Neuroblastoma is the most common extracranial solid malignancy in the pediatric population, accounting for over 9% of all cancer-related deaths in children. Autophagy is a cell self-protective mechanism that promotes tumor cell growth and survival, making it an attractive target for treating cancer. However, the role of autophagy in neuroblastoma tumor growth and metastasis is largely undefined. Here we demonstrate that targeted inhibition of an essential autophagy kinase, unc-51 like autophagy kinase 1 (ULK1), with a recently developed small-molecule inhibitor of ULK1, SBI-0206965, significantly reduces cell growth and promotes apoptosis in SK-N-AS, SH-SY5Y, and SK-N-DZ neuroblastoma cell lines. Furthermore, inhibition of ULK1 by a dominant-negative mutant of ULK1 (dnULK1K46N) significantly reduces growth and metastatic disease and prolongs survival of mice bearing SK-N-AS xenograft tumors. We also show that SBI-0206965 sensitizes SK-N-AS cells to TRAIL treatment, but not to mTOR inhibitors (INK128, Torin1) or topoisomerase inhibitors (doxorubicin, topotecan). Collectively, these findings demonstrate that ULK1 is a viable drug target and suggest that inhibitors of ULK1 may provide a novel therapeutic option for the treatment of neuroblastoma. Mol Cancer Ther; 17(11); 2365–76. ©2018 AACR.
Neuroblastoma is a cancer of the primordial neural crest cells, which give rise to the sympathetic nervous system, and generally occurs in infants and young children. Neuroblastoma is the most common extracranial solid tumor in the pediatric population, accounting for 7% to 10% of all pediatric cancers and over 9% of all cancer-related deaths in children (1). Clinically, tumors can occur anywhere along the sympathetic nervous system and symptoms can develop from compression of vital structures, including the spinal column, from tumors that arise in the parasympathetic ganglia. However, the majority of tumors arise in the adrenal gland, which can progress to high-stage tumors that infiltrate local organ structures and metastasize to lymph nodes, bone marrow, and the liver. The behavior of neuroblastoma varies widely from spontaneous remission to aggressive metastatic disease. Several prognostic factors are used to stratify neuroblastoma into very low-risk, low-risk, intermediate-risk, and high-risk classes (1, 2). These factors include the age at diagnosis, DNA index (ploidy), MYCN amplification, and tumor stage, which is based upon the surgical–pathologic International Neuroblastoma Staging System (INSS; refs. 1–3). Distant metastases are detected in approximately 50% of patients at diagnosis, and most commonly occur in the bone, bone marrow, lymph nodes, and liver (1). Therapeutic strategies for high-risk neuroblastoma are aggressive, consisting of multiagent chemotherapy, surgery, radiation, and myeloablative chemotherapy regiments with subsequent autologous bone marrow transplant (2). However, despite these treatments, over half of patients with high-risk neuroblastoma will relapse (4), and the 5-year overall survival rate is only 50% (5). Thus, there is an urgent need for new therapeutic options, particularly for patients with high-risk neuroblastoma and those with recurrent or relapsed neuroblastoma.
Macroautophagy (hereafter referred to as autophagy) is an evolutionarily-conserved degradation process that maintains cellular homeostasis through regular turnover of dysfunctional proteins and organelles (6, 7). Autophagy is upregulated by cellular stressors such as nutrient deprivation and provides an important cell survival mechanism by facilitating the recycling of essential nutrients, preventing the accumulation of misfolded proteins and reactive oxygen species (ROS), maintaining organelle function, and regulating intracellular signaling pathways (7). Importantly, autophagy can support established tumors by generating metabolic fuel and reducing oxidative stress (7, 8), and inhibition of autophagy can reduce growth and metastasis in certain tumor types (9), such as in pancreatic and lung cancer (10–13). This role of autophagy in cancer survival has made it a popular therapeutic target during the development of new anticancer agents (14, 15). The limited research available indicates that autophagy promotes resistance to several chemotherapeutic agents used to treat neuroblastoma, such as vincristine and doxorubicin, among others (16–18). However, the role of autophagy in neuroblastoma tumor growth and metastasis is still unknown.
