Background
Uveal melanoma (UVM) is the most common primary intraocular malignancy in adults. As a highly aggressive form of ocular cancer originating in uveal melanocytes, it poses a significant threat to ocular vision and even causes cancer-related death. Although radiation and surgery are effective therapeutic strategies for primary tumors, up to 50% of patients subsequently develop metastasis [
1]. UVM is highly susceptible to metastasis to the liver with a median survival of 6–12 months [
2,
3]. Despite advancements in treatment modalities, the prognosis for patients with metastatic UVM remains poor because of its treatment resistance, making it one of the few cancers with stable mortality over the past three decades [
4].
Radiation or genotoxic chemotherapy exerts lethal effects by causing DNA damage. Activation of the DNA damage response (DDR) in cancer cells can ensue [
5]. Recent evidence has begun to suggest a consensus on the upregulation of DDR proteins in UVM [
6‐
8], potentially explaining the well-documented resistance of UVM to radiotherapy and chemotherapy. If DNA repair is incomplete, cancer cells cannot maintain genome integrity and undergo apoptosis and death [
9,
10], showing that defective DDR may be an attractive target for UVM treatment.
Studies have shown that the embryonic microenvironment can revert cancer cells into non-cancerous or less aggressive states, such as metastatic melanoma cells and myeloid leukemia cells [
11,
12]. These discoveries have garnered attention regarding the embryonic microenvironment in the field of cancer treatment. As embryonic stem cells (ESCs) can provide and maintain a microenvironment similar to the embryonic microenvironment, they have been shown to have suppressive effects on a variety of tumors [
13,
14]. Similarly, we have previously established an embryo-like microenvironment using mouse ESCs and indicated that such a microenvironment can effectively induce cell cycle arrest and cell apoptosis of UVM cells both in vitro and in vivo through inhibition of PI3K signaling without damage to normal somatic cells. [
15] As inhibition of the PI3K-AKT pathway can downregulate DNA damage repair factors, increasing the sensitivity of cancer cells to radiochemotherapy [
16], ESCMe has the potential to modulate the DDR of tumors.
Due to the pivotal role of DDR as a determinant of resistance to therapy in UVM [
7‐
9], inhibition of DNA repair mechanisms can render cancer cells more vulnerable to genotoxic insults, leading to increased DNA damage accumulation and compromised cell survival. This study used bioinformatic methods to detect the expression of the core DDR factors, aiming to screen the genes closely related to the development and prognosis of UVM through clinical big data. Furthermore, we sought to investigate the hypothesis that ESCMe can inhibit the DDR of UVM in vivo using xenograft mouse models. We focused on assessing the impact of ESCMe on key DNA repair pathways in UVM. Additionally, the effects of different ESCMe, including ESCs and ESC-conditioned medium (ESC-CM), on DNA repair modulation were also explored. By elucidating the interactions between ESCMe and DNA repair processes, the findings from this study hold the potential to pave the way for the development of innovative strategies that enhance the efficacy of conventional cancer treatments, such as chemotherapy and radiotherapy.
Cell cultures
Human uveal melanoma cell (C918) was obtained from KeyGen Biotechnology Company (China) and cultured in RPMI 1640 medium (Corning, USA) supplemented with 1% penicillin–streptomycin (Gibco, Japan) and 10% fetal bovine serum (FBS; Corning). Mouse ESCs were gifts from Professor Peng Xiang from Sun Yat-sen University. ESCs were cultured by knockout Dulbecco’s modified Eagle’s medium (Gibco) containing 10% FBS, 0.1 mM non-essential amino acid (Gibco), 1% GlutaMAX media (Gibco), 0.055 mM 2-mercaptoethanol (Gibco), 5 × 105 units leukemia inhibitory factor (Millipore, USA), and 1% penicillin–streptomycin.
