Dihydroceramide desaturase 1 (DES1) promotes anchorage-independent survival downstream of HER2-driven glucose uptake and metabolism

Abstract

Oncogenic reprogramming of cellular metabolism is a hallmark of many cancers, but our mechanistic understanding of how such dysregulation is linked to tumor behavior remains poor. In this study, we have identified dihydroceramide desaturase (DES1)-which catalyzes the last step in de novo sphingolipid synthesis-as necessary for the acquisition of anchorage-independent survival (AIS), a key cancer enabling biology, and establish DES1 as a downstream effector of HER2-driven glucose uptake and metabolism. We further show that DES1 is sufficient to drive AIS and in vitro tumorigenicity and that increased DES1 levels-found in a third of HER2+ breast cancers-are associated with worse survival outcomes. Taken together, our findings reveal a novel pro-tumor role for DES1 as a transducer of HER2-driven glucose metabolic signals and provide evidence that targeting DES1 is an effective approach for overcoming AIS. Results further suggest that DES1 may have utility as a biomarker of aggressive and metastasis-prone HER2+ breast cancer.

Keywords: breast cancer; cell survival; metastasis; oncogene; sphingolipid.

FIGURE 1

ECM detachment induces SL accumulation in breast epithelial cells but not in HER2+ BC cells, in part through differential DES activity. (A) MCF10A and SKBR3 cells were cultured in monolayer or suspension for 48 h. Cell death was determined by trypan blue exclusion assay. (B) MCF10A and SKBR3 cells were cultured in monolayer (A) or suspension (S) for 24 h. Whole cell lysates were analyzed by SDS-PAGE and western blot for the proteins shown with β-Actin as the loading control. (C) MCF10A and SKBR3 cells were cultured as in (B). Extracted lipids were analyzed by tandem LC/MS mass spectrometry and normalized to total lipid phosphate. (D) Isogenic 10A-Vec and 10A-NeuT cells were cultured as in (B) and analyzed as in (C). All data are representative of at least three independent experiments and reported as mean ± SEM (*p < .05, **p < .02, ***p < .01).

FIGURE 2

Deregulation of SL levels in ECM-detached HER2+ BC cells is in part through differential DES1 activity. (A) Schematic of the de novo SL pathway. (B) MCF10A (left) and SKBR3 (right) cells were cultured as in (B) and labeled with C17-dhSph for the indicated times. C17-Cer was extracted and analyzed as in (C). (C) MCF10A and SKBR3 cells were cultured as in (B). In situ DES activity was measured by addition of C12-dhPPS for 1 h. Extracted lipids were analyzed by tandem LC/MS. (D) MCF10A, HMEC, and SKBR3 were cultured for 48 h. DES1 and DES2 expression was measured by qRT-PCR using actin as a reference gene. (E) MCF10A and SKBR3 cells were treated with negative control (si-AS), DES1 (si-D1_A, si-D1_B) or DES2 (si-D2) for 48 h. DES1 and DES2 expression was measured as in (D). (F) MCF10A and SKBR3 cells were treated as in (E). In situ DES activity was analyzed as in (C). All data are representative of at least three independent experiments and reported as mean ± SEM (*p < .05, **p < .02, ***p < .01).

