Ceramide kinase promotes tumor cell survival and mammary tumor recurrence

Abstract

Recurrent breast cancer is typically an incurable disease and, as such, is disproportionately responsible for deaths from this disease. Recurrent breast cancers arise from the pool of disseminated tumor cells (DTC) that survive adjuvant or neoadjuvant therapy, and patients with detectable DTCs following therapy are at substantially increased risk for recurrence. Consequently, the identification of pathways that contribute to the survival of breast cancer cells following therapy could aid in the development of more effective therapies that decrease the burden of residual disease and thereby reduce the risk of breast cancer recurrence. We now report that ceramide kinase (Cerk) is required for mammary tumor recurrence following HER2/neu pathway inhibition and is spontaneously upregulated during tumor recurrence in multiple genetically engineered mouse models for breast cancer. We find that Cerk is rapidly upregulated in tumor cells following HER2/neu downregulation or treatment with Adriamycin and that Cerk is required for tumor cell survival following HER2/neu downregulation. Consistent with our observations in mouse models, analysis of gene expression profiles from more than 2,200 patients revealed that elevated CERK expression is associated with an increased risk of recurrence in women with breast cancer. In addition, although CERK expression is associated with aggressive subtypes of breast cancer, including those that are estrogen receptor-negative, HER2(+), basal-like, or high grade, its association with poor clinical outcome is independent of these clinicopathologic variables. Together, our findings identify a functional role for Cerk in breast cancer recurrence and suggest the clinical utility of agents targeted against this prosurvival pathway.

Figure 1. Cerk is spontaneously up-regulated in recurrent mammary tumors and following acute HER2/neu down-regulation

(A) qRT-PCR analysis of Cerk mRNA expression in primary and recurrent HER2/neu, MYC, Wnt1, and Akt1-driven tumors. (B) qRT-PCR analysis of Cerk mRNA expression from MTB/TAN tumors at primary, 48 hr deinduced, 96 hr deinduced, and recurrent time points. (C) qRT-PCR analysis of Cerk mRNA expression in primary MTB/TAN tumor cell line in vitro removed from doxycycline 1 or 6 days. (D) qRT-PCR analysis of CERK mRNA expression in BT474 cells and (E) SKBR3 cells treated with 100 nM Lapatinib or DMSO for 1, 2 or 3 days. qRT-PCR analysis of CERK mRNA expression in (F) BT20 cells and (G) primary MTB/TAN cells treated with 10 µM Adriamycin or vehicle for 4, 8, 16 or 24 hr. *p<0.05, **p<0.01, ***p<0.001.

Figure 2. Cerk protects tumor cells from apoptosis upon HER2/neu down-regulation in vitro

(A) Representative fluorescence images and (B) quantification of MTB/TAN primary tumor cell lines expressing empty vector, Cerk or G198D-Cerk withdrawn from doxycycline 72 hr then stained for cleaved caspase-3 (red; Hoechst = blue). (C) Representative fluorescence images and (D) quantification of MTB/TAN primary tumor cell lines expressing empty vector, a control scrambled shRNA, shCerk1 or shCerk2 withdrawn from doxycycline 72 hr then stained for cleaved caspase-3. (E) Western blot of MTB/TAN primary tumor cell lines expressing empty vector, Cerk or G198D-Cerk withdrawn from doxycycline 48 hr, blotted for cleaved PARP, quantified in (F), and cleaved caspase-3, quantified in (G). (H) Western blot of MTB/TAN primary tumor cell lines expressing empty vector, a control scrambled shRNA, or shCerk2 withdrawn from doxycycline 48 hr, blotted for cleaved PARP, quantified in (I), and cleaved caspase-3, quantified in (J).

Figure 3. Cerk protects tumor cells from apoptosis upon HER2/neu down-regulation in vivo

(A) Schematic of competition assay and timing of doxycycline treatment and tumor harvest. (B) Percentage of eGFP-positive and mCherry-positive cells was determined at timepoints in (A) up to 28 days post-deinduction. Data represent mean ± SEM. (C) Representative fluorescence images of primary tumors, 96 hr deinduced tumors and residual lesions 28 days post-deinduction. (D) Quantification of cleaved caspase-3 staining in primary orthotopic empty vector, Cerk and G198D-Cerk tumors at 48 and 96 hr post-deinduction. (E) Representative fluorescence images of primary tumors, 48 hr and 96 hr deinduced tumors stained for cleaved caspase-3 (red; Hoechst = blue).

Figure 4. Cerk promotes mammary tumor recurrence in the MTB/TAN mouse model

(A) Schematic of recurrence assay and timing of doxycycline treatment. (B) Recurrence-free survival of nu/nu mice harboring primary orthotopic tumors from MTB/TAN primary cell lines expressing empty vector or Cerk induced to regress by doxycycline withdrawal. (C) Mean growth rate of control and Cerk overexpressing recurrent tumors. (D) Recurrence-free survival of nu/nu mice harboring primary orthotopic tumors from MTB/TAN cell lines expressing empty vector, wildtype Cerk or the kinase dead mutant G198D-Cerk induced to regress by doxycycline withdrawal. (E) Schematic of modified recurrence assay with no primary tumor formation. (F) Recurrence-free survival of nu/nu mice harboring MTB/TAN primary cell lines expressing empty vector, Cerk or G198D-Cerk withdrawn from doxycycline 48 hr post-injection. (G) Mean growth rate of control and Cerk overexpressing recurrent tumors.

Figure 5. High CERK expression predicts decreased relapse-free survival in women with breast cancer

Forest plot representation of 5-year survival estimates and hazard ratios for relapse-free survival of individual datasets of human breast cancer patients.