Animal experiments
We injected 1 × 10
6 C918 cells subcutaneously into the right flanks of male Balb/c nude mice as previously described [
17]. We collected ESC-CM from cultured ESCs every day and then filtered through a 0.22-mm filter (Millex, USA). ESC-CM was stored at –20 °C. ESCs were collected after culture and resuspended in PBS prior to injection into tumors. Mice were randomized to receive treatment with ESCs, ESC-CM, or phosphate-buffered saline (PBS) when the tumor volume reached 150 mm
3. ESCs (5 × 10
5 cells/tumor in 200 μl PBS), ESC-CM (200 μl/tumor), or PBS (200 μl/tumor) was administrated at 2 different sites peritumorally every 7 days. When ESCs are differentiated, their ability to reverse tumor is significantly reduced or completely abolished. ESCs with a suicide gene, herpes simplex virus thymidine kinase (HSV-TK), controlled by ganciclovir (GCV), were constructed previously. GCV (Sigma, 2 mg/mouse in 200 μl PBS) was injected intraperitoneally on day 5 of every treatment cycle to eliminate the differentiated ESCs and avoid the formation of teratomas. We have previously demonstrated the elimination of ESCs by GCV. After 3 treatment cycles, the mice were euthanized in an enclosed chamber filled with carbon dioxide, followed by cervical dislocation, and their tumor tissues and surrounding skin tissues were examined.
RT-qPCR
We used RNeasy Fibrous Tissue Mini kit (Qiagen) to isolate the RNA from the UVM tissues and surrounding skin tissues following the manufacturer’s instructions, then quantified the total RNA by absorption at 260 nm as previously described [
17]. Next, a PrimeScript™ RT Master Mix (Takara, Japan) was used to generate cDNA, which was used for qPCR with SYBR® Premix Ex Taq™ (Takara) in a StepOnePlus thermal cycler (ABI, USA). The GAPDH gene was served as the internal reference.
Immunohistochemical (IHC) staining
Immunohistochemistry was performed for tumor and skin tissues according to the standard procedure using the following primary antibodies: XRCC6 (ab92450; Abcam), LIG4 (ab193353; Abcam), PARP1 (ab191217; Abcam), DNA-PK (#38168; Cell Signaling) and XRCC5 (WH0007520M2; Sigma-Aldrich). Slides were imaged on a Pannoramic Digital Slide Scanner (3DHISTECH, Hungary) and analyzed by Image-Pro Plus 6.0 (Media Cybemetics, USA). Paraffin sections were deparaffinized and hydrated in PBS. Following blocking of endogenous peroxidase, sections were pretreated with appropriate buffer, if necessary. Thereafter, sections were incubated with primary antibody at 4˚C overnight. Following washing three times with PBS, sections were subsequently incubated with MAX-PO. Peroxidase activity was visualized with diaminobenzidine (DAB).
Terminal dUTP nick end‐labelling (TUNEL) assay
TUNEL assay was used to detect apoptosis in tumor tissue. It was performed following the manufacturer’s (KeyGen’s) instructions and analyzed by Image-Pro Plus 6.0.
Statistical analysis
The survival times were compared using Kaplan–Meier analysis, and the p value was calculated using the log-rank test. Statistical analysis was performed using GraphPad Prism software. A 2-tailed unpaired Student t-test was used for analyses comparing only 2 groups, and analysis of variance and an appropriate post hoc test were used for analyses comparing more than 2 groups. Statistical significance was set at p < 0.05.
Discussion
UVM is notoriously resistant to both radiation and DNA-damaging agents, which is closely related to its high DDR activities, underlining the importance of targeting the DNA repair capacity as a potentially effective therapeutic option, either alone or as a sensitizer for other treatments [
6‐
8]. The present study showed that a low level of NHEJ pathway-related genes predicted better prognosis in UVM patients and ESCMe could significantly impair the DDR in UVM tissues by downregulating the NHEJ pathway. Previously, we found that the ESCMe suppressed the proliferation, invasiveness, and tumorigenicity of UVM [
17]. Combining these results, we can conclude that the ESCMe could inhibit various malignant phenotypes of UVM, indicating a promising and attractive prospect of ESC-based therapy to enhance the sensitivity of UVM to radiochemotherapy and improve treatment outcomes.
