Skip to main content
NCBI home page
Neuro-Oncology logoLink to Neuro-Oncology
. 2025 Apr 10;27(7):1715–1728. doi: 10.1093/neuonc/noaf096

BCOR loss promotes both retinoblastoma growth and susceptibility to IGF1R inhibition

Su-Chan Lee 1,✉, Satoshi Nakata 2, Lujain Alaali 3, Kaixuan Wang 4, Pei-Chi Tsai 5, Khoa Pham 6, Brent A Orr 7, Quynh T Tran 8, Laura Asnaghi 9, Eric Raabe 10, Charles G Eberhart 11,12,✉
PMCID: PMC12417828  PMID: 40214614

Abstract

Background

BCL-6 corepressor (BCOR) loss-of-function alterations are common in clinically aggressive retinoblastoma. The study aim was to determine if BCOR loss promotes the growth and dissemination of retinoblastoma cells, and identify the pathways it regulates in these retinal tumors of childhood.

Methods

Gain- and loss-of-function strategies were used to modulate BCOR levels in a panel of retinoblastoma cell lines, and the effects on proliferation, clonogenicity, apoptosis, and migration were assessed in vitro and in murine xenograft models.

Results

BCOR knockdown or knockout in retinoblastoma lines with high protein levels increased tumor growth, invasion, clonogenicity, and chemoresistance in vitro, while increased expression in low BCOR lines slowed growth. Growth of retinoblastoma xenografts was similarly sensitive to BCOR gain or loss. BCOR reduction resulted in upregulation of IGF1 and activation of IGF1 receptor (IGF1R) signaling, and the effects of IGF1R inhibition were dependent on BCOR level. In vitro, reduction of retinoblastoma growth and induction of apoptosis by the IGF1R inhibitors linsitinib and AEW541 were also significantly stronger in cells with low BCOR as compared to controls. Both linsitinib and AEW541 suppressed colony formation in a dose-dependent manner in BCOR knockout or knockdown cells. Finally, high BCOR levels rendered retinoblastoma xenografts insensitive to linsitinib.

Conclusions

Loss of BCOR function is associated with more aggressive retinoblastoma cell line growth and chemoresistance, at least in part due to increased IGF1R signaling. Inhibiting IGF1R pharmacologically had a marked anti-tumor effect in aggressive retinoblastoma lacking BCOR, suggesting it as a new therapeutic target, although this still needs to be confirmed in clinical samples with BCOR mutations.

Keywords: BCOR, IGF1, IGF1R, retinoblastoma

Key points.

  • BCOR suppresses retinoblastoma growth and invasion in vitro and in vivo.

  • BCOR loss results in increased IGF1 expression and activated IGF1R signaling in retinoblastoma, along with decreased chemosensitivity.

  • IGF1R inhibitors suppress retinoblastoma xenograft growth in a BCOR-dependent manner.

Importance of the Study.

BCOR is increasingly being implicated in central nervous system tumorigenesis, and molecular changes leading to BCOR loss of function are the second most common alterations in retinoblastoma after Rb mutation. Prior studies showed that BCOR loss was predominantly found in clinically aggressive retinoblastoma, and here we use cell culture and xenograft models to confirm that BCOR suppresses tumor growth and invasion in vitro and in vivo, along with modulating response to chemotherapy. The increased IGF1 production and IGF1R signaling associated with loss of BCOR appear to drive these tumorigenic effects, as IGF1R inhibitors can reverse them when BCOR levels are low. The study thus suggests new chemotherapeutic possibilities for children with aggressive retinoblastoma, along with a biomarker (BCOR) to guide their use.

The BCL-6 corepressor (BCOR) is increasingly being implicated in central nervous system (CNS) tumorigenesis. CNS tumors with BCOR alterations predominantly arise in children and include CNS tumor with BCOR internal tandem duplication (BCOR-ITD), as well as medulloblastoma and retinoblastoma with molecular changes causing loss of BCOR function.1-7 Pediatric and adult gliomas with BCOR and BCORL1 fusions have also been described.8-10 Finally, BCOR alterations have been detected in a range of other predominantly pediatric cancers, and associated with poor prognosis, underscoring their potential importance.1,11,12,13,14

BCOR regulates normal growth and differentiation in developing organs, including the brain and eye. Genetic alterations resulting in BCOR loss of function are found in 2 rare disorders associated with small eyes: Lenz microphthalmia and oculofaciocardiodental syndrome.15,16 Engineered zebrafish, Xenopus, and murine models have a requirement for BCOR in the developing brain and eye.16,17 BCOR is also required for maintenance of pluripotency in human embryonic stem cells.18 Finally, BCOR loss-of-function mutations in induced pluripotent stem cell (iPSC) lines inhibit neuronal differentiation and promote proliferation of stem and progenitor cells.19

We focus here on the role of BCOR in the retinoblastoma—the most common pediatric intraocular cancer, with approximately 8000 new patients and almost 4000 deaths annually worldwide and an estimated 3-year survival rate of just over 50%.20,21 While RB gene loss of function is a defining molecular feature of retinoblastoma, BCOR mutation has been identified as the next most common alteration in these tumors.22 Retinoblastoma cases with dual RB and BCOR mutations recently described by Afshar and colleagues were all large group E tumors which spread into the optic nerve or brain, suggesting that BCOR loss is associated with more aggressive biology.5 Molecular analysis of 96 retinoblastoma by another group found BCOR alterations only in “subtype 2” tumors with increased stemness features and metastasis.23 A recent study by Stachelek and colleagues confirmed the relatively common nature of BCOR loss-of-function alterations in retinoblastoma, and also identified non-synonymous variants in other histone monoubiquitylation genes, including DDB1, RNF20, and PCGF3.6 Overall, these studies suggest that loss of BCOR function represents a common genetic progression event in retinoblastoma associated with aggressive clinical behavior. However, the effects of BCOR loss on retinoblastoma have not been examined in a controlled setting, and the mechanism(s) by which BCOR changes might promote retinoblastoma growth and/or spread are not clear.

BCOR was identified and named based on its ability to bind BCL-6 and facilitate transcriptional repression.24 It is also a member of the noncanonical PRC1.1 complex, which acts to monoubiquitinate histone H2A at Lysine 119 (H2AK119ub1) in a H3K27me3-independent fashion.18,25 BCOR alterations lead to increased IGF2 expression in both medulloblastoma with BCOR loss of function and in CNS tumor with BCOR-ITD, but it is not clear if IGF1 or IGF2 are induced in other tumors showing BCOR alterations.1,26 It is also unclear if BCOR-ITD results in a gain of function, loss of function or neomorphic phenotype in neural cells and tumors.27 Nothing is known about the mechanism(s) by which BCOR might modulate the pathobiology of retinoblastoma.

Retinoblastomas often cause blindness and can invade into the optic nerve or choroid, facilitating metastatic spread to the brain or systemic sites.28 In order to identify new therapeutic targets for aggressive retinoblastoma, we sought to model the effects of BCOR alterations in vitro and in vivo. Loss of BCOR function was associated with more aggressive retinoblastoma growth, at least in part due to activation of IGF1 receptor (IGF1R) signaling via upregulation of the IGF1 ligand. Targeting IGF1R pharmacologically had a marked anti-tumor effect in aggressive retinoblastoma lacking BCOR, suggesting it as a new therapeutic target.

