Abnormal growth of smooth muscle-like cells in lymphangioleiomyomatosis: Role for tumor suppressor TSC2

Abstract

The TSC1 and TSC2 proteins, which function as a TSC1/TSC2 tumor suppressor complex, are associated with lymphangioleiomyomatosis (LAM), a genetic disorder characterized by the abnormal growth of smooth muscle-like cells in the lungs. The precise molecular mechanisms that modulate LAM cell growth remain unknown. We demonstrate that TSC2 regulates LAM cell growth. Cells dissociated from LAM nodules from the lungs of five different patients with LAM have constitutively activated S6K1, hyperphosphorylated ribosomal protein S6, activated Erk, and increased DNA synthesis compared with normal cells from the same patients. These effects were augmented by PDGF stimulation. Akt activity was unchanged in LAM cells. Rapamycin, a specific S6K1 inhibitor, abolished increased LAM cell growth. The full-length TSC2 was necessary for inhibition of S6 hyperphosphorylation and DNA synthesis in LAM cells, as demonstrated by co-microinjection of the C-terminus, which contains the GTPase activating protein homology domain, and the N-terminus, which binds TSC1. Our data demonstrate that increased LAM cell growth is associated with constitutive S6K1 activation, which is extinguishable by TSC2 expression. Loss of TSC2 GAP activity or disruption of the TSC1/TSC2 complex dysregulates S6K1 activation, which leads to abnormal cell proliferation associated with LAM disease.

Figure 1.

(A) LAM cell morphology. LAM cells show typical smooth muscle–like cell morphology. The phase contrast image shown is representative of five LAM cell cultures derived from different patients with LAM. Image was taken on an Olympus IX microscope (original magnification: ×200). (B–J) Smooth muscle α-actin expression. Serum-deprived LAM-1/1, LAM-2/7, LAM-3/12, LAM-4/29, LAM-5/52, ASM cells, HVSM cells, HLFs, and HBFs were subjected to immunocytochemical analysis with anti-smooth muscle α-actin FITC-conjugated antibody (SM α-actin, green) followed by incubation with 4,6-diamidino-2-phenylindole to detect nuclei (blue). Images were taken using a Nikon Eclipse TE2000-E microscope (original magnification: ×400) and are representative of three separate experiments.

Figure 2.

Primary LAM cells have increased DNA synthesis. LAM-1/1, LAM-2/7, LAM-3/12, LAM-4/29, LAM-5/52, HVSM cells, HBFs, and HLFs were serum deprived for 48 h and pretreated for 30 min with 200 nM rapamycin or diluent. Cells were left unstimulated (Basal) (A) or were stimulated (B) with 10 ng/ml PDGF for 16 h. BrdU was added, and 24 h later, DNA synthesis was assessed by immunohistochemical analysis of BrdU incorporation as described in Materials and Methods. BrdU incorporation represents the percentage of BrdU-positive cells compared with the total number of cells. Data represent means ± SE of three separate experiments. *P < 0.001 for LAM versus HBFs, HLFs, and HVSM. **P < 0.001 for rapamycin-treated cells versus diluent-treated cells by ANOVA (Bonferroni Dunn).

Figure 2.

Primary LAM cells have increased DNA synthesis. LAM-1/1, LAM-2/7, LAM-3/12, LAM-4/29, LAM-5/52, HVSM cells, HBFs, and HLFs were serum deprived for 48 h and pretreated for 30 min with 200 nM rapamycin or diluent. Cells were left unstimulated (Basal) (A) or were stimulated (B) with 10 ng/ml PDGF for 16 h. BrdU was added, and 24 h later, DNA synthesis was assessed by immunohistochemical analysis of BrdU incorporation as described in Materials and Methods. BrdU incorporation represents the percentage of BrdU-positive cells compared with the total number of cells. Data represent means ± SE of three separate experiments. *P < 0.001 for LAM versus HBFs, HLFs, and HVSM. **P < 0.001 for rapamycin-treated cells versus diluent-treated cells by ANOVA (Bonferroni Dunn).

Figure 3.

