POLQ mediated end-joining promotes DNA damage tolerance in neuroblastoma

Highlights

• •POLQ is highly expressed in high-risk MYCN amplified neuroblastoma.
• •POLQ expression predicts poor survival outcomes.
• •POLQ inhibition sensitizes neuroblastoma cells to anti-cancer therapies.
• •POLQ deletion halts neuroblastoma tumor growth and metastasis.
• •Constitutive POLQ expression in non-tumorigenic neuroblastoma cells causes therapy resistance.

Abstract

Segmental chromosomal alterations, including 11q deletion and 17q gain, are the strongest predictors of pediatric neuroblastoma relapse. These alterations result from unbalanced translocations linked to erroneous DNA repair. Breakpoint sequence analysis suggests that these abnormalities arise through a nonhomologous end-joining mechanism, indicating a potential therapeutic opportunity. While the exact components driving this DNA end-joining process in neuroblastoma remain unknown, polymerase theta-mediated end-joining (TMEJ) has been implicated in other homologous recombination-deficient cancers. Our previous work demonstrated that high-risk neuroblastoma expresses elevated levels of TMEJ components, including DNA Ligase III and Ligase I, while downregulating classical NHEJ factors (DNA Ligase IV and Artemis). Here we show that POLQ, a critical enzyme in TMEJ, is significantly upregulated in neuroblastoma. POLQ knockout impairs proliferation, enhances sensitivity to DNA damaging agents, and reduces tumor growth in vivo. These findings suggest that POLQ facilitates DNA damage tolerance in neuroblastoma and represents a viable therapeutic target.

Introduction

Neuroblastoma (NB) is an aggressive pediatric malignancy arising from developing neural crest stem cells, representing the most common extracranial tumor in children [1]. Despite intensive multimodal therapy, high-risk NB remains a major clinical challenge, with relapse rates exceeding 50 %. The genetic landscape of NB is characterized by both numerical and segmental chromosomal alterations, with 11q deletion and 17q gain serving as critical predictors of poor prognosis. These segmental aberrations, which frequently occur in treatment-resistant tumors, are thought to result from defective DNA double-strand break (DSB) repair mechanisms that contribute to genomic instability and tumor evolution [2,3].

The integrity of the genome is safeguarded by two major DSB repair pathways: non-homologous end-joining (NHEJ) and homologous recombination (HR) [4]. However, in cancers with HR deficiencies, such as high-risk NB, alternative repair mechanisms become indispensable for cell survival. One such mechanism is polymerase theta-mediated end-joining (TMEJ), an error-prone process that utilizes POLQ to facilitate DNA repair under conditions of genomic stress [5]. Our prior work demonstrated a distinct pattern of DNA repair protein expression in high-risk NB, including the suppression of canonical NHEJ factors (LIG4 and Artemis) and upregulation of alternative TMEJ components (LIG3, LIG1, and PARP1) [6]. This suggests a functional reliance on POLQ-mediated repair for maintaining NB genomic stability.

POLQ, a highly conserved helicase-polymerase, has emerged as a key player in tumor adaptation to genotoxic stress. Elevated POLQ expression has been observed in multiple cancers and is associated with poor clinical outcomes due to its ability to drive error-prone repair, foster chromosomal rearrangements, and promote therapeutic resistance[[7], [8], [9]]. Elevated POLQ expression correlates with poor outcomes in multiple cancers [7,[10], [11], [12]], yet its role in neuroblastoma remains unexplored. We hypothesize that POLQ facilitates DNA damage tolerance and tumor maintenance in neuroblastoma . Given the distinct DNA repair landscape in high-risk NB, we hypothesized that POLQ serves as a critical mediator of DNA damage tolerance and tumor maintenance. To test this, we investigated the functional role of POLQ in NB DSB repair, evaluating its impact on proliferation, response to DNA-damaging agents, and tumor progression. Our findings provide new insights into NB repair pathway dependencies and highlight POLQ as a potential therapeutic vulnerability in this aggressive childhood cancer.

Materials and methods

Cell lines and cell cultures

IMR-32 (RRID:CVCL_0346), SK-N-BE(2) (RRID:CVCL_0528) and SH-EP1 (RRID:CVCL_0F47) cells were obtained from American Type Culture Collection (ATCC) and LA1–55n (ECACC Cat# 06,041,203, RRID:CVCL_254) was obtained from Sigma Aldrich. All cells were maintained in Eagle’s minimal essential media supplemented with 10 % fetal bovine serum. The medium for IMR-32 was further supplemented with 1 mM pyruvate, 0.075 % NaHCO3, and non-essential amino acids (Life Technologies). All cell lines were maintained in a 5 % CO₂ incubator at 37 °C. All cell lines were authenticated in December 2022 at the University of Michigan DNA sequencing core by analyzing short tandem repeats and matching it to the ATCC and Sigma Aldrich corresponding cell line databases. Cells were tested for mycoplasma contamination at regular intervals. Cells were maintained in culture only up to 20 passages.