Methods

Cell Culture, Plasmids, and Reagents

Y79 and WER-RB1 human retinoblastoma and Omm1, Mel202, and MP46 human uveal melanoma cell lines were cultured in RPMI-1640 supplemented with 50 IU/ml of penicillin, 50 µg/ml of streptomycin, 1% of l-glutamine, and 10% of heat-inactivated fetal bovine serum (FBS). HSJD-RBVS-10 was cultured in tumor stem medium (TSM) medium (Supplementary Methods). The BCOR CRISPR/Cas9 knockout (KO) plasmid was purchased from Santa Cruz Biotechnology, Inc. Lentivirus was produced by transfecting T293 cells with VSV-G envelope plasmid, D8.9 gag/pol plasmid, and the plasmid containing the gene of interest as described above using Lipofectamine 2000 (#11668500, Thermo Fisher Scientific) per the manufacturer’s instruction.

Generation of BCOR Knockdown and KO Cells

For BCOR knockdown, we utilized a scramble shRNA (SCH002) and a plasmid targeting the 3′-UTR of the human BCOR gene (TRCN0000236074) obtained from Sigma. The shBCOR lentivirus was generated in T293 cells. To achieve BCOR KO, WERI-RB1 cells were seeded by introducing 5 × 105 cells into T25 flasks, each containing 3 ml of antibiotic-free RPMI medium supplemented with 1% FBS growth medium. After a 24-hour incubation, we prepared the transfection mix using the BCOR CRISPR/Cas9 KO plasmid (h) obtained from Santa Cruz (sc-409976) and UltraCruz Transfection Reagent (sc-395739), with concentrations of 3 µg plasmid DNA and 15 µl of Transfection Reagent following Santa Cruz protocol. Subsequently, we confirmed transfection 24 hours post-transfection using a fluorescent microscope equipped with green fluorescent protein (GFP) detection. Within 72 hours, cell sorting was performed through flow cytometry in the Core Facilities at Johns Hopkins.

The 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) and Cell-Titer Blue Assays

Cells were seeded onto 96-cell plates at a density of 1000 cells per well and then incubated for 24 hours. For all experiments, 6 independent wells were analyzed for each experimental condition and time point, and the overall experiment was repeated the number of times indicated in each figure legend. Cells were treated with vehicle or the indicated concentrations of linsitinib, NVP-AEW541, carboplatin, and melphalan diluted in complete medium for 3-5 days. Cells were further incubated with MTT solution (4-hour incubation at 37 °C) or Cell-Titer Blue solution (overnight incubation at 37 °C). The products formed by the addition of the MTT solutions were dissolved in dimethyl sulfoxide (DMSO), and the absorbance was measured at 570 nm. The fluorescence was measured at 560/600 nm for the Cell-Titer Blue solution-added plates. The data are presented as a percentage of the control group.

RNA Extraction and Quantitative Real-Time PCR (qPCR)

RNA extraction from cell lines was carried out using TRIzol (Invitrogen). qPCR was performed as previously described,27 with primer sequences described in Supplementary Methods. All reactions were carried out in triplicate on QuantStudio3 (Applied Biosystems), using SYBR Green (Applied Biosystems) fluorescent dye. The relative fold change was calculated based on the formula R = 2−(ΔCt sample − ΔCt control).

RNA Sequencing and Analysis of Cell Lines

Total RNA was extracted by using the Trizol reagent and sequenced and analyzed by using standard techniques (Novogene Bioinformatics Institute) after being assessed for quality and integrity using a Bioanalyzer (Agilent 2100). All samples had an RNA Integrity Number score of higher than 8.5. The final library was sequenced using Illumina HiSeq 4000 at pair-end 150 bps. Sequences were mapped to the human genome GRCh37 (hg19), with more than 94% verified to have a Phred score ≥Q30. For quantification of gene expression level, FeatureCounts v1.5.0-p3 was used to count the read numbers mapped to each gene; Fragments per kilobase of transcript per million mapped reads (FPKM) of each gene was then calculated based on the length of the gene and the read count mapped to this gene.

Differential expression analysis of the BCOR KO and control groups (with 3 biological replicates per condition) was performed using the DESeq2 R package (1.20.0). The resulting P-values were adjusted using the Benjamini and Hochberg’s approach for controlling the false discovery rate. Genes with an adjusted P-value ≤.05 found by DESeq2 were assigned as differentially expressed. The clusterProfiler R package was used to test the statistical enrichment of differential expression genes in Kyoto encyclopedia of genes and genomes (KEGG) pathways. Reactome database pathways with corrected P-value <.05 were considered significantly enriched by differentially expressed genes. For gene set enrichment analysis (GSEA) analysis, the tool http://www.broadinstitute.org/gsea/index.jsp was used.

RNA Expression Analysis of Surgical Retinoblastoma Samples

RNA sequencing data were extracted from 13 retinoblastoma cases in the SJG4K cohort, which was approved by the St. Jude Institutional Review Board with informed consent obtained in all cases.29 Among those, 4 cases had mutations in BCOR. Data were downloaded from the SJ Cloud portal (https://platform.stjude.cloud/data/cohorts). Briefly, RNA-Seq BAM files were mapped to hg38. For alignment, the STAR v2.7.1a 2-pass mapping approach was used. Feature counts were computed using HTSeq as described in this workflow (https://stjudecloud.github.io/rfcs/0001-rnaseq-workflow-v2.0.0.html#specification). The output files contained a count of the number of reads overlapping each genomic feature, in this case, genes as specified in GENCODE V31. The differential analysis between BCOR-mutant vs non-BCOR-mutant group was performed using the DESeq2 package in R (v.1.44.0). The “apeglm” log2 fold change (LFC) shrinkage was used to shrink the large fold changes from genes with less statistical information, informing a prior distribution which allowed all estimated LFCs to be compared across samples.

For GSEA, LFCs of BCOR-mutant vs non-BCOR-mutant group comparison were computed and used as the ranking metric for GSEA Preranked (v3.0). GSEA was performed to determine if the members of a given gene set were enriched among the most differentiated genes for the comparison. The ranked gene lists were tested against KEGG (v2023.2) and Hallmark gene sets (v2023.2) with 10 000 permutation. Significance was defined by the false discovery-adjusted P-value of <.05.

Colony Formation and Invasion/Migration Assays

To assess colony formation, 1000 cells per well were mixed with agar and placed on base agar in 24-well plates. After solidification of the top agar, compounds diluted in complete medium were added, and plates were incubated for 2-3 weeks until colonies were visualized in the vehicle group. MTT solution was then used for staining, and colonies were counted using the ImageJ program.

Transwell plates coated with gelatin for migration or Matrigel for invasion were seeded with 5 × 104 cells. Conditioned media from NIH3T3 fibroblast cells (for WERI-RB1 cells) and RPMI media containing 20% FBS (for Y79 cells) were placed in the lower chambers. After 16 to 72 hours of incubation, with or without linsitinib, the migrated and invaded cells on the bottom side of the gelatin- or Matrigel-coated membranes were fixed with methanol and stained using crystal violet solution.

Western Blot Analysis

Cells were lysed in with radioimmunoprecipitation assay (RIPA) buffer (Sigma) with protease inhibitor cocktails and phosphatase inhibitor cocktails (Roche). Equal amounts of proteins were subjected to Sodium Dodecyl Sulfate–Polyacrylamide gel electrophoresis (SDS-PAGE) and electrically transferred to polyvinylidene difluoride (PVDF) membranes (Bio-Rad). Membranes were blocked with 3% bovine serum albumin (BSA) in Tris-buffered saline (TBS) containing 0.01% of Tween-20 for 1 hour. The membranes were incubated with primary antibodies diluted in 3% BSA in Tris-buffered saline with tween 20 (TBST) overnight at 4 °C the following antibodies: BCOR, pIGFR, IGFR-β, pERK, ERK, pAkt, Akt, IGF1, IGF2, and cleaved PARP, and then incubated with secondary antibodies for 1 hour at RT. Membranes were washed 3 times with TBST and visualized by using an enhanced chemiluminescence (ECL). For all figures with western blots, representative images from one of at least 2 biological replicates with similar results are shown.

