Supplementary Figure 14 shows that CP-d/n-ATF5 induces anoikis of neuroblastoma cell lines

Depletion of ATF5 induces anoikis and decreases metastasis of neuroblastoma cells. A and B, Viability of BE(2)-C and SK-N-DZ suspension cells in poly-HEMA–coated plates, expressing Dox-inducible shATF5-1 or shATF5-2 at different timepoints after Dox treatment. Mean ± SD. C and D, Quantification of anoikis of BE(2)-C and SK-N-DZ suspension cells cultured as in A and B at different timepoints after Dox addition. Mean ± SD. E, Quantification of whole-body bioluminescence flux (photons/second) in mice 24 hours after intracardiac injection of BE(2)-C-shATF5-2 cells, + Dox (n = 9); −Dox, (n = 9). Mice were maintained on drinking water containing Dox (2 mg/mL) from 3 days before injection to the time of euthanasia. F, Bioluminescent images at 24 hours after intracardiac injection under conditions described in E. G, Quantification of bioluminescence of blood from mice collected 12 and 24 hours after intracardiac injection under conditions described in E. H, Quantification of apoptosis of BE(2)-C-shATF5-2 CTCs isolated from mice 12 hours after intracardiac injection and treatment ± Dox (see E and Materials and Methods), + Dox (n = 5), −Dox, (n = 5). I, RT-PCR analyses of ATF5 and ACTB in circulating BE(2)-C-shATF5-2 cells isolated after 12 hours. J, Time course of whole-body bioluminescence flux in mice treated as in E. + Dox (n = 8); −Dox, (n = 9). The mice were monitored for metastatic spread by bioluminescence and euthanized at day 31. K, Quantification of total flux (photons/second) by ex vivo liver bioluminescence at the time of euthanasia (day 31). L, Quantification of bioluminescence in bone marrow homogenates at the time of euthanasia (day 31). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Supplementary Figure 9 shows that BMF knockdown rescues ATF5 loss-induced reduction of anchorage-independent cell viability

Supplementary Figure 4 shows that decreased ATF5 expression does not alter the invasiveness of BE(2)-C cells in vitro

Supplementary Figure 18 shows that CP-d/n-ATF5 treatment decreased viability and induced anoikis of SK-N-DZ CTCs

Supplementary Figure S8 shows that BMF overexpression reduces the anchorage-independent viability of BE(2)-C and SK-N-DZ cells.

Purpose: The mortality in patients with MYCN-amplified high-risk neuroblastoma remains greater than 50% despite advances in multimodal therapy. Novel therapies are urgently needed that requires preclinical evaluation in appropriate mice models. Combinatorial treatment with highdose radiotherapy (HDRT) and immunotherapy has emerged as an effective treatment option in a variety of cancers. Current models of neuroblastoma do not recapitulate the anatomic and immune environment in which multimodal therapies can be effectively tested, and there is a need for an appropriate syngeneic neuroblastoma mice model to study interaction of immunotherapy with host immune cells. Here, we develop a novel syngeneic mouse model of MYCN-amplified neuroblastoma and report the relevance and opportunities of this model to study radiotherapy and immunotherapy.Materials and methods: A syngeneic allograft tumor model was developed using the murine neuroblastoma cell line 9464D derived a tumor from TH-MYCN transgenic mouse. Tumors were generated by transplanting 1 mm3 portions of 9464D flank tumors into the left kidney of C57Bl/6 mice. We investigated the effect of combining HDRT with anti-PD1 antibody on tumor growth and tumor microenvironment. HDRT (8 Gy x 3) was delivered by the small animal radiation research platform (SARRP). Tumor growth was monitored by ultrasound. To assess the effect on immune cells tumors sections were co-imuunostained for six biomarkers using the Vectra multispectral imaging platform.Results: Tumor growth was uniform and confined to the kidney in 100% of transplanted tumors. HDRT was largely restricted to the tumor region with minimal scattered out-of-field dose. Combinatorial treatment with HDRT , PD-1 blockade significantly inhibited tumor growth and prolonged mice survival. We observed augmented T-lymphocyte infiltration, especially CD3+CD8+ lymphocytes, in tumors of mice which received combination treatment.Conclusion: We have developed a novel syngeneic mouse model of MYCN amplified high-risk neuroblastoma. We have utilized this model to show that combining immunotherapy with HDRT inhibits tumor growth and prolongs mice survival.