A major challenge for targeting autophagy in any cancer is the lack of potent and selective pharmaceutical autophagy inhibitors. The efforts to pharmacologically target autophagy have relied heavily on bafilomycin A1, chloroquine, and hydroxychloroquine, which target not only autophagy flux, but also the endolysosomal pathway by inhibiting late endosome and lysosome function. Recently, a small-molecule kinase inhibitor called SBI-0206965 has been developed that targets an essential autophagy kinase, unc-51 like autophagy kinase 1 (ULK1) (19). The autophagy process is regulated by more than 30 autophagy-related genes (ATG) through four discrete steps: (i) the initiation of autophagosome biogenesis, (ii) nucleation of the phagophore, (iii) the expansion of the phagophore to a mature autophagosome, and (iv) the fusion of the autophagosome to the lysosome (6, 7). Importantly, initiation of autophagy is predominantly mediated by ULK1, a mammalian homolog of yeast Atg1 that is the only serine/threonine kinase among the ATG proteins (20). ULK1 forms a complex with multiple regulatory subunits including ATG13 and FIP200, a mammalian counterpart of yeast Atg17. This complex is negatively regulated by the mTOR complex 1 (mTORC1) through hyperphosphorylation of ULK1 in nutrient-rich environments, while AMP-activated protein kinase (AMPK) binds and phosphorylates ULK1 to activate autophagy under conditions of energy scarcity (21, 22). Upon activation, ULK1 induces the next step of autophagy, nucleation of the immature autophagosome, by phosphorylating the downstream BECN1 complex. Thus, ULK1 is an upstream kinase that regulates the initiation of autophagy, making it an excellent drug target to inhibit autophagy.
SBI-0206965 has been previously characterized as a potent and selective inhibitor of ULK1, exhibiting an IC50 of ULK1 kinase activity at 108 nmol/L (19). Importantly, SBI-0206965 was demonstrated to synergize with nutrient deprivation and mTOR inhibitors to induce apoptosis in several cell types (19). Although still in the early stages of development, SBI-0206965 provides new opportunity for pharmacologic inhibition of autophagy that can potentially translate to the clinic. In this study, we establish the important role of ULK1 and autophagy in neuroblastoma growth and response to chemotherapy. We determine that SBI-0206965 exhibits cytotoxicity across multiple neuroblastoma cell lines and that genetic inhibition of ULK1 significantly reduces neuroblastoma tumor growth and metastasis in vivo. These results demonstrate that ULK1-mediated autophagy plays a key role in neuroblastoma tumor growth and progression and that ULK1 is a viable drug target for the treatment of neuroblastoma.
Materials and Methods
Human cell lines SK-N-AS (ATCC CRL-2137), SK-N-DZ (ATCC CRL-2149), SH-SY5Y (ATCC CRL-2266), and A549 (ATCC CCL-185) were purchased from ATCC. Cells were cultured in DMEM/Ham's F-12 50/50 Mix containing l-glutamine and 15 mmol/L HEPES supplemented with 10% (v/v) heat-inactivated FBS and 1% antibiotic and antimycotic solution at 37°C in a humidified incubator with 5% CO2. All cell lines were passaged in our laboratory for fewer than 6 months before use and periodically authenticated by morphologic inspection and Mycoplasma testing. Where indicated, cells were starved in amino acid- and FBS-deficient medium (Wako, catalog no. 048-33575).
The ULK1 inhibitor SBI-0206965 was purchased from Xcessbio Biosciences (catalog no. M60268-2) and was dissolved in DMSO. Doxorubicin was obtained from Selleck Chemicals (catalog no. S1208), topotecan from Alexis Biochemicals (catalog no. 350-133-M001), INK128 from Active Biochem (catalog no. A-1023), Torin1 from Thermo Fisher Scientific (catalog no. NC0418592), and TRAIL from vendor VWR (catalog no. 10787-196). Subsequent dilutions of stock solutions of compounds were made in culture media just before use. In all experiments, the final concentration of DMSO did not exceed 0.1% (v/v), a concentration that is nontoxic to the cells.