Animal Experiments

All animal experiments were performed according to protocols approved by the Johns Hopkins University Institutional Animal Care and Use Committee. Mice were fed standard mouse chow and water ad libitum and housed in temperature- and humidity-controlled facilities with a 12-hour light/12-hour dark cycle. For flank xenograft experiments, WERI-RB1, WERI-RB1 BCOR K/O, Y79, and Y79 BCOR overexpression cells, which were diluted in equal amount of medium and Matrigel, were injected subcutaneously into the right and left flanks of female nude mice. When tumor volume reached 50-100 mm3, the mice were randomly grouped and administered with vehicle (phosphate-buffered saline, PBS), or linsitinib (25 mg/kg), every other day orally. Tumor growth was determined by measuring the short and long diameters of the tumor with a caliper, and body weight was also measured once a week to monitor toxicity. To assess the orthotropic tumor growth, 1000 retinoblastoma cells were injected into mice eyes and then incubated for 1-2 months until tumor was visualized. To measure tumor size, eyes and associated tumors were removed from mice, and their volume was calculated from measurements in all 3 axes using calipers.

Results

BCOR Loss Promotes the Aggressiveness of Retinoblastoma In Vitro

Among the 6 retinoblastoma lines examined, the highest BCOR expression was found in WERI-RB1 cells, while RBVS1 and RBVS10 cells had moderate levels of protein (Figure 1A). In contrast, BCOR protein was not detected in Y79, RB143, or RBT5 retinoblastoma cell lines. Next, we used 2 different loss-of-function strategies to evaluate the ongoing requirement for BCOR, by knocking it down using shRNA,or knocking it out with CRISPR/Cas9 gene editing. In WERI-RB1 cells, KO resulted in complete loss of BCOR protein and significantly increased growth in vitro (Figure 1B, C). In RBVS10 cells, reduction in BCOR protein levels using shRNA also significantly increased growth (Figure 1D, E). In contrast, when BCOR was overexpressed in Y79 or RB143 cells, growth was significantly slowed (Figure 1F-H). Reduced levels of BCOR in WERI-RB1 and RBVS10 were also associated with a dramatic increase in soft agar colony formation (Figure 1I). We were unable to determine if increased BCOR levels suppressed colony formation in Y79 or RB143 cells, as these lines did not form colonies in our preliminary experiments (data not shown).

Figure 1.

(A) Variable BCOR protein levels were observed across six retinoblastoma cell lines. (B, C) BCOR knockout in WERI-RB1 cells was confirmed by western blot, and led to significantly increased growth. (D, E) Partial knockdown of BCOR using shRNA in RBVS10 cells also enhanced cell growth. (F, G) Overexpression of BCOR in Y79 cells suppressed growth. (H) Inhibitory growth effects were observed in RB143 cells upon BCOR overexpression. (I) Decreased BCOR expression in WERI-RB1 and RBVS10 cells resulted in increased colony formation in soft agar.

Open in a new tab

BCOR loss promotes growth and clonogenicity of retinoblastoma cells. (A) Heterogeneous BCOR protein levels were present in the 6 retinoblastoma cell lines analyzed by Western blot. (B, C) BCOR knockout (KO) in WERI-RB1 cells using CRISPR/Cas9 gene editing was associated with significantly more rapid growth as compared to parental controls. (D, E) Partial BCOR protein knockdown using shRNA in RBVS10 cells was also associated with a significant increase in growth. (F, G) BCOR overexpression (OE) in Y79 cells with very low baseline levels significantly decreased growth. (H) Similar growth effects were seen in RB143 cells after increased BCOR expression. (I) Decreased BCOR in WERI-RB1 and RBVS10 cells was associated with dramatically increased colony formation in soft agar. (Two-way ANOVA tests with means ± standard deviation (SD) shown, *P < .05, **P < .01, ***P < .001, ****P < .0001. For all growth curves and colony assays, data from n = 6 wells per condition and timepoint are shown, and each experiment was repeated at least once with similar results.)

KO of BCOR dramatically increased the ability of WERI-RB1 cells to migrate or invade soft agar in transwell assays (Figure 2A). We saw a statistically significant, more than 100-fold induction of both migration and invasion after BCOR was lost (Figure 2B). In contrast, overexpression of BCOR in Y79 cells significantly slowed their migration (Figure 2C, D).

Figure 2.

(A, B) BCOR knockout in WERI-RB1 cells led to a significant increase in cell migration and invasion in transwell assays. (C, D) In contrast, BCOR overexpression in Y79 cells suppressed migration. (E) BCOR-deficient WERI-RB1 cells exhibited increased resistance to carboplatin and melphalan when cell mass was assessed. (F) In intraocular xenografts, BCOR knockout tumors grew rapidly and extended beyond the eye, unlike control tumors which remained localized. (G) Quantification of tumor size confirmed significantly larger tumors in the BCOR knockout group. (H) Histological images showed that BCOR knockout tumors filled the entire eye, while control tumors were confined to the retina and vitreous. (I, J) In subcutaneous xenografts, BCOR loss in WERI-RB1 cells promoted tumor growth over two weeks, whereas BCOR overexpression in Y79 cells suppressed it.

Open in a new tab

BCOR modulates retinoblastoma invasion in vitro and growth in vivo. (A, B) A dramatic and statistically significant increase was seen in the number of WERI-RB1 cells which migrated or invaded in transwell assays after BCOR knockout. (C, D) In contrast, cell migration was suppressed in Y79 cultures after BCOR overexpression. (E) BCOR knockout in WERI-RB1 cells was associated with significant resistance to carboplatin and melphalan. (F) BCOR KO intraocular tumor xenografts grew more rapidly and extended outside the globe, while those generated from control WERI-RB1 cells were smaller and remained contained within the eye (n = 5 animals per group). (G) Measurement of the enucleated globes and associated tumors confirmed the significantly increased size of xenografts lacking BCOR (n = 5 per group). (H) Low power microscopic images (left panels) show WERI-RB1 control tumors involving the retina and focally extending into vitreous, while BCOR KO tumors fill the entire eye (tumor—asterisk, retina—arrowhead). (I, J) In subcutaneous xenografts, BCOR loss in WERI-RB1 cells promoted growth (n = 7 per group), while BCOR overexpression in Y79 cells suppressed growth (n = 5 per group). (Two-way ANOVA with means ± SD shown, *P < .05, **P < .01, ***P < .001, ****P < .0001. For in vitro growth curves, data from n = 6 wells per condition and timepoint are shown, and each experiment was repeated at least once with similar results. For invasion and migration assays, at least 2 wells and 2 fields per well (n ≥ 4) were evaluated per condition.)

Finally, we assessed the effects of BCOR loss on the response of retinoblastoma cells to standard chemotherapies. WERI-RB1 KO cells were significantly more resistant to both carboplatin and melphalan than the parental line (Figure 2E).