Supplementary Figure 13 shows that CP-d/n-ATF5 reduces the anchorage-independent viability of neuroblastoma cell lines

CP-d/n-ATF5 induces anoikis and inhibits neuroblastoma growth and metastasis in vivo. A, Effects of CP-dn-ATF5 on the expression of ATF5 and apoptosis-related BCL-2 proteins in suspension cells. Immunoblot of proapoptotic and antiapoptotic proteins in BE(2)-C and SK-N-DZ suspension cells at 72 hours after CP-d/n-ATF5 treatment. β-Actin was used as a loading control. Densitometric analysis was performed using ImageJ. B, BMF knockdown inhibits CP-d/n-ATF5–induced anoikis of BE(2)-C and SK-N-DZ cells. Adherent cells were treated with or without siBMFs (50 nmol/L) as indicated for 24 hours and then seeded in nonadherent plates. CP-d/n-ATF5 (200 µmol/L) was added, and 72 hours later, anoikis was evaluated. C, Kaplan–Meier analysis of survival of mice bearing BE(2)-C tumors, treated with vehicle (n = 7) or CP-d/n-ATF5 (50 mg/kg; n = 7). Treatment was started 7 days after cell implantation, once daily for the first 3 days, and then twice weekly. P < 0.01, log-rank (Mantel–Cox). D, Quantification of liver bioluminescence flux (photons/second) by ex vivo imaging at the time of euthanasia from mice bearing BE(2)-C tumors described in C. Vehicle (n = 7), CP-d/n-ATF5 (n = 6). E, Quantification of bioluminescence in bone marrow homogenate measured at the time of euthanasia (RLU/# BM cells, RLU: Relative luciferase unit). Vehicle (n = 7), CP-d/n-ATF5 (n = 6). F, Quantification of CTC measured at the time of euthanasia by bioluminescence of blood. Vehicle (n = 7), CP-d/n-ATF5 (n = 6). G, Twelve hours after intracardiac injection of BE(2)-C cells, blood was collected for (left) quantification of CTC by measurement of blood bioluminescence and (right) for measurement of apoptosis in isolated BE(2)-C CTC. Mice were treated with CP-d/n-ATF5 (n = 9) or vehicle (n = 9) immediately after intracardiac injection. H, BE(2)-C cells were injected intracardially and then treated with vehicle (n = 8) or CP-d/n-ATF5 (n = 10). Whole-body bioluminescence flux was then monitored for subsequent tumor cell growth. I, Representative bioluminescence images showing metastatic growth of BE(2)-C in mice treated with vehicle or CP-dn-ATF5 at day 38 of the experiment described in H. J, Quantification of total flux (photons/second) by ex vivo liver bioluminescence at time of euthanasia (day 38). K, Quantification of bioluminescence in bone marrow homogenate at the time of euthanasia (day 38). L, RT-PCR analysis of ATF5, BMF, and ACTB (internal control) in BE(2)-CTC from K. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Supplementary Figure 5 shows that overexpression of ATF5 promotes anoikis resistance of CHLA-255 in vitro

Supplementary Figure 11 shows that FOXO3 silencing rescues cell viability following ATF5 depletion

Overexpression of ATF5 promotes anoikis resistance in vitro and in vivo. A and B, ATF5 overexpression, by immunoblot, in suspension culture of BE(2)-C and SH-SY5Y cells, 72 hours after transient transfection with pCCL-GFP or pCCL-ATF5. ACTB was used as a loading control. Densitometric analysis was performed using ImageJ. C and D, BE(2)-C and SH-SY5Y suspension cell viability and anoikis at 12, 24, 48, and 72 hours after transfection, respectively. At each timepoint, the comparison is between cells overexpressing ATF5 and control cells expressing GFP (100%). Mean ± SD. E, Quantification of mouse whole-body bioluminescence flux (photons/second) 12 hours after intracardiac injection of BE(2)-C cells overexpressing ATF5 (n = 8) or GFP (n = 7). F, Images showing increased bioluminescence of animals described in E indicate tumor cell survival 12 hours after intracardiac injection. G, Quantification of bioluminescence of blood collected 12 hours after intracardiac injection as in E. H, Quantification of apoptosis of BE(2)-C CTC isolated at 12 hours, GFP (n = 7); ATF5 (n = 7). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Supplementary Figure 17 shows that CP-dn-ATF5 reduces tumor growth and metastasis of SK-N-DZ in vivo

Supplementary Figure 1 shows that ATF5 is expressed in neuroblastoma cell lines.

Supplementary Figure 3 shows the decreased expression of ATF5 in + Dox tumors

ATF5 depletion induces BMF in suspension culture. Immunoblot of BCL-2 family proteins in BE(2)-C cells (A) and SK-N-DZ cells (B) expressing Dox-inducible shATF5-1 or shATF5-2, in adherent and suspension cultures, respectively, 72 hours after treatment ± Dox. β-Actin was used as a loading control. Densitometric analysis was performed using ImageJ. C, RT-PCR analyses of ATF5 and BMF in detached BE(2)-C cells 72 hours after ±Dox addition. RNA was isolated, reverse transcribed, and PCR was performed. ACTB was used as an internal control. D, RT-PCR of ATF5 and BMF in circulating BE(2)-C-shATF5-1 cells isolated 12 hours after intracardiac injection ±Dox (2 mg/mL) in drinking water. ACTB was used as an internal control.