POLQ knockout by CRISPR-CAS9

The All-in-one CRISPR/Cas9 vectors containing both gRNA and expressing Cas9 enzyme driven by the CMV promoter and the donor cassette flanking with homologous arms and the RFP-BSD-selectable marker was obtained from OriGene (Cat#KN212566RB). CRISPR guide RNA specifically targeting POLQ sequence is forward gRNA GATTCGTTCTCGGGAAGCGG and reverse gRNA AAACGGCGGCGTTCAGAATC. LA1–55n cells were co-transfected gRNA and donor vector using Lipofectamine 2000 transfection reagent (Invitrogen Corp, CA), and transfected cells were selected by Blasticidin (BSD, Cat#A1113903) (5 μg/ml). The single clones were isolated and expression of POLQ was detected by immunoblotting and RT-PCR.

POLQ overexpression

POLQ expression plasmid was obtained as a generous gift from Dr. Alan D. D’Andrea (Dana-Farber Cancer Institute, Harvard Medical School). Low POLQ expressing SH-EP1 neuroblastoma cells were transfected with pcDNA3.1(Hygro)myc-hPolQ-Flag (RRID:Addgene_73,132) plasmid using lipofectamine 2000. Two stable transfected clones positive for POLQ expression were selected for further studies.

Quantitative real-time PCR and RNA sequencing

Total RNA was isolated from cells using the NucleoSpin RNA kit (Machery-Nagel) according to the supplier’s instructions. The isolated total RNA was sent to Novogene for sequencing and analysis. For RT-PCR isolated RNA was then reverse transcribed using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Real-time PCR was carried out on 25 ng cDNA using Fast SYBR™ Green Master Mix and Lightcycler 480 (Roche). All reactions were done in 10 μL volume. POLQ primer sequence forward: 5′-CGCTGAGCGTCAAGCTATCA-3′ and reverse: 5′-TTCGTGACAATCCTGTTTGGT-3′. GAPDH primer sequence forward: 5′- GTCTCCTCTGACTTCAACAGCG-3′ and reverse: 5′-ACCACCCTGTTGCTGTAGCCAA-3′.

Western blot

All immunoblots except for POLQ were carried out as previously described[6]. For POLQ analysis, immunoblotting was performed using 3–8 % Tris-Acetate gels (Invitrogen). For transferring, Tris-Acetate gel was incubated for 15 mins in 10 % methanol containing transfer buffer before proteins were electro‐transferred to methanol activated immobilon PVDF membranes. Membranes were blocked with 5 % milk in TBST buffer for 1 hour and then incubated with 1:1000 dilutions in 5 % milk in TBST overnight at 4 °C with primary antibodies anti-POLQ (Thermo Fisher Scientific Cat# PA5–69,577, RRID:AB_2,688,722), anti‐PCNA (Cell Signaling Technology Cat# 2586, RRID:AB_2,160,343), anti-pSMAD1/5/9(Cell Signaling Technology Cat# 13,820, RRID:AB_2,493,181), anti-SMAD5(Cell Signaling Technology Cat# 12,534, RRID:AB_2,797,946), anti-GAPDH(Cell Signaling Technology Cat# 3683, RRID:AB_1,642,205). Membranes were then washed with TBST (10 min × 3), incubated with IgG HRP-conjugated secondary antibodies anti-rabbit(Cell Signaling Technology Cat# 7074, RRID:AB_2,099,233) and anti-mouse(Cell Signaling Technology Cat# 7076, RRID:AB_330,924) respectively at 1:5000 dilutions in 5 % milk for 1 h at room temperature, and washed with TBST (10 min × 3). Chemiluminescence signal with ECL substrate(cat# 34,577, ThermoFisher scientific) was scanned on iBright FL1000 (Invitrogen).

Immunofluorescence

DNA-DSBs were assessed by quantifying the rates of dissolution of yH2AX foci. Immunofluorescence staining was performed on cells cultured in glass chamber slides (Lab-Tek II Chamber Slide w/Cover RS Glass Slide Sterile; Nalgen Nunc). Cells were allowed to attach for 24 h and irradiated at given concentrations and allowed to recover for another 24 h. At the end point, cells were fixed with 4 % PFA, permeabilized with 0.5 % Triton X-100, blocked with 10 % normal goat serum for 15 min at room temperature and then immunostained with yH2AX(Millipore Cat#05–636, RRID:AB_309,864) primary antibody at 4 °C for overnight (diluted 1:200 in 5 % normal goat serum containing 0.1 % Triton X-100; (Cell Signaling Technologies) and Alexa Fluor 594 (Thermo Fisher Scientific Cat# A32742, RRID:AB_2,762,825) goat anti-mouse secondary antibody at room temperature for 1 hour (diluted 1:2000 in PBS; Life Technologies, Carlsbad, CA). The nuclei of cells were stained using Hoechst for 15 min at room temperature and mounted with ProLong Gold Antifade Mountant (Thermo Fisher Scientific). Images were taken on Olympus Microscope and quantified using ImageJ (RRID:SCR_003070) software (Rasband, W.S., ImageJ, U. S. National Institutes of Health, Bethesda, Maryland, USA, https://imagej.net/ij/, 1997–2018).

Cell proliferation and viability assays

Cell proliferation and viability was measured using the MTS/PMS assay. Cells were seeded at 5 × 10³ density in 96 well culture plates. For proliferation, each day cells were incubated with MTS/PMS solution at 37 °C for 4 hrs., followed by absorbance measurement at 490 nm. For etoposide (Etopophos) sensitivity experiments, cells were treated with etoposide (Etopophos) at concentrations mentioned for 72 hrs. At end point, cells were incubated with CellTiter 96® AQueous Non-Radioactive Cell Proliferation Assay (Promega) for four hours at 37 °C followed by absorbance read at 490 nm using a BioTek microplate reader (BioTek). IC50 was quantified with GraphPad Prism (RRID:SCR_002798) software.