BCOR Loss Promotes the Aggressiveness of Retinoblastoma In Vivo

We next assessed the effects of BCOR KO on the growth of WERI-RB1 orthotopic and subcutaneous xenografts. Nine weeks after injection, intraocular tumors lacking BCOR appeared markedly larger than those formed from control cells (Figure 2F). In addition, while all controls grossly appeared confined within the eye, several tumors in which BCOR was knocked out had invaded outside the globe. Measurement of volumes of the eye and associated neoplastic mass after the mice were sacrificed at 9 weeks confirmed that xenografts lacking BCOR were significantly larger than those in eyes injected with control cells (Figure 2G). Microscopic examination revealed that 2 of the control WERI-RB1 tumor xenografts were small, with only focal collections of neoplastic cells in the vitreous and retina; 1 was moderately sized; and 2 were somewhat larger, but with neoplastic cells confined to the intraocular space. In contrast, all 5 xenografts formed by BCOR KO cells filled the eye and invaded into surrounding tissues or onto the ocular surface (Figure 2H). Mitotic activity was significantly higher in the xenografts generated by BCOR KO cells (P = .01; Supplementary Figure 1). Subcutaneous flank xenografts from WERI-RB1 KO cells were also much larger than parental controls, while overexpression of BCOR in Y79 cells suppressed their growth (Figure 2I, J). Taken together, these results suggest that BCOR acts as a tumor suppressor in retinoblastoma, and its loss can promote aggressive growth and invasion in vitro and in vivo.

BCOR Loss Increases IGF1 Levels and IGF1R Signaling in Retinoblastoma

We next investigated gene expression changes after BCOR loss in retinoblastoma cells in order to identify possible effectors modulating growth. RNA sequencing of WERI-RB1 BCOR KO cells and controls identified 1647 protein-coding and non-coding genes upregulated after BCOR loss with an adjusted P-value of <.05 (Supplementary Table 1), while 1333 were significantly lower in the BCOR KO cells (Supplementary Table 2). KEGG gene enrichment analysis was also performed (Supplementary Figure 2). Given the previously reported upregulation of IGF2 and IGF1R signaling in brain tumors showing BCOR loss, the induction of PI3K-Akt noted in our cells on KEGG analysis was of interest, as it can be activated downstream of IGF.26 While we did not identify upregulation of IGF2 in WERI cells after BCOR KO, IGF1 was identified as a possible candidate activating the PI3K pathway, with an approximately 20-fold induction of mRNA identified by RNA sequencing (Supplementary Table 1, Padj = .018; Supplementary Figure 3A). Both IGF1 and IGF2 bind to the IGF1 receptor and activate downstream signaling pathways, including mitogen-activated protein kinase (MAPK) and PI3K-Akt, in many tumor types.30

In order to determine the functional significance of IGF1/IGF1R signaling in retinoblastoma and its relationship to BCOR, we assessed baseline ligand mRNA expression in our cell line panel. The highest level of IGF1 mRNA was present in Y79 cells lacking BCOR, while RBVS10 showed moderate expression of IGF1 and the other 4 lines low expression (Figure 3A). Thus, while some lines showed an inverse relationship between BCOR and IGF1 (ie, Y79, WERI-RB1, and RBVS1), others did not, suggesting regulation of IGF1 expression can depend on multiple factors. In more direct experiments testing the relationship using WERI-RB1 and RBVS10 cells, KO or knockdown of BCOR was associated with significantly increased levels of IGF1 mRNA as assessed by qPCR (Figure 3B, C). While baseline BCOR levels were lower in RBVS1, partial BCOR knockdown was associated with a statistically significant increase in IGF1 mRNA in these retinoblastoma cells as well (Figures 1A and 3A; Supplementary Figure 3B). In contrast, increased expression of BCOR in Y79 and RBVS143 cells reduced IGF1 mRNA levels (Figure 3D; Supplementary Figure 3C). We also confirmed increases in IGF1 protein in WERI-RB1 BCOR KO cells and RBVS10 knockdown cells as compared to controls using western blotting and enzyme linked immunosorbent assay (ELISA) (Figure 3E and data not shown). In contrast, IGF2 protein was not detectable. These studies support the concept that IGF1 expression is suppressed by BCOR, and that IGF1 plays a more important role in retinoblastoma response to BCOR loss than IGF2.

Figure 3.

(A) Baseline IGF1 mRNA levels varied across cell lines, with high expression in Y79 and RBVS10 cells. (B, C) BCOR knockout or knockdown led to significantly increased IGF1 mRNA expression in WERI-RB1 and RBVS10 cells. (D) Conversely, BCOR overexpression in Y79 cells reduced IGF1 mRNA levels. (E) BCOR loss also increased IGF1 protein levels in WERI-RB1 and RBVS10 cells as shown by western blot, while IGF2 was not detectable. (F) Western blot analysis revealed that BCOR reduction in WERI-RB1 and RBVS10 cells induced phosphorylation of IGF1 receptor (pIGF1R) and pERK, while pAKT changes were variable. BCOR overexpression in Y79 suppressed pIGF1R and pERK levels. Histone methylation markers H3K36Me2 and H3K36Me3 changed in parallel with BCOR levels.

Open in a new tab

IGF-1 receptor signaling is induced by BCOR reduction. (A) Baseline IGF1 mRNA levels varied between retinoblastoma cell lines. (B, C) BCOR knockout or knockdown resulted in a significant increase in IGF1 mRNA. (D) In contrast, the high baseline IGF1 levels in Y79 cells were markedly decreased when BCOR expression was increased. (E) BCOR knockout also markedly increased IGF1 protein levels on western blots, while IGF2 protein was not detectable. (F) Western blot analysis showed induction of phosphorylated IGF1 receptor (pIGF1R) following BCOR reduction, while BCOR overexpression suppressed pIGF1R. Activated pERK showed a similar pattern of change, while changes to pAKT were more heterogeneous. Levels of H3K36Me2 and H3K36Me2 paralleled those of BCOR (Two-way ANOVA with means ± SD shown, *P < .05, **P < .01, ***P < .001, ****P < .0001. For all quantitative PCR studies, data from n = 3 wells per condition are shown, and each experiment was repeated at least once with similar results.)

BCOR loss and induction of IGF1 were also associated with increased IGF1R activation. Phosphorylated IGF1R (pIGF1R) increased 8- to 11-fold when BCOR was knocked out in WERI-RB1 or down-regulated in RBVS10 cells, while overexpression of BCOR in Y79 markedly suppressed receptor activation (Figure 3F). Similar changes were seen in WERI-RB1 and Y79 flank xenografts in which BCOR was knocked out or induced (Supplementary Figure 3D). BCOR loss also resulted in increased phosphorylation of ERK in WERI-RB1 cells, and on longer exposures in RBVS10 cells (Figure 3F and data not shown). In contrast, BCOR overexpression decreased activation of ERK protein in Y79 cells. Changes to Akt phosphorylation were less clearly associated with BCOR levels, suggesting that perhaps the MAPK signaling pathway plays a more important functional role downstream of BCOR in these cells than protein kinase B (AKT). Expression of IGF1R, AKT1-3, PIK3CA, and ERK1-2 genes was also dysregulated at the mRNA level in multiple cell lines in vitro when BCOR was modulated up or down, possibly reflecting feedback loops (Supplementary Figure 3E). Interestingly, both H3K27Me2 and H3K27Me3 levels were also dependent on BCOR status (Figure 3F).

We next sought to confirm that BCOR mutations were associated with increased IGF1 levels or activation of IGF1R/MAPK signaling pathways in human surgical tumor specimens. Our SJG4K cohort contained 13 retinoblastoma cases with both mutational and mRNA expression data, and 4 of these had BCOR mutations predicted to cause loss of function. A relatively small, non-significant increase in IGF1 mRNA level was present in tumors with BCOR mutations as compared to controls (Supplementary Figure 4A). GSEA analysis showed an increase in both IFG1R and MAPK signaling in tumors with BCOR loss of function due to mutation. A positive enrichment score (>0.35) was seen for both pathways, although the adjusted P-value was >.05 (Supplementary Figure 4B, C).