Xenograft mouse model of neuroblastoma
All animal studies were performed according to the guidelines established by the Institutional Animal Care and Use Committee (IACUC) at the Penn State College of Medicine (Hershey, PA). A xenograft neuroblastoma model was generated by injecting 4.0 × 106 SK-N-AS cells into the subcutaneous tissue in the flank of NOD SCID Gamma (NSG; Jackson Laboratory, catalog no. 005557) male and female mice, aged 6–8 weeks. Cells stably expressing a firefly luciferase gene (luc2) were injected in a 50:50 mixture of PBS and Matrigel basement membrane matrix (Thermo Fisher Scientific, catalog no. CB-40234). Mice were imaged for luciferase expression on a weekly basis for 5 weeks via a Xenogen IVIS bioluminescent imager. Mice were injected with 5 μL/gram body weight of 30mg/mL d-Luciferin (Gold Biotechnology, catalog no. LUCK-1G) in PBS, 5 minutes prior to imaging. Photon flux was calculated using region of interest (ROI) measurements of either the primary tumor site, or the ventral thoracic area for lung metastasis. Tumor volume was also measured using calipers and calculated as (length2 × width)/2. At the experimental endpoint, mice were euthanized and tumors were harvested for ex vivo analysis and subsequent histology.
Tail vein injection metastasis assay
A mouse model of neuroblastoma metastasis was created by injecting 5.0 × 105 SK-N-AS cells stably expressing the luc2 gene into the tail vein of 6- to 8-week-old male NSG mice. Mice were imaged on the ventral side for luciferase expression on a weekly basis for a total of 8 weeks via Xenogen IVIS imaging, as described above. At the experimental endpoint, mice were euthanized and tumor, bone, and liver tissues were harvested for ex vivo examination and subsequent histologic analysis. Lung, abdominal lymph nodes, and other abdominal organs were examined for metastatic lesions, but none were found.
Treated and untreated cells were lysed in RIPA lysis buffer containing protease and phosphatase inhibitors and subjected to immunoblotting with primary antibodies: Flag (Sigma-Aldrich, catalog no. F1804); p62 (American Research Products, catalog no. 03-GP62-C); β-actin (Sigma-Aldrich, catalog no. A5441); LC3 (Novus Biologicals, NB100-2220), PARP (Cell Signaling Technology, catalog no. 9542), cleaved caspase-3 (Cell Signaling Technology, catalog no. 9661), Phospho-AKT (Ser473; Cell Signaling Technology, catalog no. 4058), phospho-4EBP1 (Thr37/46; Cell Signaling Technology, catalog no. 2855), 4EBP1 (Cell Signaling Technology, catalog no. 9644), AKT (Cell Signaling Technology, catalog no. 4691), ULK1 (Cell Signaling Technology, catalog no. 8054S) followed by fluorophore-conjugated secondary antibodies and detection with a LI-COR Odyssey CLx Imager.
Tumor and liver tissues were harvested and fixed in 10% formalin and paraffin embedded. Sections were deparaffinized, hydrated, and boiled in citrate buffer for antigen retrieval. VECTASTAIN Elite ABC Kit (catalog no. PK-6101) was used for Ki67 (Novus Biologicals, catalog no. NB500-170) and cleaved caspase-3 (Cell Signaling Technology, catalog no. 9661) staining. Ki67 scoring was performed blind by two independent observers. Liver sections were stained with hematoxylin and eosin.
Viability and apoptosis assays
Cell viability was measured by PrestoBlue Cell Viability Reagent (Invitrogen, catalog no. A-13262) as per the manufacturer's instructions. Caspase-8 and caspase-3/7 activities were measured using Caspase-Glo 8 (Promega, catalog no. G8200) and Caspase-Glo 3/7 (Promega, catalog no. G8090) Assay Systems, respectively. For flow cytometry cell death analyses, treated cells were washed once with PBS and stained with 5% APC Annexin-V and 7-AAD (BioLegend, catalog no. 640941) for 15 minutes before flow cytometry analyses. All data were normalized to their nontreated controls.
