Research ArticleCancer

Binding of the sphingolipid S1P to hTERT stabilizes telomerase at the nuclear periphery by allosterically mimicking protein phosphorylation

See allHide authors and affiliations

Science Signaling  16 Jun 2015:
Vol. 8, Issue 381, pp. ra58
DOI: 10.1126/scisignal.aaa4998
  • Fig. 1 SK2-generated S1P regulates hTERT protein abundance and stability.

    (A) Endogenous hTERT protein abundance in A549 cells transfected with SK1 or SK2 siRNA. Scr, scrambled. (B) Immunoprecipitation for FLAG and Western blotting for hTERT in lysates from wild-type (WT), SK1-deficient, or SK2-deficient MEFs expressing FLAG-hTERT or vector (Vec) in the absence or presence of CHX (50 μg/ml; 0 to 4 hours). A densitometry from three blots shown below. (C) Quantification of FLAG pull-down and Western blotting for hTERT in lysates from SK2-deficient MEFs cotransfected with hTERTWT-FLAG and either SK2WT or SK2G212E with or without CHX for 4 hours. a.u., arbitrary units. (D) Effects of ABC294640 (ABC) on 17C-S1P formation in the nuclear (Nuc) and cytoplasmic (Cyto) fractions of A549 measured by liquid chromatography and mass spectrometry. Veh, vehicle. (E) Cytoplasmic and nuclear fractionation of A549 cells were performed after treatment with ABC294640 and analyzed by Western blotting with antibodies against lamin B and calnexin nuclear and cytoplasmic markers, respectively. (F and G) Effects of ABC294640 treatment (80 μM) on hTERT abundance in A549 (F) or H157 and H1650 cells (G) at 0 to 8 hours as measured by Western blotting. In all panels, blots are representative of three independent experiments, and data are means ± SD from three independent experiments.*P < 0.05, **P < 0.01, ***P < 0.001 by Student’s t test.

  • Fig. 2 SK2-generated S1P interacts with hTERT by lipid-protein binding.

    (A) Ectopic expression of hTERTWT-FLAG in A549 cells (right) treated with 17C-sphingosine was measured by immunoprecipitation for the FLAG tag and assessed for bound 17C-S1P (data in fig. S1J). (B) Overlay of the hTERT model with the crystal structure of tcTERT. Distinct domains of TERT are shown in color, and the S1P interacting residue Asp684 located at the interface of the palm and thumb domains is in red stick. TRBD,TERT RBD. (C and D) Binding of biotinylated S1P (B-S1P) to FLAG-tagged WT or mutant hTERT in A549 cells (C) or GM00847 cells (D) assessed by avidin bead pull-down and Western blotting for hTERT. (E) In vitro binding of tcTERT to either S1P [POPC/POPE/S1P (70:20:10)] or LPA [POPC/POPE/LPA (70:20:10)] vesicles relative to the control [POPC/POPE (80:20)]. (F) Kd analysis of tcTERT interaction with S1P vesicles. (G and H) In vitro binding of biotin-tagged S1P to FLAG-tagged WT or D684A mutant hTERT (0.4 mg/ml) partially purified from A549 cells. In all panels, data are means ± SD from three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 by Student’s t test.

  • Fig. 3 Interaction of hTERT with SK2-generated S1P colocalizes with lamin B at the nuclear periphery.

    (A) PLA detection of the subcellular localization of the S1P-hTERT interaction in A549 cells transfected with either control (Scr) or SK2 siRNA (upper) or treated with vehicle or ABC294640 (lower). Nuclei were counterstained with DAPI. Scale bars, 20 μm. DMSO, dimethyl sulfoxide. (B) PLA detection of S1P-hTERT binding in GM00847 cells transfected with vector, hTERTWT-FLAG, or hTERTD684A-FLAG. (C) PLA for S1P binding to hTERT compared to HDAC3 or SET (nuclear proteins) in A549 cells. (D and E) Colocalization of S1P (red) and lamin B (green) in the nucleus as assessed by immunofluorescence confocal microscopy (IF-CM) in WT or SK2-deficient MEFs (D) or A549 cells transfected with control (Scr) or SK2 shRNA (E). Scale bars, 100 μm. In all panels, images are representative of three independent experiments, and data are means ± SD from three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 by Student’s t test.