Finally, to determine if BCOR loss resulted in increased IGF1 expression and activation of IGF1R signaling in tumors arising outside the CNS, we introduced shBCOR into 3 uveal (ocular) melanoma cell lines (Omm1, Mel202, MP46). In all of these, BCOR knockdown was associated with decreased H3K27 methylation and increased IGF1R phosphorylation (Supplementary Figure 5A). Western analyses also showed increased pAKT and pERK, and significantly higher levels of IGF1 mRNA were also present (Supplementary Figure 5B, C).

The Effects of IGF1R Inhibition Are Dependent on BCOR Levels

To help confirm the potential tumorigenic effects of IGF1 signaling in retinoblastoma, human recombinant IGF1 (hrIGF1) was used to treat WERI-RB1 retinoblastoma cells with high BCOR and low endogenous IGF1, resulting in increased phosphorylation of IGF1R and the downstream proteins Akt and ERK (Figure 4A). In addition, adding human recombinant ligand (hrIGF1) to cultures promoted the migration of retinoblastoma cells over 50-fold (Figure 4B). These data support the general concept that increased IGF1 can promote aggressiveness of retinoblastoma through binding to IGF1R and activating oncogenic downstream signaling pathways.

Figure 4.

(A, B) Western blot and migration assays show that recombinant IGF1 promoted IGF1R activation and enhanced migration of WERI-RB1 cells in a dose dependent fashion. (C) The IGF1R inhibitor linsitinib suppressed pathway activation in Y79, WERI-RB1, and RBVS10 cells that showed low BCOR expression and high baseline pIGF1R levels. (D, E) In all 4 cell lines assessed, growth inhibition by linsitinib and AEW541 was more pronounced in retinoblatoma cells with low BCOR compared to controls. (F, G) Apoptosis induction by IGF1R inhibition was examined using western blots and found to be dependent on BCOR status, only increasing in WERI-RB1 and RBVS10 cells with BCOR knockout or knockdown. (H) Linsitinib also significantly suppressed tumor invasion in BCOR-deficient WERI-RB1 and Y79 cells. (I) Both linsitinib and AEW541 reduced colony formation in WERI-RB1 BCOR knockout cells in a dose-dependent manner.

Open in a new tab

The effects of IGF1R modulation on retinoblastoma cells are dependent on BCOR. (A, B) The addition of recombinant IGF1 to WERI-RB1 cultures promotes IGF1R activation and tumor cell migration. (C) The IGF1R inhibitor linsitinib suppressed pathway activity in Y79, WERI-RB1, and RBVS10 cells with low BCOR and high pIGF1R. (D, E) Inhibition of cell growth by linsitinib and AEW541 was also significantly greater in cells with low BCOR as compared to controls. (F, G) Apoptotic induction following IGF1R inhibition was dependent on BCOR status. (H) Linsitinib treatment also significantly inhibited tumor invasion in WERI-RB1 and Y79 retinoblastoma cells lacking BCOR. (I) Both linsitinib and AEW541 suppressed colony formation in a dose-dependent manner in WERI-RB1 BCOR knockout cells. (Two-way ANOVA with means ± SD shown, *P < .05, **P < .01, ***P < .001, ****P < .0001. For growth curves, data from n = 6 wells per condition are shown, and each experiment was repeated at least once with similar results. For invasion and migration assays, data shown represents at least 2 wells and 2 fields per well (n ≥ 4) per condition.)

To directly investigate if the BCOR loss makes retinoblastoma more reliant on the IGF1R signaling pathway, 2 small-molecule IGF1R inhibitors, linsitinib (OSI-906) and AEW541, were used to treat tumor cells in culture. Linsitinib treatment of Y79 and WERI-RB1 BCOR KO cells with significant pathway activity decreased levels of pIGF1R, pAkt, and pERK (Figure 4C). When comparing the therapeutic effects of IGF1R inhibitors on control WERI-RB1 and RBVS10 cells with high BCOR protein and low IGF1R activity to those in which BCOR was reduced and IGF1R activated, inhibition of the receptor slowed cell growth in the BCOR KO or knockdown cells dramatically more effectively than in controls (Figure 4D). In contrast, overexpression of BCOR made Y79 and RB143 cells less sensitive to linsitinib (Figure 4D). A second IGF1R inhibitor, AEW541, also showed significantly better inhibition of WERI-RB1 growth when BCOR was knocked out (Figure 4E). Finally, the response of the 3 uveal melanoma lines tested to linsitinib was also dependent on BCOR levels, with knockdown associated with significantly greater growth inhibition (Supplementary Figure 5D-F).

These anti-growth effects are due at least in part to increased apoptotic response to IGF1R inhibitors. As shown in Figure 4F, IGF1R inhibitors increased cleaved PARP protein levels in BCOR KO WERI-RB1 and BCOR knockdown RBVS10 cells in a dose-dependent fashion, but do not raise levels of this apoptotic marker in paired controls. Similar effects were seen with AEW541 in WERI-RB1 (Figure 4G). The effect of IGF1R inhibition in the Y79 line was also consistent with this concept, as BCOR-low parental cells showed an increase in cleaved PARP, while no induction was seen after BCOR overexpression (Figure 4G). In WERI-RB1 BCOR KO and Y79 cells with elevated IGF1R signaling, linsitinib also significantly inhibited tumor invasion (Figure 4H). Finally, both linsitinib and AEW541 suppressed colony formation in a dose-dependent manner in WERI-RB1 KO cells (Figure 4I).

IGF1R Inhibitors Suppress Retinoblastoma Xenograft Growth in a BCOR-Dependent Manner

We further assessed the relationship between BCOR status and response to IGF1R inhibitors using xenograft models. WERI-RB1 and WERI-RB1 BCOR KO retinoblastoma cells were used to form flank xenografts modeling metastatic disease and incubated for 4 weeks before initiation of therapy. Additional randomly grouped cohorts of mice were treated in parallel with linsitinib for 5 weeks. Linsitinib treatment markedly decreased the xenograft size of WERI-RB1 KO cells lacking BCOR but had no significant effect on the growth of parent cells expressing BCOR (Figure 5A, B). Orthotopic WERI-RB1 xenografts lacking BCOR were also responsive to linsitinib, with significantly smaller eye masses in the treated mice (Figure 5C).

Figure 5.

(A, B) Linsitinib treatment significantly reduced the size of the 7 excised subcutaneous xenografts derived from BCOR knockout WERI-RB1 cells as compared to 7 controls, but had no significant effect on 5 xenografts from parental control cells expressing BCOR as compared to an equal number of control-treated xenograft. (C) In orthotopic models, linsitinib also significantly slowed the growth in volume of WERI-RB1 tumors lacking BCOR, but not those expressing BCOR. (D, E, F) In excised Y79 flank xenografts, linsitinib markedly inhibited tumor growth in parental cells, but not in tumors with enforced BCOR overexpression.