Data were analyzed using Graph Pad Prism and statistical software SAS version 9.4 (SAS Institute, Cary, NC, USA). Student t-test (two-tailed) was used for single comparisons. Group differences were evaluated using ANOVA or repeated-measure ANOVA models. Data were considered statistically significant when P < 0.05.
SBI-0206965 reduces cell growth and promotes apoptosis in neuroblastoma cell lines
To validate that the ULK1 inhibitor, SBI-0206965 (Fig. 1A), suppresses autophagy in neuroblastoma, we assessed autophagic flux in SK-N-AS cells treated with SBI-0206965 compared with DMSO control. Total ULK1 protein level was reduced in SK-N-AS cells treated with SBI-0206965 compared with control cells treated with DMSO, particularly in starvation conditions, as reported previously (19). Furthermore, as expected, treatment with SBI-0206965 resulted in an accumulation of p62, an autophagy substrate that is widely used as a reporter of autophagic degradation (23), in both complete and starvation medium (Fig. 1B). SBI-0206965–treated SK-N-AS cells also displayed a decrease in LC3-II accumulation in the presence of bafilomycin A1, which blocks lysosomal degradation and the autophagosome-lysosome fusion, under starvation conditions (Fig. 1B). Together, these results demonstrate that SBI-0206965 reduces LC3 lipidation and autophagic flux in SK-N-AS cells. To determine whether pharmacologic inhibition of ULK1 exhibits cytotoxic effects in neuroblastoma, we treated three different neuroblastoma cell lines (SK-N-AS, SH-SY5Y, and SK-N-DZ) with SBI-0206965 (Fig. 1C). Western blot analysis of SBI-0206965–treated neuroblastoma cells revealed an accumulation of p62, verifying that SBI-0206965 inhibited autophagy in all three cell lines (Fig. 1C). Importantly, SBI-0206965 treatment increased PARP and caspase-3 cleavage in all neuroblastoma cell lines, indicating that SBI-0206965 promoted apoptosis (Fig. 1C). The influence of SBI-0206965 treatment on cell growth compared with DMSO control was measured over a 72-hour time course. In normal cell culture conditions, treatment with SBI-0206965 significantly reduced cell growth (Fig. 1D–F). As autophagy is known to promote survival under starvation conditions, the neuroblastoma cell lines were also treated with SBI-0206965 in nutrient-deprived culture medium, which further enhanced the cytotoxicity of SBI-0206965 (Fig. 1G–I). Furthermore, flow cytometry analysis of SBI-0206965–treated SK-N-AS cells displayed an increase in annexin-V staining compared with cells treated with DMSO control (Fig. 1J and K). This effect was further increased under nutrient deprivation, where SBI-0206965 treatment significantly enhanced cell death beyond starvation-induced apoptosis, as reported previously (ref. 19; Fig. 1J and K). Similarly, both SH-SY5Y and SK-N-DZ cells also displayed increased annexin-V staining upon SBI-0206965 treatment compared with DMSO control (Fig. 1L). Together, these data indicate that SBI-0206965 inhibits autophagy and promotes apoptosis in multiple neuroblastoma cell lines.
Inhibition of ULK1 by dnULK1 promotes apoptosis in SK-N-AS cells
To further validate ULK1 as a drug target for neuroblastoma, we generated SK-N-AS cells that express a kinase-dead dominant-negative ULK1K46N gene (dnULK1), as well as control SK-N-AS cells expressing an empty vector (empty). Expression of dnULK1 resulted in an accumulation of p62, indicating inhibition of autophagy (Fig. 2A). Although SBI-0206965–treated cells exhibited higher levels of apoptosis in complete medium conditions, both SBI-0206965 and dnULK1 promoted apoptosis compared with control, as evident by increased cleaved caspase-3 and cleaved PARP (Fig. 2A). In agreement, the enzymatic activity of caspase-3/7 and caspase-8 increased in both dnULK1-expressing cells and SBI-0206965–treated cells compared with control, which was particularly evident in starvation conditions (Fig. 2B and C). Furthermore, annexin-V staining was increased in dnULK1-expressing SK-N-AS cells compared with empty vector control (Fig. 2D and E). Together, these data demonstrate that inhibition of ULK1 by either dnULK1 or SBI-0206965 promotes apoptosis in SK-N-AS cells, validating ULK1 as an attractive target for neuroblastoma treatment.