  • Fig. 4 S1P-hTERT binding prevents the ubiquitination and proteasomal degradation of hTERT.

    (A) Western blotting for hTERT to assess protein stability in A549 cells pretreated with S1P, lactacystin, or vehicle followed by the SK2 inhibitor ABC294640. (B and C) Pull-down for hTERT and Western blotting for ubiquitin (Ub) showing the effects of ABC294640 on the ubiquitination of endogenous hTERT in A549 cells (B) or of FLAG-hTERT in the presence of the protease inhibitor lactacystin in A549 cells expressing HA-Ub (C). (D) Pull-down for FLAG and Western blotting for hTERT showing the stability of FLAG-tagged WT or mutant (D684A) hTERT in the presence or absence of CHX or MG132 in WT MEFs. (E and F) Pull-down for FLAG and Western blotting for hTERT showing the stability of FLAG-hTERTWT expressed in WT (E) or SK2-deficient MEFs (F) in the presence of CHX alone or with MG132. In all panels, blots are representative of three independent experiments, and data are means ± SD from three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 by Student’s t test.

  • Fig. 5 Binding of SK2-generated S1P protects hTERT from MKRN1-mediated degradation.

    (A) MKRN1 knockdown in A549 cells stably transfected with control (Scr) or SK2 shRNA. (B and C) Western blotting for endogenous hTERT stability after MKRN1 knockdown in stable shScr and shSK2 A549 cells treated with CHX (B) or ABC294640 (C) for 4 hours. In (C), bottom blot is actin loading control. (D) Pull-down for FLAG and Western blotting for ubiquitin in WT MEFs coexpressing FLAG-hTERT and either WT or RING mutant (H307E) MKRN1 and pretreated with MG132 for 2 hours. (E) Coimmunoprecipitation of FLAG-hTERT with V5-MKRN1 in A549 cell extracts in the presence of S1P or LPA (5 μM). Samples shown are from the same representative blot but not in contiguous lanes. (F) Coimmunoprecipitation of FLAG-tagged WT or mutant (D684A) hTERT with MKRN1 in A549 cell extracts in the presence of S1P or C3-O-CH3-S1P. (G) Effect of siRNA-mediated MKRN1 knockdown on the anchorage-independent growth on soft agar exhibited by A549 cells stably transfected with SK2 or control (Scr) shRNA. In all panels, blots are representative of three independent experiments, and data are means ± SD from three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 by Student’s t test.

  • Fig. 6 SK2-generated S1P prevents hTERT-MKRN1 interaction by mimicking hTERT phosphorylation at Ser921.

    (A and B) Effects of WT or catalytically inactive (G212E) SK2 on endogenous hTERT-MKRN1 interaction (A) and proliferation on soft agar (B) in A549 cells, either untreated or treated with GA. (C) Colocalization of FLAG-tagged WT or mutant hTERT (red) with lamin B (green) detected by IF-CM in GM00847 cells. Scale bars, 20 μm. (D) PLA detection of the interaction between MKRN1 and FLAG-tagged WT or mutant hTERT in GM00847 cells. (E) Western blotting to assess the stability of WT or mutant hTERT in the presence or absence of CHX. In all panels, images are representative of three independent experiments, and data are means ± SD from three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 by Student’s t test.

  • Fig. 7 SK2-generated S1P-hTERT plays key roles in the control of senescence, telomere damage, and tumor growth.

