The correlation between low MOXD1 expression and worse patient prognosis is a neuroblastoma specific feature.

(A) Kaplan-Meier survival curves with log-rank P value for patients with neuroblastoma in the SEQC cohort (n = 498) stratified into quartiles based on the MOXD1 mRNA expression levels. (B and C) MOXD1 expression in patients with neuroblastoma (SEQC cohort) stratified according to the INSS stages (B) or low-risk or high-risk (C). (D) MOXD1 expression in patients with neuroblastoma (SEQC cohort) stratified according to the age at diagnosis (<18 months versus >18 months). (E) Detailed view of gains (red) and losses (blue) on chromosome 6 from high-risk patients in the Depuydt cohort (n = 556). The black line indicates the chromosomal location of MOXD1 (6q23.2). Maximum size of aberration was set to 180 Mb. (F) Kaplan-Meier plot of overall survival of patients with high-risk neuroblastoma (Depuydt; n = 539). Patients stratified by loss of MOXD1 or no loss of MOXD1. Patients with distal 6q mutations not affecting MOXD1 and patients lacking overall survival information were excluded. (G) Representative images of tumor cores stained for MOXD1 in different age groups [≤ 18 months versus > 18 months at diagnosis (left)] and tumors with loss of 6q (right). *Statistical analysis of significant correlations between MOXD1 expression and age at diagnosis. (H) MOXD1 expression in patients with ganglioneuroma versus neuroblastoma. (I) Expression of MOXD1 in malignant melanoma versus neuroblastoma; data extracted from the Cancer Cell Line Encyclopedia. (J) Expression of MOXD1 in the primary or metastatic malignant melanoma. (K) Expression of MOXD1 in colorectal cancer of stage 4 versus stages 1/2/3. (L) Expression of MOXD1 in localized versus metastatic breast cancer. (H) to (L) Number of patients (n) is depicted in the graphs.

MOXD1 is expressed in MES-like neuroblastoma cells.

(A) Mapping of MOXD1 mRNA expression in the t-distributed stochastic neighbor embedding (tSNE) of human neuroblastoma single nuclei data by Bedoya-Reina et al. (17). nCx (2, 3, 5, 7, 8, and 9) are selected clusters from the abovementioned published data. (B) Expression of MOXD1 in the cohort from Gartlgruber et al. (15). The groups include MES (mesenchymal), NMNA-HR/LR (MYCN nonamplified high-risk/low-risk), and MYCN (MYCN amplified). Number of patients (n) is depicted in the graph, and P values were generated by analysis of variance (ANOVA), followed by Fisher’s least significant difference (LSD) test. (C) The MOXD1 mRNA expression in neuroblastoma cells with ADRN and MES gene signatures. The number of samples (n) for each cell phenotype is depicted in the graph, and P value by t test as indicated. (D) RNA-seq–derived expression of MOXD1 in the isogenic neuroblastoma cell line pairs. (E) MOXD1 mRNA expression in three neuroblastoma cell lines with an ADRN phenotype (SK-N-BE(2)c, SH-SY5Y, and IMR-32), one with a mixed phenotype (SK-N-SH), and one with the MES phenotype (SH-EP) as assessed by qPCR. Error bars denote SD from n = 2 to 3 biologically independent repeats per cell line. (F and G) Expression of MES- (F) and ADRN- (G) associated genes as assessed by qPCR. n = 3 biologically independent repeats.

MOXD1 knockout (KO) in neuroblastoma cells increases tumor formation in vivo.

(A) MOXD1 expression score analysis by RNA-seq [data from De Wyn et al. (2021)] of sympathetic ganglia from TH-MYCN mice (at 6 weeks of age, tumor tissue at the site of sympathetic ganglia was analyzed), and the corresponding sympathetic ganglia from wild type (WT) mice without tumors (n = 4 for each group and time point). The graph shows the mean expression score ± SD. (B) Schematic description of the CAM assay. Created with BioRender.com. (C) Confirmation of CRISPR-Cas9–mediated knockout of MOXD1 at protein level in neuroblastoma SH-EP cells by immunofluorescence. DAPI (4′,6-diamidino-2-phenylindole) was used to counterstain the nuclei. (D) Survival of chick embryos as presented by the percentage of viable embryos at the indicated time points. Eggs implanted with control (CTRL) cells are marked in black and MOXD1 KO in red. Implanted eggs: n = 28 CTRL and n = 38 MOXD1 KO. (E) Number of eggs with detectable tumors as presented by percentage at D14. There are no error bars due to the absolute numbers. Implanted eggs: n = 28 CTRL and n = 38 MOXD1 KO. (F) Weight in milligram (mg) of the dissected tumors. Error bars indicate SEM and P value determined by t test. Note that there are few eggs involved in this analysis due to the high number of dead embryos and spread tumors that could not be completely dissected. Weighed tumors at D14: n = 5 CTRL and n = 2 MOXD1 KO. (G) Representative images of the CAM in the respective group. Tumor area marked by dashed lines.