Inhibition of ULK1 by dnULK1 reduces SK-N-AS xenograft tumor growth and promotes apoptosis
The pharmacologic properties of SBI-0206965 in animal models are currently unclear and require further development. Thus, to determine whether ULK1 is a viable drug target for neuroblastoma and to determine the effects of ULK1 kinase on neuroblastoma tumor growth, we utilized a genetic approach to inhibit ULK1 in neuroblastoma xenograft mouse models. Specifically, SK-N-AS cells stably expressing dnULK1 or empty vector were injected into the subcutaneous tissue of NSG mice. A luciferase reporter gene (luc2) was also introduced into the cells to allow for noninvasive monitoring of tumor growth. Weekly measurements of the primary tumor site in the flank revealed that the dnULK1 xenografts grew significantly slower compared with the control empty vector xenografts (Fig. 3A–C). Consistently, ex vivo assessment of both tumor weight and volume showed a significant reduction in the dnULK1-expressing group after 4 weeks of growth (Fig. 3D and E). Importantly, primary tumors expressing dnULK1 contained higher levels of p62, indicating that autophagy was suppressed in vivo (Fig. 3F). Furthermore, dnULK1-expressing tumors displayed an increased cleavage of PARP and caspase-3, indicating that dnULK1 promoted cell death in vivo (Fig. 3F). An increase in cleaved caspase-3 was also detected by IHC in dnULK1 tumors compared with comparable areas in control tumors (Fig. 3G). Interestingly, for reasons that are not currently clear, only an increase in the 19-kDa fragment of caspase-3 was detected in tumor lysate by Western blot analysis, whereas the level of the 17-kDa fragment remained similar between groups. Similarly, although cl-PARP 89 kDa is slightly increased in dnULK1 SK-N-AS xenografts, we observed significant increases in the 55 kDa and 42 kDa PARP cleavage fragments. These fragments are generated from PARP cleavage by lysosomal proteases, such as cathepsin B and D, and are indicative of necrosis (24, 25). Thus, loss of ULK1 function may increase both apoptosis and tumor necrosis in vivo. In addition, dnULK1-expressing tumors demonstrated a significant reduction in Ki67-positive cells, indicating reduced cell proliferation in vivo. Together, these data indicate that the loss of ULK1 function promotes tumor cell death and reduces tumor growth.
Inhibition of ULK1 by dnULK1 reduces metastatic tumor growth and improves overall survival in an experimental metastasis mouse model of neuroblastoma
As metastasis is detected in approximately 50% of patients at diagnosis and is generally associated with worse prognosis (1), we examined the effects of ULK1 inhibition in a mouse model of neuroblastoma metastasis. Here, SK-N-AS cells expressing luc2 and dnULK1 or empty vector were injected into the bloodstream of mice via the tail vein and metastatic burden, which occurred primarily in the liver, was monitored by luciferase expression over 8 weeks. Importantly, mice receiving dnULK1-expressing SK-N-AS cells survived significantly longer compared with control (Fig. 4A) and exhibited a significant reduction in growth at the metastatic site (i.e. liver; Fig. 4B and C). Interestingly, the pattern of tumor growth in the liver was drastically different between the two groups (Fig. 4D). At the time of sacrifice, the livers in the empty vector group were enlarged, displayed large fluid-filled cysts, and lacked defined solid tumor masses (Fig. 4D and E). Comparatively, livers in the dnULK1-expressing SK-N-AS group were smaller with solid tumor masses and contained significantly fewer cysts (Fig. 4E and F). At this point, it is unclear how engrafted empty-vector SK-N-AS cells cause the cystic liver phenotype and why inhibition of ULK1 kinase significantly reduces this effect.