DSBs, which can be induced by radiation and other types of genotoxic noxae or during replication are mainly repaired by homologous recombination (HR) and NHEJ [
18‐
20]. Evidence suggests that NHEJ activity was upregulated and was the main mechanism for repairing DNA damage in UVM [
7]. The inhibition of the NHEJ protein DNA-PK is even lethal to UVM. In the previous study, we demonstrated that UVM cells treated with ESCs had higher expression levels of p21 [
17]. p21 induction is essential for the onset of cell cycle arrest in the DDR, arresting cells at the G1/S transition and giving cells time to repair critical damage [
21]. If DNA repair is incomplete, such as when NHEJ-related factors are decreased or DNA damage is extensive, the repair of DSBs can be compromised and the cell undergoes apoptosis [
9,
10]. The data presented here, demonstrating the inhibition of functional NHEJ in UVM tissue of ESC-treated mice, was confirmed by gene expression analysis of PCR and immunohistochemical detection. It would therefore be reasonable to consider that the increase in tumor apoptosis by ESCs may be implicated in the upregulation of p21 and downregulation of NHEJ.
Apart from its important role in NHEJ, PARP-1 also engages in other DNA repair mechanisms, such as base excision repair (BER), nucleotide excision repair (NER), DNA mismatch repair (MMR), and maintenance of replication fork stability [
22‐
24], regulating cell death in the event of excessive damage. Several studies have indeed reported that an elevated level of PARP-1 expression was observed in UVM and was associated with a shorter overall survival time and disease-free survival time [
6,
8,
25,
26]. It is possible that PARP-1 inhibition can suppress damaged DNA repair and improve tumor killing [
27]. Therefore, induced downregulation of PARP-1 by ESCs could lead to multiple compromised DNA repair pathways and a reduced ability to repair DNA damage, suggesting ESCs as an attractive therapeutic option alone or in combination with other treatments of UVM.
Although DDR inhibitors have shown promising results in cancer treatment, they can disrupt the normal DNA repair mechanisms in healthy cells, leading to the accumulation of DNA damage, causing cell cycle arrest or apoptosis, potentially interfering with the ability of normal tissues to regenerate and repair, especially in rapidly dividing tissues such as the bone marrow, skin, and gastrointestinal tract [
28‐
30]. These limit the clinical application of DDR inhibitors [
31]. Surprisingly, this study shows that while the expression of NHEJ pathway-related genes in tumors decreased, the levels of XRCC6 and PARP1 were enhanced in the surrounding skin tissue, indicating an upregulation of DDR activities. This enhanced DDR response may efficiently remove DNA damage and maintain genome stability in normal cells, avoiding apoptotic pathways triggered by DNA damage, facilitating faster cell cycle turnover and promoting cell proliferation [
32]. This was consistent with our previous research showing that ESCs could markedly suppress tumor growth and enhance the proliferation of adjacent skin tissue [
17]. By promoting efficient DNA repair, normal cells with elevated DDR levels are better equipped to cope with various genotoxic stresses, such as radiation or chemical-induced DNA damage. This increased resistance to DNA damage can contribute to the survival and maintenance of normal cells, potentially mitigating concerns about severe side effects of cancer therapy.
It is worth noting that the suppressive effects of the ESCMe on the DDR activities of UVM were much more significant in the ESC treatment group than in the ESC-CM treatment group, which was in agreement with our previous finding that ESC treatment showed a superior therapeutic effect to ESC-CM treatment in terms of proliferation, invasiveness, and tumorigenicity [
17]. Direct cell-to-cell contact with ESCs appears to be more effective in inhibiting the NHEJ pathway and triggering apoptosis in tumor cells than exposure to their secreted factors. Furthermore, the secreted factors of ESCs showed no obvious effect on the DDR of normal cells. This may be due to the direct signal communication via cell–cell contact, which is the main mechanism by which ESC influence the DDR activity of both cancer cells and normal cells.
Nevertheless, this study has several limitations. First, this study only illustrated the effect of ESCMe on DDR in subcutaneous UVM models, the conditions of which could be different from that in orthotopic models. Further research is needed to explore the influence of ESCMe on orthotopic UVM models. Second, the experiments in this study were only performed on the C918 cell line. Further studies should be conducted in more UVM cell lines. Third, because of the lack of clinicopathological data on UVM in relevant databases such as TCGA, we were unable to find more data for effective analysis. Clinical data of more UVM patients need to be further collected.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.