FIGURE 3

DES1 is required for in vitro AIS and colony formation. (A) SKBR3 cells were cultured in monolayer for 48 h with DMSO or 4HPR (5 μM). In situ DES activity was assessed using C12-dhPPS. (B) SKBR3 cells were cultured as in (A). Extracted lipids were analyzed by tandem LC/MS mass spectrometry and normalized to total lipid phosphate. (C) SKBR3 cells were cultured in monolayer (Att) and suspension (Susp) for 48 h with DMSO or 4HPR (5 μM). Cell death was determined by trypan blue exclusion assay. (D) SKBR3 cells were treated with negative control (si-AS), DES1 (si-D1_A, si-D1_B) or DES2 (si-D2) prior to culture in suspension for 48 h. DES1 levels were analyzed by SDS-PAGE and western blot with GAPDH as the loading control. (E) SKBR3 cells were cultured as in (D). Lipid levels were analyzed as in (B). (F) SKBR3 cells were treated with siRNA prior to culture in monolayer (Att) and suspension (Susp) for 48 h. Cell death was analyzed as in (C). (G) SKBR3 control (Vec) or DES1 Crispr cells (ΔD1_A, ΔD1_B) were plated in monolayer for 24 h and in situ DES activity measured as in (A). Inset—DES1 protein levels were analyzed as in (D). (H) SKBR3 Crispr cells were cultured and analyzed for lipids as in (B). (I) SKBR3 Crispr cells were cultured in monolayer (Att) and suspension (Susp) for 48 h. Cell death was analyzed as in (C). (J) SKBR3 cells were plated in soft agar with DMSO or 4HPR (5 μM) for 12–14d. Colonies were stained and visualized using an EVOS microscope. (K) SKBR3 Crispr cells were plated in soft agar for 12–14 days. Colonies were analyzed as in (J). All data are representative of at least three independent experiments and reported as mean ± SEM (*p < .05, **p < .02, ***p < .01).

FIGURE 4

Decreased AIS following DES1 loss is not through activation of apoptotic pathways. (A) SKBR3 cells were treated with negative control (si-AS), DES1 (si-D1_A, si-D1_B) or DES2 (si-D2) prior to culture in suspension for 48 h. Protein was extracted and analyzed by SDS-PAGE and immunoblot for cleaved PARP using GAPDH as loading control. (B) SKBR3 cells were cultured as in (A). Caspase 3/7 activity was assayed using a commercially available kit and normalized to protein. (C) SKBR3 control (Vec) or DES1 Crispr cells (ΔD1_A, ΔD1_B) were cultured in monolayer (Att) and suspension (Susp) for 48 h. Cleaved PARP and GAPDH were analyzed by SDS-PAGE as in (A). (D) SKBR3 Crispr cells were cultured as in (A). Caspase 3/7 activity was assayed as in (B).

FIGURE 5

HER2 maintains post-translational DES1 activity through PI3K signaling. (A) MCF10A, HMEC, and SKBR3 cells were cultured in monolayer (Att) or suspension (Susp) for 24 h. DES1 levels were analyzed by SDS-PAGE and western blot with GAPDH as loading control. (B) MCF10A, HMEC, and SKBR3 cells were cultured in suspension for 24 h. DES1 expression was analyzed by qRT-PCR using actin as reference gene. (C) MCF10A cells were culture in monolayer or suspension with Veh (DMSO), MG132 (1 μM) or bortezomib (5 nM) for 24 h. In situ DES activity was probed using C12-dhCCPS with substrate and product levels analyzed by tandem LC/MS. (D) SKBR3 cells were cultured in suspension for 24 h with DMSO or Lap (1 μM). In situ DES activity was assessed as in (C). Whole cell lysates were analyzed for the proteins shown as in (A). (E) SKBR3 cells were treated with negative control (si-AS) or HER2 (si-HER2) siRNA prior to culture in suspension for 24 h. In situ DES activity was assessed as in (C). Whole cell lysates were analyzed for the proteins shown as in (D). (F) Isogenic 10A-Vec and 10A-NeuT cells were cultured as in (B). In situ DES activity was assessed as in (C). Whole cell lysates were analyzed for the proteins shown as in (D). (G) SKBR3 cells were cultured in suspension for 24 h with vehicle (DMSO), LY (10 μM), Wort (2 μM), PD (10 μM) or U0 (2 μM). In situ DES activity was assessed as in (C). Whole cell lysates were analyzed as in (D). (H) Isogenic 10A-Vec and 10A-NeuT were cultured in suspension for 24 h with the inhibitors shown. In situ DES activity was assessed as in (C). Whole cell lysates were analyzed as in (D). (I) Schematic of results linking PI3K but not ERK to HER2-driven DES1 activity. All data are representative of at least three independent experiments and reported as mean ± SEM (**p < .02, ***p < .01).