Evasion of anoikis, a form of apoptotic cell death that occurs when cells experience prolonged detachment from the extracellular matrix (26), contributes to circulating tumor cell (CTC) survival and metastasis (9, 27). As autophagy is known to promote anoikis resistance, we tested whether inhibition of ULK1 sensitizes neuroblastoma cells to anoikis. Indeed, both dnULK1 and SBI-0206965 treatment enhanced caspase-3/7 and caspase-8 activity in suspended SK-N-AS cells over 48 hours (Fig. 4G and H). In agreement, inhibition of ULK1 by dnULK1 or SBI-0206965 in suspended SK-N-AS cells increased cleaved-PARP and cleaved-caspase-3 protein levels (Fig. 4I). Moreover, inhibition of ULK1 resulted in an increase in annexin-V–positive cells (Fig. 4J). Together, these data suggest that targeted inhibition of ULK1 kinase promotes anoikis, which may contribute to metastasis suppression by reducing CTC survival.
SBI-0206965 sensitizes SK-N-AS cells to TRAIL but not mTOR inhibitors or topoisomerase inhibitors
Finally, we aimed to determine whether SBI-0206965 synergizes with other chemotherapeutic agents, as SBI-0206965 was previously reported to synergize with mTOR inhibitors to target lung cancer cells, and because inhibition of autophagy enhances the efficacy of conventional therapeutic modalities (14, 19). First, we tested whether SBI-0206965 sensitizes SK-N-AS cells to two FDA-approved topoisomerase inhibitors (doxorubicin and topotecan) that are used clinically to treat neuroblastoma. However, no significant increase in cell death was observed in combination treatments (Fig. 5A and B). Similarly, when SBI-0206965 was combined with the mTOR inhibitors INK128 or Torin1 (28, 29), no significant increase in cytotoxicity was observed (Fig. 5C and D). In agreement with cell viability, cotreatment of SBI-0206965 with INK128 or Torin1 did not significantly increase PARP cleavage or annexin-V staining (Fig. 5E and F). As these findings are in opposition to those previously reported (19), we validated that mTORC1/2 signaling was inhibited by INK128 and Torin1 in SK-N-AS cells, as evident by reduced phospho-4eBP1(T37/46) and phospho-AKT(S473) levels (Fig. 5F). We then tested SH-SY5Y and SK-N-DZ neuroblastoma cells, as well as A549 lung cancer cells, which were previously reported to exhibit cytotoxic synergism to cotreatments of SBI-0206965 and mTOR inhibitors (19). As expected, A549 cells displayed a significant increase in annexin-V staining and PARP cleavage when cotreated with SBI-0206965 and INK128 or Torin 1(Fig. 5G and H). However, as with SK-N-AS cells, both SH-SY5Y and SK-N-DZ exhibited limited additive effect to cotreatments of SBI-0206965 and INK128 or Torin 1 (Fig. 5G and H), further demonstrating that SBI-0206965 does not exhibit synergism with mTORC1/2 inhibitors in neuroblastoma. Interestingly, A549 cells display higher levels of phospho-AKT(S473) as compared with SK-N-AS, SH-SY5Y and SK-N-DZ cells, and SBI-0206965 reduces phospho-AKT(S473), particularly in neuroblastoma cells (Fig. 5I). The downstream target of mTORC1, phospho-4eBP1(T37/46), remained unaffected (Fig. 5I). Thus, the lack of combined effect of SBI-0206965 and mTOR inhibition in neuroblastoma may be due to reduced activity and dependence on AKT signaling in neuroblastoma cell lines, as well as due to off-target effects of SBI-0206965 that reduce AKT activity through an undefined mechanism. In support of this, expression of dnULK1 in SK-N-AS cells did not reduce phospho-AKT(S473) (Fig. 5J), indicating that SBI-0206965–mediated reduction of phospho-AKT(S473) is likely independent of ULK1 function. Furthermore, treatment of dnULK1-expressing SK-N-AS cells with mTOR inhibitors promoted apoptosis at levels similar to that of SBI-0206965–treated SK-N-AS cells (Fig. 5K). These results indicate that SBI-0206965 has increased cytotoxicity compared with dnULK1-mediated inhibition of ULK1 in neuroblastoma due to off-target inhibition of AKT, and that the use of a more specific inhibitor of ULK1 can sensitize neuroblastoma cells to mTORC1/2 inhibitors.