Open in a new tab

IGF1R inhibitors suppress retinoblastoma xenograft growth in a BCOR-dependent manner. (A, B) Linsitinib (Lin.) treatment markedly decreased the xenograft size of WERI-RB1 KO cells lacking BCOR but had no significant effect on the growth of parent control (CT) cells expressing BCOR (n = 7 and n = 5 per group, respectively). (C) Growth of orthotopic WERI-RB1 xenografts lacking BCOR was also significantly slowed by linsitinib (n = 5 per group). (D, E, F) Linsitinib significantly decreased tumor growth of parental Y79 flank xenografts, but not in tumors with BCOR overexpression (n = 5 per group). (Two-way ANOVA with means ± SD shown, *P < .05, **P < .01, ***P < .001, ****P < .0001)

Flank xenograft studies using control Y79 retinoblastoma cells lacking BCOR and paired Y79 cultures in which BCOR was overexpressed were also performed, with linsitinib sensitivity varying based on BCOR level (Figure 5D, E). The IGF1R inhibitor significantly decreased tumor growth in the Y79 control tumor group. However, there were no significant changes in tumor growth between controls and linsitinib treatment in Y79 cells when BCOR is overexpressed. Protein extracts from the xenografts showed reduced phosphorylated IGF1R after linsitinib treatment (Supplementary Figure 6). Overall, these data indicate that low levels of BCOR in retinoblastoma cells are associated with induction of IGF1R activity, which makes these tumors sensitive to small molecule inhibitors of IGF1R signaling.

Discussion

Our combined gain- and loss-of-function studies indicate that BCOR suppresses proliferation, invasion, and clonogenicity of retinoblastoma cells in vitro, as well as the growth of tumor xenografts in vivo. Retinoblastoma cells with low BCOR are also more resistant to chemotherapies such as carboplatin and melphalan. Taken together, these data suggest that BCOR functions as a tumor suppressor in retinoblastoma. Our findings are novel for retinoblastoma, where both the functional effects of BCOR loss and the downstream mechanisms which facilitate them have not been previously reported. Indeed, relatively little is known about the role of BCOR in the retina overall, although transgenic mouse models have shown that BCOR loss in the post-natal murine retina can modulate photoreceptor differentiation.31 Understanding the role of BCOR is of potentially great clinical relevance, as it represents the second most common molecular change in retinoblastoma, and loss-of-function alterations are strongly associated with aggressive biology and poor clinical outcomes.5,6,22

Mechanistically, we focused on the role of IGF1R signaling activation after BCOR loss. This was due to prior studies in which HGNET-BCOR was shown to have elevated IGF2/IGF1R signaling, as well as elevated IGF1 mRNA levels in our WERI-RB1 KO cells with alterations in PI3K-AKT identified on KEGG analysis.26 While human surgical samples molecularly profiled for both BCOR mutation and mRNA expression are limited, our preliminary analysis of 13 cases also suggests possible activation of IGF1R and MAPK in BCOR-mutant retinoblastoma. However, the association between BCOR and IGF1 signaling is not as clear in human samples as in our cell line models, and pathways other than IGF1R may also play important roles in promoting aggressive behavior of BCOR-mutant retinoblastoma.

IGF1R appears to represent a new therapeutic target in retinoblastoma, particularly aggressive tumors with compromised BCOR function. Pharmacologic inhibition of IGF1R suppressed the growth, invasion, and clonogenicity of retinoblastoma cultures. Decreased growth following IGF1R blockade was associated with induction of apoptosis. Similar effects were seen in orthotopic and flank xenografts in mice. Importantly, these therapeutic effects using small molecule IGF1R inhibitors were limited to lines with low BCOR levels and elevated IGF1R signaling, as evidenced by the analysis of 3 paired retinoblastoma lines engineered to have high or low BCOR expression. BCOR thus represents a possible biomarker predicting therapeutic response to IGF1R inhibition.

IGF1R is phosphorylated by binding IGF1 or IGF2, resulting in activation of the phosphatidylinositol 3-kinase (PI3K)/protein kinase B (AKT) signaling pathway or the MAPK signaling pathway.30 In our lines, MAPK activation was more consistently seen after BCOR loss. IGF1 is important in a range of cancers, but its role in retinoblastoma has not been investigated aside from a single study in which exogenous IGF1 could partially rescue the anti-tumor effects of Sonic Hedgehog (SHH) knockdown in WERI-RB1 cells.32-34 In cancers outside the eye, IGF1R signaling has been associated with resistance to standard therapies, and our findings with carboplatin and melphalan treatment suggest the same is true in retinoblastoma.30,35

A range of prior studies have implicated IGF1 and IGF1R in brain and eye development. In humans, low neonatal IGF1 levels are associated with reduced brain and retinal growth.36 IGF1 also plays an important role in the formation and growth of retinal structures in human stem cell-derived organoids.37,38 In the eyes of teleost fish, IGF1 is expressed by cone photoreceptors and promotes the proliferation of rod progenitor cells.39 IGF1 also promotes proliferation in primary cultures from rat neural retina, and IGF1 deficit in mice results in age-related vision loss.40,41 The fact that IGF1/IGF1R signaling can modulate the proliferation of retinoblastoma is therefore consistent with known developmental roles.

Current therapeutic approaches to retinoblastoma generally focus on chemotherapy, delivered either systemically or locally via intra-arterial or intravitreal routes, with radiation reserved for rare advanced cases due to toxicity.42-46 However, chemotherapy-related toxicity can also be problematic, recurrences due to resistant tumor clones occur, and improved treatments are clearly needed.42,43,47 Our findings suggest that IGF1R inhibition—possibly in combination with standard chemotherapy—may represent a novel therapeutic option for retinoblastoma patients with tumors that threaten vision or life.

The IGF1R inhibitor teprotumumab is now in routine clinical use to treat thyroid eye disease, with significant improvements in proptosis, diplopia, and quality of life, as well as good tolerability, providing further support for targeting this pathway clinically.48 Treatments targeting IGF1R signaling in cancer have been generally well tolerated in clinical trials, including pediatric populations, although efficacy as a monotherapy has not yet been demonstrated.30,49,50 This may be due to inadequate patient stratification using biomarkers, and our discovery that BCOR status can predict response could lead to more precise therapeutic targeting in retinoblastoma patients. New liquid biopsy technologies could facilitate minimally invasive evaluation of BCOR status as a biomarker for response to IGF1R inhibition.51,52

Finally, these studies have potential implications for understanding the effects of BCOR alterations in brain tumors. Two previous brain tumor studies reported that loss of BCOR function led to increased IGF2 expression in medulloblastoma and embryonal lesions designated at that time high-grade neuroepithelial tumor (HGNET)-BCOR.1,53 The WHO CNS tumor classification now recognizes the majority of the latter lesions as “CNS tumor with BCOR internal tandem duplication (ITD).” Vehwinger and colleagues have also shown that IGF1R inhibitors can slow the growth of PhKh1 cells derived from a HGNET-BCOR with elevated IGF2 expression and IGF1R activity.26

Our work suggests that IGF1 plays a more important role than IGF2 in activating IGF1R signaling in retinoblastoma, but because the downstream signaling appears similar, therapeutic approaches targeting this pathway may be useful across all 3 tumor types. Additional CNS tumors with genetic alterations involving BCOR and BCORL1 include newer groups with EP300::BCOR, EP300::BCORL1, and CREBBP::BCORL1 fusions, which generally arise in adolescents and young adults and show predominantly glial phenotypes.9,10,54 It remains to be determined if these result in loss of BCOR function or activation of IGF1R signaling. Given the limited number of retinoblastoma with both mutational and mRNA expression data available, it will be important to examine IGF1 expression and IGF1R activation in larger numbers of BCOR-mutant and wild-type surgical specimens in order to confirm our laboratory findings in patients.

Supplementary material

Supplementary material is available online at Neuro-Oncology (https://academic.oup.com/neuro-oncology).

noaf096_suppl_Supplementary_Figures_S1-S5_Tables_S1-S2
noaf096_suppl_supplementary_figures_s1-s5_tables_s1-s2.zip (5.7MB, zip)

Acknowledgments

L.A. is currently employed at the National Institutes of Health. This article was prepared while L.A. was employed at the Johns Hopkins University. The opinions expressed in this article are the author’s own and do not reflect the views of the National Institutes of Health, the Department of Health and Human Services, or the US government.