MOXD1 knockout (KO) accelerates tumor penetrance in zebrafish.

(A) Summary of MYCN-TT and MYCN-TT + MOXD1 KO zebrafish with tumors. P value by Fisher’s exact test as indicated. (B) CRISPR mutation efficiency in MYCN-TT + MOXD1 KO and MYCN-TT zebrafish as determined by a tumor sample analysis on Miseq and analyzed using CRISPResso2.0 software (http://crispresso2.pinellolab.org/submission). (C) Confirmation of CRISPR-Cas9–mediated knockout of MOXD1 at protein level in tumors dissected from MYCN-TT only and MYCN-TT + MOXD1 KO zebrafish. Staining of MOXD1 by immunofluorescence. DAPI was used to counterstain the nuclei. (D) epresentative images of zebrafish at 5, 17, and 27 wpf. At week 27, each fish was photographed from both sides (right and left lateral sides). Cancer cells were visualized by fluorescence. Note that the images are not from the corresponding fish at each time point, but randomly picked from the whole population. (E) Hematoxylin staining to visualize the tissue of MYCN-TT and MYCN-TT + MOXD1 KO zebrafish tumors as well as staining with phospho-Histone H3 (PH3) visualizing mitosis. Arrowheads denote mitotic cells.

MOXD1 overexpression in ADRN-like cells prolongs survival and reduces tumor burden in in vivo mouse models.

(A and B) Confirmation of MOXD1 overexpression by lentiviral transduction at the mRNA (A) and protein (B) level in neuroblastoma SK-N-BE(2)c cells by qPCR and immunofluorescence, respectively. DAPI was used to counterstain the nuclei. Error bars denote SD from n = 3 biologically independent repeats. (C) Time to tumor formation in mice subcutaneously injected with SK-N-BE(2)c cells transduced with an empty control vector or an MOXD1 overexpression (OE) vector determined by days to reach a volume of 200 mm³. The line indicates the mean and P value by Student’s t test as indicated (n = 10 in MOXD1 OE and n = 7 in CTRL). (D) Kaplan-Meier plot of the overall survival in mice of the indicated subgroup of SK-N-BE(2)c (n = 10 MOXD1 OE, n = 7 CTRL). P value by the log-rank (Mantel Cox) test as indicated. (E) Confirmation of MOXD1 overexpression by lentiviral transduction at the mRNA level in neuroblastoma SK-N-SH cells by qPCR. Error bars denote SD from n = 3 biologically independent repeats. (F) Percentage of mice with tumors from the indicated subgroup of SK-N-SH cells (n = 7 per group). P value by log-rank (Mantel Cox) test. (G) Kaplan-Meier plot of overall survival in mice of the indicated subgroup of SK-N-SH (n = 7 per group). P value by the log-rank (Mantel Cox) test as indicated. (H) Confirmation of MOXD1 overexpression at the mRNA level in neuroblastoma 691-ADRN cells by qPCR. Error bars denote SD from n = 3 biologically independent repeats. (I) Kaplan-Meier plot of survival in mice of the indicated subgroups of 691-ADRN cells (n = 6 per group). P value by the log-rank (Mantel Cox) test as indicated.

MOXD1 enriches for pathways associated with embryonic development and its expression is restricted to SCPs.

(A) Schematic illustration of in vitro grown SK-N-BE(2)c cells [MOXD1 overexpression (OE) or control] and in vivo tumors generated from the same in vitro grown SK-N-BE(2)c cells injected subcutaneously, collected for RNA-seq. Created with BioRender.com. (B and C) Heatmaps of the identified DEGs between wild-type (control) and MOXD1 OE SK-N-BE(2)c cells in vitro (B) and in vivo in mice (C). (D) Heatmap with unsupervised clustering analysis of MES core signature genes adapted from Thirant et al. (16) of the samples in (C). (E) Heatmap of MES core signature genes adapted from Thirant et al. (16) of the 25% of neuroblastomas with lowest MOXD1 expression versus the 25% neuroblastomas with highest MOXD1 expression in the SEQC cohort. (F) Gene ontology enrichment analysis of the DEGs in (C). (G) Mapping of MOXD1 expression on single cells from mouse neural crest and Schwann cell linages from Kastriti et al. (28). (H and J) Visualization of MOXD1 by cell-type clustering on a single-cell analysis of (H) mouse neural crest from Furlan et al. (44), (I) developing mouse adrenal medulla from Hanemaaijer et al. (23), and (J) human adrenal lineage from Jansky et al. (24). Color scale shows log₂-transformed expression. BC, bridge cells; ChC, chromaffin cells; cNC, cranial neural crest; M, melanocytes; Nb, neuroblast; SC, Schwann cells; SCP, Schwann cell precursors; SN, sympathetic neurons; SRG, suprarenal sympathetic ganglion; tNC, trunk neural crest.

Graphical visualization of MOXD1 in healthy and tumor development.

Created with BioRender.com.