As TNF-related apoptosis-inducing ligand (TRAIL) is a potent apoptosis inducer that has been proposed for the treatment of pediatric malignancies (30), we tested whether SBI-0206965 synergizes with TRAIL to kill neuroblastoma cells. When TRAIL was combined with SBI-0206965 in SK-N-AS cells, a significant increase in apoptosis was observed compared to treatment with SBI-0206965 or TRAIL alone, as measured by annexin-V staining (Fig. 6A and B). Moreover, TRAIL sensitized SK-N-AS cells to SBI-0206965 at concentrations as low as 10 ng/mL (Fig. 6C). In agreement, expression of dnULK1 also increased TRAIL-induced apoptosis in SK-N-AS cells, which was further enhanced in starvation conditions (Fig. 6D). Furthermore, TRAIL treatment induced autophagic flux in SK-N-AS cells (Fig. 6E), indicating that autophagy may be upregulated to suppress TRAIL-induced apoptosis, as inhibition of ULK1 sensitizes SK-N-AS cells to TRAIL treatment.
Neuroblastoma is a type of cancer that develops in early nerve cells of the sympathetic nervous system and most often affects infants and children younger than 10 years of age. There are about 700 new cases of neuroblastoma each year in the United States that account for about 7–10% of all cancers in children and more than 9% of all pediatric cancer–related deaths (1–3). There is an urgent need for the development of more effective therapies for neuroblastoma. Increasing evidence suggests that dysregulation of autophagy contributes to the pathogenesis and drug resistance of cancer (8, 9, 14, 15). Here, we demonstrate that targeted inhibition of an essential autophagy kinase, ULK1, promotes tumor cell death as well as reduced proliferation in two mouse models of neuroblastoma, indicating that the loss of ULK1 is both cytotoxic and cytostatic. Although markers of apoptosis were enriched by dnULK1 expression in vitro, expression of dnULK1 in animal models enhanced both apoptosis and necrosis markers, indicating that the mechanism of cell death may be altered and multifaceted in an in vivo setting. Nevertheless, inhibition of ULK1 function reduced neuroblastoma xenograft tumor growth and metastatic burden in the liver, as well as significantly increased survival in a mouse model of neuroblastoma metastasis. This decrease in metastatic disease, in particular, demonstrates that inhibition of ULK1 may have important clinical implications in neuroblastoma, as metastasis is detected in many patients with neuroblastoma at diagnosis and predicts worse prognosis (1–3). Thus, targeting ULK1 kinase has the potential to provide a novel therapeutic approach for the treatment of neuroblastoma. It is encouraging that the recently developed small-molecule inhibitor of ULK1, SBI-0206965, significantly reduces cell growth and promotes apoptosis in several neuroblastoma cell lines. Future work characterizing and optimizing the pharmacologic properties and on-target efficacy of SBI-0206965 in animal models will be an important next step in determining whether SBI-0206965 is a viable pharmaceutical that can translate to the clinic.
Inhibition of autophagy has been shown to enhance the efficacy of conventional therapeutic modalities in patients with advanced solid malignancies (31). Moreover, SBI-0206965 has been previously demonstrated to synergize with mTOR inhibitors against lung cancer cells (19). In this study, SBI-0206965 has limited additive effect on the cytotoxicity of either mTOR inhibitors (INK128 and Torin1) or conventional chemotherapies (doxorubicin and topotecan) in neuroblastoma cell lines. However, SBI-0206965 was observed to reduce AKT(S473) phosphorylation, a predictor of poor outcome in neuroblastoma (32), indicating that SBI-0206965 reduces key oncogenic pathways in neuroblastoma. Moreover, the influence of SBI-0206965 on AKT activity may explain why we fail to see synergistic effects from cotreatments of SBI-0206965 and mTOR inhibitors. In support of this, expression of dnULK1 did not reduce AKT(S473) phosphorylation and sensitized SK-N-AS cells to mTORC1/2 inhibition, resulting in levels of apoptosis similar to that of SBI-0206965 treatment. Although it is possible that the reduction in AKT phosphorylation is directly due to off-target effects of SBI-0206965, previous profiling of SBI-0206965 kinase selectivity did not report AKT as a target (19), suggesting that AKT may be downregulated through off-target inhibition of other kinases, such as SRC. Thus, further investigation into the specificity of SBI-0206965 in neuroblastoma is warranted. However, this off-target effect of SBI-0206965 provides duel inhibition of both ULK1 and AKT activity, which may be advantageous for treating neuroblastoma.