TMEJ assay

LA1–55n control cells, POLQ KO-Q6 and KO-Q63 cells were seeded in 6 well plates at 80 % confluence. Cells were allowed to attach for 24 h. Cells were then transfected with pCBASceI(RRID:Addgene_26,477) 0.8 ug/ml and EJ2GFP-puro(RRID:Addgene_44,025) 0.8 ug/ml using lipofectamine 2000. At 72 h post-transfection cells were trypsinized and GFP quantification were assessed by flow cytometry.

Colony formation assays

For POLQ knockout colony formation assay, 100 cells/well were seeded in 12 well tissue culture plates. Colonies were allowed to form for 14 days. At end point, cells were fixed with 4 % paraformaldehyde briefly followed by 0.5 % crystal violet stain.

For SH-EP1 cells colony formation assays, floating cells were collected and plated back in 6 well tissue culture plates and allowed to form colonies for about 14 days. Once colonies were observed in POLQ overexpression groups, tissue culture plates were stained with crystal violet as mentioned above. All colony plates were imaged using iBrightFL1000 (Invitrogen).

In vivo studies

Male severely immunocompromised NOD SCID gamma (NOD.Cg-Prkdc^SCIDII2rg^tm1Wjl/SzJ) (Cat#005,557) mice were purchased from The Jackson Laboratory between 4–6 weeks of age, and were housed 5 per cage with food and water ad libitium on a 12:12 light:dark cycle. Mice were allowed to adjust to new environment for 2 weeks before any experiments were initiated. All experimental protocols were approved by the institutional animal care and use committee (IACUC) at University of Michigan under protocol no PRO000008248 .

For POLQ knockout in vivo studies, LA1–55n control and POLQ KO-Q6 cells were injected into the left adrenal gland of NSG mice (n = 7 per group) utilizing ultrasound guidance as previously described [13]. Tumor engraftment and growth were assessed by ultrasound weekly. At the end point, mice were euthanized as per IACUC protocol standards, tumors were harvested, and necropsy was performed by a veterinary pathologist at the University of Michigan Laboratory of Animal Medicine (ULAM).

Immunohistochemistry

All xenografted animals and resultant tumors were examined and characterized histologically by a board-certified veterinary pathologist at the Unit of Laboratory Animal and Medicine (ULAM). Primary xenografted tumors, along with tissues typical of neuroblastoma metastasis (lymph nodes, cortical bones, liver, lungs), were collected and characterized histologically by routine hematoxylin and eosin (H&E) staining.

Public dataset gene expression analysis

Publicly available, Academic Medical Center cohort; Gene Expression Omnibus (GEO) (RRID:SCR_005012) database accession no GSE45547, DOI:10.1038/cddis.2013.84 (http://r2.amc.nl) was utilized to study gene expression of POLQ in previously analyzed cohorts of neuroblastoma tumor samples. Data were analyzed and downloaded from the R2 database website and formatted for publication as shown. Expression cutoff for each repair gene was determined by the R2 Kaplan–Meier Scanner. Scan modus was used for cutoff determination with a minimum group size of 10 to determine the best p values, and p values were corrected for multiple testing (Bonferroni).

GSEA and DAVID analysis

The difference in expression profile was pre-ranked based on LogFC value with a cut-off of 300 bp to eliminate the signal from noise or background. The pre-ranked file was loaded into the Gene Set Enrichment Analysis (GSEA)(RRID:SCR_003199) software tool (Broad Institute, Inc., Massachusetts Institute of Technology, and Regents of the University of California) for analysis of the regulated pathways that are enriched in a positive or negative manner. The list of the top 20 gene sets are represented by a table based on their normalized enrichment score (NES).

For DAVID (RRID:SCR_001881) analysis gene IDs from pre-ranked file were uploaded into https://david.ncifcrf.gov/tools.jsp website for identifying enriched pathways. The list of the top 20 enriched gene sets are represented.

Statistical analysis

All experiments were independently repeated for a total of three trials. Results were analyzed with ANOVA and Student’s t-test. All data are presented as the mean ± the standard deviation (SD) of three experiments, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Results

POLQ is overexpressed in neuroblastoma and predicts poor survival outcomes

POLQ is an essential mediator of the TMEJ pathway, which is typically repressed in normal human tissues but becomes upregulated in homologous recombination (HR)-deficient cancers [14,15]. Recent studies have underscored the role of TMEJ in facilitating DNA damage recovery in cancer cells, particularly in those with defects in classical DSB repair pathways [14,16]. Our previous research identified a distinct DNA repair landscape in high-risk neuroblastoma, characterized by deficiencies in canonical NHEJ proteins (LIG4, Artemis) and compensatory upregulation of alternative TMEJ factors (LIG3, LIG1, PARP1) [6,17]. Given these findings, we hypothesized that neuroblastoma exploits POLQ-driven TMEJ for survival and resistance to genotoxic stress.