(A) Rapamycin inhibits S6 phosphorylation in LAM cells. LAM-1/1, LAM-2/7, LAM-3/12, LAM-4/29, LAM-5/52, and ASM cells were serum deprived, and LAM-1/1 cells were preincubated with 200 nM rapamycin or diluent for 30 min. Control ASM cells were treated with rapamycin followed by stimulation with 10 ng/ml of PDGF. Cells were fixed, and immunocytochemical analysis with anti-phospho-ribosomal protein S6 antibody (P-S6, red) was performed followed by incubation with 4,6-diamidino-2-phenylindole (blue). Images were taken using a Nikon Eclipse E400 microscope (original magnification: ×400) and are representative of three separate experiments. (B) Ribosomal protein S6 is hyperphosphorylated in primary LAM cells. Serum-deprived LAM-1/1, LAM-2/7, LAM-3/12, LAM-4/29, LAM-5/52, ASM cells, and HBFs were stimulated with 10 ng/ml PDGF or diluent and subjected to immunoblot analysis with phospho-ribosomal protein S6 (P-S6) and S6 antibodies. Images are representative of three separate experiments. (C) Quantitative analysis of S6 phosphorylation. Phosphorylation levels of S6 were calculated using Gel-Pro analyzer software and normalized to total S6 protein levels. P-S6 in ASM cells was taken as a one fold. Data are means ± SE of three separate experiments. *P < 0.01 for LAM cells versus ASM cells and HBFs. **P < 0.01 for nonstimulated cells versus PDGF-stimulated cells by ANOVA (Bonferroni Dunn).

Figure 3.

(A) Rapamycin inhibits S6 phosphorylation in LAM cells. LAM-1/1, LAM-2/7, LAM-3/12, LAM-4/29, LAM-5/52, and ASM cells were serum deprived, and LAM-1/1 cells were preincubated with 200 nM rapamycin or diluent for 30 min. Control ASM cells were treated with rapamycin followed by stimulation with 10 ng/ml of PDGF. Cells were fixed, and immunocytochemical analysis with anti-phospho-ribosomal protein S6 antibody (P-S6, red) was performed followed by incubation with 4,6-diamidino-2-phenylindole (blue). Images were taken using a Nikon Eclipse E400 microscope (original magnification: ×400) and are representative of three separate experiments. (B) Ribosomal protein S6 is hyperphosphorylated in primary LAM cells. Serum-deprived LAM-1/1, LAM-2/7, LAM-3/12, LAM-4/29, LAM-5/52, ASM cells, and HBFs were stimulated with 10 ng/ml PDGF or diluent and subjected to immunoblot analysis with phospho-ribosomal protein S6 (P-S6) and S6 antibodies. Images are representative of three separate experiments. (C) Quantitative analysis of S6 phosphorylation. Phosphorylation levels of S6 were calculated using Gel-Pro analyzer software and normalized to total S6 protein levels. P-S6 in ASM cells was taken as a one fold. Data are means ± SE of three separate experiments. *P < 0.01 for LAM cells versus ASM cells and HBFs. **P < 0.01 for nonstimulated cells versus PDGF-stimulated cells by ANOVA (Bonferroni Dunn).

Figure 4.

S6K1 is constitutively active in primary LAM cells. (A) LAM-1/1, LAM-2/7, LAM-2/8, LAM-2/9, ASM, and HVSM cells were serum deprived, and cell lysates, equal in protein content, were probed with anti–phospho-Thr389 S6K1 (P-Thr389), phospho-Thr421/Ser424 S6K1 (P-Thr421/P-Ser424), or S6K1 antibodies. Images are representative of three separate experiments. (B) In vitro S6K1 activity was measured in serum-deprived LAM cells, ASM cells, and HVSM cells. *P < 0.001 for LAM cells versus ASM and HVSM cells by ANOVA (Bonferroni Dunn).

Figure 5.

Akt and ERK activity in LAM cells, ASM cells, and HBFs. Serum-deprived LAM-1/1, LAM-2/7, LAM-3/12, LAM-4/29, LAM-5/52 cells, ASM cells, and HBFs were stimulated with 10 ng/ml PDGF or diluent, and immunoblot analysis of cell lysates, equal in protein content, was performed with phospho-Akt (Ser-473) (P-Akt), Akt (A), anti-p44/p42 MAP kinase, or anti–phospho-p44/p42 MAP kinase (Thr202/Tyr204) (B) antibodies. Images are representative of three independent experiments.