FIGURE 6

HER2-driven DES1 activity requires glucose uptake and metabolism. (A) SKBR3 (left), 10A-Vec/NeuT (right) cells were cultured in suspension (Susp) with DMSO or WZB117 (20 μM) for 24 h. In situ DES activity was probed using C12-dhCCPS with substrate and product levels analyzed by tandem LC/MS. (B) SKBR3 (left), 10A-Vec/NeuT (right) cells were cultured as in (A) with DMSO or BAY (1 μM). In situ DES activity was analyzed as in (A). (C) SKBR3 cells were treated with negative control (si-AS) or GLUT1 (si-GLUT1) siRNA prior to culture in suspension for 24 h. In situ DES activity was assessed as in (A). GLUT1 mRNA was assessed by qRT-PCR using GAPDH as reference gene. (D) 10A-Vec and 10A-NeuT were cultured and analyzed as in (C). (E) SKBR3 (left), 10A-Vec/NeuT (right) cells were cultured as in (A) with vehicle (dH2O) or 2-DG (20 mM). In situ DES activity was analyzed as in (A). (F) SKBR3 (left), 10A-Vec/NeuT (right) cells were cultured as in (A) with DMSO or 3-BP (25 μM). In situ DES activity was analyzed as in (A). All data are representative of at least three independent experiments and reported as mean ± SEM (*p < .05, **p < .02, ***p < .01).

FIGURE 7

High DES1 levels are associated with worse outcomes and increased tumorigenicity in HER2+ BC. (A) DES1 alterations in total and HER2+ breast cancers of the TCGA dataset. (B) Analysis of RFS, OS, and DMFS in HER2+ breast cancers from www.kmplot.com stratified by high and low DES1 expression. (C) Whole cell lysates of 10A-Vec and 10A-DES1 cells probed as shown with GAPDH as loading control. (D) 10A-Vec and 10A-DES1 cells were cultured in monolayer for 24 h. In situ DES activity was probed with C12-dhCCPS, with substrate and product levels analyzed by tandem LC/MS mass spectrometry. (E) 10A-Vec and 10A-DES1, and -DES2 cells were cultured in monolayer (Att) or suspension (Susp) for 24 h. Levels of dhCer were analyzed by tandem LC/MS mass spectrometry. (F) 10A-Vec and 10A-DES1 cells were cultured in monolayer (Att) or suspension (Susp) for 48 h. Cell death was determined by trypan blue exclusion assay. (G) 10A-Vec and 10A-DES1 cells were cultured in 3D for 8–10d. Cultures were stained with anti-laminin (red) and DAPI (blue) and visualized by confocal microscopy. (H) Whole cell lysates of 10A-Vec/Lac, 10A-Vec/DES1, 10A-NeuT/Lac, and 10A-NeuT/DES1 cells were probed as shown with GAPDH as loading control. (I) 10A-Vec/Lac, 10A-Vec/DES1, 10A-NeuT/Lac, and 10A-NeuT/DES1 were cultured and analyzed as in (F). (J) 10A-Vec/Lac, 10A-Vec/DES1, 10A-NeuT/Lac, and 10A-NeuT/DES1 were cultured for 6–8d and stained as in (G). (K) (left) Whole cell lysates of SKB-Vec and -DES1 cells were probed as shown with GAPDH as loading control; (right) Cells were cultured in monolayer for 24 h and in situ DES activity probed as in (E). (L) SKB-Vec and SKB-DES1 cells were plated in soft agar for 6–8d. Colonies were stained and visualized using an EVOS microscope. All data are representative of at least three independent experiments and reported as mean ± SEM (*p < .05, **p < .02, ***p < .01).