BMF is required for ATF5 depletion–dependent anoikis induction and is upregulated by FOXO3 induction after ATF5 knockdown. A, BMF expression, by immunoblot, in suspension cells 72 hours after transfection with empty vector (EV) or BMF expression vector in BE(2)-C cells and SK-N-DZ cells. β-Actin was used as a loading control. Densitometric analysis was performed using ImageJ. B, Quantification of anoikis 72 hours after transfection with BMF expression vector or EV in BE(2)-C and SK-N-DZ cells. C, Knockdown of BMF in BE(2)-C and SK-N-DZ cells. Adherent cells were transiently transfected with siGFP (negative control), siBMF-1, or siBMF-2. Twenty-four hours later, cells were seeded in nonadherent poly-HEMA–coated plates. A total of 96 hours after transfection, lysates were immunoblotted. Untreated (UT). β-Actin was used as a loading control. Densitometric analysis was performed using ImageJ. D, BMF knockdown suppresses anoikis in detached neuroblastoma cells. Quantification of anoikis in detached BE(2)-C and SK-N-DZ, 96 hours after transfection with siGFP (negative control), siBMF-1, or siBMF-2. BMF knockdown inhibits induction of anoikis following ATF5 depletion. BE(2)-C (E) and SK-N-DZ (F), expressing shATF5-1 (left) or shATF5-2 (right), were transfected with siGFP (negative control), siBMF-1, or siBMF-2, and 24 hours later seeded in poly-HEMA–coated plates. Dox was added to deplete ATF5, and 72 hours later, anoikis was measured. −Dox (open bars), +Dox (closed bars). G, ATF5 knockdown in BE(2)-C and SK-N-DZ suspension cells elevates FOXO3 expression. Immunoblot analyses of ATF5 and FOXO3 in suspension cells expressing shATF5-1 or shATF5-2 at 72 hours ± Dox addition. β-Actin was used as a loading control. Densitometric analysis was performed using ImageJ. H, Immunoblot analyses show that FOXO3 knockdown with siRNA prevents BMF induction in response to ATF5 knockdown. BE(2)-C and SK-N-DZ cells, expressing shATF5-2, were transfected with siFOXO3-1 or siFOXO3-2 and 24 hours later were seeded in poly-HEMA–coated plates. Dox was added, and 72 hours later, cell lysates were immunoblotted. β-Actin was used as a loading control. Densitometric analysis was performed using ImageJ. I, Anoikis induction by ATF5 knockdown requires FOXO3. Percentage of BE(2)-C-shATF5-2 (left) and SK-N-DZ-shATF5-2 (right) cells undergoing anoikis at 72 hours, ± Dox addition and ± FOXO3 knockdown. −Dox (open bars), +Dox (closed bars). Cells were treated as described in H. Mean ± SD. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Supplementary Figure 16 shows that CP-d/n-ATF5 inhibits growth of BE(2)-C tumors and increases apoptosis

Supplementary Table 1. List of antibodies used in this study

Abstract MYCN-amplified neuroblastoma often presents as a highly aggressive metastatic disease with a poor prognosis. Activating transcription factor 5 (ATF5) is implicated in neural cell differentiation and cancer cell survival. Here, we show that ATF5 is highly expressed in patients with stage 4 high-risk neuroblastoma, with increased expression correlating with a poorer prognosis. We demonstrated that ATF5 promotes the metastasis of neuroblastoma cell lines in vivo. Functionally, ATF5 depletion significantly reduced xenograft tumor growth and metastasis of neuroblastoma cells to the bone marrow and liver. Mechanistically, ATF5 endows tumor cells with resistance to anoikis, thereby increasing their survival in systemic circulation and facilitating metastasis. We identified the proapoptotic BCL-2 modifying factor (BMF) as a critical player in ATF5-regulated neuroblastoma anoikis. ATF5 suppresses BMF under suspension conditions at the transcriptional level, promoting anoikis resistance, whereas BMF knockdown significantly prevents ATF5 depletion–induced anoikis. Therapeutically, we showed that a cell-penetrating dominant-negative ATF5 peptide, CP-d/n-ATF5, inhibits neuroblastoma metastasis to the bone marrow and liver by inducing anoikis sensitivity in circulating tumor cells. Our study identified ATF5 as a metastasis promoter and CP-d/n-ATF5 as a potential antimetastatic therapeutic agent for neuroblastoma. Significance: This study shows that resistance to anoikis in neuroblastoma is mediated by ATF5 and offers a rationale for targeting ATF5 to treat metastatic neuroblastoma.