Figure 6.

Co-immunoprecipitation of TSC2 constructs with TSC1. Serum-deprived ELT3 cells were co-transfected with Myc-tagged TSC1 and GFP-tagged TSC2 constructs: full-length TSC2, TSC1-binding domain of TSC2 (HBD-TSC2) (1–460 amino acids of TSC2), TSC2 without TSC1-binding domain (ΔHBD-TSC2) (461–1784 amino acids of TSC2), C-TSC2 (1114–1784 amino acids of TSC2), or N-TSC2 (1–1113 amino acids of TSC2). Coexpression of TSC2 constructs and Myc-TSC1 was detected with anti-GFP (top panel) and anti-Myc antibodies (middle panel), respectively. Immunoprecipitation with anti-Myc antibody was performed, followed by immunoblot analysis of immunoprecipitates with anti-GFP antibody to detect GFP-tagged TSC2 constructs co-precipitated with Myc-tagged TSC1 (bottom panel). Images are representative of two separate experiments. Rectangles indicate GFP-tagged TSC2 constructs.

Figure 7.

Inhibition of ribosomal protein S6 phosphorylation and cell proliferation require the N- and C-termini of TSC2. Serum-deprived LAM-1/1 (A), ELT3, and ERC15 cells (B) were microinjected with GFP, GFP-TSC2, or co-microinjected with GFP–N-TSC2 and GFP–C-TSC2 followed by immunostaining with anti-phospho-ribosomal protein S6 (P-S6, red) and anti-GFP antibodies (green); yellow fluorescence results from the co-localization of P-S6 and GFP in cells. Images were analyzed on a Nikon Eclipse E400 microscope (original magnification: ×400) and are representative of three separate experiments. Arrows indicate microinjected cells. (C) Quantitative analysis of microinjection experiments. Data represent the percentage of P-S6–positive cells per total number of microinjected cells; S6 phosphorylation of GFP-microinjected cells was taken as 100%. Data are means ± SE of three separate experiments. *P < 0.001 for GFP-TSC2 or N-TSC2 + C-TSC2 versus GFP by ANOVA (Bonferroni Dunn). (D) Serum-deprived LAM-1/1, ELT3, and ERC15 cells were microinjected with plasmids expressing GFP as a control, GFP–N-TSC2, or GFP–C-TSC2 or co-microinjected with GFP–N-TSC2 and GFP–C-TSC2 followed by BrdU incorporation assay. BrdU incorporation represents the percentage of BrdU-positive cells expressing GFP-tagged constructs compared with the total number of transfected cells. BrdU incorporation of GFP-microinjected cells was taken as 100%. Data represent means ± SE of three separate experiments. *P < 0.001 for N-TSC2 + C-TSC2 versus GFP by ANOVA (Bonferroni Dunn).

Figure 7.

Inhibition of ribosomal protein S6 phosphorylation and cell proliferation require the N- and C-termini of TSC2. Serum-deprived LAM-1/1 (A), ELT3, and ERC15 cells (B) were microinjected with GFP, GFP-TSC2, or co-microinjected with GFP–N-TSC2 and GFP–C-TSC2 followed by immunostaining with anti-phospho-ribosomal protein S6 (P-S6, red) and anti-GFP antibodies (green); yellow fluorescence results from the co-localization of P-S6 and GFP in cells. Images were analyzed on a Nikon Eclipse E400 microscope (original magnification: ×400) and are representative of three separate experiments. Arrows indicate microinjected cells. (C) Quantitative analysis of microinjection experiments. Data represent the percentage of P-S6–positive cells per total number of microinjected cells; S6 phosphorylation of GFP-microinjected cells was taken as 100%. Data are means ± SE of three separate experiments. *P < 0.001 for GFP-TSC2 or N-TSC2 + C-TSC2 versus GFP by ANOVA (Bonferroni Dunn). (D) Serum-deprived LAM-1/1, ELT3, and ERC15 cells were microinjected with plasmids expressing GFP as a control, GFP–N-TSC2, or GFP–C-TSC2 or co-microinjected with GFP–N-TSC2 and GFP–C-TSC2 followed by BrdU incorporation assay. BrdU incorporation represents the percentage of BrdU-positive cells expressing GFP-tagged constructs compared with the total number of transfected cells. BrdU incorporation of GFP-microinjected cells was taken as 100%. Data represent means ± SE of three separate experiments. *P < 0.001 for N-TSC2 + C-TSC2 versus GFP by ANOVA (Bonferroni Dunn).