Contributor Information

Su-Chan Lee, Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.

Satoshi Nakata, Department of Neurosurgery, Gunma University, Maebashi, Japan.

Lujain Alaali, Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.

Kaixuan Wang, Department of Biochemistry and Molecular Biology, Johns Hopkins University School of Public Health, Baltimore, Maryland, USA.

Pei-Chi Tsai, Department of Biochemistry and Molecular Biology, Johns Hopkins University School of Public Health, Baltimore, Maryland, USA.

Khoa Pham, Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.

Brent A Orr, Department of Pathology, St. Jude Children’s Hospital, Memphis, Tennessee, USA.

Quynh T Tran, Department of Pathology, St. Jude Children’s Hospital, Memphis, Tennessee, USA.

Laura Asnaghi, Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.

Eric Raabe, Division of Pediatric Oncology, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.

Charles G Eberhart, Department of Ophthalmology, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA; Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.

Conflict of Interest

None declared.

Funding

This work was supported by the Children’s Cancer Foundation. Additional support came from an unrestricted grant from Research to Prevent Blindness to the WIlmer Eye Institute.

Author Contributions

Planning of study: S.L., S.N., E.R., and C.G.E. Sample and data collection and analysis: S.L., S.N., L.A., K.W., P.T., K.P., B.A.O., Q.T.T., L.A., E.R., and C.G.E. Editing of manuscript: S.L., S.N., L.A., K.W., P.T., K.P., B.A.O., Q.T.T., L.A., E.R., and C.G.E.

Data Availability

The data generated in this study are available within the article and its supplementary files.