The TNF-related apoptosis-inducing ligand (TRAIL) has been implicated as a potential anticancer therapeutic agent for its ability to strongly trigger the extrinsic pathway of apoptosis by binding to its receptors (TRAIL-R1/2) to promote cancer cell death without causing significant toxicity in nonmalignant cells (30, 33–35). Moreover, TRAIL receptor agonists (TRA) are well-tolerated and demonstrate some therapeutic efficacy in a phase I clinical trial of pediatric patients with solid tumors (36). However, the majority of clinical trials using TRAs have provided limited efficacy due to tumor cell resistance to apoptosis induction (35, 37–41). To translate the promising preclinical data of TRAs to the clinic, current research is focused on the development of more effective TRAs (42) and the elucidation of the mechanisms of TRAIL resistance to develop potent TRAIL sensitizers (33, 35). Here, we demonstrate that SBI-0206965 greatly sensitizes SK-N-AS neuroblastoma cells to TRAIL treatment, suggesting an opportunity for an effective cotherapy development. Reversal of TRAIL resistance by autophagy inhibition has been previously demonstrated in many other cancer types including T-cell leukemia (43), colon carcinoma (43, 44), breast cancer (44, 45), lung cancer (46), cervical cancer (47), pancreatic cancer (48), and thyroid cancer (49). Moreover, increased autophagic activity promotes TRAIL resistance in breast cancer (45), and autophagy is stimulated by TRAIL treatment to promote survival in lung cancer cells (46). Thus, our results are in agreement with previous reports that autophagy inhibition promotes TRAIL sensitization. The mechanisms through which SBI-0206965 sensitizes neuroblastoma to TRAIL are currently unclear. Further investigation into these mechanisms will advance our understanding of both TRAIL resistance and the crosstalk between ULK1 and apoptosis signaling.
In summary, we establish that inhibition of ULK1 kinase in neuroblastoma promotes apoptosis and reduces tumor growth and metastatic disease. To our best knowledge, this work is the first to demonstrate the antitumor effects of targeting an essential autophagy gene in neuroblastoma mouse models. These data suggest that inhibition of ULK1 kinase is a viable therapeutic option for the treatment of neuroblastoma and that further development and characterization of ULK1 inhibitors such as SBI-0206965 is warranted. Future research characterizing the bioavailability, specificity, toxicity, and efficacy of SBI-026965 in neuroblastoma mouse models will be an essential component of preclinical development. In addition, it will be interesting to determine whether the observed reduction in neuroblastoma growth and induction of apoptosis are simply due to autophagy inhibition or due to noncanonical functions of ULK1 (50). Regardless, inhibition of ULK1 kinase significantly suppresses tumor growth and metastasis and promotes overall survival in animal models of neuroblastoma, making it a promising drug target for the treatment of neuroblastoma.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Conception and design: C.M. Dower, N. Bhat, B.A. Miller, H.-G. Wang
Development of methodology: C.M. Dower, N. Bhat, H.-G. Wang
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): C.M. Dower, N. Bhat, M.T. Gebru, L. Chen
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): C.M. Dower, N. Bhat, M.T. Gebru, C.A. Wills, H.-G. Wang
Writing, review, and/or revision of the manuscript: C.M. Dower, N. Bhat, C.A. Wills, B.A. Miller, H.-G. Wang
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): C.M. Dower, N. Bhat, L. Chen, B.A. Miller
Study supervision: H.-G. Wang
We thank the members of the Wang lab who gave insight and technical help. We are grateful to the Penn State College of Medicine's Flow Cytometry Core and the Department of Comparative Medicine. This work was supported in part by NIH grants CA171501 (to H.G. Wang) and GM117014 (to B.A. Miller) and the Lois High Berstler Research Endowment Fund and the Four Diamonds Fund of the Pennsylvania State University.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.