To test this hypothesis, we examined mRNA and protein expression levels of POLQ in a panel of neuroblastoma cell lines and evaluated gene expression in patient tumor datasets. MYCN-amplified neuroblastoma cell lines (SK-N-BE(2), IMR-32, LA1–55n) derived from children with relapsed disease exhibited significantly elevated POLQ expression at both the mRNA and protein levels compared to non-tumorigenic SH-EP1 cells (Fig. 1A and 1B). Notably, these high-POLQ-expressing cell lines represent the classic aggressive neuroblastoma phenotype, with increased tumorigenicity and poor clinical prognosis. Conversely, SH-EP1 cells, which lack MYCN amplification and are non-tumorigenic, displayed significantly lower POLQ expression. SH-SY5Y cells, which represent an intermediate-risk neuroblastoma phenotype, exhibited moderate POLQ expression [18]. These findings provide strong evidence that POLQ upregulation is a hallmark of high-risk neuroblastoma and implicates TMEJ as a key DNA repair pathway in these tumors.

Fig. 1.

POLQ overexpression found in high-risk neuroblastoma cell lines and patient tumors, predicts poor survival outcomes in neuroblastoma. (A & B) neuroblastoma cells with high-risk genotypes (SK-N-BE(2), IMR-32, and LA1–55n cells) expressed significantly higher mRNA, detected by qRT-PCR (A) and protein levels of POLQ in Western blot normalized to GAPDH (B) compared to non-tumorigenic SH-EP1 cells and intermediate risk SH-SY5Y cells. (****p < 0.0001, **p < 0.01, *p < 0.05). The R2 microarray analysis platform was used for analysis and is the data source. For POLQ expression in patient tumors and Kaplan-Meier survival curves we used the study with accession noGSE45547. (C) POLQ expression in neuroblastoma patients across INSS stage was investigated. Stage 4 and Stage 3 groups of patient’s tumors have high expression of POLQ when compared stages 1, 2 and 4. (D) POLQ expression between MYCN amplified and non-amplified tumors demonstrates high POLQ expression in MYCN amplified neuroblastoma tumors when compared to non-MYCN amplified. (E) Scan modus was used for cutoff determination with a minimum group size of 8 to determine the best P value, and P values were corrected for multiple testing (Bonferroni). Gene expression was correlated with survival. Kaplan-Meier curve revealed that high POLQ expression had worse overall survival probability (29 %) compared with patients with low Lig3 expression (81 %), cutoff 2294.7, P = 4.3e-13. (F) Kaplan-Meier curve of event free survival revealed that high POLQ expression had worse event free survival probability (39 %) when compared to low POLQ expression patients (75 %), cutoff 1559.4, p = 2.48e-12.

To further investigate the clinical relevance of POLQ expression, we analyzed transcriptomic data from 649 neuroblastoma patient samples obtained from the Gene Expression Omnibus (GEO, accession no GSE45547). Using a stringent statistical approach with Bonferroni correction, we observed a striking correlation between POLQ expression and neuroblastoma stage. POLQ levels were significantly higher in high-risk Stage 4 neuroblastomas (n = 214) and intermediate-risk Stage 3 tumors (n = 91) compared to low-risk Stage 1, 2, and 4S tumors (Fig. 1C). This pattern suggests that POLQ expression escalates with disease progression and aggressiveness. Given the established role of MYCN amplification as a poor prognostic marker in neuroblastoma, we next assessed whether POLQ expression is linked to MYCN status. POLQ expression was markedly higher in MYCN-amplified tumors (n = 93) relative to non-amplified cases (n = 550) (Fig. 1D). This finding suggests a functional interplay between MYCN-driven oncogenesis and POLQ-mediated DNA repair adaptation.

Crucially, high POLQ expression was strongly associated with adverse patient outcomes. Kaplan-Meier survival analysis revealed that patients with high POLQ expression (n = 67) had poor long-term survival probability of only 29 %, whereas those with low POLQ expression (n = 409) exhibited an 81 % survival probability (Fig. 1E). Similarly, event-free survival (EFS) analysis demonstrated that patients with high POLQ expression had an EFS probability of only 31 %, in contrast to the 69 % probability observed in patients with low POLQ expression (Fig. 1F). These results provide compelling evidence that high POLQ expression defines a subset of neuroblastomas with aggressive clinical behavior, MYCN amplification, and inferior survival outcomes.

CRISPR-Cas9 POLQ knockout impairs TMEJ DNA repair and slows proliferation in neuroblastoma cells in vitro

Previous studies have established that high-risk neuroblastoma cells rely on the TMEJ pathway for survival, with upregulation of key repair components, including LIG3, LIG1, and PARP1 [6,17]. Given the demonstrated lethality of TMEJ disruption in neuroblastoma and other malignancies, including hepatocellular carcinoma and esophageal squamous cell carcinoma, [8,19] we sought to define the functional consequences of POLQ knockout in neuroblastoma. To achieve this, we generated CRISPR-Cas9-mediated POLQ knockout in high-POLQ-expressing LA1–55n cells. Following gene targeting, successful POLQ knockout clones (KO-Q6 and KO-Q63) were validated via genomic PCR , which confirmed the insertion of a donor cassette at the targeted loci (Fig. 2A). Quantitative PCR and immunoblotting further confirmed the near-complete depletion of POLQ transcript and protein levels in KO-Q6 and KO-Q63 clones compared to parental LA1–55n cells (Fig. 2B and 2C). Given the established role of POLQ in TMEJ, we next assessed whether POLQ depletion impaired DSB repair efficiency in neuroblastoma. Using the EJ2-GFP reporter assay, which quantifies TMEJ-mediated repair of an I-SceI-induced DSB, we observed a significant reduction in GFP-positive cells in POLQ knockout clones compared to controls (Supplemental Fig. S1). This finding demonstrates that POLQ deletion disrupts the repair of DSBs through the TMEJ pathway, reinforcing the dependence of high-risk neuroblastoma on POLQ-mediated repair mechanisms.