Figure 7.

Inhibition of ribosomal protein S6 phosphorylation and cell proliferation require the N- and C-termini of TSC2. Serum-deprived LAM-1/1 (A), ELT3, and ERC15 cells (B) were microinjected with GFP, GFP-TSC2, or co-microinjected with GFP–N-TSC2 and GFP–C-TSC2 followed by immunostaining with anti-phospho-ribosomal protein S6 (P-S6, red) and anti-GFP antibodies (green); yellow fluorescence results from the co-localization of P-S6 and GFP in cells. Images were analyzed on a Nikon Eclipse E400 microscope (original magnification: ×400) and are representative of three separate experiments. Arrows indicate microinjected cells. (C) Quantitative analysis of microinjection experiments. Data represent the percentage of P-S6–positive cells per total number of microinjected cells; S6 phosphorylation of GFP-microinjected cells was taken as 100%. Data are means ± SE of three separate experiments. *P < 0.001 for GFP-TSC2 or N-TSC2 + C-TSC2 versus GFP by ANOVA (Bonferroni Dunn). (D) Serum-deprived LAM-1/1, ELT3, and ERC15 cells were microinjected with plasmids expressing GFP as a control, GFP–N-TSC2, or GFP–C-TSC2 or co-microinjected with GFP–N-TSC2 and GFP–C-TSC2 followed by BrdU incorporation assay. BrdU incorporation represents the percentage of BrdU-positive cells expressing GFP-tagged constructs compared with the total number of transfected cells. BrdU incorporation of GFP-microinjected cells was taken as 100%. Data represent means ± SE of three separate experiments. *P < 0.001 for N-TSC2 + C-TSC2 versus GFP by ANOVA (Bonferroni Dunn).

Figure 7.

Inhibition of ribosomal protein S6 phosphorylation and cell proliferation require the N- and C-termini of TSC2. Serum-deprived LAM-1/1 (A), ELT3, and ERC15 cells (B) were microinjected with GFP, GFP-TSC2, or co-microinjected with GFP–N-TSC2 and GFP–C-TSC2 followed by immunostaining with anti-phospho-ribosomal protein S6 (P-S6, red) and anti-GFP antibodies (green); yellow fluorescence results from the co-localization of P-S6 and GFP in cells. Images were analyzed on a Nikon Eclipse E400 microscope (original magnification: ×400) and are representative of three separate experiments. Arrows indicate microinjected cells. (C) Quantitative analysis of microinjection experiments. Data represent the percentage of P-S6–positive cells per total number of microinjected cells; S6 phosphorylation of GFP-microinjected cells was taken as 100%. Data are means ± SE of three separate experiments. *P < 0.001 for GFP-TSC2 or N-TSC2 + C-TSC2 versus GFP by ANOVA (Bonferroni Dunn). (D) Serum-deprived LAM-1/1, ELT3, and ERC15 cells were microinjected with plasmids expressing GFP as a control, GFP–N-TSC2, or GFP–C-TSC2 or co-microinjected with GFP–N-TSC2 and GFP–C-TSC2 followed by BrdU incorporation assay. BrdU incorporation represents the percentage of BrdU-positive cells expressing GFP-tagged constructs compared with the total number of transfected cells. BrdU incorporation of GFP-microinjected cells was taken as 100%. Data represent means ± SE of three separate experiments. *P < 0.001 for N-TSC2 + C-TSC2 versus GFP by ANOVA (Bonferroni Dunn).