References

  • 1. Sturm  D, Orr  BA, Toprak  UH, et al.  New brain tumor entities emerge from molecular classification of CNS-PNETs. Cell.  2016;164(5):1060–1072. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Ferris  SP, Velazquez Vega  J, Aboian  M, et al.  High-grade neuroepithelial tumor with BCOR exon 15 internal tandem duplication—a comprehensive clinical, radiographic, pathologic, and genomic analysis. Brain Pathol.  2020;30(1):46–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Kutscher  LM, Okonechnikov  K, Batora  NV, et al.  Functional loss of a noncanonical BCOR–PRC1.1 complex accelerates SHH-driven medulloblastoma formation. Genes Dev. 2020;34(17–18):1161–1176. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Francis  JH, Richards  AL, Mandelker  DL, et al.  Molecular changes in retinoblastoma beyond RB1: findings from next-generation sequencing. Cancers (Basel). 2021;13(1):149. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Afshar  AR, Pekmezci  M, Bloomer  MM, et al.  Next-generation sequencing of retinoblastoma identifies pathogenic alterations beyond RB1 inactivation that correlate with aggressive histopathologic features. Ophthalmology.  2020;127(6):804–813. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Stachelek  K, Harutyunyan  N, Lee  S, et al.  Non-synonymous, synonymous, and non-coding nucleotide variants contribute to recurrently altered biological processes during retinoblastoma progression. Genes Chromosomes Cancer.  2023;62(5):275–289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Kooi  IE, Mol  BM, Massink  MP, et al.  Somatic genomic alterations in retinoblastoma beyond RB1 are rare and limited to copy number changes. Sci Rep.  2016;6:25264. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Torre  M, Meredith  DM, Dubuc  A, et al.  Recurrent EP300-BCOR fusions in pediatric gliomas with distinct clinicopathologic features. J Neuropathol Exp Neurol.  2019;78(4):305–314. [DOI] [PubMed] [Google Scholar]
  • 9. Tauziede-Espariat  A, Uro-Coste  E, Sievers  P, et al. ; RENOCLIP-LOC. CNS tumor with EP300::BCOR fusion: discussing its prevalence in adult population. Acta Neuropathol Commun. 2023;11(1):26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Wu  Z, Rajan  S, Chung  HJ, et al.  Molecular and clinicopathologic characteristics of gliomas with EP300::BCOR fusions. Acta Neuropathol.  2022;144(6):1175–1178. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Astolfi  A, Melchionda  F, Perotti  D, et al.  Whole transcriptome sequencing identifies BCOR internal tandem duplication as a common feature of clear cell sarcoma of the kidney. Oncotarget. 2015;6(38):40934–40939. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Santiago  T, Clay  MR, Allen  SJ, Orr  BA.  Recurrent BCOR internal tandem duplication and BCOR or BCL6 expression distinguish primitive myxoid mesenchymal tumor of infancy from congenital infantile fibrosarcoma. Mod Pathol.  2017;30(6):884–891. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Astolfi  A, Fiore  M, Melchionda  F, et al.  BCOR involvement in cancer. Epigenomics. 2019;11(7):835–855. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Terada  K, Yamaguchi  H, Ueki  T, et al.  Usefulness of BCOR gene mutation as a prognostic factor in acute myeloid leukemia with intermediate cytogenetic prognosis. Genes Chromosomes Cancer.  2018;57(8):401–408. [DOI] [PubMed] [Google Scholar]
  • 15. Moleiro  AF, Oliveira  JS, Grangeia  A, et al.  Ocular severe involvement in oculofaciocardiodental syndrome: description of a case series. Eur J Ophthalmol.  2023;34(1):NP6–NP11. [DOI] [PubMed] [Google Scholar]
  • 16. Hamline  MY, Corcoran  CM, Wamstad  JA, et al.  OFCD syndrome and extraembryonic defects are revealed by conditional mutation of the Polycomb-group repressive complex 1.1 (PRC1.1) gene BCOR. Dev Biol.  2020;468(1-2):110–132. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Ng  D, Thakker  N, Corcoran  CM, et al.  Oculofaciocardiodental and Lenz microphthalmia syndromes result from distinct classes of mutations in BCOR. Nat Genet.  2004;36(4):411–416. [DOI] [PubMed] [Google Scholar]
  • 18. Wang  Z, Gearhart  MD, Lee  YW, et al.  A non-canonical BCOR-PRC1.1 complex represses differentiation programs in human ESCs. Cell Stem Cell. 2018;22(2):235–251.e9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Puigdevall  P, Jerber  J, Danecek  P, Castellano  S, Kilpinen  H.  Somatic mutations alter the differentiation outcomes of iPSC-derived neurons. Cell Genom. 2023;3(4):100280. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Chantada  GL, Qaddoumi  I, Canturk  S, et al.  Strategies to manage retinoblastoma in developing countries. Pediatr Blood Cancer.  2011;56(3):341–348. [DOI] [PubMed] [Google Scholar]
  • 21. Global Retinoblastoma Study  G.  The Global Retinoblastoma Outcome Study: a prospective, cluster-based analysis of 4064 patients from 149 countries. Lancet Glob Health. 2022;10(8):e1128–e1140. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Murphree  AL, Triche  TJ.  An epigenomic mechanism in retinoblastoma: the end of the story? Genome Med.  2012;4(2):15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Liu  J, Ottaviani  D, Sefta  M, et al.  A high-risk retinoblastoma subtype with stemness features, dedifferentiated cone states and neuronal/ganglion cell gene expression. Nat Commun.  2021;12(1):5578. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Huynh  KD, Fischle  W, Verdin  E, Bardwell  VJ.  BCoR, a novel corepressor involved in BCL-6 repression. Genes Dev.  2000;14(14):1810–1823. [PMC free article] [PubMed] [Google Scholar]
  • 25. Hanot  M, Raby  L, Volkel  P, Le Bourhis  X, Angrand  PO.  The contribution of the zebrafish model to the understanding of polycomb repression in vertebrates. Int J Mol Sci.  2023;24(3):2322. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Vewinger  N, Huprich  S, Seidmann  L, et al.  IGF1R is a potential new therapeutic target for HGNET-BCOR brain tumor patients. Int J Mol Sci.  2019;20(12):3027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Nakata  S, Yuan  M, Rubens  JA, et al.  BCOR internal tandem duplication expression in neural stem cells promotes growth, invasion, and expression of PRC2 targets. Int J Mol Sci.  2021;22(8):3913. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Fabian  ID, Onadim  Z, Karaa  E, et al.  The management of retinoblastoma. Oncogene.  2018;37(12):1551–1560. [DOI] [PubMed] [Google Scholar]
  • 29. Howell  CR, Bjornard  KL, Ness  KK, et al.  Cohort profile: the St. Jude Lifetime Cohort Study (SJLIFE) for paediatric cancer survivors. Int J Epidemiol.  2021;50(1):39–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Hua  H, Kong  Q, Yin  J, Zhang  J, Jiang  Y.  Insulin-like growth factor receptor signaling in tumorigenesis and drug resistance: a challenge for cancer therapy. J Hematol Oncol. 2020;13(1):64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Langouet  M, Jolicoeur  C, Javed  A, et al.  Mutations in BCOR, a co-repressor of CRX/OTX2, are associated with early-onset retinal degeneration. Sci Adv.  2022;8(36):eabh2868. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Shanmugalingam  T, Bosco  C, Ridley  AJ, Van Hemelrijck  M.  Is there a role for IGF‐1 in the development of second primary cancers? Cancer Med. 2016;5(11):3353–3367. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Tominaga  K, Shimamura  T, Kimura  N, et al.  Addiction to the IGF2-ID1-IGF2 circuit for maintenance of the breast cancer stem-like cells. Oncogene.  2017;36(9):1276–1286. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Song  Z, Du  Y, Tao  Y.  Blockade of sonic hedgehog signaling decreases viability and induces apoptosis in retinoblastoma cells: the key role of the PI3K/Akt pathway. Oncol Lett. 2017;14(4):4099–4105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Galal  MA, Alouch  SS, Alsultan  BS, et al.  Insulin receptor isoforms and insulin growth factor-like receptors: implications in cell signaling, carcinogenesis, and chemoresistance. Int J Mol Sci.  2023;24(19):15006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Hellstrom  A, Ley  D, Hansen-Pupp  I, et al.  Role of insulinlike growth factor 1 in fetal development and in the early postnatal life of premature infants. Am J Perinatol.  2016;33(11):1067–1071. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Zerti  D, Molina  MM, Dorgau  B, et al.  IGFBPs mediate IGF-1’s functions in retinal lamination and photoreceptor development during pluripotent stem cell differentiation to retinal organoids. Stem Cells.  2021;39(4):458–466. [DOI] [PubMed] [Google Scholar]
  • 38. Mellough  CB, Collin  J, Khazim  M, et al.  IGF-1 signaling plays an important role in the formation of three-dimensional laminated neural retina and other ocular structures from human embryonic stem cells. Stem Cells.  2015;33(8):2416–2430. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Zygar  CA, Colbert  S, Yang  D, Fernald  RD.  IGF-1 produced by cone photoreceptors regulates rod progenitor proliferation in the teleost retina. Brain Res Dev Brain Res.  2005;154(1):91–100. [DOI] [PubMed] [Google Scholar]
  • 40. de Figueiredo  CS, Raony  I, Medina  SV, et al.  Insulin-like growth factor-1 stimulates retinal cell proliferation via activation of multiple signaling pathways. Curr Res Neurobiol. 2023;4:100068. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Rodriguez-de la Rosa  L, Fernandez-Sanchez  L, Germain  F, et al.  Age-related functional and structural retinal modifications in the Igf1−/− null mouse. Neurobiol Dis.  2012;46(2):476–485. [DOI] [PubMed] [Google Scholar]
  • 42. Schaiquevich  P, Francis  JH, Cancela  MB, et al.  Treatment of retinoblastoma: what is the latest and what is the future. Front Oncol.  2022;12:822330. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. Raval  V, Parulekar  M, Singh  AD.  Emerging new therapeutics for retinoblastoma. Ocul Oncol Pathol. 2022;8(3):149–155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44. Shields  CL, Bas  Z, Laiton  A, et al.  Retinoblastoma: emerging concepts in genetics, global disease burden, chemotherapy outcomes, and psychological impact. Eye (Lond). 2023;37(5):815–822. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45. Feng  ZX, Zhao  J, Zhang  N, Jin  M, Gallie  B.  Adjuvant chemotherapy improves survival for children with massive choroidal invasion of retinoblastoma. Invest Ophthalmol Vis Sci.  2023;64(11):27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46. Chintagumpala  M, Piao  J, Gombos  D, et al.  A multi-institutional feasibility study of intra-arterial chemotherapy in children with retinoblastoma. A Children’s Oncology Group study (COG ARET12P1). Pediatr Blood Cancer.  2024;71(1):e30718. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47. Manukonda  R, Narayana  RV, Kaliki  S, Mishra  DK, Vemuganti  GK.  Emerging therapeutic targets for retinoblastoma. Expert Opin Ther Targets.  2022;26(11):937–947. [DOI] [PubMed] [Google Scholar]
  • 48. Subramanian  PS, Cho  RI, Kahana  A.  Efficacy of teprotumumab therapy in patients with long-duration thyroid eye disease. Curr Opin Ophthalmol.  2023;34(6):487–492. [DOI] [PubMed] [Google Scholar]
  • 49. Galifi  CA, Wood  TL.  Insulin-like growth factor-1 receptor crosstalk with integrins, cadherins, and the tumor microenvironment: sticking points in understanding IGF1R function in cancer. Endocr Relat Cancer.  2023;30(10):e230031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50. DuBois  SG, Krailo  MD, Glade-Bender  J, et al.  Randomized phase III trial of ganitumab with interval-compressed chemotherapy for patients with newly diagnosed metastatic Ewing sarcoma: a report from the Children’s Oncology Group. J Clin Oncol.  2023;41(11):2098–2107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51. Berry  JL, Pike  S, Shah  R, et al.  Aqueous humor liquid biopsy as a companion diagnostic for retinoblastoma: implications for diagnosis, prognosis and therapeutic options. five years of progress. Am J Ophthalmol.  2023;263:188–205. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52. Schmidt  MJ, Prabakar  RK, Pike  S, et al.  Simultaneous copy number alteration and single-nucleotide variation analysis in matched aqueous humor and tumor samples in children with retinoblastoma. Int J Mol Sci.  2023;24(10):8606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53. Paret  C, Russo  A, Otto  H, et al.  Personalized therapy: CNS HGNET-BCOR responsiveness to arsenic trioxide combined with radiotherapy. Oncotarget. 2017;8(69):114210–114225. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54. Yamazaki  A, Arai  Y, Fukuoka  K, et al.  Diffusely infiltrating glioma with CREBBP-BCORL1 fusion showing overexpression of not only BCORL1 but BCOR: a case report. Brain Tumor Pathol.  2022;39(3):171–178. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

noaf096_suppl_Supplementary_Figures_S1-S5_Tables_S1-S2
noaf096_suppl_supplementary_figures_s1-s5_tables_s1-s2.zip (5.7MB, zip)

Data Availability Statement

The data generated in this study are available within the article and its supplementary files.

Articles from Neuro-Oncology are provided here courtesy of Society for Neuro-Oncology and Oxford University Press

RESOURCES