Fig. 2.

POLQ knockout by CRISPR-Cas9 induces cell cycle arrest in high-risk neuroblastoma LA1–55n cells. (A) Genomic PCR results shows POLQ targeting in clone KO-Q6 and KO-Q63. The upper panel diagram of genomic PCR Primer design, two set of primer were utilized to detect the left and right integration junctions (LF, LR: Forward and reverse PCR primer to amplify the left integration junction; RF, RR: Forward and reverse PCR primer to amplify the right integration junction). Lower left panel show left integration junction PCR products and lower right panel show right integration junction PCR products amplified from blasticidin selected colonies in 1.5 % agarose gel. (B) Immunoblot validation of POLQ expression knockout in KO-Q6 and KO-Q63 colonies. (C) mRNA POLQ expression was measured in LA1–55n control, KO-Q6 and KO-Q63 colonies (*** p < 0.001) (D) Cell growth time was observed using MTS/PMS assay in Control and KO (KO-Q6 and KO-Q63) cells. Cell growth was reduced in POLQ KO groups when compared to Control (**** p < 0.0001) (E) The ability to form colonies of LA1–55n Control and POLQ KO (KO-Q6 and KO-Q63) was investigated. The number of colonies in POLQ KO (KO-Q6 and KO-Q63) groups were significantly less than the control group.

To determine the impact of POLQ knockout on cell viability, we performed proliferation and clonogenic assays. POLQ knockout significantly attenuated LA1–55n cell proliferation compared to controls (Fig. 2D), supporting a functional role for POLQ in neuroblastoma cell growth. Consistent with this, POLQ-deficient cells exhibited markedly reduced clonogenic potential, forming significantly fewer and smaller colonies than their parental counterparts (Fig. 2E). These data indicate that POLQ loss compromises the ability of neuroblastoma cells to sustain proliferative capacity, likely due to the accumulation of unresolved DNA damage.

POLQ deficient neuroblastoma cells exhibit enhanced sensitivity to DNA-damaging therapies

Given the essential role of POLQ in TMEJ-mediated repair, we next sought to determine whether POLQ-deficient neuroblastoma cells exhibit heightened sensitivity to genotoxic stress. To assess this, we treated POLQ knockout (KO) and control LA1–55n cells with etoposide (Etopophos), a topoisomerase II inhibitor known to induce DNA double-strand breaks. POLQ depletion markedly enhanced sensitivity to etoposide, as evidenced by a significant reduction in IC50 values from 262.8 nM in control cells to 21.4 nM and 19.1 nM in KO-Q6 and KO-Q63 cells, respectively (Fig. 3A). These data strongly suggest that POLQ knockout compromises the ability of neuroblastoma cells to mitigate etoposide-induced DNA damage, leading to increased cytotoxicity.

Fig. 3.

POLQ silencing sensitizes high-risk NB cells to anti-cancer therapies. (A) Effect of etoposide on LA1–55n control and POLQ-KO (KO-Q6 and KO-Q63) cell lines was assessed using MTS/PMS assay. LA1–55n control and POLQ-KO (KO-Q6 and KO-Q63) cells were treated with varying concentrations of etoposide for 72 h. The IC50 of POLQ-KO (KO-Q6 and KO-Q63) groups decreased significantly when compared to the control group. (B) Sites of DNA damage were measured with yH2AX foci expression post 4 Gy irradiation at 2 time points (0.5 h and 24 h). At 0.5 h no difference was observed between control and POLQ-KO (KO-Q6 and KO-Q63) groups. At 24 h the number of yH2AX foci/cell was significantly elevated in POLQ-KO groups relative to the control group suggesting the failure to repair DSBs in the absence of POLQ in these high-risk NB cells. Scale Bar: 10 um.

To further characterize the impact of POLQ loss on therapeutic vulnerability, we evaluated the effect of ionizing radiation on DNA damage accumulation in POLQ-deficient neuroblastoma cells. LA1–55n control, KO-Q6, and KO-Q63 cells were subjected to 4 Gy of irradiation, and the kinetics of DNA damage repair were assessed by quantifying γH2AX foci formation at 30 min and 24 h post-irradiation. At the 30-minute time point, γH2AX foci levels were comparable across all groups, indicating that initial DNA damage induction was similar. However, by the 24 \h time point, γH2AX foci persisted at significantly higher levels in POLQ knockout cells compared to controls (Fig. 3B), suggesting impaired DNA repair and prolonged retention of unrepaired DSBs. These findings provide compelling evidence that POLQ deficiency compromises the ability of neuroblastoma cells to efficiently repair DNA damage, rendering them more susceptible to both etoposide-induced cytotoxicity and radiation therapy. Given the profound impact of POLQ depletion on treatment sensitivity, these results underscore the therapeutic potential of POLQ inhibition as a strategy to enhance the efficacy of genotoxic therapies in high-risk neuroblastoma.

