein was detected by Coomassie Blue staining of SDS-PAGE gels. (B) LC-MS/MS analysis of FLCN peptides. Extracted ion chromatograms for FLCN peptide AHSPAEGASVESSSPGPK with no phosphorylation (a), one phosphorylation (b) and 2 phosphorylation sites (c). Intensity corresponds to the absolute intensity of the highest peak in the chromatogram. No normalization was performed. (C) FLCN deficient UOK257 cells were reconstituted with FLCN WT, FLCN double phosphoinactivating mutant S62/73A or FLCN double phosphomimetic mutant S62/73E. Exogenous WT FLCN and the phosphomutants (S62/73A and S62/73E) are expressed in both the nuclear and the cytoplasmic cellular fractions. (D) FLCN deficient UOK257 cells and the isogenic cell lines reconstituted with either FLCN WT, a tumor-associated mutant (FLCN ?F157) or the phosphomutants (S62/73A and S62/73E) were cultured overnight in the absence or presence of nocodazole. (E) UOK257 FLCN WT cells growing asynchronously (Asyn) or arrested during G2/M with nocodazole were treated with either DMSO or cycloheximide (10 μg/ml). Protein stability was determined by Western blot at various time points over the course of 8 hours (V – UOK257 vector only cells; WT – UOK257 FLCN WT cells). Lysates were resolved in SDS-PAGE and FLCN protein detected with anti-FLCN antibody. Actin serves as a loading control.

Figure 6. FLCN phosphomimetic mutations inactivate the effect of WT FLCN on cell cycle.

(A) A FLCN phosphomimetic mutant (S62/73E) results in more rapid progression through the cell cycle compared to the FLCN WT, while the FLCN phosphoinactivating mutant (S62/73A) retains the ability to delay cell cycle progression. Representative cell cycle profiles from UOK257 cells and the isogenic UOK257 cells expressing FLCN WT, or the phosphomutants S62/73E or S62/73A, collected 12 hours after release from thymidine. UOK257 cells expressing FLCN WT or S62/73A have more cells in G2/M and fewer cells entering G1 compared to the FLCN null (vector only) and the S62/73E cells as measured by PI staining. (B) UOK257 (black line) and their isogenic counterparts reconstituted with WT FLCN (red line) or FLCN S62/73E mutant (blue line) were synchronized by double thymidine block. Cell cycle progression after release from thymidine was analyzed by FACS. Vertical axis indicates the percent of all cells registering in the specific phase of cell cycle. Horizontal axis indicates hours post thymidine release. The average of three experiments is presented (n = 3) in each panel; bars correspond to standard error of the mean (SEM). (C) Same as in (B), except that the reconstituted phosphomutant is FLCN S62/73A (purple line). For both (B) and (C), * indicates a statistically significant difference between UOK257+ vector only compared to UOK257+ FLCN WT at the corresponding time point; # indicates a difference between UOK257 FLCN WT and the UOK257+ phosphomutant (S62/73E or S62/73A); $ indicates a difference between UOK257+ vector only and UOK257+ phosphomutant (S62/73E or S62/73A) (One-way ANOVA, Tukey's post-test, p<0.05).