    (A) β-Gal staining in primary lung fibroblasts (PLFs) [at passages 7 (P7) to P17] that stably coexpress pCDH and control shRNA (shScr) or SK2 and either shScr or shTERT. (B) Telomere damage assessed by the TIF assay (γ-H2AX and TRF2 colocalization) in PLFs at P12, cotransfected as in (A). (C) β-Gal staining in SK2-deficient or WT MEFs at P5. Positive cells were counted from three to four fields. (D and E) Telomere damage as assessed by the TIF assay (D) and senescence by β-gal detection (E) in SK2-deficient MEFs transfected with WT or mutant (D684A) hTERT. (F) Volumes of allograft tumors derived from LLC cells stably transfected with shScr or shSK2 in the flanks of age-matched WT or SK2-deficient mice. (G) Volumes of xenograft tumors derived from A549 cells stably transfected with vector, hTERTWT, or hTERTD684A in mice treated with vehicle or ABC294640 for 21 days. In panels (A) to (E), data are means ± SD from three independent experiments; in (F) and (G), data are means ± SD from four mice each, with each mouse containing two tumors. *P < 0.05, **P < 0.01, ***P < 0.001 by Student’s t test. (H) A model of our findings, revealing a nuclear lysophospholipid-mediated mechanism through which S1P binding to a pin-pointed region in hTERT functions as an allosteric phosphomimetic and stabilizes hTERT. Various ways to reduce S1P abundance or binding may induce the degradation of hTERT and, in turn, the acceleration of telomere damage and senescence in tumors.

Supplementary Materials

  • www.sciencesignaling.org/cgi/content/full/8/381/ra58/DC1

    Fig. S1. Roles of SK1 versus SK2 on hTERT abundance and 17C-S1P binding.

    Fig. S2. S1P-TERT binding is mediated by the C′3-OH of S1P and the hydrophobic region of TERT between the thumb and finger domains.

    Fig. S3. hTERT colocalizes with lamin B at the nuclear periphery.

    Fig. S4. Detection of stably expressed hTERT in MEFs.

    Fig. S5. Effects of S1P-hTERT binding on hTERT-MKRN1 interaction and growth inhibition in response to GA treatment.

    Fig. S6. S1P binding might mimic protein phosphorylation of hTERT at Ser921.

    Fig. S7. Effects of SK2-generated S1P on hTERT-dependent senescence.

    Fig. S8. Regulation of senescence by S1P-hTERT binding in wild-type or SK2-deficient MEFs.

  • Supplementary Materials for:

    Binding of the sphingolipid S1P to hTERT stabilizes telomerase at the nuclear periphery by allosterically mimicking protein phosphorylation

    Shanmugam Panneer Selvam, Ryan M. De Palma, Joshua J. Oaks, Natalia Oleinik, Yuri K. Peterson, Robert V. Stahelin, Emmanuel Skordalakes, Suriyan Ponnusamy, Elizabeth Garrett-Mayer, Charles D. Smith, Besim Ogretmen*

    *Corresponding author. E-mail: ogretmen{at}musc.edu

    This PDF file includes:

    • Fig. S1. Roles of SK1 versus SK2 on hTERT abundance and 17C-S1P binding.
    • Fig. S2. S1P-TERT binding is mediated by the C′3-OH of S1P and the hydrophobic region of TERT between the thumb and finger domains.
    • Fig. S3. hTERT colocalizes with lamin B at the nuclear periphery.
    • Fig. S4. Detection of stably expressed hTERT in MEFs.
    • Fig. S5. Effects of S1P-hTERT binding on hTERT-MKRN1 interaction and growth inhibition in response to GA treatment.
    • Fig. S6. S1P binding might mimic protein phosphorylation of hTERT at Ser921.
    • Fig. S7. Effects of SK2-generated S1P on hTERT-dependent senescence.
    • Fig. S8. Regulation of senescence by S1P-hTERT binding in wild-type or SK2-deficient MEFs.

    [Download PDF]

    Technical Details

    Format: Adobe Acrobat PDF

    Size: 546 KB


    Citation: S. Panneer Selvam, R. M. De Palma, J. J. Oaks, N. Oleinik, Y. K. Peterson, R. V. Stahelin, E. Skordalakes, S. Ponnusamy, E. Garrett-Mayer, C. D. Smith, B. Ogretmen, Binding of the sphingolipid S1P to hTERT stabilizes telomerase at the nuclear periphery by allosterically mimicking protein phosphorylation. Sci. Signal. 8, ra58 (2015).

    © 2015 American Association for the Advancement of Science

Stay Connected to Science Signaling

Navigate This Article