POLQ knockout suppresses neuroblastoma tumor growth and metastasis in vivo

To extend our in vitro findings, we next sought to determine whether POLQ depletion impacts neuroblastoma tumor progression and metastatic potential in vivo using our well-established orthotopic xenograft model[13]. For this, ∼2 × 10⁵ POLQ knockout (KO-Q6) or wild-type LA1–55n control cells suspended in media containing 50 % Matrigel were injected into the left adrenal gland of NOD/SCID-Gamma (NSG) mice. Tumor growth was monitored weekly by high-resolution ultrasound.

Wild-type control xenografted tumors exhibited rapid growth, with a mean tumor volume of 653.5 mm³ at day 37 and a high engraftment rate (86 %, 6/7 mice). Necropsy revealed robust local invasion and microscopic evidence of lung metastases, indicative of aggressive disease progression. In contrast, POLQ knockout tumors were significantly smaller, with an average volume of only 207.0 mm³ at day 37, and exhibited a reduced engraftment rate of 43 % (3/7 mice), suggesting that POLQ depletion impairs tumor establishment and expansion (Fig. 4A, 4B). Notably, no lung metastases were observed in the POLQ knockout cohort, reinforcing the critical role of POLQ in promoting neuroblastoma dissemination.

Fig. 4.

POLQ knockout halts high-risk neuroblastoma progression in vivo. LA1–55n Control or KO-Q6 2 × 10⁵ cells were injected into left adrenal gland of NSG mice (n = 7). Tumor growth was monitored weekly using ultrasound. (A) Tumor growth curves measured by ultrasound demonstrates that POLQ absence slows tumor progression (****p < 0.0001). (B) Representative picture of harvested tumors. (C) Immunoblot demonstrates POLQ and PCNA expression was down in a KO-Q6 harvested tumor when compared to a control tumor. (D) Histologic images of control tumors with local infiltration into adjacent musculature. (E) Micro-metastasis to the lungs (arrows). Scale Bar: 200 um.

To validate the molecular consequences of POLQ deletion in vivo, we performed Western blot analysis on excised xenografted tumor samples. As expected, POLQ expression was undetectable in knockout tumors, confirming the efficiency of genetic deletion. Importantly, PCNA (Proliferating Cell Nuclear Antigen), a marker of tumor proliferation that positively correlates with POLQ levels in primary neuroblastoma tumors (Supplemental Fig. S2), was significantly reduced in POLQ knockout tumors compared to controls (Fig. 4C). This reduction in PCNA expression further supports the conclusion that POLQ is essential for sustaining tumor proliferation in neuroblastoma.

Collectively, these data provide compelling in vivo evidence that POLQ depletion suppresses neuroblastoma tumor growth and metastatic spread. These findings underscore POLQ as a critical driver of tumor progression and highlight its potential as a therapeutic vulnerability in high-risk neuroblastoma.

Knockout of POLQ augments BMP expression and activates smad1/5/9 signaling in neuroblastoma

To delineate the molecular consequences of POLQ depletion and to identify downstream signaling pathways, we performed RNA-sequencing (RNA-seq) on POLQ KO (KO-Q6, KO-Q63) and WT LA1–55n cells (n = 3 per group). Differential expression analysis identified a striking transcriptional shift upon POLQ loss, with multiple genes involved in differentiation, apoptosis, and immune response significantly upregulated (Supplemental Fig S3). The 50 top upregulated and the downregulated genes common in both KO-Q6 and KO-Q63 cells are listed in Fig. 5A. Among the top upregulated genes common to both knockout groups was Bone Morphogenetic Protein 2 (BMP2), a member of the TGFβ superfamily that induces growth arrest and differentiation in neuroblastoma through activation of Smad signaling [20,21]. Given BMP2’s well-established role in promoting terminal differentiation, we hypothesized that POLQ depletion may shift neuroblastoma cells toward a less proliferative, more differentiated state. To validate this, we examined the phosphorylation status of SMAD1/5/9, key effectors of BMP-mediated signaling. Western blot analysis revealed a significant increase in phosphorylated SMAD1/5/9 levels in POLQ knockout cells compared to controls (Fig. 5B), supporting our hypothesis that POLQ loss promotes BMP-Smad activation. POLQ knockout cells exhibited a marked reduction in the expression of key stemness-associated markers, including SOX2, Nanog, and Oct-4, which are known to be negatively regulated by Smad1/5/9 activation [20]. Additionally, genes characteristic of human fetal erythroid tissue at early developmental stages (HBQ1), as well as stem cell-associated genes LRRC34 and TERT, were significantly downregulated in POLQ-deficient cells (Supplementary Tables S1 – S2). These findings suggest that POLQ deletion promotes neuroblastoma differentiation and disrupts stem cell-like properties.

Fig. 5.

Knockout of POLQ augments expression of BMP and activates Smad1/5/9 signaling pathway in high-risk neuroblastoma LA1–55n cell. (A) Expression profile of knockout POLQ vs control in LA1–55n cells. The top 50 upregulated and downregulated genes in response to POLQ knockout are listed. (B) Western blot of POLQ knockout cell lines shows increased phospho-Smad1/5/9 relative to total Smad5 and GAPDH as loading controls. (C) Important enriched gene sets reflecting the consequences of POLQ knockout.

Functional annotation analysis using DAVID confirmed that the most significantly enriched gene sets following POLQ knockout were involved in cell differentiation, apoptosis, and immune response, while downregulated pathways were associated with unfolded protein response, oxidative phosphorylation, and fatty acid metabolism. Moreover, Gene Set Enrichment Analysis (GSEA) revealed strong enrichment of genes within the TGFβ signaling pathway, interferon-alpha response, apoptosis, and the p53 pathway (Fig 5C). Importantly, gene ontology analysis further demonstrated a negative regulation of stem cell proliferation, reinforcing the concept that POLQ loss induces a differentiation-like phenotype in neuroblastoma (Supplementary Tables S3 – S6).

Taken together, these findings establish that POLQ depletion reprograms neuroblastoma cells towards terminal differentiation via BMP-Smad1/5/9 activation, concurrently downregulating stemness-associated genes and promoting apoptosis. These results highlight POLQ as a potential target for differentiation-based therapeutic strategies in high-risk neuroblastoma.

Constitutive POLQ expression confers resistance to DNA-damaging agents in neuroblastoma

POLQ overexpression has been associated with adverse clinical outcomes in many cancers [7,12,22]. Given the observed role of POLQ in sustaining neuroblastoma proliferation and DNA damage repair, we next investigated whether enforced POLQ expression confers resistance to genotoxic therapies. To test this, we engineered SH-EP1 cells, a non-tumorigenic neuroblastoma model, to stably overexpress POLQ (POLQ-8, POLQ-12) via transfection with a pcDNA3/POLQ expression plasmid. Successful overexpression was confirmed via Western blot (Fig. 6A).

Fig. 6.

Gain of POLQ by stable transfection in SH-EP1 cells. (A) immunoblot demonstrates overexpression of POLQ in clone POLQ-8 and clone POLQ-12. (B) upper panel: SH-EP1 control cells grow as flattened epithelioid cells and do not aggregate or float. Overexpression of POLQ (POLQ-8 and POLQ-12) tended to cause cells to aggregate or float (lower panel shows colony formation of floating cells in condition medium from control or overexpressed POLQ cells). The experiments were repeated three times. (C) SH-EP1 Control, POLQ-8 and POLQ-12 cells were seeded in 96 well plate and either not treated or treated with 5 uM etoposide. At 72 h post treatment the percentage of cell death was measured using the MTS/PMS assay (****p < 0.0001). (D) DSB resolution as evidenced by decreased γH2AX foci at 24 h after 4 Gy irradiation in POLQ-8 and POLQ-12 group when compared to control cells.

A phenotypic consequence of POLQ overexpression was the dramatic alteration in SH-EP1 cell morphology. While control cells adhered well to tissue culture surfaces, POLQ-overexpressing cells exhibited a notable aggregation phenotype, suggesting potential changes in adhesion properties or intercellular signaling. To assess whether POLQ overexpression altered clonogenic potential, we collected conditioned media from both control and POLQ-overexpressing cells and performed a colony formation assay. POLQ-8 and POLQ-12 cells exhibited significantly increased colony formation compared to controls (Fig. 6B), underscoring the role of POLQ in promoting neuroblastoma cell survival and expansion.

To directly evaluate whether POLQ overexpression impacts sensitivity to DNA-damaging agents, we treated SH-EP1 control, POLQ-8, and POLQ-12 cells with etoposide and quantified cell viability using the MTS/PMS assay. POLQ-overexpressing cells exhibited significantly enhanced survival following etoposide treatment, with only 16.9 % cell death compared to 30.1 % cell death in control cells (Fig. 6C). This suggests that POLQ enhances neuroblastoma cell resistance to chemotherapy.

Next, the effects of POLQ overexpression on radiotherapy was assessed. To induce DSBs, both SH-EP1 control and POLQ overexpressed cells were irradiated at 4 Gy. After 24 h of 4 Gy radiation, POLQ overexpression led to decreased γH2AX foci, indicating more efficient DNA repair and reduced accumulation of unrepaired DSBs (Fig. 6D). These findings strongly support the role of POLQ in conferring resistance to genotoxic insults by promoting more efficient DNA damage resolution.

Collectively, these results establish that POLQ overexpression enhances neuroblastoma cell survival, clonogenic potential, and resistance to both chemotherapy and radiotherapy. Given the strong association between POLQ overexpression and poor clinical outcomes in multiple cancers [6,21,22], our findings further highlight POLQ as a major contributor to therapy resistance in neuroblastoma and a prime candidate for targeted therapeutic intervention.

Discussion

The results presented in this study highlight the critical role of POLQ-mediated end-joining (TMEJ) in promoting DNA damage tolerance in neuroblastoma, a pediatric malignancy characterized by segmental chromosomal alterations, and poor patient outcomes. We demonstrate that POLQ is overexpressed in high-risk neuroblastoma, particularly in MYCN-amplified tumors, and correlates with inferior overall and event-free survival probabilities. MYCN amplification is well-established as a driver of aggressive neuroblastoma phenotypes, promoting proliferation, metabolic reprogramming, and resistance to genotoxic stress [23]. Our results strongly support that POLQ overexpression further exacerbates these aggressive features by facilitating TMEJ, an alternative DNA repair mechanism that sustains tumor survival under genotoxic stress.

Our study expands on prior work demonstrating that high-risk neuroblastomas exhibit deficiencies in classical NHEJ repair factors (LIG4, Artemis), instead upregulating TMEJ components (LIG3, LIG1, PARP1) to compensate for defective DSB repair [6,17]. This repair plasticity confers a selective advantage to cancer cells, allowing them to bypass traditional repair constraints and propagate under genotoxic stress. Here, we chose LA1–55n cells for this study due to their known deficiencies in both P53 and LIG4 [20,24,25]. P53 is a critical regulator of NHEJ and HR DNA repair pathways [26,27]. Our findings extend this paradigm by identifying POLQ as a critical driver of this adaptive repair mechanism in neuroblastoma, supporting its role in maintaining genomic stability in HR-deficient contexts, as observed in breast and ovarian cancers [7–9,22].

Using CRISPR-Cas9-mediated knockout, we demonstrate that POLQ depletion profoundly impairs neuroblastoma cell proliferation and colony-forming ability, accompanied by a significant reduction in TMEJ repair activity, as evidenced by the EJ2-GFP assay. These findings validate that neuroblastoma cells are functionally dependent on POLQ for DNA end-joining and survival. Moreover, POLQ knockout enhances neuroblastoma sensitivity to genotoxic agents, including etoposide and radiation, aligning with reports from other malignancies where POLQ inhibition sensitizes HR-deficient cancers to DNA-damaging therapies [15,16].

The impact of POLQ loss on tumor progression was striking in our orthotopic xenograft model. POLQ knockout tumors were significantly smaller, exhibited lower engraftment rates, and lacked lung metastases, underscoring the essential role of POLQ in driving tumor progression and dissemination. This is particularly notable given the high metastatic potential of LA1–55n cells, reinforcing the notion that POLQ inhibition represents a compelling therapeutic strategy for high-risk neuroblastoma. Importantly, our study represents the first in vivo evidence supporting POLQ as a neuroblastoma target, further strengthening the rationale for pursuing POLQ inhibitors in clinical settings.

A key discovery from our RNA-seq analysis was that POLQ knockout reprograms neuroblastoma cells towards terminal differentiation via BMP2 upregulation and Smad1/5/9 activation. BMP2, a member of the TGFβ superfamily, is known to suppress neuroblastoma proliferation and promote neuronal differentiation [18,19]. Consistent with this, POLQ-deficient cells exhibited significant downregulation of stemness-associated markers (SOX2, Nanog, Oct-4) and genes implicated in early developmental plasticity (HBQ1, LRRC34, TERT). This aligns with studies in glioblastoma where inhibition of DNA repair factors, such as DNA-PK, induced differentiation and sensitized tumors to therapy [28,29]. These findings suggest that POLQ depletion not only impairs neuroblastoma proliferation but also disrupts the undifferentiated, stem-like phenotype that drives tumor aggressiveness. Given that TMEJ activity is highest in stem-like, therapy-resistant cancer cells, POLQ inhibition may represent a dual-faceted strategy to both impair repair capacity and promote differentiation-based therapy.

While our study provides compelling evidence for POLQ as a viable therapeutic target in neuroblastoma, we acknowledge several limitations. First, our in vivo studies were conducted using a single cell line, LA1–55n, which, while highly aggressive, does not fully capture the genetic heterogeneity of neuroblastoma. Future studies should evaluate POLQ inhibition across a broader panel of neuroblastoma models to confirm the generalizability of our findings. Additionally, while we demonstrate that POLQ knockout sensitizes neuroblastoma cells to etoposide and radiation, further research is needed to explore synergistic interactions between POLQ inhibitors and standard-of-care therapies, including alkylating agents and PARP inhibitors.

Given the growing interest in POLQ inhibitors as synthetic lethal agents in HR-deficient cancers, our findings provide a strong preclinical rationale for extending POLQ-targeted therapies to neuroblastoma. Recent advances in small-molecule POLQ inhibitors open the possibility of clinical translation, particularly in combination with genotoxic therapies [30]. Additionally, our findings suggest that POLQ inhibition could be leveraged to promote neuroblastoma differentiation, potentially reducing relapse rates and improving long-term patient outcomes.

Our study positions POLQ as a key regulator of DNA repair dependency, therapy resistance, and neuroblastoma progression. Targeting POLQ represents a promising therapeutic strategy, with the potential to enhance treatment efficacy, impair metastasis, and promote differentiation in high-risk neuroblastoma. Further clinical investigation of POLQ inhibitors in pediatric oncology is warranted to translate these findings into tangible patient benefits.

GRANT support

This work is supported by funds from Hyundai Hope on Wheels and the Department of Surgery, Pediatric Surgery Section, University of Michigan, Ann Arbor, Michigan.

CRediT authorship contribution statement

Sahiti Chukkapalli: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Keyonna Williams: Writing – review & editing, Writing – original draft, Formal analysis, Data curation. Biao Hu: Writing – review & editing, Writing – original draft, Data curation, Conceptualization. Kimber Converso-Baran: Data curation. Olivia Tussing: Data curation. Patrick O'Brien: Writing – review & editing, Visualization, Supervision, Formal analysis, Conceptualization. Nouri Neamati: Writing – review & editing, Writing – original draft, Supervision, Conceptualization. Erika A. Newman: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization.

Declaration of competing interest

Authors declare no competing interest.